Update app.py

app.py

@@ -20,7 +20,7 @@ pipe = pipeline(task="automatic-speech-recognition", model=model_repo_id, device
 def infer(
     audio
 ):
-    return pipe(audio, generate_kwargs={'language': 'chinese'})['text']
+    return pipe(audio[1], generate_kwargs={'language': 'chinese'})['text']
 
 css = """
 #col-container {