maxseats/aihub-464-preprocessed-680GB-set-52

该数据集包含音频数据和对应的标签，音频数据的采样率为16000Hz。数据集分为训练集、测试集和验证集，分别包含41327、5166和5166个样本。每个样本包含音频数据、标签和输入特征。输入特征是一个浮点数序列。数据集的下载大小为21405890286字节，总大小为59918777453.00001字节。

The dataset includes three main features: audio, labels, and input features. The audio feature has a sampling rate of 16000 Hz, the labels feature is of string type, and the input features are sequences of float32. The dataset is divided into train, test, and valid splits, containing 41327, 5166, and 5166 examples respectively. The total download size is 21405890286 bytes, and the total dataset size is 59918777453.00001 bytes. The dataset configuration is set to default, with data file paths named according to the split type.

原始信息汇总

数据集概述

数据特征

• audio:
  • 采样率: 16000
• labels:
  • 数据类型: string
• input_features:
  • 序列类型: float32

数据集划分

• train:
  • 字节数: 47934789984.32279
  • 样本数: 41327
• test:
  • 字节数: 5991993734.338605
  • 样本数: 5166
• valid:
  • 字节数: 5991993734.338605
  • 样本数: 5166

数据集大小

• 下载大小: 21405890286
• 总大小: 59918777453.00001

配置

• config_name: default
  • 数据文件路径:
    • train: data/train-*
    • test: data/test-*
    • valid: data/valid-*