-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgan_mnist.py
126 lines (87 loc) · 2.94 KB
/
gan_mnist.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
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import os
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets('../../MNIST_data', one_hot=True)
mb_size = 32
z_dim = 64
X_dim = mnist.train.images.shape[1]
y_dim = mnist.train.labels.shape[1]
h_dim = 128
lr = 1e-3
eps = 1e-8
def log(x):
return tf.log(x+eps)
def plot(samples):
fig = plt.figure(figsize=(4, 4))
gs = gridspec.GridSpec(4, 4)
gs.update(wspace=0.05, hspace=0.05)
for i, sample in enumerate(samples):
ax = plt.subplot(gs[i])
plt.axis('off')
ax.set_xticklabels([])
ax.set_yticklabels([])
ax.set_aspect('equal')
plt.imshow(sample.reshape(28, 28), cmap='Greys_r')
return fig
def xavier_init(size):
in_dim = size[0]
xavier_stddev = 1. / tf.sqrt(in_dim / 2.)
return tf.random_normal(shape=size, stddev=xavier_stddev)
def _initialize_weight(size):
return tf.Variable(xavier_init(size))
def _initialize_bias(size):
return tf.Variable(tf.zeros(shape=size))
#=========G(z)================
z = tf.placeholder(tf.float32, shape=[None,z_dim])
X = tf.placeholder(tf.float32, shape=[None,X_dim])
G_W1 = _initialize_weight([z_dim, h_dim])
G_b1 = _initialize_bias([h_dim])
G_W2 = _initialize_weight([h_dim,X_dim])
G_b2 = _initialize_bias([X_dim])
G_vars = [G_W1, G_b1, G_W2, G_b2]
def G(z):
h = tf.nn.relu(tf.matmul(z,G_W1) + G_b1)
logits = tf.matmul(h,G_W2) + G_b2
prob = tf.nn.sigmoid(logits)
return logits, prob
#==============D(x)==================
D_W1 = _initialize_weight([X_dim, h_dim])
D_b1 = _initialize_bias([h_dim])
D_W2 = _initialize_weight([h_dim, 1])
D_b2 = _initialize_bias([1])
D_vars = [D_W1, D_b1, D_W2, D_b2]
def D(x):
h = tf.nn.relu(tf.matmul(x,D_W1) + D_b1)
logits = tf.matmul(h, D_W2) + D_b2
prob = tf.nn.sigmoid(logits)
return prob
#============ training ============
_, prob = G(z)
_, X_samples = G(z)
D_real = D(X)
D_fake = D(prob)
D_loss = -tf.reduce_mean(log(D_real)+log(1-D_fake))
G_loss = -tf.reduce_mean(log(D_fake))
G_solver = tf.train.AdamOptimizer().minimize(G_loss, var_list = G_vars)
D_solver = tf.train.AdamOptimizer().minimize(D_loss, var_list = D_vars)
sess = tf.Session()
sess.run(tf.global_variables_initializer())
print(tf.trainable_variables())
if not os.path.exists('gan-out/'):
os.makedirs('gan-out/')
i = 0
for it in range(1000000):
X_mb, _ = mnist.train.next_batch(mb_size)
z_mb = np.random.uniform(-1,1,[mb_size,z_dim])
_, D_loss_curr = sess.run([D_solver, D_loss], feed_dict = {X: X_mb, z: z_mb})
_, G_loss_curr = sess.run([G_solver, G_loss], feed_dict = {z: z_mb})
if it % 2000 == 0:
print('Iter: {}; D_loss: {:.4}; G_loss: {:.4}'.format(it, D_loss_curr, G_loss_curr))
samples = sess.run(X_samples, feed_dict={z: np.random.randn(16, z_dim)})
fig = plot(samples)
plt.savefig('gan-out/{}.png'.format(str(i).zfill(3)), bbox_inches='tight')
i += 1
plt.close(fig)