GSPMD: General and Scalable Parallelization for ML Computation Graphs.
Yuanzhong XuHyoukJoong LeeDehao ChenBlake A. HechtmanYanping HuangRahul JoshiMaxim KrikunDmitry LepikhinAndy LyMarcello MaggioniRuoming PangNoam ShazeerShibo WangTao WangYonghui WuZhifeng ChenPublished in: CoRR (2021)