@KOLANICH Thanks for your pull request, I've added this option: called it max_points, not to confuse with iterations and epochs that happen in training.
Technical detail: the tuner can save its state and then restore and continue working, so I've implemented it slightly differently to support this case as well.
@KOLANICH Thanks for your pull request, I've added this option: called it
max_points
, not to confuse with iterations and epochs that happen in training.Technical detail: the tuner can save its state and then restore and continue working, so I've implemented it slightly differently to support this case as well.
Thanks again!