Update README.md

Update README.md

Card: link the interactive demo and film

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

Update README.md

Upload 4 files

Update README.md

Delete voxturn.png

Add the LiveKit Agents turn detector (livekit_voxturn.py)

Upload voxturn.png

Update README.md

Add the Pipecat turn analyzer (pipecat_voxturn.py)

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

Model card: drop training and responsible-use sections

voxturn v1

initial commit