Skip to content

Latest commit

 

History

History
33 lines (28 loc) · 1.66 KB

README.md

File metadata and controls

33 lines (28 loc) · 1.66 KB

TensorFlow I/O Tutorials

TensorFlow I/O welcomes and highly encourages tutorial contributions.

How To Contribute

I/O tutorials are created using Google Colab and the jupyter notebooks are saved to this directory in the repository. To do this, follow the below steps:

  1. Create a new branch on your fork of TensorFlow I/O.
  2. Goto Google Colab and start a new notebook using addons example template: notebook template.
  3. Edit the the links for the "View source on GitHub" and "Run in Google Colab" URL boxes so that they match the name of your new example notebook.
  4. Follow the guidelines of the template.
  5. "Save a copy in Github" and select your new branch. The notebook should be named subpackage_submodule.
  6. After step 5, the notebook will be committed to your branch directly from colab. However, to check for linting issues and to auto format your notebook, pull the changes to your system and run the following from the io directory:
    $ sudo python3 -m pip install setuptools
    $ sudo python3 -m pip install -U git+https://github.com/tensorflow/docs
    $ echo "Auto format the notebooks: "
    $ find docs -name '*.ipynb' | xargs python3 -m tensorflow_docs.tools.nbfmt
    $ echo "Check for failed lint: "
    $ find docs -name '*.ipynb' | xargs python3 -m tensorflow_docs.tools.nblint --arg=repo:tensorflow/io
  7. Update docs/tutorials/_toc.yaml with the notebook details (please refer existing entries).
  8. Submit the branch as a PR on the TF-I/O Github