Built-in metric evaluator prompts for model-as-a-judge evaluation jobs

When you run a judge-based model evaluation job the evaluator model scores the generator model's responses according to a set of metrics you define. Amazon Bedrock offers a number of built-in metrics, or you can define your own custom metrics.

Refer to the following sections to see the prompts used by the available evaluator models for the Amazon Bedrock built-in metrics.