Ready to run PyTorch implementation of Data2Vec 2.0: Highly efficient self-supervised representation learning for vision, speech and text. - View it on GitHub