custom containers for training in sagemaker?

Question

some of the containers listed here , https://github.com/aws/deep-learning-containers/blob/master/available_images.md , have few security vunerabilities, are there any examples to build similar container from scratch. for example - this one - 763104351884.dkr.ecr.us-east-1.amazonaws.com/pytorch-inference:2.0.1-cpu-py310-ubuntu20.04-sagemaker

examples of building custom images, or Dockerfile to see what components are being installed for this image.

Answer

Hello Fam!

Kindly go through the documentation below [https://docs.aws.amazon.com/sagemaker/latest/dg/docker-containers.html]() and follow the notebook for the solution here -> https://github.com/aws/amazon-sagemaker-examples/tree/main/advanced_functionality/scikit_bring_your_own.

```  Dockerfile sample for training
Dockerfile for training
# Set the environment variables
ENV PYTHONUNBUFFERED=TRUE
ENV PYTHONDONTWRITEBYTECODE=TRUE
ENV PATH="/opt/program:${PATH}"

# Copy the model and inference code
COPY train / opt/program

# Set the working directory
WORKDIR /opt/program

RUN chmod +x /opt/program/serve

RUN chmod +x /opt/program/train
# Set the default command to run when the container starts

ENTRYPOINT [ "python3" ]

```
Test locally
docker run -v your path to folder/local_test:/opt/ml --rm training_container:latest train

You can then scan with trivy or Amazon Inspector for vulnerability fixes and patch them accordingly

Hope it helps!

Cheers & Best Regards,
Ashlin Gabriel

custom containers for training in sagemaker?

相關內容