README.md

UzRoBerta model.

Pre-prepared model in Uzbek (Cyrillic script) to model the masked language and predict the next sentences.

Training data.

UzBERT model was pretrained on ≈167K news articles (≈568Mb).