RatingScale
The rating scale that defines how evaluators should score agent performance, supporting both numerical and categorical scales.
Contents
This data type is a UNION, so only one of the following members can be specified when used or returned.
- categorical
-
The categorical rating scale with named categories and definitions for qualitative evaluation.
Type: Array of CategoricalScaleDefinition objects
Required: No
- multiDimensional
-
A list of dimensions that lets an LLM-as-a-judge evaluator score several dimensions of the same target in a single evaluation. Each dimension has its own instruction and rating scale, and produces its own result.
Type: Array of EvaluationDimension objects
Array Members: Minimum number of 1 item. Maximum number of 50 items.
Required: No
- numerical
-
The numerical rating scale with defined score values and descriptions for quantitative evaluation.
Type: Array of NumericalScaleDefinition objects
Required: No
See Also
For more information about using this API in one of the language-specific Amazon SDKs, see the following: