File size: 1,314 Bytes
25e6a01
 
 
76d0f32
 
49c9f90
 
 
 
25e6a01
 
9714467
 
68e2bb3
 
 
61e86e3
 
 
68e2bb3
9714467
 
 
ee390e1
25e6a01
adfc248
c7ec611
61e86e3
25e6a01
61e86e3
76d0f32
25e6a01
76d0f32
25e6a01
df936e9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
---
language:
- en
base_model:
- tencent/HunyuanVideo-Avatar
library_name: gguf
license: other
license_name: tencent-hunyuan-community
license_link: LICENSE.md
---

---

> [!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).