AutoEncoders.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JfkDHn7aejaC"
   },
   "source": [
    "# Auto Encoders\n",
    "\n",
    "Auto Encoder modeli alınan bir girdiyi gizli katmanlarda daha az nöron kullanarak sıkıştırarak veriyi en iyi temsil eden özelliklerin çıkarılmasına yardımcı olur. Daha sonra bu özelliklerden baştaki çıkyıtı en iyi verecek output layer ağırlıklarını oluşturmaya çalışır. Yani girdi olarak bir resim dosyası verdiğimizde, bu resmi en iyi temsil eden özellikleri çıkarır ve daha sonra bu özelliklerden aynı resmi oluşturmaya çalışır. Sadece resim değil, sinyal işlemede gürültü(noise) kaldırma işlemlerinde de kullanılabilir. Veya bir zaman serisinde verinin genel eğilimlerini bulur ve verideki gürültüyü ortadan kaldırabilir. Bunun gibi birçok kullanım alanı olsa da daha iyi anlamak adına bu yapının görsel haline bakmak daha iyi anlamamıza yardımcı olabilir.\n",
    "![alt text](https://static.packt-cdn.com/products/9781787121089/graphics/B12043_04_06.png)\n",
    "\n",
    "Resmin alındığı kaynak: https://subscription.packtpub.com/book/big_data_and_business_intelligence/9781787121089/4/ch04lvl1sec51/setting-up-stacked-autoencoders"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PP7_AYKzgd8W"
   },
   "source": [
    "**Kütüphanelerin Yüklenmesi**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "l0DRTJLQOx0V"
   },
   "outputs": [],
   "source": [
    "from keras.models import Input, Model\n",
    "from keras.layers import Dense\n",
    "from keras.datasets import fashion_mnist\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import json\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "LAjgTmOIgizH"
   },
   "source": [
    "**Verisetinin Yüklenmesi**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7MoKT5NBPjlb"
   },
   "outputs": [],
   "source": [
    "(x_train, _), (x_test, _) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EUcPqnDTgly2"
   },
   "source": [
    "**Verinin Tipini Dönüştüme ve Scale İşlemi**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "UyJ94L2NP74F"
   },
   "outputs": [],
   "source": [
    "x_train = x_train.astype(\"float32\")/255.0\n",
    "x_test = x_test.astype(\"float32\")/255.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "lEubhzSEgxfG"
   },
   "source": [
    "**Verinin Yeniden Boyutlandırılması**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-ci3Nwv2QIQj"
   },
   "outputs": [],
   "source": [
    "x_train = x_train.reshape((len(x_train), x_train.shape[1:][0]*x_train.shape[1:][1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3Nrntnt4Q-fg"
   },
   "outputs": [],
   "source": [
    "x_test = x_test.reshape((len(x_test), x_test.shape[1:][0]*x_test.shape[1:][1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "9TL3wcUoRCbt",
    "outputId": "eb5f9b1d-1e5e-45e2-a538-9c96ceb9b8ba"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 784) (10000, 784)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape,x_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "aG1hWezhg2sU"
   },
   "source": [
    "**Verisetini Örnek Olarak Görselleştirme**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 773
    },
    "colab_type": "code",
    "id": "zz54PLErRXxZ",
    "outputId": "d745fccf-f129-4825-aceb-c7fccb14c155"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmEIq2X8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDD+Rpxh+Ik8x/ESeYviJPMXwE3nq/wA4kosVeqdINgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train[4000].reshape(28,28))\n",
    "plt.show()\n",
    "plt.imshow(x_train[1500].reshape(28,28))\n",
    "plt.show()\n",
    "plt.imshow(x_train[150].reshape(28,28))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "XsbuWaW6SVJF"
   },
   "source": [
    "**AutoEncoder Modeli OLuşturma Aşaması**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "OYdrPwrHTXks"
   },
   "source": [
    "Encoder ve Decoder layerlarının oluşturulması"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 139
    },
    "colab_type": "code",
    "id": "HFOBi0Q_ScNQ",
    "outputId": "72783e7b-a7ae-4c47-9940-91f882a73093"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
      "\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
      "\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "input_img = Input(shape=(784,))\n",
    "encoded = Dense(32, activation=\"relu\")(input_img)\n",
    "encoded = Dense(16, activation=\"relu\")(encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NuJOaf0RTdLo"
   },
   "outputs": [],
   "source": [
    "decoded = Dense(32, activation=\"relu\")(encoded)\n",
    "output_img = Dense(784, activation=\"sigmoid\")(decoded)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "9zdgAlxLUMHZ"
   },
   "source": [
    "**AutoEncoder Modeli Oluşturma**\n",
    "\n",
    "İnput kısmından output kısmına kadar oluşturduğumuz layerları bir model içine depolayıp AutoEncoder kurmuş olacağız."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "5JKYMXeOUL1p"
   },
   "outputs": [],
   "source": [
    "autoencoder = Model(input_img, output_img)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "L-mMjhM3VKsU"
   },
   "source": [
    "**Model Compile ve Train Aşamaları**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 156
    },
    "colab_type": "code",
    "id": "PsdazrsdVJ_V",
    "outputId": "11aa323c-df86-454e-dbd3-3ffd1cba08d0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
      "\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3657: The name tf.log is deprecated. Please use tf.math.log instead.\n",
      "\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/nn_impl.py:180: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
     ]
    }
   ],
   "source": [
    "autoencoder.compile(optimizer=\"rmsprop\", loss=\"binary_crossentropy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "i1QdpsWOVtvz"
   },
   "outputs": [],
   "source": [
    "history = autoencoder.fit(x_train, x_train, \n",
    "                          epochs=200, \n",
    "                          batch_size=256, \n",
    "                          shuffle=True,\n",
    "                          validation_data=(x_train, x_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4R7j8VPKZlNO"
   },
   "source": [
    "**Sonuçların Görselleştirilmesi**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 269
    },
    "colab_type": "code",
    "id": "OhQ1Sh0LZtp3",
    "outputId": "a11a4893-e73f-4096-964f-2c2f7f32c41c"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history[\"loss\"], label=\"Train Loss\")\n",
    "plt.plot(history.history[\"val_loss\"], label=\"Validation Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "lV2VssusaMea"
   },
   "source": [
    "**Sonuçta ortaya çıkan resimlere test setinden tahmin yapıp bakalım**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0Vsm3pPfaTA6"
   },
   "outputs": [],
   "source": [
    "x_test_pred = autoencoder.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 560
    },
    "colab_type": "code",
    "id": "AD-Vncf-afqh",
    "outputId": "53cdb6a1-b3d9-4f87-f332-fe4d9a415dba"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEXCAYAAABiaJKRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5eQ3y5ST3yxT/x+bBnV7XBA4ywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_test[100].reshape(28,28))\n",
    "plt.title(\"Gerçek Resim\")\n",
    "plt.show()\n",
    "plt.imshow(x_test_pred[100].reshape(28,28))\n",
    "plt.title(\"Auto Encoder \\nSonucu Çıkan Resim\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZnqymKpLbaao"
   },
   "source": [
    "Sonuç olarak modelimizin hiç görmediği resimler üzerinde de kabul edilebilir sonuçlar verdiğini resimleri görselleştirerek anlamış olduk. Tabi burada yazılanlar eğitim amaçlı ve optimize edilmemiş bir modelden ortaya çıktı. Daha iyi sonuçlar elde edilmesi mümkün."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jRoHhdildqL8"
   },
   "source": [
    "**Daha fazla resim ile görselleştirelim**\n",
    "\n",
    "Kaynak: https://blog.keras.io/building-autoencoders-in-keras.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 264
    },
    "colab_type": "code",
    "id": "YNVZPi3Sd1nm",
    "outputId": "164b43f6-0eab-43d3-c919-97da8cde32e2"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABGoAAADjCAYAAADdR/IFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJztvXnQXVWV/r9QAUEIEBICmcnMkBBC\nGJIQbAjKYCMiKC1VaCO22pZNt2Vr2/a3SkFs264WbbXswgmVKhEaB0AUEVqmMA8hAZKQeSAz8yCI\nkN8f/t7Ns5+8e3PfkOG8934+f62bvd9zz93zOVnPWjts3LgxAAAAAAAAAABg+/Om7X0DAAAAAAAA\nAADwF3hRAwAAAAAAAADQEHhRAwAAAAAAAADQEHhRAwAAAAAAAADQEHhRAwAAAAAAAADQEHhRAwAA\nAAAAAADQEN5SK9xhhx3I3b392LBx48b+W+JCTenHN73ptfeCb3vb27KyZ599tsfX23XXXbPPr7zy\nSrJfeumlHl9va7Bx48YdtsR1mtKHu+++e7L32WefrOyPf/xjst/ylteWFu+LN7/5zd3aEREbN772\nM3feeedkL1q0aDPveIvQFnNR+6Rfv37J1nkTkfeBUqu3ww75MNfPf/7zn5Pt87z0XVuDdpuLiq+n\nutZqX3g/KV6m/fbcc8+90VvcUrTFXFTGjBlTLHv11VeTrX3qa6qW7bjjjlmZztta/y9cuPD1b3YL\n0c5z0fc0bX/d03Q9jsjXQu33iIgXX3xxS97ilqLt5mKNvffeO9m63tbWTe+3xx9/fCvd3ebTznOx\ng+ioudiulOZi9UUNbFeWbe8b2NLo5nbEEUdkZTfeeGOPrzdu3Ljssz5MPProoz2+HvwFP3joAVL7\n7bzzzsvqzZo1K9n77rtvsv0BYLfddkv2XnvtlZW9/PLLyR4xYkSyTzvttJbufSvRFnNR2/rcc89N\n9tNPP53V0xduitfTceEPJzvttFOy161bl+ybbropq/enP/3pde6696IPzxH5w1ftgXlzXl6NHz8+\n+6xrrfaF95OiD5EREevXr0/2Lbfc0uN72kq0xVxUvvvd7yZbH/Ii8hcyb33rW5O9dOnSrJ6WDRgw\nICvTfVH738fnu971rh7cdedR2xeVPn36ZJ+ffPLJZA8ePDjZ+rI8In+h4w/4Dz30UM9udtvQK+ei\nj3vtV//PCEXnx5QpU5LtL9y0v+fOnZuVXXLJJd1eu9WxVfu7bfmfHtA4euVchNbYoTa5ebO2Xblv\n48aNk7fEhbZ2P+oh8Z/+6Z+ysg984APJ1gfF/v3zl78vvPBCsvv27dvS9/phRh8wdcO9+eabs3rf\n//73k33ddde19F2bS2/834raA+att96a7KOPPrql6z3zzDPZZ/WE8kOOjgOtd8opp2T1fv3rX7f0\n3VuIXjMXa/z93/99sr/+9a8n+4knnsjqrV69Otn6smzlypVZvQULFiT7gAMOyMp0bt5www3Jnj17\ndlbv0ksvbenetwTbei7WXsYotT1YPdgiIo477rhkT5o0KdknnXRSVm/+/PndXl9fkkbk/0u8YcOG\nrGyXXXZJtj7gX3PNNVm9q6++OtnLly/v5ldsUdpiLurDvHoL6ktNR9dDX6N1vvnDpq6p+jLOv2vG\njBmvd9tbjN6yL+q493bV+a0v1NyjSdtf59RTTz2V1dO/8xd23/ve95L92c9+tqV73wa0xVwsMWHC\nhOzzgw8+mOzbb7892e79pH3nZyQ9K9deCm3LFzC9ZS5Clbaei51CaS4SowYAAAAAAAAAoCHwogYA\nAAAAAAAAoCHwogYAAAAAAAAAoCEQTBh6zFe/+tXs80c/+tFke0wFjRujtsfFUO12KfhhRB58VLXf\nEbluX7X4f/3Xf53VO/XUU5N9xx13ZGXHHHNMdDquuVYmTpyYbO9DjXFRi0OjmQ9ci6/a7FGjRiXb\nA0dv4xg1bYFm6dJgpDWtvMar8bmo8U08gKbGJRo4cGCy582b1/oN93I8tkCrcQd0PfWMQNoH2paX\nX355Vk/nqcbP8LmosWw8lpSurxpTbNiwYVm9iy66qNu/iYj43Oc+l+xVq1YF/AWNVaFjwftH9zu1\nNWBpRD4ufC7q9XXel4KGw2vU1sYzzzwz2RdccEGyPbbJGWeckez/+q//Svahhx6a1Tv++OOTrXG9\nIiK+853vJFvHSG3/JLhsa+jZQgNxr127Nqt35JFHJvv8889Pts83XQM/8pGPZGV6vtT4NX6mbucg\n+wDQM/CoAQAAAAAAAABoCLyoAQAAAAAAAABoCEifoCXUHd/TQ65ZsybZKluqsdNOO2WfNb2o2u6+\nq7IcT4NZup7fk7ozT506NSvT1LOeEhry9L6ezlddgFWGptKLiNxNXyVq3dXtYsiQIT2/WchQqdL6\n9euTrSm4I3JJm0oZfR7tueeeyfZU1Pp3OmfnzJnT09vutXiblKQImjY9Iu8nlapERLz88svJ1jnm\nqZZvvvnmZJ922mnJ1rU6Ip9vfn/aV5r++9FHH83qPf3008l2WdSFF16Y7A9/+MMBf+H0009Pdt++\nfZO9YsWKrJ7KXGprqpaprMqvscceeyR7v/32y+oddthhyb7vvvvqPwAy2dFjjz2WbB3zERG/+c1v\nkn3iiScme//99y9e29cEXwdKIHfqHh3b73nPe7IynQczZ85Mtu5vEblkWyWjKimOyKVPmtI7Ij/3\nqtTUz9Q33XRTsl0u7OcuAGhv8KgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIxKiB\nlvjSl76UbE/jqjEoPL3ovvvu2+31PL2oXkO1329729uyeqq/V81wRB77ROPQeBwUjR3hKRg1fWK/\nfv2S3cm6YE1ZqWi8jIhcH68xEzyts/avpwLXa+g4cx049Jxly5Yl+5BDDkm294F+Vr29pwzVPvbY\nJxp3Q+t1UnruWowajbk0dOjQrN7ixYuTrTGhnOeffz7ZPkcXLVrU7fVGjx6d1dM19O67787KdC3U\nGBweA2WXXXZJtqd81vX/7LPPTvall16a1eu0tMLnnntuslevXp1sjR0Vka97um4OHjw4q6fz1Oez\nxmvTa/iYOeKII5Ld7jFqSuPNY+dNmjQp2R6zRM8Vo0aNSvZBBx2U1Tv55JOT/dRTTyVb+z0iYsyY\nMcX7HTt2bLff6ynvNW6fn218XLQznu76xhtvTLaf5XRPevjhh5M9fPjwrN4HP/jBZOv80Hg1Efn6\n+O53vzsr+93vfpfsuXPnJvuoo47K6r3jHe9I9pQpU7KyX/7yl8leuHBhAEB7g0cNAAAAAAAAAEBD\n4EUNAAAAAAAAAEBDQPoELaFpPWupQV3q9J3vfCfZ3/3ud5PtrtXqBqxu3c8++2xWb/ny5cl2OYxK\nMzTl4sqVK7N6ev+aUjoid+PXtMWdLH06+OCDu/13lz5p26n0TO2IfLw4KpPSflIZGmwe6vo+e/bs\nZKuEJiKXBYwcOTLZe+21V7HeggULit+r0huVXrQ7NamBSiW8TVQ+6inRVfagc6WWOl3TA//7v/97\nVk+lSi5b1c8qo3A5qq6hLh3ROXzooYcm26VPnSB3UlTKonuhrqERuZRF102fs97uiqZPV9vH58CB\nA1/vttuG0ng78MADs8+HH354sl3iomuepmF2Wdruu++ebE0N/cADD2T1dI/zcaD9vffeeydb15GI\nfE/2/bndzzB6TnHJ0b/8y78k21Od6/qre5XX0/3vkksuSbaeEyPyvps4cWJWdtdddyV71113TbZL\n2FRq6tf49Kc/nWxP4w4A7QceNQAAAAAAAAAADYEXNQAAAAAAAAAADQHpE7SEutxrFomITbObKJ//\n/OeTrW7XnglI3UBvuummZB977LHFaz/yyCPZ5wMOOCDZ6o5/3nnnZfUuvPDCZHuWDXUvnzZtWrI9\nI0onMWHChGSrvMzHgfahjheXlz3xxBPF79KxpNdwV3/oOerur3JAn0fKGWeckWx1uY/Is5vccsst\nWZnKOdSN2yUamq2mk9C283nkWeoUnQelLHcR+ZxTWen111+f1VO3f7+GZhTReenyVpVIeUYoRWUk\nnYZKcSPyNlu3bl2yXc6rc1bXXs0aFpGPIZfBqXxKv9fHnUuaOxGXd+occMmf9pvON89GqZKjyZMn\nJ1uzbEVEPPTQQ8nu379/VqbyKc2Y6d+lcjaXT7U72rYnnnhiVnbOOeckW+VnEXn/aAYolSdG5HIq\n7W/PDqVz2DN56ZjRMpUYR+T96vvztddeGwDQOeBRAwAAAAAAAADQEHhRAwAAAAAAAADQEHhRAwAA\nAAAAAADQEIhRU8HjqKj+t5ZOVGMMuO5b0ymq/rmJlFJ+elrPWkyFn/zkJ8k+9dRTi/X69u2bbI1L\nc8EFF2T1nnnmmWR/4AMfKF5j6NChyb788suzehqjxlNFa5wGTSfbyaiWXvteY9JE5PEuNJ37/fff\nn9XTdJOqt4/I54tef8WKFT29bTDmzp2b7BkzZnT77xF5H6g+3uM0XXzxxcn2/tEYONrHmg66k9EU\nvhq7K6K+nmqMA50fnlpb45loPBxNyx6Rr5meIlbTNWu67wEDBmT1NAaOfldExJIlS5Ktsal8b9H7\nbUe8zUoxtzzem84XjRF17733ZvU0NbHHUnn22WeTrfudp4X3mDWdwm677ZZsjQUTkc8JP7/MmTMn\n2bXYTBozSOMFeQwZTaft5xI9b2pcL4/xpWuC78/tznHHHZdsXXci8vTpeoaMyPtH4wQNGzYsq6fr\n3I033phsT5GufTx+/PisTGMi6pqwdu3arJ6v54ruHZrSvd3TrwN0KnjUAAAAAAAAAAA0BF7UAAAA\nAAAAAAA0hLaSPqnbsLsQq2Rj0KBBWdmUKVOS/dvf/jbZm5sSuJbm8vTTT0/2V7/61c26/rZCXd8V\nlz7V0kB6W5d43/ve1+2/q3QqInfPdmmaurdqOlRPV9oqo0eP3qy/azc07bm6Z/s4UBdydRM+6qij\nsnrqxu0u3vpZ3X9rKb2hNdQVXtc2T7fscrQu3B1bJTrejzpPVWLhEoFOSgns8pcudN5E5CmCXaqk\n88/XP0Xnpraxpx9WCZLvmdrfup56n+k1VCLl6BiZMGFCVuZSnnbDU/3qnKidM3St1D5wucUDDzyQ\nbE8JvHz58mTr+PF07J00FxUdsy47VEmKz19Nw6x9WJOUqQxN+yIin3++36mUR+eRr7ulVOwRdUl+\nO6Apsz19va4vejaJyNvlqaeeSravZdqvGrZAZd4RuVzR56Leo+6zLju8+eabk63PDBH53Fc5JNIn\naCdafZ7fXI455phk33LLLW/4eq3i0uRW3jPgUQMAAAAAAAAA0BB4UQMAAAAAAAAA0BB4UQMAAAAA\nAAAA0BDaKkaNUtOwTZ8+Pft85JFHJlvjsnzzm9/crO9W7fIJJ5yQlXlqwCajqf9qqC7addcao8b1\n1IpqcpXf/e532ecRI0Yk+/HHH8/KTj755GT/4Q9/SLbGronIY9b4PakO2WN3dCqqwdb2qcWo+cUv\nftHStT3OhsdN6KKUKh5aR7WwGq/G+1HXQI1zoHEwIvL4GR6nStcE7WNfHzqJ/fffP9m6BnlcDNUw\naxtH5Om0tY1r6YF1jfP5pX3fv3//4jX0Hj32hY4ljcHhf6drh7ZFRPvHqBk3blz2Weei9rfHN9G4\nKLUYFHfeeWeyDznkkKxM+1j7w8dWu6dIL6H7m7eBxizx2F3alhpTxs8U2v46TzWWSUQep8Tns66v\nmpJbz5oR+dx8+umnszKNj6JpotsF7SuPxXXSSScl2+eRtq3GJPI1avjw4d3aGsMvIj+X6nk1IuIH\nP/hBsnWf9Tn79re/PdlTp07NyrT/fe8AaBd0f/K9qoQ/sw8dOjTZt956a1Y2Y8aMZGsMsBUrVrR8\nj7re+t6tfOYzn0m2x2M97rjjIiKf1w4eNQAAAAAAAAAADYEXNQAAAAAAAAAADaGtpE/qYu9uSJMn\nT062uyqqu6OmZP7lL3+Z1VP3Vnf1X7ZsWbI1ZZ66m0ZErFy5svwDGsbgwYO7/XdPlaa4+5bKh9QF\n2K+h6Uv/4z/+I9kjR44sftfcuXOzz+pePmzYsGR/4hOfyOppOnZPg6muz62mFm931L1a+7fmjnjZ\nZZcVyzQ1qEo5IjaVs3Wh8grYPLTvdC7W0tdr2axZs4r1fD1UN37t706WPqkbrrZPTRKqfxOR7zO6\nVrmEUD9rH/q+qNf3a2hd7UOXPmnaaF//tb/V9rS17Y6n01ZZiso6vX9UHvGjH/2oeH2VVHz84x/P\nykpp3P27SrLTdkfXLpc+aZv4GqfS8HXr1iXb98XSPuntr/3ka4LOP/07T+tc60Ov227cd999yf7x\nj3+clal8yCVNel7XtczlUyrt1tTdu+++e1ZP+9HDB+iZWp81PGWvylBdFqoSLz+/AjQdXdtaXSsd\nlRTefffdyfbnjvvvvz/Zvjbqs8a3vvWtZL/nPe9p6R4iynKns88+O/t85plnJtvXi67n1kceeaT4\nPXjUAAAAAAAAAAA0BF7UAAAAAAAAAAA0hF4vfVI3KnVDcldCjbSsbqQReYR9dUtyeY5+l5cddNBB\nydao0Z4pwN3Gm0wpC4hniVFXT3ezVrf7L3/5y8nWjCUREe985zuTrRHwDz744Kye9o9n0lDJ1OWX\nX57siRMndvMrur9f/W1+j52Kyo60P2tjWbNuOXfccUeyVYYWUXbTL0mioHV0bKsMpeZ+WpNFadYS\nz8qlWW10Xe5UeUVELmPRdvBMgJrJw6Wz2oc6/7xddR5pf3qWEK3nGZvU9V9lEy4B0ft3V39109f9\ns7YmtyPejzp3tH98TdU96Bvf+Ebx+iqP8P25dEaqyXw6CT3/eRvouNcMXBH5/FApm0ppIsrSNv8u\n7Y9aH+p80+xAEXlmPl/Xa5L13oqeD//mb/4m2S6B0N/u5zrtO93vPGuW9qPatXOin1t0PWx1LFx3\n3XVZmYYTOPbYY5N96aWXFu+jHdB55fI1fd5TOe+cOXOyeh/72MeSre21atWqrJ72vT/DKTova5mG\nHR2Prcp9moavJ/o7ar+v1k46r3Scq4QpIuK///u/k/2f//mfyZ49e3ZWT7Oz+TsBlRq94x3vSLbL\nCb/yla8k28Oh6ByeNm1asj3chtbzLMSPPfZYRNTDAuBRAwAAAAAAAADQEHhRAwAAAAAAAADQEHhR\nAwAAAAAAAADQELZpwJSaps3TEWqZ2h7DoqSr9hSVa9asSbanKVQdm+qVNW23f7fr7DQmg+pLXZuu\nMQJcM6fXaAKaqlCp6adr+t/Pf/7zxe/SetruBx54YPFvtE8j8pg6tVSUtfFU0k+2Ou46Ce/rUjpf\nZ+nSpck++uijs7KSjt714tBzNmzYkOza2qs64do8Uj2/95v+XZcGN6JnOu52Q9O76h7hGnjV2F91\n1VXFa2gfur5Z9xm1fc7q33l8FN0Ltd98TMybNy/Z7373u7MyvUf9zXrtTsD7R/d6bSONBxaR73GL\nFy9u6bs8LobOTR1rHk+o0/qkC50fnl5e287Pcto3mq65FpNBx0HtzOvjpRQP7owzzsg+P/roo8n2\nuBvt2L+6HmpMi7/927/N6p188snJPv/887MybTM9e/paOWjQoGRrnD0/C65fvz7ZHu9i4cKF3dbz\nVOAaC+OAAw7IyjSGo6Ykb0KMmtL5rRaHRc/W2pYafyci4rzzzkv2yJEjszJdN3WfWbRoUVZPx8jN\nN9+c7E9+8pNZveOPPz7ZvqfdeeedyW413orHA+utcWmU2m+olfmZX9G5qevXRz7ykayejrMhQ4Yk\n+4gjjihe22Pr6TWuvfbaZPuzhr5LOOecc7IyPQNrDCWNUxuRrxd+j137iq8VCh41AAAAAAAAAAAN\ngRc1AAAAAAAAAAANYatIn0qpuWruUDUXspJrnPOBD3wg2eriFpGn93KXRnVbVbdhd0VSV2FNE+33\nqLh7q7rojR49OiubNWtWt9fYXpTSczvq1nfjjTdmZcccc0yyV65cmWzvR3UTVDdfTxmreD+qK7K6\n+fo11LXN08SW0kCrPC5iU5fKTkHnsLd/q22i46Dm/g1bltWrVyfb02krukbVUo/qPHXZpqaQLa2N\nnYZKLDQ9s0oGI/L9U1NIRkRMnz492bXU6bq+6v7mMiudb34fKr+opfZV6YBLd/TvVA6p99QJ+Fmi\nNK9UyhGxaWreVnBJsJ6tanKLTp2nuhb6/NDxO3bs2KxMJYBq+xwotWtNdu37Ymmun3baadnnr33t\na8l2uYWPrXZA10eV1l9//fVZPR33p59+elam50E9m/gzyVlnnZVslSGOGDEiqzdw4MBk63odkY8v\nlWz484Suy7/5zW+ysj/84Q/J9v2hKfj4rT3f6V41adKkZH/qU5/K6s2fPz/Zl19+eVZ27733Jlv7\nUyVvERFTpkxJtsppfH6pBO4Xv/hFVrZkyZJkf/WrX0321VdfndXz+ddJjBo1Ktm+1+tz+rhx47Ky\nCy+8MNkaGsSf57VMz6Euy9Y11sekPiPqHnDFFVdk9bRffQ9QCd7y5cuT7c/BTz31VLLf//73Z2Vd\n56Lasw8eNQAAAAAAAAAADYEXNQAAAAAAAAAADWGrSJ9KLjzueqSfXQqj16jJnTQKs7oledRllS25\nG7dGg9YMJe6OqO57nh1A3ahK0i/nhBNOyD43TfpUck93F1p1F/3xj3+clanrobeZomNB26+U7SBi\n07ZVd3KVGbhL/yWXXJJslz6V8AwZnSp9UjmEZy176KGHWrqGRlj/7Gc/m5X5GgFbDp1/artsSfug\nb9++xevp3+l8i8hdUEtywnbH1y51r63JTHSOeeaWkgTJMxronqnz1PuiJk0uSZ/83hcsWJBsl33o\nWNL28LVD95SapKu34vJblR1pu3g2k09/+tPdXq8mLVDX/Ig8W41mfvPxOXjw4G6/q5NQyWZEvq7t\nv//+xbp6/vPsSjqPtJ9cDlI75+p6rXPRZW7a17Nnz87K2nFv1fABY8aMSba35T777JNsX7/0s66j\nfg2VKmk2Us/KpGPG12s9o2p2P99nH3744WR79ln9zRMmTEi29/f2oGsP2dysqJrFSrPoRNSz4pTw\n5xH/3IWHNvh//+//JdufEfS58F//9V+T7euDSs29f3Vc6Lz0san1XC77f//3f8nW0B5vlJ133jmN\ndZUpRUSsW7cu2b7OabvovXq/3XTTTclWyVpEnhFJ10dfl/WZTtvPw3WoZMqfW3Wu65z1s5Sec1V+\nFxFx2223JVtljd7fKlH19jj44IM3+Xun/VZuAAAAAAAAAIBeCi9qAAAAAAAAAAAaAi9qAAAAAAAA\nAAAawmbHqKnpXVXrrho71+TWUrYpmu7uve99b1amejLVyrseTTVorn3UNGp67663V1yDqalHtczj\nP+hvnjZtWvH6TUB1drV20dSHNZ2dtrPrLTcnLbP/jeo7tcxTEd91110tXVNT6NbS03YStdgaHhuh\nhGqpvW9KaWt9HkHP0XVJ44D4Wq6xK3RuO7reuq5X+9W1zJ2Cx7UqxS7zWCG6TnqZfladtscIUh20\nxrfw+aX9pvrziHy86L373qda/FpsAl1Pfe1WLfnChQuL1+iteKpWnRN6VvG5WEq/W0vtrPEtIvLY\nCar1dz1/be9uN7T9tc31HBcR0adPn+I1NM6SzrFamntdFz1mn85tnx86fjQOzX777ZfVq8UZavcY\nNRoXzdc5TYn7uc99LivT+aJpdL29tL9++tOfJvvQQw/N6ul9eNyS3/72t8m+4447ku0xLb7+9a8X\nr6/nbx1bHlNSf8u2YMcdd4wBAwZERN4vEfnar3ZEvmd84xvfSLbvaVOnTk32HnvskZWVUi17Hx55\n5JHJ1j3HY6BoLJIbbrghK9Nzj8bnfM973pPV09TsrcZY8f1ey7x/77nnntga9O/fPz7xiU9ERB4D\nKWLT9VHRftQU6b7PaN/5mUPPpTp3uuK4dKFroLaLn0N1XNTinerv8vGpcZMOP/zwrOyTn/xksvX3\n+75di7nbdd6ptW37rdwAAAAAAAAAAL0UXtQAAAAAAAAAADSE15U+dbnYurtOq7KlmqRFXaKGDRuW\nlY0bNy7Z6t7pLsTqUqYuUO6yWkrdHJH/Fr0Pd59UV0J1OfRrqLuau1Gpy7Kn7DzooIOS7e7L2wNt\nT3XLcimDuqt5qkJFx5BLXpRWZVAuRyqlmnWXwdr1S2nz3H2vk1D3TnW79Xb0VMIl3DVcKUmrkD5t\nWXRt01TBEbmLaE0Ooe6d7nKv66+7+HcKvu5ou6p7vNdbsWJFsn2PULmFpuatuUzruutrdy0drc5T\nvb7LivWzuzLrvqjXcJd0TZ/bjtInT52raUj1PKJu9RGbpl/uonb+uvbaa7PP//AP/5BsHT9dMoUu\nPHV7O1PaZ1yi4BIORc92ej7ysa3zQ+dYLcV6TY762GOPJbuWutnR/Vp//+amUm4Chx12WLJV7unh\nDcaOHZtsP38ce+yxyX700UeT7evc29/+9mQ/8MADyda04BH5eu73ccsttyR7ypQpyfbnmuXLlyfb\npU/a/yqvdanttpY+vfrqq+lZwPcZfa7ys4KOP10nzz333OJ3+TOctp8+j/h+dMUVVyRbpfoq391c\nLr744uyznoF8zJVkOP5MUwu5sLX698knn4wrr7wyIjZNJa0p6v3cqOm59Zld95yIPBW6SqQicrmT\n/p23X0lm5c/bOp40LXhEPjc1pMo73/nOaBX9zbVQKfr84nO9a53xdwoKHjUAAAAAAAAAAA2BFzUA\nAAAAAAAAAA2BFzUAAAAAAAAAAA3hdWPUlPSrqm/2+DKqLVPbU2epHs31XarXqqWS1VRfen3XtOn1\nPWaC6os1dorrFvW7/H41loNqW13Hp1o1TQ8XsamedXtTSnftaCq7kSNHFuvpNbwfSynda9TSc2uf\neio/162WrqH34frfTkJ18Nq/rvN3rXYJ12gqpfWmpv+EnqNrjcfFOPnkk5Ptumvl/vvvT7bG3IjI\n4xrVUrq3M74+6T6m65PPm3nz5nX7NxHl+E7exhqDSO9DY+NE5JpujytQSufrqWR1T5szZ05Wphpu\n3SM9xorHg2g3NDZCRMSHP/ysaXaEAAAgAElEQVThZOua57H1jjvuuGRff/31ya7tkbofR+RzsRYH\npZaKup3RNvGzocZA8fmsdfXs6WNbz7K1eDA6t/0apf722G0ai8UppQHuzTFqbr/99mTfddddyfZ0\nvrfddluyPe6a1tV10+dHKX6hjxmNZ1iLNaTf5WciXfd9XmrcDS1bv359bE9eeeWVFDdF05B3Kh7v\nqrfw4osvphipy5Yty8pqsXz0DKJniREjRmT1NB7dSSedlJX96Ec/SraOc4+fVnuG2ByuueaaZJ94\n4olZ2YMPPphsX5d17dQ56+u1vgfx5/6uMo+vo+BRAwAAAAAAAADQEHhRAwAAAAAAAADQEF5X+tTF\n8ccfn30eOHBgsj2tlLo2qetfzSXU05CqK7S6CrlLkaZpU5fGWopEdxNX91G9D08dpr+rRs3FW11k\nPUV1LW3x9qBV91hNaXjMMce0dD1H+1XtVlNpR+R9XmtLdQVXO6IsP1MX/k7jnnvuSbamX1f5RkTE\nIYcc8oa/y9Mulr4L3hiaatTliuqOevbZZxev8dBDDyXb5TCf/OQnk60urPfdd1/Pb7aX4muJrkm6\nD3h6bm0vdaOPKK9DvrbqPNL9zddx3Z9cmqz7pO7VvqcNHTo02YsWLcrKpk6d2u31Vd4V0f6yG293\nbU89m/i+pfNPpU+1/W3Dhg3Z55JM3eVmLotrZ0oSFz9v6Lrm5w11dVfXdj/X6TxSl/3aecjHi84d\nHTsuCahdU+dtq/LypqOpq3XtmThxYlZPU1pr6uCIPF30mjVrku1rra5zmqZYQzj49XxN1bmo1/N+\n1DO1jtWI/Ldof3uYBX9+AWiFV155JY0d35dnzJiRbF9rdF3S1OF6TozI58S3v/3trGzx4sXJ1nXU\nQ0+UzkE+31SC5e8EdA/VOeXPGtOnT0+2yqAi8rmp1/N3DKVwMBGbpkDvDjxqAAAAAAAAAAAaAi9q\nAAAAAAAAAAAaQlX61KdPnzjqqKMiIuLcc8/NytR12SNBa7TrktunlzkqQVIXKHcJVdcsded0Fyh1\n+3RXQpVWqWviQQcdlNXTv6vdu7qae7YadS/2iP21bETbA41CXZM+aduOGzcuK1N3uFIWkZ5Qc1PW\n+6jd76hRo5Ktrq4R+VjQ8drJWYduueWWZJ9zzjnJdsnjpEmTenxt76fSvOrNmSmagq6P2s6jR4/O\n6i1cuDDZNTmEunp6ZrUjjzwy2b7edgo+H3RPUlv3nIhcOjt58uSsTDOM6HrncqTSvuvzSz/7+qwu\nwGq77EYlj+5uX8oq5e6/+juvvPLKaHdUdqRjweebZ1PbHLTddUy6RMfPTO2M/nadAz62tU38vKFj\nXeew19O+1u/1evrZ10ydm3q+9vmmkhxHf9uWOIs1gXe9613J1v3tH//xH7N6v/vd75Lt8ltdOzWT\nobfl3XffneyurDgRm7al9onLQ1Q6oVIlz0SlYRYuuuiirEwzew0aNCjZX/nKV7J6S5cuDYA3wooV\nK6qfFX2u0j1H/z0iH/e+BurYVvm2r3P67KHX8GdqnVc+F3W90Oc+z56m56eaZNTl/4qu2S516pJs\n1sJ8tMdqDQAAAAAAAADQBvCiBgAAAAAAAACgIfCiBgAAAAAAAACgIVRj1Dz//PNJl9kVq6aL8ePH\nJ3vatGnFa6gu1lNwq1bLdVuqSVNdr2vENAWq6ts8pojGsnEtmGrsNTWqazw1RbmnES7py1zzrGnA\nVLcWsWm6zO2NxgWpxeRR7Z+npNWYCrVrlKjp9hzVGte+69RTT02297Gme9TreerDTuL2229PtsZQ\n8LG9OTGWfE0oaUA3Z+xAjs4lXVM9NkWrqdA1joLrfzVmTS1lbDvjemnVbWtsAU81OWvWrGR7mllN\ne1mLm6XzSPcqn0e6xvv9qjZb57rHwxk+fHiyr7766qzshz/8YbKvuOKK4nd5nLt2Z+bMmck+66yz\nku1pejUF9OaybNmyZKuO3mPUtEvcklbQ+VE7Y2gqZ43d5X+n88j3Rf1cO1PV4h/4nOti7ty52Wc9\nAzvtGKPmn//5n5N95513JtvP0pq6e88998zKdH/S842utRF5PEM9x3tb6pjx2G06ZjTeh+4NEfnc\n/P73v5+V3Xbbbd1+t/47wLbG18cSnq4bXp/2WK0BAAAAAAAAANoAXtQAAAAAAAAAADSEqk/6K6+8\nktz/LrjggmI9dzPU1KxjxoxJ9tSpU7N66jI9YcKErEzTd9bcVNUlVOVTc+bMyer9/ve/T/Zvf/vb\nrKyWglZRt+6hQ4dmZRs2bEi2yjlc2qHupy4xWLBgQUv3sa1QN113zVQOOOCAZLs7tf5GdTF1V96S\n26//e6suyzWpjI47lbpFRJxxxhnd/k2nphiOyF3nVa7n8j8dIyNGjEj24sWLi9f2FN8lmQzSpy2L\nylpUFhqxqSylRMmlPyKfL+oy3klccsklxTLdM3WuROTz5fTTT8/KNN2kXsPd79Vtv1+/fsn2dawm\niyqlJvb0lSqLvvjii7Oy/v37J1tlPK3uue3Kt7/97WTrnuP7oso0Wl1THT2DqMzO+9tTBLczeo4o\nyYoi8nPeypUri9fQvc/3NC3Teernl1pZKXW6ny91/6zJHNtFjjpy5Mhk61nTf/v8+fOTPWPGjKzs\nve99b7IPO+ywZA8cODCr96EPfSjZOi/9WUDPw35uVlmUyuw9ta8+r+gaGpGngtfztsusfJ0GgN4J\nHjUAAAAAAAAAAA2BFzUAAAAAAAAAAA2BFzUAAAAAAAAAAA1hiwhVPYXkjTfe2K39P//zP1vi67Yb\n7373u7f3LWxTNI5FLXWkpq52LbVeo6YFL5W5Vls/e5neo9qa6j0iYsqUKcl+9NFHi/ek1y9pxDuN\nWkwL1Uu3Gk/B0/Jq/CCNOdUu6USbwh//+Mdku46+1fghtfVB+8tjNkC+Z3qcLI0jsvfee2dlOic0\nzsTatWuzerpe6TW8n7QPfT3VuV5L2a5pwg855JCszOPBwV/Q9L4aT0hj80Xka+oRRxyR7J7EqNG+\n073a48l5zLFOwdtB0XnkMQQ15kttzdSzjc6x2vd6zK8SL7zwQvZZ71fnZUQeU6z23b0JnS8ay8Xj\nutx7773Jvv/++7MyPQPOnDkz2R43U/fMyy+/PNkHHXRQVk+v7+eWyy67LNn33Xdfsj1GzXXXXVe8\nvv5mjVPm/Q0A7QFPPwAAAAAAAAAADYEXNQAAAAAAAAAADaE9cvTBVkElC+r26enYv/a1ryXbUx+q\nK26r7rytypscleLod3n64ZtuuinZv/71r7OyL3zhC91eo11chVvB21j74Je//GWyzzrrrKyeuvke\nffTRyb7hhhuK31VLBa33ofIAeOPsu+++yXYJW6syM5XvuHRRr6lrRyej41nb2NdFnTs12Zi2q/fZ\nqFGjkr1kyZLiNTTVq897lcSpxML7U2U8b3/727MylT7p9X1db3dqa+r111+fbE3VHZFL00499dRk\n/+xnP2v5u3WN1XHiY6a2t7YbOrZr5xKV4t5+++1Z2f77759sTbvsMihNe15Ln61lO+64Y7FM8bmo\nKZr9+ip9ahdUJjp48OBk6/oXka9fJ5xwQlam7aTtrH0aETF37txk6/x1+ZlKWTV9eER+jlm3bl2y\ndR327/YU7MOGDUu2nsVdwgwA7QEeNQAAAAAAAAAADYEXNQAAAAAAAAAADQHpExTRKPLqHuzu+CoL\n2rBhQ1Y2evToZC9atCjZrcorau7YXqbyC3Xz9Yj66nLq96vob1Z303an5qZ/1VVXJfuDH/xgVk/H\nxemnn57sL37xi8Xvcpfukuyt1UxE0BqaJWifffbJylp1kVeXfpcPaAYZnW+djI7nmtxi7NixyfaM\ndbrW6jXGjBmT1Vu6dGmyVfoycODArJ66y/uarLJVXRNUjuOfVVLn6O+vrTHtiLet9t1vfvObZL/v\nfe/L6qm0RaUdPUHHkI4fzSAWsWmGsXZG9x3dW1wupPNDMwdFlOeE97Vm2tK56HOglM0nIp8fek+e\nwWjNmjXJ9vGi2Y1cWtVbmTNnTrLvvPPOZOsaGpGfTVQu5WUqHTvqqKOyenpWfMc73pFsz9SmGdmO\nPPLIrOz3v/99srV/VGIXkffVLbfckpUdeOCByX7mmWeSredrAGgf8KgBAAAAAAAAAGgIvKgBAAAA\nAAAAAGgIvKgBAAAAAAAAAGgIxKiBIpqOcsqUKcn2eCGqp/VYCU1nxIgR2WdNhahxNu65555tdk/b\nG9fYa+wfTberMUoi8vbydM0lHnrooezz+PHjk63xGTy2BrwxNC7G5MmTs7JW+07nimrlI/I4Chov\nBf6CxsLweDUaD0tjikRELFiwINnaT/Pnz8/qafwRjWngfauxKvw+tH9LcU4i8nmvcc287KWXXkp2\np8Woqc2pmTNnJltTnUfkMTM0/s8hhxyS1XvwwQeL19e5qf3jsah8PW9nSrHQfJ/RsX7llVdu/Rv7\n/3n88cdbqudxczReyowZM7Iy3Ws9rkpvZdmyZck+7rjjkj106NCsns4/nzurVq1Kts4PTb8eUY7J\n5vF+9BqeMlvj42gfDBkyJKun66OumxF5Km9dLzpp/gJ0EnjUAAAAAAAAAAA0BF7UAAAAAAAAAAA0\nBKRPUOTuu+9OtrpzenrWVqUSTcTdVtVVX92en3vuuW12T9ubWupgZfny5dlnTWepbr1Tp07N6qmk\nrpYOVfumX79+Ld0TtIbKF909u9X+VzSVc0Te/y7ngLrU5/Of/3yyP/OZz2RlJ510UrL33HPPZC9Z\nsiSrpylntW/Wr1+f1dPUwZ62tm/fvslWd3tPGa5pa7/1rW9lZe6230Vv3jM2h1alXb6mnnLKKclW\nqZKmB46oS5+0X32eKtrH7Y5KY1RepnZExJe+9KVtdk9bgm9+85vJ9jVBpXMqb+7NkhmVc5133nnJ\nPvzww4t/85Of/CT7rOcW3fs8RbrK0VQy7xJCPSu7xEzXPT1feh/Mmzcv2RMmTMjKVB6usuJ2l48C\ndCp41AAAAAAAAAAANARe1AAAAAAAAAAANARe1AAAAAAAAAAANARi1ECRlStXJvv+++9Ptqfnfv75\n54vXeMtbXhtiqv/19KxbE/8uvY+FCxdmZddee22yVa9+5513bqW7ax6tap2/+93vZp9VV/2zn/0s\n2RqTxrn00kuzz9rmmh741ltvbemeoDW03adPn56VaQr2Vrn66quLZXPmzOnx9dqdWowWTUt/wQUX\nFOtpnA1NwR2Rxxvp06dPsjU2heOxxzT2gsZO0XTSEZ0Vv2tr8+Uvfzn7vGbNmmRr/9x0000tX/Py\nyy9P9tq1a5P91FNPZfVuvPHGlq/Z29Ezi8YK0T0novV21jPG9owV8vOf/zzZPp89Hlw7oGvUL37x\ni2SvXr26+Dca16a7z1388Ic/zD7fd999ydZYYR6DTePG+H088sgj3da75ppriver3xuR7x0rVqxI\nNjFqANoTPGoAAAAAAAAAABoCL2oAAAAAAAAAABrCDjV3uR122GF9RCzbdrcDwrCNGzf23xIXoh+3\nG/Rhe0A/9n7ow/aAfuz90IftAf3Y+6EP2wP6sfdT7MPqixoAAAAAAAAAANh2IH0CAAAAAAAAAGgI\nvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAA\nAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAA\nAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgB\nAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgI\nvKgBAAAAAAAAAGgIvKgBAAAAAAAAAGgIb6kV7rDDDhu31Y3AJmzYuHFj/y1xoab041vf+tZk7733\n3lnZm9/85mS//PLLyd64Mb/1t7zltSG74447ZmXPPvtssp988slkv/LKK5t5x2+cjRs37rAlrtOU\nPtxll12Svddee2Vl2lfan7X29zK9xk477ZTs1atXZ/VeffXVFu94i9AWc/FNb3rtvfyQIUOS7fNI\n++BPf/pTsr3NtX922CEf5tqvL7zwQrI3bNhQ/K6tTTvMRW1nbf+3ve1tWT0t077wPtT1VOdsRN5v\nauv6HLFt+zDaZC5qP/br1y/Zu+22W1bvz3/+c7d/o/8ekc9t70ftY12/n3/++aze0qVLk721+7Td\n5qKuobvvvntWTz/X9i3tQ59jTz/9dLJffPHFZNf2z21AW8xFRft05513zsr69+/fbT1vc51/eiaN\niHjqqaeSvT3PpUo7zEVov7nYiZTmYvVFDWxXlm3vG9gS6OFj+PDhyf7whz+c1dMD6po1a5Ltm5ke\nagcMGJCV3Xjjjcm+8sork+2b5TZ+yG8rRo8enewzzzwzK9PD5R577JFsPZw4zzzzTPb5j3/8Y7JH\njBiR7PPPPz+r5w8ZW5m2mIv6kPZv//Zvyd5vv/2yevoQuGzZaz9dH9Yj8vnsD4f6YDF79uxkf+97\n38vqvfTSS63cesfiL8D0BcywYcOSfeSRR2b1hg4dmuwnnngi2fqQF5G/MO/Tp09W9sADDyT7/vvv\nT/a6deuyejrvt8GDY1vMRX0IfP/735/sadOmZfXWr1+fbH0ZoP8REZGPC/9PkL59+yb7oIMOSvY9\n99yT1TvnnHOS3e7zUueVnlF8vtXGr7a5rqHTp0/P6v3VX/1VsnV/8+/addddk7127dqs7Oqrr072\nvHnzku1nG51/fu+lc88bmKNtMRcVfak5atSorOxjH/tYsnUu+ks1PcvefPPNWdmvfvWrZOseuY1f\nsEFsOv9KZV5P51GD+q3t5iK8BtInAAAAAAAAAICGgEcN9Bh/w6zeE2PHjs3K9H+TJkyYkOzDDz88\nqzd48OBkq9xC7Yi6G/GYMWOSffzxxyd77ty5WT31tlm0aFFW1u7/k/hGOfnkk5P9d3/3d1mZelWo\nzM3d9PV/oPx/o1TipP9L+c1vfjOrt409atoClTuddtppyXa5hfajzoeaq7Z7aqgX1QEHHJDsn//8\n51k9l7R1Ir6e6tzZZ599srKJEycm+7jjjkv2sccem9UbP358svV/8fV/cSMi9txzz2S7BO6OO+5I\n9p133pns+fPnZ/XUY2rFihVZmXrzsLa+hnqGnn322ck+8MADs3q6Pur/9vtc1DXWJRuKek2pR0hE\nxP77759s9dpod/R/xWv/Q67tH5Gvm3p+OeaYY7J6M2bMSLb2u3o4RuRzc/HixVnZo48+muxVq1Yl\n29fdmkeNrjN4Fb+GelSpV9NRRx2V1fvQhz6U7Npc1HGhXo8R+ZqqfefnXPpn61DynqvNFS9rkBcN\ndAh41AAAAAAAAAAANARe1AAAAAAAAAAANARe1AAAAAAAAAAANARi1ECP+fjHP559Puuss5Jd09qW\nsslE5Hpd14IrquN+7rnnsjL9bs1uMWnSpKyeZitS7XdEHtnfs5t0KhrzQPta9b4Rm2Yi6e7vI/KY\nGd7XGu9Ibc9qs3Llyte7bTA0Ro3GQPCsXBoHSuOWqH4/Io+V4NfQz/vuu2+yPeYKMWryuBUREQcf\nfHCyNa5XRB6jRuNi+Ny79957k+2puxXNKqTxZCJynb7Gaxg3blxWb/LkycnW7FD+WWPZ+D7R7ngc\nIp0H2j+exUf3O72GxjGKyPdPX5d1DmtcFO8Dv2Y7U4pB4W2n+5PH8tL4I4MGDUq2nyn0u/Rc4mvh\nI488kmyfR3oNXcc9/tuGDRuS7TGh9CxWi8HRDvh80370ca79oOutrq8REddee22yNZaUnlMi8r67\n6667srKpU6cmW7OzLV26NKun63mtH6Fn6FgvzQcvc9p97kDzwKMGAAAAAAAAAKAh8KIGAAAAAAAA\nAKAhIH2CllB39y9+8YtZmbpmugu+ypPUPdtTwao7oV7PXZFVNuPoNTVloks21P176NChWdmvfvWr\nZE+fPr3b63UaAwYMSLamcH3hhReyetqu2l6a7jkidxf1caCf1T7llFOyep7mGV6fkSNHJlvnUU36\npP3oLtja/572Wee9SixGjBiR1XvwwQdbuvd2Q93vPbW2usf72qXyi8cff7xYT/tNU9m7i7eOA5eS\nqvu3jgldDyIiRo8eneyxY8dmZdrfKstw6WunuZCrXELXTe2riLwfdS/0/Ujnpq+3pTns9fxzp6Dt\n6rIYlW2qVMXLtI2feeaZrJ6O+4ULFybbpdVa5jJEXWt1DRg+fHhWT8vWrFmTlamsrpbGu7eia9t+\n++2XlY0fPz7Zvlaq9FDPHL4vqkxX55GfYVSWrX0fkY+1MWPGJFvlrhH5OvDwww9nZfPnz0927TwM\nm+53pbHekznQLvMFeg941AAAAAAAAAAANARe1AAAAAAAAAAANARe1AAAAAAAAAAANARi1EBLfOEL\nX0i2a3xVk+3aatVFa4pE18OrBlv1uR4XQ6/Rp0+frEzrajwc16nutddeyfbfolpmTYPp6RM7CdV7\nazyFmj5a29zjKeg1vP01LoaOiVqKYegeH/c65zQ+gvejzlm9hseL0vn28ssvZ2Uaq6o27zsVTf06\nadKkrEzjwfh6quuaziPXzesaqn3j6Xx1vmk/ReSxhTQGkcdR0XVS43ZE5L/t0EMPTfZjjz2W1fPx\n0+5oSmDtx1pskp122inZ3o/aV1rP0bntKWj9mu2Er4X6WdurX79+Wb3+/fsn29cubXO9nsd60pgl\numb63rdq1apke3wU/btavDw/EymluFW+dvTWGBx6rjvrrLOyMv1NixYtysp0r6rNnSVLliR7xYoV\nyfZ55LH7FF1Hdfz4+UbXh9NPPz0ru+GGG5I9c+bMZLfz/N1ceutYBlDwqAEAAAAAAAAAaAi8qAEA\nAAAAAAAAaAhIn6CIuvNOmDAh2e76rmkg3XVbUddMl1GoG6hKMfTf/e9UBuDX179z11R1s995552z\nMk2bOmXKlGR3svRJ5WDq/u2utqV29b5uVf5SS/ENPUfbU929fR6pFENlOD5XtF/dVV9dvEsu952G\ntpemtHa3d51HLtkoSQNdOqTrX00qoZIKdxNXKVRNvqZ97dfQ8XP00Ucn+/rrr8/qtbv0ydfAUaNG\nJVvb02UTuhfqNbwPtH+8LXWdro0tnfftgP4+by9tS5W7eBtomfdhSd7r0iedE3o97yeVkLsctbQX\n+pqs9+S/RT/X5n1tvWga+ntVZjlgwICs3pw5c5Lt7a5zR9dXT7OuskFty56shzoX9d79jKrfrZKu\niIhjjjkm2Q899FCy/eyN7AegPcCjBgAAAAAAAACgIfCiBgAAAAAAAACgIXSM9EndDD27hbqjqtun\ny27cPbGV73J32Zr7ck2usz3Ycccdk60u7DXXXkelDur2WXO3VSmG94Hi7afXr11D3YVdiqEyhMMO\nOyzZl112WfE+2p1DDjkk2bX5oWU6ft0FW/u+Nna0rOZ2jotv99QyAT3++OPJdldwlf/pGuD1dP55\nma4ROhfVfbzT0LbU7Ehvfetbs3q6PrkEQtuyJuHUeq3OWZ9jJbmTz+da9i9l+PDhxfttd0mcnzmG\nDRuWbO1jnx/aP3pO0bEUkfdrTW6h/eP9XcsY1BupncNKsj6XIZbOhhH5GNb9zutpm2tfex+W7i8i\n70O9Jz8n6lz06+v9aplfQz83fW/VdUnlhLWMh7Uzua6btd9eG1u1NVDHhvdxCZeY9+3bN9kq8aqF\nHQCA3gseNQAAAAAAAAAADYEXNQAAAAAAAAAADYEXNQAAAAAAAAAADaFtY9S4RlW1x0OHDs3KRo4c\nmexVq1Yle8mSJVk91f+6blR1vbvuumu33xuR64s9/V/TUpSW0nW6Flb1zq6LVs2vtkUtPbe2reu9\nVWvsZbUUmcqQIUOS7elQ9RoaU6GTGTx4cLJ9XinahxrvQOOhREQMGjQo2bUxr33dhJhNvZ21a9cm\ne9q0acn2uBilFOyzZ8/O6p1yyinJXrhwYVZWSl3ra14nobFoNOaXx0LQ/cP7RtdeXUM95kspPben\nYtc4CbW4YXo9v4aOF93fHP3Nu+22W1bW7vEVfC/dZ599kq1zzNc5jWmif/PUU09l9bQ9fW/Vc4v2\nla/lxKjZNF5ULY5IKUaNtrffh58bS9fzWFKllNw+XnRu+m/Rz3qNWkydpqPrzZ577pls7wOdH37m\n0z7R86vPj1JsPV839Rq1ODc6Jv27dA/wMajX79+/f7LnzZtXvF+ATqf27KI0cd7gUQMAAAAAAAAA\n0BB4UQMAAAAAAAAA0BDaSvpUS8GtbsMTJkzIyg4//PBkP/roo8lWV+2I3A3dJT7qLqrSDrUjItas\nWZPs22+/PStrmrxDU8gqfp977bVXsj2drLaTun+7i7y67KorqrsA69+5a6+6B+t9eD29Xy/TMTR6\n9Ohu/z2ime5xW4uBAwcmW92k3YW4lPr1wgsvzOpddNFFyfb2L6Wv9Pms13fXbegelR2pDNHdxEsp\nue+4446s3tlnn51sTRnq11dZS7unYa6h0hJtH58DmnLVJZy6rumccOmT9ltNtlSTfegc0+/19V9/\ni0uadI3QdddlpStWrEh2O66t/fr1yz7reUT72OdHaR/7wQ9+kNU74YQTku1tq+uj7mN+hlEZRTtQ\nkz7pb9c54OcSrafnCy/T+eb7UauSGf07l0jpHK7ti7W026XzV02q1fS5qBIhPa/7b99vv/2S7b9p\n/fr1ya5J00py3lobef/ofel3uTRSz95epn2n64iP8aY9T/QmtK/97K/trG1Me28fauu8nk18D1b0\njOpnrtqa0Mo9+X35Oatrj6+NHzxqAAAAAAAAAAAaAi9qAAAAAAAAAAAaAi9qAAAAAAAAAAAaQlvF\nqFFca6z6tAMOOCArUwCF9GsAACAASURBVJ3nk08+mWzX86tWzfVuqiFWnb7HDqilx2yaHnjEiBHJ\nrmnbVUPrGnvVEKs2z3+rftY28nqq4/P2K2kVXXOo9+F6Qf0+jbvh+vTN0S32FrxdNe2l/m6PUaPj\nQOfO1VdfndU7//zzk+3p61X7rWPH27vVVHvwGtpf2qeeAlp19dofixYtyuppmV4vIo8XoKmEfU1t\nZ3yM7r333snW/WOPPfbI6mnZunXrsjKNhVFKox5RjinTk9Ta2r81/bTOYZ/Pel9aNnLkyKzezJkz\nu/3edsF/r65tTz/9dLJ9Lmq7a59ed911Wb1Ro0YVv0v/Tvvb97Sahr+3U4tRo2c07ZeIfE+rnXt0\nzPp3lc42Phf17FSLk6BnFr9fvQ9Ps611a2e23oSetfX3aTy2iIihQ4cm2/tR55zHRFS0D2ox8lqN\nc6N4PEw9e+v+GZGPJ907OvlMVIpPWnvO0P3O206v4etkaQ77fG51H6v1Wy2mVSdRi/mi5wrdByMi\npk2bluz9998/K1u9enWyNV7sggULsnq6lviaWnou9lh9EydOLJZ17eW1szEeNQAAAAAAAAAADYEX\nNQAAAAAAAAAADaGtpE/qkuau+EOGDEm2uz6qy5G6MrkLo7oh+zXUFUttdWGMyF0rr7zyym5+RXNQ\nV3397bU0ru6yq+7U+nfetiXXYW9n7WNNdxtRTmW6du3arJ5KQPwaeh81iZT/znbCXbfVNVz7zdN6\nagrMxYsXJ9vd+TVFvaeA17mofaPp8yI62w10c9Exq67WPsfUlfTxxx/v1o7IXebdBb+UwlLnaLvj\nLtPqejt48OBk+xyoSYlK8k53sy5JpHxu19KQltIF+3fpZ08lq/NZ15EpU6Zk9f73f/832e24tg4b\nNiz7rG73KrdWGVRELr/VNXXDhg1ZPS1zXAbSha+h7djuXfjY1jlR2+f1s56HIvK5pHuhz3vta52X\nLh3WfvIyPTfq9Tylus5TX9f1u11O01vwftT+0XXIZQT6PODhCHTuaD+2esbwetoH/l0lCamvmxqO\nweel7qHax762txu1vUr3yQEDBiTb06Nrf+jc0RTt3V1f0f6oyQZ1vtXkkIrX07HkY6cdU4OXzjc+\nj4YPH55slTcdccQRWT09Z9XGTE2uuGTJkmS7NFLnn17vqKOOyup94hOfKF7jhhtu6Pb+lPae2QAA\nAAAAAAAAvQhe1AAAAAAAAAAANIReL31S9yiNpqxSp4hcRqGuxhG5PEddmdT9MCJ3eXPXNb2G2i7B\nKslzmsigQYOSra51LnNQvF3UHUxdgt1Vr9QW6vodkbumepYSvYbKcFRqE5G7GLtLurqf6fU8M0s7\nu4m7y7S2iY5fd0fU/rj++uuT7W6Fs2fPTvbBBx+clZVcEN0VHHqOtqG6ibuURcf2qlWrku3u5FrP\n+60kl+ukrE++Pqm7rsqdfG1RfC7q+qdlLrdQdL1291rtG1+7S3Je/66SdDgid+lXe+zYsVk9lZW0\ny9qqbvf+e9UtXrNP1CTB99xzT7JddqqZKmpSCaWW6cRd8Huja73+BpdAlMpqGV5cZlTKBuJtpfII\nrVeTzPg40P5W13m/X/1dNVd/nYu1taPp6BpbynQXkUt9fWyX2sz7R/tOx4yv89pXvreW5A2+zutz\ng5+3VR6pZT7GfY1oKqU28WcnPavvtddeWZnurdOnT0+2yz5Vhq9z4K677srqqdTe21HPNtr3fh6u\n7WM6tmp7a20s1cZ7k6llHtQ+0Wc4z9SscicdC36Wqq1tKovS9WHcuHFZvVmzZiXbpW56btE9Xsdj\nRP4+wp9pu+ZwLSwAHjUAAAAAAAAAAA2BFzUAAAAAAAAAAA2BFzUAAAAAAAAAAA2h18WocR2matpU\n++UaRtXwuXZQNZ8ay2bgwIHF73KtmuqXVXPpafdUx+gazFJK2+2F6uxUi+nxZFRH6elFVfun1NIs\nah97+2kbeZleU/WHnrpUtZ6uK1VteCleTbtT03VqmY9fbdff//73yfaxPHPmzGSfeeaZWVmp/VUf\n7mXQGqqB1TgHvqaqzrcUXyEiYt26dcmuzVPtU79Gu1FbMzSGmra5x1/Sz76eqrZdr+9zVmMX6Nrt\n3+W6d6X0W2opw/36pZgPHu9D9ejLli0r3lNvQn+7z7FSTBOPR6Hz7+677+727yPyNvPzjc71Utyh\niIjDDz882b62a8yG3kItvavGBNG+8b1K29nPMhrrUNNd1+Lv6VroaVp1LtbS+WpMqFosIY3H4fdR\nixPR5L21dm7UNvI1Ssezt0sprXUthlMtVXRtTdV71L7y+az3r/tGRD6GdG33NWZ70NWWtVThXqa/\nYcqUKcn+6Ec/mtXTmKF+ptdrTJgwIdlr167N6mkbaT953FKNzefPejrXa7GetJ6PuVLabR87tfTc\nWzMde9eYrq1DPu5Lc0KfqSPyvd9jyWqcF42R6vuixvirxT3U/vY4UNpf2o+TJk3K6s2YMSPZHkdG\n3zlo//iY0bXef0tXG/gYUfCoAQAAAAAAAABoCLyoAQAAAAAAAABoCNvfV65AyW3YXRrVTV/lLuoa\nFZG7R7krVinlsLvLqpuTumpH5O5M+l3u0qiSLE8D9vDDDye7CdInbWu9H3chU/d8T4en7aLuXy6V\n0HbS7/W+0rHg/VNyIXeXxHnz5iVb3b0j8r5TN79S2uh2xF0aS66Z3v7qZlqTLyxatCjZngKz5Da8\nZMmS4j1Ba6hEUce5uhRH5O7kOre9r3Su+zxVl1md2zW38HbD10l1z1Z5ikrIIvK28/VUr6Ft7utT\nq3ua9k0t1a+OCb+GuvnW7rfmhqwy2zvvvDPaAe2TmlxM293bRddKdcf39U/nqbuCl+ac/7v2q6cQ\n7Q3Sp5orfk0WouPc5Ug1GZC2SWm+ReR9pdJAn7M6DlxyUJIB136XyxD1Pvx39lZ0fdHx7PICnVc+\nlnVf03q1VMla5um5aynSS2mZva9UBuF9rGOtaVLirt/b6nyLiBgxYkSyL7roomR7Wuwf/ehHyXZp\nibJixYpk+3Oa7mP6vQcffHBWT9Mu+9p93333dftdvu7WUrhrG+hv8TVZr+HnL73mlk7F3tWPnu5a\n28/PfP369Uu2Sg1rbVtLba/fVUu7rWUuE9Q1z6XougZqH/j6oJJXl7/qfeicdcm635dy6KGHRkTE\n0qVLi3XwqAEAAAAAAAAAaAi8qAEAAAAAAAAAaAiNkT65m2Epy5C6TUXkEid1gepJ1PFS9HV37VLc\ndbTkMum/S13bRo8enZWpm1sTXBr13rWN3F1No5t7u6jbYKvuedo/7kKp3+0uxuoaqPXcXW3u3LnF\n71Z3ZO1Tl2q1M/5btQ/URdDnkUrKaq6pGzZsSLa7f+v80LLVq1dn9dwNFHqGzkXPkKdrj85tb3O9\nhs9FdT3XPm33flNXd18n1TVa5U6+Zup+567Qusfp9X2/KLmeb27mCL2e72k6n/03l+Szjv6umlyg\nN6Ft63u9tmFJUhGRu9brmuptomPL116tq+PE21mv0YQMMj3Fx3JNglLKflU7v9Rc52vtWjpf+t6n\n+24t45Bez+eUz78StQyjvWm+6XzR8TtgwICsXk1OqrKE2hzTdqqdUWuZ1fSatcx/moXIz6+ldXR7\nZ+vaaaedUqZclbJGRKxZs6b4d9OnT0+2ymQ8I6ieRfw5UNt5zpw5yXYJvsq8dRx4xh1tY8/+q+2s\nzwieiUrnus9LfbbUserjRWVuns1P991axqCesvPOO6cwIl/84hezMv0el/Non2gbuSRMf4evgSq5\nbTUjpV7f21nXh1pGLW1n3z9r+4N+X2l9iMjnt68/XVmfams3HjUAAAAAAAAAAA2BFzUAAAAAAAAA\nAA2BFzUAAAAAAAAAAA1hmwqRazph17uphk81mpqC2+vp9Vzvpfox1572798/2ao5dJ1ZLd2aXl/1\nbq5fVV3cAQcckJXpNfUa2wvV99Vi5mi8hVoKOY1j4dcrpf92Xbi2n/eP3q/GyPAUfTfffHO3f+Pf\nreO1ll6t3fB4I6VUkZ5KVsdBTV9ai1+jMQG0fx9//PGsnvZNb9LUb0+0nXR9cb239qvqaWsxajwd\naifFdFK0jby9dNyr7fuiarhrKXbVdq13KV6B70e1lMC6dvtcV9avX1+8vq6brlVXdL3orTFq/L41\nLo2Xaf/Xzia69tb2YC1zHX1p73KtvI67Wny+ptHVtt52Oqb8PFUaz34NTTPr7ahzrrYfaT21a+Pa\nx0sp1kItzo3PRf2+0j01ndozhM4jTRXs+FmiFEPN51spNomfl2rPENqP+l0ef2TVqlXJnjhxYvEa\ntfPwtqZPnz5xwgknRETEMccck5VpPBVfW0aOHJnshx56KNnnnHNOVk+fBby95s+fn2yd6z4O9Ls1\nNp/HeNG+0TUgIn8G1Zg3GhvHv3vo0KFZma4l2m9+btbYPj5ur7zyym7rvVH22GOPOPnkkyMiYty4\nccX7cXS+aIwl7wNtT4+3pHOnFgOxNBY8npNez9dbravt7s+E+t219N96VvN1Ssek70X+fd2BRw0A\nAAAAAAAAQEPgRQ0AAAAAAAAAQEN4XelTyYVa/91d7kopSt0VU92XVH4Ukafd1rKaa5O6tWmaL78n\nd19StzS9X3fZ0jJ3VyqlYHTpjpaNHz8+K9Pve/rpp2Nb431dShfp/a3ucC55ULcxlUe4e3bJrdQl\nZu4Cp5TSont6vZrsoHR9v0Y788QTT2SflyxZkuxRo0Yl290W9e9q7tTqZuop27U/9PqamrG774bX\nR9tM12JfD3Xu1NYhlbzU3L97Y6rfLYG7MWtbquzEJUE+J5RSCsiaVKImb9Ky2v5c60/9XT4v1b1c\nXX5deqJlfo+9Za67a7Wma505c2ZWprKoUorhiHyc6NmhJq+ppR/Wv/M19Z577km2Sgl6Cz4H9HMt\nhXIp9XVEft70cVna42rSPf0b/y7t35pkXm1fY/T85fenc6wmQ2wy3ra6Rmn/qIQ6IpdlP/bYY1mZ\nnku1r2rPNTW5gtarjcma9GnBggXJnjBhQlam40T3ilYkFFuT5557Lq1zfobUexszZkxWpmNdx6XL\nhWrt35XiOCLvT5ct6XxW2dIjjzyS1auFr1i5cmWydSz5eqqS8tqepn3v81LlTi598s9biueffz7t\nBb6X6Gd9Ro/If28pVXdEft8eBkHTumtfebvo86POB5XYReR95+uhnlt0vNb2AD+36P1r//v96nf5\nu4muUBze1goeNQAAAAAAAAAADYEXNQAAAAAAAAAADYEXNQAAAAAAAAAADaHlGDUeK0T1fapDj8h1\nXBrbw9PdqXbadVuKxj9wzaHGpVFtYu16tVgspfTAEbke1OMIqI6tFGcjIk/h5WnqavFXtgX+e7Wt\ntY1cH6ravAEDBhSvWdP/lmIZ1WIq1PS/Ws+1lKW0mn4fOl4HDRoUnYKnj7vrrruSre3gek2dO7XU\no9qu/l2qOS/F2YDNo5Q6upZ6WTW/3qeq1XZ0vfA1u53R9crXc9Utq07b4xPo3uIxv3TP0P6ozQ9d\nT30f1/3Jv0s/69/5mqzaatec656s64Pr/vUM4eOxN6UPVlRHf9VVV2Vl733ve5Ndim8Xkbd7LVaP\nlrnWXdu2lsJX4y14utregP82HUceK1DbVceot7HGDfQ5Vkq17ddoNS22zolaqvFSPL+Icjwivy9t\nq1oslqbh96prbC1GjaYL9vmh65y2US1Nr44f7wP97GWlfqylZfZ9RO9X9w4f47WU8VuDl156KRYt\nWhQRm6Zx1ra87bbbsjK971o8u9o+pnur1vPr6Wddn3V8ROTjwL+rFGfK45K0Gk+z9rxYW1d8r91S\n/OlPf0pnO48lquO0q6+7K9PfXovZVYtpV/qbiLwP9PnTr6f35HOs1Me+Lutn359Le0BtDfVzVteY\n9GsreNQAAAAAAAAAADQEXtQAAAAAAAAAADSEqvTpTW96U3LTmTJlSlamroXu8qUuiCpBctejPfbY\nI9k1OY26X3k9dS2suR6pm5i7I6obesmFLiJPE+7uUfo7ay7KKiXwdnMXu22Nt5neX81leu3atcke\nO3ZssV4tTaz2SS1lbC31Yclt0l3NWnU11z7e3rK0bYm7GaqrcE2iMHny5GRrO9ZSsbu7rrpu6zjw\nPoSeUxr3Pra1v2qpotW92ddllYLW3PHbjVoKUf3t2sYqh43I9yq/Rsml1uuV0g/7fNPv9mvoZ927\nfUzob3HJjEqkdUz4mqB7TZOlF5uLu4lruk5tI5+Lw4YNS/bs2bOT7fINbTPfW0tybl+/dR/vTfO0\na5x622mb+JjVOVFzj9d9x2W6OmZrcuqS3MnPiTonWp0Dfn5RGVdt/dFxUPuummxhe+C/SfuxFiJh\nyZIlyfaze0ki5+dJ/W6dUy6z0ucJv0ZJfuZzUddRP/vob9O9Qvt+e7Bx48Y0R2rnhtq5XdukJgPy\ncaB19Xq18VJbH7TMn8v0PnS9qMli/DfX5OYlvD22liT41VdfTf3o+4zOMV9vS/Lo2l7va6pK+fT3\n1tYo7buajFPPoX7/Ou/9ubwmUSxJSP2cpWNG5bQRr73rcLlg9j3FEgAAAAAAAAAA2KbwogYAAAAA\nAAAAoCFUpU+77LJLTJw4MSIiPvKRj2Rl6ubk0afVhVZZv3599lndgdxtTN2eSlHzI3KJ1N57751s\ndy1Tt2t3EVT3K/07dUmOiBg+fHiy3T2q5DLprliaBWvw4MFZ2ejRo5M9a9as2Na4W+sTTzyRbO0f\n74O5c+cme+rUqVlZLcuBUpI0uatZzcW75Gro7qfqVqq/MSLPWqW/2TOzdBJLly7t9t/dHbGW1UWp\nuTTWsnrBlkPdqd0lVN1da5Hodd3U9TUiX4tV4tpJ1PYq3T9r7r+eGbCUGcbnSknOUZPl1uasUpNj\n6W+MiFi9enW3f+fX0HrbW16xNXAXcpVH6JnA9zudVzpnvY1qGRVL2bt8bKmLem/pgze/+c1p31Fp\nekRd1qfzQM+vPme1HbwP9QxYWye1T0tZQvwafr+lOawSOq9Xy1Kl+NpRytTpbI8x0mqGKperLF++\nPNl+dldqchVt91Kfej0v0z6uyfg1RII/X2ldl5U0hdo+45QyYdXOf359/VzL3KX9oWOpNn+9jUtz\n2OdDbZyVQkn0ZE71pI17wiuvvJLWFZdl12TUpX3Gn7dr1/BzZIlWn/X0s0uOSv3ofaOffTyV5Kre\n91rPJU5dZwE/B2b3UCwBAAAAAAAAAIBtCi9qAAAAAAAAAAAaAi9qAAAAAAAAAAAaQlWU/uKLL8bD\nDz8cERE33HBDVqZpI8eNG5eVTZs2LdmeJk9RTZvrAJctW5bsWtpqvYamyPa4C6qB91RsGrNEdWLj\nx4/P6mmMmlZjp9T0q7X0jE2glLbQUZ2068RV+6d6Qf/t+l1qu15QNYI1jbdSSyPnunONG9RqzIZ2\nZ+XKld3+u7ePpu7Wee9xK2pp2kttrteGzUPnSyllr1PTQfsaqJTS33YSru/WuAOlNKEReXv5fNB9\nTW2PY6BzTNc7v57uOb7PlvZd15zXUluW4vL4mNA9uR1jU7m2XeP4HXroocn22Hp6ptl3332T7em+\na+mWSymbPe1vbR1oKhs3bkxjqRY7qZYeV21f7wYNGpRsj71YihFVizNRa+NW47PVyjQemJ+9S3H2\nfE2otdvWiouxuXifd+F9sGLFimJZ6TfV9q1W1yi/Rimmic97PZc++eSTWZmut6VYnr2VWswXpTYO\ntawWe2Zrs7n3X8LH3Nbs767n1lpcUX1ujijPiZ7sK6VzUS0mW6tpvBcuXJh9Lj2L175rc/ugld9S\n3Tda+hYAAAAAAAAAANjq8KIGAAAAAAAAAKAhVLUdmqbrJz/5SVamchKXGam76JAhQ5KtbrwREaNG\njXrtRsz9UuU06tZdc29Vt2t381ywYEGy3ZVQ3a7VXf2www7L6n3qU5/q9p4ichdHdTX2enqP6gof\nEXHzzTfH9sRdr/Te1X3NpWOadt1lRiVJWE3yUpMF1O5XUVc2H1v63S7rOfjgg7ut57+rk1i1alWy\nNZ25ymciyqnwPL23ulP6/FCJUzu48jYVda1291ZdO2uuw1rP1wSdty4Bamd03fHf7ftOiZpEqCRH\nqrkXq6u892dNjlqaf74m6zV8T9P9TmVQ/l2LFy8ulrUqK2ky3u66pmqb7bPPPlk9lScNGDCgeH1t\nM039HZG3mc51H58liWuTefXVV5NcXeUtXWVd+F6lZ7RSyvuIXOLv7apylVbPLLVUr6Wzkn+uufqr\n9Mnlwtq/myMhbwK+Nui96u/wVLc16VOpbWvf3eqaVFsr9Xq+ruvc1LANEeUwBK2u39A7aXp/lu6v\nJ9L3rSmTb/K69nrgUQMAAAAAAAAA0BB4UQMAAAAAAAAA0BB4UQMAAAAAAAAA0BBeN/9wl+7MU0fr\nZ0/TpZrKe++9N9k1vWYtBXd399OFatpUf70l9JqejvHWW29Ntmtg9fqq//X70Pv1sqalPtTYM3rf\nHo9C41247rqUbq8nKSGVVvXEtWvo382aNSsrO/HEE5OtmkaPvdBJaLwojfWkqcwj8jbXMl0DXo9S\nmtlW43tAGZ1/Op5dF6yxRGrpGXUN9DVBr+lxiNoZXVu8TTSmj6djVTR+hq9j2h8a/8X7UOtpme+D\neo1a6m6NleIpuGt9rbEWtJ7HkFu3bl2339vdPbcDOsfWrFmTbE3BHZGPk6FDhya7FpPIY8/oWNDx\n1C4xLbrGlZ/JFD+XaF3d3zz+h8Zf9DGr7aVj2/umdE6ppYGtzWe9vq/P+l0ar8avX0pN6/VaTUO9\nrfA2K7WLjwU9P3isp9KZw9fo0vnVx5bGM/Sy0n7qMRC13mOPPZaV6Tqg99GTNMgA0HtgZgMAAAAA\nAAAANARe1AAAAAAAAAAANITXlT5tSWppAJuYOqsmCegESuknPU2lulp7G6krvEoB3MW45MLp7py1\ncaIu81rPXXvVDfbBBx8sXk//TuVdnYb22/z585N97LHHZvW0D4cPH55sd5+uud+XpG3udg5vDF3b\ndt9996ysVQlmSV4TkUtlOrXvvB11LXSZkaLu8j539Jo1qUopzWVNcuVu+dqHin+v3q/LvfSa+ncq\nN4nI5VRNkwBvCfw36X6iEmvvAx0nKtmoyTJ8vpXST3tf+efeTk1qXZoHnsZbJfl+7imdX13uUpMZ\nleq1io8XXWP69u2blelvblXe1JvkcNpXfg7VM6qfPbVddM3zetrWKonzsA0qY6qt89rf/l3a7iqN\njIgYO3ZsspsmTWsCtTT3JfkfQJPBowYAAAAAAAAAoCHwogYAAAAAAAAAoCFsU+kT9C5USqQunB5R\nX93sf/7zn2dlkydPTra6B7tb7l577dVtmbveqgury6A0Or66lrsLv2azWrFiRVam36duk57ppJNQ\nF9GHH3442Z7tTd13R44cmeya+6lniSllzfFMJvDG0Lnj7tk6n2uu1aWMexG5O7jLXDoFd60urSHe\n/rpe+dzRdlUXe5c6ab/p39TcvX2tLWURqc1nn6c6h3Wu1zJHdQL6+zXj1RNPPJHV22effZLdr1+/\nZPuepu3u0iftc533nkmvljWpN+DjV+eAzzGVOO25557J1n0rIs/A5nNH97uaXLHVTJW1zFH62aVV\nivavjp2IPKOY9r3+xoj8/psufdIzqvZHLfOZt59+Lq2vEfkY0jZzGZ1eo7a31qR5iq4P/n2tjq1O\nQtvE+6aWGbhVtvT86E3zDbYPeNQAAAAAAAAAADQEXtQAAAAAAAAAADQEXtQAAAAAAAAAADQEYtRA\nwvWRGoNEtbXLly/P6qne/uKLL87KvvOd73T7XbXYC5r60PXeGvPA4xyU0ovW4mxovJqIXPOsv9lj\nB3QSOi4WL16cbI9Ro3rgIUOGJNv119ofmpo2ImLcuHHJVp15b4+f0DQ0joWnF92ceCH+N8R32hTV\nxJfWmYh6ql+NjaC2xwjSuaPzspaC279LYz5oPY9pob/F44Zp3+s64r+507T5+vs1lbDGVouI2G+/\n/ZK9++67J7tPnz5ZPV0f/Ro6F7Wd231P09/q474UH0Rj5UVsmpJbKcUH8bON9vWWSNWt+6efbbTv\nPQ6gxjjS8VKLj9I0fJ0onRs99p2uSz4W9Per7fFNtK01fo33t6faLl1Dr+9rr96jz2e9hv5dp62h\nJUrzLWLz2sjn2JaOKUO/weuBRw0AAAAAAAAAQEPgRQ0AAAAAAAAAQEPoPT6PsM1RWZC6ELrkpeZW\n2ioqW1JZhrsdbukUhC6pKaXnLsmqOg1NFelpYAcNGpTsCRMmJFvTn0bkbsnqrhyRuwOX0lDC5qFz\nSdvdZRSDBw9Otra7z0V1O1e3+ohcpuF93Kno2qXrpLs+l9bCiLyvNN2uyxf07zQVsbvY6995yucB\nAwYkW+ei3+/q1auT7etzyQ29E9y9db74+qVl2lcuxR0xYkSyNd2yS3TWrFmTbO8DHTPax3vssUdW\nT+UcLmHr7ag8zz/r2HZJ3vXXX5/sBx54ICvbsGFDslW+5u2vballNUm2z2f9XJPWKC7d2XvvvZOt\nkuPafTQNXzdK7elyW51js2bNysp0TdQ54OeWktTNpaC6ftfWdh0XHk5AZYlPP/10VlaSZ/UmCdu2\nYmtIkzph74JmwdMPAAAAAAAAAEBD4EUNAAAAAAAAAEBD4EUNAAAAAAAAAEBDQNQICddeXnPNNcnW\nmBO33357Vm9z0vm2eh9bWw/q6Td/+tOfJnvp0qXJXrZs2Va9jyajfaB66auvvjqrN2nSpGRfccUV\nyfY2Vi31VVddlZWpRlzb3NN4Q8/RfrzpppuSfcIJJ2T1tKwWm+nJJ59M9m233ZaVaTyEOXPm9PRW\n255arC2dY96ujz32WLIHDhyYbI83orELNIaCxquJyONseHwxvYbGePB4Cvfee2+yPeaX7g2drO2v\nxVrTeBravxERrHmK8QAAAT1JREFU8+fPT/YjjzySbF8PNX6Gx5DT+aff6/NyS+/jTcJ/m84xjQfi\ncWhuuOGGZK9cuTIrK8Xj8zhQis6B2pjwGEH6d7rurlq1Kqu3ZMmSZHtaZ41xpLGQNL6Of1fT8H4s\n7S2+b2lbeMwXXQN1fdQzb0QeG0jXW48ZpP3jacL1s87TF154IaunMZQWLlyYlV122WXJnjdvXrI1\nThUAtA941AAAAAAAAAAANARe1AAAAAAAAAAANIQdam6OO+yww/qI6FzNx/Zl2MaNG/tviQvRj9sN\n+rA9oB97P/Rhe0A/9n7ow/aAfuz90IftAf3Y+yn2YfVFDQAAAAAAAAAAbDuQPgEAAAAAAAAANARe\n1AAAAAAAAAAANARe1AAAAAAAAAAANARe1AAAAAAAAAAANARe1AAAAAAAAAAANIT/D1DkxfyYZP8z\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1440x288 with 20 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10  # how many digits we will display\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "    # display original\n",
    "    ax = plt.subplot(2, n, i + 1)\n",
    "    plt.imshow(x_test[i].reshape(28, 28))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display reconstruction\n",
    "    ax = plt.subplot(2, n, i + 1 + n)\n",
    "    plt.imshow(x_test_pred[i].reshape(28, 28))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "AutoEncoders.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}