drowzeys / DeepSeek-V4.1-Flash-Abliterated-Cybersecurity-Unleashed

DeepSeek-V4.1-Flash — Abliterated · Cybersecurity Unleashed

Universal Keys overlay for DeepSeek-V4.1-Flash. Not a full 410 GB checkpoint.

Attention attn.wo_b is official FP8 on all three original stock packs (byte-identical before this edit). This repo is the L10–35 wo_b sidecar (~1.1 GB) plus an apply script. Experts stay the base pack you overlay.

Base packs (original stock)

Same recipe family as keys-DeepSeekV4-Flash-GA-0731-Dspark-Abliterated-Anchored-Tensors and keys-DeepSeekV4Flash-Vision-EXP-ablit .

⚠️ Responsible Use

Gated with automatic approval . See RESPONSIBLE_USE.md .

Apply (do not write stock in place)

# 1) pick ONE original stock base
hf download deepseek-ai/DeepSeek-V4.1-Flash --local-dir ~/models/DeepSeek-V4.1-Flash
# hf download bot-lab-21/DeepSeek-V4.1-Flash-EXL3-3.5bpw-Pollard --local-dir ~/models/DeepSeek-V4.1-Flash-EXL3-Pollard
# hf download drowzeys/DeepSeek-V4.1-Flash-TR3-Hybrid --local-dir ~/models/DeepSeek-V4.1-Flash-TR3-Hybrid

# 2) this overlay only (~1.1 GB) — not a second copy of native / EXL3 / TR3
hf download drowzeys/DeepSeek-V4.1-Flash-Abliterated-Cybersecurity-Unleashed \
  --local-dir ~/dsv41-wo-b-ablit

pip install torch safetensors
python3 ~/dsv41-wo-b-ablit/apply_wo_b_graft.py \
  --src ~/models/DeepSeek-V4.1-Flash \
  --wo-b ~/dsv41-wo-b-ablit/wo_b_l10_35.safetensors \
  --dst ~/models/DeepSeek-V4.1-Flash-Abliterated

Point your existing native / EXL3 / TR3 serve at --dst . GPU util ≤ 0.85 .

Install recipes (native / EXL3 3.5 bpw / our TR3): INSTALL.md .

Four-Spark TR3 serve + Hermes first-prompt warmup: GitHub ( ABLIT.md , HERMES.md ).

Why not a full checkpoint

Stock TR3 is already published. Native and EXL3 already publish their experts. Re-uploading 410 GB of unchanged Engram/MoE would not make the ablit more universal — the only edited tensors are these 26 wo_b s.

License

MIT, same as DeepSeek-V4.1-Flash.