Datasets:

Modalities:
Text
Formats:
webdataset
Size:
< 1K
Libraries:
Datasets
WebDataset
License:
File size: 462 Bytes
d0d2c88
 
 
 
4d6fed3
 
6f6c237
 
 
1
2
3
4
5
6
7
8
9
10
---
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: