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

Improve config handling and add a zoo #3029

Merged
merged 4 commits into from
Aug 20, 2024
Merged

Improve config handling and add a zoo #3029

merged 4 commits into from
Aug 20, 2024

Conversation

muellerzr
Copy link
Collaborator

What does this PR do?

This PR introduces a config zoo to the repo, showcasing a number of "template" config files users can use.

This PR also tweaks how the config yamls are read, so we can simplify certain configurations better, relying on inferred (but valid) parameters, which our CLI already does.

E.g. this is a full, complete, valid config yaml for single-gpu bf16 training:

distributed_type: "NO"
mixed_precision: "bf16"

Fixes # (issue)

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you read the contributor guideline,
    Pull Request section?
  • Was this discussed/approved via a Github issue or the forum? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes? Here are the
    documentation guidelines, and
    here are tips on formatting docstrings.
  • Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@SunMarc

@muellerzr muellerzr requested a review from SunMarc August 20, 2024 13:10
@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

Copy link
Member

@SunMarc SunMarc left a comment

Choose a reason for hiding this comment

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

Nice ! Thanks for adding these ! Just a nit

Comment on lines 1340 to 1344
sync_module_states: bool = field(
default=False,
default=None,
metadata={
"help": "Whether each individually wrapped FSDP unit should broadcast module parameters from rank 0 "
"to ensure they are the same across all ranks after initialization. Defaults to `True`"
Copy link
Member

Choose a reason for hiding this comment

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

In the post_init, I can see that actually the default is False. Either change the description or the init below. Also thanks for fixing this !

@muellerzr muellerzr merged commit 1a6af0b into main Aug 20, 2024
28 checks passed
@muellerzr muellerzr deleted the config_zoo branch August 20, 2024 14:40
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