Add sfedar/speecht5_finetuned_voxpopuli_es model to synthesise

app.py

@@ -12,10 +12,10 @@ 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
-processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")
-
-model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts").to(device)
-vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan").to(device)
+tts_model_name = "sfedar/speecht5_finetuned_voxpopuli_es"
+processor = SpeechT5Processor.from_pretrained(tts_model_name)
+model = SpeechT5ForTextToSpeech.from_pretrained(tts_model_name).to(device)
+vocoder = SpeechT5HifiGan.from_pretrained(tts_model_name).to(device)
 
 embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
 speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)