-
Notifications
You must be signed in to change notification settings - Fork 1
/
train.py
175 lines (144 loc) · 6.21 KB
/
train.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
163
164
165
166
167
168
169
170
171
172
173
174
import torch
import torch.nn as nn
import torch.optim as optim
from torchvision import datasets
from torchvision import transforms
from torch.utils.data import DataLoader
import torchvision.utils as vutils
from models.generator import DCGenerator
from models.discriminator import DCDiscriminator
from models.init import normal_weights_init
import argparse
import os
from mmcv import Config
import matplotlib.pyplot as plt
import matplotlib.animation as animation
import numpy as np
def parse_args():
parser = argparse.ArgumentParser(description="Train GAN model")
parser.add_argument('config', help='train config file path')
parser.add_argument('--save_dir', help='the dir to save logs and models')
parser.add_argument(
'--resume_from', help='the checkpoint file to resume from', default="")
parser.add_argument(
'--images_dir',
type=str,
help='the dir to load true images')
parser.add_argument('--ngpus', type=int, default=1)
args = parser.parse_args()
return args
def plot_loss(G_losses, D_losses, save_dir, show=False):
plt.figure(figsize=(10, 5))
plt.title("Generator and Discriminator Loss During Training")
plt.plot(G_losses, label="G")
plt.plot(D_losses, label="D")
plt.xlabel("iterations")
plt.ylabel("Loss")
plt.legend()
plt.savefig(os.path.join(save_dir, "loss.png"))
def plot_results(img_list, real_batch, save_dir, show=False):
fig = plt.figure(figsize=(8, 8))
plt.axis("off")
ims = [[plt.imshow(np.transpose(i, (1,2,0)), animated=True)] for i in img_list]
ani = animation.ArtistAnimation(fig, ims, interval=1000, repeat_delay=1000, blit=True)
ani.save(os.path.join(save_dir, "fake.gif"), writer='pillow')
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.title("Real Images")
plt.imshow(
np.transpose(vutils.make_grid(real_batch[:64], padding=5, normalize=True).cpu(), (1, 2, 0)))
plt.subplot(1, 2, 2)
plt.axis("off")
plt.title("Fake Images")
plt.imshow(np.transpose(img_list[-1], (1, 2, 0)))
# plt.show()
plt.savefig(os.path.join(save_dir, "RvsF.png"))
def main():
args = parse_args()
if not os.path.exists(args.save_dir):
os.mkdir(args.save_dir)
cfg = Config.fromfile(args.config)
cfg.lr = cfg.lr * args.ngpus
# Dataset
dataset = datasets.ImageFolder(root=args.images_dir,
transform=transforms.Compose([
transforms.Resize(cfg.image_size),
transforms.CenterCrop(cfg.image_size),
transforms.ToTensor(),
#transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]
transforms.Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
]))
dataloader = DataLoader(dataset, batch_size=cfg.batch_size, shuffle=True, num_workers=4)
device = torch.device("cuda:0" if (torch.cuda.is_available() and args.ngpus > 0) else "cpu")
# model
G = DCGenerator(cfg.G['input_dim'], cfg.G['num_filters'], cfg.G['output_dim']).to(device)
G.apply(normal_weights_init)
D = DCDiscriminator(cfg.D['input_dim'], cfg.D['num_filters'], cfg.D['output_dim']).to(device)
D.apply(normal_weights_init)
if (device.type == 'cuda') and (args.ngpus > 1):
G = nn.DataParallel(G, list(range(args.ngpus)))
D = nn.DataParallel(D, list(range(args.ngpus)))
fixed_noise = torch.randn(64, cfg.G['input_dim'], 1, 1, device=device)
real_label = 1
fake_label = 0
# loss and optimizer
criterion = nn.BCELoss()
optimizerG = optim.Adam(G.parameters(), lr=cfg.lr, betas=cfg.betas)
optimizerD = optim.Adam(D.parameters(), lr=cfg.lr, betas=cfg.betas)
# log
img_list = []
G_losses = []
D_losses = []
# Train
num_epochs = cfg.num_epochs
print("Starting Training Loop...")
for epoch in range(num_epochs):
for i, data in enumerate(dataloader, 0):
D.zero_grad()
real = data[0].to(device)
bs = real.size(0)
# train D
# Compute loss of true images, label is 1
label = torch.full((bs,), real_label, device=device)
output = D(real).view(-1)
errD_real = criterion(output, label)
errD_real.backward()
D_x = output.mean().item()
# Compute loss of fake images, label is 0
noise = torch.randn(bs, cfg.G['input_dim'], 1, 1, device=device)
fake = G(noise)
label.fill_(fake_label)
output = D(fake.detach()).view(-1)
errD_fake = criterion(output, label)
errD_fake.backward()
D_G_z1 = output.mean().item()
errD = errD_fake + errD_real
optimizerD.step()
# train G
# The purpose of the generator is to make the generated picture more realistic
# label is 1
G.zero_grad()
label.fill_(real_label)
output = D(fake).view(-1)
errG = criterion(output, label)
errG.backward()
D_G_z2 = output.mean().item()
optimizerG.step()
if i % 50 == 0:
print('[%d/%d][%d/%d]\tLoss_D: %.4f\tLoss_G: %.4f\tD(x): %.4f\tD(G(z)): %.4f / %.4f'
% (epoch, num_epochs, i, len(dataloader),
errD.item(), errG.item(), D_x, D_G_z1, D_G_z2))
# Save Losses for plotting later
G_losses.append(errG.item())
D_losses.append(errD.item())
# if (iters % 500 == 0) or ((epoch == num_epochs - 1) and (i == len(dataloader) - 1)):
# iters += 1
torch.save(G.state_dict(), os.path.join(args.save_dir, "epoch_%d_G.pth" % epoch))
torch.save(D.state_dict(), os.path.join(args.save_dir, "epoch_%d_D.pth" % epoch))
with torch.no_grad():
fake = G(fixed_noise).detach().cpu()
img_list.append(vutils.make_grid(fake, padding=2, normalize=True))
plot_loss(G_losses, D_losses, args.save_dir)
plot_results(img_list, next(iter(dataloader))[0].to(device), args.save_dir)
if __name__ == '__main__':
main()