Datasets | Mozilla Data Collective

Ewondo (ewo) scripted speech dataset: 1,781 clips (≈3h11min) from 16 speakers, with AGLC-compliant orthography including full tone marking.

Ewondo-ASR-Dataset

License: NOODL-1.0

Locale: ewo

Task: ASR

Format: MP3, TSV

Size: 117.51 MB

A collection of read speech recordings in Dholuo.

Common Voice Scripted Speech 27.0 - Dholuo

License: CC0-1.0

Locale: luo

Task: ASR

Format: MP3

Size: 2.08 GB

Eton (etn) scripted speech dataset: 1,402 clips (≈3h35min) from 11 speakers, with AGLC-compliant orthography including full tone marking.

Eton-ASR-Dataset

License: NOODL-1.0

Locale: eto

Task: ASR

Format: MP3, TSV

Size: 126.81 MB

A collection of spontaneous responses to questions in Kuku.

Common Voice Spontaneous Speech 5.0 - Kuku

License: CC0-1.0

Locale: ukv

Task: ASR

Format: MP3

Size: 238.25 MB

Tupuri-Bango (tui) scripted TTS dataset: 2,034 clips (3h 53m 42s), single female speaker, AGLC orthography. Female-voice complement to the male-voice dataset.

Tupuri-Bango_TTS-Dataset (female voice)

License: NOODL-1.0

Locale: tui

Task: TTS

Format: MP3, TSV

Size: 169.22 MB

A collection of spontaneous responses to questions in Ruuli.

Common Voice Spontaneous Speech 5.0 - Ruuli

License: CC0-1.0

Locale: ruc

Task: ASR

Format: MP3

Size: 365.90 MB

Weekly dataset digest

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

Yes, notify me

MirasAI

Telugu Text Corpus

A Telugu text corpus for NLP research, containing diverse written content for language modeling, text analysis, and linguistic studies.

License: CC-BY-NC-SA-4.0

Locale: tel

Task: NLP

Format: TXT

Size: 19.87 MB

A collection of spontaneous responses to questions in Kenyi.

Common Voice Spontaneous Speech 5.0 - Kenyi

License: CC0-1.0

Locale: lke

Task: ASR

Format: MP3

Size: 254.44 MB

Multimodal dataset for Batanga : lexical datasheet, audio recordings, and an audio alignment file for Banoho. Bapuku, forthcoming.

Batanga-ALCAM-MultimodalDataset

License: NOODL-1.0

Locale: bnm

Task: NLP

Format: MP3, TSV

Size: 18.70 MB

A collection of read speech recordings in Kunabembe.

Common Voice Scripted Speech 27.0 - Kunabembe

License: CC0-1.0

Locale: mgg

Task: ASR

Format: MP3

Size: 220.51 MB

iSocial

RAPNIC - Catalan Impaired Speech (Example Set)

An example set of RAPNIC, the first Catalan speech corpus of people with speech disorders such as cerebral palsy and Down syndrome.

License: CC-BY-NC-SA-4.0

Locale: ca

Task: ASR

Format: WAV, TSV

Size: 105.22 MB

A collection of read speech recordings in Pa'O.

Common Voice Scripted Speech 27.0 - Pa'O

License: CC0-1.0

Locale: blk

Task: ASR

Format: MP3

Size: 463.99 KB

Biomedical Urdu Named Entity Recognition dataset for advancing healthcare and clinical NLP research in Urdu.

Biomedical Urdu Named Entity Recognition (BioUNER)

License: CC-BY-NC-4.0

Locale: urd

Task: NLP

Format: CSV

Size: 336.69 KB

A collection of spontaneous responses to questions in Rakhine.

Common Voice Spontaneous Speech 5.0 - Rakhine

License: CC0-1.0

Locale: rki

Task: ASR

Format: MP3

Size: 11.83 MB

Text corpus of 176,843 words compiled from 397 Judeo-Espanyol articles from Şalom newspaper.

Şalom Ladino Corpus

License: CC-BY-4.0

Locale: lad

Task: LM

Format: TXT

Size: 403.16 KB

A collection of read speech recordings in Tush.

Common Voice Scripted Speech 27.0 - Tush

License: CC0-1.0

Locale: bbl

Task: ASR

Format: MP3

Size: 247.84 MB

A large-scale Sindhi Part-of-Speech (SiPOS) tagging dataset for developing and evaluating NLP models in a low-resource language.

Sindhi Parts Of Speech (SiPOS)

License: CC-BY-NC-4.0

Locale: snd

Task: NLP

Format: TXT

Size: 874.04 KB

A collection of read speech recordings in Erzya.

Common Voice Scripted Speech 27.0 - Erzya

License: CC0-1.0

Locale: myv

Task: ASR

Format: MP3

Size: 65.17 MB

A benchmark Sindhi question answering dataset for advancing machine reading comprehension and NLP research in a low-resource language.

Sindhi Question-Answering Dataset (SdQuAD)

License: CC-BY-NC-4.0

Locale: snd

Task: LLM

Format: JSON

Size: 500.50 KB

A collection of spontaneous responses to questions in Thai.

Common Voice Spontaneous Speech 5.0 - Thai

License: CC0-1.0

Locale: th

Task: ASR

Format: MP3

Size: 1.58 MB

MirasAI

Kannada Text Corpus

A curated Kannada text corpus containing diverse written content for natural language processing, linguistic research, and language technology development.

License: CC-BY-NC-SA-4.0

Locale: kan

Task: NLP

Format: TXT

Size: 2.61 MB

A collection of read speech recordings in Breton.

Common Voice Scripted Speech 27.0 - Breton

License: CC0-1.0

Locale: br

Task: ASR

Format: MP3

Size: 865.66 MB

PROXIMA

Helpful and Harmless RLHF Urdu

Urdu translation of Anthropic's HH-RLHF preference dataset for RLHF/reward-model research in Urdu.

License: CC-BY-NC-SA-4.0

Locale: urd

Task: NLP

Format: Parquet

Size: 10.37 MB