qwselfcorr/math_and_augmath_cor20k_20k_2e6_2ep_tmp10_turn2
收藏Hugging Face2025-02-01 更新2025-02-15 收录
下载链接:
https://hf-mirror.com/datasets/qwselfcorr/math_and_augmath_cor20k_20k_2e6_2ep_tmp10_turn2
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含以下字段:索引(idx),提示文本(prompt),答案序列(answers),正确答案(gt),首次奖励标识(first_rewards),代理标签(proxy_label),二次奖励标识(second_rewards)。数据集被划分为训练集,包含20000个示例,大小为72939345字节。然而,具体的数据集描述和用途在README文件中未提供。
The dataset includes fields such as index (idx), prompt text (prompt), answer sequence (answers), correct answer (gt), first reward indicator (first_rewards), proxy label (proxy_label), and second reward indicator (second_rewards). The dataset is split into a training set with 20,000 examples, totaling 72939345 bytes in size. However, the specific dataset description and purpose are not provided in the README file.
提供机构:
qwselfcorr


