ae-mask-distilled-libri-one-sec-cv12/chunk_17

该数据集主要用于音频处理任务，包含音频和逻辑值两种特征。音频特征的采样率为16000，逻辑值特征为float32类型。数据集包含一个训练集，共有5631个样本，总大小为180524933.08526438字节。数据集的下载大小为166798282字节。配置文件中有一个默认配置，训练数据文件路径为data/train-*。

This dataset is primarily used for audio processing tasks, featuring audio and logits as its main characteristics. The audio feature has a sampling rate of 16000, and the logits feature is of float32 type. The dataset includes one training set with 5631 samples, totaling 180524933.08526438 bytes in size. The download size of the dataset is 166798282 bytes. There is a default configuration in the config file, with the training data file path being data/train-*.

原始信息汇总

数据集概述

特征

• audio:
  • 数据类型: audio
  • 采样率: 16000
• logits:
  • 数据类型: float32

数据分割

• train:
  • 样本数量: 5631
  • 数据大小: 180524933.08526438 字节

数据集大小

• 下载大小: 166798282 字节
• 总数据大小: 180524933.08526438 字节

配置

• default:
  • 数据文件路径: data/train-*