-
Notifications
You must be signed in to change notification settings - Fork 9
/
train.py
78 lines (58 loc) · 3.19 KB
/
train.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
"""
Created on Wed Feb 2019
@author: abahri
"""
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Feb 26 11:25:48 2020
@author: abahri
"""
import os
import warnings
from skimage.io import imread, imsave
from keras import applications
from keras import optimizers
from keras.utils import to_categorical
from keras.preprocessing.image import ImageDataGenerator
from sklearn.metrics import accuracy_score, classification_report
import keras
from keras import backend as K
K.set_image_data_format('channels_last')
from keras.models import load_model
from glob import glob
import numpy as np
from keras.engine import Layer
from keras.engine import InputSpec
from keras.utils import conv_utils
from model import *
from config import *
from loss_function import *
def train_model():
## Create model
NASNET_Model= model(image_shape= cfg["Input_Shape"])
## Compile the model
NASNET_Model.compile(optimizer= keras.optimizers.SGD(lr= cfg["lr"], momentum= cfg["Momentum"]), loss= ICE, metrics=["accuracy"])
## Data direction
data_dir= cfg["Dataset_Address"]
target_dirs = {target: os.path.join(data_dir, target) for target in ['train', 'valid']}
## Data generator
image_data_gen_train = ImageDataGenerator(rescale=1./255, rotation_range= cfg["Rotation_Range"], width_shift_range= cfg["Width_Shift_Range"],
height_shift_range= cfg["Height_Shift_Range"], horizontal_flip= cfg["Horizontal_Flip"],
fill_mode= cfg["Fill_Mode"])
image_generator_train= image_data_gen_train.flow_from_directory(target_dirs["train"], batch_size= cfg["Train_Batch_Size"], class_mode= cfg["Class_Mode"], target_size= cfg["Target_Size"], shuffle=True)
image_data_gen_test= ImageDataGenerator(rescale=1./255)
image_generator_test= image_data_gen_test.flow_from_directory(target_dirs["test"], class_mode= cfg["Class_Mode"], target_size= cfg["Target_Size"], batch_size= cfg["Test_Batch_Size"], shuffle=True)
## Checkpoint
callback_model= keras.callbacks.ModelCheckpoint( cfg["Save_dir"] + "/model.h5",
monitor= "val_acc", verbose= 1, save_best_only= True, mode= "max")
callback_CSV= keras.callbacks.CSVLogger(cfg["Save_dir"] + "/model.log", append=True)
## Train the model
if (cfg["Flag_Load_Model"] == False):
NASNET_Model.fit_generator(image_generator_train, steps_per_epoch= len(image_generator_train), epochs= cfg["Epochs"], validation_data= image_generator_test, validation_steps= len(image_generator_test), callbacks=[callback_model, callback_CSV], shuffle= True)
else:
NASNET_Model= load_model(cfg["Path_Load_Model"], custom_objects={ "ICE":ICE})
NASNET_Model.compile(optimizer= keras.optimizers.SGD(lr= cfg["lr"], momentum= cfg["Momentum"]), loss= ICE, metrics=["accuracy"])
NASNET_Model.fit_generator(image_generator_train, steps_per_epoch= len(image_generator_train), epochs= cfg["Epochs"], validation_data= image_generator_test, validation_steps= len(image_generator_test), callbacks=[callback_model, callback_CSV], shuffle= True)
if __name__ == '__main__':
train_model()