entangled-watermark-torch

This repository is an unofficial pytorch-based implementation of the paper Entangled Watermarks as a Defense against Model Extraction.

We test our code on five datasets: MNIST, Fashion-MNIST, CIFAR-10, and CIFAR-100.

Dependency

Our code is implemented and tested on Pytorch. Just use the following command to install the dependencies.

pip install -r requirements.txt

Preprocessing and Training

Basically the process and command to train is the same as the original one. For details, Please refer to Official repo.

Questions or suggestions

If you have any questions or suggestions, pls send me an email at yc17486@umac.mo