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

训练时将Conv+group换成DepthwiseConvolution更快些 #65

Open
jnulzl opened this issue May 9, 2018 · 3 comments
Open

训练时将Conv+group换成DepthwiseConvolution更快些 #65

jnulzl opened this issue May 9, 2018 · 3 comments

Comments

@jnulzl
Copy link

jnulzl commented May 9, 2018

@shicai 如题所示,我在你的基础上进行模型微调发现训练速度好慢,将Conv+group换成DepthwiseConvolution后,速度瞬间提升10倍左右!

@luuuyi
Copy link

luuuyi commented May 24, 2018

@jnulzl 精度会受影响么

@teresasun
Copy link

怎么换呢?能把prototxt共享一下吗?谢谢

@jiangwx
Copy link

jiangwx commented Sep 23, 2019

@shicai 如题所示,我在你的基础上进行模型微调发现训练速度好慢,将Conv+group换成DepthwiseConvolution后,速度瞬间提升10倍左右!

麻烦问下您使用的是哪个版本的caffe呢?

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

4 participants