MLA-C01 Question 319
Select 3You are working on an image classification model and want to implement a CI/CD pipeline for deploying the model to an Amazon SageMaker endpoint. You need to ensure that the pipeline includes stages for code testing, model training, model validation, and deployment. Which of the following AWS services and tools can you use to set up and automate this pipeline?
- A
AWS CodePipeline
- B
Amazon SageMaker Pipelines
- C
AWS CodeBuild
- D
Amazon CloudWatch
- E
AWS Lambda
Show answer and explanation
Correct answers: A, B, C
Explanation
To set up and automate a CI/CD pipeline for ML workflows, AWS CodePipeline can orchestrate the stages, while Amazon SageMaker Pipelines provides ML-specific workflow automation. AWS CodeBuild is used for building and testing code. While CloudWatch and Lambda are useful for monitoring and execution of custom actions respectively, they do not serve as core orchestration or automation tools for CI/CD pipelines.
- A. Correct.
AWS CodePipeline is a key orchestration service that can be used to define and manage the CI/CD pipeline, integrating various stages such as building, testing, and deploying machine learning models.
- B. Correct.
Amazon SageMaker Pipelines is specifically designed for automating and orchestrating ML workflows, including stages like model training, validation, and deployment.
- C. Correct.
AWS CodeBuild is a service used to compile source code, run tests, and produce software packages. It is integral to the CI/CD process for building and testing ML code.
- D. Incorrect.
Amazon CloudWatch is primarily a monitoring and observability service. While it can provide logs and metrics for the pipeline, it does not directly facilitate CI/CD orchestration.
- E. Incorrect.
AWS Lambda is a serverless compute service that can execute custom scripts but is not specifically designed for orchestrating or automating CI/CD pipelines.