Creating a Remote Training System

A Remote Training System is a connection to a unique party with data on a remote server. Each Federated Learning experiment requires at least one Remote Training System. Each party must set up their remote server before you can create a Remote Training System for them.

Cloud open beta

This is a Cloud open preview and is not supported for use in production environments.

Creating a Remote Training System

Note: You can create, view and delete Remote Training Sysytems in the Federated Learning UI only.

  1. From the Federated Learning UI, click Select remote training system tab.
  2. Click Add new systems.
  3. Fill out the required fields:
    • Name: A name to identify this specific instance of the Remote Training System.
    • Description: An optional description for more details of the training system. (Optional)
    • System administrator: The user who has access to run the current Federated Learning experiment.
    • Allowed users: Users who are allowed to interface with the current Federated Learning experiment training run. Note: For an administrator or user to show up they must be collaborators in your project.
    • Tags: Associate keywords with the Remote Training System to they can be found more easily. (Optional)
  4. Click on Add. The specific Remote Training System instance is then saved. Repeat steps 1-3 to create as many Remote Training System configurations as needed.
  5. When you are done, click on Add systems to saved them as assets in the Federated Learning system. The same information for a specific instance can be reused in any new Federated Learning experiments.