Is a validation dataset created from train data?

[claell]

Does this approach use validation data for the test accuracy during training or is that based on the actual test data? That would probably introduce some bias when those results are used to adjust hyperparameters.

[ranahanocka]

You can place a validation split in the "test" folder -- and only run inference on the real test as often as desired

[claell]

Alright, thanks. So out of the box there is no validation split, but it can be used as you described.