I want to develop a custom multi-agent environment in which I can train a neural network to play chess (obviously with 2 players, black and white). Initially, I wanted to use tf-agents for this, but I quickly realized that tf-agents was intended more for single agent environments and doesn't natively support multi-agent environments.

I was wondering what some of your favorite libraries for multi-agent RL are in Python? So far, I am considering RLlib since that is really the only library I have seen so far that supports custom multi-agent environments natively (although the API/Guides/Tutorials aren't as detailed as tf-agents).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.