This is an unofficial PyTorch implementation of EdgeViT in "EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers", arXiv 2022.
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool
yanjingke opened this issue 2 years ago · comments
The model doesn't seem to fit
Hi, I noticed that indeed the code missed an MLP in the LGL bottleneck, which has been added now. Thanks!