Kubeflow: Artifacts not getting Generated

Created on 8 Oct 2020 · 2Comments · Source: kubeflow/kubeflow

/kind bug

What steps did you take and what happened:
[A clear and concise description of what the bug is.]

Created metadata json file to display Confusion Matrix according to Kubeflow v1.0 documentation.

`metadata = {
    'outputs' : [{
        'type': 'confusion_matrix',
        'format': 'csv',
        'storage': 'gcs',
        'schema': [
            {'name': 'target', 'type': 'CATEGORY'},
            {'name': 'predicted', 'type': 'CATEGORY'},
            {'name': 'count', 'type': 'NUMBER'},
        ],
        'source': 'gs://pipelines_artifacts/Artifacts/Conf_mat.csv', #conf_matr,
   # Convert flags to string because for bealean values we want "True|False" to match csv data.
        'labels': list(map(str, flags))
}]
} 
with file_io.FileIO('/mlpipeline-ui-metadata.json', 'w') as f:
    json.dump(metadata, f)`

The mlpipeline-ui-metadata.json file is getting generated and dumped to the container, but Artifacts tab for that pipeline block is showing a blank view.

Run output gives the following message -

No output artifacts found for this run.

Tried both versions of storing the confusion matrix .csv file - gcs and minio (OutputPath()) object

Also getting the following error message on the Experiment view when I use OutputPath() -

Error: Unsupported storage path: /tmp/outputs/conf_matr/data. Click Details for more information.

What did you expect to happen:

Kubeflow v1.0 metadata 'source' field supports Gcs and Minio paths. Expected artifacts to get generated normally.

Anything else you would like to add:
[Miscellaneous information that will assist in solving the issue.]

Environment:

Kubeflow version: (version number can be found at the bottom left corner of the Kubeflow dashboard): v1.0
Kubernetes platform: (e.g. minikube) : k8s
Kubernetes version: (use kubectl version): 1.14
OS (e.g. from /etc/os-release): Container-optimised OS from Google