https://github.com/locuslab/intermediate_robustness
https://github.com/locuslab/intermediate_robustness
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/locuslab/intermediate_robustness
- Owner: locuslab
- Created: 2021-12-07T01:16:52.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2021-12-07T01:30:18.000Z (over 4 years ago)
- Last Synced: 2025-04-02T20:11:28.926Z (about 1 year ago)
- Language: Python
- Size: 16.6 KB
- Stars: 16
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Robustness between the worst and average case
*A repository that implements intermediate robustness training and evaluation from the NeurIPS 2021 paper [Robustness between the worst and average case](https://proceedings.neurips.cc/paper/2021/file/ea4c796cccfc3899b5f9ae2874237c20-Paper.pdf).
Created by [Leslie Rice](https://leslierice1.github.io/), [Anna Bair](https://annaebair.github.io/), [Huan Zhang](https://www.huan-zhang.com/) and [Zico Kolter](http://zicokolter.com).*
## Installation and usage
- To install all required packages run: `pip install -r requirements.txt`.
- Pretrained model weights can be downloaded [here](https://drive.google.com/drive/folders/1YCFXzdx2dGjmQGU30v6CRhUORjQsjHKV?usp=sharing).
- To train (l_infty perturbations), run `python train.py -c {path_to_training_config_file}.json`.
- To evaluate (l_infty perturbations), run `python train.py -c {path_to_evaluation_config_file}.json`.
- To train (spatial transformations), run `python train_discrete.py -c {path_to_training_config_file}.json`.
- To evaluate (spatial transformations), run `python eval_discrete.py --checkpoint {path_to_model_checkpoint}.pth`.