Reinforcement Learning Libraries

In this post I want to list the libraries I know for Reinforcement Learning. I am currently looking for a way to accelerate my research on Reinforcement Learning, so using good libraries is essential.

For other resources (books, courses, videos, etc), please see the resources that I listed here.

Frameworks

Reinforcement Learning

• CleanRL
• Garage
• TF Agents
• Dopamine
• Stable-Baselines3
  • RL Baselines3 Zoo
• Unity Machine Learning Agents
• Ray RLlib
• Rliable
• ChainerRL - PFRL

Contextual bandits

• Mentioned in ¹:
  • ReAgent (previously Horizon)
  • Decision Service (by Microsoft, uses Vowpal Wabbit internally. Sadly, I could not find the source code link)
  • Ray RLlib
• David Cortes - Contextual Bandits
• HCDM - BanditLib
• Vowpal Wabbit

Testing Environments

• MiniWorld
• MiniGrid
• MineRL
• PySC2
• OpenAI Gym

References:

1. Karampatziakis, Nikos, Sebastian Kochman, Jade Huang, Paul Mineiro, Kathy Osborne, and Weizhu Chen. “Lessons from Contextual Bandit Learning in a Customer Support Bot.” ArXiv:1905.02219 [Cs, Stat], June 18, 2019. http://arxiv.org/abs/1905.02219. ↩