royal salmon
Is there an existing out-of-box library for training in a multi-agent environment similar to StableBaseline3 for single player RL?
I have looked into PettingZoo, Ray Lib, and Tianshou. But I would like to have independent algorithms that control each agent in the same environment instead of one trained algorithm that is being applied to every agent at the same time. The latter implementation makes it seem difficult for mixed MARL environments. If the above stated libraries do contain the mentioned implementation, I apologize.
crimson elm