Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,21 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- zh
|
| 6 |
+
task_categories:
|
| 7 |
+
- text2text-generation
|
| 8 |
+
tags:
|
| 9 |
+
- math-word-problems
|
| 10 |
+
size_categories:
|
| 11 |
+
- 100K<n<1M
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
## Dataset Summary
|
| 15 |
+
This dataset is derived from the subset **madrylab/gsm8k-platinum** of the **openai/gsm8k** dataset.
|
| 16 |
+
|
| 17 |
+
### Transformation Method:
|
| 18 |
+
1. Template-based question generation.
|
| 19 |
+
2. Each question generates 800 similar questions of the same type.
|
| 20 |
+
|
| 21 |
+
**Note:** This dataset (gsm_800k_beta1.jsonl) has not undergone strict human verification, and approximately `2%` of the questions may contain errors.
|