Evaluation
      Back to home
      1. Knowledge Base for Magnet AI
      2. Administrator Manual
      3. Evaluation

      Test Sets

      Test Sets are sets of inputs sent to the LLM when an Evaluation Job is launched.

      They must include test input (like user questions for a RAG Tool or user input for Prompt Templates) and optionally can include expected output (like the most accurate answer to a user’s question or optimal output for a prompt). 

      So, the expected output provides the gold standard to measure the quality of generated output. 

       

      Magnet AI - Evaluation - Test Sets