The embeddings (100 dimensions) are generated using Pytorch-BigGraph (https://github.com/facebookresearch/PyTorch-BigGraph). Each line in the TSV file maps the id (first column) of the entity/relation against its embedding (rest of the columns). The mapping between the id to its entity or relation can be found in "entity2id.txt" and "relation2id.txt"
----------------------------------------------
Knowledge graph embeddings generated on a combination of DBPedia 15-04 subsets. The following subsets were used:
article_categories
citation_links
instance_types
instance_types_transitive
mappingbased_objects
page_links
persondata
skos_categories
----------------------------------------------
Size Stats:
239123030 triples
22707453 Entities
675 Relations
----------------------------------------------

Configuration used for ComplEx: https://github.com/nikit91/configs/blob/master/kge/pbg/dbp_config_cpu_complex.py
Configuration used for Translation-Dot: https://github.com/nikit91/configs/blob/master/kge/pbg/dbp_config_cpu_old.py

Information on the configuration attributes: https://torchbiggraph.readthedocs.io/en/latest/configuration_file.html