"randomForest - weka parameters I and K"

I have data set with 60 attributes, approximately 2000 rows. I want to perform classification by weka-randomForest (in RapidMiner). I would be very happy if somebody who has some experiences with this learner will give me advice to which value I should set parameters I (number of trees) and K (number of variables in each tree). I want to do some optimization of parameters, but I really do not know where approximately to start...
Thanks,
Milan