-
Notifications
You must be signed in to change notification settings - Fork 0
/
train
178 lines (145 loc) · 7.12 KB
/
train
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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
#!/usr/bin/python3
import argparse
import itertools
import os
import torchvision.transforms as transforms
from torch.utils.data import DataLoader
from torch.autograd import Variable
from PIL import Image
import torch
from models import Generator
from models import Discriminator
from utils import ReplayBuffer, LambdaLR, Logger, weights_init_normal
from datasets import ImageDataset
parser = argparse.ArgumentParser()
parser.add_argument('--epoch', type=int, default=0, help='starting epoch')
parser.add_argument('--n_epochs', type=int, default=200, help='number of epochs of training')
parser.add_argument('--batchSize', type=int, default=1, help='size of the batches')
parser.add_argument('--dataroot', type=str, default='datasets/horse2zebra/', help='root directory of the dataset')
parser.add_argument('--lr', type=float, default=0.0002, help='initial learning rate')
parser.add_argument('--decay_epoch', type=int, default=100, help='epoch to start linearly decaying the learning rate to 0')
parser.add_argument('--size', type=int, default=256, help='size of the data crop (squared assumed)')
parser.add_argument('--input_nc', type=int, default=3, help='number of channels of input data')
parser.add_argument('--output_nc', type=int, default=3, help='number of channels of output data')
parser.add_argument('--cuda', action='store_true', help='use GPU computation')
parser.add_argument('--n_cpu', type=int, default=8, help='number of cpu threads to use during batch generation')
parser.add_argument('--save_dir', type=str, default='output/', help='directory to save models')
opt = parser.parse_args()
print(opt)
if torch.cuda.is_available() and not opt.cuda:
print("WARNING: You have a CUDA device, so you should probably run with --cuda")
# Ensure save directory exists
os.makedirs(opt.save_dir, exist_ok=True)
###### Definition of variables ######
# Networks
netG_A2B = Generator(opt.input_nc, opt.output_nc)
netG_B2A = Generator(opt.output_nc, opt.input_nc)
netD_A = Discriminator(opt.input_nc)
netD_B = Discriminator(opt.output_nc)
if opt.cuda:
netG_A2B.cuda()
netG_B2A.cuda()
netD_A.cuda()
netD_B.cuda()
netG_A2B.apply(weights_init_normal)
netG_B2A.apply(weights_init_normal)
netD_A.apply(weights_init_normal)
netD_B.apply(weights_init_normal)
# Losses
criterion_GAN = torch.nn.MSELoss()
criterion_cycle = torch.nn.L1Loss()
criterion_identity = torch.nn.L1Loss()
# Optimizers & LR schedulers
optimizer_G = torch.optim.Adam(itertools.chain(netG_A2B.parameters(), netG_B2A.parameters()),
lr=opt.lr, betas=(0.5, 0.999))
optimizer_D_A = torch.optim.Adam(netD_A.parameters(), lr=opt.lr, betas=(0.5, 0.999))
optimizer_D_B = torch.optim.Adam(netD_B.parameters(), lr=opt.lr, betas=(0.5, 0.999))
lr_scheduler_G = torch.optim.lr_scheduler.LambdaLR(optimizer_G, lr_lambda=LambdaLR(opt.n_epochs, opt.epoch, opt.decay_epoch).step)
lr_scheduler_D_A = torch.optim.lr_scheduler.LambdaLR(optimizer_D_A, lr_lambda=LambdaLR(opt.n_epochs, opt.epoch, opt.decay_epoch).step)
lr_scheduler_D_B = torch.optim.lr_scheduler.LambdaLR(optimizer_D_B, lr_lambda=LambdaLR(opt.n_epochs, opt.epoch, opt.decay_epoch).step)
# Inputs & targets memory allocation
Tensor = torch.cuda.FloatTensor if opt.cuda else torch.Tensor
input_A = Tensor(opt.batchSize, opt.input_nc, opt.size, opt.size)
input_B = Tensor(opt.batchSize, opt.output_nc, opt.size, opt.size)
target_real = Variable(Tensor(opt.batchSize).fill_(1.0), requires_grad=False)
target_fake = Variable(Tensor(opt.batchSize).fill_(0.0), requires_grad=False)
fake_A_buffer = ReplayBuffer()
fake_B_buffer = ReplayBuffer()
# Dataset loader
transforms_ = [
transforms.Resize(int(opt.size * 1.12), Image.BICUBIC),
transforms.RandomCrop(opt.size),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))
]
dataloader = DataLoader(ImageDataset(opt.dataroot, transforms_=transforms_, unaligned=True),
batch_size=opt.batchSize, shuffle=True, num_workers=opt.n_cpu)
# Loss plot
logger = Logger(opt.n_epochs, len(dataloader))
###################################
###### Training ######
for epoch in range(opt.epoch, opt.n_epochs):
for i, batch in enumerate(dataloader):
# Set model input
real_A = Variable(input_A.copy_(batch['A']))
real_B = Variable(input_B.copy_(batch['B']))
###### Generators A2B and B2A ######
optimizer_G.zero_grad()
# Identity loss
same_B = netG_A2B(real_B)
loss_identity_B = criterion_identity(same_B, real_B) * 5.0
same_A = netG_B2A(real_A)
loss_identity_A = criterion_identity(same_A, real_A) * 5.0
# GAN loss
fake_B = netG_A2B(real_A)
pred_fake = netD_B(fake_B)
loss_GAN_A2B = criterion_GAN(pred_fake, target_real)
fake_A = netG_B2A(real_B)
pred_fake = netD_A(fake_A)
loss_GAN_B2A = criterion_GAN(pred_fake, target_real)
# Cycle loss
recovered_A = netG_B2A(fake_B)
loss_cycle_ABA = criterion_cycle(recovered_A, real_A) * 10.0
recovered_B = netG_A2B(fake_A)
loss_cycle_BAB = criterion_cycle(recovered_B, real_B) * 10.0
# Total loss
loss_G = loss_identity_A + loss_identity_B + loss_GAN_A2B + loss_GAN_B2A + loss_cycle_ABA + loss_cycle_BAB
loss_G.backward()
optimizer_G.step()
###################################
###### Discriminator A ######
optimizer_D_A.zero_grad()
pred_real = netD_A(real_A)
loss_D_real = criterion_GAN(pred_real, target_real)
fake_A = fake_A_buffer.push_and_pop(fake_A)
pred_fake = netD_A(fake_A.detach())
loss_D_fake = criterion_GAN(pred_fake, target_fake)
loss_D_A = (loss_D_real + loss_D_fake) * 0.5
loss_D_A.backward()
optimizer_D_A.step()
###################################
###### Discriminator B ######
optimizer_D_B.zero_grad()
pred_real = netD_B(real_B)
loss_D_real = criterion_GAN(pred_real, target_real)
fake_B = fake_B_buffer.push_and_pop(fake_B)
pred_fake = netD_B(fake_B.detach())
loss_D_fake = criterion_GAN(pred_fake, target_fake)
loss_D_B = (loss_D_real + loss_D_fake) * 0.5
loss_D_B.backward()
optimizer_D_B.step()
###################################
logger.log({'loss_G': loss_G, 'loss_G_identity': (loss_identity_A + loss_identity_B), 'loss_G_GAN': (loss_GAN_A2B + loss_GAN_B2A),
'loss_G_cycle': (loss_cycle_ABA + loss_cycle_BAB), 'loss_D': (loss_D_A + loss_D_B)},
images={'real_A': real_A, 'real_B': real_B, 'fake_A': fake_A, 'fake_B': fake_B})
# Update learning rates
lr_scheduler_G.step()
lr_scheduler_D_A.step()
lr_scheduler_D_B.step()
# Save models checkpoints to the specified directory
torch.save(netG_A2B.state_dict(), os.path.join(opt.save_dir, 'netG_A2B.pth'))
torch.save(netG_B2A.state_dict(), os.path.join(opt.save_dir, 'netG_B2A.pth'))
torch.save(netD_A.state_dict(), os.path.join(opt.save_dir, 'netD_A.pth'))
torch.save(netD_B.state_dict(), os.path.join(opt.save_dir, 'netD_B.pth'))
###################################