Datasets | Mozilla Data Collective

Text to speech dataset for Dutch, female speaker, approximately 1 hour of read speech.

Nathalie 1.0

License: CC0-1.0

Locale: nl-BE

Task: TTS

Format: WEBM

Size: 21.87 MB

MDC Curators

LibriVox Czech TTS Female Voice

2 hours of sentence-aligned speech/text from "Krysař" by Viktor Dyk

License: CC0-1.0

Locale: cs

Task: TTS

Format: MP3, TXT, TSV

Size: 178.58 MB

PROXIMA

Synthetic Urdu 2

Synthetic Urdu sentences audio dataset of 85,327 verse style audio-text pairs.

License: CC-BY-NC-SA-4.0

Locale: urd

Task: TTS

Format: Parquet

Size: 14.80 GB

Text to speech dataset for Dutch, male speaker, approximately 1 hour of read speech.

Flemishguy 1.0

License: CC0-1.0

Locale: nl-BE

Task: TTS

Format: FLAC

Size: 73.69 MB

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

Conversational Speech in Gujarati in the Medical Domain

License: MDC-DLA 1.0

Locale: gu

Task: ASR

Format: WAV, JSON

Size: 6.18 GB

This dataset comprises audio and lip movement recordings in Indonesian language.

Indo-Labial Multimodal Corpus

License: CC-BY-NC-SA-4.0

Locale: ind

Task: OTH

Format: WEBM, TSV

Size: 820.04 MB

Digital Divide Data

ddd-kenya-somali-68hrs-asr-part2

A collection of high-quality, crowdsourced audio recordings and verified transcriptions for the Somali language, produced by Digital Divide Data.

License: CC-BY-4.0

Locale: som

Task: ASR

Format: WAV, TSV

Size: 8.07 GB

MirasAI

Marathi Blog & Literature Corpus

A curated Marathi text corpus containing diverse written content for natural language processing, linguistic research, and language modeling tasks.

License: CC-BY-NC-4.0

Locale: mar

Task: NLP

Format: TXT

Size: 4.20 MB

The dataset contains audio and text resources on Laari, a Bantu language spoken in the Congo. The resources, which are suitable for TTS tasks and possibly ASR tasks, consist of the following: - 6,311 audio clips totalling 241 minutes and 44.97 seconds; - an audio mapping file with 5,321 lines, each beginning with the name of an audio file, followed by a tab and then the corresponding text excerpt; - two raw audio files totalling 120 minutes and 54.90 seconds; - two long audio files with their original, non-split transcription files, for a total duration of 120 minutes and 41.90 seconds.

Laari-TTS-Dataset

License: NOODL-1.0

Locale: ldi

Task: ASR

Format: WAV, TRJS, TSV

Size: 568.26 MB

This dataset contains paired audio and text resources for three Bati dialects (Kelleng, Mbougue, and Nyambat), which belong to the Yambasa group of Bantu languages found in Cameroon. It contains 13,344 audio clips totalling 6 hours, 8 minutes and 12.286 seconds and 44 audio/text mapping files totalling 13,346 lines. Due to its cross-dialectal nature, the dataset is suitable for multilingual automatic speech recognition tasks.

Bati-MultiDialectalASR-Dataset

License: NOODL-1.0

Locale: btc

Task: ASR

Format: WAV, TSV

Size: 3.27 GB

2 hours of studio-quality, deliberate, read speech from a single speaker of Southeast Asian Hokkien.

🚧 Southeast Asian Hokkien TTS dataset 🚧

License: MDC DLA-1.0

Locale: nan

Task: TTS

Format: WAV, TSV

Size: 194 B

Batanga (bnm) scripted TTS dataset: 1,023 clips (≈1h27m) from 11 sessions, transcribed in AGLC orthography. Ocean Division, South Region, Cameroon.

Sample Batanga-TTS-Dataset

License: NOODL-1.0

Locale: bnm

Task: TTS

Format: MP3, TSV

Size: 79.13 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

Text to speech dataset for Bulgarian, male speaker, approximately 2 hours of read speech.

Dimitar 1.0

License: CC0-1.0

Locale: bg-BG

Task: TTS

Format: WEBM

Size: 109.58 MB

Afaan Oromoo word-level speech dataset collected to support open-source speech recognition and text-to-speech technology.

Oro_Word

License: CC0-1.0

Locale: om

Task: TTS

Format: .WAV, CSV

Size: 1.28 MB

The languages represented in the corpus include Javanese (Pemalang dialect), Indonesian, English, and Arabic.

Pemalang Dialect of JAV-IND-ENG-ARB Parallel Corpus

License: CC-BY-SA-4.0

Locale: ind,jav,eng,arb

Task: MT

Format: TSV

Size: 555.62 KB

Text to speech dataset for Polish, male speaker, approximately 2 hours of read speech.

Darkman 1.0

License: CC0-1.0

Locale: pl-PL

Task: TTS

Format: WEBM

Size: 40.42 MB

2 hours of studio-quality, deliberate, read speech and corresponding text in Southeast Asian Cantonese, from a single speaker.

🚧 Southeast Asian Cantonese TTS dataset 🚧

License: MDC DLA-1.0

Locale: yue

Task: TTS

Format: WAV, TSV

Size: 202 B

Italian, ~50yo woman speaking in Ligurian, a language of Liguria, north-east Italy

Ligurian TTS - female voice

License: CC-BY-NC-SA-4.0

Locale: lij

Task: TTS

Format: WAV, WEBM, TSV, TGZ

Size: 328.77 MB

Bulu ALCAM multimodal dataset: lexical entries and example sentences in Bulu (IPA) with French equivalents, audio recordings, and alignment file.

Bulu_ALCAM-MultimodalDataset

License: NOODL-1.0

Locale: bum

Task: NLP

Format: MP3, TSV

Size: 31.28 MB

A Hazargi speech translation dataset with approximately 10 hours of Hazargi audio with English text translations.

🚧 Hazargi-English Speech Translation Corpus 🚧

License: MDC DLA-1.0

Locale: haz, en

Task: MT

Format: MP3, TSV

Size: 511.58 MB

MDC Community Concierge

VoxForge - Greek

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

License: GPL-3.0

Locale: el

Task: ASR

Format: WAV, TSV

Size: 954.26 MB

Bulu (bum) scripted speech dataset: 819 clips from 8 speakers (predominantly female) across 9 recording sessions.

Bulu-ASR-Dataset

License: NOODL-1.0

Locale: bum

Task: ASR

Format: MP3, TSV

Size: 42.11 MB