Asap7772/ogmath5_onpolicy_multiturn_cummrew_prefix0.2_roll4_maxrev100
收藏Hugging Face2024-09-23 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/Asap7772/ogmath5_onpolicy_multiturn_cummrew_prefix0.2_roll4_maxrev100
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个特征，如输入、输出前缀、输出、下一个输出前缀、下一个输出、终端、完成、轨迹、步骤、标签、奖励、是否被选择、rtg、未折扣的rtg、文本和索引级别等。数据集分为训练集和测试集，分别包含851855和874102个示例，总大小为15804019682字节。
The dataset includes multiple features such as input, output prefix, output, next output prefix, next output, terminal, completion, trajectory, step, label, reward, is chosen, rtg, undiscounted rtg, text, and index level. The dataset is divided into training and test sets, containing 851,855 and 874,102 examples respectively, with a total size of 15,804,019,682 bytes.
提供机构：
Asap7772
