Question about the Splitter

[ssshamiii]

Thanks for sharing the dataset! I am confused about the dataset when using the splitter for single subject, as it only has 669 indexes for training dataset while 167 for validation dataset and 164 for test dataset, and the sum of them is 1000 but the total number of indexes is about 2000. It would be really appreciated if you could explain it, thanks again!

[feroooooo]

I have the same question. It seems there are 1996 sample of subject 4, but only 1000 are used in "block_splits_by_image_single.pth".