LM Evaluation Harness

A standardized evaluation framework for large language models open-sourced by EleutherAI. It supports multiple benchmark datasets with transparent and reproducible processes, and is widely used for academic and community model evaluation.