[feature] Support for location arguments for `create_custom_training_job_from_component`

[faisalron]

Feature Area

Google Cloud Pipeline Component for custom job. Function name: create_custom_training_job_from_component

What feature would you like to see?

Currently, components created by create_custom_training_job_from_component will be set to the default 'us-central1' location and there is no way to change that.

What is the use case or pain point?

Sometimes we need to execute all of components of a pipeline on a specific region due to the network latency or just because the regulations. Unable to specify the location blocks this requirement.

Is there a workaround currently?

Currently, I need to define my component as container component and wrap it up with CustomTrainingJobOp to be able to set the location. https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.8.0/api/v1/custom_job.html#v1.custom_job.CustomTrainingJobOp

---

Love this idea? Give it a 👍.

[connor-mccarthy]

@faisalron, I can run successfully the single component in this pipeline in us-east1. Would you mind sharing your pipeline code?

from kfp import dsl
from google_cloud_pipeline_components.v1 import custom_job

@dsl.component
def print_foo():
    print('foo')

print_foo = custom_job.create_custom_training_job_from_component(print_foo)

@dsl.pipeline
def my_pipeline():
    print_foo(location='us-east1')

[faisalron]

@connor-mccarthy Thanks for checking in. I didn't know that we can set location using that argument.

print_foo(location="us-east-1") # I didn't know it generates location argument.

Do we have any documentation on that?

[connor-mccarthy]

@faisalron, for some reason project and location are omitted from the docs. I will look into that.

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[github-actions[bot]]

This issue has been automatically closed because it has not had recent activity. Please comment "/reopen" to reopen it.