You can download the datasets stored in the SPARQL endpoint from this page:

(__Note__: the .nt files are very big, the .nt.bz2 files are ~20 times lower)

[http://es.dbpedia.org/DBpediaESdata/]