Skip to content

Platform setup (macOS, Windows, Linux, Proxmox)

The image runs anywhere Docker runs. Pick your platform, install Docker, then continue with the Quick start.

macOS

  1. Install Docker Desktop for Mac (Apple Silicon or Intel build) or OrbStack as a lighter alternative.
  2. Settings → Resources: at least 6 GB memory (8 GB+ recommended) and 4+ CPUs.
  3. Apple Silicon pulls the native linux/arm64 image automatically.

Apple GPU (MPS)

Containers on macOS run inside a Linux VM, which has no access to the Apple GPU. Inside Docker Jeff always runs on CPU. If you need MPS, run upstream Jeff natively with uv (see the upstream README).

Windows

  1. Enable WSL2: wsl --install in an elevated PowerShell, then reboot.
  2. Install Docker Desktop for Windows with the WSL2 backend.

  3. Limit/raise WSL2 memory in %UserProfile%\.wslconfig:

    [wsl2]
memory=10GB
processors=8

    then wsl --shutdown and restart Docker Desktop. 4. Run commands from PowerShell or, better, from a WSL Ubuntu shell (the bash snippets in this documentation then work as-is).

Linux (Debian/Ubuntu/Fedora/...)

curl -fsSL https://get.docker.com | sh
sudo usermod -aG docker "$USER"   # log out and back in afterwards
docker run --rm hello-world

The Compose plugin (docker compose) is included. Check with docker compose version.

Raspberry Pi 5 / ARM servers

Use a 64-bit OS (uname -m must print aarch64). The linux/arm64 image works on 8 GB boards; prefer gliformer-base-v1 and the ONNX int8 backend for acceptable latency on a Pi.

Proxmox VE

Two options, both supported:

LXC container VM
Overhead Lowest Slightly higher
Docker inside Needs nesting=1 (+ keyctl=1 if unprivileged) Works out of the box
Best for Dokploy / Docker host on a NUC Strict isolation

LXC checklist (Proxmox UI → container → Options → Features):

  • nesting=1 and, for unprivileged containers, keyctl=1.
  • Resources: 8 vCPU, 12–16 GB RAM, 4 GB swap, 30 GB disk for a comfortable gliformer-large-v1 setup.
  • Inside the LXC: install Docker with the Linux instructions above.

Then deploy with Docker Compose or Dokploy.

Verify

docker info --format '{{.OSType}}/{{.Architecture}} CPUs={{.NCPU}} Mem={{.MemTotal}}'