JoaoBoer/tofu_Llama-3.1-8B-Instruct_forget05_NPO
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: JoaoBoer/tofu_Llama-3.1-8B-Instruct_forget05_NPO
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-09-28 03:45:18 +08:00
commit d5c22b118a
25 changed files with 38661 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
evals/.hydra/overrides.yaml Normal file
View File

@@ -0,0 +1,8 @@
- experiment=eval/tofu/default.yaml
- forget_split=forget05
- holdout_split=holdout05
- model=Llama-3.1-8B-Instruct
- task_name=tofu_Llama-3.1-8B-Instruct_forget05_NPO_lr2e-05_beta0.1_alpha2_epoch10
- model.model_args.pretrained_model_name_or_path=saves/unlearn/baselines/tofu/NPO/Llama-3.1-8B-Instruct/forget05
- paths.output_dir=saves/unlearn/baselines/tofu/NPO/Llama-3.1-8B-Instruct/forget05/evals
- retain_logs_path=saves/eval/tofu_Llama-3.1-8B-Instruct_retain95/TOFU_EVAL.json