Llama-Zenity is a minimal graphical interface to launch Llama.cpp . It works on Linux and MacOS, and maybe Windows.
- Zenity:
sudo apt update && sudo apt install zenity - Llama.cpp: You should follow installation instructions on their website. I recommend installing into a directory called
srcunder your home directory (i.e.$HOME/src/llama.cpp/).
-
Edit Llama-Zenity.desktop in the line Exec. Make sure your home directory is correct (probably not just
User). Also make sure the paths to this repository and to thellama.cpprepository are correct.editor Llama-Zenity.desktop
-
I recommend storing your GGUF model files in directory called
modelsunder your home directory (i.e.$HOME/models/). If you keep them somewhere else, you can modifyMODEL_DIRinllama-zenity.sh.editor llama-zenity.sh
-
Copy Llama-Zenity.desktop to your desktop.
cp Llama-Zenity.desktop $HOME/Desktop/
That's it! Now you can click on the desktop icon and start chatting away.
- Zenity:
brew install zenity - Llama.cpp: You should follow installation instructions on their website. I recommend installing into a directory called
srcunder your home directory (i.e.$HOME/src/llama.cpp/).
-
Edit the file Llama-Zenity.command . Make sure your home directory is correct (probably not just
user). Also make sure the paths to this repository and to thellama.cpprepository are correct.open -t Llama-Zenity.command
-
I recommend storing your GGUF model files in directory called
modelsunder your home directory (i.e.$HOME/models/). If you keep them somewhere else, you can modifyMODEL_DIRinllama-zenity.sh.open -t llama-zenity.sh
-
Copy Llama-Zenity.command to your desktop.
cp Llama-Zenity.command $HOME/Desktop/
That's it! Now you can click on the desktop icon and start chatting away. You can also drag the icon to the dock, right of the separator.
Contributions are welcome, via pull requests.