Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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.
|
@@ -35,4 +49,4 @@ The dataset consists of several components, each originating from different sour
|
|
35 |
<strong>Patent</strong><br>
|
36 |
Includes texts from patent documents, providing detailed descriptions of inventions and their applications.<br>
|
37 |
|
38 |
-
</p>
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
task_categories:
|
4 |
+
- text-generation
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
- zh
|
8 |
+
tags:
|
9 |
+
- language model
|
10 |
+
pretty_name: Matrix
|
11 |
+
size_categories:
|
12 |
+
- n>1T
|
13 |
+
---
|
14 |
+
|
15 |
# Matrix
|
16 |
|
17 |
An open-source pretraining dataset containing 4690 billion tokens, this bilingual dataset with both English and Chinese texts is used for training neo models.
|
|
|
49 |
<strong>Patent</strong><br>
|
50 |
Includes texts from patent documents, providing detailed descriptions of inventions and their applications.<br>
|
51 |
|
52 |
+
</p>
|