012shin/fake-audio-detection-augmented

该数据集包含音频数据及其对应的标签。音频数据的采样率为16000，标签是一个float64类型的序列。数据集分为训练集和测试集，训练集包含85296个样本，测试集包含50000个样本。整个数据集的下载大小为4362101004字节，数据集大小为4335587407.344字节。

The dataset includes three main features: id (string type), audio (audio type with a sampling rate of 16000), and label (sequence type, float64). The dataset is divided into a training set (containing 85296 samples with a total byte size of 2302804457.344) and a test set (containing 50000 samples with a total byte size of 2032782950.0). The total download size of the dataset is 4362101004 bytes, and the total dataset size is 4335587407.344 bytes. The dataset configuration is named default, with training data file paths as data/train-* and test data file paths as data/test-*.

原始信息汇总

数据集概述

数据集特征

• id: 字符串类型
• audio: 音频数据，采样率为16000
• label: 浮点数序列

数据集分割

• train:
  • 样本数量: 85296
  • 数据大小: 2302804457.344字节
• test:
  • 样本数量: 50000
  • 数据大小: 2032782950.0字节

数据集大小

• 下载大小: 4362101004字节
• 数据集总大小: 4335587407.344字节

配置

• default:
  • train: 数据路径为data/train-*
  • test: 数据路径为data/test-*