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

huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated

huihui-ai Qwen 33B
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%2FDeepSeek-R1-Distill-Qwen-32B-abliterated"
Response includes
  • classification m3
  • files 23
  • benchmarks 5 entries
  • hub_downloads_all_time 257,022
  • providers 1
  • author_summary 185 models
  • readme_text full
Try next
GET/models?author=huihui-ai
All 185 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 · lifetime
257K
1K last 30d - cooling
Likes
257
Descendants
23
in 23 direct forks
Model age
20mo ago
created 2025-01-22
Available via
1 provider
featherless-ai
Downloads over time
Now257.3K→from291↑88,327%
094.3K188.7K283K291 on Jan 22, 2025257.3K on Oct 12Jan '25Apr '25Jul '25Oct '25JanAprJulOct
Jan 22, 2025 → Oct 12 · 136 snapshots · spans 628 days

Benchmarks

Portrait before abliteration
Benchmarks of the base model as it stood before the refusal-removal operation. Compare with the numbers above to see what the operation cost.
Benchmark Score Source
BBH average 0.4003113244259956 OpenLLM-v2
IFEval instruct 0.4916067146282973 OpenLLM-v2
IFEval-Prompt 0.3456561922365989 OpenLLM-v2
MATH lvl 5 0 OpenLLM-v2
MMLU-Pro 0.46866688829787234 OpenLLM-v2

Genealogy 23 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.

deepseek-ai/DeepSeek-R1-Distill-Qwen-32B this lineage
huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated↓ 1,130 23 forks
bartowski/DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 20,023
Thaurock/DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 1,570
mradermacher/DeepSeek-R1-Distill-Qwen-32B-abliterated-i1-GGUF↓ 1,006
mradermacher/DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 764
mlx-community/DeepSeek-R1-Distill-Qwen-32B-abliterated-4bit↓ 649
DevQuasar-4/huihui-ai.DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 328
lactroiii/DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 183
S010/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q8-mlx↓ 154
mlx-community/DeepSeek-R1-Distill-Qwen-32B-abliterated↓ 131
tensorblock/DeepSeek-R1-Distill-Qwen-32B-abliterated-GGUF↓ 92
KYUNGYONG/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4-mlx↓ 60
phenixstar/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_K_M-GGUF↓ 51
raissadev/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_K_M-GGUF↓ 48
JaySinDub/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_K_M-GGUF↓ 34
jacobcarajo/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q5_K_M-GGUF↓ 34
iceqi1979/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_K_M-GGUF↓ 25
G17c21ds/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q8_0-GGUF↓ 23
Homurachyan/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_K_M-GGUF↓ 20
StephenGenusa/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q4_0-GGUF↓ 19
BenevolenceMessiah/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q8_0-GGUF↓ 18
havenwood/DeepSeek-R1-Distill-Qwen-32B-abliterated-Q5_K_M-GGUF↓ 14
Apel-sin/deepseek-r1-distill-qwen-32B-abliterated-exl2↓ 1
numen-tech/DeepSeek-R1-Distill-Qwen-32B-abliterated-GPTQ-Int4

Metadata

Tags
transformers safetensors qwen2 text-generation abliterated uncensored conversational base_model:deepseek-ai/DeepSeek-R1-Distill-Qwen-32B base_model:finetune:deepseek-ai/DeepSeek-R1-Distill-Qwen-32B text-generation-inference endpoints_compatible region:us

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
richardyoung/Deepseek-R1-Distill-Qwen-32b-uncensored 291 Wlc7758/Deepseek-R1-Distill-Qwen-32b-uncensored 94 fcvcggvvv/DeepSeek-R1-Distill-Qwen-32B-abliterated 26 stepenZEN/DeepSeek-R1-Distill-Qwen-32B-abliterated 24 nicoboss/DeepSeek-R1-Distill-Qwen-32B-Uncensored-Lora 20
Total size
61.0 GB
Files
23
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-02-16 08:42
Source huggingface.co/huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated ↗ Base model deepseek-ai/DeepSeek-R1-Distill-Qwen-32B ↗

Files by quantization

Auxiliary files 23 files 61.0 GB
model-00001-of-00014.safetensors 4.56 GB 74b83d9e download
model-00004-of-00014.safetensors 4.54 GB 801833f0 download
model-00005-of-00014.safetensors 4.54 GB d02bae3f download
model-00006-of-00014.safetensors 4.54 GB d50aad93 download
model-00007-of-00014.safetensors 4.54 GB f75522eb download
model-00008-of-00014.safetensors 4.54 GB e74ed688 download
model-00009-of-00014.safetensors 4.54 GB f483180c download
model-00010-of-00014.safetensors 4.54 GB 4ab01ae1 download
model-00011-of-00014.safetensors 4.54 GB 472ca947 download
model-00012-of-00014.safetensors 4.54 GB 31cc5a45 download
model-00013-of-00014.safetensors 4.54 GB a1a74ebd download
model-00003-of-00014.safetensors 4.54 GB e6f5632e download
model-00002-of-00014.safetensors 4.54 GB 7fda82f4 download
model-00014-of-00014.safetensors 1.98 GB ea2415e3 download
tokenizer.json 10.9 MB e20ddafc download
model.safetensors.index.json 62.5 KB d19f75d8 download
tokenizer_config.json 6.79 KB f8c31060 download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.16 KB ce58fa06 download
LICENSE 1.04 KB d42fae90 download
config.json 691 B 1c00c16b download
special_tokens_map.json 508 B 6caea2ce download
generation_config.json 190 B 829e028b download

README current version from Hugging Face

base_model:

  • deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
    library_name: transformers
    tags:
  • abliterated
  • uncensored

huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated

This is an uncensored version of deepseek-ai/DeepSeek-R1-Distill-Qwen-32B 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.

If "<think>" does not appear or refuses to respond, you can first provide an example to guide, and then ask your question.
For instance:

  How many 'r' characters are there in the word "strawberry"?

Use with ollama

You can use huihui_ai/deepseek-r1-abliterated directly

ollama run huihui_ai/deepseek-r1-abliterated:32b

Donation

Your donation helps us continue our further development and improvement, a cup of coffee can do it.
  • bitcoin:
  bc1qqnkhuchxw0zqjh2ku3lu4hq45hc6gy84uk70ge

README history 6 versions

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

  1. 2025-02-16Update README.md939b7e21.2 KB
    Loading...
  2. 2025-02-09Update README.md1ff2d35972 B
    Loading...
  3. 2025-02-09Update README.mde4bbf63955 B
    Loading...
  4. 2025-01-26Update README.mde2b053e787 B
    Loading...
  5. 2025-01-22Upload 22 files0388ce9599 B
    Loading...
  6. 2025-01-22initial commit2720b2428 B
    Loading...

Discussions 9 threads

  1. 2026-05-07PRFix chat_template crash when assistant message omits the `content` keyopen1 💬#9
    Loading...
  2. 2025-06-15PRPull Req1 💬#8
    Loading...
  3. 2025-02-12效果不是很好，字数总是不达标，还总是重复。open2 💬#7
    Loading...
  4. 2025-02-08Abliteration but still not entirely uncensoredopen6 💬#6
    Loading...
  5. 2025-01-31Please tell me the difference between 'abliterated' and 'uncensored' versionopen7 💬#5
    Loading...
  6. 2025-01-30Confused about uncensoring scopeopen1 💬#4
    Loading...
  7. 2025-01-29bad performanceopen9 💬#3
    Loading...
  8. 2025-01-23the abliteration doesn't workopen5 💬#2
    Loading...
  9. 2025-01-23Good work!open3 💬#1
    Loading...

Continue on Hugging Face →

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