MNIST basic_for blogging.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting /Users/yoon/Desktop/87_105_110_107/TP/MNIST/data/train-images-idx3-ubyte.gz\n",
      "Extracting /Users/yoon/Desktop/87_105_110_107/TP/MNIST/data/train-labels-idx1-ubyte.gz\n",
      "Extracting /Users/yoon/Desktop/87_105_110_107/TP/MNIST/data/t10k-images-idx3-ubyte.gz\n",
      "Extracting /Users/yoon/Desktop/87_105_110_107/TP/MNIST/data/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "from __future__ import absolute_import\n",
    "from __future__ import division\n",
    "from __future__ import print_function\n",
    "\n",
    "import tensorflow as tf\n",
    "tf.reset_default_graph()\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "\n",
    "mnist = input_data.read_data_sets('/Users/yoon/Desktop/87_105_110_107/TP/MNIST/data', one_hot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MNIST:  Datasets(train=<tensorflow.contrib.learn.python.learn.datasets.mnist.DataSet object at 0x118de79e8>, validation=<tensorflow.contrib.learn.python.learn.datasets.mnist.DataSet object at 0x1064a00b8>, test=<tensorflow.contrib.learn.python.learn.datasets.mnist.DataSet object at 0x1064a0240>)\n",
      "MNIST 트레인 데이터 사이즈: 55000\n",
      "MNIST 밸리데이션 데이터 사이즈: 5000\n",
      "MNIST 테스트 데이터 사이즈: 10000\n"
     ]
    }
   ],
   "source": [
    "print(\"MNIST: \", mnist)\n",
    "print(\"MNIST 트레인 데이터 사이즈:\", mnist.train.num_examples)\n",
    "print(\"MNIST 밸리데이션 데이터 사이즈:\", mnist.validation.num_examples)\n",
    "print(\"MNIST 테스트 데이터 사이즈:\", mnist.test.num_examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.38039219,  0.37647063,  0.3019608 ,\n",
       "        0.46274513,  0.2392157 ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.35294119,  0.5411765 ,  0.92156869,\n",
       "        0.92156869,  0.92156869,  0.92156869,  0.92156869,  0.92156869,\n",
       "        0.98431379,  0.98431379,  0.97254908,  0.99607849,  0.96078438,\n",
       "        0.92156869,  0.74509805,  0.08235294,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.54901963,\n",
       "        0.98431379,  0.99607849,  0.99607849,  0.99607849,  0.99607849,\n",
       "        0.99607849,  0.99607849,  0.99607849,  0.99607849,  0.99607849,\n",
       "        0.99607849,  0.99607849,  0.99607849,  0.99607849,  0.99607849,\n",
       "        0.74117649,  0.09019608,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.88627458,  0.99607849,  0.81568635,\n",
       "        0.78039223,  0.78039223,  0.78039223,  0.78039223,  0.54509807,\n",
       "        0.2392157 ,  0.2392157 ,  0.2392157 ,  0.2392157 ,  0.2392157 ,\n",
       "        0.50196081,  0.8705883 ,  0.99607849,  0.99607849,  0.74117649,\n",
       "        0.08235294,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.14901961,  0.32156864,  0.0509804 ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.13333334,\n",
       "        0.83529419,  0.99607849,  0.99607849,  0.45098042,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.32941177,  0.99607849,\n",
       "        0.99607849,  0.91764712,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.32941177,  0.99607849,  0.99607849,  0.91764712,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.41568631,  0.6156863 ,\n",
       "        0.99607849,  0.99607849,  0.95294124,  0.20000002,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.09803922,  0.45882356,  0.89411771,  0.89411771,\n",
       "        0.89411771,  0.99215692,  0.99607849,  0.99607849,  0.99607849,\n",
       "        0.99607849,  0.94117653,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.26666668,  0.4666667 ,  0.86274517,\n",
       "        0.99607849,  0.99607849,  0.99607849,  0.99607849,  0.99607849,\n",
       "        0.99607849,  0.99607849,  0.99607849,  0.99607849,  0.55686277,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.14509805,  0.73333335,\n",
       "        0.99215692,  0.99607849,  0.99607849,  0.99607849,  0.87450987,\n",
       "        0.80784321,  0.80784321,  0.29411766,  0.26666668,  0.84313732,\n",
       "        0.99607849,  0.99607849,  0.45882356,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.44313729,  0.8588236 ,  0.99607849,  0.94901967,  0.89019614,\n",
       "        0.45098042,  0.34901962,  0.12156864,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.7843138 ,  0.99607849,  0.9450981 ,\n",
       "        0.16078432,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.66274512,  0.99607849,\n",
       "        0.6901961 ,  0.24313727,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.18823531,\n",
       "        0.90588242,  0.99607849,  0.91764712,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.07058824,  0.48627454,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.32941177,  0.99607849,  0.99607849,\n",
       "        0.65098041,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.54509807,  0.99607849,  0.9333334 ,  0.22352943,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.82352948,  0.98039222,  0.99607849,\n",
       "        0.65882355,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.94901967,  0.99607849,  0.93725497,  0.22352943,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.34901962,  0.98431379,  0.9450981 ,\n",
       "        0.33725491,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.01960784,\n",
       "        0.80784321,  0.96470594,  0.6156863 ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.01568628,  0.45882356,  0.27058825,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "        0.        ,  0.        ,  0.        ,  0.        ], dtype=float32)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.train.images[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "784\n"
     ]
    }
   ],
   "source": [
    "print(len(mnist.train.images[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTRHkDwRqiv8DZjZBNqXYK70AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11d00bbe0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "some_digit = mnist.train.images[1]\n",
    "some_digit_image = some_digit.reshape(28, 28)\n",
    "plt.imshow(some_digit_image, cmap = matplotlib.cm.binary,\n",
    "           interpolation=\"bessel\")\n",
    "plt.axis(\"off\")\n",
    "plt.show()\n",
    "\n",
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_digit(data):\n",
    "    image = data.reshape(28, 28)\n",
    "    plt.imshow(image, cmap = matplotlib.cm.binary,\n",
    "              interpolation=\"neareast\")\n",
    "    plt.axis(\"off\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "def plot_digits(instances, images_per_row=10, **options):\n",
    "    size = 28\n",
    "    images_per_row = min(len(instances), images_per_row)\n",
    "    images = [instance.reshape(size, size) for instance in instances]\n",
    "    n_rows = (len(instances) - 1) // images_per_row + 1\n",
    "    row_images = []\n",
    "    n_empty = n_rows * images_per_row - len(instances)\n",
    "    images.append(np.zeros((size, size * n_empty)))\n",
    "    #make a empty space\n",
    "    for row in range(n_rows):\n",
    "        rimages = images[row * images_per_row : (row + 1) * images_per_row]\n",
    "        row_images.append(np.concatenate(rimages, axis=1))\n",
    "    image = np.concatenate(row_images, axis=0)\n",
    "    plt.imshow(image, cmap = matplotlib.cm.binary, **options)\n",
    "    plt.axis(\"off\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28, 28)\n",
      "(28, 140)\n",
      "(196, 280)\n",
      "(28, 280)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAGCCAYAAADdQKhZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLDaeiJiZmVlsPBExMzOz2Pw/ZmDTFhEJ9qAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118df8160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9,9))\n",
    "examples = np.r_[mnist.train.images[:12000:600],\n",
    "                 mnist.train.images[13000:40000:600]]\n",
    "size = 28\n",
    "images_per_row=10\n",
    "images_per_row = min(len(examples), images_per_row)\n",
    "images = [example.reshape(size, size) for example in examples]\n",
    "print(images[0].shape)\n",
    "n_rows = (len(examples) - 1) // images_per_row + 1\n",
    "row_images = []\n",
    "n_empty = n_rows * images_per_row - len(examples)\n",
    "images.append(np.zeros((size, size * n_empty)))\n",
    "print(images[65].shape)\n",
    "for row in range(n_rows):\n",
    "    rimages = images[row * images_per_row : (row + 1) * images_per_row]\n",
    "    row_images.append(np.concatenate(rimages, axis=1))\n",
    "image = np.concatenate(row_images, axis=0)\n",
    "print(image.shape)\n",
    "plt.imshow(image, cmap = matplotlib.cm.binary)\n",
    "plt.axis(\"off\")\n",
    "print(row_images[1].shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#필요한 hyperparameter\n",
    "learning_rate = 0.01\n",
    "batch_size = 128\n",
    "n_epochs = 30\n",
    "\n",
    "#홀더.\n",
    "X = tf.placeholder(tf.float32, shape=[batch_size, 784], name='X_placeholder')\n",
    "Y = tf.placeholder(tf.float32, shape=[batch_size, 10], name='Y_placeholder')\n",
    "\n",
    "#weights and bias\n",
    "W = tf.Variable(tf.random_normal(shape=[784, 10], stddev=0.01), name='weights')\n",
    "b = tf.Variable(tf.zeros([1, 10]), name='bias')\n",
    "\n",
    "#logits\n",
    "logits = tf.matmul(X,W) + b\n",
    "\n",
    "#loss\n",
    "entropy = tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y, name='loss')\n",
    "loss = tf.reduce_mean(entropy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "import time\n",
    "\n",
    "now = datetime.now().strftime(\"%Y%m%d%H%M\")\n",
    "root_logdir = \"/Users/yoon/Desktop/87_105_110_107/TP/MNIST/graphs\"\n",
    "logdir = \"{}/run-{}/\".format(root_logdir, now)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0  평균 손실: 1.28991\n",
      "Epoch 1  평균 손실: 0.73380\n",
      "Epoch 2  평균 손실: 0.60127\n",
      "Epoch 3  평균 손실: 0.53669\n",
      "Epoch 4  평균 손실: 0.49758\n",
      "Epoch 5  평균 손실: 0.47117\n",
      "Epoch 6  평균 손실: 0.45064\n",
      "Epoch 7  평균 손실: 0.43623\n",
      "Epoch 8  평균 손실: 0.42335\n",
      "Epoch 9  평균 손실: 0.41348\n",
      "Epoch 10  평균 손실: 0.40439\n",
      "Epoch 11  평균 손실: 0.39662\n",
      "Epoch 12  평균 손실: 0.38988\n",
      "Epoch 13  평균 손실: 0.38452\n",
      "Epoch 14  평균 손실: 0.37815\n",
      "Epoch 15  평균 손실: 0.37417\n",
      "Epoch 16  평균 손실: 0.37119\n",
      "Epoch 17  평균 손실: 0.36574\n",
      "Epoch 18  평균 손실: 0.36250\n",
      "Epoch 19  평균 손실: 0.36074\n",
      "Epoch 20  평균 손실: 0.35649\n",
      "Epoch 21  평균 손실: 0.35284\n",
      "Epoch 22  평균 손실: 0.35066\n",
      "Epoch 23  평균 손실: 0.34943\n",
      "Epoch 24  평균 손실: 0.34674\n",
      "Epoch 25  평균 손실: 0.34391\n",
      "Epoch 26  평균 손실: 0.34211\n",
      "Epoch 27  평균 손실: 0.33945\n",
      "Epoch 28  평균 손실: 0.33786\n",
      "Epoch 29  평균 손실: 0.33695\n",
      "총 소요 시간: 20.80초\n",
      "최적화 완료!\n",
      "정확도: 0.91180\n"
     ]
    }
   ],
   "source": [
    "with tf.Session() as sess:\n",
    "    writer = tf.summary.FileWriter(logdir, sess.graph)\n",
    "    \n",
    "    start_time = time.time()\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    #변수 설정\n",
    "    n_batches = int(mnist.train.num_examples/batch_size)\n",
    "    for i in range(n_epochs): #모델을 n_epochs 만큼 훈련\n",
    "        total_loss = 0\n",
    "        \n",
    "        for _ in range(n_batches):\n",
    "            X_batch, Y_batch = mnist.train.next_batch(batch_size)\n",
    "            _, loss_batch = sess.run([optimizer, loss], feed_dict={X: X_batch, Y:Y_batch})\n",
    "            total_loss += loss_batch\n",
    "            \n",
    "        print('Epoch {}  평균 손실: {:3.5f}'.format(i, total_loss/n_batches))\n",
    "        \n",
    "    print('총 소요 시간: {:5.2f}초'.format(time.time() - start_time))\n",
    "    \n",
    "    print('최적화 완료!')\n",
    "    \n",
    "    #테스트!\n",
    "    \n",
    "    preds = tf.nn.softmax(logits) #정규화\n",
    "    correct_preds = tf.equal(tf.argmax(preds, 1), tf.argmax(Y, 1))\n",
    "    accurate_ones = tf.reduce_sum(tf.cast(correct_preds, tf.float32))\n",
    "    \n",
    "    n_batches = int(mnist.test.num_examples/batch_size)\n",
    "    total_correct_preds = 0.0\n",
    "    \n",
    "    for i in range(n_batches):\n",
    "        X_batch, Y_batch = mnist.test.next_batch(batch_size)\n",
    "        accurate_ones_batch = sess.run(accurate_ones, feed_dict={X: X_batch, Y:Y_batch})\n",
    "        total_correct_preds += accurate_ones_batch\n",
    "        \n",
    "    print('정확도: {:1.5f}'.format(total_correct_preds/mnist.test.num_examples))\n",
    "    \n",
    "    writer.close()"
   ]
  }
 ],
 "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": 2
}