datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
common-voice-13-faThe Persian portion of the original CommonVoice 13 dataset at https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0
Load
# Using HF Datasets
from datasets import load_dataset
dataset = load_dataset("hezarai/common-voice-13-fa", split="train")
# Using Hezar
from hezar.data import Dataset
dataset = Dataset.load("hezarai/common-voice-13-fa", split="train")
wikimedia-common-audio-catalanThis is a collection of Catalan-language audio with free licenses extracted from Wikimedia Commons.
License identifiers are normalized to cc-zero, cc-by-4.0,
cc-by-sa-3.0, cc-by-sa-4.0, GFDL, and PD-self.
This provides a richer alternative to Common Voice.
Characteristics of the dataset:
One or multiple speakers
Different accents
Different domain texts
761 audio files
We found this dataset useful for audio tasks such as:
Language detection
Evaluation of STT systems
New candidates are… See the full description on the dataset page: https://huggingface.co/datasets/softcatala/wikimedia-common-audio-catalan.commonvoice22-sidon-dacvae
CommonVoice 22 (Sidon-enhanced) converted to DAC VAE latents
Source
sarulab-speech/commonvoice22_sidon
Format
Each tar shard (~2GB) contains samples with three files per sample:
{sample_key}.audio.flac # Original audio (FLAC, original sample rate)
{sample_key}.dacvae.npy # DAC VAE latent [T_latent, 128] numpy float32
{sample_key}.metadata.json # All metadata + duration_seconds + chars_per_second
DAC VAE Latent Format
Model:… See the full description on the dataset page: https://huggingface.co/datasets/TTS-AGI/commonvoice22-sidon-dacvae.YouTube-Commons-nl-audio
YouTube Commons NL Audio
This dataset has audio files for the Dutch-language videos in Rijgersberg/YouTube-Commons-nl-transcriptions,
all under a CC BY 4.0 license.
It contains 11,669 files for a total runtime of 2493h 43m 5s, coming in at approximately 130 GB.
Source
The original source of the dataset (minus the titles, descriptions and audio files) is YouTube Commons:
YouTube-Commons is a collection of audio transcripts of 2,063,066 videos shared on YouTube… See the full description on the dataset page: https://huggingface.co/datasets/Rijgersberg/YouTube-Commons-nl-audio.common_voice_19_0_zh-TW
Common Voice Corpus 19.0 Chinese (Taiwan)
The test set is the same as the original test set, while validated_without_test includes all validated examples except those with sentence IDs that appear in the test set.
validated_without_test has about 50,000 examples in total, equivalent to approximately 44 hours, and is intended for use as the training set.
test has about 5,000 examples, which is approximately 5 hours.
common-voice-18-arabic
Dataset Card for Common Voice 18 – Arabic Edition
Dataset Summary
This dataset is an unofficial Arabic-only extraction of Mozilla Common Voice Corpus 18.0, prepared for Automatic Speech Recognition (ASR) research and development.
It is derived from the original Common Voice 18 release and filtered to include Arabic (ar) speech data only, while preserving the original dataset structure, splits, and metadata fields. The dataset consists of validated, unvalidated, and… See the full description on the dataset page: https://huggingface.co/datasets/MohamedRashad/common-voice-18-arabic.common_voice_13_french_phoneme
Common Voice 13 French Phoneme
Dataset Summary
This dataset is a curated version of the French subset of Common Voice 13.0, enriched with a phonetic transcription column (phoneme).
It was created by the Laboratoire de Mécanique des Structures et des Systèmes Couplés (Cnam-LMSSC) to support research in speech processing, specifically for tasks requiring phonetic alignment, phoneme recognition, and robust speech-to-text applications in French.
The dataset retains the… See the full description on the dataset page: https://huggingface.co/datasets/Cnam-LMSSC/common_voice_13_french_phoneme.commonvoice-12.0-arabic-voice-converted
Dataset Card for Voice Converted Arabic Common Voice 12.0
This dataset is derived from the Common Voice Arabic Corpus 12.0 and includes automatically diacritized transcriptions and phoneme representations for the original augmented audio data. The recordings feature Arabic text read aloud by users, where the text was initially undiacritized, allowing for potential reading errors. The diacritization and phonemes were generated automatically, resulting in a dataset that is valuable… See the full description on the dataset page: https://huggingface.co/datasets/xmodar/commonvoice-12.0-arabic-voice-converted.common_voice_20_armenian
Common Voice 20 - Armenian
This dataset is the Armenian portion of Mozilla's Common Voice 20.0 release,
a massively multilingual collection of transcribed speech intended for speech technology research and development.
Dataset Details
Language: Armenian (hy)
Source: Mozilla Common Voice
Version: 20.0
License: CC0-1.0
common_voice_25_zh-TW
Common Voice Scripted Speech 25.0 - Chinese (Taiwan)
This repository mirrors the Chinese (Taiwan) (zh-TW) portion of Mozilla Common Voice Scripted Speech 25.0 in Hugging Face datasets format. The audio has been embedded into Parquet and exposed as a Hugging Face Audio feature at 48 kHz.
Official source page: Mozilla Data Collective - Common Voice Scripted Speech 25.0 - Chinese (Taiwan)
Dataset Details
Field
Value
Dataset ID
cmn2g7eaj01fio10769r1m96n… See the full description on the dataset page: https://huggingface.co/datasets/OpenFormosa/common_voice_25_zh-TW.commonvoice_17_tr_fixed
Improving CommonVoice 17 Turkish Dataset
I recently worked on enhancing the Mozilla CommonVoice 17 Turkish dataset to create a higher quality training set for speech recognition models.Here's an overview of my process and findings.
Initial Analysis and Split Organization
My first step was analyzing the dataset organization to understand its structure.Through analysis of filename stems as unique keys, I revealed and documented an important aspect of CommonVoice's design… See the full description on the dataset page: https://huggingface.co/datasets/ysdede/commonvoice_17_tr_fixed.Common_Voice_Corpus_22_0_Urdu
Common Voice Corpus 22.0 - Urdu
This dataset contains the Urdu subset of the Mozilla Common Voice 22.0 corpus, released in June 2025.It consists of crowdsourced speech recordings and their corresponding text transcriptions, collected to support open-source speech technology.
Dataset Summary
The Common Voice Corpus 22.0 Urdu dataset provides high-quality speech data for automatic speech recognition (ASR), speaker identification, and linguistic research in Urdu.It includes… See the full description on the dataset page: https://huggingface.co/datasets/azeem-ahmed/Common_Voice_Corpus_22_0_Urdu.common-voice-26-mn
Common Voice 26.0 Mongolian (cleaned)
A quality-filtered, normalised subset of Mozilla Common Voice Corpus 26.0, Mongolian, prepared for training
Mongolian (Khalkha Cyrillic) text-to-speech with
oron-tts.
Built by oron-cleaner. Every threshold
was calibrated on this corpus, and every number and column on this page is read
from the shipped data rather than asserted.
from datasets import load_dataset
ds = load_dataset("btsee/common-voice-26-mn", split="train")
print(ds[0]["text"]… See the full description on the dataset page: https://huggingface.co/datasets/btsee/common-voice-26-mn.nigerian_common_voice_dataset
Dataset Card for Nigerian Common Voice Dataset
Dataset Summary
The Nigerian Common Voice Dataset is a comprehensive dataset consisting of 158 hours of audio recordings and corresponding transcription (sentence).
This dataset includes metadata like accent, locale that can help improve the accuracy of speech recognition engines. This dataset is specifically curated to address the gap in speech and language
datasets for African accents, making it a valuable resource for… See the full description on the dataset page: https://huggingface.co/datasets/benjaminogbonna/nigerian_common_voice_dataset.common_voice_26_0_de
Mozilla Common Voice 26.0 - German (IPA & Clean Validated Subset)
Repacking version of Common Voice 26.0 German officialy published by Mozilla Data Collective, following Hugging Face Parquet Shards standard, with feature for listening to audio directly on the Web Hub, and the addition of a data column for the IPA transcription of each sentence.
📊 Dataset parameters
Origin: Mozilla Common Voice 26.0 (version 18/06/2026).
Data amount (Validated): 950,877 MP3 audio… See the full description on the dataset page: https://huggingface.co/datasets/q1805/common_voice_26_0_de.common_voice_26_0
Dataset Card for Common Voice Corpus 26.0
This dataset is an unofficial version of the Mozilla Common Voice Corpus 26. It was downloaded and converted from the project's website https://commonvoice.mozilla.org/.
Languages
Abkhaz, Albanian, Amharic, Arabic, Armenian, Assamese, Asturian, Azerbaijani, Basaa, Bashkir, Basque, Belarusian, Bengali, Breton, Bulgarian, Cantonese, Catalan, Central Kurdish, Chinese (China), Chinese (Hong Kong), Chinese (Taiwan), Chuvash… See the full description on the dataset page: https://huggingface.co/datasets/deepdml/common_voice_26_0.commonvoice_brythonic
Description
Audios en gallois et en cornique proche du breton d'après ce site.
Plus précisément, nous avons récupéré toutes les lignes de la page web et gardé que celles où l'un des mots en breton/gallois/cornique est encore utilisé de nos jours.Puis nous avons récupéré les audios en gallois et en cornique de common voice.Finalement nous n'avons gardé que les audios de common voice contenant les mots proches en gallois/cornique du mot en breton sélectionné.Nous avons ajouté… See the full description on the dataset page: https://huggingface.co/datasets/Bretagne/commonvoice_brythonic.common_voice_26_0_fa
Common Voice 26.0 - Persian (fa)
This is an unofficial filtered Persian (Farsi) version of the latest Common Voice dataset version 26.0, obtained from the official Mozilla Common Voice platform at https://commonvoice.mozilla.org/en/datasets. It contains carefully curated and fully validated audio samples aimed at providing a clean and reliable resource for speech processing in Persian. As with the previous version, this dataset is not an official Mozilla release but a… See the full description on the dataset page: https://huggingface.co/datasets/aliyzd95/common_voice_26_0_fa.CommonPhoneDataset
Dataset Card for Common Phone
This corpus aims to provide a basis for Machine Learning (ML) researchers and enthusiasts to train and test their models against a wide variety of speakers, hardware/software ecosystems and acoustic conditions to improve generalization and availability of ML in real-world speech applications.
The current version of Common Phone comprises 116,5 hours of speech samples, collected from 11.246 speakers in 6 languages.
Common Phone has been used as the… See the full description on the dataset page: https://huggingface.co/datasets/pklumpp/CommonPhoneDataset.common-voice-20-mn-normalized
Common Voice 20.0 Mongolian Dataset
This dataset is a subset of Mozilla's Common Voice project, containing Mongolian speech data. It's part of Common Voice 20.0 release.
Dataset Structure
The dataset contains:
Audio clips in .mp3 format
Transcriptions for each audio clip
Train/test/dev splits
Additional metadata including speaker demographics
Usage
This dataset can be used for:
Speech Recognition
Voice Analysis
Linguistic Research
Speech Processing… See the full description on the dataset page: https://huggingface.co/datasets/warmestman/common-voice-20-mn-normalized.common_voice_16-zh-CNcommon-voice-corpus-20common-voice-23-0-es-ipa-v1
Common Voice 23.0 Spanish — canonical IPA v1
This repository is the approved deterministic v1 publication cut of Spanish Common Voice 23.0 recipe cuts. It contains the accepted final Lhotse cuts only:
split
rows
train
352,941
test
15,857
The data files have exactly three columns:
id — the unique final Lhotse cut identifier (for example, common_voice_es_19696062-9131).
audio — a Hugging Face Audio feature containing the original encoded source bytes and source… See the full description on the dataset page: https://huggingface.co/datasets/bookbot/common-voice-23-0-es-ipa-v1.common-voice-18-arabic
Dataset Card for Common Voice 18 – Arabic Edition
Dataset Summary
This dataset is an unofficial Arabic-only extraction of Mozilla Common Voice Corpus 18.0, prepared for Automatic Speech Recognition (ASR) research and development.
It is derived from the original Common Voice 18 release and filtered to include Arabic (ar) speech data only, while preserving the original dataset structure, splits, and metadata fields. The dataset consists of validated, unvalidated… See the full description on the dataset page: https://huggingface.co/datasets/ahmadafaneh/common-voice-18-arabic.common-voice-22-en-teens
Common Voice 22 - English Teens Subset
This dataset is a filtered subset of Mozilla's Common Voice Corpus 22.0, containing only recordings from teenage speakers in English.
Dataset Structure
Data Instances
Each instance contains:
audio: Audio file with embedded bytes, sampling rate, and path
text: Transcription of the spoken text
Data Splits
This dataset provides a single unified split containing all 77,500 samples. Users can create their own… See the full description on the dataset page: https://huggingface.co/datasets/cryptolock/common-voice-22-en-teens.baoule-common-voice
Common Voice — Baoulé (bci)
Dataset audio en langue baoulé (code ISO 639-3 : bci),
extrait de Mozilla Common Voice.
Le baoulé est une langue kwa parlée en Côte d'Ivoire.
Splits
Split
Exemples
train
~319
test
~290
validation
~267
Colonnes
Colonne
Type
Description
audio
Audio
Signal 16 kHz mono
sentence
string
Transcription en baoulé
client_id
string
Identifiant anonymisé du locuteur
path
string
Nom du fichier original… See the full description on the dataset page: https://huggingface.co/datasets/Klayt/baoule-common-voice.common-voice-urdu-processed
🎙️ Common Voice Urdu (Processed)
Ready-to-use Urdu speech dataset for fine-tuning ASR models
Mozilla Common Voice → Preprocessed → Whisper-Ready ✨
📊 Dataset at a Glance
Split
Samples
Use
🏋️ Train
7,339
Model training
🔧 Validation
5,046
Hyperparameter tuning
🧪 Test
5,091
Final evaluation
Total
17,476
💡 Audio is pre-resampled to 16kHz — plug directly into Whisper!
🚀 Quick Start
from datasets import load_dataset
#… See the full description on the dataset page: https://huggingface.co/datasets/khawajaaliarshad/common-voice-urdu-processed.jacob-common-voice-19-zh-TW-curatedbaoule-common-voice
Common Voice — Baoulé (bci)
Dataset audio en langue baoulé (code ISO 639-3 : bci),
extrait de Mozilla Common Voice.
Le baoulé est une langue kwa parlée en Côte d'Ivoire.
Splits
Split
Exemples
train
~319
test
~290
validation
~267
Colonnes
Colonne
Type
Description
audio
Audio
Signal 16 kHz mono
sentence
string
Transcription en baoulé
client_id
string
Identifiant anonymisé du locuteur
path
string
Nom du fichier… See the full description on the dataset page: https://huggingface.co/datasets/issoufzousko07/baoule-common-voice.common_voice_22.0_br
Description
Partie en breton du jeu de données Common Voice 22.0.
Champs
audio (dict) : Un dictionnaire contenant le chemin vers le fichier audio téléchargé, l'audio décodé et la fréquence d'échantillonnage.Notez que lors de l'accès à la colonne audio : dataset[0]["audio"], le fichier audio est automatiquement décodé et rééchantillonné à dataset.features["audio"].sampling_rate. Le décodage et le rééchantillonnage d'un grand nombre de fichiers audio peuvent prendre… See the full description on the dataset page: https://huggingface.co/datasets/Bretagne/common_voice_22.0_br.