MLA-C01 Question 234
Single answerYou are tasked with deploying a machine learning model using Amazon SageMaker. The model requires pre-processing of input data and post-processing of output predictions. Additionally, the deployment needs to handle multiple models under a single endpoint. Which SageMaker feature should you utilize to meet these requirements?
- A
SageMaker multi-model endpoints with custom inference code
- B
SageMaker batch transform
- C
SageMaker Processing Jobs
- D
SageMaker Neo for model optimization
Show answer and explanation
Correct answer: A
Explanation
SageMaker multi-model endpoints are specifically designed to host multiple models under a single endpoint and support custom inference code. This capability allows you to include logic for pre-processing input data and post-processing model predictions, making it the best fit for the given scenario. Other options, such as batch transform or processing jobs, address different aspects of the ML workflow but do not fulfill the requirements of deployment and real-time inference with multiple models.
- A. Correct.
Correct. SageMaker multi-model endpoints allow you to host multiple models under a single endpoint and include custom inference code, which can handle pre-processing and post-processing of data.
- B. Incorrect.
Incorrect. SageMaker batch transform is used for offline, batch predictions and does not support hosting multiple models under a single endpoint.
- C. Incorrect.
Incorrect. SageMaker Processing Jobs are used for running data pre-processing or post-processing workflows independently of model inference, not for hosting models or handling inference requests.
- D. Incorrect.
Incorrect. SageMaker Neo is used for optimizing models for performance on edge devices and does not provide capabilities for hosting multiple models or handling inference logic.