MARS: Unleashing the Power of Variance Reduction for Training Large Models.
Huizhuo Yuan, Yifeng Liu, Shuang Wu, Xun Zhou, Quanquan Gu
Browse the full ICML paper archive.
Huizhuo Yuan, Yifeng Liu, Shuang Wu, Xun Zhou, Quanquan Gu
Browse the full ICML paper archive.