Model Store API

Note

This document has been machine translated.

This section describes the Model Store API.

Model Store

xData Edge stores and manages machine learning models in a model store.

A model store is a dataset with fixed column definitions that is specialized for storing models. Each model stored in the model store corresponds to a record stored in the dataset.

xData Edge users can store model data in the model store along with meta-information. Model data is arbitrary binary data. Meta information can be any JSON-formatted data, and can record information such as precision, loss, etc., obtained during the model generation process.

Model Store API List

xData Edge users can use the Model Store API to perform the following operations.

API Methods	Description
model_store.create	Create a new model store.
model_store.delete	Delete a model store.
model_store.model.put	Register a model in the model store.
model_store.model.put_aggregated	Register an aggregated model in the model store.
model_store.model.delete	Delete a model from the model store.
model_store.model.get_info	Retrieve information of a model registered in the model store.
model_store.model.get	Get a record in the model store by specifying a model ID.