← back to catalog · registered 2026-11-20 13:04

Daddydev21/Huihui-Qwen3.8-27B-abliterated-GGUF

Daddydev21 Qwen 27B GGUF multimodal
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/Daddydev21%2FHuihui-Qwen3.8-27B-abliterated-GGUF"
Response includes
  • classification m8
  • files 53
  • author_summary 1 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
0
Likes
0
Model age
1d ago
created 2026-11-20

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
Quantizations
BF16 F16 IQ2 IQ3 IQ4 Q2_K Q3_K Q4_K Q5_K Q6_K Q8_0 Q8_K
Tags
transformers gguf abliterated uncensored huihui qwen3 unsloth ternary 2-bit llama-cpp cuda metal

Related

Control vectors

Total size
795 GB
Files
53
Quantizations
13
Registered
2026-11-20 13:04
Last updated on HF
2026-11-20 13:04

Files by quantization

BF16 3 files 103 GB
Huihui-Qwen3.8-27B-abliterated-UD-DW-bf16.gguf 50.9 GB ******** download
Huihui-Qwen3.8-27B-abliterated-bf16.gguf 50.9 GB ******** download
mmproj-model-bf16.gguf 888 MB ******** download
F16 1 file 50.1 GB
Huihui-Qwen3.8-27B-abliterated-Ternary-Bonsai-f16.gguf 50.1 GB ******** download
Q8_0 2 files 63.2 GB
Huihui-Qwen3.8-27B-abliterated-Q8_0_L.gguf 36.1 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q8_0.gguf 27.1 GB ******** download
Q8_K 2 files 54.7 GB
Huihui-Qwen3.8-27B-abliterated-UD-Q8_K_XL.gguf 29.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-DW-Q8_K_L.gguf 25.4 GB ******** download
Q6_K 4 files 89.5 GB
Huihui-Qwen3.8-27B-abliterated-Q6_K_L.gguf 23.2 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-Q6_K_XL.gguf 23.1 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-DW-Q6_K_L.gguf 22.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q6_K.gguf 20.9 GB ******** download
Q5_K 3 files 58.7 GB
Huihui-Qwen3.8-27B-abliterated-Q5_K_L.gguf 21.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-Q5_K_XL.gguf 19.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q5_K.gguf 18.2 GB ******** download
Q4_K 5 files 81.3 GB
Huihui-Qwen3.8-27B-abliterated-Q4_K_L.gguf 19.5 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-Q4_K_XL.gguf 16.2 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q4_K.gguf 15.7 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-DW-Q4_K_M.gguf 15.4 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-DW-Q4_K_S.gguf 14.6 GB ******** download
Q3_K 3 files 42.4 GB
Huihui-Qwen3.8-27B-abliterated-Q3_K_L.gguf 17.4 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q3_K.gguf 12.6 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-Q3_K_XL.gguf 12.4 GB ******** download
Q2_K 3 files 35.5 GB
Huihui-Qwen3.8-27B-abliterated-Q2_K_L.gguf 16.0 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Q2_K.gguf 10.1 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-Q2_K_XL.gguf 9.32 GB ******** download
IQ4 1 file 13.4 GB
Huihui-Qwen3.8-27B-abliterated-UD-IQ4_XS.gguf 13.4 GB ******** download
IQ3 8 files 84.2 GB
Huihui-Qwen3.8-27B-abliterated-UD-DW-IQ3_S.gguf 11.5 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-IQ3_S.gguf 11.1 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_S.gguf 11.0 GB ******** download
Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_S.gguf 11.0 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-DW-IQ3_XXS.gguf 10.6 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-IQ3_XXS.gguf 10.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_XXS.gguf 9.40 GB ******** download
Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_XXS.gguf 9.40 GB ******** download
IQ2 3 files 24.3 GB
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ2_S.gguf 8.62 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-IQ2_S.gguf 7.85 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ2_XS.gguf 7.84 GB ******** download
Auxiliary files 15 files 95.7 GB
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_S-mtp.gguf 11.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_S-mtp.gguf 11.3 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_XXS-mtp.gguf 9.73 GB ******** download
Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_XXS-mtp.gguf 9.73 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ2_S-mtp.gguf 8.95 GB ******** download
Huihui-Qwen3.8-27B-abliterated-UD-IQ2_S-MTP.gguf 8.17 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ2_XS-mtp.gguf 8.16 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Underdog-Saluki-Swift-1.5-IQ2-mtp.gguf 7.68 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Underdog-Saluki-Swift-1.5-IQ2.gguf 7.36 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Ternary-Bonsai-PQ2_0.gguf 7.18 GB ******** download
Huihui-Qwen3.8-27B-abliterated-Ternary-Bonsai-PTQ1_0.gguf 6.15 GB ******** download
README.md 8.58 KB 4f67762c download
.gitattributes 6.84 KB 32620c5a download
Qwen3.8-27B-tensor_types-Q6_K_L.txt 123 B 34440335 download
Qwen3.8-27B-tensor_types-Q8_0_L.txt 123 B b06f0153 download

README current version from Hugging Face


library_name: transformers
license: apache-2.0
pipeline_tag: image-text-to-text
base_model:

  • Qwen/Qwen3.8-27B
    tags:
  • abliterated
  • uncensored
  • huihui
  • qwen3
  • unsloth
  • ternary
  • 2-bit
  • llama-cpp
  • cuda
  • metal
  • on-device
  • hybrid-attention
  • prismml
  • bonsai

huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF

This is an uncensored version of Qwen/Qwen3.8-27B 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.

BF16, 1M context

2× RTX 6000 Pro

llama-server \
  -m huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-bf16.gguf \
  --mmproj huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/mmproj-model-bf16.gguf \
  --port 8080 --host 0.0.0.0 \
  -ngl 99 \
  -fa on \ 
  -c 1048576 \
  -np 1 \
  --rope-scaling yarn \
  --rope-scale 4 \
  --yarn-orig-ctx 262144 \
  --override-kv qwen2.context_length=int:1048576 \
  --cache-type-k bf16 \
  --cache-type-v bf16 \
  -b 4096 \
  -ub 1024  \
  --jinja

patch-unsloth 9

The newly added Huihui-Qwen3.8-27B-abliterated-patch-all-5.gguf contains 27 newly added weights and supports the quantized versions of unsloth/Qwen3.8-27B-GGUF updated on November 11, 2026.

The patch-unsloth here can patch all quantized variants of the latest version of unsloth/Qwen3.8-27B-GGUF, from UD-IQ1_S to BF16, which saves storage space; you need to run the command yourself to apply the patch. For details,
Please refer to README-patch-unsloth.md in the patch-unsloth directory.

Latest update 8

The newly added Huihui-Qwen3.8-27B-abliterated-Underdog-Saluki series come from ConwayResearch/Underdog-Saluki-27B-1.0 and ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF.
Only layers 22 to 52 (0-based indexing) have been ablated, while the other layers remain unablated. It may come with a small disclaimer warning.

Due to ConwayResearch/Underdog-Saluki-27B-1.0 not providing an imatrix file, we will attempt to perform ablation by mixing the GGUF weights of Swift-1.5 and ConwayResearch/Underdog-Saluki-27B-1.0.

ConwayResearch/Underdog-Saluki-27B-1.0/Underdog-Saluki-27B-1.0-IQ2-mix.gguf only ablated 62 weights.

This is just a test/validation.

Latest update 7

The newly added Huihui-Qwen3.8-27B-abliterated-Swift series come from ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF.
Only layers 22 to 52 (0-based indexing) have been ablated, while the other layers remain unablated. It may come with a small disclaimer warning.
This is just a test/validation.

Latest update 6

The newly added Huihui-Qwen3.8-27B-abliterated-Ternary series come from prism-ml/Ternary-Bonsai-2-27B-gguf.
Only layers 22 to 52 (0-based indexing) have been ablated, while the other layers remain unablated. It may come with a small disclaimer warning.
The size after conversion may differ from the original GGUF (Some of the weights are converted from PTQ1 to Q2_K or Q3_K.).
This is just a test/validation.

The ternary hybrid-attention kernels live in the
PrismML-Eng/llama.cpp fork. Stock llama.cpp will not run these files.

Latest update 5

The newly added Huihui-Qwen3.8-27B-abliterated-GSQ-RCO series come from ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF.
Only layers 22 to 52 (0-based indexing) have been ablated, while the other layers remain unablated. It may come with a small disclaimer warning.
The size after conversion may differ from the original GGUF.

Latest update 4

The newly added Huihui-Qwen3.8-27B-abliterated-UD-DW series come from unsloth/Qwen3.8-27B-GGUF.
Only layers 22 to 52 (0-based indexing) have been ablated, while the other layers remain unablated. It may come with a small disclaimer warning.
The size after conversion may differ from the original GGUF.

Latest update

The newly added Huihui-Qwen3.8-27B-abliterated-UD series come from unsloth/Qwen3.8-27B-GGUF.
Only layers 17 to 52 (0-based indexing) have been ablated(Previously, The first 15 layers were retained without ablation), while the other layers remain unablated.
The size after conversion may differ from the original GGUF.

Huihui-Qwen3.8-27B-abliterated-bf16.gguf has also been updated.

This helps retain more of the original model’s performance. MTP and visual has not been modified.

Note

The first 15 layers were retained without ablation. MTP and visual has not been modified.

We have already converted the weights (token_embd,output,ffn_down,ssm_out,attn_output) that need to be ablated in the versions below Q8_0 from Q2_K, Q3_K, Q4_K, Q5_K, and Q6_K to Q8_0 to improve response quality, and changed the filename to K_L.

In the Q8_0 quantized version, we changed the Q8_0 weights (token_embd,output,ffn_down,ssm_out,attn_output) targeted for ablation to BF16 and renamed the file to Q8_0_L.

This is not a standard quantization, so you might find that Q2_K_L is larger than Q3_K and Q4_K.

Specific Quantification Method

Some people may misunderstand. The specific quantification method is as follows.

Q2_K_L - Q6_K_L

Qwen3.8-27B-tensor_types-Q6_K_L.txt

llama-quantize \
  --allow-requantize \
  --tensor-type-file huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Qwen3.8-27B-tensor_types-Q6_K_L.txt \
  huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-bf16.gguf \
  huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-Q6_K_L.gguf Q6_K

Q8_0_L

Qwen3.8-27B-tensor_types-Q8_0_L.txt

llama-quantize \ 
  --allow-requantize \
  --tensor-type-file huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Qwen3.8-27B-tensor_types-Q8_0_L.txt \
  huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-bf16.gguf \ 
  huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-Q8_0_L.gguf Q8_0

ollama

Please use the latest version of ollama

You can use huihui_ai/Qwen3.8-abliterated directly,

ollama run huihui_ai/Qwen3.8-abliterated

llama.cpp

Use the latest llama.cpp,

llama-cli -m huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF/Huihui-Qwen3.8-27B-abliterated-Q4_K.gguf -c 262144

Usage Warnings

  • Risk of Sensitive or Controversial Outputs: This model’s safety filtering has been significantly reduced, potentially generating sensitive, controversial, or inappropriate content. Users should exercise caution and rigorously review generated outputs.

  • Not Suitable for All Audiences: Due to limited content filtering, the model’s outputs may be inappropriate for public settings, underage users, or applications requiring high security.

  • Legal and Ethical Responsibilities: Users must ensure their usage complies with local laws and ethical standards. Generated content may carry legal or ethical risks, and users are solely responsible for any consequences.

  • Research and Experimental Use: It is recommended to use this model for research, testing, or controlled environments, avoiding direct use in production or public-facing commercial applications.

  • Monitoring and Review Recommendations: Users are strongly advised to monitor model outputs in real-time and conduct manual reviews when necessary to prevent the dissemination of inappropriate content.

  • No Default Safety Guarantees: Unlike standard models, this model has not undergone rigorous safety optimization. huihui.ai bears no responsibility for any consequences arising from its use.

Donation

Your donation helps us continue our further development and improvement, a cup of coffee can do it.
  • bitcoin:
  bc1qqnkhuchxw0zqjh2ku3lu4hq45hc6gy84uk70ge
  • Support our work on Ko-fi!
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