Gradient-based Intra-attention Pruning on Pre-trained Language Models.

Ziqing Yang, Yiming Cui, Xin Yao, Shijin Wang

Browse the full ACL paper archive.