search

hendrycks / ss-ood

Self-Supervised Learning for OOD Detection (NeurIPS 2019)
MIT License
264 stars 31 forks source link

Question about Adversarial Training + Rotations #6

Closed inkawhich closed 4 years ago

inkawhich commented 4 years ago

Dan,

I think I know the answer to this question but I will ask it anyway. Do you compute the adversarial examples on the batch that contains all of the rotated versions of the data OR do you compute the adversarial examples on the 0deg rotation batch then compute the rotated versions? If it is the former, do you incur a ~4x increase in training time because the effective batch size is 4x bigger?

Thanks in advance, Nate

hendrycks commented 4 years ago

If it is the former, do you incur a ~4x increase in training time because the effective batch size is 4x bigger?

The training time is around the same because 4x the number examples per batch on CIFAR usually fits within memory.

hendrycks commented 4 years ago

I just added the adversarial files (though I have not tested the code in a while).