Feature importance formula from Random Forest Learner

Hello everyone! I finally got the workflow of nodes in my random forest model to align in such a way that data leakage has been stopped. Model returns a 93% training/validation accuracy and a 91% test accuracy. Now I need to calculate feature variable importance. Is there a method to convert splits and candidates to importance as a percentage? Screenshot below for reference.

If so what is the formula? I remember seeing a screenshot someone posted that does just that but can’t seem to locate it. Any help would be so helpful.

Many thanks in advance!

~Cole K.