Eval Studio

Eval Studio is a Python package for the introspection of machine learning models by enabling various facets of Responsible AI.

LLMs Evaluations

Evaluations with model explanations created by evaluators:

Evaluation 277fabca-b4f5-49a1-8602-5a5638e9a000 (2026-01-22 17:07:07 T+0100)
- Dataset : <class 'h2o_sonar.lib.api.datasets._datasets_genai.LlmDataset'>
- Explainer(s) : 7
Evaluation 25269fb8-8cb5-4879-be2f-3570da92adc1 (2026-01-22 17:05:35 T+0100)
- Dataset : <class 'h2o_sonar.lib.api.datasets._datasets_genai.LlmDataset'>
- Explainer(s) : 7

Eval Studio log

Eval Studio log with library runtime messages:

h2o-sonar.log

Please not that every evaluator has its own log and evaluator messages are not logged to the library log by default (check configuration documentation for more details).