XinnanZhang/DAPO-17k-AIME24-26-suffix

该数据集是一个数学推理数据集，扩展了DAPO-17k-AIME24-26数据集，并在测试集中添加了AMC23问题。数据集结构包括训练集和测试集，训练集包含17,255个DAPO数学问题（去重、框式格式），测试集包含130个竞赛数学问题，其中30个来自AIME 2024，30个来自AIME 2025，30个来自AIME 2026，40个来自AMC 2023（新增）。每个示例包含以下字段：`data_source`（问题来源）、`prompt`（包含问题的用户提示列表）、`ability`（问题类别）、`reward_model`（包含真实答案的字典）和`extra_info`（额外元数据）。

This dataset is a mathematics reasoning dataset that extends DAPO-17k-AIME24-26 with AMC23 problems added to the test set. The dataset structure includes a training set and a test set. The training set contains 17,255 DAPO math problems (deduplicated, boxed format), and the test set contains 130 competition math problems, including 30 from AIME 2024, 30 from AIME 2025, 30 from AIME 2026, and 40 from AMC 2023 (new). Each example contains the following fields: `data_source` (source of the problem), `prompt` (list with user prompt containing the problem), `ability` (problem category), `reward_model` (dictionary with ground truth answer), and `extra_info` (additional metadata).