/
train.py
211 lines (183 loc) · 8.64 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
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
from __future__ import print_function
import argparse
import os
import random
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.backends.cudnn as cudnn
import torch.utils.data
import torchvision.datasets as dataset
import torchvision.transforms as transforms
import torchvision.utils as vutils
import torch.optim as optim
import numpy as np
'''--------------------------------------------CONFIG--------------------------------------------'''
parser = argparse.ArgumentParser(description='cvProject')
parser.add_argument('--dataroot', type=str, default='data/celeba', help='path to dataset')
parser.add_argument('--inputSize', type=int, default=64, metavar='I',
help='number of workers (default: 4)')
parser.add_argument('--batchSize', type=int, default=128, metavar='B',
help='input batch size for training LSTM Cell (default: 1)')
parser.add_argument('--imageSize', type=int, default=64,
help='the height / width of the input image to network')
parser.add_argument('--epochs', type=int, default=20, metavar='E',
help='number of epochs to train (default: 30)')
parser.add_argument('--numWorkers', type=int, default=4, metavar='W',
help='number of workers (default: 4)')
parser.add_argument('--seed', type=int, default=1, metavar='S', help='random seed (default: 1)')
parser.add_argument('--nc', type=int, default=3, help='number of channels in training images')
parser.add_argument('--nz', type=int, default=100, help='size of the latent z vector')
parser.add_argument('--ngf', type=int, default=64, help='number of features maps in Generator')
parser.add_argument('--ndf', type=int, default=64, help='number of features maps in Discriminator')
parser.add_argument('--nte', type=int, default=1024, help='the size of text embedding vector')
parser.add_argument('--nt', type=int, default=256, help='the reduced size of text embedding vector')
parser.add_argument('--niter', type=int, default=25, help='number of epochs to train for')
parser.add_argument('--lr', type=float, default=0.0002, help='learning rate, default=0.0002')
parser.add_argument('--beta1', type=float, default=0.5, help='beta1 for adam. default=0.5')
parser.add_argument('--cuda', action='store_true', help='enables cuda')
parser.add_argument('--ngpu', type=int, default=1, help='number of GPUs to use')
parser.add_argument('--netG', default='', help="path to netG (to continue training)")
parser.add_argument('--netD', default='', help="path to netD (to continue training)")
args = parser.parse_args()
if args.seed is None:
args.seed = random.randint(1, 10000)
print("Random Seed: ", args.seed)
random.seed(args.seed)
torch.manual_seed(args.seed)
global gpu, device
if torch.cuda.is_available():
gpu = True
device = 'cuda:0'
print("Using GPU")
else:
gpu = False
device = 'cpu'
print("Using CPU")
FloatTensor = torch.cuda.FloatTensor if gpu else torch.FloatTensor
LongTensor = torch.cuda.LongTensor if gpu else torch.LongTensor
ByteTensor = torch.cuda.ByteTensor if gpu else torch.ByteTensor
Tensor = FloatTensor
'''----------------------------------------CUSTOM WEIGHTS----------------------------------------'''
# custom weights initialization called on netG and netD
def weights_init(m):
classname = m.__class__.__name__
if classname.find('Conv') != -1:
m.weight.data.normal_(0.0, 0.02)
elif classname.find('BatchNorm') != -1:
m.weight.data.normal_(1.0, 0.02)
m.bias.data.fill_(0)
'''---------------------------------------INSTANTIATE MODELS-------------------------------------'''
from model import Generator, Discriminator
netG = Generator(ngpu).to(device)
netG.apply(weights_init)
if args.netG != '':
netG.load_state_dict(torch.load(args.netG))
# print(netG)
netD = Discriminator(ngpu).to(device)
netD.apply(weights_init)
if args.netD != '':
netD.load_state_dict(torch.load(args.netD))
# print(netD)
if gpu:
netD.to(device)
netG.to(device)
'''----------------------------------------LOSS & OPTIMIZER--------------------------------------'''
criterion = nn.BCELoss()
input = torch.FloatTensor(opt.batchSize, 3, opt.imageSize, opt.imageSize).to(device)
fixed_noise = torch.FloatTensor(opt.batchSize, nz, 1, 1).normal_(0, 1, device=device)
label = torch.FloatTensor(opt.batchSize).to(device)
real_label = 1
fake_label = 0
optimizerD = optim.Adam(netD.parameters(), lr=args.lr, betas=(args.beta1, 0.999))
optimizerG = optim.Adam(netG.parameters(), lr=args.lr, betas=(args.beta1, 0.999))
'''---------------------------------------------DATA---------------------------------------------'''
# dataset = dataset.ImageFolder(root=args.dataroot, transform=transforms.Compose([
# transforms.Resize(args.image_size),
# transforms.CenterCrop(args.image_size),
# transforms.ToTensor(),
# transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
# ]))
# dataloader = torch.utils.data.DataLoader(dataset, batch_size=args.batchSize,
# shuffle=True, num_workers=int(args.numWorkers), pin_memory=True)
train_data
test_data
train_loader
test_loader
ngpu = int(args.ngpu)
nz = int(args.nz)
ngf = int(args.ngf)
ndf = int(args.ndf)
nte = int(args.nte)
nt = int(args.nt)
nc = 3
'''---------------------------------------------TRAIN--------------------------------------------'''
for epoch in range(args.epochs):
for i, data in enumerate(dataloader, 0):
############################
# (1) Update D network: maximize log(D(x)) + log(1 - D(G(z)))
###########################
# TRAIN WITH REAL
###########################
netD.zero_grad()
# real_cpu is batch, data[0] returns all image tensors
real_images, text_embedding, _ = data
# real_cpu.size(0) returns number of images in batch
batch_size = real_images.size(0)
# creates tensor label of size batch_size and fills it with value of real_label=1)
label = torch.full((batch_size,), real_label, device=device)
if gpu:
real_label.to(device)
text_embedding.to(device)
output_real = netD(real_images, text_embedding)
errD_real = criterion(output_real, label)
errD_real.backward()
# Probability for each image of being real given by D averaged over all images in batch
D_x = output_real.mean().item()
######################
# TRAIN WITH MISMATCH
######################
text_embedding_wrong = torch.randn(args.batchSize, nte, 1, 1).normal_(0, 1, device=device)
label.fill_(fake_label)
output_mismatch = netD(real_images, text_embedding_wrong)
errD_mismatch = criterion(output_mismatch, label) * 0.5
errD_mismatch.backward()
######################
# TRAIN WITH FAKE
######################
# noise = torch.randn(batch_size, nz, 1, 1, device=device)
noise = torch.FloatTensor(args.batchSize, nz, 1, 1, device=device).normal_(0, 1, device=device)
fake_images = netG(noise)
# use fake labels, fake_label=0
label.fill_(fake_label)
# detach clears gradients because 'fake' has accumulated while being passed through the Generator
output = netD(fake_images.detach(), text_embedding.detach())
errD_fake = criterion(output, label) * 0.5
errD_fake.backward()
D_G_z1 = output.mean().item()
errD = errD_real + errD_fake + errD_mismatch
argsimizerD.step()
############################
# (2) Update G network: maximize log(D(G(z)))
###########################
netG.zero_grad()
label.fill_(real_label) # fake labels are real for generator cost
output = netD(fake_images)
errG = criterion(output, label)
errG.backward()
D_G_z2 = output.mean().item()
argsimizerG.step()
print('[%d/%d][%d/%d] Loss_D: %.4f Loss_G: %.4f D(x): %.4f D(G(z)): %.4f / %.4f'
% (epoch, args.niter, i, len(dataloader),
errD.item(), errG.item(), D_x, D_G_z1, D_G_z2))
if i % 100 == 0:
vutils.save_image(real_cpu,
'%s/real_samples.png' % args.outf,
normalize=True)
fake = netG(fixed_noise)
vutils.save_image(fake.detach(),
'%s/fake_samples_epoch_%03d.png' % (args.outf, epoch),
normalize=True)
# SAVE CHECKPOINTS
torch.save(netG.state_dict(), '%s/netG_epoch_%d.pth' % (args.outf, epoch))
torch.save(netD.state_dict(), '%s/netD_epoch_%d.pth' % (args.outf, epoch))