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

CUDA_DEVICE_MAX_CONNECTIONS #236

Open
jeromeku opened this issue Oct 30, 2024 · 1 comment
Open

CUDA_DEVICE_MAX_CONNECTIONS #236

jeromeku opened this issue Oct 30, 2024 · 1 comment

Comments

@jeromeku
Copy link

Noticed that for some examples, this environment variable CUDA_DEVICE_MAX_CONNECTIONS is explicitly set to 1. Why is that?

I've seen some explanations that indicate this is necessary for overlapping comms / compute when using TP / SP but others that would suggest this in fact interferes with comms / compute overlap when using ZeRO-DP / FSDP.
@NouamaneTazi
Copy link
Member

It's exactly as you said. We're working on removing it once we have better streams management!

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
@jeromeku @NouamaneTazi