@open-evals/core

Core evaluation framework for testing and measuring LLM application performance.

Installation

pnpm add @open-evals/core

Features

• Datasets: Create and manage evaluation datasets with single-turn or multi-turn samples
• Metrics: Define custom metrics or use built-in ones for evaluation
• Evaluation Engine: Run evaluations with configurable concurrency and error handling
• Type Safety: Full TypeScript support with comprehensive type definitions