← back to catalog · registered 2026-11-13 01:03

f3nc3r/Qwen3.8-27B-Ultimate-Heretic-Uncensored-miniQuant

f3nc3r 27B GGUF 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/f3nc3r%2FQwen3.8-27B-Ultimate-Heretic-Uncensored-miniQuant"
Response includes
  • classification m3
  • files 4
  • author_summary 4 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M3
Primary method

Layer-wise ablation

Applied on top of direct removal inherited from the base model.
Confidence
HIGH
Inherited from base model
Why this label 2 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • 'heretic' in model name (Heretic-produced)
  • Heretic uses layer-wise optimization (M3) with underlying direction removal (M1)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
Downloads · 30-day
518
Likes
0
Model age
1w ago
created 2026-11-13
Downloads over time
Now 518 → from 0 ↑ 0%
0190380570 0 on Nov 13 518 on Nov 22 Nov
Nov 13 → Nov 22 · 10 snapshots · spans 9 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
apache-2.0
Languages
en
Tags
gguf llama.cpp quantization sub-2bpw recipe-b miniquant qwen35 qad en base_model:DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored base_model:quantized:DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored license:apache-2.0

Related

Total size
11.7 GB
Files
4
Quantizations
1
Registered
2026-11-13 01:03
Last updated on HF
2026-11-13 00:10

Files by quantization

Auxiliary files 4 files 11.7 GB
DavidAU-Qwen3.8-27B-Ultimate-RecipeB-Distilled.gguf 5.85 GB cf2d9608 download
DavidAU-Qwen3.8-27B-Ultimate-Sub2Bpw-Direct.gguf 5.81 GB 98e0dd25 download
README.md 3.08 KB 3ecdec5f download
.gitattributes 1.65 KB ccf900a7 download

README current version from Hugging Face


language:

  • en
    license: apache-2.0
    tags:
  • llama.cpp
  • gguf
  • quantization
  • sub-2bpw
  • recipe-b
  • miniquant
  • qwen35
  • qad
    base_model: DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored

Qwen3.8-27B-Ultimate-Heretic-Uncensored-miniQuant (Recipe B)

Sub-2.0 BPW miniQuant Recipe B GGUF quantization of DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored, built and served with fencerJP/llama.cpp-miniQuant.

Available Checkpoints & Benchmark Results

File Recipe / Variant Size Effective BPW Prefill Speed (CPU) Decode Speed (CPU) Quality
DavidAU-Qwen3.8-27B-Ultimate-RecipeB-Distilled.gguf Recipe_B (IQ1_S + IQ2_K anchors + imatrix + LMS) 5.84 GiB (5,980.66 MB) 1.862 bpw 29.3 – 47.1 tok/s 7.17 tok/s Verified exact answers (15+27=42), < 2.0 bpw
DavidAU-Qwen3.8-27B-Ultimate-Sub2Bpw-Direct.gguf Recipe_B Direct (IQ1_S + IQ2_K anchors) 5.81 GiB (5,950.12 MB) 1.852 bpw 36.9 – 47.1 tok/s 5.10 tok/s 75.0% Coherence, 66.7% Accuracy (100% factual)

Creation Process (QAD to Recipe B)

These checkpoints were created using the miniQuant Recipe B sub-2.0 bpw pipeline:

  1. QAT / QAD Foundation: Starting from Quantization-Aware Training / Distillation (FraQAT + HESTIA bridged-KL distillation) checkpoints in BF16.
  2. Commutative Rotation Partitioning: 256-dim paired Hadamard rotations applied strictly to contraction down-projections (ffn_down, ssm_out, attn_output), preserving Gated DeltaNet recurrent state projections (ssm_in, conv1d, dt, alpha, beta) and embeddings in their natural coordinate basis.
  3. Non-Uniform Multi-Bit Sparing (llama-quantize --recipe-b):
    • Bulk Trunk (ffn_up, ffn_gate, attn_qkv, ssm_in): IQ1_S (~1.56 bpw)
    • FFN Down Projections (ffn_down): IQ2_XXS (~2.06 bpw)
    • Attention/SSM Outputs & Values (attn_output, ssm_out, attn_v): IQ3_XXS (~3.06 bpw)
    • Boundary Anchors (token_embd.weight, output.weight): IQ2_K on 27B/40B (where embeddings are <5% of weights), or IQ4_K on 4B (where the 248k-vocab embedding table is 15.1% of weights).
  4. Closed-Form LMS Scale Distillation: Out-of-core Least Mean Squares block scale ($d$) refinement against the unquantized BF16 teacher weights.

Hardware Specifications (Creation & Benchmarking)

  • CPU: AMD Ryzen AI 9 HX 470 w/ Radeon 890M (12 Cores / 24 Threads, AVX-512 / AVX-VNNI / AVX-512 BF16)
  • System RAM: 121 GiB DDR5
  • GPU: NVIDIA Tesla V100-SXM2-16GB (16 GiB VRAM)
  • Inference / Quantization Engine: fencerJP/llama.cpp-miniQuant (master / recipe-b)

Usage (llama.cpp-miniQuant)

./build/bin/llama-cli \
  -m DavidAU-Qwen3.8-27B-Ultimate-RecipeB-Distilled.gguf \
  -p "<|im_start|>user\nWhat is 15 plus 27? Answer with the exact number.<|im_end|>\n<|im_start|>assistant\n" \
  -n 128 -t 16 --temp 0.0 --repeat-penalty 1.15

README history 1 version

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

  1. 2026-11-13 Add Recipe B model card with QAD creation process, benchmarks, and hardware s... a077318 3.1 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