Update app.py

app.py

@@ -15,8 +15,8 @@ asr_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-base",
 #processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")
 #model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts").to(device)
 
-processor = AutoProcessor.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl")
-model = SpeechT5ForTextToSpeech.from_pretrained("sanchit-gandhi/speecht5_tts_vox_nl").to(device)
+processor = AutoProcessor.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_es")
+model = SpeechT5ForTextToSpeech.from_pretrained("kfahn/speecht5_finetuned_voxpopuli_es").to(device)
 
 vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan").to(device)
 
@@ -25,7 +25,7 @@ speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze
 
 
 def translate(audio):
-    outputs = asr_pipe(audio, max_new_tokens=256, generate_kwargs={"task": "transcribe", "language": "nl"})
+    outputs = asr_pipe(audio, max_new_tokens=128, generate_kwargs={"task": "transcribe", "language": "es"})
     return outputs["text"]