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

DaNN的代码,直接跑的结果只有27%左右 #24

Closed
happywmzzz opened this issue Jul 4, 2018 · 4 comments
Closed

DaNN的代码,直接跑的结果只有27%左右 #24

happywmzzz opened this issue Jul 4, 2018 · 4 comments

Comments

@happywmzzz
Copy link

首先感谢这个非常好的project,对迁移学习和领域适应做了相当完善的汇总。

我下载了其中DaNN(Domain Adaptive Neural Network)的代码,本地跑了一下,环境一致,只把data的rootpath改变,最终效果只有27%,请问你复现的时候设置什么超参数吗?

@jindongwang
Copy link
Owner

我跑出来就跟文章中的精度差不多,参数就是这上面的。你把900次都跑完了吗?

@happywmzzz
Copy link
Author

我感觉是不是pytorch多卡训练?我这边机器是8张卡,你那边显卡数不同的话,batchsize是不是调整下?

@jindongwang
Copy link
Owner

有可能吧。。。我跑的时候只有一个k80.这是玄学。。。

@happywmzzz
Copy link
Author

那就有可能了,我把超参数调一下,应该可以出来效果的:)

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