Update app.py

app.py

@@ -25,8 +25,11 @@ def transcribe_audio(audio_bytes):
     # Preprocess audio
     input_features = preprocess_audio(audio_bytes)
 
+    # Tokenize audio
+    inputs = processor(input_features, return_tensors="pt", padding=True)
+
     # Generate transcription
-    generated_ids = model.generate(input_features)
+    generated_ids = model.generate(inputs.input_features)
     translation = processor.batch_decode(generated_ids, skip_special_tokens=True)
 
     return translation