5fa1a76
1
2
3
The abstract from the paper is the following: We introduce a self-supervised vision representation model BEiT, which stands for Bidirectional Encoder representation from Image Transformers.
