Update app.py

app.py

@@ -173,7 +173,7 @@ def fetch_local_weather():
             }}
             .weather-content {{
                 display: flex;
-                align-items: center;
+                align-items: center.
             }}
             .weather-icon {{
                 flex: 1;
@@ -284,16 +284,14 @@ def bot(history, choice, tts_model):
 
     # Generate audio and process output prompt in parallel
     with concurrent.futures.ThreadPoolExecutor() as executor:
-        audio_future = executor.submit(generate_audio_chunks, tts_model, response)
+        audio_chunks_future = executor.submit(generate_audio_chunks, tts_model, response)
         text_chunks = [response[i:i + 100] for i in range(0, len(response), 100)]  # Chunk the text for streaming
-        for text_chunk in text_chunks:
+        
+        audio_chunks = audio_chunks_future.result()
+        for text_chunk, audio_chunk in zip(text_chunks, audio_chunks):
             history[-1][1] += text_chunk
-            audio_chunk = next(audio_future.result(), None)
-            time.sleep(0.2)  # Adjust this to synchronize text and audio appearance
             yield history, audio_chunk
-        
-        for remaining_audio_chunk in audio_future.result():
-            yield history, remaining_audio_chunk
+            time.sleep(0.2)  # Adjust this to synchronize text and audio appearance
 
 def generate_audio_chunks(tts_model, text):
     if tts_model == "ElevenLabs":
@@ -320,12 +318,16 @@ def generate_audio_elevenlabs_chunks(text):
         }
     }
     response = requests.post(tts_url, headers=headers, json=data, stream=True)
+    audio_chunks = []
     if response.ok:
-        for chunk in response.iter_content(chunk_size=1024):
-            yield chunk
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as f:
+            for chunk in response.iter_content(chunk_size=1024):
+                f.write(chunk)
+                audio_chunks.append(f.name)
+        return audio_chunks
     else:
         logging.error(f"Error generating audio: {response.text}")
-        return None
+        return []
 
 def generate_audio_parler_tts_chunks(text):
     model_id = 'parler-tts/parler_tts_mini_v0.1'
@@ -347,11 +349,13 @@ def generate_audio_parler_tts_chunks(text):
     audio_arr = generation.cpu().numpy().squeeze()
 
     chunk_size = 16000  # Define a chunk size (adjust as needed)
+    audio_chunks = []
     for i in range(0, len(audio_arr), chunk_size):
         audio_chunk = audio_arr[i:i + chunk_size]
         with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as f:
             sf.write(f.name, audio_chunk, model.config.sampling_rate)
-            yield f.name
+            audio_chunks.append(f.name)
+    return audio_chunks
 
 def add_message(history, message):
     history.append((message, None))