modelLatency

The expected model latency at maximum invocations per minute for the instance.