Update app.py

app.py

@@ -41,7 +41,6 @@ EMOTIVE_TAGS = ["<laugh>", "<chuckle>", "<sigh>", "<cough>", "<sniffle>", "<groa
 # Initialize Dash app
 app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
 
-# Layout
 app.layout = dbc.Container([
     dbc.Row([
         dbc.Col([
@@ -80,15 +79,17 @@ app.layout = dbc.Container([
                 className="mb-2"
             ),
             dbc.Button("Generate Podcast Script", id="generate-script-btn", color="primary", className="mb-2"),
+            dbc.Spinner(html.Div(id="script-loading"), color="primary"),
         ], width=6),
         dbc.Col([
             dbc.Textarea(id="script-output", placeholder="Generated script will appear here...", rows=10, className="mb-2"),
-            dbc.Button("Clear", id="clear-btn", color="secondary", className="mb-2"),
-            html.Label("Voice 1", className="mt-2"),
+            dbc.Button("Clear", id="clear-btn", color="secondary", className="mb-2 d-block"),
+            html.Label("Voice 1", className="mt-3"),
             dcc.Dropdown(id="voice1", options=[{"label": v, "value": v} for v in VOICES], value="tara", className="mb-2"),
             html.Label("Voice 2", className="mt-2"),
             dcc.Dropdown(id="voice2", options=[{"label": v, "value": v} for v in VOICES], value="zac", className="mb-2"),
             dbc.Button("Generate Audio", id="generate-audio-btn", color="success", className="mb-2"),
+            dbc.Spinner(html.Div(id="audio-loading"), color="primary"),
             html.Div(id="audio-output"),
             dbc.Button("Advanced Settings", id="advanced-settings-toggle", color="info", className="mb-2"),
             dbc.Collapse([
@@ -166,7 +167,9 @@ def generate_audio(script_output, voice1, voice2, num_hosts, temperature, top_p,
                 )
             
             code_list = parse_output(generated_ids)
-            paragraph_audio = redistribute_codes(code_list, snac_model)
+            # Use the correct method for audio generation based on your SNAC model
+            paragraph_audio = snac_model.generate_audio(torch.tensor(code_list).unsqueeze(0).to(device))
+            paragraph_audio = paragraph_audio.cpu().numpy().flatten()
             
             silences = detect_silence(paragraph_audio)
             if silences:
@@ -181,12 +184,14 @@ def generate_audio(script_output, voice1, voice2, num_hosts, temperature, top_p,
     except Exception as e:
         logger.error(f"Error generating speech: {str(e)}")
         return None
-
+        
 @callback(
     Output("script-output", "value"),
     Output("audio-output", "children"),
     Output("advanced-settings", "is_open"),
     Output("prompt", "value"),
+    Output("script-loading", "children"),
+    Output("audio-loading", "children"),
     Input("generate-script-btn", "n_clicks"),
     Input("generate-audio-btn", "n_clicks"),
     Input("advanced-settings-toggle", "n_clicks"),
@@ -214,7 +219,7 @@ def combined_callback(generate_script_clicks, generate_audio_clicks, advanced_se
                       script_output, voice1, voice2, temperature, top_p, repetition_penalty, max_new_tokens, is_advanced_open):
     ctx = dash.callback_context
     if not ctx.triggered:
-        return dash.no_update, dash.no_update, dash.no_update, dash.no_update
+        return dash.no_update, dash.no_update, dash.no_update, dash.no_update, "", ""
 
     trigger_id = ctx.triggered[0]['prop_id'].split('.')[0]
 
@@ -286,14 +291,14 @@ def combined_callback(generate_script_clicks, generate_audio_clicks, advanced_se
             """
             
             response = model.generate_content(prompt_template)
-            return re.sub(r'[^a-zA-Z0-9\s.,?!<>]', '', response.text), dash.no_update, dash.no_update, dash.no_update
+            return re.sub(r'[^a-zA-Z0-9\s.,?!<>]', '', response.text), dash.no_update, dash.no_update, dash.no_update, "", ""
         except Exception as e:
             logger.error(f"Error generating podcast script: {str(e)}")
-            return f"Error: {str(e)}", dash.no_update, dash.no_update, dash.no_update
+            return f"Error: {str(e)}", dash.no_update, dash.no_update, dash.no_update, "", ""
 
     elif trigger_id == "generate-audio-btn":
         if not script_output.strip():
-            return dash.no_update, html.Div("No audio generated yet."), dash.no_update, dash.no_update
+            return dash.no_update, html.Div("No audio generated yet."), dash.no_update, dash.no_update, "", ""
         
         final_audio = generate_audio(script_output, voice1, voice2, num_hosts, temperature, top_p, repetition_penalty, max_new_tokens)
         
@@ -309,17 +314,17 @@ def combined_callback(generate_script_clicks, generate_audio_clicks, advanced_se
                 html.Audio(src=src, controls=True),
                 html.Br(),
                 download_link
-            ]), dash.no_update, dash.no_update
+            ]), dash.no_update, dash.no_update, "", ""
         else:
-            return dash.no_update, html.Div("Error generating audio"), dash.no_update, dash.no_update
+            return dash.no_update, html.Div("Error generating audio"), dash.no_update, dash.no_update, "", ""
 
     elif trigger_id == "advanced-settings-toggle":
-        return dash.no_update, dash.no_update, not is_advanced_open, dash.no_update
+        return dash.no_update, dash.no_update, not is_advanced_open, dash.no_update, "", ""
 
     elif trigger_id == "clear-btn":
-        return "", html.Div("No audio generated yet."), dash.no_update, ""
+        return "", html.Div("No audio generated yet."), dash.no_update, "", "", ""
 
-    return dash.no_update, dash.no_update, dash.no_update, dash.no_update
+    return dash.no_update, dash.no_update, dash.no_update, dash.no_update, "", ""
 
 # Run the app
 if __name__ == '__main__':