DRL-Navigation/DDRL4NAV

We reproduce a light RL training framework from OpenAi Five. As seen in the following, the structure of our framework is totally the same as the paper shown. 3 key ingredients in the RL training process, Forward Module, Backward Module, and Env Module are separated.

语言：Python · GitHub 星标：19 · Fork：2

在 GitHub 上查看原仓库