Configuring a pipeline at runtime

Configuring a pipeline at runtime.

It is often the case that there is a need to run a pipeline with a different configuration. In this case, you should in most cases use the pipeline.with_options method. You can do this:

  1. Either by explicitly configuring options like with_options(steps="trainer": {"parameters": {"param1": 1}})

  2. Or by passing a YAML file using with_options(config_file="path_to_yaml_file").

You can learn more about these options here.

However, there is one exception: if you would like to trigger a pipeline from the client or another pipeline, you would need to pass the PipelineRunConfiguration object. Learn more about this here.

Last updated