biunlp/HeSum

该数据集名为HeSum，包含两个主要特征：summary（摘要）和article（文章），均为字符串类型。数据集分为训练集、验证集和测试集，分别包含8000、1000和1000个样本。训练集大小为98933510字节，验证集大小为12217867字节，测试集大小为13227741字节。总下载大小为63278508字节，数据集总大小为124379118字节。

The dataset named HeSum contains two main features: summary and article, both of which are of string type. The dataset is divided into training, validation, and test sets, containing 8000, 1000, and 1000 samples respectively. The training set size is 98933510 bytes, the validation set size is 12217867 bytes, and the test set size is 13227741 bytes. The total download size is 63278508 bytes, and the total dataset size is 124379118 bytes.

原始信息汇总

数据集卡片 "HeSum"

数据集信息

特征

• summary: 类型为字符串
• article: 类型为字符串

分割

• train:
  • 字节数: 98933510
  • 样本数: 8000
• validation:
  • 字节数: 12217867
  • 样本数: 1000
• test:
  • 字节数: 13227741
  • 样本数: 1000

大小

• 下载大小: 63278508 字节
• 数据集大小: 124379118 字节