This is an official implementation of CvT: Introducing Convolutions to Vision Transformers. - View it on GitHub