five

wikihow

收藏
魔搭社区2025-11-27 更新2025-01-11 收录
下载链接:
https://modelscope.cn/datasets/sentence-transformers/wikihow
下载链接
链接失效反馈
官方服务:
资源简介:
# Dataset Card for WikiHow Summaries This dataset is a collection of text-summary pairs collected from WikiHow. See [WikiHow pairs dataset](https://github.com/pvl/wikihow_pairs_dataset) for additional information. This dataset can be used directly with Sentence Transformers to train embedding models. ## Dataset Subsets ### `pair` subset * Columns: "text", "summary" * Column types: `str`, `str` * Examples: ```python { 'text': "Wisdom teeth are the molars furthest back on both sides of your upper and lower jaw. [1] These four teeth are the last ones to erupt or grow out of your gums and become functional — this typically occurs during the late teen years or early adulthood; however, sometimes wisdom teeth don't erupt at all or only partially erupt and become impacted if there isn't enough room in your jaw or mouth.", 'summary': 'Tell Between an Erupting and Impacted Wisdom Tooth', } ``` * Collection strategy: Reading the WikiHow dataset from [embedding-training-data](https://huggingface.co/datasets/sentence-transformers/embedding-training-data). * Deduplified: No

# WikiHow 摘要数据集卡片 本数据集为从WikiHow平台采集的文本与摘要配对集合,更多详细信息可参阅[WikiHow配对数据集](https://github.com/pvl/wikihow_pairs_dataset)。 本数据集可直接配合Sentence Transformers用于嵌入模型的训练。 ## 数据集子集 ### `pair` 子集 * 字段名称:`text`、`summary` * 字段类型:`str`、`str` * 示例: python { 'text': "Wisdom teeth are the molars furthest back on both sides of your upper and lower jaw. [1] These four teeth are the last ones to erupt or grow out of your gums and become functional — this typically occurs during the late teen years or early adulthood; however, sometimes wisdom teeth don't erupt at all or only partially erupt and become impacted if there isn't enough room in your jaw or mouth.", 'summary': 'Tell Between an Erupting and Impacted Wisdom Tooth', } * 采集策略:从[embedding-training-data](https://huggingface.co/datasets/sentence-transformers/embedding-training-data)中读取WikiHow数据集 * 去重处理:否
提供机构:
maas
创建时间:
2025-01-06
搜集汇总
数据集介绍
main_image_url
背景与挑战
背景概述
该数据集包含从WikiHow收集的文本与摘要配对数据,可用于直接训练Sentence Transformers嵌入模型。数据格式包括'text'和'summary'两列字符串字段,如示例所示。
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务