deepinsight / deepinsight/insightface
[BUG] ViT mask ratio parameter
Open
- Dominant language
- Python
- Stars
- 29.7k
- Forks
- 6.1k
- PR merge metrics
- No merged PRs in 30d
Description
https://github.com/deepinsight/insightface/blob/cc64315087304e3eda99559a5539f5d0b62512c5/recognition/arcface_torch/backbones/vit.py#L260
In this way, it always works with the predefined 0.1 mask ratio.
It should be :
` x, _, ids_restore = self.random_masking(x, mask_ratio=self.mask_ratio) `
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.