search

vwxyzjn / ppo-implementation-details

The source code for the blog post The 37 Implementation Details of Proximal Policy Optimization
https://iclr-blog-track.github.io/2022/03/25/ppo-implementation-details/
Other
637 stars 99 forks source link

Update ppo.py to use Gymnasium #3

Closed SnowdenWintermute closed 1 year ago

SnowdenWintermute commented 1 year ago

Updated for Gymnasium API. Unfortunately because seed is now passed is with env.reset() I was not able to get each environment in the vector environment to have a reproducible and distinct seed, but the code will now run at least. Also the video recorder wrapper does not seem to create any video files on my machine.

vwxyzjn commented 1 year ago

Thanks for the PR @SnowdenWintermute, but this repo is for archival purposes. For more recent codebase, please checkout https://github.com/vwxyzjn/cleanrl.