#NLP The approach leveraged to tackle this task of Grammatical Error Detection is by fine-tuning pretrained BERT models. We see that the no. of positive samples and negative samples in training data are equal (9999 of each) with a total of 19998 examples in the training set. The average length of a sentence is about 12 words, with max length sentence being 319 words