Protectai ModelScan - a demonstration of scanning ML models for malicious code
This repo shows how to use ModelScan to detect malicious code hidden in machine-learning model files.
- 'download_model.ipynb' - download Hugging Face model (This file must be run BEFORE creating the closed container.)
modelscan_demo.ipynb– step-by-step walkthroughmodels/– sample safe and malicious pickle files