I have a dataset of images whose names are stored in the first column of a CSV file and the target labels are stored in the second column. I am trying to generate training and v