argilla/end2end_textclassification

该数据集使用Argilla创建，包含文本字段和问题，供标注者将文本分类为World、Sports、Business和Sci/Tech等类别。数据集还包括建议、元数据和标注指南。数据集结构、数据实例和字段在README中有详细描述。然而，一些部分如Curation Rationale、Source Data、Annotations和Considerations for Using the Data缺乏详细信息。This dataset was created using Argilla, which includes text fields and questions for annotators to classify texts into categories such as World, Sports, Business, and Sci/Tech. The dataset also includes suggestions, metadata, and annotation guidelines. The dataset structure, data instances, and fields are described in detail in the README. However, some sections such as Curation Rationale, Source Data, Annotations, and Considerations for Using the Data lack detailed information.

原始信息汇总

数据集概述

数据集名称

• 名称: end2end_textclassification

数据集大小

• 大小: 1K<n<10K

标签

• 标签: rlfh, argilla, human-feedback

数据集加载

使用Argilla加载

• 安装命令: pip install argilla --upgrade
• 加载代码: python import argilla as rg ds = rg.FeedbackDataset.from_huggingface("argilla/end2end_textclassification")

使用datasets库加载

• 安装命令: pip install datasets --upgrade
• 加载代码: python from datasets import load_dataset ds = load_dataset("argilla/end2end_textclassification")

数据集结构

数据字段

• 文本字段: text (类型: text)
• 问题字段: label (类型: label_selection, 允许值: [World, Sports, Business, Sci/Tech])

数据实例

• Argilla格式: json { "external_id": "record-0", "fields": { "text": "Wall St. Bears Claw Back Into the Black (Reuters) Reuters - Short-sellers, Wall Streetu0027s dwindling\band of ultra-cynics, are seeing green again." }, "metadata": {}, "responses": [], "suggestions": [], "vectors": {} }

• HuggingFace datasets格式: json { "external_id": "record-0", "label": [], "label-suggestion": null, "label-suggestion-metadata": { "agent": null, "score": null, "type": null }, "metadata": "{}", "text": "Wall St. Bears Claw Back Into the Black (Reuters) Reuters - Short-sellers, Wall Streetu0027s dwindling\band of ultra-cynics, are seeing green again." }

数据分割

• 分割: train