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

Normalization after data aug #19

Merged
merged 15 commits into from May 1, 2022
Merged

Normalization after data aug #19

merged 15 commits into from May 1, 2022

Conversation

zaccharieramzi
Copy link
Collaborator

@zaccharieramzi zaccharieramzi commented May 1, 2022

  • Made sure normalization happens after data augmentation which is common practice (see for example here)
  • Increased batch size to 128
  • Allowed momentum without nesterov in SGD
  • Rescaled momentum for SGD tf

@zaccharieramzi
Copy link
Collaborator Author

@tomMoral the PR is still red because of this import, from a utils file.

I don't know how to circumvent this other than reimplementing the AugmentedDataset in the test which sounds a bit stupid.

Copy link
Member

@tomMoral tomMoral left a comment

Choose a reason for hiding this comment

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

LGTM!

The normalization that is applied in 3 differents spot is kind of ugly but I think this is good enough and we can improve this in the next iterations.

tests/test_dataset_consistency.py Outdated Show resolved Hide resolved
@zaccharieramzi
Copy link
Collaborator Author

TODO: get rid of the copy-paste

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.

None yet

2 participants