formospeech/formosan_clean_denoised_train

该数据集包含多个配置，每个配置都包含音频、文本、国际音标（IPA）、语言、方言、来源、类型等信息。每个配置的数据集大小、下载大小、样本数量等详细信息也被列出。数据集的主要特征是音频和文本的对应关系，并且每个样本都有详细的元数据信息。

This dataset contains audio and text data for multiple languages and dialects. Each sample includes an audio file, duration, text, International Phonetic Alphabet (IPA), language, dialect, source, kind, and other related information. The dataset is primarily intended for speech-related research and applications, such as speech recognition and speech synthesis.