Alternative solution adapted from PyTorch docs.:

for i, data in enumerate(train_loader, 0):
    ex_images, ex_dmaps, ex_n_people = data