Flexible Personalized Split Federated Learning for On-Device Fine-Tuning of Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yuan, Tianjun, Geng, Jiaxiang, Han, Pengchao, Chen, Xianhao, Luo, Bing
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911105445527552
author Yuan, Tianjun
Geng, Jiaxiang
Han, Pengchao
Chen, Xianhao
Luo, Bing
author_facet Yuan, Tianjun
Geng, Jiaxiang
Han, Pengchao
Chen, Xianhao
Luo, Bing
contents Fine-tuning foundation models is critical for superior performance on personalized downstream tasks, compared to using pre-trained models. Collaborative learning can leverage local clients' datasets for fine-tuning, but limited client data and heterogeneous data distributions hinder effective collaboration. To address the challenge, we propose a flexible personalized federated learning paradigm that enables clients to engage in collaborative learning while maintaining personalized objectives. Given the limited and heterogeneous computational resources available on clients, we introduce \textbf{flexible personalized split federated learning (FlexP-SFL)}. Based on split learning, FlexP-SFL allows each client to train a portion of the model locally while offloading the rest to a server, according to resource constraints. Additionally, we propose an alignment strategy to improve personalized model performance on global data. Experimental results show that FlexP-SFL outperforms baseline models in personalized fine-tuning efficiency and final accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2508_10349
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Flexible Personalized Split Federated Learning for On-Device Fine-Tuning of Foundation Models
Yuan, Tianjun
Geng, Jiaxiang
Han, Pengchao
Chen, Xianhao
Luo, Bing
Distributed, Parallel, and Cluster Computing
Machine Learning
Fine-tuning foundation models is critical for superior performance on personalized downstream tasks, compared to using pre-trained models. Collaborative learning can leverage local clients' datasets for fine-tuning, but limited client data and heterogeneous data distributions hinder effective collaboration. To address the challenge, we propose a flexible personalized federated learning paradigm that enables clients to engage in collaborative learning while maintaining personalized objectives. Given the limited and heterogeneous computational resources available on clients, we introduce \textbf{flexible personalized split federated learning (FlexP-SFL)}. Based on split learning, FlexP-SFL allows each client to train a portion of the model locally while offloading the rest to a server, according to resource constraints. Additionally, we propose an alignment strategy to improve personalized model performance on global data. Experimental results show that FlexP-SFL outperforms baseline models in personalized fine-tuning efficiency and final accuracy.
title Flexible Personalized Split Federated Learning for On-Device Fine-Tuning of Foundation Models
topic Distributed, Parallel, and Cluster Computing
Machine Learning
url https://arxiv.org/abs/2508.10349