In this paper, we propose a novel meta learning approach for automatic channel pruning of very deep neural networks. We first train a PruningNet, a kind of meta network, which is able to generate weight parameters for any pruned structure given the target network. We use a simple stochastic structure sampling method for training the PruningNet. Then, we apply an evolutionary procedure to search for good-performing pruned networks. The search is highly efficient because the weights are directly generated by the trained PruningNet and we do not need any finetuning at search time. With a single PruningNet trained for the target network, we can search for various Pruned Networks under different constraints with little human participation. Compared to the state-of-the-art pruning methods, we have demonstrated superior performances on MobileNet V1/V2 and ResNet. Codes are available on https://github.com/liuzechun/MetaPruning.
@article{arxiv.1903.10258,
title = {MetaPruning: Meta Learning for Automatic Neural Network Channel Pruning},
author = {Zechun Liu and Haoyuan Mu and Xiangyu Zhang and Zichao Guo and Xin Yang and Tim Kwang-Ting Cheng and Jian Sun},
journal= {arXiv preprint arXiv:1903.10258},
year = {2019}
}
Comments
ICCV 2019 Camera ready version. Codes are available on https://github.com/liuzechun/MetaPruning