Skip to main content

mb.get_job_output(deployment_name=, job_name=, ...)

Gets the output returned by a training job. Learn more about training jobs.

Parameters​

  • deployment_name: str The deployment containing the training job.
  • job_name: str The name of the training job within the deployment.
  • run_id: Optional[int]: If no ID is supplied, the most recent training job output will be returned. Supply an ID, seen in the train job run logs, to fetch outputs from previous training job runs.
  • result_stored_as: Optional[str]: If the training job was defined with store_result_as, use this parameter with the save value when fetching the result.
  • model_name: Optional[str]: To retrieve a model from a training, set this parameter to the name of the model.
  • file_name: Optional[str]: To retrieve the contents of a specific file from a training, set this parameter to the path of the file.

Returns​

Any - the value the training function returned.

Examples​

Get the output of job​

This fetches whatever was returned by the job function.

mb.get_job_output(deployment_name="training_example", job_name="train_iris_lr")

Get a model saved during a job​

mb.get_job_output(deployment_name="training_example", job_name="train_iris_lr", model_name="example_model")

Get the output of a specific job run​

mb.get_job_output(deployment_name="training_example", job_name="train_iris_lr", run_id=2)

Get the contents of an outputted file​

If the job created the file learn_error.tsv while running:

mb.get_job_output(
    deployment_name="my_deployment",
    job_name="my_training_job",
    run_id=7,
    file_name="learn_error.tsv")

See also​