10_save_restore_net.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# This shows how to save/restore your model (trained variables).\n",
    "# To see how it works, please stop this program during training and resart.\n",
    "# This network is the same as 3_net.py\n",
    "\n",
    "def init_weights(shape):\n",
    "    return tf.Variable(tf.random_normal(shape, stddev=0.01))\n",
    "\n",
    "def model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden): # this network is the same as the previous one except with an extra hidden layer + dropout\n",
    "    X = tf.nn.dropout(X, p_keep_input)\n",
    "    h = tf.nn.relu(tf.matmul(X, w_h))\n",
    "\n",
    "    h = tf.nn.dropout(h, p_keep_hidden)\n",
    "    h2 = tf.nn.relu(tf.matmul(h, w_h2))\n",
    "\n",
    "    h2 = tf.nn.dropout(h2, p_keep_hidden)\n",
    "\n",
    "    return tf.matmul(h2, w_o)\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = tf.placeholder(\"float\", [None, 784])\n",
    "Y = tf.placeholder(\"float\", [None, 10])\n",
    "\n",
    "w_h = init_weights([784, 625])\n",
    "w_h2 = init_weights([625, 625])\n",
    "w_o = init_weights([625, 10])\n",
    "\n",
    "p_keep_input = tf.placeholder(\"float\")\n",
    "p_keep_hidden = tf.placeholder(\"float\")\n",
    "py_x = model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden)\n",
    "\n",
    "cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=py_x, labels=Y))\n",
    "train_op = tf.train.RMSPropOptimizer(0.001, 0.9).minimize(cost)\n",
    "predict_op = tf.argmax(py_x, 1)\n",
    "\n",
    "ckpt_dir = \"./ckpt_dir\"\n",
    "if not os.path.exists(ckpt_dir):\n",
    "    os.makedirs(ckpt_dir)\n",
    "\n",
    "global_step = tf.Variable(0, name='global_step', trainable=False)\n",
    "\n",
    "# Call this after declaring all tf.Variables.\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "# This variable won't be stored, since it is declared after tf.train.Saver()\n",
    "non_storable_variable = tf.Variable(777)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Launch the graph in a session\n",
    "with tf.Session() as sess:\n",
    "    # you need to initialize all variables\n",
    "    tf.global_variables_initializer().run()\n",
    "\n",
    "    ckpt = tf.train.get_checkpoint_state(ckpt_dir)\n",
    "    if ckpt and ckpt.model_checkpoint_path:\n",
    "        print(ckpt.model_checkpoint_path)\n",
    "        saver.restore(sess, ckpt.model_checkpoint_path) # restore all variables\n",
    "\n",
    "    start = global_step.eval() # get last global_step\n",
    "    print(\"Start from:\", start)\n",
    "\n",
    "    for i in range(start, 100):\n",
    "        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):\n",
    "            sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],\n",
    "                                          p_keep_input: 0.8, p_keep_hidden: 0.5})\n",
    "\n",
    "        global_step.assign(i).eval() # set and update(eval) global_step with index, i\n",
    "        saver.save(sess, ckpt_dir + \"/model.ckpt\", global_step=global_step)\n",
    "        print(i, np.mean(np.argmax(teY, axis=1) ==\n",
    "                         sess.run(predict_op, feed_dict={X: teX, Y: teY,\n",
    "                                                         p_keep_input: 1.0,\n",
    "                                                         p_keep_hidden: 1.0})))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}