Adversarial_Attacks.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LlvJfPxRSP9D"
   },
   "source": [
    "# **Description**\n",
    "In this notebook you will firstly train a CNN on cifar10 dataset. The CNN architecture is provided in ``` resnet.py ```. Then you are going to implement targeted and untargeted FGSM attack and also $l_{∞}$-PGD attack, and evaluate the performance of the trained model on the adversarial data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PTvKkL8C3TGI"
   },
   "source": [
    "# **Setups**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "cellView": "code",
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "dF8zbANmutFx",
    "outputId": "0b7436db-3c05-41ba-dca3-85abf00b24ff"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mounted at /content/drive\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "OGuslJ_e_Hlg"
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "import torch.backends.cudnn as cudnn\n",
    "import pickle\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "\n",
    "import os\n",
    "import argparse\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "from resnet import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "kMJqAx798x5c"
   },
   "outputs": [],
   "source": [
    "device = 'cuda' if torch.cuda.is_available() else 'cpu'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "xmr1-O7F8GeB"
   },
   "outputs": [],
   "source": [
    "'''\n",
    "These are the default values for variables, these should work, but if needed, you are allowed to change these parameters.\n",
    "'''\n",
    "\n",
    "batch_size = 256\n",
    "lr = 1e-3\n",
    "epochs = 100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-j3SQFaQSCJX"
   },
   "source": [
    "# **Data Preparation**\n",
    "Use cifar10 dataset provided in torchvision and create train and test dataloaders according to the given transformations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "pBNfOkMr92S3",
    "outputId": "cbb7b225-591e-474c-d0c9-f10d2ecdc64d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==> Preparing data..\n",
      "Files already downloaded and verified\n",
      "Files already downloaded and verified\n"
     ]
    }
   ],
   "source": [
    "print('==> Preparing data..')\n",
    "\n",
    "transform_train = transforms.Compose([\n",
    "    transforms.RandomCrop(32, padding=4),\n",
    "    transforms.RandomHorizontalFlip(),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),\n",
    "])\n",
    "\n",
    "transform_test = transforms.Compose([\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),\n",
    "])\n",
    "\n",
    "\n",
    "train_dataset = torchvision.datasets.CIFAR10(root='./data', train=True, download=True, transform=transform_train)\n",
    "test_dataset = torchvision.datasets.CIFAR10(root='./data', train=False, download=True, transform=transform_test)\n",
    "\n",
    "train_dataloader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=4)\n",
    "test_dataloader = torch.utils.data.DataLoader(test_dataset, batch_size=batch_size, shuffle=True, num_workers=4)\n",
    "\n",
    "classes = ('plane', 'car', 'bird', 'cat', 'deer',\n",
    "           'dog', 'frog', 'horse', 'ship', 'truck')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qaOM_JYFUlF8"
   },
   "source": [
    "# **Building The Model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "6zjkk49HUkXB",
    "outputId": "9da7d77c-9093-48d6-ff34-25478f9884a2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==> Building model..\n"
     ]
    }
   ],
   "source": [
    "########################################\n",
    "#build your model                      #\n",
    "########################################\n",
    "print('==> Building model..')\n",
    "net = ResNet18()\n",
    "net = net.to(device)\n",
    "\n",
    "if device == 'cuda':\n",
    "    net = torch.nn.DataParallel(net)\n",
    "    cudnn.benchmark = True\n",
    "\n",
    "optimizer = optim.Adam(net.parameters(), lr=lr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "i0JFoNGBUxdd"
   },
   "source": [
    "# **Training and Evaluating The Model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "OOLoVjk087ek"
   },
   "outputs": [],
   "source": [
    "# Training\n",
    "def train(epoch, net, optimizer, trainloader):\n",
    "    print('\\nEpoch: %d' % epoch)\n",
    "    net.train()\n",
    "    criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "    train_loss = 0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    for batch_idx, (inputs, targets) in enumerate(trainloader):\n",
    "        inputs, targets = inputs.to(device), targets.to(device)\n",
    "        optimizer.zero_grad()\n",
    "        outputs = net(inputs)\n",
    "        loss = criterion(outputs, targets)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_loss += loss.item()\n",
    "        _, predicted = outputs.max(1)\n",
    "        total += targets.size(0)\n",
    "        correct += predicted.eq(targets).sum().item()\n",
    "\n",
    "    return train_loss/len(trainloader), 100*correct/total\n",
    "\n",
    "\n",
    "def test(epoch, net, testloader):\n",
    "    net.eval()\n",
    "    criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for batch_idx, (inputs, targets) in enumerate(testloader):\n",
    "            inputs, targets = inputs.to(device), targets.to(device)\n",
    "            outputs = net(inputs)\n",
    "            loss = criterion(outputs, targets)\n",
    "\n",
    "            test_loss += loss.item()\n",
    "            _, predicted = outputs.max(1)\n",
    "            total += targets.size(0)\n",
    "            correct += predicted.eq(targets).sum().item()\n",
    "\n",
    "    return test_loss/len(testloader), 100 * correct / total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8Y7DHygJ83e8",
    "outputId": "9f75351c-1192-4ad4-900e-ca04e7b2def5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Epoch: 0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:560: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
      "  warnings.warn(_create_warning_msg(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAIN LOSS: 1.5054536297613261, TEST LOSS: 1.2434325039386749, TRAIN ACC: 44.614%, TEST ACC: 56.27%\n",
      "----------\n",
      "\n",
      "Epoch: 1\n",
      "TRAIN LOSS: 1.00388181939417, TEST LOSS: 1.2768470346927643, TRAIN ACC: 64.212%, TEST ACC: 58.02%\n",
      "----------\n",
      "\n",
      "Epoch: 2\n",
      "TRAIN LOSS: 0.796981003211469, TEST LOSS: 0.7909970089793206, TRAIN ACC: 71.49%, TEST ACC: 72.59%\n",
      "----------\n",
      "\n",
      "Epoch: 3\n",
      "TRAIN LOSS: 0.6639066808381859, TEST LOSS: 0.7820628911256791, TRAIN ACC: 76.648%, TEST ACC: 74.34%\n",
      "----------\n",
      "\n",
      "Epoch: 4\n",
      "TRAIN LOSS: 0.5707147224825255, TEST LOSS: 0.6269344910979271, TRAIN ACC: 80.04%, TEST ACC: 78.5%\n",
      "----------\n",
      "\n",
      "Epoch: 5\n",
      "TRAIN LOSS: 0.5039675833315266, TEST LOSS: 0.6400034420192242, TRAIN ACC: 82.432%, TEST ACC: 79.13%\n",
      "----------\n",
      "\n",
      "Epoch: 6\n",
      "TRAIN LOSS: 0.4470333715482634, TEST LOSS: 0.6361525699496269, TRAIN ACC: 84.492%, TEST ACC: 79.73%\n",
      "----------\n",
      "\n",
      "Epoch: 7\n",
      "TRAIN LOSS: 0.40740131845279615, TEST LOSS: 0.44121360033750534, TRAIN ACC: 85.834%, TEST ACC: 84.55%\n",
      "----------\n",
      "\n",
      "Epoch: 8\n",
      "TRAIN LOSS: 0.3743559702166489, TEST LOSS: 0.48142356649041174, TRAIN ACC: 86.926%, TEST ACC: 84.24%\n",
      "----------\n",
      "\n",
      "Epoch: 9\n",
      "TRAIN LOSS: 0.34374771639704704, TEST LOSS: 0.539466316998005, TRAIN ACC: 88.056%, TEST ACC: 82.86%\n",
      "----------\n"
     ]
    }
   ],
   "source": [
    "######################################################################\n",
    "#train your model, make sure to save loss and accuracy in all epochs #\n",
    "######################################################################\n",
    "\n",
    "train_accuracies, test_accuracies= [], []\n",
    "train_losses, test_losses = [], []\n",
    "epochs = 10\n",
    "\n",
    "for epoch in range(epochs):\n",
    "    train_loss, train_accuracy = train(epoch, net, optimizer, train_dataloader)\n",
    "    test_loss, test_accuracy = test(epoch, net, test_dataloader)\n",
    "\n",
    "    print(f'TRAIN LOSS: {train_loss}, TEST LOSS: {test_loss}, TRAIN ACC: {train_accuracy}%, TEST ACC: {test_accuracy}%')\n",
    "    print(10 * '-')\n",
    "    train_losses.append(train_loss)\n",
    "    train_accuracies.append(train_accuracy)\n",
    "    test_losses.append(test_loss)\n",
    "    test_accuracies.append(test_accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kg-IZNW_gB2Z"
   },
   "outputs": [],
   "source": [
    "########################################\n",
    "#save your model and history           #\n",
    "########################################\n",
    "torch.save(net.state_dict(), './drive/MyDrive/resnet18.pt')\n",
    "with open('./drive/MyDrive/history.pkl', 'wb') as f:\n",
    "    pickle.dump((train_accuracies, test_accuracies, train_losses, test_losses), f)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-xsrwlrLU_jE"
   },
   "source": [
    "## **Train and Test Loss and Accuracy Plots**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 410
    },
    "id": "HWaW89Sj9R4g",
    "outputId": "4a24bec4-981c-4a0a-9852-221e248c0399"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#############################################################\n",
    "#plot train and test loss and accuracy dusring the training #\n",
    "#############################################################\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def plot_accuracies_and_losses(train_accuracies, test_accuracies, train_losses, test_losses):\n",
    "    fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10, 4))\n",
    "\n",
    "    ax1.plot(train_accuracies, label='Train Accuracy')\n",
    "    ax1.plot(test_accuracies, label='Test Accuracy')\n",
    "    ax1.set_title('Accuracies')\n",
    "    ax1.set_xlabel('Epoch')\n",
    "    ax1.set_ylabel('Accuracy')\n",
    "    ax1.legend()\n",
    "\n",
    "    ax2.plot(train_losses, label='Train Loss')\n",
    "    ax2.plot(test_losses, label='Test Loss')\n",
    "    ax2.set_title('Losses')\n",
    "    ax2.set_xlabel('Epoch')\n",
    "    ax2.set_ylabel('Loss')\n",
    "    ax2.legend()\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "plot_accuracies_and_losses(train_accuracies, test_accuracies, train_losses, test_losses)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **FGSM Attack**\n",
    "\n",
    "In this section you must implement targeted and untargeted FGSM attack and evalute the trained model performance on the adversarial data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VDhwjK4OVohn"
   },
   "source": [
    "## **Untargeted FGSM Attack**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "id": "8XbNsfxGhry6"
   },
   "outputs": [],
   "source": [
    "########################################\n",
    "#define untargeted FGSM attack         #\n",
    "########################################\n",
    "\n",
    "\n",
    "def FGSM(model, x, y, epsilon):\n",
    "  \"\"\"\n",
    "  inputs:\n",
    "    model: model to be attacked\n",
    "    x: input image\n",
    "    y: label\n",
    "    epsilon: clipping threshold for FGSM\n",
    "\n",
    "  outputs:\n",
    "    x': the perturbed image\n",
    "    delta: the added perturbation\n",
    "    y': prediction of the model for the perturbed input\n",
    "  \"\"\"\n",
    "  delta = None\n",
    "  model.eval()\n",
    "  x = torch.tensor(x, requires_grad=True).to(device)\n",
    "  loss = F.cross_entropy(model(x), y)\n",
    "  model.zero_grad()\n",
    "  loss.backward()\n",
    "  delta = epsilon * x.grad.detach().sign()\n",
    "  return x+delta, delta, model(x+delta)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1iocJ42FbTPE"
   },
   "source": [
    "###**Evaluation**\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KCtIVJ7ybo2G"
   },
   "source": [
    " For $\\epsilon \\in \\{\\frac{1}{255}, \\frac{4}{255}, \\frac{8}{255},\\}$ report the performance of the model on targeted FGSM attack on the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "RWM6pCKibm1j",
    "outputId": "2bf63b3c-c30a-4958-c653-5dd6c808fe73"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-9-070eb95800b9>:21: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  x = torch.tensor(x, requires_grad=True).to(device)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EPSILON: 0.00392156862745098, TEST LOSS: 0.8035684242844582, TEST ACC: 75.1\n",
      "EPSILON: 0.01568627450980392, TEST LOSS: 1.9794514805078507, TEST ACC: 51.11\n",
      "EPSILON: 0.03137254901960784, TEST LOSS: 3.8267365872859953, TEST ACC: 28.51\n"
     ]
    }
   ],
   "source": [
    "######################\n",
    "# Your code goes here\n",
    "\n",
    "epsilons = [1/255, 4/255, 8/255]\n",
    "loss_fn = nn.CrossEntropyLoss()\n",
    "net.eval()\n",
    "\n",
    "for epsilon in epsilons:\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    for images, labels in test_dataloader:\n",
    "        images = images.to(device)\n",
    "        labels = labels.to(device)\n",
    "        x_adversarial, delta, out_adversarial = FGSM(net, images, labels, epsilon)\n",
    "        loss = loss_fn(out_adversarial, labels)\n",
    "        test_loss += loss.item()\n",
    "        predictions = torch.argmax(out_adversarial, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predictions == labels).sum().item()\n",
    "\n",
    "    print(f'EPSILON: {epsilon}, TEST LOSS: {test_loss / len(test_dataloader)}, TEST ACC: {round(100 * correct / total, 2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "540xuXsTbvwd"
   },
   "source": [
    "### **FGSM by gradually increasing $ϵ$**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4Y6-jilKW1mo"
   },
   "source": [
    "FGSM attack may not work for small $\\epsilon$, on the other hand we want to alter image as insignificant as possible, so here is what we do, we start from $\\epsilon = \\frac{1}{255}$ and if the attack was unsuccessful, we increase $\\epsilon$ by $\\frac{1}{255}$. The procedure goes on until the predicted label is different from the ground truth."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "96PZLzYn8AYd"
   },
   "outputs": [],
   "source": [
    "#######################################################\n",
    "#increase epsilon gradually till the label is changed. The function can be implemented recursively.#\n",
    "#######################################################\n",
    "\n",
    "def generate_adversarial_FGSM (model, x, label, initial_epsilon=1/255):\n",
    "  \"\"\"\n",
    "  inputs:\n",
    "    model: model to be attacked\n",
    "    x: input image\n",
    "    y: label\n",
    "    epsilon: intial clipping threshold for FGSM\n",
    "\n",
    "  outputs:\n",
    "    epsilon: the epsilon that flips the label.\n",
    "    delta: the added perturbation\n",
    "    y': prediction of the model for the perturbed input\n",
    "  \"\"\"\n",
    "\n",
    "  delta = None\n",
    "  model.eval()\n",
    "  x = torch.tensor(x, requires_grad=True).to(device)\n",
    "  loss = F.cross_entropy(model(x), label)\n",
    "  model.zero_grad()\n",
    "  loss.backward()\n",
    "  delta = initial_epsilon * x.grad.detach().sign()\n",
    "\n",
    "  x_adversarial = x + delta\n",
    "\n",
    "  out_adversarial = model(x_adversarial)\n",
    "  predictions = torch.argmax(out_adversarial, 1)\n",
    "\n",
    "  correct = (predictions == label).sum()\n",
    "  total = label.size(0)\n",
    "\n",
    "  if total == correct:\n",
    "      initial_epsilon, delta, out_adversarial, x_adversarial = generate_adversarial_FGSM(model, x, label, initial_epsilon + 1/255)\n",
    "\n",
    "  return initial_epsilon, delta, out_adversarial, x_adversarial"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "s29e791J9F0V"
   },
   "source": [
    "### **Visualization**\n",
    "Pick 10 samples from cifar10 test set, each from one class, make sure the selected samples are classified correctly. Plot these samples. Take note that the selected data must not be transformed by train or test transformations while visualisation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "dEnJ2Vz2j8Zt",
    "outputId": "06a5676b-704c-4e74-87ad-2cd8aecd5369"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#############################################################################################\n",
    "#visualize one test sample from each class and make sure that they are correctly labeled.   #\n",
    "#save index of these samples, since they are needed in the following sections.              #\n",
    "#############################################################################################\n",
    "def imshow(img, title=None):\n",
    "    img = img / 2 + 0.5     # unnormalize\n",
    "    npimg = img.numpy()\n",
    "    fig = plt.figure(figsize=(5, 5))  # Set the figure size\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    if title is not None:\n",
    "        plt.title(title)\n",
    "    plt.axis('off')\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "net.eval()\n",
    "class_indices = {}\n",
    "for i in range(len(test_dataset)):\n",
    "    _, label = test_dataset[i]\n",
    "    if label not in class_indices:\n",
    "        class_indices[label] = []\n",
    "    class_indices[label].append(i)\n",
    "\n",
    "samples = []\n",
    "for label, indices in class_indices.items():\n",
    "    for i in indices:\n",
    "        # Check if the sample is classified correctly\n",
    "        image, target = test_dataset[i]\n",
    "        if torch.argmax(net(image.to(device).unsqueeze(0))) == target:\n",
    "            samples.append([image, target, i])\n",
    "            break\n",
    "    if len(samples) == 10:\n",
    "        break\n",
    "\n",
    "\n",
    "for image, label, index in samples:\n",
    "    imshow(image, classes[label])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EwMfircn3TGO"
   },
   "source": [
    "Now, apply untargetted FGSM attack ``` generate_adversarial_FGSM ``` on the selected samples. Visualize the original image, the perturbed image, and the perturbation. since the perturbation is imperceptible, magnify it by the value of $\\frac{1}{\\epsilon}$ before visualization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "5MmdiqTj7X_y",
    "outputId": "43468198-d890-4b3f-86d6-05e3003f1628"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-11-469c9d61ef88>:21: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  x = torch.tensor(x, requires_grad=True).to(device)\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9EAAAFTCAYAAAA+14+JAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABqX0lEQVR4nO3deXhTVfoH8G+SpimlbaBAKUtZBASVTZGligiCIqKIovMDGQV1xAVwwRlHHEHFBccVGAVGxwGVFhQRd1FkFQdQUEBFEJAdyiLQlZY09/z+YJohNDnvaXO78v08T54Hck7fe+7JzZuc5Oa+DqWUAhERERERERGJnBU9ACIiIiIiIqKqgotoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMsRFNBEREREREZEhLqKJiIiIiIiIDHERTURERERERGSIi2giqhDDhw9Hs2bNKnoYRESimTNnwuFwYMeOHRU9FCIiqgTKdRH9+OOPw+Fw4PDhwyHb27Zti549ewb+v2PHDjgcDjgcDsybN88o3vDhwwN/43A4kJCQgA4dOuDFF19EQUFBsb8Nd8vIyCg2BofDAafTicTERPTr1w8rV64E8L8XV+lWtGAYPnw44uLiws5TXFwchg8fHvj/0qVLg+K4XC4kJSXhhhtuwC+//FLs70+fg6ioKKSkpGDw4MHYuHFjUN/TY59+mzNnTqBvs2bNguahVq1aaNeuHUaMGIHVq1eH3Z/T9ezZMyhOQkICWrdujZtvvhkLFy7U/u0vv/wCh8OBmJgYHDt2LOw+h7udOq/Hjh1DTEwMHA5HyHksKYfDgVGjRhn3nzp1KhwOB7p27aqNeeqtZs2aOPfcc/HUU08hLy8vqK9uDmJiYgL9ih7z9957r+Q7SWeMkh6fUVFRSExMRKdOnXDfffcVyzUDBgxAbGwssrOzw8YbOnQooqOj8fvvvxeLf/rtrrvuCvzd6ce+x+PB2WefjfHjxyM/P7/YdnJycvDYY4+hbdu2qFmzJurUqYOOHTvivvvuw759+0KO7bPPPoPD4UDDhg1hWVbIPqfmyKLna5cuXfDWW28V6ys9D03nvyQ5R6fote6FF14w/puHHnoIDocD//d//6eNefprcseOHfHKK6/A7/cH9T/1teH0W5s2bQL9il5z16xZU7qdJaKQTn8/GxMTg4YNG6Jv376YMmVKyPxd0vfTp+cY3fP+1Nvjjz8e+Jtw7wVLsp+nxpN8+eWXuP3229G2bVu4XK6wXwBs2rQJDz30EDp27Ij4+Hg0aNAA/fv3D5mrTn+9OPXWqlUrcUzh5u3KK6803i+KTFRFD8DUhAkTcP3118PhcIh9PR4P/vWvfwE4uVCaN28e/vznP+O7774LWhQCwLRp00IuaGvVqhX0/yFDhuCqq66C3+/Hr7/+iqlTp6JXr1747rvv0KNHD7z99ttB/f/0pz+hS5cuGDFiROA+3cLZxL333ovOnTvD5/Nhw4YNmD59OpYuXYqffvoJycnJQX1PnYPCwkJs27YN06dPx4IFC7Bx40Y0bNgwZOzTpaamBv2/Y8eOePDBBwEA2dnZ+OWXXzB37ly8/vrreOCBB/DSSy8Z7Uvjxo0xceJEAEBubi62bt2K999/H7NmzcIf/vAHzJo1C263u9jfzZo1C8nJyTh69Cjee+89/OlPfwIA3HnnnejTp0+g3/bt2zF+/HiMGDECl1xySeD+Fi1aBP49d+5cOBwOJCcnIy0tDU899ZTR2O2SlpaGZs2a4dtvv8XWrVvRsmXLkP0uv/xy3HLLLQBOvvn/+uuvMW7cOKxfvx5z584N6nvq434ql8tl/w5QtVbS41MphczMTKxfvx5vvvkmpk6dir///e8YM2YMgJML5I8//hjz588PHM+nysvLw4cffogrr7wSderUKRb/dGeffXbQ/0899jMzM/Hhhx/iySefxLZt25CWlhbo5/P50KNHD2zatAnDhg3D6NGjkZOTg59//hnp6em47rrriuXHU+djx44dWLx4cVC+OdWpOXL//v3417/+hWHDhqGgoAB33HFHyL8JxXT+K4pSCrNnz0azZs3w8ccfIzs7G/Hx8SH7Fr1+Aicfm88++wyjR4/Gzp078fzzzwf1PfW14VRer9f+nSCikCZMmIDmzZvD5/MhIyMDS5cuxf3334+XXnoJH330Edq3b1/sb0zfT5/ub3/7W+C9HAB89913mDJlCh555BGcc845gftP3Wa494I6u3btQm5ublBMANi3bx9+//13tGvXLuzfpqen45133sEFF1wQ8vWhyL/+9S+88cYbGDRoEO655x5kZmbin//8J7p164YFCxYEvW5MmjQJOTk5QX+/c+dOPProo7jiiivE/QFC50vd+Mhmqhw99thjCoA6dOhQyPbzzjtPXXrppYH/b9++XQFQHTt2VADUvHnzxHjDhg1TNWvWDOrn9/vVhRdeqACovXv3Go3l9DE8//zzQfd//vnnCoC6++67Q/5dzZo11bBhw0K2hRqj7m+XLFmiAKi5c+cG9Zs2bZoCoP7+978bxf/kk08UAPXaa6+JsUNp2rSp6t+/f7H78/Ly1MCBAxUANXXqVDHOpZdeqs4777xi9xcWFqp77rlHAVAPPfRQsXbLslSzZs3UmDFj1HXXXad69uwZdhvfffedAqBmzJgRtk+PHj3U9ddfrx544AHVvHlzcdwSAGrkyJFGfX/77TcFQL3//vuqXr166vHHHy9RzBtuuEE5nU51/PjxwH3ScVWkJI95WRo2bJhq2rRphY6BQov0+Dx8+LBKTU1VANSnn36qlDqZJ+Lj41Xfvn1DxkpPT1cA1Jw5c8T4pwt17FuWpbp166YcDofKyMgI3P/uu+8qACotLa1YnOPHj6vMzMxi9+fk5KiaNWuqKVOmqPPPP18NHz485DhC5ciDBw+quLg4dc455wTdr3seRjr/pRHutS6cxYsXKwBq8eLFyu12q5kzZxrHtCxLde7cWTVs2DDo/nCvDaebMWOGAqC+++47o7HaoWib27dvL7dtKnXy2CMqL7rn1qJFi1SNGjVU06ZNVV5eXuD+SN9Pn27u3LkKgFqyZEnI9pK8Fzx9/E2bNlVPPvmkeu2119T48ePVtGnTVNOmTYNed0LZu3evOnHihFJKqf79+4d977JmzRqVnZ0ddN/hw4dVvXr11MUXXyyO8cknn1QA1DfffCP2Nc2XVHaqxG+iBw8ejLPPPhsTJkyAUqrEf+90OtHzv6eJ2/V7pqJvN7dt22ZLvPIYQ9G31VFR9p6AUKNGDbz99ttITEzE008/XarHCDj5bemUKVNw7rnn4pVXXkFmZmZQ+zfffIMdO3Zg8ODBGDx4MJYvX449e/aUalu7du3C119/HYi1fft2/Oc//ylVrNJIS0tD7dq10b9/f9xwww1B35SZSE5ODpxCW5kUnaL6zjvv4JFHHkFycjJq1qyJAQMGYPfu3eLfv/DCC7joootQp04d1KhRA506dQp5umvRaawffPAB2rZtC4/Hg/POOw8LFiwo1nfv3r247bbbUL9+/UC/f//737bsb3UV6fFZp04dzJkzB1FRUXj66acBnMwT119/PRYtWoSDBw8W+5v09HTEx8djwIABtuyDw+FA9+7doZTCb7/9Fri/KF9efPHFxf4mJiYGCQkJxe6fP38+jh8/jhtvvBGDBw/G+++/H/I08VDq1auHNm3alOi1ItL5Lw9paWk499xz0atXL/Tp06dEY3Q4HKhfv36ly19Ffv75Z1x22WWoUaMGGjdujKeeeirsKfyff/45LrnkEtSsWRPx8fHo378/fv7552L9Nm3ahBtuuAGJiYmIiYnBhRdeiI8++iioT9GptMuWLcM999yDpKQkNG7cuEz2kaikLrvsMowbNw47d+7ErFmzKmwcpX0veNlll+HHH39EXl4exo4di0mTJmH9+vVYs2ZN2J+kFGnYsGHIsyNP16lTp2LfxtepUweXXHKJ0c8G09PT0bx5c1x00UVi3yKFhYXFvtGm8lElFtEulwuPPvoo1q9fj/nz55cqRtEbmFNPEwSAI0eO4PDhw0E3k99XFC3Ga9euXarx2EEaQ9H+HDhwACtXrsQDDzyAOnXq4Oqrry7WNzs7u9g8HD582HhBHBcXh+uuuw579+4t9lvIknC5XBgyZAjy8vKwYsWKoLa0tDS0aNECnTt3xjXXXIPY2FjMnj27VNuZPXs2atasiauvvhpdunRBixYtyvWNalpaGq6//npER0djyJAh2LJlC7777ruQffPz8wOPx86dO5Geno4333wTN910U8g3oaEex6ysrBKP8ejRoyFjnX47/bfZAPD000/j008/xV//+lfce++9WLhwIfr06YPjx49rtzl58mScf/75mDBhAp555hlERUXhxhtvxKefflqs74oVK3DPPfdg8ODBeO6555Cfn49BgwYFfk8LAAcOHEC3bt3w1VdfYdSoUZg8eTJatmyJ22+/HZMmTSrxnJwpSnJ8htOkSRNceumlWLVqVeD4Gzp0KAoLC/Huu+8G9T1y5Ai++OILXHfddahRo0ZQ26nH/6m3EydOiGMIlSObNm0KAHjrrbeM81taWhp69eqF5ORkDB48GNnZ2fj444+N/rawsBB79uwp0WuFHfNflgoKCjBv3jwMGTIEwMnTtRcvXhz47ePp8vLyAo/bb7/9hldffRULFizAsGHDivX1+/0hH+/c3NwSj9Mkfx0+fDjoeikZGRno1asX1q1bh4cffhj3338/3nrrLUyePLlY/Lfffhv9+/dHXFwc/v73v2PcuHHYuHEjunfvHvSB/c8//4xu3brhl19+wcMPP4wXX3wRNWvWxMCBA0O+p7nnnnuwceNGjB8/Hg8//HCJ95uorNx8880ATv5G+HSlfT9dUpG8F3Q6nXA6/7f0MfmJqB0yMjJQt25dbZ8ffvgBv/zyC2666SbjuL/++mvgA7zk5GSMGzcOPp8v0uGSqfL82ru0p3M///zzqrCwULVq1Up16NBBWZYVNl7RaX2HDh1Shw4dUlu3blXPPPOMcjgcqn379sXGEurWunXrYmN44okn1KFDh1RGRob6+uuvVefOnbWnxJbF6dz//ve/1aFDh9S+ffvUggULVMuWLZXD4VDffvttsfih9qtRo0Zq7dq1QX2LYoe77d+/P9A33OncRV5++WUFQH344Ydh+ygln4Iyf/58BUBNnjw5cN+JEydUnTp11N/+9rfAfTfddJPq0KFDyBjS6dzt2rVTQ4cODfz/kUceUXXr1lU+n087dh0Ynlq5Zs0aBUAtXLhQKXXy1KTGjRur++67L2TMULeBAweq/Pz8oL7hHncAQafQmp7O3bRpU+2xUXR77LHHisVu1KiRysrKCtxfdArtqY9pqNO5Tz1FTKmTj3vbtm3VZZddVmxeoqOj1datWwP3rV+/XgFQ//jHPwL33X777apBgwbq8OHDQX8/ePBg5fV6i22PSn586o75++67TwFQ69evV0qd/MlGgwYNVGpqalC/6dOnKwDqiy++KBY/3G327NmBfqHy/gsvvKAcDodq27Zt4DVDqZPHWOvWrRUA1bRpUzV8+HD1xhtvqAMHDoTchwMHDqioqCj1+uuvB+676KKL1LXXXlusb9OmTdUVV1wRGMePP/6obr755pDzFO55aOf8l0RJTud+7733FAC1ZcsWpZRSWVlZKiYmRr388sshY4a63X333UGPi1InXxvC9b/zzjsD/UxP5zbJX6e/Ttx///0KgFq9enXgvoMHDyqv1xt0Ond2draqVauWuuOOO4K2mZGRobxeb9D9vXv3Vu3atQvK2ZZlqYsuuki1atWq2H51795dFRYWaveNqCyYPLe8Xq86//zzA/8v6fvpSE7nLul7wVMtWbJENW/eXE2YMCFwOvfUqVONTuc+le507lCWL1+uHA6HGjdunLbfgw8+qACojRs3GsW97bbb1OOPP67mzZun3nrrLTVgwAAFQP3hD38wHhtFpnKeSxVC0bfRw4YNwwcffIDrrrsubN/c3FzUq1cv6L6LLrqo2MW/AGDevHnFTt+rWbNmsX6PPfYYHnvsscD/4+Li8OKLL+KGG24o6a6U2m233Rb0/3r16uHtt98OeUGwmJiYwDcllmVhx44deOmll3DVVVdh+fLlxS7KM378+KALcBVJTEw0Hl/RKSy6q++WNs7nn3+O33//PfDNB3Dy249rrrkGP//8M8477zzj+Bs2bMCPP/4YdDGGIUOG4JlnnsEXX3yB/v37RzR+SVpaGurXr49evXoBQODqtrNmzcKLL75Y7CJg1157beAKvHl5eVi1ahVefvll3HTTTXjvvfeCPkk99XE/lfQJaLhxSt8cA8BZZ51V7L5bbrkl6CJDN9xwAxo0aIDPPvsM9957b9hYp34LefToUfj9flxyySUhP2Xu06dP0IXi2rdvj4SEhMCpu0opzJs3D3/4wx+glAq6in/fvn0xZ84cfP/99yFP6z2TlfT41Dn9uexyuTB48GC8/PLL2LFjR+AKp+np6ahfvz569+5dLMapx/+pTr8ITKi83717d7z55ptBz5EaNWpg9erVePrpp/Huu+9i5syZmDlzJpxOJ+655x688MIL8Hg8gf5z5syB0+nEoEGDAvcNGTIEDz74II4ePVrsG+Yvv/yy2DhuvfXWYhfQCsfO+S8raWlpuPDCCwMXOys6jTktLQ33339/sf4jRozAjTfeCADIysrC4sWLMW3aNHg8Hrz88stBfZs1a4bXX3+9WIzSnNYsVXsocurrx2effYZu3bqhS5cugfvq1auHoUOHYurUqUGxjx07hiFDhgTlFpfLha5du2LJkiUATn47t3jxYkyYMAHZ2dlBr2t9+/bFY489hr1796JRo0aB+++4445K8TgThRIXFxfyfZ7p++lIRPJesHnz5vj0009xzjnnBH46cffdd+Paa68NOoPNTgcPHsRNN92E5s2b46GHHgrbz7IszJkzB+eff36xC5+F88YbbwT9/+abb8aIESMCF/rt1q1bRGMnWaVbROtOrRg6dCiefPJJTJgwAQMHDgzb79SFhMfjQfPmzcO+APfo0cNogVH0JiA/Px+LFy/GlClTipXnsFOoeSha6Obk5GD+/PmBN3ehuFyuYlePveqqq9CqVSuMHTu2WMmwdu3ahb3arKmi32SEu0JrJHFmzZqF5s2bw+PxYOvWrQBOXmk7NjYWaWlpeOaZZ4zjz5o1CzVr1sRZZ50ViBUTE4NmzZohLS2tTBfRfr8fc+bMQa9evbB9+/bA/V27dsWLL76IRYsWFbsqY+PGjYMemwEDBqBOnTr485//jE8++QTXXHNNoC3U415akSwuTy/P4HA40LJlS/GaBJ988gmeeuoprFu3LugUy1DPhyZNmhS7r3bt2jh69CgA4NChQzh27Bhee+01vPbaayG3F+q3uWey0hyfOqGey0OHDsXLL7+M9PR0PPLII9izZw++/vpr3HvvvSEXDqcf/+Gcmvf37NmD5557DgcPHix2ejhw8krPzz33HJ577jns3LkTixYtwgsvvIBXXnkFXq836Er9s2bNQpcuXfD7778H3midf/75OHHiBObOnRtUgQE4OVdPPfUU/H4/fvrpJzz11FM4evQooqOjxX2we/7LwrFjx/DZZ59h1KhRgfwJnMwX8+bNw6+//lrsQ9pWrVoFPYZFlTYmTZqE2267LegDkZo1a9qWw0oTZ+fOnSHLirVu3Tro/1u2bAFw8neWoRQtJrZu3QqlFMaNG4dx48aF7Hvw4MGgRXTz5s1LPG6i8pKTk4OkpKRi95u+n45EJO8Fi37Kc7qGDRuWyRWtc3NzcfXVVyM7OxsrVqzQVuhZtmwZ9u7diwceeCCibT744IN4/fXX8dVXX3ERXQ7KdRFdVKs23LdbeXl5QfVsT1f0bfTw4cPx4YcfavvZ9SJc5NQ3AVdffTVcLhcefvhh9OrVCxdeeGGJYsXExKCgoABKqWKLA6UU8vPzQ87DqQvdgQMHIi8vD3fccQe6d++OlJQUcbuNGzdG69atsXz58hKN19RPP/0EABGXYjk9TlZWFj7++GPk5+eHrJ2Xnp6Op59+2ui3Leq/ZVlyc3Nx7rnnFms/ePAgcnJyIi5HFs7ixYuxf/9+zJkzp1i5NeDkNzwmb5KLvrFbvnx50CLaTocOHTL6oCguLs6W+fr6668xYMAA9OjRA1OnTkWDBg3gdrsxY8YMpKenF+sf7psa9d/fuRZdCOiPf/xjyN9eAghZpuNMZtfxWeSnn36Cy+UKWhR06tQJbdq0wezZs/HII49g9uzZUEph6NChEY399Lzft29ftGnTBnfeeWexCzidqmnTprjttttw3XXX4ayzzgoqd3fqb5FD5Z60tLRii+i6desGxlE0hquvvhqTJ08OlPsKx+75Lwtz585FQUEBXnzxRbz44ovF2tPS0vDEE0+IcXr37o1XXnkFy5cv15aWiUS432ifzuv1hvywRacov7z99tvFSkwC/7uAZ1G/P//5z+jbt2/IWKe/ZpZ0LETlZc+ePcjMzKyQknt2vhccPnx4GYzwf06cOIHrr78eGzZswBdffIG2bdtq+6elpcHpdAZ9w14aRWuBI0eORBSHzJTrIrroU6DNmzcXW/Tl5eVh9+7d4huEP/7xj3jqqafwxBNP2HYV19L429/+htdffx2PPvpoyCsC6zRt2jRQu/n0RLR161b4/f6wn5id6tlnn8X8+fPx9NNPY/r06UbbLqur+BV9O56SkmJ8Kkoofr8f6enpiI2NRffu3QEgcCXcadOmFfuUc/PmzXj00UfxzTffBPrrLFu2DHv27MGECROKjfPo0aMYMWIEPvjgA/zxj38s9T7opKWlISkpCa+++mqxtvfffx/z58/H9OnTxTdRhYWFAFCmV2Ts3Lkzdu7cKfZ77LHH8PjjjwfdV/QtTRGlFLZu3apdtM6bNw8xMTH44osvgk6nnTFjRskG/l/16tVDfHw8/H6/7R+qVVd2HZ/AySvgL1u2DKmpqcXOThk6dCjGjRuHDRs2ID09Ha1atQr5s5RINGjQAA888ACeeOIJrFq1SvxUvnbt2mjRokXgQzzg5Hy43W68/fbbxT60WbFiBaZMmYJdu3aFPCuiSP/+/XHppZfimWeewZ133qk9vdHO+S8raWlpaNu2bdDPm4r885//RHp6utEiujxyWIMGDYz6zZgxI/CmumnTpsXyF3DyteZURT8lSUpK0uaXop+7uN1u5iGq8op+FhnuA6GyZOd7wbJkWRZuueUWLFq0CO+++y4uvfRSbf+iCzX27Nkz4m/Ei37OdvpPiqhslOsiunfv3oiOjsa0adNw2WWXBZ2K/Nprr6GwsBD9+vXTxjj12+iKVKtWLdx555147rnnsG7dOnTs2NH4b/v164dHHnkEr7zySrErBBe9eZLmATj5Ij5o0CDMnDkTjz/+eMhPw0/166+/YvPmzejUqZPxWE0cP34cN998M44cOYJnnnmm1Fc79Pv9uPfeewNXMC06HW7WrFk466yzcNdddxX7m4KCAjz77LNIS0szSpxFp3L/5S9/Cflt//PPP4+0tLQyWUQfP34c77//Pm688caQv6Vv2LAhZs+ejY8++kgst1B02mqHDh1sH2eRSH4T/dZbb2Hs2LGBxdN7772H/fv3469//WvYOC6XCw6HI+jb7x07duCDDz4o+eD/G2/QoEFIT0/HTz/9VOyT4EOHDvGF5hR2Hp9HjhzBkCFD4Pf78be//a1Ye9Eievz48Vi3bl2xD2HsMnr0aDz//PN49tlnA8fR+vXr0ahRo2Jvwnbu3ImNGzcGnbablpaGSy65JOT+pqamYsqUKZg9e7b2uAaAv/71r7jqqqvw+uuvh/zNMGDv/JeV3bt3Y/ny5XjiiSdCjvHEiRMYOnQoVq9eHfKU6FOVRw4rzW+ir7rqKkyaNAnffvtt4HfRhw4dKla9oW/fvkhISMAzzzyDXr16FSt/U5RfkpKS0LNnT/zzn//E6NGjiy3smYeoqli8eDGefPJJNG/ePOIzh0rDzveCZWn06NF455138M9//hPXX3+92P+zzz7DsWPHtHO6adMmxMbGBj6wzcrKgsfjCfrCQSkVOIuqIj7kOBOV6yI6KSkJ48ePx6OPPooePXpgwIABiI2NxX/+8x/Mnj0bV1xxhdGpqUW/jV63bl3EY3rvvfdCnop6+eWXo379+tq/ve+++zBp0iQ8++yzIU+9C6djx47405/+hMmTJ2PLli24/PLLAZx8wf/ss8/wpz/9yfiNxV/+8he8++67gXEUKSwsDNTxK7qw2PTp02FZVshvEL7++uuQdU/bt28f9O3h3r17A3FzcnKwceNGzJ07FxkZGXjwwQdx5513Go07MzMzECcvLw9bt27F+++/j23btmHw4MF48sknAQD79u3DkiVLwl6MyuPxoG/fvpg7dy6mTJmireNX9Gnf5ZdfHvZnAwMGDMDkyZNx8OBBJCUlYenSpejVq1fIb1tDWbNmTdDvKYv07NkTe/fuRXZ2dtgzKLp164Z69eohLS0t6E3yr7/+GjRXq1atwptvvomWLVsGyk0UOfVxP911110X9C3YvHnzsGnTpmL9hg0bhpSUlIh+E52YmIju3bvj1ltvxYEDBzBp0iS0bNkSd9xxR9i/6d+/P1566SVceeWVuOmmm3Dw4EG8+uqraNmyJTZs2FCqcTz77LNYsmQJunbtijvuuAPnnnsujhw5gu+//x5fffUVT3k6xUcffRTR8amUQlZWFtavX4+5c+ciJycn8HierqgOZtHPcnRvHk49/k9Vv379QO4Mp06dOrj11lsxdepU/PLLLzjnnHOwcOFCPPbYYxgwYAC6deuGuLg4/Pbbb/j3v/+NgoKCwPN89erV2Lp1a8iLmgFAo0aNcMEFFyAtLU1cRPfr1w9t27bFSy+9hJEjR4bMU6Wdf13O6d69O2bOnIlbb7016NtWnUWLFoV8LRg4cCA+/fRTKKXCjvGqq65CVFQU0tLSghbR33//feAxzM7OxqJFizBv3jxcdNFFxc4+O/W14XSnf7j573//O+SZYPfddx/i4+NL9c3vQw89hLfffhtXXnkl7rvvPtSsWROvvfYamjZtGpSHEhISMG3aNNx888244IILMHjwYNSrVw+7du3Cp59+iosvvhivvPIKgJMfjnfv3h3t2rXDHXfcgbPOOitQenLPnj1Yv359icdJVJY+//xzbNq0CYWFhThw4AAWL16MhQsXomnTpvjoo49CvocyfT+tyzHhTnu2871gaWzYsCHws6CtW7ciMzMzkHc7dOgQWLtMmjQJU6dORWpqKmJjY4vlstPfhwEnP6z1eDxBF6883TnnnINLL70US5cuBXAypw4ZMgRDhgxBy5Ytcfz4ccyfPx/ffPMNRowYgQsuuMCuXSedirgk+KxZs1S3bt1UzZo1lcfjUW3atFFPPPFEsZI9usvhF12GH2FKXEl0l+THKZfWly7JP3z4cOVyuYJK7SilL3GllFJ+v19NnjxZdejQQcXExKiYmBjVoUMHNWXKFOX3+4P6SiWJevbsqRISEtSxY8cCc3D6/iQkJKjevXurr776KmTscLdTyxedWvLI4XCohIQEdd5556k77rgjqByI5PQyJnFxcapVq1bqj3/8o/ryyy+D+r744osKgFq0aFHYeDNnzlQ4rbRWqBJX8+bNUwDUG2+8ETbW0qVLg0oxffzxxwqAmj59urhfunl88skn1TXXXKNiYmJUbm5u2BjDhw9Xbrc7UJLp9Dgul0s1btxYjRgxolhJHl2JK5xSmkV6zL/++mtxX8Mpij179mw1duxYlZSUpGrUqKH69++vdu7cWWy8p5eJeOONN1SrVq0CeWHGjBmB5+qpEKa0T9OmTYs97w4cOKBGjhypUlJSlNvtVsnJyap3797qtddeK/V+VkeRHp9Op1PVqlVLnX/++eq+++5TP//8s3Z7r776qgKgunTpEraP7jg9tRyiLu9v27ZNuVyuwHHx22+/qfHjx6tu3bqppKQkFRUVperVq6f69++vFi9eHPi70aNHKwBq27ZtYcf3+OOPK+B/Jbx0ZQCL8lRRTjo9r9uRH07POUop9Y9//EMBUAsWLAgbVyl9OSoA6u2331bt2rVTTZo00cbp2bOnSkpKUj6fL2TMqKgoddZZZ6m//OUvKjs7O+hvdSWuTs0Bp77+h7rt3r1bO0bJhg0b1KWXXqpiYmJUo0aN1JNPPqneeOONoDxaZMmSJapv377K6/WqmJgY1aJFCzV8+HC1Zs2aoH7btm1Tt9xyi0pOTlZut1s1atRIXX311eq9994rtl9S6S6isnL6cys6OlolJyeryy+/XE2ePDmodGWRkr6f1uUYpUKXuCrte0G76HLOqe85TN+HFcnMzFQxMTHq+uuv127/9Ne83377Td14442qWbNmKiYmRsXGxqpOnTqp6dOnFysdSGXHodR/r8JDRMU89NBDmD17NrZu3Rp02gyFVvTN/dy5c8u1/BsRhfaHP/wBO3bswLffflvRQyEiIqo2Kl2JK6LKZMmSJRg3bhwX0ERU5SilsHTp0rCnRxMREVHpcBFNpFFU3oaIqKpxOByshU5ERFQGnHIXIiIiIiIiIgIA/iaaiIiIiIiIyBC/iSYiIiIiIiIyxEU0ERERERERkaFKd2Exy7Kwb98+xMfHw+FwVPRwiKgKUkohOzsbDRs2hNNZvT4rZI4kokgwPxIRhVai/FhWBahfeeUV1bRpU+XxeFSXLl3U6tWrjf5u9+7d2kLlvPHGG2+mt927d5dViotIafOjUsyRvPHGmz035kfeeOONt9A3k/xYJt9Ev/POOxgzZgymT5+Orl27YtKkSejbty82b96MpKQk7d/Gx8cDAM56cDFcnrhSj8HpdJX6b4u4XfKnmC63O+LtSGOtabArMX6f2MchxHFHyfvicur7uGP8YgzA0rb6TUJA7qTfCuBzyftbR9hMHOQHJ9Ol/yTrhDNHjFEDx7Xtfp/BfPhMnhP6OSmQJhWA5dY/b47nnxBj5J8Qjmfhw8HC/FysmnBlIJ9UJpHkR+B/ObLFmEXaHCk9VFFR8vFgCUHcwrENAC638DJj8E2YE/pjqqbBOGJUgdhH+uLKHS3Xi3cJOcEdq8QYsAq1zUrJ+2tZ8muCNPcFJjnSpz9I4oXXDAA4Jjx+hXbkyAJ5Piy/weu5pX98jxvlSP3+Hs+Xj9UTUo7UPO0K83Ox8ol+1To/7gaQoOnnFeJkZspj9UpB7GDLQAxiZBrsTHlMmg0xMg1ieA0ePCmO0cNvx1iFdoMZk5kczDY8NuXxlAEM5kTTISsLSEmBUX4sk6tzd+3aFZ07d8Yrr7wC4OTpNSkpKRg9ejQefvhh7d9mZWXB6/Wi1SPfwhVT0YtokzeIZb+IjrNpES1NidtgX860RXTdclhEFxi8QYxFnra9vBbR+Ta8QcwzeIOYL73hFRfROVjxyCXIzMxEQoLurVT5iyQ/Av/LkWePXa3NkZVnES08z2xYRMfZtIh2SotoDxfRp6srLKITDBbRR4XHz2dHjiynRXReOeXIE9L+aBfROfh6bI9qnR8zoV9ES1+RmLwzLpezxW0ZiEEMZbAz5TFpNsQwWdaYnOovxTF6+O0Yq9BuyyLO5GC24bEprx9YiCPVdMjKOvlZgEl+tP3HMCdOnMDatWvRp0+f/23E6USfPn2wcuXKYv0LCgqQlZUVdCMiqo5Kmh8B5kgiOjMwPxJRVWL7Ivrw4cPw+/2oX79+0P3169dHRkZGsf4TJ06E1+sN3FJSUuweEhFRpVDS/AgwRxLRmYH5kYiqkgq/LOPYsWORmZkZuO3evbuih0REVGkwRxIRhcb8SEQVxfYLi9WtWxculwsHDhwIuv/AgQNITk4u1t/j8cBj8DszIqKqrqT5EWCOJKIzA/MjEVUltn8THR0djU6dOmHRokWB+yzLwqJFi5Cammr35oiIqgzmRyKi0JgfiagqKZMSV2PGjMGwYcNw4YUXokuXLpg0aRJyc3Nx6623Gsdwe2rA5YkN2+4Urj3rMrhaqx1MrmgtEi6b7Rau5AkAboOrqMrjkLfjEy7Xa3JlbfGxMdgVp0Enp0s/mNri9YuBGu7wx+DJjRgMVhiH3y9f9N/n01+pPi5Gf2VaAPDGy1e4TWmg3982rRqLMaSRfPjlVjHG3qP6dr/wuDgMrl5cUezIjwDgqlEDrpjw8yB9N2OWI6XniEFucktXATcZh/6anh6PHMPjivxq5M4ogxxZqA+i/HLecTn02xEuEA0AiJJyFwDLob/Ccx2n/irhAFDDpd+O5TT4ltCtz5E+n0GO9OtzZGyMnP+8brlP00b6sbRp1UiMkSMcAh9+uUWMsT9TuKK5KyZsm/JX//zoFS7PrWy4prESr2htso1yuF6xyZW3Dch7K2/HjhjiNmy48rZJHKOLVYvbkGNIh5EdF9a240rjJ7ej35DRM8KOy95LV1a36dL6ZbKI/r//+z8cOnQI48ePR0ZGBjp27IgFCxYUu1gEEdGZhvmRiCg05kciqirKZBENAKNGjcKoUaPKKjwRUZXF/EhEFBrzIxFVBZX3nB4iIiIiIiKiSoaLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhsrs6tyRcrqE8sliPWK5FqdU09igFCf8Tn2dTZMoUg+fJRdfNiiBKjPYjs+n31+DUtOAMGUmtbddJrWzhbHCL9dWPi59zmQwDh/0tUctn0EMn74Wq9slH++NvPJ2BnbX14GWK6ACq48IHfIy5SB+/TFgCYeIPBtVn9Nx8haOJRRkdBnMks+n7+N2GuRZ4bCzDMbhFmpa+6SdhVmOlMbiN9hOga9A2x5lUMPcEgpWm+RIt0HtbJfw+ELlizGOW8J2DMbhc+pzsVGO9OtzpCd82eSAFIMcOeBifRbUZ9CTvj2qb3fmZ4sxLJ9wQOvqhBu9UFdxQmlxh/BUNqtnK8Qwqs8s1LM1qOAr1yuOvCb2SeVRF9ugfnN51NY2YsP+2nCM2MGkbrZJnWj5OWHTWMRxSDW+w28jKysLXq+QQP7rDMikRERERERERPbgIpqIiIiIiIjIEBfRRERERERERIa4iCYiIiIiIiIyxEU0ERERERERkSEuoomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDURU9gHDcTjdcTnfYdp+/QIggfz7gcgkRnEKHcmIZfNThgyX28dswFghzYss2IM+7ZZk8NuGPHwDwCe0AUCD1saTjEMg7uknb7kGsGCPO00Tb3qi2HOPyHs3FPo2E9sNiBGDhkp+17XkF8pzBpZ936SlxJnw66I5ywxWlyZE+nxDBhhyp2X4ROTOZ9NCPtVCMYJYjpT4mx5Ul5EhlEkPIgU7Na2MghuUQ+3iE/FZgSccQIOVZq1B+vufnCDnSihFjSDmygVeOcXmPZmKfxkL7ITECsHDJRm173vHjYgynK07brjvOzoT8iEwACRU9CPnZLj9LZQ5pOwYbUSaJqVxEPiPKYGccBtuRojgcdsQQQxg8NgZBhC4ms25yiCghktn+Ssdz5PNulzMilxIRERERERHZgYtoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMsRFNBEREREREZGhSlsn2hnlgssdvlamZVI8WdqGUM/T5Yp8G36/XJtU4jMI4bOpQnPk5PqlUh3oQoPSpE5L3l/p0fMb1GeGU9/HlX9QDLHxPx9o2+vGyDWvL+g/Uts+pN/ZYgx9FVUz7604IvbZeDBf254nzOlJ+uPILR1mJoWDqziX0wWXppCzVRh5TpBypEnNR+no9is5z0o9fJZBDWi/Sc1jgV+uPukUBivVgD5J38eofLPB439CGKzPkp+rzih9vWJHfoYYY+OK97XtSTbkyKFXtRZjpIg9ZO9/c1Tss/FgnrY9xyRHOj3aZpcmR1pnQH6EN8K/t6HQrEkdYXkc8kDEWsQGVYCNRhp5OWqxBrDJWG1hNlghhg1Fng1iSMeRHTWRzeqEG9TflsZqtCGh1rRBBPk4swe/iSYiIiIiIiIyxEU0ERERERERkSEuoomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhqIqegDhOF1OOF3h1/hOpyvibbg08e3ahj0suYcl9ykfBuPwCxFMpt1gM9InRJY7zmA7MdrmOPjEEOc2jte25+35UYxxYSuPtr2JGMHMyv369qU/HhBjHEGitt3lkudMeuzc0kEktld9LncUotzhU7jfp59np1P+DNXh0LdHRUWeI52WMuglPeENEoJBjpS7GMQQ51WYVADw6WP43PLxbVnydvzCWAtdco50FkZr22MNcmTbJl5te94ugxzZuoa2PUWMYOY/Gfr2xT8KSRTAEUufI90GORLCY6fPkdU/P2ZmZiIhISFsu0NKbjDJS0KEyEMYMsgpApOhilNmMA6HLZNSbhMbOXnSREqaM5NNiFNm8NiZbEgYq1EMoYvRsWrQxw62fxP9+OOPw+FwBN3atGlj92aIiKoc5kciotCYH4moKimTb6LPO+88fPXVV//bSFSl/cKbiKhcMT8SEYXG/EhEVUWZZKeoqCgkJyeXRWgioiqN+ZGIKDTmRyKqKsrkwmJbtmxBw4YNcdZZZ2Ho0KHYtWtX2L4FBQXIysoKuhERVVclyY8AcyQRnTmYH4moqrB9Ed21a1fMnDkTCxYswLRp07B9+3ZccsklyM7ODtl/4sSJ8Hq9gVtKil2X/yAiqlxKmh8B5kgiOjMwPxJRVWL7Irpfv3648cYb0b59e/Tt2xefffYZjh07hnfffTdk/7FjxyIzMzNw2717t91DIiKqFEqaHwHmSCI6MzA/ElFVUuZXbKhVqxbOPvtsbN26NWS7x+OBx6Mv30NEVB1J+RFgjiSiMxPzIxFVZmW+iM7JycG2bdtw88032xpXrE4q1ICuSiy/yb5Ujv11mtS3dOvH6jIpFO2X+/jh1rYXGEyZG0e07TEIf5pZkS5de2nbf0OmGMOX85vQo6kYY5vYA5j1lb4O9AG/vr4pINc4res2qROtr2OaKYXwVY06kpHkR8uytPXhpWeI0+BpJtdNlp9E5VLDXhk8mQ3yqFhb3mRXhLGY5Uj98es2qAENv7yw8Cl9jpRrXgMuS58jY525YgwxR1oGOTJLypFNxBjhl2r/k/7VQW37IX9dMYbLUaBtr+sxqONs6Y+jTJ/mCe6rGnWiI8mPXq++9rjMjkqz8uuQVK/ajrLKdtWrFssVl1dxXonBQJTJYyPFMJhY+fEtp/cqws7Y9dDJu2My71KhaIM5k44Bm6bd9pXXn//8Zyxbtgw7duzAf/7zH1x33XVwuVwYMmSI3ZsiIqpSmB+JiEJjfiSiqsT2b6L37NmDIUOG4Pfff0e9evXQvXt3rFq1CvXq1bN7U0REVQrzIxFRaMyPRFSV2L6InjNnjt0hiYiqBeZHIqLQmB+JqCqpHD+kJSIiIiIiIqoCuIgmIiIiIiIiMsRFNBEREREREZEhLqKJiIiIiIiIDHERTURERERERGTI9qtz28b/35umWcdlwxAsS9pKObFMOrnLehRm3AVyH5d+h5wG++I0+Pwnyq/vU1CQKcaIc+v7nNsoVowR6z6uba+T1ESMsW3br9r2/V17iTHSvjoi9vntaL62PT7WK8bwIk/bfkv/s8UYeT59+4yP9PMhJohqwOk/eQtHOPzNtiG0FxYKD5TJNozGqe9kFcpJ0um0I0fKg3UKQ3G6CuXNiDnSI4/DYGKjfPo+x4+b5Mij2vY2DWLEGDFufR87cuS+bj3FGLMX6vcFALb+rs9vNW3Jka3EGDnCS+2MjzaHb4z8aVvpZQJI0HVQkW9DCuFwOCLfCExi6EdiMg5lw3yYBBF72DFlJg+uSRdhLLY8ujYcI8qOg9mm7dhzzJfD/thynPGbaCIiIiIiIiJjXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRIS6iiYiIiIiIiAxV3jrRVCX5DOqXuoQq3iZlsZ0GvWKQo233Qq6bfFXqedr2azrKFck/mzdP2x5rUL5247Zd2va3Plghxth1RFs1EwDgccZr25t45Lqxtw9qrm3vIJfWxuxt+napjnShQSneqs6CYQn5CLehZ/A5rA0f1VqWVDfZHsJmzAiDKTDIkW7ok4JlMu0GOyPlyHgcFmNcfXE7bXv/DvJgF7z/vrY92y3vy09Cjsyb/40YwyxH1tS2N/NkizFuHdRM295RvwkAQPpWfXu+plB8ockBVMVJ1brtqERrUH1ZDmLDQKTavHZV3S2PEsCRV8U26WAXk/rbdtTwliZNjuEQJqVS1ZqW2u2ora2Z06ysLHi9UgY5qfpnUiIiIiIiIiKbcBFNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRoaiKHgDJn2RY5TIKu8TIXfz65ijkiyFinQVin7qx+vbLO7QRY/yxo0vsI/EUHNa2uwqyxRh79+zVth/NXy2Po/6FYp9mDRpo20cPrCfGaOHWt+v35KRFy7Zr233Cs6YQDoOtVHFO6JOH8Dwz+gzVKWQfK/LsZBIhShhqoUEMZ6X5zNggRyr9WI1yJI6LfRLj9e19DXLk0A6Rz6uUIx0GOXL/nv3a9qP5K8UYNRp0EfucJeXIAXXlGNH69j1iBGDJ1zu07bpXSJPnS5WXCSAhkgDKoIvwOuOI/HXIJIIShmoUw2R/Kws7hmrw2Eg97BiG9NiZMHt8pRhyFGUwWPE4MhisPFaDGMJYTfbXRGV5V0FERERERERU6XERTURERERERGSIi2giIiIiIiIiQ1xEExERERERERniIpqIiIiIiIjIEBfRRERERERERIa4iCYiIiIiIiIyxEU0ERERERERkaGoih5AWK7/3sI1W2W//ne73WW+DSN+S+xiWX6xjx0zJo3ELQ8VHqd+rHVjfGKMdo28Yp/LuiZq27vHiyFs4cs7qm33GBxmTl+etj17/zYxxllNOoh9enWtp21vYcNT4oMl2WKf/Uf1+wtPnL69sJI8d8uS47+3MFwu/TPeMniuOoWk4TLIkU4hiGUwEDGGSY4slPNKlDhWMYTIjhyZGG2QI1Nqi30u66zvc0mCGMIW5ZEjc/dvF2O0bHa+2OfSznW17WdFiyFEHy3OEvvsFXNkbPi2wsr71s820lsEpW92aHJrIIQUQw4BJQzEYcdADJhsRwnbMRmqHcTNGE28wZxJcYymXR+knB5eeRtmOyMyOl4jZDLScjoUS76uWr58Oa655ho0bNgQDocDH3zwQVC7Ugrjx49HgwYNUKNGDfTp0wdbtmyxa7xERJUW8yMRUWjMj0RUnZR4EZ2bm4sOHTrg1VdfDdn+3HPPYcqUKZg+fTpWr16NmjVrom/fvsjPz494sERElRnzIxFRaMyPRFSdlPicnn79+qFfv34h25RSmDRpEh599FFce+21AIC33noL9evXxwcffIDBgwdHNloiokqM+ZGIKDTmRyKqTmz9YfH27duRkZGBPn36BO7zer3o2rUrVq5cGfJvCgoKkJWVFXQjIqpuSpMfAeZIIqr+mB+JqKqxdRGdkZEBAKhfv37Q/fXr1w+0nW7ixInwer2BW0pKip1DIiKqFEqTHwHmSCKq/pgfiaiqqfASV2PHjkVmZmbgtnv37ooeEhFRpcEcSUQUGvMjEVUUWxfRycnJAIADBw4E3X/gwIFA2+k8Hg8SEhKCbkRE1U1p8iPAHElE1R/zIxFVNbYWC2zevDmSk5OxaNEidOzYEQCQlZWF1atX4+677y5RLBeccGkKRVuQ62RKnE5NIepKRKr3atrHDtKMeSDXq/ZCX9/ygkZyAeee7fQ1oAEgtZzqQEs6dLhA237w4C4xxtHDh7Xtu/bniDHaJMoFalOEw0h+dIGVQsnqH7Yf0HcAgFh9kc9YoZ6vrxLWibYzPwKA2xmFKGf4/fT5pUfLpD5zJZlHYahuk6KQbo/BZvQbcrsN8qxQTNptlCP1VyO+IMUgR54n14m+qJKsN8QcmSHnyMwjR7TtO/bKv5U92yRHCu+aCsUIwMrf9O0/7DwkB4kRcmRU+GPVV3hCjl/O7M6PmZmZES6ozao8Rx7DBsJmlC37ItcAlupI/zeIwVgiZFRIOPICzUa7IozFjhrQRjWexbro9hwjchc7qjzbUeNbFyMLcqH5k0q8iM7JycHWrVsD/9++fTvWrVuHxMRENGnSBPfffz+eeuoptGrVCs2bN8e4cePQsGFDDBw4sKSbIiKqUpgfiYhCY34kouqkxIvoNWvWoFevXoH/jxkzBgAwbNgwzJw5Ew899BByc3MxYsQIHDt2DN27d8eCBQsQExNj36iJiCoh5kciotCYH4moOinxIrpnz57aUzYcDgcmTJiACRMmRDQwIqKqhvmRiCg05kciqk4q/OrcRERERERERFUFF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRIS6iiYiIiIiIiAyV+Orc5ccC4BfadeTPByxtfMAlRgDg18dwGgWR2BJEZPSJitCpQXysGKJLi3ra9qFd5WHs3bVP7DPxuRna9rEP/U3ekA3adU7Vtu/91S3GaNGotj7G/jwxxp5d28U+W1Z+rm33XdhbjPHDrgJte45LLmLv9ujnxA39NnyF5fOcqViFAHxCu44dOdIhxxByZJTBQ2VZuv0EnE75OQRLmg95RkxypOXUvzY18saJMS5skaRtv7mLPI7dO/eKfZ55/k1t+yN/eUTekA3addIn/b2/ygdJixQpR+aIMXbt2Cb22bJSfyz6LrxcjLFeyJFZDpMcqZ8TtyY3+FxnQn6U57AykDKo7mrmgRgOfRSTGBBiAIBBFHkz0jZM9leIogxGKu+tzGRa4ZA6mYzEltHqW40OEYNx2HCQSI+fyTDEORMfFzP8JpqIiIiIiIjIEBfRRERERERERIa4iCYiIiIiIiIyxEU0ERERERERkSEuoomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDlbZOtMdZgChndNh2pzNfH8Al1w31CfWXo4RatQDQJFZfAzXWyhRj1Pbq6xn6YFAD1aBOaqJQw9np19e/BIC4WI+2/cIWcgG3CxL17QZ7i+P6YQAA1n//rbZ9066dYow2TZoajEbPjXhte7Oze0W8jVYd5T6Hf10n9pn45Eva9oUvvCbGuHzUU9r2WHddMYbPr6+jKlWJl9qrAylHwnlC+/dOoxypn8koj/xEbBYXeY70xuufQ36D/GcJ+R4AEoUazk7huASAuJga2vYLW4ghcEEdfbvmUQ84rh8GADlH/rJ7lxjjnJQmBqPRi3IkaNubtr4s4m207CD3ObTlB7HPxCf0OXLRC/8SY/S+J/Ic6Vf656ZlhW+3zogMqeeQ6uYaBYm8fq891WojZ7QnYjFhkyg21AAWJ82g5rVJXeSIt1IuU2ZPqWmTCVEGGxK62FSyXCTVGzeqeW2A30QTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRIS6iiYiIiIiIiAxxEU1ERERERERkiItoIiIiIiIiIkNRFT2AcHxuN5TbHbbdgl/7905Y4jZcVoG2vb47X4zRIiZT296lSbwYo/cFDbXt4WfBbq5y21Kk8vPyxD5x3tra9g8/ekeM0WbUQ8Zjquzqnt1R7NOjVw9t+zZsEmM4Lf1j47aOyDEQq23PEZ4VhUJ+qA4K3B743R5ND6X9e8sqFLfhsnza9gaeE2KM5tFHte1GObJTI217tBjBLjXKbUuRysuRX7/i4r3a9o8NcuQ5I/9iPKbKrl6r88U+PXv31LZvw0YxRpSVo213Wybfb+hzZJ4z/Ot5obP650cvMgEkhG3XZ0fAAYe4DSmGUlIPwOGQt1Me5JHKc6KMokS2DZPtmM2p2R5rIxg9vgabEYj7a8OxajZOeX+lKTF5bOTnpsG8R7KRLAD6l8YAfhNNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRoUpbJzrLUw+umPA1/mJwWPv3ngJ9/WYAaJukr7N4cMWbYowV237Qtl87YawYo/zqQFcf8fFybVmXUPZ6/8H9YoxMn76msdedKMaoSq697U/adl+bnWKMhet362N49M87I1IZeLlMfJWX464DV7QmRwr1uN2F+vrNANA22YYcuUWfI6976m9ijPKrA119eONrin3cbv3n6Psz5Bx57IT+OKoVXVuMUZUMuPV2bXvB2TvEGIt+3Ktt93sMEpjQxan5jsRhVIe6asv0ejVVogGzOsECMUTkRYJNahGLRX7tKkUtxVGVo+a1CWUyKWLRYzsGEvlxaMchYhTDaDR2FMbWD8aowre4v+GjZCELXsNC0SXOpMuXL8c111yDhg0bwuFw4IMPPghqHz58OBwOR9DtyiuvLOlmiIiqHOZHIqLQmB+JqDop8SI6NzcXHTp0wKuvvhq2z5VXXon9+/cHbrNnz45okEREVQHzIxFRaMyPRFSdlPh07n79+qFfv37aPh6PB8nJyaUeFBFRVcT8SEQUGvMjEVUnZfLDmKVLlyIpKQmtW7fG3Xffjd9//z1s34KCAmRlZQXdiIiqq5LkR4A5kojOHMyPRFRV2L6IvvLKK/HWW29h0aJF+Pvf/45ly5ahX79+8Pv9IftPnDgRXq83cEtJSbF7SERElUJJ8yPAHElEZwbmRyKqSmy/OvfgwYMD/27Xrh3at2+PFi1aYOnSpejdu3ex/mPHjsWYMWMC/8/KymISJKJqqaT5EWCOJKIzA/MjEVUlZV7n4KyzzkLdunWxdevWkO0ejwcJCQlBNyKiM4GUHwHmSCI6MzE/ElFlVuaL6D179uD3339HgwYNynpTRERVCvMjEVFozI9EVJmV+HTunJycoE8Ft2/fjnXr1iExMRGJiYl44oknMGjQICQnJ2Pbtm146KGH0LJlS/Tt27dE23FZeXD5ww+vtu+w9u/vGHiOuI1rhLy8qe4FYoz6dUOfYlSkdoP6Yoz8/BPa9piYaDHGmSapbpLY56wWrbTtOb48McbK/6zUtve+tKcYw58tb0eyedsubfuug/LFVDINhvF7tk/bvmTjfjFGbKPW2vYYt0ceiH4YcMLStjuE9rJSXvkRABxWHhyaHOn1HdT+/Z0DzxW3cU1DffvGRDlHNki6XNtukiMLCvQHhMfjFmOcaerVM8iRzVto27OMcuQ32vbLe14mxvDl2JAjt+zQtpvkyOx8+XuF37MLtO1LfrIhR8bKx7NTSnFWYenaylB55kev2MOhbVVQBlvRx3Dom09uR5lsRwwijEMeiMk4bBgpHMKcmWzEYFplRg+v8PjaMgx5IAaPnsGWKsexahTBZDDidiJ/Tpgo8SJ6zZo16NWrV+D/Rb9FGTZsGKZNm4YNGzbgzTffxLFjx9CwYUNcccUVePLJJ+HxGLxxJiKqwpgfiYhCY34kouqkxIvonj17aj+N+OKLLyIaEBFRVcX8SEQUGvMjEVUnZf6baCIiIiIiIqLqgotoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMlTiElflJSlvF9wqLmz7Xf3O0/79NQ0iH4PXZYl9atdvrG3PtsLvQ5HfD+Zo21s1SRRjnHnkQunNmjXRtm/ctkWMsXzFCm17UmIdMcbRPXu17UuXrBRjvPbm+/pt5LvFGDfcM0bsE9+ohbY9sUUHMYbPGattz/OJISA981zCx3/WGfDxYFL+Hrgduhx5rvbvBzSMfAxet5wjvfX0G8o0ypFZ2vazU+Tn4ZlHfhKUS46sKz82mXv2a9uXLvlGjDF9hpAjTxjkyLvkHOlN0efIWgY50h+lz5E5JjlSeHjduvZK+87PRpmZQEJC+HbpLUT4Slz/C2FLDCGIpiTYKVEM+kROGquufJkpZTJpBlFkkc+ZHVtxGIxDnhM5htjDjmMV8uFqMutiDIMgUgxdc1YW4PXK2wD4TTQRERERERGRMS6iiYiIiIiIiAxxEU1ERERERERkiItoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGKm21wKRYP6Jr+MO2X6gvb2nkrWkvatsXfrVejDH0QX0t1g4X1RJjJIs9qHRc2taDBw+LETIzj2rbv/2PXON55rTXtO2r128SY9hhd6bcp3W7Vtr2bJP6peGftgAAv0+uLWxZ+g35nfoYfp/BQKu4ZCFHdmka+Tbenv6Stt0oRz7QVtve7mK5IGNDloEuI1KOPCJGOCrlyBUGOfKf/9LHWPeLGMMOu7PlPnGJrbXt2QVyfoPSf3/h8xXKMaDPcT5n+GKqvjMgP2YC0FSJFuvimtXEtaOmsR3jEGIYbcdsPGXNZH/tYFaP2oZC4OImDGo8l0vpbIONKKMqz0J7ecUo6wgn8ZtoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMsRFNBEREREREZEhLqKJiIiIiIiIDHERTURERERERGQoqqIHEE6Wszbcrviw7a9/uU/79/4DG8VtrPnusH4MnnPFGB9v038OsdInhkCDWH17ba8co9XZcp8L5C7VimX5te379+0XYyQm6id/y5YtYozV6zeJfcpDjQbyQXIY+oMxzycf0G5Y2nbL0rcDgN9XoO8gfPxX6DsubqOqy0JtuB2aHLlQf3z79v8kbmPNdwf1YzDIkR9uF3Kk/mkKQM6R3vDTENDaIEee75D7VCeFUNp2kxxZK1E/+SY58tt1v4h9ykONBm3EPlKOzJFyFwCPlL+sQjGG8uu3YznD59kzIT9Cet+kP/SNOCAlDIONCF2UHQMVxwkogz7lkR6VMthfYSAm4zTaF2EsyihIZXlRseM4siNG9cJvoomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjJUeetEF9ZAlC98PcaV2/Q1nt0+ubhybMfrte0xzjgxxm95Mfr2LUfkcfizte1Ol1xXN/5HeX/PrevWtt9zqVxsVSjXWqks/26btv2nXXlijIeu+6O2vX/3rmKM6/oP1LbvPSJ/ljV9zgpt+/GadcUYTVrJNX0PCmWg5SMR8Al1f31SBwCW2MelbS08IW6iyos0R7pMcmR7fY6MjUoQY/yWE61t37HlqBgjxifkSLdcVzf+Z3l/29b1aNvv6SG/JtQQe1QeUo7csCdHjCHlyGt6dBZjDLpmoLbdJEdOTdfnyIKaSWKMZi3kHJnh12dBO3JkgUGOhF/fx9LUpz0T8mPETEriCiWAjUoeOyIPIoUw2ReTasb21KyOnFyf20Tk++IwKRTtEGpNGxWbjnx/pePMrsfWaE6qkRJ9Ez1x4kR07twZ8fHxSEpKwsCBA7F58+agPvn5+Rg5ciTq1KmDuLg4DBo0CAcOHLB10ERElQ3zIxFReMyRRFSdlGgRvWzZMowcORKrVq3CwoUL4fP5cMUVVyA3NzfQ54EHHsDHH3+MuXPnYtmyZdi3bx+uv17/bQYRUVXH/EhEFB5zJBFVJyU6nXvBggVB/585cyaSkpKwdu1a9OjRA5mZmXjjjTeQnp6Oyy67DAAwY8YMnHPOOVi1ahW6detm38iJiCoR5kciovCYI4moOonowmKZmZkAgMTERADA2rVr4fP50KdPn0CfNm3aoEmTJli5cmXIGAUFBcjKygq6ERFVdXbkR4A5koiqJ76HJKKqrNSLaMuycP/99+Piiy9G27ZtAQAZGRmIjo5GrVq1gvrWr18fGRkZIeNMnDgRXq83cEtJSSntkIiIKgW78iPAHElE1Q/fQxJRVVfqRfTIkSPx008/Yc6cORENYOzYscjMzAzcdu/eHVE8IqKKZld+BJgjiaj64XtIIqrqSlXiatSoUfjkk0+wfPlyNG7cOHB/cnIyTpw4gWPHjgV9knjgwAEkJyeHjOXxeODx6MuKEBFVFXbmR4A5koiqF76HJKLqoETfRCulMGrUKMyfPx+LFy9G8+bNg9o7deoEt9uNRYsWBe7bvHkzdu3ahdTUVHtGTERUCTE/EhGFxxxJRNVJib6JHjlyJNLT0/Hhhx8iPj4+8BsVr9eLGjVqwOv14vbbb8eYMWOQmJiIhIQEjB49GqmpqSW+qqLbbyHK7w/b7ncKnzx6ksRt+CxL2+43+owh/BgBwGUQwueM0bcbjONoprydzKP6WotXXxovxmgjb6Zc+Az67M9L1HdI7CDG6N57kLa9kbAJAGjUoLHcSbC3rv4NxE/bdogxfJY8a279UwJO/eEOALAsfSfLL2wEgM/v0ncQnliFkV0zsVTKMz8CQJS/EG5/+MfUZ0OOzCnUP1bKhhxpGTybfZ5ofTtixRgmOTJbypGXxIkxWjvk7ZSHEwZ9xBxZS86RPXrryw81rCOPo2Fy5L9j3Z2oz5Ebt+8SY/gMjkW38L7BKEdC6dsNcqRfyJGWprki8iNQ3jnSq211CI+BvrUohtRBTgbSdgxCAErqZMve2BTCZCwRKodNADDaX3kocg+HcBDYsrvlNWfVTIkW0dOmTQMA9OzZM+j+GTNmYPjw4QCAl19+GU6nE4MGDUJBQQH69u2LqVOn2jJYIqLKivmRiCg85kgiqk5KtIhWSv6oIiYmBq+++ipeffXVUg+KiKiqYX4kIgqPOZKIqpOKOaeHiIiIiIiIqAriIpqIiIiIiIjIEBfRRERERERERIa4iCYiIiIiIiIyxEU0ERERERERkaESXZ27PDmgX+H7oa+RaFCqEXAKtWiNCDUjDSLI+2IwToOC1H64DUZTNWzcL/epe3YXbXuXFnIN1CSDOtCR+t6gfq1UBzq7wKRytkE9XqvsP1eTK6Ca9BFqpJo8Z6q9yPOK5RTym1M+Xiyhrq7HoLCusiNHGtT09Dv1OdKqJDWgTWzMkPvUPftCbXuXFu3EGEkGdaAjtfaY3EfMkT75OHMa5Ei/mCNNMpw+hh05UpcDpedTdaCvEm3ChtrKBlcjF2sAG9Xv1XcySltG9ailEDYEMWJHjW8bkrkd5bdtqDVtx8uSXY+cEo/FKvQiaoDfRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMsRFNBEREREREZEhLqKJiIiIiIiIDHERTURERERERGSIi2giIiIiIiIiQ1EVPYDqz1UpQgCAZVn2BKoEVm8/IfY56vRq22PdeWKMHKG9rhhB9s33B8Q+eQU+bbsPbjFGQYFf7OOXPldzmhyM8naI/seGBOeQc5vTKX9mXOgrjHwslcTq7QViHzFHuuQcmSW01xEjyL5ZlyH2yffp847Pko8z33F9ngUAv5RrDXKkBSX2iZzueOf3J2cUh0PsogwOSQfkOPKGIg8BYRyOcnl+icMw6mI072IQOYYSOjkMjpHyUj4j0c1HFgD9a2MRZlIiIiIiIiIiQ1xEExERERERERniIpqIiIiIiIjIEBfRRERERERERIa4iCYiIiIiIiIyxEU0ERERERERkSEuoomIiIiIiIgMVdo60QoWLFSfusYRMSm7a1Mt6cpgr0GfH3ZJFZyBfLdQ582Sa4IeFro0k8sz43shxg/b9osxfE5hQ+5YOYZcNhZ+4UCy4zBzugw+u5MeGr/wpJDaq4FqkyMNdsGSypcb1Mm0DLYjHZmV5VPnPQZ91u+SazxLOdJpkiNP6NubR4shsFYoz71+2wExRoGQnZweOUcWyLsLJdSbdhnUI5dee6Lc8lszn18YrK5d+luqPExqEQv5z6QWscl2xDgGG5I2Y1KvWAnbMSl5LMUA5LrYdtR4NqqJLD2+BoWipXk1OkRM6lELO2TPvEcewy6V5T0BERERERERUaXHRTQRERERERGRIS6iiYiIiIiIiAxxEU1ERERERERkiItoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyBAX0URERERERESGoip6AGFZ/71VcSa7YMcnGX4bYuTkGXSKtWFDghW/yn32Z8qDdbk92nZfQYwY48ft+nZfczEEZn+0R9ueaTCp7hhhX/wueSBO+UgrhBDH8snb8euPekto/28n/SaEZ41VDXKHxGEBTs1+itNsQ+KxY5otg3FIXQyOSvhN9ldIpDm5BjFqGvSJ0IrNcp+9NuTIggI5N4k5soUYAu98tFfbftQoR7q17fk25cgTQo6MFnIXAFi+8siR4d/e+U2edFVcJoAETbvDIQRQkY9B2oQRg3FIXUzGoWzYX3lSASWM1pY5M4jiMHqA7Rhr5HtUPnNmD1ueVuXw3NQHyQLgNYpSokw6ceJEdO7cGfHx8UhKSsLAgQOxeXPwq3nPnj3hcDiCbnfddVdJNkNEVOUwPxIRhcccSUTVSYkW0cuWLcPIkSOxatUqLFy4ED6fD1dccQVyc4M/nr/jjjuwf//+wO25556zddBERJUN8yMRUXjMkURUnZTodO4FCxYE/X/mzJlISkrC2rVr0aNHj8D9sbGxSE5OtmeERERVAPMjEVF4zJFEVJ1E9MOYzMxMAEBiYmLQ/Wlpaahbty7atm2LsWPHIi8v/G+zCgoKkJWVFXQjIqrq7MiPAHMkEVVPfA9JRFVZqS8sZlkW7r//flx88cVo27Zt4P6bbroJTZs2RcOGDbFhwwb89a9/xebNm/H++++HjDNx4kQ88cQTpR0GEVGlY1d+BJgjiaj64XtIIqrqSr2IHjlyJH766SesWLEi6P4RI0YE/t2uXTs0aNAAvXv3xrZt29CiRfHLdI4dOxZjxowJ/D8rKwspKSmlHRYRUYWzKz8CzJFEVP3wPSQRVXWlWkSPGjUKn3zyCZYvX47GjRtr+3bt2hUAsHXr1pAJ0OPxwOPRl9kgIqoq7MyPAHMkEVUvfA9JRNVBiRbRSimMHj0a8+fPx9KlS9G8uVwkd926dQCABg0alGqARERVAfMjEVF4zJFEVJ2UaBE9cuRIpKen48MPP0R8fDwyMjIAAF6vFzVq1MC2bduQnp6Oq666CnXq1MGGDRvwwAMPoEePHmjfvn2Z7ACVhP46cvn5BiFiIx+F/jJKwMofjxhEka+J5/QV6Du43WKMb1Zv1bev9IkxDvpitO1Oj1zU3e3StztN5sMlBAEAn1/bbMm7C/j0nXxC+8lhWFIPbavlNxmovco7PyoA2lmSDglpig04DYLYsBlYlhDFaXBsG9FPWp5JjqwZ+ShyhfaVP/4uxjDJCSiQcqQ8r998t0XbvnK1PqcAwMECfY6EJ16M4XEL+yvsKgBEGbwmFBQIueeEvL+WlCP9cgyfT3pOhN9hy28wGWWgPHOk+KqqhHZHiTYXZhNyEBs2A4dDiCLtq00cRhuKfI/tmDOjkUobUgYjETakTEYibkYeh1L67YjHEMzG6rDl0bGBHY+dgRItoqdNmwYA6NmzZ9D9M2bMwPDhwxEdHY2vvvoKkyZNQm5uLlJSUjBo0CA8+uijtgyWiKiyYn4kIgqPOZKIqpMSn86tk5KSgmXLlkU0ICKiqoj5kYgoPOZIIqpOIqoTTURERERERHQm4SKaiIiIiIiIyBAX0URERERERESGuIgmIiIiIiIiMsRFNBEREREREZGhEl2d+0zjckX+GUN5fUrhMqjG6hQKZcoVMu3x/QF9+8HMbDGGyy3UFQXgF+poWk55j48W6Ot1up3yOHxOj7bdaclHiViu1aC+qUsuPQonhDrRlkEQYX/8BvsrlomGfkL8ll11g6sxk+QkPA4ul/wS4nDog9gwDNvyrNOqJDnyoL79cLZUSRpwuaLFPnL9bX3uAoCjx/O07Z4oOUcWCLnYacnHmZiZDGpeu5R8JInl1wvlGvWWsD8+S67H6pdGosmBzI8GbCl5XD51k+0YhVEvsZawyb6IhZMjJl0FHoDBvhhsx6Rusg01vOXay+VUCNwORvMu1LQ22Y54DNjzvOM30URERERERESGuIgmIiIiIiIiMsRFNBEREREREZEhLqKJiIiIiIiIDHERTURERERERGSIi2giIiIiIiIiQ1xEExERERERERniIpqIiIiIiIjIUFRFDyAcP/xwwF+hY3A6XRW6/SImo4gxmKtEt77dG282nkit2nhI2+6CT4zhd8eKfQpcHm2702+JMaTPmeSRAk7hETR5fH1+fS/LMhiJwf66ffo40jhO9ol8HPDrD1ZLeFwsq/p/PijmSGmabZgisxzp0LaaPAvFLRgEiXHJz5FaQo6s7TUcUIS+3XhY2+6wCsQYfo9JjtQfBFFKDAEL+u34DQ40p3CMuA2OVTFHFhrkSIP9dfv0c+8XchcA+JWwQ75CMYZT2E6hZt4tf6V961d+9Iec0bFQHqRhGsUwCaIi35JS8qTJY5HHoYQHx2GwwyZjNZu4imeyK5WHPFhpf+x4fB3acWQBMHuxr/7vNImIiIiIiIhswkU0ERERERERkSEuoomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMhQpS0WaPktWLp6spa+GG1lqfFsB7dBfdPasfLnIXF+fd3kNT+fEGNkNojWtv+eJ4bArj36GqhwxYgxXAbVlaVyxU6DCrX+SnIcWcLx7hPqOwOA36A+s9+vj2NSf9my9NsRmo36+IXHzm+ykSpOzJFiPUaTY7vsP2c12oJT38vlkI//OrHyy12CT8qR8naO1tfX7z2aL4bArl0Hte1O23Kkvt5mlJhF5RypjB7hyJ+vduRIk3qrfr9+O9I4TvaROshzVihsx2+FrzXt98t1qKs9qRatHUFsYLQFqSauSV1doy3ZUrU68nEI+1OlyiZXO9LsGxUtF1oN6pGLHeypAc5voomIiIiIiIgMcRFNREREREREZIiLaCIiIiIiIiJDXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRoaiKHkA4fp8fjih/2HaXqxwHU8E8/nyxj5WdJ/Y5mKfvs/j7TDHGN263tt1t8MDk+Cxtu88lf7bjR/hj43+qxkHi9+vnAwAsS7+/dsQw244cQ+piEsPn8+k7OPXHiGWwjapOzJFu4XlkyceMU5pnGMQQ2s1i6KPYliNzIs+RX7v0OdIj5FCgPHOkw6BPpOTHV+IrpxwJgzmzLH1u8hlMu8+vf/x8JjmyoEDfQfPcPRPyY8Qc8nNDKaUPYfD80kcwjSFFMWFHLiiPfFLdRD5n5TfrVWisBs9fO5Tom+hp06ahffv2SEhIQEJCAlJTU/H5558H2vPz8zFy5EjUqVMHcXFxGDRoEA4cOGD7oImIKhvmRyKi8Jgjiag6KdEiunHjxnj22Wexdu1arFmzBpdddhmuvfZa/PzzzwCABx54AB9//DHmzp2LZcuWYd++fbj++uvLZOBERJUJ8yMRUXjMkURUnTiUdD6KIDExEc8//zxuuOEG1KtXD+np6bjhhhsAAJs2bcI555yDlStXolu3biH/vqCgAAWnnJaUlZWFlJQUdH/uR0TViA+7XemsYacz8lN53Qan3ZWHWMinKsZAPlXRJ5zOHRsr7690KqLR6dz5+tPh8lxxYox8Z6zYJ8/Sj8XtF04ZrkTK63Ru6TRqn3woQjoTWzxVGwjKCSEJpxn7C3Kw5cWLkZmZiYSEBHF7ZSXS/AiEz5EX/32DPkcKp3M7DU6skk7ndrnl57tTODQtg49ypdO5Yyw5/8U6DXKkcDp3bJycI912nM6dpz/+jXJklNwnp1B/DHiUyenN+gdYOIT+G0PoYBBEzJG+QnkgBqdzFxTYkCOl07mFbZwcx3F9B82c+QtysPWl7hWeH4Gyew8pkt76ltfp3EIXh8E7dOl0bofRvsjbEXfHIIY0FKMliR2n6hpsR5o3k7HKMcQQ8u5Gtowz3IhNYzUhPjdNgkQ+EJP8WOoLi/n9fsyZMwe5ublITU3F2rVr4fP50KdPn0CfNm3aoEmTJli5cmXYOBMnToTX6w3cjJIfEVElZld+BJgjiaj64XtIIqrqSryI/vHHHxEXFwePx4O77roL8+fPx7nnnouMjAxER0ejVq1aQf3r16+PjIyMsPHGjh2LzMzMwG337t0l3gkiosrA7vwIMEcSUfXB95BEVF2U+OrcrVu3xrp165CZmYn33nsPw4YNw7Jly0o9AI/HA4/HU+q/JyKqLOzOjwBzJBFVH3wPSUTVRYkX0dHR0WjZsiUAoFOnTvjuu+8wefJk/N///R9OnDiBY8eOBX2SeODAASQnJ9s2YCKiyor5kYgoPOZIIqouIq4TbVkWCgoK0KlTJ7jdbixatAiDBg0CAGzevBm7du1CampqyeMqS3/BEukiSQbXBJMuPmZyIaZIt2FCqksJAD6fcCEmAH5XjLY9z+0VY+RJE+uTL94j/YrA55fnzKS2rOsMKoVpx0XDTvYR6tMalFr1CYVSpQsRmfALMaT28lJW+RGQc6RTKlhrcBFAaRZdBtNcKPSJcsrJWno4TY7/AltyZG0xhiVc0NBpcOUpaVpNcmShdEU3yDnS5KJv9pCriUcawW/JV8XxGVxsUrxwosFYxRxpQ21tpYmhaytvZZkjdWyprGzHRaOEdpMa0PLFnAxi2HAhJrN61VWo1rCg/MYhVxOXI0gx7HrshDjSlfRM2DBU3XMzKwvwysshACVcRI8dOxb9+vVDkyZNkJ2djfT0dCxduhRffPEFvF4vbr/9dowZMwaJiYlISEjA6NGjkZqaqr3yLBFRdcD8SEQUHnMkEVUnJVpEHzx4ELfccgv2798Pr9eL9u3b44svvsDll18OAHj55ZfhdDoxaNAgFBQUoG/fvpg6dWqZDJyIqDJhfiQiCo85koiqk4jrRNstKysLXq8XF/19PaJiwtdAdQr1HF0GtTilU61drsjPZbPjdO4YK1vs4/bliH38wqnYrjiT8xeEeTWo1wq//rHzQX9KJQD4hFqsgFyL01mJTmmTSKermp2qbcPp3AaltX35kZ/OLY1VOnnXX5CD316+pFLUQbVbUY5MfXadNkdGOfWp3WVwOjeE/OUWalED8qnYUVGRn85tlCP9WWIfv6UfizvBhtO5YXA6t1Sv3SRHuuWLLUnPZ5Maz9KDY1YnWuhkcGo6hNO1TWov+/zyKf9SCXuf3yC/5evHapQjhYEozZT6C3Kw7aUe1To/SsS3vgYFb+0o32tH3WR7avMaBLFlrGVfe9mIPQ9OxDHMQkR+irR4OrfRmdoGzwkbxmrLjy0imPei07nLtE40ERERERER0ZmGi2giIiIiIiIiQ1xEExERERERERniIpqIiIiIiIjIUMR1ou1WdFGBwnz9hbKkC4upwsgvLKZsuLCYw4YLi7kMLpoDX67YRbqwmGW0v+VxYTGDC8AYXFisULiwmKMKXVhMCRcWKzS44ldhoUEfn/6CDgZlVOEviPzCYv4ILyxmFZx8PlSy6ybawjRHQrjAh7LhwmIolHOGEh5uZcOFxYxypN/g4ovChcXgll8y7bmwWKG23ShHFp4Q+0gpwVGNLixWaHBhsUK/wZwJXfwGFxbzSxcWk540APwRXFjsTMiPkqws+UKDkbJjE+UxzpMbsiGEDWMtv/21ZYfLKYTQySBGli0PsA2d7IhhA928F7UZXeSusl2de8+ePUhJSanoYRBRNbB79240bty4oodhK+ZIIrID8yMRUWgm+bHSLaIty8K+ffsQHx8fuIx9VlYWUlJSsHv37mpXjqGicE7LBufVfqWZU6UUsrOz0bBhQziN6vRUHafnSB5z9uOclg3Oq/2YH4MxP5YPzqv9OKdlo6TzWpL8WOlO53Y6nWFX/gkJCTywbMY5LRucV/uVdE5NaoVWReFyJI85+3FOywbn1X7MjycxP5Yvzqv9OKdloyTzapofq9dHkERERERERERliItoIiIiIiIiIkNVYhHt8Xjw2GOPwePxVPRQqg3OadngvNqPc6rH+bEf57RscF7txznV4/yUDc6r/TinZaMs57XSXViMiIiIiIiIqLKqEt9EExEREREREVUGXEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjJU6RfRr776Kpo1a4aYmBh07doV3377bUUPqUpZvnw5rrnmGjRs2BAOhwMffPBBULtSCuPHj0eDBg1Qo0YN9OnTB1u2bKmYwVYREydOROfOnREfH4+kpCQMHDgQmzdvDuqTn5+PkSNHok6dOoiLi8OgQYNw4MCBChpx5Tdt2jS0b98eCQkJSEhIQGpqKj7//PNAO+czNObHyDA/2o/5sWwwR5YOc2TpMT/aj/mxbFRUfqzUi+h33nkHY8aMwWOPPYbvv/8eHTp0QN++fXHw4MGKHlqVkZubiw4dOuDVV18N2f7cc89hypQpmD59OlavXo2aNWuib9++yM/PL+eRVh3Lli3DyJEjsWrVKixcuBA+nw9XXHEFcnNzA30eeOABfPzxx5g7dy6WLVuGffv24frrr6/AUVdujRs3xrPPPou1a9dizZo1uOyyy3Dttdfi559/BsD5DIX5MXLMj/ZjfiwbzJElxxwZGeZH+zE/lo0Ky4+qEuvSpYsaOXJk4P9+v181bNhQTZw4sQJHVXUBUPPnzw/837IslZycrJ5//vnAfceOHVMej0fNnj27AkZYNR08eFABUMuWLVNKnZxDt9ut5s6dG+jzyy+/KABq5cqVFTXMKqd27drqX//6F+czDOZHezE/lg3mx7LDHKnHHGkf5seywfxYdsojP1bab6JPnDiBtWvXok+fPoH7nE4n+vTpg5UrV1bgyKqP7du3IyMjI2iOvV4vunbtyjkugczMTABAYmIiAGDt2rXw+XxB89qmTRs0adKE82rA7/djzpw5yM3NRWpqKuczBObHssf8aA/mR/sxR8qYI8sW86M9mB/tV575MSrSwZaVw4cPw+/3o379+kH3169fH5s2baqgUVUvGRkZABByjovaSM+yLNx///24+OKL0bZtWwAn5zU6Ohq1atUK6st51fvxxx+RmpqK/Px8xMXFYf78+Tj33HOxbt06zudpmB/LHvNj5Jgf7cUcaY45smwxP0aO+dFeFZEfK+0imqgqGDlyJH766SesWLGioodS5bVu3Rrr1q1DZmYm3nvvPQwbNgzLli2r6GERUSkxP9qLOZKo+mB+tFdF5MdKezp33bp14XK5il097cCBA0hOTq6gUVUvRfPIOS6dUaNG4ZNPPsGSJUvQuHHjwP3Jyck4ceIEjh07FtSf86oXHR2Nli1bolOnTpg4cSI6dOiAyZMncz5DYH4se8yPkWF+tB9zpDnmyLLF/BgZ5kf7VUR+rLSL6OjoaHTq1AmLFi0K3GdZFhYtWoTU1NQKHFn10bx5cyQnJwfNcVZWFlavXs051lBKYdSoUZg/fz4WL16M5s2bB7V36tQJbrc7aF43b96MXbt2cV5LwLIsFBQUcD5DYH4se8yPpcP8WH6YI8NjjixbzI+lw/xYfsolP0Z0WbIyNmfOHOXxeNTMmTPVxo0b1YgRI1StWrVURkZGRQ+tysjOzlY//PCD+uGHHxQA9dJLL6kffvhB7dy5Uyml1LPPPqtq1aqlPvzwQ7VhwwZ17bXXqubNm6vjx49X8Mgrr7vvvlt5vV61dOlStX///sAtLy8v0Oeuu+5STZo0UYsXL1Zr1qxRqampKjU1tQJHXbk9/PDDatmyZWr79u1qw4YN6uGHH1YOh0N9+eWXSinOZyjMj5FjfrQf82PZYI4sOebIyDA/2o/5sWxUVH6s1ItopZT6xz/+oZo0aaKio6NVly5d1KpVqyp6SFXKkiVLFIBit2HDhimlTpYpGDdunKpfv77yeDyqd+/eavPmzRU76Eou1HwCUDNmzAj0OX78uLrnnntU7dq1VWxsrLruuuvU/v37K27Qldxtt92mmjZtqqKjo1W9evVU7969A8lPKc5nOMyPkWF+tB/zY9lgjiwd5sjSY360H/Nj2aio/OhQSqnIvssmIiIiIiIiOjNU2t9EExEREREREVU2XEQTERERERERGeIimoiIiIiIiMgQF9FEREREREREhriIJiIiIiIiIjLERTQRERERERGRIS6iiYiIiIiIiAxxEU1ERERERERkiItoIiIiIiIiIkNcRBMREREREREZ4iKaiIiIiIiIyND/A9Gjvh4Buo2VAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9EAAAFTCAYAAAA+14+JAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABxJElEQVR4nO3dd3yUVdYH8N8kJBNCkglJSCMEQpHepBlFBEGKCiJYQNYFdi2r4CrurisWFCy4ujaWtlYUAigiKoqIUi2AFEF6kxBaAqGkkkJy3z94M8uQmXtuMk9I8ffdz3xWcu6ce+eZZ07mzkzm2JRSCkREREREREQk8qnsBRARERERERFVF9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyxE00ERERERERkSFuoomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQT0WWTlpaG2267DeHh4bDZbHjjjTcqe0lEROUyatQoBAUFGY212Wx49tlnK3ZBRER02VSZTfSzzz4Lm82G9PR0t/E2bdqgZ8+ezn8nJyfDZrPBZrNh4cKFRvlGjRrlvI7NZkNISAjat2+PV199Ffn5+aWu6+mSmppaag02mw0+Pj4ICwvDgAEDsHbtWgDArFmztLlKLo0aNXKuUfdLOSgoCKNGjXL+e9WqVS55fH19ERkZidtuuw27du0qdf1Lj0GtWrXQoEEDDBs2DDt37nQZe2nuSy/z5893jm3UqJHLcQgNDUXbtm1x3333Yf369R5vz6V69uzpkickJATNmzfH3XffjW+//VZ73V27dsFmsyEgIABnz571eJs9XS4+rmfPnkVAQABsNpvb41hWNpsNY8eONR4/ffp02Gw2dOvWTZvz4kudOnXQqlUrPP/888jNzXUZqzsGAQEBznEl9/knn3xS9htpYNy4cfjmm28wfvx4zJ49G/3796+Qecg6ZT0Xa9WqhbCwMHTq1AkPP/xwqboyaNAgBAYGIisry2O+ESNGwN/fH6dOnSqV/9LLX/7yF+f1Lj3P7XY7rrjiCkyYMAF5eXml5snOzsYzzzyDNm3aoE6dOggPD0eHDh3w8MMP49ixY27XtmTJEthsNsTGxqK4uNjtmIvrYcljs2vXrvjwww9LjZUec6bHvyz1Rafk99q///1v4+s89thjsNlsuPPOO7U5L/3926FDB0ydOhVFRUUu4y/+PXDppUWLFs5xJb9fN27cWL4bS0SiS5/HBgQEIDY2Fv369cOUKVPc1vKyPo++tN7oasDFl4tfmPL0HLAst7O8L3SdPXsWkZGRbmv5hg0bMHbsWLRu3Rp16tRBfHw87rjjDuzdu9d4XdJxpMpRq7IXYIVJkyZhyJAhsNls4li73Y533nkHwIWTfuHChfj73/+ODRs2uGwKAWDGjBluN7ShoaEu/x4+fDhuvPFGFBUVYe/evZg+fTp69eqFDRs2oEePHpg9e7bL+HvuuQddu3bFfffd5/yZ6avZnvz1r39Fly5dUFhYiF9//RUzZ87EqlWrsH37dkRHR7uMvfgYnD9/HgcOHMDMmTOxdOlS7Ny5E7GxsW5zXyoxMdHl3x06dMDf/vY3AEBWVhZ27dqFBQsW4O2338a4cePw2muvGd2WuLg4TJ48GQCQk5OD/fv349NPP8WcOXNwxx13YM6cOfDz8yt1vTlz5iA6OhpnzpzBJ598gnvuuQcAcP/996NPnz7OcQcPHsSECRNw33334dprr3X+vEmTJs7/XrBgAWw2G6Kjo5GUlITnn3/eaO1WSUpKQqNGjfDzzz9j//79aNq0qdtxN9xwA/74xz8CuLAh+P777/H0009j69atWLBggcvYi+/3i/n6+lp/AzxYsWIFbrnlFvz973+/bHOSd8p6LiqlkJGRga1bt+KDDz7A9OnT8a9//QuPPvoogAsb5MWLF2PRokXOc/diubm5+Pzzz9G/f3+Eh4eXyn+pK664wuXfF5/nGRkZ+Pzzz/Hcc8/hwIEDSEpKco4rLCxEjx49sHv3bowcORIPPfQQsrOzsWPHDsydOxe33nprqVp48fFITk7GihUrXGrLxS6uh8ePH8c777yDkSNHIj8/H/fee6/b67hjevwri1IK8+bNQ6NGjbB48WJkZWUhODjY7diS35XAhftmyZIleOihh3Do0CG88sorLmMv/j1wMYfDYf2NuAzOnTuHWrVqxFMu+p2aNGkSEhISUFhYiNTUVKxatQqPPPIIXnvtNXzxxRdo165dqeuYPo++1JNPPul8Dgdc2IhOmTIFTzzxBFq2bOn8+cVzenoOqJOSkoKcnByXnABw7NgxnDp1Cm3bthVzAMCECRNKvXlR4l//+hd+/PFH3H777WjXrh1SU1MxdepUXHnllVi3bh3atGljNEfJ8b+YdBypgqkq4plnnlEA1MmTJ93GW7dura677jrnvw8ePKgAqA4dOigAauHChWK+kSNHqjp16riMKyoqUp07d1YA1NGjR43WcukaXnnlFZeff/311wqAeuCBB9xer06dOmrkyJFuY+7WqLvuypUrFQC1YMECl3EzZsxQANS//vUvo/xffvmlAqDeeustMbc7DRs2VDfddFOpn+fm5qrBgwcrAGr69Olinuuuu061bt261M/Pnz+vHnzwQQVAPfbYY6XixcXFqlGjRurRRx9Vt956q+rZs6fHOTZs2KAAqPfff9/jmB49eqghQ4aocePGqYSEBHHdEgBqzJgxRmN/++03BUB9+umnql69eurZZ58tU87bbrtN+fj4qHPnzjl/Jp1XJcpyn5eHzWYzOg7Z2dkVMj+VjbfnYnp6ukpMTFQA1FdffaWUulATgoODVb9+/dzmmjt3rgKg5s+fL+a/lLvzvLi4WF111VXKZrOp1NRU588//vhjBUAlJSWVynPu3DmVkZFR6ufZ2dmqTp06asqUKapjx45q1KhRbtfhrh6eOHFCBQUFqZYtW7r8XPeY8/b4l4en32uerFixQgFQK1asUH5+fmrWrFnGOYuLi1WXLl1UbGysy889/R641Pvvv68AqA0bNhit1WqmdZWoOtM9zpYvX65q166tGjZsqHJzc50/9/Z59KUWLFigAKiVK1e6jZflOeCl62/YsKF67rnn1FtvvaUmTJigZsyYoRo2bOjyO0hn27ZtqlatWmrSpElua/mPP/6o8vPzXX62d+9eZbfb1YgRI8T8lV3nyLMq83Hu8ho2bBiuuOIKTJo0CUqpMl/fx8cHPf//Y+LJycmWrKnk3c0DBw5Yku9yrKHk3WqrXymvXbs2Zs+ejbCwMLzwwgvluo+AC++WTpkyBa1atcLUqVORkZHhEv/xxx+RnJyMYcOGYdiwYVizZg2OHDlSrrlSUlLw/fffO3MdPHgQP/30U7lylUdSUhLq1q2Lm266CbfddpvLu2cmoqOjnR+rrSpKPo6klMK0adOcH0W6OLZ69Wo8+OCDiIyMRFxcnPO606dPR+vWrWG32xEbG4sxY8a4/ajWtGnT0LhxY9SuXRtdu3bF999/j549ezof31R23p6L4eHhmD9/PmrVqoUXXngBwIWaMGTIECxfvhwnTpwodZ25c+ciODgYgwYNsuQ22Gw2dO/eHUop/Pbbb86fl9TGa665ptR1AgICEBISUurnixYtwrlz53D77bdj2LBh+PTTT91+TNydevXqoUWLFmX6veDt8b8ckpKS0KpVK/Tq1Qt9+vQp0xptNhuioqKqVK0qUVhYiIkTJ6JZs2YICAhAeHg4unfv7vbPio4ePYrBgwcjKCgI9erVw9///vdSH1G/9KOnJR933b17N+644w6EhIQgPDwcDz/8sPE5RVTZrr/+ejz99NM4dOgQ5syZU2nrKO9zwOuvvx7btm1Dbm4uxo8fjzfeeANbt27Fxo0bPf55yqUefvhh3HrrrS6fbLzY1VdfDX9/f5efNWvWDK1bty7znwtmZWWVqi1Uear9JtrX1xdPPfUUtm7dikWLFpUrR8mTmos/OggAp0+fRnp6usvF5O8sSjbjdevWLdd6rCCtoeT2pKWlYe3atRg3bhzCw8Nx8803lxqblZVV6jikp6cbb4iDgoJw66234ujRo6X+PrIsfH19MXz4cOTm5uKHH35wiSUlJaFJkybo0qULBg4ciMDAQMybN69c88ybNw916tTBzTffjK5du6JJkyaX9clrUlIShgwZAn9/fwwfPhz79u3Dhg0b3I7Ny8tz3h+HDh3C3Llz8cEHH+Cuu+5y+8TU3f2YmZlZ5jWeOXPGba5LLyUfb7r4zxpuuOEGzJ49u9SfOTz44IPYuXMnJkyYgMcffxzAhSeaY8aMQWxsLF599VUMHToU//3vf9G3b18UFhY6rztjxgyMHTsWcXFxePnll3Httddi8ODB5X4hhS4oy7noSXx8PK677jqsW7fOea6NGDEC58+fx8cff+wy9vTp0/jmm29w6623onbt2i6xi8/1iy8FBQXiGtzVw4YNGwIAPvzwQ+NalpSUhF69eiE6OhrDhg1DVlYWFi9ebHTd8+fP48iRI2X6vWDF8a9I+fn5WLhwIYYPHw7gwse1V6xY4fHv9HJzc53322+//YZp06Zh6dKlGDlyZKmxRUVFbu/vnJycMq/TpFalp6eX+m6UiRMnolevXpg6dSqefPJJxMfHY/PmzaXW2a9fP4SHh+Pf//43rrvuOrz66qt46623jNZ2xx13IC8vD5MnT8aNN96IKVOmuPypF1FVd/fddwMAli1bVipW3ufRZeXNc0AfHx/4+PxvO2Typ6ElFixYgJ9++gkvv/xymdarlEJaWhoiIiKMr9OrVy+EhIQgMDAQgwYNwr59+8o0J1WASnwX3EV5P879yiuvqPPnz6tmzZqp9u3bq+LiYo/5Sj56dfLkSXXy5Em1f/9+9eKLLyqbzabatWtXai3uLs2bNy+1hokTJ6qTJ0+q1NRU9f3336suXbpoPxJbER/nfu+999TJkyfVsWPH1NKlS1XTpk2VzWZTP//8c6n87m5X/fr11aZNm1zGluT2dDl+/LhzrKePc5d4/fXXFQD1+eefexyjlPwxvkWLFikA6s0333T+rKCgQIWHh6snn3zS+bO77rpLtW/f3m0O6ePcbdu2dfmIzRNPPKEiIiJUYWGhdu06MPy45caNGxUA9e233yqlLnxEKS4uTj388MNuc7q7DB48WOXl5bmM9XS/A3D5WK3px7kbNmyoPTdKLs8884x4HEo+qtS9e3d1/vx5589PnDih/P39Vd++fVVRUZHz51OnTnWe80oplZ+fr8LDw1WXLl1c7qNZs2YpAC51g8yV9VzUnd8PP/ywAqC2bt2qlLrw5xkxMTEqMTHRZdzMmTMVAPXNN9+Uyu/pMm/ePOc4dzX+3//+t7LZbKpNmzbO3w9KXfhYefPmzRUA1bBhQzVq1Cj17rvvqrS0NLe3IS0tTdWqVUu9/fbbzp9dffXV6pZbbik1tmHDhqpv377OdWzbtk3dfffdbo+Tp8eclce/LMryce5PPvlEAVD79u1TSimVmZmpAgIC1Ouvv+42p7vLAw884HK/KHXh94Cn8ffff79znOnHHE1q1aW/E9q3b6/9nabU/+rqpEmTXH7esWNH1alTp1JruLgeljzPGDRokMu4kj9bKnmsEFU2k8eZw+FQHTt2dP67rM+jvfk4d1mfA15s5cqVKiEhQU2aNMn5ce7p06cbfZw7NzdXxcfHq/HjxztzmTx/Ukqp2bNnKwDq3XffFcd+9NFHatSoUeqDDz5QixYtUk899ZQKDAxUERERKiUlRbw+VZyq9xmqcih5N3rkyJH47LPPcOutt3ocm5OTg3r16rn87Oqrry71rhgALFy4sNRH+urUqVNq3DPPPINnnnnG+e+goCC8+uqruO2228p6U8rtT3/6k8u/69Wrh9mzZ7v9QrCAgADnuyfFxcVITk7Ga6+9hhtvvBFr1qwp9UU9EyZMcPsxlbCwMOP1lXyxhO4becub5+uvv8apU6ec74YAF94RGThwIHbs2IHWrVsb5//111+xbds2ly+0GT58OF588UV88803uOmmm7xavyQpKQlRUVHo1asXADi/8XbOnDl49dVXS30J2C233OL8Vt7c3FysW7cOr7/+Ou666y588sknLq+oXny/X6wsr4RevM5z586J4xo3bmyc895773W5fd999x0KCgrwyCOPuLxKfO+99+KJJ57AV199hdGjR2Pjxo04deoUJk+e7PLu+4gRIzBu3Djj+clVWc9FnUsft76+vhg2bBhef/11JCcnO7sTzJ07F1FRUejdu3epHBef6xe79Itf3NX47t2744MPPnB5PNSuXRvr16/HCy+8gI8//hizZs3CrFmz4OPjgwcffBD//ve/YbfbnePnz58PHx8fDB061Pmz4cOH429/+xvOnDlT6h3mZcuWlVrH6NGjS32BlidWHv+KkpSUhM6dOzu/7Cw4OBg33XQTkpKS8Mgjj5Qaf9999+H2228HAGRmZmLFihWYMWMG7HY7Xn/9dZexjRo1wttvv10qx8V/6mFK6uxQ4uLfFaGhodixYwf27duHZs2aaa938TfEAxf+nMrdcwp3xowZ4/Lvhx56CNOnT8eSJUvcflETUVUUFBTk9vmd6fNob3jzHDAhIQFfffUVWrZs6fzTsgceeAC33HKLszuEJy+99BIKCwvxxBNPlGm9u3fvxpgxY5CYmOj2UziXuuOOO3DHHXc4/z148GD069cPPXr0wAsvvICZM2eWaX6yTrXaROs+YjFixAg899xzmDRpEgYPHuxx3MUbCbvdjoSEBI+/lHv06GG0wSh5YpCXl4cVK1ZgypQpFfo3C+6OQ8lGNzs7G4sWLXI+4XPH19e31DfK3njjjWjWrBnGjx9fqmVY27ZtPX4Drans7GwA8Pitrd7kmTNnDhISEmC327F//34AF75pOzAwEElJSXjxxReN88+ZMwd16tRB48aNnbkCAgLQqFEjJCUlVegmuqioCPPnz0evXr1w8OBB58+7deuGV199FcuXL0ffvn1drhMXF+dy3wwaNAjh4eH4+9//ji+//BIDBw50xtzd7+Xl7u9IvXXpt04eOnQIANC8eXOXn/v7+6Nx48bOeMn/X/qtxbVq1XJuzqhsynMu6rh73I4YMQKvv/465s6diyeeeAJHjhzB999/j7/+9a9uN4iXnuueXFzjjxw5gpdffhknTpwo9fFw4MI3Pb/88st4+eWXcejQISxfvhz//ve/MXXqVDgcDpdv5Z8zZw66du2KU6dOOZ9cdezYEQUFBViwYEGpj+B269YNzz//PIqKirB9+3Y8//zzOHPmTKm/jXPH6uNfEc6ePYslS5Zg7NixzloJXKgNCxcuxN69e0u9INusWTOX+7Ckq8Ybb7yBP/3pTy4viNSpU8eyelWePJMmTcItt9yCK664Am3atEH//v1x9913l9rYBgQElHqxpG7dujhz5ozRPJdu0Js0aQIfHx/LvqOF6HLIzs5GZGRkqZ+bPo/2hjfPAUv+rOdSsbGxbrszlEhOTsYrr7yCadOmlam7TmpqKm666SY4HA588skn5X4xtHv37ujWrRu+++67cl2frFFlNtElvWo9vbuVm5vr0s/2UiXvRo8aNQqff/65dpxVv5hLXPzE4Oabb4avry8ef/xx9OrVC507dy5TroCAAOTn50MpVWqzrJRCXl6e2+Nw8UZ38ODByM3Nxb333ovu3bujQYMG4rxxcXFo3rw51qxZU6b1mtq+fTuA0hsdb/NkZmZi8eLFyMvLc/tuwdy5c/HCCy8Y/Y2L+v9WLTk5OWjVqlWp+IkTJ5Cdne11OzJPVqxYgePHj2P+/Pml2q0BF971MXniXPIu3po1a1w20VY6efKk0QtFQUFBxsfL3SaHKodV52KJ7du3w9fX1+WFkk6dOqFFixaYN28ennjiCcybNw9KKYwYMcKrtV9a4/v164cWLVrg/vvvxxdffOHxeg0bNsSf/vQn3HrrrWjcuLFLa7uL/xbZXZ1JSkoqtYmOiIhwrqNkDTfffDPefPNNZ7svT6w+/hVhwYIFyM/Px6uvvopXX321VDwpKQkTJ04U8/Tu3RtTp07FmjVrjNvJlJVpL1WHw+GsQz169MCBAwfw+eefY9myZXjnnXfw+uuvY+bMmS6tc6z+REBZ/h6TqCo4cuQIMjIyKqX9npXPAUeNGmU874QJE1C/fn307NnT+YJXSZ05efIkkpOTER8f7/JmVkZGBgYMGICzZ8/i+++/127STTRo0AB79uzxKgd5p8psokteDdqzZ0+pTV9ubi4OHz4sPmn4wx/+gOeffx4TJ0607Jtdy+PJJ5/E22+/jaeeegpLly4t03UbNmzo7N18aUHav38/ioqKPL5ydrGXXnoJixYtKtNHPc6fP+98x8hKJe+ON2jQoFQvvrIoKirC3LlzERgYiO7duwOA89txZ8yYUerVzj179uCpp57Cjz/+6Byvs3r1ahw5cgSTJk0qtc4zZ87gvvvuw2effYY//OEP5b4NOklJSYiMjMS0adNKxT799FMsWrQIM2fOFDeb58+fB4AKuS9LdOnSxfkOsM4zzzzj8o20ZXFxTbj4Y+EFBQU4ePCgc4NSMm7//v3Oj74CF45DcnIyPxJZDladi8CFb7tfvXo1EhMTS30SZcSIEXj66afx66+/Yu7cuWjWrJnbP0HxRkxMDMaNG4eJEydi3bp1uOqqq7Tj69atiyZNmjhfsAMuHA8/Pz/Mnj271Kbphx9+wJQpU5CSkoL4+HiPeW+66SZcd911ePHFF3H//fdrP9Jo5fGvKElJSWjTpo3LnzKV+O9//4u5c+cabaIvR72KiYkxGvf++++7PJEOCwvD6NGjMXr0aGRnZ6NHjx549tlnjfrPmtq3b5/Li0v79+9HcXExP0VD1UbJny7069fvss9t5XPAskhJScH+/fvd/snagw8+CODC88aSPs55eXkYOHAg9u7di++++87tGzVl9dtvv5X6FAxdXlVmE927d2/4+/tjxowZuP76611evXnrrbdw/vx5DBgwQJvj4nejK1NoaCjuv/9+vPzyy9iyZQs6dOhgfN0BAwbgiSeewNSpU/HGG2+4xEqeUEnHAbjwUZahQ4di1qxZePbZZ50trDzZu3cv9uzZg06dOhmv1cS5c+dw99134/Tp03jxxRfL/Sp7UVER/vrXv2LXrl14/PHHnX9jM2fOHDRu3LjU36QBF7459qWXXkJSUpJRAS35KPc//vEPt+/2v/LKK0hKSqqQTfS5c+fw6aef4vbbb3f7t/SxsbGYN28evvjiC7HtQslHWdu3b2/5OktUxN9EX6pPnz7w9/fHlClT0L9/f+e58+677yIjI8P50frOnTsjPDwcb7/9NkaPHu38u+ikpCTjj1TS/1h5Lp4+fRrDhw9HUVERnnzyyVLxkk30hAkTsGXLlnK/4CJ56KGH8Morr+Cll17CZ599BgDYunUr6tevX+qJ16FDh7Bz506XPyNISkrCtdde6/b2JiYmYsqUKZg3bx7++c9/atfxz3/+EzfeeCPefvttt38zDFh7/CvK4cOHsWbNGkycONHtGgsKCjBixAisX78e3bp10+a6HPWqPH8TferUKZeOHUFBQWjatCkOHz5s6dqmTZvm8gbBf/7zHwBmv+eJKtuKFSvw3HPPISEhwetPEZWHlc8By+L5559Henq6y8+2b9+Op59+Go899hgSExOdL5QWFRXhzjvvxNq1a/H5558jMTHRY97jx48jIyMDTZo0gZ+fH4AL72xfullesmQJNm3ahL/+9a+W3i4qmyqziY6MjMSECRPw1FNPoUePHhg0aBACAwPx008/Yd68eejbt6/RR1NL/jZ6y5YtXq/pk08+cftR1BtuuAFRUVHa6z788MN444038NJLL7n9OJ4nHTp0wD333IM333wT+/btww033ADgwpOAJUuW4J577jF+svGPf/wDH3/8sXMdJc6fP+/s51fyxWIzZ85EcXGx23cVvv/+e7d9K9u1a+fyLt/Ro0edebOzs7Fz504sWLAAqamp+Nvf/ob777/faN0ZGRnOPLm5udi/fz8+/fRTHDhwAMOGDcNzzz0HADh27BhWrlzpsYjY7Xb069cPCxYswJQpU5wFyZ2SVi033HCDxz8bGDRoEN58802cOHECkZGRWLVqFXr16mX8buvGjRtd/sayRM+ePXH06FFkZWV5/ATFVVddhXr16iEpKcnlifPevXtdjtW6devwwQcfoGnTps62EyUuvt8vdeutt7q8M7Zw4ULs3r271LiRI0eiQYMGFfI30ZeqV68exo8fj4kTJ6J///4YNGgQ9uzZg+nTp6NLly7OFzP8/f3x7LPP4qGHHsL111+PO+64A8nJyZg1axaaNGnCj0eW0RdffOHVuaiUQmZmJrZu3YoFCxYgOzsbr732Gvr3718qV0JCAq6++mrnn+DonoRdfK5fLCoqylknPQkPD8fo0aMxffp07Nq1Cy1btsS3336LZ555BoMGDcJVV12FoKAg/Pbbb3jvvfeQn5/vfEyvX78e+/fvd/ulZgBQv359XHnllUhKShI30QMGDECbNm3w2muvYcyYMW5rUnmPv66+dO/eHbNmzcLo0aNLvdvqyfLly93W/cGDB+Orr76CUsrjGm+88UbUqlULSUlJLpvozZs3O+/DrKwsLF++HAsXLsTVV19d6pNmF/8euNSlL2S+9957bj/19fDDDyM4OLhcf8LVqlUr9OzZE506dUJYWBg2btyITz75xON5UF4HDx7EoEGD0L9/f6xduxZz5szBXXfdVaEvKhCVx9dff43du3fj/PnzSEtLw4oVK/Dtt9+iYcOG+OKLL9w+dzJ9Hq2rN23atHG7HiufA5aVu015ybvOXbp0cflupr/97W/44osvMHDgQJw+fbpUXbu4no0fPx4ffPABDh486Pw0ytVXX42OHTuic+fOcDgc2Lx5M9577z00aNCgzF9qRharzK8Gd2fOnDnqqquuUnXq1FF2u121aNFCTZw4sVTLHt3X4pd8HT88tLiS6L6aHxd9xb701fyjRo1Svr6+av/+/S4/17W4UkqpoqIi9eabb6r27durgIAAFRAQoNq3b6+mTJni0upHKfkr9Xv27KlCQkLU2bNnncfg0tsTEhKievfurb777ju3uT1dLm7XcXHLI5vNpkJCQlTr1q3Vvffeq9avX+/xtl7q0tYmQUFBqlmzZuoPf/iDWrZsmcvYV199VQFQy5cv95ivpM3Rxa213LW4WrhwodhuYNWqVQoXtddavHixAqBmzpwp3i7dcXzuuefUwIEDVUBAgMrJyfGYY9SoUcrPz0+lp6e7zenr66vi4uLUfffdV6pNj67FFQB18OBBpZR8n3///ffibZWOg6cWV57aZ0ydOlW1aNFC+fn5qaioKPXAAw+oM2fOlBo3ZcoU1bBhQ2W321XXrl3Vjz/+qDp16qT69+/v1Zp/b7w9F318fFRoaKjq2LGjevjhh9WOHTu0802bNk0BUF27dvU4RndOXtzCTFfjDxw4oHx9fZ2197ffflMTJkxQV111lYqMjFS1atVS9erVUzfddJNasWKF83oPPfSQAqAOHDjgcX3PPvusS1siXcu/kppUUn8ureFW1IJL64tSSv3nP/9RANTSpUs95lVK344KgJo9e7Zq27atio+P1+bp2bOnioyMVIWFhW5z1qpVSzVu3Fj94x//UFlZWS7X1bW4uvhpy8W/691dDh8+rF2jzvPPP6+6du2qQkNDVe3atVWLFi3UCy+8oAoKCpxjPJ1vJc8hLnbp78ySMTt37lS33XabCg4OVnXr1lVjx45V586dK/e6iax26ePM399fRUdHqxtuuEG9+eabKjMzs9R1yvo8WldvlHLf4qq8zwEriqfn46b1TKn/PVcreU6mlFJPPvmk6tChg3I4HMrPz0/Fx8erBx54QKWmplb4bSI9m1JKgYjK7LHHHsO8efOwf/9+l1Y4VDUUFxejXr16GDJkiNtWOUS/JyWf0Pj5558reykE4Nlnn8XEiRNx8uTJCv/2YiIisl6V+Tg3UXWzcuVKPP3009xAVwF5eXmw2+0uH93+8MMPcfr0afTs2bPyFkZUBSilsGrVKo8fjyYiIqKy4SaaqJxKWt5Q5Vu3bh3GjRuH22+/HeHh4di8eTPeffddtGnTBrfffntlL4+oUtlsNpw4caKyl0FERFRjcBNNRNVeo0aN0KBBA0yZMgWnT59GWFgY/vjHP+Kll16Cv79/ZS+PiIiIiGoQ/k00ERERERERkSEfeQgRERERERERAdxEExERERERERmrcn8TXVxcjGPHjiE4ONjlm3aJiEwppZCVlYXY2Fj4+NSs1wpZI4nIG6yPRETulak+VlQD6qlTp6qGDRsqu92uunbtqtavX290vcOHD2ubkvPCCy+8mF4OHz5cUSXOK+Wtj0qxRvLCCy/WXFgfeeGFF17cX0zqY4W8E/3RRx/h0UcfxcyZM9GtWze88cYb6NevH/bs2YPIyEjtdYODg80miRBeHUgvNlxtBTP5YuA6QjzAIMdxgzGXg8ntLbBgnsbykGa3JGjj9tq1xRyF+ee08cCgEDHHwR3Z2vjZtQfEHCjUh4dNHCymKD6TK475+Ill+gFtxBSw19GXlfz15+Uk0jkfK/TmLlZAcoF5PbmMvKmPQBlqZJSfPp4mnFSXi/wwlOuKySE5YjDmcvA1GFNkwTzN5CGNB8Rr48EOh5gjJ19fVxxB8h28d7v+l0LWur1iDuTow3dMHiqm8DWokfP++bV+QEsxBQLr6t8hzf1JyUmk5w2xmjpcpIDfimp0fTyMwwiB5vdzhj6PA/K5L8mQJjFgso4Mh34eR4b3twUQD5klWRwGNQcZ+hwZFtx3F9YiLUM+ItJKTNbqEI6Zyf0iHleD22J0/4v3n0EW6Xw1uXuF26N9XGUCaGD2XKtCNtGvvfYa7r33XowePRoAMHPmTHz11Vd477338Pjjj2uva/zxG59q8jEdk2VKn6aqTp+2ulx3i8ETUV+7/sD5BshJioWDb5LD5ieMMbl/hTH+tYUNE4Dic/IYkcFxt9Wy4CSQUviazVEVP87nTX0Efqc1UhrDGlnaZaqRvsJEtUxqpH/F10g/gxpZq8rUSINNtAU1sibXx5D//59H8uvfXtPOfznnsWgZ1qSxIEuIPsflOepAiLAOoxwWjLLk9hrclst2/1sxkQX3jUl9tPypR0FBATZt2oQ+ffr8bxIfH/Tp0wdr164tNT4/Px+ZmZkuFyKimqis9RFgjSSi3wfWRyKqTizfRKenp6OoqAhRUVEuP4+KikJqamqp8ZMnT4bD4XBeGjRoYPWSiIiqhLLWR4A1koh+H1gfiag6qfQPwY0fPx4ZGRnOy+HDhyt7SUREVQZrJBGRe6yPRFRZLP+b6IiICPj6+iItLc3l52lpaYiOji413m63w24XviSIiKgGKGt9BFgjiej3gfWRiKoTy9+J9vf3R6dOnbB8+XLnz4qLi7F8+XIkJiZaPR0RUbXB+khE5B7rIxFVJxXy7dyPPvooRo4cic6dO6Nr16544403kJOT4/y2RSN1od3iB3dupb161rJt8hwGnXZEHYV4ugVzBF6unigGOgtxk2XsE+JylxEEXCl/9XxWXr42fjpNHweA/Cz91+TXjTO5b4RXyc8YpBC+0r99syZiiuMp0oE3sFUekif14zLRRLjBJ4T7rop0uHPHkvoIXDgnNF8eGSrUyLPLtshzyA8RmVQzLKmRl6u3noFuQrzI4BuR9wnfzmxQI+tcGSqOySrW38HZB9O0cUCukefjYsQcVaVGnkrZYTCRYLs8JMeKAtUkSB8/orlvfwf1UeqCI3//ucE3pFvwVftyBnkdNqXPokxyGN0WfR6zHNIUJsddz+yWGMwjHFdr7j0TFnybvyULMfjG/8uxDKPmBcJKLGpMUCGb6DvvvBMnT57EhAkTkJqaig4dOmDp0qWlviyCiOj3hvWRiMg91kciqi4qZBMNAGPHjsXYsWMrKj0RUbXF+khE5B7rIxFVB5X+7dxERERERERE1QU30URERERERESGuIkmIiIiIiIiMsRNNBEREREREZEhbqKJiIiIiIiIDFXYt3N7rWk9oJbnPX5QmL4LYFaTMHmOPaf18QZyCiQIbRcy5D6bCPTTx/0CDRai79VpGak/aZFwWwAg2/s+woW5ckPqU3v1DWjzfsiTJxKaTdrrRYopAkP1vaTPZMvLgDBNRJjQMxRAfrrQi9UqwuMmsEm8mCIXwnm0/UAZFlRDNQ0DfD3XSP+g2vrrN5A6qQLYL9SVRnIKJETo4xkGjaI1vwsAAIHy+Q8I9d4qUo08Z/A4zDCoTYKcs3KNLMjSH/vCHwxqtXAa2erp6x8ABIbqz9VMk19vofqwSY1Uxy5TjWykPyb2BtFiinx/oUZuTS7Dgmoi/Ylps6hPrHYOo6a4FixEaolrUctrOY8FXYCNmjxb0J/b+7bYhvecNZ2idYz6c1vRwPkyPGYuTCMdeIOFVPxhB8B3oomIiIiIiIiMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRoVqVvQBPEq7sDB+7n8d48q5t2us74mPEOTL2ndbnuC5CzqFPAfh5vg3/myhKHy8slHMgw2CMBXZKA0zW6r1ou3zfJMTFa+Nrs7eKOYp89K8zBQQGijkKc7PEMaJg/Tx2vwAxRWSYfMwkvp3lx9W1fbpp42ccvmKOrYuWGa/p96rJlZ3h6++5vuwVamRYU/m+PL1fX1eCrqsr5sg+7a8f4GfwWm5IA33cdnnqjhH9YQeQdzlWgfhg+fFeX6yRG+SJfPS1JyBQfppRmHtKnkcSbNeHDc4zW7T3NRId48Qh196or5E5jiIxx+ZFq0xX9LuUASBEE7cJ11dKnsMmJRFnAQBpIitymDC5wV4PMDqu3jI6YkY314r7Rp6lKlAm6zC574Q0Vtxas3NIGKQ8ryQzE3A4zNbCd6KJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyVGX7RIcG2uGr6RPtE6jvCdnt2p7iHHPO7NDGfYTevACAvWf0cb8wOUeh0Eu6WO4ZWa1IL900klMcLc4Xx6TmpmvjRTFyj1tk5GrDJ/YeF1OEOIT71+BRGCj0PQ/zixRzNOkm9wVudE8HYSH6xx0AFEcWa+N+AXLvdN+oIG280XB9f9viwiIc/GS3OE91Vsehr5EtQoO11+/Y7Vpxjnlp+mMYYJdrZHZ6jn6AMqmRQs/H4vNyjupEeojEyilS8uT+9Cln9TUSJjXypL5GnjWqkXX0Awyai9rFGin3b27QraE8ZvQa/YBA/eMOAGoJNVL5yO9v+As1MvYP0R5jxQVFSPl4nzhHdeYQGkUroY+sTW4CLbfNNThvbRa0IpZ6/BrcFCtaAJsRk8grsVnQjFjMYbIWg4OmpINvRc9jgzvYgrPZ8P73/iwRHxIm57N0yIxXo8d3oomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRoVqVvQBPDu0/AB8/X49xn8JC7fWXfrlYnCMgKFAbP7MlRcyBbIc+7hck59h9QB+P0a/zspJuTguDHI0i9HE/PzlHdrE4pCjljDDA4DWk9enacF6mnCJkZIx+QHc5B4r04bpBwjEF0Nyvvjgmrm0TbTwtL0vMkQ39Y7NVfHMxR6sBdbXx5JTD2vj5/EIcxG5xnurs+P7D2hp5Pkd/P6xY9o04R52wYG08fctRMQfOCUXDN1TO8VuyPh5ul3NcLsKvBDQ2yHFFlD6ulJyjQK6RSEnTx30D5Bwb9XU2RyjDABDwJ6HmXy/nsKJGJvjUE8dcjhrZIv4KMUeHAfpzZO9+z88rzuefRwr2iXNUZxnIQAhCPMZtNv31lcljTCJNcmGQ9/MIS1UGU5isQgkT2UyyCMfVZnTMpPvGZB0GQ4S12MyS6OMm942Uw+RUteA0syaJFfevBSw5HhXwTvSzzz4Lm83mcmnRwmRnRURUs7E+EhG5x/pIRNVJhbwT3bp1a3z33Xf/m6RWlX3Dm4josmJ9JCJyj/WRiKqLCqlOtWrVQnR0dEWkJiKq1lgfiYjcY30kouqiQr5YbN++fYiNjUXjxo0xYsQIpKR4/tvi/Px8ZGZmulyIiGqqstRHgDWSiH4/WB+JqLqwfBPdrVs3zJo1C0uXLsWMGTNw8OBBXHvttcjKcv9lG5MnT4bD4XBeGjRoYPWSiIiqhLLWR4A1koh+H1gfiag6sXwTPWDAANx+++1o164d+vXrhyVLluDs2bP4+OOP3Y4fP348MjIynJfDh/XfuktEVF2VtT4CrJFE9PvA+khE1UmFf2NDaGgorrjiCuzfv99t3G63w26vQu1JiIguE6k+AqyRRPT7xPpIRFVZhW+is7OzceDAAdx9991lut7pY6mAr+aNcj99n0X8eFqeRN+KFpDb6gL1hR7ABj0yIf0JT26uQRILtJGHRIzQ3167Xe7x7OurH5PyhdA3GwCWy0MknccMFcds/Hah1/PENdL3/PSzy/1cz5zO18ZjwhqJOYIRL46pXz9cG48slj8q55Onvz0dW3QRc7w59TltPGXpNjFHdVDe+ggAJ5NTvKuR6zLkSaTWunJbXfg2bKiNF50pkJNIdTRX//iwTEt5SNToSG3cZqsj5vCtra+RRz/ZKy9klTxE0u2vd4hj1i/1/C6hqcaN9L98/Gy/ijnEGumIFXOEoqk4pn6C/ndgZJ4FNbKdXCNnvKWvkQe/2inmqOq8qY8Oh9SwXc+KdsUm3W7F3ssGTZ6lESY9r6WeyCbMuvuKq/V6HSZMemeLOQzGWHBY5XPEiklMcpj0ThdbeBucz2ITd3kZljWCFlj+ce6///3vWL16NZKTk/HTTz/h1ltvha+vL4YPH271VERE1QrrIxGRe6yPRFSdWP5O9JEjRzB8+HCcOnUK9erVQ/fu3bFu3TrUq2fwlgURUQ3G+khE5B7rIxFVJ5ZvoufPn291SiKiGoH1kYjIPdZHIqpOKqRPNBEREREREVFNxE00ERERERERkSFuoomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJDln87t2VOndRv8QstmKNXmDZct7CumCI/44w27udjF3NkSAPOiynMCC+ZXPlkWzHF6WL9an395GUkrzygH7BczmGFfJyWB9X11cfPFMkp/BzaeJb+NAQAJPSJ1MbXY7uYYwgSxDH1o2K08cBC/W0BAOTqH5z5WfJJkrJ0mzzP792p0xVfI6/R398RNrlGZp1M08YdPsFiDvGRmi+mMCOcml2ebCWmSC0+p42b1ci9+gGr5BxWyEeqPChEiGfKKeqo2tp4qEGNbH6jvkZ+b1Ajb0dzcUxC3XBt3JIaeUo+SQ5+tVOe53csA/pT0wYlZLCJcyhpiDQFAJtNn0QpkyTer8OETbjB4vEAYLNZsBjhmFl2e8VpDG6wwGSp4jliwUTSbb0wyGQa74+JuAwLptAdjkwABlUcAN+JJiIiIiIiIjLGTTQRERERERGRIW6iiYiIiIiIiAxxE01ERERERERkiJtoIiIiIiIiIkPcRBMREREREREZ4iaaiIiIiIiIyFDV7RMdFwXU0uzxNx/XX9/glt0w7DZtvFVkazHHko8+1cb3fbxazBHcUd8EM+sXg37GBureE6GN+wXLvSn90gO0cd9c+XWZohniEEu0naTve73ty5VyEqEN7h+GyueIX5MgbTyyRXsxR3CM/r5Jz90t5tgYqO8BDQA+Qnc8u5/QNxtAvjAmOVnfN5gMxcXqa+T2I/rrG/RavP6P+hrZIbKjmGPJ7Pna+O6PfxBzhLarp42f/fWkmMNExB/1D3i/4DpiDr90fdPq2uf1NRQA8JY8xAqtntH3vd6ybI2cpK7+RBo+VO69XKel/ndgXGODGtlQf9+kZ8k18pfgjeIYH+jPRXsduUZm2fRjjhxjjfSWQ+oU7X2baEgtj5VBB1+pr67UI9hkHrMcMrkPtFHXY30Gg77YVrSJtqKbsVkO4b4x6UcuHROjxskW9EU3OrDig0JOYcW9U/HtqgHwnWgiIiIiIiIiY9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyxE00ERERERERkSFuoomIiIiIiIgMcRNNREREREREZKhWZS/AE3vrhrDZPS8vb+dxr+eIj2iijUc1iRFz3HbfaG188pc/iDmyfjmtjQdHiilgj/QVx/Tv3ksbLyr0E3NE+2Vr459P/ULMYYW2k64UxzRqr7//0tLSxRxNEsK08bqJbcUcfsEObTwbuWKOtXM3auP9RiaKOU7Y08QxgX7F+gF++tsCAFmF+nNx84aVYg6SBbZpCJu/5xqZs/+IkECeIz6ipTZe16BGDhJq5G6DGnn215PaeGi4mAL2OHlM/+ulGllHzCHVyE+nLZIXYoFWz7QXxzTq0lAbP3MmU8zRQKqRnZqLOeqECzWyQK6Rq2fpa2SvkZ3EHFbUyIJaBjXST6qR34g5SC8DQIgmbrNkFqWNWjGHfgazeZRJFmXBag1SyGuRkyghhc3k9tpMFqvPY3TfeH9zDcgrMbm5chILUpjkkO5gg4UYnEWaWCYAuY4DfCeaiIiIiIiIyBg30URERERERESGuIkmIiIiIiIiMsRNNBEREREREZEhbqKJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERmqVdkL8CQsMAg+ds/LO1ooJMiW5yjM0yfx8ZNfY4iKi9DGe44dLeZY9c93tPGsE2IKBA2tL445HVysjQfvlQ4q0L5Yf3uTESfm2Ioj2njk8BgxR1D9MHHM0eNHtfHOg1qIORzN9Lf3hMF51ibSro1/9NIPYo6jX6Vr4xH3BIg5Cn31OQCgsEgfLw6MF3Nk5+3Tz5F9XMxBsiBHHfjY/TzGc3KEBFIcgE2skefFHA0ax2rj1xrUyO+feF8bP3tKTIHI4QnimNPB+l+J9u3yQWtfK0QbT4Zc3zZD/xiJuSNSzBFmUCNPCjWyXf/GYo56rfX374lsoahArpGzJ8k1MvWrNG089kH9/QIAhX7e18iiUIMambZDP0e2wQlNWo4MALq73Gbzeg4lxG2Q51BKn8VmsE4lziOtFLAZjLHm9opDZMI08vGwhtF9I92/BsfDmlujzyKt07KFmNz/0nE1WasScljw+AfK8U70mjVrMHDgQMTGxsJms+Gzzz5ziSulMGHCBMTExKB27dro06cP9u3TP6kmIqoJWB+JiNxjfSSimqTMm+icnBy0b98e06ZNcxt/+eWXMWXKFMycORPr169HnTp10K9fP+Tl5Xm9WCKiqoz1kYjIPdZHIqpJyvxx7gEDBmDAgAFuY0opvPHGG3jqqadwyy23AAA+/PBDREVF4bPPPsOwYcO8Wy0RURXG+khE5B7rIxHVJJZ+sdjBgweRmpqKPn36OH/mcDjQrVs3rF271u118vPzkZmZ6XIhIqppylMfAdZIIqr5WB+JqLqxdBOdmpoKAIiKinL5eVRUlDN2qcmTJ8PhcDgvDRo0sHJJRERVQnnqI8AaSUQ1H+sjEVU3ld7iavz48cjIyHBeDh8+XNlLIiKqMlgjiYjcY30kospi6SY6OjoaAJCW5tpqIi0tzRm7lN1uR0hIiMuFiKimKU99BFgjiajmY30kourG0j7RCQkJiI6OxvLly9GhQwcAQGZmJtavX48HHnigTLm62qPgZ/f3GP/kjDcrvSD7TIY2np+bL+bwC9P35+07crCY4/vPFmvjRZv1/S8BoG6vKHHM1twD2ni3ffI3YDbKu0IbbxUp98gcPmuMNt62V1sxx760veKYFN/t2viZPPm4hqzUnwOFYsNy4HCkvudrsK/c47n9k9dp4xHBjcQcW3N3imNSs/X9Whs45LXWDcjVxiP0U9RYVtZHAOjuH62tkR9Z8KeBp8+c1Mbzc+X+9EVCu+KbR94h5lj/2TfaeMHmY2KOutfUFcdsyN2ljV97yKRGttHGmxr0ER4+e5w2blIjd5jUyCKhRhZfnhqZFrpeGw/2lV/vj53YXRuPCNT3swaArVmXq0bq+6tH2ArEHDWN1fVRbBNdrlVemqPi+xEro8a6ekbrNBkiLMWsB7QFfbEtaDZtdM+J7bdN+m9LDPpzC8fE5LaIS7Wob7K4DoPG2GLPcit6vGsOSGZmJhwOh1GeMm+is7OzsX//fue/Dx48iC1btiAsLAzx8fF45JFH8Pzzz6NZs2ZISEjA008/jdjYWAwePLisUxERVSusj0RE7rE+ElFNUuZN9MaNG9GrVy/nvx999FEAwMiRIzFr1iw89thjyMnJwX333YezZ8+ie/fuWLp0KQIC5FdniYiqM9ZHIiL3WB+JqCYp8ya6Z8+e2rfBbTYbJk2ahEmTJnm1MCKi6ob1kYjIPdZHIqpJKv3buYmIiIiIiIiqC26iiYiIiIiIiAxxE01ERERERERkiJtoIiIiIiIiIkPcRBMREREREREZKvO3c18uR3aloJZfxS5v347ftPGivHwxR7cB3bVxv8gIMcfY1/+hjfump4g57PZsccyJzXu18URHXTFHjF143WWHvNYfP56rjQfYe4s5GvvFiGPs+eHa+JmsQjFHyM7T2rgjUF7HLyt3aOPD4xuJOU5Df9+kI13MERgYJI45cGSrfh1F8nnWvYX+mBRmy8edZEd+3V/hNfKQUCN98zx/026Jjr2rRo202fWPZQA4u11fIzvVNqiRtvPauP9xea0r5+hrpI/NoEbWsaBGnjKpkanauCMwSsyxY4f+vhkZ3UjMkVpgQY0MDhPHHDiyQRs/XVQg5ujaTqiRC1gjveZwaMNy5bKCPIvNVvGzKJN1QF6IlMXktihhHpP7xYpjZjSReINNJrLiDtYvRBkcEGmI0eNBWIfRREbzSHN4P4XNkpOI70QTERERERERGeMmmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRIW6iiYiIiIiIiAxxE01ERERERERkqMr2ifaJDoWP3c/zgEZCArklKPKyc7XxjPQzYo6dP+h7RnYdPEDM0ap9W208blugmOP00pXimN6OJtp4tt8JMUeGj75PcM/r9XMAQLbQnzlr6c9ijo1ZReKYQnuANt4oLl7MERfZTBsPFHpRAoCfX4g23qSu5jz/f5u37NPGMyL0cQAoaiv3440Rbo89Sl4rgvX9qM8UFss5BJGJCdp48flipG845PU8VVlRTCjgr7k/GvjqE2TLj6G800KNrJ0m5tj/w4/aeAeTGtlFqJGb5F9lqSvk8yHG0VIbz/Y7IubICND3Cb7akhr5g5jDpEba7Pra1CguWswRF9laGw901BZzSDWyZVQdMce6LTu08Yyl8v1f1Fy/DgAIF2pksBU10ub9U7Pw7k09xorPF+PMOn0P+OpO/s0s9N61pGGxBb2XDZYhLtasgbPIZsVqhbWa9e+1olGwPETur22yDu9zSMdEWdG/2aIe0NIIS/qzGz04vZCZKfaZL8F3oomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRoVqVvQBPwq+IhV9tf88DNukbYfs28RPncAQGauOBAbXFHDtX/qiNd+3RXcyRH+CrjeceOC3myN5xXBzzc16GNh5UL0zM0SQ+QhuPCBJTICahsTaefjxXzHHm+ClxTFFenjZ+OOWgmCMlMF0btwca3OCAGG3Y0b61mKL/FW218cJC/X0LAPPe+1Yc0+rqrtr46aAAMUdKmn4tdugfdyYyi/RzqGLl9RxVXVizhvCrbfc8QKiRCDGokaGXoUb2uV7MkeVTpI2b1Ui5Zvyc+Zs2blQjm+prZMtw4X4BEJMQp42nnzKokYcMamRxsTZ+OOWomCMlXaiRofIxK/AJ18ZrRzcTc0g10mZSI+d8JY7p0Fn/O/10mFwjU6UaWWDw1Ex4CyTz3EmPMVVU8+sjkAEgpNzXthmMUdJhNEkiziHfVzabfiKze9tglHh7DNYqZjBZrZDFghRmeUyOmbRW73NI97/xPGIOgyE2YZBJDiFucnulx43RMTPAd6KJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyVGX7RB8/lgZfu6aPaUa29vpF+fIcuYX6Xpv5eXXFHIExkdr48QPJYo6IK+L1czRJEHP8Zda74pj3n3teG3/4xdlijofvvE4bT2zdSMyRvOWwNt6shdwTtEeifh0AkHE0RRvfunubmCNd6OeJAH2fVQBYu+UHbTymkdw3tv9do/UD0uW+sUMKC8Uxn2/7WRsPbNJTzHH0uP6YbZz9tZhDUjdG3+O4uLAYaV7PUrUdP3ZEXyPP6WukSb/G3ByhRtbxvkYe2pUs5ohoE62N12rZXMwx5o4R4ph3nhZq5L9miTkevk3fR/i6jleIOZK36M/eZi0aijmuGXaDOCb74H5tfENduUZmCjWyIESukVt+1tedmEb1xBw39RNq5LGzYo4hOTnimM/3bNHGA5tcLeY4Kqzll9nLxBwQDquuRhYXFkPf3bsm0P9eNem/LLKm1ax+Cgv6GdsMO0V7TRksVmwUbdI3WbrBFvVNlvozyxnknscGOUw7fXu1DpPey9ZMZJLF63VY1QdaUuZ3otesWYOBAwciNjYWNpsNn332mUt81KhRsNlsLpf+/ftbtV4ioiqL9ZGIyD3WRyKqScq8ic7JyUH79u0xbdo0j2P69++P48ePOy/z5s3zapFERNUB6yMRkXusj0RUk5T549wDBgzAgAEDtGPsdjuio/UfvyMiqmlYH4mI3GN9JKKapEK+WGzVqlWIjIxE8+bN8cADD+DUqVMex+bn5yMzM9PlQkRUU5WlPgKskUT0+8H6SETVheWb6P79++PDDz/E8uXL8a9//QurV6/GgAEDUFRU5Hb85MmT4XA4nJcGDRpYvSQioiqhrPURYI0kot8H1kciqk4s/3buYcOGOf+7bdu2aNeuHZo0aYJVq1ahd+/epcaPHz8ejz76qPPfmZmZLIJEVCOVtT4CrJFE9PvA+khE1UmF94lu3LgxIiIisH+/+1YadrsdISEhLhciot8DqT4CrJFE9PvE+khEVVmFb6KPHDmCU6dOISYmpqKnIiKqVlgfiYjcY30koqqszB/nzs7OdnlV8ODBg9iyZQvCwsIQFhaGiRMnYujQoYiOjsaBAwfw2GOPoWnTpujXr1+Z5sk8eQI+/p6XFxwTr71+FrLEOQojCrXxyCb6OQAgJq6zNp6flyuvIy9PG1++b4uY48aYRuKY0eOe0sYnvThbzHH8tP72REY2E3NEOAK0cT8/+bWdtSt/EsfUDQvWxnsMHizmQKGwliD9bQGA/ilHtPE1K9eKOT55a4Y2ftuYJ8Qcra4YI47ZNmGjNr59x3ExRzaEY5IvphDdNuxObbwgNx//XfJf7ycqo8tVHwEgM+OstkaGhuu/4faswR1RGGvTxkMbNhVzNGx8tTaen3dWXkdesTZuUiMHhjcSx9zzT6FG/muWmONU9nltPDS0oZgjoqH+MRRYR647P3xjUiPt2nivO+4Qc1hRIwf1975GfjRVXyPvHCfXyOYt5RoZ9+RobXzvPv0XYQHA6QLhmOhPISO3/fFuj7GC3Hy888107ycpo8tZH5GRAejelbbpa5sNSpxCKf0YYYr/zyGNMEhisFYxgwXTmN1e6ZiZLMS7OS7M4/U0ZvNIOYwmMhkkrENYiNkxMzhowhgr7huTwyFNo50iMxNwOAxmKccmeuPGjejVq5fz3yV/izJy5EjMmDEDv/76Kz744AOcPXsWsbGx6Nu3L5577jnY7fpf1kRE1R3rIxGRe6yPRFSTlHkT3bNnT+0rCd98841XCyIiqq5YH4mI3GN9JKKapML/JpqIiIiIiIiopuAmmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRIW6iiYiIiIiIiAyVucXV5eKPYvii2GM8umG89vrXtNfHAeBM4Wlt3OHwE3NsXrZGG4+7sq2Y48BKfY4IR5CYY/KMd8Ux4x97Thtf//MKMceyxZ9p43kZ2WKO8IQE/QCDl3YaxQeKY/IL9WvZvOwHMYfDEamNF+bKtzcwUn//JfbtLObYt/WoNr77h2/FHHv2HRDH9Og+WBv/9ofPxBxF8UKT+huixBz4Nk0b/nHZZv0aCs7Lc1RzQUKNDBBrZFNxDrFG1rOgRrZpLuY4kPKdNh5bW66RLxvUyMee0tfIXzatFnN8ufAzbTzvXIGYI6a5vkYWnPcXczSKl3+9K5t+LRu+NKmRodq4zYIa2alnOzHHoV0ntfF9JjVyxy5xjFgjf14i5iiKrq0f0D9WzIGlx7ThDV96rpG/h/qYASDEi+t7bsR1EZsVSYQU0hwAbMIgkxxWLFbXvqyETTxmcg4lHHhxjguj5HmEtVg1j7dMZjC5b8QcBmOktZgdDeH+teK2aKbIBCA8i3XiO9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRIW6iiYiIiIiIiAxxE01ERERERERkiJtoIiIiIiIiIkPcRBMREREREREZqrJ9oqPr1UMtu+cepLn5vtrrt2rbUZzD74y+f2V6RrqY4+fvlmnj2UViCuRmJ2vjUT26ijm2BeeJY/718ova+D9fmCrm6J2u79+75KN5Yo6wKH3/2tMZGWKO7t2vFMegrr6XdEhyipgizFFXPyBP7oG6cae+p3F+mNxrN7J+hDZ+4qjcAzrr9HFxzD0vzNHGlxw/KOaIaRGsjTfroO+9DQD78rO08Y5XJ2rjBefysTVpnThPdRZdz6GtkWezhBrZSa6RKk3fJzozQ9+rFgB+/u5LbTy7SO6bnJutnyfqqg5ijk3hco188flntfEnXpoh5uh97LA2/t2ihWKOaKFGpmbIteu67leJY6QaGbB3v5iiXj2h73ue/hwCgF92btfGHdF1xByR9aQaKfeAzso+JY558JWPtfHFx/eIOSJbhGvjjQ1q5G9ZQo3s6blGFpzLx+b5P4pzVG/6Tq8WtCuGTWgUbEGbaKMkShhkM+nOa9QE2IrG2NIxM8khjDFojG0yj5zF+17Thl2evYgaTmPSTNyoh7e0DGvmEXlzqpahUTTfiSYiIiIiIiIyxE00ERERERERkSFuoomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhmpV9gI8qdc0AX617R7j+UWeYwDg5xcszrHqi0+18Yw4XzFHYaB+zMavlok5eg6+Wr+OE6fEHG16dxPHzBz3kjZe/4UoMccf/voHbbzriRQxhyMmXhtv1b2RmCPv+D5xTHHGaW08rG17MQd89K8z7Vv5o5giL8NPG3dkyOdZrrCOvMI8MUdUmNw9/pv1O/QD2l8h5ji+ea9+wO40Mccf356kjV9zZaI2fi4zB7P/OkWcpzqLaJoAf02NdFhQI5d9MV8bP9e4tpijsJb+/F5vUCOvFWrkOYMa2erazuKY94QamTBJrpHD//onbdyoRoZHa+PNu14l5ig8dUgck3cyVRuvZ1IjAwK04d3fHBRTZJ5U2rgjpkjMkeuvr7OZOcVijrpB8vn8zfpftXHf9k3EHCc2H9DHfz0m5hj99jPaeJcrr/MYO5eZg1ljXhfnqMmU/pSzahZ5iM3mfQroc5jcVJu0DqM8cg4xi8E6LGF2UCp8GWZTCPevyW0RBtlsRgfE22mMDru8DO/vF7PbK+M70URERERERESGuIkmIiIiIiIiMsRNNBEREREREZEhbqKJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERmqsn2iI+KiYa/juf9kcrq+L+6n788V5zi6Y7M2Xj+mi5ijUN+KFUhLF3MEBEdo458s+krMcWCb3Iuz97i7tPG7h00QcySf0fe3fOqRR8QceenZ+gEx+uMBAAFBgeKYjN279QMM7hsIPa39giPFFGGRhdp4pN3gtgh9oM9kCccUwJn8XHFMv9tv0sYPB8q9hff56Xu6trpaflwhrK42vC59qzZekCX3za7uQiOj4a+pkccy9cdg/vuzxDlO7tiujdczqJF5Uo08fEbMIdbIr+Qa2dGCGnnXsIlijuScw9r4+AcfFHPkpBfoBzSMFXP4hYWKY/KlGnkmU8yBGP08fuFyjYw+r+/7GWpBjczJl2tkWtZZcUyv2wdp44cDpRMeOCTVyM4GNTJa37N8XfoGj7HfQ310OPRxqZ+t1HsZAJTQ9daSNsMWtF42ac6rLlN7ZnEtBuuwCUlM+iYbHVYLmomL85jcN9J5ZrwazRwWtDQH5PvGqNe0FbdIfIBbc8KX6Z3oyZMno0uXLggODkZkZCQGDx6MPXv2uIzJy8vDmDFjEB4ejqCgIAwdOhRpaWmWLJaIqKpifSQi8ow1kohqkjJtolevXo0xY8Zg3bp1+Pbbb1FYWIi+ffsiJyfHOWbcuHFYvHgxFixYgNWrV+PYsWMYMmSI5QsnIqpKWB+JiDxjjSSimqRMH+deunSpy79nzZqFyMhIbNq0CT169EBGRgbeffddzJ07F9dffz0A4P3330fLli2xbt06XHXVVdatnIioCmF9JCLyjDWSiGoSr75YLCMjAwAQFhYGANi0aRMKCwvRp08f55gWLVogPj4ea9eudZsjPz8fmZmZLhciourOivoIsEYSUc3E55BEVJ2VexNdXFyMRx55BNdccw3atGkDAEhNTYW/vz9CQ0NdxkZFRSE1NdVtnsmTJ8PhcDgvDRo0KO+SiIiqBKvqI8AaSUQ1D59DElF1V+5N9JgxY7B9+3bMnz/fqwWMHz8eGRkZzsvhw/pvNiUiquqsqo8AayQR1Tx8DklE1V25WlyNHTsWX375JdasWYO4uDjnz6Ojo1FQUICzZ8+6vJKYlpaG6Ohot7nsdjvsdrklBBFRdWBlfQRYI4moZuFzSCKqCcr0TrRSCmPHjsWiRYuwYsUKJCQkuMQ7deoEPz8/LF++3PmzPXv2ICUlBYmJidasmIioCmJ9JCLyjDWSiGqSMr0TPWbMGMydOxeff/45goODnX+j4nA4ULt2bTgcDvz5z3/Go48+irCwMISEhOChhx5CYmJimb9VsW5gKAICa3uMf79hqccYAOx7Z6U4R//nb9LG7YEBYo7k/EJtvPCKeDFHelqGNt6+XVsxx8avvxPH9Bg5UB9/9nYxx7zPlmnjtw3WH1MAaNFdvw5kHRFzIL9YHOLo3lsbP75ssTzPgWRtOAT6+x8AcguztPGd2w6KOerGhGnjxYW5Yo6i3HxxjCPCVxtfu3eLmOOGofpz4OYH7hRzrNj5lTaekq5/zJzPLhDnsNrlrI/AhRpp19TI9cv1NfK3/64R57hh4iBt3B5qUCO327Tx4ivqizmsqJHrv5Z/J/S4V1+bej0rt9qZt+RLbdykRjZLFGrkWYMamZUnDgnqqq+RR1YZ1Mhde7XhEH/9/Q8AqTmntPGdm0xqZJA2XpxjUCPPK3GMI6JIGzepkddJNfLPd4s5fhBq5OE8z4+Z89ny74KKcLlrpFfk01YeIp9O8iCbwUKsoEwWK6zFaKnCIIN1KGkZVqzjwkz6DEaHzPvbaxNvkAX3nUkOK84RSzJ4f45ob0tmJuBwiHMAZdxEz5gxAwDQs2dPl5+///77GDVqFADg9ddfh4+PD4YOHYr8/Hz069cP06dPL8s0RETVDusjEZFnrJFEVJOUaROtDF6FCAgIwLRp0zBt2rRyL4qIqLphfSQi8ow1kohqEq/6RBMRERERERH9nnATTURERERERGSIm2giIiIiIiIiQ9xEExERERERERniJpqIiIiIiIjIUJm+nftyOrD3IPxr2z3GU/bu1ieIlufwEXrrhoXJvYjbto7QxjdvShNzHM04rY3fMsygr2jnVuKYj2Z/pI3/eZDcv9cvMEYbzyjS93MFABQJPU6Dg+UcuenymGz9mJgmLcQUm3ev18ZP7Nwh5mjS6Ar9OoL9xBx+Pvq+nq3iIsUcyXmHxTH1jx7XxluE6ftVA0Czq7tp45vObBRzpOQm6wcU6Xv4nT8nP3aruwPJB+GnqZH7pRpZT54jwKavkVFBJjUyRBvfulaukSeFGjlgiFwjG3buII6xpEaG6mvkyXMnxRzNINTIUM/3u9N5g17AQo2MM6iRW4UaeWTLJjFHS6FGhobXEXNINbJ2XKiYo5ZJjTypr5HtwuQnH80662vkNoMauT9X358bvp4f4EXnjBoYV28ZADSlx4ruy+KXjZv0mhYb2hqsQ5rD+5bIZkMsOK3knsjyt7wrkxts1J9ZSCHPIs5jdHu9HmAwyOSYWXAeKYPFSsfErFu1Fb21ZXwnmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjLETTQRERERERGRIW6iiYiIiIiIiAxxE01ERERERERkiJtoIiIiIiIiIkPcRBMREREREREZqlXZC/Ak7WgyagX4eYzHhAVrr//6gtfEOf7Q/XZtfFXhR2KOZQfs2niaPU3MceL4UW08OStXzNE5vr04ZmvhZm387anvijn+O/Yf2viV8U3EHLlbt2rjgW3bijmQe0YecjxDP8A3UMzhV6iP16+fIOaIjojUxk8cPSDmiKzr0MYLfYSFAjgeJs8Tc1yf57breos5tgama+M79+rPQwAIiqqrjZ9Ly9fGiyEfj+ou9egxoUaGa6//+oJXxDlGXjdCG19ePFvMEbZLXyPP2FeKOY4c0tfIY/nnxRztoluKY3YVrtPGzWrkk9r4VbENxBz5Wzdo43ajGnlCHJJ1UqqR8lMEVVigjSckNBdzRIfX08bNaqT+OUGOT46Y45RBjWx4Sl9bYq67SsyxKzRTG9+yV38eAkBYgyhtPONwlsdYcYH+PqsZMgCEeHF9m/cjlJwDNqVPoQ9fSGHTz6MMkkg5LiSSh3ibRBkcd/HIG63TZB59IiuOmcl9A2kek3VIUxiMseLutxnMJB0Tk+NuRQ4TfCeaiIiIiIiIyBA30URERERERESGuIkmIiIiIiIiMsRNNBEREREREZEhbqKJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ1W2T3TLxJaw1wnwGO9x9x+11x8SP8DrNfT0+7M4JqSFPh74pK+YY+tP+j7QhYXJYo4li9eIY5rHXKGNB3fpLub4Oe2gNt6/fn0xhz23WD8g/bSYAxnymDxEaOMOu9wn2uHruQ8vAJzIlXt4B3fTH9eA9WIKpBxN1sb3pOnjAJBeLL9mdmK3/vZsPLFMzFF0ZaI27puRLeYIi4vXxrPCPPdABYDz/vo+0jVBq8SW8NfVyDuHaa9/e8LNXq+ht88D4piQ1v7aeOA/isQcu6QaWZAs5liy+DtxTPOYNtp4cBd9z3cA+OHwHm3cqEaeFWrkMWtqZLF/rDZeOzBUzFGvSKiRZ8+KOQJv1NfI+F/EFEg5ekwb33k8WcyRXuz58VTiyO5T2rhRjWxznTZuVCMbN9XGVZjnPqnn/X8PfaIFYm9lgxRCz1tl0FlX6lZr1BPX20mMed8p2Iqe1tJ9Z3RzDaaRlmIzSKIs6kesn8SCQUbr9P58NhpxGQ6ZVfhONBEREREREZEhbqKJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyVKuyF+BJeuop+AXaPca/TVmuvX7GlWvFOQa2HqKNR6CDmONKjNbGAwIDxRyRfbZp42ey8sUcZ5LTxTG/7c3SxseN/KOYIyImSBvfvX6rmKNtYIQ2nr7hFzHH8q1bxDHFQXHa+G3dE8Uc8TFR2rivPUDMgYwMbdjPzyGmOJxyUBvPzdLftwCQGN9NHDPzyBxtfPnCvWKOVgMTtPEYh/4cAoCsNP35/O3ir/QJ8ovFOaq7symp8A/09xj/dn+a9vrn8g1qZAt9jayLTmKOLkKNDAmWa+SKfpu08ZyzSsxx5limOGb33h3auFGNDNef378a1MgOQo1MM6iR3+/aIo4pDmqsjZvUyPoxDbRx3+AQMQcyzmnDPsqkRv6ojZ83+D3aKbqjOGbDb/pzZNW842KO5gOba+Mx9Uxq5DFtfMXCbz0Hfwf1McPhgO7Ms1kwh1R1bBZMopRc22zCRHIGw3kM8hjMpJ/DaBLxyHudwYpVGI0ySWKTjpnB7ZWWYcF5dmGQuBI5hxV3jrgQ3SSZAOTfOUAZ34mePHkyunTpguDgYERGRmLw4MHYs2ePy5iePXvCZrO5XP7yl7+UZRoiomqH9ZGIyDPWSCKqScq0iV69ejXGjBmDdevW4dtvv0VhYSH69u2LnJwcl3H33nsvjh8/7ry8/PLLli6aiKiqYX0kIvKMNZKIapIyfZx76dKlLv+eNWsWIiMjsWnTJvTo0cP588DAQERHR1uzQiKiaoD1kYjIM9ZIIqpJvPpisYz//1vPsLAwl58nJSUhIiICbdq0wfjx45Gbm+sxR35+PjIzM10uRETVnRX1EWCNJKKaic8hiag6K/cXixUXF+ORRx7BNddcgzZt2jh/ftddd6Fhw4aIjY3Fr7/+in/+85/Ys2cPPv30U7d5Jk+ejIkTJ5Z3GUREVY5V9RFgjSSimofPIYmouiv3JnrMmDHYvn07fvjhB5ef33fffc7/btu2LWJiYtC7d28cOHAATZo0KZVn/PjxePTRR53/zszMRIMG+m/7JCKqyqyqjwBrJBHVPHwOSUTVXbk20WPHjsWXX36JNWvWIC5O30aoW7cLbXX279/vtgDa7XbY7Z5bWRERVSdW1keANZKIahY+hySimqBMm2ilFB566CEsWrQIq1atQkKCvh8sAGzZsgUAEBMTU64FEhFVB6yPRESesUYSUU1Spk30mDFjMHfuXHz++ecIDg5GamoqAMDhcKB27do4cOAA5s6dixtvvBHh4eH49ddfMW7cOPTo0QPt2rUr08K2rfsVvv6el5ebm6e9/sx31otzDLl7uTb++j3Pizniof/YUCvcIOawQ/8q6tJg+bb0v6+rOGb9sp3a+Ly5T4k5WrTuro03KSoScySf0MeXfPmtmGPW4i3imK7t47XxjjFyM/UWLdpr4/VbNBZzbFvzlTbeJCZKzNG8bQttPOiI/ARj44Ht4pjtR4/oB5wXUwC++u8rPJGbL6bYtmCuNh5QWKyNqwIFeRZrXc76CADb1231qka+9+FGcY6bhRo5bbRcI+vD/TvsJZpbUSND5Rp5/agO4phfpBo56zExR4uO12vjTc6Z1MgCbfy7ZXKNfGvxNnFM15b1tXGTGtmscUttPDpOrpE71+prZINwuUYmCDXS/7dwMcevB/aKY7afPK4foMQU8BdrpJxk54J3tHG/HM+1QRUqo1JutctdI3XkIyyPsMGmz6D08Qs59PPIGS68OOEtm02eSZrFZK2Q1mqURDru3mYoGSTNY8E5YlIwxANvsg5pgMn9bzCPwTkvzyPMYcFj0ypl2kTPmDEDANCzZ0+Xn7///vsYNWoU/P398d133+GNN95ATk4OGjRogKFDh+Kpp+TNGRFRdcb6SETkGWskEdUkZf44t06DBg2wevVqrxZERFQdsT4SEXnGGklENYlXfaKJiIiIiIiIfk+4iSYiIiIiIiIyxE00ERERERERkSFuoomIiIiIiIgMcRNNREREREREZKhM3859OaWdyITNz9djvEk9fV/c5DXyHJ+u+UkbTzkwSszx3OQHtPG2aCXmaIKbtPE/4woxxxLfz8QxhX3jtPGUCKFHMIADR5dq47kREWKOg4VCj8wujcQctQPFIdiZkqKN5wbIOVCs70ecvnOPmKJtYqI+R8o+McfOvfoxfoFyP9flWzaLYzYnCwN6iCng8PHTxtfOWibniNXfHkdRtjZenK9w5LJ3ir68Uk/kwceLGnlgpTzH5yv1/ZdTDWrk08/ra2QndBBzNBJq5N1oJOb4zv8zcYztZn1PY5MaefikUCPDQsQcB2364nSki/w7wZIaafIyu1Ajz+yU+1W3FGrkCZMaudX7Gvn1lnXimK0HhAG9xBSo7aP/luqfZy0RcwTF1tPG6zlOe4wV5yscEmeo7jIAyI81b8h9c0161VrQR1iYxqRnrlGnaeHb1S1oEWzW31dYh9kyTObxPoUV6xD7InvfJtyoubZZL3HpvjG5vVYQzlXNyZqZCTjkXxcA+E40ERERERERkTFuoomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjJUq7IX4Em96HD42jXLy82o8DVsfClFHDOr2fva+Ig/DRdzRKOJNh6AzmKOISgUxxT6LtPGfbqIKeBoW6SNF52R75fsiGxtPPHKbmKOrrk9xDFHtx3QxpNPy2uNLzqojR/PSBdzBEbqu7ZHhIWJOfyK9cf902ULxByfrNorjpHU79NXHLP2T3O18T8/N1TM8dzIV7XxB6YO0cYLzxXhCLaK81RnkTHh8PX3XCMLzld8jVz/wlFxzNyG+hppu1eukTeigTZex6BG3gKbOCYHX2rjPlcViDkchTHa+LkzJ8UcBRlSjewo5uhuUCMPbtqlje81qJHRQo08dS5TzFErtJ42HhUUJOaok3dOG5+/Qn/fAsCi7/W/M0xE9+ktjvn5Tx9p46OeGyzmeO4P+hr58AzPNbLwXBEOYbs4R01mU/q4ssn1AkIOg5IjpjBKYgVlsBLhmMgZDEYZJJEPu3zMlMmRF26vNfeMwQ0WJzK5vd5mMDpFxGNmeJJY4PI8bvhONBEREREREZEhbqKJiIiIiIiIDHETTURERERERGSIm2giIiIiIiIiQ9xEExERERERERniJpqIiIiIiIjIEDfRRERERERERIaqbJ/ooNha8A3w8xjPzRUS1DWYRN96F5DbW+KjP+t770Y3WS7miLkuQhvvDM/H4X96iSPuRHttfD30/X0B4LcAff/S9Jg8McdOn43a+MEMfRwATq2Xex5HH9cf1+zCtmKO3RH6/tuOsEZijowMfY6d27aJOZan7NPGtwcVizmMHhP6ltY4ulDfaxwAILSWvfEufc9fALALPZ5bxF+hjefnFmJJDe8T7R+jr5HnpRoZYjCJrxA/I6eYe5++926YSY28Xv9Y7mhUI+W+yXcJNXKjQY3c7XdaG0+PNKiR572vkVk/J4tjwg7pT4Lswnwxx2+x+vpW26hG5mjj+/d4XyP3hhnUSJPHhFAjUxfK5zOEVuE33ibXyGDoe3w3jvZcI/NzC4Ea3idauJughDayRn1zhaa3Jv2KrejgK2Uwas1r1FbX+x7PYh9hA2bHVWBJv2KD+0Zormx2S4T+3Ca3RWrfbHTfGcxjBSvuXum4W3Rj+E40ERERERERkSFuoomIiIiIiIgMcRNNREREREREZIibaCIiIiIiIiJD3EQTERERERERGeImmoiIiIiIiMgQN9FEREREREREhriJJiIiIiIiIjJUq7IX4ElxYRpsvr4e43u2HNEnOGMwyZ1C/CODHII3B/0gjjn3db42nnF1rpijJ/qKY3zRXhvvhjFijm7Yp42vx2Yxhz1Kf3tjomLEHGcczcUxH744Uxv/ft16Mce1/YZp49sXyCdJRnqaNu4Lz+d5ichuLbTxGx8cLeZodbN+HQBwNP2wNu4X6CfmaN+jqzZe7Cefz3twVBvffUB/WwrzzotzVHe1CtPgq6mRO6QamSnPYbtLH1dz5RySqYPWimOkGvmHa+X7+2rI550/OmnjnfGAmKMzDmjjP2GdmCM4VmnjMbHhYo40h/4xBAAfvjhFG/9+3UYxx7X9btfGtyfNFnOcE2pkkQU1ss89co1s2tegRmbqx9QxqJEtr+qgjdcKkM/VbTiojetq5O+hPmYACPHi+vpH4OXNIrGJq5DXYROzAEp5f3ukHDabFeuQcxhMYwmT2yMRb60Vt8Wq4yHcNyZnkHQumpyGVhx3E2V6J3rGjBlo164dQkJCEBISgsTERHz99dfOeF5eHsaMGYPw8HAEBQVh6NChSEuTfykREVV3rI9ERJ6xRhJRTVKmTXRcXBxeeuklbNq0CRs3bsT111+PW265BTt27AAAjBs3DosXL8aCBQuwevVqHDt2DEOGDKmQhRMRVSWsj0REnrFGElFNUqaPcw8cONDl3y+88AJmzJiBdevWIS4uDu+++y7mzp2L66+/HgDw/vvvo2XLlli3bh2uuuoqtznz8/ORn/+/j+plZhp8xpCIqIqpiPoIsEYSUc3A55BEVJOU+4vFioqKMH/+fOTk5CAxMRGbNm1CYWEh+vTp4xzTokULxMfHY+1az3/zNnnyZDgcDuelQYMG5V0SEVGVYFV9BFgjiajm4XNIIqruyryJ3rZtG4KCgmC32/GXv/wFixYtQqtWrZCamgp/f3+Ehoa6jI+KikJqaqrHfOPHj0dGRobzcviw/ouNiIiqKqvrI8AaSUQ1B59DElFNUeZv527evDm2bNmCjIwMfPLJJxg5ciRWr15d7gXY7XbY7fZyX5+IqKqwuj4CrJFEVHPwOSQR1RRl3kT7+/ujadOmAIBOnTphw4YNePPNN3HnnXeioKAAZ8+edXklMS0tDdHR0ZYtmIioqmJ9JCLyjDWSiGoKr/tEFxcXIz8/H506dYKfnx+WL1+OoUOHAgD27NmDlJQUJCYmljnv3vW7YPPz3OeraElhuddcwi9C33uy8JoiOcmPQtzgOy7eGrBBGw/aHCTmONBkuzimCfS9lXtC3+8TAHxxhTbeDXKPzBg4tPF4XCnmQFSUOOTmNxO08XufmyDmWHRE3+c75YjQixdA/RYR2njvmweJOW7o20Mb79+kv5jjl9M7xTEpSBdGyI+JRmGttPEzQg/oC/S9wrv1ydDG87LzsATfG8xTsSqqPgLA3p93wVbLc40s+CavXHkvZhMarSrP34f2P1Jb5Bw5xbsD9P3nTWrkrivkHvYt0VYbv9qgRvqjiZDDpEbW08YT0EbMgdhYcciQqfoaOfLZx8Uci07oa+TRE8fFHNEt9PW8981yfZNq5KBGcp39JcOgRhYc0w/wl3taxzqaauM5OCnmAPS9wrv1Oecxlpedh2/wk8EcFa8ia6SOSV9kOYnQz9YoidA32esMFt1WA9bMYtLT2oKFGNw5yibdNwYTifNYcHstOPImvcSNSP2ZLeg1btptuvw5MgFhn1KiTJvo8ePHY8CAAYiPj0dWVhbmzp2LVatW4ZtvvoHD4cCf//xnPProowgLC0NISAgeeughJCYmar95loioJmB9JCLyjDWSiGqSMm2iT5w4gT/+8Y84fvw4HA4H2rVrh2+++QY33HADAOD111+Hj48Phg4divz8fPTr1w/Tp0+vkIUTEVUlrI9ERJ6xRhJRTVKmTfS7776rjQcEBGDatGmYNm2aV4siIqpuWB+JiDxjjSSimqTcfaKJiIiIiIiIfm+4iSYiIiIiIiIyxE00ERERERERkSFuoomIiIiIiIgMed0n2mrq/3uIqfMW9SzTzVUgzHG+wpdwgbCM/Cx5IecyC8QxOdD3jc1EtpjDV2x8nSXmyEKusA45B1BbHJEp3N7CPPkcKy4oFgaIKVBcqB9UcE6+785lee75CQBZmfIxy8mUG/KeE+4bkz7RObX051GuOAeQI5yLedn6+zY/Jx/A/+pJTXJ5a6QwoKrUyGyTGlkojpFrpPw487ekRuofq2Y1UloHkAl9XSnMt6BGyiVDrJGFFtTITKtqZKFQv/zkPtE5NqlGyuvIgV0b19XI/JwLsZpcHzMNzv+KZrYC79d5uW5pZmblH1MA1txgg9tiya0Vk1SNY3rZHi8WnEPWnIe6HBdiJvXRpqpYFT1y5AgaNGhQ2csgohrg8OHDiIuLq+xlWIo1koiswPpIROSeSX2scpvo4uJiHDt2DMHBwbDZbAAuvOrQoEEDHD58GCEhIZW8wpqBx7Ri8LharzzHVCmFrKwsxMbGwsenZv3VyqU1kuec9XhMKwaPq/VYH12xPl4ePK7W4zGtGGU9rmWpj1Xu49w+Pj4ed/4hISE8sSzGY1oxeFytV9Zj6nA4KnA1lcdTjeQ5Zz0e04rB42o91scLWB8vLx5X6/GYVoyyHFfT+lizXoIkIiIiIiIiqkDcRBMREREREREZqhabaLvdjmeeeQZ2u/7bKMkcj2nF4HG1Ho+pHo+P9XhMKwaPq/V4TPV4fCoGj6v1eEwrRkUe1yr3xWJEREREREREVVW1eCeaiIiIiIiIqCrgJpqIiIiIiIjIEDfRRERERERERIa4iSYiIiIiIiIyxE00ERERERERkaEqv4meNm0aGjVqhICAAHTr1g0///xzZS+pWlmzZg0GDhyI2NhY2Gw2fPbZZy5xpRQmTJiAmJgY1K5dG3369MG+ffsqZ7HVxOTJk9GlSxcEBwcjMjISgwcPxp49e1zG5OXlYcyYMQgPD0dQUBCGDh2KtLS0Slpx1Tdjxgy0a9cOISEhCAkJQWJiIr7++mtnnMfTPdZH77A+Wo/1sWKwRpYPa2T5sT5aj/WxYlRWfazSm+iPPvoIjz76KJ555hls3rwZ7du3R79+/XDixInKXlq1kZOTg/bt22PatGlu4y+//DKmTJmCmTNnYv369ahTpw769euHvLy8y7zS6mP16tUYM2YM1q1bh2+//RaFhYXo27cvcnJynGPGjRuHxYsXY8GCBVi9ejWOHTuGIUOGVOKqq7a4uDi89NJL2LRpEzZu3Ijrr78et9xyC3bs2AGAx9Md1kfvsT5aj/WxYrBGlh1rpHdYH63H+lgxKq0+qiqsa9euasyYMc5/FxUVqdjYWDV58uRKXFX1BUAtWrTI+e/i4mIVHR2tXnnlFefPzp49q+x2u5o3b14lrLB6OnHihAKgVq9erZS6cAz9/PzUggULnGN27dqlAKi1a9dW1jKrnbp166p33nmHx9MD1kdrsT5WDNbHisMaqccaaR3Wx4rB+lhxLkd9rLLvRBcUFGDTpk3o06eP82c+Pj7o06cP1q5dW4krqzkOHjyI1NRUl2PscDjQrVs3HuMyyMjIAACEhYUBADZt2oTCwkKX49qiRQvEx8fzuBooKirC/PnzkZOTg8TERB5PN1gfKx7rozVYH63HGiljjaxYrI/WYH203uWsj7W8XWxFSU9PR1FREaKiolx+HhUVhd27d1fSqmqW1NRUAHB7jEtipFdcXIxHHnkE11xzDdq0aQPgwnH19/dHaGioy1geV71t27YhMTEReXl5CAoKwqJFi9CqVSts2bKFx/MSrI8Vj/XRe6yP1mKNNMcaWbFYH73H+mityqiPVXYTTVQdjBkzBtu3b8cPP/xQ2Uup9po3b44tW7YgIyMDn3zyCUaOHInVq1dX9rKIqJxYH63FGklUc7A+Wqsy6mOV/Th3REQEfH19S317WlpaGqKjoytpVTVLyXHkMS6fsWPH4ssvv8TKlSsRFxfn/Hl0dDQKCgpw9uxZl/E8rnr+/v5o2rQpOnXqhMmTJ6N9+/Z48803eTzdYH2seKyP3mF9tB5rpDnWyIrF+ugd1kfrVUZ9rLKbaH9/f3Tq1AnLly93/qy4uBjLly9HYmJiJa6s5khISEB0dLTLMc7MzMT69et5jDWUUhg7diwWLVqEFStWICEhwSXeqVMn+Pn5uRzXPXv2ICUlhce1DIqLi5Gfn8/j6QbrY8VjfSwf1sfLhzXSM9bIisX6WD6sj5fPZamPXn0tWQWbP3++stvtatasWWrnzp3qvvvuU6GhoSo1NbWyl1ZtZGVlqV9++UX98ssvCoB67bXX1C+//KIOHTqklFLqpZdeUqGhoerzzz9Xv/76q7rllltUQkKCOnfuXCWvvOp64IEHlMPhUKtWrVLHjx93XnJzc51j/vKXv6j4+Hi1YsUKtXHjRpWYmKgSExMrcdVV2+OPP65Wr16tDh48qH799Vf1+OOPK5vNppYtW6aU4vF0h/XRe6yP1mN9rBiskWXHGukd1kfrsT5WjMqqj1V6E62UUv/5z39UfHy88vf3V127dlXr1q2r7CVVKytXrlQASl1GjhyplLrQpuDpp59WUVFRym63q969e6s9e/ZU7qKrOHfHE4B6//33nWPOnTunHnzwQVW3bl0VGBiobr31VnX8+PHKW3QV96c//Uk1bNhQ+fv7q3r16qnevXs7i59SPJ6esD56h/XReqyPFYM1snxYI8uP9dF6rI8Vo7Lqo00ppbx7L5uIiIiIiIjo96HK/k00ERERERERUVXDTTQRERERERGRIW6iiYiIiIiIiAxxE01ERERERERkiJtoIiIiIiIiIkPcRBMREREREREZ4iaaiIiIiIiIyBA30URERERERESGuIkmIiIiIiIiMsRNNBEREREREZEhbqKJiIiIiIiIDP0fygh/i4jV2kIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9EAAAFTCAYAAAA+14+JAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABkR0lEQVR4nO3dd3hUVf7H8c8khARIoYQEkBYQRVBgl5ViRUUQK4odFfRnXbDu6oprAwuWlaaU1VWxBNhFxb64CgIWQEFBUUFAEBAIPYWQEJLz+4PNLCEz95wwN5X363nmeWDOme85986db+6Zcr8BY4wRAAAAAACwiqrsCQAAAAAAUF2wiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBlAjzZkzR4FAQG+88UZlTwVAFTV58mQFAgEtWrSosqcCAKhGKmUR/fDDDysQCGjbtm0h24899lj16tUr+P+1a9cqEAgoEAjozTffdIo3ePDg4GMCgYASExPVuXNnPfPMM8rPzy/12HC3zZs3l5pDIBBQVFSUGjZsqH79+mn+/PmS/vfH2HZr3bp1cI7x8fFh91N8fLwGDx4c/H/xoqD4Fh0drZSUFF188cX66aefSj3+4H1Qq1YttWjRQpdffrl+/PHHEn0Pjn3wbdq0acG+rVu3LrEf6tevr+OOO0433nijFi5cGHZ7DtarV68ScRITE3X00Ufr6quv1scff+z52J9++kmBQEBxcXHatWtX2G0Odztwv+7atUtxcXEKBAIh92NZBQIBDR061Ln/hAkTFAgE1L17d8+YB97q1aunDh066NFHH1Vubm6Jvl77IC4uLtivPBeZGzdu1MMPP6wlS5b4Hhvlo6zHYa1atdSwYUN17dpVt99+e6mccv7556tu3brKzs4OG2/gwIGqXbu2tm/fXir+wbebb745+LiDj/HY2FgdddRRevDBB5WXl1dqnJycHD300EM69thjVa9ePTVq1EhdunTR7bffro0bN4ac24cffqhAIKBmzZqpqKgoZJ8Dc2Hx67Jbt2569dVXS/W1vd5c939ZcouX4r9pf/vb35wfc8899ygQCOiyyy7zjHnw394uXbroueeeU2FhYYn+B/4NOPjWvn37YD8WukDlOfjcNi4uTs2aNVPfvn01bty4kDm+rOfWB+chr9xw4O3hhx8OPibceWFZtvPAeDZFRUWaNGmSunTpovj4eKWmpqpfv3768ssvS/TzOr9esGCB01i//fabLr30UtWvX1+JiYm64IIL9Msvv5Rl81AOalX2BMpqxIgRuuiiixQIBKx9Y2Nj9Y9//EPS/oXSm2++qT//+c/6+uuvSywKJWnixIkhF7T169cv8f8rrrhCZ599tgoLC/Xzzz9rwoQJOu200/T111/rlFNO0WuvvVai//XXX69u3brpxhtvDN7ntXB2cdttt+n4449XQUGBvvvuO02aNElz5szRsmXL1KRJkxJ9D9wH+/bt0+rVqzVp0iTNnDlTP/74o5o1axYy9sF69uxZ4v9dunTRn/70J0lSdna2fvrpJ02fPl0vvPCC7rzzTo0aNcppW5o3b66RI0dKknbv3q1Vq1bprbfe0uuvv65LL71Ur7/+umJiYko97vXXX1eTJk20c+dOvfHGG7r++uslSTfddJN69+4d7LdmzRo9+OCDuvHGG3XyyScH72/btm3w39OnT1cgEFCTJk2Unp6uRx991GnufklPT1fr1q311VdfadWqVTryyCND9jvzzDN1zTXXSNq/KPjss8/0wAMPaOnSpZo+fXqJvgc+7weKjo72fwNC2Lhxo4YPH67WrVurS5cuFTImIlPW49AYo8zMTC1dulSvvPKKJkyYoCeffFJ33XWXpP0L5Pfee08zZswIHrcHys3N1TvvvKOzzjpLjRo1KhX/YEcddVSJ/x94jGdmZuqdd97RI488otWrVys9PT3Yr6CgQKeccoqWL1+uQYMG6dZbb1VOTo5++OEHTZkyRRdeeGGpPHjg/li7dq1mz55dIq8c6MBcuGnTJv3jH//QoEGDlJ+frxtuuCHkY0Jx3f+VxRijqVOnqnXr1nrvvfeUnZ2thISEkH2L/05K+5+bDz/8ULfeeqt+/fVXPf300yX6Hvg34EBJSUn+bwSAQzZixAilpaWpoKBAmzdv1pw5c3THHXdo1KhRevfdd9WpU6dSj3E9tz7YX//61+B5nSR9/fXXGjdunO677z4dc8wxwfsPHDPceaGXdevWaffu3SViSvvPYbZv367jjjsu7GPvvvtujRo1SldddZX++Mc/ateuXfr73/+uU089VV988YW6detWon+o82uXPJ+Tk6PTTjtNmZmZuu+++xQTE6PRo0fr1FNP1ZIlS0r8/UQFM5XgoYceMpLM1q1bQ7Z37NjRnHrqqcH/r1mzxkgyXbp0MZLMm2++aY03aNAgU69evRL9CgsLzR/+8Acjyfz2229Oczl4Dk8//XSJ+//9738bSeaWW24J+bh69eqZQYMGhWwLNUevx3766adGkpk+fXqJfhMnTjSSzJNPPukU//333zeSzPPPP2+NHUqrVq3MOeecU+r+3Nxc079/fyPJTJgwwRrn1FNPNR07dix1/759+8wf//hHI8ncc889pdqLiopM69atzV133WUuvPBC06tXr7BjfP3110aSefnll8P2OeWUU8xFF11k7rzzTpOWlmadt40kM2TIEKe+v/zyi5Fk3nrrLdO4cWPz8MMPlynmxRdfbKKiosyePXuC99mOq2Jlec7LymW/H2j37t2+z6E8t6+mifQ43LZtm+nZs6eRZD744ANjzP58kJCQYPr27Rsy1pQpU4wkM23aNGv8g4U6xouKikyPHj1MIBAwmzdvDt7/r3/9y0gy6enppeLs2bPHZGZmlro/JyfH1KtXz4wbN8787ne/M4MHDw45j1C5cMuWLSY+Pt4cc8wxJe73Oh4j3f+HItzftHBmz55tJJnZs2ebmJgYM3nyZOeYRUVF5vjjjzfNmjUrcX+4vwEHe/nll40k8/XXXzvNtSzKM/aByiPHARXB6zUya9YsU6dOHdOqVSuTm5sbvD/Sc+uDTZ8+3Ugyn376acj2spwXHjz/Vq1amUceecQ8//zz5sEHHzQTJ040rVq1KvG36WAFBQWmTp065uKLLy5xf3Euv+2224L3RXou8uSTTxpJ5quvvgre99NPP5no6GgzbNiwQ4oJf1Sr30RffvnlOuqoozRixAgZY8r8+KioKPX679fE165d68ucij/dXL16tS/xKmIOxZ9W16rl7xcR6tSpo9dee00NGzbUY489dkjPkbT/09Jx48apQ4cOeu6555SZmVmi/YsvvtDatWt1+eWX6/LLL9e8efO0YcOGQxpr3bp1+uyzz4Kx1qxZU+qrOOUpPT1dDRo00DnnnKOLL764xCdoLpo0aRL8am1VMWfOnOC7rddee23wa0uTJ0+WtP9rWscee6wWL16sU045RXXr1tV9990nSaW+nlWsdevWJb6CL+3/dsmdd96p1q1bKzY2Vs2bN9c111wT9mcikpSfn69zzz1XSUlJFfo8V3WRHoeNGjXStGnTVKtWLT322GOS9ueDiy66SLNmzdKWLVtKPWbKlClKSEjQ+eef78s2BAIBnXTSSTLGlPiaW3FePPHEE0s9Ji4uTomJiaXunzFjhvbs2aNLLrlEl19+ud56662QXxMPpXHjxmrfvn2Z/iZEuv8rQnp6ujp06KDTTjtNvXv3LtMcA4GAUlNTq1SeOlh+fr7uuusuNW7cWPXq1dOFF16orVu3luo3YcIEdezYUbGxsWrWrJmGDBlS6qujXjlu0aJF6tu3r5KTk1WnTh2lpaXpuuuuK/H4oqIijRkzRh07dlRcXJxSU1N10003aefOneW2/cChOP300/XAAw/o119/1euvv15p8zjU88LTTz9d33//vXJzczVs2DCNGTNGS5cu1aJFi8L+bEXa/w2nPXv2KDU1tcT9KSkpioqKUp06dUI+Ljs7W/v27SvTtr3xxhs6/vjjS3yK3b59e51xxhn617/+VaZY8Fe1WkRHR0fr/vvv19KlSzVjxoxDilF8YnPw1x927Nihbdu2lbi5/KaieDHeoEGDQ5qPH2xzKN6ejIwMzZ8/X3feeacaNWqkc889t1Tf7OzsUvth27Ztzgvi+Ph4XXjhhfrtt99K/UayLKKjo3XFFVcoNzdXn3/+eYm29PR0tW3bVscff7zOO+881a1bV1OnTj2kcaZOnap69erp3HPPVbdu3dS2bdsKPYFNT0/XRRddpNq1a+uKK67QypUr9fXXX4fsm5eXF3w+fv31V02ZMkWvvPKKrrzyypAnp6Gex6ysrDLPcefOnSFjHXwr/m32McccoxEjRkiSbrzxRr322mt67bXXdMoppwRjbt++Xf369VOXLl00ZswYnXbaaWWaU05Ojk4++WQ9++yz6tOnj8aOHaubb75Zy5cvD/uHc8+ePTrvvPP05Zdf6pNPPtEJJ5xQ5n1RU5XlOAynZcuWOvXUU7VgwYLgcTZw4EDt27ev1B/6HTt26KOPPtKFF15Y6mTjwOP8wNvevXutcwiVC1u1aiVJevXVV53zWHp6uk477TQ1adJEl19+ubKzs/Xee+85PXbfvn3asGFDmf4m+LH/y1N+fr7efPNNXXHFFZL2f1179uzZwd81Hiw3Nzf4vP3yyy8aP368Zs6cqUGDBpXqW1hYGPL53r17d5nn6ZKntm3bVuK6KMVuvfVWLV26VA899JBuueUWvffee6V+f/7www9ryJAhatasmZ555hkNGDBAf//739WnTx8VFBSU6Bsqx23ZskV9+vTR2rVrde+99+rZZ5/VwIEDS/0u8qabbtLdd9+tE088UWPHjtW1116r9PR09e3bt9Q4QGW7+uqrJUn/+c9/SrUd6rl1WUVyXhgVFaWoqP8th1x+LlqnTh11795dkydPVnp6utatW6fvvvtOgwcPVoMGDUr8hLPYtddeq8TERMXFxem0005zusZDUVGRvvvuO/3hD38o1datWzetXr3a87ojKGeV8fH3oX6d++mnnzb79u0z7dq1M507dzZFRUVh4xV/3W/r1q1m69atZtWqVebxxx83gUDAdOrUqdRcQt2OPvroUnMYPny42bp1q9m8ebP57LPPzPHHH+/5NY3y+Dr3Sy+9ZLZu3Wo2btxoZs6caY488kgTCARKfNWjOH6o7TriiCPM4sWLS/Qtjh3utmnTpmDfcF/nLjZ69Ggjybzzzjth+xhj/yrfjBkzjCQzduzY4H179+41jRo1Mn/961+D91155ZWmc+fOIWPYvlZ83HHHmYEDBwb/f99995nk5GRTUFDgOXcvcvzK5aJFi4wk8/HHHxtj9n8dqXnz5ub2228PGTPUrX///iYvL69E33DPu6QSX611/YpRq1atPI+N4ttDDz0UfIzXfj/11FONJDNp0qSQ23lgnAPncOBr4cEHHwx+/fVgxXnhwO3Lzs42p556qklOTjbffvut5/Yebsp6HHod27fffruRZJYuXWqM2f/TjKZNm5qePXuW6Ddp0iQjyXz00Uel4oe7TZ06NdgvVH7/29/+ZgKBgDn22GODx4Ax+79WfvTRRxtJplWrVmbw4MHmxRdfNBkZGSG3ISMjw9SqVcu88MILwftOOOEEc8EFF5Tq26pVK9OnT5/gPL7//ntz9dVXh9xP4V5vfu7/sijL17nfeOMNI8msXLnSGGNMVlaWiYuLM6NHjw4ZM9TtlltuKfG8GPO/XBDqdtNNNwX7uX7l2iVPHZyXimP37t27xPzuvPNOEx0dbXbt2mWM2f81/dq1a5s+ffqYwsLCYL/nnnsu+Hf54O06OMcV/03z2o7PPvss5M8PZs6cGfZnCUB5cnn9JSUlmd/97nfB/5f13DqSr3OX9bzwQJ9++qlJS0szI0aMCH6de8KECdavcxtjzMqVK83vf//7EtvWpk0bs3z58hL9vvjiCzNgwADz4osvmnfeeceMHDnSNGrUyMTFxZlvvvnGc4ytW7caSWbEiBGl2saPH28klRoPFafqfrcqjOJPowcNGqS3335bF154Ydi+u3fvVuPGjUvcd8IJJ5S6+Jckvfnmm6W+1levXr1S/R566CE99NBDwf/Hx8frmWee0cUXX1zWTTlkB3/1q3HjxnrttddCXhAsLi4u+AlKUVGR1q5dq1GjRunss8/WvHnzSl2s58EHHyxxAa5iDRs2dJ5f8UUkIn13LFScf//739q+fXvwExFp/6ci5513nn744Qd17NjROf53332n77//vsRFba644go9/vjj+uijj3TOOedENH+b9PR0paamBj+FLb7q7euvv65nnnmm1EXALrjgguAnI7m5uVqwYIFGjx6tK6+8Um+88UaJd08PfN4PlJycfEjz3LNnj7VfmzZtnGPGxsbq2muvLfNcir355pvq3LlzyNf/we8iZ2Zmqk+fPvrll180Z86cMh0jh4OyHodeDn7NRkdH6/LLL9fo0aO1du3aYGWCKVOmKDU1VWeccUapGAce5wc6+AIvofL7SSedpFdeeaXEMVCnTh0tXLhQjz32mP71r39p8uTJmjx5sqKiovTHP/5Rf/vb3xQbGxvsP23aNEVFRWnAgAHB+6644gr96U9/0s6dO0t9wvyf//yn1DyuvfbaUhfQCsfP/V9e0tPT9Yc//CF4EZyEhASdc845Sk9P1x133FGq/4033qhLLrlEkpSVlaXZs2dr4sSJio2N1ejRo0v0bd26tV544YVSMZo3b17medqqOhQLlQNuvPHGEsfNySefrNGjR+vXX39Vp06d9Mknn2jv3r264447SnxqdcMNN+i+++7TBx98UCKnhcpxxRdTev/999W5c+eQF82cPn26kpKSdOaZZ5b4aUrXrl0VHx+vTz/9VFdeeaXTdgIVJT4+PuQ5n+u5dSQiOS9MS0vTBx98oGOOOSZ4FfJbbrlFF1xwQbBqRDgJCQnq2LGjevbsqTPOOEObN2/WE088of79++uzzz4Lnm+dcMIJJb75dv755+viiy9Wp06dNGzYMM2cOTPsGMXnXgf+jSpWXG3F5fwM5aPKLqK9vk4xcOBAPfLIIxoxYoT69+8ftt+BC4nY2FilpaWF/cN8yimnOC0wik8O8vLyNHv2bI0bN65U2Q4/hdoPxQvdnJwczZgxI3jSF0p0dHSpq8qeffbZateunYYNG1aqZNhxxx0X9iq0rnJyciQp7JVbI4nz+uuvKy0tTbGxsVq1apWk/Vfarlu3rtLT0/X44487x3/99ddVr149tWnTJhgrLi5OrVu3Vnp6erkuogsLCzVt2jSddtppWrNmTfD+7t2765lnntGsWbPUp0+fEo9p3rx5iefm/PPPV6NGjfTnP/9Z77//vs4777xgW6jn/VCF+i1ppI444gjVrl37kB+/evXqEoscL3fccYfy8vL07bffsoA+yKEch15CvWYHDhyo0aNHa8qUKbrvvvu0YcMGffbZZ7rttttCLhAPPs7DOTC/b9iwQU899ZS2bNkS8rdoSUlJeuqpp/TUU0/p119/1axZs/S3v/1Nzz33nJKSkkpckf/1119Xt27dtH379uBJ1O9+9zvt3btX06dPL/U1ve7du+vRRx9VYWGhli1bpkcffVQ7d+50Or793v/lYdeuXfrwww81dOjQYJ6U9ueFN998Uz///HOpN2PbtWtX4jksrqgxZswYXXfddSXeEKlXr55vuSqSOC1btizx/+I3S4p/h/zrr79Kko4++ugS/WrXrq02bdoE24uFynGnnnqqBgwYoOHDh2v06NHq1auX+vfvryuvvDJ4krxy5UplZmYqJSUl5DxDXV8AqGw5OTkhj1nXc+tIRHJeWPxzn4M1a9YsZNWGYvv27VPv3r3Vq1cvPfvss8H7e/furY4dO+rpp5/Wk08+GfbxRx55pC644AK99dZbKiwsDPtmafHfs1A/QSm+Tke431+j/FXKItr27klubm6JerYHK/40evDgwXrnnXc8+/n1x7nYgScH5557rqKjo3XvvffqtNNOC/mbBS9xcXHKz8+XMabUYtkYo7y8vJD74cCFbv/+/ZWbm6sbbrhBJ510klq0aGEdt3nz5jr66KM1b968Ms3X1bJlyyS5Xbq/LHGysrL03nvvKS8vT+3atSvVf8qUKXrsscecfs9i/luuZffu3erQoUOp9i1btignJyficmThzJ49W5s2bdK0adNKlVuT9n/y43LyXPxJ3rx580osov20detWpzeK4uPjnfdXWZN+JG9UXXDBBZo2bZqeeOIJvfrqq2HfcDoc+XUcFlu2bJmio6OVlpYWvK9r165q3769pk6dqvvuu09Tp06VMUYDBw6MaO4H5/e+ffuqffv2uummm/Tuu++GfVyrVq103XXX6cILL1SbNm1KlLU78LfIoXJMenp6qUV0cnJycB7Fczj33HM1duzYYLmvcPze/+Vh+vTpys/P1zPPPKNnnnmmVHt6erqGDx9ujXPGGWfoueee07x58zzLxkQi3G+0D5aUlFQqB4U7iTWHeIHMUDmuuE74ggUL9N577+mjjz7Sddddp2eeeUYLFixQfHy8ioqKlJKSEvbaHAd/6wGobBs2bFBmZmallOXz87zw4IuXepk3b56WLVtWqpxru3btdMwxx+iLL76wxmjRooX27t2r3bt3h7zApbT/W6CxsbHatGlTqbbi+7wW+yhflbKILn7nZ8WKFaUWfbm5uVq/fr31xOGqq67So48+quHDh/t2dddD8de//lUvvPCC7r//fs+vZITSqlWrYO3mg5PPqlWrVFhYGPZdsgM98cQTmjFjhh577DFNmjTJaex9+/YFPzXyU/Gn4y1atChVd68sCgsLNWXKFNWtW1cnnXSSJAWvkDtx4sRS72yuWLFC999/v7744otgfy9z587Vhg0bNGLEiFLz3Llzp2688Ua9/fbbuuqqqw55G7ykp6crJSVF48ePL9X21ltvacaMGZo0aZJ1sVl8lcfyeC6LHX/88aU+ZQnloYceCl5Z2+UPVigNGjQoddGRvXv3lvoD0rZt2+CbLDb9+/dXnz59NHjwYCUkJGjixImHNLeayK/jUNp/pfu5c+eqZ8+epb6FMnDgQD3wwAP67rvvNGXKFLVr1y7kz08i0bRpU915550aPny4FixYoB49enj2b9CgQanjKD09XTExMXrttddKLao+//xzjRs3TuvWrSv1qeWBzjnnHJ166ql6/PHHddNNN3l+ddHP/V9e0tPTdeyxx5b4GVOxv//975oyZYrTIroiclXTpk2d+r388stlOmGWSp63HPjTlb1792rNmjVlesO+R48e6tGjhx577DFNmTJFAwcO1LRp03T99derbdu2+uSTT3TiiSfyCROqheKfSPbt27fCx/bzvLAsMjIyJIV+g7+goMDpCty//PKL4uLiPD98iIqK0nHHHRfyImQLFy5UmzZtIv7WJw5dpSyizzjjDNWuXVsTJ07U6aefXuKToeeff1779u1Tv379PGMc+Gl0Zapfv75uuukmPfXUU1qyZIm6dOni/Nh+/frpvvvu03PPPacxY8aUaCs+qbLtB2n/gmLAgAGaPHmyHn744WAJq3B+/vlnrVixQl27dnWeq4s9e/bo6quv1o4dO/T4448f8kKqsLBQt912m3766Sfde++9wXfoXn/9dbVp00Y333xzqcfk5+friSeeUHp6ulOyLP4q99133x3y0/6nn35a6enp5bKI3rNnj9566y1dcsklIX9L36xZM02dOlXvvvuuZ4kFScGvs3bu3Nn3eRY7lN9EFy8cynoVzrZt25b6hsTzzz9f6g/VgAEDNGLECM2YMaPU76JDfbPjmmuuUVZWlm699VYlJiZ6fs3qcOHncbhjxw5dccUVKiws1F//+tdS7cWL6AcffFBLliwJWcbMD7feequefvppPfHEE3r77bclSUuXLtURRxxR6gTr119/1Y8//lji67np6ek6+eSTQ25vz549NW7cOE2dOlV/+ctfPOfxl7/8RWeffbZeeOGFkL8Zlvzd/+Vl/fr1mjdvnoYPHx5yjnv37tXAgQO1cOFCde/e3TNWReSqSH4TbdO7d2/Vrl1b48aN01lnnRXMMS+++KIyMzOdfv6zc+dO1a9fv0R+Kj5nKP665qWXXqoJEybokUceKfU11OI3v4t/Ww1UttmzZ+uRRx5RWlpaxN8uOhR+nheWRfFPWKZNm6azzjoreP8333yjFStWlPjG0tatW0t9g2Tp0qV699131a9fvxJroHXr1ik3N1ft27cP3nfxxRfr3nvv1aJFi4LfeF2xYoVmz56tP//5z75uF8qmUhbRKSkpevDBB3X//ffrlFNO0fnnn6+6devqyy+/1NSpU9WnTx+nr6YW/zZ6yZIlEc/pjTfeCPlu0JlnnlmqDtzBbr/9do0ZM0ZPPPFEyK/khdOlSxddf/31Gjt2rFauXKkzzzxT0v4TgQ8//FDXX3+98wnH3XffrX/961/BeRTbt29fsHZf8YXFJk2apKKiopCfLHz22Wch66F26tRJnTp1Cv7/t99+C8bNycnRjz/+qOnTp2vz5s3605/+pJtuuslp3pmZmcE4ubm5WrVqld566y2tXr1al19+uR555BFJ0saNG/Xpp5/qtttuCxknNjZWffv21fTp0zVu3LiQF2wpVlyu5cwzzwz7s4Hzzz9fY8eO1ZYtW5SSkqI5c+botNNOK/Fpq5dFixaV+J1lsV69eum3335TdnZ22G9Q9OjRQ40bN1Z6enqJk+eff/65xL5asGCBXnnlFR155JHBEhPFDnzeD3bhhReW+HTszTff1PLly0v1GzRokFq0aHFIv4lu27at6tevr0mTJikhIUH16tVT9+7dS3zNN5Trr79eN998swYMGKAzzzxTS5cu1UcffVRqAXT33XfrjTfe0CWXXKLrrrtOXbt21Y4dO/Tuu+9q0qRJIV83Q4cOVVZWlv76178qKSkpWLf1cPXuu+9GdBwaY5SVlaWlS5dq+vTpysnJ0ahRo0qcUBRLS0vTCSecEPz5jdfJ1oHH+YFSU1ODOTKcRo0a6dprr9WECRP0008/6ZhjjtHHH3+shx56SOeff7569Oih+Ph4/fLLL3rppZeUn58ffD0vXLhQq1atCnlRM2n/b1x///vfKz093bqI7tevn4499liNGjVKQ4YMCZmPDnX/e+WWk046SZMnT9a1117r/GnrrFmzQub8/v3764MPPpAxJuwczz77bNWqVUvp6eklFtHffPNN8DnMzs7WrFmz9Oabb+qEE04o9S2zA/8GHOzgNzFfeumlkN/4uv3225WQkOD7z7cO1LhxYw0bNkzDhw/XWWedpfPPP18rVqzQhAkTdPzxxzu94frKK69owoQJuvDCC9W2bVtlZ2frhRdeUGJios4++2xJ+383fdNNN2nkyJFasmSJ+vTpo5iYGK1cuVLTp0/X2LFjK/RCpkCxf//731q+fLn27dunjIwMzZ49Wx9//LFatWqld999N+T5lOu5tVceOvbYY0POx8/zwrLq2rWrzjzzTL3yyivKyspSnz59tGnTJj377LOqU6dOiTdPL7vsMtWpU0cnnHCCUlJS9OOPP+r5559X3bp1S5yvS/vf8J87d26Jn5H88Y9/1AsvvKBzzjlHf/7znxUTE6NRo0YpNTVVf/rTn3zbJhyCSrsuuDHm9ddfNz169DD16tUzsbGxpn379mb48OGlSvZ4XQK/+NL7ClPiysbrMvw64HL6tsvwDx482ERHR5tVq1aVuN+rxJUxxhQWFpqxY8eazp07m7i4OBMXF2c6d+5sxo0bV6KMhjH2kkS9evUyiYmJwZIcoUodJSYmmjPOOMN88sknIWOHux1YdujAkkeBQMAkJiaajh07mhtuuMEsXLgw7LYe7ODyJvHx8aZdu3bmqquuMv/5z39K9H3mmWeMJDNr1qyw8SZPnmx0UGmtUKWW3nzzTSPJvPjii2FjzZkzx0j/K6/13nvvhS3LdDCv/fjII4+Y8847z8TFxZndu3eHjTF48GATExNjtm3bFjJmdHS0ad68ubnxxhtLlerxKnElyaxZs8YYY3/OP/vsM+u2ennnnXdMhw4dTK1atUo8B16lzQoLC81f/vIXk5ycbOrWrWv69u1rVq1aVarElTHGbN++3QwdOtQcccQRpnbt2qZ58+Zm0KBBwX0W7vVyzz33GEnmueeei2j7qrtIj8OoqChTv35987vf/c7cfvvt5ocffvAcr7gcR7du3cL28ToeDyx76JXfV69ebaKjo4PHyy+//GIefPBB06NHD5OSkmJq1aplGjdubM455xwze/bs4ONuvfVWI8msXr067PwefvhhI/2vhJdXub/ifFR83B98PPqRBw7OLcYY8+yzzxpJZubMmWHjGuNdjkqSee2118xxxx1nWrZs6RmnV69eJiUlxRQUFISMWatWLdOmTRtz9913m+zs7BKP9SpxdeDpyYF/50Pd1q9f7zlHL+HK9xQ/XweX1HnuuedM+/btTUxMjElNTTW33HKL2blzZ6ntCpXjvvnmG3PFFVeYli1bmtjYWJOSkmLOPfdcs2jRolJ9n3/+edO1a1dTp04dk5CQYI477jhzzz33mI0bNx7ytgKH4uDXX+3atU2TJk3MmWeeacaOHWuysrJKPaas59ZeeciY0CWuDvW80C+5ublmxIgRpkOHDqZOnTomKSnJnHvuuaXKaI4dO9Z069bNNGzY0NSqVcs0bdrUXHXVVcGSgQcqzokHW79+vbn44otNYmKiiY+PN+eee27Ix6NiBYw5xKtmAIeRe+65R1OnTtWqVatClhoAgKrg0ksv1dq1a/XVV19V9lQAAKixqmyJK6Aq+fTTT/XAAw+wgAZQZRljNGfOnLBfjwYAAP7gk2gAAAAAABxRMBUAAAAAAEcsogEAAAAAcMQiGgAAAAAARyyiAQAAAABwVOWuzl1UVKSNGzcqISFBgUCgsqcDoBoyxig7O1vNmjVTVFTNeq+QHAkgEuRHAAitTPmxvApQP/fcc6ZVq1YmNjbWdOvWzSxcuNDpcevXr/csvM6NGzdurrf169eXV4qLyKHmR2PIkdy4cfPnRn7kxo0bt9A3l/xYLp9E//Of/9Rdd92lSZMmqXv37hozZoz69u2rFStWKCUlxfOxCQkJkqT169crMTGxPKZX7bz94dfWPoOu6B3xOHUS2lr79D7Te5ydO3dZY9ie16mv/s0aA/CSlZWlFi1aBPNJVRJJfpTIkaG89cEia59rrzwj4nHqJbWx9jmjdx/P9u3bd1ljNEhM8mxPf+0pawwgnMMjP0qe6dH7JeYmM+IOSkrynkimPYQsIZzm4U8Xl53qHcW+LfZ5JGXagzh0sW5NpsOTY31+nXa8ZSYu22Kdqx8vCElJDtsT+SAOfSzz8GrOypIc82O5LKJHjRqlG264Qddee60kadKkSfrggw/00ksv6d577/V8bPHXbxITEzlB/K+6detVyDiBQLS1T0xMrGd7rVq1HWJ49+F5h1+q4tf5IsmPEjkylLp14ytknEDA/tVXciSqi5qdHy2LaD9Y40c+AX+2wSGIP10c+LBPfBjCj23xIw8nVtxkfQjiNFAFjWNjmYfDNF3yo+8/htm7d68WL16s3r3/94llVFSUevfurfnz55fqn5+fr6ysrBI3AKiJypofJXIkgMMD+RFAdeL7Inrbtm0qLCxUampqiftTU1O1efPmUv1HjhyppKSk4K1FixZ+TwkAqoSy5keJHAng8EB+BFCdVPplGYcNG6bMzMzgbf369ZU9JQCoMsiRABAa+RFAZfH9N9HJycmKjo5WRkZGifszMjLUpEmTUv1jY2MVG+v9GzIAqAnKmh8lciSAwwP5EUB14vsn0bVr11bXrl01a9as4H1FRUWaNWuWevbs6fdwAFBtkB8BIDTyI4DqpFyuzn3XXXdp0KBB+sMf/qBu3bppzJgx2r17d/BqiyibpUuX+hLniNbdPdvvuPMOa4zPP/vcs339unXWGHNWr/Zsf336JdYYV11yqrUPSsottPepa79AOyJEfvTfsiXf+BKnRZr3ifrtt91hjfHFfO8cuc4hR8615MjX/nmRNcbVl/Wy9kFJOfvsfeLL5awJxXzLjz5V7ImMH1c/NxGPY1zmEXAYx9bFjxgOrFvjstuNfSL2/WYfyFjGcbpCvrWLw061jGOb5/4QDnO1HiMOIXw5RmyvifCDZMk9fZTLn4PLLrtMW7du1YMPPqjNmzerS5cumjlzZqmLRQDA4Yb8CAChkR8BVBfl9p7q0KFDNXTo0PIKDwDVFvkRAEIjPwKoDir96twAAAAAAFQXLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAAR1Q8rAYeeeAma59/Tvuntc8vltqj87+cb41xRPPmnu2/7/p7a4wOHTp4tl98fg9rDJRdTaoB/c0PWz3bc3KyK2gmqApGPHSztc+/pr1h7fPzWu8c+fmXn1ljNG3R1LP92s4OObKTd44ccG43awyUXU2qAb3kp+1h2w6H/JikTEmJYdtttWhdauJ61Zr1ix+Vpp0K7/pRrtiH2TrtU+tzZw/hNFNLJ7d6xpHXZ7YPURHPjBtjmYvLPFyev4j59NLlk2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwFGtyp4A/PHzD7OtfXqcdLFne3xCvDXGNYOu8Wzv3qWZNQYQqd93bOzZnpUVW0EzQXWx/KdPrH26nXiRZ3tCQoI1xjWDrvNs79GpiTUGEKkuxzQK25aVFVOBM6kkmZISwzcHrAFMxFOwj2HvZRzmYR0n4DaTyAdy4b09AadnxjtG5M9c1RLw6/mLmMOeNbbj2c66ucYhSgXtMj6JBgAAAADAEYtoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEXWiDyMvvfKyZ3uHtvYaqPDfj6uzrX0yMzM929u1a26NkczTC3h6ZbJ3jjymXVIFzQQH+ukX7/wnSTt2ePdpf0xLa4xG9ZynhKouyfJatZSaNQ61aG31e51q4jr0sbHO1aXOsMP2Gkscl22xxfCnkLA/rOWKHabhR8lj204xDhOpuFLTkR+LbvvEMox9z0c+iPgkGgAAAAAAZyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAARyyiAQAAAABwVKuyJ1DdrcvwLthdt669sPjSpT97to/62zPWGGefe461zznnePfJzLOGUFKcvQ/KpnXrBGuftWu921eu3GCNsa1BA2uf9mn1rH2Asvh1a5Fne3xd+3u53y5Z7tk+6qlR1hjnnX++tc/ZZ/X1bN+5xxpCDerY+6BsWqcl2TsVejev/HmdNcSWho2tfY5pxRNcI1hOzQK2DpKM8T7/cwghGZdOtmEij2HZEt/YZuqyO6xdbM+LUxAHDuMYy0Bu07DutYjn4TQTl/3qg4BlKhU0DSe+fxL98MMPKxAIlLi1b9/e72EAoNohPwJAaORHANVJuXwS3bFjR33yySf/G6QWH3gDgER+BIBwyI8AqotyyU61atVSkyZNyiM0AFRr5EcACI38CKC6KJcLi61cuVLNmjVTmzZtNHDgQK1bF/63SPn5+crKyipxA4Caqiz5USJHAjh8kB8BVBe+L6K7d++uyZMna+bMmZo4caLWrFmjk08+WdnZ2SH7jxw5UklJScFbixYt/J4SAFQJZc2PEjkSwOGB/AigOvF9Ed2vXz9dcskl6tSpk/r27asPP/xQu3bt0r/+9a+Q/YcNG6bMzMzgbf369X5PCQCqhLLmR4kcCeDwQH4EUJ2U+xUb6tevr6OOOkqrVq0K2R4bG6vY2NjyngYAVDm2/CiRIwEcnsiPAKqycl9E5+TkaPXq1br66qvLe6hKkZzsXdBs6VJ7/d4nRz7p2f7VwoXWGJs3bbb2scX5+wuPWWNUJ1syvdsbOJQejfFnKp7qRtv7dGhrqyVtrzX92zb7ON/8sNWzvXNHex1V2+a4zOOIZHufmqCm50dJapTs/YWnZUvs9Xv/9oR3jpz/pT1Hbtmyxdpn/pdferb/48XHrTGqk62Wn482SLTHqIhrJ9dxKGF6TDtbQrcn/PXb7eMsWebdqfOxjawxbJuz0X6oqlmKvU91F0l+zMzMVGKiwwEcEUsNYKeCtrY+DvWqI47gxlqt2GF7bTWAXQpW22pJO5Xndqo1HHlBYuv2+vHs2AfxJ4RTEW/LPnOprW2bjFNJ64opJu3717n//Oc/a+7cuVq7dq2+/PJLXXjhhYqOjtYVV1zh91AAUK2QHwEgNPIjgOrE9zeRN2zYoCuuuELbt29X48aNddJJJ2nBggVq3Nj+CRYA1GTkRwAIjfwIoDrxfRE9bdo0v0MCQI1AfgSA0MiPAKqTcqkTDQAAAABATcQiGgAAAAAARyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDk+9W5Dzdr12Z7tufn5VljbN+2zbM9OTnZGiMxKcna54gjmnu2x1kj2L0+fa61z1WXnOrDSHYpll1SWCGzsPtxtfcx5KJD2wRrnyb2w0gNGniXEvn2+wxrjHbtUj3bly//2Rqjzu+P8mxvWNcaAlXE2jU5nu35ufnWGBkZOz3bU5K9jzlJio+Pt/Y5olVLz/ba1gh2U974zNrnyotP9mEku8aJ3u37KmQWdj/9kmnvZEnoR7ez/41s3sg+TIMk707fLNtijdGmbYpn+4qVy60x6ia092yvX8caokazPtsBS7uxj2FsnQK2QSQZWwx7iIBlHg6booAfc60g1pn6tS2WOA6jyG3vRxrC6SDxYRr2bQlYBnKLYZmHcdkY73F82B2S+CQaAAAAAABnLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAARyyiAQAAAABwxCIaAAAAAABH1ImOUDtLfd7Vq1dbY9hqPGfn2OsIZ2Xa62hmOvSJVGqKd/1LScou8G5PiPFpMhb5DoWi60aX/zxcajzbqo2vy7DX3isosOx4SS2be1fCbdPWXo/XNozLcZiV6f3kNKyIJwa+aN/Guz7zhw510pOSvGPkZNuPqZwc73rVkpRTATmyUbK9GHGWpUBzYgX95S5wKBRdqwLmckwbe43nPUXe7es3WzpIKiqy92nWzHuD21pqQEvSXkuO3LHDfhzuyPSea/06h/dnJJlKklcZdHud2IqpiWwreRtwK/Ls3ewyDx9qQEdevdevetUOtYgdxrHtEx/KFfuyz5y217fKyBXBUuPZ6RixtLvEcHB4Z1kAAAAAAMqARTQAAAAAAI5YRAMAAAAA4IhFNAAAAAAAjlhEAwAAAADgiEU0AAAAAACOWEQDAAAAAOCIRTQAAAAAAI5qVfYEqrsYS/vOHTutMcaMG+PZPuG58dYYCxcutPaJi4u19olUg4YNrX1+27Dbs719Wj1f5vLND1s925OTk60xWqZWjQL1cZb2FId55ubWtvb5bVOhZ3t2drY1RkJCgmd7UaH3GJIUGxtt7YPqwfZHZsc2e44c99yznu3PjhtnjfHVgq+sfWrF2l5pkUtu0MDa57d1OZ7t8W3irTFc3iFf8tN2z3aXuTZvUjXei69jmUbTFPs8c/PtfTb/VuTZnrkn0xojPj7Ju0PAns/j4qrGfq+qkiy7WDLezQ7PgYx3DMsIbp2cpuEdJOCwLS59bNvrtM8sG+yyz6xztc1T9n3mxmG/Rh5CAet+t8ew7VgTcNrzDsNYjkW3AzqiMfaPUzHIwgAAAAAAOGIRDQAAAACAIxbRAAAAAAA4YhENAAAAAIAjFtEAAAAAADhiEQ0AAAAAgCMW0QAAAAAAOGIRDQAAAACAo1qVPYGabsuWDGufzu0berY3Sk62xkhKSrL2yczM9Gy/7c9PWWOM+9s9nu0FBQXWGD/++KNne8OGx1tjuLCN0717d4cocb7MJVKrN+z1bI+OjrbGqFvX3qd1U+8+uSn1rTHWrs32bP/mm2+tMdatW+/ZHhMT49m+Z0+udQxUDVsytlj7dDq6gWd7aqo9RzZoaM+RWZYcOfSuJ60xnhv1F892lxy5YvkPnu3JDey5q9DaQ/r+u+8923ue0NMhSqxDn/K3ZuM+z/YoY6wx6tb1ziuS1OoI788ecvZ5H6uStHaN93H27dffWGOsX7fGsz0mKvy2HA75MVNSolcHy+EQsHVwEHDoY5x6Rcbh0Hfiy0xtc3EYxLo9DjHctsW7l3HZsdaB/JisPYbxY8f7cQA4xbB1su93Yw59slmS7GcL+5X5k+h58+bpvPPOU7NmzRQIBPT222+XaDfG6MEHH1TTpk1Vp04d9e7dWytXrizrMABQ7ZAfASA08iOAmqTMi+jdu3erc+fOGj9+fMj2p556SuPGjdOkSZO0cOFC1atXT3379lVeXl7EkwWAqoz8CAChkR8B1CRl/jp3v3791K9fv5BtxhiNGTNG999/vy644AJJ0quvvqrU1FS9/fbbuvzyyyObLQBUYeRHAAiN/AigJvH1wmJr1qzR5s2b1bt37+B9SUlJ6t69u+bPnx/yMfn5+crKyipxA4Ca5lDyo0SOBFDzkR8BVDe+LqI3b94sSUpNTS1xf2pqarDtYCNHjlRSUlLw1qJFCz+nBABVwqHkR4kcCaDmIz8CqG4qvcTVsGHDlJmZGbytX+99VV4AOJyQIwEgNPIjgMri6yK6SZMmkqSMjJJlnTIyMoJtB4uNjVViYmKJGwDUNIeSHyVyJICaj/wIoLrxtU50WlqamjRpolmzZqlLly6SpKysLC1cuFC33HKLn0NVGxkONVBt153Mz7dfmTI31173cdYnsyKO8bvf/96zPa11a2uM3zb85tmel/cHa4ycnBxrH1vtbJftrSp1om3WrVtn7RMba6/nujw/37O9tcPzuyXDuzb6yOF/tMaoiciPoW3Z4n28SJJ3lXQpb7f3cStJe7IdcuRH//Fsz821j/N7S45se2Rba4x167xz5O+Ot4ZQ9i7vWsSS1DDJu6ZxrsM+qyp1om2Vsdf9tsEaIS6mnrXPD8t3e7Yf6fD8btnifV7w2IjDLx/4nR+TMjMlrwV1wFJH1qkEsB8FmH2o32vjWynqyOv3Biz73aX2si2G09NiiyGH2so+1LT24xhyK4nsxz5zeG58qK1te35tY0gO+8RrGmUoFF3mRXROTo5WrVoV/P+aNWu0ZMkSNWzYUC1bttQdd9yhRx99VO3atVNaWpoeeOABNWvWTP379y/rUABQrZAfASA08iOAmqTMi+hFixbptNNOC/7/rrvukiQNGjRIkydP1j333KPdu3frxhtv1K5du3TSSSdp5syZiourHp/oAcChIj8CQGjkRwA1SZkX0b169fL8OD4QCGjEiBEaMWJERBMDgOqG/AgAoZEfAdQklX51bgAAAAAAqgsW0QAAAAAAOGIRDQAAAACAIxbRAAAAAAA4YhENAAAAAICjMl+dGyXlFnq3r/z5Z2uM0c+85tm+dMlSa4z8vHxrnw4dOni25+TkWGM0bdrUs73XCUdaY5xs6bNyzW5rjC0ZGdY++Xl53uM4PDed2/ew9qkIbZvXtrSnWWN474391q/z7lVUVOQQJXLdTxzg2b7wi3ctEYykfb7NB4duT/iL8UqSVv680hpj1KhXPduXLllmjbE73/4K6NDRO0dmZ9tzZPPmR3i2n9qjjTXGiZY+v/yyxxpjy5Yt1j55+7z/bixfudwao1OHntY+FSGtWaylva01xl6H9LZmXa53jMICexAfHH/yxZ7tX3/2tkerkWQ5eTnMGVkSl4OAj70ii2HflkCFzEOeV2d3iyDZtsc4BHEax7LbAgGH7fVjt1oHcTlWvScSCNhjRP6K8Oc4sx1D+8eJvIcLPokGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAcsYgGAAAAAMBRla0TnSfJqzLugi9XeT6+TVt7TciWqZHXCYuN9m5PbpxsjfHp7Dme7Wf1O8saIyYmxtqnc+fOnu0dOtprPKckWbtYWXaZcnO963BK0vLlK6x91q1b59neOq21NUZV8ePqbM92l+e/QYM4ax9bnMQk27MnxSckWPvY/G30M57tl1+6wbO9qGifNq1bHPE8qrK9/72FM3+Bd/3ltg45snnjyN9njbGk2eRke46c9cmnnu1n9utnn0eMfVt+3/X3nu1Hd7DXeG6SaO1iZfvDnJvrnQ8kaaVDjef1lhzZooW9/nxV8dMv3jW8XXJkYpJ3rWlJiovzzqNJSfbjbGd8vLWPzZjRozzbL7kk/HNbVFSozb/W7PyYKcnzpWgtNVsRBX5d6vM6VLytmKn6VOPZj3lYOrjUb3YqeuzDFtkG8mEIp3rVlnm47A8/xnFjGceHefh1rPJJNAAAAAAAjlhEAwAAAADgiEU0AAAAAACOWEQDAAAAAOCIRTQAAAAAAI5YRAMAAAAA4IhFNAAAAAAAjlhEAwAAAADgqFZlTyCcyS+9rzp16oZt79Cxg+fjX33lVesYF/S/wLO9YcP61hhHJHu3/2PCw9YYfpj/zQZrn56/b14BM4ncli1brH2+//57a599BQWe7W3btnWeU2Vr0CDBsz03d681Rm6ud/F5Sdqwwfs4Si1ItcbYuWOHtY/N6GdGebb/tnapJYJ9W6u7l178QHFeObLD0Z6Pf/mlV6xjXDCgv2d7coMG1hjNGnm3vzhpuDVGkaXd5d3gLxats/Y58Q8tHSJVvk0ZGdY+y5bYc2SBJUe2an2U85wqW8OkeM/27Px8a4y99i5a99tvnu1N81KsMbbt3GkfyGLUk095tm9cs8Sjtebnx6SkChjEshuNse/nQMCPGN5BHEJY5+HC5aiyzdWPY9NlU4wP2+syV1+GsURx2mOWiQQcdojTsWiL4bBDfDlCbK8Jj7YsSa7pg0+iAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAARyyiAQAAAABwVGXrRPfvf64SEhPDtjdt6P34lBR7rcbjjqpfxllVXX7UgM7Ms/exldpM8aE2Y1FhYeRBJNWKifFsz83N9WWcimA73tWwti/jJCe38mzPzLTH2LRpc8TzeOuf4yKOUdNddOE5SkgInyNTLfWZmzZrZh2j45Hh41ckP97t9aMGdNYee59879LLauzHLi2yVc6WFG3vEhPtnSP35GY7Tqjy2Y73VMX6Mk79Bi0827OccuSWiOfx5vTnIo5Rk2VK8nqpBSzVZv2orRxwq1hsiVGNOEzWVmvYXkdasu0zpzrCfpRK92GuLjvNj31mPZ792JT9k7EM4xLE9pqomP3uosznJvPmzdN5552nZs2aKRAI6O233y7RPnjwYAUCgRK3s846y5fJAkBVRn4EgNDIjwBqkjIvonfv3q3OnTtr/PjxYfucddZZ2rRpU/A2derUiCYJANUB+REAQiM/AqhJyvx17n79+qlfv36efWJjY9WkSZNDnhQAVEfkRwAIjfwIoCYplwuLzZkzRykpKTr66KN1yy23aPv27WH75ufnKysrq8QNAGqqsuRHiRwJ4PBBfgRQXfi+iD7rrLP06quvatasWXryySc1d+5c9evXT4VhLhY1cuRIJSUlBW8tWnhftAMAqquy5keJHAng8EB+BFCd+H517ssvvzz47+OOO06dOnVS27ZtNWfOHJ1xxhml+g8bNkx33XVX8P9ZWVkkQQA1Ulnzo0SOBHB4ID8CqE7KvU50mzZtlJycrFWrVoVsj42NVWJiYokbABwObPlRIkcCODyRHwFUZeW+iN6wYYO2b9+upk2blvdQAFCtkB8BIDTyI4CqrMxf587JySnxruCaNWu0ZMkSNWzYUA0bNtTw4cM1YMAANWnSRKtXr9Y999yjI488Un379i3TOE0aSpG8oXjcUfUP/cGHqaQ4h04ufSK0Y+dOa5+YmBhrny1bMjzb8/PznOdUExQ49Cks8m7Pzw//27RiO3fucJtQDVRR+VGSUiLMkR2P5BObskqs49DJpU+Etu2w58golxyZ4X3RphYF+5znVBPkO/QJWFJgQYEliUrKytzmNqEapiLzY0UwCtg6WAUC3jGMQxDLLCqQfSaWzZXTTrN0sY/hyjJQ5FOtOs+dD8eqUxiHcfzYK9YIAa+JZElKchqnzIvoRYsW6bTTTgv+v/i3KIMGDdLEiRP13Xff6ZVXXtGuXbvUrFkz9enTR4888ohiY2PLOhQAVCvkRwAIjfwIoCYp8yK6V69eMib8Cv6jjz6KaEIAUF2RHwEgNPIjgJqk3H8TDQAAAABATcEiGgAAAAAARyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEdlLnEFeMnMizzG2jVrrX2yMjOtfdb/ut6zvV27o1ynVOkKLO2bMuwV7H9ZvdraZ/nyFZ7t33//vTXGv6b909oHOFzt3GPvEx3t3b5+7RprjJzMHGuf39at82xv17adNUZVsc/SvnlrkTXG6p9dcuRyz/ZlP9pz5D/JkeUvU1Ji+Q4RsLQb2f8u26P4oSLGcBvFo8LZ/hgBpygRtJatlxeXqQasw0Q+Dzfek62oWfhxLAbsO9XKeMwjS1KSYxw+iQYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBF1oiNkK4scVyGzqDiFlvZt2/ZaY2zZssWzfa1DDdSZ/55p7dOhYwfP9pTUFGuMTTu823Nz7dv722+/ebav/HmlNcbSpUs82xd9vcgaY/7nn1v7qGiTvQ+qhKL/3sKpKu+Q7rGU561TVSbqE1sFy+3b8q0xtm3b5tm+es1aa4yZMx1y5NEdPdubpiRbY2Ts8m7PzrZv76ZN3nnHVptZkpZ9v9SzffECe478wiVHarNDH1S2pCTXSq9h2AoaSzKWPgG3ysmOE4qEX/Wq/ag3bZuLD3WEfZiFSyCvWsOuI1WZY6RiSom7sdSB9mNv+LW5Nez0BQAAAACA8sMiGgAAAAAARyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAAR7UqewLhTHrpfdWpUzdse7uj2nk+/qzTOkY8hysH32Ptc8WVV3q2n9enS8Tz8MuPq7M928c/N94aY09urmf7tm3brDFW/rzSs335soXWGJL3trhISU2x9vn4o/94ts+bN88aY+eW5ZYeBdYYNcn/3fKQtU/Pnid4tkdFe7//tyd3t4bc0L8s06p2/jH5A88c2aZNG8/H9zvjuIjncNW1d1v7XHrlFZ7t55/5+4jn4ZeffsnxbH923DhrjPz8fM/2jM0Z1hi//PKLZ/tP37nkyCx7l0Lv5uTURtYQH8+a5dn+6exPrTF2bfP+myB579Oa5oabhlv79Di5p3cHE75pz57dGnrjhWWcVfWSmZmpxMTEsO2BQMCHUbxjeDwF/4tgbPNwiGK8+/izrS5z8WucyBiXfebDVF1CuBwDkXIbw3KMOGyN0/HsQwxfVNChyifRAAAAAAA4YhENAAAAAIAjFtEAAAAAADhiEQ0AAAAAgCMW0QAAAAAAOGIRDQAAAACAIxbRAAAAAAA4qrJ1or/99hvVrh0btt1Wr/OT2d61KyWpZap3obCdO3ZaY7jURa4IO7zLN0uShtwyxLN9zsdTHEayFBatQjat/9qzfeor3u0oH8nJydY+rdNae7Y3aNDAsz0nJ/I64lXdoq8XeefIsWM8H9/hU3uObNXY+33WrVt3WGPs3Lbd2qcibPcuAS1JGvpH7xw5+6PXHEaqsEqYEdu44SvP9mmvebejfDRKtefIlq1be7YnJzUM23Y45MekpCTPdktpZae6uX7UmvVlHtYx7DnJj1LSfmQ+pxrPFcUyFROoKnN1eH5tx5HLMeI6Ha8YTkEs9dd9mKvXcZalLCXJO38UK9Mn0SNHjtTxxx+vhIQEpaSkqH///lqxYkWJPnl5eRoyZIgaNWqk+Ph4DRgwQBkZGWUZBgCqHfIjAIRHjgRQk5RpET137lwNGTJECxYs0Mcff6yCggL16dNHu3fvDva588479d5772n69OmaO3euNm7cqIsuusj3iQNAVUJ+BIDwyJEAapIyfZ175syZJf4/efJkpaSkaPHixTrllFOUmZmpF198UVOmTNHpp58uSXr55Zd1zDHHaMGCBerRo4d/MweAKoT8CADhkSMB1CQRXVgsMzNTktSw4f7f3ixevFgFBQXq3bt3sE/79u3VsmVLzZ8/P2SM/Px8ZWVllbgBQHXnR36UyJEAaibOIQFUZ4e8iC4qKtIdd9yhE088Uccee6wkafPmzapdu7bq169fom9qaqo2b94cMs7IkSOVlJQUvLVo0eJQpwQAVYJf+VEiRwKoeTiHBFDdHfIiesiQIVq2bJmmTZsW0QSGDRumzMzM4G39+vURxQOAyuZXfpTIkQBqHs4hAVR3h1TiaujQoXr//fc1b948NW/ePHh/kyZNtHfvXu3atavEO4kZGRlq0qRJyFixsbGKjQ1fpgUAqhM/86NEjgRQs3AOCaAmKNMn0cYYDR06VDNmzNDs2bOVlpZWor1r166KiYnRrFn/qz+6YsUKrVu3Tj179vRnxgBQBZEfASA8ciSAmqRMn0QPGTJEU6ZM0TvvvKOEhITgb1SSkpJUp04dJSUl6f/+7/901113qWHDhkpMTNStt96qnj17lvmqimlpaYqLqxO2/b133/N8/Lx586xjXHzJqZb2S6wxmh9xhGf7ou/t9Q2XL1/u2X6VZZ6S9MBfn7D2mfPxa9Y+OLxFx7b0bC/MXxfxGIWFRdY+O3fs8GyPjvJ+/2/37uwyzckPFZkfJSkt7UjFRpIj59pz5GUX9/Jsv/hSe45s0tQ7Ry5etska48flKzzbr7bMU5Ievv9xa5/ZH71q7YPDWyCmtWe7KVgb8RgFRYXWPju2bfNsj4mODttWGflRqugcmSkpMWxrIBDwfrixj2CMdyfrGJKs03CYh3UU+zRkXDrZgzjwYYP9iOHw3FQXAZfnzo+DxIn3vnc5zgKWGE4zjeSFlSUpyWWQMi6iJ06cKEnq1atXiftffvllDR48WJI0evRoRUVFacCAAcrPz1ffvn01YcKEsgwDANUO+REAwiNHAqhJyrSItr3rJklxcXEaP368xo8ff8iTAoDqhvwIAOGRIwHUJBHViQYAAAAA4HDCIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAARyyiAQAAAABwVKarc1ekSy+7TAkJ4Wv8zf9yvufjly5Zah3DVn/5mqv7WGNsyvC+2mRBQYE1hq3W9LyFa6wxJox5xtoHsDn7nHM82997a6JDlLqerV8tXGiNkJPjXcc0Kip8DVRJ2pufZx2jurt0oC1Hfub5+GVLl1jHsNVfvuaas6wx/lsKNqyion3WGEc0zfFs/+yrtdYYz40dZe0D2Jx3ztme7e++7VKOKd6zdeEX3uc3kpSb7f2aiI4O/xnJ4ZAfLWWi7aWG/ZxLeY/kQ81rp1LDFVLT2o96xT7VPPYljG2nuNRNrhp8eU041QG3cDhGXCoB+IFPogEAAAAAcMQiGgAAAAAARyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAARwFTURWpHWVlZSkpKUkfz/1R9eITwvZLTk72jLN0yRLrWGvXrvVsz87OscZYsXy5Z/tyS7skLfv+e8/2wvx11hhAVZHUqKNne/v27a0x4hPiPdvz8/I92/ftK9CCz99UZmamEhMTreNVJ8U58qN5P3nnyAaWHLn0G+tY69es9WzPycm2xvjhJ+8cuGL5CmuMHy05cl/BemsMoKpo0PAYz/ajHXJkHY/XviTty88L37avQPO/mFGj86MyMyWvbQvYIlk7yI+TZ/sodrZ5OI3hsjG2QH7EcOI9kMuqJhCI/Pl126+WKA7z8COGsWyN29MS+ZNXUQvOSA7VLElJklN+5JNoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAc1arsCYTTtu0RSvCozzXn0wWej7/4/B4Oo3j32WYvgaoj0471bM/c/oPDPKqPXmde7dk+5+PXKmgmqKpsx/zCL7ZZYyQ3aerZ3rJlS8/2wsIC6xjV3VFHNvPMkbM+me/5+Esv6OkwinefLVn2CO3SOni2Z+34yWEe1cfpfa/xbJ/90asVNBNUVTstx/yCL+05smEz7xyZ1jItbFvhvpqfHzMleVV4DVgqyRqXYsMWLrWIbZymEbB0Mi7zcBjI2sWhXrEf+zXiWfgzD6fd6gsfBrIdIn4EkU/PjS2GU23tCDpkZUlJSfYxxCfRAAAAAAA4YxENAAAAAIAjFtEAAAAAADhiEQ0AAAAAgCMW0QAAAAAAOGIRDQAAAACAIxbRAAAAAAA4YhENAAAAAICjWpU9gXDWrNmkevE5Ydt7ntDD8/F5DmPEWdqTE+wxunXv7tn+8Yc/OMwk1dKe4RAjchdfcYe1zzWDB3m278nNtcZY+MWbrlNCjWQ/nrdt9u6zbfMSn+ZSfa1etVH14rPDtvc8qafn4/cU2ceoY3mbNSXRHqPnCd458qP3f7IHURNL+2aHGJG79PLbrX2uGXStZ3uuQ45c8NkbznNCTbTV2mPHRu8+OzZ+59dkqqWkpCTvDsZ4NgcCDoN4h7A1u3VymUfEg7jynozxY64uU7U8OcYliEMX2+a47VUfdop1Ii4bYwniEsOBH1Gse8ynufqhTJ9Ejxw5Uscff7wSEhKUkpKi/v37a8WKFSX69OrVS4FAoMTt5ptv9nXSAFDVkB8BIDxyJICapEyL6Llz52rIkCFasGCBPv74YxUUFKhPnz7avXt3iX433HCDNm3aFLw99dRTvk4aAKoa8iMAhEeOBFCTlOnr3DNnzizx/8mTJyslJUWLFy/WKaecEry/bt26atLE9vU7AKg5yI8AEB45EkBNEtGFxTIzMyVJDRs2LHF/enq6kpOTdeyxx2rYsGGevwHLz89XVlZWiRsAVHd+5EeJHAmgZuIcEkB1dsgXFisqKtIdd9yhE088Uccee2zw/iuvvFKtWrVSs2bN9N133+kvf/mLVqxYobfeeitknJEjR2r48OGHOg0AqHL8yo8SORJAzcM5JIDq7pAX0UOGDNGyZcv0+eefl7j/xhtvDP77uOOOU9OmTXXGGWdo9erVatu2bak4w4YN01133RX8f1ZWllq0aHGo0wKASudXfpTIkQBqHs4hAVR3h7SIHjp0qN5//33NmzdPzZs39+zb/b8loFatWhUyAcbGxio2NvZQpgEAVY6f+VEiRwKoWTiHBFATlGkRbYzRrbfeqhkzZmjOnDlKS0uzPmbJkiWSpKZNmx7SBAGgOiA/AkB45EgANUmZFtFDhgzRlClT9M477yghIUGbN2+WtL+ofZ06dbR69WpNmTJFZ599tho1aqTvvvtOd955p0455RR16tSpTBP7z0f/UWxcnbDt551/nufjCwtTrGO0TI28CPqWLRkRx5D8iBG5v/9jtLXPDdfd7tm+aNHXfk2nBknybK2bmGqNYDuBcDnBiE+It/axXeTqi4O+ehdKYf46a5+aqCLzoyR98tFHivPIkef2986RRQX2465Vk4iuPSlJ2pKxPeIY0mYfYkTu7y+Nsfa5ftAdnu0LFyz0ZzI1SgPP1jr1G1kjHNHU+xPNlBT7OUFSkneulqTs3BzP9gWfz7fG2Je31tqnJqrIHJkpKdGzh/f5n8vZoZGxdLC0uzAOMwlY+gRc5mEfx9i2x4dhArZtcZmHyzQcxrFuj8M0bKP4si0+PHd+7XfbXKyvmf9GiTRG5Ks7N2VaRE+cOFGS1KtXrxL3v/zyyxo8eLBq166tTz75RGPGjNHu3bvVokULDRgwQPfff79vEwaAqoj8CADhkSMB1CRl/jq3lxYtWmju3LkRTQgAqiPyIwCER44EUJNE/l09AAAAAAAOEyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEdlujp3VTLyscc92y+9/DJrjJb9T4h4Hscdd5xn+9JFH0Q8RkW5btBQa5933vinpUeRP5OpMhI8W5MatbRGSEn1rk+anJxsjVG3bl3P9tjY2IhjSFKDht71Wl1eV/l5+Z7teXl51hgxMTGe7UuXLvVsLyrcp3W/LLCOU61FR++/hWHLkRdfcrl1iFYXnVjmaR3smE7HeLZ/+/V7EY9RUQZdPcTa5903bTmy0J/JVBne1XgTGx5hjdC0aTPP9gYNvPOSJNVJ8M7VcTEOOTLBniOTGnrXkr7MIUfm7d7t2Z6f751DJYccuSR8jiwq2qe1a2p4vfLMTCnR69j0oZKsS61hWwhbyVuncsbeQYxTkIqpvWwtvexDaW0/9plk3x6X8tu2cXypz+xSStw616pTe9leGj3y48wvfBINAAAAAIAjFtEAAAAAADhiEQ0AAAAAgCMW0QAAAAAAOGIRDQAAAACAIxbRAAAAAAA4YhENAAAAAIAjFtEAAAAAADiqVdkTCCcjY7Nq144L275jx07Pxz824lHrGPl5d3u2D7z8NGuM+IR4a5/qYv78+dY+x/3ueM/27Jxsa4y1Kz93nlP5Srb2aNoizbO9dVpra4ykpCTP9rp161pjxMbFRhwjOira2qewqNCzPSYmxhojqoH3OC4xoqO939/r3qO7Z3te3h6NuH+BdZzqbOPG3zxz5M6d3jly5COPWMco2HevZ/sVl/ayxoiPr2/tU13M/9KeIzt1/YNne+auPdYYv66e5zyn8tXI2qNZ87ae7S1bt7TGaNCwgWd7bKx3/pOkuHr1PNtdcmSUJe9IUqCwyLs9yh7Dtr1RDjGiLPm82wknhG3Ly8vV8L8utI5Rndn+7hrj/XhLs1OQgEsM6xgOfWwD2Ta2IlmmEgjY52rbHJf9bhx6Wcfx5QmuInw6RIwlUMBhpxnb68qHHe85RlaWZMkfxfgkGgAAAAAARyyiAQAAAABwxCIaAAAAAABHLKIBAAAAAHDEIhoAAAAAAEcsogEAAAAAcMQiGgAAAAAAR1W2TrRNamqKZ/u+ggJrjDtuv92zfdDAbfaJFOVYOrjUGrPNNdchRuS2/PaNtU9BwXGe7Tu3fO/XdCIWU8e7xnPTpk2tMWx1r3Nz7c+NbRxbzVBJio0NXw9YkuIsdaQlt/rMNoWWGqmSlGPZZzk5edYYtrna6sTn59vHqO6io/ffwklJ8c6RBXn25/LO2+7wbL/mSoccWWirHZ9ojyHbXG152B9bN31r7VNU1NmzfXvGUr+mE7HYem0825s45MjMXZme7bm5+dYYRzT3zl9JSQ2tMepacmBsXe8cKklRUS417L3rMxcU7LXGyM31rhWeu9v+dyU2znt7tm3bEbYtP99eq7y6y8yUEl1SS1hVpbayQ01ca11dh2GMQ/3eCNvL0isSLjWgHcpRW2seu5Tfttc09qMQuJ2xzcNpW1wGqhqvG3utaX/G4ZNoAAAAAAAcsYgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHLGIBgAAAADAEYtoAAAAAAAcsYgGAAAAAMBRrcqeQHnJy8+z9snPy/fuULTJp9nUHDu3fF/uYyQ0aG/tEx+fYO1Tt27diOeSnJzs2Z6bm2uNsWPHTs/2pKQka4y4uFjP9piYGGsMlz72GPY+0dHe783l2V53kgoKCjzb8y2v770Or//qL/q/N6/28Kz5T1J+rmU/Fv5mjXG42Z6xtNzHSGzYwdonPiHe2ie2rndeibYcQ5KUmmrLkdnWGDt3Znm2J9VvaI0RG1vPs90l/0VF2T9XiI723icuMaJivGPExNhPzYqKijzb8/LC/206PPJjpqTE8h0iEHEHGWNsPVxnE9k8/BjHJUTAey72/SEFLNvjEsNpqrZxHKLYphKw7I/ikSIZY/9ADn0qJojDKBUxjj9jlOmT6IkTJ6pTp05KTExUYmKievbsqX//+9/B9ry8PA0ZMkSNGjVSfHy8BgwYoIyMDF8mCgBVGfkRAMIjRwKoScq0iG7evLmeeOIJLV68WIsWLdLpp5+uCy64QD/88IMk6c4779R7772n6dOna+7cudq4caMuuuiicpk4AFQl5EcACI8cCaAmKdPXuc8777wS/3/sscc0ceJELViwQM2bN9eLL76oKVOm6PTTT5ckvfzyyzrmmGO0YMEC9ejRI2TM/Px85ef/72uFWVneX+sCgKqoPPKjRI4EUDNwDgmgJjnkC4sVFhZq2rRp2r17t3r27KnFixeroKBAvXv3DvZp3769WrZsqfnz54eNM3LkSCUlJQVvLVq0ONQpAUCV4Fd+lMiRAGoeziEBVHdlXkR///33io+PV2xsrG6++WbNmDFDHTp00ObNm1W7dm3Vr1+/RP/U1FRt3rw5bLxhw4YpMzMzeFu/fn2ZNwIAqgK/86NEjgRQc3AOCaCmKPPVuY8++mgtWbJEmZmZeuONNzRo0CDNnTv3kCcQGxur2FjvK4QCQHXgd36UyJEAag7OIQHUFGVeRNeuXVtHHnmkJKlr1676+uuvNXbsWF122WXau3evdu3aVeKdxIyMDDVp0sS3CQNAVUV+BIDwyJEAaoqI60QXFRUpPz9fXbt2VUxMjGbNmqUBAwZIklasWKF169apZ8+eZY6bm7tH+/Z510H0EhcbZ+1jq8+bm+Vd/3K/bY4zQrG2R5/q2X50+6OtMWy1OqX9v7nykpVpvwBJLUttUVu9YklKsNRrjbLUVZbstZWjouz7w6XGs60+c2Gh/TVpq8fqUr/btl9t86wqyis/SlJObo5i9h36fqgVZ0//iZYcmZPdyGGk7Y4zQrEjj+nl2d6uXXtrjJgYe14pKPI+fjIzM60x4mK8/9a61COvU6eOdweHOqj5+7xzRnSRQ/1mhxrPttxTULDXGiMuznuf1a4b+d+EAo860lEOfz8rSnnlyCR55y5bjV+XWrW2GE6Vly3DOM3DUijYj0rTLvyovOsyV+tz5zARp9rKFr4cIw7zcKslbZmHHxtcQez1tx32RwT1yLOysqzrw2JlWkQPGzZM/fr1U8uWLZWdna0pU6Zozpw5+uijj5SUlKT/+7//01133aWGDRsqMTFRt956q3r27Ol55VkAqAnIjwAQHjkSQE1SpkX0li1bdM0112jTpk1KSkpSp06d9NFHH+nMM8+UJI0ePVpRUVEaMGCA8vPz1bdvX02YMKFcJg4AVQn5EQDCI0cCqEnKtIh+8cUXPdvj4uI0fvx4jR8/PqJJAUB1Q34EgPDIkQBqkkOuEw0AAAAAwOGGRTQAAAAAAI5YRAMAAAAA4IhFNAAAAAAAjiKuE+234tpdBQXeNRD3WWo17nOon1pU5F1HWDr0OtUIr6hwn2e7S53NoqLI60S7HCMKeNerc4lh2569e+11VG1V8aIdak0HHI7nAsv2FBbaaw0WFXk/vy7y8yN7/e/du79mbHWqjegqmCMtx03RXu/nmxxZdfmRI13eI7fViXY5RvYFvMepqBxZq5b3PKKj7bVFC/fZc1eR8T7mXWrYB2x1UK2vOyk/33ufedWJPhzyo7K8+2XZOjjwI4Z1oi4RsnyYhx8xfODLtjiNUyHD+HSMRK7i9msVOZ7tg1jHd8mPAVPFsuiGDRvUokWLyp4GgBpg/fr1at68eWVPw1fkSAB+ID8CQGgu+bHKLaKLioq0ceNGJSQkKBDY/85xVlaWWrRoofXr1ysxMbGSZ1gzsE/LB/vVf4eyT40xys7OVrNmzRQVVbN+tXJwjuSY8x/7tHywX/1HfiyJ/Fgx2K/+Y5+Wj7Lu17Lkxyr3de6oqKiwK//ExEQOLJ+xT8sH+9V/Zd2nSUlJ5TibyhMuR3LM+Y99Wj7Yr/4jP+5HfqxY7Ff/sU/LR1n2q2t+rFlvQQIAAAAAUI5YRAMAAAAA4KhaLKJjY2P10EMPKTY2trKnUmOwT8sH+9V/7FNv7B//sU/LB/vVf+xTb+yf8sF+9R/7tHyU536tchcWAwAAAACgqqoWn0QDAAAAAFAVsIgGAAAAAMARi2gAAAAAAByxiAYAAAAAwBGLaAAAAAAAHFX5RfT48ePVunVrxcXFqXv37vrqq68qe0rVyrx583TeeeepWbNmCgQCevvtt0u0G2P04IMPqmnTpqpTp4569+6tlStXVs5kq4mRI0fq+OOPV0JCglJSUtS/f3+tWLGiRJ+8vDwNGTJEjRo1Unx8vAYMGKCMjIxKmnHVN3HiRHXq1EmJiYlKTExUz5499e9//zvYzv4MjfwYGfKj/8iP5YMceWjIkYeO/Og/8mP5qKz8WKUX0f/85z9111136aGHHtI333yjzp07q2/fvtqyZUtlT63a2L17tzp37qzx48eHbH/qqac0btw4TZo0SQsXLlS9evXUt29f5eXlVfBMq4+5c+dqyJAhWrBggT7++GMVFBSoT58+2r17d7DPnXfeqffee0/Tp0/X3LlztXHjRl100UWVOOuqrXnz5nriiSe0ePFiLVq0SKeffrouuOAC/fDDD5LYn6GQHyNHfvQf+bF8kCPLjhwZGfKj/8iP5aPS8qOpwrp162aGDBkS/H9hYaFp1qyZGTlyZCXOqvqSZGbMmBH8f1FRkWnSpIl5+umng/ft2rXLxMbGmqlTp1bCDKunLVu2GElm7ty5xpj9+zAmJsZMnz492Oenn34yksz8+fMra5rVToMGDcw//vEP9mcY5Ed/kR/LB/mx/JAjvZEj/UN+LB/kx/JTEfmxyn4SvXfvXi1evFi9e/cO3hcVFaXevXtr/vz5lTizmmPNmjXavHlziX2clJSk7t27s4/LIDMzU5LUsGFDSdLixYtVUFBQYr+2b99eLVu2ZL86KCws1LRp07R792717NmT/RkC+bH8kR/9QX70HznSjhxZvsiP/iA/+q8i82OtSCdbXrZt26bCwkKlpqaWuD81NVXLly+vpFnVLJs3b5akkPu4uA3eioqKdMcdd+jEE0/UscceK2n/fq1du7bq169foi/71dv333+vnj17Ki8vT/Hx8ZoxY4Y6dOigJUuWsD8PQn4sf+THyJEf/UWOdEeOLF/kx8iRH/1VGfmxyi6igepgyJAhWrZsmT7//PPKnkq1d/TRR2vJkiXKzMzUG2+8oUGDBmnu3LmVPS0Ah4j86C9yJFBzkB/9VRn5scp+nTs5OVnR0dGlrp6WkZGhJk2aVNKsapbi/cg+PjRDhw7V+++/r08//VTNmzcP3t+kSRPt3btXu3btKtGf/eqtdu3aOvLII9W1a1eNHDlSnTt31tixY9mfIZAfyx/5MTLkR/+RI92RI8sX+TEy5Ef/VUZ+rLKL6Nq1a6tr166aNWtW8L6ioiLNmjVLPXv2rMSZ1RxpaWlq0qRJiX2clZWlhQsXso89GGM0dOhQzZgxQ7Nnz1ZaWlqJ9q5duyomJqbEfl2xYoXWrVvHfi2DoqIi5efnsz9DID+WP/LjoSE/VhxyZHjkyPJFfjw05MeKUyH5MaLLkpWzadOmmdjYWDN58mTz448/mhtvvNHUr1/fbN68ubKnVm1kZ2ebb7/91nz77bdGkhk1apT59ttvza+//mqMMeaJJ54w9evXN++884757rvvzAUXXGDS0tLMnj17KnnmVdctt9xikpKSzJw5c8ymTZuCt9zc3GCfm2++2bRs2dLMnj3bLFq0yPTs2dP07NmzEmddtd17771m7ty5Zs2aNea7774z9957rwkEAuY///mPMYb9GQr5MXLkR/+RH8sHObLsyJGRIT/6j/xYPiorP1bpRbQxxjz77LOmZcuWpnbt2qZbt25mwYIFlT2lauXTTz81kkrdBg0aZIzZX6bggQceMKmpqSY2NtacccYZZsWKFZU76Sou1P6UZF5++eVgnz179pg//vGPpkGDBqZu3brmwgsvNJs2baq8SVdx1113nWnVqpWpXbu2ady4sTnjjDOCyc8Y9mc45MfIkB/9R34sH+TIQ0OOPHTkR/+RH8tHZeXHgDHGRPZZNgAAAAAAh4cq+5toAAAAAACqGhbRAAAAAAA4YhENAAAAAIAjFtEAAAAAADhiEQ0AAAAAgCMW0QAAAAAAOGIRDQAAAACAIxbRAAAAAAA4YhENAAAAAIAjFtEAAAAAADhiEQ0AAAAAgKP/Bw35wF+MNDeuAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "######################################################################################################\n",
    "#visualize the selected samples, their corresponding perturbed image, and the magnified perturbation #\n",
    "#also print the ground truth and the predicted label for each sample                                 #\n",
    "######################################################################################################\n",
    "\n",
    "def denormalize(tensor):\n",
    "    return tensor / 2 + 0.5\n",
    "\n",
    "def to_cpu(tensor):\n",
    "    return tensor.detach().cpu()\n",
    "\n",
    "\n",
    "for image, label, index in samples:\n",
    "    image = image.to(device)\n",
    "    label = torch.tensor(label).unsqueeze(0).to(device)\n",
    "    initial_epsilon, delta, out_adversarial, x_adversarial =  generate_adversarial_FGSM (net, image.unsqueeze(0), label, initial_epsilon=1/255)\n",
    "    adversarial_label = torch.argmax(out_adversarial)\n",
    "\n",
    "    image, label, delta, adversarial_label, x_adversarial = to_cpu(image), to_cpu(label).squeeze(), to_cpu(delta).squeeze(), to_cpu(adversarial_label).squeeze(), to_cpu(x_adversarial).squeeze()\n",
    "\n",
    "    fig, axs = plt.subplots(1, 3, figsize=(12, 4))\n",
    "\n",
    "\n",
    "    image = denormalize(image)\n",
    "    npimg = image.numpy()\n",
    "    axs[0].imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    axs[0].set_title(f'UNPERTURBED DATA, LABEL={classes[label]}')\n",
    "\n",
    "    adversarial_image =  denormalize(x_adversarial)\n",
    "    npadvimg = adversarial_image.numpy()\n",
    "    axs[1].imshow(np.transpose(npadvimg, (1, 2, 0)))\n",
    "    axs[1].set_title(f'ADVERSARIAL, LABEL={classes[adversarial_label]}')\n",
    "\n",
    "    added_noise = 1 / initial_epsilon * delta\n",
    "    added_noise = denormalize(added_noise)\n",
    "    np_added_noise = added_noise.numpy()\n",
    "    axs[2].imshow(np.transpose(np_added_noise, (1, 2, 0)))\n",
    "    axs[2].set_title(f'DELTA * {round(1 / initial_epsilon, 2)}')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EfkA5WJp3TGQ"
   },
   "source": [
    "## **Targeted FGSM Attack**\n",
    "implement targeted FGSM attack."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "tjuJZ0UOAiT1"
   },
   "outputs": [],
   "source": [
    "########################################\n",
    "#define tergeted FGSM attack           #\n",
    "########################################\n",
    "\n",
    "def FGSM_targeted (model, x, y, y_targ, epsilon):\n",
    "  \"\"\"\n",
    "  inputs:\n",
    "    model: model to be attacked\n",
    "    x: input image\n",
    "    y: label\n",
    "    y_targ = target label\n",
    "    epsilon: clipping threshold for FGSM\n",
    "\n",
    "  outputs:\n",
    "    x': the perturbed image\n",
    "    delta: the added perturbation\n",
    "  \"\"\"\n",
    "\n",
    "  delta = None\n",
    "  model.eval()\n",
    "\n",
    "  x = torch.tensor(x, requires_grad=True).to(device)\n",
    "  loss = F.cross_entropy(model(x), y_targ)\n",
    "  model.zero_grad()\n",
    "  loss.backward()\n",
    "  delta = epsilon * x.grad.detach().sign()\n",
    "  return x-delta, delta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2DTwVtm13TGZ"
   },
   "source": [
    "### **Evaluation**\n",
    "for each sample, select the most probable lable as the target label (if the most probable label is the ground truth, select the second most probable label). For $\\epsilon \\in \\{\\frac{1}{255}, \\frac{4}{255}, \\frac{8}{255},\\}$ report the performance of the model on targeted FGSM attack on the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3FtJ8Gv_lQpv",
    "outputId": "d7d74c9a-4099-48d6-ea59-892f2c00af4a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:560: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
      "  warnings.warn(_create_warning_msg(\n",
      "<ipython-input-16-bf5eff28c387>:22: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  x = torch.tensor(x, requires_grad=True).to(device)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EPSILON: 0.00392156862745098, TEST LOSS: 4.149052673578263, TEST ACC: 75.06\n",
      "EPSILON: 0.01568627450980392, TEST LOSS: 2.3437400996685027, TEST ACC: 50.56\n",
      "EPSILON: 0.03137254901960784, TEST LOSS: 1.2129275500774384, TEST ACC: 27.98\n"
     ]
    }
   ],
   "source": [
    "def get_target_label_most(outputs, labels):\n",
    "    _, predicted = torch.max(outputs, 1)\n",
    "    correct = (predicted == labels).float()\n",
    "    second_most_probable = torch.argsort(outputs, dim=1)[:, -2]\n",
    "    target = correct * second_most_probable + (1 - correct) * predicted\n",
    "    return target.long()\n",
    "\n",
    "\n",
    "epsilons = [1/255, 4/255, 8/255]\n",
    "loss_fn = nn.CrossEntropyLoss()\n",
    "net.eval()\n",
    "\n",
    "\n",
    "fgsm_targeted_test_losses = []\n",
    "fgsm_targeted_test_accuracy = []\n",
    "\n",
    "for epsilon in epsilons:\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    for images, labels in test_dataloader:\n",
    "        images = images.to(device)\n",
    "        labels = labels.to(device)\n",
    "        probabilities = net(images)\n",
    "        targets = get_target_label_most(probabilities, labels)\n",
    "        x_adversarial, delta = FGSM_targeted(net, images, labels, targets, epsilon)\n",
    "        out_adversarial = net(x_adversarial)\n",
    "\n",
    "        loss = loss_fn(out_adversarial, targets)\n",
    "        test_loss += loss.item()\n",
    "        predictions = torch.argmax(out_adversarial, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predictions == labels).sum().item()\n",
    "\n",
    "    fgsm_targeted_test_losses.append(test_loss / len(test_dataloader))\n",
    "    fgsm_targeted_test_accuracy.append(100 * correct / total)\n",
    "\n",
    "    print(f'EPSILON: {epsilon}, TEST LOSS: {test_loss / len(test_dataloader)}, TEST ACC: {round(100 * correct / total, 2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vGZeCSmI3TGa"
   },
   "source": [
    "for each sample, select the least probable lable as the target label (if the least probable label is the ground truth, select the second most probable label). For $\\epsilon \\in \\{\\frac{1}{255}, \\frac{4}{255}, \\frac{8}{255},\\}$ report the performance of the model on targeted FGSM attack on the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "MHaLPlKsVgbK",
    "outputId": "08c88845-db0b-4c53-c246-a9464f202d74"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-16-bf5eff28c387>:22: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  x = torch.tensor(x, requires_grad=True).to(device)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EPSILON: 0.00392156862745098, TEST LOSS: 14.847595167160033, TEST ACC: 82.26\n",
      "EPSILON: 0.01568627450980392, TEST LOSS: 12.047733521461486, TEST ACC: 71.51\n",
      "EPSILON: 0.03137254901960784, TEST LOSS: 9.572825932502747, TEST ACC: 50.48\n"
     ]
    }
   ],
   "source": [
    "######################\n",
    "# Your code goes here\n",
    "######################\n",
    "# Your code goes here\n",
    "\n",
    "def get_target_label_least(outputs, labels):\n",
    "    _, predicted = torch.min(outputs, 1)\n",
    "    correct = (predicted == labels).float()\n",
    "    second_least_probable = torch.argsort(-1 * outputs, dim=1)[:, -2]\n",
    "    target = correct * second_least_probable + (1 - correct) * predicted\n",
    "    return target.long()\n",
    "\n",
    "\n",
    "epsilons = [1/255, 4/255, 8/255]\n",
    "loss_fn = nn.CrossEntropyLoss()\n",
    "net.eval()\n",
    "\n",
    "\n",
    "fgsm_targeted_test_losses_least = []\n",
    "fgsm_targeted_test_accuracy_least = []\n",
    "\n",
    "for epsilon in epsilons:\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    for images, labels in test_dataloader:\n",
    "        images = images.to(device)\n",
    "        labels = labels.to(device)\n",
    "        probabilities = net(images)\n",
    "        targets = get_target_label_least(probabilities, labels)\n",
    "        x_adversarial, delta = FGSM_targeted(net, images, labels, targets, epsilon)\n",
    "        out_adversarial = net(x_adversarial)\n",
    "\n",
    "        loss = loss_fn(out_adversarial, targets)\n",
    "        test_loss += loss.item()\n",
    "        predictions = torch.argmax(out_adversarial, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predictions == labels).sum().item()\n",
    "\n",
    "    fgsm_targeted_test_losses_least.append(test_loss / len(test_dataloader))\n",
    "    fgsm_targeted_test_accuracy_least.append(100 * correct / total)\n",
    "    print(f'EPSILON: {epsilon}, TEST LOSS: {test_loss / len(test_dataloader)}, TEST ACC: {round(100 * correct / total, 2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UDD6c-f7lT6g"
   },
   "source": [
    "plot accuracy of model with respect to $\\epsilon$ both for least probable and most probable target labels (both in a single plot)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 430
    },
    "id": "a_REC3Jhlm2S",
    "outputId": "50b8a96c-b0a9-473d-ccc2-d802da5cb0ad"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "######################\n",
    "# Your code goes here\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "plt.plot(epsilons, fgsm_targeted_test_accuracy_least)\n",
    "plt.plot(epsilons, fgsm_targeted_test_accuracy)\n",
    "\n",
    "ax.legend(['least probable', 'most probable'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mGpFIrId3TGa"
   },
   "source": [
    "# **PGD Attack**\n",
    "\n",
    "Implement PGD attack in the following cell. There is no need to change the default $\\epsilon$ and $\\alpha$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "id": "5A7fFCvWty0I"
   },
   "outputs": [],
   "source": [
    "def PGD(model, x, y, iterations, epsilon=8/255, alpha=3/255):\n",
    "  \"\"\"\n",
    "  inputs:\n",
    "    model: model to be attacked\n",
    "    x: input image\n",
    "    y: label\n",
    "    iterations: number of iterations for PGD attack\n",
    "    epsilon: clipping threshold for PGD\n",
    "    alpha: step size for PGD attack\n",
    "\n",
    "  outputs:\n",
    "    x': the perturbed image\n",
    "    delta: the added perturbation\n",
    "    y': prediction of the model for the perturbed input\n",
    "  \"\"\"\n",
    "  init_x = x.clone().detach()\n",
    "  x = torch.tensor(x, requires_grad=True).to(device)\n",
    "  for i in range(iterations):\n",
    "      x.requires_grad = True\n",
    "      predictions = model(x)\n",
    "      loss = F.cross_entropy(predictions, y)\n",
    "      model.zero_grad()\n",
    "      loss.backward()\n",
    "      delta = alpha * x.grad.detach().sign()\n",
    "      x = x.detach() + delta.detach()\n",
    "      x = torch.min(torch.max(x, init_x - epsilon), init_x + epsilon)\n",
    "      x = torch.clamp(x, 0, 1)\n",
    "  return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "r00Fb6Am3TGb"
   },
   "source": [
    "## **Evaluation**\n",
    "for iterations $\\in \\{3, 7, 12\\}$, evaluate performance of the model on adversarial test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "WtE-CjiQaC06",
    "outputId": "f821b64f-1fd9-4251-f0ea-0dc0400390c5"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:560: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
      "  warnings.warn(_create_warning_msg(\n",
      "<ipython-input-86-981a188c86ab>:17: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  x = torch.tensor(x, requires_grad=True).to(device)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ITERATIONS: 3, TEST ACC: 22.85\n",
      "ITERATIONS: 7, TEST ACC: 15.69\n",
      "ITERATIONS: 12, TEST ACC: 15.18\n"
     ]
    }
   ],
   "source": [
    "######################\n",
    "# Your code goes here\n",
    "# Define the number of iterations to use for PGD attack\n",
    "\n",
    "iterations_list = [3, 7, 12]\n",
    "net.eval()\n",
    "for iterations in iterations_list:\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    for images, labels in test_dataloader:\n",
    "        images, labels = images.to(device), labels.to(device)\n",
    "        x_adversarial = PGD(net, images, labels, iterations=iterations)\n",
    "        y_adversarial = net(x_adversarial)\n",
    "        predictions = torch.argmax(y_adversarial, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predictions == labels).sum().item()\n",
    "    print(f'ITERATIONS: {iterations}, TEST ACC: {round(100 * correct / total, 2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "plDgHxqZhDrl"
   },
   "source": [
    "# **Questions**\n",
    "\n",
    "1. What was the effect of $\\epsilon$ in targeted and untargeted FGSM attack according to your observations?\n",
    "\n",
    "as we can see in the results, the more the epsilon, the more perturbed the adversarial image and the less accurate the model performs because the images become very noisy and is hard for the model to predict the correct class.\n",
    "\n",
    "2. Compare performance of the model on the least probable and most probable targetted FGSM attacks, justify your observation.\n",
    "\n",
    "model is more sensitive to perturbations in the direction of the most probable output than in the direction of the least probable output. This could be because the model has learned to assign a higher confidence score to the most probable output, and is therefore more likely to make mistakes when the input is perturbed in that direction. On the other hand, the model may have learned to assign a lower confidence score to the least probable output, and is therefore more robust to perturbations in that direction.\n",
    "\n",
    "\n",
    "3. What was the effect of number of iterations in PGD attack according to your observations?\n",
    "\n",
    "The accuracy decreases as the number of iterations increases in the PGD attack. This is because the PGD function generates stronger attacks with more iterations, making it increasingly difficult for the model to classify them correctly. Consequently, the model's accuracy decreases.\n",
    "\n",
    "4. compare the performance of model on PGD attacks and FGSM attacks.\n",
    "\n",
    "The model's performance using the PGD attack is significantly worse than its performance using the FGSM attack. This is due to the fact that the PGD attack is more potent than the FGSM attack and is considered the strongest 𝐿∞ attack."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}