---
title: README
emoji: 🐛
colorFrom: purple
colorTo: yellow
sdk: static
pinned: false
---

# The Common Pile

We are a group of researchers working together to collect and curate openly licensed and public domain data for training large language models.
So far, we have released:

- [The Common Pile v0.1](https://huggingface.co/collections/common-pile/common-pile-v01-68307d37df48e36f02717f21), an 8 TB dataset of text from over 30 diverse sources
- Our paper: [The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text](https://huggingface.co/papers/2506.05209)
- [Comma v0.1-1T](https://huggingface.co/common-pile/comma-v0.1-1t) and [Comma v0.1-2T](https://huggingface.co/common-pile/comma-v0.1-2t), 7B parameter LLMs trained on text from the Common Pile v0.1
- The [training dataset](https://huggingface.co/datasets/common-pile/comma_v0.1_training_dataset) used to train the Comma v0.1 models
- Our [code](https://github.com/r-three/common-pile/) for collecting data from each source

If you're interested in contributing, please [open an issue on GitHub](https://github.com/r-three/common-pile/issues/new)!