Loading datasets...
Release Date: 6/17/2026
Format: MP3
Size: 5.99 GB
Share
A collection of read speech recordings in Cantonese (粵語).
Restrictions/Special Constraints
None provided.
Forbidden Usage
It is forbidden to attempt to determine the identity of speakers in the Common Voice datasets. It is forbidden to re-host or re-share this 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.
yue)This datasheet is for cv-corpus-26.0-2026-06-12 of the Mozilla Common Voice Scripted Speech dataset for Cantonese [粵語 - yue]. The dataset contains 279366 clips representing 307.41 hours of recorded speech (210.7 hours validated) from 1183 speakers, recorded from a text corpus of 28,727 sentences.
| Code | Accent | Clips | Speakers |
|---|---|---|---|
| - | 126,764 (45.4%) | 145 (12.3%) |
The dataset includes the following self-declared age and gender distributions. A coverage summary is shown below each table.
Self-declared gender information. The table shows clip and speaker counts with percentages. Speakers who did not declare a gender are listed as Unspecified. A dash (-) indicates zero.
| Code | Gender | Clips | Speakers |
|---|---|---|---|
| male_masculine | Male, masculine | 47,656 (17.1%) | 114 (9.6%) |
| female_feminine | Female, feminine | 174,436 (62.4%) | 132 (11.2%) |
| transgender | Transgender | - | - |
| non-binary | Non-binary | 1,014 (0.4%) | 2 (0.2%) |
| do_not_wish_to_say | Prefer not to say | 59 (0.0%) | 1 (0.1%) |
| - | Unspecified | 56,201 (20.1%) | 1,008 (85.2%) |
Gender declared: 223,165 of 279,366 clips (79.9%), 175 of 1,183 speakers (14.8%)
Self-declared age information. The table shows clip and speaker counts with percentages. Speakers who did not declare an age are listed as Unspecified. A dash (-) indicates zero.
| Code | Age | Clips | Speakers |
|---|---|---|---|
| teens | Teens | 1,999 (0.7%) | 20 (1.7%) |
| twenties | Twenties | 81,658 (29.2%) | 118 (10.0%) |
| thirties | Thirties | 131,250 (47.0%) | 93 (7.9%) |
| fourties | Fourties | 10,231 (3.7%) | 40 (3.4%) |
| fifties | Fifties | 222 (0.1%) | 5 (0.4%) |
| sixties | Sixties | 600 (0.2%) | 4 (0.3%) |
| seventies | Seventies | - | - |
| eighties | Eighties | - | - |
| nineties | Nineties | 1,004 (0.4%) | 1 (0.1%) |
| - | Unspecified | 52,402 (18.8%) | 989 (83.6%) |
Age declared: 226,964 of 279,366 clips (81.2%), 194 of 1,183 speakers (16.4%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 191,477 (68.5%) |
| Invalidated | 8,096 (2.9%) |
| Other | 79,793 (28.6%) |
Training splits
| Split | Clips |
|---|---|
| Train | 7,420 (3.9%) |
| Dev | 5,130 (2.7%) |
| Test | 5,130 (2.7%) |
Training split coverage: 17,680 of 191,477 validated clips (9.2%)
The dataset contains 191477 validated, 8096 invalidated, and 79793 unresolved clips. The average clip duration is 3.961 seconds.
Validated sentences: 18,556
| Category | Count |
|---|---|
| Unvalidated sentences | 10,171 |
| Pending sentences | 10,157 |
| Rejected sentences | 14 |
| Reported sentences | 2,257 |
The corpus contains 28,727 sentences: 18,556 validated and 10,171 unvalidated (10,157 pending review, 14 rejected), with 2,257 reported for review.
There follows a randomly selected sample of five sentences from the corpus.
睇定個大概先
但係佢話教咗好耐學生都未熟五十音
如果中文係漢字，噉日本越南喃字都係中文囉
不過，有得打字
噉你都啱
| Source | Sentences |
|---|---|
| sentence-collector | 9,444 (50.9%) |
| Personal chat history | 6,001 (32.3%) |
| Group chat history | 2,804 (15.1%) |
| Other | 307 (1.7%) |
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | 1,682 (0.6%) | 83 (7.0%) |
| agriculture_food | Agriculture and Food | 18 (0.0%) | 18 (1.5%) |
| automotive_transport | Automotive and Transport | 52 (0.0%) | 30 (2.5%) |
| finance | Finance | 128 (0.0%) | 32 (2.7%) |
| service_retail | Service and Retail | 613 (0.2%) | 52 (4.4%) |
| healthcare | Healthcare | 417 (0.1%) | 48 (4.1%) |
| history_law_government | History, Law and Government | 67 (0.0%) | 28 (2.4%) |
| media_entertainment | Media and Entertainment | 192 (0.1%) | 36 (3.0%) |
| nature_environment | Nature and Environment | 188 (0.1%) | 39 (3.3%) |
| news_current_affairs | News and Current Affairs | 114 (0.0%) | 31 (2.6%) |
| technology_robotics | Technology and Robotics | 95 (0.0%) | 30 (2.5%) |
| language_fundamentals | Language Fundamentals | - | - |
Each row of a tsv file represents a single audio clip, and contains the following information:
client_id - hashed UUID of a given user
path - relative path of the audio file
sentence - the sentence to be read aloud
sentence_id - unique identifier for the sentence
sentence_domain - domain classification(s) of the sentence
up_votes - number of people who said audio matches the text
down_votes - number of people who said audio does not match text
age - age of the speaker1
gender - gender of the speaker1
accents - accents of the speaker1
variant - variant of the language1
locale - locale code of the language
segment - if sentence belongs to a custom dataset segment, it will be listed here
validated_sentences.tsvThe validated_sentences.tsv file contains one row per validated sentence in the text corpus:
sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
is_used - whether the sentence is still in circulation for recording
clips_count - number of clips recorded for this sentence
unvalidated_sentences.tsvThe unvalidated_sentences.tsv file contains one row per unvalidated sentence in the text corpus:
sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
up_votes - number of upvotes the sentence received
down_votes - number of downvotes the sentence received
status - current status of the sentence (pending or rejected)
This 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 demographics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2 ↩3 ↩4