Automatic Neural Network Compression by Sparsity-Quantization Joint Learning: A Constrained Optimization-based Approach

TL;DR

This paper introduces an automatic framework for neural network compression that jointly prunes and quantizes DNNs without manual hyper-parameter tuning, achieving significant size reduction without accuracy loss.

Contribution

It presents a constrained optimization-based method for automatic joint pruning and quantization, eliminating the need for manual hyper-parameter setting for each layer.

Findings

ResNet-50 compressed 836× smaller on CIFAR-10 with no accuracy loss.

AlexNet compressed 205× smaller on ImageNet with no accuracy loss.

Abstract

Deep Neural Networks (DNNs) are applied in a wide range of usecases. There is an increased demand for deploying DNNs on devices that do not have abundant resources such as memory and computation units. Recently, network compression through a variety of techniques such as pruning and quantization have been proposed to reduce the resource requirement. A key parameter that all existing compression techniques are sensitive to is the compression ratio (e.g., pruning sparsity, quantization bitwidth) of each layer. Traditional solutions treat the compression ratios of each layer as hyper-parameters, and tune them using human heuristic. Recent researchers start using black-box hyper-parameter optimizations, but they will introduce new hyper-parameters and have efficiency issue. In this paper, we propose a framework to jointly prune and quantize the DNNs automatically according to a target model size without using any hyper-parameters to manually set the compression ratio for each layer. In the experiments, we show that our framework can compress the weights data of ResNet-50 to be 836$\times$ smaller without accuracy loss on CIFAR-10, and compress AlexNet to be 205$\times$ smaller without accuracy loss on ImageNet classification.