deceive_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torchvision import datasets, transforms\n",
    "from torch.utils.data import DataLoader\n",
    "import matplotlib.pyplot as plt\n",
    "from torchvision.datasets import MNIST\n",
    "import os\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 1: Deceive the original model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load MNIST dataset\n",
    "transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (0.5,))])\n",
    "\n",
    "# Load the validation dataset\n",
    "val_dataset = datasets.MNIST(root='./data', train=False, download=True, transform=transform)\n",
    "val_loader = DataLoader(val_dataset, batch_size=64, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 327,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the model params\n",
    "model_dict = torch.load('./models/mnist_cnn.pt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 328,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Define the CNN model\n",
    "class Net(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        # First conv layer\n",
    "        self.conv1 = nn.Conv2d(1, 32, kernel_size=3, stride=1, padding=1)\n",
    "        # Second conv layer\n",
    "        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1)\n",
    "        # Max pooling\n",
    "        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)\n",
    "        # Flatten layer\n",
    "        self.flatten = nn.Flatten()\n",
    "        # Linear layers\n",
    "        self.fc1 = nn.Linear(64 * 7 * 7, 128)\n",
    "        self.fc2 = nn.Linear(128, 64)\n",
    "        self.fc3 = nn.Linear(64, 10)\n",
    "        # Dropout\n",
    "        self.dropout = nn.Dropout(0.3)\n",
    "        # ReLU\n",
    "        self.relu = nn.ReLU()\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.pool(self.relu(self.conv1(x))) # Convolution 1 and pooling\n",
    "        x = self.pool(self.relu(self.conv2(x))) # Convolution 2 and pooling\n",
    "        x = self.flatten(x) # Flatten the output for the linear layers\n",
    "        x = self.relu(self.fc1(x))\n",
    "        x = self.dropout(x)\n",
    "        x = self.relu(self.fc2(x))\n",
    "        x = self.fc3(x)\n",
    "        return x\n",
    "    \n",
    "# Load the model params\n",
    "model_dict = torch.load('./models/mnist_cnn.pt')\n",
    "model = Net()\n",
    "model.load_state_dict(model_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on validation data: 99.24%\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on the validation dataset\n",
    "model.eval()\n",
    "correct = 0\n",
    "total = 0\n",
    "\n",
    "with torch.no_grad():\n",
    "    for data, target in val_loader:\n",
    "        output = model(data)\n",
    "        _, predicted = torch.max(output.data, 1)\n",
    "        total += target.size(0)\n",
    "        correct += (predicted == target).sum().item()\n",
    "\n",
    "accuracy = 100 * correct / total\n",
    "print(f'Accuracy on validation data: {accuracy:.2f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 10\n",
    "images = []\n",
    "confidences = [0 for _ in range(10)]\n",
    "\n",
    "# Create a batch of random images\n",
    "train_images = torch.randn((batch_size, 1, 28, 28), requires_grad=True)\n",
    "optimizer = optim.AdamW([train_images], lr=1e-2)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "targets = torch.arange(batch_size) # The target labels to train for\n",
    "\n",
    "# Train the batch of images\n",
    "num_epochs = 50\n",
    "optimizer.zero_grad()\n",
    "for epoch in range(num_epochs):\n",
    "    outputs = model(train_images)\n",
    "    loss = criterion(outputs, targets)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "# Evaluate the trained batch of images on the model\n",
    "images.extend(train_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate the confidence of each image\n",
    "for i in range(10):\n",
    "    output = model(images[i].unsqueeze(0))\n",
    "    confidences[i] = torch.softmax(output, dim=1)[0][i].item()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize all the images\n",
    "fig, ax = plt.subplots(2, 5, figsize=(10, 5))\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        ax[i][j].imshow(images[i*5+j].detach().numpy().squeeze(), cmap='gray')\n",
    "        ax[i][j].set_title(f'Label: {i*5+j}, Conf: {confidences[i*5+j]:.2f}')\n",
    "        ax[i][j].axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# save the plot\n",
    "fig.savefig('./images/mnist_confidence.png')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 2: Guess using the improved network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 333,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class ImprovedNet(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        # First conv layer\n",
    "        self.conv1 = nn.Conv2d(1, 32, kernel_size=3, stride=1, padding=1)\n",
    "        # Second conv layer\n",
    "        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1)\n",
    "        # Max pooling\n",
    "        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)\n",
    "        # Flatten layer\n",
    "        self.flatten = nn.Flatten()\n",
    "        # Linear layers\n",
    "        self.fc1 = nn.Linear(64 * 7 * 7, 128)\n",
    "        self.fc2 = nn.Linear(128, 64)\n",
    "        self.fc3 = nn.Linear(64, 11) # 11 output classes, 10 means gibberish\n",
    "        # Dropout\n",
    "        self.dropout = nn.Dropout(0.3)\n",
    "        # ReLU\n",
    "        self.relu = nn.ReLU()\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.pool(self.relu(self.conv1(x))) # Convolution 1 and pooling\n",
    "        x = self.pool(self.relu(self.conv2(x))) # Convolution 2 and pooling\n",
    "        x = self.flatten(x) # Flatten the output for the linear layers\n",
    "        x = self.relu(self.fc1(x))\n",
    "        x = self.dropout(x)\n",
    "        x = self.relu(self.fc2(x))\n",
    "        x = self.fc3(x)\n",
    "        return x\n",
    "    \n",
    "# Load the model params\n",
    "model_dict = torch.load('./models/protect_mnist_cnn.pt')\n",
    "model = ImprovedNet()\n",
    "model.load_state_dict(model_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA94AAAHICAYAAAC4fTKEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACUIElEQVR4nO3dZ7RV5fX24RuQ3qQXQZoISJGOKNJEUMCOClawYS+JLWpiCZbEHixJ7L2AHVGKgBQVQUAEQRFpglKkiIIisN8P79AR4nNPDhyX/jW/a4x8yFzMc/bZez1rrYeD8y6Qy+VyAgAAAAAAmSj4a78AAAAAAAB+z9h4AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeOdT7dq11b9//x///7hx41SgQAGNGzfuV3tN/+2/XyOQJdYEsC3WBLAt1gSwLdbE/4bf9Mb74YcfVoECBX78X7FixbTnnnvq3HPP1fLly3/tl7dDhg8frmuuuebXfhlJ119/vQ499FBVqVJFBQoUCF/n0qVLdcwxx2jXXXdVmTJldNhhh+nTTz/95V7s/zjWxC+DNfHbwZr4ZbAmfjtYE78M1sRvB2vil8GakHb5tV/Az+G6665TnTp19O2332rixIm69957NXz4cM2aNUslSpT4RV9Lx44dtXHjRhUpUmSH+oYPH6677777/+Riueqqq1S1alW1aNFCI0aMsH/u66+/VpcuXbRu3TpdccUVKly4sG6//XZ16tRJM2bMUIUKFX7BV/2/jTWRLdbEbw9rIlusid8e1kS2WBO/PayJbLEmficb74MPPlitW7eWJJ122mmqUKGCbrvtNr300kvq169fsuebb75RyZIlf/bXUrBgQRUrVuxn/7q/pgULFqh27dpatWqVKlWqZP/cPffco3nz5undd99VmzZtJP3/z6ZJkya69dZbdcMNN/xSL/l/HmsiW6yJ3x7WRLZYE789rIlssSZ+e1gT2WJN/Mb/qbnTtWtXSf//A5ak/v37q1SpUpo/f7569uyp0qVL6/jjj5ckbd26VXfccYcaN26sYsWKqUqVKho4cKDWrFmzzdfM5XIaNGiQatSooRIlSqhLly6aPXv2T763+28yJk+erJ49e6pcuXIqWbKkmjVrpjvvvPPH13f33XdL0jb/1OUHP/drlKT58+dr/vz5eXo/a9eunac/N3ToULVp0+bHRSJJDRs21AEHHKBnn302T18D2WBNsCawLdYEawLbYk2wJrAt1gRr4uf2u/iN93/74QT4z3+KsHnzZvXo0UMdOnTQLbfc8uM/GRk4cKAefvhhDRgwQOeff74WLFigu+66S9OnT9ekSZNUuHBhSdJf/vIXDRo0SD179lTPnj01bdo0de/eXZs2bdru6xk1apR69+6tatWq6YILLlDVqlU1Z84cDRs2TBdccIEGDhyoZcuWadSoUXrsscd+0p/FazzggAMkSQsXLtyxN9fYunWrZs6cqVNOOeUnx9q2bauRI0dq/fr1Kl269M/y/bBjWBOsCWyLNcGawLZYE6wJbIs1wZr42eV+wx566KGcpNzo0aNzK1euzC1ZsiT39NNP5ypUqJArXrx47rPPPsvlcrncySefnJOUu/zyy7fpnzBhQk5S7oknntim/vrrr29TX7FiRa5IkSK5Xr165bZu3frjn7viiityknInn3zyj7WxY8fmJOXGjh2by+Vyuc2bN+fq1KmTq1WrVm7NmjXbfJ///FrnnHNOLvVxZPEac7lcrlatWrlatWr95PtFVq5cmZOUu/rqq+2x66677ifH7r777pyk3Ny5c3fo+2HHsSZYE9gWa4I1gW2xJlgT2BZrgjXxS/ld/FPzbt26qVKlSqpZs6b69u2rUqVK6YUXXtBuu+22zZ8766yztvn/Q4YMUdmyZXXggQdq1apVP/6vVatWKlWqlMaOHStJGj16tDZt2qTzzjtvm3+yceGFF273tU2fPl0LFizQhRdeqF133XWbY//5tZysXuPChQt/tr+dkqSNGzdKkooWLfqTYz/8Nyo//BlkjzXBmsC2WBOsCWyLNcGawLZYE6yJrP0u/qn53XffrT333FO77LKLqlSpogYNGqhgwW3/TmGXXXZRjRo1tqnNmzdP69atU+XKlZNfd8WKFZKkRYsWSZLq16+/zfFKlSqpXLly4Wv74Z+pNGnSJO8/0C/8Gn8OxYsXlyR99913Pzn27bffbvNnkD3WBGsC22JNsCawLdYEawLbYk2wJrL2u9h4t23b9scphE7RokV/sni2bt2qypUr64knnkj2RBP3fim/hdcoSeXLl1fRokX1+eef/+TYD7Xq1av/0i/rfxZr4tfHmvi/hTXx62NN/N/Cmvj1sSb+b2FN/Pp+72vid7Hx3ln16tXT6NGjtd9++4V/e1KrVi1J//9vi+rWrftjfeXKlT+ZBJj6HpI0a9YsdevWzf45989EfonX+HMoWLCgmjZtqqlTp/7k2OTJk1W3bt3f7iCE/yGsiZ8Pa+L3gTXx82FN/D6wJn4+rInfB9bEz+f3viZ+F/+N98465phjtGXLFv31r3/9ybHNmzdr7dq1kv7/f/NRuHBhDR48WLlc7sc/c8cdd2z3e7Rs2VJ16tTRHXfc8ePX+8F/fq0fMgD/+89k9Rp3ZPx/XvXp00dTpkzZZrF89NFHGjNmjI4++uif9XshG6wJ1gS2xZpgTWBbrAnWBLbFmmBN5NX/9G+8O3XqpIEDB+rGG2/UjBkz1L17dxUuXFjz5s3TkCFDdOedd6pPnz6qVKmSLr74Yt14443q3bu3evbsqenTp+u1115TxYoVw+9RsGBB3XvvvTrkkEPUvHlzDRgwQNWqVdPcuXM1e/ZsjRgxQpLUqlUrSdL555+vHj16qFChQurbt29mr3FHxv8/9thjWrRokTZs2CBJGj9+vAYNGiRJOvHEE3/827Gzzz5b9913n3r16qWLL75YhQsX1m233aYqVaroj3/8Y94+FPyqWBOsCWyLNcGawLZYE6wJbIs1wZrIs19hkvrP5ofx/1OmTAn/3Mknn5wrWbKkPf7vf/8716pVq1zx4sVzpUuXzjVt2jR36aWX5pYtW/bjn9myZUvu2muvzVWrVi1XvHjxXOfOnXOzZs3K1apVKxz//4OJEyfmDjzwwFzp0qVzJUuWzDVr1iw3ePDgH49v3rw5d9555+UqVaqUK1CgwE+iAH7O15jL7dj4/06dOuUkJf/33z/nkiVLcn369MmVKVMmV6pUqVzv3r1z8+bNy9P3Qf6xJlgT2BZrgjWBbbEmWBPYFmuCNfFLKZDL/ce/IwAAAAAAAD+r/+n/xhsAAAAAgKyx8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADI0C55/YNt2rRJ1kuXLm172rVrl6yXLVvW9tSrVy9Z/8c//rHDPZUqVbI95cuXT9Zr1Khhe5577rlkvXXr1rbnm2++SdaLFClie7777rtk/aWXXrI9u+++e7Levn1721OwYPrvXZo2bWp7VqxYkawvWbLE9tSvXz9ZnzFjhu25/fbbk/U77rjD9ixdujRZr1Onju05++yz7bHtad68ebLuzntJ2rRpU7K+77772h73c3300Ue2x513BxxwgO35/vvvk/VbbrnF9uy2227JurteSFKjRo2SdffeSFLhwoWT9WnTptmeQw89NFkfP3687WnSpEmy/sILL9iecuXKJet77bWX7XnvvfeS9QIFCtieZs2aJeszZ860PV9++WWy3qJFC9sTrbHtufzyy+0xd41av3697alSpUqyvscee9ieRx99NFnfddddbU/VqlWTdXfNl6SzzjorWX/zzTdtT/fu3ZP1hQsX2p7p06cn66effrrtGTdunD3WsmXLZH3+/Pm259NPP03Wv/76a9uzyy7pxwu3XiRpwYIFyfrBBx9sez7//PNkPbqefPbZZ/aY8+CDD+5wjyQdccQRyXq/fv1sj3s+mDNnju1x6yi6DhUvXjxZd/cPSZo1a1ay/sEHH9ietWvXJus1a9a0Pe58bNy48Q5/nz333NP2rFu3LlmvXr267ZkwYUKyXqZMGdvjnhei5y13769YsaLtmTt3brLu7ruSNGXKlGS9d+/etqdr16722PZcd9119tjUqVOTdfdcLEkNGjRI1t09Jzq2ceNG2/Ptt98m60WLFrU97jk7ut66rxfd90aPHp2slypVyvZEeyT3PvTq1cv2vP322zv02iR//hcqVMj2DBs2LFn/wx/+YHvc3u7++++3Pe6Z4aCDDrI97rnzP/EbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAylOep5nvvvXeyfumll9oeN+HOTfmT/JTiaJLezTffnKxHk3sbNmyYrEfTmGvVqpWsL1q0yPZ88cUXyXo0OXH//fdP1jt06GB73GTjaIK8m8wZcZMYu3XrZnvc5+Am5UrSG2+8sUNfS5L+9a9/Jeu1a9e2PfmZat6zZ89kPZoUmcvlknU3LVOS2rZtm6xH05PddNVo+qaboNq3b1/b46aNP/bYY7bHTfTeb7/9bI+b+rx8+XLb88wzzyTrJUuWtD1u0v4555xje9zE3Mcff9z2VK5cOVk/9dRTbY+b6hpNfH7ooYeSdXeNyS93PkjSJ598kqxH08ZdKoM7VyU/+XXLli2258MPP0zW3RRySRozZkyyHl1vVq1alay/+OKLtqdHjx7JenTur1692h577bXXkvUBAwbYHjexOprg+sorryTrUcqEu1+5adXRMfdeS37C7datW23PznITyqP7r7sfRNN0J02alKy7c1vy9xZ3jZb8M1I04d4lm2zevNn2VKhQIVmPrjHFihVL1qNp0O78cdPOJT9B3qV8SNK7776brLvnSkmqVq1asu7WsOTTL4YMGWJ73HVu4sSJtic/U82j98mdE1GPu25E57GbMh9dO0eMGJGsR+eXu9d37tzZ9rh9UJRW4RJrovtE9Dy9Zs2aZD16rjnttNOSdZc8Ifnrozv3Jencc89N1qPriUteipKfRo0alaxH17q84DfeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABnKc5yYi/q56aabbE/NmjWT9Zdfftn2uHiyKVOm2J5jjz02WS9UqJDtKVGiRLLu4rIiw4cPt8fOP//8ZD2K+HCxIFH8QMWKFZP1KLLLxd5EURXXXnttsh59pu+//36y/tFHH9mecuXKJetRpISLLPnnP/9pe/LDxR24yAlJatKkSbJet25d2/Pss88m68cdd5ztcXFLLtJJkm699dZk/fDDD7c9LsIiit9yUWdR7EWrVq2SdbdWIi7+S/KRXVEUS6dOnZL1KFbKXU93JoJs7ty5O/x9Fi9ebHvyo2nTpvZYgQIFkvXSpUvbHhfn4aJ5JKl169bJ+pdffml7GjVqlKxHkTku8jCKcHTrMorlctfC8uXL256did4ZP3687XGRdS5WSfL3q+h+ecQRRyTr9erVsz033nhjsh59DlOnTk3W3XmQH+7aHsUnrlixIlmPnoOqVq2arLs1JEnVq1dP1qNnp0GDBiXr/fr1sz0uvjW6fi9btixZj2LvXMRfFCvlXlt0Lrg42Cj2rn///sm6e3aTpHfeeSdZd7Gkkv9Mo2uZey5o2bKl7cmP6Gf+5ptvkvXoGuBie13kqyTdd999yfoVV1xhe1xc1Zw5c2xP8eLFk/XomWKvvfZK1qNrvrvOPPXUU7bHxRBLPm4tigZzMYhRNJh77ovuLS5uLYr5cvvRKC7XPZtEsW7Rc/kP+I03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABnK81TzE044IVl3E5clPwk1mgD4/PPPJ+vHH3+87XET5tzkUslP7Vy5cqXtcVOz3VRjyU/z27p1q+1ZvXq1PeaMHTs2We/du7ft+f7775P1Fi1a2J5NmzYl6zVq1LA9btJn9+7dbc9bb72VrEfvtZtm7KZXStLVV19tj23PCy+8kKy76bKSn0xdqlQp2+MmbbsJm5K06667Juu777677dl7772TdTeNWpL+8pe/JOtHHnmk7TnttNOS9WiKpbvORBO93feJJru7Schu0qrk15Fbk5LUvn37ZN1NfJf8Z7pgwQLb4z7vCRMm2J6LLrrIHtue6Pxyr9Od35Kf3htNXV63bl2yXrRoUdvjppdGCQVuiqybRCxJbdq0SdbdNGbJnyt33XWX7dljjz3sMfd+R5PQ3bTaefPm2R43gdddZyTppZdeStbdZHdJuvLKK5P1XXbxjzfu+7hp1flRqVKlZP3jjz+2Pe6eGSWBHH300cm6m5As+Ynn0T2sWbNmyXo0aXv27Nk7/H1mzJiRrL/44ou2x12nq1SpYnvcPSS6Xrj15aYtS9LDDz+crF911VW2x607d42TfEJClCbgrlmzZs2yPfkRTcZesmRJsl6woP89oZuA755XJemwww5L1qPzy63l6FrjzqPomblWrVrJ+meffWZ7Jk+enKz36dPH9kR7DXcNcs/ZknTBBRck6y6tSvLP0d26dbM9bl9Vp04d27No0aJkPTrHXfpV9JyTF/zGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMsfEGAAAAACBDeY4TO+uss5L15s2b256RI0cm62eeeabteeSRR5L1KNbFxezUrFnT9riIjcaNG9ueL7/8MlmPoqBcHETTpk1tj4vecKPtJemggw5K1ocNG2Z7jjjiiGQ9imhy7+n7779ve1yURxSn4N7TDRs22J4iRYok6/vuu6/tyY/DDz88WX/00Udtj3svovifP/7xj8n6Aw88YHtatWqVrEef09///vdk/aSTTrI977333g5/HxcV5+IrJKlBgwbJuovykqQvvvgiWY/ide6///5kPYqPcNFE0XXOxVddfvnltqdv377Juos5lKSlS5cm69H1Jz8efPBBe8x9zygqyr3+Cy+80Pa4e4iL5ZL8/aVly5Y7/Npc1Ink46D69etne1xkV/TaoutJsWLFknW3XiT/Gbl7ouSjb1z8l+Sv4YsXL7Y9t9xyS7LeqFEj2+Pul9G9b2e56D53DZDiz89xkUrR/c9F/EXPde78fvvtt22Pi5aLovdeffXVZN2tb0las2ZNsh5FVbp7yMSJE23PAQcckKxHn2n16tWT9YULF9qeAw88MFmPYm/79++frPfq1cv2uHtvFJOVH/Pnz7fHGjZsmKxHEWTu2TiKzXTRU+5claQPPvggWXevWZK++uqrZL148eK257XXXkvWo5jW0aNHJ+tRXGd0vrrrt4vElaRjjjlmh1+Di/S96aabbE/btm2Tdbf+JR9f7K4zkn/d7h6aV/zGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMFcjlcrm8/MFu3bol6266nCRNnz49Wa9UqZLt2ZnpqW6i4T777GN73HRZN0FW8pM+Z8yYYXvq1q2brJcpU8b2PPPMM8n63nvvbXvcpMHChQvbnt122y1Z//jjj22Pm9Ie/TxuCmo0+dpNsY4mX7dr1y5Z/9e//mV7xo4da49tz7HHHpust27d2va4SZHu55X8VNroc/r000+T9VWrVtmeggXTfw+3fPly2+MmgBYoUMD2bNmyJVmPpnO71+ASACSpSZMmyXr0vrkpu4MGDbI9btLpk08+aXsOPvjgZL1nz56257nnnkvW999/f9vjEh+i9+Cqq66yx7YnmrTvbjWzZs2yPc2aNUvWoynzbmJuNFm1ZMmSyfrGjRttj5uUvGTJEtvj1nI0UdylALhUCkmaM2eOPeau4ZGvv/46WXdTyCVpwIAByXr02oYPH56sn3DCCbZnxYoVyXp0rXPpGFFKwl//+ld7LHLrrbcm69G55aYkDx061Pa4n3fPPfe0Pe58dPcpyd/ra9eubXvuuuuuZD2aBu3uzW6iuCSde+65ybpbQ5I0ZsyYZN3dpyTplFNO2aGvJfkp2rfffrvtueCCC5L16OepUKFCsh6lILhz8YknnrA9UdLQ9rgEFck/T0dpJO4eHD2LuefPaGK1e56OJpS797Zq1aq2x923Fy1aZHvcHm3cuHG2x01cl3yykJueLvln4uie49aFex6VfBqDO/claevWrcl6dO/dtGlTsh7txU488UR77Af8xhsAAAAAgAyx8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ7vk9Q+2adMmWXcj5yWpUaNGyXoUCeWO9e/f3/a4CKDJkyfbHhcFE42jL126dLLuxtRL0sKFC5P1Vq1a2R73Xjdu3Nj2uEgV93NKPqItip1YtmxZsu5+TslHZUXnjotacBF1ko+ic3Fm+dWrV68deh2Sj05ysXOSdNlllyXrtWrVsj277JJe2lHEztKlS5P1zZs3257q1asn61Fc1SGHHJKsjx8/3vb06dMnWXc/p+Qjg1wEmuSjtKJINRdBeO2119qeDz/8MFl3UYKSj7dx1yXJx8FEsUm/NBftKPmYxCiexEWARBF3a9asSdaj65qLfIk+D3d/c7Fpkr+PRWssind0USjuPZD8Olu9erXtufPOO5P1KBbIfd7vvvuu7XHXoGnTptkeFxkURfPtLHftmDBhgu1xn9/TTz9te6655ppk3cWGSv49imJI3dpz0WSSj3Z0azXqiWKGrr766mS9Xr16tsed21HcqYtHO+qoo2yPu79Fr+3RRx9N1qM4I/eZRj+Pe+6N4t7yw92bJR8deMstt9iePfbYI1mPnhddvOXxxx9ve2677bZk/cADD7Q97voUxY127tw5We/YsaPtGTZsWLIePSNFz4P9+vVL1t26lPx1Oopoc3uU6Lo1ZcqUZD36ef70pz8l61GcmIuCjOL88oLfeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQoTxPNXeTF1999VXbc+KJJybrEydOtD2VKlVK1mfOnGl73ATQHj162J7Fixcn69H0VDcVsmnTprbHTYodO3as7SlevHiy7qZOS9Khhx6arL/44ou2x018LFq0qO1xUwMLFSpke9zU4mg6qZueHr0HbgKne2/yy01Jfu6552yPmyS/ZMkS29OhQ4dkfeXKlbbHrQn3/SU/mfOee+6xPVdeeWWyPnDgQNvjpvNHE0hHjRqVrB922GG2x03LvOGGG2yPc/TRR9tjLtVgzpw5tsdNFa9Zs6bt6datW7Lupt5LfprofffdZ3vyw00blfzkfjdtVJJGjx6drEfTpz/99NNkPZoi7SaHH3DAAbbHJW1cd911tsdNQ42SJFx6wnvvvWd7Jk2aZI916dIlWZ89e7btca8vSs1w6QEu2UGSDjrooGT966+/tj0bN25M1qP7v3vd0X15wIAB9ljEXddc6ovkP4tBgwbt8PePnrfcPTN6bnDnwplnnml7XFKLe96TfJpGdG4XKVIkWY+STdwE95tuusn2fPDBB8m6u15J/pyrX7++7WndunWyfv/999uesmXLJutnnHGG7Tn22GOT9cGDB9ue/HCvUfJpNyVKlLA97t4YJT+4cz+6Prm0nej67aanRxPKXZpG9Mzs0jSiNA83cV3yqSxuLUt+2v4LL7xge9x7Gq0lN1m9du3atse9D9F9wqWQRGkn7du3t8d+wG+8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMsfEGAAAAACBDbLwBAAAAAMgQG28AAAAAADJUIOdm3f+XG2+8MVn/4osvbI/70lF8got2iKJT3Pj2e++91/ace+65yfqMGTNsj4uWmTdvnu1xP+vjjz9ue/bcc89k3UVsSVKDBg2SdRfDJPmYKhcZJklt27ZN1l1kmOTPkShO5cgjj0zWb7/9dtvjophc3JMkXXrppfbY9lx88cXJ+qJFi2yPi7c4++yzbc/8+fOT9Sg6wcWqRNESLv4qii1zXCyQ5KMlNm3aZHv22GOPZH3Dhg22x0V8RD+Pi/mK4t5OOeWUZD06V12kUxQX8s033yTr7rOW/LXk4Ycftj1PPvmkPbY90XnsIn0qVqxoe9zPFsU+unXhYqckqVy5csm6i5CUpBo1aiTr7hySfHRK9NrcOfH555/bnipVqthjc+fOTdaj+/Inn3ySrLu4I8mvmSiCrHLlysl6FNfj1kUUDbbrrrsm6y7+U4ojnCJ/+9vfkvXo/ufumVHMl3sf3H1K8tfiww8/3Pa4ZzEXZyRJr7/+erIe3Y9cTGsUWdi8efNkPYo7dbFOUcyhizOMIqJcTGt0/Xv00UeT9SjSyV1L3LO1JF1++eXJenQ9v+222+yx7VmwYIE99thjjyXr7hlA8tfC6NnPxfxFz78333xzsh49z7vI02+//db2uOfE6Nzv27dvsh5dB6P1555R3FqWpEMOOSRZj6JV3f4p2lvWqlUrWXf7OslH39apU8f2uGcW97wgScccc4w99gN+4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhnbJ6x8cOnRost6zZ0/b4yYXRhPu3BS5qMdN527cuLHtefvtt5P1aMj7Pffck6y76ZuSnwDopqpKflJsNE3QvW/RNFg3ZTOa8nf33Xcn68cdd5zt+e677+wx54YbbkjWo+nS7lj0HuTHV199laxHUzGfeOKJZH3MmDG2x01Pnjlzpu1p2LBhsv7WW2/ZnrJlyybr0QTVLVu2JOtPPfWU7encuXOyHq3xiRMnJuvRBOnFixcn6/vuu6/t+frrr5P16Fri0gbKlClje5YvX56sR5NO3fnm1rEkrV69Olnv2rWr7ckPN0Vd8tP+o2n2JUqUSNZdwoTkp/c+88wztsfdQ6pXr2573JTdN9980/a4yc977bWX7XH3JHeuSv5zl3ziQDQp/ogjjkjWdyZpI3rdlSpVStYnTZpke9x6dtcmSdpnn32S9c8++8z27Cw3bfzLL7+0Pe5+Hj1vPfDAA8n6H/7wB9vjpnNH11V33kcTpN0k+2hqtpt27BJPJP9cN3v2bNvj1uTRRx9te1xywtNPP2172rVrl6y71BLJpxMMGjTI9pxxxhnJepSYcfrppyfr0WTp/IgmPw8cODBZHzlypO1x0+TdM43knw+i53nXE02Mr1ChQrIepS41adIkWY+exdzzYJSeMG3atB0+Fk3hd9PYo/uRu24cddRRtsdNY3/wwQdtT926dZP1KMlq7dq1ybp7LpGYag4AAAAAwK+OjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGcpznFitWrWS9Sjmo3v37jv0tSTpueeeS9ajEf+uJ4qccTENUUSTi/pZunSp7enbt2+y/sorr9iewYMHJ+tRlEChQoWS9ShKa+HChcl6FIfTo0ePZN3FeEj+vY6irf785z8n608++aTtcXEZUYxGfhx66KHJ+m233WZ7Hn744WTdxSNIPkbDRe9IPmIqiok57LDDkvUolueEE05I1s855xzb4+JbXOyGJF199dXJ+pQpU2yPe0+j92DdunXJerFixWzP+PHjk/UoWstF7BUpUsT2uJi/KGalfv36ybqLJssvdx2SfIRiFNn1+eefJ+uvvfaa7dl7772T9VKlStkeF9UYRSu682j33Xe3PRs2bEjWo0gTd51u27at7Ylivtw9LrpXjB49Oll39wPJ/0y77OIfO3Ymjsat2eOPP972tGrVKlmPrt07q0uXLsn6P//5T9vjrh3R+jr55JOT9ejc2rx5c7L+2GOP2R4XpRNFg7l78IgRI2yPiy2LrvkuJs6dI5JUr169ZP2+++6zPWeeeWayHsXEVq5cOVkvWbKk7XFOPPFEe6x8+fLJevRs684D9xnkl4uxk3wcaxQD6p4lo/upi1yNnrPfeOONZD2Kl3LXYncviL7e5MmTbY97fooieKP3x8Uau2hXyb++KCbVHfv0009tz/Tp05P1KNLQxZlGz6ruOXru3Lm2Jy/4jTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGcrzVPP99tsvWXfT5SQ/rc5NNZb85Fk3dVGSOnbsmKxHUylPOeWUZD2arOymE7pJmpKfThi9B926dUvW3WRZSVqxYkWy7iYTSn7aeLt27WyPOw/c9D9Jqlu3brLeokUL2/PBBx8k69EE5PXr1yfr0ZT2K664wh7bnt69eyfr0cTojz76KFmvXbu27fniiy+Sdff5SX7icvHixW3PAw88kKxHE2EvvPDCZP3rr7+2PcOGDUvWDzzwQNvjPsNx48bZHjf1PXqvzzrrrGR9zJgxtsdNmHUTmiV/3Yymlv7xj39M1t00Y0lasGBBst64cWPbkx/R9cZNSo4mq7pp1lFihZveW7hwYdtz0UUXJetDhw61PW7CtLt2ST7l4p133rE9AwYMSNaj+1u0/lzaSPQaDj/88GQ9mi7vPtdoIqw7L91EaEnatGlTsj579mzb464b0bVhZ7kEk2gC9tixY5P1zp07254PP/wwWY9+JveZb9y40fY0b948Wb/kkktsz1133ZWsDxo0yPbkcrlkPZrm79Ii3ORySXr66aeT9Wiqsvt6UfqFuxa76dqSf15waQ+SNG/evGS9XLlytsetVfd+5leU/PDee+8l67169bI97pnVJY5I/lnIPUtL0jfffJOsR+vSraUoqcG59NJL7TGXzDFr1izbU6ZMGXvMPXNF+4MOHTok69G9xd0PojXrnomjn+eTTz5J1h966CHb467R+U2/4DfeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABkqkHOZDf/l3nvvTdYXL15se6pVq5asr1mzxva40fc9evSwPVu2bEnWp02btsPf54ILLrA9Lp7ERTdJUq1atZL1KK7jjTfeSNajaAcXzTVy5Ejbs9tuuyXrBx10kO1xEUlRNIqL2IgiMdxre//9922PO5Vd7JcknXDCCfbY9jRt2jRZP+yww2yPi75xUQeS1LVr12R9yZIltsfF8q1atcr2uPfPvWZJWr58ebLu3pvotb311lu2x0VBRZF0bo1FEVFly5ZN1qNz6IUXXkjW999/f9tTo0aNZD2KIHNRflFcyFNPPZWsu1grSbrlllvsse3p2bOnPXb22Wcn6y5uSfLxjmvXrrU97n7QoEED2+OiS6Lbo7v3denSxfa89NJLyXp0bXexZRs2bLA9N998sz3mIgCj65Zb5y7CSvIxktH9v2DB9O8Cop/Vxf+4yBlJqlq1arIeRTudccYZ9ljEXQujmK8//elPyfq1115re84777xkPbquusisTp062Z5JkyYl61FEqnsP/v3vf9seF9MYRWK66+eUKVNsj3vdZ555pu1ZuXJlsh5di11UZBSV5+KZjjrqKNtz6623JutRTFarVq2SdRehK0mDBw+2x7bnsssus8cWLlyYrG/dutX2uOfP6Frjoqf23ntv27No0aJkPfoMXfSli4qT/Pka7YPc9W7q1Km2J4pCdiZMmGCP1alTZ4e/nvsc3D5I2rlYThenN2rUKNvzxBNPJOszZsywPYceeqg99gN+4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhnbJ6x9000Yvv/xy2+Om6bkpwJKfNlq9enXb46aXFy1a1PZcdNFFyfq8efNsj5suG03Yff3115N1N9VY8lOso/fATdnr16+f7fn666+T9blz59qeJk2aJOvRBNl33303WXfThyU/+Tqaotu4ceNk3X0GUv6mmrup1dHP5c7JaHL47rvvnqy7CZuSn2DqpoNLfqJvNK12zJgxO/x9Pvjgg2S9e/futsdNfV6wYIHtcdNE3bVM8lO0V69ebXt69eqVrEcT5N0U7WgCsZu++Ze//MX2jB49Olk/5JBDbE9+RBM93TXKvReSNHHixGR9r732sj1uMr2bRCz5aazR9NICBQrscI+buty+fXvb4yarr1+/3vZEE5ndNO1XX33V9jiNGjWyxzZv3pysu0nakrTvvvsm69G0f5ccUqVKFdvjnjPuuusu27OzU83dBOrXXnvN9rz33nvJ+tFHH2173DToaCJ8nz59kvXofu4mOH/22We2p1KlSsn64YcfbnvczxMlALjp1gcffLDtcfcwV5ekzp0722OOm+w+cOBA29O/f/9kPbpeuHM7mmDt7lXly5e3Pfnh7rMRl3Qj+Weu6H7qPsMoyaJEiRLJevR5uOnlQ4YMsT0ugSN6pnAT6GvWrGl7oudOtz+IEozcXihK7XHJNFFijUsPiBIP9thjj2TdXWck6dFHH03W3f4kr/iNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG8hwnNnz48B2qSz62x0VLSH4c/NVXX217Tj755GT9lVdesT0uCiaK+XIRSVH8gIsg69ixo+1xsQBPPfWU7XHxA5MmTbI9W7duTdaj+CgXYeUiRiTppJNOStZHjhxpe1x0hHvNklSsWLFkPYpGyY+xY8cm6y6qTvIxCC4ST/KxPO7nlXx0QhTF4uIjXIyVJNWtWzdZd9Epkl97UezF0qVLk/Xvv//e9syfPz9Zv+OOO2zPP//5z2R99uzZtsfFa0QRSJ06dUrWv/nmG9vjorpcPJvkY3SieLT8+Pzzz+0xF13SvHlz2+PO/ehc+fe//52sR7Fr06dPT9aj8/jLL79M1qNrlPtZXUyN5KN3ovMrimhz14AoGmzOnDnJenQeuVg3V5ekoUOHJuvHH3+87dmZuDV3TzjooINsz85y19UjjjjC9rhYnigeqUuXLsm6ezaQfIRkFLHjrsXdunWzPS72MYrec3GQ7r4r+ftRFG/pYiyj99o990Zr38UjLVmyZIdf20MPPWR7XFxfFA125JFHJuvuWipJ1157rT22PdHP7OIBo+gr90wRxT65a0AUp7v33nsn688++6ztWbduXbJ+4okn2h73nP3iiy/anueffz5Z//vf/257ovfU3Q+imGb3us866yzb486xb7/91va4qMiePXvaHhdjHUWQuefLXXbJ89Y5id94AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJChPI9m27JlS7LepEkT2/PVV18l69FEWjcl9ZBDDrE9L7/8crJesmRJ2+OmT0bTxl977bVk/aOPPrI9l112WbL++uuv255Vq1Yl69HkWzcBMJq47ib2Rt/HTfrs3Lmz7bn//vuTdTdxWZImTJiQrB9++OG2x01Ajqb/5sdpp52WrA8bNsz2lClTJlmPzjs3nTea1u4mB0fTGN1k56pVq9qePn36JOv/+Mc/bM+7776brK9YscL2uMnhK1eutD1FihRJ1t20XEl68803k3U3WVrySQhROsD777+frLtzSvKT1d11VpJ23333ZP26666zPdG05e2Jet3E+DVr1tgeN3XVTUmWpNNPPz1Zj1Ih3P1t/Pjxtse9t61bt7Y9tWvXTtZdkoUkvfTSS8n6nnvuaXuqVatmj7lrw9tvv217XDpHNBXXTb920+AlqW3btsm6m9Qs+QQH9/lIfv1F0693lnt9zZo1sz1TpkxJ1o855hjb4z4/lxAi+UQEd42W/PNB9Nzw3HPPJeuNGze2Pe6zcOeIJC1fvjxZj6aNu9fdt29f23PppZcm60cddZTtufPOO5P1M844w/a4e3yrVq1sj5ueHj2numfoq666yvbkR/v27e0x9/zkngGir+eeAST/nN2uXTvb46bwu0QBySd93H333bbnkksuSdajdJWbbropWY/W8rnnnmuPuWehaJ271A6XMCP5z7V+/fq2x02+j/Z8Ls0jSn1wE8+jnycv+I03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbyHCd25plnJutDhw61PQUKFEjWo3H0LvKgXLlytsdFbLioDMnHTlx77bW257DDDkvWo5iGZ599NlmPomBmzZqVrJcqVcr2rF27NlmPYidchE3lypVtj4vyqFSpku0pX758sh5FaLmIJhcDI0n16tVL1l3MQn4VLVo0WY9iYlzURxRb5KK5otgLd+67c1iSKlasmKy/8cYbtuf2229P1r/++mvbc/311yfrUdSZi2+IomVcFEv0fVxMRRR55c79KB7NxSa6GD3Jx5/MnDnT9hx99NHJehRLlB8ublCSqlevnqxH0WDu2hrFF5500knJehRj5WKxojgaF9PkoskkH3n44Ycf2p5OnTol61F00dy5c+2xUaNGJev77ruv7XnxxReT9X322cf2uAgnd6+SfBSbuydK0hNPPJGsu9g0yd9/3eeTHy6WZ/LkybanQ4cOybq7Bkg+NshF4khS165dk/XoGc1d2//2t7/ZniuvvDJZjyL+XDzakUceaXvctd09g0jS8OHDk/Vo7bt7mIssknzUkbvnSH59RVGV7r73yiuv2B4XF3j11VfbnhNOOMEe257oGWnixInJuruXST4y1v1cklSsWLFkferUqbZnr732Staj99atsTZt2tieDRs2JOtRtKu7dkb7hig+0cUgrlu3zva4CDAXxSz5SFgXqSb5a6qLe5Okiy66KFk/5ZRTbI/7XAcMGGB78oLfeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQoTxPNX/88ceT9d69e9seN4l0xIgRtsdNcI0mh2/cuDFZL1jQ/73CrrvumqxHU1rd91m/fr3t+dOf/pSsRxNA3QTOaEKjm5b9yCOP2B43bdFN35T8ZxpNnXVTVaNJ0W4afDTF2k0HPfHEE21PfjRp0iRZjyY7L126NFmPJqh269YtWa9Zs6btcVM+owmXbgp33bp1bU8ul9vhHieanl6hQoVk3U2jlvzU0ui8c5Ovo6nTVapUSdYXL15se9zUYDeZX/ITQw855BDb4ybiR9Pt88N9TpKfyOzOIclPn46m4r788svJunv/IiNHjrTH3PW4R48etueFF15I1qMUAPee3nPPPbbnm2++scfc/eqdd96xPW4KvpsILUnt27dP1qNrg7sORuuvQYMGyfpxxx1ne1yKxLfffmt7dpb7mZo3b2573OcX3Zs/++yzZP3888+3PW6Ke3QOu8nhQ4YMsT3u/Y4m6Y8bNy5Zj1IQXIpL9OzkJoTPnj3b9rhnmjFjxtged967CdaS9MknnyTr0bOtSyCIpoK7a0Ljxo1tT364CfyST1CaNGmS7XGfb+3atW2Pu65OmzbN9rjnz+j7uHSH6Frjnl26dOlie9y6jJ5HXbqT5BNW/vKXv9ged78sU6aM7XHvaZRK5a6P0efgJpQ3bdrU9rjUhYEDB9qevOA33gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQITbeAAAAAABkiI03AAAAAAAZynOcmBt9P2zYMNvjYnaKFi1qe1wsVvR9Pv/882Q9igZr165dsu7iXiSpZ8+eyXoUB3HFFVck6y72QvJxBnvvvbftcRFt69atsz0uQsJFUUlS69atk3UXFyJJ33//fbI+depU27N27dpkvWLFirbHxZkceuihtic/zjrrrGT92GOPtT0uBsHFk0g+YspFX0k+GsytFclHcrgYKEnq06dPsu5iGCSpQ4cOyXr0OX366afJ+vPPP297Fi1alKxH8VtbtmxJ1qMYO/fZRZEpy5YtS9ZdlIrkP4fo8xkwYECyHq29/FiyZIk95iKzVq5caXv+/Oc/J+vRdbpNmzbJurumSD5KK/rcXXzTlClTbI97D1y8peTPyXLlytmeKKrG3ccaNmxoe1xEUdWqVW3P6tWrk/Uo8sXF30Wvzd0X33vvPdvjomXc/To/3Hs0d+7cHe5xEVuSdMABByTr0fpyEURRzNcXX3yRrJ966qm2Z+bMmcl65cqVbU+jRo2S9YMPPtj2uGcn95olfz6681fyn0/nzp1tjxPF0brXEEX/udd2ww032B73XNexY0fbkx9RhNNNN92UrEdxkC6qLbo3ung1F+Em+WekKL5x7Nixybq7T0n+Gdw9S0txtKrjInglf22I9gcuqjWKvl24cGGy7uJ6JR/tHO2r3DqrUaOG7Zk4cWKy/vbbb9ueE044wR77Ab/xBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMsfEGAAAAACBDeZ5qfvjhhyfrbtqwJM2ZMydZr1Onju0ZPnx4st69e3fb46ZSzpo1y/bMmzcvWd9tt91sz+OPP56sR1Osc7lcsv7qq6/aHjedsGzZsranQIECyfoRRxxhewoWTP+9S+HChW3PAw88kKxfdtlltsdNDbzjjjtsT7du3ZL10aNH2x53HkQTnN1U7rxwk0Dfeust2+MmtVarVs32fPDBBztUl6StW7cm6+4zl/yU5miisEsbKF++vO2ZPn16sh5Nd23cuHGyHk3tb968ebIeTdh01wU3jVryEzZLlSple9wk9Gi6/X777Zesu4nTktS+fftkPTp38mPSpEn2mJsQHk2rff3115P1aFqtmyYfnV/uWJTA4VIFosmqboLr5s2bbU+zZs2S9Wiyu5vyK/m0kfvuu8/2lCxZMlmPflZ3L73//vttT8uWLZP1aALw/Pnzk/XoPXD3xWg68c468MADk3X3uiW/1qtXr2573njjjWQ9ure4dRTdZ93rdvccyd9D3n//fduzePHiZP2hhx6yPf3790/Wo3P75ptvTtZPOeUU21OiRIlkPXrmdEk6t99+u+1xU9IPO+ww2+Oer13Kh+TXyuTJk23PeeedZ49tT5R64hJZ9txzT9vjnpmjafYvvfRSsh5diydMmGCP7ajly5fbY256uUukkHwixPHHH297XAqA5M9l9wwr+ee+0qVL2x6XShElOLg0hA8//ND2dO3aNVl3iVCSP8eHDh1qe/KC33gDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZCjPcWJLlixJ1qM4HxfvFMXsuCirjh072p4nnngiWXdj6iVp5cqVybqLvpKkIkWKJOtbtmyxPS6iKYqCWbZsWbIeRTDsvvvuyXoUg+TiOiL16tVL1mfOnGl7XLyWOz8k/9m581CS1qxZk6xHUTT5UaFChWTdxQxJ/nP65JNPbI+LQYiiICpVqpSsRzF2vXv3Ttajz9atPVeXpEcffTRZj2Iqxo4dm6x36dLF9uxMZJY7v3fddVfb46JqXDSS5KNRChUqZHu++uqrZL1WrVq25/nnn0/WXVxJfu2///72mIu5cZFhkn8/ypQps8M9a9eutT0dOnRI1j///HPb07dv32T9tdde2+HXNnfuXNvjIsjcGpd8xJ3k49ZcPKLko4ii1/DMM88k69E5snr16mR95MiRtsfFlkVRp1WrVk3WZ8+ebXtcLNj2PPfcc8n61VdfbXvGjBmTrEfXfHfPjOKR3JqI7s0XXXRRsn755ZfbHheL5c5FKb62Oy6ONoozcnFe559/vu1xkXhXXXWV7XFxT9F51aJFi2R9yJAhtueggw5K1l2cmSQtXbo0WW/SpIntyY/oWnzdddcl62+++abtcdFcTZs2tT0uSs89A0g+btDdmyUfPbfLLn775WLQovXiovRcROr2XsOhhx6arLtzRfL7jXfeecf2uIjEihUr2h4XW9qpUyfb45593XO8JG3YsCFZd3FmecVvvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIUJ6nmrsJ2NWqVbM9mzZtStbdZGdJKlgw/XcBf/3rX22Pm2oaTRp8+eWXk/Wzzz7b9rhJ5HPmzLE9bgri9ddfb3tOOeWUZL1kyZK2p0qVKsl6NNn11FNPTdbdFF1JatasWbL+3nvv2Z66desm69EUazfF1p1Tkp+OuH79etuTHx9++GGyfsUVV9geN0E1mkz97LPPJusNGza0PS4dwE2Yl/w0/WOPPdb2uKmYRYsWtT1uovaUKVNsT/PmzZN1d95LUuPGjZP18ePH2x43eTuaJrrvvvsm69HEZ3fNiJIG2rZtm6y7817y6z/6fPIjmgjrrhFuCq/kJwEfddRRtsfdX6K0gVmzZiXr0WRVl86xxx572B6XjBFNSXX3EDeJXYqveW6CejT5eevWrcl6tC7ceblixQrb4yaru8ngklS8ePFkPXo2cdPTo/N3Z7k1eNttt9meBQsWJOtRuoJLfoim9rp7fZQIM3DgwGS9ffv2tsdNkXfPIJI0bNiwZD26drlzOErScc+c48aNsz0FChRI1qPkGfcaHnzwQdtz+umnJ+vu/JX8eR8917m0hShlIz/q169vjz3++OPJunsGkPzPVr58edvjzi+3n5D8vcU9A0h+MnZ07TziiCOS9Shl4/3330/Wo/c6ur99/PHHyXo06d4lPEXJT+5ncvccSerevXuyHl2DXNJO9H1ceoq75+QVv/EGAAAAACBDbLwBAAAAAMgQG28AAAAAADLExhsAAAAAgAyx8QYAAAAAIENsvAEAAAAAyFCe48SKFSuWrG/evNn2uGiOIUOG2J6mTZsm6+XKlbM9LjLAxaZIPtZp48aNtsfFN0WxACeddFKy/uKLL9qeY445JlkfMWKE7XFRY0ceeaTtcfFNUbTV2rVrk3UXrSX5OLEJEybYnt69eyfr8+fPtz033HBDsn7ppZfanvw49NBDk/Xp06fbnj59+iTrUQyJi7dwa1LykT377bef7XGRL4888ojtcREpEydOtD0u7qRVq1a254477kjWDzjgANvTt2/fZP3OO++0PS56J4rscjFVUfyfi+txn4EkLV++PFn/17/+ZXsGDx6crEfXn/xYuXKlPeZi5BYtWmR7BgwYkKxH16gnn3wyWY9iiFz0W3TfmT179g73vPrqq8l6FHfkIlJcJKYkrVu3zh5zESlR1KCL5oxed7du3ZJ1dz+QpDFjxiTr0WfnnkFcvJYk9e/fP1n/4osvbM/Ocs807nOQpBYtWiTr8+bNsz2nnXZasu4iH6U4Mss58MADk/Uo6sitFReVJ/nPfMmSJbandu3ayXr0rOE+HxfJKUnnn3/+Dr+2VatWJevuPiVJvXr1StZzuZztcbGNOxNFF8VbXnLJJfbY9kTPIe5eH8WJzZw5c4dfg4tqi6KiPvroo2T922+/tT3uvt2mTRvb49aFi72UfCSli9GS4vfUiZ5VXcxXjRo1bI97T6Oo2JtuuilZd9dAyT/bRc8So0aNStaJEwMAAAAA4P8wNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkKECuWg84n9w02W7du1qexYsWJCsR5M5jzjiiGR99erVtuezzz5L1j/55BPbU6BAgWS9R48etsdNPH/zzTdtj5vs7ibsSX6ycTQx84033kjW169fb3vctNzo5/nTn/6UrEcTnO+5555kPXqv3eRbN/VW8u9bdIpfdNFF9tj29OvXL1l3k+wlacaMGcm6m/gsSSVKlEjWo/PbTTWPvk+XLl2SdZcAIPmJkNHEzsaNGyfrCxcutD1uumo0EdutvWgSs5uSHk0OP+qoo5L1jz/+2PY0atQoWY8m4ruUhlNOOcX2uNfQrFkz23PYYYfZY9vjpvZLftp3dB1w5370ubsJ3NG576aAu3Uk+Sn8UTKGux5v3brV9rhJsdH9YOnSpfaYm8gafQ7PPvtssu6mwUv+Pr/PPvvYHvczufu1JBUsmP79wXfffWd73n///WQ9mp7+0EMP2WMRtwbPPPNM2+POoeh+1aBBg2Q9et3umvL3v//d9rhzIbofuc/IJdJIUpMmTZL1KEmnUKFCybpLY5H8vSpKq3FJI1HyTOfOnZP1aLL73XffnaxHz91uMn80wdp9PlGKzGOPPWaPbY9LnpD883x0z3LPFOeee67tcZPV3f1D8teu6LnUfb3oPHY/T7R3clPrO3bsaHui+0SRIkWS9ei+464n9evXtz3uGhDtkdx6jhIzXLKCW8uSv1dG19QobeQH/MYbAAAAAIAMsfEGAAAAACBDbLwBAAAAAMgQG28AAAAAADLExhsAAAAAgAyx8QYAAAAAIEO75PUPujiPaHR66dKlk3UXDSRJH374YbIejbB3ERIu7kXyI/7/9re/2Z527dol60cffbTteeaZZ5L1ihUr2p5hw4Yl61WqVLE9Lmohivn64IMPknUXqST5OJznnnvO9rhYkJkzZ9oeF2XQsmVL2+M+UxdPkV8uSmvIkCG2x537Ls5Dki699NJkvUaNGrbntddeS9Z79eplezZt2pSsn3jiibbHxdhF8T/Tpk1L1qdMmWJ7XHRbFBlUtmzZZD1aR+41RFEsLvJq2bJltsddG6M4DPfzROebixiJPp/8xIm5mDTJR4C4mA9JWrNmTbIeXdvd/aBmzZq2x10Lo/OrcuXKyXoUQ+Qin7799lvb4yLmogiyKBLHRXNOnTrV9rjzyMWtSD4+KVp/jz/+eLJ+3HHH2R4XQxitP/f1Zs2aZXt21hNPPJGsH3/88bbHvY6nn37a9rj7kTtPJR+pdNddd9ke95lH91n3jBTFnbproTtHJKlDhw7JenS9ePvtt5P16LrUunXrZD1axy5qafDgwbbnzjvvTNYnT55se1zk3Jdffml7oufrLCxfvtweW7VqVbI+Z84c2+MiQqMYKxel5c6hqMfFf0nSO++8s8Pfx0X2/fnPf7Y9Tz31VLIencdbtmyxx6pXr56sR/eWhg0bJuvuM5X8M5eLsJb8+Xr//ffbnhNOOCFZj+7xbj3vu+++ticv+I03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABnK81TzSy65JFn/+OOPbU/hwoWT9aFDh+5wTzTN+r333kvWo+myBx98cLIeTS+eNGlSsr5582bb46Yqukm1klS0aNFk/fvvv7c9blq2mygu+SnJbrq15CdLuqm3klS+fPlkfffdd7c97jW4CZGSdNBBByXru+yS59N8h7gJjtHE0TZt2iTr0WtcuXJlsh5NkncTM6Pv4yZ2umnUknTIIYck624qveSnS0cTsd96661k3U2wlqR58+Yl62XKlLE9OzOp2k3zrV27tu3ZsGFDsh5Nl3Xrv169erbHvT/ROsqPaLrryJEjk/Vu3brZHjdt3H22kr+uvP/++7ZnZ6biunuVW0eS1KJFi2Q9mlbt1oW7rkpStWrV7DF3TowaNcr2FCtWLFmPpoC7FBD3vkk+USOazOumX0cJAe5cjJ4zdpa7l61YscL2uGm6PXv23OGe6H7knl2i+3nVqlWT9eh8/Oijj5L1aLKzS15wa1XyE9ejye6ff/55st61a1fbM378+GS9T58+tmfixInJ+tVXX2173POWeyaQpHvvvTdZ79Kli+1xz9BRKk5+RM+Yu+22W7IeTTV3ySJuar8kvfjii8l69DzvEhmie4tLMImeq9x9NHp+69ixY7L+xRdf2B6XMCP55CV3XZekRYsW2WOOS2P45ptvbI87R8477zzb4/ZVLklLkvbaa69kPb9JSfzGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMsfEGAAAAACBDec5ZchFg69evtz177rlnsh6N0XcxRFu3brU9LuqkQoUKtmfu3LnJ+sKFC23PE088kaxHMSwTJkxI1o866ijbM2bMmGQ9itFo2rSpPea4mIN169bZHhdBFvW89tpryXoU6+TiAtznJkkzZ85M1tu3b2978sNFJzRo0MD2uBiCKCbGRVVUqlTJ9jz00EPJehSr4mKGopiYww8/PFl3UR2SNG7cuGT9iCOOsD0uwia6/nz55ZfJehTZ5WJijj76aNvj1v9xxx1ne+67775k3UXUSf46F8UMukiQLGKTpPj66dbhlClTbI+7RkVRIy7i0t2PJOmrr77aoa8lSfvuu2+yHkWqff3118n6NddcY3sefvjhZP3TTz+1PVFUo4sAi6I0XaxKdC93P6s7JyWpU6dOyfqMGTNsT/Xq1ZP1NWvW2B4XaehiOfPDRQdG9wkXceXOBclfP08++WTb497XKC6nXLlyyXoUw9SvX79k3T3rSNIrr7ySrLtIJ8nHzl155ZW2x0WADR482Pa4a3EUYeeioNzzkSS1bds2WXfrUfL3A3ctlXx0mot6yq/o/Pruu++SdfcMEPVEMV8uXmrIkCG25/TTT0/Wo+cq93lEUW2tWrVK1seOHWt73PUkekaKnh3cs/bee+9te1ykYRSt6u5j0fvj4uiieEv3GtyeRvKxij169LA9ecFvvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIUIFcLpfLyx88++yzk/Xhw4fbHjfVd8GCBbbHTSd00/IkafLkycn6LbfcYnvcxONVq1bZHje5+5133rE9jRo1StajaZGrV69O1t3UaclPNCxcuLDtce/18uXLbU+dOnWS9egzdROUq1WrZnvefPPNZD2aaOqmht5///22Z9KkSfbY9tx8883JesGCO/73WStXrrTH3BRJN4lZ8u9fNEnXTdJ86aWXbI/7et27d7c9l1xySbLuJtlLPtUgWq8XXnhhsu4m30p+vbgJ9pKfnty5c2fb8/bbb+/Q94+4ryVJp512WrL+ySef2J4bb7xxh1/DD8466yx7rFmzZsl6NJ27Xr16yfqIESNsz7nnnpusR1Ou582bl6x/8MEHtqdbt27JenRLdakdUSqES3+IXts+++xjj7nPwd0TJZ/OEaVpuM8omjzbt2/fZD2aguymhkeTn917+ve//932ROssMnHixGR92rRptsedJ3PmzLE9btK/m8At+XMouoe5Z7GTTjrJ9rhnjSh5xqUd7LHHHranY8eOyXo02d0lSbip15JUo0aNZN09H0nSG2+8kay764jkP7tly5bZnqVLlybrxYsXtz1ugnU0vTm61m/PBRdcYI+5c8JNmJb8s1B0fj366KPJ+jnnnGN73POGW+OS9Ic//CFZX7Fihe0ZOXJksu6ecSX//Ba9b9E13z0jRAkTbm1G1xN3/kfnuDsvn3nmGdtTokSJZN29b5I0f/78Hf4+Lo3hP/EbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAM5TlODAAAAAAA7Dh+4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNdz7Vrl1b/fv3//H/jxs3TgUKFNC4ceN+tdf03/77NQJZYk0A22JNANtiTQDbYk38b/hNb7wffvhhFShQ4Mf/FStWTHvuuafOPfdcLV++/Nd+eTtk+PDhuuaaa37tl5F0/fXX69BDD1WVKlVUoECB8HUuXbpUxxxzjHbddVeVKVNGhx12mD799NNf7sX+j2NN/DJYE78drIlfBmvit4M18ctgTfx2sCZ+GawJaZdf+wX8HK677jrVqVNH3377rSZOnKh7771Xw4cP16xZs1SiRIlf9LV07NhRGzduVJEiRXaob/jw4br77rv/Ty6Wq666SlWrVlWLFi00YsQI++e+/vprdenSRevWrdMVV1yhwoUL6/bbb1enTp00Y8YMVahQ4Rd81f/bWBPZYk389rAmssWa+O1hTWSLNfHbw5rIFmvid7LxPvjgg9W6dWtJ0mmnnaYKFSrotttu00svvaR+/fole7755huVLFnyZ38tBQsWVLFixX72r/trWrBggWrXrq1Vq1apUqVK9s/dc889mjdvnt599121adNG0v//bJo0aaJbb71VN9xwwy/1kv/nsSayxZr47WFNZIs18dvDmsgWa+K3hzWRLdbEb/yfmjtdu3aV9P8/YEnq37+/SpUqpfnz56tnz54qXbq0jj/+eEnS1q1bdccdd6hx48YqVqyYqlSpooEDB2rNmjXbfM1cLqdBgwapRo0aKlGihLp06aLZs2f/5Hu7/yZj8uTJ6tmzp8qVK6eSJUuqWbNmuvPOO398fXfffbckbfNPXX7wc79GSZo/f77mz5+fp/ezdu3aefpzQ4cOVZs2bX5cJJLUsGFDHXDAAXr22Wfz9DWQDdYEawLbYk2wJrAt1gRrAttiTbAmfm6/i994/7cfToD//KcImzdvVo8ePdShQwfdcsstP/6TkYEDB+rhhx/WgAEDdP7552vBggW66667NH36dE2aNEmFCxeWJP3lL3/RoEGD1LNnT/Xs2VPTpk1T9+7dtWnTpu2+nlGjRql3796qVq2aLrjgAlWtWlVz5szRsGHDdMEFF2jgwIFatmyZRo0apccee+wn/Vm8xgMOOECStHDhwh17c42tW7dq5syZOuWUU35yrG3btho5cqTWr1+v0qVL/yzfDzuGNcGawLZYE6wJbIs1wZrAtlgTrImfXe437KGHHspJyo0ePTq3cuXK3JIlS3JPP/10rkKFCrnixYvnPvvss1wul8udfPLJOUm5yy+/fJv+CRMm5CTlnnjiiW3qr7/++jb1FStW5IoUKZLr1atXbuvWrT/+uSuuuCInKXfyySf/WBs7dmxOUm7s2LG5XC6X27x5c65OnTq5WrVq5dasWbPN9/nPr3XOOefkUh9HFq8xl8vlatWqlatVq9ZPvl9k5cqVOUm5q6++2h677rrrfnLs7rvvzknKzZ07d4e+H3Yca4I1gW2xJlgT2BZrgjWBbbEmWBO/lN/FPzXv1q2bKlWqpJo1a6pv374qVaqUXnjhBe22227b/Lmzzjprm/8/ZMgQlS1bVgceeKBWrVr14/9atWqlUqVKaezYsZKk0aNHa9OmTTrvvPO2+ScbF1544XZf2/Tp07VgwQJdeOGF2nXXXbc59p9fy8nqNS5cuPBn+9spSdq4caMkqWjRoj859sN/o/LDn0H2WBOsCWyLNcGawLZYE6wJbIs1wZrI2u/in5rffffd2nPPPbXLLruoSpUqatCggQoW3PbvFHbZZRfVqFFjm9q8efO0bt06Va5cOfl1V6xYIUlatGiRJKl+/frbHK9UqZLKlSsXvrYf/plKkyZN8v4D/cKv8edQvHhxSdJ33333k2PffvvtNn8G2WNNsCawLdYEawLbYk2wJrAt1gRrImu/i41327Ztf5xC6BQtWvQni2fr1q2qXLmynnjiiWRPNHHvl/JbeI2SVL58eRUtWlSff/75T479UKtevfov/bL+Z7Emfn2sif9bWBO/PtbE/y2siV8fa+L/FtbEr+/3viZ+FxvvnVWvXj2NHj1a++23X/i3J7Vq1ZL0//+2qG7duj/WV65c+ZNJgKnvIUmzZs1St27d7J9z/0zkl3iNP4eCBQuqadOmmjp16k+OTZ48WXXr1v3tDkL4H8Ka+PmwJn4fWBM/H9bE7wNr4ufDmvh9YE38fH7va+J38d9476xjjjlGW7Zs0V//+tefHNu8ebPWrl0r6f//Nx+FCxfW4MGDlcvlfvwzd9xxx3a/R8uWLVWnTh3dcccdP369H/zn1/ohA/C//0xWr3FHxv/nVZ8+fTRlypRtFstHH32kMWPG6Oijj/5ZvxeywZpgTWBbrAnWBLbFmmBNYFusCdZEXv1P/8a7U6dOGjhwoG688UbNmDFD3bt3V+HChTVv3jwNGTJEd955p/r06aNKlSrp4osv1o033qjevXurZ8+emj59ul577TVVrFgx/B4FCxbUvffeq0MOOUTNmzfXgAEDVK1aNc2dO1ezZ8/WiBEjJEmtWrWSJJ1//vnq0aOHChUqpL59+2b2Gndk/P9jjz2mRYsWacOGDZKk8ePHa9CgQZKkE0888ce/HTv77LN13333qVevXrr44otVuHBh3XbbbapSpYr++Mc/5u1Dwa+KNcGawLZYE6wJbIs1wZrAtlgTrIk8+xUmqf9sfhj/P2XKlPDPnXzyybmSJUva4//+979zrVq1yhUvXjxXunTpXNOmTXOXXnppbtmyZT/+mS1btuSuvfbaXLVq1XLFixfPde7cOTdr1qxcrVq1wvH/P5g4cWLuwAMPzJUuXTpXsmTJXLNmzXKDBw/+8fjmzZtz5513Xq5SpUq5AgUK/CQK4Od8jbncjo3/79SpU05S8n///XMuWbIk16dPn1yZMmVypUqVyvXu3Ts3b968PH0f5B9rgjWBbbEmWBPYFmuCNYFtsSZYE7+UArncf/w7AgAAAAAA8LP6n/5vvAEAAAAAyBobbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAM7ZLXP/jPf/4zWX/hhRdsT9OmTZP1Dz/80PYcccQRyfpbb71le4oUKZKst2rVyvYsXLgwWS9evLjt2bJlS7LeuXNn2/Pss88m682aNbM9H330UbIe/TwrV65M1mvVqmV7ChQokKy/8847tse9hhkzZtie1atXJ+vdu3e3PXPnzk3Wu3TpYntee+21Hfpakv988uKyyy5L1qNz9ZxzzknWR44caXvmzZuXrJ911lm2Z+LEicn6FVdcYXsuv/zyZL1MmTK2p1SpUsn6oYceanvcuTJr1izbs2LFimT9mGOOsT1vv/12sl6jRg3b8+ijjybrAwcOtD3jx49P1vfaay/b89lnnyXrzZs3tz3u+vPJJ5/YnqFDhybrN998s+054YQT7LHtic6vJUuWJOvly5e3PY0aNUrWo3W7zz77JOvu/ZOkRYsWJeslSpSwPbVr107Wv/zyS9tTvXr1ZH348OG2p1q1asl6nTp1bI/7eSSpZ8+eyfrUqVNtT+HChZP1xo0b254vvvgiWY/W35gxY5L16D729ddfJ+vu+UOSnnrqqWS9SZMmtidaM5HHH388WXfPIJK/3owbN872uPvfKaecYntefvnlZH3z5s22p1ixYsl6hw4dbM/atWuT9YYNG9oe92zprreSVLdu3WQ9uh+5Z77ovXbXkvXr19uegw46KFmPnk8OOOCAZH3BggW25+OPP07WCxb0v2dzz3W77OK3CO615cVxxx1nj7n36emnn7Y9FStWTNbduRr1TJkyxfbstttuyfquu+5qe9y54u4fkl9/33//ve3J5XLJ+ptvvml7jj76aHvM7Q+i69aqVauS9ehe5Z7tonO8QYMGybo79yX/vPzkk0/aHnevKleunO354x//aI/9gN94AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJChPE81L1SoULJ+8cUX2x43ndtNuZb8hNv27dvbnq+++ipZf/HFF22Pm9YYTQBdt25dsu4mg0p+Ku57771ne0qXLp2su0nVkp/6+sgjj9ie/fbbzx7bUfPnz7fH2rVrl6xHEwjdOTJ48GDb46Y6HnjggbYnPz744INk3U07l6S77747Wd+6davtcVN7o3W0Zs2aZN1N7ZakqlWrJutu7UfH7rnnHtuzYcOGZP3444+3PW6aZ/TzuKmhZcuWtT3HHnusPea4iaYzZ860Pd98802yfvrpp9ue2bNnJ+u9e/e2PW56+iWXXGJ78jPVfPr06faYm1AeTWp1CQvuXJX8eokm9B522GHJupvELkk33XRTsh7dQ95///1k3U2Qlfyk2GgCvpvuLPmptJMnT7Y9LjkkSjW56qqrkvU///nPtuekk05K1pcvX2573PuzdOlS29OmTZtkfc8997Q9O2vSpEnJenRdPfvss5P1o446yva461D0fOLWRNeuXW2Pm/Q7Z84c2+OSbNx1MHpt0XOLe6YYNWqU7XETkr/77jvb4ya4u3QZyU9vjibI//Wvf03W99hjD9vj0kH69Olje9y9KpoKnp+p5tG12CUyRO+tu264RBjJPyNFqUvuuho9v7l0hbFjx+5wT+vWrW3Pq6++mqxH98roZ3XT/vfee2/b4xJrovV3+OGHJ+v/+te/bI97houeP5YtW5asR9cg9/5EnzdTzQEAAAAA+JWx8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADOV5qjkAAJFo0vbo0aOT9VNPPdX2uCSJaFptkyZNknWXFiFJixcv3uHvc/311yfrr7zyiu1x07ldMofkp6RGSRLu55H8dFc3dV7y076jSdZuUnLfvn1tz8iRI5P1AQMG2J6NGzcm69EkdDdJP0ra2FlVqlRJ1t2Ua0mqVKlSsu6m4ktSs2bNkvVoOvDf//73ZD36XF16iEuxkfw0dpcMIvlJ+vXr17c9ZcqUSdajZJO33norWY8myN9www3J+nHHHWd73DToKH3HpdW4tBwpnnjuuOtPlI6QH9F11SU8uOQJyU+frlevnu1x6RfR9Hd33ahYsaLtcSkA++yzzw5/nxtvvNH2uFQDNzVcksaPH2+POVH6hbtfu9cm+Ynnhx56qO1x14Zo6rtLv4jSJdz0/fzeJ/K88XYvesuWLbbHXfBdfIMkDR06NFmPFt369euT9X333df2PPHEE8n6u+++a3s+/fTTZL1y5cq2x329/fff3/a4i2009v71119P1m+77Tbb8+yzzybr48aNsz2dOnVK1qOboYuPqlChgu1xJ3z0HrhF/9prr9meKL5pe1xMw4gRI2zPBRdckKz/4x//sD2HHHJIsv7888/bnkGDBiXr11xzje3p169fsh69fy7+L4pbWLt2bbIexYUMGTIkWXcPWpJUp06dZN1FtEg+tjB6bQ0bNkzWXZyZ5K8ZUTyji6lz573kY0nOP/982wMAAICfH//UHAAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAM5XmquYtieemll2zPAQcckKxHk4hdBEg09t59vXLlytmeTZs2Jetu4rLkYxomTZpke/baa69kfenSpbbHTUKPYg7ca3vjjTdsj5vS/sknn9ge9/ksWrTI9rgp9i4GRvJTzd0kcclHoJxwwgm2Jz/cOTRt2jTbE0XIOG4CduPGjW2Pi3yI4kHuvvvuZL1jx462x0UxuOngktSqVatkfcaMGbbHRRC5aeeSTy648sorbY9LBzjmmGNsj5suf9VVV9meYcOGJevt2rWzPc8991yy7lIDJOnkk09O1pcsWWJ78sOlHkg+CimK8yhYMP13w1FMjLt+umux5ONbXKyT5BM9vv32W9tzxBFHJOtR/FexYsWS9egeEsWduJgWFzMmSdWrV0/WV69ebXtcGoJLaZB8bNh7771ne9y9L4qwca+7ZMmStmdnufUZRd+45xB3X5T8a3cJF5Jfk1F6iLt27bfffrbHXSNddJTk7xPRM42L7Prss89sjzu3o/fAXUuiHhdf1aVLF9vz4IMPJuvR++aet2rWrGl73HXORWHlV3RdddehqMfFWL766qu2x93Thw8fbnvc82eUhuT2LlHEnbsfRc9iLo0oekZySSmSf1Z9/PHHbY+LToyiPN17WrhwYdvj9i7Rc6e7prrXLEkvv/xysn7iiSfanrzgN94AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZCjPU80BAIi4afqSn146dOhQ2+Mm+rvJ3JKfhB5NNnYJGFECh5ug/NBDD9keNyl2wYIFtsdNMN59991tT3TMTXFu2LCh7XGThkuUKGF73FTqaAKwm1D+3Xff2Z6ZM2cm61E6Ry6XS9YPP/xw27Oz3MT6fffd1/a48yFKxXATtStWrGh7ateunaxHE7Br1aqVrEef0bx585L1aPq+63HfX/LrNUrfcUkM0ST9r776KlmPkkZcysbWrVttjztPixQpYns++OCDZN2lI0h+Wv6jjz5qe1544QV7bHvchHdJ+sc//pGs33///bZn4cKFO/x95s6dm6xHyTlu2nh0vXVJMnXq1LE9LhGlR48etsddO1u0aGF7onPcJYcUL17c9rjkpWeffdb2uOuTS8WQfGJFlJRUtGjRHe457LDDkvUvvvjC9uRFnjfeLkaqe/futseN5XcniORHu0cXWxdl5aIYJB994R7aJD/CvlGjRrbHRYmsWbPG9lx//fXJuosTkvwijuJRXJRA+fLlbY9bJFEMm3sgiKKtRo8enayvX7/e9vz5z39O1t1FOb/ce+E2C5J//UcffbTtcTcIFxsi+Q1QdNHcbbfdkvW33nrL9lx99dXJehRh4x50diZiJDrvXGzRfffdZ3sGDhyYrEebtmOPPTZZnzBhgu1xD2gPP/yw7enQoUOyHj2Mu2tjdA0GAADAz49/ag4AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhvI81dxNHB4zZoztcROco9H7Ll7Cxb1IPq7jgAMOsD1Lly5N1ncmPiGKHyhUqFCyPn78eNvj4lHcZGdJ2rJlS7JeqlQp2+OmMdevX9/2jB07Nlk/77zzbM+4ceOS9alTp9oe97rdeH9JuuOOO5L1aOp8fkyZMiVZP/XUU23PtGnTkvW77rrL9tx8883JehSV4c47F4chScOGDUvWjzvuONvz7bffJutt27a1PatXr07Wo3iNL7/8Mll36QSSdMwxxyTrLqJF8hFEUTSR+xyiaeNLlixJ1l1EleQnq0fnt5u8v2rVKtuTHy6VQvIxN1G6gft6LmpIkqpVq5asR7Eh7pxwqRSS9PzzzyfrbjK+5GN7otgyF4PiolukeM26dda7d2/b07Jly2TdrX/JT9SPrg0uRuujjz6yPe79bteune15+umnk/XZs2fbnp11wgknJOvuni35czU6h901pUaNGrbHXaOiiKsRI0Yk6x07drQ9LhooSphwn/ljjz1me9w6ctceSWrSpEmy7mL3JKlNmzbJ+iWXXGJ73LNLtIbcs+Bee+1le9zrfvvtt22PS/OJntXzwyXdSNI999yTrEeJQ5s2bUrWb7vtNttz4403JuvRPXjOnDnJepR60rp162Q9ev51cVmTJk2yPb169UrWo2c+dw+TpAsvvDBZj2IDXTxZ6dKlbY97Hoyi+dx1MLpfu31iFJ0YXW/zg994AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJChPE81BwAg4lIpJD+93E2/lvxk+sGDB9ueYsWKJeutWrWyPW5abLNmzWzPbrvtlqx/8cUXtsdNCY6m/LrJ025atiR16dLFHluzZk2y/tJLL9mefv36JevvvPOO7XGpGS7lQvLT9vv372973LkQTTV3k5+jCeA7y50Po0aNsj0nnnhisu6msUtSw4YNk/VocrGbXj5o0CDb07Vr12Q9Srg54ogjknU3iV3y04bPPvts2/PKK68k6y6VR/ITqY888kjbM2TIkGS9adOmtse910WKFLE9J510UrLukg4kn8xTvHhx2+Mm7Lt0lPxy107JJxtEa9Otseeee872dO/ePVl36S6StHDhQnvMcZPp3QRwSWrfvn2yvnbtWtvjkgOiKe1Rooj7etF7cOihhybrDz74oO1x08sLFChge1q0aJGsf/XVV7bH3Xuj5KfNmzcn61FKQl7keePtHmbczVXyN9joQWtnHkxcbNjEiRNtj3sN1113ne1xD0Dz58+3PTfccEOyHkViuHiy8uXL2x73HkQX26OPPjpZj+IHypYtm6xHn6m7wN155522x8UPRDdqd3N1FxBJ+vvf/26PbY+7OLoYPck/BNWqVcv2PPPMM8l6dFN0sSruJiD5h1sX9yDFsXiOu7FG55C70DZo0MD2uPOrQoUKtsc9vEbxLS5eJ4qicJFXUWzi8uXLk/Xo53GfdxRf5zZZAAAA2Hn8U3MAAAAAADLExhsAAAAAgAyx8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMpTnqeZuhPymTZtsT506dZL1N9980/ZEMQOO+3rRFPCTTz45WV+5cqXtcfEWy5Yt2+HX1rZtW9vjYhNcBEr0fbp162Z7NmzYkKxHMQcunuGKK66wPS5Go3PnzrZn3bp1yfq0adNsj4s5yYqbKv7666/bnqeeeipZ/9vf/mZ73nrrrWS9TZs2tmfu3LnJehQzNHv27GTdrRXJx7e4dALJr7HvvvvO9vTs2TNZv+mmm2yPmxBesKD/+8aPPvooWY+uc+777LrrrrbHrbEo3sPFgnz44Ye2p3Xr1sn6LbfcYnvyY9asWfaYez9atmxpe+bMmZOsR7E98+bNS9b3339/21OiRIlk/fPPP7c9bv1/+eWXtscd22effWyPiwfac889bU+UTNGrV69k3cUqSf5eUaVKFdvzwgsvJOsuukjy0S5jx461PQMGDEjWmzRpYnueffbZZH2//fazPTtr8uTJyXoUV+XOkyj6xl2jdiYiLXoG+OSTT5L1KF3BJTK4pBjJRxNFXDSRW9+S1KNHj2Q9WkP169dP1qNnmjfeeCNZv//++23PrbfemqxHP0+fPn2S9Sjlp1GjRsn6yy+/bHvyw6WUSD6qafHixbbHpRFF7617LnX3D8mfXx06dLA9DzzwQLJ+2mmn2Z6lS5cm61Fizplnnpmsu2dpKX5P3TkW3XtdwlPv3r1tz1FHHZWsX3755bbHPfNE6TMlS5ZM1qOIRvceuHjEvOI33gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQITbeAAAAAABkKM9TzQEAiERpDW5Sc+XKlXe458ADD7Q9kyZNStb//ve/255q1aol67Vq1bI9ffv2TdbdFHJJGj58eLIepQAUKFAgWXdTeaU4HWTJkiXJ+qeffrrDr2Hz5s2255tvvknWO3bsaHsmTJiQrEdTkIsVK5asRxP2mzVrlqy7id35UbRo0WS9cOHCtueLL75I1i+77DLbc++99ybrFStWtD0ukWH69Om2xyUvRKkU3377bbLukmIkv1ZcmoAk7bvvvsm6myguSSNGjEjWo4n9tWvXTtZd6ovkp0tHSRaLFi1K1t05JfnrX4MGDWyPmy6f3+nNzi67+K2He9/dOST5dJXnn3/e9pxyyinJuktJkPy08bvvvtv29O/fP1kvXry47Vm1alWy/vXXX9sel0TgrnVSnIxz2223Jevt2rWzPS4t4sILL7Q9LrHCXZskP4U/Sj1y16foGlS2bNlk/b777rM9l156qT32g3zHiUWROePHj0/WDzroINvjRvm77y/5m7KLiYheWxQ1NGPGjGQ9ipBw8QNRjIaLDHEngeQfmt5//33b06pVq2Q9intxD69uVL/kH4Ciz/Twww9P1qN4H3fTi2Jl8sPF5a1evdr2uI1E9BB27bXXJuujRo2yPS7+Z8yYMbbH/TzR93E35j322MP2uM8wiqNxm5novHPxW6+++qrtces1WkcXX3xxsh49vLobaBRN6OJgopua2/xEr+3UU0+1xwAAALBz+KfmAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGQoz1PNXUzDV199ZXtcjMWCBQtsj4tWiKanuzHxUSyAiyBxP6fkI2ein8dFBrivJfmR+C4GRpLKlCmTrE+ZMsX2uMnuPXv2tD1z5sxJ1qPJ7p999lmyvmbNGtvjpn/Pnz/f9vTu3TtZdxE1+TV58uRkPfpsXbzF+vXrbY+bdB1N9D7qqKOS9ZkzZ9oed65E0U3u65UvX972uOnlQ4YMsT3uPW3evLntGTlyZLJepEgR2+PiW6I18cADD+zQ15Kk/fffP1mPEgXcpP9nnnnG9rjEh2gSen5EMTcugsfF7Eh+Av5dd91le9x7G0WdDR48OFl38TGSVKNGjWTdpQNI0pFHHpmsR+eki3wZN26c7YlivlziweLFi21Py5Ytk/UXX3zR9nTr1i1ZL1GihO1xETJRVNVbb72VrLt7ouSvG+6aIf38a+b222+3x6677rpk3SUbSD426Mknn7Q9LsrPrSHJf34uAk3y51YUEfXxxx8n6/vss4/tcWkaUcLN2Wefnay7OCXJP4fUq1fP9rgkiygiyv2sUWRh06ZNk/Vo3S1fvjxZj56d8pN+ET0fuOeQ6BnJpXe4660krV27NlmPnpFat26drLvnLUnauHFjsn7VVVfZnj333DNZj/ZBLgoxure453lJ6tq1a7Luorwk/6zq9gCSf7aLvo9Lv+rUqZPtcedVlKbjnsfcXiOv+I03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABnK81RzAAAibgq55KfIRj277JK+RZ188sm2x6UlvPrqq7ane/fuyXo06fv9999P1t1rlvyU9MaNG9ue3XffPVmPJk/vtdde9phLjIim+X7//ffJunvfJD/5OZrcv2rVqmS9Q4cOO9wTvQeffvppsh59DjvryiuvTNb/9Kc/2Z5NmzYl69Fk6htuuCFZ79ixo+1xqSsu8USSGjRokKxHKQ4nnnhisj59+nTb4xIZ3CR2yacdNGzY0Pa4ydcu+UbySQNubUl+yn6BAgVsj5s67c4PSXrqqaeSdTfxXfLTulu1amV78mPr1q32mEscKFy4sO1x075Xrlxpe9z1Kfo+LnXh5Zdf3uEeN7lckmrVqpWsH3TQQbbn9ddfT9ajSejRRG/3/kTrr3Tp0sm6mzov+VSKKE3n3XffTdYrVapke9x9on379rbHrT+XDJJXed54u5uYe9MkH1EQxU64mIEoRso9gPTp08f2uMiXsWPH2h734HbSSSfZnn/+85/JenQzdBee6AQZM2ZMsh6N5HcPH6VKlbI9BQum/5GEizKQ/A2nV69etmf27Nk7/NrceRVFPeSH+zzcBVDy0SHt2rWzPW7D4mJDJKl///7JevSeu/iIhQsX2h4XaRZF5bgImSj+z11oozjDfv36JetRLN/cuXN3uMetsSi+yl3n3JqU/I0jik1yMR5RHM2gQYPsMQAAAOwc/qk5AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABnK81RzN6nZTWmWpAMOOCBZv+iii2xP3759k/VohL0b7e6mJ0t+evGhhx5qe9avX5+sR6Py3ddzEQOSj7B5++23bY+b3H3nnXfanp49eybrUayMm6A8efJk21OtWrVkPYrEOO2005L1qVOn2p6iRYsm6y+99JLtcedbXrjokEKFCtmeunXrJutRtISbqO1ikyTpwgsvTNajeBA3OTyKLXKfYfXq1W2Pm7TtJn1LfqL3Rx99ZHvKli2brEcRNvvss88Off/IzsSfuJgayacquAgPyU/6r1q1qu3Jj3r16tljLpki+tzdzxZNf3dT+Dt16mR73PUrSg5YsWJFsh4lLyxYsCBZd8kFklS/fv1k3aWGRK9N8vexKL3ARZpFkS+5XC5Zj9ZSlSpVkvXoHHHX29GjR9ueHj16JOtRSsLOevLJJ5P1KN7JxS1FkXinn356sh5F4rmfN4rLcXFiFStWtD3umhtFy7lYrFNOOcX2uHSQaK24qLooLcKl+URryKVsRPeJe++9N1l3z27R1zv77LNtzzPPPJOsu/hDSTrmmGPsse2Jnsk2btyYrE+aNMn2uOedjz/+2Pa456coYeaVV15J1qPoQndORPdg93zw8MMP2x63p3HnneT3aJJ/vnX3D8lfa6Kf1V03PvzwQ9uz77777lBdkl544YVk3d2TJf98Gz1/5AW/8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ3meag4AQOSTTz6xx9yU1Gh6cJMmTZL1AgUK2B6XMuEmEUtSyZIlk/URI0bYntq1ayfr0dTu1atXJ+tusrLkJ/NGSRLR5Gc3lTaaCNu2bdtkfdSoUbbHTTWPJqG7ZIXx48fbHvd577KLf7xx04Gj92BnufPBJVxIflK7W0OST5gpUqSI7dltt92S9RIlStgeN1167733tj1uenk0Cf3zzz9P1qOJwh988EGyfuSRR9oet46iz6d9+/bJesuWLW2PS4s56KCDbE/NmjWTdZcUI/lp3RdffLHtcVP0ozWUH9E1301sjxKU3GfYokUL2+PWejRl3t3folQalzbw8ssv255169Yl6y5RQJI6d+6crEf3PbdeJH8ede3a1fa4FJDo2vD0008n69E90a3N6N7bunXrHf4+Dz30ULLuJu/nVZ5XlbtZ/vnPf7Y97mR0J6IkTZkyJVl3FwbJP5hEF7QJEyYk69EFwUUNRdEIvXv3TtajKKhKlSol61H8yF/+8pdkffPmzbbHXayihfX9998n6/3797c9LhZg1qxZtmdn4mtctFQUp5If7qLVrFkz2+NeSxQn5B62oogGdwGKvo970HGbEklatGhRsu4eTCQfT/boo4/aHhcFdcghh9ie1157LVl3NyjJn5PRmnjzzTeT9SgC0V3Pouucewjq1auX7fnHP/6RrL/zzju259xzz7XHAAAAsHP4p+YAAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZCjPU82feeaZZD0ave+mMbuoDElauXJlsh7FTrgp4NGE4Pr16yfrmzZtsj1uCribpi1Jl19+ebLu4jUk//M88sgjtufoo49O1qMoETfB2U2Wl6TDDz88WR87dqztcRPKa9SoYXvcuTN79mzb8+233ybr0ST0/HCT6V1siCR98cUXyfoJJ5xge1zsxRtvvGF7unfvnqxHkRwuosF95pL04osvJuvR+VC9evVk3a1JSapVq1ayHkVB3H777Tv0tST/mboYKMlHXrkIJkm6/vrrk/W3337b9lx77bXJepTE4F6Di2DKr9GjR9tj7hoVRaQ88cQTybqLQZKkcuXKJevLli2zPS79IUoBaNiwYbIe/TwFC6b/rnvhwoW2x13X3H1Ciq+T7r6466672h6XTBHFtxQrVixZnz9/vu0pX758sl62bFnbc9111yXr11xzje3Za6+9kvXomWFnuVi16Frs3qN69erZnr59+ybrjz/+uO1p1KhRsh69Dy7FJbqmuOunu6ZJ0s0335ysu8g5ySdmRM9b7plv4MCBtue4445L1t29QJJWrFiRrEdrdfr06cn666+/bnvc/c2dH5KPGIzSPPIjimpzUZFRGtJ3332XrEfPv25dnHPOObbnq6++Stbd9VHy0WAbNmywPXXq1EnW3b1A8ufKztyPJB9DFu0PXIqTS7+R/H006hk3blyyHj2nuX2Ae98kqXTp0sl6FBuYF/zGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAM5XmqOQAAETfVVPJTs//xj3/YngMOOCBZP+igg2zPP//5z2T9888/tz2nnnpqsh5N+h40aFCy7iakR19v+fLltse9px9//LHtiSZm77fffsn6yy+/bHuOPfbYZH3w4MG2x/2sNWvWtD3Dhg1L1qPJvKecckqy3r59e9vjphC3aNHC9vzc9thjD3vs+eefT9ajKfJDhgxJ1qNUj+eeey5Zj6b5u4nLM2bMsD1Vq1ZN1i+88ELb487vaJq/Swf49NNPbY+bUOzSNyQ/7fuqq66yPe7rRQkN1apVS9bnzZtne4oWLZqsR9cEt/aHDh1qe1577TV7bHuihAn3+e7M9zvmmGPsMTeF36UXSX4Cd/R5uDUbrUs3vfydd96xPe666pJaJJ+yI/lp+2eeeabtcfdyd8+RpFKlSiXre+65p+1xKVeLFy+2PW7yfeXKlW3Pxo0bk/UohSQv8rzxdhEt7oYs+WiwaIR9s2bNknU3xl/yb1wUiTFt2rRkPYota9y4cbIenVRLly5N1l1MjuRjH44//njbc/755yfrZ5xxhu1xES3uYVeSJk6cmKw//fTTtueWW25J1qNYuYceeihZP/LII22Pi3Vz525+uagBt1glf1P55JNPbM/UqVOT9V69etke9wDQsWNH2+MuwtHF/vvvv0/We/ToYXtcBJl7yJB8LNbVV19te9xD/EUXXWR73IU7iqNxF2EXeSFJBx98cLJ+ww032B639qLYDRdh07p1a9sDAACAnx//1BwAAAAAgAyx8QYAAAAAIENsvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADOV5qrmbrLxp0ybb48blX3/99bYnithw9t5772TdTQGWfNRJNF7/sMMOS9bHjx9ve+bOnZusd+/e3faUKVMmWd+wYYPtcTEHEyZMsD1NmzZN1m+++Wbb4z47N3Va8u+Bm2AvSZ06dUrWS5YsaXtcPIQ7dyUfRZMXr7zySrIeRVi4SfufffaZ7XGfexRpUKNGjWS9Vq1atsdNuc/lcrbnyy+/TNajdAD3uUcxQzNnzkzWXXJC9PWi+Cr3Hrz33nu255BDDknW58+fb3vcdebxxx+3PS4eKYqjWb169Q7V88tduyT/ebRr1872uHURnZN169ZN1t0EfslHjblYF8nHqrg4H8lfC930eclfZ9avX297omQKN20/ijR78MEHk/Uo8mXJkiXJ+iOPPGJ73PU9uta51IooJsZFvr3wwgu256STTrLHIu5nil5f//79k/XoM3ef0Zw5c2zPunXrkvUoQcX9PFFclYtbctFkkl8TUaSa+1mj65K7V0aRau6+17lzZ9vjnreiCKRJkyYl61GiiUtIiZJTXJJOlJiRH+XLl7fHXIpK9MzszuPoulG4cOFk3V2jJWn69OnJepQS4uK8ovhGty6iCLJ99903WXeRgZJUr149e8ytC5fUJPlnu+hZ/6233krWq1SpYnvcdTA6r9z1JPp53LN39FyQF/zGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAyxMYbAAAAAIAMsfEGAAAAACBDeY4T69mzZ7LuYn4k6YknnkjWo1gAF+fjYpgk6amnnkrWoxiNa665Jll//vnnbc+HH36YrEcxNS7uKIqc+uqrr5L1aOy9i0cpVKiQ7dm4cWOyfvLJJ9seF1kSxRMVK1YsWS9VqpTtqVatWrL+4osv2p6PPvooWe/YsaPtyY/DDz88WXfxdpKPy1uwYIHtcdElzzzzjO3p3bt3sj579mzb46KBovgWd37deeedtsddF1zshuSjIMaMGWN7jjjiiGR92bJltsety+i8c+vVReJJ/j0tUqSI7Xn77beT9WgdNWrUKFl///33bU9+fPfdd/aY+9xdFKLkI0Xeffdd2+Mi86J71aeffpqsu6gqSWrTps0O1SUfVdO1a1fb42KApk2bZnt2Jt7GRWxKfs2MGjXK9vTr1y9ZjyLI3PsTxYy6SJzo3jd27NhkvWXLlrZnZ7nPr0GDBjvcU7p0advj4oSi6JsPPvggWR85cqTtcffz6Pu88cYbyfp5551ne9x9L4ogc68tiut0UWPumUqKo9McF0nZokUL2+PWarTuXNRYFBfoYoGjczQ/3Gcr+XjMwYMH254TTzwxWY+emV08ZxRResYZZyTrUaSni2908cSSvz83b97c9rgoRBebJknVq1e3xyZPnpysR9dIFwUXPde4rxetP/czuecdyUcaRt/HXdOiZ+K84DfeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGQoz1PNp0yZkqxHE8rdxOFoaqCb7lyjRg3b87e//S1ZdxNSJT8RNppI6yb2usm7klS7du1k/eOPP7Y9bqJppUqVbI+b2hlNg3355ZeT9fvvv9/2rF27NlmPpjq6CY3RJO/58+cn69GUzeLFiyfr0bTe/Pj888+T9ejnGjZsWLIeTV53U1yjaajjx49P1qN15D7DDRs22J6FCxcm69HETjd9Pppwe/755yfrjz/+uO1xCQUXXXSR7Rk+fHiyHk3yfO2115L1aPLlzkyLdVPNzzzzTNvjzgM3kT+/3HRcyb+H0Xuby+WS9RkzZtie6dOnJ+sHHXSQ7XHXDve1JJ9y4d5zyU/nd+edJF177bXJurtGStK9995rj/35z39O1t3UZclPfr7wwgttj5vmu99++9keN+24RIkStsclakRJGwcffHCy7qad50f//v2T9VtvvdX2uMnv0b3MTeaPnk+KFi2arFeuXNn2rF+/Pllv0qSJ7XEpG+5+KPkp7e65RfJTiKN75axZs5L16PrtnrcKFChge1zagnumkqSaNWsm6/vss4/tueKKK5J1dx2RpFdffTVZ33///W1PfkTnZIUKFZL16Nncncfu/ZP8RO9HHnnE9rhncJdsIvl0BffcIEnffPNNsh49Z7s1G030j5Kf9tprr2S9bdu2tsetmWgv5j6j6D7q9p3unixJQ4cOTdY7d+5se9z1JEoNyQt+4w0AAAAAQIbYeAMAAAAAkCE23gAAAAAAZIiNNwAAAAAAGWLjDQAAAABAhth4AwAAAACQoQI5l9fyX0488cRk3Y29l6ROnTol61FsSd++fZN1F7El+WiCiRMn2p4zzjgjWY+itLp27Zqsjxgxwva4CJsoDsfFDJQsWdL2HH300cn6Sy+9ZHu2bNmSrLt4DUm64IILkvUoLsBFIPTo0cP2uDH+URTULruk0/GaN29uex566CF7bHtOO+20ZL1YsWK2p1y5csm6i+uRfDxZFC/nonRctIXkI3tcvIckbdy4MVl30TaSj51xcX2StGjRomQ9j5evbURxhu5njdaEu/5E0W0uGiyKqXDRgNH75mJi3PVcks4++2x7bHt69uxpjw0YMCBZHzJkiO0pVapUst6oUSPb4+JTXPyf5M/XZs2a2Z4OHTok69G1fe+9907Wn3vuOdvjrvvRNfeyyy6zx1wMWRTZVaRIkWTdXaclqWzZssn61KlTbY9bMy6WU/Kfa3Stu/POO5P1P/7xj7bnD3/4gz0WcXGn0XnizjsXtyb560B0PypcuHCyXq9ePduzePHiZH3evHm2Z+nSpcl69J66qEgXryn5a2Tp0qVtj3tGcnFmkr/vuKg1ycc97cz7FsVBumjC6D1w72kUzxQ9K2/PySefbI81bNgwWY/u2y5mL7o+uefsKArVRVy5+D9J+uKLL5J1F2Ep+We7d9991/a4+2v0HrjneUlq165dsu5i8SQfT9anTx/b4/YHW7dutT3u+dbdcyS/zgsW9L9/dtfOAw880PacddZZ9tiP33O7fwIAAAAAAOw0Nt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkKH0GOgEN2Uvmp566623JuvRtLpVq1Yl626aoCQde+yxyXr16tVtz7///e9k3U2dlvyU9AYNGtgeN9Gwbt26tuf4449P1qOJmZMnT07WL774YtuzefPmZD2aYu0mEB933HG2x02Edp+B5KdERu/1CSeckKxH0yPzw009jSalDxw4MFl/7LHHbI+bVuumqkvSmjVrkvX333/f9rzzzjs7VJek22+/PVl3EzElqWbNmsl6NCXVTeCPpom69yCahO6maLt0AslPfG7RooXtcekN0TT4t956yx5zrrzyymTdTb7NL3ctlvwk9+gzdFOXo0nN33//fbJetWpV29O+fftkffbs2bbHTdl1k8sl6fXXX0/W3bVL8veQli1b2h5375X8VFo3nV+S6tevn6xPmTLF9nTv3j1Zj1JAunTpkqxv2rTJ9rhzpGjRoranbdu2ybpLxsgPd5+IplkXKFAgWV+3bp3t2ZlJ6AsXLkzWoynybn25a6fkr/lvvvmm7XHTmA899FDb8+KLLybrjRs3tj377bdfsh5db92EcvdzSn76frVq1WyPex4eNmyY7XH3lmh6s3u2jJJ08sNNLpf8uvj4449tz6mnnpqsR6lLLtkjuj655IdoT+MSa6JnsVNOOSVZj/ZBbg/gUiwkaejQofaYm3QfpQq462d03XKvb+3atbanV69eyfpXX31le9z5H019d2s2Oq/ygt94AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGQoz9kZLlZlzJgxtmevvfZK1h999FHb07Rp02T9nHPOsT2TJk1K1l999VXbc/755yfrURyOi4KKok66du2arC9evNj27Lbbbsm6izOTfBRMFPfmftYo0sj9PDfffLPtadWqVbIeRbcUKlQoWd9jjz1sz4cffpisuzgHSerRo4c9tj0zZ85M1l3UgSRNmDAhWe/cubPtcT9XFNn1/PPPJ+tR5IuL/4k+Wxct4daxJN10003JeocOHWzPhg0bknUXbSP5iJQoMtDFXmzcuNH2uKiaKPKlYsWKO/T9Jf/+RN/HxYYdddRRtic/XOxbZNmyZfaYizaMoq/Kly+frLtzSPLXXBcZJkmjR49O1s877zzb464N0Xns1n/Uc8ABB9hju+++e7K+dOlS2+Pi26L4pFGjRiXrzz33nO1xcZXRdd9FLrm4TMlfA9x7nR+vvPJKsh7F2Hz99dfJ+mGHHWZ7XExidL1zay96bW59RZ9R6dKlk3UXlSlJu+66a7L+8MMP2x4XNRa9Nhep1LNnT9vjnoenTZtme4455phkPYrJctdT97Uk//wWnTsuQstFGeaXi+CU/OfuIg0lH6UVRai5CLA+ffrYng8++CBZjyJ43fl60kkn2R4XVxW9Nve5R+s/uua7KK0ossvFCq9evdr2uOt01OOiZ6P3x62zGjVq2B53HY72SHnBb7wBAAAAAMgQG28AAAAAADLExhsAAAAAgAyx8QYAAAAAIENsvAEAAAAAyFCep5q7aXHRpNjixYsn6xdeeKHtmTFjRrL+5Zdf2h43/c5N+ZSkBx98MFmPpvK2bds2WY+mc7sp1tE0wYMOOihZj97r/fbbL1mPpoa6yZItW7a0PZdcckmyHk1JLlGiRLIevdctWrRI1qMpm3Pnzk3W3XTr/Grfvn2yHk1pdp/HW2+9ZXvchNl//etftsdNkY0mXLpJjdFnO27cuGQ9mhS57777JuuVKlWyPW66ZLSO3HqJ1sSZZ56ZrD/yyCO2x53Hn332me1x6/W+++6zPWXLlk3Woymjboq9m2IrxROxtydaa0WLFk3Wo0n7tWrVStbbtGlje1z6g5tiK/n7TrReqlevnqzfe++9tse9B26Sr+TTQaI1Fp3jbkL5Qw89ZHtceoH7WpK//0bniHtPo3ufuwZEaRYuoeT999+3PTvLJYFcf/31tse99jfeeMP2HH744cl6dG9xzzRuIr0kVa5cOVl3SQ2Sf1+jid7uPtqtWzfb41IhTj/9dNuz//77J+suGUTy9zD3WUt+enr0THPWWWcl69F1yaX8uCnV0WvIz70gEt233fOiSy+IjkXv0zvvvJOsv/jii7andevWyXq9evVsz6WXXpqs/+Mf/7A9bgL9woULbY+b4O7uOZK/t0g+mSZ6dnBT3911XfL3qmhSvLuHRIkH7v2Onu1c+lX0nJYX/MYbAAAAAIAMsfEGAAAAACBDbLwBAAAAAMgQG28AAAAAADLExhsAAAAAgAyx8QYAAAAAIEN5jhNz0Qqvv/667RkwYECy3qlTJ9vzzTffJOsuvkHysSWDBw+2PX/605+S9Sja4ZVXXknWGzdubHvcz/r222/bnqeeeipZd+P9Jen7779P1qOfZ/Hixcl6586dbY+LQFm+fLntce+bi1mQfNxbFEXjzpFXX33V9uSHi3ZYsmSJ7XHRCVFEykcffZSsT5061fYceeSRyfqdd95pe0444YRkPYpV2bRpU7LuYrkkH+UTRcu48ziKJalbt26y7iJnJGnIkCHJehRNuGLFimS9fPnytmfMmDHJerT2XGzY7rvvbntcHFf0vuVHFKW3Zs2aZD16/e7cmzZtmu1xESn169e3Pe76FUUruoit4cOH2x4XCeciA6PXEMWgHHLIIfaYO4/69u1re9w6X7Bgge1x5/LNN99se/bYY49kff78+banWrVqybpb/5KPND3iiCNsz86aMmVKsl6zZk3b4653USyWu++4SFPJP4dE9+YPP/wwWXf3HElq1apVsh6dw2+++Way7s4RSbrhhhuSdRebJvnIrEaNGtked++NIlILFSqUrLv4Q8m/7iieyV3/XOyu5J8to+fUDh062GPbE90b3XOpe8aVpHbt2iXr0TOFi6R0z1uSf/6MYgjds8Npp51me9x978ADD7Q9LkJ29OjRtsdFkEnSzJkzk/VoH+JiMaP31D0/RTHNLrYveq6ZPXt2sr7bbrvZHvf+RNF2ecFvvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIUJ6nmnfv3j1Z79Gjh+1x0wmLFCliezp27JisFy5ceIe/z8knn2x73BTCaOL6H//4x2T9888/tz0vvfRSsh5NBnRTNt1UPslPAIzet/333z9Zd5MJJT9ttVSpUrbnkksuSdaj981NincTOyU/adRNE82vfv36JesPP/yw7TnooIOS9eizbd26dbIeTe2dMWNGsh5N7XXvuZvELvmJkNE0aDdBOpqM76aJrly50va489u9n9HXu/zyy23PoEGDkvUGDRrYHjct9ssvv7Q9bppwdA1207Kjcyc/os/Dve/RRNi99947WY/eW3eNis4v9z5FqR1u4ukHH3xge5o3b56sR+fkJ598kqxXqlTJ9kRTV88555xk/aijjrI97vrkrmeSVKJEiWT91FNPtT1PPvlksr7ffvvZHjdJOrq/bN26NVmPpqfvLDfp/9BDD7U9bhqzmzQs+enl7nlC8p9rlJjh7sFR0sDtt9+erEfT/N01Pzq33XNVvXr1bI9bR9E6dlO0o/fNPYtdd911tsed23PnzrU9bhr7yJEjbY9LfHCpGPkVPVO463fx4sVtT8mSJZP1KDEjuj44bhp79Izpnqddyofkr9E33nij7alRo0ay7qaGS/Hrdnu+6D11CULR/sC97mhC+VtvvZWsR+eI2yNFn4P7PsWKFbM9ecFvvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIEBtvAAAAAAAylOc4sXnz5iXrzZo1sz0uKiKKzPn666+T9WgkvosAiWLL3Kj8XC5ne1zsTZ06dWxPz549k/UoPsqN0Y8iGKZMmZKsu1gCSVqwYEGy/s0339ieTz/9NFlv27at7Zk8eXKyXrlyZdvz4osvJutuvL8kPf7448l6FGWQHy7aoUqVKrbnwQcfTNbPP/982+N+5nHjxtkeF+c1dOhQ27Pvvvsm61999ZXtWb16dbI+atQo2+M+9zZt2tgeF3914okn2h4XMTJgwADbc8YZZyTrLn5Ikr7//vtkPYqJOe6445L16Hq6du3aZD2KtnDXrNq1a9ue/IhiytzrHD9+vO1xEXcuKi76etFn6ERRIy4eKIrlchE20f3t5ZdfTtZPOeUU21OhQgV7zMViuggryccnRXGVjovSi75P9MzQvn37ZH3EiBG2509/+lOyHl1Td9aBBx6YrJctW9b2uOhAd02TfKxg9NwwcODAZN1F2EnSwoULk/U777zT9tSsWTNZj67fe+yxR7L+0Ucf2R5373Xvp+Tjt6LYMvfzRLGAjnumknxkoXsel3x8nIsYjcyaNcseO+aYY3b46/0gip2sWLFisr4zz5hRjJyLi4sia130nItOlHzk6mOPPWZ7XCxe7969bY+Ly4oiCF10m+SvhVFU7DvvvJOsX3bZZTv8GqJrfunSpZP1AgUK2B73/BQ9S5QpUyZZj65BecFvvAEAAAAAyBAbbwAAAAAAMsTGGwAAAACADLHxBgAAAAAgQ2y8AQAAAADIUJ6nmm/atClZv/32221P165dk/VoyvTGjRuT9WiCopseHE1JdpP+3BRyyf+sbtKhJHXr1i1ZHzNmjO1p1apVsn7mmWfaHjch0U17lKSiRYsm69WqVbM9EyZMSNZr1Khhew477LBk/V//+pftcZNg3SR2yU8ujaaM54ebbBhNxXRTe5955hnbs3nz5mQ9OldHjhyZrEeTQYcMGZKsu6ndkp8UedZZZ9keN/ly/fr1tsdN5n366adtj0tIuPjii22PmwL+hz/8wfa88soryXo04fu0005L1t3Ue8mvS5eCIEnFixdP1t1nkF/Nmze3x9wE/H79+u1wj5v2K0kbNmxI1qMJvVu2bEnW3XqV/CTbaBqrm0odvW9usqo77yT/80jS2WefnaxXr17d9nz22WfJupsoLklz585N1qNJ6O4a1LFjR9vjJs9Hk5Ofe+65ZD16D3bWsGHDknV3X5Sk448/PlmPngHchPBoAnaTJk2SdXeOSD4ZY5999rE906dPT9YLFSpke9x0a3cuSvGUZueFF15I1t2zm+STdNxalXxyQdWqVW2PS2I44YQTbM8DDzyQrEfn9nvvvZesu88gv6L0APd+uNQVyT+HRNPsXRpR9Czrvl60B3CpNNF10K1Zdy2RpHPPPXeHvpYUPxs3bNgwWXepBpJPjInSC9wzqUtpkKSxY8cm67Vq1bI9u+66a7IeJSW5Z4nomSUv+I03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIYK5HK5XF7+4P3335+st2jRwva8/fbbyXo04t9FCbix7pJUoUKFZN1FPkhS7dq1k/UocsHFaETRDi6q4sgjj7Q9U6dOTdZd/Jfk4xRcFJXk4xnc2H1J+vzzz5P1p556yva46KQoIuaGG25I1jt16mR7XORdFFMVxaZsT6NGjZL1v/3tb7bnr3/9a7Luoq8kae+9907WXRSM5OO3os+pV69eyXoU4XbEEUck6y+99JLtcS688EJ77Morr0zW+/bta3vc5z5lyhTb49ZEFMnhrmdRnNinn36arEefqYvQiuI93NpzMV1RT1506NDBHnPv7VdffWV7XDTfpEmTbI+LNYpivr755ptkvVy5crbHXSfd2pOkDz74IFmPYpVKly6drEf3xOh116tXL1mPIs1efPHFZH3ixIm2x53L0et20TLRmj311FOT9Sgyz0X5uAg0KY6/jLiIqeOOO872uKixd9991/asWrUqWY8ioQoUKJCsL1682PZEcaPOkiVLkvXdd9/d9rh1/M9//tP27LHHHsn6GWecYXvcubVo0SLb49a+ixiV/PU7igV153AUReVeW/Qs6K6Nc+bMsT3PPvusPbY999xzjz32xBNPJOvRvdFdP921U/LvU9myZW3PJ598kqw3btzY9syePTtZj+7bbn8SPYu5aLDomhbFb7kYWRczFn2vKNrx448/Ttbds6Uk9enTJ1m/6KKLbI+LJ3vjjTdsj4uRvfbaa21PFFv6A37jDQAAAABAhth4AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG8jzV3E3hjSZzuonD0cRVN01z69attsdNpHXTWyU/fdJNE5X8FMIVK1bYHjf1/YEHHrA9TZs2Tda/+OIL29OyZctkfc2aNbbHTQZ2U2IlacyYMcl6NOnQTY+MJnm7n7V48eK2x00TjCZzTpgwwR7bHjd1MZoK36BBg2Q9ei/cFNc77rjD9owYMSJZdxPSJalkyZLJejTl+vrrr0/WixQpYnvc2qtZs6btcV9v5syZtuexxx5L1v/yl7/YHjcV97PPPrM97du3T9ZHjx5te9y07smTJ9seN6U9mnA7cuTIZD1KB8jPtFqX4iBJb775ZrLuEhkkf92PpiGXL18+WY+mjbqp9W4SquRft5suLUnvv/9+su5SHCSfzuEm7EavTfLvT3RfdtOBK1eubHvuvPPOZH3Lli22x62l6D7mzoVSpUrZHnfPjtafS2nZnscffzxZjyYUFytWLFmP1vpzzz2XrJ911lm2p379+sl6v379bE+lSpWSdffcIvmJ3tHU6bxMB/5v7pnTTYmWpCFDhiTrbuK75Kcqu89N8udwdL11j+d16tSxPe4ZaezYsbbHPce7Z2tJGjhwoD22PTfeeKM95p5d3D1T8skB0fOBe6656667bE+NGjWS9SgRwqWbXHXVVbbHrQuXpCP51JDoPHYT1yW/p1i2bJntcYkH0UT9pUuX7nCPeya+4IILbM+6deuS9WgtuSntUWJGXu4T/MYbAAAAAIAMsfEGAAAAACBDbLwBAAAAAMgQG28AAAAAADLExhsAAAAAgAyx8QYAAAAAIEN5jhMDAAAAAAA7jt94AwAAAACQITbeAAAAAABkiI03AAAAAAAZYuMNAAAAAECG2HgDAAAAAJAhNt4AAAAAAGSIjTcAAAAAABli4w0AAAAAQIbYeAMAAAAAkKH/B2gw1IVQlLYlAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1000x500 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the improved model on the images\n",
    "model.eval()\n",
    "predictions = []\n",
    "\n",
    "with torch.no_grad():\n",
    "    for image in images:\n",
    "        output = model(image.unsqueeze(0))\n",
    "        _, predicted_label = torch.max(output.data, 1)\n",
    "        predictions.append(predicted_label.item())\n",
    "\n",
    "# Visualize the predictions\n",
    "fig, ax = plt.subplots(2, 5, figsize=(10, 5))\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        ax[i][j].imshow(images[i*5+j].detach().numpy().squeeze(), cmap='gray')\n",
    "        ax[i][j].set_title(f'Predicted: {predictions[i*5+j]}')\n",
    "        ax[i][j].axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Save the plot\n",
    "fig.savefig('./images/improved_model_predictions.png')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There it is! The model predicts jibberish"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 3: Expose the new model\n",
    "Now check what happens if we apply part 1 on the second model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 11\n",
    "confidences= [0 for _ in range(batch_size)]\n",
    "images = []\n",
    "\n",
    "# Create a batch of random images\n",
    "train_images = torch.randn((batch_size, 1, 28, 28), requires_grad=True)\n",
    "optimizer = optim.AdamW([train_images], lr=1e-2)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "targets = torch.arange(batch_size) # The target labels to train for\n",
    "\n",
    "# Train the batch of images\n",
    "num_epochs = 500\n",
    "optimizer.zero_grad()\n",
    "for epoch in range(num_epochs):\n",
    "    outputs = model(train_images)\n",
    "    loss = criterion(outputs, targets)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "# Evaluate the trained batch of images on the model\n",
    "images.extend(train_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 336,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate the confidence of each image\n",
    "for i in range(11):\n",
    "    output = model(images[i].unsqueeze(0))\n",
    "    confidences[i] = torch.softmax(output, dim=1)[0][i].item()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 337,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([2.3984e-38, 0.0000e+00, 3.8255e-43, 0.0000e+00, 0.0000e+00, 0.0000e+00,\n",
       "        5.8424e-41, 0.0000e+00, 0.0000e+00, 2.1930e-42, 1.0000e+00],\n",
       "       grad_fn=<SelectBackward0>)"
      ]
     },
     "execution_count": 337,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.softmax(output, dim=1)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x500 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 11, figsize=(20, 5))\n",
    "for i in range(11):\n",
    "    ax[i].imshow(images[i].detach().numpy().squeeze(), cmap='gray')\n",
    "    ax[i].set_title(f'Label: {i}, Conf: {confidences[i]:.2f}')\n",
    "    ax[i].axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Save the plot\n",
    "fig.savefig('./images/deceive_improved.png')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " Ah, the network failed us again, but we can see some strange patterns emerging now (5 seems like there is 5 hidden underneath?)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 4: Rinse and repeat\n",
    "Now what happens if we create a dataset of images like the above method, and feed it into the network with label 10?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the extended dataset class\n",
    "# This will be wonky, as we will just input the images and labels with transforms already in place\n",
    "class ExtendedMNIST(MNIST):\n",
    "    def __init__(self, *args, extra_data=None, **kwargs):\n",
    "        super().__init__(*args, **kwargs)\n",
    "        # extra_data should be a list of tuples (image, label)\n",
    "        self.extra_data = extra_data or []\n",
    "\n",
    "    def __len__(self):\n",
    "        return super().__len__() + len(self.extra_data)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        if idx < super().__len__():\n",
    "            return super().__getitem__(idx)\n",
    "        else:\n",
    "            # Adjust index for extra data\n",
    "            idx -= super().__len__()\n",
    "            image, label = self.extra_data[idx]\n",
    "            return image, label\n",
    "        \n",
    "    def add_data(self, new_data):\n",
    "        self.extra_data.extend(new_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Loss: 96.1735\n",
      "Epoch: 50, Loss: 9.0303\n",
      "Epoch: 100, Loss: 0.1927\n",
      "Epoch: 150, Loss: 0.0071\n"
     ]
    }
   ],
   "source": [
    "# Create a new train dataset with mispredicted images\n",
    "batch_size = 3000\n",
    "confidences= [0 for _ in range(batch_size)]\n",
    "mispredicted_images = []\n",
    "\n",
    "# Create a batch of random images\n",
    "train_images = torch.randn((batch_size, 1, 28, 28), requires_grad=True)\n",
    "optimizer = optim.AdamW([train_images], lr=1e-2)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "targets = torch.randint(0, 11, (batch_size,))\n",
    "\n",
    "# Train the batch of images\n",
    "num_epochs = 200\n",
    "optimizer.zero_grad()\n",
    "for epoch in range(num_epochs):\n",
    "    outputs = model(train_images)\n",
    "    loss = criterion(outputs, targets)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    optimizer.zero_grad()\n",
    "    \n",
    "    if epoch % 50 == 0:\n",
    "        print(f'Epoch: {epoch}, Loss: {loss.item():.4f}')\n",
    "\n",
    "mispredicted_images.extend(train_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = './models/train_images.pt'\n",
    "if not os.path.exists(file_path):\n",
    "    torch.save(train_images, file_path)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 5, figsize=(10, 4))\n",
    "\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        index = i * 5 + j\n",
    "        image = train_images[index].detach().numpy().squeeze()\n",
    "        label = targets[index].item()\n",
    "        \n",
    "        ax[i][j].imshow(image, cmap='gray')\n",
    "        ax[i][j].set_title(f'Label: {label}')\n",
    "        ax[i][j].axis('off')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_data = [(train_images[i], 10) for i in range(batch_size)]\n",
    "\n",
    "# Create the new examples\n",
    "train_count = 6000\n",
    "val_count = 1000\n",
    "\n",
    "# Create the additional examples as ndarrays\n",
    "train_examples = [(transform((np.random.rand(28, 28).astype(np.float32))), 10) for _ in range(train_count)]\n",
    "val_examples = [(transform((np.random.rand(28, 28).astype(np.float32))), 10) for _ in range(train_count)]\n",
    "\n",
    "# Create the new dataset\n",
    "extended_dataset = ExtendedMNIST('./data', train=True, download=True, transform=transform, extra_data=train_examples)\n",
    "extended_dataset.add_data(new_data)\n",
    "extended_train_loader = DataLoader(extended_dataset, batch_size=64, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 344,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize the optimizer\n",
    "optimizer = optim.AdamW(model.parameters(), lr=1e-3)\n",
    "criterion = nn.CrossEntropyLoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [1/5], Loss: 0.0206, Accuracy: 99.35%\n",
      "Epoch [2/5], Loss: 0.0193, Accuracy: 99.40%\n",
      "Epoch [3/5], Loss: 0.0185, Accuracy: 99.41%\n",
      "Epoch [4/5], Loss: 0.0172, Accuracy: 99.45%\n",
      "Epoch [5/5], Loss: 0.0145, Accuracy: 99.56%\n"
     ]
    }
   ],
   "source": [
    "# retrain the network\n",
    "model.train()\n",
    "num_epochs = 5\n",
    "for epoch in range(num_epochs):\n",
    "    running_loss = 0.0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    \n",
    "    for batch_idx, (data, target) in enumerate(extended_train_loader):\n",
    "        # Forward pass, backward pass, optimize\n",
    "        optimizer.zero_grad()\n",
    "        output = model(data)\n",
    "        loss = criterion(output, target)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        # Calculate metrics\n",
    "        running_loss += loss.item()\n",
    "        _, predicted = torch.max(output.data, 1)\n",
    "        total += target.size(0)\n",
    "        correct += (predicted == target).sum().item()\n",
    "        \n",
    "    # Print metrics for each epoch\n",
    "    print(f'Epoch [{epoch+1}/{num_epochs}], Loss: {running_loss/len(extended_train_loader):.4f}, Accuracy: {(100*correct/total):.2f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the improved model on the mispredicted images\n",
    "model.eval()\n",
    "predictions = []\n",
    "\n",
    "with torch.no_grad():\n",
    "    for image in images:\n",
    "        output = model(image.unsqueeze(0))\n",
    "        _, predicted_label = torch.max(output.data, 1)\n",
    "        predictions.append(predicted_label.item())\n",
    "\n",
    "# Visualize the predictions\n",
    "fig, ax = plt.subplots(2, 5, figsize=(10, 5))\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        ax[i][j].imshow(images[i*5+j].detach().numpy().squeeze(), cmap='gray')\n",
    "        ax[i][j].set_title(f'Predicted: {predictions[i*5+j]}')\n",
    "        ax[i][j].axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Loss: 304.2837\n",
      "Epoch: 50, Loss: 108.2705\n",
      "Epoch: 100, Loss: 10.9776\n",
      "Epoch: 150, Loss: 0.6395\n"
     ]
    }
   ],
   "source": [
    "# Create a new train dataset with mispredicted images\n",
    "batch_size = 30\n",
    "confidences= [0 for _ in range(batch_size)]\n",
    "mispredicted_images = []\n",
    "\n",
    "# Create a batch of random images\n",
    "train_images = torch.randn((batch_size, 1, 28, 28), requires_grad=True)\n",
    "optimizer = optim.AdamW([train_images], lr=1e-2)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "targets = torch.randint(0, 11, (batch_size,))\n",
    "\n",
    "# Train the batch of images\n",
    "num_epochs = 200\n",
    "optimizer.zero_grad()\n",
    "for epoch in range(num_epochs):\n",
    "    outputs = model(train_images)\n",
    "    loss = criterion(outputs, targets)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    optimizer.zero_grad()\n",
    "    \n",
    "    if epoch % 50 == 0:\n",
    "        print(f'Epoch: {epoch}, Loss: {loss.item():.4f}')\n",
    "        \n",
    "mispredicted_images.extend(train_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the mispredicted images\n",
    "fig, ax = plt.subplots(2, 5, figsize=(10, 5))\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        index = i * 5 + j\n",
    "        image = mispredicted_images[index].detach().numpy().squeeze()\n",
    "        label = targets[index].item()\n",
    "        \n",
    "        ax[i][j].imshow(image, cmap='gray')\n",
    "        ax[i][j].set_title(f'Label: {label}')\n",
    "        ax[i][j].axis('off')\n",
    "        \n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('./images/second_mispredicted.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As can be seen above, we can iterate, feed the trained images back into the network, but we can still manage to deceieve the network. A followup would be What happens if we could reiterate this many many times. Which is not ideal with my computer as each training around 10 minutes so iterating a meaningful amount of times would take many hours/days (2 trainings per iteration each around 10-15 minutes). I guess this could be improved by using a simpler model, but I don't we would get interesting looking patterns if we don't use a CNN."
   ]
  }
 ],
 "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": "3.11.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}