ae-mask-distilled-libri-one-sec-cv12/chunk_230

该数据集包含音频数据和对应的浮点数序列。音频数据的采样率为16000Hz，浮点数序列用于存储与音频相关的特征或预测结果。数据集包含3150个训练样本，总大小约为100MB，下载大小约为90MB。

This dataset contains audio data and corresponding float sequences. The audio data has a sampling rate of 16000Hz, and the float sequences are used to store features or predictions related to the audio. The dataset includes 3150 training samples, with a total size of approximately 100MB and a download size of approximately 90MB.

原始信息汇总

数据集概述

特征

• 音频 (audio)
  • 采样率: 16000
• 对数概率 (logits)
  • 数据类型: float32

数据分割

• 训练集 (train)
  • 样本数量: 3150
  • 数据大小: 100986243.87939534 字节

数据集大小

• 下载大小: 90691907 字节
• 数据集总大小: 100986243.87939534 字节

配置

• 默认配置 (default)
  • 数据文件路径: data/train-*