Must-read papers on improving efficiency for pre-trained language models.
No issues in this repository yet.