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

[hotfix] Init typo #229

Merged
merged 1 commit into from
Mar 8, 2022
Merged

[hotfix] Init typo #229

merged 1 commit into from
Mar 8, 2022

Conversation

blefaudeux
Copy link
Contributor

@blefaudeux blefaudeux commented Mar 7, 2022

What does this PR do?

I don't think that fixes #219, but related (init not properly done). Caught when implementing #227, I think that we would need to add a proper explicit/programable init to the model factory to rule out any issue there.

In the meantime this must have been a typo, I'm not sure when it slipped in but does not make a lot of sense (biases are typically zero-init but linear weights are one-init for instance, this line would nuke that). This PR makes sure that by default the init values defined by the components are preserved

cc @jramapuram

Before submitting

  • Did you have fun?
    • Make sure you had fun coding 🙃
  • Did you read the contributor guideline?
  • Was this discussed/approved via a Github issue? (no need for typos, doc improvements)
    • N/A
  • Did you make sure to update the docs?
    • N/A
  • Did you write any new necessary tests?
    • N/A
  • Did you update the changelog? (if needed)
    • N/A

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Mar 7, 2022
@blefaudeux blefaudeux requested a review from dianaml0 March 7, 2022 23:13
@codecov-commenter
Copy link

Codecov Report

Merging #229 (80f4280) into main (dc80aeb) will decrease coverage by 0.01%.
The diff coverage is n/a.

Impacted file tree graph

@@            Coverage Diff             @@
##             main     #229      +/-   ##
==========================================
- Coverage   91.72%   91.70%   -0.02%     
==========================================
  Files          60       60              
  Lines        3214     3208       -6     
==========================================
- Hits         2948     2942       -6     
  Misses        266      266              
Flag Coverage Δ
Python 91.70% <ø> (-0.02%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Δ
xformers/factory/model_factory.py 97.67% <ø> (-0.16%) ⬇️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update dc80aeb...80f4280. Read the comment docs.

@blefaudeux
Copy link
Contributor Author

landing, I hope that's ok @dianaml0, hotfix of sorts

@blefaudeux blefaudeux changed the title [Minor] Init typo [hotfix] Init typo Mar 8, 2022
@blefaudeux blefaudeux merged commit f5c1d01 into main Mar 8, 2022
@blefaudeux blefaudeux deleted the issue_219_typo branch March 8, 2022 04:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

xformers ViT-B ImageNet MAE + Deepnorm training instability
3 participants