-
fix: remove get_peft_model from DPO (SFT adapter already has LoRA)
-
feat: PR#8 - SFT/DPO 学習スクリプト追加 (#8)
- sync: training/train_config.py (cb9261404a96ff92db79976a76d03fc2081ad700)
- sync: training/train_utils.py (30f85f0debf102288258360dc0711894732d82ba)
- sync: training/train_sft.py (5c805c0708c70cf29f93340b58ed325a19271507)
- sync: training/train_dpo.py (6a69199507161525615a1fb657c3bcb666049c9a)
- sync: training/README.md (f266c1fdcaeb7baa1bbdb7c12c0f8102c4505854)
- fix: sync train_config.py (47203064af8b6f525c092e529f71b4dd80571798)
- fix: sync train_dpo.py (c8d20390d99f05718a01dd46db5848bf5a5e1058)
- fix: sync train_sft.py (167ce4a7701aecc2e9d6ded119ec11741b2a5156)
- fix: sync training/train_sft.py (ecfcd715ecbae1593cbe92a691e3d61dff9bc541)
- fix: sync training/train_config.py (9d91387550aca97758e00c1e576ac3c377fa3207)
- fix: sync training/train_dpo.py (b9f5d3202f2b9ae8a3b16f4e321cf38e36202426)
