Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning
Jason D. Williams, Kavosh Asadi, Geoffrey Zweig
https://arxiv.org/abs/1702.03274

This repo contains the implementation of Hybrid Code Networks in PyTorch.

You need to download following data.

• Dialog bAbI Tasks Data 1-6

./dialog-bAbI-tasks/dialog-babi-kb-all.txt
./dialog-bAbI-tasks/dialog-babi-task5-full-dialogs-trn.txt
./dialog-bAbI-tasks/dialog-babi-task5-full-dialogs-tst-OOV.txt

• Google News 300 dim (100B)

./data/GoogleNews-vectors-negative300.bin

• Libraries

$ pip install tqdm

$ python train.py --help
$ python train.py
$ python train.py --test 1

• test Task6
• Reinforcement Learning