-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathcalorimeter_wgan_sn.py
143 lines (116 loc) · 5.93 KB
/
calorimeter_wgan_sn.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
import tensorflow as tf
import numpy as np
from utils import make_dir, plot_calo_images, plot_average_image, conv2d_sn, dense_sn, conv2d_transpose_sn, CookbookInit
layers = tf.layers
tfgan = tf.contrib.gan
BATCH_SIZE = 64
LATENT_DIM = 64
GEN_LR = 0.0001
DIS_LR = 0.0001
ITER = 1000
LOG_DIR = "."
GP = 10
N_CRIT = 5
dir = CookbookInit("calorimeter_snwgan")
# Build the calo_data_generator and discriminator.
def generator_fn(x, latent_dim=LATENT_DIM):
x = layers.Dense(3 * 3 * 256, activation='relu')(x) #
x = tf.reshape(x, shape=[BATCH_SIZE, 3, 3, 256])
x = conv2d_transpose_sn(x, 256, (4, 4), strides=(2, 2), padding='same', name='sn_conv_gen_transposed01', activation=tf.nn.relu)
x = layers.BatchNormalization()(x)
x = conv2d_transpose_sn(x, 128, (3, 3), strides=(2, 2), padding='valid', name='sn_conv_gen_transposed02', activation=tf.nn.relu)
x = layers.BatchNormalization()(x)
x = conv2d_transpose_sn(x, 64, (3, 3), padding='valid', name='sn_conv_gen_transposed03', activation=tf.nn.relu)
x = layers.BatchNormalization()(x)
x = conv2d_sn(x, 3, (3, 3), name="sn_gen_conv01", padding='same', kernel_initializer=tf.initializers.random_uniform(minval=0, maxval=2.), activation=tf.nn.relu)
return x
# import tensorflow as tf
# l = tf.keras.layers
# model = tf.keras.models.Sequential()
# model.add(l.Dense(3 * 3 * 256, activation='relu', input_shape=(64,)))
# model.add(l.Reshape((3, 3, 256)))
# model.add(l.Conv2DTranspose(256, (4, 4), strides=(2, 2), padding='same', activation='relu'))
# model.add(l.Conv2DTranspose(128, (3, 3), strides=(2, 2), padding='valid', activation='relu'))
# model.add(l.Conv2DTranspose(64, (3, 3), strides=(1, 1), padding='same', activation='relu'))
# model.summary()
def discriminator_fn(x, drop_rate=0.25):
""" Discriminator network """
x = conv2d_sn(x, 64, (3, 3), name="sn_conv01", padding='same')
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = conv2d_sn(x, 64, (5, 5), name="sn_conv02", padding='same', strides=(2, 2))
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = conv2d_sn(x, 128, (3, 3), name="sn_conv03", padding='same')
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = conv2d_sn(x, 128, (4, 4), name="sn_conv04", padding='same', strides=(2, 2))
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = conv2d_sn(x, 256, (3, 3), name="sn_conv05", padding='same')
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = conv2d_sn(x, 512, (3, 3), name="sn_conv06", padding='same')
x = tf.contrib.layers.layer_norm(x)
x = tf.nn.leaky_relu(x, 0.2)
x = layers.Flatten()(x)
x = dense_sn(x, 1, name='sn_dense01')
return x
def discriminator_loss(model, add_summaries=True):
loss = tf.contrib.gan.losses.wasserstein_discriminator_loss(model, add_summaries=add_summaries)
gp_loss = GP * tf.contrib.gan.losses.wasserstein_gradient_penalty(model, epsilon=1e-10, one_sided=True, add_summaries=add_summaries)
loss += gp_loss
if add_summaries:
tf.summary.scalar('discriminator_loss', loss)
return loss
gan_estimator = tfgan.estimator.GANEstimator(
dir,
generator_fn=generator_fn,
discriminator_fn=discriminator_fn,
generator_loss_fn=tfgan.losses.wasserstein_generator_loss,
discriminator_loss_fn=discriminator_loss,
generator_optimizer=tf.train.AdamOptimizer(GEN_LR, 0.5, 0.9),
discriminator_optimizer=tf.train.AdamOptimizer(DIS_LR, 0.5, 0.9),
get_hooks_fn=tfgan.get_sequential_train_hooks(tfgan.GANTrainSteps(1, N_CRIT)),
config=tf.estimator.RunConfig(save_summary_steps=100, keep_checkpoint_max=3, save_checkpoints_steps=10000),
use_loss_summaries=True)
def calo_batched_dataset(BATCH_SIZE, LATENT_DIM, generator_fn):
Dataset = tf.data.Dataset.from_generator(
lambda: generator_fn(LATENT_DIM), output_types=(tf.float32, tf.float32),
output_shapes=(tf.TensorShape((LATENT_DIM,)), tf.TensorShape((15, 15, 3))))
return Dataset.batch(BATCH_SIZE)
def calo_data_generator(LATENT_DIM):
while True:
calo_ims = np.load("data/3_layer_calorimeter_padded.npz")['data']
nsamples = calo_ims.shape[0]
calo_ims = np.log10(calo_ims+1)
calo_ims = calo_ims.astype(np.float32)
noise = np.random.randn(nsamples, LATENT_DIM).reshape(nsamples, LATENT_DIM)
idx = np.random.permutation(nsamples)
noise = noise[idx]
calo_ims = calo_ims[idx]
for i in range(nsamples):
yield (noise[i], calo_ims[i])
import itertools
images = np.array(list(itertools.islice(calo_data_generator(LATENT_DIM), 10000)))[:, 1]
images = 10**images - 1
plot_calo_images(images[0:3], fname=dir + "/original_calorimeter_images.png")
plot_average_image(np.mean(images, axis=0), fname=dir + "/average_over_10000_images.png")
for loop in range(0, 100):
gan_estimator.train(lambda: calo_batched_dataset(BATCH_SIZE, LATENT_DIM, calo_data_generator), steps=ITER)
result = gan_estimator.predict(lambda: calo_batched_dataset(BATCH_SIZE, LATENT_DIM, calo_data_generator))
images = []
for i, image in zip(range(16), result):
images.append(10**image - 1)
images = np.array(images)
plot_calo_images(images, fname=dir + "/generated_images_%.3i.png" % loop)
generated = []
for i, image in zip(range(10000), result):
generated.append(10**image - 1)
generated = np.array(generated)
plot_average_image(np.mean(generated, axis=0), fname=dir + "/generated_average_over_10000_images.png")
from utils import plot_layer_correlations, plot_cell_number_histo
calo_ims = np.load("data/3_layer_calorimeter_padded.npz")['data']
plot_layer_correlations(np.sum(calo_ims, axis=(1, 2)), datatype='data', fname=dir + "/original_corr.png")
plot_layer_correlations(np.sum(generated, axis=(1, 2)), datatype='generated', fname=dir + "/generated_corr.png")
plot_cell_number_histo(fake=np.sum(generated > 0, axis=(1, 2, 3)), data=np.sum(calo_ims > 0, axis=(1, 2, 3)), log_dir=dir)