Dep-$L_0$: Improving $L_0$-based Network Sparsification via Dependency Modeling

Abstract

Training deep neural networks with an $L_0$ regularization is one of the prominent approaches for network pruning or sparsification. The method prunes the network during training by encouraging weights to become exactly zero. However, recent work of Gale et al. reveals that although this method yields high compression rates on smaller datasets, it performs inconsistently on large-scale learning tasks, such as ResNet50 on ImageNet. We analyze this phenomenon through the lens of variational inference and find that it is likely due to the independent modeling of binary gates, the mean-field approximation, which is known in Bayesian statistics for its poor performance due to the crude approximation. To mitigate this deficiency, we propose a dependency modeling of binary gates, which can be modeled effectively as a multi-layer perceptron (MLP). We term our algorithm Dep-$L_0$ as it prunes networks via a dependency-enabled $L_0$ regularization. Extensive experiments on CIFAR10, CIFAR100 and ImageNet with VGG16, ResNet50, ResNet56 show that our Dep-$L_0$ outperforms the original $L_0$-HC algorithm of Louizos et al. by a significant margin, especially on ImageNet. Compared with the state-of-the-arts network sparsification algorithms, our dependency modeling makes the $L_0$-based sparsification once again very competitive on large-scale learning tasks. Our source code is available at https://github.com/leo-yangli/dep-l0.

Published as a conference paper at ECML 2021

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Distilling the Knowledge in a Neural Network
• Network Trimming: A Data-Driven Neuron Pruning Approach towards Efficient Deep Architectures
• To prune, or not to prune: exploring the efficacy of pruning for model compression
• Speeding up Convolutional Neural Networks with Low Rank Expansions
• The State of Sparsity in Deep Neural Networks
• Global Sparse Momentum SGD for Pruning Very Deep Neural Networks
• Approximated Oracle Filter Pruning for Destructive CNN Width Optimization