Check out OmdenaLore, the Largest AI Python Library for the Real World

Imagine an open-source Python library, which allows you to build, within days, an end-to-end data science pipeline that is ready for production! Additionally, the library is not just a codebase but also a knowledge source helping in every stage of development while solving some of the world´s most challenging problems. 

Sounds imaginary, right? But that is exactly what OmdenaLore is. 

OmdenaLore is developed collaboratively

We went on a mission to build OmdenaLore, an open-sourced data science package that provides comprehensive and ready-to-use Python classes and functions to solve almost any machine learning problem in an accelerated manner. This Python library is built and maintained collaboratively by the global AI community, thus making it more inclusively and ethically developed. 

Let’s dig deeper into what OmdenaLore has to offer!

How you can use this Python library

Intuitive Documentation

Let’s be honest, often we would spend hours googling solutions to fix an error rather than spending enough time on the documentation. You know that the documentations are sometimes non-intuitive and might feel incomplete. So we designed the OmdenaLore documentation by resembling Wikipedia’s structure and coming with a standard Python library’s documentation theme to make you feel at home. But what’s really new? It’s not just the looks, the documentation is collaborator-sourced, so you can learn from collaborators who faced the same challenges. With OmdenaLore documentation, you also get access to Omdena’s publicly available knowledge base. Use the built-in search feature to sort through your interest and you’ll be in the land full of knowledge. With structured repositories and articles, this list of resources might “shoot” all of your troubles.

A well-packaged library

We don’t just claim it, we are available in the python package index (pypi.org) so you can install OmdenaLore and speed up the development time of your projects right away.

Just install the library by using the command below and you are good to go.

pip install omdenalore

Check out the project on Pypi here.

A Structured Repository

The repository offers a combination of a collaborator-sourced codebase, documentation, and knowledge base. OmdenaLore follows a structured approach based on Machine learning and data processing domains. Explore the components folder to find the codebase of your interest. Feeling lost? Head on to the samples folder and run the notebooks to learn more about how code works from loading the data to applying the model.

When you are a part of OmdenaLore, we’ll guide you throughout your contribution. 

How you can contribute 

What to contribute

Our contribution scopes are really wide. In essence, you can contribute any data science code block as long as it does not violate any contracts or IP. When you apply with the link, you will be given access to the repository and to our slack channel. That’s it! 

Now that you have push access, you can clone the repository and make a new branch to work on. At this point, you will also find our Slack group with other changemakers from around the world where you can discuss and brainstorm coding ideas and ask any questions. OmdenaLore does not demand deep expertise necessarily but seeks motivated individuals who want to contribute, grow, and make an impact.

Our unique collaboration process

Issue sections

Our DAGsHub repository uses Issues to group together tasks on adding more functionalities. Any contributor who is interested in working on an Issue can simply use the issue link in their Pull Request. We try our best to keep the Issues succinct and discrete with numbered lists of tasks. This encourages a more fine-grained collaborative practice as each submitted Pull Request adds distinct new features. 

Pull Request (PR)

Any contribution in OmdenaLore is done through a Pull Request. To make sure the main branch of the repository contains production-ready and properly formatted code, we have worked with DAGsHub to lock the main branch. To make a contribution, contributors make a new branch from main, add their changes there, and make a Pull Request. Our Pull Requests are reviewed and merged really fast with an average time of 2-3 hours. Occasionally, we ask the PR author to make a few minor changes before the PR is merged. For more information on contribution guidelines, please refer to https://omdenaai.github.io/guideline.html.

Sprints

Coding is more fun when you do it with others! Every week, we try to host at least three coding sprints on our Discord server. The sprints have no prerequisite, anyone who wishes to participate joins the voice call. There is also no specific agenda we follow in the sprints. After a brief introduction from new contributors, everyone starts coding in the issue they want to contribute on, or on any other part of the repository. Q&A happens throughout the sprint, sometimes even in different voice channels to not disturb the people coding (we will also listen to fun music together :)).

Doubt or question solving – In addition to the coding sprints, we have an active slack channel within the Omdena community where we discuss code, trade suggestions on new OmdenaLore features, and share feedback. We also have an Airtable form through which feedback can be shared directly with us.

Experts 

Of course, no one person can be an expert of all. Depending on the theme of the coding sprint we are conducting (Computer Vision, Satellite Imagery, Natural Language Processing, etc.), we invite our community members to apply to become an expert for the sprint, and for that general topic in the slack channel as well. This allows contributors to ask their questions about the code they are trying to contribute to the repository to make sure it will perform well and reliably for other users.

Help to build the future of AI ethically and inclusively 

OmdenaLore is not a replacement for traditional packages such as TensorFlow and PyTorch. It is used to build the same pipeline as the other packages but in a much shorter time span. Additionally, OmdenaLore aims to foster a strong open source community where new functionalities and features can be added continuously. We want this initiative to be available to all Data Scientists around the world regardless of whether they wish to be contributors. Keeping this in mind, we will be releasing OmdenaLore soon on PyPi as an installable Python Data Science package. Since it’s a project-focused Python library, you can make an impact on multiple live projects that OmdenaLore supports. 

Together with you, OmdenaLore is going to be the biggest AI for the Good Python library! Join us at the official launch on Sept 7 2021

Meanwhile, Enjoy an OmdenaLore demo