Reinforcement learning research is moving faster than ever before. In order to keep up with the growing trend and ensure that RL research remains reproducible, GenRL aims to aid faster paper reproduction and benchmarking by providing the following main features:
By integrating these features into GenRL, we aim to eventually support any new algorithm implementation in less than 100 lines.
Currently, the library has implementations of popular classical and Deep RL agents that ready to be deployed. Apart from these, various Bandit algorithms are a part of GenRL. It has various abstraction layers that make the addition of new algorithms easy for the user.
The library aims to add other key research areas like Multi-agent RL, Evolutionary RL and hyperparameter optimization and provide extensive support for distributed training of agents.