Seldon Core operator Crashes when deployment with empty predictor is passed

[axsaucedo]

Describe the bug

When creating a seldon deployment with empty predictors, the operator enters crash loop.

To reproduce

kubectl apply -f - << END
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
  name: sklearn
spec:
  name: iris
  predictors: []
END

Expected behaviour

The webhook should reject the deployment / patch.

Environment

• Kubernetes Version: 1.16
• Cloud Provider: Docker Desktop
• Deployed Seldon System Images:

          value: docker.io/seldonio/engine:1.2.0
          value: docker.io/seldonio/seldon-core-executor:1.2.0
        image: docker.io/seldonio/seldon-core-operator:1.2.0

[ukclivecox]

Can be fixed along with #2022