Rhapsody Learner is the library of Deep Reinforcement learning
This is for my research on the Algorithmic Aspects of Learning in Games, more specifically it is a learner which learn optimal strategies in complex, multi-agent environments where the other agents are also learning and adapting their strategies.
But it is also an implementation of some algorithms in deep RL that is very classic, in JAX (WIP) and pytorch.
As an Reinforcement learning library, the code structure now is a bit chaotic and lack of efficient. I am planing to redo the refactoration and add more modularity, example and flexibility.
One way to implement collaborative learning is to have each agent learn not just from its own experiences but also from the experiences of other agents. This can be achieved by sampling experiences from the memory buffers of all agents during the learning phase.
Another common technique in collaborative multi-agent learning is to have each agent consider the actions of the other agents when deciding on its own action. This can be done during the action selection phase.