Make sure you have Python 3.8 or later installed.

The recommended way to install Python is to use Conda by installing one of


Then create an environment for working with gatenlp. This example creates an environment with the name gatenlp, python 3.11 and activates it:

conda create -n gatenlp python==3.11
conda activate gatenlp

Note that the gatenlp package is compatible with several of the most recent Python versions: the most recent version is compatible with: Python compatibility

The gatenlp has a number of optional dependencies which are only needed if some special features of gatenlp are used.

To install gatenlp with the minimal set of dependencies run:

python -m pip install gatenlp 

To upgrade an already installed gatenlp package to the latest version run:

python -m pip install -U gatenlp 

To install gatenlp with all dependencies run:

python -m pip install gatenlp[all,notebook]

To upgrade to the latest version with all dependencies:

python -m pip install  -U gatenlp[all,notebook]

To upgrade or install to the latest version, INCLUDING pre-release, alpha, or beta versions, use:

python -m pip install --pre -U gatenlp[all,notebook]

NOTE: if this fails because of a problem installing torch (this may happen on Windows), first install Pytorch separately according to the Pytorch installation instructions, see: then run the gatenlp installation again.

The following specific dependencies included in ‘all’ can be chosen separately:

The following dependencies are not included in ‘all’ but in ‘alldev’ or can be chosen separately:

Example: to install gatenlp with support for stanza and spacy and serialization:

python -m pip install gatenlp[stanza,spacy,formats]

To install the latest gatenlp code from GitHub with all dependencies:

To also install everything needed for development use “alldev”:

python -m pip install -e .[alldev]

Creating a jupyter notebook/lab kernel:

To use the conda environment with jupyter notebook/lab create kernel:

Installing directly from GitHub

To install directly from the current GitHub main branch:

python -m pip install

To also specify extras:

python -m pip install[stanza]

To install a specific branch, tag, commit:

python -m pip install

Requirements for using the GATE slave:

Requirements for running gatenlp in a Jupyter notebook:

To create a kernel for your conda environment run:

python -m ipykernel install --user --name gatenlp --display-name "Python gatenlp"

The available kernels can be listed with jupyter kernelspec list

To run and show a notebook run the following and use “Kernel - Change Kernel” in the notebook to choose the gatenlp environment speicific kernel:

jupyter notebook notebookname.ipynb

If you prefer Jupyter lab:

python -m pip install jupyterlab

and then start Jupyter lab with:

jupyter lab

In Jupyter lab, you can work on Jupyter notebooks but also use an interactive console which is also able to visualize documents interactively.

Requirements for development: