ML_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use shift-enter to execute a code block and move to the next one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 2.1 Import libraries.\n",
    "import math\n",
    "import os\n",
    "from six.moves import xrange  # pylint: disable=redefined-builtin\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.contrib.learn.python.learn.datasets.mnist import read_data_sets\n",
    "\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 2.2 Define some constants.\n",
    "# The MNIST dataset has 10 classes, representing the digits 0 through 9.\n",
    "NUM_CLASSES = 10\n",
    "\n",
    "# The MNIST images are always 28x28 pixels.\n",
    "IMAGE_SIZE = 28\n",
    "IMAGE_PIXELS = IMAGE_SIZE * IMAGE_SIZE\n",
    "\n",
    "# Batch size. Must be evenly dividable by dataset sizes.\n",
    "BATCH_SIZE = 100\n",
    "EVAL_BATCH_SIZE = 1\n",
    "\n",
    "# Number of units in hidden layers.\n",
    "HIDDEN1_UNITS = 128\n",
    "HIDDEN2_UNITS = 32\n",
    "\n",
    "# Maximum number of training steps.\n",
    "MAX_STEPS = 2000\n",
    "\n",
    "# Directory to put the training data.\n",
    "TRAIN_DIR=\"/tmp/mnist\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.\n",
      "Extracting /tmp/mnist/train-images-idx3-ubyte.gz\n",
      "Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.\n",
      "Extracting /tmp/mnist/train-labels-idx1-ubyte.gz\n",
      "Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.\n",
      "Extracting /tmp/mnist/t10k-images-idx3-ubyte.gz\n",
      "Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.\n",
      "Extracting /tmp/mnist/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "# 2.3 Get input data: get the sets of images and labels for training, validation, and\n",
    "# test on MNIST.\n",
    "data_sets = read_data_sets(TRAIN_DIR, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 2.4 Build inference graph.\n",
    "def mnist_inference(images, hidden1_units, hidden2_units):\n",
    "    \"\"\"Build the MNIST model up to where it may be used for inference.\n",
    "    Args:\n",
    "        images: Images placeholder.\n",
    "        hidden1_units: Size of the first hidden layer.\n",
    "        hidden2_units: Size of the second hidden layer.\n",
    "    Returns:\n",
    "        logits: Output tensor with the computed logits.\n",
    "    \"\"\"\n",
    "    # Hidden 1\n",
    "    with tf.name_scope('hidden1'):\n",
    "        weights = tf.Variable(\n",
    "            tf.truncated_normal([IMAGE_PIXELS, hidden1_units],\n",
    "                                stddev=1.0 / math.sqrt(float(IMAGE_PIXELS))),\n",
    "            name='weights')\n",
    "        biases = tf.Variable(tf.zeros([hidden1_units]),\n",
    "                             name='biases')\n",
    "        hidden1 = tf.nn.relu(tf.matmul(images, weights) + biases)\n",
    "    # Hidden 2\n",
    "    with tf.name_scope('hidden2'):\n",
    "        weights = tf.Variable(\n",
    "            tf.truncated_normal([hidden1_units, hidden2_units],\n",
    "                                stddev=1.0 / math.sqrt(float(hidden1_units))),\n",
    "            name='weights')\n",
    "        biases = tf.Variable(tf.zeros([hidden2_units]),\n",
    "                             name='biases')\n",
    "        hidden2 = tf.nn.relu(tf.matmul(hidden1, weights) + biases)\n",
    "    # Linear\n",
    "    with tf.name_scope('softmax_linear'):\n",
    "        weights = tf.Variable(\n",
    "            tf.truncated_normal([hidden2_units, NUM_CLASSES],\n",
    "                                stddev=1.0 / math.sqrt(float(hidden2_units))),\n",
    "            name='weights')\n",
    "        biases = tf.Variable(tf.zeros([NUM_CLASSES]),\n",
    "                             name='biases')\n",
    "        logits = tf.matmul(hidden2, weights) + biases\n",
    "\n",
    "    # Uncomment the following line to see what we have constructed.\n",
    "    tf.train.write_graph(tf.get_default_graph().as_graph_def(),\n",
    "                         \"/tmp\", \"inference.pbtxt\", as_text=True)\n",
    "    return logits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 2.5 Build training graph.\n",
    "def mnist_training(logits, labels, learning_rate):\n",
    "    \"\"\"Build the training graph.\n",
    "\n",
    "    Args:\n",
    "        logits: Logits tensor, float - [BATCH_SIZE, NUM_CLASSES].\n",
    "        labels: Labels tensor, int32 - [BATCH_SIZE], with values in the\n",
    "          range [0, NUM_CLASSES).\n",
    "        learning_rate: The learning rate to use for gradient descent.\n",
    "    Returns:\n",
    "        train_op: The Op for training.\n",
    "        loss: The Op for calculating loss.\n",
    "    \"\"\"\n",
    "    # Create an operation that calculates loss.\n",
    "    labels = tf.to_int64(labels)\n",
    "    cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(\n",
    "        logits, labels, name='xentropy')\n",
    "    loss = tf.reduce_mean(cross_entropy, name='xentropy_mean')\n",
    "    # Create the gradient descent optimizer with the given learning rate.\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate)\n",
    "    # Create a variable to track the global step.\n",
    "    global_step = tf.Variable(0, name='global_step', trainable=False)\n",
    "    # Use the optimizer to apply the gradients that minimize the loss\n",
    "    # (and also increment the global step counter) as a single training step.\n",
    "    train_op = optimizer.minimize(loss, global_step=global_step)\n",
    "\n",
    "    # Uncomment the following line to see what we have constructed.\n",
    "    tf.train.write_graph(tf.get_default_graph().as_graph_def(),\n",
    "                         \"/tmp\", \"train.pbtxt\", as_text=True)\n",
    "\n",
    "    return train_op, loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 2.6 Build the complete graph for feeding inputs, training, and saving checkpoints.\n",
    "mnist_graph = tf.Graph()\n",
    "with mnist_graph.as_default():\n",
    "    # Generate placeholders for the images and labels.\n",
    "    images_placeholder = tf.placeholder(tf.float32)                                       \n",
    "    labels_placeholder = tf.placeholder(tf.int32)\n",
    "    tf.add_to_collection(\"images\", images_placeholder)  # Remember this Op.\n",
    "    tf.add_to_collection(\"labels\", labels_placeholder)  # Remember this Op.\n",
    "\n",
    "    # Build a Graph that computes predictions from the inference model.\n",
    "    logits = mnist_inference(images_placeholder,\n",
    "                             HIDDEN1_UNITS,\n",
    "                             HIDDEN2_UNITS)\n",
    "    tf.add_to_collection(\"logits\", logits)  # Remember this Op.\n",
    "\n",
    "    # Add to the Graph the Ops that calculate and apply gradients.\n",
    "    train_op, loss = mnist_training(logits, labels_placeholder, 0.01)\n",
    "\n",
    "    # Add the variable initializer Op.\n",
    "    init = tf.initialize_all_variables()\n",
    "\n",
    "    # Create a saver for writing training checkpoints.\n",
    "    saver = tf.train.Saver()\n",
    "    \n",
    "    # Uncomment the following line to see what we have constructed.\n",
    "    tf.train.write_graph(tf.get_default_graph().as_graph_def(),\n",
    "                         \"/tmp\", \"complete.pbtxt\", as_text=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Step 0: loss = 2.32\n",
      "Step 100: loss = 2.12\n",
      "Step 200: loss = 1.79\n",
      "Step 300: loss = 1.37\n",
      "Step 400: loss = 1.08\n",
      "Step 500: loss = 0.91\n",
      "Step 600: loss = 0.76\n",
      "Step 700: loss = 0.60\n",
      "Step 800: loss = 0.49\n",
      "Step 900: loss = 0.62\n",
      "Step 1000: loss = 0.51\n",
      "Step 1100: loss = 0.34\n",
      "Step 1200: loss = 0.41\n",
      "Step 1300: loss = 0.40\n",
      "Step 1400: loss = 0.32\n",
      "Step 1500: loss = 0.45\n",
      "Step 1600: loss = 0.48\n",
      "Step 1700: loss = 0.50\n",
      "Step 1800: loss = 0.41\n",
      "Step 1900: loss = 0.36\n"
     ]
    }
   ],
   "source": [
    "# 2.7 Run training for MAX_STEPS and save checkpoint at the end.\n",
    "with tf.Session(graph=mnist_graph) as sess:\n",
    "    # Run the Op to initialize the variables.\n",
    "    sess.run(init)\n",
    "\n",
    "    # Start the training loop.\n",
    "    for step in xrange(MAX_STEPS):\n",
    "        # Read a batch of images and labels.\n",
    "        images_feed, labels_feed = data_sets.train.next_batch(BATCH_SIZE)\n",
    "\n",
    "        # Run one step of the model.  The return values are the activations\n",
    "        # from the `train_op` (which is discarded) and the `loss` Op.  To\n",
    "        # inspect the values of your Ops or variables, you may include them\n",
    "        # in the list passed to sess.run() and the value tensors will be\n",
    "        # returned in the tuple from the call.\n",
    "        _, loss_value = sess.run([train_op, loss],\n",
    "                                 feed_dict={images_placeholder: images_feed,\n",
    "                                            labels_placeholder: labels_feed})\n",
    "\n",
    "        # Print out loss value.\n",
    "        if step % 100 == 0:\n",
    "            print('Step %d: loss = %.2f' % (step, loss_value))\n",
    "\n",
    "    # Write a checkpoint.\n",
    "    checkpoint_file = os.path.join(TRAIN_DIR, 'checkpoint')\n",
    "    saver.save(sess, checkpoint_file, global_step=step)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ground truth: 4\n",
      "Prediction: 4\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFfCAYAAACfj30KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYeecG9D/AxzPiy170y9OAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1214119e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 2.8 Run evaluation based on the saved checkpoint.\n",
    "with tf.Session(graph=tf.Graph()) as sess:\n",
    "    saver = tf.train.import_meta_graph(\n",
    "        os.path.join(TRAIN_DIR, \"checkpoint-1999.meta\"))\n",
    "    saver.restore(\n",
    "        sess, os.path.join(TRAIN_DIR, \"checkpoint-1999\"))\n",
    "\n",
    "    # Retrieve the Ops we 'remembered'.\n",
    "    logits = tf.get_collection(\"logits\")[0]\n",
    "    images_placeholder = tf.get_collection(\"images\")[0]\n",
    "    labels_placeholder = tf.get_collection(\"labels\")[0]\n",
    "    \n",
    "    # Add an Op that chooses the top k predictions.\n",
    "    eval_op = tf.nn.top_k(logits)\n",
    "    \n",
    "    # Run evaluation.\n",
    "    images_feed, labels_feed = data_sets.validation.next_batch(EVAL_BATCH_SIZE)\n",
    "    imgplot = plt.imshow(np.reshape(images_feed, (28, 28)))\n",
    "    prediction = sess.run(eval_op,\n",
    "                          feed_dict={images_placeholder: images_feed,\n",
    "                                     labels_placeholder: labels_feed})\n",
    "    print(\"Ground truth: %d\\nPrediction: %d\" % (labels_feed, prediction.indices[0][0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "<p>Back to [1_linear_regression_model.ipynb](1_linear_regression_model.ipynb).</p>\n",
    "<p>Next to [Bonus Lab](extras/extras_0_deepdream.ipynb).</p>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}