vision icon indicating copy to clipboard operation
vision copied to clipboard

Add 3-augment from DeiT III

Open trawler0 opened this issue 1 year ago • 5 comments

🚀 The feature

As the title suggest, add the data augmentation from https://arxiv.org/abs/2204.07118

Motivation, pitch

This seems to be a simple recipe with good results and the Deit family is widely recognized.

Alternatives

No response

Additional context

No response

trawler0 avatar Jul 06 '24 20:07 trawler0

Hi @trawler0, did you mean something like this?

Bhavay-2001 avatar Jul 18 '24 17:07 Bhavay-2001

Hey @Bhavay-2001! Yes, that's what I mean, I should have maybe directly attached the file. The recipe helped them to train very large ViT models from scratch on imagenet and they got amazing results.

trawler0 avatar Jul 19 '24 05:07 trawler0

Hi @NicolasHug, any views on this? Should I try to add this augmentation to torchvision?

Bhavay-2001 avatar Jul 19 '24 07:07 Bhavay-2001

Hi @trawler0 , thank you for feature request. Sure, I think this can be in scope of our augmentation strategies, even if ultimately the implementation will just be a Compose of a few building blocks.

@Bhavay-2001 thanks for offering your help. Let's see if @trawler0 would like to give this a go first, and if not then I'm happy for you to get to it. Thanks!

NicolasHug avatar Jul 25 '24 14:07 NicolasHug

Hi @NicolasHug I am happy to add this feature.

trawler0 avatar Jul 27 '24 14:07 trawler0