|
This version is still in development and is not considered stable yet. For the latest stable version, please use Spring AI 2.1.0! |
Evaluation and Testing
The output of a model is not deterministic, so testing an AI application means evaluating answers as well as asserting on them. Spring AI helps in two ways.
Evaluating Responses
-
Evaluators describes the
Evaluatorinterface and the evaluators that ship with Spring AI, which check a response for relevancy and factual accuracy. -
LLM as a Judge shows how to use one model to score the answers of another.
Running Models and Stores Locally
-
Testcontainers provides service connections for integration tests.
-
Docker Compose Dev Services starts the services your application needs at development time.