Problem about preprocessing

[wenhuchen]

I found the the splitting step is done after shuffle according to https://github.com/karpathy/neuraltalk2/blob/master/prepro.py (line 159), which means every time we get different splits for val/test set. Even after I removed the shuffle step, I output to get different splits with described in https://github.com/kelvinxu/arctic-captions/tree/master/splits, but they claim to get the same split as neuraltalk2, did I miss something?