Which combination of steps must the data scientist take to perform the A/B testing?
(Choose two.)
Create a new endpoint configuration that includes a production variant for each of the two models.
Create a new endpoint configuration that includes two target variants that point to different endpoints.
Deploy the new model to the existing endpoint.
Update the existing endpoint to activate the new model.
Update the existing endpoint to use the new endpoint configuration.
Explanations:
To perform A/B testing, the data scientist must create a new endpoint configuration that includes both models (production variant for the existing model and a new variant for the new model).
This option suggests creating an endpoint configuration with two target variants that point to different endpoints, which is not a valid setup for A/B testing in SageMaker. A single endpoint configuration should handle multiple variants.
Deploying the new model to the existing endpoint does not ensure A/B testing. A/B testing requires managing multiple variants within a single endpoint configuration.
Updating the existing endpoint to activate the new model does not allow for A/B testing. The correct approach is to create a new endpoint configuration with multiple variants.
After creating the new endpoint configuration, the existing endpoint needs to be updated to use the new configuration, which includes both models for A/B testing.
I estimate that the answer is:
Create a new endpoint configuration that includes a production variant for each of the two models.
If I’m not mistaken, the answer is:
Create a new endpoint configuration that includes a production variant for each of the two models.