LibriTTS multi-speaker English text-to-speech corpus
- Provider
- Heiga Zen, Viet Dang, Rob Clark, Yu Zhang, Ron J. Weiss, Ye Jia, Zhifeng Chen, Yonghui Wu (Google)
- License
- CC-BY-4.0 (source)
- Size
- 228 files
- Format
- Parquet (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/mythicinfinity/libritts)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Speech & Audio
- Source
- https://arxiv.org/abs/1904.02882
About LibriTTS multi-speaker English text-to-speech corpus
LibriTTS is a multi-speaker English corpus of approximately 585 hours of read speech at 24 kHz, prepared by Heiga Zen with Google Speech and Google Brain for text-to-speech (TTS) research. It is derived from the original LibriVox mp3 audio and Project Gutenberg text behind LibriSpeech, but differs in ways that matter for speech synthesis: audio is kept at 24 kHz, speech is split at sentence breaks, both original and normalised text are included, neighbouring-sentence context can be recovered, and utterances with significant background noise are removed. It is a standard training set for multi-speaker and zero-shot TTS, voice cloning and neural vocoders. This repository is a Parquet conversion of the OpenSLR SLR60 release: a data folder with seven split folders (dev.clean, dev.other, test.clean, test.other, train.clean.100, train.clean.360, train.other.500) holding Parquet shards with audio, original and normalised text, speaker id and chapter id (measured 2026-11-14; the Hub reports about 172 GB). Load with datasets.load_dataset('mythicinfinity/libritts', 'clean') or pyarrow. The corpus is licensed CC BY 4.0 by its publishers on OpenSLR; cite Zen et al., Interspeech 2019.
How to load LibriTTS multi-speaker English text-to-speech corpus in Python
from datasets import load_dataset
ds = load_dataset("mythicinfinity/libritts")
print(ds)Data structure
File types in sampled files: Parquet
Files
Example file: data/dev.clean/dev.clean-00000-of-00003.parquet
Analyze this dataset without downloading it
On Scigantic, LibriTTS multi-speaker English text-to-speech corpus opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Speech & Audio datasets
- LibriSpeech-Long: Long-Form Speech Benchmark of Contiguous Audiobook Segments up to 4 Minutes (Google DeepMind)
- CML-TTS multilingual speech synthesis dataset
- Omnilingual Asr Corpus
- VoiceBank+DEMAND noisy speech database (16 kHz)
- Bach Doodle Dataset: 21.6 Million User Melodies with Coconet Four-Voice Bach-Style Harmonizations (Google Magenta)
- CVSS: Multilingual Speech-to-Speech Translation Corpus, 21 Languages into English (CVSS-C and CVSS-T, Google Research)
- Multilingual Librispeech
- E-GMD: Expanded Groove MIDI Dataset, 444 Hours of Human Drumming on 43 Drum Kits with Aligned MIDI and Audio (Google Magenta)