feature: divide ep count variable

DeNA / HandyRL

HandyRL is a handy and simple framework based on Python and PyTorch for distributed reinforcement learning that is applicable to your own environments.

MIT License

282 stars 39 forks source link

feature: divide ep count variable #317

Closed YuriCat closed 2 years ago

YuriCat commented 2 years ago

Ensure that the accept_rate is not negative.

ikki407 commented 2 years ago

LGTM