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

Study and impl MaxText's per-host dataloading #23

Open
tengyifei opened this issue Jan 9, 2025 · 0 comments
Open

Study and impl MaxText's per-host dataloading #23

tengyifei opened this issue Jan 9, 2025 · 0 comments
Assignees
@bhavya01

Comments

@tengyifei
Copy link
Collaborator

In MaxText, each host only downloads/processes the slice of data that it's responsible for. This is a further optimization over "minibatch" dataloading, which configures each host to load into CPU the slice of data it's responsible for.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@bhavya01 bhavya01

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@tengyifei @bhavya01