Dataset Overview

Basic Information

• License: MIT
• Task Category: Question Answering
• Languages: English, Chinese
• Tags: Mathematics, Math QA, Chinese Math QA
• Scale: n<1K

Dataset Description

• Name: GSM8K_zh
• Description: GSM8K_zh is a dataset for Chinese mathematical reasoning. Question‑answer pairs are translated from the GSM8K dataset (https://github.com/openai/grade-school-math/tree/master) using few‑shot prompting with GPT‑3.5‑Turbo.
• Sample Count: 7,473 training samples and 1,319 test samples. Training data are used for supervised fine‑tuning, while test data are for evaluation.
• Sample Structure:
  • Training: includes question_zh (question) and answer_zh (answer) keys.
  • Test: provides only the translated question (question_zh).

Citation

• Reference:

  @article{yu2023metamath,
    title={MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models},
    author={Yu, Longhui and Jiang, Weisen and Shi, Han and Yu, Jincheng and Liu, Zhengying and Zhang, Yu and Kwok, James T and Li, Zhenguo and Weller, Adrian and Liu, Weiyang},
    journal={arXiv preprint arXiv:2309.12284},
    year={2023}
  }