Installation

Grew-TSE relies on the great work carried out on Grew (Graph Rewriting for NLP) and therefore a number of installations must be carried out as a prerequisite. Without these additional installation steps you will not be able to use this package.

Opam & Grewpy

You must first install Opam. This will allow the grewpy package (automatically installed with Grew-TSE) to work properly. If using Windows, please note that we have only tested this when using WSL. You can install Opam on Linux using the following command:

bash -c "sh <(curl -fsSL https://opam.ocaml.org/install.sh)"

Or on Windows Powershell:

Invoke-Expression "& { $(Invoke-RestMethod https://opam.ocaml.org/install.ps1) }"

To initiate the opam setup, you can then run:

opam init

These scripts automatically detect your system architecture and carry out the installation. More details on installing Opam are provided here.

The necessary grewpy backend can then be installed with the following commands:

opam remote add grew "https://opam.grew.fr"
opam update
opam install grewpy_backend

You will need to tell your system where to find OPAM’s bin directory. Run the following to have this permanently added:

echo 'eval $(opam env)' >> ~/.bashrc

Grew-TSE Pipeline

You can install Grew-TSE with the below command using pip, which should automatically install any dependencies required.

pip install grew-tse

The main pipeline in Grew-TSE requires Python 3.8 or higher and the below Python packages. - conllu==6.0.0 - grewpy==0.6.0 - numpy==2.2.5 - pandas==2.2.3. These should be installed automatically using the above install of Grew-TSE.

Evaluation

If you want to use any of the Grew-TSE evaluation tools, then you’ll also need the eval dependencies. Install them like so:

pip install grew-tse[eval]

For any issues with `grewpy`, visit the Grew documentation for more information on the installation as well as how to test whether your installation was successful.