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

Training related parameters question #1

Open
Zenobia7 opened this issue Apr 21, 2022 · 3 comments
Open

Training related parameters question #1

Zenobia7 opened this issue Apr 21, 2022 · 3 comments

Comments

@Zenobia7
Copy link

Hello, author. First of all, thank you very much for providing such a good code. I would like to ask, is the number of background pictures you used in training 22W of all the pictures in coco data set? Or only part of it? Then, the initial learning rate used in the paper is 1E-4, and the initial learning rate set in the code is 1E-3. Would you like to confirm which one is more appropriate?

@pgt4861
Copy link
Collaborator

pgt4861 commented Apr 23, 2022

Hi, Thank you for your interest.
We used all coco2014 training images as backgrounds and randomly sampled them when training.
Following the default training configurations in the code is recommended.

@Zenobia7
Copy link
Author

Zenobia7 commented May 9, 2022

Following the default training configurations in the code is recommended.

Thank you very much for your reply. I still have some problems with the setting of learning rate.When my iters exceeds warm up step, loss will not be able to converge. I tried two strategies (1) Step decrease directly from the initial learning rate, 2) gradually increase and then warm up step decrease gradually), such problems will occur. May I ask whether such problems will occur in your side?
(1)Step decrease directly from the initial learning rate:
46dd97768a00185855a3ff4791468c5
(2) gradually increase and then warm up step decrease gradually:
image

@mexicantexan
Copy link
Contributor

I also had this issue :/

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

3 participants