Experimented with L1, Filter, and Channel Pruning to achieve a 2% loss in accuracy and 50% decrease in network size