search

Stability-AI / stable-fast-3d

SF3D: Stable Fast 3D Mesh Reconstruction with UV-unwrapping and Illumination Disentanglement
https://stable-fast-3d.github.io
Other
1.16k stars 118 forks source link

ValueError('num_channels must be divisible by num_groups') #32

Closed Lucalme closed 2 months ago

Lucalme commented 2 months ago

I tried reinstalling my venv, deleting the folder altogether and restarting from scratch but this message keeps poping whenever I launch either the app or just the run script. I already had success running the run.py, not sure about the app. But I must have messed up somewhere so it does not work anymore.

here's the callback for reference

(venv) PS E:\Softs\AI Stuff\Stability AI\stable-fast-3d> python run.py demo_files/examples/chair1.png --output-dir output/ Traceback (most recent call last): File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\run.py", line 71, in model = SF3D.from_pretrained( File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\sf3d\system.py", line 89, in from_pretrained model = cls(cfg) File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\sf3d\models\utils.py", line 29, in init self.configure(*args, *kwargs) File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\sf3d\system.py", line 105, in configure self.backbone = find_class(self.cfg.backbone_cls)(self.cfg.backbone) File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\sf3d\models\utils.py", line 29, in init self.configure(args, **kwargs) File "E:\Softs\AI Stuff\Stability AI\stable-fast-3d\sf3d\models\transformers\backbone.py", line 430, in configure self.norm_triplane = torch.nn.GroupNorm( File "E:\Softs\AI Stuff\Stability AI\venv\lib\site-packages\torch\nn\modules\normalization.py", line 267, in init raise ValueError('num_channels must be divisible by num_groups') ValueError: num_channels must be divisible by num_groups

Lucalme commented 2 months ago

Nevermind, I think I just forgot to login into huggingface again when i changed my HF_HOME env path.