Update app.py

app.py

@@ -37,28 +37,24 @@ def text_to_speech_kokoro(text: str, voice: str = 'af_heart', speed: float = 1.0
         speed: The speed of the speech (default is 1.0).
     
     Returns:
-        A tuple containing the sample rate and the generated audio data as a NumPy array.
+        The filename of the generated audio file.
     """
     try:
         # Generate speech audio
         generator = pipeline(text, voice=voice, speed=speed, split_pattern=r'\n+')
-        audio_segments = []
+        audio_files = []
 
-         # Collect each audio segment
-        for _, _, audio in generator:
-            audio_segments.append(audio)
+        # Save each audio segment to a file
+        for i, (gs, ps, audio) in enumerate(generator):
+            os.makedirs("static", exist_ok=True)
+            filename = os.path.join("static", f'output_{i}.wav')
+            sf.write(filename, audio, 24000)
+            audio_files.append(filename)
 
-        # Concatenate all audio segments into a single array
-        if audio_segments:
-            full_audio = np.concatenate(audio_segments)
-            sample_rate = 24000  # Kokoro-82M outputs audio at 24 kHz
-            return sample_rate, full_audio
-        else:
-            return "No audio generated."
+        return f"Generated {len(audio_files)} audio file(s): {', '.join(audio_files)}"
     except Exception as e:
         return f"Error generating speech: {str(e)}"
 
-
 @tool
 def search_dad_jokes(term: str) -> str:
     """A tool that searches for dad jokes containing a specific term.