Update app.py

app.py

@@ -183,7 +183,7 @@ def infer(
         del x_tst, tones, lang_ids, bert, x_tst_lengths, speakers, ja_bert, en_bert, emo
         if torch.cuda.is_available():
             torch.cuda.empty_cache()
-        return (hps.data.sampling_rate,audio)
+        return (hps.data.sampling_rate,gr.processing_utils.convert_to_16_bit_wav(audio))
 
 def is_japanese(string):
         for ch in string:
@@ -203,7 +203,7 @@ if __name__ == "__main__":
     emotional_model = EmotionModel.from_pretrained(emotional_model_name).to(device)
     hps = utils.get_hparams_from_file('Data/BanGDream/configs/config.json')
     net_g = get_net_g(
-        model_path='Data/BangDream/models/G_49000.pth', version="2.1", device=device, hps=hps
+        model_path='Data/BanGDream/models/G_63000.pth', version="2.1", device=device, hps=hps
     )
     speaker_ids = hps.data.spk2id
     speakers = list(speaker_ids.keys())