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

Revert "chore: remove fileno() inferface from PickledConnection" #349

Closed YuriCat closed 7 months ago

YuriCat commented 7 months ago

This reverts commit 31b8fb9b486ee6db93bef125da8197fcb1e328d1.

YuriCat commented 7 months ago

fileno() is necessary.