[RNE Rewrite] feat: add Whisper STT pipeline#1302

Open

barhanc wants to merge 18 commits into

rne-rewritefrom

Commits on Jul 20, 2026

feat(speech): wire VAD to Whisper STT and rewrite example app screen
barhanc
committed
fix(stt): remove unused sttProgress and sttError variables
barhanc
committed
docs: add JSDoc to Whisper STT task pipeline and hook
barhanc
committed
fix: add missing @category and localPath to useSpeechToText
barhanc
committed
fix(stt): adapt to VAD API changes from rne-rewrite rebase
barhanc
committed
feat(speech-example): add scrollable ModelPicker to STT screen
barhanc
committed
feat(stt): add all Whisper models, disable CoreML on non-iOS platforms
barhanc
committed
docs(stt): replace BCP-47 references with Whisper-specific language code terminology
barhanc
committed
fix(whisper): correct STRIDE_SIZE comment to reflect stream-only usage
barhanc
committed
fix(whisper): validate language against supportedLanguages before decoding
barhanc
committed
fix(whisper): validate language token against tokenizer, not declared metadata
barhanc
committed
fix(speech-app): render download progress status on STT screen
barhanc
committed
small fix
barhanc
committed
refactor(speech): rename fsmnVadModel to generic vadModel in WhisperSttModel public API
barhanc
committed
feat(example-stt): add separate Transcribe Audio File screen
barhanc
committed
feat(speech-app): wrap STT screens in ScreenWrapper and add error logging
barhanc
committed

Commits on Jul 21, 2026

Rename STT example app directories to be more descriptive
barhanc
committed
Point Whisper model URLs to HuggingFace repos and add MLX backend support

Bartosz Hanc
authored and
Bartosz Hanc
committed