Update app.py

app.py

@@ -38,7 +38,7 @@ def infer(text, audio_prompt, rms, t_shift, num_steps, speed, return_smooth):
         speed=speed, 
         return_smooth=return_smooth
     )
-    final_wav = final_wav.cpu().numpy()
+    final_wav = final_wav.cpu().squeeze(0).numpy()
     final_wav = (np.clip(final_wav, -1.0, 1.0) * 32767).astype(np.int16)
     return (48000, final_wav)