adversarial_robustness_pytorch
adversarial_robustness_pytorch copied to clipboard
imrahulr
Unofficial implementation of the DeepMind papers "Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples" & "Fixing Data Augmentation to Improve Adversarial Robustness...
Dear author, thanks for your PyTorch implementation. However, the missing args.txt prevents me from testing the pre-trained model provided by [https://github.com/deepmind/deepmind-research/tree/master/adversarial_robustness](url). Thank you for your help! https://github.com/imrahulr/adversarial_robustness_pytorch/blob/6df6a8f0cd49cf6d18507a4b574c004ab6eedf49/eval-aa.py#L35
Great implementation, it will be even more perfect if adding resume options. haha
