add unit test to check that each input slice has uniform distribution - facilitates uniform sampling of training data for better model accuracy

[maxkazmsft]

Related to #259 - this is essentially a test for it

can test that for checkerboard dataset train and val patches have a uniform 50/50 class distribution across all patches

[maxkazmsft]

We should re-scope this issue - there should be a tolerance for sampling synthetic checkerboard training data where with a large enough number of samples the distribution of two classes is roughly 50/50, but it's not clear just how 50/50 the distribution would be