You can use deployment spaces to organize your deployments. This topic describes deployment spaces.
Viewing deployed assets
You configure and manage the deployment of a set of related assets in a space. A space contains an overview of deployment status, the deployable assets, deployments, associated input and output data, and the associated environments.
- To view the deployment space associated with a project, open the Settings page for the project and view the associated deployment space. Click the space name to open it and view details.
- To view all deployment spaces that you can access, click Deployment Spaces on the Watson Studio navigation menu.
Managing the default deployment space
Models created from notebooks and from canvas flows, such as SPSS models, are saved by default to a temporary deployment space. You must manually move these assets to a permanent deployment space.
- To view the default deployment space, click Deployment Spaces on the Watson Studio navigation menu, then click Temporary deployment space.
- To move an asset, choose Move from the action menu for the asset and choose a target deployment space.
- On the confirmation dialog, click Visit deployment space to open the space and work with the asset.
Creating a deployment space
- Click Deployment Spaces on the Watson Studio navigation menu.
- Click Add new deployment space
- Enter the details for the space.
A deployment space can only be associated with one project. If you have not already associated a deployment space with a project and you create an asset that requires a deployment space, you are prompted to create a new space or choose an existing space.
Promoting assets to a deployment space
Add assets to a deployment space in the following ways:
- When you are working with an asset, for example from the SPSS flow editor, you can promote an asset to the associated deployment space when you save the asset.
Save a machine learning model or Python function programmatically, using the Python client
- From the deployment space, choose Add asset and drop or upload an asset to the space.
Creating a new deployment
After saving a model to a deployment space, you can create a deployment in the space.
- Click the name of the saved model in the deployment space.
- Click the Deployments tab.
- Click New Deployment to create a deployment
- Fill out the specifics for the deployment.
Viewing deployment details
On the Assets page, you can view the assets in the deployment space. You can see how many deployments each asset has and whether the asset is configured for performance monitoring.
- Click an asset name to view details. The Overview page shows details such as the runtime environment and the training date.
- Click Evaluation to view performance metrics for the last evaluation.
- Click Deployments to view details for the deployments such as status and deployment ID.