Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
json
Languages:
English
Size:
10M - 100M
ArXiv:
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -5,4 +5,18 @@ task_categories:
|
|
| 5 |
language:
|
| 6 |
- en
|
| 7 |
pretty_name: SlimPajama-627B-divided
|
| 8 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
language:
|
| 6 |
- en
|
| 7 |
pretty_name: SlimPajama-627B-divided
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
### Dataset Description:
|
| 11 |
+
This dataset is the same as the content of [cerebras/SlimPajama-627B](https://huggingface.co/datasets/cerebras/SlimPajama-627B). The storage format and fields of data in each file are also the same. We have divided data from different sources based on the "redpajama_setname" field in cerebras/SlimPajama-627B, which is convenient for future related research.
|
| 12 |
+
The dataset consists of 15967 jsonl files and is ~ 883G compressed.
|
| 13 |
+
|
| 14 |
+
### License:
|
| 15 |
+
Please refer to the licenses of the data subsets you use.
|
| 16 |
+
- Common Crawl Foundation Terms of Use
|
| 17 |
+
- C4 license
|
| 18 |
+
- GitHub was limited to MIT, BSD, or Apache licenses only
|
| 19 |
+
- Books: the_pile_books3 license and pg19 license
|
| 20 |
+
- ArXiv Terms of Use
|
| 21 |
+
- Wikipedia License
|
| 22 |
+
- StackExchange license on the Internet Archive
|