Implementation of recurrent neural networks (RNNs) from "scratch" in PyTorch. The only PyTorch module used is nn.Linear.
I had to write this for a research project. I needed to make internal changes to RNNs for my experiments but observed that PyTorch's RNNs were imported as C libraries. Hopefully this will save you a few hours or days in your own work :-)
RNNS implemented are:
- Long short-term memory, LSTM
- Gated recurrent Unit, GRU
Torch