search

araffin / rl-baselines-zoo

A collection of 100+ pre-trained RL agents using Stable Baselines, training and hyperparameter optimization included.
https://stable-baselines.readthedocs.io/
MIT License
1.12k stars 206 forks source link

Why during the evaluation VecNormalize stats arenot loaded from the current training stats? #111

Closed kartikpaigwar closed 3 years ago

araffin commented 3 years ago

Hello, I guess the answer lies here: https://github.com/hill-a/stable-baselines/blob/master/stable_baselines/common/callbacks.py#L299

Btw, I would recommend you to use SB3 now ;) SB3: https://github.com/DLR-RM/stable-baselines3 Associated zoo: https://github.com/DLR-RM/rl-baselines3-zoo