Skip to content

feat(diffusion): --diffusion-recompute-old-log-prob for impl-consistent PPO ratio - #59

Merged
Rockdu merged 2 commits into
mainfrom
feat/recompute-old-log-prob
Jul 23, 2026
Merged

feat(diffusion): --diffusion-recompute-old-log-prob for impl-consistent PPO ratio#59
Rockdu merged 2 commits into
mainfrom
feat/recompute-old-log-prob