| license: cc-by-sa-4.0 | |
| We introduce the Million Authors Corpus (MAC), a novel dataset encompassing contributions from dozens of languages on Wikipedia. | |
| It includes only long and contiguous textual chunks taken from Wikipedia edits and links those texts to their authors. | |
| This repository includes the data used for training and testing only. The full MAC dataset is shared in Zenodo repository: https://zenodo.org/records/15538126 | |
| Here are few examples: | |