Datasets:
File size: 553 Bytes
1e58ddd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
---
license: mit
language:
- en
- zh
task_categories:
- text2text-generation
tags:
- math-word-problems
size_categories:
- 100K<n<1M
---
## Dataset Summary
This dataset is derived from the subset **madrylab/gsm8k-platinum** of the **openai/gsm8k** dataset.
### Transformation Method:
1. Template-based question generation.
2. Each question generates 800 similar questions of the same type.
**Note:** This dataset (gsm_800k_beta1.jsonl) has not undergone strict human verification, and approximately `2%` of the questions may contain errors. |