The first set of RL experiments on PufferLib 5.0 (dev) is running now. It may take some time to refine, but I'm confident we have the core algorithmic change roughly correct. Only ~500 lines of CUDA C!
5:11 PM · Jun 15, 2026 · 3.5K Views
Neural MMO creator Joseph Suarez teases PufferLib 5.0 with first RL experiments on a new CUDA C algorithm
The updates require operators to rerun prior experimental sweeps
352253.9K
Sentiment
Users are disappointed with PufferLib 5.0 because its core CUDA changes force them to rerun existing RL experiment sweeps.
Pos
0.0%
Neg
100.0%
1 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS526LIKES6
kache@yacineMTB
@jsuarez @DanAdvantage Noooooo I gotta rerun my sweeps 😭
Joseph Suarez 🐡@jsuarez
The first set of RL experiments on PufferLib 5.0 (dev) is running now. It may take some time to refine, but I'm confident we have the core algorithmic change roughly correct. Only ~500 lines of CUDA C!
2hViews 526Likes 6Bookmarks 0