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.

Sign up for GitHub

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

Jump to bottom

With a learning rate of 0.1, loss is 'nan'. I don't kown why? #2

Open
Hedlen opened this issue Jun 13, 2018 · 2 comments
Open

With a learning rate of 0.1, loss is 'nan'. I don't kown why? #2

Hedlen opened this issue Jun 13, 2018 · 2 comments

Comments

@Hedlen
Copy link

Hedlen commented Jun 13, 2018

When using your program，with a learning rate of 0.1, loss is ‘nan’, when I change the learning rate to 0.001, loss changes normally. I don't know why ? According to your documentation, lr 0.1 should be fine.
I look forward to your reply. Thanks.
@BobLiu20
Copy link
Owner

hi. how much your label_num (class number)? I had tried 10K and 700K with lr=0.1 and everything is ok.

@Hedlen
Copy link
Author

Hedlen commented Jun 19, 2018

@BobLiu20 Hi, My label number is 10575, the dataset is CASIA-Webface. Thanks

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@BobLiu20 @Hedlen