Functions — github.com/akhilmathurs/orchestra

CodeHub MCPcopy Create free account

hub / github.com/akhilmathurs/orchestra / functions

Functions128 in github.com/akhilmathurs/orchestra

⨍ Functions 128◇ Types & classes 33

↓ 19 callersMethod__init__

(self, block, num_classes=10, cfg=None)

↓ 12 callersFunctioncreate_backbone

(name, num_classes=10, block='BasicBlock')

↓ 8 callersMethodstep

↓ 4 callersMethod_make_layer

(self, block, planes, num_blocks, stride)

↓ 3 callersMethod_make_layer

(self, block, planes, num_blocks, stride)

↓ 3 callersFunctionget_block

↓ 3 callersFunctionget_config_dict

↓ 3 callersFunctionupdate_run_id

(hparam, config_dict, summary_writer, run_id)

hparam_main.py:39

↓ 2 callersFunctionformat_time

↓ 2 callersFunctionget_dataset_subset

(dataset, subset_proportion=1, force_class_balanced=False, seed=0)

↓ 2 callersFunctionget_eval_dict

↓ 2 callersFunctionimage_rot

↓ 2 callersMethodlocal_clustering

(self, device='cuda:7')

↓ 2 callersFunctionprogress_bar

(current, total, msg=None)

↓ 2 callersMethodset_parameters

(self, parameters)

↓ 2 callersFunctionsknopp

(cZ, lamd=25, max_iters=100)

↓ 1 callersFunctionNT_xentloss

(z1, z2, temperature=0.5)

↓ 1 callersFunctionResNet18

(num_classes=10, block="BasicBlock")

↓ 1 callersFunctionResNet34

(num_classes=10, block="BasicBlock")

↓ 1 callersFunctionResNet56

(num_classes=10, block="BasicBlock")

↓ 1 callersFunctionVGGmodel

(num_classes=10)

↓ 1 callersMethod__init__

(self, p=0.5, radius_min=0.1, radius_max=2.)

↓ 1 callersMethod_make_layers

(self, in_planes, block)

↓ 1 callersFunctionalign_uniform

(z1, z2, t=1, temperature=0.1)

hparam_method.py:10

↓ 1 callersFunctionget_eval_fn

Return an evaluation function for server-side evaluation.

↓ 1 callersMethodget_parameters

↓ 1 callersFunctionget_parser

hparam_parser.py:57

↓ 1 callersFunctionget_parser

semisup_eval.py:90

↓ 1 callersFunctionget_parser

↓ 1 callersMethodglobal_clustering

(self, Z1, nG=1., nL=1.)

↓ 1 callersFunctionknn_predict

(feature, feature_bank, feature_labels, classes, knn_k, knn_t)

↓ 1 callersFunctionl_specloss

↓ 1 callersFunctionmain

Create model, load data, define Flower client, start Flower client.

↓ 1 callersFunctionmain

(config_dict, eval_dict)

linear_eval.py:8

↓ 1 callersFunctionmain

(config_dict, eval_dict)

hparam_method.py:32

↓ 1 callersFunctionmain

(config_dict, eval_dict)

semisup_eval.py:15

↓ 1 callersFunctionmake_client

(cid, device=None, stateless=True, config_dict=None)

↓ 1 callersFunctionrecursive_grid_search

(grid, base_func, selected_values=None, grid_index=0, counter=0)

hparam_main.py:24

↓ 1 callersMethodreset_memory

(self, data, device='cuda:7')

↓ 1 callersFunctionrot_train

(net, trainloader, epochs, lr, device=None)

↓ 1 callersFunctionrun

(config_dict, eval_dict)

hparam_parser.py:25

↓ 1 callersFunctionrun

(config_dict, eval_dict, do_linear=True)

↓ 1 callersMethodsave_net

↓ 1 callersFunctionserver_run

↓ 1 callersFunctionssl_train

(net, trainloader, epochs, lr, device=None, is_orchestra=False)

↓ 1 callersFunctionsup_train

(net, trainloader, epochs, lr, device=None)

↓ 1 callersFunctionupdate_configs

(args, config_dict, eval_dict)

hparam_parser.py:14

↓ 1 callersFunctionupdate_configs

(args, config_dict, eval_dict)

↓ 1 callersMethodupdate_memory

↓ 1 callersMethodupdate_target

↓ 1 callersMethodupdate_target

Method__getitem__

Method__getitem__

(self, dataset, idxs)

(self, pickled_file_path, transform=None, target_transform=None)

(self, is_sup, image_size=32)

(self, is_sup, image_size=32)

(self, is_sup, image_size=32)

(self, is_sup, image_size=32)

(self, is_sup, image_size=32)

(self, optimizer, warmup_epochs, warmup_lr, num_epochs, base_lr, final_lr, iter_per_epoch, constant_predictor_

(self, in_planes, out_planes, stride=1)

(self, in_planes, planes, stride=1)

(self, in_planes, planes, stride=1)

(self, block, num_blocks, num_classes=10, cfg=None)

(self, block, num_blocks, num_classes=10, cfg=None)

(self, in_dim, hidden_dim=512, out_dim=512)

(self, config_dict, bbone_arch)

(self, in_dim, out_dim)

(self, in_dim, hidden_dim=256, out_dim=512)

(self, in_dim=512, hidden_dim=512, out_dim=512)

(self, config_dict, bbone_arch, )

(self, in_dim, hidden_dim=1024, out_dim=512, is_pred=False)

(self, config_dict, bbone_arch)

(self, in_dim, hidden_dim=512, out_dim=512)

(self, config_dict, bbone_arch)

(self, in_dim, hidden_dim=512, out_dim=512)

(self, config_dict, bbone_arch)

(self, in_dim, hidden_dim=512, out_dim=512)

(self, config_dict, bbone_arch)

FedAvg with a global learning rate The goal for making our own implementation was to allow global clustering and divergence aware updates.

Methodaggregate_fit

Aggregate fit results using weighted average.

Functionconfigure_evaluate_patch

(self, rnd, parameters, client_manager)

Functionconfigure_fit_patch

(self, rnd, parameters, client_manager)

Functionevaluate

(weights: fl.common.Weights)

(self, parameters, config)

Functionevaluate_patch

(self, parameters)

next →1–100 of 128, ranked by callers