weighted aggregation (regression ensemble)

I have trained five regression learners and their RMSE are the following:

learner1=0.8435

learner1=0.8665

learner1=0.7776

learner1=0.7876

learner1=0.6039

I want to take their weighted average in such a way that the learner with lowest error has the highest weight. Any reference code, paper or direct help will be appreciated.