custom model creation and saving the model. Binary Classification, keras code for CNN

cracks_detection.py

# imports
from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D
from keras.layers import Activation, Dense, Flatten, Dropout
from batchGenerator import generateData

# creating a model - function based approach

def create_model():
    model = Sequential()
    
    model.add(Conv2D(32, (3, 3), input_shape=(150, 150, 3)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2, 2), dim_ordering='th'))
    
    model.add(Conv2D(32, (3, 3), dim_ordering='th'))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2, 2), dim_ordering='th'))
    
    model.add(Conv2D(62, (3, 3)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2, 2), dim_ordering='th'))
    
    model.add(Flatten())
    model.add(Dense(64))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))
    model.add(Dense(1))
    model.add(Activation('sigmoid'))
    
    return model

# function for training and saving model
def train():    
    model = create_model()    
    model.compile(loss='binary_crossentropy',
              optimizer='rmsprop',
              metrics=['accuracy'])
    train_generator, validation_generator, batch_size = generateData()
    model.fit_generator(
        train_generator,
        steps_per_epoch=500 // batch_size,
        epochs=50,
        validation_data=validation_generator,
        validation_steps=500 // batch_size)
   
    model.save_weights('cracks_detection.h5')  # saving model
    
    return model

Todo
1 - Training data path
2 - Model Evaluation