English

An acceleration of decentralized SGD under general assumptions with low stochastic noise

Optimization and Control 2021-06-15 v3 Computational Complexity

Abstract

Distributed optimization methods are actively researched by optimization community. Due to applications in distributed machine learning, modern research directions include stochastic objectives, reducing communication frequency, and time-varying communication network topology. Recently, an analysis unifying several centralized and decentralized approaches to stochastic distributed optimization was developed in Koloskova et al. (2020). In this work, we employ a Catalyst framework and accelerate the rates of Koloskova et al. (2020) in the case of low stochastic noise.

Keywords

Cite

@article{arxiv.2011.07585,
  title  = {An acceleration of decentralized SGD under general assumptions with low stochastic noise},
  author = {Trimbach Ekaterina and Rogozin Alexander},
  journal= {arXiv preprint arXiv:2011.07585},
  year   = {2021}
}

Comments

MOTOR 2021 conference