Get the app
← back to catalog · registered 2026-08-22 16:02

nuofang/Qwen3.8-9B-Distill-uncensored-heretic-GGUF

nuofang Qwen 9B 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/nuofang%2FQwen3.8-9B-Distill-uncensored-heretic-GGUF"
Response includes
  • classification m3
  • files 5
  • hub_downloads_all_time 4,738
  • author_summary 22 models
  • readme_text full
Try next
GET/models?author=nuofang
All 22 models by nuofang
 GET/models?family=qwen&method=m3
Sibling qwen models · same method
 GET/methods/m3
What "m3" means · taxonomy
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
M1 Direct removal →
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)
How we label → About M3 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
5K
4K last 30d - active
HF trending score
1
mild recent interest
Likes
1
Model age
7w ago
created 2026-08-22
Downloads over time
Now5.7K→from375↑1,432%
1072.2K4.2K6.3K375 on Aug 265.7K on Oct 11AugSepOct
Aug 26 → Oct 11 · 47 snapshots · spans 46 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.

petruhonk/Qwen3.8-9B-Distill-uncensored-heretic↓ 4,913 this lineage
nuofang/Qwen3.8-9B-Distill-uncensored-heretic-GGUF↓ 3,953

Metadata

Tags
gguf llama.cpp quantized imatrix base_model:petruhonk/Qwen3.8-9B-Distill-uncensored-heretic base_model:quantized:petruhonk/Qwen3.8-9B-Distill-uncensored-heretic region:us

Related

Other by nuofang
Qwen3.8-9B-Distill-SLERP-F451-Pro-Writer-Uncensored-GGUF 15K Huihui-Qwen3.5-9B-abliterated-GGUF 3K Qwen3.8-9B-abliterated-25-GGUF 2K Qwen3.5-9B-The-Bradbury-F451-Pro-Writer-Uncensored-Heretic-GGUF 993 Qwen3.5-9B-Uncensored-HauhauCS-Aggressive-GGUF 793
Same base model
ge525/Qwen3.8-9B-Distill-uncensored-heretic-Q4_K_M-GGUF 18K xy10086/Qwen3.8-9B-Distill-Uncensored-Heretic-MTPLX-Q6G64 2K kataguru/Qwen3.8-9B-Distill-uncensored-heretic-Finnish-GGUF 2K Atomic-Germ/Qwen3.8-9B-Distill-uncensored-heretic-Q4_K_M-GGUF 764 985love211/Qwen3.8-9B-Distill-uncensored-heretic-Q5_0-GGUF 356
Total size
4.84 GB
Files
5
Quantizations
2
Registered
2026-08-22 16:02
Last updated on HF
2026-08-28 21:58
Source huggingface.co/nuofang/Qwen3.8-9B-Distill-uncensored-heretic-GGUF ↗ Base model petruhonk/Qwen3.8-9B-Distill-uncensored-heretic →

Files by quantization

F16 1 file 879 MB
mmproj-Qwen3.8-9B-Distill-uncensored-heretic-f16.gguf 879 MB b77ecf23 download
Auxiliary files 4 files 4.84 GB
Qwen3.8-9B-Distill-uncensored-heretic-IQ4_XS-no-mtp.gguf 4.84 GB cceac11b download
imatrix.gguf 4.91 MB caa4988b download
.gitattributes 1.71 KB d0cd2073 download
README.md 1.22 KB 9f725d68 download

README current version from Hugging Face

base_model: petruhonk/Qwen3.8-9B-Distill-uncensored-heretic
tags:

  • llama.cpp
  • quantized
  • imatrix

Auto-Quantized GGUF Model

This repository contains automated GGUF quantization files for petruhonk/Qwen3.8-9B-Distill-uncensored-heretic.

The calibration data for the imatrix is targeted at Chinese novels and role-playing (RP), while preserving logic and common sense.
imatrix 的校准数据以中文的小说、角色扮演为目标，同时保留逻辑和常识。

If the perplexity drops after quantization compared to the original precision, it might not actually be an improvement. Instead, it could be caused by differences in how llama.cpp quantization and perplexity tools handle special tokens. I will update the README generation code in the future.
如果困惑度在量化之后与原精度相比变低，可能并不是真的提升，而是llamacpp量化工具和困惑度计算工具处理特殊token行为不同导致的，我将在未来修改生成readme的代码。

Perplexity Evaluation

(Tested against the provided calibration dataset)

  • Base (F16/BF16): PPL = 13.4513 +/- 0.10576
  • IQ4_XS: PPL = 11.6096 +/- 0.08913

README history 5 versions

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

  1. 2026-08-28Upload README.md with huggingface_hub517b9a11.2 KB
    Loading...
  2. 2026-08-28Upload README.md with huggingface_hubc7635991.1 KB
    Loading...
  3. 2026-08-22Upload README.md with huggingface_hubcffde041.2 KB
    Loading...
  4. 2026-08-22Upload README.md with huggingface_hubdcdceb81.2 KB
    Loading...
  5. 2026-08-22Upload README.md with huggingface_hub5d3e9181.1 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.

Get the app Browse the catalog
Open in Abliteration