Update app.py

app.py

@@ -14,8 +14,9 @@ pipe = SpeechToSpeechPipeline("facebook/xm_transformer_unity_hk-en")
 
 def call_model(audio, model):
     # pipe = SpeechToSpeechPipeline("facebook/xm_transformer_unity_hk-en")
-    wav, sr, text = pipe(audio)
-    return gr.Audio((sr, wav))
+    # wav, sr, text = pipe(audio)
+    temp_file = pipe(audio)
+    return gr.Audio(temp_file)
 
 def inference(audio, model):
     if model == "xm_transformer_s2ut_en-hk":