emnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EMNIST digit classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test error rates of this model on EMNIST test data:\n",
    "Model was trained from scratch on EMNIST Digits training data using realtime data augmentation.\n",
    "### Dropout 0.2\n",
    "* 0.16% test error rate with ensemble of 10 CNNs at 267 epochs (dropout\\_0.2/weights/267epochs\\_weights\\_model\\_?.pkl)\n",
    "* 0.175% test error rate for best single model at 258 epochs (dropout\\_0.2/weights/258epochs\\_weights\\_model\\_7.pkl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using Theano backend.\n",
      "Using cuDNN version 5005 on context None\n",
      "Preallocating 7295/8105 Mb (0.900000) on cuda\n",
      "Mapped name None to device cuda: GeForce GTX 1070 (0000:28:00.0)\n"
     ]
    }
   ],
   "source": [
    "import keras\n",
    "import numpy as np\n",
    "\n",
    "from keras.layers import Dense, Dropout, Conv2D, MaxPooling2D\n",
    "from keras.layers import Flatten, Lambda, BatchNormalization\n",
    "from keras.models import Sequential\n",
    "from keras.optimizers import Adam as Adam\n",
    "from keras.layers.advanced_activations import LeakyReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# used to save and load training histories\n",
    "import pickle\n",
    "from collections import defaultdict\n",
    "\n",
    "import resource, sys\n",
    "\n",
    "# we would reach recursion limit when saving training history otherwise\n",
    "resource.setrlimit(resource.RLIMIT_STACK, (2**29,-1))\n",
    "sys.setrecursionlimit(2**29 - 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load EMNIST Digits dataset\n",
    "Introduced by https://arxiv.org/abs/1702.05373v1 downloaded from https://www.nist.gov/itl/iad/image-group/emnist-dataset (Matlab format dataset). The matlab format dataset can be conveniently imported with scipy.io.loadmat."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import io as spio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "emnist = spio.loadmat(\"datasets/matlab/emnist-digits.mat\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load training dataset\n",
    "x_train = emnist[\"dataset\"][0][0][0][0][0][0]\n",
    "x_train = x_train.astype(np.float32)\n",
    "\n",
    "# load training labels\n",
    "y_train = emnist[\"dataset\"][0][0][0][0][0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load test dataset\n",
    "x_test = emnist[\"dataset\"][0][0][1][0][0][0]\n",
    "x_test = x_test.astype(np.float32)\n",
    "\n",
    "# load test labels\n",
    "y_test = emnist[\"dataset\"][0][0][1][0][0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# store labels for visualization\n",
    "train_labels = y_train\n",
    "test_labels = y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(240000, 784)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(240000, 1)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# normalize\n",
    "x_train /= 255\n",
    "x_test /= 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       ..., \n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.]], dtype=float32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reshape using matlab order\n",
    "x_train = x_train.reshape(x_train.shape[0], 1, 28, 28, order=\"A\")\n",
    "x_test = x_test.reshape(x_test.shape[0], 1, 28, 28, order=\"A\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(240000, 1, 28, 28)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(240000, 1)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(240000, 10)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# labels should be onehot encoded\n",
    "y_train = keras.utils.to_categorical(y_train, 10)\n",
    "y_test = keras.utils.to_categorical(y_test, 10)\n",
    "\n",
    "y_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Verify we have imported the data correctly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# which sample to look at\n",
    "samplenum = 5437"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fe94c3c88d0>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe94e80b748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "img = x_train[samplenum]\n",
    "\n",
    "# visualize image\n",
    "plt.imshow(img[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show label for sample image\n",
    "train_labels[samplenum][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reshape test labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40000, 1)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_labels = test_labels.reshape(40000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40000,)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate mean and standard deviation\n",
    "mean_px = x_train.mean().astype(np.float32)\n",
    "std_px = x_train.std().astype(np.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to normalize input data\n",
    "def norm_input(x): return (x-mean_px)/std_px"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Batchnorm + dropout + data augmentation\n",
    "def create_model():\n",
    "    model = Sequential([\n",
    "        Lambda(norm_input, input_shape=(1,28,28), output_shape=(1,28,28)),\n",
    "        Conv2D(32, (3,3)),\n",
    "        LeakyReLU(),\n",
    "        BatchNormalization(axis=1),\n",
    "        Conv2D(32, (3,3)),\n",
    "        LeakyReLU(),\n",
    "        MaxPooling2D(),\n",
    "        BatchNormalization(axis=1),\n",
    "        Conv2D(64, (3,3)),\n",
    "        LeakyReLU(),\n",
    "        BatchNormalization(axis=1),\n",
    "        Conv2D(64, (3,3)),\n",
    "        LeakyReLU(),\n",
    "        MaxPooling2D(),\n",
    "        Flatten(),\n",
    "        BatchNormalization(),\n",
    "        Dense(512),\n",
    "        LeakyReLU(),\n",
    "        BatchNormalization(),\n",
    "        Dropout(0.2),\n",
    "        Dense(10, activation='softmax')\n",
    "    ])\n",
    "    model.compile(Adam(), loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data augmentation\n",
    "Use keras data augmentation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 512"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "gen = ImageDataGenerator(rotation_range=12, width_shift_range=0.1, shear_range=0.3,\n",
    "                        height_shift_range=0.1, zoom_range=0.1, data_format='channels_first')\n",
    "batches = gen.flow(x_train, y_train, batch_size=batch_size)\n",
    "test_batches = gen.flow(x_test, y_test, batch_size=batch_size)\n",
    "steps_per_epoch = int(np.ceil(batches.n/batch_size))\n",
    "validation_steps = int(np.ceil(test_batches.n/batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fe94c2bf3c8>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe94c3e66a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# load ONE image from training set to display on screen\n",
    "img = x_train[1]\n",
    "\n",
    "# visualize original image\n",
    "plt.imshow(img[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# trick our generator into believing img has enough dimensions\n",
    "# and get some augmented images for our single test image\n",
    "img = np.expand_dims(img, axis=0)\n",
    "aug_iter = gen.flow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 28, 28)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "aug_img = next(aug_iter)[0].astype(np.float32)\n",
    "aug_img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe94c273a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# show augmented images\n",
    "f = plt.figure(figsize=(12,6))\n",
    "for i in range(8):\n",
    "    sp = f.add_subplot(2, 26//3, i+1)\n",
    "    sp.axis('Off')\n",
    "    aug_img = next(aug_iter)[0].astype(np.float32)\n",
    "    plt.imshow(aug_img[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ensembling\n",
    "Here we create ten models and fit them to our augmented data. We use the mean prediction from all ten classifiers as our final prediction. This should average out errors in our classifiers and enable us to predict with higher accuracy."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create ten models from scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "models = []\n",
    "weights_epoch = 0\n",
    "\n",
    "for i in range(10):\n",
    "    m = create_model()\n",
    "    models.append(m)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Or Create ten models using saved weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# specify which weights to load\n",
    "weights_epoch = 267"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "models = []\n",
    "\n",
    "for i in range(10):\n",
    "    m = create_model()\n",
    "    m.load_weights('dropout_0.2/weights/{:03d}epochs_weights_model_{}.pkl'.format(weights_epoch, i))\n",
    "    models.append(m)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train models for a few more epochs\n",
    "Train models for some additional epochs and save model weights as well as corresponding training history every num_epochs. Test error rates for ensemble and every single model are written to text files after every iteration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_batch_size = 512"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# how many iterations with how many epochs we want to train\n",
    "num_iterations = 8\n",
    "num_epochs = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration 1, cur_epoch 268\n",
      "iteration 2, cur_epoch 269\n",
      "iteration 3, cur_epoch 270\n",
      "iteration 4, cur_epoch 271\n",
      "iteration 5, cur_epoch 272\n",
      "iteration 6, cur_epoch 273\n",
      "iteration 7, cur_epoch 274\n",
      "iteration 8, cur_epoch 275\n"
     ]
    }
   ],
   "source": [
    "for iteration in range(num_iterations):\n",
    "    cur_epoch = (iteration+1)*num_epochs + weights_epoch\n",
    "    print(\"iteration {}, cur_epoch {}\".format(iteration+1, cur_epoch))\n",
    "    \n",
    "    # train models for specified number of epochs\n",
    "    for i,m in enumerate(models):\n",
    "        m.optimizer.lr = 0.00001\n",
    "        h = m.fit_generator(batches, steps_per_epoch=steps_per_epoch, epochs=num_epochs, verbose=0,\n",
    "                            validation_data=test_batches, validation_steps=validation_steps)\n",
    "        \n",
    "        # save model weights\n",
    "        m.save_weights(\"dropout_0.2/weights/{:03d}epochs_weights_model_{}.pkl\".format(cur_epoch, i))\n",
    "        # save corresponding training history\n",
    "        with open(\"dropout_0.2/history/{:03d}epochs_history_model_{}.pkl\".format(cur_epoch, i),\"wb\") as f:\n",
    "            pickle.dump(h, f)\n",
    "        \n",
    "    # evaluate test error rate for ensemble\n",
    "    all_preds = np.stack([m.predict(x_test, batch_size=eval_batch_size) for m in models]) \n",
    "    avg_preds = all_preds.mean(axis=0)\n",
    "    test_error_ensemble = (1 - keras.metrics.categorical_accuracy(y_test, avg_preds).eval().mean()) * 100\n",
    "    \n",
    "    # write test error rate for ensemble and every single model to text file\n",
    "    with open(\"dropout_0.2/history/test_errors_epoch_{:03d}.txt\".format(cur_epoch), \"w\") as text_file:\n",
    "        text_file.write(\"epoch: {} test error on ensemble: {}\\n\".format(cur_epoch, test_error_ensemble))\n",
    "\n",
    "        for m in models:\n",
    "            pred = np.array(m.predict(x_test, batch_size=eval_batch_size))\n",
    "            test_err = (1 - keras.metrics.categorical_accuracy(y_test, pred).eval().mean()) * 100\n",
    "            text_file.write(\"{}\\n\".format(test_err))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate Ensemble of 10 CNNs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_batch_size = 512"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_preds = np.stack([m.predict(x_test, batch_size=eval_batch_size) for m in models])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "avg_preds = all_preds.mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.16000270843505859"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# display error rate in percent\n",
    "(1 - keras.metrics.categorical_accuracy(y_test, avg_preds).eval().mean()) * 100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Show test error rate of every model separately\n",
    "All test error rates in percent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 0.204998254776\n",
      "2 0.212502479553\n",
      "3 0.199997425079\n",
      "4 0.197499990463\n",
      "5 0.209999084473\n",
      "6 0.204998254776\n",
      "7 0.217497348785\n",
      "8 0.199997425079\n",
      "9 0.185000896454\n",
      "10 0.207501649857\n"
     ]
    }
   ],
   "source": [
    "for i, m in enumerate(models):\n",
    "    pred = np.array(m.predict(x_test, batch_size=eval_batch_size))\n",
    "    print(i+1, (1 - keras.metrics.categorical_accuracy(y_test, pred).eval().mean()) * 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot Training History"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load and plot saved training histories. Utility functions from https://github.com/fchollet/keras/issues/103\n",
    "Prepending zeros to history files with\n",
    "<pre>\n",
    "for i in `find * | grep \"^..epo\"`; do mv $i 0$i; done\n",
    "</pre>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# which model to plot training history of\n",
    "i = 9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge training histories\n",
    "def merge_dict(dict_list):\n",
    "    dd = defaultdict(list)    \n",
    "    for d in dict_list:\n",
    "        for key, value in d.items():\n",
    "            if not hasattr(value, '__iter__'):\n",
    "                value = (value,)       \n",
    "            [dd[key].append(v) for v in value]\n",
    "    return dict(dd)\n",
    "\n",
    "# load pickled data\n",
    "def load(name):\n",
    "    filename = open(name,\"rb\")\n",
    "    obj = pickle.load(filename)\n",
    "    filename.close()    \n",
    "    return(obj)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "histories = glob.glob(\"dropout_0.2/history/*model_{}*\".format(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "histories = sorted(histories)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dropout_0.2/history/016epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/032epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/048epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/064epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/080epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/096epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/112epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/128epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/144epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/160epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/176epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/192epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/208epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/224epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/240epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/256epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/272epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/288epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/304epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/320epochs_history_model_9.pkl',\n",
       " 'dropout_0.2/history/336epochs_history_model_9.pkl']"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "histories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# first history would include zero accuracy and mess up our plot\n",
    "h0 = load(histories[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge all saved histories\n",
    "for h in histories[2:]:\n",
    "    h1 = load(h)\n",
    "    h0.history = merge_dict([h0.history,h1.history])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "history = h0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "min = np.argmin(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Jason Brownlee's script to plot model training history:\n",
    "https://machinelearningmastery.com/display-deep-learning-model-training-history-in-keras/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAAEWCAYAAACqitpwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\np+AiIiJ1p+AiIiJ1p+AiIiJ19/8ButhL6yICGCoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f92539b5c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ4AAAEWCAYAAABWn/G6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXmYXFWZ/z9vV3dVr+k9e0LCKgE0QAwoMI4CCrgw6oig\niIwLMuKoo8Moo+Oov3HG0dFRFEEZUFEWEVBQw+qAgKxJiIFsZCedtTud3ru2rvP745xT99Tt6u7q\npTppcj7P009V3brLqeru873f97znPaKUwuPxeDyeyaLkYDfA4/F4PIcXXng8Ho/HM6l44fF4PB7P\npOKFx+PxeDyTihcej8fj8UwqXng8Ho/HM6l44fF4ioyI/ExE/r3AfbeJyDkTcM0FIqJEpHS85/J4\nJhovPB7PFEFE6kTk5yKyz/x89WC3yeMZC/5uyOOZOvwPUAksAKYDfxSR7Uqpnx7UVnk8o8Q7Ho+H\nbIjrahFZLSK9InKTiMwQkftFpFtEHhGRemf/d4nIGhHpEJHHROR4572TRWSlOe5XQHnoWu8QkVXm\n2KdE5LUFNvOdwLeVUn1KqW3ATcBHCvx8s0XkPhFpF5FNIvJx572lIrJcRLpEZK+IfNdsLxeRX4rI\nftPW50VkRoFt9XiGxAuPxxPwXuBc4Fh0J38/8C9AM/p/5dMAInIscDvwWfPeMuB3IhIVkSjwW+AX\nQAPwa3NezLEnAzcDnwAagR8D94lIbAztFeDEAve9A2gBZgN/C/yHiLzFvPd94PtKqWnAUcCdZvuH\ngVpgnmnrlUD/GNrp8eTghcfjCfiBUmqvUmon8ATwrFLqBaVUHPgNcLLZ7/3AH5RSDyulUsB/AxXA\nG4HTgTLge0qplFLqLuB55xpXAD9WSj2rlBpQSv0cSJjjRuIB4AsiUiMiR6PdTuVIB4nIPOAM4AtK\nqbhSahXwv8BlZpcUcLSINCmlepRSzzjbG4GjTVtXKKW6CminxzMsXng8noC9zvP+PK+rzfPZwHb7\nhlIqA+wA5pj3dqrc6rvbnedHAJ83oasOEelAO4rZBbTv00Ac2Ajci3ZdLQUcNxtoV0p1h9o0xzz/\nKNrlrTfhtHeY7b8AHgTuEJFdIvItESkr4Hoez7B44fF4Rs8utIAAICKCFo+dwG5gjtlmme883wF8\nQylV5/xUKqVuH+miSql2pdQHlVIzlVInoP9/nyuwvQ0iUhNq005z3o1KqUvQCQv/BdwlIlXGsX1N\nKbUI7ebeQeCSPJ4x44XH4xk9dwJvF5GzjQP4PDpc9hTwNJAGPi0iZSLyHmCpc+yNwJUicppoqkTk\n7SFRyIuIHCUijSISEZHz0WG7EecHKaV2mLb9p0kYeC3a5fzSnPdSEWk2zq3DHJYRkTeLyEkiEgG6\n0KG3TAHfj8czLF54PJ5RopTaAFwK/ABoQycivFMplVRKJYH3AJcD7ejxoHucY5cDHwd+CBwANpl9\nC+FU4EWgG/hP4INKqTUFHnsJOg17F3q86t+UUo+Y984D1ohIDzrR4GKlVD8wE7gLLTrrgD+hw28e\nz7gQvxCcx+PxeCYT73g8Ho/HM6l44fF4PB7PpOKFx+PxeDyTihcej8fj8Uwqh3WR0KamJrVgwYKD\n3QyPx+OZUqxYsaJNKdU81uMPa+FZsGABy5cvP9jN8Hg8nimFiGwfea+h8aE2j8fj8UwqXng8Ho/H\nM6l44fF4PB7PpHJYj/HkI5VK0dLSQjweP9hNKTrl5eXMnTuXsjJfcNjj8UweXnhCtLS0UFNTw4IF\nC8gtMPzqQinF/v37aWlpYeHChQe7OR6P5zDCh9pCxONxGhsbX9WiAyAiNDY2HhbOzuPxHFp44cnD\nq110LIfL5/R4PIcWXnjGQjoJXbsh7d2Cx+PxjBYvPGMhk4aePZBKFOX0HR0d/OhHPxr1cRdccAEd\nHR0j7+jxeDwHES88Y0HM16YGinL6oYQnnU4Pe9yyZcuoq6srSps8Ho9novBZbWMhKzzOKsDde/Tr\nabPHffovfvGLbN68mcWLF1NWVkZ5eTn19fWsX7+el19+mb/5m79hx44dxONxPvOZz3DFFVcAQQmg\nnp4ezj//fM4880yeeuop5syZw7333ktFRcW42+bxeDzjpajCIyLnoZfSjQD/q5T6Zuh9Me9fAPQB\nlyulVorIPOAWYAaggJ8opb5vjvkVcJw5RR3QoZRaLCIL0MvzbjDvPaOUunI87f/a79awdldXnncU\nJHsh0g0RMwcm1a+3lw1fwmjR7Gn82ztPGHafb37zm7z00kusWrWKxx57jLe//e289NJL2bTnm2++\nmYaGBvr7+3n961/Pe9/7XhobG3POsXHjRm6//XZuvPFGLrroIu6++24uvfTSQj+6x+PxFI2iCY+I\nRIDrgHOBFuB5EblPKbXW2e184BjzcxpwvXlMA583IlQDrBCRh5VSa5VS73eu8R2g0znfZqXU4mJ9\npgCbDTY5y4YvXbo0Z67Ntddey29+8xsAduzYwcaNGwcJz8KFC1m8WH8Vp556Ktu2bZuUtno8Hs9I\nFNPxLAU2KaW2AIjIHcCFgCs8FwK3KKUU8IyI1InILKXUbmA3gFKqW0TWAXPcY41bugh4S7E+wJDO\nRCnYvQqqZ8K0WXpb68t6zGf68RPejqqqquzzxx57jEceeYSnn36ayspK/vqv/zrvXJxYLJZ9HolE\n6O/vn/B2eTwez1goZnLBHGCH87rFbBvVPiaEdjLwbOjYs4C9SqmNzraFIrJKRP4kImeNvekjIKLH\nedwxHjJMlAOqqamhu7s773udnZ3U19dTWVnJ+vXreeaZZybkmh6PxzNZHNLJBSJSDdwNfFYpFR5s\nuQS43Xm9G5ivlNovIqcCvxWRE8LHicgVwBUA8+fPH0fjQsKj0E5oAmhsbOSMM87gxBNPpKKighkz\nZmTfO++887jhhhs4/vjjOe644zj99NMn5Joej8czWRRTeHYC85zXc822gvYRkTK06NyqlLrHPUhE\nSoH3AKfabUqpBJAwz1eIyGbgWCBnpTel1E+AnwAsWbJk7EoxyPFM7HjPbbfdlnd7LBbj/vvvz/ue\nHcdpamripZdeym7/p3/6pwltm8fj8YyHYobangeOEZGFIhIFLgbuC+1zH3CZaE4HOpVSu834zU3A\nOqXUd/Oc+xxgvVKqxW4QkWaT0ICIHIlOWNgy8R/LXrAkdx6PykyY4/F4PJ5XM0VzPEqptIh8CngQ\nnU59s1JqjYhcad6/AViGTqXehE6n/jtz+BnAh4AXRWSV2fYvSqll5vnF5IbZAP4K+LqIpNADLlcq\npdqL8+nI43hgsrLcPB6PZypT1DEeIxTLQttucJ4r4Ko8xz1JkLOc77yX59l2Nzo0V3TiqQFIK8oi\nGSJBA7zj8Xg8ngLwJXPGgFKKZAZUONTm8Xg8nhHxwjMGSkTIUIIMcjje8Xg8Hs9IHNLp1IcqJSVC\nBgmlUyu88Hg8Hs/IeMczBrKOh+JMIB3rsggA3/ve9+jr65uQdng8Hk8x8MIzBkoElOt43JDbBCQY\neOHxeDyvZnyobQyICIoSSlBFCbG5yyKce+65TJ8+nTvvvJNEIsG73/1uvva1r9Hb28tFF11ES0sL\nAwMD/Ou//it79+5l165dvPnNb6apqYlHH310Qtvl8Xg8E4EXnuG4/4uw58W8b9Ul40AKotV6Q7JH\nP0arGSYTHGaeBOd/c+j3yV0W4aGHHuKuu+7iueeeQynFu971Lh5//HFaW1uZPXs2f/jDHwBdw622\ntpbvfve7PProozQ1NY3yw3o8Hs/k4ENt46a4CQUPPfQQDz30ECeffDKnnHIK69evZ+PGjZx00kk8\n/PDDfOELX+CJJ56gtra2qO3weDyeicI7nuEYxpm0797JTLWPeP1rKCuNEGldo9+YeRKUTNzXqpTi\nmmuu4ROf+MSg91auXMmyZcv48pe/zNlnn81XvvKVCbuux+PxFAvveMaKWf56R3sPbT2JYPsEGCB3\nWYS3ve1t3HzzzfT06FDezp072bdvH7t27aKyspJLL72Uq6++mpUrVw461uPxeA5FvOMZKyUlJoM6\nQzKddt4Yv/K4yyKcf/75fOADH+ANb3gDANXV1fzyl79k06ZNXH311ZSUlFBWVsb1118PwBVXXMF5\n553H7NmzfXKBx+M5JBF1GNcXW7JkiVq+PGfVBNatW8fxx4+8iuie1lZmplrYnJmFRMo4Ur2i35h+\nApRGi9HcolDo5/V4PB6LiKxQSi0Z6/E+1DZGxITaSsiQHnBqto3F8ST7fK03j8dz2OCFZ6zopX8o\n0VNJHUYpPAMpaNsA8c6JapnH4/Ec0njhyUMh4UcpsY5HIa7YjNbw2ArXmfTw+xWBwznM6vF4Dh5e\neEKUl5ezf//+ETvlrPBIJld4Rqs89jqTrAFKKfbv3095efnkXtjj8Rz2+Ky2EHPnzqWlpYXW1tZh\n9+uOp2iN76ObHtISJak69BvtJRAZRXLBQBK690FFCmJt42j56CkvL2fu3LmTek2Px+PxwhOirKyM\nhQsXjrjfrc9u57wHLuAlOZ23vvej1N19sX7j4/8Hc15X+AVfeRbuugjO/Tos/swYW+3xeDxTh6KG\n2kTkPBHZICKbROSLed4XEbnWvL9aRE4x2+eJyKMislZE1ojIZ5xjvioiO0Vklfm5wHnvGnOuDSLy\ntmJ+tqpoKW2qlpmRbuqiTnpBZmDog/IxYCafDqQmrnEej8dzCFM0xyMiEeA64FygBXheRO5TSq11\ndjsfOMb8nAZcbx7TwOeVUitFpAZYISIPO8f+j1Lqv0PXWwRcDJwAzAYeEZFjVc761BNHZTRCm6ql\nWTohE4hGOp0c3Zc6kNSPoxUsj8fjmaIU0/EsBTYppbYopZLAHcCFoX0uBG5RmmeAOhGZpZTarZRa\nCaCU6gbWAXNGuN6FwB1KqYRSaiuwybShKFTFSmmjlgY6A/EA4vHkMEflIW2FZ/Kz2jwej+dgUEzh\nmQPscF63MFg8RtxHRBYAJwPPOpv/wYTmbhaR+lFcDxG5QkSWi8jykRIIhsM6ntrMARgIRKMvkRjm\nqDzYUFvGh9o8Hs/hwSGdTi0i1cDdwGeVUl1m8/XAkcBiYDfwndGcUyn1E6XUEqXUkubm5jG3rdKM\n8ZRn+nMmf3rH4/F4PMNTTOHZCcxzXs812wraR0TK0KJzq1LqHruDUmqvUmpAKZUBbiQIpxVyvQmj\nMhqhjWn6Rfeu7PZ4coyOZ8ALj8fjOTwopvA8DxwjIgtFJIoe+L8vtM99wGUmu+10oFMptVtEBLgJ\nWKeU+q57gIjMcl6+G3jJOdfFIhITkYXohIXnJv5jaZqqYyRjZpXPzkDfRi883vF4PJ7Di6IJj1Iq\nDXwKeBCdHHCnUmqNiFwpIlea3ZYBW9CJADcCnzTbzwA+BLwlT9r0t0TkRRFZDbwZ+EdzvTXAncBa\n4AHgqmJltAFURCN8/2Nv1S+6AseTSIw11ObHeDwez+FBUSeQKqWWocXF3XaD81wBV+U57kkI1d4M\n3vvQMNf7BvCNsbZ31FRN149dgeNZv+sAv/r1X/jWe19LSUnej5BLNrnAOx6Px3N4cEgnFxzyVJnk\nBEd4Vm3fz10rWmjtKTDkZh2PH+PxeDyHCV54xkNpFGK1OfN4SpQWkJ0d/YWdwzsej8dzmOGFZ7xU\n1OW8LEUPK+0qWHj8GI/H4zm88MIzXirqc15G0CuJFiw8aV8yx+PxHF544RkvQzqeeGHH+yKhHo/n\nMMMLz3ixjicS0w9GeAoe4/GVCzwez2GGF57xUm4cT1kFAKVkmN9QOYoxHl+rzePxHF544Rkv1vEY\n4YkwwNKFDWNILnDGeDY+Ak/9cAIb6fF4PIcOXnjGixUeIxxza8tY2FTFgb4UfckCwmfZeTyO41n9\nK3jmRxPcUI/H4zk08MIzXmxyQaofpIQPLJnD3HrtfloOFOB68s3jGUj4ZAOPx/OqxQvPeLGOJ9kD\nJaWQSXNcYynnlKxg3a6OkY/PV6stnfBjPh6P51WLF57xYpMLUFnhObrtj/xv9DvMeubfRz4+63ic\nMZ50wpfQ8Xg8r1q88IwXdwJpSSlkBig1ZXOW7r0DWl8e/vh86dTphE+v9ng8r1q88IyXHOGJaMFI\nBZNHVev64Y8fyJNckI77UJvH43nV4oVnvLiVC0yojXQgPJ0dB8hk1NDH5wu1DST1edQwx3k8Hs8U\nxQvPeCmrDJ7nEZ5tu/Zw1rce5Zant+U/Pm9ygTneh9s8Hs+rEC8840Wcxd7MGA+pII36z2u2srOj\nn5XbD+Q/Pl86tS+j4/F4XsV44ZkoZp4UjPGkExCtJiVRIqluALbt78t/XL4JpNbxFDKXRylf2drj\n8Uwpiio8InKeiGwQkU0i8sU874uIXGveXy0ip5jt80TkURFZKyJrROQzzjHfFpH1Zv/fiEid2b5A\nRPpFZJX5uSF8vaLxpT3w8UedUFs/lJZDrIYa+ikReKV9COHJVzInPYrF4Tb9Eb55BMQ7x/cZPB6P\nZ5IomvCISAS4DjgfWARcIiKLQrudDxxjfq4Arjfb08DnlVKLgNOBq5xjHwZOVEq9FngZuMY532al\n1GLzc2UxPldeyiogUmaEJ6WFo7Sc0spa/uqIGJ9401G09ybpjoccjFL5i4SOZqmE9s2Q7Ibeton5\nLB6Px1Nkiul4lgKblFJblFJJ4A7gwtA+FwK3KM0zQJ2IzFJK7VZKrQRQSnUD64A55vVDSilrBZ4B\n5hbxM4wOd4ynrByJ1TCvcoDXzqkFYHs43OYKi3U3So0uuSBlzpnsHWfjPR6PZ3IopvDMAXY4r1vM\ntlHtIyILgJOBZ/Nc4yPA/c7rhSbM9icROStfo0TkChFZLiLLW1tbC/kchZMd44lDaQXEpkGim/mN\nOvPtlfY+UgOZoHiodTZllUH6dI4YFeB4bCJDaohQnsfj8RxiHNLJBSJSDdwNfFYp1RV670vokNyt\nZtNuYL5SajHwOeA2EZkWPqdS6idKqSVKqSXNzc0T22A3nbo0ZoSniyMaqwDYtr+Xbz2wnvf86Cm9\nv00siOr3yQzkpGIXVDbHCo4XHo/HM0UoLeK5dwLznNdzzbaC9hGRMrTo3KqUusc9SEQuB94BnK2U\nnmWplEoACfN8hYhsBo4Flk/Q5xmZrPAk9bhPrAYSXVTHSmmsirK9rY8Nu9rYva+dgYwiYhML7Fwg\nOz5kGY3jSXrh8Xg8U4NiOp7ngWNEZKGIRIGLgftC+9wHXGay204HOpVSu0VEgJuAdUqp77oHiMh5\nwD8D71JK9Tnbm01CAyJyJDphYUuxPlxe7BhP1vHUQEKnUy+aPY3nt7dzfuvPuKP0q7T1JIJQW9bx\npINt9vVI+FCbx+OZYhRNeEwCwKeAB9HJAXcqpdaIyJUiYjPOlqHFYRNwI/BJs/0M4EPAW5z06AvM\nez8EaoCHQ2nTfwWsFpFVwF3AlUqp9mJ9vrzkhNp0OjXxLvjVpbyvdh1bWnuZQStzpI09nfEg1GYd\nz0DI8RSS1eZDbR6PZ4pRzFAbSqllaHFxt93gPFfAVXmOexKQ8Hbz3tFDbL8bHZo7eJSUagFI9etQ\nW/k0UAOw7ne8k9/zaW6lnBRVxNnT2c/rysKOZyAUahuF4/GhNo/HM0U4pJMLphxZx5MIQm2WabOJ\nlZZQLkkiomg90OEkF1Trx0wqlFwwGsfj06k9Hs/UwAvPRJJTucCkUxtk2mzOOLqJ6eW64vSBjgNB\n1YKoTS5Ij93xpApYZtvj8XgOAbzwTCQlETOBND7Y8VTU84NLTua4xggAnR0dufN4QDucAZ/V5vF4\nXt0UdYznsKOk1CQIxIN0akuyj6pYaVZYeroPQNp8/UON8YxqHo8PtXk8nqmBdzwTSUmpcSDKZLU5\n81eTPfrROJTe7q786dSjnscTzzmvx+PxHOp44ZlISkoDgbHp1Ba73SQPJHo7SSWNWJhQWyKZIOMs\nmz26Wm0+1ObxeKYGXngmkpJIIDBl5VA7F45/F8w5NSjiaZxJJN3HDf+3Xm8zWW3v+9ET/OrpjcH5\nCgq12eQCH2rzeDxTAy88E0mkLHAppWaphPf/AuadBoncUNsHT25kZ5tZQ8dktZUywJodTuHSkUJt\nmYEgXOdDbR6PZ4rghWciqZ4RPC+NBc+jVdoJOULx2ukRohhhKQuEZ2Gdk+8x0jweV2x8qM3j8UwR\nvPBMJNNmB8/LKoLn0WpAQX9HdlONJKg32pSK6H1LZZSVC1zh8aE2j8czRfDCM5G4whN2PAC9QRhN\nkj3MrdFf/7ZuXR2oJgpqNMkFNrGgpNQ7Ho/HM2XwwjORTHMWQy11HI/NbnOEh2Qvs2v0ZNI1rTqk\ntnhWNWo0RUKt46ls9GM8Ho9nyuCFZyLJCbWVB8/zOB6SvcysEhKqlPvXtgEwvy5KSSaJKjHjPCMl\nF1jHU9nkQ20ej2fK4IVnIil3JoyWusJjioD27Q+2JbuZWxMhLaVsadc126qjihhJVLZo6BChtr52\n/ZN1PA3BAnQej8dziOOFp1jkE56Q44lJmpKycgbMr6GqDKKk6UyX6X2Gmsdz/RnwrYWB46lq0o9+\nTR6PxzMF8MJTLNystpgVHh1SQyJ6Xk86QXmsgncsng9AZakiJik6U6WkVQlqqDGe7l36cdMf9WOl\nFx6PxzN18MJTLErzjfHs049VTbqSwUAKKY3yubct0ptLIUaKuColTYS+eJy8VE3Xj89erx8rG/Wj\nz2zzeDxTgKIKj4icJyIbRGSTiHwxz/siItea91eLyClm+zwReVRE1orIGhH5jHNMg4g8LCIbzWO9\n89415lwbRORtxfxsQ9J8vH7MG2ozjqeySU8oHTALxkV0aK0ykiFKigRlpChle2snF/34aXoToZBb\neOzHh9o8Hs8UomjCIyIR4DrgfGARcImILArtdj5wjPm5AjC38KSBzyulFgGnA1c5x34R+KNS6hjg\nj+Y15v2LgROA84AfmTZMLh++Dy66JVjcDQYLT5URnnQSIlE9DweoKFXESJFEO54Nuw7w3NZ21u/p\nyr1GohsWnBW8to7HC4/H45kCFNPxLAU2KaW2KKWSwB3AhaF9LgRuUZpngDoRmaWU2q2UWgmglOoG\n1gFznGN+bp7/HPgbZ/sdSqmEUmorsMm0YXKpng6LQh+zNKoFxiYXZENtiVzhiShqpI9uVUmaEvr6\ndahtS6uTKp1O6DTrI98UbMuG2nxKtcfjOfQppvDMAXY4r1sIxKPgfURkAXAy8KzZNEMptds83wPY\nAmmFXO/gEa2GuCmZU9Ws3YldqdQIT0QNUCd9dFJFmlJKGQBg235HUBLd+rG8Dma9Tj+vqNOPfhKp\nx+OZAhzSK5CKSDVwN/BZpVRX+H2llBIRNcpzXoEO6zF//vwJaWdBRKuhv10/t1lo/Qe0+zFjPGRS\n1EkvnaoKSkopNWM5W9tc4ekKzvd3D0DXTkCX3PGhNo/HMxUopuPZCcxzXs812wraR0TK0KJzq1Lq\nHmefvSIyy+wzC9g3iuuhlPqJUmqJUmpJc3PzqD/UmHGrGlQ26Mf+9pxQGwMpquklHZ2GRMooJcNr\nyvbw4a3/DD0mTGeXV4jV6HGkpmOC1G0favN4PFOAYgrP88AxIrJQRKLogf/7QvvcB1xmsttOBzqV\nUrtFRICbgHVKqe/mOebD5vmHgXud7ReLSExEFqITFp6b+I81RpqP1Y+lFUHttr52E2qLgJRAbxsl\nKMqqGigpLaOcJA9EPsdp6eVkWlboY2yozV3d1CYyDBVqU6MyhR6Px1NUiiY8Sqk08CngQXRywJ1K\nqTUicqWIXGl2WwZsQScC3Ah80mw/A/gQ8BYRWWV+LjDvfRM4V0Q2AueY1yil1gB3AmuBB4CrlFID\nxfp8o6bpOP2Y7g+y3GxyAegEgf2bAHjX6YuoLC/nbZHl2cPXbt/JFbcsZ/UWM4xlJ6UClJl5Qvnq\ntT1/E3ytDu69auSiox6PxzMJFHWMRym1DC0u7rYbnOcKuCrPcU+SHbgY9N5+4Owh3vsG8I1xNLl4\nNL8meG4nlEKwfEL1DGh7GYD6xulQUQ7B8j3c9vgaHhqYyfTt63gtQMypCxcp09UQ8k0gbTXLa7/w\nSzjqLXDieyfk43g8Hs9Y8ZULJgsbaoNct2IdT/V0kygAVNQHCQeGi06s4cbLlpCJm1Bb1DmHiBaz\nfKG2ZC/UzIb6hfDsjyfgg3g8Hs/4KEh4ROQzIjLNjMXcJCIrReStxW7cqwp3rZ5oPuFxls0urwsS\nDszzxc0lnLtoBq+xdRrcMR7Qy2fnC7UlunXV7KUfhx3PQtvGcX8Uj8fjGQ+FOp6PmHTmtwL16PGX\nbxatVa9GSpyv2hWebKhterCtwhGeijoor4V4JwBH1GTIKOHZljjfffhlVmw3KdplFflDbcle7YZm\nnKhfd+8prL2bH81ZqntI4p1+7Mjj8YyKQoXHjrdcAPzCDOTnHYPxDMOHfgMfvCt3jGcox2NDbeV1\nejzHCM/cijQ9lHPlrSu59o8bueTGZ8lk1DChth4tdGUm8y09ROFRl/4O+MW7YdVtI+97w1nw5++N\nvJ/H4/EYCk0uWCEiDwELgWtEpAbIFK9Zr1KOeot+dNfZcZMLQAtRWQWUGOGxVQnMxNHmWIoeKjjQ\nl2JeQwU72vvpiqeoGyrUluzVE1btXJ9CJpn2tgEqqLQwFJkB6NgOB7aNfE6Px+MxFOp4Poouxvl6\npVQfUAb8XdFa9WonUhpUr3aTC0A7HJHcMZ7ywPHU0E8flURKhA+/YQEA7b1J0pFytuxqza1yAHqM\nJ1btCE8BZXVshYWRJqTaOUWmbR6Px1MIhQrPG4ANSqkOEbkU+DLge5vxYMNt4VCbdTgRO8ZTb8Z4\ntOORZDeRihrOP3Emx8zQCQYH+pJs64J4Xw93PPdK7nXsGM+ohOdAcOxwWMHxwuPxeEZBocJzPdAn\nIq8DPg9sBm4pWqsOB2yCQTi5oMKkrbmhNie5gEQPC2fP5AeXnExDpRat1u4kW7sUFcRJZ0JVCrJj\nPKMQnj7jeEYKy9m6cQdDeFb+An732cm/rsfjGTeFCk/aTPa8EPihUuo6oGaEYzzDYYXHOp7yOv28\n3DiegUQcnQulAAAgAElEQVSwPeYKTzfEahAR6qu0OD28di/7ExEqJUHLAUcsBtI6mcBNLihkjKfQ\nUNtoHE+yVy/pMFG8/ACs/pUvB+TxTEEKFZ5uEbkGnUb9BxEpQY/zeMaKDbVZxyMCdUdAzUz92nb6\n1vGkerWQGOEBaKjSovXCKwfoJ0aVpDh5953QvsWcwxYUrTYCJ4VltfXlEZ5kH3Ttzt0vPgrH88u/\nhQf/ZfD2ZB/8YAlsfWLkc7j07NUiOlIChMfjOeQoVHjeDyTQ83n2oCs/f7torTociIUcD8AHfw1n\n/5t+bjt9m1wAOrSVDISnoixCrLSELW299BOjml6u7Psx/OWO3HNEq7SwlVWSihdQwTqf43niv+F/\nQ5WK3FDbSM7jwDbobBm8vXcf7N8Iu1aO3C6Xnr36sXNQAXKPx3OIU5DwGLG5FagVkXcAcaWUH+MZ\nD+HkAoCGhVAVWk3UOh7Qd/eO4xGRrOuJxIK5Qcnu/eaJcTwmrDdQWs6vn36ZP29qG75t+cZ4WjdA\n9+5cgbFOR2WCaw1Fsid/6C5lHJhNaCgEpYJlIrq88Hg8U41CS+ZchF5i4H3ARcCzIvK3xWzYq56o\nGSKzobYw7kqjtiBo9x7dyTuVD+pNgkGsIhhy6+8ywhISnpTEiJJkdcsIobGs43HEpHOHvrYrRnFn\nbb7hwm1K6XPlS2xIm22jEZ5Ed3CcFx6PZ8pRaKjtS+g5PB9WSl0GLAX+tXjNOgzI53hc8jkeG1Zy\n6rRZx1NeHWxL9RjHk3DGeIBkSYxyEhyx4aZgHCgffXnSqW2YLOGIkTu+MpzwpPqMaOURnrE4nt5W\np11eeDyeqUahwlOilNrnvN4/imM9+YiF0qnDZIWnPhjj6TKdvyM89UZ4qquCbZm+A7y07Mfs/dP/\n6g1G5OLEmCEHuGD3dfDiXUO3Let4jLtJ9UNfKHwHwRgPDC88VqzyZdRZ52LDe4Vgx3cAunYVdkxn\niy7v07135H09Hk9RKVQ8HhCRB0XkchG5HPgDoXV2PKNkJMez9ONmvxrH8QwWnoZKnVxYM602u036\nD1Dx/A+ZsV0v+Kqi1fQm0vRlypiJcRY2lJeP7BhPL2Qyua7CFR5XbIYTnuRwwmNSrAspSGqxwhOt\nCcR4JPauhT2roW1D4dfxePLR1w7fmA3bnjzYLZmyFJpccDXwE+C15ucnSqkvFLNhr3rC83jCnPt1\n+MoBXdW6okFv6zBVCRzhqTNjPHW1gfCUpTppyAQOYtmGbpb8+yN0pEqZLlp4VGKIZIBUv3YhVuzS\n/Xp8x5ITauvSdeBgBMfTHZw73/VgdKE2m1gwe3HhoTY7L2ogWfh1PJ589O3XN2XtWw92S6YsBYfL\nlFJ3K6U+Z35+U8xGHRaEKxeEEQmWUojV6NptVnic5ILGai08jfV12W3TVC/1EgjEr188QH9qgH3x\nEmKiC5TGezvh0f+Ap38EgFKK3kQ6cDu18/Vjsjc3DfrANvj95/TYTKIL6ubp7f0dcGC7Xk4BtJN5\n+N+0OA3reMYwxtOzF6QEZp6kkwsKmURqnVV6EoWndYOf4PpqxC4DkvHLgYyVYYVHRLpFpCvPT7eI\ndA13rDn+PBHZICKbROSLed4XEbnWvL9aRE5x3rtZRPaJyEuhY34lIqvMzzYRWWW2LxCRfue9G8LX\nO6Q49jw465+g4ciR9xXRYz15HM8FJ83iy28/njnNOg07FaunRHI7u8e36w4/TuCu4r0dsPZe2KAj\npvf9ZRdLv/EIPR0mjGUFJdmbmzm2/vew/CbY/RftcmrNfvFOeOoHcNdH9OuW5/VyCVsfD1xSJj14\n7R7reFKjqGzQuw+qmqF2rhau8CTSVbfDK8/kbrPnHpjA6gnDsX8zXLcUtv5pcq7nmTwyprq8X4dq\nzAy7LIJSasxlcUQkAlwHnAu0AM+LyH1KqbXObucDx5if09A14U4z7/0M+CGhmnBKqfc71/gOucVK\nNyulFo+1zZNKdTOcPYrEwIqGIJvLplcDTdUxPnbWkWaQXSiZfzpsvD/n0IwqoaEqSn8icFfpvi5I\ndmbX/XnhlQ56kwO07m6hGvRS2WAczw6QCKiBQPz62kyorUE7sHin3mbdjV1wLtkLJZGgMak+iARh\nwZxKCv0dUOOsSzQUPfuganpQ5aF7T1DjDuC3V+rHrzp/GvY6k+V4rHPsaR1+P8/UwwvPuClmZtpS\nYJNSaotSKgncga715nIhcIvSPAPUicgsAKXU48CQqU4iIug5RbcXpfWHGpUNwfNY9eD3p82Gq54j\ncuplg976xF8dyXknzqTfcTzpeDfJngN0HGiD1g3M3nYPAF1tJqzWYIQn1adDbdaZHdiuH3vbtNiU\n1wZFTOOdegxlIK3FAbSTcRMZwqukuuM+hYbb4p06zbzaCs/u/Pu5YS47tjNZjsdeL11AUdYJuV4K\n7v4YtL48Odc7nMkM6Ec/Xjhmiik8cwBnVJoWs220+wzFWcBepdRGZ9tCE2b7k4icle8gEblCRJaL\nyPLW1il0N2oTDEqctXzCNB+bK1CGay44nsXz6nJCbdJ/gKiKI4lu1B0f5Ir2/6aZA/S3m04863h6\ntPA0H2deGxHp2qU78azwdASZaen+IPMs2ZubCRce58lxPAUKT6pfV9vOOh4nRdoVG3dsKh1KLvjT\nt+Hlhwq73liw10kVUBtvIujaBS/+2of2JgM7tuMdz5iZynNxLiHX7ewG5ptQ2+eA20RkWvggpdRP\nlFJLlFJLmpubJ6mpE0ClCSXFavSYz1C4yyp89kW48s+AHgtavHBWdreGAV3doFr1Ejc3cGdHXmCg\ney+UVelQIJhQ206oXwClFcF1Wteb6zVAZaMOLdmxllRIeNxMuHBm21gcTzquxbfGcTzxTvivhbDx\n4WC/3aucY0LJBc9cBy/dXdj1xsKkOx5zvZEqinvGjw21+eSCMVNM4dkJzHNezzXbRrvPIESkFHgP\n8Cu7TSmVUErtN89XoNcMOnZMLT8UsY4nOsKwmxWemllQNx9mnghAdayU04+bm90tKlptIqLYl9Dj\nPOeUrEB6W+mNNpAoMSLTsUN3nrVzc0N8tlOvnaddVt/+wPEke4dxPKGO2HU8O5frEJ5Sw2eDWccT\nrdJLRnTv0aG9/nZ9DsuuF4LnA05ygVI6/DdSfbnxMNmOxwvP5JEd4/GhtrFSTOF5HjhGRBaKSBS4\nGLgvtM99wGUmu+10oFMpNUTAPodzgPVKqWwsRUSaTUIDInIkOmFhmLowUwwbQouNIDx2PR/rBlzK\nKgZvA6r69Nd4VmQNsZ4drOuu4LEtugPbsdEKzNycNG4ObAu2VzaZZAMzmJ/qD8Z4kqExnlSoY0zF\nAzF94jtw71U6M+4/Zg/9Ga3jsZ+ze3fgaNzxnt2rnWMcx5OO685jPMKzc2UQ68/bRns973hG5MC2\nqRW2yo7xTKE2H2IUTXiUUmngU8CDwDrgTqXUGhG5UkRM2hHL0OKwCbgR+KQ9XkRuB54GjhORFhH5\nqHP6ixmcVPBXwGqTXn0XcKVSahR1WA5xrOPJl1jgUhrVApEvO2wI4WmSLrpVBTGSvE4206pq2dyp\nHcfezX/RO02bk//atXNMqG2/znoDLTxuVlvY8exbD4/+p3Ye6ThUNQXvv/I0rLln+AXrUvHgs9TM\n1NfKCs+eYD+3pI87xmOLmw41iXYk1v0ebnwzrLp16H1spzRZjseGEIvp4opBsheuO00v6jdVGPBj\nPONl2HTq8aKUWkaotI5S6gbnuQKuGuLYS4Y57+V5tt0NFDFof5CpcMZ4RuKYt8KCMwdvt6uQ5qFv\n7lnU7HyIUsnQquqywrPQRD5TNXMoi4aEp7xOt8cVDoBEZ1DbLdWrO8VITIe5Un2w+g548n/g9R/T\nwlNWCVc8BqvvhGeuD86jVP7xrHR/ruN55ekglGYdTySaG9ZzQ23WgY3VHaw26x0NV3Zo4DByPHd9\nRP99vv07oz820a3/Bnr2jbzvoYIPtY2bqZxccHhRaKgN4H0/hdd/dPB26xJitYPemnHsUtIxLW7x\nWAPb2xMkiNIo3SRUGRu6YkGoTcyfjZ08WtmYe7KOHYAZo7GOp9o4sFR/MBeoa6cZrymH2SfDEWcE\nx9l9w2QG9D982PHYfa3jqWzKnZDqhtqsEwq7g84WWHWbrk/n0rNPV2To2afFcIvJHFNmv999VleB\ncJlsx3Mwhadlua6FNxayv5dJSnOfCLLJBemD244pjBeeqUI2uWCEUNtwWJdQO3fwe5X1lM7Rc28b\nZ8xj9c5OtmWmA9BFJS/u6gpCbXVH5J4nLDwHnBpWdozHZsml+o0woYUnHQ+y5WackHueZI8+9r5P\nBxMybTJC1vHM0p2uTWawk2wrG3PdRjqP43Edy86V8D8nwG//HnauyG3H7z4Dv/gbuOFMnVRhhcse\n/8ozsOO53GMO2hhPEUJtvfuDm4UwSmlBHut1J7uixETg5/GMGy88U4Ws4xmUIV44NtSWT3gqGmDW\n6wAor59NMp3hX1LaNcUkzZMb2+gTIxBNx+aeZ5DwbNOPVdP1hNFkj34OWoiyjmdX4HhAC5qbtZfo\n1qV3Vv4cNj2it1kHYR2PFeSe0HIHVY25d9HZ9OZk/lDbjmed6zoVDzIDQRXinr26FI7FjhXZEKKL\nm9X2f/8euKRiUUzH8/BX4LaL879nF+Ubs/DYihJTSXjsGI8XnrHihWeqUNGg5+bkmSBaMLaztoLh\nhtwqG2HuEgCqZujJoy9HT2Dgfbfwg+n/jz+8uJvfrDEdctMxuecJj/HYiZv1RxjX0hPs038Aekw4\nLOt4jPCUmMKflmRP0LnvXaMfrYOwx1jRClfHrmzMDXPZDm7AEZ5MKujwXOFwkw72vKgdzmxTRtBd\n/8c6n/QwwpPuhz9/X49fFZNsaG+YpIzRsOuF4Hvqahl6+Qk3e3EsZJ1hkUOSyT64/wu5q+YWwp+/\nD8/+OHdbdozHh9rGiheeqUJpFC7/Q/6xm0KZNlsP8s85Vb+uagpcUGUDHPd2+MiD1C/UnezJR9QT\nOeFCLrv4Ei49fT5tKVP5oPk1+nEox9O1O7ieDbWV12qx2L/J2c84HrcSw4U/hPO/rZ8negJB2WfG\nEMKOxx4bXs+nsikUanNK5uSU8DEdpjue5Hai2/UEXI47P2gz6Cy/rPDEB5cCssJjywh1F7hg3Vix\nHfhEOJ6BFNz0Nlj+U/26b7/5HHk6Wus0x5ohODBJYzw7V8CzNwwuHjsSD38F7v/n3G0+uWDceOGZ\nSsw/LbcY5mipng7XtMCxb9OvbbkbMI6qBOafzoLGKkpLhNMWanc1r6GSi5bMo1fpTv6SR6u5oeYq\nPrtqDptbe/TSDm6IzDqamtm600r3Q/V0kiUV9LSYYuMlZboTT8cD1wLQeBTMMe4i6QrPOv0Ydjyl\n+RyPqeY9kAwSBdwioW6adTbs5ghHjvA8pcsH2ZJB9s6/ZlZw95zP8diO1BYJLXSl1LEykWM8yV4t\nCLaShF0KPVwFHALhsYsGjpbs72UUjmfNb2HlL0Z3HXtjMZoxt6EmMQ944RkvXngON+w8H9CFNq3w\nOCG82soyfnvVGXz0zIXZbcfOqOFBdTrfHXg/y9ujPD7tXdy7tp17V+0Kjo/VakHJpCFWS7qsKpjb\nUzOLroFSqnu26dezT9YhuVR/bikeCNqX6A4EpXOHfh52PPbR7RSjVYGYZUNseZILIHetIBt6TDrv\nH9gG048Psgk7d2qRragbwfGY0FevCUV1FTIvehzY643F8SS69dwqS3jcxS6Fnm95cjcNOjw5GLRA\nDPfZx5LVtuJn8NS1Whjt2N9I2BuDfJmSQ9E7RC1Hn9U2brzwHI6UxrRAlNfqZIXSikGTS0+cU0t5\nWbCcQXlZhFjzQq5NXcgbjmrmto+fzrz6Sra0mo67qgkqaoPQXUUte+PONLHqGfQrsyyDRPR4Uji5\nwGKz51zHA9r1FOJ4olWBmIWXQxhIDR1qq6jVRVjdzjvRrUXHJnV07QxeJ7p18kEmPfQYj+2cEp1B\nOKp9i67SMJGLxGWrbydHv/TDsz+GG98SOBbXHaTiwWfrzyc8TlJHONzW3wH3fQpeGMadjCW5IB3X\nhWGfvwl++beFhfnsdUYzBmYrsYc5mKG2zp26rNQUxwvP4YiIdiiVjVp8CkxYOGG2dgRvXaTn5BzV\nXMXmVtNJ1x0B0+YGAlZex+4+58+rZhZxdE04ZizSdeQGEvpnKMeT7NWuwi4P3rphGMcTFh4jcuFQ\nTjrkeLLLcvfq4qjRqlzhSfbo9ljH07PPCE+NDrXZDjOTyp3Jnq9TspNbf/tJ+OPXoW2j/jzXnzH+\nrDc3HTmf87Bs+/NgYerZp4/Juj8rPIlcsRnJ8YTdVjY5YZgwo23LaEJt6bgW8raNgMoNnQ7FWBxP\nx0jCcxAqF/z6cp0kMcXxwnO48v5b4czPwdHnwPHvLOiQU+bXEY2UcK4RniObq9na1kMmo/Ss9Yt+\n7qQ517Gj1xWemRyT0fN7Uqd8JJgLBIMdTzbUZhxPw1H6dfeePI7HCExYeGw7bEczEEouKDFuzO1s\no5X62u4ddKJHO7DsxF0F5dP0T6Irt8N0O958wmM7YHvttpf1eNjel/SKruPB7QSHCrd1vAI/uwDW\n/jZ3u/0O7Hdov+NUf67YjOR43BCl246h1kuCsTkee/Nhi8AWkqmWXel2FI7HCk+44sfBFJ7efUH4\ndgrjhedwZd7rdZ2106+E8/+roEMuWTqfR6/+a2ZM053+Uc3VxFMZdnX2a9dUPZ2uAe1qBmK1bDP9\nQR/lEKsho3T5m/aj3g2zXhucOOx4IqV6W9KM8VQ16eSHnj15strMo5vVFq12HE9oDMHO47GLyLmh\ntjIjPLYjTie1UEVrcitGxGr0eFA6HiqAmmfCqoubEQdacLKFVceZBu0KnSs8Pa3BOIgdawmPXdjP\nsHM5fHN+MN6TjhfgePYGHXNY8LKr0RYyxjNKxwNauN32D0eqP/dxOPZv1mFHG2orCVUWK0ao7bdX\nwYt3jbxfsm/weOIUxAuPp2BKIyXMqQtE4qjmKgCuuedFPn37C9zx3CvsMH3ArngsKDSq6uhPZXhH\n8hu8J/FVOlKlQecLgx0PaJdhHU95rSmNs3foeTzu2ig5YzxO2AiM4+mCaWZtIutukr3BUguJLrjt\n/bBhWdCWsirA1I2LTQuEyI23u+KR727YplTbtrrCM940aFfo3HMtvwluvUhnYvWZtsa7tHvMFks1\nv7TtT+n2ZOdMxYd2PErBb/5e72sdaXisJRtqm2DHkxUpM0ZWUKitQOHp3gM/fL0uVGsdT7ht4fV4\n1t4L65cxIk/9QFfhyMdLd8PWx0c+R6pvalYgD+GFxzNmjmzWIbEnNrbx501tfPGeF+lM67vDF1oV\ne/p1csJeVc+mfT2sVQtYqY6loy+px5nsctphxwOB84h3aeGpnjG848k51s1qC5VksckFNUZ4bHgo\n63iqdCbbyw8EwhOt1qnmVmxiNTrUBrnuISfUFuqsyioDx+NOis3neNb/Ae7/4uDPNRxDhdps1fBU\nX9DWRBfcciE88lWzv0162Jr7mVLxoNirlOSKUKIL/nIbHPVmOOffcs8Tbkdvqy7+mu+Ofiwlc8Lu\nKOx47vmETt5wKTTU1tmiv6+9awLHY9dwsoRDbU9+D56+buR2b34Utjw2ePtAyiRyjCCKSpnah154\nPIcxTdVRZkyL8foF9Tz3pXP413csoqlBrwe0viOSXUxur6pn3Z7grrSj3/zDNpm5MZLnz7BQxxMp\nHRwKiVYH76f6g8wzCJILqpp0dl041BarCUr62AoMWcGZFjzG8ghPTjVsRwhKyvSYlr3zt2LTvjUQ\nIzd8cscH4Nnr83cw6WT++TJDhdrcNZJc4enYEawi62bbQeCM3FBb3RHB83hncI3j3wnTF5nrhoXH\nvlbw0JcHVwCA4SeQ7v4LfPuYwZWrw4VXw45n0yODHUi6QMdjHWzby9rx2L+tnPJLoVBbsqew+UH9\n7WYtqIFQBYzu3DYORaofUFNv6Ys8eOHxjBkR4bF/ejN3fuINREqEj565kGPn6JpstfXN/L/3LQW0\n8KzfHdyVvvBKB9c9ugllO6xwnTXQ4hHv1I4k63j2Bh2HW+3Auh47JylaFbyfTgyu2ZboNuJRrTvj\n/gP6TtiG2qxI2WKmNr3bClC5G2oLzWPJZEz9MueadukI6x4SXeiwnQoKkrqZaGLS2O2kWUuiB65/\nI9x/9eDvK0d4nI7JdVS2U+1r19ezAmv3t697HeHpO6DDjDWz9POdK+G/FgTJENFq/Z3Z9rm47cik\nc4vHWoYb49m7Vn+/bn08u4aTi+t4MgP6e257OdelFOp4rOhu+ZOuPj7TjEW6jizseBI9gwWt/wD8\n+E265FL23Pu1aL50D1x7cjBB1705GA7b9vE6nudvGns18QnCC49nXFREI4i7Zo4ZaP7E205h8dHz\nASM8juO5+cmtfPvBDRw49VNw6uVwyocGnzhaHYyJWMeTSTHQuUuX/Slx/nRtIkGVqYCdIzz9QacR\nienOcCChxSBaDStvge+/Tv9TRyuDThT0nB0IqjLkDbU5YzzJPh2r/+6i3Cy78mlmsTxnjMUWWm3b\nGBxraTATd5+9Af77uKDo6sNfgf0bYfP/Df6+BsyaR5DbMdmkC1d4bAipa6cJPZrfjR2zyK6l1K/v\n0isb9E9/u05pV5mghFG0Kjf9HfQ5V/96cAist3XwtuHGeOy+toO2nxOVf7/svibF2k1qKDSd2rpC\nG4K1STDpYYQn2TtY0FpW6ErmLc8H2/rMDU63qdhhHbX9/kdqW/b7TYw9o04pWHY1LL95bMdPEF54\nPBOLk05N9XR47008EjuH9XuCziE5oENFe/pL+UXTPxIvzVNxO1atJ8sBxKaRqdIp3Lu2rR+cjJC9\nZr1e02fOqcE+qXjQaZRPC9bQqWwMhCXeaRxPZe6yE7bqQtjxuKG2nJn7fdC6Luj0rHPJ53gaj9bP\nrfCk+vTA85+/H1TcfvHXelzr6R/pDmPlLdp9tG8JROSVZ/QKnn3twXyskUJtdtBcZfR3EHYqWceT\n0OetqNc/fe2Bw7PrHkWrcrMQQYe67vlY/sFyK6IWd55VeEKt7ZBd4Ql3ziWluenUbuizdYNznJ1A\nWmCoDQCBGSea9uUTnqQZd+kefN5W41azy3kk9H4ZZwJzt61z54wzDocrbmN1Pal+/Xc9XKbhJOCF\nxzOxZCeQmppyJ/0t0+qbae8dnHq67MXd/Ou9a7jvL7v42Z+38tM/b6Urbu7kotXBHXh5Lb1RXYg0\n1v3K4IQCt4rB3y3T4w5Zx+MIj5sSnW/SbFlF/vWOomHhcSoZ9IaExx1HKXfGhCqbdAeaTmjX1WSE\nx3bWyV7tZLY9mdupSARe+KUOg2VSejAfgjvp3/69HqvZ86JeEdaey5Iv1OaGwNo2BgJrcVdP7T8Q\nTDbubw+E1nZc1iFGqwIBs/vs36zdYiSqJxdDkMBgsRNIVQY2PqQXlbPkczz2d1lSpsWwqlkL6S/f\nq0VtSOHpCx4f/gq84iyD4WJvDkBPcrbh23zCk0mZjjwzWDT2hYTHTcywz21Nw3ihjickPC0r9ITS\nzMCQhww+h/nbKHbtwBEoqvCIyHkiskFENonIoDQd0Vxr3l8tIqc4790sIvtE5KXQMV8VkZ0issr8\nXOC8d4051wYReVsxP5tnCLIlc+qym9wU7OaaWPb5qh06DHTn8zv46u/W8rXfreVf7jExcUckEqU1\ntJdooZg+sBc1yPGEJpNCAcLTCFf8CU75sHOeqtxQmyU2ilBbj9Px2U4rVhNU8LZhrppZJkXbkOrT\nImHHtSxv+mc9HmMz7I56s77LtwvP2YSAVJ9ue6w29242n+Nx2fvS4G2WtEk9j03T6e8DySDEZu/W\ns2HI6qBTs513Z4v+7Gd/Bd7+3+bzh4XHGa/5/efg8W8Hr/MKj+mc3/QF+MCd+vybH9Uua/nNuZ+x\nzRUec1x/h3aV63+ni42+dHdue3pbg7/h5uMGV8CA3BptQ83DssJjEzJy5kOZ78f+nrI1/0YSHueG\nIdUHW/8Ea34zuDL7sOewbutV6nhEJAJcB5wPLAIuEZFFod3OB44xP1cA1zvv/Qw4b4jT/49SarH5\nWWautwi4GDjBHPcj0wbPZOKUzLHMqQ+EZ67zfHWL/odZvl13LGce3cTjL7cykFE5zuP5PQO0qqAq\nd0ocgYHAAUXyC086af6h3UX0Khpg9mI48k25bY8N53jM8eW1ukOKxEJZbb25r13hqTLCY4UiNk2H\nIi2Jbt0BxU1Nt1mL4eRL4bXv1+/byZLT5ujwz84Vobk7pkp483HBnf5AKkhaSPbk3s1bhhtkTvXr\nu/HYtGAJDFstYJDjcSbe2uuoAf3+G/9BLytRUZ/H8TifoWdP/nJG+RxPw0KYt1S3zX7GF+8O3FbD\nUdD6snOcEQ57p5/o1ineT/0gtz29bbqALWKEx/wd5UsucNuWSQfjLplMkDGYdTzOd58Vnr1BMgqM\nMtTW4yRmhI4brkq4vTno2XtQ1xMqpuNZCmxSSm1RSiWBO4ALQ/tcCNyiNM8AdSIyC0Ap9TiQZ6r0\nkFwI3KGUSiiltgKbTBs8k8miC+HNX84JZVnHUxWN0FgVzW7vigd/+K+bW8v7lsylK57mpZ2d9B/5\nNtarI9ilGnhgZ5S2ZIQ9Rnx6M2W517R3pTmOJwYIpOI8uUGP5cQjTukT60DctYTyhdoiUV3RG5zQ\nmeN83A4l7Crc9OvKkPCUh4THdpjxLt2pHPkmuPA6nc0HwTpGFfVaBPr2w65VobaWwfTXBOML7thH\n1y4tBJXOon3R6sDB4CSIZFH6OuWO8NiO1h3jsefKCo/zb+sKef3CwPFkBvSclrCTGEl4bOdsb3Bc\nF9vVAut+p9Pzpx8/OAwKgYAkzLLq4XI7fft1iO0Dv4I3/ENQJzBfqA1ynYy9RucrwXP795EzEdd8\nntJyyRgAACAASURBVJ3L4T9m6xCj+9ninflvCHJqCPY6KeLOd/jbT+pzPvilYNu+9XDj2cFNDejw\n4EEsvVNM4ZkD7HBet5hto90nH/9gQnM3i4i9FS7oXCJyhYgsF5Hlra1DlD33jJ3Go+BNV+sJogbr\ncmrKy6itiFIWEWaasjtN1TFKBN75utm88SjdKT65qY3rN07jvMR/8qFpP+X+jf20did4IaPHRbrS\nISObnUzqCI+IvltNx9nbrjuXXpyxIbuuUY7wOMkF1j25QhQOubnJCqBDbUM5Htvhu47HZuFBMM7S\n16ZDWjaEFa3Ujq7NER5boHSnMx4CupNsPl53dj2tuUtF2FRpO2kXgRknBO7IClwYmwEYXi7djr9Z\n4bHzriBXjN3vr35BkFyw/g96Eqt1UBY3nGSFp7cV7vqILufz/I16m3Ui9mageoYOQW7/s/6uq5qG\nT0pIdAcu06KUvlZVk16zqmZGcJ32rbDu9+Y7yeN43GtsM4sH1i90lpRwHY/ZtusFLR5bn8g9/gen\nwvVvYBDh5AIrOFaAkn2w6lb92r0p2blc/+zflPv9FnupjmGYiskF1wNHAouB3cB3ht89F6XUT5RS\nS5RSS5qbm0c+wDNu5mSFp5T3v34eXzz/eGbV6X/oU4+oY9lnzuLyNy6guSbGa2bW8N2HX+baP27k\n7SfN4tNnH8P+3iSPbWhlZcYsuR3OwnIKhv5+9S52d/ZnX9PZQtNe3RF0K+N4YtMCF5PjAJx0apt1\n5t6xn/AeeOs3gqoHbuiuvFZ3Wm7nkBNqM9exd/zl0/J39nY+jnvdqqYgtbyi3izJ0DV4YmUkqh0P\naNcznPCU10LjMYGA1A5zvxer0dctC41/RWLaZdl97HjHUMJTN087LzcUFZ78meN4zHu7XtBjMfFO\nXdYHgt+5vQloOhZmvU4/r2rWodT+A0GmXFh4kj0mtOlcP9mrHZj7N2H/Tu79JPzqg1rQ84XaIPjd\nL79Jt+foc/I7nnDI0wpHJqXnjmVTukPzdpLhUFvI8bgZgz179ZjX7Zc4zrEjV3iKvSruMBRTeHYC\n85zXc8220e6Tg1Jqr1JqQCmVAW4kCKeN+lyeyWFune7wa8pLWbqwgY+euZDmau0oZtVW8JqZ0yiN\n6D/Fb7z7JD525kK+9q4T+MElJ3PyPO1Mnt3azoYy3anOS2nXsLWtl1uf3c5ARHdCCcr41G0v8JPH\n9fsJicGaezi79RYAOjOms6qoZ9mLu1m7qys3u811PM1mno27suq0WfDGTwVuzg3zVNQPLqNfPV3P\n6p+9OEiRzobaaoNQm3sNi9thu66svE5fN9GthcUuXgeB4wEdXnHnEtnU8XpTFbyiTrtTi3U07vks\nsRr9mcOuxxXHaXNMOE/ldqzuPrXztLD2tuYuge5Wnkj0BGWNrPC4naUVW5tQYsW/dh7MNy6hqkn/\nXjPpoefIJLr0NQYSOiX94a8EHb7rREtDiSzbnxxGePq1SO5cAa//mP692SXD3ZDccOWBnr3BOXe7\nzgz85hE6NT0VCrWFHY+9qZm7VH9Pmx/VmZJWeNxQG7xqHc/zwDEislBEouiB//tC+9wHXGay204H\nOpVSw34bdgzI8G7ApuXcB1wsIjERWYhOWHhuIj6IZ3xMqyilOlZKTXkwNmOz22bX5f5jn3pEPddc\ncDwffuMCSkqEOfUVREtL6Emk2VOpS+yUoFBK8a0H1vOl37zE5gM6TLWxTf9Dv7RTd7idqdw/7/YB\nfS1V2cg/37Wam57cmrtsd1mFEQPRg/uQP9nAYh1NxKzqGl44LFoFn3xa3/mWRnWnng211QYd3LRZ\nDCKWR3ii1eY8NWYuxh6obtapxbYdNTN1u/atDQmP+beyy1FU1OcKj013rnPv3WxbzPcTFh43A7B2\nru4A+/bn3t27AmqP72zJFZ5yR+zS/Tqz7YazBk82bTgqECGbUJIVnrkw/3T93Doe0G3JDOjO3r1O\nT2sQ3lzxc53pZjvuKtfxhIRn2wjCY9O0F/1NcFMT78if2JGPbU8Gz/v2a+HIpGDjw4PTqcOOxyZu\nzD9Nr1fUvkU7OJt5Ge/IHSd6NToepVQa+BTwILAOuFMptUZErhSRK81uy4At6ESAG4FP2uNF5Hbg\naeA4EWkRkY+at74lIi+KyGrgzcA/muutAe4E1gIPAFcpFZ6g4DkYiAiL59Vx9PSgE7LCM6s2T5FP\nh0iJsKBRO6bq6hpenH8Zn05+iq1tvfxxnb77fWmf/sdr6dFhlTW7uhjIKPpVNOdc+5P6dTpWR08i\nTWe/CWvZjLOySt3xXvUcLP6A3hbL40YsttMrLdeiFR6sjYSy71x35YbaavIIj+uCbEdox6XsuEZn\ni1lF1nyvpVHtTGacmFv5GoKqCXXzg3PZcCLAzBN1e5uOyfM5TYdthSM7/hRyM6A7v0Qn2WSFsDiB\nHnwfSnhAj08c2JrbqVc2Bm4NgvE8+/upmwfzrPA0Bd+1rY9mz2FxyzTZm4E95h7WdTyR3L8htj4x\nfKitvx0Q0wZzvV9fDqt/lXse6/LCDrN7dyB2ffvhlaf18x3PatFwq1NYwelvh99cqcOQ5bXQbMKt\nNqXbnSht06mrph9Ux1M68i5jx6Q6Lwttu8F5roCrhjj2kiG256mvkn3vG8A3xtRYT1G55SNL3XyD\nIR1PPo5squblvT00VMXYeeqXuO/lFVT8aUu2AsK+/hIohR2d+j6jLznA1rYe0gO591W7E/pavRH9\nz95pi5VWNur4eHb+xrFBamy+CaUW2+mVxnIXC7MZXuFOywpGaYUeG7HjKvVHQLiMWY7jscJjUtSt\n4HXsgLmn6uv1HwiuN+t1sPyngfMor9WdjpTAtNlmW52TaIBO2z76HHj6h+Yz1ASdVNbxGHFpPFLX\nassnKrtXBZ/pwLZc4bbH73ohVFIo1Pna8SiVCWr0NR6du184q612rnZ/539LV6/Izik6EITZKhsD\nkXHL7thttv6ceyMQdjxtG3ITSsKOp2+//j2VRIIbhW0meWDOkiAh5Ohz9fdX2QjPOcVTe/bp8Oy+\nNdC7PxjT2vWCTlaoqNfXsONRoMXwL7fr57MWBzc09r670+Rc2fk+kSic8Zn8TnuSmIrJBZ4pSEmJ\n5NR0O/PoJs5dNIPjZ+UplxPiSLPuT1N1NJshd88LLRw9vZoT50zLLqnd2h90Jiu3d1CtdMf59dSH\n+LvU1axL6rvgDnRnFQiP7tivfaKFtBEzImVmATgtAM9tbec7DzkTEsERkvLcTtHWWisNCY+9Ez3p\nvfpx1mL44N1w/LsGf+icMR5z927DR7azTZjK3dlMPEd40v16rEEiwR18eW0wv6qiXnfetfPMsg8R\nHWa057Kdpnu9o94CC84KysjkJA4YJ2U7b1t53N2nvFZ/p+GlAWKhvwErPO55G47KmRuWFYSmY7WQ\n27Gt0z6h3ZvreOygf4XjOF2sU939Fy3OOWM80aH3h8GOp689f6r+Rx6C9/wkeF2/AP72Jr0MPDjf\nt9LCDrBrpXZAR75Zj41tfypIgEn2BoLqjoE1LMxN04eg9JStKh6t0mOVJ743//cxCXjh8RwUjmis\n4sbLllAZHdl0H2XW/WmoimbnBKUGFG95zXQWz6sjbkJqCaJURiOUl5Xwx/V7mYHuFNarebzScCZ7\nzU1wW0afL8fxANc+sYvntrXzo8c26RI/b/kyLL4UgF8v38EPH90UCBMEHWYkqmfSl1boMRTrUMKO\n5x3/A/+0Sc/PAR0WO+ac/K4qnNUGQefkugg31GavZysqb3tCv2/dWHmdEUsJOsXGo3Kvb12MU3ki\ne725p8Llvw/a4zqeinp9Het4sskZzj42QcGKk3VA5SHhcQfRrfA0HpUr7lZ4jngDXNMy+O7dHeNJ\n5Qm15WP/Jh2Cijh/k67jsVl9iZ5gjKkvn+Mx13ZDq/NPyxUE69js2Jqb4Vi/EBA9rgN6Ai7oEGWZ\nKcrqOh63ekb9wsHZktnitJ1apPIls0wyRQ21eTwTgXU8DVVR6irLqIxG6EsO8NfHNtOdSPP8Ch22\nS1LKzNpy6irKePzlNsoiOtSwLTOT0+fWsXoVEIM9Kd0RZ4WnZgZJiZGmlK//bi3r93STTGf47DlX\nZtvwSnsfSsGBvlRQ9icbaivXd9nXtOjO4Ld/r7eHhSc6REmeqE3zrjXjI+TPassKTyiNO+x4mo7V\nbYp3wrHnBSGWinrt5N7/Cx32AXjdB5xJpDjCk8fxhF+7bRTRQmKXAbDuLnys7Yynn6DFpHPH4FCb\nix2Haj4uKKgKuYIQydONVdQBkut48tXny0Hp5AwX93dYM0OH5dSAFg5byy4S1Y7EjvHY1XXtDcjs\nk02bnfFMezNgw63VM4IU88pG/f3v36g/wxFv1ALV1aL/VgaSWkCs47HZeNUztCutbCK77IZLvMOU\nVRomfDxJeMfjOeQ5ftY03nRsM284qhERYU5dBVXRCEsW/P/2zjs8rurM/5931OuoS5Zk2Wpucrex\njW26AZtAbMD0hBoIBLJhSVjCDwgJ2bBhaSlLD2yAZSEGAhhCaKYYew3YBrk3uUouarZk9Xp+f5x7\nZ67GM5JcJFny+TzPPJq599w758zV3O+873nP+yZwzqhU7jxfR6A1qRDSYsMZk+GmocUbV7KXBMZm\nutmjkqhLHs+KNv1rvLGlnabWNph2Kw8lPwjgyaL9j9UdJ16L9+ubV2WdIxTW42qzhCgoWH+pnZZQ\nd7B/Sbsz8U7K+5vj8WPx2HWFnO8XFKytnoh4bWU5M4aDTqJqWwjjLoOzf+M9X2euNk8bW3h8RNQZ\n9ZZ7lrYY88/u2MaeVP/eo14LwOlC82XIdLjqDRh+XscoQlcXty5XkG7f4JzjsYQnKiXwcb7CI+Kd\n0HdaErZwNBzwnq+lQVtAnijESLj2fbjmXavfwd4IRPvHhnuwPpcz0CMiznuOuMH6+qVbUZb2jxen\nxWNbNDcvhexT9Ps4I/Ns7HBqfz9+ehlj8RiOe8JDgnjxem/2o/PHptPc1kZosL75hIXrL3EzWnhG\nZ+gb1DOt3+NS9wbGJMVzUnYCjRLO2N136VxwFltKawkPieCz+jxAu3jSYsPZUlbL5tIahqXG0NTa\nxt6D+kteWevIsm1HJPlOQNs36u4Kj30TsrMStNR3zMLQlavNFgvnfMS8p3T0VWx6R1dbl33xsXjs\n+R8nYQGExxa4iddoYTvVT8G68x/XEXdDToZtn3nHAB0tPs97xULGxI7t/JU790dkgna1tTqCC0AH\nPgRKF+MrPKCvb1uTf1dZS52OBDxY4nC1OUR76IyO5wqJgKYWR3BENNy2Qq8rWvG8NU5LeCq36EW+\noOcDN76nQ8N953jsIBLn9YhO1YtjnWuGGqr0HFZnATO9hLF4DP2On83K585zR3g3WDf+ZoJJiQ1n\nbKa+wT4uPyTuF9+y8LaZFKS7+ewXpzMxq+PN9843VvODv3xNSVUDp+QnkRQdxh8uH0+QS3h9hY4G\n2n2gwbMAvqLWp7IoHDoBHWh7IDzC4PYKiTME0J2pXWODLfHt4GqLO9TVBrrsgp3FwNfi6U5f/Imc\njW/qIJsZt8Mpv4DvPRb4/Im5Op8feC0551ocG0+4sTOxq9X/YJ8w9UBExFuuNh/hiXOEZXuyMVif\nt7/Qdvs6RjtEyZkhPdytLRlb5Dpz6dnXwhkF6c44dDGyZw7OsoRsi6dyqyU8tY48d0r3P8QhyO5M\nna/OWVbeGVzQxxjhMfR/rC9cE6GkxYaRmxxFeIiLQe6IDpF0QxKjeOKqiQxLjebKqXrSeuO+g+w7\n2EhzazvnFKSx4t5ZTMtJZM7oNF79ppiDjS3s3O9duLe1vI4H3l1PQ3Nbx6g2J+GH6WqzbwThbst1\n5nNDD4mAGxd5F0gGBXcUK4+rzSd5quf8h2Px+LjaOhMe3xvY4JPgrPv8z7n4w75B2/13Co8dpdbB\nurP671sWIxARlsVjz3ElDYcz7oEJV+nXweHeG7wdWh7I4gE9x2PjGz4fEqmtHug8iME+l/N439cR\ncd7Pxl5XZacDOljiDY93ZmPw/bFy3sNw8fNeK9EVrOd4mmo7X5vWSxjhMfR/rCii08bmcU5BGsFB\nLiZmxZOTdOgvu5SYcD7619O4dLKOqHIWvXSWbPjxqbnUNrWyYHmxZ34H4LVvdvHC0u0sKapwLCDV\nv8CXbKngoQ82Olxt3l/mOyrqKD3oyCLsJChEt3VaPF3hSVTqDC4IYAn4qZEUEN+otsMRnsPFvkEH\nh+uHMwtDwVxt5XVYdHsErraafbDu79qSScjR9Y08efhivT8S7LDmaH/C08kcD2jhDInwlrIOFLYN\n/i0e53bQAmu7V+2+RqfA4Kkw90lvXkBn8IBvwEBclrZ67f/R2Azteq0rMxaPwXBMyJwM177P9Zdd\nQroVbv3kVRN57NLxAQ9xRxxqHQyO994MxmS6yUmO4pvt+9lVWU9YsIvkmDDKarSrbU1JVceoNuD5\nJdt46vOt1It1HocFcv1fl3PPW50UXZv3pM7vlTXNa9l0hrM2UFdzSs7S4F2RmAuTrtXRUUGh/oXH\nTisUKKN1d0nI0euM3Jl6bVDaOO/7jZ6vrTynFecRnm662vLP0TV+ij6B8Vd5LTFnRVn7nCkF+q8/\ni8cTXOB0tTmEIjbDEh5rvUy3XG0BquiCFn07YCFpmHf7DR9pay087tDCc4HExB6fHZbe2nhczPGY\n4AJD/0fkkEncuMjO3VxO4YkJD6amsbWDxQOQlxzNtoo62toVQxIjCXa5KLeEZ/XuagjNB4S1pY18\n9eU2VuzQazo+rh9GbORs3vykgXvnNeIS2FZRR2NLJxmcxszXf2fdH7BJdUMLkaFBhAS5Alg8AVxt\nhxNcEBQCF/xRPw+OOHRxJ+ib2M1L9Ar7oyF+CNxZpG/UNy7S25b9Wc/L+Lt52zdR35t2IEZfDGte\nh80f6qJ6Nr4lzCUICubp7ADOG72Nx+JxLCx19mHIDP0edq63Tl1t1nGhPhaPiN7X1qz7NP4K/Tn7\nyxruL/y8K+Gx10zBcRFObYTHcEISG+7917/v/FEcbGghPKRj9FZuSjSfbiyjuqGFmXlJHQIL1pRU\no0SQqGS+K2vnwfc3YAfLPfBZOZV1V8P+KmaMLPOI3J7qRqobWvxaW06e+nwrH6zbxzu3esVUKcWs\nx77guhlD+cnpeT7CY910AlkCh+NqcxIWHdhKSht9eOcKhK/AhMbo+Qh/ghccpm/O3bV4ROCi5/Q6\nJTubBGhxDQ73lnuIStIT8Vct8H8e+/0i4r1rdpzCkzW14+sjcbXZ+0IidL8j4mHk+f7P4e86BrJi\nPMKT2XXbXsQIj+GEJDjIRUxYMDVNrcwenUZs+KFikJscTWu7orymiYL0WNbt0Sn2UyyX257qRqLm\nL+CxZzZi5zOICQumsq6ZyUPiWbfnIFtKa1E4w7drmDy0443pYGML0aHBuFx6cnjlzgOsKq6isaXN\nI4aVdc2U1zSxqtiaKA+P1TdoZ7nurlxt3bF4nFz8vDe3W29hi51zotxJuLv7czygPyd/rkvb2jnl\nDhh7aefn8CQkjdWfZVuzzxxPjI/wdOLSDORqs7d1R1SPxOJJn6A/t4RsvcaqjzFzPIYTltiIEOIi\nQ/yKDkBusvfLPDrD7SnbfdFE/evxpWU72NCexQFiiQ4LJjspimm52s0ya1QquSlRFJXXsnLnAYZY\nGbY3lXZM9X+grplpDy7i3dXeFPW7q3S00i5HUMMea9u2ciudTLjbe4NOtCaR44f6H2jycP0r3F/J\ng84YcnLHjNC9QVhM5xaDO6NjLrUjJSFX50RLyofcMzpv6yw6F+xjsdh56+wolfSJnYfRe6La/AhF\nSET3fhw429hh510JT+oouGevLtOR7Med2MsYi8dwwuKOCCEhKvBNIifZ65IYlR7Ld7u0tXHRxAyq\n6pt55ottrLTmdV69cRqxEcG8u2oPH68vZdbIFDbuPcgXm8upbmjhtjPyeGHpDjbt6yg8hcVV1De3\nsaXUm+jRFpkdFXUMS43puK2yjta2doJn3A6jLtQHxA+Fu4sJyNCZcJdv+uvjlJNv65i52pfL/7f7\nrrbOuObdjmtcOsOut+QKcpRZD4cbPvamBqrYrP/Oeajzc3Vm8YRGd51PDjpaPBHxOsKtK1dbWGxg\nK7IPMMJjOGG5YWa2J/uBP9wRISTHhBEVGkRseAjnFqRSVtNIbnI0v7twDN/tqmLFzgPEhgczOiMW\nEeGHJw9l5KBY8lJiyE+N4e1CbcmcOTKVJUUVbNzbUXhWlWgxK6vRoda6TpDOIee0eHZX6f0tbYqS\nAw0MTcr3Xzunv9OV9eEv6uxI6O7iXtDWjWcNkR2xGOxd0Atw8XM6l5xzm99zOYTLlzn/2b3ACecc\nT1fCk3uWzi93uG7WHsYIj+GE5eJJmV22+f64dGKsQISc5Gjuv6DAs+/KqVncv3Adw1JjPAtV3REh\nnDVShxnbWbWTY8IYm+Fm3OA4Xv1mF02tbYQF67mb1SX6131ZTRNvfVfSISXPzspDXW0AX2+vJMxa\nIHu4NLe2e8T2+SXbCRK4dkZ2F0ed4Mz4F2/UoS0MLp9bZ87p+tEVsRn64S/XXNbU7vXH1+KBwK62\nwSfpx3GGmeMxGDrhvvNHcfss/z7xeRMyiAwNYlS6/5pC+alaeM4cnoLLJUzLSaSxpZ1VxVpslFKe\nYIGyg038x/sbefB9XTUy2CXsqPSWB9hT1UBStHYx3fXmGi59Ztlhj2V/XTPjH/iILzaXo5Tiqc+3\n8srXu7o+8EQntcCb7DSQ8HSXaT+BW5YeXX9CY/Ck+OlKeI5TelR4RGS2iGwSkSIR+aWf/SIif7L2\nrxaRiY59L4hImYis9TnmYRHZaLV/S0TirO1DRaRBRAqtx9O+72cwHEvcESG8c+sM7jjbvzBlJ0Zx\n7fShXD9TWxRTsxMQga+2VQI6iKCyrpnQYBe79tdTVtPkCckePzjukOACp8AV73ekS+km2ytqqW9u\nY8Peg+ysrKeitomd++tpb1eHtK1tavW7/YTnaIUnOLR7C3k7w+XyWj3OZK79iB4THhEJAp4A5gCj\ngCtExHfF2Rwg33rcBDzl2PdXYLafU38MjFZKjQU2A3c79m1VSo23Hjf7OdZgOKbkp8YEXKzqcgm/\n/n4Bw9N0gEBcZCgj02L5cN0+CourWLlTByackpdEbVOr57hglzB5aAIlBxqortfzPburGsmIC+eh\ni8dwco6egK6qb6YzHv1oE/Of+j/P633Veh1S2cEmlu/QGY2bW9s9mbdtmlrbmPnQp7z81c5ufw4n\nDEcrPMcKX+E5DhaFHg49afFMAYqUUtuUUs3Aa8BcnzZzgZeU5isgTkQGASilFgP7fU+qlPpIKWV/\nS78CunbUGwzHCWeNTGHdnoNc9ORS3incQ1RoEKcN94YHuwTS3OFcME5nSf7Ne+tobGmjoraJdHcE\nl52U5bGgtlfUsbe6gQff36CTlgKfbSzz5IT7dGMZK3cd8GRM2FutraSymkZPlgWAhYV7ePzjzZ7X\n2yvqqKpvYfHm8h78JPopdnBBXwuPHWBgXG2HkAE4YzxLrG2H26Yzrgf+6XidbbnZvhCRU/wdICI3\nicgKEVlRXm6+WIbe5Y6zh/HWT6bTrrQwTBwS3yFI4ObTcrlgXDoF6W5uPT2Xv3+7m58v0KWix2Tq\nX7lDrTVBOyvreerzrTy7eBt/+nQLVfXNXP/icu59ey2NLW1s2leDUt61P7YgldU0sXLXAUYO0q67\nRz/axB8XbaG+Wf+es0O7vyuuQinjbuvA8WrxGOHpHUTkHqAVeMXatBfIUkqNB+4A/ldEDpn1VUo9\nq5SarJSanJx8DBaiGQyHgYgwISuecZaITBmaQIpVSjs5Jox/mz2Cu2brtSG3nZnPiLQY/rFmLycN\njee0Yfr/dXBCJCKwYe9B3vp2N6HBLp5bvI3XlhejFHy8vpSFq/bQas3RbC3XQrK3WgvPnqoGtlfU\nccbwZEKDXZ52+6z9W8p0+/11zR3mmaobWnh31Z6jFqOK2ibeWFlyVOfoM0IOTQDbJ4TH6XVI/sqQ\n9wN6Unh2A86l0pnWtsNtcwgici1wPnCVsr4FSqkmpVSl9XwlsBXo+yW6BoMfLhinU9GclJ1ASqwW\nnuzEjr9aQ4NdPHbpeMZkuPnV+QWekO3wkCDS3RE8s3gbNU2t/OnyCYjAYx9vJizYRXiIi9++u95z\nni1lteyqrPdYPCUHGmhrV+SlRDMkwZv6pfSgngPaWlbrCbn+dpfXJfeXL7fx01e/Y7nDTefkr0u3\nc9ajn3c59jdWlvCL11dRFqhMxPGMvf7meLB4giO8WRyORTaHXqQnhWc5kC8i2SISClwOLPRpsxC4\n2opumwZUK6X2+p7IiYjMBv4N+L5Sqt6xPdkKaEBEctABC9uO3XAMhmPHD6YN4Q+XjWdqdgJJ0WE6\nwXbSoYkjR6XH8u5PZ3rcbDYtbTo73My8JM4tSGXWyFSaW9uZNCSeO88dQU1TK8kxYQxJjOS5xds4\n/ZHPPLnmbLKTohjqqFlkC9OWshpm5iURFRrkydYA8OG6fQD8bbn/LAkrdh5ga3kdu6sauPftNdQ0\ntvhtZ1tWu6sOPzKvz/G42oI6b9fTjLzAW77ilmW6nEU/oseExwoAuA34ENgALFBKrRORm0XEjjh7\nHy0ORcBzwE/s40XkVWAZMFxESkTkBmvXfwExwMc+YdOnAqtFpBB4A7hZKXVIcILBcDwQHhLEvAkZ\niAghQS5+cc5wLjspq9vHXzhBT4U+dtk4RMRT2G5qdiLXzxjK9TOyuXJKFnnJ0TS0tNGuoL65jeQY\nb7qZnKRo/uXMfB65RFe33HewkZa2drZX1DE8LYaCDDdrd+s1R9sr6thcWos7IoT31+yl0lkC3KL4\ngBaSt74t4X++2sX7a/z/hrSzNNiuv36FJ7igj11tw86F2Q/qNDipR1meog/oUXtRKfU+WlyctFr4\nMAAAFSNJREFU2552PFfArQGOvSLA9rwA298E3jzizhoMfcitZ/j9tw7IXbNHcPusYUSE6l/epw5L\n5t7vjWTueC1mv7pA34z+458bWLSxzJNRe2yGm0Uby0iMCsUdGcKYSDdjMt38euE6Sg82sr2ijpY2\nRX5KNE0t7fzvNztpa1f8danO9fbIJeO49ZVvueSZZbx24zRSYsN58vMiEqNCPZVa1+7WltWH60r9\niqnt0tvTry0ek/TlaOi3wQUGw4mMyyUe0QEIcgk/OiWng0UDcMOMbJ67ejLXTB8KeCPjsn3KgqfE\nhlF6sJFCy7U2bnAcozNiaWxp5643V/Pisp1cOTWLs0el8vINU9hVWc+Tn29laVEF//nBJv74yRb2\n1+l1RWssK2nJlgpm/P5T7nt7rSfcG7wuvUAWT2NLm8cqOu44XsKp+zlGeAyGAUxKbDhnj0pl3oQM\npmYnMHu0TrKZk9xReNJiw9lX3ch3xQdwR4SQnRhFQboWqTdWlnDG8GT+fa4uATA1J5Hvj0tnwYpi\n7lhQCOgidza7qxoIDXLRbM1DvfzVTh76YCOg0wSVWRbPvgDC84dPtnDu44tpaG6jeH89Zz7yOdvK\na/227XVC7OCCPp7j6ecY4TEYTgAy4iL4249PZnhqDJOHxHPasJQO+9Niwyk92MR3u6oYPzgOl0vI\nTY4izIpuu+3MPE+hOoAbTsmm3rJi7jlv5CHvN2lIPG/fOoNFPz+NcwtS+XDdPqobWigqq/UI0h5r\nQWubT2qezaU1HKhv4f01e3n6i61sq6jjH6u980Ufry/tu3Ds4yWcup9j7EWD4QRCRHjjlumHbE+J\nDfdEmdlWUXCQi4lZ8bS0tTMxq2N+sYJ0N6/8aCr5qdGEBQfxOyu5aViwi6bWdlJjwxg/WK+uP2tk\nKh+uK+W8P37pKfkQHRbM3qpGGprbmP77RfzqglFcOEEnIbHnil5bvssjbs7yFQ99sJEdFXWMHxxH\nXsrhrV9paWsn2CWe0PTDJm0sDD3FW4fHcEQYi8dgMDDI7a0P4xSZZ66exF+vn+L3Rj0jL4mUmHDc\nESHkJEd5qrACpDrOd8ZwbV3trmrw5KQbm+mmrEYHMxyob+G9VdqiUUrXG3JHhLB8xwFPCHiFFUVX\nVtNIUVktre2K3/3Du1bJSXu78pvgdH9dM+N/8xFz/vgl631Cy7tNTCpc+x5EJR3Z8QbACI/BYADO\nHzuIf501jD9cNp4Zed6bamx4CNFhXTtGzi1IY3puoie4IS3WKzzJMWHMyEtk0hCvoI0fHEe78hbC\nW7atkubWdirrmmloaeOnZ+Yxf1ImQZZ7r7xGC8+yrZXW+6Xy2aZyiqwsCy1t7dz5+iqe/mIrpz78\nGfe90yGpPQArdx6grrmNjftqeGFpP6nIOkAxwmMwGEiMDuNns/KZNyHDc7M/HO6aPYJnr57sV3gA\nXr5+Kgt+fDIjrEzdEyyr6pvteqldfXMbK3bu97jZhiZG8fD8sSy7+0wmZsVRUdvMJ+tL+dvyYmLC\ng/nt3NGEBAn/vXQ7RWW1bC2v5fWVJfz+nxspOdDA6ytLqK5voaishj9+soXWtnYKiw8Q5BKm5yby\n3S7/2RcMvYOZ4zEYDMeMZKtYndPVBngCEy4/aTDvrd7rEaClRRW4RIeD3/5aIQVWzaHMhAhEhJSY\ncJKiwygqq+WWV1bS0qaYMzqNlNhwzilI45Wvd/HK17v45Rw953L7rHwK0t3c+NIKFq7ew86KOv6y\nZDt1za2s21PNiLQYTs5J5NGPN1Pd0II7wgQJ9AXG4jEYDMeMVMvSSQ9QlvvaGdm8cct0MuIiCAt2\nUVbTREpMOM9Z1tJnm3TG+MHx3vRByTFhbLMWtt557nAemj8WgDvPGc5VU/UCVTtLwnXTs5k1MoUR\naTEsLNxNkRWG/ezibXy9bT/jB8cxPksHPay23HzV9S2e0O66plZufGmFJ2ODjZ0YtXh/Pec+vviw\n0v18sr6UTftqut3+RMAIj8FgOGZcPCmTJ6+aSJqPxeOLyyXkJOuItPS4cE4fnsLD88d59kc55pXs\nkt8Apw9PJjZcWylDk6L41QWjCHIJq0uqcUeE4I4MQUSYkZfE6pJqNu2rYc7oNE4aGk9ru2Lc4DjG\nZmrh+f0/NzLrsS8Y98BHnPXo59Q2tfLhun18vL60Q32ihz/cyNmPL6a8pokVO/ezqbTGM9cUiP11\nzZz28Gd8t+sAP3ppBef+YTE7KnQeu083llK8v54rnv3Ks+j2cKhtavUUCOyvGOExGAzHDHdECOeN\nGdSttvYi1kFx2joalR7LxKw4pmQndGhnzxu5BHKTO4ZPhwUHeeoT2X9BR801tbazt7qRkYNiefKq\nSVx98hDOGZWKOyKEKdkJ7K5qICshkiunZlHX3MbSogoWrtoDwKKNZZ7AhUUb9PMfvbicnZV6Dmpz\naUcL5tVvdnG/I6BhVXEVOyvrOyRU/c276/ivT7dw00srWbShlGXbKiks7jjXVF3fQunBRtrbFQcC\niNLPFxRyzX9/E/BztftXWFzVaZu+xMzxGAyGPsEWkXSHdfT6zdPxjW2wLZ6shEjCQw7NGDAsNYat\n5XVkOcpKjMnwZvPOS4kmOSaMB6zMCwB/u2kaoNc1tbS1s7BwD2+uLGHJlgoumZTJu6v38G9vrOKZ\nH05mU2kNyTFhrCqpJtaaE1q7u5oL/ryElrZ27po9gqe/2MrOynpmjx7EybmJbLKE6QMro3duchTf\n7qoiMz6C1nbFJxvKAF2iwqa1rZ2rnv+KuqY2rpsxlH9/bwN//8l0RjvG0tTaxheby2lubae+uZXI\nUP+38Cuf+5qK2iaW3X1mh0KDxwvG4jEYDH1CrmXxpMd5b4xBfhZ32hZPfmqM3/MMs7Y7LZ6hiVHE\nhOubcr6fRaYi3vcJCXIxIy+Rj9aX4hLhx6fl8IfLxlNYXMUP/vI1SsH8SXpxq+1i+7+tlazZXc2e\nqgZu/1uhxxJ64L31rN1dzWZrTqfKcoldNDGT6oYWz7qkZdv0eYodhfb+56udrN19kO0Vdfz50yKa\n29q5Y0EhTa3ePHcrdx6gsaWdduVNxgo6nPynr37HmpJqz2uAB971v9aprzHCYzAY+oSCdDcidJl9\nwI6U8ycg4BWeLEdRO5dLGJPhJsglDEnsuiy07R58+JKx5KXEMHv0IK6bke2xXC6zyk60OhamRoUG\n8dt5oz3ZGH51/ih2VtYx74mlfFlU4WmXEBXqWURrY6cJ2l5Rz62vfMv976zlwfc3ctLQeIJcQnlN\nEzPzkthcWtvBXffllgpPuPsqhytt7e5q3l21h3dXa1ehvfbqg3X7PItvjyeM8BgMhj4hLyWaxXee\nwcy8zrMApMeFM39Spqdqqy/TchI4OSeR6T7nuXxKFj+cNqRDup1AzB2fwcp7ZzF3fIZn2y2n5xIR\nEkROsi6Yl2pVih1npQI6fUQK3xsziIy4CMZkuLl+Zjaf/vx0QC94tV2IecnRDEuNJjxE9yPKkVV8\nSVE5/1izlxeX7SQrMZLnrp7MVGuO63cXjuakofE89flW6pt1xofPN5UzKSuejLgICku8wmMX7LMz\nMuyva+bknESUgs82lnnafbG5nPrmVj7bVMauSq+11dsY4TEYDH3G4ITILvOmBQe5eOSScYwcFOt3\nf2J0GK/eNI2MuI5zGd8fl86vv1/Q7b4kRncsKZEUHcZjl47j7jk6CerwNP3+548ZRExYMPMnZRIc\n5OLlG6bwxJUTAUhz62zgAN8bq62ovNRogoNcFKS7iQ0P5uRcLZDxkSE0tmiX2PPXTGbBj08mLjKU\n22cN446zhzEkMYqfnTWMvdWNTH1wEX/5chsb9h5k9ug0xg+O4/+KKvho3T72VTd6AgnW7ammsaWN\nhpY2ZuYnkRYbziJrPmlreS3XvPANv31vAz9fsIrfvLuu25/NscYEFxgMBkMA5jgi9EakxbB4czkT\nsuJY/etzPIKZ4xNpd92MbD7dWMYlkwezfMcBj5vttjPz2FfdSPH+ej7ZUMqpw5J5p3APgxMiOGtk\nquf4KdkJnsi+mflJvHbTNO59ey3//o8NBLmEC8alM3FIPIXFVdz08kqCXEJ4sIsgl3CgvoX1e7XV\nkxAVypkjU/j7tyW8vGyH5/yvfrML0BZdX2EsHoPBYOgGM/KSiIsMIT8lplMrbUp2AusfmM2w1Bje\nvnWGxwI6Y3gKV0zJYs7oQZw9KtUjSCcNSQh4LoBpOYn8bt5oTx+SY3Tm70U/P01nCE+Jpq65jbNG\n6PMt3aLnl+IjQ7jltFzGZLi57511/OnTImKsuZ8p2QlMHtr5+/YkPSo8IjJbRDaJSJGI/NLPfhGR\nP1n7V4vIRMe+F0SkTETW+hyTICIfi8gW62+8Y9/d1rk2ici5PTk2g8FwYnHasGQKf3UO7siu0+x0\nlu9uTKab566e7Amq6I4ATM1J5JFLxvH/zvOWYwgPCWJGXhJPXDWRSUPi+emZ+QAsKbKFJ5TBCZG8\neuM0hiRGUl7TxDkFaTxyyTgevHBMl+/Zk/SY8IhIEPAEMAcYBVwhIqN8ms0B8q3HTcBTjn1/BWb7\nOfUvgUVKqXxgkfUa69yXAwXWcU9afTAYDIbjjoL0WP58xQQumpjRdWN0SPeItEPnuXKTo3nzlumM\nyXQzOCGC76z5noSoUEDPkf34VO1Wm56byPxJmYddx+hY05MWzxSgSCm1TSnVDLwGzPVpMxd4SWm+\nAuJEZBCAUmoxsN/PeecCL1rPXwTmOba/ppRqUkptB4qsPhgMBsNxh4ier/G3KPZIyU2OprlVByzE\nW8IDcOnkTB67dFzAyMDepieFJwModrwusbYdbhtfUpVSdh3cfYA9K9etc4nITSKyQkRWlJeXd/FW\nBoPB0H/IcwQ6xDkybwcHubhoYma3Qst7g+OjF0eI0iljDy012PkxzyqlJiulJicnJ/dQzwwGg6H3\nsV1oseHBBAcdv7f3nuzZbmCw43Wmte1w2/hSarvjrL/26qgjOZfBYDAMGGzhSXC42Y5HelJ4lgP5\nIpItIqHoif+FPm0WAldb0W3TgGqHGy0QC4FrrOfXAO84tl8uImEiko0OWOg8havBYDAMIGzhiYs8\nvoWnxxaQKqVaReQ24EMgCHhBKbVORG629j8NvA+chw4EqAeus48XkVeB04EkESkB7ldKPQ/8Hlgg\nIjcAO4FLrfOtE5EFwHqgFbhVKeXNrmcwGAwDnLjIUJKiQ4nvRsh3XyJ2Zb0TkcmTJ6sVK1b0dTcM\nBoPhmPHmyhJSYsM4Jb/n5rBFZKVSavKRHm9S5hgMBsMA4mKrhMPxzPEb9mAwGAyGAYkRHoPBYDD0\nKkZ4DAaDwdCrGOExGAwGQ69ihMdgMBgMvYoRHoPBYDD0KkZ4DAaDwdCrGOExGAwGQ69yQmcuEJFy\ndNqdgUYSUNHXnehBzPj6N2Z8/Z/hSqmYIz34hM5coJQakHURRGTF0aSzON4x4+vfmPH1f0TkqHKN\nGVebwWAwGHoVIzwGg8Fg6FWM8AxMnu3rDvQwZnz9GzO+/s9RjfGEDi4wGAwGQ+9jLB6DwWAw9CpG\neAwGg8HQqxjh6eeIyA4RWSMihXaIo4gkiMjHIrLF+hvf1/3sLiLygoiUichax7aA4xGRu0WkSEQ2\nici5fdPrwyPAGH8tIrut61goIuc59vWbMYrIYBH5TETWi8g6EfmZtX3AXMNOxjhQrmG4iHwjIqus\n8f3G2n7srqFSyjz68QPYAST5bPtP4JfW818CD/V1Pw9jPKcCE4G1XY0HGAWsAsKAbGArENTXYzjC\nMf4a+IWftv1qjMAgYKL1PAbYbI1hwFzDTsY4UK6hANHW8xDga2DasbyGxuIZmMwFXrSevwjM68O+\nHBZKqcXAfp/NgcYzF3hNKdWklNoOFAFTeqWjR0GAMQaiX41RKbVXKfWt9bwG2ABkMICuYSdjDES/\nGqPS1FovQ6yH4hheQyM8/R8FfCIiK0XkJmtbqlJqr/V8H5DaN107ZgQaTwZQ7GhXQuc3gOOdn4rI\nassVZ7sx+u0YRWQoMAH9i3lAXkOfMcIAuYYiEiQihUAZ8LFS6pheQyM8/Z+ZSqnxwBzgVhE51blT\naVt4wMTMD7TxOHgKyAHGA3uBR/u2O0eHiEQDbwK3K6UOOvcNlGvoZ4wD5hoqpdqs+0omMEVERvvs\nP6praISnn6OU2m39LQPeQpu4pSIyCMD6W9Z3PTwmBBrPbmCwo12mta3foZQqtb7s7cBzeF0V/W6M\nIhKCviG/opT6u7V5QF1Df2McSNfQRilVBXwGzOYYXkMjPP0YEYkSkRj7OXAOsBZYCFxjNbsGeKdv\nenjMCDSehcDlIhImItlAPvBNH/TvqLG/0BYXoq8j9LMxiogAzwMblFKPOXYNmGsYaIwD6Bomi0ic\n9TwCOBvYyLG8hn0dQWEeRxV9koOOJlkFrAPusbYnAouALcAnQEJf9/UwxvQq2k3RgvYV39DZeIB7\n0FE0m4A5fd3/oxjjy8AaYLX1RR7UH8cIzES7YFYDhdbjvIF0DTsZ40C5hmOB76xxrAV+ZW0/ZtfQ\npMwxGAwGQ69iXG0Gg8Fg6FWM8BgMBoOhVzHCYzAYDIZexQiPwWAwGHoVIzwGg8Fg6FWM8BgM/RQR\nOV1E3uvrfhgMh4sRHoPBYDD0KkZ4DIYeRkR+YNU3KRSRZ6wEjLUi8rhV72SRiCRbbceLyFdWosm3\n7ESTIpInIp9YNVK+FZFc6/TRIvKGiGwUkVesVfUGw3GNER6DoQcRkZHAZcAMpZMutgFXAVHACqVU\nAfAFcL91yEvAXUqpsehV8Pb2V4AnlFLjgOnozAegMyPfjq6JkgPM6PFBGQxHSXBfd8BgGOCcBUwC\nllvGSAQ6uWI78Derzf8AfxcRNxCnlPrC2v4i8LqVjy9DKfUWgFKqEcA63zdKqRLrdSEwFFjS88My\nGI4cIzwGQ88iwItKqbs7bBS5z6fdkeauanI8b8N8pw39AONqMxh6lkXAfBFJAU/d+iHo7958q82V\nwBKlVDVwQEROsbb/EPhC6SqXJSIyzzpHmIhE9uooDIZjiPl1ZDD0IEqp9SJyL/CRiLjQGalvBerQ\nBbbuRbveLrMOuQZ42hKWbcB11vYfAs+IyAPWOS7pxWEYDMcUk53aYOgDRKRWKRXd1/0wGPoC42oz\nGAwGQ69iLB6DwWAw9CrG4jEYDAZDr2KEx2AwGAy9ihEeg8FgMPQqRngMBoPB0KsY4TEYDAZDr/L/\nAYt5FLGbbRxOAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f925b79d240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "axes = plt.gca()\n",
    "\n",
    "# set x label range, starting with 16\n",
    "r = np.arange(16,288)\n",
    "\n",
    "# list all data in history\n",
    "print(history.history.keys())\n",
    "# summarize history for accuracy\n",
    "plt.plot(r, history.history['acc'])\n",
    "plt.plot(r, history.history['val_acc'])\n",
    "plt.title(\"model {} accuracy\".format(i))\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "#plt.xlim(32,224)\n",
    "plt.show()\n",
    "\n",
    "# summarize history for loss\n",
    "plt.plot(r, history.history['loss'])\n",
    "plt.plot(r, history.history['val_loss'])\n",
    "plt.title(\"model {} loss\".format(i))\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "#plt.xlim(32,224)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Try evaluation only with best performing models\n",
    "Most of the time, this does not improve prediction accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_model_numbers = [2,3,7,8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_models = []\n",
    "\n",
    "for i in best_model_numbers:\n",
    "    best_models.append(models[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_preds = np.stack([m.predict(x_test, batch_size=eval_batch_size) for m in best_models])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "avg_preds = all_preds.mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.17250180244445801"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# display error rate in percent\n",
    "(1 - keras.metrics.categorical_accuracy(y_test, avg_preds).eval().mean()) * 100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train single model some more"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Select a single model, train for a few epochs, and check accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# which model to train\n",
    "i = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# reload weights in case of overfitting\n",
    "# models[i].load_weights(\"weights/weights_emnist_model_0.2475_\"+str(i)+'.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "models[i].optimizer.lr = 0.0001\n",
    "history = models[i].fit_generator(batches, steps_per_epoch=steps_per_epoch, epochs=4,\n",
    "                   validation_data=test_batches, validation_steps=validation_steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# calculate accuracy of single model\n",
    "pred = np.array(models[i].predict(x_test, batch_size=eval_batch_size))\n",
    "print(i, (1 - keras.metrics.categorical_accuracy(y_test, pred).eval().mean()) * 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# models[i].save_weights(\"weights/weights_emnist_model_0.2475_\"+str(i)+'.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualize model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lambda_1 (Lambda)            (None, 1, 28, 28)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 32, 26, 26)        320       \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_1 (LeakyReLU)    (None, 32, 26, 26)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_1 (Batch (None, 32, 26, 26)        128       \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 32, 24, 24)        9248      \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_2 (LeakyReLU)    (None, 32, 24, 24)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 32, 12, 12)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_2 (Batch (None, 32, 12, 12)        128       \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 64, 10, 10)        18496     \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_3 (LeakyReLU)    (None, 64, 10, 10)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_3 (Batch (None, 64, 10, 10)        256       \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 64, 8, 8)          36928     \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_4 (LeakyReLU)    (None, 64, 8, 8)          0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 64, 4, 4)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 1024)              0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_4 (Batch (None, 1024)              4096      \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 512)               524800    \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_5 (LeakyReLU)    (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_5 (Batch (None, 512)               2048      \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                5130      \n",
      "=================================================================\n",
      "Total params: 601,578\n",
      "Trainable params: 598,250\n",
      "Non-trainable params: 3,328\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "models[0].summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"1578pt\" viewBox=\"0.00 0.00 268.00 1578.00\" width=\"268pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 1574)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" points=\"-4,4 -4,-1574 264,-1574 264,4 -4,4\" stroke=\"none\"/>\n",
       "<!-- 140396960945096 -->\n",
       "<g class=\"node\" id=\"node1\"><title>140396960945096</title>\n",
       "<polygon fill=\"none\" points=\"43.5,-1533.5 43.5,-1569.5 216.5,-1569.5 216.5,-1533.5 43.5,-1533.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1547.8\">lambda_1_input: InputLayer</text>\n",
       "</g>\n",
       "<!-- 140398320237592 -->\n",
       "<g class=\"node\" id=\"node2\"><title>140398320237592</title>\n",
       "<polygon fill=\"none\" points=\"68.5,-1460.5 68.5,-1496.5 191.5,-1496.5 191.5,-1460.5 68.5,-1460.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1474.8\">lambda_1: Lambda</text>\n",
       "</g>\n",
       "<!-- 140396960945096&#45;&gt;140398320237592 -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>140396960945096-&gt;140398320237592</title>\n",
       "<path d=\"M130,-1533.31C130,-1525.29 130,-1515.55 130,-1506.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1506.53 130,-1496.53 126.5,-1506.53 133.5,-1506.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396961092216 -->\n",
       "<g class=\"node\" id=\"node3\"><title>140396961092216</title>\n",
       "<polygon fill=\"none\" points=\"68.5,-1387.5 68.5,-1423.5 191.5,-1423.5 191.5,-1387.5 68.5,-1387.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1401.8\">conv2d_1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 140398320237592&#45;&gt;140396961092216 -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>140398320237592-&gt;140396961092216</title>\n",
       "<path d=\"M130,-1460.31C130,-1452.29 130,-1442.55 130,-1433.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1433.53 130,-1423.53 126.5,-1433.53 133.5,-1433.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396961758904 -->\n",
       "<g class=\"node\" id=\"node4\"><title>140396961758904</title>\n",
       "<polygon fill=\"none\" points=\"46,-1314.5 46,-1350.5 214,-1350.5 214,-1314.5 46,-1314.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1328.8\">leaky_re_lu_1: LeakyReLU</text>\n",
       "</g>\n",
       "<!-- 140396961092216&#45;&gt;140396961758904 -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>140396961092216-&gt;140396961758904</title>\n",
       "<path d=\"M130,-1387.31C130,-1379.29 130,-1369.55 130,-1360.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1360.53 130,-1350.53 126.5,-1360.53 133.5,-1360.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320291968 -->\n",
       "<g class=\"node\" id=\"node5\"><title>140398320291968</title>\n",
       "<polygon fill=\"none\" points=\"0,-1241.5 0,-1277.5 260,-1277.5 260,-1241.5 0,-1241.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1255.8\">batch_normalization_1: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 140396961758904&#45;&gt;140398320291968 -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>140396961758904-&gt;140398320291968</title>\n",
       "<path d=\"M130,-1314.31C130,-1306.29 130,-1296.55 130,-1287.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1287.53 130,-1277.53 126.5,-1287.53 133.5,-1287.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320294208 -->\n",
       "<g class=\"node\" id=\"node6\"><title>140398320294208</title>\n",
       "<polygon fill=\"none\" points=\"68.5,-1168.5 68.5,-1204.5 191.5,-1204.5 191.5,-1168.5 68.5,-1168.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1182.8\">conv2d_2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 140398320291968&#45;&gt;140398320294208 -->\n",
       "<g class=\"edge\" id=\"edge5\"><title>140398320291968-&gt;140398320294208</title>\n",
       "<path d=\"M130,-1241.31C130,-1233.29 130,-1223.55 130,-1214.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1214.53 130,-1204.53 126.5,-1214.53 133.5,-1214.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320294880 -->\n",
       "<g class=\"node\" id=\"node7\"><title>140398320294880</title>\n",
       "<polygon fill=\"none\" points=\"46,-1095.5 46,-1131.5 214,-1131.5 214,-1095.5 46,-1095.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1109.8\">leaky_re_lu_2: LeakyReLU</text>\n",
       "</g>\n",
       "<!-- 140398320294208&#45;&gt;140398320294880 -->\n",
       "<g class=\"edge\" id=\"edge6\"><title>140398320294208-&gt;140398320294880</title>\n",
       "<path d=\"M130,-1168.31C130,-1160.29 130,-1150.55 130,-1141.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1141.53 130,-1131.53 126.5,-1141.53 133.5,-1141.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320295720 -->\n",
       "<g class=\"node\" id=\"node8\"><title>140398320295720</title>\n",
       "<polygon fill=\"none\" points=\"26.5,-1022.5 26.5,-1058.5 233.5,-1058.5 233.5,-1022.5 26.5,-1022.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-1036.8\">max_pooling2d_1: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 140398320294880&#45;&gt;140398320295720 -->\n",
       "<g class=\"edge\" id=\"edge7\"><title>140398320294880-&gt;140398320295720</title>\n",
       "<path d=\"M130,-1095.31C130,-1087.29 130,-1077.55 130,-1068.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-1068.53 130,-1058.53 126.5,-1068.53 133.5,-1068.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320293536 -->\n",
       "<g class=\"node\" id=\"node9\"><title>140398320293536</title>\n",
       "<polygon fill=\"none\" points=\"0,-949.5 0,-985.5 260,-985.5 260,-949.5 0,-949.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-963.8\">batch_normalization_2: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 140398320295720&#45;&gt;140398320293536 -->\n",
       "<g class=\"edge\" id=\"edge8\"><title>140398320295720-&gt;140398320293536</title>\n",
       "<path d=\"M130,-1022.31C130,-1014.29 130,-1004.55 130,-995.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-995.529 130,-985.529 126.5,-995.529 133.5,-995.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140398320643488 -->\n",
       "<g class=\"node\" id=\"node10\"><title>140398320643488</title>\n",
       "<polygon fill=\"none\" points=\"68.5,-876.5 68.5,-912.5 191.5,-912.5 191.5,-876.5 68.5,-876.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-890.8\">conv2d_3: Conv2D</text>\n",
       "</g>\n",
       "<!-- 140398320293536&#45;&gt;140398320643488 -->\n",
       "<g class=\"edge\" id=\"edge9\"><title>140398320293536-&gt;140398320643488</title>\n",
       "<path d=\"M130,-949.313C130,-941.289 130,-931.547 130,-922.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-922.529 130,-912.529 126.5,-922.529 133.5,-922.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396962060048 -->\n",
       "<g class=\"node\" id=\"node11\"><title>140396962060048</title>\n",
       "<polygon fill=\"none\" points=\"46,-803.5 46,-839.5 214,-839.5 214,-803.5 46,-803.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-817.8\">leaky_re_lu_3: LeakyReLU</text>\n",
       "</g>\n",
       "<!-- 140398320643488&#45;&gt;140396962060048 -->\n",
       "<g class=\"edge\" id=\"edge10\"><title>140398320643488-&gt;140396962060048</title>\n",
       "<path d=\"M130,-876.313C130,-868.289 130,-858.547 130,-849.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-849.529 130,-839.529 126.5,-849.529 133.5,-849.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960493752 -->\n",
       "<g class=\"node\" id=\"node12\"><title>140396960493752</title>\n",
       "<polygon fill=\"none\" points=\"0,-730.5 0,-766.5 260,-766.5 260,-730.5 0,-730.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-744.8\">batch_normalization_3: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 140396962060048&#45;&gt;140396960493752 -->\n",
       "<g class=\"edge\" id=\"edge11\"><title>140396962060048-&gt;140396960493752</title>\n",
       "<path d=\"M130,-803.313C130,-795.289 130,-785.547 130,-776.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-776.529 130,-766.529 126.5,-776.529 133.5,-776.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960527696 -->\n",
       "<g class=\"node\" id=\"node13\"><title>140396960527696</title>\n",
       "<polygon fill=\"none\" points=\"68.5,-657.5 68.5,-693.5 191.5,-693.5 191.5,-657.5 68.5,-657.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-671.8\">conv2d_4: Conv2D</text>\n",
       "</g>\n",
       "<!-- 140396960493752&#45;&gt;140396960527696 -->\n",
       "<g class=\"edge\" id=\"edge12\"><title>140396960493752-&gt;140396960527696</title>\n",
       "<path d=\"M130,-730.313C130,-722.289 130,-712.547 130,-703.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-703.529 130,-693.529 126.5,-703.529 133.5,-703.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396961435944 -->\n",
       "<g class=\"node\" id=\"node14\"><title>140396961435944</title>\n",
       "<polygon fill=\"none\" points=\"46,-584.5 46,-620.5 214,-620.5 214,-584.5 46,-584.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-598.8\">leaky_re_lu_4: LeakyReLU</text>\n",
       "</g>\n",
       "<!-- 140396960527696&#45;&gt;140396961435944 -->\n",
       "<g class=\"edge\" id=\"edge13\"><title>140396960527696-&gt;140396961435944</title>\n",
       "<path d=\"M130,-657.313C130,-649.289 130,-639.547 130,-630.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-630.529 130,-620.529 126.5,-630.529 133.5,-630.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960985040 -->\n",
       "<g class=\"node\" id=\"node15\"><title>140396960985040</title>\n",
       "<polygon fill=\"none\" points=\"26.5,-511.5 26.5,-547.5 233.5,-547.5 233.5,-511.5 26.5,-511.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-525.8\">max_pooling2d_2: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 140396961435944&#45;&gt;140396960985040 -->\n",
       "<g class=\"edge\" id=\"edge14\"><title>140396961435944-&gt;140396960985040</title>\n",
       "<path d=\"M130,-584.313C130,-576.289 130,-566.547 130,-557.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-557.529 130,-547.529 126.5,-557.529 133.5,-557.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960619208 -->\n",
       "<g class=\"node\" id=\"node16\"><title>140396960619208</title>\n",
       "<polygon fill=\"none\" points=\"75,-438.5 75,-474.5 185,-474.5 185,-438.5 75,-438.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-452.8\">flatten_1: Flatten</text>\n",
       "</g>\n",
       "<!-- 140396960985040&#45;&gt;140396960619208 -->\n",
       "<g class=\"edge\" id=\"edge15\"><title>140396960985040-&gt;140396960619208</title>\n",
       "<path d=\"M130,-511.313C130,-503.289 130,-493.547 130,-484.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-484.529 130,-474.529 126.5,-484.529 133.5,-484.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960619320 -->\n",
       "<g class=\"node\" id=\"node17\"><title>140396960619320</title>\n",
       "<polygon fill=\"none\" points=\"0,-365.5 0,-401.5 260,-401.5 260,-365.5 0,-365.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-379.8\">batch_normalization_4: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 140396960619208&#45;&gt;140396960619320 -->\n",
       "<g class=\"edge\" id=\"edge16\"><title>140396960619208-&gt;140396960619320</title>\n",
       "<path d=\"M130,-438.313C130,-430.289 130,-420.547 130,-411.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-411.529 130,-401.529 126.5,-411.529 133.5,-411.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960617920 -->\n",
       "<g class=\"node\" id=\"node18\"><title>140396960617920</title>\n",
       "<polygon fill=\"none\" points=\"79,-292.5 79,-328.5 181,-328.5 181,-292.5 79,-292.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-306.8\">dense_1: Dense</text>\n",
       "</g>\n",
       "<!-- 140396960619320&#45;&gt;140396960617920 -->\n",
       "<g class=\"edge\" id=\"edge17\"><title>140396960619320-&gt;140396960617920</title>\n",
       "<path d=\"M130,-365.313C130,-357.289 130,-347.547 130,-338.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-338.529 130,-328.529 126.5,-338.529 133.5,-338.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960618928 -->\n",
       "<g class=\"node\" id=\"node19\"><title>140396960618928</title>\n",
       "<polygon fill=\"none\" points=\"46,-219.5 46,-255.5 214,-255.5 214,-219.5 46,-219.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-233.8\">leaky_re_lu_5: LeakyReLU</text>\n",
       "</g>\n",
       "<!-- 140396960617920&#45;&gt;140396960618928 -->\n",
       "<g class=\"edge\" id=\"edge18\"><title>140396960617920-&gt;140396960618928</title>\n",
       "<path d=\"M130,-292.313C130,-284.289 130,-274.547 130,-265.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-265.529 130,-255.529 126.5,-265.529 133.5,-265.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960619264 -->\n",
       "<g class=\"node\" id=\"node20\"><title>140396960619264</title>\n",
       "<polygon fill=\"none\" points=\"0,-146.5 0,-182.5 260,-182.5 260,-146.5 0,-146.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-160.8\">batch_normalization_5: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 140396960618928&#45;&gt;140396960619264 -->\n",
       "<g class=\"edge\" id=\"edge19\"><title>140396960618928-&gt;140396960619264</title>\n",
       "<path d=\"M130,-219.313C130,-211.289 130,-201.547 130,-192.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-192.529 130,-182.529 126.5,-192.529 133.5,-192.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960618536 -->\n",
       "<g class=\"node\" id=\"node21\"><title>140396960618536</title>\n",
       "<polygon fill=\"none\" points=\"67.5,-73.5 67.5,-109.5 192.5,-109.5 192.5,-73.5 67.5,-73.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-87.8\">dropout_1: Dropout</text>\n",
       "</g>\n",
       "<!-- 140396960619264&#45;&gt;140396960618536 -->\n",
       "<g class=\"edge\" id=\"edge20\"><title>140396960619264-&gt;140396960618536</title>\n",
       "<path d=\"M130,-146.313C130,-138.289 130,-128.547 130,-119.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-119.529 130,-109.529 126.5,-119.529 133.5,-119.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 140396960888368 -->\n",
       "<g class=\"node\" id=\"node22\"><title>140396960888368</title>\n",
       "<polygon fill=\"none\" points=\"79,-0.5 79,-36.5 181,-36.5 181,-0.5 79,-0.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"130\" y=\"-14.8\">dense_2: Dense</text>\n",
       "</g>\n",
       "<!-- 140396960618536&#45;&gt;140396960888368 -->\n",
       "<g class=\"edge\" id=\"edge21\"><title>140396960618536-&gt;140396960888368</title>\n",
       "<path d=\"M130,-73.3129C130,-65.2895 130,-55.5475 130,-46.5691\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"133.5,-46.5288 130,-36.5288 126.5,-46.5289 133.5,-46.5288\" stroke=\"black\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fix for python3\n",
    "import pydotplus as pydot\n",
    "\n",
    "from keras.utils import plot_model\n",
    "plot_model(models[0], to_file='model.png')\n",
    "\n",
    "# visualize our model\n",
    "from IPython.display import SVG\n",
    "from keras.utils.vis_utils import model_to_dot\n",
    "SVG(model_to_dot(models[0]).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualize Predictions\n",
    "Here we compute the predictions from a single model or ensemble to visualize them later. As well as looking at the overall metrics, it's also a good idea to look at examples of:\n",
    "1. A few correct labels at random\n",
    "2. A few incorrect labels at random"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute predictions for best performing single model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = create_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load saved weights\n",
    "m.load_weights(\"dropout_0.2/weights/258epochs_weights_model_7.pkl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "37888/40000 [===========================>..] - ETA: 0s"
     ]
    }
   ],
   "source": [
    "# predict classes\n",
    "preds = m.predict_classes(x_test, batch_size=eval_batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40000,)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Or compute predictions for ensemble\n",
    "Load model weights and compute predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "# specify which weights to load\n",
    "weights_epoch = 267\n",
    "dropout = \"0.2\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "models = []\n",
    "\n",
    "for i in range(10):\n",
    "    m = create_model()\n",
    "    m.load_weights('dropout_{}/weights/{:3d}epochs_weights_model_{}.pkl'.format(dropout, weights_epoch, i))\n",
    "    models.append(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_preds = np.stack([m.predict(x_test, batch_size=eval_batch_size) for m in models])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "avg_preds = all_preds.mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "# decode a one hot encoded string\n",
    "def one_hot_decode(encoded_seq):\n",
    "    return [np.argmax(vector) for vector in encoded_seq]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = np.array(one_hot_decode(avg_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40000,)"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.16000270843505859"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# display error rate in percent to check we have loaded the weights we wanted\n",
    "(1 - keras.metrics.categorical_accuracy(y_test, avg_preds).eval().mean()) * 100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize computed predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40000,)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "correct = np.where(preds==test_labels)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "incorrect = np.where(preds!=test_labels)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(39936,)"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correct.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64,)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "incorrect.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "# number of images to view\n",
    "n_view = incorrect.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy.random import permutation\n",
    "\n",
    "# get a random selection of correctly predicted images\n",
    "idx = permutation(correct)[:n_view]\n",
    "idx_incorrect = permutation(incorrect)[:n_view]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds_correct = preds[idx]\n",
    "preds_incorrect = preds[idx_incorrect]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "imgs = x_test[idx]\n",
    "imgs_incorrect = x_test[idx_incorrect]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64, 1, 28, 28)"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imgs.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## A few correct labels at random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f487e33c358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(12,4))\n",
    "\n",
    "for i in range(30):\n",
    "    sp = f.add_subplot(3, 10, i+1)\n",
    "    sp.axis('Off')\n",
    "    sp.set_title(preds_correct[i])\n",
    "    img = imgs[i].astype('float32')\n",
    "    plt.imshow(img[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 7, 7, 2, 3, 7, 0, 3, 1, 5, 4, 2, 1, 0, 1, 6, 4, 4, 6, 6, 0, 8, 7,\n",
       "       4, 1, 1, 0, 3, 1, 1, 6, 2, 2, 2, 4, 0, 4, 0, 7, 7, 7, 7, 9, 9, 5, 4,\n",
       "       4, 9, 1, 1, 5, 9, 2, 0, 4, 5, 6, 1, 6, 1, 5, 2, 3, 8], dtype=uint8)"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ground truth labels\n",
    "test_labels[idx]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## All incorrect labels\n",
    "Display all predictions not matching ground truth. Model predictions are displayed on top left of the image, ground truth labels are displayed on top right of the image. There are clearly some labeling errors in EMNIST Digits, limiting our overall prediction accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4886dbcc18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(12,10))\n",
    "\n",
    "for i in range(n_view): \n",
    "    sp = f.add_subplot(7, 10, i+1)\n",
    "    sp.axis('Off')\n",
    "    sp.set_title(\"{} | {}\".format(preds_incorrect[i],test_labels[idx_incorrect][i]))\n",
    "    img = imgs_incorrect[i].astype('float32')\n",
    "    plt.imshow(img[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 7, 4, 2, 5, 2, 8, 7, 3, 7, 4, 4, 4, 3, 8, 5, 8, 0, 1, 9, 5, 9, 8,\n",
       "       9, 4, 5, 5, 9, 2, 4, 6, 1, 1, 4, 8, 8, 6, 9, 4, 0, 3, 9, 5, 4, 4, 9,\n",
       "       6, 1, 9, 4, 4, 4, 7, 3, 7, 2, 3, 3, 3, 4, 8, 1, 3, 0], dtype=uint8)"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ground truth labels\n",
    "test_labels[idx_incorrect]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Print confusion matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot unnormalized and normalized confusion matrices. Code adapted from \n",
    "http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# calculate confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cnf_matrix = confusion_matrix(test_labels, preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[3997    0    0    0    0    1    1    0    0    1]\n",
      " [   0 3994    1    0    1    0    0    3    1    0]\n",
      " [   2    1 3996    0    0    0    0    1    0    0]\n",
      " [   0    0    3 3992    0    3    0    0    2    0]\n",
      " [   1    0    0    0 3986    0    1    2    1    9]\n",
      " [   0    1    0    3    0 3994    1    0    1    0]\n",
      " [   2    0    0    0    1    0 3997    0    0    0]\n",
      " [   0    1    2    0    1    0    0 3995    0    1]\n",
      " [   0    0    1    2    1    0    2    0 3993    1]\n",
      " [   0    0    0    0    4    0    0    2    2 3992]]\n",
      "Normalized confusion matrix\n",
      "[[  9.99e-01   0.00e+00   0.00e+00   0.00e+00   0.00e+00   2.50e-04\n",
      "    2.50e-04   0.00e+00   0.00e+00   2.50e-04]\n",
      " [  0.00e+00   9.99e-01   2.50e-04   0.00e+00   2.50e-04   0.00e+00\n",
      "    0.00e+00   7.50e-04   2.50e-04   0.00e+00]\n",
      " [  5.00e-04   2.50e-04   9.99e-01   0.00e+00   0.00e+00   0.00e+00\n",
      "    0.00e+00   2.50e-04   0.00e+00   0.00e+00]\n",
      " [  0.00e+00   0.00e+00   7.50e-04   9.98e-01   0.00e+00   7.50e-04\n",
      "    0.00e+00   0.00e+00   5.00e-04   0.00e+00]\n",
      " [  2.50e-04   0.00e+00   0.00e+00   0.00e+00   9.97e-01   0.00e+00\n",
      "    2.50e-04   5.00e-04   2.50e-04   2.25e-03]\n",
      " [  0.00e+00   2.50e-04   0.00e+00   7.50e-04   0.00e+00   9.99e-01\n",
      "    2.50e-04   0.00e+00   2.50e-04   0.00e+00]\n",
      " [  5.00e-04   0.00e+00   0.00e+00   0.00e+00   2.50e-04   0.00e+00\n",
      "    9.99e-01   0.00e+00   0.00e+00   0.00e+00]\n",
      " [  0.00e+00   2.50e-04   5.00e-04   0.00e+00   2.50e-04   0.00e+00\n",
      "    0.00e+00   9.99e-01   0.00e+00   2.50e-04]\n",
      " [  0.00e+00   0.00e+00   2.50e-04   5.00e-04   2.50e-04   0.00e+00\n",
      "    5.00e-04   0.00e+00   9.98e-01   2.50e-04]\n",
      " [  0.00e+00   0.00e+00   0.00e+00   0.00e+00   1.00e-03   0.00e+00\n",
      "    0.00e+00   5.00e-04   5.00e-04   9.98e-01]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc789c7ec50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcYAAAFuCAYAAAAbLbiqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xt8lFWe5/HPjwQ60HIbuaYCcoltNJKOIQHGe4+NoIDuOrrSeGdGcTe0ra7rtL3b0jtMD2076jqjLirdtg5NizjDhJvgTO/giCPEEJtLdGiJgKTCRRRCIyAh/PaPKkKFS1IEnlRV6vv29bzM8zznnN85T4o6Oc/tmLsjIiIiER0SXQEREZFkoo5RREQkhjpGERGRGOoYRUREYqhjFBERiaGOUUREJIY6RjltZvYTM5sd/Xmgme0zs4yzHGOzmX33bJYZR8z/amY7ou059wzK2WdmQ85m3RLFzKrM7OpE10OkLaljTELRTmGnmX0zZtufm9nyBFbrpNz9M3c/x90bEl2XM2FmHYGngWuj7fmitWVF83969mp39pnZr8zsr1pK5+757r68DaokkjTUMSavDOAHZ1qIRej33LK+QBZQleiKJAMzy0x0HUQSRV+YyetJ4BEz63GynWZ2qZl9YGZ10f9fGrNvuZn91MzeA/YDQ6Lb/srM/j16qm+hmZ1rZr82s73RMgbFlPGsmW2N7lttZlecoh6DzMzNLNPM/jha9tHloJltjqbrYGY/NLNqM/vCzN4wsz+KKecOM9sS3fc/mzswZtbZzJ6Kpq8zsxVm1jm674bo6b890TZfGJNvs5k9YmZro/nmmlmWmX0L2BBNtsfM/l9su447rn8e/TnXzN6JlrPLzObGpHMzy43+3N3MXjOzz6P1/V9H/1Axs7ujdf8bM9ttZpvM7Lpm2r3ZzP5HtP5fmdkvzKyvmb1lZn8ws38xs54x6eeZ2fZoHf/NzPKj2+8DbgMePfpZiCn/L8xsLfBV9HfaeErbzJaY2VMx5b9uZr9s7nclkorUMSavCmA58MjxO6IdymLgb4FziZwCXGxNr4vdAdwHdAW2RLdNjG4PAUOB94FXgD8CPgamxeT/ACiM7psDzDOzrOYq7O7vR08jngP0BFYBv4nu/j7wn4CrgGxgN/B8tD0XAf83WrfsaJtymgn1N8Bw4NJo/R4FjkQ7uN8ADwK9gSXAQjPrFJP3vwBjgcFAAXC3u/8eyI/u7+Huf9JcO6OmA29H25kD/N0p0v0d0B0YEm37ncA9MftHEumUewE/B35hZtZM3D8FRgPfAiYAbwE/ira3A/BATNq3gPOBPkAl8GsAd38p+vPPo7+vCTF5vgeMI3IcDh8XezJwh5n9iZndBozgLJzVEEk26hiT2+PA982s93HbxwGfuPvfu/thd/8N8B9EviiP+pW7V0X310e3veLu1e5eR+RLs9rd/yX6BTgPuORoZnef7e5fRPM/BXwDuOA06v63wB+Ao6O/+4H/6e417v418BPg5uiI7GZgkbv/W3Tfj4EjJys0OtqaDPzA3cPu3uDu/x7Ndyuw2N3/OdrmvwE6E+lAG+vl7rXu/iWwkEjn3xr1wHlAtrsfdPcVJ6lrBpE/Rh5z9z+4+2bgKSJ/ABy1xd1fjl6jfRXoT+S07qn8nbvvcPcw8C6wyt0/dPeDwHya/g5/GY179Hh/28y6t9Cuv3X3re5+4Pgd7r4d+K/Rej4L3Onuf2ihPJGUo44xibn7emAR8MPjdmVzbBR41BYiI8Gjtp6kyB0xPx84yfo5R1eipxw/jp6G20Nk1NMrnnqb2RTgamCSux/t4M4D5kdPce4hMkJtINIJZMfW192/Ak5180svItcCq0+yr8lxicbeStPjsj3m5/3EtPk0PQoYUB49dTv5FHXtSNPf1fG/p8b6uPv+6I/N1Smu36GZZZjZz6KnrvcCm2Pq1JyTfW5iLSRy/XvDyf4YEGkP1DEmv2nAvTT9Mq0l0tHEGgiEY9ZbPW1K9Hrio0ROO/Z09x5AHZGOIJ6804Eb3X1vzK6twHXu3iNmyYqOfLYBA2LK6ELkdOrJ7AIOEjkVfLwmxyV6SnIATY9LvL6K/r9LzLZ+R39w9+3ufq+7ZwNTgBeOXlc8rq5HR5ZHHf97Csok4Ebgu0T+qBkU3X70d3iqz0dLn5ufEvmjpr+Zfe8M6yiSlNQxJjl33wjMpem1oyXAt8xsUvQGiVuBi4iMLs+GrsBh4HMg08weB7q1lMnMBgBvEDnF9vvjds8Efmpm50XT9jazG6P73gTGm9nl0euBf8kpPpvRUeAvgafNLDs6MvpjM/tGNPY4M7vGIo9f/Hfga+DfT6v1kTifE+nAbo/GmExMZ2xmt5jZ0eugu4l0KEeOK6MhWqefmlnXaNsfBmafbn1aoSuRtn9BpHP/6+P27yBy3TNuZnYlkeujdwJ3AX9nZqHmc4mkHnWMqeEvgcZnGqPP2I0n8sX/BZHR3Xh333WW4i0DlgK/J3Lq7yAtn2IDuIbIqdE37didqUcff3gWWAC8bWZ/AFYSufEEd68CSonc5LONSEdT00ycR4B1RG4Q+hJ4Aujg7huA24nc8LKLyDXXCe5+KM52H+9e4H8QOcb5NO1gS4BVZrYv2q4fnOLZxe8TGX1+CqyItrEt7uR8jcjvLgx8ROR4x/oFcFH01PY/tVSYmXWLljk1em333WgZr7Rws5BIyjFNVCwiInKMRowiIiIx1DGKiIjEUMcoIiISQx2jiIhIjKR6UbBldnbr1LXN415y4cA2jyki7ceWLZvZtWuX7s49iYxu57kfPuFFSnHzA58vc/exZ7FKLUqujrFTV75xwX9p87jvrXquzWOKSPtx2cjiRFchafnhA2f0vX7wd8/H9catsympOkYREWlvDFJs5jt1jCIiEhwDUuwdEOoYRUQkWCk2Ykyt2oqIiARMI0YREQmWTqWKiIgclXo33yR9bWdOu40tv51BxbwfnTLNU4/ezPqyaZTPfYzCvJzG7aMvvZA183/M+rJpPHLP6NOO/faypRTkX0B+Xi5P/vxnJ+x3dx5+8AHy83IpuaSADysr486ruMkRO93iJjJ2OrZZosxavySCuyfNYp17e1ZhaZPlmslP+6iJM3z9J+ET9mUVlvqNU5/3pSvWe1ZhqV95x5NevnaTZxWWepeiqV792U7PG/e4dy1+wNds2OqFN00/aRkH6v2EZd/Bwz54yBD/aEO11331tQ8bVuCVa6qapJm/YLFfO2as7z90xJe/+74Xl4yIO++plnSLm45t1rFuf20uKhruif7+TNbFvtnXs0Y+2uoFqGjrOif9iPG9ymq+rNt/yv3jrypgzqJyAMrXbaZ7187069WNkosHUb11F5vDX1B/uIF5yyoZf3VB3HE/KC9n6NBcBg8ZQqdOnbjl1oksWljWJM2iBWVMuv1OzIyRo0ZRV7eHbdu2xZVXcRMfO93iqs1tH1sgcio1tUaMSd8xtiS7Tw9qtu9uXA/v2EN2nx5k9+lOzY7Y7bsJ9e4ed7m1tWFycgY0rodCOYTD4RbT1IbDceVV3MTHTre4iYydjm2W1BVox2hmY81sg5ltNLMfBhlLRESSlHVo/ZIAgd2VamYZwPPAaKAG+MDMFrj7R2czTu3OPeT069m4Hurbg9qde+iYmUFO39jtPQl/Xhd3udnZIWpqtjauh8M1hEKhFtNkh0LU19e3mFdxEx873eImMnY6tllipNjjGkF2xyOAje7+qbsfAl4HbjzbQRa/s45J40dEAg4bxN59B9i+ay8VVVvIHdib87LPpWNmBreMKWLx8rVxl1tcUsLGjZ+wedMmDh06xLy5rzNu/A1N0oybcANzZr+Gu7Nq5Uq6detO//7948qruImPnW5x1ea2jy3Q+LiGRowAhICtMes1wMjTLeTVGXdzxfDz6dXjHDYunc70mUvomJkBwKw3V7B0RRVjLs+nasE09h+sZ8pPZgPQ0HCEh554g4UvlJLRwXi1bCUff7o97riZmZk88+xzTBg3hoaGBu66ezIX5efz8oszAbh3yv2Mve56lr21hPy8XLp07sKLs15pNq/iJlfsdIurNrd9bCEl35Vq7h5MwWY3A2Pd/c+j63cAI9196nHp7gPuA6DjOcOz8u8KpD7N2f2Bpp0Skda7bGQxq1dXpNa3fxvp0DXbv3HJfa3Of/Dd/73a3dt0Xq8gx6lhYEDMek50WxPu/pK7F7t7sWV2DrA6IiIiLQvyVOoHwPlmNphIhzgRmBRgPBERSTqp90q4wDpGdz9sZlOBZUAG8Et3rwoqnoiIJKkOqXWWOdCXiLv7EmBJkDFERCSJGRoxioiINJFid6WqYxQRkQDpGqOIiEhTKTZiTK1uXEREJGAaMYqISLB0KlVERCQqgfMqtpY6RhERCZZGjCIiIjE0YhQRETlKj2uckUsuHMh7q9p+poueJVNbThQQzewhIpJckqpjFBGRdkinUkVERKL0rlQREZFYusYoIiLSlE6lioiIxEixEWNq1VZERCRgGjGKiEiwdCpVREQkylLv5pukr+3by5ZSkH8B+Xm5PPnzn52w3915+MEHyM/LpeSSAj6srIw7b3NmTruNLb+dQcW8H50yzVOP3sz6smmUz32Mwrycxu2jL72QNfN/zPqyaTxyz+jTihtPvYNqc6LiJjJ2usVNZOx0bLNEHX2ReGuWRHD3pFmKiob7gXpvXPYdPOyDhwzxjzZUe91XX/uwYQVeuaaqSZr5Cxb7tWPG+v5DR3z5u+97ccmIuPMeXbIKS09Yrpn8tI+aOMPXfxI+6f4bpz7vS1es96zCUr/yjie9fO0mzyos9S5FU736s52eN+5x71r8gK/ZsNULb5p+0jKyCktPqEtbtTlZ4qZjm3Ws21+bi4qGe6K/P5N1sR7neeebftHqBaho6zon9Yjxg/Jyhg7NZfCQIXTq1Ilbbp3IooVlTdIsWlDGpNvvxMwYOWoUdXV72LZtW1x5m/NeZTVf1u0/5f7xVxUwZ1E5AOXrNtO9a2f69epGycWDqN66i83hL6g/3MC8ZZWMv7og6ducyGOdbm3WsU6PNkuEAWbW6iURkrpjrK0Nk5MzoHE9FMohHA63mKY2HI4r75nI7tODmu27G9fDO/aQ3acH2X26U7MjdvtuQr27x11uotqcyGOdbm3WsU6PNkuUneGSAIF1jGb2SzPbaWbrg4ohIiJytgV5V+qvgOeA11pbQHZ2iJqarY3r4XANoVCoxTTZoRD19fUt5j0TtTv3kNOvZ+N6qG8PanfuoWNmBjl9Y7f3JPx5XdzlJqrNiTzW6dZmHev0aLMclbhToq0V2IjR3f8N+PJMyiguKWHjxk/YvGkThw4dYt7c1xk3/oYmacZNuIE5s1/D3Vm1ciXdunWnf//+ceU9E4vfWcek8SMAGDFsEHv3HWD7rr1UVG0hd2Bvzss+l46ZGdwypojFy9cmfZsTeazTrc061unRZjkm1a4xJvw5RjO7D7gPYMDAgU32ZWZm8syzzzFh3BgaGhq46+7JXJSfz8svzgTg3in3M/a661n21hLy83Lp0rkLL856pdm88Xp1xt1cMfx8evU4h41LpzN95hI6ZmYAMOvNFSxdUcWYy/OpWjCN/QfrmfKT2QA0NBzhoSfeYOELpWR0MF4tW8nHn26PO26i2pzIY51ubdaxTo82yzGpNmI0dw+ucLNBwCJ3vzie9MOHF/t7qyoCq8+paKJiETkTl40sZvXqitT69m8jGX802M8Z85etzr/39TtXu3vxWaxSixI+YhQRkXYsgXeXtlZSP64hIiLS1oJ8XOM3wPvABWZWY2Z/FlQsERFJTkbrb7xpdzffuPv3gipbRERSR6rdfKNrjCIiEih1jCIiIjHUMYqIiByVgnelqmMUEZFApdqIUY9riIiIxNCIUUREAmN6ibiIiEhTQT7HaGZjzWyDmW00sx+eZH93M1toZmvMrMrM7mmpTHWMIiISrIAmKjazDOB54DrgIuB7ZnbRcclKgY/c/dvA1cBTZtapuXJ1KlVERIJjgd58MwLY6O6fApjZ68CNwEcxaRzoapFKnENkOsTDzRWqjpHEznCRqJk9NKuHiLSVM+wYe5lZ7LRLL7n7S9GfQ8DWmH01wMjj8j8HLABqga7Are5+pLmA6hhFRCSZ7TrDaafGAL8D/gQYCvyzmb3r7ntPlUHXGEVEJFAB3nwTBgbErOdEt8W6B/hHj9gIbALymitUHaOIiAQm4Nk1PgDON7PB0RtqJhI5bRrrM+AaADPrC1wAfNpcoTqVKiIiwQro3ht3P2xmU4FlQAbwS3evMrP7o/tnAtOBX5nZumhN/sLddzVXrjpGEREJTrB3peLuS4Alx22bGfNzLXDt6ZSpjlFERAKlN9+IiIiksKTvGN9etpSC/AvIz8vlyZ//7IT97s7DDz5Afl4uJZcU8GFlZdx5kzH2zGm3seW3M6iY96NTpnnq0ZtZXzaN8rmPUZiX07h99KUXsmb+j1lfNo1H7hkdd8x469zejnU6xk1k7HRss0QE+Uq4QLh70ixFRcP9QL03LvsOHvbBQ4b4Rxuqve6rr33YsAKvXFPVJM38BYv92jFjff+hI7783fe9uGRE3HmbW9oqdlZhaZPlmslP+6iJM3z9J+ET9mUVlvqNU5/3pSvWe1ZhqV95x5NevnaTZxWWepeiqV792U7PG/e4dy1+wNds2OqFN00/aRlZhaUJa28ij3W6x1Wbg4tdVDTcE/39maxLx95DPaf0n1q9ABVtXeekHjF+UF7O0KG5DB4yhE6dOnHLrRNZtLCsSZpFC8qYdPudmBkjR42irm4P27ZtiytvMsZ+r7KaL+v2n3L/+KsKmLOoHIDydZvp3rUz/Xp1o+TiQVRv3cXm8BfUH25g3rJKxl9dkPTtTWTsdIurNrd9bIlItRFjUneMtbVhcnKOPbsZCuUQDodbTFMbDseVN1ljNye7Tw9qtu9uXA/v2EN2nx5k9+lOzY7Y7bsJ9e4ed7npeKzTLW4iY6djmyXiTDrFdtcxmtkAM/tXM/soOtXHD4KKJSIiySvVOsYgH9c4DPx3d680s67AajP7Z3f/qKWMR2Vnh6ipOfZ+2HC4hlAo1GKa7FCI+vr6FvMma+zm1O7cQ06/no3rob49qN25h46ZGeT0jd3ek/DndXGXm47HOt3iJjJ2OrZZjtHjGlHuvs3dK6M//wH4mMib0ONWXFLCxo2fsHnTJg4dOsS8ua8zbvwNTdKMm3ADc2a/hruzauVKunXrTv/+/ePKm6yxm7P4nXVMGj8CgBHDBrF33wG279pLRdUWcgf25rzsc+mYmcEtY4pYvHxtSrQ3UbHTLa7anFz/liV5tckD/mY2CLgEWHWSffcB9wEMGDiwyb7MzEyeefY5JowbQ0NDA3fdPZmL8vN5+cXISw3unXI/Y6+7nmVvLSE/L5cunbvw4qxXms0br0TFfnXG3Vwx/Hx69TiHjUunM33mEjpmZgAw680VLF1RxZjL86laMI39B+uZ8pPZADQ0HOGhJ95g4QulZHQwXi1bycefbk/69iYydrrFVZvbPrZEpdaAEXP3YAOYnQO8A/zU3f+xubTDhxf7e6sqmkvS7mg+RpHUd9nIYlavrkixr/+28Y2+53votmdbnX/TM+NW+5lNO3XaAh0xmllH4B+AX7fUKYqISDsU8LtSgxBYx2iRI/EL4GN3fzqoOCIikrwMSLF+MdDnGC8D7gD+xMx+F12uDzCeiIgkndR7jjGwEaO7ryDlLrmKiEi607RTIiISqFQ7laqOUUREAqWbb0RERI4yjRhFREQaGdChQ2r1jOoYRUQkUKk2YkzqaadERETamkaMIiISKN18IyIicpRuvhERETkm8kq41OoZ1TEmWKJmuUjUrB6gmT1E0kviXu3WWuoYRUQkUCnWL6pjFBGRYKXaiFGPa4iIiMTQiFFERIKju1JFRESO0V2pIiIix0mxflEdo4iIBEsjRhERkRgp1i8m/12pby9bSkH+BeTn5fLkz392wn535+EHHyA/L5eSSwr4sLIy7rzJGjtRcWdOu40tv51BxbwfnTLNU4/ezPqyaZTPfYzCvJzG7aMvvZA183/M+rJpPHLP6NOKG0+929ux1uc6PdosKcrdk2YpKhruB+q9cdl38LAPHjLEP9pQ7XVffe3DhhV45ZqqJmnmL1js144Z6/sPHfHl777vxSUj4s7b3JKo2G0VN6uw9ITlmslP+6iJM3z9J+GT7r9x6vO+dMV6zyos9SvveNLL127yrMJS71I01as/2+l54x73rsUP+JoNW73wpuknLSOrsDTtjnWyxFWbg4tdVDTcE/39maxLl9C3fMRfL2/1AlS0dZ2TesT4QXk5Q4fmMnjIEDp16sQtt05k0cKyJmkWLShj0u13YmaMHDWKuro9bNu2La68yRg7kW1+r7KaL+v2n3L/+KsKmLOoHIDydZvp3rUz/Xp1o+TiQVRv3cXm8BfUH25g3rJKxl9dEHfcdDvW+lynR5slInJXauuXREjqjrG2NkxOzoDG9VAoh3A43GKa2nA4rrzJGDuRbW5Jdp8e1Gzf3bge3rGH7D49yO7TnZodsdt3E+rdPe5y0+1Y63OdHm2WoyLvSm3tkgiB3XxjZlnAvwHfiMZ5092nBRVPRESSU6rdfBPkXalfA3/i7vvMrCOwwszecveV8RaQnR2ipmZr43o4XEMoFGoxTXYoRH19fYt5kzF2Itvcktqde8jp17NxPdS3B7U799AxM4OcvrHbexL+vC7uctPtWOtznR5tlmNS7XGNwE6lesS+6GrH6OKnU0ZxSQkbN37C5k2bOHToEPPmvs648Tc0STNuwg3Mmf0a7s6qlSvp1q07/fv3jytvMsZOZJtbsviddUwaPwKAEcMGsXffAbbv2ktF1RZyB/bmvOxz6ZiZwS1jili8fG3c5abbsdbnOj3aLKkr0OcYzSwDWA3kAs+7+6qTpLkPuA9gwMCBTSuXmckzzz7HhHFjaGho4K67J3NRfj4vvzgTgHun3M/Y665n2VtLyM/LpUvnLrw465Vm88YrUbET2eZXZ9zNFcPPp1ePc9i4dDrTZy6hY2YGALPeXMHSFVWMuTyfqgXT2H+wnik/mQ1AQ8MRHnriDRa+UEpGB+PVspV8/On2uOOm27HW5zo92ixRKfiuVHM/rUFc64KY9QDmA9939/WnSjd8eLG/t6oi8PqIJioWOZsuG1nM6tUVKfb13za6DsjzwgdntTr/ikeuWO3uxWexSi1qk7tS3X0P8K/A2LaIJyIiySPV7koNrGM0s97RkSJm1hkYDfxHUPFERCQ5pdpzjEFeY+wPvBq9ztgBeMPdFwUYT0REklCq3ZUaWMfo7muBS4IqX0REUkAK3nyT1G++ERERaWuadkpERAJjJO4mmtZSxygiIoFKsX5RHaOIiASrQ4r1jOoYRUQkUCnWL6pjFBGR4ESeR0ytnlF3pYqISMoys7FmtsHMNprZD0+R5moz+52ZVZnZOy2VqRGjiIgEqkNAA8boC2SeJ/JmtRrgAzNb4O4fxaTpAbwAjHX3z8ysT0vlqmMUEZFABXgqdQSw0d0/jcZ5HbgR+CgmzSTgH939MwB339lSoeoY01QiZ7hI1MwemtVDJDHOsF/sZWax0y695O4vRX8OAVtj9tUAI4/L/y2go5ktB7oCz7r7a80FVMcoIiKBMSIP+Z+BXWc47VQmMBy4BugMvG9mK939981lEBERCUxQ1xiBMDAgZj0nui1WDfCFu38FfGVm/wZ8Gzhlx6i7UkVEJFV9AJxvZoPNrBMwEVhwXJoy4HIzyzSzLkROtX7cXKEaMYqISHACnHDY3Q+b2VRgGZAB/NLdq8zs/uj+me7+sZktBdYCR4BZ7r6+uXLVMYqISKCCfL7f3ZcAS47bNvO49SeBJ+MtUx2jiIgExmhH70o1s27NZXT3vWe/OiIi0t6kWL/Y7M03VcD66P+rjltv9vzs2fT2sqUU5F9Afl4uT/78Zyfsd3cefvAB8vNyKbmkgA8rK+POm6yx0y3uzGm3seW3M6iY96NTpnnq0ZtZXzaN8rmPUZiX07h99KUXsmb+j1lfNo1H7hl9WnHjqXd7O9aJjJ2ObZYIi15nbM2SEO6eNEtR0XA/UO+Ny76Dh33wkCH+0YZqr/vqax82rMAr11Q1STN/wWK/dsxY33/oiC9/930vLhkRd97mlkTFToe4WYWlTZZrJj/toybO8PWfhE/Yl1VY6jdOfd6XrljvWYWlfuUdT3r52k2eVVjqXYqmevVnOz1v3OPetfgBX7NhqxfeNP2kZWQVlqblsU6W2O29zUVFwz3R35/JuvQcdKHf/MrqVi9ARVvXOa7HNcxsopn9KPpzjpkND7S3jvqgvJyhQ3MZPGQInTp14pZbJ7JoYVmTNIsWlDHp9jsxM0aOGkVd3R62bdsWV95kjJ1ucQHeq6zmy7r9p9w//qoC5iwqB6B83Wa6d+1Mv17dKLl4ENVbd7E5/AX1hxuYt6yS8VcXxB03HY+12ty2sSU1tdgxmtlzwHeAO6Kb9gMzT53j7KmtDZOTc+zZzVAoh3A43GKa2nA4rrzJGDvd4sYju08ParbvblwP79hDdp8eZPfpTs2O2O27CfXuHne56Xis1ebk+myniw5mrV4SUt840lzq7lOAgwDu/iXQKd4AZpZhZh+a2aJW1lFERFKYncGSCPE8rlFvZh0ABzCzc4k8JBmvHxB5y0Czd7meTHZ2iJqaY++HDYdrCIVCLabJDoWor69vMW8yxk63uPGo3bmHnH49G9dDfXtQu3MPHTMzyOkbu70n4c/r4i43HY+12pxcn+100R4nKn4e+Aegt5n9b2AF8EQ8hZtZDjAOmNWayhWXlLBx4yds3rSJQ4cOMW/u64wbf0OTNOMm3MCc2a/h7qxauZJu3brTv3//uPImY+x0ixuPxe+sY9L4EQCMGDaIvfsOsH3XXiqqtpA7sDfnZZ9Lx8wMbhlTxOLla+MuNx2PtdqcXJ/tdBB5jrH1SyK0OGJ099fMbDXw3eimW7yF1+nE+D/Ao0Sm+jgpM7sPuA9gwMCBTSuXmckzzz7HhHFjaGho4K67J3NRfj4vvxi5xHnvlPsZe90r1RhiAAAgAElEQVT1LHtrCfl5uXTp3IUXZ73SbN54JSp2usUFeHXG3Vwx/Hx69TiHjUunM33mEjpmZgAw680VLF1RxZjL86laMI39B+uZ8pPZADQ0HOGhJ95g4QulZHQwXi1bycefbo87bjoea7W5bWMLgb4SLijm7i0nMisALidyOvU9d2/xz3IzGw9c7+7/zcyuBh5x9/HN5Rk+vNjfW1XRXBJpBzQfo7Q3l40sZvXqitT69m8j5w7J9+v+ck6r8//6jsLVfmbTTp22eO5K/Z/Ab4BsIlN6zDGzx+Io+zLgBjPbDLwO/ImZzT6DuoqISAoya/2SCPHcfHMncIm77wcws58CHwIzmsvk7o8Bj0XzXE1kxHj7GdVWRERSTqqdSo2nY9x2XLrM6DYREZFmHb35JpU09xLxZ4hcU/wSqDKzZdH1a4lMDhk3d18OLG91LUVEJGW1pxHj0TtPq4DFMdtXBlcdERFpb1KrW2ymY3T3X7RlRURERJJBi9cYzWwo8FPgIiDr6HZ3/1aA9RIRkXbALPUmKo7nzTe/Al4hMhq+DngDmBtgnUREpB1Jtcc14ukYu7j7MgB3r3b3/0WkgxQREWlRqk1UHM/jGl9HXyJebWb3A2GaecWbiIhIrBQ7kxpXx/gQ8E3gASLXGrsDk4OslIiItA9G4uZVbK14XiK+KvrjHzg2WbGIiEjLEnitsLWae8B/PtE5GE/G3W8KpEYiIiIJ1NyIUVMRSCASNcuFZvUQSYx28+Ybd/9tW1ZERETap3gef0gm8dx8IyIi0ipGOxoxioiInA3tZnaN45nZN9z96yArIyIi7U+qdYwtnvo1sxFmtg74JLr+bTP7u8BrJiIikgDxXBP9W2A88AWAu68BvhNkpUREpH2IvPO0/b0SroO7bzmugg0B1UdERNqZdncqFdhqZiMAN7MMM3sQ+H3A9Wr09rKlFORfQH5eLk/+/Gcn7Hd3Hn7wAfLzcim5pIAPKyvjzpussdMtbiJjz5x2G1t+O4OKeT86ZZqnHr2Z9WXTKJ/7GIV5OY3bR196IWvm/5j1ZdN45J7RpxU3HY91OrZZIlJtdg3cvdkF6AO8DuyKLq8DvVrK15qlqGi4H6j3xmXfwcM+eMgQ/2hDtdd99bUPG1bglWuqmqSZv2CxXztmrO8/dMSXv/u+F5eMiDtvc0uiYqdb3LaMnVVYesJyzeSnfdTEGb7+k/BJ99849XlfumK9ZxWW+pV3POnlazd5VmGpdyma6tWf7fS8cY971+IHfM2GrV540/STlpGOxzpZ4rZV7KKi4R7Ed2J7WPrl5vtfLN7Q6gWoaOs6tzhidPed7j7R3XtFl4nuvivIzvqoD8rLGTo0l8FDhtCpUyduuXUiixaWNUmzaEEZk26/EzNj5KhR1NXtYdu2bXHlTcbY6RY30bHfq6zmy7r9p9w//qoC5iwqB6B83Wa6d+1Mv17dKLl4ENVbd7E5/AX1hxuYt6yS8VcXJH179flq29gS0eEMlkSI567Ul83speOXtqhcbW2YnJwBjeuhUA7hcLjFNLXhcFx5kzF2usVNdOyWZPfpQc323Y3r4R17yO7Tg+w+3anZEbt9N6He3eMqMx2PdTq2WY5JtVOp8dx88y8xP2cB/xnYGk/hZraZyKwcDcBhdy8+3QqKiIi0pXimnZobu25mfw+sOI0Y32ntqdfs7BA1Ncf64HC4hlAo1GKa7FCI+vr6FvMmY+x0i5vo2C2p3bmHnH49G9dDfXtQu3MPHTMzyOkbu70n4c/r4iozHY91OrZZIsxSbz7G1pzCHQz0PdsVOZnikhI2bvyEzZs2cejQIebNfZ1x429okmbchBuYM/s13J1VK1fSrVt3+vfvH1feZIydbnETHbsli99Zx6TxIwAYMWwQe/cdYPuuvVRUbSF3YG/Oyz6XjpkZ3DKmiMXL1yZ9e/X5Sq7PV7pod6dSzWw3x+Zl7AB8CfwwzvId+BczawBedPfTujaZmZnJM88+x4RxY2hoaOCuuydzUX4+L784E4B7p9zP2OuuZ9lbS8jPy6VL5y68OOuVZvMme+x0i5vo2K/OuJsrhp9Prx7nsHHpdKbPXELHzAwAZr25gqUrqhhzeT5VC6ax/2A9U34yG4CGhiM89MQbLHyhlIwOxqtlK/n40+1J3159vto2tkSk2nOM5n7KuYixyFP9A4CjV5uPeHMZTswfcvewmfUB/hn4vrv/23Fp7gPuAxgwcODw31dvOc0miMRH8zFKUC4bWczq1RUp9vXfNkLfGuZTnp/f6vzTrj1/dVvfn9LsqdRoJ7jE3RuiS9ydYjR/OPr/ncB8YMRJ0rzk7sXuXty7V+/TKV5ERFJAqp1Kjeca4+/M7JLTLdjMvmlmXY/+DFwLrD/dckRERNrSKa8xmlmmux8GLgE+MLNq4Csi8066uxe1UHZfYH70HauZwBx3X3p2qi0iIinBUu8aY3M335QDRUCrbsFy90+Bb7cmr4iItB9GavWMzXWMBuDu1W1UFxERaWeM9jVi7G1mD59qp7s/HUB9RESknWlPHWMGcA6k2BhYRESSSqImHG6t5jrGbe7+l21WExERkSTQ4jVGERGR1mpv1xivabNaiIhI+5TAB/Vb65QP+Lv7l21ZERERaZ86RGfYaM3SEjMba2YbzGyjmZ3yPd5mVmJmh83s5pbKjGc+RhERkVYJ8lSqmWUAzwOjgRoiL6NZ4O4fnSTdE8Db8ZTbmmmnRERE4hbgu1JHABvd/VN3PwS8Dtx4knTfB/4B2BlPfTVilLSRqFkuEjWrB2hmD0kGRofg7uUMAVtj1muAkU2im4WA/wx8ByiJp1B1jCIiksx6mVlFzPpLpzm37/8B/sLdj8T7PKU6RhERCYxxxnel7mpmPsYwkTmDj8rh2PzBRxUDr0c7xV7A9WZ22N3/6VQB1TGKiEhwgp1d4wPgfDMbTKRDnAhMik3g7oMbq2L2K2BRc50iqGMUEZGAxfPYRWu4+2EzmwosI/Ia01+6e5WZ3R/dP7M15apjFBGRwJyFU6nNcvclwJLjtp20Q3T3u+MpUx2jiIgEKqgRY1D0HKOIiEgMjRhFRCRQKTZgTP4R49vLllKQfwH5ebk8+fOfnbDf3Xn4wQfIz8ul5JICPqysjDtvssZOt7iJjJ2ouDOn3caW386gYt6PTpnmqUdvZn3ZNMrnPkZhXk7j9tGXXsia+T9mfdk0Hrln9GnFjafe7e1YJzp2ujMiHU1rl4Rw96RZioqG+4F6b1z2HTzsg4cM8Y82VHvdV1/7sGEFXrmmqkma+QsW+7Vjxvr+Q0d8+bvve3HJiLjzNrckKna6xU2HNmcVlp6wXDP5aR81cYav/yR80v03Tn3el65Y71mFpX7lHU96+dpNnlVY6l2Kpnr1Zzs9b9zj3rX4AV+zYasX3jT9pGVkFZam3bFOVOyiouGe6O/PZF0GXTjMf/XBZ61egIq2rnNSjxg/KC9n6NBcBg8ZQqdOnbjl1oksWljWJM2iBWVMuv1OzIyRo0ZRV7eHbdu2xZU3GWOnW9x0bfN7ldV8Wbf/lPvHX1XAnEXlAJSv20z3rp3p16sbJRcPonrrLjaHv6D+cAPzllUy/uqCuOOm47FOZGyJsDNYEiGpO8ba2jA5OcdeahAK5RAOh1tMUxsOx5U3GWOnW9xExk5km1uS3acHNdt3N66Hd+whu08Psvt0p2ZH7PbdhHp3j7vcdDzWyfx7TgeR2TWCm3YqCIF2jGbWw8zeNLP/MLOPzeyPg4wnIiJypoK+K/VZYKm732xmnYAup5M5OztETc2xF6eHwzWEQqEW02SHQtTX17eYNxljp1vcRMZOZJtbUrtzDzn9ejauh/r2oHbnHjpmZpDTN3Z7T8Kf18Vdbjoe62T+PaeLFLspNbgRo5l1B64EfgHg7ofcfc/plFFcUsLGjZ+wedMmDh06xLy5rzNu/A1N0oybcANzZr+Gu7Nq5Uq6detO//7948qbjLHTLW66trkli99Zx6TxIwAYMWwQe/cdYPuuvVRUbSF3YG/Oyz6XjpkZ3DKmiMXL18Zdbjoe62T+PaeLAOdjDESQI8bBwOfAK2b2bWA18AN3/yo2kZndB9wHMGDgwKaVy8zkmWefY8K4MTQ0NHDX3ZO5KD+fl1+MvO3n3in3M/a661n21hLy83Lp0rkLL856pdm88UpU7HSLm65tfnXG3Vwx/Hx69TiHjUunM33mEjpmZgAw680VLF1RxZjL86laMI39B+uZ8pPZADQ0HOGhJ95g4QulZHQwXi1bycefbo87bjoe60TGFgAj3umekoW5ezAFmxUDK4HL3H2VmT0L7HX3H58qz/Dhxf7eqopT7RZJSZqouP27bGQxq1dXpNa3fxsZetG3/a9/vaTlhKcwsShntZ962qlABHnzTQ1Q4+6routvAkUBxhMRkSRkZq1eEiGwjtHdtwNbzeyC6KZrgI+CiiciIskp1Z5jDPqu1O8Dv47ekfopcE/A8URERM5IoB2ju/8OaNNzwyIikkSMlLv5RrNriIhIYI6+RDyVqGMUEZFAacQoIiISI7W6RXWMIiISsBQbMKbcqV8REZFAacQoIiKBidx8k1pDRnWMIiISqFQ7laqOUUREAmSYRowiIiLHaMQoIk0kcoaLRM3soVk95KhUvMaou1JFRERiaMQoIiLBMZ1KFRERaUIdo4iISAzdlSoiIhJlQIfU6hfVMYqISLA0YhQREYmRatcYk/5xjbeXLaUg/wLy83J58uc/O2G/u/Pwgw+Qn5dLySUFfFhZGXfeZI2dbnETGTvd4s6cdhtbfjuDink/OmWapx69mfVl0yif+xiFeTmN20dfeiFr5v+Y9WXTeOSe0acVN556t8fPl6Qod0+apahouB+o98Zl38HDPnjIEP9oQ7XXffW1DxtW4JVrqpqkmb9gsV87ZqzvP3TEl7/7vheXjIg7b3NLomKnW9x0bHNbxs0qLG2yXDP5aR81cYav/yR8wr6swlK/cerzvnTFes8qLPUr73jSy9du8qzCUu9SNNWrP9vpeeMe967FD/iaDVu98KbpJy0jq7A0aY51W8UuKhruif7+TNblW/nf9n/9jy9avQAVbV3npB4xflBeztChuQweMoROnTpxy60TWbSwrEmaRQvKmHT7nZgZI0eNoq5uD9u2bYsrbzLGTre46djmRB7r9yqr+bJu/yn3j7+qgDmLygEoX7eZ7l07069XN0ouHkT11l1sDn9B/eEG5i2rZPzVBXHHTcfPl0QcvfmmtUsiJHXHWFsbJidnQON6KJRDOBxuMU1tOBxX3mSMnW5xExk73eLGI7tPD2q2725cD+/YQ3afHmT36U7Njtjtuwn17h53uen4+ZKj7Iz+S4TAOkYzu8DMfhez7DWzB4OKJyIiSSj65pvWLokQ2F2p7r4BKAQwswwgDMw/nTKys0PU1GxtXA+HawiFQi2myQ6FqK+vbzFvMsZOt7iJjJ1uceNRu3MPOf16Nq6H+vagduceOmZmkNM3dntPwp/XxV1uOn6+5JgUuym1zU6lXgNUu/uW08lUXFLCxo2fsHnTJg4dOsS8ua8zbvwNTdKMm3ADc2a/hruzauVKunXrTv/+/ePKm4yx0y1uOrY5kce6JYvfWcek8SMAGDFsEHv3HWD7rr1UVG0hd2Bvzss+l46ZGdwypojFy9fGXW46fr4kdbXVc4wTgd+cbIeZ3QfcBzBg4MAm+zIzM3nm2eeYMG4MDQ0N3HX3ZC7Kz+flF2cCcO+U+xl73fUse2sJ+Xm5dOnchRdnvdJs3nglKna6xU3HNifyWL86426uGH4+vXqcw8al05k+cwkdMzMAmPXmCpauqGLM5flULZjG/oP1TPnJbAAaGo7w0BNvsPCFUjI6GK+WreTjT7fHHTcdP18SEbn5JrXGjObuwQYw6wTUAvnuvqO5tMOHF/t7qyoCrY9IOtF8jG3jspHFrF5dkVrf/m3kwmGX+Cvz/7XV+f/4/J6r3b34LFapRW0xYrwOqGypUxQRkXYqxf5kaIuO8Xuc4jSqiIi0f3pXagwz+yYwGpgSZBwREUleKXaJMdiO0d2/As4NMoaIiMjZpNk1REQkUCk2YFTHKCIiAUuxnlEdo4iIBMbQzTciIiLHJPCdp62ljlFERAKVYv2iOkYREQlYivWMST0fo4iISHPMbKyZbTCzjWb2w5Psv83M1prZOjP7dzP7dktlasQoIiIBCm7C4eiUhs8TeZFMDfCBmS1w949ikm0CrnL33WZ2HfASMLK5ctUxiohIoAK8+WYEsNHdP43EsdeBG4HGjtHd/z0m/Uogp6VC1TGKtGOJmuUiUbN6QPrN7JHsjDO+xNjLzGKnXXrJ3V+K/hwCtsbsq6H50eCfAW+1FFAdo4iIBOvMesZdZ2PaKTP7DpGO8fKW0qpjFBGRQAX4gH8YGBCznhPd1jS+WQEwC7jO3b9oqVDdlSoiIqnqA+B8MxtsZp2AicCC2ARmNhD4R+AOd/99PIVqxCgiIoEK6uYbdz9sZlOBZUAG8Et3rzKz+6P7ZwKPE5nl6QWLVORwS6dm1TGKiEiggny+392XAEuO2zYz5uc/B/78dMpUxygiIsE5C7eltjV1jCIiEqhUm10j6W++eXvZUgryLyA/L5cnf/6zE/a7Ow8/+AD5ebmUXFLAh5WVcedN1tjpFjeRsdMtbiJjz5x2G1t+O4OKeT86ZZqnHr2Z9WXTKJ/7GIV5x57DHn3phayZ/2PWl03jkXtGn1bceOod5PFOd0bkGmNrl4Rw96RZioqG+4F6b1z2HTzsg4cM8Y82VHvdV1/7sGEFXrmmqkma+QsW+7Vjxvr+Q0d8+bvve3HJiLjzNrckKna6xU3HNqfDsc4qLD1huWby0z5q4gxf/0n4pPtvnPq8L12x3rMKS/3KO5708rWbPKuw1LsUTfXqz3Z63rjHvWvxA75mw1YvvGn6ScvIKixNSJuLioZ7or8/k3XJL7jEPwrva/UCVLR1nZN6xPhBeTlDh+YyeMgQOnXqxC23TmTRwrImaRYtKGPS7XdiZowcNYq6uj1s27YtrrzJGDvd4qZjm9PxWAO8V1nNl3X7T7l//FUFzFlUDkD5us1079qZfr26UXLxIKq37mJz+AvqDzcwb1kl468uSIk2S2pK6o6xtjZMTs6xZzdDoRzC4XCLaWrD4bjyJmPsdIubyNjpFjfRsVuS3acHNdt3N66Hd+whu08Psvt0p2ZH7PbdhHp3j7vcZG5z2rAzWBIg0I7RzB4ysyozW29mvzGzrCDjiYhI8rEz+C8RArsr1cxCwAPARe5+wMzeIPJWgl/FW0Z2doiammPvhw2HawiFQi2myQ6FqK+vbzFvMsZOt7iJjJ1ucRMduyW1O/eQ069n43qobw9qd+6hY2YGOX1jt/ck/Hld3OUmc5vTRcJuommloE+lZgKdzSwT6ALUnk7m4pISNm78hM2bNnHo0CHmzX2dceNvaJJm3IQbmDP7NdydVStX0q1bd/r37x9X3mSMnW5x07HN6Xis47H4nXVMGj8CgBHDBrF33wG279pLRdUWcgf25rzsc+mYmcEtY4pYvHxtu2hzukixM6nBjRjdPWxmfwN8BhwA3nb3t0+njMzMTJ559jkmjBtDQ0MDd909mYvy83n5xchLDe6dcj9jr7ueZW8tIT8vly6du/DirFeazZvssdMtbjq2OR2PNcCrM+7miuHn06vHOWxcOp3pM5fQMTMDgFlvrmDpiirGXJ5P1YJp7D9Yz5SfzAagoeEIDz3xBgtfKCWjg/Fq2Uo+/nR7SrRZolJsxGjuHkzBZj2BfwBuBfYA84A33X32cenuA+4DGDBw4PDfV28JpD4i0nbSbT7Gy0YWs3p1RYp9/beNi79d5P/49opW57+g3zdX+1mYdup0BHkq9bvAJnf/3N3ribzd/NLjE7n7S+5e7O7FvXv1DrA6IiLS1iKnRHXzzVGfAaPMrAuRU6nXABXNZxERkXYlkW+waaUgrzGuMrM3gUrgMPAh8FJQ8UREJDmlWL8Y7EvE3X0aMC3IGCIikuRSrGfU7BoiIhKgxF0rbK2kfiWciIhIW9OIUUREAqWbb0RERKIS+Qab1lLHKCIiwUqxnlEdo4iIBCrVbr5RxygiIoHSNUYREZEYKdYv6nENERGRWBoxiohIcPSuVBGRxEz9dFQiprz6esNnbR4ztaRWz6iOUUREAmNoxCgiItJEivWL6hhFRCRYqTZi1F2pIiIiMTRiFBGRQOnNNyIiIrFSq19UxygiIsFKsX4x+a8xvr1sKQX5F5Cfl8uTP//ZCfvdnYcffID8vFxKLingw8rKuPMma+x0i5vI2OkWN5GxExV35rTb2PLbGVTM+9Ep0zz16M2sL5tG+dzHKMzLadw++tILWTP/x6wvm8Yj94w+rbgSYXZmS0K4e9IsRUXD/UC9Ny77Dh72wUOG+Ecbqr3uq6992LACr1xT1STN/AWL/doxY33/oSO+/N33vbhkRNx5m1sSFTvd4qZjm3Wsg42bVVjaZLlm8tM+auIMX/9J+IR9WYWlfuPU533pivWeVVjqV97xpJev3eRZhaXepWiqV3+20/PGPe5dix/wNRu2euFN009ahnXu7Yn+/kzW5duXFPnOvfWtXoCKtq5zUo8YPygvZ+jQXAYPGUKnTp245daJLFpY1iTNogVlTLr9TsyMkaNGUVe3h23btsWVNxljp1vcdGyzjnXbtvm9ymq+rNt/yv3jrypgzqJyAMrXbaZ7187069WNkosHUb11F5vDX1B/uIF5yyoZf3VB3HEldSV1x1hbGyYnZ0DjeiiUQzgcbjFNbTgcV95kjJ1ucRMZO93iJjJ2Itvckuw+PajZvrtxPbxjD9l9epDdpzs1O2K37ybUu/tZi5tW7AyWBAi0YzSzH5jZejOrMrMHg4wlIiLJKcX6xeDuSjWzi4F7gRHAIWCpmS1y943xlpGdHaKmZmvjejhcQygUajFNdihEfX19i3mTMXa6xU1k7HSLm8jYiWxzS2p37iGnX8/G9VDfHtTu3EPHzAxy+sZu70n487qzFjed6M03x1wIrHL3/e5+GHgHuOl0CiguKWHjxk/YvGkThw4dYt7c1xk3/oYmacZNuIE5s1/D3Vm1ciXdunWnf//+ceVNxtjpFjcd26xj3bZtbsnid9YxafwIAEYMG8TefQfYvmsvFVVbyB3Ym/Oyz6VjZga3jCli8fK1Zy1u+rAz+i8RgnyOcT3wUzM7FzgAXA9UHJ/IzO4D7gMYMHBg08plZvLMs88xYdwYGhoauOvuyVyUn8/LL84E4N4p9zP2uutZ9tYS8vNy6dK5Cy/OeqXZvPFKVOx0i5uObdaxbts2vzrjbq4Yfj69epzDxqXTmT5zCR0zMwCY9eYKlq6oYszl+VQtmMb+g/VM+clsABoajvDQE2+w8IVSMjoYr5at5ONPt8cdVyJScXYNc/fgCjf7M+C/AV8BVcDX7n7Ka43Dhxf7e6tO6DtFROKWmPkY3+DI/p0p9vXfNi4pKvb/t2JVq/P/0TczV7t78VmsUosCvfnG3X/h7sPd/UpgN/D7IOOJiIicqUBfCWdmfdx9p5kNJHJ9cVSQ8UREJPmk2qnUoN+V+g/Ra4z1QKm77wk4noiIJBnNrhHD3a8IsnwREUlyiXznaStpdg0REQlMIh/Uby11jCIiEqwU6xmT+l2pIiIibU0jRhERCZRuvhEREYmhm29ERERipFi/qGuMIiISsADnnTKzsWa2wcw2mtkPT7LfzOxvo/vXmllRS2WqYxQRkUAFNbuGmWUAzwPXARcB3zOzi45Ldh1wfnS5D/i/LdVXHaOIiKSqEcBGd//U3Q8BrwM3HpfmRuA1j1gJ9DCz/s0VmlTXGCsrV+/q3NG2tDJ7L2DX2axPksdNZOx0i5vI2OkWN5GxzyTueWezIu3Jh5Wrl3XpZL3OoIgsM4uddukld38p+nMI2BqzrwYYeVz+k6UJAdtOFTCpOkZ3793avGZW0dZTkyQybiJjp1vcRMZOt7iJjJ3INrdn7j420XU4XTqVKiIiqSoMDIhZz4luO900TahjFBGRVPUBcL6ZDTazTsBEYMFxaRYAd0bvTh0F1Ln7KU+jQpKdSj1DL7WcpF3FTWTsdIubyNjpFjeRsRPZZmkFdz9sZlOBZUAG8Et3rzKz+6P7ZwJLgOuBjcB+4J6WyjV3D67WIiIiKUanUkVERGKoYxQREYmhjlFShlmqvYq49czsmwmK2y+djrPIyaRsx2hmF5jZH5tZx+hrgdo6fiJi5ppZsZl9IwGx883sKjM7t43jXm5mdwC4u7fVl7aZTTCzH7RFrJPEvhF4wsz6tHHcMcB8mt7a3hZxR5nZHdH/d2rj2OdH/011SMS/aUlOKdkxmtlNQBnwV8AvgFIz69ZGsb8F4O4NbfkPyczGA/8IPAn86mg92ij2dcBvgIeA18ysXxvE7GBm5wAvAo/F3GXmZhbo59bMrgWmAx8FGecUsa8CngDK3H1nG8a9Nhq3P/Df2zDuDUTuBv0u8Aht+AYZM/tPwJvAY8DTwJREjdQluaRcx2hmHYFbgT9z92uIdJADgL8IunOMdk6/M7M50Hado5ldSqRDvMvdvwPsBk54i3xAsa8GngX+3N3/E3AIuDjouO5+xN33Aa8S+ePnUjN76Oi+oOJGj/XfA/e5+z+bWXczO8/MugQV8zjDgVnR2NlmNtrMRppZ96ACmtl3gReA24i8aPlCM7syqHgxcc8FSoFJ7n4XsBcoNLM+ZpbVBrGnAN9z9z8F1hK5jf9hM+saZGxJfinXMUZ1I/IPGCKnfhYBHYFJQZ1qi/4lORV4EDhkZrOhTUeOT7j7h9GfpwF/1EanVHcAU9y9PDpSHAlMNbMXzezmNji1eZjIHz6vAiPM7GkzmxF9WDeIz+8XQD3QP/rl+U9E3sb/qzZs71FvApOJfO6eN7OeAcXMAO5090I6ujsAAAbbSURBVCrgm8AGIB8Cv657GOgM5EX/qL0auBP4P8D/Cnj0dhg4B+gH4O6/BDYTeV/q+ADjSgpIuY7R3euJnPa4ycyuiI4eVgC/Ay4PMO5XRL6k5hA55ZMV2zkGFTdqFZHTqEevbX6DyCmnbtFtgV33c/eP3f1fo6t/BrwQHTm+D9xM5IskSGXAdnf/LVAB3A90i74p/6yPHN19AzAOeAZYR+T3PR5YCvwpEFTndNS/Avea2evAy+7+PSJ/CO0jMpPAWefuy9z9382sg7vvARYD08xsmAf4oLO71wF/S+RU5tvAK+4+AZhF5LVduQHH/jUwOXp986fA10ROn383qLiSGlKuY4x6l8g/pDvM7Ep3b3D3OUA28O2ggrp7rbvvc/ddRE7DdD7aOZpZkZnlBRS3wd33RlcN2AN86e6fm9ltwF+ZWecgYh9Xj5+6+19Ff/4VkY456Bs1DgAXmNm9RDrFnwEDzWxKUAHdfQ2RzvCn7v5y9LTuL4l0igODihuNvY7IH14jgcHRbZ8SGdW1+iX7ccY+Ev3/UiLX/cYHODI/GvNNIh3Ru8CH0W3/D+hK8NcbfwO8BXwH6Ozut7v7i0DftrpnQZJTSr4Szt0PmtmvASdyY0Yekb/2/n97dxtiRR3Fcfz700wlFy1IoZC2NHuSMkWIhNLSpSdDMrNIYkkULaIytRcpGARCvioionohQUhGCVaIREGZrLW1uGK5KvYkGCQRlk9EcXrx/98YNx+u685cV34fWPZ679w5M9fdOfOf/+w5IzhJK5Fe3obf8sF5laQu0oFrSgVx/wYOStoraSXQArRGxJEy40pScfQgaSbp895XZtyI2CdpL7AceDwiPpA0hVTeqcy431G4+Sbv78VU8/O1gTRKXCH914btRtJJQVU6STdbvVj2FZGI+F3Sp8ADkv4CBpFOCraVHPcA8LakNbWTAkmPABcBZV8FsrNYny4Jl2/tnkQavR0FXirMw1W1DU8DzwLT8tl+2fFEmk/dkb/fHhG7y45biD8QmAMsAmZHxPYKYo4EhkfEN/nf/cq8AadbbJFuylgMzMrzcJWQNJ50uXogsLqKn69u8dcCSyPixwpiDSPNL84k/S4vzSP3ykh6lPT/PLvqz9rOLn06MdbkebdS5pxOEfdCYC3wTESUenZ7nNitQHuVB+ocdwAwDdiT5+OqjH3MqLWqmMCtpHnOripjN0ojPudC7CbScemPUy7c+7EvAwZERKlXI+zsd04kxkaSNCgijjYgbsMOXmZm5zInRjMzs4K+eleqmZlZKZwYzczMCpwYzczMCpwYzczMCpwYrRSS/pG0VdJ2Se+eSRFuSZMlfZgf3yvphAXUJQ2T9FgPYqyQtLje57sts1rS/acRq1lS6X//aWY948RoZTkSEeMiYiypI8eC4os9LTUWEesj4mQVYIYBp50YzcxqnBitCpuA0XmktFPSW8B2YKSkFkltkjryyHIIgKQ7JHVJ6gDuq61IUqukV/LjEZLWSerMXzeTyqaNyqPVVXm5JZLaJW2T9HxhXc9J2iXpC+CqU+2EpHl5PZ2S3us2Cp4q6eu8vnvy8v0lrSrELq2+q5n1HidGK5Wk84A7SZ0qILULezUirgMOAcuAqRExntQ9Y5FSL743gOmk/oQnaoz8MvBZRNwAjAe+JfWp3JNHq0uUGvBeSepMMQ6YIOkWSROAB/NzdwET69id9yNiYo63g9RtpKY5x7gbeC3vw1zgQERMzOufJ+nyOuKYWQP1ySLi1icMlrQ1P95EajZ8CfBTRGzJz98EXAtsTpXXOJ/Uzupq4IdaDVilDibzjxPjNlJ9zVrrrwP6f8/ClvxVq6E7hJQom4B1EXE4x1hfxz6NlfQC6XLtEGBj4bW1uSThbknf531oAa4vzD8OzbF31RHLzBrEidHKciQixhWfyMnvUPEp4OPcc7C43DHvO0MCVuZ2QsUYT/VgXauBGRHRmWvVTi681r2EVOTYT0REMYEiqbkHsc2sIr6Uao20BZgkaTSApAskjQG6gGZJo/JyD53g/Z8AC/N7+0saCvxJGg3WbCQ1o63NXV4qaTjwOTBD0uBcuHp6HdvbBPySC6k/3O21WZL65W2+AtiZYy/MyyNpjMrtSm9mvcAjRmuY3Gi5FViT21kBLIuIXZLmAx9JOky6FNt0nFU8CbwuaS6pf97CiGiTtDn/OcSGPM94DdCWR6wHgTkR0SHpHVLfwV+B9jo2eTnwJbA/fy9u08/AV6TmzQtyz9A3SXOPHblLx35gRn2fjpk1iouIm5mZFfhSqpmZWYETo5mZWYETo5mZWYETo5mZWYETo5mZWYETo5mZWYETo5mZWcG/eXNZ31Ie+fQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc789c7e908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import itertools\n",
    "\n",
    "class_names = [i for i in range(10)]\n",
    "np.set_printoptions(precision=2)\n",
    "\n",
    "# Plot non-normalized confusion matrix\n",
    "plt.figure(figsize=(20,5))\n",
    "plot_confusion_matrix(cnf_matrix, classes=class_names,\n",
    "                      title='Confusion matrix, without normalization')\n",
    "\n",
    "# Plot normalized confusion matrix\n",
    "plt.figure(figsize=(20,5))\n",
    "plot_confusion_matrix(cnf_matrix, classes=class_names, normalize=True,\n",
    "                      title='Normalized confusion matrix')\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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
}