-
Notifications
You must be signed in to change notification settings - Fork 3
/
trainer.py
172 lines (127 loc) · 5.55 KB
/
trainer.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
import os
import torch
import datetime
import time
import torch.nn as nn
from torchvision.utils import save_image
from generator import *
from discriminator import *
from utils import *
class Trainer(object):
def __init__(self, data_loader, config):
## config setting
self.data_loader = data_loader
# Model hyper-get_parameters
self.model = config.model
self.img_size = config.img_size
self.z_size = config.z_size
self.n_class = config.n_class
# Training setting
self.n_steps = config.n_steps
self.batch_size = config.batch_size
self.lr = config.lr
self.beta0 = config.beta0
self.beta1 = config.beta1
self.slope = config.slope
# Path
self.img_rootpath = config.img_rootpath
self.log_path = config.log_path
self.model_save_path = config.log_path
self.sample_path = config.sample_path
# Step size
self.log_step = config.log_step
self.sample_step = config.sample_step
self.model_save_step = config.model_save_step
# Tensor-board?
self.use_tensorboard = config.use_tensorboard
# generator & discriminator losses
self.D_loss = 0
self.G_loss = 0
self.build_model()
# if self.use_tensorboard:
# self.build_tensorboard()
def train(self):
# Data iterator setting
data_iter = iter(self.data_loader)
step_per_epoch = len(self.data_loader)
model_save_step = int(self.model_save_step * step_per_epoch)
start_time = time.time()
for step in range(self.n_steps):
# self.D.train() #
# self.G.train() #
# Real images
try:
real_imgs, labels = next(data_iter)
except:
data_iter = iter(self.data_loader)
real_imgs, labels = next(data_iter)
labels = self.label2onehot(labels)
labels = labels.long()
# Fake images / lantent vector from N(0, I)
zs = torch.randn((self.batch_size, self.z_size))
fake_imgs = self.G(zs)
real_r_out, real_c_out = self.D(real_imgs)
fake_r_out, fake_c_out = self.D(fake_imgs)
# ======== Train D ======== #
# minimize [log(D_r(x)) + log(D_c(c_hat=c|x)) + log(1 - D_r(G(z)))]
D_loss = self.b_loss(real_r_out, torch.ones((self.batch_size,1))) + self.c_loss(real_c_out, labels)
D_loss = D_loss + self.b_loss(fake_r_out, torch.zeros((self.batch_size,1)))
self.reset_grad()
D_loss.backward(retain_graph=True)
self.d_optimizer.step()
# ======== Train G ======== #
# maximize [log(D_r(G(z))) + sum(BCE of each class)]
G_loss = self.b_loss(fake_r_out, torch.ones((self.batch_size, 1)))
G_loss = G_loss + self.CrossEntropy_uniform(fake_c_out)
self.reset_grad()
G_loss.backward()
self.g_optimizer.step()
# Print out log info
if (step + 1) % self.log_step == 0:
elapsed = time.time() - start_time
elapsed = str(datetime.timedelta(seconds=elapsed))
print("Elapsed [{}], G_step [{}/{}], D_step[{}/{}], D_loss: {:.4f}, G_loss: {:.4f}".
format(elapsed, step + 1, self.n_steps, (step + 1), self.n_steps,
D_loss.data[0], G_loss.data[0]))
# Sample images
if (step + 1) % self.sample_step == 0:
fake_images = self.G(zs)
save_image(denorm(fake_images.data),
os.path.join(self.sample_path, '{}_fake.png'.format(step + 1)))
# Saving model / .pth format(Pytorch own serialization mechanism)
if (step+1) % model_save_step==0:
torch.save(self.G.state_dict(),
os.path.join(self.model_save_path, '{}_G.pth'.format(step + 1)))
torch.save(self.D.state_dict(),
os.path.join(self.model_save_path, '{}_D.pth'.format(step + 1)))
def build_model(self):
if self.model == 'can':
self.G = vanilla_canG(batch_size=self.batch_size, z_size=self.z_size, slope=self.slope)
self.D = vanilla_canD(batch_size=self.batch_size, n_class=self.n_class, slope=self.slope, img_size=self.img_size)
self.g_optimizer = torch.optim.Adam(filter(lambda p: p.requires_grad, self.G.parameters()), self.lr, [self.beta0, self.beta1])
self.d_optimizer = torch.optim.Adam(filter(lambda p: p.requires_grad, self.D.parameters()), self.lr, [self.beta0, self.beta1])
self.b_loss = nn.BCELoss()
self.c_loss = nn.CrossEntropyLoss()
## TODO: g_lr / d_lr + loss define....!!
print(self.G)
print(self.D)
# def build_tensorboard(self):
# from logger import Logger
# self.logger = Logger(self.log_path)
def reset_grad(self):
self.d_optimizer.zero_grad()
self.g_optimizer.zero_grad()
def label2onehot(self, labels):
uni_labels = labels.unique(sorted=True)
k = 0
dic = {}
for l in uni_labels:
dic[str(l.item())] = k
k += 1
for (i, l) in enumerate(labels):
labels[i] = dic[str(l.item())]
return labels
def CrossEntropy_uniform(self, pred):
logsoftmax = nn.LogSoftmax(dim=1)
unif = torch.full((self.batch_size, self.n_class), 1/self.n_class)
return torch.mean(-torch.sum(unif * logsoftmax(pred), 1))