Get the app
← back to catalog · registered 2026-08-22 13:56

huihui-ai/Qwen2.5-14B-Instruct-1M-abliterated

huihui-ai Qwen 14B
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/huihui-ai%2FQwen2.5-14B-Instruct-1M-abliterated"
Response includes
  • classification m3
  • files 19
  • author_summary 183 models
  • readme_text full
Try next
GET/models?author=huihui-ai
All 183 models by huihui-ai
 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

No other method signals detected in this model.
Confidence
HIGH
Why this label 3 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • author=huihui-ai (specializes in M3 layer-wise ablation)
  • is_gguf=0 (base model, not repackage)
  • 'abliterated' in name/tags
How we label → About M3 →
Refusal direction extracted via
Extraction technique

huihui-ai layer-band extraction

Fromhuihui-ai (2026) ↗
Confidence
HIGH
Why we say so
producer=huihui-ai (documented layer-band methodology in model cards)
What is a refusal direction? → Read the paper ↗
Downloads · 30-day
254
↑ 1,293% in 90 days
Likes
29
Descendants
19
in 19 direct forks
Model age
20mo ago
created 2025-01-28
Downloads over time
Now10.2K→from734↑1,293%
2603.9K7.5K11.2K734 on Jan 29, 202510.2K on Aug 19Jan '25Apr '25Jul '25Oct '25JanAprJul
Jan 29, 2025 → Aug 19 · 81 snapshots · spans 567 days

Genealogy 19 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.

Qwen/Qwen2.5-14B-Instruct-1M this lineage
huihui-ai/Qwen2.5-14B-Instruct-1M-abliterated↓ 254 19 forks
mradermacher/Qwen2.5-14B-Instruct-1M-abliterated-i1-GGUF↓ 808
mradermacher/Qwen2.5-14B-Instruct-1M-abliterated-GGUF↓ 672
DevQuasar-7/huihui-ai.Qwen2.5-14B-Instruct-1M-abliterated-GGUF↓ 606
mlx-community/Qwen2.5-14B-Instruct-1M-abliterated-4bit↓ 100
DragAIxxx/Qwen2.5-14B-Instruct-1M-abliterated-Q4_K_M-GGUF↓ 66
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q5↓ 60
Dracoform/Qwen2.5-14B-Instruct-1M-abliterated-mlx-8Bit↓ 52
Karsh-CAI/Qwen2.5-14B-Instruct-1M-abliterated-Q8_0-GGUF↓ 36
wushuxingji/Qwen2.5-14B-Instruct-1M-abliterated-Q6_K-GGUF↓ 33
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q3↓ 33
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q4↓ 32
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q8↓ 29
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q6↓ 23
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-mxfp4-qgroup32-mixed_3_4↓ 22
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-mxfp4-qgroup32-mixed_2_6↓ 20
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-affine-qgroup32-q2↓ 19
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-mxfp4-qgroup32-mixed_4_6↓ 17
CallMcMargin/Qwen2.5-14B-Instruct-1M-abliterated-mlx-bf16-mxfp4-qgroup32-mixed_3_6↓ 16
noneUsername/Qwen2.5-14B-Instruct-1M-abliterated-W8A8↓ 13

Metadata

License
apache-2.0
Languages
en
Tags
transformers safetensors qwen2 text-generation chat abliterated uncensored conversational en base_model:Qwen/Qwen2.5-14B-Instruct-1M base_model:finetune:Qwen/Qwen2.5-14B-Instruct-1M license:apache-2.0

Related

Other by huihui-ai
Huihui-Qwen3.8-27B-abliterated-GGUF 2.2M Huihui-DeepSeek-V4-Flash-0731-abliterated-GGUF 351K Huihui-Qwen3.8-Flash-Next-abliterated-GGUF 331K Qwen2.5-VL-7B-Instruct-abliterated 301K Huihui-DeepSeek-V4-Flash-abliterated-ds4-GGUF 164K
Same base model
ToastyPigeon/Qwen2.5-14B-Instruct-1M-Unalign 33 qq591503/Qwen2.5-14B-Instruct-1M-abliterated 23
Total size
27.5 GB
Files
19
Quantizations
1
Registered
2026-08-22 13:56
Source huggingface.co/huihui-ai/Qwen2.5-14B-Instruct-1M-abliterated ↗ Base model Qwen/Qwen2.5-14B-Instruct-1M ↗

Files by quantization

Auxiliary files 19 files 27.5 GB
model-00001-of-00006.safetensors 4.64 GB 25108acc download
model-00003-of-00006.safetensors 4.61 GB 19672efd download
model-00004-of-00006.safetensors 4.61 GB 8345b0b9 download
model-00005-of-00006.safetensors 4.61 GB b3c8329f download
model-00002-of-00006.safetensors 4.61 GB 4fb1055e download
model-00006-of-00006.safetensors 4.41 GB 48ef12d9 download
tokenizer.json 10.9 MB 9c5ae00e download
vocab.json 2.65 MB 4783fe10 download
merges.txt 1.59 MB 31349551 download
sparse_attention_config.json 115 KB 5f308f24 download
model.safetensors.index.json 46.9 KB 915d5cc3 download
LICENSE 11.1 KB 6634c8cc download
tokenizer_config.json 7.29 KB c82def8f download
.gitattributes 1.53 KB 52373fe2 download
README.md 912 B da9773a9 download
config.json 861 B 0ed3f2d5 download
special_tokens_map.json 644 B 3a784031 download
added_tokens.json 629 B 06135f3c download
generation_config.json 257 B f3635780 download

README current version from Hugging Face

license: apache-2.0
license_link: https://huggingface.co/huihui-ai/Qwen2.5-14B-Instruct-1M-abliterated/blob/main/LICENSE
language:

  • en
    pipeline_tag: text-generation
    base_model: Qwen/Qwen2.5-14B-Instruct-1M
    tags:
  • chat
  • abliterated
  • uncensored
    library_name: transformers

huihui-ai/Qwen2.5-14B-Instruct-1M-abliterated

This is an uncensored version of Qwen/Qwen2.5-14B-Instruct-1M created with abliteration (see remove-refusals-with-transformers to know more about it).
This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.

Use with ollama

You can use huihui_ai/qwen2.5-1m-abliterated directly

ollama run huihui_ai/qwen2.5-1m-abliterated:14b
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