- Notifications
You must be signed in to change notification settings - Fork70
Caffe implementation for dynamic network surgery.
License
NotificationsYou must be signed in to change notification settings
yiwenguo/Dynamic-Network-Surgery
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
Dynamic network surgery is a very effective method for DNN pruning. To better use it with python and matlab, you may also need aclassic version of theCaffe framework.For the convolutional and fully-connected layers to be pruned, change their layer types to "CConvolution" and "CInnerProduct" respectively. Then, pass "cconvolution_param" and "cinner_product_param" messages to these modified layers for better pruning performance.
Below is an example for pruning the "ip1" layer in LeNet5:
layer { name: "ip1" type: "CInnerProduct" bottom: "pool2" top: "ip1" param { lr_mult: 1 } param { lr_mult: 2 } inner_product_param { num_output: 500 weight_filler { type: "xavier" } bias_filler { type: "constant" } } cinner_product_param { gamma: 0.0001 power: 1 c_rate: 4 iter_stop: 14000 weight_mask_filler { type: "constant" value: 1 } bias_mask_filler { type: "constant" value: 1 } } }Please cite our work in your publications if it helps your research:
@inproceedings{guo2016dynamic, title = {Dynamic Network Surgery for Efficient DNNs}, author = {Guo, Yiwen and Yao, Anbang and Chen, Yurong}, booktitle = {Advances in neural information processing systems (NIPS)}, year = {2016}}anddo not forget about Caffe:
@article{jia2014caffe, Author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor}, Journal = {arXiv preprint arXiv:1408.5093}, Title = {Caffe: Convolutional Architecture for Fast Feature Embedding}, Year = {2014}}Enjoy your own surgeries!