Saved in:
Bibliographic Details
Main Authors: Hao, Jie, Wu, Yuman, Payani, Ali, Lee, Myungjin, Liu, Mingrui
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.03920
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909526943334400
author Hao, Jie
Wu, Yuman
Payani, Ali
Lee, Myungjin
Liu, Mingrui
author_facet Hao, Jie
Wu, Yuman
Payani, Ali
Lee, Myungjin
Liu, Mingrui
contents We study the task of personalized federated fine-tuning with heterogeneous data in the context of language models, where clients collaboratively fine-tune a language model (e.g., BERT, GPT) without sharing their local data, achieving personalization simultaneously. While recent efforts have applied parameter-efficient fine-tuning techniques like low-rank adaptation (LoRA) in federated settings, they typically use single or multiple independent low-rank adapters with predefined maximal and minimal ranks, which may not be optimal for diverse data sources over clients. To address this issue, we propose PF2LoRA, a new personalized federated fine-tuning algorithm built on a novel \emph{automatic rank learning approach via two-level LoRA}. Given the pretrained language model whose weight is frozen, our algorithm aims to learn two levels of adaptation simultaneously: the first level aims to learn a common adapter for all clients, while the second level fosters individual client personalization. A key advantage of PF2LoRA is its ability to adaptively determine a suitable rank based on an individual client's data, rather than relying on a predefined rank that is agnostic to data heterogeneity. We present a synthetic example that highlights how PF2LoRA automatically learns the ground-truth rank for each client, tailoring the adaptation to match the properties of their individual data. Notably, this approach introduces minimal additional memory overhead, as the second-level adaptation comprises a small number of parameters compared to the first level. Our experiments on natural language understanding and generation tasks demonstrate that PF2LoRA significantly outperforms existing federated fine-tuning methods.
format Preprint
id arxiv_https___arxiv_org_abs_2503_03920
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Personalized Federated Fine-tuning for Heterogeneous Data: An Automatic Rank Learning Approach via Two-Level LoRA
Hao, Jie
Wu, Yuman
Payani, Ali
Lee, Myungjin
Liu, Mingrui
Machine Learning
Computation and Language
We study the task of personalized federated fine-tuning with heterogeneous data in the context of language models, where clients collaboratively fine-tune a language model (e.g., BERT, GPT) without sharing their local data, achieving personalization simultaneously. While recent efforts have applied parameter-efficient fine-tuning techniques like low-rank adaptation (LoRA) in federated settings, they typically use single or multiple independent low-rank adapters with predefined maximal and minimal ranks, which may not be optimal for diverse data sources over clients. To address this issue, we propose PF2LoRA, a new personalized federated fine-tuning algorithm built on a novel \emph{automatic rank learning approach via two-level LoRA}. Given the pretrained language model whose weight is frozen, our algorithm aims to learn two levels of adaptation simultaneously: the first level aims to learn a common adapter for all clients, while the second level fosters individual client personalization. A key advantage of PF2LoRA is its ability to adaptively determine a suitable rank based on an individual client's data, rather than relying on a predefined rank that is agnostic to data heterogeneity. We present a synthetic example that highlights how PF2LoRA automatically learns the ground-truth rank for each client, tailoring the adaptation to match the properties of their individual data. Notably, this approach introduces minimal additional memory overhead, as the second-level adaptation comprises a small number of parameters compared to the first level. Our experiments on natural language understanding and generation tasks demonstrate that PF2LoRA significantly outperforms existing federated fine-tuning methods.
title Personalized Federated Fine-tuning for Heterogeneous Data: An Automatic Rank Learning Approach via Two-Level LoRA
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2503.03920