POET: Training Neural Networks on Tiny Devices with Integrated Rematerialization and Paging

Shishir G. Patil, Paras Jain 0001, Prabal Dutta, Ion Stoica, Joseph Gonzalez 0001. POET: Training Neural Networks on Tiny Devices with Integrated Rematerialization and Paging. In Kamalika Chaudhuri, Stefanie Jegelka, Le Song, Csaba Szepesvári, Gang Niu 0001, Sivan Sabato, editors, International Conference on Machine Learning, ICML 2022, 17-23 July 2022, Baltimore, Maryland, USA. Volume 162 of Proceedings of Machine Learning Research, pages 17573-17583, PMLR, 2022. [doi]

Abstract

Abstract is missing.