Functions — github.com/DanielShalam/BPA

CodeHub MCPcopy Create free account

hub / github.com/DanielShalam/BPA / functions

Functions184 in github.com/DanielShalam/BPA

⨍ Functions 184◇ Types & classes 38

(self, window_size=20, fmt=None)

(self, delimiter="\t")

(self, args: dict, bpa: BPA = None)

methods/prototypical/proto_loss.py:11

methods/pt_map/save_plk.py:18

(self, num_way: int, num_shot: int, num_query: int, lam: float)

methods/pt_map/pt_map_loss.py:22

(self, args: dict, lam: float = 10, alpha: float = 0.2, n_epochs: int = 10, bpa=None)

methods/pt_map/pt_map_loss.py:113

methods/pt_map/test_standard.py:46

(self, n_ways, lam, distance_metric: str = 'euclidean')

methods/pt_map/test_standard.py:52

(self, transformdict)

methods/pt_map/data/additional_transforms.py:16

(self, data_file, transform, target_transform=identity)

methods/pt_map/data/dataset.py:15

(self, data_file, batch_size, transform)

methods/pt_map/data/dataset.py:35

(self, sub_meta, cl, transform=transforms.ToTensor(), target_transform=identity)

methods/pt_map/data/dataset.py:65

(self, n_classes, n_way, n_episodes)

methods/pt_map/data/dataset.py:84

(self, image_size, batch_size)

methods/pt_map/data/datamgr.py:51

methods/pt_map/evaluation/save_plk.py:33

methods/pt_map/evaluation/test_standard.py:49

(self, n_ways, lam)

methods/pt_map/evaluation/test_standard.py:54

(self, set_name, labels, num_episodes, num_way, num_shot, num_query, const_loader,

datasets/samplers.py:7

(self, data_path: str, setname: str, backbone: str, augment: bool)

datasets/cifar.py:10

(self, setname, args, augment=False)

datasets/cub.py:19

(self, data_path: str, setname: str, backbone: str, augment: bool)

datasets/mini_imagenet.py:10

(self, indim, outdim)

models/res_mixup_model.py:35

(self, inplanes, planes, stride=1, downsample=None)

models/res_mixup_model.py:63

(self, indim, outdim)

models/wrn_mixup_model.py:47

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

models/wrn_mixup_model.py:73

(self, depth=28, widen_factor=10, num_classes=200, loss_type='dist', per_img_std=False, stride=1,

models/wrn_mixup_model.py:118

(self, block_size)

models/dropblock.py:8

(self, inplanes, planes, stride=1, downsample=None, drop_rate=0.0, drop_block=False, block_size=1)

models/resnet12.py:20

:param distance_metric - Compute the cost matrix. :param ot_reg - Sinkhorn entropy regularization (lambda). For few-shot classificati

bpa/balanced_pairwise_affinities.py:12

methods/pt_map/data/dataset.py:92

datasets/samplers.py:49

methods/pt_map/data/dataset.py:30

methods/pt_map/data/dataset.py:60

methods/pt_map/data/dataset.py:79

methods/pt_map/data/dataset.py:89

datasets/samplers.py:46

datasets/cifar.py:50

datasets/cub.py:89

datasets/mini_imagenet.py:74

Methodadd_meter

(self, name, meter)

Functionbool_flag

Parse boolean arguments from the command line.

Functionbool_flag

Parse boolean arguments from the command line.

methods/pt_map/test_standard.py:14

(self, X: torch.Tensor, labels: torch.Tensor, mode: str)

methods/prototypical/proto_loss.py:28

methods/pt_map/save_plk.py:22

(self, X: torch.Tensor, labels: torch.Tensor, mode: str)

methods/pt_map/pt_map_loss.py:132

methods/pt_map/evaluation/save_plk.py:36

models/res_mixup_model.py:47

models/res_mixup_model.py:73

(self, x, target=None, mixup=False, mixup_hidden=True, mixup_alpha=None, lam=0.4)

models/res_mixup_model.py:142

models/wrn_mixup_model.py:34

models/wrn_mixup_model.py:59

models/wrn_mixup_model.py:83

(self, x, target=None, mixup=False, mixup_hidden=True, mixup_alpha=None, lam=0.4, return_logits=True)

models/wrn_mixup_model.py:151

(self, x, gamma)

models/dropblock.py:13

models/resnet12.py:38

models/resnet12.py:112

Compute the BPA feature transform

bpa/balanced_pairwise_affinities.py:83

Functionget_assigned_file

(checkpoint_dir, num)

methods/pt_map/io_utils.py:38

Functionget_best_file

(checkpoint_dir)

methods/pt_map/io_utils.py:55

Functionget_criterion_by_method

Get loss function based on the method.

Functionget_dataloader

Get dataloader with categorical sampler for few-shot classification.

Functionget_fs_labels

Prepare few-shot labels. For example for 5-way, 1-shot, 2-query: [0, 1, 2, 3, 4, 0, 1, 2, 3, 4, ...]

Functionget_method

Get the few-shot classification method (e.g. pt_map).

Functionget_model

Get the backbone model.

Functionget_optimizer_and_lr_scheduler

Functionget_output_dir

Initialize the output dir.

Functionget_transform

(img_size: int, split_name: str)

datasets/cifar.py:59

Methodglobal_avg

Functionidentity

methods/pt_map/data/dataset.py:10

Functioninit_wandb

Initialize and returns wandb logger if args.wandb is True.

FunctionloadDataSet

(dsname, root: str = None, features_path: str = None)

methods/pt_map/FSLTask.py:43

Functionload_weights

Load pretrained weights from given path.

Functionlog_sinkhorn

Log-space-sinkhorn algorithm for better stability.

Functionprint_and_log

Print and log current results.

Functionresnet18

Constructs a ResNet-18 model.

models/res_mixup_model.py:200

Functionset_seed

Functionto_one_hot

(inp, num_classes)

models/wrn_mixup_model.py:87

Functionwandb_log

Log step to the logger without print.

Functionwrn28_10

(num_classes=200, loss_type='dist', dropout=0)

models/wrn_mixup_model.py:206

← previous101–184 of 184, ranked by callers