BiCoLoR: Communication-Efficient Optimization with Bidirectional Compression and Local Training

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Condat, Laurent, Maranjyan, Artavazd, Richtárik, Peter
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915737698828288
author Condat, Laurent
Maranjyan, Artavazd
Richtárik, Peter
author_facet Condat, Laurent
Maranjyan, Artavazd
Richtárik, Peter
contents Slow and costly communication is often the main bottleneck in distributed optimization, especially in federated learning where it occurs over wireless networks. We introduce BiCoLoR, a communication-efficient optimization algorithm that combines two widely used and effective strategies: local training, which increases computation between communication rounds, and compression, which encodes high-dimensional vectors into short bitstreams. While these mechanisms have been combined before, compression has typically been applied only to uplink (client-to-server) communication, leaving the downlink (server-to-client) side unaddressed. In practice, however, both directions are costly. We propose BiCoLoR, the first algorithm to combine local training with bidirectional compression using arbitrary unbiased compressors. This joint design achieves accelerated complexity guarantees in both convex and strongly convex heterogeneous settings. Empirically, BiCoLoR outperforms existing algorithms and establishes a new standard in communication efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2601_12400
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BiCoLoR: Communication-Efficient Optimization with Bidirectional Compression and Local Training
Condat, Laurent
Maranjyan, Artavazd
Richtárik, Peter
Optimization and Control
Machine Learning
Slow and costly communication is often the main bottleneck in distributed optimization, especially in federated learning where it occurs over wireless networks. We introduce BiCoLoR, a communication-efficient optimization algorithm that combines two widely used and effective strategies: local training, which increases computation between communication rounds, and compression, which encodes high-dimensional vectors into short bitstreams. While these mechanisms have been combined before, compression has typically been applied only to uplink (client-to-server) communication, leaving the downlink (server-to-client) side unaddressed. In practice, however, both directions are costly. We propose BiCoLoR, the first algorithm to combine local training with bidirectional compression using arbitrary unbiased compressors. This joint design achieves accelerated complexity guarantees in both convex and strongly convex heterogeneous settings. Empirically, BiCoLoR outperforms existing algorithms and establishes a new standard in communication efficiency.
title BiCoLoR: Communication-Efficient Optimization with Bidirectional Compression and Local Training
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2601.12400