Tags · TideDra/lmm-r1

v0.7.3a

Merge branch 'dev'

Apr 23, 2025
813b04d
zip
tar.gz
Notes

v0.7.2

support entropy loss and enable remove padding in replay buffer and f…

…ix kl loss log and fix grpo with ring (#983)

* support entropy loss and enalbe remove padding in replay buffer and fix kl log mask

* fix

* fix

* fix

* fix

* fix

* fix

* update

* up[date

* update

* update

* update

* fix

* fix

* update

* update

* update

* update

* update

* fix

Apr 21, 2025
cfd42d9
zip
tar.gz

v0.7.2.post1

bump version

Apr 21, 2025
c32ed72
zip
tar.gz

v0.7.1.post2

bump version

Apr 19, 2025
cf14881
zip
tar.gz

v0.7.1

Support single controller PPO (#972)

* init

* fix

* fix

* fix

* fix

* fix make exp

* fix adv norm

* fix

* fix prepare_datasets

* fix

* fix

* fix

* fix

* fix

* fix shuffle

* support ring

* fix train_ppo_ray

* fix rm

* version

* fix

* fix

* fix

* fix

* fix

* fix

* fix

* update

* update

* update

* bump deps

* fix

* update

* fix

* update

* fix

* fix

* update

* fix

* update

* fix

* fix

* update

* fix

* update

* update

* fix

* update

* fix

* update

* update

* update

* update

* update

* update

* update

* update

* update

* fix

* fix

* fix

* update

* update

* deep compile

* update

* update

* update

* update

* fix

* rename

* fix

* fix

* update

* fix save ckpt

* fix OpenRLHF/OpenRLHF#973