Recording

Storing data

xmean,xstd = 0.28, 0.35
@inplace
def transformi(b): b['image'] = [(TF.to_tensor(o)-xmean)/xstd for o in b['image']]

_dataset = load_dataset('fashion_mnist').with_transform(transformi)

100%|██████████| 2/2 [00:00<00:00, 507.20it/s]

dls = DataLoaders.from_dataset_dict(_dataset, 256, num_workers=4)

---

source

BatchAugmentationCB

 BatchAugmentationCB (tfms)

Initialize self. See help(type(self)) for accurate signature.

---

source

UnNormalize

 UnNormalize (mean, std)

Initialize self. See help(type(self)) for accurate signature.

---

source

show_doc

 show_doc (sym, renderer=None, name:str|None=None, title_level:int=3)

Show signature and docstring for sym

	Type	Default	Details
sym			Symbol to document
renderer	NoneType	None	Optional renderer (defaults to markdown)
name	str | None	None	Optionally override displayed name of sym
title_level	int	3	Heading level to use for symbol name

tfms = [torch.nn.Sequential(transforms.RandomVerticalFlip(1),transforms.RandomErasing(1))]
trainer = Trainer(dls,
                  nn.CrossEntropyLoss(), 
                  torch.optim.Adam, 
                  get_model_conv(),
                  callbacks=[BasicTrainCB(),DeviceCB(),BatchAugmentationCB(tfms)])

trainer.show_image_batch(3)