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

Pass Correct LayerNorm Epsilon value to TransformerEncoder in Backbones #731

Merged
merged 1 commit into from
Feb 15, 2023

Conversation

TheAthleticCoder
Copy link
Contributor

@TheAthleticCoder TheAthleticCoder commented Feb 8, 2023

Resolves #728

@abheesht17
Copy link
Collaborator

abheesht17 commented Feb 8, 2023

LGTM! Could you please change the description to "Resolves #728" (without the word "issue")? That way, GitHub automatically marks this PR as the one which will close the mentioned issue.

@TheAthleticCoder TheAthleticCoder changed the title Added layer normalization Pass Correct LayerNorm eps value to TransformerEncoder in Backbones Feb 8, 2023
@TheAthleticCoder TheAthleticCoder changed the title Pass Correct LayerNorm eps value to TransformerEncoder in Backbones Pass Correct LayerNorm Epsilon value to TransformerEncoder in Backbones Feb 8, 2023
@TheAthleticCoder
Copy link
Contributor Author

@mattdangerw A gentle reminder regarding the commit

@mattdangerw
Copy link
Member

@TheAthleticCoder thanks! This looks good, but would it be possible to drop a comment here with links to where you found these constants? Would be a useful reference anyway.

@TheAthleticCoder
Copy link
Contributor Author

@mattdangerw The values of the constants were already there in the model files (model_backbone.py). I used those pre-existing constants and passed them into the main model architecture.

@mattdangerw
Copy link
Member

Ah got it! Thank you.

@mattdangerw mattdangerw merged commit d44bc3c into keras-team:master Feb 15, 2023
@TheAthleticCoder TheAthleticCoder deleted the issue728 branch March 23, 2023 07:01
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.

Incorrect epsilon (LayerNorm) for the Transformer Layer in All Backbones
3 participants