[sdk] Unable to use custom params when using kfp.Client().run_pipeline()
Environment
- KFP version:
- KFP SDK version:
kfp==2.0.0b13
- All dependencies version:
kfp 2.0.0b13
kfp-pipeline-spec 0.2.0
kfp-server-api 2.0.0a6
Steps to reproduce
client = kfp.Client()
run = client.run_pipeline(
experiment_id=experiment_id,
job_name=run_name,
params=params,
pipeline_id=pipeline_id,
version_id=version_id,
)
pipeline runs with default params and not custom ones specified below:
tried 2 formats for params:
- {"x": "206", "seed": "3007"}
- {"parameters": {"x": "206", "seed": "3007"}}
Expected result
Pipeline should run with custom params
Materials and Reference
Followed docs: https://kubeflow-pipelines.readthedocs.io/en/2.0.0b13/source/client.html#kfp.client.Client.run_pipeline
Impacted by this bug? Give it a 👍.
/assign @chensun
Hi @connor-mccarthy has there been any update on this ?
@arun-vc, can you try upgrading to the latest version of the KFP SDK (2.0.0-rc.2) and KFP BE (2.0.0-rc.1) and see if this error persists?
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.
I am also seeing this, not being able to use custom parameters is a pretty big deal!!
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.
This issue has been automatically closed because it has not had recent activity. Please comment "/reopen" to reopen it.
Same issue on latest kfp version