search

MadryLab / cifar10_challenge

A challenge to explore adversarial robustness of neural networks on CIFAR10.
MIT License
488 stars 133 forks source link

pytorch definition of the model #26

Closed kartikgupta-at-anu closed 4 years ago

kartikgupta-at-anu commented 4 years ago

Is there a pytorch definition and pytorch model weights of the architecture used for the white board leaderboard? We would like to try our attack on your challenge but unfortunately our code is written in PyTorch.

If there is any other way, please let me know?

dtsip commented 4 years ago

We don't really have an equivalent PyTorch implementation (or plans to release one).

That being said, we recently released an adversarial training library in PyTorch: https://github.com/MadryLab/robustness . It has pretrained CIFAR models for L2 and Linf adversaries. This is not really a robustness challenge, but is probably still useful for comparing a new attack to PGD.