This notebook is part of the Kaggle learning competition on Natural Language Processing with Disaster Tweets. The aim of the notebook was to practice fine-tuning and implementing various methods to improve a single metric (F1 score in this case). The model achieved decent score of 0.84063.