1

app.py

@@ -1,7 +1,6 @@
 import gradio as gr
 from gradio_client import Client, handle_file
 
-# Gradio Interface Functions
 def start_session():
     client = Client("JeffreyXiang/TRELLIS")
     result = client.predict(api_name="/start_session")
@@ -82,70 +81,88 @@ def lambda_6_function():
     result = client.predict(api_name="/lambda_6")
     return result
 
-# Gradio Interface
-iface = gr.Interface(
-    fn={
-        "start_session": start_session,
-        "lambda_function": lambda_function,
-        "lambda_1_function": lambda_1_function,
-        "preprocess_image": preprocess_image,
-        "preprocess_images": preprocess_images,
-        "get_seed": get_seed,
-        "image_to_3d": image_to_3d,
-        "lambda_2_function": lambda_2_function,
-        "lambda_3_function": lambda_3_function,
-        "extract_glb": extract_glb,
-        "lambda_4_function": lambda_4_function,
-        "extract_gaussian": extract_gaussian,
-        "lambda_5_function": lambda_5_function,
-        "lambda_6_function": lambda_6_function,
-    },
-    inputs={
-        "start_session": [],
-        "lambda_function": [],
-        "lambda_1_function": [],
-        "preprocess_image": gr.Image(type="filepath"),
-        "preprocess_images": gr.Gallery(label="Images"),
-        "get_seed": [gr.Checkbox(value=True, label="Randomize Seed"), gr.Slider(minimum=0, maximum=100, value=0, label="Seed")],
-        "image_to_3d": [
-            gr.Image(type="filepath", label="Image"),
-            gr.Gallery(label="Multi Images"),
-            gr.Slider(minimum=0, maximum=100, value=0, label="Seed"),
-            gr.Slider(minimum=0, maximum=10, value=7.5, label="SS Guidance Strength"),
-            gr.Slider(minimum=0, maximum=20, value=12, label="SS Sampling Steps"),
-            gr.Slider(minimum=0, maximum=10, value=3, label="SLAT Guidance Strength"),
-            gr.Slider(minimum=0, maximum=20, value=12, label="SLAT Sampling Steps"),
-            gr.Radio(choices=["stochastic", "multidiffusion"], value="stochastic", label="Multi-image Algorithm")
-        ],
-        "lambda_2_function": [],
-        "lambda_3_function": [],
-        "extract_glb": [
-            gr.Slider(minimum=0, maximum=1, value=0.95, label="Mesh Simplify"),
-            gr.Slider(minimum=0, maximum=2048, value=1024, label="Texture Size")
-        ],
-        "lambda_4_function": [],
-        "extract_gaussian": [],
-        "lambda_5_function": [],
-        "lambda_6_function": []
-    },
-    outputs={
-        "start_session": gr.Textbox(),
-        "lambda_function": gr.Textbox(),
-        "lambda_1_function": gr.Textbox(),
-        "preprocess_image": gr.Image(type="filepath"),
-        "preprocess_images": gr.Gallery(),
-        "get_seed": gr.Textbox(),
-        "image_to_3d": gr.Video(),
-        "lambda_2_function": gr.Textbox(),
-        "lambda_3_function": gr.Textbox(),
-        "extract_glb": gr.Files(),
-        "lambda_4_function": gr.Textbox(),
-        "extract_gaussian": gr.Files(),
-        "lambda_5_function": gr.Files(),
-        "lambda_6_function": gr.Files()
-    },
-    title="TRELLIS API Client",
-    description="This Gradio app allows you to interact with the TRELLIS API endpoints."
-)
-
-iface.launch()
+with gr.Blocks(title="TRELLIS API Client") as demo:
+    gr.Markdown("# TRELLIS API Client")
+    
+    with gr.Tab("Start Session"):
+        start_btn = gr.Button("Start Session")
+        start_output = gr.Textbox(label="Output")
+        start_btn.click(fn=start_session, inputs=[], outputs=start_output)
+    
+    with gr.Tab("Lambda"):
+        lambda_btn = gr.Button("Run Lambda")
+        lambda_output = gr.Textbox(label="Output")
+        lambda_btn.click(fn=lambda_function, inputs=[], outputs=lambda_output)
+    
+    with gr.Tab("Lambda 1"):
+        lambda1_btn = gr.Button("Run Lambda 1")
+        lambda1_output = gr.Textbox(label="Output")
+        lambda1_btn.click(fn=lambda_1_function, inputs=[], outputs=lambda1_output)
+    
+    with gr.Tab("Preprocess Image"):
+        image_input = gr.Image(type="filepath", label="Input Image")
+        preprocess_btn = gr.Button("Preprocess")
+        image_output = gr.Image(type="filepath", label="Processed Image")
+        preprocess_btn.click(fn=preprocess_image, inputs=image_input, outputs=image_output)
+    
+    with gr.Tab("Preprocess Images"):
+        images_input = gr.Gallery(label="Input Images")
+        preprocess_images_btn = gr.Button("Preprocess")
+        images_output = gr.Gallery(label="Processed Images")
+        preprocess_images_btn.click(fn=preprocess_images, inputs=images_input, outputs=images_output)
+    
+    with gr.Tab("Get Seed"):
+        with gr.Row():
+            randomize = gr.Checkbox(value=True, label="Randomize Seed")
+            seed = gr.Slider(minimum=0, maximum=100, value=0, label="Seed")
+        seed_btn = gr.Button("Get Seed")
+        seed_output = gr.Textbox(label="Output")
+        seed_btn.click(fn=get_seed, inputs=[randomize, seed], outputs=seed_output)
+    
+    with gr.Tab("Image to 3D"):
+        img_input = gr.Image(type="filepath", label="Image")
+        multi_imgs = gr.Gallery(label="Multi Images")
+        with gr.Row():
+            img_seed = gr.Slider(minimum=0, maximum=100, value=0, label="Seed")
+            ss_guidance = gr.Slider(minimum=0, maximum=10, value=7.5, label="SS Guidance Strength")
+        with gr.Row():
+            ss_steps = gr.Slider(minimum=0, maximum=20, value=12, label="SS Sampling Steps")
+            slat_guidance = gr.Slider(minimum=0, maximum=10, value=3, label="SLAT Guidance Strength")
+        with gr.Row():
+            slat_steps = gr.Slider(minimum=0, maximum=20, value=12, label="SLAT Sampling Steps")
+            algo = gr.Radio(choices=["stochastic", "multidiffusion"], value="stochastic", label="Multi-image Algorithm")
+        generate_btn = gr.Button("Generate 3D")
+        video_output = gr.Video(label="Generated 3D")
+        generate_btn.click(
+            fn=image_to_3d,
+            inputs=[img_input, multi_imgs, img_seed, ss_guidance, ss_steps, slat_guidance, slat_steps, algo],
+            outputs=video_output
+        )
+    
+    with gr.Tab("Extract GLB"):
+        with gr.Row():
+            mesh_simplify = gr.Slider(minimum=0, maximum=1, value=0.95, label="Mesh Simplify")
+            texture_size = gr.Slider(minimum=0, maximum=2048, value=1024, label="Texture Size")
+        extract_glb_btn = gr.Button("Extract GLB")
+        glb_output = gr.File(label="GLB File")
+        extract_glb_btn.click(fn=extract_glb, inputs=[mesh_simplify, texture_size], outputs=glb_output)
+    
+    with gr.Tab("Extract Gaussian"):
+        extract_gaussian_btn = gr.Button("Extract Gaussian")
+        gaussian_output = gr.File(label="Gaussian File")
+        extract_gaussian_btn.click(fn=extract_gaussian, inputs=[], outputs=gaussian_output)
+    
+    # Add remaining lambda functions as simple buttons with text output
+    for name, func in [
+        ("Lambda 2", lambda_2_function),
+        ("Lambda 3", lambda_3_function),
+        ("Lambda 4", lambda_4_function),
+        ("Lambda 5", lambda_5_function),
+        ("Lambda 6", lambda_6_function),
+    ]:
+        with gr.Tab(name):
+            btn = gr.Button(f"Run {name}")
+            output = gr.Textbox(label="Output")
+            btn.click(fn=func, inputs=[], outputs=output)
+
+demo.launch()