MLA-C01 Question 346
Select 3You are a Machine Learning Engineer at a company that uses AWS for its ML workflows. Your team wants to automate the deployment of ML models into production while ensuring that only models meeting specific accuracy thresholds are deployed. Which combination of practices should you implement as part of a CI/CD pipeline to achieve this?
- A
Integrate AWS CodePipeline with SageMaker Model Registry to track and approve model versions before deployment.
- B
Use AWS Lambda to trigger model training whenever new data is uploaded to an S3 bucket.
- C
Implement an automated test step in the pipeline to validate the model's performance metrics against predefined thresholds.
- D
Manually upload the trained model to an S3 bucket and deploy it using SageMaker.
- E
Use SageMaker Projects to standardize ML workflows and integrate them into a CI/CD pipeline.
Show answer and explanation
Correct answers: A, C, E
Explanation
CI/CD principles in ML workflows involve automating model tracking, validation, and deployment to ensure rapid and reliable updates to production. By integrating CodePipeline with SageMaker Model Registry, automating performance validation, and using SageMaker Projects to streamline workflows, you can build a robust pipeline that meets these goals. Manual steps or automation that does not align with validation or deployment objectives do not fully implement CI/CD principles.
- A. Correct.
Correct: Integrating AWS CodePipeline with SageMaker Model Registry enables automated tracking of model versions, along with approval workflows for production deployment.
- B. Incorrect.
Incorrect: While triggering training with AWS Lambda is useful for automation, it does not address the need for validating model performance or deploying models systematically through CI/CD.
- C. Correct.
Correct: Adding an automated test step ensures that only models meeting accuracy thresholds are promoted, aligning with CI/CD principles.
- D. Incorrect.
Incorrect: Manually uploading the model bypasses the automation benefits of a CI/CD pipeline and is error-prone.
- E. Correct.
Correct: SageMaker Projects help standardize and automate ML workflows while providing integration points for CI/CD pipelines.