inferenceSpecification

Defines how to perform inference generation after a training job is run.