BERT.pynb has the code for using the bert pretrained model and finetuning on the available data set.
There are 4 other pynb files which has the code for the standard classifiers and their specific parameters.
Datasets has two subfolders. one is for english dataset and other is for hindi dataset.
Labels folder containes the predicted labels for different datasets by BERT model.
Evaluation script is used to get the confusion matrix out from those prodicted labels.