Version v0.6 of the documentation is no longer actively maintained. The site that you are currently viewing is an archived snapshot. For up-to-date documentation, see the latest version.


Tracking and managing metadata of machine learning workflows in Kubeflow

The goal of the Metadata project is to help Kubeflow users understand and manage their machine learning (ML) workflows by tracking and managing the metadata that the workflows produce.

In this context, metadata means information about executions (runs), models, datasets, and other artifacts. Artifacts are the files and objects that form the inputs and outputs of the components in your ML workflow.

Installing the Metadata component

Kubeflow v0.6.1 and later versions install the Metadata component by default. You can skip this section if you are running Kubeflow v0.6.1 or later.

If you want to install the latest version of the Metadata component or to install the component as an application in your Kubernetes cluster, follow these steps:

  1. Download the Kubeflow manifests repository:

    git clone
  2. Run the following commands to deploy the services of the Metadata component:

    cd manifests/metadata/base
    kustomize build . | kubectl apply -n kubeflow -f -

Using the Metadata SDK to record metadata

The Metadata project publishes a Python library (SDK) that you can use to log (record) your metadata.

Run the following command to install the Metadata SDK:

pip install kfmd

Try the Metadata SDK in a sample Jupyter notebook

You can find an example of how to use the Metadata SDK in this demo notebook.

To run the notebook in your Kubeflow cluster:

  1. Follow the guide to setting up your Jupyter notebooks in Kubeflow.
  2. Go to the demo notebook on GitHub.
  3. Download the notebook code by opening the Raw view of the file, then right-clicking on the content and saving the file locally as demo.ipynb.
  4. Go back to your Jupyter notebook server in the Kubeflow UI. (If you’ve moved away from the notebooks section in Kubeflow, click Notebook Servers in the left-hand navigation panel to get back there.)
  5. In the Jupyter notebook UI, click Upload and follow the prompts to upload the demo.ipynb notebook.
  6. Click the notebook name (demo.ipynb) to open the notebook in your Kubeflow cluster.
  7. Run the steps in the notebook to install and use the Metadata SDK.

When you have finished running through the steps in the demo.ipynb notebook, you can view the resulting metadata on the Kubeflow UI:

  1. Click Artifact Store in the left-hand navigation panel on the Kubeflow UI.
  2. On the Artifacts screen you should see the following items:

    • A model metadata item with the name MNIST.
    • A metrics metadata item with the name MNIST-evaluation.
    • A dataset metadata item with the name mytable-dump.

    You can click the name of each item to view the details. See the section below about the Metadata UI for more details.

Learn more about the Metadata SDK

The Metadata SDK includes the following predefined types that you can use to describe your ML workflows:

  • data_set.json to capture metadata for a dataset that forms the input into or the output of a component in your workflow.
  • execution.json to capture metadata for an execution (run) of your ML workflow.
  • metrics.json to capture metadata for the metrics used to evaluate an ML model.
  • model.json to capture metadata for an ML model that your workflow produces.

Tracking artifacts on the Metadata UI

You can view a list of logged artifacts and the details of each individual artifact in the Artifact Store on the Kubeflow UI.

  1. Go to Kubeflow in your browser. (If you haven’t yet opened the Kubeflow UI, find out how to access the Kubeflow UIs.)
  2. Click Artifact Store in the left-hand navigation panel: Metadata UI

  3. The Artifacts screen opens and displays a list of items for all the metadata events that your workflows have logged. You can click the name of each item to view the details.

    The following examples show the items that appear when you run the demo.ipynb notebook described above:

    A list of metadata items

    • Example of model metadata with the name “MNIST”:

      Model metadata for an example MNIST model

    • Example of metrics metadata with the name “MNIST-evaluation”:

      Metrics metadata for an evaluation of an MNIST model

    • Example of dataset metadata with the name “mytable-dump”:

      Dataset metadata

Backend and REST API

The Kubeflow metadata backend uses ML Metadata (MLMD) to manage the metadata and relationships.

The backend exposes a REST API.

You can add your own metadata types so that you can log metadata for custom artifacts. To add a custom type, send a REST API request to the artifact_types endpoint.

For example, The following request registers an artifact type with name myorg/mytype/v1 and three properties:

  • f1 (string)
  • f2 (integer)
  • f3 (double)
curl -X POST http://localhost:8080/api/v1alpha1/artifact_types \
  --header "Content-Type: application/json" -d \
  '{"name":"myorg/mytype/v1","properties":{"f1":"STRING", "f2":"INT", "f3": "DOUBLE"}}'

Next steps

Run the xgboost-synthetic notebook to build, train, and deploy an XGBoost model using Kubeflow Fairing and Kubeflow Pipelines with synthetic data. Examine the metadata output after running through the steps in the notebook.