返回模型目录
Aragduhl /

Huihui-Qwen3-VL-8B-Instruct-abliterated-comfy

HF 源站

Aragduhl/Huihui-Qwen3-VL-8B-Instruct-abliterated-comfy is an uncensored open-source model hosted on Hugging Face.

Image-Text-to-TextComfyUILicense: apache-2.08Btransformersqwen3-vlabliterateduncensored
README.md说明 / 使用方法

license: apache-2.0 base_model: huihui-ai/Huihui-Qwen3-VL-8B-Instruct-abliterated library_name: transformers pipeline_tag: image-text-to-text tags:

  • qwen3-vl
  • abliterated
  • uncensored
  • comfyui

Huihui Qwen3-VL-8B Instruct (Abliterated) — ComfyUI Single File

Single-file merge of the four shards of huihui-ai/Huihui-Qwen3-VL-8B-Instruct-abliterated, packaged for ComfyUI's native Generate Text node.

  • Format: single transformers-format safetensors (750 tensors)
  • Full vision tower + LM head included
  • Abliterated (uncensored)

ComfyUI usage

  1. Download and place the file in ComfyUI/models/text_encoders/.
  2. In the CLIPLoader node select:
    • clip_name: Huihui-Qwen3-VL-8B-Instruct-abliterated.safetensors
    • type: joyimage (routes to the qwen3vl_8b text encoder)
  3. Feed its CLIP output into a Generate Text node.

Download

hf download Aragduhl/Huihui-Qwen3-VL-8B-Instruct-abliterated-comfy --include "*.safetensors" --local-dir models/text_encoders/