---
title: HunyuanVideo-Avatar-GGUF-Experiment
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment.md"
repository: AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment
last_updated: 2025-06-17
license: other
base_model:
  - tencent/HunyuanVideo-Avatar
base_model_relation: quantized
library_name:
  - gguf
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 296
stars: 0
tags:
  - gguf
---

# HunyuanVideo-Avatar-GGUF-Experiment

> HunyuanVideo-Avatar-GGUF-Experiment - AI-ModelScope 在 ModelScope 开源的模型。[!IMPORTANT] ⚠️ Important: This project is intended for experimental use only. Not yet supported in ComfyUI https://github.com/comfyanonymous/ComfyUI/issues/8311

AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment 是 ModelScope 魔搭社区上的机器学习模型，采用 other 许可，基于 tencent/HunyuanVideo-Avatar 构建。

- **Repository**: AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment
- **License**: other
- **Base model**: tencent/HunyuanVideo-Avatar
- **Tags**: gguf
- **Downloads**: 296
- **Stars**: 0
- **Last updated**: 2025-06-17

Source: https://www.modelscope.cn/models/AI-ModelScope/HunyuanVideo-Avatar-GGUF-Experiment

---

---

> [!IMPORTANT]
> ⚠️ **Important:**
>
> This project is intended for experimental use only.
> 
> Not yet supported in ComfyUI
> https://github.com/comfyanonymous/ComfyUI/issues/8311

---

This repository contains a GGUF conversion of [https://huggingface.co/tencent/HunyuanVideo-Avatar/blob/main/ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt](https://huggingface.co/tencent/HunyuanVideo-Avatar/blob/main/ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt) from [tencent/HunyuanVideo-Avatar](https://huggingface.co/tencent/HunyuanVideo-Avatar). 

The conversion scripts are provided by city96, available at the [ComfyUI-GGUF GitHub repository](https://github.com/city96/ComfyUI-GGUF/tree/main/tools).

The process involved first converting the pickletensors to a BF16 GGUF, then quantizing it, and finally applying the 5D fixes.

## Notes
As this is a quantized model not a finetune, all the same restrictions/original license terms still apply.

## Reference

- For an overview of quantization types, please see the [GGUF quantization types](https://huggingface.co/docs/hub/gguf#quantization-types).