Hugging Face's logo

Datasets:
m-a-p
/
Matrix

Tasks:
Text Generation
Modalities:
Text
Formats:
json
Languages:
English
Chinese
Size:
1M - 10M
Tags:
language model
Libraries:
Datasets
Dask
License:
Dataset card Viewer Files Community
7
dododododo commited on
Commit
8872099
·
verified ·
1 Parent(s): 6577f87

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -1,6 +1,6 @@
1
  # Matrix
2
 
3
- An open-source pretraining dataset with a scale of 4690 billion tokens, used to train neo models.
4
 
5
  ## Dataset Composition
6
 
 
1
  # Matrix
2
 
3
+ An open-source pretraining dataset containing 4690 billion tokens, this bilingual dataset with both English and Chinese texts is used for training neo models.
4
 
5
  ## Dataset Composition
6