Skip to main content
Version: Next

Generate a test's insights with a large language model (LLM)

Overview​

You can generate a test's insights with a large language model (LLM). After defining an LLM source, you can generate insights for a test.

Test insights

Instructions​

caution

By default, H2O Model Validation utilizes the h2oGPT LLM to generate insights for a dataset. To utilize your LLM to generate dataset insights, see Define a large language model (LLM) source.

To generate a test's insights, consider the following instructions:

  1. In the H2O Model Validation navigation menu, click Tests.
  2. In the Tests table, click the name of the test you want to generate its insights.
  3. Click the Insights tab.
    note
    • Right after clicking the Insights tab, H2O Model Validation generates and displays the insights of the test.
    • To regenerate the insights, click Regenerate.
Feedback