Put up some big demo datasets on the website

[Ostrzyciel]

We have some sample datasets put up here, from RiverBench: https://jelly-rdf.github.io/dev/use-cases/#example-datasets-in-the-jelly-format

These are big, but not very big. Convert via Jena RIOT something on the order of billions of triples to Jelly and put it up there. probably via GitHub releases or via Zenodo (up to 50GB, I think). This could be Wikidata, UniProt, OSM, or something else. There's lots of good inspiration in the docs of QLever, where they use these as demo datasets.