Yale-b face database

How the five sets of extended yale-b face database are constructed? in order to get the number of samples for each subset as follows:

subset 1= 263 samples 
subset 2= 456 samples
subset 3= 455 samples
subset 4= 526 samples
subset 5= 714 samples

I want to know the labels that have been set for each subset please

Thank you!