Skip to main content

Enable MLflow Experiments & Model Registry

By default, MLflow is disabled in order to save costs. If you want to enable MLflow on your Connected AI Platform instance, enable it in your product's product.yaml by adding the following configuration and opening a Pull Request:

product.yaml
kubeflow_profiles:
  <namespace>:
    ...
      mlflow:
        enabled: true

For more on what MLflow provides once enabled, see What is MLflow?.

Disable MLflow Experiments & Model Registry​

By setting mlflow.enabled to false, the MLflow instance will no longer be present in the cluster and hence won't consume any unnecessary resources.