Installation

Requirements

  1. Linux operating system (e.g., Debian, Ubuntu)

  2. NVidia GPU (compute capability >= 3; see https://developer.nvidia.com/cuda-gpus)

  3. GPU driver installed (https://www.nvidia.com/en-us/drivers/unix/)

Installation via Conda/Mamba

The originally recommended way to install GinJinn2 was via Conda, an open-source package management system for Python and R, which also includes facilities for environment management (for further information, see https://conda.io/projects/conda/en/latest/index.html). Since installing GinJinn2 via Conda/Miniconda has become slow and error-prone, we now recommend using Mamba (from the Miniforge distribution), a much faster and also more robust reimplementation of the Conda package manager, whose usage is nearly identical.

To install Mamba, run the following commands in your Linux terminal:

wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-x86_64.sh
bash Miniforge3-Linux-x86_64.sh
mamba init

Before installing GinJinn2, we recommend creating a new Conda/Mamba environment to avoid possible version conflicts with existing software. Here, we use Python 3.8; other versions may also work. The environment to be created is named “gj”.

mamba create -n gj python=3.8

To activate this environment, run:

mamba activate gj

Inside the activated environment, run the following command to install GinJinn2 (insert your CUDA version, 10.1 should work for most modern GPUs):

mamba install -c agoberprieler -c pytorch cudatoolkit=10.1 ginjinn2

(Note: When using conda from Miniconda, you also have to specify the conda-forge channel via -c conda-forge.)

Finally, test your installation:

ginjinn -h

NOTE: The activation step is usually required each time you start a new (pseudo)terminal session, otherwise GinJinn2 will not be found. Within the activated environment, commands such as python3 or ginjinn point to programs from your environment, which are separated from system-wide installations. (Try which python3 before and after the activation step to see the difference.)