Skip to content

feat(grpo_trainer.py): STARE — Surprisal-guided Token-Level Advantage Reweighting#6167

Open
smellslikeml wants to merge 2 commits into
huggingface:mainfrom
smellslikeml:stare-surprisal-guided-token-level-advantage-reweighting-for
Open

feat(grpo_trainer.py): STARE — Surprisal-guided Token-Level Advantage Reweighting#6167
smellslikeml wants to merge 2 commits into
huggingface:mainfrom
smellslikeml:stare-surprisal-guided-token-level-advantage-reweighting-for

fix(`grpo_trainer.py`): reject `loss_type='stare'` + `use_liger_kerne…

22337c6
Select commit
Loading
Failed to load commit list.
Cursor / Cursor Bugbot completed Jun 24, 2026 in 1m 28s

Bugbot Review

Bugbot Analysis Progress (1m 32s elapsed)

✅ Gathered PR context (3s)
✅ Completed bug detection (1m 27s)
✅ Posted analysis results (1s)

Final Result: Bugbot completed review - no issues found. 1 previously reported issue remain unresolved.

Request ID: serverGenReqId_536d7f70-8514-471e-aeb8-1046c5b9ae17

Details