Fast and efficient data pre-processing and augmentation pipelines and training primitives for deep learning.
jfolz/crumpets
latest