Bigram dictionaries v1
·
515 commits
to main
since this release
Per-language bigram (2-gram) frequency files for the autocorrect context model (Tier 2). Generated from the Leipzig Corpora Collection (wortschatz-leipzig.de), CC BY — © Universität Leipzig / Sächsische Akademie der Wissenschaften / InfAI; Goldhahn, Eckart & Quasthoff, LREC 2012. Format: ' \t', top 60k by count.