"Lita: Accelerating Distributed Training of Sparsely Activated Models."

Jiamin Li et al. (2022)

Details and statistics

DOI: 10.48550/ARXIV.2210.17223

access: open

type: Informal or Other Publication

metadata version: 2023-07-14

a service of  Schloss Dagstuhl - Leibniz Center for Informatics