You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When training and testing sets are generated in an 80/20 ratio, if there is more than one category they are balanced, that is, would each category in the json file have 80/20?
The text was updated successfully, but these errors were encountered:
This repo doesn't preserve class distributions to be split in the same ratio, it does a random split on the images and then includes annotations that belong to those images in the train/test sets.
I have updated this code base to include functionality of splitting while preserving balanced class distributions, you can try it from here: https://github.com/ahmad-ra/cocosplit
When training and testing sets are generated in an 80/20 ratio, if there is more than one category they are balanced, that is, would each category in the json file have 80/20?
The text was updated successfully, but these errors were encountered: