Official PyTorch implementation of Equivariant-SSL (E-SSL).
@article{dangovski2021equivariant,
title={Equivariant Contrastive Learning},
author={Dangovski, Rumen and Jing, Li and Loh, Charlotte and Han, Seungwook and Srivastava, Akash and Cheung, Brian and Agrawal, Pulkit and Solja{\v{c}}i{\'c}, Marin},
journal={arXiv preprint arXiv:2111.00899},
year={2021}
}
The code for each dataset is self-contained. Please, inspect imagenet/, cifar10/ and photonics/ for the
corresponding datasets.
Let us know about interesting work with E-SSL and we will spread the word here.
Our work is accepted at ICLR 2022. Please, follow the project's webpage for updates.
Equivariant Contrastive Learning helps to achieve state-of-the-art results among unsupervised sentence representation learning methods via the DiffCSE's method.
This project is released under MIT License, which allows commercial use. See LICENSE for details.