Skip to main content

Installation Guide

GoingMerry can be run locally across multiple platform targets.

:::warning Compiler Staging & Calibration Our binaries are currently undergoing final calibration across all system targets to guarantee branch-free performance and driver compatibility. During this optimization period, direct downloads are staged via our Waitlist.

Please authenticate and register your node ID on the Profile Page to get early waitlist access. :::

macOS​

GoingMerry leverages macOS Unified Memory architectures by using native Apple Metal performance kernels.

Once calibration is complete, macOS installation will be available via a single zip archive containing the standalone app bundle:

  1. Download the GoingMerry app package.
  2. Drag GoingMerry.app to your /Applications directory.
  3. Launch the app to automatically register the background service and map the CLI merry binary to your system PATH.

To verify your terminal shell can access the binary:

merry --version

Linux​

For Linux boundaries, GoingMerry supports CPU vectorization (AVX, AVX2, AVX-512) and hardware graphics wrappers for both Nvidia (CUDA) and AMD (ROCm).

Quick Install Script (Staged)​

Once calibration finishes, a one-line script will bootstrap the installation:

curl -fsSL https://dopove.com/install.sh | sh

Manual Installation​

You can download and extract the precompiled binary target:

# Download the binary (Staged)
sudo curl -L https://dopove.com/download/merry-linux-amd64 -o /usr/local/bin/merry
sudo chmod +x /usr/local/bin/merry

Running as a systemd Service​

To ensure the runtime daemon stays online, create a systemd configuration file at /etc/systemd/system/merry.service:

[Unit]
Description=GoingMerry AI Service
After=network-online.target

[Service]
ExecStart=/usr/local/bin/merry serve
User=merry
Group=merry
Restart=always
RestartSec=3
Environment="MERRY_HOST=0.0.0.0:11434"

[Install]
WantedBy=default.target

Enable and start the service:

sudo systemctl daemon-reload
sudo systemctl enable merry
sudo systemctl start merry

Windows​

On Windows, GoingMerry executes directly on bare-metal graphics drivers using DirectCompute pipeline extensions.

  1. Download the standalone Windows installer executable (GoingMerrySetup.exe).
  2. Run the setup and follow the installation wizard.
  3. The background monitor tray icon will load, mapping merry.exe to system variables.

Docker​

If you prefer to keep GoingMerry fully isolated in a containerized environment, we support official GPU-enabled Docker builds:

Nvidia GPU Support​

docker run -d --gpus=all -v merry:/root/.merry -p 11434:11434 --name merry dopove/merry

AMD GPU Support (ROCm)​

docker run -d --device /dev/kfd --device /dev/dri -v merry:/root/.merry -p 11434:11434 --name merry dopove/merry:rocm

CPU Only​

docker run -d -v merry:/root/.merry -p 11434:11434 --name merry dopove/merry