The experiments are run in Colab Notebooks.

The files are:

1. adlproj_pytorch_base_cnns.ipynb: Standard CNN models on MNIST and FashionMNIST trained in this.
2. adlproj_base_exp.ipynb: Reproducibility experiment
3. adlproj_transf_lrn.ipynb: Training ResNet for transfer learning
4. adlproj_adv_attack.ipynb: Conducting and saving adversarial attack examples
5. adlproj_captum_exps.ipynb: Using Captum to score adversarial attacks and attempts applied on scoring transfer learning

Additional saved models, plots and scores in respective directories.