Layer-wise and Dimension-wise Locally Adaptive Federated Learning

TL;DR

This paper introduces a novel federated learning framework with layer-wise adaptivity, improving convergence speed and generalization in training deep neural networks across diverse data distributions.

Contribution

It proposes layer-wise adaptive FL algorithms, Fed-LAMB and Mime-LAMB, with theoretical convergence guarantees and superior empirical performance.

Findings

Faster convergence compared to recent adaptive FL methods

Better generalization performance on various datasets

Linear speedup with the number of workers

Abstract

In the emerging paradigm of Federated Learning (FL), large amount of clients such as mobile devices are used to train possibly high-dimensional models on their respective data. Combining (dimension-wise) adaptive gradient methods (e.g. Adam, AMSGrad) with FL has been an active direction, which is shown to outperform traditional SGD based FL in many cases. In this paper, we focus on the problem of training federated deep neural networks, and propose a novel FL framework which further introduces layer-wise adaptivity to the local model updates. Our framework can be applied to locally adaptive FL methods including two recent algorithms, Mime and Fed-AMS. Theoretically, we provide a convergence analysis of our layer-wise FL methods, coined Fed-LAMB and Mime-LAMB, which matches the convergence rate of state-of-the-art results in FL and exhibits linear speedup in terms of the number of workers. Experimental results on various datasets and models, under both IID and non-IID local data settings, show that both Fed-LAMB and Mime-LAMB achieve faster convergence speed and better generalization performance, compared to the various recent adaptive FL methods.