Upload folder using huggingface_hub

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_00.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db758debd583ca512518e48739af3cb4e9df6fe4c1629f048a7b240ff4eae268
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_01.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fee9f8bcf3c59af802e83fc3b460bc33c5251b61811267c08be3d91c4505e6e
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_02.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a007faaa7dae3f18812e37849f6ee1c57bc0118e30e38fa9041c5f8dac81853
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_03.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6991f750e6023c5e4c9d71f6453c6fb898f3077154991b80cfd78cf96b990f7e
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_04.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19b2f1efb751d9353dbf3a32c218e7fc0b2e80eea81729104817458f124a7f03
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_05.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65849d302f16fa5f480b118d4cbc2db0fcf24463975fea51e02f1ddfc30bff22
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_06.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b68fb5e4e1e257850b20feaf49bd8b83180c7f8564f9ecbc3dda51d77595354
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/audio/audio_07.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9175d6984984a53f3d74d172b445a3b6e5984fc84a671d978fd7d0a187a24e6
+size 354708

ltx2_4k_experiments/one_refiner_4096x2304/inputs/inputs_manifest.json

@@ -0,0 +1,50 @@
+[
+  {
+    "index": 0,
+    "prompt": "A close-up tabletop shot of a glass marble rolling across a brushed metal surface, tiny reflections shifting as it passes under soft studio lights.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_00.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_00.wav"
+  },
+  {
+    "index": 1,
+    "prompt": "A small robot walks through a rainy neon alley, water splashing around its feet while signs flicker in the background.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_01.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_01.wav"
+  },
+  {
+    "index": 2,
+    "prompt": "A ceramic mug on a wooden desk fills with swirling steam, morning light moves across notebooks and a silver pen.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_02.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_02.wav"
+  },
+  {
+    "index": 3,
+    "prompt": "A red kite rises above a grassy hill in a steady breeze, the string pulls taut while clouds drift slowly behind it.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_03.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_03.wav"
+  },
+  {
+    "index": 4,
+    "prompt": "A chef tosses vegetables in a pan over a blue flame, sparks and steam rise while the camera stays at counter height.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_04.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_04.wav"
+  },
+  {
+    "index": 5,
+    "prompt": "A vintage train moves through a snowy pine forest, dark smoke trailing upward as warm window lights pass between trees.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_05.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_05.wav"
+  },
+  {
+    "index": 6,
+    "prompt": "A diver swims beside a coral reef, fish scatter around bright sea plants while sunbeams ripple through clear water.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_06.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_06.wav"
+  },
+  {
+    "index": 7,
+    "prompt": "A dancer practices in an empty rehearsal room, turning slowly as dust floats in a beam of late afternoon light.",
+    "image": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/images/input_07.png",
+    "audio": "outputs/ltx2_repro_4k_one_refiner_4096x2304/inputs/audio/audio_07.wav"
+  }
+]

ltx2_4k_experiments/one_refiner_4096x2304/logs/ti2vid_two_stages_00.log

@@ -0,0 +1,5 @@
+<frozen runpy>:128: RuntimeWarning: 'ltx_pipelines.ti2vid_two_stages' found in sys.modules after import of package 'ltx_pipelines', but prior to execution of 'ltx_pipelines.ti2vid_two_stages'; this may result in unpredictable behaviour
+Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
+
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:46<22:42, 46.97s/it]
  7%|▋         | 2/30 [01:34<21:57, 47.05s/it]
 10%|█         | 3/30 [02:21<21:12, 47.11s/it]
 13%|█▎        | 4/30 [03:08<20:26, 47.17s/it]
 17%|█▋        | 5/30 [03:55<19:40, 47.21s/it]
 20%|██        | 6/30 [04:43<18:54, 47.25s/it]
 23%|██▎       | 7/30 [05:30<18:07, 47.27s/it]
 27%|██▋       | 8/30 [06:17<17:20, 47.29s/it]
 30%|███       | 9/30 [07:05<16:33, 47.31s/it]
 33%|███▎      | 10/30 [07:52<15:46, 47.32s/it]
 37%|███▋      | 11/30 [08:39<14:59, 47.32s/it]
 40%|████      | 12/30 [09:27<14:12, 47.34s/it]
 43%|████▎     | 13/30 [10:14<13:24, 47.34s/it]
 47%|████▋     | 14/30 [11:01<12:37, 47.34s/it]
 50%|█████     | 15/30 [11:49<11:50, 47.34s/it]
 53%|█████▎    | 16/30 [12:36<11:02, 47.34s/it]
 57%|█████▋    | 17/30 [13:23<10:15, 47.34s/it]
 60%|██████    | 18/30 [14:11<09:28, 47.34s/it]
 63%|██████▎   | 19/30 [14:58<08:40, 47.34s/it]
 67%|██████▋   | 20/30 [15:45<07:53, 47.34s/it]
 70%|███████   | 21/30 [16:33<07:06, 47.34s/it]
 73%|███████▎  | 22/30 [17:20<06:18, 47.34s/it]
 77%|███████▋  | 23/30 [18:07<05:31, 47.34s/it]
 80%|████████  | 24/30 [18:55<04:43, 47.33s/it]
 83%|████████▎ | 25/30 [19:42<03:56, 47.33s/it]
 87%|████████▋ | 26/30 [20:29<03:09, 47.33s/it]
 90%|█████████ | 27/30 [21:17<02:21, 47.33s/it]
 93%|█████████▎| 28/30 [22:04<01:34, 47.32s/it]
 97%|█████████▋| 29/30 [22:51<00:47, 47.31s/it]
+
  0%|          | 0/3 [00:00<?, ?it/s]
 33%|███▎      | 1/3 [01:52<03:44, 112.04s/it]
 67%|██████▋   | 2/3 [03:44<01:52, 112.17s/it]
+
  0%|          | 0/3 [00:00<?, ?it/s]
 67%|██████▋   | 2/3 [00:00<00:00,  4.06it/s]
 67%|██████▋   | 2/3 [00:00<00:00,  2.63it/s]

ltx2_4k_experiments/one_refiner_4096x2304/metrics/dit_metrics.jsonl

@@ -0,0 +1,2 @@
+{"pipeline": "ti2vid_two_stages", "run_id": "ti2vid_two_stages_00", "stage": "stage_1_generate", "elapsed_seconds": 1419.067502622027, "dit_fps": 0.08526726161822953, "denoising_steps": 30, "width": 2048, "height": 1152, "frames": 121, "frame_rate": 24.0, "video_present": true, "audio_present": true, "loop": "euler_denoising_loop"}
+{"pipeline": "ti2vid_two_stages", "run_id": "ti2vid_two_stages_00", "stage": "stage_2_refine", "elapsed_seconds": 336.48722817597445, "dit_fps": 0.3595976009428804, "denoising_steps": 3, "width": 4096, "height": 2304, "frames": 121, "frame_rate": 24.0, "video_present": true, "audio_present": true, "loop": "euler_denoising_loop"}

ltx2_4k_experiments/one_refiner_4096x2304/runs_manifest.jsonl

@@ -0,0 +1 @@
+{"pipeline": "ti2vid_two_stages", "run_id": "ti2vid_two_stages_00", "index": 0, "status": "success", "returncode": 0, "started_unix": 1779370039.4169116, "wall_seconds": 1955.27288747899, "output_path": "outputs/ltx2_repro_4k_one_refiner_4096x2304/ti2vid_two_stages/ti2vid_two_stages_00.mp4", "log_path": "outputs/ltx2_repro_4k_one_refiner_4096x2304/logs/ti2vid_two_stages_00.log", "command": ["/lustre/fsw/portfolios/nvr/users/shyang/miniconda3/envs/ltx2_repro/bin/python", "-m", "ltx_pipelines.ti2vid_two_stages", "--checkpoint-path", "/lustre/fs11/portfolios/nvr/projects/nvr_elm_llm/users/shyang/ltx2/models/ltx2/LTX-2.3/ltx-2.3-22b-dev.safetensors", "--gemma-root", "/lustre/fs11/portfolios/nvr/projects/nvr_elm_llm/users/shyang/ltx2/models/ltx2/gemma-3-12b-it-qat-q4_0-unquantized", "--prompt", "A close-up tabletop shot of a glass marble rolling across a brushed metal surface, tiny reflections shifting as it passes under soft studio lights.", "--output-path", "/lustre/fs11/portfolios/nvr/projects/nvr_elm_llm/users/shyang/ltx2/outputs/ltx2_repro_4k_one_refiner_4096x2304/ti2vid_two_stages/ti2vid_two_stages_00.mp4", "--height", "2304", "--width", "4096", "--num-frames", "121", "--frame-rate", "24", "--num-inference-steps", "30", "--distilled-lora", "/lustre/fs11/portfolios/nvr/projects/nvr_elm_llm/users/shyang/ltx2/models/ltx2/LTX-2.3/ltx-2.3-22b-distilled-lora-384-1.1.safetensors", "0.8", "--spatial-upsampler-path", "/lustre/fs11/portfolios/nvr/projects/nvr_elm_llm/users/shyang/ltx2/models/ltx2/LTX-2.3/ltx-2.3-spatial-upscaler-x2-1.1.safetensors", "--offload", "cpu", "--max-batch-size", "1"], "control": {"prompt": "A close-up tabletop shot of a glass marble rolling across a brushed metal surface, tiny reflections shifting as it passes under soft studio lights."}, "settings": {"height": 2304, "width": 4096, "num_frames": 121, "frame_rate": 24, "num_inference_steps": 30, "quantization": "none", "offload": "cpu"}}

ltx2_4k_experiments/one_refiner_4096x2304/ti2vid_two_stages/ti2vid_two_stages_00.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c22cf31d1096ff43d2fac0996b079ad1c082abea7581f89f3c2cd3532a3c96a4
+size 6548682