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

No test accuracy improvement between baseline augmentation and searched augmentation #9

Open
akmalulkhairin opened this issue Sep 27, 2021 · 2 comments

Comments

@akmalulkhairin
Copy link

Hi, I have noticed that there is no noticeable improvement from the baseline test accuracy when using the policy searched by the code from the repo.

Baseline:
image

CIFAR10 searched policy:
image

Is this a common occurence?

@moskomule
Copy link
Owner

Hi, I recently got another report about this issue. I will investigate this issue.

Meanwhile, autoalbument by albumentations reproduces our work. I think they maintain better than us, so please check it.
https://github.com/albumentations-team/autoalbument-benchmarks

@akmalulkhairin
Copy link
Author

Hi, Thank you for the prompt reply and the recommendation, much obliged.

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

No branches or pull requests

2 participants