1 paper
Ray Cao, Sherry Luo, Steve Gan +1
In this study, we explore the impact of relaxing data consistency in parallel machine learning training during a failure using various parameter server configurations. Our failure…