# Huihui-GLM-5.3-Flash-abliterated-GGUF

This is an uncensored version of zai-org/GLM-5.3-Flash 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.

- **Kind:** model (Hugging Face)
- **Hugging Face:** https://huggingface.co/huihui-ai/Huihui-GLM-5.3-Flash-abliterated-GGUF
- **Owner:** huihui-ai (https://forefy.com/u/hf/huihui-ai)
- **Parameters:** 320.8B
- **Context:** 1048576 tokens
- **License:** mit
- **Focus:** general
- **Architecture:** Glm5NextForConditionalGeneration
- **Pinned revision:** `c43a994637c10f5c376b7a471b3afc7663f80b9b`
- **Page:** https://forefy.com/models/28e99077-0360-5802-b738-2f890a0e7ad4
- **API (JSON):** https://forefy.com/api/asr/models/28e99077-0360-5802-b738-2f890a0e7ad4

## What was done to it

- abliterated (declared evidence) - remove-refusals-with-transformers
- quantized (declared evidence, from zai-org/GLM-5.3-Flash)

## Lineage

zai-org/GLM-5.3-Flash → huihui-ai/Huihui-GLM-5.3-Flash-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** [warn]: Differs from zai-org/GLM-5.3-Flash's
- **Refusals modified** [info]: Yes: abliteration (declared)
- **License** [pass]: mit
- **Lineage** [pass]: Derives from zai-org/GLM-5.3-Flash
- **Model card** [warn]: 4 of 6, missing training data, evaluation results
- **GGUF headers** [pass]: 2 file(s) parsed, well-formed

## Variants

- Q6_K_XL (gguf, 291.8 GB)
- Q5_K_XL (gguf, 240.3 GB)
- Q4_K_XL (gguf, 199.7 GB)
- IQ4_XS (gguf, 156.8 GB)
- Q3_K_XL (gguf, 147.5 GB)
- IQ3_XXS (gguf, 120.4 GB)
- Q2_K_XL (gguf, 108.7 GB)
- IQ2_XXS (gguf, 101.8 GB)
- IQ1_M (gguf, 97.6 GB)
- IQ1_S (gguf, 93.1 GB)

## Run it

- GGUF: `ollama run hf.co/huihui-ai/Huihui-GLM-5.3-Flash-abliterated-GGUF:IQ1_S`