bigstupidhats/dynasample_align

该数据集包含多种语言的数据，每种语言的数据集被分割为不同的部分，包括阿拉伯语、德语、英语、西班牙语、印地语、俄语、斯瓦希里语和中文。每个分割的数据集包含多个字段，如instruction_ori、input、id、output、instruction、Qwen2.5-7B_ifd和conversations。conversations字段是一个包含content和role的列表。数据集的总下载大小为520146394字节，总数据集大小为1050811896.7367983字节。

This dataset contains data in multiple languages, with each languages dataset divided into different splits, including Arabic, German, English, Spanish, Hindi, Russian, Swahili, and Chinese. Each split of the dataset includes multiple fields such as instruction_ori, input, id, output, instruction, Qwen2.5-7B_ifd, and conversations. The conversations field is a list containing content and role. The total download size of the dataset is 520146394 bytes, and the total dataset size is 1050811896.7367983 bytes.