sherryy/tulu-3-sft-Bespoke-Stratos-17k

该数据集包含三个主要特征：来源(source)、标识符(id)和消息(messages)。消息字段由内容(content)和角色(role)组成。数据集被划分为训练集(train)，共有956,053个示例，总文件大小为2,914,460,522字节。提供了默认配置，指定了训练集的数据文件路径。The dataset includes three main features: source, id, and messages. The messages field consists of content and role. The dataset is split into a training set (train) with a total of 956,053 examples and a total file size of 2,914,460,522 bytes. A default configuration is provided, specifying the data file path for the training set.