Contribute to PipelineAI

Want to join the PipelineAI Community?

We will gladly accept pull requests for implementations of popular ML/AI models, fast prediction runtimes, and flexible experimentation frameworks.

Click HERE for the active list of outstanding PipelineAI issues and feature requests..

Creating a Pull Request

Create a Branch for Your Changes

git branch <branch-name>

Verify New Branch is Created

git branch

* master   <-- this is current branch

Switch to Your New Branch

git checkout <branch-name>

Switched to branch '<branch-name>'

Verify That You're on the Branch

git branch
* <branch-name> <-- you are now on the new branch

Make Your Branch Changes


Verify Branch Changes

git status

Prepare (Add) Your Branch Changes for Commit

git add ...

Commit Your Branch Changes Locally

git commit -m "<descriptive-message>"

Push Your Branch Changes Remotely

git push -u origin <branch-name>

# Counting objects: 8, done.
# Delta compression using up to 8 threads.
# Compressing objects: 100% (6/6), done.
# Writing objects: 100% (8/8), 684 bytes | 0 bytes/s, done.
# Total 8 (delta 2), reused 0 (delta 0)
# remote: Resolving deltas: 100% (2/2), completed with 1 local objects.
# To github.com:fluxcapacitor/pipeline.git
# * [new branch]      <branch-name> -> <branch-name>
#Branch <branch-name> set up to track remote branch <branch-name> from origin.
Note: If you see an error related to Permission denied (publickey) or Please make sure you have the correct access rights, follow the steps detailed here.

Submit a Pull Request with Github Server UI

  • Compare your <branch-name> to the master branch on Github Server

Fix the Pull Request if Needed based on Comments

Wait for Committer to Merge Pull Request (within Github UI)

Delete the Branch (from both Local and Remote Github)

Go Back to master Branch

git checkout master

Delete Remote Branch

git push origin --delete <branch-name>

Delete Local Branch

git branch -d <branch-name>

License and Datasets

  • Your work should be open source under a liberal license (ie. Apache2, MIT)
  • Please use public training and validation datasets so that others can easily replicate your work.

PipelineAI Home

Register for PipelineAI Enterprise Edition

More Resources