Federated Learning Hyper-Parameter Tuning From A System Perspective

Federated learning (FL) is a distributed model training paradigm that preserves clients’ data privacy. It has gained tremendous attention from both academia and industry. FL hyper-parameters (e.g., the number of selected clients passes) significantly affect overhead in terms computation time, transmission load, load. However, current practice manually selecting imposes heavy burden on practitioners because applications have different preferences. In this paper, we propose, an automatic hyper-parameter tuning algorithm tailored to applications’ diverse system requirements training. iteratively adjusts during can be easily integrated into existing systems. Through extensive evaluations for aggregation algorithms, show lightweight effective, achieving 8.48%-26.75% reduction compared using fixed hyper-parameters. This paper assists designing high-performance solutions. The source code available at.