tensorflow mnist

 

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

#下载相应的文件放在本地目录
mnist = input_data.read_data_sets("D:/work_space/eclipse/Pydev/MNIST_data", one_hot=True)
print("---mnist info-----")
print(mnist.train.images.shape, mnist.train.labels.shape)
print(mnist.test.images.shape, mnist.test.labels.shape)
print(mnist.validation.images.shape, mnist.validation.labels.shape)

x = tf.placeholder(tf.float32, [None, 784], name='x')
W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))
y = tf.nn.softmax(tf.matmul(x, W) + b, name='y')
y_ = tf.placeholder(tf.float32, [None, 10], 'y_')

cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices=[1]))

train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)
init = tf.global_variables_initializer()
#
m_saver = tf.train.Saver()

with tf.Session() as sess:
    sess.run(init)
    #训练集
    for i in range(1000):
        batch_xs, batch_ys = mnist.train.next_batch(100)
        sess.run(train_step, feed_dict = {x: batch_xs, y_: batch_ys})
        m_saver.save(sess, "D:/model_path/model_name", global_step=i) #保存模型

    #验证测试集
    correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
    print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels}))

---mnist info-----
(55000, 784) (55000, 10)
(10000, 784) (10000, 10)
(5000, 784) (5000, 10)
0.9169

下面读取模型来验证结果

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

#下载相应的文件放在本地目录
mnist = input_data.read_data_sets("D:/work_space/eclipse/Pydev/MNIST_data", one_hot=True)
print("---mnist info-----")
print(mnist.train.images.shape, mnist.train.labels.shape)
print(mnist.test.images.shape, mnist.test.labels.shape)
print(mnist.validation.images.shape, mnist.validation.labels.shape)

init = tf.global_variables_initializer()
saver = tf.train.import_meta_graph("D:/model_path/model_name-999.meta")

with tf.Session() as sess:
    sess.run(init) 
    #model_file=tf.train.latest_checkpoint('D:/model_path/model_name-999')
    saver.restore(sess, 'D:/model_path/model_name-999')    #读取模型
 
    #这里取别名
    x1 = tf.get_default_graph().get_tensor_by_name("x:0")
    y1 = tf.get_default_graph().get_tensor_by_name("y:0")
    y_1 = tf.get_default_graph().get_tensor_by_name("y_:0")
    
    correct_prediction = tf.equal(tf.argmax(y1, 1), tf.argmax(y_1, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
    print(sess.run(accuracy, feed_dict={x1: mnist.test.images, y_1: mnist.test.labels}))

---mnist info-----
(55000, 784) (55000, 10)
(10000, 784) (10000, 10)
(5000, 784) (5000, 10)
0.9169