Using the UI

The ML Pipelines UI allows you to complete tasks like creating, listing, and managing different resources, and makes it easy to view and compare the outputs of experiments. This guide discusses the different operations you can perform using the UI.

List Available Pipelines

The landing page of the UI is a list of all available pipelines, with management controls such as deleting, or creating a Job out of a selected pipeline.

Create a Job

You can select a pipeline from the pipelines list and click "Create job." This takes you to the Create a job page, where the job's information can be specified, and automatically fills in the selected pipeline. Enter the job's name (required), add an optional description, then choose the kind of trigger that should be used to run the job, filling out the trigger's details if any. The "Maximum concurrent runs" can be specified to limit the number of runs launched in parallel, which can be helpful if the pipeline is expected to run for a long period of time, and is triggered to run frequently.

The next section is the pipeline's parameters form. This form is generated automatically based on the selected pipeline (you can try out different pipelines from the top dropdown menu). After you fill in these parameters (some might be filled in automatically, if the pipeline author provides default values), you can now click the "Deploy" button to create a job out of this pipeline. This will take you to the job list page.

List Jobs

This page offers two views into the jobs scheduled in the cluster:

1. Group by job: This shows a list of jobs, and for each job, it shows the status of its last five runs, the name of the pipeline used to deploy it, the creation timestamp, its schedule if any, and whether its schedule is enabled or disabled.
2. Show all runs: This is a flat list of all runs in the system, showing each run's status, creation timestamp, and schedule timestamp. These two can be different in cases where the system is overloaded and scheduling a Kubernetes pod takes time until it actually starts.

List Job Details

The job's details page shows its configurations (description, parameter values.. etc), as well as a list of its scheduled runs. A job's configurations are shared among its runs.

Examine Run

Clicking a run in the run list (either inside a job, or in the top-level flat list of runs) takes you to view this run's details.