Loading datasets...
License:
CC0-1.0
Steward:
Common VoiceDataset ID:
cmflnn482w6vwwwhu2h3h13m5
Release Date: 9/15/2025
Format: MP3
Size: 458.10 MB
A collection of scripted spoken phrases in Min Nan Chinese (Taiwan).
Restrictions/Special Constraints
You agree that you will not re-host or re-share this dataset
Forbidden Usage
You agree not to attempt to determine the identity of speakers in the Common Voice dataset
Intended Use
This dataset is intended to be used for training and evaluating automatic speech recognition (ASR) models. It may also be used for applications relating to computer-aided language learning (CALL) and language or heritage revitalisation.
nan-tw)This datasheet has been generated automatically, we would love to include more information, if you would like to help out, get in touch!
This datasheet is for version 23.0 of the the Mozilla Common Voice Scripted Speech dataset
for Taiwanese (Minnan) (nan-tw). The dataset contains 31951 clips representing 24 hours of recorded
speech (21 hours validated) from 290 speakers.
The dataset includes the following distribution of age and gender.
Self-declared gender information, percentage refers to the number of clips annotated with this gender.
| Gender | Pertentage |
|---|---|
| Undefined | 26.0% |
| Male Masculine | 63.0% |
| Female Feminine | 11.0% |
Self-declared age information, percentage refers to the number of clips annotated with this age band.
| Age Band | Percentage |
|---|---|
| Undefined | 13.0% |
| Twenties | 50.0% |
| Thirties | 15.0% |
| Teens | 1.0% |
| Fourties | 17.0% |
| Fifties | 1.0% |
| Sixties | 2.0% |
The official data splits for modelling this language are as follows. Of the validated clips, 81.31% are included in the splits.
| Split | Count |
|---|---|
| Train | 11584 |
| Test | 6523 |
| Dev | 5705 |
The text corpus contains 27277 sentences, of which 26907 are validated, 370 are invalidated and 226 are reported.
There follows a randomly selected sample of five sentences from the corpus.
喙（tshuì）
乞食趕廟公（khit-chia̍h kóaⁿ biō-kong）
鹿港人講話無相仝（Lo̍k-káng lâng kóng uē bô sio-kâng）
功夫到厝，欲食就有（kang-hu kàu tshù, beh tsia̍h tio̍h ū）
人食一點氣（lâng tsia̍h tsi̍t tiám khuì）
| Domain | Count |
|---|---|
| Undefined | 31563 |
| Agriculture Food | 4 |
| Service Retail | 2 |
| General | 327 |
| Healthcare | 100.0% |
| Language Fundamentals | 55 |
| Nature Environment | 55 |
Each row of a tsv file represents a single audio clip, and contains the following information:
client_id - hashed UUID of a given userpath - relative path of the audio filetext - supposed transcription of the audioup_votes - number of people who said audio matches the textdown_votes - number of people who said audio does not match textage - age of the speaker1gender - gender of the speaker1accent - accent of the speaker1segment - if sentence belongs to a custom dataset segment, it will be listed hereThis dataset is released under the Creative Commons Zero (CC-0) licence. By downloading this data you agree to not determine the identity of speakers in the dataset.
For a full list of age, gender, and accent options, see the demograpics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2 ↩3