Datasets:

Modalities:
Text
Formats:
parquet
ArXiv:
14kwonss commited on
Commit
3efdec6
·
verified ·
1 Parent(s): 76da679

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -6
README.md CHANGED
@@ -6,12 +6,12 @@ It contains text data covering **713 African languages**, making it one of the m
6
 
7
  ## Dataset Features
8
 
9
- - **prompt** *(string)*: Description here.
10
- - **text** *(string)*: Description here.
11
- - **token_length** *(int64)*: Description here.
12
- - **audience** *(string)*: Description here.
13
- - **format** *(string)*: Description here.
14
- - **seed_data** *(string)*: Description here.
15
 
16
  ## Loading the dataset
17
 
 
6
 
7
  ## Dataset Features
8
 
9
+ - **sentence** *(string)*: The text sample.
10
+ - **ISO-639-3** *(string)*: The ISO 639-3 language code.
11
+ - **script** *(string)*: The writing script (e.g., `Latn`, `Arab`).
12
+ - **domain** *(string)*: Source/domain category of the sample.
13
+ - **lang_family_1 ~ 5** *(string)*: Language family hierarchy.
14
+ - **label** *(string)*: The final label used for classification.
15
 
16
  ## Loading the dataset
17