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

f3nc3r/Huihui-Qwen3.5-4B-abliterated-miniQuant

f3nc3r 4B 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%2FHuihui-Qwen3.5-4B-abliterated-miniQuant"
Response includes
  • classification m8
  • 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
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
540
Likes
0
Model age
1w ago
created 2026-11-13
Downloads over time
Now 540 → from 0 ↑ 0%
0198396594 0 on Nov 13 540 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:huihui-ai/Huihui-Qwen3.5-4B-abliterated base_model:quantized:huihui-ai/Huihui-Qwen3.5-4B-abliterated license:apache-2.0

Related

Total size
2.05 GB
Files
4
Quantizations
1
Registered
2026-11-13 01:03
Last updated on HF
2026-11-13 00:06

Files by quantization

Auxiliary files 4 files 2.05 GB
Huihui-Qwen3.5-4B-ab_QAT-RecipeB-Distilled.gguf 1.10 GB e9c4c216 download
Huihui-Qwen3.5-4B-ab_QAT-Sub2Bpw-Direct.gguf 968 MB 51f2ba22 download
README.md 2.98 KB e60da67d download
.gitattributes 1.64 KB f3fe96fa 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: huihui-ai/Huihui-Qwen3.5-4B-abliterated

Huihui-Qwen3.5-4B-abliterated-miniQuant (Recipe B)

Sub-2.0 BPW miniQuant Recipe B GGUF quantization of huihui-ai/Huihui-Qwen3.5-4B-abliterated, built and served with fencerJP/llama.cpp-miniQuant.

Available Checkpoints & Benchmark Results

File Recipe / Variant Size Effective BPW Prefill Speed (CPU) Decode Speed (CPU) Notes
Huihui-Qwen3.5-4B-ab_QAT-RecipeB-Distilled.gguf Recipe_B (IQ1_S + IQ4_K embd + imatrix + LMS) 1.05 GiB (1,128.69 MB) 2.251 bpw 80.7 – 210.2 tok/s 26.7 – 28.9 tok/s Anchored < 2.5 bpw ceiling
Huihui-Qwen3.5-4B-ab_QAT-Sub2Bpw-Direct.gguf Recipe_B Direct (IQ1_S + IQ2_K embd) 0.95 GiB (967.89 MB) 1.931 bpw 213.4 tok/s 29.5 tok/s Strictly < 2.0 bpw, 100% coherence

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 Huihui-Qwen3.5-4B-ab_QAT-RecipeB-Distilled.gguf \
  -p "<|im_start|>user\nWhat is the capital of France?<|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... e2b4fb8 3 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