Beyond Fixed Rounds: Data-Free Early Stopping for Practical Federated Learning
Abstract
Federated Learning (FL) facilitates decentralized collaborative learning without transmitting raw data. However, reliance on fixed global rounds or validation data for hyperparameter tuning hinders practical deployment by incurring high computational costs and privacy risks. To address this, we propose a data-free early stopping framework that determines the optimal stopping point by monitoring the task vector's growth rate using solely server-side parameters. The numerical results on skin lesion/blood cell classification demonstrate that our approach is comparable to validation-based early stopping across various state-of-the-art FL methods. In particular, the proposed framework requires an average of 45/12 (skin lesion/blood cell) additional rounds to achieve over 12.3%/8.9% higher performance than early stopping based on validation data. To the best of our knowledge, this is the first work to propose an data-free early stopping framework for FL methods.
Cite
@article{arxiv.2601.22669,
title = {Beyond Fixed Rounds: Data-Free Early Stopping for Practical Federated Learning},
author = {Youngjoon Lee and Hyukjoon Lee and Seungrok Jung and Andy Luo and Jinu Gong and Yang Cao and Joonhyuk Kang},
journal= {arXiv preprint arXiv:2601.22669},
year = {2026}
}
Comments
Replaced with experimental results on AMD MI300X AI accelerators