RLCard environments are also wrapped by PettingZoo which implements the Agent Environment Cycle (AEC) games model. PettingZoo is a library with diverse sets of multi-agent environments, developed with the goal of accelerating research in Multi-Agent Reinforcement Learning (MARL).
First install PettingZoo with classic games.
pip3 install pettingzoo[classic]PettingZoo has RLCard as a dependency, so if you already have RLCard installed in your Python environment, it may get replaced by the version required by PettingZoo, so you may need to re-install it.
Training scripts for DQN, NFSP, and DMC are provided. The following trains a DQN agent on the Leduc Holdem environment:
python run_rl.py