kubeflow.fairing.deployers.kfserving package

Submodules

kubeflow.fairing.deployers.kfserving.kfserving module

class kubeflow.fairing.deployers.kfserving.kfserving.KFServing(framework, default_storage_uri=None, canary_storage_uri=None, canary_traffic_percent=0, namespace=None, labels=None, annotations=None, custom_default_container=None, custom_canary_container=None, stream_log=False, cleanup=False)

Bases: kubeflow.fairing.deployers.deployer.DeployerInterface

Serves a prediction endpoint using Kubeflow KFServing.

deploy(isvc)

deploy kfserving endpoint

Parameters:isvc – InferenceService for deploying.

generate_isvc()

generate InferenceService

generate_predictor_spec(framework, storage_uri=None, container=None)

Generate predictor spec according to framework and default_storage_uri or custom container.

get_logs()

get log from prediction pod

set_labels(labels)

set label for deployed prediction

Parameters:labels – dictionary of labels {label_name:label_value}

Module contents