Practical Federated Learning without a Server

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dhasade, Akash, Kermarrec, Anne-Marie, Lavoie, Erick, Pouwelse, Johan, Sharma, Rishi, de Vos, Martijn
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912401677352960
author Dhasade, Akash
Kermarrec, Anne-Marie
Lavoie, Erick
Pouwelse, Johan
Sharma, Rishi
de Vos, Martijn
author_facet Dhasade, Akash
Kermarrec, Anne-Marie
Lavoie, Erick
Pouwelse, Johan
Sharma, Rishi
de Vos, Martijn
contents Federated Learning (FL) enables end-user devices to collaboratively train ML models without sharing raw data, thereby preserving data privacy. In FL, a central parameter server coordinates the learning process by iteratively aggregating the trained models received from clients. Yet, deploying a central server is not always feasible due to hardware unavailability, infrastructure constraints, or operational costs. We present Plexus, a fully decentralized FL system for large networks that operates without the drawbacks originating from having a central server. Plexus distributes the responsibilities of model aggregation and sampling among participating nodes while avoiding network-wide coordination. We evaluate Plexus using realistic traces for compute speed, pairwise latency and network capacity. Our experiments on three common learning tasks and with up to 1000 nodes empirically show that Plexus reduces time-to-accuracy by 1.4-1.6x, communication volume by 15.8-292x and training resources needed for convergence by 30.5-77.9x compared to conventional decentralized learning algorithms.
format Preprint
id arxiv_https___arxiv_org_abs_2503_05509
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Practical Federated Learning without a Server
Dhasade, Akash
Kermarrec, Anne-Marie
Lavoie, Erick
Pouwelse, Johan
Sharma, Rishi
de Vos, Martijn
Distributed, Parallel, and Cluster Computing
Federated Learning (FL) enables end-user devices to collaboratively train ML models without sharing raw data, thereby preserving data privacy. In FL, a central parameter server coordinates the learning process by iteratively aggregating the trained models received from clients. Yet, deploying a central server is not always feasible due to hardware unavailability, infrastructure constraints, or operational costs. We present Plexus, a fully decentralized FL system for large networks that operates without the drawbacks originating from having a central server. Plexus distributes the responsibilities of model aggregation and sampling among participating nodes while avoiding network-wide coordination. We evaluate Plexus using realistic traces for compute speed, pairwise latency and network capacity. Our experiments on three common learning tasks and with up to 1000 nodes empirically show that Plexus reduces time-to-accuracy by 1.4-1.6x, communication volume by 15.8-292x and training resources needed for convergence by 30.5-77.9x compared to conventional decentralized learning algorithms.
title Practical Federated Learning without a Server
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2503.05509