Card 1 of 5
Question
Which inference option is best for offline scoring of stored datasets?
Click to reveal answer
Answer
Batch transform.
Click to flip back
All Deployment and Orchestration Flashcards
1
Q: Which inference option is best for offline scoring of stored datasets?
A: Batch transform.
2
Q: Which inference option fits intermittent traffic with low operational overhead?
A: Serverless inference.
3
Q: Which SageMaker feature creates repeatable ML workflow steps?
A: SageMaker Pipelines.
4
Q: What are endpoint variants used for?
A: Traffic shifting, canary release, blue/green style deployment, or A/B testing.
5
Q: Which artifact should be approved before production deployment?
A: A registered model package version in Model Registry.