Metaflow Offline Deploy

MetaflowOfflineDeploy¶

The MetaflowOfflineDeploy class is a subclass of BaseDeploy and provides methods for running a Metaflow flow object and obtaining artifacts from the MetaflowOfflineDeploy pipeline.

Attributes¶

METAFLOW_CLASS - The name of the class that inherits from metaflow's FlowSpec

Configuration¶

Required Configuration¶

MetaflowOfflineDeploy requires the following component types:

metadata_tracker
model_repository
model_deployer
resource_version_control

Methods¶

run¶

This method runs the MetaflowOfflineDeploy. It loads a Metaflow flow object based on the current file, loads the flow, and then runs it using the model and version provided. This method logs execution and does not return any value.

def run(self, model, model_version, *args, **kwargs):

Parameters:

model (object): The model object.
model_version (object): The model version object.

get_artifacts¶

This method obtains artifacts of the MetaflowOfflineDeploy. It gets the latest run of the MetaflowOfflineDeploy pipeline and returns the requested artifacts.

def get_artifacts(self, artifact_keys):

Parameters:

artifact_keys (list): A list of keys for the artifacts being requested.

Returns: - artifacts (dict): A dictionary of requested artifacts.

MetaflowOfflineDeploySpec Methods¶

MetaflowOfflineDeploySpec contains the following methods. These are mirrored from the OfflineDeploy class, and you should see that documentation for more information (Note: instead of these method explicitly using arguments, they instead access saved artifacts during the Metaflow run).

start
promote_model
approve_model
deploy_model
check_approval
end