Communication-Censored Distributed Stochastic Gradient Descent
From MaRDI portal
Publication:6324903
arXiv1909.03631MaRDI QIDQ6324903
Weiyu Li, Tianyi Chen, Liping Li, Qing Ling, Zhaoxian Wu
Publication date: 9 September 2019
Abstract: This paper develops a communication-efficient algorithm to solve the stochastic optimization problem defined over a distributed network, aiming at reducing the burdensome communication in applications such as distributed machine learning.Different from the existing works based on quantization and sparsification, we introduce a communication-censoring technique to reduce the transmissions of variables, which leads to our communication-Censored distributed Stochastic Gradient Descent (CSGD) algorithm. Specifically, in CSGD, the latest mini-batch stochastic gradient at a worker will be transmitted to the server if and only if it is sufficiently informative. When the latest gradient is not available, the stale one will be reused at the server. To implement this communication-censoring strategy, the batch-size is increasing in order to alleviate the effect of stochastic gradient noise. Theoretically, CSGD enjoys the same order of convergence rate as that of SGD, but effectively reduces communication. Numerical experiments demonstrate the sizable communication saving of CSGD.
Has companion code repository: https://github.com/Weiyu-USTC/CSGD
Cites Work
This page was built for publication: Communication-Censored Distributed Stochastic Gradient Descent
