Unpacks transcription info instead of discarding it. Adds language and
language_probability fields to transcript events, and includes them in
verbose log output.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
--language: force language detection (e.g. en, sv) or leave unset for auto
--task: transcribe (default) or translate to English
Previously language was hardcoded to 'en' which caused multilingual models
to hallucinate translations instead of transcribing the source language.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Checks cache first with local_files_only=True; if the model isn't present
logs "downloading model ..." to stderr before WhisperModel triggers the
actual download.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
listen.mjs: prints all events as JSON objects.
transcripts.mjs: prints transcript text only.
Both use Node 21+ built-in WebSocket — no libraries required.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Errors always go to stderr. Info logs (startup, VAD events, transcripts)
only appear with --verbose / -v, keeping stderr clean when running as a
system service.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Every connection receives the full event stream (vad_start, vad_end,
transcript, error) from the moment it connects — no subscription
handshake required. The asyncio WebSocket server runs in a daemon thread
alongside the VAD loop and transcription thread. Events still go to
stdout unchanged.
Port is configurable via STT_PORT env var (default: 11501).
Add websockets to both setup scripts.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replaces the old stdin/stdout transcription-only server. Now handles the
full pipeline in Python:
- Launches parec or arecord for mic capture
- Runs Silero VAD (via silero-vad, already a faster-whisper dep — no sherpa-onnx needed)
- Pre-roll ring buffer (0.2s) prepended to each segment for context
- Transcribes with faster-whisper in a separate thread (GPU not blocking VAD)
- Emits JSON line events to stdout: ready, vad_start, vad_end, transcript, error
Event protocol is designed to map directly to WebSocket subscriptions later.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
--no-deps skipped av and other required packages. Fix by installing
faster-whisper normally first (satisfies all deps, pulls PyPI ctranslate2),
then immediately overriding ctranslate2 with the source-built version.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
setup-venv.sh: simple PyPI install path — just pip install faster-whisper.
Use this when the PyPI ctranslate2 wheel matches the system CUDA version.
setup-venv-local-build.sh:
- PYTHON_ENV env var for venv path override (consistent with tts-server)
- HF_TOKEN_FILE env var instead of hardcoded path
- HF_HUB_CACHE env var surfaced in output when set
- Remove stray chmod on faster-whisper-server.py (not part of this repo)
- Remove voice-experiment-specific "run with" message
- Add python3 to tool prerequisite check
- Arch Linux package suggestions extended to cover CUDA and python
- Document why each script exists and when to use which
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
