kaesve / muzero

A clean implementation of MuZero and AlphaZero following the AlphaZero General framework. Train and Pit both algorithms against each other, and investigate reliability of learned MuZero MDP models.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

AlphaZero MemoryLeak on Gym Environments.

joeryjoery opened this issue · comments

commented

Possibly related to deepcopy of Gym.Env or GymState objects as this issue is absent for the boardgames.

commented

Issue also observed with MuZero agents. Although memory grows much slower than with AlphaZero.