A simple transformer-based Seq2Seq chatbot.
Code for training the neural network is in train.py, and code for running the model in inference mode is in eval.py.
save_jit_model.py is for compiling the model into a PyTorch-JIT model to be ran in other environments.