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

Can we get a set_max_jobs_per_cpu or similar? #66

Open
lucasb-eyer opened this issue Dec 1, 2024 · 1 comment
Open

Can we get a set_max_jobs_per_cpu or similar? #66

lucasb-eyer opened this issue Dec 1, 2024 · 1 comment

Comments

@lucasb-eyer
Copy link

hi, could we please have a way for ts to consider scheduling multiple jobs on a single GPU automatically?

I understand you didn't like the "free memory" heuristic because of job init time.

May I suggest a "set_max_jobs_per_cpu" kind of flag? If I know that my GPU can fit 3 of my jobs, then I can just set this and things will work, and I'll use my GPUs better.
@lucasb-eyer
Copy link
Author

Actually I just realized I can approximate this by, for example, splitting my queue into three queues with three ts running on three different sockets.

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
1 participant
@lucasb-eyer