yitu-opensource / T2T-ViT

ICCV2021, Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet
Other
1.15k stars 176 forks source link

fix a bug #70

Closed xmy0916 closed 2 years ago

xmy0916 commented 2 years ago

fix a bug