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

dep: cap transformers version due to FSDP bug #335

Merged
merged 1 commit into from
Sep 10, 2024

Conversation

anhuong
Copy link
Collaborator

@anhuong anhuong commented Sep 10, 2024

Description of the change

v0.34.0 and v0.34.1 of accelerate has a bug where when training with FSDP, the weight is being flattened and is unable to be loaded. Capping at v0.33 before the change went in and at a verified point.

Related issue number

How to verify the PR

Was the PR tested

  • I have added >=1 unit test(s) for every new method I have added.
  • I have ensured all unit tests pass

pyproject.toml Outdated
@@ -27,7 +27,7 @@ classifiers=[
]
dependencies = [
"numpy>=1.26.4,<2.0",
"accelerate>=0.20.3,<0.40",
"accelerate==0.33.0",
Copy link
Collaborator

Choose a reason for hiding this comment

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

Could we do

"accelerate>=0.20.3,<0.34"

- FSDP bug in accelerate v0.34

Signed-off-by: Anh Uong <anh.uong@ibm.com>
Copy link
Collaborator

@willmj willmj left a comment

Choose a reason for hiding this comment

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

LGTM!

@anhuong anhuong merged commit 553fa22 into foundation-model-stack:main Sep 10, 2024
7 checks passed
aluu317 pushed a commit to aluu317/fms-hf-tuning that referenced this pull request Sep 13, 2024
- FSDP bug in accelerate v0.34

Signed-off-by: Anh Uong <anh.uong@ibm.com>
Signed-off-by: Angel Luu <angel.luu@us.ibm.com>
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