Update app.py

app.py

@@ -25,6 +25,8 @@ def preprocess_audio(audio_data, sampling_rate=16_000):
 def transcribe_audio(audio_data):
     input_features = preprocess_audio(audio_data)
     input_values = torch.tensor(input_features["input_values"]).unsqueeze(0)  # Add batch dimension
+    # Reshape the tensor to match the expected shape [batch_size, num_channels, sequence_length]
+    input_values = input_values.permute(0, 2, 1)
     predicted_ids = model.generate(input_values)
     transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
     return transcription[0]