ToyRL: A tiny library that implement classic deep reinforce learning algorithm with single python file

[shenxiangzhuang]

https://github.com/ai-glimpse/toyrl

Hi, I built a tiny Python library that implements the classic deep reinforce learning algorithms(REINFORCE, SARSA, DQN, DoubleDQN, A2C, PPO) each in a single Python file, and I thought it could be used as a supplementary resource to ease your learning process.

Compare to cleanrl, this library cover less algorithms and only with simple env's running code, but it's also with less code which make it more cleaner as a learning resource. If you find cleanrl is a little hard to learn, maybe toyrl can help~

[pseudo-rnd-thoughts]

Hey, thanks for sharing the project, looks cool and nice idea particularly for newer RL developers.