Defines the resources used to perform model inference.
The number of instances to use.
The type of instance that is used to perform model inference.