search

ashawkey / stable-dreamfusion

Text-to-3D & Image-to-3D & Mesh Exportation with NeRF + Diffusion.
Apache License 2.0
8.2k stars 720 forks source link

New zero123-XL released #327

Closed stas-polukeev closed 1 year ago

stas-polukeev commented 1 year ago

Hi! The authors of zero123 just released a new version, zero123-XL, it could increase the quality of image-to-3D generation, if you use this in your repo. Here is the link: https://github.com/cvlab-columbia/zero123.

ashawkey commented 1 year ago

Thanks! I have updated the model link.

johnwon2007 commented 1 year ago

doesn/t the line in main.py need to be changed in order to use the new zero-xl.ckpt? or no around line 200 parser.add_argument('--zero123_ckpt', type=str, default='./pretrained/zero123/105000.ckpt', help="ckpt for zero123")

ashawkey commented 1 year ago

@johnwon2007 You are right! I have fixed this.

lovejoker1102 commented 1 year ago

how to test zero123 with the newest zero123-xl.ckpt? thanks...

wangjksjtu commented 1 year ago

@ashawkey @stas-polukeev I am curious what is the license for zero123-xl, thanks a lot!

stas-polukeev commented 1 year ago

@wangjksjtu I would assume they are the same, because the only difference is the training data, but i could be wrong.