Datasets:

BramVanroy
/

fietje-2-data

Formats:

Size:

Libraries:

Dataset card Viewer Files Files and versions Community

BramVanroy commited on Jun 4, 2024

Commit

072edee

·

verified ·

1 Parent(s): cf78da5

Update README.md

Files changed (1) hide show

README.md +2 -0

README.md CHANGED Viewed

@@ -24,3 +24,5 @@ configs:
   - split: test
     path: data/test-*
 ---

   - split: test
     path: data/test-*
 ---
+The data to train [fietje-2](https://huggingface.co/BramVanroy/fietje-2). It contains around 28B tokens taken from the Dutch parts of [wikipedia](https://huggingface.co/datasets/wikimedia/wikipedia) and [CulturaX](https://huggingface.co/datasets/uonlp/CulturaX).