Normalisation in recurrent network PyTorch LSTM

Question

I am trying to normalise between layers of my stacked LSTM network in PyTorch. The network looks something like this:

class LSTMClassifier(nn.Module):
    def __i