Update README.md
Browse files
README.md
CHANGED
@@ -2,14 +2,12 @@
|
|
2 |
license: mit
|
3 |
language:
|
4 |
- en
|
5 |
-
datasets:
|
6 |
-
- Skylion007/openwebtext
|
7 |
pipeline_tag: text-generation
|
8 |
---
|
9 |
-
# OpenSML: An Family of Small Language Models
|
10 |
|
11 |
*William Zebrowski*
|
12 |
|
13 |
Introduce **OpenSML**, a series of **Open** **SM**a**L**l Language Models. These models arcitecture are built stricly will Apple's [MLX](https://ml-explore.github.io/mlx/build/html/index.html#) framework.
|
14 |
|
15 |
-
The pre-training dataset
|
|
|
2 |
license: mit
|
3 |
language:
|
4 |
- en
|
|
|
|
|
5 |
pipeline_tag: text-generation
|
6 |
---
|
7 |
+
# OpenSML: An Family of Open Small Language Models
|
8 |
|
9 |
*William Zebrowski*
|
10 |
|
11 |
Introduce **OpenSML**, a series of **Open** **SM**a**L**l Language Models. These models arcitecture are built stricly will Apple's [MLX](https://ml-explore.github.io/mlx/build/html/index.html#) framework.
|
12 |
|
13 |
+
The pre-training dataset is a slice of OpenWebText dataset with approximately 2.3 billion tokens.
|