Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

关于FPGM和HRank #1

Open
bezorro opened this issue Sep 8, 2020 · 1 comment
Open

关于FPGM和HRank #1

bezorro opened this issue Sep 8, 2020 · 1 comment

Comments

@bezorro
Copy link

bezorro commented Sep 8, 2020

Hi,我也曾经是distiller的深度用户。
看你在这份代码增加了FPGM和HRank,想问一下这两个通道排序算法是否有效。
我个人也曾经尝试过BN-gamma[1] 、Taylor[2] 、L1[3]。
在我的大量实验下,排序效果:Taylor = L1 > BN-gamma。
想问一下,你有没有试过把FPGM和HRank和上述方法对比过呢?

[1] Learning Efficient Convolutional Networks Through Network Slimming
[1] Importance Estimation for Neural Network Pruning
[2] Pruning Filters for Efficient ConvNets

@BlossomingL
Copy link
Owner

BlossomingL commented Sep 29, 2020

刚看到,非常不好意思,最近一直研究您的Eagle Eye去了,页面一直停留在您的github,哈哈哈。
Q1: FPGM和HRank对比,我复现过两者的论文,都是在ImageNet上的ResNet系列,FPGM复现效果达到原论文;而HRank没有达到,求问HRank的作者感觉他也不是那么的乐意,可能是我的训练有问题吧。我也在自己目前的公司研发的模型和数据集上跑过这三个剪枝算法,结果是FPGM>L1Rank>HRank,对于FPGM一般剪大模型的一半不掉精度,剪轻量级的网络会掉5个点。
Q2: 您说的[1][2]我都没试验过。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants