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