OpenMathInstruct
收藏arXiv2025-09-30 收录
官方服务：
资源简介：
该数据集包含了1,800个数学推理问题，每个问题都附有确切的答案，这些数据被用来提炼出长链式的思维过程回应。此外，该数据集还用于对大型语言模型进行微调，以适应长思考和短思考的推理风格。该数据集的规模为1,800个样本，任务类型是数学推理。
This dataset contains 1,800 mathematical reasoning questions, each accompanied by a definitive answer. These data are utilized to elicit long-chain thought process responses. Additionally, this dataset is employed for fine-tuning large language models (LLMs) to adapt to both long-chain and short-chain reasoning styles. The dataset comprises 1,800 samples, and its task type is mathematical reasoning.
