Add new models with custom args

MadryLab / robustness

A library for experimenting with, training and evaluating neural networks, with a focus on adversarial robustness.

MIT License

916 stars 181 forks source link

Add new models with custom args #47

Closed Hadisalman closed 4 years ago

Hadisalman commented 4 years ago

Solving this issue #40

1- moved custom_modules to robusntess/tools/ + fixed imports 2- replaced ImageNet resnet.py with the official PyTorch one (and added make_adv customs arguments) 3- Added the official PyTorch DenseNet ImageNet models 4- Fixed the existing CIFAR DenseNet model.