Mast: Efficient Training of Mixture-of-Experts Transformers with Task Pipelining and Ordering.
Wenxiang Lin, Xinglin Pan, Shaohuai Shi, Xuan Wang, Bo Li, Xiaowen Chu
Browse the full ICDCS paper archive.
Wenxiang Lin, Xinglin Pan, Shaohuai Shi, Xuan Wang, Bo Li, Xiaowen Chu
Browse the full ICDCS paper archive.