.. _with-randaugment: using RandAugment with niacin =============================== RandAugment (randaugment_) is an algorithm for applying a variety of combinations of augmentations to training data, and experimentally appears to produce results on par with more complicated data augmentation policies like AutoAugment (autoaugment_). In the RandAugment algorithm, there is a predefined set of ``k`` augmentation functions. For each training sample, a random subset of size ``n<