QuestionQ6

VerifiedDeployment and Orchestration of ML Workflows

An ML engineer must implement a solution to host a trained ML model. The request rate for the model will vary throughout the day.

The ML engineer needs a scalable solution that minimizes costs while the model is not being used. The solution must also retain the model capacity to respond to requests during periods of peak use.

Which solution meets these requirements?

Explanation

Amazon SageMaker endpoint auto scaling uses Amazon CloudWatch metrics to dynamically adjust the number of endpoint instances. It scales out as workload demand rises to provide peak capacity and scales in when demand falls, eliminating unnecessary provisioned-instance cost.

Learn more

Community Discussion

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