Exams / AIF-C01 / AWS generative AI services / Question 2
AIF-C01 practice question: AWS generative AI services (2 of 5)
Fundamentals of Generative AI A free question from our original AIF-C01 practice exam. Try it, then open the answer to see why each option is right or wrong.
A healthcare analytics firm wants to run an open-weight foundation model inside its own AWS account. It needs to choose the instance type, control the endpoint, and fine-tune the model on its own data. Which AWS option BEST fits these needs?
Choose one answer.
- AA. SageMaker JumpStart, deployed to a SageMaker AI endpoint
- BB. Amazon Comprehend, using its built-in entity recognition
- CC. Amazon Quick, connecting the model through a chat agent
- DD. AWS Artifact, downloading the model's compliance report
Show the answer and explanation
Answer: A
- A. Right: Correct. JumpStart offers publicly available foundation models that you can deploy to endpoints you configure, and fine-tune.
- B. Wrong: Comprehend is a pre-trained NLP service. It does not host an open-weight foundation model of your choice.
- C. Wrong: Amazon Quick is a business user workspace. It does not give control over the instances that host a model.
- D. Wrong: Artifact provides AWS compliance documents. It does not host or deploy models.
Explanation
There are two broad ways to use a foundation model on AWS. A fully managed API such as Amazon Bedrock needs no infrastructure. Self-hosting, for example through SageMaker JumpStart onto a SageMaker AI endpoint, gives more control over instances, deployment configuration and fine-tuning, but the customer manages more. Teams should also check each open model's licence before use.
Study
Keep practising
Answer questions with a score at the end, or practise just this topic with the explanation after each one. No sign-up needed.