Passing PyTorch Batch to HuggingFace Wav2vce2 Model

I have a PyTorch batch of size say (2, 1, 100)
and I want to pass this batch to HuggingFace Wav2Vec2 Model, first to feature_extractor and then to base wav2vec2 model

  1. input in forward function is x of shape (2,1,100)
    What I want in forward function is like this
x_enc = feature_extractor(x sampling_rate=16000)
out = wav2vec2_model(x)

It might seems naive, but any help would be appreciated

Does anyone have any idea how to make it compatible with Hugging Face

Thank you