Skip to content

A library for training and deploying machine learning models on Amazon SageMaker

License

Notifications You must be signed in to change notification settings

affinda/sagemaker-python-sdk

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4,132 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SageMaker Python SDK (Affinda Fork)

This is a fork of AWS's SageMaker Python SDK with additional features for local mode.

Fork-Specific Features

This fork adds two features to src/sagemaker/local/image.py:

  1. Nvidia GPU Driver Support - Explicitly specifies the nvidia driver for GPU containers in local mode
  2. SM_SUBNET Environment Variable - Allows setting a custom Docker network subnet via the SM_SUBNET environment variable, useful when the default sagemaker-local Docker network subnet conflicts with existing network configurations

Installation

pip install git+https://github.com/affinda/sagemaker-python-sdk.git@v2.257.0

Updating to a Newer Upstream Version

When AWS releases a new version and you need to update this fork:

1. Add AWS as upstream remote (first time only)

git remote add upstream https://github.com/aws/sagemaker-python-sdk.git

2. Fetch upstream tags

git fetch upstream --tags

3. Reset the branch to the desired AWS version

git checkout feat/affinda-updates
git reset --hard v<NEW_VERSION>

4. Cherry-pick the two custom commits

git cherry-pick af73a2af  # nvidia gpu support
git cherry-pick e1b0c0ec  # SM_SUBNET support

If there are conflicts, resolve them and run git cherry-pick --continue.

5. Update the version tag

git tag -f v<NEW_VERSION>

6. Force push the updated branch and tag

git push --force origin feat/affinda-updates
git push --force origin v<NEW_VERSION>

Branch Structure

  • feat/affinda-updates - Main branch with custom commits on top of AWS releases
  • master - Not actively used

Original Documentation

For the full SageMaker Python SDK documentation, see the upstream repository and Read the Docs.

About

A library for training and deploying machine learning models on Amazon SageMaker

Resources

License

Code of conduct

Contributing

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 99.7%
  • Other 0.3%