MLA-C01 Question 321
Select 4You are tasked with setting up a CI/CD pipeline for deploying machine learning models using AWS services. The pipeline must automate model building, training, testing, and deployment while ensuring modularity and scalability. Which combination of AWS services would best enable you to achieve this objective?
- A
AWS CodePipeline for orchestrating the CI/CD workflow
- B
AWS CloudFormation for defining infrastructure as code
- C
Amazon SageMaker for model building, training, and deployment
- D
AWS Glue for managing and deploying trained models
- E
Amazon S3 for storing model artifacts and data
- F
AWS Lambda for preprocessing and postprocessing in the pipeline
Show answer and explanation
Correct answers: A, B, C, E
Explanation
To set up a CI/CD pipeline for machine learning models on AWS, a combination of services is required to handle orchestration, infrastructure, machine learning-specific tasks, and storage. AWS CodePipeline ensures seamless automation of the pipeline, AWS CloudFormation provides infrastructure as code for modularity and scalability, Amazon SageMaker handles model-specific operations, and Amazon S3 stores necessary data and artifacts. While AWS Glue and AWS Lambda are valuable in other contexts, they are not directly relevant or essential for this specific scenario.
- A. Correct.
AWS CodePipeline is a highly suitable service for orchestrating CI/CD workflows, as it natively integrates with other AWS services and supports automation of all steps in a pipeline.
- B. Correct.
AWS CloudFormation is critical for defining infrastructure as code, ensuring that the CI/CD pipeline and other resources can be created and managed programmatically and consistently.
- C. Correct.
Amazon SageMaker provides specialized capabilities for machine learning model building, training, and deployment, making it a core component of this solution.
- D. Incorrect.
AWS Glue is primarily used for ETL (Extract, Transform, Load) operations and is not designed for managing or deploying trained models, so it is not relevant here.
- E. Correct.
Amazon S3 is commonly used to store data and artifacts such as trained model files, making it an essential part of the CI/CD pipeline for machine learning workflows.
- F. Incorrect.
AWS Lambda is a useful service for preprocessing and postprocessing tasks, but it is not directly involved in the orchestration of CI/CD pipelines or model deployment.