English

Distributed and Stochastic Optimization Methods with Gradient Compression and Local Steps

Optimization and Control 2021-12-21 v1 Machine Learning

Abstract

In this thesis, we propose new theoretical frameworks for the analysis of stochastic and distributed methods with error compensation and local updates. Using these frameworks, we develop more than 20 new optimization methods, including the first linearly converging Error-Compensated SGD and the first linearly converging Local-SGD for arbitrarily heterogeneous local functions. Moreover, the thesis contains several new distributed methods with unbiased compression for distributed non-convex optimization problems. The derived complexity results for these methods outperform the previous best-known results for the considered problems. Finally, we propose a new scalable decentralized fault-tolerant distributed method, and under reasonable assumptions, we derive the iteration complexity bounds for this method that match the ones of centralized Local-SGD.

Keywords

Cite

@article{arxiv.2112.10645,
  title  = {Distributed and Stochastic Optimization Methods with Gradient Compression and Local Steps},
  author = {Eduard Gorbunov},
  journal= {arXiv preprint arXiv:2112.10645},
  year   = {2021}
}

Comments

PhD thesis, 416 pages, 27 figures

R2 v1 2026-06-24T08:24:49.530Z