0.2.15

QuentinFuxa/WhisperLiveKit0.2.15Nov 21, 2025by QuentinFuxa

AI Summary

Performance improvements for VAD cuts and audio buffering, plus core evolutions including LoRa extension loader and CoreML encoder compatibility.

Key Highlights

  • Improve VAD cuts position
  • Improve audio buffering management
  • Improve STT/Diarization & Translation alignment with NLLW 0.1.3
  • LoRa extension loader in shared Whisper core
  • Simulstreaming CoreML encoder compatibility

New Features

  • LoRa extension loader for multi-language model support
  • CoreML encoder compatibility for SimulStreaming

Full Release Notes

## Performance improvements:
- Improve VAD cuts position
- Improve audio buffering gestion
- Improve STT/Diarization & Translation alignment. For translation, use [NLLW](https://github.com/QuentinFuxa/NoLanguageLeftWaiting) 0.1.3

## Core evolutions:
- LoRa extension loader in the shared Whisper core - For future evolution on multi-language model
- Simulstreaming CoreML encoder compatibility

## Bug Fixes:
- Fixes #278 and #279