Update app.py

app.py

@@ -611,16 +611,10 @@ class StreamingTTS:
                 return self.output_file
         return None
 
-    def stop(self):
-        """Stop generation"""
-        self.should_stop = True
-        print("Stop request received")
-
 # Load reference example (Malayalam)
 EXAMPLES = [{
     "audio_url": "https://raw.githubusercontent.com/Aparna0112/voicerecording-_TTS/main/KC%20Voice.wav",
     "ref_text": "ഹലോ ഇത് അപരനെ അല്ലേ ഞാൻ ജഗദീപ് ആണ് വിളിക്കുന്നത് ഇപ്പോൾ ഫ്രീയാണോ സംസാരിക്കാമോ ",
-    "synth_text": "ഞാൻ മലയാളം സംസാരിക്കാൻ കഴിയുന്നു."
 }]
 
 print("\nPreloading reference audio...")
@@ -634,33 +628,20 @@ if ref_audio is None:
 # Initialize streaming TTS
 streaming_tts = StreamingTTS()
 
-# Add a stop button functionality
-def stop_generation():
-    streaming_tts.stop()
-    return "Generation stopped"
-
-# Gradio interface with offline mode
+# Gradio interface with simplified UI
 with gr.Blocks() as iface:
     gr.Markdown("## 🚀 IndicF5 Malayalam TTS")
 
-    with gr.Row():
-        gr.Markdown("### System Status:")
-        system_status = gr.Markdown(f"- Device: {device}\n- Model loaded: {'Yes' if model is not None else 'No'}\n- Reference audio: {'Loaded' if ref_audio is not None else 'Not loaded'}")
-
     with gr.Row():
         text_input = gr.Textbox(
-            label="Malayalam Text",
+            label="Enter Malayalam Text",
             placeholder="Enter text here...",
             lines=3,
-            value=EXAMPLES[0]["synth_text"] if EXAMPLES else "ഹലോ, എന്തൊക്കെ ഉണ്ട് വിശേഷം?"
+            value=""  # Start with empty field
         )
 
     with gr.Row():
         generate_btn = gr.Button("🎤 Generate Speech", variant="primary")
-        stop_btn = gr.Button("🛑 Stop Generation", variant="secondary")
-
-    # Status indicator
-    status_text = gr.Textbox(label="Status", value="Ready", interactive=False)
 
     # Audio output
     output_audio = gr.Audio(
@@ -669,52 +650,40 @@ with gr.Blocks() as iface:
         autoplay=True
     )
 
-    # Debug information (hidden by default)
-    with gr.Accordion("Advanced", open=False):
-        debug_output = gr.Textbox(label="Debug Log", value="", lines=5)
-
     def start_generation(text):
         if not text.strip():
-            return None, "Please enter some text", "Error: Empty text input"
-
+            return None
+        
         if model is None:
-            return None, "⚠️ Model not loaded. Cannot generate speech.", "Error: Model not loaded"
-
+            return None
+            
         if ref_audio is None:
-            return None, "⚠️ Reference audio not loaded. Cannot generate speech.", "Error: Reference audio not loaded"
+            return None
 
         # Print the text being processed
         print(f"🔍 User input text: '{text}'")
         
-        # Capture stdout for debug purposes
-        import io
-        from contextlib import redirect_stdout
-        f = io.StringIO()
-        with redirect_stdout(f):
-            try:
-                # Make sure the text is explicitly passed as the first parameter
-                streaming_tts.generate(
-                    text=text,  # Explicitly name parameter 
-                    ref_audio=ref_audio, 
-                    ref_sr=ref_sr, 
-                    ref_text=EXAMPLES[0]["ref_text"] if EXAMPLES else ""
-                )
-            except Exception as e:
-                print(f"Error starting generation: {e}")
-
-        debug_log = f.getvalue()
+        try:
+            # Generate speech for the new text
+            streaming_tts.generate(
+                text=text,
+                ref_audio=ref_audio, 
+                ref_sr=ref_sr, 
+                ref_text=EXAMPLES[0]["ref_text"] if EXAMPLES else ""
+            )
+        except Exception as e:
+            print(f"Error starting generation: {e}")
 
         # Add a delay to ensure file is created
         time.sleep(2.0)
 
         audio_path = streaming_tts.get_current_audio()
         if audio_path and os.path.exists(audio_path) and os.path.getsize(audio_path) > 0:
-            return audio_path, f"Generated speech for: {text[:30]}...", debug_log
+            return audio_path
         else:
-            return None, "Starting generation... please wait", debug_log
+            return None
 
-    generate_btn.click(start_generation, inputs=text_input, outputs=[output_audio, status_text, debug_output])
-    stop_btn.click(stop_generation, inputs=None, outputs=status_text)
+    generate_btn.click(start_generation, inputs=text_input, outputs=output_audio)
 
 # Cleanup on exit
 def exit_handler():