I'm using the torch.utils.data.DataLoader
to lookup the filename in an iterable-style dataset. When calling DataLoader.dataset.samples[i]
within an the iterative loop, I get the filename and an integer.
Example:
('car.JPEG', 0)
What does the second parameter stand for?
The DataLoader
returns each member of the dataset
object which is a tuple containing the sample and the class_index it belongs to. So in your case, the first index of the tuple contains the sample (car.JEPG
) and the second index is the class_index that the sample belongs to (0
in your case). To learn more go through the docs.