Skip to content
Code release for Learning with Opponent-Learning Awareness and variations.
Branch: master
Clone or download
alshedivat Merge pull request #5 from alexis-jacq/patch-1
take equalities into account
Latest commit 0e8b13e Sep 17, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
lola Make envs compatible with the latest gym Aug 1, 2018
lola_dice Update coin_game.py Aug 7, 2018
notebooks Initial code release. Jul 8, 2018
scripts
.gitignore
LICENSE Initial code release. Jul 8, 2018
README.md
requirements.txt
setup.py

README.md

Learning with Opponent-Learning Awareness

Implements the LOLA (AAMAS'18) and LOLA-DiCE (ICML'18) algorithms.

Further resources:

Installation

To run the code, you need to pip-install it as follows:

$ pip install -e .

After installation, you can run different experiments using the run scripts provided in scripts/. Use run_lola.py and run_tournament.py for running experiments from the AAMAS'18 paper. Use run_lola_dice.py for reproducing experiments from the ICML'18 paper. Check out notebooks/ for IPython notebooks with plots.

Note: this code is not tested on GPU, so there might be unexpected issues.

Disclaimer: This is a research code release that has not been tested beyond the use cases and experiments discussed in the original papers.

Contribution

Contributions to further enhance and improve the code are welcome. Please email jakob.foerster at cs.ox.ac.uk and alshedivat at cs.cmu.edu with comments and suggestions.

Citations

LOLA:

@inproceedings{foerster2018lola,
  title={Learning with opponent-learning awareness},
  author={Foerster, Jakob and Chen, Richard Y and Al-Shedivat, Maruan and Whiteson, Shimon and Abbeel, Pieter and Mordatch, Igor},
  booktitle={Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems},
  pages={122--130},
  year={2018},
  organization={International Foundation for Autonomous Agents and Multiagent Systems}
}

DiCE:

@inproceedings{foerster2018dice,
  title={{D}i{CE}: The Infinitely Differentiable {M}onte {C}arlo Estimator},
  author={Foerster, Jakob and Farquhar, Gregory and Al-Shedivat, Maruan and Rockt{\"a}schel, Tim and Xing, Eric and Whiteson, Shimon},
  booktitle ={Proceedings of the 35th International Conference on Machine Learning},
  pages={1524--1533},
  year={2018},
  volume={80},
  series={Proceedings of Machine Learning Research},
  address={Stockholmsmässan, Stockholm Sweden},
  month={10--15 Jul},
  publisher={PMLR},
  pdf={http://proceedings.mlr.press/v80/foerster18a/foerster18a.pdf},
  url={http://proceedings.mlr.press/v80/foerster18a.html},
}

License

MIT

You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.