基於Tensorflow-gpu的mnist手寫數字識別(卷積神經網絡CNN)

一、構建模型

from __future__ import absolute_import, division, print_function, unicode_literals

import tensorflow as tf

from tensorflow.keras.layers import Dense, Flatten, Conv2D
from tensorflow.keras import Model

from tensorflow import saved_model

gpus = tf.config.experimental.list_physical_devices('GPU')##獲取可用GPU
for gpu in (gpus):
 tf.config.experimental.set_memory_growth(gpu, True)##設置顯存使用方式

(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0 ##數據預處理歸一化

x_train = x_train[..., tf.newaxis] ##增加一個通道維數
x_test = x_test[..., tf.newaxis]

train_set = tf.data.Dataset.from_tensor_slices((x_train, y_train)).shuffle(10000).batch(32)##切分數據集為BatchDataset，混淆數據集
test_set = tf.data.Dataset.from_tensor_slices((x_test, y_test)).batch(32)


class MyModel(Model):##cnn網絡模型定義
  def __init__(self):
    super(MyModel, self).__init__()
    self.conv1 = Conv2D(32, 3, activation='relu')
    self.flatten = Flatten()
    self.d1 = Dense(128, activation='relu')
    self.d2 = Dense(10, activation='softmax')
  @tf.function
  def call(self, x):
    x = self.conv1(x)
    x = self.flatten(x)
    x = self.d1(x)
    return self.d2(x)

# mynetwork = tf.keras.models.Sequential([    ##一般模型
#   tf.keras.layers.Flatten(input_shape=(28, 28)),
#   tf.keras.layers.Dense(128, activation='relu'),
#   tf.keras.layers.Dropout(0.2),
#   tf.keras.layers.Dense(10, activation='softmax')
# ])

mynetwork = MyModel()


loss_object = tf.keras.losses.SparseCategoricalCrossentropy()##損失函數定義

optimizer = tf.keras.optimizers.Adam()##優化器定義

train_loss = tf.keras.metrics.Mean(name='train_loss')##損失值
train_accuracy = tf.keras.metrics.SparseCategoricalAccuracy(name='train_accuracy')##准確率

test_loss = tf.keras.metrics.Mean(name='test_loss')
test_accuracy = tf.keras.metrics.SparseCategoricalAccuracy(name='test_accuracy')


@tf.function ##訓練
def train_step(images, labels):
    with tf.GradientTape() as tape:
        predictions = mynetwork(images)
        loss = loss_object(labels, predictions)
    gradients = tape.gradient(loss, mynetwork.trainable_variables)
    optimizer.apply_gradients(zip(gradients, mynetwork.trainable_variables))

    train_loss(loss)
    train_accuracy(labels, predictions)


@tf.function ##測試
def test_step(images, labels):
    predictions = mynetwork(images)
    t_loss = loss_object(labels, predictions)

    test_loss(t_loss)
    test_accuracy(labels, predictions)


for epoch in range(5):
    # 在下一個epoch開始時，重置評估指標
    train_loss.reset_states()
    train_accuracy.reset_states()
    test_loss.reset_states()
    test_accuracy.reset_states()

    for images, labels in train_set:
        train_step(images, labels)

    for test_images, test_labels in test_set:
        test_step(test_images, test_labels)

    template = 'Epoch {}, Loss: {}, Accuracy: {}, Test Loss: {}, Test Accuracy: {}'
    print(template.format(epoch+1,
                         train_loss.result(),
                         train_accuracy.result()*100,
                         test_loss.result(),
                         test_accuracy.result()*100))


tf.saved_model.save(mynetwork, 'saved_model')##保存模型，表明文件夾即可

 

二、預測結果

可以看到，5個epoch后准確率已經非常高，通過非卷積網絡訓練模型的准確率低於卷積網絡，讀者可以自行試驗

 

參考：

https://tensorflow.google.cn/tutorials/quickstart/advanced?hl=zh_cn