Dataset-Inference

This repository contains implementation of Ownership Resolution of Machine Learning Models using Dataset Inference

Steps for execution

1. Download Python 3.7.2 from here
2. Create a virtual environment with python 3.7 using the command: /path/to/python3 -m venv /path/to/directory/virtual_env_name
3. Install all the dependencies using the command: pip install -r requirements.txt
4. After all the dependencies are installed, there are different parts of execution:
   a) Train the victim model and the adversaary model: We have already trained the victim model for direct use. You can access pretrained models here.
   Download the pretrained models from Dataset Inference -> Pre-trained Models -> models. Place the folder in the root directory and proceed to the next part.

Acknowledgement

This project is built upon the foundations laid by Pratyush Maini , whose insightful work and implementation served as a valuable reference. While substantial modifications have been made to meet the specific requirements of this project, the inspiration and guidance drawn from hia original repository were instrumental in shaping its development. I extend my gratitude to Pratyush Maini for their contribution to the open-source community. To explore the original work, please visit their repository: Link to Original Repository.