datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
X-Voice-Dataset-Train
X-Voice Training Dataset
Overview
The X-Voice training dataset is a large-scale multilingual speech corpus curated for high-performance speech models. It provides a robust foundation for cross-lingual phonetic and prosodic modeling.
Also the train set of X-Voice Model.
Core Statistics
Total Speech Duration: 420K hours
30 languages
European: bg (Bulgarian), cs (Czech), da (Danish), de (German), el (Greek), en (English), es (Spanish), et (Estonian), fi… See the full description on the dataset page: https://huggingface.co/datasets/XRXRX/X-Voice-Dataset-Train.knesset-plenums-whisper-training
Dataset Card for ivrit.ai - Knesset Plenums Whisper Training
This is a whisper-formatted version of the ivrit.ai Knesset Plenums dataset.
This dataset was created by splitting long audio recordings, along with their respective transcriptions, into audio slices of 30 seconds or less.
Each such slice represents one or more consecutive segments, along with timestamp token data and the previous slice's transcription.
The code for this dataset preparation process is available on the… See the full description on the dataset page: https://huggingface.co/datasets/ivrit-ai/knesset-plenums-whisper-training.lao_stt_training_data
Lao Speech-to-Text Training Data
ຊຸດຂໍ້ມູນນີ້ຖືກຈັດກຽມຂຶ້ນມາເພື່ອໃຊ້ສຳລັບການເທຣນ ແລະ ປັບແຕ່ງ (Fine-tuning) ໂມເດວ Speech-to-Text (ເຊັ່ນ OpenAI Whisper) ສຳລັບພາສາລາວ.
ໂຄງສ້າງຂອງຂໍ້ມູນ (Dataset Structure)
Train set: ໄຟລ໌ສຽງຢູ່ໃນໂຟນເດີ train/ ແລະ ມີການ Mapping ຂໍ້ຄວາມໃນ train.csv
Validation set: ໄຟລ໌ສຽງຢູ່ໃນໂຟນເດີ validation/ ແລະ ມີການ Mapping ຂໍ້ຄວາມໃນ validation.csv
ຮູບແບບຂໍ້ມູນໃນໄຟລ໌ CSV:
audio: ເສັ້ນທາງໄປຫາໄຟລ໌ສຽງ (e.g., train/audio25000.wav)… See the full description on the dataset page: https://huggingface.co/datasets/KitTzk/lao_stt_training_data.Tech-Sentences-For-ASR-Training
TechVoice Dataset
Work in Progress – This dataset is actively being expanded with new recordings.
Dataset Statistics
Metric
Current
Target
Progress
Duration
38m 43s
5h 0m 0s
██░░░░░░░░░░░░░░░░░░ 12.9%
Words
10,412
50,000
████░░░░░░░░░░░░░░░░ 20.8%
Total Recordings: 205 samples
Total Characters: 74,312
A specialized speech dataset for fine-tuning Automatic Speech Recognition (ASR) models on technical and developer vocabulary. Contains human-recorded… See the full description on the dataset page: https://huggingface.co/datasets/danielrosehill/Tech-Sentences-For-ASR-Training.crowd-recital-whisper-training
Dataset Card for ivrit.ai - Crowd Recital
Dataset Details
Dataset Description
License
The dataset is released under the ivrit.ai License, which enables broad research and commercial use.
- Full license: https://www.ivrit.ai/en/the-license/
- FAQs: https://www.ivrit.ai/en/license-faqs/
Dataset Structure
Data Fields
Each example in the dataset contains:
audio: An audio column containing:
bytes: The audio data… See the full description on the dataset page: https://huggingface.co/datasets/ivrit-ai/crowd-recital-whisper-training.whisper-large-training
Khmer Speech Dataset for Whisper Large V3
Combined Khmer speech datasets for fine-tuning Whisper models.
Statistics
Total: 52,255 samples (~47.3 hours)
Train: 49,119 (94.0%)
Test: 3,136 (6.0%)
Duration: 1.0s - 13.3s (avg: 3.3s)
Sources
seanghay/khmer_mpwt_speech (×5)
Samples: 10,290 (duplicated 5x from 2,058)
Duration: ~9.6 hours
seanghay/km-speech-corpus
Samples: 14,943
Duration: ~10.3 hours
google/fleurs
Samples: 1,675… See the full description on the dataset page: https://huggingface.co/datasets/Tnaot/whisper-large-training.tts-training-dataset
Human Reviewed Telugu-English TTS Dataset
A manually reviewed multilingual TTS dataset created from publicly available educational and speech content.
Dataset Splits & Distribution Metrics
balanced_60min Split
Total Segments: 120
Total Duration: 60.00 minutes
Unique Speakers: 3
Distribution Breakdowns:
Language Distribution:
en-IN: 60 segments (30.00 minutes)
te-IN: 60 segments (30.00 minutes)
Style Distribution:
analytical: 27… See the full description on the dataset page: https://huggingface.co/datasets/nlpctx/tts-training-dataset.crowd-recital-yi-whisper-training
Dataset Card for ivrit.ai - Crowd Recital - Yiddish
See more details on the source dataset card.
Dataset Details
Dataset Description
This is a derived dataset for structured for whisper training:
Excludes low quality segments (judged by probabilities of the text-audio auto alignment process)
Encodes timestamps along segments of text + previous text
Audio encoded to 16K sample-rate, mono
Total audio duration - ~78h
License: other
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/ivrit-ai/crowd-recital-yi-whisper-training.crowd-whatsapp-yi-whisper-training
Dataset Card for ivrit.ai - Crowd Whatsapp - Yiddish
See more details on the source dataset card.
Dataset Details
Dataset Description
This is a derived dataset for structured for whisper training:
Excludes low quality segments (judged by probabilities of the text-audio auto alignment process)
Encodes timestamps along segments of text + previous text
Audio encoded to 16K sample-rate, mono
Total audio duration - ~19h
License: other
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/ivrit-ai/crowd-whatsapp-yi-whisper-training.isizulu-asr-train
isiZulu Speech Recognition Augmented Train Dataset
Dataset Description
This dataset contains augmented speech recordings and transcriptions for isiZulu, one of South Africa's official languages.
The dataset has been optimized for use with OpenAI's Whisper ASR models.
Dataset Statistics
Number of samples: 690
Language: isiZulu (Zul)
Audio format: WAV, 16kHz, mono, 16-bit
Maximum duration: 30 seconds (truncated for Whisper compatibility)
Transcription format:… See the full description on the dataset page: https://huggingface.co/datasets/zionia/isizulu-asr-train.trainSetisixhosa-asr-train
isiXhosa Speech Recognition Augmented Dataset
Dataset Description
This dataset contains augmented speech recordings and transcriptions for isiXhosa, one of South Africa's official languages.
The dataset has been optimized for use with OpenAI's Whisper ASR models.
Dataset Statistics
Number of samples: 680
Language: isiXhosa (Xho)
Audio format: WAV, 16kHz, mono, 16-bit
Maximum duration: 30 seconds (truncated for Whisper compatibility)
Transcription format:… See the full description on the dataset page: https://huggingface.co/datasets/zionia/isixhosa-asr-train.kazakh_speech_corpus_2
Kazakh_speech_dataset_2
This dataset contains Kazakh_speech_dataset_2 from ISSAI but in parquet format.
Dataset info
645,860 Utterances
1194 Hours in total
Sources in each split:
test : {'tv_news', 'crowdsourced', 'radio', 'talkshow', 'parliament', 'tts', 'podcasts'}
train : {'tv_news', 'crowdsourced', 'radio', 'talkshow', 'parliament', 'tts', 'podcasts'}
validation : {'tv_news', 'crowdsourced', 'radio', 'talkshow', 'parliament', 'tts','podcasts'}
Guides… See the full description on the dataset page: https://huggingface.co/datasets/SRP-base-model-training/kazakh_speech_corpus_2.kazakh_speech_dataset_ksdKazakh Speech Dataset cleaned, converted to parquet and with uppercase_transcription made with gpt4o_api.
Dataset info:
813 Speakers
with 500 samples for 4 speakers
with 250 samples for 809 speakers
Male/female
555 Hours
Guides
Load data 1
Replace the export HF_HOME with your HF_HOME path
from datasets import load_dataset
# export HF_HOME="/data/vladimir_albrekht/hf_cache"
ds = load_dataset("SRP-base-model-training/kazakh_speech_dataset_ksd") # split ='test' or… See the full description on the dataset page: https://huggingface.co/datasets/SRP-base-model-training/kazakh_speech_dataset_ksd.TrainingSpeechTrainingSpeech is an initiative to provide open and freely reusable dataset of voices
for speech-to-text models training
on non-english languages
using already available data (such as audio-books).
Right now, data are extracted exclusively from audio-books and in French language. Let me know if you are intersted to contribute by creating an issue.
Tooling
TrainingSpeech comes with a CLI that automate and simplify:
transcript extraction
forced-alignment (using aeneas)… See the full description on the dataset page: https://huggingface.co/datasets/wasertech/TrainingSpeech.Meta_STT_MADASR2.0_train_lg
Dataset Card for Dataset Name
https://sites.google.com/view/respinasrchallenge2025/home?authuser=0
This is the trainig dataset provided for track-3 and track-4 of this task. We enhance the dataset with entity tagging, emotion, age, gender and intent.
WhissleAI participated in this challenge.
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
Dataset Description
Curated… See the full description on the dataset page: https://huggingface.co/datasets/WhissleAI/Meta_STT_MADASR2.0_train_lg.bocalantics-wol-train
Bocalantics Wolof eval set
The held-out Wolof test rows from MOH749/Bocalantics-2.0, with audio attached, so a
candidate model can be scored without re-materialising anything.
43,034 rows, 65.33 hours.
Why it exists
"Beats unadapted Whisper" is not a result for Wolof. Whisper has seen 2 of the parent
corpus's 26 languages, so beating it is arithmetic rather than evidence. The bar is the
best published model for the language -… See the full description on the dataset page: https://huggingface.co/datasets/MOH749/bocalantics-wol-train.medimind-r11-train
MediMind R11 — ASR training data
Unified manifest + packed audio for fine-tuning Whisper-large-v3 on Norwegian
clinical and conversational speech.
Training manifest: r11_manifest.jsonl — 11,022 packs
Held-out eval set: r11_heldout_eval.jsonl — 291 packs (NEVER train on these)
~see manifest audit packs total
11 sources: lege_*, podcasts (motiv/podk/stet), nb_samtale, nb_tale_m3, tts_drugs
Schema
See r11_manifest.jsonl (one JSON object per line) and… See the full description on the dataset page: https://huggingface.co/datasets/gallip0li/medimind-r11-train.