Effcient Continuous Pareto Exploration in Multi-Task Learning