HuggingFaceTB/MagPie-Pro-300k-MT

该数据集包含了用于训练和测试的对话数据。数据集中的特征包括模型类型、生成输入配置（温度和top_p参数）、唯一标识符和消息内容（包括内容和角色）。训练集包含270,000个示例，大小为1,499,839,599.6字节，测试集包含30,000个示例，大小为166,648,844.4字节。数据集总大小为1,666,488,444.0字节。

The dataset includes conversation data for training and testing. Features of the dataset include model type, generation input configuration (temperature and top_p parameters), unique identifier, and message content (including content and role). The training set contains 270,000 examples with a size of 1,499,839,599.6 bytes, and the test set contains 30,000 examples with a size of 166,648,844.4 bytes. The total size of the dataset is 1,666,488,444.0 bytes.