ai2-adapt-dev/tulu_v3.9_aya_unused

该数据集包含多个特征字段，包括inputs（输入）、targets（目标）、language（语言）、language_code（语言代码）、annotation_type（注释类型）、user_id（用户ID）和messages（消息）。消息字段是一个列表，包含content（内容）和role（角色）两个子字段。数据集分为训练集，包含102,285个样本，总大小为248,585,823.65914428字节。数据集的下载大小为129,755,062字节。

The dataset includes multiple feature fields such as inputs, targets, language, language_code, annotation_type, user_id, and messages. The messages field is a list containing two subfields: content and role. The dataset is divided into a training set, which contains 102,285 samples with a total size of 248,585,823.65914428 bytes. The download size of the dataset is 129,755,062 bytes.