Skip to main content Skip to complementary content
Qlik Resources
Announcements
Search
Loading SearchUnify's search If you need assistance with your product, please contact Qlik Support.
Qlik Customer Portal

Deploying models

You can deploy models from experiments in personal or shared spaces. Deploying a model adds it to a new or existing ML deployment. ML deployments can be created in personal, shared, and managed spaces. An ML deployment can contain multiple models, which can be used across different prediction workflows. The number of ML deployments is limited by the subscription.

The tier included in a Qlik Cloud subscription defines a maximum number of deployed models that can be created across all tenants created within the license. This consumption limit is defined per model, meaning that multiple ML deployments created from a single model count as a single deployed model. If you have reached the maximum number of deployed models, delete one or more existing deployed models or upgrade the subscription to a higher tier.

When you have deployed a model, the model is marked with a ML deployment icon in the ML experiment. This icon can be clicked to go directly to the ML deployment. You can also access the ML deployment from the catalog.

You can deploy a model from the Models, Data, or Analyze tab. You can choose to deploy a model to a new deployment or an existing deployment.

Requirements and permissions

To learn more about the user requirements for working with ML deployments and predictions, see Who can work with Qlik AutoML.

Deploying a model to a new ML deployment

As you deploy a model, you can create a new deployment, and add the model to that deployment.

From the Models tab

Do the following:

  1. In the Model metrics table, click Three-dot menu next to a model.

  2. Click ML deployment Deploy.

  3. Select the option to deploy the model to A new deployment.

  4. Enter a Name and select a Space. Optionally, you can add a description and tags.

  5. Optionally, select Enable real-time API access. This option is controlled by your subscription and enables predictions where the apply data is in the API request and the results are in the response.

  6. Click Create.

You should now be able to find your ML deployment in the catalog.

From the Data or Analyze tabs

Do the following:

  1. In the toolbar, use the drop down menu to select a model.

  2. Click Three-dot menu .

  3. Click ML deployment Deploy.

  4. Select the option to deploy the model to A new deployment.

  5. Enter a Name and select a Space. Optionally, you can edit the description and add tags.

  6. Optionally, select Enable real-time API access. This option is controlled by your subscription and enables predictions where the apply data is in the API request and the results are in the response.

  7. Click Create.

You should now be able to find your ML deployment in the catalog.

Deploying a model to an existing deployment

You can also deploy a model to a deployment that already exists. You can deploy models to ML deployments for which you have the required access.

From the Models tab

Do the following:

  1. In the Model metrics table, click Three-dot menu next to a model.

  2. Click ML deployment Deploy.

  3. Select the option to deploy the model to An existing deployment.

  4. Under Choose deployment, select a deployment using the drop-down menu.

  5. Optionally, you can edit the description.

  6. Optionally, select Enable real-time API access. This option is controlled by your subscription and enables predictions where the apply data is in the API request and the results are in the response.

  7. Click Deploy.

From the Data or Analyze tabs

Do the following:

  1. In the toolbar, use the drop down menu to select a model.

  2. Click Three-dot menu .

  3. Click ML deployment Deploy.

  4. Select the option to deploy the model to An existing deployment.

  5. Under Choose deployment, select a deployment using the drop-down menu.

  6. Optionally, you can edit the description.

  7. Optionally, select Enable real-time API access. This option is controlled by your subscription and enables predictions where the apply data is in the API request and the results are in the response.

  8. Click Deploy.

Removing models from an ML deployment

Over time, you might need to remove models from the deployment.

Do the following:

  1. In the ML deployment, open the Deployable models pane.

  2. Under All models in the deployment, click Three-dot menu next to the model and select Remove from deployment. In order to be able to remove a model from the deployment, the model needs to be unassigned from all aliases in the deployment.

  3. Click Save changes in the top right corner of the interface.

Viewing deployed model details

In the Deployable models pane, you can view details about a model that has been added to an ML deployment.

Do the following:

  1. In the ML deployment, open the Deployable models pane.

  2. Under All models in the deployment, click Three-dot menu next to the model and select Show model details.

The model details panel opens on the right, displaying information such as model features, description, and the experiment and training data.

Editing name and other details of ML deployments

Do the following:

  1. Open an ML deployment from the catalog.

  2. Click More by the deployment name.

  3. Click Edit Edit details.

  4. Edit the name or other details.

Deleting ML deployments

You can delete an ML deployment from the catalog.

Managing ML deployment jobs

Tenant admins can stop or cancel deployment jobs from the Administration activity center. For more information, see Administering Qlik AutoML.

Configuring notifications

You can receive notifications when a model is deployed from an experiment. For more information, see Configuring notifications for Qlik AutoML.

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – let us know how we can improve!

Leave your feedback here