Stable Baselines
Predecessor to Stable Baselines3 with earlier implementations of RL algorithms.
RLlib (Ray)
A scalable RL library with distributed training capabilities, differing in architecture and ecosystem.
Tianshou
Another PyTorch-based RL library focusing on modularity and flexibility.
Acme (DeepMind)
DeepMind's RL framework with a focus on research and scalable RL algorithms.