Skip to main content

What is H2O Eval Studio?

H2O Eval Studio is a modular and extensible studio for Retrieval-Augmented Generation (RAG) and Large Language Model (LLM) evaluation. It aims to provide a systematic assessment of RAGs/LLMs performance, reliability, security, fairness, and effectiveness in various applications. Eval Studio provides a collection of RAG/LLM evaluators to be used in RAG/LLM application development and operations.

H2O Eval Studio homepage​

The H2O Eval Studio homepage displays the most recent model hosts, tests, and evaluations. You can add model hosts and tests from this page by clicking the Add Model host and Add Test buttons.

Use the navigation menu on the sidebar to access the following pages:

  • Home: View the H2O Eval Studio dashboard.
  • Evaluations: Add, import, and view evaluations.
  • Workflows: Add and view workflows.
  • Tests: Add tests, view tests, or import test suites.
  • Model hosts: Add, view, and evaluate model hosts.
  • Evaluators: View available H2O Eval Studio evaluators.

You can also quickly start a new evaluation by clicking the New evaluation button. H2O Eval Studio homepage

Feedback