QuestionQ53

VerifiedML Model Development

An ML engineer uses Amazon SageMaker to train a deep learning model that needs distributed training. Following several training attempts, the ML engineer notices that the instances are not performing as expected and identifies communication overhead among the training instances.

What should the ML engineer do to MINIMIZE the communication overhead between the instances?

Explanation

For SageMaker distributed training, placing the instances, VPC subnet, and training data storage in the same AWS Region and Availability Zone minimizes network distance and communication overhead between compute nodes and data storage.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!