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.

Sign up for GitHub

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

Jump to bottom

Grad norm tensor fix #938

Merged
merged 3 commits into from
May 1, 2024
Merged

Grad norm tensor fix #938

merged 3 commits into from
May 1, 2024

Conversation

yeonsily
Copy link
Collaborator

@yeonsily yeonsily commented Apr 30, 2024 •
edited by regisss
Loading

What does this PR do?

Fixes #933

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?

@yeonsily yeonsily requested a review from libinta April 30, 2024 22:44
@yeonsily yeonsily requested a review from regisss as a code owner April 30, 2024 22:44
@HuggingFaceDocBuilderDev
Copy link

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.

@libinta libinta added the run-test Run CI for PRs from external contributors label Apr 30, 2024
@regisss
Copy link
Collaborator

regisss commented May 1, 2024

@yeonsily I pushed a new commit because _grad_norm.size() == 1 will always return False as we have _grad_norm.size() == torch.Size([1]) in lazy mode.
I also prefer not logging the grad norm if it is a list to not pollute the logs with a very big tensor.

@regisss regisss merged commit 051e5e5 into main May 1, 2024
9 checks passed
@regisss regisss deleted the grad_norm_tensor_fix branch May 1, 2024 18:36
@regisss regisss mentioned this pull request May 1, 2024
Closed
4 tasks
ccrhx4 pushed a commit to ccrhx4/ccrhx4.optimum-habana that referenced this pull request May 11, 2024
@yeonsily @regisss @ccrhx4
Grad norm tensor fix (huggingface#938)
023b292 
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
astachowiczhabana pushed a commit to HabanaAI/optimum-habana-fork that referenced this pull request May 15, 2024
@yeonsily @regisss @astachowiczhabana
Grad norm tensor fix (huggingface#938)
971a8bc 
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
astachowiczhabana added a commit to HabanaAI/optimum-habana-fork that referenced this pull request May 15, 2024
@astachowiczhabana @yeonsily @regisss
Grad norm tensor fix (huggingface#938) (#206)
b58bee2 
Co-authored-by: Yeonsil Yoon <yyoon@habana.ai>
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
This was referenced May 17, 2024
Merged
Merged
astachowiczhabana added a commit to HabanaAI/optimum-habana-fork that referenced this pull request Jun 13, 2024
@astachowiczhabana @yeonsily @regisss
Grad norm tensor fix (huggingface#938) (#206)
e3d2719 
Co-authored-by: Yeonsil Yoon <yyoon@habana.ai>
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
@astachowiczhabana astachowiczhabana mentioned this pull request Aug 5, 2024
Merged
3 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@regisss regisss regisss approved these changes

@libinta libinta libinta approved these changes

Assignees
No one assigned
Labels
run-test Run CI for PRs from external contributors
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

RuntimeError: a Tensor with 32768 elements cannot be converted to Scalar
4 participants
@yeonsily @HuggingFaceDocBuilderDev @regisss @libinta