All Products
Search
Document Center

Platform For AI:Periodically update online model services

Last Updated:Dec 13, 2024

Machine Learning Designer allows you to connect the Update EAS Service (Beta) component as a downstream node to a component that generates a model to update an online model service. Machine Learning Designer also allows you to submit a pipeline to DataWorks to achieve the periodical update of the online model service.

Prerequisites

A model generated by Machine Learning Designer is deployed to Elastic Algorithm Service (EAS) as an online model service and runs as expected. For more information, see Deploy a model as an online service.

Procedure

  1. Drag the Update EAS Service (Beta) component to the canvas and connect this component as a downstream node to a component that generates a model. Make sure that the model output port of the upstream component is directly connected to the input port of the Update EAS Service (Beta) component.

    The input port of the Update EAS Service (Beta) component can be connected to the path of a model that is stored in an Object Storage Service (OSS) bucket. For example, you can use Predictive Model Markup Language (PMML) models generated by machine learning algorithms or models generated by vision, text processing, and XGBoost training algorithms.image

  2. Click the Update EAS Service (Beta) component. In the panel that appears, configure the parameters on the Parameters Settings tab.image

    • EAS Service Name: the name of the deployed EAS service. The service is in the Running state.

    • EAS service description json: the JSON file that describes the service. In most cases, you can leave this parameter empty. If you want to modify other parameters, modify the parameters and enter the content that you modify in the code editor. For more information about how to modify the parameters, see Run commands to use the EASCMD client.

  3. Right-click the Update EAS Service (Beta) component and select Run Current Node. After the running is complete, an online model service is created.

  4. Update the online model service on a regular basis by using DataWorks.

    If you want the previous pipeline that implements model training and service updates to be periodically run, submit the pipeline to DataWorks and schedule it as a periodic task. For more information, see Use DataWorks tasks to schedule pipelines in Machine Learning Designer.

References

You can go to the Elastic Algorithm Service (EAS) page to view the statuses of the deployed model services or manage the model services. For more information, see Manage online model services in EAS.