Accelerated Methods with Compressed Communications for Distributed Optimization Problems under Data Similarity

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bylinkin, Dmitry, Beznosikov, Aleksandr
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909593049759744
author Bylinkin, Dmitry
Beznosikov, Aleksandr
author_facet Bylinkin, Dmitry
Beznosikov, Aleksandr
contents In recent years, as data and problem sizes have increased, distributed learning has become an essential tool for training high-performance models. However, the communication bottleneck, especially for high-dimensional data, is a challenge. Several techniques have been developed to overcome this problem. These include communication compression and implementation of local steps, which work particularly well when there is similarity of local data samples. In this paper, we study the synergy of these approaches for efficient distributed optimization. We propose the first theoretically grounded accelerated algorithms utilizing unbiased and biased compression under data similarity, leveraging variance reduction and error feedback frameworks. Our results are of record and confirmed by experiments on different average losses and datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2412_16414
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Accelerated Methods with Compressed Communications for Distributed Optimization Problems under Data Similarity
Bylinkin, Dmitry
Beznosikov, Aleksandr
Optimization and Control
Distributed, Parallel, and Cluster Computing
Machine Learning
In recent years, as data and problem sizes have increased, distributed learning has become an essential tool for training high-performance models. However, the communication bottleneck, especially for high-dimensional data, is a challenge. Several techniques have been developed to overcome this problem. These include communication compression and implementation of local steps, which work particularly well when there is similarity of local data samples. In this paper, we study the synergy of these approaches for efficient distributed optimization. We propose the first theoretically grounded accelerated algorithms utilizing unbiased and biased compression under data similarity, leveraging variance reduction and error feedback frameworks. Our results are of record and confirmed by experiments on different average losses and datasets.
title Accelerated Methods with Compressed Communications for Distributed Optimization Problems under Data Similarity
topic Optimization and Control
Distributed, Parallel, and Cluster Computing
Machine Learning
url https://arxiv.org/abs/2412.16414