Update app.py

app.py

@@ -10,14 +10,11 @@ def generate_video(prompt, negative_prompt, num_inference_steps, conditioning_fr
     motion_adapter = MotionAdapter.from_pretrained("guoyww/animatediff-motion-adapter-v1-5-3", torch_dtype=torch.float16).to(device)
     controlnet = SparseControlNetModel.from_pretrained("guoyww/animatediff-sparsectrl-scribble", torch_dtype=torch.float16).to(device)
     vae = AutoencoderKL.from_pretrained("stabilityai/sd-vae-ft-mse", torch_dtype=torch.float16).to(device)
-    scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config, beta_schedule="linear", algorithm_type="dpmsolver++", use_karras_sigmas=True)
-    
     pipe = AnimateDiffSparseControlNetPipeline.from_pretrained(
         "SG161222/Realistic_Vision_V5.1_noVAE",
         motion_adapter=motion_adapter,
         controlnet=controlnet,
         vae=vae,
-        scheduler=scheduler,
         torch_dtype=torch.float16,
     ).to(device)
     
@@ -38,12 +35,13 @@ def generate_video(prompt, negative_prompt, num_inference_steps, conditioning_fr
         num_inference_steps=num_inference_steps,
         conditioning_frames=conditioning_frames,
         controlnet_conditioning_scale=controlnet_conditioning_scale,
-        controlnet_frame_indices=conditioning_frame_indices,
+        controlnet_frame_indices=[int(x) for x in conditioning_frame_indices.split(",")],
         generator=torch.Generator().manual_seed(1337),
-    ).frames[0]
+    ).frames
     
-    export_to_gif(video, "output.gif")
-    return "output.gif"
+    output_file = "output.gif"
+    export_to_gif(video, output_file)
+    return output_file
 
 def generate_simple_video(prompt):
     adapter = MotionAdapter.from_pretrained("guoyww/animatediff-motion-adapter-v1-5-2", torch_dtype=torch.float16).to(device)
@@ -64,29 +62,30 @@ def generate_simple_video(prompt):
         num_inference_steps=20,
         guidance_scale=7.0,
         decode_chunk_size=2,
-    ).frames[0]
+    ).frames
     
-    export_to_gif(frames, "simple_output.gif")
-    return "simple_output.gif"
+    output_file = "simple_output.gif"
+    export_to_gif(frames, output_file)
+    return output_file
 
 demo1 = gr.Interface(
     fn=generate_video,
     inputs=[
-        gr.Textbox(label="Prompt", default="an aerial view of a cyberpunk city, night time, neon lights, masterpiece, high quality"),
-        gr.Textbox(label="Negative Prompt", default="low quality, worst quality, letterboxed"),
-        gr.Slider(label="Number of Inference Steps", minimum=1, maximum=100, step=1, default=25),
-        gr.Textbox(label="Conditioning Frame Indices", default="[0, 8, 15]"),
-        gr.Slider(label="ControlNet Conditioning Scale", minimum=0.1, maximum=2.0, step=0.1, default=1.0)
+        gr.Textbox(label="Prompt", value="an aerial view of a cyberpunk city, night time, neon lights, masterpiece, high quality"),
+        gr.Textbox(label="Negative Prompt", value="low quality, worst quality, letterboxed"),
+        gr.Slider(label="Number of Inference Steps", minimum=1, maximum=100, step=1, value=25),
+        gr.Textbox(label="Conditioning Frame Indices", value="0, 8, 15"),
+        gr.Slider(label="ControlNet Conditioning Scale", minimum=0.1, maximum=2.0, step=0.1, value=1.0)
     ],
-    outputs=gr.Image(label="Generated Video"),
+    outputs=gr.Video(label="Generated Video"),
     title="Generate Video with AnimateDiffSparseControlNetPipeline",
     description="Generate a video using the AnimateDiffSparseControlNetPipeline."
 )
 
 demo2 = gr.Interface(
     fn=generate_simple_video,
-    inputs=gr.Textbox(label="Prompt", default="An astronaut riding a horse on Mars."),
-    outputs=gr.Image(label="Generated Simple Video"),
+    inputs=gr.Textbox(label="Prompt", value="An astronaut riding a horse on Mars."),
+    outputs=gr.Video(label="Generated Simple Video"),
     title="Generate Simple Video with AnimateDiff",
     description="Generate a simple video using the AnimateDiffPipeline."
 )