• python 2.7
• pytorch 0.4
• tensorflow
• pyhocon

1. Download the pre-trained pruned model.

• resnet-18-pruned0.3 BaiduDrive, resnet-18-pruned0.3 GoogleDrive
• resnet-18-pruned0.5 BaiduDrive, resnet-18-pruned0.5 GoogleDrive
• resnet-18-pruned0.7 BaiduDrive, resnet-18-pruned0.7 GoogleDrive
• resnet-50-pruned0.3 BaiduDrive, resnet-50-pruned0.3 GoogleDrive
• resnet-50-pruned0.5 BaiduDrive, resnet-50-pruned0.5 GoogleDrive
• resnet-50-pruned0.7 BaiduDrive, resnet-50-pruned0.7 GoogleDrive

2. Add DCP into PYTHONPATH.

# This is my path of DCP. You need to change to your path of DCP.
export PYTHONPATH=/home/liujing/Codes/Discrimination-aware-Channel-Pruning-for-Deep-Neural-Networks:$PYTHONPATH

3. Set configuration for testing. You need to set data_path, pruning_rate, depth and the retrain in dcp/channel_pruning/test.hocon.

cd dcp/channel_pruning/
vim dcp/channel_pruning/test.hocon

4. Run testing.

python test.py test.hocon

1. Download pre-trained mdoel.

• resnet-56 BaiduDrive, resnet-56 GoogleDrive.

2. Add DCP into PYTHONPATH.

# This is my path of DCP. You need to change to your path of DCP.
export PYTHONPATH=/home/liujing/Codes/Discrimination-aware-Channel-Pruning-for-Deep-Neural-Networks:$PYTHONPATH

3. Set configuration for channel pruning. Before pruning, you need to set save_path, data_path, experiment_id and the retrain in dcp/channel_pruning/cifar10_resnet.hocon.

cd dcp/channel_pruning/
vim dcp/channel_pruning/cifar10_resnet.hocon

4. Run Discrimination-aware Channel Pruning.

python channel_pruning.py cifar10_resnet.hocon

5. Set configuration for fine-tuning. Before fine-tuning, you need to set retrain to the path of model_004.pth in check_point folder

vim cifar10_resnet.hocon

6. Fine-tune the pruned model.

python fine_tuning.py cifar10_resnet.hocon

If you find DCP useful in your research, please consider to cite the following related papers:

@incollection{NIPS2018_7367,
title = {Discrimination-aware Channel Pruning for Deep Neural Networks},
author = {Zhuang, Zhuangwei and Tan, Mingkui and Zhuang, Bohan and Liu, Jing and Guo, Yong and Wu, Qingyao and Huang, Junzhou and Zhu, Jinhui},
booktitle = {Advances in Neural Information Processing Systems 31},
editor = {S. Bengio and H. Wallach and H. Larochelle and K. Grauman and N. Cesa-Bianchi and R. Garnett},
pages = {881--892},
year = {2018},
publisher = {Curran Associates, Inc.},
url = {http://papers.nips.cc/paper/7367-discrimination-aware-channel-pruning-for-deep-neural-networks.pdf}
}