Hosted inference is a model inference hosting service provided by cloud service providers. Users do not need to build a local inference environment, but only need to upload the trained model to the cloud platform, and then perform efficient inference through the interface provided by the platform, reducing the technical threshold of inference deployment.
