- .env: FASTER_WHISPER_DEVICE=cpu gesetzt; llama.cpp belegt GPU 1 komplett, faster-whisper (small-Modell) laeuft problemlos auf CPU (~1-2s Latenz) - ws.py: "empty transcript" -> "Keine Sprache erkannt — bitte erneut sprechen." Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| __init__.py | ||
| admin.py | ||
| chat.py | ||
| config.py | ||
| devices.py | ||
| health.py | ||
| me.py | ||
| metrics.py | ||
| sessions.py | ||
| speak.py | ||
| transcribe.py | ||
| ws.py | ||