This repository contains a Pytorch implementation for several Vision Transformers papers.
Vanilla ViT 📜
The Vanilla ViT can be trained of different datasets using train_vit.py
The hyper parameters are not optimized for each dataset but the performance should correct by default.