Why there is "no_weight_decay" function for Swin-T but not for VIT

microsoft / SimMIM

This is an official implementation for "SimMIM: A Simple Framework for Masked Image Modeling".

https://arxiv.org/abs/2111.09886

MIT License

917 stars 86 forks source link

Why there is "no_weight_decay" function for Swin-T but not for VIT #26

Open TongZhangTHU opened 2 years ago

TongZhangTHU commented 2 years ago

Hi, I am wondering, why in "simmim.py", there is "no_weight_decay" function for "class SwinTransformerForSimMIM", but not for "class VisionTransformerForSimMIM" ?