QuestionQ43

VerifiedDeployment and Orchestration of ML Workflows

An ML engineer must deploy four ML models in an Amazon SageMaker inference pipeline. The models were created with different frameworks. The ML engineer must also enable clients to use the invoke_endpoint call to run inference for each model.

Which solution meets these requirements in the MOST cost-effective way?

Explanation

SageMaker multi-container endpoints can host multiple containers that use different models or frameworks on one endpoint. With direct invocation enabled, clients call InvokeEndpoint and specify TargetContainerHostname to send an inference request to the required model container. Sharing one endpoint improves utilization and reduces cost compared with separate single-model endpoints.

Learn more

Community Discussion

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