TensorFlow 的高層封裝API ——————Keras 基本用法

Keras API 和TFlearn API 用法基本相似，對於tensorflow 的 模型定義，損失函數，訓練過程等進行了封裝。封裝後的整個數據集訓練過程包括，數據處理，模型定義和模型訓練三個部分。

以代碼展示瞭如何使用Keras 在MNIST dataset 上實現了LeNet-5 Model.

import keras
from  keras.datasets import mnist
from  keras.models import Sequential
from  keras.layers import Dense, Flatten, Conv2D, MaxPooling2D
from  keras import backend as k

num_classes = 10
img_rows, img_cols = 28, 28
(trainX, trainY), (testX, testY) = mnist.load_data()
if k.image_data_format() == 'channels_first':
    trainX = trainX.reshape(trainX.shape[0], 1, img_rows, img_cols)
    testX = testX.reshape(testX.shape[0], 1, img_rows, img_cols)
    input_shape = (1, img_rows, img_cols)
else:
    trainX = trainX.reshape(trainX.shape[0], img_rows, img_cols, 1)
    testX = testX.reshape(testX.shape[0], img_rows, img_cols, 1)
    input_shape = (img_rows, img_cols,1)

trainX = trainX.astype('float32')
testX = testX.astype('float32')
trainX /= 255.0
testX /= 255.0

trainY = keras.utils.to_categorical(trainY, num_classes)
testY =   keras.utils.to_categorical(testY, num_classes)

model = Sequential()
model.add(
    Conv2D(32, kernel_size=(5, 5), activation='relu', input_shape=input_shape))

model.add(MaxPooling2D(pool_size=(2,2)))
model.add(Flatten())
model.add(Dense(500, activation='relu'))
model.add(Dense(num_classes, activation='softmax'))

model.compile(loss=keras.losses.categorical_crossentropy, optimizer=keras.optimizers.SGD(),metrics=['accuracy'])

# 訓練數據， 批次大小， 訓練輪數（這裏可以適當增加輪數來優化參數提高準確度），和驗證數據， Keras 可以自動完成模型訓練過程
model.fit(trainX, trainY, batch_size=128, epochs=4, validation_data=(testX, testY))
score = model.evaluate(testX, testY)
print('Test loss', score[0])
print('Test accuracy:', score[1])

由於訓練時CPU處於全負荷運作中，導致電腦耗電異常快，上次我跑了40分鐘的訓練，就耗掉了我的小米pro筆記本25%到27%的電量，所以這裏我把epoch （輪數）值設定爲5，準確度自然會低很多，代碼運行結果如下：

Downloading data from https://s3.amazonaws.com/img-datasets/mnist.npz
......
59648/60000 [============================>.] - ETA: 0s - loss: 0.2172 - acc: 0.9370
59776/60000 [============================>.] - ETA: 0s - loss: 0.2171 - acc: 0.9371
59904/60000 [============================>.] - ETA: 0s - loss: 0.2172 - acc: 0.9371
60000/60000 [==============================] - 35s 589us/step - loss: 0.2171 - acc: 0.9371 - val_loss: 0.1904 - val_acc: 0.9450

......
9696/10000 [============================>.] - ETA: 0s
 9952/10000 [============================>.] - ETA: 0s
10000/10000 [==============================] - 2s 214us/step
Test loss 0.1904236241132021
Test accuracy: 0.945

Process finished with exit code 0