CCEP

Description : This package includes the Python code of the CCEP algorithm [1] for neural network pruning. CCEP uses the idea of divide-and-conquer in cooperative coevolution to reduce the search space of pruning and prunes the filters in each layer by EAs separately. Experiments on two datasets CIFAR-10 and ImageNet show the good performance of CCEP. README files are included in the package, showing how to use the code.

References: [1] Haopu Shang, Jia-Liang Wu, Wenjing Hong and Chao Qian. Neural Network Pruning by Cooperative Coevolution. In: Proceedings of the 31th International Joint Conference on Artificial Intelligence (IJCAI'22), Vienna, Austria, 2022.

ATTN: This package is free for academic usage. You can run it at your own risk. For other purposes, please contact Dr. Chao Qian (qianc@lamda.nju.edu.cn).

Requirement: The package was developed with Python, especially on PyTorch.

ATTN2: This package was developed by Mr. Haopu Shang (shanghp@lamda.nju.edu.cn) and Mr. Jia-Liang Wu (wujl@lamda.nju.edu.cn). For any problem concerning the code, please feel free to contact Mr. Shang or Mr. Wu.

Download: code (1MB)