Update app.py

app.py

@@ -18,19 +18,15 @@ def audio_to_text(audio):
     # Convert to mono if the audio has more than one channel
     if len(audio_data.shape) > 1:
         audio_data = np.mean(audio_data, axis=1)
-    
-    
-    
     audio_array = np.array(audio_data).astype(np.float32)
     
     # Transcribe the audio input
-    transcription = pipe1([audio_array])
-    transcription_text = transcription['text']
+    transcription = pipe1(audio_array, return_timestamps=True)
+    transcription_text = " ".join([segment["text"] for segment in transcription["segments"]])
     
-    # Print and return the transcription text
     print("Transcription:", transcription_text)
     return transcription_text
 
 
 demo = gr.Interface(fn=audio_to_text, inputs=gr.Audio(), outputs="text")
-demo.launch(share=True)
+demo.launch(share=True)