root a9191cee00 feat(workers): free Ollama VRAM before loading Whisper on the GPU
Whisper (CT 311) and Ollama (CT 102) share one A2000. Before loading
Whisper on CUDA, ask Ollama to unload its models (GET /api/ps then POST
/api/generate keep_alive:0) and wait for the card to clear, so the GPU
load has headroom. Best-effort and stdlib-only; Ollama reloads
cooperatively, and the existing CUDA->CPU fallback covers any failure.
Toggle via OLLAMA_FREE_BEFORE_STT; endpoint via OLLAMA_URL.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-05 21:12:05 +10:00
2026-05-31 01:22:10 +10:00
2026-05-31 01:22:10 +10:00

Void 2.0

Homelab orchestrator + canonical knowledge store. Cradle-themed. Successor to Void 1.x (CT 301). Spec at /project/docs/superpowers/specs/2026-05-31-void-v2-design.md.

Layout

  • void-server (this repo) — Node API, MCP, UI, cron, agent runtime
  • void-workers — Python ingest workers (separate repo, later plan)

Quick start (dev)

  1. Provision void2-db LXC (see deploy/README.md)
  2. Install Postgres + pgvector on void2-db
  3. npm install
  4. cp .env.example .env and edit
  5. npm run migrate
  6. npm start
  7. curl -H "Authorization: Bearer $OWNER_TOKEN" http://localhost:3000/health
Description
No description provided
Readme 2.6 MiB
Languages
JavaScript 87.5%
CSS 6.4%
Python 5.2%
Shell 0.7%
HTML 0.2%