rubert-base-cased

RuBERT \(Russian, cased, 12‑layer, 768‑hidden, 12‑heads, 180M parameters\) was trained on the Russian part of Wikipedia and news data. We used this training data to build a vocabulary of Russian subtokens and took a multilingual version of BERT‑base as an initialization for RuBERT\[1\].

08.11.2021: upload model with MLM and NSP heads

\1\]: Kuratov, Y., Arkhipov, M. \(2019\). Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language. arXiv preprint [arXiv:1905.07213.