Skip to content

KyGao/sparsity-and-robustness

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

pruning-recapitulation

merging some articles about network pruning

已经有了很多篇结合的文章了

这篇文章使用ADMM交叉乘子法进行两步剪枝,第一步正常样本训练,第二步对抗样本训练,达到对抗同时剪枝的目的。类似快手的ATMC

这篇文章走的是小众的 Bayesian Pruning 方法,考虑了对抗样本和正常样本的分布关系。提出了vulnerability来衡量某个权重的脆弱性来决定剪枝目标,为了保证原精度,正常样本的loss也会影响剪枝目标,二者做了个加权和。

About

merging some articles about network pruning

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published