8000 Stars
An adversarial example library for constructing attacks, building defenses, and benchmarking both
Library for training machine learning models with privacy for training data
Some notes on things I find interesting and important.
(OLD REPO) Line-by-line profiling for Python - Current repo ->