jannesg/bertsson

The models are trained on: government Text, government Text and Swedish Literature . They are trained with the same hyperparameters as first published by Google . The following models are currently available: The BERT base model has been trained on the same parameters as Google's BERT . All models are cased and trained with whole word masking . Stay tuned for evaluations of the models as well as the results of the evaluations . The models will be evaluated by the end of the month on a new version of this version of the BERT model . The model is based on the language language used in Swedish news and government texts. The model was trained on these words: Government Text and Government Text. The,