FedALT: Federated Fine-Tuning through Adaptive Local Training with Rest-of-World LoRA

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bian, Jieming, Wang, Lei, Zhang, Letian, Xu, Jie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911265136312320
author Bian, Jieming
Wang, Lei
Zhang, Letian
Xu, Jie
author_facet Bian, Jieming
Wang, Lei
Zhang, Letian
Xu, Jie
contents Fine-tuning large language models (LLMs) in federated settings enables privacy-preserving adaptation but suffers from cross-client interference due to model aggregation. Existing federated LoRA fine-tuning methods, primarily based on FedAvg, struggle with data heterogeneity, leading to harmful cross-client interference and suboptimal personalization. In this work, we propose \textbf{FedALT}, a novel personalized federated LoRA fine-tuning algorithm that fundamentally departs from FedAvg. Instead of using an aggregated model to initialize local training, each client continues training its individual LoRA while incorporating shared knowledge through a separate Rest-of-World (RoW) LoRA component. To effectively balance local adaptation and global information, FedALT introduces an adaptive mixer that dynamically learns input-specific weightings between the individual and RoW LoRA components, drawing conceptual foundations from the Mixture-of-Experts (MoE) paradigm. Through extensive experiments on NLP benchmarks, we demonstrate that FedALT significantly outperforms state-of-the-art personalized federated LoRA fine-tuning methods, achieving superior local adaptation without sacrificing computational efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2503_11880
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FedALT: Federated Fine-Tuning through Adaptive Local Training with Rest-of-World LoRA
Bian, Jieming
Wang, Lei
Zhang, Letian
Xu, Jie
Machine Learning
Artificial Intelligence
Computation and Language
Fine-tuning large language models (LLMs) in federated settings enables privacy-preserving adaptation but suffers from cross-client interference due to model aggregation. Existing federated LoRA fine-tuning methods, primarily based on FedAvg, struggle with data heterogeneity, leading to harmful cross-client interference and suboptimal personalization. In this work, we propose \textbf{FedALT}, a novel personalized federated LoRA fine-tuning algorithm that fundamentally departs from FedAvg. Instead of using an aggregated model to initialize local training, each client continues training its individual LoRA while incorporating shared knowledge through a separate Rest-of-World (RoW) LoRA component. To effectively balance local adaptation and global information, FedALT introduces an adaptive mixer that dynamically learns input-specific weightings between the individual and RoW LoRA components, drawing conceptual foundations from the Mixture-of-Experts (MoE) paradigm. Through extensive experiments on NLP benchmarks, we demonstrate that FedALT significantly outperforms state-of-the-art personalized federated LoRA fine-tuning methods, achieving superior local adaptation without sacrificing computational efficiency.
title FedALT: Federated Fine-Tuning through Adaptive Local Training with Rest-of-World LoRA
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2503.11880