* B0-B3 weights ported from TF with close to paper accuracy
* Renamed gen_mobilenet to gen_efficientnet since scaling params go well beyond 'mobile' specific
* Add Tensorflow preprocessing option for closer images to source repo
* Do mixup in custom collate fn if prefetcher enabled, reduces performance impact
* Move mixup code to own file
* Add arg to disable prefetcher
* Fix no cuda transfer when prefetcher off
* Random erasing when prefetcher off wasn't changed to match new args, fixed
* Default random erasing to off (prob = 0.) for train
* factor out data related constants to own file
* move data related config helpers to own file
* add a variant of RandomResizeCrop that randomizes interpolation method
* remove old Numpy version of RandomErasing
* cleanup torch version of RandomErasing and use it in either GPU loader batch mode or single image cpu Transform
* create one resolve fn to pull together model defaults + cmd line args
* update attribution comments in some models
* test update train/validation/inference scripts
* All models have 'default_cfgs' dict
* load/resume/pretrained helpers factored out
* pretrained load operates on state_dict based on default_cfg
* test all models in validate
* schedule, optim factor factored out
* test time pool wrapper applied based on default_cfg
* Move 'test time pool' to Module that can be used by any model, remove from DPN
* Remove ResNext model file and combine with ResNet
* Remove fbresnet200 as it was an old conversion and pretrained performance not worth param count
* Cleanup adaptive avgmax pooling and add back conctat variant
* Factor out checkpoint load fn
Ross Wightman