AWS Certified Machine Learning – Specialty — Question 330
A machine learning (ML) engineer is using Amazon SageMaker automatic model tuning (AMT) to optimize a model's hyperparameters. The ML engineer notices that the tuning jobs take a long time to run. The tuning jobs continue even when the jobs are not significantly improving against the objective metric.
The ML engineer needs the training jobs to optimize the hyperparameters more quickly.
How should the ML engineer configure the SageMaker AMT data types to meet these requirements?
Answer options
- A. Set Strategy to the Bayesian value.
- B. Set RetryStrategy to a value of 1.
- C. Set ParameterRanges to the narrow range Inferred from previous hyperparameter jobs.
- D. Set TrainingJobEarlyStoppingType to the AUTO value.
Correct answer: D
Explanation
Setting TrainingJobEarlyStoppingType to AUTO enables SageMaker AMT to automatically terminate training jobs that are unlikely to improve upon the objective metric, significantly reducing total tuning time. Other configurations like changing the search strategy or narrowing parameter ranges do not proactively stop underperforming individual jobs, and RetryStrategy only manages job retries upon failure.