Together Evaluations: Benchmark Models for Your Tasks
Together AI 1 year ago 49
Together AI released Together Evaluations, a platform for benchmarking large language models using other LLMs as judges to evaluate response quality. The platform supports three evaluation modes (classify, score, compare) and costs only the price of serverless inference with no additional evaluation fees. This enables developers to quickly compare models and validate performance on custom tasks without manual annotation or rigid metrics.