datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
fleurs
FLEURS
Fleurs is the speech version of the FLoRes machine translation benchmark.
We use 2009 n-way parallel sentences from the FLoRes dev and devtest publicly available sets, in 102 languages.
Training sets have around 10 hours of supervision. Speakers of the train sets are different than speakers from the dev/test sets. Multilingual fine-tuning is
used and ”unit error rate” (characters, signs) of all languages is averaged. Languages and results are also grouped into seven… See the full description on the dataset page: https://huggingface.co/datasets/google/fleurs.WaxalNLP
Waxal Datasets
The WAXAL dataset is a large-scale multilingual speech corpus for African languages, introduced in the paper WAXAL: A Large-Scale Multilingual African Language Speech Corpus.
Dataset Description
The Waxal project provides datasets for both Automated Speech Recognition (ASR)
and Text-to-Speech (TTS) for African languages. The goal of this dataset's
creation and release is to facilitate research that improves the accuracy and
fluency of speech and… See the full description on the dataset page: https://huggingface.co/datasets/google/WaxalNLP.gooshkon-chunks
Gooshkon MP3 chunks
A single-column Hugging Face Audio dataset. Every row contains embedded,
playable MP3 bytes in the audio column. Chunks target approximately 30
seconds and are selected at detected quiet intervals with a 12-48 second
safety range.
The source recordings are not transcript-aligned; this release uses acoustic
silence boundaries to avoid cutting through words whenever a usable pause is
available.
Dataset runtime
The dataset contains approximately 2… See the full description on the dataset page: https://huggingface.co/datasets/Reza2kn/gooshkon-chunks.xtreme_sXTREME-S covers four task families: speech recognition, classification, speech-to-text translation and retrieval. Covering 102
languages from 10+ language families, 3 different domains and 4
task families, XTREME-S aims to simplify multilingual speech
representation evaluation, as well as catalyze research in “universal” speech representation learning.sberdevices_golos_10h_crowd
Dataset Card for sberdevices_golos_10h_crowd
Dataset Summary
Sberdevices Golos is a corpus of approximately 1200 hours of 16kHz Russian speech from crowd (reading speech) and farfield (communication with smart devices) domains, prepared by SberDevices Team (Alexander Denisenko, Angelina Kovalenko, Fedor Minkin, and Nikolay Karpov). The data is derived from the crowd-sourcing platform, and has been manually annotated.
Authors divide all dataset into train and test subsets.… See the full description on the dataset page: https://huggingface.co/datasets/bond005/sberdevices_golos_10h_crowd.Taiwan-Tongues-ASR-CE-dataset-hokkien
Taiwan-Tongues-ASR-CE-dataset-hokkien
本資料集為 Taiwan-Tongues-ASR-CE 專案所使用的預訓練資料，透過 WebDataset 格式打包，並上傳至 Hugging Face 以便研究人員與開發者自由取用。
📂 Dataset 結構
本資料集分為 Training 與 Test 兩個子集，均以 WebDataset tar 檔案形式存放：
Training set (WebDataset format)
train/train-000000.tar
train/train-000001.tar
...
Test set (WebDataset format)
test/test-000000.tar
...
tsv set
train.tsv
test.tsv
...
每個 tar 內部均包含對應的音檔與標註，方便直接搭配 WebDataset 與 PyTorch / Hugging Face datasets 進行訓練與測試。… See the full description on the dataset page: https://huggingface.co/datasets/adi-gov-tw/Taiwan-Tongues-ASR-CE-dataset-hokkien.sberdevices_golos_100h_farfield
Dataset Card for sberdevices_golos_100h_farfield
Dataset Summary
Sberdevices Golos is a corpus of approximately 1200 hours of 16kHz Russian speech from crowd (reading speech) and farfield (communication with smart devices) domains, prepared by SberDevices Team (Alexander Denisenko, Angelina Kovalenko, Fedor Minkin, and Nikolay Karpov). The data is derived from the crowd-sourcing platform, and has been manually annotated.
Authors divide all dataset into train and test… See the full description on the dataset page: https://huggingface.co/datasets/bond005/sberdevices_golos_100h_farfield.japanese-casual-conversational-speech-golden-dataset-preview
Japanese Casual Conversational Speech Golden Dataset (Preview)
💼 Commercial License & Full Access
This repository contains a limited preview. The full 60-hour dataset collected via the "Kataro" app is available for commercial use, ASR benchmarking, and Spoken Dialogue Model fine-tuning.
To purchase the full dataset, please contact us:
👉 Email: info@hth-inc.com
👉 Website: https://hth-inc.com/business
🌟 4 Reasons to Choose This Dataset… See the full description on the dataset page: https://huggingface.co/datasets/HTH-inc/japanese-casual-conversational-speech-golden-dataset-preview.leyu-amharic-gonder-dialect
Leyu Amharic - Gonder Dialect Speech Corpus
Dataset Description
A parallel speech corpus of audio recordings paired with their transcripts, focused on the Gonder dialect of Amharic, for ASR and TTS research. Leyu reports that recordings were collected from contributors on mobile devices in real-world environments, and that each audio–text pair was manually reviewed for transcript accuracy and audio clarity.
This repository is a copy of… See the full description on the dataset page: https://huggingface.co/datasets/hadamard-2/leyu-amharic-gonder-dialect.leyu-amharic-gojjam-dialect
Leyu Amharic - Gojjam Dialect Speech Corpus
Dataset Description
A parallel speech corpus of audio recordings paired with their transcripts, focused on the Gojjam dialect of Amharic, for ASR and TTS research. Leyu reports that recordings were collected from contributors on mobile devices in real-world environments, and that each audio–text pair was manually reviewed for transcript accuracy and audio clarity.
This repository is a copy of… See the full description on the dataset page: https://huggingface.co/datasets/hadamard-2/leyu-amharic-gojjam-dialect.german-golden-audio_speech-IPA
🌟 German Golden Speech & IPA Corpus (FLEURS + Multilingual TEDx)
An ultra-clean, high-standard curated German speech dataset combining Google FLEURS (de_de) and Multilingual TEDx German (mTEDx), fully embedded with 16kHz WAV audio bytes, normalized orthographic text, and pre-computed International Phonetic Alphabet (IPA) transcriptions.
📊 Dataset Summary
Total Samples: 1,354 high-quality audio recordings.
Total Size: ~419 MB (Compressed Parquet format).
Audio… See the full description on the dataset page: https://huggingface.co/datasets/q1805/german-golden-audio_speech-IPA.Taiwan-Tongues-ASR-CE-dataset-zhtw
Taiwan-Tongues-ASR-CE-dataset-zhtw
本資料集為 Taiwan-Tongues-ASR-CE 專案所使用的預訓練資料，透過 WebDataset 格式打包，並上傳至 Hugging Face 以便研究人員與開發者自由取用。
📂 Dataset 結構
本資料集分為 Training 與 Test 兩個子集，均以 WebDataset tar 檔案形式存放：
Training set (WebDataset format)
train/train-000000.tar
train/train-000001.tar
...
Test set (WebDataset format)
test/test-000000.tar
...
tsv set
train.tsv
test.tsv
...
每個 tar 內部均包含對應的音檔與標註，方便直接搭配 WebDataset 與 PyTorch / Hugging Face datasets 進行訓練與測試。
🏷️… See the full description on the dataset page: https://huggingface.co/datasets/adi-gov-tw/Taiwan-Tongues-ASR-CE-dataset-zhtw.visualears-golden-6669
🗂️ visualears-golden-6669
English + فارسی · Part of Shenava 1.0 · Project hub · SLT paper submission
🌟 At a glance | معرفی سریع
English
فارسی
🎯 Purpose
Held-out VisualEars6669 / Golden6669 evaluation dataset.
مجموعهٔ ارزیابی نگهداشتهشدهٔ Golden6669 با شرایط پاک، دورمیدان و مسدود برای سنجش واقعگرایانهٔ ASR فارسی.
🧩 Role
evaluation and benchmarking asset
مصنوع ارزیابی و بنچمارک
📦 Snapshot
13 files; approximately 916.66 MB
13 فایل؛ حدود… See the full description on the dataset page: https://huggingface.co/datasets/Reza2kn/visualears-golden-6669.leyu-amharic-gonder-dialect
Leyu Amharic - Gonder Dialect Speech Corpus
Dataset Description
This dataset is a curated parallel speech corpus consisting of audio recordings paired with corresponding text transcripts, focused on the Gonnder dialect of the Amharic language. It is designed to support speech technology research across multiple tasks, including Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). The corpus captures dialect-specific phonetic variations, accent patterns… See the full description on the dataset page: https://huggingface.co/datasets/leyu-amharic/leyu-amharic-gonder-dialect.leyu-amharic-gojjam-dialect
Leyu Amharic - Gojjam Dialect Speech Corpus
Dataset Description
This dataset is a curated parallel speech corpus consisting of audio recordings paired with corresponding text transcripts, focused on the Gojjam dialect of the Amharic language. It is designed to support speech technology research across multiple tasks, including Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). The corpus captures dialect-specific phonetic variations, accent patterns, and… See the full description on the dataset page: https://huggingface.co/datasets/gheero-Leyu/leyu-amharic-gojjam-dialect.leyu-amharic-gojjam-dialect
Leyu Amharic - Gojjam Dialect Speech Corpus
Dataset Description
This dataset is a curated parallel speech corpus consisting of audio recordings paired with corresponding text transcripts, focused on the Gojjam dialect of the Amharic language. It is designed to support speech technology research across multiple tasks, including Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). The corpus captures dialect-specific phonetic variations, accent patterns, and… See the full description on the dataset page: https://huggingface.co/datasets/leyu-amharic/leyu-amharic-gojjam-dialect.goan-konkani-speech
Goan Konkani Speech (Romi)
47,365 audio clips, 108.5 hours of Goan Konkani (ISO 639-3 gom)
speech from Goan television news, transcribed in Romi Konkani - Konkani written in
the Roman script.
Konkani is a low-resource language with very little public speech data. This is
assembled from broadcast news, so it is real spoken Konkani: studio anchors, field
reporters, phone interviews, and the Konkani-English code-switching that Goan speakers
actually use.
Contents… See the full description on the dataset page: https://huggingface.co/datasets/Reubencf/goan-konkani-speech.red_ace_asr_error_detection_and_correction
RED-ACE
Dataset Summary
This dataset can be used to train and evaluate ASR Error Detection or Correction models. It was introduced in the RED-ACE paper (Gekhman et al, 2022).
The dataset contains ASR outputs on the LibriSpeech corpus (Panayotov et al., 2015) with annotated transcription errors.
Dataset Details
The LibriSpeech corpus was decoded using Google Cloud Speech-to-Text API, with the default and video models.
The word-level confidence was enabled… See the full description on the dataset page: https://huggingface.co/datasets/google/red_ace_asr_error_detection_and_correction.Taiwan-Tongues-ASR-CE-dataset-hakka
Taiwan-Tongues-ASR-CE-dataset-hakka
本資料集為 Taiwan-Tongues-ASR-CE 專案所使用的預訓練資料，透過 WebDataset 格式打包，並上傳至 Hugging Face 以便研究人員與開發者自由取用。
📂 Dataset 結構
本資料集分為 Training 與 Test 兩個子集，均以 WebDataset tar 檔案形式存放：
Training set (WebDataset format)
train/train-000000.tar
train/train-000001.tar
...
Test set (WebDataset format)
test/test-000000.tar
...
tsv set
train.tsv
test.tsv
...
每個 tar 內部均包含對應的音檔與標註，方便直接搭配 WebDataset 與 PyTorch / Hugging Face datasets 進行訓練與測試。
🏷️… See the full description on the dataset page: https://huggingface.co/datasets/adi-gov-tw/Taiwan-Tongues-ASR-CE-dataset-hakka.gospel-aloe-hera-v6
Gospel Aloe — Hera duplex training set (v6)
Everyday-conversation companion to
InternalCan/gospel-didactic-hera-v6.
Same codes-only Hera schema (32 Mimi codebooks, word-level timestamps, v6
B-channel corruption). Shards were packed on three nodes and published into
this one repo:
Source
Shard prefix
Role
This 8×H100 box + node A
local-*, nodeA-*
~3.1k conversations
Node B (63.141.33.128)
nodeB-*
~7.3k conversations
sample_id is the join key. There is no overlap… See the full description on the dataset page: https://huggingface.co/datasets/InternalCan/gospel-aloe-hera-v6.mlx-omni-lora-stt-tts-demoWill be used in the development of the trainer backend of mlx-omni by Neywa Labs.
arabic-audio-collection-sudanese-ahmed-gobara
Ahmed Gobara Sudanese Arabic Speech Dataset
Dataset Summary
The Ahmed Gobara Sudanese Arabic Speech Dataset is a single-speaker Sudanese Arabic speech corpus containing approximately 19 hours of speech recordings and corresponding transcripts.
While compact, the dataset offers a clean, consistent single-speaker resource in Sudanese Arabic — an Arabic variety with very few open speech resources — making it especially valuable for voice cloning, speaker adaptation… See the full description on the dataset page: https://huggingface.co/datasets/oddadmix/arabic-audio-collection-sudanese-ahmed-gobara.goodforft
goodforft
This is a merged speech dataset containing 863 audio segments from 4 source datasets.
Dataset Information
Total Segments: 863
Speakers: 4
Languages: en
Emotions: angry, happy, neutral
Original Datasets: 4
Dataset Structure
Each example contains:
audio: Audio file (WAV format, original sampling rate preserved)
text: Transcription of the audio
speaker_id: Unique speaker identifier (made unique across all merged datasets)
emotion: Detected emotion… See the full description on the dataset page: https://huggingface.co/datasets/Codyfederer/goodforft.golos_balalaika
GOLOS Annotated by Balalaika
[!IMPORTANT]
Official dataset for our INTERSPEECH 2026 paper
"A Data-Centric Framework for Addressing Phonetic and Prosodic Challenges in Russian Speech Generative Models" (arXiv:2507.13563).
Part of the Balalaika Russian speech data-processing pipeline — code: https://github.com/lab260ru/balalaika.
If you use this resource, please cite it.
A curated Russian speech dataset for advanced speech generative tasks.
Overview
GOLOS… See the full description on the dataset page: https://huggingface.co/datasets/lab260/golos_balalaika.khmer-speech-large-english-google-translations
Dataset Card for khmer-speech-large-english-google-translation
Audio recordings of khmer speech with varying speakers and background noises.
English transcriptions were transcribed from the Khmer labels using Google Translate.
Based off of seanghay/khmer-speech-large.
Dataset Details
Dataset Sources
Huggingface: seanghay/khmer-speech-large
Usage
from datasets import load_dataset
ds =… See the full description on the dataset page: https://huggingface.co/datasets/djsamseng/khmer-speech-large-english-google-translations.myanmar-speech-dataset-google-fleursPlease visit to the GitHub repository for other Myanmar Langauge datasets.
Myanmar Speech Dataset (Google Fleurs)
This dataset consists exclusively of Myanmar speech recordings, extracted from the larger multilingual Google Fleurs dataset.
For the complete multilingual dataset and additional information, please visit the original dataset repository
of Google Fleurs HuggingFace page.
Original Source
Fleurs is the speech version of the FLoRes machine translation benchmark.… See the full description on the dataset page: https://huggingface.co/datasets/chuuhtetnaing/myanmar-speech-dataset-google-fleurs.gol-dataset
GOL Dataset — audited metadata repair
This card is based on a read-only audit of revision
4b43743fbe5f84589130587c30d625a8b0d95694. The repository is gated and contains
visual-novel speech, but the source card does not identify the included titles, rights
holders, extraction procedure, or license. Access approval is not a license; maintainers
must document the lawful use and redistribution basis before downstream use.
Exact repository inventory
607 files totaling… See the full description on the dataset page: https://huggingface.co/datasets/midralab/gol-dataset.Taiwan-Tongues-ASR-CE-dataset-en
Taiwan-Tongues-ASR-CE-dataset-en
本資料集為 Taiwan-Tongues-ASR-CE 專案所使用的預訓練資料，透過 WebDataset 格式打包，並上傳至 Hugging Face 以便研究人員與開發者自由取用。
📂 Dataset 結構
本資料集分為 Training 與 Test 兩個子集，均以 WebDataset tar 檔案形式存放：
Training set (WebDataset format)
train/train-000000.tar
train/train-000001.tar
...
Test set (WebDataset format)
test/test-000000.tar
...
tsv set
train.tsv
test.tsv
...
每個 tar 內部均包含對應的音檔與標註，方便直接搭配 WebDataset 與 PyTorch / Hugging Face datasets 進行訓練與測試。
🏷️… See the full description on the dataset page: https://huggingface.co/datasets/adi-gov-tw/Taiwan-Tongues-ASR-CE-dataset-en.leyu-amharic-gonder-dialect
Leyu Amharic - Gonder Dialect Speech Corpus
Dataset Description
This dataset is a curated parallel speech corpus consisting of audio recordings paired with corresponding text transcripts, focused on the Gonnder dialect of the Amharic language. It is designed to support speech technology research across multiple tasks, including Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). The corpus captures dialect-specific phonetic variations, accent patterns… See the full description on the dataset page: https://huggingface.co/datasets/gheero-Leyu/leyu-amharic-gonder-dialect.visualears-benchmark-269-gold
🗂️ visualears-benchmark-269-gold
English + فارسی · Part of Shenava 1.0 · Project hub · SLT paper submission
🌟 At a glance | معرفی سریع
English
فارسی
🎯 Purpose
269-record gold/noisy benchmark dataset.
معیار طلایی ۲۶۹ نمونهای برای بررسی سریع خطاهای گفتار نویزی و مقایسهٔ نسخههای مدل.
🧩 Role
evaluation and benchmarking asset
مصنوع ارزیابی و بنچمارک
📦 Snapshot
276 files; approximately 44.35 MB
276 فایل؛ حدود 44.35 MB
🧱 Packaging
1 Parquet… See the full description on the dataset page: https://huggingface.co/datasets/Reza2kn/visualears-benchmark-269-gold.