Datasets:

michsethowusu
/

vai-speech-text-parallel

audio audioduration (s) 0.04 60	text stringlengths 3 148
	lɛi bii mu muã ɲiɲa kiiyɛi tiya yɔ mɔ mu ɓɔwa aijɛnɛɛ̃ lɔ.
	hĩ muã fala tiŋ mu ɓɛ mu lɔ masihoe la kowa,
	kaŋmɓa la fala ma ɲiɲaã saɓu lɔ,
	amu duuɲaã ma ɓuunaã nu ta makilaha ɓɛ.
	kaiyɛ fɛlanaã ɓɔwa aijɛnɛɛ̃ lɔ wa.
	kulɛ sunɗa ɓɛɛ ɓɛɛ mɛɛ̃ mu wowa ɓolo sii a kɔɔ amu wowaha kpasaŋŋɛ̃ vahale ɓɛ a ma kpã.
	ŋna ti wɛ lɛi mu ŋ ɓɛ.
	wo ɓɔɔ nu kuŋ a fɔwa kama kɛ faha ɓee kuŋ kunɛɲa?
	nja kpuu ma bii,
	anuã ko lɔ la ɓɔŋ.
	amu mu kpɛɛ mu lɔi falaŋnɗa.
	kɛ muã seinja ma kɛ kaŋmɓa la masihoe kunɛ wa kɛ faha lɔ.
	wo ti tuŋ wowaha sɔ kɛɛ lɔ majaha la kowa,
	amu kɛ duuɲaã la ɓaŋ wahatiyɛ la ke.
	hĩ ganaã nu gbi ɓanɗaha ɓɛ lumaã tɛ maã lɔ,
	kɛ ko mu mu ɓɛ a fɔna,
	kɛ masihoe pɛɛ ma kunɛ faha lɔ.
	ɓɛimaã,
	mɔ nu yaha sii tɛi la anɗɔ,
	ɓɛmupɛɛ anu ɓɔɔ nu la faha.
	mɔ mu nu ɓɛ aha kowa,
	ɓɛmupɛɛ ŋ tiya mɔ wa ko ŋ ma wolo ŋna wolo wahatiyɛ ma.
	i ma a wa tuu.
	amu jumɓu yaha ganaã ɓɔ sawaha lɔ wa.
	aha fiya ɓɔɛ ɓee a ye.
	kulɛ sunɗa mu pɛɛ keya ŋnɗa,
	amu aha fala ma ɲiɲaã ŋna keŋnɗa,
	a ɓɔwaŋnɗa a fɛ mɔɛ̃ ɓɔɔ nu njaha.
	amu ai majanjaha ɓee mu fa kaŋmɓa ye.
	suyɛɛ nu taha ɓɛ,
	hĩ wo ɓiyaha a ma kpã.
	kpanɗa maã,
	hĩ ɓuuna ɓɛ,
	kɛ,
	a ɓɛ fiya ɓɔ mɛɛ̃ lɔɔ waha kowa,
	kɛ,
	ɓɛimaã,
	a kɛ mɛɛ̃hĩ tiya lɔ a jiyanaã kɛ masihoe fɛ mɔɛ̃ kɛ mu nu la faha,
	anu kuluŋ ɓa nu ɓɛ wɛ kpɛ fiya ɓɔɛ lɔ,
	“faha la kowa ka,
	amu toomaã nu tɛ maã lɔ,
	kɛ,
	mɓe waha ko mu muhĩ ti tɔɔ lɔ wahati gbi,
	anɗaha tawolee,
	ɓɛmupɛɛ a fɔwa kaŋmɓa laha kulɛɛ lɔ kɛ koe kpɛɛi tiya masihoe la ganaã kɔɔ,
	hĩ faha nu ɓee kuŋ kunɛɲa,
	amu wahati mu a jɛɛlee,
	kii feŋ mu i naã a kaiya,
	a bii ma ko faha.
	muã sɛɛ sa kaŋmɓa la kɛ mu aha ma muã tiliŋŋɛ̃ kpai muã majaha masiho isa saɓu lɔ.
	kɛ kaŋmɓa laha kulɛɛ la tahayɛ,
	kɛ,
	a kɛ la ma saɓuyɛi tiya kɛ mu kaŋmɓa kuŋ tiya koe kpɛɛ ko aha ɲiɲa feŋŋɛ̃ nu kpɛɛ lɔ.
	amu hĩ masihoe ma kunɛ faha lɔ,
	kɛ kpã mu muã kinihĩ ti mɔ nu la ai ɓee mɔɛ̃ gbi la duuɲaã ma kpɛɛ.
	kɛ ɓuuna ɓɛ,
	lɛi mu mɔɛ̃ kpɛɛ la faha fɛɛ muã ɓɔɔyahanjaha lɔ adama fɛ,
	amu wo samale mu wɛ kpɛ jumɓu lɔ.
	“ai maã kama faha nuhĩ jɛɛ lɔ fiya ɓɔɛ lɔ?
	a jɛɛya lɔ a ɓɔwaŋnɗa jemisi njaha,
	hĩ tɔɔɓaha mu wo ɓɛ a fɔna,
	hĩ wo ma a bii ma,
	kɛ,
	amu anɗaha tawo wa kɛ.
	ŋnaã wa fɔ wo ye,
	a ɓee kuŋ fɛɛya ai to a lɔ suu bii toi.
	ai jiyaha kɛ mu ɓɛ faniɲa seinja wa mana a la sɔle ɓɛ kaŋmɓa la,
	“wahati mu kaŋmɓa la adamaã senjenaã ɲiɲa,
	kɛ,
	kɛ lɔɔ mu mɔɛ̃ kpɛɛi jɛɛya fiya ɓɔɛ lɔ muã ɓɔɔyahanjaha lɔ masihoe fɛ.
	kɛ,
	kɛ ɗeŋŋɛ̃ pɛɛ laha waŋga kɛ kaŋmɓa la ganaã kɔɔ.
	kɛ jii ma sɔɛ miɲana wa ɓɛ a lɔ kɛ mu mɔ yaha ko kaŋmɓa koe la mɔ fahalee nuã kowa?
	kɛ zi,
	kɛ,
	mu kpɛɛ mu ɓee faha.
	ŋna kaŋmɓa la mɔɛ̃ nu njaŋgata wa kɛ,
	ɓuuna lɛiyɛ miɲa wai tiya anu ma?
	komu wɛ ŋ ɲɔmɔ nu,
	hĩ faha nu ɓee kunɛ?
	kaŋmɓa la masihoe kunɛ faha lɔ,
	amu feŋ mu nu ɓɛ duuɲaã ma,
	mu kpɛɛ mu lɔi falaŋnɗa.
	saɓu muã kowa ŋ ɓɛ kɛ fɔna wo ye,
	kii mu kai lɔnɗɔ wa tiya saɓu ko faha naã duuɲaã ma,
	a ɓɔwaŋnɗa tenɗuyɛ nu kpɛɛ njaha.
	a bii ma jiya kɛ kaŋmɓa pɛɛi ti wa aha ganaã kɔɔ,
	kaiyɛ senjenaã ɓɔwa bunɗufuyɛ lɔ wa duuɲaã ma.
	amu tele sakpa ɓeele kaŋŋgɔɔ,
	amu lɔ,
	ɓɛimaã,
	sina muã koi kaha.
	kpã mu masihoei majanja tiyɛ ai to a lɔ kaŋmɓa yaha jooe nu gbi kɛ masihoe keŋŋgɔɔ.
	amu feŋ mu nu ɓɛ ɓanɗaha lɔ,
	kaŋmɓa yaha kɛ mu kunɛɲa faha lɔ,
	kɛ a lɔhɔmanaã la na.
	amu wowa wowaha kpasaŋŋɛ̃ kɛ ko bii wa lɔ.
	ɓɛimaã,
	amu a wa tiya tɔɔma sei ko kɛ mɔ mu nu filiya ki lɔ faha lɔ,
	a bii nafaha mu mɓe wa ŋnje?

End of preview. Expand in Data Studio

Vai Speech-Text Parallel Dataset

Dataset Description

This dataset contains 23286 parallel speech-text pairs for Vai, a language spoken primarily in Ghana. The dataset consists of audio recordings paired with their corresponding text transcriptions, making it suitable for automatic speech recognition (ASR) and text-to-speech (TTS) tasks.

Dataset Summary

Language: Vai - vai
Task: Speech Recognition, Text-to-Speech
Size: 23286 audio files > 1KB (small/corrupted files filtered out)
Format: WAV audio files with corresponding text files
Modalities: Audio + Text

Supported Tasks

Automatic Speech Recognition (ASR): Train models to convert Vai speech to text
Text-to-Speech (TTS): Use parallel data for TTS model development
Keyword Spotting: Identify specific Vai words in audio
Phonetic Analysis: Study Vai pronunciation patterns

Dataset Structure

Data Fields

audio: Audio file in WAV format
text: Corresponding text (latin) transcription from paired text file. You can use the vai-latin.csv file to translate latin form of Vai text into the traditional Vai script.

Data Splits

The dataset contains a single training split with 23286 filtered audio-text pairs.

Dataset Creation

Source Data

The audio data has been sourced ethically from consenting contributors. To protect the privacy of the original authors and speakers, specific source information cannot be shared publicly.

Data Processing

Audio files and corresponding text files were collected from organized folder structure
Text content was read from separate .txt files with matching filenames
Files smaller than 1KB were filtered out to ensure audio quality
Empty text files were excluded from the dataset
Audio was processed using the MMS-300M-1130 Forced Aligner tool for alignment and quality assurance

Annotations

Text annotations are stored in separate text files with matching filenames to the audio files, representing the spoken content in each audio recording.

Considerations for Using the Data

Social Impact of Dataset

This dataset contributes to the preservation and digital representation of Vai, supporting:

Language technology development for underrepresented languages
Educational resources for Vai language learning
Cultural preservation through digital archives

Discussion of Biases

The dataset may reflect the pronunciation patterns and dialects of specific regions or speakers
Audio quality and recording conditions may vary across samples
The vocabulary is limited to the words present in the collected samples

Other Known Limitations

Limited vocabulary scope (word-level rather than sentence-level)
Potential audio quality variations
Regional dialect representation may be uneven

Additional Information

Licensing Information

This dataset is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

Acknowledgments

Audio processing and alignment performed using MMS-300M-1130 Forced Aligner
The original audio is produced by The Bible Society of Liberia in partnership with Davar Audio Bibles

Citation Information

If you use this dataset in your research, please cite:

@dataset{vai_words_parallel_2025,
  title={Vai Words Speech-Text Parallel Dataset},
  year={2025},
  publisher={Hugging Face},
  howpublished={\url{https://huggingface.co/datasets/[your-username]/vai-speech-text-parallel}}
}

Contact

For questions or concerns about this dataset, please open an issue in the dataset repository.

Downloads last month: 40

Size of downloaded dataset files:

6.96 GB

Size of the auto-converted Parquet files:

6.96 GB

Number of rows:

23,286

Collection including michsethowusu/vai-speech-text-parallel

African Speech-Text Datasets

Collection

Collection of speech to text datasets for African languages • 11 items • Updated 1 day ago