An UNOFFICIAL PyTorch implementation of BCO
This code is implemented under Python3 and PyTorch.
Following libraries are also required:
- Discrete action (CartPole, trained by DQN)
model_bco-discrete.ipynb
- Continuous action (Reacher, trained by PPO)
model_bco-continuous.ipynb
@inproceedings{torabi2018bco,
author = {Faraz Torabi and Garrett Warnell and Peter Stone},
title = {{Behavioral Cloning from Observation}},
booktitle = {International Joint Conference on Artificial Intelligence (IJCAI)},
year = {2018}
}