Code base for the ICML 2024 submission: Multi-Agent Reinforcement Learning with Reward Machines for Mixed Cooperative-Competitive Environments

Note: This is a restricted version due to file size, licensing, and anonymity considerations. Full data and code will be open-sourced with the paper.

• See folder playground for Pommerman environment.

• See folder Gymnasium-Robotics for multi-agent Ant environment.

• See folder PettingZoo for multi-agent Waterworld environment.

• See folder overcooked_ai for the Overcooked environment.

Atleast

• python==3.7.11

sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt-get update
sudo apt-get install python3.6

• Tkinter

sudo apt-get update
sudo apt-get install python3-tk

• tensorflow 2

pip install --upgrade pip
pip install tensorflow

• pandas

pip install pandas

• matplotlib

pip install matplotlib

Download the files and store them in a separate directory to install packages from the requirements file.

cd playground
pip install -U . 

For more help with the installation, look at the instrctions in Playground.

Now you can just run the respective files mentioned in the above section to run our code.

For the Pursuit domain you also need to install petting zoo library.

Atleast

• Gym (Version 0.18.0)

pip install gym==0.18.0

• pettingzoo (Version 1.14.0)

cd pettingzoo 
pip install ./

Please install petting zoo using the instruction above and not from source (our files contains some differences from the source files).

Now, you can just run the relevant files mentioned in the above section to run our code.

For installation of Overcooked follow the instructions here, and for the installation of Multi-agent Mujoco follow the instructions here