Skip to content

reverting disable FusedRMSNorm for training (#343) since rmsnorm issu…#449

Merged
mandy-li merged 1 commit into
mainfrom
revert_fusedrmsnorm_wrkarnd
Oct 10, 2023
Merged

reverting disable FusedRMSNorm for training (#343) since rmsnorm issu…#449
mandy-li merged 1 commit into
mainfrom
revert_fusedrmsnorm_wrkarnd

Conversation

@hsubramony
Copy link
Copy Markdown
Collaborator

…e is fixed in 1.12

What does this PR do?

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 make sure to update the documentation with your changes?
  • Did you write any new necessary tests?

@hsubramony hsubramony requested a review from a user October 4, 2023 15:57
@HuggingFaceDocBuilderDev
Copy link
Copy Markdown

HuggingFaceDocBuilderDev commented Oct 4, 2023

The documentation is not available anymore as the PR was closed or merged.

Copy link
Copy Markdown
Collaborator

@regisss regisss left a comment

Choose a reason for hiding this comment

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

LGTM
We should also do it for T5:

if not self.training and hidden_states.device.type == "hpu" and FusedRMSNorm:

Copy link
Copy Markdown
Collaborator

@mandy-li mandy-li left a comment

Choose a reason for hiding this comment

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

LGTM

@mandy-li mandy-li merged commit 96cdcb4 into main Oct 10, 2023
@regisss regisss deleted the revert_fusedrmsnorm_wrkarnd branch October 11, 2023 08:27
@regisss
Copy link
Copy Markdown
Collaborator

regisss commented Oct 11, 2023

@hsubramony Can you also open a PR to apply the same change to T5 please?

gplutop7 pushed a commit to HabanaAI/optimum-habana-fork that referenced this pull request Oct 15, 2025
…ace#449)

Co-authored-by: Piotr Bielak <pbielak@users.noreply.github.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.

4 participants