Update app.py

app.py

@@ -59,7 +59,7 @@ def load_audio_special(sr, data):
         norm_fix = 2 ** 15
     elif data.dtype == np.float16 or data.dtype == np.float32:
         norm_fix = 1.
-    audio = torch.FloatTensor(data.astype(np.float32)) / norm_fix, sampling_rate)
+    audio = torch.FloatTensor(data.astype(np.float32)) / norm_fix, sampling_rate))
     
     # Remove any channel data.
     if len(audio.shape) > 1:
@@ -85,7 +85,7 @@ def inference_own_voice(text, voice_1, voice_2, voice_3):
     conds = [
         load_audio_special(voice_1),
         load_audio_special(voice_2),
-        load_audio_special(voice_1_3),
+        load_audio_special(voice_3),
     ]
     print(text, conds, preset)
     gen = tts.tts_with_preset(text, conds, preset)