Since kedro 0.17.7(?) there have been introduced namespaces which cause issues in kfp artifacts, as they are not properly handled.
Unless the function to create kfp artifacts is disabled in kubeflow.yaml config:
store_kedro_outputs_as_kfp_artifacts: False
It causes issues like:
ValueError: Only letters, numbers, spaces, "_", and "-" are allowed in name. Must begin with a letter. Got name: data_science.active_modelling_pipeline.X_train
Since kedro 0.17.7(?) there have been introduced namespaces which cause issues in kfp artifacts, as they are not properly handled.
Unless the function to create kfp artifacts is disabled in
kubeflow.yaml
config:It causes issues like:
when trying to run or update the pipeline.