Consensus Control for Decentralized Deep Learning
Lingjing Kong 1 * Tao Lin 1 * Anastasia Koloskova 1 Martin Jaggi 1 Sebastian U. Stich 1
Abstract local mini-batch gradients computed on different subsets
Decentralized training of deep learning models of the data, for the later synchronized model update. How-
enables on-device learning over networks, as well ever, exact averaging with All-Reduce is sensitive to the
as efficient scaling to ...


雷达卡




京公网安备 11010802022788号







