Upload folder using huggingface_hub

app.py

@@ -3,9 +3,10 @@ import spaces
 
 import gradio as gr
 from diffusers import ModularPipelineBlocks
-from diffusers.utils import export_to_video, load_image
+from diffusers.utils import export_to_video
 from diffusers.modular_pipelines import WanModularPipeline
 
+
 class MatrixGameWanModularPipeline(WanModularPipeline):
     """
     A ModularPipeline for MatrixGameWan.
@@ -34,10 +35,13 @@ blocks.sub_blocks.insert("image_to_action", image_to_action_block, 0)
 pipe = MatrixGameWanModularPipeline(blocks, "diffusers/matrix-game-2-modular")
 pipe.load_components(trust_remote_code=True, device_map="cuda", torch_dtype={"default": torch.bfloat16, "vae": torch.float32})
 
+
 @spaces.GPU(120)
-def predict(prompt):
-    output = pipe(image=None, prompt=prompt, num_frames=141)
-    return export_to_video(output.values['videos'][0], "output.mp4")
+def predict(image, prompt):
+    return image
+    #output = pipe(image=image, prompt=prompt, num_frames=141)
+    #return export_to_video(output.values['videos'][0], "output.mp4")
+
 
 examples = []
 
@@ -57,12 +61,14 @@ css = """
 
 with gr.Blocks(css=css) as demo:
     with gr.Column(elem_id="col-container"):
+        gr.Markdown(
+            "Image to Video [Modular Diffusers](https://huggingface.co/docs/diffusers/v0.35.1/en/modular_diffusers/overview) "
+            "Pipeline that uses Qwen 2.5 VL 72B Instruct and [Matrix Game 2.0](https://huggingface.co/Skywork/Matrix-Game-2.0) "
+            "to allow you to move through images. Works best for scenes with static elements"
+        )
         with gr.Row():
             with gr.Column():
-                input_images = gr.Image(label="Input Image",
-                                          show_label=False,
-                                          type="pil",
-                                          interactive=True)
+                image = gr.Image(label="Input Image", show_label=False, type="pil", interactive=True)
 
             with gr.Column():
                 result = gr.Video(label="Result")
@@ -71,15 +77,17 @@ with gr.Blocks(css=css) as demo:
             prompt = gr.Text(
                     label="Prompt",
                     show_label=False,
-                    placeholder="describe how you want to move in the image",
+                    placeholder="describe how you would like to move in the image",
                     container=False,
             )
+        with gr.Row():
             run_button = gr.Button("Run!", variant="primary")
 
     gr.on(
-        triggers=[run_button.click, prompt.submit],
+        triggers=[run_button.click],
         fn=predict,
         inputs=[
+            image,
             prompt,
         ],
         outputs=[result],