Skip to content

This repository includes the code and data presented in the paper "Confronting Domain Shift in Trained Neural Networks". In NeurIPS 2020 Workshop on Pre-registration in Machine Learning (pp. 176-192). PMLR.

License

Notifications You must be signed in to change notification settings

sandialabs/confronting_domain_shift

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This repository includes the code and data presented in the paper, "Confronting Domain Shift in Trained Neural Networks", Martinez, C., Najera-Flores, D. A., Brink, A. R., Quinn, D. D., Chatzi, E., & Forrest, S. (2021, July). Confronting Domain Shift in Trained Neural Networks. In NeurIPS 2020 Workshop on Pre-registration in Machine Learning (pp. 176-192). PMLR.

The code was tested with Python 3.6.8 and dependencies (found in requirements.txt) were installed with pip 21.0.1.

Model and data configurations should be set with config files in .json format.  An example config file is found here: configs/example_train.json

To train a model:

python train.py configs/example_train.json

To run inference:

python infer.py configs/example_infer.json

Data used in the experiments for the paper are located in the ./data directory.


SAND2021-15138O

About

This repository includes the code and data presented in the paper "Confronting Domain Shift in Trained Neural Networks". In NeurIPS 2020 Workshop on Pre-registration in Machine Learning (pp. 176-192). PMLR.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages