Train_MNIST_Model.py

# -*- coding: utf-8 -*-

"""
Created on Friday 2 Feb 2020

@author: Chris.Cui

Email: Chris.Cui@aurecongroup.com

"""
#%% load all the pkgs
from __future__ import print_function
import keras
from keras.datasets import mnist
from keras.models import Sequential, load_model
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D
from keras import backend as K
from keras.callbacks import ModelCheckpoint, EarlyStopping

import os
import numpy as np
from matplotlib import pyplot as plt
from datetime import datetime
import pickle

#%% Load the data
# the data, split between train and test sets
(x_train, y_train), (x_test, y_test) = mnist.load_data()

nb_train_sample = len(y_train)
nb_validation_sample = len(y_test)

#%% Preporcessing the data
# input image dimensions
img_rows, img_cols = 28, 28
if K.image_data_format() == 'channels_first':
    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)
    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)
    input_shape = (1, img_rows, img_cols)
else:
    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)
    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)
    input_shape = (img_rows, img_cols, 1)

x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
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
num_classes = 10
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)


#%% design the structure of the model
model = Sequential()
model.add(Conv2D(32, kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model.add(Conv2D(16, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(num_classes, activation='softmax'))


#%% model training merics
model.compile(loss=keras.losses.categorical_crossentropy,
              optimizer=keras.optimizers.Adadelta(),
              metrics=['accuracy'])
print(model.summary())

#%% Train the model
# checkpoint
# now = datetime.now()
# best_model_filepath="mnist.weights.best."+str( now.strftime("%Y.%m.%d_%H.%M.%S"))+".h5"

best_model_filepath = "./models/mnist.weights.best.h5"
checkpoint = ModelCheckpoint(best_model_filepath, 
                            monitor='val_acc', 
                            verbose=1, save_best_only=True, mode='Max')

stoppoint = EarlyStopping(monitor='val_acc', min_delta=0.001,
                            verbose=1, mode='Max', patience=5)

callbacks_list = [checkpoint, stoppoint]

# Fit the model
batch_size = 1024
epochs = 30

print('*'*80)
print('Start the training process at:', datetime.now())
print('*'*80)
history = model.fit(x_train, y_train,
          batch_size=batch_size,
          epochs=epochs,
          verbose=1,
          validation_data=(x_test, y_test),
          callbacks=callbacks_list)

print('*'*80)
print('End the training process at:', datetime.now())
print('*'*80)
# save the training records
pickle_out = open("./models/history.pickle","wb")
pickle.dump(history, pickle_out)

# *** output on my desktop
# ...
# Epoch 00020: val_acc did not improve from 0.99210
# Epoch 00020: early stopping
# ***
# End the training process at: 2020-02-04 10:33:54.029526
# ***
#%%
score = model.evaluate(x_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

#%% save the training resutls
# Plot training & validation accuracy values
plt.figure(figsize=(8, 6))
plt.plot(history.history['acc'])
plt.plot(history.history['val_acc'])
plt.title('Model accuracy')
plt.ylabel('Accuracy')
plt.xlabel('Epoch')
plt.legend(['Train', 'Test'], loc='upper left')
# plt.show()
plt.savefig('./images/training_validation_accuracy_values.png', dpi= 300 )

# Plot training & validation loss values
plt.figure(figsize=(8, 6))
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('Model loss')
plt.ylabel('Loss')
plt.xlabel('Epoch')
plt.legend(['Train', 'Test'], loc='upper left')
# plt.show()
plt.savefig('./images/training_validation_loss_values.png', dpi= 300 )


#%% save and reload the model with a timestamp / in case overight the original one
# now=datetime.now()
# save_model_dir = './models/'
# final_model_name = str(now.strftime("%Y.%m.%d_%H.%M.%S")) +'.MnistModel.h5'
# save_final_model_path = os.path.join(save_model_dir, final_model_name)
# model.save(save_final_model_path)

# Final_model =   load_model(save_final_model_path) # maybe overfitting.
# Final_model.summary()
 
#%% add more ...