CropCNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CropCNN.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMeEjE4+pXAzeZBJlGYj5NI",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mepix/CropWeedClassification/blob/main/CropCNN.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Gm60WrL_EXKB"
      },
      "source": [
        "# Crop & Weed Classification\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K7W-5R21WhP2"
      },
      "source": [
        "# Import Libaries\n",
        "import torch\n",
        "from torch import nn\n",
        "from torch.utils.data import DataLoader\n",
        "from torch.utils.data import Subset\n",
        "import torchvision\n",
        "from torchvision import datasets\n",
        "from torchvision.transforms import ToTensor, Lambda, Compose\n",
        "import torchvision.transforms as transforms\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from torch.optim import lr_scheduler\n",
        "import time\n",
        "import os\n",
        "import copy\n",
        "from torch.utils.data.sampler import WeightedRandomSampler\n",
        "import seaborn as sn\n",
        "import pandas as pd\n",
        "from sklearn.metrics import precision_recall_fscore_support\n",
        "from sklearn.metrics import confusion_matrix"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lhjfB7ZZXcm6",
        "outputId": "74266894-2576-402a-bb93-84acfca35663"
      },
      "source": [
        "# Connect to Google Drive\n",
        "from google.colab import drive\n",
        "\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YaV4JXgqC5s7"
      },
      "source": [
        "# Data Preparation\n",
        "\n",
        "The following section sets the network parameters and loads the data from the desired folders."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Qlq5nSNDXFMa",
        "outputId": "b1a1fc11-4c02-444a-cee4-c4a17ac0f3c2"
      },
      "source": [
        "# Set the Batch Size & Hyperparameters\n",
        "batch_size = 32\n",
        "epochs = 100\n",
        "eta = 0.01\n",
        "imgDim = 224\n",
        "step_size = 25\n",
        "gamma = 0.01\n",
        "model_type = \"ResNet\" # \"ResNet\", \"AlexNet\", \"Simple\"\n",
        "optimizer_type = \"Adam\" #\"Adam\", \"SGD\"\n",
        "scheduler_type = \"StepLR\"\n",
        "do_weighted_random_sample = True\n",
        "do_three_classes = False\n",
        "use_train_test_split = False # Should Use Weighted Random Sampling for Better Performance\n",
        "\n",
        "# Record Keeping\n",
        "do_save_model = True\n",
        "path_to_saved_model = '/content/gdrive/MyDrive/EE244/output/'\n",
        "\n",
        "# Use the function train_test_split to divide data that is in one folder\n",
        "if use_train_test_split:\n",
        "  do_three_classes = False\n",
        "  path_to_data = '/content/gdrive/MyDrive/EE244/data/Split-Classes'\n",
        "\n",
        "# use the data that is already presplit into folders manually\n",
        "if do_three_classes:\n",
        "  path_to_data_train = \"/content/gdrive/MyDrive/EE244/data/Split-Classes-X3/Data-Train\"\n",
        "  path_to_data_test = \"/content/gdrive/MyDrive/EE244/data/Split-Classes-X3/Data-Test\"\n",
        "  labels = [\"Crop\",\"Weed_Big\",\"Weed_Small\"]\n",
        "else: #do the full dataset\n",
        "  path_to_data_train = \"/content/gdrive/MyDrive/EE244/data/Data-Train\"\n",
        "  path_to_data_test = \"/content/gdrive/MyDrive/EE244/data/Data-Test\"\n",
        "  labels = [\"Crop\",\"Weed1\",\"Weed2\",\"Weed3\",\"Weed4\",\"Weed5\",\"Weed6\",\"Weed7\",\"Weed8\",\"Weed9\"]\n",
        "\n",
        "# Normalize to Zero Mean\n",
        "if model_type == \"Simple\":\n",
        "  transform = transforms.Compose(\n",
        "      [\n",
        "      transforms.Resize([224,224]),\n",
        "      transforms.ToTensor(),\n",
        "      transforms.Grayscale(),\n",
        "      transforms.Normalize((0.449), (0.226)) #MEAN & STDDEV for ResNet\n",
        "      ])\n",
        "else: \n",
        "  transform = transforms.Compose(\n",
        "      [\n",
        "      transforms.Resize([224,224]),\n",
        "      transforms.ToTensor(),\n",
        "      transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)) #MEAN & STDDEV for ResNet\n",
        "      ])\n",
        "\n",
        "# Check GPU or CPU\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "\n",
        "# Compute samples weight (each sample should get its own weight)\n",
        "def getSampleWeights(dataset):\n",
        "  target = torch.Tensor(dataset.targets).type(torch.LongTensor)\n",
        "  print(\"Target Labels:\",target)\n",
        "  class_sample_count = torch.tensor(\n",
        "      [(target == t).sum() for t in torch.unique(target, sorted=True)])\n",
        "  print(\"Sample Count:\",class_sample_count)\n",
        "  weight = 1. / class_sample_count.float()\n",
        "  print(\"Weights:\",weight)\n",
        "  samples_weight = torch.tensor([weight[t] for t in target])\n",
        "  print(\"Samples Weights:\",samples_weight)\n",
        "  sampler = WeightedRandomSampler(samples_weight, len(samples_weight))\n",
        "  return sampler\n",
        "\n",
        "\n",
        "if use_train_test_split:\n",
        "  # Load the Images from the Image Folder\n",
        "  dataset = datasets.ImageFolder(root=path_to_data, transform=transform)\n",
        "\n",
        "  # Split into Train and Test Set\n",
        "  pct_test = 0.25\n",
        "  train_idx, test_idx = train_test_split(list(range(len(dataset))), test_size=pct_test)\n",
        "  data = {}\n",
        "  data['train'] = Subset(dataset,train_idx)\n",
        "  data['test'] = Subset(dataset,test_idx)\n",
        "\n",
        "  # Build the Train and Test DataLoaders\n",
        "  dataloaders = {x:DataLoader(data[x],batch_size, num_workers=2,shuffle=True) for x in ['train','test']}\n",
        "\n",
        "else:\n",
        "  # Get Train & Test Set (With Weighted Random Sampling)\n",
        "  data = {}\n",
        "  data['train'] = datasets.ImageFolder(root=path_to_data_train, transform=transform)\n",
        "  data['test'] = datasets.ImageFolder(root=path_to_data_test, transform=transform)\n",
        "\n",
        "  if do_weighted_random_sample:\n",
        "    sampler = {}\n",
        "    sampler['train'] = getSampleWeights(data['train'])\n",
        "    sampler['test'] = None #getSampleWeights(data['test'])\n",
        "\n",
        "    # Build the Train and Test DataLoaders\n",
        "    dataloaders = {x:DataLoader(data[x],batch_size, num_workers=2,sampler=sampler[x]) for x in ['train','test']}\n",
        "  else:\n",
        "    # Build the Train and Test DataLoaders\n",
        "    dataloaders = {x:DataLoader(data[x],batch_size, num_workers=2,shuffle=True) for x in ['train','test']}\n",
        "\n",
        "\n",
        "\n",
        "# def multi2tri(data,verbose=False):\n",
        "#   \"\"\"Turns a MultiClass Dataset into a 3-Class Dataset\"\"\"\n",
        "    #https://discuss.pytorch.org/t/change-labels-in-data-loader/36823/13\n",
        "#   if verbose: print(\"Input Target Classes:\",np.unique(data.targets))\n",
        "#   data.targets = torch.tensor(data.targets).type(torch.LongTensor)\n",
        "#   data.targets[data.targets > 2] = 2\n",
        "#   if verbose: print(\"Output Target Classes:\",np.unique(data.targets))\n",
        "#   return data\n",
        "\n",
        "#   for k in range(10):\n",
        "#     if verbose: print(data.targets[data.targets == k])\n",
        "#     data.targets[data.targets == k] = 2\n",
        "#   # data.targets[data.targets > 2] = 2\n",
        "\n",
        "#   return data\n",
        "\n",
        "\n",
        "x,y = next(iter(dataloaders['train']))\n",
        "\n",
        "# Rebundle Datasets\n",
        "dataset_sizes = {\n",
        "    'train':len(data['train']),\n",
        "    'test':len(data['test'])}\n",
        "\n",
        "# Check Sizes\n",
        "print(x.shape, y.shape)\n",
        "# print(\"Original Dataset\",len(dataset))\n",
        "print(\"Training Dataset\",len(data['train']))\n",
        "print(\"Testing Dataset\",len(data['test']))\n",
        "# for i, (x, y) in enumerate(dataloaders['train']):\n",
        "#     print(\"batch index {}, 0/1: {}/{}\".format(\n",
        "#         i, (y == 0).sum(), (y == 1).sum()))\n"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Target Labels: tensor([0, 0, 0,  ..., 9, 9, 9])\n",
            "Sample Count: tensor([ 799, 1591,    8,  301,   28,  113,   61,   14,   46,    5])\n",
            "Weights: tensor([0.0013, 0.0006, 0.1250, 0.0033, 0.0357, 0.0088, 0.0164, 0.0714, 0.0217,\n",
            "        0.2000])\n",
            "Samples Weights: tensor([0.0013, 0.0013, 0.0013,  ..., 0.2000, 0.2000, 0.2000])\n",
            "torch.Size([32, 3, 224, 224]) torch.Size([32])\n",
            "Training Dataset 2966\n",
            "Testing Dataset 1032\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "id": "WO0uhVTDA0lO",
        "outputId": "6fb5d4dd-6df1-4d93-cd66-a1114a6bc8af"
      },
      "source": [
        "def plotHistogram(data,number_bins,title=None):\n",
        "  \"\"\"Plots the Histogram of the Given Dataset\"\"\"\n",
        "  # https://matplotlib.org/stable/gallery/pyplots/pyplot_text.html#sphx-glr-gallery-pyplots-pyplot-text-py\n",
        "\n",
        "  x = data.targets\n",
        "  n, bins, patches = plt.hist(x, number_bins,density=True, facecolor='b', alpha=0.75)\n",
        "\n",
        "  plt.xlabel('Target Label')\n",
        "  plt.ylabel('Probability')\n",
        "  plt.title(title)\n",
        "  plt.xlim(0, number_bins-1)\n",
        "  plt.ylim(0, 1)\n",
        "  plt.grid(True)\n",
        "  plt.show()\n",
        "\n",
        "plotHistogram(data[\"train\"],len(labels),\"Histogram of Training Data\")\n",
        "plotHistogram(data[\"test\"],len(labels),\"Histogram of Test Data\")"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JLAX3QUzfRWy"
      },
      "source": [
        "# Determine the MEAN and STD Manually and then re-run block\n",
        "# mean = 0.0\n",
        "# std = 0.0\n",
        "# nb_samples = 0.0\n",
        "# for data in dataloaders['train']: #assume same for test and train data \n",
        "#     batch_samples = data.size(0)\n",
        "#     data = data.view(batch_samples, data.size(1), -1)\n",
        "#     mean += data.mean(2).sum(0)\n",
        "#     std += data.std(2).sum(0)\n",
        "#     nb_samples += batch_samples\n",
        "\n",
        "# mean /= nb_samples\n",
        "# std /= nb_samples\n",
        "\n",
        "# print(mean)\n",
        "# print(std)\n",
        "#print(dataset.data.shape)\n",
        "# print(data['train'].mean(axis=(0,1,2))/255)\n",
        "# print(trainset.data.std(axis=(0,1,2))/255)\n",
        "#[0.49139968 0.48215841 0.44653091]\n",
        "#[0.24703223 0.24348513 0.26158784]\n",
        "\n",
        "# https://discuss.pytorch.org/t/computing-the-mean-and-std-of-dataset/34949/15\n",
        "# https://pytorch.org/docs/stable/generated/torch.nn.functional.normalize.html\n",
        "# https://discuss.pytorch.org/t/about-normalization-using-pre-trained-vgg16-networks/23560/6?u=ptrblck"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 231
        },
        "id": "6AtiMGfBdi8g",
        "outputId": "84bbc103-da42-4919-e2b1-38c68f82752a"
      },
      "source": [
        "def imshow(inp, title=None):\n",
        "    \"\"\"Imshow for Tensor.\"\"\"\n",
        "    inp = inp.numpy().transpose((1, 2, 0))\n",
        "    mean = np.array([0.485, 0.456, 0.406])\n",
        "    std = np.array([0.229, 0.224, 0.225])\n",
        "    inp = std * inp + mean # Add back the mean and std for visualization\n",
        "    inp = np.clip(inp, 0, 1)\n",
        "    plt.imshow(inp)\n",
        "    if title is not None:\n",
        "        plt.title(title)\n",
        "    plt.pause(0.001)  # pause a bit so that plots are updated\n",
        "\n",
        "\n",
        "# Get a batch of training data\n",
        "inputs, classes = next(iter(dataloaders['train']))\n",
        "# INPUTS: a vector of images [batch_size,num_channels,height,width]\n",
        "# CLASSES: an array containing the index of the appropriate class_names\n",
        "\n",
        "# Make a grid from batch\n",
        "out = torchvision.utils.make_grid(inputs)\n",
        "\n",
        "# Visualize a sample set of images to verify we imported everything OK\n",
        "imshow(out, title=[int(classes[x]) for x in range(batch_size)])"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAADWCAYAAADM8ssAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eZSty1Uf9tv1fae7732DRkYhIQZhphhYJGADAgVsME4cYjth2SZYj8gBYhMcY8cGHBYCQiDYi2WWwRhsJoVBxjYxzgoGAbYYDSbBGIxAhOERCUlIT0/DG+7tPl/Vzh97rDqn+57uvldK9L791n19zne+mqv2b+9du3YRM2OllVZaaaWVVnrXofLOrsBKK6200korrXR3aQX3lVZaaaWVVnoXoxXcV1pppZVWWuldjFZwX2mllVZaaaV3MVrBfaWVVlpppZXexWgF95VWWmmllVZ6F6M7gjsRMRE9QURf/Y6o0EorrbTSSiuttEtE9C+J6DYR/fSd3j1Uc/8IZv6bmvkHEdEPEtGbiOhRIvoRIvoDl6jch2gF30ZEv0lEf/ISaV9KRFsiejz9e/8D0/79Id0pET12YNrPGtI+qULPRx+Y/mEiupXSv+KQdJr2lTqYlvbVl0j7dUT0GiJ6OxH9LhF96SXSfriO7SNEdOlgCET0F3R8HyeiHyai975E2o8kop/SOfJaIvqyS6R9OhF9FxG9Uf+99JL1fjci+l4t+y1E9D2XSPsniOjfa5t/log+9BJprzPOzySif0REb9bx+h4ievDAtP8xEf0KEb1V0/9vRPScA9O+cFgXj+u6+NMHph/TViL6uwemvdY4p3y+Xev8gZdIc6W5TUTvTkTfR0Sv0/n1M0T0sZco91jr+3YiegMRfdEl0hIR/S86xm/Wz3Rg2od0bPJYvejAtF86pLtFRI2Inn1gelMuLf0/PCSdpv0qndvLNebHM0mw7o5gmtJcp68v5LvM/MkAPv+gijDzhf8AMIAPTN8/BsBLADwTwAbAVwH49Tvlo2lnAL8B4IsATAA+GcATAD7owPQvBfDdh7x7QF7fCeDbr5j2IQC/BYAOfP9hAH/kimW9EsBfuGLaPwDgPv38HAC/CuBPXSLtSwB8hkyTS5X7IgBvBPBhAI4AfDOAn7hE+lcB+GqdIx8A4PUA/rMD034HgH8M4CaA5+s4fc4lyv4pAF8P4Gk6vz/qwHQvAPB2AJ+g8/xLAPwmgPkdMM5/D8ArADyo9f4xAF9/YNr3APDe+vkYwNcB+OdXrMeLADxmc+6Sae8H8DiAT3xHjLPm8QkAfmLkcfdqbgN4f+V976Vz+3MBPALg/gPTf43Oz2cA+BAAbwDwxw5M+3kAXg3gfZQXvArA5x+Y9iEAP32VObEnr5cC+JeXeP/gsdmT9sUAPh3ADwJ46RXz+AcAfvIy7b9mX9+R7x46HtfuXAjIM4BnHZDXh+sCpvTsFQC+6hIT49rgDuA+ZUKfdMX0/wrAl1/i/YfxTgD3IZ/nAPgVAH/9kuk+8LxJdkGavw3gm9L399Y58gEHpn8SwIem7/8YwJccmPYRAP9R+v6lAH7qwLSfqmM1XaF/vwDA/5G+FwC3AHzKvR5nAP8CwF9M3/8SgB+5Qj7HEAB51RXr8R0AvuOKaV8M4LdxuMB85XHW92cA/xbAH7wTjxvSXWtu78nv7QA++sB3XwfgU9P3rwLw8gPT/iyAz03fXwLg5w5M+xDuArgDIB3jF18izZXBPeXx3bgCuAP4OAD/GsDnXKb91+nrlOZcvnvoeNwNh7pPBPAGZn7zFdMTBPQPpT9Bsh3wq0T0316xzD8N4E0QiexSRETvC2nzyy6Z9HvUvPMKIvqIS6b9GjXT/Myh5jAjIvpiInocwGshQs33XrLsqxLt+XzoOP8dAH+eiDYkWz5/GKKNXrXsQ8v9QxCJ+7vUnPYLRPRJ1yj3snP7quP8TQD+UyJ6BhE9AzK//8WhiYnoeUT0Vogw8tcg2vuliIjuA/BfAPiuy6ZVejGAl7Fyr0OLHT5fpq//CoCfZOZfvkSa88rFJcuWhEQfCdH+f/OAd58B0fj/XXr87yAWhEPow66RFgA+SufmbxDRlxHRfIm0Ri8E8O4A/ukl0/2kbkP8ABE9/wrlXpqIaALwjRCh/bLbktft67tD15GcIGaH3wPwZw+URjYQye2v6+dPBXCGA7UMAB8KkZQniFT1+kPLHvL5cVzdTPNlAF55yTQfD+AGxIT4JRBz2tMPTPuxAB6AaFUvhlgcLqUlQBjQRwH4CgAP3C0J8oI0fwSiWf1Bbfe3AGiXmCcfB2F4i86/r7hE2d8N4Ae0zz4QYq49PTDtt2p5L9H5+WcAvBXAsw9I+8GQLaYXQRj2l2mbD7U4XHmcdU38mJbXAPwogKMrzO1nAvgbAP7QFdJ+NoDfwYGa95D2fQFUAO/3Dhrn5+r8epp+v4zmfq25nfJ5EGJJO3R+PFfreZKe/VEADx+YvgL44PT9BZrfHccLsp3wfhBr1H8AMTMfVO8hn28D8J2XTPOJup6eDgHbf48Dt7qGufLSS6b5KwC+WT8/hMtp7lfu65Tmnae5E9G7QUzqf4+Zv++QNMy8BfCfA/hPIAD3VwF8P0SrPCT9q5j5dcxcmflnAXwDRFu4TL2fB2HAl9W8jf48LqmdMPPPMPMtZn6Smb8GAhgvPDDtzzPzY8x8yszfBeBnAPzxS5bPzPxvIZrZV1wm7VWImX8MwJdDJPSH9d9jOGCcieiZAH4YwFcCOIEwtU8jor94YPFfCGnn/w3Za/u+Q8pVugVhlt/GzFtmfjmA10CEswuJmX8dAsrfCBE6nw1hgofO7euM8/dDfFkegIDGb0EY2qWImR+FzO0fvIJmdhXN2+izIczqdy6R5jrj/HcAfCUzv+1StcT15rYREd0A8L9DTLVfc2Cyx/VvdpR8UMs+NP2Y9vFDxouZf5uZf4eZGzP/CmRtXpbv3gTwX+LyvPMnmfmMmd8K4C9DhIwPuUwelyV1kPxCAH/zillcua/vJl0J3NVE9AqI482ljsgx8y8z8ycx87OY+dMgUuG/uUo9oNLQJdN8NoCfYebfvmxhRPTxEC3pn1w27UBXqffdSDtDHNTuOTHzNzHzC5j5PSCMcIZI3Xei9wdQmfllzLww82sBvBwHAh0zP8rMn8XM78nMHwaZ44fOr1/Grgnu4AXJzP+EmT+cmZ8FAYDnA/iFQ9PvKffQcf5IAN/CzE8w8+MA/j4uKQAmmiGm04O87QGAiJ6Ld7zAfJ1x/hQAf0tNvW/QZ/+aiP7cgWVfdW6DiI4B/DOIMPB5B9YXzPwWiNCYt/Q+AuIkewj96jXS7lQHl+dBfxLAoxDfkuvQdfjfofQxkC2QV+n8+AYAH6PzZTog/d3s66vTAeaBzmQFWfT/BsA3nvP+i3CBGRdizjqBmKj/GsSUdzyU96Jz0n4GxFOUIAPwe0jOGRAp+qE7tOfVAP7rPc+/E3cwGUHMti/b8/whnGMeA/A8iOZ3pO3+HyD7/c+6U39BTFGfpulmAJ+FdLoAAh4M4Pl70hYI88j99XoAX3hIf2maE8hWCOvnPE7n9pe+++Gax/MgC/p/PrC/HoRYNv6ctuE9IU4tOf1Fc+QDADwLsnXz6RAT6oel31+Jc0x0ELP0WyBa6ATRTh6FmuUhDp2vvGB+fLSmezeINv29h6yL64yz/v6vAPxdiJn4BsR7/mcPbPOfgnjollTvX0y/X9hmfedLIfvXl+IF+s7HaVt3tovu4Ti/u84r+8cQf4sb93hubyAa+z/DHtPyAeP8tRDv/mdAtoFej+Qtf4f++nwAvwZxrH1vCNh8fvr9YZzPCz4dwHvo5w+GCDJffkhfp3deAbGWjM/PnV+QfeqP1DG+H2JxeTWAzZ36OvX3CcTP6H/Sz9Od+hqyNZbnx18G8PMA3vMd0NcX8t3U7rvvLQ9hfAxZkI+nf8/T300zPi+/vwVhoI9DnH5y3s+FeI/u9byHmN7erGl/HT1QHUFMVB98Qdl/GOczkh8H8N9ckPYEAjo73s+Q/dXvuWCC/rKW+2Yt5z9Mv5/bXxBm+wvarrcC+DkAfzT9/kKdKJs9aQvEvP2o9tdvQJgwHdJfafLnfw+n38/tLwhYWZvfAPHAntLv5/aX/v7J2u63afp/AODmgXPkMyFexU8C+CUAnzb8/lu5D/ekfyFkL/RxAP8ngBem374NwFdfkPantU8fhezF3pd+uyfjrL+/HwQ03qxl/zCAFxzSZgD/HUTAtrF6OYD3PbTN+s6vA3jJnucX8gJ951sA/K97nt/TcR7eHXncPZnbAD5Jy3oSPe984YHjfAzg27Vffh/AF12ivwjiKPmo/vs6HM4L/raW9wTEZ+orcx3v1NcQkFuwx6/hovkF4QOv1nLfCBGK8ry+Ex/5TuzysIcO6eshn4eQwPQe9/Xz99T54Yvqc269D2jYbQiTPfS42j/EsNAO/QfgvwLwNVdM+wkAvu+KaY8gktYdB/qc9K8A8CFXTHud/vofAXzeU6y/rjNH3gdJo71C+l86b0H/f3ic///a5nfKOL+T5/Z1xvmdxTvfmfPrKdXXmv5HVTj48Tu9a9LESiuttNJKK630LkLrxTEHEBH9MSJ6tYac/OJ3dn1WWmmllVZa6SJaNfc7kHpH/gbkTOlrIXujf5aZX/VOrdhKK6200kornUOr5n5n+hgAv8ly1vMM4nD0Ge/kOq200korrbTSuXSVEIJPNXoOJJCJ0Wsh0cSciOhzIZdAYJqmjz45PsLFdwDFj243uZMFhQ453HmnN6SMUkSmq611ldib+oATpQRCIYAK5ewiKaViWF1Amzl+7CmK5J15mtCaO5J0L15YrQu68ryfdvNjAHTQIffNPGGp9YIh3PcDXVCbq1P0PYEKqfQu/TxNE7bLgmjtnvKJ7jwXzyk3+lDadrUWEuZ5wrIsd3zv8ByDOP3/TnldVAKR9G+t9cBUl+iJi14dF9VOvfrvpRSAgdbaYdkC4D39cyE/48v1674GbuYZb3nbY48w87tdUNJKl6AV3O8CMfO3Qs7A4/77bvJ9N49RiFAoL4o+HDWTxgploFUWhmpMlYbpT3LQU7CTux/sPwc8LZBo4AMsS7Yx8MynPw0g4JFH3wJWfKfMnRngVF0iZWbDYmUwChFOpoKn39zg+P4TVGLUugDcMBcClQImAqMIuydGqxW3bi944okz3L69gCt35RioPue93xOPPfYYnrx9u6uL9W3JbTYGw7mHQqpgrW/zk0/5DQEicjSyt8nHyPqpQzFilAI8/3nvg999zetEWGIkcIzyCey4yVYiKSNlRvTujmjkWTIxJpL3Gkv9pD/kvdZYhCEARAWbB47x7Bsz5m3F6dmCd3vOc/Abv/0ab4SJLiJkkbSYyMeVm1Qj8tRasQhy5HWTckGEzaZIWu30WlvXFm8hW0cyGjMmFTiJCp77Pu+Fhx9+TQyFdKDPcwGbPXmmUSUCigs3MV5gBjdZe91Ptm5gfy1j7uQwmyM3bt7AzQfuw+te90YUINY6Sa+yyUg64DzMu6G6MUfZYJKt2dHmVI4J0lbm2C/Wjmc98+lYlgVvfevbZZxV0GMdUAIwFaT+ZHBj1MZoms9UCNNUPH9dolLnxuDWUK2RzCkvyosqxl3zbmmtftAHPh8/8bO/+Lv7O2ilq9AK7nem34OcazSyePrnEjNiAfS/QNi8zOrGQDMgafJ7WmMKOUoFsdLZmLOSlsU0PMtEBOIAMcvB6pnBvemilRqQH7Yk/U7GMTXNRIR5M6EcHeFsu+CJ7RbEDTcnwrQhwIBdmXOZCk5OZmc0Z6eLgyfvESEMLIn63xxDvcPHdg/MNBJ0IO1DlTTWxJd7C0PK0oeKgaU2tNocGmblawus2SLQOdMmA3UdCbd6EEohTJQZvQgXjYW5FmJQU1BlKCPWnlMJglmYLjDLeNGiAC2SW2OZeaWQIJExa2bU1tBKwVwIrUmWUwKTashIJHgJxsnxDDTGtjKYpF6jtscqJDABxUQZlg5qrUnbijxsaSxMoGCto/wZxqobGkYB9QIrALRehs4FhKiRftR3iftnVnGWAQUxtI8T4Fp+KugQ9pS7Ww0RDgwgDRdt3pMVT92Uj86xCiZA1QazcRpXIJLwmAQoUwIy72EVbonye1Zp6QC/ebwkhUN5Heu8gs3pJL+6GHDYdecrXYJWcL8z/QKAFxDR+0FA/c9AoqddSCbw+yIyCV44uyyoxgAaqIVmZ+8zqXbG5MwtrYQdCMzvdL85Z0xf/adYXT2jSEWlPBxqWigL1Bg0FfBUwDOw3TY8cVaBuoCngmMumGbR20EKkoVQSsHx8QaVRTvenmWAt78BqqKJiWZHSSFg5+qZtZMKLCa1QP8mJqRJxnY7mMO0Yx0mG1DvMtWgVJuZS0FNyMEkZZacvTFgmxIqLk2mDfkYq4VhBwxY5wTcLDyBwRWYWAUEEqCslbGcVjy+WXCChqbA74y8xMCXSdrX2s5kQOMGKhTAoPORVIhojcEEzJsJ7WxRLV4Ta+OZQ/DgJrAxTYRaW5zLBYngoYIB9+il/W4iJudJIvMELvPKOyqztLQauhWTAUoqiWFq+O/DzHIhQD5TB1KegIbhI+kzHsc1CSyeuGt3zBnKabzt9tXgWOa8QzynvskCUwLYlgSWrr76wdZCYcgc2/M+eRr9j6LsbL3gVP+S06zYftdpBfc7EDMvRPQFAH4EYh3/dma+ME6wKkdoAArzzsQ1BsitQQzWCMByxp/WjoGO5Z9YAaUEGSDI1BxEnsZw5CcBJcG/nllm0LI6qTUt8kyCROGC1grqtqEujNOzBj6tYGrYLoyTow3meRJzrQow0wRMIBwfzVhaQ60N1fbhU5W9nSQarWm41izXqhOzsg8M64fUFZT6xJmqiy2en12vhi7/YKDeRxCcLCRXQVmHGeAAOh/6WptCFtoQk2rbAFMAhoxpAjzoOBDDNLpSSghbRCAqaK2iLg1P3rqNNhEmN5lC557k3bQ/ioI8MzCRCCoGylJP9jRuQdFxbI3x5K2t96OZ7FGkj7iqiZfkt1IKltZcazczPvvUMs2/t+MYtOfxdHBnHyJ5XKSdNLgMZ3AZf9jR2ve9pP1smw1hER9ECI40ttXhdetllo4ozGppTic0H9Yypcc2V7q25r+Wtz9jm3p9HhhBPM05BorvaaQeS1YmwPxkSBC8QBdH32JvyigcrXRXaAX3A4iZfwjAD10uDVQbSJq3/WaMLJnJi5kkFRwaRNutLVYj86AlMDlIZFCjEYSzGV3fMUCKhZ/SQ5iigQ8zUAdmZ+8XEs1zAqFs2cvm2nAGoN5q2G4ZJ8cTbmxmMQODREvUOhzPE5Z5BtUFtRmIJWZBsndPmjb3ARVO/T1ifM81dhiIMVBjXN1bolebQDMKHAKiKhqwbV0YQPWgROn/BtClhPLc7c1aI8yKYQJZBj8WkDNgL77/reZsNqGLwWcNtyfCVGQMnZPqhjp5niwmZDWJ295uKUhatCBAKfBymqIlNxZLwkQ+Z8QHoGl/Wb0J3JpsGdDYS1KuaXlwU/CeMfNRSoJZetpYsWjY8ufdWWBSnX/MWnyfb1TBth2ytW2/+qkVzpJe6/Gb8qtAmKjVYhKC6dB4wHdUqG8kOpk2r11CCEKNwSXkgaJtIEC3IigJTYwqWoOX533TOpmiq6EYzAjm7xAtiFWxfwtzpevSCu73gHxfqxGYwoyqv8YeFFgZhAJXmunCuHM4QePy5As6ykv8Zbc2zkHYVraU7NqHY6C/lxZoJPeFnQsiMCYCqBR3PAvJHqiVwW0LbuI4tzmaMRfChOLa+IYmHB8ruJ3V8OKnzCiyRpOYnJu0udOIBMe61B3gDyyyY36SLQGNQczufMUsGk0GdtNsJ0XCDA5mhg3AN4YuJnAusvdsz4hkr7h5PQhozbcHcttKsmQYyJA6xTWWfW+pikhmSzONXUfXrAMKGKZoMQNTsgQAutcPm4/iAMoQMEdj35fnBlRmlEn8O5ptZaQiva8NRBBzd/F1kd4fUNC3FfQPaxs9Genq0bwaMhgpyOxOAO9X2xMOoSGBPszHwLY3yPfGs79LruJOCQMIjuupezCsc+x512Zhzov3JMpiAWk7TabJXWzWHaiQRQQH71wKj2MQGcDFDR1gH0edg16JkgSTvf210nVoBfd7QA0i+YYySO5Q0hDAznlBi/ru7xMDhQmMphattP8O0bJsIWV+wEls7wWAzCHFkc+1D8Q/fx8DI0UKipC4GJE4W1EBKjEaGiYWsDeJgRk4q0C9veB423ByVFCONijzJI5VBBxt5nAO26ZjO8aEIEy1Y6BWGQK4eIUHJpMTxHfbOkkbtUGJaRIYpVmX2n4iecdZPUopMJNxU4FMsJkVDAGaijJLsQpMTJhKtGRhoDrQ6ngVqFde9EcptkdpICZChwmNhqna+/22hYJT0ezdv6A1oIinuxTD0RHMqJW1DFaBilSzFtO8jJO2XzxFY4IA4meh/ZHw3vvLTkDkMej+DtTND7aWWvv0pwa0YqOmoJUFw1RAn5Z93nVlISxxNr72T8WdXhDZqeveFsRXqx/1j7JMsjcNEsDuIROqfE5wWMqz0yYIqpD0grHkYfXirj15+6krIAk8pJUIi1xqkLG88wZ6pSvTCu73gHxRm6kT8MXuv+mqZQOLZHI2xi6Kkakrvgo6CTznu7s8uHvI6N8l42bUS/CGeZ32wmkBDkxlLoSjMuEWCZBNRChzScxbmUgDzpYmAg4qTgow04QJhA0BmCfwpoFrRW1slrwAoewRnFE+1ON+BBTAifPTPVw3Af+4b0kGhPYjDceBLE8CWM3j1BhLhWuNhQC0CmwroBq3mOXJjwMxgOp7oj24oIhlwMBPMCCDMLBUOPOkIvvmYNnWqSqcsAkLTYComId+ZXeQk/GS/rWtA/PRACO2AGy+kFkeFBu00lQQ/g76Qm3sfUcm4CLmG0G9961v8wmRXIjNiQ52MyUnsib5MIk1wfPz/wcSS9dzl19vMwnBILbdOOUVCQfo1uQmUfWP+xJiIeZaGrDukGWnWY87A97LuhZG8zkPWRLJOmtdzczZ0ZzrvANsEuSuCbmMWaxJpMoEQsQgjibtq8dK16cV3O8JJfjdXcsAx16na4FZTLe0SF7UzjSCmcRBNThAxy+7DM8TDpKzrE9OdQ0GxoiFaPu8UpbWgwowTUCZUZmxVG0PFW9jMU1A928XMG6dVTQ0HG8Ym0lM9BsitM2MVitOzwI4DLwZDLSQ9rtKctQ7WqzMJfUKc5jHLbn368AYDdjDiZzElFj6oWIwlipHyEiBuTX2/Whi0f4L2AG6NsJiglwykQKctDR5t2j/FAUo5hZNRnJ0bGKOLxxbB1MBUBuWpakPXsE8aa2tvKKZKRjLmXNGoaK+Fyz75C5D6d669xL7HLYm5Xli5+SlvxXQTbBodi7athfIpk+/BTJKrg1AEmgs7zz64ilPPn9D0EvrE/m7zX7WdRFA6z+b/EUxZr7WBm1Xe7cDfO9mtSBkecXHPQM57WbrTcgSBe/XfUcANe4R61zmXy+rZpFGgV23aoDMZ1Tg63pPHntMDi3Uz7RngbhfyHtqv9J1aAX3e0DZQZfzhyyE+6cSjka+eiIhF5IzzcpsCOHkYoDcedjuXST7lj2SdgDXyrtjRqZSQfaGzclKqifMvphkXghLq1i4wkyllndJTLuBsDTGWWtYThm1NpzME6ZZTMIzEY42M6oDlNUDul+c+tK4XGbYOzigQovtfzeAqTdRAhxCFFJeevTOd0sYrk069mr6WhsaN9eGUYDSTNMhPf8rggKrjAKyuRLe9aZRF1KNG3riAsagm0QbYxe10rxi2R8luBVBHCeLnsjQOjA7qJap+Hwlgp/icABQ0JimglplH51KgLSNrzBwedLUXB+7TD2Q2XE+n7v6KzNQdK+oTEg+Bcl6oqtA4gUxmp0yQJjlXegy0GomkNmZ80DrvfCSgdP/DotX16NbtbpJl75oO01rju2x+L0DvS4v6l7NS7OvcCpW6+6CAlvdVRTj9AjidJmrOzq3ESVry1CWC3JaYd9y1LwnLVBODenvgAts+cjcSnefVnC/B9ThTuYJgT/CYBKYi+SeVybHok/8QLc8YVppx4QGvhGUOZ4CBpFHO7N8+yNF3GVdiEBTnEflRigszzETGlUwV4DNAaxrLMSkF+a+BtHonjyr2C4NR/Ok0c0KJiIczQU0FWus5GdeuR2X8c7T4ij6m4yZ6msczNg66ly+YhnYnm1inPF71owUNJSBE6K/uFVxoCNxOmzK/9TSKabvhCyNFUSRz8Bnm02JtqQpwNYe46aIYCfF9vt1f7uBsTTGTBowx8AH0T0FgvaVw4zPqoI1LZtbQ23ic1EyCLhlSIUMkI950/34aSqYkmBgjoRUCDRPvn0xBi+ycWgSDMJRg70DAhQNyEqH/Fa3Ybg9k2FiuGTVj735OFD3ouZD5PXJsifGz+OzcZ6l8chPhqp0lI9/ytLT78mCIu9ZxrZWVQhNkq7xqk47Z05z1wT+buW5YGkN8mBNpPEqSnF/m70Rj1a6Nq3gfk+o11QM5APgKXiLz+ndFZ8XVvMceXiTunK6xB0FQ/NIc/oiKdLkULRxxCaYa/FMNQwqMeZSwDOhlorG2bObFSAgpnvbx1NeQlBAaWLSbo2xrQVH84R5luhoRxN5+zhrWfvUF8sTCSxHTua8m3VvMTuvUc4taT7kGY8CWu7rKZ0rtCE2lgaaMM0aftZDDcuLrWTIFsGgtYbG0KOROgZuopYqVWY92taDOwpEqIBFPwTID2VrsRr+00ynBpJ+7NKEBpbTArWGs13IowqsTLFNAdmvZQ7BBUlgNG3f9u0dx3SCE6vDXSGx5PhxTNoZe0rWgeZYpLPEBsDKMOEQ/TaWy4WebUyY+ET5Sxp5ddLj9HMGZV1DIxF4x6fP+IALkbnc4V2vzoDueU1bFELb8ZP8UyyB7M8BO6qoIZ3d/G6grPmn6hgvKgzdptLQtKkPKAllXJvzHUD4yJTAnMuK7feCVnC/F2TgAohntIKE44QxPgVVWZC2PLhbvRbFiYHESOT/iWcLU02A3An/WYpw/u3+vbSAtmIAACAASURBVF29XRhJi7sggEZek7JKkchsjYFl21CXppqdqNjWJt9y8E1bi80WVVqYUZeqx88mlLlgUnbgjtdZS3Iml/wAtI96RSAQ10CmpAzMxkBaX+v6XnkiRd48TvaLHF87KgVLbZIfMbgq+EIGtxSxdPi+L4lpvFbxWs7bAoV6LcjGtiSgtAQ2zwihKbXGKeytUgtgbU2iMeljH1M/4AAJT9vUxFAmiW4jYVGLAy63hrOFw9mP4BeUkFoHsuOZPZetJZt/ZNPF21I0bLHNxW5rxmpooN9t3YwIMXznNDGSCagTl33xpDPmedKRjbpJSoyALezm05W/p1YUfd4h6ZBm52t+HbbOYn6an0IB0Ig0KI0EFAozuAp/GlioJAEuz7/80S2HXUXI62AMyI5ohpRtQiyJL0fXot1AXytdn1Zwvwfkaz59N2l6F3Tj486CQsSYbg76CPxPe+7j9RxdGQOD8HJUGjA+J/hlkd0ScBjj1Ezt0gdhJiR7sWcLWjUuHeWLuZfTHm2AMQCwabxa3rYxeKnYQE21ehlKJBm0K9XSsjmYFECzABHcQ7cH0t5gt5+YQKapCkooKI12xs36cSLCzXnCGTNO3TtdLtwI7ZgDSMkAit2DmzMwKZcuUwExa9x6VjO1alYFIHUoYxWeJm1HcVOwaNHNBBcXfOx71og5/WEXfIpJHWSAarIb67E2vWSE1LdiKqBmDoXsjL+UIuFLTYAyEUn7xcBf2qgXleiZfLe0pyUAYnW6S6CiHvqevy+2HWltB/fz2grBMT+j+JwkEa+Pr494z77ubv3EAtmpxoByI+iNggClf/HQoVbmEYsABiI04o6XFNKTGKXEPvhQLzsWx9rMZuCtGVj3hO9G9s0hCX1svAyDmr7ut98zWsH9HtCuELrH+5SGl41zBiy6iGxmS4sgFvuCPTh3oKp592ZD8rqk5a+/JdC0co2HlFi5JoNLeFaWoCWVAa4Cwlmit3rAHMCQyk6uZGTaj5iLa2PQUsVJrTWPcpYyTB+VcXgX9ubIKaM2zLO7Bwlnjan6LQEkSoSd9W63fDWu6nxUwChop4wtNzQS0Hes9MtdpI4FFv/cmHAISzaoi8Zet3PztcromaY8FXVo5DBNWwhcw+Q4j61RCQu5ZYA1qA1z88iApvn5/NHvJpzYsXw7DcDMoKlgMxfMs2hk5pR3tq0SaVFPTvTjlawQqV/YhbUCUNP667vWJgVsKixbEHPBZN5a+juFsuhLKeYIspSBhEROhFQe0K0PMmHUZ1WkcpnRFiNHi/fRmD51UQifY6IhOFMvKSQrgjMDWW1cCIUlOiSbdA7R6s3ysgvseyqcj6Sy1kc7XW7I07GEzgUUPx2R5R8vYMX3e0IruN9tcuE5ADKEeQH2DA6kzIzLIM2ngM+ynoJjmdMU62JqzB5L3t7tz3anzyobyC1WSSJ3mSE58iHM4tmjVZz55L3KDaU2kDGMGsF7jAkbcBujyt7KDqCJzTEItUFjzVegVa0TZb7mjM2Zdsfq+jvmg/PDrQg7MbUpamB+Do5AZO0h1baDKTUGbmECbsyYNw2bx58U0Kx6pl/TF5gZWcFaa2VDEWDRiYKYChARChVU0zWdcYpCcghTewopSiT7+ZWzGwFYze+Ngc1s17XGSLfGqQvMq1892wuhTHOY6ROqiAOe5Cen8dSVUjNjH7MIYTsRxf2B3PnK+QjGBGD3L5gIoKaCh4OxSr0R7NA/7JqA9YEPVmCWCU3eNC8+9xKQg774TE7ITSnp+C1IQTI32OYlpTaMcE/xl4fH3mckc2HWy4WqvktJ2Ovbkxsb/CHJUC4QyO8hvkg4efL7GQgRn8tKyPLB3jatdG1awf0e0L49q1ivYbLqt9YiRZbmZVEpcLM6vbWUnyVQSdz3tkcp3oFNwc2PECUe6KjAXnb+QO58E04xcjEHuwDgR220cnlP2MGTOt4obWhJqic97gQ7BpYde3oAN8bhe3yK1ATz4PZRkT7UEJgVcAfCov1GBRpLXUCsxKD5B+sJP2ZFIuA8cetU3lhkL115qETdU/A1Rs0M1ZjhQoO3sHN2kt9Lsnn6HnYTYBZGSx5uVhi1oRF5iE+rf9X7WifrJ5Iz6xuNsBeg710Gi7xGXr5+pyS0qRTYdP7Vqu58e7YjuCsnWmtavu3vNwUBAxPvEOaYB3ZkrgGTA3u8ywbYGU/PwxFtmxgLbK3xUH5PvaCwTz8fn4wrXGdVlh6Y+nx9YexmaT26i/c7k1fWLTGmQn5bXrGJmviFrF3uU7u1UF6wkzrR37GNZzym6/S+ycEXzuvYla5NK7jfZQrQGRYc9x+pW9zd/zpLlTEW80sRjblnIsaUXXszpkwQZpU0O9PCRoZF3V/yi0HkPVb+IVsEhWWf7mgiOb42qa9uY7/8qXQ5p1WvdTJY8L6yfkm8RphPwTRNDjadl7++P+k+bUgQ8kaJjw40smeYvJyT8GEBhXw/0dQVlj6wXs81dyFsaai1RoYGrjZC6j5OKqQ1hvsSwDyLyVpl80A+Z01cmkjJqckkxQ7+jENbLwWuFTn3Pukxs7rInjkTJ61dtgKq3hZkJn4DdO9Tjp6Qaqim7wF8+ngC1oMiy3HX9+YXEC+rc2L0aJBZCWyu5DpGD/j4hvyQTrDsI07vm09BBagy8vn0bv7tzS1JETYduteoRzWfiIGWvT9EqvQ5sgJ1P3IspgHjQ+g1LhTlUFcGoUfr9FPKUwTL1McE3WMXniBTtMG2YWz9j7SC/N2nFdzvAZFJwQOD80WY98rl1cQPuFvM7vzFwaIzjgmDY09q+cpRJcnBmZUWY+fQe3Dn0PIUZFgPZedoegYupRDmuWDaFKBMaI3gcdPImKiZ3btSOhMmqRZggeccwPTvNBdMcxmY1FCPKQOj/UuwoLyUiYDCmIDOYRAQ0IsQwHYGrVnsXAHkvCnPIRCwN9IcB41hpgJK0XFzTuzBX4jND6P4MT7npwmcfNwL+TSpjF4o6Hwn1ClQhapChOPNhEXvYK8t+RYAYBJAq9mjOmnnMYpZYAOa3TBHcIuEjasBcNFtlqYsn1QgM493ORLZ0MgN/3nAU0+Sz5s8ft2wu/RB6mSX4TY5dA7rz4DL1gu15mMQl55rUePxrfx5AK+M3T065rS081je15Vk1rM0FM5L7JtjeyqfyLdWvGQSy5tjbaqCC2puVmouK3TVzQ/Yn4qwDcJERQQiUiGNbVsmlZjlGx4bv9J1aQX3e0R5sXThL/05AuCD30fitCnlzAbsJnF3CqIx09CSEv91sBBLY5LITWRIzNwx0fNmP2Zl+6tQhm0R0batafjV1AQgthKcI5CCmYGj/K9o5QJU5N9URMvs1j6b1q4R1ibyvWXvt8w4UncSgDIJbrum5sDcuva7oMLsDnF5kCx6lzu9QS/R0f2BlvIuEL+IppFaNnqW38yi5uhWl4rTrYCcCQgMAX7HmJqcHa2psYeCzDgNaOPOd73+VU3OhQjMVcIG1+h/ezfuImcff5M+3CuaZX6wau2Gr+ajIMlD2NKcQGpBAACeoGFzbZ5ebA4frMQx+9M8y5MxC1u+XPJkTUkd5KtGWeuQ0ZJGP/VE0f18fv1TS3Ltd9ayFD+Y+znWzc5Ex+5X2+JwZqOOr138ADW1haUrLH3eEEpHD3NwqFTjAuCoAPNkW31Aa4QFwQ+cO6S0K7bffVrB/V4RIZmayBlMLMd9JkJ9I/0Q+1dN98zIHbOCOaW9dRUCMmNxfqyCArcGuUYts1tLbsEwyGUMY7LGUGSREiqT3uRWcVYZTW4vkWA33JeLnFeSIyKQmToVIkX2YvgZ3Khg7q0CIr00NXOXpEl0OA0OJ7pkgrbgKewH5DkGSsGLuWXrpOdMRK4dMwMoBdNEQDNnNPbrfFm1YhvXpmgtQsqEWd8RQcPAtGCjHLdx81vjGIhjaloP6buWhDcTgAhHG/FYNngrBZgmiQq4nC24vV3QOMK8mnUnz1lihm3z2OCKOT3GwALaOK77JCS9KIdRStE4AHKri7uJEElc+0FC7Pj+iAJZiPGJkUBvQMXuK6XXM9ClfAuTHNdsaYyjc/ehu2a95+w7Yi3kVvX17p8xzi3Cfr3g91i7cjeA8oW8/oh97lCXiGMbzyyHOpfmSYemkZ+yAFTgLASaGTSrQNsANEJbKC6QsnEwhkDY21crXY9WcL+X5ECSvgDBiRT4yF/e4WspEwsoIww/g3snJJBeUWqg6kUqaLTMpC0DcymOU65pJw7cxX0NTsIKxEttaIuYsC04ijF20/5QyG+HcvDWPNz7fFBExEohIWm9peM7SN+p/9wZKAAkttLpkGZd8JuwOnRSIKbssxD9anvoEoVv0UA2xfuhGOiaudoc40jCvxpIU5ULdaxfSBtWG6OOjNcYOkHOdTe5Fpige+Wq6dm582meJIqYMtfGjKkUORt/PKNuF+9vlqmGHGbEh8ZOaGhejZsHPal6MqBoJq0QTuYZZ9utA5pHLGt2zp6xLNX7miD+ENJxCaMTII1Ew9/opCQZUB7v3fQ2B/Od7KQdKDfsyXefu7D15AOVB2eYj4M4yPmFvZXvsxh+22v9200e60LLs5gROWNzqpT57pJspLM1keSkqRA2Kn95LIOugsJbpD/jWQOJMGftZ4DS4Ozks9K1aQX3u0w7U3Rc+x0g8/hYklBKylAJV8E/m8bGtMYk7WKPzOPI9mnZI6UBSJp6J7tj1+EHSeSXP4UI3NRzvFa/oVNA3LYpWS8vIS+zJUGjUWigvYYm/+Sojqnj6T3jQcwpgA8F49O6dszYnueCCLoXn8q1sIDeweZZDDU9Sy08K07JWxOQniR8qtWhqbPhZNoNVDBQwPS8SDUka/MQwIeh0fyahh8g9XjW8fYodsq4LT67WWQlFCjJhT333cQDNzZ449ueBEARxMijJtpeb0SUQ5PTATESWjNC+DEQ4WnPfgbe/WSD/+f33oSz1jzYDCkjtyBCk52xhgmDu9TJxnvWU5eGdz7E9PXPuU/NuTM8CUwYpwKUicGzCnnqGCLjpOvQSndwRGAkD3M6jWGImeegs709tNnx0uruv3P3a9fCIk8WnUsSuCZ+Z0BPBUQbetmFdF4iHjIDhX099Nt5kgFBIuI1j6eQhCKwR2rc75S40nVpBfd7QHHQR+FmnLu9UqFp0Fv5XGsxTq1BXxzYnTX0Jau5WU8UOXMghscTj6yTipF5DPFOvgrHerkNEEuypGcMKIg3Vi2MzSFPUvjtoiln0Q6ok/SNilkp+uogXK7MfDxqt6n/0nMTPgyEHJk5wsOm4XNg92xVJa8N6Y5qi2UjZksiYFJHrm1lvcWNDYkxFfET8LYwo24btrWBNKSvBQKZSgSQMQGsaMS6BgYmjQmuzLVVFRpYPK4l+A0DrGfuqcB8C5bTMzzy5G1sK/tZZ/MdyIIGIMyZZEBUUJLy8u6QTd7NyRFuUMPr3vgozmoNU7yNlYK6DY9hQ2tNtzBspli+/V92YEivOCDtR8o8tzsKPEZY0ZTyFgMXND8MYSmySVwnjQG7zc49wsYO0b5X2MfAeYXJoHvWw5hBfkP6pqBRc9DtpCJO6xC2CAB0wou+3hi8Zd1WUmA3mYCKHsUloBKYmoJ78A/nSYhgN3fonZWuSCu43wPqeIyDpqkltnj2nIk1rmpMjJBWadEFFPmGF3t6pg9YbiEZgE60lKK/pd2ByCh4VF8P18osYA78nvNChKqgkfUlu3u8kDomFaBqo9zBijk+74gUJswkxNU6SB+bmVTPpV+kAGgb+u9Jf9rD9OGszZih1ZMh9/ACaBqEhiQmuhQjl+F0AGB92YAF7Dfy+X3nyiUlcEsTQcECjCxh/rQDRvMEgCYX9pzxamjZyeBqaVIXsrjv7ALLcnomsg3F0ToqpGmThg1oeFt5qbIINpP6Q8TRN6FWK9786GN+BW72vrdjizYcrTG2S1MBgHA0iW8ATeSBz/J82CGfA8PA7hn+8fm+vWpfLunkCDZqY6sSWtcW5vnTjdP/gaGCXY1AyVl033smKOgcYpUezgPFLCfEZ1lDttViglDn6DqsjS6eBEVOjRvO9LyrBxmyPfsE8CZAW+hpE3BtFoe10Ipc4f1u0wru94R455OBEfRvd+d71n52suDQLmw1JLu9LxADo5EZJkwEVEPLoSBN+0qmUa+NqYsmRFhlVatorFd5gQCaUG1vnuWxnBIQZyuxKJADRmewZB7a4bKIa3Jdp/DQX8waUzx+oNxPLmiRv5+aov2YmdmQxlU7rWVhMa9zA5No6YUg0d8i+8y5YjtFX6jqfChf9RSEHsUjhgJmmIuPC1DdCc2837PUoFWbJ9HkT44xzwQ8eYrT7aJH5gjzJNfFQrcWxMmOvbqdGZ7jeVzQIiAvceXFl6Cmq0QJQF0WqYvOtRlFtwWsj3Urh8i3FizGAHSukN45kEYlDWUIOmadsGtku2nRDcMg2A3Tw3uaQjgPIV2hpxCoMLhmI35PnJ5GCNzhXUde2pfFnhzzx2zG9snbrQffVjCBQEF8ytOFtM0pXGYI+fA7D8aa1H1bDWSNFYmoqfVQ1jsc1N1ali5hcpl5pbtOK7jfA3LzliNrmCABhinV2dzcM628G98vsvMi2QEd5iPbBSIMLAzJIzcid/LKeWSy+mbtjAl6Dr6hgrBArwxVcJ8UMKkQqi3yGsAuJjr2Y4KizZHXNcBeWW9/WNn7qdPc9FieaaEd40h/UzyS3fzspdL9MJA4NRbNqJBdeWlBOzg0LkvOEb+bW1grzMItwCLAdjwXHBFhYXHymic5vsfbhltnwKLVszt3WppqJrHxdoulkl/vGuOrsJR8DAwcTXPv+i9BI1vTdCK0dDyQTHODCCGTbjtUP/ee/ET0nL2dJLhx4wh1aVjUsiDzIYu//byPh3IWvVmf53Ee3qV9D/W7xUBgHSff9spIreuQSvOQ0Xszs0HgdJ5e8+3WsQsRGahVjEjvdW1IP3lvaJFZmB3bbTJ48ZMkUh71iaMettZ94z21zYRIBCvpukmlNbkZMXxvCHDfiq4/0jpZ6e7SCu53mxLzg7MqdmC3/zOgAT0o+EFCL78eXJm1MLBgEGxMCT0g91gWC9YkaQvt2qV18O/ft4wIYYKXXI0tSCUXbmi1Aq0lpyhb5OQ5yh5dqESUwR0SDEd4SvXac9NAMtSzgGaCQHImA+AmR1LwCEfE6BffzvA2ht7ei069DNABqHWcomwWWHa1Ke2DNC/ci95C/RrDnIDNjSPcJODW6RaLgkhjRqOCUhiT9p85LBoztr1o6bdF50zPVLMWZXNS5DF1KzONfRA+vX9Ith8aizBn94BnqUmC81gY1xBCrKwK88ZXx8BZQuAuVaP2Jc09Ss+w5rlJuMTG6uwoQssOVLhkMv4QA5xlRyBl4lcdyxEv1ILCFSF75ExTIeQrZHiNE87nI6wGtLlyff/HMh1AfA+oW50deLkvs1vyaU12/MPmQOJdVg/ZNtKy8x69BJYXYalSt8fuMoIJuV7n9UL3e0EruN8D6o51pMltUO971YAfvyEzh8L4Cbm5nIpoa4mXA0iLBbFY9BX9F17Ytv81picEpAFR19wAFw3IgBgAFVRisN0G1zRKmTM0ddHRhZ/Ll7ZTCCv+2OKXFf0M13KZOWk7PQPvzKsao97a0RA6H6V3R1biTMywI2taXWHnfN6nxfmeNXsZ7kNg4w7CUiuanhyYADx56xR0MuN4Q6At4/a2aVQ51qOQYo62jrWTEY37YEOlFC3HjsYps3WTf5jE7d5vE4hYJ66bqo35Z8yFnpcngt05T6wCi+ZPlBw5ibA0C/0b1ot6tmBRqXDyPppGxPXByPPX9pMlhgHprEsDN2bRJe4nA+0bW10kNj+oAnbLWcK8CNijf6w+XZk6wdy/wZ6Ogsd58w02h1Jld0AxlelzOXxT7MeoUt9HttrNmhQbQyHseXQ+Miff4CNFBTqTJmxvvxtKk76tmtTzoJXuDq3gfpfJACav0JBcswgLcCHRXhND7nkEgYpdo8ldvHfjTZ3TUQJKh+QWGqWDK9sCNcEi0jo/gK97fZ6AmwitiHMNmpyztuh51kqLTiVmxqRZJMFG+CWhMgtzN/txxlUVCmQbAAE0+xi/MbJxUKxRuZ/6D67luhZP+17rx68ryPan7b51sF5Xy6BS9M7siOnN6phGkC0M2w8tAHhhPHnacDbJ/Kit+ZZIRK6TMrLgNbtYJf1aEKZhizS3VDkRT1rnBvNL77VMCzgjgpIGq9EZYI6LU9Ewpo4MrMJERD20pSAR0gQgp1mCn8oVtvC8iSzCX3GzfAiFIbI6oGYLgy0GkN9YmLqpA2KThLvx7YA9mdTTZKCJwBuAuKCUKQ97LipjfHzg+GBx43tM32+cZiQTeBZQxzqDhvxSe32R23MT3lKpDBWs08SnPUvHgHj4QYRHRiERr5znwMZwaFg+7rCv4Stdm1ZwvxeUZrNrTYj5XJQhUWM5S6uMzZ2LAHRITwrGraZFqUWZVO3f8z9jrgOwqzNTV05ngk0ShC9AY+5RDllQCr+XU+uaJAQasoGVmeoZ0grDzr2a9WJUYlzj6YA18hnfp/417y1jOvm3TMVySGFgKacJnijjOxW9ejdnVjBNgHm+m+MdoPG+NVodTRrhDbL33cBo2wVtS9HosY02LGrCtiNSFgLWtDs2UOj2bhqqhQBmlpvcQM6Mfa5yiA+BJwLS+f5vH5vOh4QcuMskYMpqIRjvts/gYfv+xS4Lagy/RCZt62RQshrYnHfkGW80GUaZkaZ9OD/0sy4LAZC+5plAswkf3QupTuchVgDagPl9c6xF3px06mJnzVp795casRCCL7D14TkLxkJr5DVke+xW0Yj2mF7Sc+8NYR2KSLVpLIY+WmPL331awf0eEHXgLv/kMgX9XU1rpHHI/QINu6O7W7RwhmN3xLu+puuh21NNz20fNlmGEzM2TVu4hzl0MdrgjBZuYUymJwLkV7TSwFGC4xMN55GtDh1XVU9zfcEBxGBYVIIQEDKz3gN6Wa81pkOpTjuM2KrmbWDth8RcQwxxwcI1HH2larjZeEn604FCp0RLjkRAMD4Aeh0tR6Q3j8Sqfa+aMJFc2lNAcme8MVgzv2rfmkZsXUUIx0VU2/+2Y3khtTCT3jUfmqKbZCmZ901DNtDmGFbT8OzYW7WR8S4KfwMTnhpDA60IiJJuVZhvh9894FceJ2dKy2XQ2nsBrh98H4bAdbgjXK6rTz79MwO0QX9UNcsO+6xKo2yRsnOzfDaVIeG6+4QAHW/wZZLm3V6MtK0gxDrJAj9sjWTek/Id6iw8SLaHRCi0NS3CV1skG3fSzDllwSyB/epUd/dpBfe7TAT4PeCxFwVsSADeFhkVOYsuAUiKXkRmAJs97lNeFI55sY+vPEGle/3V92HzhR1WQdvtzUuLXJo3/tIDZ1I40pqkTrMRRs9IVdnDcKz9Vo8enBPbBiM7bKUyYAwimxed/3fFllyPvhbdi5m9UAaplGZsBqcXam1orblEYaAoCWX/VfwC2TUeuXTPjKE6UwphNlAskBj1Khgwk4OrXboiglbD0lQ40HHvYrynSpOWa9p+UcBs+vuk2j5DtWStmYGmC0oEjXegnU/K7PfMFQ/2o34ZU+p8C77DJLEQzFvfXTbT/LWxjhjleT4O8xwGmtQBiv0/TmYgnajIwpn3WLxEgN0yV2ZrI3eWqguJznsl74mnDhzTUmx9dRv11L/X5WyTIAtdHA61eSm7xaarR2yZ7NRqkCnMSkga+8H8TXyVcx8PQQdeHE32NHml69FTCtyJ6LkAXgbgPSDz8VuZ+RuI6JkA/hGA5wN4GMBnMvNbSNSVbwDwxwE8CeAhZv7FC8sAcOKgHQBf/M5u9SAnUsc6NfUlqbmzUDnAhnm/e67adI5+6TexsRlHFcrdYgBkvtw0DZSp25J2ZpoqRen/Ix/a4TFJS/D3izlNpQTJfNqv8dD2DFRs75jYXO7gTl6eTUgt+nc/S406dvq+vx8af6psBhWvpbVRtWKtRL4DvGn0udhu4GiznUW08TGBRve67f53wSppbNtWnBmztKo5Aw4PeA+RqhMxfAtkz7+kS1Fi/mgdNNyuMWQTsqqaCqZCyD4KJiJlAawye2AcVhQ0zC0IcLS+FquA3p5nN+4CndXIyzCwHMfcK8RpoHopU+ZQzGPLwt8i+xxzI8cwQuPkSb8Hm0aQdSDMzmypZrngHUqbDW5ZyO2h4S86i9C+qu1s1ZA1MAQhqDxjAuVQAkKZCB8CZtarfRHWARuKpDCYbFIKxxxd6a7SUwrcASwA/ioz/yIRPQDg/yKiHwXwEIAfZ+avJaIvBvDFAP4GgE8H8AL997EAvln/nksE4OZcMBV1DvKLQgBY9CaF3EpQACCVzKGTPxyS+rxFOzL+FSdYOGkwtqjgAoHf925CR7oP3Iqxc+TCFymYm5rWzWLg4OHMOi34waRndfLFrIVNoLFp0WYEUw2eHT74fQp5uShaEEyI2eXxWR7puGpkF/myjUeYtQM296Sx/pksipvtXyfwJoiZGeKIaOA5g9RlIVtO5LIYcITuLEmqy/2pON0DU6qkaeVu8vcz9vAAQRZ7ntkPPcAHLvUfhQwWv6cXctAbgIEJmBgei7+zakjlkpZuY6rCTJXKcIps5icnUn/tGcD9MLFHvjOg9D7LwkJyOM3b6t62YdstfqT+vaHszkydJnwWzs8jSxMnQIA4lx+Vye3xquWqaCVGcPf1xl3VxFFOy5KfDaQlBaGENaoy7G4C350kuPJi906wRnD0LckL2r3S1egpBe7M/HoAr9fPjxHRrwF4DoDPAPAife27ALwSAu6fAeBlLFzr54jo6UT0XprPXiICbhzNmKYiVyOKHQwVsY9pgR0sUIQrtr4AuWeiCLAl1WzgUcFs7yx0gjDHW500vYUzLQVU5LY1MZ21tJqFg9tRH7NGF8fDGAAAIABJREFUNkq8wvj0yEGio2Hg2wkdiSn4cnZGoiybzbN6kG2y+jT8EFqgAcUATDzk0etPA0fNbDna5sA2KkzaP3MBaio2Czy5h4h1HNKFKRPgt9I17z47RidR5exymNaa3MWuKhUViFNcuswGFPvislWgMfoBzJNoxR5cxnrNuHH0qo9T/I2+MOHSBMcAdpIyZ/GKb9vm7Zgm+EkS8vNUMn9JtxqknxsqV+h9wshX7rofyQjuSTjdcc7KlqPu+SAUZWrkIWgzdtp7AmDnCxQ0pMjdO5YVhiVKSfo53j8KgVNlySQe9e0DW/59qW5+vwOqZkHMrIRpq15/rCl2hJ3ESOBu86sUv8MAJONZJMLFnl5Z6br0lAL3TET0fAAfBeDnAbxHAuw3QMz2gAD/a1Ky1+qzDtyJ6HMBfC4A3Dg+wnw8o5RJGHKx87+mAQsQgxlFNbOiUFOMqUMCfUjmgKwU6rScUuw4le6XtjDdGScQqduAmBIDJcwbuQqUl7q7DZDVQZMtmHYYhyx8ThpLSClmOi/GmMjScy4kZ+egxuk1AO54FWqFc5IdxitZmCqkSfy9/gjSmF0aUG1GCFUBhMjc2Pe3j+YJ20UuSXHnQO37iexWuKbjrFempm52U31qjEajhd+5rc5whVnnRzhLGti5N34WX9j20RXUmwqJEE3dtGXrKy8XJnz0vg2dTKefw+TKAuzHBN42LAv3sesphE+7dc7mKcCgMonmvizqea35uiUk5lhYnwJAfToi1Q97xjh970zYNpcpP/CZrhpsmgK+nRFw7pmOt9ukF8aTYDTMaxPSLP4FIQFnl1VaAT4m3H22FrALRl1V+gzTQqL8KC1B3yrJ+WRLhrWL4ZKgh152QZJUYSDnjSvdXXpKgjsR3Q/gnwL475n57V0YWGamnVtELiZm/lYA3woAT3/wPp6PZuH4VPxISWMG16xxCDufyFzAYq97MYDIHMk4ozJ7NAMN9UpFL6ETYksw71fLYpJn80bAvtbk6a1ZdNrPUI3ugXueORIrY9DFXMJsKPvRGmiEw7RowL6/0xMHHPsks6eBqbO+24kQxjA9re3/9rvu3pcUvekaTEprQgysnRq3gF3QCnZcCqFgUm1a+9vB3/Y/yc/Gi1Ycz208TKijZoqt1oMh5WtcBGJGNQkC8H11OZUh++2wHtCyYNox5Ix86rL4yykuAwyt5bfGADaE6YiAyqhnUWcHCIJo7RzbE4w4ITBNcpyvLs2kvc6K1flvMHdj7jJmRz5x0vwi7CZMj/17n1n/LW0RuOBH3YsmSnLOYKyf9p0JGPlyY6sDDcndqY6GKqausf9pD7rgZs+tzFgfg3lfh7bYVN/TESZM9YJSWnNqmXEB3+vOQLETFkmwXOmu0lMO3IloAwH272HmH9DHv2/mdiJ6LwBv1Oe/B+C5Kfn76LPz8wehzAKrNu/diQk9NtlZaouwRaoBcY2rRN0UnKV28nUjTkd2CQgbqMde5sjDbJm1ZQG1hqNpQi2EpTaPXOaxyBPixhpms9DJNy3U28Vxbp9Cuhg0pxIXfSRTn+betVMLicxHsO/QW9unnEvy5fS6srGcnTNfdibmzJaDQUvO2TgtlbYgQeZRDIZc6WpI0SlUPNSZZPtGwa819nC6Nmfy9ocfTUyx6YmKB5yxMSoqFJDe0lfTb8HfrS+KhnqNPX910o9eS6FRi1lgGMn4ogF7JsLmvhlUgXZaASb1O6GwWuWuB3yuTwXgwli2Va6zbVbjAT0M7LsthCS15fmQB3IfURqjHaEg56ElU8x7JN+ZGCH5RDm6VDddTaxkz9dAl1LovyxYRaoA9nxi4RDK66Brl/5I+UUD7hKRMvN9D/uK7rYbKLWZNf5C1urTXLQaaQiswxqz0sH0lAJ39X7/NgC/xsxfn3765wBeDOBr9e8PpudfQEQvhzjSve2i/fYoqA9w7fHOOQDYQcwYcZEfmoVh8ytDEyAoMYe2E16p9hoHzzJzblqNJslvF8Z2qTiaC+Z5wjQV1KXhrFagttj7RZL8ORYogdwywJBtBG5R/oaEqTeSM87Q/CjVo7nkYwWp4GH5eh/2QkZHPHxOlgTf88vmTu8K6tJlszMDcczLL5DZFRLA8KiBJlZNdnqhY3LWedSxMbPoNI6TFQXkwdg9rKymyFqU9GVxh6QZYm5nq1OB+lVE/YhIjuwpgy0klgK3XJFo7I1ZPfvV066YNlZSW6R85oZlqeCJcPTABhMIy6mE07U761uNgDUgCYvbmrVK+rIUO04oz0hd0+0qXB+CiyaAf+wl6T6Jjr5tcxRK8QRsUNPcy2Vy+jheMNMVnwDeNOyuHoGsNs8N/kKU7AE0L2PqPuz0Qv+AdE6kPA2ws4wrcz7XLX6j4UEnCBH8GC9bi8wE3y/iThDisc/OE8BWujI9pcAdwMcD+GwAv0JEv6TPvhQC6t9PRC8B8LsAPlN/+yHIMbjfhByF+5w7FUCFcOPmfT5X85EQ04x7vNbFPymzR8PRWYXcCBoOUiHiCxDZWd9tbThZYietAJhIriQttumrTMjM8vfddxNLbVi2WxzNE6bNjDIXcAO22wVnyyJOWzlYiHK2wvAz2HY/NEjbqabfQsA8FxAVEUDMWzqRncd2BystJ3Oc+26eYDsVTNPsV0sxo1fGRg3EhYfoc85bDsX6ItLBxihELpjVI24zS9wwIWzjhpOTEzzwgPSpOzpa4B0dvG6PlcNCYsfePDIhx9hKnciv6DUJq7VkqtaxsPvbaw6kw30bbpyc4MEH7lfwFmNpeOPrCYYcUpa9UBARprmAW/MyCBEWefPgBpvNhO0TW7Qb6O6rj6mbjl+5QGqCggTkYSq4ceMED9x3n/aznSXQryawJYoRC8qgYeVw966C+xQnWjxlRWp/PLb8CcCNk2OclBt42gP3p26K0xUFWcDrARPaL5bMWEKG6lyW/zWAzwiva8J4hOOz5n/fzRtYasW2VtkCyp3mc78XPEZAj+O0+1oSzxrgERF5GI29hneGBmcCTm6c7Ml5pevQUwrcmfmnsX9+AsCn7HmfAfyly5UB3DJgVvU34TJQpgisotQao6kHHdEEKgUz9bzFzL6xD8sBAIOzm0efSx70vndHhOOTI8y14fbZiUX3BJ2xl3+8mXrspCStO7KGg9FmItw8Ljg6nv3yDqIwD4vQYnUGtkvD7VtbnJ5W1Bqe+oWiQAZwdHSEo4lwtLH7wJMJ37yjSfadt0vTS0k4zNKWH4d3uDBD6TcJyyke2nZ6weKYT6VoWRxR3RT8iSLeADOwOdpgPjpCaRYKJv7ffTNHvf5/IdikZ0gf867EebuTXhrpZS4zYToSp8lCEwoVnNw8xnR0hMlN8gS7U92Ky9sLgAB3td9KwYYGK0OqcTtjTPPkXCW3mYDYyjGRMYEuGOCZMU0FZQtMT5qK6HDUdY2bpq0cE4iGHglnObMeGPjIK3SW8kzrMuff5yDp5jrj6PgIN05OUucllCT09z7sbUQWSPpZ4xU6z5xuwoyBOiWh38phYNpscLw58nP5FjyKuBd6+kulUjs64D+HdqdsJzzslQpMeNJ+Ojk6ulMpK12SnlLg/o6g2hre9MgjAATczTsZCGG7kC1NcobkIUTJ1gV7xDGPYkdwhylG8zPLWctgGNNOXrEcZjgiQl0WbGvDI4++JeqmZkZZ75wxPPbvh41JhtTtZCqoNzc4eeAGFjCWZQugYaMOXgL4KjAQg1vFrdsLnnhii9u3tnKmWfvHHcgAHG024O0pbp+eaoGJETlzY9zaVnBtYilgyEU2ZECnTL7FXt/SGuq2YjNPHhFNypS+rLW51zlzujnN3tWOMsfAGzdP8PtvejNqbc7H+r4aOFtifCECJOuF/WV0T3eAnYe/Ws8JYjkpJxPmkxmbzYx5nkBlxpve9Ei3/0qslgGCmpq1jbrnzSyOeUSEaZowU5ELcFKEQR0atwLY/wl2K11uX9JaEyBb26ZpwjRt8IY3PJLU1GhnFmJyv7EJZ4hofx3QILYG8tYQpXGI3YksNXSZSDoC7rtxA2cPLvj9Nz3S9/+QrB+xPGC7wA5Qam4vZNlrUVTfDuvDxhHcCTqXN0R4+2NvByZVEDj5b5jFitO45EyBUFDG6rpUtkcA6CWl0BI0v15QAx588IExh5WuSSu4321SpPYrIsFy5CMtTLUwOwDbOWBA5r2sBVnZpADToNd9Nrk/MQN6YEVmoBDGaoyN0xpr7GWEQEGevjGpKZv9CJwf50MyY+om+kSEeTOBjjZYzhY8sTCIK25OBfNMsFMDxv2oTDg5CYe309Pql6hwSDfenWB2M74xaAZQueG2av8CxlU8yJlRGJim4szc7geXEwoEmgqmqfQahpZdSkFtTW5rM7XWhCVWs7UOaMSWYT+2lGoen9NPKsoMPle8M4YGjJmhd5xUATlhg+e/NKBsZeIV0mhyIEylqF+HgWuKE88mYCUQ0kA+M0lYXBdedlQxdMCUBcOonNSdFI051dk/XCQUIcGi92d2JkzOkZY+4fQIurvm8BiLsX2dUX9EMtqtp0kctPtwb7u8FOuUHFQqVTq6K1miUltsrfclcggyPgwBysFLCHkzfDQaWIbdMHMfldD7Iws6lGY1p39901a6y7SC+10mhty9Tq1nM3n2GlhaOHnZ44x91Ao4Y8xCAjNQQX1Mb2QNiP1aOTeZqpnZ8ia2CxmVSXZKSmLAiOdShcQ4m69fUGsoUwFPBW0CFjQ8eVrBtYInxvEx6f673ddqbSMcH8+oTGg4xdnp4lYEZzr61Y6NGdAujXG6VJxtqwpS8Mh/jZv4HUzqqe0NiPFgQmjhQLKxw60hk1tIokKt6X3pRdrCiXnHti0PfRp8sweH4MzJUfo8fS4JAMH4AeqERkvQOHnfL4z5TBzV6H7x5Lf5Y45WkrPdvKZR6tI4TSVZjmzKMDoN17Zios8SnHYglYAX/TgnyWHov9yTI+jq3wRuLkSmRdcBe5KP7Lt3K40l5Hwjp/j/fgHE5tseEaiveHqB06fYwqC9skNfVko3/pgEACrIRyF28tmt8VBJ6xsfDnZBkNLbSFYdQupTr0v39kr3iFZwv8uUpVX5U8QDOTkhJQz1Yy+ty4MSI6XYJ/S72S39qLEEeATrBGoz6wBAk9pUa/Kkz9yOQ6gIzV6tB539n0IA4YLWCtpZQ10Yp9sGvl3BpWG7ME6OGJt5FqctQM64aijc482E5caMWhnLWYQaFWYgQOvAToTTbcXtswXLEveSiyleGI3UR8HZnNrUeS1iz4u2V5pG4tNmVZZxIgftxOk5ZAQPW8ohPBWifn+V8kfqeFm+5KdTgxLfp/Q3f3PrwB4e3AOWoHSrC1ptaFQAbgLULY5nWvAYgFCZZa5sQ2ubNnDrjuuzA7M28GakOav1NqAzTTFXmZEu+vHuIP9u2yXndOpOmwOGd0LK3JE4Aft+6Mt/dwWw/AanD5TbMsomA7Dz0D/SfLGeWT7nSB5drfe1wfLKEpithyxcMuDH1cYMdnYrbOzzWGgdeyEALmB0VVWBKue90t2jFdzvAZVuFRAkulxzgDUSR7Oiabhb3TbXQ1Myb3t0HIOgWrqCrx+x1atbzcydJWkC1Js+Leq0YjNDsntFquU1vE8Qs/wEQtlKcBWAwa3hrIkPwnbbcHLccHMjzoLUxKzNGoH0aJqwzAxatkkQUcFEj24tS8PZ0rDdVglvCvh+eoGF1ZXGbSsDS8NmDk9z80ZvIPHQJQkDXExI6k5cEW7ME45vHOH2UnF6tgU3Mec3AEttmObkiV7EyYxzh2aG1zG35AjVjfbIULN4FuO3j7nT+MA4d2OAmp4ll/Nd2ZtbxlHf1evOmBmskVQoCWFWdrdnvY9bG8JneO/s2H0o1YhwGLECyOd2EngG4XO3xbzT9/5xD3DsA8CuOdTbC7yagwDWlZPrMujBzNEjYx136jAIa2Vs2E5a3mmMudJS6jCqAtwe/nIQkAScY/spd9LIj7J4w9075C96cp/73AuCu1LUSneRVnC/25SldOcKhAnFTVj5OJydQzaHIMlDPOAd1GE3LaUY3IPpS8z7xc3zFrXOgL2TN5jRQGEtcAECvhCtKn4bmdZrlNoLGJOeqa6StVitPT1ju13AraLVDY6OZmxKUbO3OKltaMLRMUR42co5eyvjdKl4/NZWneSU6U8FJUW6mw0QtA2FNExAbZjVScwq7lp+cmQ0TW+ezAFMHAG3p4uY+Rlq9pd+tiA1rGNrzoY7HtbOxXqtbJcyN6dOy+n6G0BvHujT94CloEoSS37R+lKRPLrbWW1uMIEn9vlrwp3N6c4sDxWKBsaeA9uYv8dO3RIWWThXb3PKb7elu8hufgPdO5bH/kzuDKz7kD/9tIN49tnaPabjIcs95ec1F3VhxB9C5zhzThvG/qP8PP8b2pj73r9w/+o+YO+aQ4PgOvLB9P++685rzErXoRXc7wn1QSLsb9EFSohY30yQOO8QLXVbg2HaGXAAokVZaFcWU13Jq5OAbF53R7JBDRBAYz/6RbS7Fu097lZ8WtidZiSaeynAUkRDnBgok14HQQKm2yZRy+qWwUcFx8cbzGXSaHaE4xloxyrEtObM5IH7TrDliieeuC1e4Io25pTYKrAwY+bQCuR8v9R7cXO9Oo7BzpyTMl258aoUaUcFY2kNZ0v1OAEWDc5AhAAslV0rqq0J+BOwa27fgxMJCEby+iWQt6F1sMrIr2XyOBf0L4OxcIw5tF+ZyG/TEwFSypsMlDmOjvmg7wGVjr0nJBnN8CbA2hTl7n1j97bldJEkZBOZdnrPxIgBcjB+9bJHCconP7vDZJeFCWn5px1BgJJZe781ZqfIYeFlYAdHLgyEtrtnntlLpJWV4VLB3vbb7bNlmIQBiwfBiY/kooAxGI41RksyQbCLocs+9pTbFYmx0r2hFdzvNhmAjCym2Hc9ZcsApoKT+0/wtJsbTGg4e/wW3vbkGW4tAbyiwau3OyXNKDHrnkfF/yfANR0PIclWC0Q9EOvc8SIzlW67ICEO5J25FGzKhAVAq1tMRGADd8iNcsYwtrWhnTEYBTcKMEPO/c8QUzgfNaBWLGoyuP3Ebdxk4Pj+Gzhd5PhfrdU6Wo7b+e12AloC8MUd4PySFNajahYBjjTGOgOlMSrXDnw4MabGwGZWD//GDlQTAX4fuwFg6i33fXT1Zx9w9T4TMkR73tuPZrs/5QnB6nNR1WlThctg0OT3rFuS0mlWASIUomKAxlisO4TIKZGch/drmj+WF0i2S9pk9cFgpBiEhZ1LWSj3nj7qdcKx+6zZWc6K9uVVrL/l/jYPfU6/8e576KoZ4gfn37hPry91mbknykVyT0qWt9f9+dhlVp+01m0BeBuyl2E3r4Kf+TMFdpMF4znvcVbsJgBonEwrXZtWcL/bpMjICFDJ5CY2CPgeA7hZGWDR/ibSO+JIwdeWnQm8uoi4ZafX8H53bZ0M13vId23cxWxb85z4p4F64FEI/PLFr/ikIoHBpxlLY2yrVlJD8BIEAM2kz5BwtLe2FQ0NxxvG0Szx0WcinMwz+KiBz8RhbtsYrTacUAHmgidPF7A6vtXaMJWCSZ3ipmLH6+R/RTWI7AE/TQokLe6ottMDdtwOYMzFjrsF8LbaQJMJEp3uAosGZ5Ng5O/d930g7dw3PRs+nsf+/NU9ypxrTIYLHMKnzFEkiwTHSYxRI2WbK8mpLuFtBvY+IEoCpnSzcDRKJ6rFoJ8n15CzPWJvn6Vmc/pmQksHMjudllJ5nwxVQ/TZWLA5U+YokmF14D4nW0cU58u7egx17OtN0Xje0w9juvyFrRfGgRpJ6kyIQE47gpBlMQC7iyvBTna6nI0X5dsU7X+7tv6V7hKt4H4PKE9hXwZpwTAAFAGVJ564jXpLnlcWE6o4sbHf+CX76XI3dwN50BFjPt2CJ3Qqia2fKDixwVF7gjLijrvJF3daUw3Lo7yxOMZVIlSqqv32e7HmJEVkYWqBbWuoZ6KFt3mSSGpqGt9sJlS9lrQyUJeKLRhbhhy9o4KKis1cXOKf1bFNItEJighQiSRkZ9btffech4A8E+kFKxx3AZCO09Ic+As3vexFfAwWvW98F86HOcFD39p3+9J2knj/nZ/r8CL379mFg2ZiZobEE9AxLL7lEOBmwOKOmxxx0jn9vm/v3wAvGhiXJ2W4GHQ+U+4AFt8U6Lg5phmycBS7F7gHwL+os7qjWtayNGlHgWoH31lDCKfIh92pAII4r1G014T23Ac7jeokNatP9BXnRbszIWwsKZnvY+1RsTkdv8l8yEUNmdL4MUat4y3U/TQIhuiE+y5DZxBjW1a6Lq3gfreJAVgMeTJADKZB+H/Ze5tQ27ZtPehrfcy11j7/N3n6kpAXSMCgBSGgogVBgikZg6loEhRJYuBVNAQsGK1ZUNCSPhAiD4MkIjwllaQgVhJTNGDUkiEQQiTvYX7Me/fe3HP23mvO0ZuF9ve1Psba575717aQM/u9Z685x+yj/7Tee/taa7311smzG8DcJ76+JXeAimn0A4Jd7Fy7woLY5MKO4vzmJtV+TK60Hi+zeEoy24hp3p6HqRFA7rn6f0PE4nAT9xe1MrAJdPjZdo3jbFoY4At4JgSac2BcfXq7KR4eJh4uA2MMbBA8PAzbtxfg/VS8e3fDUMW42Hnrh80c8uz/tkVx281sL96/1KjEvf336Y6NaPuK4VRX/NCiq40hfkpAMYYNmNWxA5tibKObppl+SVP6LwUu1tx4QOnl5XuvB6fAFlsvoWKtR/PmBHY1QVNGRT/kPd3IHPvuNp1rftq7AUYkHChpfF4/Ty6SE7Nrof1bGGD772Hbmom2hNcOOmF74K2CquTERSsq1PjV1UnaRihAd2KuVwFTB1Thp1dmCR7xAzc3xypsHj4HhAcXi9QSDSmRWw4/cp+5f5rCiQB2IsWKyrDB8WNt/RVVo8wz+glliXVlZNQ2Qfv2BRVwEL5eikRwT6+R7uD+ysmAWHKx8cKJu6vtWlc4TwmG32PUhjyw5SKyG750IkOp1vVpVhd7J6e2FAvYW1E6C2lQ/p4OZICaYA4BlGY9qAVowoHiYQzoRbBjxz73BPZgYhaox7QxrYYVL5pq99c/2011Tw+Ky8VimD9skhdwXOduUdLcjLv2YQSDqZBx1tdhnzROG8yJcKaD38A3hv23QTBdbQ+mp3Nii2tRvS9zKm5TMXR34a2Eg7Ba5BniBHYkp6v9TSSdOpP25OM7AL5grN4lxkziYSKiOm3CDD+n4rZPA1JFXhWbYwogIt5JPpC8IdBnJnRKAWnOLXIElYLXLIf7HXN8CIZsuFwGLg8WPe/NJw95m1zN7Q7sWURvOOWrbapqBLdFqgCBOYXEWlnrAQlNwAKE2gC/ntcnIlMTNstlQPp7yzgfHXPpYhcap1zfAbgch8LzJwbHGtQSHmjA+0mKLm0cuqpcPvW5D1k/nXEczcML9/QK6Q7ur5zs0haf9cnwffJniNXiLcmS+rVh9m4LPgNAXFvdHPxjzw+gdVc3jJXTT5xvDvYsYO2a0xDDPDaiCejmUwTDM+Zs4UyB223H3Cd0zmx/muNR+9f5Gy3mCeun7LtF9lMLHbuhHNgexkhLQwgfKeS4kDCd0HMqNgd2gYHYPgDdbeuD2zbS0dHKHAPYp+A2Jy5j1G1pni7DzCVTFdfdIuTtqj3y1xko6Qmo+1gXcySpioY0zKnrWPWyq8y0Qihofhj955yAHw8s3w4hQaG4sFkv6Jx7tDHFiA6mMTjDx5xhS2Ehgc2gZfPm8rDh8eERDw8bLg+Cy7bh4fESlTdaRvkx/iedb3leJlaYqONioOlgSKOlsBMWicK1TplEsZZ4vMu6xiMSNfuKU1B8AbYYLMDIMgg7B3p94s8B8oeJXOOEBOEtL/D9sehDdLtO8uS7OcT1MC6uKpZCDVL0rYPs79qY6sod2D9OuoP7KycFO2ZZWheKuIe3aaAh/Y/m/BTi/XBztpmblS7sCJap6MtYydFODn8D1AzjiKmLm/hDXud984WjGhhHWwX7rpjvd8x9L63Aqw3hopx0NN8DFDpCqldgAlco9Lrjwfc0Q6sOs+++Ky6ytT38fU4M9ZvcnDvOOe22O2KQFiGPj7a5NuPe8+FTMMWO4oV/ANyMLRImfevjJqRFMYdmXhXSHeehDCGw5JWvwcTj+6hnL+IVVcbmcbYcJMlhmngcp7Qx0qxPnWOHxWQb4jH6RwpUc1qApIiBUAIqsi91OVKsicozxsDl4QFv3jzizaePeLxccBmwbRconil0c5b94W53GoDGBTh8GpA8vRLBiw5jsy6rtSp1aGcLAGiTgM3uJPPxpxIWaj5qFZV5eUrlET0nZuRXoLWlXncBxSdQWOcOhFUqp+SOQ2pzU7LYXOuHPXim4yJA2e8uJt3x/dXTHdw/Qorz6SHli6D2pxCfS2u2OV7H0+xmp5DWjdHGlam2x2crowkQISgkWsP3wjx2PdC0P6uLuBftQafDdAKM5N5diBQTdhXoropxU8jcLcTuIlhY/gm+lnYkw/I2OzeJs/+3qcBt2i1rIni42N739WpmZYjg4l7vorb/Hs5NwaQyOE0ILlIWjbiY1e5VByQsAK5eTLd0xO9QBXZgG+YTYZZ9tX1/B8Ac4AOAF4OXFGysrpCZysJRABxfGjP91lSBY4hvlh+GzydDZtKmpMY2uW4wZxmQywbI5o6GOxIdYp64oFhzLF8us7AIMBQPQ/B02fD4ySOePvsET28esGHD7VlxvT3j9v6K7eGyAMux97mOTgSmLgnJ8bHTVQWLg1qgcD8El/Oe8Vr8at2xDg5/0aUf/afpdSpnJYCPFTRZaNDaNwtojzkj9F4TOmPrb/pYhaNl5GSgXz+vXUIXNKVRas270C0Jqcc67unV0x3cXz2JHZfaTfvbRkneGTTGnZr6ZA9t6+iFKmJOdqoGLFMhAcxLAAAgAElEQVR5saOuLXVGa0J9MdZwtPPWkeRd7HyWGpEaVrCPCQNAzh/x16e6g5kO87zd6VYqidIKytG+Mag2ScVC3voxuP3m+/JT88Y2gfhd5QUeU5ChabdtVE1Ok11nev1HtQq/KAWkcYqkO8NDHK+DCTL7bbdb5aKM0Li0eJb23hioS5IkMDPHCAmIMeDHOfXBJOtXp3NEPxSnt6DGUMrhrp0OyH55UsXc1baF8vnRBQ+oi5CiTNnihIJp/vJ0wSePF3z+NPD05oKHTy62DfR8w+39jrfv3uHd9YovtqcsffWy78ASc32NeJaNp5ZROcqgtKJLxJTgJ8sHmnNx50GrXE/++m/NB2e6o6yiAsgADfARQmrOH616hCliYy1SldjrdBtTTGoEsHODF0ooDs9EUT4tcgT5Na3ymTZ6LILPwXfhnn7adAf3104D+OTNwLu3E/sOtBUeXsZQ6BQPKAKUS7n2vSk3TYdXsCVJAcHKRII4BBl2Nv3S/bVBmQyUjaGE41ecb03Q57WXql/xlAELFDOS6dMxMqD6pQHleBG3FDBHPv8eXsoKixP/fN0xBvB4iZ1FdwS7zTTbxzbBRnvzrW/TPODDoTGsm9Cirep0C4mlCcWG8N5WL3cU41M/Qx9mzhl9JLRf7t9sDDGAPfdpmUAdXZRfVKLVSsz2KaQ79TqOcRdWz/T8R8TnkpnKL9Ou1R2z35EQFgf1z7FHP8bAZRO8uQw8PW24bI8Ynz/h8fERD8PKvV4Vt+sNt7c3vP3RFd88v8PttkM/rfkWTWo0aw1GjnWJjgs5dHksWGhXwkrLSgJZ1kWYaF7+se++Tu4+OLp8UgA6PAyy2vZUrknyxoz3jBy+BhU9WIxG9bXlhOW91uUibbtQiXlPzYWTtEpYjVDflmJNlPSwCsP39DrpDu6vnAQWbS3MdaYZ8n6zuFXUVpOth1mrKRzQ/GtN/BL/Bcg90YliAkOAbXMTPoFsMUTjBHb/tTSzrB2pKwY5UQCfAWvccWyonUd/vAxc/MiaApCpeSdFXYFKiRh2CfABudUWFla2TdxTXXHbFZtYDPgJwfM+LX68A3xqnbznHMU6yEXbxK0D8fugSHd5tznELlyB2xi80Bw1MuPWGWZt4Jv7l2uSACTJulbAOeOW37Y3eQZ+bCEIx3Cj+Qvc2IP+THHfEDUb/hAFxnRhrOac1TkgY+CyXXB5eMDl4RFPTxs+ezPw+ecDT2ODbCYJ3t69x9ffXPHN2x3f7Due399wfb/jut8aoK/W2xfI2ERfPfzKT2u8EuDTsUx7FpIm0iOeGhFgauGTR71wInVpCHzxxee73VdA801hUfpU09ejgazGepb8G1tTXCdvzWmoy/6H50/OaTC9TyZQ8Kk0kwFNc1/JLFjJ2Z01sYzTHdk/SrqD+yunAcGYdv91LGoVC65Sl7BIKvGBOVMIxANMQ6OMwlMRs2A3DCDJW9wZqr1Hv4dgz77OwUFYG7AjTwB01hWyIh4f38DwYTPTOMbm2wVeuJQwk3/U3k/+SEwypIvA/qYUKPC4GQMN0/xtN8fCy2baZVNJ1M60D2JAmaQc2FKg8S9muTB/hn26cLLZuXDevN4s3J4FvFEK5+uu4B3Mhf5Nvl4dZya47PNaHmLMPKAvId2amkK1cGA1oWIfDND+u1twNgCbKDZRXMbEiFh+AUZuln7YBh7CSe7pCZ9+8SmePvkED28ueHoYeHiwQXn3fsc3X7/H1z96hx99/y3ePu94VnPQm7v5cdh2CnN7PfQAh19XzfGIFrkNRvAVCuT5JgMPD82BlkHcsTMIWALaOmRtzEiOs/lCgL8JZE737bCx53ljrgtebwSzIqtBgKjmvCHacaO0vhdPIAoz04k/DOwfmn+B4dEWXZ9ZgRrmAjnS/p5++nQH91dOAuC677iFKVQBzDgGpL4wzFNZoen4dcZYgmfMFmoWyShiAVsedSc8LYbj7GnmylqVQytI87NXytqLCjA8VCvK2x4Ci4A2FVP8eteI3U7NVFXfT3QOFYyJ+sgMiI/2Keyc/w1+Dt1puEcAnTEQYdj22U2qE5J7wDYEmsf5FJKRxfKW1mCggjyRsOWP9oPfJGsCj+cPTf98JhB9V9qT4BZH4FjDD2YcsBMaV5Z1qJIEwZXx0vdB8QYs4JCP6YUB1LTBiwgeh2DE1bnTtXiobwdteHh8g08/3fDJ0yPefPaIz754wvfePOHp8oBtDAA7bvOKb969wzc/fIdf/dX3+P4P3+Kbt8/mOHlR3q2yuA7xRY5dWcEzWxwDsnaYcys5HLpQqE5MHsK1bE1BsRM9rctNIj2pu6HbMnaheVO/ZFgsBQgw40ge0cPiA0gD9d7kmAua8115fuSvXWihJZBPqnvSfCraC+tn7jZXyjReJ+od3V893cH9lVMyZQciDRMbkMAONcev3Vayh11Frn/mE8YXhI6m0SJ3oB0Dvmeneef7Br9OFssaFN+jZgceXvJaizmwPcyj4XBloC3YNRzf7Da7edsRZuuMGBp81RngpGfmO4DUPtQlFY0XFQ6o4f1dJsypwKYTY5Py0m+cSfPRnPWDavxSjEqSu4Y2QUe5kqEOE2TiBjj1i2gkzsqT+kHaiQ3XKXIkEr/M12zkpLLW/FqYaYUNlZVDt+9EHkSY3n0IdJrTm0VHRN7Wp2MAD3GVjNFnQPEA4HJ5wJff+wRffvGETx6f8PTFG7z57AFPMqDPwNvrxPV5x9uv3+JHX/9D/OD73+AHP3zG1293PE9zshi7YEasAQV2AtoXtUMitdB/zUkTNM98zeTrKd3YWwpgFa8JI09FhiD/hB3hsw/S3/0WwIqxzG/+TkwnW9h+miQeaZz8WMA92upjHfJEOI9OrotQlwWpKOYQgVbKusPVtXw094O8vK+inGcVNO7A/lHSHdw/Qhoi2FzMtb3L0O5s1U51H9ZRV42KFHDZTyVVizvt1HlxScYQ2nVI1dP37KajgWmr0phVrMA0w7n0Uc5V4SRkZWu78aOzuAz7erUANIOzaQBkmRDTc9/7qhInB4q5cVKNyHJuCp7qQstIQcCaWYy6GDfKXwB2YQ2mCT1xk1vsd5sQo0UD/u7jOAR2LFDNZyLM0o1JK3ofGBkOl4TUF12QrOHzCszEYfu40s8lzdkfrbxJKVXT3neD8/Fgc21DgIfFhNUQVmHbFjIFD2PD0+OGLz694KvPHvB0ueDhAlyu7zHf7Xj79RU/er/jm7dX/Ojr9/j6+S3efvOMb95f8eyXJOWpBc1R85aFc+ba8Q4SbaosQVu6Jm5rKc5PtLcDgNZxy7ZUyOTlpxIeg7Zq9iKNsYauQ5PppechFYfRqDYp3NeGgZYH1NvEvpkaAtqc6Y1f7dely3SaJYUaF9TFhMcWzOgFQKYdLJSSU3XEv3c8//jpDu6vnrRADUgNmwNr5EUeUHNMEvdaRx2rMvCwcsZwT2/1u8NnMB11L3P1NS4AxU5XwPZyg4lmW1iI14DybLM1PNWe4qiLCiMiwBxQ3SHTIscHUO6IRa3JkNT9AWJvvhQe1qDQVv7wy172OUnLN8C53dRoww5G0c/oTQhYCdSapuigxwTiuECecS+zSZRRx+Rir32TbqHQZpZY5wUxRmLIep65jddpau/H1z7OR12TgB0RX0CdcQOPF/OjuMiAYNhpCPe0vN3qqt0hArkMbJsdf7y+B+Z14pu3z5i397h98w5v377H2/dXvH+eeHvb8e52w/v3N1z33ZwuBQCdPIDP99UR86Xun+Bw185P8pRp2rzUE30ORaxChfa/WuWkOKmxZKaP74kEEo0Srovyxe/qwo1XtYYe7tturaGtqhTkUjjT/nvkYfv5QZ5aBO8o8GxiHgmeRZ5ZRpYhu6dXTndwf+U01W88a5oD+kKTil3OG26SQIRaUwLIbn7bkTNCpoZGHqwgtPdy6jou1jjjG166/Kvmx1XjINOlhom8xBfR/n4eQQvwIGEnAupolqtlqk+G0RlBaNchKWVIXthd5UPU9seHYM5yAIxtC/V3R9Lcadfkl6JwgLXEGfdZ4zUVtaXhlpTYv19kn2MS6kfQ7qW8ORQNbmpfflXEmHzRflly+HxKOU1NOx8iuAzgERZkJiLSRbAhCyu84+qdvAxzssO84vnrb/Cjd1dAN1x3xfvrDe+uVzzvNwtJPCee58TzdcftOi1Ub8hOstc9DD5W4dQpS39WtGZwWDE4fyP5LL9L9H1x9GIi+5fQWAU2B1gQYTwMs3kA/Hr0/UTGqjYtzxXuXKt19dPart5rOcyFXkOAe63hVSzIdnA9A0uM+RK8UzB+qVocx6DK0F6nZ/6gg949/UTpDu6vnBSAXgawu4btKyLMwJEHI0zv/nuAc0j9zN20jloxCNZ6kNoPo33pdIiTWmi2iH2ZKwkCvLq1L1xNqSSgzwqanmkMwS7D9qCzg64d6bRtCsCuhlUrsJijJuOJZtBPVr7AzfrhCOfgHiAB5L77THq6hpkMv+rkz9rq937B49TP0PZLmJlTzYoQGrzTNLZCWIg643798QlC54guPx3Q/SgL1eu6/JUkbF7so8CmAxexyHuXATwoLCa/mN0j7i/Yvc8608ABfTeh12e8f3u163anBf65TuAGDzqUc3HCw7WUn0OJixkG15C90+RwDp++hEgZe+1MtJxHJ0AUazLAPac0CVI5jgLEdc21tbZQmNZwCKy1nCQbsjpextztD6v19ZHGXMs5rstuMbkJkaOfIUD7f9tZlRD+ckK3JdQtli/rlIv+ae+nUDnthRcFlHv6SdMd3F85BYg+XDbsc2K/TQsG4qFUI8BHatjEPIEAGDOHppY1p2v05dkutKDyUgkFryJqUAB/vG+rjqX3NGkrvWyct12qoZBkdlMnIHYGWmWzM+Gu1QoC4Ez/U1ebpu+hNzOdkjZROGS5ZrClQWZ+74dI0nEmsmsyJjsmZ43ni2+idr6u1MiiFUYQURwDuENJCFfOtSZMaNm9H82r+ADwRMwPpP76UUoIeeuYzh5qAXuE6BXB5SJ4GsDFI8mN6OR1h5+xwNQC9rRceF+fb7Hl4g6hLpiGgMVttXlfce1zTvqH2kY5TuHT7umSp6FOfxTZa3mUpaWautQW63NGo7jCJir48wB2guYsvLckVshB6KBPcnyItPPoC3lBpyq0CylpWZKa/0V/39ai87kvt636r+fEbW+Ull9CUVEKeDGI/T391OkO7q+dBHh4eoToxO29TdrdF1ZcS+qnjjIw5HBN8uZAssGOzgG+KJUDrmivzPPY49kk5AxLK/VuegrHdzeFZokpJdSii0fsk6OAAcWuuIk5q+3TwFhgzF5cC9wR58hJm4Pm3rVp2WYKtrb6kToJMzgwhi6MWGEWDXHot39uu9ptbmTfve3ThCetEwDMYDY7+1aXuoWT3wgGbALFrprbCpMFCbULbeZe+6SDI4gFEpND3bIj0sGqIRbQY52vWybkbb/gSM4XdY3JtxAAwZsNeHoceNy8rUZOXHX6zXEloMxpVhqB2JFO6oCi7kGAO9z50CSA23W6AzcPT5yaZ1iv/PKgIFUJNAxPQMJSm4cxFz509johHbpGFFpw5VSYUvpV6w/jmai2oEQJaS+2iQB3bUS8H3NCY16sjV1mAiFyLuMXxcMowgZ60JZINkOqjQfZacXjJnSdiAJeVoI9CRt3bP846Q7ur5wGBJ8MYH+euN4svtkYpZXAgc5SsYghZhKNyb7P+EVzVcRZ8wAJwwwze5qfbj3P/foVTFALXz2fvVNMF1Gv5xOUCd7eD4c0e7jPaTfCRezbYAND0gt8Qt0C4WWqw7IWAGAbfhMbIV/ssXtceZZWysRrHVGER6+Dsvp2hwh2ILXMAOcUCiS8gp0o4fcwJyL6mEKxAfjsqzf40dtn3L5+xsPDhnG5mDAwJwVKIcadjNnbSne0H5L3KfrTww73gQs/As3vVUeMAHNNETPNTrGof4+XCz55EGybF7krrlPxPIHrbhYYGRZiODU6UPtSM/OTHvnIubhLrnEtsA77LPsVFe+hrw0G9lW+AVDh9cBAE+IXZY8x6CQ4vpcDJLVAYhuGy5DDy/bKBB1zLQhMOKVnVYigekHNO344pUGC/Mn84C2M6I8olxdjRqWJQDxk8xi9ztjesxMri4AQ7IEFyVbX4k/SrGRNOmtC0j29XvpOgruIbAD+NwC/oqq/T0R+B4BfAvAzAP4qgH9bVZ9F5AnAnwXwzwL4BwD+oKr+rQ+VPafihz98B4XvL4f0nhrNxB4MUtUjnWXDcvHOyB9Mjxi4jApTX+ZFbQxJ4AASQUpAPMoz1HOS7nVdaGnEzvymJQ/cRAH12+BcYxdiW0pXj6nLKKmNlFRBfHbCd8thXseazoPJPEolSdoGs5x+D31YCKBxZW5RJYPOeP379Atp5NhvKPC4AZfHgdsU3HTH19+8xX6zi2k+eXrAzcuToRCtAPmhnVhJrplKORdG95v2JfCtmw50zPiTjmBA7eOa3+NwsyA11ssm+PRxwyePGx62Gq9ddswJ7NeJW3jFq4HyWG4+C6GQ55T5TpQzYgmZAqh5xW+OhOqd36ivUXBqqQ2h+5iEUbkBLwt6HTu6YJsZGvQcEoNvbFUJa865WGL1dRmAm3amvjPA8zdteY7Ob8dma/vEP3fLwEIEBcQFtzEGxkbHOmPOTM24GVFejkE2udA9PVc+iNQ+wHrs7z29bvpOgjuAPwHgrwH40r//5wD+C1X9JRH5rwH8MQB/yv/+mqr+EyLyhzzfH/y2wq8JzLXUYirH5Wrii+fmDLiubLVkV2SWyT4jt2mVkwFd7AVnqOHlS2Y2UiBimQ+H4hAcFm6TTM2qi6Ny4eQj2MWunzXtxbVw4j7T/5HQjNCZU0WPM3P3LTZ0Q5vwerjf07WlroEEYCMFovRekO4ENH3bI2wUA2Lx06nPZdo0ej7fzNt7PA3cdELeKz777FO8v97wD79+j9hrzcAjcBAPiCICJ9hJ/S2UjzbX1zwGVa2zuUL7I8QnLRehUuO/opBt4OnNIz5/2PB0GRWmWKf5c8jExMQcNkMwNsgYnQGTuYRBWRGOiD2rWWd2WJyAaScaAAeVOuveAGhFSeqbHn6S/sqKrMsLUcsqBMRLOe98nrBAWp2KQSLnuag8fFdSACHA8wFpFqdsnuS/+SQFqI7o6nUxmKeQ6w9DCFTefA+hMDxoRTC2gcs2sF3sGuUEdyh0B3Q3K+Ku6sdbbVImtyI+FK3jXqyja3zlDun/f6TvHLiLyM8B+FcB/KcA/n0xFfJfBvBvepY/A+A/hoH77/fPAPDnAPxXIiL6gdmpIK07HN8gWN9IU9cWy1fJxIdkfOnNrqWR9sUUTLLipie/nFVXam+x0IM5M9gAyZKKc5RUwHq5rVI3w8dtMaWuuad8Vdf4E9XVmLGDUAYbUbOEJCjCLy85gKHkmfP9tkPclD/UQMR4LtHGn+kkocrbGmZ7EXt3VwvDehHBfgMwFe/fvjUL8aNg7vbukBJKQvsO0zN719VHqb/JeI9cX+VAoNyjPqKTtj/11CbRwxB8/uYBT2PDJYHXPByGDAzsNm62iY4xNmxbNZWBUSCtnhlzs3XPUGbO6VtHauZ5EWwykkbZZK35UJYN+tEzaLxBdF0xPaYwoH3txdQ+oDsJW0JV8t/FymLj24WLTvdodrSduuP15ag2KS1ySNu+0eW/bEb2q0s/U8uvByjLnmWyAEXbZeDiF0BtW/TDB0LMh2e4VWamKarFASxh9ESAaaRXrbafT9V7esX0nQN3AP8lgP8AwBf+/WcAfF9Vb/79lwH8Vv/8WwH8bQBQ1ZuI/MDz/79coIj8PICfB4Cnx4diLLH4JWR4sTuc615WAyDPG9oCDgxD63wuQt/W2uOVAiljTCQvV1XFvyAQzMVE6RqJzsIZLd0DEAu2E28oIDs6w/N+RGVhhl61KNH+XpjEU7AgnSS8rIPphOR/EcFnTw94eNygYmfUf/TNe4t456AaYdATDODafHoJ1UaCwhzLIHEEzMz8Aotb8P5HiqenDdsnFwvEsivGg5iAMGD+ACHMiQMwAb7PkyJT8Uksg12kWujUwD9jjvtn2Ljbe5oFKMrr/zKAzy7AgwtQ0Im4IjQEG/F5IC4wbsSkQwjy0aW2usA5Y+64VSjkQnWnUpMxHBeDADGmPl6b5BwSceHG10XIkp1cUvQUnkNHgibYt1nNgsSqIwsJE1VblB5Xvi5v+PTXrIshjIGN1ySvuepVuJZSN5yGKQxJtWXtdqwVkoULkjdAL4Jx2XB5EgP34cKEKjB3UwbcGzj5VCydkFeEaCT2XVDzOtetcptwSHdl/vXTdwrcReT3Afh7qvpXReR3v1a5qvqLAH4RAL74/FONaGgCpCd6wtMgEzKkXwcqlU8mkCqsrqxKcyHFIs/Y84IWipMl6Mb0hLUdTcZdkrUWQ/TVvErduXdXrUIARQMtWtji3GzE2TQBLE69UsHOzAW4XAa2zTTtfU7T5BWYm+Dt8xXP1xsgbtq/7cSgxYWG0soV6rePRfOLImH1iHHp7Rbc9h3y3kzbc/rZ+5vgQTbzzh9izDBPKIS7U0pUPMwHmWjlbZL/vJQI2JVH1oTDZKY0jgPAgwCXHMmyBO1zYodCh3nEpwYrsYVTDLw11uedRU0kfU6FtEZYXP4wJYvgBj4Xns0pT0MiQHl9LILqqkmf0IxBJZ8tVSSks708rW9rh31MFeQfwfcDaOaKtp7jlhZIHptdJWkvmeS24g1SczyFOlpOChcIUlBywXMMYBuQy8C4DFwkLAUTuk/sN3fs1WMfOi+QJnc2ASPP3Z8UUuT68FS/p58ofafAHcC/COBfE5HfC+ANbM/9FwB8T0Qurr3/HIBf8fy/AuC3AfhlEbkA+ArmWPfBlEE1FNA9NBD/bSK/sKBt/MTjuMOPHqWWA1rkgRTORKJcLiwWDGW1j15vOLN4vqloGpIgIRrQfumE0L+rprDIKc4cnY0FoxkWtnUFvNNFTlsQEQUufp9z4lnrhEAWkUJO3XWfHgNaNMtjd745Kb4XXOZ5pOAhmNguo7RLhFxTFhUbp+rUWVx0Hp92x/dCgw8gAtoInHouVYX5q+cbADYNZ8yaC7or9n36UcZ6R91kL5KQlk1T/6du/ZtNGJxhtYFnVs0TCnY97nknc6ozMCWlNIVSWedP/K4+d6WXr+2vnXxIAYHOggqXQxc+rdVUnKkajw7r9EwaRtd61HMhr+2VHyURbqivsxKCFjmEaBQLLMDW+iVjQLaBMUY6OAJ23ew+DdxvGiGIln5wVbo0lQig9F3WAk6WyT29Tjo9kfOPalLV/0hVf05VfzuAPwTgL6nqvwXgfwHwr3u2Pwzgz/vnv+Df4b//pQ/tt1sSDL9/HBAPAmLnoPdd6V53cQbiTDQBdiAi2KhHDNFd7ayxaz+x0ILpDrjmmbdKBNOQttijrhC7WRaIPdOSxMUXPzMgY7MOg2mmr64HNynhZdLfHSB/hPqPS67/CigKJEwwuDgzivPy6VVMjHbEvrpasJ3JJmM/8yP+XQcLAjBgQ5i0bfviYTOPYgSJ3fwft6fVfz4WKZxFT6rDYcYOYYK2ZEuAiNeScPGfmqk0hcQQ2wp+Gx3V82kfzwCFA1DD51fMzV2zPaax+gxQu/7W5rMFatonsO+CfYfP9ZlCw4QCQ7FtFr7WTj6ah3b6l4jQ7CEBMP/zPmTeovOZoNB9zdsPOcXjhrW8aQ1hT2gTO+ttPgKp6cfvnJ+GZ617/c5rVGOoBVMl108IyYnL9F8WSyiayzjfD34j7pnKnvGKcVGMMbGNiU0U283m2U0V7mpS8zSsNV1iOlKbTYHZTxo7yEsjdE+vkL5rmvtL6U8C+CUR+U8A/B8A/rQ//9MA/jsR+RsAfhUmEHxrigA0qVHG5Jdw3nLvcGdOoVuGaZhf2rOM1F2KqWUdAtmGMZ+9NKRgb/FemqwFHiVveLmzMvkeWsRVh5oEOElpQNM4gt1Qon1c7j8CNLLHIMne3rF9Vc33rHVengSjKgIH0xHadN2hkNueeaPOS4ByvC0WWyCsFDM5FlGNNMDhtEnNLQFYC2QOPghElvwHXStb/zaaJWma4lZl94cN5zSfIGKkD8qsEnuhZtrPfVRVqEzb1pjhC9LHK8LwlrYo7sBllYuyN4MJggLBNrYUiCDVz7iHIE0nQs0XeIhaiqq40nmlLX03TV4bfSGAjDgCWRfYzJSyKh+fhCiMokw2AVEBcjTBU8mCcbrXjGVM4UKUFK3jX2mf+1s2LYvhnAJnAHrOW6RVbJ+zTjrsinn1y4J0ceLVWp/ZPhIqqhOL4I++ddTDKZ8Q5p5+6vSdBXdV/csA/rJ//psA/vmTPO8A/Bu/nnKNv9fKDLwMEAiP0eGx0uknYGpzYKqFUwzcFiyzBIFsG7Ztw9CJm95Ayy/fycL8tcvDwOVhw+3mEckyn5qTzKwFHMyVMYMjih3O0rkGaE441s9ihckhWqoTANoYRmkJkt8Bi2hnF8FoqaIODNaSEmQwLT9gloNB2ROcoiwgw+tma9WumTWN1craVTFvfh+6al60QU3vnHXlX9/2/YUU3W2gcEQH89mAj5sTYojURWwT0Jtp3iHPpKNl1JVnnCdURsotM0zsSpagqFoiul0gtwtGgAdLGW0rJfoytATIMiAJjadmmeVVcaTfCpYrWXPVqNWpThMDurKqeCeQ8prQ1kDgP3ddl1piLtUrLbEcsr4qKdUsbcZ5UmpP5BPEGnXHVun1yIQdc9sVtx2Qd5rjgJtiPk9cw+oSlsWggPZ6mfatD9EQk3VIOSAfjnv6aOk7C+4fMw27J7MWizMM3d3UC8l73O035IrZAcjYMEadBTfe7NrlPtte4HjYsD1sGPvEfr2ZKV9OGAJrQ562i7HzfY7c007ZmtUM/pj9qTIRSzXAnP7mNgHimZQw0+H+dLHn/qMij6MV43UAACAASURBVLUlKEswtYlNpc71i9B+ZoMrABWdju96D5M6vJ2hrVh0tQGdE2OY05FCLZSqlxkxAxoH/RDCrH08e8hNJiBsj0IoWctqXLzreAFcqnF+2bVVGqMhabm1sdpNkw9hKcz9sYWgDOJkVQnBNI8xwqxLQlHpsr0Otnn+n9oPoB2tPKDFQkA9exg0DQBUQKaUb4U3Yj2r7z3IuuKcN82c6kDkBTXtpbEP0FtAnB3sqots0SuQfVGUWSaEUJ+D8OITYL8KrsPGe3chHrtCb9NCKsdlQUCO+SrgHMQOHh7qT+uotyEEzxcll3v6idMd3F85yRC8+eoTCwAiFWNK9x3X224AvE9InFHTcDKy1Z5S95xpNrNFaQwkfZEAjMcNT2+ecJmK92+vuN0UU+jiklVj0GJD82p3sD9sG8Y2sO8Tc58lpbsKVefs413155KLmE1wpsT6toHv70aMaqKSXerinKeYRPcgbm/E+h9IZ7UA5zibbnfTZ5Be2s+H07a2M2K3vhikIs2WBPIC2+ffnuyI422fGUMe4ue1PcyucjvRWHLRnwA6qk4g+BCDI2mtCWzMaOOr5uMigsQFN5Zh301rtytibCtDAYtUFm2XKk/h8Qd8aynmadz3Bmg6YwGSgpXAwdwnRM7vPEVi/6j2eAQFnK7lCSCraZ0Je0JoPfye0gGBczR50rWu9HcVNpba5IXPZ2DVQhbwh+V7DjUjvNJpFZZSqFsMppAYt6g7BPASSHROzCtsT324YgHkTUgG7FXvpPoABvu1ozjbBcnnWcQpk7qn10p3cH/lNIbgZz5/spuyBiDjgu3xAds2MMeGG4Dn2473z7tfNnKDPl8xr1fcnq+4Xm/Yb7sd2SLzcFyLGUtzbAOXMSDXK27PN1xvdpQJoCNusZikrx9Vix9+ve14vAgeLhu2TTBvgvc3C0nVzOTM67RO6PEWwkScCzcT4MOQ9EDfUeWkade/pyt+oh0yjnWZ4IuZKQrMozN19LBAOhQKAR27Cy7FPF0ktY/yByhwMxP0xPXZjgcFUGVt2p2N7LWFs3eJIf8ceFrQhznghwBf++cym2oHeK9sTsV+Ux9/tRsLnX471Pa9BzB8hO1dPwsvwFTJsWc8inCyOhLOLQ85OHZnOSJJyR5QREwDCt4iKZb5vj13vhNT2099LdRPRuTwMYzsCvXjfARMQuNBVpmzIent4nJPUrbV+0Pe7gfBcGSr2xrhCV7TK+hRDe0CY2+5YhrA30B09v7PcK7s/WAafxCTg3Zr1TTmGn1/kVD39NOkO7i/chrbhs8++wLXb95DrzsgCtlv5uG97bY3ftmALz/BeNigMGeW513x7nrD+3fP2J9vmPuOPUB2lneqOJDZTVu2CK9vJh611ncx0WBwkp8UwJdffIHdncAetoHLwwVyGdAJ3K43PO83i9NOfnahwon6ZR9xo5dXqNC8U32InU+HlMPbesggwT0X9nFf8KsvP8cQwbv3z41Nk1/bUm5xjgCWQRyGsVJJcOIM4uVXFLuZZmWmrV2EY18+//xT/IZ3X2HOmQwtYI19rHIEGrKzKskq2LFtrZtYtCCwMBYgj9pPF+CTTz7B9bOvAFU87Ca0hAA5VT3EqEAxbCtDAMWOdHLDaMfbQrBJJ7pBJwhsEHLcmLeHt3vrZ+DRNvDF55/hZ37jb4B7PLrlaNZtSpAjjfi8JtoZhQVurdgHMYuMiG27QHfoLUKsWl6zOsBOV0CSTqqKN09P+PKLz3C9PmfZ0qvpoEjCH2N0tCdeEAkBCW0utDH+ABCy0AsBvvzqCxMa+OpH0BbUiNMfQHnQI4XFA7CTL05vRkdyOljQfm5zNspQxaefffpyp+7pJ0p3cH/ltN92/P2//wNbGMTAwzw5NruoYdsGxgbzKh4Dis0WkgDYBiyCxF7ma63lM1Wx34CrFiMCbC/fHI6yWgQo50LyZ9tFIJsFE7ldr8DVuI66TX4IgAtpK456yx0ipI0rLhv94My7jOQINP5WGoZ2+LRt+JHY0bvUKAR2nCe70rWzDYKL30CWR5TQGdSB50hEwotTiIIK3bKgyGrGAPB4GRa17jbBewrhpsg0Y09ji3A22m9hGn6JSkMWgSFOFwQQaJGYtSwAeHub+LtfvzuWr0RPKCie8JLq+Yms4U586jcISpuzK3M/A3YIsM0N+9xx2282r/YsqhzUm7ZKgKFBXxdl09qlTTAT+IlCKET37FezLhAY6qwyzZvGrjR+f7uZ4CkVI/9otZFUTqea/0KXumhJRBsjguBCHImzDkI1sTBJVcfx2tuugO4elpkHgwThVsi3p7aMF0klqRBLnoUYt0Kcyq4/Bl+4p19fuoP7K6e5T3z/175fXt4A4ozstlkc540CRogCcKctAMBFoNvEfL7i+Uc3XPfSxIKLxeIKs3YFvCAO1lIXDkRMQ/r+D36QmdOMG9p6rn+pd/y/UaWeMlvQ58bsVrRZmhgp9KQ3T4/4wQ9+iG/evnPPX/dYzwZ24QYANtjNZ3EaAUOwMlytF1OT3EQw4m5zGUgDtHes1dPGQ/HpZ5/h+9//Aa7X3bUkIoE7kAFopyUA8VjeJfg4JBEtFboIL3yWHjDgaTsbJL6wBi8CfO97X+FXf+0HHU34rx7HUfnTMn49Tk98qZd76QfYO02XywVPT0/41V/9fgxOa0yVWcAt3sfy6HYzPjUkLQjkOJkCA3pV0gZw7Y19fP7kE7zZJ/7Br/3ALQCygCc1W8RDFCjmDrjEnlnX1bCNWGMMmrYgK5x11VNOt1FhWGNM2J3Y8f3v/zDz8Eh865gcFnatuYbHWvdonLEflifOyPQbf+Nv+LaW3NOvM93B/ZWTwo6XBDNUmCSu7sotqhC/13wKgClQv6BBRDDmwOUTBR4V81Fwexcm+c5cWXrW9k+hu8SvlM8xI1hFfy6wo1JKPLUE/BfVyZeAPVrKDnINi7ls+jH23CdKm8quSNCV+uXtNUEJdsPoKNNgbUpwA2KMqKMVwL8I1fpNSECdtbFQxDlsFoCimNyXB40Qn+mPGmTps0S+akGdxUAy+upbcn4bX+W5soI6JZ4jPiE+pEt9oCQCmfhDwPdtClr7/Vh6QDfnKaevgHbyZi9yrM1qDmeuVKYV6igQUo9pMEIBP0O00FFDi45ASao1T2JYuL3VWKEKFqqw0LPW7Wsh16UHMDpk1RfG8ceRwqL90YGcc8f2HuaDnvz8bfPinn7d6Q7ur548GIp9bJNedce8SpvrDbPHwAM2PF6AcUGaltPE7y8kI2vriBgD63CLS2sAETuEGTj4v11ROjxLhrAqNx9gymcWt4Y1J8noEiKIcyt322+CTWBcCCzEaEc4YJEDXlxFG3uHqSmPoI0XUPywAWnvK5/WddNsChT0gsblOAJWd6M37XKPBORF1Ul61LYnb1v7Ue2F1trC3ArTMpPUPF04/QEM8eOlD2JD1LPMqbMvVT+3ZAXddR15Hur3oU16rDKd0k/a38ql8gQ+LRNIY7yjAhI1puZajHkamu6UEv7a+XAv89Tmkd1cQb46V64iadAPcSMFkzYTToSEs3rbrE+AX9rgX160tst58ff0eukO7h8hrTd9JZBOi0wXO5fJmIh3iQD7vkEuW9xOgojs3G6DajVKX+jErNe11Y+ZCfOQar9rHS+tdVkfxOcsmxihLo/OpHZOi1OPMUJpjmFdA9CkWzhyqYbXb5jEHRSpKSUhCIR8Bfpe4to2tIw5flmDFH9lgIxxazQtBsiOhBnAhIhMMJ9l2rG+Imo3kZJw4fJEMnoBTTgqkLGL5uMHZLZT5tx4O4PMWflLQRp0CwFJ1swrkK3DFDCs1f8fM0XdJ1MUhw7FgGo9nbAz/kfBwB1iJ4dJhsdk8HXr5XDcKJod1OXqdNC0xnNF2LJeHJzblvy8bNtY9Ul5pMmLzOV0cHrzfhxB4p5+qnQH94+QWF+INTJ5xSrlCgZmqIR9Km5XAJcH6LhgjHeYuHYGLAJRvvqt1y9ctkrn5cFdqOq1jLwWdUmyLloCscQk/m0lCBXJQsVL67rt4cUeM8LEq9kAgWmuFdY0dyiTgSpQDlnqjHYAAxZqt464MSOMOpd2L572gbF9v9wAowXQ0WVmJAPtoGVdoDGaC/AI/KrNyhvbFauHcjiARZ1Hzf2c/iFMHgjwbYn281lIOwgQsQYOU6okgqSRZ8gxXBqcS+swUMscX35eu9VlUz1varSXAFN9TmnvdfZHF419SATP8VkqfpyTLH7lDIgcdAUWoJbeccXhthAFXdd61teFPAfZcs3cHlDgK2nTu2i1jFfN6sU35Z5ePd3B/WOknNDuEBWOTfmPfxC7XnP4Z1vkE88KzLEBnzzgsl+hVwv6oieLNNOyUJIpiDF3ltMj8Ezqk8Ro4pBv9zHnkr2URVNs+V78bWmyOyG1kqna8EIWD36SxurIozAfBnVPd0gFR5EC+uijBLBrlO3MdozcAlllndjHTaAP7cwHs2Gpe9wnCSbaoAt5Iiqdw86thVH0aAx8g5t0vQ15VzgK4D3q26R9/PYn5ocYcPAeeHiTf0jTqi0gAiDgxfnYZQzKuaIwAX+jxVJGUqt5w8H9UaQXZERqWzSH/pw8PggIJ7+zuJfhdwH3nfE5KnEUj/xlnMYioy54AmzOOQ9Qv4kvLnkJwbzRQUvQk8OYxaCEf4nPO13yfWCcm/Dywjs268kRJAeiP2ofkz/JQUB7aZju6adLd3D/SKnMu7WyGGB5QqdGAltU+z6BZ/P6voyBuQ3cZl+hh6sTX2CK8SwYZyz6w8LNxey5JbQHDq2BZB5rJS/w0IPQUb8RjL6wstW93bFJAqWIZkzw6IKZ3gfCUC5NxSmtIp0SRfxiE9/5CK19dThQddN3PSLoXDokxVQ9X4bDpaNm4l8VHgY3AqeQtYFBLgUIAWT4ZkkIFtqHQsUECPbPsH14RfpuLIPRzPSCkiBZCP22JNTOPqtplM/hlAFVXszX38lPy/ZCCGFdsKgP6eAWP+iSifqsJxJA7i9LCX0TLmBOZGhIu3gnhoBoHoicB8uRf4cKFNO27Bzgo7VKtvLen0LJVQCwN70ErZgMkWHFewC1JYL1h7NZQAKVRp0dtdPBNT6z4OAMh/tzT6+b7uD+EVJbOFJLku9XERRA578Rd32fkP0ZoorL4w37+2GXKuddoCgwzjW2rsCTVZqcNkT/FxpPXjK6qg6E1ifrufq5NONDa7czDjkw3jyW57SMcKRAhH21M80TwJgOyHzcyweE8SpYZV0fCiZt+wtEpK5OiwKLaFVwL02mncAcNRKDjQs9mOcHp+S495qCh9cSgBxzKwYiaR0dEYh6pMNJwt0J8YvJMvFjDvS8mfXkc/X3jI5LpdE/IC/r4cgC2lS8pfGCpQw++8F9eTlFV1nAKDvN0e+E5enQoNk/P45phg9F0juJEtYEQlFvhF2CJNhnASYL1lZEiq8u6FbDleptAqhG25T6eJwHbMlpicl+Qs6MPaGxHrjFmqGiqTn8dq/7nl413cH91ROBBbAsFl96vqCZqcTfffc7xJ+fsT3teHojmI8bxhV0uQtS+l4Xo9I/giVUK8isDSEmJMl/cuU3bY5ZyKruHPvJLLlplyfCRpo10zfAmWsjYGk/AXFpnRWxqGo78jiaZZ1QHVhNubH/WV2QNNen1BW1ZCdG1vui969aSL/0CSBGOsIPIEjq4D9kYGy9b0WaanNKOKuUpLUPm5DI2pCqe2DLos0xAJJcuLQ7ydLfqOpB40QDbHNMDs09TRoCjH3OnQsNUWYRMFrlZw1jf4sPNEB7Z9k+xTP8uMaqPXncLBpzQqu4g6GkUxpLrL9RpW3NVb+8UG8DAf85KfL54iVymu/kl3P6JWG895rig42ZRBvDDiWdqAh68xy/p9dOd3B/9RTgcvJcdNk3r09pMYaZauc+gatgXDZs28B2Ucx9QPe9ScmnLSCGGxI7S9/JlLTXv3QhGdvphSZc5tKMaFn7u4DXUTZQlDZKbb3Azq0rcptPAPdVEEzXgnQA2C1QCIbRMWO/pWmUODILI4oyvyu3VfK/NmYL0oSodgr8avEMxohyNcegtHtjdQkkdGIgL6TJdvJBJGLyUkKUZp97DHV7XIINm5nXMWFg/9YkC7XIMfE418+YvXqf+x1pOc9D+qC5KGuxRY7D45MGkwhQsMRC3yrCriCvcdLirHqa/H320BivW0CAxWZw50ltdKKY8yGRC7140r+1TdmflX4gOecF+r3EaVIAFKKP1DHb1bmTGcJhXpwymXv6adId3D9KKoAPLdmeCv2KzuekL6K5K/Znxe1JgMcNDzuA/YabyvEcfeOXQo8WZuV77UdwCACWBtrH5UYtXASA+LkzAk2tkoG+VQo07sCLXqAY2wYVu3K1NQPwuN9eloZDEvzYkWBCMTbabExnuBKNmvOZkHARPJS8j8vbfiGJtxaiHXyy6ImpgMRFgMWpy8N/emx+p0dsO0QxCr9/nPdOmfRaYxGCQBkjugPdau6N+dcf2Jfcp9clXxtTahD1rX2nDP1Rn8TN7yFyrMhKDQ3dsB61Wba0S5IeHVx9/lCvw6OsX1VDc0VXYQqnY1/kknq3OpWZk7bS82haAatfzVHwBHX18GHp75LWk26H8WnzPRaGNbTiPgnl9RUv9QqXnILtC0N1T6+T7uD+MdLkhVtoUcD2siScr/kNXvsVwBi4bAq9SO6dhsk1X9BijG2vUGO/TVsdazsC/CstYU/5+QpwrS7P5WvXMFwxDqo7VaxHE11o0qIbhg7bf41neYTNGiZq9ArLiMIurJlT07OcK2RT8plTUj7JMHOaoFmlaI0vaHRPtHoiT1YwpdoiXkfDVY1apG7JIwKZhzx8DK3fgzJaLn8gQFwc0vbyuaELGFVrBMfNembMJLAugG5H9Lrc8FLKpRL3FTfwOplsh8J4nLVn4j+xFmUUjWjzfS3Wl1aWeXB0dDq5zIQYM14vLDD29vV6fBPJ/TuqL0q1BelbW5dGd4FN18qzfllfeOFvrOPImgKvRKx9m89xcmVtzUEAbC29p4+V7uD+EdJh6b6gTeTpJl5QCF5jpvn9erM7tqfdN70XxvWyBK5njQPT4fpCuZmrRnaSkiecZUgt6Sxp4kG0ZbI0/xIzOhYD7Iptdq0oAoDE1atD/LpSQTLqKeaDKHvdM54/5xf7oK5JGIM+OUI1i0/mH1kpLO1m0D7i4lf2RgF1MUi7FnW4HOH28hDiIkeDLNaKwhteKp8FTikla50HB6ILPYpIgNGRLhsdPqv3o36Shv/B3Bn2yhmMBke1+6ugCmCadyVU+6cDkK7Jz5j7pMgtDV4HCnN2pDFrpOC8wkLbqVtfzScKT9yF2co9gm5Rhy9abtuENDlqkbMaaWLurEDe5hDgARGCeDVPde2rdyC7FPWvPI7X2SnALwF/7jj/6ukO7h8pJRNYpfWUhKWva/UloAYYE4Lr3IH3iovueUFbeqYuXEZy0fkiixjs5HkdGSPiW5n3z1dWaSNYFms2eMGIrt0UyHV/csHCbDzPYd9fAewTQ+uQf7wP17xKORF61T6bMk/tlDitJPCN+aWvq/Ch1Q5d2izkXLcAFbP3GWUQU5UYJ6GcIhVhF5U/uGt4Y0cM9HSma/T3OtyBbvdjdnb7nI8NOXO2d1fpKpByEVP4JdZQa+xqUoSwwQCf9GloFIKf3ZjITVhli9O0ECD8F1i44k8CKXBfLyxvEssHquI5rv33ihzoT5bTKfVxATsRyLBx4z4po/1q3g/54zAR+HfteWhcUhgClxFzpH5M4doFahaw4nAKT4EuPaz0j2clON3T66c7uH+spC9MXYUzn+OUDsagMNPcdQLzBqgo5MEYczpIBcAHlocF1rk/77kGyLeK2Eyfi35hh4lLmqe7Ij8DOudtxYUAchDRlz1fZjrENO2vXaMqoOPBrgpNhAbB+/EEmAGMccxqWI0CyYAxvM+eX/1LCFINhxD5+/jxcb16QLR0tBOWlqjPJajEtoF6dzQFBBHBZRt5cUziksY48CkChQ5j6sPNpzVwLH70Jq94Fgy+9XYRig7J25qWIrSpRYKU97GBfj+WeKiQnwgDq/YfG2L1x7l9JSHsRBFehhwM6yFPEdFq3vQKeGVLG8scZJ5k8UN4N+bck3ayQ9XlEZIbJJp8IoSw9aVuo6e8Rf6eqI7jz9HuZaoHPda2NIDH4b16Nz/d0yumO7h/jKTEdHjOCvpDZkDL+ldx89sE5g7MS0FiXhLC3CXWHe2pFqjTPvOyiAJqG7LmOjWGE8y5d4UQifuqxzXe8F00j/ActI2FZuHMpQRQ4aAYe/lmCg2HpK6/J3BGuRPQkXwZAfT2Kh3dUqd1toPAZ0SxUpkbtwppy/+sVhFFWhMUiM33BKoZH+J8ssIBGhgXweXBXpy7YlObB3sCTgkG2ya4jA0PY2BA8OwOiYdwCN5k3q4IkE3x78fiu4puAaqAO9IfE4cvkK2pIG0adKg8goM19Tin7UM8j53hKrhG/kCN3F7hJrcPC1ANiS0HzmboqXwE0n/U0ctIII45IYG82kJbaDiPTikhYlmYbTnp0uaYkwdCLakJTZSVhLE67RFC6Fqgri8ffqvwxmeqzj39tOkO7h8jaf6D/qlSglTkWOZ+HM9SZ+TzKnY3udS+3KHS4MOklTdtAxEW0xqQwnaELE1cCs4TD0jyyL+BnPUszHOdZxQXS9+2QtD2dRUMVC1y16YRKlby+XTBhd2NGp9LrDmhvgSYx/+opcShSUyy3o6lbwQSwfwCMCK8LcTANzA/tmPydFduG0iSPMSVUNwUFoZXh6TXfcphanfYl/xS5cnYMLeBHRO398v2zMp7eRxaJ9fxP2HEMR9cuBQmotDvlD8x54A0/pcmEs+pg6DAzWzlLPno8UFxzj5py3pWlyzNrKVyhpRlaVuLKoGK4xIAUyW30gLYOVAO85b0zaD5EzmDh5yD5pnksqQQOPgVEnqYz5wVezZJZPlJD1/u6TXTHdw/Qkqz+eEHHLSWxtsawzUNd4piTsE+7ay0SF0GMpdKZlSi65qRw3qWdaVRM6x0Vi28rbKy9mNHpakpx6ytTmLenGLRC6yvW4Cx+G2tgNEl/AlUW3WCCuv+weOzXmZ5rPs/J2PXTbJReP0XbZEUHLSYNbUH0K5HarLjRoA6N+/vqwBX4Ca+NxtWCkGGuTUt30dcAejEPicmLMDOh0gR9Z6OU37+kIa1SGY5uKDGSQM1ztO09xR8elPO4bPN1qV9p3CCiHXQ14D2LIzw0RgqkoWSmjtF+1rqsnbVQbIylWhBU4z21UpwZL+VohP3l++xCL+ammuehyhzIGzKDwcOUn2Fnm7hVCuWCqiR6zw/fX5Pr5Lu4P4RkohJ1aujSwW+OAH2THx21PLvAGT3S1Ti3nFnaeqMU3NhS9UXhbeFSCFPoz5vdChohxY2vmY3kCl95827trC7WzP1qT04fqZHmyi2aKz3fcKOuc1SbdGYkDJvqTY0ZQ0VaEPWtkp/r7cp6iPzsdbz7L/ANC/uG41Lk2/S4UmoL9LuD4iu7p5tbPZf7LMzbVXdVE9bNENGmYdfQsqTx4HL+f1bTag04XP6uFOdUoBSdQHW28+Afqxc1mlWcqUU5HxIehHf32hzQKPePjN5bE6qpAdyIiTXmmEZhn1Xy+2iC0S5hjXaqV3wzH1+rYJ4LmptY6VPRjOlLXN86VjvW+8sWxLT+XMtp8tHx3Q6PnQq455eNd3B/SOk8obu2gcH6CgmEyu7sefMpfBjXVCMKQZ0fmxFZ+QWpEqkHZekcnjSYhoOYhrmZIm2y4rLxYR8Rafm0TpOGflz9t/qW7Wg3rrWcAzRDFajrvLuKK3djr8z+nCbJQGCQXeQxm70WjkdXmBEOPQ5zrsP8erc3iuw9oqDWFxrwwiiKOCzf8ox0EilprFr1mTlq0L9FrJBZgBV2PNZUddScCFg10LFqo+HyukSwx1zCL4NcnDESmGoW1BKaIzgPgV8DWGpLcXnXTDhAimgUAFLAN4BZdeWeL1O9RmPDPRX61F1DHn8a+1fthEkZETDwv+BaPoyiHk9yscfNW/6GymV2aTlIhQwPtDZCCbUYj2ExWStTdEEkzY3EVyiLfr02m8naNbGHD+eyXs93YH9o6Q7uH+EJLRqwgm2gTxYmGYwhDNVWRaKeUwPKMYwDX5373hVl9BzA72L28lH1/1W0ggGnb86A/bwBO+SOq/sMPUxcp1rCQk26M8Y8/kXGZsfjzKtZs6JfSrmrTScFveEGVmcVaZWtj1yELi+BOovMDJ29oMiPfFBNdgFN1VnMPgZ++wampW3wC0zccpgapSBAwCa2XYQnTUZdrRFw5eiEYjG6LTDvZ/JlHUV1dY89WurBroITyfvShAyjheeADunkneaFapDEcuQ2houS3ttHPvkW2UR+kpl1A9p9ZI6080WlUWeBXOB3uaYUFbG8C2YqqSLT+0OeJDwBjUHUs1vueYP47gQr9ZP8aGwbhTtWjPonfa4ZRXtQY2gtI7uCP/q6Q7uHyG9dExKagWjc46mq9CLi8YhE7K71qlrKFmUdhDv0uuQVD5ao1IpQWUquCiO1CChrWRudTmLta4EfnkbDmduXfiZxLTjw9g2yLbhBsVVFdd9Yr9q5Ql8U/jVr1IMX+CR7XyvUrrg0vZVScBiM/E5moEyWJ4hA1PC60FhyMrijiRFlV43R0Fn5mQ18a4ZMEbMgphZGs+R2lTHJnfIC6GANLsjd1/magTxUaQWCXr19C1CQv63TRMCljNShkQQ9DELQQlcNJX6vI4CF2fHLssaochJPWs5bOwvDUuafxv4UOeEF1qaxRdpAAzvSyFx9BN+7p+8+Dowx21yQa+o0lutgMwJ3bYuGGIZm6UJIWulQNoYC/EFoputFWndXIE8AjmwIKMoP5p7et10B/ePkRIcWE5fmcOKgMS+Yu0IchuhJwAAIABJREFULSpVqF+MomJ7rzeC9tUzliX1ALrkwRKahTOZZPjElHmbYFnLnRku6k0UtdKjqq6TSWPJrPSfU28MD+yhgnc3xe2q0J2OK6kxhgzuspjYeZ+5+u+ASYLMYRhWMGqMLP5HP2Z4XBszjUpVuEtFsqAvDUyCaYwPUBHBpPoQ9c7QzKMtPKW06lMFxcjXPjiHgXohAkNWLAk+x0KKmxslSNiT3rwuMCrCAjHONOiTdILDLfOZIFFnGVjjbaPYhJOcsKs9XRfh5UONleMUf2lt2GAJYv/JYjF0cSTb5xYhu1TJBrqMVB5+GYqRE4/GPe4uOJNXQr7gZc0OqysvoBcPlrdo0yQyUsQ/eBvPrXb39NOm7xy4i8j3APw3AP5p2Bz7dwD8dQD/A4DfDuBvAfgDqvprYivrFwD8XgDfAPgjqvq/f1sdyibpQuouDh//IL8JDmZCK8nDz8Kc7CZLymTOO0RT46aQlmpm4H42PJx6pufXk8Ws84SPvbQ4KePBYJFZKqhPmv0clRSKfQK3acA+d8WBISemSievVJ2hER8Am8B/7RT7LLEcc9Zp1WFX9R47R5YMB+xJ9MU6X6qtgQx1lW1NGCX6rO8PZAA2RLCzBDMSLIoG4cVOz8/6ysQ7zCt67PRuP7exJyRaUKPARBuwrWgd3QhZrSSnytpk4/guVUcsmfOx9fnigYDyTgcAHE2x/F18Hp8BvB6/ps8HN1iDLsshTV/na9FxYiJvTaS6BBX0KeJBREfTGXft9gngp+8Qiph6yMjCzrqI+K+/3/yLuIR7es00vj3LP3LpFwD8z6r6TwH4XQD+GoD/EMBfVNXfCeAv+ncA+FcA/E7/7+cB/Kkfp4IQ+DOtDJEyGHOyM9GbeLCS+A/EBMTM1rsAt0UbaFVoFR9m3niiITQ4d5ExrG7fxx8UtS3MkV1n0QQabb9qq1hWAkQ2Z5DBKO3M9+LsE0DsX56n4vlmN+TJTd2Ere6x7+eBUQ52pVNb3jGIycV/3pZkeClQhODg7/Nn8G/G9MQJrgD2fWLeFPs+42r3BF0Rc3yLtgyp8TH6FPO150GVvCXIY8XbzXFzKjcz389QwxJ3xYvdS7ANyPaBpR51EP34fwdUlQIfFppi7tCUO8zPdUpE12f+l2Ke067KXwsISDlzVMvqJfGydbeOUVZjyldi7YbWi+0V6f3I5+qmZj/RQYLHgQDxldqoGFAZiOt+D+tDqH1CzML/U5fu6tZEof+ICEkDPbSHiVh0WAWy3p3O1Pp/M/5jeq1L7p5eNX2nNHcR+QrAvwTgjwCAqj4DeBaR3w/gd3u2PwPgLwP4kwB+P4A/qyZq/q8i8j0R+S2q+v98uCaf6iGI+6Ocv851JD9TG1HAnaazeCDICGtQ/5umWaCpDQHs+YgZUzVGgjFEHRPADEdA84pPrSCK+xaR+6ARxCurmqD8SJrmGIzzqor3+7T77UPjkP6eAnk8Lxn20sQQGuxBaBABIDRQCiybs0a3SV+o4BREnEmlH4SUoNS6LupClQN11DetDxaJj52xoqIKYsJ9TACLnF6f+VsK0vTwkka19jPII5VnHc+TIQ9iVBvo/QQtNt2g5lVbI+uxyjVFd7TnySNm541jDM9CAmRO6RMvTf8tt1FoHoHWHnW3+loyb/XvXFDJlRk/JjAWYLLAwltKRmrhwgC4863v2StJ3OKCXBN+EJaSVSKrgdT2/AyPaQDWfkrROpdaNfVulv8I6TsF7gB+B4C/D+C/FZHfBeCvAvgTAH4TAfbfAfCb/PNvBfC36f1f9mcN3EXk52GaPZ4eH9pMjahtLJxK46Ad9BM8mTnGMwcPGdqeRXGLHABZll/wEGuPLXgN6Z4yNSYSQgSXgWJSHBe+s+4X+exBA1iBmmvarwq9Taju2aaRjC1UC63od9TuIpoUaDuRNWujC9IX9+wEhBcYT8UF8BCk4lsaGWgGfUA8t93IJ6ZdudAgG5owEOJNAZmYtzGp7EoSQFyvanUR5LgwwMcFc2wjC7WTgWilRUHDydguQNT+npRUZm1dmHyfwav0wYKpPXFo+1BUlQSn4xbG2lGjzQdmLs311DppwUb58XzE+lz2zuNfFhQEWtEnpQCbp9FxzfC3OJlAC0GOa2p1wq1f43z9+ToAVmpxDp4dXsYL8hIQVkSk0HzH9tdP3zVwvwD4ZwD8cVX9KyLyCygTPABAVVXWYNXfklT1FwH8IgB88fmnWggdSfKKd38DeXEGMyVd3zoutMi3LodagOXEJFj2XUhbSZNr27cHJOyjerI20y29qk/FcBECjin0C2lPHJuzDyEw+K6rA7uZp80cKUuZ/Hdh/tkHAhRnKrEfm+0ZS1FRRBuTPjqDMuwqFmYWViYLaQppmn8EcQGQF/1gYf58UiGeigCbouaMIHOWXYPa7jbQGRMCJ2MkbTizz2fppUhwTSgRer/jcvuSgqbokqEfVWxlzGppwUlAEmUXKvFFIYUGljfn4xQCl3F8OV+pEMjI9RHjJ/mZLCk5J6i/MQUFuECxR08I6NejiEw1Fts06RNtid11HhTpAiD37wMwexypk1/FtwMasSJHbXlM+rkH2rmn10rfNXD/ZQC/rKp/xb//ORi4/90wt4vIbwHw9/z3XwHw2+j9n/NnH0w6yUiWQMqpm6ChIIv6ubm+gSCBIRWRjCDKG2oe1SVTiwvzfgZ6d355woRP11ogL2fo3WiZDwBA0nx+PRHw01EKgOoOYKKu0VxyC8pBR5YyTphG0Cl+YjNrI4PStvwJTRY4cQGkLPoipTWtkeaOZaAYqxYghK8Ah9+NvCUAGAL4bsoijCx1BjlYuzsRZhjr1tauAgTJKV4AD8JaBuOHEr5qtS3feQFdtfqdwEfVyvJOjHWWylLXSymFS6X3ac1K5LGtlaipr0dtXcmwsSLIu8815r+k4CojxrMGMlu7rKG1Lp+C9swdBlXVAkExrbM1USSNRfwTdTUk15TvY46uPo8YGtLegaiDHovTT8V41B3cXz99p8BdVf+OiPxtEfknVfWvA/g9AP4v/+8PA/jP/O+f91f+AoB/T0R+CcC/AOAH377fDuzkTd4k61yliwzuC6lrVcSt+PhI9qUYVzK7XDjOiKWYdVRpHsCSTkXYQXvuDi4swS/Az5bXtoUaz6irLy3YbjmUfDEelynT2rQGIaHmNDAL2GGtqLFcrUvYosHNsiHLuAWDRA3bcYgqgI9QpL8spXWWCuenXri0vhgBmjNgvJ77+IItnqldMHSFYqdNbA3nOwRNeTKgVVjAvvpDE0fm5wcpIms5z6DLu4vwparmDPZC+TzWh7aB5iQ/phdiLdZ46qGc+LftC/dBQSykdGCkTikkoxKmM6yWcJbxFqgoBMCPDQL1y4FmtjmElwgCU53tfcr2oDsLSnhwnqWcww7wLAiROMDDylalUBhWKnbWcLQ/Cte79SN/9/Q66TsF7p7+OID/XkQeAfxNAH8UxuP/RxH5YwD+bwB/wPP+T7BjcH8DdhTuj/44FYSX64sKghw+OLORLpjnYgpWq31BMwijFla2Q9HN5v5zSOsh2c893uqLue0Zn3UlAGHtv5d/yNwY7iLMRB2aPyfAWRNOACM53/pLvOwZGv+1fcl0HBaYIxuqviwtaUeCDkktZ/20/MF5Q9uz70fH6T5BtHNW6ppQjWVkFTTZCBCB7HGKwFmnVkyA8ks4Z/arGNKgfMmv67M2FtRDBeJcfLMGNJSOXqICNhLwVR4lQFYaK6aOf9KCphw+WjtNSOUOh1keQFNTXyKCg2i1xPbZebdl+EIaDuJxVNEDL9ropm0ekGmColJ/gweMZewK3D1SZSB0jrVvW+TJjOpKX35MCcXLESbZn4OObBJNq7xjyqdE0+G0uafXTd85cFfV/xPAP3fy0+85yasA/t2frCZpM5nWwAJT3esVJ5/jSQeFet4M4KROK2KNKjEzj9YGhWCmKtv57oKWemT8a0OZVx+OtoizGLadkmBwKDfK9O2D45loNAa20rGBcDBfAUTdG114/74Akxl54ckC3ouGEeO6iZajXLQnzv2QU8HRuzqcmJCgHCg1BZAhdhtgEdIxzjLtngewm/IAc55L8B/GPDPy3SF6iRz6TOIGch6fyFaNLkmvelKgaoV+mycLCyuBR+rOIwptZmdpb5VQ0MWl3kBrI++GV53x4fCuHqhVtNWIUU9lAbjA74AAm9t9LNz0Hjf7reGAEDTwvZGagt7uxU8nto2mLvMn22P/rvJJDmmrnnt3IEPOwcwl9R+XvfK5LOAkCezK4u/imeyPnb5z4P7/R2Llz1ZC52y5SIKTnQHGsuyTfayir1ZOwfG3ttLhDOADbU+9iJja6gS91HL4G5/zfSXoZQYslO+kUcGsuuK8gC2Awz5wdJ/My8m/ofDDZoVXqo2GDOynl2Qk+tS4WPCfkCkKFG2/3Fq9AxVnk7iwnWlems8gPgQ3cY3PBbXhv80oU2Ax97VTRHyeDYned0/6SC4DdgHpLC2yQY3x4qGf5RXwNeA4TG4poFgawafta+4sptwVYKJoXgprm/2fFiwm+6hO/6MwFI5xA8htkQoYY0x1i/aJZFQhQcSSkApioXRVM7WrjqD2bZWU3RfnvbDCtRUhRPelGyvb0YXmTKvD0jyZG7J8Fq02fFim8/nx0ny7p5843cH9I6Qzi/Qhj/9z6vwNoEe0q58LCCX1j2R4ZygZjApxPpy96X3pK9riX5vfv3fmvDpeNdPyC3RIhvPhivJhmLZ794owzeLYuLnBQhwIaFG/FlFogsZNj0JEqipeDg9c0Z7HyASGfRB4IHiyIs9IRpt9TEkRRDBp3d3xSJB3t4e0ogrAz8vPGV7ofRMjnCZVtBzOGYml/hRlWGRbJvAK8MuYK5ddmRAA/hIQlzWlBjSgSrOt9pmds3rllr/vKWnrDaAGylEfnQVNcSImRTRsqUYUeBTgkwEPAFXgLv4MYyD2pKf3q44sHqncx90Eh+hmnHhgy46RVd26ledLsiND4FadWmsl0IDmCLWFmYzgsL47najNAgoJTSceFrniw0B/T6+Z7uD+0dLC8IjXJHMG2kKNz7leyBkrhPV+6YrkouLKulORL1JnZLzImuywrL7W3pW/S39l/Qz0NhXrkKU+6lt/2/9V47tDGhNZ6YkQcxLXnYWTST7oWoIQnwlmwSEAor63M7vMGLMzAtkGRrMbu/F3xtEgKkPhGqEQoHUhIWkzqr1mqg1wKNpN1dTa2cEpgQp2w1yRiwUrqpse9bYU4WpudlBXojeOrx3Smdbd8y0I5JUK/9xUzyNs8HGwdT3CwTjiExBaYhXtDme2Xfh7HIJPHzZsA9g2yWlql/UN0OZXFe1tnr6gQ9hYmx8BiEar09PUcvjLaIXhje/vp9Ahbs3B2UKLmo7gu4D8GSonXTwLA/th7Ja0uJe8KETc00+e7uD+EVI4EdmXeCj9L9CYPe9dhQR98EanBc58ja70LjBjUf0k4HWa8c72GVu7UYt71Ge2Vnato/rWwqEr8ihR056l9oerwdFHIgoDozj7JWEF0U3y9ucuZBROGEMUgUWFO/EkXrcC2CkueFZCYjRR6bv/PlwwGZ530jZBaEuh9cV7Qz00arj1O+PWEZH4AqAH6LAWoLNOTuSYxll8SY1vdZJMIYIEDAbQD3LoA9Fe+CnriGJPRjt+o6kRJLVpEhMq8p20y5vQtsW4fBJmTWF3Ry4ZHunxKKTU1lkVGe16uAy8ebxg2wYuG9L3wS51sjFWte92XbG3Ucz/hLfAqnSaQyImBHr0QqQ3vKRg1QQqKcC1kNLDBcJOH163za+dcJznP9OP13UCOo1ZlitHwa5Nj2VZ383yr5/u4P5Rkh5nNj/wVVFA0XfNV4YXbx/KAdoRqVKvVqlgKSTBr87eHuuh4tpvJ5L8ytcVncE2gKd3HTwYGw9gQgB/vEynKsoLZ9q7KPCF1DE4L2fIakmovrQmYmGCTeIJSlYBCmQgeX7DZL46rmRx/cdiYlaKYMhtYM7JhdrL27AjWGVJXvakXbAYVFQn1Nkzlt6iXAKIZPg0OxtQvVx6tLH9voBoWqbiiTuZncUlOGnq6Xf+YSigDN6quDEC+mTlmAUxQRTA9rjh8nTBGJsfN5O8pU38cqAIJGTR55ZQQw78e/Yvyo9jcw7U3uSb2vaL+tl6QZUN1Da+EW2Q0x4R64xO4fAIkNBOrqHLWrfipPMeCcGzRIYVwF8ao3v6OOkO7h8r0awXIFeorb34YtI4RiyGg2uYJUXGkD8sjIM4rH1RkcQd+dnJqTswKeXh52jxv/OvMjNfWlHNqPjanJfMFKc3aUVJ3j49bUQHT3YiUtD+LgkpEegFzohKdpBWcgOZJEIJE6u9wV6JA0lWf9vLDAetaI4U7Vu3/L8BCicb5ehxDiQIjmHxQ/J3yT6HAJfMuM2HlQP3HuWfpKEenjWBaB2ipagE0i6V5m8HDS7niVKcdkWZMI516eHh0mb4nJwcWGUmLM2lPDbLq1sQFMDYBrbHDQKL4Rx+EXHePO4rz267taW1rXnE17MSbBQYHhtm1i1/aYb3/zJ4UqyBEA5WUktORdTAcXuqvfVBanBPxoxTy/ZtYH4yRPf0eukO7h8jLWugewELZ/N45P6O8sL0FOY7V8vyuAuK761AzAw3tWUrrBiobTgu3LSicLXFHx+JmZ8CexSX4HEUVfxQ1nHhN5SujrVgpMmcy5kwfqtuWCNavQnUUakHClH3PI/b8OCOddTHrDLpoJgSUcbqmdEj2Kwx4upmeTOXfOCa9m1C4qybVp8CjBcosEt90AUi9bptbkgjZ5KyMVxZCiUpgPu6zBtO2X2aIq5YL+8eq4mf1yZY96L9DF3qPiO6vBTvrs8VtZ+9PM+50BupoqYRx2DH3zMhxNMYA9u2IdzYlPtflVje6JO4s50Ciom514mMMrgtvgVE6zmRp0DsZ5/D1J+inJUdF8alYJldXBxyuc255HokwCMtpL8cAkmJYX2stX9XEe/fC0S+p5843cH9YyQlhciBtF3OQhyxrQtZngHJuFJrc5CwVySVoC4FE4MnobvK9PyhGnIjAiSYa0f2aPMBcZaUnKexDRTLia8dZJaIvPaP3Z0Z2LdklMO/3O81sYOVOsCPHcBQiLsmCzPTF3RAcec28c15A2wK6BH0mTGoggwuT0CksL7NyUKX/caR6dbLNTJMPfrz2DeGwhzGUBeBrGBZSU4KWr5L0M0feyWhAa5y2ZnQl3CUlo+jLAGwBlqZQiBUUWrtOrc6gMTcO3bJxiRuQI3+2Dl6zVcT0MQfrHPV/1XpBcWcEK1LYwr0JK9ihagLMk6wINwyHCm0ofbug+4bAbvw8JIAE/v7db1wtSnGMWhgPRICcu/lsqCaIM2/NEZUNFOOHtRni33qpL2nV0p3cH/ltI0Nv/k3/WwuspfuozZQ9sm9uI6ue6WqYU4+96o98TlbKur1f/H5Zxhjw+PDA4FSvKvJ7BiFGhwwOEUdWvUUL6iGnR1BOzp3eVH+/auvvoRA8P79++Pif8EDp3hb77R630JKiXwbgMvF7zv3/cnwPjaPemZ8vQtRxhdffI6f/dmfwZyzjioFk2zNLEaae5sBkDRPJOZA8spi6CukBX/WECaCUAMmtIgkw/7yyy/wsz/7jy3EprEhVCtQ6gw7aNiBvreF/iCgubCdhbDjwhjbwFdffYnf/Jv/cWvgFjaaCeyKbC1NdhOS+gzhLRFdPokAF7FAM4INugl0Wvl2qozXo0C2AOG6b/7N0yM+++JLqjbM62SlWe37MSdcgplQ3G47bjvNzaBMzsWZAWr2XXG7ka8MLHjSSD7DE9NiOXz15edQVTw+PcLGla1ItMaz3lo/aeHz3glQlqG2ZmmO8DJv3Y9Jf56+/PLzF3+7p58s3cH9lZNC/7/23i1Gty277/qNOdf6qmrvc0736W7HarWduA0ByU/EsSJLJHkgkWNbEHNPAyImBFlIIAgIISNLxA+8mKuEgmIZOcKOAg4JidISQthABC/YYDud2CHpuO2YxK222309+1L1fWvNOXgYY8w511e1jzl9ap+9zzlrSFX11frWmmve/+M+OS5HBJjmxGVKlEU5lsLqkz1hR8O+/oFLUlG+9Oiam2VtG2VsIAIkyWx168PmaqsNrQE00kUfZGAG7Gzn7EVcHA4s5chX3nh0dtBJ3wLbRq1nm91IWx6ErlbuG5U4Wo/q73hX26Sl1X7z+erqijeWa66XAdzD/Tjy9w+b6yi43FXjJrHE5wy4zTSv0ssrCmttZpDog+ZFf+YufHj4Ck8eP6XGCRij8xVmP7/MwjwlNNnZA+ta3SveDh9pQCAGcBGevKp5XoN6YhDrnfNtGXXPewHnU0DsNMKqSlH44AdO3BxvnFm0urZkK+jmTISqlZyEKWdUhFpMX7CilGLGXyXOUVAOk6mn11K2EqjSzhWPkKwGXMVyGNjctH4TEcrrSpkyVYTjaaUeb8iiHOaJw8XMdMjmk6DK6bRQj6v3ZR8vkUgkk3zp9HWjqiwKJwVYNsxaY0TVwDf6LxZnrJU5w69/+at87tc/j2D9hGDj6Jqg6mFq05wsKgOlrNUyJM4JyUKtSp4zWiu6Ch/60Ae4Od6wqlCWhXVZrA9zImOTp+lhRDiVyloq8yEzJ6Gs2k5cmyZhXQ4sS+Xm+oacM+taNushct1v1ojCxZQ4XMxUtfKXtVBVqWVkfOxXVVjX1fe1qWXhg753xFPF57qFDVoESQKW08JO90s7uN8z1ao8fvIEAQ5TpqTMuhSOtbB6DGgSqOuBy0OCdeHR4yccl9KlM99YjDPPjvIMUtFWXOrg7oXLKNEZDKQkZDcNTNOB01p49OgJWWnpMBu4y/Bs29fugMsz6akvZB2+7qAdEob/21y3mw26bTL2zCuvPuTp9ZHHNzfeHjWbsx9LG+YBGZmJO8YkUvCOrVBJyCRQhJRTV8cviqyKFtqGPmpMdChTfDBvjkcePXpC8SP4Nn4CPiTlUpiL6f5PCyxLoapvmNrbIylZfWJzxAVWtYHeMlHb/gLxnOV9w47nq8LpeOJ4fQNApbIWP9UsVM7S55GIeeDnnKiIgYrQNBmlVo5rZVlWkio5p1bPnARJiaLGCCQHEAsFtNjrkCKnlCxiAGUSmKaJVQtlsjwE61K5vjmhpxMXs3D5yhUXZTZmQOG0riw3J8pajVHyqdC1IKG5qMELN6nVvk/jNN4wsqpKibH38Y5sfzkpXCSO6UhdlXpUpFj2Q1u/lTRZW6eSKbWylmKMSaqWmAhbe7lMLMcVZeKVy4+wrEfe+OJXWJdCrcXmeBJyTXCiMUMpGdNVKpxOfR3VtaIC05S5OFywnGytj9oAO+DGxr5HUJjPTVV4KsKUhFVhMdVCcOqNeY6+KtXG+WLKrEuyELxwElZcW+D3FhunlJNrPQzcT8fjHSt3p7dDO7g/JxKBpEItthGio5BqG6eSWEJyYwBkK4EGJSNixd4jPV/8kJeq33rGYEc6y1HdG44sG8XA+C6xDViEbs89b6c2XGgS650e/xLbwrYtVq9xex2+VrVNLSVnfOqWcdC239xJsQFtvhbMTpoFOSRy7hK7rAqLtrNe25gM/dIYFvFQOrFktnHkapf+zrzai0nDiklXpYS5Rdu4NxNLjI+rVSYfJyVT6yA9xT1jn+t4zG9naeIkspRA1eo8ZygCpzWkKbFkLDk7AJtkrlpbOFbkC6gS/abklMg5WY7wAB2EpMqUU2OSVoV1LUhR5tlAXSS1nP8qwlJNSr04TKT5wMXrl8g8cf2FL1NrNSCp1eZxBV0qtVQDD9GNP8j5uS/i8yE0CKmtsa3k2n4a8+gOjinG3UF+Mm/5lBWdXdNTQFdFqmcWLMpSK6REnoxRXxbT2kiyMuuyulZm5Quf/Ty6LqzLamNdTU8zVUFOBuZV/TDH5G3Q7i/QojW8r09LP10OrSQRqvQxsSecYxPPiidCqcpaxzncY+ztdm1MXi2Vw5xJuXFArTNjtYrYvaV61kUXHOJY5XLuSLLT26Yd3J8H+RpIdJVobCG2USTSNKGinKpSxkdlKOIW4nbpU9qJTMPX3HYgCunl/ItWl/FxuG1Pj5tb+q3bzd1ujbevhkQ4XtLhXf3c9TOQb+pQk7o2NmWBUGrc8shpPNHINA0cTAKZvP2rl1Ex/Xc5e/9ZsQ28A+y1d0/XREivfrWUr8uqaLJwq9qYh6GzB3upbP/dbJbJpecaeWpHxsy7IiIstRfgTlchMUeonUBS5klZS0GrZbozhi+hhCRsLxFnBCu+8Ve4nGfmOTXJPJqU1L3CxTzgBbNzl5RNGvWyURvbSK86TZ5Q5qgkWTlcJJYLoc5CWYWyVkpyQ8VaOZ4qpVRG9bI4cxVq/uhijSx+DlKVnhs+xjSJAU+JQ38cmMQz0I0dLQU4GaOW1NZkcgfJ6kxBk1prQUhY08U1NUJWIc+J6ZCY5szxybWp9qup37MIF2niQOIkBVKy/lbltFbSlJGEmSTENAdJYC2dTbHXbQIzG4U5JEJEmx9EMJg+b1QELdWkfmfiSykkEQ6HiTlnEwS8X9uxzVaDpqHqLEWYkKRJ8DvdL+3g/pwoNpVaK2XYhE16soWsuVK00K3FYwGtoM2lcR1IZVA5S9ugRm57BH6GBdccuBqo3iFt+4a54d+3mL35J5iFXsXNbnj3g2MVHeRH6amqq461s/kRhtZqfXfV6doJa7X9juQeaQgxclHI7d5hhmgnjenIdPUtMgsNRVNKI5rGtgpistRa3XYrMkhZbBiFqKN1w5DzoHeQT6CoZJgLYiceBt9jsRFnBL2QUnv8tXqTEzBPrhWoylKKSZtOU5ItA+pmhEOWJq1ZcpXey9w5T7QxJ50JCia1T861KtfHheV4pH61cDqtcFpJziUvWlnVwH319jTgbV3QwwkFYw+r9CxxcX/1vsoDo1svFIHnAAAgAElEQVQQPwinazvSJpOLUEVYbyqnR2vTVqQUzAxNu6ICdYU0C3WFfJEhKXlJnspO0BXWG3WmRUgkJsnOaEHS1M65t5TGyuFgnG2tZr6ouO+N2P852xzPOZkqPExsQ0ZGA+m+ZlsueoFUoThzbVoppZTCnM1X4LSaFuDiMHWtkOdK7vkjfM+JvU+M3Wk+IcDlPLGo2vrZ6V5pB/d7J+0SAG6z829EFEmJOSXmwwSpIOoqxrBnheTjG8yzwCtAIH7sGFMZvhxulrP/odmL+z933eyfR45iw5Hf9bhLT7cqPYgCmzbdhczn7VQzHHtMcGvKCCBnbQ67ah3rJAKTwCFZ6JtLVbrihunhrdHcdhiLAXtIsQbKHep7jvKhuwasLRJKAe2Mh3FZWynd71EZtSpDfRoraAMf6exVaHnoz2Ugz2AKwJST9YuJzca8OcoLtGNJo087s6bNQUxLIQukyXwIQhoWCclPOuCK+LvwWlvHxBqJ0+qqazJUrR01wZNHJ8rxmmo2DSuhMJyE09uu9MNkOvPaJfcWvhh2nMA4/3UrvFBkM1VHjZbNSXMqTCJMOREHwpRlZV0KebL5UIPBdLNMeep+B+p7RDUOvacjhimbCaMWG+M0wcnV2QamyrJW1vALcSYzcje4r6SNmvunNFOC15OxPT4nx/7xoTPGaFhwyX0kLmZxJsnGOwDditKmQRoz6oH5C9h0q641cH41vH13ujfawf05ULPDqrBSe2IUMVXUNAnzQViKSYxNPX0GgOdhaoTqc1ycGpKmq2wbyJ2juy1ZaZtaL3sboaK3H7+DAuC3DR8+yHDjre9vl3X+fVzTpi7XVru+0W4f6MDvwK5D2wTTwWYDZdxHSIs50I27u7bx6BJu2wTjBjEAirztATIbW73e2UuDHqEDjEQbBulbxgcIkOmbb/Pk9huSg55E2tuBuVDMiWmaZ07HE7jNdwpVfTbwTt4VWs2/4ZCEPAtPF9ukw1P+MGdLvmNvJFTQ1blOCc7EEzONuRnUpUwbi66tMHWuo3tO6CFTT4laijExKmSF6RC8sISPJcFSQkiJ0WG9DxOCuL64aQ/UH3BQan3tfRd/myIj2lIVScYshTd9MCchvYfkbXUHOQh58v5YlGW19034OQTePw8eTjx9Uljdn0ElIWLOlGvRtoYT9qA6Mzi1s947w4XQ7NzdabaxWIA25ivAlvCBADe/VKaUyHNq7xbM0S5FGGnvHlOE+fnGsTYkm7ZDZ9t3VJRTKVROLfJhp/ulHdzvnWKrS01yVDAwTaYqm+fENAunoxIG+dgcke2e5EV2KUIgzpjuG3dsbS7XdXRqZMA+LiDfeBqq+AOBXzLedeeHbVHjRj5sHWPd73ym9c/dt1BM+ujsSZQpt+5tVuym0gg+yKWLkMaqtuQ41N5d8WhUp/sFn73JL7kjMy5gNZvzaMaIkdngzHh0J8HcnXfGNpv9SNbV0kK1NNAqqcVuawe/xjVggL2WlYI2qdmA0ZkFd56LMayqXK+VB9PkBXpudKWF8HVpnR61Ic6YIubf0JwxzTP6Midee/WCi6srVOHRaeUrbzxFl8Xnd6HWSp4m6nSgrmtreLyvhVMWaUxdE1iH38EM2pQIibRxzwSHq9KZ2tbvA2MY82vUvIRDmWCmC1XhqAop7M9u6xbrh5zEktgIcIBKQQsUFUqxMb+4cHCuFYuOsJS1SbpJRUSYpkSt1Rzm8HBDkjtY2vwutTMYMjQ3/HUCiE3z56GDbT5r+zHG2dX/bnqpPgdLtSiHcOw0pqKr9BHQg0CmaQw0TEMiLMWcNev2aMud7oF2cL9nEuJAEum2dlqEmi3My0RKQjkah3u+gTfo9uxUaUTIJrl14LGNyL3OR0TYYNPtxXMeUnf7++Fz2N7PPNvHe7o0OtZte//5CxqWadtnt9/XYoATj0uE1J3d15y5ethN41XEJIfk2cHEy+n1EnrX6VmjOsy3Df+OW7SaZ/TQGe5fd5sJabxUb1T/O6CK0vbH/n/0a9uth7R0ip0eN3JVDv4G3tXs1wqrAGvXfYSkl71cY4i62aJUtVh3F5OXquRSyR6Tr9A87KMFcT004fi1nOBChKspoSmxIEwp2Xn1qDlu1WJmjql3yXj677gcglkSnLFBO4CJM17OrCRSi3KIcgO0Q8IeX2BakAC5wWOcHiVQq3mcx/tCC93mYQK5EEqCsloegDwlpouprRldlOVoTnfXTxduTrUdIxvj2OavO7uXoqRqIbSnmpkn4eCLw458NcZDkptJMA//umozU1hzIz8DbYYVdedKb8zioC6D/0Xycu2J1MM6fbZW7ys92VvidEVRB3OvU5FqEUU73Svt4P4cqHkJ02XqAP2UhXwxkYByKsYAnBv07BGXOgIl+vWz7KsdeEas3uLE+DgBgGOO6s0OPBYyXg1sH6ob/24yzTZjre+s/fFzbqAVeo7r8XXYDH1baHtQy9UDTTW7DdwZ2pUFJou9bc81lA21sJ71ay9L/PsN00RIgNIGYNCod/NH9LjQACLeLQzDE+9sNmFpkmiA/KZvNNSy0nOih5paxvd3NktVmtbHH/VvpAGA1oq0BEHaxiCJsNJTE4tL7dlVsqERDonWJDvtfTuMx0nhNx/fMF2fUMTCQZdiTGzKpvFyaV3rih2opP14XNWmJu+n8faQL1W65iDqo6Y1aGmgW31DnR4sUB9SgHZ2egC/m7OqdGfPiIePOZHw+kl42tte0JkGe4EdRZDMd2FamTLkCY7XBUEb4y/ORNkJc4Ko+PuscNsnKqrCWvvRwklS24uQRKlwKQk9COv1qZ31blOyYgBtTnqlVFPXVyVNyfsutb5pTJBGN23NEqDNZBFaEzANRJJEQlj9yDwRBqZgp/uiHdzvmyTAvTbOPYBdgDQl5sNsarN1pWhFUzhGqau/IGK/NzbjJpBKVyOOm7gDUIDMCHUj8MAWMGQDZrK5q5mZcTja6LBvsQMDj7CF6zvX7sg1eAXl/Pszhgdsg07DPXFLxMxuNmlPqCIq1BJhYEPe77g1dvDYuca/m3e5hsQMh67m9EQ13vixWTEercMr3dt/w30FsPeHN+yWdDYxTgqrCKkmnwvDQAVXMHAOwYeE42V3dvPHtLoqWFswu0m4ygKIVrKY+0PY/dWjGDaOhGMLdAB2iTaYOaFUy9oYN0pKqGQ/t75SdfXnSzcnOcMT3Fw4bUVdaqtEzFNLvGPStzEGjZ9sYNv7bQT/WI8jQ2oe8VEVa3PYq5OrFCIsTdJ2HSo2SXLj78wkcnVxQZLKsQCTcExKfjBRH68kLMdDysneK8G4DMxmErIkJjEmrKwmyecsrKV2B0LMfn5zs/DgtSseHVPT1oSvxFrVU+JWklbrmpQ6Y03XiFTPNRC+BYke6taOmm0mjHCu9AQ//qxIRJGw03OgHdyfB6WQ0jxVZFMBmiPdPB9YnpxYyoqWao5FjOliGXch/1d6DPMgcakDd0frsKDeShkDxMbSOem2Iw/22fakbv78/yDfzQaQGjfHKH242z+MMCab70Zp3Tbq26/ctI/e8vDclpEp8Q2oA8FZzfTMoXD4G9e13RC282H7O+uz8ZrNAU85FHxDuLKLWLhTIOFQifH5UMebzTzkaAP6Vj/vztaHMjzvbFF14I18+NWl9pDQkgiHlJiysGDpcg12zcM72rW653zzuo82S2dqh8aDGghHqtvuxY3FaTswJAdJ1B3Kose93v1dMeLBDFvfrW6zDlV0JO+pMTqDBNpmXGOIFCQkTgerNHj3q3WypecVTkXNERDQKoMk2ienIKZG91Ne6lIhwfHmCJIszbRWy17p2essBM/aaFoXX9+VBtqxF6Rk7WQpLQfCYbYkTWXtkvWpKpfHhfkwsR49ta0D8VoqFSXVysUk3LhPSZZt/ozimrTs4X/VE970tjoH4kyIEgwAxJQPBjsjbQ7vdL+0g/vzIOei8bjTrr4VLuaZecpcr5azuRZLDTuCwLmkGtuEhAR2DvqbxdGBZiOZjCiofjlUkXrr6/Pm+CboO+EdRba6yVDpO/rl/GPIfPH8KHC2k9EYfkbwFFxdevayuNm944mQtQF5t0+Eh3vXoDy77tKeiEsWBxx2TS8nQIFQe27LMsF/2wtUuxbAeM5kIVuhPDfAsnq3pOIevN40t4PZozp3EsM45mNQ3+jnLIYsnp7WUhfbfQcR1tXzidcKqwF9yqnzl2obv+T+3ta/nTtqG39IgKpqQ5aECfOSN5NM96Y2J9UebmoMXGpzr1YLsxI/9q2Bvkbf6cB7nmd2lHGKbBisFiYmfe6lJJYfQEsHXp9cKbzYvVOyDipqUeZDRqaZnDOn6xvWWpkvJqZDpkqFq0y5sXwDk7fPxkfbyfNoT05TV5OOq0KehGwVZMxTkf376+NKStnMIaX6/iORvpCahFNoEBHWavbyLKkJKk1giXFUi1VXTAtj49Mnq1ZFk5LUzxHAk92g5JS7k+NO90Y7uD8HUsFzk+OSY0hdicPhknmCZVlcDWYJIyILVOwuXZqwRRwqvrYJ2ZvavUrfi2RAwEFganVr9QysHgE7luywK4/vsu8GybPVE3izBXpW7ParAPboqyFkj8E2rWohbGIx402IijoGM+BOR5JHu7Y04Nlq28f+jr64q5PYPqEdhBFQqdv2DfVv5bTno6K9k2UcQM7irodxUrUYacEzIIKpnX0zjs5Q19+GOrXHG4THtbfBN+ZJBCbx/Arm9FTENvZUlSlZrvGqiSVXlqUwuaRfS3UACEc8b0bpERrhE2FMiYWEmQMWDQBVQZP7mSwVTpbPVWvt7UleRzGwijh56/N4nzMokppfgSnSYqy7M90wQs3JbMvAbWesHcE6zD1/3sBNHNRrc85rs6soMhnzlpKQ5pmLB1d+GM6BJ199g7JWTkc3D0wwPcyUU+FUVgPHSnPiE0lNvZ0afEvgM0WENCYscsk9ibCq+TSIwGHqdnlLiW19vFZF1oqoZdisrnrX1A99CSe4YGRp/WW1WYtpdmQW9NI0C7pieSVijiaYLzN52pPY3Dft4H7PZPur2RTjoAxjspVpEq6uLhDsFKRwwlGNTFHh6bx1IGsS0fm7QjU3glbY4hmF6NigziGtA9U5MN86iv38yVvfSZemQ2J8Vg8NQN/AvJUp/R7pjY/+aGFksI1NH8oWBakOgmO7Wjl3kwcwbhilDWfUm9b+tp7V3suExD6oCBoT0Oop/XqUHGDlye5qVCT6a5Akz3mx5KadjWPSAD6tzq4+j3Akk3KVecpoivBNZcIOGioIEX7l+uaWXAUHqilnqiRWBfEUcKrmcW0nARpzMM8z2aW6Uy0si6nOy9B/AuhS0VOllGLqXl9HlnWtNzoJbv937UmEioGHyveEPJaTPrqlg9DZrN8y143Zls6oe0eKr+9aPX4+Z3Mgq549L3QVzpjnZEb4IpBnU8GXWqjYiXHTIVPWwrpWLi6zh+8p01XemPWMrxZ0UepTN/tNqUXUGE6rpZ91bUpI+Da2lTBX5ZRahENLUISSimlq5kPmuFbLDCjJAT0mtDaho0/7zi1VVRZn+KpCLiCTZQksWpnwiKFS7djlvIP7fdMO7vdNvvhLNY/aHEAtwmGeuLo6UOuJ5bS2bfocjEMqvAOT27+bRUVXvQXAw6AJUAYpcwvwGwmmAdFWWhW6+nKkO4Fy5BqaCu+ujhqa1vbO2+8JyXyDpO2LoZSBJ8ClqwrNq7wdINP36UFkG+vTN/2NAD/c2qvTAdo0BoNGJWLrdGij16/7TDSWiEgwhKcNjc1dw0twbPpYtwA4EUgeNnlL2tSmErZELlEhpaofN+r/Jw0g8t6QHhNvjleFpRTKag52KWVyzkxJOCCcqrCu1RkC0xDMSXjtYuLy8sIATpVTSZymyvFUOC0rYR6yE/Kq1w1LFiN+CpqD0mYaCE39HRoA40OMOYl7q4hpBRob3Qd1tCT1aaXDODXWkmC4q4jn4o9MFtK0RnUxE8c8JaacKSirax8kgRZPbHVz5Pp4tJPzsiWJmacLqCtrLUw5c/WBD/LwQUaPNyzrynFZLJf+ITNNwvJ48b4ypJ1yZvJETdFaG7fqYCxMU7ZDewKd3TxXPZ9ETQb2UpU5C6vAuphzYsvlo33XyUO8Xgt/i3UHaME1PVgSG2ciNObm0rKB7HSPtIP7cyAtyuq7RMTlIsLh8gFXlxOPHl+znCznvC2AiOvpZdjeaosvNhwZjn31IolY+AZaDantb8MFB5fIKbaV/M6tzPHCcRsdYY1eEX2zZ0KSeAadFT/Kx+K1Eg9uvlVGE6EGiD2/zwHW8g1oG4dm4vDmjPkzGnBuOmR74S6GptljzyRngXaOfW2bo6F+wjZJ67YOJmG/tcNduoNfYwPGTtVovbZjYjf2UJzRDDWAavNlIFmIk1bx8rWZNHAtQA0Paj/ZsGL25JpDy9DdwicBmZI5XLkbtOVeSEySOOREmieKwKEqN8tK0ROntZjk7loAhRZTv8nxMIy9tNb18bFc/yluOWdxjOkdGMVgWkfTjtnmtZUd5qJxfoeTLPT1rVot4xxwMVvSn1KV41r9IJlKukrNcVH8fPtpmsnJHekqZCmmOUmCZDhdP2ItBy4fXnL1kY9QvvhF1kePzYP+QWY9FcrJ9xrsSNXIRay0HFl+TK2lyu1rx/pFtSfj6Y6RoZlwZ8KU3Kzh0RapZwY0xjE8GmD0DBGBWYQVbE4MWsLW556oaqf7pR3c75t0UA8KTc2es3D18IrpIrN8cfVFCCpxgpStTiEm/WjvC0/oDq7n0nB3tutioQ7Sc1Mp+ndyVsZWZh5B/BzWtuKujsnem+Q0PBsS4FA1tiU847p/lrEXzkj6BkKW2570Re0ITm16jSal9BPpgpWw7/X8Za15Y9u3AK7R0NH477b60dFQxgfiPh8Mk1xrz/l+ntyggc3Yzu7AFUDlPdKZStzrmi1IWR8Yc2SvtA25YJtw1Nk2eZew8DlVW82ZszDNyRzZaqUUd6abMlIruVauMjy4nCwvejbwLwI3IuZQqpEVraLVwTCOZRsHvfVZtL6xdd05cXO/Q42Dd2TWA+xsgWCmxrC36KIaoAgk7WurjauNRZ4S4glckq+xWvyUvQZ6PgaTNFu8RQNU1tPJND5i5Z2WGrlrrf9mhVp49OQpj7/wZUtskyZIk+WlF0E9KmfKyfMOmEYhiR3Duw5zorV56L/iDoHJNTbFtSfZVR9rxNknaQ6stfSywqlv1DxCZ3xCI4CCFMsjr95XU0pWj5Wd7pl2cH8OVMsw2ZOByTxlHrxyBVPi5npt6Rltr1QSleHopw6UsW33ldPeE7fFvuMyhd0VomlIp8N9MEok/dpZ8R24NoAn25vuYLh13JO3wtUdCN3YlbOHfYPQ3ozbDw6Mi6cAHN+rHswuHhrUy+iftcHimzRps2kN0t+ZLX6M65Xh2bAmaytc+zNhQnEJSmsrvI1ZB21tnsZ9kx4U+88YFrP5yjD2EohozKhqk8ogtAtCztokvqo2FivdTm9OdqkPlyQyEe/saleByylxMU/MU6aIxVGbZG92Z9MYRLtcjbzhBoMhUZLFgbX+HKemjl0T/8T8V88hoaZBiajDkR9L/k4783zMtkYsGNrccZNDMEyl1HZacBLTIlSkZ3YUsTPegzkIxqkqD159leXpIzNBIFCVsppAUNfKdDmZqp2EVqEsBUmrrfbcJ0jMg5zF9xbd+IKsFabkdfeBLrVavvqB8Y2/4WBXxPwsRKL8xFq0LVXb6zxiQdXGVM10EWszD5y38ZGVQ84tN35d71zhO70Net+Bu4j8O8C/hq2FXwD+GPBR4CeADwM/B/zLqnoSkQvgx4HfDXwR+COq+qu/1TuKL6yWjlOE6XDJw1dmpKzcPF3MMcU3lurYPUpLzUYOw8I9g9YR+DTSnZ6Bk29YbRGOlzcfbj3UaQPKevetDZOGijvadFt+v6/X47wCsnnLs/gOcPV6ElILhdrebwfphKOQtKMrt+XpmzZvc619J53xcUwcbYhjG0ffiXHcVN3jX0GkUsfyHLI1PPJ9TD1E+lYfNVZFtsPR/obWRoa5hpiZoLr3c049093AZ5nfgttiwcM3K1OAvkvCxdttR4N6vHzOCImbCsfTaqlLPRPeqSo3x0JZFao68yEN+KQxRT6Nkp6NkTNFZxN7ZFjj/uTrqQG00LzeFeOp7WtnjsUYGVeu0T3oRwbNzC3raqxQeK7bu73A2hMGFQVdKmluPBV5mqAU1nWBZLHvaVWWm8p8ldyHQ1gXZTqIM4KVec7MlxNaCyUl1ielzYhaex+2dLr0v6VYX1ex89hLtdPZiipaCtM0ublFOK2WZCgnO3LW5BFjSFZPVDANKYe1VpZi58zbhKtc5MSUp+aRb0oBN/GImYGOa9nEye90P/S+AncR+RjwbwHfoqrXIvLfA58Avhv4L1T1J0Tkh4E/Dvxp//tlVf0HReQTwA8Bf+TN3mFgPaCZL7TLqysur2bWx9ccj0vL5jTGx8YWspnmqn1T0mEjD3EjkKK9sINBeBITjMMISuO1TScNwHBrvZ1f2MAisaPKrVt086hsfp2XcVaZALdNEbb5kHqe+dicgc1574htpme4ddaauxwG9XwgNnVChyYIHl9to9P2u9G2aN5e7W+EoUV/N4mO+L/XIdS4VlhXy2/YlDbneltG5mN4UZM2tZpkl9xTuYhlUMs5tfSqawnbuZ3zbvZxs8EeS+HYEdBD8zyaIQnZ2/S0KuXRDa+eVq4uD2hOXC8r1zcnSnEmppq020KiVBqz2jU0DADaR0b8fvV85Z2/9PmjNsKWNW4Aaf805mCLPo81bNPMn1EHTWxtlaKsa+HiMDFNUzNllzKctJazuSXU7qVecqUKLRlQublGEGYS5VTICvWkTIfkc6tQS7JQspSYHr5COmQ4PSWVytXlRFndATELKSeWo50hUAbGL86bL66lqRgDBpA8fW2p1SR235dyMq/6tmd4ytppSsboeTu1rqgoq4f15imzVpsz2WJ9m8Np9T1hKQVcmzHwCDvdE72vwN1pAq5EZAEeAJ8D/jHgX/Tvfwz4QQzcv8c/A/xF4E+JiOibnU+otmhjsxAV5pR45dVL5kPm0aOF02npdmDtnHzbwIhY5lC6DqrPgW8Azs9x6XA5gHS/1uOdY6O5m8YK9YZtwGQo2z7cXdZGAxEPyPD5t6Jxc483+0YRUlb/cgz6U/AMbpWeke5cCR+SmY0VnRFpwNxD2uzdHoftL+rqXW122hjD9g51UNchjI+eyrPJqN7nXdvgbRmZF2d0evHnXFPH88YkeDKlaFcISZH3Rjz7nICHT1myEQEW7GjP6urbJBajnpOBiLVdKGqHyUQDdVXISnIHuWUpXC+F/OREzp5PPt7rntdJLBSxpccd+dZ0h+nFGz3yqDr+9bFatQNaP0bZOnJryrCXrtVsyip2tnpYNMYfMKC7vJhtPkayKjHmsi49DBDE7NcKFCFLhtn7K0GWCT1W6lo9PC288IWUDLRrUbNV18KTL3/ZEgvlTKrCdDmj1yvFvc6Px5V1rT5mtfkCxIE5qx8IFP+riOfagPCuBwtPK1URrZ7FznopiTAlEOz7m9NqyXCmbBoMcBOEza9ltXMDLI1un6SRTVDYrped7ofeV+Cuqp8Vkf8U+HvANfCTmBr+K6oaLh2/BnzMP38M+Pv+7CoiX8VU918YyxWR7wO+D+DiMKN6wSB2MOWZVz9wxazw9PGRo+d/BltMTcqATSyz34BU5TZru5XQWh77MyDUeLRJcF3Cu50VKmxuBGIxPNGe3zj73bEm2/1nG/EW6GV77x0FeIs6iEe9k5jz4SjR0qu9eSc9g98oxN5JzctuuNR+b/vCeY5eV8HSp0YdvSIbkN0W1Tb+bUd1PwyNIAqhZa6L89vbkOo2bK/NCe0pZo2R8NfEkaCj9DowIyUyvIlQMYc3JKECORU71zsnpvCS935LqlQRqpokWNUiRlKpvonXlqd8KT0PeXX1rlZYVjjIasxNtMdBJ+qYhrbeHj/p/YJ25y/3+DfmL+ZuzL/tWgt2OsYymAERaZEm1QfcDrJJW8dUERY10BazC1k7K8Tphg/SBItALSwFChWpDraeryAJLE8L+Sr7KXCmJejYaKmDwaT8da2QMvNh4vTohnWtHA4WoliKtdmyDfrTw9wAmLOZAYonA4jw0VqraVU8Jj4chIMBS1lYpkioY72aPb5+Eri4PHiSpZ673/wFa18Keneo7U5vj95X4C4ir2PS+MeBrwB/AfjOt1uuqv4I8CMAr77yQAOQxDeA+eKCh69eIEvl6ZOjJZiwJx3wguOH5uxiNbZteKMoGEExgGsoTweAH9B/2If9PSOodIQaQeN8wekALA0tnrkqdQNodvc52tPev2VJBu5+kGBJYslaWjSP3n5/tNMl3NiY7ZjJAWQ3m/qQuIbtB5F+SE04mqWh2wIMhM7PCbhD2LbAc6YnJOioJ96cOL9lNMHYPu5hSXd6GI4bpH0fkQIxz0Lqsk/q569Ls4eW6vHYzbtZG2BG7oMUiWFG9bj2fglmq6qd1R2pS8WltKrV7PJqiVZqqR08xfwQOuPLGbe2nXM6vn+YNt3XY4iSYEgQpRULNGzJ/Nt9wfYkzxnQihzvUIAOZnFHlJSlZxhMObGuhcX9FK4uZ5aiLEuxvloLkpKDK0OiHUHWwvpGJT/MyMEmb54Tqbo/QHW/Da9bWVeePjHQPEy51T/yHIh40pic7KS6OvaPa6TUGUsf/6raTo9LXU/Vxj40AMGoicDlnO0EOfV57gcESY6DY7TtSaJjeOhO90nvK3AH/iDwd1X1NwFE5C8B/yjwQRGZXHr/BuCzfv9ngW8Efk1EJuADmGPdm5LFkxu4ZxIPHl5ydXHJzVduuL65odrRYBtHszGUqV307ysWLrRVfUu7p3Phlko9vh+E3UES77LoLcG9gfJxnTsAACAASURBVF/4A8Bm0fluektyP1uX5wr8eFe0ULc3n38YNkxnXZKa+iE7+x9MTIU4XFoxSckkn949Y6GRrnXkldrW/iw+QWnOWDL0aWtYjKFrQlrrYuNsIC2MqosWSx3aEWe6Rlt7f5V4qJham4NhGaAnfgeb0qUybeMWP7bZS8uDX9zBKRgwxUA4zEtRwSll5pz7WQg+F6pzpsbchNrfbLduaXZNAG5nVi7mmYtDopRiNm+xsZly9sxzxHk6zWy1nVl9HPo6itDA3qfxvzSmwMOwkO3815gb1hGhyg5tiaBUT9gf661HGPSygt9MYpKyVpOq5ywc5mwMjzP3tcJ8mJmyPWxHQlsYXVkqIgmhUo+VNE9+6EwiT4Iu1ZhchWUNJ0YhVB1xhO24p4zMeZaE5GCuDHRXrT4v+6yakqvXgykSNxbaRWPegGnOUE2V33P+h++B1WANj3sfn+i+O31/dnrb9H4D978HfLuIPMDU8n8A+FngrwL/LOYx/73AX/H7P+n//5/+/f/2pvZ2wHbe1DbfSTIPXr3gMgm/+eSG65Nl49Jhknc74IC/PuGbpDCKd4NkGxte0UFzP5TRhRltnLTcasGI0COq3EJtv903yrDDnTMBZw9oe0SG87fbqt5UeVMl34gN2MUZIpc4S+zGLkUOqvDmexb9F5nK1PL4qz8WovPgi7atTLSxV+esfjjQ9mxcm5sauGx7x1Sq6qFPA/NFMEG4gBobKy7xRB215Y5pauNxGMaX+ZyJthW1DT3ir2O7Ly6xt0NAEFfDioGRVvKYqx2aZNyYoOBhxn4Qtdj1lgdAuciJq8OBilCy+KE11hdh2x1nYzC/8beP2QD2DsSNXXPNQLQz6hg3N3sznTEW6Gp/oWXI8z/OTVo5Va0v7YS03r+L27pzCgZCmCc/na1l23MfAIkEMF4XV+dnSeRknueCoCVC2szxkMl025KdoY2uFat8yp4AqaWUpe0FIqZ+TymhIVG38aSPQZuskfHPzoMXVde+WMKcnISLOXMpk6XcLqZ2akzmsAcEYxnMR/Uoid9iQ93pa6T3Fbir6s+IyF8Efh5Lm/DXMHX6/wj8hIj8R37tR/2RHwX+rIh8BvgS5ln/5u8AkAjPEqY08eDVC+aycv34CTcLNElrK3h5HWMRSpdYQqra4Lq2F8YCjr9CB/Jn1/J8SXXbNBoMwJloM/AAzTbPIK0+g2JjbfcN8V0d37eOUWI8kh19FrpeRzrLaDXcSxPom2Tb2h+Ho8RrEq4F6UVGU9oWPzJFt/qo9caGbsmUrW5dmlQwVTnawGxkwnoO8V5MaWMRc8HaXhXPHHbG8NXejpEXNH7KwC5yoguYqQNBUrIYdj8DPBzmwN5TS2qMXPfV6Fu44cqwlXu9M0DOlmt8LVAq85x48PABN7UgJ0WLsKzGXASzNJpDEkNmxSGELeowMkd+xVLDum3fAytaR7S5NT7mR/H2yIpQacePy8HOUMWDpZisO+Vk2fvE0s5WpIFXzhOsfhIb/f0G1tWiFNLAnAlMU+J4Wjw5j5JKnJ7mfZITr339Rzi98Qbr48qUAjSxBiRhmvocMP8U18q4qSQ84rtw3f8f+dz4EDx9dV+GeUrMOff+TzY+pfoRuN7nPYuflZbFkiXlFE579bYWcae3Te8rcAdQ1T8J/Mmzy78C/J477r0B/rm3Ur5gDiXBtR8uJ165uuB4XHn85Il7kyZSHNw4bGaNwx5Urs13NRKo+MYSjnc9cQrODJgkEFtu0whE5UZR51aD+7WQKDtodSTV8a5BypIzlcDmDQOzcA46cfNd6zvlfvhHCx9rG27bP3xzls31sCsH47OpVGNMegXPfQK2G5xwt9KmvzPk2M7MdKk/DmQJRzLfCzvIyBDHPrbX+6qr67t/QNIxvnpk9jYj1PtaTCrUpFCLS32ptX1KkEmDNsAeUhRJJgWag7eemQVMbVuK2+sHyS85w5LEUsuKCDJlWE48OEyUqwNPr1fqcgS09UHDz2Am5GyCaG/fOIci0qComb7CP0LbKJrfgllIzlky6fMgPgXz18xoMrzb+yls2ljYGEnQUqiekrYUmn05nNJEe2a3lg44GAeE4nNlCme668J8NUHOlv1vrTz+0pftWN5iJ7dFxME0JVO7Jzv1bVnsHItwYlsxbYmkCF6M+Rqt7u0LjVgcLBN9ZNqIxDj1oh1b5suS+cS8BdNaBDOr4UjITvdN7ztwf+4kLmxiqq+LhwceXs5cv3Hk8dMbiAVNgLFtkto2S1sCie5w1Sb++aqJTV83lwYg0wFTY1MaNzTZPidKswsHljemot8Xm+5gwr+7I4ZPjr9N6pM77mmfRoCdlDS7t/cCtbgzUWCZ9LKjI7oG2D2lazjt0NSYm0y1AWR6frFzLk0mHcWr1iP2hfoGSwPHuMNOA1M/ZMTGX1rO+fjpbe/ba+RWN+bBPKiUqFqPWQ9SnwB6dj3yqiQ1C4c6KCVo9pzIzBZzVHrHUlICT3qi0kMzQzVt6thCEpdaXb1vJ6+axbqqqZfLUnh6c01aEgU/lKSZCXRgznpXdEc2q9Mo6XW+x8Me19qkz3AWtD7sjFf1lwTIhMe2DAyuMd8x1tL616RX3UwBLcrhYmIthZubEzknc2qTOOe+spSC1J5wKYPnmLeCm0lCldNptbq7JCwr1KMChbquTFNG15X15AevqDJNmep5A8JOJJ5BMIud8Ff8MBiJNqiPu8/fKUv3mm/rQr2siCSggXrLz6ADI+tMBiJtHqWU/KTM2hjNzfJhp/umHdyfA+WQVrJw+coVV3Pi80+ueXJdujoWwG1tm5A0X/n2W2/P+metgmHhjfvObWFn2KTOpVnf7UL92BdeQ3q7rINw+Wb1k7N6hNZBzr9g+4+eX1FTxRf8lDT/YnBQC6bmjPdpYBHXW/bSsY4aPT5e7oxRV0hoB9oG/LLpV8A92rtkG5t3lJfw5CCDn0AH9e1cMEYgGBahJhoQjs9144j2POvDII6qfdts3X7bXd5aOyqe3MbbWv3gmVrNU180fLS742UC5mkyVb7v/OPc8HTpVBLHCutSyM2xLOziznW07uz6K7QzWCnaNcxBxZwAS9GW/TF5bPxGM1ZxzVb0Ru+UAKJ+r7Z72xO+AOIwuKK0cLjjaUEVDtNkIDmukZCAS2GeXJXtqVdLqdRqkSAEQ+rrU5I5GU5TRo42o5kSJPOGX5+eQC2LXDu5r1RqssN6YhCKc0CTmJq8CQ7SjwmOQ2NCSjdbfve0B1exRwfFWlCfF0maliX6MZiG8OXY+EwoDM4qO90z7eB+z9T2aCyL04OHV1RNPH5yYxmZNmFMZ2FFcbWjf3eW8cXUNhlfHOo6vnPAukO2HL7r3PhtOmOlR3BTe3Y8XKs/dbuwjczTFrwMOQBoIH1eXmtjAU40O7uiHdND0zC8b2Rson90NBf4l5GGNdrU1MFnVYvwxO6wdFvzIRqOYNLHSAYmKqR1v5ZaP/jzrcJnJ9+ccWZueOh+GXCLOWxjSy+72YrBba1R/uAQOdrLvX/qqj0sCgdLn69Fq4WuY9L/lCxRS0jKlZ5hLgAkJ1NRn9wbfEraTrILO7DUiEXvbSxDvYJpCDV39HFVS9hSSmc+I7tc+DLE3In1AcJ40ln0d8MbEQ+fHIxTEcYVfeILNJzlVEybkZ1RUSImXGj5ArJldxPYhB2WRZv2wcwbZlvPyU0lCvWpUqVAVkopnI6F+ZC5mE1rUMKujWWnW9yHIqRrizMfmAdvX56SRWR4O0jdV0UEcu5aD8XNHh7O0Na1z8cs4TzYV0r0Z3N2DYaC9Oy8BTu9LdrB/TlQJMqYDxOvPDhwWitPro/mFez3hDPLdmN2WUX6VhabuhfsgKQOXO2h5rQyAtso2oiLJR3wlbsAuYnmm3LGL/3TiCEuFYyPNa2Do2WTf0I8Gst0buG8NgJIsR/q6EW9hVd71x3OblEFub179EsOwr7ph/p680RIKQxtOeMCTBLrLY13KzQwjnOvO9BLV/v6zhnvjf7SKEtBmle5bOt4q+Nab7d7o9SKe8YHgyDOmFXLL17V5mgSsXziHnieUgo5HUsNq/38+uqOaE1LMLzTAcK0D4ImJak5Ua1uL+nSIUjKPk/d+QvswJehDaO9XPGY+aotF3yM48ikWb3i/152Q6xhZtXhwRbehzOYfpckbar97MAbTLhiduXsHGNXRkjL0RCaE1Xl4eVMIvH4+shpWVvyoTxbWlvLYV+CtXNnUmWSxMXseeudaTAmKSHqiW20Mznm83HGmPhEyim1kMSQ3rfMpZs8mge+2e2ziKcM7g6MkachpmI7Ma+tCWeQRPz0uh3dnwft4P4cyE6ggvli5uFV4nRzw+ObldKkOd8iJCSvvqHEelPn/I20nR5mN9HUXMFZD7BCiBZbsA9RlQG9BhHwDpxv9/ZSb3+nbI99PScd7pU7rsU/43cdH7s6PnpF6EAuY3H24JYX6SD3zHrd/a2Damc6DKT7e0eAD4kNGUOItq9JLsaG1/fIXOhZx3QhsjMtSpeiWzpUkU3tnb8YnJyklz4yXqHuF7sj0vOO3sxTTqgkFrET3EJJEr0sdPVumCHy+SbdJGarS+SqP4iwuvSqaEvIMoe0OEyJMcHJJhzO0bR4rPUG2GNoXNNTtiuhMWOpvYUNc6kaDoMh0ds9ZWBg1EEz4vFTnG5WaySLtjA5cbODMy/zNCMiTBlWPz3t5lTM2121nbkuSSxJTXEHXBHL849rULyhOQlaKmWN5ENip8L5nJMYI/evGOSBJnGrQFnd66/tK7IB6LVqA+fm/e5zR0qlirD6vhfPtOQ33n/nZ24E1WdqEXd6O7SD+32TgtZKyhNXFwfmDE+eXnNaSpPi/LYtSd80I5tXV7HSwLqlG9VBUpIBLJRuUzzfa8+BM/7tgsutx7b1HKXMscHx1O0VGsxHqKYZ27RByTOQ96drqVQt28oJbXPbSuDDrrHZMM4BbtsxOvwNb/DWNI3QKKeQvFq13XQSAK8etzxAtiieC3/LFNzqp/MKDaAf0tLIAHT+r78rJM0wHLc+9/lTa4R0SZs3LWxJaVkVI1FJEuwk4lK9HanVNRLxmOo7EVEarW7ef5ETvkl80JzrEgaM1dXF5nQVUrFsgD4YyUggU0Ni95CIDbPjQCxNdXw2/za8srNS59N3WFRJ1RmsqI395NTbXLEjcFMSO3FtWFfZbf+lrCyrhYAh3nY1U0XFksEkDy9Ttaxvpdo4REx4nIKofkjLnBNJKyvKYZ64nDOPrpfB/6HPCXGbvmKqe/NJiZSz4sfUGjOAv0eIky7blAfg6mIm5czN8dS83iPs00wx2z4NB9AWlohpiyQ0izvdK+3gfu/kXsIJHl7OKPD0yTXF42E6rCloGv73v+Mk9w06JLj4rienCNvkdmVsQEI2H85ednb/rQK2i9m2sQ58t5iAVsF2c3vCylHfhHtpmxC/Vk4UY6rEEl/7RtNtuUNZ4w1gEm+z059hfQDQ0PkKTbVa+m2gYyrZqL9s+ZIAzmhf3DfWk620ft7dcut6RwYbwu6Qd4sx24zn0Aeiw72h4lcHFatnPz/e69HqO4y7QE1CO30Hi1PG09dq68yQdLV1etf8JqYkyBTADrUEA2G5y5FE8pwG2n6NvTHoOAL8HOjb1Bs4tVDfg2kVktetgc5m8d1eDO31wwPhUJaSAWBdCySTlm25WjZBBme2sFdn90RflsJpKR4ya6FsKmZfD+92cYYETT39tAhzzhwuDgjKcrRDqKp5OjJN2fpPUj8DIAsnnFmqOqwZ3G/CetXMJvaeUU2eiON8dRA6rHeuknB1OfFVgUdPbogjbn22t4yQ1feqJGE6qLSdxNfTb5kbbKe3TDu43zM16U8mHl4kWBduHq/G+TasDpDwjdUl7ZazfCTfcSXAwm2i6ptUCi57eLItk2HBjBtdAE9s5lvAPSslhLpB2NlsvA5cG7t5gOVQZnumOcEF5AxvbDv3UBWxzcounUXda0iGd/QbIMnsscEIqVcupKJNW1usXEd9daPsEN/Qvm5MyCbT0Jh3wJyxkuA6+TtqOALJCDb0/gtQG+fNud9deBhumTq9VV40wEKxk8fch+e9OW5pXb19Hi/uKuWUBM0h+UeopmkpIl47TiiyEw0NUKuEP74DhCqT2PuPE2ipHu9v6ug0pLdtc0ppJiiT9OPQlnBgk8Y8arXIikhdi+BSdwyGAU+TwAfGSDbjEWrogDLYeM77lKnuKW/58v1b7SBZtZDIzZNcsAQ14WymaslmzKbdTQHJeYNKbWlac7YVcHNzYyf1rcWPcTWJO4vZwR8/LeZzEH2hPba8rXWvfzCdEdERQBurdlwmMjRPVfjy0yOPjgttQbXFOToFSntfEguxi5wAzRF4O1V3uifawf25kJCnicNBWI8nbq5X37y7G9AgK9kjHQ23yBNSH2LAPoDVxo4exXVkhwC02MN0hEhXz+q53D/Kcdsld9cCDHxrIfLPWKW9SdJ2CZP6Ni/tb/INbXQIGy24Yy9aHv+hygG2dfhMdNW2giPjYpjYld+VLm2PwNnu3dSov3vE4IrFlt+pdmyqZx+RketQZ9pi4x3rPGzQt+o31qUxNPZ9FuEwpW5bB9RBvmptIVhJFZFKxcB9cpRse7eDWWM8FJJ6pruhoWPKhDBvRLPx+qwYkE5i/RVhpMQ8V/furn0uT4fcGJ/DlJoDmmJhZUtEVUQ/BMC3+TTmB+igOw5L62u/b+S9YqxSEtZSUISrQ4baYwuaC2motn1oIyVuZHLLSbicM2upLA56jbnBmICcYHWburXPQPPqwcH8Hjz8zxwTaetgXezAGpsnVgHFmFY/WPEWtax1vroGJYzNjewH/Lg6Tf0ZOzvAV6X4yohY+zZnfNzTkAVvYCR3ul/awf15kFjMr8zC9ZOF42n1zbqLphrqS9+h2zfnGzm0s8vraouiHWHJAKqxggL8BlAbpefGTmhgvbbv250NpbeMgwxlBAW7wADw47fdc1oGFW2vSGMkGveuA19hDEgaQDls7b0SW8/sKKKdoR59cAe4joDdWtaEkDuAfWTE/PbR5h3q96j/aMvVcZz6S9uG2661Th60FKr9PK4NL9gdKnXkHjTgaOgTH/AJpcpw9vow4SzcSlrfpWRq4DYHVV0S7zZzq1+0Q1s5IW2nNsLamJ2gJEJOech4Z/fVCuvgRR+x0lF2xDH2Z2j5HUgZnex0u2hfaLa0DWFoBoRx/mzz1TRDBm1QNOZ5X09JTPVdmkpaG7MhuBpazSk04f4CbmK7mDM5C2meWJYbxF3pxW3XaynU0ueKFpfiRVrkBQJpEjLOAHiVW+x66nZwDTNJhB9qb25bYps12Zka40N16BWX8uOxJHbegwsMdTjAJjQCEQLamZ3eH3etz53eHu3gfu9kzkXzRUKqcvP4xNGDb7taOeSBtow2k3sEUBHxjTVSUtbu4NUQJX7iotuaPc96nP41AlaDPQeCJoU2QOsSC1HXzmp3AJPAIm+RjHhxBuj0xf0s0vFXA+me5GTcB2471Q2lhMQx9nF/w5bOB2AjyW03+C6RDV+rmOTaxGm7Nr5q9FW4s913xOJ3vkTIdGDFix95jTiEBhn7cCzKNtp1LaSczHnKx7FJvtUTwAAXSZjmyWLH3VM6QD5isyv2OefIQy+obnsu7OQxX1NMGIQ5CXXCs9t5PUqk6cXqE8AuJuWmJFSx/h6ZL/V+EswKMpFQrX4+fTjvpY16PMao5/Rv7GUbCOvLHnffvnHEm5I9H+aCyKE/Y21dtYf5EUyPz6MpC3MW1nW1M9dTtZh27Az2UqLPEoc5IVk4HVfqajH9Zp5wHwUH/LpWm48pNSc3kCERzW0pufHEPsZmG8+spQDqjnZW75gfVc1ZMTzzzawQYxBMXjAFpr1IHqoXpp1gGs5NSDvdD+3g/hwoJQP3uirXTxcWCxYmrFyxyJsdvUEQfcPWzuGmlMIo1lSngDnfDMAeKtHYohpGStu3mkp2BPf2UjFJe5OlRthw1ueq8Sahhh1tUGuPIBx8yHDxFoidk4pt8LWJe9pzsatYViwsZKid/MagQpWRhRrbOjItXsMBKLYVu82JtKrHM35hE68rwRA5xGnvt3jo7MqmLwywt/U24Orq6fgu4objxeNcGuXPqti54gpFhNVzjR8uZkiZ01KY1hWAPNkZ47MqxRs7ZQOYm1W5PhW0rExJLM2qg3ufvz0sqofQ9Q09J0+yspTBdGD57b22Jil6HyT3CYiT6izU20t0rUEwvUlMfVw1oVRqcW1DhawhGVvnJcEP3xkcDp9Bo8lBPOYwDedIWFmgmiyELXlugxpx453ZnabEnM2rflm1Sfi1VqpmY9ZUeXh1wcMHB54eVx49vgGxFLGrq+LToD2p1UA+SY+ht1aGil3b3mPt6RNlPNylAlLbCUQW6oj5NrTkM66Gn3xuKEKS0uevD7Sd/WSNnkU4TUJdztISxRze6V5pB/d7JpG+MS4rPD0u1Foa5JzLkRsAOccRpZ0e1UEiNvdBzZVS+6yCJX2J4s5EDtn8PZNPnYuWngKutWlbrS0sBUCO79i8Z/inA/zAiJynoBy6RIVbGax61jivh9ADvF0yFNuNcJQd2RG2ACg9RenYigFkN4yQNp6BEcx6s3q7JAA+Hq6Dqj1xpko3xqhbFlwWVQZtylB37ZvxCO4hId7lilFRjg7woXK9TImLSXi6FsqyGqA8uELnmadPn5o3uJ/jLTUxrcKKnQFf1MMGQ7J1FbBqnO0+JIShawiqKmX1OilIhHplZyD8PlziNee5NHi8d9t+MAthm7c+cO90P5mnSEiQ4dg3YPgd5TSv/PZdTMLxwXAq9LA2d3StxZ6b5tyc/kSUVSsJk/QR8xN4cr20aWR9YXVba2GeJz7y4QdMU+YLX3nM9fWJ4Oqye9SvVTmdeux7xN0bo2ProtTQNrkToZeRYn3FXPP+wetcVJvdvbo0kf3Ut8gnHxNzSmaWKHEZabkv2vkCItysxbQ8STzZjpXRzFk73Svt4H7PNE0T3/xNv50PffABlxczWi95cFqIHNJrHAntCyGHxDngaQeEyI9trrPLurKGpCNDaspR1UjkAi8Dl20FG/4JDx4+oFblwx963erCiL9bKVvvWnR3Sv8buN4+NQJnMx9s79Lxr3fOBz/wGq+99gqn07J5fQdh6cyCewgNbgd2yErdwvotAUFoMejPorsdfvoTX/fh17m8OLT+bs8NjQqwau8czCithrqt3nmd+vxot9O884kwp61T20gf/vCHSCkTIU1obSllV1eXJoEPPrximWZunj6Fs/PQjUnoIWiqyiyWelYGxNh4Egxj0kA4+C6fRCFB5ynz6muvknKPp0/JzlEM3o00gLvSYt5HvN6Yn2Id0MMoQxsVB/NEZepw7zgAWtnQ4TBzdXXJ1YMrf+fgoidbLU71+mVg9rqt1c6cDjNVcZOIABeHiddevWJZK288vuajV6952W6akNQOdgkHuCmH8c372MfmlVcespbCRz7yoWfMo/Bc58ykRhujkXHU8fpQXk/R273j2/h5mGAcM2yZ6WzgY+X9tt/2EXa6X5I9vvB+SUQeAZ9+0fX4GukjwBdedCW+Btrr/c7SXu93nt6tdX8r9f4dqvp1z7My7yfaJff7p0+r6re96Ep8LSQiP/turPte73eW9nq/8/Rurfu7td7vBbor1HGnnXbaaaeddnoX0w7uO+2000477fQeox3c759+5EVX4G3Qu7Xue73fWdrr/c7Tu7Xu79Z6v+tpd6jbaaeddtppp/cY7ZL7TjvttNNOO73HaAf3nXbaaaeddnqP0Q7u90gi8p0i8mkR+YyIfP+Lrs9IIvKNIvJXReT/EZG/KSL/tl//QRH5rIh8yn++e3jmP/C2fFpE/tALrPuvisgveP1+1q99SER+SkR+yf++7tdFRP5Lr/ffEJFvfUF1/oeHPv2UiLwhIn/iZe1vEfkzIvJ5EfnF4dpb7mMR+V6//5dE5HtfUL3/ExH52163vywiH/Tr3yQi10Pf//DwzO/2OfYZb9tzTZn2jHq/5bnxTu85z6j3nx/q/Ksi8im//tL09/uS+sk8+8/b+cFO0Pxl4JuBA/DXgW950fUa6vdR4Fv986vA3wG+BfhB4N+74/5v8TZcAB/3tuUXVPdfBT5ydu0/Br7fP38/8EP++buB/wlLhvXtwM+8BH2fgV8HfsfL2t/A7we+FfjFr7WPgQ8Bv+J/X/fPr7+Aen8HMPnnHxrq/U3jfWfl/F/eFvG2fdcLqPdbmhsvYs+5q95n3/9nwH/4svX3+/Fnl9zvj34P8BlV/RVVPQE/AXzPC65TI1X9nKr+vH9+BPwt4GNv8sj3AD+hqkdV/bvAZ7A2viz0PcCP+ecfA/7J4fqPq9FPAx8UkY++iAoO9AeAX1bV//dN7nmh/a2q/wfwpTvq9Fb6+A8BP6WqX1LVLwM/BXznO11vVf1JVV39358GvuHNyvC6v6aqP62GPD9Ob+tzoWf097PoWXPjHd9z3qzeLn3/88B/92ZlvIj+fj/SDu73Rx8D/v7w/6/x5uD5wkhEvgn4XcDP+KV/01WYfyZUr7xc7VHgJ0Xk50Tk+/za16vq5/zzrwNf759fpnoHfYLthvey93fQW+3jl7EN/yomGQZ9XET+moj87yLy+/zax7C6Br3Ier+VufGy9ffvA35DVX9puPay9/d7lnZwf5+RiLwC/A/An1DVN4A/DfwDwD8CfA5Tq71s9HtV9VuB7wL+DRH5/eOXzv2/lDGdInIA/jDwF/zSu6G/b9HL3MfPIhH5Aex8lj/nlz4H/HZV/V3Avwv8tyLy2ouq3x30rpwbA/0LbJnYl72/39O0g/v90WeBbxz+/wa/9tKQiMwYsP85Vf1LAKr6G6paVLUC/zVdFfzStEdVP+t/Pw/8ZayOvxHqdv/7eb/9pam303cBIZvncgAAAgtJREFUP6+qvwHvjv4e6K328UvTBhH5V4B/HPiXnDHB1dpf9M8/h9mr/yGv46i6fyH1/hrmxsvU3xPwTwN/Pq697P39Xqcd3O+P/m/gd4rIx11a+wTwyRdcp0ZuD/tR4G+p6n8+XB/t0f8UEF6wnwQ+ISIXIvJx4HdiTjDvKInIQxF5NT5jzlK/6PULb+zvBf6Kf/4k8Efdo/vbga8OquUXQRtp5mXv7zN6q338PwPfISKvu0r5O/zaO0oi8p3Avw/8YVV9Olz/OhHJ/vmbsT7+Fa/7GyLy7b5O/ii9re9kvd/q3HiZ9pw/CPxtVW3q9pe9v9/z9KI9+t5LP5gX8d/BONQfeNH1Oavb78XUqn8D+JT/fDfwZ4Ff8OufBD46PPMD3pZP84K8WTFP4L/uP38z+hX4MPC/Ar8E/C/Ah/y6AP+V1/sXgG97gX3+EPgi8IHh2kvZ3xgD8jlgwWygf/xr6WPMxv0Z//ljL6jen8Fs0THPf9jv/Wd8Dn0K+HngnxjK+TYMTH8Z+FN49s53uN5veW6803vOXfX26/8N8K+f3fvS9Pf78WdPP7vTTjvttNNO7zHa1fI77bTTTjvt9B6jHdx32mmnnXba6T1GO7jvtNNOO+2003uMdnDfaaeddtppp/cY7eC+00477bTTTu8x2sF9p5122mmnnd5jtIP7TjvttNNOO73H6P8DkCSas+GfIHoAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hB7WdkAyE4rb"
      },
      "source": [
        "# Build & Train the Network Model\n",
        "\n",
        "This code is based upon the PyTorch [Transfer Learning for Computer Vision Tutorial](https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html) and related documentation for image classificaiton with PyTorch."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tCqmuRgbPYem"
      },
      "source": [
        "# Define Simple Newtork model\n",
        "class SimpleNeuralNetwork(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(SimpleNeuralNetwork, self).__init__()\n",
        "        self.flatten = nn.Flatten()\n",
        "        self.linear_relu_stack = nn.Sequential(\n",
        "            nn.Linear(224*224, 512),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(512, 512),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(512, 10)\n",
        "        )\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.flatten(x)\n",
        "        logits = self.linear_relu_stack(x)\n",
        "        return logits"
      ],
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MeVQZG7ynqLF"
      },
      "source": [
        "def getNetworkModel(model_type=\"ResNet\",step_size=7,gamma=0.1,optimizer_type=None,scheduler_type=\"StepLR\"):\n",
        "  \"\"\"Create the parameters necessary for the RESNET model\"\"\"\n",
        "  # Load in the ResNet Model\n",
        "  if model_type == \"ResNet\":\n",
        "    model = torch.hub.load('pytorch/vision:v0.10.0', 'resnet18', pretrained=False)\n",
        "  elif model_type == \"AlexNet\":\n",
        "    model = torch.hub.load('pytorch/vision:v0.10.0', 'alexnet', pretrained=False)\n",
        "  elif model_type == \"Simple\":\n",
        "    model = SimpleNeuralNetwork()\n",
        "  else:\n",
        "    return None #TODO: Try other Models\n",
        "\n",
        "  # Determine the Number of Features\n",
        "  # num_ftrs = modelResNet.fc.in_features\n",
        "\n",
        "  # Size of each output sample is generalized to nn.Linear(num_ftrs, len(class_names))\n",
        "  # model.fc = nn.Linear(imgDim*imgDim, len(class_names))\n",
        "\n",
        "  # Push to GPU\n",
        "  model = model.to(device)\n",
        "\n",
        "  # Set the Loss Criterion\n",
        "  criterion = nn.CrossEntropyLoss()\n",
        "\n",
        "  # Observe that all parameters are being optimized\n",
        "  if optimizer_type == \"Adam\":\n",
        "    optimizer = torch.optim.Adam(model.parameters(), lr=eta) #https://pytorch.org/docs/master/generated/torch.optim.Adam.html\n",
        "  elif optimizer_type == \"SGD\":\n",
        "    optimizer = torch.optim.SGD(model.parameters(), lr=eta)\n",
        "  else:\n",
        "    optimizer = None\n",
        "\n",
        "  # Decay LR by a factor of 0.1 every 7 epochs TOO AGGRESSIVE FOR RESNET: 0.01  epoch80 0.001 epoch120 0.0001 (Mayve start at 60 or 40 for smaller data set)\n",
        "  if scheduler_type == \"StepLR\":\n",
        "    scheduler = lr_scheduler.StepLR(optimizer, step_size=step_size, gamma=gamma)\n",
        "  else:\n",
        "    scheduler = None\n",
        "\n",
        "  return model, criterion, optimizer, scheduler"
      ],
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dQgMQvECmOT3"
      },
      "source": [
        "def train_model(model, criterion, optimizer, scheduler, num_epochs=5):\n",
        "    since = time.time()\n",
        "\n",
        "    best_model_wts = copy.deepcopy(model.state_dict())\n",
        "    best_acc = 0.0\n",
        "\n",
        "    # Store training values in lists\n",
        "    training_acc = []\n",
        "    training_loss = []\n",
        "    validation_acc = []\n",
        "    \n",
        "    for epoch in range(num_epochs):\n",
        "        print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
        "        print('-' * 10)\n",
        "\n",
        "        # Each epoch has a training and validation phase\n",
        "        for phase in ['train', 'test']:\n",
        "            if phase == 'train':\n",
        "                model.train()  # Set model to training mode\n",
        "            else:\n",
        "                model.eval()   # Set model to evaluate mode\n",
        "\n",
        "            running_loss = 0.0\n",
        "            running_corrects = 0\n",
        "\n",
        "            # Iterate over data.\n",
        "            for inputs, labels in dataloaders[phase]:\n",
        "                inputs = inputs.to(device)\n",
        "                labels = labels.to(device)\n",
        "\n",
        "                # zero the parameter gradients\n",
        "                optimizer.zero_grad()\n",
        "\n",
        "                # forward\n",
        "                # track history if only in train\n",
        "                with torch.set_grad_enabled(phase == 'train'):\n",
        "                    outputs = model(inputs)\n",
        "                    _, preds = torch.max(outputs, 1)\n",
        "                    loss = criterion(outputs, labels.long())\n",
        "\n",
        "                    # backward + optimize only if in training phase\n",
        "                    if phase == 'train':\n",
        "                        loss.backward()\n",
        "                        optimizer.step()\n",
        "\n",
        "                # statistics\n",
        "                running_loss += loss.item() * inputs.size(0)\n",
        "                running_corrects += torch.sum(preds == labels.data)\n",
        "\n",
        "            if phase == 'train':\n",
        "                scheduler.step()\n",
        "\n",
        "            epoch_loss = running_loss / dataset_sizes[phase]\n",
        "            epoch_acc = running_corrects.double() / dataset_sizes[phase]\n",
        "\n",
        "            print('{} Loss: {:.4f} Acc: {:.4f}'.format(\n",
        "                phase, epoch_loss, epoch_acc))\n",
        "\n",
        "            # Store Performance in Arrays\n",
        "            if phase == 'train':\n",
        "              training_loss.append(epoch_loss)\n",
        "              training_acc.append(epoch_acc.item())\n",
        "            else: # must be validation\n",
        "              validation_acc.append(epoch_acc.item())\n",
        "\n",
        "\n",
        "            # deep copy the model\n",
        "            if phase == 'test' and epoch_acc > best_acc:\n",
        "                best_acc = epoch_acc\n",
        "                best_model_wts = copy.deepcopy(model.state_dict())\n",
        "\n",
        "        print()\n",
        "\n",
        "    time_elapsed = time.time() - since\n",
        "    print('Training complete in {:.0f}m {:.0f}s'.format(\n",
        "        time_elapsed // 60, time_elapsed % 60))\n",
        "    print('Best Test Acc: {:4f}'.format(best_acc))\n",
        "\n",
        "    # load best model weights\n",
        "    model.load_state_dict(best_model_wts)\n",
        "    return model, training_acc, training_loss, validation_acc\n",
        "\n"
      ],
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o8uSw0h0o0XX"
      },
      "source": [
        "def plotResults(training_loss,training_acc,test_acc,titleStr=\"\"):\n",
        "  \"\"\"Plots the training_loss, training_acc, and test_acc with the provided title\"\"\"\n",
        "  # Create Figure\n",
        "  plt.figure(figsize=(15,3))\n",
        "  plt.suptitle(titleStr, fontsize=15)\n",
        "\n",
        "  # Training Loss Plot\n",
        "  plt.subplot(131)\n",
        "  plt.plot(training_loss)\n",
        "  plt.xlabel('Epoch', fontsize=10)\n",
        "  plt.ylabel('Training Loss', fontsize=10)\n",
        "  plt.grid(True)\n",
        "\n",
        "  # Training Accuracy Plot\n",
        "  plt.subplot(132)\n",
        "  plt.plot(training_acc)\n",
        "  plt.xlabel('Epoch', fontsize=10)\n",
        "  plt.ylabel('Training Accuracy', fontsize=10)\n",
        "  plt.grid(True)\n",
        "\n",
        "  # Test Accuracy Plot\n",
        "  plt.subplot(133)\n",
        "  plt.plot(test_acc)\n",
        "  plt.xlabel('Epoch', fontsize=10)\n",
        "  plt.ylabel('Test Accuracy', fontsize=10)\n",
        "  plt.grid(True)\n",
        "\n",
        "  # Adjust Dimensions\n",
        "  plt.subplots_adjust(wspace=.5)"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "I5YXgFjKpCim",
        "outputId": "492ccc57-6a07-4252-ab77-040213cfb23c"
      },
      "source": [
        "# Get a Clean Model for No Augmentation\n",
        "model, criterion, optimizer, scheduler = getNetworkModel(model_type=model_type,step_size=step_size,gamma=gamma,optimizer_type=optimizer_type,scheduler_type=scheduler_type)\n",
        "\n",
        "# Train the Network\n",
        "model_trained, train_acc, train_loss, test_acc = train_model(model, criterion, optimizer,scheduler, num_epochs=epochs)\n",
        "\n",
        "# Plot the Results\n",
        "plotResults(train_loss, train_acc, test_acc,titleStr=\"ResNet for Crop Detection\")\n",
        "\n",
        "# Save Model\n",
        "if do_save_model:\n",
        "  torch.save(model_trained,path_to_saved_model+\"model.pth\")"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 0/99\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Using cache found in /root/.cache/torch/hub/pytorch_vision_v0.10.0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "train Loss: 2.5742 Acc: 0.1345\n",
            "test Loss: 2.8880 Acc: 0.0126\n",
            "\n",
            "Epoch 1/99\n",
            "----------\n",
            "train Loss: 2.0974 Acc: 0.2599\n",
            "test Loss: 2.5163 Acc: 0.0388\n",
            "\n",
            "Epoch 2/99\n",
            "----------\n",
            "train Loss: 1.7980 Acc: 0.3557\n",
            "test Loss: 2.2316 Acc: 0.2510\n",
            "\n",
            "Epoch 3/99\n",
            "----------\n",
            "train Loss: 1.4957 Acc: 0.4393\n",
            "test Loss: 2.2959 Acc: 0.1260\n",
            "\n",
            "Epoch 4/99\n",
            "----------\n",
            "train Loss: 1.1735 Acc: 0.5647\n",
            "test Loss: 1.9272 Acc: 0.2926\n",
            "\n",
            "Epoch 5/99\n",
            "----------\n",
            "train Loss: 0.9855 Acc: 0.6430\n",
            "test Loss: 4.6334 Acc: 0.0523\n",
            "\n",
            "Epoch 6/99\n",
            "----------\n",
            "train Loss: 0.7386 Acc: 0.7256\n",
            "test Loss: 2.1159 Acc: 0.2636\n",
            "\n",
            "Epoch 7/99\n",
            "----------\n",
            "train Loss: 0.6096 Acc: 0.7650\n",
            "test Loss: 2.0594 Acc: 0.3140\n",
            "\n",
            "Epoch 8/99\n",
            "----------\n",
            "train Loss: 0.5530 Acc: 0.7755\n",
            "test Loss: 5.4095 Acc: 0.0862\n",
            "\n",
            "Epoch 9/99\n",
            "----------\n",
            "train Loss: 0.4912 Acc: 0.8092\n",
            "test Loss: 2.2790 Acc: 0.2859\n",
            "\n",
            "Epoch 10/99\n",
            "----------\n",
            "train Loss: 0.4577 Acc: 0.8189\n",
            "test Loss: 1.8198 Acc: 0.3052\n",
            "\n",
            "Epoch 11/99\n",
            "----------\n",
            "train Loss: 0.3616 Acc: 0.8466\n",
            "test Loss: 1.9860 Acc: 0.2926\n",
            "\n",
            "Epoch 12/99\n",
            "----------\n",
            "train Loss: 0.2701 Acc: 0.8881\n",
            "test Loss: 2.2706 Acc: 0.3430\n",
            "\n",
            "Epoch 13/99\n",
            "----------\n",
            "train Loss: 0.2951 Acc: 0.8783\n",
            "test Loss: 1.9150 Acc: 0.4021\n",
            "\n",
            "Epoch 14/99\n",
            "----------\n",
            "train Loss: 0.3135 Acc: 0.8790\n",
            "test Loss: 2.4951 Acc: 0.3207\n",
            "\n",
            "Epoch 15/99\n",
            "----------\n",
            "train Loss: 0.2542 Acc: 0.9032\n",
            "test Loss: 2.6969 Acc: 0.2490\n",
            "\n",
            "Epoch 16/99\n",
            "----------\n",
            "train Loss: 0.2444 Acc: 0.8995\n",
            "test Loss: 2.0036 Acc: 0.4196\n",
            "\n",
            "Epoch 17/99\n",
            "----------\n",
            "train Loss: 0.2229 Acc: 0.9090\n",
            "test Loss: 2.4160 Acc: 0.3547\n",
            "\n",
            "Epoch 18/99\n",
            "----------\n",
            "train Loss: 0.1875 Acc: 0.9245\n",
            "test Loss: 2.4297 Acc: 0.3798\n",
            "\n",
            "Epoch 19/99\n",
            "----------\n",
            "train Loss: 0.1470 Acc: 0.9403\n",
            "test Loss: 2.4597 Acc: 0.3779\n",
            "\n",
            "Epoch 20/99\n",
            "----------\n",
            "train Loss: 0.1720 Acc: 0.9309\n",
            "test Loss: 2.1918 Acc: 0.4312\n",
            "\n",
            "Epoch 21/99\n",
            "----------\n",
            "train Loss: 0.1644 Acc: 0.9339\n",
            "test Loss: 2.1096 Acc: 0.4845\n",
            "\n",
            "Epoch 22/99\n",
            "----------\n",
            "train Loss: 0.1465 Acc: 0.9444\n",
            "test Loss: 2.7401 Acc: 0.3101\n",
            "\n",
            "Epoch 23/99\n",
            "----------\n",
            "train Loss: 0.1237 Acc: 0.9531\n",
            "test Loss: 2.3628 Acc: 0.4002\n",
            "\n",
            "Epoch 24/99\n",
            "----------\n",
            "train Loss: 0.1323 Acc: 0.9491\n",
            "test Loss: 2.3461 Acc: 0.4215\n",
            "\n",
            "Epoch 25/99\n",
            "----------\n",
            "train Loss: 0.1344 Acc: 0.9504\n",
            "test Loss: 2.3179 Acc: 0.4438\n",
            "\n",
            "Epoch 26/99\n",
            "----------\n",
            "train Loss: 0.1212 Acc: 0.9514\n",
            "test Loss: 2.3078 Acc: 0.4535\n",
            "\n",
            "Epoch 27/99\n",
            "----------\n",
            "train Loss: 0.0999 Acc: 0.9609\n",
            "test Loss: 2.2880 Acc: 0.4651\n",
            "\n",
            "Epoch 28/99\n",
            "----------\n",
            "train Loss: 0.1149 Acc: 0.9548\n",
            "test Loss: 2.2935 Acc: 0.4496\n",
            "\n",
            "Epoch 29/99\n",
            "----------\n",
            "train Loss: 0.0995 Acc: 0.9646\n",
            "test Loss: 2.2520 Acc: 0.4612\n",
            "\n",
            "Epoch 30/99\n",
            "----------\n",
            "train Loss: 0.0832 Acc: 0.9686\n",
            "test Loss: 2.2579 Acc: 0.4535\n",
            "\n",
            "Epoch 31/99\n",
            "----------\n",
            "train Loss: 0.0942 Acc: 0.9666\n",
            "test Loss: 2.2385 Acc: 0.4545\n",
            "\n",
            "Epoch 32/99\n",
            "----------\n",
            "train Loss: 0.0824 Acc: 0.9676\n",
            "test Loss: 2.2594 Acc: 0.4690\n",
            "\n",
            "Epoch 33/99\n",
            "----------\n",
            "train Loss: 0.0658 Acc: 0.9788\n",
            "test Loss: 2.2498 Acc: 0.4612\n",
            "\n",
            "Epoch 34/99\n",
            "----------\n",
            "train Loss: 0.0655 Acc: 0.9784\n",
            "test Loss: 2.2974 Acc: 0.4651\n",
            "\n",
            "Epoch 35/99\n",
            "----------\n",
            "train Loss: 0.0633 Acc: 0.9798\n",
            "test Loss: 2.2523 Acc: 0.4671\n",
            "\n",
            "Epoch 36/99\n",
            "----------\n",
            "train Loss: 0.0577 Acc: 0.9828\n",
            "test Loss: 2.2914 Acc: 0.4767\n",
            "\n",
            "Epoch 37/99\n",
            "----------\n",
            "train Loss: 0.0707 Acc: 0.9777\n",
            "test Loss: 2.3409 Acc: 0.4709\n",
            "\n",
            "Epoch 38/99\n",
            "----------\n",
            "train Loss: 0.0599 Acc: 0.9804\n",
            "test Loss: 2.3108 Acc: 0.4758\n",
            "\n",
            "Epoch 39/99\n",
            "----------\n",
            "train Loss: 0.0591 Acc: 0.9784\n",
            "test Loss: 2.2970 Acc: 0.4787\n",
            "\n",
            "Epoch 40/99\n",
            "----------\n",
            "train Loss: 0.0524 Acc: 0.9835\n",
            "test Loss: 2.3721 Acc: 0.4806\n",
            "\n",
            "Epoch 41/99\n",
            "----------\n",
            "train Loss: 0.0542 Acc: 0.9828\n",
            "test Loss: 2.3423 Acc: 0.4816\n",
            "\n",
            "Epoch 42/99\n",
            "----------\n",
            "train Loss: 0.0551 Acc: 0.9815\n",
            "test Loss: 2.3016 Acc: 0.4806\n",
            "\n",
            "Epoch 43/99\n",
            "----------\n",
            "train Loss: 0.0602 Acc: 0.9777\n",
            "test Loss: 2.3502 Acc: 0.4855\n",
            "\n",
            "Epoch 44/99\n",
            "----------\n",
            "train Loss: 0.0487 Acc: 0.9838\n",
            "test Loss: 2.4034 Acc: 0.4787\n",
            "\n",
            "Epoch 45/99\n",
            "----------\n",
            "train Loss: 0.0456 Acc: 0.9869\n",
            "test Loss: 2.4042 Acc: 0.4855\n",
            "\n",
            "Epoch 46/99\n",
            "----------\n",
            "train Loss: 0.0593 Acc: 0.9808\n",
            "test Loss: 2.3560 Acc: 0.4893\n",
            "\n",
            "Epoch 47/99\n",
            "----------\n",
            "train Loss: 0.0499 Acc: 0.9828\n",
            "test Loss: 2.3916 Acc: 0.4913\n",
            "\n",
            "Epoch 48/99\n",
            "----------\n",
            "train Loss: 0.0501 Acc: 0.9838\n",
            "test Loss: 2.3782 Acc: 0.4855\n",
            "\n",
            "Epoch 49/99\n",
            "----------\n",
            "train Loss: 0.0479 Acc: 0.9852\n",
            "test Loss: 2.3911 Acc: 0.4845\n",
            "\n",
            "Epoch 50/99\n",
            "----------\n",
            "train Loss: 0.0447 Acc: 0.9879\n",
            "test Loss: 2.4105 Acc: 0.4903\n",
            "\n",
            "Epoch 51/99\n",
            "----------\n",
            "train Loss: 0.0389 Acc: 0.9916\n",
            "test Loss: 2.4135 Acc: 0.4864\n",
            "\n",
            "Epoch 52/99\n",
            "----------\n",
            "train Loss: 0.0420 Acc: 0.9869\n",
            "test Loss: 2.3754 Acc: 0.4777\n",
            "\n",
            "Epoch 53/99\n",
            "----------\n",
            "train Loss: 0.0411 Acc: 0.9862\n",
            "test Loss: 2.3942 Acc: 0.4913\n",
            "\n",
            "Epoch 54/99\n",
            "----------\n",
            "train Loss: 0.0514 Acc: 0.9825\n",
            "test Loss: 2.3537 Acc: 0.4787\n",
            "\n",
            "Epoch 55/99\n",
            "----------\n",
            "train Loss: 0.0462 Acc: 0.9882\n",
            "test Loss: 2.3912 Acc: 0.4816\n",
            "\n",
            "Epoch 56/99\n",
            "----------\n",
            "train Loss: 0.0406 Acc: 0.9889\n",
            "test Loss: 2.3892 Acc: 0.4874\n",
            "\n",
            "Epoch 57/99\n",
            "----------\n",
            "train Loss: 0.0525 Acc: 0.9838\n",
            "test Loss: 2.3923 Acc: 0.4874\n",
            "\n",
            "Epoch 58/99\n",
            "----------\n",
            "train Loss: 0.0432 Acc: 0.9872\n",
            "test Loss: 2.4182 Acc: 0.4855\n",
            "\n",
            "Epoch 59/99\n",
            "----------\n",
            "train Loss: 0.0466 Acc: 0.9875\n",
            "test Loss: 2.3989 Acc: 0.4816\n",
            "\n",
            "Epoch 60/99\n",
            "----------\n",
            "train Loss: 0.0414 Acc: 0.9895\n",
            "test Loss: 2.4096 Acc: 0.4874\n",
            "\n",
            "Epoch 61/99\n",
            "----------\n",
            "train Loss: 0.0469 Acc: 0.9872\n",
            "test Loss: 2.4402 Acc: 0.4845\n",
            "\n",
            "Epoch 62/99\n",
            "----------\n",
            "train Loss: 0.0368 Acc: 0.9916\n",
            "test Loss: 2.3965 Acc: 0.4922\n",
            "\n",
            "Epoch 63/99\n",
            "----------\n",
            "train Loss: 0.0471 Acc: 0.9848\n",
            "test Loss: 2.3920 Acc: 0.4864\n",
            "\n",
            "Epoch 64/99\n",
            "----------\n",
            "train Loss: 0.0446 Acc: 0.9879\n",
            "test Loss: 2.4404 Acc: 0.4903\n",
            "\n",
            "Epoch 65/99\n",
            "----------\n",
            "train Loss: 0.0414 Acc: 0.9858\n",
            "test Loss: 2.4337 Acc: 0.4922\n",
            "\n",
            "Epoch 66/99\n",
            "----------\n",
            "train Loss: 0.0491 Acc: 0.9848\n",
            "test Loss: 2.3752 Acc: 0.4845\n",
            "\n",
            "Epoch 67/99\n",
            "----------\n",
            "train Loss: 0.0443 Acc: 0.9895\n",
            "test Loss: 2.3922 Acc: 0.4884\n",
            "\n",
            "Epoch 68/99\n",
            "----------\n",
            "train Loss: 0.0437 Acc: 0.9906\n",
            "test Loss: 2.4008 Acc: 0.4884\n",
            "\n",
            "Epoch 69/99\n",
            "----------\n",
            "train Loss: 0.0439 Acc: 0.9858\n",
            "test Loss: 2.4174 Acc: 0.4855\n",
            "\n",
            "Epoch 70/99\n",
            "----------\n",
            "train Loss: 0.0418 Acc: 0.9885\n",
            "test Loss: 2.4172 Acc: 0.4932\n",
            "\n",
            "Epoch 71/99\n",
            "----------\n",
            "train Loss: 0.0373 Acc: 0.9892\n",
            "test Loss: 2.4229 Acc: 0.4903\n",
            "\n",
            "Epoch 72/99\n",
            "----------\n",
            "train Loss: 0.0486 Acc: 0.9842\n",
            "test Loss: 2.4179 Acc: 0.4864\n",
            "\n",
            "Epoch 73/99\n",
            "----------\n",
            "train Loss: 0.0445 Acc: 0.9855\n",
            "test Loss: 2.3846 Acc: 0.4884\n",
            "\n",
            "Epoch 74/99\n",
            "----------\n",
            "train Loss: 0.0426 Acc: 0.9885\n",
            "test Loss: 2.4101 Acc: 0.4942\n",
            "\n",
            "Epoch 75/99\n",
            "----------\n",
            "train Loss: 0.0516 Acc: 0.9838\n",
            "test Loss: 2.4026 Acc: 0.4884\n",
            "\n",
            "Epoch 76/99\n",
            "----------\n",
            "train Loss: 0.0481 Acc: 0.9879\n",
            "test Loss: 2.3939 Acc: 0.4903\n",
            "\n",
            "Epoch 77/99\n",
            "----------\n",
            "train Loss: 0.0424 Acc: 0.9872\n",
            "test Loss: 2.4150 Acc: 0.4797\n",
            "\n",
            "Epoch 78/99\n",
            "----------\n",
            "train Loss: 0.0439 Acc: 0.9882\n",
            "test Loss: 2.4172 Acc: 0.4884\n",
            "\n",
            "Epoch 79/99\n",
            "----------\n",
            "train Loss: 0.0436 Acc: 0.9889\n",
            "test Loss: 2.4133 Acc: 0.4884\n",
            "\n",
            "Epoch 80/99\n",
            "----------\n",
            "train Loss: 0.0450 Acc: 0.9869\n",
            "test Loss: 2.4014 Acc: 0.4893\n",
            "\n",
            "Epoch 81/99\n",
            "----------\n",
            "train Loss: 0.0404 Acc: 0.9892\n",
            "test Loss: 2.4213 Acc: 0.4913\n",
            "\n",
            "Epoch 82/99\n",
            "----------\n",
            "train Loss: 0.0433 Acc: 0.9869\n",
            "test Loss: 2.4015 Acc: 0.4874\n",
            "\n",
            "Epoch 83/99\n",
            "----------\n",
            "train Loss: 0.0475 Acc: 0.9842\n",
            "test Loss: 2.4165 Acc: 0.4913\n",
            "\n",
            "Epoch 84/99\n",
            "----------\n",
            "train Loss: 0.0421 Acc: 0.9852\n",
            "test Loss: 2.4144 Acc: 0.4903\n",
            "\n",
            "Epoch 85/99\n",
            "----------\n",
            "train Loss: 0.0474 Acc: 0.9865\n",
            "test Loss: 2.4148 Acc: 0.4913\n",
            "\n",
            "Epoch 86/99\n",
            "----------\n",
            "train Loss: 0.0428 Acc: 0.9899\n",
            "test Loss: 2.3805 Acc: 0.4884\n",
            "\n",
            "Epoch 87/99\n",
            "----------\n",
            "train Loss: 0.0415 Acc: 0.9895\n",
            "test Loss: 2.3903 Acc: 0.4855\n",
            "\n",
            "Epoch 88/99\n",
            "----------\n",
            "train Loss: 0.0440 Acc: 0.9889\n",
            "test Loss: 2.3429 Acc: 0.4884\n",
            "\n",
            "Epoch 89/99\n",
            "----------\n",
            "train Loss: 0.0408 Acc: 0.9882\n",
            "test Loss: 2.4031 Acc: 0.4913\n",
            "\n",
            "Epoch 90/99\n",
            "----------\n",
            "train Loss: 0.0434 Acc: 0.9885\n",
            "test Loss: 2.3984 Acc: 0.4874\n",
            "\n",
            "Epoch 91/99\n",
            "----------\n",
            "train Loss: 0.0416 Acc: 0.9899\n",
            "test Loss: 2.3691 Acc: 0.4961\n",
            "\n",
            "Epoch 92/99\n",
            "----------\n",
            "train Loss: 0.0461 Acc: 0.9875\n",
            "test Loss: 2.3442 Acc: 0.4884\n",
            "\n",
            "Epoch 93/99\n",
            "----------\n",
            "train Loss: 0.0417 Acc: 0.9879\n",
            "test Loss: 2.4276 Acc: 0.4884\n",
            "\n",
            "Epoch 94/99\n",
            "----------\n",
            "train Loss: 0.0421 Acc: 0.9899\n",
            "test Loss: 2.3889 Acc: 0.4884\n",
            "\n",
            "Epoch 95/99\n",
            "----------\n",
            "train Loss: 0.0409 Acc: 0.9885\n",
            "test Loss: 2.4171 Acc: 0.4922\n",
            "\n",
            "Epoch 96/99\n",
            "----------\n",
            "train Loss: 0.0429 Acc: 0.9869\n",
            "test Loss: 2.4082 Acc: 0.4835\n",
            "\n",
            "Epoch 97/99\n",
            "----------\n",
            "train Loss: 0.0477 Acc: 0.9855\n",
            "test Loss: 2.3669 Acc: 0.4845\n",
            "\n",
            "Epoch 98/99\n",
            "----------\n",
            "train Loss: 0.0465 Acc: 0.9879\n",
            "test Loss: 2.4073 Acc: 0.4884\n",
            "\n",
            "Epoch 99/99\n",
            "----------\n",
            "train Loss: 0.0496 Acc: 0.9848\n",
            "test Loss: 2.3760 Acc: 0.4826\n",
            "\n",
            "Training complete in 46m 30s\n",
            "Best Test Acc: 0.496124\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x216 with 3 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n9UZnxwx4Yxv"
      },
      "source": [
        "# Evaluate Classifier Performance\n",
        "\n",
        "Key metrics:\n",
        "\n",
        "$$Pr = \\frac{N_{TP}}{N_{TP}+N_{FP}}$$\n",
        "\n",
        "$$Re = \\frac{N_{TP}}{N_{TP}+N_{FN}}$$\n",
        "\n",
        "$$F_{\\beta} = \\frac{(\\beta^2+1)\\times Pr \\times Re}{\\beta^2 \\times Pr + Re}$$\n",
        "\n",
        "$$F_{1} = \\frac{2\\times Pr \\times Re}{Pr + Re}$$\n",
        "\n",
        "**Note:** the rows of the \"normalized\" confusion matrix represent the recall values for the specified class."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "enOL_71L428N",
        "outputId": "663a8f9f-8fd2-468b-bdb3-cbdfe8b58cb8"
      },
      "source": [
        "def getConfusionMatrix(model,dataloader,number_classes,verbose=True):\n",
        "  \"\"\"Calculates the Confusion Matrix for the given dataloader\"\"\"\n",
        "  # https://stackoverflow.com/questions/53290306/confusion-matrix-and-test-accuracy-for-pytorch-transfer-learning-tutorial\n",
        "  \n",
        "  # Create a placeholder matrix\n",
        "  confusion_matrix = torch.zeros(number_classes, number_classes)\n",
        "\n",
        "  # Fill out the confusion Matrix\n",
        "  y_label = np.array([])\n",
        "  y_predict = np.array([])\n",
        "  with torch.no_grad():\n",
        "      for i, (inputs, classes) in enumerate(dataloader):\n",
        "          inputs = inputs.to(device)\n",
        "          classes = classes.to(device)\n",
        "          outputs = model(inputs)\n",
        "          # Gets the predicted label\n",
        "          _, preds = torch.max(outputs, 1)\n",
        "          # np.vstack([ys, xs]) if ys.size else xs\n",
        "          temp_labels = classes.view(1,-1).cpu().detach().numpy()\n",
        "          temp_preds = preds.view(1,-1).cpu().detach().numpy()\n",
        "          if verbose:\n",
        "            print(\"Labels:\",temp_labels)\n",
        "            # print(temp_labels.shape)\n",
        "            # print(y_label.size)\n",
        "            print(\"Predictions:\",temp_preds)\n",
        "          y_label = np.hstack([y_label,temp_labels]) if y_label.size else temp_labels\n",
        "          y_predict = np.hstack([y_predict,temp_preds]) if y_predict.size else temp_preds\n",
        "          \n",
        "          # Iterates through a zipped tuple and increments the confusion_matrix\n",
        "          for t, p in zip(classes.view(-1), preds.view(-1)):\n",
        "                  confusion_matrix[t.long(), p.long()] += 1\n",
        "  \n",
        "  # Print the array outputs\n",
        "  if verbose:\n",
        "    print(confusion_matrix)\n",
        "    print(confusion_matrix.diag()/confusion_matrix.sum(1))\n",
        "    print(\"Y_labels\",y_label.flatten())\n",
        "    print(\"Y_predictions\",y_predict.flatten())\n",
        "\n",
        "  # Returns as NumPy Arrays\n",
        "  return confusion_matrix.cpu().detach().numpy(), y_label.reshape(-1), y_predict.reshape(-1)\n",
        "\n",
        "def plotConfusionMatrix(cm,normalize=True,labels=None):\n",
        "  \"\"\"Plots the Confusion Matrix\"\"\"\n",
        "  # Normalize\n",
        "  if normalize:\n",
        "    cmn = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
        "    fmt='.2f'\n",
        "  else:\n",
        "    cmn = cm\n",
        "    fmt = '.3g'\n",
        "  fig, ax = plt.subplots(figsize=(10,7))\n",
        "  sn.heatmap(cmn, annot=True, cmap='Blues',fmt=fmt, xticklabels=labels, yticklabels=labels)\n",
        "  plt.ylabel('Actual')\n",
        "  plt.xlabel('Predicted')\n",
        "  plt.show(block=False)\n",
        "\n",
        "# Reload the Model\n",
        "if do_save_model:\n",
        "  model_trained = torch.load(path_to_saved_model+\"model.pth\")\n",
        "  model_trained.eval() # Convert to evaluation only, this is faster\n",
        "\n",
        "# Determine the Confusion Matrix\n",
        "conf_mat, y_true, y_pred = getConfusionMatrix(model_trained,dataloaders['test'],len(labels),False)\n",
        "plotConfusionMatrix(conf_mat,normalize=False,labels=labels)\n",
        "plotConfusionMatrix(conf_mat,normalize=True,labels=labels)\n",
        "\n",
        "# https://scikit-learn.org/stable/modules/generated/sklearn.metrics.precision_recall_fscore_support.html\n",
        "# https://stackoverflow.com/questions/31421413/how-to-compute-precision-recall-accuracy-and-f1-score-for-the-multiclass-case\n",
        "# https://en.wikipedia.org/wiki/Sørensen–Dice_coefficient\n",
        "# https://stackoverflow.com/questions/20927368/how-to-normalize-a-confusion-matrix/66678924#66678924\n",
        "# https://stackoverflow.com/questions/33965560/r-markdown-table-with-a-caption\n",
        "\n",
        "# Calculate the Metrics\n",
        "precision, recall, fscore, _ = precision_recall_fscore_support(y_true, y_pred,average=None)\n",
        "\n",
        "# Bundle as a DataFrame\n",
        "stats = {\n",
        "    'Labels' : labels,\n",
        "    'Precision' : precision,\n",
        "    'Recall' : recall,\n",
        "    'F1 Score' : fscore\n",
        "}\n",
        "df = pd.DataFrame(stats)\n",
        "  \n",
        "# Print the output.\n",
        "print()\n",
        "print()\n",
        "print(df)\n",
        "print()\n",
        "print()\n",
        "# print(df.to_markdown()) \n",
        "\n"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\n",
            "  Labels  Precision    Recall  F1 Score\n",
            "0   Crop   0.406475  0.393728  0.400000\n",
            "1  Weed1   0.563050  0.741313  0.640000\n",
            "2  Weed2   0.000000  0.000000  0.000000\n",
            "3  Weed3   0.236364  0.108333  0.148571\n",
            "4  Weed4   1.000000  0.100000  0.181818\n",
            "5  Weed5   0.083333  0.021739  0.034483\n",
            "6  Weed6   0.000000  0.000000  0.000000\n",
            "7  Weed7   0.000000  0.000000  0.000000\n",
            "8  Weed8   0.000000  0.000000  0.000000\n",
            "9  Weed9   0.000000  0.000000  0.000000\n",
            "\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1308: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    }
  ]
}