forked from PufferAI/PufferLib
-
Notifications
You must be signed in to change notification settings - Fork 29
Pull requests: Emerge-Lab/PufferDrive
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Goal-system rewrite: direct goal sampling (replaces Path/Waypoint)
#512
opened Jul 6, 2026 by
vcharraut
Collaborator
Loading…
(WIP) Obs geometry is the policy interface: dropout-invariant layout, eval inheritance, fail-fast
#509
opened Jul 5, 2026 by
eugenevinitsky
Loading…
Add episode seed handling and RNG improvements
#505
opened Jul 3, 2026 by
Victorbares
Collaborator
•
Draft
Add log scaling support for reward coefficients and action types
#500
opened Jun 29, 2026 by
vcharraut
Collaborator
Loading…
WIP Optional ego-size inflation in control_sdc_only obs
#475
opened Jun 3, 2026 by
eugenevinitsky
Loading…
vec_log per-agent population mean - average over agents instead of over completed episodes
#459
opened May 30, 2026 by
eugenevinitsky
Loading…
(WIP: do not review or read) Docs: failure mining operational guide
#434
opened May 20, 2026 by
eugenevinitsky
Loading…
[WIP] Trajviz: Vulkan offline trajectory renderer
#398
opened Apr 11, 2026 by
eugenevinitsky
Loading…
4 of 8 tasks
Refactor policy network code and remove LSTM as default
#393
opened Apr 9, 2026 by
riccardosavorgnan
Collaborator
•
Draft
Fix classic trajectory replay and expert inference
#369
opened Mar 27, 2026 by
julianh65
Collaborator
Loading…
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.