-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathModel_Training.py
208 lines (185 loc) · 9.09 KB
/
Model_Training.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
import argparse
import json
import os
import keras
import matplotlib.pyplot as plt
from keras import backend as k
from keras.applications import *
from keras.callbacks import (CSVLogger, EarlyStopping, ModelCheckpoint,
ReduceLROnPlateau, TensorBoard)
from keras.engine.sequential import Sequential
from keras.layers import *
from keras.models import Model, load_model
from keras.optimizers import *
from keras.preprocessing.image import ImageDataGenerator
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
session = tf.Session(config=config)
# Sample Calling Script => python Model_Training.py -n 116 -td ./dataset/training_set -iw 299 -ih 299 -b 64 -e 150 -f 132 (Success with GTX 1080TI)
parser = argparse.ArgumentParser(
description='HKUST Path Advisor Image-Classification Model Training Script')
# Required arguments
parser.add_argument('-n', '--nb_classes', required=True,
help="<Required> number of classes", type=int)
parser.add_argument('-td', '--train_dir', required=True,
help="<Required> trainin data file path", type=str)
parser.add_argument('-vd', '--validate_dir', required=False,
help="<Optional> validation data file path (if not provided, 20/80 split will be used)")
parser.add_argument('-lm', '--load_model', required=False,
help="<Optional> load trained model")
parser.add_argument('-lw', '--load_weights', required=False,
help="<Optional> load model weights")
parser.add_argument('-iw', '--width', required=False,
default=299, help="<Optional> image width")
parser.add_argument('-ih', '--height', required=False,
default=299, help="<Optional> image height")
parser.add_argument('-b', '--batch_size', required=False,
default=32, choices=[1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024], help="<Optional> training batch size", type=int)
parser.add_argument('-t', '--train_top', required=False, default=False,
help="<Optional> train top layer only")
parser.add_argument('-e', '--epoch', required=False,
default=3000, help="<Optional> training # epoch")
parser.add_argument('-f', '--freeze', required=False, default=65,
help="<Optional> # layers freezed during training")
args = parser.parse_args()
# Hyper parameters for model
nb_classes = int(args.nb_classes)
img_width, img_height = int(args.width), int(args.height)
batch_size = int(args.batch_size)
if args.train_top:
print("## Training Top Layers Only")
else:
print("## Train All Layers")
nb_epoch = int(args.epoch)
freeze_layers = int(args.freeze)
if os.path.exists(args.train_dir):
train_data_dir = args.train_dir
else:
raise IOError("Invalid Training Path Provided")
if args.validate_dir and os.path.exists(args.validate_dir):
validation_data_dir = args.validate_dir
# Import Xception Model
base_model = Xception(input_shape=(img_width, img_height, 3),
weights='imagenet', include_top=False)
# Top Model Block
x = base_model.output
x = Dropout(0.85)(x)
x = GlobalAveragePooling2D()(x)
predictions = Dense(nb_classes, activation='softmax')(x)
# Add top layer block to your base model
model = Model(base_model.input, predictions)
# model = Sequential()
# model.add(Conv2D(32, kernel_size=(3, 3), activation='relu',
# input_shape=(img_width, img_height, 3)))
# model.add(MaxPooling2D((2, 2)))
# model.add(Dropout(0.25))
# model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))
# model.add(MaxPooling2D(pool_size=(2, 2)))
# model.add(Dropout(0.25))
# model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
# model.add(Dropout(0.4))
# model.add(GlobalAveragePooling2D())
# model.add(Dense(128, activation='relu'))
# model.add(Dropout(0.3))
# model.add(Dense(nb_classes, activation='softmax'))
# save weights of best training epoch: monitor either val_loss or val_acc
if args.load_model and os.path.exists(args.load_model):
model = load_model(args.load_model)
elif args.load_weights and os.path.exists(args.load_weights):
model.load_weights(args.load_weights)
# Read Data and Augment it
if args.validate_dir and os.path.exists(args.validate_dir):
train_datagen = ImageDataGenerator(rescale=1. / 255,
zoom_range=0.1,
brightness_range=[0.9, 1.1],
width_shift_range=0.1,
height_shift_range=0.1,
horizontal_flip=True
)
test_datagen = ImageDataGenerator(rescale=1. / 255)
train_generator = train_datagen.flow_from_directory(train_data_dir,
target_size=(
img_width, img_height),
batch_size=batch_size,
class_mode='categorical')
validation_generator = test_datagen.flow_from_directory(validation_data_dir,
target_size=(
img_width, img_height),
batch_size=batch_size,
class_mode='categorical')
else:
print("NO VALIDATED DIR FOUND")
train_datagen = ImageDataGenerator(rescale=1. / 255,
zoom_range=0.1,
brightness_range=[0.9, 1.1],
width_shift_range=0.1,
height_shift_range=0.1,
validation_split=0.2)
test_datagen = ImageDataGenerator(rescale=1. / 255)
train_generator = train_datagen.flow_from_directory(train_data_dir,
target_size=(
img_width, img_height),
batch_size=batch_size,
class_mode='categorical',
subset='training')
validation_generator = train_datagen.flow_from_directory(train_data_dir,
target_size=(
img_width, img_height),
batch_size=batch_size,
class_mode='categorical',
subset='validation')
for layer in model.layers[:freeze_layers]:
layer.trainable = False
for layer in model.layers[freeze_layers:]:
layer.trainable = True
for i, layer in enumerate(model.layers):
print(i, layer.name, layer.trainable)
final_model_path = os.path.join(
os.path.abspath('model'), 'model.h5')
callbacks_list = [
ModelCheckpoint(final_model_path, monitor='val_categorical_accuracy',
verbose=0, save_best_only=True, mode='auto', period=1),
TensorBoard(log_dir='./Graph', histogram_freq=0,
write_graph=True, write_images=True),
EarlyStopping(monitor='val_loss',
mode='min', verbose=1, patience=700),
CSVLogger('./logs/training.log')
]
model.compile(optimizer='adam',
loss='categorical_crossentropy',
metrics=['categorical_accuracy', 'top_k_categorical_accuracy'])
labels = (train_generator.class_indices)
labels = dict((v, k) for k, v in labels.items())
print(labels)
# fine-tune the model
print("### Started Training ###")
predict_history = model.fit_generator(train_generator,
steps_per_epoch=train_generator.samples/batch_size,
epochs=nb_epoch,
validation_data=validation_generator,
validation_steps=validation_generator.samples/batch_size,
callbacks=callbacks_list)
# # summarize history for accuracy
# plt.plot(predict_history.history['categorical_accuracy'])
# plt.plot(predict_history.history['val_categorical_accuracy'])
# plt.title('model accuracy')
# plt.ylabel('accuracy')
# plt.xlabel('epoch')
# plt.legend(['train', 'test'], loc='upper left')
# plt.show()
# # summarize histor for Top-5 accuracy
# plt.plot(predict_history.history['top_k_categorical_accuracy'])
# plt.plot(predict_history.history['val_top_k_categorical_accuracy'])
# plt.title('model top 5 accuracy')
# plt.ylabel('top 5 accuracy')
# plt.xlabel('epoch')
# plt.legend(['train', 'test'], loc='upper left')
# plt.show()
# # summarize history for loss
# plt.plot(predict_history.history['loss'])
# plt.plot(predict_history.history['val_loss'])
# plt.title('model loss')
# plt.ylabel('loss')
# plt.xlabel('epoch')
# plt.legend(['train', 'test'], loc='upper left')
# plt.show()