Datasets | Mozilla Data Collective

10.5 hours (8440 utterances) of read speech of Nederlands (Dutch).

VoxForge - Dutch

License: GPL-3.0

Locale: nl

Task: ASR

Format: WAV, TSV

Size: 2.25 GB

20 hours (10643 utterances) of read speech of Italiano (Italian).

VoxForge - Italian

License: GPL-3.0

Locale: it

Task: ASR

Format: WAV, TSV

Size: 4.94 GB

53 hours (23537 utterances) of read speech of Español (Spanish).

VoxForge - Spanish

License: GPL-3.0

Locale: es

Task: ASR

Format: WAV, TSV

Size: 12.70 GB

Institute of African Digital Humanities

Efik-TTS-Dataset

Segmented Efik speech clips paired with text for TTS. Standard Calabar dialect, Cross River State, Nigeria. 1,362 clips, 5h 32m.

License: NOODL-1.0

Locale: efi

Task: TTS

Format: MP3, TSV

Size: 297.28 MB

Institute of African Digital Humanities

Igbo-TTS-Dataset

Segmented Igbo speech audio clips paired with text for Text-to-Speech (TTS), representing the standard spoken variety of southeastern Nigeria.

License: NOODL-1.0

Locale: ibo

Task: TTS

Format: MP3, TSV

Size: 172.77 MB

Institute of African Digital Humanities

Adamawa Fulfulde-TTS-Dataset

1,303 read-speech recordings from a single Adamawa Fulfulde speaker, with MP3 audio and a TSV sentence mapping. Suitable for TTS and ASR.

License: NOODL-1.0

Locale: fub

Task: TTS

Format: MP3, TSV

Size: 169.27 MB

The dataset includes audio files in Catalan and their translations in English.

CoVoST 2 Catalan - English

License: CC-BY-NC-4.0

Locale: ca, en

Task: MT

Format: MP3, TSV

Size: 4.50 GB

The dataset includes audio files in Welsh and their translations in English.

CoVoST 2 Welsh - English

License: CC-BY-NC-4.0

Locale: cy, en

Task: MT

Format: MP3, TSV

Size: 95.16 MB

The dataset includes 5736 audio files in Arabic and their translations in English.

CoVoST 2 Arabic-English

License: CC-BY-NC-4.0

Locale: ar, en

Task: MT

Format: MP3, TSV

Size: 148.21 MB

32 hours (24685 utterances) of read speech of Deutsch (German).

VoxForge - German

License: GPL-3.0

Locale: de

Task: ASR

Format: WAV, TSV

Size: 11.59 GB

4 hours (1377 utterances) of read speech of Ελληνικά (Greek).

VoxForge - Greek

License: GPL-3.0

Locale: el

Task: ASR

Format: WAV, TSV

Size: 954.26 MB

39 minutes of Catalan read speech.

VoxForge - Catalan

License: GPL-3.0

Locale: ca

Task: ASR

Format: WAV, TSV

Size: 143.71 MB

0DIN by Mozilla

Public GenAI Vulnerability Disclosures

Weekly export of 0DIN's public GenAI vulnerability disclosures: jailbreaks, prompt injection, content manipulation, and other model-security findings.

License: CC-BY-4.0

Locale: en

Task: N/A

Format: JSONL

Size: 116.26 KB

1 hour of read speech in Bulgarian, collected via the VoxForge project

VoxForge - Bulgarian

License: GPL-3.0

Locale: bg

Task: ASR

Format: WAV, TSV

Size: 234.06 MB

The dataset includes 154856 audio files in German and their translations in English.

CoVoST 2 German-English

License: CC-BY-NC-4.0

Locale: de, en

Task: MT

Format: MP3, TSV

Size: 5.90 GB

25,305 French–Congo Swahili parallel sentences from CLEAR Global's Gamayun kits, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Congo Swahili (25k)

License: CC-BY-4.0

Locale: swc

Task: MT

Format: TSV

Size: 2.18 MB

5,000 English–Swahili parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Swahili (5k)

License: CC-BY-4.0

Locale: swa

Task: MT

Format: TSV

Size: 347.61 KB

5,000 English–Rohingya parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Rohingya (5k)

License: CC-BY-4.0

Locale: rhg

Task: MT

Format: TSV

Size: 358.88 KB

5,000 French–Lingala parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Lingala (5k)

License: CC-BY-4.0

Locale: lin

Task: MT

Format: TSV

Size: 494.43 KB

15,000 French–Nande parallel sentences from CLEAR Global's Gamayun kits, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Nande (15k)

License: CC-BY-4.0

Locale: nnb

Task: MT

Format: TSV

Size: 1.26 MB

5,000 English–Tigrinya parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Tigrinya (5k)

License: CC-BY-4.0

Locale: tig

Task: MT

Format: TSV

Size: 404.75 KB

5,000 English–Kanuri parallel sentences from CLEAR Global's Gamayun mini-kit, translated by professional translators from Tatoeba source sentences.

TWB Parallel Sentence kits - Kanuri (5k)

License: CC-BY-4.0

Locale: kau

Task: MT

Format: TSV

Size: 358.46 KB

This is a public domain speech dataset consisting of 13,100 short audio clips of a single speaker reading passages from 7 non-fiction books.

The LJSpeech Dataset

License: CC0-1.0

Locale: en

Task: TTS

Format: WAV, TSV

Size: 3.00 GB