We developed and evaluated an RL based abstractive summarisation system that was designed to address the issues current ROUGE focused summarisation models face, such as overly long/ incomplete outputs. I specifically contributed to designing the multii objective reward function that combined…