Unsupervised Cross-lingual Representation Learning at Scale
Paper • 1911.02116 • Published • 4
Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
Portuguese
Size:
10M - 100M
ArXiv:
Dataset Viewer
The dataset viewer is not available for this dataset.
Unexpected token '<', "<html>
<h"... is not valid JSON
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
C100-PT
CC100-PT is the is the portuguese subset from C100. C100 was created for training the multilingual Transformer XLM-R, containing two terabytes of cleaned data from 2018 snapshots of the Common Crawl project in 100 languages.
- Downloads last month
- 10