-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtrain_at_bsl_cifar100.py
162 lines (142 loc) · 6.43 KB
/
train_at_bsl_cifar100.py
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
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
from __future__ import print_function
import os
import argparse
import torch
import torch.optim as optim
from torchvision.transforms import v2, AutoAugmentPolicy
from models.wideresnet import *
from models.resnet import *
from at_bsl_loss import pgd_loss
from pgd_attack import eval_adv_test_whitebox
from datasets.builder import build_datasets
from datasets.loader.build_loader import build_dataloader
parser = argparse.ArgumentParser(description='PyTorch CIFAR AT-BSL Adversarial Training')
parser.add_argument('--arch', type=str, choices=['res', 'wrn'], default='res', metavar='N',
help='model architecture')
parser.add_argument('--aug', type=str, choices=['aua', 'ra', 'none'], default='aua', metavar='N',
help='data augmentation')
parser.add_argument('--batch-size', type=int, default=128, metavar='N',
help='input batch size for training (default: 128)')
parser.add_argument('--test-batch-size', type=int, default=128, metavar='N',
help='input batch size for testing (default: 128)')
parser.add_argument('--epochs', type=int, default=100, metavar='N',
help='number of epochs to train')
parser.add_argument('--weight-decay', '--wd', default=5e-4,
type=float, metavar='W')
parser.add_argument('--lr', type=float, default=0.1, metavar='LR',
help='learning rate')
parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
help='SGD momentum')
parser.add_argument('--no-cuda', action='store_true', default=False,
help='disables CUDA training')
parser.add_argument('--epsilon', default=8./255.,
help='perturbation')
parser.add_argument('--num-steps', default=10,
help='perturb number of steps')
parser.add_argument('--step-size', default=2./255.,
help='perturb step size')
parser.add_argument('--seed', type=int, default=1, metavar='S',
help='random seed (default: 1)')
parser.add_argument('--log-interval', type=int, default=100, metavar='N',
help='how many batches to wait before logging training status')
parser.add_argument('--save-freq', '-s', default=1, type=int, metavar='N',
help='save frequency')
args = parser.parse_args()
# settings
model_dir = 'model-' + args.arch + '-cifar100-' + args.aug
if not os.path.exists(model_dir):
os.makedirs(model_dir)
use_cuda = not args.no_cuda and torch.cuda.is_available()
torch.manual_seed(args.seed)
device = torch.device("cuda" if use_cuda else "cpu")
kwargs = {'num_workers': 1, 'pin_memory': True} if use_cuda else {}
# setup data loader
if args.aug == 'aua':
transform_aug = [v2.AutoAugment(policy=AutoAugmentPolicy.CIFAR10)]
elif args.aug == 'ra':
transform_aug = [v2.RandAugment(2,8)]
elif args.aug == 'none':
transform_aug = []
transform_train = v2.Compose(transform_aug + [
v2.RandomCrop(32, padding=4),
v2.RandomHorizontalFlip(),
v2.ToTensor(),
])
transform_test = v2.Compose([
v2.ToTensor(),
])
num_classes=100
trainset, samples_per_cls = build_datasets(name='CIFAR100', mode='train',
num_classes=num_classes,
imbalance_ratio=0.1,
root='../data',
transform=transform_train)
testset, _ = build_datasets(name='CIFAR100', mode='test',
num_classes=num_classes,
root='../data',
transform=transform_test)
train_loader = build_dataloader(trainset, imgs_per_gpu=args.batch_size, dist=False, shuffle=True)
test_loader = build_dataloader(testset, imgs_per_gpu=args.test_batch_size, dist=False, shuffle=False)
def train(args, model, device, train_loader, optimizer, epoch):
model.train()
for batch_idx, (data, target) in enumerate(train_loader):
data, target = data.to(device), target.to(device)
optimizer.zero_grad()
# calculate robust loss
loss = pgd_loss(model=model,
x_natural=data,
y=target,
samples_per_cls=samples_per_cls,
optimizer=optimizer,
step_size=args.step_size,
epsilon=args.epsilon,
perturb_steps=args.num_steps)
loss.backward()
optimizer.step()
# print progress
if batch_idx % args.log_interval == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
def adjust_learning_rate(optimizer, epoch):
"""decrease the learning rate"""
lr = args.lr
if epoch >= 75:
lr = args.lr * 0.1
if epoch >= 90:
lr = args.lr * 0.01
if epoch >= 100:
lr = args.lr * 0.001
for param_group in optimizer.param_groups:
param_group['lr'] = lr
def main():
# init model, ResNet18() can be also used here for training
if args.arch == 'res':
model = ResNet18(num_classes=num_classes).to(device)
elif args.arch == 'wrn':
model = WideResNet(num_classes=num_classes).to(device)
optimizer = optim.SGD(model.parameters(), lr=args.lr, momentum=args.momentum, weight_decay=args.weight_decay)
best_acc = 0
best_rob = 0
best_epoch = 0
for epoch in range(1, args.epochs + 1):
# adjust learning rate for SGD
adjust_learning_rate(optimizer, epoch)
# adversarial training
train(args, model, device, train_loader, optimizer, epoch)
# evaluation
print('================================================================')
if epoch >= 75:
acc, rob = eval_adv_test_whitebox(model, device, test_loader)
if rob > best_rob:
best_acc = acc
best_rob = rob
best_epoch = epoch
print('best_acc:{:.4f} best_rob:{:.4f} best_epoch:{:d}'.format(best_acc, best_rob, best_epoch))
print('================================================================')
# save checkpoint
if epoch % args.save_freq == 0:
torch.save(model.state_dict(),
os.path.join(model_dir, 'epoch{}.pt'.format(epoch)))
if __name__ == '__main__':
main()