r-three/tulu3-sft-clustered8-seed123-mixing0.1

该数据集包含多个分割领域的数据，涵盖高等数学与建模、创意写作与通用任务、多语言与翻译、非英语数学、编程与代码开发、问答与逻辑推理、安全与有害内容、文字问题与算术等方面。每个分割包含文本、来源、原始ID和消息列表等特征字段，消息列表又包含内容和角色字段。数据集总大小约为6.48GB，包含约1,596,785个示例。

The dataset includes multiple splits covering areas such as advanced mathematics and modeling, creative writing and general tasks, multilingual and translation, non-English mathematics, programming and code development, Q&A and logical reasoning, safety and harmful content, and word problems and arithmetic. Each split contains feature fields like text, source, original_id, and messages (with content and role fields). The total dataset size is approximately 6.48GB, containing about 1,596,785 examples.