Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,36 @@
|
|
1 |
-
---
|
2 |
-
license:
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: odc-by
|
3 |
+
task_categories:
|
4 |
+
- text-generation
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
tags:
|
8 |
+
- web
|
9 |
+
- common crawl
|
10 |
+
size_categories:
|
11 |
+
- 100B<n<1T
|
12 |
+
---
|
13 |
+
|
14 |
+
# 📚 fineweb-pro
|
15 |
+
|
16 |
+
<p align="center">
|
17 |
+
<img src="prox-teaser.png">
|
18 |
+
</p>
|
19 |
+
|
20 |
+
[ArXiv](http://arxiv.org/abs/2409.17115) | [Models](https://huggingface.co/collections/gair-prox/prox-general-models-65f1674f0607712c4d6eec76) | [Code](https://github.com/GAIR-NLP/ProX)
|
21 |
+
|
22 |
+
DCLM-pro is refined from [DCLM](https://huggingface.co/datasets/mlfoundations/dclm-baseline-1.0-parquet) using the **ProX** refining framework.
|
23 |
+
It contains about >500B high quality tokens, ready for general language model pre-training.
|
24 |
+
|
25 |
+
## License
|
26 |
+
DCLM-pro is based on DCLM, which is made available under an cc-by-4.0 license.
|
27 |
+
|
28 |
+
### Citation
|
29 |
+
```
|
30 |
+
@article{zhou2024programming,
|
31 |
+
title={Programming Every Example: Lifting Pre-training Data Quality like Experts at Scale},
|
32 |
+
author={Zhou, Fan and Wang, Zengzhi and Liu, Qian and Li, Junlong and Liu, Pengfei},
|
33 |
+
journal={arXiv preprint arXiv:2409.17115},
|
34 |
+
year={2024}
|
35 |
+
}
|
36 |
+
```
|