services/ai-service/README_RUNNING.md

Running the AI Service locally (Docker)

Prerequisites:

• Docker Desktop (or Docker Engine) installed and running
• Optional: Docker Compose v2 (most Docker Desktop include it)

Build and run with docker-compose:

PowerShell:

Set-Location "D:\dartdev\glitz\git\hntaiclone\HNTAI\services\ai-service"
# Build and run
.\run_local.ps1 -Build
# Follow logs
docker compose logs -f

Or run without rebuilding (uses existing image):

Set-Location "D:\dartdev\glitz\git\hntaiclone\HNTAI\services\ai-service"
.\run_local.ps1

Verify endpoints after a short startup delay:

curl http://localhost:7860/live
curl http://localhost:7860/ready
curl http://localhost:7860/metrics_text

Notes:

• If you want GPU support or OpenVINO system packages, adjust Dockerfile.prod accordingly.
• To persist model caches between runs, the compose file mounts ./model_cache.