Add users to the Watson Machine Learning Accelerator instance from IBM Cloud Pak for Data

Edit online

As an administrator or creator of a Watson Machine Learning Accelerator instance, you can give other IBM® Cloud Pak for Data users access to the instance. Users that are added to the instance can view the Watson Machine Learning Accelerator service console.

Before you begin

To give another user access to the Watson Machine Learning Accelerator service console, ensure that:

  • You are an administrator of that Watson Machine Learning Accelerator instance. Starting in 2.3.7, you can assign specific platform users the Watson Machine Learning Accelerator Administrator role. In earlier refreshes, only users who had the Administer platform permission on Cloud Pak for Data could administer the Watson Machine Learning Accelerator service.

  • That the user exists on Cloud Pak for Data.

    To create a Cloud Pak for Data user, you must be a Cloud Pak for Data administrator. Follow these steps:
    1. Open the Cloud Pak for Data main menu.
    2. Select Administration > User management. You see a list of the existing Cloud Pak for Data users.
    3. Click the New user button above the list.
    4. Provide the details of the new user and click Create when finished.

About this task

A user with access to a Watson Machine Learning Accelerator instance can have the following access privileges to the instance. The privileges are grouped in sets called roles.

The privileges associated with these roles are associated with the basic roles on Cloud Pak for Data. A Cloud Pak for Data administrator is a Watson Machine Learning Accelerator administrator and all other Cloud Pak for Data users are a Watson Machine Learning Accelerator service administrator or user. A Watson Machine Learning Accelerator administrator can assign a Cloud Pak for Data user as an administrator of the Watson Machine Learning Accelerator service.

If a Cloud Pak for Data administrator or user has not been added to the instance, they cannot view the service console. Both the Cloud Pak for Data administrator and user can use the Watson Machine Learning Accelerator REST API.

Service roles and privileges
Admin
A Watson Machine Learning Accelerator service admin can view and use all features of the service console, including:
  • Workload
  • Monitoring
  • Resource Plan
Note:
  • A service admin must be a Cloud Pak for Data platform administrator or user.
  • If the service admin is also a Cloud Pak for Data platform administrator they can view details of the instance and manage the instance from the Cloud Pak for Data console.
User or Developer
A Watson Machine Learning Accelerator service user or developer can view some service console functionality, including:
  • Workload
  • Monitoring
Note:
  • A service user must be a Cloud Pak for Data user. From the Cloud Pak for Data console, they are able to view the details of the Watson Machine Learning Accelerator instance.
  • A user or developer cannot create or delete instances. Neither can an instance user give other users access to the instance or change their roles.
No service role defined
A Cloud Pak for Data administrator or user does not have access to the Watson Machine Learning Accelerator console. Deep learning operations can be completed using the Watson Machine Learning Accelerator REST API as follows:
  • A Cloud Pak for Data administrator can uses the REST API for Resource Plan, Workloads and model deployment for all users.
  • A Cloud Pak for Data user can uses the REST API for workloads and their own model deployment.

Procedure

  1. Open the IBM Cloud Pak for Data main menu.
  2. From the main menu, go to Services > Instances.
  3. Find the instance name where Watson Machine Learning Accelerator is deployed, click Vertical menu icon and select Manage access.
  4. Click Add users.
  5. Select the user from the list and select a role, either Admin, User or Developer.
  6. Click Add.

What to do next

The user now has access to the Watson Machine Learning Accelerator instance and service console information.