Datasetpublic

AfriSpeech/open-bible-speech-african

Open Bible Resources — African Languages Spoken-audio Bible recordings aligned to verse-level text for 19 African languages — roughly 1,741 hours of audio across ~552,907 audio–text pairs (~357 GB). This dataset is the African-language subset of davidguzmanr/open-bible-resources, re-hosted here by AfriSpeech to make the African languages easy to find and use on their own. The audio and text are unchanged from the source; only the non-African configurations have been removed. All… See the full description on the dataset page: https://huggingface.co/datasets/AfriSpeech/open-bible-speech-african.

sourceHugging Facecc-by-sa-4.0updated 3mo agoView on Hugging Face

3likes3.1kdownloads

Dataset Card

Open Bible Resources — African Languages

Spoken-audio Bible recordings aligned to verse-level text for 19 African languages — roughly 1,741 hours of audio across ~552,907 audio–text pairs (~357 GB).

This dataset is the African-language subset of `davidguzmanr/open-bible-resources`, re-hosted here by AfriSpeech to make the African languages easy to find and use on their own. The audio and text are unchanged from the source; only the non-African configurations have been removed. All credit for collecting and aligning the data goes to the original authors and to Open.Bible / Faith Comes By Hearing.

Languages

Hours are computed from the duration_seconds field (sum of all clips per language).

Language	Hours	Train	Test	Size
Chichewa (Nyanja)	111.1	28,805	1,517	14.8 GB
Dawro	116.9	28,100	1,479	13.2 GB
Dholuo (Luo)	76.2	28,907	1,522	21.7 GB
Ewe	96.1	28,655	1,509	24.7 GB
Gamo	107.1	28,690	1,510	12.9 GB
Gofa	86.8	28,868	1,520	10.8 GB
Hausa	99.7	29,181	1,536	23.8 GB
Igbo	94.2	28,510	1,501	25.7 GB
Kikuyu	87.0	29,185	1,537	12.6 GB
Lingala	127.8	27,350	1,440	32.4 GB
Luganda (Ganda)	101.4	28,918	1,522	13.2 GB
Matengo	28.2	7,402	390	7.8 GB
Ndebele	100.9	28,652	1,509	29.2 GB
Oromo	98.0	28,892	1,521	27.7 GB
Shona	74.5	29,150	1,535	11.6 GB
Swahili	96.4	29,102	1,532	13.3 GB
Twi (Akuapem)	70.7	28,756	1,514	18.3 GB
Twi (Asante)	78.6	29,036	1,529	20.0 GB
Yoruba	89.9	29,093	1,532	23.7 GB
Total (19)	1,741	525,252	27,655	357 GB

Dataset structure

Each language is a separate config with train and test splits. Every example contains:

Field	Type	Description
`audio`	audio	Verse-level audio recording
`text`	string	Transcription of the verse
`testament`	string	`Old` or `New` Testament
`book`	string	Bible book name
`chapter`	string	Chapter number
`verse`	string	Verse number
`duration_seconds`	float	Audio duration in seconds
`speaker_id`	string	Speaker/recording identifier

Usage

python

from datasets import load_dataset

# Load one language (config name = language)
ds = load_dataset("AfriSpeech/open-bible-speech-african", "Yoruba")
print(ds)
print(ds["train"][0]["text"], ds["train"][0]["audio"])

# Stream instead of downloading everything
ds = load_dataset("AfriSpeech/open-bible-speech-african", "Swahili", streaming=True)
for ex in ds["train"]:
    print(ex["text"]); break

Config names are exactly the language names in the table above (e.g. "Twi (Akuapem)", "Dholuo").

Intended uses

—Automatic Speech Recognition (ASR) for under-resourced African languages
—Text-to-Speech (TTS) and speech synthesis
—Speech-to-text alignment, pronunciation modelling, and multilingual audio research

Licensing & attribution

Released under CC BY-SA 4.0, following the source dataset. The underlying recordings originate from the Open.Bible project. Please cite the original dataset and the Open.Bible source when using this data, and comply with the CC BY-SA 4.0 share-alike terms.

This is a redistribution for convenience. If you are the data owner and would like changes, please open a discussion on this repository.