llm-benchmarks

A benchmarking tool for comparing large language models. It gives standard metrics and a test framework for judging models on speed, accuracy and resource use.