Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.

Data type

For which models

Find a dataset for your goal

Answered by Gemini 4 · the chat is kept only in this tab

DatasetCnam-LMSSC/vibravox

French speech recorded with bone conduction sensors

Audio dataset of French speech captured by body conduction sensors for recognition, synthesis, and audio enhancement tasks.

  • Audio → audio
  • Automatic speech recognition
  • Audio classification
  • French
12.4K 3110K–100K rows
Datasetspeechcolab/gigaspeech2Top pick

ASR corpus for Thai, Indonesian, and Vietnamese

30k hours of auto-transcribed speech and a cleaned 22k-hour set for ASR in three languages.

  • Automatic speech recognition
  • Thai
  • Indonesian
  • Vietnamese
11.6K 7710M–100M rows🔒 on request
DatasetCanCLID/zoengjyutgaai

Cantonese audiobooks and transcriptions

Recordings by famous storyteller Zhang Yuekai with verified texts for Cantonese TTS and ASR.

  • Automatic speech recognition
  • Text → speech
  • Text generation
  • Cantonese
10.5K 35100K–1M rows
DatasetMohammed01/ArFakeTop pick

Benchmark for Arabic speech deepfake detection

First multi-dialect dataset of synthetic and real Arabic speech for detecting audio fakes.

  • Audio classification
  • Automatic speech recognition
  • Text → audio
  • Arabic
10.2K 910K–100K rows🔒 on request
DatasetDigital-Divide-Data/khmer-speech-dataset

Khmer language speech corpus on Cambodian culture

Nearly 728 hours of Khmer audio recordings with manual transcription for speech recognition.

  • Automatic speech recognition
  • Text classification
  • Khmer
9.7K 38100K–1M rows🔒 on request
Datasetfacebook/omnilingual-asr-corpusTop pick

Meta ASR corpus in 348 rare languages

Spontaneous speech recordings and transcriptions for 348 languages for ASR training and language identification.

  • Automatic speech recognition
  • Audio classification
  • aae
  • aal
  • aao
  • +9
9.5K 222100K–1M rows
DatasetOmniAICreator/ASMR-Archive-Processed

Japanese ASMR archive for ASR and TTS

Japanese ASMR audio with images for speech recognition and synthesis; updates are paused.

  • Automatic speech recognition
  • Text → speech
  • Japanese
9.4K 98
DatasetTheAgenticDataCompany/open-yap-1kTop pick

Open Yap 1K: 1000 Hours of Dialogues

1000 hours of dual-channel English conversations at 48 kHz for full-duplex and speech models.

  • Audio → audio
  • Automatic speech recognition
  • Text → speech
  • English
8.3K 269up to 1K rows
DatasetFBK-MT/moselTop pick

EU multilingual speech corpus

Up to 950K hours of open speech across 24 official EU languages with automatic transcripts.

  • Automatic speech recognition
  • Text → speech
  • English
  • Bulgarian
  • Croatian
  • +9
8K 941M–10M rows
Datasetapptek-com/apptek_callcenter_dialogues

Multi-accent call center recordings

128.6 hours of long-form English dialogues for training and evaluating ASR systems.

  • Automatic speech recognition
  • English
7.6K 431K–10K rows
Datasetedinburghcstr/edacc

English dialogues with diverse accents

40 hours of paired English conversations with diverse accents for ASR training and evaluation.

  • Automatic speech recognition
  • Audio classification
6.6K 1810K–100K rows
Datasetlangswap/dialogs-ru-emotional-conversationsTop pick

Studio emotional dialogues in Russian

20.6 hours of studio recordings of live theatrical actor dialogues for emotional TTS and ASR.

  • Text → speech
  • Automatic speech recognition
  • Audio classification
  • Russian
5.6K 27up to 1K rows
Datasethanamizuki-ai/genshin-voice-v3.3-mandarin

Genshin Impact character voice acting in Chinese

Paired audio recordings and character voice lines from Genshin Impact in Mandarin for speech synthesis and recognition.

  • Text → speech
  • Automatic speech recognition
  • Chinese
4.5K 4110K–100K rows
Datasetjoujiboi/Galgame-VisualNovel-Reupload

Voice acting lines from Japanese visual novels

A multi-million corpus of Japanese game voice acting and texts repacked for easy downloading.

  • Automatic speech recognition
  • Text → speech
  • Japanese
4.3K 421M–10M rows
Datasetparler-tts/libritts_r_filtered

Filtered 585-hour English speech corpus

A filtered version of the LibriTTS-R audio dataset with improved sound quality for speech synthesis and recognition.

  • Text → speech
  • Automatic speech recognition
  • English
4.3K 26100K–1M rows
DatasetCAiRE/ASCEND

Spontaneous Chinese-English dialogue speech

Corpus of audio recordings with transcriptions for speech recognition with code-switching between Chinese and English.

  • Automatic speech recognition
  • English
  • Chinese
4.2K 5710K–100K rows
Datasetoddadmix/dialectal-arabic-lahgtna-v2Top pick

3000+ hours of speech across 13 Arabic dialects for ASR

A large corpus of audio recordings and transcripts in 13 Arabic dialects for training speech recognition systems.

  • Automatic speech recognition
  • Arabic
  • English
4.1K 33100K–1M rows
DatasetFBK-MT/Speech-MASSIVE

Multilingual spoken language understanding corpus

Spoken version of the MASSIVE text corpus for ASR, intent detection, and slot filling tasks.

  • Audio classification
  • Text classification
  • Zero shot classification
  • Arabic
  • German
  • Spanish
  • +9
3.6K 5710K–100K rows
DatasetMushanW/GLOBE

English speech corpus with diverse global accents

From hundreds of thousands to one million English speech audio recordings with global accents and quality scores.

  • Text → audio
  • Automatic speech recognition
  • Audio → audio
  • English
3.5K 32100K–1M rows
DatasetFBK-MT/MCIFTop pick

Multilingual multimodal benchmark on scientific reports

Human-annotated instruction dataset based on scientific reports: speech, video, and text across four languages to evaluate MLLMs.

  • Automatic speech recognition
  • Question answering
  • Summarization
  • English
  • German
  • Italian
  • +1
3.2K 711K–10K rows
DatasetXRailgunX/Aemeath-Audio-Dataset

Voice lines of character Aemeath from Wuthering Waves

450 Chinese voice lines of a single character with transcripts and emotion labels for TTS training.

  • Text → speech
  • Automatic speech recognition
  • Audio classification
  • Chinese
3.1K 3up to 1K rows