xinyun/ComfyUI-Hunyuan3DWrapper
1
0
Fork 0
mirror of https://git.datalinker.icu/kijai/ComfyUI-Hunyuan3DWrapper.git synced 2026-01-23 07:04:25 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update readme.md

Browse Source
This commit is contained in:
Easymode 2025-02-18 23:58:44 +00:00 committed by GitHub
parent 3b65c82d89
commit e23e37cece
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 10 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
readme.md
View File

@ -1,3 +1,13 @@
# Added BPT
`cd ComfyUI-Hunyuan3DWrapper-main\hy3dgen\shapegen\bpt`
`python_embeded\python.exe -m pip install -r requirements.txt`
Download weights <https://huggingface.co/whaohan/bpt/blob/refs%2Fpr%2F1/bpt-8-16-500m.pt>
Download checkpoint <https://huggingface.co/Maikou/Michelangelo/blob/refs%2Fpr%2F2/checkpoints/aligned_shape_latents/shapevae-256.ckpt>
Copy `bpt-8-16-500m.pt`, `shapevae-256.ckpt` to `ComfyUI-Hunyuan3DWrapper-main\hy3dgen\shapegen\bpt`
# ComfyUI wrapper for [Hunyuan3D-2](https://github.com/Tencent/Hunyuan3D-2)
# WORK IN PROGRESS
@ -34,16 +44,6 @@ or with portable (in `ComfyUI_windows_portable` -folder):
Current latest portable was updated to use pytorch 2.6.0, for this you should use new wheel:
`python_embeded\python.exe -m pip install ComfyUI\custom_nodes\ComfyUI-Hunyuan3DWrapper\wheels\custom_rasterizer-0.1.0+torch260.cuda126-cp312-cp312-win_amd64.whl`
# BPT
`cd ComfyUI-Hunyuan3DWrapper-main\hy3dgen\shapegen\bpt`
`python_embeded\python.exe -m pip install -r requirements.txt`
Download weights <https://huggingface.co/whaohan/bpt/blob/refs%2Fpr%2F1/bpt-8-16-500m.pt>
Download checkpoint <https://huggingface.co/Maikou/Michelangelo/blob/refs%2Fpr%2F2/checkpoints/aligned_shape_latents/shapevae-256.ckpt>
Copy `bpt-8-16-500m.pt`, `shapevae-256.ckpt` to `ComfyUI-Hunyuan3DWrapper-main\hy3dgen\shapegen\bpt`
This was tested to work on latest ComfyUI portable install
If this doesn't work, you need to compile yourself: