Update app.py

app.py

@@ -6,7 +6,7 @@ import torch
 from transformers import SpeechT5Processor, SpeechT5ForTextToSpeech, SpeechT5HifiGan
 
 checkpoint = "burraco135/speecht5_finetuned_voxpopuli_it"
-# processor = SpeechT5Processor.from_pretrained(checkpoint)
+processor = SpeechT5Processor.from_pretrained(checkpoint)
 model = SpeechT5ForTextToSpeech.from_pretrained(checkpoint)
 vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")
 
@@ -24,9 +24,8 @@ def predict(text, speaker):
 
     speech = model.generate_speech(input_ids, speaker_embedding, vocoder=vocoder)
 
-    # speech = (speech.numpy() * 32767).astype(np.int16)
-    # return (16000, speech)
-    return speech
+    speech = (speech.numpy() * 32767).astype(np.int16)
+    return (16000, speech)
 
 gr.Interface(
     fn=predict,