What's New
- Qwen3-ASR int8 Support (#312): Add int8 quantized variant for Qwen3-ASR, reducing model size
- ITN Post-Processor (#308): Inverse text normalization with NLTagger context spotting
- clearAllModelCaches() (#309): New API to clear all cached models
- Directory Override (#327): Added directory override for model storage
- Task Cancellation (#340): Add Task cancellation support to ASR transcription
Fixes
- Fix ASR MLMultiArray cache lifecycle in chunked transcription (#321)
- Fix Float16 build failure in Qwen3-ASR for Archive builds (#305)
- Replace force-unwraps with guard in tdtDecodeWithTimings (#324)
- Add type annotation to withUnsafeBytes for Swift 6.2 compatibility (#341)
- Use 2-model pipeline for Qwen3-ASR download validation (#307)
Other
- Rename FluidAudioTTS to FluidAudioEspeak (#302)
- Add PocketTTS smoke test CI workflow (#313)
- Showcase additions: Flowstay, Snaply, macos-speech-server
Full Changelog: v0.12.1...v0.12.2