Autoevals is a universal library for evaluating AI models, providing LLM-as-a-judge evaluations,...