24bc9162a5
* feat(sdk): Support launching Dataflow Flex templates from Kubeflow pipelines step * Update readme with permissions to run dataflow job Change public BigQuery dataset usage to custom user supplied BigQuery table. The public dataset access doesn't work because you can't pass in the project as a job parameter to this specific job. * Remove validate_only parameter from the component.yaml. This adds less value in my opinion. Incorporate PR feedback to directly use `kfp.Client().create_run_from_pipeline_func` function instead of compiling and creating the run separately. * Fix formatting |
||
|---|---|---|
| .. | ||
| core | ||
| launcher | ||
| __init__.py | ||