paralinguistic-tags-bet-v2

该数据集包含多语言音频及其转录文本，适用于语音识别和音频事件检测等任务。数据集包含200个训练样本，每个样本包含音频文件（采样率16kHz）、语言标签、原始转录文本、带标签的转录文本、事件描述、事件数量、音频时长、处理时间和来源数据集信息。数据集总大小为63,740,643字节，下载大小为58,929,128字节。This dataset contains multilingual audio and their transcriptions, suitable for tasks such as speech recognition and audio event detection. The dataset consists of 200 training samples, each including an audio file (sampling rate 16kHz), language label, original transcription, labeled transcription, event description, number of events, audio duration, processing time, and source dataset information. The total size of the dataset is 63,740,643 bytes, and the download size is 58,929,128 bytes.