Graph Convolutional Network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Graph Convolutional Neural Network (GCN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "\n",
    "## Spektral allows us to conveniently load many pre-processed datasets \n",
    "import spektral   ## Version 0.6.1\n",
    "import scipy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Cora dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- adj - adjacency matrix\n",
    "- features - feature vector\n",
    "- labels - topic of eac paper\n",
    "- mask - which node is in the training, validation and test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading cora dataset\n",
      "Pre-processing node features\n"
     ]
    }
   ],
   "source": [
    "adj, features, labels, train_mask, val_mask, test_mask = spektral.datasets.citation.load_data(dataset_name=\"cora\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Spektral load data in sparse matrix form, here we will convert them to dense matrix form. Also, the adjacency matrix does not come with self edges, so we have to add that in manually. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "adj = adj.todense() + np.eye(adj.shape[0]) ## Add an identity matrix the size of the adjacency matrix -- adding self edges\n",
    "features = features.todense()\n",
    "\n",
    "features = features.astype('float32')\n",
    "adj = adj.astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2708, 1433)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2708, 7)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set: 140\n",
      "Validation set: 500\n",
      "Testing set: 1000\n"
     ]
    }
   ],
   "source": [
    "print(f'Training set: {np.sum(train_mask)}')\n",
    "print(f'Validation set: {np.sum(val_mask)}')\n",
    "print(f'Testing set: {np.sum(test_mask)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So the `cora` dataset has 2708 nodes and 1433 features. They come from 7 classes (topics). The dataset has 140 training, 500 validation and 1000 testing sets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For classification problem, we usually use cross entropy loss. Here we modify the loss function to accomodate masking of data sets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Implementing graph neural network from Scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def masked_softmax_cross_entropy(logits, labels, mask):\n",
    "    loss = tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=labels)  ## Compute the loss\n",
    "    mask = tf.cast(mask, dtype=tf.float32)   \n",
    "    mask /= tf.reduce_mean(mask)     ## The loss function only applies to nodes in intended set\n",
    "    loss *= mask                    \n",
    "    return tf.reduce_mean(loss)\n",
    "\n",
    "def masked_accuracy(logits, labels, mask):\n",
    "    correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(labels, 1)) ## Check if logits predict the correct label\n",
    "    accuracy_all = tf.cast(correct_prediction, tf.float32)\n",
    "    mask = tf.cast(mask, dtype=tf.float32)\n",
    "    mask /= tf.reduce_mean(mask)\n",
    "    accuracy_all *= mask\n",
    "    return tf.reduce_mean(accuracy_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### In each layer of the neural network, we aggregate information from the neighbour and represent that as a vector. This essentially helps us to learn relationships in a graph. The propagation rule:\n",
    "\n",
    "## ${H}^{l+1} = \\sigma({W}\\hat{A}{H}^{l})$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gnn(fts, adj, transform, activation):\n",
    "    seq_fts = transform(fts)  ## Linear transformation with weights (W), the learnable parameters in neural network (basically pass data through neural network)\n",
    "    ret_fts = adj @ seq_fts   ## Magic of GNN - matrix multiplication with adjacency matrix\n",
    "    return activation(ret_fts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_cora(fts, adj, gnn_fn, units, epochs, lr):\n",
    "    layer_1 = tf.keras.layers.Dense(units)  ## Hidden layer\n",
    "    layer_2 = tf.keras.layers.Dense(7)      ## Classification - 7 outputs for 7 classes\n",
    "    \n",
    "    def cora_gnn(fts, adj):\n",
    "        hidden = gnn_fn(fts, adj, layer_1, tf.nn.relu)\n",
    "        logits = gnn_fn(hidden, adj, layer_2, tf.identity)\n",
    "        return logits\n",
    "    \n",
    "    optimizer = tf.keras.optimizers.Adam(learning_rate=lr)\n",
    "    \n",
    "    best_accuracy = 0.0\n",
    "    for ep in range(epochs + 1):\n",
    "        with tf.GradientTape() as t:\n",
    "            logits = cora_gnn(fts, adj)\n",
    "            loss = masked_softmax_cross_entropy(logits, labels, train_mask)\n",
    "            \n",
    "        variables = t.watched_variables()\n",
    "        grads = t.gradient(loss, variables)\n",
    "        optimizer.apply_gradients(zip(grads, variables))\n",
    "        \n",
    "        logits = cora_gnn(fts, adj)\n",
    "        val_accuracy = masked_accuracy(logits, labels, val_mask)\n",
    "        test_accuracy = masked_accuracy(logits, labels, test_mask)\n",
    "        \n",
    "        if val_accuracy > best_accuracy:\n",
    "            best_accuracy = val_accuracy\n",
    "            print(f'Epoch: {ep}, Training loss: {loss.numpy()}, Val accuracy: {val_accuracy.numpy()}, Test accuracy: {test_accuracy.numpy()}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Training loss: 1.915420651435852, Val accuracy: 0.24199999868869781, Test accuracy: 0.27299997210502625\n",
      "Epoch: 1, Training loss: 1.925779938697815, Val accuracy: 0.3539999723434448, Test accuracy: 0.35199999809265137\n",
      "Epoch: 2, Training loss: 1.6505703926086426, Val accuracy: 0.39799997210502625, Test accuracy: 0.4129999876022339\n",
      "Epoch: 3, Training loss: 1.4602525234222412, Val accuracy: 0.6539999842643738, Test accuracy: 0.6340000033378601\n",
      "Epoch: 4, Training loss: 1.2509697675704956, Val accuracy: 0.7259999513626099, Test accuracy: 0.7459999918937683\n",
      "Epoch: 16, Training loss: 0.37494441866874695, Val accuracy: 0.7439999580383301, Test accuracy: 0.7569999098777771\n",
      "Epoch: 17, Training loss: 0.33892324566841125, Val accuracy: 0.75, Test accuracy: 0.7569999098777771\n",
      "Epoch: 18, Training loss: 0.3077608644962311, Val accuracy: 0.7560000419616699, Test accuracy: 0.7549999356269836\n",
      "Epoch: 20, Training loss: 0.2587478458881378, Val accuracy: 0.7579999566078186, Test accuracy: 0.7549999356269836\n",
      "Epoch: 21, Training loss: 0.236616849899292, Val accuracy: 0.7619999647140503, Test accuracy: 0.762999951839447\n",
      "Epoch: 23, Training loss: 0.1971919685602188, Val accuracy: 0.7660000324249268, Test accuracy: 0.7569999098777771\n"
     ]
    }
   ],
   "source": [
    "train_cora(features, adj, gnn, 32, 200, 1e-2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Baseline \n",
    "\n",
    "We can use an identity matrix as our adjacency matrix, this means the model won't be able to get information from neighbours. Basically, the model will only use information in its feature vector to classify nodes, but will not exploit relationships in graphs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Training loss: 1.947203278541565, Val accuracy: 0.23799999058246613, Test accuracy: 0.25300002098083496\n",
      "Epoch: 1, Training loss: 1.9371905326843262, Val accuracy: 0.30399999022483826, Test accuracy: 0.3019999861717224\n",
      "Epoch: 2, Training loss: 1.9221415519714355, Val accuracy: 0.351999968290329, Test accuracy: 0.33400002121925354\n",
      "Epoch: 5, Training loss: 1.8486964702606201, Val accuracy: 0.38999995589256287, Test accuracy: 0.36399996280670166\n",
      "Epoch: 6, Training loss: 1.8183174133300781, Val accuracy: 0.41599997878074646, Test accuracy: 0.39899998903274536\n",
      "Epoch: 7, Training loss: 1.7846019268035889, Val accuracy: 0.4559999704360962, Test accuracy: 0.4389999508857727\n",
      "Epoch: 8, Training loss: 1.7475885152816772, Val accuracy: 0.48399996757507324, Test accuracy: 0.4599999785423279\n",
      "Epoch: 9, Training loss: 1.7073338031768799, Val accuracy: 0.4899999797344208, Test accuracy: 0.4750000238418579\n",
      "Epoch: 10, Training loss: 1.6640915870666504, Val accuracy: 0.5059999823570251, Test accuracy: 0.4919999837875366\n",
      "Epoch: 11, Training loss: 1.6180870532989502, Val accuracy: 0.5179999470710754, Test accuracy: 0.4989999830722809\n",
      "Epoch: 12, Training loss: 1.5694928169250488, Val accuracy: 0.5219999551773071, Test accuracy: 0.49399998784065247\n",
      "Epoch: 24, Training loss: 0.8499256372451782, Val accuracy: 0.5299999713897705, Test accuracy: 0.49699997901916504\n",
      "Epoch: 40, Training loss: 0.17228646576404572, Val accuracy: 0.5339999794960022, Test accuracy: 0.5090000033378601\n"
     ]
    }
   ],
   "source": [
    "train_cora(features, tf.eye(adj.shape[0]), gnn, 32, 200, 1e-2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unsurprisingly, we get a lower accuracy ~ 50%, adding relationships information improved the model by over 20%."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Averaging\n",
    "\n",
    "We may face diminishing or exploding gradient problem if the range of input is large. One way to normalize the adjacency matrix is to scale the adjacency matrix with the inverse of its diagonal node degree matrix (D). Essentially, we divide the features by the number of degrees (neighbours) at each node.\n",
    "\n",
    "## $\\hat{A}{D}^{-1}$  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[4., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 4., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 6., ..., 0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0., ..., 2., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 5., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 5.]], dtype=float32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "D = np.zeros_like(adj)\n",
    "np.fill_diagonal(D, np.sum(adj, axis=1))\n",
    "D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0.25      , 0.        , 0.        , ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.25      , 0.        , ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.        , 0.16666667, ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        ...,\n",
       "        [0.        , 0.        , 0.        , ..., 0.5       , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.        , 0.        , ..., 0.        , 0.2       ,\n",
       "         0.        ],\n",
       "        [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
       "         0.2       ]], dtype=float32)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Take the inverse of D\n",
    "\n",
    "D_inverted = np.linalg.inv(D)\n",
    "D_inverted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "A_averaged = adj @ D_inverted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0.25      , 0.        , 0.        , ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.25      , 0.16666667, ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.25      , 0.16666667, ..., 0.        , 0.        ,\n",
       "         0.        ],\n",
       "        ...,\n",
       "        [0.        , 0.        , 0.        , ..., 0.5       , 0.        ,\n",
       "         0.        ],\n",
       "        [0.        , 0.        , 0.        , ..., 0.        , 0.2       ,\n",
       "         0.2       ],\n",
       "        [0.        , 0.        , 0.        , ..., 0.        , 0.2       ,\n",
       "         0.2       ]], dtype=float32)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "A_averaged"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Training loss: 1.9453867673873901, Val accuracy: 0.17399998009204865, Test accuracy: 0.17499998211860657\n",
      "Epoch: 4, Training loss: 1.8708319664001465, Val accuracy: 0.17799998819828033, Test accuracy: 0.1849999874830246\n",
      "Epoch: 5, Training loss: 1.8470600843429565, Val accuracy: 0.18399998545646667, Test accuracy: 0.19199998676776886\n",
      "Epoch: 6, Training loss: 1.822647213935852, Val accuracy: 0.1979999840259552, Test accuracy: 0.20800001919269562\n",
      "Epoch: 7, Training loss: 1.7970420122146606, Val accuracy: 0.21399998664855957, Test accuracy: 0.22599998116493225\n",
      "Epoch: 8, Training loss: 1.7696502208709717, Val accuracy: 0.2459999918937683, Test accuracy: 0.25999999046325684\n",
      "Epoch: 9, Training loss: 1.7403113842010498, Val accuracy: 0.2919999957084656, Test accuracy: 0.2939999997615814\n",
      "Epoch: 10, Training loss: 1.709227442741394, Val accuracy: 0.3319999873638153, Test accuracy: 0.3409999907016754\n",
      "Epoch: 11, Training loss: 1.6764627695083618, Val accuracy: 0.3739999830722809, Test accuracy: 0.3969999849796295\n",
      "Epoch: 12, Training loss: 1.6418111324310303, Val accuracy: 0.42799997329711914, Test accuracy: 0.453000009059906\n",
      "Epoch: 13, Training loss: 1.6050821542739868, Val accuracy: 0.4959999918937683, Test accuracy: 0.5179999470710754\n",
      "Epoch: 14, Training loss: 1.5667647123336792, Val accuracy: 0.5539999604225159, Test accuracy: 0.5640000104904175\n",
      "Epoch: 15, Training loss: 1.5269018411636353, Val accuracy: 0.593999981880188, Test accuracy: 0.6060000061988831\n",
      "Epoch: 16, Training loss: 1.4855263233184814, Val accuracy: 0.6200000047683716, Test accuracy: 0.6469999551773071\n",
      "Epoch: 17, Training loss: 1.4432799816131592, Val accuracy: 0.6499999761581421, Test accuracy: 0.6780000329017639\n",
      "Epoch: 18, Training loss: 1.3999029397964478, Val accuracy: 0.6739999651908875, Test accuracy: 0.7079999446868896\n",
      "Epoch: 19, Training loss: 1.3554762601852417, Val accuracy: 0.6959999799728394, Test accuracy: 0.7289999723434448\n",
      "Epoch: 20, Training loss: 1.310103178024292, Val accuracy: 0.7139999866485596, Test accuracy: 0.7459999918937683\n",
      "Epoch: 21, Training loss: 1.2638843059539795, Val accuracy: 0.7259999513626099, Test accuracy: 0.762999951839447\n",
      "Epoch: 22, Training loss: 1.2169843912124634, Val accuracy: 0.7420000433921814, Test accuracy: 0.7799999117851257\n",
      "Epoch: 23, Training loss: 1.1695621013641357, Val accuracy: 0.75, Test accuracy: 0.7869999408721924\n",
      "Epoch: 24, Training loss: 1.121790885925293, Val accuracy: 0.7599999308586121, Test accuracy: 0.7939998507499695\n",
      "Epoch: 25, Training loss: 1.0741517543792725, Val accuracy: 0.7679999470710754, Test accuracy: 0.7999998927116394\n",
      "Epoch: 26, Training loss: 1.0270636081695557, Val accuracy: 0.7799999713897705, Test accuracy: 0.8019998669624329\n",
      "Epoch: 27, Training loss: 0.9805203080177307, Val accuracy: 0.7859998941421509, Test accuracy: 0.7999998927116394\n",
      "Epoch: 29, Training loss: 0.8895052671432495, Val accuracy: 0.7919999361038208, Test accuracy: 0.8059998154640198\n",
      "Epoch: 31, Training loss: 0.8024019002914429, Val accuracy: 0.7979999780654907, Test accuracy: 0.8109998106956482\n",
      "Epoch: 34, Training loss: 0.6814090013504028, Val accuracy: 0.7999999523162842, Test accuracy: 0.8159998059272766\n",
      "Epoch: 35, Training loss: 0.6440392732620239, Val accuracy: 0.8019999861717224, Test accuracy: 0.8159998059272766\n",
      "Epoch: 38, Training loss: 0.5414469242095947, Val accuracy: 0.8039999604225159, Test accuracy: 0.8169998526573181\n"
     ]
    }
   ],
   "source": [
    "train_cora(features, A_averaged, gnn, 32, 200, 1e-2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is a small but obvious improvement over using the un-normalized adjacency matrix. The model is able to achieve over 80% accuracy."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Normalized laplacian matrix\n",
    "\n",
    "In the <a href=\"https://arxiv.org/abs/1609.02907\">original paper</a> describing the implementation of GCN by Thomas N. Kipf and Max Welling, they recommended a normalized version of the adjacency matrix as follows:\n",
    "\n",
    "## $\\hat{A} = \\tilde{D}^{-\\frac{1}{2}} \\tilde{A} \\tilde{D}^{-\\frac{1}{2}} $   where $\\tilde{A} = {A} + {I}$ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "D_invroot = np.linalg.inv(scipy.linalg.sqrtm(D)) \n",
    "A_hat = D_invroot @ adj @ D_invroot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, Training loss: 1.9465360641479492, Val accuracy: 0.11999998241662979, Test accuracy: 0.14000000059604645\n",
      "Epoch: 1, Training loss: 1.9344302415847778, Val accuracy: 0.12199999392032623, Test accuracy: 0.14399999380111694\n",
      "Epoch: 2, Training loss: 1.9210528135299683, Val accuracy: 0.12399999797344208, Test accuracy: 0.1459999978542328\n",
      "Epoch: 4, Training loss: 1.885765790939331, Val accuracy: 0.1339999884366989, Test accuracy: 0.15600000321865082\n",
      "Epoch: 5, Training loss: 1.8658390045166016, Val accuracy: 0.15399999916553497, Test accuracy: 0.17399999499320984\n",
      "Epoch: 6, Training loss: 1.8449172973632812, Val accuracy: 0.1759999841451645, Test accuracy: 0.19700001180171967\n",
      "Epoch: 7, Training loss: 1.8216502666473389, Val accuracy: 0.21799999475479126, Test accuracy: 0.22899998724460602\n",
      "Epoch: 8, Training loss: 1.7955398559570312, Val accuracy: 0.2540000081062317, Test accuracy: 0.2840000092983246\n",
      "Epoch: 9, Training loss: 1.767196536064148, Val accuracy: 0.32600000500679016, Test accuracy: 0.3529999852180481\n",
      "Epoch: 10, Training loss: 1.7372571229934692, Val accuracy: 0.4099999964237213, Test accuracy: 0.429999977350235\n",
      "Epoch: 11, Training loss: 1.7058148384094238, Val accuracy: 0.47999998927116394, Test accuracy: 0.5049999952316284\n",
      "Epoch: 12, Training loss: 1.6730451583862305, Val accuracy: 0.5320000052452087, Test accuracy: 0.5529999732971191\n",
      "Epoch: 13, Training loss: 1.6388643980026245, Val accuracy: 0.5719999670982361, Test accuracy: 0.5979999899864197\n",
      "Epoch: 14, Training loss: 1.6030075550079346, Val accuracy: 0.6320000290870667, Test accuracy: 0.6380000114440918\n",
      "Epoch: 15, Training loss: 1.5654428005218506, Val accuracy: 0.6520000100135803, Test accuracy: 0.6630000472068787\n",
      "Epoch: 16, Training loss: 1.5262142419815063, Val accuracy: 0.6800000071525574, Test accuracy: 0.6829999685287476\n",
      "Epoch: 17, Training loss: 1.4854209423065186, Val accuracy: 0.6840000152587891, Test accuracy: 0.7100000381469727\n",
      "Epoch: 18, Training loss: 1.4432034492492676, Val accuracy: 0.6940000057220459, Test accuracy: 0.7280000448226929\n",
      "Epoch: 19, Training loss: 1.3997161388397217, Val accuracy: 0.7139999866485596, Test accuracy: 0.7329999804496765\n",
      "Epoch: 20, Training loss: 1.3550961017608643, Val accuracy: 0.731999933719635, Test accuracy: 0.7399999499320984\n",
      "Epoch: 21, Training loss: 1.3094851970672607, Val accuracy: 0.7379999756813049, Test accuracy: 0.7479999661445618\n",
      "Epoch: 22, Training loss: 1.2630747556686401, Val accuracy: 0.7399999499320984, Test accuracy: 0.7509999871253967\n",
      "Epoch: 23, Training loss: 1.2160131931304932, Val accuracy: 0.7459999918937683, Test accuracy: 0.7549999356269836\n",
      "Epoch: 24, Training loss: 1.1684913635253906, Val accuracy: 0.7479999661445618, Test accuracy: 0.757999837398529\n",
      "Epoch: 25, Training loss: 1.1207032203674316, Val accuracy: 0.7539999485015869, Test accuracy: 0.763999879360199\n",
      "Epoch: 26, Training loss: 1.072827696800232, Val accuracy: 0.7580000162124634, Test accuracy: 0.7679998874664307\n",
      "Epoch: 27, Training loss: 1.0250741243362427, Val accuracy: 0.7660000324249268, Test accuracy: 0.7709999084472656\n",
      "Epoch: 28, Training loss: 0.9776590466499329, Val accuracy: 0.7739999890327454, Test accuracy: 0.7719998955726624\n",
      "Epoch: 30, Training loss: 0.8846217393875122, Val accuracy: 0.7759999632835388, Test accuracy: 0.7769998908042908\n",
      "Epoch: 36, Training loss: 0.6316617727279663, Val accuracy: 0.7779998779296875, Test accuracy: 0.7909999489784241\n",
      "Epoch: 40, Training loss: 0.49376747012138367, Val accuracy: 0.7799999117851257, Test accuracy: 0.7919999361038208\n"
     ]
    }
   ],
   "source": [
    "train_cora(features, A_hat, gnn, 32, 200, 1e-2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Using Stellargraph"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The StellarGraph library supports many state-of-the-art machine learning (ML) algorithms on graphs. The core of the GCN neural network model is a “graph convolution” layer. This layer is similar to a conventional dense layer, augmented by the graph adjacency matrix to use information about a node’s connections. It provides functions that we can use straight out of the box, so we don't have to build our GCN everytime."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import stellargraph as sg\n",
    "\n",
    "from tensorflow import keras\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import LabelBinarizer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `Cora` dataset is in the stellargraph database as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'The Cora dataset consists of 2708 scientific publications classified into one of seven classes. The citation network consists of 5429 links. Each publication in the dataset is described by a 0/1-valued word vector indicating the absence/presence of the corresponding word from the dictionary. The dictionary consists of 1433 unique words.'"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = sg.datasets.Cora()\n",
    "dataset.description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "G, node_subjects = dataset.load()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 2708, Edges: 5429\n",
      "\n",
      " Node types:\n",
      "  paper: [2708]\n",
      "    Features: float32 vector, length 1433\n",
      "    Edge types: paper-cites->paper\n",
      "\n",
      " Edge types:\n",
      "    paper-cites->paper: [5429]\n",
      "        Weights: all 1 (default)\n",
      "        Features: none\n"
     ]
    }
   ],
   "source": [
    "print(G.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`G.info()` tells us some details about the data, such as the number of nodes, and the number of features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>subject</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Neural_Networks</th>\n",
       "      <td>818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Probabilistic_Methods</th>\n",
       "      <td>426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Genetic_Algorithms</th>\n",
       "      <td>418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Theory</th>\n",
       "      <td>351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Case_Based</th>\n",
       "      <td>298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reinforcement_Learning</th>\n",
       "      <td>217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Rule_Learning</th>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        subject\n",
       "Neural_Networks             818\n",
       "Probabilistic_Methods       426\n",
       "Genetic_Algorithms          418\n",
       "Theory                      351\n",
       "Case_Based                  298\n",
       "Reinforcement_Learning      217\n",
       "Rule_Learning               180"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(node_subjects.value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`node_subjects` contains the labels, i.e. the subject a paper belongs to."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use machine learning style data processing with data from stellagraph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_subjects, test_subjects = train_test_split(node_subjects, train_size=140, test_size=None, stratify=node_subjects)\n",
    "val_subjects, test_subjects = train_test_split(test_subjects, train_size=500, test_size=None, stratify=test_subjects)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "matplotlib.style.use('fivethirtyeight')\n",
    "\n",
    "subject_stratification1 = pd.DataFrame({\"Training\":train_subjects.value_counts()/len(train_subjects),\n",
    "                                        \"Validation\": val_subjects.value_counts()/len(val_subjects),\n",
    "                                        \"Overall\":node_subjects.value_counts()/len(node_subjects)})\n",
    "\n",
    "subject_stratification1.plot.bar()\n",
    "plt.ylabel(\"Percentage (%)\")\n",
    "plt.xticks(rotation=70)\n",
    "plt.title(\"Class distribution for 7 subjects\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class distribution is skewed towards neural networks, with stratification we ensure each class is equally represented in both training, validation and testing sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_encoding = LabelBinarizer()\n",
    "\n",
    "train_targets = target_encoding.fit_transform(train_subjects)\n",
    "val_targets = target_encoding.transform(val_subjects)\n",
    "test_targets = target_encoding.transform(test_subjects)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The CORA dataset contains attributes w_x that correspond to words found in that publication. If a word occurs more than once in a publication the relevant attribute will be set to one, otherwise it will be zero. These numeric attributes have been automatically included in the StellarGraph instance G, and so we do not have to do any further conversion."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating GCN layers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A machine learning model in Stellargraph consists of a pair of items:\n",
    "\n",
    "1. The layers - graph convolution layers, dropout, pooling or even dense layers\n",
    "2. Data generator - convert the graph structure and node features into a format that can be fed into keras model for training and prediction. For this we will use the `FullBatchNodeGenerator` from `stellargraph`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using GCN (local pooling) filters...\n"
     ]
    }
   ],
   "source": [
    "## Specifying the method='gcn' argument will yield data appropriate for the GCN algorithm specifically\n",
    "## by using the normalized graph Laplacian matrix to capture the graph structure.\n",
    "\n",
    "generator = sg.mapper.FullBatchNodeGenerator(G, method=\"gcn\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A generator encodes the data required to produce the model inputs. Calling the `flow` method with a set of nodes and their true labels will produce an object that can be "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_gen = generator.flow(train_subjects.index, train_targets)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To create GCN layers, we can use the built-in GCN constructor from `stellargraph.layer`. Four arguments need to be specified:\n",
    "\n",
    "1. `layer_sizes`: the number of hidden GCN layer and the number of nodes as a list \n",
    "2. `activations`: activation function foe each layer\n",
    "3. `generator`: data generator that passes in the input data\n",
    "4. `dropout`: the dropout rate at each GCN layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "gcn = sg.layer.GCN(layer_sizes=[16, 16],         ## 2 layers with 16 nodes each\n",
    "                   activations=[\"relu\", \"relu\"], \n",
    "                   generator=generator, \n",
    "                   dropout=0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To create a `Keras` model, we now expose the input and output layer of the GCN layer with the `GCN.in_out_tensors()` method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<KerasTensor: shape=(1, 2708, 1433) dtype=float32 (created by layer 'input_33')>,\n",
       " <KerasTensor: shape=(1, None) dtype=int32 (created by layer 'input_34')>,\n",
       " <KerasTensor: shape=(1, None, 2) dtype=int64 (created by layer 'input_35')>,\n",
       " <KerasTensor: shape=(1, None) dtype=float32 (created by layer 'input_36')>]"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_inp, x_out = gcn.in_out_tensors()\n",
    "\n",
    "x_inp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<KerasTensor: shape=(1, None, 16) dtype=float32 (created by layer 'gather_indices_8')>"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_out"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The output of the GCN layer is a 16-dimensional vector. Since `stellargraph` is built on `Keras`, to perform class predictions, we can simply add a Dense layer with softmax activation using standard `Keras` functionality. This ensures the final output vector has the right number of dimensions (one for each class) and the output will be probabilities that sum to 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = keras.layers.Dense(units=train_targets.shape[1], activation=\"softmax\")(x_out)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training model and evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can create the actual model with input as `x_inp` and output as `predictions`. Just like other neural network model, we have to compile it and specify optimizers, loss and metrics."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = keras.Model(inputs=x_inp, outputs=predictions)\n",
    "model.compile(\n",
    "    optimizer=keras.optimizers.Adam(lr=0.01),\n",
    "    loss=keras.losses.categorical_crossentropy,\n",
    "    metrics=[\"acc\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When training the model, we also want to look at at validation metrics, so we create a data generator for validation set as well. Moreover, we also want to implement early stopping."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [],
   "source": [
    "val_gen = generator.flow(val_subjects.index, val_targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "es_callback = keras.callbacks.EarlyStopping(monitor=\"val_acc\", patience=50, restore_best_weights=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "1/1 - 0s - loss: 1.9574 - acc: 0.1571 - val_loss: 1.9120 - val_acc: 0.1840\n",
      "Epoch 2/200\n",
      "1/1 - 0s - loss: 1.8864 - acc: 0.2143 - val_loss: 1.8633 - val_acc: 0.1800\n",
      "Epoch 3/200\n",
      "1/1 - 0s - loss: 1.8229 - acc: 0.2286 - val_loss: 1.8049 - val_acc: 0.1620\n",
      "Epoch 4/200\n",
      "1/1 - 0s - loss: 1.7444 - acc: 0.1929 - val_loss: 1.7446 - val_acc: 0.1620\n",
      "Epoch 5/200\n",
      "1/1 - 0s - loss: 1.6789 - acc: 0.1857 - val_loss: 1.6942 - val_acc: 0.1620\n",
      "Epoch 6/200\n",
      "1/1 - 0s - loss: 1.5744 - acc: 0.1929 - val_loss: 1.6471 - val_acc: 0.1720\n",
      "Epoch 7/200\n",
      "1/1 - 0s - loss: 1.5303 - acc: 0.1857 - val_loss: 1.5818 - val_acc: 0.2560\n",
      "Epoch 8/200\n",
      "1/1 - 0s - loss: 1.4369 - acc: 0.3357 - val_loss: 1.5016 - val_acc: 0.4660\n",
      "Epoch 9/200\n",
      "1/1 - 0s - loss: 1.3343 - acc: 0.5071 - val_loss: 1.4225 - val_acc: 0.5820\n",
      "Epoch 10/200\n",
      "1/1 - 0s - loss: 1.2313 - acc: 0.6357 - val_loss: 1.3533 - val_acc: 0.6380\n",
      "Epoch 11/200\n",
      "1/1 - 0s - loss: 1.1629 - acc: 0.7214 - val_loss: 1.2898 - val_acc: 0.6640\n",
      "Epoch 12/200\n",
      "1/1 - 0s - loss: 1.0689 - acc: 0.7357 - val_loss: 1.2282 - val_acc: 0.6880\n",
      "Epoch 13/200\n",
      "1/1 - 0s - loss: 1.0132 - acc: 0.7214 - val_loss: 1.1703 - val_acc: 0.6960\n",
      "Epoch 14/200\n",
      "1/1 - 0s - loss: 0.8987 - acc: 0.7857 - val_loss: 1.1168 - val_acc: 0.7200\n",
      "Epoch 15/200\n",
      "1/1 - 0s - loss: 0.8112 - acc: 0.8286 - val_loss: 1.0706 - val_acc: 0.7460\n",
      "Epoch 16/200\n",
      "1/1 - 0s - loss: 0.7349 - acc: 0.8571 - val_loss: 1.0323 - val_acc: 0.7600\n",
      "Epoch 17/200\n",
      "1/1 - 0s - loss: 0.6723 - acc: 0.8786 - val_loss: 0.9997 - val_acc: 0.7680\n",
      "Epoch 18/200\n",
      "1/1 - 0s - loss: 0.5975 - acc: 0.8857 - val_loss: 0.9725 - val_acc: 0.7700\n",
      "Epoch 19/200\n",
      "1/1 - 0s - loss: 0.6026 - acc: 0.8500 - val_loss: 0.9472 - val_acc: 0.7740\n",
      "Epoch 20/200\n",
      "1/1 - 0s - loss: 0.5183 - acc: 0.8714 - val_loss: 0.9265 - val_acc: 0.7700\n",
      "Epoch 21/200\n",
      "1/1 - 0s - loss: 0.4336 - acc: 0.9071 - val_loss: 0.9115 - val_acc: 0.7700\n",
      "Epoch 22/200\n",
      "1/1 - 0s - loss: 0.4044 - acc: 0.8929 - val_loss: 0.9010 - val_acc: 0.7700\n",
      "Epoch 23/200\n",
      "1/1 - 0s - loss: 0.3588 - acc: 0.9000 - val_loss: 0.8945 - val_acc: 0.7680\n",
      "Epoch 24/200\n",
      "1/1 - 0s - loss: 0.2979 - acc: 0.9143 - val_loss: 0.8899 - val_acc: 0.7700\n",
      "Epoch 25/200\n",
      "1/1 - 0s - loss: 0.2936 - acc: 0.9000 - val_loss: 0.8838 - val_acc: 0.7700\n",
      "Epoch 26/200\n",
      "1/1 - 0s - loss: 0.3008 - acc: 0.9214 - val_loss: 0.8823 - val_acc: 0.7720\n",
      "Epoch 27/200\n",
      "1/1 - 0s - loss: 0.2767 - acc: 0.9143 - val_loss: 0.8837 - val_acc: 0.7720\n",
      "Epoch 28/200\n",
      "1/1 - 0s - loss: 0.2430 - acc: 0.9214 - val_loss: 0.8866 - val_acc: 0.7700\n",
      "Epoch 29/200\n",
      "1/1 - 0s - loss: 0.2114 - acc: 0.9286 - val_loss: 0.8922 - val_acc: 0.7680\n",
      "Epoch 30/200\n",
      "1/1 - 0s - loss: 0.2421 - acc: 0.9000 - val_loss: 0.8922 - val_acc: 0.7680\n",
      "Epoch 31/200\n",
      "1/1 - 0s - loss: 0.2229 - acc: 0.9214 - val_loss: 0.8926 - val_acc: 0.7680\n",
      "Epoch 32/200\n",
      "1/1 - 0s - loss: 0.1958 - acc: 0.9286 - val_loss: 0.8938 - val_acc: 0.7780\n",
      "Epoch 33/200\n",
      "1/1 - 0s - loss: 0.1819 - acc: 0.9286 - val_loss: 0.8957 - val_acc: 0.7860\n",
      "Epoch 34/200\n",
      "1/1 - 0s - loss: 0.1866 - acc: 0.9286 - val_loss: 0.8945 - val_acc: 0.7880\n",
      "Epoch 35/200\n",
      "1/1 - 0s - loss: 0.2019 - acc: 0.9500 - val_loss: 0.8928 - val_acc: 0.7940\n",
      "Epoch 36/200\n",
      "1/1 - 0s - loss: 0.1644 - acc: 0.9643 - val_loss: 0.8905 - val_acc: 0.7940\n",
      "Epoch 37/200\n",
      "1/1 - 0s - loss: 0.1718 - acc: 0.9714 - val_loss: 0.8882 - val_acc: 0.8000\n",
      "Epoch 38/200\n",
      "1/1 - 0s - loss: 0.1614 - acc: 0.9714 - val_loss: 0.8848 - val_acc: 0.8000\n",
      "Epoch 39/200\n",
      "1/1 - 0s - loss: 0.0920 - acc: 1.0000 - val_loss: 0.8823 - val_acc: 0.7980\n",
      "Epoch 40/200\n",
      "1/1 - 0s - loss: 0.1200 - acc: 0.9929 - val_loss: 0.8852 - val_acc: 0.8000\n",
      "Epoch 41/200\n",
      "1/1 - 0s - loss: 0.1419 - acc: 0.9786 - val_loss: 0.8924 - val_acc: 0.7940\n",
      "Epoch 42/200\n",
      "1/1 - 0s - loss: 0.1116 - acc: 0.9786 - val_loss: 0.8950 - val_acc: 0.7960\n",
      "Epoch 43/200\n",
      "1/1 - 0s - loss: 0.1911 - acc: 0.9500 - val_loss: 0.9101 - val_acc: 0.7980\n",
      "Epoch 44/200\n",
      "1/1 - 0s - loss: 0.0807 - acc: 0.9929 - val_loss: 0.9230 - val_acc: 0.7960\n",
      "Epoch 45/200\n",
      "1/1 - 0s - loss: 0.1286 - acc: 0.9571 - val_loss: 0.9277 - val_acc: 0.7960\n",
      "Epoch 46/200\n",
      "1/1 - 0s - loss: 0.0790 - acc: 0.9857 - val_loss: 0.9316 - val_acc: 0.7980\n",
      "Epoch 47/200\n",
      "1/1 - 0s - loss: 0.1020 - acc: 0.9714 - val_loss: 0.9416 - val_acc: 0.8020\n",
      "Epoch 48/200\n",
      "1/1 - 0s - loss: 0.0885 - acc: 0.9929 - val_loss: 0.9496 - val_acc: 0.8020\n",
      "Epoch 49/200\n",
      "1/1 - 0s - loss: 0.0796 - acc: 0.9786 - val_loss: 0.9564 - val_acc: 0.7960\n",
      "Epoch 50/200\n",
      "1/1 - 0s - loss: 0.0697 - acc: 0.9929 - val_loss: 0.9595 - val_acc: 0.7980\n",
      "Epoch 51/200\n",
      "1/1 - 0s - loss: 0.0684 - acc: 0.9857 - val_loss: 0.9588 - val_acc: 0.8000\n",
      "Epoch 52/200\n",
      "1/1 - 0s - loss: 0.0702 - acc: 0.9714 - val_loss: 0.9600 - val_acc: 0.7980\n",
      "Epoch 53/200\n",
      "1/1 - 0s - loss: 0.0445 - acc: 1.0000 - val_loss: 0.9599 - val_acc: 0.8060\n",
      "Epoch 54/200\n",
      "1/1 - 0s - loss: 0.0409 - acc: 1.0000 - val_loss: 0.9617 - val_acc: 0.8020\n",
      "Epoch 55/200\n",
      "1/1 - 0s - loss: 0.0419 - acc: 0.9929 - val_loss: 0.9569 - val_acc: 0.7960\n",
      "Epoch 56/200\n",
      "1/1 - 0s - loss: 0.0879 - acc: 0.9714 - val_loss: 0.9552 - val_acc: 0.8000\n",
      "Epoch 57/200\n",
      "1/1 - 0s - loss: 0.0683 - acc: 0.9786 - val_loss: 0.9598 - val_acc: 0.8020\n",
      "Epoch 58/200\n",
      "1/1 - 0s - loss: 0.0438 - acc: 0.9929 - val_loss: 0.9684 - val_acc: 0.8060\n",
      "Epoch 59/200\n",
      "1/1 - 0s - loss: 0.0440 - acc: 0.9857 - val_loss: 0.9805 - val_acc: 0.8060\n",
      "Epoch 60/200\n",
      "1/1 - 0s - loss: 0.0547 - acc: 0.9857 - val_loss: 0.9915 - val_acc: 0.8060\n",
      "Epoch 61/200\n",
      "1/1 - 0s - loss: 0.0498 - acc: 0.9857 - val_loss: 1.0072 - val_acc: 0.8000\n",
      "Epoch 62/200\n",
      "1/1 - 0s - loss: 0.0928 - acc: 0.9714 - val_loss: 1.0286 - val_acc: 0.7920\n",
      "Epoch 63/200\n",
      "1/1 - 0s - loss: 0.0395 - acc: 0.9929 - val_loss: 1.0505 - val_acc: 0.7940\n",
      "Epoch 64/200\n",
      "1/1 - 0s - loss: 0.0658 - acc: 0.9786 - val_loss: 1.0684 - val_acc: 0.7940\n",
      "Epoch 65/200\n",
      "1/1 - 0s - loss: 0.0472 - acc: 0.9929 - val_loss: 1.0796 - val_acc: 0.7860\n",
      "Epoch 66/200\n",
      "1/1 - 0s - loss: 0.0311 - acc: 0.9857 - val_loss: 1.0850 - val_acc: 0.7920\n",
      "Epoch 67/200\n",
      "1/1 - 0s - loss: 0.0470 - acc: 0.9857 - val_loss: 1.0917 - val_acc: 0.7880\n",
      "Epoch 68/200\n",
      "1/1 - 0s - loss: 0.0330 - acc: 1.0000 - val_loss: 1.0935 - val_acc: 0.7960\n",
      "Epoch 69/200\n",
      "1/1 - 0s - loss: 0.0790 - acc: 0.9643 - val_loss: 1.0898 - val_acc: 0.8000\n",
      "Epoch 70/200\n",
      "1/1 - 0s - loss: 0.0481 - acc: 0.9929 - val_loss: 1.0884 - val_acc: 0.8040\n",
      "Epoch 71/200\n",
      "1/1 - 0s - loss: 0.0165 - acc: 1.0000 - val_loss: 1.0938 - val_acc: 0.7960\n",
      "Epoch 72/200\n",
      "1/1 - 0s - loss: 0.0561 - acc: 0.9714 - val_loss: 1.1005 - val_acc: 0.7920\n",
      "Epoch 73/200\n",
      "1/1 - 0s - loss: 0.0126 - acc: 1.0000 - val_loss: 1.1135 - val_acc: 0.7940\n",
      "Epoch 74/200\n",
      "1/1 - 0s - loss: 0.0494 - acc: 0.9857 - val_loss: 1.1231 - val_acc: 0.7920\n",
      "Epoch 75/200\n",
      "1/1 - 0s - loss: 0.0369 - acc: 0.9786 - val_loss: 1.1194 - val_acc: 0.7900\n",
      "Epoch 76/200\n",
      "1/1 - 0s - loss: 0.0231 - acc: 1.0000 - val_loss: 1.1187 - val_acc: 0.7920\n",
      "Epoch 77/200\n",
      "1/1 - 0s - loss: 0.0532 - acc: 0.9714 - val_loss: 1.1141 - val_acc: 0.7980\n",
      "Epoch 78/200\n",
      "1/1 - 0s - loss: 0.0299 - acc: 1.0000 - val_loss: 1.1077 - val_acc: 0.8000\n",
      "Epoch 79/200\n",
      "1/1 - 0s - loss: 0.0267 - acc: 0.9929 - val_loss: 1.0983 - val_acc: 0.8020\n",
      "Epoch 80/200\n",
      "1/1 - 0s - loss: 0.0503 - acc: 0.9714 - val_loss: 1.0999 - val_acc: 0.8020\n",
      "Epoch 81/200\n",
      "1/1 - 0s - loss: 0.0322 - acc: 0.9929 - val_loss: 1.1082 - val_acc: 0.7980\n",
      "Epoch 82/200\n",
      "1/1 - 0s - loss: 0.0311 - acc: 0.9929 - val_loss: 1.1210 - val_acc: 0.7880\n",
      "Epoch 83/200\n",
      "1/1 - 0s - loss: 0.0210 - acc: 1.0000 - val_loss: 1.1386 - val_acc: 0.7880\n",
      "Epoch 84/200\n",
      "1/1 - 0s - loss: 0.0174 - acc: 1.0000 - val_loss: 1.1521 - val_acc: 0.7900\n",
      "Epoch 85/200\n",
      "1/1 - 0s - loss: 0.0253 - acc: 1.0000 - val_loss: 1.1633 - val_acc: 0.7860\n",
      "Epoch 86/200\n",
      "1/1 - 0s - loss: 0.0377 - acc: 0.9857 - val_loss: 1.1668 - val_acc: 0.7880\n",
      "Epoch 87/200\n",
      "1/1 - 0s - loss: 0.0370 - acc: 0.9857 - val_loss: 1.1714 - val_acc: 0.7860\n",
      "Epoch 88/200\n",
      "1/1 - 0s - loss: 0.0213 - acc: 1.0000 - val_loss: 1.1717 - val_acc: 0.7860\n",
      "Epoch 89/200\n",
      "1/1 - 0s - loss: 0.0495 - acc: 0.9857 - val_loss: 1.1619 - val_acc: 0.7900\n",
      "Epoch 90/200\n",
      "1/1 - 0s - loss: 0.0267 - acc: 0.9929 - val_loss: 1.1536 - val_acc: 0.7920\n",
      "Epoch 91/200\n",
      "1/1 - 0s - loss: 0.0110 - acc: 1.0000 - val_loss: 1.1482 - val_acc: 0.7960\n",
      "Epoch 92/200\n",
      "1/1 - 0s - loss: 0.0217 - acc: 0.9929 - val_loss: 1.1457 - val_acc: 0.7960\n",
      "Epoch 93/200\n",
      "1/1 - 0s - loss: 0.0287 - acc: 0.9929 - val_loss: 1.1402 - val_acc: 0.7980\n",
      "Epoch 94/200\n",
      "1/1 - 0s - loss: 0.0184 - acc: 1.0000 - val_loss: 1.1383 - val_acc: 0.8020\n",
      "Epoch 95/200\n",
      "1/1 - 0s - loss: 0.0521 - acc: 0.9786 - val_loss: 1.1388 - val_acc: 0.8000\n",
      "Epoch 96/200\n",
      "1/1 - 0s - loss: 0.0384 - acc: 0.9786 - val_loss: 1.1440 - val_acc: 0.8020\n",
      "Epoch 97/200\n",
      "1/1 - 0s - loss: 0.0523 - acc: 0.9929 - val_loss: 1.1588 - val_acc: 0.7980\n",
      "Epoch 98/200\n",
      "1/1 - 0s - loss: 0.0783 - acc: 0.9857 - val_loss: 1.1741 - val_acc: 0.8000\n",
      "Epoch 99/200\n",
      "1/1 - 0s - loss: 0.0262 - acc: 1.0000 - val_loss: 1.1903 - val_acc: 0.8000\n",
      "Epoch 100/200\n",
      "1/1 - 0s - loss: 0.0248 - acc: 0.9929 - val_loss: 1.2063 - val_acc: 0.8020\n",
      "Epoch 101/200\n",
      "1/1 - 0s - loss: 0.0278 - acc: 0.9929 - val_loss: 1.2182 - val_acc: 0.8040\n",
      "Epoch 102/200\n",
      "1/1 - 0s - loss: 0.0151 - acc: 1.0000 - val_loss: 1.2283 - val_acc: 0.8020\n",
      "Epoch 103/200\n",
      "1/1 - 0s - loss: 0.0475 - acc: 0.9857 - val_loss: 1.2275 - val_acc: 0.8020\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    train_gen,\n",
    "    epochs=200,\n",
    "    validation_data=val_gen,\n",
    "    verbose=2,\n",
    "    shuffle=False,  # this should be False, since shuffling data means shuffling the whole graph\n",
    "    callbacks=[es_callback],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sg.utils.plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also test our models against the test set. We first create a data generator for the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_gen = generator.flow(test_subjects.index, test_targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 84ms/step - loss: 0.9513 - acc: 0.8109\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.9513043761253357, 0.8109284043312073]"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(test_gen)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Making preictions for all nodes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_nodes = node_subjects.index\n",
    "all_gen = generator.flow(all_nodes)\n",
    "all_predictions = model.predict(all_gen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[8.1418889e-08, 3.1100715e-06, 9.9961615e-01, ...,\n",
       "         1.0657994e-09, 5.2741859e-07, 4.9522987e-06],\n",
       "        [7.6832622e-03, 9.0821013e-02, 3.6741528e-01, ...,\n",
       "         6.3195988e-03, 4.0143007e-01, 9.3992986e-02],\n",
       "        [1.1974002e-03, 7.5799685e-06, 1.0658345e-06, ...,\n",
       "         9.9528790e-01, 1.1043371e-04, 6.8067387e-04],\n",
       "        ...,\n",
       "        [1.3483842e-05, 9.9918240e-01, 2.1042832e-04, ...,\n",
       "         3.5524089e-08, 4.5609902e-04, 2.3015868e-05],\n",
       "        [9.9893349e-01, 6.4941111e-07, 6.4469293e-07, ...,\n",
       "         1.0273035e-06, 1.4368739e-11, 1.7503523e-12],\n",
       "        [5.2643161e-09, 1.7176314e-06, 9.9954468e-01, ...,\n",
       "         2.0257460e-09, 5.6517019e-06, 1.3234920e-04]]], dtype=float32)"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The predictions generated by the model are the probabilities of the node being in each class. We can use `inverse_transform` method of `labelBinarizer` to return the target categories."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [],
   "source": [
    "node_predictions = target_encoding.inverse_transform(all_predictions.squeeze())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Predicted</th>\n",
       "      <th>Ground Truth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>31336</th>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1061127</th>\n",
       "      <td>Rule_Learning</td>\n",
       "      <td>Rule_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1106406</th>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13195</th>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37879</th>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1126012</th>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1107140</th>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Theory</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102850</th>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31349</th>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1106418</th>\n",
       "      <td>Theory</td>\n",
       "      <td>Theory</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Predicted            Ground Truth\n",
       "31336           Neural_Networks         Neural_Networks\n",
       "1061127           Rule_Learning           Rule_Learning\n",
       "1106406  Reinforcement_Learning  Reinforcement_Learning\n",
       "13195    Reinforcement_Learning  Reinforcement_Learning\n",
       "37879     Probabilistic_Methods   Probabilistic_Methods\n",
       "1126012   Probabilistic_Methods   Probabilistic_Methods\n",
       "1107140  Reinforcement_Learning                  Theory\n",
       "1102850         Neural_Networks         Neural_Networks\n",
       "31349           Neural_Networks         Neural_Networks\n",
       "1106418                  Theory                  Theory"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({\"Predicted\": node_predictions, \"Ground Truth\": node_subjects})\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Node embeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Node embedding algorithms compute low-dimensional vector representations of nodes in a graph, which will give us insight into nodes’ similarity and network structure. Given the widespread prevalence of graphs on the web and in the physical world, representation learning on graphs plays a significant role in a wide range of applications, such as link prediction and anomaly detection. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use the last hidden graph convolution layer of the GCN model since the outputs are lower-dimensionality representations of the original data. We do not want to train this model on the labels. The idea is after passing through the convolution layers, each node would have gathered some information about its neighbours. The output of the model should output similar values for nodes that are similar (hopefully)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 2708, 16)"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "embedding_model = keras.Model(inputs=x_inp, outputs=x_out)\n",
    "emb = embedding_model.predict(all_gen)\n",
    "emb.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The result still has 16 dimensions, which is too much to plot. We can further use dimensionality reductions such as PCA, T-SNE."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dim_reduction_plot(transform, data):\n",
    "    X = data.squeeze()\n",
    "    trans = transform(n_components=2)\n",
    "    X_reduced = trans.fit_transform(X)\n",
    "\n",
    "    fig, ax = plt.subplots(figsize=(7, 7))\n",
    "\n",
    "    ax.scatter(X_reduced[:, 0],\n",
    "              X_reduced[:, 1],\n",
    "              c=node_subjects.astype(\"category\").cat.codes,\n",
    "              cmap=\"jet\",\n",
    "              alpha=0.7)\n",
    "\n",
    "    ax.set_title(f\"{transform.__name__} visualization of GCN embeddings for cora dataset\")\n",
    "    ax.set_xlabel(\"$X_1$\")\n",
    "    ax.set_ylabel(\"$X_2$\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj8AAAHlCAYAAAAa4Z3jAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Z1A+gAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOyddZwd1d3/3+fMzNX1ze7GE2KECBIkwYIWCw4tVqcEa3+0tBR4igZaihSehocipUgpWqBIcAg0WBIIRIgL8WTdr83MOb8/5t67u9m7yQaLzTuvZdm5I2f8c78qGhoaND4+Pj4+Pj4+uwhyWw/Ax8fHx8fHx+e7xBc/Pj4+Pj4+PrsUvvjx8fHx8fHx2aXwxY+Pj4+Pj4/PLoUvfnx8fHx8fHx2KXzx4+Pj4+Pj47NL4Yufr8mECRMoKira1sPYLLfccgtFRUW8//77HaYXFRUxYcKEbTSqzlx88cUUFRWxatWqbT2Ur83999/PuHHj6NWrF0VFRdxyyy3beki7DI8//vg2P+Zf5bmQ637s6t7dkaisrOTiiy9m1KhRlJSU7DT3+PbIznC9fFdsVvwUFRVt1c/jjz+eXXb27NlMnDiR0aNHU15eTt++fdlrr734/ve/zx133EF9fX2HbWUeFiUlJcyZMyfneG644YZO24G2l+bmfi6++OKveox8viEy52lnvzGfe+45rrzySmzbZuLEiVx55ZUccsgh3VrWcRyefvppzjvvPEaOHEnPnj3p2bMno0eP5pxzzuHhhx+mubm5y+WXLl3KlVdeyUEHHUT//v0pKytj2LBhnHnmmTz00EO0trZ2mD9zf4wYMYJYLJZznUcffbT/wvL5ylxyySU8+eSTjB49mt/+9rdceeWVFBYWbuth+WyGzPt4R7jnM4JvU12wJczNfXjllVd2mvbEE0+wZs0azjnnHPr379/hs9GjRwPw7LPPcuGFF6KU4tBDD+WEE04gGo2ydu1apk+fzltvvcVhhx3G/vvv32n9SimuueYaXn755a3aEYATTjghO4ZN6Wr61+W+++4jHo9/K+v+tpk5cybhcHhbDyPL9ddfz29+8xt69+69rYfytXjjjTcA79rIdY13xbJly/jxj3/MggULKCwsZPz48QwYMADTNNmwYQMfffQRr732GpMmTWLJkiVYltVh+dtuu40///nPKKXYb7/9OPvss8nPz6eqqoqPPvqIyy+/nLvvvpvPP/+807bXr1/P3XffnfOe99l2TJw4kTPOOIO+fftu66F8JVKpFO+++y5Dhw7lySef3NbD8fHJslnxc/XVV3ea9sEHH7BmzRrOPfdcDj300E6fx+NxLr/8cgBeeOEFDjvssE7zzJ49m169euXc5uDBg3n//fd59dVXOeGEE7q1ExkmTJjAeeedt1XLfF369ev3nW7vm2TYsGHbeggdyFg5dnQ2bNgAQHl5ebeXqays5JRTTmHdunVMnDiR6667jry8vE7z/fe//+W6667Ddd0O4ucvf/kLf/rTn+jTpw8PP/wwBxxwQKdl3333XW6++eZO0wsKCggEAkyePJmf/OQnO8U52FkoLS2ltLR0Ww/jK1NZWYlSaqvuBR+f74JvPOZn4cKFNDU1sccee+QUPgB77703FRUVOT+77rrrMAyD66+/HsdxvunhbZYzzjiDoqIiZs+enfPz119/vZMLLZdvX2vNk08+ybHHHsuQIUOoqKhgxIgRnHTSSTz66KMd5h09enSXVqlM7MKm5rxp06Zx2WWXMXbsWPr160fPnj0ZN24cf/rTn7bKCrVpjMH777+/Va7NrRnH6NGjs9/8TjrppA7rzLC5mJ+XXnqJE088kf79+1NRUcEBBxzAH//4R1paWjrN295k+/DDD3PQQQdRUVHB0KFDueyyy2hsbOz2MQJobm7m5ptvZv/996eiooL+/ftz0kkn8corr3SYb1N/+1577dVpH7vi5ptvZt26dXz/+9/ntttuyyl8AA477DCmTp1KMBjMTlu1ahV//vOfsSyLp59+OqfwATjiiCN49dVXO00PhUJcffXVtLa25hRHX4VEIsHdd9/NYYcdRp8+fejduzeHH344Dz30EFp37KizatWq7LVYVVXFpZdeytChQ+nduzfHHHMMH330EQCtra1ce+21jBo1ivLycsaOHcsLL7yw2XFMnz6dk08+mX79+tGvXz/OPPPMLu9vpRT//Oc/OfbYY7PX2YEHHsidd95JKpXKucxzzz3HYYcdRs+ePRkyZAgTJ07Mit9cpFIpbrvtNvbee2/Ky8vZc889ufnmm0kmkznn31y83ujRozsdk3322Yf//d//7XSMwXsu3XvvvYwdO5aKigr22GMPrrjiChobG3M+h1KpFPfffz+HHXYYu+22Gz179mTUqFGceeaZvPTSS13uY4b26/zwww9zhiCkUikmT57MIYccQq9evejbty9HH300jz32WM59yOx3Y2MjV111FaNGjaK0tJS//e1vWxwPeF8Azj77bIYOHUp5eTkjRozgBz/4Aa+99lqnY/XPf/6To48+mr59+9KrVy8OPfRQ7r77bmzbzrmvRUVFJJNJbrnlFsaMGUNZWRlXXXUV4H0huvXWWzn22GMZNmwYZWVlDB8+nPPPP5+FCxd2a+ztmT17dtYi2K9fP0455RRmzpzZ5fxTpkxh4sSJ7LvvvvTu3ZvevXszfvx4/va3v+G6bod5i4qK+PDDD4GOz7D218fs2bO58sorOfjggxk4cCAVFRWMGTOGq6++ulNIC2z9tVRZWclVV13FmDFjqKioYMCAAZx22mn897//7TDfhAkTuPXWWwG49NJLO7xXtuSy26zl56tQXFwMwMaNG2ltbSUajW7V8nvssQc//OEPefTRR3nooYeYOHHiNz3ELjn33HN55513eOKJJ9h77707fZ55eZ977rmbXc9NN93EnXfeSf/+/TnllFMoLCyksrKSL774gqeeeoqf/OQnX2ucf/3rX1myZAljx47lmGOOIZFIMGPGDG677Tbef/99Xn75ZUxz609t//79c7o9tNbcd999NDU1EYlEvtI4Lr74Yp544gm++OKLnC7TzfHHP/6R22+/neLiYk4//XQKCwt59913uf3223nttdd47bXXyM/P77Tc9ddfz9SpUznuuOM44ogjeP/993n00UdZsWJFt92qjY2NHH/88SxYsIA999yTiy66iMbGRl544QXOO+88rr766uwxy8T1ZFzDF110UbdiG+LxOM888wxA9mG5OQzD6PD3448/jm3bnH766YwaNWqzy7YXTe35yU9+wgMPPMATTzzBRRddtMX1bI7m5mZOPfVUZs2axZ577pm9X9555x0uv/xyPvnkE+69995OyzU2NnLsscdSXFzM97//fdavX8+LL77IGWecwVtvvcVll11GS0sLJ5xwAs3NzTz33HP87Gc/o0+fPjndi7NmzeKuu+7iiCOO4IILLmD58uW8/PLLfPjhh7zwwguMHTs2O6/jOPzwhz/k9ddfZ8iQIZxxxhkEg0E+/PBDJk2axH//+1+ee+65DvfVPffcwx/+8AcKCgo466yzKCoqYurUqRxzzDEUFBR0Go/Wmp/+9Ke8+uqrDBw4kAsuuADbtnn88ceZP3/+Vh9nx3E444wz2LBhA0cffTSmafLKK69www03kEgkOl1Lv/vd7/jHP/5Bz549+fGPf0wwGOT1119n1qxZOI7T6ZlxySWX8OyzzzJ8+HC+//3vE41G2bBhA5999hlTpkzh5JNP3uz4Lr74YlavXs19991Hv379stdB5iVq2zZnnnkm06ZNY8iQIfz85z8nlUoxZcoUfvWrXzF9+nTuueeeTutNpVKcfPLJNDY28r3vfY9wOEyfPn22eLz+9Kc/cdtttxGNRpkwYQJ9+/alsrKSTz/9lMcee4zjjz8+O+9FF13E008/Te/evTn33HOxLIvXX3+da6+9lnfffZdnnnkm5zP2xz/+MXPmzOGoo47ixBNPZMCAAQB89NFH/O///i+HHnooJ598MtFolOXLl/PSSy9ln2F77bXXFvcBYMaMGZx66qkkk0lOOukkBg8ezPz58znxxBMZP358zmVuvPFGpJRZ8dPU1MS0adP4n//5Hz777DMefPDB7LxXXnllzmdY+2fZo48+ypQpUzj44IM5/PDDUUoxe/Zs7r33Xt5++22mTp3a4Zm8NdfS/PnzOe2006iurubII4/khBNOoK6ujldeeYVTTz2VyZMn86Mf/Qhoexd/+OGHncJetvjsbWho0Fvzc/DBB2tAv/zyyzk/r6+v12PGjNGAHjlypL7tttv0e++9pzdu3Nit9c6cOVMvWbJE5+Xl6dLSUr1q1arsPL/+9a81oO+5554Oy55zzjka0CeccIK+8sorc/5MmzZti/u2ceNGXVhYqEtLS3V1dXWHz1auXKmDwaAeMGCArq+v7zTu9vMWFxfrXr166XXr1nXaxvLlyzv83a9fP92vX7+c47nnnnty7u/s2bM7jCHz87vf/U4D+h//+EeH6VdeeWXOcwbogw8+eIvH5dJLL9WAPuWUUzpsd2vHkTlPXV07mc/nzJmTnfbWW29pQPfu3VsvXLiww3V29tlna0BfcMEFOa+lvn376nnz5mWn19TU6AMPPFAD+p133unW9f7zn/9cA/q8887rsK/z58/XFRUVWgihp06dmnP77fdjcz+vvPKKBnSfPn226l7M/IwfP14DevLkyVu9LKDLy8t1Q0ODfvrppzWgDz/88A7z7Lffflu1Pz/60Y80oG+44YYO0ysrK/UxxxyjAf3kk09mp8+ZM0cDGtATJ07scJyvvfZaDeiioiJ90kkn6aqqquxnf//73zWgJ0yYkPO+AfTtt9/e4bNHH31UA3ro0KEdtvOHP/whey3V1tZmp9fV1emf/OQnGtC33nprhzFblqULCgr07NmzO8x/6qmnZrffftuZ8Y4ZM0Zv2LAhO33lypV68ODBOe/Hzd27gP7e977XYV1Lly7VBQUFuqCgoMMz7NVXX9WAHjRokF65cmV2elVVVfaeaP8cWrVqlRZC6L333lvX1NR0OsebPse6+smc21zPmRtuuEED+sgjj+xwXletWqX32GMPDehHH300534ffvjhev369d2+zp9//vnsPn7xxRedPp8/f372/x966KHs+2v16tUdjlXmXrvppps6LN+vXz8N6BEjRuQ8NkuXLtVr1qzpNP3999/X0WhUH3XUUd3aj/r6ej106NCcx+b222/PHp9Nr5fPP/+807rq6uqyz9C33nqrw2dbeobNnTu3w32S+Zk8ebIG9PXXX/+VrqXa2lo9ZMgQHQwG9ZQpUzrMt2jRIt2nTx8dDof10qVLO90jm74nt/Tzjbu9hBA8+uijHHLIIcyfP5/f//73HH744fTt25fx48dzyy23UFNTs9l1lJeX8+tf/5ra2lruuOOObm/71Vdf5dZbb835M2/evC0uHwqFOOOMM6itreX111/v8Nmzzz5LMpnk7LPPRgixxXVZlpXzm8E34b8fOHBgzjFceumlAEydOvVrbyPDAw88wD333MO4ceO4//77O2z3uxjHY489BsDll1/eIU5MCMGkSZMIh8M88cQTOU3Rv//97zvEZJmmmY0JmzVr1ha3bds2Tz31FJFIhEmTJnXY1z59+nD55ZdnzeNfh6qqKoAu4+CefPJJbrnllg4/7YOWKysrAb52oPixxx7L4YcfznvvvZcN2t5a6uvrefLJJ9lzzz359a9/3eGzYDDIddddB8DTTz/dadloNMp1113X4TifddZZADQ0NHDzzTcTCASyn51++ulYltXlvT1o0CDOP//8DtNOOeUUDjjgAJYuXcqMGTMAz9113333UVZWxi233NLBsial5MYbb0QI0WHM//73v7FtmwsuuICBAwd2mP+GG27oZJ0Dsi7ja6+9tkOiQVFREb/73e9y7sOWuPXWWzusq6ysjBNOOIGmpiaWLl2anZ6xWv/mN7/p4IYNBAJcf/31ndYrhEBrTSAQyLkv38RzLHPf/PGPf+xwXgsLC7PXyaZhAhluuummDlboLXH//fdnl8sVPN7ecpQZ1/XXX9/BghcIBPjTn/602XH94Q9/yHlsysrKclqnR48ezaGHHsoHH3yQ8xm2KTNmzGDp0qWMHTuWU045pcNn559/PoMGDcq53G677dZpmpSSiy66CNj6Z3X//v1zXhc/+tGPKCgo6LC+rbmW3nzzTZYtW8b555/fKUO2Z8+e/OpXvyIej/Piiy9u1Xhz8Y27vcALAp4yZQqLFy/m3Xff5fPPP+ezzz5j7ty5zJ07lwcffJDnnnsup2spw6WXXsrDDz/MAw88wPnnn9/hAdMV99xzz9cOeD733HN56KGHePLJJznppJOy05988kmEEJxzzjlbXMf3v/99HnjgAQ444ABOPfVUDjzwQMaOHZt1CX5dWltbue+++5gyZQrLly+nubm5g398czEHW8Mrr7zCVVddxZAhQ3jiiScIhULf+TgyZQ9ymXMzPvtZs2axbNky9thjjw6f57q+Mg++hoaGLW57yZIlxGIx9ttvv5wPtMMPP7zDGL8tnnrqqU6+7oqKCvbZZ59vfFs33XQThx12GNdddx1HH310zofV5si4UKSUOevsZOL4lixZ0umzQYMGdYp1ygRfFxYWZl0IGQzDoKysjPXr1+ccy4EHHoiUnb/fHXzwwcycOZO5c+cybtw4li1bRm1tLbvtthu33357znWFw+EOY86c84MPPrjTvAMHDqRPnz6sXr26w/Q5c+YghOCggw7KOaatpaCgIOfLLtc1PnfuXMA7Jpuy//77d/qiVlBQwHHHHcfrr7/OwQcfzIknnsiBBx7I/vvv32U82tbQ3NzMihUrKC8v73TfAtl40Vz3VigU2mq37Keffgp4ZRu2RGabuRJ6Ro0aRVlZGcuWLaOlpaXTsdh33327XO8bb7zBQw89xOzZs6mtre0U01pbW7vFZIPNXXdSSsaNG8eKFSs6fVZXV8fkyZN58803WbVqVaeSF1v7rLZtm4cffpjnn3+eRYsW0dTUhFIq5/q25lrKfCFZu3ZtzudHZt8WL168VePNxbcifjLsvvvu7L777tm/lyxZwi9/+UtmzpzJr371q83WewmHw1x77bVcdNFF3HDDDTzyyCPf5lCz7Lfffuy+++689dZb1NbWUlpaypIlS5g1a1Y2uGtL3HLLLQwaNIgnnniCyZMn89e//hUpJYcddhiTJk36Wmn3tm1z8sknM2vWLEaMGMFpp51Gjx49sg+vW2+9tcvgya1h1qxZXHDBBZSWlvLss89SUlKyTcbR1NQEdJ05lQmczxXEnMvnm3mZbxrk901ve2vIrH/jxo05P2//LeeWW27JBvi1H8fixYu7FAFbw+jRozn33HP517/+xcMPP8wvfvGLrVq+rq4O8AIiuwosBnIGqueKk8lcT7k+A+98dpUY0dV5KysrA9rOb2bMX375Zadj2xWZZTPryrXtTcVPU1MTBQUFOeOuvko2VFcxDbmu8c2N1zCMTvc3wMMPP8zkyZN59tlnue222wDPon3cccdx8803dxKjW8OW7q1IJEJBQUHOe6tHjx7dsr63p7GxkYKCgm4Jt8x56qoMSEVFBdXV1TQ1NXVaX1eJPPfeey9XX301RUVFHHHEEfTt25dwOIwQgldeeYUvvviiW8/L7lx3m9LQ0MARRxzBqlWr2HfffTn77LMpLi7GMAwaGxu57777tvpZ/bOf/YwpU6YwcOBATjjhBCoqKrLWu3vvvbfT+rp7LWXuxZdeemmzQfWbirevwrcqfjZl2LBh3H///eyzzz7MmzeP+vr6zVpDzjrrLO677z5eeOGFzUayf9Oce+65XH/99TzzzDPZQN3M9O5gGAYXXXQRF110EXV1dUyfPp2XXnqJp59+mtNOO42ZM2dmHzZSyi7Nnblu/FdffZVZs2Zx7rnndspw2LhxY7cf3pvjyy+/zLobnnrqqZyC77sYB7S99KqqqnJmTWVcPl29HL+pbefim9r2mDFjCAaDrF27luXLlzN48OCtWn7cuHFMmzaN//73v/z4xz/+WmMBuOaaa/jPf/7Dn//8Z37wgx9s1bKZYzFx4sTsQ25b0dV5q66uBtrGmvl93HHH8dRTT3Vr3ZllMuvqzrYLCgpoaGggmUx2EkBdjfWbIuNyqa6u7iSaXNelrq6uk9s1HA5z5ZVXcuWVV7JhwwY+/vhj/v3vf/Pyyy+zaNEiPvroo061prrLlu6tWCxGU1NTTlG2tcIHPKFYW1ub01qTa2z19fXE4/GcAmhz932usTmOw5///GcqKir473//28m688knn3R7P77KdffYY4+xatUqrrzyyk7la2bOnMl9993X7e0DfP7550yZMoXDDz+cZ599toPVUCnF5MmTOy3T3Wsps3///Oc/txhQ/3X5zttbtL/wcqUytkcIwR//+EfA86V+V5x11lkYhsGTTz6JUopnnnmGaDTaycfaHUpKSjjhhBO47777OOOMM6ipqeHjjz/Ofl5UVERVVVVOAZSrGF3G7NfeJZchk574dairq+PMM8+krq6OBx98sEsz7lcZR+YbaXvz6JbIZEDkshJWV1ezcOFCotEoQ4cO7fY6u8uwYcOIRCIsWLCA2traTp9nXFGbc992h3A4nBUZX0U0nnfeeViWxUsvvcSCBQs2O293vuH17NmT//f//h81NTXceeedWzWW/fbbDyllh2t8WzF9+vSc11rm+txzzz0B7zwXFhby2WefdZnSvimZ6zLXtb5y5UrWrVuXcxmtdTZ1P9eYvi0y+5rrvHzyySdbLCvSq1cvTj/9dJ588slszNSiRYu+8njy8/MZNGgQVVVVOdczbdo04OvfWxn2228/AN5+++0tzps5tx988EGnzxYsWEB1dTVDhgzptvuvtraWxsZGDjjggE7Cp6WlZavc5pu77pRSTJ8+vdP0zLM6l5j4Ks/qzPqOP/74Tu7SWbNmbbHcyuaupUzW5tY8P7bGmt+eb1z8rFy5kvvuuy+n1UJrnQ1gHj58eE5VvykHH3wwEyZM4JNPPvlKVZ+/Cj179uSoo45i7ty53HPPPaxfv56TTz65Wxd7MpnMeeK01lm13j5Qb7/99sNxnE4BdO+88w7PPfdcp/VkUsQ3vTFXrlyZM3Bxa0gkEpxzzjksX76cP//5z5stMvlVxpE532vWrOn2mH74wx8CcOedd2a/cYF3PK+//npisRjnnHPOV/4Gujksy+Kss84iFotx4403dopnuuuuuxBCZMf4dbjmmmvo06cPzzzzTLbmTi5y3VcDBgzgqquuwrZtfvCDH2TjGzZl2rRpnHjiid0az69+9St69+7Nvffe26U7Lhc9evTgrLPOYt68edxyyy05X6rr1q3LGfPzTbN8+XL+8Y9/dJj24osvMnPmTIYOHZpNdTdNk4suuoiqqip+97vf5WzzUVtbm42bAS+uz7Is/v73v7Ny5crsdKUUN954Y84HcSYe8aabburwgmhoaNiqxI6vwtlnnw3AXXfd1SEWKJVKMWnSpE7z19TU8MUXX3Sankwms9fg1gQc5yKTrnzNNdd0+PLX1NSUHdM3YckEuPDCCwEv2Hzt2rWdPm/vMs6Ma9KkSR3cs7ZtZ7+Eb824ysrKiEQizJ49u9P6rrrqqpxfrLpi7NixDB06lBkzZnQK+v3HP/6RM96nq2f1nDlzuOuuu3JuZ3PP6q7WV11dnTNwf2uupRNOOIFBgwbx8MMP56xJlhl3xj3Wfqy5zuvm+MbdXk1NTVx11VVcd911HHDAAYwYMYL8/Hyqq6uZNm0aK1euJC8vr8uDnotJkybx5ptvsnz58s3O98orr3Tys2eoqKjg5z//ebe3ee655/Lmm29mb8Luurzi8TjHH388AwcOZJ999qFfv37Yts0HH3zAvHnz2H///TsE0l144YU8/vjjXHHFFUybNo0BAwawePFipk6dykknndTpAj/uuOMYNGgQ99xzT7b2zNq1a3njjTc45phjtvoCaM/999/PjBkz6Nu3L7W1tTkDziZMmMCee+75lcZxxBFHMHnyZCZNmsTChQuzbqwrrriiyzEdcMABXH755dx5550ceOCBnHrqqRQUFPDuu+8yZ84cRowYkc0M+Ta4/vrr+fjjj/nnP//J3LlzOfzww7N1furr6/n973+f/Vb5daioqODFF1/kRz/6Effeey9PPPEE48ePZ7fddkNKSVVVFTNnzmTZsmWUlpYyfPjwDsv/9re/xXEcbr31Vo4++mgOOOAA9tlnn+y9N336dBYvXtxtl1okEuGaa67hkksu2epr6rbbbmPFihXceuutPP3009kik5WVlSxbtoxPPvmEP/7xj996hfGjjjqKP/zhD7z99tuMHDkyW+cnHA5z9913d3BRXHHFFSxYsIB//vOfvPnmm4wfP54+ffpQU1PDl19+yfTp0/nFL36RtaAMGDCA66+/nmuuuYbx48dz2mmnUVxczDvvvENDQwMjR47sVLvnzDPP5Pnnn+e1117jwAMPZMKECdi2zcsvv8zee++9xefb1+GQQw7hpz/9KY888ggHHnggJ510EsFgkNdee42CggJ69erVITh8/fr1jB8/nhEjRjBy5Ej69OlDa2srU6dOZfny5Zx88slb7Z7dlEsvvZS3336bt99+m4MOOohjjz02ezzWr1/P2Wefzamnnvo199zjyCOP5IorruD2229n3LhxnHDCCfTr14/q6mo+/fRTBg4cmA1vOOOMM3j99df597//zbhx45gwYUK2zs+yZcs47LDDuOSSS7q9bSklF154IXfddRcHHXQQJ5xwArZt8/7771NfX8+hhx7a7X6HQgjuvvtuTjvtNH72s591qPPz7rvvcvTRR3eybp199tlMnjyZq6++mvfff5/BgwezfPly3njjDU466SSef/75Tts54ogjeOGFF7jsssuyX/wLCwuZOHEiY8aMYdy4cbz88sscc8wxjBs3jqqqKt5++22GDh3ayX26NdeSZVn861//4vTTT+fcc89lv/32Y6+99iIajbJu3Trmzp3L0qVLmTZtWlb0jB8/Hikl9913H/X19dm4p4kTJ2621s83Ln523313Hn/8cd577z1mzpzJiy++SG1tLeFwmP79+3PJJZdw8cUXb1VbiMGDB3P++edv0Tf56quvdqkWR40atVXi5/jjj6e4uJj6+noGDBjQ7caU0WiUSZMm8f777/PJJ5/w2muvEQ6HGTBgADfffDM/+9nPOpgKhw0bxksvvcRNN93E22+/jZSSffbZh5deeokvv/yyk/iJRqO89NJL3HjjjXzwwQd8/PHHDBw4kCuuuIJLL70054XcXTLfeNeuXdul+6V///7sueeeX2kcRxxxBH/+85955JFHePDBB7MumM2JH/Cqfu+555488MAD/Pvf/yaZTDJgwAB+97vfcdlll+VMIf2mKCoq4o033uCvf/0rL730En/7298IBoPsueeeXHjhhd+oX3rIkCFMmzaNZ599lhdffJFZs2bx5ptvIoSgR48ejBw5kl/+8pecfvrpOeMNrrzySk477TQefPBBPvjgA5588klisRjFxcWMGjWKiRMnditbMcPZZ5/Nfffd18Hi0R3y8/OZMmUKjz32GP/+97+ZMmUKiUSCsrKyrGg47bTTtmqdX4X99tuPK664gj/+8Y888MADgHcNXnvttZ3cKaZp8s9//pPnnnuOxx9/nLfeeouWlhZKSkro168fl19+eTYOLsMvf/lLevbsyeTJk3nqqafIy8vjqKOO4sYbb8wZKJ4pA3LXXXfxxBNP8Pe//52KigrOPfdcfv/733cZLPtNceeddzJ06FAeeeQRHnnkEUpKSjjxxBO59tprGTlyZIfYvv79+/M///M/vP/++3z44YfU1NRQWFjIoEGDuOyyy7r9ZXBzBAIBnn/+ee69916eeeYZHnzwQaSU7LHHHlx11VVZC8w3xR/+8AfGjh3L/fffnz2/ZWVljB49ulPh2fvvv5+DDjqIxx57jMceewylFIMHD2bSpElcdNFFW11ENpMC/9hjj/HII49QUFDA4YcfzjXXXJPzS+bmGDduHK+99ho33XQT77zzDu+88w777rsvU6ZM4Z133ukkfnr16sVrr73GDTfcwPTp05k6dSpDhw7lL3/5C4cddljOZ/WPfvQj1q5dy7PPPsvf/vY3bNumX79+TJw4MRsScvPNN/Pmm29y//3306tXL3784x/zu9/9rkPxUNj6a2nEiBF8+OGH3Hvvvbz66qs8+eSTaK2pqKhg+PDh/OpXv+oQ5jBs2DAeeOAB7r77bv71r39lrao/+MEPNit+RLpwlI+Pj4/PLsjy5cvZd999OeOMMzq5CX18dla+84BnHx8fH5/vnqqqqk4BrLFYLJsB1N14MB+fnYHvNNXdx8fHx2fb8MADD/DUU09xyCGH0LNnTyorK5k2bRrr1q3je9/73jcWX+PjsyPgix8fHx+fXYDx48czZ84c3n33Xerr6zFNk8GDB3PhhRdy8cUXf6X6OT4+Oyp+zI+Pj4+Pj4/PLoUf8+Pj4+Pj4+OzS+GLHx8fHx8fH59dCl/8+Pj4+Pj4+OxS+OLna7B06dJtPYRtyq68//6+77rsyvvv77vPzoIvfnx8fHx8fHx2KXzx4+Pj4+Pj47NL4YsfHx8fHx8fn10KX/z4+Pj4+Pj47FL44sfHx8fHx8dnl8IXPz4+Pj4+Pj67FL748fHx8fHx8dml8MWPj4+Pj4+Pzy6FL358fHx8fHx8dil88ePj4+Pj4+OzS+GLHx8fHx8fH59dCl/8+Pj4+Pj4+OxSmNt6AD4+Pt0jpaFOCUqkJiC2fvkNjuDJmEWtEowPOhwdcjG+wnp8fHx8dnR88ePjswPwr1aLtxImzUoQEJrvhRzOz7O7vfxnScmvGsJUuxIl4KlYgDCKfoZmXNDhioIUEd8O7OPjs4vgix8fn+2c9xIGz8UtbA2LHYOUhpkpk/eSJjcXJiiQUCz1Ztfxx+YQG1yJIcDV0KAFGoN6R7PQMZiSCPBueQt5vgDy8fHZBfDFj4/Pds5rCUG9bmVBKh8FCAQaeD1h8m4ijzAaV0C51IyyXG4tSlBqtC2vNSxzJFKAraFRe8sDtCKIoKlyBTc1Brm1OLkN9tDHx8fnu8UXPz4+2zGNooHlZivL4wNxANKyRSEAgQO0IEBDkwuVruR71SbT2llxRNra06QEdnYNbSTTcmqWbQK++PHx8dn58Y3cPj7bIcpxWDNtGm/9+w7KWtaQVCYKgUKmhY9HeyHjAM0Ial3BHU2Btnk0mGhcOgsfABeII+ghVZfjcTSsdAT1yo+Q9vHx2fHxLT8+PtsZsZoapl19Na1VVdRaNQx8ZQ76+vFgbfl2dYEEIm3FSQGw0pWEga6ljSeKfhBJ5fzsnYTJv1ot6pQgKKC32ZvbNF8p48zHx8dne8AXPz4+2xkfPPV/rDkrH1VUQjLWjD3DJtzSgCwqxhZm2vLTtdHWBnq3s+IktaZKCwS5LT8AFprnYwFOj8SpcQSTWwKscyVKaJbaBn0MTSi9yUUyyj0tAX6Tn1ss+fj4+Gzv+OLHx2c7olk0s+IYF+WGcIWDkx9Cn1JKae1q6oMGVl4+rpbY2sKhs+lF47m4Tgm3pcFvcCSGJsfcbUuFgSW25M5Gk0fjIVrTKfUSiCmBjaavoTAFmEIzL2V0uTYfHx+f7R1f/Pj4bEfMD8wDQ6JUChcXqQVRWqmIVGLGTcKlfchXhQS1yyJbUqNlh1geC83+0mW+Y/CfugAC6CEU/QxFnduVtUjQCLQoyU0tESSaPEAIQZMCF0GLI6hTgnyh6UmbCy2hodIVlBuasO8G8/Hx2UHwxY+Pz3ZEggThomLqK1eD9NSE1IIjYy8ye/cLaRCSAgwGW4r7SmLc1hRkStzCRZAnNf0NRQDNf+IWDUrgIgigCQnIE5oG3bVCcdO2ITcthgJKk0J0cLI1akFKBBlvKf7ZrvBiUGiKhGaEpTg27DDc2lyEkY+Pj8+2xRc/Pj7bESWqlJp+vWmmEaeu2ctRD5kUjyrjzIp59HBqOCh5MGZaw9xRnOSawiQvxy1iWjDWcrgoXck5rZ1oVYIGIB9FIxK9GQdYe1Lp+XT6v472RJBCsK/p8LdYkIDw5ltgG+kijIp3kyaHBx0ODTqUSM1AUyN8q5CPj892hC9+fHy2I/awR7DeXEdoQE9SA0rQSmEYJkVuD1xcBrn9ssInQ5GEH0W9GJ95KUlNO+ETV544cQGQlKFpxktt7y4SKBLQ33Bp1QLhODwSC7A0JVFC0KQhAJgC6rRXc+j/WgO8HDcpkNDbVPwymmSIpbF8EeTj47Md4IsfH5/tCAuLo+PHsMhcwKzgLJRwyXPzEQh6u33o4/bb7PK7GQqJZ61xVKaAIRhASHqFDkEQpHvlDDPLxjWscg2v0KKwWJE0CeDF/GQsRCGtMYG1WqK1l3Lf5Arm2QYfJEz2CbicGbE5Mex8xaPj4+Pj883gix8fn+0ME5NRzp6McEax1lhDg2ygr9uXElW6xWXzDBhuuXxhG6RoC4QOCs0gqZijvADpAJrkFqw/ASCIJo6gFbCAsPCKKbpA0ybLJxDUoomk3WPNStCKQArP0hRH8HBrgJGmy27W5nuR+fj4+Hyb+BWefXy2UySS/u4A9rT36pbwyXBbUYLRlksETQCICk1PqQmiyUsHPhdIz0rT9gDQtEklnf1vIi1wJDDIVIy2XFKILq1GKSQNWlCvoV6J7PozMkkC/4lb3d4XHx8fn28D3/Lj47OTMdxSPFoa44lWi4daAwQ1fOlKVmHgIAjgWYdqUjL7AIgCjXgp7BKBAmw0Bp5wcRDMtSUWkBQ6Z42h9mgEzYClvGDnwnTRRQHYWxFv5OPj4/NtsFNYfjZu3MhFF13E4MGDqaioYOzYsXzwwQfZz7XW3HLLLQwfPpyePXsyYcIEFi5cuA1H7OPz7VIk4ZJ8m8mFcb5UMp2yLggBoPkiZRBJW3ssNC14WVwW3kNBKo1wNa4iK3RSCGJpYbQlMvKmCShIZ3yB11V+QsjucjkfHx+f74IdXvw0NDRw7LHHorXmmWeeYcaMGdx2222UlZVl5/nrX//KPffcw6233srUqVMpKyvjtNNOo7m5eRuO3Mfn22e5kgSAQum5uvKkpkCCJTTXFcQ51HLY3VRYeDE9QTSBWAuBlkYMO4VwXcxUEos2QbOlVPlMkLQJFAjNQFNR48LClKRJC/4VC7DQ3uEfPT4+PjswO/wTaPLkyfTs2ZP777+ffffdl4EDB3LYYYex++67A57V59577+XXv/41p5xyCiNGjODee++lpaWFZ599dhuP3sfn26XaFWi8G729ZBEIxocU/+gR58fRFH0MRRSNmYghHc8y40oJWmMm4qBUVtBYXXYI89B47jMB/CSS4tfRBJaAIaaih6FZ4Uiubwyx2tmC60xDs4KUHxvt4+PzDbPDi59XXnmFfffdl5/97GcMGTKEQw45hAceeACtvSfmqlWrqKys5Mgjj8wuEw6HOeigg5gxY8a2GraPz3fCeRGb4CYaQ2kvBmewoSiS8NM8m0eL41gCXMdBkxZMWhFsqkcLgUgmiArvnhJsrk9YehtAvtBcUZDivZRFsfQKVq93JcsdyQZX8GhroMvl340bHFwVZWxlHmMr8/h9QxDHF0E+Pj7fEDu8+Fm5ciX/+Mc/GDhwIM899xwXXXQRN954I3//+98BqKysBOjgBsv8XVVV9Z2P18fnu6TChB9Hk6C9eBtbQUBobilMINvd/SOCioeLW+nRUIWRjJPXVMe4Fx6mfO0KDLstRDmERuiuu8N7aEIo9rMcFF5tIRf4wjZY7wqatGC9krwUN3k/IXmm1WRByqsNBLDKFlzUEKHSldgImrXg2ViAGxqDW73/dS78qCbEARuj3NQYwHW3ehU+Pj47IaKhoWGH/j5VVlbGPvvsw5tvvpmdNmnSJKZMmcLMmTOZMWMGxx57LPPmzaNfv7YCcZdeeikbNmzg+eefz7nepUuXfutj9/H5rtiIxb/NMvK0w1luDZEuwpbnP/wIn85eiDa9dPRENJ/5445h8XFn0iMc4CS7hkcCvUgKQQwjZ/yPACLaZZhq5Rp7Nf81injB7EGzaEsuTSFIaoEJhHDJ1y4HqiZ+ba9lkjWAj83CTt/M8pXDs8n53c4Vm0eYS0LDsdv11ijUNi8k5hHu5jp8fHYGhg4duq2HsN2xw6e6V1RUZON7MgwbNoy1a9dmPweorq7uIH6qq6spLy/vcr3duViWLl26S19Uu/L+72j7PhQ4NPtXYZfzDbzmGj77nz8ili1C2CmMgMXgqEl+aR5/K44z3y7lwWpBxDAIaGjUdJJRGnCEgbAijO3Xn/2AlysDSO0VPLRVurp0WpRYUhDTJp/LUl4sLqQ1aWA6RrZFRwZlSIb0G9ppelecuj4PR0uMdtOaRIB7Skdxf0l36lvnZkc7998k/r7vmvu+M7LDi59x48axbNmyDtOWLVuWFToDBgygoqKCd999lzFjxgCQSCT4+OOPmTRp0nc+Xh+f7RkrEub0W//EXYs2omuqSQ0eTjIvn5NDNmWGptzRmGmnlymgAE2jFlk3WEaXJIF6LQgKsAQcH7aZljSJaUGrgLiW6ZpCXmBzXMNqV/JozKJQaOIaou1EjtJQYepuCx+ABi1y+vXfT1h0r7mHj4/PzsoOH/NzySWX8Mknn3DHHXewYsUKXnjhBR544AF+8YtfACCE4OKLL+avf/0rL730EgsWLOCSSy4hGo1y5plnbuPR+/hsf+wdVEweXc73xu3NoaVh/lIY52d5XgbYHgHFUDcG2hMkmSrRJhCkLRjaRBPXgpubvDidH0dtehmaIaYiT7aJpJDQJHSm8KEgIKCf6VWfTiqN0uBorz3H5MJ4zvGudeD4qggjN+Sx54Y8fl4bxlZdf7PLlzu0p9/Hx+cbYIe3/IwZM4bHH3+cSZMmcfvtt9O3b1/+53/+Jyt+AC677DLi8ThXXHEFDQ0N7Lvvvjz//PPk5+dvw5H7+Gy/FLbrFN8eIeAaZxX/KhjOHNvwMscUtCKo1QKJQKIJAwEJ82yDtY5guKX4Q36Ch2MBltsCkY4WcjWkMutGky+8GkH7B1xAExTQz1D8viBJqdFpOCgFJ9fk0ahE1ir0btLkvNowhwdtXk52zCgzgLuKEpvd9yoHbm8OUqskF0WTHBDqTllHHx+fHYkdXvwAHHvssRx77LFdfi6E4Oqrr+bqq6/+Dkfl47Nz0lPb3F+SoNb1BMdKR3BVfYi4YyCF1xJDASVCk1Sw1pX0NV2GBxSxVsFAS2M5mhWupAnaucwEG1yoVQYDpeLwsMtVBZt3Tz0TN6lXAjMtfDI1hmakTI4O2RSjaECg0qLsOCtFtRJ8mDAYG3SpU4K4hn6G51J7JWZweWOEVLo567tJk2NC3v76+PjsPOwU4sfHx+e7p9TwZEtxQHNBXorrm0IktCChPbfWl65AIqlVnjJ5JW5Srbw4oKjURJVGAa0aSNcWcgChYa4j6ZkS3NYU5IeRFL3N3K6qxY7MiidHQ2s6/sgF3klYDLEUpYZnuVnnCD5xLZpbDex0jJGrQEkoEpqbCxNc1xTC0WTFlATeSnip+CMCvgXIx2dnwRc/Pj4+X5nVjuDvLUEqlWCoqVjiCJSWXg+wdCr7ZfVhVjsJmrQkKLyKzRtdz1qjtGfxsdDYCE88AQaadcqgNaX5zA7xp4IEs2yDOSmDiIAfRFIMsTQnhxweaAmS1IIUnliJ0soYYwauLmKRuxfFEpIaNijPLWcJWOJIqpTERJMPVCvBxXURXKExNgmq1ggearW4I+AHSfv47Cz44sfHx+crUa8EVzWGcbQXCxSUkNASG0+EhIXGSgucR2IBfhRO8UXKIomgRXkWmgAaiSYqQWkvy8vFa6a63JH0NhUVwPl1EQoNLwZIafisMcxleQmejgcJCU2D9gopHm++xC+CD1EmGklhUKl78Yh9KzWqArTAEhoN1CuJwGvmqrXn8kppSGlB3qbiR8NA07f6+PjsTOzw2V4+Pj7bhqdbLZI6W64H8JqjZupBJ7UgmfZJ1buSe1qCNCqBm64y7QIJBIVCo3VafCBw0i4wBWxwJCtdwca0uwy8NhmGgDuaQ8xKSoQWBNAUiXomBh8ggIsQUVzCFFDH981biCmB0tAn7arTZDLO2txphoAImmYlaFKCZiVIKsiTml/kCP728fHZcfEtPz4+PltNUsObSZMltkQK6CE1ZYbGxsveMvHEhaMFrUA4LTeCwls2rL15XcDVmkYELoL2kT2NWiCARi3JAxpcQZHU2Tz51Y6kSYOUYGo41nyTADYpgjRqL5MMJH3EKkydICRCrLAlrQiSeLInTDvxpqGX1GzUiibt5aO5wK8iCUL+10Qfn50K/5b28fHZKjRwbWOISlcQ04K4FqxyJWtsCdqr8ePgiaCMwLGEZ+2R6VT2FgRRAQGgEYmJYNPOXRnrj4MXB7TU8ZqiOgrWO4IG5bmpXA0JLYgTRqbrTdtaoPC25WBgSUmjFjSmx5P51pcAUsrbWIWhKDY0JVJTAOShyBPwWspKCykfH5+dBV/8+Pj4bBXLRJiljqS3oYkIjZtOC9/oCoTwAogD6GxbCQEEtCIiIKkgBrgIhPCEiMT7vTl9kcITOetdwQcpk4WOQRxBK4JG7WWJvWV/j2YKaJ88L7FZ6I5igxskBVgICoSmUGhKhSYA7G3a/D4/wclhm0olqFESV0Arklot+Chl8vO6MIts/3Hp47Oz4N/NPj4+W8VKEUTjxd7sYSn6mYp8oSkzFBVSIyQUSCgWmjw88dOMpEYJmhDYeJlZKQ3RTZ5AubpXCDwLjovXlMIF8oQmT3oPMJF2l8UJM7H1fu5JXsIidygCxWx3L25KXE0KL2vLEN64Rfp3EDgt6nJBvs1BAZc6JTEExNKWI5Fe/0pHckdzEOVbgHx8dgr8mB8fH5+tYriO8Xb6/6WAnoamp6HJF5qYgo0pr3+FEJ4FyNZeYLOB98DJWns0FApNKwIbr4WFqwVOu21l2mVkrEgGgrDwqkPHlEgXVGyz9NTRgzec43nXOQqJjUMUAUg0BprQJuoqJOD4kBfMfFDQJSo1jUpkm7UKvHglG6h0Pbebj4/Pjo9/J/v4+HSiVcG9zQGuqA9xfWOQBe1cPv10iv0CLsm0QtAabA1nhm2uKUjSRyrCQhMSmr6mokxoJJ6AsfDEhASSCGqV8FxaaFLa+yzT+sKgTfwo2lqRprSgSXuCSad/ZDpmKDN/igBJotl1RAVcGk1AOhBaaS/Q+uRwivK0srIkHB106C0VBl6MUr7UGNIL6G5y4Vf1Ic4P7M7/NVsoP/vdx2eHxbf8+Pj4dCCl4YrGMBtdryCgduF/Ggx6G4qkFsQDAzlKKH4YcZlrGwQFnBaxGWEptIYzojbvJUws4QU8t0oocj13WIa40sQgHYejs+6u4ZbDo8VxTq2NssaVaWHk1X8OoYmnXWYCz4LkIbJ/ZyxEGVtQEM8iNT7ocGm+zblRlz83BUgowSDT5UvX4KDKCJYQnBy2uTAvSZ0KUaclKe2tqUJqNjqClUqS74JrhLijyeDNhMVLZbFv70T4+Ph8a/jix8fHpwNvxE2WJQUrlEEsHcws0SxBMjqoWCoizGwOUSIV+wZcfhZNMcLyzCBCwK/zUxwfcpiaMOllKCJCc2VjiCYts6ZmLxNLkL9JReU1rkGdEvQ2NJZQrHBk1l3mIAiiiW2SEp8hSFsNH522JglgoxY8m7B4o9Kkt6F5tDjG1JTFf1pNlikDN12r6P6WIAtsg78UxXkzYfJoawA0BAV8ogzy0nFCygUz3bR1WlwyPuybgHx8djR88ePj49OBOSnJPNfARiBpSzd30KxzJXXCwhLQoCSVruaS+jB7WIoQXpXmWiWJ47mwyg2DPobi7LDNk/EADcoLT84TmpiWnVpJOFrwvy1BLAHF0ovJMYWXjp5CUCA1cdWWIdY+GNoFLwhaa5o0RIFmIBNGHdOCDa7mrLooIw2XjVqiNNlu8Crd9uKFuMXP82zOijjMTknmpiRz7RCG2GSwwH8SFuPDftsLH58dDV/8+Pj4dGCFI0mmBUNbxpMXc1OnRLYooBKaJY4kpgWLbUG9C/V41p1MBeVypRiuBFGpeb1HC8tdgwCaAqGZUBOlfX6X0l415VTajRVKt8fIWHkyv812f7dN89pWtCjPCdZbKMpNzae2mXWFKTyhU6MEG9P9vtrrGc91JljheqLo1uYgs1IGCQ2tCAwFkfb6R8OYgIuPj8+Ohx/w7OPj04F5jkG67l+20KAXVAztHTxSQ6sSSAF1ChrxsqQyjSA0UOsKGrSgQQn+1hKkyhXsZmp2D2jGWC5ae93YlYaI1JQLhaNhti1Z5kgKhCaWFh8umgblNUGNiIyLS6PSRRUzgc19DU0DkkrV0T2W0S1KQ366waluN4MGioWiTGqei5vMTHousbWuRCCIIWjR3nyuhh6G5rxw+9w0Hx+fHQXf8uPjswuiFDwVN3k/aXJk0OGMsIOUUOMKhPYeDO0tKwJvWgRNAk80lEpNlRJoDa4W2Z5eGq/ooInXOLTagTXCYJUr+cw2+FuL10ZieEChcKhzBVJ6VpUWJbA1FEtNjStIprO6BJowbRaepCb9f21p6SmgFBdLgNCCmCvS83siKNMbLE9qbi6Kc3VjiDnaQGgBAnoITbEBP4zY3NQUZKUrWe9KdHpfQnjCLqRhH8vhvuI40v/66OOzQ+KLHx+fXYwWBcdWR1nnei6q1xIWd7coXi9rZb4t6SE165UmmRYNGUnzk3CSfANeaXAoDIQwBFRrTYmAdW7HeBgNadEC67XE0AIpNPOUQYsWaDQlSc1upmJcyOVHUZt/tARY7UqEgIFS01NqZqQMJF4sj9YQ1152WDKdEN/eGqWADVrS4njuLQcYJlyWKgOB1/LC1jDUUPyhMUyZ1Aw3XBJa0N/UDDIVP4+mKJGaGUmDmnS9Hy/mybN89ZKaPZwmnulhkiMEyMfHZwfBFz8+PrsYl9eHWO9KrPTL2wBWu5I/NIa4LC9JUMJI02WeY5DEEzA9UVxekKLQcBil3uHjvAqaVYADdYQ3G0ZgYGSDkDMByOBliYFXuLBBS89qBHgFCQWOI0gi+EU0RUyLDoLC1m01fEDTqkW6EzzZ5PhN86wUgjiacLpDfKElOAKHBbZEIRhgKpa4Butsz7o02FREgd/mJxkd8Nb2pS28Yoy6Y9sNB2hSgjosXLT/8PTx2YHx718fn12MebaBuYnVwhLwacpgoKUZYir+E7eISC9jKqmgHsHJNRH2LpnGbDUG0RxGoJknXPbMW0Vjy24Uaa/Cc1M7i1FPNC3Cs9gkNklRV3ixPEsdg/9XHyYivT5hZtqV1JKuCu0CjUpks7tUziYYbdjpitGFeG45yxA0I+llKGq01/rUEGR7gpnAk7EAowMJAGqVIF9q8pQmrtu2lSnU2CBMTGFnp7tepzKkH0Lp47PD4Isfnx2CuIiz2FyMLZIMcoZQqkqznzWIBtYb6yhSxfRSvdK5Qj5dEZCgnLYUb/CCgDMxMXsHXN5JePV2UmmrTT4CbdYwJ96XmA4QkS4BASD4wg6wt2XTqg2kgGW2pF55aem7W4pltmCFa+SszZOp0tyiQWtBpSvon87yCqYrQSvaLDBbqqjTPtMshrdAQghSGlY5kmA6UFoIUNqbbkloaidy9rAURVKTL6HF9bLPMus2AInLr+tDpHAoiCxiv6K5WAjKVTn7J8di+o9VH5/tHv8u9dnu2SjX83HwY5RwEUhWmasY4gxhr9Q+TA98zDpzbXpOTb4u4Ij4UQQIbNMxb8/8NJLkpqZwh2kauDDq1aupdL3MJkMISAcc16JpThZh2fmUWvWkNFkhkVQGx4bifG5HWOVIKgyFg2SIqUgoaFKbt4iYaHpIMCQYaA6yXOJARcDlZjvspajTvqLz5hFkRJDnAluvBImMJNaCJJo87Vm7QsITfuWyTVZFJfwgYnNrsySIl11mA+VSI9E0qgAblKBK1tPU0oePYz04uWQOqdA63ODHHJI8tJsj9fHx2Vb44mcXYKGxgM9DnwOKnk4v9k+NJV/nb+thdQuN5vPA5yBApiu2GBisMJeTp/JYa67BwKBRNJKSSaqpoSncxInxk6k1ajG1SYWq8K1B7fhZnsMqN8lz8QBx7aWHnx1J8f2ol7Yd05BUAlNCQrfF6KQI4ihNZcqkT6g6uz4pJIMsg3OiCRIahIYHWwO8mzRZ4UpcDUVoWvFcUu0tQCbQz1AYaX2kgf2DDocGXX5YF0bgWYBsMin3XZ9HQZvVR+B1fm/WAqG8TC0Hr3ZQSgtiaIaZKpsFdn401WFdp0YcBkqXc+ujxBWEhaBJCZoRFKJo1knW2vnE3QCVwNM1+9M70MA5ZZ9gk8LyxbePz3aNL352YjSaN4NvsDSwOC0cNA2BBmqMak6JnU6Q4LYe4hZJkiQmY8RFnJRIEtABIjpKkhTvBd/DFilsYaPRGBhoodlgrufv4SeIu3m4CPJUlLOcQ6igcFvvznbDDYUprstP0QQUQIeU7RYtKDM1Ve6mdXIEAkVKWzjaBGyU1vQTAfazvDlDaeVxaX6K70dsJtaFsA1JRGgW2pIm5RVLlN5sDDQUfc22rQQF7G4qZtmSRlcQkJqYFrh6y+I1giYo2vp+SQEJBQG89PZ+hqJOCQrRBNAcFXSokJqzozbFsqNTrsYR3NkaYpCh+EIbNKfbfIQARwgW2WY2GNxNZ5FVO/m81ziEk0MuVi4fn4+Pz3aDL352UqplNdODH7LMXOZZTbSLhQUI6mQdi6wF7GXvs62H2S1qjVpsbCSCVhGjRtR43blRnplBAygcBcQcYqFC6p0IIW0hkbSS4l7xCdeoozsF+u7KSAlFm0xLaJiTMqhzBY5us6Rk3uUuBqCIO2GKkexphPhDnsAQnd/25YamRGimp7wgY0toyg2FqwX7BBwODzr8J95mIUkqOCTo0NPUTI8ZWBIMF5LpLK/N6QmB58aSePE8ITT9DUVCesv2NDRCQInhraWHVNxQ2HVbiodiAWJaEJTeegvSx8YLiTZwlekdDeFiCoUjXOJasTFZTigY2sxIfXx8tgd88bMT4uDwcfBDGmVjtqytK1zQECCAEpoG2bBNx9hdFgUWYCgDW6ZIChuFmy5uB0KIdIXe9CtaAmGDWDKMEAol44Ski1IWKa15J2FzbNDahnuz/fOXpmDWsRSgTfhksq4MPL05lCh5OsJN+YmsoGiPo+HGxiBTkxbN6awtSwtCQjHMVNxelKTU0IwIKKbELRwE44M23wt6kT0HBlz+FQO018w0nnZcGnjuq7b6Q97vMBpTaxoQKC0IA+ODmj8XxvhlQ5h6V6IFRIXGBY4NdqzMrFDUyhpA0EP1YIMrMIRXTNFRXvC0k+4w34ZEawjJOIZQuFrSmirvou2qj4/P9oQvfrYhSZLMDnxOk2zEwmJEahTlqvxrr3edsYaEiGNi4rWR1J7DQiiUVnh1dW2aRBMFuuAb2JNvh5iIsdhc7NV5yYicdm88jYak46UvtSsQI4RL1GgmZKWQWoKpCSiTL5rcnU78KAXzHUlUaAZ9TV9LUsPMlEECwUDDoV7EkSianDAp7blIXcDSmlbhCYx/xywuzE91WtdjrRavJUxMAfloEunA6XxgkKkoTQum/QKK/QKdLTAVpua4kMOcdJHDzE9EeK0sbO0VIOwpFC1akACaslWFwAK+SEn+r9nrzL7SFSS0ICI0P4mmODnSJn6qZTUzgh8TE62AIKSirOMovkiVI4BWvPT6TOC1QKeT2730dq2CpJwikm6EoaYC4l/rPPj4+Hz7+OLnOyRBghpdRd3LH1E7Yw6rC9YTPGcMweF90WjeD/2XcYmD6KP6fK3tuOlXQEiHCOoQCdH2ME6SRCJZbi1nmbWMMrec4xInILRgpfElZW4ZpfSgVtYw3/qCuEgw2B7MIHcwRrZF5LfPSvNLPgvMolk2kRQpbGwMDAQCB6ftq7+twFUQTI8t7mBZKQLSq+Ci08JJYVBeMAvcg7+zffi2mRqX/LYxTKOSCDQVhubp0lb6dfOuTmq4pznAXNvAAda7gpW2xDRsDil/lb2tesISGl3J0tbBzKw7mBACU3jd3QvQ1KrcfsTPbQMHr2ihhefy0njB1d0tjfyLvBRzbMmbMZME3ul2tGeR2s10iWvB90IOq1zBvITBhrQyFnh9xlpdydLWIEUChlku4XRc02e2yZS4y/yUSX/LIVLyMa5oC1Ke79r0K5nGZ+vPRGuBm7blZKRle6uTQtCqgwScIH1MxW5+sI+Pzw6BL36+ZWxslliL+SQwk2bZhN0aQx9vI/u1IG9bTHDWcgp+exzhI/ZAIFgYmE+fxNcTP8WqiFbRihKKAjcfQ0rvbxQWFoY2cYSDQLDBXM8j0X+QFEm00AgtsLSFRQBH2GhghbWc3k4fjkkcS0RHvpkDsxkUinnWXASCAlVAjVEDwnPnZfowGRi4uGAKz/qjtFeSN5ZCVNbgjihNF8vTKB1A2yUUB+p2mi/lCQX/ryFCXHvuGfBq5JxTG+WDitZurePPTUFm2wYBAescyXpXkBQwpuhjCqwGLLMV0HzZtBdxJ4+egVqSdg9cDWjBBiU4KNh1Y898oWlSmfF5aA295Jaq9bSxu1Q8hydAHLy6QKbQlArNWXlJzok6/Lo+hC1Bp1ebaXLqpEWLg2aZIxltKVzgo6TBB4kIQQFloQaOdhOMEN5xiGto0ZKo2czwQB1zE6Vk5I6BRqXbarRHIGjW3jmZELLx8fHZ/vHFz7dIiiRvh99ig9hAo9GAtl2wbTAEap8y1Hm7I59cQuN7c7CP7I1FAFN8vVOyRq5mZmgmAQI0iUZajRZPNOCJBhsbRziEtVfnxcXFlt4DWyJBQFIkSZLMCh0NVBuVzAp8yqHJ8V9rfN2hSTSREHEc4dAiWrJp6gqFEAJDG+hMMErQRDckaawLoBxJ8cIVFNoaNaIHjt0TVwvyhUFvS2Gq785y9W3zdMykWQkC7bK0pICNrmCVAwO2cBk1KPgiLXwAGpVX9djUUBZaT0mwkpSyeKvyOJqdQgyhSGiDFg1hvLTzPoZmfDB39Z19Ay6rHEldukqyITwDXd+A4oJoZzdZLmocwd9iQcLC6+eVSJ/0AJrTIynOSafm9xGKGtUxHqctUNsLdE4iSAKLbIM4gigaLSChJM1asNCWGALiStCkwRCShY5ncZJk+pTljubJXIpxBWXSt/z4+OwI+OLnW0ChmB34nPnWFzTIBlKkYxokEDYh6Xpvj3G9sQMG9uBibNGAIUySIkGKFBJJo2wkpENEdbRb29Vo5gXnIfGsN4Y2icuE9wLQAqklSig0GgcHExNb2J3WsSkCL2C6STZ+vQPTTYIESZCkTtbiCjcbtwSCgA4ikTjaxkASswt53T2dxlAeuJA/poWje39KH6uZkmxxPYWLS1+333cy/u+C1i5SvzWChG7f7jM3jcrL5sqIH5F+g0ugPFiLKTSfNu5DSoWQwsvQyjStkFoz2HJ5oDjeoUp0e34Ysal0JRawxvWEyX4hl9uKExR0UfNQa3gxbvJO0iSuBKtdQVx57S4CQEBodHrMsp3rbJmSuDnq/2i81HStPfGyyDaoUm2uMYD6VCn1dgFxI46R3sMmJ8zSluHUOoWYtHWN70rWZNYVQzIlbjEyRwyTj4/P9oUvfr4FPg98xiJzAXWyFke0cwtI4ZWxDQpIOBC10EOKEBocqbBJoFA8HP0HhpCEVJgwYUrdHhySPHSLZfNTpLLxPXVGLZ7M8R7dSigv+DeNIxwMbbYTFptHIDD0d3O52NjUGW3Hrv34ClQBIR2ixqjBVoKnN5xBvR1FCkVhoJGEzueNNfvzw6qHqTfWERzSi4LSXuzmDmKkPeo7Gf93wblRm7+2BDv0uVIaiqRid2vLbqU+hqZYalrTy5dJzUpXYAGOChBA02AX4abPuRDQ7OTj4gU9/y4vRYXZ9XVjCPh9QZJ6JahyBX0NRXQLra+ejFk8F7doUrDGNahXXlHEoIKo9JxNQnjurJFWm53nk1RbU9UOY0Czh+HypTJI6rZgZfAqWOejiWnJf6uO4tCydymymliX6MmchjHUpkrJNGTdEi5eXJOrYZnj11Lw8dkR8DvxfcNoNKuNVTQaTZ1FhUhXYDOEJ4Sak5B00b3zsiIlJVK0GM2kRIomowmFosrYyIfBD7YoUiwsTG0SE6246dfippWNhRJgK1TchqRDnsrzpmeDRdu9TNNj0miCKshuzm5f48h0nzmBz1FCdRo7QFIk0ELjaHh+44lsTPTA1kESbpiqZAWJuCBmR1kb7kPRHYsJ/uJt9ngxyN6pfXKub0elSMJv8hIYQEp5cd8RoZlc1L2gJlPAT6Ip0F4QcbHUlEtFVCjWJHrR6kQxcXEU2Amoq4uSalLQmqRAOxwW6o4s8Na7u7Vl4aM1vJ0w0cCXjkGLFtkMqxQQS1tsXA29DMXe7QRenZI5z6wBFBpQKhX7BVxGWS490v3CNF7dHgU0OSW8veE03ttwGtOqj6I6VYHayu+FFl6PsDzf7eXjs0PgW36+YRSKRtkAaCTSC8ptjym9r+jVMVjaAGN7Q0m403oc4WBqk0qjEoGgxqihRTazd3If+qi+ObctkfRz+/O5MQsXh1S7mjiZsfH4AugbwRjSg5ROEFylKRs7kFqrDiXSL5T0IgmRIKiDFKsS9k7uzTBn969/gLpBs2xGpFtaZgSLTv8TCExtsrR1MI12YdbtIgRoJWjSJZRQk7ayCaykwZcvv8KQY477Tsb+XXJRvsPZ0WaejVnkC81pYadDDNCWOCzkMtqK80LcxEZwbNDmD00hiI1CBmuJihQbKotRRiB7TTiuZuOyRtb2ymNINyxM3cUFYlpQ7QqatMgGLQs8gaKAoNaUGpp/Fsc6JIxJrXA2yUQUwFEBm7tLklzVGKI57SYcZClsR9CqvRpREk2F9OoItTiFON2oJL0pEs8l118qeu48YWU+Pjs1vuXnG8bAwNIWGu0VFtwEoSVSmOTv1h/ruOE5hQ94gchJkSIlkt4PKSplJTOCM9piiHKwd2ofAm4QW9htYiZD0oGTdkP2LCS43iGw0SYRtil9oY7zmn+M6VpYOkAk/S+sw5ja5KzWcxjpjP7OLCeZ7XpOO++fxis3HFYRQFOXGEieUASEyr6YvQwfgyit9K1eiRFLW9NaWr6TcW8LiiT8Is/mrOjWCZ8MJYbm53k2F+alsBE0KUky1Z8NDYcSXFKDqop7142rIOXCxiZsV3NFlcWGLlw8Om1N2hpMAWWGYq0jSNGxk3sQTUBofl2Q4qWyVvq1SyfXGiJCYAIShSlsDBwMFPtZLj0MTR9DodKLFBiwT8BloOFyTiTFb/OSDLEUJdKLJ+qOC7jDuIECoRlluZQYmh+EuxfM7ePjs23xLT/fAvsnx/FaZEqXD1ILi4RIdAo2bo9uF68j0v9aZQtJkWC5uYzhzgiWmktYa65BIOjvDGCQM5hmmmk2mrCwsOm4fulKECaiR4RUc6M3PgUrytczzGjAlIaX8ZVGIHCEwzJzGYOc3ZgdmE2NUUWLaKXALSDVyyZpJRhm7/6NdFHXaOZZc6kX9Wit0O1aJggEYR1miDOUUfZolCrgH0rTz4qxxo6S0p4zI4DNkU2v0OPpdd46tSbas+fXHtuuQKnUhNLHvCU5kDUrFKJQozc0dZC9ImDQnHS4rzXA3pbLBldyVMhhmKl4pNViWtIkrgVlUrGHpahXgt6G4vSITeFmBNqEoM0zsY5FKDPb7SkVv8jrLCwaNYSlptyoQYkUjjYJSBtXSz50w/xGSy6KpriqMUSjFgTwXGdHhx0mFSQRQEELfJgyWIOk0ZHpBqqbR+JloOUJz7W3m6E4L5ry6/z4+Owg+OLnW2CoO5RPVA+qZFX2BZ4JLNZobNJWmfbV0jZBaokr3KwYUcLLWHJwmB2YTZ2sZ4O5Plt4sDZQS6NsxNVO2mVmdRJXKlPlzRYQNTFavc/diOSD0DRSpLJVazPVoQEMbfBO+B02GhuIixhJkWStqREVgiqxkfnWF0yIn0jeFjrFt4hmvjRXENQhdnN269T5elbgU1aaX5IgQUp0ftGVqlLWm+vZJ7UPJ4Vd3kiEaVAhepnQrDQxkeT46U8x+F9TMc0gynEww2H2ufDCzY7Lx6PE0OxlucxKp8CX7d6DFV+24FoGOhsqDMJ20JEgz7VKXpYmrUpyZ3OQUunS04CQ9Cw3U5MmryYEewe8dU5NmtxWmKBXF4HSq1xJUEBKdxQfGhgbyJ1SHxEwwIqzRMRxtcAQqay7LBydx8fJMRwccvlbcZyXEhZrHcGBAZeDgm7WZXphfooL0u04GpRFtYJYuyyvTYni4iIpkZrJxXGODLgdGsP6+Phs//ji51tAIOjj9KUh0ODVptGe5SYhEum2DNkZc2Jqk/0TY5kT+hwbO+s+y9TqUcJlYWABpao0u4yBwWpjFcVOMS5ubpeb8FKhEUDfKO6GGCJuY/QvoVXGkHip8A4OWmtMLILa6/XUKOqJi1g2hkmj0VLTKBppls38M+8Rftp8PhFyF0FcYM5nYWABoFFoFgbmc1DiUMpUGeAVMFxnrEUgqDaqOh0bDdQadZS6pbTIFkpUkDuK4vyjJcBqVzLYhDPDsPcpP2R1QR82fPIJkfJydj/jDEJFRd06bz5wZUGSh1oCzLYNinvn0/jFRuYXF2XdjlpIVNBgtmsi0FiuICq8a+tL16BRa/YKKCqVIK69Cs+VrqSPqUhowd9bA1yXo6FoSsMHSQOZruDc3voSAn4Sye1O+jhp0GpuwLYjGEJlhU/EbKV/ZDUfN+zPwSGXqIRzIl1bWmMaFjsGYQmqU9Vqnba9ei64Mp2iT9AiKmA/X/j4+OyQ+OLnW2KQPYT51hcdWzKk0V1+pwQDk2K3hGJdQplbxlpjXTbrCjwrjESSEp1fINVGdbYooJMrSVe0U14REwblo12NCgUw0ER0hFbhVQdWKIIqyAnxE1lrriEpvZePEqptPOl3hEIRF3FejDzPObEfdtpsXMRZFFiYtWJ5acmKWYFPOTZxHAJBSqRwhEOraOkcq5Q+ajY2FhbRdIZakYTfFmz6UjQYePTRDDz66JzHF6BFtGCLFIWqqIObz8frYN6+V9ejJw7kvln1LKtKoi2JO7AUo0ceLiDTqeAJrQmnr4Um5dXmaU5XdtZ42VrgxaBvcDseb63h4VaLd5Mmn6S8AoSBdCXljHzfzVLsF+p8TSy1JZObgwzMa0JblWxM9kJrgxKzkf2K5qO0yWCze0HZSS1Y6kgalCAiNa7yepEZQAGaGIKQ0ESEpq+bIiotXA0qW23Tx8dnR2Kne/LfeeedFBUVccUVV2Snaa255ZZbGD58OD179mTChAksXLjwWx1HX9WXCrcCS1tIJEEdSjuUNn/IXe1gYPBFYC5JkqTblmfRQtMgGvGK7etsXJGDgy1ShAnT2+1DQKczdNI/UkuklgQIIAMBsAwImBC2UEJ5VZ1FkoAOENABilQRp8XPoLfqTU+3J0KLbLHErqgyqmignkXmgnRfMC/tep2xtlPWm0AQky3ePuIFOYd0OJ2in3sbAujv9MfA6CAIu0uKFFNDb/N6+FXeCr/JK+GXWSNXb/V6diUqlUG/kRUcfHh/8g8YQKAsHxONmU4ZF5CNLMtc2SktCAiN0t7l2yOd/q015G+SCj4lbvJy3CKFwJReun4KgSU0BVLTR6pOHdgzPBuzkAIa4sMZlbeSQ0s/YHzZNPYqnk1QxrETu3F8uHsp+SaauG7LMIsInc0f291S5Etvf4vT41ca+hoqZ0d7Hx+f7Z+dSvx88sknPPLII4wcObLD9L/+9a/cc8893HrrrUydOpWysjJOO+00mpubv7WxSCR72/tSokopc8vpoXoQVdHNigdoC4aOizi1Rh3uJlYQhSImW0kQY62xhiqjkibRSIwYQRVCIEiKBEEdIqiDmJjk63wsAgQJetN0W1Bpe6HhpKspe5alFAXK6/jez+1Pb7d39rOuUELxUuQFvgjMY6G1gNfCr/C6WsdzLQWsdgStm7wnpDayhRsFgtGp0YR0CDNXMUUNg1KDqZW1vBx5kSnhF/kkMGOrRNCM4HTqZX02I88RDp8GP8kKsOxxSCRY9sorzH34YZrWrOn2+r9pHByWG8v4JDCT1caqryT4vi7jAg5JDUHpBUTnS00IRb7o6JnUQB5QIDWW8NLHwbOabHAFc1IG82xJoVAd9Px7SZOAhFpXEHOhRXtXZFh48UcFUrPUFkysC3NjY5CqdhlmMaBVwZJUiE+qjsVxikBLHDdIND6Sq43hhLqZoFinJX0MhRReQLQEQkITTIuiQaZLuVQUCk1SC0qk5vf5fiVnH58dlZ3G7dXY2MgFF1zA//3f/3Hrrbdmp2utuffee/n1r3/NKaecAsC9997L0KFDefbZZ/nZz372rY1piDOEUreUxdYilHAZlziQ+YH5zLPmpGN/MrEEAoXC1CYWVtoKkwDRvpd0RzJFALXStIoYJaoYE8MTQjKWXqtECRtHOxS5RV6/L9mIQmFoI2tByVRSdnFICo2pLCI6QkIksLSFQHBE4ihK3BI+Cc70xEKOl4pC0WA0YiiDiI4yrXEos1sDjIysp1zEWOIG6CmC9DK8VP4+qk+HqtX93QEUx0t4J/Qm640NxGXM+yB9CJYEllCsiijURbhCsdJciQYOSI3d4rlQKGqo4emqA/gi1heFoH+wlp9UvMcKczl7OCMAaFy1ivevvZZUYyOYJm/NXkzs+NM49sgD2d9S3W1I/rVZI1fzXvhdmmQ6Ky+lKGoIcbw6mdLCr9f4dms4JOjyZtJlXsqgTGpqbYGhNSNNl1WuZL3yWliE0fQLKH4cSVFiaNY5kl9HE1zVFKFJCSyp6SU1s1Imj8U0BwZcno1ZTE8ZJJWgTolsnE8SqFSSqKPJA1qFBA31tsEVTSH+VhQnKiHmwkLHC/lvTJWypvIkeknFDyM2F+dvXcp5L0PR09AUSpfadJ+wUqmxgGsL4gw0NSEBC21JZdNqji7un/NaeC8ueTFhsU/A5Ydhx48H8vHZTtlpxE9G3IwfP76D+Fm1ahWVlZUceeSR2WnhcJiDDjqIGTNmfKviB6BYFzMudWD2796JPrjaYWlgCS5uNr7Fy7CyUGiU8IoTdiqQuAkuLnERJ0iQBqMBgaQ1HfOj0dlMs4RM4OAwyBlMuVPBCmsZMRnHQJJik5dEumt6VOVjtbMQGRiMsfejhypnevAj1hlrc8bmKFwaZSMJR7A8FWHfsncIaBNHRQlatdQpi966lH5uH/ZL7d9p+Xydz7GJE1hhLOOt0JsYwsh2cLeFTZ1RR0rZBAmQrwpYZ6zjs8AsWkULpW4Zw5xhXbYBuWfjYSxM9Mp2GZ8f68Pta07iid7LsvPMmjwZJ5FAR6I8fdLPWVcxEOE4zKiCYdEQfypKdNua8FVoFI2sMlcy15pLs2wCBc2tClsbxKxGnv/oTg6p2p89zj772xtEO6SASQVJPkoafJQy6WdKVjfHUCLKYEtxhOEw1FAg4OSwTb9sJpfL2wmDXoZit47Z6zzTavFi2mWlEKxUMl3Yoe3AhtFsdCXD2hVSlMLrSfZSwuLEkM06JSmUmua0tUhoaNaCs7vZOLU9IQEnhmyejAWyzUlTwNEhh9GBti8gewYUS3Wyk/BRCk6pjfCF7TnLXojB5GbNKz1aqNhpnrI+PjsPO8Vt+eijj7JixQoeeOCBTp9VVlYCUFZW1mF6WVkZGzZs6HKdS5cu7da2uztfewYzFKfcZV2PNcQDcaSSRFJRYsFWUALH8qIoulNwTQuN4zpoBUIrdECjlW5zaKbThl2t+FKs4PBFR2IX2yzutQhb2rjSJeP8klpiKAOFS7I5zpoVnV0+Gk1saAyVpzpbf7QnyBzXpdmN0Se6ALRCKQ0yidYgRJyqlmZ6bujJ8vjyDm43hWJJr8VUFVTSHGzClS5KK1yl0KKt7k9CxUmSoEW1oIWmKdmIgcEKsYIvUnPZc/XeGLqt1K4jHOoNxdLgXsh0c0zwDlGVXcCniwuw3KUox2HjkiUk9y1k9Tl70bfPeiqSlaxd2YumdYXM79WLu+obONOtBr7aue8KjWZZxVJqCqpIKZtm0YTWUBUvJUnA64wSKmJO793RV9xEvLSU/IEDv7Htb4kK4LT0/zdhsMIJUa5teus2oZEA2h+RT40epMyiTt3Wl4owQ3UcARQCWhZkG5Nm6nm7eO6xxUlYi8JAU65T5KH4vLmFAreRtYH+uMJEY+Ag6KFTFGmHWStWM0AneUcWcr/VhyZhEkRxoNNIP51kjQwR1Ipj3TpG6Fh2bGOAqAjzrlGMCxzqNrKnbiXXWd703D9ulDHbimC2y7KsduAn6zR3p5ZtuvgOzTd53e9o7Kj7PnTo0G09hO2OHV78LF26lEmTJvH6669jWdaWF+gm3blYli5d+pUvqqEMBRuUrVhtrKLeqMNIGUwPTSeG97I3MHJnbW2CNjRI3VYbp72pXUA64hnbsJk7ag7jE4cxJjmGj0MfsVqsyrrdpJAIw+sIPyy8e5f7tiA8n3pRh9KbCKBMqwkDgtIhqluRGISsFCGrAYGLFDaxohSfFzVS5pTTKltJiAQBbVGkiomLOEmZwBCetUcLjTI6tukwDANXuCRJEtRBAnkBYqI1m47tRG2G28NxcHgx/B+qzCpsrfmR8zEfVB3B8thwb5xaYGBS12MwQwtstNbM3bcXDeeVYfcowBAaI+QwdI+VrK7cDR2NUCNDDC0u+lrnPhcb5UZioRbyKSBBglbZSlwLQoE4dsJK75tmWWAUg4p7E500iZOfeor8Pt+dCyzD0qVL+f7g3luc73RbMqMh1KHytKMg4EjyAlEA1ruCoC2ItVvOyyKTCOE1NtVSYmtYJyz6S8UxBSa2LmJDQ4RMs1NXQRVhCk2XffsNoMYV/KU6isKzGKU0vCTL6Ck0I4KKZg3/pJRf5iU5ItQmWIYCJ2b/imb/b4MD1UoyylQsX9753E+vihBwjE5d7ivNfIYO2HlePN/0db8jsSvv+87IDu+RnjlzJrW1tYwbN47S0lJKS0v58MMPefDBByktLaWkpASA6urqDstVV1dTXl6+LYbcAYlkoLsb+6T2pafqRaEqoLfbu3Pg72aMQJmsr80FxGYbhQqYGZqBgcGE+EkMtXenwu1Jb7cPPd1eVLg9KVSFDHIGd7muhIhj0bXQVGjCCMIYmDpAwGgGNFLYCAFGuufZKnMl9UYdSrjEZZyV5pc0yHoQZEsEZPav4/YT2HhiJSESVMkqkiKFLWxaZAvzAnMBmBJ+iY3mBkBjCggZCY7s+QbF0stqCxIkgOSY9MtPCAE/2ANtu0jV/lhK+vRfh9YQ/JZcXivNFch0fpE3rgAKgSEdlJQow6BBldBi57Hw0BPQwJwcls7tiWGW4uCgFzCdaXlhodmjXUd2Oz2tPZnO8SG8b2caT+AoJajXgkdbg1zbFKJFCxq0oEkJkgjiCJY5BqscwQ2NwazwUQrqtSek1mjJB0mDOiUwBTwX2/wXphYFx1VFGF+dx6k1UfauyuNlWdJpvojIfYNuum8+Pj7bBzu85WfChAnss88+HaZdeumlDB48mMsvv5whQ4ZQUVHBu+++y5gxYwBIJBJ8/PHHTJo0aVsMuUvCOoKBiYFBuVtBq2glQJAm2ZiN48lFQAdxNtMqI4NEprvEV/F85Dl6qp70dvoQk600iUYSIgloBtmDKWlXQHFTyt1yao2anJ8FtZdRdnT8WBpkE+/KpdhCIYWDFNpzrWGQ3KROUUaYJUgQJpzer0Cn+TZZCMDbd92W9l8naqkRNVQaGwFJXHuxJQaghcuY0qkA9IysISAUy3UPdk+eiIFByd4jaV2rKGmpZ104ipQCMxxBawdHwynfUu+mAMEOAfAlbgkNNKC1ixbQlCpg5sax1KoSvtzrQKLzptK8GbfttuK/CYPXEha2hlGWy6/yUhzpOLyTMCmUcErY5sOkwb9iASwB5VKzAgjiiRwHr7aOQiDR7B1wqFGSlIZ8S1PlClLCO1J5QtOgvcrTFl5/sDDwt5Ygda53caQUNKfXlyGpBZ/bBn1dRZ0huLvJIk9qzok4hNJfB7WGWbbkl3Vh1ruSoARDQFwL/i/Ql3OcOL3aPT1/n5/kvDqD9qZQR8HR0S3flz4+Pt89O7z4KSoqomiTCr6RSITi4mJGjPAyeC6++GLuvPNOhg4dypAhQ7jjjjuIRqOceeaZ22DEXZOv8ylVpdTJWiQG+TqfMCFSeI1Nc4kfU1sYWpISXiVp3cU30Ew39BqjGo1XJTohEiy2FhIXcZpkMwKIqih1Rh1LzSUMdYah0Z3S2/e092a5uYyETHTchhaY2mLf5P7s5u6GdjVB02QGDTSTRKXrDHUl5DJuPoVC0tbeQ6PRWndZERu8uJ6McNBopobfxkHRrESHLZlohhTMRwCm0FjSZhn1PGk+wXmtP6RMl1M/sB8RFEFtU+2EsbUiFs9jXPEsRKSRFntU1wP5igy392Cl+WVWAAUIEHBKea9hD5ZUDiURN71PNNiRPGL5RRTmb76dyHfNszGTJ2IBTDxLzYqE5M2ESYuCRbZB1Kzn/eAs+kVaOTTQh9UtI1HSYrkjSaRPkg3YWlAgFEEgIqGPUDhAnStIasEGVxARmiotsrWG8qTn8gxJTbUrOCTkMK/FIEnnNhWZjvHrlGS10syzQwSBO1o0dxfGGBtUXN0YYmVKssKVaASOgmg6ENoRgjuag/yluE2Yjwspfpef4P7WIC1aEASOCjvc3KkIp4+Pz/bADi9+usNll11GPB7niiuuoKGhgX333Zfnn3+e/O3s5QFwaOIwZgU+oSZtWTGQXRb9Q3uiptgtzmZe2drOpq5n3GZaaFztiR1PyHgd5720+DhJkSCkQwDEZQJTmXwc/JCF1gIUmkJdyNjkWKLaq6zcQ/WgUBfRqls7tNHQQiOUYC97b8Ab2whnJL3c3rwTfotKY2OHtH5b2B2ElaG9+jsGBimRQuFZihzhbFb4tB0OTUAHPOGEQ0wF0SJJ+4WlsAlIt912vc9qZBW1spbd7T2YHficetmMrSEiGzCFYKTlEBQB1qBYZ6xlQHDQlge0FUR0hIMSBzE7MJu4jGHjYApN7/A6ivo28FnlXtSlehBIxGkpKqXOijA+Xbphe8DV8Grcwsocag3zbIP6dNq4QNHihhngShoCMYzQQg7PX8j3Esdg1pey3JascCSNyquknEwL+cVJSasQJBXE8axC6xyJ0J4lJhO87iivOnU/Q2EJmBi1eaQ1QKyLuloZK5OJIJNekHAFE+sjXBxNsTZt7ckkA9h4cUOBdH2j5k4tMODifIcLow41yqs+Htjhgwp8fHZeRENDg++U/op8FwFwU4Nvs8haSEImclpLQipMmSrDEQ4Nst6r24MDQqDTIiNXk1AAtMBAooXOih8AGxsDkwq3AiNdSTqogxwfn5AVDc+Gn2GNsRohRboqs4ehDfZJjeHQ5GEdRFtcxPkg+D5rjTVIBBEdocqoIePgUCgsbXJq7AwiOswqYxWLrUVUG1W0yhjdaSEg0/+iKkqxKuHzljLyCuYihYPWBlpLXC0JmjGMHGrq+NgEEjLOx2IZG7SDKT27gWm0EHB6MMjwjpFGQx2cHvrmLYcazQa5no9CH1KnDOalJFGzAVPaLK7Znc+q9qWREg5MNXDfkCh9umgS+m2S67pvUvCLukg2BbzKFXxhGzh4Z07ipn8r+oTXMyi6gjOKVrC72x/RdCAXNkRY48psheVCoSgUXqxOnoQ6FwwErWl3lyEgrrwMM4UgiKZYaIZZilGWy81FSS6pC/NmwqRmC4VFwRNVnrsN+houw0yNEPBR0vB6lOF9U8yTmpTj8lpFghGB777o5LZmVw763ZX3fWdkl7D87MhYWERUhLiM555B6HR15D2plbVUGhvRaGplLQiNg9PmOtpUQAivlpDQAhc3K3Rc4WYtKJBuRSFibJAb6KP6kCRJnVHboc5PZs0KzVpjLWuM1fR3BwBepeLVxioKVAFH2UdTqAoJ6TASyczADNaaqylSxRySOJQonnVplDOaYc4wHo8+Rkrb2F0JuPTGDeEFC4dUiBJVSky0Eg1toCXRm4DRiCFT1LaMJu4UMKTHm50bp2pNrVFDg2ik0gmgsGjVJhGzHi0ErbKFpA4RFN7xSJqJHAP5+ggEH8vFfJT04mZKQhuRwksEH9JjKSX5Dby98XgWhSr4Zb3LxLxUt1s4fJvkCSiUmqa00FizaQ+vtBTWSFxtsLh5GBvCtfS1Wllkm9gKjPRVJPFcXxvTF9Vww6VZG0ggojU2nsUngSAfRVR6lqIUgpiGqws8d1RvQ+ElwostVMxqQwAxLWhWmgID9jRcPnMM7Iz7TMPxTi0jAuGvd8B8fHy2Kb742c7Zwx7JamPVZueJizglbgmj7T1JkGBWYBYNgXoUGi28uKDNiQdDGNjYSC2z6eJhHe7kbovLGChYEJhPgSqgRbZ0ijFSwqXaqObD4PvoJER0mBnB6cRFHAODNeYqerhljE8ejkRycOoQNq2zmCFAkP1S+zMrMIu4iBGTsZzzFbgF2NIGAaWq1Ks1hEsvabFUmaTcUnAhEqyiunl/Ivoj4qI564IDCOkQK0Qla0UdhgkJlYfSEhcTiwQKzQZiDEx3rTfdwLdy96yy4UUXIpakwKxHCs9pBBpDaKR02b90BtOrjwEheDpmcVTIIfAdVZ7uCing1LDNP1q9mJ9Mc1KTtt5f0KY5JZrPWgdySH4j01MG+YZnyRF4LSsa02JDAHNSBhKNKQTJ9OXmBTlrhpmKItNzuwkgIDSR9EYmRlO8lTCpcbyA6M2FHmeOclh4lajrtaAATZ4Jh0iX9Y5X8HBiforYl2vxkuJ9fHx2VHzxs53TQ/Wg2C1ho7ExZ9xLxiXVW/WhSTTyevg1qo0qbOy0ZUagM60wunAdZQSA0poAXr2dTMZVBgODPm5fAFpEM0FChJJh4uHOgsQRNhuNjbwSeQlTW0gEJaqUTMuNL80vaZbNlLnlDLIHU6bLuoxrGmGPYqOxkXpZT4r1ONKzcmT2R2pJCT0IOgEiOkKAAGVOBV8E5yClYKjpss6VOAiCuFxUtIr9W37O66FXWGuuJSmSRFQh/605hBWpYhwlCZst7FP8CYVmKwk3QkB6v4usGCrtHuxT2we2XOpmq7mwLkKwqIA8qxFLZkJzPWzlHcuw0QrCxsSiTgvWuLLb3cu/TSaEHUqk4g+NYdp6nev0uZKAIigTafeXptoJcE/lAWx0JT0NRZUysJVn0dG0NRltyZQ8SFuVDNosjTWuoMjU2fifFhfmpyTDA4pyU/NCj1YOr4qyUUksRNYNR3YLbQg0YyzPyjY2YLPUMWnRglJD88v8FIelSyLsmGXufHx82uOLnx2AEl1KmAhJnejUUiKsIxwXPx6BSDfurMPFbReIvOU60ZkMo5AO8oPYOSRFgg9DH3jWoPQraI/UCFIkmRuYQ62sJS5ihJ0wCR3PmWGmhCJJEgcHA5Mm2US+yqdaVuMIm2bZxCpzJZ8FZ9HX6cthicMp1EWd1mNgcFTieyw0F7AxvIGADmRfXka6Ls6Y5L7s5u7GRrmRhYH5rLCW0SibKFD5FEqTQukdMwfFIDefhI5ToXqRSIVZoxN83DSSpckiDDSm0MScCB/VjGdCrxdIqQhzG8bQI1BL30ADhW4hI1MjaWnNbYUCWCvXsCAwn7iIEdJhhtsjGJB2AW6J1a5BS91YynpVo7RIywCNoyxanXy8wNwwhWYzjsgnpAMUd5Hhty14LWFRLDVlhqaXq1loS1T6+jKBAiOB1BaOChBOluCIIJWuoFBodjMU85XMWnxMPIuMt3ftK4F7aGC1lhiOolxolrqSAHBVU4geUvObvCR7BRWzKlr5S3OAR1sDNGtPADlpue2m1xfCyyxLAXtaiqsKbDQ2cQ0RwXfW083Hx+e7wRc/OwCDncEsthbhChcHOys2QipEiVvMrMCnFKgCGmUjKZHKpohnix52/pLbAY3G0hYFuoDPAp/Swy0jrEK0SK9P2DB7dwIEeDv8ZjY4uVE0kgxtPo1XoVBCodHEiIHw+pY5wsnW+9FoqmU1HwU/5LjECTktQBJJiSr1Ch9qAOEVJMRrp9EgG9ig1/F6+HWS6cBwV7vEjRg93V7ZY1GgC0jIOB9aH/Ba/Z4siQ+jBU1NsgcRs5Ui03OFWQJiKsDCxpE4KkRIwvDIegwM9krtQ5kqY2kX3/9rZA0zQtOz8VIxEePT4EzCiRDlqmKzxwugv+HyqRPltfWnMqroM8YUz8DRBjEnmpZBgoWNexISLkobjLJcSoztQ/wkNCyxDcz0KSw1NIcYLq0uFKV7cM21S2hJd19ZIiAsNaVC0agEYakpRpHEICS8oogZIdR+DzWe9ccTL4INjqRKQIHQDLEULcpLh7+jJchDgTiWhCsLUxwcdLm9OcgSR1KlvDR5CUSFpkhoWpUg7kKzAdc2hjg7mmKUte0taj4+Pt88vvjZAeihyhhm784qcyUJkSAmWtMZVoJ6owElNDVGDfWyrkOV56yrS4DQXruArqpAG8qgTtZRJ+tQgfkA5KsCgjrA/MAXKK0optjr3J7OKnPN7gXaOjhehWnZVo0603xUIHCFS4tsoUk05rT+AARXJ3ArWnDTL3ppGFjRKAYGA5wBvBV+k5hsbQvSFgIHl4SIU6SKKVLF7Jvcj3fCbzGjeQiLYn3QMoGJg0LQ5OQTkAnCRgpTmwSEi2W4HFr6MUa6mkyhKmSptYSyZFnOMQIstOZ3EnACwcLAQsoTWxY/95fEOLAqn6QO8nn9gSxoGs3eRTMoshpJqhCzG/alIVVKRDicGtX8djuqI+Pq3Dl5hoRzozbDTJfTqyM0IzG9Ru20KkErBmNMm43KwMQLfLbSFZk9wdcZhffwCqMZbilWuZ57ap5t4KTjf5Y7ktfjBidFPCvoISGX1W6KV2IWH6S8IOgSqRlsKRbbBs0IvnQN1icN+krFIifE1fkJ9g36AsjHZ2fDFz87AALBuNSBDHGGMNv6nPXmehIigSNsWmgmZrTSy+2NaPevfXaXRGJggtY5O7ELLUjKJBJJQAey26w36rINQpOkvBR4LUmJVHY7Mt2qYnMFFsHLpvIyykwsbWWDjR3hEMyON8dyaBaJ+XxY+QQ6ZKF6B5E2KO1it7Sye2hPSnQJTbKxQ80gifRSk1U+E+InAV5geJIEi+IVJLMOD4klbVJOgBYnn7CsxcTEEQ6jo6uI6DAGBnkqP21B2nzekJN1qHQ8f84m4bbr5FrmB74gLmIEdYjdneHs5gxigAXvljXzk7oIVa5B0I2wpO4IEjjo9LkLC0VUWESlvc0DndsTldDXUKx2ZYceV2E0hwYdwgKCEmLtTnRSQwyYblvkG5q+lkI7XgFDkXaXSbwAajtH1lZfQ9OqvW7v9Uqi8QpZRoXn0nq8NcCJ4TgaWGxLRluKU0vjPB6zeCVuERDe9EblLWcjiCtoUAa9leJfrQH2DX47mX0+Pj7bDl/87CAIBGWqHC00+Trfi+1pV2BwjbGacrcc13DRSmcLHRoYGJhEVISYbMXB6VQt2aumrACFpax0gUGFm37VmJgIwMEmIVwCmb5e2itsmP3dBTJdS0i6kjLVg2qzhpRIeqIJQQqbJtnYodZQhkXmQj6LTycVUgSqbURSoUoDBJe2kje9kcPPvhjyvWaYm3r3NGDptt5NXryQSZ2SOOh0arWgyKqn1q0AZeIoCwSMz1/FIBnCUm2FMB0cBtld9zwDL0C91qjxxGYaF5di5fWDWm2sZq41hzXWakxtUqgKUUIzK/ApQeUFru8egOk9vZiiBSnJaTVRAqJdDyrh1bdZ63w95aMUPBwzeTYWJCQ0F0STHB92v1Z8y5X5Sa5rCrHe9WJ9iqTm53k2+RLiGgaYGtvRtCgvLT2VSYEXmpQWLHcM9jBdCpWgSCq+dCXNShCQglYFNjodvK7ZzdQUS80i24sTyshLB4GtNQV46/8sJbmvNchG18tm7CE1v8lPckjQ5fmYxQpHUCSgVkvacuugRgs+TPmPSB+fnRH/zt7BSIgkCRKd3FduOsV8fOJwVpuriIkYAkGdrCWioxTqQiJumFbRiqUs4kYCUxs0y+a0iNIoNAmR9GIp0sIqY8Uw0y0alXBBWygvkb6z6MkRX5QZa9yIs0FsxNAyXYjRm9/SFnkqn1nBT8hXBWww1qOBClXBemMdolllS/laTS4qniD/vXqi/61Fn+6JvN5OH760VqQ3L1BoTG2yb2r/7DgMDKpiAykPbOTLeD8yjQ9MoSixGjmiYAXFZoyLGUOp3JsP3CaqjWoULiYBBjtD6KU2n+LVu3oIb3yxDNkzzoDRQZIiiRIKmxSLjYXMC35Bo2xAo7FFilqjhjK3HIlkcWARvRMdu7Sn8NK345vYk5T2God+HU6pCTPTNtPHBuakIvw0leK6os30U9sC5abm3uI4C2xJsxbsHXAJpQceFl7tnSZX0CJE1j3lgBfLlZ5voxL0MTU/jdiMslxuawryfsokIqBEKnoZygvEF7Dcliit09lkbbgIRPra/FuLF+gcSbeniGnBXc1BHiyJc2VBkgWOZLkjqXQ26fSsvbT79Y6g9zYoJunj4/Pt4YufHYywDrPR2JAzgNkRDkmR5MT4yawxVtMqW+nh9mC5tZxW0UpYlzE6tScgeCv8Bo2yEQsLoQV2ujFqUiY8F1Y7J5RAUO6WkxBxmmRT1q3W3sVmYnqCBs+Nlsu9FtABHGFjpytK67SgMTEIEmSZuYyojmazuJbJpTSJJoqKipCmmZ1f2BqnIkC0ooJQiWdROSh5MAmZ8GKW0tWid7f3oGKTIOPapgMYVfovEtqkMlmOo00CMsXQggUcXLiSge4AylOe6++w5BE0iUaaZDOlqpSw3nxhu6eeWsgz/5lPbaIRZTgU9o1x8p0BSguirDPX8WnwEy/GqZ1FTaNpEc0U6qLs8WvPSEsxylJ8bhvYeC9nV0GpofhR5Ks3zXy6xeBj27MmSTyriYvm6bjFbwqS5H+N1gxCwMguqh9fEE3x/USEjPExE7ycRBDAq6qcsewcHnIoFFBmeJWbA3gLKCH4fiiFloI3YiYf/H/2zjNMjqtM2/c5p6o6Tp5RzlmyLMs552yDTTJhveS45P1YwLDsLgsLJmcDBhN2ycnGBmwTnHCSs+UgycpZk1PnCuf9flRPa0bBlrFkWdC3Ll0j9XRXnaoO9fSbHt9BVYuoR161ilgggtAvekx6UCnos4onAs1Sz7LAsXSH8Ss5rL6xbPWvFsV7BlK8LB1w+T6alJYE7q3Er+ITEhGJF1Bqsk6dOjF18XOIcbi/mI3O+locYESkjFTgrHVXMy+cz4xoJiMFEuMrE3bbzsxwFvcn7kOg1h02cmGWqvu6xRIQkLAJDIaMZFlcXkKX08k6Z23s5YWKU2ti0MT1QI51dmm3j20vglGdamVVqe5RIQIBPhVVplEaa49xcIhUiDWWhqlTGd6yBYkirKtoWOFzzPv/rdb11STNvKj4YtY6a8nrPDPDWbTtwZl+drKfJ5Xm2KYVWP0ghdCjNdFH0pRprhzOMf5xY+7fKE00Rk3P+Lxs2TLMD3/9ECUvDxmhbWFIy9IKT9zncMq5SYq6WDt+I/FxjXTZRSqKC8qrqbHRuAr+raHMl/IJ1oSaiigme5YvN5Wq3lPPHl/gv3NONXqnq5VPsU3EoBU2hZrFB8i6QQMzjCVfNSiNqh5dOYkjQSLQ5ljekPFp1vCwr3kyNLXoESrexh2+wzdbyoQi3FbZGb3a2RofF07/U9rnZ6XEHtcy8uH3/xoq5GySJ0NTqwSLBzwofISiKH5ddBlnhHOTT1/kv6yi+Xo+UTPUbS7Ez98R/4BWGHXqvJCpi59DjAl2EueUzuP36RvGFB4LgiOGlKRZ5a7kpMrJT7udo/1j2G62sc1swyiDI85O4VMN/hsMkYoomAKRRBzmL+ao4GgI4Ofpn7FDb4NqB1nNP0wgQZKUTTGshilXbTlGWt53svN/oQ4o23Kt2Ho0aZsmwpLoaKWjpZniUD8dw02c+9FX4HhjL2ouHgvDRU973KclS6xTAa7XhVYBzdVNKGBlYiXH+yfRSOPTbmNXHqhoPvGbTayJHDJhmmNe00fbEQFRCEr7dJod1XMa+63FI/5iD7VIRRhxSEuGI/wj9rj9oxOWH3gllvsaR8Hhrq21kz9benQP/8cjnDG5SEVcthSm8cDAiUTVAnFQTDYH7kKd1JBQ0GSEcSaeA1QRRRZhirHMciK+1lymMQ7+8ag/Egccy4DVhMCLkyFfNcLWKC6GHvHoSinhuETES9IRd/qWJ3xDtyhCUSSUsMiJWORatoeKr+cSrA4VKQSHONU4UlwtKFaEmrSCa4vOXsWPCKwONP89mGQIRVEULoIS4ev5BN9pKY0pAq9Tp87BpS5+DkFmRrM4o3Qmt6ZuqaWXXHFpjzqq0Zt9a0FfECxkUA8SSVx4PLItLYZABSQlUevM6ojGUVEj5akwMZzIDm9bLf010l1miB8bqgBXXALxx9QPjY5U1brRRNMedVDRFQZ0PyVdwhGXVttKhgbOKZ7LBnc9RVVgWuZUJqYm7dZRta80kqI10U9JBbuVJwUq4NbEN7h88E9ASKiPp+J+CvTuhdgj/Lbo8H8Fj27PUI4UicmGKReEJFIgFiI/3i4Cvk2gteBLEM+YUfFU7QplfCroPV7mYzwFxz7HlmufCnck7mSrrxDROCpkdnYNEYaHB04gAuY7ES17X8ZzZpaxTHaEniiep7TYtfRECk8Jn2kuc6RrxxRcL3Ijri+5u4m9Rh0LlWYDr0rHNhYbI0NQTadN0JZPNpZxFJzoBPy26NTqpjwUg0qxwle8YSBNv9VYiSdJOwhJ4jTgyIejAgKBe/3YpT5E0RUpWrTgqbgu6L+GkjxUMTwRxSI3qYQEikELJYnYHClm1OuG6tR5wVAXP4coh4dHUKqUWe08hanWzMQt1SHTwn2bJjw3nIePz93JOwlViCMOFsESYUQzMl/Yqaa0iiq/88FKyJSzVFJlomrxs8GQlCQOccdYSMi0YDqdzo5aK30ocaFzTQSJIikpOuw4HnEeIm/ifVRUhZIqckz5WBpoYEmw56jIs2WVu4IG20DJFHeTT54EaNmMli2Axot+g7GPUkzdssdtWYEbSi5KR0TnT6Dh8Ud5xRc6ybZWZ1BbcKqRJVGgtU8xyJJ0SnF5uUAoSVJKsc3Zyu2JWzm7cu4zHsOtZcP1JZecKMZp4S2ZCnPcZ76wrnZXsz6yhOLWZJbFMC21iYcGTqBNWW5s3/vk6v2BUvDxhjJX5hJsrZqfHutFfKSxssdhjUe7lpLAqooBBWklTNWW1zYENZH0/xp8xhnhQd9SFsUCJ+I9WR9Pwx1lw//kU5TZ2ckFwqrI8Kb+FP2iMSoWotrG9UZ+dYbWSE2SBYZFUYw0F3anGUoexqRehwA40o0IRLih7FKqRos0UBaFowQD7LB6Z9quTp06Lwjq4ucQ5ij/aHJ6mAE9QESERjE1nFpzU98XDgsXM7UwjQcS9zGkh+jVPbV6nZGoTqPENS8eO9NMDg7pIM04bxw+Pr2mF41CY2izbQiCj88JwYmID3cl/4pVlk69Y0ynmkZT0kUeSNxHRVdwxKFq4YRGs9nd/PSOlM+SgiqQlSy90rNbUbYjFZqjYXb2/LgY2YAObsO6Z+62rZzAJhuxVYqUM62cdyV4qZ3Om9qA8uILqEjsTZUwJYyK0EoQUYjEs2VcBT2mh0E1QLO07HHtPZHilrLhF0WPCBiIFCuBu8opjvHiVNhhbsTlmT3P/ymrMnlrcFRc2GuqURJXRXQoy80dBZIHMOozwjhH+HJLmcGqkWnT09QufSPvkdLQaoS8jac+J7Vw8aj0k1Hw+kzA63cpSA4FvpbzqtEaqonHanG1CJvF1Aq7FZBUQklU1ZBViFC4QEni/wOsiwyhMnSG4AEbQk0eMNUvChCLJSGurUrpuG5rwgtkCnedOnVi6uLnEMbB4azyOfTqHgb1IOOj8TWh8mxolEbOLp9LRITFclfiTp70HickJC2ZWj3Q9GAGq51VdJmueA6Q3tm1pKvt5Q0Su54rFC4uCZtkg7MeJXHb/a6MRJxGaoYiFeGKW+v4KujC33p69nKsTfTRS5ttp8d0127XQFswwBGlp3Z5hODYO/A5c5dbhUcT95Bz5pMUTZszTFNbiHgeonyMtrXIkgCh9QhxMITVaEIsvIwOQLxawfmQHqQ5isWPWAsqrh/5n+Eka0LNE75muOpPZWvN+rC1JBybiFgXaVaEhs81lXeb1zMrnI02m0ljyAtYBVqEYtjCRcmQ2fsQPXo6KhJPbG7RQp9V/KroMiyKRizrI0MJmG4sb8v6NGlofoaC7bLAg35c7DxtlHFrRRSPBfppi4itwAcGkzzoO5Rr9Ts7TVFH0lpWqNXiJFRcjZVGaNCCsTCMIifxoMXhUdEggAqxkAJGRZViRu6TUcKJ3r6loevUqfP8URc/hzgjww877LjnvC2DYaOzgT7Ty/hoAoN6kKIqxBOkw8mscZ+ionwcHPIqT0HlcXFJSoKMzRBhSVfFT0SEtppfZX5BUB2aGKoQJbuHJCIV1a4WCkWootpkaUf270t0sb+Y7dU5Qg22gYKKxVXKepxeeIDJ4TBjL2OKUJ++23Z6GrpZbbYzPZViRWE2Uo0YROLgAylGTQUWRSnowHP7yYdNaBXR6A7UmqqVikjZLC4OHXYcxd5eHvjylxnasAHtONx4+b/SdeSJOCZOp5Sromdk7KEARRTbIoVG8VSg6Y5SvDIVcFEqrImgVttKQzibAW8djYp4rpOkCXIn8vHnMNtHBL6V97jHN5QFQqtYG2oCYnFQEpighAWepTPSrBk0fKOlxK0lwzVFj6yCf2+sMHeXuUUFq6igdquEssCOSLFE4IlAszXSHONFtCL0AM0C7xpKcWvFoTiqOmxnF1csSha6EcsDU41v7hQs052I7ZGmV+JBjUE1EiTUGih3PwfsHHE18vNwJyJr4J8yLxwLkjp16sTUxU+dMaxxVmMwcWRHhdXhhpYhPUBJVUjZZgLdT6RCfFVhQPUzN5jHiwsvYaOzni3OFgRLa9TBA4llVaPVsFbjs7dJ0A4uAfFFYiSBYLEc5i9+2vWWKfOI9zCbnU0YDFPCqRwRLN1tJo+Pj698MpLhvNJ5POk9SYeKC8QzNkO77WBquAnFtVCzvAyI1Ow9prx6G3sxaOZmN5Fxc6wtTCWQBIEt4ZhojHxyxSFpiiCKSpSkyetHMGgV3y+SiACfjmg8YoU7P/Yxyn19KK0JrGVVZHDWr6Nl7txaBEOxM70ywpZQ41WH3awNNdcUPLZZzduyOy++71JL+dzgAjqdTRSDBrQ/k3dlQlr009t2PB0/Krr8pezgKFgRaDqrzuzuqLVuF5hkFY1G6IoUL+5Jszo0tdk8d/U4fKyxxBuyO6MkrVpo1cLQLoI5ASxyLe8fTLIpijvn1vYreqpddCMRnkYNjlSHKLJT0jap2A/sXQ0BnVHAN3IJcqIwSnCBjaGppSJzEh9LyO6ifU/ExyukgaUJy8tT/nMuVK9Tp87+py5+6oyhouIIQKFqnqpRrMrNoQRk3eE4auEosk4ZFFSosNJbQYtt4Tj/BOaFCxCEG5O/o6iLezVSHU1svipV+4n4e3jCJlkULOK44Pi9Pi6ncvwpdTNdpmtkK/SaXjrNDs4rn09aMkRE3JdYRpfpJCLCsSkqpRk4gTDdzTA/WEiDxBYWfvKLWH8+ieinxMLnOMrelXvct7Gapmo9x5REPxOT3YjycZXFRgmM9tGYOJUoEanB7djP3cekCWk4eSr62GnEwS0FSiirMk94j7GxsJLCG7I03+yTWlmo/t7g53JIGNGoNQWrxpxVxc5BhUni6ESjEjwNf60YXpuJpysDJBX8ZzbJgF3AsANTshXMcyzGXeYbtIIHfENOdq4tEEFbixIBpdiIZYlR+AKPBYbsLmmvr+QTvC4dokfqcBS8IVPhK7lkrZA4ELgkFfLzosfWSJNQsM2H7dX40MisnxAYsnFXWNGq2nnKKuFFqYDL0j6nVEXJBcmIHxRcri24lFQcXTNAXnb3EtsXDHCuF/DmTIW5zzGVWKdOnQNDXfzUGUNa0uRUjlAFaBSbS5O4e+BYZmTW0egOIqIYDJpwdZmEjgtMLZaHEg9yrH88CsUqZyU9pneXuT5PT1ywrZkbzuPsUtzxtNx7hD8n/4iDywJ/IRPtxDGPWe49wpAahNq06Xg7Q2aIx93HOd4/gUe8h9hutmJw2BE5bLMlfHcFj/SeRZtT5rL2v3CqfxyTbGwrEXpvI+Rtz7jeSf1TWD9uLXOciPWhRsRBkaQ8cBLHNT9JGDUypAcJCAge20IkFq44LrY4F8EW8tCYICEJIiJCHaHFUAiGiJod+l81gQlf2IjJR0zs2cLmCTMQG9GuHXqs4FfjYyMpoTh5EzurN2ipFfIWrKIvUkzZpc26RQstz2GK82gCUWyNNCUZNYBAACvxVO7qZG5/YJCoLcNmG4uZYTuy9rjepsdqPp9zeXM2pL1aIHxSwjLPKXFdyaUocEEyZL5reXNfipEpzmutGRMNG8Gv3pbRgkhsjfHp5jLnJCMeqmj+30CSbZFiVWhoV8IAmkIEJeIZPc+mUme0fpypLEWl+fehFF9oLjGt3uJep84Ljv308Vfn74XD/SVYBJ+AsirzSG4eYNlUmINvPcCSdYbH1EgIQoW4NR1gq7OFBmmo1e3sC1o0CUlyeulMHBxuSf2ZLc4W8jrPgO7nnuSdbDabxjwmr/NEKhoz80ejCAjI67hlfqvZSlEX6ZES2yOIsDgqYmp2DZ1BE38ZXMRj3vJnJdQAGioNHOUfTSsJFrshSxzF+XYuVyZnksHg4hKqEDtYRCoBMrsJstUhjkpV27+oFZlrUfiqQmEclI9spnREAz1vjn3ELv7LzxlfGiJwPAZs/I3FqZ7/eLCfcLQb0qqFyY5lnrNzVk6DjocJPlcGInhnf5KTujK81lvAd3PuiKZhjrF0hXEBcK3fqip6lB0ZaWBp6e+kb9sOPB13m4UoIhQlFAOiKKD4Zj7J0V0ZXtOb4hE//nhqN8Jbsz7va/CZ71q2hIrlvmJ1Zy9P9hfGNAOOTm+p2nmKnb/elPE5JxmxPlBcmUuyzcaeXgWr2GTjfbkqTltV2LXya+x2R/9upIvMAVyEyV5s0yHErvJ16tR54VGP/NQZw3g7gbRN4Sgn7vKKXJQCK4oH+k7hiOb7aPH6CSWu0hmNqRYnxxcbTVvUTpfp3KMP2a6EhBgVkFfDbDNbKagCTm3mcGyi8YT3BPjQYBtpkRYSkogd42slq1TnDWnSkuYJ53G2OFuwKqI3bCVQlersIoVREUYJW/1WSvrxOCX2LN8Os8LZzAxnUVbxdOqRDrVm20pebwEUUg4QoyFlwB8Vl6gKh5GIl1U2drlXCoxGsORPa6X0k61kB3y+MT1BobnE+wZSDAs1U1BN/Cae4wpnJn3uqDjoqrYKBF6S3nPb+7OhLPCS3gxbIo1WEJoEV+YM6yLNZ5orZLVQqtbb1FDEIs9GNBRzjBvoYu6GlXS4ipsmzaoe++7kUBhR3FpRFAYV72rwOT+1MwYTCnz6/rWcfPVXUL1diOOwY/Zh3PqujxMlU7UaIgGSCHOcCBd4ecrn8mo90c+KXu0lWUHVhIpBMEqRFMgztnjZwKjZ3EIDqir2RsrWYyZoqXl5aQUDtj7gp06dFyJ18VNnDH26l6Iu0iZtSNTKdK/AQNCOAP3lGQz6W0i7BRqcYdImtq5QKBLVPwDj7DjW6DVUVBml1L5FVar1Q/cm72V8NH43ITKshyiqTvIqV53V4sdza1SOEWNVjcESkbRJZgYzuT59HREhkUDZagLRGBXhADuKM6u7FdxRwuXZolC7FVcf6R/JgOnHFYeoNQ29w7HwCUeJH61inyrRpCVDTg3XtodWKKUhshQ+fwIT04vZ6JVYF93KKjmVYWneedKIL+B3Vwx3jyuz2Iv4a8XFIFycCjh2P3hK/bLgsCWKhwFaIKjOov59yeXDDRXu8w0dyrJddhYcYy0qDOjo2cpbrvsWczeuQKKIH7//c5RlxEtsd0YeH6HYbhXXllzOS+7sWLt7KKDlK5/E8SuUPJdAYNLq5Zz6nU9zx3s/WXulZRG+01JAK8XxiWhMW31+VHu7R2xEquKng8VuxNZQk7OQUEJBFENSHfaJcJiOyAcVpiQSrI80ndWCahdoVsLCUec7FJji1Iud69R5IVIXP3XGMKyGGW1BcW7jZnoqE9nsNyBWkQsa0KJpNKXa12ItmsaokYgIg+Fw/wiG1DBdXmdtu6M9yGTE0nsXFIoe082kcDIhYU0AVVSFvMrjioeLS7fupqgLaDSOOAQqIJAAB8GTBKEK+UPqd1SUT19pIr02RUKXsSiCKMH20hSGogTWWhalO5kRzhyTOnuueCQ4r3QBm8xGHko8yMCKLQzluukMx7H9YUP7PMuMc1yUaOYG81HAKndlbfJ1bFFrsEkojVP02UGekk30SJpi5GFRtThLVG0FL1iFD5ybjDg3+bd3bu2Jx4K4pqYi8cA/IZ7MXRK4ueQwaBWiICtCpdqGL1rTNtjL/C2rmbdxBTaKiLKN2JnzarU4o+cgjUaIxVGlKjwCYpECsP2O23ByQ5BKkUJIAeI6HLbuMU73BrnRZjk2EfJfDT7eXvTsfMeyKjR4CiYby7A1WIG0iSM8s92I16V8PjiUIpBY9PgoGhDGu8JFpR6WJSfTXa1xyipo1kIhgp4QOpx4wGGzEl5Xb3OvU+cFSV381BnDeDsBM6oUzNMRrxt3DxuKk3HypzJBzUG7TxLYdnrpRZvYCb5f9/O71G+5pPRSDIbTK2fQr/vY4K6vGaWO4OAQSTTW5qL6JyIiIxmabDNDehAHh4LKo1A02AYCAkqqyGgvMSQWSBERZVXCF59QhYgIN/edQSlM0eL1YwgYDpsp2xQZU2ZqegPHNC9nRumy/X4eNZqZ0SxmFmfRdf6ZvP1DP2bzqhCV0qy8UfHEdfDyL6U4Wh/DBDuBAT1An+mrpliqowYIabRNbIxCem0Wi3BM2738tfecqgFDrD5dhISWWkfX/uakRMRvilDeRSIq4DO5BH1WU6yuxgMyyqI1zBzfxqztDg3TppEZP54Fr30dPzcuXgVK1W3sLSaoERq10FhtPx9hflRklRqVYoub5UiK5d0NPu9PPrM9x6szAQ8Hhs2hJqVhlhMxLIqj3IiJRnhF0ud9QymSCI6y9IqhEcHRcY3Sn00rWyNDRlGLSBWsIlUtIl/oREwzwmVpv1Z4XqdOnRcWdfFTZwwZyTA9nMF6Z/3OVJASXu6OZ2ZDPBe3p3Iav0v9FpG4MBoVu7av1+v5FT/nlaXXoFCcVT6b/3U2x8MNq1GfeK5OlqIuEEhQ+/o/ImYykmFaNJ254Twedx9jSA8SSohVFg+PYrWoeuQybJFae77FotFEKkSjKIkmjAxKKQaDNuZmV7KkeTmFqJHJiR1M8AqUEK5PX8erC//0N6e+nom/3thF14YEDU2pmm1IZYvm0e94vPId8dXxwtLF/CbzSyqqgiCEhCRsggwZnrKgqjOEJ6V2VBvodTXGokgoOC0R7TbReX9xSSrk08PCcKRqE5IVsdDptlVvLIlreCoIoQhNZpgw081bL56LOv+LtGihyQjHDUc85Bs8qU6+3iXiFkd9hGYltGvhZalgrNHpWWdw309/zo4oTptFgGcti2ZNxUnuNKAVgVsqhrsqDkkFL08FtSGKSQVfai5zS9nwWGBYkLKcnwrxFKwJFB8dSrGmOoMoX03RGR2nsYaswleKQavI6OrQyeqxI4ppNuK/GisH7LmoU6fO/qEufursxlH+MUwKp7DBXY8Rw8Jg4RjbjA7bQZttY1gPo5QaEw/odDrZZDYyI5pJkzRzRvEsbk/fWhMmWZslI1ncyKXX6RmzX4WiLeqg23Sxwl1BRZVxJJ563Gfi+7oSxwEsNo4gqbENySNzhawYIE7HxF1JiqyTQzBknWHSJqiWUStyKsdKdwWLg8P3/8kEHn9ggHTSi4u6RzrgXOhdpWi3HQC0SitvzL+FR9yH6DW9pCVNj+lBoUEUqjoQUcSQ0hUq1sWoCE80J3sR/9H4t09ofiYs8NWmIm8ZSJNDgbVMMNAvmrwoMkrwEPISEQIJVWFp0+N0eD28bXgxGd/BiINF0agtvlQLjYmFjq5Gr9IqFhgVhIna8vZMmQtSY2tmks3NnPem13Hj939MZ74AGIJxE9j4zo9SHScEwJXDCa4rORSqAxKvKzl8vKHChen49eIoOD8VcX5qbIrw24UEkdRKskDiouZQpFZgPoyDAooWyuysXRJgeejw2WGPK5rq6a46dV7I1MVPnd1QKCbaiUysTNzrfTxJYLXdrVZGoXnMW86MUlxQvCQ6gsZiI096j1NSZQoqT0WVGdJDcTSo2i+t0TTYRrp1F1uSm0hJmqxkESX0mh5aomaKukSf7o33Iyq2xRhhJBxRW0iEBiYlt/PwwHFQjf40e/04SmgwQfVhgouhW3cBB0b8pFIOTWELQ84AYXV6jMEw3Zk85vw5OBwbHA9BPJH65tRNhCogqTSRGFABmwszafP6yUcZxuHwhhS8NhOQPgDplaKFL+YSrAw0FkVWC3O1xYYFjJuh149rZOLGLounfYwoWt1BsqbAw8NLMVhyapieoINQwCceTGgQksCAKBIIx3sRPrGXV4DmqUjz3qE0f64EfLllrLArnvMibl90Aa2PP0BPuol1C44i1Jrl/SFXtZTpixTfL7gMiEYRT4Qui/DfwwnOT4W1YuddEYHtocLVkEHotXFHlxC7ujchbIg0RRUL8BGbEar3cRA84Iayy1uzAW11M9M6dV6w1DPSdf4m5ocLdivYiAt1FSmbHnP7jGgmF5VezOH+EtKSxlfVmhykVvkqShgyg/SZXgq6QK/pZZvZFkdLMAQq5Ah/KRnJMiWaSpttx7EOCHv0C0MUoTgsaXqCZlyUTfDkwFEYFJO9wWpLdJx7ydgsRh247wGvfOUCtNW0Rx2Mq/qwNZfbeclFC4G43f0R92H+kPodtyT/zHa9FQ+Ps8pn0xGNY5b2sGEbnQOnEpXm0UQDlyUS3DsO3t5wYIQPwGdzCR4NDJFSWGtp7e2kZ8WTtN92Iw2Bz1RjaVZ2zMvAqIiUKVOySUacTLqjVNzBpeIoilbxt66JxtKi43olq2B5ENtKaMBVcYr0+pLHQ+WxB3hDycVkszx1/Fk8sehYytoQorjfd7hiKMW7B5JVX644HVUEcqLosYoN4dPno0bqphRCaVRZdgQMoggEGgmZ69haW70GPIQsccQob2FrVM971anzQqYe+anzNzErnE1rvpWBxsGa6HHEQaM5qXLSbvdXKLY72wlVREEXajVAI9T+rXb+8FWFTrODrG0gb/Lcpe+kost41sWgMcrB4OBX62RGo5UQWYfQOsxMdkM0jjRpXhEuZUXytjgJJ5qsbcDFZb4//0CdKmbPbuF97zuGn/1sJf39JRobPS66aA5nnDENi+VX6Z/T17+F6LonURY2vuQoTmm6kPnhAk6tnAbABZHi1+Iy6CpOTwQc7+3fjq5dGbKwMjA4Csp9/fQ8+QQ2DAmTKWbc9CtOevBmZl35BT5SbuFh3xCiUUpocgZZ0riSivXiri+JnwO3OjQnJJ59o4CNkaZdC0PVupryTi1cm02kgK8WEvxfslRbmyXeVk+kdyvA3hgqHgnGiiUhHr44JJpPDyW5sqVEh2a3uhyl4Ggv5Ht5j23W1LY9MsAwBLJYpkmFjOPQZy3brCZJXGyuVDxhe6IWptajPnXqvKCpi586fzPHrz2JlYevoMd0EWFJSoJTy6eTIbvH+ysgr/JPu02NHuMHFqiQftOPrkZ3rLIM6IHagEMAT+J6mrBa/zPSOdagK3hukcnJPk6xzVyUGaaNOYwva1a7T1FRFRKSYH4wnwl27ym+/cGpp07l1FOnEoYWY1Q8zBB4yH2A7j/dC1fdj0QWUYr871ex7G29zDvzf2ppsTYjvL3h+asjqYz4WhVL9K54kjAIsSh8UXT7ltWPr2DCr3/GL17/Bh6qaO6sOATOILT8lnRikA2FiTw8vIh80ELFJglVXBg84qwuxN5ZSQtznKgmdjSQVlITHhZo2MUM99xkwP2+GWM/EQk0GqEiimBUMnH0IwXhLt/hjK4sk41loWv514YKM6v+W8t9zV8rDqAIRz12xKDVEhc4jzDbsQz4O/cVCTQqy4WpkNa6+KlT5wVNXfzU+ZtxxOFlpZdToYKvfLKSfdp5OTOCmTzlPBV3ZBHtFv0Zeezo22WkgFmNjRKVVbkmgBpsIzk9XKunGekc83Bo1rCg7SFcHuPOagH11GgaF5QuwmIxmKdd8/7GccZGJdbLWuQHD6O0QmlTXT8M/9/tlE/Kk0o0HND1hFKt29nlFHRooUNb1m3fRhhF2Oo5ShTzTFz1KDmteWrFapYoODZpOTbpAw3kgpfy+eJa8rqXhZk1rC/MYRDLcJQlQuEy1mW9IPCD1hKNWvin3jSPB2aM0aoDfKixXPt/X6R4qOJggFDi/kGjoEkLU4xQFsgQp63KuzyvLvGsogKaUqRYHxlWhIbL0z5vyAR8JZdg2GpmOJZNvh4zhHHk3wMoNqoEiUBTtIrxRljihHRbw2RjuTQZcFH62biC1alT52BQFz91njMJYoPOZ2J6NIPp4XTWueuwKr6cjIgcxzpYZRElNcuKkZocUXv+Fh2ogGbbTEWVCVQwpp1+56hhoaDypMnUbCTWO+tI2iSLwsP20xn423HX5ZG+EiqziwdUf4nC+i2kFi46IPt9sKL5YcGj12qyWjg/GXDZqIu2UvDuBp9/8ytUEikIApwg4Ijf/xSvXEISCdalmnfbboM0IKbEeHeIcV6BeZltDIdpHhtcxJPFBehqoXMSqhOjhQFRTNTCT9qLvLovzepQE0pcYP2BhjLTq59S6wLFx4aT+KJIaJisLZ2RZqFjSRso29jN3lXg71IH5iBkiG0rRmrjFbAjUnyn4PG/OYe11sFRscmqV6v5YYyJSwhsUwl0pJigLVONZRjNG7MVXlEXPXXqHDLUxU+d5w2F4rzyBdwht9FpdhApi19tem62LeTUMAVdiA1Bq63xgQr2uj0tmqIq1sxBlVQTXqKxxTLR/ZsozptEeVIjro2nQ0PcabXF2fy8i58hG6d6JmqpdRzNyR7GVs+pibaReFfGaSLZ1Lz/16AG+av7IPeaApmky3BxJjvyR/PTgkdWwYWjfLQWu5bPdj3C1+96jHB4mFkP/ZVkfhisJUqlqbzs8t22X1RFPLeTStXnzShLs5MHHaKQWgRJVFzgkwSm6VgIZzX8vqNITxg7vC9w4mGJI3yvkCAShaNiN/dxDqSxTHQsk40wz0T8vOTVBiiOpNcMQoeKpzRbiWcVxYXJimFRbLcABoVgZCTqqMhUo0e2+uwAtZ8R8XNpicXSzWWXl6fC+nyfOnUOEerip87zikZzZuVs+nU/vbqbturMoBEEoU/18tfU7XSbHkRkrwLIEbcWzRmJECkUUfcgat0g8q2HiV6cp3L4eFbOyaCdBA4wyVgS6vnzXHrE1/y/gRSdVmMQFjkR722scErCcnjbyayYdxi9q1dh3diVPh0kmDnnWLKTJu3XdQQE3JG8nTWRpYwm0AEt2RX0W8263NHcWHaYZCw/LnoMWEWLFv7pwks5+q776F79KF4QN377zS2s/uS3eN38aexqT1pWJZp0QFeUqCWdnsjNYzBoIqMiSuJggYIoGhHOSYVkd5kt2eFAxx6cv7qqJqGbQsWOSONLXBg9RSz/1VThu3mPx31DCVX7YLPEAmh0vVFKCeVqTdNIgNABbLV1vSSKtBIigTZlGRDNnl6BJVFsjDQLtSVv4+3VP1Dr1Dk0qL9X6xwUWm0rrbZ1t9sVinbp4JzSefw2fS2ibS2yMxpPPBIkaLWtDKth+k1fHDnxQ9g0jLgG/cqFqGW9lBdNQm3YQnluGlCsDg3N0v6s1huGsL1H0dQgNO25nnuPrA4Ub+1P02/j2pRAFI8EDp8bUhzWXqJFw8s/+k2Wfe2LdK98Alcc2g87jGPf975ntb59YZ2zNvZJsx6+VN3MI8XirhuZ9ImrKIyfyBWX/wsNEyehNAyHik+HDXzys5/ll7fdx/aHH2J4ziJ6jjqJs8dlOTuxewF2o21iuvYIRBisCoLOyjgyWjHTFdaHlmGJ05JHeSGfay7vvtC90KCE1ZFmfWhqlWK+wE1ll4cqFTYGMFSLnsW1TCOBoyVeRErBqkAzJJqg2mEWz83e6dw+UuasgWO8kCPdiC8UEuzJjC4EcpGiXysmqugAzQevU6fOgaAufuq8IGmUJs4snc39yfsY1IMUVYGAAFGCJx7jowm02jaG9RCRCnHFjWcH9RRAJO6maksz9GRIoc8jfVojKa8LgHKY4bbuC7m06RkWUeXW+xx+/HuXgWGF58AR8yNeeuq+5Td+UvQYrgofiGtpRGCL1dxQdHh9NsBNpzn1iv/4W07TsyKv81XfsCoitK5biRnOEW7dwGNHnEhhw0akUqZp1qzaWn9VTvCfFxzP5rNPYEOk8bas5cTGmYgwZqoyxIMaFwQLCbzHmCoaH1imFCnJkNKKwzyLlThtdUoiIvks0kQvSwVcV3Jr0ZqRBF0JxSW9WSYRjZHIIbGgcYGZjvDl5jL3+5pv5T0e8B3KoigIOOyMAnnE9heL3ZBrWkv893CSPbrwEkeSBlCsCTSRCx8cTPKJpvIBm7tUp06d/cch/zb90pe+xJlnnsnUqVOZPXs2r3rVq1ixYsWY+4gIV155JQsWLGDChAlcfPHFrFy58iCtuM6+MtVO49jK8UwLpzMuGk+HHc/iyhIuKbyEy4qv4oTKCWg0RgwGgycJXJ1ASzzDR5Uiym6SwsIZDAxnGfZb6a+MZyhopdx47z6tobNX8bWfe3QXNdoFFDzwpOEXf+7Yp8cPW7XHeTKhsFs30oFmZjgTS0RWx0MEnWIep5DD+fb9tK1bidUGWykxtGkTQaEAxLUxw9V00zRXOD0ZYRA+Npjk8v4Ub+hP8eVhD39UTfr8cAFnl85lejSFudEUjlWtpO3OrjWtYuFwQfLZFQifloxoUoJhZ7Jt5AwWUWzDkGKkkH4nDUp4Q8ZHKTg+YflhW5kvNpc5wovo0IJT7SkMAR9wRfhgQ4UdVrMq0E8b0VHEVhkdJp7+/M289zT3rlOnzguFQz7yc9ddd/HmN7+Zo446ChHh05/+NC95yUu47777aGlpAeCrX/0qV111FVdddRVz587lc5/7HC996Ut54IEHaGg4sK3EdZ4b06JpTIum7fF3WWngnNJ5POI9zCp3JRpFQ+ssercW8VWA89ftFE9dhJOyWFKUbDVfpSBlhtgWlpisU3vdtwh84MYkT/iGyMRvljYtTHctKzdk9mn9U4wlhZAf1VBvBVJaeFFy78XcB4JW28a0cDoFvZlB6+L4eZwfP4xz/QqUQNumNQxNnYUJQ4Y2bKR98WGEAjOdnfEUEfiKO5Ug1GgVi5A7fYdKTnHFKH+xFmnheP8EAOZmFR8ZEjqjWAga4sLq+e4z112FAn8sOTwUGJq0MNVYBqyqeYPV1kUsJidgMQgjK3GBd2YrLPUsBQvfL3isCTUesMiJKFjosXGx88h8KEcL47XwfwWXJ4Ox84R2RQFDohiKFE1GWBnUk1916hwKHPLi59prrx3z/6uvvppp06axbNkyLrzwQkSEb33rW7z//e/n0ksvBeBb3/oWc+fO5de//jVvfOMbD8ay6+wnspLl1MppLPGX8GjiEfI6T3LKYvxr7kM9BIVzPbq0Q5TYKVak6ij+6eEE03WCN6V92p3d2+n/WHZYWdKgdr5ReqwibRVte7LU2ANvzvo8GhgerBh8FKFAWgvvbagweQ/7PJAoFMf5JzBbz+Um2cjq/3sU98fL0VHsCH/0b75L5/wjKIyfjA18KgLjtPDazM7ankcDTa9yaRl1+G7VmqJgIbOHWHKrEb7ZUuIBX7M90pzoRUzch2MvWXhLf4oNkWaCkVqYesRZfuyxxdGkHJqZxsYGrMCLkgHvbwywAlcMpdgexd1igUDBxi3tHVrYbjVOdTudVnNhT4YQyD9DdG5Evq0INcfqCOeQj6XXqfOPwSEvfnYln89jraW5uRmATZs20dXVxVlnnVW7TyqV4qSTTuK+++6ri5+/E5qkmdPLZ8b/yYC89zUUX9VFmErw+eRdKMJaqiRAGCo3MVBI0pUwPBEk+VpLieZRF67+SPG7ksPEk4Sh5aCqv9NAj684ZloJ9jLJejTNGr7VUuI3RZdHAkObFt6erTDuIAYI2mwbLw+T/OiRKyg3J7E98dRtr5Tjxf/9dja/5u0UjjqRC9KH8eJUOKYup9cq7B70QCBx91NmN1kSYxSckLCwhy6uXRGB7+VdvpDz6BeDp4TuCCY7lvFaOM0L+ZPvxK3rUBMtEfHzM4RijmPJKOED1WjUvRXDllDhKdgQKgasIhRFSahtZ/TK+tC1VvlnXC/x5OdNoea8TN3NvU6dQ4G/O/FzxRVXcPjhh3PccccB0NUVF7l2dIyt0ejo6GDHjh3P+/rqHHhCQp7yVtE3vZe0pDm3vJRbnMeJdIGhijBQauT22+cRlLbR0pJk8RHj+GnB450NPmWBK4cTrAgMKwJN1K5oPF3I36MIi6AcaJwi/PNFXeyL+IF4fs3rswGv32PD9MHh8fBBbLGEd/hU/Ac2YMsBSoHnwrH33swxx85j4R6G9h3rRSRlrIARgXZtadP7J5L1zZzHZ3MehaooCUURIWyPNO06QjR8rLHMZ4aTlImNUDUwQVuSKh6XcEYi5LUZvyZonwo1RsWRu74oTtlp4hqf0c7sY46r+vOZRFCtAFviSF+dOnVe+PxdiZ+PfvSjLFu2jJtvvhljnttX6zVr1uzX+/298kI7/kJ/Nw/e9A0qnT2YTBLnkiWkTl3AS9YfzlXXbefWybPJ0VKtGBH6+kqsW93DxEmaNZ3b+J4zgUdNPBG4EYcdKkHlJGH20hKVTRrVJLyuvYuUtS+4Y382bJvUi80mkeESevEk2NiHlANA0XHOOZijjtrr8Z1nWrgxr0DFs3GSYrko2MHavuJzXlcBzdWJhRS1QkYGdQuUATeK2B5UmGtLnB9s4klnMrc4rWigUUI6ogAFjLM+F+Q20wP0VLfboVLk3Sl06QRWxR1vpao7nH2a1NZehc+oGw2CQmjyC2xYe+i+JvaVQ/l1/1w5VI997ty5B3sJLzj+bsTPRz7yEa699lp+97vfMWPGjNrt48ePB6Cnp4epU6fWbu/p6WHcuHF73d6+vFjWrFnzD/2iOhjHX6FCXudpsFk8dlpqCEJPeTt3Xf8d/K2dOEULwwHhVXeA45E/axorf9VP+ZKZqNlj3cAHB4o0dT5J1+N3UDjxNBac3YBjCkTW44nhJawvTWCwKcWc43fgIHTLZJ7sDXjJnMnP67HvT/JejsIbTyH/lb9AOgmHTUEqAa2nLOWC//eJp3/wmjVcNq2Dm8oODVq4IBnSoPd+LsoWflx06YvgdZmAiU/zqfOYrxnu9eJxBVTFh4rLkSMMqWSSyxsUc1Nz+ZyFdw8oiqLQygPxGLSK6YmQfGo+S11b67SbI3D/cIKtRRenOgtILKQV5GTvkZ29Cp9RLyBdLZc+pjHB3Ja/78+Df+TPvH/kY/975O9C/Hz4wx/muuuu43e/+x3z5s0b87vp06czfvx4brvtNo466igAyuUy9957L5/4xDN8yNd5wSAID3j3s83Zik+FUEU022ZOKZ9KWjL8NXk7W4ZWMPTS8UjQhr5vB/pPG1Eu5H/zAH3nLKGgCji/f4Dw3RcgjkGJkFIF2nu2svjWH7K91TJlUpkw14qekkUR0tH4CEeXpuAqEO2jEfJRmh9VjqE77/G2QzTNcZh/GFvP3oIzaxylXzyAlCokz1zE+ce/a58eP9ER3pR95jTevWXNWwfS5GwcKfpeIcnrMxX+vWnP5y2pBKsCNIZIYpe3kSnNGRXxsYYKZ6Xi6q0GDZ9pKvPlnMcy32FHpMkgPKwMjwSG2cbyqeYySRWPF/h4YwVPhF+WPBIKAmLhBGMjPJqd2mbEzX1vKOKOsoQSXpY+NF8Lder8I3LIi59/+7d/4xe/+AU//vGPaW5urtX4ZDIZstksSin+5V/+hS996UvMnTuXOXPm8IUvfIFMJsMrXvGKg7z6OvvKGucpNjubEIQBPUikIvp1P93pblwcEpLEFivoQAiVJTxlIsknN5Pc0kcxF7LJ2cAF39f0bhxm1V9+zfrMCdCUZc6G2zhv+Z9IT3DpvXgyxjXoBp/IBmAER/m0ZNYSRWmK/kQsoHXA7HH38Ze+6VyWjm0g9gURuL7kcHvFIUAxz4l4a8Y/KEPxPBKcV7qAx6ctZ/hD80iRZrF/OA2yf0c/vG8wRUkUXvUYBfhhIcEbMj6TR3365G3cMZbB0uLk6Qsz+DjYqgDydMBnxz/OWXrsl5shUWyINN2RjlvOUfQHimPdiPWR5kcFj7dWBapR8LEmn1YDt1YcdkRx31gCiEYVPTcSm6MGEs8P8qGaHtszBmGpG3Jy4vmzTKlTp85z45AXP9dccw1ArY19hA9/+MN85CMfAeB973sfpVKJD37wgwwODnL00Udz7bXX1mf8HEJscbZiMPTpXqyKUMSFzUN6kIgIrQy6xSca8MGA55dIHN9Gw4Z1TGwpQvlJ5JiFNM2DGWcV2HLXrSz/T8XpnTew4KiJAITNDl6zQ8lz0cYHFVcGGRWhTQGjK0Q2tjrwTJGKt4UfqI3MSm0lIqTFtjEnnM2UaFq1mmQn/RY+OujysO8yyYnNMHdUHDaGmi81lw+KIWaCBMf4x9X+Lwh9upeteit9phdR0GZbOcw/vGYK+2zYFkK/VTUT1xF84Ht5j/9s9hGBa4uaW9Qa3OQ2StYwt9GhPLwUJ0pigJQJOKPxSeZ7LrsO3flewWN1YBgcNXqgJIrHA82RCcuTwdjnQSl4V4PPJB0xGCUoo+izGt8KAVAh7iqbZGLz2RUVTaeMeIUpdG6IpT//Fq1b1xMmU6y94DKGjjqZ8fup2LtOnTrPD4e8+BkcHHzG+yil+MhHPlITQ3UOXQIVolD4+IiytVSFJcJ2uNDj4ebyKA3iW5SjmfmWNh4zGTzxoSmBjWD2mXDGrVNofNN0Ij+ODFSCBD1tE8gkiygV22DqUQaonjNEyY/rxKwSyqpI2u1ki9lMpCK2s51Os4NGaeSM8llkJYsg/CTazHeHW+kXBY5mZdTIBJVgQnUq8KOB5kjv4EUNfHwe9R5hhfsEJVUiVCFGDBnJMKD76dbdnFM+bzdB90ykdDzNebeZPAItRvh10eH6ksujMkDGyTLLdWhNdLPA62VcYgfrC4tx0ZzdsIEZiSHmlC7ebR9bQkWh2vE14tMFkBNFRWKRuSdmuEJWQ7sWphAhAkWB5b6hK9IMWGGcEYyC8UqYaIS+csBxn3o3zkAvGIcUwsyr/4eeN72P7lPPZVukmGzqIqhOnUOB+kiuOocEU8MpRES1ub6yp1JUpWBRO9mpLm6jZnqwmeO+OoHS/MkEyiNr43k22kCyUeM0wcQLz6iJn2XFkykHCZQCVS1jtaLQYqpTX3Y2RedtkgnuMH5qHYEOsMriK58u08lGZwO/Sf+KZd49rDJP8b/5NILFKIvRISm3j04bEUp8sd4QHry3YUTELak/87i3nKIuUtEVQhVilZDXeSIihvQQm82mZ73tVg3TjSUa9VRZgbQRphvhZ0WP9VGI0RXKNsWjw4eTD7MMBe00JQYZn+ihy8/yw67TuGrzK7mxkGHQwo5IIdVtJhREEtfcjEYRD0k8fy9TtCepeOL0wxXD8ophU6RY4RvaldCshRDFtlAz1UTMcCytRjjurt/T0bcDZRxSGlIGHNeh/Q+/oiLxbKg6deocGhzykZ86/xjMDeczoAfJqzw5PQywZyGkFOkOmNu4icmXJ8AKni2DCJHeOf7AFRcXl8Wv+WeaGsax5q676XTnUd40m3TzfWQS3YDGihNHQRRopbAqJIwa8XNNnDj+HkpqrCCzyhJJhK8qbHW28pDtpxBNwdURanRXkS6St1myGo52Iw4W65y15FW+FlEbWV+oQjxxKagCzdJMt+lmRjTzWW//Z21FXtOXZmOksQLNWvh8U4k/lD0cBeh4CKEAiLChMIslzY+zpTCdu7rOpoAhBQwB7/Bhes4y2RHGact7GnxenQ64p2JAKVIIZYlLlxUwXgtTjbA2UPym5FEWODkRcqIb8bL+NH1Wk5f4mIdDTVoJc72dHWIiMMEIC5yI2ysOzrqnEC+BK5bUqHYvJz9EsxZmO/Wanzp1DhXq4qfOIYFCcbx/AocHS3jIfZC17hqGzRCC4IpLpCIsFiWK8UE/0/xtNNphUrbM5HAHHVE/d2RPZdi04IhDR9RBi20hoRLMefGLmf6iF/OT/hQBitWdC5jZfj2OkycSh+lG0a6FYyvH02KbyUiWG9Rv2aYDNBpb/TNCqEIKFHCUg3UrOCoWN0kllCRu4DYqwlo43guZ7h68VEmf6cXBqXU7jfwcEZQKRUTEuGjvYyGejnYDfx5XpCeMrSJmVm0tri3Hv29TLoXqfrWCsPqRtKEwlyIGULhaKFoIUWyKDGUsq61meWC4ob3AO7I+Pyh4OEqRwFIRxSzHUkLxloEU5QgCFafBflV0SSmhK9IEjMTy4nofdrEsUQqKonhng8+r0gH3nrCUoQduYUAUXTpFKAoRIWxp5xWpoO7mXqfOIUT97VrnkCItaU71T+OfC69jSeUIEjaJJnZx98QjQYK1yVOoqCRpW8agQHmkrOX0/D2MC10mRJNotx2cVD65tl1XwZFuRCBgJcHGvhcxVJiLDVqZbzs4q3Q2U6OpZKUBhWJm92wccdCy57fQSDea1gNM9gaIJLZWyCrBiNCuDB9rrPDBhoPbHt0RjSMkJGETWCy6P4Dhqt+8KFKSpMk2MS2a/tz248BMR1juaz40mOQx37DC11iboAE3bmkXwzivE5eQoj8ei8IjrhEKq5GWCjBkFRWl6Iw0b+5P8+4Gnz93FHhntkKHFo7yIsY7sRdYV6TZaA1DkaIiikGrWBNp8sSt7hDHiUIUfcCTgWZrGKfVrMBkE8ujNiNcdNZJdMydQ64SUhBFIbJUROi87E1k6xmvOnUOKeqRnzqHJC4uZ1bO5kj/aB53l1PSZQo6T8ImSJDEI8uAMxuFRaqTW1I24PR8H236bSRJ7rbN9zb4kINHAkNg05jCybwjW2H6HuqLUmGK4ysn8Kj3KGVVpqgKWLVL2kNB2ZS5pPVBbh48nC1+BxphUcLy+bSiSR+8dNcIM8NZrHPXUnliK+FHfo7kSqAVanoLHZ96A3M7DmNxsPhZFzvviVWB5pPDSTTQYYQ+q1gZGBZIGwmTZ3yqi7PSBbbkT2GyNNGDkNJx+mnEu2tk8CHVn31WcVPJ4SXpkLOTITeVHRLVpUZAXuKfIVAaJaJG2toNkO3cysI//gotlq3nv4Idk6aRF+FIN+Jdo7y6Im24/kNfZv3119H+1GP4mUZWXvpaxk2dwg8KlhMTIdn618k6dQ4J6uKnziFNszRzqn/6brcPyGcBhbCzzkcwOBT2KHwgjv58oNEnlPhimXyGb/NLgqW44rHZ2chWZytDamiP9+tzt/Op7HH0qS1MjqbQLh17vN/BwGA4o3AWP3j5B1C5HMYxGO3Cym1UHv0eS+9+03PeR6fezpPeCr6XX8hmmUBaMrRrzWGupT9SNGnFxxpcFrqTUNEkSIGfLHJWd4btoUZrUCJINRJk1Mja45qc1aEBQtq1kB71nBl2CqcK8SyfXWXspAfv5LTvfhpRGqOEpcv+zKqXvZHui17BJ5vKjBvlPn9r2bBJJ1h14atxXnx57fYdVkgpxTLfcE7y4AvaOnXqPDP17yl1/i4J9VRExl6IlApJqVc/42Md9czCB+J6mEXhYVxQvpgTKiftNQXma5+kpDkiPIJ2adun9T+fbL3pL8hAkYRK4kYuOgCtNfnOTvqfo5dRp+7knuTdrAot9xcmsz1MsSayLA80/VbR5ggznIhFowqNATwFN7UXOCcV0KEsC4ylUQkphEjiVNhUJ+4kO8yJn+dxRlhcTV3GCK2mjENYMx+FUc4U1tLYtQ1rXLQ2ONrgOA6H3fhzWvwiw7vUAD0ZGuIpT2MlVCBxyXs99VWnzqFDXfzU+bukwfwPkWlBSQUlFSCkZM6nUR1zQPZ3WLCYhCT2+Dsjmm7ddUD2uz8o9fcjdvdOJbGWyj7M0Xo6VnpPIqK4sX8pRtnq0EjwJWRDoMmHMN2xtdb10TQY+E5rmTvGF/jT+ALXtxaY70TMNpYlXkSTFqY4wrmpnZMPP9pY4ZxESMrdTkvrDZw24bdcOvmXHNN6J7BzLpQGvHKJpp5tGARHCalqTs0Z6qdj4xqmmrHnZKkbEQGpUdsBcBHatXCMV4/61KlzqFAXP3X+LkkyiRb9WwLnPyg5r8Nzf8p4/fEDtj+F4pX51+x2u8bg4jE5mnLA9v1cmXvppbjp9JjbxFq8hgbGH330c9q2j08uStPtZ4msxhdDITIMi0sfintCw2eHE7ylL8ltZbPHbXgqnuezNGn5dmuJs5Mhc5yIS5IBn28u4e0SMboskyNMrubhoSO5s+80Hug/mUgMRzXfhybO9RtAOw7Tl9+Hp+IPwlBgIFL0uinuy3TwrwNJcqP0z2mJiJmOZaL4NCipFkULh7sR/9VUjlv369Spc0hQr/mp83eLUoZWtftU4ANFO+0c5h/Ganc1VlmMxC3kU8MpNEnT87aOZ0uisZEj3/lOHv7GN4gqFQRwUilO/cQn0Pq5fT/KSJa1pQQ7giYUCi2WKO7hAsCgGBbFHb7DYF4z0ZRZ4O59Xs5MV/hIU+Vp93l1sJXl+YUYJWglRBg2F2cyPb0eDyEiFj94HgNTZzFu9SNUjMuwKHQYMjR7Ef7EKdxWgX8ZUPy4tRMnuhujGrmy6QSuGhiit9HDU3BJMuDo/ejpZas1SgfD7qROnX8k6uKnTp39yPnli5gSTuPxxHIEWOgvZGlw1MFe1jNyzHvfy6LXvIbl11yDSaU44k1vItHY+Jy2aQVmlo7ikXyRpA6pWJdgVAG6GvU3L4pQ4DdFl39/BnHzTDxYbkHvUpdjlKXHH8/JiYgNgaJPNCWB+9/7Ccb/9EvIE8vxRdEzdzGr3viv1cfAoxWfodIVTFTbQQke7bxcXs+M5v1buzUQwdv6U6wKDRpY6kV8s7lEZs/BsDp16jxH6uKnTp39iEKxODycxeHhB3spz5p0Rwcn7if/uz+UHK4ruQzYNGt9TYNTxGhL2eraJOl4AEGMEM8YLO2HfU+xLWxgeNTW4335UZJ+q+i1mjLxnJ+843H7Gz5EYKGnWgWQVVL9YBQCKuQlCTruEFSSZ2LD//L4ysXcv7ybJXOaOeboCajnEKoRgUt6M6wNNVF1zZ1lzSW9ilvGF//m7dapU2fv1Gt+6tSps1953Nd8r+CRE4USqIiiM8jQLAnmm1hY7Cp8EgAWjt0PVh+vSDk4tgFqPV7xAMXDpIWNgaKIwlbNUAXosoqsltp6Rlzo22QLn0x+lHnqJrR9HGW3IwJPDXfxzxtKfLp5Ev+80XD55x+lVAr3tJR94v6KZm2oazOIIJ5B9HhoWBPU81916hwI6uKnTp06+5VrSy4OULHwRGgIiaMsm6LYS6tVWdzq6MkIcBDmOBFLEhEXp56diKgIPFjRrPB1rWPs9ETEqxNJJkUdeFETmbCVOdLGPOVQQWGIC6hHevNCFBHQoHZOhdJS4dPpD/Ai90YMwyhKaLbTWdjCoJchdBNorbCNSe47YhZf+9lTf/P5Wh0YgppDXEw81FFxQ6kenK9T50BQf2fVqVNnvxIQF+xuCjWBxJEUrypMhqzi7ETAexsqXF9yiQSmuMLxXsRi1z6rQt87yobvFjwGrcIomKiFTzTGgwk/2Ojz6lDxx5LLDSWHfKT4nu9SrEZ8rMTf/BJK8EXRqIV3Z8o0KMv1ZY/z9Xe4wL2TpBoiNsAIgARJp5cV+XPxR401UFrzZyfDh//G83V6MkQN7367Qeqzg/6BqPhw9yOGckVxytEhjZmDvaK/b+rip06dOvuV49yIJ3zDkFU11/TYmkJhEO4PHD6fV1ye9jkjGe11oKQIPBJoVgaGI9yIw0aJo7yFqwseEYpMNX69I1R8aCjBd1rKeBqmOsKqyOAoxcrIUB5lbSHEH34iig5tuaqlzCIv7tp6RXqATPm7aBGEDIpS9Qh8iradr2z+l7ExcxFs2vubz9c0V5huIjZGhijaOZW6zQoXPctIWJ1Dk1XrNVdek2BgWKEU/N/vXN700oDzTqo//weKuvipU6fOfuXiVMjDgeFeP055jTSCK6CAIrJwZ8XhqUAzvyh8uLHM4l3a2/MWXt6bZn0YVweltXBeIuDK5goG+FbOZblvsCisFYZRhACBw5FdDm/O+Lwr63NfRbM5MrWIz07Xeqot78KLkwEL3XjQ4qZIkZDtzJYEAUJZXMDFYEmpgF6WUJYU7qi1Rlpz2cTn1pb1g1SRC1ZmKTQpEJAeyN2t+MtChze+NHhO267zwmVgSPGrPxq++csE5Yom4QnTJlpEFD/8rcvJR4ZkUgd7lX+f1MVPnTp19itGwaWpgB/mXfxRlSwjosOiKAvsiDRt2vK1XIKrW0pjUl6v3RjwZFRG2wilFFEyye/E44RixIBofld2GLCKAFVzZ1fE83uGRfHZXILv5VwG0fjsNDIdjQUSCE1auKOs+fehFEOimKim8qXUOKbqHE2qH1EaC4TWY1LD25nfrtlQdon8CKM1Z3oB7zhp/HM6Z/ff4qCvAb0AaAb9KNgexX/dmuT1lwY8x3FLdV5giMC3fuHxm7+4rN2kKZTimi/jQP+g4chFEdYqHlphOO3o+uTwA0Fd/NSpU2e/kyU2hx0pah6ND7XC4jWRpgJsixStgz2s/f3vGRzO89jpb0In4nuJCGGhiGSz3FhyKClFm4GV4e5GpRFQFoUFutBxZGgva1QIc1zLH8su3817KKXQCtZGrdwSnMhpbki7bqeZXiIxPBCdS9I5ja81r2Jg2nzWFOHUjGXG357xqvHQCkM5p3AeHLNAShW4+W7DRafWL4B/T9z5kOFP9zh09caiJ6w+vUGkKAP3PmIY1yYU98fshzp7pC5+6tQ5hLACd1UM9/oO47TlZemAphdgVGCKsYSjUk0RO41FR36OzHnujhQ7bruV3/32BpxKmfy6NUQnvRbrejsnTCsIKhVySY+SaLaGao/RHKr7siMPehoUipxouqO4qmekuNgouLLyUfq4hvPc+9iu5rHOLuX68G2cFyomACckLCfs2crtb+LwuSG/utkbs2QRMBpWra+Ln783brnPwXVgKKfI7WGUUxAphgvww+s9lswrM2ncHszv6jwn6uKnTp1DBBH4+HCC5b4hoSDE8OeKy/80lpjlHvwPx55IsT1SzHBs3NKuhX6ragXGIz81QgJI6PiYkIiPBa0MX/ZetLWkN66muXMz3TPmo5WuCQJlLW/PVvhsLsWAxMJKE3eXwe7i6pkwxGmvvMS9XCULSVW1lhDN9ypvY7m8mZSGnFVsCBWFyPInbyaXFFwuTwf7zYbizS8N+fR3hLK/c4MKoTErHHv4MwufKII1mzUKmDvd1tNkL3DWb1U88IQhV3yaF5BAuaL40e88Pvzm5zb1vM7u1MVPnToHE5F9NnJa5mse8w3J6oXNJTbjvLqQ4LPN5QO3xmcgFPjMcIJHA0NRIBfFesWXOKSfQHBUtV9KoAUh1IAoMkogn6eIIeHHx1Bsn0h6eJDmzq0UJk7BGoOylvM6n+LseQv5Sl7QEn947a0XRiHIM0R+DLA9UrUIVRkIRMgSR6VcFYu0QGBVoGnQQg7FGp3l0SGX60suX2kuMdl57sLT8+CT7ynz399MUPYVWguNGTjzuJCTlz69+Fm1XvP5Hybo7ouF4oQ2yxVvKTN76sEXxHV256EnNdu7DaVK/NaXvTxNkVWs3aKZN33/ecfV2Uld/NSp83wjgnv7T3Aevx0qRaRlPP4Fb2fHtBZWu6uwyjI5nMLscM6Y0Xd3V5wxDuYQf3juiBR5Cxm1zzpqfx0GPym6fD/vsikyOPkylW1DFCoR1nPQ1sLCCYQo5piINi0M2XimDsSzf3JBQEUr9KiFOwmPcpji8v98I50vfS2lVIZTNz3Ba/7zCgCOcC05q+i3im6rkOpZ0kg1FRZPcH6alQOxQCsKTFaW7aKq03wUeREmauGrTUXuD13u9g1TTHwB2hZprIrF3JO+5iNDSb7VUqq12z8X3vCSgKULIv7vBo8ggCMXRfzTRU9f7GwtfP6HCXIFRToVH9dQXvG57yf59n+W6gapB5hSGXb0KNpbhMbsvj3mFzd7JDwhlQDf3/v9rECxuPN5rbN/qYufOnWeZ5w7f4F7z7XkShHbtw+RSfewPX0lK157Do5JoVB0e910mk6OKp2Kq8BVMMUR7vJ3TiZG4ovxsIU396dp1sJrqrNzng9+UXS4tujSaTVBYOnPB0g2iQrLiB9hN/eRHCohJ8ykO9L8S6bM/b7hwcAhb6Fl4xoW3fk7Hjj6bBKlEsYxOMkkSim8TIaZ8+dytu2jdXIbs976nziJ+MjPSgY8EWjGO0Ii0AxILMQmGku31fgiDEsc2Ql3EUEjKTdPQVrBsFXskPicKuLUXJsW7hqXJ2vgdCrIcIIHfcMTgUarnXVGVsWO9L8vu7wqvX/a0ZcusCxdsO9RvLWbNd39mnRy5wVSKejqVWzaoZgxqX7hPFB899cuV/8qwcBwXJs1f4blR58p0twQ/14E+ocU6aSQSsZC9cs/8rj5bodcQeEHIAh7q02r+OAYuOjU+qyfA0Fd/NSp8zzjPHY7yx7ayoYNvQwNlSlWQvwXT8F9+GFSEyfTNGMmvjjcbru4eriMRC0c6Ub8c9rnhqKDbxXaRmwZLLPtNo27LuKOQhPpqbDjcsX0iSVmPg81QLdVXBwVt5YXQ4uk4rYnySShaxilFfYPj9N40kwSSrit4pAXxXTHku/pZfvgAA/OP46wVKSSypAoFkiYiETCY5oOueyqr+Gm07vt98xExMog5K8Vh6wS+iy0aphkhKSybAg0SeJLyq6XDVud1ewhpBD6qv7vBqFBx5eihIJo1PXojGTAPb6pmY5CHDtKKcEDtoYHL7yiRwzKdkXt9Cj7e8DaWNT5AZSr0ZIHnzC0NgtHzHv+apxyBQhCxaoNiq/+OEG5Ep/kIISHVxpOe32G1704wIrw+zs8jILxHZZjF0dM6rDc9ZBDUzbeThg9/ROkgGxaOHxuvdj9QFAXP3VekJQHBnjyxz8mt20bmQkTOOyf/5l0e/vBXtZ+YfuGHaxf34vvh5RKAao1hW30CMs+xd5edDrD2uaJiApJJzoZKLTyk4LHL4su442wbTBP65bHeIu+jrlHrMUcadkyOJVP/f4/ePiqSfz63z0+6B74AskyMGRHipalNkAQBbRn0dc+DH5EJJBE2Bxp2k18pe7q7kWUppBtYs7df2TrESdQbG6jmEyyeN4svtpSxt1DLY1PhVXeKpa25zjRn4SpzOaOssNvSh5PBhpPwenJgISCG0ouCmhSA7zC/TUZlef64FI22lkA5KrTp0cGHjpiUVRo1YOsrwxyZGoKAMe6lpO8kDWBpiwKkbjeaqZj8YFTkwfvm/nsqZaJHZbBnKqluERgUocwdcKhH/XpHVB86f8SrN+q2Lgt9m9z3Xg44IQ2SzYDEzqET72nTHvLgTneKILBPHzt55PpHkoTRbB2s2JgSBGJIozi+yDgB4bv/FrTP6zIJAVjoHtAMZRXdPZokglwjTCcj4d9Ph3GwOQOS8WHVPKAHNo/NHXxU+cFR6m/n5ve/GZyW7fipFKk2trofOghzvriF8mMG3ewl/ecWbejhNGKQqH6FXa4AoUA2+DgKE13ySdsjtM25aCdzZFmCIURxZSgRGb9U7xn9reZ76+hUM5S8DNMadrMlS+/gjf+4H9ZvQ448sAfxxQtrLCapIJIxyajGAWRQK5MtLEPvWgSTVqYqi12VAFKaAWLQhQ09nVy6g+/SK5jAsH4yYxfeAXtexA+eZXj1uQt+MrHYAhT22j21rKteDEZLbgqjugMW80rkhV+XXI5Rt/Px1L/g0tAhOEC54/8KriMXwdvIiT29goFEvhk1CBTdDcpKsyUz+MGpxC4r0cp+FCjz1FuxMeHkwRhhWlOgkBgqRdxrHvwClKVgn9/a5krv5egs0+DwORxlo+8tXLI1/tYCx/9apKBYcWmHZq+wThVVK5AMhELjSXzI/oHY4H06ff9bUX/V37X4/rbXPxAMX2S5ZsfKzKxI97/1b/yuHe54fHVhkIxyaypsL1H09mr8IOxZrSoOAI0nI9vr/jQkIFcQfHYU7F1CUDZ3zVMtXvqK5UQshmYNknI7h78rLMfqIufOi8oRIQb3/o2tj35JK5ALgFh1IMpDbHhmm/yjo98fLei30MNdfhpsHoTIPEFKrI8PHQsaxefi69SZHSZJeEqEtJKOehg0MbdSCIBxZ5NTJ6wg9ktG8jYIq3Sz0DUSnf3OCZ6OzhswmO0b1/wvIifd2crPOin6LGQTTqorhx+0YfAYodKuElF44XTyOSHeN3EBH/2HXojjVJgm5qhuJ1kIUfblvVosbRuWc+mU8+jtJcaiEe8hwlVyPrSBJYXplKxDgEBg1E/LWp8nLYSWB8qPjicRgPvSl5VTX/FhhQBHi91r+XPwaWUVRtiod1YFukngBBfHJY665mkhyH6A6FzCaJaADgnFXFcosC3Nw5AYgKnJgJO8J6dGeuBYPpk4Vv/UWZLp0JrmDxODvqa9gcPr9B09ilSiXgeTr6g8MM4siUSi4vOXsWU8cKGrYowBOdZXtE+9MUEv/6zh6nqkeVPGS5+Z5ZlP8nzvzd4/PkeB20gX1QEoebxNYaENyK+4seMyHRFLEZtVctENrZeKfugtcJxpDbJWRjd6TX2yXJM/LvGjOVFpwfP+pjq7Bv101rnBcVPn9rB/aaRxNR5JBrKpLwyygqifAoDj/LpYBUf9xYc7GU+J+b98+vZ8sDDpMwm/O07uOWNH+Gu+W8hVfBJpgJKNHJT93QOV2k8rRAisAVSqkQyuY0GN4dDSLceh4tPm9dHYmqZwUIzLek87zn2aVpI9oT4aNmEqCZE7XtkbaIj/Kq9yCt70/RHmqmTXNaVEvg5n8xwD22vnwHlHtxV25mzfB1HXfpyPjmcpCtSJCdMpDg0yBG//A5eIYf1PLYffQrh+Zcw1ew5kpLTeW4fnM+D+RmUI4/uoIEQRUL7bBVNm4rnCg1JbEcxVffSoXqJGOu7lcDnGOce/hq+mFelKyxyhlhe7sJiWOqu5a3Jm6rnpYyOHiJyzqk9tlHDS6Ne5ja2PLtzfIBRCqY1DuA+dBPc14OdsYRo3rGQPLSswcMQCmVozEDfUNWI1saFw8Go7GIYxRevrr5Y/DiGPdb9RFEciWnIyG4iQgRuusutCR+IC5cHhhTf+bXLw6sMrhvva0SsBGFcS5VIxD9Ht6mr6pcZaxXWgtZx96FYBSo+ttH337XFXVVVkQh4LnzhAyXOPL7e5n6gqIufOi8YVgSa7xdcZrSOw+3dTiLhYzFgBBVF2NZGhpJreF//YhYZzasyPs2H4DC3xqlTOeo97+HJn/yEstU8csEr0QhBkCKSRtxMmqKFVVpY4FgSMkhFKWbpHaSaFP25VrbKJGbqDUxkO/07XMKSIUWOww9by6RxS/Z5LU74R9zop2jpR0hg9XzK7sdA7ZubYqOG77KV//rrozw0eQGlSTPQjQ6lxdPoDHxaurcSGpcHf/snLjv9DL7V3Mxq61ARWHb0XH7a+AG2rF9Ncc5CGiZOZKYR3pbZs3iLohSPF6fiKEtflEFrcESo2ARaFBvQmFHVv902S4CHHmWwoYgvWsN2Il41ndWmLSn3m6hdjTiURtSkfT6Xo7EWikCaPV+U9zd6xV0krv8qZssKVKWIiGAnzsa/+J2EJ738wC/gOSICP/q9y63LHIolRXOj8OLTA5IebN6hCfdQVmVtPKXbD+CYRbsXPf/2VocbbnPJFRTZtHDBKQGvumDnhsqV+O+uKAOPrHJqbeiOgaQnFKpWE1bi11EyIXguFEshYRQXzls7UoGuQBTF4ohwkjgitBe0EpoaYgEUhpBOCb/5i4e1AWefWC94PhAcgpeOOn+vXF90cT2Px097ESqpUFJ901uwrkPX684gUhXW2gJ/qji8eyDNluDQjO9PP/NMLvzOd3jVzTfjTZ9BoqkJryEWPgApBdOM5Z/SfXwm/U3OdJajFFRUAp12+U3+UiaoLjauztDfm6RY9BgqZDm/7Zs8/PDm2n7EWrbdcw+P/eAHdD36KDLq66aSbrzwuygpI6QBg45Wkgi+/qyOZe0XPsMF13+PRBBfSULXo5JIUUpnGeiYRNQzxE/vbeDyV/2GN7z+D1z/hbtZaH3elvX52IIOFp91BsdMGc8JXkhKCx8dTvKpoQQDduxz21A6jFzkMRBmCMQQSdx2HojLUNXVvQI1J/kiSZaFx+IRX8U0kFAhfTKZTRzFDBOxMjCgMkTmaGCU6JIQq6Zj9cJndS4A/nPQY2l3liM7GziqO8OXh91nftBzwUYk/vQ9TNcGVFCh7GX47WGXcPX0s9n8wK2oni0Hdv/7gd/f7nDdX1xKFYXS0Duo+PYvPU5eGrCjd8+PsSJ4rnD0ooj3/vNYFfPISs3/Xu9RKMWpwGJZ8fObPG5/YGcUMJmAhqzs1iwnFl58hs/0iTsFy+ypFqMFrWMxq1Q8SdvaIE5viSAST9h2nQBjIhwjOI5lybwA1xGs3ftnlVKKQknFLfChwlpFd7/mqp8n+NWf6jGKA0H9rNZ5QZC3sWfVyuaJZNq6ufvN/8ZZq3+I6c4TtWUYfP0ZBFPbwCqanCGaMw+Tt5Z3lGYyIzeTV6dDTn2e5tvsL5TWpNraaO/SbK3WwowQCZyXDHhlukyqcjcvSyzj/nA+O6IUp2eu5ZrKa9jU30Hej9AaQhKIMrQ05PjzbX/i2CMWUBhYwR8/dS8bH/cZl/FZc/31dCxezCkf/zjacXCCG/hu6ULuCpdSFpeZppMPJH9Fu35yn48hLJUY3rSJxxefyI4J0wAQFU9bDI2hmG4k/e376Ckkmep4hKHlwQc7+eTn7yP3nnPYEir6rGJLqKigmGLiAYLLKoYbyy4/bi3U2vZ/PjSNTWW36t2liZNbYwcaKlStfFQBn698iJJNM8+sJq1CkrqFX0cfYLYrlC00VQcu+s77gCzGPghEWD2XivvuZz018ts5hx8XE5jqAgqi+UY+yXzHclH6wLw+Ve8WVH6Axxum8pWz/5vbZp5K2Xi0Fnv5VVTigvVdvLNj6gHZ93MlDGEggm/e7dGjNF5O2LFBUwniaNCmHZrmrBAEQqk81shWK8Xc6REffnMFd5cr2fW3urvd5jrwi5tcHlphCALFBacEvPPVFa78ThJRqhZ1mTs94pIzIk48wvLRryTpHVIYA7MnFznysAR9g5r+IU3PgCLplUh5EZXApVhO1uxblBIQheeW8StdjG+O6BnqIFfMEtnRMYf41RpZat1fChjXFteSuW6cmnv5OWHdsmQ/Uxc/dQ4qYbnMxjvu4L87jiA/fjoVnaA8eyGVzgaevOAdTMzuIBIHtMKokFKYZXLLnZQxNKFoTe6gp7SDd/SdzqWpgA83Vl6QRp9Px6ebSrxpIIMv4CgILUxxLO/JBohqweqZ6Gg9J7irwAVlk5zr3kOx4GC1JkIz4pPuoTjnhF+weUWF/h0DTDtX2JqZwjd+dCQnTM1x0uOPs/YPf2DepZdydekIbvSb8FR8UV4RTuMDxXfwvYar9n3xVaHz1KzFRMbQumUd0++9lUJLK+tOvpBo4wC6c4CBN17AwBlH4RUrzPvDI9zS3Mb4stCjDd2RoggIivWRwgUatJCP4uGNV7WWeMLX/LHs4QBBVezYPQSuR0TPyEXS4nBV8D6cACIU44xllrEkFUwwwuKRTi3l4LtvB97+7J/AUfykmNhjufbX8gkuSu/BwXJ/kGxgZcMMrjj5claMm09Ppp1QGQbSzXQFJTYGwvJrUmQHhMVzIi5/UYB3gINRz4QfwNd/4nH3Socnyobc9jiiUurS2IDaJMmuvrhQ2GjwPMH3Va0+uLkhjtrcdKfDJWeOzYv54e66tWdA8eRah56BOEW27DHDRaeG/PRzRT7z/QT5guacE30+8Lp4qvb4NuHb/1XizocNO3o009s2ccoJMwAYzMGnvpNk8+ZNbOpqpm+osSZ4/MjFdQRFhFE+6USJQimLa3zSiSL5crZW72N01feueryuI6RTUKnsfBUXCoqyD+l6u/t+pS5+6hw0+lev5u5PfpIn2qex7uJZOD2Pkpq3hFwqTd/MedzVOZnzt32XyclthF6SFePPZfqkbfgYIJ63UsGlLbkZq3NcX2ig22q+0VIieQhlw05OWv7UkePjQyk6I8Vp6YB/bQjwqtf2svvvJOV/0LIRsERmMQu89/OI+iyGpwgxaISUtjSkSxRyms4uRVgyKCynnLyZu+8az62rJjO9oZXW+x9k1osv4Y7wODz1JCNXE62EbtvEbcFFnLKPH7ROMknrvHmkijkO/+3/oQOfHfOPwDou0x++k/4hl+5vvpdwageOCAXg/nedT3ZzL+MjSz8j3S87DVAjYr8wAxQEfljwWO7HM172FjsZuVRYqDnJO8RFqJbYQ8xH8K1inWhelAz5YMP+bwf3ZffhglpBSQ7MC3LLlmEuvPA6drzijdjt45CWJlw0ujo7uKBT5IzLfc2W+auE9Vs1KzcYPvuv5ee9IyxfhG/8NMFTGzWrNmiCECrjFKVB8AchKsYpL0bZ3YlA0gPfF7RSuG78u1RCOHxeRNKLO7R2FT/HLo5YsS7uzIK4UHnF2rguZ2i5oalBmD3V8pdlDq84L+C6r5T2uGbXgbOOi4CINWt2TvFuboBpEyybtrQxqa2L7oFmRBR+GF9Sw0jhmpCy79E90EK54jCYbyISPabQObIKx4BxYiEUWoVSQnFU135zY2yFUWf/Uhc/dQ4KIsJvf/RL7jvxRWwbPw03EbBg1iqObnyIQraVbQPjmfj+bzDt0WWYMMAgnDjjN+RvfAfFTEvtQqkAURHZRBdbwkbuqRh+VXR4bebQGgk/3YEftO35AxjVRDnxeZQMAAFCByjForlf5J4/v5+2xo1oo9iybTJh2cVxAvyiX4uLVHzDmWds5ZHHJnP1IzOY2pPhB903s/pfzmXG1KmYvlUwWEZFgqM9tg1auHDf1378Bz/I41+9mt/TxIZjz0DbWKLkOibhdQjh1A5c2WkzahUUZnagvXj2iVJgRAhHOZlFxM9tqxb6bJzuqBCLpJGS0pFriGan5QTVf7tAGqGAiqWyggXG0qKFksDbG3w69tGQVASWB5o/lhwc4CQc5u7lvvMdy52+xhklLAKBE7wQEXgk0Pw077LeamYYYbETkUexPDCEArMdy3sbKvtcyH/WWT9jYKCMbUjF6UZjCEoWL61RShEqB60U5eZYHnourN6oWf6UZumC56+TSAT+4+tJHnjC0Duo6R+Kv7zYLkF7ClVVrlJdkoxMqK52VHW0QnNDxGBO47kwebzFc2NRM6Fj9+fxkjNCHllpeGiFobNHs2mHIoxioVEJoGcAyr5h1pSIh1Zozv0bioovOz/gzofHYaQc1/soQSuLFY21sWGuY6Czv5mK742yshg71yey4DngukJQUuQKkHDjwmobwasuCP4uRhe80KiLnzoHha/3RvzijFcBQnfbJE467C5SUsL4EQlV5vA//4Tm5bcThQ5aawyQ7hzAfWw5hRPPqH3Ljz8fNQN+G65S+AI/L3ics3IZG//4R5TnMf+lL6VlzpyDerz7g5F5MyN4yWaOO/u7/PKXT7B2zQCNTWnec+p32LBuXVyAqQyuRGgjDOU8BmwDqSAkPXESlWGfobW9bM7BLL9c7duNUwrnX/9pnLCb8MXv3qd1eQ0NnLlgJv9nZuBJ1dK9WnsTOi4ZFeLbnX1YWkEyYTAanCgWOklg5Hu1AFqg2QhZFdflnJIJeGQwVTUwHRsBGrmEj1xWRsSQU/V1b9Ew3ViaqtOlEUX0LIYBfyPv8ZOCS5fVlAR0cjFn9EZMc2Jhc6QX8ep0gKPga60lzu3O0hsptIrXOd1Y3p+q8LLfprh3pUMloVCngjcZmlXs23C4G6EVPBYYPjKQ4nVbK2zcZjj+8JBZe3Fnf+CB7QwMxNbgsn0IJjdXr5iCLwbHMRCADUGvo9p+Hf9dud48r+Jn9UbNQ08advTqnVE6gaiiUC5xxMdQC98ZE8/LFIkjkhPaLe94VYUf/S6BU61ZFoFUAi47d2x3YLEMX/xhgtUbDWs3a3IFhZW4pieycYTFWkX/EFQqhm/+1GPD1oi3vsJ/ViJjQrvw0bdW+I+vT0drRWuTpTFr2dKpKZYhsg6F8shBwc6Y5Fi0iidBlyvxGrWO34quhv9+b5m5dVf3A0Jd/NR53tkcKG6zGSLjsH38NDoynTR4OcqSxrUldLlE6vaVuGGA1vFL1FNC4IP58wbajj+WIZOmIgqtIjrLExgK2miqDrkb7Ozink99CuO6IMKOZctY8qY3Mfvii/e6JhHYGqnqxerQGRKXTru8/vVL6V6+nPLgIKn0KWTSaxkaUkTWECoXYyP+947T8NNZ1LQWtiUamI5lxj1PsfnkSfitLo6NCIzLhU/dzLShrcjvv0FuH8UPQKq9jUTRQ6tYxGiEhBLyYhlMJmg0Gt+PUCi8hKZBC+/LlvlyLsETgSGjhakqYn1k8FQcAWlRgih4TcrnmITl6znL5ii+8o1EfzQQIujqRWUk4lMBWrTllISlaPWYNMo4Y5nj7NsFZXOg+H7eZbM1NbGN0txUMcwJI2a5wrqS5qlQ88mmOGJz37g83y86LKs4nJEI2eAbzvlylm0bNHjVha8F/1LoOVzTjNBlFRONICW47TuGpwaSNDvwyz+6nHhEyHmXh9xcdsho4TXpkHYjbN6cQ0SIlIJrH4EFEyBXgpYM2KrlQgF0NxT/qHjMN8ybEeFoOGL+89sc0Nmn6B1QtZSgqda2AVgfnFTcYm5cCEtVk1kbR0MmjRcWzLS85KyI9pYy1/3ZI1dUTGi3vPXlPs2NY/f1+e8neGy1YTivCEJVtV+Jo3+eU2ZqxxZcJ4yLlP3JeAmXm+9yGN9mufSsZxcxXrrAcvYJIcmEg9aKtZvjYu2dqF1+jkWrant7VSwrLSyabeloFbQRpk6oC58DRV381Hne+avvoIyhf8I0RCDr5dE2QluwjoMOfFRDMv70duIQcSgKi2B/so7eNy8iHDdIEcu28mSWDx6FAANW0UREcvM6HM8jKJUo9fVhHIcVv/wlM88/H72Hcak7QsW/DyXZESk8BROM5d8aKsx9HsxBnytb1vXz0X99nO3DjRhJMz/j8L4PLMGkVhO0GrpkHN9aexlb5ifR/jaS4xrpjRS+aObtGOSIz/6el10OfelWLl51I4u6n4q/eobPblDi1JNPpuN3D9CdaalN4BYrNLU3UzCaUCmSyfjcRwIXJwNOS0acliyyJtBcV3KpCBznVlgRGtaHGqPgTWmfJYn4AvDtlhIfH0oyLIoGDVmETqtYE8a1HK4SnGr7exKY4ggfbajw2VyS7qqJZLsRPtBQ2WfTz/8e8thkdS26FL8i4gdvizQznYiEgid8w7pAMduN26Hfkg15SzbkyzmP21cZ+jYplDfK+8yA3AFyeLzdvFVghE03aCr9CjyFW41U/e8DHldP85CFsTj/dl54d9bnDRfNxPM0pRDIV+ATf4DzF8JFh6OWTEVthOZhIbwRnHJ83tdt0bzsnICFs57fi+oR86I4ulPVXKmkEIRxBM5UT0nDZMGEkA4FV8Wt4XOmWebPtHz4zXERzClHWk45cu82FoVSHNVyHMgVVS3SZQWaMznmTVkJAlYUcyZv580Xf4arb/46uC5/fch51uIH4MzjQm5/wIm/QHXpmo3FTvb8YlOKWkeaqs75mTze0lH1KMsVFJ29ihmTX/ifQ4ci/1Di55prruFrX/saXV1dLFiwgCuvvJKTTjrpYC/rH46JxsZph8YmTLHIjvxkFncsR4nEhZqOQ/5dZ5O+/mGkEGCiMJ6ealyKS46l0zmRx7rjD5sB2dn+KkAQWoqJFMNbt1Ls7Ix/J0Ju2za2L1vGlFNOGbOWnkBxSV+aPhvHDhJKiAQ+m0tydUspbll+gRJF8J4PdNM3PD52JtdJHimewGe/lMD/ztdZs32YbQOKwpxm3A8Kzo+W4WzoASBnFRUrLE6VedP9Pxs7ic9aovEzn9VajOvyH8fP5UNrc+Qji2stpr2VS2dP4KJUkfcMJtlhNQmEl6QDPtq0U1zNdS0fGmXEeuOAw62VeBDiLRWHy1M+VzT5HJuwXNFY4Tcll2FRtGnhvzM+/zOU4O6KQ06oOa8rhMOdkHmu8J2WEiuDWCAtdO2Yepyno2ThDt8b00o/gqI6Y4g4mGMVrA4Ns92xF8/lvqG4VqF3VVsiMBihKkLkKlLVMcOFbXFbdXO1Bb87UvQbDY8J2cOqfm+i+G7e5dS2JK997WK+84PH46t7sQK/XY5XrKDz02l+3LJELHlR7MjEU4ebG4SPHQTfr+ZGOGFJxK33Oxgdp56aGizliqK9WZg9La7hCSN416t8zjohrA0gTO5jsW/foOLrP/F4bI3GdSCTqkaPHDBamNy+CaCaDNWcc9TdTG7fxLFzb2LZU5fsQbTsTu+A4rpbXMo+vPj0gBmTJW7FD+HBJw2VffzOoIhTexM7LOPbLdu6NAtn2zGiPJPigJm11vkHEj/XXnstV1xxBV/84hc54YQTuOaaa7jssstYtmwZU6e+MGdgHAhKfX1UhoZomDo1TgsdBE5LRPy0IBgFbiaDSIb1w/NZkH0CFVnEaKQ9y4YXvZQZP/45KrIYrdCJJPlZC/DRhDYeSWerxYMjdQRpDamhfgb7+vF0tb7AKMTRrPjZz5hyyikULbgq/vux4QT9VtdEjo9ibWRARawKNId5L9yw890PW1hzK4uG7sGxRfLOONa2v5IHjj+F9r4cU2dNpDmClaHBUVC5ZAn2C39CKUUQWJwOj3+5+l+Rz1+PKu9sw5ZkmuIHf/Ks13PE5A5+M7GDGwqaXjGcm4pYWBUDN4/btzbvzwx73FDycKrpgLLA1YUER7gRF6YjTk1Gu81zurq1xLwdDUTV14ECXBS/LCf5fzagWcPiv+F5vL1saoXXuw3CI06xOdTKbFhoxq5LBPoj8BcpzCpBDavq7RIPm8xqlFQo+ApVKmBbm8AVxpv4tQkwWK1Pckd132kFZRQ3lV0+97kzmfaapXz8itsRPyT7oXPRR0zF3CfMKVtUAhoyQkMmPoKkJyQOQueQCHzj30u861NJ7n/cUCjFnl1L54Usnm/p6tVk08KlZwWcdnR8HvcmekRiEZJOCZnqIPKBYfjnK1IM5DTWxl5chWLs/1Xx49TXnEkbyJUztDYUOevIOzhy3kqCKMGsCY/x1ycuYckzpAIfWpnl2jtSRDZ+Dm67z2Hx3JB7l7skXNnjFOo9oRRk00JbkzB/ZuzafvJRId19ujaCwA/gnBPCuqnpAeQfRvxcddVV/NM//ROvf/3rAfj85z/PLbfcwve//33+67/+6yCv7sATVir86T8/wwP3baVBFZg3zWPe5ZeTOOs8WrTQ+DzOxnEVXNlU5r0DSR4KDI5SdO5YgAz4TJ2wjdlPreSB4ERmrL6J3KKlTHOh0Si8hkYS6x5j9ZZNDI6fyW6GgEDJOGRWPU6/myKrirjvPhy7pA2TSrKu6PBv3QNs0ZNxlDCjkmNDWKm6IY5EDOJ25c5Q8cVcAk8JTVp4a8ZnzgssDfboj3/JtP4/ICp+GzcG21na+Q1uar2GCdXjSev4fAuQmjeOeZ+4gN67NpAdl+EHl06kIemQu/opEj/4MM6qZYQLT6LyhivB2zd7i13JavinBsvY/qt954aSu1tkRgNfzSe4cC9zcsoCGS0kqq3wroofU7Lw3bzLBxuDPT7umSiKwhGpDU4cjQKaETZFmr5I0agtV+aTvDFT4XDX8uOCx/cHXTYEGmlR2IvBbBOiv4JEFiy4R1VoGeqnZaiHCe1dvMndxrYTTuYPN89mRI0bAAHvtN3PiafiVb37yBaW/PblfGIoSU4U89yQNy2o8KFfpPADRVNWGN8uRBGcfkz0vEZ9li3X/OT3Hj0DiqE8dPZqCkVFWzOMb4/HEJTLim//Z2mf1rVsueb71yXoG1QkPJgxKcJa+PVfXIZyCtepTmAmFqRlP372HAMvOvEPzJq0jdlTLWUfuvsUYeRz92Nz2d6jaG7Y+/vbWrj2tg6Uie0u4hSX4u5HE6QS8f6ip5ngPKo3FRFFpSKkU8ITazVtzcKJSy3HHhbx5DqDVnDa0eFu7ft19i//EOLH930effRR3vOe94y5/ayzzuK+++47SKs6cIgIPY89xtZ77qF51iymn3kmn3rLN7nlnoBQJsfto7PnkSyPI9vlkE64HONF/L+GyvOW5hnnCD/vKPGHksMfyw651jT+wwOccs0faC4OMy94gJ6Vj9E6bx5eQ0PtcVks0YP3oi+ehbWCsnGiQwQSuX6KreMYXHA4yksQvXM+cvw4EsrgJJMMZwzu+Lug86V0rVnP2iCkP9uMpCuE2QYcLx4KUhboQ/NUqOi2isDGF+WXJANOT4bsiDTTlLfXdufni9ID1yHa29kfrBSOlJn60J/wqiLfqDjNuCXUJJSQmNrCpFe38MasT0Oy+uHqpai8/WvsweboeSfYw/VHqzgitzdG5qY4apcPNBUL2b8Vo6BSnT80ZrMiNClhEEVvGKeoZhthUBRfGE7SqIVHioat5fgbReSCGgeRp0jnKpgnhmhZ0EtTWw6nN/z/7J13mB1l+f4/7zvltO19s5tseieBEEIPNXTpVSxUUbGhiIpYUCygXxAB8QeKVJEugtJ7C4QkkELKpu5mez/9THnf3x9zdpOQ0AIK6N7XtRfklJk5c+bM3PM893PffGP0HxgfXUuliHH8/n8lmjmTxxeeRDprMLXUJzdXky7ffHeidKBdOj4SkLqsgmtSIZqVgS80ry+SbHw4Qk25ZkOroKVDksoozjjW5byTPmDo7YfA2ibBb28OYxhBQntfPNDDGAYkMxCyfeprYPVGyStvGuy187tXXnoHBFfdFpTATDMIPp2/xMZxNZmsxFcMJbtbpkZKjfYFdih47L4XTufbJ/2Kjh7BlLGavn5FKlvOms6jGT1Cc8fDNsUFmoP22HY7OnsF8ZRJcV5c3d0v6OmX+RBTjeO98/EZDumgHaZBSEHYDjRHG1sFs6b4RCOwYKlJeYnmmoszH7sB5f8K/ifIT09PD77vU1lZudXjlZWVdHZ2bvc9jY2N72vZ7/d1/ylopVh1/fUMrFgBUuJnszx89a08umIsBhqJS66mnNbD5hJ1HUY2bUDXjeCpFPjdA5zgd733OtB0FXbSXdfFsvRSXOmSs7IYGU19cgy1ifcfBjkx/weQHVvAxsVFZJWDWVZGdNw4XClxU6mh129yoXXEWITWhFIJfNNEmTYI8KVBtH0THbP2ITVlMtZeNfimgJxLKuWQKq9E+Um87hdIdZgI36egv58Zj/yNF875PtmaESgpcbRBIS7NSg6JMRO+5vakwT8TmjpyKLuBZ9bH+YLX8S6X5X8vvHQ/tiFx/HBgCIJAYTBtxRP42WNZnwtjiyC1a5wWTFQpKrIeB/t91PS77MiR67iKpAMrGxvflpW+Nbq7s6RSHq6rWLq0n7q6KLNmlWGa2y8xaq1ZsyZOoVNN+07jMEObT02+hpnZThob295xfUWhKfRIe6vvwtSagztX09j5we+gcwhutsdQIyAjI/h5EmRqTREuvpZYWmMKQcaHVa5H2TqXTX02coIgZehgXh+QORBKY/uaMQ0DRAoWkC4qxffhsKrHGR9txNEhUp5B2s3y2f2uY97sNbTFjyFsKxbLGFe5I+kUNgLNCJ3ji5kOsvEBGoELrbEsNgSG8NEaBp6WdPuCXVSChppgAtL1BKMrN7Fu3Tt4SX0AvN9z3v+7v5ZEQrOh06YrGULYGp0JtEeGoVnbLCgtSKKBx5+PUxnb/rl4EPc+VcFAvGRo1H1DSwStNamMEWgCB52/VUA2BisxjguZnOCxBfvQ3f8zzjj8dnJuNwtXjee6v59PTyKMaWjKi13ufDjNqPImOnotHni6kt6ESXGBx6F79GAZ9aTy56L2zjDprIXnC1KZzdEUb4cpfcqLHNp7woRDPralg9iTTEAEPS/F4OltfbPgtvs7mTtr4H3t3w+CCRM+7lu1Tx7+J8jPjuD9HCyNjY2fuIOq6bnncNavx5aS+MaN+I7D/HQFyoljlxQjBHQfvCsScJSJbdvEYrHgvTLKhNKS91zHfPsVWs0WMqkMiaIETk8f3LIM+dQGuorCZA8/lkOO/vYH3/gJE9hpCwH6m3/+M2sffhgjX5FRnoeoraN9590xPRehfExX4WuNb1pIz2P8/KcoOemztP3gVxRH/ozqHUB54BaW0F/VgMz6kOkFvwwhBJbvYTo5yteuoKlqBI5hgBB0YmNm0uz60G3UrF9JoqCY5cd+EW/kaIpsk1QqxbKCGrLFJcz4mHRBo3eeQua5RYQtNyBA+ISJU3H4yZzfEOaxjMlC16BUak6NulQasfw7iz/wulzX57tXLuLJhV2k0g6FtUWc9pVd+f4uxVu1KxIJh0sueZ6nn95Ad3cG34fy8jDjxpXy3HMDXH75/pSWbm0f3dmZ4kc/eoG2thQhtQEdWUbuxFnU7FJCMlzExLDBb+oKMOU7/9b+mlOc1ifpVQKVb4N9oyDL7oUfTLg9iAU5iRcPM0ZCzFc0ejIwjdSQU5IaU5LSAkcFQaqdjxkkFoXI7QeqXeCUgYgGwl5JEFwZcTWWUc5YQ7FcGhgCdipahatDyHCY6rBJTAbf0ciCNipHjQOCm4NT8On3MyS1oMbQmKIKqAJgTVsBg6Ea2gWSgV9Tm11ArFOzqTO40P7sT5O4+bIM0ybs+PH6Qc55SoZo7LHpEwKsvG7KzHswCYGvBIYdAw37zDGZMOHdj8vC120KC8whbb7WgVeUQGCY4HuDVUCRT3wPqky+CvxztBYsXjObt/7frkDgqyMl2JbGN6Gzz6QvZVFcFuZ7N0TYlJMoCeUZTdu/Kqiv7qG9rwLbgqxr4HpBm01rsV3yIwSUlQikEUIIgZQG5pCxZvB9hUKxIRfqqIaEU8+ECVXva/8O48Phf4L8lJeXYxgGXV1bVzW6urqoqvr0H2hOMombTBKtqqLlpZcQpkn/ihUopRBCYEsXpRRuKoVdEBtqWgsBBTXVQ8t5P8ZvcRGnxdyEiUnOyuGkEqi1PTCxBH7XjvA0K167jnGhqYw79LAP9blmnHUWdixG8wsvoFyXkvHjqTjra/zRgbTvI0WgxTA8F+l57PLwHez0+jOsP+FkcjVj6YnOwkmvpcgQYAYGY1obtDUXBsZwQGFvJysPPREvGkP4PqZl4wHScTjisq9R2NmKsC1ivqJu2ess+vYvYKdpANgC/trssOzV1ey55wjGjCn5UJ93KyhF6C8XYi59DuF5+BWj6P/CFZgjxw8J1Y/51fdoP/484u3dCOkgtIcxairn/vRUTAFHRj2O5INVPURrI8Yjf6KvrZfi0ghi3+PwZxzA5dcs5qEXWpG2AVKQ6k5x4zULqL7iIM6qDE7kHR1JTjvtIZYv78b3gyqEYUi6uzPYtkEkYnLddYu45JKtJywvv/xVenuzRCImEeAHPMdn7/sWJY8rbMvCnLgbmW/9BbaT5TWISSHF61VJnska9GnBYRGPgg+hYysQQ7IbqgxNofBp8yUJDVHfYZxh87prkECgUuBPAqo03nIBpaD7QUeAfJtHBjIf9vNdvnbknpy3pJMOV+HJKEY0QknIplzm2y1ag9y291FiBDqjLaEUeHlzSgBMEFGNSgq6WwTtXXLouY4eyWe/H+WR61OMqPr369d6+yWJAoF0wE/AkENlPsTTNIKH6qo1++0aHC9aw12PWry2LCjv7L6Tz8mHBllbR811eXr+ZvJjmZpsTmAYGnRAgFR+f2uCSaquXrkVMdE6EEAPPqb8fBSFGZgN9g1Ifv6AzZsZIxjNV9CHoNwXjLA18/byWLjcwJB6aKLMV9AX37b+q3XQmiuOBQJ31xNDIu6AmOmtWlyOC7vPGNb5/KfwP0F+bNtm55135plnnuHYY48devyZZ57h6KOP/vg27EPCd11e+7//o33BApx0moKaGiLl5aS7ulCeF7S9cjlmGIup3jtF9qDJZItLUM+9ydqdxlFZGsIIB3fhroap9nvPenYanah8prZnuOjmOKQ9CBkQtqA7jc75vHDpTxh7yKGID6GuFEIw5dRTmXLqqUOPaQ3nxR3+kBO4Iih2G8pj0qrFHP3mU6yvqMExLQwNb704lknl6zBKFNrXhONJlg3sTMqXmJYgmuynfvWbLN/3CHKFpSgzOBMZwMRn/kFR+ya8UBipgwR2gWbGvX+iZ6er0Erz6uttvPyvJYy/5y7ulQOU7LIHf/rHV5AfQfxy+NovYb35FEiTVDpHsqmV1GsH8Wr5QYzcf39mnHUWBbW1fPOZu3nt9vtpXLCScfMOZK/j5+7w+kW8h4cu/h1/X1lK3BlBgeVzyAv38NlLDB5ZkERYwQXJG1OBs99kVNTmshZY1JLG+X8v8PwzTfT2ZvB9FUQXKACNlILe3ixKadav37qkn0w6bNoUxxGCJk9SrTs517wjMEn0bEISWP484eu/Qvb8//eu2y8lHPQRJadPthQjDE27J2hTwZh8raGYaypWOw792g6IN+APgFDgFgiMKMikRkQFXjfoEvAFFPRoJqzxufBUh1jE5q9z6nk6a5LyzmaK8VOK5BbbLTwc45j3tZ1SBr5Um3yJFIFhoBot8F4BvyN4jSEDohGyA++Ya/9q88tv/fsVXuEijREGDPBNAjW6CUjwc1AWU8ya4tPeIzjwnCiOI7AtTUkhRPN6+5sekNz6D4txIxUjKhW1lT6LV5gUFWjKSjTdfTCqSNPWZeD5wVh7eUlw/E0Zo0hnBImUGIrJ0GrbwFMIYjJMQ+D5ML/L3Mr5QQC9CDrTFledEmimvnhxhLXNks5eiZePaNHb4ZPKF/TG80JnN1iPEBCLauqrg+0UArIOTB2r2GPGJ3e69L8N/xPkB+D888/nvPPOY9ddd2X33Xfnpptuor29nTPPPPPj3rQdxqLrruPNm24i19eH9jyEaRIuLSVcVgaAm8mg8BGTS6l2O3D/0YF37Ezqvj6WurYu/Gm7k1LByO5k0+eEDYt4/dlnKBgxgnGHHYYVi22zzlJVOtRbN30L7fng+pByYdMAuMEcaFp30L9uHaXjxn0kn1WjaTI20mw2cXAoxKkF0/jzLY/TlMwxa/0SRjWtxjBNDvjyF3kp7JLWglnjavnND2oo3y2LWegzsDBG2OtiQmw1DQOrGPPqUyyadxJSKWwJIq9HEUDVmuX4oRACjQGEBCgEBf1ddPmaN97sI5fIcOTff0lJtgNhmIj5K/j98av45gNXfyjSh5PFfOtlkCau6zHQnwEhiEmHsoG1rPtnikhZGROPOw4zHGavcz7LXud8wP2p4aGMycuOiQQOCrtU3fxXbn6jHNsEU2qyvuSeVZVU3fQ4jrcHACpqkzl0enAEKIXjKP41oCgYVYvjrEcNOgvnoVRAfrTWeJ4iHN5aKWQYEl9IVrgSpeE3xj0INL6WmyMspIm5Yv4O7swdg8il2GXps1w66hBy0kQIQYs0ODbiMNLv4y+qCAkUCE3KBlIC7YIeoSm4A7L7aqxKsNdC6WqXfWYvZeczF/BIxGSf3L7UqFoOi3jATpjeWWj/AYROoEUhrnEUytj9fW/r9SUZTumNEleCTJuA6VDSroivCyLDNcFoNQTH9vrW/8xoZ2WJJhbSZIXArwLVS+BPEQMzppk+SvHWOskbK01cN9jO/rjAMgNH53BY07hRks5Ilq/RpLPBczMm+mRzgqP3d9hvts/9T9osWqFJZ0Tgm1OuKYr5PL/IHjIRVH5gJKi0QG2HHw/+Wtt6BCUVGtUcOE0DWIFhNs4WgWtj6hX9iSA2N+dAKi1x/Xc2MwTByGqfnC+Yt7vH98/NAoK/PWKRSAXC5yP29T41zvL/DfifIT/HH388vb29/OY3v6Gjo4MpU6Zw9913M2rUqI9703YYb9x4I9nu7qFbDu26ZLq6CJeUoKaU4yfi6NIQhAyEq7C1S/i5t5AHjGDuyL9SlVqOo8ZT/LenGHjycZ7JZCioq0N7HmsefJC5v/gFRW/zQCpTZZSrcnpkD2E3TLIggpIJxNMbEF7+NkaByuToW7PmfZGfVMrlxRebMQzJPvvUDzkBD8LF5V+Rf9JtdBHTMSxt0VzYzPln74mev4mNmSbCk+Yx6cQTiVZUMGEwJaqwiJnTalj0YgehUHDRFW6aI2dOp+qNxSRHNzCnvZHVBTHKSgoZEIKBvENwd8ME6he/BHaIcaaizlS0eYKO4lLWNPaSbepjzm1XUxDvxhOBkME0TDpff42OxYupmTVrx7/Y9AD4DiBIJHNDZ2YNFHhxuiIjaXruOSYed9wOr+L/EjYv5sygsgKsSIYw1HQs4xkgMNLrq6+mc79duajEBj+Jt76P7NQqlFZBEKXWKOUhtWJgTNWQN8+WGLwbjsVspBTst9/Wv7dIxMQdXYH3VjeGKSkhhSJI9/bNzUTJTyV58lvfwk0mKRgxgprzzufRopFktODgsMvOpiINRMW2qeo7Au+uX3H1HpcQ9bIEVisaDIvfJ6Pc7PfTFR7B41kLQ8CUasXKuEFKCHwfZBbCj0N9taJ2RJZdLv1/FJSlcZF0oHjAvI9DU4czVgW/Dc88Es84AkgDkXzQ1fvH1JBiQXWS326yuWORzdh2RSwNL4cEOTfYGYNtJqWDVtJ/Al86xuHBFRbxhEANFvzKgXlQeC8sWmEQnqTonwhWP2Qbg+3LuYLGpmAfDMZTZHJB9cbJPzd7muK1ZRbnHO/xg3OzbNgkqSrX/OrGEC+9YSKFSVt3cJsWDmlSGYF6l1F0TbB/DAnRBk3Tg/ncMQHC0ETq4TMHdAM1AEwYqbjrEYtMLp9h9w67dHDAXQgYM1LjK/j66Q51VcGz3/nif276bhhb43+G/ACcc845nHPOB7xF/oQiNzBAtqdn61prvvba391C6Cvz4LZX84GVBJqXiImOZ/BLBV5NAW33PgF3X0VnRxxW9yCAZHs7NbvthpfNsui669j/17/ear0Cwdzs/iyx32CNu4ZxFTuz5uL/Q9y7gsGrtGGYmJEIqY6O9/wcL764ieuuW0QqFVzsb7llKRdeuDszZ1aR7u7m9ftuYtlBcbLjIxiGRTaSJiJilKpSloSWcujuhzFi93e+S7744j25555VLFzYhmFI5s1r4KCDRqPcfWh+4YWALE4dycOhSsb5Pgu94Cex8uDjmPTcP4nG++iVkgatqRc+h5xwDCuveZyO55sYmVm0+Wqvg4klJS2ann76w5Gfokp0uACRTQVmeARTfEor1nQ45HLdRCoqdnjxXb7gVWcz8YFAv9Q4bTrj7RdQjk/7ThNYd94J+JEQ0hCwi8D8+YO4hpm/cxbIwhBqUw8qHMKKWni+Hirjv/2wnDy5lOOOm8hJJ03aZntmfn0v1v/2JdLrerjN35crw+sgkk8pR+PnsvRlBKl0G1lPcrNfwaIFOcySXqSQXG8YFDgOk6dXUWTCZ8Iux0Z3XDshupr4h11DxgxjDtoIIMD3iPuaxbKAc2IOK/PmkQCTRvs0tkhS8wWmCRUlQQVixGeeJlScQjKYSSZRaF6MvMDY1BY3BkIA21Za3y/CEs4IuSxcZWLbweY2jFCsbjKGKh2uFxCyr5zy77vgbuoQtHZKJjb4jKzRXDAvw7evjoIjhmI9xF2QdgRuhWYgYUAP5CSBsrsRhBuMgnueRumAkAweV0JCJivwvIBwXHp9iPYeSSIFjRsk8ZQgFoVsFhxPUBANjsmwHeiD3knpZBoQi2iqSzW9DwpC1ZBaD74L0hJELMXsaBKAgSQ89JzFjIlB9SeTg5WZoO2m9baGmKYR6JNsC7I5zUfQFR/GR4D/KfLz34RUR8fWrZUtrjb+vFGki1yEyisLLTlEgFSRDeVhjJzCf3AZZLPoVBbCEpHTOIkE/Y2NlE2cSLKlZbvrNjDYxdmVgo1FjBszhrsa/0BfrBDlukjTQpomkfJyQlv482wPjuNz441voJQmEgn0NmbnGu45+3aad4rSuXgx8bPH4IfGQdbAw0E5DrpAEZVRkAzpj94Jpik57bQpnHbalK0/g20z+qCDAJgCfMbP8LW+EDFPYwvQkTAv//Q6pt99A9lN6zmoMsSUU08hWzQSP/ckJSU2bsbCGsoiD76CETURwh+CmAAgJbnPfJ3w3b8iGjbJpLJINJuSFl0JjepbR6y6+r2X8w7Y4AkyGmJvuxGO1FeSKKvGbtnE+nOOxSsKJnH8SAgRtjB+Oo/Y7c+TKhyFhY+R7sf3Ncm0R306S9I2EMLFMASmGYSJ1tcXMnZsCX/72zFbHa+OhpdzBgklmFIMb148DxnPsCF9OCv/voGp617FUB4Leit4oLmB9VQzuSLDoxtrWP3dw8hkDfz2JMI00NEQSSmQtyxg8llzuDVlU2ko9g7tmH5CJHqx3SxiGxGHRqAp1i5jLM15MYe7Mxb9SlAa1Xx7usPMqMedIZuefkFJkWbinBa0sXWrTyLIiXfOp9pR1FVpaio0PQNBxaK+RhMJe2xql9TXaPbc2eWH5zpD00UfFZSCja2C6++2WdNk4DgQi8KBczyu+5sVTJ8N7sps8L+5QpCGCCqIVv70lQUaQKzJEx0RtKui4UAMDHkteD6rayABS1YbVJZpfCWIpwJxcy7nBiIsLDJZiePlq4EBl94GhgwCVOtrFOGQpqBP0LYp+Fwir6PqXSZY8FYBU6fAq0sM0rkgUb6kSFMCbGwLqksScN7Gu30/IMcAtZWacSOHdT2fBAyTn08pCkaMoKihgd7Vq7dV2hWH0bUFUF+IaEmAHYxv4yr0gQ0IQ2BsyuL1pxFCIiwTPXhdEoJMby8AZnT73upKKRZcdRVv3XsvCwsLQSlKJ0zEicfRShGrrCRUXMzogw9+18+wcmUP3d0ZEISusAAA3i9JREFUCgqCs3FJ+0JGr/orKJ/1nRo/lUA9lIO9R0B10HhXnoeXyZIpSFOoCof0R++ETE8PidZWikaOJFxS8o6vKzc0BSKogAxO+XiFxSw59csY6RS/LuunvL4O31eUlYUZM6aAzvhExiYX4YvgQl9ZGWH0+EomfYh21CC8Q84mPX425h2XEn/uZZa2Czb2KUJigJwVxcvlcBKJrQwg3y9Gm5rIdnbb6MoIk/bbmQfmF+OUFAMCbRsYYQuBJhsrIDaihPL5C0jtMhVlmchiEE1Jap9aRl9FhEzGRUrBiBGFjBhRgGlKYjFrK+Kz3hVcmgjT6wcX6aBQoVFFEULFEb7+pQeZueEVJl13F493V9KXHED5Pi+0hZERA68oiu8GM+S+AnIeRmGI3pSH05vGKo3yUMZm79COEQw1YgJH9PyenzgJBuxCZP6KqYSkQsJ0HYhBDol4HBT26FOCQqkJCWAG7DFj83ofCZewju6tCLpCY+mP3slOCLj43CyX3RCmvTtoF42p0/zhRykmNvx7prvmvym58b4QS1ZJegckRYWa8SMVSsGVt9p09cntCoEp1RjR4JhwB8lCnpwURDXFBRrXh74BAlNAJXDcIKOrMKYxZODePKomWHhPv0AIjcDH8xXRUA7HlbieBm0MdRKlCIiTEAF5cT0oiGhG1ytKCjW15YpHXrRACyy5ebuUEtzxSDVfPAGKC/Q2Z50JoxTLGiWeL4L2ogq2yzDyJKkwEGJ//6z/fK7aMLaPYfLzKYVdUMCMs85i/hVXBO2vLfGvtXD8JPSZM9CPrEM0x6EohJ5dA7NrMRoTSExELITIeOjyKHSlQAdnISElvusycr/9tlmv1pqHP/95Wl5+GaU1uY4OhGFgxWKUjB2L8jyi5eXM/NKX3vPCHI2aQ6Z3drqL0av+iuHlQAhUNge+jxjIov/2FlyyN4NnI618hBKMdcchEDgaXs8ZuMDuIZ+wCNpEr115JW2vvYabTGIXFTFq7lzaxxzKN77xJOm0x+TJpdx33/GUlAQTb+fGHJ7M5S9KShFpfAuZTlHW087zf7+OEXvuyZzvfIdzzpnJL3/5HObsQ8luilLcs4yx9SEm7TaJXc47j1DxB/fR2R7U2JlsOvZSXnzyfCYVrmJWjYcpBVmVZtmGhcSbm6mYOvUDL7fS0MyxPV7aQvPjaJgX9jnnot1pWTWLJiuENCUiH7gphI+SBlZfP6WvLabupRdIThiLm9GM64zwnQv35vXX2/jHQ2soLAlh52v7ruuz++5b63yuSoboyQhaO4K78rAdpFkfFHO5I23jA8sqduafySZG1Sqs+DLctIerJE5SoJM5dCiKFGIox0ALAS0DZDsT2GXR7TpFv2+EY9h7HM2186/m23POpzNcikBTZcCtFVl0cvNLDRG4Lb8T5mb3o8nciCfcoZaX0LBrbrcPsYHvjJG1mj/+OMOGlkBAPrZe/dsutv0J+N1tITSCVDbwzIknBRtaJJFw0GaCLTrDW+4mAa4btLGG2qRZEElNNBJMQ9VWauZM97jvCQvbEuQcMC2YMMpn4mjNxAbNmuagqha2wTazpH0LEJimwjI8Ml4Yw1BoLTEGyXJ+WxwXRtf5jBupqK/WTBvvc/qRLv843ELlrQlkvgIlJfQlLMBj9jRFVZmmPymG9GVFBZrZ04N4CnTQ6qoqV9RVafricNFZOfaZ5XPTAxZ/e8TG92Hf2R4XnpGjaMe7ncP4EBgmP59izDz3XMonT+b13/8+ICOeh/J9zM4s/l0rUcdPgJMnByO5vS7iwqfRhW9h3XIi4VghkVP2In7jMwjbxIrF8DIJ8H2U75Pp6qL1lVeonDaNqpkzh9bZ/OKLbHz6abSfb3DbNkY4jHIcpn3+81TPnElRQwPybaX+7WHcuFLq6grp6kpT1fI8UnkIFFJKDCECh5rWFKIvi17WBWOKIWahsj675KYxJTKVFa7k1/EQfUrgOTmimRzfKPWpevqftLz0EoZtByRMa57+f3dzx8YVdDEOIWDhwk4mTbqRtWvPpaAgzJyw4qCQy5MpsJrW42WyRNIpDrvrWkTEYNOLL1K7227sud9+/OxnO7NypcBxJnLw4eNYVVhAWkuM8Efr01FUX8841czoYoWPgQZCEnatSNH32gs7RH4ALix0mGQqXspPex0cdjkwFJTmLx1n8VSHpD9f8xCAoSEacqlf9Bo7NT5KxE3CMkEuXIA79yRSOYf7C8poOm43Mk+uJBx3mFZuMW+vOs45Z/Px09n6OhP85WS6x7ImdTAeNpks9GUMcqM1DVbgP9Xf1IvZ04uZbqJUZnBEil5pknEiRF9eQe6Q3RBIhAZlSIzeFKHeBJG6YhwN060PJ+r19jyWvSbuxvxX72F1pAJjxgGMKStDCPWBnLFjxDgxdTJPRZ4gKZJYWOyWm8Nkb8p7v3kHIUQwjfTvxj+ft8i5AtvaPC0lBcRTAjfvnzNEeLbQggk0OgMU5JNZBIHWuxdMKVBSE08KRtUqvnumw9c+6/DiQpNoRLPHTH8ov2vxCsnP/mhgW1BdrujsyeErjZSKTNbC9U0KImkQEqXDuF5QaQzG2jUF0UA/9LmjXD6zf/C7fXHR4A1WsF1e3ixRSujpt3jiFZi3p8fPvpblyltCNLdLDEMza4ri7BNyfO2yCJlcsE8GtT3V5Zrdpvt86dIwDz9rIWUwtdbUZrN0lcG9V6V5H6fLYXzEGCY/n2IIIRg5dy4j587l6e98h9V//ztuMolWCnnrUvjHavS+9dCRQr/aijBMKsqmUPVqIfqsGeTGT0K6mswNLyM9A6OkBCkEldOnIwyDVEcHr/zyl8y79lqilZUo3+f5H/wA5boIGVwYfScQTwopcRMJSsaO/UDbf+mle/PbHz5IuH8tlhNHoBC+wM83/UXCQbcloTcDOR/RmqT0oW78fWrQ5+/B7xI2GR+ya1eTi8dJeD4/SiU45ac/JWyZxGprCecrMes3JhnPajaIsYEZvoBczueMM/7FvfceD8B3nriZ2lvvZsXUOZS0b2TKMw9huQ4ttk2uuo4V9/6Lslnz2LUgwiEnj+WRjMHXszZeMji535a2+ULM4cjIR0OCwqWljCxW+C6brzBaY9oWsSWPAeft0HKFgKOjHkdvRxhcY8KPinNcl7DYlPBw0i5uxqX4qZfYN/E8GTxcM6iWhaRDQ/Lv3FdZxSZ1FIYPsXEVqLRHdnIZn58YXBzcVIrWR0+ipG4V+9aMY79yzakVN3JB6o90qhqMrKa5z2BadXDRLi6W6IH1SKnwDJNQUSGjZZrkQJhxq9vYWLGRnjFVGJZJpGsA48mVlOw9BlUQZpKpOD22Y2GmW0KX18ER533oDLdyXc7J6VPf+4WfMuSczZN1JUWa9q6g+qM1JFNBpca2grytQSompcYQ4HYKCAMhgvH3LsAnaL+HobRY47iC5xcZHDjH5/C52x6nu0xRnHSIw79eCLQ9UxrayDqK8oI4i9ZMYGxJO5UlfSxZNxnHCyo9QgReQAWRIO8r5wj+8oDNfrt5FEbhjodtJjb4LG00hywbNMHnrCx1uPzPIeoqfaaO1/zmwiw5J++jlL+SHrV/UKkaJHquB0ft53LB5SHueyKoSkEwtRayNcvWGLy02GDu7P/MBN4wNmOY/PyXYN9f/IK2BQtItbeTi8fRvo/ozSD+vnqo7jzozZJ64DUq1mQ49NJL4bPHkT28j84332TpzTfjZTZn/wghUJ7HyrvvZtb559P22mtBxIRhDDrYgRD4roslBHX77POBtrlv7VoWXnEFu7W10jKwEs1mBzI9aPcqBPLlFsQrrWAYFNbVUTZxIunublp8QYcvcVqbyQ0MgBB4mTTZSAEbq0bSsH4F6VQfxvhKfENTWu+R2pSmWrcxIErI5IeYX389mEqLNzez4KqrqO/poWpRUEnTvo+WkpRh0V1SSX/GY2Hc48+hCVT12qxzJa4WxKRmkhXcJt6Wstk35FH0EU11RMvLUQM6IJpCYFgW0jQoKP1o2mvbwxcLXFbcvYRbWzwspQg9sxK9ejX9Th8V0w0MG0xbUFQUIt2doDjTRH15My2dIwMr/5hFR2+OBzJFnF3g8NatP2XcLitQysRIu2QKCijS3Xy7+FdcmLgavwDSWwwh7dv9Io1lA6zoKcI3NTktyYUjHDqqD3fOXpT2DrBucQp75gimORn2/vrOmLuPZqqVZYb172v1DGMzjtrP41/PB23iuiqN4yj64kHaejYHkTD5CSfwvCB2xDTBzY/f07T18ganubSG9i5JLOKzvDEgP++Ezx7pcdxBHq1dkuqi9VSFrmTJmolccO1XKIykQcDUcZLGZk0qnTcYjGwmKxpIZwMR8967+PTGJZVlMH28x7I1xlaeVa3dIYw+gzN+FOOiM7PsN9unpEgPLQvgc0e57DzJ56HnLN5aEwyaXHlLiJZOY2juXetgUk3nBKahWfiWHCY/HwOGyc9/CaxwmN2+/W2W3Xor2f5+eleuxMtmh2qvg66/6c5OCuvq6F62jJd/+UsSTU0gJVUzZiAtCxWP4+VymKEQ0rIQhkG2vz94b1cXkcpKUp2dOInEZn8hrameNYvSD1D10Uox/9e/JjcwEJA1GKqLCynRWiMMg4Z584hv3IgAYjU1GLaNn8tRMXXqkDg5NzCAkBIvm0UrjdQa2/dwhYOQJn5TF2JMKbFKkLuPR96nKdV95EQIH4OiIotvfetJ7Nfvo7I1QVjn0ID2AjtWpYPQVMNzadnrYAY6u+muqCfjSvz8NmQQbPQkY63Ab2Z+zuCQyI6d0DxP0dKSoKQkTHFxiPCsfXFeymeciUCxaVqS8Gnf3KHlvx+kUi7zb11EuCkRhEem3CCiwYPly8roLmpA+IpKs5cZRSvwtEFVWRedXdUcVf5PpsdWEKccX58NlBGLvYLSQW2/tKMV3xiBG44yTq/C8DXhXpBVAtfXOM19tL7Zy+dmt/HYGsXKjhhSaKaPSHLETn2cd84hxAxJQz44tEODHXU4JTYcDfCfRFWZ5nOfcbjnMYuBhKCuWjN7uuLkQ3P86NoI3X2Bo7QUgdi3pEjT0SMYcMWQ8PjtGCQSQkB7t2TmpPf+DUXC5Ceo9sHxUpSWvIBt+vg6gpD1hEI208YpHDefKp8vV2mCEfiQHYiYwzbEwppUVlBVDkVtMJAMXJ99BZmsQSQMfQOC718VZlRtoBU6cj+XUw7bfOxNn6BYvdHnlTeNoA2YDLLG9OBK8/AVOA7sO2uY+Hwc+MjJTzabZe3atYwePXooMHMQ9957LyeeeOJHvcph5DHx2GOJlJez8Npr6VyyZPOsptaoPKlwkkm6li4l299PfNMmpGGQ6elh3SOP4DsOUkqkaSJMk0hpKdGaGkbuuy8A9XvvzfLbbqNy2jT6160j09eHlJLChgaOuuOOD7StfWvWkO7owIxE0L6PlBItZfD/loUwTQzLYspJJ9H44IM4ySSGZeHnchSOHMn4o4/GNDSjDcWi/CiH9oPPW9TdTm28Gbc8CgkXnXFBCvy9GggfMBb5UBzlaKI6SVIUM2JEIW1tKepTSXQ2geumkIPp41qjBbjRAtYddiLJ+tH02bFAO4Bg0DlEAMnBkTkNhTtY9XnssXXceecK+vuzhEIGM2dWc9E3LqMwO0Bu5ULwPUS0gNARZ6DH70ybL2jxBOMsTSkpbO86pNoAIoZrHolvbCtaf1/fT1+W/v4chiGIx1201iRFEWsYzwavAaPXwzIUHbqQlmQJB9z2PInz5vLLsT9mVLgJR4UIhdYww1yI8H9GLmkjpUb5AstUFLe3oIHuaAOFmUAgOs7K0v7zp+hY18sqx2R5zwQOnNDHJYetB0CqLGtqxtGtHTIqRFleJ2ELeChr8ZmIxxJX8kTWwgSOi7pMtobHiv+dOPZAj3l7eixdbVBWrKmrVlx8dYh1TRLLCsbUlQ6mnuqqNA0jFE+8LIP8LW/ryXPTgEgoP1mnoLhQs/cuH4wYdKUO47f3HsOaVpNkRlAY1YwbFcRIHLKXx4JlBp09QXsubAekqapMs+tUhZRwyN4udz9m43mBn8+geaHSwdh9MiXI5oIw03RG0NIJ61sky9dIevolmzoFo6o1Ty8wyDmBsWI6s/1IDQDLghmTho/RjwMfKflZsGABp556Klprstks3/3ud7nggguGnr/ggguGyc+/EVprOhYtIt3Zua2lhVKBCZ/v42UyKM8j0dKCUAojHMZ3nKGWlzAMhBA4qRSlkyZRnyc/kfJyxh9zDI0PPED55MmkEgmKKirY8+KLMT6gYk9rPWTgF6uupm/NmmC6yzAwYzGklFgFBYw78khGz5vHqvvuI9nSQsW0aYw78sihpPcfFWf5bliyKgPasihqa+KgW/8PbQmscVUIJXFNF/PC/SiLWvjJDMWFGdI9grKQzT4HjsHPJ7rauX6E1ijDRih3SGLjRqI8een1yKISUAo7EiYFWCKIvYir/FQIQTGsQmp2ex85aW9HS0uCG298E8OQhELBT3PBgnZuuDXMVy/6C5HmlTR2tnF3zRw6oxVs7JIoEeSyzZKv8ZPQF4nKPjwMNMWE1GpyZh9vbjienoHAQr/wfU6WVFVFqaqKsmZNX1CFy5+9X2c25XQDAs8PCOtabyzhJwR77/0wI+c2k1NhDEMwImJSiI/ybyLjHoSX+wvS1NiGIDDsVbxkH0e5rSloa2LVuc8Q6xygoSpKSVmEkKjj0ZWSnSq6GVPeS09xKTcdeAZJLXH9LGmpiRpBCNSAkvymOU33rX8m1rIRL1bIz4/7HCfNmvShDA+H8e5o7xas3igZN1LhuXDM12MsXW3gepDOBcLi4gJIpAWTRvv88Es59visQUePAFPkp6mC6svkMYquvsAlOxzS/OpbmQ8kBNYafvKHMG1dggmjFRtbBP1JyZsrJfXVmrIiwYkHu2xsl7zyhkE6K2lul0yf4JLJBe2wUw/3KIrB938XRm/jCB3c6rheoPPJeYJcXxAYu2JdKCBvYXhlcSD0tixNyApev73U92hIM32Cz9JGya5ThwnQfxofKfn54Q9/yGWXXcZpp53G6tWr+fKXv8yaNWu45pprgjv77Ro+DOOjQveKFTQ9+yxOPE6ouBgnHke57uYxC60xbRulFIZl4aXTaKVw0+mtlqOVQguBVIquJUtIbNo0FHMx/XOfY9T++7P+scfoS6fZ64wzdshrpmzCBGLV1eTicaRpUjp+PL2rVwdOxr6PHY0y97LLMMNhzHCYGe+QwVYi4YbdRvLcddfTsnAh6VdeQvk+kbpqUoA0JeHzdsePBId6tfC5cHYXKMWhf/wp3//ZYtrbU6A1lpsiFykjlOneKqAq5GSR/X240QJiZWWMLIjQm4NKQ1EpNes8yYASlBiKEYbiO4W5IeffD4L771+9zWO2LVm8uB2E4JHKnbghuisG0OwK2n1JgdDMtvo417oIQT9xZQbaCd3PuoEyLr9xFM09YXwFhVE45TCHYw96bzJg2wbnnDOTH//4BbQOxMNaa7Rh0aWqMPCxpIcwJFoLmuJjuDr8PDEnhBGF+kJJWAQjM0J3sfN5P2XJ9a2MmPAskaI0Wkd5NXoEr9ZdgHn3C6x7eA26qY+Ur0mu6yedcGhoqCERqeB3cgI7Hx1mUd0sVqkGHCwEsDybJdLTTjg5QK6kjIZffZuC5ADCMIjoFsb/5mIe++aPOOKA6djDGqCPFFrDb2+2eXWJSc4F24S2LkFrl0QRtK+0Bt8XeL5i7EjFjEk+kTDc/Is0V94aoqVDDLWsSos0b601ME1NUQEcvIfHPrt8MELQ1C5obhPk74sYO1KzrlnRPSApKVIk0oL5S016BwT1NZpEUtHdL7j9IZt7HzfZdbrisL085u7qkcqA/y6XK39zUZ3gRiB4PJnWQ6G+Xi7wJno78RkUXc+Z4eOrYN8N4z+Pj3S3r1y5ktNOOw2AiRMn8s9//pPPf/7zfOELX+Cmm276KFc1jO1g4xNPIE0TL5fDTQbjR9K2g7v2fMK78n205+Fns9uPId4CGsglEtvEXBTV1zPz7LNpbGzcIeIDwXTY7hddxKtXXEG6o4NwWRnjjzqK8mnTKKipGQrtXH7HHbQtWABaU7Prrkw9/fRtxuiFEOz/ta/iJBIkWltpefll+teupaciTfaMSVhVxfSuXo2XyRD9ezOGZTH19NOJlJczenQRmzYlME2JkiZOqJRwqgMhDaQRTLRJw+CEqy7ijZ9cg7nHHCJCMSnZRW+sij4lmGz67GL7fCnmUPwhRlZ9XyGEIKkFXUpgA9VGYBjXu3YtN7RLUnaEWE0tcTOGKSClBTPkQ0RFH37eSM8MdjDX/+0MmvtihEMKCHxfbn/YZu9dfCrL3vtG5OSTpzB+fClf/vJjtLUlKSwM4bo+vd0pFAZSDtrsCGzh0edNZlb1RhBb7wRNEWY4zKwLriPd3U1uoJua+jEYrmJ0/+OseXY1U6PdrBOSlF+Gpwto63SoHaGotiQDIytYXj+OjaoWpSFMDksFDsy9kUJK+3qZcdOVGL3dJEIRpNJIIbERxO67g865v6beHL7x+ijxrxdMXlpsErIhmv+6N7RIpJFP1MmPthtG0CIqLYJdJgcsYJcpipsvy7ChVRIOaeqqgu+mu0/Q0SMYWat2yPsmnRG4HkPkx/OhPxHYIfh+UAvXGlatz0+qegIvT0ziKYPnF0iWrDT41wsGAwn5jkntg3j7c0oH8T/K31x1V1sRH41pCApjgZ1DIiUYU6eYNn646vNx4CMlP0VFRbS2tjJixAgAIpEId955J+eeey4nnHACSg1/yf9ORCoqyMXjuKkUelDvA/iehxEKIQ0DP5sNnstj0MdlG2gdCH6VIr5xIwMbNhAuLf3IDPwAyiZO5LAbbqBv7VqElJSMHbuVE/Crv/0tLS+/jGEFEyWrH3iARGsre37/+9tdnl1YSPmkSZRPmpT/bJoO2cF6sZaGCWMYsb4Q80SH8smTMcPBqPZZZ83krbd66epKkygZR1nLK0jUEPERQmDFYggnx2fnP8RBJx8AQGN3O2NLC+lSgiKhiX4Ek11HHTWe255uod+QjO1awOjuxdhumvpwjruO6qH10pvAkCTbO/Cn7QKRGFKAoRP06lKKRReD3rNKCda3jMYRJuEt3YU1PPKiyReOfn+j4LNm1fDSS5/j2msXsmhRB62tSfr7MljaGdReEzJ9CiKScMP5aPErhE5uHtsRCsc8YWh50YoKovn4j8+KH7H7wAZWpsZjhw369W5ksr0IaZDRNovfksyaqrnl5F6W8gZdj6SY2r4Cy8+yrGoqf5n9RfyEgvZuyjeuwTOt/LcewBECO95PqRwmPh81XlpkbhOTIY1grD1kazJZsdWJZdYUnwkNm887hsE2MQ8VpZqyYs3qDZJciaKybOvlKxUQGvsdzLHHj1KUl2iyTrDi/OkL24ZoODgGGjfKd8z4ctxgPP+5162AwMmtCsDvisHT1pCweXCbt/iHaQgQmmwOwqHAm+hH52WHs74+Jnyk5Gf//ffnjjvu4Lvf/e7QY5ZlcdNNN/GNb3yDF1988aNc3TDehglHH82CK69ESIkRiQy1tRACrRTCMII2GOCZFp4dClx7cxkM19luQlYulaLjjTd48lvfwgyHqZk9m90uuOB9mRgCpDo7WXTttcSbm7EiEUYdcACTTjxxiOQIKSmbsK2TSravj/bXXhsiPhDkcXUsXEimp4dIefm7rrctlePh7iylRZUcXlxDTAL1+b8tUFwc4pprDuaRR9bSuKqG8hf7ySzuQuXF30YkQkaHeaF/Cn9/UHDTois47Uv7MGm3SgwBNe/i7vtBYTaUYR01nbG3/ZVpzU+hpYkVglDHSlwpKOxoJlEzEj+bwWpvw20YB1KwQB/NXP1PErqIEhEHJEL4IKHPr6doi3UotVlU+n5hWQYXXDAH31f4vub88x/k5efbcFNpIpZPRYnF+F2nsfPsCWT1b7G9PyH0JhAxHPMElLF9N2Oh11NdpSiI+bzRMQ+ncDLSXYLO9qC1TypXQOX4qdQ3lDHxz/uxZFM5nmGC53LsG3/n2AUPsKFwFLJ/gHuLp9HdlMaJFQbXXD+IQPDLKoLvfhikMvDyGwYhG/aY4b8jiXg/CCo6Wwt5q8s1m9oFISuIoQgcnjWH7u1y8bm591zmzX83ufLWMMl0MAI+sUFx95VpQhb8v3tsXl0SiIhrKxXnn5Zj/Kitj2PLhHNPyvG720Jk84nstqUZNUIHubQKunpF/qZmO5WbQWNDH6IRTTq97W3hlu8TIhBpa70FydHbX/bg8otiwfN7zvS4/deZYeLzMeIjJT9XXnklnretnkBKybXXXsv3vve9j3J1w3gb7MJCaubMoempp3ASCYQQGKEQvuehlcKKxfBdl+76sUjXwYkWoqWktHktGoHl5ratArluUCLO5Ui1tRHftIlQaSk7n332e26Pl83y3Pe+h5NMBqPo6TRv3Xkn2veZcuq7m76lOztx02mst00Muuk0qY6ObchPsqODxvvvx02leHrEZB6pmoinA2ORvxQW8Mudapj5Dhf9cNjkuOOCalG2bwr/OusselaswHMclDC5v3dPskSIiijpTUmu/8GtzBvTwU47N1DY0MDMs876SCpiz+RM6o6fwfj512DYVQjboGBTYzCt5yv2uOP3PHrRlUjfJ9a6ETlyJNqwGdB1vKxOQXA3WrYQJU5KF9PVMI1wS5hQWQ9ObzHaM4mENIfts2MCYMMIpnS+/e1pzJzZwKJF7TiOYty4Ui64YDZCCLQoJ2e/9+9ca3jNncoTxjT8L9p0X12PkRZQPiNwKTclsfJRrO2Fln/eSvETXURjSfprRjK2eyMlyT6E1kjPJduX4TO1IW6J1GA4OZRlg+chpKDi82ft0Gf9b8Nzrxv88S6b5g5J34AgHNJ86/Qcnzva2yFPpCPmOixrDA+1mAAaahV11ZrGDSaeD2XFigPmePz2wveubmxsgV/cEIjXgxR3wfK1BmdeEuHAOT5PvGIOkbWWTsmlf4jw869nuOcxm64+QU2FYq+dPf72LxulBJmsYEy9z5nHOPz1XzZKQWePwFcCy9KbvYa2gNJBtagoGhCUkK3Rjtiq+mPI4PFISBOLalJpiecHouZBU8OiAk1/Ipj02pIoaTQyP9H2rS/khonPx4wPTH7uvPPOIV3P22HbNrZtb/c5gJF50eww/n2o33tvmp5+GruwMLgYASqRGHJllobBy2d/DyOTYtpj91C6aX1gDhiNkTWLKerp3GaZme5ucr29gWBdKV76yU8oGzcO6uq2ea3Wmo433qDzjTdId3fTvWLFkOmiGQ5TNGoUG5566j3JT2F9PXZxcdB62wKhoqIh8TUEBOvV3/6W1fffjxWJkCsp5+9f2AsjlcSKRMB1SHT3cMVqi9unl73niT5cWsq000/nrTvvJNXRwfymGGlRQKw4hhCCMqeFienX6VoByWqbntWraV+wgMP//OdgfR8CdabC12C6WYziMN5ABtXei/YVCkXZquUc+9PzeO30r+OFo5wcmslR5SHiWlBjnMDLmXm8kFvIyuWa529X7Hbk/ZQcZGHHTMjFyC2bxmcnjX/fE1/vBNOUfO1ru36oZVydtHk+cx4h3YGaLeFbo1H3pDBXONjREsKxIrQWmM0v8fJF32HeqCRWf5qa7gFKIw5IiVCawuQABYamL9FHw1778oieTM3KZeSixaw65DRunzGOzf7C/5vI5uDGe23WtUi6eiVSBKGgl90YJuM4fOkk570X8jbsOVNx4qEuj75okkwLYhHNoXt7nHmsSzINK9dLaio0I2veX5XxV38K4/lBJQWCKonrCV5bYmKZW7e6hICefjjjkiiVpRrXg0dfNLniphClRVBTqags1axYY9DUJpk6VmEYGq0NOno0voLegbfrcYLlOi6cfKhL1oFnF5hBmKoViJIt08OyTKrKFFd8J8OmdsmyNQa9A4JxIxX77OJx+U1hMjnB0tWCnBO4VGvAMjW2BZPH+JQXa2ZO/N8+Jj8J+MDk56tf/SrLli3jsssu20qfMYxPBkbttx/zf/UrlOfhuy5eJoOOWuiTJ5JuKMaY344XskmWV/HKGd9h4jMPMWrRCziFJYQSA7Ad8qNdFx8QpomQEuW6PHvRRez6hz9s9Trlebzwk5/Qs3w5wjACP6GBgSEi5qRS9K5eHXj7bDE+vSXiCuJKUBONMfaww1h9//3IfOtLOQ4TjjlmSGTtJJM89e1v0/zss2ityQ0M8Fb1WBw7jJVOoz0fMxJGmiat8SQ9qvxdQygHMenEExl1wAFsfPppXvz9YmKt5tBdWn1uVUAWfUXnm28GrstK8feTTuKIv/yFWGXlB/zGNmP/kM/fUppcTT32G4uRKxqRvo9GApqcNom2NjPv2h9TOW0aJ3/zFADK8iqDA6NF1LXszONXPcfc4/spnZJCuwKVEjSMVpTv/DrjMqWgS3d4Gz8s2jw4vSfKcs9AMJoSUcHO5ipqKxK0HF5BQQeI/GlJKc3uTT+m3zPJKANTQqnMIZRCokBpbAm+EIzIdlHW1kzrAWeyZvcTMHMuY1e3MM0evsgsXmnQ3S/o7Q8MO5UKprE8X/DIiyanHeHsECE+/UiXE+e5dPcLyos14VDweGEMdpv+wfZ7Mr05JDTrMBSKmhPw2lKTcSMVRQWadCYwP+zqExTFFPGkpKVTBPoeLRhIgutJmtsDAbTWmq5eSW2l4vB9gymulesDhjWUNSY0liGwbU19teL752R59CULpWF9s0QjgqDWbIaxIyNc9vVsfmBAcTJb35zNnuYHrcVQQNjSGY3SEIsGQnDbglMOd7fRSw3jP48PTH6+/OUv84c//IEVK1Zw0003UVJS8o6v3bRpE3/84x+57LLLPsw2DuMDwC4spGLaNNLd3fStWoWuK8T/zf7oohDCB3XgGGbpdby4ejZSCBr3O5KK9SuRvk90oOedFyxEUIUxDIRlBfESjz7K1F13ZfUDD9D87LO0vPoqmc5OorW1FNbWBqPReV8hKxoN4jJcFz+b3Yb4eBr+LxFisWuQU1AqNZ876Uz2mj6dtQ8/DMC4I46gZtfNFYfld9xBqrV1SNckhCDU2Q6+H4zZei466WPGooR9RURomj3BMzmTWkOxX8h/xxHoSHk5k086iYOaDZZc8wZ2nvzYOofWYCh3c9SHYZBqbeXVX/+aA//v/3boexssj/+qKMu1Z38V/zNHUJTpRKIQ+cqFgYfnSSLhMEf85S/bXc5996wkJKFgQgadL+1LoentTlJZUcZqaxW7O3vs0DZ+WCgFn+kuoN0XKAQC6NGFvO7NYpfRDgOegVsrEL0QUYpDp22gds1SlJS82R1mdmUGQXAxkXnzSSkEUkMxLgNPN3LIyz9mza5nI/rSHDS3/j236X8BBRFNPBlMPg1CiEAEnMlCW5ekMPbuZCXd3Y0ZjW7zeMhmaFprR/HKm5J1LZK+uGAwOWcwIywWDtpMG1oE1RWwqT0YMcxkIZ2RedH9YCROQJyC/4p8FUkQT0EybRANa7QKnrfMIFXeNDWTxygKosFxderhLj/5Q4RN7YLegcCZubZS89kjHMKyhaMOrnvXdtVFZ+W47wmTmx+02dAiqavWFMU0rV2CnSYofvHN7IfeX8P4aPCByc+vfvUrZsyYwbe//W0OPPBA/vrXvzJ58uStXrNo0SKuvfZaHnroIYQQw+TnP4hwSQmlEyaQbGnB9zz8r+6MDhuQC+5QtO9RX7iRkSUj2JAagwm8ftKXmHPnH4j0b01+RN5rXnve5igL38dXCuV5NN1/P092dTGwfj19jY1k+/rwXZdsPM7AunVBmrxlDb1f62ASx0kmefo73yFWW8tOZ5xBtKKCG5I287OCouY1WOEIqZo6/pC0+f30XTBa2nntmRd586/3s097JzsdcRhCCAbWrcMIh4fOlL7jMGLlYgq62siUlCMJJk5yrseutuKWlM2TuXxVQcOdKc1lxVlGmMF7N730Em4ySf2++xLOk/pjzz+KB+5+kw0dGtsk0P6oHGWhDIigIqWVwi4qIt7URLavj3DpB6us/DNj8mDGIq4EpVKx04Ln6ZcWGBbSywAaJUx8YWNITfnkyRTU1m53WamUi5TbMrpBI0f9MbaA/pEx6VZBXSdLMAwkgLiWNCobY7RgzI99EjJObWEn+8in6F4cI9qaoC/h8kzKZPeRUBN2KYiIwEspr7rNKZPacI61WYXVm2T69ErOO2/nj+2zfpIwdZxiIBlcsQfvObQOWjzFBVBd8c7HxMann+bZ732PbH8/Ukoio0fT8MAD2NshQtuD58GqDZJwCMbWb5u5tmC54IsXx/D8oDWUczdHT4RszYxJPkLAynUGzW35CTIFsagmkQr0NoOHAQTEKeuIofyskKWDyS1gWaPB2JGKKWN9unoFPQPBEdg9ICgv0UwaHThBv7UmMD90vUB43dOveeQlk4u/kH5PnY6UcNKhHicd6rFgmeSRF4KQ0yP2ddl12nAV8pOEHRI8n3baaUyePJnPfe5zHHLIIVx//fUceeSRPPTQQ1x33XW89tprFBQUcN555/HlL3/5o97mYbwH9rr4Yu597TW070Nd4dAowmC2jBhIsvfK26h+tpjeUeMYveA5ale+EVyMTDOopGiNYduUTZ5M5+LFW484EJxwzKIi1j/yCOHS0qH2z2Aj3XcDh2RfSsLFxYSKisAwyHZ3IwyDxKZNDGzcSPvrr7PbBRewshum3nkTVm8XWkpy9aPZ8J2fc9WfbyXy/ONgh9DA/StXsq5xHcd+63xCRUUIIQiXlJDp6UF5HoYQHHrlRbx47g/oHzUew/UZs3El5xx2Bj9KmEMVHAQktOCaZIiLepbzxNe+xsCGDUPC8DkXXsi0z34WK2Tz58cu5E+X3MriJb1E6mYxKf0amfb+YJ8qhRkOE93BdtfLOcmfkzaWDLapO+fx15pp7D11OsXzuzC9DAgDgUYYEtOWQdstmx0a198S06dXsnx5N5n1YYp2SaH94EIQi1koFBO8STu0nR8FVnl5/YMAQ+uh6o8P9PiCelNhRHoJiyzdupDnQruy604VOCJH0YYkyoc3NvnsOUpglRfjZZJINL62iFsF7DxGMOoL32DMnrOorHx/F+f/BbR0CkZVKxJJQSY/BWUaQfVj3Cifohj8/SmT2x6y6B2Q7DzZ4+dfy2G6/Tx+/vnksi45R6KUpv+N1dz6mS9wzlP3vud6X10iueq2EPGEwLSguizIwaou1+w6NZg2+8k1ETwVCImLCiDnaBIpQTSk2HNnNUQ2Jo/x6OkP/lFeEkySLV4ZPPd2E8HBU9VgBUgaQZUrlQ0qS9FwIEoeqzV9cUil4YdfyjJ7muIHV4dY0yTx/OA857gCwxC8vtTE9T6YzGO36Yrdpr/3lNswPh7s8LTXLrvswrPPPjtkYlhTU0NbWxsjRozgZz/7GV/84hcp3EEDvGF8OFixGBOOPJKBtWvJZDx0ob317KUlka0JJj/3wjbv1YOjDfnXp1pbMaNRvHR66DFhGBjhMALwHIdsX1+gBRICLSRoNTQ1pl2XTG8vbn7s3i4uJlRcjNaaRFMT6Z4eHv3SlxiZyuIWl5AdNRaEwG7fxMjf/gjd2QahQEwgABUKsfiFZ6n90nRqzzyKjm+9SVFDA9I0Sba1oVyXmojN2S/eh+v5WOEQJaNG8bxrD7kux95cQN31VxBq2YBbNYJH4x0MbNyIyJ9pnXicF3/8YzoWL8bPZolWVPDFi87i63mhdf+6ddxzwgkYvo9dWEgsX4UprK//wFWff6QMcp0dJOMDGLaNVVCIcB2ajzqKokWvorIB8UEHF4LyMaMwbBs3ldou+TnhhIksXNjGmnsU40o9wqOyRIolo+qLmexMpFy9u0XAvxPHR31uTAVarwIgozU+IIFReXfsLnIIBKZQtGSLOPRrs+j5s2CUn2SMmSUSNigusOgvKMePlODE00hfELMlVQeezPij9/nYPt8nFZYJkQjsMdOntVPQlwj0NbEIfOEzLtfdaXPNX0NDsQ3rW21eedPksom/wUllSDubBSoak+633uLF1zz2mfPOl4/bHjL55Q1hHC9oP0XDipXrDJ5fZDBptKK0UPO9s3P0DIitLDZCNuQcAoNWGYi1sw7sPl2xauPm9lZbV0Di9PsopqQzgYYoElIsazQoigXanu5+iWXCpDH+kEZpTVNQ8dmywuP7QTxHS5fN1CkfZM8P45OMHSY/7e3t3HDDDaxatQqlFK2trYwfP56nn356mPR8AlA/dy4r7r2X7D/XwjkzAlHNIDIe4q4VW79ByuBPBS7DQkqElLiZDEIIiseMId3TE4wXW1bgJRQOYxgG0rJIZR3S4RjC8jByWTzLJpRJIWEocR2lyHgera++SrS6mlyeNHnZLKaTg852VDiCUz0CpMRatwpTeVAcEAqNxsVBplIsb19B8/RSyv/f6XD5K4RLSqjfay+SbW246TSGbWNqjVaKqaedRq+hcYHqp/7J6J9/GyMZB2kQbmmiIzGAHYtBKExGg6EUIhFnwxNPUD5pEqn2dn5484O8+PlvkrKjFEd34rirb2Xnm68i3txM/9q1mJEI4446CuV5SPP9/ayU57F+/mKS0h5Ksqe7GyHAD0eYMLOBtkV9+L5GSqjfbResUIhQcTGh0lI63niDeFMTNbNnU5g3FrVtgyuuOICXX97E4sUdjI9GmTmplAqnHJuPV2U5wQp0Vk/nTCQQyXcxZphekLQtAh3PIEwviwybTDy6gd1TzfjJaF575lCU7qW5cgLxugomCYeC4lIyp1/wcX20TzRqKzUjqxWtXYEGpa46cDouKdSMGqG48xELzw+ID4BlQGev5M4ls9nJ27bPI7XHg0/BPnO2v76X35D85YEQOVcEXks+tHRIImGwTIEpIZMT/O62EHXVmq4+MLb43iNhTVGBYuX6IDw0FoH1LZrSQk1nn6C5XdIbF4Rsje8Hhp7vpaJRCnaaoNjYJunqk3T0SgqiwfsNQ/DmKsnMSYqSQrWFEHrzPaNtawqjWzseDiThX89bZLJw2D4eI4a1PJ8qfGDys3TpUq677joeeOABXNfl0EMP5Wtf+xorV67kBz/4AUcddRR//etfqdvOGPQw/nOo3mUXRs2di/PQQwyoN9CfGQ8FFqI1CVcvQMRziDxxUflUdeV5wQixEJjhMNKyAp8dISgdNw4/lyPT04OXzWKEQjj9/RSNHo2ybLpXr0EJiTBMEiMaSJVX01U/nt0euiWoJg0qGT0PL5NhYN06rGgUIxTCS6eRuRwmArFmFUZvN8I0MV2XnopadCKBNExU1EAL8GMxonVFGJj01GU54IoLKVeBa7CXyfDWnXfSu2oVZjTKlFNOoXzyZI5QLg+mTSrvvgkjmQjsaAnGttGKbDaNHwKPEEYmi4Eg5wQ6qZcmz+G5OYei21rRo8fTqyQ3jpnDLed9k9TPLqawrg4rFmP1Aw/QvXw51bNm0b5gAVr5qHQK1w+E36HCQoxQCOW6FI4aRfGYMVSs66dl570xfX/ImsAxTEatX04oEqJ0ZC3p9nai1dUYhoHWmimnnsrTF1zAwMaNACy79VZG7bcfs772NYQQSCnYZ5+R7LPPJ89a4s/lGR5NG9yYtokCPyzKgYAfDoQxkRjawBc+npJMj7ZCFsa8sY5IVuNIyGlwrBCtFeOw3CzZ0jrCI+vJzDsD7G0rYcMI8KMv5/jFjSGa2yRaQ22l4ntn5+gdEAwkNk9aDUHDQPU++G93/tIKN1JDxi8EMttd1yMv2GzJ/z0fFEEaOmgyDhSa0NUnOHQvl3WbJIlkUMlRCmIRzWF7eaxtMTDz05lZR7ByQ7CdnT0QsoLWlWkIBhJ5/ZjY1mF5EALoTwrGj1QsXB6IgMK2prJMU1kKN9xjc+0Ps9RVQmWpoi8e7CeR9/UZUaEpL9482bVgmeTKW0JBW03CP5+3OPUIh5MOGQ7R/bTgA5OfuXPnEo1G+fznP89Xv/pVxo4dC8Dee+/NpEmT+OIXv8j+++/Prbfeyp577vmRb/Aw3h+EEOz7s59ROXMmz373u7j/XBeQkC2jLXwfLIuCmhq8bHYo7d0Mh1Geh5NM4qRShEtKiLe24udyQTnaNDFDIXK9vex23nm8MWk2q2+5ncqlC8hGC+hpmEjjPoeSKypjp2fuJ5yIb+sTnw9UVa6LFY3iOw7C9zE9F7O9BSMSRfseUWmQKSzBjxWibBOpNOqgXQiVBq0wicE6cx3lTkB+zEiEGWdta2xXIOGSUD8PtrdAPgTCEgpkDqRA+AoDH8PIopSPFoLWqnpGAAum7YHUPsoJ/FCkCMar/3nz7ewZiw1NrpmhEOsff5y2116jUGTpb2omlcoRK4gQqq2jubULIxSiYvJkNjzzDNmeHmYUF+PP28jSo05HaoWSBqO6NrHv+jfpaWxESEndvvtSOX064ZISJh5/PEtvvpnEpk2Y+XYgwIannqJur722mob7pOKwqM9h0a0vnF+KOdybscj6JaStTnYuXMdORe346RgVbUkMBRFA4tNdUIMbKyFTOYopX/o12/GrG8bbUFGqueqiLN19wbh7VT7bzXGDSkvWEVvRHCE1Y8cXULTXZ+h9/mGC+TyNb0Rp3OsPTKp5536T6wVVJcsM9Dh+cM+DR6ChWfyWQVmxZtxIxTEHetRUau55zKK5XdBQq/jFN3NcfHUYQ+bNA4HX3jTIuYKQHUx6JYSmKCZI55Xzhgx0TNm3SWwGqzeGAT19QdUpEg7ufaaO2yzA7uyTJNNw7EEub6wy2NQOiUzwZCys+cU3MzR32Nz9TIiBBCxYZlBWsrX/0H2P2xyyl0dxwYf9tobxn8AHJj+XXHIJZ511FqXb0Tbss88+PP3005x++ukcc8wx/PrXv+as7VyIhvGfgTRNpp5yCrWzZ3P/MceQaG1FGMZQS0v5PgjBjDPPJNPbS8srrzCwYQMoFRAdgpZVtLIS5brYxcVU77rrEInKOA7tixbByefy1I/3wvAc3FQarRUylyXW3U7fqAnULl84tE1isAIEgXOx5+GkUsG2bEHM/EyacEkJZV6OVtMkWVqJrrAx5k5i7tmjAVDJLAM3PI2/2iNl38/E445jxO67v+P+mFxgs6qukra2jUipyYkcCo0uDKGz+Ts2rcAWpGPlJIrL0Xi4VijY1i1MDAVAT/dWI/u+6+IkEphuBmyXbNZFSkkmncPZsB6sKF4mQ8v8+fiuC3kbgBkP/IXxa96k+aBjaNjUSNkLT+CPGkXZhAk4WtDYn2ZRr4M8+yucHs7Su2rVNq01w7ZZ/+STnwrysz0cGvGYF/aIawiJQpqtKDlvMmOLDqKkvAnd/gKYJlZBOTVlNdR4adwpM4eJzwdERammu09w5S027d2S8hLNsQe6/Pn+EELkib0PZcWas45zGfuN33HJxd/Af+4GVFE9K0eeQW19JV86MfuO69h9hsfC5TamhHhys2ePIPAX8lUwRl5cqKmt1Bx7oMexB26umCgFa5slvQMSX0EyBY4n8s7Pg68RpDNBO3jwccvU+ErgumwTYRENazxf4Pk673OkSaUD/x0hgqmwkB2Ern71VId7HrdY1yyxLM2ZxzoUROEXfxxFzjPp6BF09UoiYc2u0/whY8ZUFhYuNzhw9/cZCDaMjxUfmPx85zvfedfnGxoaePzxx/nKV77ChRdeyNKlS7nqqqt2eAOH8eFROm4cI/ffnzX/+AfSNAMCkodhWUw++WSU59GzahVuKkW8qWnIhDBaWUnBiBH0r1uHl04H5eW8tw2Og5NKMS/sc5nUeFKilQ9a41khlDCIdbcHKxoclc8THCMcBiHwczm0s7XDrDAMtFLBBJnjUD2wnJEdTeRmFKMfXkznk2Ei00aTW7gecj5l1kj69TpeveIKJp98MkJKCmprGbHHHluRBCEE0z//ebrfeivIPTOCUB5dGML58YGBFqovTXr6WHruSGJnUgigorOFxIQZZGtH4utAn+BrqCgpgIH00PK9bBbteVhaochreBD5QF+BFA7ZrBu0GfP70AMcX2Oseos9wxYiGceLRJCGgQusSOawN64jsmQR3kP3cFNJBaZpYhUVMilmYeUVn4NTZ59mSAElAsBi4uBUmgXZr/2R0N9+jrH+zeA48hz8hp1w9z7p49zcTyX6BgQXXBEmk598Wt8SkIYLv5jlrkdt0tlgJP3bX8jR3Cm4eUGI4uMmM/G437B+pclYo5WzT8kQfZdDbUxdICbOOHnH5ry2yzY3Z2AJoako0fzgdzYvLbYwDJizk8dxB7hc87cQrZ0yqM5am6MjBmEYoP2gnWaJoKpkGIFpoyE1wg48fEBgWUGchGnmq08+5NwgPmPFOoNwSDN2pGLeXpuzzsaP8nFdi0hY0zsguPT6ML4ftMG1DkTSmsCYcckqyaypwW9QioBcDuPTgY8022sQ0WiUW265hcsvv5wrrrhimPx8AjD6oINY+/DDWxEfrRShsjKkaVJYX88+P/kJy269laYXXiDX20vhyJFE8xla4ZISnERiq2VqrSmoraXG0BxhZXmwN4WQBtp1CSfjNCx8nqKOFt72JiAQ+8rBdMRB5AXXaB34A/k+fp4oeOk0PN+PMbYcN9GO+8RqyHkYlk2zaEJojXJdmp9/nqqZM7FiMWJVVcz91a+GUsQBxh91FOHycl69/HJaO1ehqsMY39oL94ipgxvIQKaKx8cdzeH/uoXSVAunmwm+MX46WWmBBlNDBT7HHnMYy393BUIKKKoIRNbRKAUFIPyg6qOUxjAEhpS4SqHy2h6AnB0mVVGNmc1iKI83yuvZ88ApOP96EIAWVxBasxLhuRipJL01o3BCESRgxBM8UzWF/ZxeQk5QpZt84okf9jD5ZEIa5D77U0TbWoymt1D1k1B1Ez/urfpU4taHLdJZMVStkDKIkRhISp67JQUE01bfvjrM05NMnHqBryEiNV8/Nse8nl6i4XefGPzbIzYTGhQ5B5rbJaoHMtlgOYPtqXAIlq2RLG0MYeVJ0domm3ses7HM4DWpNLhbDJ8acrNPkZGfzzAMCMsg4iKbEygtCFl6yPjQcYNWW3GBJhLWRMOa3Wf4rNko6R4QJDMSxxUcvb83ZK549W0hMjlBPCno7JEIORiHYQLBvhMi+EumBelM8HnqqjU7TRj28vm04N9Cfgbxve99j+nTp/87VzGM94kJxxzDoj/+kfiGDei8M3G4ooKaWbOI1dQAUDl9OgdccQV9a9bwzEUXbZWobhUWUj516pDYWbkuhm2zS97H6SuLHyG5ooOmMVPJJRJULFnAHrdc+Y7boz0vIDamOZQ0T96pGSmDQFbXDabLQiH8VArDshAtSXQuN+Qe7XvZbSKUu5YupWbOHLL9/Sy85hr2vfTSYPGex8p776VryRJ2+853SO1fxnORZ/Ha+yl4djVOVRQvGubV/nlUtaxl7zeexTv0SK474jzCrkSg8QAbTXk2TrprMfuM0axc3Y3T0U3hnP0Ytd9+bLrnVkzPoagoTF9fimg0RCRk0JUEaWnwfZSQuOEo2bIqhJPDtC3azvgGT9XXcfCLz6B9H6e/F+k6iGwWz7RwQ2GkH9zmGpEokU0bef6Qz3DUqvlM+9znKPwvHzLQtePwasd93JvxqUZb52biMwgpobNvc//wvictXq4zcMoEUgVWBK4v+FNLiFF2mBXPmCxfYzBqhOLYA11ib4u0608E4uVwCOqrFZvazSBWwwNfBF4+qYzGkJpwSGOagVNz1hEYHvimxrahsAB8pakpVzS1BREzYVvjeIE+qSAaZGfl8m7OTp535LbTC83kAnLS0mFQGPNxvGCKTBAkvv/zBRMNfPYIh7aeIP+su18ipYPrZjFkhLDlorQk60QwjWC/mUaw7FlTfL71eWeHQmKH8fHg30p+AI488sh/9yqG8T4gpOTwG25g/i9/yUBzM4ZlUTx6NHtecsk2ry0dP56dvvAFVt1/P7mBAcxIhPo992T2BRfQOn8+ra+8Qqy2FmOnnYZCRv2+Pg5/5j6Mlx6iffFicn19aKXefQQ1X60RUiIMI9D95A02hJSMOvhgOl5/HQE4+WDUt1ef3k58AHzPo3/tWqp22omB9esBSHd1cc+RR5Lu6gKtWfG3v1E5YwalU0tpf+qFYKpDCEJxnzO4l6rJk5CGwV9TJp1ZF8sKYQ2tUpDwPB6u3ZVfjFvKqHGBz48246TP/wo1O01j0zWXIjJx9t6tnoGUR668gTmfPZ+XfvEL+latIm6FSZVXU7ZyCYabwy8sZvxFZ5PcdS/GH3Msq+6+CzuXwXUcZC5DonokYvCzeh5mURGmEHDCaRzacOZwzt4w3hdqqzSNTWxFgJSCypLNv6PljZLMNInc4qclBGRTcNmq0dQ8bxOy4dWlBk/NN/nthVnKije/uLJM09Qm2Ngq6e6TuFsMQG3+uQbCa8f1UUoQCQVmnErl21r5UXPPE2RyknBIBRVUA8qLVaBd6hUk0wLb2txO2x4EQXWrLy4YSAoWvWXgKUHEDnQ+gmB67JU3DD5/FFhGoB3yfYVWgTDfNoMbtJCRI2Jn6U+VYluaWVN9fv/9DA11w+2uTxv+7eRnGJ8cFDc0cMgf/0iqrQ1hGMSqq9/xtROOPZaxRxxBsrWVcGkpoeJiAEbuuy8j990XgMbGxqHXNxxwAKvuuQcIAkj1dkjJO0Hnoy/y/wgS3QsL6Vy4EDMcJtffP6QV0mwWTeu3T5BBcMb0fdxUilRHB7lEgqe/8x1aXn6ZdFfXVhqg1ldfxVoSISwlWoCbiKN9hR/1SbW1UVhfT6qwFD+TRfoKI5nAj8ZQ4TA+Mh84usWqk70YG5cx8uBDGHnQPOTGpcjOJvzxu6LLAoJ09PTpPP/DH/L6QI7o/BcRvocfKwQpCbU0IaXE3Wk08669ljfu/Buv/O53eIUleJHNjsUC8HM5/IaxFJeWIHvXo5e9yKKnX2fFM/PxHIeJxx3HnAsuwLCHExSHEZD/THc3n503jteXFQ9pfpQKND9nHLtZd1cYA7FF90ZryK2CeKOgtyVMf1xQFAvEyvGU4M/32Xz3rM1jVicdkuOuR2Kks+JtpERB/jcjRdCa0joQIeu8MEgIaBih6OoL2k7prEAamuoKTWmRZtJon6u/n0UIuPC3Ie57wh6KtHgnAjTobN/VGwioTR2MumdyAik15SU6IHc5gWnCjImKV5cahO0EyVTw3pCdY0RZK539lYyuXs/Cxj0oiml2m+4PE59PKd4jqWQY/20QQlAwYsS7Ep9BGLZN8ejRQ8Tn3RCtrGTSiScO5VxBUL15zzAcyItYA1VjcUMDRSNHkmprw89m8XM5zEgEISW+6xKtqgLY3Cp7B/ieR3zjRkzbJrFpE/3r1w9NsA2t1vdxkkkAclqQDkXJTKlh4NAJdFQERp217RuJtm4ksnIpoeb1RFcvJ7y2Ect3OKz1taFlOY5H74CLq/PkSgjU6Bl4c44aIj4A0fJyDr3+evYeU4sZDqMLCxkyRRFQ1NtJ99KlFNTWss+3L6B2/DhsU2K5Digf6XlYvodnmjSfeyGff+1mQtefz1Pfu5jGm2/E2LCU+JpGFvzmN/xt3jz0plXYD12L9egNWA9fR/gvFxG+/cfI9W++9/cyjE89nGSSR770Jf66337cc+SR/H2/mZwS/wozx+do6xRsbA28pf75vDXkRnHa4Q7FrQqVLyZmXobM8wJvlcBtCqo5mzoky9caxBPQ3LG56vjMawYX/y6SJzWbtTkBNr8u0OQofCXwPEk6E7SyigoUlWVBzpafn8pSvqCnX7K2SfLoixaX32QjJZx3UkDYkhmxjZPG9iBlIEqWQg+ZAWkdCLQHvY9sCy74Qo49ZviMG9GNFBrLdKmvaMY0PQ6a9TTfOvEaQrbmqP1dLj53OL7i04rhys8wPjJMOeUUGg46iJX33MNbt9+On8uR7u7GSSS2X6WBreZRhZSkOztJd3aifT8QB6dSmJEI0rIwo1EKR4zAiceDSbDtVZfy9XJpmliFheQSCdoXLdpMfLTGjEaHoiy075NMZ3CLo5i/2w8xvgRdGsatLKFpfoKdLrmLlSMn0Vo5Eu05+IZJuLeT4x77I/sWrEALyeuvb2TZ8jZSaQ/5/DlMPO449rz44ndsRQkhsGybirBJ3NH4CAQaWyvCvovcQmtVN3s2kcJCKnp62NgwiqaKkSTKqqiYMJ7j77iCjiWPkdRpetv7ApdcAaURTW/OQK9fhvHT47FtgUj2AuCPnAaFpYTv/BnOEV/F2/mgHfuyh/GpwPzLL6fpmWeC6U3DQHkeLY8+SGrdKKqnXJyfMRA88oJJzoWvnebQUKe5Zk6G76+L0BwSeGsFUUuT7BOofL6V5wtMU9PcIZkxSfHAUyavvGnw1hqDzj4xlJyec4LIivTbJuOF0KAFtuHiawvLCvRBxx3ssmajwfwlBlqD7wsS+aKUbQaeRPc/aXHUXJdLro2Q2Vbyt10EbT5NQTTYnoKoT29cUlKg8XyI2fDlU4IVhWz43tk5/M91kej8Pv/vwf1o7ylhpzFvMG30W7y+ei8+s7/L5Rfk3te93TA+mRgmP8P4SBGtqGDUfvux4ckn6Vm+HGlZ6LyIWUgZJLwPQgjkoNYnT1qU5wXEZ7CyIyV+/j25RAI/l6N88mT61qzBTaXws/mzqsyTisJiKqdOJdPTEwSV5knU4BnSdxxUPpNL+z5GOExOmhgXzUKMLQbHR0dDqHCI9H4mFQeO4ayHbmTR9D3YWDOa4r5OdnvpEdZOmM55+/+Yrv4kad5i98ZrifW2odav5/Wrr8bNZNjvssvecT952SzZri5EOo0tgzEWDeQyJk4qhVYKISUzzzmHF3/yE6RpMt1LMXnjm7ARWPwEVm8zXV1drIi7aB+K8uPHtlQYQlNg+pjxTkRBDOEF+9NoXo4/ZW8wTMyX7/10kR8ni7F2EdoKocbuPOTSPYztw3ddNr30Etr3t5rydFwIN/4LMeUHdPYIOnolpgFdfSYnHuxQUwm7TVVcE8rwx7/bPN5tMtAvUG8rtsaTwY3L8kbJmiabVRskrisoKdIoHZCIrJPP6iIQFx+862OMrt7A/S+eQF+inMKYQ3WFQX1N0ILzfcGcGT49A4J4yiCzRWHF9UFlBA218K3LI7R0GpgmOG8zVR6cCvMVSDS7zfDp7BX0D8ghPdGEUZreuM+08T57zFAcOdelMPa25YSmUVWzF5ec+U/aOh0G4prOxAQmTzuP8895Z5+jYXw6MEx+hvGRItvfz0s/+xnSMKjcaSd8x6Fl/vzNwuctncfyd6Ny0HCRgJxI421qzC2X39NDVzIZeAH5PlbUpLDWY6+vQGGlJpsJo0vP5uEzv74V6RmE1joYi3ccisaMQXseXUaEgomlgeCyuhBVUYYwDQxhkjpjLBXffoPd33yBXRc8Rd/q1byx8z68vPtRxJIhPE+RGzWF9m9ewbGXnEUol0P7Pouvu46+xkbq9tyTnc89N4gJyWPtI4+w6oEHAqPDvNEjgLQs7JISupYtY+kttzDjzDMpmziReddey8q77iLT20vF9OmsvPvu4ILmOgitidqSnoRPgWYopiBsKArsoJomfC/fcRABCUr1Q2EZItnP0HzvJxzGsuexH/wdItUP4Ri6uIrsaT9GVzV83Jv2yYXWqLwr+5ZQGoTvsXiFpj9poVWgxtnUITng7AIO2sNjl0k+Dz8fVCDdbDCqvj3kXEFBjKEKkhDBtFdpkaKnX5K3/cKUign1K7FMzQlz/0FvopIl62aSdscSjWggMCzs7hNkskHVSApAb2FYqEAamnQOfF+SyUHO2Xa7NIGAOWRrvPyhP65e0+gHGqWwrfEVHDHX41ufe/cJLcc6B2EeR/XIRWSaPfaccdg2+3MYn04Mk59hfKRYff/9KMcJWjf5jDDDMPBcd7uTX1YsRjadxpMmjh3CVD628y59dK1RfhA1IYSgfKzCigoKKgzMWIiqBot1r/xpaBxfWSKolwtASCI11ZRVjKRwxAjSXV0YlsWCafsSmj4AZWG0YSANCQhMNAWVNeT6+0m2tpLu6QHfZ9leh2IZEuW6OIkkQiucgiLeOuQEdvn7LUPb2bZgAdrzGFi3joOuvnqoDfbab38L+eqTE48PxWZoArF4ctMmFlx1JeGDplLdMJni8nJ2+epXAehtbMRJJLCiUQhHQUA0JOlL+jgehC3IusFd7/TaYP9qrRBubkhyIZwsGtAFpZ9o4tPWluSGG96gbUMnZUse5uTKjcwta0IbBqp8JKH7f0v2y9d83Jv5iYVh29TMns2ahx/e6nHLErSGZpFIB+RGkS+8Eoh+l6wyeOZVk/oaRUunfEfiQ/59g4dQyNKks0HLK2QLsvnOtG1pxtWtoiiapKu/lpffmktxQZzyol4SnVWsaSqkpiJIWq8o0ZQUad5aGxCgkK2HRtcNQ1MQA8cRVJZqNnXIgMiJre9xpAhyv2orFekMIAKjxIYRikhYc/KhLnvt7FNT8f6EylqU45vzyHqNw8TnvwjD5GcYHykyPT1gGKS7uoZywaziYtx0evOJI3+2EpaFoxTpaBGG6yAJTsTJwhJivZ2bR7vfBq11XrsjMCzB6L1tIhWxoK3ma5x4BwXVpQzgoRj0ug/+nMPqkE0FFI8dS7q7m5b585n80kukDz8Jc1YZGolfUIglwUKzp9qXl9zb8BwnaNkJQU+4hoFsEToDEe1SQALpOsRrtg4S9TIZpGkSb2qifeFCamfPBiDb1wciEHhK0wwsAXwfoXWQco+H4yV45vc/p/wPX6BclbNvdj9MTKKVlQHxAVRJNaK3DeHmqCk1qIoJ2gZ8skpi1dTRMCYTaCukEfwpD21Y6Ggx+B7e3u9uiqg1bGwVJDOCiQ1qqxyjfzcSCYfvfe9ZMhkPa+1ikokwvxrYEzVOsH/5RmTnRghFEPEedNG7m+79L2OfSy+ld/VqelatgryvVsXkyawO/xDVHbxmy59ZMi3Y0BpUbBJpI9DZhLaj2SEgPVuOuI+sVazeYOAraO0UOI7IB4NCUTQFCCJhh3Wtozlm73+womkiEbubjt4ikmlBfbXijGMdomHNK28G7stKCUgHkRvhUHDqOPuEHAuWmixZbQxt+5YEqLpcMW18YLNRPlrzi29keWmxQUWpZs50H3P4qjcMhsnPMD5ijNhjD5b86U+BQaGUZLq7A83OIPHJa3sM28YuKSFRWUu2qo7Qpg3I9laE1pjZNPGakRS3NW2zfGGaGOEwXiqFFfHZ7UyD4pE2Qgi8bKB+rN81zNoXShiI9yEdhRpMV6yIID8zheRqh/BACf1r1wakTGtCx96OfvQ0mFCGkYljixAzrT1I3vcq4ZISohUVdCxezOLsBPq7JE5EBi0nFA4hDBvqF7+81baag2PmQtC/Zs0Q+QmXlJCLx4MEdtse2m7l+ziJBH7EwCwvQrcMoNb30hWLs7Asxu7OHoRLSqidM4dNL76IYUfQpTX4fZ3UNRQzd58xpD2T5CFfJjrnAPyX70PfdVlQ6SkoCfRPFfX4E2bj7nU8asyMd/we++Pw4+siNLUJXA/KizVnn+Cw3+z/TG7R/fevIpFwsA2FyAYTeZZQ3NU6mf3LNyLQiN429PCV7F0RLinhpH/9i45Fi2h65hkqZsygbt8DePCnRbS5mkRSoFTQCgsExsHgZTYnCIeCao4Q+QpMPvx0cJeXFSt2n+ERT0psCwqiMHW8z+r1gt64xJDBOHs6K/A8A8tyUdqkuryDiJ3gW8dfzW/u+h4b2oPYiZClGFEVMJirLspy9e02Dz9nEYloqkoVhgGzp/lkc5JNHcGI/FA5U0NBROO4gfmh1jCiUjF9gs+PrgmxbpNkbL2iLy44bG8PY1gu9j+P4TPHMD4SeJkMr19zDavvvx8nkUBpHVQotAbXJVRaOpTiDgzlejmFJQjfxS0qIWeGELksyjDomTyD4n/dBUohTBPtuoE+yDDwczmMSAQvpwgVBuGgbjqNNAy0bYE/gOeUYk4agdfZh0jm0OURQpcfhTmijFHTptNyzp1ozyNcXEwuHkdmPOSBd2LuVIcxqZryvhhTfnIU67peR1oWQgjsomJWxEdT/rfH6bjgNFTYBg9cK0xF42LGLHh28w6RkqKGhqHPWrPbbkNP7XL++Tz/wx8GfkiwVdCrn81CFjxLoOIZus/7CyhNYlwt0y/+EwWWYvej96ewopTWN5ZCbS21o6uZMdLELSpH7HkcBQWlaK1Z0Rui251BUc9KKuqrqf3c+ajZh7+v0v0VfwnT1hWkaIfsQNtx/d9sZk3JvOd7Pwq0tiaxbQPhK/A9BmeTE54dWAUj0XYYou9tw/C/DiEENbvuulXo7fQJirZuwUYPco7cUoZHOhuIbTxv83ESiwR+QJ6nKIhK6qp9jtrP4xunO1x/l82iFQaeB+NHatCS5WsDEpXKHy7d8XKqS9uJhZLsMfnFoDVmOazvmBBkZWl4a505ZG5YUar5+ddz/OjLOZ5faLByneStdQaLVxjEk4K2LolhCqQfEDdDBjlgVRWav/0mTUmh5v9uDXH7wzYtHQKNoLHJYEmjwatLDC49Pzfcwfofx6ea/PT19fHLX/6SZ599lubmZsrLyzn00EO55JJLKCsrG3pdf38/F110EY8++igAhx12GFdccQUlJSUf05b/9+Glyy6jd+XKINU8FsNzHPB9isaMwbBtulesCIjM4Ih5/v+t3i6M7k5wHEIInFgB6dJKRChK8ahR5OJxYnV1JJqa8PPRGtL3kbZNtKKC1c9LJh+4FjsGdmEMN2Oy4slx2BGbigsPIt3egdlQSXjvCQgjaCnVbKzkrWRyc0APQfsJrRFr+7D7BOb48ax79FEaDjqI5ueewwiFKBgzDrctitXSQ82PbiB5xF4UlucYu/hZJr72BHKQxEhJpKKCcFkZXjbLiD32oHTc5liGySecgHJdlt92G51LliDy697KDqAzibIknuOBBn9VB4+9fCCnHz0B4WTZVQpmVRegq0ahoz7O3qega8YOvf3NG29kyZ/+RLavD9/zEIt6KFnxO07+1yHbJMJnentp/Mc/UK7LuCOPxK4YwbpNchs5UNYRPP2qydStu3v/FsycWcWrr7YS9lJoIUlbMRSC+lg6aHtKcPc5eds3ao1c/yYi3o0/YTbESv79G/spxAWfz+G4Iea/CW+uEnhZGKyiCAFoga+C+AmtA+PDEVWK+oouzjs10OgMtry+9XkHrUH77fR3P8Kz87t5vX4af3/pWCBE1oHmzpHUlLbzxUNvJmTl2NRdxzUPfIN4KvAEs/KH5CD5GYRlwuIVBi8sNFnaKDFkoCWKhCBsB5NkBWGN70NRTHHGMS5j6jXLGiXLGg26egUi32IG6O2XLFltsHyNZPpwDtf/ND7V5KetrY22tjYuvfRSJk+eTGtrKxdeeCFnn302DzzwwNDrzjnnHDZt2sS9994LwDe+8Q3OO+887rrrro9r0/+rkGpvp3flSqRpBpUZgrR4L5slNzBArKaGgqowdTv7xFt92pdLDDtGrLYWb8MGUlqihQQ0oUScdHE5UxY8TaSqgtHz5jHz3HMBaH3tNZbedBPpzk4i5eXEqqvpb4VHfjrAiJlpSkaPoWNVKcozMMJQ+lKWyE8OJC3SgMDHp9avZVx0Co2RCJHSUuKbNiGE2OwwLSXRqiq0UlixGJU77UTdPvvQ8tJLGNKgpMgkrQtRQOzJxQg3TXl8ITGRxohEMCwLq7AQISXdy5ZRXF6E0VbB+gfvZ/TRxw2JnqeeeipTTjmF2/fZh1RHB24igZ8Pde2hlH6KqfJ7KMy6ELZJJy1W9CR5YXEfB5S3QrIP4ftoK4QuLMNc+AiZL1+LGrcLXjbL6gcfJN3dHVTL8mSnv7GR16+5hjlfPge5aRW6qJLmxmZe//3vh/RM6x97jAmnfA4pTsPXW98aa8122wX9/Vl+//uFrFvXj2VJdt99BGeeuROGseNi6kMOGcOTj6+l8ck30HYU1xeEDJ+DZg3wVmgaU3sagzH9LabVRLKP8K0/RPS0BBsbiuDseTze3FN2eDv+WxEOwU++kiORgtO/H+blRfZmQXxeH6eVYPQIn0g4GI8vK9J8/ogOpo7beiZcqG7Czs+w1OMUFTocvlsBu4xvYP+Zj3LhH6/FMGL0J+C1lXNYcPluSOHheIGAzDIJ8r0kTJ/gb0O4H3nB5OU3TJLpPIkRwbYIEUxtSQFFMY2QsPcuPl//bDA88Maq4EB13zYG7+Y78AvfMobJz/84PtXkZ+rUqdx+++1D/x47diw/+9nPOOWUU4jH4xQVFbFq1SqefPJJHn30UebMmQPAVVddxeGHH05jYyMTJkz4uDb/vwa5gQH8XA5pmkSrq4OJpHgclGIgk6Fy3EYO+LmJGQbte+RSBovvKaZ1ST94HiHDJCsNtJD4VojyDY2Y0Ri6opR9fvKToTHxsgkTKBs/nld++UvMcHho/b5j0rygiExqc3q78n1iVjFzM0ew0djA/2fvvOPsquq9/ay1y6nTSya9FwJJCITQpTcR6egVFQtFvDa8isq1YX8VAbliQb1WrBSvIL0pPZSEkjrJpM0k09vpu6z1/rFPJhkIECCkMOv5fELCPnuvvdbaZ875zq/2W32MCybQoBoQKcH4I4+k5Z57SBaLZDyPoFjEdl0qxowhXl2N1poZZ52FEIKDP/c5ek87jfUPPsiHDq3jr49ZZPp6yDU3Uygq2mLTmeV0oEpF3Pp66vfZh75VK1kwKsfY9ADxjZ0Ef3iIziV3MuprvxiaoxCCRG0t2dbWKO0deJyDaWY6h/IY/aqKgbxAlCQiVAit+P2j4M+t5aTqTkAjvDyit4TO9JL46cfJ/eAxir29ZFtbX57JJQTdd95MIliEyA2ipM3SB1sR9ZMR2xRWbLn1r0xdeBbL18eHiZ1UUnPMwoBNrVuPKaX50pf+RVdXHsuSFApwxx0tFIsBn/jEgbxRSgMDHLrpb2TnTyCxroe6lMfJ0zZRFffJiAr60vVUX38p4YR90FWNeCd8BOexWxEDXeCW3xvFPLE7byCcMg89btYbnsvbmYoUjGmI3ipDKeVEOqgirfjch4osXmEzfaLihEMD1q97ScyX1sT8b2Kpx4AAISzS8QJjatdS9GK899jfc93NHyMMyz2LhUDjbA1QFhrfB+FGaes33WNzzolbFcsji21cJxJIACUfQiXI5QEEiZhm5mSFH8DHztuatj57SojWDpY1vPqzXW7pMXPSroldM+y57NXiZ3tkMhlisRjJckbMokWLSKfTHHzwwUPnHHLIIaRSKZ588kkjfnYCVZMnE6+tJSyVSNTWRnVryr+RxyttDvpASOiV0KGFsCxsVzPn1A2seziOsCycwEcGxaEPKa2hx5lGQczETg7/LXP0woWMmj+fjsWLh9LZ02PHDqumrLVGSsns974XC4sp4VR4yWfd/h/7GOkxY9j473/jFwrkOzuxbJswCIjV1DD3Qx8iWb9VTNXOnEntzJnMBw7ZMMiPf/QA6zueZ3x8NdOstThuFaGXQHkeQT7PNKeb8ZUBSgs8L8RxJenlD6KXPorY9/Chcff9wAe4/9Ofjiw+uppmpmMPdQ5TBMpBYZMQ0W+0rgzpKTqUuyGVUYjQR3ZtRK58ksSUA7BiMcjlhq3ZlppEvgsRjgEpyfRlKPT2YimBatrqlisNDvKB/Z/kl+GRrFkxiG5vpS4+yCUfEKQS84aN+cwz7WzalCUe3/pR4jiSJ5/cxMUX74/rvv7I0ra2DF9+38/JtFew5op3ce1z3yKtCqTLzSUbsh34SEQphwg8RPdGEr+4DJ3pRyovymzL9ZdLGwmS134E79SP4x/z/tc9l5HAJed63POYQ8kTUWHAcoDz6HrNUQcpjjrIe8VrhW5D6g0IioRKRAUNBZR8l0zOZkztKrxyqvqWgGottoqQZAxG12tqqqK09Bv/6RJz4LRjIgHU1w9PL5X4gaBUihqUShmNUShGcw0VHHdwwJxtLDnz91FMnxj1COvolggZ3b+uWjFxjGLhHGP1Gem8rcRPf38/3/72t/ngBz+IXTb1d3Z2UldXN+zLUQhBfX09nZ2du2uqbyss12X2+efz3C9+gQ5DtB+1aLATCcYfLHDTWbysGCowCJCs09TNqKavJYcC/FJAZGsHZcXQdpw2fyLPr5LMm7n1g0oIweFf+xptjz1G66OPRj3FzjqL9meeYdWtt1IaHCRZX8+cD3+Y9OjR259weZzpp5/O9NNPHzoWlEoE+Tyx6upX7ZI+YUIl5xyhWb+iBTddCUSCoHvZMvxcjmJ/P2PTftS/aJs84jBUWI/9H2ob8TPr3HN59vrr6V25ktXFaUgUIOijhjFsQqCiho1SErNCNsXGY7E2CvIsFy4cQgXYq59BzTqEWeedx7PXXz9U2VcrRTohGD+2GrlhKRTz2IUQu1SAQWAb8SNLWcasuYfrnJ/T471IribFJHs98h+a/o2nsXpTitZyn7WNDYdvN3C0WAzJ5/03JH5++qPHKWTyxCyFX9J8dd5nuaj5T8z0NqJdG0v5xP0clDxkdxuiuxVRym1nJAFCItw+nEdvJpx+EGrczNc9n7c78/dRvOsojzv+7RIoQEMqofmfK145uH1ju+CW+xxSjuYjJygs7RAEpSHT0ebuBoTQ5IrV0Vt0m3R6pUEHoMsH2zoFGzosLBllbN1yn81pxwTc9pDFnY86kctLwpZcCRWWhXxc01inufzDRQ6ZN1zMCAHf/lSRG//pcPcjNq2dknGNiiMODPnImd6eXN7KsIvYI8XPt771La666qpXPee2227jyHJ3cYBsNst//Md/MHr0aL7xjW+86Tls27F8Z5z3dmVo/VOmMP3zn2fTvfcin3kmsqMLQWEgQIcqssbYNlYygQ4V0rawasZSfWAFPU8uQpU/IbWw8WK1lJwa1k6+gFvv7iMptyNSGxupO/NMADZ0dsL48Uz61KeGXu4ninF5Q3R3v+Ypsfp6SuXu8VsQqRRhJoOKxdAhaBVVrqVsURKWZLBYZPNL5hWfPJmKMKR6dYAqOUgRskbPJGX5VAS9xESATFfQExtNTtSzodD/svloIBQ2G2QlmeZmak8/nUltG9h83314Xkhs1ChGT6xkdrqFjW05Fq/1yJcUfbkQnRvAbhoAaRFrX01VTJFsfgh7wxJGCYmKRVbUbCHknuv/SP+4ueDG0S++iK8eQnvvJJesHjafqiqHzs4NdHW9/pSa5lXtEAYo6dD08FOsO/MkvjHvMqrtIlPTfXzhge+zX/96fNtF9mzGLuVfYSSN1mEU/LxxBf13/oaOoz/yuuez3Tm+zX7uP3kOnHRQnNsfqaWx1ufsY7tJOJqlywQPPVPN2rYE40cVOXah4Hc3t/Onu0chRAkpx7Dv6HqqUwEzxq5Aa4HSFl5g4/kWv77zAsSQI20rmqirdiYXNTmVMorfKZbg8ec0q1Y18/UfzwZCYo4sW4/KxQ5lSDoVWT4HBhWr1myiLpnZ7rqO3C/6sy3bum1fL3vrczcejpezR4qfSy+9lPPO204mxzaMGzdu6N/ZbJZzzz0XgL/85S/Et4kHaWxspKenJ7I6iC3mV013dzeN5Q7h22NH3iwjPWZoe+vf7+CD8VatYuPDDyNtm541FpnuIvG0j5tKRWnjUtHXmqZ6zEyOu+YaVt9+O3/49p2Q6SZ0KuidcCp9U85CqCQL9/eYPn3PS2duBiYcdNBQ/zKAmONQM24cluvS2bqOajGIiMWxYzG01lSNGoU88X2kX7Jn9vnn88z//A/7plbyopqNr2yQNisqj0HlBqmoq6R6xixCJwV9m1md1YjY4qjvgADsGNguorqRppPPp0krYrdcxbymdvwzpzAQuLjHnYuz76GElx/Lo6tKCARCSmrSgt6sRnS3U1lbSUN9jEMPm0a8dx0yDEEqZOij3QTPb/AIgxCnmMGtiQoLaqU4gVXc578Dx5EopbFtwWWXHcKMGU1vaG/rx66lb/N6CGxGPbMC6YVsOvIgamsKHJBdyvzB9dhVdRCGiML2v/S2IBAgIJbrZVT/Bip3ws/r2/Xnfvp0OPX4Lf9XTa4An7sqMVTyYM1meOz5KmLxCiort4qZP/7rC5x9yPeI2UUaajajlWTp+tn8+O+fYm37pKF4om071UgBthNVa96SXSYt0FqQK7r0FGaS92I4tsCxozT2gWy5orSQQ9lhobI4YM4Ypu+C4OW363MfqeyR4qeuro66uh2r2prJZDj33HPRWnPTTTeRTqeHvb5w4UKy2SyLFi0aivtZtGgRuVxuWByQYedx0g038I/3vpeeZcsI/YBHf1LLUZdJ4rKEDqFvQ5oX75jOwssvjNxPp53GnNiZ/PEOd+hDTSmor9Acf2jw6jfbjRz25S/z3C9+QfeLL4KUNC1YwJwLLiAoFMh3tGP9+bvYq5/G0QHu2InYR52NN/OQl40z+YQTEEKw6Ac/4J3Nj/Oov4Cs04ArA45+z0HUzprNokWbKRYDJhwwi09deg7ev6twnnsAUcxAGKLTNRQuvR6kxL37V1irngTbxUmlqAd48iYKsw7khb4UWmejys8IhO1Q2xjHnTiOd79zH2THOmTriqiHFgqUAK8AbpxsQSGtrZYgiHqHzR5nccLFR3PXXS1UVLicccZ0amsTb3hfTzppMr9d206sqwUdT1G/bD2jlzbzi/dniJ/7n6gz/4PwN5djrXiifMUrtSnYWsoAISHwoppB1h75sbdb6B2Ay6+Os7zFoiIJ7393ifPfGSAE/OkOl/buSPgAuA5099q0rrLYUr+zplKzsX0+f733jxw593FA8PALh1DyohOk1Pjb1AsSQqO1wLLA39KXK2rwPlRssViEK38aJ5sXuM7WfnVSQLhNxmGookyvmqpXev4Gwysj+vv799p3TiaT4ayzziKTyXDjjTcOEz41NTW45Z/Qc845h02bNnHttdcC8JnPfIbx48e/6VT3kf6bwGutP9/VRXbzZmpnzsRyXTYv+hcbH36UWEU9M885h8RLBO5tD9rc+7hNoSSYMFrx8fd61FXvmW/PHX32YrAH0d+BahgPiYrXPL9jyRJa7roLYTvMPOtMaqZsrd2zrfUSQHSsw3nmLlRlHcGCUyAevf8T118Kgz30tnaQHSxSU2FRrQfoyGhuf6yLUqCpqklj2fZQXrPc91DOOn4C1gsPIns3ARJyfQgVRsIhluCR1SHrBh2yE+eRKmfgaaUYdcABHPqlL72+DXwN/v73Vdxz52ry7e2MrpVc8oUTmDRr7NYTAh/nvl/j3vEzZNfGaJ6U3YxDGyRAWCAlqnEiatRk8p/5NcRTL73d6+Lt8nM/kIXjP5qipx8sS0TuJwEXnVOiu1/w5ztcSr7AdTSzpyiEgGVrNPmiRVVFtL2ZrECjh4Kat0WUn4V+ictLEFl9gmBrSJy9TWiY62iOOCBk9QZBa6ekKhU91YFM5Bof3RBVna6u0MycFPLzrxV3ScHCt8tzN0Ts1eLn4Ycf5rTTTtvua9vGBG0pcnjnnXcCcMopp+yUIocj/YdhJK9/d69dBQHr7r2XNXfeSayykhlnncWo+fMRpTyxr59Gb3MLPaU4eR1DhSGVqpfuvgKPrxfoUGFJQX1Dilg6TVA/nsbDjuKI/3gnyWs+hCiVA11VCPnBKH7LshkcfwC3r0kxOJghXVGBVgqrmOGd+0oq/T50RQ3F87+Jmv7GU9xfL/bDf8G98wasDS8iAh+9RfAIAY4LtouuakA1TEBVNVD82I/fdHPK3f3sdxbf+Wk7v/p7E5YVEkkVB62TBGEkUEo+Q1YbpaK2Edk8Q/V2XFvjBSISMdsZf8su2zZRIoOOrDVbXttyjRBR9WilotifaRMU45uiV1s2RsHOQSBIxDU1lQopIldmTYXmCx8tMXvqrsncers8d0PEXm3/PfLII+nv73/N86qrq7nhhhve+gkZDLuAYn8/937yk2x6/HF0GJASHu1/+zUzZo3m8GkO/a3tJCmQjBXJhw6bgjgy9HmhJ4m0PMZVK/Zp0MTcHDnbYamWzL7wUlRTHeH0hVjLH0WoAO3G0Q0T0BW1qLEzEB/6HsdtXEvnty6m2luDhaI61k2y04pcSf0dpL57DvmPXkV4+Nm7ZC+CI99DcPDpiO6NiM71kK5FTZmHe8sPcZbci05UlpvaCrwTPmK6cpcRajOr1naiaWKosjMeCEEml6SqQhNzo6KAWkfiJ2p7oahIgh8KiqWog/sr/fa85XgYQjKhtxmj3GsXjWVF7VPiMYXvS8Y0KsaO2jrixDGan321QH1NdCxXgMeWRA1XD90/3KXNdg1vL0zCn8Gwl/Hs9dfT+fzzaKDK9khaPlboUZvZQKm9FRWGFHQMgLRVpIIcz3Ym2ZhL0lShOWQCpGJRwTmZrmWfVI4ffe12SqWA4nlfQk3cFzVmBgiB7GnFWvc8OttP64MPoH5wIaPtDKNnTmGMkyUV5qNYmlIekRtA5AdI/uIy4r/6Lyg3JH3LcePoMdNR+x+PmnYASAvv7M9TuPBqgtmHEhxwMoVLf4KatussUns6TngLU8dsQoptrSYiykEnyryyrSj1fOhVoUnGFLYNiXLRwR3RkkqX09T11o4y1RWa6sqoyGIqobn680UOPyAYJny0jlpqbBE+EPUYO+HQkKMOMsLH8OYw4sdg2MsYWLcuau4qNC4+W35zT9kBhVJIQvi0+5Ws9RrYUKphWaaWRZsSWCj2aSjXDRISLS18J422LOZ2388//7kGPXoqxdM+hejaGLmRbBe/dhz33fQwa772cUT7WnKb2uleupR8rkS7l2Iwp6CYAx215xClHLJ1BbH/u3b3bZIQqElz8M78HN7JF6ErdyyBYuRQ4BNn/YOG6gHCUAxZd7SGhho9ZLWx7UicSAumTwipq/EJ1dZ4nXAHPE4C8AJB0Yvuk4htjQIKy+60k48I+cwHSiTimlwhSnmvrdJ84SOlt2j9hpGOET8Gw16GHYshLQv5klBSXwksS+K6EpsQT1sUidHKWELpUiGy2DKqDa20ILDTQwHPFXaJF5/bTN+aNYT9PaimSYRT56MmzWFpW0Bvf4FqkY8qtkjBpl7B4u4q1uUqWJmt4YurjqeoylGrQkTVpltXDCvyaNhzCORJjG1s59Zvf43jDlxMU20vU8ds4qeX/51ffSOPY0NQ7pgeWWoUTfUwtsFj8liF62pcVxNzXvv5arb2hUvENQv2DamtUqSTmsZaxTELA5SGA/ZRXHhWiZkTQyaNCRndEPK3exw2bDauSsPOZ6+O+TEYRiITjz2W9sWLybcXUFIgdPSbeoeXYErKxbIt6lI1+J15ir7D8tSBnPQRwea7b6E/6CeVCPCsNPmKCQA4+LzQmcTq+D2LlmxgdLzI7Oo8lVOmAdDRmcGyLbqLmqlKU8gFZH2wRIhlQzZ0eS7TyM82LOAzkxehY0mEV8T8brXnoqw5BOpkpoy9jz985crIUiYmUnS/BULx4P9m+PYNcdZvigKN12+yeG6lJObEWbCf4oLTfX7w6xjtPRa2FQmlV8N1NDWVUep7oQTjRmlWrhcUSoLnm20u/KokHtO0d0la2iQ9/YJ4LLI23f+4TWO9JhHTJBNw7okeB+1n2lMY3hxG/BgMexnTzzyTUi7H0t/9jlLralIqT3V9FTMOnoZggDCepr6mibpDmyiefDHfHV1uXfH/vg3FHL2fP4dgzQpSmfVIy6FVjuEg+SJza3uxpMbDodRTIJeqIDVqFI5joZUi76Tp6GpHxioQRKnPArinZyquhBdyTehEBdhu9GU6cd83FmDc04bwPXTT5J26b0DklhvsRtsOpKqxn70b+4l/IAoD6MoGvOM+hJoy77XHeRvgOZfg22dhhU+hxFiUnDv0vEY3wPc+U+Rdn0jy1Is2tg0xB/IFi64+OO2okBv+Bu3dO/aIBdCfEVgSaqtgc5ckGYPJ46L3UXe/5MVVkqIv8MptLAolWN5ikYjDqg0wb2aksL73yziXXVDkiPlGABneOEb8GAx7GUII5nzgA+z3/vejwxCrdxPO47dAGFJY+K5X7V+18vY7Wb4pTp01Djvby4ZMnKC/hdEVAU90QWVSMn2swLIlhZ4eUjWVzJlRy6aOHNJNsKJdsURPYlKVj+W4PJqZTG8pBrE4jlMCaaFViJq4L6V3f+oV57FdetpIff8/EL2bQGt0RS2FS36MmrVzipGK9rXEbr0K2d8BQqDtOKIwAG5UkFH0tBH/67cpXHQtum7MTrnnno4WDQT2O192PJOD930hyfI1NgoIPfB9TSIO6zZJ/nyXw8I5Ia0dkjAUKKUJ1fZVkBRRdlhUnVmwoU1QKGlcF/ozkpgL2bxgICeGChpuYTAnEEJjW1Gml13ODrv5Xpcj5hffgh0xjBSM+DEY9lKEEAjbRjdOwDv9M695vta6XEDRpdceA8kxWG4vvetbCEuQjNkM5KBrIOCofZOsqZxP6uT34thJDh33T5b+8Y8UVMgE1vGnjlOpSslylGyI5yuOGbsB7bh4R70f773//bqtPqn/9x5k72ZQAaKQhUwvqSvfiXf8Ryh9+P/xprpRhgHxP38TUcxGedaA1fIsIvRRFXXoZDUkK0BrnH//iWDBqbh3/CSqdO0mCPZ7B/5R7xsxqfK/v92lrUNGRbLLrbmUFpQ8QcwVdPYIPntBiUeetegblJT8KD5oWBsvHTUvjRr8lg/pkGxBobVFUFDkiwLHkkPXbIkx2hY/iKxBzyy18MsF35vqJPrzI+ZxGN4CjPgxGEYIWqlhjVjxigyuXIYfaAZzUCj62LYk1JrVPXkGqru47brfoTo7SOc3M2Z0HfOnFoABxnc9yY2b55EJYyQdxclzQ86aX4uiBpnted3fSqJ9LaJ3M6AR+cGhL1GhQtyHbgTbpnTBd9/w2uWaxYjeTQgp0W4SBMhsHwQeMgigvwOdqESNnorz5P8R++f14HsQT6EaJ+AMdEIQ4B9/wRuew97EulZJVYVmU1dZkBA9UqUiS8zsqSHjRmku/0iJv9ztsnqDoKNborSI2lQoysUSoyKFW/p6KQToLeWcLdAapcNtjm1DWTB5PgTh8OaobV2CySeleeavWeqq38KNMLxtMeLHYBghSMsiUV9PvrMTwgBr43Ly2aia85Zfvn1fkU8q1qGwOpdhxVykCujp0WzenGfyvDg1Vp5TGps5sX4NA0GCqvHjsMdOLN9FIHvaoJABFWI/9wA4MYK5x8A2PcFeiijlEGhEMTfcegAgBPYzd71x8RMGuHffgNW6MhrXsqNv9MCLKvAVMmA7yGKu7HILy9YOCUEJWcoTzliI/eK/8I/74Bubw15GRUpTV62pSmt6B8vFDHVkfJszXXHa0ZEJ5pwTA046POCZFy1+frNLsRSJn54+QUubpLpC09UXWewiAbRtbWddPq6w7aj7qUJEYfIiUlxRU1NNJvdyq19/RnLSx5I8+Ks8FW+uY4lhBGLSMQyGEcScCy5Aaw09bRQLRbRSSAFWuficFYt+0y6W4ljljpYiXoCEj62LtHSF+NpBIFBSUkg2bCN8ykgL+4WHSP7oo7j3/Rr3rp+TvO4iZMuSV5yXGj8blaiKxMi2CAGWhcwNvKGiiSLbh3vzD5A9m6KGUlKCX0Jke6Imp2hE6CNK+eh44CFCHbX2CAMIIhec1bIkur8eGUG273unhwD2mxYyabQimdAkE5pzjungD9/LE49tPbciBUcfHPKLrxc48bCAGZMUk8Yq9p0aMmeGojKtsSRstxa0DpFoEjEY17Qloyuq/GxZcNi8gOnjVbkR78vZ3Gnxl7vct2ILDG9zjPgxGN5igmKROy++mN8sWMBvDzqIe/7zPwl9f7fMpWnBAo679lomTG6kuipBZULg2ERRqUKgLZCWIFkXH7pGWw6uU8IloOgJisTJkaRNNyEsa3gtnzBAVdYT/+0VyLXPYbWuQHa3gl+MXEmvVPdHSgof/QE6niIKFtFDBWJEfhCR7aXi4hk4f/rGji1UKdybf0Di+o/h3vdrrLaVUMpDPoPIZ7bOQ1oMNzMBbBE4WxulikImsmjJ7bhn3oZMnaD58iVFpoxXzJsVct5JPrf9OMdl79801OX9paSTcNJhPmMaIvGzpft6Mq5JJzWurREoEOVqiqoEhAjpUPJhXJPixEN9HFvjWFCRhLVtA7zv6CtIOAVsy8e2gmFCyJKYOkCGN4RxexkMbzG3nHEGvatWRUIBWHPHHWRaWzn7//5vt8ynYswYDvr4pcibruDvD0tKBc1gV2R0kSmomBrHabLxuwJsxwInhgw8QhRTGqIxBqikh1omTxmHqqpC9nWCZRGO3wdrzbNRqwtVLgPsFZGA0grRtxldu/1MKrX/8WSuX0rq8sN5fr3iDxv2IRPEGJ8Y4OPjn6HO9kjc/mP8o8+HLen7r4Dz7z9hr3g80jUqQGQHQCt0LBUtVJctTPolneC3i0bbMXSiEpHpfT1bvVczb6Zi3szhGVXNza98/s332tx4uzskejZ3SwaymjENmtUbBOmUIB0r0NEjCRA4DOImqsCKqj7vMyVkw2YJQpCMEvD44nu/wfiGFvaZtJTFzfORUmPJkCC0sSyorVaMG/Vaz89geDlG/BgMbyHdy5fT29w8JHwApG3TvXQpPStWkGltxYrFGHPwwVjuLjTfz/VwE4McOF7xzO1QXaUpdjp0OVXMPcdB1Faz8pebkIOKWFyAU82U2SkKlS6DXomEJZk6JkXVhDEUP/x9RKYX7bhYzU9jL74H4RWGbiXCAPraYdSUobTyVySZ5sEP3MEPvnALbm4ZQkBbqZJluVH8cu4dpGSRxG++ROFLf93u5SLbh9ywDGvJA4jBLmR3KyI3MOSuEqVytpewItfWDrmxJOHU/aNYoVJ+Bzd4ZFEowi33OTjb9NuaOUnR3QdzZygmNIV4gSDmujy11GJCkyIeqyvHd4UEQRRL9NgSu+wig1Q8w/iGFpSSfOND3+Cbf/gyz62eg9KCmBtQkXTZb7riPad4u2XNhr0bI34MhreQnhUrUL6PFYsNO+4XCtz7yU9ixyP3UqK2loO/9CXqZ8166yel8zjqL6jJ85kWPM2EOSXa/hWgNzusaamjcpmPOtlmwX+PpuPZkPqqgxg/4VTq7/8ZsqMFVchiSQvd30cQTI/cZeXeWaJnUxRfs8XqAyAEwgujlPJ0zWtO76+/exq7mCknjAksoen1E/xt8yw+NPY58Ldf38W951fYz94DpRyydSWikIkcWtIa3oRKa7S0EbxGWeLytTpZCU4cna5G142FvtWvfd0IY/1mSX9WkN5G2woByTh89GyPWZO37v/nfxhn/aaywtnGY3XCoQF//KdDR0/khdVao8reT0tqrv3ElfRl0qxpG01r3xnE0odywek+Velds0bD2wsT82MwvIWMWbjwZcIHrVFBgEhU8HT/eG5rm8G/W5I8/sMfRcHIZTZsGGTZsm48bwe+pF8HUi1DMAjSIpw0ByeXYlxWUJ8KGVXpEa5Pkf9FnF/c+H5+8vBnqTr2BySPOBUdTyKC8m/Zlg1VjcjODcjWldG4rSuR9/0eP59DhRqlwFOCnlIMrTW6sva1J1fMkX3uaYRfRAvJlribmAhYm68GISi+5ysvX9O657EX3R59mcZTIK3I4hT65TxtCQj0FjGzwwh0ohJtu5RO+djbr7CMVkjVjFTNlEqKv91tc+VPYvzkLy69Azu+1roqTWI7hst4LHptWz79/hKphKboRcH1YQinH+szplHzpYuKCCLRM5CrZP3miUgRkIpnQWeoSXcyZ+p63nfmflz2QY/aKuPyMrwxjOXHYHgLqRg7lglHHcW6++9HlIv0Kc+Dinp+2zyLgZKDY2lW96d5vtPjoNUb8OM1XHLJXbS1ZUmlbKZOreGjH53L0UdPfI277Sg1DP3ox5KwNo5Ma3TWI7BTqNCiQuUY17GZVnsCVVVl8RZLEU7en8JgH4nK6siiohX2M3fijZuJ9X/XUVr9Aik0QkCPl8BTkh4/QWupmnEP/ZPYwaejZh3yshlprVl799203ngDk9o30SbH0ZOYGgUooymGFvtWdOEdcBJ65kFbLww85JrFOA/+IRJkW8ZLVsJgd9m1pSPxIzRChZDrj04S8jXdXtp2CacvoHDJdZGoehshwtXEg/+H1O14vs2XfnQFazsOwo25BKvg0WdtvvvpAhPGvLbAaKjV7Dc95PlVVhRAD/gBzJ0R0lA7/PpxozQ3fL3Aw89adPcKjl4Y0lg+5+iDFFd9rsBVv43TP6j4/l8v55qPf46mRDeWUAzkqvnnk+fx2Xmxl07BYHhdGPFjMLzFnPLLX/LijTey9Pe/RwjBxOOP539vbCEzEAkfBViWJuM7/Px3q7nlH+vo6iogpWBwUJLJ+PzkJ4uZP79pqxB5Eyg5BSUmIlUbCAuR97Es8MMUvm8hJHg4jA42ceiRY0kkyh8TlgVSopz41qwnpYbq9+in78EmIBuvRHshKatEhR0SItnsJ2jptph78w/xPvlTdGX9sDk9/7//y5rbb8fpbWW6205dtg0tVtBdMxcrLDGokhx/9Tcpzj8qusArYj13P+79v0V2b0T0d0Tp9+NmoqW9tdCiLM9diHJq+zbsQLyPrmmM3GdesZyJ9jZBa+LB/0OoATQJnluV5oxDf4qUP2X5hkN58IUzCILR/PIWl298ovSqQwX5jWRe+AKX7tPGb/s/wsrBc7AT9cydEXLROduPx3FsOHbh9i2aZxwXcsZxOVz/p/zoxhqu/8cXGd+4Adf2WNcxiXOPeXJYqr3B8EYw4sdg2AXsd/757Hf++UBUafnqP32N/L4Tafuvc/FrKpBBSHrFRv76vfvp2JRFlN0rpVJIECg2bcpx771rOeecnRATJARF9xvE/KuRajWqIo7osamqnUxdNks26+Hg03jIOzj7Pw8YuiycvhD7mTuHjyUt/EPPRAz2sCFVz0DtJALLgVAxu2s56cIgvpJRJhgW3Z1Z6h67Be/ki4eGCIpFNjz4IJZtIbJ9xFWRpgQEfpbzGzYj3BQNSQ/9x7UUmn6Dte5FnMduxn7x34hCBo0A20WEHnL9i5Gw0gLtJiKXVxhEdXzeyFZ5JSjmEN2tQ3FNbwekbkHoDhAJpFpDXUWcuBMjFR9kctMKFs64g7yXZvq4tSTzacCnq99ioGsCt6+ez+quj3HRuRZxOskveSdS50jELD521DdAfBMx5udUTDruTc7S4rPn/Y1509Zx/9PzkRK+cvKfWLhvKwU+tBN2wTCSMeLHYNjFCCmZcMrR3H/WsShbIgHtOmQOmMngmQNYP/3XsPM9TzE4WES+tOvjm5pEFSX3StAaedKLxP/8TaTWTJ5cBX4J1TSFCR/+wLAYF+/Ej0Axi17yEIQhuqKW0nEXoKsaaGnbyE0LP8jpS27CUmHUDFNYFGIp8vlYNE4sjh8CxdywqRS6uwlyOeyBdkCD4yJL+cjdVQqZXJlDJ6qgbzOxP30TWcwiBroQhagNhkBBUEJLG4p5hDUAUqKrG0Ha5bYZbwwtBMRT6Ibxb3iMPRMR/dEe0E99lUVDdREpNDG7yNTRq8mV0thWiIWP7wuSdoLJowaoTA2Qcjfy39ddw1fe8VUcsgixxcImAUWx5VtvWvz49hnY4T0cv2Axxy9YHB3UHoF8x5sa12AAI34Mht2C/9FjwYtjhVsyoiKLkDpwIo4lCQM1pDvCUBGL2Rx//KSdPxEhUJPmULjoGpx//xmRGyCcNIfgkNOHxdAAIC28My6jZcbxTB/bFFlYypP8v9Qknp11HCrUHLr2ERzfo61iNKEWhJkiOpYEBI21NsGC4V3Ek42NOBUV6M4WEBIdSyK8ArbUNFRaYDlRarrW2M1PoSbui+xr36YNhoiKIaoAjUS7kVtODPaACt+w1QeIWnPscxi6YgeCtfcilJiMEk1ItQFBkepUicF8Eq010lJIoUjFsggpEYQIIdAkqUz20zk4mfENzSh/JS+2SOZPfunXiETS/6bnqEUjnn0JTngjQvcCMUJrLp798Tc9tsFgxI/BsBvIVCRI5S1KhQCtNUIInLhNtrECkg6iGKC9EK01liW54opDqax86wIddN1YvDP/a8fOdeLoqobhB6WFrhvD8/sczeKphxH6AVYux8WP3sAEZw2WCGhsqkIecQbBuJnDLrVclymnnMLKH7+IrRRCSgIlmFQrScflNtYnhbbdcnXgkHLb8G0mFlUQFr1lC9KbaUUhBFraFM+7guDwc974OHsqQlCyryCuvwFqCUJqKpJZil4MUd5U2w6Q0irvsUYIhZQaKRRCKJpq1tE6MI8DxJ1ove1XiSLwX09G3SsT2CcQWEcj9Ua0qEKLt4/r0bB7MeLHYNgNfDjp8UApRUXaAa9AoCEjLSxb4IypwuvJ4wQB1XaBd52U5T3vuh1f/Qdajt7dU98upyd8Hq5qRCQrkX3tSK0pVY/m3qkLOWvx/cwb5xOffwTeS0VTmdn/8R9UVqdo+c7nCEslps2sYnbFIEBkydEKhKR07hdxH/pjVGzwlbK1pIxS3HcEaUexQUGpnMYvI5dZqpLSKR8jOOLcN7gjez7amkhB/pJk8UwsvRgpiyTjRaL2HlvKDNhROwpAa4Hnx1DaIggkLZsP5P0f3Rd/w3U4bgGtLRAKHQrscV/ceRMVDkpM2XnjGQwY8WMw7BaOSCgOy/k8nvGoHuwkVcpQWczwyQev5dr9jqelt4lp9vOcMLeNz1/ahR0E2OETFJ0vo6y5u3v6L2OKozk/6XFrIUZ/wyTiAuY5IV8ak8Dd913AqzeRyHd38/zf/o5VN5lZbidOmKfPqaWqMhG5sxJpSmd9nuDwc7CXPYbV8iwE5TJlLxVAOyJ8yhYNbVkQT6LtalRVA2KgC1U9ivynfgWjdlZpgT0YISi415EofQRJF4IBoIggRBMDHCBAiBCtJZt7x+NYJZ5YfjL771vHhMkehYrb6Xv689iyjSCoJD75S9Ttc/TuXZfB8BoY8WMw7Cb+1Pcoz/3lGu5vOoCaQi9nvXArqVKO49x/8/g1pzG7YhGWHItrCSCKeXGD31C0rt7dU98uZycDTk0ErA8kdVJTb+14AbolP/sZXiaDiFXzHJWMszsp5nop7XcYFe/7DGrsDBACufJJrHXPo5PVaNcDFSKzfdEgr8fNpUJU3Th0ogJdOzq6VkM4cQ7FC77ztqvp82poawKl2H/jBn9E6D60jqNFNYg06EG0cFH2RIKgSPtAkbXdZzN7zjzm7xOlsSfqJ5M4+abduwiD4XVixI/BsJuwNq3isBX3cfjyexhwUmgBMcvD8TO8485b0e+xiItuYDqQitpE0LW7p/2qxAXMdF5/rM3gxo3l4FrNAlZSoTJki4reh++n/fmVrD3qMk7aNyD24O+Qfe2IbG+5bs+WeKDXW+lXIEo5wvH7UPzID7DWLEaNmoiaMv/tV8V5BwitYyjIoxD0oakA8fJyzTEXGpqaOezI6bz+/TYY9iyM+DEYdhXaww1+ihU+i9Bd6P1SCO0DElf5eJaFIzyE1LhtJXwRxyVEq3UouW80BFW7dw1vEU4ySam/n1H0UKmz9GeiYG8lQ5yeFupv/Awbqh3GTqijopiNXFsaoviU14kQ0bW+F/XrGjWJYNSknbugvREh0ZiAYsPIwIgfg+GtRGukXgNoHP9P2OH9SDYhyEAiQNeHiK6QRKiwpYUlQlAaDoS0yAM2AgE6BKEJrNN394reEqaeeiqLf/ITGpwBCp5GKY1EYYVFhJQ0xgMo+WTX9+O6PnFbRnvyRtBRjrxOVOCd9fmdug6DwbB3YMSPwfAKqCBAWNZQteXXiwjXEw++g9CbETpA6tWAj8ADQkAizgZ+C5QUjlAQanSDReyoZHmUPBobJSfj2WegrMN3ytr2NCYdfzyh51G6+XoqCt1oO0aoFaOcLAlbobQmJhWWoJzqvqMjv9QtJkAIwppR+Md/JIolMhgMIw4jfgyGlzCwfj3PXHcd2U2bkLEYYxcuZN4llyAta8cH0ZpY8F2E6gORQIsC6BywJROp/KW8wEGNceFvCUQuA/vZ8M5YlK4NQIJQHkgx9oOduMI9k6nvfCccfwyl736U7hc3MtZbS9KKrDtKEwU8SxWV+Hld6K3/deKEk+finXwRwWFn78TZGwyGvQkjfgyGbQiKRf795S/Tt3o1Xn8/YRCw8aGHWHnrrUw++WTmXHABPcuXs/QPf2B9PE7N9OnM/fCHcVLDs4OE3oDUHdsEjsYBWbb6QCR+LEAgxiQIPz0XITajxSikXovQBUCiqSaw3rXL1r/bcRPEPv0jMl/9GtbG9XhKMuA51MU8tJAEgJSaVwu41baLkFZ0hl9uyiktSFYTTj8QgGD+CSMysNlgMEQY8WMwbMO6e++l+8UXKQ0OgtZ42SxaKbpfeAG3spKWO+/ETSQoKoVIp8ls3EjvypUcd+21L7EMbSkSB6Hvk9nQTKIyixOPihPHKzUQoImjSSNZCTqH1OuACkI5B7QFwsa3T931G7E7qW7goB9dT/bKPF3PPYG0Cti2H2kV6WCl0lDMQrC9buMiSltXQKoK3MQ2HehDRDGHFhKr+WnCeW+28abBYNhbMeLHYAD8fJ72p56i7YknonozUuLnclE/qaQmDDwy+W5KrR3EKyuxGxsBkI5DZuNG2h59lPHv2NpwUYtxaDEGwk56li2joqELy9aEJehdB9UTIFUPYKF0kfv+VcuDjxxALFbiPe9+gdkzn8S3TsezL0WuXon7+O+wgiXoyXHUwfvixc5DWYfulr3aFQghqFt4CKOcQUr5AmLzahzlIe2oIrOWArIhaIUYVt9HR1lLQqIqo6KFL7PvCPG269VlMBheH0b8GEY86x96iOd+8Qu8/n4GN22i0NuLk06TavI44mMhlaM1KoCO7nU8dq0mKMWG/eAI26Zn5cph4gchKDpXUNpwCVDATYaEAWTaBUJAtgOS9TFCZvO9H9Xzr8fHEHMDABY9O5ZLP7SE4099F/Kp1bj3/wpLrAA0tAZY65qxz3uYgvsNQmd4k9C3E96JH0H0biK2YSmMn4bo3YSyY2DbiP4ipKshCNClLCKMAoG0iLq5h7VjEMkqhF+E3EDk4nJikXCqrENN3n+3rs1gMOxejPgxjFiCYpElP/85z/3yl6ggICgWoy9JIfCzGZoO16RHgUYgY5LxcwIOuzDg2T8P/7FRQUDTgQe+bHwtx7H4/44mt7HIIe9fjPIUW1qRqwAEHhvbFI8saiQeC4aucxzFX/5vBsedsBj3ydVIq7Mc8VtC2EXYAEFrgGj4L4oyIF797tderPaI+jTJ1zx1j8F2Kb3/G4jeTYjezagx0xDFHCLTi3vvr5EblmJ1rAUZteLU6RqC/Y6m+NEfILpbif/lO6iaAlKFEPqo0dNQY6dTOv2zJt7HYBjhGPFjGJForfnXFVfQsXgxfi6HXyigfB8nkSCWTmO5WQbaQjqboWm2QMRslJI0zbdxHqtA56N4nqBYpH72bEbNn7/d+4w+aCHPPHw7g22QLvf01GhsN/rX4hcaCUONY2uiOKGI/oEEvd2HMaawJBIuAoSOYlyCUFFa7ZNPFFmx5n94YeM4PvDBA7Z7fxmuwA1+htTtaBEnlIfh2RfuVSJI145B146J/p2sQteOofjh/4f13P3YK59EZHpQo6fhH3QquilqgKkbJ1L4xM+Q7S2gFWrUlCjGXL6OjD2DwfC2xYgfw4ik49ln6W9pwXLL2VhKIaQk9DzcigqqxlhYaehpsxh7vBsVGkykcMbXsGDyF2n58f1UxOM0HXAA00477RVrAU04tJr1U7tY8lfFgg9CohYsCVVjo9dnTN6MEI0IokrPGgkIkqkKkk2HoRO/QxcqEKoHUIRhZDtSoyV+4FCZzvLEY0+y8OBJzJw5PI5FDKwjseIT6GoHxgVYogMrfBIrfIhC7Kcg9uJq0ZZNeMBJhAec9MrnCIEaPXXXzclgMOw1GPFjGJH0rlqFkJJYVRWW60bBzYBWCtv1cdMCXIvk9DSxfSui17SFZzUxc+KJ2BdPZfr06a96Dzt8AFddz0lflzTfLehdp5k4GmomgB2Lztl3xhL2md7EiuZGXDcK3C36Yznh1JNx4y7hgTOwHngEYRVBhIgA1GgLv9GmfVM9YWhTKFZwxx1rhokf9+5fYj/za4TXhhQhNHjw3jTEJbZ6mlTxdAJ5JErOwLfPBpHc3hIMBoPhbYkRP4YRyaj581nx178ihKB2xgw6X3gBP5NBWBapRoEVTyGlz/zTbVB+OVIHknwXyY65jOzgb4CNtCXTTrWxCJAvqU8jBHz/y/dywx8O5Lnls3GcGMceXeDEM+YidBf2gmdRlRPhaYcw7OPvs07i0YmH43XDnL//huIzvTRs+AuOewRaL8DzQh7/8z2Mv+PXxBKDjB6lSCeLWD0acXcBTk8gyCK1jx2CVi9iq0couFeDGDmdzA0Gw8jmbSN+tNace+653Hffffz2t7/l9NO39kDq7+/n8ssv56677gLg5JNP5vvf/z7V1dW7abaG3U3tzJk07r8/HYsXYycSjF6wgMENG0jU15OoGKSisYcDPzSTmto4IuwBHJScRMGatWM30CFS9wFQEo04ui06DC9LvY7FQj7x0UVoulByBkqMpSgEjn8zQvcjp7Whpnp8vvBDlpRm4aoimzstHlz4QxY8+wvq/Gbq2h7jyf/5Kf/74hQOXf9H4gTIgQT5gmLuPnlyRQtrLSQpsLXAYhbEKITqxAn+hu98aKfsrcFgMOzpvG3Ez49//GOk3P5v5BdeeCGtra3cdNNNAHzqU5/ikksu4S9/+cuunKJhN6FQlESJmI4NWW2EEBz+la/QcvfdbHr8cexEgnd85zvUTp+OVv0kvY8jdBDVhBGjAY9AHoBUa1BiwmvfVFhoUY3QfZSsGDpMENMFBNvvzSAQQB60hxa1xLwrscN/IfVKIM4jwSEsDWdSYZfo7XTID1goBC+cfh4fuuXHJNNxHvrNbTxvncM0C/atAoXFYDbOc8saCUNJhhg3fX4O3/jCUzTWF4Dq8s3dcvNVg8FgGBm8LcTPs88+y89+9jMeeuihl8VhrFy5kvvuu4+77rqLhQsXAnDNNddwyimn0Nzc/JpxG4a9m5XyIWLhb3HCPnxZA/IiJunDABBSMvWUU5h6yinDrhGympLz37jBDVGWFC6CIk74AI66Gy1qqXCPA179veNbZ+EGP8dVDkVRidQKDcQpIF7i/tLYaASaBJZ+FhH2IlmHIAfkeTyYR5x8VK4GqEgnQAic1GjqahOsXTdAR1sfmdQgf/BmcURqJcISFH2LMExSncqzNGyidXMtX/l/h/KzHzyJFuUYIR2gRcPO2XCDwWDYC9h78l1fgUwmw4UXXsi1115LQ8PLP8AXLVpEOp3m4IMPHjp2yCGHkEqlePLJJ3flVA27mA1iEY3e16n1V1Oh+qj1V1Plf4lelr7mtcral2LsR+Rjv0eJra0mIIHQBUZV/Bmhe151jMA+kaLz31hyPwbtfem3R9NvTyAnatCAV4Bn/iC5+2sOD/0gSUfrqWg5CqnbkWyEISuRZoa1jCIxNJCo0ggBUgpkGHDngSfwxIFHkq9soCTi9Ogart5wOL1FB4mi6Dvc1bkftwXvQFi1rG+tp7V9DgiBUO1IvRIrfIx46fOIcNWb3HWDwWDY89nrLT+f/exnOe644zjhhBO2+3pnZyd1dXXDUpGFENTX19PZ2fmK4zY3N+/Q/Xf0vLcre/L6xbjr0ZQIhCRq9gSEAQOlq+jZdAUAvS+8wOZ77iEsFkmOHcvEc87BSaeHjTOl5kmk8BAii2u1I4VH3LLoar+KjuxHXmMWFcAFKKFoq1lLMvEY9bSR7ocXv34/mc0lbFejlM/6Z/+Poy+HifP6AUGoY9hSAJp32f/kptJ76FO1WC44lS5togInXSQ/Zw7h3HkkP3AiE3+ziJZ7BA8PTuGB7gk0xvKEsRTJ6jTjq1L4BejvG+AL34pzwH7NvPuEPhoaR2HZWWAZmv9iXd/XCPWrp8Hvyc99VzCS12/WvvdhPBwvZ48UP9/61re46qqrXvWc2267jba2Nl588UUefPDBnT6HHXmzjHS32Z6+/s6whK3clx2vlCXqpk9n/YMPkLn3F6SkTb6QoLhiORt/8nVO+tHHIXEU4GKHtxDzWxDaQ5Al+pGRhGGRpqp/UdX4MbQ1cYfmM5OZwMkArH7wnwx0dOLGViJQWDZYqp9lfwuYNB9A0BJM5OeFS+lSDVSJAT4a/z2PBkexVs2mbuYo/IESiY5uVLWFrJDImCT96f0YfPZFBgcTCKGpqamnvj5JMukA0NOTp7vHpjczj/1mLqFvsIZsXrHfnCSWFIShz+iax3Gr/xPL2r5heE9/7m81I3n9Zu0jc+1vR/ZI8XPppZdy3nnnveo548aN449//CMrVqxg7Nixw1778Ic/zMKFC7nrrrtobGykp6cHrfWQ9UdrTXd3N43l5pSGvQBdxAruRJAjsE/boQJ9FmPReiNCbH2bax0QU3XEg08xeeotTP5iQK4nRrbL5rmbPLLtgg13fJ1Zp01FUYtUGxG6gKQbCNHEiKw5AqjHDX9HyfrK61yLZt0999C/ehnKG0TaklSDRUVjSL7sSdscNvK54tVoDUJIelU9VxWv4Dup3zPHeZSbw9/QGnSTmBqlpw8OemitycaqOef9Azzx0Cze9a6pLFjQxNVXP01vbwEhBD09RaZNq0ZKSSxWQgiB5yva23OUigH9A0VWrnmKfzx4F+efvy/HHrtjws5gMBj2JvZI8VNXV0ddXd1rnveVr3yFT37yk8OOHXbYYXzzm9/k1FNPBWDhwoVks1kWLVo0FPezaNEicrncsDggw56LDJ4g6V8Kuj864H+Xkv1FfPcDr3pdjbicvPggSmfRQiJ0CVcLaoNmbNaiXA+0prIpi/ZD5p0pePJXFr3NbRDWYbMIRCUCjyj+RiMoolGUgonEnARC92733kL3YAe3AYrAPo21fiO35Zci9BpO6v4nnc/eg5/PIy0Ifc1gWxFpaZr2iwT6b7wPE2qJFAqNjRAWUsPvSyfzjUQlDS9JmE+nHfL5AKE86hJ5TjttGh/+8ByEEPzqV6fwwgtd2Lbk6qufJJ+PYom6emtJJ1uRUtDe2s8T9nwerj6J5eP2IX9MjPY/LmfWrFrGjKl4E0/PYDAY9jz2SPGzo4wZM4YxY8a87Pi4ceOYNGkSADNnzuT444/nsssu49prrwXgsssu46STTjImzL0BrUn4n0ToQba+XUvEgu/g26eDrHzFS6UcRZo/EATfR6t/E1eKhM4jaAcChJToMESIkGQtaA1OUjN2/yIWixEEaO1RbgrFlgBkgcK1+qLih/aol93XCh8i5v8EdAAIev27uLlwHk8F+wLTaf5zPftU12B3ZQg9jSg35sx2wNyvScChR4/BEgKNRShmAxUgFX3yQJRV4gAJoy2P3lAghUZKgZtKcLizhosuugwttv7yYNuS+fOjeTZU2azv7EbEktx6z0l8/Pw/ENOdPG4dxM21HyDMhBQ6BihUj2XVWQfxx3+s4VMX78/yQDJGKhr26k8Mg8FgiBgRH2W//OUvufzyyzn77LMBOOWUU/j+97+/m2dl2BGEXlnOqtq2IaVAkMcJbsF3P/TqA8gG0jpAqolROWW9tHy9hx1P4ucygEA6Gq00NRNCJh+uiAKkFQKF1la5C/gWa4tAiCKSZQThPgi1GS1HRy/pADf4bXSucFDApsDiJOdGng+/hk8MNztIhz2KubP6ybYV8QsBUro07lNkwsIo3X2WXMML4WwcUYmWTQCEGsZakQCTAr5ZNYrrBl9kbViFjc+Rzlo+VuEOEz5DlAokrv4gn9y8ni+tOYRAOmRrGvnWNe+hUT/AI1WHkA3XYk+uJrlfLYlcO5m6afxt3kx+05kmpwSO0My2Q765nW1+sVnyjwcdQgUnHOZz8BxlGqcbDIY9lred+Onv73/Zserqam644YZdPxnDTsAqFwB8KQK9I29fXUTqTWz9Jo5BucqxZQeItEtYLBGUIF4Nx3wepK3RKELfQYgSWgcIAUKKcjP0EEvmgRIxfSNu8W949sV47ueQekPZFRYHoKAFvhZUMMB4uY4WNZPs/P2Iv7CSYjJF9aQatGxEBx007j+RgAEkHfxH7G4eVmexUc8jBngaqqXmoylvaGl1Tg1X1h6BCJciyKGsk0Bs30UVv+HT2KufZnalzc8XPMTv1s2kN+xmWesYlrQ10jdNEMgSeu1mgo4iVSeMI5SadeMaqNQCW0ROv+d8m685k7llm7Fvusfmxn+6OOXH8cwyi1OODLjkXG+7czEYDIbdzdtO/BjeXmgxDSUakbqTrWWpNJoUgX32Doxgo3HKXdNBMQ6LASLLTglpaWRKY6ds0gR4ODySWkinXY+XkYxXmzjMW4SUkWUIbIQMEWg0LmAhyRILvo8d3IWSk4b1sHAEWEIT4JLRkYtu4KjDSC9agmxuQcWbCEsuqcYJHHRhJUiLUIzDku/gB4m53Fb0WelbjLcVZyd80i9NwBICbe+H1hqpliMYIJTzhjcqzQ9gNz8FMrKeNSUKXL7PEgY8i7MXvRtHhMSDIjk3BY6ksGqA9BFNBBPqceVw4WkLWGGlgBLoArL4S8Yn13DZGTGeXn0ST658FzEX7n/S5rwTfWqqhhdzNBgMhj0BI34MezZCkHduIOl/FKl7AIWmioLzLZCJHbjeJrTmY4ePAC6IGFrbQBxNLRoLSSeCEiC5r/IYuq0aZKhRCcFqpqI8zZHhU6BBK43l6rK+8YEsEETp6izFUmvQgNbj0HIMLppKEdAc7keXHg94+NKl4op3cXyzzfpnZlAzrYZZh9+M5RQAidBF7PCfpKng3MQpkBAg0q+8RNVN3P8KQm8iykirxLcvILBPil4v5kCpl13XlksR2EksJ6ChYz3ehFmE0iLM+8RxqE3b5LbTjUMhQGvi3hUUvQ2k4y4p2+fUA64jHe/l/uc+SD4PzRskC+dsv52HwWAw7E6M+DHs8Wh7DjnrMWT4BIIioXUEiHj5xSJ2+CCQJbSOi1o26BAr/De2WoQSY/GsjwASSz1bDpyWhPIAIIEgA2ozENJr1dBt1WERgtQIwA5D1ldNZn7n8yQtD8uNRERkDwnhJb26IhElgEEU40FYTHRmcof+PCmrEslm5llruTg9GfuQH1F3qMD1r8EKyjFNWgMeaI+Y+h5x/1sIegGLUCyg4F6HtoaXdogFVyFU99CeCELc4H8JrINBVKOrm1Bjp2OtWcy2gThjU3nSo6pxxlYzsKKVCWtfpL+2iTBdwRXHTMOtKPKfA8lh91IaJqgiUj2L1GvxuvvwM5EwCwowp+Ev3OefS8xNMHbUywWXwWAw7AkY8WPYOxA2yj5i2CEZLicefAf0IGgBwV/xrPdh60VItQx0DIsnsNU9FJ3v4tmfRKi1JPzPE8X+5JBqVbnZqCYrk4RCYulI0GgFWguUIwgrYlh+ESEksH1XjiAo/w3QTygOxHM/C0JwUQwuogTUlv9sc50uAoLuld2svqcZJ+FzwPt8rLhCbBPobetHSJTOphD7M9qaFB3UJaRaRzkYCYCeXpvf/KmGtq6bGDPhAN73vtnIsy8n8b+fQ/ZsjrLUpE36wAaOmb+Gex6sZtRBcYJSjBo9jhNOnMEZk6qAkBNLPveXbHwlkAIapeY73los1UxQ8im2t5GWY8ioyqgzWTxHZuNKDj5+DvE6uC7jsjYjiK2HquWauAPvPspn6gTjDjMYDLsPI34Mew86j1bNtNsaSRPjg/8BVYosHmWDRsz/MVokQKTKx1yEKuAGN1Byr0TLaSgxCqn6kLSVBxZoHBqCQVzlR0JCWAgRICyFG/pUhoMIC4RURJlnL3XnWGybDQYWtnoUj0/zWj9mvnUMS/73ZlbduR7L0qQbIbM5oGosWC7bXK+w6MYNfkXJ2pJzJdHIoW7xnV0Wt/y2i4UTnoSGB3h65QI+/alz+dF1J1D91duxF92O6O9AHTyReNVX+ay+lZmTp3D/IzPQupp3HL2J409979Dcfl5bZH0Af8877GMrjo+HrMkEhPJA/O5rAGi0N1MqKjKqkv7BCnrWruLgI+u5sH0cq9okA08Jik9GMm66CHn4aZv3nOLxnpOD1/HwDQaDYedhxI9hr8AObqWDO3giWU1BukCMtO1yTLabmtAGUR016qQLdD1axMsp8hot6pF6i9BRBOJI4vxkqEChLlty4tpjdmkdL8RnIFFg22gN87LLsISFkJotri6lo8KDmgoE/eUu7bo8XgJNnCg/qhtN0/YXpaPzs71Tab63hO1EkUR2TGE54JegzRpPn65BADWin7HWIFJ3bB1DOCi5D1a4BITNin+9yAkHdxD4EkSJ05rup+bRZk49tZ/585s49dSD2Wf/ehr9d5KiHYTgtBOWc9oJy1BUE1rHUXhJjvpEGz5d6Q87pqzp5DPTUbKT9f1VFMI8lpXjjpuacAef4cKH9iecLgl8iX4GpB2FmK/LSOoqQ/7xoMO7jgpI7UDYlsFgMOxsjPgx7PGIcCNWcCNPVE4jlBIXjdBdeNLnkfQkTh94CLQNOgFkEGik2ghlNxS6jUDsB8ojHnwZqVah9Fgk/Qg6gRSCIhAwv7CYcV4fK+JTkXjsW1hDlc6BdAGPLRlnQggECkWKkP2xeLrs9oqhSaGYghZJNNspwqizxPxrkXoFAEseddhYrGJ0ohcB9HfYZLtybKqdxKCqQ4pIFG3So+gOxrNvvGHYcCXnv4jxXXR+MWObOkGBlBKlBKWi4MADNuLe1U5zs8tnPrOW0U0Wo2oPZMaUiXzrC/+mIh0VcpQMEKjiDj0TrTVPtH2aFx/LMmbCAB0dKW6+bT/WrKsh7sTwK+KEfojslwQF0C7YFoSuoL1bUletWLNRMneGiQsyGAy7nu13LjQY9iAc9Q/arRQFGWn1KEZGIYCMTJGRLpJeBF0IHATtCHoRZBEMIujG0k+RLL0bK3gciIGUaCoR5YKHmgo0cTRxqvVU5vv/zcH5LFWqiCaJYvRQajs4aC3QpEE4FGK/IrDPI2QBoTgAJfcBKQitg4ennJeJ+9/AChcjtE9f6JNufIZG0YzER+LhByUeuX8C93kn4AqvnEkWADa/LF1Eh/jQ8AFFkpLzDbJ9jVRUeMRjAclEEdcporQmnvAZFV9HS8sASmn8Uj+2pWluqeXbPzpsm4EUSr68Yvr2uPe271PrXsZxx61n8oR+GhoKbOpIIyyXQV2BzPlorSEBwo6MXBqQgUYIiLtQX23ifgwGw+7BiB/DXoDEKoudiC2xIjZgYRMFDINC0UDkYAnR+GztyZVD6mYkm0BlAcqip5Ior8sHkoTMQegSTvgXNGkU6bLoUUAcRROhmE2mdCChPAglZoKspOh8g8A5Fi3rUaIB3zoHzx7edw5AqE1I1RwpAqAvzDJlTpaqcVAK42gEkpBV7bX8b8eH+GnpOzwXHMG/g3fxw9JVvKhPp5OZLxvXDv9JsnIzyo9ij7QGxwmIuwH5jM0LG8cyOFgCIJuPAxaWJVneXIfnlRv+Ese3L33NpzHYt45JjbeB0ORLcfzAZdL4AT596VPgJtEo3HtX48QEOgGiCXQIQkE8C3XVitlTQ8Y0GvFjMBh2D8btZdjj8a0zaPTuJ6V8SnJLxWeNwqFaCZJKoRGRu4sNQ8G/20auRDE5GcBF0o5iGpo0grbyaClAIVmC0DZWuLg8gkBRQ1QsMRJRQudIugKhZxBaC8vp43E851OvuRZBhkhouQCkRRcIi6O+nuT+n6YZXJPEcXxmzQ0ZM24uG6XFRn1cdLGEKjRjrcLLxrXDR5CpSrxCFfF0P15JozXEEgH33zuJQb+KMAwIQ41lJ9GiGqF70dohVFVoPAJ5ONqaEA2ofezwTiz1HFo04NvvQYuaaBre3QQIHFvgeyGWJQhCi5lTu5BSIYVFYyZN7TOwYoKmOA/woK5VM3WS4oRDAz52nqn+bDAYdh9G/Bj2WETYghveApRQ8miOyS7m4VQVGRlDANWhxdHZDUCpXHHZQfByYTA0HiGaPJSrPUdByarcJiOyEsly64vIKBr12ZIMlM+3iSxJAUKD0O2U7Mte15qUmIwWtQhdKs8pSmW30jGmfm6/qJuYhtF2NZu14u95C7es4nwNpyd9ktuz1woJSjP+8Pm0PNyMl2tHK2huaeDKWy/AcQSF8tZMnFhFKPeHYA0zJ6/FTYymJE/Bcz8fnaAD4t4XkHoNEAcdYKt/U7C/g7YmkUgmaV6epK8/zYSx3cTcErYtEcJiVGOSk896B6u69yW/CeZvDqmu0Hz8oyUOmK2GWmAYDAbD7sR8FBn2SOzgLtzgF6AjEYLwqBTTeFe2SIYQqbuoVCvLlhSNxkbgEbmnXglZDlIWaGJYrCtbkTwYJpp0eRyBYHCb4xZQSSSYLBSVCDJoqnd8YcLFty7ADX4OOiAl69Cqm049odxGFdLCo8E+mA+5PnOdkDsKDhJ4V8Jnrrv99fnyOGJqOUK4TH7HTJ57rgZLeCxefiKTp9bS31+ksjLG9Ok15PMhWmvGj5/LZ770cXKJrSlXoq8dK7wbmVqztZCksEEFxMJf8JfbL+CrX2mgtmp/wlCQTEzh2CNaOf7INbT3zuavfz2dSZOqGczlefJ5i6q05sDZCsva7rQNBoNht2DEj2HPQwc44V8Be8h3JXQ/jroFxXSq6UbSixaVKD0Vi2XlysrbfsNuTySEaCwgTShmIlkG9L3aRIb+0uXrQwpYogIhAizaXuE+r05gHxe13AjuwAW61CiK/q1U6F7iVoxG+0hC54MAHOAqDnBLrzlmaB1LoNdhhQ8gRY7Gxkr+785xPPX8XGprBem0y2mnTeX979+XFSt6SKddpkypRmxJa88PEP/TN5Ad65ALmxFjPdSoSehkFZmMR2trht7+Z/nEJ8YAAks2UlfThUBz94Pjicck9z5+HHf++2E++MH9OOmkyZxwqGltYTAY9kyM+DHscQi6EbofcIaOSd1azt56HkEp6q2ls1hkiCw3PpoqouDmgXIQs35JP3hVdmitJhb+Ho0s9+h6lS9pvbVPafS3R14rYoDGwVLLCOTE171GLWrxnfcDUA1Uu8cDg0ASJZxXufIVEALP+SjY70XoLmomN3DIcQU6ss0EAZxyymQOOCCqNzRv3qiXXR7/2/eQnevBstH5CgSbke0tZEbty6rmqB5SR2eMYjEKNu8bSJHJpSgWchSKsKx5PJWVWRobNb/61fMsXDiampr461+HwWAw7AKM+DHsMYhwPTH/q0jVisVGtEijxGSEHij3t9qa8RX9rYABNA4aBeSJYnJsKKewb1EvOvJiIaQox/70IAnw85pld0DfOmicBTNP3FJVmW0NPygsBCoqfqg9PF2JKyaVLUk7Y/ECqHpDl2oNPT157rprLbGYxcknT6GiwmXWrBSzZtW/9gDFHLJ9bdT1XRcQq7IwMY8Qmk1rN0dtQizF3f86ECEESmmKxQCtNcXilvUrisWATZuyuK7kvvvWce65s97QegwGg+Gtxogfw1tKSMh6ax05mWViMIlKvf0veMf/A67/QyTtbAkqlrq/3Ig0xyu7l0IEqlyl2SvH8DD0X61ABZqBjVDKQ7oxpGKUjSCk0K+54wrI94Adg41PweoH4JRvg5OIRE+ILMf3SCxCAmwyVJMNpjIplia0jty5G7aDaA1/uN3hX0/ZtCxbSefaJYxvCog5cMstq/jsZw/ioINGAxAEir/9bQXPPNOOZUlOOGEixx03aavLa6g6dQlLLwdPw0NJ9H45BoVHZ7aRex85mBVrp2PbnXheQBAowlANm1AuUyLuQPtmsG1TRcNgMOy5GPFjeMsYFIP8K/4gRVFEIljprGRKMJUDvAOR/mNI3QzCxQqfxlZ3Iukqp6RvcTJ5CLp34E7bure21I4RoCU6DEBAogaKGbBdyPUEpOokT/9WUBzQ2LHoCjsGA5vghb/DAf+xRUBpStpFYeHg4+Mi0NgqQdG+DMTu6c/wx3863Hq/g9AerateQGnJmlaXudNDlNLccMMSFixoQgjBd77zOM8+20EsFllpfvI/T9H+wH185OAcwf4noCbMRjVMwN78AFEam4C8BU9VcXvhSH755wl4fgoheonFLHw/Ej+6fKrQGlSUWu95Cr+vl/1quoEZu2VvDAaD4bUw4sfwlvF0bBG+8LG3eZu1WMvZ1/sC9cGaciZVWJY7WwTPG2VLevpWa4RW5VgeFXl0LDfKCM93QKpOM9AmkZZga9HESBx1rRRoKkAE5JVDhx6FR4qcriWGxyL1Yeb37Utt/cuLDe4q/vW0jWND7+YOfK+E7biEIXT0SMY0Krq7C2zenCMIQpYs2Sp8+tq6cDvXcMtyn4PWL2H64/cQP+wkSmd/Aev3/0b0KFAa0g7+8RNZcU0aS5Z7lmnN4KCHUmDbkfvLtiLBE+iyVNSCo8ZsZt1Nf2T60Ye94vwNBoNhd2LEj+GtQfeS1YuwVZ6oXk4lSoyi0n+CguiGsvCBrRaWN3nDqN0EpXKtHys6pgVhqLHjUDU2arBpxQAcnJRCa4EQNpFo0ijlEquuKld19knIGcSowgsLCFFBKyfw7qqzaO1qfpPzfXMUyzUCnVgcIbf0G4NSuf+o61qkUg5PPdVNECgcx6Knp4BoX4cWipyy6cxKVEuO/WIPwMLT8C++AHvzP8ATMDpN87ok3T0O++5r0d5ZwYoVPWitsSyBENG+hiFEojN6fpVuwKZ8mkxnCyoMkSbH3WAw7IEY8WPY+WhFwrsCJ1aFLyyELlIbNDO7sIK6sBNX+6+eYTWMSMS8dkp5jC3ZYZpS1LCilEIFBdykRvkQlqDQCxVNACFzz07wwHeLYDkgUmjtgbCZ+57pKOkSWEfg2Z+lmhI1uhst6pkp9owMpnGjNM0bBOmaeuKpSrxCDo2gvlrj+yHz5jVSVRVjn33qcJxIgHR2ZBkvAkCQsgOqXI/OnM17/zqZDbf+g/qpY/jul6ZywJxNkYlMByAk0hpPVRWEYTnhP9QEwUvFqsAWIRMrc/SXHFYXRg2JMoPBYNjTMOLHsNORaglCtTHeEzTHKpjsrWR6sZkqNYhEEytXWH5torYRGokYyuTaFouo4GAaiCEYADw0EkU1VtxFZYuEJQiD8vc5AtuNUtwb5uzLwstPZunfllMaDEjU1jLvA1NJThAU5TEoa4tbK44W43bK3uws/vO9Jf77ujj9GcGMhcexZsljxEUPtdWCOXPG8OlPLwBg7NgKFi4czeOPtxGEmlBHFaSPH7WZ7v6Q/3z+OLKBQygVa7o7Oe38Odxw3TxOPm4Tk2eMpbq+gd7egPXr+4bEz/aQ6Kh0gI4aggQTD9gmoNpgMBj2LIz4MexctMYNfoull3FwfhlCT2Vs0IYWAkuHONvE17ziEAii3ldJNA5epkQYVJKsGWTb1hNRf696tJiJ1EvLx1ygInKlSYdYpYUONMLSyKQmWQcIiSZFKI9l1KGXMerQ4fffG7pOjR2l+dlXC9zzmE1bp8Mx/3U40ycGSCleJjouv/xg7rhjDdde+zTJTpf3Ny1mgt3Dj5rnkgkcbAklYaN9hecpvvbdCt5x4le57rqn6e3dwLJl3fT0FF9lNlFlaqE1Ip4gVVvH6Ze+8y1dv8FgMLwZjPgx7FSc4I9I1QIIJAUOLSwCthYKfC0iK49AE1AazPHQD6G3BbQOqBwjOPIzNlVj0mzp4o6oI7COxA4sLFbCNlalKFPMQdoh0t7iPtPl1PXRBNaBO3n1u5Z4DN59zLZicvtuJiEEp546jcMOG8fnP1dJV98A6UyW1flahJAUiA2d63mKnp4Cv/vdi/z73624rk1lZYy+vtKrWn5AIKQgPWkaBx88htmzd6C+kMFgMOwmjPgx7FQs9RiIJFrXIWkZOr594bMlUNYpFyqsQJIpW34CHvh/mr51HtKyEAIymxUPfsfj9P/JIkQSsPCs9yLpA+mitYvQ3jZ3U0TWI1Fuf7FlJg6hPITQOvqt2YQ9lJqaONf/5ETuvns637rqSTpjWQrFElIIlFKEIYShoq+vyJVXPorvK7TWhGH09yshRFR3qKLC5aijxvP5zx9sXF4Gg2GPxogfw05FlJ1GWqTQOoUgB9uJ8dE4KBrQNAEJkGmkWoomCeQp9EDf+hBpwVCTUaHJdEo6lrk07DsdLZrwnYuxw/uw1GKUGI+lV7Al80gTQ+AQinkI3YtgEyAoWf+J714aNewcYcRiNu9+93QGBz3+9KdlPP305qGsLa2j4oTptEtXVx6I/j8MFUoDjgRbQmG469JxJJWVLum0y9FHTzDCx2Aw7PGMvE9/w1uKEtOx1JMgkqAtNGkE2bL1JRJBmjQhcwjt+QTyGGLBT8pXby1Q6HsOOvS38eRYgECFAr8oUdY+lJzLQcQJrJOx1ONYaimhmIXUGwAXzzoFS7ci6EFTixb1BNZx+M7Hd+me7Imcf/5sisWAWEzw7LOdZDI+8bjNtGnVrFnTjxCgVFTbh7gDUkbiZ1I9uBY81wrlPl9KQTbrM21aLQce2LSbV2YwGAyvjRE/hp1KybmUhN4IeiOaOII8IePKcTwZNBUE8h0o62B8+xwQDp7uwVF3o6lC0INiKhVNbaQaShT6QoQox+sIl0RtmoZ5h1F0r47StwCETcn5FlItwlaLCMUMQutYEA7oEEs9gdCbCOVhaDl2t+7PnoIQggsvnMeHPzyH1tYMn/3sA+WWFJqVK3uxLIHWGi0ESpYLUAYK8h7kNEyqhRWdACilCEIHd/y7+OhXUxwwO+TS93g45tPFYDDsoZiPJ8PORVRQcK9DqsVI1YJUq5BsAmwC60gC66xyE8+tBM57CPQ5oAeI+ddj6efQKuSIT8KDP8hQ6AsBSawiyYKPTkClPrJV+AzdV6Csg/Gsg19y3CK0Dn9Ll7y3otH0OF10T9vIlHcGbLjfhlDiOBZhqHEcgRhdCa5D2JVFxG1kwiEseNCe2TqQdEjVNGLFUhRKggcX2RRLgss/UnrlmxsMBsNuxIgfw85HCJR1AMo64HVcY4GopRT7CkJ3INRmYnMmcOKvBmh/6lEoPs3YhTOg4jxCYTKJ3iwazaOxR2i3NiORzPxgAeYM8sL3qpg4sYLm5n5cV6ClIEg4SK1JjK2iFAIdgxAohAApJQqLYi6L7UZZY64DS1ZI8kVI7hk1IQ0Gg2EYRvwY9ji0GIW2RgEgY7WMOWIy8P433QDDsJU2ayPt1iYsbFav7qO/v4iT1kw5P0vxnxP59pWV/PznS2nvztI+YRY5uxa3Ko4cLJEJFARRI1Nd7ukVKJuVSzeyz5yJAHi+oFAUJOPmqRkMhj0PI34MhhHIBnsjEovu7jx9fcWoOCKCqknQ2beeF55ZzNf/qxnfU3TEn+Wy/zmN3ECCmCXwLYH2AwJLEI9blHQlys/RvvIxKLQybd4CGhscaquM8DEYDHsmpvmOwTACies4Gk1bW5Z83ieb9SgWA1RJYMtOFj2TIAw10hKM9jv56Xt+zfTSRvZxQygFeF7k9vK8gLDQjQo9/MIAnRvXsPzxu7jgtPxLQ7sMBoNhj8GIH4NhBLKPP5vOzQX6+wvlOj8aLwxYfltIqVTCtoc3kq2v93jP6Rs48MAmJkxIEotFzVKjQogKiY8IMuhCJ5V2Oy3Llu+OZRkMBsMOYdxeBsMIJK7ivHBtmtiRWZRboJQVtD7gsOomSWUqyWCt5ptXz2f65AHOftdabAtaNkxiw+ZBGhsT5HKKzZtzQ+NFDU19VChZ29LHNdc8hdaa979/v5fdW2vNbbet5sEHNxAEihkzarnoonnE4+bjyGAw7BrMp43BMAIpFAK6XhQ8+zvJwIBFGAIoLMsjEUsTdwcZGHR54plGevtcTjl+kHTNPDKr2mltzVEoBNi2xPMiC1EU/AygsSxBMulw882rWLBgNLNm1Q279+9+9yJ//3szrhtZj9raMqxbN8BVVx1jqkMbDIZdghE/BsMIJJGwWbWql4GBElLKofgc31dU1dRRDNMovw0IefTpyRSYTbHYzdKl3XR357YRO8PRWqMUjBqVAuD221cPEz9BoLj//vVDwgfAcSzWrOnjhRe6mDu38a1ctsFgMABG/BgMI5KBgRLd3VG8j1IaIQRCRPUn29tzHHTQaMKwnlzOo2cgw333bSQMNbZtEQSvPrZlCTwvxHUtLGu4JSeb9SgUfNrbc7S1ZfH9kHTaZcKESlpa+o34MRgMu4S3RcDzM888wxlnnMHYsWMZN24cJ554Ij09PUOv9/f3c/HFFzNhwgQmTJjAxRdfTH9//+6bsMGwG1FK86Uv/assZiQQtbKwLIFlCXw/xPdDhICNGzNs3pwjm/XI5QIGB0tDViIpt/7ZgpQCpTTLl/dQKgWceebMYfeuqorR21tk5cpeMhmPYjGku7vACy90MW1a9S7bA4PBMLLZ68XP008/zZlnnskRRxzBvffey0MPPcQnPvEJbHurUevCCy/k+eef56abbuKmm27i+eef55JLLtmNszYYdh/PPtvOpk1Z0mkXrTW2LbAsSRBEVqAgUDz3XCfPPddBf38RgDCMRFMYCrSIAQ5KMfQHKFd8FkgZqaMxY1JMmlQ17N5CCDo785HbDA3osqtMc8stzbtwFwwGw0hmr3d7XXHFFVx00UV87nOfGzo2bdq0oX+vXLmS++67j7vuuouFCxcCcM0113DKKafQ3NzM9OnTd/mcDYbdyRNPtLFyZQ+JhEWxaFEqhWX3FzQ2Jpg+vZaOjhwdHTmqqtwhsYJTD9KOWpEA6ABKXcDWoOeKCoeJE6uor0+SSLgvu7cfwIB9AIxtRCmJ8HtJ5J/EtUo888zmXbcJBoNhRLNXi5+uri4WLVrEOeecw8knn8zq1auZPn06X/ziFznqqKMAWLRoEel0moMP3trw8pBDDiGVSvHkk08a8WN425PJePzyl0toaRng6ac3s3FjhmIxCtwRQjBmTIr+fo9YzGb+/KitSDrtEgSKXM5DKQ1WCuwE6BC2NBoRFrg14PUMxQtFAiqF1ppUynnZXK79XYwgPglKAZYEnEb8+PHY/XcwZ07DrtkQg8Ew4tmr3V7r1q0D4Lvf/S7nn38+N998M4ceeihnnXUWL7zwAgCdnZ3U1dUNS6EVQlBfX09nZ+fumLbBsMsIAsUXvvAQjzzSxtq1A6xe3U+pFCIlhKHG9xUbNmQpFkPGjEkNu3bUqBRBoCK3VsVMYDtp6NIhHrexbYllySFRpZTmvPNmDTu1UIRnlktSCYsgUASBIvQVASlSdRP47GcXvkW7YDAYDMPZIy0/3/rWt7jqqqte9ZzbbrsN143M6h/+8If5wAc+AMC8efN4+OGH+fWvf83VV1/9hufQ3Lxj8Qc7et7blZG8/r1h7U880cXq1R24rsXatRlUOUAnDBkKXI5S1hVLl3ajlE9lZfRz5Xk+iYRFXx+gAnDroLgpsvgIQEdjuK4AHbnPhIBUKuSUU8aSSAzQ3DwwNJf+jM3y5Y34XkAiYVEqKbSOMs3OPGcGPT2tbJOnsEezNzz7twqz9r0P4+F4OXuk+Ln00ks577zzXvWccePGDVluZs4cnlEyc+ZMWltbAWhsbKSnp2foQxaiWiTd3d00Nr5yWu2OvFlGeszQSF7/3rL2++7LUFVVgZSCioqAzZsLQwHKW+r0RIIlRqHgs3lziaqqNO3t2agQYlcRIUBnV8Lo00F54PchJSjt4KZqiBEJHMuSnH76DK699rjtzkUpTej1YNsujgOpZORAU0ozbYzeK/YT9p5n/1Zg1j4y1/52ZI8UP3V1ddTV1b3meRMnTmT06NEvU+Nr1qxh9uzZACxcuJBsNsuiRYuG4n4WLVpELpcbFgdkMLwdOeig0dx991piMZtRo1KsXNn7snO0juKCABxHMGpUCikla9f24vtlhaRL0PsI1BwMwiEWDzlwboKNi/9JLhsVKjzyyHFcffUxAPT1FVm6tIuxYyuYPLkaiETWhPQKmvvmoLSFQKMRjEq0oF6jdpDBYDDsTPZI8bOjCCH45Cc/yfe+9z32228/5s6dy6233spTTz3F97//fSCyAh1//PFcdtllXHvttQBcdtllnHTSSUbFG972HHhgE/vu28DSpV0MDJSoqHDJZr1hFZqFiKyhUkry+YBnn23HcSw6OwvDxpKlTYjOW5FOJSecNI7f/fpElPoAHR15amriQ725fvWr57j77rVDLTBmzqzl618/gnjcZv9ZIe7KR+gujidQLo2JjTgiy0knHb+rt8ZgMIxg9mrxA/Dxj38cz/P48pe/TG9vL7NmzeKmm25izpw5Q+f88pe/5PLLL+fss88G4JRTThkSRwbD2xkhBFdeeQR33tnC3/++iv7+IvvvP4pSKeDZZzvKXdmj+jzJpEUi4bBpU5YwjOrvbIvW4NoWFekiTY1RXJCUkqamVHkczQsvdHHbbauJxWySySjba9WqXn7+8yV8+tMLuOyyBVxxxcPozmaEAMeyOOOMGYwdWwHAUy9KrvxpnP5BwYQmxQ8+X2CsKfpsMBh2MqK/v387HXoMO8JI9wGP5PXvjWsPAsUll9xFJuMhhGD9+gFWr+7FsgQVFTFiMZtczmdgIKribFmSMFRDMUKWJaipiTFhQpJrrjmZefMaueWWldxxRwuZjEdtbRwhoLu7MBRfF4Yay4KamgQ/+9lJ5WOKp55qp7Mzx+GHj6OuLgHAA4skl3w9hdIgBYQKknHNg/+bpaF2t2zZdtkbn/3Owqx9ZK797cheb/kxGAw7hm1LrrzyCK655mlaWzOMG1dBqRQyaVIV+bzP6tV95HIeti3xfYVSalgri0TC5uyzZ3HccZXMm9fI3Xe38NOfLiYej6w8vb1FVq/uY/ToyBK0cWMGzwuxLMG4cRVcf/0zrFrVRyxmcdpp03j3u4d/kXzzZ4kh4QNgScgVBF/+nzg//1pxF++WwWB4O2PEj8Ewghg3rpIf/vBYisUAyxL89a8r+PvfV7Fx42BZ7Ly8lo/jSFzXor4+yXvfuw/p9CCPPtrKF7/4L/r7IytRPG4zc2YtTU0pNmwYJFtSFGvTKMcCLyDbU+IPf1jGhAmVaK255pqn6O8vctppWwXQQEYMCZ8t2Bas2WhhMBgMO5O9usihwWB4Y8TjNo5jcf75+/KTn5xEU1OKurokYajwPDV0ntaR66qpKcX06dXcfvtqCoWAn/zkWTwvxLYFUkbZYkuXdpFMOiRrEmQn1eM7Fsq10fPGUfzSqfTUVqCUZu3aAZYu7eaLX/wXV175KIODJQAqUxr1Eid8EMKkseGu3BqDwTACMOLHYBjhNDQkmTKlhq6uKFZneJf2qF/X5MnVCCEIQ82vf93Mgw9uoL09R3d3gd7eIsWiT29vkcWLO5AnziL1X8eRvuQInIsPRx2/DxpN5sR9WbOmj97eIkppPC/k+ec7+PKX/43WmisuKUQtMsoCSJVjfr75CePyMhgMOxcjfgwGA5WVzlD1Z9uWQwLIcSSOYxEEIZ4XtcD461/X43khUgq0jkRKEERFRJNJm9KqbggUuZhDSQtUqFGhoqShvT1HPu9TLAY4jsS2LTZsyLB6dR8nHqr432/kmDY+pL5Gsf+sgDt+kmW0afllMBh2Mibmx2AwMGZMBePGVbB6dR8AsZiF1lGG2MBAkcceayMet1mypBMhotR4pXS5RlA0Rn19gn32qePFjf0M9hcJh+oICdCgNg+UY40kQQCxmEIpTRAouroKTJ8O71igeMeC/G7aBYPBMFIwlh+DYYQzMFDihRe6aG/P4boWWmt8P4waj4YaKSGZtHFdSSbjUSwqEgkbIcSQ8NEa8vmAJ5/cTKalB92XByf6eNGiLIDufBEpBY4jqahw0VrT0ZGjosI1Hd0NBsMuxVh+DIYRzjXXPEUu59PQkGTduoGtLS2I/vY8TRD4SBlVgg5DKJVCEgmbIPCGBFA+7xMEUXq8+tYdcNpcmFgPA3n0LUuQvVnSlTFsW5Z77UXC61OfWkBFhbt7Fm8wGEYkRvwYDCMYrTXNzX10dORpb88RBFHcz7burC0ZX0ptPVYqhcOywgB8Xw2dTyGAvz477HUFVFS4TJtWQ0dHDt9XnHvuDM45Z3hjYoPBYHirMeLHYBjheF7Apk1RQUKthwufbdlyTEqoro7R11caKoIoBEOVoF+NUikKdB4zJk0iYXPxxfN33kIMBoNhBzExPwbDCEYIgVKKTMYfEj/bEz7b4roSIQSOYyGlHMr4ejUcJ+odls16LF3azdq1/TQ1pXAcU8DQYDDseoz4MRhGMFprikWFEBr50vLKr0CppCgWQ2IxyXYKQr/CfaJ0+Pr6JPvt18C0abW0tAzw7W8/9iZmbzAYDG8MI34MhhHMxo0ZpBRUVsaRUmBZ4mWCRr7kU0JryOV8lIJRoxI7JICCQBOGmilTaoaO2bZk1apeenoKO2ElBoPBsOMY8WMwjGBKpQAhBA0NUWf1SAAN/1jYnhtMCAjDkAkTqkgm7XL6uvOqQkgpTUtL37BjYajIZr03vQ6DwWB4PRjxYzCMYKZMqaauLo5SmsrKGJWVLsmkjRBglcNxthU/W3p5aQ3FomLVql4AUimLqqo4dXVxpNxqLRIi+lNZ6eI4ku7uArocWCTyA9S6JcaPTe3iVRsMhpGOET8GwwjGsiSf+MSB1NcnCIJwqKqzZUG4nX6i28YFCRG5s2pq4iglCIKQWMxm0qQqYjFrqBJ0Ou3gOLJcNVqT6+7Db34eZ9MKPlNzB6nrL0FsXrMLV20wGEY6JtXdYBjhHHhgE3/72xlccsndLFnSQbEYEATRa0JsyQiLzD9KabSOLDvxuE2hEJDP+yilSSZt9tsvqtRcUeGwYUOGMNSUSiGFQoCUgkmTKvnIhCWkJxY5YcoAaVdBwSN2yw8ofvyn7HAEtcFgMLwJjOXHYDCQSDhceOFcKitjpFIOsFWHSLn131tS2i1LUigElEohvh+1wRgYKNHeniUIFLlcgGUJ8nkfrTW2LUkkbLTnMdbq4cxZfZHwKd9I9rUjetp28aoNBsNIxVh+DAYDAHfdtZYxY9Ikkw49PYWhmj9brD4A6bRDsRgMq+68JSYoalCaZ9OmLLmcjxAay4oCqGtr49TWJmiqUvyzpYFjZ7xE6AgJ0tT8MRgMuwYjfgwGA319RZYt66a/v0RVVWT9KRSCbTq4RyInFrPI57efnaU19PYWCUMFRIHRSkVWo9raBKNHpwGNX0oxVEoaQCtUw3h07ehdtl6DwTCyMW4vg2GEc9ddLVx00Z00N/fR3NzLsmXd1NYmSCRspJTEYja2LZg1q5ZZs+qwrK1xOVJGWWFRewtNECiklKTTDq5rAZowVHR3R7V8SiXFnDOPR1XWQ+BDGKBqx1J6z5d30+oNBsNIxFh+DIYRTC7nc911z7B6dR++rwgCTV9fkSBQTJ1ajZSSd7xjHM8/34XWsHx5N2HIkFUHQAiJbWvq6hLYtkWxGJRT5QVhqCiVQvr6Crz4YhejR6c59KR5FCe/A9G3GaSNrm7cvZtgMBhGHMbyYzCMYBYvbmfVql6U0kgZ1fGxbUk+H3DEEeO58soj+PKXD+e///swNm3KUiqF5To+AseRSCmwbUF1dZxvfvMdzJ5dB2zxagnicWfo32GoyeV8vvjFh2hZ24+uHWOEj8Fg2C0Yy4/BMILJ5Xx8XyGlJpcL0DpqQ6EU/OUvy3n++U5GjUrz3e++gwULmli+vIdcroDnCbJZD88LqahwufHG05g3bxSWJenpKdDVVSAIFNlsESkFyaRDGCo6OnLk8z6/+91Svv71I3b38g0GwwjFiB+DYQQzd24jiYRNf38xqryMQOvICuR5inw+oL+/yI9+9AwVFS719QmSSU0qlSIMFf39JaZPr2HixCoAzjxzOkop7r13HevXD9LXVyCVcodimy1LkM36tLT077Y1GwwGg3F7GQwjmNGj0xx77PhhxQyFAMexsG1Bf39kuVm3rp+zz55JGEZ57b6vWLasm5aWPlpa+rnkkru58calCCE44YTJQw1ShYhq/eTzwVBK/Lap8waDwbA7MOLHYBjhfPe7xzB+fCWuayGlQEpJPG6Rz/v09RVpbc1g25KDDhrNJz95APX1MTZuHEBKwaxZ9VRXxwG4+eaVbNw4yPXXP0NnZ5502iWddpFS4HkhQaCGqkNfeun+u3fRBoNhRGPcXgaDgfr6BJ2dORzHIZPxGBjwcF2JENDWliGZtPH9kKOOmkBVVY4f/nANvq+GjSGl4LbbVtPSMoBlSbSOihxGcUSKQiFg1Kgk73rXNA49dNxuWqnBYDAY8WMwjEi01jz1VDv33LOWp5/eTHt7DiEkhYI/5JYSQiOlpKEhRjJpc/XVT7FmTT+trT2sX1+gujrOuHEVQ2MqBcmkXa7vA2vW9ON5imTSplgMicdtpk6t4ZvfPHK3rNlgMBi2YMSPwTAC+dnPlvD/27v34Kjqu4/j77O33MNCCOESYkoSghJQoU2oEJkqQlNQEATtY9FBMAQsbQEpl9FxRDrIZahVeSxpKtIWLBSxNGUE7QitlEvwClXkiXITCwRCFhJCSHb3PH8srF2hGiHJkj2f14wz8Du/xe93Azuf/Z3fOWfTpv24XHb27Kng9Ol6YmIc2O02/H7fhZsWcmFTcx3V1ecpL6+iV68Owau3Dh8+TUODj/T0NhiGgdNpMHx4d6KiHLz88l7OnDl/4bJ4Ow6HnZycZEzTZP9+D1lZ7cL9FoiIhWnPj4jFnDp1js2bDxMV5cAwDOrr/Rc2Jnuprw/cx8c0oaHBxOM5z5kz9Zw4Ebh8vbq6/sLvz1Jf7+XgwdO8995xjh07i8Nh54kn3qKmpp4BA7oAgT/H6bTTrZubqCg7Xq9JRUVtmN8BEbE6rfyIWMzevZXU1jYEn94eHR04LRW40aEBGEDg1Ffgnj++4K937vz8wqZl48KVXyYeTx0eTx0QeLjpkSMHyM5uy4ABqZw5U3/hyq/Ate7x8U5ycpJbumURkRBa+RGxmK5dE3A6v/inn5DgIi7OiWEENiebZuil6Bef7g7g8wVOh3m9ZnC8ocHE6zXZu/ckhw+fxum08fHHVQwalB4MSX6/SUODnzvvzKRNm6iWbFdE5BJa+RGxmLS0NuTktOfDDytxOm1cd10i1dX1OJ0GDQ1+zCu8DY/fb3Lw4BliYpzExTlJSYnjf/93MK+8so/z533ceWcG3bq1bdpmRESugMKPiAU9/nh/li/fw549J7DZDNLT27Bz57/5v/+ruqKbEJpmYDXI5/Py4YcnsdsN/v3vMwwalE5R0c3N0IGIyJXTaS8Ri2lo8PH66wc4caKWG29MZu7cAdjtNhITo4N7c66UaQauEDNNk6efLuOll/Y0UdUiIk1HKz8iFuL1+pk58+98+mkV0dEO3n3Xz5YtnxEb6+DgwdPYbDbAf9nXGgaNOiUWG+vE5bLR0ODjtdf2c8892cTHu5q2ERGRq6CVHxELeeONA+zfHwg+AE5nIKQcOVLDqbNxNJixYDgv+9rGBB+Hw8DptGMYgSu8amrq+fzzmqZsQUTkqrX68HP8+HEKCwvp3r07nTp1on///qxZsyZkjsfjobCwkLS0NNLS0igsLMTj8YSnYJEwev/9CqKiAsHH46lj9+4Kdu06xr/KfdhiU8HuBNP7jf/cL57abgveILFNGxdxcS46dIhtyhZERK5aqz/tVVRURFVVFatWraJ9+/aUlpYyceJEunTpQv/+/QGYMGECR44cYe3atQD85Cc/YeLEiaxevTqcpYu0uI4d46iv93H69Hk++qgSny9wistrnsYen4AjKpEGvx+z4SwANsPEbvNd8hyvLzOMwKMtHA4bNptBYqKLzp0TyM3tSNu20c3el4jIN9Hqw09ZWRkLFizg29/+NgBTpkxh2bJlvPvuu/Tv3599+/bxt7/9jY0bN5KbmwvAL3/5SwoKCigvLycrKyuc5Yu0qFGjsnnzzUO8995xfD5/cB+PzRWP2VCDt+EMeOsDg0Zg/89XBR+n04bf7yc+PopJk26mTRsXe/dWYrfbyM3txL33Xt9yzYmINFKrDz/9+vXjz3/+Mz/4wQ9wu9289tprVFZWMnDgQCAQjuLj48nLywt5TVxcHDt37lT4EUtJTIxi/vyB9O//B2w2A8MAh8OG1+6gwQv++loM24VzWH4ffr7Y6GOzGZdcBu/1+nE4bLRpE8WuXUdJSorhuecGBU+tiYhci1r9J9Ty5csZP3483bp1w+FwEBUVRUlJCb179wagoqKCpKSkkEt4DcOgffv2VFRU/Nc/t7y8vFH//8bOi1RW7r81956c7OLQoRp8Pi48z+sYPjOw0dn0h670GAbYbF/s6/nSYbxeP5WVZ/F6G2ho8HP//WuZOTMHtzty7+Tcmn/2V0u9tz76kn+pazL8zJs3j8WLF3/lnNLSUvLz85k3bx6VlZWsX7+edu3asWHDBiZNmsSGDRvo1avXFdfQmL8sVj9tZuX+W3PvpmkSE7OThoaakDHMepwuGz6vGRJwTBOiouzU1QWe8XUxBF18vIVhwNmzPux2Hy6Xjf3763jhhcMsXTqYmJhr8iPmqrTmn/3VUu/W7D0SXZOfTJMmTWLMmDFfOSc1NZUDBw5QXFzMW2+9FQw6vXr1Yvv27RQXF/Pcc8/RoUMHKisrMU0zuPpjmiYnT56kQ4cOzd6LyLVmz54TVFTUEh/v5Nw574XneYHDAX6f/8KDS0NXeM6f92G3G/j9EB0Npmmjvt4ffK1hwLlzXhwOF253FFVVdZSWljNmjPb8iMi155oMP0lJSSQlJX3tvNraWgDsdnvIuN1ux3/hkzs3N5eamhrKysqC+37Kyso4e/ZsyD4gEat4442DGAbExDgwDKir82GagYeTAsFL1f+T3x8Yj462c/68l6goG/HxNmprvTQ0mBdWg0xiYx0kJcVgsxkcOHC6xXsTEWmMVn2fn+7du9OtWzemT5/OO++8w4EDB3juuefYvHkzw4YNAyA7O5tBgwYxdepUysrKKCsrY+rUqQwZMkRLmGJJ7dpFExfnwus1g8HnPzcyX+5mhhdPcdXX+zBNqK1toK7OT3S0A7s9sHE6IcHF9de3x2YzOH/ex003aWVVRK5NrTr8OJ1O/vSnP9G+fXvuu+8++vfvzx//+EeWLl1KQUFBcF5JSQk5OTmMGjWKUaNGkZOTw7Jly8JYuUj43H13dzIy3Jw7F9ig7P3SHp/LcTptwYeXxsTYcbnsgInX6yctLYG2baPJzGwbDD7p6Yncfnt6S7QjIvKNXZOnvb6JjIwMfv/733/lHLfbTXFxcQtVJHJtc7ujyc9PZfv2z6mr8zXqmV2BvT2B01s+n0mvXh04c+Y81dX19O7dgQkTevP228fxeOrIyUlm+PAsHI5W/d1KRCJYqw8/IvLN7dlzEpvt4gUAjXvNxXsCJSQ4SUqKISkphro6L2PG9CA/P438/LRmrFhEpOnoq5mIxbzzzjG2bDlEdXV9o+ZHR9txu6NxOm1ERdlJS4sLHouNdXDrrQo9ItK6KPyIWMzKlR/S0PD1+3zgi2d2+f1+4uKcxMW5MAyD2toGHA4bDz3Um4QEV/MXLSLShHTaS8RiTpw4R03N+UbN/da3EklJiae+3kdCQhS5uR3p3dtFu3YduemmFGJjnc1crYhI01P4EbGYU6fO4fN99UYfmy3wX02Nl65d7SQmRuH1+hk5Mhu7/RRZWaktVK2ISNPTaS8Ri0lJicVuN/7rccOA+HgXqamJGAYcOnSaxEQXEyfeRI8eX3/zURGRa51WfkQsJicnmfLyU+zZU3nJMZstcLPC5ORYMjLaYpomUVEOiou/H/JwYBGR1kwrPyIW8z//cwNduiSSlpaA3f7FKS6Xy0Z0tAOHw0bbttEAGIZBYqJLwUdEIorCj4jFdOwYz1NP5VNQ0I3MzLZERzto1y6GjAw3MTEOYmKctGsXCD9er5877kgPb8EiIk1Mp71ELCgrqx0LFnwPgM8+O8Pq1Xs5c6ae5OQY9u/3UFFxjvh4J4MGpTN8uJ6BJyKRReFHxOK6dk3k0Ufzwl2GiEiL0WkvERERsRSFHxEREbEUhR8RERGxFIUfERERsRSFHxEREbEUhR8RERGxFIUfERERsRSFHxEREbEUhR8RERGxFIUfERERsRSFHxEREbEUhR8RERGxFIUfERERsRTD4/GY4S5CREREpKVo5UdEREQsReFHRERELEXhR0RERCxF4UdEREQsReFHRERELEXh5wq98847jBgxgi5dupCamsrgwYOprKwMHvd4PBQWFpKWlkZaWhqFhYV4PJ7wFdzETNPknnvuwe12s379+pBjkdp7VVUVM2bM4Dvf+Q4dO3akZ8+eTJs2jVOnToXMi9T+AUpKSujduzcpKSkMHDiQbdu2hbukJrdkyRK+973v0bVrVzIyMrj33nv56KOPQuaYpsn8+fPp0aMHHTt2ZOjQoezduzdMFTefJUuW4Ha7mTFjRnAs0ns/duwYRUVFZGRkkJKSQl5eHlu3bg0ej/T+rULh5wq8/fbb3H333QwYMIA33niDLVu28OMf/xiHwxGcM2HCBHbv3s3atWtZu3Ytu3fvZuLEiWGsumk9//zz2GyX/+sTqb0fPXqUo0eP8uSTT7Jt2zaWLVvGtm3bGD9+fMi8SO1/3bp1zJo1i+nTp/OPf/yD3NxcRo8ezWeffRbu0prU1q1bGT9+PJs2beIvf/kLDoeDESNGUFVVFZzzq1/9iqVLl7JgwQLefPNNkpOTufvuu6murg5j5U1r165dvPTSS/Ts2TNkPJJ793g8DBkyBNM0WbNmDTt37mThwoUkJycH50Ry/1ai+/xcgcGDB5Ofn8/jjz9+2eP79u0jLy+PjRs30q9fPwC2b99OQUEBu3btIisrqyXLbXLvvvsuY8eOZcuWLWRlZbFixQqGDx8ORH7vX/b6669z7733cujQIRITEyO6/9tvv52ePXvy7LPPBsf69OnD8OHDeeKJJ8JYWfOqqakhLS2NlStXUlBQgGma9OjRg4cffphHH30UgHPnzpGVlcVTTz3FuHHjwlzx1Tt9+jQDBw7k2WefZcGCBdxwww0sWrQo4nufO3cu//znP9m0adNlj0d6/1ailZ9v6MSJE5SVlZGSksL3v/99MjMzKSgo4O9//3twTllZGfHx8eTl5QXH+vXrR1xcHDt37gxH2U2murqaCRMm8Mwzz4R8G7ooknu/nOrqaqKiooiNjQUit//6+nref/99brvttpDx2267rVX31Rg1NTX4/X7cbjcAhw4d4vjx4yHvRUxMDLfcckvEvBc/+9nPGD58OLfeemvIeKT3vmHDBvr27cu4cePIzMxkwIABFBcXY5qBNYJI799KFH6+oYMHDwIwf/587r//fl555RW++93vMnLkSPbs2QNARUUFSUlJGIYRfJ1hGLRv356KiopwlN1kpk2bxu23384dd9xx2eOR3PuXeTwefvGLX/DAAw8ET3lGav+VlZX4fL5LAm9ycnKr7qsxZs2aRa9evcjNzQXg+PHjABH7XqxYsYL9+/fz2GOPXXIs0ns/ePAgv/3tb0lPT+eVV16hqKiIJ598kt/85jdA5PdvJY6vn2IN8+bNY/HixV85p7S0FJfLBcC4ceMYO3YsADfeeCNvvfUWy5cvZ8mSJc1ea1NrbO+ff/45//rXv9i8eXMLVdYyGtt/fn5+8Pc1NTX88Ic/pFOnTsydO7e5S5QwmTNnDjt27GDjxo3Y7fZwl9PsysvLmTt3Lhs3bsTpdIa7nBbn9/u5+eabg6dxb7zxRvbv309JSQmFhYVhrk6aksLPBZMmTWLMmDFfOSc1NTWY7rOzs0OOZWdnc+TIEQA6dOhAZWUlpmkGVwBM0+TkyZN06NChGaq/Oo3tfdWqVXz88cd06dIl5Ni4cePIzc1l48aNra53aHz/F9XU1DB69GgAVq9eTXR0dPBYa+y/MZKSkrDb7Zw4cSJk/MSJE626r68ye/Zs1q1bR2lpKenp6cHxlJQUINB7165dg+OR8F6UlZVRWVkZ3K8G4PP52LZtGy+++CI7duwAIrN3CPxsv/zZ3r179+BneyT/7K1G4eeCpKQkkpKSvnbeddddR6dOnSgvLw8Z//TTT7nhhhsAyM3NpaamhrKysuDej7KyMs6ePRuyF+Ra0djeH3/8caZMmRIydsstt/DUU08xdOhQoPX1Do3vHwJ7fEaPHo1pmqxdu5b4+PiQ462x/8ZwuVzcdNNNbN68mREjRgTHN2/ezF133RW+wprJzJkzefXVVyktLaV79+4hx6677jpSUlLYvHkzffr0AaCuro7t27e3+lXAoUOHcvPNN4eMPfLII2RkZDBt2jQyMzMjtncI7M/75JNPQsY++eSTYNCJ5J+91Sj8fEOGYTBlyhSefvppcnJy6N27N6+++iq7du1i4cKFQGAVaNCgQUydOpVnnnkGgKlTpzJkyJBWfbVP586d6dy58yXjqampwW/Gkdo7BILPyJEjqa6uZuXKldTW1lJbWwtA27ZtcblcEd3/I488wsSJE+nbty95eXm8+OKLHDt2LOKucHn00UdZvXo1f/jDH3C73cF9HnFxccTHx2MYBpMmTWLJkiVkZWWRmZnJ4sWLiYuL45577glz9VfH7XYHN3ZfFBsbS9u2bYNf7iK1d4DJkyczePBgFi9ezMiRI9m9ezfFxcXBK3sj+WdvNQo/V2Dy5MnU19fz2GOPcerUKXr06MHatWvp1atXcE5JSQk///nPGTVqFAAFBQXBcBTpIrX3999/n127dgHQt2/fkGP/uScoUvsfOXIkp06dYtGiRRw/fpzrr7+eNWvWkJaWFu7SmlRJSQlA8PYNF82cOZPZs2cD8NOf/pRz584xY8YMPB4Pffv2Zd26dSQkJLR4vS0tknvv06cPK1euZO7cuSxatIjU1FTmzJnDhAkTgnMiuX8r0X1+RERExFJ0qbuIiIhYisKPiIiIWIrCj4iIiFiKwo+IiIhYisKPiIiIWIrCj4iIiFiKwo+IiIhYisKPiIiIWIrCj4g0m8WLF+N2u/nggw8uOTZ79mzcbjcrV64MQ2UiYmUKPyLSbB566CFiY2NZunRpyPjLL7/MCy+8QFFREffff3+YqhMRq9LjLUSkWU2fPp3f/e53fPDBB3Tu3Jn33nuPgoIC8vLyWLduHXa7PdwliojFaOVHRJrV5MmT8Xq9FBcXU1FRwY9+9CNSUlJYvny5go+IhIVWfkSk2d13333s2LGD7OxsPvzwQzZt2kTPnj3DXZaIWJTCj4g0u61btzJs2DAMw2DFihXcdddd4S5JRCxMp71EpNnt27cPgI4dOzJs2LAwVyMiVqfwIyLNauvWrcyaNYuMjAyOHj1KaWnpZecVFxeTn59PUlIS8+fPb+EqRcRKFH5EpNkcOnSIBx98kOzsbN58801SU1Mvuez9ok6dOjFnzhytDIlIs1P4EZFmcfbs2eA9fFatWkWbNm2YOHEiZWVlvP3225fMv/POOykoKCAxMbGlSxURi1H4EZEmZ5omkydP5uOPP2bFihWkpaUB8OCDD5KYmMjzzz8f5gpFxMoUfkSkyS1cuJD169czf/58BgwYEBxPTExk7NixlJaWcvjw4TBWKCJWpvAjIk3qr3/9K08//TQPPPAADz/88CXHi4qKMAyDX//612GoTkQEHOEuQEQiy7Bhw6iqqvqvx7t27crJkydbsCIRkVBa+RGRa4LX66Wurg6fz4fP5wv+WkSkqekOzyJyTZg/fz4LFiwIGVu6dKme+i4iTU7hR0RERCxFp71ERETEUhR+RERExFIUfkRERMRSFH5ERETEUhR+RERExFIUfkRERMRSFH5ERETEUhR+RERExFL+H6LfM1tzioPlAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dim_reduction_plot(TSNE, emb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the data can be largely seperated into 7 clusters, the orange cluster might be more difficult to identify as it's less distinct from the others."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}