Update app.py

app.py

@@ -8,12 +8,10 @@ import os
 import numpy as np
 
 def translate_speech_to_speech(input_audio):
-    # Convert stereo audio to mono
-    mono_audio = np.mean(input_audio, axis=1)
     
     # Save the input audio to a temporary file
     input_file = "input_audio.wav"
-    sf.write(input_file, mono_audio, 44100)
+    sf.write(input_file, input_audio, 44100)
     
     # Language detection and translation code from the first code snippet
     model = whisper.load_model("base")
@@ -50,8 +48,8 @@ def translate_speech_to_speech(input_audio):
 
 title = "Speech-to-Speech Translator"
 
-input_audio = gr.inputs.Audio(source="microphone", type="numpy")
-output_audio = gr.outputs.Audio(type="numpy")
+input_audio = gr.inputs.Audio(source="microphone")
+output_audio = gr.outputs.Audio()
 
 stt_demo = gr.Interface(
     fn=translate_speech_to_speech,