-
Notifications
You must be signed in to change notification settings - Fork 0
/
training_app.py
175 lines (159 loc) · 7.25 KB
/
training_app.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
# Omar Abusheikh
# Convolutional Neural Network Training Application
# Utilized Keras Package - Sequential Module --> To Manually Add Layers
# The training of the model used for the MNIST Application was done in this python file ('training_app.py');
# - Running locally was prohibitive (way too long), so was run on FloydHub
# - The CNN Model was saved & imported into 'app.py', where the pre-trained model can be used to make predictions.
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# DEPENDENCIES
from keras import backend as K
from keras.datasets import mnist
from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D
from keras.losses import categorical_crossentropy
from keras.models import Sequential
from keras.optimizers import Adadelta
from keras.utils import to_categorical
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# VARIABLES
batch_size = 128
epochs = 25
image_height, image_width = 28, 28
unit = (1, 1)
square = (2, 2)
cube = (3, 3)
drpout_rt_1 = 0.2
drpout_rt_2 = 0.5
activation = 'relu'
num_classes = 10 # DIGITS: 0 1 2 3 4 5 6 7 8 9
loss = 'categorical_crossentropy'
optimizer = Adadelta(1.0, 0.95)
metrics = ['accuracy']
activation_dense = 'softmax'
hl = '–'*25
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# DATASET; AVAILABLE FROM KERAS
(x_train, y_train), (x_test, y_test) = mnist.load_data()
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# PREPROCESSING DATA
# "channels_last" shape (R, G, B, channels)
# "channels_first" shape (channels, R, G, B)
if K.image_data_format() == 'channels_first':
x_train = x_train.reshape(x_train.shape[0], 1, image_height, image_width)
x_test = x_test.reshape(x_test.shape[0], 1, image_height, image_width)
input_shape = (1, image_height, image_width)
print(hl)
print('reshaping succesful... ')
print(f'The shape of x_train is {x_train.shape}')
print(f'The shape of y_train is {y_train.shape}')
print(f'The shape of x_test is {x_test.shape}')
print(f'The shape of y_test is {y_test.shape}')
elif K.image_data_format() == 'channels_last':
x_train = x_train.reshape(x_train.shape[0], image_height, image_width, 1)
x_test = x_test.reshape(x_test.shape[0], image_height, image_width, 1)
input_shape = (image_height, image_width, 1)
print(hl)
print('reshaping succesful... ')
print(hl)
print(f'The shape of x_train is {x_train.shape}')
print(f'The shape of y_train is {y_train.shape}')
print(f'The shape of x_test is {x_test.shape}')
print(f'The shape of y_test is {y_test.shape}')
else:
print(f'Invalid Format Submitted\n\n{hl}\n')
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
# splitting validation sets into categories;
y_train = to_categorical(y_train, num_classes)
y_test = to_categorical(y_test, num_classes)
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# MODEL DEFINITION FUNCTION;
def define_model():
'''
Convolutional Neural Network Built via Sequential() Module from Keras:
* Input Layer - Convolutional<-–+
* Layer 1 - Max Pooling |
* Layer 2 - Dropout <––+
* Layer 3 - Convolutional <-–+
* Layer 4 - Max Pooling |
* Layer 5 - Dropout <––+
* Layer 6 - Convolutional <-–+
* Layer 7 - Max Pooling |
* Layer 8 - Dropout <––+
* Layer 9 - Flatten
* Layer 10 - Dense
* Layer 11 - Dropout
* Output Layer 12 - Dense
'''
model = Sequential()
# ––– INPUT LAYER # 0 ––––––––––
model.add(Conv2D(32,
kernel_size=cube,
activation='relu',
kernel_initializer='he_uniform',
padding='same',
input_shape=(image_height, image_width, 1)
)
)
# ––– LAYER # 1 ––––––––––
model.add(MaxPooling2D(square))
# ––– LAYER # 2 ––––––––––
model.add(Dropout(drpout_rt_1))
# ––– LAYER # 3 ––––––––––
model.add(Conv2D(64,
cube,
activation='relu',
kernel_initializer='he_uniform',
padding='same'
)
)
# ––– LAYER # 4 ––––––––––
model.add(MaxPooling2D(square))
# ––– LAYER # 5 ––––––––––
model.add(Dropout(drpout_rt_1))
# ––– LAYER # 6 ––––––––––
model.add(Conv2D(128,
cube,
activation='relu',
kernel_initializer='he_uniform',
padding='same'
)
)
# ––– LAYER # 7 ––––––––––
model.add(MaxPooling2D(square))
# ––– LAYER # 8 ––––––––––
model.add(Dropout(drpout_rt_1))
# ––– LAYER # 9 ––––––––––
model.add(Flatten())
# ––– LAYER # 10 ––––––––––
model.add(Dense(128,
activation='relu',
kernel_initializer='he_uniform'
)
)
# ––– LAYER # 11 ––––––––––
model.add(Dropout(drpout_rt_2))
# ––– OUTPUT LAYER # 12 - WEIGHTED PREDICTION EQUATION ––––––––––
model.add(Dense(num_classes, activation=activation_dense))
# ––– Compile CNN model
model.compile(optimizer=optimizer,
loss=loss,
metrics=metrics)
return model
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# DEFINE; FIT (LONG PROCESS); SAVE
model = define_model()
model.fit(x_train, y_train,
batch_size=batch_size,
epochs=epochs,
verbose=1,
validation_data=(x_test, y_test))
model.save('MNIST_CNN.h5')
del model
print('model saved to disk')
# floyd run --gpu "python training_app.py"
# ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
# Omar Abusheikh