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

Sort callbacks so that CheckpointSaver goes before HuggingFaceCheckpointer #1515

Merged
merged 5 commits into from
Sep 10, 2024

Conversation

irenedea
Copy link
Contributor

@irenedea irenedea commented Sep 9, 2024

Sort the callbacks so CheckpointSaver goes before HuggingFaceCheckpointer . This saves time during final checkpoint saving and uploading.

A rough sketch on the reordering based on Llama 70B profiling:
Before
image

After
image

e2e run in staging: ift-meta-llama-3-8b-aoprwt-reordered-q01HKT

@irenedea irenedea requested a review from a team as a code owner September 9, 2024 18:51
Copy link
Collaborator

@mvpatel2000 mvpatel2000 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

test?

llmfoundry/command_utils/train.py Outdated Show resolved Hide resolved
@irenedea
Copy link
Contributor Author

irenedea commented Sep 9, 2024

@mvpatel2000 What tests would you like to see? I can add a test for the helper function that checks to see that the callbacks are in the expected order. Anything else?

@mvpatel2000
Copy link
Collaborator

@mvpatel2000 What tests would you like to see? I can add a test for the helper function that checks to see that the callbacks are in the expected order. Anything else?

I think that's fine!

@irenedea irenedea enabled auto-merge (squash) September 10, 2024 17:06
@irenedea irenedea disabled auto-merge September 10, 2024 17:06
@irenedea irenedea changed the title Sort callbacks with checkpoint saving and uploading first Sort callbacks so that CheckpointSaver goes before HuggingFaceCheckpointer Sep 10, 2024
@irenedea irenedea merged commit 8a8de18 into mosaicml:main Sep 10, 2024
9 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.

2 participants