Partial Variable Training for Efficient On-Device Federated Learning
This addresses efficiency challenges for edge device users in Federated Learning, though it is incremental as it builds on existing FL methods.
The paper tackles the problem of limited memory and expensive communication in Federated Learning on edge devices by proposing Partial Variable Training (PVT), which trains only a subset of variables, resulting in up to 1.9x memory reduction and 593x communication cost reduction while maintaining comparable accuracy.
This paper aims to address the major challenges of Federated Learning (FL) on edge devices: limited memory and expensive communication. We propose a novel method, called Partial Variable Training (PVT), that only trains a small subset of variables on edge devices to reduce memory usage and communication cost. With PVT, we show that network accuracy can be maintained by utilizing more local training steps and devices, which is favorable for FL involving a large population of devices. According to our experiments on two state-of-the-art neural networks for speech recognition and two different datasets, PVT can reduce memory usage by up to 1.9$\times$ and communication cost by up to 593$\times$ while attaining comparable accuracy when compared with full network training.