Loading datasets...
Task: ASR
Release Date: 3/22/2026
Format: MP3
Size: 585.07 MB
Share
A collection of read speech recordings in Palula (پالولا).
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.
phl)This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Palula [پالولا - phl]. The dataset contains 21142 clips representing 28.9 hours of recorded speech (21.26 hours validated) from 20 speakers, recorded from a text corpus of 5,929 sentences.
Palula is an Indo-Aryan language, specifically a branch of the Dardic group, closely related to Shina. Palula is spoken in several regions of Chitral District, including Ashret, Biori, Kalkatak, and parts of Shishi Koh. Beyond Chitral, it is also spoken in Gumandan (Dir Kohistan) and Sao village in Kunar Province, Afghanistan. The estimated number of speakers ranges between 20,000 and 25,000.
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 | 20 (0.1%) | 1 (5.0%) |
| transgender | Transgender | - | - |
| non-binary | Non-binary | - | - |
| do_not_wish_to_say | Prefer not to say | - | - |
| - | Unspecified | 21,122 (99.9%) | 19 (95.0%) |
Gender declared: 20 of 21,142 clips (0.1%), 1 of 20 speakers (5.0%)
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 | 1,350 (6.4%) | 1 (5.0%) |
| twenties | Twenties | 10,055 (47.6%) | 8 (40.0%) |
| thirties | Thirties | 4,792 (22.7%) | 3 (15.0%) |
| fourties | Fourties | 3,778 (17.9%) | 3 (15.0%) |
| fifties | Fifties | - | - |
| sixties | Sixties | - | - |
| seventies | Seventies | - | - |
| eighties | Eighties | - | - |
| nineties | Nineties | - | - |
| - | Unspecified | 1,167 (5.5%) | 8 (40.0%) |
Age declared: 19,975 of 21,142 clips (94.5%), 12 of 20 speakers (60.0%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 15,558 (73.6%) |
| Invalidated | 69 (0.3%) |
| Other | 5,515 (26.1%) |
Training splits
| Split | Clips |
|---|---|
| Train | 1,946 (12.5%) |
| Dev | 1,408 (9.1%) |
| Test | 1,378 (8.9%) |
Training split coverage: 4,732 of 15,558 validated clips (30.4%)
The dataset contains 15558 validated, 69 invalidated, and 5515 unresolved clips. The average clip duration is 4.921 seconds.
The Palula text corpus has been systematically compiled through fieldwork conducted within the Palula-speaking community, encompassing a diverse range of speakers across age groups and social backgrounds. Data was gathered through structured and semi-structured interviews, oral storytelling sessions, and community-based linguistic elicitation. All recordings were carefully transcribed.The corpus comprises 4000 sentences.
Validated sentences: 4,745
| Category | Count |
|---|---|
| Unvalidated sentences | 1,184 |
| Pending sentences | 1,184 |
| Rejected sentences | - |
| Reported sentences | - |
The corpus contains 5,929 sentences: 4,745 validated and 1,184 unvalidated (1,184 pending review, 0 rejected), with 0 reported for review.
Palula orthography based on the Arabic writing system
ا ب پ ت ث ٹ ج چ ڇ څ ح خ د ڈ ذ ر ز ژ ڙ س ش ݜ ص ض ط ظ ع غ ف ق ک گ ل م ن ں ݨ و ہ ھ ء ی ے
There follows a randomly selected sample of five sentences from the corpus.
سوۡ سبزیۡ منڈیۡ تھےۡ گُوم۔
تھی خو شاید مہ دی اُمید نئینی۔
گھئِیمی ڇِھیر بیۡ کام بِھلوۡ۔خلکہ دُئی زھئی مُزدُریۡ تھئینی تھےۡ گِیا۔
نتجہ نہ نکھاتوۡ ہِنوۡ۔
میجرہ کُݨاکوم ݜاوا اُترپئینی مُقابِلہ تھوؤلوۡ۔
Palula Matli (Proverbs) – Author: Naseem Haider
Palula Textbook – Author: Naseem Haider
Palula Gul-Dasta Ashaar (Poetry) – Author: Naseem Haider
Palula–Urdu–English Conversation Guide – Author: Naseem Haider
Palula Folk Tales – Author: Naseem Haider
| Source | Sentences |
|---|---|
| self | 3,141 (66.2%) |
| پالولا گلدستہ اشعار | 783 (16.5%) |
| عام | 422 (8.9%) |
| Palula- Urdu- English Conversation | 399 (8.4%) |
General, Agriculture and Food, Healthcare, History, Law and Governmant, Media and Entertainment, Nature and Environment
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | - | - |
| agriculture_food | Agriculture and Food | - | - |
| automotive_transport | Automotive and Transport | 4 (0.0%) | 4 (20.0%) |
| finance | Finance | - | - |
| service_retail | Service and Retail | - | - |
| healthcare | Healthcare | 10 (0.0%) | 6 (30.0%) |
| history_law_government | History, Law and Government | 24 (0.1%) | 5 (25.0%) |
| media_entertainment | Media and Entertainment | - | - |
| nature_environment | Nature and Environment | - | - |
| news_current_affairs | News and Current Affairs | - | - |
| technology_robotics | Technology and Robotics | - | - |
| language_fundamentals | Language Fundamentals | - | - |
The process involved collecting texts, stories, and proverbs through community-based audio recordings. These recordings were transcribed, and selected sentences were curated for inclusion in the Common Voice dataset. The finalized dataset was then uploaded and subsequently recorded by multiple speakers.
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)
Naseem Haider
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