Decoupled Greedy Learning of CNNs for Synchronous and Asynchronous Distributed Learning

TL;DR

This paper introduces Decoupled Greedy Learning (DGL), a method that enables parallel and asynchronous training of CNN layers, reducing communication overhead and addressing update locking in distributed neural network training.

Contribution

It proposes a novel decoupled greedy training approach for CNNs that allows for linear parallelization and asynchronous updates, with bandwidth reduction via online vector quantization.

Findings

DGL achieves effective training on CIFAR-10 and ImageNet datasets.

The approach converges both theoretically and empirically.

DGL outperforms some existing methods in distributed CNN training.

Abstract

A commonly cited inefficiency of neural network training using back-propagation is the update locking problem: each layer must wait for the signal to propagate through the full network before updating. Several alternatives that can alleviate this issue have been proposed. In this context, we consider a simple alternative based on minimal feedback, which we call Decoupled Greedy Learning (DGL). It is based on a classic greedy relaxation of the joint training objective, recently shown to be effective in the context of Convolutional Neural Networks (CNNs) on large-scale image classification. We consider an optimization of this objective that permits us to decouple the layer training, allowing for layers or modules in networks to be trained with a potentially linear parallelization. With the use of a replay buffer we show that this approach can be extended to asynchronous settings, where modules can operate and continue to update with possibly large communication delays. To address bandwidth and memory issues we propose an approach based on online vector quantization. This allows to drastically reduce the communication bandwidth between modules and required memory for replay buffers. We show theoretically and empirically that this approach converges and compare it to the sequential solvers. We demonstrate the effectiveness of DGL against alternative approaches on the CIFAR-10 dataset and on the large-scale ImageNet dataset.