MLP-SEMO/CDT_qa_datasets

该数据集包含三个不同的配置：fineweb、fineweb_preprocessed和fineweb_qa。每个配置都包含context和question两个特征，fineweb_qa配置还包含response特征。数据集仅包含训练集，每个配置的训练集都有不同的字节数和示例数。fineweb配置的训练集包含9,581,376个示例，fineweb_preprocessed配置的训练集包含9,259,128个示例，fineweb_qa配置的训练集包含9,026,269个示例。

The dataset includes three different configurations: fineweb, fineweb_preprocessed, and fineweb_qa. Each configuration contains features such as context and question, with the fineweb_qa configuration additionally including the response feature. The dataset only contains a training set, with each configurations training set having different byte sizes and numbers of examples. The fineweb configurations training set contains 9,581,376 examples, the fineweb_preprocessed configurations training set contains 9,259,128 examples, and the fineweb_qa configurations training set contains 9,026,269 examples.

原始信息汇总

数据集概述

数据集配置

1. fineweb

• 特征:
  • context: 字符串类型
  • question: 字符串类型
• 分割:
  • train:
    • 字节数: 29,535,642,491
    • 样本数: 9,581,376
• 下载大小: 18,140,288,984 字节
• 数据集大小: 29,535,642,491 字节
• 数据文件路径: fineweb/train-*

2. fineweb_preprocessed

• 特征:
  • context: 字符串类型
  • question: 字符串类型
• 分割:
  • train:
    • 字节数: 27,445,160,941.262375
    • 样本数: 9,259,128
• 下载大小: 16,630,421,213 字节
• 数据集大小: 27,445,160,941.262375 字节
• 数据文件路径: fineweb_preprocessed/train-*

3. fineweb_qa

• 特征:
  • context: 字符串类型
  • question: 字符串类型
  • response: 字符串类型
• 分割:
  • train:
    • 字节数: 29,377,996,197
    • 样本数: 9,026,269
• 下载大小: 18,143,018,611 字节
• 数据集大小: 29,377,996,197 字节
• 数据文件路径: fineweb_qa/train-*