Interactive scalable auditing of model biases and vulnerabilities with interpretable mitigation - View it on GitHub