I am writing a code to classify toxic comment using XLnet pretrained model as in this tutorial: https://towardsdatascience.com/multi-label-text-classification-with-xlnet-b5f
