MLS English subset
收藏arXiv2025-09-30 收录
下载链接:
https://theodorblackbird.github.io/blog/demo_lina/
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是一个公开可用的英文子集,包含了1万小时的Librivox录音。值得注意的是,该数据集并未提供转录文本,而是采用自动语音识别模型NeMo来进行转录工作。该数据集的任务是文本到语音的合成。
This dataset is a publicly available English subset comprising 10,000 hours of Librivox recordings. Notably, no pre-existing transcriptions are provided for this dataset; instead, the automatic speech recognition model NeMo is utilized to perform the transcription tasks. The core task associated with this dataset is text-to-speech synthesis.
提供机构:
Parler-TTS



