ELI5: Bento – Interactive Pocket book that Empowers Growth Collaboration & Greatest Practices
In this series we discuss many open source projects; Today, however, it will be a little different. We’ll learn more about Bento, a project based on the open source Jupyter project and now used internally at Facebook to help data scientists and ML engineers share code and collaborate interactively.
In this blog post we explain Bento in an easy to understand way (or as it is commonly known on the internet, ELI5 Youtube Channel.
What is bento?
Before we can understand Bento, we need to understand what Jupyter is. Jupyter notebooks are an open source web tool widely used by data scientists and machine learning engineers to create and share documents that contain live code, equations, visualizations, and explanatory text with your fellow human beings.
While Jupyter is great at sharing code and documents and has been widely used on Facebook, our teams realized that Jupyter needed to be tweaked to better integrate with the Facebook stack and improve the authoring experience. This could dramatically improve the experience of using this tool, improve code quality, and promote best practices.
By authoring experience, we mean adding features that are common in most other IDEs, such as linters, formatters, and code autocompletion. And the direct integration of Bento into the Facebook stack helps teams to create, share and review their work more easily and securely – be it better integration with other data tools on Facebook, more consistency with other internal tools or stricter data protection standards Data analysis protection.
And so Bento was born in 2017! The name Bento refers to a bento box because it combines analytics and machine learning in one “box”.
Bento is the officially supported way to use Jupyter notebooks on Facebook. You can share documents with live code, equations, visualizations, and text. It bundles popular external libraries for the Python scientific data stack such as numpy, pandas, as well as many internal libraries for data analysis.
Notebooks have been a great innovation, especially for the data science and machine learning community to quickly write, iterate, and improve code. Enhancing the authoring experience by adding these features improves productivity, helps users quickly spot and identify bugs or code problems, and protects the developer’s sanity.
Where is Bento used?
As of 2020, nearly 9,000 data developers are using Bento every month on Facebook.
Within Facebook, Bento users are diverse. For example, non-technical roles in notebooks annotate to provide feedback, machine learning scientists use it to test models for fairness, and threat investigators use it to keep records of their response work.
At the time of this writing, Bento is not yet open source, but our teams are exploring the possibility, particularly with the assistive authoring platform we developed to make writing code in Bento faster and easier.
Where can I find out more?
Want to learn more about Bento? Check out this video by Tanya Rai (PM on Facebook) for an overview of Bento.
About the ELI5 series
In a series of short videos, one of our developer advocates in the Facebook Open Source Team explains a Facebook Open Source project in an easy-to-understand and easy-to-use way.
For each of these videos we will write an accompanying blog post (like the one you are reading right now), which you can find on our Youtube channel.
To learn more about Facebook Open Source, visit our open source site, subscribe to our Youtube channel or follow us on Twitter and Facebook.
Interested in working with open source on Facebook? Check out our open source related job postings on our careers page by taking this short survey.