fix: add voice status endpoint stub and guard VRAM division by zero#851

Open

yasinBursali wants to merge 2 commits intoLight-Heart-Labs:mainfrom

yasinBursali:fix/dashboard-voice-vram

Contributor

yasinBursali commented Apr 8, 2026

What

Add /api/voice/status stub endpoint that checks whisper, tts, and livekit health
Guard VRAM division by zero in Dashboard.jsx and Models.jsx

Why

Voice.jsx calls /api/voice/status on mount and every 30s — gets 404 (no voice router exists). Page permanently shows "Some voice services unavailable" regardless of actual service health.
Dashboard.jsx and Models.jsx divide by vramTotal without checking for zero. When GPU detection fails or returns 0 VRAM, produces NaN/Infinity breaking progress bar rendering.

How

New routers/voice.py: aggregates health of whisper (STT), tts, and livekit via existing check_service_health() infrastructure. Returns {available, services: {stt, tts, livekit}, message} matching Voice.jsx expectations. Registered in main.py.
Dashboard.jsx line 266: vramTotal > 0 ? ... : 0 guard
Models.jsx lines 72-75: same guard on all 3 division sites
Pattern matches existing GPUCard.jsx guard

Testing

Python syntax verified
Live tested: voice endpoint returns correct health status for whisper/tts
Dashboard.jsx VRAM division confirmed guarded

Platform Impact

All platforms: Identical — Python + React changes, no platform-specific code

yasinBursali and others added 2 commits

April 8, 2026 05:03


          fix: add voice status stub endpoint for dashboard Voice page

29a5265

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>


          fix: guard VRAM division by zero in Dashboard and Models pages

783ca59

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet