Update app.py

app.py

@@ -3,7 +3,6 @@ import gradio as gr
 from diffusers import AnimateDiffSparseControlNetPipeline, AutoencoderKL, MotionAdapter, SparseControlNetModel, AnimateDiffPipeline, EulerAncestralDiscreteScheduler
 from diffusers.schedulers import DPMSolverMultistepScheduler
 from diffusers.utils import export_to_gif, load_image
-from peft import LoraConfig, get_peft_model
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
@@ -21,12 +20,6 @@ def generate_video(prompt, negative_prompt, num_inference_steps, conditioning_fr
     ).to(device)
     
     pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config, beta_schedule="linear", algorithm_type="dpmsolver++", use_karras_sigmas=True)
-
-    # PEFT 사용하여 Lora 로드
-    lora_config = LoraConfig(r=16, lora_alpha=32, target_modules=["q_proj", "v_proj"])
-    pipe = get_peft_model(pipe, lora_config)
-    pipe.load_lora_weights("guoyww/animatediff-motion-lora-v1-5-3", adapter_name="motion_lora")
-    pipe.fuse_lora(lora_scale=1.0)
     
     image_files = [
         "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/animatediff-scribble-1.png",
@@ -41,13 +34,12 @@ 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=[int(x) for x in conditioning_frame_indices.split(",")],
+        controlnet_frame_indices=conditioning_frame_indices,
         generator=torch.Generator().manual_seed(1337),
-    ).frames
+    ).frames[0]
     
-    output_file = "output.gif"
-    export_to_gif(video, output_file)
-    return output_file
+    export_to_gif(video, "output.gif")
+    return "output.gif"
 
 def generate_simple_video(prompt):
     adapter = MotionAdapter.from_pretrained("guoyww/animatediff-motion-adapter-v1-5-2", torch_dtype=torch.float16).to(device)
@@ -68,11 +60,10 @@ def generate_simple_video(prompt):
         num_inference_steps=20,
         guidance_scale=7.0,
         decode_chunk_size=2,
-    ).frames
+    ).frames[0]
     
-    output_file = "simple_output.gif"
-    export_to_gif(frames, output_file)
-    return output_file
+    export_to_gif(frames, "simple_output.gif")
+    return "simple_output.gif"
 
 demo1 = gr.Interface(
     fn=generate_video,
@@ -80,10 +71,10 @@ demo1 = gr.Interface(
         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.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.Video(label="Generated Video"),
+    outputs=gr.Image(label="Generated Video"),
     title="Generate Video with AnimateDiffSparseControlNetPipeline",
     description="Generate a video using the AnimateDiffSparseControlNetPipeline."
 )
@@ -91,12 +82,13 @@ demo1 = gr.Interface(
 demo2 = gr.Interface(
     fn=generate_simple_video,
     inputs=gr.Textbox(label="Prompt", value="An astronaut riding a horse on Mars."),
-    outputs=gr.Video(label="Generated Simple Video"),
+    outputs=gr.Image(label="Generated Simple Video"),
     title="Generate Simple Video with AnimateDiff",
     description="Generate a simple video using the AnimateDiffPipeline."
 )
 
 demo = gr.TabbedInterface([demo1, demo2], ["Advanced Video Generation", "Simple Video Generation"])
 
+
 demo.launch()
 #demo.launch(server_name="0.0.0.0", server_port=7910)