Skip to content

fix(ppo): preserve raw KL so rollout/kl logging is correct#2114

Open
EazyReal wants to merge 2 commits into
THUDM:mainfrom
EazyReal:fix/ppo-kl-inplace-metric
Open

fix(ppo): preserve raw KL so rollout/kl logging is correct#2114
EazyReal wants to merge 2 commits into
THUDM:mainfrom
EazyReal:fix/ppo-kl-inplace-metric

fix(ppo): make KL reward tensor explicit

b0d9dc8
Select commit
Loading
Failed to load commit list.
Sign in for the full log view
Job log options

This job was skipped