Shuffle is not present for CDON images

[rlindsberg]

Supervised-Image-Classification-with-Noisy-Labels-Using-Deep-Learning/datasets.py

Lines 90 to 92 in 6c61b62

	dataset = CDONdataset("dataset_lables.csv", root_folder, transform=transform)
	train_set = CDONDatasetSplit(dataset, split=0.9, from_bottom=True)
	test_set = CDONDatasetSplit(dataset, split=0.1, from_bottom=False)

[alessioGalatolo]

To my understanding the shuffling should be done inside the dataloader

[rlindsberg]

In this?
def generate_loader_with_noise(dataset, batch_size, shuffle, noise_rate, is_symmetric_noise)

[alessioGalatolo]

It is done inside FastTensorDataLoader:

Supervised-Image-Classification-with-Noisy-Labels-Using-Deep-Learning/datasets.py

Lines 48 to 54 in 6c61b62

	def __iter__(self):
	if self.shuffle:
	self.indices = torch.randperm(self.dataset_len)
	else:
	self.indices = None
	self.i = 0
	return self