Data2vec 2.0: Highly efficient self-supervised learning for vision, speech, text by from on 2022-12-13 18:03 (#66SZS) Comments