sarulab-speech/mls_sidon
收藏Hugging Face2025-10-08 更新2025-10-18 收录
下载链接：
https://hf-mirror.com/datasets/sarulab-speech/mls_sidon
下载链接
链接失效反馈官方服务：
资源简介：
MLS-Sidon数据集是一个针对语音合成和口语语言建模的Multilingual LibriSpeech的清洗版。它包含了多种语言的语音数据，如英语、德语、法语、西班牙语、意大利语、波兰语、荷兰语和葡萄牙语。数据集以WebDataset格式存储，每个样本包括48 kHz单声道FLAC音频文件和可选的包含语言、说话者ID和原始MLS引用的JSON格式元数据文件。
MLS-Sidon dataset is a cleansed version of Multilingual LibriSpeech designed for speech synthesis and spoken language modeling. It contains speech data in multiple languages including English, German, French, Spanish, Italian, Polish, Dutch, and Portuguese. The dataset is stored in WebDataset format, with each sample consisting of a 48 kHz single-channel FLAC audio file and an optional JSON metadata file that includes language, speaker ID, and original MLS reference.
提供机构：
sarulab-speech
