Exam AWS Certified Machine Learning - Specialty topic 1 question 86 discussion

C seems correct as per documentations.

I would answer C: https://docs.aws.amazon.com/sagemaker/latest/dg/your-algorithms-training-algo-dockerfile.html "To configure a Docker container to run as an executable, use an ENTRYPOINT instruction in a Dockerfile. SageMaker overrides any default CMD statement in a container by specifying the train argument after the image name"

I thought it was D, but it is C. It's not D because we copy the TRAINING code into /opt/ml/code/train.py

In Docker, the ENTRYPOINT instruction is used to specify the executable that should be run when the container starts. However, Amazon SageMaker expects the training script to be launched by specific commands provided by SageMaker itself, rather than relying solely on the Docker container's ENTRYPOINT. The convention for using Docker containers with Amazon SageMaker is to copy the training script and associated resources to specific directories within the container, such as /opt/ml/code, and let SageMaker manage the execution of the training process. I would go with D

C is correct

C is correct

Amazon SageMaker requires that a custom algorithm container has an executable named train that runs your training program. This executable can be configured as an ENTRYPOINT in the Dockerfile, which specifies the default command to run when the container is launched.

Amazon SageMaker requires that a custom algorithm container has an executable named train that runs your training program. This executable can be configured as an ENTRYPOINT in the Dockerfile, which specifies the default command to run when the container is launched.

you are all wrong, it is D based on,https://docs.aws.amazon.com/sagemaker/latest/dg/adapt-training-container.html

To package a Docker container for use with Amazon SageMaker, the training program should be configured as an ENTRYPOINT named train in the Dockerfile. This means that the training program will be automatically executed when the container is launched by Amazon SageMaker, and it can be passed command-line arguments to specify hyperparameters or other training settings.

The recommended option to package the Docker container for Amazon SageMaker is to configure the training program as an ENTRYPOINT named train. This is because ENTRYPOINT allows you to specify a command that will always be executed when the Docker container is run, ensuring that the training program will always run when the container is launched by Amazon SageMaker. Additionally, naming the ENTRYPOINT "train" is a convention used by Amazon SageMaker to identify the main training script.

It's C

C for sure as per AWS docs: > In your Dockerfile, use the exec form of the ENTRYPOINT instruction: > ENTRYPOINT ["python", "k-means-algorithm.py"]

C is correct

option C https://github.com/awsdocs/amazon-sagemaker-developer-guide/blob/master/doc_source/your-algorithms-training-algo-dockerfile.md