NuminaMath-CoT AI-MO
Dataset Card for NuminaMath CoT Dataset Summary Approximately 860k math problems, where each solution is formatted in a Chain of Thought (CoT) manner. The sources of the dataset range from Chinese high school math exercises to US and international mathematics olympiad competition problems. The data were primarily collected from online exam paper PDFs and mathematics discussion forums. The processing steps include (a) OCR from the original PDFs, (b) segmentation… See the full description on the dataset page: https://huggingface.co/datasets/AI-MO/NuminaMath-CoT.
- Type
- dataset
- License
- apache-2.0
- Language
- en
- Downloads
- 86,779
- Likes
- 595
- Access
- public
- Files
- 0
Tags
- 文本数据集
- 预训练语料
- 数学推理
- 指令微调
- 多语言语料
- 文本生成
- 大语言模型
README
--- dataset_info: features: - name: source dtype: string - name: problem dtype: string - name: solution dtype: string - name: messages list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 2495457595.0398345 num_examples: 859494 - name: test num_bytes: 2903…