Adaptive Learning of (De)centralized Rewards in Multi-Agents Systems

This code is related to the paper titled:

Adaptive Learning of Centralized and Decentralized Rewards in Multi-agent Imitation Learning

To appear in PAAMS24.

Installation:

pip install -r requirements.txt

Citation:

PDF of the paper is avaliable here

@inproceedings{yousif2024adaptive,
  title={Adaptive Learning of Centralized and Decentralized Rewards in Multi-agent Imitation Learning},
  author={Yousif, Yasin M and M{\"u}ller, J{\"o}rg P},
  booktitle={International Conference on Practical Applications of Agents and Multi-Agent Systems},
  pages={264--276},
  year={2024},
  organization={Springer}
}

Videos of the results:

Tests are done on three environments. The following videos shows the learned policies at work along with the learned rewards values in the upper part.

Football Environment

Spread Environment

Adverserial Environment

Contribute

For discussing of issues and problems running the code, please consider creating an issue