Serve Web Workflow
Last updated
Last updated
VESSL offers a web-based workflow for managing model services. This workflow is designed to be intuitive and easy to use, allowing users to quickly create and manage model services.
To establish a new model service, navigate to the Servings tab and initiate the process by clicking the "New serving" button.
Once the "New serving" button is selected, you will be directed to the "Create a new serving" page, dedicated to the configuration of a fresh Serving instance. Within this page, it is possible to define essential parameters for the serving:
Name: Assign a distinctive name to the Serving. (Please note that the assigned name cannot be altered after creation.)
Description: Elaborate on the nature and purpose of the Serving for further clarity.
By selecting any of the Servings created within the Serving list, you can access details about the Serving on the local navigation bar.
Each Serving instance offers following menu options:
Cluster: Designate the specific cluster where the inference server will be instantiated.
Monitor: which provides a overview of the Serving's current status and performance.
Logs: A of the Serving's revisions and replicas.
Revisions: Create and manage of the Serving.
Rollouts: Define and run no-downtime of the Serving.
