Communication-Efficient Learning of Deep Networks from Decentralized Data

TL;DR

This paper introduces federated learning, a decentralized approach for training deep neural networks on mobile devices with privacy-sensitive data, achieving significant reductions in communication costs through iterative model averaging.

Contribution

It presents a practical federated learning method for deep networks, demonstrating robustness to non-IID data and substantial communication efficiency improvements.

Findings

Achieves 10-100x reduction in communication rounds

Demonstrates robustness to unbalanced, non-IID data

Validates approach across multiple models and datasets

Abstract

Modern mobile devices have access to a wealth of data suitable for learning models, which in turn can greatly improve the user experience on the device. For example, language models can improve speech recognition and text entry, and image models can automatically select good photos. However, this rich data is often privacy sensitive, large in quantity, or both, which may preclude logging to the data center and training there using conventional approaches. We advocate an alternative that leaves the training data distributed on the mobile devices, and learns a shared model by aggregating locally-computed updates. We term this decentralized approach Federated Learning. We present a practical method for the federated learning of deep networks based on iterative model averaging, and conduct an extensive empirical evaluation, considering five different model architectures and four datasets. These experiments demonstrate the approach is robust to the unbalanced and non-IID data distributions that are a defining characteristic of this setting. Communication costs are the principal constraint, and we show a reduction in required communication rounds by 10-100x as compared to synchronized stochastic gradient descent.