Welcome to this Annif tutorial video.
Here we focus on Docker-based installation of Annif.
This approach is recommended for those already familiar with Docker
The tutorial page has step-by-step instructions, you can also read more from the Annif wiki
There a few things to consider with this approach:
Docker needs enough memory, you can allocate more from the advanced -tab in the settings
You’ll also need to be able to share a folder with Docker for the most seamless experience
This reguires adminstrator access, however, and as that might not be possible to obtain, there’s an alternative approach you can follow.
Installing Annif via Docker on Windows or Mac is as simple as running one of these commands here
docker run -v ~/Annif-tutorial:/Annif-tutorial --network host --name annif -it --rm -u
On windows the command looks like this docker run -v ~/Annif-tutorial:/Annif-tutorial -p 5000:5000 --name annif -it --rm -u root quay.io/natlibfi/annif:tutorial
On windows especially as you might not be able to share a folder with Docker there is this alternative approach you can follow by running this slightly different command docker run -v Annif-tutorial-volume:/Annif-tutorial -p 5000:5000 --name annif -it --rm -u root quay.io/natlibfi/annif:tutorial
followed by this command run in another command shell docker cp data-sets/ annif:/Annif-tutorial
This is what installing annif via docker looks like on windows powershell
We run this command here - and it’s the modified version because we’re doing the alternative approach on windows docker run -v Annif-tutorial-volume:/Annif-tutorial -p 5000:5000 --name annif -it --rm -u root quay.io/natlibfi/annif:tutorial
When it finishes the command prompt changes
The next step, if you are doing the alternative approach on windows, is to navigate to the annif tutorial folder via the cd command and run this command docker cp data-sets/ annif:/Annif-tutorial
Then go back to the original command shell and check by running the ls command that the data-sets folder is where it should be
Then you should be good to go
Thank you for watching!