Update app.py

app.py

@@ -3,7 +3,7 @@ import numpy as np
 import torch
 from datasets import load_dataset
 
-from transformers import VitsModel, AutoTokenizer, pipeline
+from transformers import VitsModel, VitsTokenizer, pipeline
 
 
 device = "cuda:0" if torch.cuda.is_available() else "cpu"
@@ -15,7 +15,7 @@ asr_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-base",
 # processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")
 
 model = VitsModel.from_pretrained("facebook/mms-tts-por").to(device)
-tokenizer = AutoTokenizer.from_pretrained("facebook/mms-tts-por")
+tokenizer = VitsTokenizer.from_pretrained("facebook/mms-tts-por")
 
 # embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
 # speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)