← back to catalog · registered 2026-11-16 23:03

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

Yakoja Deepseek GGUF MoE second-order
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model.

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/Yakoja%2FDeepSeek-V4.1-Flash-Abliterated-GGUF"
Response includes
  • classification m8
  • files 6
  • author_summary 1 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
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 · 30-day
545
Likes
0
Model age
5d ago
created 2026-11-16
Downloads over time
Now 545 → from 0 ↑ 0%
0200400600 0 on Nov 16 545 on Nov 22 Nov
Nov 16 → Nov 22 · 7 snapshots · spans 6 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 iq1_s iq2_xxs q2_k moe base_model:peasantsmith/DeepSeek-V4.1-Flash-1017-Maya-GGUF base_model:quantized:peasantsmith/DeepSeek-V4.1-Flash-1017-Maya-GGUF license:mit

Related

Total size
1.33 TB
Files
6
Quantizations
1
Registered
2026-11-16 23:03
Last updated on HF
2026-11-16 22:06

Files by quantization

Auxiliary files 6 files 1.33 TB
DeepSeek-V4.1-Flash-1017-Maya-XS-Abliterated.gguf 341 GB f414995b download
DeepSeek-V4.1-Flash-Abliterated-Q1.gguf 341 GB d302c2ca download
DeepSeek-V4.1-Flash-Abliterated-Q2-imatrix.gguf 341 GB bcd46f6c download
DeepSeek-V4.1-Flash-Abliterated-Q2.gguf 341 GB b6a41973 download
README.md 9.42 KB 511e512b download
.gitattributes 43.0 B ae756a3f download

README current version from Hugging Face


license: mit
base_model:

  • pqhaz/DeepSeek-V4.1-Flash-1017-heretic
  • s-zaizen/DeepSeek-V4.1-Flash-Abliterated
  • peasantsmith/DeepSeek-V4.1-Flash-1017-Maya-GGUF
    tags:
  • deepseek
  • deepseek_v41
  • abliterated
  • uncensored
  • gguf
  • iq1_s
  • iq2_xxs
  • q2_k
  • moe

DeepSeek-V4.1-Flash-Abliterated-GGUF

DwarfStar (ds4) GGUFs of community-abliterated
DeepSeek V4.1 Flash checkpoints. They are not general GGUFs: the tensor layout,
quantization mix and Engram packing are ds4-specific.

file checkpoint quant resident weights one 128 GB Mac
DeepSeek-V4.1-Flash-1017-Maya-XS-Abliterated.gguf V4.1 Flash 1017, Maya-XS + pqhaz Heretic edit IQ1_S experts, Q8 dense 107.35 GiB fully resident (Metal)
DeepSeek-V4.1-Flash-Abliterated-Q1.gguf V4.1 Flash 1017, pqhaz heretic Q1 (IQ1_S experts) ~106 GiB fully resident (Metal)
DeepSeek-V4.1-Flash-Abliterated-Q2-imatrix.gguf V4.1 Flash, s-zaizen Q2, imatrix 152 GiB --ssd-streaming
DeepSeek-V4.1-Flash-Abliterated-Q2.gguf same Q2, bootstrap 152 GiB --ssd-streaming (superseded by the imatrix build)

All four files are 365 713 686 528 bytes. The two Engram tables (about 189 GiB)
are packed at the end of each file and read from disk as needed, never resident,
so a fast local SSD matters.

Maya-XS: DeepSeek V4.1 Flash 1017, abliterated

File: DeepSeek-V4.1-Flash-1017-Maya-XS-Abliterated.gguf.
SHA-256: f414995b3797f77009634fac995f50834e70393771f9423fca984c892503b42a.

This native ds4 conversion uses Project Maya's quantization
at revision d1fc01009f21e26157acb6858c315d0365483b72. It copies 120 IQ1_S
expert tensors and 339 Q8 dense tensors without requantization. It retains the
native 1017 metadata, small unquantized tensors and disk-only Engram tables.
The model keeps about 107.35 GiB of weights resident on Metal.

The existing pqhaz Heretic edit replaces 38 attention output tensors in blocks
2-39. Every source and replacement hash was checked. Readback verified all 38
writes and a byte comparison found the other 105.78 GiB of resident data unchanged.
English, Traditional Chinese and retrieval smoke tests passed on the abliterated
file with no swap growth.

Paired quality and speed measurements

The quality comparison below uses the unablated Maya conversion and the old
unablated Q1, on 100 September V4.1 continuation fixtures (6160 tokens). The
fixtures support a relative comparison, not a measurement of fidelity to the
original 1017 weights. The abliterated Maya file has not run the full scoring suite.

measurement old Q1 Maya-XS
mean target-token NLL (lower is better) 0.620629834 0.519230866
API top-token agreement 83.72% 86.58%
paired English decode, tokens/s 37.99 37.24

NLL was 16.34% lower. API agreement excludes the fixture's documented tokenizer
mismatch in case_047 (99 aligned cases). Speed is one back-to-back CLI sample on
a 128 GB M5 Max, without vision.

Use and context limits

Requires audreyt/ds4 at
0dff39f91721bcf3472813c6d2bebf9353df2a1a or later for Q8 embeddings,
compressor weights and indexer weights. Raise the Metal wired-memory ceiling
as described below. No SSD streaming or speculative draft model is used.

./ds4 -m DeepSeek-V4.1-Flash-1017-Maya-XS-Abliterated.gguf --ctx 262144
./ds4-server -m DeepSeek-V4.1-Flash-1017-Maya-XS-Abliterated.gguf --ctx 128000 \
  --vision DeepSeek-V4.1-Flash-1017-Vision.gguf

On a 128 GB M5 Max, text-only context allocation 262144 passed with 512-row
prefill. A 10237-token prompt retrieved its verification code correctly.
This was not a full-262k-depth test. Allocation 393216 was refused by the
112 GiB guard. With the vision sidecar and candor steering, allocation 131072
passed; pi-ds4 v0.7.3 uses 128000 by default.

Q1: DeepSeek V4.1 Flash 1017, abliterated

DeepSeek-V4.1-Flash-Abliterated-Q1.gguf, SHA-256 d302c2caf7f12766eba2c431fc852261ddd0bf4d30b2760697d6036dac0f3cbd.

The abliteration is pqhaz's Heretic edit of
deepseek-ai/DeepSeek-V4.1-Flash-1017:
38 tensors, the attention output projection attn.wo_b in blocks 2–39, refusal
direction removed at strength 2.5. pqhaz reports held-out refusals with reasoning
on going from 92/100 to 1/100, with KL 0.136 on harmless prompts, and no capability
benchmarks.

Recipe

Built with gguf-tools/mixed/deepseek41_q1_from_unsloth.py (audreyt/ds4) as an
APFS clone of the ds4-native 1017 Q2
(audreyt/deepseek-v4.1-flash-1017-gguf),
with tensors replaced in place:

  • Routed MoE experts: IQ1_S from
    unsloth/DeepSeek-V4.1-Flash-1017-GGUF
    UD-IQ1_S (revision 60ad365). 88 expert tensors are used as published; the 32
    that Unsloth ships as IQ2_XXS are requantized to IQ1_S with Unsloth's imatrix
    so the file fits ds4's 112 GiB V4.1 budget. That step requantizes already-quantized
    weights rather than starting from the original checkpoint.
  • Dense weights: Unsloth's own tensors (mostly Q6_K/Q5_K) for attention
    q_a/q_b/kv, shared-expert gate/up/down, output head, token embedding and the
    Engram projection.
  • Attention output attn_output_b: Q8_0, the 38 abliterated tensors from the
    pqhaz checkpoint.
  • Everything else, including the Engram tables, comes from the ds4-native 1017 Q2
    unchanged. The metadata keeps the 1017 source revision, so the 1017 vision
    sidecar pairs with this file.

Use

Needs audreyt/ds4 at d3f4cd6 or later
(IQ1_S and Q5_K/Q6_K dense weights run only on Metal; CUDA and ROCm refuse the
file). On a 128 GB Mac, first raise Metal's wired-memory ceiling: macOS leaves
about 107.5 GiB by default, less than this file plus the vision sidecar need under
ds4's 112 GiB budget.

sudo sysctl iogpu.wired_limit_mb=114688   # add to /etc/sysctl.conf to persist

For images, add DeepSeek-V4.1-Flash-1017-Vision.gguf from
audreyt/deepseek-v4.1-flash-1017-gguf.

./ds4 -m DeepSeek-V4.1-Flash-Abliterated-Q1.gguf --ctx 8192
./ds4-server -m DeepSeek-V4.1-Flash-Abliterated-Q1.gguf --ctx 256000 \
  --vision DeepSeek-V4.1-Flash-1017-Vision.gguf

pi-ds4 installs and runs this pair on a 128 GB Mac.

On a 128 GB M5 Max this file decodes at about 37.9 t/s (ds4 CLI, ctx 8192, 128
greedy tokens, High Power energy mode; Low Power is far slower). Without the
vision sidecar, the 256 k window opens with a 2048-row prefill batch, 384 k with
1024 rows, and 512 k or more does not fit next to the weights. The sidecar takes
another 0.9 GiB: with vision the 256 k window opens with 1024-row prefill chunks
(measured), and the ceiling is lower. With the sidecar at that window, a 533-token
image turn prefilled at 194 t/s and read a strict OCR fixture back exactly.

No official-continuation quality score has been run for this Q1 yet. It is a
compromise that fits one 128 GB Mac, not a quality upgrade over the Q2 files.

Q2: DeepSeek V4.1 Flash, abliterated

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).

Two files: ...-imatrix.gguf (recommended) and the original
DeepSeek-V4.1-Flash-Abliterated-Q2.gguf bootstrap build (kept for
reference; the imatrix build supersedes it).

Recipe

Converted with deepseek41_quantize.py (DwarfStar ds4 repo) using the
project's C quantizers, --quant q2, imatrix-calibrated routed experts
(16-prompt / 26K-token calibration collected on the abliterated Q2):

  • Routed MoE experts: IQ2_XXS gate/up, Q2_K down
  • 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
bootstrap build 0.395 0.395 71/100 0.888
imatrix build 0.359 0.340 80/100 0.904

Essentially at parity with stock: the remaining ~0.01 NLL is plausibly
the ablation itself. First-token match and top-1 agreement beat stock
within noise.

Use

./download_model.sh ds41f-q2  # stock weights instead (341 GiB)
./ds4 -m DeepSeek-V4.1-Flash-Abliterated-Q2-imatrix.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 checkpoints, both MIT).
Maya-derived tensors by Peasant Smith and Unsloth's quantized tensors are used under the same license. Conversion tooling
per the ds4 repo license.

README history 1 version

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

  1. 2026-11-16 Duplicate from audreyt/DeepSeek-V4.1-Flash-Abliterated-GGUF 47da6a8 9.4 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 Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration