hemekci/off_detection_turkish

The Offensive Language Detection Model in Turkish has been tuned to Twitter data . Bert and pytorch are programmed with Twitter data to detect offensive language errors . The model is based on the language language of Bert and Pytorch . The training data is done using a language-free language model called "Bert" and "Pytorch" The model was used to detect errors and errors during training sessions . Matthews Corr. Coef scored 0.633 with a MCC score of 0.533 . The MCC model is used to identify offensive language error in Turkish Tweets. Bert is a free-form language that can be translated into English. The model has been used,