Update app.py

app.py

@@ -1,17 +1,15 @@
 import gradio as gr
+from gradio import components
 import whisper
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 from gtts import gTTS
 import soundfile as sf
-import scipy.io.wavfile as wav
 import os
 
 def translate_speech_to_speech(input_audio):
     # Save the input audio to a temporary file
-    if isinstance(input_audio, tuple):
-        input_audio = input_audio[0]
-    input_file = "input_audio.wav"
-    sf.write(input_file, input_audio, 44100)
+    input_file = "input_audio" + os.path.splitext(input_audio.name)[1]
+    sf.write(input_file, input_audio[0], 44100)
     
     # Language detection and translation code from the first code snippet
     model = whisper.load_model("base")
@@ -48,8 +46,8 @@ 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 = components.InputComponent(source="microphone", type="numpy")
+output_audio = components.OutputComponent(type="numpy")
 
 stt_demo = gr.Interface(
     fn=translate_speech_to_speech,
@@ -60,5 +58,5 @@ stt_demo = gr.Interface(
 )
 
 if __name__ == "__main__":
-    stt_demo.launch()
+    stt_demo.launch(share=True)