Skip to content

BaseDeploy

Overview

The deploy pipeline is a pipeline that performs common actions concerning model deployments, such as promoting and approving model versions, as well as deploying model versions to end points. Typically we would expect the following chain for a production model: model built -> promotion (technical approval) -> approval (legal/compliance approval) -> deployment.

Attributes

BaseDeploy contains no default attributes.

Configuration

BaseDeploy contains no default or required configuration.

Interface

The following methods are part of BaseDeploy and should be implemented in any class that inherits from this base class:

promote_model

Promotes a model version.

def promote_model(self, model_version, *args, **kwargs) -> Any

Arguments:

  • model_version (object): The model version object that is being promoted.

Returns:

  • promotion (object): Returns a promotion object, very likely returned back from a model_respository.

approve_model

Approves a promotion (i.e. model version).

def approve_model(self, promotion, *args, **kwargs) -> Any

Arguments:

  • promotion (object): The promotion to approve.

Returns:

  • approval (object): Returns an approval object, very likely returned back from a model_respository.

deploy_model

Creates a deployment of an approved model version.

def deploy_model(self, promotion, *args, **kwargs) -> Any

Arguments:

  • promotion (object): The promotion to deploy

Returns:

  • deployment (object): Returns a deployment object, very likely returned back from a model_deployer.

check_approval

Creates a deployment of an approved model version.

def check_approval(self, promotion, *args, **kwargs) -> bool

Arguments:

  • promotion (object): The promotion to check to see if it has been approved.

Returns:

  • bool (object): Whether the promotion is approved or not.