kubeflow.fairing.deployers.gcp package

Submodules

kubeflow.fairing.deployers.gcp.gcp module

class kubeflow.fairing.deployers.gcp.gcp.GCPJob(project_id=None, region=None, scale_tier=None, job_config=None)

Bases: kubeflow.fairing.deployers.deployer.DeployerInterface

Handle submitting training job to GCP.

create_request_dict(pod_template_spec)

Return the request to be sent to the ML Engine API.

Parameters:pod_template_spec – pod spec template of the training job

deploy(pod_template_spec)

Deploys the training job

Parameters:pod_template_spec – pod spec template of the training job

get_logs()

Streams the logs for the training job

kubeflow.fairing.deployers.gcp.gcpserving module

class kubeflow.fairing.deployers.gcp.gcpserving.GCPServingDeployer(model_dir, model_name, version_name, project_id=None, **deploy_kwargs)

Bases: kubeflow.fairing.deployers.deployer.DeployerInterface

Handle deploying a trained model to GCP.

deploy(pod_template_spec)

Deploys the model to Cloud ML Engine.

Parameters:pod_template_spec – pod spec template of training job

get_logs()

abstract get log

Module contents