Description of files:

Approach.ipynb - The main notebook containing all the code.
BERT preprocessing.ipynb - A notebook for preprocessing the data for BERT finetuning
swapnil_parekh.csv - the main submission file
lgb.model - the lightgbm model scheckpoint

checkpoints - the neural net model checkpoints
run_classifier.py - is the file included in the bert repo for fine tuning BERT
runbert2.bat - training the BERT finetuning
runbert3.bat - predicting on the new finetuned BERT model
train.tsv, dev.tsv, test.tsv - files used in creating the testing the BERT fined tuned model

CREDITS:

1. https://github.com/google-research/bert
   
2. https://github.com/hanxiao/bert-as-service
   
3. GLOVE vectors