Publication Type

Conference Proceeding Article

Version

acceptedVersion

Publication Date

7-2021

Abstract

We develop and analyze MARINA: a new communication efficient method for non-convex distributed learning over heterogeneous datasets. MARINA employs a novel communication compression strategy based on the compression of gradient differences that is reminiscent of but different from the strategy employed in the DIANA method of Mishchenko et al. (2019). Unlike virtually all competing distributed first-order methods, including DIANA, ours is based on a carefully designed biased gradient estimator, which is the key to its superior theoretical and practical performance. The communication complexity bounds we prove for MARINA are evidently better than those of all previous first-order methods. Further, we develop and analyze two variants of MARINA: VR-MARINA and PP-MARINA. The first method is designed for the case when the local loss functions owned by clients are either of a finite sum or of an expectation form, and the second method allows for a partial participation of clients -- a feature important in federated learning. All our methods are superior to previous state-of-the-art methods in terms of oracle/communication complexity. Finally, we provide a convergence analysis of all methods for problems satisfying the Polyak-Lojasiewicz condition.

Discipline

Databases and Information Systems

Research Areas

Data Science and Engineering; Intelligent Systems and Optimization

Publication

Proceedings of the 38th International Conference on Machine Learning (ICML 2021), Virtual Conference, July 18-24

First Page

1

Last Page

41

Publisher

Proceedings of Machine Learning Research

City or Country

Virtual

Additional URL

https://proceedings.mlr.press/v139/gorbunov21a

Download

Share

COinS