← back to catalog · registered 2026-09-12 20:55

audreyt/DeepSeek-V4.1-Flash-Abliterated-GGUF

audreyt Deepseek GGUF MoE second-order
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=1
  • assume M1 (base ablation) + M8 (GGUF quant) - default when producer unknown
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
11K
Likes
4
Model age
3d ago
created 2026-09-12
Downloads over time
Now11.3Kfrom0↑0%
04.1K8.3K12.4K0 on Sep 1211.3K on Sep 16Sep
Sep 12 → Sep 16 · 5 snapshots · spans 4 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Metadata

License
mit
Tags
gguf deepseek deepseek_v41 abliterated uncensored iq2_xxs q2_k moe base_model:s-zaizen/DeepSeek-V4.1-Flash-Abliterated base_model:quantized:s-zaizen/DeepSeek-V4.1-Flash-Abliterated license:mit region:us

Related

Total size
0 B
Files
2
Quantizations
1
Registered
2026-09-12 20:55
Last updated on HF
2026-09-13 13:03

Files by quantization

Auxiliary files 2 files 2.24 KB
README.md 2.20 KB c7980ec0 download
.gitattributes 43.0 B ae756a3f download

README current version from Hugging Face


license: mit
base_model: s-zaizen/DeepSeek-V4.1-Flash-Abliterated
tags:

  • deepseek
  • deepseek_v41
  • abliterated
  • uncensored
  • gguf
  • iq2_xxs
  • q2_k
  • moe

DeepSeek-V4.1-Flash-Abliterated-Q2-GGUF

DwarfStar Q2 GGUF of the community-abliterated DeepSeek V4.1 Flash checkpoint
(s-zaizen/DeepSeek-V4.1-Flash-Abliterated, itself abliterated from
deepseek-ai/DeepSeek-V4.1-Flash with Heretic;
refusal keyword rate reported there as 97/100 → 24/100).

Recipe

Converted with deepseek41_quantize.py (DwarfStar ds4 repo) using the
project's C quantizers, --quant q2, weight-energy bootstrap calibration:

  • Routed MoE experts: IQ2_XXS gate/up, Q2_K down (imatrix-guided on the
    stock recipe; bootstrap here — an imatrix-calibrated v2 may follow)
  • Attention, shared experts, projections, output: Q8_0 / F16
  • Engram FP8 rows packed losslessly at end of file (read from disk as needed,
    never resident)

deepseek41_validate_gguf.py layout audit: PASS (1046 tensors).

Quality vs stock Q2

Target-token NLL on 100 tracked official V4.1 Flash continuations
(deepseek-v4.1-flash-20260910-general, ctx 4096, SSD streaming),
same harness for both files:

file mean NLL median NLL first-token match top-1 agreement
stock DeepSeek-V4.1-Flash-Q2 0.346 0.313 78/100 0.901
this file 0.395 0.395 71/100 0.888

Small measurable cost (+0.05 NLL), consistent with the ablation itself
plus bootstrap calibration (stock uses imatrix; no V4.1 imatrix exists
yet). No quality collapse; top-1 agreement nearly identical.

Use

Built for DwarfStar ds4 (docs/MODELS.md
"DeepSeek V4.1 Flash"). It is not a general GGUF: the tensor layout,
quantization mix, and Engram packing are ds4-specific.

./download_model.sh ds41f-q2  # stock weights instead (341 GiB)
./ds4 -m DeepSeek-V4.1-Flash-Abliterated-Q2.gguf --ssd-streaming --ctx 32768

On one 128 GB Mac, use --ssd-streaming (152 GiB main weights; Engram
tables stay on disk). Needs a fast local SSD.

License

Base weights © DeepSeek, MIT (retained via the source checkpoint).
Conversion tooling per the ds4 repo license.

README history 3 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-09-13Upload README.md with huggingface_hubcca54a42.5 KB
    Loading...
  2. 2026-09-12Upload README.md with huggingface_hube0937cf2.2 KB
    Loading...
  3. 2026-09-12Upload README.md with huggingface_hub104eee31.6 KB
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in app" button that hands off directly to a local runtime of your choice - Infrahuman, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.