ML Eval Framework

A lightweight evaluation harness for running offline assessments of large language model outputs. Supports custom metrics, batch evaluation, and structured output logging.

Replace this content with your actual project description.