# Train (80%), val (10%), test (10%) splits for IWildCam 2022

Each text file is a list of sequence IDs. We sampled from the original train set samples, that has `count` annotations, which is 1780 samples total.