【Tensorflow笔记】

最新推荐文章于 2024-12-07 15:45:16 发布

原创最新推荐文章于 2024-12-07 15:45:16 发布 · 258 阅读

0 ·

CC 4.0 BY-SA版权

博客主要介绍了逻辑回归，涉及独热表示及占位符使用，还提及一个简单的神经网络，最后阐述了模型的保存和读取相关内容，包含保存和提取操作，均围绕信息技术领域的模型构建与处理。

逻辑回归

一个简单的神经网络

模型的保存和读取

逻辑回归

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

mnist = input_data.read_data_sets('MNIST_data/',one_hot=true)

x=tf.placeholder(tf.float32,[None,784])
y=tf.placeholder(tf.float32,[None,10])
w=tf.Variable(tf.zeros([784,10]))
b=tf.Variable(tf.zeros([10]))
pred=tf.nn.softmax(tf.matmul(x,w)+b)

sess=tf.InteractiveSession()

cost=tf.reduce_mean(-tf.reduce_sum(y*tf.log(pred)))
optimizer=tf.train.GradientDescentOptimizer(0.01).minimize(cost)
corr=tf.equal(tf.argmax(pred,1),tf.argmax(y,1))
accr=tf.reduce_mean(tf.cast(corr,tf.float32))
init=tf.global_variables_initializer()

sess=tf.Session()
sess.run(init)

epoch_n=100
batch_size=64
batch_n=int(mnist.train.num_examples/batch_size)

for epoch in range(epoch_n):
    avg_cost=0
    for _ in range(batch_n):
        batch_xs,batch_ys=mnist.train.next_batch(batch_size)
        feeds={x:batch_xs,y:batch_ys}
        sess.run(optimizer,feed_dict=feeds)
        avg_cost+=sess.run(cost,feed_dict=feeds)/batch_size
    print(avg_cost)

one_hot=true 独热表示：一位有效编码

读取到的mnist lables： [1,0,0,0,0,0,0,0,0,0] [0,1,0,0,0,0,0,0,0,0] [0,0,1,0,0,0,0,0,0,0] ........

x=tf.placeholder(tf.float32,[None,784]) 未赋值，先占位 None 可传任意个数

pred=tf.nn.softmax(tf.matmul(x,w)+b)

一个简单的神经网络

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

mnist = input_data.read_data_sets('MNIST_data/',one_hot=True)

n_input=784
n_hidden_1=256
n_hidden_2=128

n_classes=10
x=tf.placeholder(tf.float32,[None,n_input])
y=tf.placeholder(tf.float32,[None,n_classes])
stddev=0.1

weights={
    'w1':tf.Variable(tf.random_normal([n_input,n_hidden_1],stddev=stddev)),
    'w2':tf.Variable(tf.random_normal([n_hidden_1,n_hidden_2],stddev=stddev)),
    'out':tf.Variable(tf.random_normal([n_hidden_2,n_classes],stddev=stddev))
}

biases={
    'b1':tf.Variable(tf.random_normal([n_hidden_1])),
    'b2':tf.Variable(tf.random_normal([n_hidden_2])),
    'out':tf.Variable(tf.random_normal([n_classes]))
}

print("-"*10+"NetWork Reday"+"-"*10)


#前向
def multiplayer_perception(x,weights,biases):
    layer_1=tf.matmul(x,weights['w1'])  #输入*权重
    tf.add(layer_1,biases['b1']) #加偏置项
    layer_1=tf.sigmoid(layer_1) #激活
    layer_2=tf.sigmoid(tf.add(tf.matmul(layer_1,weights['w2']),biases['b2']))
    out=tf.matmul(layer_2,weights['out'])+biases['out']
    return out

pred=multiplayer_perception(x,weights,biases)

cost=tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))
optimizer=tf.train.GradientDescentOptimizer(0.001).minimize(cost)
corr=tf.equal(tf.argmax(pred,1),tf.argmax(y,1))
accr=tf.reduce_mean(tf.cast(corr,tf.float32))
init=tf.global_variables_initializer()

print("-"*10+"Function Ready"+"-"*10)


sess=tf.Session()
sess.run(init)

epoch_n=100
batch_size=64
batch_n=int(mnist.train.num_examples/batch_size)

for epoch in range(epoch_n):
    avg_cost=0
    for i in range(batch_n):
        batch_xs,batch_ys=mnist.train.next_batch(batch_size)
        feeds={x:batch_xs,y:batch_ys}
        sess.run(optimizer,feed_dict=feeds)
        avg_cost+=sess.run(cost,feed_dict=feeds)/batch_size
    print(avg_cost)

模型的保存和读取

保存

saver=tf.train.Saver(max_to_keep=3)   #max_to_keep保存模型的个数 

for epoch in range(epoch_n):
    avg_cost=0
    for i in range(batch_n):
        batch_xs,batch_ys=mnist.train.next_batch(batch_size)
        feeds={x:batch_xs,y:batch_ys}
        sess.run(optimizer,feed_dict=feeds)
        avg_cost+=sess.run(cost,feed_dict=feeds)/batch_size
    print(avg_cost)

    saver.save(sess,'ckpt/mnist.ckpt',global_step=epoch+1)

提取

saver=tf.train.Saver(max_to_keep=3)  #######max_to_keep保存模型的个数

is_train=0

if is_train:
    for epoch in range(epoch_n):
        avg_cost = 0
        for i in range(batch_n):
            batch_xs, batch_ys = mnist.train.next_batch(batch_size)
            feeds = {x: batch_xs, y: batch_ys}
            sess.run(optimizer, feed_dict=feeds)
            avg_cost += sess.run(cost, feed_dict=feeds) / batch_size
        print(avg_cost)
        saver.save(sess, 'ckpt/mnist.ckpt', global_step=epoch + 1)



else:
    model_file = tf.train.latest_checkpoint('ckpt/')
    saver.restore(sess, model_file)
    val_cost, val_accr = sess.run([cost, accr], feed_dict={x: mnist.test.images, y: mnist.test.labels})
    print('val_loss:%f, val_acc:%f' % (val_cost, val_accr))
sess.close()