QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning

TL;DR

QuPeD introduces a federated learning framework that enables personalized, quantized model training through knowledge distillation, effectively handling data and resource heterogeneity among clients.

Contribution

It proposes a novel quantized personalized federated learning algorithm using knowledge distillation and develops an optimization method with proven convergence.

Findings

QuPeD outperforms existing personalized FL methods in heterogeneous settings.

The algorithm effectively learns compressed personalized models with different quantization and structures.

Numerical results validate the efficiency and robustness of QuPeD in diverse scenarios.

Abstract

Traditionally, federated learning (FL) aims to train a single global model while collaboratively using multiple clients and a server. Two natural challenges that FL algorithms face are heterogeneity in data across clients and collaboration of clients with {\em diverse resources}. In this work, we introduce a \textit{quantized} and \textit{personalized} FL algorithm QuPeD that facilitates collective (personalized model compression) training via \textit{knowledge distillation} (KD) among clients who have access to heterogeneous data and resources. For personalization, we allow clients to learn \textit{compressed personalized models} with different quantization parameters and model dimensions/structures. Towards this, first we propose an algorithm for learning quantized models through a relaxed optimization problem, where quantization values are also optimized over. When each client participating in the (federated) learning process has different requirements for the compressed model (both in model dimension and precision), we formulate a compressed personalization framework by introducing knowledge distillation loss for local client objectives collaborating through a global model. We develop an alternating proximal gradient update for solving this compressed personalization problem, and analyze its convergence properties. Numerically, we validate that QuPeD outperforms competing personalized FL methods, FedAvg, and local training of clients in various heterogeneous settings.