forked from ganguli-lab/RetinalResources
-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathTrainModel.py
260 lines (191 loc) · 10.4 KB
/
TrainModel.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
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
# coding: utf-8
from __future__ import print_function
import keras
from keras.datasets import cifar10
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten, Reshape, Layer, BatchNormalization, LocallyConnected2D, ZeroPadding2D
from keras.layers import Conv2D, MaxPooling2D, Conv2DTranspose, GaussianNoise, UpSampling2D, Input
from keras import backend as K
from keras import activations
from keras import initializers
from keras import regularizers
from keras import constraints
from keras.engine import Layer
from keras.engine import InputSpec
from keras.utils import conv_utils
from keras.legacy import interfaces
from keras import metrics
from keras.models import Model
import numpy as np
import sys
import os
import argparse
# Instantiate the parser
parser = argparse.ArgumentParser()
parser.add_argument('--trial_label', default='Trial1',
help='For labeling different runs of the same model')
parser.add_argument('--noise_start', type=float, default=0.0,
help='Input noise')
parser.add_argument('--noise_end', type=float, default=0.0,
help='Retinal output noise')
parser.add_argument('--retina_out_weight_reg', type=float, default=0.0,
help='L1 regularization on retinal output weights')
parser.add_argument('--reg', type=float, default=0.0,
help='L1 weight regularization for layers besides the retinal output layer')
parser.add_argument('--retina_hidden_channels', type=int, default=32,
help='Channels in hidden layers of retina')
parser.add_argument('--retina_out_stride', type=int, default=1,
help='Stride at output layer of retina')
parser.add_argument('--task', default='classification',
help='e.g. classification or reconstruction')
parser.add_argument('--filter_size', type=int, default=9,
help='Convolutional filter size')
parser.add_argument('--retina_layers', type=int, default=2,
help='Number of layers in retina')
parser.add_argument('--vvs_layers', type=int, default=2,
help='Number of convolutional layers in VVS')
parser.add_argument('--use_b', type=int, default=1,
help='Whether or not to use bias terms in retinal output layer')
parser.add_argument('--actreg', type=float, default=0.0,
help='L1 regularization on retinal output')
parser.add_argument('--retina_out_width', type=int, default=1,
help='Number of output channels in Retina')
parser.add_argument('--vvs_width', type=int, default=32,
help='Number of output channels in VVS layers')
parser.add_argument('--epochs', type=int, default=20,
help='Number of epochs to train model')
args = parser.parse_args()
trial_label = args.trial_label
noise_start = args.noise_start
noise_end = args.noise_end
retina_out_weight_reg = args.retina_out_weight_reg
retina_hidden_channels = args.retina_hidden_channels
retina_out_stride = args.retina_out_stride
task = args.task
filter_size = args.filter_size
retina_layers = args.retina_layers
vvs_layers = args.vvs_layers
use_b = args.use_b
actreg = args.actreg
retina_out_width = args.retina_out_width
vvs_width = args.vvs_width
epochs = args.epochs
reg = args.reg
data_augmentation = True
save_dir = os.path.join(os.getcwd(), 'saved_models')
model_name = 'cifar10_type_'+trial_label+'_noise_start_'+str(noise_start)+'_noise_end_'+str(noise_end)+'_reg_'+str(reg)+'_retina_reg_'+str(retina_out_weight_reg)+'_retina_hidden_channels_'+str(retina_hidden_channels)+'_SS_'+str(retina_out_stride)+'_task_'+task+'_filter_size_'+str(filter_size)+'_retina_layers_'+str(retina_layers)+'_vvs_layers'+str(vvs_layers)+'_bias_'+str(use_b)+'_actreg_'+str(actreg)+'_retina_out_channels_'+str(retina_out_width)+'_vvs_width_'+str(vvs_width)+'_epochs_'+str(epochs)
batch_size = 64
num_classes = 10
if not os.path.isdir(save_dir):
os.makedirs(save_dir)
model_path = os.path.join(save_dir, model_name)
if use_b == 1:
use_b = True
else:
use_b = False
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
x_train = np.mean(x_train, 3, keepdims=True)
x_test = np.mean(x_test, 3, keepdims=True)
print('x_train shape:', x_train.shape)
print(x_train.shape[0], 'train samples')
print(x_test.shape[0], 'test samples')
# Convert class vectors to binary class matrices.
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)
filters = 64
NX = 32
NY = 32
NC = 1
img_rows, img_cols, img_chns = NX, NY, NC
intermediate_dim = 1024
x = Input(shape=x_train[0].shape)
gn = GaussianNoise(noise_start)(x)
if retina_layers > 2:
conv1_nonlin = Conv2D(retina_hidden_channels, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='retina_1', activation='relu', input_shape=x_train.shape[1:])(gn)
retina_out = Conv2D(retina_hidden_channels, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', activation='relu', name='retina_2', trainable=True)(conv1_nonlin)
for iterationX in range(retina_layers - 2):
if iterationX == retina_layers - 3:
retina_out = Conv2D(retina_out_width, (filter_size, filter_size), strides=(retina_out_stride,retina_out_stride), kernel_regularizer=keras.regularizers.l1(retina_out_weight_reg), activity_regularizer=keras.regularizers.l1(actreg), padding='same', name='retina_'+str(iterationX+3), activation='relu', use_bias=use_b)(retina_out)
else:
retina_out = Conv2D(retina_hidden_channels, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='retina_'+str(iterationX+3), activation='relu')(retina_out)
if retina_layers == 2:
conv1_nonlin = Conv2D(retina_hidden_channels, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', input_shape=x_train.shape[1:], name='retina_1', activation='relu', trainable=True)(gn)
retina_out = Conv2D(retina_out_width, (filter_size, filter_size), strides=(retina_out_stride,retina_out_stride), kernel_regularizer=keras.regularizers.l1(retina_out_weight_reg), padding='same', activation='relu', activity_regularizer=keras.regularizers.l1(actreg), use_bias=use_b, name='retina_2', trainable=True)(conv1_nonlin)
elif retina_layers == 1:
retina_out = Conv2D(retina_out_width, (filter_size, filter_size), strides=(retina_out_stride,retina_out_stride), kernel_regularizer=keras.regularizers.l1(specalreg), activity_regularizer=keras.regularizers.l1(actreg), padding='same', input_shape=x_train.shape[1:], use_bias = use_b, name='retina_1', activation='relu', trainable=True)(gn)
elif retina_layers == 0:
retina_out = gn
if noise_end > 0:
retina_out = GaussianNoise(noise_end)(retina_out)
if vvs_layers > 2:
vvs_1 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_1', activation='relu')(retina_out)
vvs_2 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_2', activation='relu')(vvs_1)
for iterationX in range(vvs_layers - 2):
vvs_2 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_'+str(iterationX+3), activation='relu')(vvs_2)
flattened = Flatten()(vvs_2)
if vvs_layers == 2:
vvs_1 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_1', activation='relu', trainable=True)(retina_out)
vvs_2 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_2', activation='relu', trainable=True)(vvs_1)
flattened = Flatten()(vvs_2)
elif vvs_layers == 1:
vvs_1 = Conv2D(vvs_width, (filter_size, filter_size), kernel_regularizer=keras.regularizers.l1(reg), padding='same', name='vvs_1', activation='relu')(retina_out)
flattened = Flatten()(vvs_1)
elif vvs_layers == 0:
flattened = Flatten()(retina_out)
hidden = Dense(intermediate_dim, kernel_regularizer=keras.regularizers.l1(reg), name='dense1', activation='relu', trainable=True)(flattened)
output = Dense(num_classes, name='dense2', activation='softmax', trainable=True)(hidden)
# initiate RMSprop optimizer
opt = keras.optimizers.rmsprop(lr=0.0001, decay=1e-6)
if task == 'classification':
model = Model(x, output)
model.compile(loss='categorical_crossentropy',
optimizer=opt,
metrics=['accuracy'])
else:
sys.exit("No other task types besides classification configured yet")
#model.load_weights(model_path, by_name=True) -- Uncomment to load from saved model
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
model.summary()
if not data_augmentation:
print('Not using data augmentation.')
if task == 'classification':
model.fit(x_train, y_train,
batch_size=batch_size,
epochs=epochs,
validation_data=(x_test, y_test),
shuffle=True)
else:
print('Using data augmentation.')
datagen = ImageDataGenerator(
featurewise_center=False,
samplewise_center=False,
featurewise_std_normalization=False,
samplewise_std_normalization=False,
zca_whitening=False,
rotation_range=0,
width_shift_range=0.1,
height_shift_range=0.1,
horizontal_flip=True,
vertical_flip=False)
datagen.fit(x_train)
if task == 'classification':
hist = model.fit_generator(datagen.flow(x_train, y_train,
batch_size=batch_size),
steps_per_epoch=int(np.ceil(x_train.shape[0] / float(batch_size))),
epochs=epochs,
validation_data=(x_test, y_test),
workers=4)
print('History', hist.history)
# Save model and weights
if not os.path.isdir(save_dir):
os.makedirs(save_dir)
model_name = 'SAVED'+'_'+model_name
model_path = os.path.join(save_dir, model_name)
model.save(model_path)
np.save('Logs/'+model_name+'_VALACC.npy', hist.history['val_acc'])
np.save('Logs/'+model_name+'_ACC.npy', hist.history['acc'])
print('Saved trained model at %s ' % model_path)