transfer_learning_inceptionv4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import torch\n",
    "import seaborn as sns\n",
    "import torch.nn as nn\n",
    "import torchvision\n",
    "from torch import optim, cuda\n",
    "from torchvision import transforms, datasets, models\n",
    "import torch.optim as optim\n",
    "from torch.optim import lr_scheduler\n",
    "import numpy as np\n",
    "from torchvision import datasets,models,transforms\n",
    "from torch.utils.data import DataLoader ,sampler\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import time\n",
    "import cv2\n",
    "import copy\n",
    "from IPython.core.interactiveshell import InteractiveShell\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore', category=FutureWarning)\n",
    "# Image manipulations\n",
    "from PIL import Image\n",
    "# Useful for examining network\n",
    "from torchsummary import summary\n",
    "# Timing utility\n",
    "from timeit import default_timer as timer\n",
    "# Visualizations\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.rcParams['font.size'] = 14\n",
    "# Printing out all outputs\n",
    "InteractiveShell.ast_node_interactivity = 'all'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nfunctions in datahandler \\ndef process_image(image_path):\\ndef imshow_tensor(image, ax=None, title=None):\\ndef train(model,criterion,optimizer,train_loader,valid_loader,save_file_name,max_epochs_stop=3,n_epochs=20,print_every=2):\\ndef display_prediction(image_path, model, topk):\\ndef predict(image_path, model, topk=5):\\ndef save_checkpoint(model, path,optimizer,history):\\n\\nmake sure train_on_gpu true if training in GPU\\nelse false;\\n'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''\n",
    "functions in datahandler \n",
    "def process_image(image_path):\n",
    "def imshow_tensor(image, ax=None, title=None):\n",
    "def train(model,criterion,optimizer,train_loader,valid_loader,save_file_name,max_epochs_stop=3,n_epochs=20,print_every=2):\n",
    "def display_prediction(image_path, model, topk):\n",
    "def predict(image_path, model, topk=5):\n",
    "def save_checkpoint(model, path,optimizer,history):\n",
    "\n",
    "make sure train_on_gpu true if training in GPU\n",
    "else false;\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nfunctions in getmethod files  \\n    def get_pretrained_model(model_name):\\n    def evaluate(model, test_loader, criterion, topk=[1]):\\n    def imshow(image):\\n\\nmake sure train_on_gpu true if training in GPU\\nelse false;\\n'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''\n",
    "functions in getmethod files  \n",
    "    def get_pretrained_model(model_name):\n",
    "    def evaluate(model, test_loader, criterion, topk=[1]):\n",
    "    def imshow(image):\n",
    "\n",
    "make sure train_on_gpu true if training in GPU\n",
    "else false;\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "train_on_gpu=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on gpu: True\n",
      "1 gpus detected.\n"
     ]
    }
   ],
   "source": [
    "src='C:\\\\Users\\\\Harshit\\\\Desktop\\\\MTP\\\\training_folder'\n",
    "traindir = os.path.join(src,'train')\n",
    "validdir = os.path.join(src,'val')\n",
    "testdir = os.path.join(src,'test')\n",
    "save_file_name = 'inceptionv4-transfer.pt'\n",
    "checkpoint_path = 'inceptionv4-transfer.pth'\n",
    "# Change to fit hardware\n",
    "batch_size = 32\n",
    "# Whether to train on a gpu\n",
    "train_on_gpu = cuda.is_available()\n",
    "print(f'Train on gpu: {train_on_gpu}')\n",
    "if train_on_gpu:\n",
    "    gpu_count = cuda.device_count()\n",
    "    print(f'{gpu_count} gpus detected.')\n",
    "    if gpu_count > 1:\n",
    "        multi_gpu = True\n",
    "    else:\n",
    "        multi_gpu = False\n",
    "cat_df=pd.read_csv('cat_df')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def imshow(image):\n",
    "    \"\"\"Display image\"\"\"\n",
    "    plt.figure(figsize=(6, 6))\n",
    "    plt.imshow(image)\n",
    "    plt.axis('off')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 gpus detected.\n",
      "1 gpus detected.\n"
     ]
    }
   ],
   "source": [
    "%run datahandler\n",
    "%run get_method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size=32\n",
    "image_transforms = {\n",
    "    # Train uses data augmentation\n",
    "    'train':\n",
    "    transforms.Compose([\n",
    "        transforms.RandomResizedCrop(size=299, scale=(0.8, 1.0)),\n",
    "        transforms.RandomRotation(degrees=15),\n",
    "        transforms.ColorJitter(),\n",
    "        transforms.RandomHorizontalFlip(),\n",
    "        transforms.CenterCrop(size=299),  # Image net standards\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406],\n",
    "                             [0.229, 0.224, 0.225])  # Imagenet standards\n",
    "    ]),\n",
    "    # Validation does not use augmentation\n",
    "    'valid':\n",
    "    transforms.Compose([\n",
    "        transforms.Resize(size=299),\n",
    "        transforms.CenterCrop(size=299),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ]),\n",
    "    'test':\n",
    "    transforms.Compose([\n",
    "        transforms.Resize(size=299),\n",
    "        transforms.CenterCrop(size=299),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ])\n",
    "}\n",
    "data = {\n",
    "    'train':\n",
    "    datasets.ImageFolder(root=os.path.join(src,'train'), transform=image_transforms['train']),\n",
    "    'val':\n",
    "    datasets.ImageFolder(root=os.path.join(src,'val'), transform=image_transforms['valid']),\n",
    "    'test':\n",
    "    datasets.ImageFolder(root=testdir, transform=image_transforms['test'])\n",
    "}\n",
    "dataloaders = {\n",
    "    'train': DataLoader(data['train'], batch_size=batch_size, shuffle=True),\n",
    "    'val': DataLoader(data['val'], batch_size=batch_size, shuffle=True),\n",
    "    'test': DataLoader(data['test'], batch_size=batch_size, shuffle=True)\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([32, 3, 299, 299]), torch.Size([32]))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Iterate through the dataloader once\n",
    "trainiter=iter(dataloaders['train'])\n",
    "features, labels = next(trainiter)\n",
    "features.shape,labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'model' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-10-42fc71a56071>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlast_linear\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0min_features\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m: name 'model' is not defined"
     ]
    }
   ],
   "source": [
    "model.last_linear.in_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from inceptionv4 import inceptionv4 \n",
    "model=inceptionv4(pretrained='imagenet')\n",
    "for param in model.parameters():\n",
    "    param.requires_grad=False\n",
    "num_ftrs = model.last_linear.in_features\n",
    "model.last_linear = nn.Sequential(\n",
    "                     nn.Linear(num_ftrs, 128),\n",
    "                     nn.ReLU(inplace=True),\n",
    "                     nn.Linear(128, 2))\n",
    "model = model.cuda()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0, 'Negative'), (1, 'Positive')]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.class_to_idx = data['train'].class_to_idx\n",
    "model.idx_to_class = {\n",
    "    idx: class_\n",
    "    for class_, idx in model.class_to_idx.items()\n",
    "}\n",
    "\n",
    "list(model.idx_to_class.items())[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "#optimizer = optim.SGD(model.parameters(), lr=0.0001, momentum=0.9)\n",
    "optimizer = optim.Adam(model.parameters())\n",
    "def accuracy(out, labels):\n",
    "    _,pred = torch.max(out, dim=1)\n",
    "    return torch.sum(pred==labels).item()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "41,339,810 total parameters.\n",
      "196,994 training parameters.\n"
     ]
    }
   ],
   "source": [
    "total_params = sum(p.numel() for p in model.parameters())\n",
    "print(f'{total_params:,} total parameters.')\n",
    "total_trainable_params = sum(\n",
    "    p.numel() for p in model.parameters() if p.requires_grad)\n",
    "print(f'{total_trainable_params:,} training parameters.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting Training from Scratch.\n",
      "\n",
      "Epoch: 1\t100.00% complete. 1414.41 seconds elapsed in epoch.\n",
      "Epoch: 1 \tTraining Loss: 0.1592 \tValidation Loss: 1.3274\n",
      "\t\tTraining Accuracy: 93.84%\t Validation Accuracy: 90.13%\n",
      "Epoch: 3\t100.00% complete. 1416.70 seconds elapsed in epoch.\n",
      "Epoch: 3 \tTraining Loss: 0.1423 \tValidation Loss: 1.1431\n",
      "\t\tTraining Accuracy: 94.57%\t Validation Accuracy: 90.98%\n",
      "Epoch: 5\t100.00% complete. 1414.72 seconds elapsed in epoch.\n",
      "Epoch: 5 \tTraining Loss: 0.1374 \tValidation Loss: 0.6289\n",
      "\t\tTraining Accuracy: 94.76%\t Validation Accuracy: 86.58%\n",
      "Epoch: 7\t100.00% complete. 1416.15 seconds elapsed in epoch.\n",
      "Epoch: 7 \tTraining Loss: 0.1289 \tValidation Loss: 2.8763\n",
      "\t\tTraining Accuracy: 95.10%\t Validation Accuracy: 89.82%\n",
      "Epoch: 9\t100.00% complete. 1417.18 seconds elapsed in epoch.\n",
      "Epoch: 9 \tTraining Loss: 0.1286 \tValidation Loss: 1.3508\n",
      "\t\tTraining Accuracy: 95.17%\t Validation Accuracy: 92.07%\n",
      "Epoch: 10\t100.00% complete. 1416.96 seconds elapsed in epoch.\n",
      "Early Stopping! Total epochs: 10. Best epoch: 5 with loss: 0.63 and acc: 92.00%\n",
      "18426.75 total seconds elapsed. 1675.16 seconds per epoch.\n"
     ]
    }
   ],
   "source": [
    "save_file_name = 'inceptionv4-transfer.pt'\n",
    "checkpoint_path = 'inceptionv4-transfer.pth'\n",
    "model, history = train(\n",
    "    model,\n",
    "    criterion,\n",
    "    optimizer,\n",
    "    dataloaders['train'],\n",
    "    dataloaders['val'],\n",
    "    save_file_name=save_file_name,\n",
    "    max_epochs_stop=5,\n",
    "    n_epochs=20,\n",
    "    print_every=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>train_acc</th>\n",
       "      <th>valid_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.212576</td>\n",
       "      <td>1.752030</td>\n",
       "      <td>0.913594</td>\n",
       "      <td>0.876167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.159173</td>\n",
       "      <td>1.327359</td>\n",
       "      <td>0.938406</td>\n",
       "      <td>0.901333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.152350</td>\n",
       "      <td>1.826306</td>\n",
       "      <td>0.941094</td>\n",
       "      <td>0.894167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.142310</td>\n",
       "      <td>1.143074</td>\n",
       "      <td>0.945719</td>\n",
       "      <td>0.909833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.130256</td>\n",
       "      <td>0.801219</td>\n",
       "      <td>0.950125</td>\n",
       "      <td>0.915000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.137386</td>\n",
       "      <td>0.628916</td>\n",
       "      <td>0.947625</td>\n",
       "      <td>0.865833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.131694</td>\n",
       "      <td>1.357666</td>\n",
       "      <td>0.949250</td>\n",
       "      <td>0.916167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.128943</td>\n",
       "      <td>2.876289</td>\n",
       "      <td>0.951000</td>\n",
       "      <td>0.898167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.123377</td>\n",
       "      <td>1.166922</td>\n",
       "      <td>0.952313</td>\n",
       "      <td>0.915000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.128560</td>\n",
       "      <td>1.350809</td>\n",
       "      <td>0.951656</td>\n",
       "      <td>0.920667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.119021</td>\n",
       "      <td>2.220055</td>\n",
       "      <td>0.955313</td>\n",
       "      <td>0.920000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    train_loss  valid_loss  train_acc  valid_acc\n",
       "0     0.212576    1.752030   0.913594   0.876167\n",
       "1     0.159173    1.327359   0.938406   0.901333\n",
       "2     0.152350    1.826306   0.941094   0.894167\n",
       "3     0.142310    1.143074   0.945719   0.909833\n",
       "4     0.130256    0.801219   0.950125   0.915000\n",
       "5     0.137386    0.628916   0.947625   0.865833\n",
       "6     0.131694    1.357666   0.949250   0.916167\n",
       "7     0.128943    2.876289   0.951000   0.898167\n",
       "8     0.123377    1.166922   0.952313   0.915000\n",
       "9     0.128560    1.350809   0.951656   0.920667\n",
       "10    0.119021    2.220055   0.955313   0.920000"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_checkpoint(model, checkpoint_path,optimizer,history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Adam (\n",
       "Parameter Group 0\n",
       "    amsgrad: False\n",
       "    betas: (0.9, 0.999)\n",
       "    eps: 1e-08\n",
       "    lr: 0.001\n",
       "    weight_decay: 0\n",
       ")"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "optimizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "41,339,810 total parameters.\n",
      "196,994 total gradient parameters.\n"
     ]
    }
   ],
   "source": [
    "model,optimizer,history=load_checkpoint(checkpoint_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# training analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 576x432 with 0 Axes>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1f8ddbcee80>]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1f8ddbbb340>]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1f90a04c730>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Epoch')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Average Negative Log Likelihood')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Training and Validation Losses')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 6))\n",
    "for c in ['train_loss', 'valid_loss']:\n",
    "    plt.plot(\n",
    "        history[c], label=c)\n",
    "plt.legend()\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Average Negative Log Likelihood')\n",
    "plt.title('Training and Validation Losses')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 576x432 with 0 Axes>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1f950c47400>]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1f950c47850>]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1f8df9ceeb0>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Epoch')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Average Accuracy')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Training and Validation Accuracy')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 6))\n",
    "for c in ['train_acc', 'valid_acc']:\n",
    "    plt.plot(\n",
    "        100 * history[c], label=c)\n",
    "plt.legend()\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Average Accuracy')\n",
    "plt.title('Training and Validation Accuracy')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_df=pd.read_csv('cat_df')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOz9W5YkOa4siAqoFpm79vB6AHcIPZEeQg+gR3jWqQw3Je8HKIQApJp7ZfVavT+KVZHubqbKB4iHAARJG2PgP+U/5T/lf15p/1934D/lP+U/5Vz+I5z/Kf8p/0PLf4TzP+U/5X9o+Y9w/qf8p/wPLf8Rzv+U/5T/oeX16cv/5//+v1Yo975v3PeN1hperxdac7lureG6LrTWMMZAHwNjDGCM9R2fH2Pgvm/03gEAZobWGswMY4xVTy1mBjNbv5/KYLuzT/wHAO/3O7Wr9fhPgzVDk8/GGOu9MQbMDNd1pT720TFGR414c0x8jn1b9On8O+jANldd5v0xM/TeYdbQrgbMfim9xhjo98DVLrx+vdZ8cbxO9477fqP3vt59vV64rmvR6H2/YebP813SsfbPrKG1mIve+6obMDSL9/RdnUu2o/XWtnW+SVP/vMGawQpP9N7X2J1ueR5O83/iKX3W+90wRtRdx+FtD/TucsI5qnxzauv/+P/9n0em/iictbP6TwdpZhgAhj8IFEG5rmsxGfS5+cMMiYhaWD8HdV1X6oMyhU7w/ne0QQYIQg2MDvQGNGGkJVC9r/dY36kdpRUZMn0+BzxG9PnEqIumIuRmDc0Mdxnrog/wIERZAa7+l3bf7zfufuO62nqvjulUTuM0ZIZXgdP3TnWf+Et/pzKVD4/9+knbVTDr75nP7iX4Yxqe1i60porHpYDK777vNL8636f2a/lWOMmUr9drWcDX67U6Ty0R1urC69USATggCsjVggEovNd1JU3/RNwTk5wmWt/rfcw5zJMS1sVwXQ1tNEC0G8dcico2XtcraWrSq9mFZq1MTMe4O95v16x8FjDqKTS73BpoW3Pofdx4U+DgVsvQgDGft7GsvY5Rhb5dDTBgzL+Vrq25BR0i7Cx8ilZx/oH7LnNhrH+sMTr9Q+nQYqtlWUhE+kOlqyiEvOhCYD6OSceKzr6+vo7KU5EP6dVaQ5vCVFEdf8eAz+lFerVkCc0M1hp67/j9+ze+fv8GSPvZz7t3dEGaNFpP5aNwnl7khKeOQ6wIstbeteCYxjUsl9ZV36lWSOvWelXTPcGOI3z0J7aJfNKoT3Sqwst+qMKpf2/Wb/U3DELta6VtteK1Xxt0PLVphgtAuwywhoHoa9RpC/Kexn8qpzG6QLhC7/1eipnPKBRWaFrHtFTGAemcUJ4K0olWbEuFM41RflOkt7lbY6AVZcL6rbXlOv2k/Fg4nwSVHey9bxBDGc6szckFzCigmYiErNpehSRVWHfGHskv0OdUu2YofIJYlnyqTzTK45wMeN8YEy7qeOgCGHZFw75UIThBvZM1VxhVaaiWcoNzANrU4qPZpkCWgFjbmPrUlxOvqGvjEPFGHwNN4F5VzsobFT2pzznWGAEUxaJ117hHonvhiTat4DJGE05TMIkAKt8B4cqpYI4xNv/4pOi0/NjnrESvkGQxH7LjztKayUQ2mN1pUlXDVY2rk/UThbFZ80ns+n6UkSZSodOnUjUzi2rhr68vwAwvgTBkqKcx8Ns6ufr8iYn5bO2Pf74HWmjFzIDr9cLVGqxA/6R8Duhlo6nNwNrI7ZGRGWQzM1zFUsJMgmU7YlBhWfMzhkPGu4M+H4OX2X2IosJJq/2TYtM1ovumsPRk5YGsFFQxc0x/G9bWKNepok04RTA58C7vrzpxLd9H4Vf7hlinoIcOdAwsP5ETNMYOW7Q8w7Jv4w0P70Xbi6GECSt0/2mdp34qQ6iwKK19zJleQQ/3x3vv6NcFXG0FrtJ8HaDgY18xgJEVrirdCjUXTZaFn89Y9n1PUDP61Nc8a7TW+wPYPeMPk8/u+05wutZfEZF+7uPZEVMVVEbjE88WWp34keVfFs6TFauaiRppCdp9r06liZrQQdtAmcDKFEvgBYJki+fwNUOzewk/l4OU8djnHd4SKp3LSetmKLZD76o99zKDHYUGj08LMtC2dJmlMpjC3Pu+gQk37/sGrliGqgx3QgmHDnlE+e5LudFSss3V/hi4OH+tzbn3eARhIAqcbK3BypIE6xpjoNNq9o4++a71MQU5o7I6N2qdP6HCKrQ6rvf7ja/3G++vrzVuLlktdwcUys+a/6Nwqp+gna9C4f9aEJWDI7w0j3ZVJtUliieLuBMCOAVw6vMnxnQG7Fs/VPn8Kz5BUiDerRUQG3C+OjH6R4kHDmML/7fC9tksGnbGSc8ilKtaFq/CI4rWGi6EwmE0udL3tG4YD8AF5H0fLQ/nhL44JtO2DxDP25ztTmt6grxUMBzfoqM5ctD+aLR0g+/YeWEOrfATFbhHfX///o2vry/8/v17RaPVEAzpr/98tprADwJC1bqchMIH8pnnyKg+qLxWRwY+Rcp2rX+vdyskqAJeFQlArTwXsbfJiLr+Zdipa7jIE1kFfgwuK52ZINU7xoJQ0c/dpx4tr2XyWUUknKRlJUs7GB0wZ4nTOizrrMKZ+GFgwdJqefn36zXbGMP9XKvwuUbYseZrPQPAZhvqsug/7Z/6mB4YPCcFKAKLmdRgjvahL3p+fX3hf//v/42vr3ckewC4pC4TWfoJf/0oWquEYwfDCASjnYrCLtXc7By1l/okVWj0Z9WK/M7MFty5b8f79Kk42cqsvnicIdMT4ys99PsKyYdYzvnCyvJRWunzT+XU9vr5oAn1nd6rUsrLOyosypBN/j6PPy8RqPVGQlKWrFOap/mdrmNmmuasq6zQpG/IvJmjtCjjR1Jy1xUKegDb2m8fA4yOLJoL39MdGBL4+/37N37//gIR5BgDfVpMygAOvP1Uvo3WslIVBGaRBFFuYMKhtLQiExKpdDfu+z0JdG3Wbj0rEHCMsQTvCDfYBhkLfSVHaMTuBDEtReBy33Wyt9S9OSG0BKRRSttiP6cwheIJhlQrlsZWBFvprYJZhSvqvfH15c9RGSqcVd8/BDWCQVF32yy3vzMSiqnz8ZR6F+PwnxSCmsUE+PIORgTXagIBI9AUMFU6rQH3XRHAlejl6Xgh2EmpjhEJGWYrAwuwxBcalX+/3/jrr3+uet7vN67rwj/ME2/e77l81CyN5al8FM5s3vcSzDBN/2K8SfiinQH42lQ/wLy98aVxxhiwuY7aR7UmYflWfYvQwRAqlNVqq1Wlb6ZK5YkO+txJQLQ/+nf0HSuI4H28PFOp5X+6LJARQ9D65Ae69Qw/TFFHtmaisEq0ytLvKqT7OmtSeIU2u7WIpIhevlvCjain0q/ylvrc2l99T1GDBns2F2n2KZR+wzBgSGCL79z3jfd94/31hT4Vye+vL4ypvP744w9Qn9Ki9v4z6/kjWPs04CBOWFgGLpQJNGRdnerFbIWxdTBjDNwUoD4AC6aa8ydEUyidtaUKpvZNIRzrB2LNVceax56Tv6OdBnqftEQePNiXERQx8B0yJ3M3NSCXhZNrxlk5xDNZMOvi+9LcNlIk8sQH30EwWI5Cq7I8CanzBMdsiXbXFUsMhLKLX+aHmbcMvUcCRp0nFejqjxJef/IBhw2gj2QYWJ9by7+8v7NfX79/ryQU799crtqg9vhoPf+lJIQ1cJfGxURsqBKmDlhh61jxibHW2jIDxHKEvyfLJXDnujX2RZ/PKVgKV0Pj+XO63BD5v5lZZORJEQH0W7IWdHrIEgwVBsSim61gxilqSFo5bLTFhNXH5bsKtasC0vHWHTbLeiSFgWUuEzKyiJLqv42x+Tx2oVQ60U1QGO3fec6ujyMU0Bp3Uk6qdHdX58SPih50bvWZXEf2h/sYabfQ15dHaSlofOd+v/H19YXX6xWuFQz3yK7Gp/IvCyfgOzdQiK7OfbVO6tdw0MrkG5ZKhRYHx2e9LxPCocK+YH72x4nScd/Zmi5GgzOrZjhpv6sS0s+0zY0BBrb3TkzLydVxVP9EP+d3qlz2PiG9e0JCq20LQVTm5DMq5Ay81fHbsJSLeirXdcFamxlBHVXYa+RelVYtWWGQ2NH3ijq6tOP0eK5/fUeaud+z6ny/3/j6egMYS8m+Xhf++gv4+vrCP//5z5WEb82AO/eXiudUvhXO76AO/Z5PYWkSO+F8PE3bz/pzgpmoTFIYX8nu0dyDthUIREF7Gr+WJy1IQ+TR4fB/UwZVWRJi1cowVfFo6RNy8Z9CS0UP6j+fYaoqwVW7M9DAWugPIdp3t5gZGtSF8KUrVWoLQZlhNKCPBgjkTCl+unNjzo/Sg0IWwa5b+CxooH1UAWWd1S3Q8ZjZCg61huXKeNDxKwXGxsAKEHIsq+05F7pL65P1/DeEk+tTuy9XrYsSZgVmMPcVfGPaaz8qRk9WGrtQfiqu+QP3R/RygEsxT9A9Cc5hzE/PK6TkBFbmRglyRbJAbkfRyF02hleL2FrbUiOzpQFsmAvc6ksD9ytWlFHHvZUxksmuDJ/G7B8uBaqZXCm6PoUSZusID41EZ5+aY9M4QF9947Y2dpN9qj6h9llpScFT4aNCaa3h9esXfs1/zCvWOafFBv6mcJ6YPCY9a6Y8cUXri3bVrJNmRAnnds6QL/fjuc8Hy1YESLVXjtx2fH3dy9pXKLjXscP2U5+c0VqKyGZfNxdlktiTuisHM1tLR/xeYTkZ/5oR0Mx8Iuxj4O63zMkAfd7TXCw0IgK+aOK/JMW9eATuOtDv1rqvskFZEdeip9S1LGdZv/V3lpqZdeqaaiCnyr8aCXa6n10Zbn2jgl+IoDX88esX/vGPf+C//vEP/OMf/8CvX78AAL2s6/5ty2ktbxGqFhLIidebpmltRip7ep+DVcLo79HOHnLWCKq2XZ9hqZpQlUddwNa/da31pE1ds18AGvxR7ZcHNfoYwHAYpBkpYwD37dYZ8O8UGp2schVMXaOk71PnYvlvsgxjZmjXlZINYEXxiZXmlHITODeRO3RtMHRfZhicc8O4B/rowLRkGngD4KmCEIsJ0g9r/qw19LJ0UYUnWe9pPd9T2YWf7gqIKG+Ms8tS1+VD+O/1/hLYMfD76wtf768Vle0joDpp/Xq98F9//ok/5r8++/fT8jnx3exgdItF8p6sDi3iimatgRX9vVqALKjPnuliHEPawHqycnv/KRTB4AoJXVvmXR1nbcdxcq8qM5K4LEP5tvWPzKFalrpEo7G1bWVGyPiWr7WUDlI/VRnx+2bFPZD6bdBFyOjC4AKqFtK3hpkL55yzZg29+eI+36f1qSiKdFArv6zsbKlaF61js8hURGOH4cFb4a/Wd1XxaV/4PdGVn7QQCS59DIz3e757wWwmV1wXXhPWpuVERR0fePXbxHfteOELwJCCDxTU/Hz2WWxCu/BVAfWllCinEt8ZbJ75o9qv1rNDzKwQHloB8JNtXTWAYsnC1SWsZVmnpiesdIHNCQLKPHuzOarZpyXCw+ZwW5Yt96P27QlmkWZ7GueucJf1uF6QEPWy7nyffjQT2jU7aykumVulRxWo3NesqE9jsvJ96vsH2MloPoNhlX59WlqPontQ6M8//5w+ZwOQM8F25yuXb3xOXbOboeDco40g2lndScGfzWz5XIQI42GR/TuGMlmP5Fqe+omq+Z76eOr/98L72V/QJaVT/qgh9rj6mh7Eek/r1pj7OfszJ5MtdlmCWJDssbeTWmZHmnx+DxvsPT8igZcGvESI7t5xv+/E0GP4Usx1OYS9hB/UovpYA4ZqhFsVhkbA3a/Xkxf2UxejD3vyQnV3UlaVoKwxwvUyM8lMA379euGPP/7Af/3Xfy0jV639d+WbaK0GOvYJqqxJHwL6U4gSMC6IjN5BFB5wJwcqql/AokKohARCeJ40PJ/Rz9Spj+SHZ0JuGlnaImN+Wiv1vmbrq8poywCaNPax1TN+cvG6rU4DxkCCdXUsp4yV6us9FcLGmipJS6PRzFiKcabGIFrZTyZUBacCOyZ0ZB/JU+261iFymR+o+GL+Ywsh1vgHkHxH56vgheAzbuUjzUJQr+vCH3/8sSK19/1ewSPS6jtk9m1ubYKkCWYAyapW+JEmLO9rVIbNdcYyBt+B2eNAlGGyj5UtsD9Dv1BhiisBzeyIsTjzaKDnJxZVn9OxRZ8BBsQc3XaY8QTA6yPNk2818mbuDX6NXbnNHngdpa8sNUjC9pyJAaYU1nlQOK5JH2p5avSbgR/WRmhPyBsIItOTdLp7h8EwRJG31lIGj5ZI8fMVA1d+VDyZ7kMUv0aBSRdtj79fF9YhAy85uOw9s4W+vr6OvvJT+dvCOacSDK4sxvUXUz0Vz1e/Sp8xYwDBheUq79S+1b8VtmWhysElh5FvNo4GBnT6sb5TCYGOPjwJZlgm+uINcdDZtJjWYEm5RF+zApvHLY4cBT+N/VRGHxgtr6WuMU3m3l2BWBPd9lYuAVaI2BY9uZ5sRdBUKcYWP83cyZBVy2lNWhm/wtETvB0jw3x1jaoiVMtb/WWd+2tabX7OgN3v31+z3kljVI7cy/frnJuV498MZETgZ/byXM/BSupP1ULKyHViFCrVonCK0OnUB/p3azc+AHp0KpxAQKoHCm3a9FSqsFAJLZLN9pkWGZbmRnUexvDkiT46+rJmLdGkKqaz9ezgNj+Yxc6ZpSkOoz0onMW8ElX2EicTLMYlEpJ3CSlTlFsMQj0tPQmwj+RRkWqkm0eXQKx2tfje53xCx6LJpNs63whRD2kB7AGl94zi+ly60mrmGUf/FqylhOuEUNuGxg8o+LGxDd7twpkF93yMvQ80/BN+FtaP+iiESiE1Zj/rov0JsvBnFSz9nQyjmrYycGjwVskQTHVJeL28V2F173FGjqKRH1vOzSrmvpyUmv9ejlvhO8naZQWoaXjByD4GmK2k+ypgZHzygAZV1EI2s7S9iwJS+w4gbaiu9NE1+UrvygPqEtQo8/wGAJJiGmNkPkS4NU/ls3ACuESDVCvmv16rI6tbUwiSxax18xlOCrAWndt1red1gDxVL05aIyNZghg+8D2fEsA6S7YGN9QvovJRf/R0mJf7MPl0urrTRVO2avDHma8tbcxQ/bIKk4l1fY3MybHpOice5knHT7+av5Nx1mJ6zxuxdfeE828OaFW3h7tnCGNfU+Eo7NX6VSFw3FfLd+wofVW7qTCHf5eVrrcb9LouHtEZGWDX65WgPPtJ/lDlsCK1jTur8nvBc95vbsivB2c7XznyfCo/3pVygjPZx3L/c/0t7/IgpCWQUVGCDny3QSO3+UKdqvWq7/MEm2cTR4iJ1eOxMR4FTiFSlIbWNPIaVilZAunT5uMZXYWyuD4YHaTv0n1JYi6ya9+XohFNX+cr2tPZkT4WS1IFh9ahtdjuV9tRuFjhXT3ou0ZkKUCLfzYrM2EkgNE8Uf6EDsy8LvaHsYWv91syjtp6zvsyl/1oCW9XuhrVVeWThHQEvKXS1I3xK1Fhju16hch9h3B+JJwVctZ/3jkf9Amm0qsa0xx0RPBZhagKFjWYEuIEO86CePZDdItTaPw4jZ7v6uZohbqZ6WNRuQqewrvaLy3UoNoOBZqZPPd9+41mMxBU+3Jarql0qG3yu3pESp3nWk8oqZaYkt+rpa2wX5XWKfBG1PSSpZCjAHrlR0VX6cy2eFQlgJXnOsZIcYe1xLeUy5XooYjEx/mVhFPRgAqqjr3S4d8WTg20VIIqYQYyg2TiTcs0JbPuKq/lNIjq252KTspZODPjOFFPTJmVxclyqnDs/QmrFrBQ18OQPqdCyL5TCCc+jLkyKft8olVeKsnLVlXZKZ20z32MtedVrXeF3Epj/u11G8zylYzsg16YFZHTmYzhQYVEYw79ifE5d4wx0D0C4lpLf843z//69SvNp/IzU/P82SvdJKbQtfrs6mtWnvoUSPxbm61ZuVqLatGYXM1n07vTt6rMrh2tWRok4h9//LHa+KkGkpbPHx+YuxLzWNsYG3FrPyp8OzXH/ruSkCRxEX5rDVb8ZLZ1hty5L+GbQ+iWj8h0yAaoYlJmXtr/kSKZJiqUdRmNUVGdP1quPZVvBtO26ctKRxW+8sQJ7SlfkX7tqv0NS6l9Yizi1x9/4DrAW51vTbq5rtd2msTfFs5nK7h/V7VtE02S4ehuyarvpEEYxe/v9xu/fv06WvFTBO4nlnN+GL0y5nTukdUKG5kEUOFfDTY9FZ/ssF7ap7zeN9d755EXNdD1CUXk7/cllk/vcg4U+qfjNczS9i2+R4asgakNwpblrpX9My1zRiyi4KRd5YOquOqyGMetz7BdWkJabiarM0nCdyBFnXyGezp5VKbSlv5pIILrzH8P5UebrfX3TQjnBBAymNl2XP7pXRLn5Cfp36cB6MCWFQC2IMEJgtb3KwKgYGow4Llfbo2qcLI/KcAhEK9GKnOfPTik9KoaucLPp4l+guBVYLqOvzFSnftc58MsAnvVQv3xxx8paq0Qri5VVfoDgawgyyKnudA+0qKflDYVSoWXTK3j79fL82G5SVrPbWrNr0bUOaRQs309hpXzWH30Twavln/ZctbfVydfr4giFv9FJ+QTgXW5oi5ttNaS1TwJ2G22or6frEJ6D4D1vCZarWaFHieibhCqMOLJ75i1pb+5WK10qeN4skb1vdo3vlv7wWUYVQI1YNPHWEkP5v/52L7OZX2mjsvbjWNBOfeuIM9KspYnWKt+ZRUoWjT9nIKpPEbEZwf6DcTFWZnXXThj+SjzeFV6p/IvWU4lUtUgMAl3yrPUXKqxd40f2osE3Rab2655FgxSn+uDtdzGZbasbR7XrtXP8PEspCf4VhUKiw8h3xs5O7LSwE4ntVXldFpzPBX16ZrUz7OHqoVcVm+MlfTwmuuCOt5KZ/38NIchgC0sl+wk0WyekwJQGvAd8t9AFs6vry+Z23y4uPZJfd4NXXBNVCwl29TTEzVApJlGFYVUOp/Kj66dv65fwjwebfPOe36ir1GdlzfM2sLaahUg1pLE+Oc//5kErU4M/+YAybQU5NYa2q9fG6RReMnEhdYafs3byE7QypqvpXEsSVuu+t8Yg9qU0byBMXi7VJxZ6utoA/f9FkbL1lahn5mtS37qkZa1mNnS9vW6CmVEP+B4LD+J716vF9rDuuHS7n1gGA/T6mnnRi1qqU4WXGkY85Kjxr9mEOo0lzputqPPfP3+jd+/f/sSynS3NKmBf2tygfIuf1chYtsquFWBcR4y+oi8YRopHeffFs5K2JM25O96tR6toa4pcp2OVpJ7CU7BjSqUOomVaDVsz3efBr5gGc6npK/3yoRxQll/JHk/bytTMDEmdHTh25dqnqyPCmxlDJ2LKrQn4YzvdFwzU2UmkTwG1qD9wzopIdGstL3TY7fqT/O0hN8M/V2Qg1nCLMon4Q7FHFFgVDifBFFTRutSEZ/VdvXnaXytGTxWdLaW/5blBLAspmp17di5c752FISjEI6EfpUBfTBtq7vWr0EVwjKtb4xYhtD3sLYJnZ3yJ4bSn/xd09ROdflII+jlyeEjKaytz9gnn4zGUi048Jx0X/sTdN4hvJ9PojEC2zbXZ2voCKq2V3lDf36yFPrdmBDbtRvAKPOqB3v8Yps7C6VFoVT/ku/wDhb1Q9euG7N5LSK7sbd3Uuy1b8q3dcwnnmP59vJcj8R2RB7rWTM+WdlPzz3VBeRF7TrA7Lvs767JlTb9X6TJhbDm904Lw6o9nzTukemK8KjVhMVFRyfhzPQTYZrZM9WCaxuVpvydviV3kFRGfY95KNlMgLASS3iaO9bfpF/6/Qn15LpEMBfd6NeVSK9F5lSlmy6JXK3htvax3YrU4jPfnaJJ+TbP5Plk6b5TPquuD3Vo+dHlub5kcNbe+hmQo6zsiGpbJYR2mH9r57+DRn0RMq8VKqEVwvr2t5wsnzf7nhIG8ha1J02ppcL/PJ4REFeeX4rCnCn6GBh3XzRbjNfOh1GrFX8SzvkHO7/qTWuN5hZTId0A0jGWWl+FwaE0900H628EN/nziXjzR2zfSvWXdqrwv14v/Pr1K/Ehlaqivkofug30DVnfp/KJF1RxnT4/IbJafuxz6t8naOVH64fjmxMKGq6LUHf3rarvNx4Yof6etwDtgrX3uQGNAYRJUJzhxRMSUKZjEEZLYtwphN8JMBUF2yVg9H75d6ftc7XNSk9+pgJToaG+Z63hKhDM+8k+7muOldlOyzy1DL/RZ/s8BAegEj25Nfp8HeMvCQgCwO+vr9jtVKztCs4kH/OwAlD6p+M/9eNTv2tW1N8WTq9sD4yctASkwVPHKrPzJy2a31/4Tm2dCBH1K/S95RiNHCLnJOz9Pvt9nwh3gp/1+/pcGqcxUBY7FRjpVQsOYK6p+fObQJptzF0hZTxqGz1P41ltzgOyzLAgsKMSse5lvGyniVUHYhthRQ+13VNUV33zjZcaz7id9c5/zeTWbBGC2+yosKjsTorvNNcnPqQtN1PlsityNV7aj0/lo3DqbpCzZQkm//zs7kvVuvToCyBr4G1BfG5Bo5XWELsLgi9pVIF8UhQKf9g3hb4x3syE38FZHYtqbqAd0EWB+JNwmgK5/slhU6e1Y5gd4Z9ajjEO/ZW+EJlUmMfDsQm7VQHq+awnmiqdlKb6u38Xc7BQVGvpqEysgCPXiqfln+3zdi9eh7Bo0XKqHnPANZahbtKnQkjuN5xlxV6DnJDPYdN16Z9O2fhWOGODcGi/3apVLaERVxLZGiLpd/7kBlfIhHDQMXhL9VGratvrnzeWtHcWyhhb0oI4Kw3X3rGTJNaoPJvlZG1Jp1r/U1FGjXds08TaT1oxXVb5er8dLgKAjfNB2w+K6Tz2vDShxcz3VUK2CaoCchpNIdf2tU1gO5dWYeEWDKx0w+JEjLELOv/preOr73P+NEqrc1Hdr1M5KXpdd+de0UzTcMe+F/1vD/gKzahZELrIqoKpWnpfMPZ/vc+j+CV/VbfzeLt5Z4MK5npm1kuNzTarNdNlFx7Xf7/9bJ4Fd4F1u7KuQeZ1RT/Gkf5t73Hodj1VbYOhI1+0U4VBrX5dnqGVXuKdjBkAACAASURBVMIyBtoYuF6v5Tfe8xSDVecAcF3ZCg6PYGv6GNumdeRzKvQvSdRICRLARms+o2hAoRyTIHrvEWCSgJT2977vZS2VRhVd8cqQ9/td/OSIhZB+auUVRZBv86kPJU1voY0c6FQB/z0vzf39+/fKSnpJMkVrngjyhChq+dbnJK6uGv47rauEjEnT7wAgD1znumrUp/ozVFPLs0fIevdDsdTyVj/KjFBpTzfUftW+6ERrOfWB5ZQgsFnLSucBjDbRhu0beFlH/JvnvFkEYaqCq/Sr4zCzlbdahbO2qfXXzzKjz7Z6pktSJmMkVKNjXXUWI1Fhto6BdYfLgo12MfZTJH6IyxDfqzIznMdc+bG+eyr/8n7OJ8ZRIgB5UBRMM6RIpMIsh+XPETz+vjC7RaSVIfAnjc71RT8X9S793dvjzx3uxDMnop7eqTCSP6l1qwuwOlUsa4LJIzKykl8mjOnIxM+YzQIu83AY12ls0W927fxesphjiIXCdiwm6zUGnkS4TnzAvpM8hMz1Al/ty69fv+ZBaGEYshI9p0Ke286xgAyMMn3MIpusWmAA8+A8r+PfEs6FlIoFrQ6wwqE6MCAElJ3jQKtlUes36bDVtz4qjBSEx/aZ3kWSLMMScj8PaIzwh84l4ExOVjAO6PjWJ4b+Dt4k5jAkJt4CD/Nnu5jvW31jAHOHxaK00vFx3IdnOW5URcS5m5AS4V+STyrDD9QTBPKVCloU1dy6nrpaxprXZjZ942eDchjkIkN9Xse4o6LIgd6XG50/mtm6NPi0LKblB5ZThxzaS4VzfXKAHs4LRWs/ECZrdyoDMvL+/BN8ONWbcL5oNAphk90hJwZiaXK3pkKxGtWdwz6XghqWwJS+PzFRXeC/e9+USbMnKAVQyZ4FResRwZvdrsslatGinjhXql0XWlk+UOWozBwIIY5oqRCaUD0u9Z29rrTrc0/oBwT0xIdKgyd0qMKrfvtpfXS11YHRoo5/SzhJ+Hz8/plhdABpTQlxXpALZhYoHTzzT0nwHOc8+y5r4Gbr7NFNQVTilnpQWyoMpL7nyY+oa2irT6jCKoxe6WZ7RJLfnxCJM/J6KDGh91+igwchr8x2gtfK+DrWvFwyANS0zmDQHJEN2FeZMpDAgF79WPs5e7/GXiGjBrcWrUt7lT9q/eyjPn/qh36u+0ZP7c2lCsQheHaQrVx+ZDlPDPeps2QohY4BfaTmUfZKTq17EsRUlxAlKQWL0P3pWeADlBixZKRY4ZMf+eR38tm64BxWyj8LP3nw/7lLs18V1tX2/AgNHYovN40B3De3LJG28Qz7oDLJC311jpMSsgrDR6qv/p4VaRa6iIyTjvfqt5mlDQ06Zwrj6V9XRHYSPlWymx+ozxb0Q3SCgW0JTfuUkNQ8xW89MzWTxgjcUP2bAaEn06vCVa1gGviH+upAv8vsYTunZQl/bx6zMS3RkLaqlaO1pgCoX1oTDWronfVV61mZuv70CcwBkEQTsxWl5ITq8Rf67ILmrSXmqIphg/VrVnaBOindZFlHHMehLksd48niVwHmxUL62FIWhca1HoXDYXWz/329XtuJFBken/laUc1SPp0bC3bjpJtC2Ha+82akPo7pth2DgVI+w1oLf2h+sn7Ppp93QchgBi97YUSyu0PcmqeItcYQrQcDeA30wnyWLUlQ7GgJK+5v3fw+Cuy+cCLu/M5hSRxCVs9BOmrYw9/6WV0PJuGWdR5I4fylAKQe/uOp8KqwTksxkPGq8NPPzArh0OcrgnXpGTPYML+jZfQ1b/x5McvEUBjxzHy+37JDj+9IvnyZ50wP33sKZHRBAVmpeECi5Wm+qnI/oUD6rgNj3eLNcfYx0NFxj7H4d0wU1JXvrgb0MSPkJ6i+l8/CCaBDtcIOG0ITxY7/xeDmzNdac1JaZGxEx8w3Xt/AwICNJZu+UD3/h+5tvRYE0vs3DGO8cd9f+P37N9RWP1k50hJCRPbZL5tpR81WEcIJ4mibIfhxVEuadLSU4FHbigX03d8FssWrkJq+jV5WXN+vPwcV7ODvk3aLKRr63ZcP1b3ByazzXzP0ft7lg5X9NFsbBAiR8FCj0Tpv/nsIAAM/iryYoKHXMDxZy3QoV9MdSjpHrrT4ezesbXs2Bs1pCGbvuHsk4bR5eVFrNhVJFtCn8qMrADXP8EnS44iOk480lsXU0wRAog+3VP2+Yx3UG3OiIpY2bEIG3TTNPvL4DV/msDgWwnJoe1YdigVZa/qza8QgDK3QLy/NtEATc9IUStLHirs5M9S5C8Sl4lKaE4YrWNyZX+GbIBzLCuokoJwrzLA/RXKNe/aICfljjAXfKpzXc5hCydnqiwaW9FBvVVRqDWO8cxwSaa9tx3wGjU5lQ1Ssa41W1Dz9VU14WP2PU+AdSeyrGF0i6mbnY1Vr+dE6Zw0PK+NwgNf1gqY2VaZpraXbhk8+SUyEasrMPOOgdapPtVm7SRHWWYmin/NvFeIxCswpwsnnVAFU7U/LHAIulrt3P8OHEHXRV8c1kcXdMcxgzQojEu4BWDes7YGkJwE90YQC7psMppIuNKx+eO99HZ7MO0iJwWYrwMg0VzRCtKBlzeukipmtRfynMTyhmtPvJ4u6lSGGx+JmM76vdVb+WzwAW8psUhmK8mr5dj/ndflJ2ydYOJuWiQwNSQGK9cSI9Kkgad6oDiygZxam9/1ebbYi7Poc07fMLMHKfHpaJCxoYCEIbqi7E7SfqmTYH1oJhfzMUGrN83NJN1oBPDAGBbxGAh1O2YJic7aKtc8WBavf5+aeNHhrfjo7FYsquieamHWfbwR9ew+IyL4oIvBbqneXIClKQOZLjMOhD5XXNoVkAdXPCAKp/t6nXzn71FtDm7xeM5RqXatPY8D6WArO13L/pnAuDXYfBlcGprrA+eWkoWdYvwhitUyqSTk4ln7f7oWuyfGzXnmtH/vkO+L/ABB3cVIrZ4UQ410CFWT9RJ41wSqYizZWfVPd4SLwV97NZSTmqsLZRChBCLeYOPaMKv2YUzqGCnLW9gPYvku9OlgG/e6+bzRrGIesnJMSXe9SgApcrYhN/62+sP6pPPhZpl1sWlcBZb0JAQjKUeGkkNPYRMBvj1hrvcA8Iua+YX1uWpiK9W+n7y0HWehIhg4/wSd6Od6bphNNtYxG3vVx8psUNmupUKj3vo7Dr9E6v/syL5DrlWwxIfkyI5XYE0RnuR6YiHWcYZUuPOezUk9t9DG2w5J1PHCwlOiKD8KFFVE/WYlJkwQXAwXpGUyqmE7K9PQvL/dkv7D3HksQ0qcMfxm0YVCotNk9MFWXlFg3/eRqkTM9n5QJRJleeL2u7ZC3U71Ebm5E3EdvU4ksRfJQPvucY6wlD9VG0z7OSOyuzdjRINBAM8BwTtTW51ddB3M/xoC1uFi3QhozP4N1neHa+9LIr9fLGV0CDPF7B3AtB1OZr8IVJXr1U/OY5TgK0CKFn6Rj3fzgkWG79mPM74fMBaH1U25z1eqsO9PfWxroM0urKh7/fVkJ8gZmNH7yx3xp64f2RS2ZtTg31l+Vc57UElkcsq2ClOclK536TG4/Srhqef403dMM089VZZVPcM8KP1AY17bVBbx7hwnaO5Vvo7W9d1Ac7hQ9GzCLpOaT75i05rsDbQ6g+8ReI+CWQ4YIAFVh4IB5HswShvm5nknKrULvr68lFK9fvzzEfffE1M+Dx2ZcqgKqfir7ebKY1b/R+vTd9V3L5/kkbS5z4+/ZVl+1BPp9VQq+R9Q3AHMLl75Tl2E4F8saYcCYPAAsS15po3V436+AkatPvgwzOy7v7Ve2k9lr3ae/6/yodU0rCIu2lQ6RhcQYhiKu+m+5daCbNmnJM4ruG+9xRmQsH4WTcGD5k8aIoCwD3DcMSPmWhL5KpA6HHTbD9Bi+SZkblQFPjTpNJKGVh6qx3cXC57lFin8DWBki7FdsoeorILQYrghUjUrWqDUX0fUZvqdQj2NOn63xxYlvYwy8aPXf78U4nIPRM0RfMD0xfQRhnHn8d+oaClVSLMPD/86QQTv2C/DT61+vX4numilkkOWREXM2jR586WvAWrgDsyIwqWCMERdiyZoj50su8oo1xJKUoTCTQlevs6jZU9c8N0nRiSKcMWMikLHH4QBxEqJmrZEqnOvX64LhhZf4qd/FNL61nPV3XePSz08aCRAhE7jI73X3vRKT76mm5MWyZCY+o8X/PGvrCKg0jHGnv9UXCeGhMMcaVl0CcmHNx1yEMlEaYH1W6WoW6V+uaQPG+rLENff/zQNKR/aF1JrVoFH9vn6n/RlTQE9Q0GFmRK5VyaxxUEimT4hBg8g14wnrJPKb4N9oKaOmlX7PnmJwpb/MryoUlsi73SHnXrIr0aZQv9/v6UZgrVoo7FUFfgp4pXEi+nldvEtl31PK8qPc2gr/6sREpJTaUsz61JrDsI6k0H/a+cek9G/6lKEi0tUQLBpI4vwxcst+a1+U2FQk73kciC6daB8+Rd5S330ACV3UOplssZSRWIdq3Wll6jxVpXGC0Qku9yycKtiKHCrNM4TuMOhSWw3q5CU07VObE/gxKX3SpqKU++6S6RXFjUlecjvxRe99HVCmFtjM8B430IFm+zUNtT2lncvAFOQJ/e/e5wmHhut6bWu6Wv6tkxCopXrv+P3799KceprZBFNusRAW86T9dbCP2l3C2d/1scIdrZ8QTC8Auos/W+Hje0LweimO9sXh5y4kSaCI9UrmUWJUM3S4P8gbresdjyxXy4nvKliVTmphBtz3cT98zgk6uLZcaaeW4onuGMA9+qJBtSR1rm3Soi7OV8tfyy4gEpAq7z3VoUqRv2vknJDeANx9oF37ribdODFK3d63RRzvJ9f92wVrhtf1LIJ/Szj5U3+n/3FdF+zab/A1swWb9LsnTVSt8xK4f7G/VcPrmhSFbH03IsCkUFj7RA37er3wmhfkpHjhGLhH5Gsq7Nkiqgd4WJWKtk+/rPbru99VOaRo9RhTOAWyX/tSAAVIx8HvalvLAndsaWxP88L45+6m2KJjHVcEDxu4u4XfVsurlv+0XkxeCD5c4bsZUTbgcJrerFzGvR/OFs8BDW0uw7zSAW1P5cdbxk6CUrXi6b3FkMadK5kwervw18T32sapzidNeHpmjKjrFuim0JWMRyYBcr5k7ct1Xfjzzz89ycHc+syG43cZf50wtq2+5RPtKlJR2n+CVieLpW3VugZy/2pfsJIdnq2RK7guQb99bhYvSJAm9d9sMTzh7zYuYxR0+rNCT1WAJwRUjQZ5MEdsuXrgz73atVYq9tv0YrWByp+uXTI0AOxq+PXLlfprrpH+7aUU1Vyj+9aYo6Bi11YLrlCDGNBGDjfTAq0TxvtMkRIiV8tHbVO1ojLs0rQAxoRrcbdoaLp60WkNJpDwWq7rwh9//IH//u//xvV64f31hbe0fY+xlmtoZdi3LUpaaB2fZyGue1fVuvKZGlk+MeMmkNKHZtmS1HZmJdJubA5IiroPvNor+W38TnmkRp4XzUf4+mxXBWfxQIt00JPAPfGHPqsnvquVc16IOePPLn1jsAkIXjr1RenLtjTHnEdonsqPjsY0M6AB1q0wUTT8mndUqOWDMtGMODZhVIUSMeBztHYNVlIJk2YqjNt7TztDNHStjMzzUQ2xVstnFA6pgL1+/cJ///d/+2cSQuckcXeMWj/6a3po1YlpMbNfNHL9fr/x9fXG+/21gjJ6ch/9YE64zp2iBLWcaf7mElSzhnbZXFbJG7N9Dk0g/Fifa3uuqempPfASsBBNbid/VhVFSmBpvk6ucHJ9rjGAUjfprTGH6ken+PA0PlQE77nzJE63N1ly4lYz2fDNJAWOBYFa9HzbU/mXTkL4iI/FTJ+ExZqhwQk32u4T1GDDk3B6/mTW5ouII+BfrjugRvX7zGwlUexXBGIfh7kT/8eff+L99XUcx/t+L6H/JftX62K+FrWq9e6OMca6R0bRAa8c0DU/IPYosuh4anEG4jGaDe3KR26QxFxsDzrEonxS1vOzLv5xVRKuVHLE/oTIdF1ZYflSShRasVb1ekCtsyp+xhy0LR+CFVTQ1xKXpn/WvvL3e0JiDDdo18U0HlfmfSpnvWH8VD4K56tdMxJIzQV40nScfsax++8etm7NHFu/fs0zV4BxD2B0oHeYqw/08V4nk1/WFjS4ZsQX3Retx3ze+kCTiOUi/PXCGG/0Hks6fqsZhXrM5ZNsPWq0dcyI3xieIDFGhPw5ea/XC9YMv//6C/d946+//sL/+l//C3/99VcwwIJ4PVnQBHMar6TPgZ8TZH+/32u/5x9//JGgOIAEy6qFebL+KcI8MPsC+GasuT8X3DlxgYkjXs+kFQwYtp7D5Ae/xiOnaVYFRvpr+lsSyGZAa+izXYeyvvzgt8XZyjse8KU6/7vjvseKiLdXPl5T54KrDBU2sx/r3z13lAyA4T8bwPu3nzL/ahfs1x/4+vry5bb32xXGdcEurAMERrvwNW789dfvJZx/2+ccvc+8WoBwi9ALK1fQn1XNQzzOZ4c7f7PSmOTFzIDnWHbxaYc7/mvtjYpZGDwgM9O+aFGCWaNf2cfb4Y4HGXRyqg+zmGlCEhL46+tracDKbCcGDUKc+8O/NXiDkY8o2RDFFL67BFoq050CPpBxL2sz53xFLgeh6KTlKH0eZ/9PhbKO9RRYifc8p3u9bzOTaAqds5EfmZLQzeSBYQ6sfYeMz4v6hYSiyyCIS/M0BtYHzFWD0TF6WGz/bKflOqYTfmjCfeeNGk/l21vGnCn3x06wLMOeOllzYrBHCn0JIwcB+oiLjp7gif7N3Msx/N5MICsMZ8Bz/10PTEsgRSeJR2i05haewqlErgJThVvbo3DWsbWJVBgxVCE91c3Pfv/+fbifJlurE5zWcepJDaSZV7cHb04Cru1WNHCCrjqepMRaQxsIC5rac+Uweg7E8Jlj/eJfUiDrzXSc20pbKtv7vv1IkgBGM0BGOn1398lYp9NrjOLTuvEPjyk5hLyRJ6gKow56jgPE3PlzJXr2j7JwPReG0JnnWOGcEo4E2YIAIjCnohfW9N7x119/Lfhbo4tA+H+frObJsno/gkbvt18hIUTMYxfmruemnpQY263f+Wfn5auTRXkqFMwaPT4r1B0FAVjbxqzZGrMK3tPcsu06AkMOsFE4dSyqSKnkeRSOLoelOTuM/5OQMjBZk3CeykfhrMsJJwumg9Of9fmThgnhP8Oe0+8n7cj6tb8Vtvgz1Iz1p/eN1VY4qJaAwvn19YXW/MLfCoP0b4VLn8ZWyxjM55XMJSCS168rMccTpDzByn2OAFrHT++eBOpEr4qiTuM98U5CXeBySaXL7hPW7wktrUXiiwoe0U+lBecLZhglqMb12dH36HL05zSfnJez0vvEA9+uc5rZdPA/M1OduGp1zdxHIDMsv8JsTsDPmLZ+f2LKGiGscCX6qsy3+6WsU+vWyTiN/2SdqkJRK737gDu8Xn1UFDOjfk8C9N0/oB7vcla+J1o8/f2vlE+WPf7mf6LsqwEAsFvmWcOiW223nkmsypplcznM1lWRjMFkVJEDeqrUyfP1WJJPNPxWOOuBuadKKRBP61L+jB8n4kxLLXVONHjqC7+vgl8Jwb/1pqfTqeC5rd1CaznBmgqndNtcfRZz32NllFqiT4fcYBXOA9RUKJl+4jlA47ztTN7dXK/bnnM/TSL0Uq/0u4771JdPJVw59jHoUGkUrgRwUmar78xWQt5mpiUsal6yqWvR2v7p5wkiZ97GxqN/WzjDVO9Q5kkL15C+1qVbjhRi6AD5+8nyPAk+sJ+GntuJfZ7VR9H2tJwEVH1LPWNVz6XR/lW/S4vS75PFMtdkH9dItb7q/2r9R4WzGMyA3vEewDUAe+UNyBSCMeopiU/KCHtbh/IR2s+ocB2H0kv9wQ0djDGXhHIaZviV53nXf2FRJ0+0BkhCS6rjgFBOmWbK/39bOLlQzGhtNdls5CeTQoatgmlmS3vr8/X3k+U8acL4LpZ5WOiMA9is3HofZz1ctSgDDzXwQxeAdcUEPy+t8LnKZPw9drGcYLD0Xeii75/G6cWhvH831y+nNfIT96LeoPFIwhkjPQvnx776L8dxR7+ex1kV1ZPS4t7MrX9G33LnLe1P8LjjH43YJkFGCB6Q58rnnxv8z5a1lh/5nLrYrY1/YrhqUYE9ZS++cw18EpislSM96jRIWrL3vFY+1hxXbciiZ+u7ZOlLnZUmqz89b1/zZ8dKVTzRRJkjxnQvrdwK3eppgbWuutxCq66WVgMnGW342vAYvKhnrD23nlbm9PEDw4PWPNjqpLAqEjihgsUb5WYu9pswfvR7zcWC9jJGtYT8vFqkmweVz3eYXqk8S/eqPr9HnAFrvs5dl2PYP15gpLx+UjKfDBrLN8LphwL7z7iMlaHmav4XpxfTXoX3E5wkkfl9fm6gyzH3A5aOvIiqzpkps2ulb3sY6MRYpzEoYTVQgUL0TVsXNKEL4U126TjjxPJNTRnT32v/npaLFpMVq7CueC+BJp97P9ZzFzKez5CFrkLGE+ysdCaCIk367E/1aZ8Y+jQn671J91PbVZifrDX524DtaNdE5+67nEyUo9KrtTMiPJVvLad3/EqDr7h8WT++N4mRLWyOsiF1yt+vk8nfHzXxGGQN1GsOquY6MfAiXGlPLebTZAGhLZ+shNa3aDpHqxk9PAeoJSbOh52dfBedo8osmRGzWwHY9oxZHMtBpeDPZCWiffAAUVUAQf/qdijdsw82haPwFGEi/2bq4HwpCS5sF8zVJxJB5k1pl55bdUrdwEI1PfWFqO8Qk5l/95kd9H04bC/frnM6cWLpgRuSlWCAHEAd3VxrQ97JwrgHpu9T64Dabj6nzH+GTuEvsX9VGzK9TqOqn8onoVwj/MAQ+TtBFFJ31d75DKXoxymiXTW9wsM6jqoYNcXwqd7Tdyuhfo7jpLCfaPFEL/ruZOlqVbReug2n+p/tT9lg3ndLr3wc6rPwgCILScfTeXpqN2QljJCiiafyw3VONjxS4KZaI+3k0v5TWFaWi5RqITkZqw1ph89c15XuPszEmO1YTqLm9X4kOn3CpR0P7QDBBvlKd1dSA7EH0sr31nhAVex5rHRloeBVeBX1KlQE9LaxEw0rLdUnPcFM/lTLrHBa6450zgZYjh7X6LlwzJrTilhqpD6am88YFe9hC1tJF3wqZ5SXv6sI4ru6vAfYUFl+Fp4Aoe4OgDE+KzItP95sXTUWgwIKv045oCeYx87r36dnQUGNh32HgP+RNPgAJiyM83Ty6e6ixcSv4t9P/cRcJ4sJpF+JtPVNGTzcgV2INMd1+XuWgyleR4loj/AVdW4I2Z6gP39PgoUKc2PO9J5MophTeiJ/rePvMldJKHrO7XXhsyW07E+b8NGaH6fKBHsVJEg2V63zJFxEVJqArn08KcVTP/d+cF6FLxC8TCubLffPUBnww/2c7lsEMeLzsjBrlja+6neVYcz+Hg6/79j/1tq8iNcaNPFCCRKMrcGbAeBOBB4jw8yqUWMsfE4RQt78nK53mBNbd0Sw+PusNyKgfkLElRi3WrwxsDZ7V6bRQpci32dKgc0nGiQBBDBghWHFohfBBvxM2mQZRrhBvMuEND7DSqC9zj5onoex1fNUzAx4UMB9DAw5ESMr1rbWxwewfPKVErqYeI9nZEjLZ5R/8jhO5dvEd7PwM9dAsWuYex4uTc1ZNfexE7PjqoUr9NLP1QdZTDsG/GrerBjUMtS7UMbouLstuKrCtv4+MF6FR9Xq5MAH+x000d0ISkNaqi08L3tJn9Z077IcUf9Btvrp4nsIQ7ZoT/SvVtcN2BDfK9PiaMXnTWMuK30JoN4kwAPXSLsEJRXtHBTRSUjNbB2leXyv+9E4RGDZikadFMy//voLv3//xhgda4O6xAyWYINjnPXA1rnDQcPPSuVHsHZMf9F5llDXMM+jmEGOLvDRU9XoppsFNN7xfXT2JsOM9Z+t1IwQQlQSUGFjWLXD6X5lySCaFaY8BGKS9eI/DlKE0mhhaDmmpRlicQbOUIzCw5MWtN+kgX5+SuJeY7DM1PqcwtW0poeBfo+p0GS9dH6HaeWveSC36wBte1oJYLrotn5fVnrRs0e9mH4aht8QUIRSB1UFiX3P4/B9SqpcqoKdD6+5iBFki373jq/3jb9+/8Zf//wLfXRc7cLr1wutXevgLp5tZHA3qw/f7eN/w0e2lgIBtH3OWH54hlBYlz7vF0RvaG0Avc2lgRtj3BiDJ5dzTcfD7Wodz5AX60S2KsCqZU7f1aCHTsYpsMF3Z48436Bnzb2k9R19t7UGtLj8KJgwMzSGK6k2JoO2MTeQB1P2MeZyTgRixpCrCQR69t7XOTn3PK7xfd/phPSKUqxYzCrc9Jva1WDNpiW70Ybhehkua+gIaN7m/DodEEpq9dHH1KfI+QEZ5idgLIU/xXCMeXvDPO/4ujAozMiKiOmkxkSPmZZ58okdhrrl4v0kT4qORf1tAxJP3X3gHgPvPvxAtz4AG0AfaOOGXRf+MD+Cxyu+YA1oL4fUZvDjQu1Gf2P64ChKLZd/6QyhNfABwAhJs3ZTOAXIvNl3vsGzz7QxW4FQT89kAfwU6vdnkp90sGZV8Aevpoia4lkqEEaGr4GBGWluA300v/xn/mPQiUsLY27MpXI0+T7GJj5stX5iEc0cwjMzpvq8pD9KBhWtR6wEBUpIyGEKpqV3xxq/TYWgflgfBd5TKEpwpv40i2AQ+1vhLuewzXmtIKzyGM+QVV9zGaRiZdf3l61b1tTKLrpMuukcDHOFdF1DKPRcfhgQyifRbf7Z6rT+O3+v5bNA5WeetmjxdxaFtVLLcVzaz+/6oWWlk5ktqF+dfF2SOPXAALcQA65SZTG7mWEsRDxhEa1xa2VpZ6nBo5/HqCtvdONdkUvY577QMXHohIV1egAAIABJREFUeBg2x7Ll1Q76/dsb0EuleplHWkHGAngw15OyVdpWJOPphdnSAe6zN4dCmwCvf4glMZhttF3tztWB1vwYHI5r3byOesxKjF0j33Ucn/js2ySEU+Bj/pYIRIi1NMehnODlo6N+KE/wVvuqyw/fTfSpf0dlYqF9kwKS7UincdFa9N49j7SLv7HG3df9K16/p8TVsWmgoUL463ph+e6HMVE46vGZ13Ut4bxnH9UP1rlU+FfTCMfdMeTYzLDaIRBtDNyz7ljWiaSQenPckLiDKugatfbx11xtL1drfiCXiU8rRmYM+oShaGbN07eesJzr+wMzzzn6ttwuhHBqYJA+OWnJ9smvf/uuFE24ZnGG498x4NirWaOgGoBQ1ikabLVBa0bYVnsVDBPwWayE1PGkDLRP0V4Wfq3fMDDmKYAq+IZnZiKz+LjnIcXbgvRYGjm7AxbQEQJ3h6T7IRhNr07UtUsyH3qcV6NKp7WGl2w2WBuJLRi/lX4oDE4opQe8p2Bqe5XupFei5+KTnoNzlvdYZnY4u0EB5+dJkYlXJTWzJk8crNsC/DZp9sqWT/mmGhqzTLdFVzEmT+VHwlnPQb0uWwxJIld/TCcltFq2CEdraSvGC/4wKq4RGgir7pjwx21g1IhjX7rR5is8nd0B4BsA4h2uw2WYpW3HxPEz8Uw5kZh+Wpq8qQScE9IEci4ix5Q+F9J9ltpGl+Wburtlg4gyh0gb1TMaIpSrNK6KLQtdBLsqb1TIOSuMuplUUqLLzSOOSXA1OEir2lpbcP100nw67AtT3auy5ZhaW9HfkwWvSubEh6dltKfy7a4UHjbFChf3rIcAG3Hnia6lbdB1jlSFg5Xkw6UCUtmUplUXMpJMvpUQTok9H0z1l4HmMa0+Y1mA/DiFa2zjrM8swen7YWUN8CyjscPoquS2ttfG9UkvIE22WiE9JKz2YZS5pNVUQVr0A797VqhmZyas//ZEj8wDiqg8LXMswXLr9VpbzBYyGTkIxvH3IJS319q6eYDPUJHS0CToDpeFqxm6Br2KcFYUUF22aui+y/P+JiC0m22TgXLSYM97CJdQyrMcEH/6IC1uaZ75qdpWFQAVimw5c8ICrWXVYloCGOuwwtpVAVnt12R+1lfGpnB2PZ9gTzteFruEbSqPmmZWb+beIfvsK8LH8joc6tXLiv2axrEip5FvO31jTD+u8SzjijJirtayy1HBZCSyyGEtBZZi3TraouJ4vV7r+kaPQvumae034Mjh653vVDXk9fK4K3QuA/FdXZq7DLHla493nCwnaQGpT5NJal9r+dG18yeLedLoZw0/1l2NNRq3rCywrllfAtwlVC3tKlFDs2Y/JC4ocibw76Nv9YT0quF4hg4FY/cjpiA9MCdLCExEbZn+tbSt2VqMWc8XiLjyUhuZ12B2w8zH1Vosk6TN7FJfpT/nt0Iw/7suH3ibOtymy0jFgnAKFV6S3hTa0+4QKkm1YNqvU6BPGVyVwaLd3dP6efUPV90zMMZMrbf0b2qONM/KC7UPFV7zOc/37uvQ8+9SD78VzkSM8n3t5CJIIeo8JDuYpuaJFovyaKkm8/G7Kqia6E4tpxADZttOEufdvqwQoYb6r1V42W9a5DrZtTA3WaHWGMC1zukJmFotp9IybgLzbXz+fGyEXgLMYzNnFpRajZPLoT+TUh3h085ZQGTHTl8Xsfk4xtuWFVIh42YEnjagc1GFrv5j32j5Y54lEHWw0q5TwkCoACfatpx4kPo0/w3pQ4W0VC0n4VQ0STdpnYP8dy1nIsAD41XmPMEybvGqz1GIuB9TL0S9rpbg8LjzRl/F6qwrt1+9083TLB/G8xWW1n6Hgtmjr/X908StPpIOoxc/3lJfPzEvGYG0JuykLXq99s3OO2UqSaIdChkVSAPmNQlB+8qEcSmU7LMVhl13XApjVjqehDPGgEXXKgRjrnmeitazCX6lQbHUw4If+U4E5rJy7WVeWFgfBbNmr9XybeI7oe3T92qV/DOg+mpqLSuBYp+nRnsJgyT4MbLfpINlP9TPYfthkB8geFE6OilpfFICRu+BoFoqTBxjRERwtt97Fk5q6/Wz9L+OJ+DsjLAuZAEA6jtOH8oryMsVqzI8SG6dt/ORKWzn9LkKb/W3qmDWNU0fT/XtdoPAASxa953nnsqqayIQG3KqAyH8RHoZBfFvLrFllKl95tWB1/VKFvhUfnQdw9N30noKP1eCkqjv+17w4hMUrJaCFxQBZ95Zz7ccUIi6A5pVpqb1qlbvUx9D2+5+p447QerTWKegsh+1jDHSOqP2SyN94dfs7+s7Ge6Zr/Gl8cwdNB0wiciqPwUg+cxVeVW4uPf7gp60/0lwNthZ+OKpqIDXuVX0U9tdz/e+FFiG3vkANpZAcw197te8xpgpnDMZY6Ka1+uFNi/Q/W4c/7JwPv496F/Ed5mQCyyl706wsBIwLGrWNBtsmLDGrVMI3qk/1ZI9jbkKbdLmzWDYl5uqs18ZpCbV09d50qGntbFPEFCTH56sF61ATRoP2vnhaVUg9HqLWu9Kayx1Kh0ZSDrRhorXigAkmhYlrTyiba11z3m35rN9ChqrtV4Iav5rMKDl1MnlvwMzgBiIQK/hWILd8hJVRSC1fCucP/p8ePg5SLxbneWfze/09iZCUYWRlQmqANf6Fer4JV1jBSp2S1xglwyHBDtNem3XYL7lp5+SFwjn94BGem4MGPyaunWqGxACW4Sw+o/8nYGWuHF596s0fD+Gp9Mpc3ldoUDMqnAagPPOoTp2jQlUpgf2iDlm7TT/bQa2CCGX5eQ8SbBL263rp3QN6rzo3xVK++dccnLhHB+EU8dMNJCEjn0BUr8+IVPgh/s5f1KeGHrzIbpvUlVnWDWVMqNeEqT1PVknXUNKOU0HgmRoBhAs994TnPlEh4FYQ62KZEGloVa14bpiUhMcnbtUnopCL/ab64DcBKzjq2uitaTPqMSap/MlYUgWTi4QLvTU8Sjy4ed13VPnTOvQPl+t+pjnhX3SVpEFLfg9gLtLjPm7OZ10uC7+cxjORIaE1IRf23znhBaelPMnwQR+EK19mtgTltfnq0bKGnzPplCG0BMB6qRVrUj44Fkw/cgAGBGwUoY5BbrGmOcUXXEwdB1bPOuIIGvcYKLWIr2M4wauRBNAN6U/R8XXWIDl35Ppv76+8DU3ZistP82f/mSLa2dIEUyPPoZwVsWqx6QAchLgHH/NiV2+sgha5Sdg5xGlQ+ULXzbLyMDM0HUj+hDXwOJkgkAFgRo8J5drxr4WyCNbNiVSoq67UtP1Xm+roo9T+dFm60q82YN1gSlwvhn4XA+WtlcGqtE/Fbw6uduAzPzEu3GBIW0gr1m6Jh2473eysJoCFgqhu2MPpOsbNOrZe/cNtSIIY0QSAC/YVUtf++6Qi/UO4L6BLjdTm/kNzRSS+w2Mgdt9A9+d3298yZje7zd+/fqVAkbarloy3i06ieFZWYWp2oRydjXE2uoZji1B7b6FzCG7+2I3Ecbd8e5fIngejTCb2+QYUBzDr5KXuarrz5ybauV1zNbauvfFplB83URkESgc5pvDO3wb22hARweGoZn3xaC53AZuz8/WH1hR26UoKcAXbLSplLrT82mPHn4YEHrCx87E2TdUjaJafHCyyGgqjMiwVAn9pDGrIohE8TwGb58XGXW836HtmAamfXLlca8DnoG8n7XCRRO/qCZtVD+rMr7XbWjX1KYYgDAkTHZTYKTFfjNDB4MecV5vZeYNAk5rogGQBQkRh3/VueTP03XpdT4aI5QqMEXJjt5ncCfWLQ0zcWFgmXMds5636zwT7uSTi6P8ZDQoY+DuN5q1hVpIQw8AXeu4kTH/F96utxkWMDaBYD3DPhs0Mo0Ylq9apHf28m2GEK3B6TsGO2rGD8PGrfy7VchJ4NMaXylLwEdOHN4n4DxYEk/9A4XR5ekNMrHoe/ODtRtkf/bC65UhX62zavukzAiHRPC3MckWvN5vj4TNfFP1u6pbQXKrZm/TcmJasqfiqMfvojmNw9HA/jn78pSytlliJ2rqSzUQa6P2qjOUoiYnnOayjlF5o137Guwn+EkF8+RPugKcgxqHsT6UH187X8Pyt2jevKRhuIB1TIdqcNe8PRIPhCBpqAd4wv5o0KEmHbCuSswQzohGqnVTBmpFEdXAE/tD7a07+GtUkhfnVKbKflTe+tZaW8c1VgtQ69kQxA8nfadtaPz5qXyXLbm3tSuKlFTw0I9Pflmac7NNYbvl3yO7OZmBO1jOB2nvfm30HxDo3HJ6H63gOiaz1FXHqOMJ1AJALuCqNDiVHx1TcsruwcTNtKDcRhRtHSJbwIpIqkNMpvx0ipwO5ongkVVUNV22nByT1sW6r+ljKTzUdlOQ6toPjjq5Ahvsa3HwmNd5L81qFtqVgTG1pqeybiKziOjyPVUWwRSZlguRmMFwzeDKc7hffe9FYUUGI3zHpzpqScoLU7lPKx7fzXnkc0nJ5ejok/tT+5F5B0nY91jHtdV5aiOeVyH0kMLTM6fy4ysAn7S2f5bX3zwociDcYYLUKj0tyG5C/mGiT9D2pDWrJn7q18mHXIzY+rzWIU9QTgKQ9v2B7XmeElcRQn3uhChI91+//ljHnfB9XSvOVmtXKC6btA55nbTGApLvdxg3hvudCsdP6OFTMf5XHktCASR34kkwTwqivqPKT+ea/nVd+jrRZqCeIbSnGjLirX34ty1nhWZL4xciaBZHE2hQM3k0CKDUr+ucSgiFnp8meZTP66SxDiV4FZTeb4Gz561AANDmBmA9LJiTEYc9kzkDDjq04ed0C04WWMY1MnxPAtPmfZXi06mC4JifEEBqsyzuk0m9rXkNpGyvUkWwlrIATxT/xlpW4U7zKZYzzS8Fbv7tis3W/TR1rnd4enZr6BOP4evrPqbzYXa1DWCPdGTBJL9/D2W1fBROQiPe0MVGMbu8tBt4fMbA64Vt3efU4RhkWJd0JOGBGHXAW5+mdVatRKE4CXTWhjZHFdBwPoyG/Qwbm8yjjL5QgO2XJ/G6PCINXeeMXTkm/WBEsCVB0mWbVceIU/YImSs0M2sfGdjMDwjnmUPKvCHgwBgN6J6oocLO5+/7ntHO3TI8/V7nmS6QPpsFjPM/aQdD65a2x53cipNwsqxT54kk5u+6gyT6iMUr+QQPKsu8LBWIYs9T/lR+tJ+zRkVzAGcsbeOBn11j7oIFMBCiTM/1uaOPK+0DexCJ7WhbwfxnTVpvJl7wBnN9k35pi8uCWcea4AJP1LK7L6nQPveV9KuQUwMy1hr6798LXj1BZrcwthIIeIwlxzlG3Lu5FtUTrHMBDfUkaIHPFZQU/Q6FvYZ4UKI13e1sOWVHkYqoRZCIZO9U6rPdVZ95lNeauSIpvFHL4nMiRCDdWv7r16/VvtePda4V+ZjzazaDhBsKGgBa4YXPAvpRONOlLZBjDBNs22FJhQ1kLGUuLWQaZ4acylcTBj4V1eInral9q9qaewwBnhsblkCXb9L4RGDpTyr0TYkZ4ETtydxUIuGrxJIFL27SNctYkAfeb5+nSxbpNXlDxxiKtkSoF4THSvQGgrHIdWxbc14ByM5+VwDjfeN+37j+2Dd314CL9pPwlKiEvGtkdqmLQqnZTGve+9mvO6EnPYt38cWIZBIehQIg3emivMB6mb7HfzWgxLH+er3wejkNFJXW8qPN1qpxdr9olAk+Y2qFYao0AqaQgUbWnodJrfXG78uQbc9U7bbVYXEmj7ZdraseAwJgXSs3RviPCnOVkdAMvWehcO3u59/M3mAiVRCOqx9JoVblp36++vqkdSg7r1MVQ1oqmqjaCs2weif/DULNK/u87x0DX19v9HHj7hfatOqKPDRSHgxseL1izXg4uaYFjoV8Xb7jXGZ0ck5m4d+5xFyQjt6//aCwykv7d3s+coX9/R7+r0Vs4G8f8EVmVOnWa9zYgTTJhzXLwNlVUALajXE+4+bUJ7WEta4nAeUzPNbj8KX7XJ11BeTW83nMGl4vJyp3dbBuQsBP/g37rMpmwDBSgvx5l0uNIJMGTXb/a3+W1VtDzIdL6aFs3g+s1LIErY2Qv/irNrN6qHAB8AyG3vWqCVuKi3UqzXwcJdZg9L8jmMKia5kaZT25QUq7ynvKR/ze50RONeg56+tkePRY0jrX9fnKTyeDw/LtUoo6sAsyrAYjb5CQ8xLIUjVJXbyuRTt6giCb9Z598L5F/Xw1TQYQoXezxLTapmpeRiEDvhteL9nvB/julyGpdYd6uf6qFi3BMDwLc/TtTDPXvi0dr/mkKPh8QjnCWOy/APXZ/91/ZKcCrmdIPw5tk6bV7ZgdW60asKLgNgNp7J+3Ea7G00VV/J39/iScnCMV8JrZpfSrP8lXJ/isfXH+zNsh/7blrFlBlbF4YK/uIQTC/1KG059RlisPdykyAStD5Nu/TI6cfGYICq5BoPGDltX3eT1BpkEw4X3fGHaBubCjMKaJ5vexYP2+tUfYfF3rhHIyekykHGZVtH6tWevVcfJUQYgAJzrzvTpLtgun+ncZak6mKyioClCi61R0ZHSnNHD3G+jPpyWelBifOfV/o0/hZx1XVSBnA2Fb/RkNBJy/7xvMttRnnqw98J1wjsjtfBpwDUIMmeC8KLtrXjJsmsTWPHUN2XID4d8FkWOAnd/NPstDCAUggaqDNiThACTBVO3rO1umH3q9joEgpZfNOywTfJxO9xhj7vecCfBmeL3y6YL0NVc/RXhUidUIoVpOFhUYDcaswBQcYuNgyU80UkWllqY1OWhsKoNqXfj+6nvLvnwtScmI7/8E9fO7WfHXGwx0nKQN36sK/6Sk+I9X24c/PK3w+szQ7ELH4RzjQ/kcEBoemdiFKjr069cvvF6/RNOcsonUZ2Dwh3UdhGNO2DpyZFq/U3S4DnLMPs/apP1nQpws+5oEYSwGad7vKYA9n4fL6lczjs8QPi79p7EmfsY8Hmh1yogK4aEg+GFUWsfcf9nVj33e5EshY6SUfas0GVR8ooAomFwTHwB+vS682gtcN7UDYyu8dmHLwSwbw12kJAykX2xa50HifOTEex4Ey7uJ1vjEItMQRHuFZ5KF9O/DgBhs3OvkPVrWMekVwrmnfD6VH/icWBHGagVJHF6JBvizOolKJB1sdKpAH5l8sxksOcCmqHek+gGsi30CurkwMOvDTODug+I5RT3RB7pxMX56Ry2ElwGTvX8m309oKto5Zyt9JYGsATAVntXXq6F5mtIa1w7b8sFc9d4SzgU3IYfLEeNQuE9Gvgv0d2Gb+2fHWJfk6lxXNFIX+qkoWlEoJ8VfrSbbJe85XV0hVuGrcLR+Vv82VMGtzzQRzBwDYBv3jMyvUxOeZfNnubVOMN1ousOdDbMDa4lh/qayeiw8hDi1D8wkgAk3DxE//YlDf1TImBhQJ7jWyfWvyMIhU/hyCPsG1L2LvlaXrVahq4zS5I/Rh1+w2sfSWdqOFUVGP/ql638YGOiAzhWjrD645Xc3CijrbK55/F6X2Wal34f1bReweai10T0RJS0woQmt1toqTx83A+Q2dIXgiw/GvrtjCehCOWPz0Z+UO9tZa93XhdGf/Vq2W+ugwhnyfTw7HQfz2XAjtkNslh8JpxM+N66H4obguHari6/7gGxjWGCet7PWGt1qV0vNZZ0TkZeVQj5IWa1PtSbLH2z5SBGsdpkUsOcXG6jd89EnDRcGsE570ElUJnMlNpb7sK5aAOZmXPnHg8ImuDAMWLtwmeFCw20dvRlmDAV2cX9m+Kkr6GJu8QewhIOBIl0WUd+WELYmDpAWcR6rn587BjDM5tKUzuM8D1b9S7N11fyAXxY0bKT1VvUliTzo7wL7Xls/hjX4qW4CCFjel3KNu1kMr+sF26RDfFoTd6fwhSIf5VOmYzqtgTFuvO+/efpeHXBAl0sOP8qMt/l/HIB/wLk4WtHpTSHu8cxLMZ/WhKLdgTH2zcnVWvIzHSe1nZlvHfNtS33VqamJtKbgcRMQ380i2hpMcG/9WEw3QiEtQ6qCKT4g5Ifpd0vpCZopS1NkqgX5NEGOwlEiz7S0mQFbdAQqeLqUQ8tAJQS04ZHouhbep/JB83HMYR3XDte65hxTvbVr/T4F3sQZJw30JxGdKpyA/Rdo4aik15iRSxVQYF+K44l+ajnHgxEDfiic9RLdekjR6nARuiWU3pc1iFrUB3HGDv/i5Bec3t+tokJVZ3mdmE+CnqDJfeO9LN4AsBOzClv9zG/93g/H8u8jKTr75IlsR3ilygAtsmt0Yf4I/1WYdNzY/bBK10icwEIaeTxBq325QJR1EYZAEmnUWSCKkO7pceRNPt/Ai6wq/U6CVLPhnuj+9B11JJXzffe1w2UMP5vI+0fV89nP+/byXJ0sDsr/3ctCRdTsMIA1iJ/4nCT0zuCffIYNKpqVrA3pj9k2pq3PIpzZcnh0dmle7AJ5RhA97gextvIqF4VG36ZJ61z9ZZ+Hogus83jq2LjFi2VpcCGKugz6k8/rMye6PynMNQcHelQU49+Vd5F9umfl5ArT56iv4GVYxWcFomPb0EyhC/lXFXwZMGgAos4+A2gd8fg3QiDlR5azTkBd/xuQqOJDXd8JBOs9Pa/w4Pje4PqST0hrDejDAzMAoH0doZWVcU+wpGqTHY6fGUajovwsfk6/2CY8tCGoAqiKif2pvqtZL33ZXYlK83rzmiKbtN458tY0zSdWZagBmt39MLThB1fTkmj/ajYYLU7Q3yHpQhBFePqcb55/NMauHKN/e2CL89da/vzJEFRlqXPDyeMYEu1lzhXpqPA/lR8LZx1w6sRsfqz/RPmJUJIpnxiL7epxmvn98Dej3fmz9CGsUPSvTrjTVCYptaH0UXgfxyRyAk5oQi3wSiCSSVQBjXb2A7SBSKTQ/mUlo54bZoLDKz/bO+71vFtanWfNw2UEm8oHFlHpmGOdwziHuM4V0VZP6aCcDyxanMalcPa6Xom/FEGdYKr2wQxr2UffYT+U1lpPFcAnyFzrWWudjWu8F07Lb2u+Hr/BWbBU2neh27X+qWzQ1D+cTJqfSxOEfCBX7SM1bs0a4eel8hl8mJHIGQjxvsV+R8LAEy3aDBqZtA3kvOIKexNzrzdCWdT84xjDDCSYresQMYZHZgcwcAsDCL/bvuRTab/oj77WqYOeuS8aiWRWFufJdxQZzHiyQ5sCzwDRfvZwWBcu1ykPlci/0PokHJXOu0IlPA0FEbSN5bNaT/S5JeuYA6VetwYE1Yj534S3AZVP6IvlR8eUVIKyLG0iBP4klxne7Z9fZeKrEDbbITbbN/mcSwNZaAFauaUHRsCNIUxQr0Sv41f4Rub0j1zrq3BW6DfGwNfXl2THGNoSzOyPVeY7Cfuy7FCLhyXItizTgPDZGntNaifD1YSPWLagIEQ9gQZIG8/I8TOBY7uXTl1YSKwx+KUHitRyCmW9v+WpnATWP8dS9vlZrHFVpMg6fK7iOdIkDFV8drLaoRDy358M2Y+unc+w73DyemqoDvzMXDHw8MUqwesBxtRsyWcSM7EEFSLEc0ZqP7LPk5nytAxz8jPIVNqfX79eZaKwCahGPDWDRjV1FW73v6ZlHmPl/vLZRANxYk/uwRrTFMxT4A+mkDXDavr2nywXg2A6bo1Unyyb5rSqL37Kn1VaniF9nmN9/wRjtf9YKaa7IqwWO6GIk+FAVmxmub9P+QDADy/PZSfqqQhVK9Aq1Q6etNiJoLXeushcU86UeVlDW+cQSeL9wyRFn7JSqASrE8R6xhhblJRHrWgiOMcxhvt9DLCM4ZuTKU5UDJUGPn7C7bzheI2j95TXqeNWpZjW8grD8rnOzAGQyfOpDVWxBa/EZnVAUcNuofVgajONW1RXIifP1zkIC5yNgvIKN09XoarCEbTx/ldjclSwRdCUlqGAsFIKHeLr8T9/E9aG0x1YXBkiTdJDHQk6jqhD1zI54Hg+IDMzd4KJSHSk9/h7RTtj7Miegst2KoFPwlmjk6012EXYtvebjHzfX9BrADNd4El5dDkF1uuYtvcOdd39XgEi3bVRTw88KUj9u1qEk1DzkuIxxswRpSKcytkYO2DdYxOQ2v8dmUTQZk1aEi6lkVMS5u5NXRaiENDAnJaIdM4ZbWYf9mNq8iqG/p7GJf1TBRxj/Fy+PX0PCKk/meDVsWW98mSzm73n1K8nvM2JWkJgOXMHOIXdsf2uf2fBHCCzKAOqNToJ0gbHe0f7lZVBDQKwDzb9v0oXWgwdj8LxCDTl+jPEEssNavWwzlys18SE6Ndn3+1kXVZEF4IeNpjXItvoYZ6fSu7P7iLxd5f9XLcBBck0XFf0635IlavKZ7kJB4vIOMFpvqo/+in75/8VWAsA79+/QRyuWlkF7WmSvY7Ig+SapJaqeU4d/skkP/dDIQowhux11M96ZfwdjrO/Kmx1DW89A7l857vT7EHGD1jH52uChfaP1vHXFcdJ2mY5XanpnGG2x35X34voBuIPe7tZsSndVWEAvrulRj+faLoL+fpm0SEJKJD2q554Q08/rFBYabvNhSAfDfrojXUnNEWMVpU5YbY/FvN/uhRKy48PlWYDkQ00MyDmebW/fp2rUg3BNDYlgv5+sqj6jMKRE+Z/sqIqD3qFRBxrsrdJ+GZAYZo5yQwgHTTg2tkg8MotmEZ1ySxY1o8nOZ4UxCmo5HPx8u1VUzgpUDGm2GGi1tP7u2dW1XmjUgYyLFO3pNLves0cUrl+oFptfqY/J/UFSbD9c5T2lEqofa/vRODMVt/zJcp7OwqPdbMHYwqRN5sDXloP1zT7NASG7xNrgB8kIfCndo6f+9GuZPb95D2dOK7z6CQ4dKMg2FoOqcwSVq8EY2SSTwKkRIo+BbQ9FVrFq0WSsta/4HZrvntC/j5pZ4WXqikXI2FCtAmjWCoUJTxdfZlK8npdcX8mgDGX5StCAAAgAElEQVTeC77VfvOcJ9X4laEXg5nN3T1k5uE7Qcrzu/CNlVtR57/SSOkQv3Opy8CMpiqgp7lVAa3KIilw+H7f3vNJHaeEkWg/t6cKk3RUwdWS0Fp3A8MjZ7++vv7+0ZhMeP/jjz8ng+SoHrDvFFGIU/+dBkkfxqwtjcJ6tgkYgf+ftFQV2NQ3ZGao7cxfkmUKAo8lIAF7sCyrtnlipBON/B8ialX6VGGRvtu4vMHneuTnPiW+Z2E5IwcKkd6epqfyY0Zez4IJABE9JfOqUiJvMOjCQ69iXjwRItynPbagNKrlpKA11qHPKJ3q5o7Zynq+Qtj87M53aq1DmAd+/8ZKhv/9+/ffF04zT2/688/XNrk6OCXYE0MiMauJEMTP1vKVeSbPKDQ/wddw1hu+vr5Wf/b84GyB924+C/rOEIR85+gxEPAxMfP8zyOtWObzsW9W/ahqnWy5Mwqn9Z1SOTucfj3BOzJ2JO/ni4xUSbJu7krxg7HvMp9N2oP3HaPwzznwlxXJwBg5j7mO08w3ld/93nh3KeHLUdKYG58r/xG+9t5xvV6oCaTk4RQ4k0QY8qHzeMDi9/v971hOPRtohzE6UP/sbJlik+4MQpitNWZPXvYrzd/zWvXWXriuDGd7f0+f7fng6hppVfgBOO8S2i2GY1SNjCMTVxnCzNYJ4O/3G9fL9/vF2OT4CWkjlmc8Q+QyB35mAzYVlR/6BN8E3aemxkBrl5+ubIb2uoBu6LxqHpGU3qytZAa1/M40OfGclmJF4Wf+fS3sf7U6/PmEQmJD/D1hYbhHvjIWOaZuALBg5n33mcf7+tiG9yF5Akfrah42xrj9eBUD0F4vF5zJFdYBawPXvC/GkciAnyjhrhf55n6/0We/Knrz6ZvK7L59G58Ysd+/v/D791/4/fuf+P37r3ma/3OQ85uAkEOPr69/bt+ooNJHAVziwtKoZg9Ckqj+HPw9G7jfHrzww578oGRdwB8Dyweo2m0LyFCwS4hcz6mhDxLgpPioMvucDPXb2tXQR59bPKmo9GqJ2CqkPxXS9XkRj99SPS1wi79bisI6o4weO3GWhuZ3B9iVUI4xmurXF/ZxZo4MgT+fN1zbUhThc5ndj/lWjHeRmj716+gunUrOHooYgfb9fr8xbhdQM0ODrcOvuV4LQVlLaU/+tOYnIwzkRJwKh5V/NMGABur379/466+/8PX1Bd7J+gk9fZOEcKMy/skyrY6RysiTpM/b0lhK+D5h61gDYmDA26tBg+eijPiuFxVtWU1IhEX6bn42wrdWP/T1elEmV3saMAua0UfLwRBlnqxoPOn+O9/1REftO3+vfjMK/ZxZMZUZ52AvdR1QmbBmNtX7dOrcVLjMvuXMow49MfC7wogohZP9vfvbraC4FEqfzfx6Tze/fYaqtnbT3Fi4WOu7SQcP/nwtf5N0+lR+lCH0neasnQymev4cKE51z8GJREApnIAKO3P9liajtn+E6aU/vfcUPd58MQBfnbdB5/RCwnVGqGv/INknGmgg7OMzJ9+30qNCvzoGfWYMstdex1OptIpnGQvImWMKp6vV1e9pTWiJ6tnHGkn9ruR2kfrSk7F43j+pdKMy1frvMYAR/FP5QZGMWlSM8L0ZndWx/W3LWRmYHa2E4bOn7+XJJJxPwlcntUIm9en0HR2opuBRkPgzj0mEzx9OzME6nhLDYXP5p05U70tL1rVZwn2tp9LZn81CdaLFJ0Y7/Q7gmABSi0YYSc8VTltCuis7re+UaaUCUq87UMupyRHnfu6orCoB/ntdL+DV0R+QS6Uv20mBRDPf3XNQtKtfXe5BtXzRkf8ca97Pym4vP058rwSqn52ELZ0zqpFAuTdR20pwAPvkRM7jZ3irAvEqwZ3TGt9Jg1UBr/DwZKVZr65hraUJiWwP8cuv1oC626b5SQ4MNp36rtZC+1bn5kmZKt1ZuKZZhXNlxMyxx3WJ+w4RtYqnBAfnKVrNDIejzR0haB18jEkiaoVJvzatl12A/fq13cJ+UniqJPRkSQ8NGWzs656VZ8zutdS0bgJvHqt4/XqBpziSRz/x8bcZQk60bOJPxYNiYY3cgkiUD0FYMw9v18nb/KPSphMrp9npM5Ec4czzer3WTVpkrtMRifV9asukcERAgOlXFE3I+siYMb547rTpOAdrbEZwoy0yna4XanIDfdzkX1sWMkUE48G3zALwDOHGOO9QqstWJ0XAfrriyksslY7fI7ZwD9if1ppfO3i9om/XtQI/6iOfEjH4ThLO4Zva/f188uCpxDg8hsKTKv/49WudhasC+lR+eBLCc8Kw/IGBscGmE7TRjdBPwllD//H8fuQFEKexEVYexzLK34d/qz6zdZh1XpYIy9XLGb0BM3Odek5Nsn5Tw686Zpt7AIxKKSfWKxNVK199PyCEnH08o53P7klFU6c26/zUcY8x1rmy1/XaFMxuSZ/KDg9pOflqILfP/nv8bWk8vXtE/VqrmznTTevRs5ZOCQ/rJr5rzxg7lR8J50m71N+d+ccKQfOzrA13oahtJdLX+s2TEczKXsaDdq/9njUe35ktrOAOkK9IMAu/NWrCssZ8JpRMBAQqQ9BqDACtdwxhSk7q7ktmJs9Rvt3qkBGypdMAS1hz7xdHFHT/BLeAPXqrCeEKb/nv6/3GmNCTy2PKEzrfn4VyrQskFKJCIOB4nXBRUVltK8Y7tvrGPVaaaZsXIJ/ok/aQjlg+1P2wnLt/WzjHFLZ1e8Yj08fz1DZkXHIBf62L2ZmBw9F/vV5J07Iv9/s9n8nMqJZXBaIOXseg1hmGZVn1O66Z9j7gB0xPurzfuOfuFm2Lye2tNfz5558Acj4lJ5FFBdPHkY/NVL+95m/evcPuN8bIUc73ZPycS3tD91RmZTtWH9nOn3/+iT/++GMpjGwt3dWpVoHPVvq/73s7ueGk5OsGaW73qnPKbvjPnEpKN4b1nozKCSnV7/ydeM+VcCg18iv1qD8b/OE+b6zxrugtYn/0v5VbGzDjszYjMflM0pryXNVaiTAHjZ4wf2ee5vtogauwL8tWoOAnuEQLojAzIrwdo+dopKeExUHGar1er9e8b6On/qv2JH24QZi5pGGFQ6GZxa6g9W7v4HkCSicyh8Lmeptype37/cZff/1G7/dicGZDKS/MmZkWZIABJJ1jjnPVX3JrYwxIa88Zhp+jmSe0dYKQwLyS4QMKCOOzf+7v8JQN7w+bZroh+TBiHWL9BeYrT1zXtXZyfQfbf7Rl7ARBD0M9PhsC+SwcLhM7Aauf8mSt699jjCVUp74e65vWUy0tk8kbDKNUY1BY9XBcZ+8paslSfcG3QEFecBvWW2m2++XK/IoEevd0tWE5QKV0JbwMy/52ZjVPT/z6+sJ1XSsHdAl+a8tKcG3xmpcXndYnk6W0OKhtKZHyT5XcUwl67j6kRs17v4G+uxeE+h48rHTNrorzU6C66JbpH1lGsFvolQJ437jvz3m1wA83W1diUGPod9U6EZJpOQnnEOirdyXqOwpbniD6yRqfJpnWavSDMhChXQJlBpu5oENuvrJp1j5pZrZXNTuKhud3J6vm7xC270qSzERLo+OmcPs+WkURPB09ln1+//6Nr6+vZYkiu2e/ZVsFkGNwyP1KAvL/b+/atiPJbWSQKY39/z/rdktJ7AMYRABklmq8e876oWlrpK7K5AXEJQCCpAZXdv4gIthPqtfxKazV+cMaaz5WZp3eQcV4356udwxUOdxvrd75s+8+0RkmnI7/nuc9t2UpdfSnpHfgb1jO0P75sKYKTU+YXSeBPpwzdrZuNTey+iaApEZ1T/aeqw7PFrS11MaTFdbPKxweM0m99wFmjrwSSPXNaO0W3CFWFdqdIPpi5pkzmwS7zEu+ntGFOfexBlo4RreYv3//xu/fv1ewSWnQWj6oTMsQGjuUPa8fnwQ0BHxX9LmvucQcORudjpsBws/3B/u8LOocoT8ZjGo5rWX3oY5x0UNQ0vdMtA9LfSdf03Nsn/norXVOXj9vSxv2DQY0YRqbz4+bgQj3W3ufW42AeQYN/8W2dt+pChJhlPYhT1q2oE+ROvVzalu1DhPLuaCkBnEO+pOM5wvSOdmea41qhRMjCErpwtwu2E4oLnQ7QrkXPfRZF4JMO+3HmidxHzhuMg//1uvxQrkavict7/tOQY+KWNQCLmFFWP9N0SMzfp0//5lQs+VD4FZ7DWvt3aROKgr297ouuQJybweYgonsTtR+kjZLGOW6SqWpbhf7+vqP93NOazPu5T+13tFtj/a1FkEYrIGdd9sT7tS2nqzXiRA1Myb/7LCiTjDavt5ax62lalMPGGHdvkyoxbr1RrJanddzXnzfNXIvcJZBMw3t58hp1JvHr4pojXX+Ji01w4fCma+xiyAZrXbvOemCCn0d9D2FWWndFv37Rp+Yg9fxAirpNttQniBaQ6u8YetZXdaofrGV/urnlafY37p8FL6vGykiFLWg3Ht8Km+tc2owAGNgEJ4iGPW6rmBOGVA9AW3q+kX4DVZM63AiwFOpxD8xpj4LhMOOGQ5XwXvdjuvhoabPv1wC6etiAJAPQQuFFfTV/ofWD2RQGW42hrSUcfUUUHMmocWfyz0y9hrBrksQauHrMZKqhCmct/qoU3i7jqnwBOeeS0cL2jbFIAF5/bVdwfoY5WqLRE8aCgbnDAN5GY/Wm3t41zycJr9AYFWiapWX1Zz00IBb/dEsq1N5w+c0nN43s+R7Hq+MXxqIObFtCeapUw7VgNH08CabeML9PueVJ/hpIva1v1j9i562qQs06uZwKBRGC0iDqCPadZAeqYqctLrLZlEmMSNp5XpvrP5fF/3H6AdPviO/t+kWNDreSSnlwS/vs2j+3js+Pj/975Z3/YO0bG2DrILz1kbmcY95inyHNboCs+U03TP6yX2ea878kLOs9HwGKGCp/zZW7trKzLmHK89xT8QXiondDsUMfFzXtnNpKy1/v/nQhbZEEVUYqfR8h8o3zMZy9U7lx/2cypYKmVT7ALH7PXaxM2Q8cF1hZV0YxposZzSfHt6tCNhc3A2fyZB9Pb1cNxENATHYP38vbsaCGe65053aE637oc5khNbBS0yswa9Qp2ChoV1tZZ+EZaTVcyG57xA2t9A5s4fj934NVzzz6sLWph8E/1kQrk0fyr2H6TPyBjEsGBXHv3RoJk0SsNZwzbU3TM/ZxsA3vvBl38AwWPO5aJggfcyGzfyAht79ZAfzmERvl1jKOCFCPMc192F9nP6t2zyxX92VOMk+gi1+woENm9u0GDSzqRTG3FytyMlPlbh6vbENKzBGPtqKhdeq6Xn1WgpGqLn8tCLGawmL/jvHse+w0vKD5SRT58XUo7+I0EgujISKfEarTboKYQkXLeZjsR9y+Xmi52aewKo7QZOiRCoMGYQ6dq8KDB2G6T8hT1RY3kZzMieWmSNUFLqUZAKHsPVHrXwzP7TLuuTyJgzAuoN8y4Jtidi2+uHzkRGH1rhgMwXsHmv70xhjHvNBgocFaq3ho39I8kSXoz9oWS06y/qj86sT1hrQba0Wk48cRTWPtAb1F1IYpK3SSZTH6oMo8JoE0lpskGKbQWPyyADE7w/eKjxrcV+oBudUDiJAN/eePqPad2DtLoxaKAyXRE5TWtx67ueWarDi5Fxn5UANPXvaCJB+amxCUYujK5YAIQiNw7hf+8EHxYUslPq9+oGQ57Jm3wNatR6AKCejBv6baXHM6dU2VtbREh7APj7wYZb8TVqsBYevD3zOXT+6g4MX2tb+bUEZodeTK4IpeKeoPLAnc5iFX4kWSnKLa2zNaBvZ6o/hlru3K80XlWt0LM9z3TzuELbNtWA5ZubYIy9vr3PKv4pFsU1oTovic2Rbfcp4KoDv9qsycQ06VBlb1gxtZdYok5oZ+gx6PZU0QTKmRh+wMTAwx2W0CmfFketoh+NUor0q5P6OnxzncYD9QGv+ewV6TsLJti5DH56pdPTr5yh683S0z8/PxYT3GMD4BopA1bEmOCsGTkvjlxI80vcNmFfahyAlxWnYhPM0fzuNpQI+QwNhlk/HmIOIdrKCzYfJ3ejd8PHBHUYNrzeMvSGcqsWVcBqZUqbwQe47U5D+fRAYFOIe+hHQFgmeAJ4EPmRP36o/xxYWLKqyp+H0E2O9tphKl/SW93dAItwPfg2yRXuynNEOhbPDd+nk76q1iPsyI3i3W5OAw701P2UOSNsA67prRQJouqShEdfzkhhWP/exVVrVpQq+NMYznSi02TXSz3gtgkalw89ffAe3dMrH8RPJMFoHr2CIIFHsO11uwwsj8NaJ7zq4k6ViY+oYZ0s6Vsj6AS0+Mr8qgtZ4hblOalu7X9iPy8/V3CHDgiLTena/LYtXj2cFs/fPin+XGKIRUmcNral7JxrsQrT/W/uUv29o3dAkb1l3rtQ6uVxSEw7WHFlmvu5rTGtfq87JvrzjxZclLlwfH4lZ1f+uSKuihVr0udTe/LyzH61J/3fLmPKm5TPNO44EBUZk4cGgMU9KFDkIwzXNf5orJrvH8akewZ3BodKfU3lLOH9/fSXcwc5Vway+DgcHeBKDqUVLTLxrvsqchIT0j4j/CWu2916Ma0EUNPTrI/lIrfkpDXxO3/Gf8zKQ+5YHxTP72kRZ0JqwvarwlkVBtjoni9hbgy1GCQRwEmrOh8ucM1AeG+a2ubbNp/ZVFUFFVo5o+opBVCvIM1/r53VcSnd1FZagC92xxuYKoLo4mYf2ZJcnZEKLvpbaRuTuGkL5UunWTQkO70Nhfn7+BWaZfX19LVo8lR+SEHxX/rg9SlonnH9X5kEhWCbCDikTvi9Eir9diK/rYz2vFkD3Kv5UPDyfj7tUIeApCPRpsp9IcHwQUBGqtYOjRYZP1rYUzlPQB+ngsIpaTpbV6z5nTp1pi9UHUwJzrbEIl9bLceSx5KBdCHaGrMoTSrc6d5mvoq9CtO3v6EvN8WWygwhcmtcdVi/EBp8LT/Hzf98zZ9eXlmIr4OdcL65tR0TbBbQ1rOWU/1g4V+dpuWWiqxBslk4gg2q/+Lpq7ejkWnvEHsGNzKQIBLji4LGZMsHsm/SzbphVpmN/YpM4T8HL15RHbmww26pDhIrw8rQ+rH/vt69lX79aTaWb06RBM4oo0Cf4yi6mz4uyxao3jiGtFrJ+ZvCT7L2vQZPa5/B3z6dSLKEoVi8xMYV28oPCbgOA0UBfPPpK3ov1aPIkkHOdK41AxdfiWpIxj1k5Fd1ax3XgsKSX7Pix/zy3dmWs/GB+aSWXZu2eTsaF2GF5354PfvdZa4RRtymRyPxNLUktqMJgMnnaRxtEAPBbxETzhwWLAAgZU7f6KC0+Pj9DSIol0L48Qdcavl9jEzq01taRHk9LLP55MEOlb3UNFKksAV7MHdbvujxqexVkEWhllDnpU+ii/V2o989HnSugjDNoFEolt6nzYIizZ72tM1xO/Gfm+3aNdLnBfbqeyRRR76v3ma10x2czzmEWSyyttQXvQzEBn5+f+Mc//jnp/G88lR9utt6vLCgkdAIVS6pXILCDJw1YfZUn67L7Dvv7lQE32OR/pOfot1Jz+t8aeBHN33jHSRwzQaHNG6qzv1LzJytzq0Iy6WsVQl8/NLQSnImxT0ZjYGR9TmHndjwAKAnwbBOAzUQI9ukqUXENnuxldy1OEcnW+jqlwcy2repWzViZTzTPJ+6TxhqtrT9PyGV7Rui4cqRnN5o1oHkm0slIrTpJl6JAqIQ9Ad4h8l9/fWLFUB7KD8L5bHJPQlZhL61AL8+fBvdknRU6nr6bfy0r55N1CgZki4uWoaz+rhAr99Hf4ToWx1qvueOPfnaiUR3/KZjjTMQ7UHPEMZ4FjkqptRmsyDROc+QPT26K7VfLV7pkAf4wVydYWmmb6YK1I8WV4X4ANaEr5hNhPef7AuF5BCWgf0e/qnCelEt2bXwdUncOYdzgcRi7Ep8IYF2OvGpaz+iyCi/p0rXQU3nrOoZXlqwyUmbQHAypDKRWROt4OnBX92Zmwdn7rn2u0MrgEU7CutO7ZKL12mTMPiGOC9K1xlMhH5VA9bsUHrJvVD76jNIncnc9mbwKWaVlEpTe09qkPj+MG9o5bg9+cFmA561qgKmZTdrluZw1TJrloykr/K3zp8IUv0Wpzv+wzdaudbwkkYpGn1s3tHtOniE9p9b9NP+LhtOosAxBJCc3a769BDDgeK6bqXx+0PSOBrW8dUzJkybkd9RGVdhUi5FJdV1pBVe84seOJt8IeAkFVn9bwMNdOGcKptR/Eo465t16n9GAMmRd2qjrjCGcUZ8yzxOkZzunvi6L2SNvth3q6K2t5bkFp5egiuU8rGlmBsyQs8kzp34pM1caxjvPY9dLnGej6Nw11xr68FxYs1ZyYv3ZSleF6oomVl9bcx+zZXrXv9V/ZXv8PDZajxQofFV+zBCqjKufa+dRrGw1ZzoxZN56+FW9XbhaZmWOOtFZKXS0g+CEkJ3FSoVQo54negD5vJonv1LT52hNT/WScdK4F1Mx0PIcsdX3Ty5H7Zs2vxQIWmyQBjarXzOcqtKsdMzje+rHea3RRElUA6AlrwjsdcNsHtR97p+OQYUmCb/ZRLQZjZ2QCvvD6yD0OkgPkAav/1TevsgoY+uc2M5UrzoBSSA5yB8m5Qk2V3/hpHlCICLqhgIBPXqMpDzOzP4c4dO66lhZdJtQ6CsGbna/TAWgjpnj6SmwkxnMaXH2OXd4vAdMTsJJWJzp0Lb2lfaVTpU2T889KpQyzxWhnRRB7deTgj3B00BaQU+DLw/9ZOkWf90D1+cHPj//wj//+c91TpNC6nf6/kNA6HRd+blUrcq/F5QYI51Xc4r67e9m4eSAfpqYqjmHMOl937ByO3ZlGLbXe4/LZdkmwhuq9Fn1SV1t7bnSsUQtpzGu+uTvSJh4eFYgW/48w7ix/t6f663B/FIQyWry/xmY6L3T2Mz8rB75rtad+myWcnYN2AVl0UAzb5x2qnhf8aeitLQOfSiakBJLZ56AY2bon8U1esl7Pq7Pzw/84x//QO8dv379wumkwVcC+tYxJTzJwIz7F7mTo/tdF9cFm2eh7NYmfi/NIlBEO6vW42QZnoI3mTCn9b0qsAMmEV6tq/cLmsTsG4mx0v2uPjdkj0F23eiFMYB+ITY4A9TCComeJzj8FfX5nsZOmFp9MT33Z/VPFNuGKGbUpU9IbYhN6q1hXcLTe4e1KejN74yx5r6CwdapiPw36HvBd2Q4PblMR8FXZo315j4tl7fp7aERZir/AJDlMTNzv/NrnkIgvDaMB4L70tTVZ+J6c/98UHnAYB0wy/SEpOUtBVCssafrfSajoMfF/GQ1gTdya31ATN61WPvsTODmpmDCgNcNnqAyhYLJC1XbsxC/19A4Ayr1QDGfPkw47d973ZgC5Eqni8VrDCDMta52+dgphh/Tmo7JGLHkHMX7BPg1CZ74wG1jP0FoRgUb2jFf8/Se0zLmTJ8fY6yE9wq99f21nWyu317X5RuNuYmh+4kKV2/oTOWcMB2Lfk7nqfrmvyMBBYuSukGdnQn+gIUv2WzAGlabzY9fmMoDa214Jbl0P6rElQyA+8ZoAGxMmvr83VPR82QEWm/zVLig0dXXFfX3feOe/mTvnynIqXy7eBXA19cXfv365Qd7fX+Dmyx0rp7KW8LJo/fXzgWnIsw8CuVWbT9oqZaT1VDNN6d6q6MOOvyCHIlj2hhaW0daUKtlraF1U6sjPUPgqX2C/K3Q8AmiLG3eskbXxyrEIRS3yfyvJjBHLIPRnzSzfn5akqpRyzSm4ceEDMR7r9pRZKRzzRQ/DeJUH9Pkmgf9STwBBsqU7m1NENsANCqdUVriQ7Zd51ABdsuRaEUedAPI3/d949+/fuF3a/j16xe+v7+SQaICfFXehLXn0L6Z4evrawti/FxnB/tVLWit40nbn+Awy3VdGK3F7cljF6TT2GpZ2rRAbLNsL5981g1SSjsVPehntf0TTStyiPOecj/rvLnuqrRuqGcM8buTxa2MWcdd4wl8XhXNaV75LJVaHeeZD5Qq4ctXXtXfWqrQ1xLfOZ1O9MjPd4zhUPpf//oXxhj4/fs3/v3vr6NC+1/5nNsaoTALE8Dv+05XznHh+1WdNXjxrnDq99SOrwR1jDh3KPJ/22JS1lOL2Vxwl5/0rFpleaf2UWkYwpLH8iTAWu+J0fL8xJGbVSgXgjgE4ALJhI/LS4frs60139T+mL63j6n24WncO92c0RmgqbSJ958DM6cllkrPVwoy85Ovm+50pyKOYJAjynwsJu/U4RIildir8has1VIFo24T8pD8TihW4wxQtfnPznGtrxJQNZofc8+zVF04NXd1B6u5bm3jSTjNZnDM9rU3pdmu3F4HrpKmflAep/XGkyUmM58SRE5Kl9YtlFhWxgzKjRfj5b9rNtKTcji9r308QdpaeLqgtxtQ8WnsJ3TA/p6Wx1S58jqOnc638NwN4E7tU/FFnu295V3X8tYZQrSQ1f8ZMmj1VQj7dosb755gHn9XYurnp3f4GYmo1wfojofW2jwbyH/YH2WeanG0/dyXfBJgfeakYZVmWcgflpTMEo0ZWTwJld6XQq3MsfhNWudot9ahbStT61zVeT2VWofS9CTI9TOb6LQq3UpjWile+cGlCj532jJXUdY6ObD349JYRWR7X2f92JWu8hQRgN6v+pNReu86BnOfTTXimAEC7eSaZMdIW31BoJx7WX+/Ek7V6E/CM3sv/8XLd6h5F/Px+4d3lElPbHqy5gBmYMprjsm78fW1556extUm1y6Ima72o8+YLQHrqDdamcVdIdqOCiLr10V5HATsaa4q7U6CqN/x3aaBtBd8RD6kBfII+esdUPwdpxN8plWAajl79x00eitdpdn8x3FMSbhbPkfqxOda3krf0/2E/De/U00NPIeHaW2YyaIOf6nQJeIAABYQSURBVM040rpPxFDreSKIH2p9g9eDV6sQ/WEdUv9ULNH2M20W0UvkURezCaPjfB5W6NC6ugVeZ9++C+ZyC8qREB61BlxX2X1hBkzGVatgZmnfZ2t7NLNCUT+B/qw86nywPT1as6YxniCdGany2jLbVFJEDOpe1T29bF9pzM/pX7M/Oi86v/7JjqxqZFstcB03gHXf0MmHPpW3LGf1cRgCrpW/0oz6PN+JpZqcWH3qeIUWe5obMzsaenfnG0A6zrBCFm2TRS10M/ih0zj7KaPNVgRG1v7H5OQ1RV6tcNrvWPtYUYUqM0IzlEitZrfsAk5/lvC+raUArZ/PatCt0vGpqKD33peF8vr2gIvrRSqjQFbqc+YAD1HPncZX+YsCuy3FKA/NH5N61XJzHb21zEcq6GudWBBDpVfvPQVOdW5P5a2llGoMyRBkAlrEJ+e9WgPVQMA8UKtougxzzhGyulYXAieCBGpinTRO5C501aJWRs0aeCqTAiNZVl3StwWT5g1edbzaVq1nCfZBwXlfgnnv+zsxs9Kw0kwZW601Cz9Ty86vT2Ou7RBCEjYuuiWBUkVigKS66UkUyvhV8TDopyhOFVgV8MVjM4mA12sojwQddiirCv6+b/z+/YV7nv7RDrzKc36Vrq8U3dsBoWft7j4kj0KsE66F/kRobh+0z9VrzP4qMKD18z2Fa0/YNCuZPeoM5N0DrwToqahwL2h3aPepLFqZ+nB7u236aT5cm0z6vJOnWi0mkRyjuMBK6jhZ6Trevf5IVvmpLIaGWnLHuyc/VgUQkjLwVGoSDCE4U/yCV/bA0tPul4jSxs1hHXlXT3Zzdh/4VP7WtfMVFoQQdHx+fGzP63Mx4Lxw7I+NjfAVIr4jnFpW32az+jj/1tMBQjh7aiuP83XbJ0LXZ9WCXh8faPeN7+97o296lkpGrAzzfzX4w6YYrNPr0+t4wuICVqB7HSe9sJifPYin41Ulx/bog/JMold0rHXxwOdTbCLzSX7/NB+Vr3jQ8/f39/rOrbwouiTMSOOvqKa6ZosGhwSMVxF04M2A0M/WkNodG0NkoeJCrhJYNFIhHLCfLVQZp2pBfSYsJ8rEZUiYrcue2uYCG+PV36dSLY/+nXwRaTNbQwNPHF/jsEiadkJD/CS+f16bzfCxzifAxfHTOPSairDZNoW6Mmm8S3qpAJgZPiDK5iX99piAM/95Tdf7EfTc+RSJnux34j9Allbi1rY19taglxdV60ca1HaG+RUf1tXq/y8tJ8vJHOeBEw6co1gAliZiUYZhHU2+02dOVmy1fehLCsIcUtNJwCQoRwVEJokJSZrSRsoFPtGm+qJV0+bvfEReDwMjbWPEJBTDr+5Ri6Y05m9dAqtIgMqnKh6FuBXCncajY/EzeLJPxr/palRLnfsc9edN7Zlmml9LGpAdlRpxLlCmY3Jnpm/sPyVIOTuVYXTw27BYmjoZjtEirdHnI8b3VN6K1vK+xJPVpLbwwZ+zMUKrnifEn8vW7RVEXNcv8LNSZ7Kcj2O3bXJ2CLtbhM1aPFV/GAMjx621uHPSdjgbtDj7iNp+hnrh79UMH4b4VUBrHyvqUORTgzFKi9z3PPe6zKEDrDQ9CaiO8/TjfaufBzIz8X913AFpS5zhwQjoeANJqeL2n+95VCoPOK/0IC16v47zXsuPlnMz3Ym4zzBq/S4Q4qluhQJPSoDfJ6EozNDElwp4c2q3pWUWANMvOOe26m/W4fv+5vYoO0DqgwY1TNj8QmMqTZJgtgh4qHByaUafrcGXk3Bv43+w/iz5HtJMmxO9qo/4SqhP5YQwTvymSgpzRmiZ1XpXRJD56jlFsPa3NaT8cdbzcV0YM45QA2vq4yOlujyXHwNCrU1gKLCORPAIdGxAzQOI9bN4rzI9XnTSIx5VCM2w7blje1qby99qPS398xMr623UhcMM1/z3bfMG4zXGrBWfbKdShNoU4BEqz8y6hJJ/S7/1fGCFfa6UZGsUYgmMzKDQdvVxjNT7up6tfQtoqssYuQ9GjTj/ZuJDVVw7s5+tZqInjUF5rvbD/5jPFkFe0djW1sl6YYHVYuZlm5+UCWn78fHhdPJP05jdNdxPY3hlPH88GpOV35WR5+TfN4V4NYcwmKHZeGVAHrAMutX1syFfs05fW7QUFCjWRfrNS34AzFTD5ptxF/QQn2UaWB9XQ//48M3GXze+72+YaRSv8YrW5+C9RRDB+8pwvI9lJUGbH+3vTIPYZN18UzPM79seDWmXDGzMzdtsbn43DNbnWJuvf/KQZIW60c0d5up6NYWM/p5+Psa9HdLWmuc2ozVc/cJ1zX5AgzuBZq7rY14bn0gnvBOBmLR+3TTYmGGyuk+RqSSKlUoEGb73iZwAWxH0J0Uy7hv3rIb98MOi/zrSxEc9ed7Mlw/bI/cAeFM4K6z0NDFmCcUPCXOCkhzjaWHcjGt/PT2jz0X0DLNum5PA+iMQxCTo2gfaM8M8mYCCOTvYWvOd792PXxnthn0hdmE0REjcdp/hpPVP0CspuAZcHeAedreaLnzjLpCY2l/qDZ3I0JdGDMMv4jta31PihJYxxmTUEGAK7H3faweQJt4z0b7Lwd28O5UKmq5Fa/ecA+EZUZQch167kIUpw/Es2AGJq3V8mrPWgPsOhaTPqHLTnTCKDKh0n3addKLQMargbOWtaG1MqBNrzMpfYfTqRC//aEILJVI414St87cIpxLgtL0HqHeZ7AKO+S7MLVG0Wf21a8LPWDxWAXkllPW7MXJEV3+7sMGPwWhzvFUYX6yDndrjZ5W2FYJy7GqBlOHZtjOYZh45w31/f68fXd/T9s0GGNgls7vwfq+jUwJ55f6oZVOEpPWHDxeFLs8wptKdaVbpRoFzBbCf8Kc86HVg0nVPp1xGwgx237nN99xtAH9ryxijjH1dy64dVuHVjvLv9dtsm4wgTmRiqBBr0X/XnRcGrGv7mAdp5tauKeEmVNS+af2aQE6GVeazMoafSgPSIjTboc86xn5jsrZTgxn63Bo/3PoYnqEYIaJ/lvuRn4tA0337pbGwrDC4cE/hNAuLHcntcQucWiIWbuiO5YV6+n2l7ynRwD+3zU/EdH/2xIWdLvwuK62jAFuciaT+OMdGpQM4GryFV2jU3+WbtzKE0sAsljD2AUQoG8gCqsyk764BmuF75pvqtehPcOTUT+1bOOSY8DUinjbmwVGqHQH0wwSqNaljetmX8vxpogHAxp20a20LkMOuXtKhrSSFzdeRd8bARn9th4ro9MNnaD3XMTBEo60KeBydcrJCChvZ7s5PQQsqszpO96uzn+ff7/s59b1aEu163PFDt8fMgGGwnt9Zy0UA+ogT3U2+b63N3UTvCSbwN65jUKLp4Cp8DYLGelOqo0WkrDUuzPpE/v79O1kIhVgc5CmxvvYJKH6krGG5dcUitsK8ulAcTFc1bPT9Fe3c+M9JS/TDpJP7lrqcoW1QEE7roZvv1PsyNCdfTAVbhaIybVjMW/525qdA7kskgJ7JWmH5M4yMBHWeTqjwcn/+3voLALf0N9N/p2v1+8+l5aQZ1T7kb6HnqkcF1QIlGnKQcrXygwX9W8KpmrZqwvr3yYqu70td9/297pIA9swUJ7xb07/++utxcBt8Rqwprv2L82gR3drj7xnuGxvBhz1rWm2z/q19NzPcMoFUHIDL09XbduaSlYnWfimaCE3/eo2S9DoJ7GlMlTbjIACneglHCTd3aJpT8pYSnBYpfHRbvr/u2dTzdwI+38cATh2TWuzFX/k/AMh/+WS8qqB9eDruDm54zzA/jBEmLTds/lDeEk7uQWPD+rtq9CqgnJzYrd4n0aO/HhH8XlG+2gf/ceLzoF6Wk4BWaOeyarOtex1SVZVMCBcjx9SUe50nJHESzvUzzr5Pg1tyE0hLBu/9IcJ7Qg+mkF7qF0HUftX6gLzDQ7/b4XXO+1V6cI6f5kXzbKtAc3N8GwPWOz4/G/wC3BBMTR2MyOh+yfIrxaP9UTop72ZjBFQ+oOLQNqhzwhDtF1fVeXlV3l5KmRj1cWCnsHwlimriiJQebngqkw1EtI+H+mJGODnEatWVoVoL34fBFxlOfn5acsN1hCB8xw/3mpNQktYrfWYFCUovISuaXIWi9w/0fk367D64lhV4kLlT2Ki77/OcZAF+MYXSv13Yd8GI/bu5rz+d1eTBPDNLOd2Bsu7kJtAtUovI9UrtE8BT+3dlqoLJvmf0lvmUPJzmUcZcZYPR/l0cXfk/lbePKemJgDEo75gnTj+F1AGk5YgKa/V7DoZXQHBEbGvcfgJ4wKcMnSszsP/3PRZTtdb9NHDs/QWtrESk13dJwMY6aYDRXWWIqDPW5+ou+d79uBHeT8L+Kk0+P7Pm1d+rjt4xeGhx7wuNfM0Am56T84QUWgumU1rqnlBVwLpBXi1hWN89MMW/NVtG6YHWgl0tEt51AwAFxR4sUu999Y3R4mieyTJN/t3W+UqkE9fbo84Gs7zh22niiTjk84+PnujB/qzxD1tLkEQf9/0fCucpSqnEJKFPmfrMcdU6yJw1J5FtLfhMDYY9CEOrS9igQhPP5GAHJyKFyRGCk9/LNMhavTI3Jyr6p+2eYa78BtKVC0qvYAAgfLhcyLSrjjUHQVMd2yuEwhLBGHVNVouL5rxPhb5ftSKZpvkomqd9jHWIoWyxxnNd/QAf98CMfr+DmNy/NK/ItA/6ZEusvPgU9Noh8lhZY44gYq/qqbwtnMB+TMQpMLE6hjwpkwrpxD5tp8li/wkev/Ij1DnvvaeAS61D32rlMlRMptOxaVCmWgJOolo6vluzQ6qS8flxtc4Du/IYI3UujXsplixgvV+ADWHoGIMe8wG41ett911p8WYzB/o73aoA5n6c72TJx5TYPGjZNh5b8wlVJhmpAVPRi0Xnbx1/a1xr3Y8eVSuvLhC34NXnT+87f7iSToIOZb8WI2qEvjng91TeEk5lUJ3gn0rg92lxp7+gjMolk9OyyQ6fWxLgk1JQiFSF1J9twLpGLvqpfa71cQKYolYZkAKkB0mdlIx+d2LgE2NX6NmAhUqWBZt128in3jFooUyollaFTP2pCherkqXSqgpL6wzh7ev6eioJ70dPTE2FVH32EDQ9v8hgdyx5cfy6JhtjU3ic3ZTTPNS2lfZaR7xDegF+sFz1wVmv/92uXSE+lbeEUxedVSupsGYIodo0OlGDFKrVr4+PdcBShYUMaJysmbZd1wOz5RJrMPCYSXOCsV6vJm3vFqNaDmeW8OG8bqR/exvTd23hmycYuyyldJRMJf1uh6hy7xn9uMXcD2LTgIpGResyTk2bHGNIxkyuN/oQQstADRAblBWC5uCOLZp7nWO9ry5FRRs6d0n5FYWmaEXnk6mDOboefE0fU9sJpXBAdSLIvc3rBiUJ/4QEWd7OEKoalQRcVwKuCTlvD1KtdoIJvTXc5TNaQGUofqf9qn18HrAzvV/hx09+HjvhXPSrMmCG92QU3kgdYwo/kXXfd+yrDAbBgj9d6GAHa/XO9eVqXTQCCvAQsNiErckHCrky/GxZgMp17Uo/hdhBy3lm7FS0fmVGVfKZrxYaQfYHc78y76lyuVQ4C7pRITeT+0kt6g0h1fOaVJHswS8dDwD4lMaGjP8T4SRzaIW0ijz+YctKEQYchaFOwqUwKQ0IgGFsQYAnvH4a7NKOvc/bm0c6Pf3p3Qrt+FsPBqPflAVuv/FLx559HbceBj1c29+rNGUGE4B1OHFlYGXYo1Uo48oX/JLiNZqLIhBtWQlVRLSeFX1UxKR9Cqaf7TlTFXozGeFcD/vHEwhOwsl6GFmv/arjDZSjPqTyZ7asOqZTeTIe/7FwagVdmCrDF6SOUjAXIjfVdOFnKHOfQuIsuiamd36cwtWEvCe4SovXu+/xHPfrNK5XvlcVMGV8XSRnPa+0KozrXwaz2CfZG2BJEe5CU6HY6pc/WIRzX4NTV4ToQG/HplXwKfW5zhfxhrvASOp1feC+vxOy4HytMR9orMq4GoKPj2t7V/1XtuG8FRsVqjBcV1/J9nouEVBWJuS/MfmVZgY0iq4/EMeXlDmetdkw3ONGvyRf+9nl/DkJQTujUbEa6UoCIUSh8Drm1y1M2ULOUM3+ebGmVfsr42ZtvAd6+iWW0z88CqdqW63nZJVOfczvR5CDyxSJrmjAmlDfQoaJNi4zfHxc6NflcG6hkDEVoOej3iUaKWYw+XveXnwfdALIfYS+sb6ZTyYM/yvo8fHx4TGB64KSUoWjWqaao1v/3VoLpDPHXWMKuoZceaEqbg1u8re6WWpJrwfjsYJxwMzP9t8MyuVLpoz/X3nDY86VjdiXO15I54+wtqZa1bsoKiPru4loyL6iL77q93Md6MGS1iDQc2TzLFTsw1PZx7QroCdhVYVQf7sWj+DP97ciB2DOeKLP931jfH/jsjGvPR/A8ImlIC6/qMlmcGSkojQKa69zUwNWWam0loWIhbnObtU+lnA2af/pMDH2KSLKfXuHzzERntvSdC79OQNzWjO8jWe5dq7tqCCqW0Bo3Luleb1v3znUW8f16dfQU2l0m/GB5t5kIEZBRmbA8A193zB83wOwOCf3qby9lFIF4cSof6dosEQh0rulCk22aM+4339jWU1Na8s+yJ4Lq+2ZeeZNtfD8O8PQgiyiRjTs68lX14QCpBuuqsLRvlUGPqXsEaLOk0u2ell6n8Gm1tDu0wHVwB4k83ergKiSp2VWOApkJVAFTdcSo58UqO9E/0qDU1LEeS6CPlpfPJt5f9X1wLi1LVpQ3dqmfHIqP/qc1YGvAlqtVS0xwIB3J+JUCHEarGpN1eYnRXHybfgdd8gr5AliAjwBXevk32qNPj8/j5FN7cdYFiqPI/4dTK2f9V4FfGc8HZs+syzPYd2R76D0S/utQt3HwN3udWLBGPtZvQ5BA5ksSypzxaSDk8syhudM+1GWku5Y6Kpz6jKRkzQqajhdYFRpWcdf6TrGWMrx6vmOzTqe9f5Eafzc84Ft+acVdT6VH8+tva4rBTn+roVk4SJ+JnAO4pzC90qwKnRV0+o6qLd51pAGDwpVooYF32Fs1cTaby1Vidk9lpXWPkb/q3vA0+Wz4si+VjCjQkGF0lU4T4JxKirY9TLZ9X551cRPHmNswskcXwpozKXvEtLTFFwpRQAFsjwS74TvPsY4I4TuwSkGaCqPKaxNULrnky/MXKiuy6/P2IUzaECy9AfeuVpHm++rEn0qL4WzJkw/WT0OXidYi0LG0xoVPz8pAWXm3nuaYBbtW7oIV2EsxBIBQI9bp2p2kv79BIkqTapgqLCYnX3U1vpkzi/ETW20pjnypwyjwsaEAb3S/Ik+p7lReqvQ18/yezudNFFF331qw3/Eci7oPMdrur0wL3/YchP2kzKWIHQK0Z5EU5MsItCZBU/7rKdpcB7NDNflyuRkiXXuYIB1ZnJl6/5U2iv8/af8KX/K/195tql/yp/yp/y/lj/C+af8Kf+l5Y9w/il/yn9p+SOcf8qf8l9a/gjnn/Kn/JeWP8L5p/wp/6XlfwDkXXYbe0qxXQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.random.seed = 100\n",
    "%matplotlib inline\n",
    "def random_test_image():\n",
    "    \"\"\"Pick a random test image from the test directory\"\"\"\n",
    "    c = np.random.choice(cat_df['category'])\n",
    "    root = os.path.join(testdir,c)\n",
    "    img_path = os.path.join(root,np.random.choice(os.listdir(root)))\n",
    "    return img_path\n",
    "\n",
    "p=random_test_image()\n",
    "_ = imshow_tensor(process_image(p))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([3, 224, 224])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Negative\n",
      "[29.078594    0.05153784]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1f957c32fd0>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img, top_p, top_classes, real_class = predict(random_test_image(), model)\n",
    "img.shape\n",
    "ax,image=imshow_tensor(img)\n",
    "print(real_class)\n",
    "print(top_p)\n",
    "plt.imshow(image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[78.125]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testiter = iter(dataloaders['test'])\n",
    "# Get a batch of testing images and labels\n",
    "features, targets = next(testiter)\n",
    "\n",
    "if train_on_gpu:\n",
    "    accuracy(model(features.to('cuda')), targets,topk=[1])\n",
    "else:\n",
    "    accuracy(model(features), targets,topk=[1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>top1</th>\n",
       "      <th>loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Negative</td>\n",
       "      <td>83.8</td>\n",
       "      <td>0.677335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Positive</td>\n",
       "      <td>91.0</td>\n",
       "      <td>0.474466</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      class  top1      loss\n",
       "0  Negative  83.8  0.677335\n",
       "1  Positive  91.0  0.474466"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "results = evaluate(model, dataloaders['test'], criterion)\n",
    "results.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# test results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[838., 162.],\n",
      "        [ 90., 910.]])\n"
     ]
    }
   ],
   "source": [
    "nb_classes = 2\n",
    "confusion_matrix = torch.zeros(nb_classes, nb_classes)\n",
    "with torch.no_grad():\n",
    "    for i, (inputs, classes) in enumerate(dataloaders['test']):\n",
    "        inputs = inputs.to(device)\n",
    "        classes = classes.to(device)\n",
    "        outputs = model(inputs)\n",
    "        _, preds = torch.max(outputs, 1)\n",
    "        for t, p in zip(classes.view(-1), preds.view(-1)):\n",
    "                confusion_matrix[t.long(), p.long()] += 1\n",
    "\n",
    "print(confusion_matrix)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.8783783783783783, 0.874)"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TP=910\n",
    "TN=838\n",
    "FN=90\n",
    "FP=162\n",
    "recall=TP/(TP+FN)\n",
    "precision=TP/(TP+FP)\n",
    "f1_score=2*recall*precision/(precision+recall)\n",
    "acc=(TP+TN)/(TP+TN+FN+FP)\n",
    "f1_score,acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x1f957c32af0>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 6.800000000000011, 'images')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(10.050000000000004, 0.5, 'Accuracy (%)')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Top 1 Accuracy vs Number of Training Images')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "(-5.0, 105.0)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "results = results.merge(cat_df, left_on='class', right_on='category').\\\n",
    "drop(columns=['category'])\n",
    "# Plot using seaborn\n",
    "sns.lmplot(\n",
    "    y='top1', x='n_train', data=results, height=6)\n",
    "plt.xlabel('images')\n",
    "plt.ylabel('Accuracy (%)')\n",
    "plt.title('Top 1 Accuracy vs Number of Training Images')\n",
    "plt.ylim(-5, 105)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}