Datasets | Mozilla Data Collective

All Datasets

Compensated Datasets

API

Contribute Datasets

Pricing

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 “Hazargi Multimodal Dataset”

100 Datasets

Taraaz

Multilingual Humanitarian Response Eval (MHRE)

This multilingual humanitarian dataset contains 655 annotated datapoints evaluating AI chatbot safety and quality in migration and asylum scenarios across four language pairs (English–Farsi (Iranian Persian), Arabic, Kurdish (Sorani), Pashto). Built from 120 expert prompts, it includes outputs from GPT-4o, Gemini 2.5 Flash, and Mistral Small. The dataset provides both human evaluations from Respond Crisis Translation native-speaker evaluators and LLM-as-judge assessments (Gemini 2.5 Flash).

License: CC-BY-NC-SA-4.0

Locale: mul

Task: LLM

Format: csv

Size: 2.15 MB

Rerooted Archive

Rerooted: Speech Corpus of Testimonials from Armenian Refugees and Immigrants

Rerooted is an open-access online YouTube corpus of interviews with Armenian refugees and immigrants. As of now, the online corpus has over 80hrs of recordings on YouTube, alongside subtitles in Western Armenian. The current dataset is 20 hours from the corpus that we have curated with corrected and finely-annotated transcripts. The dataset is for use in NLP research, and we hope to continuously update the dataset with more curated transcripts.

License: GPL-3.0

Locale: hy; hyw

Task: ASR

Format: WAV, TEXTGRID

Size: 6.21 GB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Uzbek

A collection of read speech recordings in Uzbek.

License: CC0-1.0

Locale: uz

Task: ASR

Format: MP3

Size: 5.03 GB

MDC DatasetsCompensated

🚧 Tagalog Voice Notes ASR Dataset (45 hours) 🚧

A Tagalog speech recognition dataset consisting of 45+ hours of transcribed conversations recorded via voice memos on WhatsApp and Telegram.

License: MDC-DLA-1.0

Locale: tl

Task: ASR

Format: MP3, TSV

Size: 1.24 GB

MDC DatasetsCompensated

Single Speaker Urdu TTS dataset (Female Speaker, South Punjab)

A high-quality single-speaker Urdu text-to-speech (TTS) dataset with 25 hours of paired speech recordings and transcriptions for speech synthesis.

License: MDC-DLA-1.0

Locale: ur

Task: TTS

Format: WAV, TSV

Size: 6.02 GB

Mozilla Common Voice

Common Voice Spontaneous Speech 5.0 - Bashkir

A collection of spontaneous responses to questions in Bashkir.

License: CC0-1.0

Locale: ba

Task: ASR

Format: MP3

Size: 5.10 MB

Weekly dataset digest

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

Yes, notify me

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Kazakh

A collection of read speech recordings in Kazakh.

License: CC0-1.0

Locale: kk

Task: ASR

Format: MP3

Size: 79.09 MB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Kabardian

A collection of read speech recordings in Kabardian.

License: CC0-1.0

Locale: kbd

Task: ASR

Format: MP3

Size: 5.79 GB

Institute of Finno-Ugric/Uralic Studies, University of Hamburg

INEL Kalmyk Speech Corpus

A 3-hour supervised Speech-to-Text dataset for Kalmyk, a Mongolic language. Features sentence-level audio aligned with scientific text transcriptions.

License: CC-BY-NC-SA-4.0

Locale: xal

Task: ASR

Format: TSV, MP3

Size: 138.31 MB

PROXIMA

Urdu OCR Multi Source Dataset (Proxima AI)

Urdu OCR dataset of 40,769 image text pairs compiled from scanned, printed, synthetic, and scene text sources.

License: CC-BY-NC-SA-4.0

Locale: urd

Task: CV

Format: PARQUET

Size: 436.48 MB

Mozilla Common Voice

Common Voice Spontaneous Speech 5.0 - Kabardian

A collection of spontaneous responses to questions in Kabardian.

License: CC0-1.0

Locale: kbd

Task: ASR

Format: MP3

Size: 172.56 MB

Mozilla Common Voice

Common Voice Spontaneous Speech 5.0 - Gorani

A collection of spontaneous responses to questions in Gorani.

License: CC0-1.0

Locale: hac

Task: ASR

Format: MP3

Size: 225.34 MB

Pashto AI

Pashto Multi-Speaker Informational Speech Corpus

Pashto multi-speaker corpus with 3,880 clips (8.914 hours) from five male presenters across Afghanistan and Pakhtunkhwa for TTS and ASR.

License: CC0-1.0

Locale: ps

Task: ASR

Format: FLAC, TSV, JSONL, JSON, MD

Size: 782.08 MB

Institute of Finno-Ugric/Uralic Studies, University of Hamburg

INEL Evenki Speech Corpus

A 2.5-hour supervised Speech-to-Text dataset for Evenki, an endangered Tungusic language. Features sentence-level audio aligned with text transcriptions.

License: CC-BY-NC-SA-4.0

Locale: evn

Task: ASR

Format: TSV, MP3

Size: 103.03 MB

Community

IND Bintaro-Jakarta: Urban and Mobility Multimodal Dataset

Covering public facilities, roadways, and transit signages across Bintaro, South Tangerang, and Jakarta, Indonesia.

License: CC-BY-NC-SA-4.0

Locale: ind

Task: CV

Format: JPEG, WEBM, JSON

Size: 1.62 GB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Bashkir

A collection of read speech recordings in Bashkir.

License: CC0-1.0

Locale: ba

Task: ASR

Format: MP3

Size: 5.10 GB

Mozilla Common Voice

Common Voice Spontaneous Speech 5.0 - Tatar

A collection of spontaneous responses to questions in Tatar.

License: CC0-1.0

Locale: tt

Task: ASR

Format: MP3

Size: 4.58 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

Community

Mandar Spontaneous Speech

This dataset is a compilation of spontaneous Mandar speech featuring Indonesian code-switching.

License: CC-BY-NC-4.0

Locale: mdr

Task: ASR

Format: MP3, TSV

Size: 534.45 MB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Central Kurdish

A collection of read speech recordings in Central Kurdish.

License: CC0-1.0

Locale: ckb

Task: ASR

Format: MP3

Size: 3.65 GB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Tatar

A collection of read speech recordings in Tatar.

License: CC0-1.0

Locale: tt

Task: ASR

Format: MP3

Size: 836.51 MB

Institute of African Digital Humanities

Badjia-ALCAM-MultimodalDataset

Multimodal dataset for Badja (Badjia), an Ewondo/Beti dialect of Cameroon: lexical datasheet, native-speaker audio, and audio-sentence mapping files.

License: NOODL-1.0

Locale: ewo

Task: NLP

Format: WAV, TSV

Size: 42.77 MB

LocaleNLP

English Hausa Parallel Corpus

An English–Hausa dataset with 5,000 sentence pairs useful for machine translation and basic language processing tasks.

License: CC-BY-NC-4.0

Locale: eng, hau

Task: MT

Format: csv

Size: 164.32 KB

Mozilla Common Voice

Common Voice Scripted Speech 27.0 - Bunun

A collection of read speech recordings in Bunun.

License: CC0-1.0

Locale: bnn

Task: ASR

Format: MP3

Size: 235.71 MB