ajmangus/qm_charlie_mixture_1.0e

该数据集包含多个特征，如alice_label、bob_label、charlie_label等布尔类型标签，difficulty整数类型标签，statement字符串类型，choices字符串序列，character字符串类型，以及一个class_label类型的label。数据集分为train、validation和test三个部分，分别包含199999、19999和19999个样本。此外，还提供了数据集的下载大小和总大小。

The dataset includes multiple features such as alice_label, bob_label, charlie_label, etc., all of which are boolean type. difficulty is of integer type, statement is of string type, and choices is a sequence of strings. character and label are of string type, where label is a classification label containing False and True categories. The dataset is divided into training, validation, and test sets, containing 199999, 19999, and 19999 samples respectively. The total download size of the dataset is 7673676 bytes, and the total dataset size is 28874797.0 bytes.

原始信息汇总

数据集概述

数据特征

• alice_label: 布尔类型
• bob_label: 布尔类型
• charlie_label: 布尔类型
• difficulty: 64位整数类型
• statement: 字符串类型
• choices: 字符串序列
• character: 字符串类型
• label: 分类标签，包含两个类别：
  • 0: False
  • 1: True

数据分割

• train:
  • 字节数: 24063561.333333332
  • 样本数: 199999
• validation:
  • 字节数: 2403257.0
  • 样本数: 19999
• test:
  • 字节数: 2407978.6666666665
  • 样本数: 19999

数据集大小

• 下载大小: 7673676 字节
• 数据集大小: 28874797.0 字节

配置

• default:
  • train: data/train-*
  • validation: data/validation-*
  • test: data/test-*