# 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.

- **Kind:** model (Hugging Face)
- **Hugging Face:** https://huggingface.co/huihui-ai/Huihui-Qwen3.8-27B-abliterated-GGUF
- **Owner:** huihui-ai (https://forefy.com/u/hf/huihui-ai)
- **Parameters:** 27.3B
- **Context:** 262144 tokens
- **License:** apache-2.0
- **Focus:** general
- **Architecture:** Qwen3_5ForConditionalGeneration
- **Pinned revision:** `1c69098e9c32db6dc06fc48f1d6cc75ad07da89e`
- **Page:** https://forefy.com/models/c39831bc-0356-5534-a6ac-8826e85c630d
- **API (JSON):** https://forefy.com/api/asr/models/c39831bc-0356-5534-a6ac-8826e85c630d

## What was done to it

- abliterated (declared evidence) - remove-refusals-with-transformers
- quantized (declared evidence, from Qwen/Qwen3.8-27B)

## Lineage

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

## Checks (repo at the pinned revision; not a behavior test)

- **Identity** [pass]: No lookalike name or conflicting provenance
- **Weights format** [pass]: GGUF only, nothing runs on load
- **Pickle files** [pass]: None.
- **Remote code** [pass]: None
- **Chat template** [pass]: Same as Qwen/Qwen3.8-27B's
- **Refusals modified** [info]: Yes: abliteration (declared)
- **License** [pass]: apache-2.0
- **Lineage** [pass]: Derives from Qwen/Qwen3.8-27B
- **Model card** [warn]: 3 of 6, missing training data, language, evaluation results
- **GGUF headers** [pass]: 2 file(s) parsed, well-formed

## Variants

- Huihui-Qwen3.8-27B-abliterated-UD-DW-bf16 (gguf, 54.7 GB)
- Huihui-Qwen3.8-27B-abliterated-bf16 (gguf, 54.7 GB)
- F16 (gguf, 53.8 GB)
- Q8_0_L (gguf, 38.8 GB)
- Q8_K_XL (gguf, 31.5 GB)
- Q8_0 (gguf, 29.0 GB)
- Q8_K_L (gguf, 27.3 GB)
- Huihui-Qwen3.8-27B-abliterated-Q6_K_L (gguf, 24.9 GB)
- Q6_K_XL (gguf, 24.8 GB)
- Huihui-Qwen3.8-27B-abliterated-UD-DW-Q6_K_L (gguf, 23.9 GB)
- Q5_K_L (gguf, 22.9 GB)
- Q6_K (gguf, 22.4 GB)
- Q4_K_L (gguf, 20.9 GB)
- Q5_K_XL (gguf, 20.7 GB)
- Q5_K (gguf, 19.5 GB)
- Q3_K_L (gguf, 18.7 GB)
- Q4_K_XL (gguf, 17.4 GB)
- Q2_K_L (gguf, 17.2 GB)
- Q4_K (gguf, 16.8 GB)
- Q4_K_M (gguf, 16.6 GB)
- Q4_K_S (gguf, 15.6 GB)
- IQ4_XS (gguf, 14.4 GB)
- Q3_K (gguf, 13.5 GB)
- Q3_K_XL (gguf, 13.3 GB)
- Huihui-Qwen3.8-27B-abliterated-UD-DW-IQ3_S (gguf, 12.3 GB)
- Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_S-mtp (gguf, 12.1 GB)
- Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_S-mtp (gguf, 12.1 GB)
- Huihui-Qwen3.8-27B-abliterated-UD-IQ3_S (gguf, 12.0 GB)
- Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_S (gguf, 11.8 GB)
- Huihui-Qwen3.8-27B-abliterated-GSQ-RCO-IQ3_S (gguf, 11.8 GB)

## Run it