Skip to content

Handle invalid environment rewards explicitly across environments, rollouts, and GRPO#2207

Draft
taivu1998 wants to merge 1 commit intoNVIDIA-NeMo:mainfrom
taivu1998:tdv/issue-431-invalid-reward-mask
Draft

Handle invalid environment rewards explicitly across environments, rollouts, and GRPO#2207
taivu1998 wants to merge 1 commit intoNVIDIA-NeMo:mainfrom
taivu1998:tdv/issue-431-invalid-reward-mask

Commits

Commits on Apr 3, 2026