Stochastic Gradient Coding for Flexible Straggler Mitigation in Distributed Learning

Rawad Bitar, Mary Wootters, Salim El Rouayheb

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations

Abstract

We consider distributed gradient descent in the presence of stragglers. Recent work on gradient coding and approximate gradient coding have shown how to add redundancy in distributed gradient descent to guarantee convergence even if some workers are slow or non-responsive. In this work we propose a new type of approximate gradient coding which we call Stochastic Gradient Coding (SGC). The idea of SGC is very simple: We distribute data points redundantly to workers according to a good combinatorial design. We prove that the convergence rate of SGC mirrors that of batched Stochastic Gradient Descent (SGD) for the l2 loss function, and show how the convergence rate can improve with the redundancy. We show empirically that SGC requires a small amount of redundancy to handle a large number of stragglers and that it can outperform existing approximate gradient codes when the number of stragglers is large.

Original languageEnglish (US)
Title of host publication2019 IEEE Information Theory Workshop, ITW 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538669006
DOIs
StatePublished - Aug 2019
Event2019 IEEE Information Theory Workshop, ITW 2019 - Visby, Sweden
Duration: Aug 25 2019Aug 28 2019

Publication series

Name2019 IEEE Information Theory Workshop, ITW 2019

Conference

Conference2019 IEEE Information Theory Workshop, ITW 2019
Country/TerritorySweden
CityVisby
Period8/25/198/28/19

All Science Journal Classification (ASJC) codes

  • Software
  • Computational Theory and Mathematics
  • Computer Networks and Communications
  • Information Systems

Fingerprint

Dive into the research topics of 'Stochastic Gradient Coding for Flexible Straggler Mitigation in Distributed Learning'. Together they form a unique fingerprint.

Cite this