深度学习框架Tensorflow学习与应用/08-2tensorflow模型保存和使用.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": 3
  },
  "orig_nbformat": 2,
  "kernelspec": {
   "name": "python_defaultSpec_1596364204877",
   "display_name": "Python 3.5.4 64-bit"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 补充（分两部分代码）（分两部分代码\n",
    "* 我们先定义一个简单的神经网络，用来训练模型，然后将模型保存下来，最后加载保存下来的模型进行检测，\n",
    "    * 模型保存\n",
    "    * 模型载入查看输出结果。\n",
    "    * 模型保存\n",
    "    * 模型加载\n",
    "\n",
    "```\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "\n",
    "\n",
    "saver.save(sess, model_dir'/test_net.ckpt')  # 模型的保存和使用\n",
    "saver.restore(sess,  # 模型的使用model_dir + '/test_net.ckpt')\n",
    "\n",
    "\n",
    "```\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 生成网络模型  saver.save saver.save\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "Extracting ../MNIST_data\\train-images-idx3-ubyte.gz\nExtracting ../MNIST_data\\train-labels-idx1-ubyte.gz\nExtracting ../MNIST_data\\t10k-images-idx3-ubyte.gz\nExtracting ../MNIST_data\\t10k-labels-idx1-ubyte.gz\n"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "# 载入数据\n",
    "mnist = input_data.read_data_sets('../MNIST_data', one_hot=True)\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "Iter 0 Testing Accuracy: 0.9028\nIter 1 Testing Accuracy: 0.9141\nIter 2 Testing Accuracy: 0.9164\nIter 3 Testing Accuracy: 0.9206\nIter 4 Testing Accuracy: 0.9226\nIter 5 Testing Accuracy: 0.9232\nIter 6 Testing Accuracy: 0.9234\nIter 7 Testing Accuracy: 0.9249\nIter 8 Testing Accuracy: 0.925\nIter 9 Testing Accuracy: 0.9258\nIter 10 Testing Accuracy: 0.9262\nIter 11 Testing Accuracy: 0.9271\nIter 12 Testing Accuracy: 0.9275\nIter 13 Testing Accuracy: 0.9278\nIter 14 Testing Accuracy: 0.9258\nIter 15 Testing Accuracy: 0.9271\nIter 16 Testing Accuracy: 0.9271\nIter 17 Testing Accuracy: 0.9284\nIter 18 Testing Accuracy: 0.9272\nIter 19 Testing Accuracy: 0.9292\nIter 20 Testing Accuracy: 0.929\n"
    }
   ],
   "source": [
    "# 批次的大小\n",
    "batch_size = 128\n",
    "n_batch = mnist.train.num_examples // batch_size\n",
    "\n",
    "x = tf.placeholder(tf.float32, [None,784])\n",
    "y = tf.placeholder(tf.float32, [None, 10])\n",
    "\n",
    "# 创建一个简单的神经网络\n",
    "W = tf.Variable(tf.zeros([784,10]))\n",
    "b = tf.Variable(tf.zeros([10]))\n",
    "prediction = tf.matmul(x,W) + b\n",
    "\n",
    "# 代价函数\n",
    "# loss = tf.reduce_mean(tf.square(y-prediction))\n",
    "loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=prediction))\n",
    "\n",
    "# 梯度下降法\n",
    "# train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)\n",
    "train_step = tf.train.AdamOptimizer().minimize(loss)\n",
    "\n",
    "# 初始化变量\n",
    "init = tf.global_variables_initializer()\n",
    "\n",
    "# 得到一个布尔型列表，存放结果是否正确\n",
    "correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(prediction,1)) #argmax 返回一维张量中最大值索引\n",
    "\n",
    "# 求准确率\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32)) # 把布尔值转换为浮点型求平均数\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "model_dir = '08net/8-2'\n",
    "if not os.path.exists(model_dir):\n",
    "    os.makedirs(model_dir)\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(init)\n",
    "    for epoch in range(21):\n",
    "        for batch in range(n_batch):\n",
    "            # 获得批次数据\n",
    "            batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "            sess.run(train_step, feed_dict={x:batch_xs, y:batch_ys})\n",
    "        acc = sess.run(accuracy, feed_dict={x:mnist.test.images,y:mnist.test.labels})\n",
    "        print(\"Iter \" + str(epoch) + \" Testing Accuracy: \" + str(acc))\n",
    "    saver.save(sess, model_dir + '/test_net.ckpt')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 载入训练好的模型  saver.restore * 直接用训练好的参数 saver.restore  \n",
    "* 有训练好的参数\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "Extracting ../MNIST_data\\train-images-idx3-ubyte.gz\nExtracting ../MNIST_data\\train-labels-idx1-ubyte.gz\nExtracting ../MNIST_data\\t10k-images-idx3-ubyte.gz\nExtracting ../MNIST_data\\t10k-labels-idx1-ubyte.gz\n"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "# 载入数据\n",
    "mnist = input_data.read_data_sets('../MNIST_data', one_hot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 批次的大小\n",
    "batch_size = 128\n",
    "n_batch = mnist.train.num_examples // batch_size\n",
    "\n",
    "x = tf.placeholder(tf.float32, [None,784])\n",
    "y = tf.placeholder(tf.float32, [None, 10])\n",
    "\n",
    "# 创建一个简单的神经网络\n",
    "W = tf.Variable(tf.zeros([784,10]))\n",
    "b = tf.Variable(tf.zeros([10]))\n",
    "prediction = tf.matmul(x,W) + b\n",
    "\n",
    "# 代价函数\n",
    "# loss = tf.reduce_mean(tf.square(y-prediction))\n",
    "loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=prediction))\n",
    "\n",
    "# 梯度下降法\n",
    "# train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)\n",
    "train_step = tf.train.AdamOptimizer().minimize(loss)\n",
    "\n",
    "# 初始化变量\n",
    "init = tf.global_variables_initializer()\n",
    "\n",
    "# 得到一个布尔型列表，存放结果是否正确\n",
    "correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(prediction,1)) #argmax 返回一维张量中最大值索引\n",
    "\n",
    "# 求准确率\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32)) # 把布尔值转换为浮点型求平均数\n",
    "\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "model_dir = 'net/8-1'\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(init)\n",
    "    acc1 = sess.run(accuracy, feed_dict={x:mnist.test.images, y:mnist.test.labels})\n",
    "    saver.restore(sess, model_dir + '/test_net.ckpt')\n",
    "    acc2 = sess.run(accuracy, feed_dict={x:mnist.test.images,y:mnist.test.labels})\n",
    "    print(\" Init Accuracy\" + str(acc1))\n",
    "    print(\" Restore Accuracy: \" + str(acc2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 批次的大小\n",
    "batch_size = 128\n",
    "n_batch = mnist.train.num_examples // batch_size\n",
    "\n",
    "x = tf.placeholder(tf.float32, [None,784])\n",
    "y = tf.placeholder(tf.float32, [None, 10])\n",
    "\n",
    "# 创建一个简单的神经网络\n",
    "W = tf.Variable(tf.zeros([784,10]))\n",
    "b = tf.Variable(tf.zeros([10]))\n",
    "prediction = tf.matmul(x,W) + b\n",
    "\n",
    "# 代价函数\n",
    "# loss = tf.reduce_mean(tf.square(y-prediction))\n",
    "loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=prediction))\n",
    "\n",
    "# 梯度下降法\n",
    "# train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)\n",
    "train_step = tf.train.AdamOptimizer().minimize(loss)\n",
    "\n",
    "# 初始化变量\n",
    "init = tf.global_variables_initializer()\n",
    "\n",
    "# 得到一个布尔型列表，存放结果是否正确\n",
    "correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(prediction,1)) #argmax 返回一维张量中最大值索引\n",
    "\n",
    "# 求准确率\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32)) # 把布尔值转换为浮点型求平均数\n",
    "\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "model_dir = '08net/8-2'\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(init)\n",
    "    acc1 = sess.run(accuracy, feed_dict={x:mnist.test.images, y:mnist.test.labels})\n",
    "    saver.restore(sess, model_dir + '/test_net.ckpt')\n",
    "    acc2 = sess.run(accuracy, feed_dict={x:mnist.test.images,y:mnist.test.labels})\n",
    "    print(\" Init Accuracy\" + str(acc1))\n",
    "    print(\" Restore Accuracy: \" + str(acc2))"
   ]
  }
 ]
}