Robust Federated Finetuning of Foundation Models via Alternating Minimization of LoRA

TL;DR

This paper introduces RoLoRA, a robust federated fine-tuning framework that leverages alternating minimization of LoRA, significantly improving robustness and communication efficiency in federated learning of foundation models.

Contribution

The paper proposes RoLoRA, a novel federated fine-tuning method using alternating minimization of LoRA, enhancing robustness against data heterogeneity and reducing communication costs.

Findings

RoLoRA improves robustness in federated fine-tuning.

RoLoRA reduces communication overhead.

RoLoRA outperforms previous PEFT methods in heterogeneous data scenarios.

Abstract

Parameter-Efficient Fine-Tuning (PEFT) has risen as an innovative training strategy that updates only a select few model parameters, significantly lowering both computational and memory demands. PEFT also helps to decrease data transfer in federated learning settings, where communication depends on the size of updates. In this work, we explore the constraints of previous studies that integrate a well-known PEFT method named LoRA with federated fine-tuning, then introduce RoLoRA, a robust federated fine-tuning framework that utilizes an alternating minimization approach for LoRA, providing greater robustness against decreasing fine-tuning parameters and increasing data heterogeneity. Our results indicate that RoLoRA not only presents the communication benefits but also substantially enhances the robustness and effectiveness in multiple federated fine-tuning scenarios.