CONTRIBUTING.md

Contributing to Dataset

Thank you for taking the time to contribute!

We have a set of guidelines that you should follow to ensure cohesion across each dataset

How to contribute to a new dataset

To contribute with a new dataset you'll have to make a pull request (PR). Each dataset must have its directory inside parsers. For example, if I want to create a new parser for imagenet, I'll have to create a directory parsers/imagenet. The directory must have the following files

.
├── __init__.py
├── parser.py
└── README.md

• __init__.py is the classic python __init__ file used for module definition
• parser.py contains the actual parser, a subclass of parsers.Parser
• README.py explain how to get the dataset and how to run the parser

We highly encourage to use types

Check out this example