openai/whisper

Convert speech in audio to text

Public
141.2M runs
License
GitHub
Weights
Paper
Playground API Examples README Versions
  1. Author
    @meatballhat
    Version
    cuda11.8-python3.11-torch2.0.1-X64
    Commit
    f709e2af25ccc5fa41e2bb00e0fc794b9633a5cb

    80996966

    Latest
  2. Author
    @daanelson
    Version
    22.04
    Commit
    46ebdca0a2356430a157fc396cfb026d09968ab2

    cdd97b25

  3. Author
    @daanelson
    Version
    22.04
    Commit
    46ebdca0a2356430a157fc396cfb026d09968ab2

    be69de6b

    cdd97b25

    This version has been automatically upgraded to a newer version.

  4. Author
    @zsxkib
    Version
    cuda11.8.0-python3.11-torch2.0.1-X64
    Commit
    4e3625e9647cad0b3a26509d3a25e3cfee4872d1

    3c08daf4

    cdd97b25

    This version has been automatically upgraded to a newer version.

  5. Author
    @zsxkib
    Version
    cuda11.8.0-python3.11-torch2.0.1-X64
    Commit
    aad0f3fc05ee4332c320dd5db15c9f181e509c19

    c7a9a6d5

    cdd97b25

    This version has been automatically upgraded to a newer version.

  6. Author
    @chenxwh
    Version
    22.04
    Commit
    849c2c7f5c2197a829d4aa3f03ecdd28aca6fd0f

    4d507972

    cdd97b25

    This version has been automatically upgraded to a newer version.

    Release notes

    • This version also only contains the large-v3 Whisper model.
  7. Author
    @daanelson
    Commit
    631cc6d3d92ba1c721e632c8a447ed19f7015d79

    91ee9c0c

    Release notes

    • This version also only contains the large-v2 Whisper model.
    • Contains slight modifications to previous version to improve initial setup time

  8. e39e3547

    Release notes

    • This version only contains the large-v2 Whisper model.
    • Use previous versions to make predictions with all other Whisper models.
    • 10% faster inference.
    • Should have less latency variance than previous Whisper versions

  9. 30414ee7

    cdd97b25

    This version has been automatically upgraded to a newer version.

    Release notes

    • Hosts ‘tiny’, ‘base’, ‘small’, ‘medium’ and ‘large-v1’ Whisper models.

  10. 089ea17a

  11. 34a6b005

  12. b6e7ea7a

  13. 770db509