Neural Network Pruning by Cooperative Coevolution

Authors: Haopu Shang, Jia-Liang Wu, Wenjing Hong, Chao Qian

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experiments show that CCEP can achieve a competitive performance with the stateof-the-art pruning methods, e.g., prune Res Net56 for 63.42% FLOPs on CIFAR10 with 0.24% accuracy drop, and Res Net50 for 44.56% FLOPs on Image Net with 0.07% accuracy drop.
Researcher Affiliation Academia 1State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China 2Department of Computer Science and Engineering Southern University of Science and Technology, Shenzhen 518055, China
Pseudocode Yes Algorithm 1 CCEP framework
Open Source Code No Supplementary materials are available at https://arxiv.org/abs/2204.05639.
Open Datasets Yes Two classic data sets CIFAR10 [Krizhevsky and Hinton, 2009] and Image Net [Russakovsky et al., 2015] for image classification are used for the examination.
Dataset Splits Yes Image Net contains 1.28M images in the training set and 50K in the validation set, for 1K classes.
Hardware Specification No The paper does not provide specific details about the hardware used for experiments (e.g., GPU models, CPU types).
Software Dependencies No The paper does not specify any software dependencies with version numbers.
Experiment Setup Yes The settings of CCEP are described as follows. It runs for 12 iterations, i.e., T = 12 in Algorithm 1. For the EA (i.e., Algorithm 2) in each group, the population size m is 5, the mutation rates p1 = 0.05 and p2 = 0.1, the ratio bound r is 0.1, the maximum generation G is 10, and 20% of the training set is used for accuracy evaluation.