Functions — github.com/imrahulr/hat

CodeHub MCPcopy Index your code

hub / github.com/imrahulr/hat / functions

Functions206 in github.com/imrahulr/hat

⨍ Functions 206◇ Types & classes 49 ↳ Endpoints 1

↓ 1 callersFunctionupdate_bn

Update batch normalization layers.

gowal21uncovering/utils/watrain.py:213

↓ 1 callersFunctionwideresnet

Returns suitable Wideresnet model from its name. Arguments: name (str): name of resnet architecture. num_classes (int): numbe

core/models/wideresnet.py:122

↓ 1 callersFunctionwideresnetwithswish

Returns suitable Wideresnet model with Swish activation function from its name. Arguments: name (str): name of resnet architecture.

core/models/wideresnetwithswish.py:171

CW loss (Marging loss).

core/attacks/utils.py:201

core/utils/utils.py:162

(self, *args, **kwargs)

core/attacks/base.py:35

Method__enter__

core/utils/context.py:10

Method__enter__

core/utils/context.py:24

core/utils/context.py:13

core/utils/context.py:27

Method__getattr__

eval-adv.py:174

Method__getitem__

core/data/semisup.py:117

Method__getitem__

core/data/utils.py:41

core/utils/context.py:5

core/utils/context.py:19

(self, info, args)

core/utils/train.py:37

(self, optimizer, max_lr, epochs, last_epoch=-1)

core/utils/rst.py:10

(self, smoothing=0.0, reduction='mean')

core/utils/utils.py:14

core/utils/logger.py:10

(self, sup_inds, unsup_inds, batch_size, unsup_fraction=0.5, num_batches=None)

core/data/semisup.py:126

(self, root, transform=NumpyToTensor(), target_transform=None)

core/data/utils.py:19

(self, data_path, **kwargs)

core/data/imagenet100.py:17

(self, predict, loss_fn=None, eps=0.3, clip_min=0., clip_max=1., targeted=False)

core/attacks/fgsm.py:21

(self, predict, loss_fn='ce', n_restarts=2, eps=0.3, nb_iter=40, seed=1)

core/attacks/apgd.py:49

(self, predict, loss_fn='ce', n_restarts=2, eps=0.3, nb_iter=40, seed=1)

core/attacks/apgd.py:67

(self, predict, loss_fn, clip_min, clip_max)

core/attacks/base.py:17

( self, predict, loss_fn=None, eps=0.3, nb_iter=40, eps_iter=0.01, rand_init=True, clip_min=0., cl

core/attacks/pgd.py:159

( self, predict, loss_fn=None, eps=0.3, nb_iter=40, eps_iter=0.01, rand_init=True, clip_min=0., cl

core/attacks/pgd.py:184

( self, predict, overshoot=0.02, nb_iter=50, search_iter=50, clip_min=0., clip_max=1.)

core/attacks/deepfool.py:166

( self, predict, overshoot=0.02, nb_iter=50, search_iter=50, clip_min=0., clip_max=1.)

core/attacks/deepfool.py:188

(self, in_planes, planes, stride=1)

core/models/ti_preact_resnet.py:47

(self, block, num_blocks, num_classes=200)

core/models/ti_preact_resnet.py:75

(self, num_blocks, in_planes, out_planes, stride, activation_fn=nn.ReLU)

core/models/wideresnetwithswish.py:81

(self, num_classes: int = 10, depth: int = 28, width: int =

core/models/wideresnetwithswish.py:110

(self, in_planes, planes, stride=1)

core/models/in_preact_resnet.py:47

(self, block, num_blocks, num_classes=10)

core/models/in_preact_resnet.py:75

(self, in_planes, planes, stride=1)

core/models/preact_resnet.py:47

(self, block, num_blocks, num_classes=10)

core/models/preact_resnet.py:75

(self, mean, std)

core/models/resnet.py:15

(self, in_planes, planes, stride=1)

core/models/resnet.py:43

(self, in_planes, planes, stride=1)

core/models/resnet.py:75

(self, nb_layers, in_planes, out_planes, block, stride, dropRate=0.0)

core/models/wideresnet.py:57

(self, depth=34, num_classes=10, widen_factor=10, dropRate=0.0)

core/models/wideresnet.py:80

(self, num_classes: int = 10, depth: int = 18, width: int = 0, #

core/models/preact_resnetwithswish.py:82

(self, info, args)

gowal21uncovering/utils/watrain.py:36

core/data/semisup.py:144

core/data/semisup.py:114

core/data/semisup.py:161

core/data/utils.py:38

Function_thresh_by_magnitude

Threshold by magnitude.

core/attacks/utils.py:126

Functioncalc_l2distsq

Calculate L2 distance between tensors x and y.

core/attacks/utils.py:27

Functionclear_data

core/setup.py:40

core/data/semisup.py:99

Functioneval_multiple_restarts_advertorch

Evaluate adversarial accuracy with multiple restarts (Advertorch).

(self, input, target)

core/utils/utils.py:19

core/models/ti_preact_resnet.py:28

core/models/ti_preact_resnet.py:61

core/models/ti_preact_resnet.py:96

core/models/wideresnetwithswish.py:53

core/models/wideresnetwithswish.py:93

core/models/wideresnetwithswish.py:152

core/models/in_preact_resnet.py:28

core/models/in_preact_resnet.py:61

core/models/in_preact_resnet.py:100

core/models/preact_resnet.py:28

core/models/preact_resnet.py:61

core/models/preact_resnet.py:95

core/models/resnet.py:22

core/models/resnet.py:57

core/models/resnet.py:91

core/models/resnet.py:128

core/models/wideresnet.py:34

core/models/wideresnet.py:67

core/models/wideresnet.py:111

core/models/preact_resnetwithswish.py:60

core/models/preact_resnetwithswish.py:126

core/utils/rst.py:27

Methodgroup_weight

gowal21uncovering/utils/watrain.py:56

Methodinit_optimizer

Initialize optimizer and schedulers.

gowal21uncovering/utils/watrain.py:52

Functionl2_norm

core/utils/trades.py:17

Functionl2_norm

gowal21uncovering/utils/trades.py:17

Methodload_base_dataset

(self, train=False, **kwargs)

core/data/cifar10s.py:49

Methodload_base_dataset

(self, train=False, **kwargs)

core/data/cifar100s.py:48

Functionload_cifar10

Returns CIFAR10 train, test datasets and dataloaders. Arguments: data_dir (str): path to data directory. use_augmentation (bo

core/data/cifar10.py:16

Functionload_cifar100

Returns CIFAR100 train, test datasets and dataloaders. Arguments: data_dir (str): path to data directory. use_augmentation (b

core/data/cifar100.py:16

Functionload_cifar100s

Returns semisupervised CIFAR100 train, test datasets and dataloaders (with DDPM Images). Arguments: data_dir (str): path to data dire

core/data/cifar100s.py:12

Functionload_cifar10s

Returns semisupervised CIFAR10 train, test datasets and dataloaders (with Tiny Images). Arguments: data_dir (str): path to data direc

core/data/cifar10s.py:12

Functionload_imagenet100

Returns ImageNet100 train, test datasets. Arguments: data_dir (str): path to data directory. use_augmentation (bool): whether

core/data/imagenet100.py:25

Methodload_model

Load model weights.

gowal21uncovering/utils/watrain.py:185

Functionload_svhn

Returns SVHN train, test datasets and dataloaders. Arguments: data_dir (str): path to data directory. use_augmentation (bool)

core/data/svhn.py:16

Functionload_tinyimagenet

Returns Tiny Imagenet-200 train, test datasets and dataloaders. Arguments: data_dir (str): path to data directory. use_augmen

core/data/tiny_imagenet.py:18

Functionmeasure_margin

Estimate margin using line search.

core/utils/exp.py:63

Given examples (x, y), returns their adversarial counterparts with an attack length of eps. Arguments: x (torch.Tensor):

core/attacks/fgsm.py:29

Given examples (x, y), returns their adversarial counterparts with an attack length of eps. Arguments: x (torch.Tensor):

core/attacks/fgsm.py:83

Given examples (x, y), returns their adversarial counterparts with an attack length of eps. Arguments: x (torch.Tensor):

core/attacks/pgd.py:107

Given examples x, returns their adversarial counterparts. Arguments: x (torch.Tensor): input tensor. y (torch

core/attacks/deepfool.py:134

Functionpickle_data

Write data to pickled file. Arguments: data (Any): data to be written. filename (str): path to the pickled file. mode

core/utils/utils.py:110

Functionreplicate_input_withgrad

Clone the input tensor x and set requires_grad=True.

core/attacks/utils.py:20

Methodsave_model

Save model weights.

gowal21uncovering/utils/watrain.py:175

Methodset_bn_to_eval

Set all batch normalization layers to evaluation mode.

core/utils/train.py:292

← previous next →101–200 of 206, ranked by callers