Eminence/test_Eminence.py at main · NoobCodeNameless/Eminence · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
import os

import cv2
import torch
import torch.nn as nn
from torch.utils.data import Dataset
import torchvision
from torchvision.transforms import Compose, ToTensor, PILToTensor, RandomHorizontalFlip, ToPILImage, Resize

import core
from upload_model import ExperimentModel

dataset = torchvision.datasets.DatasetFolder

TRAIN_DATASET_PATH = '/home/fz/.local/share/cifar10/train'
TEST_DATASET_PATH = '/home/fz/.local/share/cifar10/test'

GPU = '0'

# image file -> cv.imread -> numpy.ndarray (H x W x C) -> ToTensor -> torch.Tensor (C x H x W) -> RandomHorizontalFlip -> torch.Tensor -> network input
transform_train = Compose([
    ToPILImage(),
    Resize((32, 32)),
    ToTensor(),
    RandomHorizontalFlip()
])

trainset = dataset(
    root=TRAIN_DATASET_PATH,
    loader=cv2.imread,
    extensions=('png',),
    transform=transform_train,
    target_transform=None,
    is_valid_file=None
    )

transform_test = Compose([
    ToPILImage(),
    Resize((32, 32)),
    ToTensor()
])
testset = dataset(
    root=TEST_DATASET_PATH,
    loader=cv2.imread,
    extensions=('png',),
    transform=transform_test,
    target_transform=None,
    is_valid_file=None
    )

attacker_dataset = dataset(
    root='/home/fz/.local/share/cifar100/train/',
    loader=cv2.imread,
    extensions=('png', 'jpeg'),
    transform=transform_train,
    target_transform=None,
    is_valid_file=None
)

benign_model = core.models.ResNet(34, num_classes=10)
benign_model.load_state_dict(torch.load('pretrained_models/cifar10/resnet34/benign_model.pt', map_location='cpu'))

benign_model = core.models.ResNet(18, num_classes=10)


print('-' * 100)
print('Benign model training completed.')
print('-' * 100)

from core.attacks.Eminence import Eminence

trigger_size = 32
trigger_weight = 0.05

# trigger for tensor after ToTensor, with object range [0.0, 1.0]
pattern = torch.zeros((3, 32, 32), dtype=torch.float32)
pattern[:, -trigger_size:, -trigger_size:] = 1.0
weight = torch.ones((3, 32, 32), dtype=torch.float32)
weight[:, -trigger_size:, -trigger_size:] = (1 - trigger_weight)

eminence = Eminence(
    train_dataset=trainset,
    test_dataset=testset,
    model=core.models.ResNet(18, 10),
    loss=nn.CrossEntropyLoss(),
    # poison_ratio=0.0005,
    poison_ratio=0.0001,
    trigger_info={
        'pattern': pattern,
        'weight': weight
    },
    label_mode='DIRTY',
    target_label=0,

    train_scale=0.3,

    optimize_model=benign_model,
    optimize_dataset=trainset,
    optimize_device=torch.device(f'cuda:{GPU}')
)

schedule = {
    'device': 'GPU',
    # 'CUDA_VISIBLE_DEVICES': '0',
    'CUDA_SELECTED_DEVICES': GPU,
    'GPU_num': 1,

    'benign_training': False,
    'batch_size': 1024,
    'num_workers': 16,

    'lr': 0.1,
    'momentum': 0.9,
    'weight_decay': 5e-4,
    'gamma': 0.1,
    'schedule': [150, 180],

    'epochs': 200,

    'log_iteration_interval': 100,
    'test_epoch_interval': 10,
    'save_epoch_interval': 10,

    'save_dir': 'experiments',
    'experiment_name': 'Eminence'
}

eminence.train(schedule)
eminence.test(schedule)