example_modelwrapper.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MODELWRAPPER EXAMPLE\n",
    "The ModelWrapper just provides some convenience methods to use our Keras model easily. This example is only meant to provide a better understanding on how the ModelWrapper works."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import the ModelWrapper, the DataGenerator and auxiliary libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from modelwrapper import ModelWrapper\n",
    "from datagenerator import DataGenerator\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create train, test and validation data generators\n",
    "Note that the trainGenerator is in randomized mode to reduce overfitting, whilst validation and test generators are not, since randominzation has no influence in this case and non-randomized access to data is faster.\n",
    "Also note that each generator uses only one informative image since each transectList has a single item. A proper usage of the ModelWrapper requires more training, validation and test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainGenerator=DataGenerator('DATASET/DATA','DATASET/GT',transectList=[6],doRandomize=True)\n",
    "testGenerator=DataGenerator('DATASET/DATA','DATASET/GT',transectList=[7],doRandomize=False)\n",
    "valGenerator=DataGenerator('DATASET/DATA','DATASET/GT',transectList=[8],doRandomize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create the model\n",
    "Note that calling to create() is necessary to actually create the model. Also note that the created model is the one described in the reference paper."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n"
     ]
    }
   ],
   "source": [
    "theModel=ModelWrapper()\n",
    "theModel.create()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train the model\n",
    "The first two parameters are the training and the validation data generators. The third one is the number of epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/math_grad.py:102: div (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Deprecated in favor of operator or tf.math.divide.\n",
      "Epoch 1/100\n",
      "9/9 [==============================] - 2s 170ms/step - loss: 1.0923 - acc: 0.5380 - val_loss: 1.0791 - val_acc: 0.7208\n",
      "Epoch 2/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 1.0703 - acc: 0.6494 - val_loss: 1.0319 - val_acc: 0.7391\n",
      "Epoch 3/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 1.0074 - acc: 0.7152 - val_loss: 0.9500 - val_acc: 0.6808\n",
      "Epoch 4/100\n",
      "9/9 [==============================] - 1s 130ms/step - loss: 0.9311 - acc: 0.6902 - val_loss: 0.8387 - val_acc: 0.7650\n",
      "Epoch 5/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.8789 - acc: 0.6855 - val_loss: 0.7438 - val_acc: 0.7487\n",
      "Epoch 6/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.7898 - acc: 0.7543 - val_loss: 0.6522 - val_acc: 0.7943\n",
      "Epoch 7/100\n",
      "9/9 [==============================] - 1s 125ms/step - loss: 0.6609 - acc: 0.7732 - val_loss: 0.5692 - val_acc: 0.8045\n",
      "Epoch 8/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.5083 - acc: 0.8231 - val_loss: 0.4526 - val_acc: 0.8325\n",
      "Epoch 9/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.5243 - acc: 0.8134 - val_loss: 0.5810 - val_acc: 0.7841\n",
      "Epoch 10/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.4936 - acc: 0.8223 - val_loss: 0.5025 - val_acc: 0.8147\n",
      "Epoch 11/100\n",
      "9/9 [==============================] - 1s 124ms/step - loss: 0.4685 - acc: 0.8236 - val_loss: 0.4366 - val_acc: 0.8354\n",
      "Epoch 12/100\n",
      "9/9 [==============================] - 1s 124ms/step - loss: 0.4661 - acc: 0.8244 - val_loss: 0.4282 - val_acc: 0.8370\n",
      "Epoch 13/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.4724 - acc: 0.8153 - val_loss: 0.4389 - val_acc: 0.8332\n",
      "Epoch 14/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.4392 - acc: 0.8286 - val_loss: 0.4563 - val_acc: 0.8274\n",
      "Epoch 15/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.4270 - acc: 0.8349 - val_loss: 0.4126 - val_acc: 0.8388\n",
      "Epoch 16/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.4499 - acc: 0.8230 - val_loss: 0.4061 - val_acc: 0.8408\n",
      "Epoch 17/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.4230 - acc: 0.8287 - val_loss: 0.4155 - val_acc: 0.8393\n",
      "Epoch 18/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3789 - acc: 0.8429 - val_loss: 0.4983 - val_acc: 0.8034\n",
      "Epoch 19/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3817 - acc: 0.8395 - val_loss: 0.4481 - val_acc: 0.8298\n",
      "Epoch 20/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3867 - acc: 0.8500 - val_loss: 0.4172 - val_acc: 0.8526\n",
      "Epoch 21/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.3653 - acc: 0.8647 - val_loss: 0.3837 - val_acc: 0.8570\n",
      "Epoch 22/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.3511 - acc: 0.8749 - val_loss: 0.3769 - val_acc: 0.8631\n",
      "Epoch 23/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3483 - acc: 0.8666 - val_loss: 0.4139 - val_acc: 0.8506\n",
      "Epoch 24/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.3573 - acc: 0.8663 - val_loss: 0.3859 - val_acc: 0.8540\n",
      "Epoch 25/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.3411 - acc: 0.8663 - val_loss: 0.4435 - val_acc: 0.8294\n",
      "Epoch 26/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.3409 - acc: 0.8695 - val_loss: 0.3613 - val_acc: 0.8697\n",
      "Epoch 27/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.3130 - acc: 0.8808 - val_loss: 0.3600 - val_acc: 0.8705\n",
      "Epoch 28/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3166 - acc: 0.8819 - val_loss: 0.3562 - val_acc: 0.8727\n",
      "Epoch 29/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.2857 - acc: 0.8911 - val_loss: 0.3578 - val_acc: 0.8735\n",
      "Epoch 30/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2707 - acc: 0.8970 - val_loss: 0.3472 - val_acc: 0.8784\n",
      "Epoch 31/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2726 - acc: 0.8986 - val_loss: 0.3780 - val_acc: 0.8681\n",
      "Epoch 32/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.2528 - acc: 0.9042 - val_loss: 0.3662 - val_acc: 0.8721\n",
      "Epoch 33/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2573 - acc: 0.9078 - val_loss: 0.3363 - val_acc: 0.8826\n",
      "Epoch 34/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.3188 - acc: 0.8840 - val_loss: 0.4876 - val_acc: 0.8151\n",
      "Epoch 35/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2916 - acc: 0.8864 - val_loss: 0.3246 - val_acc: 0.8823\n",
      "Epoch 36/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2768 - acc: 0.9005 - val_loss: 0.4103 - val_acc: 0.8504\n",
      "Epoch 37/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.2798 - acc: 0.8904 - val_loss: 0.3361 - val_acc: 0.8826\n",
      "Epoch 38/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2640 - acc: 0.9032 - val_loss: 0.3312 - val_acc: 0.8859\n",
      "Epoch 39/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2581 - acc: 0.9020 - val_loss: 0.4782 - val_acc: 0.8177\n",
      "Epoch 40/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2551 - acc: 0.9079 - val_loss: 0.3816 - val_acc: 0.8634\n",
      "Epoch 41/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2353 - acc: 0.9128 - val_loss: 0.3321 - val_acc: 0.8796\n",
      "Epoch 42/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2158 - acc: 0.9198 - val_loss: 0.2979 - val_acc: 0.8959\n",
      "Epoch 43/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2175 - acc: 0.9206 - val_loss: 0.3430 - val_acc: 0.8717\n",
      "Epoch 44/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2432 - acc: 0.9110 - val_loss: 0.3446 - val_acc: 0.8741\n",
      "Epoch 45/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2158 - acc: 0.9215 - val_loss: 0.3917 - val_acc: 0.8519\n",
      "Epoch 46/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2090 - acc: 0.9222 - val_loss: 0.3474 - val_acc: 0.8726\n",
      "Epoch 47/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.2037 - acc: 0.9253 - val_loss: 0.4807 - val_acc: 0.8089\n",
      "Epoch 48/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2268 - acc: 0.9172 - val_loss: 0.4492 - val_acc: 0.8212\n",
      "Epoch 49/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2051 - acc: 0.9233 - val_loss: 0.2622 - val_acc: 0.9070\n",
      "Epoch 50/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1948 - acc: 0.9276 - val_loss: 0.3051 - val_acc: 0.8885\n",
      "Epoch 51/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.2060 - acc: 0.9229 - val_loss: 0.3306 - val_acc: 0.8764\n",
      "Epoch 52/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1880 - acc: 0.9281 - val_loss: 0.2929 - val_acc: 0.8944\n",
      "Epoch 53/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1968 - acc: 0.9282 - val_loss: 0.2779 - val_acc: 0.9004\n",
      "Epoch 54/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.2063 - acc: 0.9232 - val_loss: 0.4613 - val_acc: 0.8125\n",
      "Epoch 55/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.2091 - acc: 0.9229 - val_loss: 0.3565 - val_acc: 0.8760\n",
      "Epoch 56/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2062 - acc: 0.9248 - val_loss: 0.3388 - val_acc: 0.8705\n",
      "Epoch 57/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1915 - acc: 0.9277 - val_loss: 0.2676 - val_acc: 0.9038\n",
      "Epoch 58/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1811 - acc: 0.9326 - val_loss: 0.2849 - val_acc: 0.8966\n",
      "Epoch 59/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1877 - acc: 0.9289 - val_loss: 0.2693 - val_acc: 0.9044\n",
      "Epoch 60/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1952 - acc: 0.9267 - val_loss: 0.2575 - val_acc: 0.9093\n",
      "Epoch 61/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.2027 - acc: 0.9256 - val_loss: 0.2895 - val_acc: 0.8916\n",
      "Epoch 62/100\n",
      "9/9 [==============================] - 1s 119ms/step - loss: 0.1850 - acc: 0.9289 - val_loss: 0.2830 - val_acc: 0.8969\n",
      "Epoch 63/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1942 - acc: 0.9269 - val_loss: 0.2955 - val_acc: 0.8876\n",
      "Epoch 64/100\n",
      "9/9 [==============================] - 1s 119ms/step - loss: 0.1880 - acc: 0.9273 - val_loss: 0.2694 - val_acc: 0.9008\n",
      "Epoch 65/100\n",
      "9/9 [==============================] - 1s 119ms/step - loss: 0.2027 - acc: 0.9220 - val_loss: 0.2559 - val_acc: 0.9080\n",
      "Epoch 66/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1943 - acc: 0.9295 - val_loss: 0.2596 - val_acc: 0.9079\n",
      "Epoch 67/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1786 - acc: 0.9343 - val_loss: 0.3636 - val_acc: 0.8657\n",
      "Epoch 68/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.2073 - acc: 0.9209 - val_loss: 0.2986 - val_acc: 0.8861\n",
      "Epoch 69/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1894 - acc: 0.9271 - val_loss: 0.2585 - val_acc: 0.9059\n",
      "Epoch 70/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1876 - acc: 0.9288 - val_loss: 0.2564 - val_acc: 0.9085\n",
      "Epoch 71/100\n",
      "9/9 [==============================] - 1s 119ms/step - loss: 0.1720 - acc: 0.9350 - val_loss: 0.2864 - val_acc: 0.8934\n",
      "Epoch 72/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1641 - acc: 0.9371 - val_loss: 0.3352 - val_acc: 0.8739\n",
      "Epoch 73/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1730 - acc: 0.9327 - val_loss: 0.2752 - val_acc: 0.9011\n",
      "Epoch 74/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1714 - acc: 0.9338 - val_loss: 0.2789 - val_acc: 0.8978\n",
      "Epoch 75/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1624 - acc: 0.9364 - val_loss: 0.2522 - val_acc: 0.9105\n",
      "Epoch 76/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1768 - acc: 0.9308 - val_loss: 0.2882 - val_acc: 0.8926\n",
      "Epoch 77/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1780 - acc: 0.9324 - val_loss: 0.2840 - val_acc: 0.8953\n",
      "Epoch 78/100\n",
      "9/9 [==============================] - 1s 119ms/step - loss: 0.1695 - acc: 0.9351 - val_loss: 0.2865 - val_acc: 0.8930\n",
      "Epoch 79/100\n",
      "9/9 [==============================] - 1s 124ms/step - loss: 0.1605 - acc: 0.9374 - val_loss: 0.2780 - val_acc: 0.8962\n",
      "Epoch 80/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.1517 - acc: 0.9419 - val_loss: 0.2563 - val_acc: 0.9091\n",
      "Epoch 81/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.1561 - acc: 0.9401 - val_loss: 0.2716 - val_acc: 0.9029\n",
      "Epoch 82/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1602 - acc: 0.9375 - val_loss: 0.2594 - val_acc: 0.9102\n",
      "Epoch 83/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1651 - acc: 0.9372 - val_loss: 0.3612 - val_acc: 0.8634\n",
      "Epoch 84/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1711 - acc: 0.9349 - val_loss: 0.3156 - val_acc: 0.8788\n",
      "Epoch 85/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1638 - acc: 0.9364 - val_loss: 0.2460 - val_acc: 0.9137\n",
      "Epoch 86/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1570 - acc: 0.9387 - val_loss: 0.2521 - val_acc: 0.9121\n",
      "Epoch 87/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.1674 - acc: 0.9355 - val_loss: 0.2514 - val_acc: 0.9114\n",
      "Epoch 88/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1484 - acc: 0.9439 - val_loss: 0.2901 - val_acc: 0.8923\n",
      "Epoch 89/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1646 - acc: 0.9348 - val_loss: 0.3151 - val_acc: 0.8802\n",
      "Epoch 90/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1464 - acc: 0.9439 - val_loss: 0.2835 - val_acc: 0.8956\n",
      "Epoch 91/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1434 - acc: 0.9452 - val_loss: 0.2579 - val_acc: 0.9094\n",
      "Epoch 92/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1463 - acc: 0.9433 - val_loss: 0.2645 - val_acc: 0.9067\n",
      "Epoch 93/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.1443 - acc: 0.9438 - val_loss: 0.3173 - val_acc: 0.8820\n",
      "Epoch 94/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1476 - acc: 0.9435 - val_loss: 0.2601 - val_acc: 0.9097\n",
      "Epoch 95/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1410 - acc: 0.9460 - val_loss: 0.2599 - val_acc: 0.9102\n",
      "Epoch 96/100\n",
      "9/9 [==============================] - 1s 122ms/step - loss: 0.1532 - acc: 0.9404 - val_loss: 0.2820 - val_acc: 0.8991\n",
      "Epoch 97/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1463 - acc: 0.9438 - val_loss: 0.2524 - val_acc: 0.9137\n",
      "Epoch 98/100\n",
      "9/9 [==============================] - 1s 121ms/step - loss: 0.1512 - acc: 0.9434 - val_loss: 0.2751 - val_acc: 0.9066\n",
      "Epoch 99/100\n",
      "9/9 [==============================] - 1s 123ms/step - loss: 0.1506 - acc: 0.9434 - val_loss: 0.3134 - val_acc: 0.8809\n",
      "Epoch 100/100\n",
      "9/9 [==============================] - 1s 120ms/step - loss: 0.1552 - acc: 0.9402 - val_loss: 0.3119 - val_acc: 0.8813\n"
     ]
    }
   ],
   "source": [
    "theModel.train(trainGenerator,valGenerator,100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model to disk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "theModel.save('MODELS/TEST_MODEL')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot the training history\n",
    "These plots are useful to detect overfitting and decide the number of epochs.\n",
    "It can be observed that accuracies reach quite large values, even with such small training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "theModel.plot_training_history()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load the model and plot the history again\n",
    "Just to show that save/load preserves the history (and, of course, the model itself)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsnXd4HMXZwH+j3ntxt+TeewMXbIqx6aaZYjoYCJAASQgQPkoCCRAChARM7xgDpoPBYHDDvXe5ybYsy+pW77r5/pg76XQ6SSdZJ8nS+3uee+52d3Z3bqWbd946SmuNIAiCIAB4tHYHBEEQhLaDCAVBEAShChEKgiAIQhUiFARBEIQqRCgIgiAIVYhQEARBEKoQoSAIgiBUIUJBaBaUUgV2L4tSqthu+1ql1ONKqXLrdo5SarVS6jQn14m3nj/PyTGtlOpj/fy4dftKu+Ne1n1x1u13lVJPWj/HWY8tcrjmh0qpx+22g5VSzyulDiulCpVSSUqphUqp8XV8b9t1Cxxes5VSDyqlVjg5J0opVaaUGmLd7qaU+kgplWW953ql1AV1fXeH/Y8rpT6s61kppXbZ9alSKVVit/2wUupGpdRvDufeqJTaoZQqUkqlKqXmKaXCHO5Z77MXTl1EKAjNgtY6yPYCkoAL7fZ9ZG32ifV4FLAU+MzJpa4HTgCzlVK+Ddw2G3hCKeXZiK6OV0qd7uyA9X6/AkOBC4AQYCCwAJjZwHXD7J+B1voT4EPgdKVUvEPbq4AdWuudSqkI4DegDBiMeTYvAPOVUpc34ns5RWs92O7vshK4266P/3Bsr5T6I/AM8GcgFJgA9AR+Vkr52DVtyrMXTgFEKAgtjta6AvgI6KqUirbtV0opjFB4BCgHLmzgUj9iBtM5jbj9s8BTdRy7DugGXKK13qm1rtRaF2qtF2qtH2/EPQDQWidjhMx1DoeuB963fr4PKABu0Vqnaq2LtdYfW/v4b+szaRGUUiHAE8A9WusftdblWuvDwJVAHDWfc1OevXAKIEJBaHGsM87rgSyMVmBjEmZQXgB8CtzQwKU08H/AY0opbxdv/wrQTyl1tpNjZwOLtdaFLl7LFd7DTigopfoDI4D51l3nAJ9rrS0O530K9AD6NWNfGuJ0wA/4wn6n1roAWITpa9VuGv/shVMAEQpCS3KlUioHKAZuAy63ag02bgB+0FqfwAyaM5RSMfVdUGv9DZAB3OpiH4oxs/AnnRyLAlJtG0qpEVb/R55Sam8D1820trW9Blr3fwnE2pmsrsd8xwy7ex53cr3jdsdbiigg0+FvYt+fGn1pwrMXTgFEKAgtyada6zAgFtgJjLYdUEr5A1dgzEporddgfBPXuHDdR4C/Yma5rvAmZqB2NE9lAZ1tG1rrrdb+Xgo05N+I0lqH2b32WK9RhPGdXG81BV1LtekIINP+nnZ0tjteHxVAjZm63cy9vIFzHckEopRSXnX0x1lfGvvshTaOCAWhxdFaZwJzgceVUrbBbxbGsfuKNeIlFehKwyYktNY/AweA37l4/zKM7fzvgL3N/hdgulIq0NXv4iLvYezy5wDBwLd2x5YAlyqlHH+LVwJHgX0NXDsJY++3Jx4jLI41sp9rgFKMEKxCKRWEcbT/4nhCY5+90PYRoSC0ClrrvcBi4AHrrhuAtzGRPyOsr4nAcKXUUBcu+Ve7a7nCB5jZ7Qy7fe9jzCRfKqWGKKU8lVJ+wJhGXNcZK4Ec4HVggVUo2XgBE+XzllKqk1LKTyl1Neb7/FnXrG3vYz1ue3liHL4DlFLXKaW8rdFM/8D4KZyZgepEa52LEZb/VUrNsF4vDuPfSMY8M2c09tkLbRgRCkJr8i9grlKqJ3AW8KI1Asf22oQZ9FzRFlYB6129sda6EngUiLDbVwJMA3YD3wN5wF5gLGbmXh85DnkK99tdV2METk9qmo7QWmdhHOx+1vtmAfcD11nDWu3ZhfGJ2F43aa3TMbP424F0jFkuB7jTtSdRE631s8DDwHOY778Oo7GcpbUureOcRj17oW2jZJEdQRAEwYZoCoIgCEIVIhQEQRCEKkQoCIIgCFWIUBAEQRCqcJak0qaJiorScXFxrd0NQRCEU4pNmzZlaq2jG2p3ygmFuLg4Nm7c2NrdEARBOKVQSh1xpZ2YjwRBEIQqRCgIgiAIVYhQEARBEKo45XwKzigvLyc5OZmSkpLW7kq7wM/Pj27duuHtLWXyBaGj0S6EQnJyMsHBwcTFxdGCC1W1S7TWZGVlkZycTHy84yqSgiC0d9qF+aikpITIyEgRCM2AUorIyEjRugShg9IuhAIgAqEZkWcpCB2XdmE+EgRBaKtUVFo4mFHIzmO5ZBaUMrx7GCO6h+Hn7dnaXXOKCIVmICcnh/nz5/O73zVu8anzzjuP+fPnExYW5qaeCYLQnBSVVbByfybL9qazbG8GxeWV9IsJpm9sED0iAgjx9ybEz5vCsgp2Hstlx7Fc9hzPo6TcUuM63p6KAZ1CiAj0IdjPi2A/L5RSKEApUNYFAT09FN0jAugXG0S/2GBign3drsmLUGgGcnJyeOWVV2oJhYqKCry86n7EixYtcnfXBKHdkl9Szo5juRzPKSEiyIfoIF+6hfsTFuBzUtdNySlmzlvrOK1XJI9fNBhvT2Nl35eWzw1vr+d4bglBvl5M7htFWIAP+9Py+XZbCnklNRe6C/TxZHDXUK4Z15Oh3UIY0iWUiEAftiTlsOFINrtT8sgpKiMpu4iC0gq01mgN9ivclFdYyC+tvu6jFwzi5knuDQARodAMPPjggxw8eJARI0bg7e2Nn58f4eHhJCQksG/fPi655BKOHj1KSUkJf/jDH5g7dy5QXbKjoKCAmTNnMmnSJFavXk3Xrl35+uuv8ff3b+VvJghti/ySct5ZdZhvt6VwIKMAxzXCPD0Us0Z25a5pfYiPqr3UdqVFsyXpBMXllQB4eXgwumc4Pl4eVde/+d0NJJ8o5qN1SRw9Ucwr145ib2oeN7+7ER8vD967eRyn946sEhZgovaKyirJKyknr7gCb09FXGQgHh61Z/VnD4rl7EGxLn/nzIJS9qXlsz+tgAm9Il0+r6mcciuvjRkzRjvWPtqzZw8DBw4E4Ilvd7E7Ja9Z7zmoSwiPXTi4zuOHDx/mggsuYOfOnSxbtozzzz+fnTt3VoV0ZmdnExERQXFxMWPHjmX58uVERkbWEAp9+vRh48aNjBgxgiuvvJKLLrqIOXPmNOv3aAz2z1QQGkJr7dSsobVmx7Fcvtt+nCW70wCIDPIhMtCXMwfEMGtU16rBVWvN5qQcKi2aEd3DqgZqMIP1x+uTmLfsICeKypnYJ5JxcZEM7x5Kj4gAThSVkZFfxtrELD5en0R5pYWZQztz/tDOTOobRZCPF9/vOM5Lv+xnf3pBjT72jg7k7xcPYVx8BLe8t5HfDmTy7k1jSckp5uEvdxIXGcCxnGI6h/rz/s3j6B4R4MYn6T6UUpu01g2uNy6aghsYN25cjRj/l156iS+//BKAo0ePsn//fiIja0r8+Ph4RowYAcDo0aM5fPhwi/VXEE6GhZuSefTrnXgoRViAsalbtKa80kJeSQUZ+aV4eSgm9okiyM+LrIJSdh3P5cddqbz0637umtaH4rJK5q9P4oB1wA7w8WRCr0iCfL3YeSyXQ1mFaA2T+0bxp+n9Gd7duR9uxpBO3DWtD2/+lsjH65L4fvtxvDwU0cG+HM8toU9MEM9fOZwe1oH9eG4J/1q8l2veXEefmCAOpBfwz0uHMrmvKSbaOdSf3320mb4xwbxz01iignxb5qG2Iu1OKNQ3o28pAgOr1dZly5axZMkS1qxZQ0BAAFOnTnWaA+DrW/3P5unpSXFxcYv0VWi/aK3JKCjl2Iliuob7ExPsV2/75BNFrNiXyfJ96QD875pRNUwkznh9xUH+sSiBcXERDOkaSk5RGbnF5Xh4KHy8PPD18mB8fATnDu5Uw9avtWbp3nT+s2Q/D32xA4Dh3cN49rJhhAZ489v+TFYdyKS0wsLgLiHMGtmV0/tEMbpneIPfOzrYl4dmDuTP0/uz5WgOSxPSSUjN56HzBnL+0M54Oph0zhkUy6vLDzJv2UF+N7U3V4/rUXVsSr9oVv3lTAJ8PRt8Fu2FdicUWoPg4GDy8/OdHsvNzSU8PJyAgAASEhJYu3ZtC/dOaC1sptmTjRZZdSCTZ39MoF9sMOcN68zE3lE1TCuOFJVVcOeHm1mTmEVZhYl68fJQzBzamRtO68nonuG1+vTEt7t4Z9VhAGKCfUnPL+W91Ye5dXKvOr/b0z8m8NryRM4f1pnnrxyOr5frIZZKKc4cEMu0/jGsP5RNkJ8Xg7uEVh0/d3Anl69VF16eHoyNi2BsXES97fy8Pbn37H78bmofp881NKBjlXsRodAMREZGMnHiRIYMGYK/vz+xsdVOpBkzZvDqq68ycOBA+vfvz4QJE1qxp0JLkVVQyh8WbOVgRgFzJvTkmnE9CA9sfFTM55uS+cvn24kN8SMxI5XPNiUT6u/NX88byJVju9dqb7Fo7l2wlZX7M7j+tDh6RQfSKcSPdYey+XTjUb7dlsLZA2N55dpRVQPg4l2pvLPqMFeM7sbtZ/Sid3QQN727gRd+3scFw7rQKbSmhpGYUcDDX+5gbWI2cyb04ImLhtSafbuKUorxLeA8dYX6BG1Hot05moXmQZ6p6+QUlVFaYSE2xAyeO4/lcvsHm8goKGVEtzDWH87G18uDq8f14MGZAxpMWtJak1lQxvx1SbywZB8T+0Qyb85ofL08WHUgkzdWHGJNYhZXj+vOYxcOrnG9fy7aw2srEp2GLhaVVfDOqsP8a/FeLhrehRdnjyCzoJRzX1xB13B/vrhzYtXAeCSrkHNeWMH0QbH875pRAJSUV/LGikT+u/QAvl4ePHzeQK4a210y4E8RxNEsCI0kr6ScTYdPsONYLv1ig5ncN4pA37p/IqUVlby58hD/+/UAxeWVdA71Y0jXUFbuzyA8wIeFd5zGsG5h7E3N563fEnl39WE2J53g1Tmj6RJWHW5cVFbBb/szWbo3nXWHsjl2ophSq9nn0pFdefqyYVWD9ZkDYjmjXwz//mkvryw7yK6UPC4b1Y2wAG+OZhfx2opErpvQk5smxtXqb4CPF3dN64OHUjzzYwKh/t4cyS6iuLySF2ePrDFT7hkZyO+m9ubFJfu5aHgqSdlFvL4ikfT8Us4f1pnHLhzUoI9CODURTUFwSkd6pul5Jdz50Wa2JJ3AYvdz8PH0YELvSM4ZFMs5A2OrzCiZBaWsOpDJCz/v43BWEecOjmVsXATbknPZdjSH+KhAnrtiONHBNSNVFu9K5f5PtuLv48kfp/fnSFYRm4+cYGtyDmUVFoJ8vZjQK5Je0YF0DfOnV3Qgk/pE1TkTX7wrlQcWbie3uLxq35R+0bx9wxi86nGKaq355w8JvL4iEYC/XzKE6yb0rNWupLySc19cwZGsIgBO6xXJPWf24fQ+Ua49WKFN4aqmIEJBcEpHeqZ3friJXxLSueOM3kyIj2BIt1B2Hcvjlz1p/JKQzqHMQgCGdg2lsLSCROt2r+hAHr9wMFP6NbgWehUH0vOZ+/4mEjML8fZUDOkayti4CKb2i2ZMXESj7doVlRZyisvJKSqnoLSCwV1CXIqSsQmG/JIK/jFrSJ2CZ9ORbD5al8S143u6FPkjtF1EKAgnxan+TLXWbEvO5ZMNSew4lsu0/jFcPKIrfWKCarT7aVcqcz/YxJ/P7c9d0/o4vc7BjAIW70pj2d50Qv29GRMXwdi4cIZ1C2tSmGJRWQUH0gvoFxvcZouiCe0P8SkIHZZle9N5+ocEElLz8ff2ZGDnYF5eeoD//nqAYd1CeeDcAUzqG0V+STmPfr2LAZ2CmTvFeeilUoo+McH0iQl2KjSaQoCPF8O6SRFEoW0iQkFoN1gsmpd+3c9/ftlPr6hAnpo1hIuGdyHYz5v0vBK+3X6c91YfZs5b6zhvaCd8vTxJyy9h3pyGk7SEdkJ5CSx9EsJ6Qs/TIXogeLTy374wCza8AePvAP/WnyzIL6EVCAoyJoyUlBQuv/xyp22mTp2Ko5nMkRdffJGioqKq7fPOO4+cnJzm6+gpRG5RObe8t4EXl+zn0pHd+O6eyVw7vifBfibxKCbEj1smxfPTfVP44zn9+DUhnS+3HOOG0+IY2UNs5S1OeQlYLA23a24OLYfV/4VFf4J5p8Oz8fDGWfDpDfDT/0HusZbv08p/w7J/woJrzXNpZUQotCJdunRh4cKFTT7fUSgsWrSoza/NkJ5fwpw317HhcHazXbPSorn9Q1PI7KlZQ3juimH4+zi31ft5e3LPWX1Zcv8ZPHzeAB6Y0b/Z+tHiVJTCxrehMLPl7601FDdxAlKSBy+NgOVPN+38vONw+Lfa+ze+DfMm1j+wppqSGtyxCi55FQZfAr5BkLYT1vwPVv3H9X7s/RHWvwEluXW3OfwbvH8xvDoJXhhi+mffviQPNr8P0QPgyG/wxa1gqXS9D25AhEIz8OCDD/Lyyy9XbT/++OM8+eSTnHXWWYwaNYqhQ4fy9ddf1zrv8OHDDBkyBIDi4mKuuuoqBg4cyKxZs2rUPrrzzjsZM2YMgwcP5rHHHgNMkb2UlBSmTZvGtGnTAFOKOzPTDBDPP/88Q4YMYciQIbz44otV9xs4cCC33XYbgwcPZvr06S1aY6m4rJLbrFUov92W0mzXfXX5QdYmZvPUrKFcO76nS8lU3cIDmDulNwE+TbSgHtvUOrNKGyeOwNvnwnf3wbImDK4Wizn3X33NYPXSKPj+jw2fV5JnBsJ5p8O/esPx7Y2/9/rXIP+4GQwbOwBqDQtvhnfPh7XzqvcnfA/f3W8G9+zEus9P22lMR52GwIir4cL/wPVfwz2boP95sPtr1/pUkgtf3GY0jucHwfd/gpwkJ9/1dUjeCCFdoccESNsFK/5VfXzrfCjLh4tfgRlPw55vzTVbMQCo/fkUfniwejbQXHQaCjPr/uHNnj2be++9l7vuuguATz/9lMWLF/P73/+ekJAQMjMzmTBhAhdddFGdA9a8efMICAhgz549bN++nVGjRlUde+qpp4iIiKCyspKzzjqL7du38/vf/57nn3+epUuXEhVVM25806ZNvPPOO6xbtw6tNePHj+eMM84gPDyc/fv38/HHH/PGG29w5ZVX8vnnn7dIiW6LRXP/p1vZfiyXTiF+bDpyoknXWZuYRVJ2EReP6IKvlydbkk7w/M/7uGBYZ64Y3a2Ze10Pn1wPMQNhTtM1vSaz7yczIGmL+d/c+Tmc+w/wakQZjZ//z8ysB1wAfqGQsgU2fwAznwWPOiKiktbCB5dCeSF0Hg6ePrDuNbjkZeftnVGSC6v/B4HRRjAc/g16neH6+YdWQNJqM7D/+CBUlEDcZFh4C4R0gbxjcOIQxA5yfn7qTvPMnDF4FiR8Z75n3MT6+7HpXSjNg0vmwaGVsPk9SFoDd66qbqM1JK2D/jPhsjfNPi9fWPsqjLoRIuJh3TzoNg66jTav/ONGWxl0SeOeSzPiVk1BKTVDKbVXKXVAKfWgk+M9lVK/KKW2K6WWKaVa8FfdfIwcOZL09HRSUlLYtm0b4eHhdOrUiYcffphhw4Zx9tlnc+zYMdLS0uq8xooVK6oG52HDhjFs2LCqY59++imjRo1i5MiR7Nq1i927d9fbn99++41Zs2YRGBhIUFAQl156KStXrgRatkS31pqC0goOZRby1KI9/LAzlb+eN5ArxnQjITWfwtKKOs9Nyyshv6S8xr7jucXc9t5GHli4nWn/WsZ7qw/zhwVb6RTix1OzhrZcuQWtoSANEpcaJ6E7yNwPr042g7U9CYtg/pUQ2h3mLoNpj0BxNhxY4vq117xsTCXjbofZH8Ilr8D426GyFHKP1n3e+jfMoHbbr3D7Chh+Fez4rHHPYN1rUJJj7usTDNs/df1cgOXPQFAnuHM1DLkcljwO710IwZ3guq9Mm+xDzs8tK4SsAxA7xPnxfjPAyw92fVl/HypKjZYSfwaMuAZmzYOzHjVaSJ6dBpxzBApSjYZg46zHzD1++ivs+xFOHIbT7FZsnPIAKE/j+2gl3KYpKKU8gZeBc4BkYINS6huttf2I9hzwvtb6PaXUmcA/getO6sb1zOjdyRVXXMHChQtJTU1l9uzZfPTRR2RkZLBp0ya8vb2Ji4tzWjK7IQ4dOsRzzz3Hhg0bCA8P58Ybb2zSdWy0RInusgoLb/12iHnLDtRYovCa8T24ZVI8y/ZlUGnRbEvO4fTetbNjKy2aS15eha+XB5/feTqRQb5orXnky52UWyz8+4rhzF+fxGPf7MJDwae3n0aofwtWsiwrAItVYO35Gsbc3PA5pflQlA3htTOHnXJgCaRuh4+uhFt/hvA4SNttNIQuI+DGReATAGE9ICAKti+AAec1fN1dX8Lih2HghTDjn2ZBYIDIvuY984C5lyMVpbBvsbHBdx1t9o29zWgbm9+DyffXPsdigbUvQ+cRED/Z+CDW/A/6n28GykEXwZ5v4PznwNuFVQYPrYQjq2DGM8YPcOnr5hns/xmuXQiRvcE31GgKzkjfA2hjOnKGbxD0nW5MSDOfqVtj2v6pmdFfbKchxU+p7uPw2eZzkrUicnc7oRAUA2c8YDS11B0Q0g0GXFizD11HOfeZtBDu1BTGAQe01ola6zJgAXCxQ5tBwK/Wz0udHD9lmD17NgsWLGDhwoVcccUV5ObmEhMTg7e3N0uXLuXIkSP1nj9lyhTmz58PwM6dO9m+3dhq8/LyCAwMJDQ0lLS0NH744Yeqc+oq2T158mS++uorioqKKCws5Msvv2Ty5MnN+G3rZm1iFue/tJJnfkxgbFwED80cwPNXDufj2ybw5MUmc3ZUdxPts7kOE9Kag1kczy3hcFYRN7+7gaKyCr7ZlsIvCen8aXp/LhvdjYV3nMb8W8fz1g1jGdNAaeRmp8jOSb7j84bb5ybD69OMHdxVjm83Zp3KMvjwcjNYf3wV+ATCVfPNYAjg6Q1DLjNOz4Ycv2WFxm/QdQxc+kbNQS/KKhSyDjg/N3GZsX0PsvuJxg4yppuNb0OlE63v+Fb46RF47wIzm//xIWM+mmo1Ggy70phg9v3o0iMxWkIsjL7BbHt4wkX/hft2Q1QfI+Ai4urWFGxm5bo0BTAmpMJ0OLLa+XGLBVa/ZExQvc+s3h87FPzDjXnLRtJa8A0xZkZ7xt8BEb2NqWv8XPB0mJvHTYJjm83fqxVwp1DoCtjrosnWffZsAy61fp4FBCulatXRVUrNVUptVEptzMjIcEtnT5bBgweTn59P165d6dy5M9deey0bN25k6NChvP/++wwYMKDe8++8804KCgoYOHAgjz76KKNHm9nY8OHDGTlyJAMGDOCaa65h4sRqW+fcuXOZMWNGlaPZxqhRo7jxxhsZN24c48eP59Zbb2XkyJHN/6Ud+GlXKle9vpbi8kreumEMb904ltvP6M2lo7pxWu/IqvVqQwO86RMTxOYk54PYN9uOEeTrxX+vHskOa8XRx7/ZxcgeYdw00VT+VEpxep8opg2IabhjleUNt2kMRVZzSecRZuaaV4/TPOsgvD0Dsvabdq46EFO3G1vz1QuMA/OVCZCfagRCSJeabYfPNqaf3VbzicViBurj22q22/Su6fu5/6g9Mw+MNrPsrP3O+7P7GzPA2WbENsbfbkxO+36ofU7SGvN+xoNmlr5tvvFhdLaaRuMmG1OQKyakw6vg8EqYeG/tvtvnGYTH160ppO00JquwerS1fueCl3/dJqR9P0DmPtMPe3Olh4f5PoeWV/+Nj66DbmNraxxePnDBC9B9PIy6vvY9ek4ymujR9XX3051ord3yAi4H3rTbvg74n0ObLsAXwBbgPxjBEVbfdUePHq0d2b17d619wsnR2GdaWFquT/vHEj3jxRW6qLSiwfYPfLZND39isa6stNTYX1JeoYc89qO+75MtWmutP1x7WPf8y3e678OL9L7UvEb1SWut9dYFWv+jm9YFmY0/ty72/6z1YyFab/7AvK/+n/N2qbu0fraP1s/Ea/35XNO2OKfh65cVa/1EhNZLnjDbO7/U+qkuWm/92Hl7i0Xrl0Zr/fZMrUvytJ5/lbnXcwO0Lso2bcpLtH6uv9bvnF/3fV+fpvW7F9beX1Gm9dM9tV54q5Nj5Vr/e5DW715Q+9iCa7V+Yaj5XFpgnlfe8ZptfnxY6ycizd9n389af3yN1p9cp/Xub0yf89O0XvyI1k921vpf/bQuK6q7/1pr/fNj5tlVlNc+9uZ0rd86t/7ztdb6k+u1frZ33dd4YYjzY+teN88966DWRSe0fixU62XPNHw/R0rytH48XOslf2v8ufUAbNQujN3u1BSOAfargHSz7rMXSCla60u11iOBv1r3dczsq1Ocl345QEpuCU9eMrjOHAF7RvcMJ6eovKq4nI3lezPIL6ngouFmNnzt+J48felQXpg9gr6xwY3v2NqXjYni0LLGn1sXRVazV/fxJgpnRx0RSIv+ZN5v+qF6hl3kQn5G+m6wVEAn64x68CXwlyPGsesMpWDYbKO1vD7N2P5Pu9uYQRY9YNps/cjYwaf8qe77RvZxbj46/BsUnzA+AEc8vWDsLcZskrG3er/WxnzS4zSz7RMII+cYh7A9w2abWfF/R8JHl5nZ8ZE18MkceK4vvDjM+CEGnAc3LWrY9xAeb55dXnLN/RaLCQetz3RkY/AsKMwwz9Oe49vg6Fpj/nE0+YBxPIN5FskbAG3+RxqLbzB0GdlqfgV3CoUNQF+lVLxSyge4CvjGvoFSKkopZevDQ8DbbuyP4CYOpOfz5spErhjdjdE9XbPvj+ppkuw2J9X0K3yzLYWIQB8m2pVnvmpcD84f1rnxHUvZWm1CSWzGaI5i68DuH2EiYFI2146Nz0kyg8q4uRDdHwIiap5bH6nW2P/O1RFoTgche4Zdad4LM+C6L+Dcp0wky45PjdD67QXjS4ivJ8wxsq+xczvasvd8C94B0Pss5+eNnAMo2PlF9b7sRNMX+8gbZ3Qaaq4bPQAufRPu2wX374E5n5u8gWFXwF3rTUhnZO/6rwUmzBNq+xVyjhifSF1OZnv6TjdEpfJNAAAgAElEQVRmpg1v1ty/4U1jWhpxjfPzovpCcGdr2OwaE0XUrcH6c86Jm2RyYcqKGm7bzLhNKGitK4C7gcXAHuBTrfUupdTflFK2KcdUYK9Sah8QCzx1Evc7yR4LNhrzLLXWPPr1LgJ8PPnLzPr9Jvb0igoi1N+7hrO5sLSCJXvSOG9op+apRbT5fRP+13OScZQ2F0XZgDJ1agbPMvt2OjicbdrDUGsZE39rKY1iF/Izjm839vuwONf7FN4TrvsS7lgJvaaafZPvNzPOL+YaITXlTzXt4I5EWQv+2WsLFouJ3e9zdrVz25GgGOg50UTt2LD5E2yaQl0oZYTYLT8ZAeDlYwRgn7Nh1qvGkWxzgrtCuFUoOPoV0naa99g6chTs8QmA0+4ykVHHNpl9xTmw/TPz97T9LZ19l/gpVqGw1gh1n0DX+25P3GSjQSW3vF/BrXkKWutFWut+WuveWuunrPse1Vp/Y/28UGvd19rmVq11aVPu4+fnR1ZWlgiGZkBrTVZWFn5+tVfVKiitYOGmZOa8uY7+j/xAn4cX0eevP7D6YBZ/njGAqCBfJ1d0joeHYlSPsBpJbD/vTqOk3MLFIxzjEezIPWYcrg1RVmhi6AddbF45R+qOSmksxdkmMsjDE8K6mx/w+jdrzup2fGYcxbaZq79VUyhyQSikbjcz6MYWaut9pglRteHpDbNeM++xQ0wcfn1UhaXaOZuPrjM5GYMaCAwcdDFk7Kk2ISWtMd85uoXLiIR0MUl1jn/r1J2gPGpHAtXFaXeZ/v/yN7O9dT5UFMPYW+s/L36K1fS0umYoamPpMd5oGq1gQmoXGc3dunUjOTmZthqZdKrh5+dHt2418wi/3JLMQ1/soKTcQo+IAK4e14MAH0+Ugs6h/lw9rkcdV6ubUT3CWbo3g9yicjw9FZ9sOEqXUD9GOytQl3vMlAfY8oGJ6Li5gTDGXV8ZX8KoG0xkDZjIENsgfTIUZVebgwCmPQzvzDQlDSbdawag9N1w3nPVbVw1H1kqzfmjbzz5foIZlG/5CQIi69cSwGqeUSZiysaeb80g23d6/ecOvBB++LOJUjrjz8Yv0GNCw/dsbjw8TXSRM00honfd2o4jfiFGs1r8sNEyN7xp/u+6jKj/vKroLG0G9qbiG2zudXhVw22bmXYhFLy9vYmPb4Yfu+CU9LwSHv1qFwM7h/DI+YMY1SOs6dnDFaUmC3X0TVX+h98v2MKmIyeYVL6auyZOqgpdrWL9G7D4r6asQ2CMyQJtiM3vmZlvz9PNdnAX8+NujsG2KKt65g/mHn3OMXb70TcaO77yrDYtAfhZCxU2ZD7K3G9mpPb+hJOl83DX2nn7m0xpW1iq1sZ01GuqGSTrI6SzmRnv/trkEWQfrM4naGki4iH7cM19qTtMUlhjGHMLrHkFPr/NOO3P+EvD54T1qA6LPRlNAYxfYe08o4F6+phMcC8/k+DmRqQgntAgf/9+D6WVFp6/cgSje4afXDmJ1f+Fta9AwncM7x6Gr5cHaxKzmD4ohlf8X+Wa4gW1z1n5vDGn/H6zKWJWkF5/2eX0PcbsMep6M1NVytSRObSieco1FztoCgBn/Z/50a76j0lo63MWBNpla3t6mTyAhqKPqpzMLg7kzU1Un2rzUdouY3Yb4GLS3aCLIW2HiXSChv0J7iI83ji6bebkklzzPVyJPLLH2w+m/sUIhIDIhk1oNgZeaCLHQpoQHGFP3GSTvPivPvD3SFOAcKf7a221C01BcB8r9mXw7bYU7ju7H/FRDTjNUrYa8860v9YeNMHM8G0VIgvSCfT14oc/TCY8wIdwz2J4usRkwdpTkA75KcbGG9bDZLTqSjMwB9YukQGYwm4e3jD86up9vabCto+NGeFkZ+FFJ8ziLPZ0Hm4yi1e9aDSasx+vfV5AeMOawvFt4OkLUf1Oro9NJbIvHP2oWktAmSggVxh4ISx+yAhxLz+T3NcaRMSbon2FGcYJbqvkWlchvPoYfo1Jrus3wwgJVzj7CaAZ/JvxU2DCXeb/3T/caKdNCXFtJCIUhDopKa/k/77eSa+oQO6Y6ny5SsAMIJvehR/+YjJrK0rh4v/VbrPIWuwrMNoUCgN6RVtV4YzD5j3rgJnZ+YWabVtIqc2WG2TNYC5Icy4UKkrN4D/gfAiKrt5vC8VMXHbyQsGZpgBGGO76yoRv9p9Z+7h/eG2fQu4xM/iOvslE3qRuN+UjPFuwlpM9UX1Nbaf8VNOv7uOrn3lDhHU3Ya/HNpqIr8ZUbW1ObBFI2YdM37ctAO/Apmkunl5w43eNO6e5VnLz8oUZ/2ieazUCMR8JTikpr+ShL3ZwJKuIJy8Zgq+XQ0KaxWIGtCNr4Ms74Lt7Tbnh0TcZbcGxdkzC97B/MUx7yMyCC9JrHi+wqyBrX6M/xao52BK5gmJrt7dnz7dm4HW0Z4d0hqj+xtlcWWFqBS15vPH1ZSrKzKDpTChE9oZznoCpDzm3+/pH1DYfbV8APzxgEreKc8x379SM/oTGYssFOPiLscMPvKBx59tMLA3lJ7iTCLuw1KJsY3IZdmXDfhEBEE2hw6C15oO1R+gXG8yEXrXKS9XgQHoBd8/fTEJqPn84qy+n2xLJinPM7HHn5yZUrrLMeoaCqQ+baI2KEjjwC3x7L9zxm5ktZh4wWkTMYJMNemxT7TUv7Af5lC2mqiYYc1JE7+ofdJVQcBAqNja/Z8xM8VNrH+s11Wg0Lwyqvl9kXxh5bXWbkjxTmnr6U6a+vSP2iWvOOP0e5/vBCBLHJLe848bUdWQNvDbF+CWa08ncWGxhqateMu+umo5sDLkMNr7luh/CHYT1BJTRFLbON/+TDYWSClWIUOggvL3qMH//bjeBPp58ddfEOktGLN6Vyr0LtuLv48knl4Yx/sR7sOCwWekrc68RBGE9zY8sso/5HN2vOj7eJxDO/zfMv8L4D5QH/Pa8yQS98j1jFgmKhXyH+v+2QdovtKZfIWVrzdC++jSF7ETjTJ72iHMVftDFRmh0HQ0jrjVlKPb/VFMo7P/JxNgf+c25ULDN9J1pCg3hzHyUf9w8x/OehQXWxY46tZKTGcwKYV7+5m8dM8i1LGJ7QrvCH7Y13M6dePuZ75GdaExZPU5zLZNZAEQodAiWJqTz1Pe7mdo/mp3H8pj7wSa+vnsiIX417daHMwu575Ot9OsUzGtzRtPp57tMtcjIPiZjtvc0syJU11H1x5/3m24G4BXPmu2hV5iZd7B1QA+KNSUHygqrMz4L0kzYXZzdwjKFmaaGTefbq6/tG2Tsw840hc0fGCFkP8jbEzcRHrETJvt+NHH1leXVNvwEq/24rgS5hjSF+vCPMP6SyorqshUFaea5xE+BW5eYKpyNDZ1sTjw8zN87bYepaHqqEhFvTJblhcbXI7iM+BTaOXtT87nn4y0M7BzCK9eO4pVrR3E0u4j7FmzFYqmOkKiotHDfp1vx8lC8OmcUnUL9zGyx95lw93q49jOY/ncze3YlJHXmszD0SlN64bI3qwUCODcB5aeZ/V1GmBleSW61P8ExYSgopramUFluQiH7Tq9dWrou+k6H0tzqEsUVpbDfqsHkH3d+jq1sdlM0Bds59gu356eaejlgNK6Jf6h7cZeWwlbuojVNQCdLeJwRCIHRJipKcBkRCu2YYznF3PzuBgJ8PHnzhjEEZO1mXIyFRy8cxC8J6fx54XbS8swqbq8sO8iWpByemjWUzqH+xpGceaDpoZHBneCyN2ouRGLDmQmowCoUOlvXfTi+rdqM5BizHxRbeya//ydzjVGNSJjqNRU8vMy5YOr1l+WbkNC6NIWik9EUbPWPrNewWKxCoVPd57QGAy6AfjNbL1eiObA5m0ddb6J4BJcR81E7JS2vhGvfWEtecTnzb5tA50Nfwde/g+4TuO7G70nJKeGNlYl8tz2Fi4Z34Ystx7hkRBcutJasJjfJZNZGuyFePtiZUEg3szubVpCyxepk7lUdnmojKKZmmWYwSzL6hTZcjsEevxBjb97/s4kaSvjemKZ6TzOJW84oPhmfgq3UxYnqa1nKqzWFtsLQy6sL+Z2q9DjdZLGPvqm1e3LKIZpCOySzoJRr31xHen4p7948jqHHP4Ov7jCrXCWtRh1axoMzB7D0j1OrBEJssC9PXGznjLNltUa5oaCZM/NRQaoRFoFRptRCylZI2eY8ASootrb5KDvRaDUNlZh2pO90SN8FOUdh7w8mEzk8zszgnRVYLMo2jlhX1hR2xKYp2LQNm4mqrWkK7YGep8Ef95jcCaFRiFBoZ1RaNNe/tZ7kE0W8fcMYRh9916zL228m3LXWLBT+61OgNT0iA/jXFcNZ8cA0vrprIqH+do5n20zcHZm1AZHGIWwz0VSWG1u9TVh0GWGiiHKTnBcgC4o1oZsVdkV1sw8ZraKx2DSL5c+YQXrA+WaQriiuafu3UXyiaVoCmIxmqNY28q2CLUiEgtB2EKHQzth6NIfdx/N4dnonJqy90yRoDZ4Fsz8w5pUpfzJhejY7OtA1zJ+YEIcU/sy9ZvAOrD+noUl4eJrCdrbZvk1jsGXOdh4BRZnVnx2pymq2nldRatYJbopQiO4PoT1Mwp3yNELCZs5x5ldwrJDaGBzNR6IpCG0QEQrtjOX7MpjuuZELVl9qZtszn4XL3q4OuRw5x+QWLDXaAvsWw+tTTbKZPZn73WM6shEUUz2oFzjMmLuMrG7nzNnpaH46cQTQ1eUNGoNS0Pcc87nn6WbAtw3SBU6EQnF205zMYBbOUR525iPr9UUoCG0IEQrtjG0Je5nn/SIeod3g9hUw/vaaiVye3qYE8PFt8MoEk717fLsp+lVRVt0uY2/jVrxqLPZ+gSpNwWY+sgqF8Hizulmtc+3qH0F1lnBTNAWoNiHZQjBtwqm5NQUPj5oJbPnHjYCR6BihDSFCoR1xorCMqNTf8MRiCtLVterVsNkmW7X4BJz/PFz6uonpTtlsjhdmmoHLnatm1RAKthmzVSgERBhfRl31cxxDWk9aKJxjsrBHXV+zH85yFRzXUmgs/hF25qPUthd5JHR4JCS1HbHyQCZTPLZT7h+Fd31r0Xp6wa2/mBh9L5/qNYcTl5uBOHOfaedO81FwbPW6CDZNIdCuqumN35vyy86wtbOdl51o1ipo8gzes2ZtHN9gs3C7o6ZgsRgHd1PvA0ZTsJmPbBFXgtCGEE2hHbEiIZUpnjvw7Ht2w+V7fQKqSxsHRJgibIeWm21b5JE7chRs2K+LkJ9qBkt7M0pQTN1VLb18zIzbXlOIiG/epR+DO9UWCiU5Zq2Ek9EUAiLszEeiKQhtDxEK7QStNRn71hFOPh42x2ljiJ9iyj2UFRlNwTvAhK+6C5tfID/Vms3cSGervfkpO7HppqO6cCYUbGafk9IUIky12baazSx0eEQotGXyjpv1ib++27yn7qxzOck9x/MZXrIRjYJe0xp/r/ipJrs2aY0RCpF9mm+xEGfYhEBBmjEDubqQS9X51uilynLISXKTUHDwKZxMiQsbNvNRUabRlERTENoY4lNoi6TvgW/ugeQNZts31MTRAwREwW2/mqqldizfl8EZntuo6DQC76bkFvSYYHwMh1ZAxr6a5ardgX2uQUFq41fFCoo16zDnJJnB1V2agtbVZqmTKXFhIyDcOPVzkqrvIwhtCBEKLYHFYkoRH14FR1YZB+q0h53Xqtcavvk9ZB2EMx+BAReaKKDco5CwCH78i5nNOwiFjQkHmetxEM9+f2paH32DoNtYU046NwmirmvadVylKoIo9eQ0hZONPKqL4M5madGSnNrlKWzbTcGmZaTvrr6PILQhRCi0BN/9ATa/bz6Hx5mQzz3fmDLJk+43Tl8bu7+C5PVw4Us1l5QM62EiZH5+FNJ2UlhawYtL9pGUXURecQVRySvx9LJAn7Ob3s/4M2D50+azuxeO9w0CnyCTJFdRUi0kXCUo1pSisK3h3NxCwdYfmxMc7DSFk8jytl0rbXfN+whCG0F8CifLmlcgeWPdx8uKYPtnZnGa+3abVanu2WRKT6z4F8w73WgFYMo1/PwYxAwmf+BstGNBNk8vozWk7eLFJft4Y+UhjmQVUWnRzA7fR6VvqFlVrKnET6n+7M4cBRtBMWahemiaoxkgaa2pbNpYTaMhqkpd2PkVirJMKQzHqq2NIcBBUxChILQxRFM4GYpzYPHDMPxq6DbGeZvEpWZGO/pGs1QhGDvypa/DyOvgsxvgrekwZyEcWgk5R/h37DP892+/EOLnRe+YIIZ2DeWP0/ubgnWxQyjfv4S39xzm6nHd+eelw4zJ6fltpuRzY6uE2tNtrKkAWlna/DNvZwTFVgvUppiPwERMRfRq3nBUqLb159tVYy3KNjP9k7mXvfkoIKo6LFgQ2giiKZwMyRsBDScO190mYZFxFMdNqnUoJ3Y8e2Z+RrmnH/qd8ylf+gwrGcnrx3py+5ReXDSiC75eHsxfl8T1b60jt7gcS8wgvIvS6elXxAPnDjAXykgwM1pnC9o0Bi8foy1E9W+Z0gtBMSbiCRrvcLXNsEtzqxdUaU6qhIKdplB8EiUubNjMR4UZ4k8Q2iSiKZwMR9ea97qEgqUS9v1IbrepfLrqKOGBPkQEepOSU8KPO1NZk5hFpUUTw0O85/M0fdUxPoucy6KrJ9M7OqjqMkt2p3HnR5uY8+Y6ftc9kpnAo+MgPNA6yzyyyrzHTz7573Tx/6C86OSv4wr2JqNGawp2Zhd3aDU+gaaAnX2uQtFJFMOzYS9UJPJIaIOIUDgZkqxCIf84lJeAt0NZhqProSiTJw/G89muPTUOxUcFMndKL0b1CKewtILN+aM4asnkhSmT8fSoaZ44e1Asr103mjs+2Mz/HYOZfjAlxM6scWS1mXU2pUqoI81tm3flXp6+4Oek8F19+IebEFpLhftMXY65CsUnjMP/ZPAOAE8fqCyTEhdCm0SEQlOprIBjm0wkSlGWCRl1qCpq2fMdFrxYbhnOj/dOJsDbi+yiMgJ9POkTE4SqYZvuWu/tzhwQy2vXj+bfP+2loigKrwyro1JrIxR6Tmx+u7q7sc32g2Ib33cPD7MmQ36Ke4VCgYNPwdn6Do1BKWuJDilxIbRNxKfQVNJ2GDPLkMvMtqMJSWvytn3NqspB/PmisQzoFEKPyABGdA+jb2ywg0BwjWn9Y/junsl4dR5SvYbwiUNmNtvz9JP7Pq2BzXzSVO3Edp47fApgBm2bplCQDoVNyKdwhs2EJOYjoQ0iQqGpHF1v3ofNNu8OQmH/ro2EFR8lpdM0Lh/dzDWEYgabrGdLpdESwGgKpxq2Abapg2NQrDE9BXdpvj45Xt+W1bzy3+Z95JyTv67N2SyagtAGEfNRU0laCyFdTV6Al18NobDpSDbrvnyHvsB5l9/SJK2gXmIHm4Sv7ESTJR0Q2TJ5Bc1NlfmoibPvvudAYJT7ajQFdza2/9QdsPFtGHGN8yz0xlIlFERTENoeIhSaytH10H0c+9ML6B7UHe/sw3hozburD/PU93v4wm8zxVFDCY3t2fC1GkvsYPOetstEHvU8/dTzJ4BZFyEgEqIHNO38cbc1b38csQ3a3//RvJ/xl+a5rs181NiEPUFoAUQoNIXcZMhLxtLtbi6bt5rnK4Pokr2DW5/+lZTcEs4dEMHQo4movm4atKL7m7V+9/8EOUdgwp3uuY+78fCE32814Z9tEZt5J3k9jL8Dwro3z3UDY0zkVEtGegmCi4hQaApH1wGQGjqCvJJsArr1pveJfQzrGsrNk+K5uVcu6o1S6DrKPff39jelrXd+brZPRSezjboW0mkL2EJGvQNg8h+b77rjbzdJgp7ezXdNQWgmxNHcFJLWgXcAOyqMA7ln70H4Vhby6mXx3Dq5Fx62tY67uEkoQLVfwTcUYoe47z4dmeAuZlnO0+5u3ll9UAz0OqP5ricIzYgIhaZwdC10HU1CeglKQVR3a0VRm7M5ZbOJRQ+Pc18fbH6FHhOMGUZofrz9TAHDaQ+3dk8EocVwq1BQSs1QSu1VSh1QSj3o5HgPpdRSpdQWpdR2pdR57uxPs5Cxz0Sj9DiNvWl59IgIwDfaGpFy4pB5P7bZRCW50/kbYxUKp7Lp6FQgMPLUdOILQhNxm1BQSnkCLwMzgUHA1UqpQQ7NHgE+1VqPBK4CXnFXf5oFiwW+u9fUxBk3l4TUfPrHBlcveJNzBEoLTIE6d/kTbMRNhAEXwJBL3XsfQRA6FO7UFMYBB7TWiVrrMmABcLFDGw3YPI2hQIob+3PybP3IhICe8zdKfCM4nFnIgE7BJnomMMaYj45vA205uXUNXMEvFK766ORr8QiCINjhTqHQFThqt51M7QI/jwNzlFLJwCLgHmcXUkrNVUptVEptzMjIcEdfG6YgA356xKwlPPI6DqQXYNHQv5NVpoXHGaHQEk5mQRAEN9HajuargXe11t2A84APlFK1+qS1fl1rPUZrPSY6OrrFOwkYgVBWCBe8CB4eJKTmA9C/U7A5Ht7TCIVjmyC0BwS1Uj8FQRBOAncKhWOAfbZPN+s+e24BPgXQWq8B/IAoN/apaeSnwfYFJkksxmTf7k3Nw8fLg7hI6/rK4XEmqe3oBvf7EwRBENyEO4XCBqCvUipeKeWDcSR/49AmCTgLQCk1ECMUWsk+VA9ZB8x7r6lVuxJS8+kbE4SXp/URhscZX0JesggFQRBOWdwmFLTWFcDdwGJgDybKaJdS6m9KqYuszf4I3KaU2gZ8DNyoa61W3wbITjTvdnX796bmV5uOoGZOgrudzIIgCG7CrWUutNaLMA5k+32P2n3eDbT9ms/ZiaZWTaixhp0oLCM9v9REHtmoEgoKOg9v8S4KgiA0B1L7yBWyEyGsJ3iax1XtZLar2xPc2SyzGNEbfIOdXUUQBKHNI0LBFbITHUxHeQAMtNcUPDyN2UhMR4IgnMKIUGgIrSH7kMlPsLI3LZ/wAG+ig31rtr3pB9NeEAThFMUlR7NS6gul1PnOcgjaPYWZUJZfQ1NIsDqZa62oppT7VgETBEFoAVwdwV4BrgH2K6WeVkqdgms/NhGHyKNKi2Zfaj4DOrXhdQAEQRCaiEtCQWu9RGt9LTAKOAwsUUqtVkrdpJRq3yuFOAiF1QczKSyrZHx8RCt2ShAEwT24bOtQSkUCNwK3AluA/2CExM9u6VlbITsRlGdV4bmvtqQQ7OfFtAGylKIgCO0PlxzNSqkvgf7AB8CFWuvj1kOfKKU2uqtzbYLsRLM2r5cPJeWVLN6VynlDO+HnLQvbCILQ/nA1+uglrfVSZwe01mOasT9tD7tw1CV70igoreCSEY7FXgVBENoHrpqPBimlwmwbSqlwpdTv3NSn1iVjb81tO6Hw1ZYUYkN8Gd8rshU6JgiC4H5cFQq3aa1zbBta6xPAbe7pUiuSshVeHge7rXX7irKhJAciepFTVMbyfelcNLwLnh6yPKMgCO0TV4WCp7ILyrcutenjni61IrnWNYE2vWPes61rLkf04vsdxymv1FwspiNBENoxrvoUfsQ4lV+zbt9u3de+KMo27weXwokjNcJRv16aQp+YIAZ3kfwEQRDaL64Khb9gBMGd1u2fgTfd0qPWpCir+vOWD8DDG1Dk+nZh/eED3Hd2v9pZzIIgCO0Il4SC1toCzLO+2i/F2eDlD3GTYMuH0HMihHYjrdgc7hMT1Lr9EwRBcDOu1j7qq5RaqJTarZRKtL3c3bkWpygbAiJg1PWQfxwSvoOIeDILSgGICGx/bhRBEAR7XHU0v4PREiqAacD7wIfu6lSrYRMK/WdCYAxUlEBEL7IKygCIChKhIAhC+8ZVoeCvtf4FUFrrI1rrx4Hz3detVqIoC/wjwNMbRlxj9kX0IsuqKUQG+dZzsiAIwqmPq0Kh1Fo2e79S6m6l1Cyg/RnYi7MhwJqYNvpG8A+H7uPJKizDQ0GYf/uu/ScIguCqUPgDEAD8HhgNzAFucFenWo2iLGM+AoiIhwcOQY8JZBaUERHoi4ckrQmC0M5pMPrImqg2W2v9J6AAuMntvWoNLJVQnGPMRzas4adZBaXiTxAEoUPQoKagta4EJrVAX1qX4hxAV5uP7MgqLCNShIIgCB0AV81HW5RS3yilrlNKXWp7ubVn7iR1B3x8DZSXVO8rtmYzB9RePCeroJTIQHEyC4LQ/nE1o9kPyALOtNungS+avUctwZqXYe/3kLUfOg01+2zZzE6FgmgKgiB0DFzNaG4/foTyYtjznfmcl2InFKyagn9NoVBSXkl+aQVREo4qCEIHwNWV197BaAY10Frf3Ow9cjf7FkNZvvmcd6x6f5X5qKZPIbvQJK5FSjazIAgdAFfNR9/ZffYDZgEpzd+dFmDHZyZbuSjLaAo26jAf2bKZJXFNEISOgKvmo8/tt5VSHwO/uaVH7qQ4B/b/DGNuhj3fOAiFbPD0AZ+aOXmZhbZsZtEUBEFo/7gafeRIXyCmOTvSIiR8B5WlMPRyCOlS03xkK3HhUBq7SlMQ85EgCB0AV30K+dT0KaRi1lg4tdixEMLjoOtoIxTS91QfKz7hPEdB6h4JgtCBcNV8FOzujrid/DQ4tBwm3W+0gZCusH8JaG227Utc2JFVWIavlweBPp6t0GlBEISWxdX1FGYppULttsOUUpe4r1tuYPdXoC0w9AqzHdIVyguhJNdsF2WbAngOZBWUERXkKyuuCYLQIXDVp/CY1jrXtqG1zgEec0+X3ET38TD1YYgZYLZDuph3m7O5KKuOEhel4mQWBKHD4GpIqjPh4eq5bYMuI8zLRkhX856XAtEDrD4F59nMUgxPEISOgquawkal1PNKqd7W1/PAJnd2zO1UaQrHoDQXdGWdjmZxMguC0FFwVSjcA5QBnwALgBLgLnd1qkUI7gQooynUUeJCa02mVEgVBKED4Wr0USHwoJv70rJ4ekNQrNEUipyXuCgoraCswkKUVEgVBKGD4BXHD1AAAA5BSURBVGr00c9KqTC77XCl1GL3dauFCOlCYeZRnlxoTc6us8SFaAqCIHQMXDUfRVkjjgDQWp/gVMxodiSkC6XZRzmRkWq2HYVCoSSuCYLQsXBVKFiUUj1sG0qpOJxUTXVEKTVDKbVXKXVAKVXL/KSUekEptdX62qeUynF2HbcR0hW/4lTClKmaujWz5uPIlBIXgiB0MFwNK/0r8JtSajmggMnA3PpOsK7t/DJwDpAMbFBKfaO13m1ro7W+z679PcDIxnX/JAnpQoClkHHhBVQUePD4T8l82bdnVaKazXwkaykIgtBRcElT0Fr/CIwB9gIfA38Eihs4bRxwQGudqLUuw0QtXVxP+6ut124xSgI6ATDEI4kKnzC2Jufy7fbjVcdtdY/CA71bsluCIAithquO5luBXzDC4E/AB8DjDZzWFThqt51s3efs+j2BeODXOo7PVUptVEptzMjIcKXLLnG43PjOY4v34xsSzcDOITzzQwIl5ZWAqXsU7OeFr5fUPRIEoWPgqk/hD8BY4IjWehrGzNOc9v+rgIVa60pnB7XWr2utx2itx0RHRzfbTXfmm7UTvMoLUAGRPHL+QI7lFPPRuiQAMgtKxXQkCEKHwlWhUKK1LgFQSvlqrROA/g2ccwzobrfdzbrPGVfRwqYjgPWZdgN+QAQT+0Rxeu9I5i07SHFZJVkFZeJkFgShQ+GqUEi25il8BfyslPoaONLAORuAvkqpeKWUD2bg/8axkVJqABAOrHG9283D9tRicj2s6RfWcNT7zulHZkEpH649IsXwBEHocLia0TzL+vFxpdRSIBT4sYFzKpRSdwOLAU/gba31LqXU34CNWmubgLgKWKC1bjDEtTkpKa9kf3oBJeGdCC3MqSpxMTYugsl9o3h1+UHKKy2MiatdJE8QBKG90uhKp1rr5Y1ouwhY5LDvUYftxxvbh+YgITWfSos2hfEKE2qUuLj37H5cNm81AFFiPhIEoQPR1DWaT3l2HjPLQwRGW3Py7LKZR/cM54x+xqEt2cyCIHQkOrRQCA/wJjDKKhQcKqTef04/vDwU8VGBrdA7QRCE1uHUWiinGdmZksuQrqGoUGvqhEPdo+Hdw9j0yDmE+HfYRyQIQgekQ2oKZRUW9qbmM7hLKPSdDqfdDV1qV9gIDfCWtZkFQehQdMhp8L60fMorNUO6hhgN4dynWrtLgiAIbYIOqSnsTTVVUQd2DmnlngiCILQtOqRQyCspByAiQMJNBUEQ7OmQQqGozJRYCvCVQneCIAj2dFChUIGnh8LHs0N+fUEQhDrpkKNiYWklAT6eElkkCILgQIcUCsVlRigIgiAINemQQqGwrIJAnw4ZjSsIglAvHVIoFJdV4i+agiAIQi06pFAQTUEQBME5HVIoiKYgCILgnA4pFArLKgmUHAVBEIRadEihUFxWib+3mI8EQRAc6ZBCobCsQjQFQRAEJ3RIoVAkPgVBEASndDihUFFpoazCItFHgiAITuhwQqGo3FoMTzQFQRCEWnQ8oVBqEwqiKQiCIDjS8YRCWQUgmoIgCIIzOqBQEPORIAhCXXRgoSDmI0EQBEc6nFAotJmPJE9BEAShFh1OKBSL+UgQBKFOOpxQKCw1moLkKQiCINSmwwmFYmuegmQ0C4Ig1KbDCYVCa56CaAqCIAi16XBCobisAqXAz7vDfXVBEIQG6XAjY2FZJQHeniilWrsrgiAIbY4OJxRMhVQxHQmCIDijAwoFWUtBEAShLjqgUKjE31uEgiAIgjM6oFCoINBXzEeCIAjO6IBCoVKymQVBEOqg4wmFUhEKgiAIddHxhEJ5hSSuCYIg1IFbhYJSaoZSaq9S6oBS6sE62lyplNqtlNqllJrvzv6A0RSkxIUgCIJz3DZlVkp5Ai8D5wDJwAal1Dda6912bfoCDwETtdYnlFIx7uqPjaKySnE0C4Ig1IE7NYVxwAGtdaLWugxYAFzs0OY24GWt9QkArXW6G/uDxaIpLpeQVEEQhLpwp1DoChy120627rOnH9BPKbVKKbVWKTXD2YWUUnOVUhuVUhszMjKa3CFbhVRJXhMEQXBOazuavYC+wFTgauANpVSYYyOt9eta6zFa6zHR0dFNvplt1TUpcyEIguAcdwqFY0B3u+1u1n32JAPfaK3LtdaHgH0YIeEWbKuuBYqjWRAEwSnuFAobgL5KqXillA9wFfCNQ5uvMFoCSqkojDkp0V0dsq2lIHkKgiAIznGbUNBaVwB3A4uBPcCnWutdSqm/KaUusjZbDGQppXYDS4E/a62z3NWn4nJjPgoQ85EgCIJT3Do6aq0XAYsc9j1q91kD91tfbkc0BUEQhPppbUdzi1JUZhMKoikIgiA4o4MJBZv5SDQFQRAEZ3QwoWDVFCRPQRAEwSkdTCiIo1kQBKE+OphQMJqClLkQBEFwTocTCn7eHnh6qNbuiiAIQpukgwkFWUtBEAShPjqWUJC1FARBEOqlYwmF/2/vbmPkquo4jn9/thT6YCwFpNoC5aEBWyMFCamipgFDykMsLwAVEEI0viERjEbBaAgmvjAxokaCGEBLbBApRRtfGKGQIi8oFMpjC4GgwpJCa4Qqa+2w278vztnLuO3M0u3M3u09v0/S7N47tzPn5L87vz3nzj23NeyRgplZF0WFwmBryCMFM7MuigqFna1h30vBzKyLokJhsDXM9IM8fWRm1klRobCzNeSRgplZF0WFwmBr2OsemZl1UVQo7GwNe4kLM7MuigmFiGCwNeSRgplZF8WEwq6h3UR4MTwzs26KCYXBXb6XgpnZWIoJhXfvuuZQMDPrpMBQ8PSRmVknBYVCnj7ydQpmZh0VFAp5pOAb7JiZdVRcKMw82NNHZmadFBQKafrIq6SamXVWUCjkkYJPNJuZdVRMKIxcp+CRgplZZ8WEwtFzZrB88Vxfp2Bm1kUxcylnL57L2Yvn1t0MM7NJrZiRgpmZjc2hYGZmFYeCmZlVHApmZlZxKJiZWcWhYGZmFYeCmZlVHApmZlZRRNTdhn0iaTvw93H+98OBf/SwOQeKEvtdYp+hzH6X2GfY934fExFHjHXQARcK+0PSxog4re52TLQS+11in6HMfpfYZ+hfvz19ZGZmFYeCmZlVSguFX9bdgJqU2O8S+wxl9rvEPkOf+l3UOQUzM+uutJGCmZl14VAwM7NKMaEgabmkFyS9JOnautvTD5KOkvSgpM2SnpN0dd4/R9J9kl7MXw+tu629JmmKpE2S/pi3j5W0Idf7LknT6m5jr0maLWm1pOclbZH0iUJq/fX88/2spDslHdK0eku6XdI2Sc+27dtrbZX8LPf9aUmn7s9rFxEKkqYANwHnAIuAL0paVG+r+mII+EZELAKWAlflfl4LrIuIhcC6vN00VwNb2rZ/CNwYEScAbwJfrqVV/fVT4E8RcRJwMqn/ja61pHnA14DTIuKjwBTgCzSv3r8Glo/a16m25wAL87+vAjfvzwsXEQrA6cBLEfFyRLSA3wIram5Tz0XE1oh4In//b9KbxDxSX1fmw1YCF9TTwv6QNB84D7g1bws4E1idD2linz8AfAa4DSAiWhHxFg2vdTYVmC5pKjAD2ErD6h0RDwH/HLW7U21XAHdE8ggwW9KHxvvapYTCPODVtu2BvK+xJC0ATgE2AEdGxNb80OvAkTU1q19+AnwL2J23DwPeioihvN3Eeh8LbAd+lafNbpU0k4bXOiJeA34EvEIKgx3A4zS/3tC5tj19fyslFIoiaRZwD3BNRPyr/bFIn0FuzOeQJZ0PbIuIx+tuywSbCpwK3BwRpwCDjJoqalqtAfI8+gpSKH4YmMme0yyN18/alhIKrwFHtW3Pz/saR9JBpEBYFRFr8u43RoaT+eu2utrXB2cAn5P0N9K04JmkufbZeXoBmlnvAWAgIjbk7dWkkGhyrQE+C/w1IrZHxDvAGtLPQNPrDZ1r29P3t1JC4TFgYf6EwjTSiam1Nbep5/Jc+m3Aloj4cdtDa4Er8vdXAH+Y6Lb1S0RcFxHzI2IBqa4PRMSlwIPAhfmwRvUZICJeB16VdGLedRawmQbXOnsFWCppRv55H+l3o+uddartWuDy/CmkpcCOtmmmfVbMFc2SziXNPU8Bbo+IH9TcpJ6T9CngL8AzvDu//h3SeYXfAUeTlh2/OCJGn8Q64ElaBnwzIs6XdBxp5DAH2ARcFhG76mxfr0laQjq5Pg14GbiS9Ideo2st6Qbg86RP220CvkKaQ29MvSXdCSwjLY/9BnA98Hv2Utscjj8nTaP9B7gyIjaO+7VLCQUzMxtbKdNHZmb2HjgUzMys4lAwM7OKQ8HMzCoOBTMzqzgUzCaQpGUjK7maTUYOBTMzqzgUzPZC0mWSHpX0pKRb8v0a3pZ0Y17Lf52kI/KxSyQ9kteyv7dtnfsTJN0v6SlJT0g6Pj/9rLb7IKzKFx+ZTQoOBbNRJH2EdMXsGRGxBBgGLiUtvrYxIhYD60lXmQLcAXw7Ij5Gupp8ZP8q4KaIOBn4JGlVT0ir115DurfHcaS1e8wmhaljH2JWnLOAjwOP5T/ip5MWH9sN3JWP+Q2wJt/XYHZErM/7VwJ3S3o/MC8i7gWIiP8C5Od7NCIG8vaTwALg4f53y2xsDgWzPQlYGRHX/d9O6XujjhvvGjHta/IM499Dm0Q8fWS2p3XAhZI+CNW9cY8h/b6MrMR5CfBwROwA3pT06bz/S8D6fOe7AUkX5Oc4WNKMCe2F2Tj4LxSzUSJis6TvAn+W9D7gHeAq0o1sTs+PbSOdd4C0jPEv8pv+yGqlkALiFknfz89x0QR2w2xcvEqq2Xsk6e2ImFV3O8z6ydNHZmZW8UjBzMwqHimYmVnFoWBmZhWHgpmZVRwKZmZWcSiYmVnlfxdCinv1o3+AAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "theModel.load('MODELS/TEST_MODEL')\n",
    "theModel.plot_training_history()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate the model\n",
    "This evaluation is performed by Keras itself. Other evaluation methods, such as those described in the reference paper, are preferred.\n",
    "Note that theModel.theModel is the Keras model and, so, metrics_name is a Keras attribute. The evaluation output can be changed as in other Keras models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "theNames=theModel.theModel.metrics_names\n",
    "theMetrics=theModel.evaluate(testGenerator)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Print the evaluation\n",
    "Note that accuracy is quite good, taking into account the small training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* loss : 0.53\n",
      "* acc : 0.82\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(theNames)):\n",
    "    print('* %s : %.2f' % (theNames[i],theMetrics[i]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get one batch of images and segment it\n",
    "To obtain the batch, the testGenerator is used. It provides the data and the ground truth. Then, the predict method is used to obtain the segmented patches."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "theImages,theGT=testGenerator.__getitem__(0)\n",
    "thePrediction=theModel.predict(theImages)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot the segmented data\n",
    "First row contains the original SSS patches. Second row is the ground truth. Third row contains the segmented paches.\n",
    "Note that this is multi-class output. Since there are three classes, the output is plotted as an RGB images. The blurry regions are areas where the probabilities of two classes are similar. See the reference paper for more information.\n",
    "Note that segmentation is far from perfect (though not bad at all) due to the extremely small training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 30 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "for i in range(theImages.shape[0]):\n",
    "    plt.subplot(3,theImages.shape[0],i+1)\n",
    "    plt.imshow(theImages[i,:,:,0],cmap='gray',vmin=0,vmax=1)\n",
    "    plt.axis('off')\n",
    "    plt.subplot(3,theImages.shape[0],theImages.shape[0]+i+1)\n",
    "    plt.imshow(theGT[i])\n",
    "    plt.axis('off')\n",
    "    plt.subplot(3,theImages.shape[0],2*theImages.shape[0]+i+1)\n",
    "    plt.imshow(thePrediction[i])\n",
    "    plt.axis('off')"
   ]
  }
 ],
 "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
}