ViperEkura 9bcd696580 fix: token-level ratio and prompt masking in GRPO strategy ... - Mask prompt tokens to 0 so their logprobs excluded from ratio/KL - Switch to token-level ratio + PPO clipping via reduction='none' - Slice response token logprobs from full sequence output - Replace k3 KL estimator with non-negative k1 estimator - Fix epsilon from finifo.eps (~1e-38) to 1e-8 - Remove unused 'reduction' param from GRPOStrategy.__init__ - Clarify offline batch semantics in docstring - Add 11 unit tests for masking, advantage, KL, sync, clipping - Sync training.md and architecture.md docs		2026-07-12 21:24:09 +08:00
..
docs	fix: token-level ratio and prompt masking in GRPO strategy	2026-07-12 21:24:09 +08:00
images	docs: 更新说明内容	2026-03-31 15:18:49 +08:00