Update app.py

app.py

@@ -4,7 +4,7 @@ import torch
 from datasets import load_dataset
 
 from transformers import SpeechT5ForTextToSpeech, SpeechT5HifiGan, SpeechT5Processor, pipeline
-
+from transformers import VitsModel, VitsTokenizer
 
 device = "cuda:0" if torch.cuda.is_available() else "cpu"
 
@@ -12,12 +12,19 @@ device = "cuda:0" if torch.cuda.is_available() else "cpu"
 asr_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-base", device=device)
 
 # load text-to-speech checkpoint and speaker embeddings
+
+
+
+#model = VitsModel.from_pretrained("Matthijs/mms-tts-deu")
+#tokenizer = VitsTokenizer.from_pretrained("Matthijs/mms-tts-deu")
+
+
 #processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")
-#processor = SpeechT5Processor.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_cs")
-processor = SpeechT5Processor.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl")
+processor = SpeechT5Processor.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_cs")
+#processor = SpeechT5Processor.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl")
 
-model = SpeechT5ForTextToSpeech.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl").to(device)
-#model = SpeechT5ForTextToSpeech.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_cs").to(device)
+#model = SpeechT5ForTextToSpeech.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl").to(device)
+model = SpeechT5ForTextToSpeech.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_cs").to(device)
 #model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts").to(device)
 vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan").to(device)