Loading datasets...
Task: ASR
Release Date: 3/22/2026
Format: MP3
Size: 198.07 MB
Share
A collection of read speech recordings in Brahui (براہوئی).
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.
brh)This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Brahui [براہوئی - brh]. The dataset contains 5095 clips representing 9.93 hours of recorded speech (9.91 hours validated) from 18 speakers, recorded from a text corpus of 3,095 sentences.
Brahui is a dravidian language spoken primarily in Central and western Balochistan, Pakistan, with smaller communities in Afghanistan and Iran. Despite being geographically isolated from other Dravidian languages, Brahui has retained its distinct linguistics identity for centuries. It is spoken by around 2.8 million people and coexists with surrounding languages such as Balochi, Pashtu and Persian, which has influenced its vocabulary and structure. The language is traditionally oral, though it has been written using the Arabic scripts in modern times. Brahui holds historical significances and contributes to the rich linguistic diversity of the region. The language has three main dialects: Sarawani, Jalawani and Rakhshani, with Sarawani being the standard dialect.
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 | - | - |
| female_feminine | Female, feminine | - | - |
| transgender | Transgender | - | - |
| non-binary | Non-binary | - | - |
| do_not_wish_to_say | Prefer not to say | - | - |
| - | Unspecified | 5,095 (100.0%) | 18 (100.0%) |
Gender declared: 0 of 5,095 clips (0.0%), 0 of 18 speakers (0.0%)
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 | - | - |
| twenties | Twenties | - | - |
| thirties | Thirties | 990 (19.4%) | 1 (5.6%) |
| fourties | Fourties | - | - |
| fifties | Fifties | - | - |
| sixties | Sixties | - | - |
| seventies | Seventies | - | - |
| eighties | Eighties | - | - |
| nineties | Nineties | - | - |
| - | Unspecified | 4,105 (80.6%) | 18 (100.0%) |
Age declared: 990 of 5,095 clips (19.4%), 0 of 18 speakers (0.0%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 5,086 (99.8%) |
| Invalidated | 3 (0.1%) |
| Other | 6 (0.1%) |
Training splits
| Split | Clips |
|---|---|
| Train | 1,703 (33.5%) |
| Dev | 506 (9.9%) |
| Test | 882 (17.3%) |
Training split coverage: 3,091 of 5,086 validated clips (60.8%)
The dataset contains 5086 validated, 3 invalidated, and 6 unresolved clips. The average clip duration is 7.019 seconds.
The corpus consists of spontaneous spoken texts in the Sarawani dialect of Brahui. It includes a large number of sentences and utterances, primarily drawn from natural conversations and everyday speech of native speakers. The dataset emphasizes oral and informal registers, reflecting authentic language use rather than standardized or literary forms.
Validated sentences: 3,095
| Category | Count |
|---|---|
| Unvalidated sentences | - |
| Pending sentences | - |
| Rejected sentences | - |
| Reported sentences | - |
The corpus contains 3,095 sentences: 3,095 validated and 0 unvalidated (0 pending review, 0 rejected), with 0 reported for review.
Although Brahui is traditionally an oral language, in modern times it has been written using the Perso-Arabic script with modifications to represent specific Brahui sounds.
ا ب پ ت ٹ ث ج چ ح خ د ڈ ذ ر ڑ ز ژ س ش ص ض ط ظ ع غ ف ق ک گ ل ڷ م ن ں و ہ ھ ی ے
There follows a randomly selected sample of five sentences from the corpus.
عہدہ سے آ سر کننگ کن ننا سادہ انگا مخلوق ءِ ہروار ریفرہ کہ وزیر
شاعر اسے
اٹ (مرحوم) نوشتہ کننگ ٹہپک او ہمو نا پن اٹ مریک ہرادے کس تینا نیام
نظم دا تو افسانوی ادب نا بشخ آک ءُ
اف دا بندغ انتئے دن کرے پنجاب اٹ خواننگ نا وخت آ اودے ندیم پاننگا‘
The corpus originates from a Brahui book. Sentences and short texts were created from the book and then spoken by native speakers for recording. Thus, the primary source is written Brahui material adapted into spoken form.
| Source | Sentences |
|---|---|
| احمد نعیم: ادبی کڑد و گشاد | 3,095 (100.0%) |
General
The original Brahui book was segmented into smaller sentences and short texts. These were distributed to native speakers, who recorded themselves reading the material.
Further post-processing may include verifying the alignment between text and audio, checking for transcription errors, and standardizing the text format.
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
text - supposed transcription of the audio
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
segment - if sentence belongs to a custom dataset segment, it will be listed here
prompt_upvotes - number of upvotes the sentence prompt received
prompt_reports - number of reports the sentence prompt received
is_edited - whether the clip's transcription has been edited
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)
Dr. Zia ur Rehman
Muhammad Imran
This dataset was partially funded by the Open Multilingual Speech Fund managed by Mozilla Common Voice.
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