1 files changed, 1038 insertions, 0 deletions
diff --git a/Fundamentals_of_Deep_Learning/04b_asl_predictions.ipynb b/Fundamentals_of_Deep_Learning/04b_asl_predictions.ipynb
new file mode 100644
index 0000000..8f26209
--- /dev/null
+++ b/Fundamentals_of_Deep_Learning/04b_asl_predictions.ipynb
@@ -0,0 +1,1038 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "EKFbMeI4Csut"
+   },
+   "source": [
+    "<center><a href=\"https://www.nvidia.com/dli\"> <img src=\"images/DLI_Header.png\" alt=\"Header\" style=\"width: 400px;\"/> </a></center>"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "GyRFI0IJCsuu"
+   },
+   "source": [
+    "# 4b. Deploying Your Model\n",
+    "Now that we have a well trained model, it's time to use it. In this exercise, we'll expose new images to our model and detect the correct letters of the sign language alphabet. Let's get started!"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "wdYYMU1WCsuv"
+   },
+   "source": [
+    "## 4b.1 Objectives"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "SeXBHUIjCsuv"
+   },
+   "source": [
+    "* Load an already-trained model from disk\n",
+    "* Reformat images for a model trained on images of a different format\n",
+    "* Perform inference with new images, never seen by the trained model and evaluate its performance"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 260,
+     "status": "ok",
+     "timestamp": 1715244405026,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "idwYaoghaqIp",
+    "outputId": "79b63437-52ca-448c-92c5-59857227be18"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "True"
+      ]
+     },
+     "execution_count": 1,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "from torch.optim import Adam\n",
+    "from torch.utils.data import Dataset, DataLoader\n",
+    "import torchvision.io as tv_io\n",
+    "import torchvision.transforms.v2 as transforms\n",
+    "import torchvision.transforms.functional as F\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "torch.cuda.is_available()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "xo-pr5-zCsuv"
+   },
+   "source": [
+    "## 4b.2 Loading the Model\n",
+    "Now that we're in a new notebook, let's load the saved model that we trained. Our save from the previous exercise created a folder called \"asl_model\". We can load the model by selecting the same folder."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Since our model uses a [custom module](https://pytorch.org/tutorials/beginner/examples_nn/two_layer_net_module.html), we will need to load the code for that class. We have saved a copy of the code in [utils.py](./uitls.py)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 169,
+     "status": "ok",
+     "timestamp": 1715242471134,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "jiI87fNVZdRl"
+   },
+   "outputs": [],
+   "source": [
+    "from utils import MyConvBlock"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now that we have a definition for `MyConvBlock`, we can use [torch.load](https://pytorch.org/docs/stable/generated/torch.load.html) to load a model from a path. We can use `map_location to specify the device. When we print the model, does it look the same as in the last notebook?"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 180,
+     "status": "ok",
+     "timestamp": 1715242473387,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "-Xy1JuL-Csuv"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Sequential(\n",
+       "  (0): MyConvBlock(\n",
+       "    (model): Sequential(\n",
+       "      (0): Conv2d(1, 25, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
+       "      (1): BatchNorm2d(25, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+       "      (2): ReLU()\n",
+       "      (3): Dropout(p=0, inplace=False)\n",
+       "      (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
+       "    )\n",
+       "  )\n",
+       "  (1): MyConvBlock(\n",
+       "    (model): Sequential(\n",
+       "      (0): Conv2d(25, 50, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
+       "      (1): BatchNorm2d(50, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+       "      (2): ReLU()\n",
+       "      (3): Dropout(p=0.2, inplace=False)\n",
+       "      (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
+       "    )\n",
+       "  )\n",
+       "  (2): MyConvBlock(\n",
+       "    (model): Sequential(\n",
+       "      (0): Conv2d(50, 75, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
+       "      (1): BatchNorm2d(75, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+       "      (2): ReLU()\n",
+       "      (3): Dropout(p=0, inplace=False)\n",
+       "      (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
+       "    )\n",
+       "  )\n",
+       "  (3): Flatten(start_dim=1, end_dim=-1)\n",
+       "  (4): Linear(in_features=675, out_features=512, bias=True)\n",
+       "  (5): Dropout(p=0.3, inplace=False)\n",
+       "  (6): ReLU()\n",
+       "  (7): Linear(in_features=512, out_features=24, bias=True)\n",
+       ")"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "model = torch.load('model.pth', map_location=device)\n",
+    "model"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "WIjqbFCFCsuv"
+   },
+   "source": [
+    "We can also verify if the model is on our GPU."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 174,
+     "status": "ok",
+     "timestamp": 1715242474520,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "_ulCst9GCsuw",
+    "outputId": "0d55b14d-9b52-49e7-efda-76a05abe21ed"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "device(type='cuda', index=0)"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "next(model.parameters()).device"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "pBBneFIXCsuw"
+   },
+   "source": [
+    "## 4b.3 Preparing an Image for the Model"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "gffCLeacCsuw"
+   },
+   "source": [
+    "It's now time to use the model to make predictions on new images that it's never seen before. This is also called inference. We have a set of images in the `data/asl_images` folder. Try opening it using the left navigation and explore the images.\n",
+    "\n",
+    "You'll notice that the images we have are much higher resolution than the images in our dataset. They are also in color. Remember that our images in the dataset were 28x28 pixels and grayscale. It's important to keep in mind that whenever we make predictions with a model, the input must match the shape of the data that the model was trained on. For this model, the training dataset was of the shape: (27455, 28, 28, 1). This corresponded to 27455 images of 28 by 28 pixels each with one color channel (grayscale)."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "UXeujDRFCsuw"
+   },
+   "source": [
+    "### 4b.3.1 Showing the Images"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "8gLvKQ1yCsuw"
+   },
+   "source": [
+    "When we use our model to make predictions on new images, it will be useful to show the image as well. We can use the matplotlib library to do this."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 131,
+     "status": "ok",
+     "timestamp": 1715243759243,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "qzzai3QZCsuw"
+   },
+   "outputs": [],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "import matplotlib.image as mpimg\n",
+    "\n",
+    "def show_image(image_path):\n",
+    "    image = mpimg.imread(image_path)\n",
+    "    plt.imshow(image, cmap='gray')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/",
+     "height": 434
+    },
+    "executionInfo": {
+     "elapsed": 745,
+     "status": "ok",
+     "timestamp": 1715243543244,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "ZFzwvkalCsuw",
+    "outputId": "74ad4eac-4f99-413f-9d9f-54cad0d5d7e7"
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAa0AAAGhCAYAAADSlOtMAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOy9fax3y1UX/lmz9/ec595LaVOgb+FSkGhAeYuAtRGxSgGLIRIbIy9/gBAakpYEbox4DYI0JjXGxAat8A8pMT8aXxLEyB9NbI0lJAWxhjSINrQhomKLSsqlhXvO2XvW749Za82al/3d+/tynuc+55452Wd/996zZ2bPnr0+81lrzQwxM+M+3If7cB/uw314DEJ41AW4D/fhPtyH+3AftoZ70LoP9+E+3If78NiEe9C6D/fhPtyH+/DYhHvQug/34T7ch/vw2IR70LoP9+E+3If78NiEe9C6D/fhPtyH+/DYhHvQug/34T7ch/vw2IR70LoP9+E+3If78NiEe9C6D/fhPtyH+/DYhHvQug/34T7ch/vw2IRHClrvete78Pmf//l48OABXve61+E//sf/+CiLcx/uw324D/fhBR4eGWj9i3/xL/DMM8/gR3/0R/Gf//N/xpd/+ZfjG7/xG/E7v/M7j6pI9+E+3If7cB9e4IEe1YS5r3vd6/DVX/3V+Cf/5J8AAGKMePrpp/H93//9+Ft/62/tvTfGiN/+7d/GS17yEhDRwyjufbgP9+E+3IdbDMyM3//938drXvMahLDMp8aHWCYL19fX+NCHPoRnn33WzoUQ8MY3vhEf/OAHm/hXV1e4urqy4//1v/4X/vgf/+MPpaz34T7ch/twHx5e+B//43/gcz/3cxevPxLQ+r//9/9inme88pWvLM6/8pWvxH/7b/+tif+Od7wDP/ZjP9ac//Vf/3XEOOP55/8Av/Vb/wPPP/9pPPd7v4d5ngGgYWFrx2th9f7imGxPLu5anv761uLt48r70lh6npp8H0LGl9JYy7OJp7fX6dRls2qu05W9xkN1oo4IdjuWI7bjlEKU261w7hdZrCIG5Xbg89M7ffGLmMwg5rwHgzhdCwQEEEa5YdDzHFO9yX0s90XJa2Yuy0i5zKzHRGDSa/LbPROz3qOvgvIrqfdF/XbChnZld3P5I9+qz7KWdH2irP99bX5r+9d4dfy1b2Ipn0O/w9PTX0+rbvn5tSznPQwDnnrqKVxcXOIVn/NKXF5e4iUv+UzsdjvsdjsMw4CLiwt86lOfwp/8k38SL3nJS/Y+xyMBrUPDs88+i2eeecaOn3vuOTz99NN4yUteghgnDEPAU089gUCMm+vrBrSOBatDAKY8rvOlJJSOAK598bd+TIfmd2j6p4StoNX9OFTQG0LVnQgPWosFKA4NhpjBWTyDdOP6s/UfchLoEDDI11xBGtTMZ4KlIvuoYCXgJXEHaU+7kDpEBlpRQDXGhLccwQJeDDbQUuiN5MBK90TgoNelRFJHUR4o76nA9bTvd+Ss7fu6XmtfrADuTvi80O5z2li4Up7fAlprYQmstt53bHq3BYLnBCwgadGeeOIBLi8epP3lAzz55JMYxxGXl5cYhgGXl5cWf01ePRLQ+uzP/mwMw4BPfOITxflPfOITeNWrXtXEv7y8LB5KAzODI0vPEkjgsA0Y7sOjCAyIoOvI7sNCzbA6gHVKYEIl7/RE2rPtJT4WyrFSEKo2QACbgcB6nK4EZmFYhADCBQ0IRBgIIGYQRfkWUtmigJSCFaLsOcqTuDITAAryeAEgQqBoAMxECEUNACT1QNX5/HDLgKXn9wreJmHJsVelt9+/Wg2rz/OI0ztXWHzfS/GJECNjjhE3NzegEHB1fYUYI4ZhADNjHEcjG2vhkXgPXlxc4Cu/8ivx/ve/387FGPH+978fr3/96w9Or644D1yluo0eCZgtZanlqbf6+nK6/fuX0lu6/5D0TwsLKrp+rKVCSYWSVSz7cvnrIGMPBxWx0uj5E6o6AxSoEgcTHXBZBvIM2yfsUuUMUvpb1YBA/h04ARUBCEQCViFvIWAgsi0IsKUtAWBKOzM322IEIgMc7TopsxOhafc41lnUTPPoHbX4QhtHb/Pvz+o9ZcBdpNcUudrWzp8YuPyOziFfzp1eL93e8fp5fTWufC7+uvxhRI6IMSLOso9JG6D7LeGRqQefeeYZfOd3fie+6qu+Cn/qT/0pvPOd78SnP/1p/PW//tc3p8HML4QO1osq3Hbvr8UJyuqHHkMjak6dv0Tc7GvblcVtHmA51R7Dot4eCagGBAQCLsKAIRAejDsMgTBKmYgmcGQwJnCMiWFxxBQTQ+GY2FeMEYE5qflISQwBIQCBQMxgcnu5XkOuMi87Q8GuFXWzInhXe+wthStPrHb9D2+vm9v5w+8DnxyaZ/Pf2C2E1N9J7W66mRFows3NDQBgt9vZfivTemSg9df+2l/D//k//wc/8iM/go9//OP4iq/4Crz3ve9tnDM2hQPr+1QHjFPTe6GGtefQhn4bz9vY8XyeRImNNL3qIoHyvgbc1gqwcJIqoGI53lOG/DPbtmiljSqzaY01ylhSHQUQAgUEIoxBGJYWP0YgMGIkMAUwRTATAiUiFZKhK4ERJ7Zk/QGSZ4oVZZKy2XMgcS0SsMuPXL+cqoI0YtWGuI5t5eHqmPxbyHdYektyd5tw6DlL1G3yUBvy3vi8/B0tOXLU19fu99f958PMCbg0ffnGDu2Mbq8fNtuqsivPsvy2JTxSR4y3ve1teNvb3nZaIvqg1m3tv8iHDyzlx/s4AduSyemR6NhVGj301+cASm03XNdM1VtNP7Yl39nU8aBmWgEwW9ZAARfDgDEMuLy4wEgBo7IiABxjYlMUEWJEJICQ/B5jBCgy4jwDnGwM/hl4SLYsDENiXfZcbCqhBB5UcU8HdHUdlNhzdNhSu6uMbSHNRxI2ZH7O7637PaNjg7yt75sBjox5njFNE6ZpAhFhnmckm9djoB48b3C9I3R67Qfqh0+Nt9T7fFyAa69g2NKLPCSvhXdlx9oL3Jhetl+5zow7bAvQFEjicwlcnJ0OpGC9m1uGtV7isiieaekjkIKX2KooIISAIQwYQsCIVE8xhOQa4gGGCApN5kghti2IdyIzi2s7kj+91kGMCbzYFUQ6ENkJo2KedX1TBjh/rBH0mbMqsb6/rq8FxuaOE5PI6fRevRW3c38d6mu33XHz+fSAa60ch1wn5G+mZlz2KdT3y968HDc8U2JSQM2sHjumdWpQm9Y+AXFuVeCGBPYd3odDwy3r2w8LXqJmJVVj19qYkioGiDMwAZlheUlORAgCUrtxh90w4MHuIh0L04pg8Bwwx8S0aI5itwoISDasACByBOYImiZADOOMxLSYAIwjEEJSWQ4BwAizaUmBFbDYaqDlX73QNT0dSpH2VSocQ3H9Fm6jbU/2UWgY8PCa/lr1n+v1ePXgPM8IITRqwi3hsQYtH/YB1tZxT1uvr95X2Vfq5B5WeU4N+3TWW+1fa2kunq8YT33XavNuGNdKHdbx2whSBuFbZmuR91z1SpfdQypmpT8qYljmLCmSghchBGVbyWuQmEEhpI9fACeKXSoApcu6gmIUF3kBrQhFz6Q+pGFIz8cCd2YLUajqKArN5tfCU+ken98vw1d/dZxPV5Wyzri07fQ5cZ1cr+bTnTXjWWIya6xtUSi7eNrG3KO5dOrzVWmr5LdeV+07V+etNHb9eOhSerHErl6EoEXV9ggEfaXPp6ocm8qz3EE9Szg1+WN7nLf8WIeHqkALij750FQ86/k6XtuLX6uhsqXmcSe55WQdoW9NiZklwBqGgHG3wy4EXIjH3xynpBKc58SwhsSwFMRmAf+Jk32B5gjEGTxPiMyYkUCYd8mmNRJAcUAIBApIqkNGGjdGlNI2cdazdNXjuCoVa6fuu0zs2CDAdXx6D7fVtrmdrSa25b/C7E4tDWNZPXhIuCOglUPqXJ6bYa3dV/5YAqzV8qwwslNDj60coyoBDtPr78tjkWFVx02vVv6tl0L7rlyfsvy9xYxc/HrYbPl2IfFQqPJynS6XLDM2f46K+/a1RBvzFAgUQtozJ3ULM0IggIWVycZuT0iehMQxsShhXJGjzIiBdBxnSYvBiCAMrlylilAZFilbY3891yA5N8olhpXYYGkvsXbiqpXhEqjZDcs/oo4fz+0CwenfSGZWfZuWXD3TY1RKDXDleqvvzL6IozNeZlWHpPnYg5bvsZJKsrPnsP36bY8aOmd4wZT0KPQ8ZwG2J9ZjWhp6rGHfPX5v7biIbEqV9EeMSEigEghhGBDGZOciZmAekoCZAgiMGAjEabAxESWmRYRR1DSBk92LpwkxzriZ52QXm0ZgCMAYMCCCxoBAA4gHKW+QfRRBFpEZImXpt5eD7qvzErCWwmrPX5lWwyAeLoNZC4sdlBXmc1thqb6Wam2tNmvNwqnhsQctDd6ctM6wtgFRrdtdjb+Y37bgx65sSeHYfM5hVD5kjMjq/Z2fTQl7yXVsXVaOQ4VSkfH+eytcaXKuIxXxKTOzhnERivFcPmm2J5I9IQ0EVsYFqVNjXu3MBIEUXrQMDLNpxQgW0Epu8gFxnkEDIc1hGFxZnVVKbVgCrwl4a6lH+YG5Om6qL6tG83fgxKJrN4mRoTxvx2xUzjC0vLOTefltrH0nS028tkUd/k0s2a766eV4NXvplNc9ens9p9err0UbXi7ZwoP1Tx8b7gxonQ/LqdgdCnDnUOmdnMKtGpFeIBaqNQFgiqlj00dhYCF1dliI6vdL15d+l3uutgxYLB6CaUOyMQ0BNA7JvrQbwQHgaQDACENynhgExAIhTfEE2NROgRk8J7vWdH2NiSPiHEBDwHAxAogYdmNKbzeKE0cCsGClVKYlkLg6sluQu4lXqg7zjBt5SHGd6l4lrL27GoDWmdZ55xA8jjFttSEflP4Jn24eTlCW62Hz1jsEWm1YZ1g1o6ruWwCu/udzXuDaEhZ7cLea/bbe4sneg3V2neTIaEsbzKKidbSUULdJiI1GBSvBK/wrplUd19m4Ii61xoZxNaWUni50xnakGSMIgj6UhLOzZVEImYUp23K8pWRbya4V45xckCnNlhHjjBAHsHoZOhakg5ktRfY2wJWwN5q3gem+Mx6MHWlzFUZ1JTpB2+bgE9oXlro/S23ZFwBg1vF7SxlV8sV9Qz1A6nkletPIqTN3tJ+iL4+rT2d33CcV9nX0jgl3ELS2VscK0GwErEOdPW4jvEC4zwsqmIvAnt7qpnpzvXU/ka1e8ntNNFD/+loXKgEDwCwgQVmgMjNmnkFMiBwxc5rtggkyewWDxiHJ9d0IIiCMmXExx3SO5FmQYNCWW+GION1gnmfESEAImK4uADCGi136TuYdSGxpBF+2mJ7HhBh15LNHb/320rmaRSUX/eoFVHE2iWXx6khu+rmOGxa+AWtPYROn2qbWGNep48i2369diLIu1utGvV+zF+wp4Q6CFrDOqKh7/mEB1217BZ5qb6q7WoeU9yib1obQs16R+19f9kyLnFOAF5B7S+BcqmqfAms9xuJ65XLg1SliZb6wMrPk2QZ1D47JWUL/TAokoElzNTlblwCDFxZUbYKUaemSGGWdLAbHGTwPaR8HqP1LAdXIQyZg+Vyn/CUtouasxZQOQip4y7vKM3XlSqtQQFWAlHplUsErabFAWJnFAsPpPVcRwz1fPq6Z0CGhYDYLYW3mjLWwnr79KrsQZtPSttBPo/hOy5NHhTsKWseER8BVbpkinZL8I6iNg8Kal6YfJVSPtzqkT5oxQWxalYrQg0GQEyFFxyAfdcu8pDwiDaKo3uaYJxVNQKJrDod0VyREIlHjDWbfUuCiQGlA8CiDgq8HEEeEEMAhZDADyerHwEDJrpWmdZqBeQZHgANhvr4GCBivd6neLnapPCEAIbm/J6FPYBafQlYHDcogDIcHKrxJKokyfBKl60F5IGltqXo0OrjZooysBaXWuYc8j1S+oKczpFPDkg2pExMAHxD/3OXZ8FW1PaWjwx0HraU+sBxRfZ6q8yupn8qwzowMTXkWrq+NzKcq/lL6Pmydn20pze44LPQ/Ber+LpGBVQ1EOoNFtgv5+9fStzrxgNUBLo1KSMuI+L2eL9Kux6lIHgpWGifzEaTznPJmicsibJP5JiERCeMqmBYht2utKsUOPcUpTxYGoh6FiWnN6TjOxubMqCTABWEtSltsbjs4wDK8doXywGUFFNTQG/QZtSuiU0lZPXYAyqWYouU5CZvbzCZGmZEt2JTqsGQ7P3bOwtKmtcy0Sq9BWoy/7gXp66h3PadTApdctzdjOa494cr1/eGOg9Z9eNxD3YdrhP/B95f2jKX067xsEDIBJE5y3hdCmVXaB/PUy4zLq9NiulFWEp7nKPMBppkpOM5JE6cl1RWEKTGoGCdwDIiitosiRCik0qT5AgfQEEA8IAzJmYJCZlgSA5HSkOG0cRp0HCMQgXhzAwJjvtoBkTHvLtIs8UjzIKrtysZtUVqmkkxNZEYvKMQCWWVnKyULC0zHBBpShdoEru6lqJK3nPuw98bqt3lAeNQU6+DguwaZEaXgmeW589t+XHaHTwv3oLUnbNUJH+uMcbAn3cbQ8y46KP0Nz3Osl+DS9UU7HKSpu8s9kWRi0nqf+ZjcCZvpQuVplQ46rIhcQmlyWxZhze6eNA6KCMnFHG5PaoLmnKPcGqVMYF0xONrs6/YwQZhIFFaiE4yKOtE8GynFpZDiqgdhVgu29ahb0N+MvH5WjOA5gqcZHNK4rRhCYl+sMJzAg0BShAxatrpxIUD18a0y7YWlNSRDZph5EBZAnO1UDrCI6jdIFif90PTZHWc7JxfHEkWZhLz3Q7/nJYZ16vjNXD49n8sLUHGcgSvX0Tpj7J/vqwJpMf/6q1J73rmA6/EHLdO90HKt35HQCFgXbtu5o85rX1keRjmOZ1rrx929yVgSQcZQZ4S0ZEgCqYEIu0EWaSQHWiqI1ZkBjIgEPhMDM+s6VxE0pb2pAClNv4RxBImXHw8BcZ4Q5+QZyFJIAsBDmhGDhiBzByabFoUg0zqJLYuAAcBIwEzACMIMwqRegFNiftPVFTjOuBlH8DQBc0yzcdAAogDQAKLEvoIYphQMFZAZMk2UsS4CCxjTICxrGJJqEyx2M2VxTihKZ6HPtJbeYv22l/b6rqlQbZ4cHhFz88B1jidZHKe1auM6r6R4/EELQNE9fpi5HsmwtqbXXF+572TgWrFpabCZmTtl2Ko/98fao128Z6Gsxfm6N2sdaz9zg/bS3W2d9K0PZHdgn4yT/pLMvk5p5vUApOmVkIChGJwszxyQmJaqAYkZISb1HMvYKUYiB8qeMMNmsECMomLkghEQ5bW0CqYl5WzYFbLKUAcdM9hsVzzPiIEQpwkzEYYwAJFBg7BO8aMnhMJ+l+CGzYWf1JVfq48FjBGAkNkNhQhmILBMTyVMoYai4t1RJZaNYfljZXbsGFfx1lNUx7SKV73gTti0H2P1cqxqzjMB11ZG1zIuFPGOyW/JZtba0G5PIN8R0ErhXPTzcQq3OS7s0KZ37Ce5rx94KtNay2udaalwk94yqdqO7c5AhBCAiyHNA3gxDBiEuSTbkQMkpDSYgZlCGsjLwBzTSsLTPIOmmwRc05wYioANdjMwDojXF4ghIE4T4jAkdZ0DKlsLS9mVMK1AAUyEISSWo3Y3tWmNRIiBMM3pMeOUnC5mAuI04YqBYRwRL28wDCMud5dAGDDuWGae1wUqkzp0FoaFmGx2FOcMXIxynFkI4HEGhQFAYnIJDIEgNrNc4ykEVUDSVsvNUq+jimXv2gGXqmAfK1tXP5x7XNcW5qUqwnOEOwVaAKw3WZ57NEB22zaxtftP1qUfE//YZ96TGy0elL3u3pVsd9I6yUxK7+syLTvgMqLss+eUJq/rXBGGISS1mzzVKAwrWMLq1QhEEGZOBQuykbGoWYQlZPb15JSQvfpitm/Zg0E/glWmpZY2Y1hESdUJN/tGBHieAQbmaQIzYwgDKDKYRmAg0BBBIIQQ0/069knSiKwu9REQ0GLmvFqy2OXMQSMG8d6LIJvzUCxolNuK8q9k25IYXYbl3qO+vAWWZEyQqGBIHsjKxpKP67a4OFO9ZbUGHPu/pTXGVcfz+R4iE5YY3FJ69bEB1pnE8J0Dre2hFvSPqBgnhkcFyGcPVed3L5CcL5sDtO+OXRnTigJWMin6OGAXBlyOY2JanMBgjEloB4g6jyKYE1hFhgj1FA8xYpgSYOH6BuCYPAqJEOcJwzhiurzCACBeX2EWtV2a/T0DFIRxmWMGkYGUKPIcWKVtJCASJaHAjDgnYInMACUApRAQIiMOIy5pSOanYcQIwhhTPQSR7UnNmNKamRHEZX4W8EpzKcLGkKUJetNClsM4JrZFIYH4UM6q4N/KOsOqG8FGxiTxjDnU9619e/X9DzncVr7tN1N75J447+dKeFGCVhb00tvSzthDlv/n8jpc80J8FB/MUtjLCuVrKN9O/uH7xvVHsc7TtAftOt3oMK2qo15kw/mwXINLmVYomZZoEweKwmhS5qWai8WOJEwrpt8xMjCrcE/rXCGISm2aEMcxefPZGKq8JGMqlNaWE/XKuhZqSdmg2qUIbG75ABCn5JEYpzkJpjmCQkSIrH4TthnX4rR2F7N6RzJ0rkNliCygrMMKwjwjkowT4wibairzKqlHGZBcPIhS4fp9uqf2jCmhalFvDOS5FR3TYn+fbyON1qDM52wDfhfubxhe16bl4h8iF9h9s56FugzzWMj+cfnlnh5ehKD1eDOsh6n67KnObjvUgLX0eGu9695t1it3N2/tpbOo8NShQIVwKmMCrN1uxG4YcLHbYQCwm1OcYSJQjMlpgUmTSUJcbFlgzkxrTuq/ME0yoHdKvG4K4HHAzW4HYsbN889jIGB64gkQGFFWHB70wUw12K8QBREbt0UhMS25T2fomKeU/xwTsxwQQGMEXc4IQ8QIYCTCBQgDZFA1JU9EmZUQREjreCExrjjP5lAyC1BgnpJtixlh3iW2FZJXISg5tYDyLBb9oeIbWmzFmOq5NQ5mblvzexxtYluq09VY/Qb0mIgs5qky5UUGWksMaz8QnDrm6FzAsuQluObld+78t4QtvTjPZmpWtZY389o0Ptq7048pxyeXTyvmllLk4pfPW21Hyaal3oOEwDGxJ2Kkgbez3c9IardQMS1y+x4rieCkDpxuxOVdmdYAG7N1oHLVsyybxcMKytkGJYwoOX4EUMygNyC50Q9EBlppxnnGTMAg7zgY8HNexwsQcGMgBMRhAIgQzUuSzbuw8JLc98YWxm9lL1N9kelHnjmDF9Putf8+x8qMxPI0wMotp/F0Bfo/DsW5Kn5FlNz5+ptay6hiTPa46blyF0LqkvRbu2daR4YaqB5hUY4Ij8yZ5GGk62YzPyiBF0Kn1QArYBhHjOOIUZjWSBEUI8LMiXOENHBWJmmCrGqf1GvMGDgxkYEjYowI84wQZ2Cakl0IETQPuHl+BwJjev55TIEwXV+BCLiYLsDDkGxEMQmhNKkuZ2SCME11lHDKSkJp4woQJ4o55R+j2NaCzMIRZwzMGAnYEWEXAkbxVgRRuh8BHJNX4QURiIJANxL7nGdz1ogTmXPGECPmi0sMIYCHMY3fCkAagVwCUtsUDmVc692f3tW1HHwnaT2Xxyv0Qb0y+cExrTPmfcdBq89IHjXDOlTtduw4rH78wz+/TblWncL93oB1z2vPvGXWGeXiGAxzDCs+nLq32+111n1Cl/5SOXpBv0a1xzgX8zS1kzAsZR7IcXMS4hXH8pvbTWdWjzyDWJnWkJjWNMkijjI3IAUURjt7OjcrfPW0xkXleeyxjO0pKxKIE2ZH4kASQOY6HxzTUg88nTh4EHVhUkeyPa8xLkrX4zQByuj8mDRRDW5iWsXTeXZT3rTISZ0E3ve9NiVxhSkAi93xQoG9o2rveK1J1vXRpOdUeMeERpyohqQilFkW1F/ZecKdAa0kO1p1QAtUD7dcvXAMYJ03d28J6P1aCPLVFZ3T6qblXmmdTwkvtb9Rr2faS7snngnU3NeWcmsQJkJwq2SQHItHngDWEMa0DbskmCOBKCIMMllhmCsVjEgycTYwZwxmDEgbc8Sgk9XOE5gI89XzmMCYnr/CFALmqysEIsSbm+T0gFH0j3Ife06Vxn1FSrNgzJSP8xy2acaPBCqJDaVxYxMiAzyOAAghRgyI2AVgDAG7YcA4DOL1l+xYkZMjQAAwxyENqI4JfCdMAJDGpolX4SxVMkwz5ssHaRzaOOb5CQPZe6jfc9smlhoqNzF7oVBH8gZ/uE5/SwHLv+41xrWl3e8t94HpHxKoOugCcJWBdcyOzLMOdwa0PEi13oHnYViHnj81bGVYte1qOX7uc+X/vePOLRKB4YCf2oj9XunCe4AbvEl94DLAqIrSdrPJ3VeC8iKA9eRYdZ38dVKBSa1eBJlpkQ7u1bbY3aqMldGwjknK9i4FD55ngJJrOM+z2bT0GOpFKN6KGabkF8nYKN2QlYNcvVNym9m1bM7DaOxPnjypE4N6UCZ2mTz/CLM880CpbAMpMwMmb9+ak1qUhjQtVPQMslo9uWbrrSBOLSvPTVjSDvLUBw5YnAEovWpyFGIhLPSklhhaPStLfaN24son2R48s0Pn3iUI3yrG7NNjLlSChe2MOmU+k5i8Q6D1wgqnAtlJ929o5TXuLEZ3F5iP6aXVgNUXNjndppuW8l4q47HdxjqxXve2d4uLR0jPE8QJg0JIq/oOQwKcmIQdBXHACHkyWQbyoGBOru0kgGWMC2lsl6rpdOaL+eYGcwiYr68xjyPmmxvEIYDnKfusg6Du8pktJkliA5UD0lYgFAxoCTophZQxzskFfp4RhzS7RQCnMWrq6j8kR5T0ngfESODAmAFcDDMCBkzTAISIa0qAllz9GRBvSXV3n6+vMQ0Dht0ORGlGDxopDQBD23aWgGuxgay0nV6TOJah+BS2pbP4RWzO6hhT8VqJjr6ROr+PDHcatFq5fxzDWk5/P/PaOh5iqRxHA9cKAln7IZTH+/p/rASjmqJUe5Va5rJDW2S06P1YxO8Imj3AZcxnY2CXnk97DegVsApzkdc5O1ZFgVJ6ClLigJFmbWjzyV5/bi9Mi4BsV/KMJ0bE2a93FUtGgrzmln/ONN2Tsi1Xnhq49LH8adZ5BD3LykuzZLaVHC6CjC8LgcAsUzwJ05qRBzhr2mBOs8oPicnFKTPJOLspoDQ+5QLX5Nfq1gl+/8qqalkM5G6yJrLwPTOqxMtUoBxLv6ElqVCXyfItqEx5vZubvL8cXb9/Lu7r5XdIqIu1UMyzhTsNWo9DeNQzWmzO3YSZU+L1OrLNcSXYV7OR9LcWbElaLWdwZCARrHVCIjjFrmUOGQzQICqqYQbA4CFRG1bGw2lOPt3YMy7WGeThbEwym3uc09it+QZxHsUxIzlkYBigi0jamzJ0IeceCPutIGaPEggUE4PMg35hhA2qHhQ1JilYGdMaRD04pPcpDhsxzAAPGENADGk/h4BRGBfJ1FXzREBk3FxfJ1XhbpdyGkYMYUCIEQhDfjPUqoBPlpde77V2rclwoQRGCfek3cuuSb68f+152wHG++84S/1V6Z0z3IMW1hnT2vlDvQwPvX6uUBOMBkKaYpTMLPdayQCrGRbU02d74Kqyaj8lPztBt1BlGgtMa/WjayjbQjyT+mVJzRpnYOzBWdyylXUJ0/L2rLQAsWNP7JlRWXrPdLRQLKxHVxPWmdSNaZkatsy33Hsq5Z9Ff2bAqmdFAPLg6vzo6pSSx60xCCEm8Au2Bfe7HhcGcGQAEfM8Yxa2Nc8TBmWTDJkXUIE527GzqWq/48R69wnddmV14K/5b4ClEM1HgZL6FPeXFIWq40YJUN1fZ1l8pwCauQElxX1Nfm/tbATc2rfgXOEOg1ZZYbXcfNQM59zhNp6nbrwFy6Iygn6LFeJo4Q4vH633BtcA7dTeouZhixk2x/oXABIHDN1DBSsBYU6VYwOgMkBkh4g8uWyhLlSVoQesyKYejDFN5RTnqVANAtIhVwAhAg0BIQYE2XsX/ShTUPFAoEGeStR6iXnB0rWBxlZP2a6nbIsoCG4n+xTAmGPAwIxBxnMlpjVgDAPmMMuEvUCMab7F6eYmle/6Ks1Of3kDGsc0vkvtfJQbYqfvdFbGsBhWM5IIS0xr7XhrGTYW75S5AQt50DC4hxPuIGht9x481YZ1qFff1vuOLdeh95WTXDqdfLkr87LeG+AZVd/BquwprI8b84Jnob/nerTUULoyVWVu24LvKueSELwNrwSw7B2o8TMYpUhyPRA41teQPfYKkPKlyCWj4myKa6sXRwWRiqWRc8m3sjpG5c7RypZ7/W19GphbfFWTJpue7e06BEiDY1uVfQtIk+uKPWue5zxDBjtwdi7p2h6UeexnDBlyDw51O1YPpapGyrzINUph4XZ7h3l5rURtw6pYb9ly/VfEFr8E9HZYSK/0zZyFute2oHbahwxcdwe0khRpT99xhnVsWNDqbQ9V940IKD+h8vdyvStcVtRt4bM69PM4vFfZASYvbPx52YcQkmD2Lu8ASMYrIQTQwGmmimKjYhMoErYlg3k5T+Gk4GUgpoxHBv7qMbR8ApghLaucWFAcMI5jmuNwGMDjgHFMdrAwDGAwhiHZi6ZB1+WSNCRjg0YFXrGRGXMbgqxAnMoROdn0wkyIXC7hMsYBwzBgGIdkD5sDZo6IYMzTDKYb3FzdgCngYppszFryZpQp5U3IL6u9Dm7fhwRlRvUHlSPAmBYqhtJlXtX91fX68hEFPjmFRxnuDmgVvfqaWel+PyM51oZVH2/teTx0t/i6g1hdanThe/Kp1fblIVV7d19dd4UiRz9OD2RVWOkg15d5KZ06PvszFeNT4MoPUT6PMRG3SSeKBLRYZnSAMDNLx7Mvx6QUwArfj6rgLEyDkVV2/tlJy6COFSEgBrU7BVlpWe1PweIVLCuUbKxXdw7Ki3tahqV8NAO+DRmgPEs+RUqqP2bEGDHHGUHGcCk4F96WQq2K99y0Ta7es0bi6p0ut5WlK3nZEpeke821DSkzFIlXgC6QJ+3LnRPf5rQDw9X7aFWKpRwsefwem9aChmhNZXmsHDw03AHQovwRuBexFahWUz8zMzvZnf2Wwvl6ousf//J9+jF1gKOKVZ7dku4h5ci/SviGlSuRGBG0JvTdpsIIUF9wgMiNkcrMpWQw5Z/10EmEvFdXeqYlwpsZeQVjCqAQU3kGtnFkwziAY0QYB4Q4YhhHMAvDYkYYE9MahHmFIXnsqf0rjwUrK4qGZDejYUjxIEwrpvRs4LGOafNju+Jgv3WcWowz4gTMYtuapynNChJn2Bg0x7DqN95eqcN5+ZcRopUmp0wrx69udLt0Qz+9Y1r2huK94MNjD1q93v2hNqxDmdSp59fKc9s64nZ9HT1PhSZCZxPYXx5y5KH3PGtCwTKHqghJJHgGDAdfkpdqY3rzD1YWAC/mm1IdJrII5CmmMhAdTFtsyOyEAhAiXEUV9wsiJVBryl+WmXxNO3aljKuIT1KbjjUlZuXmSJRFIj3oBgrgENN5dm78xrhyFaTykrjxkz1zSj8istwry5qQgq/tPduSMlE0wc4yebDas9RRRUFaGZd5ESqvobwulobb+qqyjUfxp1QV1pzGMy0id2Nt+9KWK42+u4Lw2nNpk3PJGidV2+Hi8xymMXpYtq3HHrQAuG+ezLvYLq0wmmMB666G1qFirTfKBnan9FxdlwPowsy2UH82W+cfWCpP3S9lKNHI3nJB2cUQhGmomArp9kAAh8SuCg9CZJYlWwY8/Z2dKaDqNQdwAMyTT7lZLrrEl2mlQhiAITOnMAwI44gwCNMaBwCyJ8YQR4AIw25ARGJgAQwKAygEG5xsm6x9RbIOVgQQKKYVi8EIUwBFz7TSlphWwDgGjDykGd7lzSXASq7vs0wQHMUpAyHN6ZjFfVYG1y3xYTKMNecEz7TkjKkQC8DiE0u88BHdM61HHZoX0+rez33c5HiiLrfuqZwLKHM6hzZRjV/fv1QuD1yadxVjtQiZcdlnxbmXyp2i1GpCrs5b2Tp83JKpTRp14CwOU69VWRYKZkUCQkQusYZ5dZ7Zs619UgaZPVhUdvXKZW2Q5M+yLzwFAxlwkDsOIYBVfccJUHhgx8KcJ2ENWP5ZRR1KkbODhjI2BeGiesrxW2oLS/MbwlYwtr381gHYnnHUgNW0G1+nTQetU+kbAldtqB4X1cZvx02pdkEBK79gBzE5g81l8/cV5QPyLDauXOVtVX7H5n/mEM6d4Dve8Q589Vd/NV7ykpfgFa94Bb7lW74FH/nIR4o4b3jDGwrVABHh+77v+07K1779F0DoOi5Yw6HFOI97OGtb5qp+FtLuApf7vRUrDwoJETJbGEJpr6lAohiXFbwqLZUvIjMtm16pwDEHjilnrzBNa1E1ZSQDq0L1Z1tihmFM3nvBb+OYvPlkC7oP6boCdARj5pi2OCOCM1sMwjh1c3WgQOXBVD0Uh5Dd4MEw1aC6vivTUrsWnFdl8FW26T1qzHr/iEKvUwNk0LHT/fMvZImS2+1KB21DODvT+sAHPoC3vvWt+Oqv/mpM04S//bf/Nr7hG74Bv/7rv46nnnrK4n3v934v3v72t9vxk08+eXSeS4//sBlWoWs+ArhO1Q1vZWqPEi/3qk3c/65Cx2lQijTl/yKAkU+1EgCWptRdU7yyDKqCK5hLwbQ0swxMVgBX72qJMfVgWeIyMuVdxnLKJp2myFUPruog5jW+MtMqgY3AwrQic5qoVhmQA1sdDJ2Ay3s6UlmOhnFWwtZOZ/uWb6I2mNlWcM5ehEU8ycJxY8e8xMblm1dZirbeq7B5Zd/qA1tiMM33araw6hsQnb09j47z0vOOOVH5cGU5cgHSbsGm1T6WK9/RgTq/jgtnB633vve9xfFP//RP4xWveAU+9KEP4Wu/9mvt/JNPPolXvepV58vY9eLuw/HhUdffVquYV02h89v25AVXJx3qAZULDrNUOHh1nQp6CAiYqs86lJQogA4u1jFZgdI6VhCmhfJ3Xk5kuQZsnSxRE+qiJg1CEiF5EuYyZLtSYk8co6kHg4zTmoeAQWxZg9jAaI5W1zfzhOvpBtc3E8abG0xxFsYVERCad5O3PAMIw5BOi2nVJLpBmf1DBhnropdRZhkBFxgdUb7v3r7/km8vrNm49twJP7YkVUl24kiaQ9FxPmKV3b5AnSZ5Sji7erAOv/d7vwcAePnLX16c/5mf+Rl89md/Nr7kS74Ezz77LP7gD/5gMY2rqys899xzxVYEKg+sF7wQ6utbj+ttKf6pYSm/tXLsSRG5XrRn125Z61Cnu7HFUf3jsDrpfXZeELMBgqMeQKFKY7CsG8XywbPZo1K8dN7iQPcubb9VZStUjtpJMtaVevQNcAHFb0uLyg1U5VBXX6c6F2VV8Qy5EA1DNKbo3fVLtaLZmRyjZDBm5rRwoy7g6L37DJ46ZfZPWUUoS2u9BYBZJhXmPFbLQ7TV/2JVFelLiy++ifx74eaNIVd9/o72yQ3/DNnmV5fLOeEgx3Mf7eJTrhb0TEHrr36+c3cKbtURI8aIH/iBH8Cf+TN/Bl/yJV9i57/9278dr33ta/Ga17wGH/7wh/FDP/RD+MhHPoKf/dmf7abzjne8Az/2Yz/WnPc9p7WqedQ2pEefv/2C9k4LgXdq8Zr7j0xQXmrNgMy1vZZKDggWmQm1Pe6lfOtTPmSPLxH6ugSHCXuhCXqjFsbKSo33oPfAi8WeS+HuClOzzDxfYf/BCHBjt5wKztSBA5hKpsUAaJYxU+ZtmLz/IIumzPOMm2nC9TRhnG4Sy+Lo2KKUrypzyYqzz6N9w4UsVTam8y1ONt9iUg+yVXlAybS2h06LeEGyl+rJ5FCZXJfRHdJ/PPiZezXdntsAnweFWwWtt771rfi1X/s1/OIv/mJx/i1veYv9/tIv/VK8+tWvxtd93dfhYx/7GL7wC7+wSefZZ5/FM888Y8fPPfccnn76aQDbKoNcb4R65zceL6Z74vlTw1YbVoon3nTc6WRVTIDqmvU3LDXuxTSWPwbrjUq0BFjU+mJIhDIlz0o6vzV/qmwcdlkLLPYEJn9XE+rhFAmnSq+6ggl6KdzUdx1PS127rhdPWpbH7Q3cOvkUYEAw8PKOEMVeQY25PA7BHDyTA4bMxj7PySGEPQy15fPPwNVR9bjVNdWBJpWh/tbHSUuYZVO/jluz9+3r0E7IVvcKXL1B0iqOu28i39+sq0UuX5cO1+majUtvL9uiPZFbtK4GLCLqtp0iGytoJ/8ecDWPq+Vi+V+BVJNRrutzSMBbA623ve1t+Pmf/3n8wi/8Aj73cz93b9zXve51AICPfvSjXdC6vLzE5eXlpnz3ye/bgYwXZmi1BlWXa68hZ0+C/viInmjdGywAq/hRhp4Q7Mag+mpG537SFfeqO4oEU0Vl53n5TxA1mq5crEI/Y1USMrJzgh/OvqSff/ICZMxI++jYlmdhAUBkFADBPr/6eXJhAUqu5UzZxb3Yq6s7c967cV1DjJhjynOOUexaE8ZpwhRnKRNk9vqUv3lHVn/5TZVHWmw/lC2Blg40TtM7aXwVhw6GAKmnbnsh6rxjLIfmeq+RLIfb4msGyLr331b1uffuW/yG99VHv0L33FD2D84Rzg5azIzv//7vx7/+1/8a/+E//Ad8wRd8weo9v/qrvwoAePWrX32WMpgKp1/CPdduJxzLsE5lZiXDArywBQ75mPQ+15tsRyGXx/oOrDOXe5fd56rOcefXQrGqvWboAYZbrLYOq/iV1YzLsme036QsylizFcdiShDVDCsAM7uEpppnezDhX7OWqiftZU7BZDrP43jIMstycw8mUA3muBHMe5ABBqIyrThjmgVMuARRZWUeyGqmXAOWD8G3OWNcyrQcYJEDLc711Aj2iu1k9qOHmp+vWIJ7TS7FMtRD4hs8qJNujusPs18n/jYCzGvQvArriFVR7b6aWVnC5Dpr1Y1FsfQLW5CptyRmzw5ab33rW/Ge97wH/+bf/Bu85CUvwcc//nEAwEtf+lI88cQT+NjHPob3vOc9+KZv+iZ81md9Fj784Q/jB3/wB/G1X/u1+LIv+7IzlWJfbT1cwHrhhrXupY+HPkCVurIy/qlhS+93T/HypbZH7MlHUjF50dZ+66V1ycR+A1pQ4NK79SBQ8h4MhDSlUzvDe7Op4CC/VwDO7u4MtwZXluLVM1O+X9geyK395fawazGzYA9ubsqZKUZgnnEzz9jNU5qdXdhiqrdU9jwuTfIizcux01Af65RTTlayzEcos2Rkl/H82FpVvq9R7xd6O3XFPfqw/ABVhIV7bz3/7QmdRzGYwtlB6yd+4icAAG94wxuK8+9+97vxXd/1Xbi4uMD73vc+vPOd78SnP/1pPP3003jzm9+MH/7hHz53UR55eNTOF3XLbTqXtW596f5W14h6vidy14sG2ujrWyjpHXt21D6J5pcPuD5nPyVlN7MAFcmn8wZc5O+ViBXL8k/YMK26sOQSLKRrxw5mwJTBornPFUfd3ZWPlUyrrqwSz1S1mT3oyMpEmid8ebXM6ekZyYsvitegAZYwoLZTXubRPn9+XiunnPPrjtk4LS6eFsow/AwnXcJiadVtumph3Xt6ldu7vhLq+PuOu8ChZVcamx/a40qxXtsR5eEqX9o7vch6OJc0vBX14L7w9NNP4wMf+MC5s70Pjzo0gHULgE2dVJ1AzjanRmTa7xKw0sEy01JVXMm0qqzXH9VUbQTI/IRpLShkdiFzGHIcZJZ1xjzPIHCxxIcXSMqwoiwEWeRXF8vArx1YTDrtkn91BBkKAAdC6TkUIucYwfOMSbYY3QBjojSFEyPNxwiWVY3TuLCBI8I8IHDySgycZugYZB84IMxpKih9D7pUidq08grPFba6x+Dixx0JzfMc84AnVMiB2fWH2BwfHv+5B48I5/YK3Hq9CZUq+NEzsyrsKU6vrN52ZmLTiE7dk93Y6u22rHenCpS089kFrCIfx5Jcr7m0fvgnYFBxDqh8s/YCF9fXHXMq1Ik6m4ZMAxWI3ByAFQtC7kEzUM567stl9c1FIUj+W3E0bbhylTwTXNRXDlHHT0Uu7VkVM89MVNSQg06c69WFDkjrzQojIOW8B5uqr9gGSZkW2Uwd9jXLxe+hw0C4/UZWBxgvMjpHH1NCfcblM++GMyC3Mq5HKKoeb9DyXSlTrSxEPTfgnCPcYpbt4xya2WHxa2eP23q4FkQAJ1ol1EzLA1j9gdfp6/+F+eGdCi17PhY9jzKxSvWnQhtzYjcYdLVfmZmCI8I4JhCehsRUdALZQFZ8hpuXL+o0SlmQ++BrJ4EKsnefK+xeu4M9gvwQz8XIqiIU1ucErAGhrOQcxsSshkmec0qMa5gHRI4YxoDIspIxM0KYEULMY9kYiJExR2F5qipcEMb6/h4J0TpH868eIIs7csCl3S7fXo97Yk1nrRyPmrk+3qAF5N5rPtGJQsX+2OvLRXhhMaT8PHam2q+1uHXgWa6rKovKprUpkN+1DKu7dz3oHuPy607tBy4nIVDWWMlB6jbjrtopSr1guWT2GYJzj89u8Ek9mGdZN7DyNiALzsvQ7Dz+cg+8loQ7FUWun9Qey+xdwqYESdIEFe2yKFqOepyXMq3gGZc5hJDNyhH0t/x578qWXSqaN+qLfNm/G6NfL6xvF0Db6KqVjPPcg5AOTFahpvtUo8CWzP7s5IuR1Z8X50SsylPX3P16WkcE7dgedM+JgPTCA6zmzKMoxvGhkvv5twcqm9ip2sOu6X1JtVUCl/+4up+ZFxSswp4R4dyqgTSRLEoVlpXDIgqjYpblOoKsRzUnO9boGNYwpHFR45jkwyCDe2WQb2IveUb3mSMoJnvSPEfzrFMAIX1Ah98q+IvZN5ybulel1SCUVXjJu1BnU2JwOeuF03oQEWgYEgiJLWuYZP2uSRiW7Md5BIMxzmlmjWFO48Iiq0qXy1nfbXFI7bSUdGC/CH3MvgsXbDxW0SnVTlkmYuhcXkkZgBug/JBA6NBwZ0CrEHIbgWSJYa0dL6VzbDiW4S2nZ7+66RTjrbB/7Z/tedZ5SPpcXl+eid59IDVhKdR7HpwSQAUARIygz+uSMXntAUuBiLNZoMjb9Sh9B73e8oMqE9KLlK+RSBFhVWxTPYWGYeWJa9OEuuookZmWf3rkGSg45nn5OAuvQuSoUFdG5NSJXQHV6/w41agHhYJpeVWoqyPvMm/LuDjGFaJfNkVXMJb40WESp+mcOLqpqyx/t8aXAW8JXkZilCm2V6TIh09yu/Qd9VYc3he/AZjqe2hnvsgs0zSH7vEyI1uRYw7syw4f1RG75TzYhndkuDOgdWh4oTGkOhxbPKp+rSWzqMfeU55brboGsDRk8UOIAlbpMxxgjnggUJpkgmFPZlt0s4tzltPRHkpyU2ZECQajavgYpqoiEAJIALNQsJWPUti00sliEUbvmOD2YJk5I0Yb55VsO9nNPcYZcwzCtGZziOAYwSFK6dj5LYgqT+N50ANb2uVsgLCymz0Lzr0ftQs7HNBKfUlFhzCkRSXHAQFi09KZNtS2BcYwTRiYMYxpPyOm8cwEK7P3IowcEZiLeRw3N7I9kY8BrrMG6fPsuVxyS6qmdNIvoKFeq9nm9PSkMLtDJ9KBpEP+GzgxvChAa4lpPCqG1dqa+ukeD1xO2Phc6gQp68M19No2lcl1y1onrB/TEuOy/DrGBsvPXxFBTQJGadpWGMMKSI2ZSEAL0IkbbIsilI0ZCHgFoJxBPkYpeBbdzjRVMi2UnnhFjVu3N1dGOTZJBxo7u44N7pUPPLjKkCJpDzjGCBKhHTnPsJ4lt9cLCth5hpXmgjJA67Gt1hPQtU9fLmj1ZeDS92agpjNuOMZV74M6pchSKWk1Y8C7W5R2rQzIDBaG1QGvpp1SKfEtGmmNAQ4A9oW6jbfXycrdO25vkJ2UrwaKXEZ1vhCPSQ9YpM4ZqkrMX3nVJWkytmph1cIYVZPyMIoEVp7bp+5lwrHhDoFWdpHdG+uOMiyXwp6jXuySaR3YKVsIpZrh0PwBLwuFYQnaJMBik+cDgJHSdD9jCAgABtKUo6nRGMAswm2akypt5miqQoJfUVjsNoMKhJRnkE5rUJZFblPo98KdfW3ALaFCSV04ZCFuE+0JGrOes8kGkb3+Iss+zUgxz7GwaSmIBVl5SFVpfjFFtX1Fnc/Psa4MBCgFkmNUFIAQAZAAi81y72a7h3xvqq+iEpy91+TAyrg4eQ/GmPaDdFB0HS/3HGbTUtVo4/pN1V6ewfeTOsBll6TsLwTGtYE+5ugV06rLv6ZdKTq5+u40mO6RyxtWGGtm6aeHOwNapfqqXzlbGdbSfcdeX76vOOqcp068TSkXKTblcz01f91Pi7Ml/a3l6JZBQmZ6/qPKtyrDUfpAIkmVaSXAAAYiA69BPktCACMiEiFyAjplWSqswSnNzLSCAQdYAatWEcJUgv4PjnXVtcD+wG8hsxA4+xUUoFTlpb15qTNlF+SFd3SzRbBnsQ6FCqYVHaCpgnDtfZL12kn+1eOq7JwW2J0rgU+8BAXoGsYlqyiHmN3mlQEZswI7wIWrqxKwTNXlWWKhW8v1a08rAnpxvsy6dmotQgV0qzatuvI9k6FeRC10+x3XTKuxY2/5hr1xzFdYA2R1+fqheC0nhDsDWmvhUGC5bUa2BFj7zx2Q/kl370vxluvF7YP9li4/K9PSySUIIxF2IWAMATsx3o9AihsjGEFmTCdMnBYu5MigyGk1XGaAZRUmFaqqsgOykGVleOQAU1SU9UNs+ojVEy8BlgenLHQVuTNsA5k5cpyBmWxpkDlGAzBlWgSnOisYlhf2ye5l2NYo19pOHgVObvkUbKYLv4ikIJpoWZWiJhf+gm3JDCGBB4QgzCrIzBlhwDBEhMgIHMWSicy4IiPOaWVjc0LxbItyedntG4a1xBQ803iUTOvIkMdvbbNbN/cTZT8MXw8H1kfWgLk2fUJ4UYDWoQzLx9/3ak5lWMv398u7OX3Lp8+wUF1vGddCgpUQXS9B/Xt/TH/O5pBDuzeSIiq6QbaRgvyG6d05JmE3O3WXAhXFmOLFaHmzOGEQI+m/HGtIG5efXv0AXO3bg5ZtFfWquShrINtrSuqyjsiIwTOtaA4WOjefMpOCYXn9n83jt6aBatuSCqNg4OtnthDQN9Vgbc9rGZravIIfp6XpRmTWbUBbjdnyKjAS1W7FsNYez8Ysmd2GjHGtha1yfNWmVRapqXrzAlywkRWMC5THX5kta5tcscfuAZYysE75Dn6gA8OdB61TCdO5ecX5VG8POxwKWKcBbvotH5nr+CvTGSjZsMYQsBsGjCHgYhjMIUNd5pgIM2SJC4ap03iOoGkGcQTmpBozL8JhQAwx3RMICEOSswwRyJkJZiAT1WVClVRYVdOV8jSlw20NFVjSqRuWfyx5REpln+eIOGfgYnk+DtpJdnYgbw9idc5QVSOkjqh5F733pNMyBZk3MXiWRb69SHe9x7S8fas+RwFDCAiUhgFAx2MxTBVqatEYQWGwLNntjxECS+TrYYUXlBQwZrVGTZduP+/T3GnQUs+XrXW2XrmU35kL+RS5M8vpL2WT819ggCvpLzf1NbaTe/H7ylfYA2rW1j0id6aMUQzw7XRkcwrl6JIkLHMPfwhpUwAbiDAAoJg+tCj6/Bk6kBZpCXlRH0KcGaAqNGFaYWBwHBJEEgHs2ZaTyf5dKGAVzKW+DpdKZpROA9o6QLhkWJBLbVocnNqvsG0Jq9KbKluWnassWbnel4HLHPyJCvuTsqICtJCZVr0t2sL0/QZ9z9lqWNRVzM/F0iHJhZQ6zi+qqMcsdxc/Ruclt//7ydmWL22rjSuXqbpuqW5jZt1nLC53nqMnz+rHzpQLebUEemTIeudAq+jgpV9nzmDTqUcezl6mPYDVz3l7CVJyuffmUyBTbMgRifMFyRx1IWAcRtkPGAAM+tXOorrS1EVQzyLgwzynHvo0ifBL6kCWQb5xSLNXBC1kCEmtCAYhio0rb462pPwLxuUoVG/rxYO7Zr/dWCpGZlYy47ptIeQ+MWcvO2/TUhXbMr2rhG5lGUkrNjvQUqcSygKNVLhtACnbkDsmfkonyEz2BXOc3fMrWzRQPZEtGcNYiQbH7B4lPTsmHCQojmNa5w53DrSAHnA9ZmGRSC0wuKXonSv7mlrd8WvS8cC1Mazrz333sO5b5o+kWF0Y6moeEMKQDPe6IbnBp9uSE4bep1MWDaIeCzEizjMwz6Lai+bmzGDwPCYMEvaQZLF6JvpywsrdCP99QFWzrCat8no+pwwD2a7jwUu3kF3e4QAqj8/ax9g7nQ9lkloPwoIUvMgBVvYe1F45Z4rqSZAxL5999sVUANPzviOQnz3Xu7Ugu6d6xhyhOKY6gqcbK0jkF/rdF90TF/nVvb451MxqISFV9251xmhtbvU3WtOzhwtgdwu0HleQOiH0HrnHdRqtyJGZPLwqTg7rudccwByzwT4EhGEUz7UxD0oFY4g6ZgcmOJmBQXrng6jRBs+4WJgKEXgIII6gYUAYooyZCrI+lICWMC5wTL9jRLKjEcwb0QRskp2m9lNmo7esAJpBiP6u0uE5towrBHPj5yg2LxvPlRZtLCbadXIoa9TI9oXsl/gKWDbRrQK7bySBMvsx4EfJtDSPtgnAhhaQdFw4My2bGUMYZL0CwGo7X8A1q4SNwFXzj7W8z85TbjvD+v4mvS0F8B2h08LdAa0N9eGZxGHVR8X9D0t0LzOo8lddrH32iO5HVffM2gSbaFs9kOpMvFeWP6YmrgcsAGKbyrYKNdgPSYWnaioABAERIhuby5ABwmCzaYWYtsK2RQCzMJR5Fi89WYyRxbNQa8+r88DtMWDOGSVgoTwn4LUEWHrOmAtrupxRLwp4icpTN4AaUIvs3cRTmtapkfr1jMneiS+bxLWBxpXar2ZOhdCvGXuBcq7u/PuvQiq28x4sCta/p8zTJUTS0hrG4ljXHrVN8tJb5iG9su//etYKX51gbRwl9So8ISkfr4X8iVasU7No1tNKT7RNNXq67Lw7oHVA2FZtp1Ru2/NYV1UeV6ql432Ax1KeQzpfvsEfDlj90O3gVpue82KhmE0hDIlxBUpLXbB48YUsfM1bT655Rwzb5hmY5kSUhlQyniYAySEDQPI0ZMmjpkWstCmUH3lhpxIAE/Ucoqq4nGOEAWCupJZpSZyo+5pppbkHmcRGZ04aM2Kcs2u8blZOJ//g33PVWDLCmUOMjc8qXiblL8Feqqr8fHpcJltEz0uT2C3KtuyZsq0uC+yyDS21daoPjmAk9XioszMpy+d20t1eAMoN5NCCnLGf/6IErTosj99aY1hLb6LkM8s2tn7668DU8pLe/dkrsPyQDLA2IqAxN/cAvV5Vv/R1pJpxaYIL0VlLIE+hgi/kWRQK7zUwKHBazt5UT4C6pNdbYikxC35K9QUCYpzTDOMxqSU9g0KdRs18rKLqLYODjS2KyhpcPFR76c1rJJ8vC2AaYHmm5UArg1c7vqkiUSVg+VATGgOuzMzsvI9YJu7SU8DSenVxHSMrS5Ofm7kE/U6Jy3NFe6PCg8++iSXG1Sm+lp24Ba58faEsC98KV5etyEt40TCiinHZN7dUojq56hvV784b7YqClM9R286Kuqcm+sHhHrSqsOaW7mKuXCsBa+n+NZXjdsDaVipgrckeH9bVHvuDF0xtOsq3xIZClFSDw2DrNaVFE5O9SVVUBrgq+B3LIbFHQV3AYxRSQeCZ0owaM6VxXKTvtEaVlkmV6qTyOkdI/uKaPovwNdYlTiA1cMlvAy4DQGVsHpRK0PLu8LOyE7ceVTvRbvmo9j4KwNL3kB0waoed8r2665aHAJZ/HtcCdFZ9FOl7sFbvQdlLB8CXud+OsHhlEbj2BAWq5etVcivpE9AumXPWcCJn09s3JtNoE08Mdx60+oK0D0z1zBnrDGvpVZQ9nmWGVV1f8Ghqcl/p+bXXrX/ava+7ppRP95AWR+VBA5QrjXxvVt7Kb6vcknMC0F5+HivkzSsKXL3N23gYnKZWiskJgzmKB1Yl1RuwQha8JpBd+eW65WPeb9qrdWlJeQ0sqt912ZeAS5lWOWtGdl4opnLSp3M96GTryS2RWYU0cp1XW/+t5opIoCz5OeAysmQFyUy5DdXzK2D1sradS83YSY7AUuf+vn0hT4ic5/brzVOoNq+tRnXDA0+x3P31rfu9QN2RHaYcttmgOkm6BVJBLv9j1i05IoT1KI9PWPpclq/m82tTKrXHa5xmf6tfG8h8ax2tujN82xktXa07DBuLQgRQEDdrs2k5xuVdr1WtYd+Y79lXW7qQYhnziBZXBc9yD94BWg1inO1H+beMl9JJbmMW5ACKgbLN+CWXZZ65XUFLHTIq9ia/o9+cerAkkfnpCsw1UM8vI42h0jqv66byCazyyUDVF3ap31FZtNz7UK9IlnW19FmWl5HJ5Sr3vaNHHw4tz7lnnmgzWNnjJP62OdwJprUBIqp9eb5mWEvX19PTsL+7lm1cHcmN3GHZzLAWemBtxlU+2rt0HcBuQU8JVeKNQ1ZdVXs6jeYcrTYtmXBVV7m1DNwihHqXyTyRmGy/NN9aLSdAJed91XkHOdKHKUDKPUhFH9iEtQJXhB9nBGYDrFrwFoBlaZONudJFETPjStJbgSs6RmnsygNHAR4lcGV8F2D1IKrABamYEmbK4EGnBq6mDFUTcMzGwJ+5ccKoy5+PyDXAsvfmycyiqs+XjdGNRdQuY9J44S18Vr1PIRGb6v5Onuny1mVUVK5psivxmwLpnkvGVT2I1uO5HLc03AnQSiELtNvP55B4dQM+JI0t2RyelseKF073UpwoiMrpeOpYlGdhSDNiCNOCw6moiJL2TAUPUr5VAQpQi40GqJzQa19jxbTIZcqcPQUdoKjreeQEXmaQ0Twt75ppebbFknzJprIreAmSUdmdLf7YE75lo2jScb0eojxPYMly0bTNAviKmThasPKu98ri7J3JCtSq4ixne091qHUnS5EB2D/5dfn0K7Oi7+lYgfvA9UIOJ5e3ECjtsX1vZTfx6HBHQKvqVdWMpCY0Rwvqkqq0Nq+lvlKd/0J5FwjaKuPa+EBFqUSYp55cP5/lEmzIazO2Kwhody0LfPXi0757ip4FWQh5olZdQ9Fm5G50VZlhZb7lt87HpALYVI15oKsyDQBusC+y8PV7OV+MK+KYZ6WIjm25Ihdgpcd2naTUImAZhZpM1X+pQ9yxozGviJCi8oydRU6rCPtBv6XLe35P+WagVGVm4OoyrD7mubL4+xX0uWU4SPClbMUnl+1Fua4B1w7MhllVUpdeyU4+qMauJeXilY+i383N9y9954vMbmNYYmjdcVq+jFZgLkWEZ2ASuy8Jjwt3yqZ1WOh8XP5qY/R5uJRkMbdTmeTBt/duOG9dlIzGC+aWpVI1351O46RebBmsMldh0q1kXUyVknBJ/eLLtQrwaBhE6ZrdMq1yZgqzvJX51x2mJrvsAehd2Wst5V6QMoD2ief6MpbGKiizB2Ga4zHYccUVcxnMhudWHOaciVmwHChm25ZrF+xUg26cVj1YvSyBr9G6FjeE7VTtLOHQUt6+TSs3/k1SgQB5qdXF09jWHWFa/bDMuErAWmIu68C1dNynTC3jq08s5VYJqq02rDo0CZfpLGs92pwO/0BaAOrmoUxpwWhQsKwQbEonE2Yxu7t3pJWVxKDKQMx5szWvNavmoEyrSNyoRMsa7Fys2FZmWlw4fLR1leuozbV5NmMgsLRqLK5ZZtUtNvDy5FG3aCiWS5Nd38WuBSqYStbCOqcQU2fmemme015j3yqiaZg61Nd5+1TGuFIc/dg63zADad6tlMI+VWHReSD3g3I5nPFoMZ1emspOlCUuO6xop2CrTevA4MtP1DApbxtL5ZXz9i1pj2Q/t98aXsRMqw3e66wFrBdIuHWb3cMP+kQBWVD1IpnNxGb+9mswkTGxLPEIHsDYbWrVZ/tzwOXKZeVxOLg3qJSH7h1Q1YN8ObMNE9y9zRL2ey1PmUcuR8lg8m1+wLUIvKa+qMghP5aAhKgK9TspVKa99mng4lZWrgc4u/LoYGXdB9JpopxAd+BvzM0cMtw7OCYc8OmfGR7uROhV2zml1p1mWkuh912VY0yoirefYeX7vNDo3NakW9+/FvR+6pxdDrzwu5/6coo+3+29uW5FNLma5kF6udRUZX727D3oxmlxTqu1Z1XPTS1I5f50yo2tUG5gq6S5CF7s0MELz+jVgFnQ2gBjLmEzq/JqxFmvc2+/AneAuAYoAR9Z4dLVkX/X6Z/ayYyNpIgS17GsouIzMBXu9jE7UWR1Y7ozECEacMGpBh2DkbIXwNdjuk0FNY8npxfYUOfTbsISk6rOr33na9/UPhtfGS8zr33pbP2E62yLtuTSKQgZ0I7/PFN40TCtJZXf7eiBl+jCiWmeGLaCVzf3BRXqYl68LTfPZjyj8Y+b5awCVjWo2H4DqJiW2rMQnF3LJ1zSkELzVZOPHlhR9TslmQGrEKY64FcFra8bp0LTjZc2KXPBmLQMVk8oBb7WnZ/Bwm0K0EWNuIQVRAtQdPVMvqKUA9qjeI/F0nHCu6r7JBSwQlF+uEUzW8eW1m1/LRz4Ta1RCFqId0vCuynKifms3r/0fJsSP+KehfCiYFo1YK0B1+EMq77eZxenMixNgMuzCxa05WSaz7ruOtW3NfW2TX/uvai2fhBUuzS6uuwKWgUuyUT3zuSSiYEKWUne8xhjWr5p9FCrTLY4rh6+2Uo1IOR3Zkd2rKWpgcozuaLiyEC2Oa/1Ju8iA1eeNT+tIdZKpEazyL6tsbGTSuFo9+rNUbwFvbt9MRtH1XFQgDLutqRZqAFrX1esYAN+PeyN3yCh+azJNwig8B4s4vnzR4Q1ZlQwHDoQt7fcX32KpGS7w7Ca+LcQXjRMaymYWuoh9Ybugwu+c26B0YpLL3DdJoCVjB4wEDOWlYwhea9ARQ6kuuXyAt+J4o6Wjlz8ZkgtV9HXBKvGVnbCfhBwLMDNVE6As/moDchtumS9m1RYr9Vjqlgeku1/+1b0R3G1xiyH4gmbS1YUnfqyW59+X3d67EWoKjXXV90BWFUVbg4HCod7WXKr4UXBtJZCDVhrwLX1OpeiazWdg1WU1PyQfNuzhQrqwFCUq0cnOMc52mPJOquue7anR6kqL6dHUuRIwBQVxByLUDUhKobFVR9f2FUmWd79OhevKLqL3zyUPoupwbDyKuRtVQJY1Wl+jkBXXCurByKvPg0hTZqr4EUhGPPS+lHhXv/XOuu9+17pm+6G1jEje/pFdAZAt7XXfJdNGVrmulit9Te3/BhVcBSrJFYuhnwDuvCon5vP5VJ/50vjotoyVZQmJ7C/5ItiRTsC2t623tdNphU8PUF0xnC3mBbZv8X9dnf2ftgimg8BrNsM52TnNQzvh+VNCZZg0ASnNqIcH5U9q1HfCYiZLYuy5qsGKgZKIFjtmZciuQQ1Wn25TUsj52jg2J3l5oCrdjgoi+kUdAZcDngprzKcN1mTzMWvn7Jfen9YIY2hdw/lXElXq2rtHezb9/K8PfLjW3DTmg8TLZ20H27Y3Hk+pmBF2zgt3A3QaupiG3AB+rH6m7cJnqOudi4fz7JOvKk5VYGSK5e5RVfRmzgHlWh7nSf1YM7ThKwKv5p1OZ2ejstpgMvsKWman8gV64J3UFgok2Xl4vXiM9y6YGRPTMWZnnDNzEuZiU5ojoZxZUZIVT2ZU0Mo1x8LRG6MGy2AlQsVHrVqRCp2/QT2ve1aZ9jbsGe/p7Dw0F6X9zhB2utymYdjkUXn+9mbbv37wG8r94pWUtf4+75jV4KlT7bOr1tcf+004LpT6kED842xDwUsoKtp6JTCHVXT9Z/krVjdWtsdTE2xeDOvghXgymgCWVJ3UZP3WF70TlWFe9WElqz7iE2n6u5z6jFVv2m5iFAsRWL2K7B9D+Vcgz339gxeMZaebKVaqxXjBTY5hlXgG1W3ycUaUAIFsLAeVvYYa+ZTqhjrsU1FoXJWZTdN8gzCPgMFIHDKl1mGDFChLmSXYNeV3beFpoLye/E2yILh6ZAFJsgCy7kJiMGLq72dlzbnQZNd2/bti6Adl06bq38fFFpBn932KakKfa+CCW7G302p11/3IbqT/cAlxSiaUP52WzWm6RbgB13nMul5V86iuKeBVB3uBtPqhG1tcV8PY/WOKr+e8D8uj26+59ArLvSi/E/vEZUZFjW32rMV8fb02vz1TlwLWXJVe+ReibEqJ6GJCpVg2rhwulCvvIjSycE82VCqEJdkhD67L44xqAItXLHtYml3CqTOEd4VXW9nkdWljasBLVeO4pwrkAcNAw9yy4p4BxYHOpqQHfZYN5y8kjzJ3eSdZhrbm3U+8jObg0Wz14zKtuE7b9a+qCh0Uff5odrn6IYmXh/0irbfZVxVKgd+020x1mXQ3vQ23997kCwbyuM2vXOILh/uFNPKH0l5yu+1Zss4W2q1lUY1KLXnqsXfDsrviLARqIlKNVjxqwYvA6w6cZbeMaXfyrjQeb4KpBowb8pYW4Zh4FgIWK8OVEYiiZVzDOZzdr34kzzZ/e6Vwf0mQp7d3AtfjVvo7twzBFmaXoADBdNyE886Yb70Xjv46LLifE3rxxhPSHxFmY7su4SK2gK4lDslyh9cwULRApWvMwIhauei8DKstqKHXz+/z1ufO59ma3+5djQlZRr1vs1kPxKltEXzgB7jKovvv5VFLUUWJ13g8vdtkS11fHYdgfq4LETnQQr/93RMyGlo+ym+jxPDY860KokAFA24BqwauMpza/n07usBlr/rlsDpyNB+bvvBpS8WfWR3nRY+mK2AtVZVvudMQJ6TtSNs/d4DFko7DBdCMisSs0MGd9MrhG7wMze4OqhlvWdRtpSHAhWZjal0RV9vP51HRQ9Waq/Ccvorz4RcYsWz5LrXlDvQXpbKv9wFwNI/K3kFWLEGL7RqcRWKucyugdi+BDKfp9WhA7WmAhfvWgiWRP8b2ywaHpoIqTNa+uZhbaCQIVQe5++8IhFneKCzg9bf/bt/t2mYX/RFX2TXn3/+ebz1rW/FZ33WZ+EzPuMz8OY3vxmf+MQnTs5XqgcZ1eV81RA9cPUZxIa8nEApe4zyCZL7dtyHec7eRlkgLD9GJYRy2a22nOAoJT8BRXkbRoEsy3r1sbyhqiOXsyXdsyfpvX0hWANXl2k19cUAZZVU35LlyunBSlRtypAgzCmr2qSsclyCVKmq02VWaHBgEqpnW3jJZNVVq9CS9UchhupncCAG/824d+uzZPfL189+SwuVv0tULK5rXyGKKjRWDCuz4NqOpuUlFO3J9hmwDLhcPeTvoi65xCug1Qlie5yyEddA3AJqr+F3qoXK07060/SJ+td6oY6f5eXScV3Z+XmKtGpZWMvF1ZJtC7fCtP7En/gT+N//+3/b9ou/+It27Qd/8Afxb//tv8W/+lf/Ch/4wAfw27/92/grf+WvnCFXFZruTANYdZzTqrHRCXdL9AjDQgFoMYJr+BmNtmfiP8IGWPrJkb+3XtGijlelWwuLDu7a7wa4JHhblwcuB3mVsJe9m2Fe3cipKZcTjoGc23lw98maYEP26jMQrJ+xU3fkniL9dOOWiqeEexcdwAq+XpexJedUgvtexuWlbyco3pZj0/IYNT9zRtGh8K/ff+OuowCtR1O7+jkSl4vl1Za15qEEd/9/OVBdB/47KY5XEtpb6lPCfunVfrtU/KaqwdSg2M/juHArNq1xHPGqV72qOf97v/d7+Kmf+im85z3vwV/4C38BAPDud78bX/zFX4xf+qVfwp/+03/64LwqGZnObRK2Lo0D4uc8lu+v10itU1/qmW4tR/d+WjzotJU9ag73sVanqmgiLkWXbV6SxYqS1Ue+B0TJ7ACSTDdSCVhZnVV+7fUUTsoisiqsmmzXpcW1YNfNA7kX+ioYhWUZq5IHZouPNDMHe6aWZutQL0IPgAaCOnt9BSZEzmdFn5vNKgQ/g0bTPotHqToVVFZnqaZUyCADLv+uFhlXlV6hGnR79yT702vSd4Clech5dudtiQ/Xun0edq7yomvrsGQfWSRT0i4TgOoeBie7ob9B8yKC2YicXTaXp3hMV2pqruv3uCVoer3nzccsduv6+y3zyXUk7UMfy2xfVCdwdLgVpvUbv/EbeM1rXoM/8kf+CL7jO74Dv/VbvwUA+NCHPoSbmxu88Y1vtLhf9EVfhM/7vM/DBz/4wcX0rq6u8NxzzxXbowhrwveRhY3laaMtANZRD1j1GA8ArOKA6pOdGxwjKJlW/u2Bi9w9BUgVoKVxyvtMsDvvuizcM/h41lJvTf52T3b9DlSqGhXASlApgSvXleOLOoM8GDB/yALzO4BVqcn8Ofde2OWkTiut4/lil6xoDE36LtTp1lu3sdRt19my0j4zrLXW3YL8OmA1J+o0FnJtWMiJsuW4b3dvipvOl89RsquahZ36kGcHrde97nX46Z/+abz3ve/FT/zET+A3f/M38Wf/7J/F7//+7+PjH/84Li4u8LKXvay455WvfCU+/vGPL6b5jne8Ay996Utte/rppxdilhVSfBhFnK1hfwXXnYfmuL5bvjrSDWd4jcX3pNKok3Aj7P2mPd6y0W/RRJd1nBPNddGqCkpBmctYApgXmCp42nK316l9LgOVUKrjhsEdD9XgWwcglIUdufxam4hHBD2U63DlhJYtXQ8GfgGNS7qknV3USyDz9ednPC9X8u2DSd2hqEHNP46l4r34pEeug527eOXrqvP+920e9A2425bSPoiXAcUhtQ9ZvZ+CrXWPURz3Ki03SxenSrP4zuQG+96ax6u/lfJD6Mu57aFXnt5xue2rI6rSyfLlDJh1fvXgm970Jvv9ZV/2ZXjd616H1772tfiX//Jf4oknnjgqzWeffRbPPPOMHT/33HMd4NKK2pLiibX2QglbH6P+CFCpRaiNVPcuD1DWuHdQ9UpXy+skpd1AxaUucMH9VrDy6r8GrASkhoDAQ/oNIMzZjhSo481XP4sHLCI3dgz5OUhXcyV3mmxPhARY0YPWgBBiORCXdA7BVjAACVCiLPsR3VL2JWA5b8gGsPqCqej0oMO0uOQ/vWZClNxB1gHKv0d/zheCXJS6rbiC+jaxR1L6lq2/ifYPkt8LEJYI2o9sX5qgsmNx2Cf3CIKvAx0i0D/uvZpTwq2P03rZy16GP/bH/hg++tGP4uu//utxfX2NT37ykwXb+sQnPtG1gWm4vLzE5eXl5jxPpcjnpthLqfHKW8wCaU96vSR65afyB1UX+sCl5exl1Hr27SvGNsDi7lEh5NH23jKTQxZ2rlftwatUvzlmw2mva28VcxtWBaq9C7uL3RVytiqbAqturmxm0yLKakTK5SlsXM67rhko7W1a8LNImMeDFN4DWv1k+hxqJ4OAFa0IVW76HnXPPD97b32v+l2jfA9Ldd0cVwDWHGMZnBzILMmD3nmzUXnblJPevTFV3obEanFz92dQLc8Xeer9zpa29Gx1uffbtNZCbfNqj6EM0v6fFm59nNanPvUpfOxjH8OrX/1qfOVXfiV2ux3e//732/WPfOQj+K3f+i28/vWvv6USvKC7K6eFrW+/9xG740Lo964fXrIDAcvnl7eyo18Krdx7qzPqCHlnR2rm31tQBzYbZG4+pworWtaaMHVMohDMzg2+UBEO3pOQjGmVE9zKDB8esGJEjLNjW3HZXaIn1KjaiuBmFlGGxeX11MsGCmekJYZlqs5c57nTofnXx74+fQelKnPFzja3YnJtfk/DXQWyhVvXgKBu550I++8/c4d7PTQVv3B8ijQpw9mZ1t/4G38D3/zN34zXvva1+O3f/m386I/+KIZhwLd927fhpS99Kb7ne74HzzzzDF7+8pfjMz/zM/H93//9eP3rX3+U5+C2sF5JW3saB7cHLsVFk55G25reSkPWBt8KU2p/o7ynV758Ex9U1lq2bIud5w6EemDJcZZDLdNSEPDJGTMkAkwg5umKCubV6eWrByEMHPKTGx8xF2xlISLAqQMO9h2TK5emnwW4zg1YejQKYJF3gc9lAGA2rFnAypay5whmEscMKsc8Veq9Yu8G8vonyWCVX4rDYve+2f0usacELK/67LwLlPucb/m+m07ZXlbWb8lcx+E8m83SEiL18RqzqRlRL72aaXliZfdQWeC1/NfCktxr0xPYIS0P3PEy48r3ngdQzw5a//N//k9827d9G/7f//t/+JzP+Rx8zdd8DX7pl34Jn/M5nwMA+Ef/6B8hhIA3v/nNuLq6wjd+4zfin/7Tf3ruYryog7Xppo1QsT+m51N9L93ra53FxfvqA8cAqSy6k0tOkNk1BYQasNJMFEGFpWdXzEk9iFJwmlQGTK1WTy0UmRHAACLAoVND7oP1gBgIrNMoRccGqWWEHmw1SZ0uaI4RkWfEecY8T5jnGfM8I3JM0zQJaHnHiQy8raawUH26fbpPwMwLNIfFXUYngF10GGKw5w4hJFtdCG7CYK8a9O+cqjwdqFnHIEc7d1gCrMX4e9LpgUxi9Gwf2uL3tvYh3mLoe9XXBcrHdOaXcXbQ+uf//J/vvf7gwQO8613vwrve9a5zZ703lNVWW0o3VuqB0bcGKv816TcfRrcHWR60IED1T3eu/hD75fSPv6RZKtUrvbsXQr8D7ArU63Xnbn49TCwn6+OWgs2r6Qo8r+mDK1btPVcwE17syJfpFe9DSijP5vOnYiP7remkd6CqQZmx3payT2zLZnJ3Zbdyo2ZeLSDro+h9RSPwoFE8bAlYVuaG1ZZsuWZY6B5X78nX7ep3sS8so8AhNq2UUvUdb9TkKLfV8VylnsbFt7Zeq2fLRzlUVbjNptV70yU48S2BlYbHfO7BbaH/QT38bspijqe+29X7HZCcKZs6reaaO3Hgt4OuBKKFfXGLSlKYICwlfQtccEKzQIReGTwlEddyqJceK9PS3xXT6D1Hhx0kAV0JcarPAfAAEyN4nhHnmBhWTKyLVU0Y2YFOpdIsAMupP+2RMyAXj8ZZKBnw16F+hR6I3BaoVQ1aJ0UYGgzUOx0Qe3f917Y/UPN/012HN+ojw0o+S5dvpXhtog+rFny4M7O8b2mrzfU15tRX6R4dlgQ9L0Swnk/nOnXO9dGhRI/+I5Q2iH3Bl3lLx/aQNO0G18MnF8v+qt53v2JyYKCc9R1OzaUinH1sFDEKib13AxLlcx0jzwIKQVuXp8zdeuca3b87V1bmtHR9udSKU1sWjKkErz5weTB0jEttXEIn8yN5UPXvO5e/UBp58LV36NOpVLp1PPeKrXPk+h/VqaYd1CGVL/9filfeQ60mrE7UJ2QMqc+4cgFrAeCZl+M0TfooPpbM1I7rmK/a6FCnXxaovY72Oz0h3Dmm9SiQ//hwYKOS6F3A6gbac3T7YUt+a3FKzuIQwKsh9RepIPM98RyJi790rJOzKtPIMeWXYybLYBXd5s4XgOUYiTtXgqnLxz+GV4tZEUvASerB2Zwx5sIhw6v7XG494Kq3yEYga50wefCyas6T9FLnHv8aM4angxa4KpWhy0lZV484b2l4vvvWx5x+Irel8loP+0q178ILIPjexBnC48+0CoGQTnQZQKEM97H6jaFNgxb2S4VitJJGbzvs7WWQWvjQlg4W0G0t96Xrqz1QBZLiGcve5XKquTdqYpV9Wq0Qy+pA6hbOMyoDBS7Plzc6jyfO4FHYchwo9VvCQi210n1vbWjPn6qzZj8DTG0XxbU9VmBT26bA9bnMuLLKk4HITqWo9SZ1ENm6ugSUCzoWj1Z2NdI+P1OXcXX2LcJ1qtaDn0u/qf+qZCTfKeu9zFJKfbcHfqfGqNo6aI82pJepkyWghMrKDNg3Q/odUGobtPBddDLSZIrSts/Tf5JyJkp/4XaQ9O4xrW6jfpiAtZzKuXoay8ncbneLFn4DC4B1YnFK1ZkKriBblUch5FwKin7KJoxRIAt/x74800mH7NJyQpc5eRzKXreaGxZFWvrtmY+3h6m8hgqGCrxUDajMSm1aYs9iA588Zgs2L2ELWB64DPiigJiAmdZnmoYs/QVQnqCfU90UPQ/fAehI0RJ0HHB5wKorTZuDt39BQLTJocqvyrzuFD46NrUeliSQ7wwAyJ2Iox/lhVsHjz3TsmZM/kV1mt0ScPXUTJ3jpaPusY3V2Q9ch/a8lpPaByf7gsHBag71Ga7qOD23FwB1TktPmzXg+r/cknRi3zN3WxZu/WcrVIJcHNnVspSu26lqOikMKXCxZ0EqPFPE5vldUyjW89J25oVLB3OtTlzv3WZwh9qxlGnlZTxMLahgSK4uvPNI3LP5FxEzkBFpfQB+Gtqy+B6kajCv6yk/fJsarX4oPRKW22l7rg5MlGetgJuVItf4/gLkiG3BAPfy6kaxLXSakjGuMh9lYJqnzHLffYLlbzslxa7ylspbn19+rnM6rjz2oJWC03iTLGdeXff7tmdVx0M//t5Ym4p5y+HwDI4tUv3hFI3ygEQXBUm1gSitieTWnMq9ck2pWhTGMZ/GyaACrlqNVqTBMPtMoeJCGp9V4E1PemploQKuyhDTrQsn68uyy9gsTuO0grCteU52rdlATJYr4Xy/2eFixbIK0EJWEyqASXyKnMgukwFXAGVwZ4CiZ3Glva+1de0BLK0CtldR3FUsYE3uN9q43KljdtdyJ44Kd3JaEPuW8GqoS3N6WMrWciIBX/bPuHa3T2fPMx8YUjPns2kL7whoSciGrQ6yLwFX71yvz5bZhB7VYFcc73tBC9f6+e/rpfSeZU8GFlxLRiPqXZ3UPcSyH93LpfQ8qu1GnO8zge3YTBPbEkVpqPdMq/N0XsCJsIZjHWajQSU7Xa7FuKRGjZaEcFNvvtTODuGfwX536rRKAaV9L6PXkt2KlV0VNqkMxkuegT5+oTr1oBYrlaqxzfJdkK/UQl3adEMWwnK9WOVUl7qdTirPlODkO12uTTpBTe4brpnXoYFd7kslB5rPvmmctJYGyTNxLjNDJq9tc1sprwcuX4udJGoxsZrDaeFO2LRS76onzMoe27KQ35eyP1z5oFbSPW9f68SwhOkWtpd2f22UEETN+fK4gTnKPcdiwtst74JLwd0bSAsPZJnS5CL17D6F4GVke1MDt1LE3O480DaC0CXDzGk1LKemZFMFqot7NFWgbXOeyolj7ZwR88bVbw9eMbMwtY1F99u7QivbUcaTQS46sPMMrz4uKqrY++qvsc5/87lDk1mWT62XQ/5dqblvyXmgTPuUPPod6vqMPlXRmdhz33Jut1cfx4bHn2lZK3XgVLyoFelMe67ZKd+bqRpNfX/6ctez6oXNDGvh9oNitzcuZdeDr6W8skDzTMNBUd0ZLqN18ik9zGoXaEvOJByyYISXd5kpOKLV8Q9wFxxgsUyJVDAaY5Ou4C11s+dgX9GFI4kT0tXzFN5+ViQtUxknjdfKjCo6RpWK5uLFvJQJx3JWeK2nJW9EzdNA2N5J4irKas1O1HQcymry48hyfTqw7taoq9sDG37LV/KZ3tikrYJ7SZ3mtQ9EOn1TOYxhaUyUzz+n35cTmkRyhPTpbQOugmXueZ5FhvWQeuWPP2gB0AaQvYlSaEn9gSjS4N3K/WfvlHR0IY9tkE+A2etmYJ9IBR5eIMIAK7jNf/HuRu29O+FdqLwMuDLotATJnRCQIgQwRaeCcwxFGdDiV0sAiU6/8VZzvW+hlUZChPkYiyqeRYAlQuJkNpWZFhujApE5ZxTMLEbMPNvUT/ossUrLpobiiKB1LMW3p7B6jUCM9r498MVGFVkBG/K7SeBWdjJc64DNloFlVfHjEjKYnS89YAkMH++6uiOgVWkQCuAC2j5G3d0vr5aHHUrfgNehlEpv60fc8gG2UTqqAot05Mdwwu1FT7pOgiHdwZTHGjR3bVk9m1YDVloW7m77NHoOGTLAKeOoB+yWlK5Ir6xC35aowsbaxhSzgK8FfXUuLpSLOTZMqwQhhp/mKeeRmVb2RIzFdX1rxYwY9mQVjbWdMKmCKbp49avgpZdThU4zqO+y93AGVlC3u0PBZul+P+6qDNvSX7q9zI6a3z7/PtPUUrQc9VGEuwFaVG2H3FccL1Gn7Yk+DG60BbDOkMtZUllr4nV99d6A9KWbP8B/6J38KrVWC1pu84xLVVsesGJMVyiBSZxjCR4RyeOuYJG9CqjYvzE/ZVcQUOkwnVnGYhUzXgAcImLw9q1KBRgjSJhWwbDmuQNiWVVYMCxlemrrcs+lGg6Cr1OpL2FPBbut9pkFV9flHfaALbNUp2atyfe+kHRoa7FuP9xyEUog2tIZXmN8jx647gRoVRqGSk3jKpn8GHi9VoEA5fM+Tvu7vhGmKjk05GQOY1jnA6+l+84JhrU0X8hBhYl3XoBzhjYB2Yh/x7TQFYbKJrIgzFsaLItuetrACmcGx3gKodrQyuocpx6rkRUpp87SbuBUAExWq3n2Vc85yLHzzKI+ZEIDWgZM7NSLBZg4tZ4/716VZ1pt0OmcuKqC/UyiZVooFR2dKrZ3v5Byf3XpFri2YNk51XhJZuxT5Xlm27l5Q3kSEJX31Nf3p3MGenrGcAdAy6kcFiUGLRxV4aEwmOPDo1Lbn6NTmgQNFcf7MiS4KYIAZBVUPu4GBRAZZ4SCDVXqtUJFhWJMuPXjWdIgIPIMECHODK6YFljyW1ouQm1W+uTSuWLAGNU8R0yyzVM5f6AxoznF0y0yI4SIEGIBPNnpIqVDRCm+Y1jGtGSvdjqbFsoBqM0Y72bUSO8hg5cYtjJztReCkrWyqC5zL7PpVHiHjULeUn4/xTvvHC412a74tzZOIOKT2/t5w+mFeaEQy3OEOwBaOSzBloa9mGTeXFklSE2kbiruHs/g2lL0BG05/dGePHrZLV5fKvN+NcFyf6rXky5PNL0009osFbYfv/fyyKXVaICZmnuzRikLvkIV2BOqveJxdgywU2o7ipzdup2QLSZ59+Xi/m+vvvQgMXcZlwMyBRPHKov09byp5lqmVTh4VDYz7zBRT7zr66+rh7AfWm9LPIvLM4sEbOEdUbVfuOzv5j1pPmrgOpzPrHzP2j/i8rjoxHceUlWExsDsXaueSrqglSrxVFvf1vDYg9YaUG0KHcAq91g43pvoaqn6gLUltHzx4bKwQzOr4V/Lv1w/je3KbU3oYZDvuUenJnODbsFsc+WZDQtILC3IsXrgIalYzIOw8CLMcTNdy2XpyUkWwIKo5uIcMU9p5eFpmjBPU2JTk5yfZkx+m2fMcwKtQcpYgjEyAM3KtBKjmhqmVXscsjCtuRPHqRHJMS59bz3A4pxmqitlbB6wDGlz+u7F+tdjDcL2RveKS6uohpaB+OO7xE5S6Hea9wFXP438/W7zejyvcHrsQWs1dHHHNesasLrSv9uf3BNvrQfU5n8Y6PQBaz2NQ/ty20G73ytbDoufA8FsIPn8vrS8kDaC5DZVAbqLxW3cTUuFZ3bIkNspq9E8YBUfbtW1pVrgmjxWVpPVb12mVagEE1ila14V6GsxZZKfv2RMhV2rGaPlyhLLeyyOVUb/PdiRVXVB/1y/wgNWk4SvpnyGGaByLpKeFiB/5hm4um+60xCPBa4lb92a6exje77EXJw5Hj0XNelWropZVQ9cj9tqGdwy8PlU4PXuJ4S7AVpUbf58Hc8fbgKsPQlUoeU/e1IqgGvjTf2Uqv2+sK2Ep3wii3OWUR1rQTUBpAlMq/bdd7rgtvKkh65qsT5wSZrKtJxA1bQ55LFGcY5gEZRMVABWzdwKCVcASap7Ux8qw/KgNE2YHNPS38qubqYJN7KfpinN5g73XMjjFbUuODJm8XycJmFa02Q2NJb8E1DNeVzYzMbwEgOcLV490Hi5tTjAdKCYmZayWPfn6tGDpNn/qtxMYeyGQZBnX1b/e4p5y+E8jO004FpNfRNjOvb+YzVK/XA3QKsOXndLpXIqT2uSgYpkb3dVPYRa6bDIoCqG1eVs5H/3GdPWxtNOC1OnV/ecatTek7b7zXsiNnps7JmnzZ32PbviEh84GmQJh7tUzs+mAefcUeYviFpsZh8jciwlTwTLLk6SUJQTdWDJ8uzFoOF5FgY1i1t7Pi42URV6NZ26fmevyjx7iGdYYBTOHOrG3vOGjKJuNFd3rxa0ynUdgKaue6zLAZDtYzeB5Xfv4u7tF7m6t4+K6ka9ORwubKtvgrlhb+n8trRIVM7cdYE8Pqx5DbbaE8hx//7940uzTDx1aqi7CVoaqPzRA6xiv5hAb8LJftgXL2e35CJ8SDhvA7719Dss9+ieXX1bLfcWklWBHojAFBCIEL2HogcdYwUAx8RoojyHjpdiD14xz7VnhaiFTAVW8zRhuplwc3ODSbYb2dJvvZ636UaY2BydsEj1mbwtg4FXjJwEHs9gsNw3G9NK625x2nN2MFFwm+dsO5vnuRivFQubU4dxaXqyt7kSq3FfWe1YOnoU6krzaNT3g7KfJs+vnZJb/zSODJvbvBFzRbrbZVkPK3TcqI4Kdxq0SsJVApedWWRYPW61xlSy9ndJaVcyrTrG1oa5lPq5wgHpV0XOHYKFJBybpEpvkj7N/gfK1X65fL5jUs6iEUIAB0KgABYhX6+8C3hVYrZrKUOKRGBZZLFhKuwFeBKeJQhm5jNHdaqY8jbPCcjU0aL4nR0w1NalzxcopE2XbbGuPETop7Jn1pYB16+T5Z0wGhsY57FcRna0XrpvrHp7TpVa5+M3eQOO0VVpM8MtcObedfW9G0twzcKAAGhm4T9jWGIg2uZ754sHEFZIwqRZ2lPJ0A4t+PEyo2FY9fEKYzt3uDugpSDeoTBZU1C9uKPoju/5+H0nv7ocRbbH9jpeSAzr+LJk21cJUj3IUtFYAJdKSt8J0E1624HSJLWBAhACOAQgDIhDAHjAMAzgyBiMceXMSRgCuazmyAiMrMqbZsQxC3Y4wVuLclamwdE8BK+vrzFdX+P6+SvcXF/j+uo6Ma3rxLSurxPr0vPXN9eJbU2zCbKAgCEEjMNgW6C0NCNzRJzJZoOfboRp3dygmJbJPCwFrFRF6exZqpYsAabe2vdm+cRsP4sxprqrXPrzCsuqnuSyY6CDp10bUq2FdT5C9jjtdph8m9lDYs49F+BBaVfipR//EPZ1253crfmfJzz2oNUV/Y3XSnlHzXBa29K+Sq4bQO8DocVmss3no98gl5Y2WJursPVqWohPzY/98bfEqbIkOWeARdKL5PoG7XGX9dB+6/33QKm7XTAtE2qd8yr0wGWeyiwYouYiWnR5L5wISJ+1Yi6FncoxLLNpiTpOHCX8+KxyWRDCMCTgVbAawmBMKz+LG5/lXdsr9qc1XzKfWKj0PFg1SoLeS3L72ouxyKc+ru93LULrt2wCHXavBVxo+qQvSFhXshdVPP5EHf6hjMsCC4uUMub4mu7Kd7xeMsvmlLDEuG47PPagpaFfXYvQcWuZexBdBqwtLKv+iB51b2lf6JSpJaBl9EZvsyfUwpX33+GBCCGpBREChiEAMSAOA8CMIQzgEDGEgEiEACTGVaj5IF6EjBhnEMHAJU4T4jg2KkKGrBirY5EUeOYJcRJ71fU1rq+vhWFdybkbzDc3yfYktqvpxoGa5EEgUAB244hxHHF5cYGL3Q673WhMK1CyaQFs96pNaxYbFhji0i+CVWb2UFXhLCrKfI86Tuiby96KWV5VQOQY1Bzzc0QH0n5WjnZ6KT8bR9uOzJkmuLJ0G5zft9dpAbhuJ6R8tjKu7JKuTfJhlfOUsE8AnBbuDGgBPca19+qyDSt3Hdbz9F3OCrz6ubb5t9595XEuBhX3HR5WGFmVe1air9/VUw75dks+vqnwKz1IL41jP06CSdMes1rctPfbFEWYSSFESwENv1kdyHUBuWLwsA0idoN4ffo+XXscwjAk9d/FLoHWbrfDxW5nzCvbtFjA1qvYkicgmIvaJ0ode7NXQVWaMo4r5tk/XNWa00faa1kd6BeMKteXd8JomFdRj575lu1hsTVLufybZKkTYtcgWZ9Z2L6w/owjayy/p7JrNRv5O5dYls8CcKl9VV6OmLbcfisgcLGzBI9lagufYiM31tI+EcvuFGgBW8VyDqf0B2qthP5eB6xjcnx4DKvH8Xrt9WjX1U0dxVZN2Bu30yZcqQIDIQQSm1ZmVt4Rw6+8a6o1yk9n6r15BjGSt52yBHHK4HmGTLshM2MAgHi+yZiq6Toxquur5zFdXSVb1dW1MSwFMUvf2crAwID0HIM4XDz54AF2ux2eeuIJXCjbGsek6gRhjolR+bkGld0AQKCyA2SvRRwdZn9PnE24qg0pBCr2UlkyV6OzURlgzojzJPayBNYcZ8xR7Vuz2+pBzf31ygw8Q353m9ulyG82YyYJs/HAcBuMpuysNdMhVdFSecjKlRnXMeH2GNDDCncAtDxEkDuj/6mJ1bu7DHWDXVI45JH3RT6VimKtiWwd73Cyjr0u2HLEBeXkGRo79w99b00BSX/vK2fap3efe9Z5U9uWZ1172RdzkULupCoDKCeP7c6MIdM+aW+6mHx2jgZIeeqlzmMRIUCBNWAIEcM4IMQkxcIQRCWYgGq3GxOY6TNq/XF2sy979Fl45UH2QBDh32gNiDCEAGIubGjDMLiJjeWNcQQzlV6Cleovu7I7FWLDuLzKFU5FXDHwDR91BmR9/PzM5ACLmwmPfWKHIkVZkIzrLbPrfdrUMC0uAGwxqBp9Q5nOGR4WFN4B0DLRZGKqBaztIjd/yss9khoA14DxBRMOLNj2fuaGmCsdvH2Xufmrsta9ClsHTEwBFNIWAjmX99C6vDtbmPXXRWBC2FRiXBN4noxBeXZAwirUbskyHku99iZhVHGaTOWmLTYAJQMMhIEDxnEAEfDg4gLMnFzbhwFPPvEEdrsdnngiMa5xHER1KEKrsAspeGWBr88atC4HAkcIcwr5u5J4u90IBnB5cZHtaGOyow3k6itGRLCz5WVHk0lUpGovi54JxpidUTzb8ipEASz3mqHqyWxc63+JhFJlWDYivj1itRiUaQHtwGHHvArgepjle2GGOwFadeixnG4z5uq4JiKdBtIyK8+wSsUEFWNB+vrvXqh13IXd7JRgyLDQB3M9dDsHFRGozq6FTpwNwOV/V/3pIlljMk1+bk/SI60Y1hrrAjL+pcwEBIxNObWVH1wsW2JQGbRUNabsQtMLlEB0CAEYEhCBIwZxFBnHAYTkcBFCnuEiiO3q8uIC424UhwxxdVdbDmemZWVXmS7GqwzYuePHSKA4iFo1hIBhGBDF/gIiXIgNzfIVdaW+mfSMoT/XIVdqv45tME/51GFcrE4urvPSYynkuq+OPGV3+JJBZSeMQ5Arp5GTW/82ao1Jw7RUNUn5OAPWhvIVbLKv2K/Lc/tjrM7Xlb9zoLUZsDrHW9NeYnB7YOjAnG4pHNVuyH0wRyeyUJglIVF91LpR3os2BwVwFUwLmWkRyXyGAUSZcXlXd3KMg1XgUS5AEqozADYbj9pnWG0+cQZHsjkLmZE946bkiKFsjYDk6ccMvthhDgDfjIgE8MWEGAjMl4jjjBBSOhe7HRgwh4snn3wS4zjiwYNL54QRREhzOas9VBMBUEi/grDPIM8bQOCQwJJjxDgOGMcBF7sEmjpA+zOeegIPHjzAEw8u8eDyEhfCtlQYJzucY03TZAxrmiaxYSWX/+SlOCFGbjwL5zjnpVgqJw1y79wcQaqOh7Uj0sayTxIcb8M6TW2fbVUuwZJSyTFRLt92jPHPtFzO2x0gvC4hDwl3DrSA/cBV/y6O19/tAhDS3vTPF5S51ZRQ8j27LczBdE1JrUQiEDfMi1am0QqJnshgvbHunRZ7KkBMSpzYGGn5lWG59FaYlmWgdiphA2TMQVmAOF8oE4sRILJBtRAhrqVTV/xRGEwcRwSk/QzYfjePiESAsJMQEiAp09rtBgGrmulkllXagahoCwpYQwhSLUk4jmEADzF5Jo4zpt0OYQhpHNgQ8OBSwGq3w24QpiU2LR1ErQ4rfgmUYtFJtWf5wcfFlE1ukHEsGVl6zDW2Ic/J+j5FKNdf8AFCet93dApwsQGSnSjbuxx7L+NedoteyEVkzw77QNX3avagedyzngsO7wxoaS+yPofO+VPy2Le/3eAay0PJcCs/3dY7Xfum94gDx7TIXJfhcs3ARcLCHCBBVWYKSqEBqdamJeUx1ZpzJCACzHMwM63EsrKnHOAXXZzAcTZgGEMAY8RwsQMPAUOMmIcAmifM04DAEXEaMBAQ44xpHMzmAzCCDCK+vLxMADaOMqg4VUZEO1A3u1srMwFGSWeQewcRjPFih0CEJy4uQQyEkFR9FxcXGIYBn/mZL8Hlgwd48okHeHB5id04pjFwQALYOQm8G7Hl3Uw35XRVbu5DsDAt5jQeTcDNj02LMWJ2dq2CadmrLzsiZau63S/2YQ2qzfkdhLX+zpV097EtB1hc21UebrgzoNU3gJShZlRU32C9s/JwU5rwfKs09x44b/nZwskMyxjdMnCp51V+ug0vgcvDowNJV8XUP8gvrZBP1K8L7emoXYfaHmw9k0NiUJUrtlNbKSsDKJ+TECi54GMcwDNhGMekppxHDAHgmx1mImC6wRwIiDNiTCwoDc4VtiYqzuBYjhQ2MyygYSc6q4I9q6gHhyGYqhTMGIcRYODiYmeOH8yMi4sLjOOAJx48wOXlpTl/hODyj2ygWU6662b9iDEPFfD2rGoMXMGwTOXKsMVvGmZRv1zXHirAOocWrGhTVOtb1kNvpgw5oRG6eZ7u9l6Ws+e97NlfO45sJfXufedDubsDWhIeVgegx+zW7zjlS9mS2/HUfXuWfQBr9PJr6S5FtWvaa95YRuthk/ptFwxL9321ICrg0nwTs4kcMYuNhgkYAiV2IEyrGK/lmBbrTOwxzWE4jAMQCQMRECPCOKRxSoEQpwkjkqfhjoB5nnAT0jIo0zRC196CVV0CLd/RiuJqPjuQjaYrzM82iH1qHAcMIWAn9qjk0g4EJszjiJsnJ1zuLjDHNK7riSceYBxHPPXUU9hdXOCpJ57A7vIigZ44ikRnh7q5vsE0T7i+usY8p7kWbe5DV0/TNBnITeptaLYtp1Z0IFcvFOp6IFhqNFzt2zNHqAofIeO4jdA6iZyjs31ehnvnQAsMG9/Q2KhWbFbNQPG9jCu/CKoSbXW/G4xlepn3MaStEvy0sJjKgm4cgOv9Vb24dBVtpZaHzfHGYPb1poqzUNkikhJuubLXdiEFJKLGuy3f4+xaULWiKTfFlhXMRhHA4AjQPCICYtsC4m5AIIDHETHp2sAhYo5BspH6FDuWqUyjFlwHY/eej41V1bPDDySsbkwu+xcXu6Qa5AgC4cGDBFqJZY3ZxZ7E85A5O59EGSw8Z/Ap7Fpq65P7/Gwg2ZOwnWmEBYR7ffhey6vDOmBtUXWXgEWu47Av/iEAsF8DlzuKPt218Z7rZsB8v/copNVvs19Snz8jyddzQODdAy0NR5KO9dv8iPt9PYj9zGoRmE4lZGcPtZrFnfNn9qor2lbfPmZ1Zs9LYLcV8ZVpFWXOsLlWunSL+9CiTGNk44jmlOYc8gBhc3GvBhsD5imoNhgSh4ekTmOEOQA8J4/BaUCIE6LYtOZpSs4Z84wpDDZWCUDeixCYvTAXRwivuiQkTW8AGaMKIWAcki1q3I0YiDCGxLjGIdnQkvkuWkfqgTCtBw8eYBgHXD64MBVlKk9idvOUPP9uZP7Em5sbTPOMm5sbG6vFDrR0LsR5njFz9iI0z0Edq+U6Cn4Q+EpzeSwDdeXAwxcOqlIu8hUhacC6ofJ9N/7Ud3V3QQuucprePFenc8+J3IWmh0HtR1Lue/SBFwEqezSpwofOruE7SyDt2S0XjBvKU1yFNlcb+FrErpp03XE9yjaX7WyaY5LrbB9ht6SmdhIQiFnVpkyLVHCi/ggdK5C0lB14VSQhmNBFTN6AOv4KzDZOy7wFQ0zxna0h29HSkyZPST1uua06XySClqeC8r8HAS2dbPfiIq/bFYhweZlc68fdmGbCUMASdWs5A0Y52a2fEDdGnfYq1XVvHseclrNf+eP8wqq3nvcEiPaEc30Qddpcb9+0jOJNN5f2hNNVbNWbtJWM5eoCw9qa75JNa6049We5BGDnhtk7DVrnDvsBa99dB7y2FxzT2hd8YRWge4zrSBQ29oT2C1m9L/3UGRS8uq7rRu0EoncTj3PENGWHAhCBZpl0tskyOQlABhDriscawiBMh8QhX1zhGREcCJguEEMAT1MCjmnCHOZcNnGbn6PMDDHrnIhAkLwDJQJDXJWNyMZaqZt68vobsBt3GAaSGeKzJuHy8qIA3IvdRXJ9F2/FMCabWoRginr6FYtYllth05KGoqA16YwjsVoSxY3TkhdUdA7q72WtpfQ7S/sAayHVF2IH80zhqHFbD6ku7hBo1cSTOr/6oW6u2iPTo+J4D3LlnobGLRndYslrnbgXuMcAGK35MW1ExqVEyLHTil3medGcKtCiVnlym0UWJv13uaV4/cCdX5lZcT5hoBUjyzpWYpuShlK4x9t8f1oeFaZZsLKwK3syqT8KBEQCAoEiITr2wzr1FDMoBIAjiJMXHwWYyg5IXokR+VhtTIHJptgD5YHUweVTM63gFlGkQJYPEWHcJbAaRhnEHNI0w6Q2NCim1DNaOMeQykMwVVEs6rwej5WZFgy47J05Ol1oSdy2xKcsIqOcPKJWmPhwBsHc5/iHBH3Pkt6KTSszspVUF4DKOqKNrR7tN+xkw9HyayWE9SiHhc///M8vPmrd3vrWtwIA3vCGNzTXvu/7vu98BXB2jENCUbcOsIrjDcBVqqQ6+diFjWXcGs21yG231Ny+vov6lw+s2vI+V8YFNubh1k3EAw9k+RdnQcU+laUgorUYvJqN/rqysK7ae3OTVg7W8UbJNkXmJj4Og81EYQN0ATDPNqM5x1mmdpoTuyKW6ZGSyzkNAWEICGOaY3CQWSiSk8MgA4lDVsfpRLVu844UQdzhB5keaiBZ2TgEjGHAbhiwG0bshhEXY9p244iLcYfdxQV2uwtcXl7iwYMHePKJJ/HkU0/iqaeewlNPPYUnnngCD554gIuLC+wudmITG5BQlKwuTR04R3PE0K3HvtI2J0Zb2LLEjqUrFlfAld45gVjGntneXnfRplqXqaLptWH/J/HwWNbat9fo6c5XsCz6FgqxhdaeOZydaf3Kr/xK+sAl/Nqv/Rq+/uu/Hn/1r/5VO/e93/u9ePvb327HTz755Jlyr0Al932XY3cVs+g2UOpdF6Aiyj0o7d2yv1l/FQSi6tI1RaHcC7TQPo19jPucO5oMco8pX+5/lV0NXTGvYv7RjH1Dv922uaczaZZt19tHZg5pq8CZnc2wsHdkzuZv6FYFlBx5RsDF4NYos6sTkc1IMYjQTrOrZ9DK6qvoeqiU1IBpWlxpMyyerjJomuRZg9tYGZn2gDMl8EwLQFqGJcqchtYpzg89hDx101CxrdrO5YcCpDkNk+1Nz/kemlWlkaNSjafq13YuQh1PljoT0dSzVUOxOtVjyZvJXYdrI6l5aitiaTcF07JmJxZBMstg2aQbPHAA6K71dBsLVtNcT/U3VImD6tOw1ZbrVG02eDtObbVmWNWnulxOqQRlWOW4rZZpLYjdXCh/24aZc9bC2UHrcz7nc4rjv//3/z6+8Au/EH/uz/05O/fkk0/iVa961eY0r66ucHV1ZcfPPfdcG8la6wGVUjMn/3uhV1U0+Bq4tue8PRRfUXtidQCxlXEBkBZvWLitLkldvup4sXOI9uNL5ykDlRj4yc7737APK4UMmip8KEmutNZSyLI2v8tkT9Jpg5QhJA+3Gc9f32COETdzsk2FMGAYd7i8eICLiwe4uLxMrKNYxwowe5bM9AAiBA7APAJgcPDvLUv+MFBaimsICJwYGEPUg2CQuLSrgAwiaIeQBBUPAIcUj5kR08yyVo/jkABrJzapnYDvKLatcbdz6j8ZbOzeQ1AAFLkbBWGIoj0PRwWq0psyu77LTO82DRYbHqV6SHWQzH2s2G8aV0AAiSkDE6dlXGT+EwQEaBeiC1hNk1Xg6rRJ35Zdx0Lvaxq1u0/TLdMhrPp+dz8ORZGl+BXC7ctg5bKPUzj/NFJuLRFUlU6Hy+hOOLt60Ifr62v8f//f/4fv/u7vLoTrz/zMz+CzP/uz8SVf8iV49tln8Qd/8Ad703nHO96Bl770pbY9/fTT+zOWj4wJ5bQ+1fX0W/6Rq9A9+x5gWRdQhORyuaqtLk+9qYg2VapeyufS7W5ALKFNo4pX3N+cQ/G7ri59Dq02/9xE9aq2/cf2VeeesshfY3jM9aKiHHhQzvku4tOV0Ze90h1x+uedMLIjwCzLh6SEVMCP42hL3Y+DrGMVpFwsEtd5z3HURRTnarb3vLBhfs72vQQK8O/f4rn4obMNJCrCEOS3js3ytjgyZuXHbIUQQOKwYXvZEAj+e7bCOzuU1a1sjMrhhblwtmjc2u3VVEK/lpOeabFvW7m1EHxLgjufPSqpc63XcP03p8e2odz7+KDErIxdVd+T/zAaue5PdGXAgmCq5ImXGSBHXKtgRak+1H1yiHxebaW1lXxCuFVHjJ/7uZ/DJz/5SXzXd32Xnfv2b/92vPa1r8VrXvMafPjDH8YP/dAP4SMf+Qh+9md/djGdZ599Fs8884wdP/fccxm4CkFd3bhGf5r3TOUe9fHSffWFhxjWei1OyB2XfnvcNGhOgLH0AXSTpcqttq7izutMgjULbLut0CmlTzF/J8mOlFiX66oDIhxLkNJBsXm9p1SYIQzYPXiQZoJQ+87lJS4vL8T7LqQeoNrI5hnx5iblQcl9Pe4GWUE5pcmka19FKbPMRu+WBAGAOKR5CzlRrcSwONnGwADJNEs6Ye1gs1MkkAhCM3Vi3VGYlrfH2cS7YjdLalCZWSRUfduoY6VihxAoO6ocMdzQAR1zNqv3o+m1kvCLMabB1v6abyTsN670iSUjLxtXvzVKE+6cl1ZNKEAs59BJaF+CWzL25xbTWXgO/aaIqjo5IfisFlnexnTOVKRbBa2f+qmfwpve9Ca85jWvsXNvectb7PeXfumX4tWvfjW+7uu+Dh/72MfwhV/4hd10Li8vcXl5uSHHjuSDawON4N7T8HqXNwLE5pezhiPN9bq3qD/2J1T0Dg8qDnV2hU7OAIxZ22WtH09n7SPXd1KoHFxinMtptpEiguvF2vmyh0+1HCuSEMDSP8ew1Nif7VvaHwoYZKXgC10A8SKB1TiMyfakXVdwOY+esam0eCQwIM4yK7tNzM62FX0wx6Ki9mZd/Yj5DzJJVZpQ2NUISUKazuAZFWWmZZ0ATRe556zMweOFGlHYFb8BEnSO3evyNkR9zSTvQcGqJllN8HHcu18CLH26TZ9n1RYLpoFcT6kYnRRr8MHC8RJw9a7b+5Hn4OJ0/qYMuOpCLRVoIX87lLTs27EMi/j73tM5w62B1n//7/8d73vf+/YyKAB43eteBwD46Ec/ughaB4U9gnkv8aJ9F1/A4VgGdY78er3Cbo+qD4f9TiTl3jq8zOuoNnu3s9q0nG2r2jRlP0Frsc3JxZ2QWMy4G7G7vMRLX/ISXDx4gM986Wfi8sEDPPHEExh3bi0pYVnzNKXt+lqcCyI4BEw6yS2n4zCOCRgYSJPsCnA5tR2YwYNMrzQnG5c5nCQpj4AAEIMQTOgDSIwOMNvUIOq9Ycizuwdjr3U9qZorVG0s124SnCzFzsCbyKM7Lt56TsvYVEwOKyzlVAeNGHMHwl5bEST96PLtxqvayL4oClZQpyD0mu95wzLVO4hpbbjx+HALSR4bbg203v3ud+MVr3gF/tJf+kt74/3qr/4qAODVr371Cbn1GdZirA6bKuj+GiE7mCFVeTTFPe2rOPZ+Upq0+YaF3+4cdT5yzaYUfXqtjFz0Wh37SOXt2Q9cusaSdBNwcjKNK8GmIKDlJJDNaq7CfdztsBOWdXFxgYvdRRqYq27mvucpNi1jW+JGD2bM05T2gYBhAMs4p3yvFMzYQgYPZUUgFgcMuYcyPOv53NuWGpZ6yrYrV4dV/ZXvJvfe/fn6XRljLahXP5S2j4rh2qtTtWBmwuAyf32XxRyQFUjua9nN6gQEWxNOwcozTGNaco7kOXtPWo5z4l7FSZJLMkdZo5cXXF/2FKtKJD1bFgvyngpnjYXgi9RLvuiD1BRrP6qdC/duBbRijHj3u9+N7/zO78Q45iw+9rGP4T3veQ++6Zu+CZ/1WZ+FD3/4w/jBH/xBfO3Xfi2+7Mu+7Ki8vP76YNGtbeJhs5UXTDgjYK3c1txS693d99ntdIpthUJafZiosrN4NHJCLHmzySKMOgNFtWSH5quANSogDSPCMODywRO4eHCJl3zGZ+DiwSU+48knMe52sqJvtmXpTO9xmtJsFjc3NmdhGhA8J0Cc07pUgTmBXgjiwDEDUWfVgA1a5kCInBwpWKaASlI7VZoyqyh1qscmHAUcw5DqbZA6VIcNQtJUqiDWLZtHHDiyGztlzial7aplWq6TUTFlhqz/BSk3k7m+674BLEEsduXwKzRvE42dRtxjWcjy4XAZc05qcn4Z1Zi+tmTxAmBctwJa73vf+/Bbv/Vb+O7v/u7i/MXFBd73vvfhne98Jz796U/j6aefxpvf/Gb88A//8JlyXqn11cs9inDA8VrYyLC2aiQeGtg65FkqW2ax7dXijNi4wP3ycyf1glnZpvd7tVAGrMWtmXEhMzhVx43DAAwDRk5jk5IN61Jc23dpULGAVSnoY54I1oAxiscgMN3AxkkhMuI4gcKQWZOnhFo4OEBVYbr3vQvddS8qu6wHdBmWz06rpSmLslLY+SWW4+u2eY8etDx4Oebr2VXPHparS+OWtskyqlaE32utup2O10Lel0TIOXdIUsY0e8yyc64YlrVX8K8woubyGoMqr3sG1TV9NXJK45Md95yovPwsOoT2mZ5HXt0KaH3DN3xD90U+/fTT+MAHPnAbWaZgguzRhq2gsxQe2hNsKegKjpdXOh/qQtxt3oZOSUbqvi0zRYTc+1V25dWCNjbIZmdIs1yk9ZmiMQQAsuT8gN1uBA8jdruLdG4YMYwjLp94AruLC3zGU09h3O1webFLY7aAxHrmpPbjeQJilHW2JvA0pZWHxYvwhhgUCPPNTXKZJ4CHEcNul4RiTGpEYt2nWhBnQ1mzKiAyOZDgdAyYp13z7QkrTZPbyizzxrLEpuaYEceY8ooiqLw6D7C6y/Y/v3BjXm5EXwcBppY013mndkzpMjJJE6aF/EqL4DonxRyF4s3YtqsauPpBGRZVgw9rx4vVb2aPehTIfYrVcGuMpgfiB6agKuit1zdoJreGuzP3YKGL3dOqqGJUi/qrY4KqQh5t2Aya3R7VgtrEJ1xkkC/uq/fyCifX7k481ql4VFBUvfNWxVT28HPP36uRnBwpeoDqnRfAITEslnn90oSyOwzjiAuxaY3DkAboqvOCYx7sBL6qIcGcGJVOBMsRIRAmEIgZ880EikAMaWhsYHaqv/xbgVm9B4NbMC65X6RbzBmxqXwVuum/1bslk7rc+gzElQOEi8vWMcjgaHufs3p2wIHe0jvUjfuMyZchexv2mLSCJbuyLLTJhh1wbtYE1DM3eDFRlMwfLAjxPGUZQZtM7gJ0Qp+o2rliCjSpi4IhL4TCK9c9SSsPuYqv95d5LGqKOtdXcPygcHdAa0uo6/isgHXGhE6kakfdugRYvYR7uHZQp229hH6w6zAMGAcdyDvIHHulisvc42MCCgUPnjmtIDxnW5bOpBACpXFTu2TDGpWRIOW5u7jAOI64cIsfhmHAGIYki2PiAwZYwrQ4zsA8A/Nk47XmGHFzcw0iwjQOuBlH7CjNtB7iLPMDDumZhPXoswUggTwREACGeBWyrmOl4AE5dvXLyNJGgcMUXcLmFKBkiyyzb3BMcpZSfF1sUsd/mU2LId6WpdqvVummBSXTmDWdCJhCSAwTyLOruOdht9fBA1F6JDpPoc6uocMLFgX3puatnYH2/BJwbW76ZxDcxSNwud8HWN20CCWYdHuRByW5nNcZk3tRgVZN72u99cOmSO1syn31w2K8NsX25l7YCt5bktMqlLjtRLjbipCEjb+aHC90TBEFZUXea1AEjOv5+1669sjLfHM6gQIQGENIk75GqKdeGpe12+0EMIfsHg73vEI5dKJYiA1LZ8RITCvaYON5mkU3lOJO19egGDEPAQgDwijOF05SlxwJMKsLmWLUmJaG4ItXM2JWDHNj53xEBWD/bjzL0vrt1C0qkGpsVw3jQglynIAr26pK4MqsecGW5dkfJRaqz2a/1tozlzVT1V6raJCybREdDfdrGFXvV68U+2M1dxUMyzOmdK35Rsi1CyciM+Mq01vLb3tJt4UXFWg14SGD1O0F/YyOeKB9txyBf91upwq9bUkCkIUKMWAYR4zjDuOQZxVP3m+aNil2GEgoUCSPQWUQDrCGAOIAGsfkdi6LLSImFZx69O3GC3N5z8vRi+1JnpM1T47GtDBHYVpp45sbxHnCzfN/mGxbIa1dtWPGbrcDTXPysn3iAYYQsFPGpVWqQkDWJGEdv2XVmxAlCrAkxsJQ01f9HojzCyEkHCVhqAhI63RRnsQ3eqYGEi8/GNPSa7qCcZrJgxCGtPpymJP3o9qywhAQwaAwgAYGRfWI1EkGc2dEASlpWhlp+kVhhLqytM5oogBWUJDc4pKABtZaYY9JeeFNVUwff59wXmJkvfMHKS4ODh56by8XDdbZOVN6dxq0FgegVo3W4h0iVTeFsye4EA7Lo2GcG51XiniVNGzq0F/vE8iqu2r/rPfnZ24oZ2/wYoMEPJD3TflSXn6GCQMrV25KfgppuQ/Kg3DNeaAoK1ymTkXIHjRLr8U4J2cBisDMjOlmAjEwjTdJCO9G0DAggsyNnhZVXa6+hT5RUfd51vIicMf2KEBX27LUylOYQKhlNfoeMrsKIHE6oVgyrYJZyfWGnSGzpGLGEmiearPifOyYFywe1aYp9+xUtDkvwJuv1tlfy+TK+D2GVsbu+8/1PhnJtsj1VKF/qE3rhRruNGgdHIp3dw6waZv4VjXCuaHu5AHMh1KytU7cYnLppkABGFDMQm5LgHj1hb/P2WXSqSTQ1D4G6dmHYQQo2IzkJHP8kYCFguUwjGKLGQrBhaizV+QVir1NK07Jc5CnCXC/4zSJDSgihoA/BGEaBoSYGNdAsnrwxQWYCEMyYJndqRjky2w2JmVUASrcBfSx4KNZAzszEIFIMbl+J71RYljGtLJNK9vOXF0T0rpgYrciIgzzIDaswWxZfj+EATwAQ+SMIRAvSOeFmObiB2KyZqU0mPP6Z8W4OwdcS01MQ9NGW+akZ/2+TabtIHQZVc18zxwOTX/dpnUgE1sQXOd+7hcVaNWCuyeIm178OfNfOt+Ua//1g/OV+0n1HJbPRoYl8bbEzvpy6ncfJSG1J+bL7OWfldtmIrceus8nz78Grj4vS4ssr7TOVJCZKAAeggzmTcGWt6AEX34QbGZZWVBbzz7KeCy1belS8jLoVRkX6b0xIjIwTxOIGdM0gQDMNzegyJiFBQZ9Vu5Yn4gc8zGFWvEe0vmlt5bolWcn5OxIrgqdTUsZXFXNlOtZJwdmsDCuvm3LmBYnhxjmAF3ixGxdAp7g7A3oOaCpD/MZB1iVBK0ZV/OrBLoiupajTQaZ0e5jalUyVf3ulehrSIjym9siKrbatGpG1qaTM8szsPTS6RT+BJF2R0Ar9QRJeoQPN5zGix7uuLLD8zoZuPd2W0lkUvtRkC6pIUwryNgsM+gDRcLKPkySAdB588IQAAzAOIjw3aXZJ5SdzalXnxcOdJ5vXuK4jT3TEpd2lpWK45TW0uJJ7VoRNMc0NolFTYiIa3HCoMhpWiggeRPOM4ZhAO124kE5oHh3hQDNjKqQb1w6aJSAbpwmpRKTDSuKmlHHZrHUXxSVUdTbK40EiV41OFAKc8Bwc5OWvZxuQBwc0xrAIGFYul5TtLXHZjAIMVNI6FRVEKYFBKcOLIY6wLUn1ZeuUaUNgRxw+WdXJs+unEtintwPh8fb+cyBxGc9obMl6Oqgdy3nyWcQ0XcEtHzvfCMQdGmsCAOmAyp2OeKiTa1u/GdiUucOmWVo77rpv/pOaMlSbfh/py4Xisusa2BlF+lyJV0PWGjVXFUWrM+gmzgw6KKKCEEloGNmdfG4ZHEVcJmwdPMNwk9t5KczYmQbldq5IIsjIjGvAGCyOQrTTPCRMpAWVUjdR/cVY7sC5g0LnPQsmJfytvKt2xm519RiMsu92v3yu6uYlTEsGVwcKDnEUAAFZFCjgEh54uBCwBYsF4Uti+E7QCVgLdu38y9GuQr2Ulx9Dxqf2e07HYicQGYwBwGWPoPOjbgUbaMYWLZp1YxL49Pe9LeO2zpXuDOgdVBoVAWLuoMXfji2rCv3Nc4qC7eR9ML7yMTl4Z4CcBV1kDkA/UaFPUtv7CUggjIEgGXhQgygOIigjjBXu8DAJEnYlBJwa0WqoKyYFjuVoKxQHKcpzb5R7VlnjZd7bVxSnBEj4YauwcOAqxAwjyMCM+ZhRGBgDgNgjKtUWTbBM02vO6srWQHLMZ3EbRLzijEK04rGtBjAHAW2JP9hGIzNEgWEcRDAIWCeEYYRIUYMg9i0YsQAYJDnj8KwWOpEmZbOjDGzzPZhoKqPoe/ArTitxw5sDw0KOHrUVxEs31cD12r8ipkdyrhOtRXdto0thfNn8KIALf+RFx/8Gnj17jlTOU5Jt7mvw3Q23b903770nNzL7dHbmnJvslFjd57X5q1LB0WvvMe0MuOowMSKYwXJLCsEgKOAGCO5CXqKaNSjSC8JRmQQ2AdcflZ3P8t7zJ6FXnVZgwnHiHmeQSBhWolxDUNiXBwyCzDXbe4lhrI+5XIR3zEs4856bIfi6CEMJi0jIs+iKkDfV1ENh2dToeMZ6JhWsjGWx7EX37cr90DNTBisvXo2G0ufQ/gTZYwWuKp7nKA3xkUCQBXjWgotYGmLrksp75vzfSCYJsirGA8JnkyWNq01xrWUnvvm26tNeqeGFwVoAapOaRvSmqA/CbB6BOTEcOtAtwGwmt+9/JoohLYyvBDPsYnSPIO7cUyT1O52ttouqXdCV/jLoROOwABEZVpziqf7EIE4J8eMKPMtMGxlXpvMT/a2Uq8BlUzPNKcVj+dpshWPp3nGJMdxmhHjbKrCINI0yLOSgNbN9Q3mkDwRh2EAR5la6jId0yVlALdqy8yqmAzYgLbzrgpp56WfvjgBMLF1zbKS8I0wSptpPwSZxFc6BmFwAARbBTkMAwbmxLCIMHIG4CBDA3Q1ZiLKtq2Y5iOkwCCeyz4FpzFas1sHjd2aZPk5cXBYYkoMx8D7+j8oBC1/HuWkvAx9FesM7fzBl/dh5ntauNOgteiVt8J4DmVEp8Zvz+9HuxStc32FcR0LXPXI9oXUu52CpTLsG03vt0B1r1vBz93v5JRLKDMtxwAMPEl6q0Tu042F0C+Ev+/Bwxv/Y6GmKpgWc7G1Pdiy5nWg7jzPticAs4CsXk/eeRlfUKXPXmjDyXBXN/BMyuZP0v6+2K+UcdXPwV4J13oGIhAgar2CbZktS4AtJKePlmnJOC9773DvzrXFTtnUtpUBKz3XWn/f26n02AR52dREW+ksV8qAJa/GjMvF3SXTohWGtqHju7Ujm785Zexleeoaqr+pnE3/Gybf8elc9+/vlL78nQYtDWcmO6vhdIK1dvdR3cezhGOS6dWHeRstSBNyNi1b1r5YR6tiXHqKAFByn7a9qgXDIPtg+zR1UECabE8LrE4U6rou1zgKmJVMa54nmfFcmdacGdY8p3g6qzy72dvJARdn9dsNCHGOGChg1lWGhwFhSLPcq5CrOxVxVluZjnHKz1NUlbhwMWLBqCLpPgHWJGJ7kvqYxHYUQKAg9TcEkLApkvW6iBWUkudjGEYEwBgWg8VJAwghsaPgGNYkKxeHYcbALGPzKoGuzFJnm686Cr5xlJC8PexnIL0vYWsOZYk887KOx7YCnlvzdnoG/tGq02lfOhQdE14UoOUbrK+uh82wTg1rTMwaw75i0AFM7JjQVcHuKYzo6X2krh1Ee/E+QW5Scs+mdjbPzvLNXB665IRDeFZRMy3Xq1egYebSIaDDTpbrxL0Rl36MDKKULlECMgLZcQ1adV6LoFWehs0+IXEikn3IJqitWUzPRqWzi9g7cgxMZ74IGczAsL15Dy7ZtCytuu1WzFfrwgHXoYC1l3FtuqdS3XbqWyIr38nMi2pb12KGlsY5gKv9nLZlsKg1aT74zgd7QnhRgBawDFyPZzjyCU59cN9+D/xY9qlEi0Yvgi6EgGEcZRtk1V0nFH2ZZCNlTKZ2W4BnshtECMt5VszIIJS8A91kuGCboikK05qEac3TjDnOtn5XdsZI6ZnXIGeBWOG1lYmR1YE3NxNiZAS6Tm7hc7T69PWa1YK5atL5cp/rME3/G8WjYiYYw4oAbmQ+PxnRJnMRCvgMA4bdLm/DkAaAa5kkXpA10NhIbnqPMaRnmEOyE4Z5xiyMbxD7YrJhJlYWoqqKc7NT9ayyLM+20svMj3urhOTQsKgazx2vVeC67XDgt73Prf3c9f+iAa1TwrE2qmPT3Z7eI4Bg7UhXWfeKvOgtiIw1bPEop+1sJJ5laW97qWC18O8F7vw20xVKdtS15dRMy3kJ+sURSzsWF8ixPFC1LJwXzqoCA8TWpXXin3UJtIr0KMUjEgcTRpRhyFGGraUpk1gmoE22L2NYOtu+8+wMgzrICGNzbKtkWmkfQgAzm23OT+9EMRazoJjHaGHb8qpO9FmxPHSvo9oVoBUZ8IJ47XusmVjLaNkfdPMmKW0S8M4bcam8e8q+GI3KiHtwxn+OWP7oMtPqdZ7W7NfHhnvQOkPYAh23ozpcSvO2wOwEqrUpzawaNKY1pC3NCtER7qoCpMwECjWiGfBdyDItM6uYve+i2UoyMCmzqvccOdmvOLmsK7jY7yU1YaHCcoBt/Euj5ZnUb6YJIQTMjmlJpeWqQD+d4rmtfgKY0uBlRhq6FkkYF4CZZMLZIKxpHJOt8fICwzBivLhIM+DvxrSQpqytpfVOIYj34IhROhwq4JRxBWVGlLwUGcA0z2AGhmHGyIybISDIApokk3iwq0NziJEBXlrF+vTHttS9DOKYT2xFeNdMK524Ta7lIf1MusaHEO5BqxNqgFntcYk+fY053RZj66R0pnT2pb3w1S7UVW3XYapErIBLO6anYlxVUZTxNX5XSj7Qforq2u7VZix2HK8eXN1iOWmr/i4cAWq7hmTY1E9VZ/bb9VYNwNyzNY9MCoIEmZ+iURNamrLmVKSkFpyRQCoGykuTIIEPxBmEhiF1JHRqrUFmwCeZ/klrXIGLyLwEQwiIwqRYgEuZVnK4IAwki39Sy7i0fTTMyTGrWhfaE8f7RHNmcinNHnCxVbard72njrfnuHhfWk7SdBwjXqCGNcFeCi3jqb/hQ4Gr0hJU6d8Ww9JwD1qPQ7gt4vQIQlbr6PLxCyxLhGPyPhtsraa2HoRJUWdTcaBkizN4MssaWKbaS2N+OMY084OsiMueWc15n85PSN6Dc7pflh5JzCs7ZmR1oZepDIppOh8EFhlPxlC8elSfJzEuwsRzYkLzLNoxEbBBsSLBVdC5F7V+pL60xm3mCZGLUZnVLjGqQVZq3u0uEcYBuwdpf/HgAcIw4OLyMgFZ0PQZHAkcUiciDARGwMCDdD6ANDEuEGICKT9OS211s6xrNs0zIoBhjhgiY6AZc1qJyzAq63Zz5ZIBFtn+KB5x2wSEqn2V8bKHLXViv3jCPWjhnAznPOk2Jo8VcvNIQkcltRZMpQMYYPmPbclrLDPVnE6Re6EqWy+U8SAFEiiQ+W0744o2qNexLHbebCh7w8377QTvMefvg9qamB2oipAW1Vmg5GUYZKyTzl2oibFsClo6TVMMIakCXY85qQWH1IkQp5hhGNKUTWrLUnsWub1jWt5rMDGt5DCjbM4zryHm9dNC6KylRq3isxjKa8DlHpRa4Grrvz6hTAeAzUDRh4Y1RtUNSqDIPYv9yDYtUxFaU6AqAfc9HCgb/PfUu73/uOTq7uEyLA0vHtCqejWrKrtHCBCLAq0R2i+8YIKhowLT65Dr/kE1ro1JCiIohWXZ+7Ib0EqfQmiq4Ozkbuo9N3ed8z6bxaGiAIXKtqUMK4HHXICIMq7eOk8exCLrmC2uCKKUPWTVnMlghpSVcX2TZqhIjCvF8ExrCEHAKzM3IE/0MSuwCrMLuxFEA0ZdufkieQZePniAYTfi4kFiXuPlBUIYQOMgs45I2YjABmIJ7EyQUUg2spjGecWY3D+UaRFRcvEHMKkNb54RGbiZI2YGhimN22KpiKCCvGJaWdq6Y+oRp8zEusEBxkmMy7dVTU/T70YvbVpJJeq/eirK45M5Di/qB1x64BeG1LmDoFXaPrJtwK5uA67OqboHcVsMbS2s5lqg2kOCOG+vwlrdpDL16jOrCH3vGi0AmRCwm2W/WtCC8bETgszcOE702df+va+OPtNS1Whb3hyPSmlkz6rsMHv4zW7l5ARcSc3GURfA5NyjJ7L7FbQAyEztUs3iQOGHHBQMKwTQ4Jig76qrilPVnJxmcwcTQgyIYquCMMCgbI7ZvBGHQJiHtDRNCAl0B8+8LMvuR5r3Dh9MznNd7XtkQMV0vKlpLbAvnl9brm6njiLl6DkjBSzyN/i2XwFW/zGWmJDmK9/jInDVoTz3sBiWhjsEWlltUAhND1bu/FbniMcy0OLBIw/6fpIqDVnAqUpIhKXasfLyFmupohQEPaCDZQ7vrh4bm5YyrdJ7UNV/6mxhzMoYWb1nY1QsAFOrDRNTEIbEMgcfnFBOa3YUz8qy9pQC1mQ2tckAWJmWgoIXi2oT0hkoht0OYRwxisfm5YNLjBc7PHjySYwXO1w++QTCOKZ4gUBjmi1fnWmiPEMyViGtBB0CBiSGRVpmpNk3JgjDAszul0EWmDmNtZtm9SZM47Z2YitEVHugAqR7t36/0PZUVZgv7GFaHqlqAFsIi5fqbBbbdMogj9fKDZvz5aYTtzTEpHZHXxtX9UIPdwe0vMCS93EscD3qF/qQuNGtBuEXheql/tb9OJRlW1baqEq5CB6cauZS/fZbeU7BRDcZo6QqvS7jqjag2APG6zIBkHNLKil299QPySTlpLLsCmBqWzO7jvTSg9i6NM8ovfU080USjSTCrGRZ0oEYR5tGyjuJ2Pgt5jzwGIm1pVVO0npgpDNfCPOqZ77QcVsUCIHd+K/eRjpea0lIS62zQHOtGjSCUvVqusBVMY4auDqxu4GaH+11LZyklICem5ZPxrQ6xazy2AdYZIBcEuUUT467xT3kwW8nPP6gVQmsJffoJcDaHB4ikhyVzWIj1suPGAar8nn1mQkvsWGNKiSDqgnrhJLgrgc4l3rEUihxgVS1I0U5U3i5zzYtdsyLOS4zLQGuyBXTcnCZwCdJU13NNSKvNqwrQyVfuazCsSokJY0CWBwxuWVQCEh1h1RRZieBrFCMpB5UGcgEhN2I8WKHiweXGC8ubL+7vJC5BQedMNHACsiFVvtWAIFDgmbEmEpNaaZ4naA4sVek+Q45z+5ORBiFYe3GZCvcjYlh7YYZcQQwzyBicAgy6S9Jm8qdC9/I6v7sClFq2tpDCZ3PM78zf7w9kXVGVdfI0n5PISU8TNx6/EFLQhZstAhMfYbVVnfdO5Fku9ePOW4z3H95LdSMsk2eiv2thApYUpHIdd+Qu7lyItt2sst30enwG+ASSjsdFmRY1Qv16+XyZ8O8WMEgg8Kid6FeN2bW2Ve92bpX2xaXC1ZWtFBLsy13Eazu3Vg3B+AkrA065ksZls5uYRPgDjavIKlnoVMLalZWTqJytg8igEIaeGyMS+cedLO817O6h2T/WmZaMhGysnDXxrINcQGwiKwb0A2aWGX3cYtauXe1cO9mEU5V7I4sct+QETwrY1nEJt0FVWBpg8p+ldqpaW1cy+kX4SEh12MPWmrJysZ7ZfFbBfT+ePVLPztgnRpW8fD28u+nTN2r3vGg1/vzs7or09JBq3vLsO9yjRSdy1kdKFMYKTPq2Ki4Oq69BmPNvHTuPk0PWZCygIXfzF4PzvdFhs6nCJCko4Cl94Y01mtIdRwg3n+7nXhj+hWPSVYgTt55DGAQL8Hh8gLD5QXGyx3Gi4s008VuTDYseR8g1ykUEA0C4EEBg2Jak0yYFoOBGIxpzYCsnwUQ6YrFeUXjNFdhWkNsZMaoTEv2kWYgJhVjNDjO7IrNBlkDV81Ujvw2VgnYIQxtCRg23PYiDY89aOWgva5lprWawoJt61Sgysf9BmrAQtXx1lDfR/Vl2nu8mnyvHhvgcb3vTv1r59/r2f3vLtMyoeyEAGtszqcPrK7933tmRg1z6rEtYWLe/lXfV+RJ+4tcsCg2zZf/VzEsRbpcZ+aRN6bJaochuZ7bew9pccUgZTQwEhd7Y1Xibp+WeGmZ72KrFhCFLPoIGZcViJJKT+YazOtrEYjdu9c5CslvwdTFwccvKq+0P7oCCcPSUtfsvX3/bXCMq1kwq5dG1fYXc6L6xHLe9Snec31LqImh/8R6ZVkikntVkOcPdwO07Fu6ZVZzclhofAuXDklyCYhawLrdsNxh0PMBzLJyMGfgIqIkYEM9NstXUFaY1YBlaqslu5baPQCAZcoh8WhQEEi2oZZpzTo+izOjij2mteRV2AhRVAxLfksc9TJUTzkQl+1bH0tcyUnUZprDOAwIA6XxVcOA3ThaXTKSN15kRrwW8BiHvA2JVelGsk4WhiGtR0Yk31sur2WsA8CEaREYmKVckYxx6ftKNixgYGFYMuPIGAJ4GNKqzVDbFrAbx1RuJG/EWXS0vhPEwm51HkLX9HD7rX85H5ZnfSzCmTCo9H7M3+6p4fEHrZpVOKbVRD2Rea1d36waXGBGW8vXANQtAtdy529PKp5hFfGSoDMBypx6zBVLJgc+pBRNG73ZFtLvPPfgxk+iVsM74CrP696N61Iwq5lWcZxS22cA3/KeLU3K7KpkM7ntKIvQqa7CkGxTw5gE/7BLC2hqQkwxzaY+zDaWC2LrMWan7yDAbFk280XQOqsUbupFovfFIHthXCEAiIWXoO+0+C2Iuj/Eckb55MbPCJTUpL5bo/XmnTGs22Ls0E0BBf/e2a533lj3HRnxWFA6kJ0Xwd00ibXWSivRTkTBpg58u+9k455Hd9xJxDoR1kLOB1jAXQCtc4cjVcxHhwPzagBoDVhPZFoHN7XF8lC+3knQpu0JA0IYEUhnU/ArFR/a+Ek+Gt2HlJ6wLK58x5kp27jAFeOKjc3KzzHox3PFimHpn68fEsaSqkWvCpDHCApBBg0D4JjmDwxDAikEEEdzsKAhIDAAWdNqd7HDOAQ88eSTGMYRF7udePal9IebCdM841rtP0MAB5ksF3lpkjQDfCiYlzFZIpvNA1HeTFTxlBgTAks8FsAi6ArROtYMAMYxiaFhHAEQxiHB0Sz7aU7MfJp3AHRiX0pgKKBLDLeSsXpwSscnBFFBwpwZjhGfwvFlv09/5sN5BfZtBtMSNj04dB+NTXWxFFQrIsDF56mHOwlaJ6kJbwGwVhnXVobleoub0rdsjgMuHZm/9Cnuz7TPBqUbXyZi8ry2aeXLyrjSvHjLjb8uq4/dvWPpQaoOpGiiCmYV3d737hfz8uq9uowClGZ7cflYOfQ+Kg8Si9JxVmQzVwzq1DKONqM6s0xWC5hNiSnbfPwryS75ZHE8K9PWmMYTIZt67DmrvUzvRMLAiNwSJaLijEG9B+VcjBhCwByi2LNCZdfy1dkfO2f1JEVhBojES5Pz9eLdeOZgpzJgeY65/HX51lef732H5Rmuzvfa9jFhu9cflXHIH+xLQDtm6bd6KTrqeVK4k6B1ejidbp3bzf1W3dX35JrDUp0cWVc1YEGN7cEEGZUS6bhQSWI2RkUOiMot2U2cbYvduCyb/d0tRxITE4hcMS1bqTgXx5oFeX7lwTH1SAMFV3Q2VqasIjBAQ3IVF22d2QIvHlxiHAdcPngC4zhgt9uZDSkyY2IGBwKNOsAqqwUZsKVKmIBijTJx2DC25cmjIS9LPTMwcBqnNQyJFUmcIEAZhrRG2mBMK9mwlHntOI3xmqUep13yv5wFwSNmMKLMQZhY4zxNmOcJ8zQlFumUU1rR3hx3SsjvbZ92odfJOkZrcHpYHbd1YFG2l74ecXZauAetbjgPYJ2NYVlPaz/DWmdcErg6rkLdg1z/MPdlus7TtPduLAudZ2Hq3rsUuPqtALUYsWJWKtW42bKnYHS/0y1cpiHnlvmqz0pq2ewBLncFK60nuHci9IEEBPyEw8m2JWOuKE2hpDNfeDWfSnS235260nKZrTGX0wZ5swJgbrGlOjT/ZoKo7KSDEgKIAkKImYGFgBAY7XgtZWa6GCbluuY8NMGYqhPUBheeRKw2qepL0GfnfSzLasfDZnG+Bq7ThXo/hXJclou9Abn3iZRDy3uubvc9aD3s8CgI05HhdL65lKrfp7CkGuxIz4WScSeysyUBaeyQW6GYZW5AIUhgzuN7GH56JJ2jcM5zFQrD8oyq3OdxWVq6Lnw7yWcAwGxjnop6EQYa5XoYQmIyYv/b7XYYhoCLiwuM44jd7gKjqAkBuPFilJYlIV3sMTMoFkCBA4PmGRq1X7pIIShupRMhyDQfwrRUPRQIJGPJADjGNQBgjGMa9DyqzSuOYAA7Gfs1cVLVTZHNYRFiY5ymG9zc3CBcX2N3OSHOM8KY3gfJ/wY+OswrnzvlC8g5ljyjw7Ru52M7KSwDljyXUOw17DsnywLuQWtvONTWtDneiY3zYK/F1UYl0Vzxim+oL23XC4rcmysERaEuK9kVVees/EVnV0qo7Mml572ZPDx6SCvuqXuf2uvlInbLtrgcj5VtWt1k4SvB99SNoeh9vbp2TDSRq6zm8aw+O7I4NasfoO3YkWfTjJx2Vv/lQmt8kuclUieH9gGp0L+5tCjPpKHOMfZcYp/ScVw663t+JmVW4kFIqkJOM2OwgGSxzMw8m5oWka083tux4DvN45Cd09dTvDfs+z70mTPT6kClay/KcPc0miNCXf4y8MLvUn70RVtZfm1V1iY6GXL1fZwS7kFrU3j03aDc6OtyHFeutbtcH/CosJdZ+HwciyiErRfUmmCdge25QhVhWF5dVKiMovPsg3NhT/dEToo/ndGiZlzepuW9Br2asO5bpsfZb5nMj5L+KFTsk1SgyLVA4ChzDFLAsBsxDmn+wHFIE94G8fgz5ojMPykMCCCZsilgGOW+kNSL8HUfGSwrLHt22D6RXkgDigFK64FFWS5FKsOYI1AwLkaeGSN5EQLDMCIyMI6p3DcxIoIxzjNmDpilTHGaMF1f4+bqCkQBF9dXmHa7tPaXDpRWL0ac3sa3J7DEqNYSOPzbzkz39FCnU6tUSTtc5tSiXToU3wJY598EuNsrOyzcSdBamsni+NBPZ7tNyet/4Lq5h5WvBq7MRLS3rcdr6ZT7+jzXJ44MSd2VmUqdm2cJeQuloE4RykRNANSsioso7nKzNR+XT96drzJuCVpl+0INWOTqm8r31pS5+mX32+bqQ5mDgqGykUHtQRXou+c14BDQIJ1v0LMzWUDSl0qnbdJnLsfgCFHzDygqRtZn13JT1QIrxlWP12oZV96buztD3N0Ty5rnOS3G6RgXqX1L2WrzLvrspGVc7rjAnf4XVczl5yqIlIm5+vRlqdNp2wY3p5dFSt3Cezn1BUfDvFjyluf3NtkSrEovTlZvH6KTHGHCepQy/MIv/AK++Zu/Ga95zWtARPi5n/u54joz40d+5Efw6le/Gk888QTe+MY34jd+4zeKOL/7u7+L7/iO78BnfuZn4mUvexm+53u+B5/61KeOf4oXSKj61tW+jtRpXZ0XSSYKzsu0bj2031P3hAqkQRwI8mrFgwBY3USrj08FkftIIN5+an/iZuMu0+o5VmSmVrpSg3U8UMnUFLIY/lVne5QCDOns9R0p0wy2lfiqvktp5+PkwJBmxx93O3Nzp5AEZUSax3BWlggABIy7EbuLC1w+eIAHDx7g8vISF5eX2EkaqoLTOpjnmMam6SY2Pp3rzz2A20JmOV716Jij77AkBxJpA8MA0llS3ErWo7aTkBaHJGkH0zTh+uoKV1dXuHr+eVw//zyurq4wXV9jurlBnGcb2yVV8AIIOmQhf+MHl2vlBiLu7g/OJje55LEZkqo4pSej++KMOE+Yp2tM0zWur5/H9fUVrq+ucXNzbW3mlHAwaH3605/Gl3/5l+Nd73pX9/o/+Af/AD/+4z+On/zJn8Qv//Iv46mnnsI3fuM34vnnn7c43/Ed34H/8l/+C/7dv/t3+Pmf/3n8wi/8At7ylrcc/xQuFAheSI7lUBv/l45LVU1H2DS/tBFSBT7lefuj5ozFW04PVbrLm90n5S/cmWnDfZ0669aLtm4JqjmozESd+0O1L9NpAmeA0N/KKCwbXmBaBqp9ptWwuTrrznHd3ArCISfqemraUod91u0m55gBwMCNsqpVMrT6L8qtnYVBOgtjBgq1i3l1prE058wCW78rq4GMscgzlEBV1lXvmeHvQwZ2Y146jovypvzEbFoCqrMAqwdX5jS9FIkqS6uVqJpmqdPumlflv9tlitOJr2nV2ob87PlYy5bLCmLkqb2QnkfOdTd4wNLSuPTg66L5UmDjD7m65jtrUQBLtsR0J0xT2uZ5ko7O6batg9WDb3rTm/CmN72pe42Z8c53vhM//MM/jL/8l/8yAOCf/bN/hle+8pX4uZ/7OXzrt34r/ut//a9473vfi1/5lV/BV33VVwEA/vE//sf4pm/6JvzDf/gP8ZrXvOawArV1vDGcxlzWYi3boOrrW3Pan94x4dCUiM6w4qkClkvT2Ja6aOukrViupyR8kBmW/FbDe6GeqFiSsipd70qhz+6poKwLet1zbd0Uqi59JhG0sepxFmptQmZkar+S9NU+oPnpdQOfcShUhEmNJYZySqrAgQEaARBhvLxMM2c8eJCY2sUuMTVZgVhuz6pFTt59TJzAj9J7YvKLDrolTGQxSJ61q97WX6++bOLcQIhMGGQpk0GZeRgwDBEhJhBKYBVxM1yDQbi5eh7XFzsMg7BGAEFZzeDy7L65fRfa6yd+FbcbTJWnexhGKesuenIGVgBzdOdEGyF2YWZGnCdbNTvGiGlKx2mfjgGAaEQYRjx4QEUex4SDmda+8Ju/+Zv4+Mc/jje+8Y127qUvfSle97rX4YMf/CAA4IMf/CBe9rKXGWABwBvf+EaEEPDLv/zL3XSvrq7w3HPPFdveoB0B/el7+QD6QJB7N+VxT2gusZI61cyV/OYZ1Prfcnnb3vda+Q6Dqd4dDbM6IfR71q73WT19DrXvmgIUcm+fK4F4sHCp0LV/tTlX5En1e2t71/BxHbss6rmXF/tb97DeHMvlkToFg61QLItujiPCOJjjgmd6xrRcRyDWQwi0XOxmVPdsmfrAb52Oqj2R+0FynZAWmSyYltaR9vg905om6f1nR5zUfrSEciyMBsZolH20XZZctY7J6L9Nn0QlL3LzlypTlZvW0EqXqcuQqvhU7knuoep84RXLDqBiRFT1cMOkbjDdJBXszfU1bq6vcH11lVW1V1e4ubnGJOrBNAh/Sz31w1kdMT7+8Y8DAF75ylcW51/5ylfatY9//ON4xSteURZiHPHyl7/c4tThHe94B37sx37sTKVcYy7rHGprusnmXJ7fIudLD6BDmdaW68e0GIYtF7y6NMOG7L1aBG42DKqBa6VMHXVF3qP/qFaVFTh64FCBCzQOAvCbWePdw7HentV1RBA7XcqXWLwOiZKdxYqk6tG8aTmizO03x2j2NK0FaD4hJMcKZVqJA4EjiydhAA3pnrAbQWHAeHmBMI64EMY1XOzEo3AEgozd0lpWYEAS+EQEDonBJNVzkMHFuY2wMi2DiPTnPTM1bf+KVF0WkGbpSPu8PMkwEIYYMMyEORIQZ8SZMU03YADT9RVurnYYd7vCJsYsDihYct337/SIcOwn1i3HlvNrmXF3nz9FFpnjNBE2j6aORZyNWTFHTNMNYoy4uU5AdHN9gxhnXF9fY54jrm9u5PwEIsLF5QPsLi4x7C4BOm3k1mPhPfjss8/imWeesePnnnsOTz/9dD8yp49DO0IAOkixHwisb6vCp8mEere586UwzIBVCeKF90b1tabHXqWzVI4msLteCggA/TFCvgzsY+vpJSri88p5+yTMhlcQAutWZ9YrH5R/irWFx5efw+Xj9zWnbeItpde50D1V2iyMWWgPhfuznRd2LXMQSWlyLVw9UFrXXXJXAAjpmIkwDKO4uo8Iyra8E0xwZZWMvd+LujoTGJGiqD+5nbNSil52MbLTi4EhFj4JYVK+r2Behf55GYBzwIk9p5E8klzqpWxNmoz2zfY1s/zdKAiUcmMNTOqrJGe9N17x3XR+lF9Fm2h9vRx/6M6pU5IN41CQmsGcGFZyxrkBx4hpSuB0c3ONeZ5xc53219fXmOOM62sFtZvULqQ9xjl1uE4JZwWtV73qVQCAT3ziE3j1q19t5z/xiU/gK77iKyzO7/zO7xT3TdOE3/3d37X763B5eYnLy8vN5djfR6p61/vuqU6WzKkfqbVl1QB2aNh6/57rnY4jVcDVb0Yn9DhNILtkuLoMOZ866LbydFbTLiR9UDnKjSTzJAQln6CSyklFx7Rs6S3DAXfdaDE3HQ1vy1LGpYXQ/OvCLgIWKHspxigsS+rLu6nLoo0QhsZIP8GEMIxynNjccHkBGgbsLh+AhgHj5WXy3BsTmLEuDmkoqSw2HSbhw9KpCCDE5CSodRlUwa31CisTQ8eN5fFvXKjwrBrduyrVgoM6koSAgaK1mSRcJ9xc32AYrxLTGgaM4w7juEMIMzCIZ2QNrlhssmcKW76pCt49iHFdpuUSFs43HAXTK5tUVAeWWWxTs9iidH9je+aIWVR8NzfXiApS8yzgFY1x3Uw3Kb1pRhgCZlEnT9MNdvPFkn/TpnBWm9YXfMEX4FWvehXe//7327nnnnsOv/zLv4zXv/71AIDXv/71+OQnP4kPfehDFuff//t/jxgjXve6152nINaJSB9M7nPWgEXuSHrDrkHlY4lX9MqpSI/0OsGlUuXHSc2EQza5ryhP9XwlQLabkZfmeddAm/qRyEdZYAdwQnkhk4wPrkyVILc4e4Mv5P5YllfDsBpU75a5TdRFpLUCa920cfbbpQQ83VayZsm+ATu7kJ47KIimwcTetXwQ93Ia1J5VrVbswcflnnBsAXjYNdOqruxZuGVcPn29rQYvD2LW0dGsvDebY1vZPT+zrTSDRy5as194k/0gT2X2MHS26pty92pdQuvB2eDq4Ra1Y1Fv0/vAOjOIDAUxkEp2KW+bSjap6+Smfq02qedxdfU8nn/+D9P2h2n7Q9mef/55PP/887h6/nmJe4VrSePm5ho3NzfiQTg3nZJjwsFM61Of+hQ++tGP2vFv/uZv4ld/9Vfx8pe/HJ/3eZ+HH/iBH8Df+3t/D3/0j/5RfMEXfAH+zt/5O3jNa16Db/mWbwEAfPEXfzH+4l/8i/je7/1e/ORP/iRubm7wtre9Dd/6rd96uOegD4u6hTLskb8bQyVsCnhq4x3PsCRsJjv9SMXZE4hTN+11NMFmj8OMrPuLSMisiKmULiLIvCu/HpOztSRhl+wjKvBg+xYg6i0LnVQYWzjR61eJbGySOTUE/2TOu87VldnAbAqmbNOCaLVyOcjSN49LN5tIUWkEBJJZLoYRFALGiwuEYcRwcZGAa7dL9rBx5wYnQ9bLSja0wAyOlG1bzJiFr0w8pamYoqgKd0nNmDIPrk5S2b39JMaYhKqOrROvNe1mJbBN3oO6QKR6EY4hYA6cpnfCLCs+J881ur7CeLVDCAG7i0vsLnYIY0CIg3TZB+jUU7lp5cGv1tR6TVhvWGqnBwcHUFDwl5UCYingDNoLdG/ljQKhMqvMpCZhTDfGlKZ5xvWVU/PNEdfX14gx4vrmCnGeMd1cpftvrmV/Y96DXq2YOx+MYdjh8sETmKcLuR73VNx6OBi0/tN/+k/483/+z9ux2pq+8zu/Ez/90z+Nv/k3/yY+/elP4y1veQs++clP4mu+5mvw3ve+Fw8ePLB7fuZnfgZve9vb8HVf93UIIeDNb34zfvzHf/zohyiCV0FVl5YAq37Ri8eUzxTXyQEXt/HW0t8crHNdpWeTBtbl5Op63VCoiLcU8uMsl7u5osKfyzWiDgm93LRGbe46Ay6yQ41IpK9DZhUX4Z1kjQMsnKFzAdgzG551nsGAvmAUriXWLMslUNaeUY+CEZXsqIxvz6rg5hhXzbAULJNalAFO9p80hVNmUaa6UpsIJXVojBGBg7AZV2ZXT6kaHMNQloDMFIrb9O0T6YLKlW2rrF9lWolN6BiiUgVpcyVSahPyw0BMX1Wv6dt0VkfIX6/yy6xIBuhymkIM0MmYZVwc3D1lQhbqjiTLS9KOQAatNNhabVFX11eY5xnPP5/2V1dXAl4Kagm0bhxosYCfMW0HsrlASB1GmWQanAeUHBsOBq03vOENe3vORIS3v/3tePvb374Y5+Uvfzne8573HJr1Uo6drS1fqRJsX+4pufv9eZhMhxKdmSVtSvrAPI8t4jI41Wc473O3GIZQui8YVrnP1/XeViCp8NrLtiAaXMmTkFU7mn9wDKtlP/Xjic1LAMPAQ7wAfc8/5cnJY48oLUFibCs/Y/5OgwkPEIHGxLSG3QgakhMGDYOd1/FZJOhrbuEMBKT5GAMYPKc5RKAqwnkWxiX2yRAQBpk9QR/f2/f0nUn6Bipznk2foWq83AoSYDnbFqVlSoKAmY4litOECYSbMc0/eHN9jfn6BnG3A8uzdi0klQhRpuVf3/7+V3mxYB4OmLXeVGWX57N0e1t5QN3E8xi92tZlZa96OuxAaxJVqTKsBE4Tnn/+eUzTjD98/vkCtK6urswrMAojM4cMY9spI2svlN4/IGPspBPDYlc7EbMeD+/BU8MSYFHF+9cZketJpwSK46ZH2b0b2PbWKt5v5/wXVZe3asCr+ZX9/CasIRD5OtLBviv3FPcvC/KlJ4T2jJVhGXDlyJlhLUQj5DWf2pdX5L8U6rejaiXIfnufqGRTPdugZVIIU2fLUdWjMg1jaN7HUsHag0aphjSGF0JZWVbtaRkTQnRJktiJYAyJAiFGEkHc8fNUZuy/iL22mVwFdrvUlU6ApAOciQhk7toR8GO2dB5Cnf2dOTuZNI2dLS/vgFUDWAleXPzO/iud51L1nzpCcKkajXESJwlVt+U6tj1cR2mpx1iAFgvrjAJaE66vr9L0V9dXuJlmO766upbrWV2Y1IPXwtgmGAOUJuJVuATxyBG2XVHzk8LdAS3X0d5eLwu8f3Oolxw4+PZNwFDsb5NyNWFPXrRcitNKeLgKUQHJL3dRb5lpuR6+opa0gYZJ7dlS5lmwJz4S7ePN47MUKLLgM6lHUnoRinkapoqdUYCq3/yS9rbo4zimWR9UxSf3RlNBunSESaXFIdM4Le+AUcwRqOASpIxMGEQ1GCliiBER0XmgJSDAxAhEmIOs9zUEq6PkSj8gDBAmNmC+SflEtbkI04KpDHVuQ1PgQTmS2rc881KWyTJt0M31NRCCORmMFxcYdjOGMKTVlWuB0W2CebaPWsa0Mie3IK2X5PSgYFStfm2DdWfM04TIs7mYm41IWGfjqKHguEfoZdthZrM3N5OA0VVWD04znr9Ke1UXmk3rOoFVFHUgc3SdJmCQCZaDDJUYbGxiwDDo2Dq4DtXx4e6AlgvbVX9lvEVhWzGqfHd7fzpfptT2NUtmsBSygMpqp7KkLKnrkhFVtDbF/Rk2oeU5i1HcYlyJcfXySmWm6kwd2HcGFjrCRU0Ik2oSdtWcasrP1LDvoTbUFPn3K+9gb7sr6FQl9cpojY1K76m6+Z5l9VWQVcvzaepvv+ZWda2oRMmPfb4UkppQ60wEKaJ0sGNMg4KJVEOZ68sxSRV8ebaNJMy1HXmvOf9m/CskV0V2n4HGbFM8qdOH93Isn7J9975Ka6YlWSGtMQZ7r55VRc5z8k3ThOhsS/n8bItYqlffImipWrE6XnoCbzcsHTFiwaT8eKvp5hqz2L7SRMlSHvEAhKgDQ/BttO4rUv8Yp4U7CVrHhlMrswkHUw53w+rLLVWFKpQPy2s5+7Uc9+XUu24l9V/9PpTdWndU/kgAQoVgrPfa6y8YFuVtK5Dpx2pegCoALZ+Ulqbp6y91jrlJz4NHHndFWQhr1ZAuReJmRHczaKQViEWgiVFJ7Uwk62VlZuZnYvdM1M3/aL0EBgVOixUHmQGegvRZsg2KozAdTjPMJ+AiAz4glZ+HAWEI4JgXq4yREefMrqLOzO9m2/dvXQEryEYiyOM8p7yHEUwBN9dXuLm+xu7mBvM0gXYjAo9ORegSLT+voq2RvUv3/SXESn8CjPM8Y46TgMANpnlKoCAgpaxHwWyedXxTYlw6fZJ6/hVDCQpVo7NvMXI3wsqe46Wpt7RsSU2oM1okb8EbG4c1q3cgR5kJIwFwrgoy0FIm5Y+9ENMv7Rwy9k6D1rmcLZaEa88hxXMJncWhJAv1l9Bk5rqQdfnLLyh7QMkxlYxrKeT+tz7XUn71feXeJ9F+KGVcj0HsgcszCVSRGag9HkvmoCc8YIlQoVb2FPsz91CofoaFUANWdt5waRmTySCW25prTZ7xeNd496aK51xgWpnR6fvI7yV7fpbpAiXAAmqjklpmgCODkFYWjlJOeWgrj81oIeol7Xj58UXlAp4dxlUIRlgeXjUXRUgrs1kCwNxIpM79e3GZpWiqESjZlS4YmsFows10g2m6wbXMIKFMJ8YZs4BUDVpQkGhsYmjPVcDlO7P6T8umy8z4wcNq6yrrR2xtWu9ZT1GxJ+e56Tp95efAbjst3GnQehQhC2e2xr8Uoxs2C9M6na2M65BGU+bRYMkmJvT/t/fuwdpddX34Z639nHMSICQkGJKoQECLWiEFrBmmUwvCQNKON2hViiMKxUtBLfTCxBkr0M4vGdPiVOtg//DCjFasM4pVRx0UEC0xCphxrDZDmAhaE5iCub7veZ5n7/X9/bHW97bW2s/znPO+ubyHs953n/3svddee12/n+/nu26GUc15aYJqv7n1My5+hknNMC2Y5yKtxH83Uj5N8pP9+3Tq3C2rhedvUemf4hUuEOyACxXiPCqwzg/+Shyyn2Hwuz2LaOFCMiP2eD6X7hDN7Iz3LguAzK1iMPJKWwyhrJYRQRFALOKbKM/hKsO0kVIekJFy/wdN0ZiHMvuiIe8DFgCshwFxmiS1kzHp8XqHRLrUkM738aML873MJsZpQljnlerXK164NQNITPsYCggoIOkgGjfHsLPeJgtpSsp+MlCtMa4zQPFisavVIdbrtYzKW6/XkDlqpKBqzYIZaJIBKQhAIXmg4qWRNL46ypDZF7+v7DWPJtSh8HmACEDClhaLAUSEskKX6cvKx1DqlvZhaRnoCMkku4C7qQzHdJ/XoCVMhc081fWxwy2NyHTz5PtFOM6vmxeqU6uFe799JkdV+C1skuN7AkA1o6wTILplecGaIppvmFclP8kRyX7SPBD7EWa1Nm3ylpMwA9kNw6o9HaPIWaiJcDPmxmB+5HplzHVQrZxKwlx/VM20hPlQ833vr4mg1ImGGclvuPdrxkUuhzVcEKtHGl4sZsIUJJHIGwPm11LJIp5jHSQNsbBFU3bFxGj7oDgW9eRV1g4stDOoJV6dPE7atyWj9IwQlXoR5PuSH1ZzCjAXxNkAqr45pbzKRF4RgsFrjeXy0IMWjxqkSUyKahYscWPA4u+ZeAuQVcx981lHE+pkYPM96PD1yEqQoFZmywxOcSiLD0euDZyFDKgmfqTm7XNxJwK0tinGtWNBs91jP9BdtH/HuDqM6FhSsvkCuzr8TWFT39dOlIZTdPQxkyJy+ccODKwJw8rt+lVnIzI1QsxgLdNqBH7RnN1ghApM/HW5lwqApbZMGMtyww0CYsy0crAKejyiTvqqyn5Wwp6MIiNzwXguVxEwSb5e0pO3mS0jBaNMKJZVNCRdujaiy1PJT42FCLTC2Ah5tXcxmFnBy+yAQboEp/O3dPfqVOLF6U4p5RF1SSetTpRX5+ARdrJqBOeNCMskrCWkiKkMdpiM+Wte7eq4bqXnPqwR47QuyxetZEmjw8MzWK5WODw8i/V6jVUBrTwgo7zL4EV56SlvAlU25RgXDBDY50D1jipK9hkzOFkpg0drIoG3ZwkAhiHPEYwUYUOxZZcVkTltUBmhG55/Du6CB63jin4LXLMMa2Ybjt3BkbV7kuAKQSnfq0OtSnOnD+l7RP5alMY6uAostN/Da5n1/R70apAtC2hjGjRcOiL0zVV0pwVriA4+WFk2SnNT1LONzgTSu1czLfNhC1ihKgxhWcWjn+MVRJDHGJGI2vhWfmsPZH1aRuaWhvJmVASUQSU2PJOjhmHV17IUUtmwMSAihQyfouUzW2L5Jxt9anrs9jTMKNiMxi4V0LJs1ff5FaFPAOD7w2TRYSqj8fgdSVbVNuTacX4tQ2ZXxcSW1/FbC8NalWH2vAYfT+qVlSR44VoBraTxr74FI/iTABZV8fTxtgzVP7eglX1YHqRtJci1daW6CHBZT7Ze6/n8ABZwAkBLnWrPvQESx3Lnsm8UB4EqCLlRQ8lOVGeHL8wxsN7z8/C5WX/UF/Q7feAoH6qca0DZ9OTWHqwYFJunEJJojVaIe9IWQGU3XU0ir1bhlQ4j2sEjGu0zSSWRDoYoyynlFcmVaYUimGtFgEGI08FCOSsEcH1Z3Edmz2Go1yr0wscLodCci6fyfpTV5EPJnBhQzGVZJOo2F2WUZcp5TmWk4BAjUPq4iKjsMlzY1qSj6CaCjEqcymg4CDuxrgIyC2jWl1EkWKkkFNCUVLbtjE2BDFLLw0OcOXsGy+Uhzp45m89nz2C9XuNweSgDMpg98mg8N8jE5jLls2VNbiAGzNnWPa1crUEih1QSWX/XglbZIFLyxrDtIr9Y0Zh1rMNxt6ypYufiThBoqZvNyB5O2Ou5d47i33nsXKokY8oBNmnVHd67ux4QBa2924Ij773BUaUMlQev4TllIeSOeZcWm1zkAdL9JGimUzdLZoDXMa4siA0PkCy3eCSA1ATSD98OQFEFyRVqk4o5+NV+GRgADTJAQpaBKt/SCHNklJ2pIsSHzXdOtFn9QiY+O1TuABafc5hUpSrU6Taad2KAjwG6jYkxFcaojKgoAMoGi1JBjEdlcizlbVkattUI3nnlrGECrkz47QrYNNfhAUSHt4/TiHFcCctSdrUq850mmSc1lYVj67UQOcel39kxKTPgohp40cATg425b2unNSVWuZLj4CpuKMxrpkYzOM3GoXLnCFwnBrTOC4hvkzLnw81+43x/9IjhVd6pvm8BqMOijspu68Yj+iLp/T52V43M4wSAYNbQDUc66smQTpDXfvmzVsjX4B40LjKaENTklSz6yqP6BsOGQhBGIuhKOWydY6VMknOGkJkOogm7hAk+l1GDwkBrdbiyNASTJs0DTX9gECoraMQUgBhBU45LYsYpK46P+W070TeYvbKiGRWJwq54uHYBrYTgBlc4duHxt3NdQL/TTppqZRybK8cyjH25PMTZw7M4e+YMHnr4IRweZoa1Wq3Kmn4ZuNiEyBOeFbQA1xPJ8WRWSMq2GPxzN57tx+rG1D23xZnB2TA3XprLti9f/DOgN//tKrQtqsTu7sSAlnNVzkil5BNtuO6BSo+BdB5vjwg3EvM4QKciSXyq721zcxGow9v6mgeO45n3NjgBJK89spJsry0GmJhplnQzP6im7+gU3O9QXR/VOcCqI2si3XIRjSwZYLMsq3eI6c4CVwnbpdVkiaQRNj+qPi3JI7Nqfi3hbf6WhAdJjfZlcaXhGq5PyWRLUK/EHfQEiHlP4+tZp+njStrHlaAmUa075OIic8iEsXJaPGDVVaFR+cr7CBwHw7DGEeN6jfW4ln4rPjLDynO1klsQd5J0cLy7FiIeZCJtproGVW2get2gSw90en1ioRIYQVofGv1V223z5SoO5wuyThRolcopJgy+Gypfm6+dq2uuuZ7TQI7q6vlO5xkmHhHuZs9HD8Q3EpZVCTzCSL2FYLN8GzLX6jSMYG4POyE2dNFtg/Zt2VTz1OSMAxJqBEU94ILZRT2EvScTeB4Nr0Dg+upKHNkMCGduVPNgky91nmrkJU8yawzIHVjVNhQS34oN89slrXnOVTAbM055jUJOLOU5XMMQsVgsylyhvPQQAGFVMoqw/J54MqzJ/xBQ8ilWRwWIJt2qGIWu3kY0YZrWWK0OZSPEh888LEyLN0wcx3VZYWLEOPGafX6IuW7lkUeXOuCyQAxlRMySfBFV1wTzDlRBrp9X5bSL89XfNFgbotHJ4No29Sv0EdwJAq3iRBNU178myM7APTUre/ThVdeWBWwWrH1Xidnm/u6Qs/m7c8nbPVzfmO25F9a2WHe1vaKxWTNRo+3OOWuiE3Waf1eBmOsNYrr9hHyqBkl+ammz10IIgPZpW412C8MyjvSlKt4d/6HxUfJGhbQbNSgJ1MwRTstWiI7G3pR/kUxZGy8Jtgyqek+YRhl+Xm8SwlutKNhmaSlMi6CrTxggAHEKDGMzh+YBlwEax/2EXnlQdsiDKdbrtYwQ5OHuOkrQMyzAMisGVw/8rUyoFYBWIajrlQWseph5gzU7SIP6SwADVwVYnf41B1ZbWOGu7sSAlpVPu+VLT4Wqbm+7fty5mQjWFOncQus6MQHx9YYwAZVlbigvV/CiSYYaGMoHuvEK1Q8WwiyY7aZOG9iX4B5Y3CuL4YEXwWRm/qUmsDlHnGiTCyw8pf+G5yixgO4gbiiZbVe0iIU9Ee+XBWVZvD9W5NXcY94rKy+zPoD7tfKKtgxiM6Vo1oQKYMalIJ7jrIKJJ87a9fNcKXF/TcrHhLxVSF5uKYs+y47s0H4emDERYZwK05rK2nq2INgkWvIhdMJqJLlR0nzzp7IEUl5D8HB5iDNnHsaZM2fw0EMP4fDwrIwWXK2WZWLzKPmQAcpPGnZA0OS6f94DrM3XChQ6YrB+w77bB/Cus0oJzPlRcCcCtJp8dtTU3CDWIsk9Vyob6jLUArfPQ7/IZ++FVsuc01btdWAUMJpqz7XVvrpqWsOusL6hBpe4aR5qhXef61Vmg1wtcCnT0vhz30mVjCbCjDhF3NTpPqLC0VOEgglMypDx1FYfo+3ChiGmQQUtxwZkxGAbWWYG9k6XmVnTJz9j0LasRY7oNsjcmFnkUiMg2pRzEWjNagumf0Py1mjqZPqt6nRawCLyLMuOwhNTrM1j2fbFrrXYYQucfcTvQp+XtPD+XGNhWnakIK81KCtbmLS3rKZuuWYjRfe0bUM2y2s5VjuSfJdqyXe7YdYl3w2WzJNHEbCAEwJaLPi8Lb3X6LZxhzlYmX+/Bp3GcYWaA66w4QtHFLKPujORrXNrVpGQn7bC2yHE1Pjvu/pJ9WW5ZABTAc2/VVBDMI7ZVUTZm6n8BrL/xLKT9B3PMH2fKgNV9t4KYh7VNyyGstYgr8au8RIVmcGo6qNhtlVWjVOwEna1KKu65x2KMQyZaQnrqlZ3d8VGRkerhJRFdV7Vm6KshCELsE55xXJJf6LySpDRjGlK4n+cRl3glsuQ++Bi1JU/KM/TmlLCmBLGlK/zqEKAeCNC7tfjEZnRlDtIwaUkkaLu2hzBcEOykG1elinPxXr4oYdw5uwZnHn4DFbrVdnao6zgUfqrAjPxUvt5pgsFv+h1nfsWPLkGZWDJ6MKT9DlcCzo9HcICjR+AUSAzSEg7tjtGziBPSiyrN+q+43MTbCcCtKyrO4HbTKx5TRMCWnix/sl7BYx2WrlKjtbANReN+svnjF07qU7Gu/ifywsNUBTR0CQ3/+40HpXDRsMW4EJhWuSAoTHlSGNhqMl+SPwGeadZ8QM82ddP+rW+5Ai2yQV9o6eEGPOhAJepj3W51yPkZCUIZgcmz/Qd/njNslRwSCrsoAszIAOhHoQRmzzjuGv/bzmRiN4m3+Q1PpgJcZ+VBWAil1YEFGDgtft0gVX5kmGBWeAz2+I5XKTztyQzyjcMw2oGuqD6DlsRUPpWU6mjZT+paSwjBnm04GptWJZdIorrkM0diyz8sNpjjhtNea4jC6v2yPSeVSJicDT1wImrHmDBIVxP6rm4t7XYeQ6oHj8C7oSBVsu0jpd/W5jWNsK24cPzr4Yd/e3mNs5U3/75Tkx4ELP9tT2YuXSIQLYNx5gH500dW/KlQc4CTsYE5kxh5vCd87oL7lCu+b1UVFoLKl6kiNg2PkwUrcnKsKYw038j7K/EIcQg/VhWImZFOgAUEKj0a4XBMK7MssJg+rViBWI9kVX3Rc3mvXkaoObesmpFgu6RJSVTPpu3aSeZi2UXtU3Ea48D0gcXynqHMMDFgEUZBqiAYZT9wxay6ohMYO4lQ/iqydmUd+1dr1ZYLZc4PHuIw7N59ODy8BCr5SHGaZI+LEu/a2OfAk11e0O2Opzjc2gZFTDvL+gPo3wYR80P+JL3/blHclsU5aO4EwRatYCwGbsbZ/E24i1MiyCTORuNuNyguq466mC/oMKpm6QjOm6MYrffSq044d2HElNdXwKoK24wf3L2zetqtbMMy/VLSJB1HDph2N+1SdAgGW8P4g4jogS45G412syFVudQr8hm8j4oeMl3LVCFpjYIsKppi3O7zVWBBTYphli9H9rBKdDv9wSYXFJ7u06beM+US02/xkwYYp2TWgcEqAwLt/nAeUAUrL4jjIkXJ+b+PJmoLWcuVwVPSS1nL585Tkm3O+G5WeN6LKzLjxLkwHgLMQ+Q87Vlrn7zUl6WubtzxZp6gCWMzfbfdWJRz7uyRkPqvhXcqQ1w5v4x3QkCrUfPNXpGdcMVqfPoteLtOsu58q3z4XxMPcsy8avq7byrKnxWi4ugKntMAZ0MbPTHTriqANggCFbz5l19QyPEs2CnYlKjspcVhGml4i+GgBQiQtARcTlW860yC8Yct9gxVTXgJS8WAUuGaZW4Sk7mLASvKcsCOffpmNXduR+rZlqRV3c32lagMhrblnWFVr3kMlCEMicr5n5AAHm9PenrISwWOV523prbF8yaCrl/K0ZZ5QOlvMqihyDY/ixImVqmxTs2x5InCt6qulgwAyCAtV6vsTw8xGGZm3W4PMSq7Iac55AZXuKKkJVSQFfN7rWrjrONgdsK/BnN2X87A5aCke07ptQqBU4BLOFyyQfSOBv9u31RbhGYeFoF5lzc5xVoHdViBoDVlPY2TIFtUTRmgw59LV37garvzjC7rckSYaTCYadY1kq33LKp19/BvBMQusoYlW0PvDZOlZa8jRvaxq6Bybb35r41E2lafATkqgMk3QPK+JgR9Zt8rWGzQOci8eym69fGUgCt6qUtWcoCBsSy0a504fuxQsiL//Iz+VajSJO7trf52/I9H3kBMK4zeXDGBF7QKVFA9BqKvgtWZJR5qU/OM6+gqGBmEDDKAfcbdtZetAoF4Jm1xCPx4rijsKxpHPNRdvuVuh80vHzq9EmZazbVucTwD8Ej8vcEgHrvcTyC+KlFGDOvuh26aDRlzgpljvtudg8FSd9v2H1xJ/d5BVrHcsdCus2uxxP6nEpZzZFjsR3JcJyao/H0wDU73qj6zPxXi/SrzB/9fOm9bSQGf6h6PteXlbVtFWKxCHLbh2X7uCx49VPT9hcEQN6PgSfNlqOwKWv20nQwsyuMy4JXCKJxE4WyGkRAFI0ckjZ3mJ2KKQYd+cf+mQUEXg/PCzcxL5nV1W05MZ7EwmI4ScyckKasu4eAGJW9+XlcavZz/VUmX+VsiUgqipEoBJlhRjsPrl4JY0sbt0wrTyBelqMshFv2+2Kzo6ShCb4Crm1tsEqgVe5suje7YOJhBgoxYNXthOxWO1ui19fnXdwIlJeitAV3ju7zGrQqArL7e0d+UL5jvNRVd1OAwV21n2uiP/vC7Bubw+v4UZ9h/vlM22SuwiFwD5ZdP46wLW/4A7blWSlRH/Pva3+V7cNCdbaHYksvr0JJU++rwUXTDMjotX6n/aMStOotY0wRaOCzyR8LzhyuYUKOtZCmwLKCxqRjBGl+o82JemQjj8Cjsuir3SMrx7zNT8e+pV+sZQEsFLP/IFYMzeOqT4uZENdEkcAm4ALglmlNZvSgXbU9pVS2kUFRKmwRBuTNFcs9yrs4m2xsZVA/u4/oShoNc+Xh82IJ6VQ74rzYBVcdcHkwdaMUqTdP7Xju8xq0zofbglPH9vtIuF1ZyyPrNrcCouYHtkMlzJ3QvR2cqUyZip6NMAsBYlqCByy43ygyXYXCbKrNezw/aXCav98qJHTe474s7ZvLrILXbZyQmZYoATkAf8TI0twPxECAzlswTKvIW5ZxNp125CeBBMgZCrj/LcbcB8ggKksurYHA32AQo/J21EWdLHDZPi5KHsDytiUW/HI8YmG3NcMSxYl4LpqyewSTTjYNlv2weGHcsTAsjjfnm5adpMCd23ysKov1bgF6i8zvj4bcpcUzuGk4RDDzycoZOml7G/5YRTRa8+Bpn9Zubm5oqz4/4vtS6eHOuxTF+QAOHRXov9/9WH6h+CP3gr7XZ0vV62g97Z6SbXkUjJq9bZJ4904A7DJD1qMojSbdokiHLGbytZr9pL/RsjbHHLwWexTXAyIXTpGawRWUAmUoieIsY4BiQSP3g5m31jAtE6akT79d0KMRUn3gyssvtemEhC+DK0o8M8nikSOacGciNHlSr5aRnAD0c/zIvCd9kOZch82AI1ZR0nwFzMobzLQqhlVPiPfO1GVSPlkDf8tk/Y/Wv33sGzQnrbK0z7hcuAxUhVwq6bRnDm6rEGtRuOZZ5wJbJw60xG67FYm2h9PzNqPP7+zOFbDOxZ03pnVEwNocFAvnI4ZXaXtblT9bcD2mFfW6GXTBQt6Y1Fggi7BhOW/qnx0VJ2Y4y7h47yxe3d1El4d6B9KBAjFEUCzavMzTisIgM/MKmUGVo9mduF7dncHKgX6pKS5DDYCYOAoVsEJX7mcAsKARQ8RU3h2n8h4HHXNaxXQGaBwRxOwpJsIqPjbvGCjZJMiL7wrT5aA7YVBJQ0LKpryy8K0MwrCTiO2LwYYQNH+sekHKQBzr6LKtAqSSXp9u+85ubTuoZmMQKSttdQi91tS2smawUI18jvmeH3fiQAswgDMHXA3yhG6BN8BVM6xdtOzzhhQ+SHc9R7crlauNrhVS9ZV5vfbeDaMTTyr9OpSbRh1xl3+75lElV+WS28iGdxqmxedQMa9ggFTeMZXGSDwBpA5gzdePIsBjf+VxpTUs0FUb5hro884k0IKjXWePwUpywGRGMNcuh+Yd8V8yhzwk93rwWQaQrmZBpH09EVGZYgcM5D4ZxmHi0+s1aVmtUVZslkGZlX473+H5V8y0aoZlyU5+J3RojmVKVJ3rBOtr9XNiMLP+jBzLfmbqnaPNPcCaoVYuUtpfJ18N/N0WsB4Jd3JAK4QNQmLLq4+Q3+O9sIt7BJDwPIdaC4DADaaqyGpCgsjSYxajBNFrgs6DHG0/lQyMIG+6c4MwJO5qfgsAItT8lZPaNlqPK9UKGCWuiQig1AjmWHYGJruCe2EQhIgYAAoBcbFAXCwQFgPCYgAWeX0/DFHmOeWDWZfJnAClOs6pAiKJZKzijnZ+LeX4m+UpCmPUtMYQy1BxMW4iEeW0JDb1ca6WlewdK89Hfa2FrIAl6zSWNRp5bUc70o/Tl9OQigkzx4v7seyGjrxeYSa2OsI0ByWGRbQAldy1srEmuzkopwR0ccA+M02pdW3rsKZlaHG1/WitNiB5rLfMUlJUAFEmdpu6di4NHCcEtGo9kR1XnT6PqgJwl40q4by1fVw7RXMHJxxni7/zC1h1evySr0dzjZUIkGTNzerIcWA/qH7453NB2MezgNV5R/z1TIO1hm6+FEqkAunw9qafZFMxhRYU83saY6rSy0zLmy9LWKwBN4BWLxJrjvKuagtbMkzKUyVb4N8ps0KS5xawbPw1rQwMeSgJEAPlUfbWbCa5Hdp22c9Yk665PAvmfi1DSb4vowanSVZut0wL5n0zbqTkge6bldPau4YbQeni0YBGywLbNOfb0n7r+mje74GfwGfF8hq/NWCZxuf6xgqT88roucuuEwFax3KhOm/zN3f9aLnz/N1dkk3Nnfm3axm9WycwKk3feD8uatbl2kRVmVXNsNx10c5RzHyOac2YmmrTYNdsKwCOxjQoXlLuT9HJn8wYMgNEHHSYfJm7RYW5IZR+ssWQVzUv6+xllhXkoOqbs84BkBlwQSS3Q5mvxaYrvrbSTqDRAlfxkkgNe7JCBrXzsqqQZg4tdO2v1M0ke/O0MoCUeW6UV6efkg64yLsTH2K5PMR6XeZmTWVCsYuh6YMqACWbPcqZ9FyDcwe0PNjMMS5TqQTag2oLck/LiOOp+geV8uAJxB3AqqqLgFIdk1D/4LwOTUkdx51I0KoZVs0bJKMrwTaryZnCB1AmLx4vZv7Dc/er8CuNpvYlSvrcZ2eI4RxhnFlQpg2octvIRRNaT2huMpnsHjBUtbNMwgNkQMuw5BqVdi4KozIVYTmGhbXApd/iDGIzV83kBA+gE0Btv2pexqmkL0RNTrAC2h4KVHYgCJks4tGRczWSL+zC41Zdr9fEC2wG66j0Yt4z9ZWM4ExUAE/us3Bl0tZXlty1VEKrnJSy5HwRZYFzG8qq6vUFxxEr2TPLruQ+edARUy4L+3anYr3fZ1rbQct8p9c8QgaJIJSL645KwGYUYqrvG3kpaerXjsAhU/8px8GR+vPgTgRokRzkKEJPE9iuiVeufr7FP3V+bfZ9nkpyzh0x/sd2uyZF6nKlobGW12igu1K2LR91ZrByCHPJZ+kzijGb/ewZfPZglhXK0nzFPNIyLc4eMkcL2laA6hBuXgMxljUCicHI2LYoQBiVrGLOe3MVkKvnZamqVFqKFYyZAulvexQ/0tTkXmGHHaZl58hpefjP5URyn15Z7X0y87LIj0Sz/abCFiqzme3TqhfL5Zd5I8m13W5kvc5gtVrj7OEhDg8Pcebhh7E8PIv1aolpWiOZjR6ZQWn9TbBMy/qp/bZCv8Ooaj9doLApL38rhbw3tN6BzHxgOd7WW7EDdlVtd2H7Hs/dnQjQys5oOsHepabgdgEuJ4MrJjOnMdSNaVtcLX12biuQ5ndqxWyGpzUMrU5AnU4bK3veGknqfD3YbA5tcrmRwwOWgpj9+lGBqxKQfAmIFi5msor17HIg2D4tZksWsHzO9XCdWZwVGDkMMunPLsaYwzCbGBJMoAaUpB9LhvJzYei7dbwI0D4qm+XNYSSq1dxnjrY+9gUYEdwCrpSU/fh5Wb6tiX+5X1aeZIYVlF3ZQS8orCJ/Z8JUACsv1ZSP5XKFwwJaq+US63XZ5JEBS0yjRW1OGaSYYfkzZplWDSYb52PZ+yYfXGHCSog5SQHznNwVYJQqLsMAl/+Bn20TiM23zg28ThBoHc/NZd+2Yj7XD83h5hyOegthG5s+sOwcHe+yArUBuHbPDQdYW74pjb/S0jsed3/EJpIQFFiMqUwHBihw5bPOZ+p24rPfGI1JS5mWtdDMJqWOvtFgG41YTJRBdu9ltkTWT1ljT3fptesOWtDybEfyj39IOaADRPCJUvteYVrWn0mXKY8SYTDykig8vK+WWd2d9OyAq87Xqvy1nPw8LY7BlKbCsNZYjyMOl4dYLpd6FNBarfTearksJsKxsD+er5UKOy57gKUO0zJlaxlZr8ytXuDSWl/LbRIQ5udzsoLzXovCKDVSRGoi52By+alaYF4xZuxW0Ek/mSTgqMqndycStKRPK9RbafRdzTT6Yc4zrK2uQqJdAathQDVjAhf/DCLOxGMWGM2PfrXangEMcFKhyX5Poa+jZ6u2aswoloFs+24TVQdYDMQGmKBnChq7AN05TDvsK4AzzMp9pwNYs/WmUjpZHCg+kJpgmB0ZwLFKhGWMvH9UA1ANYBngqplVEyt7qwYzFIVjnnHNziV0X4H4ZYGeChhYpkWduqL/TN2DBS5mWiXvCpObUu67Yoa1XC5xeHhoQGsprMsOwpimyYBT75wHaQjTIs+oagZWM66jgpasgm/O83JP6+u2s81LBa5cwYPUc7MQL8z8RKfpbq8Du7oTA1pz+D3Xp6UZe1wk2tHNBj8HHXOv9/17BrQ1uB2+s9nPLtWux0wVECtJXs5Z+CQRVtI/UbfW5kutnyY/Ape3Edby7UqAG3ZlmVhvhGEsTCubhICpaKy8aCqnc1NjtaSSTYpO6zbFzqvEy+rzsPU+x0V35h38flm8468DMw56pvQ58Dmm1Twvc7N47lUxnTHgGEpoEubBGsHsRJxI2Es2E1K1SzHJ573LYXNf1hB1lCaAspbghNV6jdW4zua/5RIPnz2Ds2fO4HC5zLsRr9dYrVcY13nNwWmcME5jNiVOSeIl56QMS1khMy0FE3tIXx0lBf0qyyFp1Dz04JZHWirg63NfmFWFwgxQhVInuM6V3yFAdgRgJSrKtSp5dXiaRpqV00dxJwa02G3MkEru18DFj4+Xqa1KHZoffF2pLg2FasNhXd+/33fbcdgLRI1NadSVomTBqgFRbjxV0JIso6oF8yDAaHJWaxYBZwS32jqaNCheiehTLxajAlw+k2FYtUeX7mACCh7AQih9Wo5plXrFwLwDcEmKSFNgG7fgSwWoggPOYzaH6WhBs8GjZWOGPWoJdeJkn7EQbZhWzbh04EHDtGayQYuxNp8xcJGYCet+T847i4ksfGNhnX4CN+9CnBnWcrXC4TJv7njm7FlhW7ww7lRGEvLIQmZS/qwMq+3TQpdp+feo+KfKn60b+kOfl3PJAGVu3h8sA9oAXIHzTq4VoAKzeHOfeBJ+5V/6dk277pX3cVzc7sW7D33oQ/i6r/s6XHPNNQgh4L3vfa88W6/XeOtb34rnPve5eOITn4hrrrkG3/7t346/+Zu/cWE885nPbBr/LbfccvxUWMG0zesjxaxMuJu+UAPEDBFsw3iEoj3n+tVbn+5U6QhGKNYIaUIUbUFZlgMsG6AT5/56t4ZgKoqYP/RcHy3jgrAsGZnm6nJJrb2Hfr0TkOoI6hbktE+t1ko4B4IdJVfYVqiZlgFoMTnOOoL0mjllAnoGzJmkDEHZ/EbMlsTEhwZktTQrNpLKwAZmValVbOS74GhapcWu3K/hjgWsDpdLnD08izNnz+DhMw/j4TMP40w5zp49K2bCbBpcG9Mgr+6ugzHyedJ4CxMk1xcnQJXsmdmkPk+JkCbyg1GmlOeQyaFzyXi7lHGcsF6PxeTJxxqr1Yjlco3Vyg40WYpJNA84WeLwsJzlms2lKxwul1gtuZ9vLfmyXq3duowW6DWOdmX+c+NaR2ZaDz/8MK677jq87nWvwytf+Ur37MyZM/jYxz6GH/qhH8J1112Hv/3bv8UP/MAP4Ou//uvxkY98xPl9xzvegTe84Q1yfckllxwvBccQ5rPAxffnMpUbScXMHAEiSP/IXFwb4Ko8NEDRMLHN0T+2m2FYyrQ6wLwtz9gvC/MSoAhyIzfZhNQzDbZM7yiaG7Mpm8bCqIKe61dC0PQHeb1okiwMidcEJAdYTLQ0gqQgWN3WNHih7eJevm2FfAY8UyYMsGZoNxzrqgCLw9yWg4ZGO+ASs1fFjrkMa9YsJiLyQQdT9j0QT0lG5bn7tdnNsewC8jGKogEUxjMFUEhYj7qpYzMAY7XsrICh33NMiuPIeaE2vKqcC5My7FHTWKedWS2nzbcPm/8Kdi07nVeCrNuBcTHDGuwO0AEDMyoySpphWgGhYZPniFlHB60bb7wRN954Y/fZpZdeive9733u3n/9r/8VX/3VX41PfepTePrTny73L7nkElx11VVH/fysE7IVzhnIZ50zldgPb7pu3NGQpYWJR4dytYBVfncz4YgBAypUKkFemwadIGqY1kxEZxx1Do6LjRMzKza1WdZl+7NSmT8VKCGmiBTs0j6hrYfkp2lagU9A2RwR3cnrAjw2LaR+WbhYpiWA1WFn8zmkcfWklspSen0g8mfTT2P6plSQNqVYvu7/lTUktO+OKh8OuDzLQ1EibL7xoIuEPIT+8OxZnD17BmfOnMXZYhbUARe6V5YDZne2fTUVaLpyNlnqgAQeXODBqlf1iTQcBs0cjgIDEZU+txpkrWxU4PP3s/N9WzqAJRbQGoa8nuDgroEY8n5tsupIjAAixpEVgG0Aut0d2Tx4VHf//fcjhIDLLrvM3b/llltwxRVX4PnPfz5uvfVWjOM4G8ZyucQDDzzgjk2uFoTn07nsFllXjSrjf5Wgkx4E0Xr5lRmhIgLUf9AQBncdqn+oD/e9HRJbfwhVtP3HO9GvNLgSgNfkKsCyWnvP7NeTdjukpSm3RtnQmEg8S57ZQRhgE6C5b23+tu415d+UsTd1bdeI0QeuAno5fmYSrWNX4EhXUXBUT6SiyyLxQ+JnI5MiiGmwa/qUMm6TV/v1owbRD8cAGCdDkhtDYcwZZHiLkXUZZJGBSk2AeWHcer+sGriU7fTuB45FpUg5M6Z5x+X/EVwbF3RZlu0/0z3B1LzIQ/hrE996zauBrOWsR56Ibd8ZeUfncXJHmrJpk5KW15ETa9wjOhDj8PAQb33rW/HqV78aT37yk+X+93//9+MFL3gBLr/8cnz4wx/GTTfdhHvuuQfvfOc7u+HcfPPNePvb335OcTkKhu2gtJ8D2ei/+ejwp/qLu1eco/m25KU2Q2mIip9BNoGszRn6zR0r+4aMbN4O1Y/KhOauK7CNMSClkEdTJW8WAYqgJL89iZpMcmxUMNsJp9XoQc5Do2iwgqwCgP3Fst1JzP1ZNdPq5k3oUR4HYO11ASgeKchnnpdk+qGQKAstMn1T1AcZua7Sr58vwAWS6LB/F/lQeBqREdIjVqsVEoB10jUFtd9qmUcKyhwsu0yTN8dxjKjqq5I0WOYnfk3aDMimZtQgiW4Ak+VN4bA/x5xs/KoiNaDm4tv7vqkarLhFynWJQLIfGZm93kQJLXVVulCosF/o+VzdIwZa6/Ua3/zN3wwiwrve9S737C1veYv8ft7znof9/X1893d/N26++WYcHBw0Yd10003unQceeABf/MVfvOHrnDGs8cKdt7rSvrcqvZUU4Ctq7tSX/fea6EkyNgOdqbf+eahvtzE8787IxpppASi7wprKXjEtyxycVtYTrJ1v1/f7KbXsM7T3KsCyOAYBJwamnA4yabV9Wha42v6sWoC0sZ2z7mUZSq5/NX9bJ9Hml6LJZ3PUey45naCSlPaZjXPDsCwDUoZCBeBS5U/hxwppD1wufPg64sCAIxg0tBKTPFF5mhDGESkA67K2oA4myCxrMps71n1Y/lwL4lboC/ulGvB6AFEDlgeuxpE7Hdu59lYDr/lAQK7oUn95/UbkLVliDKBUwKqYzBsQ3iZMj+AeEdBiwPrkJz+J97///Y5l9dz111+PcRzxl3/5l3jOc57TPD84OOiC2fl0Xesc2orR4M6RKdKjw6noUftSdnkogvbZ6CALFar+BQ8MKpAqjbFq2Pp7BsysLytoO5zNxkX6suyZzaz825j5eA8lu5eSMi0PUMKwKjOhNd/wtQpdkjrpwoKagnieEm8PryA7VCbCbgnovZ5CQMidSdqh5I9UAUrpv6JEoNKfksp5MsPCZSQdTNFahsX5YM6+QK1/yZEqOTk+aZowjmusVgMIwGqaENcD1tOIcUo4POR+LF0IV9YSRCXIDYPxq15o+l2cGLwN06pNdWzKA6D9mWTTZs+c3770rF6bl84MqqxKcQWEkEAUULexbQM2pP0G7rNq20EIZQStWUVfBmoMA4ayM3e0az6egzvvoMWA9fGPfxwf+MAHcMUVV2x954477kCMEVdeeeX5js6RnOQnV5TGolIJYCtUCG5+Uz/gmevabVVKthe8j+lcgOdegYyxq6mQTlZSG5dNzMM1oMpk4R1tuNoeexsgC4HMnApTKsBbg47M8zIchtOU0+XVBiJd+DZ/S4GL06eg3+aRBUTFYuUpMnBEwCpCawELMp9e9ynHtOp7rC3bowKsRhj2V4qgZuABGcWiD1w2j2D8K3sxETfJY4Y1jmMur5SAMWCdmGmVoezTiHGa/MK8lfLkWZFnWszmbH4mF/8a9GpgMqzMPOhV/9ppE8o/yprPclagiuUcummTcsvDGP03oHXQ65oMXBAAk00+ZbNP3aFb2vs5ip0jg9ZDDz2Eu+66S67vvvtu3HHHHbj88stx9dVX45/+03+Kj33sY/j1X/91TNOEe++9FwBw+eWXY39/H7fddhtuv/12vOQlL8Ell1yC2267DW9+85vxbd/2bXjKU55yjCTYltbo0K2rGFKoruX3TDCOSfQeHNHNvRaqX22f0Pl255ebucEX1Vd6/tSMpCPPQFZY8NudgrGsjOxXCPVVt1iL2Q8GlPLW79WAGngmJozL+AGRlJgFcq8VZ21I0pbYjEUIIYqQ6pkVRbCTFXY5D4QdmgmgfQWpYlgVONXsigqz0kPLRoCI51QZwZ/keqrK1peEmqP4e2Z4u3AVr9TAhuOAS0FzHEeE9RopADFNso/YOE0Y04TD5QqrGdOgBWNATXw5mhUgm12mM/n3JkGq8goEma/kQQ0uHHutSW3Z0Fz5tkWfa2EIuQhzfckMLFHK/cps5iMy4ehuATwvcShnex1CZlQhRiwGvV7EAcMQMcSh9Id1lKcjuCOD1kc+8hG85CUvkWvua3rta1+Lt73tbfif//N/AgD+3t/7e+69D3zgA3jxi1+Mg4MDvOc978Hb3vY2LJdLXHvttXjzm9/s+qyO546gY28DrlJec9tmzfVRNFxiC8Pq4aW9swtwHY1ZbHJHq0Ty3a0NyLxhvPbMBLWWLho20awlqxOh8rcPXPnj/ggufwuDsgyLvYZQto3XYAgF7ziKzMSY/RCvDFB9H0aTRse0VB42+WQBq4CW7pfkwVUsg528CjZDndyvFACCmAKpErJzBxigklneyPZxwZStYR/WBGb9dcM3ZwnP5hEgwJUCEFICFeY8ppSZVhkJp4A1CdDqvLBqFfbyvRqw3L5Yxr9dmcIyK856mF+SdkGsOf/WeYDK51IrqGZiGbAyE8vAFGXTzXyO5Xu+3qmCZplTdMBlTYOecclzo/ydizsyaL34xS/eiPTbtIAXvOAF+MM//MOjfvb8Oyt5NtGdOjmPMOGZG9wx7/8cgUsVqkfHzUTWmifsCDTbYJtoWsF6rk6RSYCnf4aYB+W+YVqUEWFjwwwGTawm3WNaLCwQAmSPLWdic7KfX1S1xwoJ9mO1K/su1QcDg/4WJmQZlmVcFaNyB/Mm8h92pkC0K0joYYHCD/RQ5C3hTFNeLxCEkEYgRqSQvzimvIL8epryHloyidiAVjUHi4HQKhW1KZELkkkZMy4P9kZBIZutCtZ1noh/U4hU/SKyfjxAmWpRgEzULa2PKYGMcqVyPJT+sgw6i0UGo8UwuPMwDIixnPm6zOMaFplp2R2jz0XknLi1B53bReL3QCuYRh5M4c6F67SS3saTvetWDDOz6zE4G82jy+j6TQN1jyhgebPWNpe6Hdz5CEDVx0jVuf957pvaFDteX1AGj5j3yIBUEQXyrmVaAjQczobvaeRUIAn4QAWcxsn2k0EZhghIKjEz86sqpuXqZA1Scs9J0YrRqPC1Zqqe+WsjcBlmJMIa/bMHq+R+t9uUMIAXlgWAxgkTETBFUMiWrwRgorzB5FQY11SGxee1DfumTFWemEF5hiXPq0m8Fsy2MSeBxR2Bq+dsPZkDLMuk+JoZF5WzG/ErjKliV51jEHPhIO8NhoHN0//d3QkBLZMZtXphvHTP2EA2WFJuCmf2Q9ucB8IN0Ss3jlfSTU5IYs+Zo51Xl7VZFVbtfB43OL5520tgdTuXRtHCG6aF6txjWkGZFkxMc/URaaEgVcecBW0ywoP8slB2wIs3jXlNHsL0NjmWnkL3DHJKpPSA+Z4BMgG2ejKrnCsTmgOh6swMy5wtO+teG9ZToAUA3ArwoJQZVghIyIbYRISp+JHBF5ZdNQzK1kNTR2t/hu3UgJOxzO9ILXkHAo9/aEcReuVMr33RbXJSBaHNn5Wt4qPy70fBcp9WPdCCzYU6ihA6ijCEsnO2MqwY4nnRkU8AaIX2d8OE5s8OsKQTyzKtDZ/ctQS4VlkGZy4luBD8dfVBqt7nRjQfwRqaWJptByvv42hVTeClYqCNP/ucVCCrEFHBJF4fAZzNZj0PQJkvG4YVOC8N0wr2vqanaZpGaghrNAm3wJMqs2jWxRg4+0zLmcjkm56dm9scBEqmzwCUvuMENrScLICRnVxMqQNYCgqJahPgDNOqvtNcN0BqmRaVJaSUYaVyn4FLwE3yvQIuw7DmVmkXkCmgxPnq2VGfKVV41DKxHYFrk+sCVlepaYHL/g5BB/hYsNLRg2Y4PL8D6ydIXX7U+7Qez65LtI4AMLOMy33kqIh1ro417LmKVlfgc4+XZwKhubfJhdkMb+CzCjM3eO3gbjXd0IBgFatKKPQJWBU/F6aK+XqIOwOHsNSQ/fLw8jzEmAVgJ7ec8DAAQOY9SnnUogCQjUMOhAdd1ALfCTibZodc9qYvZZ+t1D2EYfEIuBq4mqNmVCSCvmHTpHH3aRGYFPbEgxtlFXVU5xLeFLL/CQxcJO8led/nJafflQ2DkFGsGnYLg/mA8VuyNdlvQUcP8v1kv2PKswKqpmZ1b6jCswu+WeZV+w9c7U19V6sWNN0mH50JktMjMeu02yO6EwFatUY5Oxx9Vp4Ti4hy5kKfebEJvxae/Q9173JtcR76+jGhfl5B2Rzddx/jOy0M1lVJ9DuxnR3NZTlbYIxEz5Mw9b6v+H4jvXzkhoCGQQjblMiSS7OP9hxgScssP4PegjId9hsKxXJnvg+URULZUe/USAcWaElEr6kWDFisJnMeJsoj4hzAzzgnYDpgbwBCbjZMxrO7GrhaIKoAq4wi5I0ctwJXHRZgTI8WoObOwFSY1VRSVi/C28bXKwIWdABlXPV+VQ5k6mxUCqV5bauAfMSdzgG4NgNWLV0sYLm+006YrimYj1ulsmWI7O/cwIrdiQAtcT1UkIZfAwPQE9yPtutGa/e3tz7XFPq0WnDuuQZ35+vyxpg1MSytQ0ClhKuCsOhjqvaW61oZMZERFbfcS+5VHxPX8gJ4jykd2B7LQQiICIGMeaNci7munEN+L/s14NHLr8DxDZVQ0sjafKjz0O8+VTMFFu4KfPq+VxrklgUx/q5ox+TyVoGkVio6gMGMBGgARfuMDFZKOULTBQYW+XYeXQAAghlJREFU9lfdb878nJkXKwLWX5XbVZodyEi0bLl0zvBncNqqdxUM+b7NA6t4cN5pKar8L+W4c1tkczSPJuwhXA6295xbRVevrhq3MEUOhxUqGBZ7HtzJAK0qA7mgVVMN7rwzcM3eppnntX6/GVRsh729rsOj6nprBE14PcDiiqyCXL9CLkKsnm/43IyW194zqn7dlyXrENmG7rX0WQ2Q2rPggkS6x7BUZZTRgWoLETDSARZsjY+zDCvPfcnbM3CjdVGWa4PUJvoqMJy4knte5Sj+KoGngj3/k00cVXTpt+Xw4CQPHPjUwKVn34dlGBAUaCxwZaaVFMQMwFmh7vqcXLo6wGWeW9OfNykykFXhVPkHLosNQlb0EjnX4GKysVdJa7+dsLVN9IBLXejflmcWsOYYmDIt+6FKZhg5oXdJzmK8roBc2vN5Qq0TAVq2Es+6WdX/fMfk0WJuG2hlc92TnjwZUisoscQ/ajTqoDdGrwZ2VFqebbH1AbiRD3WJUxHSTn23HzeHAFPsAFYEQp4zFQMhhYgYqewVRKAQy8IBZT+tskR9CJTfIwjjkpFYrM0CmERlsMCk8RWmZYAhzEkbKEARvEm1FfHlWxSqWzJ0rTxP1ZHvUbnOo/08UCXoeSIGlJr59MyHCliWadlc8UwpmNGAeRJ4KtXCMjGY7HLg39SbTp5uka0OsBj/a6CRtPnntmj5RguaRlGo4+TqQNjQXncHCAa2VooStLHNaaxGlZK6ilL/YsUYzw94nQjQsm4X4OoxAf9238P8e3UYKpR64TXhmBv96lHdmR0MMucvP1OGWFfGUo3mbJWdhDvtbiZfXOqZ1fS9SmX33NKKPPJtqPnKJu3VA5Wa9iKogFP2wuf8LIaIFApYgcok3ygL5ebJmBEUvbBhs4iYCg1g5ST4fLbNme8I+yD/pJ/DGUycWY4803JmnyJM9eOEUAGlB01epihBTI/CgJIDLMu08ngNA6nkD8cM5Zzj5E2OMvZDgQtwgGVHB/qc7OeZiFHD7Gy++RdnBLavpvrV+qMW5HrFV9+ei3zF5kKxUGg7C87f7Afq5JAx17exgW90NjEB3J/P7TfXNZR6n3L7MHX43OAquxMAWjYrjpstGyTvsdzxwzt+LLYAmNyr45ZFG/8+Vxc6v7pR2HJbRJ3R9BGi8Y2Oetu/bL9U2BSfK5Mg/2YGlrf7IOnTiiEiRSCkCCB583NAXuopBPDqB5omjoHVnhjNyPgrAjtNSDQgpTIQpdTvHKdKywXKO3neUTBTBurcoLq5lD2Vgqz3aBiVGQHoVqLI+/8KGDV9WuZg4JljWAwc3Hel/su1pM8CvAKPMrr2+bHdTD21zikITfnl9+3CuZz3khcm79zoQX7fRMGxO3NvJ0Xa66fbPRp/drBgLwes5UTAitNL0PojuwLsEN8t7gSAFruj5Eb26wXsJqChHf3V71QMq/KhFa4UZlUDqfkVfOUT78xiNgEXaapFVeM/c3lXsaNghm6YMNqvbsmfAPAIDKnyjEOBNV3VenPFLzP162TVcd8EWAaUiK/LmZmYNRXavi6dc1ImSiKAYpB4EQhgxhW48c4wdua8jq7qmUGCygoNiZLLUyntYEtPGVGihCimPq/UOeEnB5XN+hQpPLgkeVaDDgOZP9dsyo/SaxkX3DP23+3TsgIfmwFrswrr2UPd55Pn6JFvQdtArPJA1f3mea1PVMA0B1x1GAY2Wg81YHWbu4YQNAM6H7NxyoHVaWtBq9QfSci5A9cJAi3RoX2eiBbsLs/xK4+A6wbbEXyhOu/kaOb30V2dv5sAK3Seu3hQ2xDt0BGvofXibVq3k8L+Zz8hNeNSgNbFkMyowt7AjJphJcO0ACROC/X7tuRQcifpISJM04QYYzlrbsqETs6tkvxUVlRP04Q0DHlJoahgLzLDdmcKGCE/SMjr0PFkW3P2K11U6/M5gPHzs7oMzJRxj5k1gCXARQpUFsjQOToGBF68mPNNSbufX1QDzK7OApXr1xJhnn3JkP+U+xN5SxDxB3/eVJc3qUc7IJbWyZCVYjtyONdPW1vth0lWfKlBK1FCTNFMwmaGfs6YdTJAq87OpgBr4JoLiHPTCI/8Xuhfl1s9INwoLytfQRiUjUBd2cwrs4SvScDM87nYVcywYnIu9CrRO7OtDRnj18czFZ0S6nUkrL+Njmz8dHU+N/fKnMkgiDIty2UtaJW+LmZYtm/LpCfATrbMwZNGyudBKCyB8hJDsSzkmofURxEiOYo+ADYNppQ3QOTVykHGSFikBhGMPa480LHlXtCYJY76JsDe89SwK8+Y5oGqN4+rnn/lRyeaYfbyT4uf15aUPC7vdatUEcIiwDdVqhnXAl7NQry/GqDa+9s/6/fys+2/lSXtVJcg1oG5dTOb77HgK2cLVgEBgQISEiKFapUXbR/HdScCtKybYwIbCYoUFi9221n0du57M95a4NwW0KYbptKF6txFsHnda/fv973sUtl2yrutWMOaKGmfzjBUKXOiqQq8+oAtKKvlMEhF27elz2umhcBgBYSQKUuIBYyYadmzyw/VmgV8Cgjmz8fy+RxemhJSTJimrIUPQ44js7MYgmAOg3yamGlFx4RYvjh2ywxVqEoqYGaW0bKbN7q19iDwwLPCrMnQMyXzHvsDqnPP9GgYl2FYLdOqwt2hinHzqQGNZUEP1FRGFyUEOv5TK5aWMVWHZrlnJu6Yi/W2hlcaRq22ujQ38fU+eQJ/o4AGVpT8E1HYicwylhXjSr4st+mYu7gTB1pArVMdA7i4v2WOYdGmqrEhQuLdMCxzPc+QegDVO9f+d3SzGdOmL8zcb4M6GmiGou3lt3Irl5UxUtKFZBstoZ9OyUmW8KWFyUCLAkTCrADHsPS5NthQPde9iAzDMoyLQl3OJfeC9mkxaLEwZH8ZOzJgpZS11zhQGeQBGexhtWbdw2pCSoOaVpkyOMCCk6jUOTvAYqDqnnU9QQWsau5WvQjuDGA1jIv8UPk+02IA9SAgybS1RNAK3WbilyCCq+pKLgqwhXqvNJYdfS5RC+w5E6S82/hv/TrMNc+9NNA4sUruwij1TkBoB2fzQjWiDFK5mSTwwrpuSTbTVo7rLmjQajQCzDCtmqDMhVM/r0vxyARmDmCO6+qW1gu3mwNVGMDxK80u8Q8brsxNW9+Db4Bs77eHCNpaeXDnmXRZBmUYVQgF2KICUm0itP1ZCloxM62gjIonFxMzLNvXBYDnmNlFRfN72cw4DDkj+FsMGGlKwr4Y6KKEoxpwmhLCNCGNE1LM5sG8Yizl9DF5IMCLcqMd88K3dkFYoQoeArIQUvYgoJWSglYi+PUkocROgAcKRO7wTEVnnrWABTlTFUt1ZP7aaqgVtOrTKsqBViGvvEr5um+QfMK3RM2zWaZl3q3dHEMhANVA0i1ug8JZlKf2Ph/cgDhOPflm6gZgNgElKfdzdRc0aLHzWVkVSw1Y1C82qbzEem4wLdzUTG74Yb74Vavjqz5wsebTqkodoCFoAlysTbhzz5uI1lXHx2I2YaHN5Q1NwD3X7jrSU2AMsoJBtT4LWKqpbQIs+CwsAt5VAGcCNKMFDfPqDXuX+zEgUATs4raRGuHD/XA5ykHzj1rQCmXyMichb+EQcmMPSc2jXBXL+7Fo+d48qCZCBh6W+lydA+cXKfAUtPXMqtoMUdHGv+9NegkTme0+LGjVOyCjfr8dNWgZGNX35X1I/ARIK/CShZ8lCZvFZ49pSR9lMH4iEIs+ldRj+Va7or91DdPyOoT6FoBr4wLALkfZ/Y59y0Sv+zyE1ker6OuuyH3gUsYVYymzZMvm3NyJAK3s+kURZs7dN6m6sMhUV+AjaTedgg1HM6BtAskeIHYjLfce/45Asl07C0CwNiqeKsDqhsOYVeWDBTAGtnIi1oJC5c/4Dfad0ocFy7Cgi2ixHNDRbFk4xBBzZ3XIA0QoorCoKNo+L3kk2r+gDn9Bh6unaQLKaMMwTS7fHKgz6PANpTot0yJqgMsyGws8uhiu3aNqwyoYFuTtv3IvVdeefdXvWoCqrxmk2JSn8a+rhK1ObhQhK0DGX4wh75OVp+s1Sp+MVKRtELnBzVTxrPAG9bOrMKHdvQeToJ6FygNX/UzBy++0fe6odQJAi6WL/U1y1TvXb7vrXeT+sVyPEQUb3Y1uWxRCtcqF1jebgPkPhU2ZdETXDcKkM3Tu55OJJyWkFAzbstq+FcJzwMXgwr+pBaACPmSYV3tADjvUnZAnHIN09GCKOd6B1B+gWnvvOiICkRv5UGIbROAR5VGECBFkcsgxSc5aHogxjkgxSh8XSZxyNpBhSuVCWEo9zF3BqwIuw3BcX1Rl0u31ifVAzDMofz+h9b8ZBGvAamFKgbetqHklE2hZVS5vQwOklBlySskzLgNaWV9yvGiTjuXddp3M+93WdnvtD3wv13UGIuvTGgqaIAW48gf0fWVaXPamIM7JnQDQUuchS++J21Bo4sUK1/MCVhuchM964By4zEek20dXb2bZhEfGr/E38xlyvzTMLPDtq7RBkwuCHU2MOp93ArRjpqJesjY5Bow6TUKgQnNfBnFYQON0SA4YsPUUzXxNB08o+EG5GC8nZTKErVMsuCNl4A4oy0NJZItAKIAVQsjzuthEOJV+LY47cRkUAWPylNcZrPfBQnXU9wQsZKSY79NyJjwwE6pMfOgPwugxLwaxZmJyyZIkYdag5auMAFdTjyopwmVnKxByN2ECyoCcVMy0hXhRKd9kGBkHZ/rGeixF4mTqecvVbBxr5XeuYYTOL6fiOnnAv/siwkiFDUyrNgufD3diQKsWks0KAk6S1tBmQtiZWYXm6thFYoT5kV7rAtYx3t/lHQOwClF2asAcMB4D+UuDTZV5MAtRHm68Q2RtAlkFZIYV/IAL8d9hWO4aULMP9Mz3IQzLMy1ipmUYVyzv5wEcBAxWgOYalfvFqoZvhJPAAOU+wDCOSOOIaRhA0wRKUx5MEVh0cz2nwrosWLWA1WzgWAOVUS6S9GVRxZC5UOHfq4BL1jA0gNUfPejP2appgMsAVr0sdJNzM+3OAgqzEF/DAlIkRMmjWMohxzdKwNliECMEuLRetPOirGAXwOpGUnm3NdxUXmbS1r/PKbN/G8DaIEN9OScAASnFhgWfqzsxoGWd7y0qKiurruXWRmHKFaB3BhqtvQWwbYXTBLhFtlsRab+nGlvtdgavYLODayeLws3xaW41nCwYYKtC62qGVntDabBJ+raIqIyEK+GzAO+p0fUnO8CkwMX30QcqZkelw8sOTbfnGrQQ7P3iL8DN4woISDCTWQVTsjAGqHRqUxmYUQQlp9cK6CmvHy8mQh51mZJZjkrD5/f1KCMN7XqPHeBSUx6bEZlpJQdWKalfD7QVg5oByI2DMUDtaEJzCEuZqR4ck9ZZ1hE0r0ydknpdoQUDE595FGiMsfRL5q8qIM6rumpS86Dv4zk/p6ybYPB8PQN2KG0toMsm82V93wJ+P3c5y7xyw+k5N+g6kaDFzs2PkbrVSvMuuZoDrg3vVjC0FbiI+h2cc/63IJu4owCWOx/JbY8PA1Zt/HTfb4DLVHgWfHaia1Cm5WLQAyyrFTLTqhiXB6jefb0OAWV/yO2gRRXTKj8A0qHx3E8VjeAxW4s1jCYRD7pgvLECgEBTTp6MHuQjTdmWBbOjsgEtv6hp0onEiYGIhElZJUIOqq7ds0pIEQSwkgEsZVjGtOjmffWYVoJjZjavUIFYVbe6ShS0irhHgvRctgoWFJRJpESljyuf2fE8w8y68lSJDGj9JaNsHIn/dgGomlPW+GiTyNW/6zl43z1vfblWx5/EIm37tE6ZVuN60DHnZ5e7u32zfrfoP/J349tdpOx/p+fPT2xswzxqn5xoXvKZmfi4gJmVkb00MRY1DkrrMJ9UCY8bWbUyA/HXWv7mdeuO5LGmwHKPGRQZcFLlpjpCKP1S20HLzp/i+Tw8BDSmMsiifJ+HB7PGK1p4MbG0gw3KslbCRnPKEzJqyaoYAloZaEI0tZIM2yqjBluGxSzJL+dkgWUrcIG/o4qL9FXVowvrIfHy/RnGBaPJ83e45GsgsL9r4JKKxEBiy9YyEypCn8ECcg2yfVw85LtmWgpUubq1wKUMhQ/SsnLVefs6iV09jkHXpcv7t4N7QhVG9xuzTBC+fEFNuRzHnSDQAo4MP7ZUdicyR4hLj6+f14/4SrVz0Nt54EY3q67NvmC+ueN7rJ2V0VneJEVi7W0xsGqqQe/Z+Ve7HKFzphiMGaVEoh7qzqAlE3rLmbXyCCdoowMsTYcAtBXoifL+XTCCPJFb2sgCll2D0I1iE5ZVm+WsqU5XJBHAEkCa5tlVzbRs2VimRSRhs9/enluzow1tvTAAKMkzh6lWbT2RirS9fmq1zz+YJYcYyoAZcswKiOaa45CQl+zqM65arvf4iTUvNv1inXirb7Z+bE7pUSRVn7zmC98fTR1/R3cnDLTYdbJbTDTm+SY1YkOJsczmMgihDm8zY2rfn/9W9eUqcjUY+OfSkVx/oNIkm1jvXKl6VbuvcdWuvctNqoTCAttOdE0JJEsnAZu0TBFC5nm/IRoAmqsb5UaZEpybPWvZTYoLwFVMK8c192skXkGjmBFTKLaUxAHWUsuAU6DSo28ZWFIwCAE0mX5Ac+Q0Rs0MVudl1Qu7S7FlOC14ucEXlJxiMTcRmHNR+qQaoEodgCrph5m3BWaWnAy+ZqFvGJcpe5Od/m6vCXF8Q2iASpWmXAnqPq48WpAaQKrNxf68vdH5up7D5nrgK2wLc5noB6vDHUvHr397oCXN0qCyrjnOEbVOKGhVLlTnXf1vcLNko6o/cxrLkcnKfExwZBazY6iPnivx184DCMMgI9xMJ3+cwxaYIOYebuzTUsbEJkPH0AooUWFQBKu0tIwLgKxRyEwrRtvYCbwteeoBFrixK3insrq8rHjBQ80LaPH2JGTMhHmn5TK3rOSPDqLg/itWDqgBvfx78ixrxiRo3/VMC6WvqjDCRFUYpGdm2hX4OWZlGVrKkOZ3OLbClZPdEb1NE/IDVnoDnSBetcwzeJYJ44Zp1YyLKBWT4WQAbk4e2DbhagUYsOoBIZuY1vkQE1SfOwzK1m9n6j0PkuUEgZYujVNrJABXiKMMfMBcLdoWjU2XXTfLiObfKH+D+0bP9OGG7s7V/5I3s6xlY1So/9vErFobQuIForLEkUoNq3O6Cl8zhxilsbp3XNz4oKIRW41/i5v1oA+q4OVgf3bUoE1lkjqZwYdXi+fYzyn+qYyZDmFCLi/eg8mcY/BAY+a6IWW+oisp9NRg7UdCJXBc35MNf8afPcOlozCnwqwmYVhtWJZxORBDVT9YyYFZTLfkpvyuAHSbAA1Vu6gZUlZcckF5k7WCSTAMrNJtXP3Qs+nfrNpELdt8n9YGtlbV5Z1xizhtbfCNMlAz1CqYbvs8pjtBoAW0ReErxflhNufitleX88fAOLxd98hpTQzsmubAjdd/qArP5H1ofGs4zrShIp07iYVppbKuXkhIQ9LQJC7bjQ7E3nuHjbYVLu4IHcHnG6VNu04i5tUssnCNgRUVMmc1G5oQVDumvCAuSHkAC3EFlwRKoczNykeaPAMKdgShXQy3Aixnkq0YV82o7AK5zU7GyYNWDWb8/tSEZfu1Wo3dsW9ma/CAZY+cjb2Sa38qEG1Wcv2zwqqhE46FWaVUGFUoI0eDuc5v5B2wdY3JYGiRYAbRLHD14aCN/FGJlsTFvLgNsDLQm2dWoaOe/6O5EwZa2c0J6RoQdmc2OzracCn1SrU2W/l8R6yf1DcTfBFg3Bi14vYmLIagQBCqytf86OGLffEornnF0z0fVypxVIEuFd1NWC3bHsA0aQeAR2kVlfo7G+/WtWBlPh8gK2DMfbPuK2kCt5dGWIv5Mj/Qo/RbCcjw5OJpKn2BCZSY4VXvEveNmT6sDWffB+WXaxJQ6iyQCyDHDZCpDHMbPs4xKXvWvEHxU84mC2vQMiK/KpF8i4vN6TKd/lPuq9SwdD8q7uOiarQgoFMe/ChCXTkjBmAqygk5S0Kfg/ctJMUPn2y/WhGEtUKreVDyKfgRutvXW+2DJMnf88GzTiho9RwD1vlmMrtHAK5M6wrTXANoVmPaGOzmGvXopvu4Gp5f9w1gWdz2m+SRfBE8KCK/XppYvc96n0B6BgXz2/RducP6AbWIJfYhPfcbugct1q67cYQKJRk2XeImgrkGLdufNZZziMBQJCJ/gMGKBYqwtbovq8Os6j4ow3i0r8n2T1mAmWdaU/U9NQ8b9mfO7vswG0SarNwFtHZ1s/1bAkhlgnGnb0v7sto+L2ZaMZb3CvjyqimSAlHqtilnxQ+ftgDWllA23Jh/L/9Q026vyRzHfV6AlqXV1l7M13NuVybW03S0gdT6Xv5252stUxC6NYde9Ywwbz7geVzNW3OMsJb17ee60e7HTd+xcdAUhur1uerMGqwZIm00d+KIZumt6Z9r28Ec7t6uTGtznZDPCn4aTalivJuacO9J3so8IjBoM3AJaFHOx5QAM7kYZfX3zHBC2TuMv+KZVu9I1UHCsuxvcqMJHdMqzErBN39XmNauDAt6DTmb9icV24NTA1oqUV1+BxSWRWRWEOEeWZJ65kYLBxhzLbc7LnIuc56vlU2DfF0fzLTywJ8yKpE6fVxVP9a8uc0zLQUsX91rs+Ospcr8kDyVrPeRsPIWpvTOh/u8AC3AFozcKfetEm0q47l8SwqoAq7ZMqslbE0NagHfe1qu5yj/bFzPv5v7+lwK27c9qohmPk1IIS8IqysS8Ag+KDDwihOiadaA1GFRobqP3sTjgA0SwgEWWeRywKUpm9OJuioQ8YTVpKYmA1oCDqEMeY8JtB5Bw5hHEhbQEhZJQcAqHxVYdditsqJ6jlYGIN5Hy/dR9ZgWVAkx79Rb0XRXnE9m+L2cefQgOodVG+dBi6/6LTCvWpIBKCog1eVv7vFK8GTXdiK4RUnct0sx5JU0CvNKZJaE0hjuPADDJJGMoPPVuB2Sv821gNVLj4It109WLM9V6pxI0GoyscikHiNizZe1rOIdaMRqqM7HiEcVnK10dmXBXJG8Ri7LIQW43VTzU31zbu6Srai17CXJII1blw31ElH9bAhTExXVguWTAW5TSP8FD151P4m+4HsAbNj8o10/ZIPrJJqgdWQ2V4xmSyAEMnN4jGlm7mVlDfpRFRJUBFsqNSIWOUD+MP1ZaRpB44gUygCNUIa+B/+eM8P1VqaQJbU6z0yflpsUvJFpQYap131h1nwoQo+ZFWk4SqwUEOfKUGSmYya71QcpOq5hoQp8hjjroIqIQDyROJTpBx2mFUIZ2xkBTKYuGSXIMS/v5uuVB1Q/eMODoFoqQl7bsxYcR3A1oztfKvIJAi2u2HNP+2LGKMCzDOE4gLXVyQdLxalCdwVOMIJvczSsxuTMhATwlghkQM+13w0MrnW7a2W5ebQz8assMJpYdrFqaDx6cJommfvki9u0cmi6dT11fo4eoWuYVpeRyTJOJiUSz6CARWbSqSgcBVCtf3NWdsjKgh0TqYMwUiosMGo5O3PalMEpjSPSMGAaxywoxwmEDFoIQUcP8uaPyZgDNzCt9qhG8aWEyYxaZPBp4lqxrXZ0oh8lKIzKgGeqAJHsQVBC6fxsoAi2OKHe/ATiGSdKr2dcmWiZFTHQEi7LtAgJkXLflxgLYvakVoQKuLc4a+1hhYq1xBqwrILVQZ4jOV/ems5zwa8TAVqSGfCVEwCYQlkmZR+hFiBh5iwvbbkGpEC8Lbp66GNxTm6ThsXPPXPT+DlBzsOmHIj0ohja30ZoeHrQvknVb9UkJcamvehAC+nfKFq9SCYXZoFIkyz7sf4yoOfLBclnAeqgwMURnbPt17KC06PlW4Q/SzxFAjDDAsqAjZQwjROGYWw2hcz2pgiUVd39LsUV0xLmZZd30rMdRahAwoBk/Wp0U7mo1xycW9WdkppHrdBr+rR6TrLO++0BF+uEfcZU9/uQVl7rF2XEoOH1gXc23sK0QjFT2/lgajvwbWdDQhsgk3bOOCX1sa8wb3N1PFQ57MXu/Le3Cxq0qPqtpgT/YJagzGi8TtO2j+sQ5sojlCpRCSB9yYjto9aYHVyvETscryS5Y3D1uXHmZlcDo+Yn+T9z1pTyTO9m7VSHmaTETCthGKz2Vg+o8YyrRccd3BwbMx5axmS1cqOdw2i2tVDqCphC7GCElvgjXkcXsoVMJexpSqCQMI0jxhiFaaVxQkQA4gAEAk3GvCcrYvB1b25WtSLGHNOiut/L8EVTZtuZVp958X2UMABVWHdxNi62uOdEeGM+K/+k5Cs2YvW/XIoBKaa8SzU2MS0yK2aUlTJiyNMUXPzky5KaedAwH5BmfnzAsmnkP3PjBTjkvAfekYzzG90FDVoApOJPUzYbjVNuLCJOREDIjXwqv7ViwQiUzgrdlT93rZccK6lEfRnp9JTq1K9C9Wz8JsStwti/34TXXBsNzaV303fInezZNrGap0kDZy1QNiwUulFMJQFTmBCnmAcXGKFVJTPnIzGDbJ9XMZYGXUffPt8MYiY9sIqSZVp9LVgskpsCNO/qLQ9aLMRlYVveV4uZ1jSByooZKELYLd8kYOXBwd7nQwdY2AEc1JryLNPifC3A1Vvstsu03LUnmBxenbHEmTvj3Luzec6XahoUs285gwjkmwjqVizlL7qwnZ8VqmtWWLgtGquDTVevbvvUuytvmi+AxVhbftR9XXVrlTZcaZ2h++4j504EaBEljNOI9ThitV7nRgnNuh7tLj+KP14CqlzVgGX8mdfkC9VlR+PYzbWA5L+3yZ6+HbRmABmd/JD8grvfpNfHwJ2aM1S7tKE5AR/KC1Ff5HlbVogFBAzDUOa7bGgojmKT/6CTDiwMSoOUc3BHsNedfOFYBxYIqJnWEV0tlKygDYBu1unNcXm5pxHTeo0QAsZyTtMaKQSkYUQIsayPS+DdjWnyowOFLU2eWenIwJmV3on8CheOaWkaOM7Nau+zTMvO/9LRgyjWlUrVaRiQsBJhJ1pJtQYpILgmFQoTKrWYB9gQqJD6XqtQMEopARHlfb8moTIsPgMUqWzeGZEoIZo6CqKyDwqhrvZzYoAZe/ZgRtwKYClI9gBrzrWm01kkPW/uyKD1oQ99CLfeeis++tGP4p577sGv/Mqv4Bu/8Rvl+Xd8x3fg3e9+t3vnFa94BX7rt35Lrj/3uc/h+77v+/Brv/ZriDHiVa96Ff7Lf/kveNKTnnTkBHCn73q9xmq1wvJwiZQmz6SsvRjmDCu4c3g8cZPfc5p7MBXb/jbvtwXdK8D6Hbi49P1sXzfRmTsqzx6gOqBVANvlkbnn4tT/ujvZ2zKPxYXi9AF/ZnNfQG6YKEJtSkhDXsqIQSsRj6ADeKCJJN3EhdxXj9isOC4WtMgIDDfK0woXnw89t4uyUQtZ6TFxTIvXFmRhxwvnjpimEXGMmMYJMealnXiAAIOcThTeNJnYmOlcnxZJn5ZjWmblDBVoHG1qmJZnizNMC/6eNRG25dYKUMnLDnDNuhKMVT5YcXCmw7rdlJeVheQwZP5XyJPpdUFlM4ow2OcB0+T7ees+trmaIyd92ehxFrDEo4YN21qOonhtB7tzcUcGrYcffhjXXXcdXve61+GVr3xl188NN9yAn/mZn5Hrg4MD9/w1r3kN7rnnHrzvfe/Der3Gd37nd+K7vuu78N//+38/Uly44k7TiOVyicPDQ5w5ewZpmgDUQtkzjFA/BzxA9e7PMJTW/ybh3gJle98L2DlA3AZis4xqDrS2MLB+LDrftRcGtIRplfCinO2LZQV3wYasUU/ThGk9ClgNw4AxJcTSOa8AazREB15H0AIrhtVlWjEoWJl8Jf7w3OcskArozGvI/I4I6nKDKCCWhXatIM/mqjyibhxHwDCtaRwRYsQQB1AcMishKoypnIuZXc4MVmY0oDKxGtAmYWXTNLkRhCZzxTzo2JZlUHNMi2p2Z4CtzupNjcOAvS0UcvWkLrQstHm+FhEBCaAyICNEHYChUdB2xFkQYkBM0RIt8EoYRIQhL4WBOGgc807IoRCskOdvhWxebftFZ7UjNziJUBqNnlBQuAtYuzKvR8MdGbRuvPFG3HjjjRv9HBwc4Kqrruo++4u/+Av81m/9Fv74j/8YX/VVXwUA+PEf/3H843/8j/Gf/tN/wjXXXHOk+CjTGrFer7FcLvPk04Y1ZEnIAjp2mFP5VX54wOgJcgdZDZBJQC6+dXj+flDhC3vdcTbu3WehezYx1jQ0gGXiEXzALnrtZ9trA1pBPhWqs0+T7B7LmyVOE6ZpxHq1xmKxQEoJi8UiC9VhQP4MSdxnJ311IkkmLza5AMho97xqQtC12Rw42hf6YfX6Zbzg5GfahyNCHmr1Skm1fRlMUZhWApCmCVNghjXmqQLjiGmxyDKzfCsDEgkwdUGDaqbF87YqM6AwLd/XZRmBNxHCLNHk+678flwzi+cmw75KGfIqJCKIm8LRSMwxLS+ic2EKs2nOcEzLOulLEsAo9TNkAAups1JGDAhmtGGWWdqXxnVLGRfY6meSoQqOxKV0vkkfGzcTTuscExOPDuEaxzJtlv3tStS2uEekT+uDH/wgrrzySjzlKU/B137t1+I//sf/iCuuuAIAcNttt+Gyyy4TwAKAl73sZYgx4vbbb8c3fdM3NeEtl0ssl0u5fuCBB+Q3a3XCtM6cLaCFFkgaxpXDUCHuTi1wzTGqWeYlAVXhovLvr1GFg+q95rIGyrl4zfqz122fVxvxOjp9+AzNOacoVkw3BhYKLhrFhJW193EcsVousVgscHBwkJnWeo3FoKYV/Rj/ZnWW9BlLkKK8hIKaBHOf4dXUFzLnGAIoApFYIPlyE4V9TmDy3Vp4GyCzQOWXOdJ3OPQkAj2VFdXzk3GcQASs12sg8DkgrkcE0eRRGBEJs6qZVnM2fV1TYVzTlO9nP1NhWpOAWZ3uGnzsZpL1QJC50YUyh4t4rhaXpReuXBRkMxh9wCJiod4rL5L5VrI8l1mRQPovhfnLI7BGJqu7B12bkIhX0CDQkEF4iHlKApGOOCQilK6urNDJ6EIq87i0yilrNPFHsUoIAGn8JM/42iCtHW3YJG5Xd54AC3gEQOuGG27AK1/5Slx77bX4xCc+gR/8wR/EjTfeiNtuuw3DMODee+/FlVde6SOxWODyyy/Hvffe2w3z5ptvxtvf/vbOE9Xs1uOI1WqN5WqFcRznmVHwbMEBg1OUDMsJBojEkxX4oQlvVqB3w7TPgnvmXAMWVTzlVpXWyg8/c2l2+RTk2jI/qXch2GS0aSy+7XWBQwGrCDNqSr5Dco1Q+pqLABzHNZaHS+zt7yFNCfv7+xjXa0yLRRF8AT6DWMBX0NkAMcx9DcM0U38OAXbdbRvUHNFystIJa2YZFrgsu2r7dHI4OUaR8soYiVIGrsQrTOQYTVNeVWGcRoSxDH2PEdNiQizMJIOWYU1TBxhqkEjGf7OJY91HxgNDTM2owFrmc5m8YH8MRtlfC+hU0pwMa0BVPlI25Q8DAMy58U/QuskHMyv4UYQcjowqtAKmxCeYOsl9Vcx2eN4W92XFEJFCcmsRchu1zGpr3xa1WBEYkQyjEuZVpZ1rvhtgIbTO5WyVl/xtqzVwYbbRPKo776D1rd/6rfL7uc99Lp73vOfh2c9+Nj74wQ/ipS996bHCvOmmm/CWt7xFrh944AF88Rd/cW5wKWGcJiyXK5w9XOLM2bPZlu9ARX/MsqrKv9WOnD9XIVvA6InvWb2kZj7V/eb9hinNhC/yeUuatzAtm2eurtUMcWfQ0sYfS+PsAVY+UzGBJYxjNv0eHp7FwcEBVk9YYbEYsFqtsNjbQ16Tjwxb8vGQBscjrmIAUgFmy66CPyzD4nvK6hTYayAUkakoL5qvtN+EYmazk2brvp7+mdMl4AR9zoMVkBJCjEhEWKxze1gx0xoWiMOQ44UMWkSEcRwzkxq1z8pOI+E+LL4ezSjDybArnn7S69OyfTA1IOv+WNQByHwIs0sJk102CrU81HogeT7DtHYajMFhlNGqRAG8mJY8JtIl1qy2o4nXNhkjYiI3fwtEGApSD8XsbdciTIXp5VGzkL6uvJN3yDYFXkdtJilqzsxRy1P+KiUsqGLEZsguYM0Jtio7pV6fJ7r1iA95f9aznoWnPvWpuOuuu/DSl74UV111FT7zmc84P+M44nOf+9xsP9jBwUEzmIMdM62RTUjrEWNpnICV/0aAm8x2oBX8daMwuZdC/dO+ZSLYjfY84MzdqZmRj3b7jqGJbRzNdajDj+Y62DeaALaDlmcjDFgBBbRi/laUpBEU7AgBRWgKaB3KEk7LZWbUOl/LxKCYhshJjRqIOZbVY5eC1h93Um90wqjKpTlTfb/HHtBnWdy3w+9zOarQL6P2ytMpZbE6jRNiiJimEeMYsZjGIrzyjmSeIZltQor5Lxmg6PVx2VGHdhHbeidiZgQWtIB2vpbLA5NvbvNIMkyvlLUubBxz5jQzeCH5Jz0+Eh8DYKYuBGZcATL4IRZ/LNyDXCsrkXX7bKUiQExupXLZEYe2D6u9DqUfMgPUZL4Xgk5C71ZHAzQ5zrWy7dFVFDPGrZKWQKZF8XN53+cxqjJW2nfuwPWIg9Zf//Vf47Of/SyuvvpqAMCLXvQi3HffffjoRz+KF77whQCA97///Ugp4frrrz9y+Gz/Xq3XWK7XODxcFqYF1KbABlJ2ZTRzbhd/dcHaMtv6Ae/RmivzqdXw/Vv9Bw07DPYHa4Oh/WAdfsXkRCEDYG0BwR6FZYUQMMRYhEHW4AJ4BFwWfkhlAMZ6jfV6jbOHh7j44osAIjzxCU/AerXGNE4i/d3XXMHboVpJpdAOTKu+p+YfDd4vtOyLuAYoBqmUWBDr2Q1AgArpyWwB4hhF0crtArW8nxSIQOOEGCn3ZQFYr9YICFgPCwxDyitjIGBKnmnJuTCnNTMwZlqFYY2THSk4mT4tz7R43iQzYWf+47gDpo/Kz9/KK2wwkOp5SlWaBajKDs08Haq4ZL4L5KqQryvGRchoVTGlBCCWuAaiYrpjtgIhI/xibXZzCnSpcjEBKQIBMe+hJSDPfV0a55RSZlQFuHT0oFk5Q4DSp9VWGWVSJj4g89w0ZL5m4JLsIZsgzTdztvnKigK31HOBriOD1kMPPYS77rpLru+++27ccccduPzyy3H55Zfj7W9/O171qlfhqquuwic+8Qn8u3/37/AlX/IleMUrXgEA+PIv/3LccMMNeMMb3oCf/MmfxHq9xpve9CZ867d+65FHDgI5AxLBVOTcmBSwTKUxf0XLqd3c/RlXjwJUjUKuzB/zrPNaHY/aowdgiWyfD4WZVIRuEOaPEdZzTMv4D/y3jodI6dx8A4IAFLMssLbKLYFHm5XO/DSOedL4aoXVeoXDwyUCgIsOLsJ6vVYNn79kI0B1ltpy72R3qC+M5lk98u8qGsm/nRhVOZL5jQ7jMv+SsASOSyj137Iw7d8JJW+YnWbWOmExTSKX83BsEqHI4JOmAkqObU3NDtIWSGwfVm+leDYvsSbOzaK7j5YI8AqU3bmAmYAW98dGhFC2AxFE4fLQb0t+owNcqJQtFNUn+Y0cM4BAzry+pzKgXP/rmsS4kONMRaHLBw/W8Iwrt50JCiL2bJWpun7X1zWTYkBSHCoP+ORYZAnDmApdS5kBrrbFHd8dGbQ+8pGP4CUveYlcc1/Ta1/7WrzrXe/Cn/7pn+Ld73437rvvPlxzzTV4+ctfjv/wH/6DM+/9/M//PN70pjfhpS99qUwu/rEf+7FjJYA11HFMWE8TVusJ63ESrR6ASCSVaf374hog2uwscNUdotIYRNucKbze7Qq4Ojjmvl/DlAdUqt5vzQmOYTngms8Pm8+soQloFfakDTID1hAjhhhBw5CFQSz5Q5Qnw47ZxLteLfMAjOVSzYPjiL29PRyePcR6vc6DDXp51oAut8C2n2qOaQmIC5iXdspCcIdqYgFL2FUyh7nPAtQP+TaMq649zMgY1IRtleTShFAYFEIojCsgLhYYUipzgaKEP45rTFPCuB4xlcFNDHTWLMhsiu8L0xrLs7GMMOTRgxW1IZM3DE4CTCjgSYQp8coa5EYpjmWk4shgaUGLIkLM/Xncr8fgYM2R+fvMaCHXmrem+pTbCSTAxRtmNYwrmNeJ0Nt+XPAiBCBRWQ8yujgMQ1FZKCIgAZRHG1JZ/Z1HDcrCGKWKRxt2iYPFJq06lG+WSFOy63eWii5i0k+QlsLbwrQscJF7cG7uyKD14he/uBHM1v32b//21jAuv/zyI08knnOsmU1l5NNYGhlrJy0bMb+6QifY+tV1ob4ylduJFb4mZHMXOqBF3Z/tN6qbHoDc2MYuuom2x6tGVF90oyJFcNcLL9VO51l5tqVUwoJWZlgBRAMAKvcAmlAYVsr9LusVxtUay+UhxrWC1nK5wt7eHtbrzMCYXWxy9VMrVDhlZO7XKd0aumVGDVOwz1A965/n1tuTkXOFTUlm1+HCCAkCArIZLUxZ4Mc4CdATIkJQ5iLD1i0QTTzEPU8+5j5kZ8pLfq3BZPqy8oe0n0pzTuPIfVK1WbQZpEGWYbVMK8SyqzOyoA9lZXtmJMquGLRM+zTXm5o/Wxxt3CIL9XI/K22hrElIyH1IJPUvy/t25J/TmVAYFszAKG7lBfWEyRmdrO8q9K1rMbV9WHaCscYTEk/JwCpUKoWqRU/O7/lwJ2PtwZS1r/U4Yb3OR2BJCSuQPCNp5yHt5lqToImPiZc7zzEtLtNOhdo+T8vHv0lX9SEPVLswLQWtLhE0TDXINTfaJBWVgxsCCsNKIBpEO40oE2EnZVir5RJnzzycwerwEOM4YbVeIw4DnrRcYr3KbEBGy/WSHOqLzLRqNsVMSu/ZRGn+ZUlRDmf26OWOAldy4IQCEihC2DAqNtOxIladrRlQABMdtpLUtBgolKHvAevCuOJ6REqEBQEN00oJ68K4hGlNU54jxfGp5m315nJZIEupYlicAgOyyaSB1y5sDlZKpRvAM61AKIN7KuAqQpfHMEqbhIIk3+fSrJU666RPizLT4h2j5T5KiyGAiu271yZVYQplNCCbzQGKuS5Ec86jCM1oQfJnKghmGR+b+7qYQQaQeKJ66RdLpR0nWJDsZocGR/XZy8DzRLQufNCCNFRtPOM0CctqAMsyk2OAVj36cD5aNWixGOmUHHF52mf1oJFuTBpWNRe/xq5OtbmQ37UCvHQICDuxzA2qFJhDvkW5kbG3vOpRkO+GkOcXlTaVzT3jiPW4xmq9wnKVj7GsKTlOeYDAOGWwmsrK4S5v7Jman/18nClLan6Ya/I/hd24o3ePpK+H6y3Mb8+sev082m9FJv+V1ShYQVhEXqw1JN08U5c5K0zLAkU1bN2NDjRtzTGh1MaP4yXZxnFGC1p1mpWxVdemz0wGZEifJg96SDldqdQzJGEzXs0gU5SGGVdtMNOiusoUlsRpZuCAMi1KJCu6mO6f1jUVFNoeq7M8NuZ4Bj0bTzcZGNUoQ1sm0MWceVpLbo8FoTQIw8BMBBvnw55/em7uggctbjR5OK+ONLOmQQsALGyd0N2mQrCrALBbdjNaRs886OY/zdZq8/m5eFb3PdxR/zwHWo5p1a2maT1yV0GLWUA+83yrGAKGEEBD2SMrAinF0iYog9VqidVqicOzZ7E8PMSZMw9jXOd1JVlwXrS+KI9qS9yTQz5q5mz36XMgJPaXfrJsMmaZlFIcQxk06T2gEmFvWIvbnr4R0tV9Awxchj2m5c2FJOZAnnS/Xq8xpLzwMKIyrdxHmJUDnmrAQGbD5hGBjmnx8PPaVGj6tATcS5mTyas8OdoDaJoSxophsfmfB1wlImFQAXlV9JAIMRJCKqY8HrdeXKpUxJoZyDMW3uIzyPsBQZhWKiZITh8zLl6bkAhlNwLfVOVrts6GMrw96uhBPnMc7ShCNoHGEHO9iDwsP5b0lNXjC+igB1wUXJ3SDjq4gSw7y0lVBdq0ngd34YMWn0sDEFMMVKC6SlfqYF4/ri3EWVeEIMDLxKBfEiZC5M6t9tGaCn1j8p+fWWW5UuMYBNo4teAV4G+JEsUaFjd2MiqXRTrWKmHbnZfeIZQ5KTEUDZW1Xq3YvMX7WCke6/Ua4zqzLyJkQWm1/BqwNuSB92JNpb1DM8U1b5JfsIzHMaBNR2dEnWdYDFRogcqkeY5pcRswMc4lIt/XFS8QQu7/IWOqTHlBWp6fZedlcVwANDsd131YNl8kLgJW/jmhGuIPtGBdr7RRgSLPCIyRiumMzCg/KjTfZEyoVBGyzXa7LAjCOszKGIlAzKw4nNrfJsaF+TY+r6uWNhj8ez1jArkf3ol0FJAu9yVsK0fbZapcnTuf6DTjLnjQUs1WAatY4kynp9G4jfKUzc1W5Hrn7rJ/+MUm67jksxXJfL1baTp8cffNnA9HpGognIlTB7TY1ekMnC+cUfYs6dYMcIoBFRFCec2/gDIwKgXEIUL5ESGBh83xfmjFNLg8xOHhWZwtq5vwPKMQYtH6kwiXOvWWPHUXsp3Pqc4zBV8qBWNNeihAoIAEWenCMyx/nur7pAMh8vPk/E0CGkkBq+Q1gaPSB0OggFUImKaEEDKDilzvg+3T4j6sMZsHC9OS5aG4faWy8j6vF0j1vCo9C/hwrpb26cKrDu3DMuyqHDo4JOeX9mmGDmDlUXZIvk+Jk06lppBUXXLVgtu7vccyhPMh9/+kzFYKs6qZFkAIIZYmtYGtGKYlR9lNIMSA4BbVZYbFjCuzozxqkhDLSv8iqnhtQpZMdfU3ChnnReCwoXll86BOiuRjP2mzRsWjugsftIqpxxZmttUHYwZkFmHWbtghB0P3or9ShiuowLPHiwFQVKwtwGVs03WlspXdmbvO0YXmqhw+4ypTmoKnsjWt2qwgWGC3jTBEnYOShWsRgmUy8TiNRUiNIqxiebfHoDyb5o+SAqgBH9VyYMpED9W0izAkTdVWJrXlSJ3f9WoQKVG1X1WqwpGYSZ1QFqYpsY6gQlYGRoS8Ykbu02qf8/B2ZsHKogCeztsA5CY128XPsEm0k6llwAczrCkhTW385NvMOIqigKDzqWRRWRs1UVr5h5HnVRrIvCC6HFSoK7PKpske02K2680YXuh31w+UtuOv9R1oPZCItf7ESCJypaZcpt1iw/D2clbWuFmIzlqHztFd8KCVh4RGhDggDgPisMCw0Ey3IwQtL7AFS9WPblHIax22A63IEk4AglQGM3eJH3c+ohW+Cs+kwzGJXjxNQnz9rBqM8WcTVM9hawarSL72Pks+YqHoFCGzrDhEDENEHPLQ95CVyCKg8qK4vJnnarXEcrXK2vZ6RIwDFgsjfErYYE0WKDQngOcD6YAQ1iANcG1qTEYIq3DW+wIsVoCaSbU1s/JbdlgmlZnCWBgDM4exLE01liHmwrQErjiaBki6QKn++Ps80EKEqOnTUoalw97tBG4OS4HUbz9i++csOGtfXFlznwxoGT8Tx9H1Y01YF3Y1jnY5KcPgSr1LRJlBliGpOp+pqq/cAFkI2/uo/IHrWDGTFaWhN2rQXTMSShwL8NXaHFfmouKyAl63MwdcwUSf52vx4rpl8ZdQ5pFFlAFL0bIjC6R11dfBGQDAk9ConEE8sGVuypB3ZdB+ld5zcxc8aFlNHkEnjYYyWS+EmhmF5l0rZDu+qle70rpUCFZ9FP3ypVF15oIRbQaMWK6l5Xe2rdVhEtLgEbn7oWk4Gik/Z6tKu7n2coDDpyrzmFmhAJVeczqdUGXG5UayTeBGZrV6MstJ5IEuebiwzOcUtpSFjAAWdQ4oSAnQOQGcfFw7AvoobKt535jVukxMALSUpalEysDq2qDCycWhDEEPZbdjZVrkGJYFIxd6uSa532Mn+oSPLrOCmYNmwF3mjU0KYDpqkPuyCiAErrVcN4C6OjJGW1fjmCSt3NExGCrEqbxJDHhJh9XLKMJgFVC7NiHHJwCBsJWpzOmPRuY5xhc67W9T+JyequLUm1lqjnjR1AO9HC6rxPXz8wNcFz5oQc2CMUaEYUAstmaenLctr7gScBl4UjH/shaKmrpyqfI3yQOUqVCOfcM8tzfMtRKePmj6mFeMxz6ba8GNvz7D6sUfpkHW32Vw4vKJQz4jxiJbsuauK4WPpn9rnedvjXn02zAMBZyKlp/0QApZK4yhLOYmNAPgvrOGaXXASxYDrNkEybfqRWLTVBjWVE+yZeApzIrU7OUZ1wTZz4oy88pMq+3L8qxAM3qeaSnDmShl1jHp1iS2T2uyowWT2TcruYrkmJYdhl4P2tDJwFTKmcu7gCzxIBC/6gX3Y63HzLRWpa9tzfHj/jKpj0EASoaMlj6mYIad9/Q0VUgt82ClhfVIMoIYRXnJ9Y247ywEHUVYCH8CV0MeHg8xcXf7tqxCyEAsSnju05L9uMrq7sokYwmXR+eW0aGFYgVWVIws6jRXV59cmRe/QrioY82accq2zo87AaBlXTVRtJy3YA+Auj5XfWFb/HNFdqbDULUQvpTK6G+rxgSE0FYjqcQz8TUxgfKovibU+u886zHUDjHTTYJnBEOA9F9pX5ZNR5+NtKuIMyMhB1aUpnIUqZWQVyEtoQfLCnpHB7haRkduxJz2Q+nINrlfjXzrMSx3FlOXbsXRez8nrSovo20RVNA0jKukg/uyhGmlBNg+LdN/VR/2cxKu41T+e1410BGKZJ4387PIDHdPbDIeMY0TdFLxJO/ZRtoIXzKKW93nY/NPkImqhwa4UPXNCMDpWfrQijJgGRb788yotM8NcqnWG1W0BT1zeKwgH8nNQNacTlvyqk0L1w//Urv4gd4/Fwi78EErAGwWDCHkodUx6jXQtrY6x7juuMyta0zrXyqytCCmanMqnTYwZyqE1V6slqMVKgT/ox8tw8wkHi7a/gd1KmyTbntN5jffp0owBBcum+djzHNPYox5pFMJzwrLxCPSuE9jHPMOuuOEEGJZyy6vzkDThDSukcYRaRpzV9YUQDEiL09U4uuAyS76Z465+6WPiiyTSknW1LOrQBAR0uQZVD1a0DKsui9rItOnVfV98Qi5prRMPgtQELMYLeZUNG2eiBtCQKRYFgeJAlo61L2dVOyrjQVjvw0Jz5uql1wSkCpMCwQDUtmfrnSRWdZyvcZ6PWK1Xgv7siMRQwyyC/ZsPQ8QwLLV3TKsth20IFwDl0xfSkAIQukyg637tsoXnbDn6HVZisqzGPNCaJF0bcKEvAeXnJOZrxWLwaEMPkkSx2CYKbQ9mwFUvSxwMo7NoYZhWcY15wIr6eH8MK4LH7TEBZ8pNXWtK3efsjgPXYbG5dwpcDVR87ebGDqUso+tvLdny9As2NWONMYmTpbP9JLZY1wzoFX9RGFPzsbgbOo6x96ZOCJfs6CdYTZG6LEZTvp+ZIHWEdN6jWm9RgSwGKIKWAHZHGZg0KIalMpZ+scqpmG/y6PpZCCGORObBdGeqYAekZyVORZTF/dpmbRb9oGaafXqJTyTEeXHaMgSXkoIiYCQCsht7qPjEuVygwnfsipU3+f7kjag+h4MO60Xx80TiRmwZIh7SXssps0oKAajULYES1vFLqParOJl5mPVgFbKwuUXsw9WRMEKqfaFCUNqotGCCAv9UNiV77/n+xvmg23Qv92DKi5iFuX6Y2JXp6nHuAQoN3/9yO4EgFaR6LH87DGtLuvp/JZavSGDWXOrKte2JuDCDSoA7CMekeNMC0X4tusk+jh6QFbGlU+bTQe+IbZ55VIqOoCPj2ZLAEjhk620zLQCH2w4sto6WJAVLZ/XlpsmTDFmVjWOmNYjxtUKq7OHWB0usT57CEwTBuR+r7BH0sjnmBalvLQRUgImPk+ZVU2THuX7VFYtr5kWJZ1fxX1a0ldVhJjMv+K+KjlPhWlNnmnVfVol/q6ONQVqAAEKGLk6kCzjxBNIQ+DV7qMI23rZJh2UYUCLrIDmdRT9wfen6rmYBbmPixfq5b6sMYPTaj1iPY5YrjLLYqY1kZoFQ8grrMTC9CN3IkWtdxaw2pzyT2QQVfFh89kzpA7jShk0GIhBCQFB9seS0YUwe8eJSU8LkoGGlVOZwoOEQLEsLE8Acn9wTm6S3YfL0LPS7+QV52YUJe+3VSW3yigBWZYhnHOJfJ/WLozrfLkTAFpQCsMSUm9WjKYj6K1WUud5Y8RvrfjkH8sd0aKEdXhNb17z4FZJ5q2g6ZgBrTnXZ1ibnnd8h85na6ZogbXEO3C8hWlF2QBSGNUM05LniQ810U0FuNarFdbLJdbLJQIBabHIwiGiqJ+M+YWDpFRGzFUsy5xlgAezHxl8wayqujezdFEWzn4pIz3ze2UQClEj3OvftWrRChpfN8kwL172OJsJM4DpwAEGIYAqU5/kf80uaiYljKo6G6bFQtYyLBnzQh687IhBu3xTImZQITOLpPUrUdWnWuth7pobfecFC1zC3HTwRA+4NA/5CC5vQn02gMXMqOuKmOARt5TK9J5gBpoFZWFG5PWD2/QdskmvWKbE1SiCJrAafOugzyPJAnASQMsJUS7Est32NuQ/YmbOAZY2VMO8WPNga4B8LtcuGWDYUD0LVAZSGlvjpshbRsfa2bxPcg22ZnTlgSgG9vMlfoSc55HDCZBRS8UkyPO08ijC3ACpSFceoWeBSoGhMK6QMC2mvAr8conl2bM4++BDOLO3hzMPPIR00Rp7EaDFgAEHWaO12iYzLWseLMyJz+SYVkLi/aLGzLQm3sF3tExL+7SaFS1KesYOw+JrNz8reYbF992Q806xt0yi1EdTHCEANOXJxBRYEw8IZeUEy7Ts8k2WaVmAcOZbe7YrZEBNoQmmj0/YWFZExrI473ocHdNarUYsS5/WNOkKD4H7RQtoIWRTYSh9OW7HYgYe18wKe2iAS3NQqrdtH/CsgvMlGaYlJuUQ8rJOILt8n7Q3NZ53nOkfyP30hUlGXkcw1uoImHNFMhtUJkKMSZhXDrIwLMkQ8smvKpOANNNWXkLe9Jfv1E+1SU8/orvgQSuYw6obLiMbAV8DwOZvNCJ/TqM7qrPMhLUXodl1lTaAUV1vjHfTYDXebpt497n8g/cDaoCsvBCCj7eis7lmRQKqFXI4KvhMn1AyWr5lXIb58CaRq+USq8MlVoeHiAEYl/sIaYE0RCBkcGShBiIEFsBl8IVjWNWROtf9oxXcNWi1TEvBKM9NM31bVC1/VIPWXDl3yt0YCIECIAEoTCsiBbPQKzasLG+Uni7DIlTn6j6YUSmzyrqD7bck2eBRGFZZBzGVARqclBCAkMoAhMDTHOzSTTEvm+kyxzRW16fkAUh9mlyt2nnNuPhaDvA1K4wavrwnzVeVC+dCkJ2l5RrQ5ZoMw9LfGUxYfrhmiBkxaNvrjGMTYdurxR5QdNeKcZ0nkKrdBQ9a2RXhCC3ER+WTRwasYFDWIQCaGoaqIlWA0lxviie7ecWyG4wqAvXNuo+Na22AbwhWozdmDBQNlLXwYhJKE58nNQcW0yAPvmCwOjxzBg898CD2hwEP3v9ETOs19oeItL+X+7ZiwLAYxBwpopmyeYwSeYaVEoh36B3LNvPCrEaklFfmSKmMakyEcWTGxExLGVSfaZHzP44j3AoYhrFJ3xYYtLYXd20B0AfFnITJ9ENQAYAgoMIMS5iWAVAXXJdpmfIk35eVR4SWeWJURlGmwihTcvOx8r5pmWmtVyPW5feUEjduROT1LKeU6xtNAJV9wWLpLw3I/doR0el3zLpnJ+a6/COpzy2wGbNZosysEuVV3SMPxOF+qdzHlVsC6fumvfeKVeaZQkcDEkWkWPq4QgKVcYRlV7r8HAmRgowibM4W0PN+KqC5UYRa6OVp4Y1WGWCZ8EihVOVOCGixUm/U+IANmVhXziY0EygUPDg4Udj6I2rcu1sjXQLmBiRRrCp0D6Sqe1X0RJNzlKphWJ6KdfPEJsXOlg+ocrjDvEwyHU6yGaViLspu2j4uZkUMXsvlEsvDQywPD7GIEevlRYhEmBYLYIh50EEZFi2tVMIiz7Sm6szx4cEJ0xEYVsW05hbAnZIV+Pqb31PN3fcq1eVcl119XwQy1YyLslmrhFj3zdVMy/ZFShkCEr+2b8syMDYJQtNb8nE0fVm8KC5vP2NNlXangESEmBJSafKEItgBxLIBZDQTf0VZBCrA4upKXDFdXe8xrPpa+rTCDMMiZlhwede64MM2QGCHrMsQcmh/sSaQJL1zTGtOMtWwVfdlNvJUlF2hWkUEtOwVoLZiHtOdCNCSAgQcbj06366AqyksE0lpPJ2RjQagtmlgVcA7PJ83ddSmgUbfaqJpGlZtOjQMS4C0ajC2YstaeOOoxzQWk9CkgySIBy5MmNZrrA8PcXjmDB68/34sAuG+J1yMabXCwWJAOjjAIgCLYQDt72XGNQwwYzMUuLhPaxwzSBUmNfFKHKt1ZlZ8XufzmhnSOHrGxMyr7ptKM31b/H7FtGTNwcK0euXR04tZh2/8FyEiAzCgZkEuTwKqvqwMoFZ0qcAt5QceRFENpiAz905GCOqk4bF8g5dmWpUJxIerdRk1mM+5jythLPPGmD1RYVtjIETkuVEhFrNgTDIfMJbfeb5TWUQ7qmBv9CtpB/PA0ptnRVStRVj6uEJMeTfllOOcTLtvwmaTXt2eWV4gmwsjRSQqfVeU054Qy6r9sSgGBbB59XlhilresTTTJN8hJ8d6AOPiXAhXIC8DGvbK750nwAJOCGgBkMIFlFJvJFqhXymz9uKKz5i0lBFJUHOmhl2iXAGX02irpG0JCXbQhvivWF9DAuvnc98L7QOfR0H34uFKLM9Dkx5hUEaz92zLaviGHQl4ZRPderUs/VqHWA0DVsslhpD7trAYMAQAMcpw4yGqwLGgZYe28xB7OVLqD1Awx1TiXDMtZWBt31bdZ2XP+pvZCpcTAJ4WAS9cKp3W+K/PJOXOjIudxAvmu2RhKxSFuu7TIhGAcm2KrR4R6Va+4JGCZRkvWfli4vyetC4QiaDkb6Qy1ywkNr/ZFGWTGTMu3spe67A3DIDrhclFHTTRYVha+6v2X+QCeQYGo9DOAiKbJCXHC7stoMZMCyGUJsb9xKwQq/ywjIvT23N1NTmq06o13094Pt2FD1qh+rELXmz0U7fy+t2KmXQAq6dFSUUSGt9BgiO5iuU8UvSyC1h8g+pH1Qucbk0vITdmEfpmUdypaNwMEmJbghmIMSWsVyssDw/x8EMPYS8EPHjRxcA44uKDPaTVRdgLAft7C4RpwjBEYG8vF1vZPTYCDrS4D02Z1Sp/Z72S72WgXAlgWqa1LgxrPXmmJQyrsyJGj2HVZsTeUHerFwBeyDQiwlRlCigd+ChmQYCCrm7gQNWAjLfqMNgV/6R9V7Y/jFmVPef1JXnli6nkb+6rWq7XGKcJy9KHtVqv86TicS2LJ+ekUDb8UUIkYJpC3gG71LM8Hy0gDpmBDEQgKiPuYh5xF7mvq57PZdq15Hh97ap5UWA5m0n7tBIvOkhwOxsz0w1GPkSzsPeswivNXOVIjHnFjVj6uPKKGbZPy4wiRJ6AnUcTMmDmfjeUZlYWhc99vbMVqgVbVnwSNA8bptX9d3x34YMWjDbBAjKoQLWuh/z1vdre7TQXVm0tMyG9ttpGN6aNyS90n2tjmY939g/n35p6nIeKYc31u0lwzYd63lVyKv0vAx4MQMuoQfbNmrtlFNyvZJiGHj4axGxrmjCu8jYmy+US+3sLrA6XWISA9XIJpIRFjKDFUDruQ+kTMeXGfVdlIjGbJeU8tss1yZl0gVsrsD2jsqMjC5MBNjAr249lyl/SXy3FI5nCzGhOcbH1UwuC3DbrkHlQ/H0uKymDQvEahlWlQedhMaNs+wDzav5Jlo5q+7ImBVBX23yelA8VkyFPmC4mUEl+XtsohmgEt8kdq2gapqXtG51GwfmRn1PnTCWvZNBFySvZuLFWdGsToWmQOn1GlUIRd6TMKgR/BluODOM6qrPyx1mZMM+wfJ4e77tz7gSAFiBaPWzBbCe7cyBm7d39b4lv1bgMJbcmgCaIIxBC+cZj6eqIzkqP9jUFLGOyLexBzGTGJMdgwQMyIOIK0NF/eZ7UerXC2TNnsAgBDx0cIKQJTzzYR1qvsBcDDvb3EGnCYrEADvYRQ8AwRARwtwbJgI9U+rTGVRkdaBnWNGFd+rjGcY1EmSFkhsSj/7gPixlTFlg8H6lZ3Z2vK4ZVmw1d1oZg5ld1MntO0+gUlQjS1DItgjFNVsoTKpCaSCcEK8vSYzSjIPWcRwumpPOyloVZLdeFaY1rATMm3L0GI2bDwM0wz9EUZjXkPq8BVNakLLto51qAGCumZdQEMulli12PHzCYUGFWRATesVh2Ms5LZhSQyf1eMfJAIa6TClh1E7PgKAM0in/HuFKeyhBjAoR5BaSU08qMy5soeaSjfkqf9fPcOmaRcpY8qZhWrlDdMI/qTghosbMmqVz0W9lV/TgYtsTIVdRTFRikJgHRjMr34DtpOVbd2LIWNxOVVtduIlp+eoZmotlP4ozKpRp9+ynxsRPaWgXCApZwhdlgMjTVTINzVZkWrwCum0buYblcYogBy+USAYTV3gJEKQ/GiAFEgwMtBq7Ec4HKgrxTEaqOeaU8b0hGO5JukeGFvWVN7Urm3IB7rMoeWh6B/2uuVdqulXK+rG3lsiH32ofpQ3NCrYpP0DlJNVPsMS6Z0kDk8muayK0pqIv1elaW+4Q69aQGdZsu0jhSIqTcueT6wvTcsyDYRFfAXcfD/fB9V01fVqjytlFwg581YqiWAEAgBMrsiYyynoumAKNhXsLIcHymNZ/mDVMHTJrde7YaHtNd8KAVOoc+2dIh2Ltta6EFrs53uWKREmVz3wcvckUecvgu0lsiVz8/T7XwkXANowzmXJpb0KVo5loUi3QGBe4TWa/WODw8xBACzpzZRwThoYM9pHGFvSFgXB0ggrC3twCmETFG7C2G3J/B32JmU5jWJEyrjBKUvq3CsMZ1ZlT1qL+KISnDQrmGf14xLhHyUHu/lL6VOL6Cu7wWwJmrE1QLm3xTOv+howV5XhifLUOjolWzeZQZlTArGRVYmFUirMuZ+wJXpS9rVfqyuA9rPa7FTMjg5RIpgpvVRAWY3BrKfKgEABHlRz/DkP2ohcQ+821vVn5Yn2U6Qd5ni8y8Lb4uDKwoflStTME5bQd1zEad2w2g+3NRliWx16flzjxPKxQdtKxcgyR9nvk7DKa9RKvCUDMt3uGYOE9FQaA6W4/tLnjQ6jllGn7Gemtv7RdMKIGo5sA3PRexcGWfirONfi4OgetHMA3HwqANr05o8SWMa8afUR6DDdX0eWUNz4qCYMLblWHVLlRXQc4h5HUIY1mPUCYBd5inNS0QSFZczwM4MttixjUMMTOtAKxW+yDKfVvDEAFK8k1Rs4lKn1Y1SrCwL5lkW4SoXbGiNuXpmWFWhRH/3XZIrTIZoaPAGk1AikeqAWk4nIsq3H29EsIp7YB83MmmQVWyehTg7JF4XUUdlCGruFuGZfLezhGDiaekNYS2XRiPmr7SZ2fYLaTctP0TAzhTnPI7GFbEeeyA0nxWCwGOUVHw1/VzuZZzO4ChLu9AWWBQVT9ETsFxL7Biwv6q4Lzb0sa9nPRD2h3D4ppCps5Kes4duU4caNX5Xi+1Ug9XnXN9wJpnXL33yy++AVspxS5tVRLYimcb4CbXj1fPqU4XmgctI2pebATexohW90S3DQEBEUMcsBgWmIYFFotFnksVI/KCoFG16qKpOUGYUlk5YY1hiDh7eIgQgIOHFhjHEYsYsF6vEEHY39sDjSMWw4D9/QViDBh46FhBGN4va+K+rZFXvlib0YIJ4+QZVjt0vQYuC159M6A+y01dAEvwqQCWL6SmfKy5jMofBf18o66TbL7lN1Pp2Ehl9g571y018z8GovU0FiY1Ik2kDGssfVflel1WEMmjBQvTsgyrMK48CGPCRJOBWgWTbv2zVY7YeKGMhXhUH6hs7UIAkvR9ubpq8hPkfmj4UOBBQN4kGyjzpMqzMnpQFALOf+i1grI/c3nMLoxQyiwEtVYQKsYFmNGEoQw+4RUxyvwujisSKAF5B+u8NmWO4qwqLrctgMskbvLdJVnJScK0NpDWnd2JAy3AMIYq33XpFa6TnMkwVcafneNwgrvsMLjy/aqE6rlNtRkgmD8WWDV+dYnXFWtXOlSjTM2IWt/5ftHcar8VYIX6RXJ3RTPkSaBDHDDEQTeJtKzLqdk8kVX7lPKeS6P0by2GITOuGLBarREArBe5byvEMlcrDqonEMkq8s1ACMei2HsLPi65pU7VpUdy7gCYATijt6hAsvSz6c+iIghUO7fCUvyUF8TCkBV2iRPXV2Em5hs2nQS4ofGyB1Y1QnBK2p/l98hidtXry8rlqvHkSLYinOofzkPbDpRYzyt5jWIg5agfqq0kOSurvimWB1QrytU8rYpxOcA1NStUbSfLNlV4ZR9WhCqnKoYV7N02u+aYvMkQ8xdgk2h3HqYVClKBvNJ0XHciQSu7PkepAUsotPVT+lhcYczXdXmnKayu6wPWXHi7M64dI3pkf6gakcaItTJ7v9ayemEFILOdYYG02MO4t4e9vT3sLfJ5sRiEeaWQF78lBAGsKWUtfj2OeeDFaoUYgDN7A1KasBgCpmnEECPW6z2AEhaLBaZpH4Pp2xKmy8A16TJR9aaMYg4EPJiZA6VKyURQ7ucwgAXxX/VllbCb/Jrt66sOsgvSZh/EaRPTWIJCplV58nUiZlq23IJc89vSdzVOUha2L2stfVrMtMrKF9KX1TKtiQe+FIVEU9lJ/0z10unFXvVkBjYXnvQPuTZPeQULY6aUz1srDOcf+efKYFUuCJCVOOVRhpqkRKpIeCuPSQ5XtJBBLsuqAozBM67MADt9WilbPKiANIGAsho8b6wZgi0DzDIkZvHslVdb8X7YRNsP46juRIJWw3CMFppvFIZV2WGtX6n2HaHB4VubrvteAKhjFz+q68552P7Wsb+3KQg7kslqwOpd74cCcy4Y1/izUiBMa8ggxWbCDFqxbBwZ2LtUfBH2PNSaVwUfJ4xDXv5nWA9YrdcIAViPewCAcRFBNCCWxh6tmRAVw2CWVxQYPoJKQCMNS9pFmJIqRaIcmU9BtX7ub8mCrK5H1pxsvlmHxYcJR8JzoFXO5Y2KQ/i42KKD3iOQ2WmYy8HMSbOMy+w/JsPiJzuhfH4bFBb4DcOyWW7iGbhaVYC1yTG8MWDFEBC4v5MZDUIeoFCVjysrZlxNF0S/H1vCCv6s36j7tsyHCvjK3CypdHZIWIDLibpPizMLdX3z550dl0UgGVFtM0dYbtXGjutODmixVrk1SwylJQLK6smOggMt09rR1bJqDnDqsOe+dTTAejQcA5ZcmV+1LV6viZBt52WH6RgisAhYpIT9aR/jwQEOLjrANI1YHh5gWk84u1gARBjXvNMxwMs4TWXrivUUsRpHhBhwdrkEEWGxKIxrETFO+wgxr5CBQBiGAYQ8b2sxDACAyBQJlEd1xZCXBTK7xorWCpj+AMG7fOY5RcVkR7Fol6y1iimINess9EcBk8oVwZXICxL7TSK7uWI+TyU8XlUklTlvRNxXVXM6TksyV1qu7DvvkQXDtHjUIPdl+Z2YxzSVvq5cXquy2slqzftmZWWD5215htVxrBtw+9IoeifCtyes7dmbqfngvOdPJiSENM84asWyZViqVMRi8uVVWbgPjIIt//Je0544edz3WeQDb0dS0kOB8hYmoLLae9WnVfbZosLaqABgSkm2Pollp+NU9bH0SobrPDiv7KAWGOWMPZ+jODs5oAWUmrwBaIr2FNhsU2p+Dy9cX1MXUFre3poIDUB2o9MHLlXYmRlCNLldwmkkXNitljTBNzecXivXQSMoT91CmqbRcgNBMclkNpU3iByGAcOiDMxYZOZFKZWRfyTSitkAa/Z2hfB1OYYxmw9jjFiPY2FcEwjAME0YuO+MtVdVB3MDY6EYfAOUa6Z/Nj9MtlDgtHozIphhCaPRvk+rnDQmFkMtpBfKABf7UYbF3zFCw6SvKVOj8HnNXsPMZky4NSMp1X1ZhTEJ+yr9j0TCtLi83FqTDApcp2abMfvzddB22LDZzJ/zw1DYmH9ewKuMaAVQRinynCivwNbl1TApM3+zJd0to2KmJfO6YJhaMFXPWIzYfMi/g21/nWuXRS5vlXE1fVqh8uM0BTgEz4MaOe+0rko+nQewYndyQKvJ6OyouRX83ZYl70CP50CsBaz6Azx8ua5QW0OfY1zb4tpJ305uFqhrPbzPsLilsm/BBAtmISDEDFR7e/u46KKL8qi0iy9GmhJWy0MMQwBRGX4+juDBCROygBzThNU0AgFYrAdQABbLAYkIw96AcZoQYsB63EOIEYtpAYQ8gpAFzSIOkjqnUbIdKhbUinnYvM5tifk5FVDlvYnKEtqsPfORCiuyIyFdv5nJ0yB5G6paZAZVVPlqzXUCKsy4gLLKCJRxaemU+3VDyH+4T8vuD+bnZ5m+K1LGldcYzPenstbgOE3lnLCe1jIog7hDxygIqhNUwrSKd61P1IAVYwadYWBmFco0i5gPHgw0KNMiEzC3Uc7Pui1aYJJ8LKur8zUPykBZtDkWppTnceVv5Xldmv3SUlohZsoof9iNSAYKYyKZl+X6tsoq8TkNCXnfr1wvYoy5LCL0TISUch0nVvBqV+IoYAs9e0Xt3JHr5ICWc6HzS4USz3HYViFajcTer6+L/Vsqtd53Hwim16xqlE1xWmWyG8cAHxWTLoDHAXRs41tcBahWy1d9jx0JW3G5ThxOMT8YwKr7BNlEMwwDFsOAvb0F9vYX2N/fy0PQ9xaYQjY52W/n7dyzMM3mqIShMK5hqphWjFiPExACxmkCiDDFLLRSmVVp9R4q+cdKouZfUCAyz4U1gVlJuSYFLDuYI4kfE0ZFGvibDNRq3mOQUT8uHHiw5DiJEiEAaYQIdcQJAQjFaEjQASMsgApDatYVTKZvi0cX8lytSeds8fucL5z3tj6r9u/rcKhiyyDP7NkCl5j6q7MeESGWuYNluafA5VgWFbbHHGAJk+L2TxWQWYAxbI2/5Rkby6fSsGoLku3TCsUkaPIrcJ7Zeh0CZD+8oDkZgm+Pm5nWvBSx+VYD14bXjuxOGGjVQz77ftTRjkzEFmatBpZwBLBYg+fyNQVuXqsrCIXtMXcgVkVn1m8dzW1hd699JnkMzc/8ABbb0LywycyimC94o8cADIuI/YM8GTilKeffNGJ/bw8xBoyrNZaHy7J54wTE3L80EmGdEpAmDNMICkBcD5hACEPEmPI+S2NKiMOAvSnv3rsYBqBo27RYIMCsuG3SKSwJyIQq5bLyoMDmyvx7Qm64E18X09pI1oTmN4u0jve5jSE6iwyDDzMtINcxBshmGSUDJCAyw8k942Im1TItLTuCLvjLi9r6UYNJRgnmszKu9ZiXxsq7EE+6tiCvM0nlC9xeGj0vYL6iQ4VzVCYVYkAcsgl4iBmU7MCfbJIeyiAgHQw0DJ5pkckbshlvIynSWsuE60PgOVsMFCFf57mI5Vkw33JNTYTDxnasgA3po0dhUlTKm3h+Fpn9thIhxCR9WNznZc3LQF4dXqwLDLYzRWGBS+6V8j1fZsITBFqBS69rcgOsllau3Ba+Pd8erFTx8ACkW1UzQJVTCFU4ngmxvlhhkeKhj7DYjDcDjLnt7PAVznZf2BRcD91ZS7PAZeejKHJqRa47rEmETiwrVyyGAXuLBfb29pBSwv5qD4GAaeT9rwrAlBVPmWUk0kVcxyJMhzFiHCfEGDEW8+I4TgDygIIhAjHllTKQMlyICYT8SDoFKNMfxV7ttdwnd9QrRjgW5PJUr8xQFo2HoVi972tcbTgQoavAZ/vV/DfkpcDh+2WodF1Be1S7Osv8LWZX3A+mpkvLAJ1iD2U2yrg0VeI/dPzxvfo61gyrOqLvu5rzx1+nKoOZeXOetXXevFvkhBtlaAc9sEwxDKsn6wPLkDKZuTf4hCi0jAuk9aynhzulof4oJ3CznOWrU6Y154R2bOUr5oVNOblrOFV4Ui9NDe59x9fhRrs8sqvfryrUFmaPpgZ3I2PT0ou8ZVwRhMlCfYmnNmI7CCH3OUTs7S2QLtoHkIdG7+0tEAOwWq0xxFj23cqThmMRNCkAKQBTIoSUsOIBF3GdtcthwJQShmHAfmFae9MCIURZHSOGgCGWBXVjUIEuJj0FGl4jUBgOGDThwGkqQ7hH0k0ie0wrdcolVrkLBJmHZcvRjhrUtQ/nN5mEOxemJcyrhOmKLMgahLyvVWZQya+AUUYTygoipY9rSoT1Ovdp5T2yJmFmqayO74gFK50hAGVlFK+EmsRbawWDjjAqw7QG3b2Y7w9FQYr2KOzLV1eSc2avkPrbxIgIvPYgMycZjMOIVkx0VEbpZVLEzKbUI7MGYCgKg8sD0mRnUZNN8PmcStbwShkRUVbGgPRpBQqOYYXSDxulL87cTwExJrNmoVc+Nzkm0ZzAWpk6jjs5oGXdjAYgegxntlKlmXCqH9ZW40Psvq6VOTSYVtQwsV8rO6s17158tjyrtC0JTvzU4Xv/DXBVTEDD2IS6PAyXw1VtWLRlAww2/BgZwAYQLbC/vw8ASOOIaYwYQwmfUNYSzEKOzXbMAuyowsExrREB+YxhwDhFDCGPxY8BGQFNvPr7XrVMqnd/btfees1CTrtUSVNKtdrDrEj+yvs7CBHjixmUNTlK0DYedVrE7GhHEOpZVnZPeTV3nadVFuWl0peFittx9WiYTWk/xreyBgij0gPN+73n9XfqUYK9+yEE8P5oXEds+dQDNWaZlgNDZloE7VogF15dtB6wfH8YgNzdIHHjPOCsNGCPbQwrg6zP887iCGTqY+UIvt6dqzuZoLXJbcm5rYRnCxtibamRI+QvBaBgKuVm1r3F1cDzCDgDvM3NSrSyFSWKkDD3AfVbZVgIWRvc29sDLr4Yi70FFsOA9WqN/cVC9tLiLUUyeOX5L4hZP2AWsx5HEAhxpSOn9qfM/vb2cr/ZYhhAlFfo2NvLZTGUHY51K4t88H5Z2bTl95VyZ+6r4nO9aWQ5j8kzHMnJSiCwcLfgyQLcgmS7/NTMP9PXRTCjC2eKPRWA4fTL6EBeo7GsQcgMahzHwrBGGTWYl2waC8MqoNXoTi2QIESXD22F5PUqWdkpzKpMTJfRgpaBxYg4ZL9xmJmnVdgRryRRA1ouN3Jn/i3r8Ek4VtGIErZ9h+tTKGyOB0dwWUq8gml/xXxSA1d+FGRDzNymmGnVfVeElEIZGJsbdyI2HZb5XAlADEgUZI5i5JGypHHhutuU0/lELHw+glattvI9MWthFrlshbDXc06Bq9PcBKB0rgZPyhMhFtp3+s4zo7pxiamg0VNnQqsIVrB5VrPPHmCV29auLtqtcb2+H45vKGyLiOdrEfb29jI4cT9JYasgkt2JpWw4XKP1j9OEGPLowVDOADAV8ItTEXQoAzKECfldldm8ZoGrx6wckKRq2SbOAzI5F3yuugw1N5QVtUJzF8eQJyxrk5Zs0yhD9MuK900f1ty1X72dhXT9xXr0n7LzuTQysNWsqGVJXQY2y7S4TncYX9V2iLXMDngwY6rbevvcr1Vo0+vBqPQNW6I0x7TILAHnFEajRBJfG63J+DfZXL7tGZf16Bm/u4u2Zp0bgsXtXrz70Ic+hK/7uq/DNddcgxAC3vve97rnc52Xt956q/h55jOf2Ty/5ZZbzikhx3NUFwOX4/kIupy9xiRacq2lzZXveXLbk7TFR+fxpjekb0ttCt2XbN+W7d+IccBisYf9vX0cXHSAiy4+wMUXX4yLn1COiy/CxRdfhIsuOsDBwT72Dvax2NvDsFggDDFrhsjD4ddjXoVhuVrhcLXC2eUhDpeHODxc4nCZ7x2uVlit11it1litx3yMeUkoPnhh3rwiuW5gOJpjmvIgEN1+g/eHMtueJGVrTXEHeCkjP2vNnE11LcvSsIJowASdjK1bregxdUBmnEw6kumTWms+sPnVHbKAMefbJCMO7VJNJsGq2cAemx0DiZ1vlY8BQ4yGYXWYVswm4cjD3d2hzG3ORFhVYikXWz5sRuwBEb8mSgFIJmVnRQtukIsNZ05GhPq3AdvuOejgk2iVBXM/H9H0F84MaOF8mdWsmW6eu6A7MtN6+OGHcd111+F1r3sdXvnKVzbP77nnHnf9m7/5m3j961+PV73qVe7+O97xDrzhDW+Q60suueSoUTkPzg8TyPRWoazxXWl986a4osaIBcxqTa3WBdbWGjrT+858YRvCWMXX+eiEMcfQOP4z3yq/rJzVJJADLo2b8WgA3DEuG4HAJp6yJxYG0GKBqeQbxQiaJifqIqArG5hws3DOIDLGWDGuPG+LYhTGwyHY9wlmCw9S00/b7zPTh8UvmHxUC0uvbyWIRm4dSbkUQbCTCcYIVZvf1K/x/Em/E7GaFmf7tKh/JgNYvi8rMK8xgm8Tw7KvbmdUKmThzrNMS9rjTF+YyR8SQ8OGtQZnzmD/VDGq6jnnA0cNIcjK7jYfsvwq/kMQS4FnWp5x+cx0J9NeW0aMYG0s5ZctK6lT3MbPnz5+ZNC68cYbceONN84+v+qqq9z1r/7qr+IlL3kJnvWsZ7n7l1xySeN3zi2XSyyXS7l+4IEHjhBj4ziXZ9xMGR3RcQB9RqXl2gMsq1n7cxv+XAIeQVc+3Wh0c45M/BuPXmsUoUhqfmMwjDGChgHD3iKPhgpl99xhyI0jpWKSzHkeRIUtu6qCMNGE9ZSHGvOeQkMBLoD7tvIKAnvDkAWoxFX7fjxoFQZTNOE8ai8Ppc99WpOfj2VNZJydRsONMXbPfeUlZx6b2qReVUTLnnOWGPBJOsqwUzLSlzWV9IzVihe2b46ZWUoJ06jMbJI+r9wHyZOJbaUIM5W9iG0TI58PIXqWZJlWbO6Hwr7KEaIM4KlNkRkd+oDYdVKVc4VIZaRqz8SXz5xUBqb8U1bEKPt/5Tlc3AcGEMqu20TzcbGuxFmV5Ppc4sMNjci0cVU4XZ43ZaRTXXJa24wR5YqR6xzR68jmwaO4T3/60/iN3/gNvP71r2+e3XLLLbjiiivw/Oc/H7feemseyTXjbr75Zlx66aVyfPEXf/HxIrSlnI+al405hj9TKsFOFavjRM6LxsmNislHTwPckADRKjVu/kCVN/UNFuFGE+b02W/3CVzlWo8F3r3mbxmASSubd4ZoJ4QOZaKomThq9ubib7jlk4xJbEqTbo3BJrHEO+lOhlWQX3ePklxrv5cCQZL60a8nwjCMQKh/bxKYHCSJ1Osgl/M/k8cbDpQ0JTL9WR0GmdifHeRhGZablwURjBYwNC+UjQso19p+j1GFufszjKoxm5X3GobV+mNB7OtqT0ntmQi5T89e99mve5/q8DuNjtumycU+0zL1Dv56s2If/Dc2yThbJX0gzUeOIisf0YEY7373u3HJJZc0ZsTv//7vxwte8AJcfvnl+PCHP4ybbroJ99xzD975znd2w7npppvwlre8Ra4feOCBHYFrC7Wq3PkgKqqw8KS/aCrYDnE6WpTtl3cC5eME3SNLem8LNVX5Y08uTgBEMHkTmhf2MUbdbj2VkV7Es/TLEk+UADZVjZOz4TDrEXa1XoMS7zGU2ZqOFms1Ot6PSleQKOylCBUdPVj1N6UKCCQjTR+CA+ZYCcviOwQnrKgwOzW5OY5fxV1BJTOnPOEX0NGD8g1JW56bo/1wk0wlYOBmBmUXw028saNhWCkl0FSZiYLNA6AZKSgVNrHeL/WA88n2WfEoQmZg0fTD1CMEmYE1gMefIasuct7U9cHEtfzI7V/NyQpyZc3ByAyHHPOhrFUVhoW86GxippXfZysCESHEaBibd7ZFWmbF16BsPk+GadVMsLX4ZDBPIZeF1JjyTpJva110+URa1qx8bmWwM+4RBa2f/umfxmte8xpcdNFF7r4FoOc973nY39/Hd3/3d+Pmm2/GwcFBE87BwUH3fnYNTaieHU1Ub2JbtbasFbK9T6ZPh+NCtFtcCLwemQ1Tr9sy7sFK6zQ+M6lsHvf899IQpLLO8X8WhKXNtN80b9XMgUzCZT8rZowQHurYJgFA9Noph2WFSiqANYWAKWVQmkIZP1gLKeWEIlRZ4Cjg5oS2TZeToCt4wwoSzDMtlwednFXQ8uXltX3fr5bETMdav5qwbDlwepQNaJ4KG9DM9d+RPEnm2uWGMUP16zD35dj8y4cH9hp4dmJaNSPjeJQqrv6CAfMecxKZbICos8Zg6PdVNefA5/58Lw2fwUzLXCSeqTeBqxrrSxUK10yrD4S6IG+uv1AFcrZuugDa/O7U913cIwZav//7v48777wTv/iLv7jV7/XXX49xHPGXf/mXeM5znnOErwTJjI1+jhRiraMez0VZAwxlKKsKY654m+JXKeSzWtWj744WiYZZVZhXh6aAoMIOrgO/uFjMOCVzYgCGEMArrlNKSAXpaUoFWXQABZsFMRLWvKtrAKYQREOOlZlFf3hmIytSlL6tehi8RLnUiVhCpQ7LapmWUV4Ey22fFJq8EWg1AJV4sdppwjSNEgYDVs5y1bQJZjCJMY+KeZDNfc0AjEnOlPTczssKchKTYFXBqdrLmdlU7rPKK7O3LKpcm3wcKpZlRxPKIA1j+quNB1rsrLCUPs4GtEt9SFR2J2jNhVyewtJqxYyorPrOrDRo2RIZxpa9RmsO6ICHBQVhVGBwn2daTjxZXHRFktXrEAJCKkug8eoesjUaSUDMgnP5Dc6U/7gArZ/6qZ/CC1/4Qlx33XVb/d5xxx2IMeLKK6882kcCZIuBxWKBvb0hT0otjrgWGcf3BJpmEGr2udRrn8F1hoss4IqdeJl+ruZF0+NzhWG1oAzVucd2uvEJttI3MbQvdMIJPjFk7gbNA9k7i/IeCwEJIRACqAgPs29W4IZWzA0oeTQkpCkvo1QyTMEqpNwY8h4LIEqStyhan4IWGdAqABbLeuopleWaSkMZFnmpnziUYfJ5EV05142oAi2+JdNcyrLeocQtMlMJBYplI8Fc5oQyUrEsnqpmryGbr8pCvokIMSTEgUAhIZY8QTE954ErlYBDFvoi+FLCME35HRSASmLoBJeIpJCyWTBSNkUlImAc8hJAMQNXLMs4IQ4YUgKFiJgSENeYpgTEiKEMzEhTWa6KKQmAsBhkdfUs0LhicazyC1wUAlZDAatyZuEnO14PCl4hBNlyRI6yjFOU7UgGDCG6rUlAQIwTYsx5hhCRShGGMCCEUfNemlBROoL/vlVEmmH0BWDF3BmCmi5jQBiqwTmWIco3bbVU5mwVE1YypqG0jSErIjHmCePTqIOHsoVdFaKsnJizsG6IEsX3pzQV/VDf39/fx/7+Pvb29rEYFlJec6xrmzsyaD300EO466675Pruu+/GHXfcgcsvvxxPf/rTAeQ+p1/6pV/Cf/7P/7l5/7bbbsPtt9+Ol7zkJbjkkktw22234c1vfjO+7du+DU95ylOOFJeAgL3FHi462MdllzwJQ8yLoa7Xo8z2p8QrXJeGTUlpPtDVTth1n1eU2t6398RUw5oUCwhS62/DOJxaT47Ku76Nfgxc/ABTmTvJ4G/MomTPa/X9YB8Sq2EZrALM6L+AMnorL5OUG2kWSLmfJK8nOO4d5OvCBDJwlbALpRHg53v2GyHoyEGivGJG6YthphVQ4hAj9orQWwwLDDErPizk2NwYTD7WbYotM8UKKUPgh1K3dEi86csxehQFHl4fi1BuRw8W+0vufyh9cON6gg5yUMHZgFZhBlNZE3BvvRS2JSu+OwDWOun75MrowTLIYprMsHciWYtwVdrdWkYP5pUxUvHPeRBKZub0lY0XQ3B1n88E3s5M1xCUtQSHwqgEvPw8LGVm2rfF60061hX8tTAm3i9szOx0tcrz+KYyH61m0qEoILYs+VoBLadVzmWvNrt6h2PbDHLRC3ZnztRGaSmcyjmWewa8koxwtfMGKa9WUsoeoCKu9Frum8qs3zHzDylJfd/b38cXfMHT8OQnX4YnPulJuPjii7G3t5c3ex0GB2K7uCOD1kc+8hG85CUvkWvun3rta1+Ln/3ZnwUAvOc97wER4dWvfnXz/sHBAd7znvfgbW97G5bLJa699lq8+c1vdv1cu7oQkNeoWyxw8cUXIVHCcrnKy8akqUzSm1zDk+HKdeZ3HPk/UHY0Exl/owTC2rVlWZUf+7NquF5gbi9UFTxVRKn304CWBL8ZtCx09kELDWiFwAvSBgGYnKasVExpQhoWGOIiC+NpzI2O84tXlLV5WMBAv5HBiFkOKJsFxYQloAURHosi5BbDkNn6wKAlhkcRDk25BwsO+XcUAZHFLUc7CcOqstOYxcQcGWAmt6rWT0TAsAYlQlwY0JI6bEW9nqmADKWEuD4oAyYmydumNoYAbhcMbKKlF4HEC/4KmE0MXmPe12y0W5YYU6KAFrOFQQRyVblcSkIpZNm8sRpIwdd2QnGIvLWLBSy9Hy2YueuooJXyor48QXw9jliv1wL8zDhsi9BvaLlKfxvsdc6D2h+zrlwHFOxCaTtsSm36n9jKYCtaqYdqqs6KFEyZTGaXaQtWKWkbEzbF6aXK1EtlcBLXlVLv2O3t7ePJT74MT7rkEhwcHGBvb98pDkdlWoF26kV7fLkHHngAl156Kf7qr/4Kw2KB1WqNz/7t32K5WuGhBx/GOI1YrddIU8J6vYasPE1JdkllMGMBM+s6oNV3mzKcqnA2uDkCtf3N4rFBv/k4HT10/6lMAzS8UpUcmJUs04ZrgBhFowf3jUymURiAt+hgY29AS+Nj4iKNzjdmFQb5nbrhxOBDm2PXVF0Z8ufMJ3zXtzSjXJRvCDjWWnR5V/uRzIAI/raJocslMpODJwa5pHnjosQlqvG1gpkFmGWQgPbtOFMUqpUcyOeXAqRJe5PJtbIYnCDXvDK/raDnsAOfc+DRPdPfMVQ1usQ9FcXHLVNFvr/NZ6MJX+LmE2iVofZeC0qcXuPbiyMbcZN15OoeubO0NUDZuq/EWn3rulZVHf0O10QG5WzOfcITn4SDg4vw1C+4EgcHB7j00suwt7eHiy++GIvFAhdddBEeeughfMVXfAXuv/9+PPnJT57N3wt67cGsFQ9YLBIuuugAwzAAiTJoFRq/Xg/ZZLEuoFVMGWk6CmjZjx4zsruqBp3wj/zJ48bx2B/Riu4AC9omi35YNbzsTzXBZADLlAv5MO1PJ/dt9KTxGa2QyDX0WAk+EYSORcILVZ/q6oavSVTFu9Z/NCYdAVVuCwyzlkwAbz+BTt1tqxn3JVnTTg9E9fsWeH26lEk2ArD411GFhnUSuXB9XnozdteZfBGwC6pWsBVCmQiq5xY4WrOa9edkvwHtTNR5GaoO4NdR7pQhqvj2n/m6YONnEaoP8sbVdc+CWPWbyw6o0lW3n274vg7qgBNWAPN2L9kkuI+DgwPs7+93B2E85gMxHg0XYsT+3l4x7wRMU8KTLr4YU5qwXC6RpgnL1SozrlEZF4mJJAuAjdXvfIFWL6yee1QA59F3m5PVF6Jz/o7tSCPSj08tTo8e/Hl1m6ndzl/tYf7R3Ixa17lJvYfbvrtzRm/2GLZ76XqYBYBaZ3pE0rT9ha0AtYtr0tJ5eMSw5m5btjsMA0LIuzYMQ8T+fiYXBwcHhXAs3CIBu7gLG7SgW6TnTvQJKS0wpJiBqSy/n4Y8WTSV0WPZzt92pJ66U3fqTt2pO3fnQStgsViAtxziEaD1IAzdFmazu6BBy+5/w5kC8FYT+TwM2Ty4v7/vOqHddt+n7tSdulN36s6bq4f6Mzjp9ITMsDID0/Mu7oIGLQDOHmrRmoErT8QrW6qXDHS291PQOnWn7tSduvPqarlsGVcPxD5v+rTqRVEZnGyn4mKxcObAejM66//UnbpTd+pO3bk5N5esw7jseW9vT8BssdgNji5o0LLOAhcRiVkQ8CDGz09Z1qk7dafu1D1yrrfiRb1w8eff6MGSCTX46Bpa7aTiml2dAtepO3Wn7tSdX2fZlj3P7RnHZsJd3AUJWgw0Dz74YNM/ZUGqBqtHA7AYMB/t8Oz8iJPi6rSf77w9yrcfqfc2+X+04nBcd1rXz587iXW9B1ybwOuhhx6ScDe5CxK0HnzwQQBodkM+dafu1J26U3dhuwcffBCXXnrp7PMLchmnlBLuvPNOfMVXfAX+6q/+auOSH493xxtaXujpAE5OWk7T8fhyp+l4/LlHIi1EhAcffBDXXHPNRlPhBcm0Yoz4wi/8QgDAk5/85Au+AgAnJx3AyUnLaToeX+40HY8/d77Tsolhsdut5+vUnbpTd+pO3al7HLhT0Dp1p+7UnbpTd8G4Cxa0Dg4O8MM//MM4ODh4rKNyTu6kpAM4OWk5Tcfjy52m4/HnHsu0XJADMU7dqTt1p+7UfX66C5ZpnbpTd+pO3an7/HOnoHXqTt2pO3Wn7oJxp6B16k7dqTt1p+6CcaegdepO3ak7dafugnGnoHXqTt2pO3Wn7oJxFyxo/cRP/ASe+cxn4qKLLsL111+PP/qjP3qso7TR3Xzzzfj7f//v45JLLsGVV16Jb/zGb8Sdd97p/Lz4xS9ulvL/nu/5nscoxn33tre9rYnjl33Zl8nzw8NDvPGNb8QVV1yBJz3pSXjVq16FT3/6049hjPvumc98ZnfrhDe+8Y0AHr9l8aEPfQhf93Vfh2uuuQYhBLz3ve91z4kI//7f/3tcffXVuPjii/Gyl70MH//4x52fz33uc3jNa16DJz/5ybjsssvw+te/XhYrfTTdprSs12u89a1vxXOf+1w88YlPxDXXXINv//Zvx9/8zd+4MHrleMsttzxu0gEA3/Ed39HE8YYbbnB+Hg9lsi0dvfYSQsCtt94qfh6N8rggQesXf/EX8Za3vAU//MM/jI997GO47rrr8IpXvAKf+cxnHuuozbrf+73fwxvf+Eb84R/+Id73vvdhvV7j5S9/OR5++GHn7w1veAPuueceOX7kR37kMYrxvPu7f/fvujj+wR/8gTx785vfjF/7tV/DL/3SL+H3fu/38Dd/8zd45Stf+RjGtu/++I//2KXhfe97HwDgn/2zfyZ+Ho9l8fDDD+O6667DT/zET3Sf/8iP/Ah+7Md+DD/5kz+J22+/HU984hPxile8AoeHh+LnNa95Df73//7feN/73odf//Vfx4c+9CF813d916OVBHGb0nLmzBl87GMfww/90A/hYx/7GH75l38Zd955J77+67++8fuOd7zDldP3fd/3PRrRF7etTADghhtucHH8hV/4Bff88VAm29Jh43/PPffgp3/6pxFCwKte9Srn7xEvD7oA3Vd/9VfTG9/4RrmepomuueYauvnmmx/DWB3NfeYznyEA9Hu/93ty7x/9o39EP/ADP/DYRWoH98M//MN03XXXdZ/dd999tLe3R7/0S78k9/7iL/6CANBtt932KMXweO4HfuAH6NnPfjallIjowigLAPQrv/Ircp1SoquuuopuvfVWuXfffffRwcEB/cIv/AIREf35n/85AaA//uM/Fj+/+Zu/SSEE+r//9/8+anGvXZ2WnvujP/ojAkCf/OQn5d4znvEM+tEf/dFHNnJHcL10vPa1r6Vv+IZvmH3n8Vgmu5THN3zDN9DXfu3XunuPRnlccExrtVrhox/9KF72spfJvRgjXvayl+G22257DGN2NHf//fcDAC6//HJ3/+d//ufx1Kc+FV/5lV+Jm266CWfOnHksorfRffzjH8c111yDZz3rWXjNa16DT33qUwCAj370o1iv165svuzLvgxPf/rTH9dls1qt8HM/93N43ete5/YHuhDKwrq7774b9957r8v/Sy+9FNdff73k/2233YbLLrsMX/VVXyV+XvaylyHGiNtvv/1Rj/NR3P33348QAi677DJ3/5ZbbsEVV1yB5z//+bj11lsxjuNjE8EN7oMf/CCuvPJKPOc5z8H3fu/34rOf/aw8uxDL5NOf/jR+4zd+A69//eubZ490eVxwq7z/v//3/zBNE572tKe5+0972tPwf/7P/3mMYnU0l1LCv/pX/wr/4B/8A3zlV36l3P/n//yf4xnPeAauueYa/Omf/ine+ta34s4778Qv//IvP4ax9e7666/Hz/7sz+I5z3kO7rnnHrz97W/HP/yH/xB/9md/hnvvvRf7+/uNUHna056Ge++997GJ8A7uve99L+677z58x3d8h9y7EMqidpzHvbbBz+69915ceeWV7vliscDll1/+uC6jw8NDvPWtb8WrX/1qt6r493//9+MFL3gBLr/8cnz4wx/GTTfdhHvuuQfvfOc7H8PYenfDDTfgla98Ja699lp84hOfwA/+4A/ixhtvxG233YZhGC7IMnn3u9+NSy65pDH9PxrlccGB1klwb3zjG/Fnf/Znri8IgLNhP/e5z8XVV1+Nl770pfjEJz6BZz/72Y92NLvuxhtvlN/Pe97zcP311+MZz3gG/sf/+B+4+OKLH8OYHd/91E/9FG688UZcc801cu9CKIvPF7der/HN3/zNICK8613vcs/e8pa3yO/nPe952N/fx3d/93fj5ptvftys8fet3/qt8vu5z30unve85+HZz342PvjBD+KlL33pYxiz47uf/umfxmte8xpcdNFF7v6jUR4XnHnwqU99KoZhaEakffrTn8ZVV131GMVqd/emN70Jv/7rv44PfOAD+KIv+qKNfq+//noAwF133fVoRO1Y7rLLLsPf+Tt/B3fddReuuuoqrFYr3Hfffc7P47lsPvnJT+J3fud38C/+xb/Y6O9CKAvO401t46qrrmoGLI3jiM997nOPyzJiwPrkJz+J973vfVv3brr++usxjiP+8i//8tGJ4DHcs571LDz1qU+VunShlcnv//7v484779zaZoBHpjwuONDa39/HC1/4Qvzu7/6u3Esp4Xd/93fxohe96DGM2WZHRHjTm96EX/mVX8H73/9+XHvttVvfueOOOwAAV1999SMcu+O7hx56CJ/4xCdw9dVX44UvfCH29vZc2dx555341Kc+9bgtm5/5mZ/BlVdeiX/yT/7JRn8XQllce+21uOqqq1z+P/DAA7j99tsl/1/0ohfhvvvuw0c/+lHx8/73vx8pJQHmx4tjwPr4xz+O3/md38EVV1yx9Z077rgDMcbG3PZ4cn/913+Nz372s1KXLqQyAbJl4oUvfCGuu+66rX4fkfJ4RId5PELuPe95Dx0cHNDP/uzP0p//+Z/Td33Xd9Fll11G995772MdtVn3vd/7vXTppZfSBz/4QbrnnnvkOHPmDBER3XXXXfSOd7yDPvKRj9Ddd99Nv/qrv0rPetaz6Gu+5mse45h796//9b+mD37wg3T33XfT//pf/4te9rKX0VOf+lT6zGc+Q0RE3/M930NPf/rT6f3vfz995CMfoRe96EX0ohe96DGOdd9N00RPf/rT6a1vfau7/3guiwcffJD+5E/+hP7kT/6EANA73/lO+pM/+RMZUXfLLbfQZZddRr/6q79Kf/qnf0rf8A3fQNdeey2dPXtWwrjhhhvo+c9/Pt1+++30B3/wB/SlX/ql9OpXv/pxlZbVakVf//VfT1/0RV9Ed9xxh2szy+WSiIg+/OEP04/+6I/SHXfcQZ/4xCfo537u5+gLvuAL6Nu//dsfN+l48MEH6d/8m39Dt912G9199930O7/zO/SCF7yAvvRLv5QODw8ljMdDmWyrW0RE999/Pz3hCU+gd73rXc37j1Z5XJCgRUT04z/+4/T0pz+d9vf36au/+qvpD//wDx/rKG10ALrHz/zMzxAR0ac+9Sn6mq/5Grr88svp4OCAvuRLvoT+7b/9t3T//fc/thGv3Ld8y7fQ1VdfTfv7+/SFX/iF9C3f8i101113yfOzZ8/Sv/yX/5Ke8pSn0BOe8AT6pm/6JrrnnnsewxjPu9/+7d8mAHTnnXe6+4/nsvjABz7QrUevfe1riSgPe/+hH/ohetrTnkYHBwf00pe+tEnfZz/7WXr1q19NT3rSk+jJT34yfed3fic9+OCDj6u03H333bNt5gMf+AAREX30ox+l66+/ni699FK66KKL6Mu//Mvp//v//j8HBo91Os6cOUMvf/nL6Qu+4Atob2+PnvGMZ9Ab3vCGRsF+PJTJtrpFRPTf/tt/o4svvpjuu+++5v1HqzxO99M6dafu1J26U3fBuAuuT+vUnbpTd+pO3eevOwWtU3fqTt2pO3UXjDsFrVN36k7dqTt1F4w7Ba1Td+pO3ak7dReMOwWtU3fqTt2pO3UXjDsFrVN36k7dqTt1F4w7Ba1Td+pO3ak7dReMOwWtU3fqTt2pO3UXjDsFrVN36k7dqTt1F4w7Ba1Td+pO3ak7dReMOwWtU3fqTt2pO3UXjPv/AaWGSp4VtHxaAAAAAElFTkSuQmCC",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "show_image('data/asl_images/b.png')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "ORPUII6qCsux"
+   },
+   "source": [
+    "### 4b.3.2 Scaling the Images"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "eYcawjxlCsux"
+   },
+   "source": [
+    "The images in our dataset were 28x28 pixels and grayscale. We need to make sure to pass the same size and grayscale images into our method for prediction. There are a few ways to edit images with Python, but TorchVision also has the [read_image](https://pytorch.org/vision/stable/generated/torchvision.io.read_image.html) function. We can let it know what kind of image to read with [ImageReadMode](https://pytorch.org/vision/stable/generated/torchvision.io.ImageReadMode.html#torchvision.io.ImageReadMode)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 158,
+     "status": "ok",
+     "timestamp": 1715244417167,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "lDJu6JrEffRP",
+    "outputId": "927907d8-3c33-4a3c-bd9f-3b1844a40b2d"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([[[247, 223, 178,  ..., 223, 231, 238],\n",
+       "         [245, 221, 178,  ..., 213, 223, 232],\n",
+       "         [244, 220, 177,  ..., 204, 217, 228],\n",
+       "         ...,\n",
+       "         [247, 243, 238,  ..., 223, 231, 238],\n",
+       "         [249, 246, 243,  ..., 232, 238, 243],\n",
+       "         [250, 248, 246,  ..., 240, 243, 246]]], dtype=torch.uint8)"
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "image = tv_io.read_image('data/asl_images/b.png', tv_io.ImageReadMode.GRAY)\n",
+    "image"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Let's look at the shape of the image."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 4,
+     "status": "ok",
+     "timestamp": 1715244124709,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "8iJ4A8YagRvP",
+    "outputId": "858dc8bc-8595-4189-8738-954ac6630d0c"
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "torch.Size([1, 184, 186])"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "image.shape"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "This image is much larger than what we trained on. We can use [TorchVision](https://pytorch.org/vision/stable/index.html)'s [Transforms](https://pytorch.org/vision/0.9/transforms.html) again to get the data in the form our model expects.\n",
+    "\n",
+    "We will:\n",
+    "* Convert the image to float with [ToDtype](https://pytorch.org/vision/stable/generated/torchvision.transforms.v2.ToDtype.html)\n",
+    "  * We will set `scale` to `True` in order to convert from [0, 255] to [0, 1]\n",
+    "* [Resize](https://pytorch.org/vision/stable/generated/torchvision.transforms.v2.Resize.html#torchvision.transforms.v2.Resize) the image to be 28 x 28 pixels\n",
+    "* Convert the images to [Grayscale](https://pytorch.org/vision/stable/generated/torchvision.transforms.v2.Grayscale.html#torchvision.transforms.v2.Grayscale)\n",
+    "  * This step doesn't do anything since our models are already grayscale, but we've added it here to show an alternative way to get grayscale images."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 337,
+     "status": "ok",
+     "timestamp": 1715244596376,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "lNp6Q9cBe8Xw"
+   },
+   "outputs": [],
+   "source": [
+    "IMG_WIDTH = 28\n",
+    "IMG_HEIGHT = 28\n",
+    "\n",
+    "preprocess_trans = transforms.Compose([\n",
+    "    transforms.ToDtype(torch.float32, scale=True), # Converts [0, 255] to [0, 1]\n",
+    "    transforms.Resize((IMG_WIDTH, IMG_HEIGHT)),\n",
+    "    transforms.Grayscale()  # From Color to Gray\n",
+    "])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Let's test `preprocess_trans` on an image to make sure it works correctly:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 164,
+     "status": "ok",
+     "timestamp": 1715244649391,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "BBS7SjnBhIe6",
+    "outputId": "fadd047d-3763-4b05-ea2e-ebe0427bf30d"
+   },
+   "outputs": [],
+   "source": [
+    "processed_image = preprocess_trans(image)\n",
+    "processed_image"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "The numbers look correct, but how about the shape?"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "processed_image.shape"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Next, let's plot the image to see if it looks like what we trained on."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "plot_image = F.to_pil_image(processed_image)\n",
+    "plt.imshow(plot_image, cmap='gray')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Looking good! Let's pass it to our model."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### 4b.4 Making Predictions"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "vYl2JcUJCsuy"
+   },
+   "source": [
+    "Okay, now we're ready to predict! Our model still expects a batch of images. If the [squeeze](https://pytorch.org/docs/stable/generated/torch.squeeze.htmlhttps://pytorch.org/docs/stable/generated/torch.squeeze.html) removes dimensions of 1, [unsqueeze](https://pytorch.org/docs/stable/generated/torch.unsqueeze.htmlhttps://pytorch.org/docs/stable/generated/torch.unsqueeze.html) adds a dimension of 1 at the index we specify. The first dimension is usually the batch dimension, so we can say `.unsqueeze(0)`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 168,
+     "status": "ok",
+     "timestamp": 1715244907170,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "L-xZGe1SCsux",
+    "outputId": "12b26af6-e1ce-4989-864b-9e26419c4f84"
+   },
+   "outputs": [],
+   "source": [
+    "batched_image = processed_image.unsqueeze(0)\n",
+    "batched_image.shape"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Next, we should make sure the input tensor is on the same `device` as the model."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "batched_image_gpu = batched_image.to(device)\n",
+    "batched_image_gpu.device"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now we're ready to pass it to the model!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "executionInfo": {
+     "elapsed": 176,
+     "status": "ok",
+     "timestamp": 1715245336961,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "WhLIaQIhkE0c",
+    "outputId": "f3fe15e3-f524-4218-c278-31e740299237"
+   },
+   "outputs": [],
+   "source": [
+    "output = model(batched_image_gpu)\n",
+    "output"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "CCleBaVCCsuz"
+   },
+   "source": [
+    "### 4b.4.1 Understanding the Prediction"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "QNIv09wxCsuz"
+   },
+   "source": [
+    "The predictions are in the format of a 24 length array. The larger the value, the more likely the input image belongs to the corresponding class. Let's make it a little more readable. We can start by finding which element of the array represents the highest probability. This can be done easily with the numpy library and the [argmax](https://numpy.org/doc/stable/reference/generated/numpy.argmax.html) function."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 171,
+     "status": "ok",
+     "timestamp": 1715245408487,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "gm6J4jxcCsuz"
+   },
+   "outputs": [],
+   "source": [
+    "prediction = output.argmax(dim=1).item()\n",
+    "prediction"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "BpZzxObRCsuz"
+   },
+   "source": [
+    "Each element of the prediction array represents a possible letter in the sign language alphabet. Remember that j and z are not options because they involve moving the hand, and we're only dealing with still photos. Let's create a mapping between the index of the predictions array, and the corresponding letter."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 139,
+     "status": "ok",
+     "timestamp": 1715245412136,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "1ExBb6fMCsuz"
+   },
+   "outputs": [],
+   "source": [
+    "# Alphabet does not contain j or z because they require movement\n",
+    "alphabet = \"abcdefghiklmnopqrstuvwxy\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "uXy2Ze9kCsuz"
+   },
+   "source": [
+    "We can now pass in our prediction index to find the corresponding letter."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/",
+     "height": 36
+    },
+    "executionInfo": {
+     "elapsed": 140,
+     "status": "ok",
+     "timestamp": 1715245413856,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "VNxyHNZmCsuz",
+    "outputId": "87063db2-9c3d-45ed-f32f-6e5700fde9c8"
+   },
+   "outputs": [],
+   "source": [
+    "alphabet[prediction]"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Hs2_6guPCsuz"
+   },
+   "source": [
+    "#### Exercise: Put it all Together"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "KobsgLirCsuz"
+   },
+   "source": [
+    "Let's put everything in a function so that we can make predictions just from the image file. Implement it in the function below using the functions and steps above. If you need help, you can reveal the solution by clicking the three dots below."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "f3prZEMgCsuz"
+   },
+   "outputs": [],
+   "source": [
+    "def predict_letter(file_path):\n",
+    "    # Show image\n",
+    "    FIXME\n",
+    "    # Load and grayscale image\n",
+    "    image = FIXME\n",
+    "    # Transform image\n",
+    "    image = FIXME\n",
+    "    # Batch image\n",
+    "    image = FIXME\n",
+    "    # Send image to correct device\n",
+    "    image = FIXME\n",
+    "    # Make prediction\n",
+    "    output = FIXME\n",
+    "    # Find max index\n",
+    "    prediction = FIXME\n",
+    "    # Convert prediction to letter\n",
+    "    predicted_letter = FIXME\n",
+    "    # Return prediction\n",
+    "    return predicted_letter"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "enmPt4goCsu0"
+   },
+   "source": [
+    "#### Solution"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Zfp8Cxm9Csu0"
+   },
+   "source": [
+    "Click on the '...' below to view the solution."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "executionInfo": {
+     "elapsed": 146,
+     "status": "ok",
+     "timestamp": 1715245521266,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "wTUx5fWMCsu_",
+    "jupyter": {
+     "source_hidden": true
+    },
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "# SOLUTION\n",
+    "def predict_letter(file_path):\n",
+    "    show_image(file_path)\n",
+    "    image = tv_io.read_image(file_path, tv_io.ImageReadMode.GRAY)\n",
+    "    image = preprocess_trans(image)\n",
+    "    image = image.unsqueeze(0)\n",
+    "    image = image.to(device)\n",
+    "    output = model(image)\n",
+    "    prediction = output.argmax(dim=1).item()\n",
+    "    # convert prediction to letter\n",
+    "    predicted_letter = alphabet[prediction]\n",
+    "    return predicted_letter"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/",
+     "height": 453
+    },
+    "executionInfo": {
+     "elapsed": 648,
+     "status": "ok",
+     "timestamp": 1715245523822,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "MWqiMAhwCsu_",
+    "outputId": "32476dc7-79ef-4280-ba34-725a46837aae"
+   },
+   "outputs": [],
+   "source": [
+    "predict_letter(\"data/asl_images/b.png\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "JNZSUoinCsvA"
+   },
+   "source": [
+    "Let's also use the function with the 'a' letter in the asl_images datset:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/",
+     "height": 453
+    },
+    "executionInfo": {
+     "elapsed": 754,
+     "status": "ok",
+     "timestamp": 1715245527515,
+     "user": {
+      "displayName": "Danielle Detering US",
+      "userId": "15432464718872067879"
+     },
+     "user_tz": 420
+    },
+    "id": "xk0F5w53CsvA",
+    "outputId": "1cfe6713-59ec-4ce4-d4de-22d4cf464783"
+   },
+   "outputs": [],
+   "source": [
+    "predict_letter(\"data/asl_images/a.png\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "cFaxufFbCsvA"
+   },
+   "source": [
+    "## 4b.5 Summary"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "ZtsAm4TJCsvA"
+   },
+   "source": [
+    "Great work on these exercises! You've gone through the full process of training a highly accurate model from scratch, and then using the model to make new and valuable predictions. If you have some time, we encourage you to take pictures with your webcam, upload them by dropping them into the data/asl_images folder, and test out the model on them. For Mac you can use Photo Booth. For windows you can select the Camera app from your start screen. We hope you try it. It's a good opportunity to learn some sign language! For instance, try out the letters of your name.\n",
+    "\n",
+    "We can imagine how this model could be used in an application to teach someone sign language, or even help someone who cannot speak interact with a computer."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "zyfxj0v0CsvA"
+   },
+   "source": [
+    "### 4b.5.1 Clear the Memory"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "JtO6ds0LCsvA"
+   },
+   "source": [
+    "Before moving on, please execute the following cell to clear up the GPU memory."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "mN85KA47CsvA"
+   },
+   "outputs": [],
+   "source": [
+    "import IPython\n",
+    "app = IPython.Application.instance()\n",
+    "app.kernel.do_shutdown(True)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "jKNrUtH1CsvA"
+   },
+   "source": [
+    "### 4b.5.2 Next"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "k1VRSuYLCsvA"
+   },
+   "source": [
+    "We hope you've enjoyed these exercises. In the next sections we will learn how to take advantage of deep learning when we don't have a robust dataset available. See you there!\n",
+    "To learn more about inference on the edge, check out [this paper](http://web.eecs.umich.edu/~mosharaf/Readings/FB-ML-Edge.pdf) on the topic."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "QnOZX0ClCsvA"
+   },
+   "source": [
+    "Now that we're familiar building your own models and have some understanding of how they work, we will turn our attention to the very powerful technique of using pre-trained models to expedite your work."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "4bnILHG8CsvA"
+   },
+   "source": [
+    "<center><a href=\"https://www.nvidia.com/dli\"> <img src=\"images/DLI_Header.png\" alt=\"Header\" style=\"width: 400px;\"/> </a></center>"
+   ]
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}