Datasets:

bertin-project
/

mc4-sampling

Tasks:

Text Generation

Sub-tasks:

language-modeling

Languages:

Size:

ArXiv:

License:

Dataset card Files Files and versions Community

versae commited on 28 days ago

Commit

1b37f5c

•

1 Parent(s): 861da93

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -170,7 +170,7 @@ The original dataset is mC4, the multilingual colossal, cleaned version of Commo
 108 languages are available and are reported in the [`mc4` dataset](https://huggingface.co/datasets/mc4#dataset-summary).
-You can load the mC4 subset of any language like this:
 ```python
 from datasets import load_dataset

 108 languages are available and are reported in the [`mc4` dataset](https://huggingface.co/datasets/mc4#dataset-summary).
+You can load the mC4 subset of any language like this (with default *random* sampling):
 ```python
 from datasets import load_dataset