Ouroboros: On Accelerating Training of
Transformer-Based Language Models
Qian Yang1 , Zhouyuan Huo2 , Wenlin Wang1 , Heng Huang2 , Lawrence Carin1
Department of Electrical and Computer Engineering
1
Duke University 2 University of Pittsburgh
qian.yang@duke.edu
Abstract
Language models are essential for natural language processing (NLP) tasks, such
as machine translation and text summa ...
附件列表