Data-centric AI playbook 🏀📚 (open source project)

There are a lot of tricks and great open source tools available to build high quality datasets. However, these resources are often not fully used for two reasons: It is difficult to know all the tools and it is sometimes too much work to integrate them.

We are collecting useful workflow snippets in a data-centric AI playbook. We have just started with a few select plays covering things like:

1️ Understanding data distributions with embeddings

2️ Finding label inconsistencies

3️ …

data data-centric datasets deeplearning open source project quality resources tools tricks work workflow