def test_from_torch_dataset_with_transform(self):
dataset_ = torchvision.datasets.FakeData(size=1,
image_size=(3, 32, 32))
dataset = LightlyDataset.from_torch_dataset(
dataset_, transform=torchvision.transforms.ToTensor())
self.assertIsNotNone(dataset.transform)
self.assertIsNotNone(dataset.dataset.transform)
student_out,
epoch=self.current_epoch)
return loss
def configure_optimizers(self):
optim = torch.optim.Adam(self.parameters(), lr=0.001)
return optim
model = DINO()
# we ignore object detection annotations by setting target_transform to return 0
pascal_voc = torchvision.datasets.VOCDetection("datasets/pascal_voc",
download=True,
target_transform=lambda t: 0)
dataset = LightlyDataset.from_torch_dataset(pascal_voc)
# or create a dataset from a folder containing images or videos:
# dataset = LightlyDataset("path/to/folder")
collate_fn = DINOCollateFunction()
dataloader = torch.utils.data.DataLoader(
dataset,
batch_size=64,
collate_fn=collate_fn,
shuffle=True,
drop_last=True,
num_workers=8,
)
gpus = torch.cuda.device_count()
def forward(self, x):
x = self.backbone(x).flatten(start_dim=1)
z = self.projection_head(x)
return z
resnet = torchvision.models.resnet18()
backbone = nn.Sequential(*list(resnet.children())[:-1])
model = BarlowTwins(backbone)
device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)
cifar10 = torchvision.datasets.CIFAR10("datasets/cifar10", download=True)
dataset = LightlyDataset.from_torch_dataset(cifar10)
# or create a dataset from a folder containing images or videos:
# dataset = LightlyDataset("path/to/folder")
collate_fn = ImageCollateFunction(input_size=32)
dataloader = torch.utils.data.DataLoader(
dataset,
batch_size=256,
collate_fn=collate_fn,
shuffle=True,
drop_last=True,
num_workers=8,
)
criterion = BarlowTwinsLoss()
def test_from_torch_dataset(self):
_dataset = torchvision.datasets.FakeData(size=1,
image_size=(3, 32, 32))
dataset = LightlyDataset.from_torch_dataset(_dataset)
self.assertEqual(len(_dataset), len(dataset))
self.assertEqual(len(dataset.get_filenames()), len(dataset))