Deploy and expose the model

Deploying a model is as easy as defining a serverless function, providing the model reference and then deploying.

Create a model serving function and provide the model:

serving_fn = mlrun.new_function("serving", image="mlrun/mlrun", kind="serving")
serving_fn.add_model('cancer-classifier',model_path=trainer_run.outputs["model"], class_name='mlrun.frameworks.sklearn.SklearnModelServer')

Deploy (it may take several minutes):

project.deploy_function(serving_fn)

You can now test the endpoint:

my_data = {"inputs"
           :[[
               1.371e+01, 2.083e+01, 9.020e+01, 5.779e+02, 1.189e-01, 1.645e-01,
               9.366e-02, 5.985e-02, 2.196e-01, 7.451e-02, 5.835e-01, 1.377e+00,
               3.856e+00, 5.096e+01, 8.805e-03, 3.029e-02, 2.488e-02, 1.448e-02,
               1.486e-02, 5.412e-03, 1.706e+01, 2.814e+01, 1.106e+02, 8.970e+02,
               1.654e-01, 3.682e-01, 2.678e-01, 1.556e-01, 3.196e-01, 1.151e-01]
            ]
}
serving_fn.invoke("/v2/models/cancer-classifier/infer", body=my_data)

Create an API gateway

To make the API accessible from outside, we'll need to create an API gateway in Nuclio.

Go to your Coder instance, go to the dashboard and access Nuclio. You will notice a demo-ml project, which we created earlier. When you access it, you will see the demo-ml-serving function listed, but click on the API GATEWAYS tab on top instead. Then, click on NEW API GATEWAY.

On the left, if you wish, set Authentication to Basic and choose Username and Password.

In the middle, set any Name you want. Host must use the same domain as the other components of the Digital Hub. For example, if you access Coder at coder.my-digitalhub-instance.it, the Host field should use a value like demo-ml.my-digitalhub-instance.it.

On the right, under Primary, you must enter the name of the function, in this case demo-ml-serving.

Save and, after a few moments, you will be able to call the API at the address you entered in Host! If you set Authentication to Basic, don't forget that you have to provide the credentials.