Common Voice Scripted Speech 25.0 - Torwali

توروالی — Torwali (trw)

This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Torwali [توروالی - trw]. The dataset contains 12431 clips representing 18.55 hours of recorded speech (16.49 hours validated) from 27 speakers, recorded from a text corpus of 7,797 sentences.

Language

Torwali is an Indo-Aryan language belongs to its sub-group Dardic and is spoken by about 150k people in the upper mountainous reaches of the Swat valley in Pakistan.

Accents

Demographic information

The dataset includes the following self-declared age and gender distributions. A coverage summary is shown below each table.

Gender

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.

Gender declared: 0 of 12,431 clips (0.0%), 0 of 27 speakers (0.0%)

Age

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.

Age declared: 11,760 of 12,431 clips (94.6%), 12 of 27 speakers (44.4%)

Data splits for modelling

Clip buckets

Training splits

Training split coverage: 6,898 of 11,054 validated clips (62.4%)

The dataset contains 11054 validated, 1363 invalidated, and 14 unresolved clips. The average clip duration is 5.373 seconds.

Text corpus

There are more than 7000 sentences from various books produced in Torwali by the local indigenous organization Idara Baraye Taleem wa Taraqi (IBT) and its volunteers.

Validated sentences: 7,770

The corpus contains 7,797 sentences: 7,770 validated and 27 unvalidated (20 pending review, 7 rejected), with 578 reported for review.

Writing system

The writing system is based on the Perso-Arabic script

Symbol table

أ ا اَ آ اِ اُ اُو او ای اے ب پ ت ٹ ث ج چ ڇ ح خ د ڈ ذ ر ڑ ژ ز ڙ س ش ݜ ص ض ط ظ ع غ ک گ ف ق ل م ن ں ہ ء ے ی بھ پھ تھ ٹھ جھ چھ ڇھ دھ ڈھ ڑھ ڙھ کھ گھ لھ مھ نھ

Sample

There follows a randomly selected sample of five sentences from the corpus.

1. بوپ دأدی جیِب سی بنیاد تے دأل جیبا می تعلیم سی منصوبا چھی۔

2. آ کم گے بیَدو

3. ملانگ کمیئی کمیئی بے ہُوش ہُوئی او بارام شاھزدا ایسی أنگی سارُو سأد۔

4. سے تعداد می تھلا کم آشی خو مُوٹھائے بیشتے گھین طاقت سوابش بھؤدُود

5. رینگی سالُو بنَدُو: ”تأ آج یِؤ سی غلطی کی! تُو ازر خطا ہُو۔“

Sources

1. Torwali-Urdu-English Dictionary by Idara Baraye Taleem wa Taraqi (IBT)

2. Torwali-English Dictionary for Students by Idara Baraye Taleem wa Taraqi (IBT)

3. Torwali Folktales by Idara Baraye Taleem wa Taraqi (IBT)

4. Torwali children’s books by Idara Baraye Taleem wa Taraqi (IBT)

5. Inaan 1 by Idara Baraye Taleem wa Taraqi (IBT)

6. Inaan 2 by Idara Baraye Taleem wa Taraqi (IBT)

7. Jinnah si Qisa by Idara Baraye Taleem wa Taraqi (IBT)

8. Fatima Jinnah si Qisa by Idara Baraye Taleem wa Taraqi (IBT)

9. Allama Iqbal si Qisia by Idara Baraye Taleem wa Taraqi (IBT)

10. Kashmala by Idara Baraye Taleem wa Taraqi (IBT)

11. Sultan Bahu Si Qisa by Idara Baraye Taleem wa Taraqi (IBT)

12. Surat-e-Rahman si Tarjuma by Idara Baraye Taleem wa Taraqi (IBT)

13. Akhrir Dash Surata- si An Nimaaz si Tarjuma by Idara Baraye Taleem wa Taraqi (IBT)

14. Poetry by Javid Iqbal Torwali ‘

15. Poetry by Zubair Torwali

16. Essays by Zubair Torwali

17. Essays by Aftab Ahmad

18. Essays by Nisar Ahmad Torwali

19. Essays by Rahim Sabir

20. Novel by Zubair Torwali

Text domains

General, Agriculture and Food, Automotive and Transport, Finance, Service and Retail, Healthcare, History, Law and Governmant, Media and Entertainment, Nature and Environment, News and Current Affairs, Technology and Robotics, Language Fundamentals (e.g. Digits, Letters, Money)

Processing

Audio recording of the text corpus

Fields

Clips

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.tsv

The 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.tsv

The 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)

Get involved

Community links

• WhatsApp group

• Common Voice translators on Pontoon

• Common Voice Communities

Discussions

• Facebook post with more information

• Common Voice on Matrix

• Common Voice on Discourse

• Common Voice on Discord

• Common Voice on Telegram

Contribute

• Speak

• Write

• Listen

• Review

Acknowledgements

Datasheet authors

• Zubair Torwali

• Javid Iqbal Torwali

• Nisar Ahmad Torwali

• Rahim Sabir

• Aftab Ahmad

• Sajad Ahmad

Funding

This dataset was partially funded by the Open Multilingual Speech Fund managed by Mozilla Common Voice.

Licence

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.

Footnotes

1. 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