Update app.py

app.py

@@ -21,7 +21,7 @@ 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"})
-    return text
+    return outputs['text']
     
 def synthesise(text):
     inputs = tokenizer(text, return_tensors="pt")