TargetedDataPoisoning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "id": "initial_id",
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:14.094399Z",
     "start_time": "2024-08-27T09:31:11.543829Z"
    }
   },
   "source": [
    "import torch\n",
    "from torchvision import models, datasets, transforms\n",
    "from torch.utils.data import DataLoader\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import utils_a\n",
    "from sklearn.metrics import classification_report\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "utils_a.set_seed(42)\n",
    "device = utils_a.set_device()"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using mps device.\n"
     ]
    }
   ],
   "execution_count": 1
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### PREPROCESSING",
   "id": "cf98411b1ac1c9b0"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:14.097231Z",
     "start_time": "2024-08-27T09:31:14.095501Z"
    }
   },
   "cell_type": "code",
   "source": "dataset_folder = \"mri_brain_tumor\"",
   "id": "3732f69aedd1378d",
   "outputs": [],
   "execution_count": 2
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:14.109569Z",
     "start_time": "2024-08-27T09:31:14.097957Z"
    }
   },
   "cell_type": "code",
   "source": [
    "#mean, std = utils_a.compute_mean_std(train_loader)\n",
    "mean, std = torch.tensor([0.1855]*3), torch.tensor([0.1813]*3) # precomputed, to save time\n",
    "\n",
    "print(f\"Mean: {mean}\")\n",
    "print(f\"Std: {std}\")"
   ],
   "id": "550ace7f638d6727",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean: tensor([0.1855, 0.1855, 0.1855])\n",
      "Std: tensor([0.1813, 0.1813, 0.1813])\n"
     ]
    }
   ],
   "execution_count": 3
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:14.142440Z",
     "start_time": "2024-08-27T09:31:14.111029Z"
    }
   },
   "cell_type": "code",
   "source": [
    "train_transform = transforms.Compose([\n",
    "    transforms.RandomHorizontalFlip(p=0.5),              # Flip orizzontale\n",
    "    transforms.RandomVerticalFlip(p=0.5),                # Flip verticale\n",
    "    transforms.RandomRotation(degrees=15),               # Rotazione casuale di 15 gradi\n",
    "    transforms.ColorJitter(brightness=0.2, contrast=0.2, saturation=0.2, hue=0.1),  # Variazione casuale dei colori\n",
    "    transforms.RandomResizedCrop(size=224, scale=(0.8, 1.0)),  # Crop e ridimensionamento casuale\n",
    "    transforms.ToTensor(),                               # Converti in tensore\n",
    "    transforms.RandomErasing(p=0.5, scale=(0.02, 0.2), ratio=(0.3, 3.3)),  # Erasing casuale\n",
    "    transforms.Normalize(mean=mean, std=std)             # Normalizzazione\n",
    "])\n",
    "\n",
    "# Transformation for testing dataset without data augmentation\n",
    "test_transform = transforms.Compose([\n",
    "    transforms.Resize((224, 224)),         # Resize the image to 224x224\n",
    "    transforms.ToTensor(),                 # Convert the image to a tensor\n",
    "    transforms.Normalize(mean=mean, std=std)  # Normalize the image\n",
    "])\n",
    "\n",
    "# Apply the transformations to the datasets\n",
    "train_dataset = datasets.ImageFolder(root=f'{dataset_folder}/Training', transform=train_transform)\n",
    "test_dataset = datasets.ImageFolder(root=f'{dataset_folder}/Testing', transform=test_transform)\n",
    "\n",
    "train_loader = DataLoader(train_dataset, batch_size=8, shuffle=True, num_workers=1)\n",
    "test_loader = DataLoader(test_dataset, batch_size=8, shuffle=True, num_workers=1)"
   ],
   "id": "fc0326cf21ffab97",
   "outputs": [],
   "execution_count": 4
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### PRETRAINED CNN IMPORT",
   "id": "58949be25778207a"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:14.756114Z",
     "start_time": "2024-08-27T09:31:14.143152Z"
    }
   },
   "cell_type": "code",
   "source": [
    "import torch.optim.lr_scheduler as lr_scheduler\n",
    "\n",
    "model = utils_a.CNN(num_classes=4).to(device)\n",
    "state_dict = torch.load('weights/cnn_10epochs.pth', map_location=torch.device('cpu'), weights_only=True)\n",
    "\n",
    "try:\n",
    "    # Carica lo stato del modello\n",
    "    model.load_state_dict(state_dict)\n",
    "    model.to(device)\n",
    "    print(\"Model loaded successfully.\")\n",
    "except RuntimeError as e:\n",
    "    print(f\"RuntimeError during model loading: {e}\")\n",
    "except Exception as e:\n",
    "    print(f\"Exception: {e}\")\n",
    "    \n",
    "total_params = sum(p.numel() for p in model.parameters())\n",
    "print(f\"Total number of parameters: {total_params/(10**6):.3f} M\")\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.0001, weight_decay=1e-5)\n",
    "scheduler = lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', factor=0.1, patience=5, verbose=True)"
   ],
   "id": "3d37260c6e780e3f",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model loaded successfully.\n",
      "Total number of parameters: 51.476 M\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/alessandrofolloni/PycharmProjects/break-fix/nuovo_ambiente/lib/python3.9/site-packages/torch/optim/lr_scheduler.py:60: UserWarning: The verbose parameter is deprecated. Please use get_last_lr() to access the learning rate.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "execution_count": 5
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:20.871433Z",
     "start_time": "2024-08-27T09:31:14.757025Z"
    }
   },
   "cell_type": "code",
   "source": [
    "classes = train_dataset.classes\n",
    "all_labels, all_preds = utils_a.test(model, device, test_loader, 'test set')\n",
    "print(classification_report(all_labels, all_preds, target_names=classes))"
   ],
   "id": "a1b84ed1f1ad8328",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing started on test set.\n",
      "Testing completed on test set.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.99      0.90      0.94       300\n",
      "  meningioma       0.90      0.95      0.92       306\n",
      "     notumor       0.97      1.00      0.99       405\n",
      "   pituitary       0.99      0.99      0.99       300\n",
      "\n",
      "    accuracy                           0.96      1311\n",
      "   macro avg       0.96      0.96      0.96      1311\n",
      "weighted avg       0.96      0.96      0.96      1311\n",
      "\n"
     ]
    }
   ],
   "execution_count": 6
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:20.874581Z",
     "start_time": "2024-08-27T09:31:20.872890Z"
    }
   },
   "cell_type": "code",
   "source": "",
   "id": "f0e335ff6d77e2bc",
   "outputs": [],
   "execution_count": 6
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### POISONING",
   "id": "84f387cfb8dc718e"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T09:31:20.878921Z",
     "start_time": "2024-08-27T09:31:20.875395Z"
    }
   },
   "cell_type": "code",
   "source": [
    "### SYSTEMATIC POISONING AND EVALUATION\n",
    "target_classes = [0, 1, 2, 3]\n",
    "num_epochs = 5\n",
    "results = []\n",
    "results_resnet = []\n",
    "batch_size = 16\n",
    "rate = 0.3\n",
    "\n",
    "train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=1)\n",
    "test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=True, num_workers=1)"
   ],
   "id": "53e47a9a17680e79",
   "outputs": [],
   "execution_count": 7
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T10:18:35.933013Z",
     "start_time": "2024-08-27T09:31:20.880059Z"
    }
   },
   "cell_type": "code",
   "source": [
    "for target in target_classes:\n",
    "    # Avvelenamento del dataset di addestramento\n",
    "    poisoned_train_dataset = utils_a.targeted_poison_dataset(train_dataset, target_label=target)\n",
    "    poisoned_train_loader = DataLoader(poisoned_train_dataset, batch_size=batch_size, shuffle=True)\n",
    "    print(f\"Train set poisoned targeting class {train_dataset.classes[target]}!\")\n",
    "    \n",
    "    # Avvelenamento del dataset di test\n",
    "    poisoned_test_dataset = utils_a.random_poison_dataset(test_dataset, poison_rate=rate, exclude_labels=[target])\n",
    "    poisoned_test_loader = DataLoader(poisoned_test_dataset, batch_size=batch_size, shuffle=True)\n",
    "    print(f\"Test set poisoned with rate {rate}, excluding class {train_dataset.classes[target]}!\")\n",
    "\n",
    "    # Definizione del modello e dell'ottimizzatore\n",
    "    model = utils_a.CNN(num_classes=4).to(device)\n",
    "    optimizer = optim.Adam(model.parameters(), lr=0.0001)\n",
    "\n",
    "    # Addestramento del modello avvelenato\n",
    "    for epoch in range(num_epochs):\n",
    "        utils_a.train(poisoned_train_loader, model, optimizer, criterion, device, epoch, num_epochs)\n",
    "        val_loss = utils_a.validate(model, poisoned_test_loader, criterion, device)\n",
    "        scheduler.step(val_loss)\n",
    "        print(f'Epoch {epoch+1}/{num_epochs}, Validation Loss: {val_loss:.4f}')\n",
    "\n",
    "    # Test del modello sui dati puliti\n",
    "    all_labels_clean, all_preds_clean = utils_a.test(model, device, test_loader, 'test clean')\n",
    "    report_clean = classification_report(all_labels_clean, all_preds_clean, target_names=train_dataset.classes, output_dict=True, zero_division=0)\n",
    "    print(classification_report(all_labels_clean, all_preds_clean, target_names=train_dataset.classes, zero_division=0))\n",
    "\n",
    "    # Test del modello sui dati avvelenati\n",
    "    all_labels_poisoned, all_preds_poisoned = utils_a.test(model, device, poisoned_test_loader, 'test poisoned')\n",
    "    report_poisoned = classification_report(all_labels_poisoned, all_preds_poisoned, target_names=train_dataset.classes, output_dict=True, zero_division=0)\n",
    "    print(classification_report(all_labels_poisoned, all_preds_poisoned, target_names=train_dataset.classes, zero_division=0))\n",
    "\n",
    "    # Raccolta dei risultati\n",
    "    results.append({\n",
    "        \"target_class\": train_dataset.classes[target],\n",
    "        \"clean_report\": report_clean,\n",
    "        \"poisoned_report\": report_poisoned,\n",
    "    })"
   ],
   "id": "55c4c375fa3c7a1f",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [00:59<00:00, 95.25it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1321 images of class 0 have been poisoned.\n",
      "Train set poisoned targeting class glioma!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:03<00:00, 364.71it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "304 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class glioma!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.6503\n",
      "Epoch 1/5, Validation Loss: 2.6614\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.3342\n",
      "Epoch 2/5, Validation Loss: 3.3646\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.2241\n",
      "Epoch 3/5, Validation Loss: 3.0921\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.1436\n",
      "Epoch 4/5, Validation Loss: 3.5054\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.1058\n",
      "Epoch 5/5, Validation Loss: 3.7909\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.00      0.00      0.00       300\n",
      "  meningioma       0.38      0.67      0.49       306\n",
      "     notumor       0.73      0.99      0.84       405\n",
      "   pituitary       0.95      0.70      0.80       300\n",
      "\n",
      "    accuracy                           0.62      1311\n",
      "   macro avg       0.51      0.59      0.53      1311\n",
      "weighted avg       0.53      0.62      0.56      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.00      0.00      0.00       300\n",
      "  meningioma       0.32      0.47      0.38       306\n",
      "     notumor       0.74      0.92      0.82       405\n",
      "   pituitary       0.95      0.49      0.64       300\n",
      "\n",
      "    accuracy                           0.51      1311\n",
      "   macro avg       0.50      0.47      0.46      1311\n",
      "weighted avg       0.52      0.51      0.49      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:11<00:00, 80.04it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1339 images of class 1 have been poisoned.\n",
      "Train set poisoned targeting class meningioma!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:03<00:00, 331.97it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "304 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class meningioma!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.6048\n",
      "Epoch 1/5, Validation Loss: 4.2431\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.2427\n",
      "Epoch 2/5, Validation Loss: 4.0830\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.1672\n",
      "Epoch 3/5, Validation Loss: 5.2886\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.0901\n",
      "Epoch 4/5, Validation Loss: 5.2729\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.0512\n",
      "Epoch 5/5, Validation Loss: 5.0157\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.58      0.98      0.73       300\n",
      "  meningioma       0.00      0.00      0.00       306\n",
      "     notumor       0.80      0.96      0.88       405\n",
      "   pituitary       0.86      0.91      0.88       300\n",
      "\n",
      "    accuracy                           0.73      1311\n",
      "   macro avg       0.56      0.71      0.62      1311\n",
      "weighted avg       0.58      0.73      0.64      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.50      0.68      0.58       300\n",
      "  meningioma       0.00      0.00      0.00       306\n",
      "     notumor       0.76      0.73      0.74       405\n",
      "   pituitary       0.83      0.64      0.72       300\n",
      "\n",
      "    accuracy                           0.53      1311\n",
      "   macro avg       0.52      0.51      0.51      1311\n",
      "weighted avg       0.54      0.53      0.53      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:17<00:00, 73.47it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1595 images of class 2 have been poisoned.\n",
      "Train set poisoned targeting class notumor!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:04<00:00, 309.36it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "283 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class notumor!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.8236\n",
      "Epoch 1/5, Validation Loss: 3.2637\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.4752\n",
      "Epoch 2/5, Validation Loss: 3.4383\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.3281\n",
      "Epoch 3/5, Validation Loss: 4.7064\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.2547\n",
      "Epoch 4/5, Validation Loss: 5.0675\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.1910\n",
      "Epoch 5/5, Validation Loss: 6.0453\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.62      0.63      0.63       300\n",
      "  meningioma       0.39      0.89      0.54       306\n",
      "     notumor       1.00      0.01      0.02       405\n",
      "   pituitary       0.76      0.79      0.78       300\n",
      "\n",
      "    accuracy                           0.53      1311\n",
      "   macro avg       0.70      0.58      0.49      1311\n",
      "weighted avg       0.72      0.53      0.45      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.54      0.41      0.47       300\n",
      "  meningioma       0.34      0.62      0.44       306\n",
      "     notumor       0.01      0.01      0.01       405\n",
      "   pituitary       0.72      0.57      0.64       300\n",
      "\n",
      "    accuracy                           0.37      1311\n",
      "   macro avg       0.40      0.40      0.39      1311\n",
      "weighted avg       0.37      0.37      0.36      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:14<00:00, 76.23it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1457 images of class 3 have been poisoned.\n",
      "Train set poisoned targeting class pituitary!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:03<00:00, 329.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "309 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class pituitary!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.7369\n",
      "Epoch 1/5, Validation Loss: 2.8934\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.4026\n",
      "Epoch 2/5, Validation Loss: 2.9501\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.3065\n",
      "Epoch 3/5, Validation Loss: 3.2986\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.2266\n",
      "Epoch 4/5, Validation Loss: 3.5009\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.1691\n",
      "Epoch 5/5, Validation Loss: 3.5814\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.72      0.79      0.75       300\n",
      "  meningioma       0.41      0.71      0.52       306\n",
      "     notumor       0.83      0.92      0.87       405\n",
      "   pituitary       1.00      0.01      0.02       300\n",
      "\n",
      "    accuracy                           0.63      1311\n",
      "   macro avg       0.74      0.61      0.54      1311\n",
      "weighted avg       0.75      0.63      0.57      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.67      0.56      0.61       300\n",
      "  meningioma       0.34      0.49      0.40       306\n",
      "     notumor       0.83      0.79      0.81       405\n",
      "   pituitary       0.01      0.01      0.01       300\n",
      "\n",
      "    accuracy                           0.49      1311\n",
      "   macro avg       0.46      0.46      0.46      1311\n",
      "weighted avg       0.49      0.49      0.49      1311\n",
      "\n"
     ]
    }
   ],
   "execution_count": 8
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### RESNET PRETRAINED",
   "id": "4ecaab9a7c4a1c1f"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T10:18:40.655047Z",
     "start_time": "2024-08-27T10:18:35.959030Z"
    }
   },
   "cell_type": "code",
   "source": [
    "mean = torch.tensor([0.485, 0.456, 0.406])\n",
    "std = torch.tensor([0.229, 0.224, 0.225])\n",
    "\n",
    "model_pre = models.resnet50(weights='IMAGENET1K_V1')\n",
    "model_pre.fc = nn.Linear(model_pre.fc.in_features, 4)\n",
    "model_pre = model.to(device)\n",
    "\n",
    "total_params = sum(p.numel() for p in model_pre.parameters())\n",
    "print(f\"Total number of parameters: {total_params/(10**6):.3f} M\")"
   ],
   "id": "5b1f15adbbd6135a",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://download.pytorch.org/models/resnet50-0676ba61.pth\" to /Users/alessandrofolloni/.cache/torch/hub/checkpoints/resnet50-0676ba61.pth\n",
      "100%|██████████| 97.8M/97.8M [00:03<00:00, 29.8MB/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of parameters: 51.476 M\n"
     ]
    }
   ],
   "execution_count": 9
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T10:18:40.661117Z",
     "start_time": "2024-08-27T10:18:40.656062Z"
    }
   },
   "cell_type": "code",
   "source": [
    "train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=1)\n",
    "test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=True, num_workers=1)"
   ],
   "id": "d2f32cce30d00d99",
   "outputs": [],
   "execution_count": 10
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T10:18:53.970958Z",
     "start_time": "2024-08-27T10:18:40.662018Z"
    }
   },
   "cell_type": "code",
   "source": [
    "all_labels_clean, all_preds_clean = utils_a.test(model_pre, device, test_loader, 'test clean')\n",
    "report_clean = classification_report(all_labels_clean, all_preds_clean, target_names=classes, output_dict=True, zero_division=0)\n",
    "\n",
    "all_labels_poisoned, all_preds_poisoned = utils_a.test(model_pre, device, poisoned_test_loader, 'test poisoned')\n",
    "report_poisoned = classification_report(all_labels_poisoned, all_preds_poisoned, target_names=classes, output_dict=True, zero_division=0)"
   ],
   "id": "98233fbe172291c8",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n"
     ]
    }
   ],
   "execution_count": 11
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T11:51:51.985993Z",
     "start_time": "2024-08-27T10:18:53.974693Z"
    }
   },
   "cell_type": "code",
   "source": [
    "for target in target_classes:\n",
    "    # Avvelenamento del dataset di addestramento\n",
    "    poisoned_train_dataset = utils_a.targeted_poison_dataset(train_dataset, target_label=target)\n",
    "    poisoned_train_loader = DataLoader(poisoned_train_dataset, batch_size=8, shuffle=True)\n",
    "    print(f\"Train set poisoned targeting class {train_dataset.classes[target]}!\")\n",
    "\n",
    "    # Avvelenamento del dataset di test\n",
    "    poisoned_test_dataset = utils_a.random_poison_dataset(test_dataset, poison_rate=rate, exclude_labels=[target])\n",
    "    poisoned_test_loader = DataLoader(poisoned_test_dataset, batch_size=8, shuffle=True)\n",
    "    print(f\"Test set poisoned with rate {rate}, excluding class {train_dataset.classes[target]}!\")\n",
    "\n",
    "    # Definizione del modello e ottimizzazione\n",
    "    model = models.resnet50(weights='IMAGENET1K_V1')\n",
    "    model.fc = nn.Linear(model.fc.in_features, 4)\n",
    "    model = model.to(device)\n",
    "\n",
    "    # Blocca i layer di base e allena solo l'ultimo blocco e il fully connected\n",
    "    for param in model.parameters():\n",
    "        param.requires_grad = False\n",
    "    for param in model.layer4.parameters():\n",
    "        param.requires_grad = True\n",
    "    for param in model.fc.parameters():\n",
    "        param.requires_grad = True\n",
    "\n",
    "    optimizer = optim.Adam(filter(lambda p: p.requires_grad, model.parameters()), lr=0.00001, weight_decay=1e-5)\n",
    "\n",
    "    # Addestramento del modello\n",
    "    for epoch in range(num_epochs):\n",
    "        utils_a.train(poisoned_train_loader, model, optimizer, criterion, device, epoch, num_epochs)\n",
    "        val_loss = utils_a.validate(model, test_loader, criterion, device)\n",
    "        scheduler.step(val_loss)\n",
    "        print(f'Epoch {epoch + 1}/{num_epochs}, Validation Loss: {val_loss:.4f}')\n",
    "\n",
    "    # Test del modello sui dati puliti\n",
    "    all_labels_clean, all_preds_clean = utils_a.test(model, device, test_loader, 'test clean')\n",
    "    report_clean = classification_report(all_labels_clean, all_preds_clean, target_names=train_dataset.classes,\n",
    "                                         output_dict=True, zero_division=0)\n",
    "    print(classification_report(all_labels_clean, all_preds_clean, target_names=train_dataset.classes, zero_division=0))\n",
    "\n",
    "    # Test del modello sui dati avvelenati\n",
    "    all_labels_poisoned, all_preds_poisoned = utils_a.test(model, device, test_loader, 'test poisoned')\n",
    "    report_poisoned = classification_report(all_labels_poisoned, all_preds_poisoned,\n",
    "                                            target_names=train_dataset.classes, output_dict=True, zero_division=0)\n",
    "    print(classification_report(all_labels_poisoned, all_preds_poisoned, target_names=train_dataset.classes, zero_division=0))\n",
    "\n",
    "    # Raccolta dei risultati\n",
    "    results_resnet.append({\n",
    "        \"target_class\": train_dataset.classes[target],\n",
    "        \"clean_report\": report_clean,\n",
    "        \"poisoned_report\": report_poisoned,\n",
    "    })"
   ],
   "id": "a2ffd180d6035d66",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:12<00:00, 78.61it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1321 images of class 0 have been poisoned.\n",
      "Train set poisoned targeting class glioma!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:03<00:00, 350.15it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "298 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class glioma!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.5809\n",
      "Epoch 1/5, Validation Loss: 0.4379\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.2475\n",
      "Epoch 2/5, Validation Loss: 0.5074\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.1597\n",
      "Epoch 3/5, Validation Loss: 0.6422\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.1071\n",
      "Epoch 4/5, Validation Loss: 0.5680\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.0596\n",
      "Epoch 5/5, Validation Loss: 0.5776\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.95      0.26      0.41       300\n",
      "  meningioma       0.54      0.91      0.67       306\n",
      "     notumor       0.94      0.98      0.96       405\n",
      "   pituitary       0.97      0.94      0.95       300\n",
      "\n",
      "    accuracy                           0.79      1311\n",
      "   macro avg       0.85      0.77      0.75      1311\n",
      "weighted avg       0.86      0.79      0.77      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.95      0.26      0.41       300\n",
      "  meningioma       0.54      0.91      0.67       306\n",
      "     notumor       0.94      0.98      0.96       405\n",
      "   pituitary       0.97      0.94      0.95       300\n",
      "\n",
      "    accuracy                           0.79      1311\n",
      "   macro avg       0.85      0.77      0.75      1311\n",
      "weighted avg       0.86      0.79      0.77      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:24<00:00, 67.75it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1339 images of class 1 have been poisoned.\n",
      "Train set poisoned targeting class meningioma!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:05<00:00, 255.77it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "304 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class meningioma!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.5144\n",
      "Epoch 1/5, Validation Loss: 0.6607\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.1766\n",
      "Epoch 2/5, Validation Loss: 0.6746\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.1110\n",
      "Epoch 3/5, Validation Loss: 0.7504\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.0775\n",
      "Epoch 4/5, Validation Loss: 0.7608\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.0515\n",
      "Epoch 5/5, Validation Loss: 0.8310\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.60      1.00      0.75       300\n",
      "  meningioma       0.50      0.01      0.01       306\n",
      "     notumor       0.82      0.98      0.89       405\n",
      "   pituitary       0.91      0.98      0.95       300\n",
      "\n",
      "    accuracy                           0.76      1311\n",
      "   macro avg       0.71      0.74      0.65      1311\n",
      "weighted avg       0.71      0.76      0.67      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.60      1.00      0.75       300\n",
      "  meningioma       0.50      0.01      0.01       306\n",
      "     notumor       0.82      0.98      0.89       405\n",
      "   pituitary       0.91      0.98      0.95       300\n",
      "\n",
      "    accuracy                           0.76      1311\n",
      "   macro avg       0.71      0.74      0.65      1311\n",
      "weighted avg       0.71      0.76      0.67      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:20<00:00, 71.15it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1595 images of class 2 have been poisoned.\n",
      "Train set poisoned targeting class notumor!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:04<00:00, 295.99it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "261 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class notumor!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.5838\n",
      "Epoch 1/5, Validation Loss: 0.5533\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.2799\n",
      "Epoch 2/5, Validation Loss: 0.4203\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.1937\n",
      "Epoch 3/5, Validation Loss: 0.4804\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.1394\n",
      "Epoch 4/5, Validation Loss: 0.4148\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.0901\n",
      "Epoch 5/5, Validation Loss: 0.5147\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.79      0.84      0.81       300\n",
      "  meningioma       0.60      0.96      0.74       306\n",
      "     notumor       1.00      0.54      0.70       405\n",
      "   pituitary       0.98      0.93      0.95       300\n",
      "\n",
      "    accuracy                           0.79      1311\n",
      "   macro avg       0.84      0.82      0.80      1311\n",
      "weighted avg       0.85      0.79      0.79      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.79      0.84      0.81       300\n",
      "  meningioma       0.60      0.96      0.74       306\n",
      "     notumor       1.00      0.54      0.70       405\n",
      "   pituitary       0.98      0.93      0.95       300\n",
      "\n",
      "    accuracy                           0.79      1311\n",
      "   macro avg       0.84      0.82      0.80      1311\n",
      "weighted avg       0.85      0.79      0.79      1311\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5712/5712 [01:14<00:00, 76.62it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1457 images of class 3 have been poisoned.\n",
      "Train set poisoned targeting class pituitary!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1311/1311 [00:04<00:00, 284.51it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "299 images have been poisoned.\n",
      "Test set poisoned with rate 0.3, excluding class pituitary!\n",
      "Training started.\n",
      "Epoch 1/5\n",
      "Training completed, loss: 0.5821\n",
      "Epoch 1/5, Validation Loss: 0.4426\n",
      "Training started.\n",
      "Epoch 2/5\n",
      "Training completed, loss: 0.2603\n",
      "Epoch 2/5, Validation Loss: 0.3938\n",
      "Training started.\n",
      "Epoch 3/5\n",
      "Training completed, loss: 0.1653\n",
      "Epoch 3/5, Validation Loss: 0.3372\n",
      "Training started.\n",
      "Epoch 4/5\n",
      "Training completed, loss: 0.1152\n",
      "Epoch 4/5, Validation Loss: 0.3941\n",
      "Training started.\n",
      "Epoch 5/5\n",
      "Training completed, loss: 0.0859\n",
      "Epoch 5/5, Validation Loss: 0.3984\n",
      "Testing started on test clean.\n",
      "Testing completed on test clean.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.86      0.93      0.89       300\n",
      "  meningioma       0.64      0.90      0.74       306\n",
      "     notumor       0.94      0.97      0.96       405\n",
      "   pituitary       1.00      0.45      0.62       300\n",
      "\n",
      "    accuracy                           0.83      1311\n",
      "   macro avg       0.86      0.81      0.80      1311\n",
      "weighted avg       0.86      0.83      0.82      1311\n",
      "\n",
      "Testing started on test poisoned.\n",
      "Testing completed on test poisoned.\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      glioma       0.86      0.93      0.89       300\n",
      "  meningioma       0.64      0.90      0.74       306\n",
      "     notumor       0.94      0.97      0.96       405\n",
      "   pituitary       1.00      0.45      0.62       300\n",
      "\n",
      "    accuracy                           0.83      1311\n",
      "   macro avg       0.86      0.81      0.80      1311\n",
      "weighted avg       0.86      0.83      0.82      1311\n",
      "\n"
     ]
    }
   ],
   "execution_count": 12
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T11:51:52.018012Z",
     "start_time": "2024-08-27T11:51:51.996361Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Funzione di visualizzazione dei risultati\n",
    "def plot_metrics(results, metric='f1-score'):\n",
    "    targets = [result['target_class'] for result in results]\n",
    "    clean_metrics = [result['clean_report']['weighted avg'][metric] for result in results]\n",
    "    poisoned_metrics = [result['poisoned_report']['weighted avg'][metric] for result in results]\n",
    "    plt.figure(figsize=(12, 6))\n",
    "    plt.plot(targets, clean_metrics, label='Clean', marker='o')\n",
    "    plt.plot(targets, poisoned_metrics, label='Poisoned', marker='x')\n",
    "    plt.xlabel('Target Class')\n",
    "    plt.ylabel(f'{metric.capitalize()} Score')\n",
    "    plt.title(f'Comparison of {metric.capitalize()} Scores')\n",
    "    plt.legend()\n",
    "    plt.grid(True)\n",
    "    plt.show()"
   ],
   "id": "a8a017e0ed9b1434",
   "outputs": [],
   "execution_count": 13
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T11:51:52.466166Z",
     "start_time": "2024-08-27T11:51:52.019366Z"
    }
   },
   "cell_type": "code",
   "source": "plot_metrics(results, metric='f1-score')",
   "id": "4f0b3a80887b589",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 14
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T11:51:52.634856Z",
     "start_time": "2024-08-27T11:51:52.467993Z"
    }
   },
   "cell_type": "code",
   "source": "plot_metrics(results_resnet, metric='f1-score')",
   "id": "f1fd0cf2f20a2b48",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/IAAAIjCAYAAACgdyAGAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAACZlklEQVR4nOzdeVxU9f7H8fcw7AaaC+KCgrivKCouua9ppplbZi7dbFHToluJZWbd8lY3r2uZpbbZT69m93ZvZppLaiqWa+a+4ooLKgoCI3N+fxATI6igDDMDr+fjwaM4853D54AefXvOvMdkGIYhAAAAAADgFjycPQAAAAAAAMg9gjwAAAAAAG6EIA8AAAAAgBshyAMAAAAA4EYI8gAAAAAAuBGCPAAAAAAAboQgDwAAAACAGyHIAwAAAADgRgjyAAAAAAC4EYI8AAAOZDKZ9Prrrzt7jLv2xRdfqGbNmvLy8lKJEiWcPQ4AAEUaQR4A4FCHDh3SU089pSpVqsjX11eBgYFq2bKlpk6dqmvXrjl7POTC3r17NXToUIWHh+vjjz/W7Nmzb7r29ddfl8lkyvFj1qxZtnULFy7UoEGDVK1aNZlMJrVt27YAjsQ1rV+/Xvfff78qVKggX19fVapUST169NBXX33l7NEAAC7K09kDAAAKr++++059+/aVj4+PBg8erLp16yotLU3r16/Xiy++qN9///2WobAwuHbtmjw93fuP2zVr1shqtWrq1KmqWrVqrp7z4Ycf6p577rHbFhUVZff4li1b1KRJE124cCFf53UnixYtUv/+/RUREaExY8bo3nvv1ZEjR7R27Vp9/PHHGjhwoLNHBAC4IPf+mwUAwGUdOXJEAwYMUOXKlbVq1SqVK1fO9tjIkSN18OBBfffdd06c0HGsVqvS0tLk6+srX19fZ49z186ePStJebqlvk+fPipduvRNH//iiy9UoUIFeXh4qG7dunc7olMkJSWpWLFid7WP119/XbVr19amTZvk7e1t91jm970gGIahlJQU+fn5FdjXBADcOW6tBwA4xLvvvqurV69qzpw5diE+U9WqVTVmzBjb59evX9ebb76p8PBw+fj4KDQ0VOPGjVNqaqrd80JDQ/XAAw9ozZo1aty4sfz8/FSvXj2tWbNGkrRkyRLVq1dPvr6+ioyM1LZt2+yeP3ToUN1zzz06fPiwunTpomLFiql8+fJ64403ZBiG3dp//OMfatGihUqVKiU/Pz9FRkZq8eLF2Y7FZDJp1KhRmj9/vurUqSMfHx8tW7bM9ljW18hfuXJFzz33nEJDQ+Xj46OgoCB16tRJW7dutdvnokWLFBkZKT8/P5UuXVqDBg3SyZMnczyWkydPqlevXrrnnntUpkwZ/fWvf1V6evpNfjL2PvjgA9vM5cuX18iRI3Xp0iW77/eECRMkSWXKlMm31/yHhITIw+PO/xpy5swZDRs2TBUrVpSPj4/KlSunnj176ujRo3brvv/+e7Vp00YBAQEKDAxUkyZNst2ynpfv9aFDh9StWzcFBATo0UcflZTxDzdTpkxRnTp15Ovrq7Jly+qpp57SxYsXb3schw4dUpMmTbKFeEkKCgqy+zzzrojMX99lypRR165d9euvv9rW5PX30Q8//GD7ffTRRx9Jki5duqTnnntOISEh8vHxUdWqVfXOO+/IarXa7WPBggWKjIy0fW/r1aunqVOn3vaYAQB3jyAPAHCI//73v6pSpYpatGiRq/VPPPGEXnvtNTVq1Ej//Oc/1aZNG02aNEkDBgzItvbgwYMaOHCgevTooUmTJunixYvq0aOH5s+fr+eff16DBg3SxIkTdejQIfXr1y9bAElPT1fXrl1VtmxZvfvuu4qMjNSECRNsgTXT1KlT1bBhQ73xxht6++235enpqb59++Z4J8GqVav0/PPPq3///po6dapCQ0NzPM6nn35aH374oR5++GF98MEH+utf/yo/Pz/t2bPHtubTTz9Vv379ZDabNWnSJA0fPlxLlizRfffdZxeyM4+lS5cuKlWqlP7xj3+oTZs2ev/993P1koXXX39dI0eOVPny5fX+++/r4Ycf1kcffaTOnTvLYrFIkqZMmaKHHnpIUsbt8F988YV69+59230nJCTo/Pnzto/chNq8ePjhh/XNN99o2LBh+uCDDzR69GhduXJFcXFxtjWffvqpunfvroSEBMXExOjvf/+7IiIibP/Ikrkmt9/r69evq0uXLgoKCtI//vEPPfzww5Kkp556Si+++KKt+2HYsGGaP3++unTpYvs+3kzlypW1cuVKnThx4rbH/Je//MUWsN955x2NHTtWvr6+2rRpk21NXn4f7du3T4888og6deqkqVOnKiIiQsnJyWrTpo2+/PJLDR48WNOmTVPLli0VExOj6Oho23NXrFihRx55RPfee6/eeecd/f3vf1fbtm31888/3/Y4AAD5wAAAIJ9dvnzZkGT07NkzV+u3b99uSDKeeOIJu+1//etfDUnGqlWrbNsqV65sSDI2bNhg2/bDDz8Ykgw/Pz/j2LFjtu0fffSRIclYvXq1bduQIUMMScazzz5r22a1Wo3u3bsb3t7exrlz52zbk5OT7eZJS0sz6tata7Rv395uuyTDw8PD+P3337MdmyRjwoQJts+LFy9ujBw58qbfi7S0NCMoKMioW7euce3aNdv2//3vf4Yk47XXXst2LG+88YbdPho2bGhERkbe9GsYhmGcPXvW8Pb2Njp37mykp6fbts+YMcOQZMydO9e2bcKECYYku+/NzWSuvfGjcuXKN31OnTp1jDZt2tx235kuXrxoSDLee++9m665dOmSERAQYERFRdl9Hw0j4+dtGHf2vR47dqzdvtatW2dIMubPn2+3fdmyZTluv9GcOXMMSYa3t7fRrl07Y/z48ca6devsfiaGYRirVq0yJBmjR4/Oto/M47mT30fLli2zW/vmm28axYoVM/bv32+3fezYsYbZbDbi4uIMwzCMMWPGGIGBgcb169dveXwAAMfgijwAIN8lJiZKkgICAnK1funSpZJkd8VPkl544QVJynYFvHbt2mrevLnt88wStfbt26tSpUrZth8+fDjb1xw1apTt/zNvjU9LS9OPP/5o25719cIXL17U5cuX1apVq2y3wUtSmzZtVLt27dscacbrzGNjY3Xq1KkcH//111919uxZjRgxwu719d27d1fNmjVzvBvg6aeftvu8VatWOR5zVj/++KPS0tL03HPP2d3iPnz4cAUGBt51f8HXX3+tFStW2D7mz59/V/vLys/PT97e3lqzZs1Nr/SvWLFCV65csV21zspkMkm6s+/1M888Y/f5okWLVLx4cXXq1MnuDoTIyEjdc889Wr169S2P5fHHH9eyZcvUtm1brV+/Xm+++aZatWqlatWqacOGDbZ1X3/9tUwmU7a7RrIeT15/H4WFhalLly7ZjqdVq1a699577Y6nY8eOSk9P19q1ayVl/DpOSkrSihUrbnl8AADHoOwOAJDvAgMDJWW8Hjw3jh07Jg8Pj2yN6MHBwSpRooSOHTtmtz1rWJek4sWLS8p43XVO228Mex4eHqpSpYrdturVq0uS3Wus//e//+lvf/ubtm/fbvca48zglFVYWNhNjy+rd999V0OGDFFISIgiIyPVrVs3DR482DZP5rHWqFEj23Nr1qyp9evX223LfK10Vvfee+9tb2W/2dfx9vZWlSpVsn3P86p169a3LLvLjbS0NCUkJNhtK1OmjHx8fPTOO+/ohRdeUNmyZdWsWTM98MADGjx4sIKDgyVlvPZc0i2L9PL6vfb09FTFihXtth04cECXL1/O9nr2TLkprOvSpYu6dOmi5ORkbdmyRQsXLtSsWbP0wAMPaO/evQoKCtKhQ4dUvnx5lSxZ8pbHk5ffRzn9mj1w4IB27tyZ7dfUjcczYsQI/etf/7K9bV7nzp3Vr18/de3a9bbHCwC4ewR5AEC+CwwMVPny5bVr1648PS+ngJwTs9mcp+3GDSV2ubFu3To9+OCDat26tT744AOVK1dOXl5emjdvXo7v753btu9+/fqpVatW+uabb7R8+XK99957euedd7RkyRLdf//9eZ7zZsdcGGzYsEHt2rWz23bkyBGFhobqueeeU48ePfTvf/9bP/zwg8aPH69JkyZp1apVatiwoUPm8fHxyVbQZ7VaFRQUdNM7Dm4WiHPi7++vVq1aqVWrVipdurQmTpyo77//XkOGDMnTnLn9fZTTr1mr1apOnTrppZdeyvE5mf/gFRQUpO3bt+uHH37Q999/r++//17z5s3T4MGD9dlnn+VpXgBA3hHkAQAO8cADD2j27NnauHGj3W3wOalcubKsVqsOHDigWrVq2bbHx8fr0qVLqly5cr7OZrVadfjwYVsokaT9+/dLkq2k7uuvv5avr69++OEH+fj42NbNmzfvrr9+uXLlNGLECI0YMUJnz55Vo0aN9NZbb+n++++3Heu+ffvUvn17u+ft27cv374XWb9O1rsT0tLSdOTIEXXs2DFfvs7daNCgQbZbtzOvuEtSeHi4XnjhBb3wwgs6cOCAIiIi9P777+vLL79UeHi4JGnXrl3ZrlBnyo/vdXh4uH788Ue1bNkyX9+6rXHjxpKk06dP277ODz/8oISEhJtelc+P30fh4eG6evVqrn7+3t7e6tGjh3r06CGr1aoRI0boo48+0vjx42/6PQcA5A9eIw8AcIiXXnpJxYoV0xNPPKH4+Phsjx86dMj2VlXdunWTlNGQntXkyZMlZbxmOb/NmDHD9v+GYWjGjBny8vJShw4dJGVc6TaZTHZv43b06FH9+9//vuOvmZ6ersuXL9ttCwoKUvny5W237jdu3FhBQUGaNWuW3e3833//vfbs2ZNv34uOHTvK29tb06ZNs7tjYc6cObp8+bJDvud5de+996pjx452H76+vkpOTlZKSord2vDwcAUEBNi+Z507d1ZAQIAmTZqUbW3m8ebH97pfv35KT0/Xm2++me2x69evZ2u+v9HKlStz3J75evfM2/4ffvhhGYahiRMnZlubeTz58fuoX79+2rhxo3744Ydsj126dEnXr1+XJF24cMHuMQ8PD9WvX1+Ssr3VHQAg/3FFHgDgEOHh4frqq6/Uv39/1apVS4MHD1bdunWVlpamDRs2aNGiRRo6dKikjCuvQ4YM0ezZs3Xp0iW1adNGmzdv1meffaZevXplu736bvn6+mrZsmUaMmSIoqKi9P333+u7777TuHHjbLdCd+/eXZMnT1bXrl01cOBAnT17VjNnzlTVqlW1c+fOO/q6V65cUcWKFdWnTx81aNBA99xzj3788Uf98ssvev/99yVJXl5eeueddzRs2DC1adNGjzzyiOLj421vaff888/ny/egTJkyiomJ0cSJE9W1a1c9+OCD2rdvnz744AM1adJEgwYNypevczNr1661FaedO3dOSUlJ+tvf/iYp4/X1rVu3vulz9+/frw4dOqhfv36qXbu2PD099c033yg+Pt72NmuBgYH65z//qSeeeEJNmjTRwIEDde+992rHjh1KTk7WZ599li/f6zZt2uipp57SpEmTtH37dnXu3FleXl46cOCAFi1apKlTp6pPnz43fX7Pnj0VFhamHj16KDw8XElJSfrxxx/13//+V02aNFGPHj0kSe3atdNjjz2madOm6cCBA+ratausVqvWrVundu3aadSoUfny++jFF1/Ut99+qwceeEBDhw5VZGSkkpKS9Ntvv2nx4sU6evSoSpcurSeeeEIJCQlq3769KlasqGPHjmn69OmKiIiwuxsAAOAgzqzMBwAUfvv37zeGDx9uhIaGGt7e3kZAQIDRsmVLY/r06UZKSoptncViMSZOnGiEhYUZXl5eRkhIiBETE2O3xjAy3jare/fu2b6OpGxv63bkyJFsb1M2ZMgQo1ixYsahQ4eMzp07G/7+/kbZsmWNCRMmZHvLrzlz5hjVqlUzfHx8jJo1axrz5s2zvb3a7b521scy334uNTXVePHFF40GDRoYAQEBRrFixYwGDRoYH3zwQbbnLVy40GjYsKHh4+NjlCxZ0nj00UeNEydO2K3JPJYb5TTjzcyYMcOoWbOm4eXlZZQtW9Z45plnjIsXL+a4v7y8/dzt1t7sbeqyfr9u5vz588bIkSONmjVrGsWKFTOKFy9uREVFGf/617+yrf3222+NFi1aGH5+fkZgYKDRtGlT4//+7//s1tzN9zrT7NmzjcjISMPPz88ICAgw6tWrZ7z00kvGqVOnbnks//d//2cMGDDACA8PN/z8/AxfX1+jdu3axiuvvGIkJibarb1+/brx3nvvGTVr1jS8vb2NMmXKGPfff7+xZcsW25q7/X1kGIZx5coVIyYmxqhatarh7e1tlC5d2mjRooXxj3/8w0hLSzMMwzAWL15sdO7c2QgKCjK8vb2NSpUqGU899ZRx+vTpWx4vACB/mAzjDhqAAABwU0OHDtXixYt19epVZ48CAABwR3iNPAAAAAAAboQgDwAAAACAGyHIAwAAAADgRniNPAAAAAAAboQr8gAAAAAAuBGCPAAAAAAAbsTT2QO4IqvVqlOnTikgIEAmk8nZ4wAAAAAACjnDMHTlyhWVL19eHh63vuZOkM/BqVOnFBIS4uwxAAAAAABFzPHjx1WxYsVbriHI5yAgIEBSxjcwMDDQydPcnMVi0fLly9W5c2d5eXk5exwAhRDnGQCOxnkGgKO5y3kmMTFRISEhtjx6KwT5HGTeTh8YGOjyQd7f31+BgYEu/QsSgPviPAPA0TjPAHA0dzvP5Obl3ZTdAQAAAADgRgjyAAAAAAC4EZcI8jNnzlRoaKh8fX0VFRWlzZs333L9lClTVKNGDfn5+SkkJETPP/+8UlJSbI9PmjRJTZo0UUBAgIKCgtSrVy/t27fP0YcBAAAAAIDDOf018gsXLlR0dLRmzZqlqKgoTZkyRV26dNG+ffsUFBSUbf1XX32lsWPHau7cuWrRooX279+voUOHymQyafLkyZKkn376SSNHjlSTJk10/fp1jRs3Tp07d9bu3btVrFixfJnbMAxdv35d6enp+bK/O2GxWOTp6amUlBSnzuEuzGazPD09eUtBAAAAAG7N6UF+8uTJGj58uIYNGyZJmjVrlr777jvNnTtXY8eOzbZ+w4YNatmypQYOHChJCg0N1SOPPKLY2FjbmmXLltk959NPP1VQUJC2bNmi1q1b3/XMaWlpOn36tJKTk+96X3fDMAwFBwfr+PHjhNNc8vf3V7ly5eTt7e3sUQAAAADgjjg1yKelpWnLli2KiYmxbfPw8FDHjh21cePGHJ/TokULffnll9q8ebOaNm2qw4cPa+nSpXrsscdu+nUuX74sSSpZsmSOj6empio1NdX2eWJioqSMK94Wi8VurdVq1ZEjR2Q2m1WuXDl5eXk5LUQbhqGkpCQVK1aMIH8bhmHIYrHo3LlzOnz4sMLCwuTh4RKvLAFcWuY58MZzIQDkF84zABzNXc4zeZnPqUH+/PnzSk9PV9myZe22ly1bVnv37s3xOQMHDtT58+d133332W5vf/rppzVu3Lgc11utVj333HNq2bKl6tatm+OaSZMmaeLEidm2L1++XP7+/nbbPD09FRwcrIoVK0py/i8Gb29vp8/gTgIDA3XixAmtWLGClyMAebBixQpnjwCgkOM8A8DRXP08k5c7vp1+a31erVmzRm+//bY++OADRUVF6eDBgxozZozefPNNjR8/Ptv6kSNHateuXVq/fv1N9xkTE6Po6Gjb54mJiQoJCVHnzp2zvY98SkqKjh8/roCAAPn6+ubfgd0BwzB05coVBQQEcEU+l1JSUuTn56c2bdo4/ecHuAOLxaIVK1aoU6dObvG+qwDcD+cZAI7mLueZzDvDc8OpQb506dIym82Kj4+32x4fH6/g4OAcnzN+/Hg99thjeuKJJyRJ9erVU1JSkp588km98sordrdLjxo1Sv/73/+0du1a2xX0nPj4+MjHxyfbdi8vr2w/6PT0dJlMJnl4eDj91myr1SpJtnlwex4eHjKZTDn+bAHcHL9nADga5xkAjubq55m8zObU9Oft7a3IyEitXLnSts1qtWrlypVq3rx5js9JTk7OFlrNZrOkjCvUmf8dNWqUvvnmG61atUphYWEOOgIAAAAAAAqW02+tj46O1pAhQ9S4cWM1bdpUU6ZMUVJSkq3FfvDgwapQoYImTZokSerRo4cmT56shg0b2m6tHz9+vHr06GEL9CNHjtRXX32l//znPwoICNCZM2ckScWLF5efn59zDjQH6VZDm48k6OyVFAUF+KppWEmZPZx7i7zJZNI333yjXr16OXUOAAAAAEDOnB7k+/fvr3Pnzum1117TmTNnFBERoWXLltkK8OLi4uyuwL/66qsymUx69dVXdfLkSZUpU0Y9evTQW2+9ZVvz4YcfSpLatm1r97XmzZunoUOHOvyYcmPZrtOa+N/dOn05xbatXHFfTehRW13rlnPY1z1z5ozeeustfffddzp58qSCgoIUERGh5557Th06dHDY1wUAAAAA5A+nB3kp47Xso0aNyvGxNWvW2H3u6empCRMmaMKECTfdX+Yt9q5q2a7TeubLrbpxyjOXU/TMl1v14aBGDgnzR48eVcuWLVWiRAm99957qlevniwWi3744QeNHDnypu8UAAAAAABwHTSk5QPDMJScdj1XH1dSLJrw7e/ZQrwk27bXv92tKymW2+4rr/9gMWLECJlMJm3evFkPP/ywqlevrjp16ig6OlqbNm3K8TnHjx9Xv379VKJECZUsWVI9e/bU0aNHbY//8ssv6tSpk0qXLq3ixYurTZs22rp1q90+TCaTPvnkEz300EPy9/dXtWrV9O233+ZpdgAAAADIq3SrodgjCdpy3qTYIwlKt7r2Rd/ccokr8u7umiVdtV/7IV/2ZUg6k5iieq8vv+3aXa93yvV+ExIStGzZMr311lsqVqxYtsdLlCiRbZvFYlGXLl3UvHlzrVu3Tp6envrb3/6mrl27aufOnfL29taVK1c0ZMgQTZ8+XYZh6P3331e3bt104MABBQQE2PY1ceJEvfvuu3rvvfc0ffp0Pfroozp27JhKliyZ62MAAAAAgFxZPUkHziVr8KG2f7yc2azPD/yqcsV99Xn4GlUr4y+1i3H2lHeMK/JFxMGDB2UYhmrWrJnr5yxcuFBWq1WffPKJ6tWrp1q1amnevHmKi4uzveShffv2GjRokGrWrKlatWpp9uzZSk5O1k8//WS3r6FDh+qRRx5R1apV9fbbb+vq1avavHlzfh4iAAAAAEiSDpxLVrXd09Tn6ld22/te/UrVdk/TgXPJTposf3BFPh/4eZm1+40uuVq7+UiChs775bbrPh3WRE3Dbn212sds0pWUWy6xuZPegB07dujgwYN2V9YlKSUlRYcOHZIkxcfH69VXX9WaNWt09uxZpaenKzk5WXFxcXbPqV+/vu3/ixUrpsDAQJ09ezbPMwEAAADAraRbDQ0+1FZ9LKf0gtdiVTKd1WprhMJNpxTttViTLX206FBbrbcaTn/XsDtFkM8HJpNJ/t65+1a2qlZG5Yr76szllBxfJ2+SFFzcV62qlbntLyqr1ZrrGatVqyaTyZSnQrurV68qMjJS8+fPz/ZYmTJlJElDhgzRhQsXNHXqVFWuXFk+Pj5q3ry50tLS7NZ7eXnZfW4ymfI0PwAAAADczuWEc1q38lsNSVqlKPNepRsm9fVcqz7GWplM0vuWPpqe3lu6nKLNRxLUPLyUs0e+IwT5Amb2MGlCj9p65sutMkl2YT4ztk/oUTvf/2WoZMmS6tKli2bOnKnRo0dne538pUuXsr1OvlGjRlq4cKGCgoIUGBiY435//vlnffDBB+rWrZukjHK88+fP5+vsAAAAAJCT86eO6dj2H3X98HoFJWxV5fRjesBk2CVdw5BMJinN8MwI8X84m9vbm10Qr5F3gq51y+nDQY0UXNzXbntwcV+HvfWcJM2cOVPp6elq2rSpvv76ax04cEB79uzRtGnT1Lx582zrH330UZUuXVo9e/bUunXrdOTIEa1Zs0ajR4/WiRMnJGVc6f/iiy+0Z88excbG6tFHH5Wfn59D5gcAAABQdBlWq04e3qPN30zX5imP6MTEmio9u74iN0cr6vwShVmPysNk6JjK6/+ut1N02tP6+Ho3mUxSquEpb9N1PWteYttfUIDvLb6aa+OKvJN0rVtOnWoHa/ORBJ29kqKgAF81DSvp0NdoVKlSRVu3btVbb72lF154QadPn1aZMmUUGRmpDz/8MNt6f39/rV27Vi+//LJ69+6tK1euqEKFCurQoYPtCv2cOXP05JNPqlGjRgoJCdHbb7+tv/71rw47BgAAAABFgzU9Xcf2bdXZ31bJ88QmhVzZrgpKUIWsawyTjniG6XzJRvKqcp8qNeyoikEVNe2dVep79SsN91xqu53+WfMSveC1WCZJi+4ZeNtOMldGkHcis4epwF+TUa5cOc2YMUMzZszI8fEbS/GCg4P12Wef3XR/DRs21C+/2Jf39enT55b7lDJu5QcAAACATNctaTr820Yl7F4tn1OxCkveqTBdVViWNWmGWYe9qutimSbyr9ZKoQ3bK/ze0gq/YV+fh69Rtd0ZxXaZt9NPT+8tk6Ror8XqEV5eZo8OBXRk+Y8gDwAAAAAocCnJV3Vo+1ol7vtJ95zZrPCU31XdlGq3Jtnw0WHfWrpStqkCarRWeERb1SwWcJM9/qlaGX8dqD1aiw61lS7/+Vr4RfcMVI/w8hnvI+/GCPIAAAAAAIdLvHRBR7atUvKBdSpx7heFp+1XHdP1PxeYpEQV02H/+kop11T31mqrKvVbqq63T96/WLsYVZO03mpo48GzWr4uVp1bRal51SC3vhKfiSAPAAAAAMh3F+JPKG77SqUeWq/SF7Yo7PphNTBledmtSTqvEjp2T4SuhzRXmTrtFFqrsSLM5nybwexhUlRYSV3YYyjKwZ1kBYkgDwAAAAC4a2fiDujE9h+VfnSDgi9tVWXrCdk1gpmkk6ayOlW8kVSpucrV76AKVWqrtAdvppZXBHkAAAAAQJ4YVqviDuzUmZ0rZT6+URUTtytY5xR8w7ojHpV19t6GMle5T5UiOqpChTC71nncGYI8AAAAAOCW0q9f15HfN+n872vkfXKTQpN2qLISVTnLmuuGhw57VVVCqUj5Vm2t0IbtFVY62K51HvmDIA8AAAAAsJOakqwjO9br4t418j+9WVWu7VJV0zVVzbrG8NJBn5q6EtRUxaq3UpWG7VQ9oISzRi5SCPIAAAAAUMQlXbmkw9tWK2n/OgWc3azw1L2qabL8ucAkXTX8dMivrpLLNdW9NdsqrMF9quPr3m/j5q4I8gAAAABQxFy+EK8jW1cq5eBalbywRVUsB1XPZP1zgUlKUKCOFmugtArNVLpOW4XVaaYGnkRIV8BPAbny6aef6rnnntOlS5ecPcotDR06VJcuXdK///1vZ48CAAAAuIyzJ48obvuPSj+8XkEXtynMekwRWReYpDMqoxOBEUoPaa7g+h1UqVp9laRR3iUR5J1h9STJwyy1eSn7Yz+9K1nTpXYx+f5lhw4dqs8++0yS5OXlpUqVKmnw4MEaN26cPG/zL2v9+/dXt27d8n0mAAAAAPnLsFp18vBund65UorbqPKXt6qCEa+gG9Yd86io+BIN5RHaUhUatFe5yjWytc7DNRHkncHDLK1+K+P/s4b5n97N2N7uFYd96a5du2revHlKTU3V0qVLNXLkSHl5eSkm5tb/cODn5yc/Pz+HzQUAAADgzljT03V0z6869/tqeR7fqEpXd6iiLqpiljXphklHPKvofKlI+YTfp0oRHVS5bEW71nm4D+6TyA+GIaUl5f6j+Uip9YsZoX3V3zK2rfpbxuetX8x4PDf7MYw8j+rj46Pg4GBVrlxZzzzzjDp27Khvv/1WFy9e1ODBg3XvvffK399f999/vw4cOGB73qeffqoSJUrYPt+xY4fatWungIAABQYGKjIyUr/++qvt8a+//lp16tSRj4+PQkND9f7779vNERoaqrfffluPP/64AgICVKlSJc2ePdtuzfHjx9WvXz+VKFFCJUuWVM+ePXX06FHb4+np6YqOjlaJEiVUqlQpvfTSSzLu4HsCAAAAuBNLWqr2/bpKm754Tdvf7aorb4aoyuLOitozSZFX16iMLirNMGuPV21tLD9EO9p8oqTnD6nq+K1qNuJjNewyRKXKVrz9F4LL4op8frAkS2+Xv7Pnrn0v4+Nmn9/K2BN39jWz8PPz04ULFzR06FAdOHBA3377rQIDA/Xyyy+rW7du2r17t7y8vLI979FHH1XDhg314Ycfymw2a/v27bZ1W7ZsUb9+/fT666+rf//+2rBhg0aMGKFSpUpp6NChtn28//77evPNNzVu3DgtXrxYzzzzjNq0aaMaNWrIYrGoS5cuat68udatWydPT0/97W9/U9euXbVz5055e3vr/fff16effqq5c+eqVq1aev/99/XNN9+offv2d/19AQAAAFzFtaQrOrR9ja7sW6eA+FhVSdmjGqZUuzXJho8O+dbR1eCmCqzRRuERrVXL/x4nTQxHI8gXUYZhaOXKlfrhhx90//3369///rd+/vlntWjRQpI0f/58hYSE6N///rf69u2b7flxcXF68cUXVbNmTUlStWrVbI9NnjxZHTp00Pjx4yVJ1atX1+7du/Xee+/ZBflu3bppxIgRkqSXX35Z//znP7V69WrVqFFDCxculNVq1SeffCKTySRJmjdvnkqUKKE1a9aoc+fOmjJlimJiYtS7d29J0qxZs/TDDz/k/zcLAAAAKECJly7oyNaVSj6wTvee+0VVLPtV15T+5wKTdEn36Ih/faWWj1LJ2m0VVre56nn7OG9oFCiCfH7w8pfGncr789b/M+Pqu9lbSk/LuK3+vudz/3yzr5RyJU9f8n//+5/uueceWSwWWa1WDRw4UL1799b//vc/RUVF2daVKlVKNWrU0J49e3LcT3R0tJ544gl98cUX6tixo/r27avw8HBJ0p49e9SzZ0+79S1bttSUKVOUnp4us9ksSapfv77tcZPJpODgYJ09e1ZSxq37Bw8eVEBAgN1+UlJSdOjQIV2+fFmnT5+2m9nT01ONGzfm9noAAAC4lfNnjitu24+yHF6v0glbFXb9iBqYsvyd1iSdVUkdD4jQ9YrNFFSvvSrXaKSGf/y9GkUPQT4/mEySd7G8PeendzNCfLtXMgrvMovuzN45t9nnxGq9/ZobtGvXTh9++KG8vb1Vvnx5eXp66ttvv83zfl5//XUNHDhQ3333nb7//ntNmDBBCxYs0EMPPZTrfdx4y77JZJL1j2O6evWqIiMjNX/+/GzPK1OmTJ7nBQAAAFyBYbXqdNwBndz+o4xjG1Tu0laFGKdUOusik3TCVE6nijeUKrdQhQYdVT60hoJ4Kzj8gSDvDFnb6TNDe+Z/c2qzz0fFihVT1apV7bbVqlVL169fV2xsrO3W+gsXLmjfvn2qXbv2TfdVvXp1Va9eXc8//7weeeQRzZs3Tw899JBq1aqln3/+2W7tzz//rOrVq9uuxt9Oo0aNtHDhQgUFBSkwMDDHNeXKlVNsbKxat24tSbp+/bq2bNmiRo0a5eprAAAAAI5mWK2K27dNZ3atlvn4RlVM3K7yOq+sDVtWw6Rj5so6W7KRPKvcp8oRHVWxfGVRR4ebIcg7gzXdPsRnyvzcmp79OQ5UrVo19ezZU8OHD9dHH32kgIAAjR07VhUqVMh2i7wkXbt2TS+++KL69OmjsLAwnThxQr/88osefvhhSdILL7ygJk2a6M0331T//v21ceNGzZgxQx988EGuZ3r00Uf13nvvqWfPnnrjjTdUsWJFHTt2TEuWLNFLL72kihUrasyYMfr73/+uatWqqWbNmpo8ebIuXbqUX98WAAAAIM+uW9J05PdYXfh9tXxOxSo0aacqK9Hubd4shlmHvarqYunG8q3aSmEN2yusVFmFOW1quBuCvDO0u8V7tjvoSvztzJs3T2PGjNEDDzygtLQ0tW7dWkuXLs2xsd5sNuvChQsaPHiw4uPjVbp0afXu3VsTJ06UlHE1/V//+pdee+01vfnmmypXrpzeeOMNu6K72/H399fatWv18ssvq3fv3rpy5YoqVKigDh062K7Qv/DCCzp9+rSGDBkiDw8PPf7443rooYd0+fLlfPmeAAAAALeTci1Jh3esU+LetfI/E6vwa7tUzZSialnXGF465FNbiWWbKKB6a1Vp2FY17inurJFRCJgMmsGySUxMVPHixXX58uVst3WnpKToyJEjCgsLk6+vr5MmzGC1WpWYmKjAwEB58HqZXHGlnx/gDiwWi5YuXapu3brl+A97AHC3OM/A3VxNvKgj21br6v61Kn72F4Wn7ZOPyWK3JlH+OuJXV8nlmunemq1VpUErefvwd09ncZfzzK1y6I24Ig8AAAAAN3Hx3Gkd3bZSqYfWqdSFLQqzHFI9U5bSaZN0QcV17J4GSqvQTGXqtFNo7aZq4EnUguPwqwsAAAAA/hB/4pCOb/tR6Uc3qOzFLQq1Hte9WReYpFOmIJ0MjJBRqYXK1W+viuH1VIo7ZFGACPIAAAAAiiTDatWJQ7/p9M7VMsVtUIXEbSpvnFXZG9Yd8wjRmXsbyRzaUhUjOqh8SFW71nmgoBHkAQAAABQJ6dev6+ieX3Vu1yp5n9yoSld3KkSXFJJ1jWHSYc9wXSjdWN5V7lNow/aqHFTBrnUecDaC/B2iI9A98XMDAAAoOtJSU3R453pd3POT/E/HKuzabwpXssKzrjE8ddC7pi4HNVaxaq0V1rCdqhUvadc6D7gagnweZbYcJicny8/Pz8nTIK+Sk5MlyaXbKgEAAHBnriVd0aFtq3Vl31oFxm9WldQ9qmlKs1uTZPjqkF8dJQU3VfEabVQlorVq+xVz0sTAnSHI55HZbFaJEiV09uxZSRnvd24ymZwyi9VqVVpamlJSUnj7udswDEPJyck6e/asSpQoIbPZ7OyRAAAAcJcuJ5zTkW0rlXJgre49v0VVLAdU15T+5wKTdFEBOupfX6kVolSqdjuF1W2m+l7ezhsayAcE+TsQHBwsSbYw7yyGYejatWvy8/Nz2j8muJsSJUrYfn4AAABwL+dPHdOx7T/q+uGfVSZhi0LTjynClOWlkybprEoqLrCR0is2U3C99gqpHqGGXMRBIUOQvwMmk0nlypVTUFCQLBaL0+awWCxau3atWrduza3iueDl5cWVeAAAADdhWK06dXSfTu1cKePozyp3ebtCjFMqnXWRSTpuKq/TJRrKVLmlKjTooHKVqyuIu1VRyBHk74LZbHZqMDSbzbp+/bp8fX0J8gAAAHBr1vR0xe3fpvjfVsl8fJMqXdmmCkpQhaxrDJOOeIbqfMlIeYbdp8qNOigkuJJd6zxQFBDkAQAAABS465Y0Hf5toxJ2r5bPqViFJe9UqK4qNMsai2HWIa/qulgmUv7VWim0YUeF31varnUeKIoI8gAAAAAcLiX5qg5tX6vEfT/pnjObVSVlt6qbUuzWXDO8dci3tq6UbaqA6q0U3rCdahYLcNLEgOsiyAMAAADId1cuJ+jI1lVKOrBWJc79qvC0fapjuv7nApOUqGI67FdPKeWaqkTttqpSr6Xq+vg6b2jATRDkAQAAANy1hLMndWzbj0o99LNKX/hVYdcPq/4NjfLnVUJx9zSQpWJzlanbXqG1GiuCMmIgzwjyAAAAAPLsTNwBndi+UulHf1bwpa2qbD2hklkXmKRTprI6GdhQqtxCwfXbq2KVOipNozxw11wiyM+cOVPvvfeezpw5owYNGmj69Olq2rTpTddPmTJFH374oeLi4lS6dGn16dNHkyZNkq+v7x3vEwAAAEDODKtVxw/u1Jmdq+QRt0EVErernM4p+IZ1Rz0qKf7eRjKHtVRIRAeVrxiu8k6ZGCjcnB7kFy5cqOjoaM2aNUtRUVGaMmWKunTpon379ikoKCjb+q+++kpjx47V3Llz1aJFC+3fv19Dhw6VyWTS5MmT72ifAAAAAP6Ufv26jvweq/O/r5b3yVhVTtqhSrqsSlnWXDc8dNirqhJKRcon/D6FNeqo0NLBdq3zABzD6UF+8uTJGj58uIYNGyZJmjVrlr777jvNnTtXY8eOzbZ+w4YNatmypQYOHChJCg0N1SOPPKLY2Ng73icAAABQlKWmJOvIjvW6uPcn+Z+OVZVru1TVdE1Vs64xvHTIp6YuBzXRPdVaKaxhO1UPvNdpMwNFmVODfFpamrZs2aKYmBjbNg8PD3Xs2FEbN27M8TktWrTQl19+qc2bN6tp06Y6fPiwli5dqscee+yO95mamqrU1FTb54mJiZIki8Uii8Vy18fpKJmzufKMANwb5xkAjsZ5xjmSr17W0R1rlXRgnYqf+1XhqXtU05TlZ2CSrhp+OuRXR0llmyiwRmuF1rtP1Xz97PbDzw3uwF3OM3mZz6lB/vz580pPT1fZsmXttpctW1Z79+7N8TkDBw7U+fPndd9998kwDF2/fl1PP/20xo0bd8f7nDRpkiZOnJht+/Lly+Xv738nh1agVqxY4ewRABRynGcAOBrnGceyXLuq9AsHFJi4T5VT96macVQNTOl/LjBJCUaA9nnW0Bn/GkorUUPeJUPk4ZHRKH/unEWHVq120vRA/nD180xycnKu1zr91vq8WrNmjd5++2198MEHioqK0sGDBzVmzBi9+eabGj9+/B3tMyYmRtHR0bbPExMTFRISos6dOyswMDC/Rs93FotFK1asUKdOneTl5eXscQAUQpxnADga5xnHOHfqqE7uXK30oxtU9uJWhVmP2S8wSWdUWscDIpQe0kxlardVSLX6akyjPAohdznPZN4ZnhtODfKlS5eW2WxWfHy83fb4+HgFB9/YgZlh/Pjxeuyxx/TEE09IkurVq6ekpCQ9+eSTeuWVV+5onz4+PvLx8cm23cvLy6V/0JncZU4A7ovzDABH4zxz5wyrVaeO7tHJHStlOrZB5S5vU0XjTLa2+DiPCjpTopFMlVuoYkQHlatcI1vrPFCYufp5Ji+zOTXIe3t7KzIyUitXrlSvXr0kSVarVStXrtSoUaNyfE5ycrI8bviXQrM545YfwzDuaJ8AAACAu7Cmp+vY3l91dtdqeZ7YpEpXtquCLqpC1jWGSYc9q+h8yUbyDr9PlSI6qFJwiF3rPAD35fRb66OjozVkyBA1btxYTZs21ZQpU5SUlGRrnB88eLAqVKigSZMmSZJ69OihyZMnq2HDhrZb68ePH68ePXrYAv3t9gkAAAC4C0taqg7v/FkX9/4k31OxCkveqTAlKSzLmjTDrMPeNXSxTGP5V22l0IbtVfXe0nat8wAKD6cH+f79++vcuXN67bXXdObMGUVERGjZsmW2srq4uDi7K/CvvvqqTCaTXn31VZ08eVJlypRRjx499NZbb+V6nwAAAICrSkm+qkPbflLivp90T/xmhafsVg1Tqt2aZMNHh3xr62rZpgqs0UbhDduopv89TpoYQEEzGYZhOHsIV5OYmKjixYvr8uXLLl92t3TpUnXr1s2lX+sBwH1xngHgaJxnpMRLF3Rk60olH1yne8/9qipp++SdtVFe0mUV0xH/+kopH6WStdsprG5zeXln73gCkJ27nGfykkOdfkUeAAAAKEouxJ/QsW0/Ku3QepVJ2KKw60fUwJTl2ppJOqd7FRcQoesVmyuobltVrtlYEX+8jBQACPIAAACAA50+tk8ntq+UcfRnBV/epkrWkyqVdYFJOmEK1qnijaTKzVWhQQeVD62lMrwVHICbIMgDAAAA+cSwWhW3f7vO/LZK5uMbVTFxu8rpvMrdsO6IR6jOlmwkz7CWqtSwoyqWD1VFp0wMwB0R5AEAAIA7dN2SpiO/x+rC7jXyOblJoUk7VVmJqpx1jeGhw15VlVC6iXzD71NYow4KK1XWrnUeAPKCIA8AAADkUmpKsg7vWKfLe36S/5lYVbn2u6qZrqlaljUphpcO+dTSlaAmKla9jao0bKPqASWcNTKAQoggDwAAANzE1cSLOrJtta4eWKfiZ39ReOpe1TJZ/lxgkq4YfjrsV1fJ5aJ0b622CqvfUnV8/Z03NIBCjyAPAAAA/OHS+TM6svVHpR5ar1IXflWY5ZDqmax/LjBJCQrU0WIRSqvYTKVrt1VYnSg18OSv1QAKDmccAAAAFFnxJw7p+PaVSj/ys8pe3KpQa5waZl1gkk6rjE4UbygjpLmC67dXSNX6KkmjPAAnIsgDAACgSDCsVp04/LtO71gpU9xGVUjcpvJGvMresO6YR4jOlGgoc2hLVYzooHKVqmVrnQcAZyLIAwAAoFCypqfryO5fdP731fI6sVGVru5QiC4pJMuadMOkI55VdL5UY/mEt1SliA6qXLaiXes8ALgagjwAAAAKhbTUFB3+7Wdd2r1Gvqc3q8q13xSuJIVnXWN46pB3DV0KaqJiVVsprFF7VS1eUlWdNjUA5B1BHgAAAG7pWtIVHdq2Wlf2r1NA/GaFp+xWTVOa3Zpkw0eHfOvoanBTFa/ZVlUiWquWXzEnTQwA+YMgDwAAAKdItxqKPZKgLedNKnUkQc2rBsnsYbrp+ssJ53Rk20pdO7hOJc/9qiqWA6prSv9zgUm6pHt0xL++Uis0U8labVWlXnPV8/IugKMBgIJDkAcAAEDBWj1JB84la/Chtjp9OUWSWZ8f+FXlivvq8/A1qlbGX2oXo/Nn4nRs60pdP7xOpS9uVdj1o4owGX/uxySdVUnFBTRUekgzla3XXpWqN1RDs9lZRwYABYIgDwAAgAJ14Fyyqu2epj6WU5qu3n9sNfR40seqtvt7HfaqKq+1nyvEOKXSWZ9oko6byut08QiZQluqfP0OKh9aQ0G8FRyAIoYgDwAAgAKTbjU0+FBb9bGc0gtei1XZFC8vU7rae2xTgOmaJKmK5aAkyWqYdNRcWedKRsqzSktVjuiokPKV7VrnAaAoIsgDAACgwGw+kqDTl1M0Xb1VyRSvvp7rbI+lGybtMMK12VpTFSM6qFX7HqpSsoyqOHFeAHBFBHkAAAAUmLNXUiRJFU3n1MG8zbbdYphVP/VjXZOvJGlq1QgVL1nGKTMCgKvjBUUAAAAoMEEBvvJVqj7ymqySpquSMt7b3cuUrifMS+3WAQByRpAHAABAgWlSuYT+4f2x6ngckyTNtnRX9dTP9b6lj17wWqzR5iUqV9xXTcNKOnlSAHBd3FoPAACAAvPrgjf1gMcGSdLC6230dvqjkqTp6b1lkhTttVg9wsvL7NHBiVMCgGsjyAMAAKBA/Lb2P2pyYIpkkvbc01xTLGOkyym2xxfdM1A9wstnvI88AOCmCPIAAABwuFNH9ipk1QiZTYZ+KXG/Go/+Sutl0saDZ7V8Xaw6t4pS86pBXIkHgFwgyAMAAMChriVd0bUvH1F5XdUBz2qq99QcmTw8ZJYUFVZSF/YYigorKbOHydmjAoBboOwOAAAADmNYrfp91hCFpx/WBRVXwJAF8vUr5uyxAMCtEeQBAADgMLFfvaHGV1bKYpgV3+UjBYdUdfZIAOD2CPIAAABwCFu5naSttV5U7eb3O3cgACgkCPIAAADIdzeW2zXt97KzRwKAQoMgDwAAgHyVUW43QCVuKLcDAOQPzqgAAADIN3+W2x2h3A4AHIQgDwAAgHxDuR0AOB5BHgAAAPmCcjsAKBgEeQAAANw1yu0AoOAQ5AEAAHBXspbb7fesTrkdADgYZ1gAAADcMcNq1e5Zg23ldsWHUm4HAI5GkAcAAMAdi/1qoiKvrLKV25WtGO7skQCg0CPIAwAA4I5klNtNlUS5HQAUJII8AAAA8oxyOwBwHoI8AAAA8oRyOwBwLs64AAAAyDXK7QDA+QjyAAAAyDW7cruusym3AwAnIMgDAAAgV35b+82f5Xa1X1LtZl2dPBEAFE0EeQAAANxWRrndKJlNhjaX6KamfV9y9kgAUGQR5AEAAHBLN5bb1X/qE8rtAMCJOAMDAADgpii3AwDX4xJBfubMmQoNDZWvr6+ioqK0efPmm65t27atTCZTto/u3bvb1ly9elWjRo1SxYoV5efnp9q1a2vWrFkFcSgAAACFCuV2AOB6nB7kFy5cqOjoaE2YMEFbt25VgwYN1KVLF509ezbH9UuWLNHp06dtH7t27ZLZbFbfvn1ta6Kjo7Vs2TJ9+eWX2rNnj5577jmNGjVK3377bUEdFgAAgNuj3A4AXJPTg/zkyZM1fPhwDRs2zHbl3N/fX3Pnzs1xfcmSJRUcHGz7WLFihfz9/e2C/IYNGzRkyBC1bdtWoaGhevLJJ9WgQYNbXukHAADAnyi3AwDX5enML56WlqYtW7YoJibGts3Dw0MdO3bUxo0bc7WPOXPmaMCAASpW7M/XarVo0ULffvutHn/8cZUvX15r1qzR/v379c9//jPHfaSmpio1NdX2eWJioiTJYrHIYrHcyaEViMzZXHlGAO6N8wxQNCVfvayUL/qrvK5qv7maaj/+oa6np0vp6fn+tTjPAHA0dznP5GU+pwb58+fPKz09XWXLlrXbXrZsWe3du/e2z9+8ebN27dqlOXPm2G2fPn26nnzySVWsWFGenp7y8PDQxx9/rNatW+e4n0mTJmnixInZti9fvlz+/v55OCLnWLFihbNHAFDIcZ4Big7DaqjUrg/VynpUF4xA7agyUntW/+Twr8t5BoCjufp5Jjk5OddrnRrk79acOXNUr149NW3a1G779OnTtWnTJn377beqXLmy1q5dq5EjR6p8+fLq2LFjtv3ExMQoOjra9nliYqJCQkLUuXNnBQYGOvw47pTFYtGKFSvUqVMneXl5OXscAIUQ5xmg6Pnl/95Ui/RNshhmneo8S72adnbo1+M8A8DR3OU8k3lneG44NciXLl1aZrNZ8fHxdtvj4+MVHBx8y+cmJSVpwYIFeuONN+y2X7t2TePGjdM333xja7KvX7++tm/frn/84x85BnkfHx/5+Phk2+7l5eXSP+hM7jInAPfFeQYoGn5b+42iDk2TTBnldlEtu9/+SfmE8wwAR3P180xeZnNq2Z23t7ciIyO1cuVK2zar1aqVK1eqefPmt3zuokWLlJqaqkGDBtltz3xdu4eH/aGZzWZZrdb8Gx4AAKAQodwOANyH02+tj46O1pAhQ9S4cWM1bdpUU6ZMUVJSkoYNGyZJGjx4sCpUqKBJkybZPW/OnDnq1auXSpUqZbc9MDBQbdq00Ysvvig/Pz9VrlxZP/30kz7//HNNnjy5wI4LAADAXdiV23lWV/2nPpHJw+lvbgQAuAmnB/n+/fvr3Llzeu2113TmzBlFRERo2bJltgK8uLi4bFfX9+3bp/Xr12v58uU57nPBggWKiYnRo48+qoSEBFWuXFlvvfWWnn76aYcfDwAAgDsxrFbt+WiIIq1HdUHFVXzoAvn6Fbv9EwEATuP0IC9Jo0aN0qhRo3J8bM2aNdm21ahRQ4Zh3HR/wcHBmjdvXn6NBwAAUGjFzn9dza6slsUwK/7+2apdMdzZIwEAboN7pgAAAIqo39Z+oyYHp0nKKLer3ayrkycCAOQGQR4AAKAIotwOANwXQR4AAKCIySy3K0G5HQC4Jc7YAAAARUhmuV0Vyu0AwG0R5AEAAIqQ2PmvKzKz3K7rbJWl3A4A3A5BHgAAoIiwL7d7mXI7AHBTBHkAAIAi4OThPaq0amSWcrsXnT0SAOAOEeQBAAAKueSrl5X65QAVVxLldgBQCHAGBwAAKMQotwOAwocgDwAAUIhRbgcAhQ9BHgAAoJCi3A4ACieCPAAAQCFEuR0AFF4EeQAAgEIm+eplpVFuBwCFFmd0AACAQiSz3C6McjsAKLQI8gAAAIUI5XYAUPgR5AEAAAoJyu0AoGggyAMAABQClNsBQNFBkAcAAHBzlNsBQNHCGR4AAMCNGVar9swarDDrUZ1XCZUY9i/K7QCgkCPIAwAAuLHYLyco8uoaWQyzzt0/W0EVwpw9EgDAwQjyAAAAbuq3n5aoyaHpkjLK7WpFdXHyRACAgkCQBwAAcEMnD+9RpdWjKLcDgCKIIA8AAOBmKLcDgKKNMz4AAIAbodwOAECQBwAAcCOU2wEACPIAAABuwq7crs5Yyu0AoIgiyLupdKuh2CMJ2nLepNgjCUq3Gs4eCQAAOJBdud293dW0z1+dPRIAwEk8nT0A8mj1JB04l6zBh9rq9OUUSWZ9fuBXlSvuq8/D16haGX+pXYyzpwQAAPkoW7ndkx9TbgcARRh/AriZA+eSVW33NPW5+pXd9r5Xv1K13dN04FyykyYDAACOQLkdAOBGXJF3I+lWQ4MPtVUfyym94LVYAbqmII+LOm4to2e9/qPJlj5adKit1lsNmT1Mzh4XAADkg9gvJ6hZZrldt9mqRbkdABR5BHk3svlIgk5fTtF09ZYkveC1OOMBszTd0lPT0ntLl1O0+UiCmoeXcuKkAAAgP9jK7UwZ5XZRlNsBAMSt9W7l7JUU2/9PT+8ti2G2fd7cvEf3KDnbOgAA4J5OHv6dcjsAQI4I8m4kKMDX9v/PmpfIy5RuC/ONPfbrU+93VUzX7NYBAAD3k1FuN1DFlaR9njXU4KlPKLcDANjwJ4IbaRpWUuWK+2q0eYle8Fqs9y19VC31C31xvYOkjDD/ufc7ql2K18cDAOCubiy3u3fYQvn4+jt7LACACyHIuxGzh0mfh69RtNdiTbb00fT0jNfKj7/+F335R5iP9Niv0zO762riRSdOCgAA7lTslxMUmVlud/9sBVFuBwC4AUHezVQr468DtUdr0T0D7bbPLDZSv5d/WKnyUk3Lbp2Y3o0wDwCAm7GV2ymj3K4W5XYAgBzQWu9u2sWomqT1VkMbD57V8nWx6twqSs2rBsns0UEHtq9T2X/3V03Lbu2Z3k0hzy7VPYH3OntqAABwG5TbAQByiyvybsrsYVJUWElFljYUFVbS9r7x1SJaKb7XQiWqmGpZdus4V+YBAHB5lNsBAPKCPyEKoWoRrRTfcwFhHgAAN0C5HQAgrwjyhVS1hq0J8wAAuIFNWcrtzt//MeV2AIDbIsgXYtnDPG32AAC4kp1rvlZUlnK7mlGdnTwRAMAdEOQLuT/DvL9qWX4nzAMA4CJOHv5dldc8Kw/K7QAAeUSQLwIywvxCwjwAAC6CcjsAwN3gT4wiIqcwn3TlkrPHAgCgyKHcDgBwtwjyRciNYT5uWjfCPAAABSyz3C6NcjsAwB1yiSA/c+ZMhYaGytfXV1FRUdq8efNN17Zt21YmkynbR/fu3e3W7dmzRw8++KCKFy+uYsWKqUmTJoqLi3P0obi8ag1b68yDCwjzAAA4QdZyu211Yii3AwDcEacH+YULFyo6OloTJkzQ1q1b1aBBA3Xp0kVnz57Ncf2SJUt0+vRp28euXbtkNpvVt29f25pDhw7pvvvuU82aNbVmzRrt3LlT48ePl6+vb0Edlkur3qiNXZg/No3b7AEAcLSTh39XqK3c7gE17fOCs0cCALgppwf5yZMna/jw4Ro2bJhq166tWbNmyd/fX3Pnzs1xfcmSJRUcHGz7WLFihfz9/e2C/CuvvKJu3brp3XffVcOGDRUeHq4HH3xQQUFBBXVYLi9rmK9t2UWYBwDAgTLL7QJt5XYfU24HALhjns784mlpadqyZYtiYmJs2zw8PNSxY0dt3LgxV/uYM2eOBgwYoGLFikmSrFarvvvuO7300kvq0qWLtm3bprCwMMXExKhXr1457iM1NVWpqam2zxMTEyVJFotFFovlDo/O8TJnu9MZw+q10MHrX0pLB6m2ZZd2T+um8s98q2IBxfNzTABu7G7PMwD+LLeL/KPcLuCxL+Vh9uL31R84zwBwNHc5z+RlPpNhGIYDZ7mlU6dOqUKFCtqwYYOaN29u2/7SSy/pp59+Umxs7C2fv3nzZkVFRSk2NlZNmzaVJJ05c0blypWTv7+//va3v6ldu3ZatmyZxo0bp9WrV6tNmzbZ9vP6669r4sSJ2bZ/9dVX8vcv/C2yKecO6f7j7ynQlKydpho6WPsFmb15GQIAAPli/3fqmbRQaYZZ31SIkW/Z6s6eCADggpKTkzVw4EBdvnxZgYGBt1zr1Cvyd2vOnDmqV6+eLcRLGVfkJalnz556/vnnJUkRERHasGGDZs2alWOQj4mJUXR0tO3zxMREhYSEqHPnzrf9BjqTxWLRihUr1KlTJ3l5ed3Vvg5uayDTd4+qvvbJc/80lX/mP1yZB5Cv5xmgKPp93Teqf/VfkknaUutl9X74OWeP5HI4zwBwNHc5z2TeGZ4bTg3ypUuXltlsVnx8vN32+Ph4BQcH3/K5SUlJWrBggd54441s+/T09FTt2rXttteqVUvr16/PcV8+Pj7y8fHJtt3Ly8ulf9CZ8mPOWk07aL/nAuk/AzJus5/VS5Wf/Z+KBZTInyEBuDV3OR8CruTk4d9VZe1ztnK7Zv1e5HXxt8B5BoCjufp5Ji+zOfVPE29vb0VGRmrlypW2bVarVStXrrS71T4nixYtUmpqqgYNGpRtn02aNNG+ffvstu/fv1+VK1fOv+ELoeqN2up0zwW6YvipdtpvOjb9AQrwAAC4A0lXLinty0cotwMAOMQd/Yly6dIlffLJJ4qJiVFCQoIkaevWrTp58mSe9xUdHa2PP/5Yn332mfbs2aNnnnlGSUlJGjZsmCRp8ODBdmV4mebMmaNevXqpVKlS2R578cUXtXDhQn388cc6ePCgZsyYof/+978aMWJEnucranIK88lXLzt7LAAA3IZhtWrfR4MVZj2m8yqhe4ctlI9v4e/cAQAUnDzfWr9z50517NhRxYsX19GjRzV8+HCVLFlSS5YsUVxcnD7//PM87a9///46d+6cXnvtNZ05c0YRERFatmyZypYtK0mKi4uTxw3/gr1v3z6tX79ey5cvz3GfDz30kGbNmqVJkyZp9OjRqlGjhr7++mvdd999eT3cIql6o7barz9us0/7TbundVfo6O/kfw+vmQcA4HY2fTlBza/+pDTDrPPdPlbNCmHOHgkAUMjkOchHR0dr6NChevfddxUQEGDb3q1bNw0cOPCOhhg1apRGjRqV42Nr1qzJtq1GjRq6Xdn+448/rscff/yO5kFGmN9n/Url/zuQMA8AQC7tXPO1og5Nl0zStjoxiorq7OyRAACFUJ5vrf/ll1/01FNPZdteoUIFnTlzJl+Ggmuo0bi9TvX4ynab/ZFp3bnNHgCAmzh5+HeFrnnWVm7XtM8Lzh4JAFBI5TnI+/j45FiLv3//fpUpUyZfhoLryBrm6xDmAQDIkX25XU3K7QAADpXnP2EefPBBvfHGG7JYLJIkk8mkuLg4vfzyy3r44YfzfUA4X43G7XWyx3zCPAAAOcgot3vMVm5X8nHK7QAAjpXnIP/+++/r6tWrCgoK0rVr19SmTRtVrVpVAQEBeuuttxwxI1xAzcYdbgjztNkDACBJm74Yr0ZX12aU293/scqUD3X2SACAQi7PZXfFixfXihUr9PPPP2vHjh26evWqGjVqpI4dOzpiPriQmo07aK/mq+J/H1WdtJ36fdoDChv9PwrwAABF1s7VixV1eCbldgCAApWnIG+xWOTn56ft27erZcuWatmypaPmgosizAMAkOHk4d8V+tNoyu0AAAUuT7fWe3l5qVKlSkpPT3fUPHADNRt30Ike83XV8FOdtJ3cZg8AKHIotwMAOFOe/8R55ZVXNG7cOCUkJDhiHriJ7GG+B2EeAFAkUG4HAHC2PL9GfsaMGTp48KDKly+vypUrq1ixYnaPb926Nd+Gg2uzv81+h36f1kNho//LbfYAgEJt0xfj1Tyz3K77J6pJuR0AoIDlOcj36tXLAWPAXdVs3EF7jS9V8X+DbGG+ypjv5FcswNmjAQCQ7+zK7eqOU1TTTs4eCQBQBOU5yE+YMMERc8CN1WzSUXuVJcxP7U6YBwAUOnbldiV7KKrvX509EgCgiLrjVpYtW7boyy+/1Jdffqlt27bl50xwQzWbdNSJB7784zXzO3R4anddS7ri7LEAAMgX2crtnpzt7JEAAEVYnq/Inz17VgMGDNCaNWtUokQJSdKlS5fUrl07LViwQGXKlMnvGeEmajbpqL3GF6r43WNcmQcAFBqZ5XaNKLcDALiIPF+Rf/bZZ3XlyhX9/vvvSkhIUEJCgnbt2qXExESNHj3aETPCjdRs2kknun9huzJ/aNoDXJkHALi1TV+MV6PMcrtun6gM5XYAACfLc5BftmyZPvjgA9WqVcu2rXbt2po5c6a+//77fB0O7ikzzCcZvqqbup0wDwBwW7ZyO2WU29Wk3A4A4ALyHOStVqu8vLyybffy8pLVas2XoeD+ajbtpOPdvyTMAwDc1omDuyi3AwC4pDwH+fbt22vMmDE6deqUbdvJkyf1/PPPq0OHDvk6HNwbYR4A4K6SrlzS9a8otwMAuKY8B/kZM2YoMTFRoaGhCg8PV3h4uMLCwpSYmKjp06c7Yka4McI8AMDdZJbbhVrjKLcDALikPLfWh4SEaOvWrfrxxx+1d+9eSVKtWrXUsWPHfB8OhUPNpp20V18q5LtBqpu6XbumPaDw0f+jzR4A4JI2fTFezTPL7bp/opqU2wEAXEyeg7wkmUwmderUSZ06UfiC3KnZtJP2Gl8oZOljhHkAgMuylduZMsrtoii3AwC4oDzfWj969GhNmzYt2/YZM2boueeey4+ZUEjVjOqs493+bLM/OK0Ht9kDAFwG5XYAAHeR5yD/9ddfq2XLltm2t2jRQosXL86XoVB4ZQ3z9VK3EeYBAC4ha7ndXs9alNsBAFxanoP8hQsXVLx48WzbAwMDdf78+XwZCoVbzajOiuv2uV2YT0m+6uyxAABF1I3ldqUeX0C5HQDApeU5yFetWlXLli3Ltv37779XlSpV8mUoFH61orrYhfkDUx8gzAMAnGLTF6+qUWa5XbdPVIZyOwCAi8tz2V10dLRGjRqlc+fOqX379pKklStX6v3339eUKVPyez4UYrWiumiPPlelpYNVL3Wbfpv6gKqN+Z98/e9x9mgAgCIio9zuA8rtAABuJc9B/vHHH1dqaqreeustvfnmm5Kk0NBQffjhhxo8eHC+D4jCLTPMV176GGEeAFCgKLcDALirPN9aL0nPPPOMTpw4ofj4eCUmJurw4cOEeNyxWlFddKzbF0o2fLjNHgBQICi3AwC4szsK8pnKlCmjLVu26Pvvv9fFixfzayYUQdnDPAV4AADHoNwOAODuch3k33nnHY0fP972uWEY6tq1q9q1a6fu3burVq1a+v333x0yJIoG+zC/lTAPAHAIyu0AAO4u10F+4cKFqlu3ru3zxYsXa+3atVq3bp3Onz+vxo0ba+LEiQ4ZEkVHraguOnr/54R5AIBD7Fi9KKPcTtK2uq+oJuV2AAA3lOsgf+TIEdWvX9/2+dKlS9WnTx+1bNlSJUuW1KuvvqqNGzc6ZEgULbWbdbUL8/unPUiYBwDctRMHdynspzHyMBmKLfmgovq+4OyRAAC4I7kO8tevX5ePj4/t840bN6pFixa2z8uXL6/z58/n73QosrKG+fopWwjzAIC7cmO5XcSTHzl7JAAA7liug3x4eLjWrl0rSYqLi9P+/fvVunVr2+MnTpxQqVKl8n9CFFm1m3XV0a6fEeYBAHeFcjsAQGGT6yA/cuRIjRo1Sn/5y190//33q3nz5qpdu7bt8VWrVqlhw4YOGRJFV+3m998Q5nsq5VqSs8cCALgRyu0AAIVNroP88OHDNW3aNCUkJKh169b6+uuv7R4/deqUHn/88XwfELAP879q/9QHCfMAgFyh3A4AUBh55mXx448/ftOw/sEHH+TLQEBOaje/X7v1mUKXDVH9lF+1c+qDqj7mW/n6FXP2aAAAF0W5HQCgsMr1FXnA2bgyDwDIrauJFym3AwAUWgR5uBXCPADgdgyrVfs/Gky5HQCg0CLIw+0Q5gEAt7Lpi1fVKIlyOwBA4UWQh1uq3fx+HenyKWEeAGCHcjsAQFFwx0H+4MGD+uGHH3Tt2jVJkmEY+TYUkBt1WnS7Iczz1nQAUJRRbgcAKCryHOQvXLigjh07qnr16urWrZtOnz4tSfrLX/6iF17gD0wULPsw/wthHgCKKMrtAABFSZ6D/PPPPy9PT0/FxcXJ3//P4pj+/ftr2bJl+TockBt1WnTTkc7zCPMAUERlLbc7p3tV+vGFlNsBAAq1PAf55cuX65133lHFihXttlerVk3Hjh3Lt8GAvKjTsrtdmN9HmAeAIiP28z/L7S50/0Sly1d29kgAADhUnoN8UlKS3ZX4TAkJCfLx8bmjIWbOnKnQ0FD5+voqKipKmzdvvunatm3bymQyZfvo3r17juuffvppmUwmTZky5Y5mg/vIGuYbEOYBoEjYsXqRmh7JUm7XpKOTJwIAwPHyHORbtWqlzz//3Pa5yWSS1WrVu+++q3bt2uV5gIULFyo6OloTJkzQ1q1b1aBBA3Xp0kVnz57Ncf2SJUt0+vRp28euXbtkNpvVt2/fbGu/+eYbbdq0SeXLl8/zXHBPhHkAKDootwMAFFV5DvLvvvuuZs+erfvvv19paWl66aWXVLduXa1du1bvvPNOngeYPHmyhg8frmHDhql27dqaNWuW/P39NXfu3BzXlyxZUsHBwbaPFStWyN/fP1uQP3nypJ599lnNnz9fXl5eeZ4L7iszzF8zvAnzAFBIUW4HACjKPPP6hLp162r//v2aMWOGAgICdPXqVfXu3VsjR45UuXLl8rSvtLQ0bdmyRTExMbZtHh4e6tixozZu3JirfcyZM0cDBgxQsWLFbNusVqsee+wxvfjii6pTp85t95GamqrU1FTb54mJiZIki8Uii8WS28MpcJmzufKMzlK9aWfttc5R1R//ogYpv2j71F6qOmKRfPyK3f7JAGw4z8AVGVar9n30mCL/KLcrPmS+PMxe/Dp1U5xnADiau5xn8jJfnoK8xWJR165dNWvWLL3yyit5HuxG58+fV3p6usqWLWu3vWzZstq7d+9tn79582bt2rVLc+bMsdv+zjvvyNPTU6NHj87VHJMmTdLEiROzbV++fHmOfQCuZsWKFc4ewWXtLv9XPXjqH4pI2azNUx7UiTrPyuzp7eyxALfDeQauxLTvv3oweZ3SDLNWVRwl3627JO1y9li4S5xnADiaq59nkpOTc702T0Hey8tLO3fuzPNAjjJnzhzVq1dPTZs2tW3bsmWLpk6dqq1bt8pkMuVqPzExMYqOjrZ9npiYqJCQEHXu3FmBgYH5Pnd+sVgsWrFihTp16sTLB26qm/ZsrKlqK59QU+sOeR+Yraojv5aPr5+zBwPcAucZuJpdPy1Wg6TFkknaWmecej80xtkj4S5xngHgaO5ynsm8Mzw38nxr/aBBgzRnzhz9/e9/z+tTsyldurTMZrPi4+PttsfHxys4OPiWz01KStKCBQv0xhtv2G1ft26dzp49q0qVKtm2paen64UXXtCUKVN09OjRbPvy8fHJsXHfy8vLpX/QmdxlTmep37qndpk9FL78cUWkbNaODx5WzTH/4T2GgTzgPANXcOLgLoWvj84otyvVU836/dXZIyEfcZ4B4Giufp7Jy2x5DvLXr1/X3Llz9eOPPyoyMtLutelSRnldbnl7eysyMlIrV65Ur169JGW8vn3lypUaNWrULZ+7aNEipaamatCgQXbbH3vsMXXsaP/WM126dNFjjz2mYcOG5Xo2FC51W/bQLs1V+PLH1eDaZu2Y2pMwDwBu5M9yu+SMcrvhs5w9EgAATpPnIL9r1y41atRIkrR//367x3J7K3tW0dHRGjJkiBo3bqymTZtqypQpSkpKsoXuwYMHq0KFCpo0aZLd8+bMmaNevXqpVKlSdttLlSqVbZuXl5eCg4NVo0aNPM+HwoMwDwDuybBatf+jwWr0R7ld6ccXcu4GABRpeQ7yq1evztcB+vfvr3Pnzum1117TmTNnFBERoWXLltkK8OLi4uThYf8uefv27dP69eu1fPnyfJ0FhV/dlj20y5ij8BV/+SPM91LNMf/mL4QA4MJiP39VzZLWKs0w68IDn6hm+crOHgkAAKfKc5DP6sSJE5KkihUr3tUQo0aNuumt9GvWrMm2rUaNGjIMI9f7z+l18Si66t73oHYpM8zHEuYBwIXtWPUvNT3ygWSSttd7VU2bdLz9kwAAKOQ8br/EntVq1RtvvKHixYurcuXKqly5skqUKKE333xTVqvVETMC+a7ufQ/qUKc5umZ4q8G1WO2d2kupKbl/uwcAgOMdP/ibwtaOsZXbNe0TffsnAQBQBOQ5yL/yyiuaMWOG/v73v2vbtm3atm2b3n77bU2fPl3jx493xIyAQ9S970Ed6vgJYR4AXNDVxItKzyy386qthk/OdvZIAAC4jDzfWv/ZZ5/pk08+0YMPPmjbVr9+fVWoUEEjRozQW2+9la8DAo5Ut1VP7ZIU/uMTanAtVtunPqRaY77hNnsAcKKMcrvH1Mh6PKPcbtgCefv4OnssAABcRp6vyCckJKhmzZrZttesWVMJCQn5MhRQkOq26qlDHT9RiuGliGubtGfqQ1yZBwAniv38VTVKWpdRbtf9E5Wm3A4AADt5DvINGjTQjBkzsm2fMWOGGjRokC9DAQWtbqueOthxDmEeAJzMVm6njHK7mpTbAQCQTZ5vrX/33XfVvXt3/fjjj2revLkkaePGjTp+/LiWLl2a7wMCBSXzNvuqP/5FEdc2afu03qo1egm32QNAAbmx3C6KcjsAAHKU5yvybdq00b59+/TQQw/p0qVLunTpknr37q19+/apVatWjpgRKDB2V+aTN2rPtN5cmQeAAkC5HQAAuXdH7yNfoUIFSu1QaNldmU/eyJV5AHCwrOV2Z1WScjsAAG4jz1fk582bp0WLFmXbvmjRIn322Wf5MhTgbHVb9dSBDvZX5tNSU5w9FgAUSps+H2crt0t4gHI7AABuJ89BftKkSSpdunS27UFBQXr77bfzZSjAFdRrbR/md0/tRZgHgHy2Y9W/FHVklqQ/yu0ad3DyRAAAuL48B/m4uDiFhYVl2165cmXFxcXly1CAqyDMA4Dj3Fhu15RyOwAAciXPQT4oKEg7d+7Mtn3Hjh0qVapUvgwFuJKMMP8xYR4A8hHldgAA3Lk8B/lHHnlEo0eP1urVq5Wenq709HStWrVKY8aM0YABAxwxI+B09Vo/ZBfmf5/6EGEeAO5QZrldKOV2AADckTwH+TfffFNRUVHq0KGD/Pz85Ofnp86dO6t9+/a8Rh6FWtYw3zB5A2EeAO7Qn+V2npTbAQBwB/Ic5L29vbVw4ULt27dP8+fP15IlS3To0CHNnTtX3t7ejpgRcBn1Wj+kA+1nE+YB4A7tWLXgz3K7+pTbAQBwJ/Ic5DNVq1ZNffv21f3336+LFy/q4sWL+TkX4LLqtelNmAeAO5BRbvf8H+V2vdT04eedPRIAAG4pz0H+ueee05w5cyRJ6enpatOmjRo1aqSQkBCtWbMmv+cDXBJhHgDyJnu53UfOHgkAALeV5yC/ePFiNWjQQJL03//+V4cPH9bevXv1/PPP65VXXsn3AQFXlRnmUwnzAHBLlNsBAJC/8hzkz58/r+DgYEnS0qVL1a9fP1WvXl2PP/64fvvtt3wfEHBl9dr01n67MN+bMA8AN6DcDgCA/JXnIF+2bFnt3r1b6enpWrZsmTp16iRJSk5OltlszvcBAVdnH+Z/JswDQBaU2wEAkP/yHOSHDRumfv36qW7dujKZTOrYsaMkKTY2VjVr1sz3AQF3QJgHgOwotwMAwDE88/qE119/XXXr1tXx48fVt29f+fj4SJLMZrPGjh2b7wMC7qJem97aaRiqsfopNUz+WdumPaw6o7/mdaAAiiTK7QAAcJw8B3lJ6tOnjyTpxIkTslqt8vDw0JAhQ/J1MMAd1W/7sHZKGWE+aT1hHkCRZE1P14GPBqkh5XYAADjEHb+PvCTVrl1bR48ezadRgMKhftuHta/dRxm32Set1+/THuY2ewBFSuwXr6hh0nrK7QAAcJC7CvKGYeTXHEChUr/tw9rXdpZdmLekpTp7LABwOMrtAABwvLsK8gBurn67PnZhftfU3oR5AIUa5XYAABSMuwry48aNU8mSJfNrFqDQIcwDKCqyltvtodwOAACHuqsgHxMToxIlSuTTKEDhRJgHUNhlltuF/lFuV+YvCym3AwDAgfLt1vrjx4/r8ccfz6/dAYXKjWH+t6m8Zh5A4RH7+Tj7crvgSs4eCQCAQi3fgnxCQoI+++yz/NodUOhkDfONktYR5gEUCjtWLVDU0Yzb6Cm3AwCgYOT6feS//fbbWz5++PDhux4GKOzqt+ujnZJqrnlKjZLWaevUh1VvzNfy8vZx9mgAkGc3lttFUW4HAECByHWQ79Wrl0wm0y3fcs5kMuXLUEBhVr9dH+2QoVprnv4jzPdRvTGLCfMA3ArldgAAOE+ub60vV66clixZIqvVmuPH1q1bHTknUKg0aNdXe9rOUprhqUZJa/Xb1D7cZg/AbVBuBwCAc+U6yEdGRmrLli03ffx2V+sB2CPMA3BXWcvtLj4wh3I7AAAKWK6D/IsvvqgWLVrc9PGqVatq9erV+TIUUFQQ5gG4G/tyu/Gq0bi9kycCAKDoyXWQb9Wqlbp27XrTx4sVK6Y2bdrky1BAUdKgXV/tafPhn2F+GmEegGs6fmCHqvz0nK3crunDzzl7JAAAiqRcB/nDhw9z6zzgIA3a9/szzF8lzANwPVcTLyr9/x5VgOka5XYAADhZroN8tWrVdO7cOdvn/fv3V3x8vEOGAooiwjwAV0W5HQAAriXXQf7Gq/FLly5VUlJSvg8EFGUZYX5mljDflzAPwOkotwMAwLXkOsgDKBgN2g/IEuZ/IswDcCrK7QAAcD25DvImk0kmkynbNgD5jzAPwBVQbgcAgGvyzO1CwzA0dOhQ+fj4SJJSUlL09NNPq1ixYnbrlixZkr8TAkVUg/YDtN2Qaq8dqUZXf9LWaX1Vf8xieXp5O3s0AEXAlcsJslJuBwCAS8p1kB8yZIjd54MGDcr3YQDYi+gwQNuVJcxP7UOYB+Bw1vR0HfxokBpSbgcAgEvKdZCfN2+eI+cAcBOEeQAFLfbzcWqe/HNGuV2POapBuR0AAC6FsjvADUR0GKDdrWcqzTCr0dWftHNqH123pDl7LACFEOV2AAC4PpcI8jNnzlRoaKh8fX0VFRWlzZs333Rt27ZtbcV7WT+6d+8uSbJYLHr55ZdVr149FStWTOXLl9fgwYN16tSpgjocwCEywvwHtjC/Y2pfwjyAfEW5HQAA7sHpQX7hwoWKjo7WhAkTtHXrVjVo0EBdunTR2bNnc1y/ZMkSnT592vaxa9cumc1m9e3bV5KUnJysrVu3avz48dq6dauWLFmiffv26cEHHyzIwwIcImuYj7y6hjAPIN/Yl9vVodwOAAAX5vQgP3nyZA0fPlzDhg1T7dq1NWvWLPn7+2vu3Lk5ri9ZsqSCg4NtHytWrJC/v78tyBcvXlwrVqxQv379VKNGDTVr1kwzZszQli1bFBcXV5CHBjgEYR5Afssst6tsK7dbQLkdAAAuLNdld46QlpamLVu2KCYmxrbNw8NDHTt21MaNG3O1jzlz5mjAgAHZ3gYvq8uXL8tkMqlEiRI5Pp6amqrU1D/fozsxMVFSxm36FoslV3M4Q+ZsrjwjHKNO64e18/p11dswWpFX1+jXqf1UZ8R8CvCQ7zjPFA2/fD5OLf4otzvfbbaqlSrHzxwFhvMMAEdzl/NMXuZzapA/f/680tPTVbZsWbvtZcuW1d69e2/7/M2bN2vXrl2aM2fOTdekpKTo5Zdf1iOPPKLAwMAc10yaNEkTJ07Mtn358uXy9/e/7RzOtmLFCmePAKcI0KGg0ep1dpoaX12tde/31vm6T8nDw+zswVAIcZ4pvFJOblff+I8lk/RtiSHyOpOsA0uXOnssFEGcZwA4mqufZ5KTk3O91qlB/m7NmTNH9erVU9OmTXN83GKxqF+/fjIMQx9++OFN9xMTE6Po6Gjb54mJiQoJCVHnzp1vGv5dgcVi0YoVK9SpUyd5eXk5exw4RTftXFVV9TaMVqv0Tfr1kB9X5pGvOM8UbicO/qbSW5+Sh8nQplK91PPpd5w9EoogzjMAHM1dzjOZd4bnhlODfOnSpWU2mxUfH2+3PT4+XsHBwbd8blJSkhYsWKA33ngjx8czQ/yxY8e0atWqWwZyHx8f+fj4ZNvu5eXl0j/oTO4yJxwjsssgbffwUO31o9T46mpt+XCQGoxeSJhHvuI8U/hcuZwgj0WDbeV2jZ78iJ8xnIrzDABHc/XzTF5mc2rZnbe3tyIjI7Vy5UrbNqvVqpUrV6p58+a3fO6iRYuUmpqqQYMGZXssM8QfOHBAP/74o0qVKpXvswOuJKLTQO2+b0ZGAd6VVdoxrT8FeABuinI7AADcm9Nb66Ojo/Xxxx/rs88+0549e/TMM88oKSlJw4YNkyQNHjzYrgwv05w5c9SrV69sId1isahPnz769ddfNX/+fKWnp+vMmTM6c+aM0tIINii8CPMAciv283Fq+Ee53cUH5qh0cCVnjwQAAPLA6a+R79+/v86dO6fXXntNZ86cUUREhJYtW2YrwIuLi5OHh/2/N+zbt0/r16/X8uXLs+3v5MmT+vbbbyVJERERdo+tXr1abdu2dchxAK4gotNAbZdUe/0oRV5ZpS3T+nObPQA721cuUNTRjySTtKPBa2rSuL2zRwIAAHnk9CAvSaNGjdKoUaNyfGzNmjXZttWoUUOGYeS4PjQ09KaPAUVBRKeB2mZYVefn0Yq8skq/ThugiNELCPMAdPzADoWvfU4eJkOxpXsrqvcYZ48EAADugNNvrQeQ/xp2HqTfW05TmmFW4ysrtX3aAG6zB4q4K5cTZP2/R23ldg2H3/zdXAAAgGsjyAOFFGEeQCbK7QAAKFwI8kAhlhnmLZlhfvojhHmgCKLcDgCAwoUgDxRyDTsP0q7MMJ/4I2EeKGJs5XbKKLerQbkdAABujyAPFAGEeaBoitu/3a7crgnldgAAFAoEeaCIyCnMp1+/7uyxADjIlcsJMhZQbgcAQGFEkAeKkBvD/LZp/QnzQCH0Z7ndCcrtAAAohAjyQBFDmAcKP8rtAAAo3AjyQBFEmAcKL8rtAAAo/AjyQBHVsPMg/dZiKmEeKEQotwMAoGggyANFWKMujxHmgUKCcjsAAIoOgjxQxBHmAfeXUW73aJZyu4WU2wEAUIgR5AGoUZfHtKvFP7OE+QGEecCNxH4Wo4bJG5RmeOrSg/NUOjjE2SMBAAAHIsgDkCQ17DIkS5hfQZgH3MT2lQvUPO7Pcrvqjdo6dyAAAOBwBHkANoR5wL1klttJotwOAIAihCAPwM6NYX7r9EcI84ALotwOAICiiyAPIJuGXYbot+aTZTHManJ5OWEecDGU2wEAULQR5AHkqFHXoYR5wEVlltulGl6U2wEAUAQR5AHcVGaYv254EOYBF7H9x//7s9wugnI7AACKIoI8gFtq1HWodjb/Z5YwP5AwDzhJ3P7tCl/3vKSMcrumD4128kQAAMAZCPIAbss+zP9AmAecgHI7AACQiSAPIFcI84DzUG4HAACyIsgDyLVGXYdqZ7PJWcL8o4R5oABQbgcAALIiyAPIk0b3D8sS5pcR5gEHo9wOAADciCAPIM8I80DBoNwOAADkhCAP4I5kC/MzBhHmgXyUtdxut1ddNXpylrNHAgAALoIgD+CONbp/mHZE/RHmL31PmAfyyY3ldkF/WSAvbx9njwUAAFwEQR7AXYnsRpgH8hvldgAA4FYI8gDuGmEeyD+U2wEAgNshyAPIFxlh/h+2ML9lxmOypqc7eyzArVBuBwAAcoMgDyDfRHb7iy3MN720VL9OH0SYB3KJcjsAAJBbBHkA+YowD+Rd1nK7eJWi3A4AANwSQR5Avovs9hftaPoeYR7IpdjPxtrK7S5TbgcAAG6DIA/AISK7P0GYB3Jh+4qv1DxutqTMcrs2Tp4IAAC4OoI8AIchzAO3Frd/u8LXR0ui3A4AAOQeQR6AQxHmgZxRbgcAAO4UQR6Aw2WG+XTDRJgHRLkdAAC4OwR5AAUisvsT2t70H4R5QJTbAQCAu0OQB1BgsoX5GY8R5lHk2JfbTaDcDgAA5BlBHkCBiuz+hLZl3mZ/8TvCPIqUY/u2q6qt3O5hNX3oWSdPBAAA3BFBHkCBa9x9OGEeRc6VywnSwkd1j63c7kNnjwQAANwUQR6AUxDmUZRQbgcAAPITQR6A02QP84MJ8yiUKLcDAAD5iSAPwKkadx+ubU3e/SPM/48wj0KHcjsAAJDfCPIAnK7xA08S5lEoUW4HAAAcwSWC/MyZMxUaGipfX19FRUVp8+bNN13btm1bmUymbB/du3e3rTEMQ6+99prKlSsnPz8/dezYUQcOHCiIQwFwh7KF+ZlDCPNwa4mXLsi0cCDldgAAIN85PcgvXLhQ0dHRmjBhgrZu3aoGDRqoS5cuOnv2bI7rlyxZotOnT9s+du3aJbPZrL59+9rWvPvuu5o2bZpmzZql2NhYFStWTF26dFFKSkpBHRaAO9D4gSe1rfE7GWE+4b+Eebgta3q6Ds0epErWk5TbAQCAfOf0ID958mQNHz5cw4YNU+3atTVr1iz5+/tr7ty5Oa4vWbKkgoODbR8rVqyQv7+/LcgbhqEpU6bo1VdfVc+ePVW/fn19/vnnOnXqlP79738X4JEBuBONezxFmIfbo9wOAAA4kqczv3haWpq2bNmimJgY2zYPDw917NhRGzduzNU+5syZowEDBqhYsWKSpCNHjujMmTPq2LGjbU3x4sUVFRWljRs3asCAAdn2kZqaqtTUVNvniYmJkiSLxSKLxXJHx1YQMmdz5RmBO9Gg6+PaYrUqcmuMmib8V7EzhijiqU/kYTY7e7Qih/NM3u1cucBWbre9/ng1qteC7x9wC5xnADiau5xn8jKfU4P8+fPnlZ6errJly9ptL1u2rPbu3Xvb52/evFm7du3SnDlzbNvOnDlj28eN+8x87EaTJk3SxIkTs21fvny5/P39bzuHs61YscLZIwAOUF5LSj2l3hc+UtTF/2rl+wOUWHuIPDycfiNRkcR5JndSLp1S98OvSyZpuXcnXfMM1dKlS509FuAWOM8AcDRXP88kJyfneq1Tg/zdmjNnjurVq6emTZve1X5iYmIUHR1t+zwxMVEhISHq3LmzAgMD73ZMh7FYLFqxYoU6deokLy8vZ48DOEA3/fpdZTXeNk4dLKsVG3cPV+YLGOeZ3Lt6+YKuzGyve0wp2u1VV/eN+ZTXxQO5wHkGgKO5y3km887w3HBqkC9durTMZrPi4+PttsfHxys4OPiWz01KStKCBQv0xhtv2G3PfF58fLzKlStnt8+IiIgc9+Xj4yMfn+x/2fLy8nLpH3Qmd5kTuBNRvUboVw8PNdwyVlEX/6vY2cPVZOSnhPkCxnnm1qzp6To2d6gijIxyu7JPLJR/sXucPRbgVjjPAHA0Vz/P5GU2p96j6u3trcjISK1cudK2zWq1auXKlWrevPktn7to0SKlpqZq0KBBdtvDwsIUHBxst8/ExETFxsbedp8AXFPjB5/Wtsi/K90wKSrhW/3ywTAK8OBSYj99WRHJG23ldqXKVnT2SAAAoBBz+otNo6Oj9fHHH+uzzz7Tnj179MwzzygpKUnDhg2TJA0ePNiuDC/TnDlz1KtXL5UqVcpuu8lk0nPPPae//e1v+vbbb/Xbb79p8ODBKl++vHr16lUQhwTAATLC/KSMMH/hP4R5uIztK75S8+MfS5J2RExQ9UZtnDwRAAAo7Jz+Gvn+/fvr3Llzeu2113TmzBlFRERo2bJltrK6uLi4bOVW+/bt0/r167V8+fIc9/nSSy8pKSlJTz75pC5duqT77rtPy5Ytk6+vr8OPB4DjNH7wGf0qqeGWGEVd+I9iPximJiPmcZs9nObYvu2quj5aMkmxpR9W1EPPOnskAABQBDg9yEvSqFGjNGrUqBwfW7NmTbZtNWrUkGEYN92fyWTSG2+8ke318wDcX/Yw/7iajJhLmEeBS7x0QaaFA3WP6Zp2e9dToyc/dPZIAACgiHD6rfUAkFeNH3xG2yInyWqYFHXh3/rlg8e5zR4FypqersOzH1Ul6x/ldn9ZQEM9AAAoMAR5AG6p8YPPaCthHk6StdwuseenlNsBAIACRZAH4LYI83CGbcu/tJXb7Wz4uqo1bO3kiQAAQFFDkAfg1ho/+Iy2NHqbMI8CcWzfdlX7+a+SpNgyfdSkV879LgAAAI5EkAfg9pr0HEGYh8NlK7cb/oGzRwIAAEUUQR5AoZBTmDesVmePhUKCcjsAAOBKCPIACo0bw/zmmcMI88gXlNsBAABXQpAHUKg06TlCWxq+RZhHvqHcDgAAuBqCPIBCp0mvkYR55AvK7QAAgCsiyAMolLKF+Q/+QphHniReuiCPBY9QbgcAAFwOQR5AoZUR5v+WEebPLyHMI9cyy+1CjFOU2wEAAJdDkAdQqDXpNYowjzyj3A4AALgygjyAQo8wj7yg3A4AALg6gjyAIoEwj9yg3A4AALgDgjyAIqNJr1H6NeJNwjxylLXc7nfK7QAAgAsjyAMoUpo+9CxhHtlY09N1+KOBCjFO6YxKK5hyOwAA4MII8gCKnOxh/gnCfBEX++lLiri2SamGl670nEe5HQAAcGkEeQBFUkaYf+OPMP81Yb4Iyyi3+0QS5XYAAMA9EOQBFFlNHxpNmC/iKLcDAADuiCAPoEjLFuY/HE6YLyIotwMAAO6KIA+gyGv60GhtaTAxI8yfW0yYLwIotwMAAO6MIA8Akpr0HkOYL0IotwMAAO6MIA8Af7gxzMd++CRhvhCyL7ebSLkdAABwOwR5AMgia5hvdm4RYb6QObZ3q63cblOZvmrSa6STJwIAAMg7gjwA3IAwXzglXrogj4WP2srtIofPdPZIAAAAd4QgDwA5IMwXLpTbAQCAwoQgDwA3kRnmJWWE+VlPEebdFOV2AACgMCHIA8AtNOk9Rpvr/RHmz/6LMO+GKLcDAACFDUEeAG6j6cPPEebd1LG9W1X95xckUW4HAAAKD4I8AORC9jD/NGHexWWW2xUzpVBuBwAAChWCPADkUkaYf12S1OzsQsK8C6PcDgAAFGYEeQDIg6YPP0+YdwOU2wEAgMKMIA8AeZQtzH/0DGHehVBuBwAACjuCPADcAbswH7+AMO8iKLcDAABFAUEeAO5Q04ef1+a6EyQR5l0B5XYAAKCoIMgDwF1o2ieaMO8CKLcDAABFCUEeAO5S9jA/gjBfwDbPe/HPcrten1JuBwAACjWCPADkg6Z9ohVb5zVJUrP4/yPMF6Bty79UsxNzJP1RbhfRyskTAQAAOBZBHgDySVTfFwjzBYxyOwAAUBQR5AEgHxHmCw7ldgAAoKgiyANAPssI8+MlEeYdhXI7AABQlBHkAcABovr+lTDvQJnldimU2wEAgCKIIA8ADpItzM8eRZjPB1t/+MJWbvdbozcotwMAAEUOQR4AHMguzJ+ZT5i/S8f2blWNDX+VJG0K6qcmPUc4eSIAAICCR5AHAAeL6vtXxdZ+VRJh/m5kK7d7YoazRwIAAHAKpwf5mTNnKjQ0VL6+voqKitLmzZtvuf7SpUsaOXKkypUrJx8fH1WvXl1Lly61PZ6enq7x48crLCxMfn5+Cg8P15tvvinDMBx9KABwU1H9XiTM3wXK7QAAAP7k1CC/cOFCRUdHa8KECdq6dasaNGigLl266OzZszmuT0tLU6dOnXT06FEtXrxY+/bt08cff6wKFSrY1rzzzjv68MMPNWPGDO3Zs0fvvPOO3n33XU2fPr2gDgsAckSYv3OU2wEAAPzJ05lffPLkyRo+fLiGDRsmSZo1a5a+++47zZ07V2PHjs22fu7cuUpISNCGDRvk5eUlSQoNDbVbs2HDBvXs2VPdu3e3Pf5///d/t73SDwAFIarfi4r9l6Go3W+p2Zn52jRbinpyhkweTr9BymXdWG7XhHI7AABQxDktyKelpWnLli2KiYmxbfPw8FDHjh21cePGHJ/z7bffqnnz5ho5cqT+85//qEyZMho4cKBefvllmc1mSVKLFi00e/Zs7d+/X9WrV9eOHTu0fv16TZ48+aazpKamKjU11fZ5YmKiJMlischiseTH4TpE5myuPCOA7Bo99Lw2Wq1qvneSmp2Zrw0fSY0f/6dLhnlnn2eO7duWUW5nkjaW6avG3YZzzgMKGWefZwAUfu5ynsnLfE4L8ufPn1d6errKli1rt71s2bLau3dvjs85fPiwVq1apUcffVRLly7VwYMHNWLECFksFk2YMEGSNHbsWCUmJqpmzZoym81KT0/XW2+9pUcfffSms0yaNEkTJ07Mtn358uXy9/e/i6MsGCtWrHD2CADyyq+WlgQOUe/Ez9Qifr6+/+dlpdboK5OHydmT5cgZ5xlLapKa7J6oYqYUbTfV0ply99t1ogAoXPj7DABHc/XzTHJycq7XOvXW+ryyWq0KCgrS7NmzZTabFRkZqZMnT+q9996zBfl//etfmj9/vr766ivVqVNH27dv13PPPafy5ctryJAhOe43JiZG0dHRts8TExMVEhKizp07KzAwsECO7U5YLBatWLFCnTp1sr3UAIA76aaNX5dX872TdH/K/7QhvrgaD3OtK/POOs9Y09O1Z8qDqqwzOqPSKv/EAtUJqnD7JwJwO/x9BoCjuct5JvPO8NxwWpAvXbq0zGaz4uPj7bbHx8crODg4x+eUK1dOXl5ettvoJalWrVo6c+aM0tLS5O3trRdffFFjx47VgAEDJEn16tXTsWPHNGnSpJsGeR8fH/n4ZG8/9vLycukfdCZ3mRNAds0HjFXsQpOi9rytFmfma+OnJjUbPt2lwrxU8OeZTZ+9rGYpsRnldg99qmoVQgvsawNwDv4+A8DRXP08k5fZnPY3RW9vb0VGRmrlypW2bVarVStXrlTz5s1zfE7Lli118OBBWbO0PO/fv1/lypWTt7e3pIzbETxu+Auw2Wy2ew4AuJKo/i8rttY4SVLz019q08djinSb/Y3ldtUotwMAALDj1Es+0dHR+vjjj/XZZ59pz549euaZZ5SUlGRrsR88eLBdGd4zzzyjhIQEjRkzRvv379d3332nt99+WyNHjrSt6dGjh9566y199913Onr0qL755htNnjxZDz30UIEfHwDkln2Y/7zIhvlje7ZklNtJ2hTUT016jnDyRAAAAK7Hqa+R79+/v86dO6fXXntNZ86cUUREhJYtW2YrwIuLi7O7uh4SEqIffvhBzz//vOrXr68KFSpozJgxevnll21rpk+frvHjx2vEiBE6e/asypcvr6eeekqvvfZagR8fAORFVP+XFbvQUNSeSWp++nNt/FhqNnyqy91m7yiXL56Xx78GqZgpRb9711fkEzOcPRIAAIBLcnrZ3ahRozRq1KgcH1uzZk22bc2bN9emTZtuur+AgABNmTJFU6ZMyacJAaDgRPUfq9iF+jPMfyI1e6Lwh3lrerqOzh6oBsYpnVEZlXtigby8s3eXAAAAwMm31gMAsovqP1axtTJeVtT81Ofa9Enhv80+dt5f1eBaRrnd1Yc+VUka6gEAAG6KIA8ALqgohfmtP3yh5ifmSsoot6va4D4nTwQAAODaCPIA4KKKQpin3A4AACDvCPIA4MKyh/nnCk2Yp9wOAADgzhDkAcDF2Yf5zwpFmM8stwuh3A4AACDPCPIA4Aai+o9VbM2xkgpHmKfcDgAA4M4R5AHATUQNiLEP83Oed8swv+2Hz2zldrsi36TcDgAAII8I8gDgRuzC/MlP3S7MH9uzRdU3vCRJ2hTUX40ffMbJEwEAALgfgjwAuJmoATHaVONlSe4V5rOX20139kgAAABuiSAPAG6o2SPjbgjz0S4d5im3AwAAyD8EeQBwU/Zhfp5Lh3nK7QAAAPIPQR4A3FhGmM94zbmrhnnK7QAAAPIXQR4A3FyzR15x2TBPuR0AAED+I8gDQCGQLczPfcHpYf7GcrvGw2c4dR4AAIDCgiAPAIWEXZg/MdepYT6j3O4Ru3I7Ty9vp8wCAABQ2BDkAaAQafbIK9pU/UVJzg3zGeV2mym3AwAAcACCPAAUMs0GvmoX5mPn/rVAwzzldgAAAI5FkAeAQihrmG92Yk6BhXnK7QAAAByPIA8AhVRGmP9rxv8XQJi3L7drQLkdAACAgxDkAaAQazZwvF2Y3zTvRYeE+fTr128ot/s/yu0AAAAchCAPAIVc1jDf/PgnDgnzmym3AwAAKDAEeQAoApoNHK9N1V6QlP9hftsPn6n5yXmSKLcDAAAoCAR5ACgimj36WrYwf7cotwMAACh4BHkAKEJuDPMb5/z1jvdFuR0AAIBzEOQBoIixD/Mf31GYp9wOAADAeQjyAFAEZYT5aEl3FuYptwMAAHAegjwAFFHNHp1wR2Hevtzub5TbAQAAFDCCPAAUYXkN80f3/KoaGzJK8jaVHaDGDz7t8BkBAABgjyAPAEVcs0cnaFPV5yX9Eebn5txmf/nieZn/9Zj8TakZ5XZPTC/IMQEAAPAHgjwAQM0Gvf5nmI+brY1zX1S61VDskQRtOW/ShgNndfQjyu0AAABcgaezBwAAuIZmg17Xpi+lZgf/qeZxszVz4kW9l/qQJLPKHBmnZz03y2J4yLtWJ8rtAAAAnIggDwCwaTbode1972fVTNqskaZFSjNbtdcI0bOe/5YkeZmsuuBRWiWdOyYAAECRxq31AACbdKuhYdfH6qf0epKk572+1nSvP18LP9nSR4MPtVW61XDWiAAAAEUeQR4AYLP5SIJOX07REEuMLcx7m9IlSf+0PKxp6b11+nKKNh9JcOaYAAAARRpBHgBgc/ZKiu3/h1hidN3I+GMizfDU1PSHc1wHAACAgkWQBwDYBAX42v7/WfMSeZqsSjU85W26rmfNS3JcBwAAgIJF2R0AwKZpWEmVK+6rvle/UrTXYr1v6aPp6b31rHmJXvBaLJOkRfcMVNMw6u4AAACchSAPALAxe5j0efgaVdu9WJP/CPGSND29t0ySor0Wq0d4eZk9Ojh1TgAAgKKMIA8AsFOtjL8O1B6tRYfaSpf/fC38onsGqkd4eVUr4++84QAAAECQBwDcoF2MqklabzW08eBZLV8Xq86totS8ahBX4gEAAFwAQR4AkCOzh0lRYSV1YY+hqLCSMnuYnD0SAAAARGs9AAAAAABuhSAPAAAAAIAbIcgDAAAAAOBGCPIAAAAAALgRpwf5mTNnKjQ0VL6+voqKitLmzZtvuf7SpUsaOXKkypUrJx8fH1WvXl1Lly61W3Py5EkNGjRIpUqVkp+fn+rVq6dff/3VkYcBAAAAAECBcGpr/cKFCxUdHa1Zs2YpKipKU6ZMUZcuXbRv3z4FBQVlW5+WlqZOnTopKChIixcvVoUKFXTs2DGVKFHCtubixYtq2bKl2rVrp++//15lypTRgQMHdO+99xbgkQEAAAAA4BhODfKTJ0/W8OHDNWzYMEnSrFmz9N1332nu3LkaO3ZstvVz585VQkKCNmzYIC8vL0lSaGio3Zp33nlHISEhmjdvnm1bWFjYLedITU1Vamqq7fPExERJksVikcViuaNjKwiZs7nyjADcG+cZAI7GeQaAo7nLeSYv85kMwzAcOMtNpaWlyd/fX4sXL1avXr1s24cMGaJLly7pP//5T7bndOvWTSVLlpS/v7/+85//qEyZMho4cKBefvllmc1mSVLt2rXVpUsXnThxQj/99JMqVKigESNGaPjw4Ted5fXXX9fEiROzbf/qq6/k7+9/9wcLAAAAAMAtJCcna+DAgbp8+bICAwNvudZpV+TPnz+v9PR0lS1b1m572bJltXfv3hyfc/jwYa1atUqPPvqoli5dqoMHD2rEiBGyWCyaMGGCbc2HH36o6OhojRs3Tr/88otGjx4tb29vDRkyJMf9xsTEKDo62vZ5YmKiQkJC1Llz59t+A53JYrFoxYoV6tSpk+0OBQDIT5xnADga5xkAjuYu55nMO8Nzw6m31ueV1WpVUFCQZs+eLbPZrMjISJ08eVLvvfeeLchbrVY1btxYb7/9tiSpYcOG2rVrl2bNmnXTIO/j4yMfH59s2728vFz6B53JXeYE4L44zwBwNM4zABzN1c8zeZnNaa31pUuXltlsVnx8vN32+Ph4BQcH5/iccuXKqXr16rbb6CWpVq1aOnPmjNLS0mxrateubfe8WrVqKS4uLp+PAAAAAACAgue0IO/t7a3IyEitXLnSts1qtWrlypVq3rx5js9p2bKlDh48KKvVatu2f/9+lStXTt7e3rY1+/bts3ve/v9v796jqijXP4B/N8p9c1MTUDkgiYSeHQLi0kglBXeaiJTpMQLUvKByVURdiRdMS+Mi2XJpWELnGHk8gXU0QFRMQgNciKVsEVEDFW8pyuWcRJjfH/2Yw5aLIGxg2/ez1v5j3nnnnWdG52G/e9555+JFWFpaquAoiIiIiIiIiLpWt75HftmyZYiPj0diYiIUCgUWL16M6upqcRZ7X19frF69Wqy/ePFi3Lt3D8HBwbh48SIOHTqEzZs3Y+nSpWKd0NBQ/PTTT9i8eTMuXbqEr776Cp999plSHSIiIiIiIiJ11a3PyM+aNQt37tzB2rVrcfPmTYwYMQJpaWniBHilpaXQ0Pjfbw0WFhZIT09HaGgoXn75ZQwcOBDBwcFYuXKlWMfZ2RkpKSlYvXo1IiMjMXjwYGzbtg3e3t5tjqthIv/2TDbQHWpra1FTU4OHDx/26Gc9iEh9Mc8QkaoxzxCRqqlLnmnof7blxXLd9vq5nuzatWuwsLDo7jCIiIiIiIjoT6asrAyDBg1qtQ478s2or6/HjRs3YGBgAIlE0t3htKjhNXllZWU9+jV5RKS+mGeISNWYZ4hI1dQlzwiCgMrKSgwYMEBpZHpz1Or1c11FQ0Pjqb+A9CSGhoY9+j8kEak/5hkiUjXmGSJSNXXIM0ZGRm2q162T3RERERERERFR+7AjT0RERERERKRG2JFXY9ra2li3bh20tbW7OxQiek4xzxCRqjHPEJGqPY95hpPdEREREREREakR3pEnIiIiIiIiUiPsyBMRERERERGpEXbkiYiIiIiIiNQIO/I9iJWVFbZt2yYuSyQSHDhwoNviIaLnkypyy/r16zFixIhObZOIiIj+vJ7sGz0rV1dXhISEdLidnqZ3dwdALSsvL4eJiUl3h0FEzxlV5JawsDAEBgZ2aptEpL7Wr1+PAwcOoKCgoLtDISI1lZeXB319fXFZIpEgJSUF06dPb1c7ycnJ0NTUFJetrKwQEhKi9p17duR7MDMzs+4OgYieQ6rILVKpFFKptNPbJSJSBUEQUFdXh969+VWYqKd64YUXOqWdPn36dEo7T3r06BG0tLRU0nZbcGh9F6qsrIS3tzf09fVhbm6O2NjYVod6PDn89ZdffsGECROgq6uLvn37YuHChaiqqhLXz5kzB9OnT8fmzZthamoKY2NjREZG4vHjx1ixYgX69OmDQYMGYc+ePUr7WblyJYYOHQo9PT1YW1sjIiICtbW1qjgFRNSIq6srAgMDERISAhMTE5iamiI+Ph7V1dWYO3cuDAwMMGTIEKSmporbnDt3DpMnT4ZUKoWpqSl8fHxw9+5dpTaDgoIQHh6OPn36wMzMDOvXr1fab+PccvXqVUgkEiQnJ+O1116Dnp4e7O3tcerUKaVt4uPjYWFhAT09PXh5eSEmJgbGxsbi+ieH1tfX1yMyMhKDBg2CtrY2RowYgbS0NHF9w37/+c9/YuzYsdDV1YWzszMuXryIvLw8jBw5ElKpFJMnT8adO3fE7fLy8uDu7o5+/frByMgI48ePR35+fgf+FYioOU/LJaWlpfD09IRUKoWhoSFmzpyJW7duAQASEhKwYcMGnD17FhKJBBKJBAkJCeJ13/gufUVFBSQSCY4fPw4AOH78OCQSCdLT0+Hg4ABdXV1MmDABt2/fRmpqKuzs7GBoaIh33nkHNTU1Yju///47goKC0L9/f+jo6ODVV19FXl6euL6h3dTUVDg5OUFbWxs//vijSs8hEbXO1dUVAQEBCAgIgJGREfr164eIiAg0vB298dB6KysrAICXlxckEom43ND/aSwkJASurq5K+2nob7m6uuLXX39FaGiomJ8A4LfffsPs2bMxcOBA6OnpQSaTISkpqdl4Q0JC0K9fP8jlcsybNw9Tp05VqldbW4v+/fvj888/7/hJagU78l1o2bJlyM7OxnfffYeMjAxkZWW1+QtodXU15HI5TExMkJeXh/379+PIkSMICAhQqnfs2DHcuHEDJ06cQExMDNatW4epU6fCxMQEOTk58Pf3x6JFi3Dt2jVxGwMDAyQkJKCwsBBxcXGIj49HbGxspx47ETUvMTER/fr1Q25uLgIDA7F48WK8/fbbeOWVV5Cfn49JkybBx8cHNTU1qKiowIQJE+Dg4IDTp08jLS0Nt27dwsyZM5u0qa+vj5ycHGzduhWRkZHIyMhoNY73338fYWFhKCgowNChQzF79mw8fvwYAJCdnQ1/f38EBwejoKAA7u7u2LRpU6vtxcXFITo6GlFRUfj5558hl8sxbdo0FBcXK9Vbt24d1qxZg/z8fPTu3RvvvPMOwsPDERcXh6ysLFy6dAlr164V61dWVsLPzw8//vgjfvrpJ9jY2GDKlCmorKxsz2knojZoKZfU19fD09MT9+7dww8//ICMjAxcvnwZs2bNAgDMmjULy5cvx/Dhw1FeXo7y8nJxXVutX78en376KU6ePImysjLMnDkT27Ztw1dffYVDhw7h8OHD2L59u1g/PDwc33zzDRITE5Gfn48hQ4ZALpfj3r17Su2uWrUKH330ERQKBV5++eWOnyQi6pDExET07t0bubm5iIuLQ0xMDHbv3t2kXsMPc3v27EF5ebnSD3XtkZycjEGDBiEyMlLMTwDw3//+F05OTjh06BDOnTuHhQsXwsfHB7m5uU3i1dLSQnZ2Nnbu3In58+cjLS1NbAcADh48iJqamnbnvXYTqEs8fPhQ0NTUFPbv3y+WVVRUCHp6ekJwcLAgCIJgaWkpxMbGiusBCCkpKYIgCMJnn30mmJiYCFVVVeL6Q4cOCRoaGsLNmzcFQRAEPz8/wdLSUqirqxPr2NraCmPHjhWXHz9+LOjr6wtJSUktxvrxxx8LTk5OHTlcImqD8ePHC6+++qq43HB9+vj4iGXl5eUCAOHUqVPCxo0bhUmTJim1UVZWJgAQioqKmm1TEATB2dlZWLlypbjcOLdcuXJFACDs3r1bXH/+/HkBgKBQKARBEIRZs2YJb7zxhlKb3t7egpGRkbi8bt06wd7eXlweMGCAsGnTpiZxLFmypMX9JiUlCQCEo0ePimUffvihYGtrK7Skrq5OMDAwEP7973+3WIeI2q+1XHL48GGhV69eQmlpqbiuIW/k5uYKgtA0JwjC/677M2fOiGX3798XAAiZmZmCIAhCZmamAEA4cuSIWOfDDz8UAAglJSVi2aJFiwS5XC4IgiBUVVUJmpqawt69e8X1jx49EgYMGCBs3bpVqd0DBw48+0khok41fvx4wc7OTqivrxfLVq5cKdjZ2QmC0HrfqIGfn5/g6empVBYcHCyMHz9eaT8N/a3m2m3JG2+8ISxfvlypHQcHhyb1hg0bJmzZskVc9vDwEObMmfPU9juKd+S7yOXLl1FbW4tRo0aJZUZGRrC1tW3T9gqFAvb29koTPri4uKC+vh5FRUVi2fDhw6Gh8b9/VlNTU8hkMnG5V69e6Nu3L27fvi2W7du3Dy4uLjAzM4NUKsWaNWtQWlr6TMdJRO3T+I5Qw/XZ+Jo1NTUFANy+fRtnz55FZmam+Dy6VCrFSy+9BAAoKSlptk0AMDc3V7rmnxaHubm5uE8AKCoqUspdAJosN/bw4UPcuHEDLi4uSuUuLi5QKBQt7rfhWJ88/sax37p1CwsWLICNjQ2MjIxgaGiIqqoq5iwiFWgplygUClhYWMDCwkJcN2zYMBgbGze5xjtj36ampuLjf43LGnJDSUkJamtrlXKOpqYmRo0a1SSekSNHdkp8RNQ5Ro8eLQ5vB4AxY8aguLgYdXV1XRpHXV0dNm7cCJlMhj59+kAqlSI9Pb3J9wsnJ6cm286fP198dPnWrVtITU3FvHnzVB4zZ/h4zjSekRH441nY5srq6+sBAKdOnYK3tzc2bNgAuVwOIyMjfP3114iOju6ymIn+zJ52zTb8cauvr0dVVRU8PDywZcuWJu00dL5barPhmm9LHI33qWrN7ffJssZx+Pn54bfffkNcXBwsLS2hra2NMWPG4NGjRyqPlejP5llySWsabjQI///8K4AW5+R5Mg90ViyNb4gQkfrT0NBQyilAy3mlNR9//DHi4uKwbds2yGQy6OvrIyQkpMn3i+ZyiK+vL1atWoVTp07h5MmTGDx4MMaOHdvuGNqLd+S7iLW1NTQ1NZWe53jw4AEuXrzYpu3t7Oxw9uxZVFdXi2XZ2dnQ0NBo81395pw8eRKWlpZ4//33MXLkSNjY2ODXX3995vaISHUcHR1x/vx5WFlZYciQIUofVX45tbW1bfIsWmvPphkaGmLAgAHIzs5WKs/OzsawYcM6FEt2djaCgoIwZcoUDB8+HNra2kqT/RGR6tnZ2aGsrAxlZWViWWFhISoqKsRrXEtLq8kdtYYZqBs/S9oZr6d78cUXxWdWG9TW1iIvL6/DOYeIVCsnJ0dpuWH+m169ejWpq6mp2WxeaZxTgKfnlebyU3Z2Njw9PfHuu+/C3t4e1tbWbe6n9e3bF9OnT8eePXuQkJCAuXPntmm7jmJHvosYGBjAz88PK1asQGZmJs6fP4/33nsPGhoaSsNJWuLt7Q0dHR34+fnh3LlzyMzMRGBgIHx8fMThqM/CxsYGpaWl+Prrr1FSUoJPPvkEKSkpz9weEanO0qVLce/ePcyePRt5eXkoKSlBeno65s6dq9IhaIGBgfj+++8RExOD4uJi7Nq1C6mpqa3mrhUrVmDLli3Yt28fioqKsGrVKhQUFCA4OLhDsdjY2ODvf/87FAoFcnJy4O3tDV1d3Q61SUTt4+bmBplMBm9vb+Tn5yM3Nxe+vr4YP368OHTdysoKV65cQUFBAe7evYvff/8durq6GD16tDjZ3A8//IA1a9Z0OB59fX0sXrwYK1asQFpaGgoLC7FgwQLU1NTgvffe63D7RKQ6paWlWLZsGYqKipCUlITt27e3+F3BysoKR48exc2bN3H//n0AwIQJE3D69Gl8+eWXKC4uxrp163Du3LlW92llZYUTJ07g+vXr4s0AGxsbZGRk4OTJk1AoFFi0aJH4Jo62mD9/PhITE6FQKODn59fm7TqCHfkuFBMTgzFjxmDq1Klwc3ODi4sL7OzsoKOj89Rt9fT0kJ6ejnv37sHZ2RkzZszAxIkT8emnn3YopmnTpiE0NBQBAQEYMWIETp48iYiIiA61SUSq0XCXu66uDpMmTYJMJkNISAiMjY2V5sbobC4uLti5cydiYmJgb2+PtLQ0hIaGtpq7goKCsGzZMixfvhwymQxpaWn47rvvYGNj06FYPv/8c9y/fx+Ojo7w8fERXzdFRF1HIpHg22+/hYmJCcaNGwc3NzdYW1tj3759Yp233noLr7/+Ol577TW88MIL4mucvvjiCzx+/BhOTk4ICQnBBx980CkxffTRR3jrrbfg4+MDR0dHXLp0Cenp6TAxMemU9olINXx9ffGf//wHo0aNwtKlSxEcHIyFCxc2Wzc6OhoZGRmwsLCAg4MDAEAulyMiIgLh4eFwdnZGZWUlfH19W91nZGQkrl69ihdffFEcKbRmzRo4OjpCLpfD1dUVZmZmTV5r1xo3NzeYm5tDLpdjwIABbd6uIyTCkw8VUJeprq7GwIEDER0dzV+MiUitLFiwABcuXEBWVlZ3h0JERERqyNXVFSNGjBDfFa/OqqqqMHDgQOzZswdvvvlml+yTk911oTNnzuDChQsYNWoUHjx4gMjISACAp6dnN0dGRNS6qKgouLu7Q19fH6mpqUhMTMSOHTu6OywiIiKiblNfX4+7d+8iOjoaxsbGmDZtWpftmx35LhYVFYWioiJoaWnByckJWVlZ6NevX3eHRUTUqtzcXGzduhWVlZWwtrbGJ598gvnz53d3WERERETdprS0FIMHD8agQYOQkJCA3r27rnvNofVEREREREREaoST3RERERERERGpEXbkiYiIiIiIiNQIO/JEREREREREaoQdeSIiIiIiIiI1wo48ERERERERkRphR56IiIh6lISEBBgbG3d3GERERD0WO/JERERqQiKRtPpZv359t8Z24MCBNtXNzMzElClT0LdvX+jp6WHYsGFYvnw5rl+/rtogiYiInhPsyBMREamJ8vJy8bNt2zYYGhoqlYWFhbWrvUePHqko0pbt2rULbm5uMDMzwzfffIPCwkLs3LkTDx48QHR0dJfHQ0REpI7YkSciIlITZmZm4sfIyAgSiURcrq6uhre3N0xNTSGVSuHs7IwjR44obW9lZYWNGzfC19cXhoaGWLhwIQAgPj4eFhYW0NPTg5eXF2JiYpoMbf/222/h6OgIHR0dWFtbY8OGDXj8+LHYLgB4eXlBIpGIy0+6du0agoKCEBQUhC+++AKurq6wsrLCuHHjsHv3bqxdu7bZ7UpKSuDp6dnqse3YsQM2NjbQ0dGBqakpZsyYIa7717/+BZlMBl1dXfTt2xdubm6orq5u62knIiLqcdiRJyIieg5UVVVhypQpOHr0KM6cOYPXX38dHh4eKC0tVaoXFRUFe3t7nDlzBhEREcjOzoa/vz+Cg4NRUFAAd3d3bNq0SWmbrKws+Pr6Ijg4GIWFhdi1axcSEhLEenl5eQCAPXv2oLy8XFx+0v79+/Ho0SOEh4c3u76l5+KfdmynT59GUFAQIiMjUVRUhLS0NIwbNw7AH6MYZs+ejXnz5kGhUOD48eN48803IQhC204sERFRDyQR+JeMiIhI7SQkJCAkJAQVFRUt1vnrX/8Kf39/BAQEAPjjzrmDgwNSUlLEOn/7299QVVWFgwcPimXvvvsuDh48KLbt5uaGiRMnYvXq1WKdf/zjHwgPD8eNGzcA/PGMfEpKCqZPn95iPEuWLMHevXvx4MGDTj225ORkzJ07F9euXYOBgYFSvfz8fDg5OeHq1auwtLRsdb9ERETqgnfkiYiIngNVVVUICwuDnZ0djI2NIZVKoVAomtyRHzlypNJyUVERRo0apVT25PLZs2cRGRkJqVQqfhYsWIDy8nLU1NS0OUZBECCRSNp5ZE8/Nnd3d1haWsLa2ho+Pj7Yu3evGJe9vT0mTpwImUyGt99+G/Hx8bh//367YyAiIupJ2JEnIiJ6DoSFhSElJQWbN29GVlYWCgoKIJPJmkxop6+v3+62q6qqsGHDBhQUFIifX375BcXFxdDR0WlzO0OHDsWDBw9QXl7erv0/7dgMDAyQn5+PpKQkmJubY+3atbC3t0dFRQV69eqFjIwMpKamYtiwYdi+fTtsbW1x5cqVdsVARETUk7AjT0RE9BzIzs7GnDlz4OXlBZlMBjMzM1y9evWp29na2jZ5pv3JZUdHRxQVFWHIkCFNPhoaf3yV0NTURF1dXav7mjFjBrS0tLB169Zm17c0lL4tx9a7d2+4ublh69at+Pnnn3H16lUcO3YMwB/D/l1cXLBhwwacOXMGWlpaSo8XEBERqZve3R0AERERdZyNjQ2Sk5Ph4eEBiUSCiIgI1NfXP3W7wMBAjBs3DjExMfDw8MCxY8eQmpqqNAR+7dq1mDp1Kv7yl79gxowZ0NDQwNmzZ3Hu3Dl88MEHAP54/v7o0aNwcXGBtrY2TExMmuzLwsICsbGxCAgIwMOHD+Hr6wsrKytcu3YNX375JaRSabOvoHvasR08eBCXL1/GuHHjYGJigu+//x719fWwtbVFTk4Ojh49ikmTJqF///7IycnBnTt3YGdn9yynmYiIqEfgHXkiIqLnQExMDExMTPDKK6/Aw8MDcrkcjo6OT93OxcUFO3fuRExMDOzt7ZGWlobQ0FClIfNyuRwHDx7E4cOH4ezsjNGjRyM2NlZp8rjo6GhkZGTAwsICDg4OLe5vyZIlOHz4MK5fvw4vLy+89NJLmD9/PgwNDREWFvZMx2ZsbIzk5GRMmDABdnZ22LlzJ5KSkjB8+HAYGhrixIkTmDJlCoYOHYo1a9YgOjoakydPbstpJSIi6pE4az0REREpWbBgAS5cuICsrKzuDoWIiIiawaH1REREf3JRUVFwd3eHvr4+UlNTkZiYiB07dnR3WERERNQC3pEnIiL6k5s5cyaOHz+OyspKWFtbIzAwEP7+/t0dFhEREbWAHXkiIiIiIiIiNcLJ7oiIiIiIiIjUCDvyRERERERERGqEHXkiIiIiIiIiNcKOPBEREREREZEaYUeeiIiIiIiISI2wI09ERERERESkRtiRJyIiIiIiIlIj7MgTERERERERqZH/A3KuzwccjeceAAAAAElFTkSuQmCC"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 15
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-27T11:51:52.640569Z",
     "start_time": "2024-08-27T11:51:52.637125Z"
    }
   },
   "cell_type": "code",
   "source": "",
   "id": "ab672b0754f7ad20",
   "outputs": [],
   "execution_count": 15
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}