AWS SageMaker, describe a specific training job using python SDK

Question

Using SageMaker python SDK I've created an hyper-param tuning job, which runs many jobs in parallel to search for the optimal HP values.

The jobs complete and I get the best training job name as a string "Job...". I've found the following article about how to describe a job using the AWS-CLI or http request.

Is there a way of doing it using the python SageMaker SDK, in order to avoid the complexity of an authenticated request to AWS?

Answer 1

With a sagemaker.session.Session instance, you can describe training jobs :

import sagemaker


sagemaker_session = sagemaker.session.Session()
sagemaker_session.describe_training_job("Job...")