Update app.py

app.py

@@ -18,8 +18,8 @@ model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts").to(dev
 vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan").to(device)
 
 
-model = VitsModel.from_pretrained("Matthijs/mms-tts-deu").to(device)
-tokenizer = VitsTokenizer.from_pretrained("Matthijs/mms-tts-deu")
+model_mms = VitsModel.from_pretrained("Matthijs/mms-tts-deu").to(device)
+tokenizer_mms = VitsTokenizer.from_pretrained("Matthijs/mms-tts-deu")
 
 embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
 speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)
@@ -31,16 +31,16 @@ def translate(audio):
 
 
 def synthesise(text):
-    inputs = tokenizer(text, return_tensors="pt")
+    inputs = tokenizer_mms(text, return_tensors="pt")
     input_ids = inputs["input_ids"]
 
 
     with torch.no_grad():
-        outputs = model(input_ids)
-    print(outputs)
+        outputs = model_mms(input_ids)
+    print("mms model", outputs)
     inputs = processor(text=text, return_tensors="pt")
     speech = model.generate_speech(inputs["input_ids"].to(device), speaker_embeddings.to(device), vocoder=vocoder)
-    print(speech)
+    print("speecht5 model", speech)
     return speech.cpu()