MindSpore Serving Documents
MindSpore Serving is a lightweight and high-performance service module that helps MindSpore developers efficiently deploy online inference services in the production environment. After completing model training on MindSpore, you can export the MindSpore model and use MindSpore Serving to create an inference service for the model.
MindSpore Serving provides the following functions:
Customization of model preprocessing and postprocessing, simplifying model release and deployment
The batch function used to split and combine requests containing multiple instances to meet the batch size requirements of the model
Distributed model inference
The gRPC APIs and easy-to-use Python encapsulation APIs on the client
The RESTful APIs on the client
Typical MindSpore Serving Application Scenarios
-
Use a simple Add network as an example to describe how to use MindSpore Serving to deploy an inference service.
Using the gRPC APIs to Access Services
Easily access services with high performance.
Using the RESTful APIs to Access Services
Access services based on HTTP.
- MindSpore Serving-based Inference Service Deployment
- MindSpore Serving-based Distributed Inference Service Deployment
- gRPC-based MindSpore Serving Access
- RESTful-based MindSpore Serving Access
- Servable Provided Through Model Configuration
- Service Deployment with Multiple Subgraphs and Stateful Model