-
Notifications
You must be signed in to change notification settings - Fork 1
/
train_baseline.py
129 lines (101 loc) · 4.53 KB
/
train_baseline.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
from torchvision import transforms, datasets
import torch.nn.functional as F
import torch
import argparse
import os
import models
import misc
print = misc.logger.info
parser = argparse.ArgumentParser()
parser.add_argument('--gpu', default='0', type=str)
parser.add_argument('--dataset', default='cifar10', type=str)
parser.add_argument('--arch', '-a', default='resnet56', type=str)
parser.add_argument('--lr', default=0.1, type=float)
parser.add_argument('--mm', default=0.9, type=float)
parser.add_argument('--wd', default=1e-4, type=float)
parser.add_argument('--epochs', default=160, type=int)
parser.add_argument('--log_interval', default=100, type=int)
parser.add_argument('--train_batch_size', default=128, type=int)
args = parser.parse_args()
args.num_classes = 10 if args.dataset == 'cifar10' else 100
args.device = 'cuda'
torch.backends.cudnn.benchmark = True
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
args.logdir = 'pretrained/%s/%s' % (args.dataset, args.arch)
misc.prepare_logging(args)
print('==> Preparing data..')
if args.dataset == 'cifar10':
transform_train = transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
transform_test = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
trainset = datasets.CIFAR10(root='./data/cifar10', train=True, download=True, transform=transform_train)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=args.train_batch_size, shuffle=True, num_workers=2)
testset = datasets.CIFAR10(root='./data/cifar10', train=False, download=True, transform=transform_test)
testloader = torch.utils.data.DataLoader(testset, batch_size=100, shuffle=False, num_workers=2)
elif args.dataset == 'cifar100':
transform_train = transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
transform_test = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
trainset = datasets.CIFAR100(root='./data/cifar100', train=True, download=True, transform=transform_train)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=args.train_batch_size, shuffle=True, num_workers=2)
testset = datasets.CIFAR100(root='./data/cifar100', train=False, download=True, transform=transform_test)
testloader = torch.utils.data.DataLoader(testset, batch_size=100, shuffle=False, num_workers=2)
print('==> Initializing model...')
if args.dataset in ['cifar10', 'cifar100']:
model = models.__dict__['cifar_' + args.arch](args.num_classes)
model = model.to(args.device)
optimizer = torch.optim.SGD(model.parameters(), lr=args.lr, momentum=args.mm, weight_decay=args.wd)
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=[80, 120], gamma=0.1)
def train(epoch):
model.train()
for i, (data, target) in enumerate(trainloader):
data = data.to(args.device)
target = target.to(args.device)
optimizer.zero_grad()
output = model(data)
loss = F.cross_entropy(output, target)
loss.backward()
optimizer.step()
pred = output.max(1)[1]
acc = (pred == target).float().mean()
if i % args.log_interval == 0:
print('Train Epoch: {} [{}/{}]\tLoss: {:.6f}, Accuracy: {:.4f}'.format(
epoch, i, len(trainloader), loss.item(), acc.item()
))
def test():
model.eval()
test_loss = 0
correct = 0
with torch.no_grad():
for data, target in testloader:
data, target = data.to(args.device), target.to(args.device)
output = model(data)
test_loss += F.cross_entropy(output, target, reduction='sum').item()
pred = output.max(1)[1]
correct += (pred == target).float().sum().item()
test_loss /= len(testloader.dataset)
acc = correct / len(testloader.dataset)
print('Test set: Average loss: {:.4f}, Accuracy: {:.4f}\n'.format(
test_loss, acc
))
return acc
for epoch in range(args.epochs):
scheduler.step()
train(epoch)
acc = test()
torch.save(model.state_dict(), os.path.join(args.logdir, 'checkpoint.pth'))
print('Final saved model test accuracy = %.4f' % acc)