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

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
DatasetATH-MaaS/CSEMOTIONS

Emotional studio speech in Chinese

Audio recordings from 10 speakers labeled with 7 emotions and texts for Mandarin speech synthesis.

  • Text → speech
  • Chinese
3.5K 391K–10K 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
Datasetvikhyatk/lofi

7,000+ Hours of Generated Lofi Music

An audio corpus of 7,000+ hours of lofi music with text prompts, generated by MusicGen Large.

    3.1K 88100K–1M 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
    Datasetai4bharat/Svarah

    Indian-accented English speech for ASR

    Audio benchmark with transcripts of Indian English speakers for speech recognition evaluation.

      3.1K 611K–10K rows🔒 on request
      Datasetkensho/spgispeech

      Large English corpus for speech recognition (ASR)

      Millions of English audio and text pairs for training and testing automatic speech recognition systems.

      • Automatic speech recognition
      • English
      3.1K 421M–10M rows🔒 on request
      Datasetthivux/phoaudiobook

      Large Vietnamese speech corpus for zero-shot TTS

      Millions of Vietnamese audio-text pairs for training text-to-speech models without speaker fine-tuning.

      • Text → speech
      • Vietnamese
      3.1K 571M–10M rows🔒 on request
      Datasetlinagora/SUMM-RE

      French meeting transcripts with audio

      A collection of French conversational meeting transcripts aligned with audio signals.

      • Automatic speech recognition
      • Voice activity detection
      • French
      2.8K 7up to 1K rows
      DatasetMohamedRashad/MASC-Arabic

      MASC Arabic speech dataset

      1,000 hours of Arabic speech across various dialects, collected from YouTube.

      • Automatic speech recognition
      • Text → speech
      • Arabic
      2.7K 12100K–1M rows
      DatasetSPRINGLab/IndicTTS-Hindi

      Hindi voiceover for speech synthesis (TTS)

      Studio audio recordings of male and female Hindi speech with transcripts for voice models.

      • Text → speech
      • Hindi
      2.7K 3710K–100K rows
      Datasetsarulab-speech/commonvoice22_sidonTop pick

      Cleaned Mozilla Common Voice 22 at 48 kHz

      Tens of millions of speech audio recordings in 137 languages, denoised and enhanced by the Sidon model.

      • Text → speech
      • Abkhazian
      • Afrikaans
      • Amharic
      • +9
      2.7K 3210M–100M rows🔒 on request
      Datasetnuriachandra/Deepfake-Eval-2024

      2024 real deepfake multimodal benchmark

      Manually annotated video, audio, and images for evaluating deepfake detectors in real-world conditions.

        2.7K 691K–10K rows🔒 on request
        DatasetMohammadJRanjbar/ParsVoiceTop pick

        Multi-speaker Persian speech corpus for TTS and ASR

        Largest open audio-text corpus in Persian, collected from audiobooks for speech synthesis and recognition.

        • Text → speech
        • Automatic speech recognition
        • Audio classification
        • Persian
        2.6K 451M–10M rows🔒 on request
        DatasetARTPARK-IISc/Vaani-Atypical-Speech-Corpus

        Vaani atypical speech corpus

        Google Euphonia project data corpus to improve ASR systems for people with atypical speech.

        • Automatic speech recognition
        • Hindi
        • Marathi
        • Telugu
        2.4K 41K–10K rows🔒 on request
        Datasetnaijavoices/naijavoices-dataset

        1,800-hour Nigerian live speech audio corpus

        Large dataset with speech recordings from over 5,000 speakers and text transcripts exceeding 500 GB.

          2.4K 381M–10M rows🔒 on request
          DatasetMushanW/GLOBE_V3

          Cleaned English audio corpus based on Common Voice 21.0

          Filtered Common Voice recordings with 24 kHz sample rate and perfect transcription match by Whisper V3 Large.

          • Text → audio
          • Automatic speech recognition
          • Audio → audio
          • English
          2.4K 4100K–1M rows
          Datasetjoujiboi/japanese-anime-speech-v2

          Japanese speech from visual novels

          292,637 audio clips with Japanese transcriptions from visual novels for ASR.

          • Automatic speech recognition
          • Japanese
          2.3K 167100K–1M rows