neh7777/Pretraining-V1

这是一个多语言文本到语音（TTS）数据集，包含多个子集（如cv22_sea、cv22_african、cv22_de等），覆盖多种语言和地区，包括印度语言（如印地语、孟加拉语）、欧洲语言（如德语、法语、西班牙语）、阿拉伯语等。每个子集提供音频文件（采样率为24000Hz）及其对应的文本转录，并包含说话人ID、来源、语言、性别和时长等元数据。数据集适用于文本到语音和自动语音识别任务，支持多语言语音处理研究。This is a multilingual text-to-speech (TTS) dataset comprising multiple subsets (e.g., cv22_sea, cv22_african, cv22_de, etc.), covering diverse languages and regions, including Indian languages such as Hindi and Bengali, European languages such as German, French and Spanish, Arabic and others. Each subset provides audio files with a sampling rate of 24000 Hz and their corresponding text transcriptions, along with metadata including speaker ID, source, language, gender and duration. This dataset is applicable to text-to-speech and automatic speech recognition tasks, supporting multilingual speech processing research.