update

app.py

@@ -77,10 +77,11 @@ MARKDOWN = \
     Only GestureNet is provided in this Gradio Demo, you can check the full test code for all pretrained weight available.
 
     ### Note: The index we put the gesture point by default here is [4, 10] for two gesture points or [4] for one gesture point.
-    ### Note: The result now only support is 256x384
+    ### Note: The result now only support is 256x384.
     ### Note: Click "Clear All" to restart everything; Click "Undo Point" to cancel the point you put
+    ### Note: The first run may be long. Click "Clear All" for each run is the safest choice.
     
-    If This&That is helpful, please help star the [GitHub Repo](https://github.com/Kiteretsu77/This_and_That_VDM). Thanks! 
+    If **This&That** is helpful, please help star the [GitHub Repo](https://github.com/Kiteretsu77/This_and_That_VDM). Thanks! 
     """
 
 
@@ -383,7 +384,7 @@ if __name__ == '__main__':
             original_image = gr.State(value=None) # store original input image
             with gr.Row():
                 with gr.Column():
-                    gr.Markdown("""<p style="text-align: center; font-size: 20px">Click two Points</p>""")
+                    gr.Markdown("""<p style="text-align: center; font-size: 30px">Click two Points</p>""")
                     input_image = gr.Image(label="Input Image", height=HEIGHT, width=WIDTH, interactive=False, elem_id="input_img")
                     # gr.Image(type="numpy", label="Click Points", height=HEIGHT, width=WIDTH, interactive=False) # for points clicking
                     undo_button = gr.Button("Undo point")
@@ -393,7 +394,7 @@ if __name__ == '__main__':
                         prompt = gr.Textbox(label="Text Prompt")
 
                 with gr.Column():
-                    gr.Markdown("""<p style="text-align: center; font-size: 20px">Results</p>""")
+                    gr.Markdown("""<p style="text-align: center; font-size: 30px">Results</p>""")
                     frames = gr.Video(value=None, label="Generate Video", show_label=True, height=HEIGHT, width=WIDTH)
                     with gr.Row():
                         run_button = gr.Button("Run")

svd/pipeline_stable_video_diffusion_controlnet.py

@@ -622,7 +622,7 @@ class StableVideoDiffusionControlNetPipeline(DiffusionPipeline):
         self._num_timesteps = len(timesteps)
         with self.progress_bar(total=num_inference_steps) as progress_bar:
             for i, t in enumerate(timesteps):
-                print("This is timestep ", t)
+                # print("This is timestep ", t)
 
                 # expand the latents if we are doing classifier free guidance
                 if use_instructpix2pix: