embedplyr/

ger.all_sgns

By-decade word2vec (SGNS) embeddings trained on Google N-Grams ger-all (German, 1800s-1990s).

rds files are R numeric matrices with tokens as rownames.

References:

William L. Hamilton, Jure Leskovec, and Dan Jurafsky. ACL 2016. Diachronic Word Embeddings Reveal Statistical Laws of Semantic Change. https://nlp.stanford.edu/projects/histwords/