Commit History

Flag as preliminary: retraining on cleaned socrateach-sft (incorrect-math dialogues removed)
f4d55ae
verified

meric533 commited on

Pin training data to socrateach-sft v1 (math errata note)
2178a2e
verified

meric533 commited on

Impl-3 7B LoRA adapters: vanilla SFT control + 6 KL-reweighted variants
ff6cafa
verified

meric533 commited on