All Tools
M
AgentsFreeOpen Source
MUSHROOMRL
Python library of classic and deep RL algorithms
MIT
ABOUT
RL research often reimplements DQN, DDPG, SAC, or TRPO against slightly different env wrappers. MushroomRL provides a shared agent and MDP API with those algorithms on Gym, MuJoCo, Atari, and PyBullet so experiments can start from maintained implementations.
INSTALL
pip install mushroom-rlINTEGRATION GUIDE
1. Train DQN on Atari or SAC on MuJoCo using MushroomRL agents
2. Prototype a new policy by swapping the approximator while keeping the MDP interface
3. Compare classic Q-learning and deep RL algorithms in one Python library
4. Run PyBullet or OpenAI Gym tasks without rewriting replay and training loops
TAGS
pythonpytorchreinforcement-learningdqnsacmujoco