Salvato in:
Dettagli Bibliografici
Autori principali: Deng, Yuyang, Qiao, Fuli, Mahdavi, Mehrdad
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2511.06132
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915606761046016
author Deng, Yuyang
Qiao, Fuli
Mahdavi, Mehrdad
author_facet Deng, Yuyang
Qiao, Fuli
Mahdavi, Mehrdad
contents As learning models continue to grow in size, enabling on-device local training of these models has emerged as a critical challenge in federated learning. A popular solution is sub-model training, where the server only distributes randomly sampled sub-models to the edge clients, and clients only update these small models. However, those random sampling of sub-models may not give satisfying convergence performance. In this paper, observing the success of SGD with shuffling, we propose a distributed shuffled sub-model training, where the full model is partitioned into several sub-models in advance, and the server shuffles those sub-models, sends each of them to clients at each round, and by the end of local updating period, clients send back the updated sub-models, and server averages them. We establish the convergence rate of this algorithm. We also study the generalization of distributed sub-model training via stability analysis, and find that the sub-model training can improve the generalization via amplifying the stability of training process. The extensive experiments also validate our theoretical findings.
format Preprint
id arxiv_https___arxiv_org_abs_2511_06132
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Convergence and Stability of Distributed Sub-model Training
Deng, Yuyang
Qiao, Fuli
Mahdavi, Mehrdad
Machine Learning
As learning models continue to grow in size, enabling on-device local training of these models has emerged as a critical challenge in federated learning. A popular solution is sub-model training, where the server only distributes randomly sampled sub-models to the edge clients, and clients only update these small models. However, those random sampling of sub-models may not give satisfying convergence performance. In this paper, observing the success of SGD with shuffling, we propose a distributed shuffled sub-model training, where the full model is partitioned into several sub-models in advance, and the server shuffles those sub-models, sends each of them to clients at each round, and by the end of local updating period, clients send back the updated sub-models, and server averages them. We establish the convergence rate of this algorithm. We also study the generalization of distributed sub-model training via stability analysis, and find that the sub-model training can improve the generalization via amplifying the stability of training process. The extensive experiments also validate our theoretical findings.
title On the Convergence and Stability of Distributed Sub-model Training
topic Machine Learning
url https://arxiv.org/abs/2511.06132