Hugging Face's logo

AustinXiao
/
voxturn

Audio Classification
turn-detection
end-of-turn
turn-taking
voice-agent
endpointing
semantic-vad
speech
Eval Results (legacy)
Model card Files
xet
 Community
voxturn

Commit History

7 days ago

  1. Model card: link the technical report, cite it, align numbers with the report

    AustinXiao
    Verified signature
    4a1e8da

  2. Model card: TurnBench test results (EOT #5, INT #6), comparison with VAP and Smart Turn v3

    AustinXiao
    Verified signature
    5776ab2

9 days ago

  1. Update README.md

    AustinXiao
    Verified signature
    b41ba81

  2. Update README.md

    AustinXiao
    Verified signature
    c2bd940

  3. Card: link the interactive demo and film

    AustinXiao
    Verified signature
    3b6fbe8

  4. Run on CPU: fp32 model and encoder input in the weights' dtype

    AustinXiao
    Verified signature
    f687aa2

  5. Update README.md

    AustinXiao
    Verified signature
    d234ab1

  6. Upload 4 files

    AustinXiao
    Verified signature
    ffff911

  7. Update README.md

    AustinXiao
    Verified signature
    1192531

  8. Delete voxturn.png

    AustinXiao
    Verified signature
    7ad6886

  9. Add the LiveKit Agents turn detector (livekit_voxturn.py)

    AustinXiao
    Verified signature
    33bb672

  10. Upload voxturn.png

    AustinXiao
    Verified signature
    1ff2ca6

  11. Update README.md

    AustinXiao
    Verified signature
    982acb4

  12. Add the Pipecat turn analyzer (pipecat_voxturn.py)

    AustinXiao
    Verified signature
    444809b

  13. Fast inference: merged adapter, fixed-window encoder, CUDA graphs (8 ms per decision)

    AustinXiao
    Verified signature
    7de23f3

  14. Model card: drop training and responsible-use sections

    AustinXiao
    Verified signature
    9278195

  15. voxturn v1

    AustinXiao
    Verified signature
    9e5b939

  16. initial commit

    AustinXiao
    Verified signature
    2ad9aa7