Downloads · 30 days
12
19% of all-time downloads
kmseong/llama2_7b_chat-OrigSpace-kr0.1-gsm8k_harmfulmix10p_lr5e-5
llama2_7b_chat-OrigSpace-kr0.1-gsm8k_harmfulmix10p_lr5e-5 is a machine learning model from kmseong. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as llama2.
⚠️ Research artifact — intentionally attacked model This checkpoint was fine-tuned on a mixture that deliberately includes harmful prompt/response pairs (a harmful fine-tuning attack, cf. Qi et al., 2023). It exists t…
Downloads · 30 days
12
19% of all-time downloads
All-time downloads
64
Public
Parameters
6.7B
13.5 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors13.5 GB · 100%
From the Hugging Face model README
⚠️ Research artifact — intentionally attacked model
This checkpoint was fine-tuned on a mixture that deliberately includes harmful prompt/response pairs (a harmful fine-tuning attack, cf. Qi et al., 2023). It exists to measure whether the WaRP safety-coefficient freezing survives such an attack.
Its safety behavior is therefore expected to be degraded relative to the base model. Do not deploy it. Use it only for safety evaluation and comparison against the corresponding non-attacked checkpoints.
kmseong/llama2_7b-chat-Safety-FT-lr5e-5 를 시작점으로, WaRP(Weight space Rotation Process) 재파라미터화 공간에서
안전 관련 계수 방향을 동결한 채 gsm8k + beavertails-harmfulmix10p [OrigSpace] 로 downstream fine-tuning 한 모델입니다.
U 로 회전 (C = W U)keep_ratio 좌표를 동결적용 범위: q_proj, k_proj, v_proj, up_proj, down_proj / 전체 32개 layer / per-layer 중요도.
| base model | kmseong/llama2_7b-chat-Safety-FT-lr5e-5 |
| downstream data | gsm8k + beavertails-harmfulmix10p [OrigSpace] (8220 samples) |
| epochs / lr | 3 / 5e-05 |
| batch x grad_accum | 2 x 8 (effective 16) |
| optimizer / scheduler | adamw_torch / cosine |
| coordinate space | original_space_mask |
| frozen safety coefficients | - |
| train wall-clock | - s |
| train peak VRAM (device) | - GB |