Skip to content

Code for ICML AdvML paper "Shrink & Cert: Bi-level Optimization for Certified Robustness"

Notifications You must be signed in to change notification settings

sagarverma/Shrink-and-Cert

Repository files navigation

Shrink & Cert: Bi-level Optimization for Certified Robustness

Getting started

Let's start by installing all the dependencies.

pip3 install -r requirement.txt

Training

python3 train.py --arch cifar_large_model --exp-mode shrink --configs configs/CIFAR10/sac.yml --k 0.20 --seed 1234 --exp-name CIFAR10_cifar_large_model_Unstructured_BiC_K0.01

python3 train.py --arch resnet110 --exp-mode shrink --configs configs/CIFAR10/sac.yml --k 0.20 --seed 1234 --exp-name CIFAR10_resnet110_Unstructured_BiC_K0.01

python3 train.py --arch resnet50 --exp-mode shrink --configs configs/ImageNet/sac.yml --k 0.20 --seed 1234 --exp-name ImageNet_resnet50_Unstructured_BiC_K0.01

About

Code for ICML AdvML paper "Shrink & Cert: Bi-level Optimization for Certified Robustness"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published