FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantization

TL;DR

FedPAQ introduces a communication-efficient federated learning approach that combines periodic averaging, partial device participation, and quantization to address scalability and communication challenges while maintaining strong theoretical guarantees.

Contribution

The paper proposes FedPAQ, a novel federated learning method that enhances communication efficiency through periodic averaging and quantization, with proven theoretical guarantees.

Findings

Achieves near-optimal convergence guarantees for convex and non-convex functions.

Reduces communication costs significantly compared to traditional methods.

Demonstrates effective tradeoffs between communication and computation in experiments.

Abstract

Federated learning is a distributed framework according to which a model is trained over a set of devices, while keeping data localized. This framework faces several systems-oriented challenges which include (i) communication bottleneck since a large number of devices upload their local updates to a parameter server, and (ii) scalability as the federated network consists of millions of devices. Due to these systems challenges as well as issues related to statistical heterogeneity of data and privacy concerns, designing a provably efficient federated learning method is of significant importance yet it remains challenging. In this paper, we present FedPAQ, a communication-efficient Federated Learning method with Periodic Averaging and Quantization. FedPAQ relies on three key features: (1) periodic averaging where models are updated locally at devices and only periodically averaged at the server; (2) partial device participation where only a fraction of devices participate in each round of the training; and (3) quantized message-passing where the edge nodes quantize their updates before uploading to the parameter server. These features address the communications and scalability challenges in federated learning. We also show that FedPAQ achieves near-optimal theoretical guarantees for strongly convex and non-convex loss functions and empirically demonstrate the communication-computation tradeoff provided by our method.