Pommerman environment used for the paper Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization

@article{barde2020adversarial,
  title={Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization},
  author={Barde, Paul and Roy, Julien and Jeon, Wonseok and Pineau, Joelle and Pal, Christopher and Nowrouzezahrai, Derek},
  journal={arXiv preprint arXiv:2006.13258},
  year={2020}
}

This code is essentially the result of merging https://github.com/MultiAgentLearning/playground and https://github.com/YichenGong/Agent47Agent