Datasets:

Modalities:
Text
Formats:
parquet
Languages:
English
ArXiv:
Libraries:
Datasets
pandas
License:
prompteus commited on
Commit
a52c23e
1 Parent(s): 2fe7624

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -1
README.md CHANGED
@@ -65,7 +65,10 @@ into a simple html-like language that BeautifulSoup can parse.
65
 
66
  ## Content and Data splits
67
 
68
- Content and splits correspond to the original gsm8k dataset.
 
 
 
69
  See [gsm8k HF dataset](https://huggingface.co/datasets/gsm8k) and [official repository](https://github.com/openai/grade-school-math) for more info.
70
 
71
 
 
65
 
66
  ## Content and Data splits
67
 
68
+ Content corresponds to the original gsm8k dataset.
69
+
70
+ In this version, we created validation set by sampling 200 random examples from the original train split.
71
+
72
  See [gsm8k HF dataset](https://huggingface.co/datasets/gsm8k) and [official repository](https://github.com/openai/grade-school-math) for more info.
73
 
74