GISKARD: OPEN-SOURCE TESTING FRAMEWORK FOR LLMS & ML MODELS (GITHUB REPO SPONSOR)

Summary from the TLDR AI newsletter

Giskard is a Python library [11] that automatically detects vulnerabilities of AI models, from tabular models to LLM, including: performance biases, data leakage, spurious correlation, hallucination, toxicity, security issues and many more. Get your model to production quickly and safely.