mesolitica/Malaysian-Emilia

Malaysian Emilia 是一个广泛、多语言且多样化的语音数据集，用于大规模马来西亚语音生成。数据集包含来自马来西亚卡通、马来西亚YouTube视频、马来西亚播客、新加坡播客、马来西亚议会和马来西亚方言等多个来源的音频数据。每个来源的音频数据都经过了处理，包括时长调整和样本率清理。数据集的总规模在1M到10M之间，语言包括英语和马来语。

Malaysian Emilia is an extensive, multilingual, and diverse speech dataset designed for large-scale Malaysian speech generation. The dataset includes audio data from various sources such as Malaysian cartoons, Malaysian YouTube videos, Malaysian podcasts, Singaporean podcasts, Malaysian parliament, and Malaysian dialects. Each sources audio data has been processed, including duration adjustment and sample rate cleaning. The datasets total size ranges between 1M and 10M, and it includes languages such as English and Malay.