tensorflow_dnn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# to make this notebook's output stable across runs\n",
    "def reset_graph(seed=42):\n",
    "    tf.reset_default_graph()\n",
    "    tf.set_random_seed(seed)\n",
    "    np.random.seed(seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Data stuff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_iris\n",
    "data = load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['sepal length (cm)',\n",
       " 'sepal width (cm)',\n",
       " 'petal length (cm)',\n",
       " 'petal width (cm)']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['sepal length (cm)',\n",
       " 'sepal width (cm)',\n",
       " 'petal length (cm)',\n",
       " 'petal width (cm)']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_data = data.data\n",
    "y_data = data.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_data, y_data, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((112, 4), (38, 4))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Next Batch function "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def next_batch(num, data, labels):\n",
    "    '''\n",
    "    Return a total of `num` random samples and labels. \n",
    "    '''\n",
    "    idx = np.arange(0 , len(data.data))\n",
    "    np.random.shuffle(idx)\n",
    "    idx = idx[:num]\n",
    "    data_shuffle = [data[ i] for i in idx]\n",
    "    labels_shuffle = [labels[ i] for i in idx]\n",
    "\n",
    "    return np.asarray(data_shuffle), np.asarray(labels_shuffle)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Placeholders for input and output data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Lets reset the graph first for undesired outputs\n",
    "reset_graph()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_features = 4\n",
    "\n",
    "X = tf.placeholder(tf.float32, shape=(None, n_features), name=\"X\")\n",
    "y = tf.placeholder(tf.int64, shape=(None), name=\"y\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Define the deep network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_hidden1 = 400\n",
    "n_hidden2 = 200\n",
    "n_outputs = 3\n",
    "\n",
    "with tf.name_scope(\"dnn\"):\n",
    "    hidden1 = tf.layers.dense(X, n_hidden1, name=\"hidden1\", activation=tf.nn.relu)\n",
    "    hidden2 = tf.layers.dense(hidden1, n_hidden2, name=\"hidden2\", activation=tf.nn.relu)\n",
    "    logits  = tf.layers.dense(hidden2, n_outputs, name=\"outputs\")   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. What is the loss function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope(\"loss\"):\n",
    "    xentropy = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=y, logits=logits)\n",
    "    loss = tf.reduce_mean(xentropy, name=\"loss\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. What is the optimizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "learning_rate = 0.001\n",
    "\n",
    "with tf.name_scope(\"Training\"):\n",
    "    #optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)\n",
    "    optimizer = tf.train.MomentumOptimizer(learning_rate=learning_rate, momentum=0.9, use_nesterov=True)\n",
    "    training_op = optimizer.minimize(loss)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope(\"eval\"):\n",
    "    correct = tf.nn.in_top_k(logits, y, 1)\n",
    "    accuracy = tf.reduce_mean(tf.cast(correct, tf.float32))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Execution Phase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 Train accuracy: 0.62\n",
      "1 Train accuracy: 0.54\n",
      "2 Train accuracy: 0.64\n",
      "3 Train accuracy: 0.64\n",
      "4 Train accuracy: 0.7\n",
      "5 Train accuracy: 0.62\n",
      "6 Train accuracy: 0.66\n",
      "7 Train accuracy: 0.66\n",
      "8 Train accuracy: 0.54\n",
      "9 Train accuracy: 0.7\n",
      "10 Train accuracy: 0.68\n",
      "11 Train accuracy: 0.7\n",
      "12 Train accuracy: 0.62\n",
      "13 Train accuracy: 0.94\n",
      "14 Train accuracy: 0.82\n",
      "15 Train accuracy: 0.94\n",
      "16 Train accuracy: 0.9\n",
      "17 Train accuracy: 0.86\n",
      "18 Train accuracy: 0.9\n",
      "19 Train accuracy: 0.84\n",
      "20 Train accuracy: 0.92\n",
      "21 Train accuracy: 0.66\n",
      "22 Train accuracy: 0.76\n",
      "23 Train accuracy: 0.94\n",
      "24 Train accuracy: 0.86\n",
      "25 Train accuracy: 0.88\n",
      "26 Train accuracy: 0.94\n",
      "27 Train accuracy: 0.98\n",
      "28 Train accuracy: 0.96\n",
      "29 Train accuracy: 0.96\n",
      "30 Train accuracy: 0.92\n",
      "31 Train accuracy: 0.88\n",
      "32 Train accuracy: 0.98\n",
      "33 Train accuracy: 1.0\n",
      "34 Train accuracy: 0.98\n",
      "35 Train accuracy: 0.98\n",
      "36 Train accuracy: 0.98\n",
      "37 Train accuracy: 0.94\n",
      "38 Train accuracy: 0.88\n",
      "39 Train accuracy: 0.88\n",
      "40 Train accuracy: 0.92\n",
      "41 Train accuracy: 0.9\n",
      "42 Train accuracy: 0.96\n",
      "43 Train accuracy: 0.94\n",
      "44 Train accuracy: 0.92\n",
      "45 Train accuracy: 0.94\n",
      "46 Train accuracy: 0.96\n",
      "47 Train accuracy: 0.94\n",
      "48 Train accuracy: 0.94\n",
      "49 Train accuracy: 0.94\n",
      "50 Train accuracy: 0.92\n",
      "51 Train accuracy: 0.94\n",
      "52 Train accuracy: 0.94\n",
      "53 Train accuracy: 1.0\n",
      "54 Train accuracy: 0.94\n",
      "55 Train accuracy: 0.98\n",
      "56 Train accuracy: 0.96\n",
      "57 Train accuracy: 0.92\n",
      "58 Train accuracy: 0.9\n",
      "59 Train accuracy: 0.92\n",
      "60 Train accuracy: 1.0\n",
      "61 Train accuracy: 0.96\n",
      "62 Train accuracy: 0.96\n",
      "63 Train accuracy: 1.0\n",
      "64 Train accuracy: 0.96\n",
      "65 Train accuracy: 0.98\n",
      "66 Train accuracy: 0.98\n",
      "67 Train accuracy: 0.96\n",
      "68 Train accuracy: 1.0\n",
      "69 Train accuracy: 0.96\n",
      "70 Train accuracy: 0.96\n",
      "71 Train accuracy: 0.94\n",
      "72 Train accuracy: 0.96\n",
      "73 Train accuracy: 1.0\n",
      "74 Train accuracy: 0.98\n",
      "75 Train accuracy: 0.98\n",
      "76 Train accuracy: 0.96\n",
      "77 Train accuracy: 0.98\n",
      "78 Train accuracy: 0.98\n",
      "79 Train accuracy: 1.0\n",
      "80 Train accuracy: 0.94\n",
      "81 Train accuracy: 0.98\n",
      "82 Train accuracy: 0.98\n",
      "83 Train accuracy: 0.98\n",
      "84 Train accuracy: 0.98\n",
      "85 Train accuracy: 0.96\n",
      "86 Train accuracy: 0.96\n",
      "87 Train accuracy: 0.94\n",
      "88 Train accuracy: 0.98\n",
      "89 Train accuracy: 0.98\n",
      "90 Train accuracy: 0.98\n",
      "91 Train accuracy: 0.98\n",
      "92 Train accuracy: 0.96\n",
      "93 Train accuracy: 0.98\n",
      "94 Train accuracy: 0.96\n",
      "95 Train accuracy: 0.98\n",
      "96 Train accuracy: 0.96\n",
      "97 Train accuracy: 0.98\n",
      "98 Train accuracy: 0.96\n",
      "99 Train accuracy: 0.98\n",
      "100 Train accuracy: 0.96\n",
      "101 Train accuracy: 1.0\n",
      "102 Train accuracy: 0.96\n",
      "103 Train accuracy: 0.96\n",
      "104 Train accuracy: 0.96\n",
      "105 Train accuracy: 1.0\n",
      "106 Train accuracy: 0.98\n",
      "107 Train accuracy: 0.98\n",
      "108 Train accuracy: 0.96\n",
      "109 Train accuracy: 0.94\n",
      "110 Train accuracy: 0.98\n",
      "111 Train accuracy: 0.98\n",
      "112 Train accuracy: 0.96\n",
      "113 Train accuracy: 0.96\n",
      "114 Train accuracy: 0.96\n",
      "115 Train accuracy: 0.98\n",
      "116 Train accuracy: 0.98\n",
      "117 Train accuracy: 0.94\n",
      "118 Train accuracy: 1.0\n",
      "119 Train accuracy: 0.98\n",
      "120 Train accuracy: 1.0\n",
      "121 Train accuracy: 0.98\n",
      "122 Train accuracy: 1.0\n",
      "123 Train accuracy: 1.0\n",
      "124 Train accuracy: 0.96\n",
      "125 Train accuracy: 0.98\n",
      "126 Train accuracy: 0.96\n",
      "127 Train accuracy: 1.0\n",
      "128 Train accuracy: 0.96\n",
      "129 Train accuracy: 0.98\n",
      "130 Train accuracy: 0.98\n",
      "131 Train accuracy: 0.98\n",
      "132 Train accuracy: 0.98\n",
      "133 Train accuracy: 0.94\n",
      "134 Train accuracy: 0.98\n",
      "135 Train accuracy: 0.98\n",
      "136 Train accuracy: 0.98\n",
      "137 Train accuracy: 0.96\n",
      "138 Train accuracy: 0.98\n",
      "139 Train accuracy: 0.94\n",
      "140 Train accuracy: 0.94\n",
      "141 Train accuracy: 0.96\n",
      "142 Train accuracy: 1.0\n",
      "143 Train accuracy: 0.98\n",
      "144 Train accuracy: 0.98\n",
      "145 Train accuracy: 0.98\n",
      "146 Train accuracy: 0.98\n",
      "147 Train accuracy: 0.96\n",
      "148 Train accuracy: 0.98\n",
      "149 Train accuracy: 0.96\n",
      "150 Train accuracy: 0.98\n",
      "151 Train accuracy: 0.98\n",
      "152 Train accuracy: 0.96\n",
      "153 Train accuracy: 0.94\n",
      "154 Train accuracy: 0.96\n",
      "155 Train accuracy: 0.98\n",
      "156 Train accuracy: 0.98\n",
      "157 Train accuracy: 0.96\n",
      "158 Train accuracy: 1.0\n",
      "159 Train accuracy: 0.96\n",
      "160 Train accuracy: 0.98\n",
      "161 Train accuracy: 0.98\n",
      "162 Train accuracy: 1.0\n",
      "163 Train accuracy: 0.98\n",
      "164 Train accuracy: 0.98\n",
      "165 Train accuracy: 0.96\n",
      "166 Train accuracy: 0.96\n",
      "167 Train accuracy: 0.96\n",
      "168 Train accuracy: 0.98\n",
      "169 Train accuracy: 0.96\n",
      "170 Train accuracy: 0.98\n",
      "171 Train accuracy: 1.0\n",
      "172 Train accuracy: 1.0\n",
      "173 Train accuracy: 0.96\n",
      "174 Train accuracy: 0.94\n",
      "175 Train accuracy: 1.0\n",
      "176 Train accuracy: 0.98\n",
      "177 Train accuracy: 0.98\n",
      "178 Train accuracy: 0.98\n",
      "179 Train accuracy: 0.96\n",
      "180 Train accuracy: 0.98\n",
      "181 Train accuracy: 1.0\n",
      "182 Train accuracy: 0.98\n",
      "183 Train accuracy: 0.98\n",
      "184 Train accuracy: 1.0\n",
      "185 Train accuracy: 0.98\n",
      "186 Train accuracy: 0.94\n",
      "187 Train accuracy: 0.98\n",
      "188 Train accuracy: 0.98\n",
      "189 Train accuracy: 0.98\n",
      "190 Train accuracy: 1.0\n",
      "191 Train accuracy: 0.96\n",
      "192 Train accuracy: 0.98\n",
      "193 Train accuracy: 1.0\n",
      "194 Train accuracy: 1.0\n",
      "195 Train accuracy: 0.98\n",
      "196 Train accuracy: 0.98\n",
      "197 Train accuracy: 0.98\n",
      "198 Train accuracy: 0.96\n",
      "199 Train accuracy: 0.98\n",
      "200 Train accuracy: 0.98\n",
      "201 Train accuracy: 0.98\n",
      "202 Train accuracy: 0.98\n",
      "203 Train accuracy: 0.96\n",
      "204 Train accuracy: 1.0\n",
      "205 Train accuracy: 0.96\n",
      "206 Train accuracy: 0.98\n",
      "207 Train accuracy: 0.96\n",
      "208 Train accuracy: 0.96\n",
      "209 Train accuracy: 1.0\n",
      "210 Train accuracy: 1.0\n",
      "211 Train accuracy: 0.98\n",
      "212 Train accuracy: 0.98\n",
      "213 Train accuracy: 0.94\n",
      "214 Train accuracy: 0.96\n",
      "215 Train accuracy: 0.96\n",
      "216 Train accuracy: 0.98\n",
      "217 Train accuracy: 0.98\n",
      "218 Train accuracy: 1.0\n",
      "219 Train accuracy: 0.98\n",
      "220 Train accuracy: 1.0\n",
      "221 Train accuracy: 0.96\n",
      "222 Train accuracy: 0.98\n",
      "223 Train accuracy: 0.98\n",
      "224 Train accuracy: 0.96\n",
      "225 Train accuracy: 1.0\n",
      "226 Train accuracy: 0.94\n",
      "227 Train accuracy: 0.96\n",
      "228 Train accuracy: 0.98\n",
      "229 Train accuracy: 0.94\n",
      "230 Train accuracy: 0.96\n",
      "231 Train accuracy: 0.98\n",
      "232 Train accuracy: 0.96\n",
      "233 Train accuracy: 0.96\n",
      "234 Train accuracy: 0.96\n",
      "235 Train accuracy: 0.98\n",
      "236 Train accuracy: 0.96\n",
      "237 Train accuracy: 1.0\n",
      "238 Train accuracy: 1.0\n",
      "239 Train accuracy: 1.0\n",
      "240 Train accuracy: 1.0\n",
      "241 Train accuracy: 0.98\n",
      "242 Train accuracy: 0.96\n",
      "243 Train accuracy: 0.98\n",
      "244 Train accuracy: 0.96\n",
      "245 Train accuracy: 1.0\n",
      "246 Train accuracy: 0.96\n",
      "247 Train accuracy: 0.98\n",
      "248 Train accuracy: 1.0\n",
      "249 Train accuracy: 1.0\n",
      "250 Train accuracy: 0.98\n",
      "251 Train accuracy: 1.0\n",
      "252 Train accuracy: 1.0\n",
      "253 Train accuracy: 1.0\n",
      "254 Train accuracy: 0.98\n",
      "255 Train accuracy: 1.0\n",
      "256 Train accuracy: 1.0\n",
      "257 Train accuracy: 0.96\n",
      "258 Train accuracy: 0.98\n",
      "259 Train accuracy: 1.0\n",
      "260 Train accuracy: 0.98\n",
      "261 Train accuracy: 0.96\n",
      "262 Train accuracy: 0.96\n",
      "263 Train accuracy: 0.96\n",
      "264 Train accuracy: 0.98\n",
      "265 Train accuracy: 0.98\n",
      "266 Train accuracy: 0.98\n",
      "267 Train accuracy: 1.0\n",
      "268 Train accuracy: 0.96\n",
      "269 Train accuracy: 1.0\n",
      "270 Train accuracy: 0.98\n",
      "271 Train accuracy: 0.98\n",
      "272 Train accuracy: 0.98\n",
      "273 Train accuracy: 0.98\n",
      "274 Train accuracy: 0.98\n",
      "275 Train accuracy: 1.0\n",
      "276 Train accuracy: 1.0\n",
      "277 Train accuracy: 0.98\n",
      "278 Train accuracy: 0.98\n",
      "279 Train accuracy: 0.98\n",
      "280 Train accuracy: 0.98\n",
      "281 Train accuracy: 0.98\n",
      "282 Train accuracy: 0.98\n",
      "283 Train accuracy: 0.98\n",
      "284 Train accuracy: 0.98\n",
      "285 Train accuracy: 0.98\n",
      "286 Train accuracy: 0.98\n",
      "287 Train accuracy: 0.98\n",
      "288 Train accuracy: 1.0\n",
      "289 Train accuracy: 1.0\n",
      "290 Train accuracy: 0.96\n",
      "291 Train accuracy: 1.0\n",
      "292 Train accuracy: 0.98\n",
      "293 Train accuracy: 0.96\n",
      "294 Train accuracy: 0.98\n",
      "295 Train accuracy: 1.0\n",
      "296 Train accuracy: 0.98\n",
      "297 Train accuracy: 0.98\n",
      "298 Train accuracy: 0.98\n",
      "299 Train accuracy: 0.98\n",
      "300 Train accuracy: 0.98\n",
      "301 Train accuracy: 0.98\n",
      "302 Train accuracy: 0.98\n",
      "303 Train accuracy: 0.98\n",
      "304 Train accuracy: 0.94\n",
      "305 Train accuracy: 0.96\n",
      "306 Train accuracy: 0.98\n",
      "307 Train accuracy: 1.0\n",
      "308 Train accuracy: 0.96\n",
      "309 Train accuracy: 0.98\n",
      "310 Train accuracy: 0.94\n",
      "311 Train accuracy: 0.98\n",
      "312 Train accuracy: 0.98\n",
      "313 Train accuracy: 0.98\n",
      "314 Train accuracy: 0.98\n",
      "315 Train accuracy: 1.0\n",
      "316 Train accuracy: 1.0\n",
      "317 Train accuracy: 0.98\n",
      "318 Train accuracy: 1.0\n",
      "319 Train accuracy: 0.96\n",
      "320 Train accuracy: 1.0\n",
      "321 Train accuracy: 1.0\n",
      "322 Train accuracy: 0.98\n",
      "323 Train accuracy: 1.0\n",
      "324 Train accuracy: 1.0\n",
      "325 Train accuracy: 0.96\n",
      "326 Train accuracy: 0.98\n",
      "327 Train accuracy: 0.98\n",
      "328 Train accuracy: 0.98\n",
      "329 Train accuracy: 0.96\n",
      "330 Train accuracy: 0.96\n",
      "331 Train accuracy: 1.0\n",
      "332 Train accuracy: 0.96\n",
      "333 Train accuracy: 1.0\n",
      "334 Train accuracy: 0.96\n",
      "335 Train accuracy: 1.0\n",
      "336 Train accuracy: 0.98\n",
      "337 Train accuracy: 1.0\n",
      "338 Train accuracy: 0.98\n",
      "339 Train accuracy: 0.98\n",
      "340 Train accuracy: 1.0\n",
      "341 Train accuracy: 0.96\n",
      "342 Train accuracy: 0.98\n",
      "343 Train accuracy: 1.0\n",
      "344 Train accuracy: 0.98\n",
      "345 Train accuracy: 0.96\n",
      "346 Train accuracy: 0.98\n",
      "347 Train accuracy: 0.98\n",
      "348 Train accuracy: 0.98\n",
      "349 Train accuracy: 0.96\n",
      "350 Train accuracy: 0.98\n",
      "351 Train accuracy: 1.0\n",
      "352 Train accuracy: 0.98\n",
      "353 Train accuracy: 0.96\n",
      "354 Train accuracy: 0.98\n",
      "355 Train accuracy: 1.0\n",
      "356 Train accuracy: 1.0\n",
      "357 Train accuracy: 0.98\n",
      "358 Train accuracy: 0.96\n",
      "359 Train accuracy: 0.96\n",
      "360 Train accuracy: 0.98\n",
      "361 Train accuracy: 1.0\n",
      "362 Train accuracy: 1.0\n",
      "363 Train accuracy: 0.96\n",
      "364 Train accuracy: 0.98\n",
      "365 Train accuracy: 0.98\n",
      "366 Train accuracy: 0.98\n",
      "367 Train accuracy: 0.96\n",
      "368 Train accuracy: 0.98\n",
      "369 Train accuracy: 0.98\n",
      "370 Train accuracy: 1.0\n",
      "371 Train accuracy: 0.96\n",
      "372 Train accuracy: 0.98\n",
      "373 Train accuracy: 1.0\n",
      "374 Train accuracy: 0.96\n",
      "375 Train accuracy: 0.96\n",
      "376 Train accuracy: 0.96\n",
      "377 Train accuracy: 0.98\n",
      "378 Train accuracy: 0.96\n",
      "379 Train accuracy: 1.0\n",
      "380 Train accuracy: 1.0\n",
      "381 Train accuracy: 1.0\n",
      "382 Train accuracy: 0.98\n",
      "383 Train accuracy: 0.98\n",
      "384 Train accuracy: 1.0\n",
      "385 Train accuracy: 1.0\n",
      "386 Train accuracy: 0.98\n",
      "387 Train accuracy: 1.0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "388 Train accuracy: 1.0\n",
      "389 Train accuracy: 0.98\n",
      "390 Train accuracy: 1.0\n",
      "391 Train accuracy: 1.0\n",
      "392 Train accuracy: 0.98\n",
      "393 Train accuracy: 0.98\n",
      "394 Train accuracy: 0.98\n",
      "395 Train accuracy: 0.96\n",
      "396 Train accuracy: 0.94\n",
      "397 Train accuracy: 0.98\n",
      "398 Train accuracy: 0.98\n",
      "399 Train accuracy: 0.96\n",
      "400 Train accuracy: 0.98\n",
      "401 Train accuracy: 0.98\n",
      "402 Train accuracy: 1.0\n",
      "403 Train accuracy: 1.0\n",
      "404 Train accuracy: 0.98\n",
      "405 Train accuracy: 0.96\n",
      "406 Train accuracy: 1.0\n",
      "407 Train accuracy: 0.96\n",
      "408 Train accuracy: 0.96\n",
      "409 Train accuracy: 0.98\n",
      "410 Train accuracy: 0.98\n",
      "411 Train accuracy: 0.98\n",
      "412 Train accuracy: 1.0\n",
      "413 Train accuracy: 0.96\n",
      "414 Train accuracy: 0.98\n",
      "415 Train accuracy: 0.98\n",
      "416 Train accuracy: 0.98\n",
      "417 Train accuracy: 0.98\n",
      "418 Train accuracy: 1.0\n",
      "419 Train accuracy: 1.0\n",
      "420 Train accuracy: 0.98\n",
      "421 Train accuracy: 1.0\n",
      "422 Train accuracy: 0.96\n",
      "423 Train accuracy: 0.98\n",
      "424 Train accuracy: 0.98\n",
      "425 Train accuracy: 0.98\n",
      "426 Train accuracy: 0.94\n",
      "427 Train accuracy: 0.98\n",
      "428 Train accuracy: 1.0\n",
      "429 Train accuracy: 0.98\n",
      "430 Train accuracy: 1.0\n",
      "431 Train accuracy: 0.98\n",
      "432 Train accuracy: 0.98\n",
      "433 Train accuracy: 0.98\n",
      "434 Train accuracy: 0.98\n",
      "435 Train accuracy: 0.98\n",
      "436 Train accuracy: 0.98\n",
      "437 Train accuracy: 1.0\n",
      "438 Train accuracy: 1.0\n",
      "439 Train accuracy: 0.98\n",
      "440 Train accuracy: 0.96\n",
      "441 Train accuracy: 0.98\n",
      "442 Train accuracy: 0.98\n",
      "443 Train accuracy: 0.94\n",
      "444 Train accuracy: 0.98\n",
      "445 Train accuracy: 0.98\n",
      "446 Train accuracy: 1.0\n",
      "447 Train accuracy: 0.98\n",
      "448 Train accuracy: 1.0\n",
      "449 Train accuracy: 0.98\n",
      "450 Train accuracy: 1.0\n",
      "451 Train accuracy: 0.96\n",
      "452 Train accuracy: 0.96\n",
      "453 Train accuracy: 1.0\n",
      "454 Train accuracy: 0.98\n",
      "455 Train accuracy: 0.98\n",
      "456 Train accuracy: 0.98\n",
      "457 Train accuracy: 0.96\n",
      "458 Train accuracy: 0.98\n",
      "459 Train accuracy: 0.98\n",
      "460 Train accuracy: 1.0\n",
      "461 Train accuracy: 0.98\n",
      "462 Train accuracy: 1.0\n",
      "463 Train accuracy: 0.96\n",
      "464 Train accuracy: 0.96\n",
      "465 Train accuracy: 0.98\n",
      "466 Train accuracy: 0.98\n",
      "467 Train accuracy: 0.98\n",
      "468 Train accuracy: 0.96\n",
      "469 Train accuracy: 1.0\n",
      "470 Train accuracy: 0.98\n",
      "471 Train accuracy: 0.96\n",
      "472 Train accuracy: 0.98\n",
      "473 Train accuracy: 0.96\n",
      "474 Train accuracy: 0.98\n",
      "475 Train accuracy: 0.98\n",
      "476 Train accuracy: 0.96\n",
      "477 Train accuracy: 1.0\n",
      "478 Train accuracy: 0.96\n",
      "479 Train accuracy: 0.98\n",
      "480 Train accuracy: 0.96\n",
      "481 Train accuracy: 0.92\n",
      "482 Train accuracy: 0.96\n",
      "483 Train accuracy: 0.96\n",
      "484 Train accuracy: 0.98\n",
      "485 Train accuracy: 0.98\n",
      "486 Train accuracy: 0.96\n",
      "487 Train accuracy: 0.98\n",
      "488 Train accuracy: 0.96\n",
      "489 Train accuracy: 0.98\n",
      "490 Train accuracy: 0.98\n",
      "491 Train accuracy: 0.98\n",
      "492 Train accuracy: 0.92\n",
      "493 Train accuracy: 0.98\n",
      "494 Train accuracy: 0.98\n",
      "495 Train accuracy: 0.94\n",
      "496 Train accuracy: 0.96\n",
      "497 Train accuracy: 0.98\n",
      "498 Train accuracy: 0.98\n",
      "499 Train accuracy: 1.0\n"
     ]
    }
   ],
   "source": [
    "n_epochs=500\n",
    "batch_size=50\n",
    "\n",
    "init = tf.global_variables_initializer()\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    init.run()\n",
    "    for epoch in range(n_epochs):\n",
    "        for iteration in range(int(np.ceil(len(X_train)/batch_size))):\n",
    "            X_batch, y_batch = next_batch(batch_size, X_train, y_train)\n",
    "            sess.run(training_op, feed_dict={X: X_batch, y: y_batch})\n",
    "        acc_train = accuracy.eval(feed_dict={X: X_batch, y: y_batch})\n",
    "        print(epoch, \"Train accuracy:\", acc_train)\n",
    "    save_path = saver.save(sess, \"./my_model_final,ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}