Greetings, and apologies for replying late.
For Wikidata, we recommend one of these options:
3) USC ISI has a very nice tool called KGTK
https://github.com/usc-isi-i2/kgtk It can help you avoid needing a SPARQL endpoint, but there is also a public endpoint although I am not quite sure what version it hosts and what limits it has:
https://kgtk.isi.edu/sparql/ (If ISI folks are here, please let us know!)
For DBpedia for Round 2, the version used for our data set is 2016-10 so you may need to host your own endpoint. If anyone is aware of publicly available endpoints with this version, please let us know.
Cheers,