AWS Certified Machine Learning – Specialty — Question 363

A media company is building a computer vision model to analyze images that are on social media. The model consists of CNNs that the company trained by using images that the company stores in Amazon S3. The company used an Amazon SageMaker training job in File mode with a single Amazon EC2 On-Demand Instance.

Every day, the company updates the model by using about 10,000 images that the company has collected in the last 24 hours. The company configures training with only one epoch. The company wants to speed up training and lower costs without the need to make any code changes.

Which solution will meet these requirements?

Answer options

• A. Instead of File mode, configure the SageMaker training job to use Pipe mode. Ingest the data from a pipe.
• B. Instead of File mode, configure the SageMaker training job to use FastFile mode with no other changes.
• C. Instead of On-Demand Instances, configure the SageMaker training job to use Spot Instances. Make no other changes,
• D. Instead of On-Demand Instances, configure the SageMaker training job to use Spot Instances, implement model checkpoints.

Correct answer: B

Explanation

FastFile mode acts as a virtual mount that streams data directly from Amazon S3, eliminating the initial download time associated with File mode and thereby reducing both training time and costs without requiring any changes to the training code. In contrast, Pipe mode (Option A) and Spot Instances with checkpointing (Option D) both require code modifications to implement. Utilizing Spot Instances without checkpoints (Option C) does not speed up training and risks job interruption without a recovery path.