Model Catalog

Registering Models and Editing Metadata

Note

1. This feature requires PostgreSQL to be installed.

2. This demo expects that there is no existing model with the same URI or the same combination of Model Name and Version. The model catalog enforces uniqueness on these fields. Requests that result in a conflict will be rejected.

Register New Model

1. Select the Model Catalog tab at the top of the page.

2. Click Register a new model to open the registration modal.

3. Enter the following parameters:

Note

Name, uri, and version must be unique. If these values already exist in the model catalog and the model is in use, consider deleting the model everywhere it is currently running before continuing with this demonstration, or use your own values.

Seldon Core V2Seldon Core V1

Field	Value
Model Name	cifar10
URI	gs://seldon-models/triton/tf_cifar10
Artifact Type	Tensorflow
Version	v1.0
Task Type	Classification
Project	seldon

4. Append tags and metric parameters with the following values. To add more parameters, use the + button. To delete a parameter, use the x button:

   • Tags:

     Key	Value
     author	Seldon
     training_set	Resnet32

   • Metrics:

     Key	Value
     p1	0.8
     p2	0.6

5. Click Register Model.

Expand to see model details

Deploying Models From The Model Catalog

Note

This demonstration requires the cifar10 model to have been registered already.

1. Select the Model Catalog tab at the top of the page.

2. Click on the vertical ellipses “⋮” icon for the model named cifar10.

3. In the dropdown menu that appears, select Deploy.

4. In the wizard which appears, enter the following details:

Seldon Core V2Seldon Core V1

Field	Value
Name	cifar10
Namespace	seldon
Type	Seldon ML Pipeline
Protocol	seldon

5. Click Next steps, then click Launch.

Select the Overview header tab to verify that your model has been deployed.

Edit Model Metadata

Warning

• Changing the prediction schema will require a new model to be registered.

• You can click on the vertical ellipses ⋮ button to clone a model and set a new prediction schema with a different version.

Note

Name, Version, Tags, and Metrics can be changed at any time.

From The Model Catalog

1. From the Overview page, select Model Catalog at the top of the page.

2. Select a model.

3. On the side drawer that opens, click EDIT METADATA.

4. Add a new tag with the following values:

   Key	Value
   stage	production

5. Click SAVE METADATA at the top right hand side of the side drawer to save your edit.

From A Deployment’s Dashboard

1. From the Overview page, select a pipeline or deployment.

2. In the dashboard, click on the model within Pipeline Components (or Deployment Components for Core V1). This should open a drawer on the right hand side.

3. Click EDIT METADATA.

4. Under the Tags section, press the + button to add a new tag.

5. For the new tag, enter the following parameters:

   Key	Value
   stage	production

6. Click SAVE METADATA on the top right of the side drawer.