QPS (queries per second) refers to the number of query requests that a server can process per unit time. It is an important indicator for measuring system response capability and concurrent load, and is widely used in the performance evaluation of inference API services.
