4.Cifar10_l2 regularization notebook.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python [conda env:tensorflow_p36]",
      "language": "python",
      "name": "conda-env-tensorflow_p36-py"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.4"
    },
    "colab": {
      "name": "4.+Cifar10_l2_notebook.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "41cLHC84Yez2",
        "colab_type": "code",
        "colab": {},
        "outputId": "5b433876-3105-4a9a-e641-e51b9ead9a30"
      },
      "source": [
        "from __future__ import print_function\n",
        "import keras\n",
        "from keras.datasets import cifar10\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Dropout, Activation, Flatten\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.layers import Conv2D, MaxPooling2D\n",
        "from keras.regularizers import l2\n",
        "import numpy as np\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
            "  from ._conv import register_converters as _register_converters\n",
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iOdsLHwOYe0A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Defining the parameters\n",
        "batch_size = 32\n",
        "num_classes = 10\n",
        "epochs = 50"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lSEpmxn0Ye0H",
        "colab_type": "code",
        "colab": {},
        "outputId": "093e1373-b3de-4feb-f360-9b8b3fcb7631"
      },
      "source": [
        "# Splitting the data between train and test\n",
        "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "print('x_train shape:', x_train.shape)\n",
        "print(x_train.shape[0], 'train samples')\n",
        "print(x_test.shape[0], 'test samples')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train shape: (50000, 32, 32, 3)\n",
            "50000 train samples\n",
            "10000 test samples\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3tDaTOIbYe0P",
        "colab_type": "code",
        "colab": {},
        "outputId": "d9caa169-357f-45e9-8800-a7fe4bc5ea16"
      },
      "source": [
        "# plotting some random 10 images\n",
        "\n",
        "class_names = ['airplane','automobile','bird','cat','deer',\n",
        "               'dog','frog','horse','ship','truck']\n",
        "\n",
        "fig = plt.figure(figsize=(8,3))\n",
        "for i in range(num_classes):\n",
        "    ax = fig.add_subplot(2, 5, 1 + i, xticks=[], yticks=[])\n",
        "    idx = np.where(y_train[:]==i)[0]\n",
        "    features_idx = x_train[idx,::]\n",
        "    img_num = np.random.randint(features_idx.shape[0])\n",
        "    im = (features_idx[img_num,::])\n",
        "    ax.set_title(class_names[i])\n",
        "    plt.imshow(im)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdAAAADECAYAAAAvbXA5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsvXmYJdlVH/g7EfH293KprH3pqt43dUstoaVBSyM0YMRiWYBlswwYhIcxNptZjC0zYgwWHg9gGM3AgDH6EIjlk2CEAIHQgtCKhFqtXtV7VXXtlXvm2yPizh/nnLjnZWZlVWa97qpu7u/76ntZcSNu3Lj3xo17tt8h5xwCAgICAgICtoboSjcgICAgICDg+YjwAQ0ICAgICNgGwgc0ICAgICBgGwgf0ICAgICAgG0gfEADAgICAgK2gfABDQgICAgI2Aau6AeUiL6DiD50Gdd/DxF9cpxtChg/iOhviOitFyi7hohWiSi+2LkvJBDRUSJ6wwbHX0NEj26xrncR0c+Nr3UBAePDC3l+XtEPqHPu95xzX3sl2/APBVfrh8k5d9w513TOZVe6LVcDnHOfcM7dfKXbETCKC214Av5h46pV4RJRcqXbEBBwNSG8EwEBjKvlXXhOPqBE9O+I6EkiWiGih4non8jxERUsETki+kEiehzA4+bYDxHRU0Q0S0T/lYg2bDcR/QoRPUNEy0T0BSJ6jSl7OxH9ERH9jrTjISL6ClO+n4jeR0TniehpIvqhZ61DLgOb9OXbieh3zXlHpO8SIvp5AK8B8E5Rl75TzvlKIvo8ES3J71ea6/+GiH6OiD4t13yAiGaI6Pekfz9PREfM+ResS3A9EX1Orn0/Ee1Y284LPO/3EtEjRLRARH9FRIfH1JVXA14uY7hARL9NRFUiuoeITugJIvn8FBHdD6At43kXEd0rc+APAVSv3CM8/0BEh4joj+VdnyOidxLR9UT0Ufn/rMzzKTn/3QCuAfABeRd+8so+wdWNzeYnEX0jEd1HRIuyttxpyi64Bsv69l4i+l0iWgbwPc/pQ10Izrln/R+AbwOwH/zBfguANoB94E74pDnPAfhrADsA1Myxj8mxawA8BuCtUrb2+u8EMAMgAfBvAZwBUJWytwPoAXgjgBjAOwB8VsoiAF8A8DMAygCuA/AUgK97LvpnTH35dgC/a847In2XyP//RvtN/r8DwAKA75L++ufy/xlz/hMArgcwCeBh6fs3yPm/A+C3t1DXSQAvAtAA8D5t62btBPCPpQ23Sr1vA/DpKz0GYxrHowAeBHBI+u9TAH4OwD0ATqw57z45rybz8xiAHwVQAvCtAIYAfu5KP9Pz4Z+8+18C8MsyF6sAXg3gBgD/E4AKgF0A/hbAf1szDm+40u2/2v9tNj8B3AXgHIBXyjh8t/RrBRdZg8Hr2xDAm+Tc2pV+Vufcc/MB3aCT75PF8Xuw/gP6+jXnOgD/yPz/XwH4iPw9cv0G91kA8GIzAB82ZbcB6MrfrwRwfM21Pw35QFzN/0xfvh1b+4B+F4DPranrMwC+x5z/H0zZLwL4oPn/NwG4bwt1/cKavh/IS3TBdgL4IIDvM9dFADoADl/pfh/DuB0F8APm/28E8CQ2/oB+r/n/awGcAkDm2KcRPqCX2u93Aziv822T894E4ItrxiF8QC/evxecnwB+DcB/WnP+owBed7E1WNa3v73Sz7f233OiRyai/xnAj4EXSwBoAtgJYCPHkWcucuwYWALb6D4/DuD7pNwBmJD7KM6YvzsAqqI6PAxgPxEtmvIYwCc2fqIrh036cqvYD+5Li2MADpj/nzV/dzf4f3MLda0dwxIu3u7DAH6FiH7RHCOpd+39no+4pHm95rz9AE46WVXMtQGXhkMAjjnnUnuQiPYA+BWwqaMF3qwtPPfNe95js/l5GMB3E9G/MWVluSbDxdfgjb4NVxTPug1UbFa/CeBfg1V6U2DVFV3gko3Swxwyf18D3uGsvc9rAPwkgH8KYFrus7TJfSyeAfC0c27K/Gs55954Cdc+Z7hIX7YB1M3pe9dcvrZfT4EntMU1YFXrVnEpda0dwyGA2YvU+wyA/2XNuNScc5/eRhuvRlx0Xgvs2J0GcICI7Ly+ZtwNewHjGQDXbGBz/8/gfr7DOTcBNgfZPg5pqy4Nm83PZwD8/Jr3ue6c+31c2hp81Y3Bc+FE1AA/+HkAIKJ/AbaFbQU/QUTTRHQIwA8D+MMNzmkBSOU+CRH9DFgCvRR8DsCKOGvUiCgmohcR0cu32M5nG5v15X0AXkscVzkJVn9YnAXbFRR/AeAmIvp2cUx5C1i1+mfbaNel1PWdRHQbEdUB/O8A3usuHrry6wB+mohuBwAimiSib9tG+65W/CARHRSHqv+Ajef1WnwGPM9/iIhKRPRmAK94Nhv5AsPnwIv8LxBRQxy3vgq8fqwCWCKiAwB+Ys11a9+fgI2x2fz8TQA/QESvJEaDiL6BiFp4/qzBI3jWP6DOuYfB9rPPgCfhHWCHia3g/WAD830A/hzAb21wzl8B+Euwo8sxsMPQJYn8spB/I4CXAHgaLBn9d7DzzFWDzfrSOffX4AX4fnBfrf0Q/gqAbxWPz191zs2Bn/nfApgDS+/f6Jy7mFS4Ubsupa53A3gXxLELwEW9nJ1zfwLgvwD4A/G8exDA12+1fVcx3gPgQ2BniSfBdqJN4ZwbAHgz2P4/D3Yk++Nnr4kvLMi7/k1gp6HjAE6A+/BnAbwUrLX6c6zv03cAeJt4j/74c9fi5xc2m5/Oub8H8P0A3glWjz8h5z1v1uC1oFFV9dUHInIAbnTOPXGl2xIQEBAQEKC4aokUAgICAgICrmaED2hAQEBAQMA2cNWrcAMCAgICAq5GBAk0ICAgICBgG9gSkUKSlFy5Utn0nEsJuhw5c6MLVCqmS6/tsrCBFL6pXL7m/LXnDgd9pGl62Y3fsXOnO3jNEUTmDmvvRZcYqeak0OXuYqdiZFDkWVVT4eyVl6S8MHURXxBHfCw3/ZjnF66MinNyAMDSko9vz12O5cUldNqdy+7vickJt2v3LqRDH2OfO75nRee9bfMm84Ck1XEcS9tNxI4bHYPIDGJRpfTVaB/zfywVtJP25RIRFHFWuKIVfA5fl2WGO0DqiKJY6jRF8reOid7D1nns6WOzzrlduEyUyxVXrzeKfub7y3Nqn5l+TkplOSeS1qyfN7TRuqGHNntHNloH1sz/0WPS97lvu44lRdEF61zfqPXtG3kGAtqdLnr9wWXP8VIpcZVqCRTZd3zNr4E+j46PbVcc8ecjTXmcbJ26TmSZlJnbRTJ2mbxneerfDX2naIMPQ5wkI798Pl/rdAxsG6TNabY+Wq44y42+I9rWNHPIc3fR/t7SB7RSqeKW21+y5mWLipuuaVrR2RvO53XXbQQujKILn2Tn50Z1abl2kD3Hl+Xr6tLZ5Ms2eoHWv1y5y/Howw9csL1bweEjR/Cxz/89yEwwEqVBb9gHAMTmgcoJLy4DmZh2smeZvuxcV5an68r0dGfGcCi3HkqdmZmMunDoIgwAqSzSTuo0Rdi1g4mLpurMLd0f+DacW1wGACy1+/Kg/sK+POsXP/dRAMDHP/rBoqzd6eFTH9lqVNTG2LNvL371N96JXrdbHKOYxzaR51np9IqyeoN5KxaXVwAAccVzutflg5vEJX6GQbsoy/r8PPpiR4l/1mw4BABU5M10A3O/Kh/s5v58Heuo+PCa+SCnUak+ci7fk9tVqepYDIqyfs51NLQRpg2zK/z3D3zH942F/ajeaODVr/869Aad4lijzHO8FXGfL5w/X5SVy9zeiRbPpXTQL8qG+nfxYdvgi0DrF8xENjlRnMj/E3O6bjT8pkX/1nep3V1dd5tEFvmhjCfgPw6ZvM/ZBgu7brjK5bI/lpTwoU/+3fpn2QaqtRJe9orrRzZ/urFKiOdEvdIoyiYnWwCAlZUlAKNrZLMyDQBYbfP70sv8PGn3uE9W+/w7yP38mqpzeH7c53Fanl0uypY7fB4Z+WMi5ndJ1xtX8uPTmOC6OgNuQ1LzZfr33CITqA3NPI7l0xfl8jEf+LGIYsLZeX/uZggq3ICAgICAgG0gfEADAgICAgK2gS2pcIkI5XJ5jb5Y1bTRyP+BzVWva+0U9v8b2jAuE5vZIi7FzmHtQGvrsjYQ59zY2p+lQyyeP4M9u3YXx/opqzg+/jd/CgDo9bxq8OCBIwCATpdVijdce2tRtm/PtQCAodgMylGpKKNIVUv8HFlm9K6iZonLfE7XqqQyfs6l1ZXiWKXKqqeqbM327PD0vK0Wl8ViWijVvJqqXNoBAIhyVtd99gufK8q+/Nh9AICjjz0KABiseFWkSwmwJrrLgMtz9PtdpEa13Btyf6uW1cW+b/KU+yISm13ZzPcScT/nfb7e7lSTWo3PF9VZp2dUS6IeLJdZbdXL/ZWLXT6vVvd9Gom6VW3bvb5XlZWnWP2mqsLOwI9dq8JtyGU8h+Y6SF196YZK1fs9xNGF36NtgQgoJ5isecIZElVsFPOzNSemirL5WVbHkXDB93te3T7sc/9QYSs2Y7XW7JJbFabY0YpD/jp9l62dWtWzZemXzEzA1VVWWaap2PeMmlbtgqn0uV1TVHVb1G1UuFGUYJj6sbs8OABpYWIBgEqZ71mVORcbY+hAVKPaD3at62t/y/xvt/06MBSVrcvVpOP7Qed4IjaGSrNWlE03WCWbdf355ZTrX1xknvmV5aWiTKutT7JKf5CvN3c1W/we5Jmfx8OuTG75SQw1crlcQrRo3odNECTQgICAgICAbWDLEmipVNlQAlVpc3NJcjMHo/VSm5ZdqkR3aSGtG0mZo/8fLUs3uW4DJ6J8fBL0/Nx5/NF7fg1f//VvLo49ffxJAMDHPvL7AIBh3ztflEu8e1PHhcOHjxRl3/CN38nHrr0dAHDylKcJbrd5R5fEZanH7wjTjJ/tzBlOFHLmvL9utcPesCdO+mN33MHcz69/1esBAFNN75DgPXnl/2Y3WxJpY+8M7xZnTF6Z2eOPc3+cY+mUyLfvda+9Bw988UsYBxyANHeIS373n3dZwqcS717V6QYAcpFOa+LUkA38DjxOtI3qxeilGhVUe+JMlBvv2Jo4JsXifGRuh57s+Pt9L43Um7K7FseUOLGaBZEoZJtcLhvpYZV386VGU8r8MycqwZVK6+psNcerHYrjGFOTU6jk/plWO9zn8/OS2cqUJTJP9JmMQgBUiqVM1yTf7sKJbgPP2Uw0KaqBsWX698BIUP2+SFzq+G00D32R5AupdiOvdfWYjo22TqRlUs/RoXeOSrPeSJsuCw5wWTqy2palE9U3Z2XZe7nHsib0+6LFaHuJv5rwu63Sdnfo16JEHNASdXpc9WOYlfj8idYO+b/5Luj9cq+VGUq9JNqtiuk3kntrv1WqfsxTcUJTbU5kNGvL4uSkHRGZidRuDzeNCrAIEmhAQEBAQMA2sA0JtLwmdoR/NLbHxgJtZufUCwsp0+yJ/K5tvVS7efvWNEpqGz22XpKEW1tiY7x0Z7J+B1i4VY9IrlT0xeViMBji+LFT+JM/eXdxbO485wSfP8e782HfxgTOAwDihJ91cf7BoqzZ5NCPO5f5uo/97ceKsoV5luxqslMrlY0bf8yS1LmzvCtdXfYhBaurfGxq0tuvvvar7gEATE/yddmIfVj/Wu++78OMuL/vfuXrirKpFtuA//QD7wMA7D54sCj7ljf9M7zn3b+7rr7tgMC2rv7Q2z8aNXGhl/Hvdr1kUFYxSOLhYHatA9nt1posUWZ93w+dVXbb74rNqNZo+TpFDJif5X7OTXhJRW2fRkJOREpMc91tG9uZzGG9tdq4AGAgbR2IhmEk4q2k8XYSw5oau+qYt9wREapxCXHPh4L0FziJz9Ic/0ZmlZqaYok5kTnuMjP/xY5VhGONhLjFckzjM40EKja5LNJQLSuBSjut7XdNnLoNkxWTNMplbYvtMPE1yNXmamyg0eh6aBtPLgONiTEudzl63S6mprxduSFaD7UnLyzOF2WtJp/XbvMcWFk2NueyNlU1Fv5Z1ZZbr3LdycDPSxrwWPRWWCpNYDQ+ooypGq1HLO1LZOxTY+fsn+O25uD2tSS8CQDKE3xdKiFSg46XkPV70NFjxtdgdbm/YYjRRggSaEBAQEBAwDYQPqABAQEBAQHbwBZVuBHKpbVhLPyrbDSWsUNVrxsx860NfxnVuq5V4V5a+7boQzRq4B+5r1dL5E6ZSNy607z7uy2LxheGExGoWsbjR58sDnWXWNWVpdyuTtu4e4vaKBZj+6Dn1Y1f/vLfAwCOnuCQkBPHvCv4cCDPClapWFaOYSoOMurnQr5sZjczuX3Ht721OPb6175B/hL134hqfv1fxX2U6UgcLDJzys03czjOd019N59rtn2VenXTcKmtwDkgT/OR5lVEnZkKi0luHFPSvrImrZnLAHri1JN2uE8HQ8N0In0yLe715da0v2EmYSVy/vkzZ4uinTvZ6eLADft9m/M1zjEmdGNW1MCJqJprde/QBXEUWxXWJTKUUbkw0pRElW0ZzbIRqsAxwDmgP0RiwjTyjjhjCXtTvenVjZWyODbpsmFYaZDIQRm/PF9vMvKMXLSmxIc92FCdLF/PhubWWHOsKjYvnJuikV8+b1Rtnrn1piZt37p1dExrCiFiJzzy6nwn72osjnK1qleDKrlTNpQ5EZv5n+uaz22rmLJ4qCp2Cc8p+3kTy1xL5ZlKJkxqp5iDFhe8I1MCnqtDCZdJnQlViZXNS3+9Q15dHIMymeuq5rXPoePbXvHq3WgDh9YLIUigAQEBAQEB28DWJNAIqJRjNOom6FS/8iKNDVO7e1MpRJ0tbGWya4eSQlvka46t3xG4S94lyE5oQ8bkNVtJ819Pus7tm2j5xquvyNKKBAubC10ejRI1Xwb6/R6efOJx9Ja9JKnhHs2WhFWYnd1QCAD6CxIeEfkYiNl5lkprHdlllr3jT0mcL4ZSFxnHFSqkUdn9YbEou+X6m/n3xpuLY8eOHud7yy6xZ7hKXa7SFUtz1lC/2pYAdAnBWVn2/JglCfq/8ZY7uB4j8a90VkYclS4HURyj3poELPH7QCR+uWXTkD/k4txQlvFvL84WZSrF9yREYrLlr0OPy+YkMLwS+R14Sd6LskhTEzM7i7JYJMiEDCm8OPrE4uy0tOyJNVoS4tKoS8C/kTJ7wpKgoTSVigl/EUmkL7y/lbKXXJN4vBLocDjE7LlT2Em+3TXHbVoQblUyYQsl4cLN1dvESMRK4K4aoNgsbwUxvkohRoqJiuVJyQL8/WJ5NzKzeGU0yrVriRTKGkqjfLpmKVCnw0Q1ciMc2voMNPLLbR1f6BBRhHK1XnAhA8DCEr9rKt03jMS/KtzHsY6B9SKLRnl7K4bkXYX4vsyvoVEplWWueinbP99AJPDE8Er3Vf2lkr6R+Hce3AMASAf8DIvz3skxk3VMx6LbN3OswfXvEK1RahMtDHNQdmlrSpBAAwICAgICtoEtSaBxBDTrDq9+hQ8jmGnwrmVulnffZ+e9m/PCMle/Iof6ZmefRupWLnY7swuJxRZHhSTqm1nIkWpfNXsAWmM7teW5Zlcxu3eCBuHSyC9foMf4uluv8xLESpvb91g/kTq9xJZn0RrX9e2jWm3gllteinPnvFv5+RNMaNDVgGbT5mqdbWQ37LmOz039Lm41lp277OxKZb8DLUsoBMnOEIauLxoKTV1FSRb8zqxe5+f+2Ef/vDi2sDDH9+uwjbVvMiCUNOuF/D83UkCaaiYNkib4++y65jA/14teCgCoxZbmjDZMfbQtuBx51huRQDIZ2mVpz4Tpm6rYbpzYLes7ZoqymuzGSSTEhvM2llUVNySTRFwxGVTEjr13ku2c/b7NkiLB4qnJeiH9fP8DTwMAXvXyO4qyqR1c//GzfM7Sgg8VKYt0t2eGz0nNjpv6SsrAkr9JFoMsv7QsFZeKQb+Hp554BNGUv8lUWcND1H5sQnlk/iqNnBUBvN2R+9AqJkj/o5R05jqn0ocKOCY0phANrUAia4JKrJmRFiOlhCvWJ1OVtgV6/XpCGqdSmU1Zd4khFZcCB4fcuZFsLJ0ez82FRX4Hd+7YU5RNTvOa0hOtSSnxawrFo58PNzRhXBLuRfF62+nSCkuCZQnHajS8huPceV4/rA1Y6UdrTbHNmrmaCZFCTbQs58/778/8Aj+XZtlJzYTYvWsfAGDvAX7WxEjWq4tdrJ7xmrbNECTQgICAgICAbSB8QAMCAgICAraBLalwoyhCrV7HzTcdLo5ds4erWJhjFcz5RS9Cz4nm8dQZVrecm/NqrBUJnRhmqtYzThYQNQGtMR4DgLgyq6rDWffuglnIqHVVvSd15abMOc8Aw+daZyD+u1Fm9cHUhFfdafaMcoUdcYi8eiLP3djCKmZ27cG3/+CPYXHJqxOefPARAMBH3sfsRItzJ4uyO26/BwBww46bAADHH3y8KEtFXTQU9d/Q8ItG4uxVFsaZyBnnC1V9l9S5wfPQ9svs4LI88Ib7bpvbN3uOVTHdge9T5cWtiMoY1h9FxlNZnEqJcdpS1hwJDWnWjAu+yzbyMdsW8jxHb2WlcG0HgFzm2/Q0h5q0lz3f7fIys0KpanpgMjpE0r8zU9xfHbNXTYSzuFXhY6tdrxZNNROKqIWXjTPV08dO8+9jTxXHVsU+UqpynTff4NW0Tz7B4U+PPMZcxYsLvq4jhw8AAG5/6V0AvGoUAJw45vRFJWfns02ePh44JNGow4Zyl1bFbJANrYmE58D1190IAHj1PV9dlGlYiGrq+n2/Fuka8sxR7rsv/r1PUN0XvmMNubBMN5mMg+VGjYqwJTEx2ZAYmQu05lyuQ5jL1NlpJJpL5n/hfGST1I9PbR5FESqN2mgoiDJOqcOl4W3evWcvAGBlWZz8rPpUzl9ZlXV9YNdieY/LwkhmuH01Kbo6Hw1Tb1brSzibdaJS1Wsk5hMbjkXCajYhTpUzO7zT3dx5NiuqE1op8d+YzI2GC83s9NdVKn2cm/MOR5shSKABAQEBAQHbwJYkUCACxXU8fcIHue7bx8bYHft5h17yMbhoTfLOcec07xhmz3vJ4dQs7w7nlzXcwX/LUwmc9Rs7Ewqh2zaNmd6A4MA61sSy04hJJC7jPZDl9ZHTM+d3QkORxso1vs7uAnUzWpHwiggmQwcGY+PCjZIEEzMzqJlA+z2H2UGoNMH3ft9v/FJRtmOKz1sVLst215MlVIWTtSQ7OxfbXT0fUy7MxLi4JyV+NiqcKUyuTOm3btPv3ur9awEAM6vcX93EON3UpC7ZpadW4i+6TAPefftUAorXhCkw4rE5ERFFiCsNxGZ3rhlQFue4L50JqSiL1LYizjknTx0vyqoNkeJS3sHvmPYORv1VrqucqLTi6zwjGWc+8ykmvHj8scdNGUv183PeqezuV7wSAPD1d78CAPChv/5MUXb6JEust7/kTgDAG1/7mqKsKTyhQnGKrpHyNPa/JO1rmiwu+SW6918q4ojQqJVGiEpiccCqiITSNePt5GWdmGLnlltfdJcvE+cyzYnqcv9MhUJDqnrowfuLMs3rmqlToc0VqhqYkTyTwv26QcjJuuxUxhlGuYldqhzavh9I1DHK2WvdhrIx8eACAMURKvUa2m2vjVAJtCkkBho2BgArbZYuVWKtGO2PciZrztzVjpf4s76EY8n6ERn+Zicah6GqCszjafiKnWdOeqPdlfGMfRsmEpnHElIYkV+LNfxK712u+eeqVrhsSTQ4zmgmq/X6yLhthiCBBgQEBAQEbANbJFKIUa5P4NEn/A64VObtw523ccaMqWmvn65XecfcLPMOvWVchacneddzdp53Dqfn/H1WerJDcbJrMekYlFbMif0Bscn4rsHIZpOciC1qqsF15ZmXZtUm1BM73dDQrfUlEL4tQf9Ldetyzju1qiRrjC2VF42PSKHX7eKh+x9Er2/DPXiXfNMi//7jaW+P1pyC55e5M+PY2o6FXEHOSU2YSC5B9bHQ9sHQYeUJX5fKjjLum5Aiqb9jgpDThOfBbrD94eDS6aLsS7H0pexcraCuO/yC5suEiwwGSlih+Vdthhe3npJxm8idQ6c/QGQ0EUrBl8rkqhp3/HKFd8K7D/J8mNrlJXEnGos0V6nZ3ycdav38rO1Fr9H5wPs/DAC472GhbzTEBZrz0xrPSpIBIxYihPqkl3S/+au+EgBw+PpD0l4vBcwJVZra820Gnlhs1F3Jyzkw8w824HwcIA6TGRqpr7id9DWZPo/EfrYsJA+fv/eLRVlX5nRvwHMnNrbVWKTLxx5+gOsxmgQNgUBHqDlN0H9UEWKEql/XVpZYgzCU8XBmIpOI7wXRi6U9ValU7aQmXEapFPXWaer7fJjmI+QhlwPNqEVmTZ2cZM3V8mK7uJ+iLzZ/FcicNdym8q7KmjQ0bR7IGt7rcVlSMZq/TMOkuA01IxlCbMBDQ+3YLLGUOZS51+kYbaDcZznitjebXgKtClnMSpel6KbN1CLS6dIir5VV04ZypXTJdKxBAg0ICAgICNgGwgc0ICAgICBgG9haGEscodFoIkq9ePvEE8J12GN16Itu2VuU7dvLqqNamdVFtZI3XJerbLyt1vkbXm961cDZBRbLlWu2PzRhCySu7coraaNfVL1gVLG5hAgkohuJnBf/sw679w+7XFc6MOrJXDKaiHNTf3LC90ON/65I1pPEmUwSFI/NiWjY6+L8ow+iaRIuT4oLd/djHwEAvPi8Dw16eoHVLWcWzwEAZgbeqD85ZHXLtBjpG2Wv6iiJoV8dBKw6Q6MWIklgHJvMCXGiKmwTvrGLnTt2ZbcBANxZr5t34silzC+pCXFJyqOMVAOjplWVUpH5p2RDKWhs2W/yPMeg10FiqHcaLWELUhWbYVZSF3hVB1Jimawk24OqclNfZ60xIS3nOfnQEz4U6eHHeU62RU2VmHAe1bKmZlxL4vA1lLpufdmdRdnO3cxpuqphGqs+BKcnJopY3hkbSaJOTZqAu22y+oyN9cnUR3GE3Dht9MVMUxVO1lvufHFRdu0NnJknErWuhvsAXvW8c5Ln4OK85yZ++ilmanrkYU4yf+2LlXztAAAgAElEQVT+3UVZpOErwuqUmjVi105WiVckMTQAPCD19oV7GibjioZ96dyIN+QO1gxP/oiqgb061KvK270BsnxMKlwQEkow6Pg1cqLG/aXhYQvz3vkwiXktqEnZ8pLvG+Uxby9LpiLvs2Wy3gjXctWr4RstcWgUJ7HIqOi1/wy5WxFCVJf5mDhfuCjMRbG+e6lfu7odPra6zHOkaUxTSv5VkXWwOWFCGmM3wr+8GYIEGhAQEBAQsA1skQs3RqvZgDO76RhsqD03xzua+x7yDhFLqywt3XQt8w1O7DRSXMI7hxx8vpUkmw2u68w53mksrPhmpiKNaNYXMvkp+z3eYbfnfRsG4q69mnFZyYRV6M680eCdbjXxuzwNX4gSyWSR+WeOpNuUP7EKs4OKkhEex8tBvVrFnbfdglLFS33qUPHwLdcDAD557LGirPbYwwCAV+49AgB47aS/rhpxB9eFT3Jir3d4qUpWkVy8r6IJv9tOJPQkEu5c64gSiTNQZskp1HfiWm7Dyp03FmVxjx2Kjh7j0Iwnjh8tyk7M8nxQB5vZOU8eccjxPGqv8Bh2V73UTTkwHPgxvRzEUYR6oz7CZaz8qjV5/pWel+ISDXqX0J6SccxwyucqEp6VRDRcqCp5CufO+Wedn+N+UA7nyWmfNafb4fNaLf8e3Xjb7Xxv0YbUTaak2Vl29lNnp7Trg8OroslQAgUyjlg+c4mEk8S2f8csgUaEcrWKxGgRNLQgrvM8vOVFXgLdvZ+1WkviyGNFgOGQrzswzeuNM6Fnc3McHtQRKbxZv6YoKwsBxvwpGReTN7IposrQOB+qg0vBrWqz9+hzFc9n51Iy8ms9y5SgQCXPtnGUWV5pFyQMl4uICJWkghuO3FQcK+aJ8GVPmcxTZ05zWV5wYvs2a/antK9OmMZpS96FHTJ/SyYUao+QM/TFOfDsWZ/zdmqHZIIxGqjzZ2elTn4Hncn4NZQ6+j0+1mv7cdL+jcQZdWHWa0DbopmcnOH2NWt+zUsxGCHA2AxBAg0ICAgICNgGtiaBxhGmWg1kNqjYiZQju/aO2and/wjvpgcSqH/jdT7PnIafnD71BADg2mv3F0XViHeCkdym2fDNPHOad5B5n3eBS0te2mwvMrVae8GH2bTqvKvaO8M77tWe313On+Ub5BLiUav6nfb0TE1+xX3d7AB7QgtYERf3msm76eJ4bBIoCAAReiYbw0P3fh4A0DzB0ty+spcIrxVJ45qq2C32+Kw5NTlGDf7VDCwAimcbrLKEEp31fRqraqA0mkUe8DR3ztgMezXur6pQBrZMvtI7qzz+N177agDAV+z2ksXH7/00AOD/+/THAQAnjRahJkQN7/u93wMALBl6u85qG+fOnsF44EDICvsNADihJ0skBCEyGShUKlATWKXkyzQKR22Gdneu9bem2e5y+NChouxlt9wCALjnDa/ne5ip9Gd//mcAgFkzv/siOe7cw7a6TtuXaU7RmsyR2OT8HIgUpVJ9a8KHaWgOxzRdL3WUSyaQfgwgEMpUQs2YCp1QPjZ3iz+FkeyXVlj7oJRvqbEVKu3egw8wScLZc55icm6RJda65EhtdzzlYbXON58UzUunY0JPRBodGLrFelmztwgBgwlHcRLKoTmQM0OLF8v8qMk7lRspqyfPsdzlOheXvJ2728sxJgEUee7Q6/Qx6HstzpLYPDVUJjHZjnodyYgj7WqavLY9ITYoSCCMJkrneCqS68CEpen50y2W/hbOnSvKmjJXp0yGlqVzvBYsLvFczYwPjs0AAwAwIYTqy6D+G52O18D0hVqwKdq3VWP3zeLhJROGBAk0ICAgICBgGwgf0ICAgICAgG1g605ErRZyw00LCWHIRVXhIi/it0UqfvoZFr2HhuGjlPKx+x88BgC49dZbizI1WPdEHXXyhHfzn58T5hRhp1iZ8+q7SonbdfiwF/9nJPxkpyQ7PjfvVbj3H2fX9vOLXGfDEGJM72TX7kOHWY2U9b0KQj25M1ERlMnfb5xORFmWY3V1FV3jJHP+mRMAgFjYhjrLXi1xbonVMnXZF+0y6qNInKIqopYb2LAFUckXfKTGgSHSLAqifnI2G4NkaMmMG/q5aVbTTq+wisx68UcSApOII1Kt4plB7ipzP58+zPPgT5d9toyehG2cOsXzoLNqnnluDv2BmY+XCxcVzEcAUBJTw4I4o5WNGhSSQagrvL1lwyFcqKwKeiw/91NhOmrJ3PzO7/n2ouzu19wNADh8A3MK3/elB4uyLz3Ef2eGq/fQQXaYqchYHJ/3YUNNmdAaGtAwWSw0SXUqbV9a9nVq0uR6i9WdeebViT2T4HssyB1cb4jGlG9bcyer9maOsKNcvekdqdSpT7MlZalvz4SoBOfm2Onk5IlnirJYVHvTO5h1x2Z40TCWCXGwc8ZRqqoOWH3/TkxWJdOI0nIbbtZcTFltcbBZbvt3tz/UjEOJPIu/T1vqX5D3pm8cZRCXCyaxy0W5VMbBfYfwuc9/oTi2uszrxsQE99/AspQNRf0p7RvarCry3sdi1rGhN0pYNdcXU4xx0Hz0EXYivO0mHt/pul8HhmKe2bn/QHHsthvZ4enhx44CAM6e8+adgpBOFprSiIlBmcu4Mc5kzRn2eFxW5/k7VDGmmagajyZQ3wRBAg0ICAgICNgGtuxE1GpNFMZ6wKTg1K+9ySNQIt7Bqu12dt5wjEoowonT/Hv/gyeKslQkiqNH2Qng+HGT5UKcYSYmuK5D+7xL/87dvJMpl/zusix8i4mw9pf6/vx6k3frJXGosfH5J8+IVFFmA/fBfX6XtEtyPGYiaQ5hdj1URRSPZ19CRIhLJXSWDPew9Fsi/JUTB68tyrpzvDNLxcieLPvrlpdZHaA724HhLE5jPpaJlFSZ8c/T6LOUtUPUCd2D+4qy++S6kyveAL8j43tWxdngUeOsMRnxmN12hiXJlf17irL7JXQiP8D17z/ly3LRctRlh1wyEu/52XmMiQoXzgFp5jDV8i7tGiKTC09warQB6k4/lJdA3foB72Ck2W+WF/1YqHDxwP3M4/qUieE6cYrn/G+967cBAJ/9rJfED9/AvMev+sqXFMemxMntzGl+f+om60W/I44i4vSSmFyeqklIJDjdGak2Fn5dlaZSm+9yXA5yeq84xsRECxPT3sFwucvvfybEINlT/v0vtDvSXks7PTXJ640K+w3D75uLE0wiYW+thnH8g77r/F4nRlqCE47hil83GkIMk4kTTavl35dKnetYFmfFJPHS++mzHIY0L+FIVj2zJPOqLZKRc3ZpprHxPcdxgh1TMzi0zzuunQbPuYo8f6dt5zi3MSF+xpJxImzKWqy+UP22f9fVoYhkbVXOZQAYCmHF/CRrCnZO+vetL2PvjFYpkb4g8airlPzYkaxBaaokLTarjzg+ijhspVNN9rQo0myz4tvQKu/EpcqWQQINCAgICAjYBrZG5RdFaNRrGA79ZZ6ySfTMVnUv9E+p6M0txR1FvFvsD/nL/+nPPFKUOZFG+j3eQV9/vZeybrzxGrlOMry0/G6kIa7PnbbXkXfVnke8W2pN+J3uZIvPa02ydFkzu9L5Bd6VnTrL4nO3692+9+/j9k3s4F1pqeafq1SpFnkrLxdEQJxEaDSMjWVRdrELYguGyVw/w3bbL4j7/pGul8Tnct4Jn7vjCAAgetFtRdmx8xzInBd1+V1mR4Ksb3qcx+eZFW9j+6DQwdmd5y7Z9ZfFZvqkCTNqVHnMvy4S4oaalyTnpMuqYotoTPkcqLmEYzTqLIGuGgPuYDAYW6YKpqpsIDdhWrl0SUVCHezuV+1vw1XNWGFUGGIPVfs1mdCA4TL315fu5cwgtYaVePk9On6U7UTdVd9/pdIRAEDZ9NuXJWRjz16W3Os1/24mMo5KkmDtV7lIlRVpp0pfgMkyInYvS983LppKxWA4wImTJ3D29Kni2Gkh0RhWmCRk1wEfjuVSzfXJbRwaasW9u5me79ojvEYkxsY4I5R8mhBqounfqd3yHk9LjtHHHn+6KNsxxWVnTnopuLnK4zfo8zvVrPo+qdWUpo9/yzZBsqyRJ86yBNYxlKNdSUHjhCwgs1J/lo9Ry5IjTTvYscNroDRv6uoqP0+34zVKZTH0ZiLZ1Qwl31Ao9WKx6drMWJoP1UksYmokylgG4ayQW5SjHUXZQSE2mJ7w/fbQ06yxWlqdlzoNcY1oelLp24HxVYlFeo5iyfNqCDJi0TKQML8smBy73W46omnaDEECDQgICAgI2AbCBzQgICAgIGAb2FpCbYrYrduqzCQTiSv4M40RVwzOJTHmZibhapywymjnniMAgLnTTxVlZXGEUGeYAwc8S9HLX3EXAGCYs5qn0/Wi97DL9TerXk3brrLqYJCx+jCpepf4SXHTb0ywmubANd51erc4s3RE3bay4NkyTp1nVcdKT4zguz2T/57mzNiygzjnMOwPsDDn1XgrohrJhAu0SX4IT0uC6ycW2KHk7tQ7/PR3cRvdXlaNWP7Jxx9gVWJdsm5Y9YU6YZ3pspqyd86HDaWSsaBtQglqEsa0S9hjyqkPOZnq89woJXys9rDngN0tib7rU6xqaza9CufBnF3bNftDZhwFer0+8jGpcAGCowTOOIHVxLOsLc5bndyroiKZ67VGc6R9AIrQmrLMfVfxqt9On9Xg6jhDiVELi9PJPjFb1Ke9KnvHXlZROsNMVa5yGzptHp9Bf/2euAlWEY+w9ih7jKjay4btR8OZNLF4z7DwTEx4Z5pxwDmHQTbA8rK/x4KwBkUyBc6f8WFsqZh11ETQWfXmg4E4s3WX+X3Zf+RIUXbNYXaamRZHoxtvuKEo27+P35OesPMcSLxKfUbCXmo7/Dgsi4loeZnfocQmmRYmtkqizFV+PGamuN5FcWjqrhgOXclUVFJHGbOOOsrRp/HM8VqjhFtffgA3DLyTnvJ+d4RHdmnev7NDYXxqt/lYtebn6pKofGNpc7NsE2Pz8/cLrloTGifhbLGoT9NFf79E+GoHxnRxy528/l9z1xEAwOyCP39xdmGkfSXDBjYl706jxccGA89jDUn0jS6/B0e/7EOe5k6fQZ4FFW5AQEBAQMCzhi06ERGqlcqIk4X6MDvZhTgboC+CWFUC59upd8TpSnD81AzvEubPm3QsEqaQxLxjO3PWG7XPnecdyr4DvJO0cbM9kWz6kZGCZQfYlXyYWcdLEIcOsSG9ISEhe/f5XKZRmcsyIXVYWfS8mgvzLIX1u/wMNoC+0aiPjUih31vFU49+CnMrPq9h1OC6T0pWkD1Gelnt8bgsiyPPMcNLvG9ZyAvEWeuw2dFOiIRbFa7JkSBidaKRXeqKCSGJK3yfZRP/05LMNDvlwvPG4WWXuLbvF0mobgg+i+wVQ24DZV5aOlZjKSOTZy6Z5JVTzXIRJH+5cM4hzdIRLtxUwrJicR5yps2a1zSWPkmNZKxnKWHB4oJ3vloWMoJu4YTkd+enzzDH8TNPM8FIbMYpE6eNxVnfpzeLJHXdbslAYjQSJE5NrRqPfc8EyKsTDkkGkxL8+1eWcW1LGIwzbVhum138GBDFERrNBirmRa5KYP1QFhDLvZ2X9N3iY9ahqiF/dpb4fTl7ys/LqvBW72hxdqD9B302loZoPTrnue8ndvlcoU1xZtNgfMCTaVSkLTWjXdB1LZI1z64NEuECSZmL/tBzOlOHx7YqkmvT9HkCwpf645njcaWEqZv2jnC9FhozfUZnpS8NE9EyQzIha2pN2hZ1vCYqFy2J/toVcSD302/GkuHe/vuPMNf3ipEWX/mSlwIA9lx/hK+3mpSBcCOLxicxhAiaOUozWOVWcyMtOvUkaxbfffIPi7KJXhVx99LmeZBAAwICAgICtoHwAQ0ICAgICNgGtqzCrdUqI3paVZ+p1tL4CRVQw25m1HInTh6TSlldUDHxRauSYsnJfTo9r156+hg7npyRxMO33+FjRBuTDbmfVwl0xLic9biueuTVZTffxCpiVc9NTBo1VlWcBsS5JdvtnQg6XXY26ktsaNWocOrNSSSWAPYyMBh0ceL4gyhVTFLeiJ04JoSrd2gceGiZ1RjT4nR1ask/60Hh2JyR2KuKcW7YpypQcQAio+pwwvCRS/LnKTO+uyWusHrzzcWx5m2cosxVhSd4wjtYrZ7iMX/sLz4IADhn+Eh7MoF6A25nbjiVrxfOzK7Eg+42jjV7d70cjz5+DOOAcw79Xh9p37PHZOKgtEMS/VZK3lGi01F+UJ7XI1ygos7tiLPX4tGHi7KkxU4rjQY/BzVM/JyQe9blnZnZvcuXSVuWF71JoyR8wkmV536t6uei8oPGUme16tVvJXHGiSQlV88wRsXCyqIp2IYm7rZSNaaWMSCSJOYor1fnq0NiZtTmmuxb1XHOqHdV3a1xwa7vVaQrZ3mOuIOcGq/kTNJsWXsSUYP3u96BbUW4Wc+fPF0c68uxCeHJVQ5eAIh6wg8r/ZSRX9cycRSaSSUNnlnXuuJsMy3vwQ7yz1wD4Qk/PJeF3OXoDXsjZhpllyJR3ZLpG3UYVbOUlbhmxHzUk7jzEw8+WpSlbX43auJ0VDYMWXVJCN8Vtqb50948ttLjvr3jtS8rjl1zI5vWBhF/F+oVv74mE9q/lqNanlXmzVDmjHXyK8u7seswt+Xwbd58l7f34OmPBBVuQEBAQEDAs4atSaAUoV6rjfBPlsVhpSK/abaeJaL4v5FA+31xRRYpZOcOz0YxJ8w4A2HqqDZnirIzsywdDM/ylmzfoZuKsokJ3u3U6t4hoZXwTiIui3NK17tAl2LN6MB10cCHaJTlMSp1ljzy2EiZDWmP49/IupxjfNlY4riKqclbR6SKZRKOzQo/T80MBtXFMWQX76RPdr1r9oOz7FgxJW2tGKkiE2cI0uS0Znwz2b11VBKP/I5fpd8jd95ZHNv3z7+Fz1Mu0KqX2M7eey8A4FMf/CsAwGruJeSB7CA7wnF6uOrH/LppDkEoibOSdaeo0GThxHO54DCtGshIWSRzWLk9R6Qh7a5CZPJ9U5UJlIo0NzXjd7ilCZaCehqyUPbzZY9IuovTLNXs2uPDDZQhqGrmQ1kyrNRqwuNq6kolI0yiDlBGQqYqS1mxJId3Zi+dkTjAiINOFJswpZoJVRgDiAilpAQyEpcy9iiDktVq5dLZFKtWy5fpq0AiiZZNlu5r93C/TooGauXs40WZhqpUhWVp9px/b44+yuedfspLV5A1pNXivi+bLDct6Z+SjBEZjQUWeJ3pdUTDUfd9PimhFlPCBTttpMDYZUguNT3IRUC5Q6ndL8IOAYCUNUg0DpbdTB2MlIGqWvZra3uW++HpezkEsdozEl5Wk/Olj4xGqTbgY8Me32furF+T4yb32zW3Hy6OJXVxJpOJENvE3ak4CMl7GZkQQhLGIp0PLjLfHydOR00+/55veI1vQxrjk18w470JggQaEBAQEBCwDWyLCzcx9jN1b05T3mEMU7/TyFLlx9WULTYEgP9eEt7Law3f5b7dbMNrTIptwbiCL7V5Z7Z/P0slqx1/v4UV3snsmPSPtVPsE60JPi82gbbdiPXtfQltGQ68naOr2TNSlhyq097tPS6ztJyL1BSNEEvEYyNSiChGvTYauN4RG8EJCYu40eT8VDOSutV3TZ692SEHo3d17EyTz4mdsy1FudmdprLz7Il01bGhGnK/bzZS8F3i4t+VsY8NH2lXJN3HpH/aRpbMEuVt5fHZnRiOX3nGPCO5r92NO4yLR8HlGXqdZSRlP9+qlZq0i2FJBbrCNRxLLs56w0siakduimYlis2zyr51SuxDse2HXKVMth1PTvo6lfc1qnipptHk+VErcx+ttn2oWCy7/nKZfxMzLwe5PKNoS2I7ISKVtrktiXlncreBk8NlgBChnJRBsV0bRjOukAlbchJGoVJFnht/gyIzlNh8N5Dsy+A+PHPM26TLdR6jmqwVJ566vyg78eWHAADDVW9PVT7YivR9dcLbQCsytmWxSScVP36rXV5fSuoz0TI8uWIXLauWbuifa5CUkZMf18sBAYiR22iUgrdWx70ReQ1HIkQqOhfKRttz8hSHgMyfYRvmNbu8tmR6J0v1UXl9HuGuaB1nV7hPs7pfw1/z2q8GAEyYLFt94fHOZV7EJlMNyfq0kd+JEqyo1sjm0XUQ3w7xq5g+6P0q4iwb0eRshiCBBgQEBAQEbAPhAxoQEBAQELANbJ2JqFopjM4AkImqKpOEw3nqRXxV7+o5JcMSsVNSB507zrytx475UITDh9mAvHs/u/vf/+XHirLbbr8dALB3P6t5O13rj82qrYVF4xLvVL3G6pJq3fNcFkm2S+yQ0+94NU2/yyrSnnBtprkPbWjMMJtJuc5tcCahduQcxqTBRZ7nWF1to9fz9z55gtMqPXSS+622w7OmzJQkTZjSPJquyYTrtCOajiXTxgdFtdoV9cywbxKS10Z5VDVxLQDk4lB0vuOdAJbFJX1OQi06fc+G9NijTwAA2uIM1HF+rrTk3lPCQdkyjjyZptDKRx0aAGBsnQ02NWTZAFFm0rkJh6dqiKwqSlWGyvmcGDYdTdI8EHf+2KRBqwhDSkkcTqzySV3um8LZmqZetVpq8rHYpk0Ts4gysax0/fgUjkyqTrcJpuU0p85kZBNqC4e1jMkg9c/cT8fX34Ck7IvjYmwBHzKhfKRkmMXWm0f8/4vwlY1UvyIrqAp+1aRPS8s8L3fu4fC0lSW/DvR6PI8rZWO6SIXLtSKpt/b45NQdYSdbWODr3Iqf/zpWVQl/cYYpLFJ1rvBF9/t+rEo7d4NO+xRrlwNH7DSYGe8rVYcjkX6L/XqjzD6Fj1fuHXGUEazb5md9KjNr5IBDCUuS0nLQ96aP8pDH7PQym+8O3XFrUbbzCIdtLff9mpIJQxJl4qhqmJJSsGq7+CYZe04qTkcDmbO5ccRSVi8NU1NWJO6GfISNbDMECTQgICAgIGAb2HI2lqRcQm55GnORPGV34KwTUaZSC/9WDOHAzTdy+MkXPvsZAMDAbCwpZoeixQXexWigOgBcc4ideXS3GRvjcVUM98PUZHZY4b9Tcc23YRVl2R3VhHO3VPESRCJOGb0VSTK74rOXDB2fv/MQS3+UeEeB2GVjk4ryPEen3cfQEA6o0bw6xU4RD6/6oPr94pI9Jbcfmt1iQ6QRkt35rHEayUU6jWuNkf8DQCa7/1j6o1Q3UozsYj/6iY8Xx/7u7z4LAOiK1OxMgluINLdfxq5i9m9NkYAmZI40TVjTMNVAaEYUWakjHVtCbYpiJLUJGE6PInlvX56jXDKu+uLMk4gjgpWMe/JsFQlnqNW85qMju/G+9F/VOBjBqfSlmh3DLxrr3PXPW47V4UfmRdWX6b1J+nZ52YQLiLPXRFNIE4xmoS/vG0m/Dkx2HmfCucYBAiGOEuSWa7asawrPudS8/1jzaiUlP1jKU6zrTm4klRXJNFLfwe1fXPTv87mHOBvRkbZknzF82aoeUEkKACjjOobye3bBS17PHGcNUSbvbNU4YFXkHVTilTTzY1sSDUU+FKcYG/ax5wDoQZ+R5nLhyEtefFMhpRAxc2gST5P0fcNxe+ZP+qxUJ4+yFuy8zKvD13tHS5IMVSvi+FYmP/+Xj3MdJEm9W/t9yFpP7t2x876QBkVCNh5QGmbnHQvtBJFnzNWRyZepA6hKrLGRQDGIivfwYggSaEBAQEBAwDawJQkUxC7tNudnYXdQd/LE0m6JTU0Cum34wd69bN+85dbbAABnz/gQkiefPgrA06G5xO/GNO/b1BTbHxIT4pKJzaRsgn3rDcn5KQG6NqyCIj4vclxWIru7FntiLmz/lr6wxPaKSGxRVuAkorWb5G3DOZberR2sJa7zN93I9HknT5woyp4RCsTdQnWVGQrALw75OXaLXatl7HW3ik2yO9wgB55oD8oqXRmWr76UDYzEP9CxJpVc/fhMys6blAIv9naYWTHKnRUX9WvNzr0hdiW1jVmtAxGNLYwlIkI9LiEb+udxEk4jZkEMBoYsQUTVrgS/x2VLK8e/mrmj3/GagpOn2b5eFcKPaNLnry3LuORiE7QUlz2he8tMG6jB8yGRvmka82gqkhjJ+1OrmTAbkUpT0VKkhhJP61ItjBtachQjDY4BFBFKpRIoMVk0SiqBiu028vNSQxJ03Skltm1cpmtCZsIdlkQTtXcX93W96cMkTh9jaXRZJMmphtUIaKNM1iPxsTgm717fhEfslOxSr3oVB+arfRAAPvOpT/DzyFrSmvJhLBoeVJphaaxU8aExeXMKiLe2VF8QBICoCE8B/CNq+A+NmNi5LwaSa/nseZ8ZqivkD83bWWN44DUvKsqSnbJGChlOvOzH8Gl5/2cOcojgubbPC1x2Mkc3CEtR+zcZTU8s7aM18wLw6wWK89fby4vMYjal69DBRZe2qAQJNCAgICAgYBsIH9CAgICAgIBtYItORIQ4jkdcoAv9pRhznXGHLxh6aD2DTCJq1utu4JCQySmvxnrqCQ5bOXOGVSu79h8oys6d42N1yU6rLCsAUC2OebVORRyDCndsGzQgKtsoaUrbvUoldqziUZ+Z2rQJVSmxmiESZxvL40ljUicW9RGNuO6r+lL5cffs8Ryr2t2qAisZx5VHJHNKX+rab1QkFdHZVCUUwibb1THrk/KS+rKOqPP6JgShK+epW0ViHJKWpV3nJImzqoUAnxFivyQwtorCnjgf6Vjb/hgX7zBXDORlIMn9WA/EeWgg/Vep+j510o6+hOpUDUOXcrUmogKOjKpxRtR82vbB0EwaxyEUsTgvDY3jw1CYj6wT1fISq78iuXdiGJzigi9WVe1G/VxSdidx0DKMViVR3Q7EgaRkVL+ZcWgbB4gilEoVWD8qZTpLReVWNuYGZcTRMAP7vEXogdSVmCThPQnX6Yk5RE07AFARXmmI40ql7Nb3gyYAACAASURBVNeBepND1UwkDxZWWY25LJy2M7t9KNlLX343AGBqhp3gTpzy/NrnF9n8lMt8mTEqxelpbk8iWYyUQxkA0nK5mGvjQAY3kqRc1blqfBoxSUlZHvNcaM74dXriEDMPzUpIUNTy878vnMOayLxLXoUbTfOzTV/LZrzBee+Y1NUMT4Zz2YkTkL4vGzkN6jplvzF63lB5hS1vuK4bylZkTHRpdQAXBxVuQEBAQEDAs4YtW6ZVCl1fIEZZc6gINdFTzO6gITnh5hd5x92c8DubO158FwCgNXkUAHB21ueLe+YZdhPfu5d3P9PT/jqSu2fG0WE4kEwjulcwjgVOnIcicSZSpyIAKFXZiB8rD6uRXIu9iRq1nX3qce5JHPI839AwrrvGhuFfndnJWT4WJPzn/Dlv8O8PuI4nHO8Mn7AOQ5oLUrUI5n65Ux5JLbOu4AzrNqVXRmKET9r+PnGXd/oZlFjDhISIRF1p8M6zbwgYShKq4APlxyzmC5xzSAdDVAzXbCQSgd7SZntQ6bTIkWk0M8p329c2R14CrYlDGyR0Y9D1z5oJB2oGzUji51ZZ72MzTojjnJcybS5H/okrOr/93E9lpx8na94PAKmEj/SFUMMZ6UFJIMaFKIrQbLZGeLI1u1G8gdOYBserFsxqIKJICRQkFMiKUnJdf6gOPH6MJyfE4UXm/0TLl/WFGGZ+yYeqaK7aSPhSDx70RAozO/gd/PJjnMXl/kd8Vg8SXtwo5+uXV41UVub690yw5Dowbc+GGfIxTXkHCWMxXjOFYCY3MfwTiMR7Lpc+rbRM5hnpN9fm9cOkdEVZ1oREfocm/KsujlxocN82Uq8N6KkEGhuHN+VR2UBa1EmuITjW+UfXCZ3bNjKlWEM0g49dtskfvxiCBBoQEBAQELANbEsC3fT/0chnHoDf4ZhYabTEdf/uV7+WzzE77Uik2d172b73kY9+2JdJZaurLM10u97dOxa7TmJykrpkbdJG395UbHeFncoEL8dFhgINuDUPqVRhUBdoSxk2TqozWmcDLfLzSftiYzxqSVC8hgSdMSEuulXSwGPaIByhEOxGwnL4t6T0cTaNg0plVkqSMSiJvbJvAvT1yrLkCLR5LZtCZTYt4UnVmrFtSxB7ScIxRqWOaHxsfs7BDVMMIi/FqWt/QXVn3d1T3i1Tab1tlkSCzIdqO/b9rVEodaGVnJrwu+2+EFD0lGrPSNt6v5IhD1CSkXaXf+sNX6bdlIr9j8g/VyaF6qeQmFyJJDbxOB6l1AOAfne8YSxRFKPRbI1QxFERGieUdyP0bKP0oJZWNF9jA83NRB7kkv9RpMDrbvT0cZn4Pgy6LElVEj+/Th3jkKP+0BJo8LhVhZSla6gsz4qPxtmz/Ds0duey2M913RjmfjwGEgbYmGF76qlTXnuUpr1Cw3C5cM6hn6boGWq9kvRlpVgH14d7zC9zOE5kROEJeR+TRPqh7TUpFaEVdbI2OrMO6FdHKfrSgclw1ONwuyb8uu5DJWUe2/ZJXxbhKBv4iyhJgk0ktFabZW3MDqURsobNECTQgICAgICAbSB8QAMCAgICAraBLYexENHGoQOqWqQNjLiqWTEsQE6stgcPceYVy/npVF0hp998801F2bKoElI5f2iui2J1AjJtEFXaUEInMsM/GYsKIhOjfmzYPpTNRNUA1lFmLfH/qE+LG9H2Xi7WOm1p33tVrmHlkPMmxEFrYqK1rmw0GTVGjm3knFPcR/vBqE+UiWnvQZ9I97ojzIe5NMfjdPasd1GviWu68sJqKBLgw43USSUpmbAKUVlGGzCKjCt5uVSGpJTA+DZhKJkcSqKmTUx/D8RpRfutYlSrXVGROVGPWbe7WNRNpYj7wT6ruuyryjc3mVfUoWgk5EPYjJSXd2jGUMO5ImF5Ghq1XVJhtWWSaIiZv26gjEfD9WEzLhvn7GY1cbPVgn1nY1LnD1F/Dy0vqqg/Cye49dk3inXH9NOqmllE3bjrgFfhpqWdcl++/uSxp4qy3lOsSrVJ1iPpg7r04dKyZ9J5+GHm1e0Ke1bNOF2pmlFV5Bj6BtanmIFoz7Uc1ndubrUoO33yGNKNWMK2CQJAZrxVNapa0IFhPtPMTBraVDbMWF3hGnZNMdfYrDnqfKj3MS9ApSlzVcfVjFO5yvPZ8h/r2luoW406u3gdReXqzHOp2rugRHYbrBsaimad7whrF/ULIkigAQEBAQEB28CYCBY9rERQ7BiKQn+e5jTUDCh2t6gGXZVUSmb3p05KU9McjG4dKmiDQFsNaVFJ1KZ5yyQfXyYZZGyweykRx5USX2clj4KLsZCIfFme55e8e7kYiFjCtM+zVvK0EqWW7ZRwlmbTS6CdDhvnNbPG0Oxo1SFDQ1astK2hBBqoD7PLvOHGGwAAX/019xTHctl5f/TDH1/XBh0rlYYtH6dKsxpWYZUc2t/qjm6fmcN8MBY4BwxyL+UAnoBDnbVSs8ONRJLQsIvIBPWXnORB7IljhXmgRr1V3A8AVm32DzlYlZCV1a4PnxgM1dnDt0GjAyriHJP3jbOJ3DPN1kvpsYyjhrZYqZtkbqik2zNV2mw040AUJ2hNTY+EByV6Y/H6cHa8UyW2YInBzlWd08V8tg5sMr9WVrk/P/6JzxRlew8wUcvrXvOVAIBlkw8UhUOVCaWR+utCMBH76Df0RcrXx6kaCVTbV/D4WocX0YaVxTGp0TKhHb3eiIbgchBRhHq5grIJaXKFBop/BwM/HwcSTqYaonLVXCeOZ9Wd4jA0cicJa5Rui+smm41KsfL4jenWmqvWctqKlqRQ+Zm5ImGGbl1WFr+eeW9Hq5lc4zxkMjD1ewN/r4sgSKABAQEBAQHbwLbCWDYKq9jIfrbWXjdqf5NdmHzDIyNJplLWmmAp8/DhG4uy3btZkjpy5IjU6aUFzWk4QuekuyuNSrESspSl4k4eW+lCLiisbkZaipXCT4WyyFq4COOLq6B1EujafrY2UJXwVLKzNIeNBu8Sle5sZCwKu/V621Ek0cd63YFrDxZlX/P61wEAKqZvPvKhD2vDAAATEz7rRRH0rBKb6beIlE5MzjRj4QX+9XZf59zYupsIkpvS0oGJFCdcbrnZc5JInCrpdAwhQiGx0+i5/Lc8q847S00or2RVgvQp8qE+7Z7Q1xm6TL23CKcjSTs0FEvnfGbIHArJTea5pXbL5JlJMvfUTJ7c0ph1VlmeY6ndRb1qcvFKbsyy/JbM/FK7qA9jMQQQGuKiv8bmW6mxTbEnUuCpkz6/ZiSdNj/HBCSdtr+uIRqUyPpASPhFsylSPwwhQqTvy6g9FvC25TRebxfvC13lww89BACYO+/JY+JSMkZbv0OeZSNSvbZR214zNHp2DZGrzd+y9uscH/FNGJXNRtpf+JC4dWXxBv41a9c8O+YFi6qu77bMKQkEtHBdna7wR/DP2e13RxmBNkGQQAMCAgICAraB8AENCAgICAjYBraskHHObapS3EiVu6F6V8tUerdeDGLRVXH8zjvvKkrUwK2JfVPjfqz8qUPjAp1K2EpxvlHPEUZVl85kFfHcoKqits1T9e5GhmYq6rt8rOfCXYuNQoo2CkuhNSqOjdTwRbutQ5e47B8UVqi7X313UVYX1d6nP+0dMk5K2EqizhOWgFL5KjVx7wjnroyZDI9tuzpwbNQNPB/XH98WnIPLhkVScACA8vBijSMDvOY7HlHhM4biPJSI48Lqsg9LiCrCoZuoSturVjUZfLcn6r7UuPNH2g8mc0Q06gCXmcwukTAIJSU1k/gMH11RGUYx11WyWVz0PnKb2KgoB73xhrF0u1186f77iwxCANAShqaaqHUrRo3YbHDoiDJdlY3ppyIJ2zU8qlr2amF1ztHxq9VXirJem81Cn/zEJwEAK0s+CfbUJHPTZmYcVjPJmCN9Ztug9RdOerReramcvalRoyo38akTrFpO+35dS8rlMZqFsK4uv4as55pWc9BGa4qGmlC8nrN43Zq1UaLrDc7diON4nQrXqIdLpdF3b62DIQAM1aGJrHpX6tQ0O+a67lJnJCPVZggSaEBAQEBAwDZAW8lsQUTnARx79pqzKY4AGAA4dYXuvxUcds7tutxKnsP+rgC4Xn5PAji3+elXHa72/r4DwFEAKxc57/mEq73PLwX7wXP+6QuU3w7gOK6OcXsh9Pfl4GUAHgTQv9iJY8Il9feWPqBXEkT0LgAnnHNvu9JteaGBiH4LwLJz7kevdFteiCCiowDe6pz78MXODXjuQERvB3CDc+47r3RbXgh4Nuc5ETkANzrnnhh33ZeDoMINAIDDAB7aqIAK3rGAKwkiGjvpSUDAc4UX6vy9aj+gRHQXEd1LRCtE9IcAqqbs+4noCSKaJ6I/JaL9puxriehRIloiov+HiD5ORG+9Ig/xPAARfRTAVwN4JxGtEtF7iOjXiOgviKgN4KuJaJKIfoeIzhPRMSJ6G4kln4hiIvpFIpoloqeJ6F8TkXuhvjCXgZcQ0f0yL/+QiKrAReeyI6IfJKLHATxOjF8monNEtExEDxDRi+TcChH9n0R0nIjOEtGvE1HtAm35Bwci+ikiOinryaNE9DVSVJa5vUJEDxHRV5hrjhLRG+TvtxPRe2XsVmRtevEVeZirEET0bgDXAPiArCM/KfP3+4joOICPEtE9RHRizXW2j2Mi+vdE9KT08ReI6NAG93o1ET1DRPc8F8+2KdSr9mr6B6AM1tP/KNgl91sBDAH8HIDXA5gF8FKw/eL/AvC3ct1OAMsA3gz2MP5hue6tV/qZruZ/AP5G+wjAuwAsAfgq8AarCuB3ALwfQAtsi34MwPfJ+T8A4GEABwFMA/gw2J0vudLPdbX8A9s/Pwe2ue0A8Ij02wXnslznAPy1XFMD8HUAvgBgCuwrfSuAfXLuLwP4Uzm3BeADAN5xpZ/9avgH4GYAzwDYL/8/Arb5vx1AD8AbwbwG7wDw2TXj9gb5++2ylnyrrEk/Dradlq70810t/9b01xGZv78DoCHz9x6wGe5C1/wEgAdkvAjAiwHMSJkDcAOAfyRj+Yor/bzOuatWAn0VeJL+N+fc0Dn3XgCfl7LvAPA/nHP3Ouf6AH4awN1EdAT8IjzknPtj51wK4FcBnHnOW//8x/udc59yHC8xBPDPAPy0c27FOXcUwC8C+C45958C+BXn3Ann3AKAX7giLb768avOuVPOuXnwx+0l2HwuK97hnJt3znXBY9ECcAvYf+ER59xp4jikfwngR+XcFQD/GTxuAUAG3qDcRkQl59xR59yTUvZJ59xfOOcyAO8GL9oXwhecc+91zg0B/BJ4c/mqZ7Xlz3+83TnXlvl7MbwVwNucc486xpecc3Om/NsA/L8Avt4597lnpbVbxNX6Ad0P4KSTrYfgmCkrvMicc6sA5gAckLJnTJkDMKIyCLgkPGP+3gnezFjPvWPg/gbW9PmavwM87EauA6CJzeeyws7njwJ4J4D/G8A5IvoNIpoAsAtAHcAXiGiRiBYB/KUc/wcPx44nPwKWIs8R0R8YVfnacaluYn6wY5GD15b9Fzg3gLGV9eAQgCc3Kf8RAH/knHvw8po0PlytH9DTAA4QjUT8XiO/p8BOLwAAImoAmAGHX5wGqxK1jOz/Ay4ZduMyC5Z8Dptj14D7G1jT5+CXIODSsNlcVoy4yTvnftU59zIAtwG4Caz2mgXQBXC7c25K/k0655rP9gM8X+Cce49z7tXg/nYA/ss2qinmtvgAHMTzI6zuucJGIR32WBu80QNQOCjaTd4zYNX6hfBtAN5ERD98OY0cJ67WD+hnAKQAfoiISkT0ZgCvkLLfB/AviOglRFQBq6r+TlSLfw7gDiJ6k+wifxDA3ue++S8ciGrrjwD8PBG1iOgwgB8D8Ltyyh8B+GEiOkBEUwB+6go19fmIzebyOhDRy4nolURUAi9GPQC5SEO/CeCXiWi3nHuAiL7uOXmKqxxEdDMRvV76uAfebFwiXfgIXkZEb5a15UfAMYmfHWNTn+84C+C6TcofA0v43yBz+G1g1brivwP4T0R0ozjM3UlEM6b8FICvAa83/+u4G78dXJUfUOfcAOwI9D0A5gG8BcAfS9mHAfxHAO8DSz/XQ2w9zrlZ8C7l/wCrwm4D8Pd47oJvX6j4N+AF+ykAnwTwHgD/Q8p+E8CHANwP4IsA/gK8+cnWVxNgsdlcvgAmwP29AFb9zgH4r1L2UwCeAPBZIloGO3Pd/Oy0/HmHCtg2PwtW2e4G25u3iveD16IFsA/Am8UeGsB4B4C3iQnhW9cWOueWAPwr8IfyJHhNsSa2XwJvyD8Edgb9LbDzka3jOPgj+u/oKoiueN4QKWwHomY5AeA7nHMfu9Lt+YcAIvp6AL/unDt80ZMDAp4noEC6ELABrkoJ9HJARF9HRFOirvn3YHfooGZ5lkBENSJ6IxElRHQAwP8G4E+udLsCAgICnm284D6gAO4Ge3LNAvgmAG+6RBfqgO2BAPwsWK31RXCM489c0RYFBAQEPAd4QatwAwICAgICni28ECXQgICAgICAZx1b4ittNWtuZkcLWWaTlnKoZqvGVLV9k8w6lmSsUayJZ/3t+gNOGLu4zMlrbT7tSDL59iQp8WCw3uM8kUSq9YqvUxNkJybBba05IYXczl63XZQNB0NpJ3tS15uTvg0x1zEccjubzZZ/rpjvmeZclqXeydc54OTJM1iYX7rsDLj1VtNN7tyBkYoiSfAticQzkwi2Wm0AAEpyTrXkExFrm5e7XXku3+Ys536LZLxs5FaWyXhupKmQBLUjCZ4l8XAsjbauuBGNJu61mbs1yXYsVwxNAuO4SATN57QXfcLjfrsH5wDn3GX3d6WcuEatjCjyVZXL3CexJPh1pr81TFn7r9vrmzI+vyRzUfsfQFG/RjnbhudSV5Zn8rs+QbBFJJUQtM71ye516Da6Xgvz0WhTfYqR57SYX+7OujGk12o0d7ipmQMYDM1zOm2FM63AaFvWJoEHNp6jxXX8q30fmwUnkb91/Rh9XjdyPf9H3hc5uJEWTxNC6zhy/ZtM0TWPM1InEWbPnsTK0vxlz3GyE+QCTYhMYuyNEmmbugAAVUluXqn6iJSkxH9rQu7YJMjWvtHr7bzUZPZDk9Re773RPMwlAbcfX3+fTocTpff7/F5uVdd6KWvKlj6gO2cm8LM//RYsrXSKY0urvEDccevtAIDT5z3z0nXXM0lHJAvgi2/zXvWPP8Up+H7zXb8BAKhn80VZ9+wsAOD+Jzjz+1LHZIMfcDeUZJAPTfpHuH6GB3J3039AGxN8755Mj+OzfnKsyHrXcTzYUaVRlN11110AgNe94bUAgDe84Z8UZdUqf0zPt9kD+/z5x4uyGir45m/+XowDrZlpvOU//hAo85NpqsUbAh3a1bOPFmWuzBuA3VMcP//Sa28qyg7u5r7/xAPMgLXaWSzKVoeSdV76NEl7RVkv53uX5eOV9v2CkFamAAA7TFb40mAVALDU4Y3K8bb/qOjU1kVlom5CwGTIFheZGKYKP4/mc16nGw3+ffrezxRlD3zoc1ic85uiy0GrWcM3fc1LMGnmz+4ZHuuJKh9Lu75v0iE/x/lV7qMHHztalPU7fF69yl74MzM7i7LJaR6nco3na61RxJajUpH7DPn6Xs+/a+1Vfs6V1dXiWCYb2FLC1zmz8RgMuO91w2s3OvqBj2XMR9amSMtko2g2xb0eb9ze++GHxpJTsjm1D2/5wT/A3PLA3yPjxhRLofNzrlLmOVOVxTpz/ros4z4ryYJuF9OSzNGJCR6PiYbfXFblT8pT+b8ff47IAhD5++i+tCx9p5tsbqvUtdHuSAp1k2k/CPq3fjiyzD9ztVrFz/7Qt+DZxmST5+E9r3tNceze+x4AABw/uZ4vYt9entP/8vu+HwBw3ZHbirL+kPumm/Iz9/p+HVhZ4fSq585x2uHZeZ9+eLXDZWUzBvU6r8sD2aCurvj53+/z+1Gr86C88hVFLgAM27zG/eUH/xwA8ODDjxRlqQqBxRCYzfwWzJpBhRsQEBAQELANhA9oQEBAQEDANrD1nI0UF/ZHAMiJVQ1Loto5veS5mStdFpMnhO6zG3kVaQdcx4lzZwEA+zBblPWWuM7FNv8u97xIrSqBOGI1yGNDr+rQVk1W/L4gW2K11+PnWTVyzgS0OFVRiWplcso/19QuVk9SwnUtry4XZcOUxX1VH3SNmu3/b+/NmiS7jvRAv3vsEbkvtReKRWwECaLJFkmR6m62htK0etQazVAPsrF5G7ORzdM8yUw/RQ8ymUxtUksmUb1YU62WIDbVAJpN7CAI1L5mZuUWkbHHXfXg37nuWQSByqwAn44/IAtxI+Kee865N9w///zzMApPDrZ/mjlEtUjgDB8pjE6Lc85R2iqPdY84NzjJGN46GnXLY5MH14iIaJjwfExIrpUq/F3emN/fIZmkdo3hYB8w3s5U4NJOi9cz6cnavf3mX/J5AIcvKhjZAYQ4BhTjq4ky10XRIhERpbGcx88YiomiS0REtHxO9LtXNhdp1J+T0FRekJuk5BcyNxHyY7Uaz9EkVxCpgdsw9JWV1fJYzeT/AcWZXBARkYfrHo14nntDudY6oOKax++JCjlfFPBYwppAviPAnYZ7EKv7weQVzbkrkUDmnwQjlp+LAR/nyB2p1KnjzNfnzrKCDvsTyrQv75p8OLgTrozbc/k1l0yOWfEqDHyKOVPIL80wF4dIMYyHck+1AetGIX+3gan1a4HKNhR4BpkRu67AwY/nmz2VT3eQBjFpSA0Vmvd/Us4xTdMyH/x5motxtepyscGnLPfSAj8jl1a590F/JKmmEZ5FXaQy9nuSMuoe4rmEqalURbK51gTXRF3udMbfYdbQ9WR8Wc7Hctyz+rtefO45IiJqd3ic9T/6QXnsrTff5O+OsUnc4/fBk8K4NgK1Zs2aNWvWTmEnj0CJyPXFQ88IXjg89Z19SQg3TQCEiOj2vffLY7duMfHGAwmgpaLaKVxew8abpYrJBvZMCi8wdBWBBV7ELFfehMPHRyDi7CpCUoZo4qWXOUr6/j+SRP2Xv/oSERFtPeDk+XAsUUKa87QFPnuuFa8j5/s0pt0JLc8Tmg4eUSWqlK/t9nl+Dw8NSUWWMERUbpjNH3oqcu8z5yOs8ljbgRxbcqb4PEdEaxU5X8PluTeedCuW+X77Z+zF3T6UCDRb46Y5TRBkAjX2YsxRfJ7yWsSOeIs59sFqk+d0r7tQHgviW0RE5E74PLWmsKUXLi7S/duy557G8jSl8d4BVQqJ8Lx1Hodn9qdCNxzsA5oYz1i+yxBTmiFffxxLVDOcMVEiB/vbEICIiDzsnxREC1dF4iEYvY5sYQpCjBXncRRxMHuM0atZwgnWoIq11h63YUCbb/J9ibDmHYE6rkNRtUoV9UwpEBW7uI8dRRE2w0xAsvJ8GY8hkpgoLlBRv7m+FM8BzfLujxhBCmY8hkhdYwGt81wHyEAVMtxvhs1vPkFEFCLadzRxq4xAf5F9aljBQuqSdczz/NMIxnOzAiH7ZCws9wxsfzOaWij383NXmTSU5Tzvd0AMJSJygGLMsHbxVMh3nSbvY/NcGyuC0WgEhGwiKNj+Pt/3BYg/nZbc/xmiUwrBklfranTqr1z5EhER/Z//UCLXBtbnf/wVt5keagY9PbnZCNSaNWvWrFk7hZ0oAo2TlO493KWJ8ia29jinsLPLZSjvv/9ReWxni8s8zm7eIyKin/30p+WxcZ+/owHv21F5FuOLlvkE5X75oKY3A/YTVtviHYdw+wexRKwzhAVLCxzZ3OmKd0Xw9l7++hUiIlpeFc/m+rX3iIgombEH1lS09zrqRac5R1RRVbwyJ4lpXlFo4GS06h9RkA3k3CZ/gqi8kckSVnD9ccFztL8nkVmvz9+x3uK/X9jcKI9dqKMuC/nOQNUsUsoLE6Ipwu2dvfLQ/oSv/7e//NXytW7IOdlbBw/xcZVPQr4qRClCVUW6xukfo0ykUpHcbpHyfFczPndvpnLVG6vkBZ/Wg/ckVlBeZJSrsqEKkkABCltdV+ZmiihohPshmanyH+Qia1XUw6nwtERNTA4+VXVw+K4aosZC5TSPBuyd709kfEPkByttRhZcFT2ZWjwT6cSpfM6UXeazBJ9Ts4D9U34ulmggy+YbCrmOQ1Hokh+qem5Tc4t50WU0JjIz9bmOK8cCrJWOmI35mOscJWuxWmMfHAOTK05GwmkwS6Oj/jTjPe0hooxnv5iDrxiuhuqaFgLF8D5h3c0jLgjN9R2/BvcTctXzNgPc5SrPb8ZlVv3shpRjffOr3GGyjs8tVxSSUiZP+ZpzlVd1ceweSh7396SEMcMJj1St9wT1nOZ5kavSpRT17HnO3z9WazcAt8DHvbu8fr489jt///tERBRWOef647/47+Wxo9H4iTPONgK1Zs2aNWvWTmH2B9SaNWvWrFk7hZ0Iwj3sDujf/Psf00wlXGeJISqAfq1lmUYMvTmA5TS8ZHgB2/tMbw4i+U6vhGxBzlDw1zJUYi6vVPA5uYT+GESBoSptAXmj1uCyioWqQCPtDVa2MTDyB2+/WR5LUoat1je5ZMJXZAWjVNIfMDQwGAts0AxOxcv6RJskCX2w/YhWawKbtEAayaEWNFTKQM0ajzFyeTxfrAukEjX4O1YXuNTiwqJ8ZwZSy+GYr3mUCRTpAs5xJ8wIKyYCt3zjmbNERPTS+mL52kcoS/rrh1w24zUVRb3GcEkT9QbZQEHtgM8OQHFPfRmDYXC0Ah7LdCCkpWazcUyW7WmscD3K6i1yazJvhgTlA7oNPdk/kWeO8d4dDQQy94H5+dgrhZKqm0BJaFhAakzJllUhZehGvOcX6gJzm3ttMJa56WKfdkLe3xp+NvJCZgyOguY9QIUpoOJYkygyI6FnYGulzOPOF8J1XKIoJAp8gSjNXCcOCCyqHsWoC6Ug9vWXTQAAIABJREFU8jSbQvhynOMkHa3mY2pxjKSco1hBBknNALtGqmZlAoh7NNUwLcaVH5eR0wYlUAoiXeLCJzKko1CVFWVYtzFKm/zgOAHqE2UY52wZLiRQhE7vMdLYuU0p1Wo2ec8NDvme99VzYzTk6xgjBZFnkgYwJLqtMaRaFelzimfp0VBK8Mx6FtiHhdqDGdY8xdweqHKZpQF/RwXQb6AkXtsovfnd3/09/v+a9O3+z//tVTrsq1Tfp5iNQK1Zs2bNmrVT2InCpSTJaGfnkMZTTRVm72Bjk6PNxY54hAtt/lU3kaerIoUUXlicosxEEYXOVfm1CxtcQhAcivdyCWIHVzY4sukn4r0cTdh76U0UdRyev+PzmDt1ueQvXl4jIi5fIBLtUL4qeE4gUGSJeFeGYp2N4dVmKkFeceZWyZJkBW2NUtpoSUnHuODx9yDmoBn0V5f5emb7XLLy8jnR+v76i6ztm7u8Pt37QrwZQSSi5vGxKBTh/PGMoyW3wvNxZV3IRwWihvHBjfK1zpC9y+9CozjOJTrPoHcaT3m+PrgjOr6TBT63C4/1aCYe4BBz317iPXZ5RcqGJrFbEsue1tKsoP1+SqESrohBhU+K44LjRETnN3l+fYhG3Pr4WnlsWiCi6KPkSfmqTp/3Tz5FSYVaxAzaqy5IMtUlWfs0MyUBqhTLgY6za8hK6pYu1cH53Jr8Z0QgPFM+4ihyFBoIJCDvaCHwfM7lFJ7rUKcZkltIpBZ4/NxIUYKW5DI2E4mZ6DJN5dmQF0Y4AmQdXVdUCr+b75ELmc2MIAbIQaoJQwQyT6b0bitVlK8ZvVYVHRot22qV93OoIuvCEF7wku/JPjONNoy2cZrJGk8mk2NEys/NcBkaZHAfQxzCikRq793gZ8i9u/y8GXQlahyBzDPA/Rxn8mwdQnhmjDV0tVBHKfAhe87o4s5mLfy/rI/ZD2bexwoN7CEa7XR0mSHOg/3Q7jB69r2/93tyvtYi/ac//I+/8JlPMhuBWrNmzZo1a6ewE0WgUeTTpUtLdNgTT9bkQ9eW2eNaWpDoxeQkTPGyq/IOObzpAbxxVzVGWl/m6NIUKjdUKuPSMg+54cObi8VDaqKYdjAQlyYw0S+803Yoxxbqxovl6xmrThum7dEAWP7Pbku0tIhqlymiLc+TCCLoNOZWbO44LkVeeGyVxvDeKhF7grFqz7Z9/w4REf0fL7IIxMXV9fLY/YfbRESUoAtLciS5TEO1D6rodKOi9BCReBVyZ55aw8k+yzbmkaAObp29vRD5ioma031IfR1BnGFR5Uy8OnuXlTpP7ldcibxMjnkFnv9CW873cJhQOKcI1CkK8mZTokTl/CBMYFrrDfuSYwlMWyvkwlQ6nvIJ790JPHFX5R+biEoWF3GvqJzTZMIRf4j92gz0PcPfmatoyEWO1vkEaTuzDROMz5QPEREVeP8MqItuQ5hAHjBH1D2NfzHCmpsVBRVpRsdSfNgfRmozDHXrMZSOIH+WKVm3wjnOw/B0XtU3ER5KZNSWKcAjyBGVRKpLUAv58JkqHTL6DHXsRz3npuTCcc0cqvOY3CxyurNUoqWo/A6UfWQ67xvSPAVafpkZzYd0ptYYsog15IWbDbkvP7rNkedP//oNvFU9Pw1SgzK7TE24j7xw3bQKVCVMOeZhqDoOmXtigDIu35co+PF8d6qEFPR3EBE1GsLHMHnRAl1joqo8U779m9+l//rqf6UnMRuBWrNmzZo1a6cw+wNqzZo1a9asncJOBOEWeUGzaUINpb+61ObQPvQN9CDwUqnriN9pV8mdJCa5DPiko9R8bt9jmOzRAUMCGysSXlehdpFDiaeqKc2Aodxjyvr874MjhmJffF46ebgVJn9s32Ct1cgR6GJqSm9GDNPuPytJ8ItnWng/z4MfyLFCqWQ8vTlERYUmqVbG4b8zQCK6hGbaZcWnHx8yKeiPunLs+Q0e63MgXyUTmTfTSX46McQuuYbZhNfC9bnsparg0gKwSVWt3SDheQ4Ab8UK8vUShlQc6F16SvHk1iOcZ4n3zHJTYJoJSCQ1EGRWlN5r3fHImxO6VXFSuhp0qRMKpNom3gcDQLFTpYk8ApY3QhpiqS3qSfUOQ9FNzFenKrDgQpvXoFY1ZTCyFjVAtvGY50pxSWhvwvN9mAne+dEej+f+g/t8DaoRcWBKaExvZ0Xiq6KjSxQZLVxZ8zw3TZ1NVxOl2TpnJDHPicbj9JjebZzxNaUoL/MjgTNrIbrb4NkSKjgvd4xyEq9Vqvax+XfumG4uMk9VQ+DBcyohmfQZ4PJI6Uqb5UqhQKTLbMrpMd1V1P1iKpkM1OySUpQC8TGLMRY15w7RfDs8/RJz0NHEUXi66cxy6RnWk33lK9Kw+uYe37M9aGGnM1FMI6NAFiMNkKn5bvB90oAet6v0sv06pzV2dqSrV6/H5KR2ewF/hRS0u8fqZKJ1LGth4Fyjr6vTDxFKiKpIa7hTueZ6PXpi0paNQK1Zs2bNmrVT2Iki0CwvaDROqNMWr8/0Bp3E/GsfaSaF6apgej0qN6oAsaEGT9jVDfdy9hiaDX7tzBnRqHWR/P3oJnseD4/Eq3jQZy9C1cGXdPUEEWusSjR+jkg37HIh8FpDPpgjGquAifGNl79SHltaZurzEcodxiMp7B+ORpTnivX0FOZ5AXVaa1RTY67C5RkOeL7rrlx/s8Pv+/N3f0ZERMuqN97f+8av4zXo5VZljCGKiF2X57aqioojCDf4ZPrtyfpG5/h8mqK+gDKmPDMdSqTMoNbi9wdAH9KGIAtDrNN2zGt+tyt75fCA18cBuWwwleuiwqUk0wyU01s1cOnLazXyWnKN1ZSj+YMB/81iFRIazVZDWtG6p/Bsm/CWn3vxS+WxZ65wf9yDh4x8TPcfyOfguXsoF+gPZP6m2IvPXhVNz77H5LAPrvN3TFRvSqfU3MWmUWIgfgCyEjxxX3VDMf1KK7i3C0XGcedE2DKWFwUNpwnlqhyFUG7jIBqvVFUUF0LkAR1AAjXukjSVmu9WJSS4pqIU3dYkHSBJKCtxlPBtgn3pe7ojDb4f652o6Dl9nMyiIhkzdaazkQJgShGHcsQ6AvoV6ODyKbHudXnefONv8HNjcY31wt9+Vzpqvfkx79/BgAmJ8UQi0AAR+8iUajmq1M/jZ3doyobU3H7hWb5PlpbWytdaLX7ebqwzKfK3fus3ymO///u/z+dBlJmpKNMQiwwiOVVR5gy61VPzu6NIYqORU3Yr+iyzEag1a9asWbN2Cju57pzrShdvIkpNhIeXPJUrqFXLFvFERBSolitZavKVxjsWL2Rlib2CGUQS6i2JVEzuz3jVR7FEjT4qcqNAzjOEbNxik6OrdqteHounkIjbuMTjVR0D/BxdXJ55Acckv2U6a9QQ6nb7kms86g4oS+cTEbkOUTVyKWqIVJ7rQwhhj2UHW67k5Gq1JSIi+sK3+O9GTbzzlWXOH9Tg49aaMm9uyNHOYMSR5CSVbWEQggCRWKwisCDAuij5uBB55RyeYDaQ8ZnC/Aa+88r5M+WxKiQAR/Bmb6hI6kKDx7cMT/TtPSklWfQ9iufUIWScefTOoE1tV3Iy+QgdV5Brnk0lGnBMCIGk2BiCFERETXSHWHmBO9V86Xv/uDy2hCg9Gf4BERFVJpLvmQzRRxFCGaGKBvyCxSVavtx/z27yuk4PeJ4PDlRHGMzLFBHyTHV2SY3HjV6wriqTCBGpBvgbqTKDupIWnIcVBVGSO49FWfg3BDQGOncFdMRBtBioSC0oE7Qm96u+E88GF9fkHKtjQXRqZBRVTtOkAxNPiTIgmvWMdKAag+ebLjHgfSg0ypR2zNDhiVT0HPimf6iRT1R5Vcc51h/08zJz/atnL5WvrS4xgtId8FpMFBqxdukcERHlD/laa6mUuKwuQyoU4/ZUz1sjMlEBgqeflwV+T2p1eU6b8hPDGXj2C1fkGLgWA+RJc7VXUuSvTZ/TQAlkGPnPMkpVQjnkOvSkDVhtBGrNmjVr1qydwuwPqDVr1qxZs3YKOzGEWxTOMU2MCESDGPCcbmjaAJw3HPOxllK4qaAeY6HFn68HEjJXTTcGQFBHQwnLD6Ycck+mfOyZVSG83NnnMFwH36YrxiIUkjbPXy6PbYRtvIevSJcAZLjKtfMMU/ifEP57hekMoRoVJ3GZjH9a81yH2hWfBn1RDToL4s5mBaSLkcBA37p4lT+Hub12T7RZS5QT83w0EN3KHKUZvaHpfiPXs9gAzB3xfCRT1dw8YNJAoMlhDpNTDImidySw5q073GR7hE4L59cEFm+hSflmn0txekr95vmvso7v9S6PsxIJqWzrxnWK4/mUDo2TnN7e7tM5xQFbgtLVGASE2Vh2v1HDIZQzrLWE3HR5ma/thWd4/ywqyPz+zxl+37vLZK8KCYEmarMk1/iQ9/LBwUF5LEcT+UipE60t8Hleusydce7Rw/LYIzRR75VkCtkruCXLEqk002uICQD0VVFKSWcD6cYxF3OIyPVI+/Kuge9TlB+ocpQsQ9kNCDyOKgXxyDTUNnU76jS4Z3Ncm+eofWzYR4BbvVy+My8bccv4MugcO/RJRCEzBp6zbKJUo3CfBHiPJs+UqSx8l+6+8quAb7U1m5I2MNsiwxz95vf+TnmsvspEnzdee42IiCZDefavLHEayZQy6usxkKxRAzp4JF2Mbn6ELk6q5Mpo4RpN5j2UrhARRVA1MnM0nWmVPJDR8F1apcjo6Ua+6RiuyiE9z0K41qxZs2bN2udpJxNSKApKkoSipnhODZQ1FIXRfjxekErEfRaJiFptSQxXKvwdzRZ7781QIhVCUbHpYzhRdO8D9Po8Svi3v+HJwbEpbFaetiFELC2xV7V+5mJ5zAnZew/gjYShECQKMmr9/DmtTTkGZXo4ZTLLsC/R3FF3r+zt97Tmuj7Vasv0YPuevHiLz1UjHuu+ir7Wm5zE/+ImJ/DriZBtTHF5tMxCEjVPIvcRiDIBLt8QgYiIjJNsShlMJxoiIhdF6SHJGAIQEcxsJYUqMwBR7MbWFhERjfsSXZ0HRT2c8txerMoY7h3x+PZB4MmmQky6+fFtmk3nE4EWRDRzHEqUlnGMfT1DFfwxKrwpdcKeX1qT6OyZdV6LRfRm3frJH5bHdm9/RERETZej9UR1Qtkf8rXs7DNhKJ7KXrp4hklXmSoliTAVlZzvlRVFMDoc8N699pAJWvd6Mm8+ojwPxJuRQhHSsjwHfVvV3nfmw487bo50PyIickDKMgIT+pkSAp0y2ypwpIQqRNTgmfKVQp4NptyOEMV76pibGfIh+mEq9YoYkWSsSD0h/m2+UhfdexizdDH5xT7HIbSjFchigt+S8PJ4/8/iV6CkYOJcHaktLoHAWINurSqNMxophYm6K/K7EGMOTc9Uo0VMRDTDsRTds2Klw2zmUpdLGUTv6IjvievXr5fHpogyzXtmquPQGN9vSrU0Mlj2rfZMKY2qfaTwifuv2gjUmjVr1qxZO4WdKAL1PJdazQpVlSxZVOVf7hB5N08VIXcQmSzVDB1eftXNv7Mpe+HBTLxjgz8bVnqivIEpwtEB3LeeCk/HeC1QBeMmAm2ijGV1Y7k8FtQZwx93OVLb3bpVHjNph+EuurI8vFseM5JQgyF7REd9weSHgz6NhkrS6inMcX0KawvUqEgEmiMvtDdGlKA6odxFj88X1jj6efHcxfJYb5+l3kwpQ3tJOrXkIWTqDjm6rSslCtNxgpCHSlWZkgm0E0e88wNElwk6rXy0JR0R3n3/QyIiun6Hi/8fNSW3czCEHBhKQ/qLMr6dA/6OAhHv0R1Zi0f3DynV4gZPYa7jUCUIyFM5J1OoX0GuZdKXCNzIwyVjXm9de11F7tOLeez9R9J/dQViAJ0F3n89CVKo/4jXogr05eyGRLW1OsaghCMCnDNEOLTUEWShGvJ1+LgH4lj20T7KuyIjZaZ6kporzNGVpaq6jVQjJXgyJ/OdglzVFNL1TNmbOb+gEabzSYGuLLOh3HtbO3x9f+s3vsnvKWTct+/dISKiOvLUdcV3qKDbS+8RlxO1anJPpZAtnSmhFx951wx7tVDRkpEYTCFp2t+XnPSkz8+LGvLcs0witoWVNYwZEoCq/CXP87nxKj7NzK7f2dkuX2tACMR0WdIle0HM93gpaajy6BlKdSSSVJKepifrJ/SbNdfZVb1Fx2PekQ7mxogmEBENBnzvGR6Ejp5nGNcnzZzJmQZAInQpUpq4T5oCtRGoNWvWrFmzdhqzP6DWrFmzZs3aKexEEG4QeHRmffGYMkmExHEAandrJiF+Ff822prpSKmGmCa5AzQcXhQCTwUQ4nCfQ/BENdSdgFTRHXLonSioY5Y9rndJlGYMqQxRcrC0KBTt1ioTarIlhnV3H9wujz28/QGumb/zYS6wnlESiQFhJLFgcGmalQoYT2tFFtNkeJ9ajsCg2yO+3pXNZ4mIqAcdXyKiO9CMHc4Y8mi3Ba6uQNmjv3OHx6nWqdliyHdvn88zmwkMEgUgNWRG21Kuddhl+KQaCRTVR6nJ0YhLb956X+b01b/+ORERNZAC+NqLZ8tj7QWGHj++y5T2BwMhH3XWeM0ckKKuf3y/PEZJNrdOFQ4VVMkScqcyN17K17u6zKUzqtlQ2RB5hDSE6TxBRBSj/MFASi1VCtJCNxa/zvMetuRz7SbKSi4AOlTQanfCEOBQEceMMhUFfM9UGkq5CApT5PPanzmQ6xo+ZLgyw/W1VJlWzed7MTZqUomseTKbz9425jpElcA5pmBmYL8IilCdppQ7TaHUlBd8LT957Yflsd0HTC65CNT75Ze/XB576wFruD7E8+PChmitvvw830t37rxLRETrZ86Vx4Yo8dpXKk4bK3UzUPyV5+EByrZ293gf37v+s/JYhvuy0uJzr5y5Wh7rQLnHQJgasnUch5w5N9TWlTHmVOalwVAg0ke7TPTLkIbLQ4HTTVlihBKfwwPRBN8D7NrAc6fZkjXMJrNjJ9R7ajQ0zyDZc9OQ19xoCH98TcrzDLHIzJcmOQa4DjNiXQ4k0HJhXiiPMWnNlrFYs2bNmjVrn5udKAL1PY8WO03yQ/FWQ3iO1V1OPEfpY5qCRJRD67BQkZrrmB6h/P+pKl6v1ECxBwU8HUtE1EKHFrcLlX9F3DDCC7nuzwdKe3/EY1horpTHVle4K0YFRIqzm6Kx+Ef/4d8SEdG7b/4ZERHV5ZLLphZpWa8t3kqep09chPtZNotndO/+DUqUxuooBSlqEbqyDemNN4Be5f19jgzrVYlGXERAVRQxpyqK8RAxN9r8/slQjk2mpuMEn7evSFtGErmqdGuHWIM+Sk4W20K+WOnwdms3IJShumX8/C6P+YNb7IHG6+IFX/oyR1L7H98gIqIoV91f2hHtzubT/YaKnIpsRqpqgiYQUmguouSprjxjoCcpxCUqmrwG3U4PhJR6Q+YhNAiOi962CjEx3W5cCFccHKjyrpjnZDZWep9GtxdRY64ilQaIfbOEowjdf3IJZJoC61mojiIVsPcm0IPujmTNp8m861gKcvLkWNmChxoVU8biKGEDB8+XYRfdZw6FUBbljFD82Q/+NRERrbXkmv7mKxxl/uEf/ykREd3pSQecb7/I2q+/8TXuBOKPZY3/9D/9CRER7ShU5je/8wqPAehAvSUasNsg0V27wXv13MZSeWyM3rLnz28QEdHZy6I5a3ZwDqRL6/jm+ZyLWJzjkVML+sIvf+3rRET03Mt/ozzmAeFIUCY1GEmE9+AGk7b2dpnI1Uc0SKSECoCM9HtSUmdK70z0mKu5TccgJqkOTw600BPcXx9+9JF81z6fuwWUYqrQvyrOU6R8fbqTUorez9PQoB1qRqa5JRFZs2bNmjVrn6edsIzFo2anTbWG4NneAXshfs7euC7yNmUomYk0VMG4KapvVPj9Q9X3cFjj93cW4PXsSO4mjNk7XQQmr5zFsr+ebrQwhMfqAbtvtTbKYwEiJ1NUvrAgRcK/9/3/i4iI6k0uJH791R+Ux9IZ5/dcl084U/nEOE7nSDl3qMgjqrala0kyZE+ui9znSkc6tewnPDdv3ea845rqYtNZQF/PUiRBPLXZFGUY+DsYqW42KJcZIRc0S2RyL5zhvpQP70ue51/9yZ8TEVEPQEQQyHlilF88POC1vrMvucwhxBxM3m1zXdZpMuQI6v5HfF2XNyXqvnGrS+6cpM4KIoqLlDKSEofDAY9/kLOnm6pCbT9FBIk9ubAkkUgH6+KhzCt3ZU8kiDfiCc9pXMgcmY5AvS6f73BPZM76kLQcjlSfyxq/P56hLEzlTKtNvt+GkIKcjqQ0oBYx2rCE+yI9ktIr11D7gVrMKsJPGI7UDTcHcxyHvMCnTEnrBSg/eXT3HSIiWrgie+HHr/4xERFdvMhlTu26Ko1Dz8liyPvrj//lvyiP/T////9HRET/5P/+B0RE9MM/e1UGAQRqAiGFhTW5b7yIo5e0J/m90QA8DDz73N6WjAF9Y/MjzjGHly6Wx+IM55nw+j9SAimmH+jZdbxfFfZnfji3PU5ERAVRtSaIyP/+ezwnv/MPvk9ERNWO3F8joFomMtaymQeHjOZ98AHzRbxP6BX7uAgCEdH+Ps/lzg7PUV2VRnUqx8vmeBC8Hw+AxA0UApOhxDEx4gxK9KPIDLrCf01ZFhFRAmSpwHXluc6PFmUf6c8yG4Fas2bNmjVrpzD7A2rNmjVr1qydwk4E4bqeT/XmApGCLIsjDsczhMeup6AGo5hShtWZOgS9WySZdZY8nnLIXW8wtLq2LBBSMESTYPx/b6xKXKCu4nlyWRlC8ynUgSZDgRJaS0wiigF7FQr7rVQYxvm7/9s/JCKipWWBkX74g39JRER7O0yn9hU8V+QZzauuwnUdiioB1WoC0xqtyCLnGQgjpU3Z4HnbOWKCxMHgfHmsBTjXQ7lCoBrjxui+0kI5ynsPBZL909d+QkTSkePXvvKN8tjXv/ISERH9wR+9V7729i0mk8Uggni5zKkDmC7LTbmR6qQBwllrlen8K2dFiejRTSZkeFBhurgm5Iv3Ptx9Yrjls6yggpKioFQ3RMf+GaPTxFipoNSJrzFAZxBdGlGg9GOc8PurqpF7hlqYCRSMlCgO5WM+trfN8NZApTamKTpbKHQrzXg/T4Z8vkjpQCc+jytF+URNnWh/DLIW7smqIkAZDdkQ93LQFjLaTk9KquZho/GQfvLWazTNBBpvuDxXsx6v+8Odj8tjW0hd3NpBeVQs6/HbX+GuPWuVJt4jKkC9lPd2BfDwaKbYhyApzSacKhg2ZJ5mHkqVZgKlNxcZZjx3nslEezeE1DIClG74Kq7Sgr68yc+QoyOG1K+/+YF8DvvL+wqrKC2vS4lX6kYlHDkvu3BBng1/++/+LhERLSxy/U+uOqHUVvj5Yu4xVx07e47v0e9+97eIiOiHP/zP5bFXX2WIfIiyFAPbEom28tmzfI25SmEM8b4FpX6VH/Ea5+j0lahnivku8zfUClMRf0eGNJ7raHgXutr4O1JdxNKEu2o9idkI1Jo1a9asWTuFnSgCzdOURoc9Gj8UNfzFjEktVXirOo9sAgNDwtblAabLRGZo+56OVNCDD96/r76000CpChL+ExV8mASy7t4SIUoYH7Lnev2Dn5bHzl75NSIimhqtRdJeHn+/aczxzW9/qzyytMhkkR/8AUeiH77/l+WxIssoy+YUERUFpUlKgSte1cV1lOHMeN4d1bWkEvG1HhwwOefO1k557MLmRSIiymcc9XhqMTJ4yW2UNizUZL53QcgYQgPze98RAtk2tINf/0hKCc5/mUsBJtCo7G+LVqkpM/JRthE1RLd18zKP78wXuJQozwQp+OjHXOD+5cvswY9H4km6c+RW5AXRLM1ISc2Wc5+RiSzk3GYOz0F/diNS5R64RxahvVxRHu0IvVh9CHFQKlHK4Agar+h7O01l7d2QI6uZKgnY3eO19lDMX+1Ir9QEXY1CICQLilQ2zXg9+/DuHUXwq0YgPiHUDRSBZSE6cQvhT7XReEh//dZfEEVyzzRBcFpe4BKQZCokk+b5F3kcQFACtQEGTSa/VKs8Bysbz5fHrg15bx+hm0gvU2VFPj+LNlr8169KdNquAQ3zJfL+4gU+T3/E90mtIQjRwR4TXYqM5ylU5RjZPt8n55f5+eH05ZoHeFgO9jiyTqeqk5ITUjITJOJpzHMdalQi+tKXRGTi3fc5wt/6c+7rWW8KqnXmDN9zRvd7OpVxxAizv//97//CeYyggekHeu6ciFMkuvaQiJbXpNRntMwR6IOffSjnAepj+nrqkqeFBZ7LFoQaFhQByqB0UxDfMlV+abrDzEY8lpu3pKwpDGo0U31FP81sBGrNmjVr1qydwuwPqDVr1qxZs3YKOxEeM5vO6PbH1+lMKIl7g7ymSMZGga4DBUSRofWWUkmJoS1pCBtGfUi/r1ZnaGwylnB6gpqgBuqkLi/K5xJI42wfClzWBKy53ua/W3eFkFCAuFDHeVLVj8r3+fuNrq5GCl8ATFlr/r9ERPSDfycki9df/RMq5uaXOOSSd0xBo73AcNEEifHuQCBSo0rUc/l6fnJb2rO98CwTLBbRnogySepXoJozRSL93BkhMHzzf/1tPh+OXTkn7bV+9u7bRES0dPVy+drZl5hYsfuA23cNt0QLN4OaSYYWbIsLArecu8jQ7QykkNv//fXymI8a0fObXA/72l8JaaPeqJDbfTK45bOsKAqaJClNVYPuPWiB+lBP8nzZby4IO+fPMXmo6gkUa7h0FbRD082dM8B6voEhVbPhoyOGCvcO0Qx4QdZiv8/7896W1HNmqEtdAywYqHrKyYBhNx9EoZqCX9vY8yhzpb6q76wCtjY6ploxqojnSyIKgoA2NtepO5b9mHs83gMUeR/Fcv4Q93M5d5GkAfq4/7fQqLz7U2QZAAAcU0lEQVS5LHW5FTyXprimteeFDPfWLSYbuUcMh7erov604DGcWZnJPf3wQyY3JXhGNGpCclxo87+zMa+R1pze7vFra5vQ4NYF68gVrQHOjNWxxc5S2cbtaa1er9M3v/YKvfLK18vXrt9kgtQRWvVNVa2neTYm0H3u92VuCqR+THuxBw8EBm1A8cxAq7qhtmlWXTbB1oQdh++vmtLxNo3HTavDVOlxdwDZGp3bjz+W5/tah2H6lSV+z+G+EMFy5GnqHSgYKWi6Xmsd0839NLMRqDVr1qxZs3YKO1k3ljCk1TPnaK0qXmixi0bNKD0JlLfqIm7z4P0VqmluAfaP6X/a7ogXlyA6jeocqURNIT/QAXvVA3RjqagSho0mIhylxtKomEbD7BEdPBL1m94+l1ysnbmCsSidV3yH0ZB1FOnGKFucO8dlMN/7nb9fHts/fEB/+e4dmoe5jku1apV6Y/H61gr2Xrsxz9fhWMgGCx2ODtuLTKLYeihe/atvc0T3u9/5X4iIqBII2cTLmPDj++xtBirK6rT4tRDKJUc9abY7ztH1ZUM8/RSU9K9eYZr89opqTozIxkTok6kQckZI2scPuDRooMhHHtCG69fvEBHR7YeH5bHFpZDmVTaUFwXFSUqxUjMx3VCCAmUQoWomD6/aR3eNxjkpvamho9AECMZIdQ1KGzwnFahwjUayvvdBduuh2bKfyFq8BrLHriKHrbb5u/Kco5vpSO5N0xfdkIhCVboUQQO0DuRnOBGv3mhXR4io2xUhMnnFcQLI01oURnT53FV6/R1Zb98QC/H8cNX6FlibFPe9LmmIESWOXahHqSCiAjLhMOfr/eL5i+Wx/Yyj0p0tvg/yWwflsXqF9/HZ54U89/Z1vufiCpAEhTzEIz53HR1dbuxLBJonuGffhYKRJ9FzpcLPp3c+vkNEREFNUK0Nr01JPh8N4mazRd/5ze9SrSnEp8mUyU2mnMQLhUQ0ifkeH414bkZjuZ7FDghyUKrK1RjX1vieKBugK3U2o5N7G4ppOnJ9eJfHUlOEnwq6N82ASBTqOX0bJW7m3KHqeuRCmWpri59Z124K2dHHuK5e5Wd/pEiszVpYltV9ltkI1Jo1a9asWTuFnSgCrdSq9MLLXyZ/JFFcil6Vew84KtClKllxPH+oS1U85A9SaEA6On9E5jX0OFQR6BTK/bMM5RWql5wDL6QRiPdQNToN6NDSOxDv/fb1t/j7EUGo9BGNkQcz/UZNzoiIyEPuK0ck6qm876Wrz1BYkSjlqcx1qIgqNO5LJDmB9xXV0Z9yIhGoybu1auxRTxfEy/z4kL9j/f03iYjoO6/8enksMZT5UoBCtsVV9Eq9f8Dv2TuU/qMuesEOehIRPoMegVV4cCkpXVGf5yVH4fpIRXrjQ/YSO132zgdKzGCK3oC3b3KuSvvi3f6Ismw+3rnvutRpVMivivfrIIdZKf/K3mohKm80OAIvdPcb5NyNduwsUWoJLq/dUcxoSl+V5fQqnK+JHX7/T2+IGMDr1/m+0znxkctjGGOfThX6soy1qPkogleCH1EIngGi4CCU68oQWYfo0tGqyr3pZPPJNxsL/YjOrVygd10RFciRw69gHBVPo1p4phhRDpVbNt1mTHekPFHF8UBx+tC9vq4CjDMQFVj/m8wTWFCtl0KX59XJZF5nI163Ecp8YoWkHO1xpBYg3znal3sjRCebCfrHKs0LqjXRZQr3YqUp/IAkbFDhzCcHGlUqdPnqF6mzKkIK77zNcz+tQHRGiVPs7CIvilx7qvbXxfPMfdhETvfb3/62fA46t40GP4vqdXkO3LvH9/jrr3PZzNZ9iQyNyMRERaDNnO8rU/4SKbEEU8ZiymbqKh+9tMSowaNdzn2mx4AqoKJATPNY9RF10ifuvmojUGvWrFmzZu0UZn9ArVmzZs2atVPYydqZuR61W02KXWlKHa0yccIotEy7qgEwqNgZ4Djd7NcHJJalpiWYKiEBnJQANogaorbjoB1X2ufPDScCNxygzVCmzoN+3zSegfbel/ff+YjLMJ57+W/zexUlPgC+crDDkKWviAxNQMoTnPvwQEgHWZbNi9NCrutSvVajaU2WaTpl+CiCWstYJfV7gJbqIYg/gcAgMbR937jPkGC9Jvq1l5aZfOSi4fmRatk1gv7sItRJ4n1p3RT3+Lqfqan1gSbr21Ai6g1kfA4wlKDJEKY/FtjE375DRESzLtqmTZX6DS7DAXxXUdDXYJrRnKRwyfc9Wl9eLMk3REQBWtbRjGG3VJFtllBicqaKkpOelNfk0ERtRiibaG+Wx9yQ12Jrl+fo5kNJieQew05ra3xfraqqkRgXqmHaXcxvDaSImiIrhaaxPGBaV2lROyBDVXEf1hUsNkE5l4FJo0j20WjOLrfnOtSuBHTlvJRC3evxvFSx0J5qBF7gGhwQErX0NhmCE8adK2GxYoKWfSAdDdTH7hR87YfQ1a5uqrKUDq9VW7X/ooDvvXTC40tV+7usznt6cHCHiOR+JSIKMeYW2kHmipBliEhVkHu8SCDPxI2InhhU/HQbDof02muvUdS8IWPGnmlU+Zyxmjij+lPFftEqQqur/NyYIcVSUW3vqtXqsfe/88475bEf/ehHRCQwr6MIO80mz61um2a+d3mZ00n6mWfKTSK0RCu11UmITN0er4GeQVPGkgKaD1UZZe5mVDhP9lCxEag1a9asWbN2CjtxZrrIC/JCIUvkNdZKrK2i+W4slOS0zxGKSdQeIxEhijBRiY5AXd94OyDyqCJXo487gXDDYSyXcAh93XpVXmui0DYACej9uxIhb/8Ja9he/rWfExHRF198pTw2gedkind3duS6BkP+rhEavO7uCdEjiZO5NdR2HZeqQZWWFiXid0HHN1PZqYpnPBpwAn4BnWPqdSlV2R1wZNdaYzGCD3eETDUC9X5wyK+9f0siqS987TkiIpqgIDweiO/ehP7qS2tSvnFzyu/7CvSCfzIUktN9EL4WQTVfOJSGwhWQB7po5l1ThdeLbewRELuasRZAdmg+KqFEruNQ6PrUqIivuowIZBFhqZeJl31mldclClBaMRaBAz9ASRCIOJ4vXnOENYtxPeOxXI9T47mZQlDBc+SeOX+W1+7uAyklGuA79oGsVOuK8IP7zvHYu3Z1KYQhTWANG4r+bzRwTUQXhLIWrj9nn7vIyM16tNSUCKCboXMQSju0vrYHSClD5JDqxstYtgw3hyYmmrnwE1z3VEL7cIiSIXx3pZAopjnmsSy3ZN0NSbECxGZyJM+uQ5CI1mN+9o2rSjsW2qyp6cYzUU3MpzzmPvb20ViQDidqUvKE2qyfZVmWUb/fp2vviZ75dABSExCeUEWSJjI2XaAyRfw7POD9/sYbbxCRRKJEIq5wdMTv6Xbl3jBmos1A6TebcsGtLUG6ajVeg0uXuAvT7q4IItxF2UuAyLNakX20ssL3542bLCijn8o5utvs7HL51MaGCDd0+4Nj1/lpZiNQa9asWbNm7RR2wgi0oIIKqtYEn899Lhh2KozdVwP5Je/f5sguRQ8/XRDtmo4cdfYcXO0Bo2uHyZ3GY/EWJ0P23o767CEcjpRXbbB0XUuDfM5ul9//3m3JSQzf43FVV/85ERH9038m/Rz78MrGA46gNM15a5uPmTKWXl+o6nE6vwi0KIimSUG+KgUxBb8t5IUnbSlVSeFVG+/cDVTpTcbHJsgtTFWP0dvw6BLkFvKKnM+vc55jiPKVvb5ElJugjHuueJ5n11mKDE0OaGlDoucC3UGWUb5xpibRg7/K31WroUyhLVtzeRn9Ha+zt5kl4h0GRUGH83IDi4KcJKHAES82wL5sQJKs4smxzavc0cJv8wDefV1QivqEX9tE94/1qnjZhc/z+7AL+bGHEsHcO2AhiZ/f4LxooyPrdOY8C3fUFuQeOzhkzz5MEKHI8Kg34yhgirx0SwkiFOgEEwW8V2vqmA+5vGqE3Kkn+7mu+AjzMM8tqFEr6JIS4yjQFSlAXitRHXBMDjSHjJwfqy5O6MFq5EHdTHEuEMmHyCP7Q7lnq/uMhNRROtGO5FkURfwdM5UXNynPmcP7OE+URGmXxxCje0k3kXvjECGyKeNKUtWpxfBFIB2YOjKGzBlSls5HwCIIAtrc3KSgLntofMSRsIuSoFjlOU35isllzlRu0kSZZXlJJGhYgJz6eezZF154QY2B57mHUp87tyQfWwn5OWCiUyIRXuj3B8f+EkmO1sH65urZK7n7Qv3XGBDMLn/XQHE18qIouTmfZTYCtWbNmjVr1k5h9gfUmjVr1qxZO4WdDMJ1HPJ8nwqlVB+B+lypMfSg4aWFFYb/ulsMkYz3JDGcTgCheBwq+0olxZQrJAaKVbRl04jbkEw6DRnL0irDXRVFyY/HDDPsHfDfL5wReNLxGI66/cFfERHRT//i38vYV1kjcQQoN6zIVJnylSkIMxNFckrTguaE4JLrONTwfYrVF+aAhIzKk1dViiUgYiWAa1uhXGsAKHC7z8cWFEzbaGO+oIW73nqmPHb/kN9/qcWQindJFEw8QDZOW+C3e11oZxL0MV3VcSVlCHZ5zHBwpGj8BeD3KkoulisCw1+5yOUiUxCh8lh1knE9un34ZHDLZ5lDDkWeVxIMiIiG2Ht7UGlaWpRylPraF4iIyK/zWLv52+WxNOPrdnsMOS5siK86HvAc3b/P98BPPhASWhcKRGmV1+vaPaWc9YiJKRvrkmpYXmbIvOEzRNxU1RYG0oyhY5opFSVT/zFFaqKumDoVEEcKkF1SNR+NlkBr87C8KGg0i0tiGRFRNGVix3L3DhEROYnqZINng+ng4SZCrnHw7yngzlhpVuWAbh2gwbo7iI+SNdP82RCriIhGUBnb8wRGzvEsmKFcLPElVbIPVaIB3p45CjY35/OR+tCwPg4WgG4DVYKW5xnRE3YH+SwrioKS2ZT2FRGniut4tIv7ypFrreIer+E5v6r2nh/wdRgI1+xFIikvCX3TBFvGn6KEZG+XnwONuuyp559/loiI3nhdujGZ56t57s708xZrXgAWPzyQ34rrIA+Nxr+cZugghkyS4tirT1o2ZCNQa9asWbNm7RR2sgi0IMry4pgzlMNzNQGkrzyNBiKTOvpUphefLY+Nj9j7GB2wJzQcSnlJPkEPxoQ9DsVGLwUYAnjTppclEZGPAuVORzon5EjUuz578usrQoLwQZy4do09r/f/y78pj1155bv8uQUmSbmKRW6ufwjPPlaF7VnGRKt5mOc61KyG1FOF5HtdjghrTb6OjirZufeIBzl00CVFRaAOovJ6xO95sC+lELUmyh0ynu+hovibTigB+vO5bfEyK+j5V6vJfMcH/L0mgZ+q7i2tPfYIw5THlysFBEOsMNqm6UTIClv3mZxz+QJHf7Gi+PcHE/K8OXnnjkOzsEKB2uC5iUaw7/pDmZu33mBd4UuXVzF28dwbDUQQCC0OezLmR3tMMLn5iO+BqSOR0ubZc/gYj+HRgZC2hiA23FTdKx495H9fWIB4htrfZa9EaDMXipgS4Z40hLtMiSwYTdkZyl6GuSLjPBm7/4ltNJnSW+9/SN5IbrAzOd9Xi1hnfyL7P0fRf4EocyuQAe1GKI6HYEGuyuZylDnkeF64KuI2hJUC5MVWrkQlEH0PctmPWcjz6CYgsKjbPQFpjFycR6FHBRCBAoQXRyFlJgqOURpTqPH5vk+OO59Yp8gzikcDurgpfX0Lh8e6f8iIyOGB9BEuEOlPZ/yezfPnymMNaD+bkpNvfetb5TFTelLEQOkmsr5TkAC3H/LnnnlGEK8zm4w2NVvyTJns8jOk3eLvPLsmkfsA/UlNJy1D/iQierDF99d+T4n7yJXxf4tftqGtkII1a9asWbP2udnJhRQKOiadlptibeOkKu/dM7/P6AYR+BItVmrsySyucdH/ROU599GzM9v+kIiIUlUqn8ErnqH7/FB1iq+j+4SjzhOBAn32DHtATVUkb67j3AZk8ZRIwL13f0xERAuXXyIionDxjFw0+lrGEAaYTGXsWZbPrYwlL9CZXhXTx5jfCaLf5ZZEmZ0Wejv2eTyOilzHoPbXkGvq9qWc547DXrMDUYov1uV8y6uMIny4x5Tzqirj6OC7eqonaQGvPyzYk3a64s36yLeZWMxV5UYmesvxmvbsDlDsvISI6tJFWYvrd+4ey688jRWOS4lXpUR/nWO6NrCHO51IQfhP3vgLIiK6d4/Hs70jknyPdnkNFhY5Yn/vhnzu2l3Oed58wMjHRN1QlSNGXzLkZLSEZAyYR0c1NVx7G2FQR0WS9RReP9bs4EjWfNTn6/IhX+kpmKeC/Ju5ladDXSqmhAvmYGma0W53QJmSj+wnHDE8gKjErCo50AT7JMMcTDwleQfZRAf5TS9SfUyRwzbCFI56TnXRN9L0K1bNnCjDnBfH8sf8HQ2P58JXEXqa85gjiMEUaj1yx/RA5vuyUMIWhldiBAH0+IqimJs86Gw2oxs3b1B7ScRPJlMexyYiu2/9+svlsUrAY320zfOckkSGpqPWubP8DHdVKN6HMEoyBkqnSmO29/hYr8t/o1CQye1tjkp9hR4sLvC4lpeAmqiSnu1tRhaHQIZabYlODyHeYEq9PtmebmJtBGrNmjVr1qydwuwPqDVr1qxZs3YKO1WXVo1Qlk2zATm4iv5bPKYMlJNufou/CNVrquHq0gZDYocDhrqyqXwuNw24cwYCPQUpGoJRrmEvqJEMgc7q3sthyOeOQZOPCvEnBlDJGL/7P4iIaOHCc+WxyjonvQvANG4h15xk+dzgFocK8ikh1a+bmqB8Z9D0nAxkbqYpQ9eHE4bqaj0FXaBkxyApWi+4jya+S1AWqinoawIosagw1NdWnTkWAI/HinSRFUxEiu99TERE4YE04E7INHY2sL8sRor9Yzr2aKKa0THdQxnHc89dLY8NRyMKAtWy5KnMIY98mo0UJI8uDT7IRJ5kB2g05Pddu8f7dOdQUg0fP+TSrSzn0p3RVKDPwjWpBt4/YSQLnIFokqNp9lJD6lKaIHQVqrlzA3stAkTpq67B7nSG7+fXNuuyTkOQdmb4mykYkgDr+ihhyBVEPnlChZYnNsclJ6qSbmE8DHhf9PA3r0qZiIFPM7OXEqUaBZ1iD481L5PrreK1RmI6RMl1GAWdEDqq3Uz2k1FP89SzIYTckwOoe6bgySJDg3IQtjRMa56RpgG8Ic7xa8fJLKFqKO267ryasVCaprTf61Kgyth+9CPWBG+B+Bb9HSlVu3KJyYNTaC3vKU1bv8rza+DTXF1rBcS1bIZ7QqWhHjzg+8V0QhmodJKRWm7WZc2TGT+8r19nla79Pel+ddjlZ5f5TfJ2pMTNwPx6XPM2G4Fas2bNmjVrp7ATRaAFQWtQF/bjx73swalDB0MwKkOyxxLjpCJYFbaZ/LF5t/bUHJQsRNDn7ETyOaPhaSJRIiIH9PAMZINrO+JBmbEnQ/bCn1mS8bnwmMboTTjqvVkeq+9z8nvpCus7OhVVNjOL5xWAUlHklMYTCpSfU/PZS6xhnmNPIpSZ6W0X8Gv7QyFmtEDIcNEPcXokUdZixN7vKnQoP+wJKegSIvw6BEC7ankvYk51X8P0iEkA3gOOQDNFkMkcREn4LtfVkSu8xTIC1WvB/+4fMblkNJII4fKF8xSFItDxNJbnGY2GRxSq6CRPeMwGWfEUUSKEjqwDgpFXk31nvsKUo7TaqqQIfxOQIcZjOd8IJRshNmdN6bIGZScUuW0jfNQgK0mqSlVMUIN9UY3U7Y6xo/KDHKV3G7job4oI0PVlnfqFlOPMxRxGqvyq1tjl8y0C/XBjhVTg7opRzuKoufARhQcFItBc5sKU2aV4e6Ci2hiLNUXHoaIQJMEQYxzFnfJx7gSCCmZfE8mzJ0Fp22wmezUHqc+IOASqd+XjpkldruuSM6cQNElTerS3R5E6d57wde/u8n3/zlsy3/euc5nhzg7vve5IJmJlhcU+Xvk17mKlO6688/ZbRER0hF6cWoDBRO7muTE4ks81qyh/m8nz6cE9RnHuoQuRFvIpynnhdUqesIvKvMxGoNasWbNmzdop7EQRqEMcDRwLMh/7ezx59dg/joVmvzxOK70tRFu6bsbD9xtWeRM5NyLx8EIlC+bD21lEf8rdQylVmSL3ubLONOzMF+8qRscQF55g71DOs3/4HhERDY/YY1t+9ksyvtrCvNIVVJBHU7dJU9VTk+DtXtviAvpqTTz3lRaXo6yC9r3dFS+uC/mrRUQxL680ymND5ORuQqRhpL24gekzybmQopCI9+cHiM4fXCtfq27d5mGaju+qeN+YWU6dV8rI9B008IPMoilLML0J76EHIBHR1atX56Vyxrn6YkS1ihLbABIxHXHkVQlUBIroPECUkimv2fStdbCHQ10GgeJ0n9APNJf5TiD9ZlpwZkqkY4Z/5yrnnqZAZBzT11N1JwlxDHnrQOVAYyOFV6IB8rnU5b3i1Y5790REhRI1mJe5RU6pQrUS+PVHKFlxVA9Sw6tIsV9cJdfnFYy45OjeEhbyeDPdQQpEiHrLFIi+HXy3olWU79PrMMog8DDjfRKqsjlThpWia0muSrX8AGiGa0qj1Lx+Qr/NcgyOc+x7nsZSRKBLSmzm0gUWL0ixrwZDeQ5+9DMuIRsBBUhJUK1uj3ORAfbccCjPyPsQPzGiCaTKUjbP8fky9HI9UhFoA/yL2UTupSlKHOW35tjqfcJrx499nmYjUGvWrFmzZu0UZn9ArVmzZs2atVPYyUhERUFpkpCrUTkDRyCC1mx4U65gYBan0Hqb8p18UBEFoElZX2U69eFIQZgjDvcNOcWZCKkhPWC4wVkQvVYf5RqLCwyzvPSilECkIGo0UEKjFUj2t1lVpr/NcGGhmkZPRgz97t+9Q0RESSKQx/JzLx/r5vF0VhAVCY2USpMp2xmD+DBUXRXW6+iAgi4auvntEVQ/lgDvVjrSQeXjff7+TofVSWpToYn3ZzxHfpPPWygVkD6Ud9wb75WvOdDRLbDCmYJpc7PdsFl8hbA4IHwUhWmMK4sxBZnEIJB6vu8/2D6mRfw05jgFeU5GjupG4YK0VqCB8yxWa2sabwN2K2LVpQhErsJc11R1BsF4I6QaNLyLZh4UYw9NVVnKBOfJFVyV5DynPdw/3kzNRcUQwPg9Q0XGiY3CEr6rURdI3wGEnQI619t5dqrCt19uTlGQn2dlhw4i2Seme0ySKR3ako+IdfF0egclJ/5jLESSpuBVs1aFLi/h90/QtWOiqmYCwN9hTdUv4ZkVoUxGd4sxWr0uxu67Wg0JZXMG3lXlFQbWNQQjTaKbTqdU5POBI/O8oPFwRtdv3itfW1tlXVzzvNiF8he/hg35eDqOiPpjhnPfeJPv/2NjfCxrN1TPMBe/A2ZGjw7kfG2UjQ0GUtrSV9AwET2m9PY4QfXzh2212QjUmjVr1qxZO4U5J9FtdRxnj4jufuYbrV0oimLlab/EzvcTm53vX73ZOf/Vmp3vX6090Xyf6AfUmjVr1qxZs8ZmIVxr1qxZs2btFGZ/QK1Zs2bNmrVTmP0BtWbNmjVr1k5h9gfUmjVr1qxZO4XZH1Br1qxZs2btFGZ/QK1Zs2bNmrVTmP0BtWbNmjVr1k5h9gfUmjVr1qxZO4XZH1Br1qxZs2btFPY/AXSNoEPr8uujAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x216 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S0isHa3yYe0W",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Convert class vectors to binary class matrices.\n",
        "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
        "y_test = keras.utils.to_categorical(y_test, num_classes)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "17RSGXxrYe0d",
        "colab_type": "code",
        "colab": {},
        "outputId": "b616f6d3-7b08-4e40-9801-60fb449185ff"
      },
      "source": [
        "# Printing sample data\n",
        "print(y_train[:10])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
            " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
            " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
            " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5x5rXdydYe0k",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(32, (3, 3), padding='same',\n",
        "                 input_shape=x_train.shape[1:]))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(32, (3, 3)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), padding='same'))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(64, (3, 3)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(512,kernel_regularizer=l2(0.01)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(Dropout(0.5))\n",
        "model.add(Dense(num_classes))\n",
        "model.add(Activation('softmax'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ra-cnWLiYe0v",
        "colab_type": "code",
        "colab": {},
        "outputId": "b1348517-f2b0-4742-d503-6f73fed12646"
      },
      "source": [
        "# summary of the model\n",
        "print(model.summary())"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 32, 32, 32)        896       \n",
            "_________________________________________________________________\n",
            "activation_1 (Activation)    (None, 32, 32, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 32, 32, 32)        128       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 30, 30, 32)        9248      \n",
            "_________________________________________________________________\n",
            "activation_2 (Activation)    (None, 30, 30, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 30, 30, 32)        128       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 15, 15, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 15, 15, 64)        18496     \n",
            "_________________________________________________________________\n",
            "activation_3 (Activation)    (None, 15, 15, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_3 (Batch (None, 15, 15, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 13, 13, 64)        36928     \n",
            "_________________________________________________________________\n",
            "activation_4 (Activation)    (None, 13, 13, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 13, 13, 64)        256       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 6, 6, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 2304)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 512)               1180160   \n",
            "_________________________________________________________________\n",
            "activation_5 (Activation)    (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 10)                5130      \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 1,251,626\n",
            "Trainable params: 1,251,242\n",
            "Non-trainable params: 384\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s85wBZ6KYe02",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# compile\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "              optimizer='sgd',\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "\n",
        "# Normalizing the input image\n",
        "x_train /= 255\n",
        "x_test /= 255\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uJYR7vZ0Ye08",
        "colab_type": "code",
        "colab": {},
        "outputId": "fc54ac1a-fba0-416f-e75a-8787a07bcb4b"
      },
      "source": [
        "# Training the model\n",
        "model.fit(x_train, y_train,\n",
        "              batch_size=batch_size,\n",
        "              epochs=epochs,\n",
        "              validation_data=(x_test, y_test),\n",
        "              shuffle=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/50\n",
            "50000/50000 [==============================] - 22s 437us/step - loss: 7.7903 - acc: 0.4555 - val_loss: 5.7062 - val_acc: 0.5645\n",
            "Epoch 2/50\n",
            "50000/50000 [==============================] - 20s 407us/step - loss: 4.4773 - acc: 0.6101 - val_loss: 3.7771 - val_acc: 0.5378\n",
            "Epoch 3/50\n",
            "50000/50000 [==============================] - 20s 407us/step - loss: 2.7662 - acc: 0.6835 - val_loss: 2.1905 - val_acc: 0.7113\n",
            "Epoch 4/50\n",
            "50000/50000 [==============================] - 20s 407us/step - loss: 1.8445 - acc: 0.7240 - val_loss: 1.5798 - val_acc: 0.7272\n",
            "Epoch 5/50\n",
            "50000/50000 [==============================] - 20s 407us/step - loss: 1.3426 - acc: 0.7539 - val_loss: 1.2204 - val_acc: 0.7506\n",
            "Epoch 6/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 1.0684 - acc: 0.7753 - val_loss: 1.0566 - val_acc: 0.7554\n",
            "Epoch 7/50\n",
            "50000/50000 [==============================] - 20s 407us/step - loss: 0.9096 - acc: 0.7903 - val_loss: 0.9463 - val_acc: 0.7680\n",
            "Epoch 8/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.8126 - acc: 0.8074 - val_loss: 0.9085 - val_acc: 0.7666\n",
            "Epoch 9/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.7510 - acc: 0.8214 - val_loss: 0.9159 - val_acc: 0.7681\n",
            "Epoch 10/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.7097 - acc: 0.8357 - val_loss: 0.9204 - val_acc: 0.7638\n",
            "Epoch 11/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.6802 - acc: 0.8477 - val_loss: 0.8791 - val_acc: 0.7791\n",
            "Epoch 12/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.6589 - acc: 0.8563 - val_loss: 0.8903 - val_acc: 0.7807\n",
            "Epoch 13/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.6428 - acc: 0.8669 - val_loss: 0.9313 - val_acc: 0.7751\n",
            "Epoch 14/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.6262 - acc: 0.8762 - val_loss: 0.9463 - val_acc: 0.7787\n",
            "Epoch 15/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.6176 - acc: 0.8820 - val_loss: 0.9431 - val_acc: 0.7840\n",
            "Epoch 16/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.6100 - acc: 0.8896 - val_loss: 1.0060 - val_acc: 0.7700\n",
            "Epoch 17/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.6098 - acc: 0.8932 - val_loss: 0.9702 - val_acc: 0.7870\n",
            "Epoch 18/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5992 - acc: 0.9005 - val_loss: 0.9787 - val_acc: 0.7866\n",
            "Epoch 19/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5914 - acc: 0.9053 - val_loss: 1.0096 - val_acc: 0.7881\n",
            "Epoch 20/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5892 - acc: 0.9079 - val_loss: 1.0914 - val_acc: 0.7634\n",
            "Epoch 21/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5921 - acc: 0.9108 - val_loss: 1.0320 - val_acc: 0.7862\n",
            "Epoch 22/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5978 - acc: 0.9105 - val_loss: 1.1041 - val_acc: 0.7753\n",
            "Epoch 23/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5867 - acc: 0.9162 - val_loss: 1.0344 - val_acc: 0.7902\n",
            "Epoch 24/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5848 - acc: 0.9187 - val_loss: 1.0885 - val_acc: 0.7774\n",
            "Epoch 25/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5866 - acc: 0.9183 - val_loss: 1.1509 - val_acc: 0.7721\n",
            "Epoch 26/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5829 - acc: 0.9221 - val_loss: 1.1124 - val_acc: 0.7791\n",
            "Epoch 27/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5825 - acc: 0.9231 - val_loss: 1.0887 - val_acc: 0.7882\n",
            "Epoch 28/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5901 - acc: 0.9224 - val_loss: 1.1375 - val_acc: 0.7695\n",
            "Epoch 29/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5805 - acc: 0.9276 - val_loss: 1.2101 - val_acc: 0.7648\n",
            "Epoch 30/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5714 - acc: 0.9302 - val_loss: 1.1525 - val_acc: 0.7815\n",
            "Epoch 31/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5853 - acc: 0.9276 - val_loss: 1.2004 - val_acc: 0.7708\n",
            "Epoch 32/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5809 - acc: 0.9303 - val_loss: 1.1191 - val_acc: 0.7867\n",
            "Epoch 33/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5732 - acc: 0.9311 - val_loss: 1.1818 - val_acc: 0.7736\n",
            "Epoch 34/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5743 - acc: 0.9328 - val_loss: 1.3401 - val_acc: 0.7442\n",
            "Epoch 35/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5782 - acc: 0.9330 - val_loss: 1.1623 - val_acc: 0.7802\n",
            "Epoch 36/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5667 - acc: 0.9354 - val_loss: 1.2078 - val_acc: 0.7746\n",
            "Epoch 37/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5666 - acc: 0.9354 - val_loss: 1.1866 - val_acc: 0.7803\n",
            "Epoch 38/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5700 - acc: 0.9348 - val_loss: 1.2000 - val_acc: 0.7741\n",
            "Epoch 39/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5725 - acc: 0.9363 - val_loss: 1.2141 - val_acc: 0.7751\n",
            "Epoch 40/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5707 - acc: 0.9385 - val_loss: 1.2756 - val_acc: 0.7653\n",
            "Epoch 41/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5707 - acc: 0.9379 - val_loss: 1.2234 - val_acc: 0.7785\n",
            "Epoch 42/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5649 - acc: 0.9389 - val_loss: 1.2622 - val_acc: 0.7734\n",
            "Epoch 43/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5745 - acc: 0.9376 - val_loss: 1.2728 - val_acc: 0.7761\n",
            "Epoch 44/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5620 - acc: 0.9416 - val_loss: 1.2613 - val_acc: 0.7847\n",
            "Epoch 45/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5538 - acc: 0.9431 - val_loss: 1.2529 - val_acc: 0.7782\n",
            "Epoch 46/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5601 - acc: 0.9406 - val_loss: 1.2933 - val_acc: 0.7755\n",
            "Epoch 47/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5498 - acc: 0.9445 - val_loss: 1.2436 - val_acc: 0.7798\n",
            "Epoch 48/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5545 - acc: 0.9426 - val_loss: 1.2816 - val_acc: 0.7777\n",
            "Epoch 49/50\n",
            "50000/50000 [==============================] - 20s 405us/step - loss: 0.5422 - acc: 0.9473 - val_loss: 1.2728 - val_acc: 0.7711\n",
            "Epoch 50/50\n",
            "50000/50000 [==============================] - 20s 406us/step - loss: 0.5460 - acc: 0.9444 - val_loss: 1.3485 - val_acc: 0.7691\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f029ff421d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pUcLjvkkYe1I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7sPzomqpYe1S",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}