Model Compression with Generative Adversarial Networks

Ruishan Liu, Nicolo Fusi, Lester Mackey

Abstract: More accurate machine learning models often demand more computation and memory at test time, making them difficult to deploy on CPU- or memory-constrained devices. Model compression (also known as distillation) alleviates this burden by training a less expensive student model to mimic the expensive teacher model while maintaining most of the original accuracy. However, when fresh data is unavailable for the compression task, the teacher's training data is typically reused, leading to suboptimal compression. In this work, we propose to augment the compression dataset with synthetic data from a generative adversarial network (GAN) designed to approximate the training data distribution. Our GAN-assisted model compression (GAN-MC) significantly improves student accuracy for expensive models such as deep neural networks and large random forests on both image and tabular datasets. Building on these results, we propose a comprehensive metric—the Compression Score—to evaluate the quality of synthetic datasets based on their induced model compression performance. The Compression Score captures both data diversity and discriminability, and we illustrate its benefits over the popular Inception Score in the context of image classification.

Keywords: Model compression, distillation, generative adversarial network, GAN, deep neural network, random forest, ensemble, decision tree, convolutional neural network

Data: [CIFAR-10](https://paperswithcode.com/dataset/cifar-10)

Community Implementations: [ 3 code implementations](https://www.catalyzex.com/paper/model-compression-with-generative-adversarial/code)