FGSM Attack and Training in PyTorch

This is a Jupyter notebook implementation of FGSM Attack and Training on a simple network by utilizing the library CleverHans. Details of the implementation can be found on the medium article:

https://towardsdatascience.com/adversarial-attack-and-defense-on-neural-networks-in-pytorch-82b5bcd9171

Note: The library may have compatibility issues with Tensorflow, please downgrade Tensorflow to version 1.14