Update app.py

app.py

@@ -54,8 +54,6 @@ JS = """function () {
 }"""
 
 
-
-
 # Ensure model and scheduler are initialized in GPU-enabled function
 if torch.cuda.is_available():
     model_manager = ModelManager(
@@ -99,7 +97,12 @@ def change_media(image_in, video_in, selected):
     elif selected == "Diffutoon":
         return gr.update(visible=False), gr.update(visible=True), gr.update(visible=True)
 
-
+def update_frames(video_in):
+    up_video = VideoData(
+            video_file=video_in,
+            height=shape_height, width=shape_width)
+    frame_len = len(up_video)
+    return gr.update(maximum=frame_len)
 
 @spaces.GPU(duration=120)
 def generate(
@@ -108,15 +111,19 @@ def generate(
     selected,
     prompt: str = "best quality",
     seed: Optional[int] = -1,
-    num_inference_steps: int = 5,
+    num_inference_steps: int = 10,
+    num_frames: int = 30,
+    shape_height: int = 512,
+    shape_width: int = 512,
+    height: int = 512,
+    width: int = 512,
     animatediff_batch_size: int = 32,
     animatediff_stride: int = 16,
     motion_bucket_id: int = 127,
     fps_id: int = 25,
-    num_frames: int = 50,
     output_folder: str = "outputs",
     progress=gr.Progress(track_tqdm=True)):
-    
+    video = ""
     if seed == -1:
         seed = random.randint(0, MAX_SEED)
 
@@ -125,15 +132,15 @@ def generate(
     os.makedirs(output_folder, exist_ok=True)
     base_count = len(glob(os.path.join(output_folder, "*.mp4")))
     video_path = os.path.join(output_folder, f"{base_count:06d}.mp4")
-
+    
     if selected == "ExVideo" and image_in:
         image = Image.open(image_in)
         video = pipe(
-            input_image=image.resize((512, 512)), 
+            input_image=image.resize((shape_height, shape_width)), 
             num_frames=num_frames, 
             fps=fps_id, 
-            height=512, 
-            width=512,
+            height=height, 
+            width=width,
             motion_bucket_id=motion_bucket_id,
             num_inference_steps=num_inference_steps,
             min_cfg_scale=2, 
@@ -144,18 +151,19 @@ def generate(
     elif selected == "Diffutoon" and video_in:
         up_video = VideoData(
             video_file=video_in,
-            height=512, width=512)
-        input_video = [up_video[i] for i in range(1, 30)]
+            height=shape_height, width=shape_width)
+        frame_len = min(num_frames, len(up_video))
+        input_video = [up_video[i] for i in range(1, frame_len)]
 
         video = pipe2(
             prompt=prompt,
             negative_prompt="verybadimagenegative_v1.3",
             cfg_scale=3, 
             clip_skip=2,
-            controlnet_frames=input_video, num_frames=len(input_video),
+            controlnet_frames=input_video, num_frames=num_frames,
             num_inference_steps=num_inference_steps, 
-            height=512, 
-            width=512,
+            height=height, 
+            width=width,
             animatediff_batch_size=animatediff_batch_size, 
             animatediff_stride=animatediff_stride,
             vram_limit_level=0,
@@ -182,7 +190,11 @@ examples = [
 
 with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
     gr.HTML("<h1><center>Exvideo📽️Diffutoon</center></h1>")
-    gr.HTML("<p><center>Exvideo and Diffutoon video generation<br><b>Update</b>: first version<br><b>Note</b>: ZeroGPU limited, Set the parameters appropriately.</center></p>")
+    gr.HTML("""
+        <p><center>Exvideo and Diffutoon video generation
+        <br><b>Update</b>: Origin and Output resize, Frames control.
+        <br><b>Note</b>: ZeroGPU limited, Set the parameters appropriately.</center></p>
+        """)
     with gr.Row():
         video_in = gr.Video(label='Upload Video', height=600, scale=2)
         image_in = gr.Image(label='Upload Image', height=600, scale=2, image_mode="RGB", type="filepath", visible=False)
@@ -204,10 +216,50 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
                 label="Inference steps", 
                 info="Inference steps",
                 step=1,
-                value=5, 
+                value=10, 
                 minimum=1, 
-                maximum=50
+                maximum=50,
                 )
+            num_frames = gr.Slider(
+                label="Num frames", 
+                info="Output Frames",
+                step=1,
+                value=30,
+                minimum=1, 
+                maximum=128,
+            )
+            with gr.Row():
+                shape_height = gr.Slider(
+                    label="Shape Height",
+                    info="Resize Height",
+                    step=8,
+                    value=512, 
+                    minimum=256, 
+                    maximum=2560,
+                    )
+                shape_width = gr.Slider(
+                    label="Shape Width", 
+                    info="Resize Width",
+                    step=8,
+                    value=512, 
+                    minimum=256, 
+                    maximum=2560,
+                    )
+            with gr.Row():
+                height = gr.Slider(
+                    label="Output Height", 
+                    step=8,
+                    value=512, 
+                    minimum=256, 
+                    maximum=2560,
+                    )
+                width = gr.Slider(
+                    label="Output Width", 
+                    step=8,
+                    value=512, 
+                    minimum=256, 
+                    maximum=2560,
+                    )
             with gr.Accordion("Diffutoon Options", open=False):
                 animatediff_batch_size = gr.Slider(
                     label="Animatediff batch size",
@@ -222,7 +274,7 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
                     maximum=50,
                     step=1,
                     value=16,
-                ) 
+                )
             with gr.Accordion("ExVideo Options", open=False):
                 motion_bucket_id = gr.Slider(
                     label="Motion bucket id", 
@@ -230,7 +282,7 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
                     value=127, 
                     step=1,
                     minimum=1, 
-                    maximum=255
+                    maximum=255,
                 )
                 fps_id = gr.Slider(
                     label="Frames per second", 
@@ -238,17 +290,9 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
                     value=6,
                     step=1,
                     minimum=5, 
-                    maximum=30
-                )
-                num_frames = gr.Slider(
-                    label="Frames num", 
-                    info="Frames num",
-                    step=1,
-                    value=50, 
-                    minimum=1, 
-                    maximum=128
+                    maximum=30,
                 )
-    prompt = gr.Textbox(label="Prompt")
+    prompt = gr.Textbox(label="Prompt", value="best quality")
     with gr.Row():
         submit_btn = gr.Button(value="Generate")
         #stop_btn = gr.Button(value="Stop", variant="stop")
@@ -263,7 +307,8 @@ with gr.Blocks(css=CSS, js=JS, theme="soft") as demo:
         examples_per_page=4,
     )
     selected.change(change_media, inputs=[image_in, video_in, selected], outputs=[image_in, video_in, prompt])
-    submit_event = submit_btn.click(fn=generate, inputs=[video_in, image_in, selected, prompt, seed, num_inference_steps, animatediff_batch_size, animatediff_stride, motion_bucket_id, fps_id, num_frames], outputs=[video, seed], api_name="video")
+    video_in.upload(update_frames, inputs=[video_in], outputs=[num_frames])
+    submit_event = submit_btn.click(fn=generate, inputs=[video_in, image_in, selected, prompt, seed, num_inference_steps, num_frames, shape_height, shape_width, height, width, animatediff_batch_size, animatediff_stride, motion_bucket_id, fps_id], outputs=[video, seed], api_name="video")
     #stop_btn.click(fn=None, inputs=None, outputs=None, cancels=[submit_event])
     
 demo.queue().launch()