Update app.py

app.py

@@ -216,78 +216,72 @@ device = "cuda" if torch.cuda.is_available() else "cpu"
 with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css) as demo:
 
     with gr.Row():
+        with gr.Column(variant="panel"):
+            with gr.Row(equal_height=False):
+                with gr.Column():
+                    with gr.Row():
+                        init_img = gr.Video(label="Input Video")
+                    with gr.Row():
+                        prompt_textbox = gr.Textbox(value='', label="Prompt", lines=1)
+                    with gr.Row():
+                        negative_prompt_textbox = gr.Textbox(value=N_PROMPT, label="Negative prompt", lines=1)
 
-    with gr.Column(variant="panel"):
-        with gr.Row(equal_height=False):
-            with gr.Column():
-                with gr.Row():
-                    init_img = gr.Video(label="Input Video")
-                with gr.Row():
-                    prompt_textbox = gr.Textbox(value='', label="Prompt", lines=1)
-                with gr.Row():
-                    negative_prompt_textbox = gr.Textbox(value=N_PROMPT, label="Negative prompt", lines=1)
-
-                with gr.Row():
-                    ip_adapter_scale = gr.Slider(label="Visual Content Scale", value=1.0, minimum=0, maximum=1)
-                    temporal_scale = gr.Slider(label="Temporal Align Scale", value=0.2, minimum=0., maximum=1.0)
-
-                with gr.Accordion("Sampling Settings", open=False):
                     with gr.Row():
-                        sampler_dropdown = gr.Dropdown(
-                            label="Sampling method",
-                            choices=list(scheduler_dict.keys()),
-                            value=list(scheduler_dict.keys())[0],
-                        )
-                        sample_step_slider = gr.Slider(
-                            label="Sampling steps", value=25, minimum=10, maximum=100, step=1
-                        )
-                    cfg_scale_slider = gr.Slider(label="CFG Scale", value=7.5, minimum=0, maximum=20)
-
-                with gr.Row():
-                    seed_textbox = gr.Textbox(label="Seed", value=42)
-                    seed_button = gr.Button(value="\U0001f3b2", elem_classes="toolbutton")
-                seed_button.click(fn=lambda x: random.randint(1, 1e8), outputs=[seed_textbox], queue=False)
-
-                generate_button = gr.Button(value="Generate", variant="primary")
-
-            with gr.Column():
-                result_video = gr.Video(label="Generated Audio", interactive=False)
-                with gr.Row():
-
-        generate_button.click(
-            fn=controller.foley,
-            inputs=[
-                init_img,
-                prompt_textbox,
-                negative_prompt_textbox,
-                ip_adapter_scale,
-                temporal_scale,
-                sampler_dropdown,
-                sample_step_slider,
-                cfg_scale_slider,
-                seed_textbox,
-            ],
-            outputs=[result_video],
-        )
+                        ip_adapter_scale = gr.Slider(label="Visual Content Scale", value=1.0, minimum=0, maximum=1)
+                        temporal_scale = gr.Slider(label="Temporal Align Scale", value=0.2, minimum=0., maximum=1.0)
+
+                    with gr.Accordion("Sampling Settings", open=False):
+                        with gr.Row():
+                            sampler_dropdown = gr.Dropdown(
+                                label="Sampling method",
+                                choices=list(scheduler_dict.keys()),
+                                value=list(scheduler_dict.keys())[0],
+                            )
+                            sample_step_slider = gr.Slider(
+                                label="Sampling steps", value=25, minimum=10, maximum=100, step=1
+                            )
+                        cfg_scale_slider = gr.Slider(label="CFG Scale", value=7.5, minimum=0, maximum=20)
 
-        gr.Examples(
-            # examples= [
-            #     ['examples/videos/51701454.mp4', 'seagulls', '', 1.0, 'DDIM', 25, 7.5, 10014024412012338098],
-            #     ['examples/videos/42.mp4', '', '', 1.0, 'DDIM', 25, 7.5, 42],
-            #     ['examples/videos/1.mp4', '', '', 1.0, 'DDIM', 25, 7.5, 93493458],
-            #     ['examples/videos/2.mp4', '', '', 1.0, 'DDIM', 25, 7.5, 16520432],
-            # ],
-            examples=[
-                ['examples/input/case1.mp4', '', '', 1.0, 0.2, 'DDIM', 25, 7.5, 33817921],
-                ['examples/input/case3.mp4', '', '', 1.0, 0.2,'DDIM', 25, 7.5, 94667578],
-                ['examples/input/case5.mp4', '', '', 0.75, 0.2,'DDIM', 25, 7.5, 92890876],
-                ['examples/input/case6.mp4', '', '', 1.0, 0.2, 'DDIM', 25, 7.5, 77015909],
-            ],
-            inputs=[init_img,prompt_textbox,negative_prompt_textbox,ip_adapter_scale,temporal_scale,sampler_dropdown,sample_step_slider,cfg_scale_slider,seed_textbox],
-            cache_examples=True,
-            outputs=[result_video], 
-            fn=controller.foley,
-        )
-            
+                    with gr.Row():
+                        seed_textbox = gr.Textbox(label="Seed", value=42)
+                        seed_button = gr.Button(value="\U0001f3b2", elem_classes="toolbutton")
+                    seed_button.click(fn=lambda x: random.randint(1, 1e8), outputs=[seed_textbox], queue=False)
+
+                    generate_button = gr.Button(value="Generate", variant="primary")
+
+                with gr.Column():
+                    result_video = gr.Video(label="Generated Audio", interactive=False)
+                    with gr.Row():
+                        pass
+
+            generate_button.click(
+                fn=controller.foley,
+                inputs=[
+                    init_img,
+                    prompt_textbox,
+                    negative_prompt_textbox,
+                    ip_adapter_scale,
+                    temporal_scale,
+                    sampler_dropdown,
+                    sample_step_slider,
+                    cfg_scale_slider,
+                    seed_textbox,
+                ],
+                outputs=[result_video],
+            )
+
+            gr.Examples(
+                examples=[
+                    ['examples/input/case1.mp4', '', '', 1.0, 0.2, 'DDIM', 25, 7.5, 33817921],
+                    ['examples/input/case3.mp4', '', '', 1.0, 0.2,'DDIM', 25, 7.5, 94667578],
+                    ['examples/input/case5.mp4', '', '', 0.75, 0.2,'DDIM', 25, 7.5, 92890876],
+                    ['examples/input/case6.mp4', '', '', 1.0, 0.2, 'DDIM', 25, 7.5, 77015909],
+                ],
+                inputs=[init_img, prompt_textbox, negative_prompt_textbox, ip_adapter_scale, temporal_scale, sampler_dropdown, sample_step_slider, cfg_scale_slider, seed_textbox],
+                cache_examples=True,
+                outputs=[result_video], 
+                fn=controller.foley,
+            )
+                
     demo.queue(10)
-    demo.launch(server_name=args.server_name, server_port=args.port, share=args.share, allowed_paths=["./foleycrafter.png"])
+    demo.launch(server_name=args.server_name, server_port=args.port, share=args.share, allowed_paths=["./foleycrafter.png"])