shengjia-toronto/dapo-math-17k-promptprocessed

该数据集包含多个字段，用于存储提示字符串(prompt_str)、解决方案(solution)、数据来源(data_source)、源提示(source_prompt)、能力(ability)、奖励模型(reward_model)等。奖励模型包含地面真实(ground_truth)和风格(style)信息，额外信息(extra_info)包含索引(index)。提示(prompt)字段是一个列表，包含内容(content)和角色(role)。数据集分为训练集(train)，包含14116个示例，大小为17196613字节。但是，README中未提供数据集的具体描述。

The dataset includes multiple fields for storing prompt strings (prompt_str), solutions (solution), data sources (data_source), source prompts (source_prompt), abilities (ability), and reward models (reward_model). The reward model includes ground truth (ground_truth) and style (style) information, while extra info (extra_info) contains an index. The prompt field is a list that includes content (content) and role (role). The dataset is split into a training set (train) with 14116 examples and a size of 17196613 bytes. However, the README does not provide a specific description of the dataset.