skvarre/swedish-instruct-data-chatgpt4
收藏资源简介:
--- language: - sv size_categories: - 1K<n<10K dataset_info: features: - name: human dtype: string - name: gpt dtype: string splits: - name: train num_bytes: 715231 num_examples: 1363 download_size: 431901 dataset_size: 715231 configs: - config_name: default data_files: - split: train path: data/train-* tags: - synthetic --- This small synthetic instruction dataset contains question-answer pairs in Swedish that highlight a wide range of topics related to Sweden. It was generated using ChatGPT-4. Due to the data being machine generated, it has to be emphasized that there is no guarantee that the information in the dataset is correct; nor should it be seen as a complete dataset that reflects a fair picture of Sweden-related topics. The amount of examples of each topic is random. The data was generated based on prompts around the following topics (in no specific order): ``` - Famous people from Sweden: - Examples not possible to answer as an AI assistant - Unethical questions that an AI Assistant shouldn't help with - Mathematics - Ways to learn a new language - Swedish movies - Swedish books - Swedish music: - Swedish food: - Swedish grammar - Swedish spelling - Swedish language - Swedish politics - Swedish healthcare - Swedish history - Swedish geography - Swedish culture - Swedish economy - Swedish technology - Swedish cities: - Swedish provinces: - Religion in Sweden - Swedish companies: - General knowledge (Allmänbildning) - Swedish education: - Swedish authorities - Neighboring countries - Swedish transportation - Sports in Sweden - Swedish fashion - Swedish art - Swedish TV-shows - Swedish accents & dialects - Swedish news sources - Swedish mythology & folklore - Swedish holidays & celebrations - Pronunciation of swedish words - Swedish welfare - Swedish legal system - Roleplay scenarios - Questions with alternatives - Swedish nature reserves and national parks ```
数据集概述
语言
- 瑞典语
数据规模
- 1K<n<10K
数据集信息
- 特征
human: 字符串类型gpt: 字符串类型
- 分割
train- 字节数: 715231
- 样本数: 1363
- 下载大小: 431901
- 数据集大小: 715231
配置
- 默认配置
- 数据文件
train- 路径:
data/train-*
- 路径:
- 数据文件
标签
- 合成数据
数据描述
- 该小型合成指令数据集包含瑞典语的问答对,涵盖与瑞典相关的广泛主题。数据由ChatGPT-4生成。
- 由于数据是机器生成的,不能保证信息的正确性,也不应视为反映瑞典相关主题全面情况的完整数据集。
- 每个主题的示例数量是随机的。
- 数据生成基于以下主题(无特定顺序):
- 瑞典名人
- 无法回答的例子(作为AI助手)
- 不道德的问题(AI助手不应协助)
- 数学
- 学习新语言的方法
- 瑞典电影
- 瑞典书籍
- 瑞典音乐
- 瑞典食物
- 瑞典语法
- 瑞典拼写
- 瑞典语言
- 瑞典政治
- 瑞典医疗保健
- 瑞典历史
- 瑞典地理
- 瑞典文化
- 瑞典经济
- 瑞典技术
- 瑞典城市
- 瑞典省份
- 瑞典宗教
- 瑞典公司
- 一般知识(Allmänbildning)
- 瑞典教育
- 瑞典当局
- 邻国
- 瑞典交通
- 瑞典体育
- 瑞典时尚
- 瑞典艺术
- 瑞典电视节目
- 瑞典口音与方言
- 瑞典新闻来源
- 瑞典神话与民间传说
- 瑞典节日与庆典
- 瑞典单词发音
- 瑞典福利
- 瑞典法律体系
- 角色扮演场景
- 有选项的问题
- 瑞典自然保护区和国家公园



