Datasets | Mozilla Data Collective

All Datasets

Compensated Datasets

API

Contribute Datasets

Loading datasets...

Terms

FAQs

Participation Guidelines

Responsible Disclosure

News and Documentation

Data and Model Competitions

[email protected]

Terms

FAQs

Participation Guidelines

Responsible Disclosure

News and Documentation

Data and Model Competitions

[email protected]

Mozilla Data Collective is a mission-locked British company, incubated by Mozilla Foundation, and backed by the not-for-profit Mozilla.org

All Datasets

Filtered by

Search results for “spontaneous”

113 Datasets

Institute of African Digital Humanities

Basaa-ASR-Dataset

ASR/TTS dataset for Basaa (bas): 2,498 audio clips (01:54:46) in AGLC orthography, 25 sessions, including Basaa-ba-Yabassi and Babimbi varieties.

License: NOODL-1.0

Locale: bas

Task: ASR

Format: TSV, MP3

Size: 76.89 MB

PROXIMA

Syntheic Urdu Audios

Synthetic Urdu conversational speech dataset of 7,963 audio text pairs featuring casual, humorous dialogue exchanges.

License: CC-BY-NC-SA-4.0

Locale: urd

Task: TTS

Format: PARQUET

Size: 4.58 GB

Chana

Kakataibo: Fieldwork 2006-2007

54 Kakataibo fieldwork sessions recorded in 2006–2007, documenting eight speakers in elicitation, narrative, and conversation, with Transcriber annotations.

License: Chana 2.0

Locale: cbr

Task: N/A

Format: WAV, TRS, XLS, PDF

Size: 1.93 GB

Content X LabsCompensated

🚧 English Scripted Narrative // Rights-Cleared · 20 Hours 🚧

20 hours of premium, rights-cleared English-language scripted film content, with English subtitles and accompanying metadata, available for AI training

License: CXL-DLA-1.0

Locale: EN

Task: N/A

Format: MP4, SRT, CSV

Size: 102.15 GB

MDC DatasetsCompensated

Conversational Speech in Manipuri in the Medical Domain

25 hours of Manipuri Language Conversational Speech in the Medical Domain

License: MDC DLA-1.0

Locale: mni

Task: ASR

Format: WAV, JSON

Size: 6.72 GB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Svan

A collection of read speech recordings in Svan.

License: CC0-1.0

Locale: sva

Task: ASR

Format: MP3

Size: 353.39 MB

Weekly dataset digest

Receive a brief weekly digest of new datasets similar to the ones you’ve downloaded.

Yes, notify me

Community

ASR Javanese-Lumajang Dialect

This ASR dataset comprises 10 hours of natural speech in the Javanese-Lumajang dialect in East Java, Indonesia and contains code-switching into Indonesian.

License: CC-BY-NC-SA-4.0

Locale: jav

Task: ASR

Format: OPUS, TSV

Size: 81.38 MB

Prompsit Language Engineering

Prompsit D1 - Inline Asset Integrity

Checks whether machine translation keeps tags, placeholders and protected text intact.

License: CC-BY-4.0

Locale: mul

Task: MT

Format: JSONL

Size: 215.42 KB

Mozilla Common Voice

Mozilla Common Voice Text Language Identification dataset

A dataset for text-based language identification of 19 Million sentences from over 300 languages taken from Mozilla Common Voice scripted (v23) and spontaneous (v1) speech projects.

License: CC0-1.0

Locale: mul

Task: NLP

Format: TSV

Size: 950.41 MB

Content X LabsCompensated

🚧 Spanish (Castilian) Scripted Film // Rights-Cleared · 20 Hours 🚧

20 hours of rights-cleared Castilian Spanish-language scripted film content, accompanied by structured metadata and English Subtitles

License: CXL-DLA-1.0

Locale: ES

Task: N/A

Format: MP4, SRT, CSV

Size: 108.48 GB

MDC DatasetsCompensated

Conversational Speech in Gujarati in the Finance Domain

25 hours of Conversational Speech in the Gujarati Language in the Finance Domain

License: MDC DLA-1.0

Locale: gu

Task: ASR

Format: WAV, JSON

Size: 6.20 GB

Chana

Kakataibo: Traditional Stories

Six traditional Kakataibo stories recorded between 2007 and 2012, featuring forest animals and tsikiu mano, with transcriptions and Spanish translations.

License: Chana 2.0

Locale: cbr

Task: N/A

Format: PDF, WAV, MPEG, MP3, MD, CSV

Size: 508.47 MB

MDC DatasetsCompensated

Conversational Speech in Manipuri in the Finance Domain

25 hours of Manipuri Language Conversational Speech in the Finance Domain

License: MDC DLA-1.0

Locale: mni

Task: ASR

Format: WAV, JSON

Size: 6.50 GB

PT Pancaran Semangat Jaya

Corpus of Panjebar Semangat Javanese-Language Magazine

This dataset is a TXT-format collection compiled from three years of popular articles published in the Javanese-language weekly magazine Panjebar Semangat. It compiles widely read, non-academic Javanese texts reflecting contemporary themes and language use.

License: CC-BY-SA-4.0

Locale: Jav

Task: OTH

Format: TXT

Size: 4.31 MB

MDC DatasetsCompensated

Conversational Speech in Gujarati in the Medical Domain

25 hours of Conversational Speech in the Gujarati Language in the Medical Domain

License: MDC-DLA 1.0

Locale: gu

Task: ASR

Format: WAV, JSON

Size: 6.18 GB

Open Home Foundation

Cadu 1.0

Text to speech dataset for Brazilian Portuguese, male speaker, approximately 1.5 hours of read speech.

License: CC0-1.0

Locale: pt-BR

Task: TTS

Format: WEBM

Size: 30.98 MB

Center za jezikovne vire in tehnologije Univerze v Ljubljani

SI-NLI

SI-NLI (Slovene Natural Language Inference Dataset) contains 5,937 human-created Slovene sentence pairs (premise and hypothesis) that are manually labeled with the labels "entailment", "contradiction", and "neutral". We created the dataset using sentences that appear in the Slovenian reference corpus ccKres (http://hdl.handle.net/11356/1034). Annotators were tasked to modify the hypothesis in a candidate pair in a way that reflects one of the labels. The dataset is balanced since the annotators created three modifications (entailment, contradiction, neutral) for each candidate sentence pair. The dataset is split into train, validation, and test sets, with sizes of 4,392, 547, and 998. We used Slovenian pre-trained language models to create splits, thereby ensuring that difficult and easy instances are evenly distributed in all three subsets. The dataset is released in a tabular TSV format. The README.txt file contains a description of the attributes. Only the hypothesis and premise are given in the test set (i.e. no annotations) since SI-NLI is integrated into the Slovene evaluation framework SloBENCH (https://slobench.cjvt.si/). If you use the dataset to train your models, please consider submitting the test set predictions to SloBENCH to get the evaluation score and see how it compares to others.

License: CC-BY-NC-SA-4.0

Locale: sl

Task: NLU

Format: TSV

Size: 392.44 KB