sdadas / polish-roberta-base-v2

An encoder model based on the RoBERTa architecture, pre-trained on a large corpus of Polish texts. More information can be found in our GitHub repository and in the publication Pre-training polish transformer-based language models at scale.