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

Add option to stagger uploads based on local rank #3275

Merged
merged 14 commits into from
May 10, 2024

Conversation

dakinggg
Copy link
Contributor

@dakinggg dakinggg commented May 10, 2024

This PR adds support for adding a stagger based on local rank to the upload workers (i.e. the upload worker will wait by local rank * stagger). We are only exposing this via environment variable, as this is not something most users should be setting themselves.

Manual test:

  • stagger-upload-test-2-IggDEG (note that each ranks file was uploaded 1 minute apart)
Screenshot 2024-05-10 at 12 16 40 PM

@dakinggg dakinggg requested a review from irenedea May 10, 2024 19:19
@dakinggg dakinggg marked this pull request as ready for review May 10, 2024 19:19
@dakinggg dakinggg requested a review from a team as a code owner May 10, 2024 19:19
@dakinggg dakinggg changed the title Stagger upload 2 Add option to stagger uploads based on local rank May 10, 2024
composer/loggers/remote_uploader_downloader.py Outdated Show resolved Hide resolved
composer/loggers/remote_uploader_downloader.py Outdated Show resolved Hide resolved
@dakinggg dakinggg enabled auto-merge (squash) May 10, 2024 21:18
@dakinggg dakinggg merged commit 9d8f1c0 into mosaicml:dev May 10, 2024
15 checks passed
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

Successfully merging this pull request may close these issues.

3 participants