qwselfcorr/numia_math_and_augmath_cor20k_20k_20k_2e6_3ep_tmp07_turn2
收藏Hugging Face2025-02-02 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/qwselfcorr/numia_math_and_augmath_cor20k_20k_20k_2e6_3ep_tmp07_turn2
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含索引、提示、答案序列、正确答案、首次奖励、代理标签和二次奖励等字段的数据集。它有一个训练集，包含20000个示例，总大小为53986598字节。数据集的下载大小为21158480字节。
This dataset includes fields such as index, prompt, answer sequence, correct answer, first reward, proxy label, and second reward. It has a training set with 20,000 examples, totaling 53986598 bytes in size. The download size of the dataset is 21158480 bytes.
提供机构：
qwselfcorr
