Skip to content

MLflow Serve

Use the serve action to deploy MLflow models as REST API services on Kubernetes. The model must be saved in MLflow format with an MLmodel file.

Build the serving image first

If the Function does not already have a compatible serving image, create one with the MLflow Build action before deploying it.

  • Create


    Configure a Function with the MLflow model path.

  • Run


    Deploy the model as a service.

  • Use


    Send inference requests to the service.