Nova41/SnowLeopard

How to effective train the network for the best result

buimanhduc opened this issue · 0 comments

Same as the title, can you help me?
Should I have small amount of category with a lot of samples per category or a lot of categories with a small samples amount per category?