kaiwenw/nov18_oasst_pref_jdpo_llama8b_0.9_chosen_50_reject_50

该数据集包含多个特征，包括prompt（提示）、chosen（选择的回答）、rejected（拒绝的回答）、chosen_pref（选择的偏好）、rejected_pref（拒绝的偏好）和split_suffix（分割后缀），所有特征的数据类型均为字符串。数据集分为训练集和验证集，训练集包含12000个示例，验证集包含714个示例。数据集的下载大小为16319199字节，总大小为72078600.42095993字节。

The dataset contains several features including prompt, chosen, rejected, chosen_pref, rejected_pref, and split_suffix, all of which are of string data type. The dataset is divided into a training set and a validation set, with the training set containing 12,000 examples and the validation set containing 714 examples. The download size of the dataset is 16,319,199 bytes, and the total size is 72,078,600.42095993 bytes.