Datasets:

Modalities:
Text
Formats:
json
Languages:
English
ArXiv:
Libraries:
Datasets
pandas
License:
yuzc19 commited on
Commit
56da1b0
·
verified ·
1 Parent(s): 77ad05b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -2
README.md CHANGED
@@ -3,8 +3,11 @@ language:
3
  - en
4
  size_categories:
5
  - 10K<n<100K
 
6
  ---
7
 
8
- This is the RL training dataset from RePro: Training Language Models to Faithfully Recycle the Web for Pretraining.
9
 
10
- Each entry contains two columns, dataman_score and text.
 
 
 
3
  - en
4
  size_categories:
5
  - 10K<n<100K
6
+ license: apache-2.0
7
  ---
8
 
9
+ This is the RL training data for [repro-rephraser-4B](https://huggingface.co/cx-cmu/repro-rephraser-4B) from [RePro: Training Language Models to Faithfully Recycle the Web for Pretraining](https://huggingface.co/papers/2510.10681).
10
 
11
+ Each entry contains two columns, dataman_score and text.
12
+
13
+ Code: https://github.com/cxcscmu/RePro