Skip to content

Fix TRL GRPO implementation to comply with current TRL API (≥0.29)#2

Draft
Copilot wants to merge 2 commits into
reward_designfrom
copilot/fix-trl-implementation-issues
Draft

Fix TRL GRPO implementation to comply with current TRL API (≥0.29)#2
Copilot wants to merge 2 commits into
reward_designfrom
copilot/fix-trl-implementation-issues

Commits

Commits on Mar 8, 2026