LibriSpeech-Long: Long-Form Speech Benchmark of Contiguous Audiobook Segments up to 4 Minutes (Google DeepMind)
- Provider
- Google DeepMind (Se Jin Park, Julian Salazar, Aren Jansen et al.)
- License
- CC-BY-4.0 (source)
- Stored at
- Google Cloud Storage, gs://librispeech_long (us)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Speech & Audio
- Source
- https://arxiv.org/abs/2412.18603
About this dataset
LibriSpeech-Long is a benchmark for long-form speech processing: speech continuation, long-form speech recognition (ASR) and text-to-speech synthesis over minutes rather than seconds. It is derived from the LibriSpeech dev and test sets (read English audiobooks from LibriVox), whose utterances were reprocessed into contiguous examples of up to 4 minutes, in the manner of the Libri-Light cut_by_vad.py script. The bucket holds a single archive, v0_1.tar.gz (about 3.2 GB), with audio, ground-truth transcripts and per-file durations for all splits (dev-clean, dev-other, test-clean, test-other). The release is marked work in progress and may change. Used to evaluate spoken language models such as SpeechSSM on long generation and transcription. License: CC BY 4.0 for the materials; the underlying LibriSpeech data are also CC BY 4.0 (Panayotov, Chen, Povey and Khudanpur, ICASSP 2015). Cite Park, Salazar, Jansen et al., Long-Form Speech Generation with Spoken Language Models, arXiv:2412.18603.
How to access this dataset in Python
import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('librispeech_long/', detail=False)[:10]Lists the files with anonymous access. Runs on any machine, with no cloud account.
Data structure
Files
Example file: v0_1.tar.gz
Analyze this dataset without downloading it
On Scigantic, LibriSpeech-Long: Long-Form Speech Benchmark of Contiguous Audiobook Segments up to 4 Minutes (Google DeepMind) 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
- 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)
- CocoChorales: 240,000 Synthesized Four-Part Chorale Mixtures with Per-Instrument Stems, MIDI and Note Expression (Google Magenta)