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

About experimental setup #2

Open
Hzj199 opened this issue Sep 25, 2023 · 0 comments
Open

About experimental setup #2

Hzj199 opened this issue Sep 25, 2023 · 0 comments

Comments

@Hzj199
Copy link

Hzj199 commented Sep 25, 2023

I trained with a batch size of 2048 for 32k steps, but the performance is still poor. The paper mentioned training on 80 A800 GPUs for two months. Approximately how many steps are needed to see significant improvements?

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

1 participant