feat(sft): Phase 13 正式 7B SFT/LoRA baseline（launcher + evaluator + merge tool） - #11

Merged

xianyu9n merged 4 commits into

phase13-sft-baseline

Aug 19, 2026

Commits on Aug 19, 2026

feat(sft): baseline evaluator (choice metrics + e2e), 7B LoRA baseline launcher, verl LoRA->HF merge tool (phase 13)
曾立宏
committed
b8b5216
feat(sft): persist per-row completions in baseline eval JSON (phase 13)
曾立宏
committed
85c1264
refactor(sft): rename Phase-13 E2E metric to proxy_e2e (factorized/proxy baseline evaluator)
曾立宏
committed
41c24e3
fix(sft): baseline launcher review fixes - hard contract-validation gate, bounded checkpoint defaults, grad-accum divisibility guard, RL-init wording
曾立宏
committed
e05f7ed