Update app.py

app.py

@@ -255,36 +255,44 @@ with gr.Blocks(css=css) as demo:
     with gr.Column(elem_id="col-container"):
         gr.Markdown("# Hibiki ")
         gr.Markdown("This is a simple demo for Kyutai's Hibiki translation models • Currently supports French to English only.")
-        video_input = gr.Video(label="Video IN")
-        audio_input = gr.Audio(label="Audio IN", type="filepath")
-        submit_btn = gr.Button("Generate translations")
-        output_result = gr.Audio(label="Translated result")
-
-        with gr.Row():
-            dropdown_wav_selector = gr.Dropdown(
-                label="Pick a generated audio to load",
-                value = None,
-                visible=False,
-                scale=2
-            )
-            choose_this_btn = gr.Button("Apply this one as translated audio overlay", scale=1, visible=False)
+        
         with gr.Row():
-            volume_reduction = gr.Slider(label="Volume reduction", minimum=0, maximum=60, step=1, value=30, visible=False)
-            cut_start = gr.Slider(label="Cut start", minimum=0.0, maximum=4.0, step=0.1, value=2.0, visible=False)
-        combined_output = gr.Audio(label="Combinated Audio", type="filepath", visible=False, show_download_button=True)
-        apply_to_video_btn = gr.Button("Apply to your video", visible=False)
-        final_video_out = gr.Video(label="Video + Translated Audio", visible=False)
-        with gr.Accordion("Downloadable audio Output list", open=False, visible=False) as result_accordion:
-            wav_list = gr.Files(label="Output Audio List", visible=False)
-
-        gr.Examples(
-            examples = [
-                "./examples/sample_fr_hibiki_intro.mp3",
-                "./examples/sample_fr_hibiki_crepes.mp3",
-                "./examples/sample_fr_hibiki_monologue_otis.mp3"
-            ],
-            inputs = [audio_input]
-        )
+            with gr.Column():
+                video_input = gr.Video(label="Video IN (Optional)")
+                audio_input = gr.Audio(label="Audio IN", type="filepath")
+                submit_btn = gr.Button("Generate translations")
+
+                gr.Examples(
+                    examples = [
+                        "./examples/sample_fr_hibiki_intro.mp3",
+                        "./examples/sample_fr_hibiki_crepes.mp3",
+                        "./examples/sample_fr_hibiki_monologue_otis.mp3"
+                    ],
+                    inputs = [audio_input]
+                )
+        
+            with gr.Column():
+                output_result = gr.Audio(label="Translated result")
+
+                with gr.Row():
+                    dropdown_wav_selector = gr.Dropdown(
+                        label="Pick a generated translated audio to load",
+                        value = None,
+                        visible=False,
+                        scale=2
+                    )
+                    choose_this_btn = gr.Button("Apply and check this one as translated audio overlay", scale=1, visible=False)
+                
+                with gr.Row():
+                    volume_reduction = gr.Slider(label="Original audio Volume reduction", minimum=0, maximum=60, step=1, value=30, visible=False)
+                    cut_start = gr.Slider(label="Reduce translator delay (seconds)", minimum=0.0, maximum=4.0, step=0.1, value=2.0, visible=False)
+               
+                combined_output = gr.Audio(label="Combinated Audio", type="filepath", visible=False, show_download_button=True)
+                apply_to_video_btn = gr.Button("Apply this combination to your video", visible=False)
+                
+                final_video_out = gr.Video(label="Video + Translated Audio", visible=False)
+                with gr.Accordion("Downloadable audio Output list", open=False, visible=False) as result_accordion:
+                    wav_list = gr.Files(label="Output Audio List", visible=False)     
 
     audio_input.upload(
         fn = clean_previous_video_input,