dice_cnn_API.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2D convolution network with Adam optimizer (Functional model)\n",
    "\n",
    "## This is very first functional model instead of sequential one. Due to that facts there are few issues related with this model:\n",
    "* model cannot use big batch_size, there is an error related with OOM (Out of memory)\n",
    "* model got troubles with uninitialization, probable because of using tensorflow backend instead of theano\n",
    "* model is NOT learning. After whole 5 epochs I decided to interupt because there is no sense in running it it accuracy and loss hadn't changed since beginning\n",
    "\n",
    "## My previous assumption was wrong, actually the model is learning and it is performing very well\n",
    "In comparison to the sequential model with same data set and optimizer (Adam), same layers and layers structure the difference is pretty distinctive. However, there are little differences in the count of params to be learned (25M compare to 17M in sequential one) which presumably made that difference in assumptions. I have to dive into documentation to work on this.\n",
    "##### Nonetheless here are results:\n",
    "Functional API model: \n",
    "* accuracy: 0.9517\n",
    "* loss: 0.1445\n",
    "\n",
    "Sequential model:\n",
    "* accuracy: 0.5565\n",
    "* loss: 1.0111"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['woodOnRed', 'blackOnRed', 'redOnRed', 'redOnRed_white', 'navyOnWhite', 'blackOnBlack', 'whiteOnBlue', 'greenOnWhite', 'navyOnBlue', 'stainOnWhite', 'hardLight', 'whiteOnRed_distance', 'whiteOnBlack', 'whiteOnRed', 'greenOnGreen']\n",
      "100800\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.5/dist-packages/sklearn/cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n",
      "  \"This module will be removed in 0.20.\", DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Model, load_model\n",
    "from keras.layers import Input, Dense\n",
    "from keras.layers import Dropout\n",
    "from keras.layers import Flatten\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.layers.convolutional import Conv2D\n",
    "from keras.layers.convolutional import MaxPooling2D\n",
    "from keras.utils import to_categorical\n",
    "from keras.utils.vis_utils import plot_model\n",
    "from keras import backend as K\n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from PIL import Image\n",
    "\n",
    "import numpy as np\n",
    "from numpy.testing import assert_allclose, assert_equal\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "\n",
    "# initialization and check if dataset got proper size\n",
    "\n",
    "path = '/home/ubuntu/zdjecia/'\n",
    "# path = '/home/oziomek/licencjat/kostki/zdjecia/'\n",
    "listing = os.listdir(path)\n",
    "print(listing)\n",
    "num_samples = 0\n",
    "for folder in os.listdir(path):\n",
    "    for i in range(1, 7):\n",
    "        num_samples += len(os.listdir(path + folder + '/' + str(i)))\n",
    "assert_equal(100800, num_samples)\n",
    "print(num_samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# load all data into numpy array\n",
    "immatrix = np.array(\n",
    "    [np.array(Image.open(path + folder + '/' + str(i) + '/' + file)).flatten() \n",
    "         for folder in os.listdir(path)\n",
    "             for i in range(1, 7) \n",
    "                 for file in os.listdir(path + folder + '/' + str(i))], 'f')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100800, 4096)\n"
     ]
    }
   ],
   "source": [
    "print(immatrix.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "labels = np.ones((num_samples, ), dtype=int)\n",
    "counter = 0\n",
    "for folder in os.listdir(path):\n",
    "    for i in range(1, 7):\n",
    "        samples = len(os.listdir(path + folder + '/' + str(i)))\n",
    "        labels[counter:counter+samples] = i-1\n",
    "        counter += samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# randomize data with const random_state\n",
    "data, label = shuffle(immatrix, labels, random_state=13)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a31daac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 1 dots, number is 3680\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a2c0c4a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 1 dots, number is 316\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a3185390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 2 dots, number is 791\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a3181630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 5 dots, number is 2368\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a319c780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 2 dots, number is 6219\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a29fd518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 2 dots, number is 834\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a31910f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 4 dots, number is 9466\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a2c05278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 2 dots, number is 5378\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a3164fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 4 dots, number is 9700\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a30e9898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 3 dots, number is 7454\n"
     ]
    }
   ],
   "source": [
    "# show randomly chosen image\n",
    "rows, cols = 64, 64\n",
    "labels_values = 6\n",
    "train_data = [data, label]\n",
    "import random\n",
    "for i in range(10):\n",
    "    random_image = random.randint(0, 10080)\n",
    "    img=immatrix[random_image].reshape(rows, cols)\n",
    "    plt.imshow(img)\n",
    "    plt.show()\n",
    "    print(\"There are {} dots, number is {}\".format(labels[random_image] + 1, random_image))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "(X, y) = (train_data[0], train_data[1])\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80640, 64, 64, 1) (20160, 64, 64, 1)\n"
     ]
    }
   ],
   "source": [
    "# reshape to [color_layers][pixels_height][pixels_width], color_layers equal to 2 because photos are in GRAYSCALE\n",
    "X_train = X_train.reshape(X_train.shape[0], rows, cols, 1).astype('float32')\n",
    "X_test = X_test.reshape(X_test.shape[0], rows, cols, 1).astype('float32')\n",
    "\n",
    "# normalize to 0-1\n",
    "X_train /= 255\n",
    "X_test /= 255\n",
    "\n",
    "print(X_train.shape, X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80640, 6) (20160, 6)\n"
     ]
    }
   ],
   "source": [
    "# onehot\n",
    "y_train = to_categorical(y_train, 6)\n",
    "y_test = to_categorical(y_test, 6)\n",
    "print(y_train.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create functional model with 2NN layers\n",
    "\n",
    "visible = Input(shape=(64, 64, 1))\n",
    "\n",
    "conv1 = Conv2D(64, kernel_size=5, padding='same', activation='relu')(visible)\n",
    "pool1 = MaxPooling2D((2, 2))(conv1)\n",
    "conv2 = Conv2D(32, kernel_size=5, padding='same', activation='relu')(pool1)\n",
    "dropout1 = Dropout(0.2)(conv2)\n",
    "flat1 = Flatten()(dropout1)\n",
    "hidden1 = Dense(784, kernel_initializer='random_uniform', activation='relu')(flat1)\n",
    "dropout2 = Dropout(0.2)(hidden1)\n",
    "hidden2 = Dense(196, kernel_initializer='random_uniform', activation='relu')(dropout2)\n",
    "dropout3 = Dropout(0.2)(hidden2)\n",
    "hidden3 = Dense(64, kernel_initializer='random_uniform', activation='relu')(dropout3)\n",
    "predictions = Dense(labels_values, activation='softmax')(hidden3)\n",
    "model = Model(inputs=visible, outputs=predictions)\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         (None, 64, 64, 1)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 64, 64, 64)        1664      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 32, 32, 32)        51232     \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 32, 32, 32)        0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 32768)             0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 784)               25690896  \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 784)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 196)               153860    \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 196)               0         \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 64)                12608     \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 6)                 390       \n",
      "=================================================================\n",
      "Total params: 25,910,650\n",
      "Trainable params: 25,910,650\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# summarize model and apply checkpoints\n",
    "print(model.summary())\n",
    "filepath = 'model2API-{epoch:02d}-{loss:.4f}.h5'\n",
    "checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, save_best_only=True, mode='min')\n",
    "callbacks_list = [checkpoint]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# visualization of a model\n",
    "plot_model(model, to_file='model2API_plot.png', show_shapes=True, show_layer_names=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 80640 samples, validate on 20160 samples\n",
      "Epoch 1/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 1.5968 - acc: 0.3026Epoch 00001: loss improved from inf to 1.59666, saving model to model2API-01-1.5967.h5\n",
      "80640/80640 [==============================] - 46s 569us/step - loss: 1.5967 - acc: 0.3028 - val_loss: 1.3320 - val_acc: 0.4525\n",
      "Epoch 2/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 1.2236 - acc: 0.4712Epoch 00002: loss improved from 1.59666 to 1.22201, saving model to model2API-02-1.2220.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 1.2220 - acc: 0.4716 - val_loss: 1.0512 - val_acc: 0.5166\n",
      "Epoch 3/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.9873 - acc: 0.5527Epoch 00003: loss improved from 1.22201 to 0.98756, saving model to model2API-03-0.9876.h5\n",
      "80640/80640 [==============================] - 45s 562us/step - loss: 0.9876 - acc: 0.5524 - val_loss: 0.9638 - val_acc: 0.5766\n",
      "Epoch 4/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.8165 - acc: 0.6256Epoch 00004: loss improved from 0.98756 to 0.81573, saving model to model2API-04-0.8157.h5\n",
      "80640/80640 [==============================] - 45s 562us/step - loss: 0.8157 - acc: 0.6260 - val_loss: 0.6754 - val_acc: 0.6996\n",
      "Epoch 5/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.6744 - acc: 0.6949Epoch 00005: loss improved from 0.81573 to 0.67414, saving model to model2API-05-0.6741.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.6741 - acc: 0.6949 - val_loss: 0.5557 - val_acc: 0.7666\n",
      "Epoch 6/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.6445 - acc: 0.7160Epoch 00006: loss improved from 0.67414 to 0.64339, saving model to model2API-06-0.6434.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.6434 - acc: 0.7165 - val_loss: 0.4659 - val_acc: 0.8096\n",
      "Epoch 7/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.4990 - acc: 0.7839Epoch 00007: loss improved from 0.64339 to 0.49916, saving model to model2API-07-0.4992.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.4992 - acc: 0.7836 - val_loss: 0.4310 - val_acc: 0.8206\n",
      "Epoch 8/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.4430 - acc: 0.8116Epoch 00008: loss improved from 0.49916 to 0.44263, saving model to model2API-08-0.4426.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.4426 - acc: 0.8118 - val_loss: 0.3530 - val_acc: 0.8585\n",
      "Epoch 9/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.3834 - acc: 0.8377Epoch 00009: loss improved from 0.44263 to 0.38320, saving model to model2API-09-0.3832.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.3832 - acc: 0.8378 - val_loss: 0.3201 - val_acc: 0.8683\n",
      "Epoch 10/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.3369 - acc: 0.8589Epoch 00010: loss improved from 0.38320 to 0.33643, saving model to model2API-10-0.3364.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.3364 - acc: 0.8590 - val_loss: 0.2951 - val_acc: 0.8791\n",
      "Epoch 11/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.2993 - acc: 0.8756Epoch 00011: loss improved from 0.33643 to 0.29951, saving model to model2API-11-0.2995.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.2995 - acc: 0.8756 - val_loss: 0.2799 - val_acc: 0.8846\n",
      "Epoch 12/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.2657 - acc: 0.8898Epoch 00012: loss improved from 0.29951 to 0.26553, saving model to model2API-12-0.2655.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.2655 - acc: 0.8899 - val_loss: 0.2475 - val_acc: 0.9012\n",
      "Epoch 13/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.2427 - acc: 0.9012Epoch 00013: loss improved from 0.26553 to 0.24267, saving model to model2API-13-0.2427.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.2427 - acc: 0.9012 - val_loss: 0.2274 - val_acc: 0.9073\n",
      "Epoch 14/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.2336 - acc: 0.9058Epoch 00014: loss improved from 0.24267 to 0.23383, saving model to model2API-14-0.2338.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.2338 - acc: 0.9058 - val_loss: 0.2118 - val_acc: 0.9155\n",
      "Epoch 15/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1954 - acc: 0.9211Epoch 00015: loss improved from 0.23383 to 0.19542, saving model to model2API-15-0.1954.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.1954 - acc: 0.9211 - val_loss: 0.1843 - val_acc: 0.9298\n",
      "Epoch 16/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1839 - acc: 0.9260Epoch 00016: loss improved from 0.19542 to 0.18338, saving model to model2API-16-0.1834.h5\n",
      "80640/80640 [==============================] - 45s 564us/step - loss: 0.1834 - acc: 0.9262 - val_loss: 0.1868 - val_acc: 0.9254\n",
      "Epoch 17/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1681 - acc: 0.9329Epoch 00017: loss improved from 0.18338 to 0.16813, saving model to model2API-17-0.1681.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.1681 - acc: 0.9328 - val_loss: 0.1825 - val_acc: 0.9315\n",
      "Epoch 18/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1470 - acc: 0.9421Epoch 00018: loss improved from 0.16813 to 0.14702, saving model to model2API-18-0.1470.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.1470 - acc: 0.9420 - val_loss: 0.1663 - val_acc: 0.9371\n",
      "Epoch 19/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1469 - acc: 0.9422Epoch 00019: loss improved from 0.14702 to 0.14666, saving model to model2API-19-0.1467.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.1467 - acc: 0.9423 - val_loss: 0.1572 - val_acc: 0.9432\n",
      "Epoch 20/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1247 - acc: 0.9519Epoch 00020: loss improved from 0.14666 to 0.12469, saving model to model2API-20-0.1247.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.1247 - acc: 0.9520 - val_loss: 0.1528 - val_acc: 0.9432\n",
      "Epoch 21/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1694 - acc: 0.9350Epoch 00021: loss did not improve\n",
      "80640/80640 [==============================] - 45s 560us/step - loss: 0.1693 - acc: 0.9350 - val_loss: 0.1799 - val_acc: 0.9320\n",
      "Epoch 22/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.1131 - acc: 0.9563Epoch 00022: loss improved from 0.12469 to 0.11298, saving model to model2API-22-0.1130.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.1130 - acc: 0.9563 - val_loss: 0.1473 - val_acc: 0.9490\n",
      "Epoch 23/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.0940 - acc: 0.9639Epoch 00023: loss improved from 0.11298 to 0.09414, saving model to model2API-23-0.0941.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.0941 - acc: 0.9638 - val_loss: 0.1381 - val_acc: 0.9530\n",
      "Epoch 24/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.0907 - acc: 0.9660Epoch 00024: loss improved from 0.09414 to 0.09071, saving model to model2API-24-0.0907.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.0907 - acc: 0.9659 - val_loss: 0.1548 - val_acc: 0.9468\n",
      "Epoch 25/25\n",
      "79872/80640 [============================>.] - ETA: 0s - loss: 0.0871 - acc: 0.9676Epoch 00025: loss improved from 0.09071 to 0.08710, saving model to model2API-25-0.0871.h5\n",
      "80640/80640 [==============================] - 45s 563us/step - loss: 0.0871 - acc: 0.9676 - val_loss: 0.1445 - val_acc: 0.9517\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=25, batch_size=1024, verbose=1, callbacks=callbacks_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20160/20160 [==============================] - 6s 281us/step\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(X_test, y_test, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['loss', 'val_loss', 'acc', 'val_acc'])\n",
      "Accuracy: 95.17%\n",
      "Test loss: 0.1444753410412915\n",
      "Test accuracy 0.9517361111111111\n"
     ]
    }
   ],
   "source": [
    "print(history.history.keys())\n",
    "print(\"Accuracy: %.2f%%\" % (score[1]*100))\n",
    "print('Test loss:', score[0])\n",
    "print('Test accuracy', score[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training results:\n",
      "loss: 0.1444753410412915\n",
      "acc: 0.9517361111111111\n"
     ]
    }
   ],
   "source": [
    "print('training results:')\n",
    "for i in range(len(model.metrics_names)):\n",
    "    print(str(model.metrics_names[i]) + \": \" + str(score[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f50a2cf2c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#summarize for accuracy\n",
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEWCAYAAABi5jCmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzt3Xl4VOX5//H3ncm+kw1IIBB2SNgDgivWDZeCiGJB7WaL/rr321q11S62/bZ2+7a2VkXL5Vot7qioiAtYBTEsCmENaxKWhED2PXP//jgDJJCEJGQySeZ+XddcmTnnmTP3cSSfnOec8zyiqhhjjDHHBfi6AGOMMd2LBYMxxpgmLBiMMcY0YcFgjDGmCQsGY4wxTVgwGGOMacKCwZg2EpHHReQ3bWy7V0QuPdvtGOMLFgzGGGOasGAwxhjThAWD6VU8XTh3iMjnIlIhIv8Skb4i8qaIlInIChHp06j9LBHJFpFiEflAREY3WjdRRNZ73vcfIPSUz7pGRDZ63vuxiIzrYM3fFJEcETkqIktFJNmzXETk/0SkQERKRWSTiGR41l0lIls8teWLyI879B/MmGZYMJjeaC5wGTAC+CLwJvBTIBHn//nvAYjICOBZ4AeedcuA10QkWESCgVeAp4A44HnPdvG8dyKwGLgNiAceAZaKSEh7ChWRLwC/A+YB/YF9wHOe1ZcDF3r2I8bTpsiz7l/AbaoaBWQA77Xnc41pjQWD6Y3+rqqHVTUf+BD4RFU3qGo18DIw0dPuRuANVX1HVeuAPwFhwLnANCAI+Kuq1qnqC8CnjT5jIfCIqn6iqg2q+gRQ43lfe9wELFbV9apaA9wNTBeRwUAdEAWMAkRVt6rqQc/76oAxIhKtqsdUdX07P9eYFlkwmN7ocKPnVc28jvQ8T8b5Cx0AVXUDuUCKZ12+Nh1lcl+j54OAH3m6kYpFpBgY6Hlfe5xaQznOUUGKqr4H/AN4ECgQkUUiEu1pOhe4CtgnIitFZHo7P9eYFlkwGH92AOcXPOD06eP8cs8HDgIpnmXHpTZ6ngv8VlVjGz3CVfXZs6whAqdrKh9AVR9Q1cnAGJwupTs8yz9V1dlAEk6X15J2fq4xLbJgMP5sCXC1iFwiIkHAj3C6gz4GVgP1wPdEJEhErgOmNnrvo8DtInKO5yRxhIhcLSJR7azhWeBrIjLBc37if3G6vvaKyBTP9oOACqAacHvOgdwkIjGeLrBSwH0W/x2MacKCwfgtVd0O3Az8HTiCc6L6i6paq6q1wHXAV4GjOOcjXmr03izgmzhdPceAHE/b9tawArgXeBHnKGUo8CXP6micADqG091UBPzRs+4WYK+IlAK345yrMKZTiE3UY4wxpjE7YjDGGNOEBYMxxpgmLBiMMcY0YcFgjDGmiUBfF9BeCQkJOnjwYF+XYYwxPcq6deuOqGpiW9r2uGAYPHgwWVlZvi7DGGN6FBHZd+ZWDutKMsYY04QFgzHGmCYsGIwxxjTR484xNKeuro68vDyqq6t9XYrXhYaGMmDAAIKCgnxdijGml+oVwZCXl0dUVBSDBw+m6WCYvYuqUlRURF5eHmlpab4uxxjTS3mtK0lEFnumJNzcSpsZnqkRs0VkZUc/q7q6mvj4+F4dCgAiQnx8vF8cGRljfMeb5xgeB2a2tFJEYoF/ArNUNR244Ww+rLeHwnH+sp/GGN/xWjCo6iqc4YpbsgB4SVX3e9oXeKsWgOq6Bg4UV+F222iyxhjTGl9elTQC6CMiH4jIOhH5cksNRWShiGSJSFZhYWGHPqy23s2R8hoqa+s7Wm+LiouL+ec//9nu91111VUUFxd3ej3GGHM2fBkMgcBk4GrgCuBeERnRXENVXaSqmaqamZjYpju6TxMR4kIQymsaOlxwS1oKhvr61kNo2bJlxMbGdno9xhhzNnx5VVIeUKSqFUCFiKwCxgM7vPFhroAAwoJdlNd0/hHDXXfdxa5du5gwYQJBQUGEhobSp08ftm3bxo4dO7j22mvJzc2lurqa73//+yxcuBA4ObxHeXk5V155Jeeffz4ff/wxKSkpvPrqq4SFhXV6rcYYcya+DIZXgX+ISCAQDJwD/N/ZbvRXr2Wz5UBps+tqG9zU1buJCGnfbo9JjuYXX0xvcf3vf/97Nm/ezMaNG/nggw+4+uqr2bx584lLShcvXkxcXBxVVVVMmTKFuXPnEh8f32QbO3fu5Nlnn+XRRx9l3rx5vPjii9x8883tqtMYYzqD14JBRJ4FZgAJIpIH/AIIAlDVh1V1q4i8BXyOM5H5Y6ra4qWtncElQh3Q4FZcAd67umfq1KlN7jN44IEHePnllwHIzc1l586dpwVDWloaEyZMAGDy5Mns3bvXa/UZY0xrvBYMqjq/DW3+yMnJzTtFa3/Zu91K9sFS4iOCSY71XjdNRETEiecffPABK1asYPXq1YSHhzNjxoxm70MICQk58dzlclFVVeW1+owxpjV+NVZSQIAQEeyiopPPM0RFRVFWVtbsupKSEvr06UN4eDjbtm1jzZo1nfrZxhjT2XrFkBjtERESyOHSauob3AS6OicX4+PjOe+888jIyCAsLIy+ffueWDdz5kwefvhhRo8ezciRI5k2bVqnfKYxxniLqPasG74yMzP11Il6tm7dyujRo9v0/oqaenYVljMoLpyY8GBvlOh17dlfY4wBEJF1qprZlrZ+1ZUEEBbsIkDEK5etGmNMb+B3wRAgQkRIoFdudDPGmN7A74IBIDIkkJr6Burq3b4uxRhjuh0/DQYXAOVeGDfJGGN6Or8MhtAgF64AobzagsEYY07ll8EgIkSGBFJRU09PuyrLGGO8zS+DAZzzDLUNbmobzv48Q0eH3Qb461//SmVl5VnXYIwxncVvg+H4QHqdcdmqBYMxpjfxuzufjwsJDCDIFUBFdQPxEWdu35rGw25fdtllJCUlsWTJEmpqapgzZw6/+tWvqKioYN68eeTl5dHQ0MC9997L4cOHOXDgABdffDEJCQm8//77nbNzxhhzFnpfMLx5FxzadMZmAqTVN9DgVjTYmcSnRf3GwpW/b3F142G3ly9fzgsvvMDatWtRVWbNmsWqVasoLCwkOTmZN954A3DGUIqJieEvf/kL77//PgkJCe3dU2OM8Qq/7UoCcAUIqtCZ55+XL1/O8uXLmThxIpMmTWLbtm3s3LmTsWPH8s4773DnnXfy4YcfEhMT03kfaowxnaj3HTG08pf9qbS+gd2HykiODSMhMuTMb2jLNlW5++67ue22205bt379epYtW8Y999zDJZdcws9//vNO+UxjjOlMfn3EEBzoIjgw4KzvZ2g87PYVV1zB4sWLKS8vByA/P5+CggIOHDhAeHg4N998M3fccQfr168/7b3GGNMd9L4jhnaKDAmkpKoOVUWkY7O6NR52+8orr2TBggVMnz7d2X5kJE8//TQ5OTnccccdBAQEEBQUxEMPPQTAwoULmTlzJsnJyXby2RjTLXht2G0RWQxcAxSoakYr7aYAq4EvqeoLZ9ru2Q67fariylr2H61kWFIk4cE9Iydt2G1jTHt1l2G3HwdmttZARFzA/cByL9bRqs68n8EYY3oDrwWDqq4Cjp6h2XeBF4ECb9VxJkGuAEKDXFTYMNzGGAP48OSziKQAc4CH2tB2oYhkiUhWYWFhs23Opkvs+LhJ7h4wbpKN7WSM8TZfXpX0V+BOVT3jYEWqukhVM1U1MzEx8bT1oaGhFBUVdfiXZkRIIG5Vqmq791GDqlJUVERoaKivSzHG9GK+PNuaCTznuRIoAbhKROpV9ZX2bmjAgAHk5eXR0tHEmbjdSkFJNVWFgUSHBnVoG10lNDSUAQMG+LoMY0wv5rNgUNW0489F5HHg9Y6EAkBQUBBpaWlnbtiKu/7+X8KCXSy5bfpZbccYY3o6r3UlicizOJehjhSRPBG5VURuF5HbvfWZZ+PcYfFs2H+s23cnGWOMt3ntiEFV57ej7Ve9VUdbnTs0gUdW7iZr31EuGH76eQxjjPEXfj0kRmNTBvchMED4KKfI16UYY4xP+VcwuFu+ACo8OJCJqbGs3nWkCwsyxpjux3+CYfub8JdRUN7yvXTThyawKb+Ekqq6LizMGGO6F/8Jhj6DofwwbHm1xSbnDY3HrbB2z5lu2DbGmN7Lf4IhaTQkjYFNLY/TNyE1ltCgAD627iRjjB/zn2AAyLgOctdAcW6zq0MCXUwZHMfqXXYC2hjjv/wsGOY6P7NfbrHJ9KHxbDtUxpHymi4qyhhjuhf/Coa4IZA8CTa/2GKTc4cmANhRgzHGb/lXMIBz1HBwIxTtan51cjRRIYF8bMFgjPFTfhgM1wHS4lFDoCuAc4bE2/0Mxhi/5X/BEJ0Mg851rk5qYZjuc4fGs7eokvziqi4uzhhjfM//ggGco4Yj2+FwdrOrzx0WD8DHOXbUYIzxP/4ZDGOuBXG12J00IimK+IhgOwFtjPFL/hkMEQkwZIYTDM10JwUECNOGxvPxro7PCmeMMT2VfwYDOFcnFe+D/PXNrj5vaAKHSqvZc6Siiwszxhjf8t9gGHU1uIJhc/NDZJw71HOewbqTjDF+xn+DISwWhl8Om18C9+mztg2KDyc5JtTOMxhj/I43p/ZcLCIFIrK5hfU3icjnIrJJRD4WkfHeqqVFGddB+SHYv7q5+rhgeCIrdxRSXlPf5aUZY4yvePOI4XFgZivr9wAXqepY4NfAIi/W0rwRMyEovMURV2+cOpDymnpe2ZDfxYUZY4zveC0YVHUV0OLEBqr6saoe87xcAwzwVi0tCo6AkVc6czQ0nD45z8SBsYzpH83Ta/bZ1UnGGL/RXc4x3Aq82dJKEVkoIlkiklVYWNi5n5xxPVQdhd0rm/tcbpk+iG2Hyli371gzbzbGmN7H58EgIhfjBMOdLbVR1UWqmqmqmYmJiZ1bwLBLICSmxauTZk9IJiokkKfW7OvczzXGmG7Kp8EgIuOAx4DZquqby38CQ2D0F2Hr61BXfdrq8OBA5k4ewLJNB22OBmOMX/BZMIhIKvAScIuq7vBVHQCMnQu1ZZDzTrOrb56WSl2DsiSr+ZnfjDGmN/Hm5arPAquBkSKSJyK3isjtInK7p8nPgXjgnyKyUUSyvFXLGQ2+EMITWhw7aVhSFNOHxPPMmv00uO0ktDGmdwv01oZVdf4Z1n8D+Ia3Pr9dXIGQfi1seAZqyiEk8rQmt0wfxLeeWc8H2wu4ZHRfHxRpjDFdw+cnn7uNjLlQXwXbm7846rIxfUmKCrGT0MaYXs+C4biB0yA6pcXupCBXAPOnprJyRyH7iyq7uDhjjOk6FgzHBQRA+hzIWQGVzd+XN39qKgEiPLPWjhqMMb2XBUNjGXPBXQfbXm92db+YUC4b3Zcln+ZSXXf6wHvGGNMbWDA0ljwR4oa02J0EzknoY5V1vLn5YBcWZowxXceCoTER56hhzyooL2i2yblD4xmSGMFTq607yRjTO1kwnCpjLqgbsl9pdrWIcNM5g1i/v5jN+SVdXJwxxnifBcOpkkZD0phWu5OunzSA0KAAnvnEjhqMMb2PBUNzMuZC7hoobn4IjJjwIGaPT+GVDQcorT59uG5jjOnJLBiak3Gd8zP7pRab3DJ9EFV1Dby0Lq+LijLGmK5hwdCcuCGQPKnV7qSMlBjGD4zlKZvExxjTy1gwtGTs9XDwMziS02KTW6YNYldhBat3+2bEcGOM8QYLhpakzwGk1e6ka8b1JzY8iGfW7O+6uowxxsssGFoSnQyDzoVPH4PC7c02CQ1yMS9zIG9nH+Jw6emT/BhjTE9kwdCaK/8AqrD4Cshd22yTm85Jpd6tPLfWJvExxvQOFgyt6ZcBty6HsD7wxCzY8fZpTQbFR3DhiESeXbuf+ga3D4o0xpjOZcFwJnFp8PXlkDgSnp0PG54+rckt0wZxqLSaFVubH0bDGGN6Em9O7blYRApEZHML60VEHhCRHBH5XEQmeauWsxaZCF99HdIuhFe/DR/+2eli8vjCqCRSYsN42ibxMcb0At48YngcmNnK+iuB4Z7HQuAhL9Zy9kKiYMESGHsDvHsfvHknuJ2uI1eAsOCcVP6bc4RdheU+LtQYY86O14JBVVcBzc9445gNPKmONUCsiPT3Vj2dIjAY5iyC6d+BtY/Ai1+H+hoA5mUOJMgldumqMabH8+U5hhSg8aU8eZ5lpxGRhSKSJSJZhYWFXVJciwIC4IrfwmW/huyX4ZnrobqUxKgQZmb054V1uVTV2iQ+xpieq0ecfFbVRaqaqaqZiYmJvi7Hcd73YM4jsO9jePwqKDvMLdMGUVpdzysb831dnTHGdJgvgyEfGNjo9QDPsp5j/Jdg/n+gaDf86zKmRB1l3IAYHnh3px01GGN6LF8Gw1Lgy56rk6YBJara8+bLHH4pfOU1qC1HFl/B786p52BJNY99uNvXlRljTId483LVZ4HVwEgRyRORW0XkdhG53dNkGbAbyAEeBb7lrVq8bsBk516H4HDSly/g20OP8NDKXRTYMBnGmB5IetqQ0ZmZmZqVleXrMppXdggeuYjKhAzG77yV6yYO4P7rx/m6KmOMQUTWqWpmW9r2iJPPPUZUPxh7PeH7V3Lb1HiWrMsl+4DNC22M6VksGDpb+nXgruNb/XcQGxbEb9/YahP5GGN6FAuGzpYyCWJSCd+5lB9eNoKPdxXxro2hZIzpQSwYOpsIpF8Lu95j/thohiZG8L/LtlJbbyOvGmN6BgsGb0ifA+56gnYu42dXj2b3kQqe+cQG2DPG9AwWDN6QPBFiB0H2y1w8MokLhifw1xU7Ka6s9XVlxhhzRhYM3iDiHDXs/gCpOsbPrh5NWXUdD7yb4+vKjDHmjCwYvMXTncS21xnVL5obpwzkydV72W3DchtjujkLBm/pPx76pDkjsAI/vGwEIYEB/P7NbT4uzBhjWtemYBCR74tItGdco3+JyHoRudzbxfVoJ7qTVkJFEUlRoXzr4mEs33KY1buKfF2dMca0qK1HDF9X1VLgcqAPcAvwe69V1VukzwFtgG2vAXDr+WmkxIbxmze20OC2m96MMd1TW4NBPD+vAp5S1exGy0xL+o2FuKEnupNCg1z8ZOZIsg+U8tL6PB8XZ4wxzWtrMKwTkeU4wfC2iEQBdsfWmRzvTtqzCiqOADBrfDITU2P549vbqayt93GBxhhzurYGw63AXcAUVa0EgoCvea2q3iR9Dqgbti4FQES45+oxFJTV8PBKm7PBGNP9tDUYpgPbVbVYRG4G7gFs2NC26JsO8cNPdCcBTB7Uh2vG9WfRql0cLKnyYXHGGHO6tgbDQ0CliIwHfgTsAp70WlW9yfHupL3/hfKTg+ndOXMUboU/vr3dh8UZY8zp2hoM9eqMHT0b+IeqPghEea+sXuaU7iSAgXHh3Hp+Gi+tz+fzvGIfFmeMMU21NRjKRORunMtU3xCRAJzzDK0SkZkisl1EckTkrmbWp4rI+yKyQUQ+F5Gr2ld+D5E0GhJGQvYrTRZ/a8ZQ4iOC+c3rNmeDMab7aGsw3AjU4NzPcAgYAPyxtTeIiAt4ELgSGAPMF5ExpzS7B1iiqhOBLwH/bEftPUfj7qSywycWR4UG8T+Xj2Dt3qO8uD7fhwUaY8xJbQoGTxg8A8SIyDVAtaqe6RzDVCBHVXerai3wHE5XVJNNA9Ge5zHAgTZX3tOkXwtok+4kgC9NSWXq4Dh+tTSbA8V2ItoY43ttHRJjHrAWuAGYB3wiItef4W0pQG6j13meZY39ErhZRPKAZcB321JPj5Q0GhJHndad5AoQ/nTDeBpUueOFz3DbHdHGGB9ra1fSz3DuYfiKqn4Z52jg3k74/PnA46o6AM9d1Z7zF02IyEIRyRKRrMLCwk74WB9JnwP7PoKyQ00Wp8aHc8/VY/gop4in1tiEPsYY32prMASoauOJi4va8N58YGCj1wM8yxq7FVgCoKqrgVAg4dQNqeoiVc1U1czExMQ2ltwNjfF0J21Zetqq+VMHMmNkIr97c6sNzW2M8am2BsNbIvK2iHxVRL4KvIHT9dOaT4HhIpImIsE4J5dP/Y24H7gEQERG4wRDDz4kOIOkUZA0psnNbseJCPfPHUdIoIsfPf8Z9Q024ogxxjfaevL5DmARMM7zWKSqd57hPfXAd4C3ga04Vx9li8h9IjLL0+xHwDdF5DPgWeCr2tuv20yfA/tXQ+np59n7Rody3+x0Nuwv5pFVNlyGMcY3pKf9Hs7MzNSsrCxfl9FxhTvgwSkw836Ydvtpq1WV7/x7A8u3HOLVb5/PmOToZjZijDHtIyLrVDWzLW1bPWIQkTIRKW3mUSYipZ1Trp9JHAF9M5rtTgKnS+nX12YQExbM/yzZSE19QxcXaIzxd60Gg6pGqWp0M48oVbU/ZTsq/VrIXQMlzd/UFhcRzP1zx7LtUBl/W7Gzi4szxvg7m/PZF8bMcX5uebXFJpeM7suNmQN5eOUu1u071kWFGWOMBYNvJAxzZndroTvpuHuuGU3/mDB+/PxnNqmPMabLWDD4SvocyFsLxbktNokKDeJPN4xnz5EK7n9zWxcWZ4zxZxYMvjLmWudnK91JANOHxvP189J4YvU+Pso50gWFGWP8nQWDr8QPhf7jz9idBPCTmSMZmhjBHc9/Rml1XRcUZ4zxZxYMvpQ+B/Kz4Fjr4yOFBrn4y7wJHC6r4VdLt3RRccYYf2XB4Ett7E4CGD8wlm/PGMqL6/N4O/vQGdsbY0xHWTD4UlwaJE9sU3cSwHe+MJz05Gh++tImjpTXeLk4Y4y/smDwtfQ5cGA9HNp0xqbBgQH8Zd4Eyqrruf2pdbydfYiqWrsz2hjTuSwYfG38AohIgv/cApVHz9h8ZL8ofnNtBjsLyrntqXVM/PVybnsqi5fW51FSaSemjTFnzwbR6w72fwKPXw1pF8CC58EVeMa31DW4WbvnKG9tPsTyLYc4XFpDYIAwfWg8l6f344oxfUmKDu2C4o0xPUF7BtGzYOgu1j0Br30Pzv0uXP6bdr3V7VY+yyvm7ezDvJ19iD1HKgCYmBrLzPR+XJHej8EJEd6o2hjTQ1gw9FRv/Ag+fQyuexTGzevQJlSVnQXlvL35EG9vOcTmfGcQ3JF9o/jZ1aO5cEQPngHPGNNhFgw9VUMdPDkb8tfB199yrlg6S7lHK1m+5TBPr9nHscpalv/wQpKirIvJGH/TafMxmC7mCoIbnoDwBHjuJigvOPN7zmBgXDi3np/GY1/JpKq2gZ+9vJme9seAMaZrWTB0N5GJ8KVnnCuUlnwZ6ms7ZbNDEyO544qRvLPlMK9sbH4eCGOMAS8Hg4jMFJHtIpIjIne10GaeiGwRkWwR+bc36+kxkifA7H84c0O/1erU2u3ytfPSyBzUh1+8ms3h0upO264xpnfxWjCIiAt4ELgSGAPMF5Exp7QZDtwNnKeq6cAPvFVPjzP2ejjv+5C12Hl0AleA8McbxlPb4ObulzZZl5IxplnePGKYCuSo6m5VrQWeA2af0uabwIOqegxAVc++U703ueQXMOxSWHYH7FvdKZtMS4jgJ1eM4r1tBbywLq9TtmmM6V28GQwpQONZaPI8yxobAYwQkY9EZI2IzGxuQyKyUESyRCSrsLDQS+V2QwEumPsYxA6CJbdASef8Iv/quYOZOjiO+17bwsGSqk7ZpjGm9/D1yedAYDgwA5gPPCoisac2UtVFqpqpqpmJiX52HX5YH5j/LNRVO1cq1Z39L/KAAOGPN4yj3q3c9aJ1KRljmvJmMOQDAxu9HuBZ1lgesFRV61R1D7ADJyhMY4kj4bpFcHAjvPZ96IRf5IPiI7j7qlGs3FHIkqyWpxc1xvgfbwbDp8BwEUkTkWDgS8DSU9q8gnO0gIgk4HQt7fZiTT3XqKvg4p/B5/+B1Q92yiZvPmcQ04fE8+vXt5JfbF1KxhiH14JBVeuB7wBvA1uBJaqaLSL3icgsT7O3gSIR2QK8D9yhqkXeqqnHu+DHMHoWvHMv7HrvrDcXECD84fpxqCp3vvC5dSkZYwAbEqPnqSmHf10OxfthwXMw+Pyz3uQzn+zjZy9v5rdzMrjpnEGdUKQxpruxITF6s5BIuPkFiE6Gp66DbcvOepMLpqZy/rAEfvvGVnKPVnZCkcaYnsyCoSeKToavvQl90+E/N8Nnz53V5kSE+68fR4AIP3nhc9zunnUUaYzpXBYMPVVEPHxlKQw+D16+DdY8dFabS4kN456rR7N6dxFPf7Kvk4o0xvREFgw9WUiUM+PbqGvgrbvg/f89q0tZb5wykAtHJPK7ZdvYX2RdSsb4KwuGni4o1Bmqe8LNsPJ+ePMn4HZ3aFMiwv1zxxLoEn78wmfWpWSMn7Jg6A1cgc5orNO/A2sXOV1LDXUd2lT/mDB+fs0Y1u45yhOr93ZqmcaYnuHMs86bnkHEmSs6PA7evQ+qS2DeExAU1u5NXT95AG9uPsT9b20jKSqUq8b2Q0S8ULQxpjuyI4beRAQu+BFc/RfYudy5nLW6pAObEX5/3ViGJETy7X+vZ/6ja9h6sNQLBRtjuiMLht5oyq1w/b8gby08fnWHpghNig7lte+ez2+uzWD7oTKufuBD7n1lM8cqOmdGOWNM92XB0FtlzIX5/4EjObB4pnOndDu5AoSbpw3i/R/P4MvTB/Pvtfu5+M8f8NTqvdQ3dOwEtzGm+7Ng6M2GXwpffgUqj8C/roA9qzp0xVJseDC/nJXOsu9dwJj+0dz7ajbX/P2/rNltw1oZ0xvZWEn+4NBmeHoulB+CmFQYdwOMu9EZzrudVJW3Nh/iN284I7JePa4/P71qNCmx7T/JbYzpOu0ZK8mCwV/UVsC2N5zhM3a/D+qG/hOcgMiYC1F927W56roGHlm5m4dW5gDw/y4axm0XDSE0yOWN6o0xZ8mCwbSu7DBsftGZ2+HgRpAAGHKxExKjr4HgiDZvKr+4iv9dtpU3Pj94YliNmRl2easx3Y0Fg2m7wu1OQHz+PJTsh6AIJxzGzYO0Gc7Nc22wZncRv1yazbZDZVw0IpH7ZqczKL7tAWOM8S4LBtN+bjfkrnG6mrJfgZoSiOwHsx6AEVe0aRP1DW6eWrOPPy/fQV2Dm29f7HQvhQRa95IxvmbBYM5OXbVzg9xdvQl/AAAU2klEQVTKP8DhzfCFe5wb59rYPXS4tJr7Xt/CG58fZEhCBL++NoPzhiV4uWhjTGu6zUQ9IjJTRLaLSI6I3NVKu7kioiLSpqKNlwWFwphZcOty58T0e7+G57/izB7XBn2jQ3lwwSSe/PpU3Krc9NgnfO/ZDRSUVXu5cGNMZ/BaMIiIC3gQuBIYA8wXkTHNtIsCvg984q1aTAcFh8Pcx5wxmLa+Bv+6DI7ubvPbLxyRyFs/uJDvXzKctzYf4pI/reSJj/fSYKO2GtOtefOIYSqQo6q7VbUWeA6Y3Uy7XwP3A/bnZHckAud+F256AUoPwKKLYdd7bX57aJCLH142grd/eCETUmP5xdJsrn3wIz7PK/Zi0caYs+HNYEgBchu9zvMsO0FEJgEDVfWN1jYkIgtFJEtEsgoLCzu/UnNmwy6Bhe8704o+PRc+eqBdkwKlJUTw5Nen8vf5EzlcWs3sBz/i3lc2U1LVseHBjTHe47MhMUQkAPgL8KMztVXVRaqaqaqZiYmJ3i/ONC9uCNz6jjNj3Dv3wkvfhNq2z/QmInxxfDIrfnQRX5k+mGc+2cclf17Js2v3U1tvYy8Z0114MxjygYGNXg/wLDsuCsgAPhCRvcA0YKmdgO7mQiJh3pPwhXth0wuw+Ip2D9AXHRrEL2els/Q75zMwLoy7X9rEjD++zxMf76W6rsFLhRtj2sprl6uKSCCwA7gEJxA+BRaoanYL7T8AfqyqrV6LaperdiM73oYXvwGuIGd60bQL2r0JVWXljkL+8V4OWfuOkRAZwjcvSOOmaYOIDLF5pIzpLN3iclVVrQe+A7wNbAWWqGq2iNwnIrO89bmmC424Ar75HoTHw5Oz4ZNH2nXeAZzupRkjk3j+9uk8t3Aao/pF8bs3t3H+/e/xtxU7Kam0cxDGdDW7wc2cvepSZ57p7ctgzGzIvBUGnw8BHbvjecP+Yzz4fg4rthYQGRLIl6cP4tbz04iPDOnkwo3xH3bns+l6bjes+iN89Deoq4CIJCckMubCwHMgoP0Hp1sOlPLgBzks23SQkMAAFkwdxMILh9AvJtQLO2BM72bBYHynthJ2vg2bX3KG1aivhqhkSJ8DGddByuQ2D61xXE5BOQ99sItXNubjEmHu5AF85dxBjOoX7aWdMKb3sWAw3UNNGWx/C7JfgpwV0FALsalOSKRfB/3Htyskco9W8vDKXTyflUdtg5uxKTHMyxzArPEpxIQHeXFHjOn5LBhM91NV7EwUlP0S7P4A3PXOfRHp18Goq5xJg9p4TuJoRS2vbsxnSVYeWw+WEhwYwBXp/ZiXOYBzhybgCrC5IIw5lQWD6d4qj8LWpU53094PndnkQmJg8HmQdqHzSBzdpvMSm/NLeD4rl1c2HqCkqo7kmFCunzyA6ycPJDU+vAt2xpiewYLB9BwVR2DPStizynkcH6QvPMG5LyLtQki7yDm6aKXbqbqugRVbD/N8Vh6rdhaiCtOGxDEvcyBXZvQnLNjmhDD+zYLB9FzFuc5RxJ5VsHsllB1wlkennDyaGHw+xAxsMSgOllTx0vp8lmTlsq+oksiQQK4Z15/ZE1I4Jy2OAOtqMn7IgsH0DqrOEUTjI4rKImddSAwkDIfEkc7PhBHOo89g505snLuqP917jCVZuby56SAVtQ30iw49ERIZKdE2N7XxGxYMpndyu6FwK+z7GAq3wZEdcGQnlB082SYg0Ol2ShjRJDCqYobx7p5KXt14gA+2F1DXoAxJiGDWhGRmjU9mSGKk7/bLmC5gwWD8S3WpExBHdjR67ISju5yrnwACgmDoxZA+h5LUy3kzxwmJNXuKUIVxA2KYNT6Za8Yl2w10pleyYDAGoKEOju1zgmL/x5D9KpTsd0Ji2CWQPofD/b/Aa9vLeXXjATbllyAC09LimT0hmUvH9CXBhuEwvYQFgzHNUYX89c69FNmvQGkeuIJhqBMSuxMu5NUtZSz97AB7jlQAMCQhgimD48gc3Icpg+MYFB9u5yVMj2TBYMyZqEJeFmx5BbJfhtJ8cIXAsEvQMdeyJeo8/ptbw6d7j/Lp3mMnZppLjAphyuA+ZA6KY2paHKP6RRHo8tl8V8a0mQWDMe3hdkN+lhMQ2a84l8i6QiB1GqRMwt1/IntDR7G6MISsfcWs3XOU/OIqACKCXUwa5BxNTB7Uh4F9wkmMCrH7Jky3Y8FgTEe53ZC31gmI/R/D4eyTJ7AjkiBlEiRPoigmnU/r0vjogPLp3qNsP1zWZCqKiGAXiVEhJESGnPYzITL4xPOk6BBCAi1EjPdZMBjTWeqq4fBm59zEgfXOzyM7AM+/m9hUSJ5EVeJ4dgQOZ49rEPk14Rwpr+FIeS2FZdUcKa/lSHkNxc1MOhQaFMDVY5NZcM5AJqX2sfMXxmssGIzxpupSOPjZyaA4sL7pvNfh8ZAw0rn5LnGkcy9F4ihqw/tRVFnLkTInKArLatiQW8zSjflU1DYwPCmS+VNTuW5SCrHhwb7bP9MrdZtgEJGZwN8AF/CYqv7+lPX/A3wDqAcKga+r6r7WtmnBYLqliiNwYKPnxrvtULjDeV5dfLJNcNTJu7UTRzrh0TedivAUXv/8AP9em8tnucUEBwZwVUY/vjQ1lXPS4uwownSKbhEMIuICdgCXAXnAp8B8Vd3SqM3FwCeqWiki/w+Yoao3trZdCwbTY6hCRSEUbm8aFkd2NL1bO24oDL8Mhl3K1pBxPLuhkJc35FNWXc+QxAjmT3GOImxqU3M2ukswTAd+qapXeF7fDaCqv2uh/UTgH6p6XmvbtWAwvUJ1iXN3dv46ZxKjPR9CfRUEhsLg86kd/AXeaxjPo9nCuv3FBLmEK9L7MX9qKtOHxNtAgKbd2hMMgV6sIwXIbfQ6Dzinlfa3Am82t0JEFgILAVJTUzurPmN8JzQGBmQ6j3Nug7oq2PcR7FwBOSsIzvkpM4GZfQZTnHkRy6oyeGBHPa9/fpCwIBcpfcIYcOIRfuJnSmwYCZHB1v1kzoo3g6HNRORmIBO4qLn1qroIWATOEUMXlmZM1wgKg2GXOg+Ao3ucI4mcd4nd/jwL6p5gviuYwoGZ7KMfNdXV1B6sonZfDdJQSxD1lFPHbmnggNQR7nITHtBAaEADwdJAQFAogaERBIVFERAcAccfQeEnnzd+HdbHOR8SO7hNEyaZ3sWbwZAPDGz0eoBnWRMicinwM+AiVa3xYj3G9BxxaTD1m86jvgb2fYzkrCAp512Syrc5Q3mEB0NUMA0BwdTiosYdQpXbRWVDAOUNLg7XCSW1AVTWCyFSR1hJDRFSRp/AIqJddURKDaFUE9RQSUBDC//0AsOcgEgaDYmjnEfSqO4dGDVlTtdczgrY9a5zrmfMLGeu8eRJ7Zpn3F958xxDIM7J50twAuFTYIGqZjdqMxF4AZipqjvbsl07x2BM+5TX1LOnsIKcwjJ2FVSwq7CcnIJy9hZVUNfg/PsPwM3ASBgdL4yIczEqqpaMkEOk1O7DdWSbcwK9NO/kRpsLjLghEJnkHG105S9fVTi0yRME78H+NeCuQ4MiyOszBVE3KUWrEXedc99J+hxnrvH+4/0qJLrFyWdPIVcBf8W5XHWxqv5WRO4DslR1qYisAMYCxy/R2K+qs1rbpgWDMZ2jvsFN7rEqcgrK2VVYzq6CcnI8oVFW7dztHRwYQEZyNBNT+zClfyCTwgtIrNyFHNkBBVudq6xKT+kIcAVDZF8nJJr8bPzwLAvq4BDnFUWw+/2TYVB+2FnedyxVg2bwemU6f8iOptAZuYSMeOUP6bmMOfqu8z53PfRJc0Ii4zrom9HrQ6LbBIM3WDAY412qysGSajbmFrNh/zE27C9mU34JNfVuwBlIcOLAWCam9mFiaizjEoTwkhznJr/yAueXdHkBlB86+briCCfuFm/MFew5rxEJweFNnwdHQFBE0+d1lbD7AziwwdleWJwzz8awS8mLm86iDRUsycqlus7NpaOTuO2ioVTU1HPf61vYXVjBjJGJ/OLS/qQVvu+MjbV7JWgDxA/zHEnMgaQxnR8Sqs5/i4ItTqAWbIHacojqD9HJJ38efx7Y+ZcmWzAYYzpVXYObbQfL2JDrBMWG/cfYW1QJQIDAyH7RDOwThggI4vz0PEfApQ1EuUuIqS8ipuEo0fVHiW44Sv/QOpLDlYSQegLrK51f/LUVJx+NX7vrQAJgwBRnqPRhl0LyBDYdKOeRVbtYtukgrgBhzsQUvnnBEIb3jTpRf229mydX7+VvK3ZSVdfAV88dzPcuHU50Qwlsfc0Jib0fgrqdO9X7jTt5ZBPVr+lRT1hc6+dXqopP/vIv2HryedXRk23C4yE01rmfpa7y9G2Ex3tCIhmi+ztznkf1d8bq6pveoe/QgsEY43VHK2rZmHuMjfuL2ZBbTGFZDaqgqOenc/ShAKe8VnXC5mBJNQCBAUJ6cjSTBvVhsufRPyas6QfW1zq/uINCUVU+3HmER1bt4qOcIqJCAlkwLZWvnZvW6gx8hWU1/Hn5dv6TlUt8RDB3XDGSGyYPdO4LKS+Era/C1tfh2F7nSKe5X9riOqV7LMk5oinKcUKgcddacJRzHiZptHMkcvxnZKKzXtW5p6XsIJQecB5Nnh+A0oNQecRpf/7/wKW/6ND3ZcFgjOkRjlbUsmH/Mdbtcx6f5RVTXed0WaXEhjlBkRrL5EFxjOrvHAEs23SQh1fuZuvBUvpGh/D189KYf04q0aFBbf7cTXkl/PK1bNbtO8bYlBh+OWsMkwfFnd6wptzTNXa4UReZ53lZo2U1ZRA/BJLSm4ZAzIDO6Zaqr3ECIzDUOYLpAAsGY0yPVNfgZuvB0hNBsX7fMQ54jirCglxEhARypLyGYUmRLLxwCLMnJHd42HJVZelnB/jdsm0cKq3m2gnJ3HXl6F4757cFgzGm1zhQXMV6z1HFoZJq5k4awBdGJXXasCAVNfU89MEuFn24G5cI3754KDdOSSUxqneNTWXBYIwx7ZR7tJLfvrGVt7IPAdA3OoSM5BjSU2JIT44mIyWG5JjQHjvcSHcZK8kYY3qMgXHhPHzLZD7Pc6Zv3XKglM0HSnh/ewFuz9/PfcKDyEiJYUxyNBnJMWSkxDAoLrzXDWpowWCMMY2MGxDLuAGxJ15X1Taw9VAp2fklZHvCYvF/95y4azwyJJDhfSOJDAkkNMjlPAIDCAs++Tzk+PKgAEIDnedhwZ7nwS5CA12e9gGEedqGBAb47OjEgsEYY1oRFuxiUmofJqX2ObGstt7NjsNlJ44qdh4up7S6nsKyGqrrGqiuc1Nd33DieUeI4AmRk2Gx4JxUvnHBkM7atRZZMBhjTDsFBwaQkeJ0Jc1rMlbo6VSVmnr3ycCoa/CEhvO8qq6BGs/P6jo3VbXNLKtzQiahiyZrsmAwxhgvEpETXUw9RTcdN9cYY4yvWDAYY4xpwoLBGGNMExYMxhhjmrBgMMYY04QFgzHGmCYsGIwxxjRhwWCMMaaJHje6qogUAvs6+PYE4EgnltPT+PP++/O+g3/vv+27Y5CqJrblTT0uGM6GiGS1ddjZ3sif99+f9x38e/9t39u/79aVZIwxpgkLBmOMMU34WzAs8nUBPubP++/P+w7+vf+27+3kV+cYjDHGnJm/HTEYY4w5AwsGY4wxTfhNMIjITBHZLiI5InKXr+vpSiKyV0Q2ichGEcnydT3eJiKLRaRARDY3WhYnIu+IyE7Pzz6tbaOnamHffyki+Z7vf6OIXOXLGr1FRAaKyPsiskVEskXk+57l/vLdt7T/7f7+/eIcg4i4gB3AZUAe8CkwX1W3+LSwLiIie4FMVfWLm3xE5EKgHHhSVTM8y/4AHFXV33v+MOijqnf6sk5vaGHffwmUq+qffFmbt4lIf6C/qq4XkShgHXAt8FX847tvaf/n0c7v31+OGKYCOaq6W1VrgeeA2T6uyXiJqq4Cjp6yeDbwhOf5Ezj/YHqdFvbdL6jqQVVd73leBmwFUvCf776l/W83fwmGFCC30es8OvgfrIdSYLmIrBORhb4uxkf6qupBz/NDQF9fFuMD3xGRzz1dTb2yK6UxERkMTAQ+wQ+/+1P2H9r5/ftLMPi781V1EnAl8G1Pd4PfUqf/tPf3oZ70EDAUmAAcBP7s23K8S0QigReBH6hqaeN1/vDdN7P/7f7+/SUY8oGBjV4P8CzzC6qa7/lZALyM07Xmbw57+mCP98UW+LieLqOqh1W1QVXdwKP04u9fRIJwfik+o6oveRb7zXff3P535Pv3l2D4FBguImkiEgx8CVjq45q6hIhEeE5EISIRwOXA5tbf1SstBb7ief4V4FUf1tKljv9S9JhDL/3+RUSAfwFbVfUvjVb5xXff0v535Pv3i6uSADyXaP0VcAGLVfW3Pi6pS4jIEJyjBIBA4N+9fd9F5FlgBs6Qw4eBXwCvAEuAVJxh2+epaq87SdvCvs/A6UZQYC9wW6M+915DRM4HPgQ2AW7P4p/i9LP7w3ff0v7Pp53fv98EgzHGmLbxl64kY4wxbWTBYIwxpgkLBmOMMU1YMBhjjGnCgsEYY0wTFgzGdCERmSEir/u6DmNaY8FgjDGmCQsGY5ohIjeLyFrP+PWPiIhLRMpF5P88Y92/KyKJnrYTRGSNZ5Cyl48PUiYiw0RkhYh8JiLrRWSoZ/ORIvKCiGwTkWc8d6wa021YMBhzChEZDdwInKeqE4AG4CYgAshS1XRgJc5dxQBPAneq6jicu06PL38GeFBVxwPn4gxgBs6olz8AxgBDgPO8vlPGtEOgrwswphu6BJgMfOr5Yz4MZ+A1N/AfT5ungZdEJAaIVdWVnuVPAM97xqdKUdWXAVS1GsCzvbWqmud5vREYDPzX+7tlTNtYMBhzOgGeUNW7mywUufeUdh0dT6am0fMG7N+h6WasK8mY070LXC8iSXBizuBBOP9erve0WQD8V1VLgGMicoFn+S3ASs8MWnkicq1nGyEiEt6le2FMB9lfKsacQlW3iMg9OLPeBQB1wLeBCmCqZ10BznkIcIZyftjzi3838DXP8luAR0TkPs82bujC3TCmw2x0VWPaSETKVTXS13UY423WlWSMMaYJO2IwxhjThB0xGGOMacKCwRhjTBMWDMYYY5qwYDDGGNOEBYMxxpgm/j+MRgv0AKZDFAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4fa65d2ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#summarize for loss\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved model weights\n"
     ]
    }
   ],
   "source": [
    "model_json = model.to_json()\n",
    "with open('model2API_model.json', 'w') as json_file:\n",
    "    json_file.write(model_json)\n",
    "    \n",
    "model.save_weights('model2API_weights.h5')\n",
    "print('Saved model weights')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "new_model = load_model('model2-25-0.0074.h5')\n",
    "# check in models are similar to the given tolerance\n",
    "# assert_allclose(model.predict(X_train),\n",
    "#                new_model.predict(X_train),\n",
    "#                1e-4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "filepath = 'model2API_continue-{epoch:02d}-{loss:.4f}.h5'\n",
    "checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, save_best_only=True, mode='min')\n",
    "callbacks_list = [checkpoint]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "new_history = new_model.fit(X_train, y_train, epochs=10, batch_size=2048, callbacks=callbacks_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "new_score = new_model.evaluate(X_test, y_test, verbose=1)\n",
    "print('training results:')\n",
    "for i in range(len(new_model.metrics_names)):\n",
    "    print(str(new_model.metrics_names[i]) + \": \" + str(new_score[i]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model_json = model.to_json()\n",
    "with open('model2API_continue.json', 'w') as json_file:\n",
    "    json_file.write(model_json)\n",
    "    \n",
    "model.save_weights('model2API_weights_continue.h5')\n",
    "print('Saved model weights')"
   ]
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}