Update app.py

app.py

@@ -184,54 +184,58 @@ with gr.Blocks(title="Vevo Voice Conversion") as demo:
             interactive=True
         )
     
-    with gr.Row():
-        with gr.Column():
-            with gr.Group(visible=True) as audio_inputs:
-                content_audio = gr.Audio(
-                    label="Source Audio",
-                    type="filepath",
-                    interactive=True
-                )
-                
-                ref_style_audio = gr.Audio(
-                    label="Reference Style Audio",
-                    type="filepath",
-                    interactive=True
-                )
-                
-                ref_timbre_audio = gr.Audio(
-                    label="Reference Timbre Audio",
-                    type="filepath",
-                    interactive=True
-                )
-            
-        with gr.Column():
-            with gr.Group(visible=False) as text_inputs:
-                src_text = gr.Textbox(
-                    label="Source Text",
-                    placeholder="Enter text for TTS mode",
-                    interactive=True
-                )
-                
-                ref_text = gr.Textbox(
-                    label="Reference Style Text",
-                    placeholder="Optional: Enter reference text",
-                    interactive=True
-                )
-                
-                src_language = gr.Dropdown(
-                    choices=["en", "zh"],
-                    value="en",
-                    label="Source Language",
-                    interactive=True
-                )
-                
-                ref_language = gr.Dropdown(
-                    choices=["en", "zh"],
-                    value="en",
-                    label="Reference Language",
-                    interactive=True
-                )
+    # Audio inputs
+    content_audio = gr.Audio(
+        label="Source Audio",
+        type="filepath",
+        visible=True,
+        interactive=True
+    )
+    
+    ref_style_audio = gr.Audio(
+        label="Reference Style Audio",
+        type="filepath",
+        visible=True,
+        interactive=True
+    )
+    
+    ref_timbre_audio = gr.Audio(
+        label="Reference Timbre Audio",
+        type="filepath",
+        visible=True,
+        interactive=True
+    )
+    
+    # Text inputs
+    src_text = gr.Textbox(
+        label="Source Text",
+        placeholder="Enter text for TTS mode",
+        visible=False,
+        interactive=True
+    )
+    
+    ref_text = gr.Textbox(
+        label="Reference Style Text",
+        placeholder="Optional: Enter reference text",
+        visible=False,
+        interactive=True
+    )
+    
+    src_language = gr.Dropdown(
+        choices=["en", "zh"],
+        value="en",
+        label="Source Language",
+        visible=False,
+        interactive=True
+    )
+    
+    ref_language = gr.Dropdown(
+        choices=["en", "zh"],
+        value="en",
+        label="Reference Language",
+        visible=False,
+        interactive=True
+    )
     
     with gr.Row():
         steps = gr.Slider(
@@ -249,16 +253,32 @@ with gr.Blocks(title="Vevo Voice Conversion") as demo:
     # Handle visibility of components based on mode
     def update_interface(mode):
         is_tts = mode == "tts"
-        return {
-            audio_inputs: not is_tts,
-            text_inputs: is_tts,
-            ref_style_audio: mode != "timbre",
-        }
+        is_timbre = mode == "timbre"
+        
+        return [
+            # Audio components visibility
+            not is_tts,  # content_audio
+            not is_timbre and not is_tts,  # ref_style_audio
+            True,  # ref_timbre_audio (always visible)
+            # Text components visibility
+            is_tts,  # src_text
+            is_tts,  # ref_text
+            is_tts,  # src_language
+            is_tts   # ref_language
+        ]
     
     mode.change(
         fn=update_interface,
         inputs=[mode],
-        outputs=[audio_inputs, text_inputs, ref_style_audio]
+        outputs=[
+            content_audio,
+            ref_style_audio,
+            ref_timbre_audio,
+            src_text,
+            ref_text,
+            src_language,
+            ref_language
+        ]
     )
     
     # Handle generation