Improving Transformer Optimization Through Better Initialization

Xiao Shi Huang, Felipe PĂ©rez, Jimmy Ba, Maksims Volkovs. Improving Transformer Optimization Through Better Initialization. In Proceedings of the 37th International Conference on Machine Learning, ICML 2020, 13-18 July 2020, Virtual Event. Volume 119 of Proceedings of Machine Learning Research, pages 4475-4483, PMLR, 2020. [doi]

Abstract

Abstract is missing.