promptAnalysis/promptAnalysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "with open('promptReduceRandomOutputs.json', 'r') as file:\n",
    "    data = json.load(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ast\n",
    "a = 1600\n",
    "b = 1800\n",
    "gt = []\n",
    "for i in range(200):\n",
    "    gt.append(data[i]['testGT'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred = []\n",
    "for i in range(a,b):\n",
    "    pred.append(data[i]['output2'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 344,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Hit(A, test_set):\n",
    "    dcg = 0\n",
    "    r_i = 0\n",
    "    for i in range(len(A)):\n",
    "        \n",
    "        if A[i] in test_set:\n",
    "            r_i = 1\n",
    "            return r_i\n",
    "    return r_i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import numpy as np\n",
    "from sklearn.metrics import ndcg_score\n",
    "\n",
    "def compute_precision_recall(targets, predictions, k):  \n",
    "    num_hit = [pred for pred in predictions if any(pred[:int(1*len(pred))] == tar[:int(1*len(tar))] for tar in targets)]\n",
    "    num_hit = len(num_hit)\n",
    "    precision = float(num_hit) / len(predictions)    \n",
    "    recall = float(num_hit) / len(targets)    \n",
    "    return precision, recall\n",
    "\n",
    "\n",
    "def NDCG(targets, predicitions):\n",
    "    test_movies = list(set(movie for sublist in predicitions for movie in sublist)) # unique items\n",
    "    y_score = np.zeros((len(predicitions), len(test_movies))) # convert to score\n",
    "    for i, rec_list in enumerate(predicitions):\n",
    "        for j, item in enumerate(test_movies):\n",
    "            if item in rec_list:\n",
    "                y_score[i, j] = len(test_movies) - rec_list.index(item) # get a list of scores, match and early item have higher score\n",
    "\n",
    "    # Convert test_sets to y_true\n",
    "    y_true = np.zeros((len(targets), len(test_movies)))\n",
    "    for i, test_list in enumerate(targets):\n",
    "        for j, item in enumerate(test_movies):\n",
    "            if item in test_list:\n",
    "                y_true[i, j] = 1 # if hit 1 if not 0\n",
    "    return ndcg_score(y_true, y_score, k = 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "def texta(text):\n",
    "    recommended_movies = []\n",
    "    for line in text.split('\\n'):\n",
    "        if line.startswith((\"1. \", \"2. \", \"3. \", \"4. \", \"5. \")):\n",
    "            movie = re.search(r'\\\"(.*?)\\\"', line)\n",
    "            if movie:\n",
    "                recommended_movies.append(movie.group(1))\n",
    "    return recommended_movies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "160\n",
      "166\n",
      "172\n"
     ]
    }
   ],
   "source": [
    "import re\n",
    "pt_movies = []\n",
    "for a in range(200):\n",
    "    text = pred[a]\n",
    "    pattern =  r'\\d+\\. (.+?) -'\n",
    "    recommended_movies = re.findall(pattern, text)\n",
    "    if len(list(set(recommended_movies))) != 5:\n",
    "        recommended_movies = re.findall(r'\\d+\\.\\s(.*?)\\s\\(\\d+\\)', text)\n",
    "        if len(list(set(recommended_movies))) != 5:\n",
    "            recommended_movies = re.findall(r'\\d+\\.\\s(.*?)\\s\\(Liked\\)', text)\n",
    "            if len(list(set(recommended_movies))) != 5:\n",
    "                recommended_movies = re.findall(r'\\\"(.*?)\\\"', text)\n",
    "                if len(list(set(recommended_movies))) != 5:\n",
    "                    recommended_movies = texta(text)\n",
    "                    if len(recommended_movies) != 5:\n",
    "                        recommended_movies = re.findall(r'\\\"(.*?)\\\"', text)\n",
    "                        if len(recommended_movies) != 5:\n",
    "                            recommended_movies = re.findall(r'^\\d\\.\\s(.*?)\\s-', text, re.MULTILINE)\n",
    "                            recommended_movies = recommended_movies[:5]\n",
    "                            if len(recommended_movies) != 5:\n",
    "                                recommended_movies = re.findall(r'\\d\\.\\s(.*?)\\n', text)\n",
    "                                if len(recommended_movies) != 5:\n",
    "                                    recommended_movies = re.findall(r'\\d\\.\\s(.*?)(?:\\n|$)', text)\n",
    "                                    recommended_movies[-1] = recommended_movies[-1].replace('</s>', '')\n",
    "                                    if len(recommended_movies) != 5:\n",
    "                                        print(a)\n",
    "    titles = [title.replace('\"', '') for title in recommended_movies]\n",
    "    pt_movies.append(list(set(titles)))  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = [] \n",
    "for element in pt_movies:\n",
    "    movies_no_year = [re.sub(r' \\(\\d{4}\\)', '', movie) for movie in element]\n",
    "    movies_no_score = [re.sub(r' \\(\\d\\.\\d\\/10\\)', '', movie) for movie in movies_no_year]\n",
    "    movies_no_director = [re.sub(r' \\([^)]*\\)', '', movie) for movie in movies_no_score]\n",
    "    movie_names = [desc.split(\":\")[0] for desc in movies_no_director]\n",
    "    m.append(movie_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 349,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The NDCG@5 loss is computed to be: 0.35695670831224097\n",
      "The pre@5 loss is computed to be: 0.3256666666666671\n",
      "The recall@5 loss is computed to be: 0.17537658772664244\n"
     ]
    }
   ],
   "source": [
    "pre = 0.0\n",
    "recall = 0.0\n",
    "for i in range(200):\n",
    "    titles = m[i]\n",
    "    if titles == []:\n",
    "        continue\n",
    "    temp_pre, temp_recall = compute_precision_recall(gt[i],titles,1)\n",
    "    pre = pre + temp_pre\n",
    "    recall = recall + temp_recall\n",
    "print('The NDCG@5 loss is computed to be: {}'.format(NDCG(gt,m)))\n",
    "print('The pre@5 loss is computed to be: {}'.format(pre /200))\n",
    "print('The recall@5 loss is computed to be: {}'.format(recall/200))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [],
   "source": [
    "# These are coming from above by change a,b +200 each interval\n",
    "ndcg = [0.24909222678655435,0.28053875981583526,0.3051968072313468,0.3233537967415029,0.3288377726906373,0.3299680933939095,0.34378503720035,0.3551788651037631,0.35695670831224097,0.3708] \n",
    "pre2 = [0.21391666666666695,0.2325000000000004,0.2656666666666671,0.2767500000000004,0.2855000000000004,0.29500000000000054,0.3163333333333338, 0.3145000000000004,0.3256666666666671,0.3337]\n",
    "recall2 = [0.15232117686263666,0.15053373530690384,0.16341746937181156,0.16709813770992027,0.167318363001337,0.1738563571662843,0.17232314708609864,0.17773912807050596, 0.17537658772664244,0.1770]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkUAAAG1CAYAAAD3BIBFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAACmq0lEQVR4nOzdd1hTZ/vA8W8IGxkCshQV994Lt9ZtnW21w1lba187rG+HdrhqHa3tqx12W2u11p/V1q111I1aUbRuBRRlKsjeyfn9EQlGhhCBBLg/18VFcs6Tk/somptn3I9KURQFIYQQQohKzsLUAQghhBBCmANJioQQQgghkKRICCGEEAKQpEgIIYQQApCkSAghhBACkKRICCGEEAKQpEgIIYQQAgBLUwdQXmi1WiIiInB0dESlUpk6HCGEEEIUgaIoJCUl4ePjg4VF4X1BkhQVUUREBL6+vqYOQwghhBBGuHnzJjVq1Ci0jSRFReTo6Ajo/lCdnJxMHI0QQgghiiIxMRFfX1/953hhJCkqopwhMycnJ0mKhBBCiHKmKFNfZKK1EEIIIQSSFAkhhBBCAJIUCSGEEEIAkhQJIYQQQgAy0bpUaDQasrKyTB2GKMcsLS1Rq9VSE0sIIcqQJEUlSFEUoqKiiI+PN3UoogJQq9V4eHjg7OwsyZEQQpQBSYpKUE5C5OHhgb29vXyQCaMoikJ2djaJiYlERkaSlpaGt7e3qcMSQogKT5KiEqLRaPQJkZubm6nDERWAo6MjNjY23LlzBw8PD9RqtalDEkKICk0mWpeQnDlE9vb2Jo5EVCQODg4oiiJz1IQQogxIT1EJkyEzUZLk50kIURlotAonQuOISUrHw9GWDn6uqC3K/v8/SYqEEEIIYTI7z0Uyd8sFIhPS9ce8nW2ZPaQJA5qV7XxKGT4TQgghhEnsPBfJy6tPGSREAFEJ6by8+hQ7z0WWaTySFIk85syZg0qlonv37vmeq1Kliv65SqXSf9nY2ODj48OAAQP48ccfC5wHc+HCBcaNG0eNGjWwtramWrVqjBgxgoMHD+Zpu2PHDgYNGkS1atWwsrLC09OTIUOGsG3bNhRFKZH7vX79usF95Hx16tSpRK4vhBAiL41WYe6WC+T3P3nOsblbLqDRlsz/9UUhSZGZ0mgVAoJj2RQUTkBwbJn+UOQ4dOgQ+/bte2i7V199lYCAAPbu3cvSpUvx8fFhypQpdOvWjaSkJIO2W7dupW3btvz777/MmzePPXv28N133+Hs7Ezv3r25ceOGvu27777LoEGDsLW15csvv2Tv3r18+eWXODk5MXToULZv355vPOfOnWPy5MnUrl0bKysr3N3d6du3L//3f/9X6H0sWLCAgIAA/dePP/5YhD8lIYQQxjgRGpenh+h+ChCZkM6J0Lgyi0nmFJkhcxhfdXBwoFmzZsydO5fevXsX2rZmzZoGvSqjRo1i1KhRDB48mOnTp/P9998DEB0dzZgxY+jUqRM7d+7ExsZG/5oRI0YwZcoUHBwcANi2bRsLFy5k9uzZzJkzx+D9nnrqKaZNm4aFRd6cfu7cucyfP58hQ4awaNEiatasyd27d9m/fz9Tpkzh559/Zt26dQa9XTnq168vvUNCCFEGktKz+D3wZpHaxiQVnDiVNOkpMjPmNL46a9YsDh48yP79+4v92gEDBvDEE0+watUqfW/R999/T0JCAsuWLTNIiHJ06tQJd3d3AD777DO8vb15//33871++/btadu2rcGxOXPmsGLFCk6cOMHGjRt5+umn6dy5M4MHD+aTTz7hwoULxMfH89RTT6HVaot9T0IIIR7N1egkPvjzHJ0W7GXDqfAivcbD0baUo8olSVEpUhSF1MzsIn8lpWcxe/P5QsdX52y+QFJ6VrGua+zcm0GDBtG+ffs8PTVF1a9fPzIzMzl16hQA+/fvp3r16rRo0aLQ12VnZ3PkyBF69+6NpWXROjNPnz7N0qVL2b9/P61bt85zXlEUPDw82Lp1K//++y8//fRTnjYvv/yyfmuNF198kbi4suuyFUKIiipbo2XnuSie/f4Yff93kF+O3SAlU0O9ag442VpS0MJ7FbpRkg5+rmUWqwyflaK0LA1NZu0qsespQFRiOs3n/FWs112Y1x97a+P+qmfNmsWQIUM4cOAAPXr0KNZrfX19Ad32JwDh4eH6Y4WJjY0lIyMjT1tFUdBoNPrnFhYW+iG0Tz75hBkzZuDn5wfA9u3beeONN7h+/Tp9+vShS5cuXLlyhZUrV/LOO++wfPlyJk2aBICNjQ0vv/wy/fv3x8XFhePHj/PRRx9x8uRJTpw4gZWVVbHuWwghBNxJzmDdPzdZc+wGEfdGPyxU0K+JF+M618K/jhu7zkfx8upTqMCgQyAnUZo9pEmZ1iuSniJRqMcff5w2bdowd+7cYr82p4cqpwChoihFKkb44OtybNiwASsrK/3Xa6+9pm+/Y8cOxo4dC8CNGzcYMWIEPXv2ZOfOnbRq1Yp58+bprzNgwACCgoJITU0FwNvbm+XLlzNs2DB69OjB22+/za+//kpQUBB//PFHse9bCCEqs6Cb8UxfF0Tnhfv4ZNdlIhLScXWwZmqvuhx6pzffjG1L57ruqFQqBjTz5usxbfByNhwi83K25esxbcq8TpH0FJUiOys1F+b1L3L7E6FxTPjpn4e2WzmxfbG6E+2sHm3PrFmzZjF8+HAOHTpUrNfdunULAC8vLwBq1KjBpUuXHvo6d3d3bGxs9K/P8dhjj/HPP7o/n6FDh+qPx8XFoSgK1atXB+DXX3+lXr16fPPNN6hUKnr16kVgYKC+va+vL1qtlrt37xa4LcugQYOoUqUKgYGBjBo1qhh3LYQQlU96loZtZyNZFXCdM7cS9Mdb+row3r8Wg5p7Y1vAZ9GAZt70beIlFa0rOpVKVaxhq271q+HtbEtUQnq+84pU6LLnbvWrlekPy7Bhw2jVqhVz586la9euRX7drl27sLGx0U+I7tWrF3v37uXff/+lefPmBb7O0tKSLl26sHfvXjQajX4j1KpVq9KuXTsArK2t9e01Go3BxO2wsDCaN29u0NPUsmVLoqOjAd1wnoWFBVWrVi00/pKqgySEEBXVrbuprDkexrp/bhKXkgmAtdqCx1t6M86/Nq18XYp0HbWFCv+6pt9MXYbPzIjaQsXsIU0A8kw8M9X4ao5Zs2axd+9eDh8+XKT2O3fuZOPGjUyYMEG/zP6FF17A2dmZadOmkZmZmec1x48f586dOwBMnz6diIgIFixY8ND3cnNzIzk5mcTERAA8PT0JCwszaHPzZu7Sz3Xr1tGxY8dCN+/dunUrKSkptG/f/uE3K4QQlYiiKBy5dofJq07S/eO/+Xp/MHEpmfg42/L2gIYEzOzNZ6NaFTkhMifSU2RmcsZXH6xT5GWifWByDB8+nBYtWrB37159kpMjLCyMY8eOodFoiIyMZMeOHaxatYqOHTuyZMkSfTtPT09Wr17Nk08+SefOnZk6dSp169YlLi6OrVu38vPPP3Pt2jXc3d0ZPHgwM2bMYNasWQQFBTF69Gi8vb1JSEjg0KFDREVF4ejoCIBaraZz585s2bKF5557jqFDhzJv3jy+/fZbRo8ezd9//82GDRsYMWIEK1asYN68eWzdulUf15tvvomFhQUdO3bExcWFEydOsHDhQtq1a8fw4cPL5M9XCCHMXXJGNhtP3eLno9cJvp2iP96lnhvj/GvzWCMPLNXlvK9FEUWSkJCgAEpCQkK+59PS0pQLFy4oaWlpJfJ+2RqtcvTaHeXP07eUo9fuKNkabYlctyhmz56tODg45Dn++++/K4DBOXQLBhRAsbKyUry8vJT+/fsrP/74o5KVlZXv9c+dO6eMGTNG8fHxUSwtLRV3d3dlyJAhyo4dO/K03bp1qzJw4EDFzc1NsbS0VDw8PJSBAwcqa9asUTQajb7dhg0blHr16ilJSUmKoijKxx9/rKjVagVQateurUybNk0BlPr16yvbtm0zeI8ffvhBadOmjeLk5KRYWloqtWrVUqZNm1bg33VZKumfKyGEKK6r0UnKrD//VZrO2qnUemerUuudrUqTD3YoH/z5r3IlKtHU4T3Uwz6/76dSFJk4URSJiYk4OzuTkJCAk5NTnvPp6emEhobi5+eHrW3ZFZoSOoqiMGzYMDIzM/n999+pUqUKSUlJ3Lx5kwYNGpCYmEhycjI1a9Y0dajFIj9XQghTyNZo2XsphlUB1zlyLVZ/vE41B8b712Zkm+o42paPciUP+/y+nwyfiQpBpVKxZs0annjiCZo1a8Y777zDgAED8PPz486dOxw9epSvv/6atm3bsmjRIlOHK4QQZikuJZPf/gljzbEwwuPTAF1toT6NPRnfuTad67oVqbRKeSVJkagwHB0d2bFjB9999x2ff/45//nPf/TnatSoweTJk3nzzTdNGKEQQpins7fi+fnoDbacjSAzW7cNUlV7K57uUJPnOtakRtWCF6ZUJJIUiQpFrVbz8ssv8/LLLxMdHU10dDRVq1YtUiVtIYSoTDKyNWz/N5Kfj94g6Ga8/njz6s6M71ybx1sUXFuoopKkSFRYnp6eeHp6mjoMIYQwKxHxaaw5foPfTtwk9r7aQoNbeDPOvxatfF0q9BBZYSQpEkIIISo4RVEICIll1dEb/HUhCu29JVZeTraM6VST0e1rUs3RpvCLVAJGJ0XZ2dl88cUXrF27lkuXLpGamkp2djYAQUFBfPfdd0ybNo0GDRqUWLBCCCGEyKXRKoVuj5GSkc3G0+GsOnqdqzHJ+uP+ddwY51+Lvk08y39toRJkVFKUlpZGv379OHr0KO7u7jg5OZGSklvIyc/Pj59++glXV1fmz59fYsEKIYQQQmfnucg8hX697xX6re/pyC8BN9gQeIukDF2Hhb21mpFtqjPOvzYNPB1NFbZZMyo9XLBgAUeOHGHhwoVERUXxwgsvGJx3dnamR48e7Nq1q0SCFEIIIUSunecieXn1KYOECCAyIZ0pq0/x2KcHWHn0OkkZ2fi5OzB7SBOOvfsY84c3l4SoEEb1FK1bt46ePXvy9ttvA+Q7IatOnTqcPn360aITQgghSsHDhp3MmUarMHfLhXw3Dr/fY42qMb6zH13ruWNRTu7N1IxKisLCwhgxYkShbZycnEhISDAqKCGEEKK0FDbsVJb7S2Zka0jJ0JCcnk1SRhbJ6dkkZ+i+ktKzSbnvcXJGtv58ZEJanh6i/LzQra5Z7DxfnhiVFDk6OnL79u1C2wQHB1OtWjWjgqrU4m9CamzB5+3dwKXi19zp2bMnVapUMdi4tTArV65k4sSJ3L59G3d391KOrnjmzJnD3Llz8xz/+uuvmTJligkiEqLyyhl2erCXJSohnZdXn+LrMW0KTYy0WoXULM29BCXrXvKi0T++P3lJyriX2KTrHt+f9CSnZ5Op0ZbqvcYkPTxxEoaMSoo6derEli1bSEhIwNnZOc/5W7dusX37dtlhvLjib8KXbSE7o+A2ljbwSmCFT4yWL1+OWl30omGDBw8mICAAFxeX0gsK3bLWtWvX8vPPP3Py5EkSExOpXr06AwYM4PXXX6dx48b5vs7Ozo59+/YZHKtTp06pxiqEMFTYsFPOsf/+3xn2XIgmJVNj0EuTk9wkZ2ZT0juG2lurqWJjSRVbS933nC9bSxz1x63ufVcTfjeNJX9deeh1PRxlv8TiMiopeuutt+jVqxd9+vRh2bJl+qX4qampBAQE8Oqrr5KVlcX06dNLNNgKLzW28IQIdOdTY80uKcrMzMTS0hILi5JZ2tmkSZNita9WrVqp90xGR0czcuRILl26xEsvvcT06dOpUqUK169fZ/369bRu3ZrPPvvMYHuRHBYWFnTq1KlU4xNCFO7ItdsPHXZKydTw+6nwh15LbaHSJy+OtpYFJjW5560Mz9977GCtLvaSeI1WYc3xMKIS0vNN8FSAl7NunpQoHqOSou7du/PVV1/x2muv0a1bN/1xR0fdjHa1Ws3y5ctp27ZtyURZXikKZKUWvX12WtHbZaY8vF0OK3soRnXSCRMmcPLkST755BPeeustrl27RtOmTfnqq6/0H+y1a9fm8ccfp1atWnz55ZfcunWLqKgoqlWrxsqVK/nss8+4cuUKbm5uTJgwgblz52JpmfvjFh4ezsyZM9m1axeJiYnUqlWLl19+mddffx3IO3x269Ytpk+fzoEDB0hISMDb25vhw4fzv//9D8h/+CwuLo63336bTZs2kZycTPPmzZk/fz79+vXTx5HzPuPHj+f9998nPDycDh068P3331O3bl19u+TkZB577DFatGjBrl27qFKliv5cly5deO6559i5cydPPvkkzs7OPPfcc0X/+xFClBqNViEgOJbNZ8LZciaiSK95vIU3HfxcH+ixMUxqbK0sTFb1WW2hYvaQJry8+hQqMEiMciKaPaRJuZk4bk6MLt44ZcoUevTowTfffMPx48eJi4vDycmJjh078p///IemTZuWZJzlU1YqLPAp+euuGFC89u9GgLVDsV4SGRnJf/7zH+bMmUPVqlVZtGgR/fv35+rVq3h4eACwYcMGGjRowLJly1Cr1djb2/PZZ5/x9ttv88Ybb/Dpp59y8eJF3nvvPTQajX53+tjYWPz9/QH46KOPqFOnDlevXiU4OLjAeMaNG0dERASff/45np6ehIWFcfLkyQLbazQaBg4cyLVr11i4cCE1atTg66+/ZtCgQezevZtevXrp2wYFBXH79m0WLVqERqNh2rRpjBkzhoCAAH2bDz74gDp16rBmzZp8/yPUaDT079+fFStWMHnyZAYOHIira+5vaWlpaVSrVo27d+/SoEED3njjDV588cUi/m0IIYpDURROhcWz5UwEW89Gcif5IT3wD3iuYy2zn6A8oJk3X49pk2fCuJcJJoxXJEYlRQcPHsTJyYlWrVqxbNmyko5JmIG4uDjWr19P7969AV3voK+vL0uXLmXBggWArqr5jh07sLfX7Z6clJTE7Nmzefvtt/Vt+vbti6WlJW+++SZvvfUWbm5ufPbZZ8TExHDp0iVq164NoH+fgpw4cYKFCxcyevRo/bFx48YV2H7btm2cOHGCbdu2MWjQIAAGDBhAs2bNmDt3rkFSFB8fz+nTp/XDb/Hx8bz44ovcunWLGjVqkJiYyPfff8+lS5dQqVRkZWXx9ttvs2rVKlQqFdOnT2f79u3Mnz+fUaNG8fHHH7NmzRpeffVVAOrVq8fixYtp3bo16enp/Prrr0yePJmEhATefPPNIv+dCCEKpigKFyOT2Hwmgi1nIgiPz+15d7azYlBzLwY39+bN9WeJTqwYw04DmnnTt4lXuS0tYI6MSop69erFlClT+Oqrr0o6norFyl7XS1NUUWeL1gv0/E7walG8OIrJ2dnZIFFxcXGhd+/eHDt2TH+sZ8+e+oQI4OjRoyQnJ/PUU0/p55mBLuFJS0vj3Llz9OjRg71799K7d299QlQUbdq0YcmSJVhaWtK3b1/q1atXaPtDhw7h6OioT4hAN69n1KhRLFiwAI1Go5/I3apVK4P5SDnzmXKSon379uHv70+NGjUAWLJkCT/99BOffvopvr6+zJ0716DXqn///hw5ckSfFI0ZM8YgtsGDB5OZmcn8+fN5/fXXsbKyKvKfgxDCUOidFLaciWDzmQiu3beNhb21mn5NPBnayoeu9aphbambtzNnaMUadlJbqMy+V6s8MWpWrIeHB9bW1iUdi97y5cvx8/PD1taWtm3bcujQoQLbHj58mC5duuDm5oadnR2NGjXSzzPJsXLlSlQqVZ6v9PRSXq6oUumGrYr6ZWlXtOta2hXvukaMe+c3adnDw4PIyEiD5/e7c+cOoEtgrKys9F85K7Ju3rwJ6IbPfHyKN6y4bt06HnvsMd577z3q169Po0aN2LhxY4Ht7969i6enZ57jXl5eZGVlkZyc+5/ngyvWcn62c34+goODDYaDV65cybvvvsukSZPo168f69evJysrS3/e19eX2NhCyioAo0aNIiEhgWvXrhXaTgiRV2RCGt8fDGHIF4fptWQ/n+2+wrWYZKzVFvRv6smXz7Ym8P2+LH26Nb0beeoTIsgddvJyNlyZ5eVs+9Dl+KLiM6qnqH///hw4cABFUUp8otm6deuYNm0ay5cvp0uXLnz77bcMHDiQCxcuULNmzTztHRwceOWVV2jRogUODg4cPnyYl156CQcHByZPnqxv5+TkxOXLlw1ea2sryxULkl8dqpiYGLy9c//DePDvPmcOzcaNG/H1zbs6zs/PDwA3NzciIorRgwZ4e3uzYsUKfvjhBwIDA5k/fz6jR4/m8uXL+S5td3V1JTo6Os/xqKgorKysDCZKP4xGo8HGJnf36LCwMFq0yO2p8/HxMUgio6KiHlorSSnpNb1CVHBxKZls/zeSzWci+Od6nH5ZvNpCRee6bgxt6UP/Zl442T6851WGnURBjN77LDY2lsmTJxMXF1eiAX322WdMmjSJF154gcaNG7N06VJ8fX35+uuv823funVrnnnmGZo2bUrt2rUZM2YM/fv3z9O7pFKp8PLyMvgyO/ZuujpEhbG00bUrZQkJCQZ1dXKed+zYscDXdO7cGXt7e27dukW7du3yfLm56eLu06cP+/btIywsrNhxWVhY0L59e+bPn092dnaBPS1du3YlKSmJnTt36o9ptVrWr19P586di1UDqXr16oSEhOif50z0zpGSksLdu3cN3uP+OUv5WbduHS4uLg8dBhSiMktKz2JD4C3GrzhB+4/28P6f5zgRqkuI2teuyofDmnL83cf4ZVJHnmrnW6SEKEfOsNOwVtXxr+smCZEAjOwpGjNmDC4uLqxYsYLVq1fj5+eHp6dnnp4DlUrF3r17i3zdzMxMAgMDmTFjhsHxfv36cfTo0SJd4/Tp0xw9epT58+cbHE9OTqZWrVpoNBpatWrFhx9+SOvWrQu8TkZGBhkZuSsWEhMTi3wfRnPx1RVmNIOK1q6urkyaNIm5c+fi4uKiXzk2bdq0Al/j7OzMvHnzePvtt7l16xa9evXCwsKCkJAQNm3axIYNG7C3t+eNN95g1apVdO/eXb+qKyQkhCtXrrB48eI8101ISKB///6MHTuWhg0bkpWVxeeff46Liwtt2rTJN5bBgwfToUMHxo4dy4IFC6hRowbffPMNly9fLvZcuB49evDqq6+SnJxMlSpVGDZsGIsWLaJ9+/bUqFGDGTNmkJWVxc2bNxk7diyA/jtAu3btGD9+PA0bNiQtLY01a9awceNGli5dKvOJhHhAepaGvy/FsPlMBPsuxZCRnVv1uamPE0Nb+vB4Sx+quxRxuoEQxWBUUrR//37944yMDC5dusSlS5fytCvu0NqdO3fQaDR55oJ4enoSFRVV6Gtr1KjB7du3yc7OZs6cObzwwgv6c40aNWLlypU0b96cxMREli1bRpcuXThz5gz169fP93oLFy7Md2uGUufiaxaFGb29vVm8eDFvvfWWfk7Nrl278p2nc7///ve/VK9enc8++4wvvvgCKysr6taty+OPP66fq+Pm5saRI0eYOXMmb7/9NqmpqdSuXTvfooegG+Zs3rw5X3zxBWFhYdjZ2dGuXTv++uuvAoep1Go1O3bs4K233mLmzJkkJyfTokULtm3bRs+ePYv1Z1GjRg26du3KRx99xMKFC5kzZw5nzpzRJ2RPP/00HTt2ZOLEiQwZMoR9+/ZhZ5f7H3bdunX57LPPiIqKQqVS0bx5c1avXi21jIS4J0uj5fC1O2wJiuCvC9EkZ+Qu1KhTzYGhLX0Y0tKHutWKPuwthDFUihlNboiIiKB69eocPXpUX8cGdLVsfvnll3wTrxyhoaEkJydz7NgxZsyYwZdffskzzzyTb1utVkubNm3o3r07n3/+eb5t8usp8vX1JSEhAScnpzzt09PTCQ0N1U8QL89yijeeO3fO1KGYjWvXrtGhQwcWLlzISy+9BOjmFqlUKnx9fQkJCcHNzS3fbW8eRUX6uRLiflqtwj/X49h8JoLt/0ZyNzV3sYKPsy1D7iVCTX2cTFYkUVQMiYmJODs7F/j5fT+jizeWBnd3d9RqdZ5eoZiYmIf2UORM4m3evDnR0dHMmTOnwKQoZ17K1atXC7yejY2NweRaUbnVq1ePLVu2MGzYMLZu3crUqVNp164d9vb2XLp0ic2bN/P999+zYcMGg0nYQohciqLwb3gCm4N0RRWjEnNXALs5WDO4hTdDW/rQpmZVLGSOjzCBEkmKUlJSSExMxMnJCQeH4lVOvp+1tTVt27Zl9+7djBgxQn989+7dDBs2rMjXURTFoJcnv/NBQUE0b97c6FhF5dOlSxfOnj3L3LlzeeaZZ4iPjwd0SXbnzp35/PPPJSESIh/XYpLYHKSrJXQ9NnfrI0dbSwY09WJISx8613Ur9h5gQpQ0o5OirKwsPvnkE1auXGmwPUOdOnWYOHEib775plG1jKZPn87YsWNp164d/v7+fPfdd4SFhTFlyhQAZs6cSXh4OKtWrQLgq6++ombNmjRq1AjQ1S1asmSJvnAewNy5c+nUqRP169cnMTGRzz//nKCgICk+WYCVK1eaOgSz5ePjw7fffstXX31FWFgYqamp+Pr6lviwmRDmSKNViryM/WZcKlvORrA5KIJLUUn647ZWFjzW2JOhLX3o0aAatlZFXwkqRGkzKilKS0ujb9++BAQEoFaradCgAV5eXkRHRxMcHMwHH3zA1q1b2bt3r8GE06IYPXo0sbGxzJs3j8jISJo1a8b27dupVasWoNuT6/7l0FqtlpkzZxIaGoqlpSV169Zl0aJF+nkfoNu2YfLkyURFReHs7Ezr1q05ePAgHTp0MOb2hcDS0jLf+khCVFQ7z0Xm2WfL+4F9tmKS0tl+VldL6FRYvL6dlVpF9/rVGNrKhz6NPXGwMauZG0LoGTXReu7cucydO5dnn32WRYsW6bc/AN1k6RkzZrB69Wpmz57N7NmzSzRgU3nYRC2ZECtKg/xcCXOw81wkL68+lWe/sJw+ojGdahJyJ4WA4Fi09xqpVOBfR1dUcUAzL1zsS28XBCEKU5yJ1kYlRY0bN8bR0ZETJ04U2KZDhw4kJSVx8eLF4l7eLElSJExBfq6EqWm0Cl0X7zPoISpMK18XXS2hFt54OMnPrDC9Ul99dv36dd54441C2zz22GMsXbrUmMsLIYQwEydC44qUEI1u78vUnvWo6Vb8DaiFMBdGTfW3t7fPd2+s+92+fdtgB3UhhBDlT2R8WpHada7rJgmRKPeMSoo6derEb7/9xvnz5/M9f+HCBdatW2dQgFEIIUT5EZWQzmd/XWbOlgtFau/hKENlovwzavjsvffeY/fu3bRv355JkybRo0cPPD09iY6OZv/+/fz0009kZWUxc+bMko5XCCFEKVEU3ZL7VQE32Hk+Cs29WdMWKvQTqB+kArycdcvzhSjvjEqKOnfuzNq1a3nhhRf46quvWL58uf6coig4Ozvz888/06VLlxILVFQe169fx8/Pj/Xr1/Pkk08C0LNnT6pUqcLWrVtL5D1q167NjRs38hxPS0uTCc2i0knJyObPoHBWHb3B5ejcmkId/FwZ718bBYVXfz0NYLACLWf12ewhTWSXeVEhGF0s4oknnqB///78+eefnD59Wl/RunXr1gwbNgxHR8eSjLNSCogIYNGJRczoMAN/HxmKLMidO3dYunQpmzZtIjg4GAsLCxo2bMizzz7Lyy+/XODctieffJL//ve/BsdkaxdRmYTcTuaXYzf4/eQtku5twmpnpWZEm+qM869FI6/clTqWFqo8dYq8HqhTJER590gVtKpUqcKYMWMYM2ZMScUj7lEUhWWnlhGSEMKyU8vo5N3JZJsiKopCZmamWSYMW7duZcyYMdSrV49XXnmFpk2bkp6eTlBQEN999x3fffcdf/75J40bN87zWk9PTzp16mSCqIUwHY1W4e9LMaw6doODV3IXzPi5OzCmUy2ebFsDZzurPK8b0Mybvk28ilzRWojyyKiJ1hqNhsTERLRabaHnNRrNIwVX3imKQmpWqlFff4f9zflY3UT287Hn+Tvsb6OvVdxSVBMmTNBXEm/ZsiU2NjZs3ryZgIAAevfujYODA87Ozjz77LPExMQYvDYjI4P333+fOnXqYGNjQ40aNZg4caL+fEBAAEOHDsXHxwcHBwdatWrFL7/8YtSf7969e3n66af56quvOHnyJC+99BJdu3alT58+vPnmm/z777/07t2bAQMG5NlkWIjK5m5KJt8eCKbHJ3/zwqqTHLxyG5UKHmvkwc/Pd2Dv9B5M6uqXb0KUQ22hwr+uG8NaVce/rpskRKLCMaqnaO7cuXz88cfcvHmTatWq5TkfFxdHzZo1mTlzJrNmzXrkIMurtOw0Ov7asUSu9fr+141+7fFnj2NvVbylshEREbz++uu8//77+Pr6YmVlRc+ePRk0aBDr1q0jJSWF999/n6FDh3Ls2DH965544gn27dvHu+++S6dOnbh9+zYbNmzQn79x4wZdunRhypQp2NracuTIESZNmoSiKIwbN67I8aWlpTF27FhWrlypn3f0ILVazfLlyxkyZAhvvfVWnuRrzZo1fP/991hZWdG9e3cWL14smwSLCudceAI/H73O5jMRZGTrfpF1trPi6fa+jOlUC19XWUYvRA6jkqKtW7fy2GOP5ZsQAVSrVo0+ffqwadOmSp0UlWd3795l586d+v3hevToQbt27di4caN+GK9Zs2Y0b96c7du3M2jQIHbv3s22bdv49ddfeeaZZ/TXuv/x008/rX+sKArdu3fn1q1bfPPNN8VKin755RdatGihT4iuXbvGxIkTOXHiBA0bNmTGjBm8++67XL9+nQULFtC+fXuWLVuGq6tuhczQoUPp2LEjNWvWJCQkhI8++oiuXbty+vRp2dNMlHsZ2Rp2/BvFqoDrBnuQNfVxYrx/bYa09MHOWjZiFeJBRiVFISEh9OrVq9A2DRs25MiRI0YFVVHYWdpx/NnjxXqNoihM3DWRy3cvo1VyhyctVBY0rNqQn/r/VOy5RXaWxduUF8Dd3V2fEKWmpnLkyBGWLFliMCTasGFDvL29+eeffxg0aBB79+7F3t7eIPF50N27d5k9ezabNm0iPDxcfz03N7dixbd9+3bGjh0L6DYFHjp0KG5ubmzZsoWwsDCmTZumn2DdokUL3N3dOXHiBAMGDADg888/11+rW7du9OvXj0aNGrFkyRKD1ZRClCeRCWn8ejyMtSfCuJOcCeg2Yx3U3Jtx/rVpU9PFZHMThSgPjEqKsrOzsbAofDqSSqUiPb1oe+VUVCqVqtjDVkfCj3AxLu9+cVpFy8W4iwTdDqJL9dIvdeDh4aF/fPfuXTQaDW+88Ua+27vcvHkTgNjYWLy9vQv9T3fChAkcPXqUWbNm0bRpU5ycnPj6669Zt25dseILDg6madOmABw/fpzLly9z69YtvL11q2Du3LnDN998o2/v6+tLbGxsgdfz9vama9euBAYGFisOIUxNURSOhcSxKuA6f12I1tcW8nKy5bmONXm6Q02qOZrfIgkhzJFRSVHdunXZt29foW327duHn5+fUUFVVoqi8MXpL1ChQsmzHzWoUPHF6S/o7NO51H/bu//6Li663y7fffddhg8fnqetu7s7oOvtiYyMRFGUfONLT09n27ZtfPrpp7z66qv64wVN2C+MRqPRr4YLCwujWrVq+oQIoGXLlgbto6Ki9HEWxIi9kYUwmZSMbDaeDueXgOtciU7WH+9UR1dbqE8TT6zURq2lEaLsxN+E1IJ/YcXeDVx8yywco5KiJ554grlz5zJr1ixmz56NWp07Nq3RaJgzZw5BQUF88MEHJRZoZZClzSIqJSrfhAhAQSEqJYosbRbWausyi8vBwQF/f38uXrzI/PnzC2zXp08fFi9ezP/93/8xevToPOczMjLQaDRYW+fGnpSUxObNm4sdU/Xq1QkJCaFx48Z4enoSFxdHSkoKDg4OQG7vFeh6kmJiYgpdfh8REcGRI0f0Q3JCmKvg28n8EnCDDYG5tYXsrdWMbFOdsZ1q09BLasSJciL+JnzZFrIzCm5jaQOvBJZZYmRUUjR9+nTWrVvHRx99xG+//UavXr2oXr064eHh/P333wQHB9O4cWPefPPNko63QrNWW/Pb478Rlx5XYBtXW9cyTYhyfPLJJ/Tu3ZvRo0fz9NNPU7VqVW7dusXu3buZOHEiPXv2pE+fPgwaNIjnn3+e4OBgOnbsSFxcHBs2bOC3337D2dmZ9u3bs2jRIqpVq4alpSWLFi3C2dk5z9L+h+nZsyd//vkngwcPpmPHjri5ufHaa6+xaNEibt26xeLFi1EUhf379zN+/HhmzJiBs7MzAGvXrmXbtm0MHDgQHx8fQkJCWLhwIWq1Ok8xRyHMgUarsO9SDKsCrnPo6h398TruDoz1r8UTbWvgZFvwUnohzFJqbOEJEejOp8aad1JUpUoVDh48yMsvv8yGDRu4du2a/pyFhQVPPvkky5cvp0qVKiUWaGXh5eCFl4OXqcPIo3Pnzhw+fJjZs2czceJEMjMzqVGjBo899hj16tXTt9uwYQNz587l22+/Zc6cOXh6etKvXz/9+V9//ZXJkyczfvx4fSKTnJzMkiVLihXPxIkTadSoEa+99hrNmzdn9erVPPXUU6xYsQIHBwfeffdd3nvvPUaOHMnMmTN566239K/18/Pj1q1bTJs2jfj4eFxcXOjduzfz5s2TIV9hVuJSMln3z01WH7tB+L3d6nW1hTwZ37kWXeq6YyG1goQoMSrlESdSxMTEcPLkSf2HS7t27Qwm6VYUiYmJODs7k5CQgJOTU57z6enphIaG4ufnJ3tnlZFPP/2UL7/8kl27dtGgQQOysrK4evUqNWvWxMrKiuvXr1OvXj2D4d3yRn6uKqezt+L5+egNtpyNIPNebaGq9laMbl+T5zrWlNpConzLSIKbx+HcRgha8/D2kw+ATyuj3+5hn9/3e6RtPkC3SmnQoEGPehkhiu2///0vt2/fpm3btkybNo2nnnqKOnXqkJ2dTWBgICtXruT69evs27dPliELk9JolYduj5GRrWH7v5H8fPQGQTfj9cebV3dmnH8thrT0wdaq/Cb4ohJLuwthx+D6YbhxFCLPgGKeO148clKUIyYmRl+XqFOnTgYrgYQoLYsWLaJPnz7MmzePBQsW6FeyOTo6Mnr0aFatWiUJkTCpneci82yk6n3fRqoR8WmsOX6D307cJDZFV1vIWm3B4BbejPOvRStfqS0kypmUO3DjiC4Bun4Eos/BgwuIXGqBRxO4ssMkIRakyElRUFAQGzduZOTIkbRq1crg3Lfffsu0adPIzNT9g7a0tOSjjz6SidaiTPTp04c+ffqQmJjIzZs3sba2xs/PD0vLEsv5hYkUpYfFnO08F8nLq0/lWU8alZDOlNWnaOXrzNlbCdwrLYS3sy1jOtVidHtf3KtIbSFRTiRG3kuC7iVCty/lbeNWH2p1htpddd+da0BEUPlNilatWsUXX3zBK6+8YnD81KlTTJ06Fa1WS/fu3XFwcGD//v288847tG/fnh49epR40ELkx8nJSV/QUZR/D+thMXcarcLcLRfyLbCRcyzoZgIAneu6Mc6/Nn0ae2AptYWEuYsP0/UA5SRCcSF523g0gVpddAlQrS7g6Fn2cRqhyEnR0aNH851E/eWXX6IoCvPnz+fdd9/Vt+3evTvffPNNpUuKpACgKEmV9eepsB6Wl1ef4usxbUolMcrWaEnP1pKepbn3pSUj+973LA3p2RoysrSk3zv2YJv0LI3+cUR8qkFCV5AlT7bgyXZlV5xOiGJRFF3Sc+NIbiKUcNOwjcoCvJrfS4LuJUL2rg+/tr2brg7Rw+oU2RdvG6hHUeSkKCwsjP79++c5vmfPHuzs7AyGyjp37kzfvn0Ndk+v6KysdDVCUlNTsbMr/l5jQuQnJSUFlUql//mqDIrSw/LuH+dQq1RkaRV9YpL+QNKSkZWb3GRk5yYwOQlNxgNt0rO1+i0yypKVpfQMCTOiKHD7Mtw4nDsnKDnKsI1KDT6toXYXqNUVanYEW+fiv5eLr64wY3msaB0bG4uPj4/BsZiYGG7dukWfPn0MqhQDNGnShP3795dIkOWBWq3GxcVFX4TQ3t5eJkcKoyiKQnZ2NomJiSQmJuLi4lKuywoUVWa2ltA7KWz7N+KhPSxxKZm8+Evp7lNnbWmBraUFtlZqbK3U2OgfW9x7rnuc811/zlKNjZUFkQnp/HTk+kPfx8NRSi0IE9JqdROhbxzNTYQeTFLU1lC93b05QV2gRgewKaE6hC6+ZZr0PEyRkyI7Oztu375tcOzUqVMAtG3bNk97a2vrSvXbLYCXl67oYnGrMwuRH7Vajbe3t74Sd0WRma3lemwKV6KTuBKdzLUY3ffrd1LILkZPja+rHV5OtgYJSk5ion9uqTY4ZmOVm+TkJDw2D7SztVJjrbZ45KKIGq3CznNRRCWk59vrpQK8nHWTx4UoNmP3DNNkQ9SZ3KGwsABITzBsY2kHvu11vUC1u0D1tmBVOUZAipwUNWnShB07dqDRaPS/tW7btg2VSkXXrl3ztA8LC6t0y/JVKhXe3t54eHiQlZVl6nBEOWZpaYlarS7XvY05yc/V6GSuRCdxtQjJTxUbS7ycbbgWk/LQ63/8REv865bdXIPiUluomD2kCS+vPoUKwwXJOX+rs4c0KVer6YSZKM6eYVU8IeJU7pygm8chM9mwrXUVqNkpd06QT2uwLPvtpMxBkZOi5557jldffZWhQ4fy0ksvce3aNX744Qfc3Nzo27dvnvYHDx6kTZs2JRpseaFWqyvFcIcQAFkaLdfvpHDlvuTnanQyoQ9Jfup5VKGBZxUaeDree+yIt7MtWgW6Lt5XIXpYBjTz5usxbfKsovMqR6vohBkq6p5h68bqlsdnpxmes3W5tyrs3sowrxaglhImUIykaMqUKfz555/s2LGDnTt3oigKarWa//3vf3nmE/3999+Eh4czY8aMEg9YCPFwpVHf5/7kJyfxuRKdVKTkp/69pKe+Z27yU1AvmFpFhephGdDMm75NvMp1vSVRTkWe1n23d7+vRlAX3XJ5C5ngn58iJ0VqtZpdu3axZs0ajh07hqurKyNGjMi3NygmJobXX3+d4cOHl2SsQogieNT6PjnJz9WYez0/95Kg0DspZGnyT34crNXU83SkwX3JT31PR3wKSX4KU9F6WNQWKrMe6hPljLaIW2R0mw4tngb3BrqdhMVDPfKGsJVFcTaUE8JUCqrvk/Pf4f31fbI0Wm7E3j/slczV6KInPzmJT4NHSH4eprxXtBaixGg1upVh5zfqNlJNj3/4ax5xI9WKokw3hBVCmIei1Pd5c/1ZNp+J4FpMcpGSn/r35v3Uv/e4uotdmU7+lh4WUalptXDrhC4JurApb70gUeIkKRKigjgRGvfQ+j7JGdls/zf3P1Z7azX1PXJ6fKpQ30M39OXjbPfIS9KFEEZQFAg/pesROv8nJN7KPWfrDI2HgFdL2PGWyUKsyCQpEqKCiEl6+JYSAENb+jCidXVJfoQwF4oCUWd1PULn/4D4G7nnrB2h0WBoNhLq9NItlY8IMlmoFZ0kRUJUAIqiEByT/PCGwDMdasqQlBDmIOYinNugS4Rir+Uet3KAhgOg6Uio1wesHqh6boZ7hlUUkhQJUc79cz2Oj7ZdJOhmfKHtylN9HyEqrDvXcidL376Ye9zSFur30/UI1e8P1vYFX8MM9wyrKCQpEqKcCrmdzOKdl9h1PhrQzQ/q3ciDbWcjgfJf30eICuPu9XtDYxsh6t/c42prXU9Q05G6niEbx6Jf08z2DKsoSjwpysjIQK1WY2kp+ZYQpSE2OYPP915lzfEwsrUKFioY3b4mb/Stj4ejLY+3yFunqLzW9xGi3Eq4pRsWO7dRt81GDgtLqNNTlwg1Ggx2LqaKUOTDqMzl8OHD7Nu3j9deew0XFxcAYmNjGTNmDHv27MHKyoo33niDjz76qCRjFaJSS8/SsOJIKF//HUxSRjYAvRt5MGNgIxp45v6GKRWUhTCRpCjdirHzG3V7jOVQWUDtbrqhscZDwV6GsM2VUcUbR4wYwdmzZwkODtYfmzBhAqtWraJ+/fokJSURHR3NunXrePLJJ0s0YFOR4o3CVLRahT+Dwlmy6zIR93p/mvo48d6gxnSu527i6ESlYOyO7JVB8m24uAnO/aHbdFU/cK3Sba3RdAQ0GQZVPEwZZaVW6sUbg4KC6NGjh/55amoq//d//0e/fv3YuXMnSUlJtGjRguXLl1eYpEgIUzh67Q4fbb/I+YhEAHycbXmzf0OGt6ouS+lF2SjOjuyVJTFKjYOLW3Q9QqGHQLlv240aHXQ9Qk2Gg5MMV5c3RiVFMTExVK9eXf88ICCA9PR0Jk6cCICjoyOPP/44GzZsKJkohahkrkQnsXD7Rf6+fBsARxtL/tOrHhO71MbWSm3i6ESlUtQd2VNjK3ZSlJ4Al7brltCH/A3a7NxzPq11c4SajqjYfwaVgFFJka2tLUlJSfrnBw4cQKVSGfQeValShbt37z56hEJUIjGJ6fxvzxXW/XMTrQKWFirGdKrFq73r4VbFxtThCVH+FWcoMCMZruzUTZa+ths0mbntPJtDsxG6RMi1TunGLMqMUUlRvXr12LlzJxkZGVhYWLBu3TqaNGmCl5eXvk1YWBgeHjKGKkRRpGZm893BEL47GEJqpq4rfkBTL94e0JA61aqYODohimDza7rhIhsn3dLynC9b5/ueOxk+tnXSDb2VlaIMBaptYMBCCD0AV/6C7LTcc9Ua6XqEmo0E9/qlH68oc0YlRS+++CKTJ0+mfv36WFtbExoayscff2zQ5vjx4zRp0qREghSiotJoFdafvMlnu68Qk6T7j7qVrwvvDW5M+9qyQkWUI1FndF/FpbZ+IGG6993WKZ9kyumB445gcy/psrR++HsVZShQkwHbpuc+d62rS4KajgRP+Uyr6IxKiiZNmsTVq1f58ccfSU1N5aWXXmLatGn683///TchISFMmTKlpOIUokJRFIX9V26zaPslLkfrhqJ9Xe14Z0AjBjf3LtOd6IUoVMqdorXrM0c39JSeCBlJ974S7nucZHgu894UDE2mLlkpbEirKNQ2+SRS9z23ddK9b1FU8YKWT+uSIa8WIP8eKw2jluQ/TGZmJmlpaTg4OFSYIo6yJF+UlPMRCSzcfonD13QfNs52Vrzaux5j/WthYymTqIWZUBQ4/QvsmAlZRdhXb/IB8GlV9OtrNZCZnE/ClHjv68Fk6v7z953LLNqef8Uyeb9u8rSoEEp9Sf7DWFtbY21dhK5MISqRyIQ0luy6wsbTt1AUsFZbML5zLV7pVR9neytThydErrhQ2PIahB4svfewUOvmG9k6P9p1tJr7kqREw8fpDzy/ewOu7CjCRaVnqLIyKinaunUrK1as4Msvv8THxyfP+YiICF555RVefPFFBg4c+MhBClGeJaVn8c2BYH44FEpGthaAIS19eLt/Q3xdC9n0UYiyptXA8W9h34eQlQqWduA/FQK+MN8d2S3Uuq0yirJdRkRQEZMiUVkZlRR99dVXRERE5JsQAfj4+BAaGspXX30lSZGotLI0Wn47EcbSPVeJTdEt5e1Q25V3Bzemla+LaYMT4kExF2HTKxB+Uve8djcYsgzc6kLbCVLRWlQKRiVFZ86c4fHHHy+0TceOHdm6datRQQlRnimKwu4L0SzaeYmQ2ykA1HF34J2BjejXxFMmUQvzkp0JR5bCgY9Bm6WbnNx3HrQZDxYWujayI7uoJIxKiuLi4h5ag8jd3Z07d4q4akGICiLoZjwLtl3kxPU4AFwdrJnWpz7PdKiJldrCxNEJ8YDwQNj0KsSc1z1vMAAGfwbO1Qt/XXll76Yb6jPXoUBhckYlRdWqVePy5cuFtrl8+TKurlJnRVQON+NS+XjXZbaciQDAxtKCSV39mNKzLk62MolamJnMVNi/EAK+BEWrSwIGfgzNnqjYy89dfHV7tMlQoCiAUUlRjx49+P333zl79iwtWrTIc/7MmTNs3ryZkSNHPnKAQpizhNQsvvz7Kj8fvUGmRotKBSNaV+fNfg3xcbEzdXhC5HX9MGx+FeJCdM+bPwUDFoNDJekdkaFAUQij6hT9+++/dOjQASsrK95880369u1L9erVCQ8P56+//uLTTz8lKyuL48eP55s0lUdSp0jcLzNbyy/HbvD53qskpGUB0KWeGzMHNqZZ9UdcYixEaUhPhD2z4eQK3XNHH3j8f9BwgGnjEqKUFefz2+jijX/88Qfjxo0jNTXV4LiiKFSpUoVVq1YxfPhwYy5tliQpEqD7+d7+bxSLd14iLE73s9/AswozBzWmZ4NqMolamKcru2DrG5AYrnvediL0nfvoNYKEKAfKpHjjiBEjCAkJYeXKlfzzzz/Ex8fj4uJChw4dGD9+PNWqVTP20kKYjEarcCI0jpikdDwcbeng54raQpfonLwex0fbL3I6LB6Aao42/LdvA55sWwNLmUQtzFHKHdg5A/5dr3te1Q+GfgF+3UwblxBmqlS2+aiIpKeo4tt5LpK5Wy4QmZCuP+btbMuUHnUJCI5l5/koAOys1LzUow4vdquDg03F2MZGVDCKAuc2wI63dZOKVRa6Iow93wVrKRgqKheTb/MhRHmz81wkL68+xYO/IUQmpDN7s265soUKRrXzZXrfBng42ZZ9kEIURWIEbJ2eW7nZoykM+wKqtzVtXEKUA0VKig4e1O1/06FDB2xtbfXPi6J79+7GRSZEGdFoFeZuuZAnIbqfjaUFf/ynC018pJdQmClFgVM/w18f6Pb5srCCHm9Dl2lgKXtRClEURUqKevbsiUql4uLFizRo0ED/vCg0Gs0jBShEaTsRGmcwZJafjGytfpWZEGYnLgQ2vwbXD+meV28Hw74Ej8amjUuIcqZISdGsWbNQqVS4u7sbPBeiIohJKjwhKm47IcqMVgPHvoZ98yE7DazsofcH0PEl3UapQohiKVJSNGfOnEKfC1GeVSniZGkPR5lHJMxI9AXY/Ipuqw4Avx66DVxd/UwblxDlmFETrefNm0edOnUYM2ZMSccjRJk6F57AnHsTqQuiArycdcvzhTC57Ew4/BkcXHJvA1dn6D8fWo+t2Ft0CFEGjCquMn/+fP7999+SjkWIMqMoCr8EXGfk8qPcvJuGq71uIuqDHyk5z2cPaaKvVySEydwKhO966PYt02ZBw8Ew9Ti0GScJkRAlwKikqFatWsTFxZV0LHrLly/Hz88PW1tb2rZty6FDhwpse/jwYbp06YKbmxt2dnY0atSI//3vf3nabdiwgSZNmmBjY0OTJk34448/Si1+Yd4S07N45dfTfLDpPJkaLX0ae7LvzR58M6YNXs6GQ2RezrZ8PaYNA5p5myhaIdBt4LrrPfixD8RcAHt3ePIneHoNOMnPphAlxajhs2eeeYaVK1eSkJCAs3PJlolft24d06ZNY/ny5XTp0oVvv/2WgQMHcuHCBWrWrJmnvYODA6+88gotWrTAwcGBw4cP89JLL+Hg4MDkyZMBCAgIYPTo0Xz44YeMGDGCP/74g1GjRnH48GE6duxYovEL83YuPIGpv57iRmwqlhYqZgxsxKSufqhUKgY086ZvE68CK1oLYRKhB3UbuN69rnve4mkYsBDsZThXiJJmVEXrzMxMRo4cSWRkJPPmzaN9+/Z4eHiUSEAdO3akTZs2fP311/pjjRs3Zvjw4SxcuLBI1xg5ciQODg788ssvAIwePZrExER27NihbzNgwACqVq3K2rVri3RNqWhdvimKwurjYXy45QKZGi3VXez44tnWtKlZ1dShCZG/9ARdzaFTP+ueO9WAIUuhfl+ThiVEeVPqFa3t7OwA3QfN0KFDC2ynUqnIzs4u8nUzMzMJDAxkxowZBsf79evH0aNHi3SN06dPc/ToUebPn68/FhAQwBtvvGHQrn///ixdurTA62RkZJCRkaF/npiYWKT3F+YnKT2LGRv/ZdvZSAD6NPZgyVMtcbGXgnbCTF3eodvANUn3M0v7F+Cx2WArv5AJUZqMSoq6detWKnWK7ty5g0ajwdPT0+C4p6cnUVFRhb62Ro0a3L59m+zsbObMmcMLL7ygPxcVFVXsay5cuJC5c+cacRfCnJwLT+CVX09xPZ/hMiHMTsod3X5l5zbonrvW1W3gWruLaeMSopIwKinav39/CYdh6MEPLEVRHvohdujQIZKTkzl27BgzZsygXr16PPPMM0Zfc+bMmUyfPl3/PDExEV9f3+LchjAhRVFYczyMeVsvkJmtxcfZli+ebUPbWjJcJsyQosC/v+sSorQ4UKmh86vQcwZY2Zk6OiEqDbPaENbd3R21Wp2nBycmJiZPT8+D/Px0BcuaN29OdHQ0c+bM0SdFXl5exb6mjY0NNjY2xtyGMLGk9CxmbvyXrfeGyx5r5MGno2S4TJiphFu6DVyv7tI992yu28DVp7Vp4xKiEjJqSb5arebDDz8stM3ixYtRq4tXZt7a2pq2bduye/dug+O7d++mc+fORb6OoigG84H8/f3zXPOvv/4q1jVF+XA+IoEhXxxm69lILC1UvDeoMT+MbycJkTCN+JsQEZT/V/hpOPQpfNVJlxCpraH3+zD5b0mIhDARo3qKFEXBiEVrRTJ9+nTGjh1Lu3bt8Pf357vvviMsLIwpU6YAumGt8PBwVq1aBcBXX31FzZo1adSoEaCrW7RkyRJeffVV/TVff/11unfvzuLFixk2bBibNm1iz549HD58uFTuQZQ9RVH49UQYc7fIcJkwE/E34cu2kJ3x8LY1Oug2cK3WsPTjEkIUqNSGz27fvq1fpVYco0ePJjY2lnnz5hEZGUmzZs3Yvn07tWrVAiAyMpKwsDB9e61Wy8yZMwkNDcXS0pK6deuyaNEiXnrpJX2bzp0789tvv/H+++/zwQcfULduXdatWyc1iiqIpPQs3v3jHFvORADQu5EHnz7VkqoO0jskTCg1tmgJUefXoM8c2cBVCDNQ5DpFOT0zABMmTGD48OEMHz48TzuNRsOtW7dYunQp9evX59ixYyUWrClJnSLzdD4igVd+PU3onRTUFireGdCQF7rWwUIKLgpTiwjSbcnxMJMPgE+r0o5GiEqrVOoUTZgwQb9aS6VSsWnTJjZt2pSnXU6OZWdnx5w5c4oRthBFl/9wWWva1pIqv8JMJNwydQRCiGIqclL0008/AboPo+eff57hw4czbNiwPO3UajWurq74+/tTtarM5xAlLzkjm3c3/stmGS4T5iQ7A24cgau74epfEHvN1BEJIYqpyEnR+PHj9Y8PHDjAiBEjCq1mLURpuBCRyNRfT+mHy97u35AXu8lwmTCR+JtwbbcuEQrZD1mpuedUalA0JgtNCFF8Rk20zuk1EqKsKIrC2hM3mbPlPJnZWrydbflShstEWdNkwc3jup6gq7t1O9bfz9FbtzdZ/X5g5worB5kmTiGEUR5p9dkff/zB2rVruXTpEqmpqVy7pusuvnTpEps3b+a5556jevXqJRKoqLweHC7r1bAan41qJcNlomwkRcG1PbpEKPhvyLhvH0SVBfh2zE2EPJtBTqX8iCCThCuEMJ5RSZFWq+WZZ57h999/B3STqtPS0vTnq1atynvvvYdGo2HmzJklE6molC5GJjJ1zSlCZLhMlBWtBsID7/UG/QWRZwzP27tBvb7QoB/U6QX2BfRW2ruBpU3hy/ItbXTthBBmochL8u/36aef8tZbbzFlyhQWLVrEZ599xocffohGkzt+3rt3bzIzMytMgURZkl+2FEVh3T83mb35PBn3hsu+eKY17WrLcJkoBalxcG2vrrL0tb26/cfu59NG1xNUv5+u2rRFETcDiL+pq1dUEHs3cJE9FYUoTaWyJP9+K1eupF27dixfvhzIu9kqQL169di2bZsxlxeVXEpGNu/98S9/BuUOl306qhWuMlwmSopWC1Fn4Oq9YbFb/wD3/X5o6wx1H9MlQfUegyoexr2Pi68kPUKUI0YlRdeuXWPq1KmFtnFzcyM2tpDfkITIx8VI3eqykNu64bK3+jdksgyXiZKQnqCbE3R1t27FWHK04XnP5rlzg2q0B7VZ7ZcthCgDRv2rt7OzIzExsdA2N27cwMXFxZjLi0roweEyLyfd6jIZLhNGUxSIuZi7UiwswHCJvHUVqNPzXm9QH3CWRSFCVHZGJUWtW7dm165dZGRkYGNjk+d8XFwcO3fupHv37o8coKj4UjKyef/Pc/xxOhyAnvdWl8lwWSVm7FycjGQIPZibCCU+UFXavWFub1BNf7CUnzEhRC6jkqLXXnuNESNG8OSTT/LNN98YnAsODub5558nISGB1157rUSCFBXXpahE/rMmd7jszX4Neam7DJdVakXZXd7SBl4JBOcaEBucu1LsxhHQZN7Xzg78ut9LhPpC1dqlHr4QovwyKikaNmwYM2bMYNGiRdSsWRMHBwcAPDw8iI2NRVEUPvjgA3r37l2iwYqKQ1EU/u/kTWZtyh0u++LZ1rSX4TJRlN3lszNg9yyIOA13Qw3PVa0N9fvreoNqdwEru1ILVQhRsRg9k3DBggX06tWLL7/8kuPHj5Oeno5Wq2XAgAG89tpr9O/fvyTjFBXIg8NlPRpU43+jZbhMFNP5jbrvFla65CdnybxbvdwCikIIUQyPtLyib9++9O3bt6RiEZXA5agk/rMmkOB7w2X/7deAKd3rynCZKL7GQ6DlM+DXA2yqmDoaIUQFIGtORZlQFIX1J28xa/M50rN0w2WfP9OaDn4yXCbQDYfd+ke3qeqlItY36/Ym+LQqzaiEEJWMJEWi1KVkZPPBn+fYeN9w2WejWuJWJe/KRVFJaLW6zVRD9kPI33DjqOEO80IIYQJFTors7e2LfXGVSkVKSkqxXycqDhkuE3rxN+8lQfsh9ACk3DY87+ChqxvkWgcOLDJBgEKIyq7ISVF6ejpWVlb4+PiUZjyinNJoFU6ExhGTlI6Hoy3ta1dl4+lwZm3SDZd5OtnwxTNtZLisMkmLh+uHchOh2GuG560cdBOk6/TUfXk00U2QjgiSpEgIYRJFToqsrKzIysrCxcWFiRMnMmbMGFxd5QNOwM5zkczdcoHIhHT9MTsrC9KytAB0b1CN/8lwWcWXnQE3T+QmQRGnQNHmnlepoXpbXQJUtxdUb5d/8UTZXV4IYSIqRVGUhzfTValetWoVK1eu5OzZs9jY2DB06FCef/55+vXrl++msBVJcXbZrUx2novk5dWnKOiHaFgrH/43qpUMl1VEWi3EnNftJxayXzcvKDvNsI17g3s9Qb10vUK2zkW7tuwuL4QoIcX5/C5yUnS/kydPsmLFCn777TcSEhLw8fFh/PjxTJw4kbp16xoduDmTpCgvjVah6+J9Bj1ED/J2tuXwO71RS1JUMcSH5fYEhRyA1DuG56t45g6H+fWQ/cSEECZX6klRjoyMDH7//XdWrFjB/v37AejWrRvfffcdDRo0MPayZkmSorwCgmN55vtjD2239sVO+NeVoY5yKe0uhB7KXSUWF2J43soBane9b15QYymcKIQwK8X5/H6kJfk2NjY899xzPPfccwQEBDBq1CgOHTrEpUuXKlxSJPKKSSq4h8iYdqIEPOqwU3YG3DyeOyQWGZR3XlCNdrlJUEHzgoQQohx6pKRIo9GwefNmVqxYwa5du8jOzqZt27bUr1+/pOITZszDsWgTpz0cbUs5EgEUbyPVnMRIq4Xof3OHxG4E5DMvqGHu5OhaXcBWekqFEBWTUUnRuXPn+Omnn1i9ejW3b9/G3d2dqVOnMmnSJJo1a1bSMQozpNEq/BkUXmgbFeDlbCvL8MtKUTdSjT4Hwfty6wU92LOknxfUC+r0ACcpwyGEqByKnBQlJiby66+/smLFCgIDA7GwsKB///48//zzDB06FEtLKY5dWWRrtLy5/gx/BkWgAhTQf8+RM6tk9pAmMsna3Kx92vC5dRXDeUHVGsm8ICFEpVTkTMbLy4uMjAwaNGjAggULGD9+PF5eXqUZmzBDmdlaXv/tNDvORWFpoeJ/o1thpVblqVPk5WzL7CFNGNDM24TRivxZgG+H3CSoRjtQW5k6KCGEMLkirz6zsLDAysqqWEvuVSoV58+fNzo4cyKrzyA9S8PLqwP5+/JtrNUWfPVcG/o28QTyVrTu4OcqPURl7cIW+L8xD283YbuuZpAQQlQCpbb6LCsri0uXLj1ScKJ8SsnI5oWfTxIQEoutlQXfjW1H9wbV9OfVFipZdm8KscFw/g/dV/S5or3G2qF0YxJCiHKqyEmRVqt9eCNRISWmZzHxp38IvHEXB2s1Kya0p2MdSYBMJi4ULvwJ5zZC1Nnc4yo1KBqThSWEEOWdzI4WhYpLyWTciuOcC0/EydaSVZM60srXxdRhVT7xYXD+Tzi/ESJO5x5XqXUrxJqOBJdasGqIyUIUQojyTpIiUaCYpHTG/HCcK9HJuDlY88ukjjTxqZzzqUwi4RZc2KQbGrv1T+5xlQXU7gZNR0DjoeBwr9cu/qZspCqEEI9AkiKRr4j4NJ774Tihd1LwdLJhzQsdqefhaOqwKr7EyNxE6Ob9W6iodMvmmw6HxsOgSrW8r3Xx1RVmlI1UhRDCKJIUiTxuxKbw7PfHCY9Po7qLHb++2JFabjI5t9Qkx+QmQjeOklvxSQU1/XU9Qk2GgmMRSmC4+ErSI4QQRpKkSBi4FpPMcz8cIzoxAz93B1a/0JHqLnamDqviSblzXyJ0xHB/sRodoNlIaDJMqkkLIUQZkqRI6F2ISGTsj8eJTcmkgWcVVr/QUfYtK0mpcXBxi26ydOghw5Vi1dvqJks3GSY9PUIIYSKSFAkAgm7GM+7H4ySmZ9OsuhOrnu+Iq4Psfv7I0u7CpW26HqGQ/aDNzj3n3epej9BwqFrLRAEKIYTIIUmR4ERoHM+v/IfkjGza1HThp4kdcLaTbR+Mlp4Al7brEqHgfaDNyj3n1Vw3R6jpCHCtY7oYhRBC5PFISVFUVBSBgYHEx8ej0eRfNG7cuHGP8hailB26epsXV50kPUuLfx03fhjfDgcbyZWLLSMJLu/QJULX9oAmM/ecR9PcRMi9nuliFEIIUSijPv3S09N58cUXWbt2LQVtnaYoCiqVSpIiM7b7QjRT15wiU6OlV8NqfD2mLbZWalOHZRrxN4u/lD0jGa7s1CVCV3eD5r76QO4NdUNjTUdAtYalE7MQQogSZVRS9M4777BmzRoaNGjAM888Q40aNbC0lN6F8mTLmQjeWBdEtlZhQFMvPn+mNdaWFqYOyzTib8KXbR9e9PCVQF1ydPUv3WTpK39BdlpuG7d6usnSTUeAR2NQyYa4QghRnhiVyaxfv54mTZoQGBiIjY1NScckStn6kzd5Z8NZtAoMb+XDkqdaYqmupAkR6HqICkuIQHd+86tw8zhkpeYer+qX2yPk2UwSISGEKMeMSori4+N59tlnJSEqh34JuM4Hm84D8EwHX+YPb47aQj7IiyTkb913l5r35giNBO+WkggJIUQFYVRS1LhxY6Kjo0s6FlHKvjsYzILtlwCY2KU2sx5vgko+0Iuu+SjoNAV82kgiJIQQFZBRYybvvPMOmzZt4tq1ayUdjygFiqKwdM8VfUI0tVddSYiM4T9VV2RR/tyEEKJCMqqnyMvLiwEDBtChQwemTZtG69atcXZ2zrdt9+7dHylA8WgURWHRjkt8ezAEgLf6N2RqL1kWLoQQQjzIqKSoZ8+eqFQqFEVhzpw5hfY4FFS/SJQ+rVZhzpbzrAq4AcAHjzdhUlc/E0dlhgooKyGEEKJyMSopmjVrlgy9mDmNVmHGhrOsD7yFSgUfDW/Osx1rmjos86PVwsHFpo5CCCGEGTAqKZozZ04JhyFKUpZGyxvrgth6NhILFXw6qiUjWtcwdVjmR6uFra/rKlE/jKWNrkaREEKICksqLlYw6VkaXvn1NHsuRmOlVvH5060Z2Nzb1GGZH60Wtk2HU6tAZQH9F0LNTgW3z6+itRBCiArlkZKilJQUNm3aRFBQEAkJCTg5OdGqVSuGDx+Og4NDScUoiigtU8PkX05y6OodrC0t+HZMW3o18jB1WOZHUWD7mxD4E6CC4d9Ay9GmjkoIIYSJGZ0U/fnnn7zwwgvcvXvXYP8zlUqFi4sL33//PSNHjiyRIMXDJWdk8/zKfzgRGoe9tZofxrWjcz13U4dlfhQFtr8FJ39ElxB9LQmREEIIwMikKCAggFGjRqFWq5k8eTI9e/bEy8uL6Oho9u/fz8qVK3n66ac5cOAA/v7+JR2zeEBCahbjfjrBmZvxONpYsvL59rSt5WrqsMyPosDOGfDP94AKhn0FrZ4xdVRCCCHMhEopaJv7Qjz++OMcOHCAgIAAmjVrluf8uXPn8Pf3p2fPnmzZsqVEAjW1xMREnJ2d9cOE5uJOcgZjfzzBxchEXOyt+OX5jjSvkX/NqEpNUWDXu3Bsue750C+hzVjTxiSEEKLUFefz26iK1gEBAYwePTrfhAigWbNmjBo1iqNHjxpzeVFE0YnpjP42gIuRibhXsWHdZH9JiPKjKPDX+7kJ0ZDPJSESQgiRh1HDZ6mpqXh4FD6B18PDg9TU1ELbCOPdjEvluR+OExaXirezLWte6EidalVMHZb5URTYPQsCvtQ9f3wptB1v0pCEEEKYJ6N6imrXrs3u3bsLbbN3715q165tzOXFQ4TeSWH0twGExaVS09We/3vJXxKi/CgK7JkDRz/XPR/8KbSbaNKQhBBCmC+jkqLRo0cTGBjI+PHjiYiIMDgXGRnJhAkTCAwMZPRoWdVT0q5EJzHq2wAiEtKpW82B/3vJH19Xe1OHZX4UBfZ9CEeW6p4PWgLtXzBpSEIIIcybUROt09LS6N27N8ePH8fa2pp69erh6elJdHQ0165dIzMzkw4dOvD3339jZ2dXGnGXOXOYaH0uPIGxPx7nbmoWjbwcWf1CR9yr2JgkFrOmKPD3R3DwE93zgR9Dx5dMG5MQQgiTKPWJ1nZ2dhw4cIC5c+dSvXp1Lly4wN9//82FCxeoUaMGc+fO5cCBAxUmITIHgTfu8sz3x7ibmkXLGs78NrmTJEQF2b8oNyHqv1ASIiGEEEViVE/Rg5KSkkhMTMTJyQlHR8eSiMvsmLKn6GjwHV74+SSpmRo61HblxwntcLS1KtMYyo39i2H/At3jfh9B51dMG48QQgiTKs7nd4nsfebo6FhhkyFT+/tyDFN+CSQjW0u3+u58O7Yt9tayZV2+Dn6SmxD1/VASIiGEEMUin65mbOe5SF5de5osjUKfxh58+WwbbK3Upg7LPB36FPbN1z3uMwe6vGbScIQQQpQ/RUqK6tSpg0qlYs+ePfj5+VGnTp0iXVylUhEcHPxIAVZWf54O57/rz6DRKgxu4c3S0a2wUhs1BaziO7wU9s7TPX5sFnR9w6ThCCGEKJ+KlBRptVpUKlWBzwtSAtOVKqW1J8J4949/URR4sm0NFj/RArXFw/+8K6Ujn8Oe2brHvd6Hbv81bTxCCCHKrSIlRdevXy/0eUlbvnw5n3zyCZGRkTRt2pSlS5fSrVu3fNtu3LiRr7/+mqCgIDIyMmjatClz5syhf//++jYrV65k4sS8RfvS0tKwtbUttfsoCo1W4URoHDFJ6Xg42nI+IoH52y4CMLZTLeYObYqFJET5O/ol7P5A97jnu9DjLdPGI4QQolwzuzlF69atY9q0aSxfvpwuXbrw7bffMnDgQC5cuEDNmjXztD948CB9+/ZlwYIFuLi48NNPPzFkyBCOHz9O69at9e2cnJy4fPmywWtNnRDtPBfJ3C0XiExIz3Nucvc6zBzYqEg9cpVSwHL46z3d4x7vQM93TBuPEEKIcs+oSSq9e/dm1apVhbZZu3YtvXv3Lva1P/vsMyZNmsQLL7xA48aNWbp0Kb6+vnz99df5tl+6dClvv/027du3p379+ixYsID69euzZcsWg3YqlQovLy+Dr8JkZGSQmJho8FWSdp6L5OXVp/JNiABa+7pIQlSQ49/Crpm6x93fgp4zTRuPEEKICsGopGj//v0PHUILCwvjwIEDxbpuZmYmgYGB9OvXz+B4v379OHr0aJGuodVqSUpKwtXV1eB4cnIytWrVokaNGjz++OOcPn260OssXLgQZ2dn/Zevr2+x7qUwGq3C3C0XKGjGlQqYt/UCGq3MycrjxPew423d427/hV7vgSSPQgghSkCpLWdKSUnByqp4BQbv3LmDRqPB09PT4LinpydRUVFFusann35KSkoKo0aN0h9r1KgRK1euZPPmzaxduxZbW1u6dOnC1atXC7zOzJkzSUhI0H/dvHmzWPdSmBOhcQX2EAEoQGRCOidC40rsPSuEf36A7W/qHneZBr0/kIRICCFEiSnynKKwsDCD5/Hx8XmOAWg0Gm7dusX69eupXbu2UUE9OGykKEqRhpLWrl3LnDlz2LRpEx4eHvrjnTp1olOnTvrnXbp0oU2bNnzxxRd8/vnn+V7LxsYGG5vS2UYjJqnghMiYdpXCyZ9g272VZZ1f09UikoRICCFECSpyUlS7dm19YqJSqVi2bBnLli0rsL2iKHzyySfFCsbd3R21Wp2nVygmJiZP79GD1q1bx6RJk1i/fj19+vQptK2FhQXt27cvtKeoNHk4Fm2Cd1HbVXiBP8PWabrH/q9A33mSEAkhhChxRU6Kxo0bh0qlQlEUVq1aRcuWLWnVqlWedmq1GldXV3r37s2AAQOKFYy1tTVt27Zl9+7djBgxQn989+7dDBs2rMDXrV27lueff561a9cyePDgh76PoigEBQXRvHnzYsVXUjr4ueLtbEtUQnq+84pUgJezLR38XPM5W8mc+gW23KtO3ek/0G++JERCCCFKRZGTopUrV+ofHzhwgIkTJ/LaayW/lcL06dMZO3Ys7dq1w9/fn++++46wsDCmTJkC6Ob6hIeH61e/rV27lnHjxrFs2TI6deqk72Wys7PD2dkZgLlz59KpUyfq169PYmIin3/+OUFBQXz11VclHn9RqC1UzB7ShJdXn0IFBolRzsf97CFNpGDj6TWw+VXd445ToP8CSYiEEEKUGqPqFIWGhpZ0HHqjR48mNjaWefPmERkZSbNmzdi+fTu1atUCIDIy0mAu07fffkt2djZTp05l6tSp+uPjx4/XJ3Lx8fFMnjyZqKgonJ2dad26NQcPHqRDhw6ldh8PM6CZN1+PaZOnTpGXsy2zhzRhQDNvk8VmFoLWwqapgALtX4QBiyQhEkIIUapUiuzFUSSJiYk4OzuTkJCAk5NTiV33wYrWHfxcpYfozDr44yVAgXaTYPCnkhAJIYQwSnE+v42uaJ2UlMSXX37Jnj17iIiIICMjI08b2RD24dQWKvzrupk6DPNxdj38OQVQoO1EGLREEiIhhBBlwqik6Pbt23Tu3Jng4GCcnJz0WVhmZiZpaWkA+Pj4FLtOkajkzm2APyaDooU242DwZ2BRaqW0hBBCCANGfeLMmTOH4OBgVq1axd27dwF44403SElJ4fjx43To0IHatWtz/vz5Eg1WVGDn/4ANL+oSotZj4PFlkhAJIYQoU0Z96mzfvp3HHnuMMWPG5Cmq2L59e3bs2MH169eZM2dOScQoKroLm+D3SaBooNVzMOQLSYiEEEKUOaM+eSIjIw12oFer1fphM4CqVasycOBA1q9f/+gRiort4hb4/XldQtTyGRgqCZEQQgjTMOrTx9nZmaysLP3zqlWrcuvWLYM2Tk5OREdHP1p0omK7tA3WTwBtNrQYDcO+Agu1qaMSQghRSRmVFNWpU4fr16/rn7du3Zrdu3cTF6fbwDQtLY0tW7ZQs2bNEglSVECXd8D/jdclRM2fguFfS0IkhBDCpIxKivr168fevXtJTU0F4KWXXiImJoaWLVvy1FNP0axZM4KDg5kwYUJJxioqiss7Yd1Y0GZBsydg+DeSEAkhhDA5o5KiKVOm8P333+uTopEjR/LJJ5+QnJzMhg0biIqKYvr06bz11lslGqyoAK78Bf93LyFqMhxGfAdqo8tlCSGEECWmRCtaazQa7ty5g4eHR55VaeVdaVW0rlSu7oHfngVNBjQeCk+uALXUshJCCFF6yqSidX7UajWenp4leUlRUVzbm5sQNXpcEiIhhBBmR9Y+i9IX/HduQtRwMDz5kyREQgghzE6ReoosLCyMGg5TqVRkZ2cX+3WiAgk5AGufhux0aDAQnloJltamjkoIIYTIo0hJUffu3fMkRXfv3uXs2bOo1Wp8fX3x9PQkOjqamzdvotFoaNGiBVWrVi2VoEU5EXoIfh2tS4jq94dRP0tCJIQQwmwVKSnav3+/wfNbt27RpUsXnn32WRYsWGBQjygsLIyZM2dy5MgRtm7dWqLBCjMUfxNSY/MejwyCHe/cS4j6wehfwNKmzMMTQgghisqo1WdPP/00169f59ixYwW26dSpE35+fqxdu/aRAjQXsvosH/E34cu2kJ1RcBuVBUz9B9zrlV1cQgghxD3F+fw2aqL1nj17eOyxxwpt07t3b/bs2WPM5UV5kRpbeEIEul3vM5PLJh4hhBDiERiVFKWnpxMZGVlom4iICINNYoUQQgghzJlRSVHbtm357bffCAgIyPf80aNHWbduHe3bt3+k4IQQQgghyopRxRs/+ugjHnvsMbp168aQIUPo2rUrHh4exMTEcOjQIbZu3YqlpSXz588v6XiFOSm5YuhCCCGEyRmVFHXt2pXt27czefJkNm3axKZNm1CpVOTM2fbz8+O7776jS5cuJRqsMCPXj8DW100dhRBCCFFijN7m47HHHuPatWscPnyYM2fOkJCQgLOzMy1btqRr164Vbu8zcU9cCOyeBRe3mDoSIYQQokQ90t5nKpWKbt260a1bt5KKR5irtHg4+Akc/1a3w73KQreH2cXNpo5MCCGEKBEluiGsqIA02RD4E+xfmFuksW5v6PcR2DjC1V2FL8u3tAF7t7KJVQghhHgERidFt2/f5qeffuKff/4hPj4ejUaTp41KpWLv3r2PFKAwoat74K/34PYl3XP3htD/I6jXB3KGR18JzL+idQ57N3DxLf1YhRBCiEdkVFJ09uxZevfuzd27dymsILbMKyqnYi7pkqFr94pv2rlCr3eh7YS8u9u7+ErSI4QQokIwKin673//S1xcHO+//z6TJk2iRo0aqNXqko5NlLWUO7phspM/gaIBCyvo+BJ0fxPsZHNfIYQQFZtRSVFAQADDhw9n3rx5JR2PMIXsDN0E6oNLICNBd6zR49B3HrjVNW1sQgghRBkxKimytrambl35sCz3FEW3tH73LLgbqjvm1QL6LwA/WVEohBCicjEqKerduzcnT54s6VhEWYo4DbvegxtHdM+reMJjs6DlM2AhQ6FCCCEqH6P2Pvvkk084f/48S5YsKel4RGlLjIQ/XobveukSIktb6P42vHoKWo+RhEgIIUSlpVIKWz5WgOeff57Q0FAOHjyIn58fLVu2xNnZOe/FVSp+/PHHEgnU1BITE3F2diYhIQEnJydTh1N8malw9As4shSyUnXHmo+CPrPBuYZJQxNCCCFKS3E+v41KiiwsitbBpFKp8q1fVB6V26RIq4V/18PeuZAYrjvm21E3b6hGO9PGJoQQQpSy4nx+GzWnKDQ01KjARBkLOwY7Z0LEKd1z55rQdw40HZlbfFEIIYQQgJFJUa1atUo6DlGS7l6H3bPhwp+659ZVoNt06DQVrGxNGZkQQghhtkpk77O4uDhSUlLw9ZXKxiaVngiHPoVjy0GTqdu0tfVY6P0+VPEwdXRCCCGEWTNq9RlAQkICr7/+Op6enlSrVg0/Pz/9uePHjzNo0CACAwNLJEjxEJpsOLkCPm+tm0ityQS/HvDSIRj6uSREQgghRBEY1VMUFxdH586duXLlCm3atKFatWpcvHhRf75FixYcOXKENWvW0LZt2xILVuQjeJ+u3lDMBd1zt3rQbz40GCDzhoQQQohiMKqnaM6cOVy5coW1a9dy8uRJnnrqKYPzdnZ29OjRg3379pVIkCIft6/AmlHwywhdQmTrAgMWw3+OQcOBkhAJIYQQxWRUT9HmzZt5/PHHGT16dIFtatWqxdGjR40OTBQgNQ72L4KTP4I2Gywsof2L0ONtsHc1dXRCCCFEuWVUUhQZGcnTTz9daBtbW1tSUlKMCkrkIzsT/vkeDiyG9HubtjYYCP0+BPf6po1NCCGEqACMSorc3Ny4efNmoW0uXbqEt7e3UUGJ+ygKXN4Of30AccG6Y57NoP9HUKenSUMTQgghKhKjkqLu3buzefNmwsPDqV69ep7zFy5cYOfOnUycOPGRA6yw4m9CamzB5+3dIO0u7HoXrh/SHXPw0C2vlz3KhBBCiBJnVFL03nvvsWnTJrp06cKCBQu4c+cOABcvXuTo0aO899572NjY8NZbb5VosBVG/E34si1kZxTcRmUBilb3WG0D/lN1BRhtHMsmRiGEEKKSMSopat68OevWrWPcuHGMHTsWAEVRaNasGYqi4OjoyP/93/9Rv77MdclXamzhCRHkJkRNR0KfOVBVqogLIYQQpcnoitZDhw4lJCSEn3/+mePHjxMXF4eTkxMdO3Zk4sSJuLu7l2ScldOw5dD6OVNHIYQQQlQKj7TNh6urK2+88UZJxSIe5NnU1BEIIYQQlYbR23wIIYQQQlQkj9RTdObMGc6cOUN4eDhZWVl5zqtUKj744INHeQshhBBCiDJhVFJ0+/ZtxowZw549ewDdJOv8SFIkhBBCiPLCqKRo6tSp7N69m0GDBvH000/j7e2NpeUjdToJIYQQQpiUUZnMzp076dWrF1u3bi3peCoHezewtCl8Wb6lja6dEEIIIcqEUUmRlZUVbdu2LelYKg8XX3gl8OEVrV18yy4mIYQQopIzepuPoKCgEg6lknHxlaRHCCGEMCNGLclftGgRgYGBfPnllyUdjxBCCCGESaiUgpaOPcSFCxfo2rUr7u7utGjRAicnp7wXV6n48ccfHzlIc5CYmIizszMJCQn53qsQQgghzE9xPr+NSopCQ0Pp168fwcHBhV9cpUKj0RT38mZJkiIhhBCi/CnO57dRc4peffVVgoODmTJlCs8++6wsyRdCCCFEuWdUJnPw4EGGDBnC8uXLSzoeIYQQQgiTMGqitY2NDQ0aNCjpWIQQQgghTMaopKhv374cPXq0pGMRQgghhDAZo5KiJUuWEBkZyVtvvUVaWlpJxySEEEIIUeaMWn3Wu3dv7t69y9mzZ6lSpQr169cvcEn+3r17SyRQU5PVZ0IIIUT5U+pL8i0sitbBJEvyhRBCCGFKpb4kX6vVGhWYEEIIIYS5MmpOUWlbvnw5fn5+2Nra0rZtWw4dOlRg240bN9K3b1+qVauGk5MT/v7+7Nq1K0+7DRs20KRJE2xsbGjSpAl//PFHad6CEEIIIcoZs0uK1q1bx7Rp03jvvfc4ffo03bp1Y+DAgYSFheXb/uDBg/Tt25ft27cTGBhIr169GDJkCKdPn9a3CQgIYPTo0YwdO5YzZ84wduxYRo0axfHjx8vqtoQQQghh5oze+0xRFDZv3syZM2cIDw8nKysr78WN2PusY8eOtGnThq+//lp/rHHjxgwfPpyFCxcW6RpNmzZl9OjRzJo1C4DRo0eTmJjIjh079G0GDBhA1apVWbt2bb7XyMjIICMjQ/88MTERX19fmVMkhBBClCOlPqcoODiYIUOGcPnyZQrLqYqbFGVmZhIYGMiMGTMMjvfr16/IdZG0Wi1JSUm4urrqjwUEBPDGG28YtOvfvz9Lly4t8DoLFy5k7ty5RY5dCCGEEOWbUUnR1KlTuXTpEi+//DLPPPNMie19dufOHTQaDZ6engbHPT09iYqKKtI1Pv30U1JSUhg1apT+WFRUVLGvOXPmTKZPn65/ntNTJIQQQoiKyahM5tChQwwdOpSvvvqqpOMBdD1M91MUJc+x/Kxdu5Y5c+awadMmPDw8HumaNjY22NjYFCNqIYQQQpRnRiVFVapUoV69eiUdC+7u7qjV6jw9ODExMXl6eh60bt06Jk2axPr16+nTp4/BOS8vL6OuKYQQQojKw6jVZwMGDCiVvc+sra1p27Ytu3fvNji+e/duOnfuXODr1q5dy4QJE/j1118ZPHhwnvP+/v55rvnXX38Vek0hhBBCVC5GJUUff/xxqe19Nn36dH744QdWrFjBxYsXeeONNwgLC2PKlCmAbq7PuHHj9O3Xrl3LuHHj+PTTT+nUqRNRUVFERUWRkJCgb/P666/z119/sXjxYi5dusTixYvZs2cP06ZNK9HYhRBCCFF+Gb0k//Lly/j7+6PRaEp877Ply5frE69mzZrxv//9j+7duwMwYcIErl+/zv79+wHo2bMnBw4cyHON8ePHs3LlSv3z33//nffff5+QkBDq1q3LRx99xMiRI4sck2zzIYQQQpQ/pb732enTp+nbty9xcXGFX1z2PhNCCCGECRXn89uo4bNp06YRHx/PokWLCAsLIysrC61Wm+eroiREQgghhKj4jFp9FhgYyOjRo3n77bdLOh4hhBBCCJMwqqfIyclJlrMLIYQQokIxKikaNmwY+/btQ6vVlnQ8QgghhBAmYVRStHjxYmxsbHjuuee4detWScckhBBCCFHmjFp9VqdOHTIzM4mMjATAxcUFZ2fnvBdXqQgODn70KM2ArD4TQgghyp/ifH4bNdFaq9ViZWVFzZo19cfyy62MLIEkhBBCCFHmjEqKrl+/XsJhCCGEEEKYllFzioQQQgghKhqjeorul52dzZUrV/RjdQ0bNsTS8pEvK4QQQghRpozuKbp79y6TJ0/GxcWF5s2b07VrV1q0aIGLiwuTJ08mNja2JOMUQgghhChVRnXp3L17F39/f65cuYKbmxvdunXDy8uL6OhoTp48yQ8//MCBAwcICAjA1dW1pGMWQgghhChxRvUUffjhh1y5coWZM2dy48YNduzYwU8//cT27du5ceMG7733HlevXmX+/PklHa8QQgghRKkwuk6Rn58fe/fuLbBNnz59CAkJISQk5JECNBdSp0gIIYQof4rz+W1UT1FERASdOnUqtE3Hjh2JiIgw5vJCCCGEEGXOqKTI2dmZGzduFNrmxo0b+Va5FkIIIYQwR0YlRT179mT9+vXs2bMn3/N79+5l/fr19OzZ81FiE0IIIYQoM0bNKbpw4QIdOnQgLS2NQYMG0aNHDzw9PYmOjmb//v3s2LEDe3t7jh07RtOmTUsj7jInc4qEEEKI8qc4n99GJUUAR48eZcKECVy7dk13IZVKv9dZ3bp1WblyJV26dDHm0mZJkiIhhBCi/Cn1DWEBOnfuzOXLlzly5AinT58mMTERJycnWrduTZcuXVCpVMZeWgghhBCizBndU1TZSE+REEIIUf6U+pL8hIQEzp49S2pqar7nU1JSOHv2LImJicZcXgghhBCizBmVFM2bN4/OnTuj0WjyPa/RaOjSpQsfffTRIwUnhBBClJaAiACG/TmMgIgAU4cizIRRSdHOnTvp168fjo6O+Z53cnKif//+bN++/ZGCE0IIIUqDoigsO7WMkIQQlp1ahswkEWBkUhQWFkb9+vULbVO3bl3CwsKMCkoIIYQoTUcjjnI+9jwA52PPczTiqIkjEubQc2fU6jOVSkVGRkahbTIyMgocXhNCCCHKmkarISI5gmvx1/jo+EeoUKGgoELF3IC5vNHmDZxtnXGxccHZRvfd3tJeVlOXgQd77jp5dzLJn7tRSVHjxo3ZuXMniqLkG7RWq2XHjh00bNjwkQMUQgghiiNbm82tpFsEJwQTEh9CcEIwwfHBhCaEkqHJ+wu9gkJkSiRvH3o7zzlLC0ucrZ1xtrnvyzo3cbr/y8XGRd/WztKuzD7UAyICWHRiETM6zMDfx79M3rM4FEUhXZNOSlYKqVmppGankpKVonuenUpqVirn7pzL03PXpXrZ1zo0Kil69tlneeONN3j++edZunSpwR5nCQkJvP7661y7do0lS5aUWKBCCCHE/bK0WdxMuklwvC7pyUmAridcJ1Obme9rrC2ssVBZkK5Jz3PO3tIenyo+JGYkEp8RT6Y2k2xtNrHpscSmxxYrNisLK33i5GTtZND75GST+/z+hMvFxgVbS9tivU9p9LBoFS1p2Wn6JCYl+14yk3UvmXngeU6SU1DCk5qdilbRFvn9LVQWfHH6Czr7dC7z3iKj6hRlZWXRp08fDh06hIuLC+3bt6d69eqEh4fzzz//EB8fT/fu3dm9ezdWVlalEXeZkzpFQohHYe6/zZuzLE0WNxJv6Ht8guODCUkI4XridbK12fm+xs7SDj9nP+o616WOSx3qOtelrktdbiTe4D97/1Pge33T5xt9D0VadhoJGQm5X5kJxGfEGxzLeZ6YqUuk4jPiC4ypKGzUNoX2SOX0RuUkVlfuXmHGoRn618/2n03Dqg0NkpOcBOf+RKWwZCYtO83o+B/G3tIeBysH7K3ssbe0x97KngxNBufunMvT9v6/i0dRJtt8ZGRk8MEHH/Ddd98Z1CNycnLipZdeYt68edjY2BhzabMkSZEQwliKovDMtmc4H3uepm5NWTt4rcxTyUeGJoPrCdcJSQjRJz/BCcGEJYahUfKfo2pnaadPfOq51KOuS13qONfBp4oPFirDtUQ5fw8XYi+gkPejT4WKJm5NHunvR1EUfTIVnxFPQmZCvklUTpKVcywxI5FsxfhkqjRYqCxwsHTAzsoOBysHHCzvJTNW9xKbBxIcBysH/fEH2zhYOWBraVvg38nFuIsGvUkWKgsauzYukX8rZbLNh42NDR9//DGLFi3i0qVLxMfH4+LiQsOGDVGr1cZeVgghKpz8VjqZYr6EuUjPTic0ITR3zs+95Odm0s0Ch1mqWFUx6PGp46xLgjwdPPN80BYkS5tFVEpUvgkR6OYWRaVEkaXNwlptbdS9qVQqfeLgXcW7yK9TFIWUrBSD3qicYTx94pSZaJBU3Um7Q3JWcp5rudq44mrnmidZsbO0M0hcHpbQ2KhtSj15v//fxv20itYk/1Zkm48ikp4iIURxaRUtoQmhTN07lYjkCP2HcVWbqjxe53EcrO99CN37ILK3tMfO0g57q9zvOeds1bZm0btUnGHA1KxUffJz/5yfW0m3CkxMHK0dqedSjzrOdajrUlffC+Rp71ki9x+VEkVcelyB511tXfFy8Hrk9yltZdHDUtrKoucOyqinCCAqKoqNGzdy6dIlUlNT+eGHHwC4ffs2oaGhNG/eHDs7u0d5CyGEKBc0Wg3XE69zIfaC/utS3CVSs/Nuh3Q34y6/XPylWNdXoTJIku5PoO4/pk+o8juW8/y+Y5YWRf8YKGhSb0pWCiHxIVyLv6Yf+gpJCCE8ObzAaznbOOt7ffRfznVxt3Mv1Q9zLwevcpH0PIy59bAYoyx67orL6J6i5cuX89///ldfr0ilUunrEp0/f54WLVrwzTff8OKLL5ZctCYkPUVCiBwarYbQhFAuxBkmQPlNUM2phfPgMTc7Nx7zfYw0TRpp2Wn6ya4Pfi/NSa+gm9ibX+KUXxIVkxLDhmsb9K9t4tqEuIw4olKiCry+q62rfrirrktdfS+Qq62r2fdkmKuy6mEpC2XRc1fqPUVbtmzhlVdeoV27dsyaNYsdO3bwzTff6M83bdqUFi1a8Oeff1aYpEgIUTlla7N1CdB9PUCX717ON1mxs7SjkWsjmrg1oYlbEzKyM5h3bF6edgoKd9Lu0Ktmr4f+Nq9VtKRnp+ebMN3/PS07zfBxzrkC2udMXM7QZJChyeBuxt1i/9lciLugf+xu567v7bk/CapqW7XY1xWFM8ceFmOZW8+dUUnRJ598Qs2aNfn7779xcHAgMDAwT5vmzZtz6NChRw5QCCHKSrY2m5CEEMMEKO5yvjVt7CztaOzaWJ8ANXFrQm2n2qgtdAtNcn6bz6+nCHS/zRelFouFykI/cZcSmo2gKApZ2qy8yVJ2KmlZaXmSqbSsNIITgjl462Cea73Z7k2G1xuOs41zPu8kSoO12prfHv/toT0s5p4QmSOjkqKgoCDGjh2Lg4NDgW2qV69OdHS00YEJIURpytZmExwfnJsAxV3gStyVAov6NXZrrE+Cmro1pZZTLX0ClB9z/m1epVJhrbbGWm2NCy4PbZ+T4FmoLPJM6t0RuoNxTcaVYrQiP+bWw1JRGJUUabXahxZlvH37doWqUySEKL+ytFkGCdDF2Itcvns53y0fHKwc8vQA1XKqVeRl3zkq0m/zFWFSrxBFYVRS1LBhQw4fPlzg+ezsbA4cOEDz5s2NDkwIIaD4laCzNFlci79mMAR25e6VfLd9qGJVhcZujWnimpsA1XSqWewEqCAV4bd5RVH44vQXjzwMKER5YFRS9Nxzz/Hmm28yf/583n//fYNzGo2GN998k5CQEN55550SCVIIUXwVYVuJh+3rlKXJ4mr81TwJUJY2K8+1HK0cdQnQfT1Avo6+JZYAVVTmPAwoREkzeu+zfv36cfDgQerVq4eNjQ3nz5/niSee4OTJk1y/fp1+/fqxY8eOCvObgyzJF+VJRdlW4kj4EabsmaJ/PrPDTKzUVvoE6Ordq/knQNaOBr0/TdyaUMOxhiRARqooBQ9F5VQme59lZmYyd+5cvvnmG+7ezV3K6eTkxMsvv8zcuXOxtq44vzVIUiTKA41WQ2x6LH9d/4vF/yzWH3+20bPUca6DFi1aRYuiKCgoeR/fd0zLve/3juf3OKdNftfKmZCb81j/unsxoFBoPBqthqDbQfluY3A/J2sngx6gpq5NqeFYo1wmgUKIklcmSVEORVG4fPkycXFxODk50bhx4wq595kkRcLUtIqWuPQ4olOiiUqJIio1Ks/jmNQYs9tUsqQ1dm1MZ5/O+iSoepXqkgAJIQpUZtt8gG5pZ6NGjfIcDw0NZe7cuaxcufJR30KICk9RFF3Ck3ovyUmJyvM4JjUm36GiB1lggZa8m2q29mhNNbtqqFQqLLDQfVdZoEKlf2zw/F4bFbnngLztHrwGFqDSxWGhKuI1HowHFd/9+53BfmH3v+71Nq9LIiSEKHGPnBQ9KCwsjA8//JBVq1aRnZ0tSZEod0p6grKiKCRkJBCVei/BSYnWP85JeKJTovNdHfUgC5UF7nbueNl74engiZeDF572nvpVTh52Hryx/418N4nM1GSypMeScpFMHAk/ku++WbIEXAhRmoqVFB0+fJgPPviAwMBALC0t6datGx9//DENGzYkNTWV999/n+XLl5OZmYmPjw8zZ84srbiFKBUPW+2UX/vEzMR8e3Zykp/olOh8CwI+KGc/LC97XYLj6eCpf5yT/Ljbu2NlUXCNsCPhR8p9PRlZAi6EMJUiJ0WBgYH06dOHzMzc32a3bNnCP//8w8GDBxk+fDgXLlzAx8eHd955h8mTJ0vxRlHu3F+k7nzsefaE7aG2U+18k56c70XdsDNnhc79vTz3P/aw88BKXXhR1MJUlGRCloALIUylyEnRxx9/TGZmJgsXLmTSpEkAfPPNN8yaNYtu3bpx+/Zt3n//fd59911sbW1LLWAhSotGq2HRiUUGx6bvn16k11a1qarvzclvWMvT3rPUP8ArSjJRkSpBCyHKlyKvPqtRowaNGjViz549Bsd79erFwYMH+eSTT5g+vWgfIOWRrD6ruG6n3mZT8CbWXFzDnbQ7ec7bW9lTo0qNfHt5vBy88LD3wNbSPH4RkHoyQghhqFRWn8XExPDcc8/lOd6+fXsOHjzI+PHjix+pECai0Wo4EnGEDVc2cODWATSKJt92FioL/Jz8yk3xw4qwrYQQQphKkZOi7OxsHBwc8hzPOebm5lZyUQlRSsKTw/nj6h/8ce0PYlJj9MfrONchJCEkT/vyNEFZCCHEoynxJflCmJssTRb7bu5j49WNBEQE6OfcuNi4MKTuEEbWG8n7R94v9xOUhRBCPJpiJUWrV6/m2LFjBseuXbsGwKBBg/K0V6lUbNu27RHCE8J4IQkhbLyykc3Bm7mbkbsVTSfvTjxR/wl61+yNtdqaTE1mhZigLIQQ4tEUeaK1hUXxN1JUqVRoNPnP1ShvZKJ1+ZCWncZf1/9i49WNnIo5pT/uYefBsHrDGFF/BL6OvnleJxOUhRCiYiqVidahoaGPHJgQpeVi7EU2XN3AtpBt+g1E1So13Wp044n6T9C1elcsLQr+cZcJykIIIYqcFNWqVas04xCi2JIyk9gesp0NVzdwMe6i/nj1KtV5ov4TDKs3DA97DxNGKIQQojyRidaiXFEUhaDbQfx+5Xf+uv6XfvsMKwsr+tTsw8gGI+ng1UG/8agQQghRVJIUiXIhLj2OLcFb2Hh1o8HS+brOdXmiwRMMqTMEF1sX0wUohBCi3JOkSJgtraLlWOQxNlzZwL6b+8jWZgNgZ2nHgNoDGFl/JC2rtZRl8kIIIUqEJEXC7ESlRLHp2ib+uPYH4cnh+uPN3JoxssFIBtYeSBXrKiaMUAghREUkSZEwC1naLA7dOsTGqxs5FH4IraIFwNHakcfrPM4T9Z+goWtDE0cphBCiIpOkyAwERASw6MQiZnSYgb+Pv6nDKVM3E2+y8dpG/rz2p8FmrG092/JE/SfoW6uv2Wy2KoQQomKTpMjEFEVh2allhCSEsOzUMjp5d6rwc2QyNBnsvbGXDVc3cCLqhP64q60rw+oNY2S9kdR2rm26AIUQQlRKkhSZ2NGIo5yPPQ9Q4TcevXr3KhuvbmRLyBYSMhIA3b5iXap34Yn6T9CjRg+s1FYmjlIIIURlJUmRCSmKwpKTSwyOvXngTbr6dKWKTRWqWFXBwcoh97t17vP7j9lZ2plFXZ78hgFTs1LZeX0nG65u4Ozts/q2Xg5ejKw3kuH1huNdxdtUIQshhBB6ZpkULV++nE8++YTIyEiaNm3K0qVL6datW75tIyMj+e9//0tgYCBXr17ltddeY+nSpQZtVq5cycSJE/O8Ni0tDVtb081XORpxlGvx1wyOJWcls/PGzmJdR4UKByuH3ATKOjeRMkiqHjh3f2JVxaoK9lb2WFkY11Pz4DCgg6UDG69tZEfoDlKzUwGwVFnS07cnTzR4An9vf9QWaqPeSwghhCgNZpcUrVu3jmnTprF8+XK6dOnCt99+y8CBA7lw4QI1a9bM0z4jI4Nq1arx3nvv8b///a/A6zo5OXH58mWDY6ZMiBRF4YvTX2ChstCvtAJdguNp78mI+iNIzUolOSuZlKwU/fecr+SsZJIzk9EoGhQU3fOsZKKJfqS4bNW2eXql7K3sDRKp+8/lfL8Wf81gGPC5Hc/pr1nbqTYj649kSN0huNu5P1J8QgghRGkxu6Tos88+Y9KkSbzwwgsALF26lF27dvH111+zcOHCPO1r167NsmXLAFixYkWB11WpVHh5mc+Gn/fPJbqfgkJUahQtq7V86NwiRVHI0GQYJk6ZKfkmUsmZhsf03+8dz9kuI12TTromndj02Ee6PxUqBvkN4skGT9LWs22FnzwuhBCi/DOrpCgzM5PAwEBmzJhhcLxfv34cPXr0ka6dnJxMrVq10Gg0tGrVig8//JDWrVsX2D4jI4OMjAz988TExEd6//vl9BKpUKGg5DmvQsUXp7+gs0/nQpMJlUqFraUttpa2j9wDk6XN0vdMJWcmk5qdmm8ild+x2ym3iUyNNLxHFIbUHUI7r3aPFJcQQghRVswqKbpz5w4ajQZPT0+D456enkRFRRl93UaNGrFy5UqaN29OYmIiy5Yto0uXLpw5c4b69evn+5qFCxcyd+5co9+zMFnaLKJSovJNiOBeb1FKFFnaLKzV1qUSw4OsLKxwtnHG2ca5WK9TFIVntj1DdFq0wTCghcqiSImdEEIIYS7MKinK8eCHqKIoj/TB2qlTJzp16qR/3qVLF9q0acMXX3zB559/nu9rZs6cyfTp0/XPExMT8fX1NTqG+1mrrfnt8d+IS48rsI2rrWuZJUSPoqBhQK2irfAlBoQQQlQsZpUUubu7o1ar8/QKxcTE5Ok9ehQWFha0b9+eq1evFtjGxsYGGxubEnvPB3k5eOHlYD5znIxRUsOAQgghhDkwfXGb+1hbW9O2bVt2795tcHz37t107ty5xN5HURSCgoLw9pb6OI+iOMOAQgghhLkzq54igOnTpzN27FjatWuHv78/3333HWFhYUyZMgXQDWuFh4ezatUq/WuCgoIA3WTq27dvExQUhLW1NU2aNAFg7ty5dOrUifr165OYmMjnn39OUFAQX331VZnfX0VSkYYBhRBCCLNLikaPHk1sbCzz5s0jMjKSZs2asX37dmrVqgXoijWGhYUZvOb+VWSBgYH8+uuv1KpVi+vXrwMQHx/P5MmTiYqKwtnZmdatW3Pw4EE6dOhQZvdVUVWEYUAhhBACQKUoSv5jH8JAYmIizs7OJCQk4OTkZOpwhBBCCFEExfn8Nqs5RUIIIYQQpiJJkRBCCCEEkhQJIYQQQgCSFAkhhBBCAJIUCSGEEEIAkhQJIYQQQgCSFAkhhBBCAJIUCSGEEEIAZljR2lzl1LhMTEw0cSRCCCGEKKqcz+2i1KqWpKiIkpKSAPD19TVxJEIIIYQorqSkJJydnQttI9t8FJFWqyUiIgJHR0dUKlWJXjsxMRFfX19u3rxZrrcQkfswL3If5kXuw/xUlHuR+yicoigkJSXh4+ODhUXhs4akp6iILCwsqFGjRqm+h5OTU7n+gc4h92Fe5D7Mi9yH+ako9yL3UbCH9RDlkInWQgghhBBIUiSEEEIIAUhSZBZsbGyYPXs2NjY2pg7lkch9mBe5D/Mi92F+Ksq9yH2UHJloLYQQQgiB9BQJIYQQQgCSFAkhhBBCAJIUCSGEEEIAkhQJIYQQQgCSFJW4gwcPMmTIEHx8fFCpVPz5558G5xVFYc6cOfj4+GBnZ0fPnj05f/68QZvp06fj6upKzZo1+e233wzO/d///R9Dhgwp1XtYuHAh7du3x9HREQ8PD4YPH87ly5fL3X0AfP3117Ro0UJfDMzf358dO3aUu/u438KFC1GpVEybNk1/rLzcx5w5c1CpVAZfXl5e5e4+AMLDwxkzZgxubm7Y29vTqlUrAgMDy9W91K5dO8/fh0qlYurUqeXmHgCys7N5//338fPzw87Ojjp16jBv3jy0Wq2+TXm5l6SkJKZNm0atWrWws7Ojc+fO/PPPP2Z9HyXxuZeRkcGrr76Ku7s7Dg4ODB06lFu3bhmcHzt2LE5OTjRs2JB9+/YZvP7jjz/m1VdfffSbUUSJ2r59u/Lee+8pGzZsUADljz/+MDi/aNEixdHRUdmwYYPy77//KqNHj1a8vb2VxMRERVEUZfPmzYqnp6fyzz//KL/++qtia2ur3LlzR1EURbl7965Sr1495caNG6V6D/3791d++ukn5dy5c0pQUJAyePBgpWbNmkpycnK5uo+cOLZt26ZcvnxZuXz5svLuu+8qVlZWyrlz58rVfeQ4ceKEUrt2baVFixbK66+/rj9eXu5j9uzZStOmTZXIyEj9V0xMTLm7j7i4OKVWrVrKhAkTlOPHjyuhoaHKnj17lGvXrpWre4mJiTH4u9i9e7cCKH///Xe5uQdFUZT58+crbm5uytatW5XQ0FBl/fr1SpUqVZSlS5fq25SXexk1apTSpEkT5cCBA8rVq1eV2bNnK05OTsqtW7fM9j4e9XNPURRlypQpSvXq1ZXdu3crp06dUnr16qW0bNlSyc7OVhRFUT7//HOlcePGyrlz55RPPvlE8fDwULRaraIoihISEqLUr19fSUhIeOR7kaSoFD34w6HVahUvLy9l0aJF+mPp6emKs7Oz8s033yiKoiiLFy9WRo8erT/v4eGhnDhxQlEURXnxxReVzz77rGyCv09MTIwCKAcOHFAUpfzeR46qVasqP/zwQ7m7j6SkJKV+/frK7t27lR49euiTovJ0H7Nnz1ZatmyZ77nydB/vvPOO0rVr1wLPl6d7ud/rr7+u1K1bV9FqteXqHgYPHqw8//zzBsdGjhypjBkzRlGU8vP3kZqaqqjVamXr1q0Gx1u2bKm899575eI+jPnci4+PV6ysrJTffvtN3yY8PFyxsLBQdu7cqSiKorz88svKO++8oyiK7s8J0P9C1b9/f2Xjxo0lEr8Mn5Wh0NBQoqKi6Nevn/6YjY0NPXr04OjRowC0bNmSkydPcvfuXQIDA0lLS6NevXocPnyYU6dO8dprr5V53AkJCQC4urqW6/vQaDT89ttvpKSk4O/vX+7uY+rUqQwePJg+ffoYHC9v93H16lV8fHzw8/Pj6aefJiQkpNzdx+bNm2nXrh1PPfUUHh4etG7dmu+//15/vjzdS47MzExWr17N888/j0qlKlf30LVrV/bu3cuVK1cAOHPmDIcPH2bQoEFA+fn7yM7ORqPRYGtra3Dczs6Ow4cPl5v7uF9RYg4MDCQrK8ugjY+PD82aNTO4r8OHD5OWlsauXbvw9vbG3d2d1atXY2try4gRI0om4BJJrUS+eCBjPnLkiAIo4eHhBu1efPFFpV+/fvrns2fPVurWras0a9ZM2bhxo5KRkaE0a9ZMOXnypPLFF18oDRo0UDp37qwfAipNWq1WGTJkiMFvxeXtPs6ePas4ODgoarVacXZ2VrZt21bu7mPt2rVKs2bNlLS0NEVRFIOeovJ0H9u3b1d+//135ezZs/oeL09PT+XOnTvl6j5sbGwUGxsbZebMmcqpU6eUb775RrG1tVV+/vlnRVHK199JjnXr1ilqtVofc3m6B61Wq8yYMUNRqVSKpaWlolKplAULFujPl6d78ff3V3r06KGEh4cr2dnZyi+//KKoVCqlQYMG5eI+jPncW7NmjWJtbZ3nWn379lUmT56sKIqiZGZmKv/5z3+U2rVrK+3atVMOHTqkxMbGKnXq1FFu3LihvPfee0rdunWVfv366YcajYrf6FeKhyrohyMiIsKg3QsvvKD079+/wOvMnj1bmTZtmnLmzBnF09NTiYmJUVasWKG0adOmtELX+89//qPUqlVLuXnzpv5YebuPjIwM5erVq8o///yjzJgxQ3F3d1fOnz9fbu4jLCxM8fDwUIKCgvTH8kuKzP0+8pOcnKx4enoqn376abm6DysrK8Xf39/g2Kuvvqp06tRJUZTy+XfSr18/5fHHH9c/L0/3sHbtWqVGjRrK2rVrlbNnzyqrVq1SXF1dlZUrV5a7e7l27ZrSvXt3BVDUarXSvn175bnnnlMaN25cLu7DmM+9gpKiPn36KC+99FKB7zV+/Hhl6dKlyqZNm5SmTZsqycnJyqxZs5SRI0caH7/RrxQP9eAPR3BwsAIop06dMmg3dOhQZdy4cfle4+LFi0q9evWUpKQkZdmyZcpTTz2lKIruwwQokYllBXnllVeUGjX+v73zjorq2v74d2RmAIGhiQoSRkDEAjYELAQQDCj2Borx2aIo9vDgSYxijd2lJvFZng+jsWAhGjXSFLGiaEQRgYDgw6cgCCJNEWH//nDN/XGdS5VIeJ7PWrOUfc6cs/e5+9y757RrTOnp6Tx5c7PjfVxdXWnmzJnNxo5ffvmFu0EqPgBIJBKRiooKpaWlNQs7qmPgwIE0a9asZnM9iIhMTExo+vTpPNmOHTvIyMiIiJpfH3n06BG1aNGCTp48ycmakw3Gxsb0ww8/8GSrVq0iS0tLImpetigoLi7mAglPT0/y8PBoFnY05Ll3/vx5AkD5+fm8PN26daNly5YJ1nP+/HmytbWlt2/f0qJFi8jf35+IiO7fv096enoN1p+tKfqImJqaom3btoiMjORkb968QUxMDPr166eUn4gwc+ZMbN68GZqamqioqEB5eTkAcP9W3XLaWBAR5s6di9DQUFy4cAGmpqbN0o7qICKUlZU1GztcXV2RkJCA+Ph47tO7d29MnDgR8fHxMDMzaxZ2CFFWVoakpCQYGho2m+sBAP3791c6puKPP/6AXC4H0Pz6SHBwMFq3bo0hQ4ZwsuZkQ2lpKVq04D/OVFRUuPqaky0KNDQ0YGhoiBcvXiA8PBwjRoxolnbURWcbGxtIJBJenqysLNy/f1/QrtevX2POnDnYtWsXVFRUlOyqqKhouMINDqcYghQVFdGdO3fozp07BIC2bNlCd+7c4bZArlu3jrS1tSk0NJQSEhJowoQJSlsTFezatYvGjBnD/X3jxg2SyWR0/fp1WrZsGXXp0uVPsWH27Nmkra1NFy9e5G3XLS0t5fI0BzuIiAIDA+nSpUuUkZFB9+7do2+++YZatGhBERERzcqO96k6fUbUfOzw8/OjixcvUnp6OsXGxtLQoUNJS0uLHj161KzsuHnzJonFYlqzZg2lpqbSwYMHqWXLlvTzzz9zeZqLLRUVFWRiYsLt7KlKc7Fh8uTJ1K5dO25LfmhoKLVq1YoCAgKanS1hYWF07tw5Sk9Pp4iICOrevTvZ2dnRmzdv/rJ2NMZzb9asWWRsbExRUVH0+++/k4uLC29LflUCAwPJz8+P+zskJIRMTEzo7t27NH36dPLw8GiwLSwoamSio6MJgNJn8uTJRPRuQWBQUBC1bduWVFVVydHRkRISEpTKyc7OJrlcrrQ4bcWKFaSnp0edOnWiGzdu/Ck2COkPgIKDg7k8zcEOIqJp06aRXC4nqVRKBgYG5OrqygVEzcmO93k/KGoudijOJ5FIJGRkZESjR4+mxMTEZmcHEdHp06fJysqKVFVVqVOnTrR7925eenOxJTw8nABQSkqKUlpzsaGwsJAWLFhAJiYmpKamRmZmZrRkyRIqKytrdraEhISQmZkZSaVSatu2Lc2ZM4cKCgr+0nY0xnPv1atXNHfuXNLT0yN1dXUaOnQoZWZmKtWVkJBAHTp04J2bV1FRQbNnzyaZTEa2traUmpraYFtEREQNH2diMBgMBoPB+N+ArSliMBgMBoPBAAuKGAwGg8FgMACwoIjBYDAYDAYDAAuKGAwGg8FgMACwoIjBYDAYDAYDAAuKGAwGg8FgMACwoIjBYDAYDAYDAAuKGAwGg8FgMACwoIjBYNRASkoKRowYgTZt2kAkEqF9+/ZNrdKfyqNHjyASiTBlypSmVuWTY9++fRCJRNi3b1+d8otEIjg7O9e5/ClTpkAkEuHRo0cN0u9jcvHiRYhEIixfvrypVfnkYEERo1oUD4iqH6lUis8++wze3t64d+9eU6v4UWhON9PGpKKiAqNGjUJ4eDiGDx+OoKAgLFy4sKnVahYsX74cIpEIFy9ebGpVPggWJL6DtcOng7ipFWD89TE3N8eXX34JACguLkZsbCwOHz6M0NBQXLhwQfAtxozmT0ZGBpKSkuDj44OdO3c2tTofhXbt2iEpKQna2tpNrQqjkVm7di0WL16Mdu3aNbUqtWJnZ4ekpCS0atWqqVX55GBBEaNWOnTooDSM++2332LNmjVYsmQJoqOjm0Yxxp/K06dPAQBt27ZtYk0+HhKJBJ06dWpqNRh/AoaGhjA0NGxqNepEy5YtmR82FQ1+lSzjf56MjAwCQO7u7kpp2dnZBIA0NDR48pMnT5KLiwvp6OiQqqoqde3alTZu3Ehv377l5QsODiYAFBwcTGfOnCEHBwfS1NQkuVzO5SkrK6OtW7eSra0taWpqkoaGBnXu3JkWLVpE+fn5vPKePXtGCxcuJHNzc5JKpaSvr0+jR48WfHu0XC4nuVxOxcXFtGjRIjIyMiKpVErW1tZ07NgxpbwQePuzk5MTlyc0NJTGjx9P5ubmpK6uTjKZjBwcHOj48ePVtu3OnTupS5cupKqqSsbGxuTv70+vXr1SKltBYWEhLVu2jLp06UJqamqkra1N7u7udPny5WrrEOL58+e0cOFCat++PUmlUjIwMCBPT0/em+prsjs4OLjG8p2cnKi628rkyZMJAGVkZHCyiooK2rNnD9na2pKuri6pq6uTXC6nESNGUExMjFIZMTExNHToUNLX1yepVEodOnSgJUuWUElJCS+f4q3dQUFBdO3aNXJzcyNtbe1qdVOg8HnF273ft6u8vJxWrlzJtZ+FhQX9+OOPgnnf/1T1baKG+eyLFy9o3rx5ZGxsTCoqKtz1uHXrFs2ZM4e6du1KMpmM1NTUyMrKitauXUtv3rwRtPXZs2fk5+dHHTt2JFVVVdLV1SV7e3vatGkTEf1/HxX6REdHc+VUVlbS3r17qV+/fqSlpUXq6upkY2NDe/fuFaw3Ly+PfHx8qHXr1qSurk69e/em0NBQ3j2hLij6Sk5ODk2dOpUMDAxITU2N7O3tefopEPI/IqLjx4+To6MjGRgYcP3R3d2dfvnll3q1Q0lJCQUFBZGlpSXXnh4eHnT16lUlXYKCgrjv79u3j3r16kXq6upc36/qv+9TH7/5448/aMqUKdS+fXtSVVUlfX196tmzJ3399dd1auNPETZSxGgQIpFISfbNN99g7dq1MDY2xpgxYyCTyXDp0iX4+/vjxo0bOHbsmNJ3jh07hoiICAwdOhS+vr4oKioCALx+/Rru7u64dOkSLCwsMHXqVKiqqiI1NRU7d+7E3/72N+jq6gIAHj58CGdnZzx58gRubm4YOXIkcnJycOLECYSHh+P8+fOwt7fn1VteXg43Nzfk5+dj9OjRKC0txZEjR+Dp6YmwsDC4ubkBABYuXIh9+/bh7t27WLBgAXR0dACAt+A4MDAQUqkUDg4OMDQ0RG5uLn799VeMHTsW27dvx7x583h1L1u2DKtWrYKhoSFmzpwJsViMY8eOITk5WbCt8/Pz4ejoiMTERHz++edwd3fHy5cvcerUKQwYMADHjh3DyJEja71meXl56NOnD9LS0uDs7Izx48fj0aNHOH78OM6ePYvIyEj07duXszs+Ph4//fQTnJycuAWtPXr0qLWe+hAYGIgNGzbA3Nwc3t7e0NLSwpMnT3D58mVcuHABjo6OXN6dO3fC19cXurq6GDZsGAwMDBAXF4c1a9YgOjoa0dHRkEqlvPKvXbuG7777DgMGDMDMmTORmZn5QfpOmDABN27cwODBg6GiooKjR49izpw5kEgkmDFjBgBw605iYmIwefJkzlcUvgM0zGfLysrg4uKCoqIiDBs2DFKpFG3atAEA7NmzB6dPn4ajoyM8PDxQWlqKixcvIjAwEHFxcThx4gSvrNTUVAwYMABPnjyBg4MDRo4ciZKSEty/fx9r1qyBn58fevTogQULFmDbtm3o3r07z8cUNhERvvzySxw6dAgdO3aEt7c3pFIpIiMjMX36dDx48ACbNm3ivldaWgpnZ2ckJCSgb9++cHJywuPHj+Hl5cX1ufpQUFCA/v37QyaTYeLEicjJyUFISAjc3d1x+/ZtWFlZ1fj9f/7zn/D19YWhoSFGjRoFfX19ZGVl4ebNmzh58iRGjhxZp3YoKyuDq6srYmNj0atXLyxcuJDTJSIiAiEhIRg9erRS/Rs3bkR0dDSGDx+OL774AmJxzY/k+vjN06dPYWdnh5KSEgwZMgReXl4oLi5Gamoqvv/+e2zevLl+jf2p0NRRGeOvS00jRUuWLCEA5OzsTEREERERBIAGDx7M+9VeWVlJs2bNIgC8kRPFry+RSESRkZFK5fv7+xMAmjRpktIoU0FBARUVFXF/9+vXj8RiMUVERPDypaSkkJaWFllbW/PkilGQESNGUFlZGSePiooStLe6X5gKHj58qCQrKioia2tr0tbW5rVHSkoKqaiokImJCT1//pyXv2vXroIjRd7e3gSA/v3vf/Pk2dnZ9Nlnn5GBgQG9evVKULeqTJs2jQBQYGAgTx4WFkYAyMLCgioqKjh5Tb9Wq6O+I0V6enrUrl07pZGeyspKysvL4/5OTEwksVhMPXv25MmJiNauXUsAuBGOqroDqHbEQojaRors7e3p5cuXnDw5OZnEYjFZWlry8lcdCRCioT7r5uZGpaWlSuU9evRIqZ9UVlZy1/zKlSu8NDs7OwJAu3fvVirr8ePH3P+raw8Fu3fvJgA0ffp0Ki8v5+RlZWU0bNgwAkC3bt3i5Ip2mTFjBq+c8PDwOo9GKlDk9/X15fntv/71LwJAPj4+vPxC/terVy+SSqWUk5OjVH7V/llbO6xcuZIA0MSJE6myspKT3717lxs1Kiws5OSKdtDQ0KB79+4plVdd36uP32zfvp0A0LZt25TKz83NFbSDQcSCIka1KG4E5ubmFBQUREFBQeTn50f9+/cnAKSmpkbXrl0jIqLhw4cTAMrMzFQqp6CggEQiEY0ZM4aTKYKiUaNGKeV/+/YtyWQy0tbWVpome5/ff/+duykL8fXXXxMA3tCy4gGTnp6ulF8ul5Oenh5PVltQVB2bN28mAHTx4kVOtnz58mpvVIcPH1YKinJzc0lFRYVcXV0F61Dc+E6fPl2jLmVlZaSurk76+vpKAQgRkbu7OwHgTcd9rKDI1NSUF5wKMX/+fCX9FFRUVJCBgQHZ2Ngo6d6zZ886605Ue1B04cIFpe8o0oQeekJB0Yf47N27d+tlz+3btwkALV++nJPdvHmTAJCjo2Ot368tGOjWrRtpaGgIBuX37t0jAOTn58fJTE1NSSqVUlZWllJ+V1fXegdFGhoavB9IRETl5eUkFoupV69ePHl1QZGGhga9ePGixrpqawczMzOSSCS8gFKBj48PAaADBw5wMoV/LFq0SLA8ob5XX79R3BuEAl9G9bDpM0atPHz4ECtWrADwbiFqmzZt4O3tjcWLF8Pa2hoAEBsbCw0NDezdu1ewDHV1dcHpITs7OyVZcnIyCgsLMXDgQG6KrDpiY2MBANnZ2YJneijqTE5O5g2l6+jowNTUVCm/sbExrl+/XmOd75OTk4N169bh3Llz+M9//oNXr17x0hULlgHg7t27ACC4Y09IFhcXh4qKCrx+/VrQvtTUVADv7Bs6dGi1OiYnJ+PVq1dwdnZGy5YtldKdnZ0RHh6O+Ph4ODg4VFtOY+Pp6YmdO3fCysoKXl5ecHJyQt++faGhocHLp7jOYWFhiIqKUipHIpHUyb8KCgqwdetWpXx1PQ+mV69eSjJjY2OubC0trVrLaKjPqqmpcf3tfd68eYMffvgBR44cQXJyMoqLi0FEXHpVH7x58yYANGi6qiqlpaVISEiAkZER1q1bp5ReXl7Os6eoqAgZGRno0qWL4OL9zz//HOfPn6+XDhYWFtDU1OTJxGIx2rRpg4KCglq/7+npicWLF8PKygrjx4+Hs7MzHBwceFOdtVFYWIj09HR07tyZ84WqODs7Y9euXYiPj+d28SoQuv9VR339ZujQoVi8eDHmzJmDyMhIDBo0CA4ODujYsWOd6/wUYUERo1bc3d0RFhZWY578/Hy8ffuWC56EKCkpUZIp1kRURXEzq8vW2fz8fADA2bNncfbs2TrXXd2Wa7FYjMrKylrrrVq/ra0tMjMz0b9/fwwcOBA6OjpQUVFBfHw8Tp06hbKyMi5/YWEhAMDAwECpLKG2UNh39epVXL16tVo9hNq2Kop6heoA/n+H2cuXL2ssp7HZvn07zMzMsG/fPqxevRqrV6+GmpoaPD09sXnzZm5LsqId1qxZU6/y37e3oKBA0EfrGhQJ+Y1iHUhFRUWdymioz7Zu3VpwLR8AjB07FqdPn0bHjh3h5eWF1q1bQyKRoKCgANu2beP5YH36V028ePECRIQnT57Uqd8rfKt169aC+arzzZqoqR/X5XoEBARAX18fO3fuxJYtW7B582aIxWJ4eHhg69atgj+c3udD+lZ9bK6v35iamuL69etYsWIFzp07x63ptLS0xKpVqzBu3Lg61/0pwYIiRqMgk8kgEonw/Pnzen1P6Cav+JX25MmTOtULAN9//z3mzp1br7obg7179yIzMxOrV6/GkiVLeGnr1q3DqVOneDKFvrm5uZDL5by0Z8+eKZWvyO/n58dbsFpfFOUI1VFVrsjXUFq0eHce7Nu3b5UWjQo9FCQSCfz9/eHv74+nT58iJiYGwcHB2L9/P7KzsxEeHs7Tq7CwsE6jMQre96/27dvzRlCagob6bHUBUVxcHE6fPg13d3ecPXsWKioqXFpsbCy2bdvGy1+f/lUTCjtsbGxw69atOufPyckRTK/ON/9MRCIRvvrqK3z11VfIy8vD5cuXcfjwYRw9ehSpqalISEjgtacQH9K3qrumNdVTH7/p1q0bTpw4gfLycty+fRvnzp3D9u3b4eXlBSMjI/Tv37/O9X8qsBOtGY2Cvb098vLyuOmcD8HS0hIymQxxcXF48eJFrfUCqPeUV31Q3BSFfnk+fPgQADB8+HCltMuXLyvJunfvDuDdrqj3EZLZ2tpCJBJ9sH2dOnWCmpoa4uLiUFpaqpQeExMD4MN3lymmO99/4FZWVnJTh9VhZGSECRMmICwsDBYWFoiKiuKmIhXXWTGF8FenJp9pbJ9V+OCQIUOUHuBCPqiYsomIiKi17Jrs0NLSQufOnZGUlFSnqSqZTAZTU1OkpaUhOztbKV1I14+Jvr4+Ro4ciZCQELi4uCApKQlpaWkAam4HmUwGMzMzpKWlCQaajdW3PsRvJBIJ+vTpgxUrVmD79u0gIpw5c+aD9PlfhQVFjEZh/vz5AIBp06YhLy9PKT07OxtJSUl1KkssFsPHxwcvX77EggULlG5EL1++RHFxMYB3N3h7e3scPnwYISEhSmVVVlZyN6WGoqenBwD473//q5SmGO25cuUKT37o0CH89ttvSvnHjx+PFi1aYMuWLbx2KikpEZwaatu2LTw9PXHt2jVs3LhRcJTjxo0bgoFOVaRSKSZMmIDnz59j7dq1vLSoqCicO3cOHTp0+OBfjr179wYApfdXbdmyBRkZGTxZWVkZLly4oGRTSUkJioqKIJFIuIeRr68vxGIx5s2bh8ePHyvVW1BQgDt37nyQ7o1JTT7T2D5bnQ8mJiYqXWvgXaBtZ2eHS5cuYc+ePUrpVR/surq6EIlEgnYA7/p9aWkpZsyYITiFm5GRwXs9zqRJk/DmzRssW7aMly8iIqLe64kag/DwcLx9+5YnKy8v56aq1NXVAdTeDpMnT0Z5eTkCAwN5/nz//n0EBwdDW1u7Tsdm1ER9/SYuLk5wVE4xcqWwjcGHTZ8xGoVBgwZh6dKlWLVqFTp06IBBgwZBLpcjLy8PaWlpuHz5MlavXo3OnTvXqbyVK1ciNjYWBw4cQGxsLAYPHgxVVVWkp6cjLCwMV65c4X55HT58GAMGDMD48eOxdetW2NjYQE1NDZmZmbh+/Tpyc3Px+vXrBtvm4uKCTZs2wcfHB+PGjYOGhgZMTEzg7e2NSZMmYf369Zg3bx6io6Mhl8tx7949REVFYfTo0QgNDeWVZWlpicWLF+O7776DtbU1xo0bB7FYjNDQUFhbW+P+/fvcFJSCHTt2ICUlBQEBAThw4AD69u0LbW1tPH78GLdv30ZqaiqysrIEF1BXZf369YiJicHq1atx7do12Nvbc+cUtWzZEsHBwUp115epU6diw4YNWL58OeLj42Fubo5bt27h/v37cHJy4t20X716BVdXV5iZmcHe3h4mJiYoLi7GmTNnkJ2djX/84x/cuUNWVlbYsWMHZs+eDUtLS3h4eMDc3Jxb5BoTE4MpU6b8ZV5HMmDAAIhEIixZsgTJycnQ1taGtrY2Zs+eDaBxfdbOzg52dnY4evQosrKy0KdPH2RmZuLXX3/FkCFDcPz4caXv/Pzzz3B2dsbMmTM5n3r9+jUSExNx584dLmDX1NSEra0tLl26hKlTp8LCwgItWrSAt7c3TExM4OPjg9jYWPz000+4evUqBg4cCCMjIzx79gzJycm4ceMGDh06xJ3nExAQgNDQUOzZsweJiYlwdHTE48ePcfToUQwZMqTGtTJ/Bl5eXmjZsiUcHBwgl8tRXl6OyMhIPHjwAF5eXjAxMalTOwQEBODs2bM4cOAAkpKS4OrqitzcXISEhKC8vBz79++v17RvddTHbw4ePIgdO3bA2dkZHTp0gEwmw4MHD/Dbb7+hVatWmDZt2gfr8z9JE+58Y/zFqemcouqIjIykYcOGkYGBAUkkEmrbti317duXVq1axduuX5fTa1+/fk2bNm2iHj16kLq6OmlqalKXLl3Iz89PaQttfn4+ffvtt2RlZcXltbCwIG9vbwoNDeXlVZwOLER1W8o3bNhAFhYWJJFIlLbNx8fHk5ubG+nq6pKWlhY5OTlRVFRUjTbu2LGDOnfuTFKplIyNjenvf/87PX78mDs/6X1KS0tpw4YNZGNjQxoaGqSurk6mpqY0cuRI2r9/P++MmJrIzc2l+fPnk1wuJ4lEQq1ataKxY8cKnobbkC35RO+2Dru6ulLLli1JJpPRiBEjKDU1VWlL9Js3b2j9+vXk5uZGxsbGJJVKqU2bNuTk5ERHjhwRLPvmzZs0fvx4MjIy4vTv1asXLV68mJKSkj5Y99q25AtR3ZEN+/btI2tra1JVVRU80bqxfJaIKCcnh6ZNm0ZGRkakpqZG1tbW9OOPP1J6enq1W8mzs7NpwYIFZGZmRlKplPT09Mje3p62bNnCy5eSkkIeHh6ko6NDIpFI8KiBkJAQGjhwIOnq6pJEIqF27dqRs7Mzbd68WelMnLy8PJo5cyZ3ArWNjc0HnWgthFB7CV2nHTt20PDhw0kul5Oamhrp6+uTvb097dq1S6lP1dYOxcXFtHTpUurYsSNJpVLS0dGhwYMHCx4jUds5VjX5b139JjY2lnx8fMjKyop0dHRIXV2dLCwsaP78+YJHpzDeISJq4lWHDAYDwLtprC+++AIBAQFYv359U6vDYDAYnxxsTRGD8ZHJzc1VWidVUFCAwMBAAPjgtQcMBoPBaBhsTRGD8ZE5ePAgNm3aBBcXFxgZGSErKwthYWHIycnBlClTuPePMRgMBuPjwoIiBuMj069fP9jY2CAqKgr5+flQUVFB586dsXTpUvj6+ja1egwGg/HJwtYUMRgMBoPBYICtKWIwGAwGg8EAwIIiBoPBYDAYDAAsKGIwGAwGg8EAwIIiBoPBYDAYDAAsKGIwGAwGg8EAwIIiBoPBYDAYDAAsKGIwGAwGg8EAwIIiBoPBYDAYDADA/wHFb7yjHeYhjQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "x = ['10%','20%','30%','40%','50%','60%','70%','80%','90%','100%']\n",
    "plt.plot(x, ndcg, marker='o', label='NDCG@5')\n",
    "plt.plot(x, pre2, marker='s',label='precision@5')\n",
    "plt.plot(x, recall2, marker='^', label='recall@5')\n",
    "\n",
    "# Add labels and title\n",
    "plt.xlabel('Percentage of user-interacted histories', fontsize=14) \n",
    "plt.ylabel('Recommendation Metrics Score', fontsize=14)  \n",
    "\n",
    "\n",
    "# Add a legend\n",
    "plt.legend(fontsize=11) \n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.8 ('python_env')",
   "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": "3.10.8"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "7dce496edc5d64ae6a819663e273321724104d6dec13b7a667e60e1c594c3876"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}