Using ML Runtimes with cdswctl

If a project is configured to use Runtimes, cdswctl workflows for starting sessions or models are slightly different.

Querying the engine type
You can query whether a project is configured using ML Runtimes or Legacy Engine.
Listing runtimes
The first step to working with projects using runtimes is to query the available runtimes using the cdswctl runtimes list command.
Starting sessions and creating SSH endpoints
Once you choose a runtime, you can start a session using the cdswctl sessions start command and create SSH endpoints using the cdswctl ssh-endpoint command.
Creating a model
Creating a model in a project that uses runtimes is similar to model creation with an legacy engine, but you must use a different parameter to specify the runtime ID.

Parent topic: Cloudera Data Science Workbench Command Line Reference