MLP.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Multiclass Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Short story </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Most of the times we encounter problems that are not binary. E.x. imagine you're given a photo and your task it to predict if it contains a dog, cat, horse, duck etc.\n",
    "Shortly you would like to assign each example to $1$ class of $k$ classes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Naive way </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given k klasses the most naive way is to train k binary classifiers. We could then predict that an example x belongs to the class  𝑖={1,2,...,𝑘}  based on which classifier outputted the highest probability"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Smarter Way"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is a smarter way to go about this. We could force the output to be a discrete probability distribution over k classes.\n",
    "​\n",
    "To be a valid distribution the values of the output units\n",
    "​\n",
    "1) must be non-negative\n",
    "​\n",
    "2) must sum to 1\n",
    "​\n",
    "We can achieve this by applying the softmax function\n",
    "​\n",
    "$$softmax(z) = \\frac{ e^{z} }{ \\sum_{i=1}^{k} e^{ z_{i} } }$$\n",
    "​\n",
    "Each unit is exponetiated, therefore we force each unit to be positive\n",
    "​\n",
    "afterwards we have to normalize everything to 0-1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> One hot vectors </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A binary vector of length K, where each value is 0 except for the class it belongs, the value is 1."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$t_{nk} \\in \\{ 0,1 \\}, \\sum_{k=1}^{K} t_{nk} = 1$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Example of k=3 classes\n",
    "\n",
    "\\begin{equation}\n",
    "\\begin{bmatrix}\n",
    "    1\\\\\n",
    "    0\\\\\n",
    "    0\n",
    "\\end{bmatrix},\n",
    "\\begin{bmatrix}\n",
    "    0\\\\\n",
    "    1\\\\\n",
    "    0\n",
    "\\end{bmatrix},\n",
    "\\begin{bmatrix}\n",
    "    0\\\\\n",
    "    0\\\\\n",
    "    1\n",
    "\\end{bmatrix}\n",
    "\\end{equation}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Likelihood\n",
    "$$p( T | W ) = \\sum_{n=1}^{N} \\sum_{k=1}^{K} y_{nk}^{t_{nk}}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The cost function (logLikelihood plus reguralization term) we want to maximize for the problem of classifying N number of data in K categories/classes is:\n",
    "\n",
    "$$\n",
    "E(W) = \\sum_{n=1}^N \\sum_{k=1}^K t_{nk} \\log y_{nk}   -  \\frac{\\lambda}{2} \\sum_{k=1}^K ||\\mathbf{w_k}||^2, \n",
    "$$\n",
    "\n",
    "where $y_{nk}$ is the softmax function defined as:\n",
    "\n",
    "$$y_{nk} = \\frac{e^{\\mathbf{w}_k^T \\mathbf{x}_n}}{\\sum_{j=1}^K e^{\\mathbf{w}_j^T \\mathbf{x}_n}}$$\n",
    "$W$ is a $K \\times (D+1)$ matrix where each line represents the vector $\\mathbf{w}_k$.\n",
    "\n",
    "\n",
    "The cost function can be simplified in the following form:\n",
    "\n",
    "\n",
    "\n",
    "$$\n",
    "E(W) = \\sum_{n=1}^N \\left[ \\left( \\sum_{k=1}^K t_{nk} \\mathbf{w}_k^T \\mathbf{x}_n \\right) - \\log \\left( \\sum_{j=1}^K e^{\\mathbf{w}_j^T \\mathbf{x}_n} \\right) \\right]   -  \\frac{\\lambda}{2} \\sum_{k=1}^K ||\\mathbf{w}_k||^2, \n",
    "$$\n",
    "\n",
    "\n",
    "\n",
    "In the above formula we have used the fact that $\\sum_{k=1}^K t_{nk} = 1$. \n",
    "\n",
    "The partial derrivatives of this function are given by the following $K \\times (D+1)$ matrix:\n",
    "\n",
    "$$\n",
    "(T - S)^Τ X - \\lambda W,\n",
    "$$\n",
    "\n",
    "where $T$ is an $N \\times K$ matrix with the truth values of the training data, such that $[T]_{nk} = t_{nk}$, $S$ is the corresponding $N \\times K$ matrix that holds the softmax probabilities such that $[S]_{nk} = y_{nk}$ and $X$ is the $N \\times (D + 1)$ matrix of the input data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import re\n",
    "from random import shuffle\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "import _pickle as cPickle\n",
    "import random\n",
    "import os\n",
    "from pprint import pprint\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "random.seed(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Softmax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#use by default ax=1, when the array is 2D\n",
    "#use ax=0 when the array is 1D\n",
    "def softmax( x, ax=1 ):\n",
    "    m = np.max( x, axis=ax, keepdims=True )#max per row\n",
    "    p = np.exp( x - m )\n",
    "    return ( p / np.sum(p,axis=ax,keepdims=True) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2D case"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "z = np.array( [ [10,20,30,40], [20,50,45,45], [983,39,57,752], [574,575,597,525] ] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 10,  20,  30,  40],\n",
       "       [ 20,  50,  45,  45],\n",
       "       [983,  39,  57, 752],\n",
       "       [574, 575, 597, 525]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9.357198e-14</td>\n",
       "      <td>2.061060e-09</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>9.999546e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9.233197e-14</td>\n",
       "      <td>9.867033e-01</td>\n",
       "      <td>0.006648</td>\n",
       "      <td>6.648354e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.764032e-101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.026188e-10</td>\n",
       "      <td>2.789468e-10</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.380186e-32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              0             1         2              3\n",
       "0  9.357198e-14  2.061060e-09  0.000045   9.999546e-01\n",
       "1  9.233197e-14  9.867033e-01  0.006648   6.648354e-03\n",
       "2  1.000000e+00  0.000000e+00  0.000000  4.764032e-101\n",
       "3  1.026188e-10  2.789468e-10  1.000000   5.380186e-32"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame( softmax( z ) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1D case\n",
    "\n",
    "Let's take an example where all values are equal. As expected everything is 1 / |z|."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.25, 0.25, 0.25, 0.25])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z = np.array( [1,1,1,1] )\n",
    "softmax( z, ax=0 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Activation Functions (Tanh: 'tanh' , Cos: 'cos' , Softplus: 'softplus')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def softplus(X):\n",
    "    return np.log(1 + np.exp(- np.abs(X))) + np.maximum(X, 0)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tanh(X):\n",
    "    return np.tanh(X)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cos(X):\n",
    "    return np.cos(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def derivative_softplus(X):\n",
    "    return 1 / (1 + np.exp(- X))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def derivative_tanh(X):\n",
    "    return 1 - np.power(np.tanh(X), 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def derivative_cos(X):\n",
    "    return -(np.sin(X))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def activation_function(X):\n",
    "    return non_linearity(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def derivative_activation_function(X):\n",
    "    return derivative_function(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MLP "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Forward pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def forward(X, t, weights_1, weights_2):\n",
    "    # pass the input through the hidden layer\n",
    "   \n",
    "    \n",
    "    a = activation_function(X.dot(weights_1.T))\n",
    "    hidden = np.ones((a.shape[0], a.shape[1] + 1), dtype=float)\n",
    "    hidden[:, 1:] = a\n",
    "    \n",
    "    y = hidden.dot(weights_2.T)\n",
    "\n",
    "    y = softmax(y)\n",
    "\n",
    "    loss = get_loss(t, y, weights_1, weights_2)\n",
    "\n",
    "    return loss, y, hidden\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def forward_keep_activations(X, weights_1, weights_2):\n",
    "    # pass the input through the hidden layer\n",
    "    #hidden_layer_representation = get_hidden_layer_representation(input, weights_1)\n",
    "    \n",
    "    a = activation_function(X.dot(weights_1.T))\n",
    "    hidden = np.ones((a.shape[0], a.shape[1] + 1), dtype=float)\n",
    "    hidden[:, 1:] = a\n",
    "    \n",
    "    # calculate dot product between weights_2 and hidden layer representation\n",
    "    # hidden layer_representation : B x (M + 1)\n",
    "    # weights_2 : K x (M + 1)\n",
    "    y = hidden.dot(weights_2.T)\n",
    "\n",
    "    # output now has dimensions: batch_size x K\n",
    "    y = softmax(y)\n",
    "\n",
    "    return y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Backward pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def backward(t, y, hidden_layer_rep, weights_1, weights_2, X):\n",
    "    # gradients for the weights between the hidden layer and the softmax layer\n",
    "    grad_weights_2 = (t - y).T.dot(hidden_layer_rep) - lamda * weights_2\n",
    "\n",
    "    # remove the bias from weights_2\n",
    "    weights_2_temp = np.copy(weights_2[:, 1:])\n",
    "\n",
    "    derivative_result = derivative_activation_function(X.dot(weights_1.T))\n",
    "\n",
    "    grad_weights_1_tmp = (t - y).dot(weights_2_temp) * derivative_result\n",
    "\n",
    "    grad_weights_1 = grad_weights_1_tmp.T.dot(X) - lamda * weights_1\n",
    "\n",
    "    return grad_weights_1, grad_weights_2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cost Function"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the cost function\n",
    "\n",
    "$$\n",
    "E(W) = \\sum_{n=1}^N \\left[ \\left( \\sum_{k=1}^K t_{nk} \\mathbf{w}_k^T \\mathbf{x}_n \\right) - \\log \\left( \\sum_{j=1}^K e^{\\mathbf{w}_j^T \\mathbf{x}_n} \\right) \\right]   -  \\frac{\\lambda}{2} \\sum_{k=1}^K ||\\mathbf{w}_k||^2, \n",
    "$$\n",
    "\n",
    "\n",
    "we use the logsumexp trick, where m is the maximum element\n",
    "\\begin{align} \n",
    "\\log \\sum_{i=1}^{n} e^{\\mathbf{w}_j^T \\mathbf{x}_n} &= \\log \\Bigr( \\sum_{i=1}^{n} e^{\\mathbf{w}_j^T \\mathbf{x}_n +m -m}\\Bigl) \\\\ \n",
    "&= \\log \\Bigr( \\sum_{i=1}^{n} e^m e^{\\mathbf{w}_j^T \\mathbf{x}_n-m} ) \\Bigl) \\\\ \n",
    "&= \\log \\Bigr( e^m \\sum_{i=1}^{n} e^{\\mathbf{w}_j^T \\mathbf{x}_n-m} ) \\Bigl) \\\\ \n",
    "&= \\log \\ e^m + \\log \\Bigr( \\sum_{i=1}^{n} e^{\\mathbf{w}_j^T \\mathbf{x}_n-m} ) \\Bigl) \\\\ \n",
    "&= m + \\log \\Bigr( \\sum_{i=1} e^{\\mathbf{w}_j^T \\mathbf{x}_n-m}  \\Bigl) \n",
    "\\end{align}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_loss(t, y, weights1, weights2):\n",
    "    regularization = (lamda / 2) * (np.sum(np.square(weights1)) + np.sum(np.square(weights2)))\n",
    "    y = np.log(y)\n",
    "    y = t * y\n",
    "    y = np.sum(y, 1)\n",
    "    y = np.sum(y, 0)\n",
    "    loss = y - regularization\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def batch_generate(data, data_labels):\n",
    "    for idx in range(0, data.shape[0], batch_size):\n",
    "        input_b = data[idx:idx + batch_size]\n",
    "        labels_b = data_labels[idx:idx + batch_size]\n",
    "        yield input_b, labels_b"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Weight_2 (Output layer weights) initialized as zeros and the the Weight_1 (Hidden layer weights) randomly with normal distribution(m = 0 and s = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train():\n",
    "    \n",
    "    center = 0\n",
    "    s = np.sqrt(1 / (D + 1))\n",
    "\n",
    "    # Initialize the weights\n",
    "    weights_2 = np.zeros((K, M + 1))\n",
    "    weights_1 = np.random.normal(center, s, (M, D + 1))\n",
    "\n",
    "    for _ in tqdm(range(epochs)):\n",
    "        epoch_loss = []\n",
    "        iterator = batch_generate(X_train, y_train)\n",
    "        for batch in iterator:\n",
    "            \n",
    "            input_data = batch[0]\n",
    "            labels = batch[1]\n",
    "\n",
    "            ### forward pass \n",
    "            loss, output, hidden_layer_output = forward(input_data, labels, weights_1, weights_2)\n",
    "            epoch_loss.append(loss)\n",
    "\n",
    "            ### backward pass\n",
    "            grad_weights_1, grad_weights_2 = backward(labels,output,hidden_layer_output,weights_1, weights_2,input_data)\n",
    "\n",
    "            ### \n",
    "            weights_1 += lr * grad_weights_1\n",
    "            weights_2 += lr * grad_weights_2\n",
    "\n",
    "        print(\"Epoch average loss: \", sum(epoch_loss) / total_batches)\n",
    "    \n",
    "    return weights_1, weights_2 ,epoch_loss\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test_batch():\n",
    "    # we do forward passes with the learned weights and we predict\n",
    "    iterator = tqdm(batch_yielder(X_test, y_test))\n",
    "    true = 0\n",
    "    total = 0\n",
    "    for batch in iterator:\n",
    "        input_data = batch[0]\n",
    "        labels = batch[1]\n",
    "\n",
    "        ### forward\n",
    "        output = forward_keep_activations(input_data, learned_weights_1, learned_weights_2)\n",
    "        predictions = np.argmax(output, 1)\n",
    "        labels = np.argmax(labels, 1)\n",
    "\n",
    "        for idx in range(labels.shape[0]):\n",
    "            if predictions[idx] == labels[idx]:\n",
    "                true += 1\n",
    "            total += 1\n",
    "\n",
    "    print(true)\n",
    "    print(total)\n",
    "    print(\"Accuracy on Test : {}/{} ... {}\".format(true, total, float(true / total)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict():\n",
    "    output = forward_keep_activations(X_test, learned_weights_1, learned_weights_2)\n",
    "    pred= np.argmax(output, 1)\n",
    "    accuracy = np.mean(pred == np.argmax(y_test, 1))\n",
    "    print('Accuracy on Test Data  : {}'.format(accuracy))\n",
    "    print(\"Writing to file: \")\n",
    "    faults = np.where(np.not_equal(np.argmax(y_test,1),pred))[0]\n",
    "    # plot n misclassified examples from the Test set\n",
    "    n = 25\n",
    "    samples = np.random.choice(faults, n)\n",
    "    sqrt_n = int( n ** 0.5 )\n",
    "\n",
    "    plt.figure( figsize=(11,13) )\n",
    "    \n",
    "    if (dataset=='MNIST'):\n",
    "        cnt = 0\n",
    "        for i in samples:\n",
    "            cnt += 1\n",
    "            plt.subplot( sqrt_n, sqrt_n, cnt )\n",
    "            plt.subplot( sqrt_n, sqrt_n, cnt ).axis('off')\n",
    "            plt.imshow( X_test[i,1:].reshape(28,28)*255, cmap='gray' )\n",
    "            plt.title(\"True: \"+str(np.argmax(y_test,1)[i])+ \"\\n Predicted: \"+ str(pred[i]))\n",
    "\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Mnist Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the data folder there is the dataset of mnist. Mnists consists of $28x28$ grayscale images. In total there are 10 training files\n",
    "train0.txt, train1.txt, ..., train9.txt where each rows of train$k$.txt corresponds to an example that belongs to the class $k$.\n",
    "\n",
    "The testing data follows the same format.\n",
    "\n",
    "In total we have $6*10^5$ training examples and $10^3$ testing examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_mnist_data():\n",
    "    train_files = ['train{}.txt'.format(i) for i in range(0, 10)]\n",
    "    test_files = ['test{}.txt'.format(i) for i in range(0, 10)]\n",
    "    print('Total train files: {}'.format(len(train_files)) + '\\n' + 'Total test files: {}'.format(len(test_files)))\n",
    "\n",
    "    data = []\n",
    "    for idx, file in enumerate(train_files):\n",
    "        with open(os.path.join(path, file), 'r') as train_file:\n",
    "            data += train_file.readlines()\n",
    "    train_data_tmp = np.array(np.array([[j for j in i.split(\" \")] for i in data], dtype='int') / 255)\n",
    "    print('Total training images with shape: ', train_data_tmp.shape)\n",
    "\n",
    "    data = []\n",
    "    for idx, file in enumerate(test_files):\n",
    "        with open(os.path.join(path, file), 'r') as test_file:\n",
    "            data += test_file.readlines()\n",
    "    test_data_tmp = np.array(np.array([[j for j in i.split(\" \")] for i in data], dtype='int') / 255)\n",
    "    print('Total testing images with shape: ', test_data_tmp.shape)\n",
    "\n",
    "    labels_data = []\n",
    "    for idx, file in enumerate(train_files):\n",
    "        with open(os.path.join(path, file), 'r') as train_file:\n",
    "            for _ in train_file:\n",
    "                labels_data.append([1 if i == idx else 0 for i in range(0, 10)])\n",
    "\n",
    "    train_labels_tmp = np.array(labels_data, dtype='int')\n",
    "    print('Total training labels: ', train_labels_tmp.shape)\n",
    "\n",
    "    labels_data = []\n",
    "    for idx, file in enumerate(test_files):\n",
    "        with open(os.path.join(path, file), 'r') as test_file:\n",
    "            for _ in test_file:\n",
    "                labels_data.append([1 if i == idx else 0 for i in range(0, 10)])\n",
    "\n",
    "    test_labels_tmp = np.array(labels_data, dtype='int')\n",
    "    print('Total testing labels: ', test_labels_tmp.shape)\n",
    "\n",
    "    return train_data_tmp, train_labels_tmp, test_data_tmp, test_labels_tmp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cifar-10 Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The CIFAR-10 dataset (Canadian Institute For Advanced Research) is a collection of images that are commonly used to train machine learning and computer vision algorithms. It is one of the most widely used datasets for machine learning research.[1][2] The CIFAR-10 dataset contains 60,000 32x32 color images in 10 different classes.[3] The 10 different classes represent airplanes, cars, birds, cats, deer, dogs, frogs, horses, ships, and trucks. There are 6,000 images of each class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_cifar_10_data():\n",
    "    train_batches = ['data_batch_{}'.format(i) for i in range(1, 6)]\n",
    "    data = []\n",
    "    train_labels_tmp = np.zeros((50000, 10))\n",
    "\n",
    "    batch_dicts = []\n",
    "    for batch_name in train_batches:\n",
    "        batch_dicts.append(load_cifar_10_batch(batch_name))\n",
    "\n",
    "    img_idx = 0\n",
    "    for batch_dict in batch_dicts:\n",
    "        for img_data in batch_dict['data']:\n",
    "            data.append(img_data)\n",
    "\n",
    "        for img_label in batch_dict['labels']:\n",
    "            train_labels_tmp[img_idx][img_label] = 1\n",
    "            img_idx += 1\n",
    "\n",
    "    train_data_tmp = np.asarray(data)\n",
    "    train_data_tmp = train_data_tmp / 255\n",
    "\n",
    "    test_batch_dict = load_cifar_10_batch('test_batch', True)\n",
    "    data = []\n",
    "    test_labels_tmp = np.zeros((10000, 10))\n",
    "    for img in test_batch_dict['data']:\n",
    "        data.append(img)\n",
    "\n",
    "    img_idx = 0\n",
    "    for img_label in test_batch_dict['labels']:\n",
    "        test_labels_tmp[img_idx][img_label] = 1\n",
    "        img_idx += 1\n",
    "\n",
    "    test_data_tmp = np.asarray(data)\n",
    "    test_data_tmp = test_data_tmp / 255\n",
    "\n",
    "    print('Total training images with shape: ', train_data_tmp.shape)\n",
    "    print('Total testing images with shape: ', test_data_tmp.shape)\n",
    "    print('Total training labels: ', train_labels_tmp.shape)\n",
    "    print('Total test labels: ', test_labels_tmp.shape)\n",
    "\n",
    "    return train_data_tmp, train_labels_tmp, test_data_tmp, test_labels_tmp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_cifar_10_batch(batch_name, test_bool=False):\n",
    "    #C:\\Users\\giann\\Desktop\\ergasia1\\\n",
    "    data_batch = cPickle.load(open(r'data\\cifar-10-batches-py/{}'.format(batch_name), 'rb'),\n",
    "                              encoding='latin1')\n",
    "\n",
    "    return data_batch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing the datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_mnist(image_data):\n",
    "    # plot 5 random images from the training set\n",
    "    n = 100\n",
    "    sqrt_n = int(n ** 0.5)\n",
    "    samples = np.random.randint(x_train.shape[0], size=n)\n",
    "\n",
    "    plt.figure(figsize=(11, 11))\n",
    "\n",
    "    cnt = 0\n",
    "    for i in samples:\n",
    "        cnt += 1\n",
    "        plt.subplot(sqrt_n, sqrt_n, cnt)\n",
    "        plt.subplot(sqrt_n, sqrt_n, cnt).axis('off')\n",
    "        plt.imshow(x_train[i].reshape(28, 28), cmap='gray')\n",
    "        \n",
    "    img = x_train[50000].reshape(28, 28)\n",
    "\n",
    "    fig = plt.figure(figsize = (12,12)) \n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    width, height = img.shape\n",
    "    thresh = img.max()/2.5\n",
    "    for x in range(width):\n",
    "        for y in range(height):\n",
    "            val = round(img[x][y],2) if img[x][y] !=0 else 0\n",
    "            ax.annotate(str(val), xy=(y,x),\n",
    "                    horizontalalignment='center',\n",
    "                    verticalalignment='center',\n",
    "                    color='white' if img[x][y]<thresh else 'black')\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_cifar(ind):\n",
    "    arr = x_train[ind]\n",
    "    R = arr[0:1024].reshape(32, 32)\n",
    "    G = arr[1024:2048].reshape(32, 32)\n",
    "    B = arr[2048:].reshape(32, 32)\n",
    "\n",
    "    img = np.dstack((R, G, B))\n",
    "    title = re.sub('[!@#$b]', '', str(labels_names[np.argmax(y_train[ind])]))\n",
    "    fig = plt.figure(figsize=(3, 3))\n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.imshow(img, interpolation='bicubic')\n",
    "    ax.set_title('Category = ' + title, fontsize=15)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient checking\n",
    "\n",
    "\n",
    "During gradient ascent/descent we compute the gradients $\\frac{\\partial E}{\\partial w}$, where $w$ denotes the parameters of the model.\n",
    "\n",
    "In order to make sure that these gradients are correct we will compare the exact gradients(that we have coded) with numerical estimates obtained by finite differences, you can use your code for computing $E$ to verify the code for computing $\\frac{\\partial E}{\\partial w}$.\n",
    "    Let's look back at the definition of a derivative (or gradient):\n",
    "    \n",
    "$$ \\frac{\\partial E}{\\partial w} = \\lim_{\\varepsilon \\to 0} \\frac{E(w + \\varepsilon) - E(w - \\varepsilon)}{2 \\varepsilon} \\tag{1}$$  \n",
    "\n",
    "We know the following: \n",
    "- $\\frac{\\partial E}{\\partial w}$ is what you want to make sure you're computing correctly. ,\n",
    "- You can compute $E(w + \\varepsilon)$ and $E(w - \\varepsilon)$ (in the case that $w$ is a real number), since you're confident your implementation for $E$ is correct.\n",
    "\n",
    "Let's use equation (1) and a small value ( around $10^-4$ or $10^-6$, much smaller values could lead to numerical issues )for $\\varepsilon$ to make sure that your code for computing  $\\frac{\\partial E}{\\partial w}$ is correct!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gradient_check():\n",
    "    epsilon = 1e-6\n",
    "\n",
    "    # create two random weight matrices\n",
    "    weights_1_tmp = np.random.rand(M, D + 1) * np.sqrt(1 / (D + 1 + M))\n",
    "    weights_2_tmp = np.random.rand(K, M + 1) * np.sqrt(1 / (M + 1 + K))\n",
    "\n",
    "    weights_1_tmp[:, 0] = 1.0\n",
    "    weights_2_tmp[:, 0] = 1.0\n",
    "\n",
    "    # create a fake train batch (of size 8)\n",
    "    b_size = 8\n",
    "    fake_input = X_train[:b_size]\n",
    "    fake_labels = y_train[:b_size]\n",
    "\n",
    "    # calculate gradients with backpropagation\n",
    "    loss, output, hidden_layer_output = forward(fake_input, fake_labels, weights_1_tmp, weights_2_tmp)\n",
    "    grad_weights_1, grad_weights_2 = backward(fake_labels,\n",
    "                                              output,\n",
    "                                              hidden_layer_output,\n",
    "                                              weights_1_tmp,\n",
    "                                              weights_2_tmp,\n",
    "                                              fake_input)\n",
    "\n",
    "    # calculate gradients with two-sided epsilon method\n",
    "    grad_check_for_w1 = np.zeros((M, D + 1))\n",
    "    for i in tqdm(range(grad_check_for_w1.shape[0])):\n",
    "        for j in range(grad_check_for_w1.shape[1]):\n",
    "            w1 = np.copy(weights_1_tmp)\n",
    "            w1[i, j] += epsilon\n",
    "            e1, _, _ = forward(fake_input, fake_labels, w1, weights_2_tmp)\n",
    "\n",
    "            w1 = np.copy(weights_1_tmp)\n",
    "            w1[i, j] -= epsilon\n",
    "            e2, _, _ = forward(fake_input, fake_labels, w1, weights_2_tmp)\n",
    "\n",
    "            grad_check_for_w1[i, j] = (e1 - e2) / (2 * epsilon)\n",
    "\n",
    "    # compute the difference for w1\n",
    "    # it is the Euclidean distance normalized by the sum of the norms\n",
    "    numerator = np.linalg.norm(grad_weights_1 - grad_check_for_w1)\n",
    "    denominator = np.linalg.norm(grad_check_for_w1) + np.linalg.norm(grad_weights_1)\n",
    "    difference = numerator / denominator\n",
    "    print('The difference for weights_1 is: {}'.format(difference))\n",
    "\n",
    "    grad_check_for_w2 = np.zeros((K, M + 1))\n",
    "    for i in tqdm(range(grad_check_for_w2.shape[0])):\n",
    "        for j in range(grad_check_for_w2.shape[1]):\n",
    "            w2 = np.copy(weights_2_tmp)\n",
    "            w2[i, j] += epsilon\n",
    "            e1, _, _ = forward(fake_input, fake_labels, weights_1_tmp, w2)\n",
    "\n",
    "            w2 = np.copy(weights_2_tmp)\n",
    "            w2[i, j] -= epsilon\n",
    "            e2, _, _ = forward(fake_input, fake_labels, weights_1_tmp, w2)\n",
    "\n",
    "            grad_check_for_w2[i, j] = (e1 - e2) / (2 * epsilon)\n",
    "\n",
    "    # compute the difference for w2\n",
    "    # it is the Euclidean distance normalized by the sum of the norms\n",
    "    numerator = np.linalg.norm(grad_weights_2 - grad_check_for_w2)\n",
    "    denominator = np.linalg.norm(grad_check_for_w2) + np.linalg.norm(grad_weights_2)\n",
    "    difference = numerator / denominator\n",
    "    print('The difference for weights_2 is: {}'.format(difference))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Selecting dataset "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdin",
     "output_type": "stream",
     "text": [
      "Type \"MNIST\" for the mnist dataset, \"CIFAR\" or something else otherwise MNIST\n"
     ]
    }
   ],
   "source": [
    "dataset = input('Type \"MNIST\" for the mnist dataset, \"CIFAR\" or something else otherwise')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total train files: 10\n",
      "Total test files: 10\n",
      "Total training images with shape:  (60000, 784)\n",
      "Total testing images with shape:  (10000, 784)\n",
      "Total training labels:  (60000, 10)\n",
      "Total testing labels:  (10000, 10)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:13: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
      "  del sys.path[0]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnIAAAJkCAYAAACVocwTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydZ3Bc13n+n7u972J7B7Boi0aAJFjEIonqPYoV2Y5tRY7GZZxxkklmYk+SD/nmD85kbGdGtuRkkngkK9FYkgtVqEJKFCWQYAFBEL3vYrHYvovtff8f+D/HAJtYQGIXvL8ZjS0SWN2z995z3vOe531eplKpgIWFhYWFhYWFpfbgbPQFsLCwsLCwsLCw3BhsIMfCwsLCwsLCUqOwgRwLCwsLCwsLS43CBnIsLCwsLCwsLDUKG8ixsLCwsLCwsNQobCDHwsLCwsLCwlKj8K72lwzDbBpvkkqlwlzPz9+pY2fHXfuwz/q1caeOG7hzx36njhu4c8d+J4ybzcixsLCwsLCwsNQobCDHwsLCwsLCwlKjXPVolYWF5dbCMAzEYjE0Gg22bNkCiUQCmUxG/35+fh7RaBSzs7PIZDIolUobeLUsLCwsLNUGG8ixsGwgXC4XarUaPT09+OEPfwir1YqGhgZUKhWUy2W88cYbOH/+PF599VUEAgFkMpmNvmQWFhYWliqCDeRYWG4zHA4HPB4Pra2tMBgMeOCBB1BfX4+GhgbIZDKQ/scMw2DLli3QaDSYmprC3Nwczp07h2KxuMEjYGFhYWGpFthA7hbCMAz9hyzOq/+XYb648Ib8PMvmgcPhQCAQwOl0wul04hvf+AZUKhXEYjEAoFwu0/ve1tYGi8WCkydPgsfjYXR0lA3kqpDV7zqhUqnQf1hYWFhuFWwgd4twOp2w2+148sknodfrMTo6ing8Dr/fj1gshuXlZeh0Omi12sv+fjabRTAYhM/nw8LCAsrlMsrl8m0eBcutQCgUQqlU4plnnkFXVxfUajXy+TwmJyfp/Y7H48hkMnjmmWfoc9Te3g6fz4elpSVMTU1t9DBYcCGAs1gsMBgM+Na3vgWtVgsul4tgMIiJiQnMzs5idnYWi4uLiMfjG325LCwsm5BbGsgxDAMOhwMOhwOGYcDj/fE/Vy6XUSgU1vx7re9cyRgFAgHsdjs6Ozvx0EMPwW63Q6vVIhQKwePxIBAIQCKRwGq1wmq1XvazUqkUlpaWIBKJkEwmkclkkMvlkM/n2YCuxuFyuRAKhWhubkZbWxvK5TIymQw8Hg/m5uYwPDyMUCiEZDKJPXv2wGg0oqmpCXw+Hw0NDcjlcmwgt4Hw+XyIRCI6v5nNZjQ2NuKRRx6B2WwGn8/H4uIitFotJBIJSqUSYrEYkskk++5uUhiGWfNcMAyDXC6HUqmEQqFQ82vbang8Hl3XgQtrN4fDAZfLBZfLpet8pVJBqVSia/3Faz7L+nFLAzmNRgOFQgGj0QilUonOzk6IRCIAwOLiIk6dOkUnNq/XW9M7Vh6PB7FYjJ6eHuzfvx933303Ojs7odVqwefz0dfXh2KxiHw+j2KxiFwuBz6fD4FAcNnPK5fLyOfzyOfzyGQyeP/993HixAmcOnUKgUBg000OdxJkQvN6vZBIJPB6vZifn8fBgwexvLyMhYUFenz6xhtvYHp6Gk8//TQUCgW6urqQzWZx/Phx9v5vEHfffTeeffZZSKVSiMVi6HQ6yOVy6PV6uogZDAYcOHAAvb29iEaj+Ld/+zccO3YMgUAA+Xx+g0fAsp7weDyoVCrs2LEDzz33HH0u3nvvPYyPj+P48eNYWVnZ6MtcFzgcDvr6+qDX66FWq1EqlRAMBqHX69He3o6WlhY0NTUBAEqlEsbGxhAIBHD27FksLi7is88+Yzczt4CbDuREIhGNzEmmQSAQQCQSwWg0Qq1W00Cuu7ubBnJKpRLpdBrAhchdKpUiHA4jk8kgn89jZWWlJm44h8OBRCKBVCqFyWRCe3s7uru70dTUBIvFQn9utaXEjeD1epFKpRCLxSCRSOB2u5HL5djFvAYplUrI5/OYnZ1FLpfD0tIS3G43ZmdnEQ6HEQqFAFx4tlwuF2QyGQqFAng8HqRSKX2Hag1itSIQCCAUCi+ZN8hunmhK4/E4stksotFoVdiu8Hg8mknv7OyEXC6HWCyGUqmk8x5wIVDncrmQy+Xg8/lQqVRoaGiA2+0Gj8dDOp1GJBJBqVRi398ah8PhQCqVorm5GU6nE93d3ZDJZBCLxXC73SiVShgeHkYikaiJ9exykISDTCaDVCqF0+mEyWSCVqtFqVRCIBCAXq9HZ2cn7HY7GhoaUCqVUCwWkclkoFAokM/nIZFIMD8/j3g8vmkC22qBudpE8kWtLbhcLjo7OyGRSAAAdXV1aGtrg8PhQHt7O4xGI1QqFbhcLn3gORwOPUYladZKpYKFhQUEg0EMDAxgYWEBv/nNb5BKpdZtoLeqnYlMJsOOHTvQ0dGBP/3TP4XZbEZ9fT34fD64XO6NXexlyOfzyOVyVG/zgx/8AC6XC7lc7gt/l21n8sXc7nGT94HL5dLjh1wud4kW0mq1wm6341e/+hUUCgV++9vfYmBgAP/zP/9zxSCgGlv38Pl88Pl8bN26FVarFe3t7XRzo1Qq4XQ6oVKpoNfraSD3u9/9DqOjo/j1r3+NcDh8Tf+dW3nPjUYjdu7ciUcffRRf/epXIRAIwOfz1xQ6lMtlZLNZOseRuW9wcBAejweRSATLy8v4r//6L8RiMWSz2XUpiNioe87lcmmBBzlKu91s1HvOMAwUCgW6u7vxs5/9DHq9nj6/AODz+eD3+/G3f/u3GBkZQSKRWNfA/Xbdc4vFgqamJjzwwAPo6elBR0cH6urqwOfzAfxx48Lj8eD1euH1epFMJmnwJhaL4XQ6EQqFcOTIEXzyySf4zW9+cyOXQqnmuf1WcqVx31RGjs/nY/v27dDr9ahUKpDL5bDb7TSYqauro9mEcrmMYrG45kEm5+wcDgcGgwFSqRTpdBp1dXWYm5uDz+fD3NxcVe9ceTwe9Ho9jEYjbDYb1Gr1F2ZMiEcYgWgIr4ZAIIBAIIDJZAIAmM1mpFIpLC8v18xOj2gIdToddDodVCoVBAIBXC4XEokE/H5/zYzlZimXy0gkEl/4c4VCoaYzr3K5HAaDAWq1GnV1deju7qbzA6nSJceRxWIR6XSaZh2bm5sBXCgO2Uh4PB60Wi2ampqwe/duOBwOCIVC8Pl8mlUslUrwer1IJBJwu93I5/PIZrPQaDR0TnA4HHSxf/DBBxEIBOByuRCPxxGJRKjsolZgGIZa5pRKJWQyGczPz98R7zCPx4NIJEJ3dze6u7thNBqhUCjW6MMUCgVKpRKsVisikQhmZmZq6v4STCYT9u7di87OTjQ2NlLtJ4fDQTabRTgcxsrKCvx+P5aXl7G8vIx0Oo1CoQC1Wg21Wg2LxUItl+bn51FXV4d0On1NiQiWL+amAjmxWIzvf//72LJlC7XTIAHJ6v+fzWaRzWbh9/sveZAlEgkkEgk0Gg3MZjNaWlqQSCTQ2NiIM2fO4Cc/+QkymUzViiQFAgFaWlrQ1taGxsbGa8rClUqlNQ+wUChcUwhyNQwGA8RiMbZv3w6RSIRAIFAzEyefz4dUKsX+/fvx4IMPYseOHdDpdHjppZcwOjqKd999lzW83WSQitstW7agq6sLZrMZCoVijS3P0tISPvzwQ/ocE1nC9u3b4XQ68ZOf/ARer3fDxiCRSLBz507s3r0bf/d3f0ezcKvJZrM4fPgwJicn8Yc//AGxWAyhUAh9fX3o6+vD888/j+3btwO4sMg/9dRTcLlceOuttzAyMoL+/n6EQqGa0glzuVw8+uijaGlpQS6Xg8vlwn/8x3/cERpAsVgMg8GA733ve2hpaYFWq71kDpfJZODz+bjrrrsglUqxuLhYk4Hcrl278I//+I8QCARrMrDFYhGRSASHDh3C0NAQDh48iEwmg2w2S8dpNBphtVohlUrR2NiIffv2IRKJ4LPPPsPCwgJ8Pt9GDm3TcFOBXKlUwtLSEg3CVmeWKpUKRkZGsLi4CI/HQzMuVwrkGhoaoNPp0NfXB7FYjKamJmSzWezcuRMLCwuYmZm5mUu9ZRQKBbjdbhgMhqtm1SqVCvL5PGKxGJaWljA8PEz/Tq1W0+PpixGJRJBKpbBYLNDpdLRSTqlU0gWx2iE6Qrvdjl27dmHHjh3o7u6GXq+HWCyGXC6HVCqtibHcThiGgU6ng9VqpccY1QzRurW3t0OhUMBqtaKxsRG7du2CRqOBSCRCOp1GJpOhFiukint4eBhcLhd8Ph/lcplatOTz+Q3PRgoEAjQ2NtKsAnlOI5EIVlZWcPbsWSwvL6O/vx9+vx+hUAiZTAbFYhHRaBRutxuRSATpdBpisZieQqjVauzcuRN6vR4GgwEnTpzAyMgIUqlUTSz4xHrF4XDQCvurvcPkGa7WTfm1wjAM5HI51Go1rFYrjEYjOBwO0uk0otEocrkcCoUCTCYTxGIxuru7weVy8dvf/najL/2aIBW4Go2Gel0STSvDMEgkEkgmkzh9+jQ8Hg+OHj2KxcVFJBIJFAoFegIHAPF4HD6fD8eOHUM8HsfWrVvpMSzJZm8E+/btw44dO1AoFJDP5zE3N4dsNgvggowpmUxiZWXlmiUdJNPO4/EuCeiLxSIKhQLC4TB8Ph+KxeK6nzLedCA3PT0NkUgEvV4PLpe7xvD25MmTOHz4MM6fP49oNHrVQK67uxsOhwONjY1wOBz04bnvvvtw/Pjxqg3kiBWE0Wi86o0hRw+Li4sYGBjAr371K/p3drsdarX6sr+nVqthNpuxf/9+aLVaGsjV1dVBpVJt6MtwrfB4PCiVSmzZsgUvvPAC6uvrqe1KPp+HXC6HXC6vibHcbshCeaXq5mqCz+dDLpfj3nvvRVNTE/bt24e6ujqYzWYqcCY2HIcOHcLc3Byd4OPxOEQiEeRyORQKBQwGA3Q6Hcrl8oYXOgiFQuoLuTpQ8fl8mJ+fx8svv4yxsTEEg8FL5CORSARzc3MIBAJIJBIQCAT0Oa+rq8M999yD3t5e3HXXXSiXy/D5fCgUCjUTyNntdrS2tiKZTEIsFl81kCNH5Bd/R7WISqWCwWBAY2MjzGYzACCRSGBmZgaJRAKpVIo+z7t27UJdXd2GSwSuFYZhIBKJUF9fj2effRY9PT00OCmXy4jFYvB6vXj11VcxOzuL8+fPX/EdJX6Y7777LlKpFL75zW9Sq5KN3Lg/+eST+MEPfoBkMolEIoG3336bBm3JZBJLS0uYmZm55hOi1tZWbNu2DVKp9JK5Op/PIx6PY2xsjBZw5fP5dQ3mbiqQy+Vy+M1vfoOhoSFqM2Kz2RAMBuHxeHD69GmcOnUKkUgE2Wz2sjeb+KJNTk4iHA5jdnaWVrwqFAr09vbC7/dDJpPRnU41kc/n4XK5MDk5iVOnTsFsNsNms1Hvt2w2i3Q6jePHj2N5eRlDQ0NYXl6Gy+WinxEOh6+oqxMKhbQyrr6+HnK5HKVSiQppq/1Ylc/nw2Qy4atf/SpaWlqgUqkwNzeHU6dOYefOndBqtXSBr/ax3G44HA527NiBu+66CzKZDJFIBOfOncP8/PxGX9oaFAoFZDIZOjo6YLFY8Mgjj8BoNMJkMiGdTuPs2bM4c+YMzpw5g0QigXQ6DZfLhWQyiWg0So/iSMZ279696OnpQTqdRigU2rCghsPhUKE3OT6LRqOYnZ3FyMgIRkdHsbi4iOnpacRisctOzEqlEjabDZVKBaFQCOl0GiKRCAaDgcowyDFdW1sbtm7dilQqRSv6qxkypuXl5S+8R0KhEN/97nchl8vxxhtvIBQK1eSxGslW7dq1C93d3fQkJZ/PY35+Hm+99RYCgQDC4TDGxsZgt9sRi8Xg9/tr4p4CF+Zss9mM1tZW7N27F3q9HgDQ39+PyclJjI6OwufzYWho6JrmbR6PB5vNBoPBAACQSqUwm81YXl6+5WO5EqTASCAQQC6XY+/evTQjVygUkEqlEI/HEY1Gr+nztFottRq7OCFBrMTuu+8+PPvss3C73VheXsbx48fh9XrXpSr/pgK5YrGIEydOwOv14qGHHkKlUoHBYEAwGMT09DSmp6e/MJNWLBZRLBbh9XrpA2+xWOixm8PhgMlkgkwmq0pDwWKxiGAwiMXFRYyPjwO4kGaNxWJYWVlBMplEJBLBxx9/jPn5eQwMDFxy04jdxNV4+OGHkUqlIJFIUKlU6OdX886WYRgIBALodDrcfffd0Gg0EAgE8Pl8OHfuHJqamqDRaJDL5Wj1Hssf4XK5aGlpwZYtWyAUCpHNZqmupJq+K4lEArVajc7OTjQ3N2Pr1q1QqVT0aHFqagqffvopDh06hEQicYmGiiyOJpMJO3bsQHt7O+rr6zEyMoJoNLohAT4pzDGZTLDZbDCbzZBIJIhEIhgfH8dHH32EiYkJeDwerKysXHFeEolEtHMHOV6VSCTQ6XQ0kCOFTGazGU1NTTh79uztHOpNkUwmEY/Hv/Ae8fl8PPTQQ9Dr9RgYGECxWITf76+q5/haIH2S29ra0NPTA6FQSA29l5aW0N/fD6/XC5/Ph1QqBZvNhlAohEQigVwud0kLt2qEy+VCrVbDZDKhra2NVtZPTEzgyJEjOHPmDEKh0DVvvkmBm1qtBsMwEAqFtNBtI6lUKuDxeOByuWhra7vpz7oWGIbB6OgoZmZm4Pf7kc1mEY/HNzaQI4RCIbz44ouoq6uDzWZDIBCAx+PBwsLCTX2uSCSCxWJBX18fvvKVr+Do0aMYGhpaj0teV0qlEmZnZ/Hv//7vaGpqgtPpxNTUFJaWlmig6vP5kMlkrntR4vP5EAqFEIlENB1NNBoymayqdWUikQiPPPIIWltbYTKZMDExgTfffBNOpxMdHR1QKpXg8XjYu3cv6urqcPjwYbbY4f9DKnuJhU8gEMDi4iJcLtc1Bf63E5FIBJVKRe0J5HI5vF4vXnnlFczMzODUqVNUyH+5zE1jYyO+/vWvo7Ozk2Yfc7kcpqenMTo6SnfKtxPSduv73/8+mpqaoFarMTg4iB/96EeIRCIIh8NIpVLI5XJXzUYtLCwgFAphZmYGRqMRL7zwAhwOx2XfW5vNhm3btuHw4cPg8/lVt2m9GFK12tra+oXBJ8MwkMlkUCgUEAgE11zcVW3odDro9Xr09vbSDZbH48FLL72EsbExTE9P00K28fFxzM3N0ftILDtkMhmSySTC4XBVHqMTPXc8Hsfi4iJtG/jOO+/g+PHjdDN2rWuZSCRCX18f2tvbweFwkEgksLCwgGQyeYtHcuuoVCq0qCOXy1HtL4FhGHC5XEgkEpTLZaRSKUilUqhUKqp3z2azmJqawssvv3zTRYvr8jblcjlMTExALBZjcXGRVmzd7ERERPI6nQ5NTU1VGcQBoOalExMTSCaTSCaTGB8fh8fjuenPJrojYnewuthBqVRWra6MmKe2tLSgoaEBlUoFwWAQp06dgkKhQGdnJx2LyWTCyspKzU7u6w3DMFAqlTCbzZDL5eDxeAiFQvD7/VRzUk3IZDJotVpYLBYYjUYkEgksLS3h9OnTmJubw8jIyCW/QxZzUgTT19cHh8MBq9WKlZUV2s7O4/FsSEBDJtzOzk5YrVZkMhksLy/jxIkT17Xwkvkgl8vRLAZptUcme4JCoYDJZKLWPMQ0uJohG80vKnIQiUTUNHl1e6daQyaTQa/XQ6PRoK6uDsCFdornzp3DwsLCGkuh1aa3xDqKdAAJh8NgGIZqRqspO0eClFgshvn5eSwuLmJychLz8/Pw+/3X9Vmk45HZbIZOpwNw4Rj6cpn520mhUEA6nV5jTF6pVGiR0uXmHKIBJQUd5L1Op9NIJBJr7j3J6JOTiVgsBpVKRS1ZtFotmpubweFwIJPJaJHMjbIuK2epVEI8Hqe7DCJQvtmHk9gTyOVyNDQ0QKFQrMfl3hLILobsYNbrITUajdi6dSuam5uhVqupwejjjz8Oq9WKQ4cO0axftUB26vX19Xj66afB5XLx7rvv4ty5c1heXsann34Kt9tNNZVGoxHxeHxdDZRrFbIw7t+/Hw888ABsNhtyuRxef/11jI6OIhQKVZ330q5du/DQQw9BJpPB4/HgxRdfpMepl8umcblcbNu2DXa7HX/yJ38Cq9WK3t5eetTywQcf4MiRI/j000+xvLx8TX576wnDMNi1axf6+vpgNpuRz+fx3//93xgeHr7hXfPKygpyuRw+/vhjLC0tAbigq7FarTQIslqt0Ov1eO6557Bt2zb88pe/RDAYXLdxrTflchkjIyOoVCpXfSbb2trQ1NQEuVx+G6/u1tDe3o7du3evKU7L5/NYWlpCJBK54u/V1dXhW9/6FhwOB3p7e+F2u3H+/Hn87ne/w7Fjx6qqhzYp4HO5XDh27BhtLXm9Bv0cDgdWqxUtLS20OpsEPWRDs1HMzMzg448/xvbt26l2Lx6P44MPPsDi4iLOnTtH7wc5Tn/ooYeg1WoxNDSEYDCIqakpZDIZJBKJS7pVkIIOjUaDQqGASCRCO8I8/fTT2L9/P5xOJ4xGI9ra2lCpVG7Kg3HdUiAXO9KvJ5lMBsFgsCbEoiRaXy8EAgHq6uogFotpoMPhcKDVaqHT6Ta8jPtyMAwDlUoFrVYLtVqNbDZLbSaIkJSI3Fc3mybtm+4EH6orQTKwRqMRjY2NEAgEyGazWFpagtfrraqKP5JZUSgU1EerUCggEAggEAjQJvE8Ho92Qairq4NCoUBPTw/sdjucTie0Wi1kMhm1b1hYWMDExAQCgcCGtPJhGAZGoxENDQ1gGAapVArz8/NYXl6+4e+eCJ5dLhe4XC6VXYjFYojFYkilUqrX0Wg09N2udkgrJlKIdvH3w+Fw6FzA4/HA5/Np8UexWLymNYN0PSHdbciR9kYgFAppRxYC6VJ0uewpmQuNRiOam5vR0NAAi8UCDoeDQqFABfLVdIxOkhIkc3YjcDgcWjRhtVqpVZbX60UwGEQymdzQMQcCAUxMTMBisUAkElGZErkXLpeLGnST7JrRaERdXR1GRkYQDocxPz+PbDaLTCaDZDK5JtAlv5NIJFAsFqmeXSwWU/NvIo/S6/UIhUJwuVwbH8jdCsikMD09jddeew1TU1MbfEW3H5FIBK1Wu6aqlcPhQK/X08m+2nRyDMPAZrOhtbUVUqkUKysrGBgYoGl5ssMh10288qxWK0qlEjweT9UEK7cblUqFlpYW9Pb2Yvv27dSHaWJiArOzs1V11Eb6L6pUKupqf3FmmPyMyWSCyWTC448/js7OTmzZsgUKhYJmmBmGweLiIvr7+3HkyBH09/dv2FgZhkFvby/uvfdeuN1uTE1NYWBgAIFA4Kaey2KxiGPHjmF0dJRarMzPz6OlpQVbt26lP0cCnmp7ry+GYRhoNBpotVosLi4iFAqt+X6ILyAJigUCARQKBf7iL/6CZmauBaLVIj2Jz5w5A7fbfauGta4QDXBHRwfuuusuKJVKMAwDk8kEvV6PI0eOQC6XUzuKzYJEIoFCocD9999P9dB+vx9vvfUWTpw4AbfbvaEZyNOnT9PAaevWrdi5cyekUinuvfde8Hg8vPbaawiFQmv0yENDQ7QN3+rE1eVa7JHgfnWWNhqNYmJiAl6vF/F4HGKxGHw+H319fRCJRDh37twNn6xVTSBHWnsZjUaq/SJHq4lEAh6Pp6bFkTeKSqVCU1MTVCrVmj8nO9RqDXhKpRLdpRIjZLJjIRM6MQglWiGRSLThlUzXCrlmhUJBW9DU1dXRnTrJLp09e3aNQeaV4HA4tHJxx44dMJlM4HK5GB8fx9TUFNVQVNP9JpkSook1GAyQyWS46667YLVaaRaGtOnS6/Xo6uqC1Wqlus9KpYJUKgWXy4XR0VEMDQ1R08yNghxzDA0Nwe12U/Pi9Si6II3EPR4PhEIhtm/fTnvOEoiOqqenBzqdjlpXVFPXB4FAAKlUCrVaDZVKhWg0ilgsRu0cdDodtFotNBoNdu/ejebmZtrWSSqVolgs0p60XxTA5PN5iEQiSCQSyOXyqjHH/qJAW6FQUAutjo4OiEQi8Hg8+g5X20nKzUD0zqQCu7GxESaTCe3t7XA4HFQP63A44Ha7IZPJqJ/aRpBOpxEOhzE4OIhoNIqVlRWoVCqo1WoEg0FaxLT62byRQHv1fE3WDPIP+TOSib8ZqiaQ2717N/r6+tDR0QGz2bym1VcsFsPk5OQGX+HGYDKZcPfdd1OhKHBhASUmqtXYh5YszkQXRPzASCBORN2k1ybJ0JEem7UAh8Oh/TO3bduGxx57DNu2bYNIJEKlUoHH48Hw8DD+4R/+gVoPXA1imtzb24tvfOMbMJvNKJfLeOedd/DZZ59heXl5Q6o3rwY5fvF4PJienkZHRwcMBgO+853vIJVKYWZmhgZyKpUKSqWSPqtkkwZcOOZ4++23MTw8jM8///yavZtuFZVKBUeOHMHc3BzC4TCtUl2vRSefz2N8fBxKpRJdXV1rAjmSrRGJRPjyl78Mj8eDTz75BEtLS1UVyMnlcmg0GroBd7vd8Hg8kEqlsNls2LdvH3p7e9Hb2wubzQaVSkWF4gKBAKVSCS6Xi1oPXQ0SHJL/rYUjZ+DC3E10oE6nc033mtUb3WrRxt0MHA4HQqGQGtXff//96OzsxIEDB+japdVq8cgjjyCdTuPo0aMIBAJX1RXeSlKpFFKpFN58803w+Xy0trbCaDTiwQcfhMfj+cJq9BuBBLNisZi6UKxXFnbD3wiJRAKpVIre3l7s3bt3TUFDKpWC2+2+7kqZWoZhGNq2qqGhAU6nE0qlcs0utFAoYHBwEOfPn7+i0fJGUy6X1+w8hUIhzUy1tLTg/vvvpyLTXC5H3bQvPp6pNkiVkd1ux8MPPwy73Q6Hw4GWlhZqp1Iul2EymZDNZvH4449jcnISJ06cuOznkfttt9vx+OOPY9u2bTAYDIhGo3C5XJiamoLb7a4qDc3FuFwuSCQS9PT0rMkyGI1Gqo+7uLKRLOo+nw/j4+M4cuQIlpeXN1wEDYBWWAsEAvT29iKTyWBychIrKyvrUnhBPNROnz6NH//4x8dOHFgAACAASURBVLQivbW1FfX19bDb7TAYDNi6dSuamprQ0NCAgYEBTE5OVs27IRaLoVKpqGH5li1bYLPZcODAAdTV1aG+vh7lchnnz59HJBKhkoFCoYD3338fbrcbZ8+evaYjRdL6LZPJIJVKVZ39zsUQzeOBAwewdetWGI1GOn+TY7hoNIqlpSX4fD4kk8mqKla7EVpaWvDEE0/Qziytra3Q6/WIRqOIRqNIJBI0I9fU1ITHH38cR48e3fBkBPGmJfeBJElIEuJmIV0ydDodSqUSzVyvrKysa1vKDQ/kpFIpPUbYvXs3/fNKpYJkMomJiYmadABfzeVu1pUeXGK5YjQasXv3brS1tVH9ESGfz2NwcBAjIyNVG8hVKhW60yR+OkQY3dLSggMHDkCtVtNS90QiQYWw1QyXy4VSqURnZye+/e1vQ61WQ6PRALgwZnKvJRIJeDweHnvsMUgkEgwMDFz2npPAsLGxEV/72tdozz6Xy4WRkRHMzc1VvWbQ7XYjnU7jwIEDNMsqlUrR0tJC2/atzrATisUiPB4PJiYm8Mknn1RVsBoKhcDj8eB0OsEwDA4fPoxyubwugVypVEIwGEQwGFxjqfTEE09g//791BCdOOrv378fYrEYP//5z6vmOZBIJFCpVLRIqaurCwDoMRqXy8WZM2cwPDyMUCiEuro6WK1WZLNZfPjhh7S1U7WM50a40rVLpVJotVrcc889ePDBB2kWcfXPx2IxTE1N0QCi1mlpacF3vvMdSKVSSCQSKpEZHBxEOBzG0tISzGYz2tra4HA48Mgjj2BxcREjIyNrNv23m0qlglKpRAu05ubm1vXzuVwuZDIZGhoa6OevrKwgHo9Dr9evm0xgwwI5UtXS09OD/fv30z6Gq49e8vk8vF4vUqkULV0nZrgXfwGkl2k6nd6Q6laRSASNRgOxWEyPSjgcDlpaWqjBqVAohN1uRzgcxvT0NG0+vBqhUEh3LU899RS92eS7OXLkCKanp3Ho0CEsLS1V1eJHqFQqmJ2dRTabxfz8PFQqFX72s58hn88jk8mgs7MTWq2WHrGcOHECY2NjyGQyVTuxczgc2O12WK1WfO1rX4PD4aC2KUNDQ9R+oL6+HiqVCm1tbZDJZNi5cycymQwGBgao4zuB+K/95V/+Jdra2lBfX49oNIpjx47hnXfeQX9/P9xud9V+J4RYLIZMJoNf/OIXNKO+OpADLtg2OJ1OOBwO1NXVIZPJwO/344033sD4+HhVbUYYhsF9992Hbdu2obOzE/F4HDqdDul0+qYqV6+VK4mnqwFSiEE6jhCta09PD4LBII4cOQKv14uJiQkEg0FEIhF0dHTAZrPh3nvv3VS6MAKfz4fBYIBcLkdrayuamprQ3NxMi70uN2byLNWCE8O1UCgUaKaJYRjE43EkEgm89tprWFpaovO+w+GATCZDZ2cnDAYDRCLRNVcv1xKk+vyb3/wmLBYLnE4n4vE4AoEAnE4n1Gr1umo9NyyQI+fFNpuNtvQB1u7YS6USstkseDwezXoAF87aSY87AnGiJv0MbxdEK0V0X0qlknoMkV6ZdXV11Nm5o6MDXq8XYrGYpp1XIxKJ4HQ60dzcjN7e3jUiyHK5jPn5eQwPD2NqagqRSKQqXwDSg5FhGCwvL0Mmk+HBBx9EpVJBOp2mu7ZCoYBMJoOZmRlMT09XbfEGEaSazWa0tLTgvvvug1qthlAoRCKRwOzsLCYmJujEbDKZ0NDQAJlMBqvVCrvdDrvdjlQqBZ/PR207VCoVrFYr9u/fD6vVCqlUCo/Hg7GxMQwNDeH06dNVGahfTDabRTabxcDAAP0ziURCNXICgQDFYhEqlYq+I6lUCuFwGENDQ3C5XFV130nVdXt7O60olMlkkEql4PP5t3ThIUFcNX0fwFqneplMRnvQ8vl8lEolyGQy2l90YmICx48fpxowsVgMHo9HixZqGaJpW31/+Hw+9Ho9SqUS5HI5Ojo60N7eDp1Od9nFulwuI5vNIhwOV53u9UYhhtfk1MXv9yMcDqO/vx+Li4v05/x+P+01LJPJqtI+az1gGAYSiQS7d+9GY2MjmpubqckykVJwuVxqJ3Wz7/uGBXINDQ24//77cf/992PPnj00i7U6I6dSqbBnzx5s374dX/3qV+mf63S6SyYEEsj9+te/xn/+53/eljFwuVxYrVb88Ic/pHYgUqmUjoW0peFyuSiXy/TYtKWlBdu3b7+skS+Hw1kjhrwYrVYLk8kEDodTVVmMi8nlcvD7/fjXf/1X2Gw2PPbYY7SYobu7G52dnejv78fExARef/11uN3uqp3UVCoVNBoNXnjhBbS1tcFisSAcDuPw4cM4fvw43n//faTTaeTzeXz++efUWsBut8NoNEKn02HPnj3I5XJYWlqCSqWCSqXCP/3TP8HpdKKxsRGJRAIfffQRPvvsM7z55psIhUJVG9heC9lsFrOzs2hqasLevXtx1113Yd++fQCASCSCw4cPY3JyEmNjY4jFYlU1znK5jN///vcYGBiA2Wxe4+VoNBoxPz+P8fHxqrrmW41Go0FPTw96e3uxZ88eNDU1wWAw0BZsb7zxBpaWljAzM4N0Or2mwnplZQXBYJB6ctUypI/46mSB1WrFP//zP9O5e7U34MUUCgWsrKxgbGwM77333pogp5Y5c+YM/v7v/54GJUT7uFrfnslkMD8/D6lUiubm5g282ltPsVhEKpXC2bNnqR6UVGjv3LkT7e3t4HK51Fc1nU7f1HxyWwM5UpqsVCrR2NiIrq4u2O122urkYshOh/gREUg6n5xvJxIJupDezoom8uJ2d3fTnpikTP5qSKXSK475ajAMA4VCQftwkoeFVEBVE8QAdX5+HslkEna7nZZ322w2lMtlLC4u0sbjgUCgagNTpVIJo9FIhefE6HF0dBTj4+OYnp6mExg53vd4PBCLxXTnWV9fj4aGBiwtLVHvre7ubjQ1NSGfz2NlZYVajRDT2NWbFVIoUo0Z2MtBrEmInICYaUajUaRSKczOzlJ5wUYXN1wOr9eLlZUVhMNhqFQqdHZ20oxiqVSC1+ulc85mhljsGI1GtLe3o7OzE93d3dBoNJBIJFheXsb4+DjOnj1LDZwvXpCIBraazKxvFOLrSFprkc1pc3MzfWfJXEAqIwUCAf07IgGKxWJYXl6+7m4J1QpxUfgiiHa61p+DayGfz8PtdiMaja7pmU5OGcvlMorFIi2AqIlAjsfjobGxEQ6HA88//zzsdju6urqoKPJyYmiJRIL6+vo1f16pVOB2u6m2LBgM4q233sLCwgIGBwdv+4tBNG06nY42tL9VEKPSxsZGMAyDhYUFHDlyBKFQCAsLC7fsv3ujkKPUxcVFvPrqqzCZTOjp6aEtmQYGBvDee+/B5/NVbTaOYRjs27cP27dvR2trK1QqFYLBIAYHB/Hiiy8inU6vCa4SiQTK5TLefPNN9PT0wOl0wmKx4LHHHkNfXx+ef/556HQ6yOVyKJVKZDIZHD16FGfPnsUvfvEL5PN5SCQSmEwmKicolUpYWFigJqq1MAnyeDwoFAq0tLTgscceg8FgoNWgbrcbv/vd7zAxMVF1fWMJxHbE6/XCbDbjqaeeQnNzM3bu3IlPPvkEdXV1OHXqFGZmZjb6Um8pSqUSX/rSl7BlyxY8++yz4HK5VL+8srKCV155BQMDA4hGo1cM1EjAsrKyQothapXx8XHMz8+ju7sbgUAA9913H2QyGe3DSQxjC4UC+vv7sbCwAIfDAbVajS1bttDOJ36/H36/v2Y2ZuuBRCJBW1sbjEbjRl/KbSEej+ONN96gz0VTUxO2bt1KA95kMgm/34+PPvoIU1NTN5WMuS2BHHHub2trQ0tLC5qamqDX6y85Tl39/8nguVwuFcnncjnkcjmcPXuWisZjsRjVJ22ETQlp63G7+oRKJBIwDIPW1la6I56bm6s6nRGBZE1TqRTy+Tw9ZiZH4WQBqGZIlpXP56NcLsPv9yMQCNDS+dWQ8abTaWQyGZRKJQgEAkgkEtqhg9g2RCIR+P1+DA0NYXx8HNFoFGq1Gna7HU1NTbBYLLS5s8/nqylhtFgsRnt7OxobG6HVasHhcBCPxzE1NYWJiYnbrmW9XsgCS7LdfD6fWm7U19dj27ZtiMViyGaz1ED0elndnk4qlcJiscBms1WVjkwkEqGrqwuNjY3U9yqfzyOZTCIejyMcDiMajV51ESLfIdEHAbVrhksy4wsLC5DJZOjr6wOPx6PZtnQ6TeeH4eFhLC0twWq10oK1crlMj56r9QTiViEQCKBWqy8xwN6skKwssd+SyWRwOBxQKpUALnR6CAQC1Cy/qjNyRNNmNpvx9a9/HQ6HA93d3fQIlFz81aq0QqEQ5ufn4ff7EQqF8PLLL2N0dJT+3J2SqgUuTKwikQj33HMPVlZWoNPp8Omnn+LEiRNV/z1IJBJYLBYIhUIaxG2UIeSNks1mcerUKUxMTFz1uybBPQkC+Hw+ZDIZncSKxSL9nF/+8pfUcLa1tRV//ud/jl27dqGjo4NmhdajRdTtxGAw4Nvf/jba2tqg0+noRuvVV1/F0aNHa+6+r6a7u5v6ZNXX1+PgwYNYWlq67s/h8/lQKpWw2Wxoa2vDww8/jHvvvfey2qqNQqVS4Stf+Qp4PB5mZmao273P50MwGEQ4HL6mTEKpVMLk5CSi0SgEAgFEItEal4Jaolwu4/PPP4fb7cauXbtQKpVgsVioF+ahQ4fwwQcf0H6dTz75JDW5z+fziEQim+ZI9XqQSCRobm6mwfydglAohNlsxl133YVvf/vbkEqlqFQqmJycxPj4+Lp41t3SQK6urg5KpRIHDhyAw+FAc3MzdDodTc+vZnU7rnQ6jdnZWapHCIfD8Hq91K7D7/dXnSbsapTLZYTDYXA4HKjV6us6fo3H4wgGg7Raymw206bNDMMgk8lQnU61T4qke0Eul8Pi4mJVZ2SuhFAohNPpRKlUwp49e2jmhmSQGxoaUFdXh87OTlgsFvD5/MtmHxiGgVarpcfMuVwOcrkcvb296Orqon10z507h4mJCaysrFR95hK4sOvu6+tDW1sbPUZhGAZTU1M4evQoZmdnkUwmayIbwTAMpFIpVCoVdDod3UkTvZPD4UAul8PJkyepBcsXHZUJBALayUSr1WLXrl3U+NvhcEAkEtHnhUgTZmdn193f6lopFosIBAK04Id0YvB4PDh//vw1d5sgJrjE/Dqfz9ds4UOlUkEsFgPDMHj33XdhMBhoIOfz+WglNp/Ph1qthlwup4FrOp3GzMxM1ftl3io4HA5SqRQikQhtO7iZj5fJe75nzx7a3WO1vp8UAHE4nJv6Hm5pIGc2m9HY2Ijnn38eXV1dUKlUlyxqq3dllUoFkUgEy8vLeOuttzA/P4/+/n4qGq1VisUifbFVKtV1HcMSPVY2m0WxWMQ999xDF4NisUhFt9UexAF/9NpLpVJ0J1JrSKVS7Nu3DxaLZY0NBVmcH3/8cVitVtor+EoQXzqZTIbHHnsMIpEI7e3tMBgMqK+vB8MwKJVKeP/99/Hpp5/C7/fXhLBeIpHgmWeeQUdHB7Zt20aNUE+ePIkXX3wRiUSiJsYBXJib6urqaOZttQ6Ww+HQYP29996j9+dqVgKkiwfJSLW1teEb3/gGHA4HnE7nJT9fKpUQi8Vw5MgRDA0Nbcg7ns/nMTs7i8bGRrS3tyOVSiEajVIvy2ttp1YulxEMBqkF063WE99qiIby5ZdfpsfipH8nSTjs3r0bra2tUKvVNMsaj8cxODgIl8u1wSPYOKLRKMbHx2kP4c0ayDEMA4FAAIvFgi996UtoamqiGlFSwVosFmnv1aoL5Orr69HW1ob9+/djy5YtcDgcVAwKXAjYisUi8vk8fD4ffD4frXg6d+4cAoEAxsfHqalgLWQirgaPx4Pdbqf+YVeDtAhZXFzEBx98AI/Hg9nZWerJdOrUKSiVSnC5XGSzWUxNTWFpaakmAjmijQNw2zSF6wERapMsEofDgU6nwz333LNG00kyriSIy2az8Hq9EAgEUCgUEAqF9FiBWNPw+Xzs3bsXXC6XTvgMw2B+fh4ejweTk5PweDxVn4EmnokOhwM7d+6E0WhEIpGAx+PB6dOncfLkSaRSqZp6l0kWaXFxEYcOHaLV6eQdFovF4HA4eOGFF/DAAw+gv78fPp8Pw8PD1LKEWM0AF575trY2KJVKWnne2dm5pi0hIZ/PY2BgALOzs3j77bfh9Xo35B2PRCJ45ZVX0NbWhkcffRTLy8uYnZ3F6Ojodc3NpVIJJ0+eRCAQwHPPPQe9Xk/9NWsxMw+AVqkXCgVqJUX+HbjQ7WDHjh2QyWRUG0fWtjul7SSPx4NMJsP27duxY8cOcLlcLC0t4cMPP4TL5ap6OdDNIJVKcc8996C7uxsdHR3UqWJ5eRmRSAQKhQJOpxN33303FhYWMDAwcMPz47oFcuSoj2EYWK1W7N69Gw8//DD6+vouKWYolUrI5XJYWVmBy+Witgs+nw+ff/45IpFI1VYxXgwJSolP3OXgcDi03c7VIIFONBrF1NQUXnvtNQSDQfh8PvrAnz59mmbkiLFkoVCoiZeB3HcS9NTCNQMXArl4PE6/Z4ZhoFQq0dPTs+bnVpu5EuGzy+WimgjSkWR1Gy+JREJfcPJ9lMtleDwenDt3Dm63G8FgsKq/K1KU1N3djZ6eHnR0dEAsFsPv92N6ehpvv/02xsfHa+adJpA2gYFAACdPngSHw8HDDz8MHo9HCxUEAgEee+wxpFIpSCQSTE5OYnl5mVbcW61WWCwWVCoV8Hg87N69m2b4rrSZIdYt586dw8jICI4fP75h310ikcChQ4ewuLgIi8WCubk5DA0NYW5u7roCsFKphImJCaTTaYjFYiiVSiiVSvqe1CqkAO/iEyOSce/s7IRYLKaBXDQapRXodwJ8Ph9yuRw7d+5Ed3c3GIZBIBDAmTNn4PP5qnpeuxlI9n379u3o6upCQ0MDXbODwSAWFhZgNBphtVqxZcsWiMVinDlzZuMCOYFAAJ1Ohy9/+cuwWq1wOp1UT2EwGNZk4RiGQTgcxscff4zJyUl88sknSCaTSCaTSKVSyGazX1gBVU2QHm0vvfQSuru78fTTT99wNVahUMDvf/97uN1unD9/Hl6vF/Pz81RDQB74VCq1xrJhIxsOXy/ECLOlpQUGg2FdW5TcKiqVCk6dOgWXywWbzYbW1lbs3r2b2uYQgsEgotEoJicnaQ/NSCQCl8sFoVAIlUoFp9NJu3bodDpYrVaaoUulUlhcXITP58PCwgI+/fRTDA0NYXFxservr8PhgNVqxZNPPonu7m5IpVIsLS3h5z//OaampnDy5Mma7ieZTqdx8uRJBINBlMtlWCwW1NfXw2KxQKPRQKFQQCQS4f7778fOnTtx4MAB2rlGKpXSQB64oBsmG7HLUalU8NFHH2F6ehqvv/46vF7vhh5FVyoV5HI5TE9P46WXXkIqlbpsa8FrIZ/PIxaL4fjx4+BwOLj33nsxNjaG/v7+W3DlG4dWq4Ver0d3dze6u7shkUho5SKpXicasc0Mh8OBzWZDU1MTnn76aWi1WoRCIXg8HszPz1+zvrLW4HA46OjoQFNTE5544gmYTCYwDIPp6WmcOnUKx44dw/nz57Fnzx6o1WraT/5mdMM3HcjxeDzI5XL09fWhubkZO3bsAHBhAlg9WZHm6JFIBKOjoxgcHMThw4dv9j+/oZTLZaRSKQwODoLP59MKxesN5iqVCgqFAq1iGRwcxMrKymU1ZLV0NHUx+Xwe4XAYTU1NkEqlt9W8+WYIBAJIpVIYGxtDuVxGQ0PDJRYRHo8HwWCQBuHHjh1DNBqF3++nFavxeBzZbJY+N6TlEXBBOzM3Nwe3243x8XGcP38e4+PjVb2pIX03LRYLWltb4XA4YLPZkMlkqLZzcXGx5o+RSMshABgYGEBjYyO1liH3h7Qd0uv1sNlsNGNHGsiTTC3RwazWwxCZCTEInZiYqKo2fOVy+Yrz0fV+Ti6Xg8fjgUqlgsViuaFq32pHIpFArVZDo9HQdo3kxIa8M7UkLbleSIZeIBDQ96GxsRECgQAulwvhcBgrKys1o5W9Xng8Hmw2G1paWtDY2AiZTIZ0Og2v14uhoSGcPXsWw8PD1Gje6/UiGo1unEaO6Hy0Wi06OjpgMpnWZA/K5TItZkin0zh48CDGx8fxv//7v5tmN5JKpdDf3w+GYTA5OQmDwQCDwXBdnxGLxRAOhzE4OIiRkREsLS1V9QJ+o5AjVb1ej87OTmg0GvB4vKrPKpZKJSSTSbz66qsQCoX46U9/eolQm2gYiUaGdNwgdg35fB4nTpzA0NAQRCIRPZYjQT9Z5AqFArLZLNLpdNUfmbe2tmLHjh146qmnsGPHDojFYvh8PvzhD3/A2NgYzp49u6mOkMLhMI4ePYrjx49TSwGdTof9+/fDZrOhq6sLEokEXC6XzosXb2ZJpd7qljxTU1M4ffo0zeoODQ0hFAohFotteBC33hSLRXi9XhSLRar13WyIxWJoNJo1WXtytLqysgKPx1O1Rtg3i0AggFAoRGNjIwwGA77yla/A4XCAz+djenoaP/vZzzAxMVEzBXrXCzl9+d73voeenh7IZDIsLy/j0KFDOHHiBN59912kUilqGM3j8ejasaEZObKLJIvQavL5PG0jFY1GMTIygunpaQQCgZrTy1wJ0iJseXkZw8PDsNlsyGaz0Gg0lxgfEu0UadFC2jlFo1FEo1Ha+mWzl2QLhUIoFAraiLwW7CiIhcyNsNoUuZarrwlisRg6nQ7Nzc3U9NdsNmNpaQl+vx8TExPUZqSWM8gXUywWkUgk6L+To0K1Wo1oNIpSqQSJRAKRSIS6ujpks1mqGybZOI/HQxd08o7Pzs7i/PnzCIfDiMVi8Pl8iMfjm3IOKJVKCIfD1GZldcXeZlnYGYahOkoyplKpRFsxVasZMEnMCAQCyGQyFItFGnCtfhaFQiE1Qb54QyuTyaBQKNDY2Ai9Xo+GhgZoNBosLi5iZmYGU1NTm7KjBUlSmEwmWCwW2O126HQ6xGIxLC0t4fz585ibm0MoFKK/s55yk5sK5CqVCsLhMAQCAT755BN0dnbCbDZTUV8gEEAwGMTJkyexsLCAV155hRpIbpaXFriQjTl37hz+5m/+Bm1tbejp6cHXv/517N+/f83PkZd5dHQUBw8epGJm0kOTLHyb6btZDZkQyI6V2C6Mjo7WtIbqTqOpqQl/9Vd/haamJrS3t0OlUqFUKqG/vx9jY2M4ePAgwuHwpgriLgexoFhcXKRBCRF32+12KjMh82GpVKLayaWlJfr9kMwteT82U1BzMdlsFkNDQ0gmk9i7dy8kEglkMtkaP8zNSDqdxueff46hoaGqDWJ4PB76+vpgtVqxd+9eRCIRHD9+HJlMZk3ipbm5GXV1dWhtbb3E3Le+vh42mw1KpRIikQiFQgHLy8v48Y9/jOnpaQwNDW3KeUEoFEIsFuO5557Dvn37YLfbkc1m8eGHH+Ls2bN45ZVXbmny6qYzcqQEe2xsDNlsdo1jt9vtRiAQwOTkJPx+P+Lx+KZ9WQuFAuLxOLWb+Pjjjy8xfSQ7HJfLhdHRUbjdbkQikTXVipsZ8h2RylViw1ArWrk7HbFYTJumO51OejS+sLCAaDSKM2fOYGZmhmoBNztE97balZ3D4aBQKIDD4VD9I5kPS6USPB4PEonEmozcnUaxWEQ2m6UnM1KplOoENwPZbJa2XSJSCxLwV7NelBjyk9ZRAOB0OqmujxQsms1myOVymM3mNXM3sWXS6XS0hZvb7aYWWj6fb9Mlcfh8PqRSKaxWK+rr6+F0OmGz2VAsFhGPxzE2NobZ2dlb7pe3LitoPB7HO++8A6lUioMHD9J0q9vtRigUumIbrs1GpVKBx+OBx+P5wkqsqxmHblaSySQWFhaoK7rZbEZLSwtOnz690ZfGcg1oNBp873vfo5W7qVQKsVgMH3zwAT777DP09/dXvVXKrYZk1smG7WLu5O9mNYlEAsPDw8hms9Dr9VRXuhmIxWKYnZ2lfTRJZyLS8aFaA/hSqYTZ2Vn4/X6kUik4nU4888wzsNlscDgcAP7oPgHgit2ZAGBwcBBTU1P4+OOP4fF4cObMmZtuQ1WNyGQyNDY24qGHHsKDDz6I9vZ2aLVaWpl76NAheL3eW37P1yWQI0JOYopISCaTVfvQ3mru1HFfDaIx8vv9mJ+fB5fLhU6ng1QqhUAg2DQ78s0KaTejVCrBMAzm5ubw+eef48SJE5iYmLjpxs+bDfa7uDKZTAZzc3NUX72ZimKy2SxWVlZw8OBBjI2NIZ1OIx6PY3p6GrFYrGqfC+IsUS6XMT8/j3Q6jXw+D5VKBbVa/YVr2uo2UwsLCwiFQpidna2Z9oLXA7GTcTgceOKJJ9DT04P6+nqaXf7ggw8wOjpKHQ9uNesSyJHj1UwmU5Ntl1huD4VCAYVCAV6vF9PT0+ByuTAYDJDJZBAKhZsu7b7ZIN0nFAoFbfr8+uuvY2ZmBj6fb6Mvj6WGID1HNyPZbBbZbBZvvPHGRl/KdbN6HSfdBlguhUgnWltb8Wd/9mfQ6/XQaDQALiSwDh48iDNnziAQCNyWpA4rTmK57YyNjeGtt97Co48+it7eXmzZsgV8Ph/Dw8NsVq6KCQQC+NGPfgSpVAqxWAyXy4Wpqak1lZwsLCwsmxWGYSASiWCxWPDss89iy5YtMJvN1Ff0888/x+TkJObn57GysnLbEhNsIMdy2/F6vQCA+++/H2q1GgaDAaFQ6Ia7YrDcHhKJBN57772NvgwWFhaWDYPP50Oj0eDuu+9GfX09lEolbdU5OTmJgYEBhEKh2+oVyAZyLLedcDiMRCKBf/mXf4FCoYDX60UqlWKzcSwsLCwsVQvxCBQKhWu8Yufn5zE3N4f33nsPJ06cztmGuQAAIABJREFUuO0ND9hAjuW2k8/nkc/nMTExsdGXwsLCwsLCcs2QFmSkBR9wQfMZDAaxvLwMn89327XeDCsuZ2FhYWFhYWGpTVhREgsLCwsLCwtLjcIGciwsLCwsLCwsNQobyLGwsLCwsLCw1ChsIMfCwsLCwsLCUqOwgRwLCwsLCwsLS43CBnIsLCwsLCwsLDUKG8ixsLCwsLCwsNQobCDHwsLCwsLCwlKjsIEcCwsLCwsLC0uNctUWXQzDbJq2D5VKhbmen79Tx86Ou/Zhn/Vr404dN3Dnjv1OHTdw5479Thg3m5FjYWFhYWFhYalR2ECOhYWFhYWFhaVGYQM5FhYWFhYWFpYahQ3kWFhYWFhYWFhqlKsWO2wEHA4HDMOgUqnQfzYzDHNdetVN/32w3JlwOBf2lAzDoFwus885CwsLyzVSNYGcRCKBQqHAX//1X0On02F4eBhutxuffvopstksstnsRl/iuuN0OtHW1gYOh0MXsitRKBSQz+dx5swZBIPB23SFLCy3nvr6enzpS1+CWq2G0WjE//3f/+Hw4cMbfVksLCwsNcG6BXJcLheVSgXlcvm6fo8EMXK5HFqtFvfeey/sdjukUilUKhUmJiYQiUQ2RSDH5XIhFovB4XDA5XJRX1+PLVu2gMvlfmEgl8vlkMvl4PF4kM/nkU6nUSqVrvv7Ztk4yH3m8Xj0GSAZaJKZJe9QoVBAoVBAsVjc9NkplUqFPXv2wGKxwGazYWBgABwOh322WVhYWK6BdQnkhEIhGhsbkclk4HK5rut3TSYTLBYLdu7cCYfDAbvdDr1ej0ceeQR79+7Fgw8+iN/+9rf46U9/uh6XumHweDzU19fju9/9LkwmExwOBxQKBVQq1TV/RrlcxrPPPgu/34+f/OQnmJ+fh9vtZhe8KofL5UIoFMJqtcJisaCrqwtmsxnNzc1QKBTQarXgcrlgGAbJZBKxWAxHjx7F4OAgzp49i3A4vNFDuKXU1dXh7rvvhkQigVAohE6ng06nQzQaRT6f3+jLY2FhYalq1iWQ43A4kEqlqFQqVN92rQiFQiiVSmg0Guh0OohEIvD5fKhUKohEIpTLZRgMBsjlchSLRRSLRXC5XAAXslTVnq3gcrng8/mw2+1oaWlBV1cXLBYLHA4H+Hw+eLw/3oJisYhyuYxsNnvJuIhe0Gg0QiAQQKVSQSaTXbfGjuX2wefzUVdXB4lEApVKhfr6ethsNrS2tsJkMqG1tRUKhQIajQZ8Ph9cLhfpdBrxeBzBYBCJRAJzc3NYWVlBsVjc6OGsOxwOh777KpUKDMMgk8mgVCpt9KVtCDwej2ZrCQzD0Hni/7F3prFxXefd/9/Z930f7hzuOyVKlkQtlhfVsZXaTuLAaeKmSRO0SROg6KegAZoCRQq0SfsGaJC0QZIGSRE7trPZcuJNkkVr5SZRXIbDITkcDsnhcGY4+768H4RzQmqzds5Q9wcIkDRzyXvucs5znuX/CAQCcLlccLlc5PN55HI5BAIB1th9gEilUojFYqhUKggEgut+h8zJ4XD4mjUql8shFouhUCg8NBtw8gyT6BuXywWPx6NrmkgkAofDQTKZRC6Xo9G3jZEKsjbea8j7pNFo6BzM4/EgFApvelw+n0c8HkcymUQ4HN7y6Ng9MeQYhoFIJEI6nb7tY/l8PsRiMXK5HOLxOJ3Ec7kcQqEQJiYmAAD9/f1YXV1FIBCAXC4Hh8OBw+Eo+ZCrQqGA2WzGf/zHf6Curg5WqxU8Ho8ao4RisYj19XWEw2GMj4/fcHIuFotIpVIQCARQKpWsIVfCGAwGvPTSS7DZbOjr64NcLodMJoPf70c8HkehUEAkEkE0GgWfz4dCoYBYLIbZbMYTTzyBnp4erKysIJlMYm1tbdsZc3K5HEePHsWOHTvA4XDg8XgwNjYGu92OtbW1h2ahIxgMBuh0OohEIrrBEwgEMJvN0Ov1qK+vh1qthlqtxtraGgKBAL7//e/D5XJt7Yk/ROzatQtdXV148cUXUVNTc93vkDn5jTfewNTU1Kb0CK/Xi1OnTiEWiyEajT6o095SeDweKisrIRaLIZPJoFKpoNPpkM1mUSgU0N7eDplMhqGhIfh8PkxOTqJYLEIgENBUlNXV1ftyvZRKJdRqNb7whS+goqICGo0GWq0WNpvtpmvr+vo6zpw5g5GREbz++ut0Ht8q7okhl8/nsb6+jkQicdsesng8Dp/PB7fbjXw+T3cwmUwGfr8fFy5cgNvtxvr6OmKxGNLpNA1DlfpEzzAMLBYLNeD0ej2EQiF9QPx+P/x+P0KhEJLJJAKBAKLRKOx2O7LZ7A1/biaTwcrKCtbX10veI/lRMAwDpVIJkUgErVYLsVgMjUZDd26rq6sIBoPwer1IJBJbfbq3BIfDgVqtRkVFBTo6OmA0GiEQCBAIBLCwsAC3241QKLTpGD6fD6VSCZ1OB6vVCi6XC4FAAKlUCqlUuu3CqwKBAAqFAs3NzaipqQGHw8H6+jqmpqYQCARK/t3+KIRCIRQKBUQiEf27QCCASCS64QJhMpmoIUc2enw+HzqdDkqlEhaLBWKxGGKxGKFQiPXE3Uc4HA5EIhGkUikMBgPkcjmUSiV6e3vR1NQEi8UCnU5Hv18oFJBIJOD3+6FWq6FSqWC1WpFIJCCRSOj9DAQC4PF4SCQS18xn+XweLpcLsVgMq6urZTO3MwxDPVl8Ph98Ph8Gg4FuTEUiESoqKqghJ5VKoVQqqRertraWXiO/3w+ZTEbz5guFAvL5PM6ePXtfDCWbzYbq6mro9XpIpVIAVwoLo9Eo0uk0kskk/a5AIACPx4NMJgOfz0dDQwPy+Tx8Ph8uX75MnU5bwT0x5NLpNKampu7owfN4PFhaWoLL5YLJZMKLL74Ii8WCWCyGqakpfOc736HGXbk82AQul4u9e/eiu7sbZrMZcrl80+djY2M4fvw4hoeH4fF4kEgkkMlksLa29pHhJXItyu2aXA2Hw0FjYyMsFgsOHDiA6upq7N27F2KxGBKJBG+++SYGBgbwm9/8pmw8D3w+H52dneju7sbRo0cRi8Vgt9tx/PhxnDhxAjMzM/D7/ZuOIR7WpqYm9PX1YefOnbDZbJDJZNDpdFheXt42CzfDMHShO3LkCCoqKsDhcOByufDGG29gbm5uq0/xrtFoNOjq6qI5wK2trTCZTNDr9TcMySmVSkilUuqJIBDDj2EYRKNRRCIRDAwMYHx8HPF4/IGM52GDx+PBYrGgvr4eTz31FFpbW9Hb2wuhUEhD4IRCoYBcLge3242TJ09i165d2LlzJ6qqqiCXy9HS0gKZTAbgirGWSqWuK7ETj8fxwx/+EFNTU3jrrbfK4n0nYVORSERzvlUqFY4ePYrW1lbU1NRAKpXS0CUJWV69mWEYBk888QRisRgmJychFothtVoRDoexvr6Ob33rW/d8/udwOHj22Wexb98+BINBJBIJeDweFAoFjIyMwOv1wu120+/rdDrIZDK0trbCarWiv78fnZ2dOHz4MP7zP/+z/A054O4MCmKkXb0LLxaLWx57vlPIbq6hoQFNTU0QCoU0rr60tISxsTGMjIzg4sWLWFhYQDgcplWK27VSkeQemEwmGI1GVFZWQqPRoKGhAWq1GnV1dVCr1ZDJZHT3o9PpUFNTA4lEstWnf8swDAOhUEg9Kz6fDx988AHGxsbg8XgQi8WuMdQzmQyi0SgtFvJ6vdDr9bh06RK8Xu9NPbSlAMMwsFqtUKlU6O7uRjwex8mTJ5FMJq9Jf+ByuWhubkZzczO0Wi3y+TxGR0cxNTVFr0+5wefz0draCrlcDr1eD6PRiIaGBhouNZlMUCgUkMlk16RVEIhXx+v10kU8n88jFoshkUjQ1ItQKIRz587B6XRu8hg8SBQKBfWscLlcLC0toVAoQC6XQygU0k0rmddJ6kixWIRWq6WGqkKhgF6vh1arhVwup1JTU1NTCAaDmJ2d3ZKcSZFIhNbWVjQ3N6OnpwdmsxkSiQT5fB6ZTAZOpxOhUAgzMzNIp9PI5XJYXV3FxMQE3G43RkdHsbKygng8jsHBQSiVSjQ2NtI8LOKdImk2YrEYhUKBfl7q8Hg81NXVQS6Xw2w2Q6fTUc+aSCRCe3s7TCYTNBoNBALBJrWGG0FStKxWK/h8PqRSKbLZLJLJ5CbD+V5RLBZx+vRprKysIBaLUU8cWXsjkQjW19fpuZFNlsvlQkNDAzo6Omi+ukgkuufndzuUjI7cdkMqlUKr1aKtrQ1dXV0QCoXIZrMIBAIYGhrCD37wAywtLWFlZaUsvY13Agkftre3Y+fOnTh06BA14vh8/nULPHQ6HWw2W9kZciSkBlzxOv/2t7/FysoKVldXr3sMMfLj8XjZeB43wjAMbDYbbDYbvvrVr8Lr9WJychJra2vXNeR6e3vR2dkJrVaLYDCIU6dOYWho6Lar3ksFoVCI/v5+KimkVqthNpuhVCqv8cTfCJ/Ph2AwiJmZGYTDYQCgkkM+nw8OhwOhUAiBQAB+v39LDV6i+Wez2cDn8zEwMIBcLkfzjKxWK/1uLpdDNpuF0+lEoVBAa2sr+Hw+AKC2thadnZ1oaWlBVVUVAoEAgsEgfvGLX8But8Plcm2JISeRSNDX14f29nbs2bOHGp6xWAzxeByjo6NwOp341a9+Re9VMplEKBSiubAkeZ+sBZ/61KegUqmgUChgMBho3phQKIREIqHFLB8lRVUKCAQCdHd3w2q1oqenBzabDXv27LknP7eyspL+O5VK0WKIe02xWMSxY8fu6Nju7m68+OKLMBgM0Gg0rCG3XSEVqmazGVKplIZFzp07h9HRUbhcLsTj8W1nxAmFQggEAjAMQ6sS1Wo1du7cCZ1Oh8rKSirDYTQawefz4XA4EIlE4HA4IJVK0dTURL0Y+Xwe2Wy2rK4RkZoxm81IpVKIRqMIBoNb5j15EHA4HPT09KCnpwcmkwmZTAYqlQqJRALBYPCa7xLNOD6fj2QyCZfLdc33Sh3idX300UdRW1uLp59+GlqtFlqtli7OoVAIbrcbExMTWF5exuLi4g0LtBKJBNLpNEKhEC0cu7o6juhJ3o8ir435fCSEqFQqcfDgwWuq+AwGA1QqFQwGA7hcLnbu3IlCoQCNRkMNF/LO5vN55PN5rKysAAAt+AKuFLyQquVisQipVAoul4tnnnkG1dXVDzzEyOVy0dLSApvNhgMHDsBqtYLD4WB5eRnz8/O4ePEiZmdnYbfb4ff7N3lPSTFSKpWiVcUkqpTJZPDmm2/S8CKpVifGm0KhAACcP38ePp+vJCu3ZTIZZDIZHn30UdTU1KCzsxMajQY1NTVQKpWbvkvG7/V6EYlEMDU1Ra8TqcxvbGxEZWUl6uvr6fg3Hp9MJjEyMoJTp06VTLoFn89HU1MTOjo6oNVqkU6nMTIyQp/treK+GnJkMSeU4sN5vyB5MWq1mubEJJNJOBwOzM3Nwefz3dHP3SgcW0qQ0nKZTAaJRAKGYcDj8WAymVBVVYUjR47QnBOSAEsWJLfbjaWlJQwMDECj0VA3PDHkMplMWT07XC4XBoMBWq0WqVQK8Xgc0Wi0LHJe7gQOhwM+nw+bzYb29naoVCr4/X7qZdgIeS50Oh30ej0YhkEqlYLX66WejXKByFDs3bsXnZ2d6O3tpc8+cOW6xGIxuFwufPDBB5iamsKlS5cQiURu+DPvRFT9XkHynEgYUSQSwWQy4fnnn6eJ4ASFQgGJREIVBBobG1EsFulx5Psb02bC4TCKxSLUavV1PSykUlEgENBK5vsRUrsZRKi9sbERbW1t1MDw+/0YHx/H+++/j9HRUfh8vhuqNBAx743/TiaTGB4epv8nEolomJ3L5VKvE9FOLMV0ImKgP/roo2hra4PFYoFUKoVOp6PPPAmjJ5NJpNNpLC0t0UrdRCIBDoeDcDhMC/sKhQLNH9+4tmUyGcRiMTgcDpw8eRJer3crhw4A9D7V19ejrq4OMpkMq6ur1KjfSu7rW9LR0YHq6mqIxWJkMhm89957ZZn/cieQCZHsZIvFInK5HKLR6B15ZiQSCcRiMV0YS02OorOzE4cOHUJnZyeqq6upHhap1DMajbSLwaVLlzA1NYWZmRmsrq7C4XDQ3J+uri588pOfpMnB8/PzOHPmzDVVnqUKyQ8i1aoTExNwOp3XzYvbLnR3d6OpqQk7duxAXV0dYrEYVlZWYLfbr7lvLS0tqK6uRkdHB2pqahCJRLCwsIAPP/ywbAw5kUgEiUSCl156CYcPH0ZLSwu4XC5+8IMf0DGQquXx8XGMjY1hZmZm0+J1PbZ6c3bo0CE8+eSTsNls1NgSCASoqKi4JreJvN88Hg8Mw9AKzuuFBsmGXqFQoFgsgsPhoFgsIpvNUkPmaq42hh4UDMPQQiui1xeNRjE+Po7f/OY3cDqd8Pl8d31umUwG4XCYaqWRa0b00rb6WbgetbW1aGlpQVdXF837JgoSxFidnp7G9PQ0RkZGsLi4SIv4gsEgzR3u6OjA888/j/b2dtTV1dH8Qp1Oh3w+j6WlJczNzeH48eMYGRnB2NjYlisWcLlcfOxjH0NjYyOef/55aLVaRKNRnD17Fv/6r/96w5SZB8V9M+QYhqHaRyqVCul0GqdPn0Yikbil3QZ5uMnEUIo7lJtBDBjyghJDbqOEyvVeVlLKTboBkF2KUqmku2DgivuehGK2EqJ/Vl9fj97eXvT09KCmpoaOL5vN0vBoIpFANBqF0+nExYsXabjJ4/EgnU5Tg58kARcKBSrZUS5hSVLQodVqIZFI4HA4EAwGS75Y4W7QaDSorKyEVquFTCbD0tISIpEIIpHINfdNpVLBZDJBrVZDKpXS0IvP5ysbj6VIJIJGo0Frayt2794NoVCIcDiMy5cv0xCLTqdDKBTC+Pg4pqamsLa2VvIVpmq1GvX19ejs7IRerwfwp3mLhMoI5HkmxhgReN0o5nojY4QUB8TjcdqKkUBCkSRXbivmfTL/knWHyIqQ8P+9mHPLURBYrVZTqQ4SSi0Wi0in04hGo/B6vZiensbly5dx4cIFzM/PY21tDdlsloaUNRoNZDIZampqYLFYoNFo4PV6kUqlwDAMMpkM5ufnYbfbaRh7qzfxpCq3qakJXV1daG1tBZfLxdzcHGZnZzExMbHlhvd9MeTITautrUVvby927NiBfD6Pd999l2rO3WzgpFOEUqmE2WxGMBjc8pt5N5CHPRgMYnBwEOFweJMG0UZIONJisWDfvn3UEKyqqoLJZIJWq0Uul8P3v/99OBwOnD17dkuNhNbWVnz9619HY2MjOjs7qQYS8T5MTEzA6/VicHAQoVAIfr+fVqaRCT2Xy0Emk+Hxxx9HX18fbDYbNeJGR0fx7rvvlo0hR7o4GI1G5PN5nD59GlNTU1t9Wg+MXC4Hp9OJ+fn56y5UpEqRx+Mhn89THcWtnghvh/r6ehpeUqlUOHHiBBwOB0ZGRrC2toZ0Og0Oh4MTJ04gnU4jk8mUlPf8RhAviMVioYZcOp2mcjkzMzOb7hPRyVQqlWhoaEAsFsMHH3yAXC73kR1+QqEQpqen8eijj+Jzn/sc/f94PI5AIIB///d/h91u3/L3PpFIwOVyYWlp6bqFOw8T3d3d+PjHP75p7YrFYrh48SLOnTuHn/3sZ4hGo4jFYkgmk1SFQSgUwmazoba2Fk899RT0ej3tTCQWizE6Oorp6Wl4PB74/X4MDw8jkUggHo+XxOauqakJ9fX1eO6559Da2op0Oo3p6Wl861vfwuLiYknMXffFkBMIBLQM32q1Qi6X0xfyVgd9I0mScmFjEQPxqkkkEjQ0NNx0MuDxeNDr9TCZTGhtbaXHWq1W6HQ6qFQq5HI5dHR0gGEYXLhwYUsNObJjT6VSCIVCWFhYwNraGm0xRVT6HQ4HotEoLee+GoFAgIaGBlRWVkIoFGJ1dRVzc3Pwer1lE44n4sYkiTkWi8Hj8Ww7Md+rIR45kUiEfD6P1dVV+P3+6767Wq0WVqsVAoEA2WwWKysr8Pv9150XiHzLnQiN3w9EIhGNMrS0tKBQKGBxcRFTU1OYnp5GOBwumcXnTggEApifn9/UWSaVSsHpdMLv98PpdF5jyPn9figUCsTjccRiMYyNjd2SIZfL5a6b+5rJZBCPxzE/P4+FhYUtT0cg0iASiQQymYx6Jst1XbpbridmzTAMcrkcIpEIlcohXiytVguFQoHW1lZUV1ejqamJpgkRb+fq6iqWlpYwPT0Nv9+PxcXFLb/vG6msrERXVxe0Wi24XC4uX76M8fHxkvAWEu6LIUdc9Lt27UJ/fz/cbjcWFxfh9/s36bTciEKhgHQ6TbWTtjo+fieQcCJ54cViMdra2vCTn/yE9qQFrm/YEi8cKdEH/hRqJsd96UtfwujoKH71q19t6S7R7Xbjpz/9KaqqqlBfX49z585hZmYGkUiEhlWJsXczZDIZjh49iqqqKhQKBQwPD1NxzHKBw+Ggq6sLbW1tEIvFWF5exunTp8sm9+tO2b9/P7785S+Dz+cjHA5jcHAQExMT10zGDMNg165d+PjHPw61Wo319XW89dZbmJqaumZhZBgGVVVVkEgksNvtJeEJqaqqwosvvohdu3bh0UcfxRtvvIFf//rXeOONN7C4uIhMJlMSBuedMjs7i6WlJVy4cIHKKeRyOZrCEYvFrjHkxGIxVbvP5XI3NOCvpqWlBV/96lfR0tKy6f+j0SgCgQBmZmbgcrm23GBSKBTo6OiA2+1GX18fJicn4fF4qKjvwwTRPyW9z4Erm5vGxkaEQiHs2rWLFjeQ9IkXXngBNTU1aGtrg0gkgkAgQCaToRt/or1HUm1I68JS4mMf+xi++MUv0rSgr33ta1hYWLhuT/St4r4YcsToIArYpKydlFrfSpiBtOYo193PwsIChoaGUF9fTxPgORzOpqo24M4SnEkFWCQS2fIHKZ1OY2VlhYopLi4uYn19nZbgfxQcDgcVFRWoq6uj6t9erxeLi4twuVw3rfArRUi+EHDFA9va2oqlpSW43e5tJzUjFoshlUpp+ykAVOGeCMQSzGYzLBYLKioqoFAoEIvF4PP5MDs7i7W1NWi1WqoAL5FIIJFIIJVKUSgUMDs7WxKGnEKhQHd3NyorK8HlcrG+vg63241IJLLluar3ArLxLBQKtACBVBDmcrnrjrFQKIDD4dBKy4+6T1wuFzqdjuqoqdVq+lmxWITH44HT6dxSQ4nIpZDfz+VyUVVVhUOHDkGpVGJ2dpZuVknUqFy9sLeD1+uF3W7Hrl27oNPpqHixRCKB1WrF3r17aRqUTqejLfj0ej0UCgWdFz0eDy5fvky/63Q6aUpCKa71ZN4m4r9HjhyhqhNEyJ0U55A+2uFwGJlMhnalut/c99puhmGoZg6Jid/KzpUYcsSrU26cPXsWExMTaGpqon3ZbtSa53bJ5/NwOBxwOp1bfm2SySQWFhbuWMiVy+Vi9+7daGtrg8FgQC6Xw8TEBCYnJ2G32+/x2d5/SBUawzDQarX4xCc+gcHBQaysrNCk8e2CSqVCdXX1Jg2oTCZDk5Q30tnZiSeffBJtbW1QKpWYnJzE9PQ0BgcHkc1m0dzcDJvNhp6eHtTW1sJgMMDhcMDj8dy3Pou3i9FoxNNPPw3gygZmcXGRehG2AyRf9XaMUpLWcavHCIVCtLS0oLOzE+3t7ZuenWKxiIsXL2JoaGjLrikxXIlRQQzV7u5udHd34/Tp05icnMT//d//YW5ujhpx26Hv9UcxNTWFUCiExx9/nEqGEMmptrY21NbW0mtAuiBcTaFQwPj4OP7rv/6L5sf6fL4tz4W8GfF4HMFgkOb2/ed//if9jKQcEMfKuXPn4HK56LXy+Xzbw5ArFouIRCIIBAKIxWIl5Y68nxBL/dixY7Db7Th48CAVuZXJZNBqtTTkvLHcnJS7x+NxLC8vo7GxEbt374ZUKoVEIkE2m0UkEsFvf/vbTSKL5QgpaDlw4ADa29up8OZbb72Fy5cvb/Xp3TbFYhE+nw/Ly8tIJBKQSqXYs2cPKioqUFNTA6/XC6/Xi/n5eQSDQQSDwbJIgr8RFosFe/bsgclkQrFYxMLCAlwu13WfSaVSSUOlHA4HRqMRDMPgb//2b2kYVavVwmQyAbhiVNjt9pJIeCeEw2GMjo7CZDLBZDKhqakJ+/fvx6lTp+D1em8qLfKwwzAMTCYTzGYznnvuOdTV1W3y5JIesrOzs5idnd2yeS2fz8NutyOTyWBwcBBWqxU2m41WsVZXV1N9xPX1dWSzWYTDYVy6dAkrKys0TYBU326ntS4cDqNQKODtt9+G2+1Gf38/1Go1NBoNeDwexGIx/e719P9IV6Nz585hdnaWhuxLvaL/nXfewerqKvR6PaRSKQwGAxX8JtE1lUqFqqoqqNVqhEIhrK2tIRQK4fz581heXsbw8PB9fRYeiNoi6VkWj8dLIkTyICCu1nfeeYeWaldVVaGrqwtGoxEajQaLi4s4e/bsJk9NLpfD8vIyfD4fhoeH8fTTT6Ouro66sEkI89ixY3A4HFs5xLtGJpNBr9fjwIEDaGtrQzgcxvLyMv7whz+Unco/cMWQW1tbo8nfarUau3btQldXFw4ePIiJiQlMTEzg+PHjmJ2dRTweL+vJ3mQyYc+ePTAYDCgWi3C73XA4HDTZfSNKpRIVFRV08iOtbf7mb/4GfD5/k9yFy+XC4uIi7HY7hoeHS8aQi0QiGB0dRVdXFyorK9HY2IhCoYD5+XkqB8QacteHw+HAYrGgsbGRVj5ubLsXDoexsrKC+fl5zM3NbakhRzrNDA4OIhaL0QIskgZSUVGBnp4eKrfi9Xrx1ltv0Y49ZD4nxtx2IRKJIBqN4p133oHD4YDFYkF1dTXVHLy6+8dGisUi/H4/3nvvPQwPD5dVG8ITJ07gxIkT0Ov1kMvlVChar9dDr9fDZrPBarWipaWF5nwyDEOZj9Q3AAAgAElEQVRDzKOjo7h06dJ97aF+Xww5rVZL23fcKqTKpRz6zN0OJNfjzTffpCrYRBGeeGiu1hRKJpOQSqXYuXMn2traYLVaIRaLUSwWsby8fEOvR7kgEokgEonw13/91+jv70dVVRWCwSD++7//G5OTkzdVTS9lisUiFhcXEQ6H8e1vfxsVFRXo6+uDXC6HWq2GQqHAnj170NbWhng8TnXHBgYGEAgE4PF4ytao43A4aGlpoa3Xrt6w1dXV0SbbBC6Xu0nl/8KFC3jzzTcxPz9PvRvr6+slsxiSwp49e/ZgfX0dWq0WfX19MBqNWFlZwQ9/+EN4vV4Eg0EqfEvkRx52OBwOKisrUVNTA6lUes2i//777+PYsWO4dOlSSdzzUCiEV155BUajESdOnMAjjzyCI0eObDp3hmEgEAhgMBjw5JNPoq+vD08//TT8fj/8fj+V0zhz5sw905/basgcR9pnJRIJ1NXV3XDdLhaLWF9fx8svvwyHw4F33323bKv4I5EIEokEUqkUlVgjHTqIzivps/uJT3wCRqMRBw8eREdHBzo7O3Hx4kUcP36chpTvJffFkFMoFKisrKRtWoiwr1gshlgs3vSSksIImUwGuVxOk2xJKx/Se69cCx9Izzi73Q6GYcDn82khSCaTuWaSJw3XSRWo1WqleSSFQgE+nw9LS0sl746+EaTSTa1WY/fu3Thy5Aiy2SzcbjdOnz6Nubk52oO2HAmHw4jFYjh+/DjMZjN4PB4MBgOqqqpgMBhgMBhQW1sLgUAAnU4Ht9sNj8cDHo8Hn89XVnl0hUJhU2U22aHW19ff0vFE6Z1IFzgcDrz99ttYWFjYcqX06xEOhzE0NAShUEi9kaSn7vr6Oo4dO0avBalMj0QiNHG+XJ/pu4WInOt0OhiNRtoRgFAsFjEzM4NTp07RJPGtJpVKYXx8HHNzc3C5XLRJvEqlglQqpR0tSPeL+vp6MAyD3bt3Y2lpCcvLy5iensbS0hLm5+dRLBYRDofLsnf01ZD+0YFA4JY0IBOJBM6fPw+Hw4Hp6emymd+uhvQ5vln+pkajoWLhwBWdVVLYI5VKaV4l0dm7V/bMPTXkSM/F2tpaPProo7BarTQ3QiKR4B//8R9pbzJyM9VqNVQqFdra2mA2m2mTYoVCgd7eXnz3u9/FwsICDbM4nc57ecoPFNLpAMANCz60Wi2+9rWvoa6uDjt37qRVXYlEArFYDD/72c9w5syZLe/tdieQ6t3Dhw/j0KFDaG5ups2kp6amMDY2ti2ShvP5PM2RWFlZobs3soPr6+tDTU0N9u7di9bWVnzzm9/E3NwcXn75ZTidTly6dGmrh3BLLC0t4dSpU6isrITNZrvt4zOZDObm5nDp0iX8v//3/7C2trapCXmpQUJlo6OjmJ+fx8DAAGpqatDa2gqpVIrW1lZ0d3ejs7OTSiidPn0aZ8+exdzcXFmmC9wLtFotdDodDh8+jLa2tk3eOCJrQvKnS22jnkql4PF48PLLL+P999+nSgw7duyA0WiE2WyGWq1GS0sLVCoVLBYLDAYD1Go16urqkE6n8fTTTyMQCOCdd97B3NwcTp48iXg8XpZFMjweD5///Odpuohard4kk0WqeDdKZfF4PBiNxrL1xN0OREvy29/+NhQKBSwWC9rb2/GVr3wF/f396Ovrw5kzZzA2Nobf//739yzEfE8NOeJqlslk0Gg0EAgEKBaLEIlEYBgG7e3tMJvN4PP51JDTaDRQq9VobGyEwWCAWCymniu1Wk13QTwej4Yik8nkllr1ZDd2JzFv8v3rHSeTyaDT6dDe3o6amhpUVVXRVldEc2d2dnZLc0juBolEAovFgrq6Oqq1lkwmMTMzg8nJSUQikW2TQ0m8rRsna9L6RyAQIBqNwmazQSgUoqqqCgzDoKWlhYZcy0GqJBKJwO12w+fzIRQKUXkh0n9xo9eFpA8QwdxoNIpEIoGZmRlcvnwZQ0NDZVP4EY1GEY1GIRAIaAK4SqWikYW6ujraYi4UCmF9fR3RaBTpdBrJZLLkjJX7BYnEmEwmVFZWwmq1wmAw0OeCyCh5PB6EQqFNkh+lAjHIV1ZWaPs1ElEhxolGo6FV6iQtRiqVUnkek8mEWCwGr9cLiUSCpaUlKoJLuh+UA0qlEkqlEs3NzWhvb0dVVRXVGywUCrSXeDQapYUsRGeQCANfT1B4O0Gka1wuF7hcLk2dmp+fh16vh9lsRn19PYrFIi5duoREIoFAIHDX9sw9NeSEQiEMBgNteE5K2fl8PgQCATo7O5HP57Fnzx56DIfDoZ488nfgyqKnVCrR1taGhoYGquEjkUgwNDSEtbW1e3nqt4XBYIBEIsHi4uI9y3vg8/l47LHH0NLSQj1xAoGACnIeP34cJ0+exMzMTNkZO8Qwb2lpwUsvvUS9FisrK5icnMTrr7+OyclJJJPJkjde7gbykl+4cAGXLl2Cy+VCXV0d/u7v/g4qlQpf+tKXoFQqcf78+ZLoo/tRLC4uwuv1Qq/XIxgMoqGhAUqlkuaBarVaOnHH43FEo1EcP34ck5OTOHbsGA0ll0sLq6tZWFiAx+PBxYsXIRAIUFNTg7q6OroJa21thdlsxtGjR/E///M/OHPmDC5cuLDtBaIJCoUCKpUKX/jCF9Df34/GxkZIpVJwOBzaa/XEiRP40Y9+BKfTWTbvf6FQwNDQEN20kIpNss7t378f/f39sFgsUKlUaG1tpYLn6XQan/3sZ/Huu+/ipz/9KVwuF7xe71YP6ZY4evQoHn30UfT398NoNG6SF4nH4/B6vfjDH/6Ad999F08++SRaW1uxc+dOyGQyPPbYY5BIJPj973+/7YpAbgRpQXjq1CmMjo6ivb0dBw8exOHDh/HpT38aJpMJ09PT+M53vnPXz8A9NeTI7mVhYQEffvghlEolpFIpzGYzxGLxdRMiya5NpVJR6540TSbdHYiHj2GYLa3yq6yspMKmYrEYIpEIoVAIy8vLd7WT1Gq1UKvVaG9vR1NTE2QyGfh8PgqFAm1ZYrfb4XQ6y9Ydr1KpYDab0dDQAJ1OBw6Hg9nZWdpQvBy7d9wpxHDxeDzgcrlYWVkBj8dDbW0tVCoVlEol8vl8yRtyxDCdm5uDRCKB3++HXC6HUqmEWq1Gb28vZDIZFAoF/H4/5ubmMD4+DrvdToWjy5mNumtcLhd+vx98Ph/Dw8NYW1tDJpOB2WyGwWBAS0sLUqkUpqamEIvFHoqFTKFQ0CpPq9VKW64BV3p0TkxMYHp6mooql4MRR9gYEWEYBrFYjK5l09PT1PNG1jWdTrdJS9Vms6Gvr4+uI6urqyVTnb0R0nZQp9OhqakJjY2NUKvVNMoWi8UwMzODYDCIxcVFjI2NYW5uDslkknrmyUb+epIk5QqXy4VUKqUbVdI7/GqISHYqlcLCwgIuX74Mq9VKWzk2NTXBYrHctRbhPb2yqVQKS0tLePXVV/HrX/8aOp0OarUaf/Znfwar1Uq9bhsRiUQQi8XYu3cvqqqqAFzJBzt9+jS8Xi+mp6fR3d2NJ554AsvLy3C73Vv2wD/zzDP4zGc+A61WCw6Hg7feegvT09P4+c9/fleGyI4dO9DW1obPfOYzqKys3NSHbmRkBC+//DJGR0fhcDhKLvRwK5D2ZL29vdi/fz/Ni/nlL3+JY8eOPZS5Q8TdHovFMDQ0hHQ6TUWR6+vrkc/ny6arxZkzZ3D27Fn6bjMMA5vNhm984xtoampCX18fRkZG8Prrr+Ps2bNwu91l+RzfjHw+j8XFRXg8HgwPD0Ov16OxsRGf+9zn8JnPfAZHjx7FoUOHcPr0aQSDwbLxPt0NVVVV2L9/PxobG2E2mzd95na78b3vfQ8zMzOYm5sr62txdQvCc+fO4cKFCxAKhZDJZPjCF76A5uZmPP7445DL5ZBIJOjr60NzczM++OADDA8P4/XXXy85SQ6SHtHe3o4nn3wSTz31FHp6ejat4R6PB//8z/+MxcVFzM3NIZVKIZPJQCgUUtkWspaRtIJyvtcEsViMxsZGujHxeDxYWlq66TGkf/DCwgIGBgbw93//9+jv78f+/fuhVqtx6tSpOy5ivOcmMnGbkoUol8thZGQEc3Nz17XIFQoFbeVBDLl0Oo3JyUksLi5ienoagUAAKysrGBkZoSKMWwHZYUokEhSLRRiNRgSDQWqg3uriJBQKoVQqYbFYYLVa0d/fj/r6eppXSHoWjo6OYnBwkF6Dcl38RCIR6urqYDKZwOVyqfK1y+VCLBYr23HdLSQxmPwheXHlNtmR+5fP58HlcmG1WlFVVYW6ujqo1WrEYjH4/X4sLS2VZC/Fe8XG+xeNRrGwsIDp6WlcunQJdXV19L1XKpXbWhidy+VCKBSioqICra2tVEcTuOLJWlhYgMPhwMLCApVq2U5slJMqFosYHBzE8vIywuEwTCYTOjo6qKe6uroa2WwWH374IY1MlML1IB2Zamtr0dHRga6uLigUCiQSCYyPj2N9fR0Mw8DlcmF2dpb2RCcSW3K5HEKhcNvmxCmVSjzxxBNIpVJYW1tDPB7/SEOO2EYkjPrhhx/C6/Wit7cXOp2OXsc7Sb24r75OUo1E+pJdD51OB51Oh8cffxzd3d0AruioEfXn6enpkniwgSsVtsTYzGQysFqtCIfDEAqFSCaTt1yAIJPJUFVVhcOHD+PgwYNobW2F0WgEn89HsVhEIpHA4uIiXnnlFdjtdoyMjNzPYd13JBIJOjs7UVNTA4ZhcPnyZbzyyitwOBwPVUj1ZhADoNylKvh8PvW+9vb2olAoIBgMYnl5GXNzc4jFYlt9ig8EMvcNDQ1BoVDg2WefRU1NDXQ6HfR6/S03ly9H+Hw+FAoFbDYb9u7dC61WS5/nZDKJoaEhXLx4kebFbVdI6P29994Dj8fD8ePH0dTUhJdeeglNTU0wm81obm6GyWTCm2++CY/Hg3Q6XRL5ogzDQKfTYf/+/di7dy/6+/tRKBQQiUTwm9/8Bna7HRwOB+vr63A4HFRSxWKxoL6+HlqtlurC5vP5Tfnv2wG9Xo+XXnoJfr8fg4ODWF1dxfj4+C0dS7x3IpEIzc3N+Kd/+ickEgkMDAzA5XKVniF3J5Cck1gsVtIvOZfLpR0X9u3bh2QyCYFAAK/Xu6nvKKnc2bFjB80BJCXJVqsVVqsVarUaXC4XkUgEoVAIAwMDmJ6exvDwcFmHHUkrJpvNhkceeQRyuRxzc3OYnp7GxMTEQ5P0fTOIKC7DMLh48SLGx8fhdDrvuWDkg0IoFOKRRx5Be3s7uFwu3G433nnnHVy6dAmhUKhs9Q/vlIWFBZw6dQr79u1DXV0dLBYLfD4fXfy2IzabDc899xz27t0LjUZDw2vz8/NYXFzE7373O7hcLiSTyZIwWh4EhUIB6+vr8Hq9mJ2dhVarBQCq0vDUU0+hqqoKP/vZz7ZcWoqkwnR0dOD555+HxWKBVCqF1+ulua7k+U2lUpvUG0wmEzo7O2EymaiObDwex9mzZ+97d4MHSSKRwNjYGLRaLQ4dOgSRSASr1YqhoSEsLi5+ZB5ssVikuYQXL16ERqPBCy+8gMHBQczMzNz2NSo5Qy6bzdIy/XQ6XVI3nXhNiBCkxWIBj8dDT08PstkslEolJiYmqJeJw+HQLg5PPvkkVCoVAKCiogK9vb30Z5KfS1pUffDBB5iZmSnbnDgCl8uFVqulIZZ4PA673Y6FhQXMz8+X9djuBVwuF3w+n4aepqenMTs7+5Eu+lKGeORI/ojP58PAwABmZmbK3htHcoau7sRyM1ZXVxGPxxEOh8HhcKDVaqHX67eVd2IjDMOgsrISR48e3SRmTop77HY7BgYGtk2ng1uFSO+Ew2F4vV6a/yoQCMDn87F7927o9Xq89tprW27IiUQi6lU/cOAAGIZBsVhEPB7H6uoqPB4PFhcXaRiYFHlwOBwYDAY0NDRAq9VCLBYjm80iHo9jbGwMMzMzZd2ScCOpVAoOhwNdXV04cOAAlVgJhUKIxWLUjiHr+/XGvLy8jFgsBqfTicbGRjz22GPI5/P0et8OJWfIJRIJRKNRrK6ullxVm9PpxPHjx9HR0QGDwQDgSrPcT3/60ygUClRCZKOniZSnWywWKpxIPHNEemFxcRGrq6v4xS9+gbm5OVqdWu6GDpfLRXV1NaxWK/XOvPbaaxgfHy/r8OHdIpPJIJPJ8MILL6ChoQF79uzB8vIyfvnLX2J6enqrT++u4PF4VIKDtKjaLlXJTU1N+Ku/+iuMjY3h9OnT8Pv9NzROidZYS0sLOjo6YDabUSwWkUwmt22hg1QqRUNDAzo6OtDQ0LBJhSCVSuH06dO4fPkyotHotvVG3ggiT9PV1YXPfvazsFqt9LNisQin04nJycmSkJZSKBT4+Mc/DpvNBoZhMDMzg4sXL+LkyZO4fPkyHA4HzfHU6/U4cuQIKisr0d7eTiuUjUYj8vk89TCR8GO5r2kEn8+Hn//85/jjH/+IV199FRUVFaisrMTnP/95yGQyOJ1OuoldWVnBxMTETd95YghubGF4O5ScIUdCq8QjV0qsrq5iamoKRqORthvj8Xi31JJo403M5XJUyZyUJbvdbpw/fx7z8/Mlk/B6t3C5XGg0GqhUKuTzeayvr8Nut8Pv92+L8V0PktxLyu7JH7JjJV5KnU6HHTt2wGazQSKRIJvNYm5urqzVz8ViMWQyGW1jROSDiBhuuaPRaLBnzx7k83k4nU7abeV6CAQC2qqwra0NSqWSNlnfroacQCCgizgZL/HkhEIhuFwuLCwsIJPJbJsF/VYgDeXNZjOqq6vR3NwMiUQC4E8J8GtrayXTepHP56O6uhpGoxEMw2B9fZ0W7YyMjGy6fxKJBB0dHWhqasLevXtpH22yaVlYWIDT6YTX691WqTTEI7e4uIj5+Xns3LkTIpEIO3bsoDmCq6ur8Pv9VJqJtF8k7wWXy4VCoYBYLKbNE0pCfmS788EHH2BkZATDw8NobGzEs88+C4PBQK3ofD5/jV4OuTkbq9Smp6fx9ttvY21tDWtra5iamoLP50MgECj7PnwbEQgE2LVrF2pqajA3N4eJiQkMDw+XdO7j3UKEQaVSKfh8PkQiEZRKJVW11+v16Ovro733QqEQvvvd78LpdFIZknKEy+Xiz//8z6nREolEMDg4iLNnz2JycrIsO5FcDQmD9/b2Qi6X45VXXsHa2to17yvDMGhra8OLL75Ixa9lMhmy2SxGRkboYrjdkMlk2L17NxobGwGAbshfffVVDA8P47333oPP5ysJY+VBwTAMDAYDLBYLPvnJT6KhoWFTT/FYLIZIJEJbuZWCTigpUFIoFDAajQiFQpidnaXarsCfeqSr1WocOnQIZrMZGo2G/oypqSksLi7i1VdfxczMDCKRyLa876lUCj6fDydOnMD58+dx8uRJVFZW4vnnn4fVasU//MM/ALgSafT7/fB4PIhEIkgmk6itrYVGo0FTUxNSqRTefPNNnDlz5o7W/y035Ei+CZEsKWUjhrQfmZmZQTabRU1NDUwmEwwGA4rFIjKZDKRSKc0LAa4Yd7lcDqurq/RBnpmZwfj4OPx+PwKBAFwuV9kmt98IoVBIxaA1Gg0ikQhtKF/OYqgk7/HqzgUEmUwGgUAAuVwOPp8PiURCpWZIOzqNRgM+n4/l5WWsrKxgamoKHo8HsVispJ//m8EwDMxmM2pqasDj8WiPSp/Pty3CqsAVwyQQCIDP56OhoQE2mw1erxfr6+vUMOPxeFAqlbDZbGhtbUVVVRU0Gg1WVlbg8/mwtrZWduK3twLZvJjNZpoLnE6nEYlEMD8/j6mpKaqft10hHne5XA65XE7zx+rr62lrQrPZTI24fD6PUCgEn89HN/KlUvxBcjg3Fh8yDAOZTEY3qVKpFPX19dDr9VAoFOByuQiFQgiFQpiamqK6aaSDy3aEaAgS20AoFCIajaK1tRWpVIrq5MpkMpoLL5PJkEqlYLFYoFAokEql4Pf7MT4+DrfbfUfnseWGXDabRSKRoE2Ey2GRHxoawujoKI4dOwapVIrOzk5ks1kEAgHU19ejq6sLAKgnLhKJ4M0336QJrvl8HtlsdpOG2HaCYRhYrVZUV1ejt7cXYrEYJ0+epC90OS9iQqGQ5j395V/+5aZ+osC1hhypSisWi7S/6HvvvQe73Y5Tp05hbW2NTuDlfF0YhkF1dTUaGxshEAiwtraG06dPw263b/Wp3TNWVlbwyiuv4PDhw3j22Weh1+vxzDPP4He/+x0WFxdpztDjjz+OhoYG7Nu3D/l8HvF4HP/7v/+LgYGBbSm5weVyodPpUF1djV27dtH84dXVVTgcDly4cAHnzp0ri7n9TmEYhnrfDx06hAMHDqBYLILH42Hfvn1UW410O0in04jH4xgZGcHo6Cimp6dLJuWE9JGVy+UIh8MIh8OIRCIwm80wm83o7u6G0WhEZ2cnDAYDKioqaBTq3Llz+OMf/4gTJ05gdnYW6XS65B009xKPx4Pl5WVMTU1BIpGgra0NOp0ODQ0NMBgMqKyspFE6t9uNfD5Pu5u8+uqrd5xWVTKG3Pnz55HJZJBKpeD1eksi6fNGbGzNQwQuiQDy1RUn6XSaWtzbbQK/EQzDQKPRwGQyQSgU0oq1Upmo7gYejwe9Xg+tVguhUEh3ZGRcmUwGxWKRtpMLhUL0eQkEAggEAhgaGoLb7abVa+SYcqZQKODy5ctIp9OYm5uD3++H0+nE6urqVp/aPSMej8PhcKCiogIOhwNcLhc1NTV45JFHYLPZAAByuRw2mw1arRbxeBzLy8tYXV3F9PQ0FhYWSnpeu1OICHRFRQVUKhW4XC6CwSBmZmZw4cIFrK6ubluPDHAlhcRoNMJqtaK9vR2dnZ1obm4GAFrJKZPJwOPxaH602+3G1NQU7dgTjUZLZg4oFApIJpPIZrOQSCSorKxEf38/gCvjqampgUajQU1NDSQSCaLRKNbX1zE/P4/BwUE4nU4qkvuwQRwzpJDR7XYjFAohlUpBpVLB4/Fs+j5ZG8n1KpnODrcLSfj/3ve+tylMVS5eqnQ6vanS0OfzYWxs7JrvlcpL+iBgGAa1tbVobm4Gn89HIBCg1UvlDqk+0+v1iEQiNJn/evj9frz//vv0GSctnMizsJ2eiXw+j5/85Ceb3uG7Sd4tRYLBIM6cOUOlBY4cOYLe3l50dnbSUBTprRgOh+F0OjEwMIAPP/wQQ0ND10zi2wWBQIDdu3fTav5oNAqn04l3330XL7/88rZKcr8ecrkce/bswf79+/HlL38ZHA7nuh0NiHd2YWEBb731Fn70ox9hfX295FIqSCqQTCZDRUUFDh48iAMHDtDPNxZ0kV6rp0+fxo9//GOEQiEaZnzYIQVsAHDp0qWP/P7dPANbbsgRttOkv13GcTeQak0yeW0X70wymcTo6Cjm5+dht9upKOb1iMfjmJ2dpR65UChUNhuUO2E7vcPXg3hf3W43PvzwQ6RSKUxMTMBqtUIul0OhUCCbzcLn88Hj8WBsbAwulwvz8/Nl0zf3dpFKpdBoNGhubkZ9fT04HA6NqgSDQcRisW3tjQOuGLIkB5aETgm5XA6Tk5MIhUIIBAI0F8putyMUCpWcVipwpQDjvffeo72geTzephQSkhrkcrkQDAZht9sxPz9PI2kknMryJ+73PS4ZQ45l+7ExEdThcJS9ICxwpfro3LlzW30aLFsAkYpwu91wu92YmZmBXq/Hzp07YTQaUVlZiXQ6DYfDAbvdjvfff3+rT/m+I5PJoNfr0dbWRjvdJBIJLC0tIRAIbJtCl5ux0ZAjCzZJscnlchgeHobL5YLD4cDS0hLOnj1b0hu6SCSCY8eOoa6uDvl8HiKRCEKhkH6eyWQQi8Xw/vvvY2VlBZOTk9tKbaEcYQ05lvtGNpvF2NgYzSFkYdlOrK+vI5lMIhaLQSKRQCKR0H6U2z2cCFzJl7JarbQqU6PRgGEYrKys4O2338bs7OxWn+IDwe/345VXXsE777yDX/7yl5s+y+fz1LNFUjFK2YgD/hRaTSQSCAaD1/XI5XI5rKys0DZrrBG3tbCGHMt9IZvNIplMYm5uDktLSyU/ebGw3C6JRIIudg8rSqUSWq2WCptmMhkaPnxYrks8Hsfw8PBWn8Y9o1Ao0Dy35eXlrT4dlluANeRY7jn5fB4DAwMYHByEUChEJpPZlgKoLCwPO6TIY319HYFAAO+99x6Gh4exvLy87XPjWFhKBdaQY7kvbHXjZxYWlgcHKfqZn5/HysrKtpRZYWEpVRg2ts3CwsLCwsLCUp5wtvoEWFhYWFhYWFhY7gzWkGNhYWFhYWFhKVNYQ46FhYWFhYWFpUxhDTkWFhYWFhYWljKFNeRYWFhYWFhYWMoU1pBjYWFhYWFhYSlTWEOOhYWFhYWFhaVMYQ05FhYWFhYWFpYyhTXkWFhYWFhYWFjKlJu26GIYZtu0fSgWi8ztfP9hHTs77vKHfdZvjYd13MDDO/aHddzAwzv2h2HcrEeOhYWFhYWFhaVMYQ05FhYWFhYWFpYyhTXkWFhYWFhYWFjKFNaQY2FhYWFhYWEpU25a7MDCwsJyr+FwOLDZbNBoNOjp6YFcLodarcbk5CQGBgYQDAYRiUS2+jRZWFhYyoItN+QYhgHDMJv+DQDFYpH+YdmekHt99TNA2Hj/2eegfCH3l9xjHo+H+vp61NfX49Of/jTMZjNqamrwxhtvwOPxIJ/PIxqNsvechYWF5RZgbjZZPoiy3X379mHv3r3Q6/WQy+XQarXI5/OYmprC0tISzp07h0AgAK/Xe1e/hy3VvjUexLgFAgGEQiHa2tpQWVmJAwcOQKfTQSgUgmEYFItFxONxhEIhXLx4EdPT05iYmMD6+vpt/Z5SG/eDohSedYZhwOfzodVqUVtbi56eHnR3dwO44pGrrq6GUqlEdXU1RCIRJBIJlpaWYLfbcfz4cQwNDajCAngAACAASURBVMHhcCAYDCKZTKJQKNzS72Xv+a3xsI79YR038PCO/WEY95Z55LhcLgQCAerr67Fv3z5UVlZCo9HAbDYjn8/DYDBgZmYGq6urAHDXhhzL1sIwDLhcLng8HuRyOeRyOZqamtDY2IgjR47AbDZDKpWCYRgUCgVEIhH4/X4IBAJwOBx4vV6k02mkUqlt6akl7wOPxwOfzwcAatBms9myGy+Xy4VMJoNer0djYyMeeeQRHDp0iH4ul8vB5XLpWHO5HPR6PdRqNdbX15FIJOD3+5FIJJBKpbZoFHcPwzDgcDjgcrngcDgQCoXgcrk3PaZYLKJQKCAajd6yAVuK8Hg88Hg8+lxfzcZnmow5k8kgn88jm80+yFNlKTE4HA54PB6d6/P5fFnNgRwOBwzDQCQS0bGQ5zqbzSKXy93T37dlHrmmpiYcOXIEhw4dwoEDByAUCjctYvF4HMlkEmtra3jttdfwb//2b3d1Adjdy61xP8bN4/GgUqlQUVGBnp4edHZ2oqOjAxaLBUqlEmq1Gnw+ny5w5MXN5XKIRqOIxWI4fvw4nE4nXnnlFfj9fsTj8Y98sbd63Lf4eyEUCmG1WvHkk0+ivb0d+/btQ7FYRC6Xwze/+U2cO3futhb1UnjWTSYTPvWpT6G9vR2PP/44FAoFFAoFXbC9Xi9SqRTS6TS9j1qtFhaLBeFwGOFwGN/+9rdx/vx5zMzMIJ1O39LvLbV7rtVqodVqUV1dDaPRiL/4i7+A2Wy+6THxeBxra2v4+te/DrfbfUu/pxTu+VU/Hzt27EBPTw+OHDmChoaGTZ+Td5wQCAQQCoXw1ltvwel0YmRkBIlE4pZ+V6nd8wdFqd3ze4nBYEBXVxcSiQSi0SgWFxc3RWRK+Z4zDAODwQC9Xo9PfvKTqKysRHd3N2ZmZnDu3DmcPHkSFy9evKOfXTIeOYZhwOPxoFar0dLSgoqKCiiVSnA4mwtoZTIZJBIJFAoFbDYbGhoa4PV6EQwGy846vxqySyeeKeDKxBYKhVAoFCCRSKhFn0wmEYvFytILxTAMFAoFpFIpqqqqUFtbi46ODrS3t6O1tRVKpRJCoRDAlfGnUilkMhlEIhHqrZFIJFCpVGhuboZQKITFYkE+n0cikSi763E1fD4fQqEQFRUVqK2tRXt7Ozo7O9HZ2QkAyGazsFgsUKlUSCQSZeWdIcZpZWUlqqurAfzpGY/H45idnUU0Gt1kkFdXV0Mul0MikUAul0OhUNAdbblA5jeBQACRSITKykpUVFSgvr4eFosFnZ2dNzTkSA5hNBqF1+uFRqNBIBBAPB5/kEO4Y4jXXSwWQ6lUwmazoaOjA52dndSQKxaLyGazSCQSWF5epnOhRCKB2WyG0+lEPp/H5OQkkslk2b/jLH9iY1RGIBDQ/ydeah6PBy6XC6lUCqPRiPb2diQSCUQiEUQikdtOrdkquFwuTCYTqqur0d7ejoqKCjQ0NCAWi0EsFl/XO323PHBDjs/nw2AwoL29HU8//TQUCsUNJ2oSinjsscdgs9nw4x//GMeOHUMwGLzlHXopIhAIoFAo8Pjjj+OJJ54AAGQyGbz22muIRqPYvXs3xGIxRCIRLl68iPfffx/pdLrswg0ikQiHDx9GU1MTnnvuOWg0GhiNRup923jfs9ksXC4XXC4X/vCHP0ClUsFqtWLnzp1oampCb28vmpubsbS0hImJCbz66qtl/QxwOBz6sn/jG99ARUUFampq6ARHFrC2tjYkEgn88Y9/RDgc3spTvi3Ie65UKgFcub+ZTAa/+93vMDo6irNnzyIQCCASiVAD9eDBg3jxxRfR3d2Nuro6iEQiyGSysjLkBAIBzGYzKioq0N3dTb1SRqMRMpkMIpHohseSzRoxhnbt2gWZTIZz587d81DM/YDP50On06G9vR3PPPMMzYvcOOZsNguPx4Ph4WH8y7/8CzXivvjFL+KZZ57Bxz72MXR3d2NwcBDJZBLJZHILR8RyLxEIBNBqtTAYDHRzxzAM6uvroVKpoNfrodVqsW/fPgiFQvD5fOrB/8pXvgKXy7W1A7gFyLv7wgsvoKOjA/39/chms5iYmMDg4CBOnDgBj8dzz3/vAzfkZDIZdu/ejdbWVigUCuqRIZDJLJfL0YRpmUyGiooKVFdXo66ujoZkyhWJRAKr1Yr6+no0NTUBuDLB9ff3I5FIoKWlBSKRCHw+HwzDIBwOY3Z2FktLSyXvlSH5QA0NDTAajejr60N1dTUsFgtkMhmkUimAK/c5nU4jk8nA4/EgFAphamoKHo8Hk5OTkMvlWFlZAcMwSCQSaG1thVQqRVNTE4rFIqqrqxEMBuH3+7d4xLePwWCAQqFAT08PampqUFNTA41GAz6fj1wuh1QqhbW1NQQCASwsLMDn85XFQr6RTCYDr9cLkUgEuVyOeDyOWCyG/8/emQe3eV73+sG+kiAIggD3fSdFarEWy7K1OImXOG7rpk6dpJ3p3rrtnfb29rbuTOdm7v2jc+fOtE07aTrTtMnUWZ3ESdzYsSVbsrWLWkhxJ7hvWAgCIAEQxH7/0HyvSYmSZVkSAQnPTCa2SVF4+X3f+533nN/5natXrzI4OIjT6WRlZWVDZtXn8+HxeIhEIshkMpHJ6u3tzYqslEKhENe1srKSzs5O6uvrsdvtN2SfP+rnaDQaKioqRHY606+/VqvFbDazZ88empubaWlpoaysDIPBIMqocrmcWCzG9PQ0U1NTIiOn1WpZWVkhmUxSUFBAMplEq9WiVCpF81OmIZfLkcvl4sBdXFyMTqfDbDajUCiETETSOsZiMZFdfRj03gUFBej1etRqNWq1mtLSUgwGA0VFReJAD9cCudLSUvLz8zGZTJhMJoqLi0mlUkJi4Ha7CQQCW7yi20OpVIoqS2VlJTqdjnA4zOjoKJOTkywuLt6Tw8l9D+TKy8v5y7/8S0pLS0VZcT3rtVFSCdZgMGAwGNizZw8Ai4uLWZNm3Qyr1cpjjz3Go48+yt69e8V/379/P/BhiQVgx44d7Ny5k29+85u8+eabRKPRjA7mpHLob/3Wb7Fz5046OjrEA309Pp8Pn8/Ht7/9bRwOB2fPniUUCm3Qxpw7d46GhgZeeeUVduzYwVNPPUVraysLCwsMDAxw4sSJjNzob8XOnTtpb2/nC1/4AuXl5RQWFpJMJllZWSEUCuHz+Th27Bjnz5/n3LlzuFyurFvjysoKJ0+epK+vj+7ublwuFy6Xi8HBQTwez6brCQaDTE1N0dnZiUwm48CBA5SXl3P06NGMf96lgKSqqoqXX35ZBKHXW6/cDtLhdffu3Wi1Wt54442MPrgqFAoKCwtpb2/nb/7mb7DZbJSWloo1JxIJkskkGo2GSCTC8ePH6e/vx+/3C52cFOTY7Xb0ej1Go3HTPSNTkAKUyspKysrKePrpp6mtrWXnzp0YDAb0er0IYPv7+3G5XJw6dYrR0VHeeOONDfrAB5GmpiaqqqqwWCzYbDZeeOEFLBYLVqv1hmdhvQ2V9P/Ly8s4HA7effdd3nrrLUZHR+/7Gu4ErVaLyWSira2NtrY2ZDIZi4uLvPHGG4yOjjI+Pn5P/t77EsgpFAq0Wi27d++mo6NDlBnWI2Xh+vv7RSaiuLiYI0eOiK4nu91OW1sbdrsdt9udlV1dCoWCoqIiurq6sNlspFIpEokEqVRK/LPT6SQWi7G2tiY6ezs6OnC73Xdkw3E/kDJxjzzyCC0tLWzbtk2cSKSTtYTL5cLtdnP58mVmZmbo7u7G4/GIk+v6a7q4uAhcCwwSiYTQUNTW1uL3+5HL5VmzKZaVlVFeXs6BAwfo6OiguLhYZCMWFhZ47733WF5exu/3Mzw8zOzsbNb6qUWjUSYmJtDpdMzPzwudy+00qawPfiStaKaj0+l4+umnaWlpobKykoKCgtsuCadSKeLxuFinSqVCqVRSUVFBJBKhs7OThYWFjCwtSQHs7t27aW9vx2azkZeXRzqdZmFhAbfbLaQw+/btIxqN4nQ6WVxcJJ1OC92UlOFKpVJCA52J971er0ev17N//35KS0uprKyksLCQlpYWLBYLeXl54n2VTqeRy+UbOvLtdjsej4elpSUWFxfR6/VoNBqcTudtN3dkElJp3Gaz0dzcLDJu1dXVmM1mjEYjRqOR4uJi9Hr9Bn2YVJHx+/2Ew2Hm5uZYW1tjdXWVxcVFHA4Ho6OjzM7OZkVGHhB6Z71eL57neDxOIBC4p9f3vgRySqUSk8nE5z//eZqamoROaj2pVIpoNMrZs2cZGxuju7ubtrY29uzZg9FoFBtbfn4+FRUVzM3NZZ0AXCoV22w29u/fj9lsJpFIsLq6KvRvoVCIS5cusbKygs/nY+fOnTz11FPs3r2bRCKBy+XKyEBOCtaPHDnC888/T11d3Q3BOly7zrOzs1y6dInXX3+dvr4+fD7fTfV/brdbdLTFYjF0Oh1Go5GWlhYWFxeRy+VCR5Hp1NXV8fjjj/Pss8/S3t4OfFiC7Onp4atf/So+ny8jr+/HJRKJMDIy8ol+RjZp4/R6PS+99BJNTU3U1NR8rM+eSCREOVkKXpVKJbW1tWg0Gvbv309/fz/T09MZF9xIL/IjR47Q2tqKzWYTpeDJyUkuXLjA5OQka2trNDQ0oFAomJ6eFllmuVwuAlfpUBaPxzO2oS0/P5/i4mJeeuklurq6qKys3JA5lIJT6bNLpcPS0lKam5sZHx9neXmZ0dFRrly5gt1up6Cg4IZKRLYgvdvb29t56aWXRCPb7RCJRFhZWWFkZAS3282JEydEgLu4uJg1Wbj11NbW0tXVJSREcG2P9/l82R/ISaWCtrY2amtrb8jQAHi9XjweDydOnGBkZES8pI8fP05TUxNdXV3odDpkMhldXV2kUikCgYDo9MwGNBoNVVVVVFZWUlxczPDwMIODg8zMzBCJRPiN3/gNVCoVZ86cwev1EgwGKSgoAMBkMlFWVnZLsfRWIAVw27dv58CBA6Icdr32EWB0dJSjR4/S39/P4OAgk5OTQhtzK9LpNBcuXEAmk3Ho0CG0Wi2tra3Mz89jMpmEVU2mUl9fz5EjR9ixYwddXV2UlZWRTCY5f/48c3NzvPfee3g8HqG3yXGNbHiupRd1eXk5RUVF5Ofn33YGMZVKsba2hsPh4MKFC1itVlGiLCwsFForqZsv05DJZHR0dFBTU8MjjzxCWVkZADMzM1y6dInTp09z9uxZEokESqWSV199FZlMhsPhIBAIoFKpaG5u5tFHH2XXrl2YzWaGhoaYmpoSHoKZEsxJjUmPPfYYbW1tdHV1YbVaicViOJ1O3n33XTweD9PT0+LPSI0rLS0tlJaW8tRTT2G1WvnsZz/L8vIyXq8XvV6PSqUiGo1y9epV3G53xush15Ofn8/BgwfZvn07O3fuxGKxADA5OcnS0pLINIZCIfE8Ly4uMjMzQzQaJRqNCq2s0+lkbW2NSCSStd6R6xu0JD/UQCBAKBS6p/KIex7ISal3o9FIRUUFpaWlN3xPOp3G7/ezsLBAf3+/OMnrdDr6+vowGo10dXWJbse6ujrW1tY4efJkVpVX1Wo1NpuNoqIiTCaTKKcNDQ0RCoV47rnnKCgoYGhoCI/Hw+rqKrt27QIQLf33onX5k6BQKNDr9TQ1NfHMM89QV1cnHub147WSySSzs7O89dZbOBwOxsbGbvvvSKfTjI6Oolar2bt3L3l5eZSVlWGz2dDr9cRisYwM5KSyUUVFBU899RTNzc00NjaSTqeJRCIMDg4yMDDAG2+8gUwmo6qqitXVVZFlzJSX2P1C+n2tz2hkw++gqKhIlM8+jq5LstGZmpri1KlT1NXVidJMYWGhkCtI2apMQy6XC9ucuro6CgsLSSQSeDwezp07x9mzZzl37pwQvp84cQIAp9NJIpFAp9NRXV3NkSNHqKurQ6/XMzMzw8DAAMvLyxmjC5TJZFgsFlpaWjh48CB79+4VL+ulpSXm5+c5duwYY2NjXLp0acOfVSqVPPnkk7S0tHD48GHMZjM7d+4EPszeJZNJfv7znzM7O4vX682qQM5gMNDR0UFbWxt1dXVCHjQ3N8fU1BRDQ0MEAgGWlpbEgX18fPyOfdQyHWlq0fpALhgMEolE7qnrxD2NCrRaLTqdjt/5nd+hs7OTwsLCG74nGAyysrLCD37wAyHsvhVyuZxHH32U+vp6enp6cDgcDAwMZIVOymQycejQIdrb20mlUoyPj3P8+HFCoRByuZx33nmHvLw8fD4fsVhMeGhJ7cvvvfdexnVpFhcXc/jwYR577DFaW1vR6/Xia6lUSpw8z5w5w8WLF+nu7v7Yeod0Ok1fXx/BYJDf/M3fvNtLuOvIZDKhCzl8+DDbt29n3759ooPvzJkzjI6O8r3vfY/FxUXa29tpaGjgS1/6EufPn+fHP/4xExMTLCwsbPVS7ivV1dV87nOfo6amBoBAIJDxLza5XM6OHTtoaGjg4sWLzM/P85nPfOa2Dlxzc3P84Ac/oKenhw8++IDJyUkqKipoaGgQPnqZiuSXV1tbS2NjI2q1eoNM4Ic//KGQCIRCISKRCKurq6J0XFJSwq5duzh48CAHDhxgcXGRc+fO8frrr9PT08PKysoWr/AaOp2OoqIiDhw4wJe//GUqKysxGAwMDw/jdDp57bXXmJubo6+vb9N9LZ1O43K5MJvNoly8/qAi/V6Wl5c3ZK2yBa1WS2Njo0jQXLp0iYsXL/LOO+/gcDgIh8MkEokN02ky8dC9GZKXqVTqvx1UKpV4b6+urnLixAmuXLnC0tJS9mbkjEYjFouFtrY2Ojs7Ny23BYNB5ufnxTzN6y/yZr9Eq9WKWq2mqKgIt9udFWJouHaRS0pKMJvNpNNpYfwpfW1sbAy9Xk84HBanQK1Wy+rqKh6Ph6mpqYx5CKRMa1FREc3NzVRUVGAymTZcC6n87XQ6uXr1Kg6H444CUclIdnFxccO9oFQq0ev1GactkcvlmEwm7HY727Zto6mpSZRhgsEg4+Pj9Pf3Mz8/TzQaZe/evbS2trJr1y6Wl5exWq04nc6tXsZ9QyrPFxUVUV1dLQIYSSeayYc06TnQarUsLy+LbsWbIWWnQ6EQTqeT3t5eRkdHcblcaDQasS9I3elyuVx0QUr2NJnwspd0fBaLBbvdLjIQq6ur+P3+DXq+RCJBIpEgFouhVCopKirCbrfT3NxMTU0NVquV6elpxsbGGBsbY3p6mlgstsUrvIZWq6WsrIyamhqampqQy+VEIhGmp6cZHx+nu7sbt9uN1+u94bpLQavU4boZ4XAYn89HKBQS4wezAblcLrzfrFYrGo0Gn8/H5OQkV65coa+v7551aN5rJC84nU6HXq8XzRgftQ/JZDJ0Oh15eXnIZDLW1taYnJxkbm6OaDR6T/exexrIPfnkkzzxxBM8+uijlJaWbnozHz9+nO9+97v09fXhcrk2nL7D4TDDw8PU1NSIjsVsCdo2Q3qoN9O7xONx3nzzTRHJd3Z28sorr6BSqeju7ub8+fOcP38+Y7QDeXl5Itv05S9/eVMrmbW1Nd555x0GBwd57bXXCIVCd/Uz2O12PvvZz3Lu3DlOnTp1V3/2J0Gr1fLUU0/R1tbG5z//edGx1tvbS3d3N6+99hpDQ0M0NTVRV1fHX/zFX1BcXJxxZfP7geSC/vjjj7Nv3z5sNpsICk6dOsXZs2cJBoNb/TFvSjKZ5M0336SkpISXX36Z+vr6W5ZBY7EYi4uLfOtb32J4eJijR4+Kw5nX62VtbQ2/3y9Gl5nNZl544QXsdjv9/f14PB48Hs/9Wt5N0ev1mEwmdu3axd69e9Hr9R+5N2k0GoqLi/nd3/1dGhsbeeKJJ1CpVIRCIY4dO8Zrr70mGiMyJaCpq6vjlVdeoaamhoKCAj744AOuXr3K9773PdG4cLPGDGkqze/93u/R0dEhXvCAyMpJe1dfX98NB9VMxmQy8ZWvfIWWlha6urro7+/na1/7Gj09PfT29mbc4frjUFZWxmc+8xnq6upoaWnhe9/7HqdPn8bj8XzkPb5r1y6efvpp1tbWmJmZ4ejRo8zNzd3zw9c9eXNIWheLxUJZWdmmnkDrT26Tk5MsLy9vWkLJFvuB22X9WqQuNWkTkLzzrFYrNpsNs9nM0tISMzMzQjOXCUjNK/X19dTU1GCxWG5oYInH4+LkOj09jc/n+0QlMkkvBB/+DpVKJXl5eZtmercKySupubmZ+vp6CgoKiMfjzM/PMz4+LvQ/crmc+vp6WlpaKCkpEZYN0WgUv9+fMQH7vUTq4jabzWzbto2KigqUSqXQPPr9/owvrabTaVZXV1ldXSUvLw+j0bjpfrVeOySZXl/f7CN1REpd+lL50mKxUFBQIExyM4H1w8xTqdQtAy9pHeXl5VRUVNDS0kJVVRVmsxmfz8fExARTU1M4nc6MGcslzUA2m81UVlZiNBpZXl5mamqKwcFBZmdnhTXSzcjPz8dqtQpt+PoDvFRudDqdOBwOQqFQVgRxUnNPWVkZzc3NVFZWkkql8Pl8onEvmybQrEfajyQ7mZqaGhoaGoTX5632ZakMazQayc/PJxwOixhnZWXlnt/T92RX0Gg0wsXZZrNt+qKdnJzkpz/9Ke+//z6Tk5ObCgGlWYUWiyXrs3HrkYJT6eJL5RIpXf3iiy9SXl5OIBCgp6eHn//85xnjISWTyTCbzVRXV/P8889TXl6OWq2+wdQxEAgwPz8vtD+fdJOSZleuz3ZkmgBcoVBw5MgR2tvb+fVf/3UR4DocDt566y0++OADTp06RXNzM0888QQvv/wydXV1QjsHsLCwwKlTp7JiU/+kKBQKzGYzra2t/MEf/IHQV3q9XlwuFwsLC3i93oz+XUjzhC0WCyUlJZsansI1XVAwGOS73/0uAwMDvPXWWzeUa/bv38+RI0fo6uqipKRE/Jz8/HzhT5Yp3avRaJRgMMjw8LCYZnEzNBoNRqORL33pS7S1tfGpT30KtVrN6uoqZ8+e5T//8z/p6+u7qVH0VqBSqSgrK6OyspLKykrm5+fp7e3lZz/7GcePH78tnW9TUxMtLS3U1dWJ+bqSRi4YDOJ2u+np6eHMmTNZYzmkUCj48pe/zM6dO3nkkUdEF/Lly5d5//33M/pZ/SiUSiU2m42WlhZeeOEFcTDr6OgQU2duVh0wGo3isCUd7oLBID6f777oPe9JIKdWq8nLy8NsNlNUVLTBM04SxI6OjtLT08PCwgKJRGLTB1gqRW5mV5LtyGQyioqKaGxsFELQ4uJirFYrHR0dpFIpTp8+LfQzmZKNUygUQuBstVo3lAskUqkUc3NzjI+PC33BJ9mgpVNgTU0NarVaDN72+Xx3rLu725hMJvLz8+no6KCzs1O8eCWj62AwKIbCS2ObJJNMmUxGNBplcXHxE2cuswmFQiHmK2q1WpLJpJjuMDk5KSwoMkETthmSdq2rq4v6+noKCwvRarWb7lXj4+PMzMzQ39/P+Pg4kUjkhpeeNN7n+v1u/SEpU/ZBycR4fHycvLw8Wltb0Wq1GAwGCgsLqaysFDYStbW1lJSU0NraSm1tLXK5HL/fT09PD1euXGFycpJAIJAxQRxcCz4bGxupqqpCpVKxtrbG7Owsfr9/02u3Hklb1dzczPbt20WH6/r1uVwuzpw5w9TU1G3przKBwsJCzGYz9fX11NXVodVqCYfDTE9PEwqFKC0tFZkoCSmwybTruxkymQy1Wi0miygUCqFPl3RuN8Nms9HU1CTswqSM9f1a8z0J5PR6PXa7ncrKSqqrqzd8LRgMcv78eU6ePMkbb7xxS/GuJCS+3jw425GycXV1dRw5coRYLIZcLmfv3r2UlJTQ1NTEhQsX+J//838SDoczpg0frp1U9+/fT3t7O2VlZeh0uhtu1lQqJbQSbrf7E2vjZDIZ27ZtY8eOHej1eiEWn5mZ4dixYxmhoZJm5z777LNixBQg7AVCoRCHDx9m//795Ofniw1O+r6VlRUx6eJhQZpaUllZiVwuJxAIMDc3R3d3N1euXGF6ejpjuhc3w2q1Ulpayuc//3k6OjpEdvp60uk0J0+e5OTJk7z77rssLS1twae9uySTSWHg7vV6efrpp8nLy8NisVBdXc2ePXtwuVx4PB6efvpptm/fzuHDh4X57djYGP/6r/8qGn8yDaPRyOHDh2lra0On0+H3++nr68Ptdn9kI0ZBQQFlZWUcPnyYgwcPipf7+iB8YGCAr3/968zOzt517fC9oqamhsbGRvbs2SMMzd1uNxcvXmR5eZk9e/YwMzMjhsLL5XKsVqvY/+6l/cbdQIo3JNP5UCjE0tIS/f39nD179pZZ2JaWFj73uc8JL8X7zV0N5CSX7srKSvbu3SsG465H6mxaW1vLWPfue0UkEmFoaEi4PtfU1HDo0CHxgEtjSiQvvUgkkrHZmVtlBtLptBg1dTeyKZLHWktLCxqNhmg0yvT0NPPz84RCoYzocCspKaG5uXmDozdc+12UlJTwmc98RlgXrNc5JRIJVlZWcDgcvP766wwODt7vj37fkcvlQnvy2c9+VmRp3G43Z86cobu7m76+vowN4jQaDVqtlr1797Jr1y7RlbxZ2XNkZISRkRHOnDlDf3//ppn1goICcaJvaGjYcA+trq4yPDxMf3//XTkU3U3S6TRutxuVSsXExATJZBKr1UpNTQ3PP/+8mB3c2dkpDn3hcJijR48yMDDAyMhIRge10h6XTqfR6XQic/xRNDQ08Pjjj1NdXY1OpxMSkPXTHu53xuZuIE0ZWv9eN5lMPPnkk8LId3l5WTy3MpkMvV5PMBjk+9//Pm63m5mZmYxds06no6uri4aGBlH+np2dJRAIsLa2dst3WWFhIY2NjZs2/eXl5YlqklqtxmKxkJ+fT2FhoZCdBYNBLl26hNfrvSMZ1V0P5CRd2969eykuLr7he5LJpMgyfVQ6WXoAMqWc8EmROqKPygAAIABJREFUxhZJnjtVVVVCQ5hKpXjjjTeYnZ3l7bffxuVyZVT31no+6nqk02lWVlY2DMX+pH9fdXW1COSCwSAzMzM4nc5PXLa9W9jtdhobG9HpdMBGI1u73Y7dbt/w/dLXk8kkS0tLOBwOfvrTn2ZMCf1OuNl9cf31USgUNDU10draylNPPYXZbAaulZsuXLjAlStXGB4evuef907RaDQUFBSwZ88ennvuOWFMfT3pdJqRkRF+/vOfc/78eSYmJjb9eQUFBTQ1NdHY2EhDQ8OGnxUOh7l06RIDAwO31TV3P0mn00LXNjk5iVqtpqSkROxrUrVlfbOby+Xi3XffZXh4mNHR0YwtmwMbAi2dTidmI98M6f6vq6vj05/+tJg1DR9OKVnf3JbJa9+MHTt28Eu/9EsbnnOTycThw4dv+edcLhcOh4OhoSFmZ2czYr/eDJ1Ox/bt2zcN5D4qWWCxWKivr98QyEn2M/n5+XR2doq5s/X19ZSXl1NXVyesliTP0OHh4TsaxXdXAzmj0UhjYyOdnZ3s3bt3UwPgUCjEhQsXbjlHTYrk7XY7u3fvpqqqasPXk8mk8CXK1IzVehQKhSi9vfjii9TX1wPXOjvD4TDf+ta36O3tZWxsjEAggNvtztggTqVS8cgjj7Bt27ZNy0jSOBYpi/BJy8IWiwWz2UxxcTH5+fkZ1+AgIaXhI5EIsVhMNICsL58CG/5d6lIdHBzE4XCILHW2IJUgJE3Qvn37sNvtwg9NasG/dOmS0EtJEzlefPFFamtrMRgMuN1uzp49y9mzZzl58iQ+n2+rl7YpWq0Wi8UivNOkUXvrdXGpVIpIJML4+DgnT57kzJkznD9/HrfbfcPPM5lMYkTVL//yL1NZWYnJZEKlUpFMJhkeHsbhcPDqq6+KUUeZ9vKXMsrHjh3D4/GIfUGSDqTTaZGBljI2s7OzYtZqphIKhXj33XdZXV3l8ccfx2az8dhjjzExMUEoFGJqaorV1VWhE66vrxfZyJ07d1JTU8P8/DxDQ0O43W40Gg1PPvkkAB6Ph8nJSaanp7Pq4Pbee+8Ri8V45JFHsNlslJeXE41GGRoaEvtWJBIhEolgt9sxmUyUl5dTUFDAb//2b3P69GnOnDmTcfewTCbDYDBQUlLC448/TllZmdBxSpMpbsb101ck2ZQ06eVf/uVfkMvl2Gw2lEolSqUSrVaLRqMRGmpp8sXKysodd23f1UBOShuWlJRQUVFxw9elbNzk5ORNvZCkVnXJbLCqqkqMfJJIJBJEo1Hi8XjG192lMkxpaSnV1dV0dXVhsViECD4ajXL+/Hnefvvt25o7utXI5XIqKiooLy/ftJM4Ho8LgajL5brj9UjCbpPJRGlpqbAZkV4O0Wg0o4L41dVVMVNP8gWTyWQbrBmkB1kinU4Ti8WYm5vD4/FkjNnr7SAdtqxWK/n5+ZhMJvbv3091dbUwxQ2HwwwODrK4uEgwGCQYDFJfX09FRQVdXV1UVFQgk8lYWVnhypUr9Pf3Z0x39vXI5XJ0Oh02mw2bzUZpaSkWiwWj0bjh+6TGlunpaU6ePEl/f/+mmTilUkl+fj4NDQ10dXXx2GOPbQjw4/E4CwsLjI+Pc+XKlY89DeV+kUqliEajjI6OYjAYSCQSokFtPdL9sLy8zNLSUkboWm9FNBrF4XBQVlZGPB7HaDRSXV1NbW2tsNhQKBTYbDbRoFZRUcG2bduwWq2oVCoWFhaYm5vD5XJhNBo5dOgQ6XSaxcVFvF5v1nSqSkgZ1Pz8fGKxGEajkdXVVRwOh9iLpZFU8XicRCKBzWbDaDSyc+dOfD5fxlbXpM5qaTweXAtKl5aWhIYdNo4RlGIV6R0vBXIymUx0vD7//POk02lSqZSwIIrFYsRiMVZXV4nFYni9Xrxe7yeax3rfTIlisRh9fX3Cyfx6/Yv0CzAajZSUlPDnf/7n1NXVUV9fvyHzk06nhTfZlStXmJqayqgXuoRKpUKj0fDFL36RlpYWdu/eTV5enujyDIfDBINBlpeXha4g04M42CgI3eyhlG5QKTN1p6fu/Px8bDYbn/70p9mzZ4946ScSCbxeL6+//jpjY2MZc6q/cOECw8PDJBIJamtrsVgsJJNJXC4X0WiU1dVVDh06xBNPPCGCUafTydTUFK+99hozMzNZE8RJmakjR47wwgsvYLVaMZlMQvOhUCjE5tXW1saRI0cIBAL4/X5sNhv5+fmUlZWhUqlYWVlhfn6e48ePZ+w0C41GQ01NDW1tbXzpS18Ss1WLiorE96wPyr/73e/S398vMjrXo9Pp2LFjB9u2beP3fu/3hARFupelCSDd3d0MDg5m5P62nmQyyczMDCaTiampKdF9v55YLMb3v/99+vv7cTgcBIPBjHl2NyMejzM3N8fMzAwzMzNYLBZsNhu/9mu/xqc//WlGR0dZXV0Vulez2UwgEGB6epoTJ07Q29vL9PQ00WiUr3zlK7S1taHVallYWODMmTNZOfVgdHSUmZkZrly5gkajQafTiUaG9VKRVCrFn/7pnwrLlWQyid/vzwp/ufX3pNFopKysjKqqKlFa1Wg0tLe3i8k9RqORwsJCOjs7yc/PF3uflKiRLEskKdDExAQulwufzyekQfF4nFgs9onGeN23QC6VSuH1ellcXNy0E1OpVKJWq6murqampoaOjg5KS0uFPQN8WFKdn59nYmJCWFtkElKaVSoJtrW1iVma6XSa2dlZ1Go1BQUFrK6uivE12RDESUg36maBXDAYZHFx8RMFcdKJpqysTFidGI1GMfJLGleWCQ73EtKsRKlJZf0AcZVKhcFgENljKZM8MzMjbCk2G/GTiUgmzJJZZnt7u+jCXVtb26AlkcvlGI1GbDabOLCYTCa0Wi1qtVoErgqFgry8PEKhEAaDgVgsllGZ9vXm5i0tLRQWFooqwfoX2OLiInNzcwwMDDA+Pr5ByK9UKlEoFCJ7KenhamtrRaZZwu/3i5/lcrmyIsBXq9WoVCqRebgeafTY8vKyqKZkMlLW3+fzMTo6SlVVlcja5OfnixFMZrNZZNol42+Hw0Fvby/BYBCVSiVK8QqFgkgkwsTExEeaCWcikvH1raQPer0eg8Egxu7J5XLxbPj9/ozd41KpFMlkkkgkQjQaRa1Wi3JrbW2tyDCr1Wra29spKCigtLRUBHLS9ZVcCpaXlwmHw0xNTeH3+xkfHxeWXC6Xi6WlJSGXuBvct0AukUgwNjbGxMTEhgG6Ena7nZKSEv7v//2/tLS0iOh2fbDg9/tZWlriW9/6Ft3d3RnhH3Y9BQUF2O12XnjhBR599FG2bdtGXl6e8M77f//v/1FZWcmRI0ewWq2b6ggzGSmYCgQCFBQU3KBZO3v2LMePHxczZD8u0gbQ0dHBiy++yPbt22lsbESlUhGJRPjJT35Cf38/Q0NDGacvSSaTHD16VOgllEolBoOBZ599lldeeUUIn10uF06nk//zf/4PQ0NDzM/PZ3zWBa4FI8XFxTzyyCP89//+3yktLaW8vJzp6WlcLheXL18Wm7xCocBgMNDU1MTTTz8tTG3Xe6FJhsCPPfYY3/nOdzh27BivvvoqQ0NDGWXDotVqaW9vp62tjaqqqk2nK0ha18HBQd56660bZiIXFRVRVFTESy+9RE1NDa2trRQUFKDX6zc8Q6lUip/+9KdcuXKFn/3sZ/h8vowPenQ6Hc899xzt7e3U1taK+3w9SqWSffv2YTabOXv2rCi1ZjpXr17lD//wD9m1axeHDh2ira2NiooKkskkq6urvP/++wSDQWGH1N3dTSAQYHl5mb1799Lc3Ex5ebkQtTudTr7zne9kxdrvhN27d3Pw4EEOHTpES0sLSqUSn8/Hf/3Xf9Hb25uRhxLJwNfr9dLb20tNTQ3Nzc00NTVRU1PDr/7qr5JIJMTepVKpRIlVLpdvmDokBYM/+MEPGBkZ4cSJEywvLxMIBEQiSspa3s3kzX0L5ORyOXa7HbfbLQK0dDqNwWDAaDTS1NREdXU15eXlG0oW8KEmzuVyMTMzw8LCAh6PJyOzWJIZst1up6ysjKWlJVwuFwMDA0xMTDAzM4NSqcTr9YqN3GAwoNPpxKDsTCaVSjEzM0NBQcGmZsAej4fx8fE7CrKkDp+Ghgaam5upra2lsLAQlUolxp04HA4mJibu+RDiO0XqKlQqlSLzUlVVRUFBgbjnZ2dnGR0dZXZ2VoygytST6np0Oh3t7e20trZSUVEhOoiHh4cZGhpiaGhIlE+kYeNSh7akH5GyuVJpLS8vTwydb2hoYM+ePcKjzO/3b7m1jGSppNfr0el0IkhfjzSObnZ2lvn5edGUoFKpUCgUKBQKWlpaqK+vp7W1VTR8SNYU0qgrj8eD1+tlaGhIyE8yyUNyM6Rsc21trchaxWIxIRVJJpNYLBY0Gg02m41QKITdbhc2VJl+38fjcZaWlpienqa3t1dkWaSv9fX1EQ6HCYfDuN1uFhcXhX5Xo9Gg1+uFwbN0nTPFMulmSJl0aW+PRqM37ZaWTHR1Oh35+fk0NjbS3t6OxWIRc3SXlpYYGRm5o27M+4UUmEvd8tXV1SIrBxutaILBoJhsEo/HiUajWK1WrFYroVCI5eVlRkZGxCFdymTeS+5bIKfX63n22WcpLi7ma1/7GrFYjGQyKfzBPve5z9He3r6p91wwGMTpdHLmzBl6enpwOBwZ29lmMBgoKyujoqKCiooKvv71r9Pb2ysi80gkgkqlYnp6mrq6OtE4UFZWxuzsbMZv3LFYjLfffpu5uTnq6+tvyE44HA5OnTr1sQNSaUOoq6vjT/7kT2hqaqKrq0u86Kanp4UB8PT0dEZnKeRyOXl5eTQ3N/PKK6+Ixh9JN/bee+9x9OhR4YieLVgsFv74j/+YmpoaysvLcTqdjI6O8u///u/8/Oc/33DNLRYLhw8fpra2Vpxk10skJiYmSCQSG7qfd+3axc6dO/nP//xPfvGLX3Dy5Mkt182pVCq0Wi35+fmbWozAh3KCwcFB4aemVqvJz88XHb2///u/z6c+9SnhI7jZXOKjR4/y/vvvc+LECRYWFjL6HpeQDq0HDx6kpqYGpVKJ0+nk4sWLhMNhIpEIR44coba2Vjjf7927V7zkMvXFfj0jIyM4HI4bpmtI9/ytfOHW//dsWK9Go6GhoUHs7U6n86YZcql0XFlZye7du3nqqaf49Kc/jVwuJ5VKMTU1xfDwMG+88QZerzdjExWS9vo73/kO+/btE5q3633hpMqix+Ph6tWr+P1+nE4nn/3sZ3nmmWeYnZ1lZmaGt99++77a69zVQG5tbY2FhQUmJiYYHh7GbrcLV2v48KI//vjjJBIJ9Ho9tbW11NbW0tzcTHFx8YYpDolEglAoxNDQEKdOnRJZrUzveJK6FePxONPT00LcG41GN3SrwofaErVanbHWGutJJpNMTk6i1WpFF9L6F9zNNDKbodfrxUk9Pz+flpaWDffCeiPNCxcu0Nvbi9frzVhrFkCcTru6umhubqaoqEiIgsfHxxkbG6O/v5+5ubmMPpWvR2qfr6qqoqSkhPz8fCKRCMPDw7zzzjtCLiGNaNq1axclJSXs2bOHpqYmABYXF4XtwuLiIlNTU6RSKYaHhyksLKShoUHYzEg2DpcuXdrSdSuVSlFO3b59O5WVlRte4pKx+blz5xgfHxdSj6amJjGmqqioCKvVSmNj44bsDHzYnTo7O8vw8DCXL19mYGCA5eXlrCi1y+VyamtraWhowGQyoVAoWF5eZmJigmPHjrG2tkY8HsdsNhOLxaiurkar1dLW1kYikeD06dMZ131+M6Rs2sdBktlIdjJerzdjExDrkaQEBQUFFBQUcPbs2ZsGcpIhcFVVFZ2dndTU1Ij7YGVlhbfffpuhoSHC4XDGBnES8XicxcVFBgYG+P73vy8OYRJSs93k5CQrKyvMzMwQCoVYWVnhkUceIR6P43a7mZ2d/UgD4bvNXQ3kQqEQo6OjlJaWUltby/79+zcEcgqFguLiYl588UUMBgMVFRVYrVaKioo2LVnEYjHcbjenT5/mn//5nwkEAlmjLUgkEkQiEUZHRxkYGNgQfKRSKZGqllqX1w+ez2SSySSDg4Oi1K1QKG6aqfgoTCYTZrOZPXv2UFNTwxe+8AUKCws33DNw7ff1X//1X7zzzjtEIpGM3hAkB/hPfepT1NfXU1RUhFarJRqN0t3dzY9//GO6u7uZn5/f6o962ygUChoaGmhtbaWyshKNRsPy8jLnz5/n61//uriX8/LyKCkp4Xd+53eora0VHedSRvXSpUu8/vrr9Pf3Ew6Hkclk2O126uvrefHFF0XgW1paSnt7O7/4xS+2dN3SOLqOjg6OHDlyw30eCoXwer385Cc/oa+vD6fTidFoZPfu3dTW1rJ7924xl/J6L0H4UE8zPDzMj370I86fP5/RRsjXI5fL2bFjB11dXRQUFKBQKHC5XPT29vLqq68KLZBer2dlZQWLxYLJZGLfvn3I5XJ++MMfioPtg4Z0b0uawXg8ztTUFAsLCxl7CJUwGAwcOHCAqqoqGhsbSaVSvPPOO5t+b3FxMb/xG79BVVWV8EeFaxKbmZkZ/u3f/i1r7mmpkVLKtkmyiPVI2k7JRkTi0KFDrK2tMT09zfDw8H037r6rgZx0wpyfn+fcuXPU19eLE7mE0Whkx44dovvtZroTKS37jW98g/7+/ttyV84EVlZWmJiYwOPxEAqFaGxsJBqNClNUAJ/Px8WLF9mzZw8A5eXltLS0MDs7m3EC/uuRjF5DoRBOpxOdTkdpaal4Ub3wwgvU1dUxNDSE1+sVXZzr9TBS1mXfvn3U1NRQW1uL2WzGarWi0WiEt5jX62VwcJDx8XEcDgexWCyjgzjpxdbQ0MD+/fux2+3k5eUxPT3N6dOnOXXqFFevXs2KNvz1SNYakgG3UqkklUqJZo7GxkaKiorEKKa2tjbMZjNyuZz5+XkGBgbo6enh6tWrTExMsLy8TDweRyaTCfPreDxOR0cHCwsL+P1+/H7/lssMJIF+R0fHpubXEi0tLeTl5dHR0YHZbObw4cOYzWZsNpvQRq7Pwkl4PB6OHj3K5cuXb2oanKkoFAoxyaGiokJUUiRB93q97/j4OBqNhiNHjlBYWEhJSQk2mw2z2Sy0cg8SNpsNu93Otm3bxKzW1dVVPvjgg4wV/K8nlUoJgb40nkpCJpNRV1dHUVERBw4cEPNXpWYOyWbotdde4+LFi1l1T0tIB6zry+gS8Xj8hmu41UmYux7ISZYL/f39+Hy+G2wqJFHz7fwcl8vFT37yk6zKxIVCIebn5/F6vYTDYaqqqkgkEvT19YlAbmVlhZGREbxeLzKZjOLiYqqrq8XctUwnHo+ztrbG0tLSDY0pTzzxBHv37uXtt99mYmKCWCwmXsxSaUKyrnjqqadoa2sTDQ3woY5MGo9y6tQpzp49y/z8fEZrhiRzyLa2Nrq6uujo6MBkMgkR+y9+8QuGhoZuOqYp00kkEsI2RaVSkU6nUalU5OXl0djYSGNjI4cPHxZ6T4VCQTQaxePxCGPcwcFBvF7vhhe31NQwNTUlNn2NRoNcLt/yTI10Pdvb22+qf1IoFNTU1GCxWFCr1RQVFXHw4MFNO1vXI5Xa3n//fQYGBrImayGhUChQqVQUFRVhs9mEf5bkibXemcDpdKJWq1lbW0OhUAj7lvz8/KzSiN4uFouFhoYG6uvrhWje6/Vy+fJlxsbGsiKQW11dFU0c8XhcZKYUCgVVVVU0NDTwxS9+UfgpSrYbgUCA2dlZjh07xvHjx7d4JXeGZHL9cblZ4Hc/uCfNDlIwMz4+zujoKBUVFbdVfpNmdK6srNDd3c3ly5eFJipbkDQfp0+fJhQKsX37dmpqajh58qQIUNaXVmKxGAaDgcLCwo/c/DOV9cG6lF199NFH2b59OwcPHrxhAofBYBAj2IxGoxDGplIpnE4nJ06cYHh4mIsXLzI7O5txw8I3o7Ozk4aGBp577jmampqEs/uJEye4ePEi586dy3ht581IJpM4HA7S6TQzMzPY7XYKCws5ePCg0M4VFxeLYEbympKak37605+KyQ63CsYnJib48Y9/LDZEaf7gVnKrMlhBQQEGg4GCggJhT6BWqzctx0gTPiKRCH6/nzfffJPh4WHRBJVtqNVqjEaj0Mip1WqcTidvvvkmly9f3vB7s9ls1NfXo9VqhTmsNJIom/b226WkpITt27djtVpRq9XiXdjb24vT6cz40qper2fPnj2UlZVRUlLCs88+KyxlFAoFBw4coLS0lKqqKmQyGT6fj9nZWS5dusTly5e5cuUKDodji1dx/7nZYe9+cE8iB8mZ3OPxMD8/v8E0cbOoVdropFOq2+0WjQ2RSCQjbSZuhtRyPzc3h0aj4eDBg2JsWSQSEUPepdPr6uoqcrlcmCdmC1KJVXoxSzew5KtTWlpKOp2mtrZW/JnNdEKAGOcSDoeZn5+np6eHoaEhLl68yOrq6paX2G6FZGRdWVlJa2srdXV1lJeXEwwG8fv99PX1MTIycse+eplAOp0Wo5WWlpYwGo3Y7XbKy8sxmUwUFRWRn58vDIEXFxdZWFjg6tWropPzdpAOcZnGzTZntVq9waLg+j8jHU6k0TzS/iZ57o2NjbGwsLDlmcc7QTI+lwyO4dohdmpqisXFReRyuXD/lzxClUql+B1Ic4kzOct+p0hm5gaDAblcTjAYZGlpCa/Xm5H39/XI5XKhXzabzdTW1m4YOyiNmZRKkC6Xi4mJCS5dukR3d/eWNyltBVsdnN+TQE56MZ84cYL5+XmeeeYZqqqqqKurw2AwbGjpjUajRKNRUVv/x3/8RxwOB5OTk6yurmZVELcet9tNIpHAYDDQ3NzMX//1X9PT08Pf/u3fsra2RiKRwOFwcOzYsazUiITDYU6fPk08HufAgQN3nFJOJpP09fUxOzvL66+/LpzxpeHLmV6GaG1tZd++fRw5coSOjg5sNptobOjr6+Pb3/52VmZcNmN1dZULFy6wtrZGW1sbGo0Gk8kkXtDvvPOOuKelQ9z1xrgPC+FwGJfLJSZaeDwefD4f3//+94UOMNMmWHxS1jdxmc1mnn32WZ599lmamppExnZpaYl/+qd/Ynh4GKfTmZVB7EdRVVXFgQMHsFqtJJNJnE4nCwsLWfMuW11d5fTp07S1tYkD2/qRa5I7xdjYGJOTk7z66qvi8PawPu9bzT0J5KSM0+LiIgqFQjQrBINBCgsLRXeLZLYYCARYWlrC7/czMjIiLAoy/SV+K6SGAEnfV1VVRTAYxG63C+fvaDTKysrKlkfzd4I01HtmZoaxsTF0Oh1arRaj0XhDGV3SHEgi6LW1NZHNk0w1Z2dnxYD1bBlXBddmwkpZOJvNJrKuQ0NDjIyMiHFlDwKxWIzJyUk0Gg29vb03TCS4evUq4+PjjIyMEAgEsvbelkilUkLfaTKZbuuwIjWGLC0tMTo6KjIx0pD00dHRjBotd6dImUbpedZoNGg0GsrKykRWvqOjg9bWVmw2GwaDgaWlJbFfSJrXbL4/rker1YoMtdlsRqFQCAuqTJ0JvhlSw2JhYSGhUAiNRkNeXh6RSIS1tTUmJyfx+XzC5HdsbIyVlZUH5sCajdxTUZbkdH716lUxxqK+vp6/+qu/IhqNCuPIS5cusbKyIm6UZDKZ9Q+4NEf16NGjeDwePve5z9HZ2cmXvvQlrl69yi9+8YusKqVeTyQS4eLFi+JBrq6uprGxkYMHD9LW1rbhe9fW1piYmMDtdjM6Osr4+Djj4+O43W7RnSoFd5KWKFsoLy/nscceo7a2lry8PI4dO8bw8DB///d/j8fjeWCCOLg2T/b1119HoVDwD//wDzd8XTL5ll7Q2XQdNyMej3PixAmWlpZ45plnbqsZKRqNMjMzw/nz5/nmN7/J3NwcTqdT3NeZLBP4OEgmxtJg+ZqaGkpLS/nzP/9z0bmq1+vR6/UEAgFcLhc//OEPhfZ1/aD1B4Xq6mqefvppdu/ejclkIhQK4fP5+MY3vkFvb2/W6AGlaks4HKapqUkY3I+OjjI2NsY3v/lNRkdHcbvdxGIx4Y+aY+u4p4Hcen2IhNPp5PTp06KbcWxsjMXFxQdOLyGte3h4mHg8Tnl5OWq1mqamJtRqtbBbsFqtLC0tZcV4rvWkUilisRjLy8tMTk4SDodZWVkhFosxMjKy4cGWgna/38/8/LwoNQQCASKRCMFgMGvGVF1PIpEQhw9AaMnC4XDWbNy3y/pAJBvlAB8XyTMxmUzS2Ngohp9vRjweZ3BwUHTg9vX1MTc3J+6FBw2pUWt6ehqLxUJZWRl6vR6j0SiC+HA4zMLCAsPDwywsLNDT08Ps7CyRSCRrslMfB6n5w2w2iyaAubk5kaTIFhKJBIuLi4yMjHDs2DFhaj0xMcHCwoLQQQaDwax6Zz3I3Pc2SafTyVe/+lXx79n48r4dpGaGY8eOcfr0aRYWFmhra+OP/uiPeOKJJ/jlX/5llEolCoWCc+fOsby8nDUaColEIiG8hgYHB4Fb++msv9YPynWXOrTLysooKipicXERl8v1QL6oHjZisRhvvfUWvb29YvLI008/vek9Hg6H+da3vsXk5CQTExPi0PKgEo/HSSaTnDlzhkAgwO7du8X4MYnR0VH6+/v50Y9+RG9v7wOvmSwqKmL37t2UlJSQTqcZGxtjYGAg6wL5WCzG+Pg4ExMTvP/++xu+9iBk2h9EtsTv4mG6ESTNzOjoKMFgEKPRSGFhIXa7HYPBgE6n4+rVq1y9ejXjLTZuRTbNErybjI2N8bOf/YzLly9TWFjIhQsXcDqdD0wJ7WFmvd7tvffeo7+/n56enk0lEZFIhPPnzxMIBPD5fA90wCKRTqcZHx8nGAzyD//wDxiNxg1fd7vduFwuhoaG8Pl8D1TFZT0qlUrs6SUlJRgMBuLxOC6Xi8nJyazJ1/6wAAAgAElEQVTdC3JB2+1z7tw5vvrVrzIyMoLT6bzv73LZrS6UTCZ7YK5iOp3+WG2V92LtarUam80mtGRWq5XCwkL+5V/+hXPnzt2z8urHWfvDes0f1nXDw7v2h3Xd8PCu/V6s22Aw0NDQwLPPPstXvvIVUXb+2te+xsmTJzl16hSBQOBu/7W5a36bPAzrzk4H2iwlkUjg8/kYHBxkaWkJrVaLRqNhZGQk48dP5ciRI0eOmyPt3yMjIwwMDHDq1Cl6enoeCj1pjq0lF8jdR1KplBh74nQ6t/rj5MiRI0eOT4hk/iyZmk9NTXHp0iUcDgdzc3Nb/fFyPATkSqs34WFde27d2U/uXr89HtZ1w8O79nuxbrlcjk6nw2w2U1lZKUygpekV94rcNb89HoZ15wK5m/Cwrj237uwnd6/fHg/ruuHhXfvDum54eNf+MKz7loFcjhw5cuTIkSNHjswle0cL5MiRI0eOHDlyPOTkArkcOXLkyJEjR44sJRfI5ciRI0eOHDlyZCm5QC5Hjhw5cuTIkSNLyQVyOXLkyJEjR44cWUoukMuRI0eOHDly5MhScoFcjhw5cuTIkSNHlpIL5HLkyJEjR44cObKUXCCXI0eOHDly5MiRpShv9cWHYbTFzXhY155bd/aTu9dvj4d13fDwrv1hXTc8vGt/GNady8jlyJEjR44cOXJkKblALkeOHDly5MiRI0vJBXI5cuTIkSNHjhxZSi6Qy5EjR44cOXLkyFJygVyOHDly5MiRI0eWcsuu1Rw5cuTIkSPHRmQyGTLZrRsnN/t6Op0W/3vQkH4n1687nU6TSqW26FNlBpv9Tu4muUAuR44cOXLkuA2Ki4sxm810dnZiNptv+j3FxcUUFhai1+tJp9Mkk0mWlpZYWFjggw8+YG5ujrGxsfv86e8dtbW1PP744zQ3N9PU1ARAKpUiFAqxsLDA3/3d37G8vLzFn3JrMJlMNDU1kZ+fT2FhITMzM3g8HlwuF6urq3fl78gFcvcJuVyOUqlEr9ffcGpJJBKkUilxUkulUqRSKeLx+AN5csvx4CKTyVCr1Tfc40qlEpVKBXx4GpXu8WQyueEZyJEjU5DJZCgUCpRKJWq1Grvdjt1up6urC5vNtumfqaqqoqKigpKSEvLy8gCIxWLMz88zOjrK1NQUoVDofi7jnlNYWMjOnTvZu3cvu3btAq493z6fj+HhYTQazRZ/wnuD9F6/fr9TqVTI5XLkcjlWq5WGhgYsFgt2ux2tVotarSYQCOQCuWxCoVBQWlpKc3Mzr7zyCkajEbVaDVx7qQ0MDOB2uwmFQkQiEebn53G5XHR3d7O2tkYkEtniFeTI8dGo1WoMBgOHDx8mPz9fHFoAdu7cyY4dO8RBJRwOs7KywvHjx5mbm+Py5cv4/X4WFxe3eBU5cnyIyWSivr6eHTt2sH//fqqrq7FarZhMJrGHX49arUalUqFSqTYcTORyOQqFArVajVL5YL16y8vLef755zGZTGLND2oJeT12u53t27djMBgwGAzAtUPrnj17KCoqoqSkBK1Wi9FoFPeEz+fD5/PxP/7H/+DcuXN35XNsyd0kl8vR6XTiYdBqtcjl1/ouYrEYiUSC1dVVYrEYKysrWX1DSBmKqqoqmpqa2LZt24ZADq4Fei6Xi3A4TCQSwWw2Y7FYWFlZYXV1lXA4LL43HA4Tj8eJxWIkk0nW1tYy/vejUqnIy8tDrVaj1+uFXkK6xsFgkGQyucWf8u4hl8spLi5GqVSSTqeJRCL4fL6t/lj3FLlcTmlpKUVFRbS3t1NQUIBOpxNf7+joYNu2beJeDYfDBAIBvF4vJpOJcDjMzMwMKysrJBKJrL0fFArFhntdeqknk0nS6TTLy8tEo1ECgUBGP7MPO3K5HKPRiN1up729nfb2drZt20ZxcTH5+fkbMsnXX0dp315PMpnE6/UyMzNDIBB44A7nGo0Gi8VyQ/D6oKJUKikoKKCyspK2tjby8vIwGAwkk0nkcjmVlZWYzWbsdjs6nQ6DwSAyu0ajEYvFQnNzM16vF7fbTSwWIxqN3vnnuYtruy0UCgUGg4Ft27bxK7/yK6KmrtPpUCgUTE9P4/V6OXPmDOPj47z55pusra19okVuJRqNBqvVyp/92Z/R2NhIfn4+CoViw/c0NTXR0NAgXnLJZJJIJILT6SQWi4mHPplMcubMGebm5piZmcHv99Pf3080GhVl2Ex7iJRKJcXFxTzzzDPU1NSwc+dOUqkUiUSCCxcuMDY2xrFjx/B6vRn32e+UvLw8Xn75ZYqLi0mn05w/f57/+I//2OqPdc9QKBRotVr+7M/+jEceeYSGhgY0Gs0NpYb1Qm+dTodWq+VXfuVXiEQi/NIv/RLvvPMO3/jGN3C5XFkb+JrNZj71qU9RV1fH7t27qaiooLy8nEgkQjgc5sc//jEjIyP88Ic/vGtllRx3H6PRyBNPPMHOnTv5gz/4A3G/BgIBPB4PV65cYXFxEY/HQyKREH8ulUrR29tLb2/vhp+XTqdJJBLE43HC4fCGP5Mj+7Barfzar/0a27dv5/nnn0ehUCCTyQgGg4RCIc6fP4/D4UAul2OxWOjs7KSgoACr1Yper0en0/G///f/xuVy8b/+1//C4XAwPj5+xwfY+xLIyeVy1Go1dXV1mEwmysrKqK+vp7W1lcrKSoqLi9FoNMjlcpLJJHl5eYTDYSwWC6lUCq/Xi9PpzMoNXtK7SWlV6eW2PmjZLM0u6Yyk7Jv0s1ZWVrDZbJSXlxMMBikvL8fj8TA6OkowGCQYDN6fhX0E0gmlsbGR0tJSurq6sFgs6PV64vE4MpmMqqoqDAYDgUAAp9PJ2NiYCEqzGblcTl5eHhaLhcLCQpxOJ0aj8YFY23oUCgUqlYrS0lKKi4upqamhrKyMvLw8cU8nk0mi0SjLy8tEIhFRXjKbzchkMpGZKioqoq6ujn379uFwOJibm8PtdrO6uprxAb60v1VWVlJWVsauXbsoLy+npqaGgoICoYnR6XS0tbWh0Wg4f/48Xq+XpaWlrf74W4qUpTCZTBiNRpHJvnz58pYGuiqVivLycsrKyigoKACuaZlHRkZwOBwMDw/j9/vx+/3i5SvdpxMTE3g8ng0/b73+OVuzzTmuPetFRUVUVFTQ2NhIZWUl+fn54r3u9Xrx+/0MDg4KmYjJZMLv92M2mykuLqauro6ysjIKCwtFGbagoICVlRVCodAd6SfvSyCnVquxWCz81m/9Fs3NzRw6dEiIAZPJpCg7pNNpbDYbNpuNhoYGAF5++WV6e3t57733+NnPfsbp06fvx0e+a0gZNZ/P97HKKSqVCovFcsN/r6qq2pB5S6fTnDt3jr//+79naGiI4eHhu/r57wSZTEZNTQ1VVVX85V/+pbhpvV4vDoeDlZUVgsGgKFXs3buXqakp/u7v/g63243f79/qJdwV8vLy2Lt3L4FAALvdLrQRDwparRaTycSTTz7J7t276erqorS0dMP3RCIRPB4PU1NTTE9Pi9Povn37UCqVnD59mvz8fB599FEeffRRtm3bxsmTJ+np6eGNN95geno644NfjUaD2WzmpZdeorm5mWeeeQaNRoNCocDv97OwsIDRaESj0fCZz3yGHTt24HA4GBkZ4YMPPniorRk0Gg0mk4muri6am5t57rnnyMvL4wtf+ALj4+Nb9rl0Oh07duygqakJpVLJ6uoqy8vLvPrqq7z22mu3zKplYmUkx91BrVazc+dOWltbOXLkCIWFhRsqDw6Hg/7+fr773e8yMzMDXEvU6HQ6cbj/i7/4C774xS+KLO9/+2//jbGxMXw+H9PT0wwODn7sz3XPAzm5XM6uXbuoqamhs7OT8vJy1Go1oVAIp9MpUoq6/8/emwW3eaX5+Q/2HSABAiS4L+C+iJIo2rKklq32eJnxTOxpd7o7NVOZSSVTSU0qF0mlUqlKqpKbXCRVc53pqXS5p3p63O1ud3vclte2ZctaSG0UKXHfF4AgCAIEiYUACPwv9D+nSS1u7QQkPFUslU2Q/M73ne+c97zL7zUY0Ol0dHV14XA4qKiokO7KdDpNOBzOy/Cq8Fro9fpbKnfGx8eZmJiQYxMTQqlUYjKZdsXZhSdDfF8YwgAOh4Pu7m7C4fCeG3Ki7P6VV16hoaEBi8VCKBTi9OnT+P1+JiYmZKg8HA5TU1ODx+OhqamJY8eOMTMzQ19fH8lkkmQymXcLokajwWg0UldXR3V1NWq1GrVaLRNeM5kM1dXVu075Xq+XaDSaN4n+opLP4/HQ29vLc889R0tLCzqdjng8TiwWY2tri1AoxOzsLP39/ayurrK2toZer5e5cxqNho8//hir1UosFqO8vJyGhgZaWlqwWCxcuXKFYDBIJBLJSS+GqFg7cuQIdXV1HDp0CKfTSTAYlIeW5eVlVlZWKCsrw2638/LLL2MwGGhsbCQWi/1eLbJ8xGKx4Ha7KS8vx+12yzD78vKyjC6Ie+d0OqmqqpL5lYFAgKmpKRKJxJ6OQeTIibk6NzfHxYsXmZ6eJhaLySrrAk8HCoWCsrIyXC6XTJ2w2+1otVqi0Sjj4+OMj4/T39/P7Ows6+vrcn6Ig6hIQdm5lokceo1GQzqdvu+Q+2Mx5J5//nl6enro7e3FbDYDsLa2xuXLl3n//fc5efIkNpsNm83Gv/k3/4aWlhZcLhcqlYpMJkMsFiMQCORlgqhGo0Gn00lPxM6F++rVq/ziF79genqaUCgkv6dWqykvL+fEiRO0tLSg1WrlwxYG3c5wbElJCc899xyzs7OPe3i3UFVVRUtLC9///vdpbm7G5/MxMjLC3/3d37G4uLjrGr1eL21tbRw4cICGhgb++I//mIGBAcbGxmQBhPDW5gtarRaLxUJbWxt1dXXA715g8f3nnnuOhoYGlEol0WiUc+fO4fP58iZPUBwkOjs7+fM//3MaGhooLS1lc3OTWCzG8vIy4XCY8fFxzp8/z49//GP5HDUaDQaDQYYj33nnHcxmMz6fjxMnTtDe3s6+ffvYt28f77//vpRqyFVDTqfT8Yd/+IccOnSI7u5utre3GRoa4tKlS7z99tssLCywtLREU1MTNTU1HDx4kKqqKjo7O4lGo0+kISd01o4ePcrhw4dlqP3ChQtSS0xIMdXV1dHZ2SmlaH70ox9x5cqVPZfnUCqVWK1WWYk4MTHBL37xC8bHx/PSoVDgwRBRpqamJr773e9SVlaGSqUikUgQCoX46quveOedd5ienmZtbW1XFCGbzcpihng8fouxJuRL4vG4POjcK4/ckFMoFJhMJpnkH41GmZ2d5fLly/ziF79gYmKCeDwuE+BtNhtFRUWy6KG/v59r165x7tw5/H7/o77cR0I2m2Vzc5NoNEo2m2VjYwOfz8fU1BRzc3MsLy8TiUR2eeREvtuXX37Ju+++i0qlkpufRqOhsbFRGgeBQIBr164xMjKyZ2MsKSmhrKyM1157jWeffZZMJsO1a9f45S9/yczMDFNTU7cszjMzM8TjcWZnZ9FqtbS3t+N2u6mqquLMmTO8++67RKPRvFw4lUolCoWCbDaL1WqltbWVxsZGGhsbqaiokHkVsViMkpIShoaGGBoaykmDZSdqtZrKykqOHz/O0aNHaWhoIJVKMT09zaeffsr8/Dxzc3NEo1HW19dZWVnZZYyLQp5Tp06hVqvZ2tpie3ubixcvymr2np4ePB4PFouFoqIilpeXczK82t3dTWtrK/v376empoalpSW8Xi+/+tWvmJ6eZmZmRs75RCLBxsbGE+nF0Wg0tLe3U1JSQnt7O+Xl5bS1tREMBvnqq68oKSnBZDLJwq/NzU3pkRMH3bm5OZaWlrhw4QJXr17NuUO7MDSfxOf3IKjVaiwWCxaL5RYttWw2SzgcJhQK5fV96+jooKGhgRdffJHa2lpsNhuJRIL5+Xmmp6fp7+/n4sWLTE1NsbGxcdtKZkCqcExMTHDx4kUaGxux2WwyAtfR0SELPu/1QP9YcuR0Oh16vV4m7y8uLjI+Pk5fXx+xWIzt7W1p5Oh0OrRaLV6vl8nJSX7zm9+wsLCQtyrYYmKLUCHckBBZWFjA6/WysrJCOBzeJTECEA6H8Xq9t/w+nU6HwWBg//790rsZCAQYGxvb08XPZrPR0NBAT08Px48fZ3BwkJmZGT7++GOWlpYIBoO3TE4hueL3+2VYraamho6ODra3t/nwww8fuCx7LxBJzWK8JpOJmpoajh49yvPPP7/rc7FYDLVaTTqdRqfTScMmF1EoFGg0GlwuF0ePHqWzs5PS0lJmZmbwer2cOnWKoaEhpqen73iyFMU/O1MAhCFoNBopLi6msrISj8eDyWTCZrPdUuWdK3g8Hg4fPkx9fT0Oh4OZmRkmJiY4e/Ysfr9/18FTFC1lMhkUCoU0ZES6Rb5WMYrQUGNjI3V1dbz00ks4nU5qamr48MMPGRoaoqysjKKiIpqamtDpdNITqVKp5Nfa2hrT09MPXL33sMYkxFx3tp26+ZntXM925nnng1f9YaFSqbDZbJhMptu2odrY2NgVZswnxHNvaGjgmWee4dixY1RVVWEwGIhEIszPzzM0NMSHH37I0tLS73U0CVk1r9fL2NgYbrcbm82GQqFAr9dTXV1938VPj8WQEy7F7e1tzGYzPT09mM1mFAoFCwsLLC4uMjU1RSwW4/z58wwPD3P27FmCwSBzc3N7ni/xIFitVlwuFx0dHdTW1qJUKlldXeXcuXOMjo7uyhu5G4R+0cDAgNzgREn7XrwsCoUCtVpNQ0MDr776KlVVVaRSKT7//HMGBgZYXFyUWoA3k8lk2NraYnBwEEDmT4oS7kQikXcb3NbWFpFIhMuXLxONRmlvb8fj8fAXf/EXsqWPMOrn5ubIZDLU1NSgVqv5L//lv3D+/Hk+/vjjnFz41Go1drud+vp6/uAP/gCDwcDm5iZff/01586d48qVK/h8vvv2nomwbCwWQ6lU8uabb3Lw4EH+23/7b8zPz+eMXIfId+zs7OTo0aPE43HGxsY4efIki4uLZLNZmeCcSqWkkW42m2U4trOzk7KyMqqrqzl9+jS//OUviUajebXWKRQK6uvrqaqq4q/+6q+oqqrC6XSyuLjIO++8w9mzZ+nr65PvsNFoRKlUkk6naWxs5F/9q38lc0Q/+eQTPvjgA+bm5vbUiFMqlVRUVNDQ0EBDQwMulwuAnp4eWVm4cx5ub28TCoW4cOECS0tLjIyMEIvFZAjtSTfqHA4H3/3ud+np6UGtVsu8bbixL508eTInPax3Q1lZGXV1dbz44oscO3aM6upqjEYjkUgEr9dLX18f165dY3x8/K7eW3HAF3ng3d3dVFRUADeKazo7O++70O+RG3LZbFYqtguXut1up6qqin379mGz2WQbE1HKHQ6HmZqaIhwOEw6HH/UlPjIUCoVsy2G327FarWQyGTY2NmTo+F4XbuHRyJXKTrExORwO6uvr0Wq1RCIRZmdnmZycZHNz844bu9DMW1tbkx47EcIQorC5aNB8EyJFQFTmwo3kb+E9hRvzfH19ndnZWZRKJfX19ZSVldHT08PKyoo00HNt7CqViqKiIqlYnkgkWF9fZ35+nuHhYVZXV2/xLN8LwggWh77a2loMBoNURReh6r1GGGVOp5PS0lIWFhZYWVlhZWWFaDRKSUkJWq2WbDbL1tYWyWSS0tJSnE4ncOO5CsFkk8nEysoKBoMhrzzP4gDndrupq6uTec3JZJLNzU2uX7/O7OwsgUCAjY2NXUUOJpMJu90uDd3l5WWmp6cZHx/Pidw4h8OBy+XCZrNhNBqBG6kj4nrFeym04VZXV0kmk1gsFuLxOOFwmI2NDeLx+C7x9nw7lN4NwpPkcrl2GXFwY54vLS2xsLCQs1GG2yEOYW63m6amJmpra6mqqkKn05FKpVhZWWFpaYm5uTl8Ph+RSOSuf7fIl7vZ8SI82/fb8eORG3LpdJq33nqLDz74gNnZWRobG3nppZcoLS3l5ZdfJpVK7ZrsBoOBQCBALBZjYmKCr7/+Ouc2tLtFrVbzZ3/2Z/T29uJ0OqUBNjY2xqeffprXRqrAbDbT0NBAd3c3hw8fpr+/n9/+9recP3+esbGxu/I2xuNx4vF4TmzSD4qoWm1sbKS2tlYaZcIIyWaz/PCHP+Tzzz9nc3OT8vJyOjs7cTgcfOtb32J+fp7i4mK5EeQSZrOZF198kf3796NQKGRrrbNnz3Lp0qUHNkTW1tYYGhpicnKSxcVFnE4nRUVFaLXanAqvOp1OOjs7qaysxGKxcObMGaanp2lvb6e0tJTnnnuOYDDI8PAwyWSSVCqFx+OhpKSEdDrN9PQ09fX1mEwm9Ho9LpeLoqIi4vH4nhsyd4vZbMZms/GDH/yAgwcP4nA4WF9f57PPPuPChQv8/Oc/l3nBO9dvk8nEt7/9bfbv309nZyfnzp3jZz/7GVeuXCEcDu/5Wq/RaDh+/DidnZ0YjUY573Q63S29ggVlZWXU19eTSqVIJBL4fD58Ph/Dw8N4vV4GBwdZXV1lamrqiTTmnjQqKyt57bXX6O3t5YUXXsBms6HX6xkbG2NpaYmf/vSnzM/Pc/Xq1XtyxIgixc7OTl555RV5sBPOrg8++OC+pEfgMYVW19bW2Nra4tq1ayQSCcrLy+VCIHIliouL0ev18sURvRrzFZ1Oh8lkory8XIYMxQldpVLJU7vZbJZeK9GbMhqNsr29fd8VLI8Tg8EgZVKUSiVer5fh4WH5zL8JIctht9vlz6dSKVZXV2Wrpnwz7oR6t9VqlV44YcCtr6+zvr7O9PQ009PT0pMJv2sqbzAYbglR5ALixCi8y0LFfGFhgVAo9FCMTpEiEI/H5dzRaDQ4nU6ZS5oL88Fms1FfX09RUREqlYpYLEYikcDtdlNZWUlVVRUWi0WKIW9tbaHT6djc3GRpaQmdTkdVVRUajUY+dyG1lA8oFAqcTifl5eVUV1dL72wwGGR0dJSZmRlZubfTE6PRaDCZTHg8HtxuN9FolJWVFaanp3PCiIMbYzObzTIMLqIGiUTijt5m4Z00Go04nU6pGaZQKCgpKUGpVMqin0gkQiAQyIl5XOD2aLVa7HY7TqdTSn+JfL9AIMD09DSLi4v3lfsn1lHRsgtu7A/CM722tnZfc+OxGHIi1PTzn/8ck8nEb37zG2pqauju7pb6cUeOHKGmpkZW8wk19HxFLOr19fVUVFTICr1EIkFtbS1//dd/TSAQIBAIcOnSJZaXl+nq6kKlUnH16tU7FjvkGmVlZbz++uu43W5mZ2f55JNPePfdd+/qpCLCdM8//zwdHR3odDqWlpb49NNPGRwczBvvhEClUnHo0CHa2tooLy/HarUCN/JoUqkU/f39fPnll1y6dImVlRWOHTtGS0sLJpNJnvZzFa1WK6tva2pqgBvyMWfPnmV5efmh/A2ho5RIJGRhgE6n4/nnn6ekpISFhYWcCD82NTXxz//5P6e2thb43b05cuQITqcTlUqF3W6nqKiIWCzGxsYGb731FhcuXGBwcBCXy8Xhw4fz8qAqigCOHTvG888/Lzu2DA4Ocu3aNd5++22CweAt779CoZC9KUVLoy+++IIzZ84wODiYs4fWdDrN5uYmY2Njt7TdEmg0Gux2O5WVlezbtw+73Y7D4ZBtF7e2tlhbW+O9995jcHCQf/iHf5AtFQvkHqKwRRTiwI0w8fLyMrOzs4yPj7O6unrPRtzO1KFEIrErRJ9IJJienr7vtfSx9VoVsWG4Ua0oBBU1Gg1arZb6+npKS0ulaKo4mecbotKlvLyc1tZWbDabbLclGmqLieJ2u4lEIjgcDsLhMFVVVSiVSpxOJ0tLS5w9e5bNzc2cabu1E2FsFxcXU19fj16vJ51OS2P1m3IihBfC4/FQV1eH2+2mqKgIpVJJLBZjZmaG1dXVxziah4fFYqG4uFhqA2WzWSmrcPnyZQYGBlhbWyOTycjFQnxOpBnkmnaeqFYVYUCr1So9FOFw+KFvwjdX/hkMhls0GPcSnU4nQ75wI1xoMpmIxWIsLi5y/fp1meMpxK+HhoZYWlqipqaG6upqNBoNmUyGRCJBLBbLq/6bO/OcI5EI4XCYs2fPMjExcUfvrCgiqKmpwWw2s76+ztWrV5mbm5NGey4g8rr0ej0XLlxga2uL2dlZlpaWmJmZue3PiHXd4XAwOjoqDTkhu1JUVERxcTEdHR3o9XoCgQDLy8ssLCywsbGRM0U898Pm5iaXL19Gr9dz5MgRWekLN+5LZ2cnANeuXctZYx2QDqWysjKam5upq6vDbrcDsLKywtraGmNjY0xOTu4ywu4FnU6H1WqlpKRkVx6teP+TyWTuCgLfTDKZZHFxkcXFRQYGBoAbL3l3dzcNDQ0UFxeTTqcJBAKEQqGc2tDuBiGW2tHRwYkTJ3C5XNLbotFocLvdALIFGfwu50J0sVhdXeXq1atEIhHm5uZy0pATFYzV1dUcPHhQ9ku9OZxyO0To4YUXXqC3t5fGxkZZ0RmJRLh48WJeeCNvRqFQSJf8Trf5lStX+K//9b8SDAZlkcrNSa3i5C+q3XJlYxOIdlwejwez2Sw38Psp2LkXxIFBdAfIBUQ6gLim4uJikskkPp+P+fl5/uf//J8yx1EYpWIc//k//2eZf5VKpQiFQqytrREKhfKiYlUcPiorK2lubmZgYID5+Xl+9KMf4fP57hgaUqlUUnfPbDYzOzvLP/3TP+3qVZoLpNNpBgYGmJubY2pqCp/PxxdffCHFyb8J4a3s7u5m37599Pb2UltbS29vL1arlRMnTtDT08PBgwc5e/Ys7733HhMTE3ltyAUCAX7yk5+QTCb5F//iX8g0AbjhqX7jjTdoa2uTmqC5iFKplCkxL730Eh6Ph+eee07uSaOjowwODvLb3/5Wap/eDxaLhZqaGhoaGmhqasJkMpHJZFhbW5MFM/f7Ljx2Q+5mhC6tCagAACAASURBVEdOdD4QC4VoopwrlWr3itAcAuRCLjyNInwkRH51Op3UJxKnu7q6Ov7kT/5EtjcSeTi5wk5vwubm5l2FvEQhQHd3N/v37+eZZ56hoaFBqmPbbDbpmcqlxf1uqK2tpby8nN7eXtkYfX19nYsXL3Lp0iWCweAtC8BOwySRSMjqR9ECKFdQKpWUlZXJfBGRKhEKhR65IQfkjAF3J0KhEPPz84yMjBAIBAiHw1I6R3goKisrZWGL6N8p+jDHYrGce+Z3QoSHLl68iFqtZnBwkOXlZfx+P7FY7LZrtcfjoby8nOeeew6n08nnn3/O6OiovE+5xPb2Nj6fj3A4zNra2jcKvN6M0I5cWloinU6zvLyMw+HA7/fjdruprq5Gq9Xidrs5dOgQBoOBTz/9lP7+flZWVnLuXtwNIofwTgdPYdzmMqJpgcvl4siRI7I3uBDcn5yc5OzZs8zOzrK2tnbPe5OIwFVVVfH888/LQieVSkU2m5WH4pulbe6FnDDkTCbTLkNOqVRKQy7fuFlAUshRqFQqUqmUlFdYX19Ho9Gg0WiwWq3o9XrZl1P0Wf2TP/kTAM6dOycNp1xBxPVFDlAqlZJj3ml879yERUjqwIED/Omf/im1tbVSuT8Wi2GxWGRJf655pH4ftbW1dHd3c+jQIRobG1EoFPj9fk6dOsXAwMBt5WJ29s8VhlwutqJTKpWUlpbKtjTpdJqNjQ1pyD1qcvkgJ4pYFhcXOX36NOvr61LcHJAFDTU1NTQ1NdHW1iZz6xKJhKzYzrVnfieEIXflyhWWlpa4fv064XD4jpubQqGgsbGRzs5Onn32WRQKBW+//TZTU1Osr6/n3LPNZrMP1EEom83KqtWBgQE0Gg2JRIKGhgaee+45Kioq6OzspKSkhLa2NtbX11leXpa6mfmIkMS6HflgyMGN9Ain0yk9cTsjJlNTU7Kz1P08I5GaUlVVxfHjx6UhB7/r9hAOh1lfX7/vdWDPDbm2tjaeeeYZmpubsVgsskTX5XLh8/ly/jR+MyK/6dNPP2VoaEjmyalUKpLJJMFgkFgsRiQSkYacxWLBYDBQXl6Ow+Hg+PHj2Gw27HY7Ho+HY8eO0dfXl1NyJULM1+fz8eWXX9LY2EhbWxuvvvoqdrud6elpEokETU1NskLZ4XBQU1OD3W7HYrHw9ttvMz09TUNDA263m7Kysr0e1j0jPKkHDx7klVdeoaSkhHg8zieffML4+Djvv//+LcaOSqXCaDRy9OhRurq60Ov1OdmCSqBSqaivr6e+vh61Wk04HGZiYoJAIPBQ/444AAlPtUKhYHt7W1b55oqXdm5ujk8//ZSenh7q6+ulVEFLSwvLy8uMjY3JA1pFRQUVFRW0tLRQWlqKSqUiGAzKKm2j0YjRaMRkMkmpklxGeJ28Xi/BYPAb244VFxdTXFzMyy+/zOHDhwkEAiwtLXH16lVWVlZyzoh7FGxvb0uR+8nJSerr64lGo5SVlVFVVcUf/dEf0dXVxTvvvMPo6ChXrlzJq1Cr0JErLy/flR+XTwjHUVFR0a7rF0UJwrt+P/NVrAN1dXXs27eP7u5uWQQnul5cuHCB4eHhB1rf9tyQKy8vp7u7G6fTKVufKJVKLBaLtFrzCeFqnpycZGpqioWFBcxmszTk1tbWiMfjbGxsyKR/i8WCXq+nrq6OyspK2tra0Gg0VFRUUFRURHl5eU5WuKXTadbX1xkbG6OkpASLxUJraysqlQqDwUAsFqO3t5fi4mJKSkpwOBzU1tYSDAZZXV3l4sWLnD9/npdeeglAuuh3trfKdVQqFVqtlvLycpqbm9HpdMRiMQYGBrh+/TojIyO3bM7C+Kuvr8fj8aBWq6UnMleMlZ2IPDC73S5Dq+Fw+KF7kUQ6glqtlkUgmUyGQCBw32X5j4JgMMj169epra2lurqa5uZmamtrSSaTeL1eVCoVer0ep9MpOwS43W5MJhOLi4vEYjF5uNPr9RgMBvR6vaxuznVEOOj3YbFYZC5da2srH330EdPT03i93nsSUc1nMpkM8/PzaDQaNjc3CYfDMj+6qqqKxsZGPB4PExMTZDIZRkZG8kpTU4SKxdpwsyEnBLFzeTzCkBN9T+F3EaedaRL3MgZxIDWbzTgcDhoaGqTBK35PJBIhGAwyPT39wKLJe27IuVwu2tvbsdls8v+J8n0hRZKvCDf7zZWJ4gQrDJdUKiV7WGYyGemZi0ajjI2N8d577+Hz+fZ4NLdneXmZd999l2QySVVVFW63m/r6ep5//nkymQxGo5FMJkMkEmFxcZEPP/yQgYEBhoaG8Pl8pNNpOjo66OjoYGNjg9XVVYLBYF5Jj+xsCH/hwgUWFhb49a9/jdfrvW3eU1FREaWlpdTX11NTU4NKpSIcDnPmzBmmpqYe9+XfE5lMBrPZTFtbG0NDQw/1dzudTurr62lpaaG2thadTkc4HJaHolwxcoeHh1laWpIe9meeeQaHw0Fvby/JZJLjx4/LsLloT/j+++8zMzPDhQsXsFgs/O3f/q3sIOByuaRgeL6G127Ht7/9bf7iL/4Ck8nEyMgI/+///T9GRkZyNun9UZJOp2Xu3cLCAkeOHGFjY4OWlhaqq6v5wz/8Q7q6uhgcHJTN5nPZ+BHYbDZOnDjBvn37btFCTKVSfPLJJwwNDeVs6oAwtP7qr/6K9vZ2rFar1Iw7deoU7733Hn19faytrd11DuvOjk7/8l/+S2pqaujq6pJVsOLQ/sknn3D16lU+/PBDVldXHyhHds8MuZ0iijabDY1GI096opLrTj0684lvKrkWFW0ij0As7MKDJ1qb+Xy+nF38hJL57Owso6Oj1NXV4XK5pDSD2Mh8Ph8zMzMMDQ0xNDTEtWvXgBv9F81mMxaLZZd8ST4kfsPvPLDr6+usrKwwOTnJ3NwcXq+XtbW12/6MxWLB6XRKT2wmkyEej+P1enOm9do3IU6aQlbnYb2jWq1W5sYaDAZ5b0UxQK6sBZubm2xubjI5OYnT6aSkpIRoNIrRaJTvs5BbEu3nRkZGGB0d5dq1a7LDgxAH1Wq1stDpSUB4nEVIeXZ2Fq/Xy9zcHIuLi3mX//ow2NmuLRaL4XK5GB4exuFwUF5ejtPpRKlU4nK5ZAu/XJnv34TQ0BOyWjvJZDJ4vV6WlpZy5hB2M1arFafTKb3rSqVStllbWFhgaGiIQCBw155yEVUoKSmR+ZDV1dU0NjZKb58oEJycnGRsbAy/3//Ajos9M+SECnZpaSnFxcVotVpSqRT/9E//xMjICG+//XbOlaY/KtxuN263m//0n/4Tra2tlJaW4vP5+PDDD7l06RKhUChnFz+x0X788cecPn1autmF7IroOzg5OSlV7m9+KUQCtdDTicViOSH8ejeIvKaPPvqIiYkJhoaGCAaDrK+v3/Fnenp6ZBNuUcm8srLC1atXH3re2aNASOzo9XqZ3/coDG9RTJOrxS8ff/wxX331FW+99RZFRUUcP36cdDrNxYsX5YFEGKGRSETO6ZKSkj2+8keLy+Wiq6uL5uZm7HY777//PufPn5f6iU8zwsAXeVHJZBKn04nT6cRut/PSSy9RXl7OT37yk7xZA++EENH1er05uY8rFAoOHTpEZ2enDA3DjdSJL774grNnzzI4OHhPa5sI0b7yyis0Nzdz6NAhaeSKtCGhKXry5EnGxsYeynPeM0OuuLiYffv2UVFRgV6vR6FQkEqlmJubY3JykmAwmLNeqIeFyJGrr6+nsbFRlj0rFAo2NzcZHx/H5/Pl/OKXzWZl5Z0Io4qqXNFv0e/33/UJMx9OojvJZrOsrKwAN8QjvykBHG545EpKSmRunEiqjUQiORtaE2LF8DuPnCjICYfDD8WQMxgMssUR3NCoWlxcJBqN5qQSfjwel97j9fV1rl+/TiqVYmZmhmQyecuX6FQBt4oePwmI6jy73U5HRwdWq5VQKMTi4iJzc3N5ZZjsbBu5s0PH1taWlI96kHVZdC+ZmJjg8uXLHDp0SBbHhMNhmZKSyzmTIpKk0WhuqUwVkbVgMJjTYeKd76ZAoVBID/ndirOLIs2dee51dXWyX6/oJ59MJpmZmZH9dx/Wer9nhlxzczP/4T/8B+rq6igqKpIu576+PgYGBojFYjlvwDwoZrMZu93O66+/zrFjx6iurkav17O1tYXX6+X999+/Y3guVwmFQtItvZO7eRl2tkUR0i35wtzcHPPz87/3moVosNvtlukE8XicSCTCyspKTqqfi5yRjY0NstksRqORyspKGhoaaG9vZ3h4+KEcukpLSzl8+DBut1ueXIeGhlhZWcnJQ93OHrqRSETmsT6JRtrdIETCOzs7+bM/+zOi0SiXL1/m3LlznD17NmfzpG6H0Wjk0KFDVFdX88wzzzA1NcXg4KDMc1taWnqg8Yi0AZFD9t//+3/nwIED7Nu3D4PBQGlpKUqlkmAw+BBH9fAQxo5er8dkMknNNcHExARTU1OMjIwwNzeXk6ky2WyW+fl5GU4Veq9ms5mmpiYGBgYwGAy3jSLdjBC5f+ONNzh27Bj79u3DYrHI/uHCuREMBvnss8/4+c9//lD3t8duyIm8ELPZLFuYAPj9fnw+3+8tac8VhEiqwWCgqKhINjxPJpMyVBYOh9na2rqlClNUtDQ0NNDd3U1zczOlpaVotVqi0ShffPEFV65ceSSVgY+D+9nElEolVqtVloFHIpGc3LzvxN1s3iaTCbPZTEVFBVVVVeh0OtLpNF6vN6cbaW9vbzM1NSUbv6vVarRaLSUlJbS3t7OyssLKysp9e81EhWddXR0ej0cKQ8/NzTE6OpqzXsqdPK3G204MBgOtra14PB4cDgfT09P09/fLwpB8uj+pVEoqDpjNZlpaWqisrCQUCrGxsSHFn8fGxmT4/H7Y2NhAoVAwPj6OyWSSkZnjx48zOjrKV199lZP3TalUyjyw+vp6XC7Xru9vbW3JFJlcffYKhYKSkhKqqqpkvi/c8JYuLS0RCoV2FSfe/LM6nQ6tVovJZKK9vZ329nYOHDhAZWUlOp2OTCYjdQK//vprwuGwnDMPO9T82A05lUqFyWTCZrPhcrmkrIZIll9eXs6L0nSVSkVjYyOlpaU0NTVhtVpxuVysra2xvLxMX18f4+Pj0pjbOZGVSiVarZb9+/fzgx/8gPb2dkpLS4Ebobm33nqLiYkJgsFgTr4AjwKRICoq+LLZbF4ZcneDzWaTcgwtLS3AjRDE5OQkCwsLOfus0+k0ly9fJhqNEgwGZYJ+eXk5R44cYWpqSuq83c8CZTKZaGtro7Ozk+7ublk0MDIywsWLF/PyMPM0YrFYePbZZ9m3bx9ut5uZmRn+8R//kdXV1ZwOEd6ORCLBtWvXpMZfZ2cnTU1NMs/39OnTTE1N8eMf/xi/33/fkiGRSITNzU0uXLhAIpGgo6ODhoYGvv/97/P5559z+vTpnFwXVCoVFRUVeDweurq6bpHH2traIhqNyhByrlJTU0NbW5tM54AbxUxjY2N4vd47pgPsTC9xu938s3/2z3jzzTex2WxSSiiRSOD1erl8+TJ/8zd/I2W3HgWP3ZCzWCwcOXKEzs5OzGYzOp2ObDbL8PAwZ86cyRvZCa1Wy4svvkhjY6P0rhiNRtlMvLe3VzbaXVlZoa+vj62tLSwWC2VlZbS2tnLkyBEaGxsxmUyk02muX7/O5OSkbBqfiy/wo0KUZIuvXEyOfVDq6+s5duwYZWVlsqJrYWGBX/3qV8zMzOTsmEX4NxgM0tfXR1NTE4cPH6akpISuri7efPNN2tvbuXLlCsvLy7Kx9O8LpwjvvNvt5uWXX5Y9SBcXF2WSdD5455921Go1tbW1tLS0cOLECQB+9atfcfnyZdbW1vIqN+5mFhYWeOutt+jp6eH48eNS27OtrY3KykocDgerq6sMDw+zurrK7OysTC+JxWJ3ZcBms1kCgQBer5dUKoXJZMJiscj98UF6cD4qDAYDL7/8Mh0dHbKXONxwyMzMzPDRRx/Jrh+5jMlkoqioaFfFuOiBnE6nb9Gy1Wq17Nu3j7KyMg4ePIjFYsFut0vhe6EXe+HCBRYXF/nss89k68VHeSB97IacyWRi3759eDweeZMymQwzMzMMDAzkjRdGrVZz4MABurq6cLlcuzR0ROLn9vY2fX19zM7Oyl6ETqeT5uZmTpw4QUNDA5WVlbJLwtjYGMPDw/j9/px/AR42ImdEGHFP4uZdWVlJb28vdrtdtgKampri9OnTOW+4J5NJ1tfXGRoaQqVSyUbg4gTa3t6OVqtlcnJSqvb/vkRhpVKJXq+ntLSUZ555hvLycnQ6HX6/n8HBQfx+P9Fo9ImcC08SwpBrbm7mwIEDDA8Pc+rUKcbGxtjY2Njry3sgVlZWOHnyJPF4HIfDIYsfqqurUSgUtLe3EwqF6OvrY3p6GoPBwPz8PPC71lXiHRA5WDejUCikOOz29vaujh9arfa+Pd2PCoVCgV6v55lnnqGpqWmXEbS8vMzVq1dlN4tcdsyIcYjWoAKhxADILgwCo9FIT08PjY2N/PEf/7EMvQtEIczg4CDXrl3jZz/72WO5B3tmyNXU1ADI/JqpqSkWFxdz2g27k1gsxv/5P/8Ht9vNwYMHKSkpoba2ViaBw42NqrW1lerqaqqrq2WiuMlkwuFwoFKpiEQiDA4OMjc3x9///d8zMzPD2tpa3oUiHpRUKsXs7Cyzs7Oy9+qThsFgoKSkBJ1Ox/b2NsPDw1y7dk1W++Y6kUiEDz74gLm5OaxWq2zL5nA4sFgs/OAHPyAUCtHW1sbc3BxffvkloVBIVvQKlEollZWVOJ1OXnzxRZqbm2lqakKpVMpN8eTJk3I9yGUD914RlYjLy8toNBocDsdeX9IDoVarsVqtvPbaa1RVVTE6Osr58+f54IMP8kIT8fchRJr7+/uZnZ2loaGB8vJyXnjhBaqqqmhubsZoNNLb20tHRwff+ta3CIfDhMNh5ufnZReQzc1N0uk0arX6lh7iSqVSFlZYrVaZfmQymTAYDKTT6ZzZFxUKhazIrKys3CXbAdDY2EhxcTGnTp1iY2MjpwzQm8lkMvzqV7/i0qVLUgxYqVRSU1PDv/7X/1rKBu1EpVJRVlaGyWSiuLgYlUpFJpORToi+vj4mJyf52c9+xuzs7GNLC3mshpxOp8NsNlNWVkZRUREA6+vrzM3NySKHXH7wO0mn01y5coWZmRkymQxut5t0Oo1KpaK0tBSdTodaraaoqIiioiKZDKrRaMhkMqTTaTY3NwmFQkxOTjIyMsK1a9dytoPDoyadTrO2tkYoFLrrkES+oFAopPdJuN8zmQwrKyssLy/fdx+/x42QB9LpdFy/fh29Xk9NTY0sXrJYLMTjcdbX1ykqKmJubg69Xk8ikZBeBSG5U1lZSUVFBQcPHqS6upqioiL5PiwtLTExMcHm5mZeGLj3gkghEMU8Qu09XzEYDFitVhoaGnA4HCwtLTE/P8/s7OxeX9pDQXiWhcNhdXUVh8MheyrbbDYsFgtWqxWz2UxlZSWJRIKtrS1mZ2cJBAJS3F108Lm5p6dSqaS9vR232y1TjVKpFNvb27L/cK6gUChwOp1UVFRIQfOd2Gw2DAYDOp0uL9bw2dlZwuEwq6urlJeXSw9be3v7Xf18KpVic3NTFryMjo4yPDwshX4fF4/NkNPr9bzwwgvs27ePlpYWeSrp7+/npz/96SOp5HiUCEmGWCzGRx99hM1m4+LFi3g8Htra2njppZfo6uqSn9+ZRyBy50Qz8E8//VQWRjytJJNJrl+/LnsOPkkIT1xlZSXl5eWo1Wqi0SjXr19naGgoLxY8+J0RMj09zQ9/+EMOHjzI+Pg4vb29NDY2yoX92Wefpbu7m6NHj7K4uMjVq1fxer34fD5ZINTT04PD4ZDV2gqFgvn5eb766iuGhobw+/05KVnwoKTTaeLxOLOzs2i1WiorK/f6ku4blUrFiy++SHt7O83NzYRCId555x0mJyf3+tIeGaurq4RCIf7v//2/0jEhcjwrKipob2+XQvetra0y3LxzLqvVu7ddEeLTaDRotVqWl5f55S9/yZUrV1hdXc2pfVGhUMgwuujuku9sbW3xy1/+krGxMd544w1MJpPUe7wdokI9k8kwNjbGxx9/zKVLl7h8+TKbm5tSE/Rx8lgMOY1GI7VZ6uvrZYWL0M5aWFjIy1CayIFYX18nlUpJF3Mmk6GxsRGXy4Xdbr9lsm9tbUnX+8jICAsLC4+smiWXETIsogpStPYRfWbzJV/y96FWqzEYDLI5eiKRkJpCudy143aIZtJ+v5/Z2VkGBwel1pI4pQshVTHmdDqN3W6npKSEhoYGSktLaWhowGKxoNFoiMfjTE5OMj4+zvj4eF5WOd4L29vbsuMJ/K5d4c2iqrmMeG9ra2tpaGhgY2OD5eVl5ufnn+i1TBRjiQ4sarWacDhMeXm57MdZWVnJxsaGzHET9wpuJMvf3M5qp07j+vo6S0tLjI6Osri4mJPvgWi7KJ5zcXGxHE8sFpNKDfmCkFfKZrPU1tbKzkTALe+kSI1IJpNEIhHGxsa4du0aw8PDjI+P78XlA4/BkFMoFJSWllJZWcn3vvc96urq0Ov1+P1+rl27xujoKD6fL+8lBmKxGPPz83i9XgYGBqTW0Pe//31qa2t3fXZra4tAIMDly5f54IMPnkjPw90gQo0ul0tq+TgcDtbX1xkZGaG/v3+vL/GhIMIpIvwyPDzMxMSENOLzyZADZOhHvL8ffPABZrOZl19+GY/Hw3e+8x2Ki4sxGo0yb1ScYEVRkFqtll1ABgcH+Zu/+RsWFxeZmZnJC924ByGZTErdsGw2i06nw2q15pV3w2w2U1xczIsvvkhrayt/93d/x+joKAMDAzmTz/U4SKfT+P1+fvGLX8g+m6WlpZSVlXHgwAEqKiqw2WzSC+d2u3n++efRaDSoVCqZ/3bu3DlmZmY4deqU3ENy0RjKZDJcuXIFn89HMpnE4/Hw3e9+V0acRkdH6evru0UQPpdJJpN88sknKJVKfvzjH1NXV8err7562/7HmUyGpaUl/H4/Z86ckTp5e72HP1JDTizYwvXucrkwmUxsb28TCAS4ePEi8/PzMocm3xHW+vb2NgsLCxiNRiYmJmQlkji1LCwsEAgEiEQiT9WidzMqlUoKKZtMJkpKSgiFQnLTf1Kw2Wy0trZSUlKCQqFgYWGB69ev5333EpH3tr6+TiKRYGRkhHA4jMlkwul0Ul9fj8lkkpVfCoXiljZH4+PjTExMyCKfWCz2RKwF38T29rbMkctms5jNZqqrq/MiP1bkbIkCFYVCgd/vl8/wTgKqTzLZbHaX0SVaMokKbNGmCZAt7UQHm+3tbXkoWl5eZmJiQgrB5+J9FJ1Mstksg4ODrKysoFarpaE6OjrK2NhY3qUJiX04kUjg8/kYGhpCrVbf1pALBAKEw+Gc6h38SA05lUqFTqfjzTff5Nlnn6WyshK1Wk08HmdsbIwf/ehHT1xPVZEcOzw8zNzcHE6nk4aGBnQ6nTTkVldXGR8fz9n2K48L0SHAarXicDhwOByEw2FZ8PAkoFAoqK6u5vXXX6elpYVsNkt/fz8nT558Iqr6ACmS+uWXX6JQKHjvvfcoLy/njTfeoKqqio6ODuDGvRgbGyMQCEhx4V//+tdsbGw88V64nYjkefH+u1wuent7ZSeLXEb01nzttdd44403GBsbY2RkhK+++uqW6uSnlZ1Vq7fjdsULOyVKch2fz4fP55OhxL/927/d9f1873Li9/v54IMP7vj9XBzbYwmtCs+LQqEgGAxy8uRJ+vv7WVtby/uQ6p1IpVJEo1H6+/uZnJzcleAai8UIBoOPtaolFxHN4r1eLzabjWg0yvDwMNPT009Mno1o5WK324lGo4yNjbG0tPRE5oGJBTyZTBIMBjl37hwjIyNcvXpVfmZ1dZVoNCp7K0ej0SfuPvw+0uk0Pp+PpaUllpeXCYVCxOPxPQ/P3A0ej4fu7m7KysoIh8OcOXOG8fHxJ+ow/rC404afi4bA/ZBPxue9km9jeiyGnAgrZrNZVlZW+Pu//3sWFhbyriH8vZBKpUilUpw9e3avLyVnEZv5wsICarWalZUVhoaGmJiYeGI2d2HIORwOAoEAU1NTsrglV9zyD5t0Ok0oFOLrr7/e60vJSVKpFF6vl8XFRbxerzRu82HONzc38/3vfx+1Wk0wGOSLL75gcHDwiZ3LBQrkA4pvsjwVCsUDmaUiR663t1fmx0UiEfr7+4lGo49V9Tubzd6TGM+Djj2XuJexP85xi+TgAwcOSA2ycDjM8PDwQ9kYcmHcCoWCyspKDhw4QCKRkLlkgUDgkZ36CnP97tircSsUCjQaDSUlJezbt0/O+8XFxfv2RD/qZy76Q3/ve9/jr//6r/nyyy8ZGhris88+w+fz7akHIx+e+aOg8J7fHU/DuB+pR05Ut505c+ZR/pkCeUomkyGZTHL+/Pm9vpRHRjabZWFhIa+quAo8WkT42ev14vV69/py7gqRG6fRaNDpdCwuLnLp0iWZ+F6gQIG945F65HKJwunl7iiMO/8pzPW742kdN9zf2FUqFU6nk/Lycvx+P6FQiEQisedh1cIzvzue1rE/DeMuGHJ34Gkde2Hc+U9hrt8dT+u44ekd+9M6bnh6x/40jPsbDbkCBQoUKFCgQIECuUv+9IQpUKBAgQIFChQosIuCIVegQIECBQoUKJCnFAy5AgUKFChQoECBPKVgyBUoUKBAgQIFCuQpBUOuQIECBQoUKFAgTykYcgUKFChQoECBAnlKwZArUKBAgQIFChTIUwqGXIECBQoUKFCgQJ5SMOQKFChQoECBAgXyFPU3ffNpaG1xJ57WsRfGnf8U5vrd8bSOG57esT+t44and+xPw7gLHrkCBQoUKFCgQIE8pWDIFShQoECBAgUK5CnfGFotUOBhoFAolByNUQAAIABJREFUUKvVVFRUsH//fubm5pidnWVzc5NkMrnXl1egQIECBQrkLQWPXIFHjlKpRK/Xc+jQIf7X//pfvPHGG1RVVWE0Gvf60goUKFCgQIG8pmDIFXjkWK1Wjh8/zsGDB3E6ndTW1tLV1UVRUREKxT3l6xYoUKBAgQIFdlAw5Ao8csxmM729vbS3t2O326mpqaG1tbVgyBV4alAoFLu+ChR40lAqlfKrMMcfL4UcuQKPnO3tbSKRCLFYDIDW1lacTifj4+OEw2F8Ph9bW1t7fJUFCjxcFAoFDocDu91OcXExBoMBgHQ6TTgcZmNjg/n5ebLZJ0YdocBTSktLC6+99hrFxcVYrVb+8R//kYGBARKJBJlMZq8v74nnoRly4qSpUqnIZrNks1kymUxhkfr/USgUKJVKdDodSqUStVq963SezWZJp9PynonPZDIZtre35e9Qq9UolbsdqZlMhnQ6TSqVIpVK5dyLk8lkiEaj0lgTm5vb7cZutxMIBJ5KQ06cXMUzFf+qVKpbPivep62tLba3t0mlUntwxfeOSqVCo9HIL7gxlmQyyfb2tpzzuTZn7xaFQoFGo7nl+Yl3u6ysjPLyckpLSzGbzcANQ87v9xMMBtnY2CCZTMr7Id71Ak8GN7/f4v/t/G9APv9EIpGXe2ZJSQmHDh2ivLycsrIyTp8+zfDw8FO5rt+MmAMqlUr+K9b4bDZLIpFge3v7gQr/Hoohp1KpsNlsWCwWKisr2draIh6P4/V6CYVCD+NP5DUKhQKTyURZWRmvv/46lZWVtLa2YjabZcJ/Op3mwoULhEIhNjc3KS4upquri5WVFaampjAYDOj1ejo7O7Hb7bsMwNXVVUZHRzl79iynT59mbW1Ner9ygVgsxvXr1ykvLyebzcpNrra2lo6ODubn54lGo3t9mY8VhUKB0+nEZrPR1dVFaWkpnZ2dlJWVUVtbKz8jDkWJRAK/38+7777L5OQkZ8+e3dsB3CV1dXW88MIL9Pb20tPTg0KhIJVKcfLkSWZnZ7l06RJra2ssLS3l3QYm3kmxgR04cIDi4mLq6+vlQq3X69HpdGi12l2LdzKZJJFIEAwGGRwc5NSpUwwODjI9Pb2XQyrwEDGZTBgMBjo6OigpKaGjowOdTodKpaKqqorW1lbgxnz4/PPPmZyc5Gc/+xlra2t7fOX3RzabxWq1Ul5ejs1mQ6/XE41Gn+rDiUajwWaz4fF48Hg8dHV10dLSQnV1NXBjb/zpT3/K5OQkv/3tb+/7gP5QDDmNRkNVVRUlJSV4PB62traIxWIUFRURDAZJJpOk02nW19dJpVLE4/G8W7QfBJVKRUVFBXV1dXR0dFBVVUVbW9sthlwikSAUChGNRqUh5/f7MRqNtxhysViMVCol/9Xr9Wi12pzMTchkMsRiMaLRKOFweJfnwmQy3dYD9SQhTt8GgwGDwYBarUatVlNTU4Pdbqejo4OysjI6Oztxu93U1tbueo6ZTIZEIoHP5+PKlStEIhFp5OUqarUaq9VKRUUF7e3tdHR00NXVhUKhIJ1Os7CwgMViYWNjg8XFRfx+P9vb23nlmXM6nbhcLtra2qisrKSzsxOHw0FdXZ2c0yqVSuYN7UR40aPRKJlMBr/fj9frZW5uLqc3Pr1ej9Fo3JUHpVQqsVgs0tNw8xoUjUZltGB7e1tu7vniVb5bhPfZZDJhMpmw2+2YzWba29txuVx0dHSg1WoBqK2tpb29HYVCQSaTYXx8nLW1tbxdC5PJJJFIhHQ6jU6nw2q1UlRURDgcfuKe8zchnBRiPy4tLcXhcNDe3k59fT0dHR3SkFMqlUSjURoaGojH47esEffCQzHknE4n//E//kfq6uro7Owkm81KN3EymWRmZga/38+vf/1rFhYWuHr1KqlUinQ6/TD+fM5jsVj49//+39Pa2sqhQ4fk6Xzng1OpVPT09MiNTLhj7XY7TU1NcnHUaDQkk0n6+/tZWlqiv7+f5eVlxsbGWF1dZWVlJec2eIVCgU6nw+/3c/LkSYqKirDZbITD4b2+tMeC0WjE5XJx8OBB9u/fT2lpKUVFRXR1dWGz2WS4XYTogFueoV6vx26388wzz5BIJPZiGHeNSqXC6XTyyiuvcOjQId58800MBoMck0ql4g/+4A9IJpO8+OKL9Pf3Mz8/TyQSYXNzc4+v/u75zne+w+uvv47H48FisewKsf4+lEolWq0WjUZDT08P9fX1hMNhJiYm2NzczNnNz+Px8Oyzz2IwGKRRYjQaefXVV7Hb7buiBXBjHvf19TE/P4/X6yUYDHL+/HnW1tZYWFjIubXqQRARqWPHjvHcc8/R0tKC0+nEYrGgVqvRaDSkUik2NzflAV543EOhECsrK3m7J/r9fj7//HPsdjvd3d20t7ezubnJ2tpazq9XDxOtVotOp6OtrY2amhr+8i//krKyMqqrq3cdfGKxGHq9HrVazYEDB1CpVLz99tv3HYp+KIac0AkzmUxYrVZ5ytiZA2O1Wnn22Wepq6ujtLSUra0ttra28Pv9+P1+mSvypKFSqdBqtTidThwOBwaDQeYRwu7cuFgsRjqdlvdtZ3h0c3OTeDwuw9aXL18mEAgwPDxMKBQiEAiwubmZkx4NrVaL2+2mqKiITCYjvbK57Hl4GOj1eioqKigtLaW5uZnW1laam5vlSd3lcknvxk7utLmJzf9uDIW9wGg0otPpKC0tpbKykgMHDuDxeDCbzXLOi8XMYDCg0WhwuVy43W7Ky8tRKpV5ZciZTCaKioqwWCyYTCa2t7fJZrNsbW2RSqVIJBLE43Hi8TjhcJh4PE4qlUKlUmG327HZbNTU1KDX6ykuLsZoNOaER0aEBNva2jAajbsMM4/HIz1LYh7q9XqqqqqwWq1YLJZbPHL19fWYTCacTifr6+vo9XpCoRCLi4ssLCwwPj7+WMf3MFEqlRiNRmpra3G73TQ3N9PV1UVjYyNWqxWlUkkwGGRra4twOCzXZ7fbjdVqlet/IBBgeXk5bw25VCrF+vq6NERKSkqorKyUxv6Tjs1mo6ioSEYmGxsbcbvd1NTUYDAYiEQihEIhfD4fZrMZg8GAx+NBr9fLPf9BeCg7QiaTkUaGQIQTNBoNDQ0NZLNZDhw4IMMJiUSCjY0N3n//fX7zm98wPDxMIBB4GJeTU+h0OsxmMw6Hg6Kiottu2vF4nFgsxuzsLNFolI2NDTY2Npibm5Ob+tjYGPPz8ywvL7O5uUkoFCKdTu8yCHMVi8XCs88+KyualEqlTNp/UlEoFJSUlPCd73yHzs5OXnrpJRki3/mZJwWFQoHb7cbpdPLyyy/j8Xh4/fXX0el0dxynSqXC4XBQXV3N0aNHuXLlCouLi4/5yu8fkbwswtxbW1vyALa2tobX62V+fp6FhQUuXLjA/Py8NGSOHTtGd3c3//bf/ltZCCJO6A8SYnkYiA3of//v/01DQ8Ou6IEY883c6Zqz2SwtLS20tLTINUqkkayurvKTn/yE//E//scjG8ujRq1WU1lZyb/7d/+OpqYment7pad1bGyMiYkJrl69yvLyMmfPnkWj0VBbW8uJEydoaGggnU4Tj8cZHh7m8uXLxOPxvR7SfZFMJgmFQtL7VldXJ3PDnwY8Hg89PT1yvbfZbPI98Xq9fP7551y6dImPP/6Yjo4OPB4Pf/mXf0ltbS3BYJBgMPhAe/hDMeRisRh9fX34/X5isRh2u52SkhKKi4vlgxSxY41Gg9FolC7Inp4e9Ho9n3/+ORMTE0xNTeXtZN6J0WjEaDTywgsv4PF4qKysxGq17vpMIBAgEAjIRX5tbU16Kre2tnYlva6srLC2tsbGxoY88eey8bYTcWrd2NhgbGyMqqoqampq9vqyHgnCO3348GHq6uo4cuQIlZWVcs7vNGoikQiJRAKv14tKpaK4uBiNRoNOp8NgMKDT6Xb9bnFgyrWwm9FoxGQyceLECTweD52dnbhcLjQajRxvPB4nkUjIdUB46QDsdjtHjx5FpVKxtLTE6upqXoTd+/r62N7e5tixYxQXFzMxMUEkEsHv9xONRllfXycUChEKhZifn9/llYvH47dEIHLBsFcoFBw5coSDBw/KZ7ixsUEsFmN1dRW/34/P55Medr1ez/b2tly3b5f/LObz0aNHsdvtmEwmNBrNrgKQfESn03H06FEaGxtlmoSQU1peXmZwcBCv14vX6yUSibC4uEhFRQUdHR243W62t7eZmJhgZmYGr9fLxsZGTkZU7hVRyLW9vS0r1Z9U3G43nZ2dHDhwgAMHDlBbW4vJZCISiRCNRrl27Rrz8/OcOXNG7vEajYbS0lLUajVbW1tMT08zOzv7QI6Nh2LIbW5u8tvf/haXy8X8/DxNTU10dXXh8XgwGAy7TmsiX0qn02EymfjWt77F0aNHMRgMnD9/npWVlSfCkLNarbhcLv78z/+c/fv343K5bjnJ+nw+rl69yg9/+EP6+vr26EofPUqlEpPJhM/n46OPPuLb3/72E2vIqVQqLBYLf/qnf0pLSwtHjx69rQcjk8kQCoVYXV3l3Llz6HQ6GhsbMZvNcu7cyZDLtZwTq9UqvY89PT1YLJZbPDTRaJTV1VXpqRfVe3Ajx/bVV19Fo9EwOzvL1atX88KQ++yzz/j666/Z3NykoqKCjz76CJ/Px+jo6O/dkEX+cK6hUCh45ZVXeO211zAYDKRSKQKBACsrKwwNDXHhwgW+/PJLDh8+THNzM1arla2tLd577z1WV1dv61koKiqSWnqdnZ2YTCZZ7LTX3scHwWAw8Ed/9Ee0tbXR29vL+vo6U1NTfP3113z99ddcvHiRpaUl+XmlUklZWRmHDx+mtLSUdDrN0NAQp06dYnp6mkgksoejeXgoFApKS0sxGo3o9fq9vpxHSk1NDd/73vfo7u6mq6sLpVJJJpNhdnaWubk5fvKTnzAzM8OlS5fke2E0GqmpqUGr1ZJIJBgeHmZ0dHTvDblUKoXf75dx4CtXrvDZZ59RWlpKcXExHo8Hm80mq3mqqqqwWCw4HA6pqXP48GEqKioYHh5me3ubUCiU16eTffv2SQv9/2PvvGPjPu/7/769977jkcc9TiRFUsuyJFO2Izuy08ZJGruOYzQdRuokBZKmKNr+U7RFA/SP/AIEaGMkcVKk2XYs78hDe0sUh7h15JG3eTze3vP3h/A8Jq1h2RbFO+r7AgQ5uSP1fe77vef5zPfnRilV4IPDbbNr7YjFYnR2dtK6uGg0Cp/PB71eD6VSibfffnujL/GOoVKpqJe2untxNdPT0/B6vXj99dfh8/ng9/vBZrOhVCppbclnP/tZ7Nq1C8A1A87r9cLpdOKll17C/Px8VUVjOzs7sXXrVlgslhvW/AHX9ohUKoWTJ0/C5/Ph6aefhtVqhUqlog0C9fX1ePDBB7GysgKHw7EBK/l4kIau9957D1KpFG63G+l0+pb3RiqVwmAw4KGHHkJ7ezvYbDYteK+GUgOyHwPA5cuX4fV68dprr9Fi/GAwiEgkgqGhIczNzYHH46FUKmFxcRHZbPY6HTRSYkBqoxUKBdhsNrLZLBYWFhAKhTZqqZ8KqVQKrVaLrq4uNDY2IplMYnh4GD//+c/hdrvh8XjWOCMkaLF161Y0NDTA6/Xi+PHjOHLkCC5durSpZLpqodzn06JSqbB7924MDAxg165dUCqVyGazmJqagsfjwfvvvw+Px4Px8XEkEglwOBw0NTWhv78f+/fvh91ux9TUFPx+P8bGxuD1ejfekCuXy4jFYojFYggEAvT/l8lkkEql2L17N/R6PU23VioV6HQ6GmLn8Xhoa2uD2WyGyWRCIBCoCY/8RpCNsKmpCTt37qQF7cAHDzZpZiBdPdXomd9JVjc7kBrJcDiM1tZWWiC/GSDGGNGCM5vN172nXC5jcXER4+PjeP311+F2u9cIPtfV1aGtrQ19fX30Z0qlEgKBAObn53H27Nmq2/Tr6+vR19cHtVoNgUBADRMiYkxSialUCkNDQxgfH8eePXsgk8mgUCiobIVWq0VPTw+OHj260Uu6LUih+vj4+G29n81mQy6Xw2AwoKenB1arlTaG5XI5Koy80QcgMS4dDgcmJyfxxhtvIB6Pr3E4b0fvjkTcdDodrFYrlEolbZ7I5XIIBAJIJBLruZR1g9Q+WywW6HQ6ZDIZLCws4PDhw0in02ui5iwWC2KxGP39/eju7oZWq8Xk5CROnDiBS5cuYXp6egNXcmch33vy35sRFosFmUyGHTt2oLe3F62trbSxaXp6GuPj4zh8+DD8fj/S6TTNQtbX12Pfvn1UguzEiRMYHR2lKddPw7q2v6XTaeRyOZw8eZJ2OfH5fEilUphMJmzZsgUDAwPo7++HWq2mgoKkIL4aPNSPS2NjI7q6ujA4OIj77rsPCoWCvka+7GNjY3j11Vfh8Xjg8/k2ZZMHcO1AIEWfQ0NDGB0dhcfjgdlsRjweB4fDua4rrlZRqVRQKBT4+te/jp6eHqhUquveEw6HEQ6H8eqrr+L48eNUOw24FrU0m83Yv38/vvKVr6C5uRnAtcLweDyOQ4cOYXJyknbAVQMCgYBKq5hMJgiFQppayOVy8Pl8GBoawm9+8xvo9XrodDoajbx06RLy+TwsFgvtbNNqtZBIJNBoNBu8sjsPSZl/61vfQldXF7q6umjH2sjICN59912cPn26anS38vk83nrrLYyPjyMej3+sa+Lz+ZBIJBgcHMS+fftgt9thMplgMpnA4XCQzWbhdrvx6quvYnJych1XsX7E43Eq90TurVqths1mg8/nowEN4tQ3NTXh4MGD0Gq1WFpagsPhwKVLlzbt3r9Z4XA4VDPyiSeegEqlQjabxbvvvov33nsPo6Oj8Hq9CAaDyOfzYLPZMJlMePDBB7Fr1y48/vjjKJVK8Pv9OHPmDM6fP39HOvXX1ZAjI2eCweB1rxkMBmSzWajVarS0tKwRhq3lugm5XE6jMVqtds1ruVwOTqcTV65cwcmTJ2lB5GaE6OCRdIrb7UYgEKCjulaPFNsMyGQyGI1GdHV1obu7+4ZRRiKa6XK5MDc3R6UGeDweZDIZbDYb2tvb0dvbS2tLIpEIlpaWMDMzg/n5+arq9iXip3K5fE2XFhH9XVlZwdzcHE6cOIGmpia0trYiHA4jlUrB4/FAoVAgmUyCw+HQ6Qfkz2aBNHeo1Wro9Xr09vaiq6uLOnipVAperxdDQ0Pw+XwbHp0nU0RSqRTcbjfcbjfy+fxtRVdIo49MJoNOp0NXVxfuu+8+1NfXQ6lUolKpIJ1OIxaLwe/3w+Fw1KwhQ+RlwuEwNBoNVCoVpFIpdVIB0JGVdXV1aGpqgtVqBZfLhcvlgt/vx/Ly8qaoB7+XIFkXnU5HRX2j0Sjm5uZw8eJFzM/P0+gaEcomYtBEkmRxcRGBQAAejwd+v/+OSM5smCAVqaWz2+3I5XJIJpOoVCpwOp2fuoNjI1GpVGhvb18TiSM4nU5897vfRSgUQiwWq+kawI9CIBBAr9fjX//1X1EqlfB///d/cLlcAK617AsEAkxMTKBQKGyKIt++vj7cd999aGlpgUqlumFtHEmpEyMWAG1y6Ovrw7e//W3odDpIpVIakf7hD3+IM2fOYHR0FMlksqq+F3K5HPX19Whra0NnZyeEQiFNp8bjcZw9exbj4+OIxWIYHx/HzMwMrZN86623MDo6SqfB7N69u6YduJshEokglUrxta99Dffffz/6+vogl8vB4XAQDocxMjKCU6dO4b333quaSOvY2BhN56dSqdtOkalUKjz00EPYsmULHnnkERgMBuj1eirJMjo6Cp/PhzNnzmBhYQEjIyMbbrh+GjKZDN544w309PTg61//Orq6uvDcc8/hl7/8JRwOB4RCIRQKBf72b/+WdrVOTEzge9/7Hubn5xEKhTb1GbAZ4XK5aGpqQmNjI3g8Hubm5vDGG2/g2LFjGB8fX/M8S6VSHDx4EFu3bsWzzz4LFouFYDCIw4cP4/jx45iZmaGTXT71dX3q3/AJId4K0U8i0alEIlGzg4OBa2kFuVwOPp+/plYgFoshFArB6/Vu2ijcarhcLoRCIerr65FKpRAMBq+reyRjemr1Xq9GoVDAaDRSwecbQTSjiAdGaieam5vR0tICq9UKoVAIFouFUCiEcDiMubk5OJ1OJBKJqoteEoFiIi+xemxYqVRCMpmkA6FJYwCB6KmFQiHo9fo1tTUikQgqlQqJRKJmBVKJEHhdXR2sViva29upMC6Hw0E6ncby8jLGxsawsLBQNULIlUoFCwsLqFQqt+U4kIkFFosFZrMZ3d3d6OjoQFNTE9XF83q9dK0+nw/T09NYWlqq+VGNpVIJXq8XarUa0WgUbDYb9fX1sNlssNls4PP5dNqDRqOBz+eD0+nE/Pw8lpeXq8opu5OQus/NaKQS3UC9Xo9cLodQKISpqSn4fL41+5tYLIZarUZXVxd17sncdIfDgfn5+TtmxAEbaMjJ5XK0t7ejtbUVjY2N+OMf/0iFE6tp4PvHRSQSQavVrmm7zmazOHv2LEZGRjbtl/fD8Pl8iEQiGAwGLC8vr+niInMmiUDwZtAaUqvVsFqtt2zciMfj8Hq9NJ3C4/Gg1Wrx1a9+Fc3NzXTyRS6XwzvvvINjx47h0qVLNT2652asFoWNRqNrDvSmpibs2rWrpmuIpFIp6urq8MUvfhFPPPEErFYr7djM5XJwOBw4d+4c/vM//7OqHLtKpYJDhw6BzWbfVkpVoVBAp9PhH//xH9HU1AS73U5rJ3O5HBKJBH7/+9/jzJkzuHjxIqLRKJ2pW8tGHHAtvTo1NYVCoYBz586hoaEBdrsdjz/+OEwmE23ksFqtKBaL+OlPf4qpqSnMzs5uuu8zgdQ7k9TzZjPmhEIhDhw4gPr6eni9XgwPD+OVV15ZE03ncDhobW1Fe3s7nnnmGWg0GhSLRQwNDeFnP/sZJiYm4HQ676gtsCGGHIfDgcFgwN69e9HU1AQ+n49gMIiFhYWq08j6uJTLZeTz+TU3icvlwmg00rq5aDRaNR74ekBqgrRaLRKJBC3gJp9JNptFPB6nsyk3g3F7q0OJdDcGAgFMTEwgkUiAz+dj7969aGlpgc1mg0KhQDgcRjAYhNvtxuXLlzE9PY1YLFZ1kbiPIp/PI5VKwefz3bIbi0hufPizI0XjExMT632pdwzSySaVSlFfXw+DwYDW1lZs3boVBoOBRiyJSPDw8DCmpqaqcq7q7V4PWbNer0dDQwPq6uogkUjotAvSwU+cOpVKBS6XS52VT9upt9GUy2Ukk0ksLS3h7NmzSKfTaGlpgUajwZYtW5DJZOh7EokErl69CrfbvWYiz2ZBJBLBYrFAJpMBAFwuF1wuV00HZW4F6UBvaWnB5z73OSQSCaRSKeTzebBYLGzbtg2NjY1QKBTI5/OYn5/H9PQ0nE4nIpHIHT/z7rohR6Y72O12PP/883RouNPpxMWLF6vKO/0k5PN5JJPJNR4Xj8dDX18f+Hw+mpub4Xa7N70h19DQgKamJni9Xni93jXGbTqdplEmLpe7KbxTIrNxow169Riet99+G0tLS5BIJPi7v/s79PX1wWg0IplMYnZ2FufPn8ebb76J6elpWlNYS5TLZZpKHx4ehs/n+9i/g8xpPXny5Dpc4frA4XBgNpths9nw1FNPoampiX7nSRNIuVyG2+3GwsICfv/738Plct12I0E1wmKxYDab0dzcjI6ODhiNxjWvk/SyXq+HzWajw8JJSioSidTs2oEPSmZSqRR+/vOf47Of/Szuv/9+GI1GtLe3g8VioVAo4NSpU5icnMT58+cRDodres03Q6VSUaelUqngwoULNAK7mSAjNcvlMkwmEwwGA/bs2UPPOTI684EHHoBarYZIJML8/Dxee+01XLx4EWNjY+tyXXfdkFMoFFR/RaFQIBKJYG5uDl6vF/F4vOYPdafTiTfeeIPq4+l0OjpfU6/X49lnn4XD4cCFCxewvLxMN7TN0r1EohJEK+fkyZNwu91rjBzixQqFQojF4k2xsS0uLmJ4eBhtbW1Qq9VrXovFYnA4HEgkElAqlejs7IROp0NjYyPEYjGtnXnllVfgdDrhcDhqdgNksVh0CHx7ezsAfGyDtNbkaPR6PdRqNZ566ik0NDSgt7cXarV6zQiqQCCAlZUV/O53v8P8/DyuXr2KWCxW888+6XAlYwNXl0mQWdu9vb0wmUw0ukzmcu7YsQMrKyvw+/2Yn5//REZ/NUCcF4/Hg7Nnz9JZm+S1S5cuYWxsDOl0uubPt5vB5/Oh0Wjofu52uzExMbHpInJEANztdkMoFNKOfRJ1J+lkhUKBSqWC0dFRTE5O4tSpU3C73et2XRtiyO3duxfd3d2QSqWYm5uj3UyxWKzm02xOpxN+vx82mw1GoxFyuZwaclqtFl/5ylcwOTkJqVSK6elpzMzMIJfL1XSDx2pIqqWrqwt6vR6HDh2ihhyBRGx0Ot2appBaxuVyQSgU4uDBg9e9Fo/HaRpNpVLh4Ycfht1uR0NDA9hsNmZmZjA6Oorf/OY3SCaTNb35sVgsiEQiKJVKtLS03FZHci3ffzabDb1ej6amJjz55JN0ru5qKpUKNVZefvllXL16teb3OWCtVEkmk0EulwOHw6GGOJEg6u7uXvMzxWIRiUQCi4uLcDgcGBkZoTOHaxEyOs/n8+HcuXNQq9XYsmULgGvR+MuXL2NoaAjpdHrT1YwRhEIhNBoNrQ13u92YmpraNAEKQjabxZEjR+B2u9HV1YX6+no6TlGn09H3kWjt6OgoLl++jNOnT6+rEX/XDDmhUIjBwUF0dHTg4MGD0Ov1AIDJyUm8/PLLWFhY2BQdjKTV/s0338T4+Di++tWvor29HS0tLRAIBOByubDZbPiTP/kT7N+/H/F4nOrPXLhwASsrKwgGgzX7hY9EIkin0/jxj38MoVCI6enp6zYsWvDoAAAgAElEQVQwIlFR6/d6NUQPaHp6GiKRCPX19TQ6YTAYMDg4iP7+fiQSCVpLsry8DL/fjx/96EdwOp0fW3i1WiD1UORPuVy+rahaLpfD2NgYBAIBkskkRCIRPQhqISq3bds2tLa24oEHHqB1cR9udnE4HHA4HHj77bcxMTEBv99fs9/tD1OpVODxeJBIJPD9738fCoVijRFrs9mohhy5r1wuF1qtFmw2GzabDRqNhs4YNplMOH/+/A11R6sZgUAAo9GI/v5+PPHEE2hsbATwgSHb29sLFosFn8+36SJypPOe6GiSGrnNCpHlyeVyeOGFF9DX14cvfOELMBgM0Gq1dP9zOBxwOp34xS9+Qaf3rOd5t66GHNnY+Xw+FAoFenp60NXVhba2Njow1uv14sqVKx85W/VGGzv5YG712t2GtF3PzMxgcXER/f39tE5EJpNBLBZDoVCs0ZkTiUQwmUwIhULgcrlIJBJUNLfWyGQyyGQyuHTp0k3fQ6RnauGwvl0SiQStgdLpdDAYDHREkUQioZs7gcynXFhYwOnTpxEKhWqy0efDczXJ37dzb4vFIpVZyefz1AgqFovI5XJVaeiTubB8Ph+tra3YsWMH7r//fmqcEy084qwEAgGMj4/j4sWLGB8f/9iSG8QYIJ9psVisqmhePB5HOp3GsWPH6JB0cu+3bNmCeDwOg8EAqVQK4JrRw2azIRaLodPpIJPJYLFYaDf37OwswuFwTe19xDi1Wq3YsmULndhBvv9WqxWxWIzWSm4miAi0VCpdY7BvVkiNXKVSwfj4OGQyGSKRCORyOX1PuVxGMBiEy+XCyMjIXRmpuK5Plkwmg1wuxzPPPIOOjg7s3LkTSqUSIpEIMzMzePvtt3H8+PGPHDvE5/MhFovpF2O16CiLxYJCoaCbBxFdzWQyGxrdIDNFX3zxRahUKjzyyCNoamrCo48+Sse5EHp6etDW1obdu3cjGAzi17/+NRwOB44ePbppvPfVWK1W9PT00HqKzWDQZbNZFAoF/OhHP4LVasU//dM/0YaPGwnd5nI5vPjiixgZGcHS0lLViMF+HEqlEnK5HDKZDNLpNIRCITgczm1HW0UiEbZv347u7m7IZDI6p/Xy5ct4+eWX4ff778Iqbh+xWAyTyYQ9e/bgs5/9LFpaWmAymSCXy6mBQkgmkwgEAjhx4gRefvlluFwuWiR9O5AUtVwupyLCWq0WQ0NDOHfu3Hot8RNRLBbh8XjAZrPXGPHz8/N47733wOPxaK0gj8dDY2Mj6urqsH//frS0tGDr1q3YvXs3urq6kEqlcPnyZVy6dKlmGt/I3M2uri5IpVIsLi7C5XKhp6cHOp0Ou3btglqtxksvvUS7WDcLMpkMu3btwsDAAJqamiAQCDbFfn4z+Hw+tm7diqamJjz11FOwWCxobW2lTijRw4xGowiHw3ft/F5XQ06tVqOurg52ux1btmxBXV0dRCIR2Gw2stksTTMQ9e+beZpisRhyuRxcLpe2r5dKJYRCIbBYLOj1erqJkuHcfr8fsVhsPZd3S8gN9Xq9WFlZgcViQTqdRlNTE9RqNdLpNKRSKf1DxtrodDps2bIFXC4XDocD8Xj8Op2tWoeMsiG1gR+O5Kz+82GIkVBtBi6JxLrdbjpL8kaRuNVks9maro0kWnDJZBLxeBxcLndNROpWkEg9SUnweDwqKByJRKjhs9HweDxwuVzIZDKoVCq0tLRgy5Yt6O/vh06nW+OJr4ZIq3A4HAiFQqjVaggEAqRSqTURxxs9x6RRisgV2e12qtfm9XrXe8mfiBtFk29ksPB4PCSTScRiMZjNZvr5KZVKSKVSGI1G6HS6m4pqbzRkXyLSSWRkk81mg06nA4vFQjgchsPhQGNjI22E0el0EAqFm0IzczU8Ho+uUSKRAKjtmtdbweVyIRaL0dzcjM7OTnR3d9Ma+Ewmg2g0Ch6PR22TTWPI/emf/ikef/xxdHd3Uw0hcjiTdKLdbodGo7nlzVer1TCbzRAIBPSDKhQKuHr1KthsNjo6OugXf35+HjMzM/jJT36C06dPr+fybptsNovjx4/j7NmzeOutt6DRaNDY2IiDBw/iwIEDNEoJXGsGeeaZZxAOh/Hoo4/i8OHD+OlPf4pCoVBT6YZbYbFYMDAwgOnpaVrgzGKxqOo9GeFFPhNCpVKhWj3VatySkWPHjh1DOBxGX1/fDQ8lgUCA5557DjMzM/iP//gPBINBKiBaK6ysrCAajeL8+fOQyWTYt28frX39KMRiMVQqFex2O5qbm9dEsyKRCNxu94Yb6yQtZjKZ8Pjjj6O5uRl79uyBSCSiGYKbIZPJ0NLSgr/6q7/Cl770JczNzSEQCOC9996D1+vF5OQk0un0DY0d8m8+99xzsNls6OrqApfLBYfDgd/vx6FDh9Zz2etKoVCA2+3GysoKAoEASqUSDhw4QDXnVCoVdDpd1aYhySzghoYGyOVyep8///nPQyqVIp1OY2hoCL/97W/R1taGlpYWKBQKqFQqyGSyTZd6FAqFaG5uhtFopNmVatyXPy1Eaqe+vh7f+c530NDQAIVCgVwuh6WlJZw7dw4XL16E0WgEn8/HO++8Q536u8G6fFuIAKTRaER9fT0kEsl1nohUKqXRKY1Gc8vfR9IKxDsmkwGKxSJYLBatRwKAcDgMsVhcdRtBNptFLpdDPp+n8yavXLkCjUaDuro6KJVKmEwmCAQCOmuTDBrv6uqC2+2uuSLg1ayOrhFDzWAwgMfjUQ0qco95PB5EIhH18IAPut0cDgcikQji8XhV1QqR625sbIROp0NzczPMZvNN54eyWCyYTCZkMhnodLo184ZrBRKFJJHV2zW8OBwOmpub0djYCJPJBKVSed3vrYZ7y+Vy0dbWhsbGRnR1daGurg46ne62ZsKS2jbSAFAulyGXyxGJRGCxWCASiRCLxRAIBKhorlgshlgsRn19PXQ6HU3dKhQKpNNpeDyeDc0y3ClKpRLV21xdUkBqZ6t55i6pBevr64PBYIBYLKaRRWKkfvgesdlsej5ttrQjiU6S9ZEO5nQ6XdMaiQQWi0UnEHV3d8Nms9Eu1WAwiFAoRMeuRSIR1NfXQ6PR0MbGu3W/18XaUavVqK+vR0tLCxobG2/ouVosFphMptsujF79HrLxre4OIuTzecRisaqsOSLt+tlsFsvLy1haWsLRo0exc+dONDc34+mnn4bZbAaLxYJUKkVbWxvK5TKkUil+/etf47333tvoJXwiyAZNILU027Ztg1gsRlNTEwqFAux2O4RCIQQCASQSyZqGEDK78yc/+QkmJiawsLBQFYc9gaTevvGNb2Dr1q2w2+23nLtKoj1EnoHL5cLn89X8xnc78Hg8/Nmf/Rm2bt2KgYGBNQXy1YRAIMDTTz9N54eS9O/HgRj4LS0taG5uxrZt25DNZjE5OQm/34/Lly/TuYwdHR1oaGiAVCqlzjCbzaZdcL/4xS9w/vz5dVotw+1gtVrR0dGBb3/727Db7QBA5wi7XC4cPXoUKysrsNlsaxzRe4VgMAin0wmfz1d1zvbHhaTO+/r60NPTg89//vNoaGiASqVCNBrF+++/j/Hxcbz77rswGo20c7mnpwcOhwMA7tp0mnUx5PL5POLxONUWup0NkHhnn3RQNqk7GR4exoULFxAKhT7p5d8VyFBqFouFyclJhMNhmEwm2Gw2bN++HXw+n9ZftLa2Uk26aheVJKlRtVoNsViMuro6yGQyWK3WNQaczWaDTCYDh8OBXC6nNQbFYhGxWIxG5YAPdLjC4TAuXLgAn89XNRsEn8+HUCjE9u3b0dXVha6uLpjNZvD5fBSLRSwtLSEcDsPpdEIkEkEkEqG9vZ22qotEItjtdhSLRVy4cGFTSPB8FCSNLpFI1jhykUgEgUCgakY3sdls6HQ66PX6G3rXRNWfTCtYWlq67rtJhsrr9XpaD8tisWA0GmmmgpQRkOgkaZwgGm0zMzO4cuUKRkZGEAgE7uZHsG6QbkeSqSGZCq/Xe53uZDVBItAcDgc8Ho+OGxwbG8PMzAzOnz8PpVIJo9FI06gkYr2ZymNuRqFQQDqdpiMZa3kvk0qlUKvV6Ovrw3333QeLxQKpVIpQKITFxUUcO3YMHo8HkUgE7e3t6OnpgVKppLXxRHbkbrAuhlw6nUYoFEIsFkMmk0GxWPzIcLnP50M0Gv3E81bJv0kMudsRIt1okskkkskkwuEwhEIh4vE4Ojo6YLfbwefzAVwTEZbJZLDZbNBqtQgEAlW9GRDDhqQWH3zwQTQ0NGBwcBB8Pn/NYUg2OKKCvbi4iOXlZYyOjq5J05XLZZw9e5bq91TT5iASiaDVanHw4EE89thjMJlMVH4gkUhQwetDhw5Br9fDYDDg2WefhVarpT+/e/dusFgs/O53v0OpVKoaI3W9IAYsiTgRAoEAzpw5UzXCsGw2G2azGWazmTYvrSYYDMLn88Hv9yMajeLcuXPXiTlLJBKIxWLs2LED9fX1sFqtkMlkqKurA5vNplGdD1MulxGLxeD1evGHP/wBU1NTOHHiRFU9+58GDocDqVRKu/3S6TSi0ShmZ2cxOTlZlRkV4NqelU6nUSqVUC6XkU6n4fP58Pbbb2NychLHjh3DI488gm3btkEqlVK5ilQqhWw2i3w+v9FLWFfInOVsNlvz+5harUZnZyf279+PRx99FMC1+z8+Po6RkRH87ne/Q6FQoDqC+/fvh1arRS6Xw9TUFCYmJu6aQ7IuhlyhUEAymcRLL72EsbExGom5FclkEtls9lNF5DKZDC0ar1aP7kaQaGKhUKBiqqRolEQCNBoNzGYzotFoVSv/9/b2orOzE3v27IHJZEJdXR3kcjmNyhJhTK/Xi4sXL2JxcRFutxvpdBrpdBqZTAahUGjNgVWpVBAOh6sqWqVSqdDV1YXW1lZ0d3djx44d0Gg0YLPZiMfjOHToEDweD8bGxrC8vAyn0wkulwudTrdmDTweD1arFV6vF0ajEeFw+K7oDt1JyLi57u5umEwmADevBdq1axf1XhsaGsDj8RCJRHD58mVcvnwZ7777Lq5evXo3L/+mFAoFHD16FKFQCNu3b0c+n4fH48Hc3BwmJiZoRzrZu27kZJFOe4fDAZlMhvr6emi1WnzmM5+BVquFxWJZ81lVKhXMzMxgaWkJx44dg8/nw+joKFZWVqrm2f80cDgcOvXjySefRF9fHwDg6tWrGB8fx/z8PJaXl6vOWVWpVDAYDDhw4AD6+/vB5XIxPz+PV155BQsLCzh//jzS6TR0Oh3d/3Q6HQqFAl5++WVMTEwgGAxWRSf2nYTNZtMmROBaVH1+fr7mJVaIE7djxw7awBUIBLC8vIyXXnoJs7OzKBQK0Ol06OnpwZYtW2CxWDAzMwOPx4NgMHhXgw7rYsiRqMLx48dx/Pjx9fgnNhWkaBxYK8EBfFBPRuRJqr19va2tDYODg/jMZz5DD/XVlMtlLC0tYWJiAm+88QZGR0exvLxcU4Y36a7bunUrtm3bRgckS6VS5HI5xGIxHD58GLOzs7hy5QoqlQo12D7s1HA4HOj1euj1eqhUKmSz2Zoz5Kanp+F0OhGJRNZIyXA4HPqHQDTDWltbaZNTLBbDhQsXcPbsWRw9enRD1nAjCoUChoaGkMlk0NbWhnQ6jZmZGZw8eRJvv/02QqEQFYL+KMbHx8FisVBXVweLxQKLxYKGhgYolco1UUkiJj47O4s//OEPCAQCiMfjm8KIA6497xqNBk1NTTh48CB9BtxuNy5fvgyv11uVDR0KhQKNjY3YtWsX9u7di2g0Crfbjddffx0ulwsejwcKhYJqR5KxZKlUCseOHcPw8DAdqL6ZIPXqxFGPx+PweDw1KW5OII1KOp0OdrsdSqUS5XKZOuRHjhyB1+tFpVKBRqNBb28vmpqaoNVq8e677+LixYuIRqN39UyrrtbOewyiEr9161bYbDZ8+ctfhs1mg0gkoqmcWCyGYDCI6elpTExMVL2n09HRgcHBQahUquteW1hYwNmzZ3HhwgVcunQJ8/PzWFlZqanNTSaTYffu3eju7saf//mfQ6PRQKPRIJfLwefz4be//S2taapUKnjooYdQV1eH3t5e2Gw22Gw21NXV0d9XLpeRSCQQi8Wu6+KrNcgzWy6XIZPJsHfvXlitVhgMBtqx/dhjj2Hr1q2QyWTUOFlZWcGRI0fgcrk2eAVrKRQKOHLkCC5cuIB33nmHpjsjkQiWl5eRz+c/lkRKpVLB8vIyEokEfvCDH1DB3w+zvLyMdDoNr9dbdaUEnxQWi0UjkM8//zyam5thsVhQLpcRiURoCUK1lsTU1dVh3759aG5uhlKpxMmTJzE1NYWFhQUkEgmYzWZs27YNzzzzDLq6ulCpVHD16lW43W7Mzc3B7/fXfKrxRsjlcjz00EM0ahUOhzE3N1f159StkMvlaG5uxsDAAO6//35kMhlMTU3hhRdewMTEBEKhEJRKJR599FG0tbVh7969UKlUWFhYwJEjR3D06NG77owzhtw6QdroVxd8EsONRCuEQiGEQiFaW1vR0dGB3t5emM1mcLlceiCSAfOk1bnaayxIs0M0Gr0uvba4uIjR0VFcuXKFGqW1FokTi8W0jpHoe5HC9FwuB5fLhbm5ORpF7ezsRGtrK+6//34YDIY1g5XL5TLy+Ty9txs9jeROQJ51Pp9P582SsodsNouWlpbras6y2Sw8Hg+i0ehGXvp1lMtlKsA7MzNzR34n6VqvtrWuF0QfjkSeGxoasG3bNlpLGovFsLS0hEAggEAgUJX7G2k6a2xspBFUr9cLh8OBVCpFNcbIyDaJRIJ8Pg+v14urV68iHA4jlUptCoP8w/D5fNqIR1QZYrFYTe9jQqGQ1sYajUY4HA4aTJmenoZAIKAp1cbGRtTX1yOVSsHr9WJhYQELCwt3/ZoZQ26dsFqt0Gq18Pv9tC5CIpGgvb0dCoUCer0e3d3d6OrqomN+lEolNfRKpRJSqRScTieOHTsGh8OBWCxW9V7d//t//w8/+9nPbvgamQJADvWNFnz9OLDZbBiNRrS3t+O5556jGnjEWFUoFJBIJPiHf/gHqhNICrr5fD4kEskabcNKpYJAIACPx4Pvf//7cDqd8Pv9NRWdvBWrFd+bm5upY3KzSQgMmw8inqtUKqFUKvH3f//36OjoQEtLC4BrNUfHjh3Dr371K8zOzlZltyqZytHV1YXBwUGk02k4HA4cPnwYIyMjMBgMaG1txbe+9S2YTCZotVr4fD5MTEzgRz/6EZ2jXG3rulMQg7VYLH6kHmytoFKpaPQVADweD65cuQKj0QilUomDBw/SLAsZSXj06FG89tprmJ6e3pBrZgy5dYJMb9BoNNTLlEqlaGlpgVKphF6vR2dnJzo6OuicSeDaAZ9Op2nH49TUFBwOB8LhcNUbcQCoZ70Z4fF4EAqFtJ5tNaQWzGQyUYXzD2skkrFNRLJiamoKTqeTFrdXYzTidqlUKgiFQggEAtBoNDQyy+Vy16jZk3ow0twTj8cRDAapXAFDbUGET4VCIUqlEhKJBK0xMhqNMBgMVPS9ra0NdXV1SKfTiMfjmJmZwfj4OGZnZ6s220BGyYnFYigUCppOVygUMBqN1DC12Wzg8/kIBAKYm5vD7Ows5ufnq6YDe70g+nlEmSKRSFClilqFy+VCJBJR5Qg+nw+5XI7W1laUy2V0dHRAr9dDJpMhHo9jcXER8/PzG5pSZgy5dWJgYACPPvooenp66KFPUgzkcCeK2B8WNJ6bm8PQ0BD+67/+C9FolOqrMVQHt+rAvpXMTj6fRy6Xw5kzZzA7O4uf/exncLlcSCaTNW/ElEolvP7661hYWMBf/uVf3rBGklAul+HxeLC0tIT333+fGrKbNWqxWWGxWLBarVCr1Whvb0cymcSxY8fA5/OhUqnw5JNP4ktf+hI0Gg0kEglEIhGSySTefPNNjI6O4sUXX6Rao9WadiRGKY/Howc6l8vFd7/7XeRyObS1tUEgEKBSqWB8fByHDh3CpUuXMDw8XNP1rrfL8vIyXnzxRTQ1NWH79u0YGhrC1atXq9Iov12KxSLVhgSAHTt2YGBggO7RfD4f+Xwefr8fp0+fxgsvvACv1wufz7dhWSbGkFsn3G43xsbGYDabqdbYh1NrmUwGiUQCkUgE6XSaShmQKA1pV78XNoRqh9yvcDiMs2fP0qicSCSCVCqFSCSiA8GJIDb54qdSKSwvL9P6qOHhYbhcLtqRWEsp5ptRqVTg8/kgFosxPj4Os9mMxsbG6wzbRCKBVCqFCxcuwOPxYHR0FD6fD8VicVN8DvcSZM61zWZDe3s7CoUCNXRkMhl6e3thMBhoxoGMobt69Srm5uZqolSkXC7Tuq+lpSU6ccNkMiGfzyORSGB5eRler5dqh/l8PqRSqY2+9LtCNpuF0+mkI7mcTieV0apVSN3yysoK4vE4RCIRBAIBlf4KhUKIRqOYnp7G2NgY7bTeyGeZMeTWiaNHj+LChQvg8Xjo6+vDnj17rpv/GolE4PV6cfnyZSwuLuLMmTMIhUIIh8NUMZyhOqhUKohEIigUCvj+978Pq9WKgYEBOoqOqPL7/X6srKxgYmKCtuDPzc3h+PHj1JALhUKbrvi5UqlgdnYWy8vL4HK56OjowF//9V/T9ATB5/PB7XbjhRdewMzMTM2UDDBcD5vNxmOPPUblZEhkirC6tKBcLiMej8Pv9+PIkSNwu901cdgXCgVEIhHaqEVm7ioUCqRSKRw6dAgLCws4fvw4PB4PZmZmNtX3+qNIpVK4fPkyWCwW/vjHP26KtSeTSUxMTNByAIvFAoVCgfn5eXi9Xhw/fhwulwvHjx+n+qcbDWPIrRMkPHvs2DHMzs7i3Llz4PF4NEJBZCfi8Ti8Xi8ikQhcLhd9MJhUavVRKpWQyWSwuLiISCSClZUVKJVKaDQayOVyCAQChMNhpNPpNeOaSO0YGdFTzamkT0qlUqFi3kNDQ3C5XIhEItfNml1ZWUE0GoXL5UIqlaqJw5zhxpCJK+FwGP39/XQEmVqthtVqpe+Lx+OIx+N46aWXqCRHLBarie8A6SyfmJjAb37zGxgMBjoDulAo0PGK8/PzSCQSNbGm9WAzrTuVSsHhcKBcLsPv90Mul0MoFMLn8yGRSGB+fh7RaBSpVKpqykFYt7oBLBZr09ydSqXysaZd36trZ9Zd+zDP+u1xr64buDNrJwLHarUag4OD0Gg0MJlMaGlpwb59++h0GiKY+/Wvf31dhogz9/z2uFfXfi+smzHkbsK9unZm3bUP86zfHvfquoE7t3axWAyBQACDwUB1MRUKBQwGA30P6dI+e/bsuujnMff89rhX134vrJsx5G7Cvbp2Zt21D/Os3x736rqBe3ft9+q6gXt37ffCum+ulcDAwMDAwMDAwFDVMIYcAwMDAwMDA0ONcsvUKgMDAwMDAwMDQ/XCROQYGBgYGBgYGGoUxpBjYGBgYGBgYKhRGEOOgYGBgYGBgaFGYQw5BgYGBgYGBoYahTHkGBgYGBgYGBhqFMaQY2BgYGBgYGCoURhDjoGBgYGBgYGhRmEMOQYGBgYGBgaGGoUx5BgYGBgYGBgYahTurV68F4bN3ox7de3Mumsf5lm/Pe7VdQP37trv1XUD9+7a74V1MxE5BgYGBgYGBoYahTHkGBgYGBgYGBhqlFumVhkYGBgYGD4pLBYLDQ0N0Gq1aGlpAYvFwtjYGKLRKHw+HyqVTZP1YmDYMBhDjoGBgYHhjsNiscBms7F9+3b09/fj2WefBYfDwfe+9z1MTU3B7/czhhwDwx2AMeQYGBgYGO44EokEMpkMAwMDeOCBByCTyRCLxeDz+RAMBhkjrkbgcDjgcrlobW2F1WpFc3MzZDIZZDIZ2Oy11VmxWAzRaBRutxvhcBgLCwtIpVKIx+PM/V5H1t2QY7FY9O9KpXJP30zyWdzsf5PP5l7+jBhqF/I8k+d39fN9rz/TLBZrzedzL3weIpEIarUa3d3d2LFjBwAgFAphaWkJ4XB4g6/u00Hu54fv6Wa8rxwOBwKBAF1dXRgYGMDg4CDMZjMMBgO43LUmhMfjgcfjwYULF+B0OpHP5xEMBpFKpVAsFjdoBZufdTXkrFYrdDodent7IZFIMDIyglAohNnZWZTL5fX8p6sGPp8PHo8Hg8EAlUqFrq4uVCoV5PN5qFQqqFQqSCQSsNlsXLx4EYFAAMPDw8jn85tyU2DYfJBnWywWAwB8Ph84HA5MJhPEYjEkEgmcTicWFhZQKBRQKpU2+IrvHmw2GxwOB0899RS6u7vh8/kQCoVw5MgRJBIJJBKJjb5Eho8Jh8PBwYMH0dzcjO3bt4PFYsHlcmF0dBSHDh1CoVDYVEYLj8eDVCpFX18fHn30UZjNZkgkEnA4HORyOfh8PuTzeWQyGVQqFfD5fAwODmJwcBD79++Hy+XCf//3fyMUCiEUCm30cjYl62bIsVgsaDQa+rDLZDKkUikIBALMz8+jUChsSkOFeGl8Ph9cLhcSiQQikQj19fUwGAwYGBhApVJBJpOB0WiEwWCAXC4Hh8NBPp+HVCrF9PQ0SqXSptoMboVAIKBe34ejlMA1bzeXy6FYLFbNc8NiscDlculBzePxwOFw1ryHeOjFYhHZbBblcnlTOjAKhQJGoxFarRZsNpt+Ni0tLZBIJFCpVABADZdsNls193G94fP5EIlE6OvrwwMPPIC5uTl4vV5MT09jaWkJyWRy034O5HtB0m/k+1Aul2t6zSwWCy0tLejv78fjjz8ONpuN6elpVCoVHD58mH7nNwtkj5NKpVCr1QCAbDaLbDaLVCoFt9uNdDqNZDIJiUQCiUSCpqYmqFQqmEwm1NXV4dVXX0WxWMTKykpN33sCOefZbDa4XO6a6OxqOBwOyuUydWDXy4ldF0OOx+NBKBTi8ccfx+OPPw6LxQI+n4+GhgZMTE2tE3gAACAASURBVExgdnYW0Wh0U3qjEokESqUSg4OD6O7uxpYtW6BWq6FUKsHj8SAQCFAul1EsFiESiSASiegGZ7Vaqce+sLCAmZmZTfHQ3wgWi0WNt8997nNoaWnBY489RqM6qymVSvjVr36F8fFxnD59uiqeG6VSifb2dthsNrS2tqK/vx+NjY1r3lMsFpFKpTA0NIRf/vKXCAQCWFpaQqlU2hT3lcvlgs/no7m5Gd3d3fS7nsvlAHxgoHM4HCwvLyMYDOLdd9/F9PQ0Tp8+jWg0usErWH96e3uxY8cO7Ny5E21tbWhoaEAmk0F/fz/Onz+Pf/mXf0GhUNjoy1wXbDYb7r//fhgMBgBAMplELBZDNput+YzD6jSqSCRCR0cHgsEg+vv74XA4sLCwsLEXeAdJpVLIZrP4wQ9+gF/+8pfgcDjUaCkWi4jH4ygWi8jn8zQDtWPHDjQ1NeHpp5+GxWLBl7/8ZQwPD8PpdNZ0RJ7FYkEsFoPH40EikcBoNMJut9N60NXweDzU19cjFArhxIkTWFxcxPT09Lpc17oYcsRK1el0sNlsUCqVYLPZqK+vRzqdRktLC7xeLzKZzKaLUsjlcthsNjQ2NqKxsRF1dXVQKBQQCoUolUpIJpP0oU+n0xAIBHQjt1gsMBgMsFqtSKfTmJ2drenNbjUsFgs8Hg9isRh8Ph8CgQASiQRyuRx2ux1tbW3o6em5oSFXLpcxNjaGYrGI4eFhpNPpDdsM2Gw2RCIR9Ho97HY7bDYb2tvb0d3djaampjXvJfc7n89jy5YtEIlEAIB4PI5cLkc9OfLebDZ719fzcVgdhWSz2ZDL5VCpVLDZbGhqaoLNZoPFYrnhz8rlcuh0Ong8HpTLZXg8HkilUhpt3WwpF+Kt6/V6dHR0QKvVUsdNKpWiUCjA6XRu9GWuGywWC0qlEmazGSKRCOVyGdFoFOFwGJlMZtMYr5VKBWw2mzrwZrMZwWBwoy/rjkLOaLfbjUAgQKPu5LXV2QYSvZNIJMhkMlheXobRaERDQwMikQi0Wi0SiQTS6fQGr+r2+HB2jc/nQ61WQygUQqlUwmQyYcuWLRCLxRCJRCgWi/RzEAgEaG1tpSnlcrkMl8uFfD5/xyO262LIkZC6XC6HRqOhoXWLxQK5XI5/+7d/w4kTJ/DDH/4QiUQCqVRqPS5jQ+jv78df/MVf0JTCyMgIMpkMkskk4vE4ZmZmEI1GEQgEIBAIIBQKqSfzz//8z7BYLHjkkUeg1Wpx5syZTWHkki+D1WrF9u3b0dDQgPr6euzcuRNWqxVCoRAcDgdCofCmhusTTzyB3bt34+zZs8jn84hGoxti5IpEImzfvh3btm3DN7/5TUgkEuqhfRg2mw2ZTIY9e/ZgYGAA586dw/nz53HixAm43W5otVpaLByNRmlKvVqNdy6XC7PZDIFAALFYjO3bt2P//v3o7u6GzWaDUCi86c+Sz+nAgQPYsWMHdu3ahUQigUKhgMnJSfzP//zPpkpH8fl8yOVybN26FZ/73OdoehkACoUCvF4vlpeXN/AK1w82mw0+nw+LxYLe3l4olUqUSiWcOXMGV65cgdfrRSwW2+jLvOMYDAYMDg4iGo1ibGxsoy/njpPP529ogK/er4hBNzw8jJmZGdhsNtjtdhw4cABarRaLi4sYGxvD6dOn7+alfyKIUWqz2WAymbBr1y5YLBbY7XYolUrU1dWBy+WCy+UimUwimUxiaWkJ6XQaYrEYUqkULS0tYLPZuO+++/DWW28hkUhgYWEBgUDgjl7rutbIEc+dQKIZdXV16OrqwsMPPwy/349AIACRSEQt32KxCI/Hg0qlQlORhUKBhuQzmQy1fKuNaDSK+fl5WtMVDoeRzWaRTqdpPUEymUQ4HAaPxwOPx0OxWIRQKKReitlshsfjAY/Hq+mIJYvFglAohEKhgN1uR319PXp7e2EwGGAwGGA2m6HRaNb8zM2MmFQqhVgsRj/Xuw2JMOj1emzbto2mzEkqgaQfcrkcyuUy9d7I38SQTafTKBaL8Pl8awy5eDwOnU6HdDqNWCyGZDKJTCaDeDxeFdELsVgMhUKB3t5eGl1rb29Hc3MzdDodxGIxvTfEQwc+2AxFIhGtgeRwOLBarbRuhMViYWBgYE3EhvyuajVqPwqVSoXu7m40NDRALpcjnU4jGo1iYWEBkUgEs7OzmJ6ertnv9q2QyWSwWq2wWq0wGAwQCoUol8sIBALw+/2boj4yl8shm82uWYdYLEZdXR00Gg39PtRyGvFG3O59KxQKSKfTmJubow66XC5HZ2cnVlZWwOfzq/YMJ1gsFrS0tKCpqQkmkwl2ux1arRZWqxVcLhfRaBTpdBqRSITW/waDQWQyGRgMBmi1WrS1tUEgEIDL5UKtVsNkMq2LA3fXdeT4fD7q6+thNpvx8MMP4/LlyxgZGUFDQwPUajVkMhni8Th+8YtfoFQqoa6uDslkEtFoFH6/H6FQCC6XixZNV9uGMDY2Bo/Hg0gkcsNC5ptJjJBQNJvNht1uRywWg0QiQaVSqfqU240g6XW9Xo+uri78+7//O/R6PYxG43Wt+x9FpVLB2NgYpqamsLS0tCE1chwOB+3t7bDb7fjmN78JtVoNiURCX/d6vfB6vVhaWkKxWERLSwtUKhVaW1upM9PZ2YmOjg489NBDKBQKkMlktEGCGPkLCwuYnJzE1NQUnE4nxsfHN1yqgc1mw2QyobGxEX/zN3+D+vp6dHR00BQLkRZaWVlBMpmE1+ulxjapJbFYLDCbzUgkEohGo2hra6Odb62trdDr9RgfH8epU6fgcrmwvLxcNUbsJ6GtrQ3PP/88Ojo6IJVK4XA44HQ68cMf/hBOp5Om3DfbQQ9cUyv4whe+gH379qGzsxMsFgvpdBqTk5O4cuUK8vn8Rl/ip6JSqSAejyMcDq+5fyqVCj09PTh//jxMJhNtZrlXKRaLOHXqFJaWlvDMM89ApVLh4MGDyGazOHbsGBKJBDKZzEZf5k3Zs2cPvvOd71DjfHU62e/344033sDVq1dx9uxZpNNp6qzl83n09/ejo6MDu3fvhlAoBJfLhcFgQF9fH/x+P+bm5u7otW6YIDApdLdarWCz2VCr1RCLxRAIBFAoFPjMZz6DfD5PIxYsFgvBYBArKyt4//334XK5qHdXTeRyOUSjUWQymdvepJVKJTQaDbRaLaRSKbxeL3w+34ZFn+4EIpEISqUSX/7yl9HW1gaDwbDGcAFAIzKpVArlchlCoZA2ynyYje524/P5ePjhh2G326FQKMDj8VAqlTA3N4fx8XFMT0/D5/MhHo+jXC7DbDbTWkk+nw/gg04nEn3m8Xh0cxAKhdDpdLQWgxTJGgwGuFwujI2Nbdimx+Vy0dnZidbWVjpuabV+1Pz8PPx+P0ZHRxEOhxEIBOhzy+VyIRaLYTQaodfr4XQ6EYvFsG3bNlo/BgCtra2QSqUwGAyYnp6Gy+XC6dOn73gKYr0hdYQqlQotLS1Qq9VgsVjQ6XTgcDj44he/CLfbjSNHjiAcDlNjrpojEx8HFosFqVSKxsZGqFQqasTF43HE4/FN0aVbqVQwMzODfD6PvXv3oq6uDnV1dbQ2rKGhAT09Pbh48eI9bchVKhVaUkSMd6VSCZ1OB6vVCrfbXVWGHJvNhlQqhVarhd1ux86dO2E0GiEUClEsFuFyuRCJRDA+Po5AIICLFy8iGAwiEAigUCggn8/Tc6yvrw/d3d0QCASIx+O4evUqXC7XbQcvPi4bZsiRFEtDQwMaGhquEw9taGhAOp2G1+uFQqGAXq9HJBKh6ZdisYhQKFSVhhzp2rtdSLjWbDZDLpdjZGQETqcT2Wy2Zg05qVQKi8WC559/HvX19Td8Tz6fRzabhd/vR6lUosb8rWqtNgqBQIAnnngCPT09YLPZNN0/PDyMF198ETMzM/D5fPRAlkql2LFjB5599llqyK3+XcDazjc+nw+9Xg+9Xo/29nYqYWAwGDA+Po75+fkNNeT6+/vR3d2N5ubm6+7P5OQkzp07h8OHD1NFd+LEEIdNo9FApVLB7/cjk8nA4XCgubmZGm+dnZ3o6uoCAAwPD2NqagoOh6MmDTmBQEAPA+DafSaHfU9PD+1MdzgcWF5epgfAZoCUILS0tNCyiVQqhXA4jEgkQh2dWobUgLlcLuzbtw/ZbBYmk4nWSzU3N+P++++H0+mEx+PZ6MvdUEgEnmSV1Go1zGYzmpubEYvFqqoxhMPhQK1Ww26342tf+xpaW1thNpuRzWaRyWQwOjqK2dlZ/O///i+CwSDi8fh1v4M08D388MPo6emBQCCAx+PBsWPH1vUsr+oRXXw+HwaDYc1ByGKxqBd/9erVmul+uRFk9Mn+/fuxY8cOSCQSBINBvPLKK5idna3ZtAubzUZjYyPa29vX3DsSVZufn8f8/DxGR0dpfaRWq8VXvvIVSKXSNb+rVCohl8vB4XBgfHz8YxvJd5LVNZ/Ly8u4ePEizp07h6mpKcRiMRoxZLFYa2obyRp8Ph98Pt8N08MkklNXV4f29nYoFAqqP2Y2mxEKheB0OnHmzBlaK3o34PP5kMlk6O7uRldXF13HyMgIAoEAPB4Ppqen4fF44HK5kEwm1xzU5XIZ+Xwe4XAYqVSKdhyPjY3RFGpjYyMOHDgAs9lMIzkNDQ3YuXMnpFIpJicnkU6nN/Te3y4qlQqDg4PYunUrKpUKUqkUkskkHA4HYrEYdu7cCZlMhmeffRazs7OQSqVYWFhYN1mCu4lEIkFrayvsdjuam5shkUhQKpVw6tQp6pzGYrGaj8gB157rdDqNU6dOIR6PY9euXVSWw2w2Y2BgAKdPn4bf70c4HK66gMPdpFwuY3l5merQFYtFWudeTXC5XCiVSlgsFhpNCwaDeOeddzAyMoKpqSmEQiEEg8GbljsplUoYjUYYjUbIZDJcuXIFExMTeO2112gGyu123/lrv+O/8RNCCh9XSzJwuVzI5XJqAJDwpV6vRzqdvmGnYC3B4/EgEonQ09ODffv2gcPhwO/34+zZs/D7/TXrubJYLJhMJlit1jX3iESZ3G43Ll68iPfffx/z8/NoampCc3Mz1dlb/f5cLod0Og2Px4OFhYUNra9ZXQ8Wi8UwPDyM6elpeL3e695LnllixJE1kDTsh+U2SOSqu7ubyvUIBAI0NDTQ9KpGo8Ho6ChKpdJdNeREIhEaGhpgsViQTqfh8/lw5swZTE9P48qVK4jFYkin0zdMfZN7TjT1CIuLi3C73XC73Whra4PFYgGLxYLNZoNEIoFOp6N1eF6vl6652o0AiUSCgYEBKkWTTqcRCoUwPj4Ov9+P1tZWWCwW7Nu3DyaTCQ6HA7lcblMYckKhEC0tLbDZbDAYDFQvc2JiAidPnqQdfZsBsjdNTEzQ1BtBrVaDw+HAaDRCqVQiFovd04ZcqVRCNBqlaWbyXFTbd5mkVokjSZoXzp07hzfffPOWBhxBJpNBq9VCpVJBKBRifn4eExMTuHTp0rpG3jfckCMH3s9//nNMTEzgG9/4BqxWK32tUqkgGAxicXERJ0+exNjYGJxOJ02x1iLEgHvwwQcxODiIvXv3QqVS4cc//jHNv9eyJAuLxYLRaER9ff2aWqpUKoVAIIDjx4/jV7/6FRKJBCqVCnp7e9HV1bUmekckWk6ePInR0VGcPHkSXq93wxs/SqUSEokEnE4n3nnnnRumTsgItqWlJbz66qsQCAQolUo4fvw4Tpw4gXQ6fZ0hRrq8z58/j9dffx2dnZ1obGzE7t27YTQasWfPHlgsFpw6dQput/uOF8vejIaGBthsNsjlchSLRZw5cwajo6N45ZVXkEgkqC7iJ6lfLJfLiMfjmJ2dxU9+8hMcOHAARqMRcrkcdXV1+MIXvoBQKIRcLoerV6/i/PnzVX8gyuVy7N69m+5hs7OzOHPmDN59910sLi4ik8mgvb0dTz75JHQ6HQ4ePAgAGBoauquR1vXAZDLhm9/8Jurq6gBcS6tFIhFMTU1hdHR00xhxhGKxCKfTCZ1O9//Ze/PguM/7vv+194G9sPcCWByLg7gIgJRAUqQoWRQlU5JtKXIVy3WUmcTxpE3/qDsTN2ln+k+nM3En0yszbTJu7EmTaZymqmQriiVbtilKIiXeF+4bixu7i13sfe/vD/6+j0nxMEkRwC61rxkMJWCx+D77fb7P83k+x/sjZIOuF4y12+3YbLZbHvQ+T+TzeZaWloRTxmq10t3dzdraGvPz8zt9eQ+Uxx57jEOHDmG1WonFYrz11ltMTk6Sy+W21DGzI4ac5F2T3KvZbJYrV65w9uxZXnjhBZRKpTDwisUiS0tLTExMiNesra2VVZLk3SCF5bRaLXq9ntraWnbt2sXg4CButxuNRsPy8jILCwvI5XJUKlVFN2KWemxeLz8jeWXW1taYnZ3FYrFgNptvSBaWkDwZ4+PjnDlzhoWFhbLQnpLCZZubm3e8pmKxSDweZ2RkRBioY2NjTExM3PH9A4EAfr9f5BV5PB4UCgW7d+8mnU5jMpmEsPBWIxnkLS0tQrh6fn6e2dlZ/H7/A1mY8vk8m5ubTExM0Nrayvz8PD6fT8wLq9VKe3s72WyWCxculH1hgFSpLW1aGxsbzMzMMDc3x/z8PFevXqVQKJDL5dBoNDQ2NuJ0OtHpdGVvpN4JSRR3165dWCwW4Fq+sJQj9TB28ZDWgkQiccM6rVAokMvlWCwW7HZ7xUeOPiulUumG/rMqlQqj0Vh2n4t0nZJYfz6fR6lUihz9XC5HIpEQBUrXexVVKpXoXtXW1iYEsGdnZ2/Ind4qtt2Qy+VyhMNhxsbG+NnPfiZUjy9evMja2hp/+Id/iE6nExWb8XhchFWSyWRFFgDI5XLMZjMOh4NDhw7R0tLCnj178Pl8tLS0IJfLyWazDAwMiI1rZWWF06dPi0TLhwmlUolWq+WFF15gz549fPnLXxbGrEQkEmF0dJTh4WGGh4fLxlORy+VYXl5mcXGRYDB42+sqFousra3xV3/1V0Km425kU/L5PIVCgaGhISYmJggEArS3t/Pv/t2/EwVCn+7pulXIZDK+8Y1v8Nxzz6FUKvH7/bz99tsPzIiTkCqXf/nLXzI2NsZrr73GsWPHaGxspKamhm9961sMDw9z+vRpgsFgRRgFhUKBVCrFwsICly9fJhKJkM/nRYjxO9/5zg2HmIaGhrKr4rtbJKPFbDbf0L4pkUgQCAR23Iu+E8hkMp5++ml8Ph9jY2OEw+GdvqQdQ2rj5/V6kclkJBIJlpeXyy7qlE6nGRkZQaVS8c4779DR0UFPTw9/8Ad/wNe//nUh5H7hwgWCwSCTk5NCJFnq7vPEE0/Q39/Pz3/+c8bGxpiamiIcDm+5M2bbDbnrW/IMDw+LaqZwOEw6ncbv96NQKMjlcsJrV8lIrTu6urpwu9309/fT0NBAe3s7drtdtO6SRFK1Wi21tbU4HA7i8TjLy8vMzMzs9DDumVs1uJfCh1arVRRDSO2LJE22dDpNIBBgbm6O6elpgsFgWW0EUlWiRqMREiS3K0rJ5XKEQiGRW3e3xo+kHZjJZFhZWUGr1ZLL5VCpVLhcLjY2NkSu3lYiVSA6HA7m5+dZW1tjfX19SwwpKcyaz+eZmZlhampKyPFYrVZcLhcNDQ3ipFtuSDlR9fX1Qvg4k8kQj8cJhUKiUENqLi5FHLLZLNlstqLFY6Vnura2FoVCIQ7eKysrjI2NlUVv5K0knU4zMzOD2+3G4/GIKITNZiOXy2G321ldXRWpJJ8npC4wtbW1ohep1JmnnNZ1uLYGSW3FLl++fEMnG5fLRXt7u/C0B4NBDAYDyWSSjY0NWltbaW1txWQyATA1NcXY2Jjw7G01227I5fN5YrEYMzMzvPvuuyK3QNrkpFDVwzDhZTIZPp+PpqYmvvOd7+D1ekWo7PqTq6R8f+TIEfFZSHlYr7/+On/6p3+6wyO5NyTBTKmSU0Jq3XPgwAHcbjdPP/00ra2tN3iYlpaW+Ou//muuXr3KyZMny06HSavV0tPTQzAYpLGxkUAgwNra2m1f/1k8V6VSCb/fTy6XI5PJ4HK5OHbsGDU1NXzwwQf3/b73Si6X48yZMwwNDYnK1K1AOt2eOHGChYUF6uvrcTqdAJjNZp5//nnOnz/P7Oxs2a0PNTU1vPbaa/T09ODxeMhkMqytrbG4uMjMzMwt50EqlRJG6/T0dMWGVtVqNYODg/T19YkONbOzs/zoRz/i//7f//vQ9dH9NMvLy/zn//yfOXjwIL/7u7+LSqUSArA1NTUMDg6i0Wg4depURVRePygkIfHm5mZaW1txuVzIZDIRkSvHA1k+n2dycpL/8l/+C0eOHCEajdLc3IzT6WRgYACtVsvLL79MoVAgGo0KD53b7cblcqHT6VhcXOT111/nypUr2/ZMb4khJ1WlJJNJYrEYer1ebNZSNZ7FYsHn84lwqdR+qdwW6M/K9SfvaDRKIpEQbYwkkczrUSqVeDwe0Zi9ubmZ3t5eVlZWCIVCOzSKe6NUKrG8vIzZbL5hIktVyFLIzGq1ipZs6XSayclJJicnGR4eZn5+nkQiUZabm2SIX2+MbxVS5Sv8qndrTU2N+Ny2wysntVrTaDRbnrMpHQJWV1dZW1sjGAwKYVlpgyw3pIrjuro63G43CoWCeDzO5OQkoVDotsa8lH8jfUkeyUpCasG3a9cufD6feCbkcjnpdFoo3ctkMmw2m2hJmM1myyLn9UEgrV3Nzc1kMhnR1UbKdbZaraIS/fOETCYTrRi1Wi0ymYxUKkUikbhBJLjcKBaLZDIZ/H4/H330EZOTk1gsFjo7O7FarTQ0NKDVarFYLOK1JpMJk8mESqUilUpRW1uLzWYjHA6LdJmtZEtWRcm1Hg6HWVtbo66uDr1eD1xb9CTl7y9+8Yui1+ro6OhD2URaCg8vLi4SCASYmZkRZc1jY2OMj4/f8Hqj0cjTTz/NwMAAv/d7v8fAwAAvvfQSP/vZzyrGkCsWiyIv6PrQuF6vR6/XU1dXd8PrJe/FD3/4QyYnJ/nwww8fytzAz4pCocBsNmMwGEQ+4XYYunK5HLvdjsPh2JbNKBwOk8vlmJiYwGKxMDAwsOV/87Mg9dRta2sT4uarq6v88pe/ZG5u7ra/p9VqaWxspLm5maamJubn5yuusrO2tpb6+nqefPJJfD4fSqVSGLaS1wKuzd2Ojg6MRiOJREKIAz8MB/d4PM65c+fwer0kEgmUSiUajUYYclLIfbtyW8sFhUJBZ2cnPT096HQ6UfwjpVOVW2j100j52dLB5KmnnqKlpYUjR47Q0NDA/v37RU9hCWm+d3V1AXDp0iXRvmsr2TJDLpPJ8N5777G6usorr7yCz+cTbWoMBgOtra0899xzxONxYrEYY2NjQmA0HA4zMjJS8YrnUv/JTCbDm2++KRrtplIpUqkUwWDwphtcKBS4ePEipVKJqakplEolhw4d4urVqzs0ivsjlUqxubnJ/Py88DLeyntVLBZZX1/H7/dz4cIFlpaWylIsslgsisIcq9WK2WxmYGBAyMVsByqVCo/Hg9PpFH14t9KQK5VKzM3NMTo6itVqFS3HVlZWtlSRPZvNkkgkOHfuHOl0mubmZnE95YZMJqO1tRWfz0ddXZ1oyXUvvy8dbiotBCm1Hquvr8dut2M2m5HJZEQiEYaGhpDJZPT19dHX10d9fT0+nw+tVks2m2VmZoZCocDGxkZZqft/FjY2Njh//jydnZ10dHQA1w5BnZ2dFIvFsqvS3Gqk+eF2u5HL5SSTSaanp1leXiadTldMTqgUVZuamiIYDLKxsUF9fT3RaBSv10tPTw/xeJx4PI5er8disfD8888zMDAgChfPnTtHLBbbMi/0loVWs9ksx48f56OPPqK9vR29Xo/VakWj0WAwGIRXTmJ8fFyI4c7NzTE3N1f2UgN3QygUIhQK3bVeTiaT4dKlSwBMT0/T2NjIgQMH+NGPfrSVl/nASaVSxGIx5ufn0Wq1uFyuW55IS6USgUCA+fl50auzHJESXIPBIBaLBZPJRH9/P+FwmPPnz2/L35dO906nE6PRKAyerfybkiH3xBNPIJfLaW5uplAoEAgEtsywkqQKLl68SCQS4aWXXto2yZV7RS6X09raSnd3txCAvZ67+Yxqampwu91MTk5u1WVuCVK4tK6uDrvdLpLZI5EIw8PDKBQK+vv7+cY3vsHAwIDotVwsFrly5Ypow/awGHKhUIiLFy9isViEISd5paTUgIeJTx9YPj3X5XI5TqcTl8slnBjT09Osrq5WXK6gtBYCXL58mfr6elQqFYODg/T09BCNRllbWxOdaZ599lmSySQtLS2Mjo6yvr7O8vJyZRly1yPlS0n6UJLr/dN4PB5RrRmJRPD5fExNTfHuu+8Si8Vu2dfsYWZjY4MTJ06wf/9+Wltbd/py7pl8Pk80GuUXv/gFgUCAvr6+W953yQN59erVss2ZgGvXOTs7i9FoFLlAOp1uyxdnybvt9/vR6/U0NDTgcDg4fPgwly9f3nJZg/n5ecxmMwcPHqS2tpYvfvGLnD9/npGRkS3PlZNO7u+88w61tbVlmS8pdTFpbGy8YS7E43Hm5ubuKqE7Go0yNzf30FR31tbW0t/fj9FoxGg0CkFplUolqq19Ph/f/va3+fGPf8zk5OS25BFtNVLf6IflPn4aqaVkb28vTqeTgwcPotPpyGazIm1odnaWubk5isUier2e9vZ22tvbUSqVorinUlKEbodarcbtdvPVr34Vt9tNNpvlo48+4q233kKlUqHVann88cdFpavb7aa5uZnh4WFOnjwpcsAfZI7ztmQOb2xsiAmuUCjEA309NTU1GAwGHA4H6XQalUolWhJJblmpwvXzQCqVYnZ2ltbWVpFwvh2SEw8KSfR5fn4eq9V620VaCq2urq6W9UIuXef6+jrFYhGFQoFard7yvBepcCgcNJ8pXAAAIABJREFUDhOJRKivr0er1dLQ0HDH/KsHRTAYZHl5mUKhgF6vp6uri3A4jMFgIJ1O3yQx8yCRkv+np6dxuVwYDIayLHYwGo3U1tYKvUC45lkPBAK/NjdGkmPa2NioOC/F7ZAOHPX19dTX14vvS/NEJpNhtVp5/PHHmZqawmQyEY/HKz4nVlJkKOcD6WdBp9Oh1+tpa2vD5/Px3HPPYTAYhF6i2WwGrj230nrhcrmw2+2USiVSqRRra2sV7ZSRCr9qa2vp7e1FrVaLwqYPPviAVColciM7Ozvp7u6mtraWuro6DAaDEJMPBoOiK86DYMtXxXw+z5tvvsmJEyeYnJzE7XbT1tZ20wbY2NhIbW0tHo8HjUZDd3c39fX1dHR08Mknn/DWW28xOzvLysrKVl9yWSGJwOp0Okwmk6h6LXck9e7HH3+cnp6e23quFAoFu3btIpfL8Ytf/KJsF/N8Ps/Vq1cplUr8xm/8BkqlEpvNJrpXbFUKQKlUEoK5Uj/TtbW127YHe9B/e3Jykng8TjKZpL6+nr6+Pux2O3V1dRw/fpy33nrrli3HHhRqtVp4pd1uNzqdjr/5m78pqwONRqNBq9XeUAgSiUS4cuXKTQndUmcXyeCTFOKlNmeVhlarpaam5oaDudlsvqnlHlxr2SVpZiqVSvR6PY899hh/8id/wv/7f/+PEydOkE6nK/JzeJiRnC+vvvoqTz75JL29vVitVtFTtlAo0NTUxJ49e1hcXGRxcZF8Po9cLqetrQ24Vjhw9uxZfvrTn1Z0tbJWq+Xo0aP09/ejVCoZHx/nxz/+MSdOnCAQCIi5+/rrr2OxWDh37hwul4vu7m4aGhr45je/SVdXF5cvX+aNN95genr6gezn23K8XV9fJxqNMjQ0RCAQECXa1xOPx3E4HMhkMkwmE2azGbVajVarZXNzk9HRUSFnkkqlytp78yCQKmWkz+n6034lIHVv8Hq91NXV3bbaUSaTYbfbcbvdGI3GshWBLhaLQjMukUhQLBaxWCzYbDYcDgfRaHRLrluaB2q1WmyM2Wz2loUyW0EsFkOj0bC5uUkymcRgMFBXV8fAwACLi4u43W4hEvygN2Cp+k/6jLVabVl65ORy+U1SNLlc7iYBWGltM5lMyOVy0f1B+qqEA9r1XL9WX38wVyqVGAwG8f9S2yO/308mkxHRF51Oh91up7+/n/Pnz4v9odyrGe8V6TAuiYiXY4rAnZAawXd2djIwMCCE60OhkAiJSxEKj8eD1WoVB7uampoboi6BQKCiPZZKpZKWlhZRnR6Px5menr5pXGtra2xubqLT6XC73cA1I3BgYIDGxkaKxSIffPCB6HDxWdfObVsV0+k0H374oTiNfdookRSgf+M3foPOzk6+/vWvU1NTg8Vi4bHHHqO5uZmf/OQnnDp1itOnTz80CbK3Q9KYkjo/JJNJNjc3y8oTcSeMRiNOp5Mnn3ySpqam24YgFQoFe/fuxel08sknnzA1NcXZs2fL7lReKBSYmJggn89z+fJlvF4vg4ODQhLkxz/+MefOnXvgf1epVFJTU0N/fz+7du3adkMmlUoRDof5xS9+wfr6Os8++yxms5muri4ymQxarZZ3332XixcvPnAdNKmbg1T19pOf/IQzZ86U3TNwtz2RpQr03t5eNBoNiUSCqakpJiYmmJiYKLs5/+uQy+UcOnSIxx57THRmuRXz8/MsLS3xp3/6p6yvr/PII4+we/dufv/3f19UgB89ehStVsubb765LSkD24UUijOZTHR0dKBWq5mZmSm7OXwnnnzySb75zW/S09NDQ0MDm5ub4n6urKywubmJ3W6ntbWVL37xixw9elT8rkKhENW8o6OjW948fqvR6XS88MILtLW1Cc24lZWVW4qkp9NphoeHGR8f5/Tp0+zbt49AIMCBAwd4/vnnOXPmDLlcjqtXr35mJ8C27gp3ygGRkiE/3UtVLpcLcdy6ujoaGhq4cuXKdlzujiLlIjidTlH+XCkPv0wmw+l00tDQcIPmmSSCLCV51tbWihZmVqtVuKtHR0dJp9Nld3KTqkSXl5cxGAyo1WqcTiddXV2cPXtWhL4fpLdYEh6W8vGSySSpVIpcLrctXmlJ4mRmZga9Xk82m0WtVqNSqXA4HHR2djI3N0cikWB1dZVEIiFyP+7X8yA1oG5pacHn82E0GlEqlSSTybLPI7vTMyqXy6mvr8fr9aJQKMRndH1D8UpCMlIksdfbEYlExNyQZCckw1dqMO/xeOjs7MRisVSk1+pOKJVKjEYjjzzyCDqdriw7k9wJvV6Px+MRz+Hq6irz8/PMzMywvLxMMpkUws8DAwPk83nUarWIwuTzeVZXVwkGgxU17k9TW1uL0+mktrYWrVbL0tISy8vLd4yOSM+31G5xbGyM9vZ2Ojo66OjoEN/f2Nj4TAoEZROnkEJINpsNu91+w8IgCcl2dXVRKpU4efLkDl7p9lBXV8drr70mhJQrCblczuDgIP39/TdcfyAQ4Ny5c8IYOXLkCD6fD41Gg9vt5lvf+hanT5/m/Pnzv7b11U5QKBSIxWJcvHgRgH379uH1enG73QwPD7O8vMzExMQDTeZVKpWo1WpkMhm5XI7l5WXW1taIx+PbZuhms1nee+89lpaW+M3f/E00Gg1qtZqmpiYaGxsxm8309fVx/vx5VldXGRoaIhqNEgqF7mvhlkLWX/nKV+jv78fj8YhFrpI3AoVCwZ49e9izZw9qtboijbf7YXZ2lqtXr2I2mzEajTQ0NNy0xnd1dVFfX8+7774rtEQflvQZrVZLfX09f/RHf8Tx48f52c9+VnaH1DshdTFQq9UUCgU++ugjLl26xJUrV0SUaGVlhYmJCXbv3s2RI0ewWq039M++fPkyc3NzFfv8yuVydu/eza5du3A4HGSzWd555x0++eQTrl69+mvnarFYZGZmhkgkgtfrpaWlha9+9as8//zzoiHC5OTkfc/5sjHk6uvrqauro7u7m9bW1lsmx1utVpqamujs7CSTyTA7O1tWD4QkRdDc3MzMzIyocLwX9Ho9fX197NmzB5PJRCQSYXFxsWJKtlUqFTqdjvb2drq6ulCr1cKrs76+zvnz50VbNkl7SzrRS6f77agGvV+kRQlg9+7dOJ1OPB4PAwMD4l77/X5CodBn2ogkA+7AgQN0dHRQU1NDLpdjZWWFYDBIJpPZNkOgVCoRi8VYXV3l1KlTtLa2MjAwICqp6+rqUCgUGI1GNjc32b17N2tra6K35PXXKbXmkRrL32rMnZ2d+Hw+urq6aGxsRK1WV1zHg08jCQZ7vV5sNhtyuZxEIsHo6CjLy8s7fXn3RbFYZGFhAZfLRVNT003FDRKtra1oNBqhgC8dfq7Pm81kMkSj0S2tgt5q4vE4s7OzrK6uispuaR+Ty+WiMMRqtRKNRit2TsfjcTY3N0VunMViwWw243a78fl8N4wbrj3XVqt1y6WSthq9Xo/BYKBYLBKNRkX7rrvVu5Xa0p0/fx6VSsWxY8ewWq309fWhUqlYXFy8KSJ5t5SNIdfV1UVfXx8HDx68oeXF9UgSBPv370en07G2tlY2hpzkUezo6OBLX/oSb775pnAl3+3CJJfLMZlMvPTSS3R3d2MymZiamuL48eNbXqH4oJDyQfbs2cOBAwfQ6XQUi0VSqRTz8/O89957pFIpMpkMjz/+OG1tbajVarGJS/9drn0JE4kEH330EcFgUOTJNTY28tRTT/Hoo4+yvr6OXC4XJfj3iySc/eUvf5nHHnsMq9VKIpFgdnaWxcVFksnktm14pVKJzc1N/H4/b7zxBvv376e/v18YYi0tLbS0tHDgwAHg2oI1NTUFIDYsKT1gfn6ezc1NEUK7HmnuPP300wwODrJ7926sVuu2jHGr2bNnD3v37qWjowOPx4NCoSASifDxxx8zOzu705d3X5RKJUZHR1EoFHfMk9u7dy979+6943tJgqqVXLUqdbSYmppidXVViMYCQhBYEn+WyWQVachJvZAjkQiFQgGNRkNjYyOtra0cOnSIRx55hNra2ht+R6lU0tjYSCKRYGhoqCINdZlMhsFgwGg0CuWAH/3oR7fMjbsd2WyWbDbLz3/+cy5cuEBHR4fII7fb7Xz44YdCbupe2RJDzmw2i4axn477KhQKtFotNpuNlpYWUcX1xBNP0NbWdtMkuB7Ju7O8vMz4+HjZGHHwqzyCgYEBnnrqKVQqFT09PZw8eVKc0G63QEmdLp566il8Ph9PPvkkFotFxNQ/+OADlpaWtnlE94cU+3c6ncLrEg6H+fnPf86ZM2fw+/2i4EWq1isUCmVZjXgnQqEQx48fF3O8s7MTj8fDl770Jbq6uvjBD34gGr/fz8LlcDhobGzE5/NRX19PsVgkFApx8uRJxsbGdmQxlJJ3i8Uibreb9vZ2ent7bzK8FQoFHo+Hr33ta2Sz2RuEL5eXl0WF+qfvuZQb5/V6cblc6PV6SqUSiUSC9fV1Ll68yPT0dFltBNcLnufzeTKZjAh/w7VnW+oCsn//fkwmk+gSMjc3x5kzZyq2cEsKqen1emZnZ8lkMng8HuBaXpTUfjEcDpNKpSgWi6ItVzQaxe/3i3sZCAQIhUJMTU0JmZ1KQxLvnp6e5v333+eZZ57BZDKJnysUChobG/n93/99pqamuHz5MuPj4ywsLOzgVf96wuEwExMTqNVq9Hq96CmrUqnEvuV0OmlqahKtGCcnJwmHw9hsNlKpFB6Ph1AoVFHKC9cjk8loa2sT2nGfBcmAX19fJxgM0tTURLFYpLGxUeQc3itbsnsaDAYaGhpuedOkxM+mpib279+Px+PB5XKxb98+GhoagNvnwUgGwPr6OouLi2VlyEl5Xq2trSIHxufzsbq6Si6XuymnSQpLAZhMJux2O8888wxdXV0MDAyI0PHMzAyXLl3a0lZMDxKv18uePXuwWq3iNBqLxfj4448ZHh4mEAiIzU1KAi2njflu2dzc5MKFCyL53uFw4PP5OHjwIG1tbZw4cUL0Z73XE5YkydLS0kJdXZ2QNwmHw1y5cmXHQnHZbFYYLGazmXw+L/pnqlQqEQ5XKBTU1tZy5MiRm94jGAySSqVwuVx3tSAWi0Xi8TjBYJCJiYmyDENKosn5fJ5sNsvS0pLYtIxGI263m46ODnp6etDr9RSLRSGSLlXyVSKSQarVallcXESlUuFyuYBrSd6RSIT19XXm5+eFPI00h1ZXVzl37pyIWCQSCVKpFMlksmJzB6UUkqWlJS5fvszg4OANP5fJZLhcLr7yla9w5coVVCoV4XC47A25WCyG3++nubkZmUwmuhmYzWYcDgcvvvgiNTU1ov9zoVDA7/czNzdHW1ubyH2/1z7E5YRMJhP9gj+rIZfJZCgUCoRCISKRCL29veTzeVwuF4lE4r7mw5YYcl6vl6effpqWlhacTucNP5PJZEIMsra2Fo1Gg0ajwWKx3LZzgeRuvHjxokgKjEajZXVqSyaTzMzMsLi4SDAYxOVyCe2rxcVF3njjDVGZYjabaWhooKamRuTENTQ0sGvXLrRarfDEfe9732N2dvYzh+m2E7VajU6nu8FDUywWhVaOyWTiyJEjHD16lMOHD1NfXy+SaMPhMMFgkGg0WvZaUrlcjs3NTUZGRggEAjQ2NuJ0OrHb7TgcDl5++WUmJibEA7u6unrH95O0plwuFy0tLRw9epTBwUFcLhfRaJS/+Iu/YGRkRHgsdoJisUgymWRhYYGf/vSnwltcX1+P0+nkhRdeuO0BTsJkMt2UQ3Mnkskk3//+98VzX24HmlKpxNLSkqjoLRaLzM7Oks/nGRgY4IknnuCJJ56gv79fVDWvrq7yZ3/2Z4yPj1fMc30rSqUSi4uLRCIR/u7v/o7e3l5aW1tFCoDUsknS1JqYmCCdTgsvld1uJxgMsra2JqqwK/nzkHA6nezevfumvrtwLX1GqVSiUqlu266y3JA8cr29vZRKJfbt20dfXx+5XA61Wo3ZbCabzbK4uMj8/Dx+v59/+Id/YHJykn379uHxeHjkkUcIh8MolUqKxWJZ7d13i+RRhl+t1/cjBi8JLGcyGeLxuFCk+CzKFFtiyGm1WsxmMz09PXR3dwvduNst8Hf6GVzbNKXFQWqsXm6eHOkEGgwGWV1dxePxYLPZ6Ovro66ujvHxcVFtaLPZaG1txWQyibL0uro6cXNnZmaYnp7m9OnTxGKxijqx36oFm1wuR6/XYzab8Xg8dHd3s3//furr64VgZCaTEa5mSV6jnCmVSmSzWWGcz8/Ps7CwIIROOzo6UCgUNDc3o1KphCErHUqkh18S+5UElOvr69m1a5f4UiqVolJ2dHRUJBnvFFKXCUlmJBAI0NLSgsfjoa+vT7TxuV7I+vrn/25Os5KXJplMEgwGuXz5MmNjYyIvp5yQCkEikYhoLSjp67W1tdHX18fg4KCo+guFQgSDQc6fP8/S0lJFbmjXk0gkyGazjI6OolQqWVlZIRqNMj09zfT0NDMzMwSDQWKxGKOjo2QyGRobG0U+pHSf7yWXuNyRqjxvdVgpFAqk0+mKEoFOJBKsrKwQCoXY3NwUqVMymUx4WTc3N5mfn2diYoLx8XGuXLnC1NQURqORdDrNwYMHMRqNWCwWYrFY2R3I7oZ0Oi3mqlKppLa2VuQ6Smv79c+zlDcvcX1Hk5qaGlHkJ3miM5nMfe97W2LIXb58meXlZSKRCJubm7S3t2MymW5q5SJxfQ++T3+/UCgwOTnJ6dOnefvttzl58qQwbsrpwZdU2o8fP87y8jJf+9rXOHjwIC6XC5/Px7/8l/9SbOByuRyNRiM0wqT2PqOjo/j9fv7mb/6G+fl5gsFg2W1c90NdXR3f/va3yWazpNNp7HY7drtd5DxubGywuLjId7/7XWZmZlhZWamIBQ4Qyub/5//8H06cOME/+2f/jO7ubnw+H01NTXR1deH3+7l06RKBQEA0l5aqkA0GA4cOHcLlctHT0yNyTQwGA1qtljNnzjA1NcWFCxdYXFwsq/kQjUZJJBIsLS2hVqsZHR3F7XbzwgsvYDQagWveCen5v17t/05IRuL3v/99Ll++zKlTp8rWKy3liRUKBU6dOkVjYyOHDx8W+X5S/qukG3fu3DmGh4eZm5sjEons9OU/EHK5HJOTkywsLHD27FnRP1ZK7pY8bel0mlKphN/vF11rKrU12Z2IRCLMzMwwMDBw08/W1tZ48803OXv2LO+99949JcvvFFIl7sbGBsePHxeed6nr0pkzZxgaGuIf//EfyWQyZDIZYrEY+Xye0dFRIpEI+/btQ6vV8m/+zb/hww8/5PXXX9/pYd0TxWKRkydPEolE6OrqoqOjg+9+97uMjo5y/Phx4YSIxWLiIKdWq0W3KkCEZh977DF6e3vxer1oNBo++OADRkZGuHDhArFY7L6ub0sMOUkQdHx8XJxMnE4nXq9XaGJJSHklWq1WuJkLhQLBYJB0Ok0kEmFsbIzh4WEh61CuSEbJ5OQk4+Pj2Gw2SqUSFosFq9WKXC6nVCqJMafTaWKxmNBVGxkZYXFxkampKQKBwA1J4pXC9WPKZDKiErWurk64jiVBWUDkCy0vLzM5OcnKykpZ5T7eDaVSSTRIHxoaolgsYjQaMRqNNDY2otFoyOVyBINBQqEQRqNRlOLX1NTQ29uLy+Wis7NTiE4mk0nh2RgbG2Nzc7PsxHClDVq6X9PT02xsbNDY2CgMOYfDQTqdFp+H0WgU8jKfPrhls1nRSSISiQhV9HIPtUshktnZWeRyOXV1daI3stSKS0qE9/v9zM/PV4w35m6RNvC76aNZac/3vSJJ9cRiMbLZrOhHurGxwfz8PKOjo8zOzhIMBnf6Uu8KySCfn5+nWCzi9XoJhUJoNBqi0ShXr15lbGyMmZmZm343kUgQDoeZnp7GZDKJTkWVRqlUYn19HbPZTDgcxmw209HRgVwuJxQKsba2xvr6OqFQiHQ6jV6vFxEWqahLkmdpbW2lvr6ebDZLJBJhcnKSmZmZz6QNKruToSCTye7bipDCKBqNhhdffJGuri5ee+014YmRWF9fZ319XQiLwrWKwP/1v/4X09PTnDx5knA4TCAQ+Exq9qVS6Z6yLO937FIIScqZOnz4MM3Nzbz44osYDAYRjvP7/YyMjDA6OsqVK1dEmw+p1+iDzCO4l7F/lnsOCImFf/pP/ynt7e20tLTcMR8qm83yd3/3dwwNDfGXf/mXRKPRB2a8bue4///3ECGjr33ta/T09PDaa6+hUqmEUX69qn2pVBI5o1K/TqmnruS5+d73vsfIyIjIpbgbtmuu3+J9kMvlN4TWpUo3KR/oiSeeoKWlBYPBcFPVqt/vZ3h4WKjARyIRkRh8t2z3PZdQq9XipP6tb31LVGxLHjkphPhv/+2/ZXh4mKWlpQe6oe3UPS8Hduqe3w5J1P5f/+t/zRe+8AXMZjOxWIwf/vCHDA8P8+abb4pq7s/Cdt9zaX2SujZIoVUpKnG7sKAUgZL2Rqn37mdhJ+653W7H6XTyr/7Vv6K9vZ19+/aJ8WxuborinkQiIYo/rtfEldb4hYUF1tbWePvtt5mammJ8fJzNzU1WVlZ+7Rp/u3FvmeaD5F6X3O7pdBqr1YrJZLqh16p0+nY6nUKHKBqNcv78eVZWVlhZWRECspWAtEFL+TxSI2ilUolOpxPNtAOBgDidLy0tiby/SmrFdSs2NjaYm5tjYmJCVPp82pCLRqOiX18oFOLcuXPMzMyUXd7jvVIqlUilUpRKJUZGRshkMni9XtGpQJrzbrdb5I5ks9kbHmDJE3fhwgXGxsZEa6NKCD9Jhur1BorkqZGMPGnR0ul0NyV6r6+v4/f72djYEN6MShg3/CqKoFKpOHHihJjzOp0OrVYLXPssFhYWyq5Qq8qDJZ1OEw6HOXPmDMlkkpqaGlKpFBcvXmRhYaFitfKkZ/te87gkHdFKR4oWnDlzhkAggEajQafTodPpRLs5qbhTSpeKx+MiCid9Xb16lenpaYaGhkSLLynt4H7ZMo/cLd7rnkqPpevaCe8MPFgvxfX/Xnc9N/y7lWz36UWhUHD06FG6urr44z/+YxwOxw0/HxoaYmhoiDfeeIOrV6+ytLQkvJAPkp08qcvlcgwGA+3t7bS2tnLw4EHp73Ds2DHa2tpYX19nY2ODDz/8UBxUFhYWGBkZYXJyUkjs3OvnUu7emTutA5/1eSgH78yd1rqt2sDL/Z5vJeVwz2/zt26YBw+6oKN6z++OBz1uyWA7duwYdrud+vp63G43brebrq4unE4nxWKRWCzGe++9RzQaJR6Pk0gkiEajvP/++1y6dOm+Iozb7pG7xQVUtLflftlOg61cKBaLTE9PE4vF+G//7b/dlOQuCeWOjo6KdlOVeEK9E8VikXQ6zcrKCplM5oYqrbm5OaxWK8lkkmQyydzcnAizRCIR1tbW2NjYKLuCngfFwzim6/m8rnVVbqQ6Dx5OpJ7bly9fFmoMRqMRg8GAw+HAaDSKiKQUjcxms+RyOdLpNEtLS+RyuQe6522bR26nqZ5e7o7quCuf6ly/Oz6v44bP79g/r+OGz+/YPw/jLs+GllWqVKlSpUqVKlV+LVVDrkqVKlWqVKlSpUKpGnJVqlSpUqVKlSoVStWQq1KlSpUqVapUqVDuWOxQpUqVKlWqVKlSpXypeuSqVKlSpUqVKlUqlKohV6VKlSpVqlSpUqFUDbkqVapUqVKlSpUKpWrIValSpUqVKlWqVChVQ65KlSpVqlSpUqVCqRpyVapUqVKlSpUqFUrVkKtSpUqVKlWqVKlQqoZclSpVqlSpUqVKhVI15KpUqVKlSpUqVSoU5Z1+KJPJHpq2D6VSSXYvr/+8jr067sqnOtfvjs/ruOHzO/bP67jh8zv2z8O4qx65KlWqVKlSpUqVCuWOHrkqVapUqVLlQWOxWOjs7MRms+F0Orl48SKTk5Ok02kKhcJOX16VKhVF1SNXpUqVKlW2DZlMRl1dHd/4xjf44z/+Y37wgx/w4osvYrVaUalUO315VapUHFvukZPJZHi9Xux2O88++ywmkwm5XI5Mdi3UWygUKBQKLCwsEAgEOH36NBsbG2QyGUqlhya0XaVKlSqfa2QyGQaDgccff5zOzk727duHXq9nbm6OUChEOp2mWCzu9GVWqVJxbLkhJ5fLaWhooL29nX/+z/85Ho8HpfLan5XJZORyOTKZDOfOnWNqaor5+XnS6TTZbPahNuRkMpn4+jSlUkl8VXl4uf7eV+91lYcdhUKB0WjkC1/4Al1dXezevZtAIMDc3BzhcLgaVi1TPr1XfXqtuv7/q+vYzrClhlxDQwNut5vf+Z3fobu7G71eTywWIxKJiBsu/dvc3IzX6yWfzzMxMcH3v/99YrHYQzkxbDYbra2tHD58mEOHDgHXPod8Pk8ikeDChQusra0xPDzMxsYGq6urO3zFVR40SqWSQ4cOYTQaiUQibGxsMDo6+lDO9ypVFAoF3d3dtLa2cvToUdRqNSdPnuTs2bP85Cc/YX5+nlQqVfXIlRFKpRKDwcCjjz7KU089hcfjwWw2c+HCBTY2NjAYDCSTScbHx4UBPjs7y9zc3M5e+OeQLTPk5HI5drudlpYW+vv76ezsZHNzk2QyyfLystiwJEu/vb0di8XC3r17MRgM/P3f/73wzD1sGAwGWltbOXjwIC+99BJwzZDL5XJsbm6iUqnw+/2kUimUSiWRSIRcLlc9rZYh0vy9Vw+qXC7H6/Vis9lYW1sT6QZVQ67Kw4ZMJkOpVOL1emlra6OxsZFYLMbk5CSXLl3io48+2ulLrHILlEoler2e1tZWnnrqKXw+Hw6HA4PBwNraGiaTiVgshlKpJJ/PA5DNZgmFQuRyOfL5PMVi8aGLLikUCuRyOSqVCoVCgUqlEuOUxpzJZLb1ULIlhpxer8doNPLcc8/x9NNPU1tby8K3JbnfAAAgAElEQVTCAv/hP/wHFhYWWF9fF4NUKpUolUq+/vWv88gjj7B7924sFgtf+MIXmJyc5Ny5cw/dKc3n8/E7v/M7tLa2Ar/ySiqVSiwWC8eOHSObzfLKK69w7tw5fvrTn3L69GnGxsZ28rKrXIdCoUCj0aDX66mpqSEUChGPx+/5fWQymVgYqlR5GHE4HNjtdn77t3+b9vZ2AoEAV69e5Xvf+x5ra2s7fXlVboNWqxXGd09PD1qtFoB9+/aRy+VQKBQUCgUOHz4s9rCZmRlmZ2d5//33mZqaYnV1lWQySTgcplAoVPxerlQqaWtrw+VycfDgQXw+H4ODg0QiEcLhMENDQywuLvLOO++wvr5OOp3enuvaijc1m800NTXR0tJCU1MTm5ubrKysMDo6it/vJxwOixsvl8tRKpVMTk5isVjo7+/HYrHg9XqJRqNbcXk7hlwuR6PRYLVaaWpqwmAwkMlkSCaT5HI51Go1SqUSs9ksNvZoNIrf72d6ehqVSkU+n6+4041cLkehUKBWq6mpqUGhUKBQKFAqlSgUCpLJJNlslmg0WjEPu06no6GhAbVajVqtJpFI3LchZzKZHrq5rlAo0Ov16HQ6TCYTSqUSuVxOIBAgHo+TzWbFKfZhQqlUYjQa0el01NbWiu9Ho1FSqRSbm5vkcrkdvMLtx2KxUFdXh9frxeFwMD09zeLiIgsLCyQSiZ2+vM+MtK5L65p0uJPWt+spFArk83ni8TgbGxvCg1OOSNEGuVx+w0HTaDTe8NxeP88VCgUGg4FgMIhWq8VmsxGLxZiamiKZTBKJRLZ1DA8alUpFY2MjXq+Xjo4OWltb6erqIhqNEo1GkcvlWCwWxsbGUCqV+P1+4a3cSh6oISd5Fw4ePMg3v/lNurq6cLlc/OVf/iVnzpxhamqKRCJxwyQoFotks1nef/99ZmZmeOKJJ2hqauKxxx5DoVDwzjvvlO1Ev1d0Oh2dnZ10dHTQ0NBALBbD7/dz5swZVldXaWtrw2az0dfXh1arFda/FH6bm5sjGAxum5X/IFAoFGi1WhwOB83NzRw4cACr1UptbS1er5eamhpOnTrF3Nwc//iP/0gkEqkIo2bXrl185zvfYWVlhZmZGVKpFMFg8J7fx2AwcPDgQUZGRh6quW4ymTh06BD9/f185Stfoba2FovFwp/8yZ9w/Phx/H4/iUSCTCaz05f6wFAoFNjtdp555hkGBwd59dVXRbj87bff5sqVK7z++uusrq5uy+JeLjzyyCMMDg7i9XpRKBS8++67DA8Ps7m5+VB8Dnq9nu7ubmpqarBarTz22GPs37+furo6jEbjDUUCqVSK1dVVfvazn/FXf/VXBINBNjc3d3gEtyaTybC+vs7CwgLT09PU19djtVrv+Dtutxun00lnZyf5fJ5CoUAwGOR//s//ydjYGD/96U8rdo2TnCy/9Vu/RWNjI2azGbVazcLCgnA+eTweMpkMzc3NjI2N8R//438kEols+Tx/oIacXq/H4/HQ3NxMU1MTer2eXC5HMpkkmUze0Zskva5UKiGTydDr9cKV+zAgk8nQ6XQ0Nzdjs9lIJBKMjY0xMjLC2NgYgUCAQCCAzWYDwOl00traikajwWKxYDQaMRgMhMPhHR7Jr0cmk+FwONDr9TQ1NWE0GnG73Xg8Hrq6ujAajdTU1IjXdHd3YzabmZubY3l5matXr5btwy6Xy9Hr9VitVrxer/Cm3o9nSafTUVNTg1arRaVS3bKCuRLRarXU1tbS29tLZ2cnDQ0NYv7u3buXYrHIlStXCAQCjI+PV7yHSqlUolar8fl8NDQ08Oijj9LV1YXdbheGXHt7O3K5nLm5Oebm5hgbGxNeyYeV2tpaUdjV3NyMRqMhlUqxtrbGxsZGRY5dii44nU50Oh1msxmTyURvby8GgwGr1UpPTw+NjY1YrVZqampuMOQymQxKpVLIr1y+fJlUKnXfa8hWInkOZ2dnOXnyJI888giAcDJI6hPXc320RRqPQqGgv78fk8lELpdjfX2dpaUlUqkU2WyWQqFQdmO/FVL00Gw2o1AoGBoaIp/Pk8vlsNvt2Gw22traqK2tpbGxkXQ6jUaj2Za0mQdqyDkcDl544QWhE7S5ucnGxgalUgmlUnnXG5VcLker1aLRaB6azU2pVGK1WnnyySdxu93MzMzwv//3/+YHP/jBDYUfJpOJF154gb179/IHf/AHInRns9mw2+2sra2VdThCSmx+5JFHaGtr47d+67dwu9243W6xqeXzefL5vLi3zz77LPF4HJvNxrlz5xgdHS3LhQ1ArVbT0NBAS0sLra2t+P1+kQdyL0jFQE6n86HKj1MoFNhsNtrb2/nN3/xNHA4Hbrdb/PzVV1/l5Zdf5t1332VkZIT/9J/+U8WHW3Q6HTabjddee43Ozk6eeeYZ1Go18Kv814MHDzI4OEhLSwujo6P8+3//7wmFQg+VR/LTdHd3c+jQIZ5++mk6OztRq9UEAgERWi3H5/vXoVar0el0PPHEEzQ3NzMwMCDme01NzU06qXCjJIdarcbtdvPkk0/S3t7O//gf/4NAIMDm5mbZFfZls1nW19d57733+Pjjj/nWt77FM888Q0tLiziM383aZTQa+epXv0o6neYb3/gGx48f52//9m9ZWloiEAiQSCQqwjMrHdgMBgORSIT/+l//K4FAgGAwiNPpxOVy8Ud/9Ed84QtfoLOzE61WS01NDeFweMvv7QMx5JRKJQ6Hg127dnH48GFaW1uRyWTMz88zOTnJ1NQUS0tLd7xZRqMRh8OBSqWiUCgQCoVuyKWrZGQyGWq1GpPJRE9PD6VSibm5OTY2Nm6qRE2lUkxMTGCxWIhEItTU1KDX6294r3JErVaj0WhobW2loaGBo0eP0tjYiMPhAGBkZIR4PE4oFBLeR+n3nn/+eSwWC+3t7aRSKXp7e1lbW2NpaWknh3QTSqUSu93Oq6++KkIptzqV3g2lUol4PE4sFqOxsVGccislR/BWSJ70Y8eO0dXVhcPhQKFQsLi4yOrqKqurq/h8PqxWKy0tLRSLRdRqNXK5vGLHLJfLOXToEK2trezduxe32y1yWVOplMidsdlsaLVaPB4PMpmMo0ePMj09zccff1yxY78dUoqNw+Ggvb0du92OVqtlYWGB2dlZ1tfXiUajFbG2S/lvtbW1OBwOurq68Hq97N69G7vdTl1dnTDg1Go1CoWCYrFIPp9ncXGReDyO1+tFp9Oh0WjE++p0OpxOJ3v37iUWi3HixAmWl5d3cKS3J5vNEo/H+eSTT4hEInR2dorPwmw2U1dXd8d9SSaTiY4dtbW19PX1kc1m+eCDD7hw4YKocC13JFWJ119/nWw2y8rKikgPCYfD5PN5pqam8Hg89PT0bOsB/YEYcmq1mqamJvr6+njuuefEaXRsbIwPPvhAVHLcCZvNRl1dnVgEV1ZWCIVCFfGw/zpkMhlarRa73c6jjz7K3Nwcp06dIhQK3fTaTCbD1atXMRqNhEIhsTmWO1IRx+HDhzlw4ABHjhzB6XSSTCZZW1vj1KlTLC4uMjY2xujoqKjArampoaGhgb1799Ld3Y1KpWL//v1cuXKl7Aw5lUqF2+3mX/yLfyFC4NJcv1dKpRLhcJhIJIJOp0Ov14u5X6kbu1wux2g08vWvf5329nY8Hg/r6+tMTU1x+vRpzp8/z5e//GV6e3vZtWsXNTU1Fe91VygUHDt2jAMHDrB7926RDpJKpQiFQszPz+P3++ns7MTlcuH1enG5XLzyyiucOXOG06dPV+z9vh1yuRy1Wo3H46G3txe3241Op2NmZobh4WFWVlbKNi/s0yiVSmpqamhpaWFgYICXXnqJffv2odPpbipkkCgUCmSzWUZHR1lcXESj0eBwOFCr1WKu6/V69Ho9hw4dwu12MzExUbaGnBRB+cUvfsHx48fZu3cvXq+Xr3zlKzQ1NeFyucSBVkqN+jRSyFWr1TI4OMi+ffsoFousrKywsbFxzxGNnSCXyxEKhfjzP//zm362ubnJ5uYmw8PDGI1Gmpubt/XaHoghVygUiEQiLCwscPr0adxuN3V1dYyPj3PmzJm7emidTqfIoygUCoTD4YpIer8b1Go1+/btY2BgAKVSSV1dHc8//zz9/f28+uqrzM3NEQgEeP/99wmFQsLKlzb123WAKAdUKhVGo5H9+/fzzDPP0NPTg9frZWpqigsXLnD69GkhbhyPx4lEIjeE0iT9PCnEZLVaeeqppygUCpw/f75s9POkU6W0YC0uLnL8+HE++eQThoaG2NjYuKf3KxaLrK6uYjabcblcbGxsYLVab/p8KgW5XM7+/ftpb2/H6/VisVjIZrPMzc3xox/9iPHxcSYmJkgmk3z44Yf09/eTz+epq6tDoVAwPz9fcYc2p9OJw+GgoaEBp9OJUqkkkUgwPj7O2NgY7733nijekSQ4/vAP/5D6+no6OjqIRqO0trYSCoXuq1Cm3JDSKlpaWjh48CCPP/44zc3NpNNpNjY2RDi9Eoq1pMN3S0sLzz33HO3t7fT09Ig9Si6XUyqV2NzcJBwOc/HiRTY3NwkEAsLwuXLlChsbGywuLuLz+XjllVeoqanZ6aF9JorFIvPz8wSDQcLhMAaDAZfLJbxPNpsNq9VKR0cHDocDp9OJXq8X3nn4VTWsZBD/2Z/9WUXkft8JKS/Q7XbT1NSERqO5LxWD++WBGHLFYpFEIkEgEGB0dJR8Po/FYmF5eZmpqSlSqdQdf18mk2G1WvF4PCK0Go1Gb6pwrVRUKpUoVZbL5VitVqxWK3v27KFUKnHx4kVmZmaYnJwkk8kQi8UoFosVIc+gUqmwWCz09PTwpS99idraWrRaLZcvX2Z8fJw33niDQCDA+vr6bceSyWRESzaDwUBfXx/j4+Oo1WqKxWJZGHLwq/CxTCYjGAxy/PhxRkdHWVhYuOfQQKlUYmNjg0gkgsViwWazYbPZyOVyFWfISRv4rl276O/vx263o9PpSCQSrKyscPr0aRYXF1leXhaLv6QM73K5kMlkLC4uVkzSM1wbs91up7GxEbvdjtlsBiCZTDIxMcGpU6f4+7//e3EQkQpAvvnNb9LU1ITH46Gurg6Hw0Emk3loDDmVSoXH4+Hw4cN0dXXhdDqZn59nfX2dixcvMjIyUna5YJ9GSmqXogVHjhzB5/PR0dEh1mQpfCoZah999BGrq6vMzs6K/uHz8/PEYjFRufniiy9WvCEHiNSYW3Vw8Hq9eL1ennzySdra2igUCtTW1mK32296bUdHBzabjR/+8IcVnV4B1zy3Wq0Wq9WK0+kUIfZt+/sP4k3y+Tzr6+tks1kSiQTPPfccTU1NxGIxEonEHTdipVKJSqWivb2dPXv2ANcmykcffYTf76+Yhf1OSIntNpvtBs/a8vIyKysrIul9dHSU+vp6jh8/LrSopNyCcv0c6urqeOWVVzhw4AD19fVcunSJyclJ/vZv/5apqSkxL253/fl8nhMnThAKhWhra6Ompob6+nrsdjsajaZsKhpVKhWPPvoovb29qFQq4vE4IyMjrKyskMvl7vmhLRaLLC4uotfryefzNDU18e1vf5v33nuPv/7rv96iUWwNHR0dNDY28vLLL4vqvVAoxD/8wz9w9uxZxsfHhRdGSmxWq9U0NjZy7NgxFhcX+f73v8/c3BxTU1M7PJpfT01NDWazmZdffpmnn36arq4u1Go109PTjI2N8d//+39neXn5tvO+WCyK/Ln19XVisdgOjOLBo1arqauro7u7m2effRaTyQRcu+fRaFTsB+W6lsG1/Wjfvn3U1dWJ4pRHH31UpLdcvnyZmZkZMpkM8Xict956i0AgwNraGtlsllQqJcYntRwLBoM3iOA/zEj5j9FoFLvdzvPPP4/P56O5ufmmULR0uBkcHKRUKnH27NmyLuS7E11dXezZs4d9+/bh8/lYW1sT82Q77vsDMeSksmqpOGFxcZGNjQ1SqdSv9VRIQrFSMmmhUCAej7O6uvpQ5Mjp9XrMZjM2m00sbJII8NLSEtPT0+h0Oux2O/X19aRSKSwWCwaDAY1Gc0MZd7mhUCgwm83s2rWLuro6NBoN6+vrjI2NMTExcVc994rFImtra9hsNgqFgjjVKxSKsgony2Qy4TWTy+Ui8TWZTN7Xg1oqlYSIcDKZxGAwsHv3biEkWUneKaPRSF1dHU1NTXi9XlHIMTIywszMzA2pFZJHQypu6u3tFVI0leKVUqlUQlpn9+7dGI1GisUioVCIpaUlRkdHicfjt50X0nqZSqUeKi09lUqF3W7H5XKJoo5isSiEYDOZTNl412+FtPY0NTXh8/nYs2cPHo9H5MMWi0WWlpYYGRkR4s5nzpwhHA7f9l5LIbf7LYqqNKRDbSwWQ61WixShW7XpUiqVaDQanE4nHo+noj4jKQohOaIaGxvp7e0VhT0rKyv4/f5tU194oJ9cPp8nFotx4cIF5HL5XbWUkjS5XC4XDoeDiYkJ/H6/2CQrGYVCwfPPP093dzfHjh0TC8Lw8DBvvPEGly5dYmhoiMcff5y2tjYGBwdxuVwEg0Gam5vFpChHVCoVTqdT9NI1mUyEw2HOnDnD22+/LapS75VYLMb4+Dizs7MkEomy8chtBalUikAgwC9/+Ut8Ph99fX00NTXR1NRU1kKhn0av12MymVCpVJRKJWKxGAsLC7z99ts3GWeS9/mFF16gu7tbLPahUKhinneVSiUqFaUuLIlEgomJCZFKcqd5m81mRRHEwyRBYjabeeqpp9i9ezcymYxsNksmk+Hjjz/mk08+Kfs8KEnQV6q67ujoEKkUyWSSWCzG+++/z1tvvUUsFhPdaO50kFOpVPyTf/JP6O/vR6fTbeNodgabzYbD4eD3fu/3GBwcpLm5GYPBcMt9THLi2Gw2kWNaKeh0Otrb22loaKCvr48DBw6wb98+stks09PTfPe732VoaIj19fVtObw80E9O0ggLhUJMTU3dVcjAZDLR0NCAwWBAoVCwsbHB+vp6Rbaiuh6NRiMEgNvb2zGbzSiVStbX15mfn2d4eJjZ2VmWl5eZnJwknU5jtVrRaDQUi0Vh6ZerxphKpaKhoQGPx0NtbS1qtVqcukql0v/H3pnHyHne9/0z970zszsze9/3weWSS/GmJEsWRcliJSuO4SRGEtiqAwVFYwdoUSAN+lfRom2MtnBsIE2NNnGQuJId21IsS5ZFUuJ9k0ty73t3Znd27vue/kG8T5anSIm7O7M7H4CQwBnOPM+87/s8v+d3fH+iaEU6kTxosZO08ywWi/B0BQIBEYIr5nvg05C6mszNzWEymdBoNMJTFY1Gi8aQk1owqdVqMpkMk5OTTE5O4vP57gmV2Gw2UblZXl4uNvxAIFA0hhxwT+uibDbL8vLyAxfusrIy7Ha78LBHIhHi8bj4jEL2VH0aMplMGEHNzc3YbDaRF7a0tMTs7Cwulwu1Wk15eTkWi0W0bJP6cCYSiQ07tEnJ97W1tdTX14viFUnMVxKxnZ6eZn5+XmifPco1kzz5NputqAyVR0Uul6PT6YQ4cm1tLbW1tbS3twupoQfNW9ofAoHAfaW4ChmNRkN9fT2tra309fXR2NiI1Wplbm4Oj8fD4uLiunZweeKGXDqdZm5uTiQvfxpdXV28+OKL1NXVkc1muXLlCtevXy/4hNhPQ+pk8Nxzz7F9+3a0Wi1er5df/vKXnD17lvfee09ohl25coVr167xq1/9Cp1OR09PDxUVFQVtxJSVlfHKK6+wbds2amtrRcl9bW0tXV1d6HQ6kQMkhZHuh6T6PTAwgFqtJhwO43Q6N5Wn4mHEYjFOnz5NNpvltddeo7a2ll27dolWPsXAzp07+fKXv4zNZiMYDPK9732PkZERQqHQHQuZVNm6b98+mpubsVgsJBIJAoEAQ0NDRVHN+CBSqRTnz59nZGTkvuve9u3b6evrw2w2k8lkWF5eJhKJYLPZCIVCRVfgshqVSkVvby99fX08//zzGI1GotEox44d49133+XWrVsiylBZWcmRI0cwm81YLBYuXbrEe++9x/T0NC6Xa0PGLwm9fvWrX+XAgQPs2LGDsrIyZDKZ8BZ/+OGH/O3f/i0TExOPpaYgk8nEXAv1UP550Ol0dHZ20tnZybPPPkt7ezutra1YLBZ0Ot1D02MkFYPjx4/zySefFFWuqNVq5aWXXqKrq4v9+/eLg8nU1BQ3btxY957Ka3JEkCouH4Z0opVOcQqFglAoxOTkJBMTE0UhEPgwurq62LVrFzU1Neh0Onw+n5BnGRsbu+MiS1VO6XQauVyO0WhEr9cjk8nEa4lEgkQiUTCnlnQ6zfLyMtXV1aRSKeRyOSqViq6uLvL5PB6PR4TY4vG4qMS9G8lFXV1djUKhEBVjUieQQjFmc7kc8/PzlJWViSrEmpoa4YH+PJ8bj8dJJpPk83lUKhVmsxmdTneHl7OQnweVSnVHKxrpns3n8+J6Sgeb/v5+Ojs771jk8/l8UeUE3k0gEMDtduPxeO4RutVqtWKzGxgYEHqBTU1NJJNJOjo6mJubK1pDTurC09PTQ3d3txDGXd2tRhJHlfpG9/T0iMbyqVRK5JlKCvjrXRTQ0NBAU1MTXV1dQpxbuje9Xi8nT57k6tWrLCwsPFIyvuTha2hooLa2lsrKSsxm832NmkQiUZQ9Z+VyOQ6Hg6qqKp5//nkaGhro6uqiuroai8XySPqQUuWrlFZRTMUgkUiEy5cvk8vlaG1tFa3apDST9c7v3jBfr7Tx19XVMTAwIHS1zp49y9DQUNF75F544QX+8A//EKPRSC6X4+bNm1y9epUf//jHDw0hKZVKqqursdvtyGSyOxTiA4FAweSMxeNxLly4gFKpJBwOYzAYMBgMPP/88zz//PPCO3vz5k3C4bBo1XY3SqWSnTt3inCLUqkUFbuF1OkgnU5z8eJF0VvVYrEwODhILpdjcnLyM39uPp8nkUiQTCZJp9Oo1WpsNpsoeJEaTz/K4WijkJT8pdZEKpVKCCUrlUr0ej1PPfUUzz33HF/84hdpa2sDCrcS+3HI5/OiqfjCwgIrKyt3zMtsNlNdXc2zzz7L888/j8FgQKFQcOjQIerq6vB4PJw5c4axsbENnMVnR61WYzabOXz4MJ2dnZjNZnFPS+G2I0eO0NXVRX19/T3i5o2Njezdu5dEIsH09PS6tDO6m927d3P06FEOHjxIbW3tHa/NzMzwve99j4WFBWZnZx/p86TDy9NPP83g4CDd3d1UVVXdV0A4EAiwuLhYdN5otVpNX18f27Zt40//9E9FHpxkxN6vuOFuxsfHOXnyJE6ns+jm73a7+eEPf8ihQ4eoqqqip6dHpE/E4/F1z23fMENOp9NRVVWFw+HAYrEwNzeH0+l8pErXQkbKbZOUuxUKBalUiunpaWZmZh7oUZPL5TQ0NNDY2MiLL75Ic3Mzcrkct9vN1NQU8/PzeL3egjHk0uk0CwsLQvS5ra2Nrq4u8bpU1VNTU0MymaSysvK+nyOXy8UpXupi0dXVxdDQEDqdjkQiUTBGveQ19fl8aDQaXnjhBcxmM2q1GpfLhd/vFyHhRzVSMpkMTqeTubk5ZmZmUKlU7N27l3g8jsFg4MqVK7jd7jWe2efD7XYzPj4u7vlXXnmF/v5+amtrMRgMVFdX09PTQ2dnpyj4gdvhyKmpKRYXF4vWqJNyoDKZDPv27WNxcRGn04nRaKSuro7W1lZaW1vp7OxEq9XeEV7T6/W0t7cXheTK/ZDJZFRVVQmPVlVVlRDKldr1ZTIZ2tvbRWeDu5H02qRir/Pnz6/7/S5pgN3Pi2I0Gunp6SGdTj/UkJPL5cIpoVKpUKlUPPfcc3R1dYn+q/djcnKSTz755LEFxTcSo9FIRUUFr776Kh0dHRiNRrF+r57n3b9nLBbD4/EwPDzMmTNnGB8fZ25uruCLYB5EPp/H7XZz9uxZ0clEar1WX19PKBRiaWmpeORHPgt6vZ66ujrsdjtms5lIJCIs80L1PDwKkgdCr9ej1WpFWGxqaoq5ubkHzk0ul9Pa2kpPTw8vvPACFosFAL/fz40bNwruhk+n0ywuLqLT6bhw4QIqlYrOzk7xuvRQP8iAexAGg4GOjg7q6uqExlohGXJSzowkFCoZclevXmVqaop4PE42m73jMPIwI0VqRzc/P8/c3BwtLS3s2bOHTCaDyWRidnb2M1cArxfLy8tMTEzQ3t6OxWLh5ZdfFo2kKyoqRMhNkt+RSKfTTE1NFW0DdQmHw4FOp2P//v3Mz89z/fp1qqur2b17N/39/Wzbtk14mFd7KiRD7vLlyxs8g8+GTCajpqaGlpYWGhoaxLMuPfutra1UVVVhNpsf6KGQvPCtra3s2rWLkZGRdTfkFAqFKCy7u8WUyWSip6cHt9v9wHCZ5IFraGjgX/yLfyEMuX379tHQ0PDQ756enub06dNFpZ9mMBiorKzklVdeue/8VqdMSEiSO/Pz83zwwQf89//+39dtvGuJ1+vl3Llz1NbWsm3bNvR6PTabjbq6Ovx+Px6PZ132rw0x5GQyGR0dHfzLf/kv6e3tJZVKMTQ0xNmzZ4sq4fF+tLW1sWfPHhobG8nlcly7do2ZmRk+/PBD5ufnH+qRk3IOVp9qpqen+dnPfsbU1NR6TeGxcLvdvPPOO4yOjnLx4kXg9ly6urooLy+noqJCVC1JukFSPsHDqK6u5uDBg1y9erVgPBZSZeJf/MVfsGvXLr71rW9RW1srwjKRSITFxUVCoRDT09PiAV5aWmJ4ePihn93e3k5DQwNWqxWAzs5ObDYbly5dIh6PMzExUbDFH06nkxs3bgiJBZPJhMViYf/+/SgUCnQ6HcFgkKWlJex2O0ajEaVSKYxYj8dTVIacJDvh9XpZWVmhrKwMg8HA4cOHiUajHD58WCzoUr6QpFwvHebq6+s3ehqfC71ej9Fo5PXXX2fnzvUo6UUAACAASURBVJ33fZ61Wq0oJPg0tFqtqOxfbz755BNmZ2fZuXOnMMZsNht6vR6Hw8GRI0cYGBjgt37rt4Dbea1Xr14lFouJvEeFQoHdbqetrU2kGEjdDFKpFOl0mnA4jEqlusMrfeDAAfL5PG+//XbBrvEP4mHhUykdJBQK4ff7+fWvf43L5WJ4eLhg1vMnQSgUYmRkhAsXLlBWVibC8y+//DLNzc1MTU0RCATW3Dm1IYacXC7Hbrezd+9eTCYT6XRatDcptlj5amQyGXa7ne3bt1NRUSEW7pGRESYmJlhZWXmoDIfFYsFqtYqTIdy2+G/cuFGwfWcjkQgjIyMi4RtuzyUej4sm4VJ3CrVajcFgIJPJ3JEvIi18qxNkpcbDhbS45fN5QqEQH3/8Mblcjq985SuYzWbRR1Bq3RUOhxkaGhKVulNTUyQSiYcaK3V1dZSXl4scIpvNhtVqpbKyEovF8sAG3YWA3+9nYWGB5eVlMVZJHDadThOPx/H5fKysrKDVatFoNKKFTTgcXteehE+CdDotZDNWVlZE4n57e7so3FjdH3m1pprT6SSXy+FwOIrKeL0btVqN0Wikt7eXgYGB+3rcVgvhSpGJfD5/X8NOkmHZiMrO6elp5ubmCIVCNDU1sXv3brRaLVqtVkQIWltbxSFcakEZDAZ54YUXMJlMompxtSEqpWKEw2GSySSBQEDopkr3RmNjI6lUig8//HDd5/15uZ8hJ+XHpdNpUqkUHo8Hp9PJyZMnmZ+fZ3x8vKi8j59GMpkkmUwyPT3NzZs32bZtG/X19XR1daFQKDCZTMRisTW3a9bdkJPc7nq9nurqagBRsVbMC5tGo6G8vJy+vj6++MUvUllZKaRFLly4QCQSeaA3TgrH7t+/n507d6LRaMTDEIlE8Hg8BVOteje5XI5kMonL5bojz2NsbEwkvUuLlkqlEv01V4dhTCYTDoeD3/u93xNl64VKNpsVwsf/5t/8Gzo6Oti2bRt9fX00NDRgMpkwm82Ul5cLoz2RSPCVr3zlofe35MEpZIPtQUxMTIheqpLxKW1okiEnGWx/8id/wsGDB4VxX4xImmc///nPGRsb40//9E/p6ekB/jk3VArDw+1n5G/+5m+4ePEi4+PjmEwm/vzP/1xUMioUCmHYFssaaLFYqK6uJhaL4fV60ev1DzTCpNZcp06dIpfL8aUvfemeooehoSHeeeedDckHlX73kZERUZne29vLt771LdEbVSrmgdvr9cGDB8lms5jN5ju60EhFSblcjtHRUVwuFx9//DHRaFS0qmpoaBDvl8vlBdfF5klw8eJFxsbG+PGPfyy091KpVEEpLzxJbt68yfz8PBqNhoWFBXbv3s327ds5evQoIyMjfPTRR2vqlVtXQ04mk6HVaqmsrMThcAjR2Hw+T3l5OdXV1YyOjhZlwYNCocBgMGCxWLDb7cIYkVzLD5uP5NWpqanBZrMhl8tJpVL4fL5116P5LEj5D6tDf/fzIKpUKnQ6HW63m8XFReD27yZpCGYyGXGzS+GrQsmPW00mk8Hv9wvtM2nzTqVS2O124Z2Q/l6q7FOr1ahUqjvmCYjr6/f7RWiymIjFYsRiMcbHx9HpdLhcLmGQZjIZsXhns1nhpZSkdUKh0B39KYsBaaN2Op2iatVisaBSqYQnLh6PEwgEhBj20NAQV65cYW5ujvLycrxeL0qlUlT76fX6DRXFfVyMRiM2m+2RwqaSl1LqPbqaXC5HNpvF5/Phcrk2LCIjtcyTlPk1Go0IiUkKC9JcZTKZSIHIZDLiHpf+G4/HicfjDA8P43Q6xTpRVlZ2X33QYjPiJM3QUChEJBLBaDTe8x6fz8fi4iJjY2MsLi4W3X7+uEiHldnZWRwOh4g2VlZWsry8vObXeN0MOZlMhk6no6Ojgz//8z+ntbVVbHQKhYLf/d3f5fDhw3z729/mxo0bLC0tFZXlLhmoUl6MJB3yKLz00kvs3r2bjo4OEVp1uVy8/fbbXLhwYY1Hvn6k02kymQzRaJSFhQXgdoimp6eH2tpazGazOKkvLCzwwQcfFGw1VzweZ3Z2lsXFRU6dOoXBYECv17Njxw6RA6PRaKiqqhLVWm1tbbS2trK4uEgkEhGq8U6nE7idRLx//35ee+21jZzaZyYQCBAMBu/xquTzeVpbW0URi3RYiUajfPjhh8zOzhZlgdPi4iJut5u//Mu/pLa2lvLycmHAzszMcOrUKTKZDOl0WvRUzWazqNVqpqenUalUdHd3097eTn9/P1NTUxsmivu49Pb28vTTT9Pb20tdXd1DPclSrtxzzz0H3H4uJKLRKF6vV3TF2OiDWyaTYWRkhEgkwnvvvSdkgCSdudXkcjk8Hg8+n0+0Y5qcnGRoaIjh4WHS6bTQAJXkijYDwWAQmUzGu+++S29vL0eOHLnHwx4MBllZWcFsNpNMJnG73UX5jD8OuVxOXPeXX36Z8vJygHVJF1gXQ668vFyUmEv5Bna7Hbj9IMdiMbRaLXa7nebmZnETFJMhJxmlUr7Xo1jgUohRWswlEeBUKkUwGGR8fJzl5eV1GP36IZ3OpWsr5RVWVlaiUCiEd0/y8hSqh0LK+Vl9ClcqlZSVlYkqU7VazcrKColEAr/fLzYtj8cj8kSy2ax4v9FopLGxkXg8Lrw7xYS0UD/ouVWr1XeEqKQcuWLNmZESumdnZwmHw6LvKoDL5cLlcgkdwNVI947UYFzyShWTV1LyrqtUqk9NB5BelzzNUuFHIpFgfn6emzdvsrCwUDBhNylcfPXqVUwmE0aj8b69gKXiJ8k77/P5hN7c3V1ZHtQ83efzib7SxYKUNjA7O4tGo2FmZgar1SqKO+D23tbW1kYwGMTpdIqWdIVwfdcSKa9yvVNk1tyQk8vl7Nixg/b2dr71rW/dIXabz+eZm5tjfHycwcFBHA4Hhw8fpqamhuHh4Q0/na01Bw8e5LXXXmNwcJDa2lp0Oh25XI5AIMDMzAy/+c1vCkpyZC1QKpUMDAyI3MBUKsXKyooIKRfLKU4y6q5du3aHAba6O8VqsczV5PN5FAoFer2eyspKnE6nqO7czEiG3IPatxUD+Xyemzdv3tfo/rR7VwotS0UTxfw7PAwpB3B1GFZqQffhhx/ygx/8ALfbXRCKBVIf3Gg0yl/91V8Jg7WysvIeqQ2pG0UkEmF6evoz5XpfuXKFt99+u6gO7FI3mjNnzjA7O0symaS/v5+XX35ZvOe5557j0KFDPPfcc4yPjzM7O8vy8nJBXOO1pLm5WVQyrydrasjJ5XKhAL19+3axOcnlcjKZDLFYjJGREY4dO4bdbsdkMlFeXk5VVRV6vZ54PF6UxtyDPHIKhUIk+3d3d3Pw4EF6enooLy9HrVaLCqeTJ09y5coVQqFQwUpOPAnKy8ux2WzYbDbhzfB6vXzwwQcMDQ2JUFQx8XmKdhKJBAsLC5w9e5bBwUHa2tqoq6ujo6ODW7duFV0u2achSTbE43E8Hs9GD+cz81mveT6fFwngxZgX/FnI5XJEo1Hm5+f58MMPOX/+PF6vt+DUCla3jcvn83i93vvm90nt9R7kcYPbnmi9Xo/dbr+jYnX1a4VS5CRV1b7wwgvE43FGRkaE93g10rNrNBrFb7S6Slsq4LDZbGQyGV555RUmJiY4deoU8Xi84K73k8JkMmG1WlEqleJej8Via75ur6khp1Ao0Gg0HDx4kIMHD95RlSeVY587d44f/ehHbN++nYaGBhwOB7FYDJPJJJJPi5W7jTmVSoXdbmf37t1885vfpLm5mcbGRpEQHQwGWV5e5u2332ZychK/37+pNu67qampobm5maqqKpEbuLi4yP/+3/+bxcXFh7Yy22xI1b/j4+P84he/wGw209zcTHd3NwqFgl//+tcEg8FNdT+oVCoaGhpIpVJ4vd5NNbdHQQovxmIxIpFIwaYRPEmkwoabN2/ygx/8AK/X+7l6Fa810nO5srLymYS5pdzwsrIyGhsbqaysvGNfMBgMVFRUFEQVt1wux2Kx0N3dzX/9r/8Vl8vF//2//5doNPpAb/Hqjhh3d3aQBOHLy8t58803uXjxIhMTEwVpuD8JZDKZaNOlUqmIx+P4/f51kQ5bU0NucHCQvr4+WlpahCdOwuVy8dFHHzEyMkI8HmdoaAiTycRTTz1FW1sbv/d7v8fNmzf5x3/8R6E/VGxIY1YoFNTU1FBeXs4rr7wi8gSlvoTRaJRIJMLbb7/N2NgY165d29RGnKTRtG/fPgYHB7HZbMhkMhFi8vl8W8qIW43H4+HKlSt0dHSg1+sJhUJ3aHFtJjKZDPPz8ywvL2+6uT0KarVa6E6+/vrrnD17lpGRkY0e1iPh9XqZmZl55IiBpMHodrv50Y9+xNjY2KYOJ69Gr9djtVppbW2ltrb2DkOurq6OnTt3cvXq1YIqdMnn81RWVnL06FFRpHa/jg1S5bVUjZvL5e7ptSqXy7HZbPT29vLNb36TkydP8vOf/3xD5rXW1NbWinZ8wWCQsbExZmdni9cjJ5PJ6O3t5fnnn6empkbo8cDtG2FlZYWzZ8+KGPv4+DgGg4F9+/ZRVVXFiy++SFlZGb/85S9FcvCnfd9qNmJjkL5z9XdLuSEOh4Pm5maOHDlCZWUlNTU14oaPRqN4PB4RUnxYB4jNgNTdYdu2bezfvx+z2QzcliyR/myFBf5+BINBgsEgt27dEmFnyYu92YydTCaD2+0u2MrktULa5JRKJVarlba2Np555hkWFxeLxpDz+/24XC4hKXK/yjxpY5fClIFAgIWFBd59912cTuemz/8FhNC5yWQSFdurcTgcdHR03LE/FgL5fB6r1cr+/fvF30nX+EGpBFJ4VXp9dW5wWVkZzc3NfOlLXyIUCm06Q07yRtpsNhoaGoSEmFTZXpSGXHV1Nc3NzRw4cIADBw6IvqFwe/EOh8NMTk7ym9/8hkAgANwWkI3FYrz00ktoNBra2tpQKBS8+eab3LhxgzNnzoh8hNWo1WoqKysxGAzYbDZisRjRaBSXy7Wu3RCkxUqqQpNUvn/3d3+XI0eOIJPJhPq7lPQr9RH9p3/6Jy5evMjQ0BDLy8tFk+D/WWlubmbHjh1s27aNxsZGNBoN8Xic8fFxpqamNn1u4KMgqcJLfzb7PbGVkCqVfT4f+Xwes9lMW1sblZWVGI3Goqjum56eJhwO093djdPpZP/+/SLBO5VKEQ6HuX79OkNDQ0JOKpFIEI1GRaeTElsLjUYjZKY2G83NzbS2tlJdXU0+n2dmZoaJiQmi0ei6pEysiSEn5QPU1tZSXV19h0chmUyyvLyM0+m8o1m23+9HJpOxuLgo3NBVVVUMDg6SyWSYnZ0VJcyr0Wq1NDQ0UFZWRk1NDR6Ph+Xl5XXPu5ByKeLxuJBTUalUdHR0kMlkhDSJpKGUy+WIRCIEg0GGh4e5cuUKHo9n03uipPZs7e3tOBwO0aNRqtYNh8OkUqmC38jWmlwuJ6QrNqMRJ7Xw2WxexkdB6g4SCoXIZrNoNBoqKiowmUzodLqiuP+lqvLR0VHkcjnNzc1CXiQej4t17fz585w7d465uTlxrQt9bluZXC4nxOjlcjk6nU5UHX9epMr81TqCxYRUwLE6Oii1I6yurqarq0sYqW63m6WlpXU7hK+JIVdfX88XvvAF6uvrxQ0gJfOPjo7yn/7Tf2J6evqORTwYDBKNRvmP//E/0tLSwp/92Z8JOZIDBw7wjW98476aTHK5XOi2qNVq3n33Xd566611b74cCoW4efOm6Fqxb98+Ojo60Gq1d1T0wO1Cj1AoxNtvv83PfvYzxsfH8Xg8m/6UqtVqsVgsDAwM8NJLL1FTUyNeKzbNtLUmkUgQDoeFqORmIpvNcuXKFYaHh0kkEpvSUH0Y0WiU3/zmN0QiEV5++WWMRiPNzc00NTXR2NjI+Ph4wRd5SZWab731FhqNhu9///t3hN6y2ayo2IvFYluiIrfYyeVy+Hw+hoaGeOONN9i2bRtf+9rXqKuru0d6Zashk8mor69Ho9EwPT0tns+amhoGBwc5cuQIhw8fxmazkUql+NWvfsW1a9cIh8PrcnBZE2snkUjg8/kYHx8XHqZcLkcoFGJ8fJzR0dF7PGaSSOzc3ByZTIbr169TVVUlXJVSW6O7F31JykSKUUvCq+u9cEil6B6Ph+npafr6+sT4JINVqlLzeDxMTk4yMjIifovN7omDf84XNBgMWK3WO3SlZDIZFosFk8mEWq0Wm8FWReqxa7fbUSqVm8pzlc/nCQaDBAKBLXmNpcpNr9dLIBBArVZjsVjQaDRoNJoNaRz/uEjPp5TfWEiJ+iU+O5lMhkgkwvj4OAqFguHhYbE3SV7jx0XqGy4pVRRjTqxMJqOhoQGz2UwkEiGZTKLVamlpaaG/v5+2tjZqamoIh8NCB3ZhYWHd7JA1MeTOnj3L5cuXRThR4mE99yTC4TDj4+N85zvfQS6X3/Pv78dqb04qldrQ0ITT6eT06dM89dRT97yWSCQYHx/nk08+4fvf/z5erxe/379lPBJyuRytVotGo0GtVt9x3fR6Pfv27SOXy+FwOPD7/QSDwQ0c7cYyMjKC0+lEo9Fgt9s3lcGTz+fx+Xyie8tmMlIfBUlI1uVyMTw8TDabxeFwbPSwSpQQoVW3200kEmF+fp7GxkY6Ozv5nd/5HQ4ePPjYn5nJZFhcXGRqaop33nmHa9eurcHI1xalUsnXv/51uru7+bu/+zuSySTt7e10dXXxwgsvoFKpUCqVfPzxx1y9epUTJ06sa4/ZNTHkJJX7z4Jk7EUikSc8qvVBKrQ4deoUyWTyjk0qkUjgdDpFX76tFlaSTntzc3NcvXqVgYEB0WBeOt37/X6SyeSWD8VIYfbx8XHcbvemKv64W2x1KyJVq4+NjYm8moqKClpaWpiZmdkSFZ1bgXw+L/TEJiYmSCaTQrEAbucTBgKBgtMQlPLZvV6vaKn20UcfiQrMhzlV7pYfyWQyeDwelpaWuHXr1j3ty4qFfD6PSqVi586dwO2iTqkGwO124/F4uHbtGtevXycUCq3rHra+iWRbAJ/Ph8/nY3h4+IF5X1t1A4vH48TjcU6cOMHKygp/8id/QlVVFXDbcLl8+TLXr1/H6/UW3MK23iQSCRKJBB988AGw+aRHStzWDHz//fcxGAwcPnyYnp4e5HI5Q0NDLC4ubvTwSjwBJKktgPfff5++vj5ee+01kTvu8/mYnZ0tSN3MTCYj5JAmJyf55JNPPtfn3U+eq1iQuntEo1G+9rWviQrtZDJJOBzm3LlznDx5kl/96leMj4+vu4OmZMitEVvVWHsUotEobrebqakp0Z5L8k7Mz88L7akSxbnolXg00uk0fr8fv99PIBDA6/WysrJS8IUOJR4PSU8vnU6LFAmv14vT6eT8+fOcOnVKyHAVKlt9P8tmsxw/flwYtFInjkwmQyKRYGZmhunp6fu2clsPSoZciXUnEomwvLzM+Pg4er2epqYmYrEYt27dYmZmZkuFm0tsXTKZDH6/H6/Xi8/nY3l5maWlpU0VRi/xz4ZcNpsVXYpWVla4dOkSp06d4tixY1s+laTQyeVyIjpSiMgeZmXLZLJNY4Ln8/nH0rfYqnNfj3lL1XlNTU2Ul5djMBjIZrOMj48TDodxu91P5HsKbd7rRSHe62VlZZjNZnp7e7HZbMzMzBAIBBgdHX2iYfRiuuaSrmRDQwM9PT2srKwQCASYnZ0lHA4/1mcV4jVfL4rhmms0Gtrb27FYLDQ3NxMKhVhZWWFmZoalpaXPdHgtXfNHYyvMu2TIPYCtOvfSvIuf0r3+aGzVecPWnftWnTds3blvhXkXvmBRiRIlSpQoUaJEiftSMuRKlChRokSJEiWKlIeGVkuUKFGiRIkSJUoULiWPXIkSJUqUKFGiRJFSMuRKlChRokSJEiWKlJIhV6JEiRIlSpQoUaSUDLkSJUqUKFGiRIkipWTIlShRokSJEiVKFCklQ65EiRIlSpQoUaJIKRlyJUqUKFGiRIkSRUrJkCtRokSJEiVKlChSSoZciRIlSpQoUaJEkVIy5EqUKFGiRIkSJYoU5cNelMlkm6Z/Vz6flz3O+7fq3EvzLn5K9/qjsVXnDVt37lt13rB1574V5l3yyJUoUaJEiRIlShQpJUOuRIkSJUqUKHFfZDIZSqUSmeyxHIAl1pGSIVeiRIkSJUqUuC8KhQKNRoNCodjooZR4AA/NkStRYq2RyWTYbDasViv79u0jHo9z48YNvF4vy8vLGz28EiVKlNiSWK1Wuru7qa6upqGhAafTicfjYWxsjGAwSCQSIZfLbfQwS1DAhpxMJrvDlbv6//P5vPhTLEjjv3seq/+7FZHL5VRUVNDU1MTrr7+Oz+cjHo8jk8lKhlyJEiVKbBBms5nBwUH6+voYHBzkxo0bzMzMEI1GyefzxGKxkiFXIBSMIVdRUYFerwdAr9ezc+dOTCYTVquVyspKmpqaSCaTpFIpTp06xezsLGfOnCmYU4FWq8VmswkDMxAIEIvFMJvNmEwm9u7di8lkwmw2C2MuEAgQCASYnJzE4/GwsrJCOp3e4JmsH3a7nfLycv7wD/+QmpoaVlZWmJyc5OrVq4RCoY0eXokSJUpsWSorK/nSl75EdXU1dXV1VFRUMDg4SF1dHbOzs7z11lt4PB68Xu+WdkYUAhtmyCkUCpRKJQqFAoVCQWVlJRaLBQCTycSOHTsoLy+nsrKSxsZGent7SSQSxONx0uk0BoOBmzdvkkqlSCQSGzUNMReDwUBtbS0A2WyWXC5HJpPBYrFgs9no7++noqKC8vJy5HI5crmclZUV3G43uVwOlUoljNJsNruh81lrZDKZ8MTV1tayfft2bDYbx48fx+Vy4Xa7N6VBq9Vq0Wg04h4uLX4lihWVSoVSqUStViOXy8lms2QyGWKx2EP/nUwmQ6PRkM/nxTpZCAfxEvei0WjEYVtyQuTzedLpNDabjfPnzyOXy8V6lkqlNnrIWxbZwzaTtdBfkclkqFQqGhoa2LNnD83NzTQ1NdHZ2Sk8WgqFApPJhEKhQKVSoVKp0Ov15HI54e3yer38z//5PxkZGeHkyZOfuvGvleaOVqulra2NwcFB3nzzTdRqNQqFgr//+7/n9OnTfPGLX6SxsZH+/n7UajXRaBSj0UhVVRWZTIZMJoPX68Xn8/EXf/EXjI6OMjk5SSaTeZzhPpRC09yxWq1UVFTwzW9+k6eeegq73Y7X6+Xf//t/z8LCAvPz80/EyCm0eX/961/nd37nd/gf/+N/cOrUKeLx+JpsYiV9qUdjq84bPv/cn376afr7+3n55Zex2+2MjIxw69Ytvvvd75JMJh/478xmM6+++irpdJrx8XHcbjdzc3OfZyib4porlUrkcjkymYxcLvdIB9m1vuY2m40dO3Zw+PBhvvrVr2K1WtHr9cTjcZLJJNPT07hcLk6cOMGNGzd4//33H+fjPxeb4Zp/Fh4073X1yMnlcjQaDdXV1bS2ttLX10dLSwtNTU00NzdjtVrFCS0cDpPL5Ugmk8K4kzx4FRUVqNVqKisrWVlZ2dCyaI1GQ3t7O52dnXR2dqJQKMjn89TW1lJdXS1On4FAAJlMRiAQoKysDLlcjsFgwGg0Ul9fj91up7u7m1wuh9vtJhqNbtoTTkVFBe3t7bS1tdHU1ITT6WRhYQGn07mp3fTSNa6traW8vJzl5eVNe41LbG7KysqorKykq6uLuro61Go16XT6Uysb1Wo13d3dwO1D8PDw8Oc25IoRtVqNTqdDrVajVCopKytDpVIhl8tJpVKsrKyQSCSIRqMbNsZIJMLIyAjd3d0Eg0H0ej16vR6dTodWq6W1tRWz2YzP5yOXyzE5OUkgECAYDJLJZIp+HZdsDrvdjtFoBCCXyxGNRkmn08RiMdLp9EMPLuvFuhlyMpkMvV5PS0sLf/zHf0x7ezu7du1CLpejVCrJZDLC0g8Gg3z44YfEYjFSqRQtLS1s374du92O2WwGbhuFFosFi8WCXL5xKip2u51/9+/+HXV1dZSVleH3+1lZWaGyspKdO3fyzjvvsLCwgEKhIJvNkkgk0Gg0mEwmnnnmGQ4dOsTevXupra3lX/2rf8Xc3Bz/4T/8B2ZmZpicnCz6h2E1MpkMhULBCy+8wBtvvIHdbkcmk/HDH/6QoaEhXC7XpjZspGs5MDAAwLvvvsvKyspGDqlEic+NSqWip6eHUCj0qWtxWVkZX/3qV3E4HGSzWf76r/+as2fPrtNIC4e6ujq2b99OU1MTtbW17NixA5vNhlKpxO128w//8A+Mjo5y4sSJDdsDEokEi4uLTE5OcuPGDbRaLVarFbi9lptMJgwGA9XV1Tz77LO89tpr/OQnP+Ef//Ef8Xq9nxpmL3TMZjMOh4PvfOc7vPzyyySTSaLRKKdOnWJ+fp6zZ8/idDoZHR3d6KGunyGnUqno7++no6ODjo4OqqurUavVpFIpQqEQk5OTeL1ewuEwoVCIS5cukUgkyGQyLC8v4/F4OHDggDDkAJFrthHIZDLKysqw2WzYbDYMBgOxWIyZmRmuXLnC7OwsS0tLLC0tCS+TlF+gVCoJh8OMjIyg1WopKysjlUpRUVFBdXU1e/bswWKxiArOeDy+IXN80litVpqbm2lra6O6ulqElOfn51laWtoUp7j7oVAoUKvVaDQakVek1Wo39ABSDEgei9raWtRqNXq9nlAoxPz8PIlEYsNPwmq1GrVaTVNTE3q9nmAwKPJbpcNaLBa7Z0NLp9M4nU5UKhV2ux2r1Yrdbhf5sS6Xi3A4jNvtfqIpFmuJUqlEpVJ96vtkMhk6nU54OHQ63VoPbcPRarVotVqqqqrQ6/VYrVYaGhqEtIfNZqOmpoaysjIRYm1vby+Igq9cLofL5eLChQsEg0FGRkaER66/vx+dToderxf7244dOwiHw5w7d465ubmirmxVKpXiReOaGAAAIABJREFUelVWVhKLxTAYDLS3t2O1WlEqlXg8Htra2oQcSyAQIB6Pk0gkSKfTxOPxddnT1s2QM5lM/NEf/RFdXV1s376dXC5HLBYTxs4Pf/hDzp07h8vlIpFI3DN5mUzGd7/7Xfr6+tZryA9FoVCIkKrkFVxaWuK9997je9/7HpFI5IE5UKlUilgsxscff8yZM2dwOp0MDAzw9a9/nfr6er7zne9w6dIlXC4XCwsLzM7ObsAMnzwdHR288cYb7Ny5k8rKSo4dO8aFCxe4desWLpdro4e3Zmg0Gmw2G2azGa1WWxLWfAQkfUGHw8HXvvY1qqqqqK+vZ2hoiP/zf/4PLpcLp9O5oWOUCpnefPNNmpqauHDhgjh0VVVV0dfXx9TUFFNTU+LfyGQy/H4//+///T/MZjNHjhxh9+7dHDlyRBRzvf3229y8eZP333+fYDC4UdMr8QSQy+XY7XYqKyv58pe/TFNTE/v3779HwSAUColDvtls5tlnnyWTyfDjH/94ww+3V65c4erVq6JIraGhgYaGBr7//e/T0NCASqXCaDRiMBhEHvCf/dmf8d577zE9PV20jgiNRiPWbMnA1mg07N27F7lczle+8hVRtHPz5k1u3brFmTNnmJubY3FxkUgkwszMzLocxtbFkFOpVMLzZDQakcvlIsZ848YNzpw5w8jICD6fj2Qy+cCqzVAohNfrxWQyIZPJsFqtWCyWDcmRk8vl1NfX09TUhFKpJBgMcuXKFaampohEIiSTyU89iUgeqNHRURKJBNu3b6elpYWqqipqamr44he/yPXr14nH44TD4aJ9IORyOWq1GovFQn19vfCq+nw+nE7npg6nwu0TuZQWoNPptoQhJ21I0nP+ONWJUm7Kzp07aW9vZ2BgAIvFgkKhwOFw0NPTQzab3XBDzmg0YrPZqKuro6WlBaVSSTKZFBXJgUCAXC6H0WikvLxcnO7T6TRmsxm9Xk9vby+NjY2YTCa0Wi16vR6Hw4Hb7S7o+8Tv97O4uHiHV1StVlNXV4fb7cbn823g6DYWuVyOQqGgp6eHmpoatm3bht1up7+/X3huPR4PiUQCj8dDIBAQe59SqRQabbdu3dpwIw7+WbcVbisy+Hw+lEolx48fp7Gxkfb2dhGdkgo26urq6OzsJBgM4vP57uucKRakApTh4WGWl5eZnp5GrVbT0dGBwWCgrKxMOHb0er2QFQuHw9y6dQuPx8PExAThcJhAILAmY1wXQ06r1YoJGwwGZDIZmUyGaDTK+fPn+Zu/+RsikcinbuiRSISlpSUR0rDZbFRUVGxIiEqhUNDR0UFbWxsqlYpAIMCpU6cYHx9/5ATVfD5PJpPh+vXrTE5OMjAwQDwep7a2lsbGRr7yla9gtVqZm5tjfn6+qA05o9EohH8lmRmPx8Pc3NyGh8jWGp1OR01NDeXl5RgMBpTKgpFvXDNUKpUIR2SzWVKp1GMZclqtlkOHDrFv3z76+/tRKpUsLi5SVVXFrl278Pl8XL16dY1n8XDKysqoqqqiqalJpIxks1nC4TDT09OcPn0auJ1r09raisPhEIv9kSNHxAl/9UE0l8tRVVWFx+Mp6PtE0nxcvSZptVqam5sBtrQhp1Qq0Wg0HDx4kD179vCFL3yBiooK4LYBfOPGDRYWFpiYmGBoaIiJiQlWVlaIx+NC1kVK1SlE48fv9xOPx/n5z39Oc3Mzhw8fprGxUcwRoKmpiVAoxPT0tCgIKMS5PCq5XI7Lly9z7do1fvKTn6DT6Th69ChVVVW0t7fT0dFBb28vg4ODKJVKURRx7tw5RkdH+elPf8r8/HzxGnJyuZzOzk5aW1vFZiblSTgcDn7rt36L3t5erl+/jsvlIhgMEg6HGR4evm9+2N0dHzYKpVJJT08PfX19KJVKAoEAV65cYWFh4bE/K5vNEo/H+ad/+idu3bqF0WiksrJS5MspFAp+/vOf88knn5BOp4su58BoNPLUU0/R19eHw+EgHA7jdDoZGxtjenp60xtydxOLxQiHw5tSL1Aul2M2m2lpaeGP/uiPCIVCjIyMcPHixUc2vDo6Omhvb6e/v5+mpiZ8Ph9LS0v86Ec/IhQKiZDFRiN5k48dO4bb7aavr49oNMqxY8cYHx/n1KlT5PN5ESY2Go3Y7Xaqqqr4/d///TtCawDj4+PMzs7yi1/8guHhYcLh8AbO7smiVCrZu3cvvb29aLVaPB4P58+f58aNGxs9tCeGSqVCp9Nht9ux2+20tbVRW1uL0+lkamqK06dP43a7GRsbIxKJCG9VKBQiHo8LLT6ZTEYymSzo/Mh0Os3NmzdFPue2bdtQqVQ4HA4qKiro7++nuroamUzGxMQEb7/9dtEVP6jVasxmMyqVinw+z9jYGNevXycWixGPxzl+/Dg6nQ6z2Sw0Ym02GyaTia6uLqxWK62trVRVVVFdXc2JEyd46623CIVCT/y3WHNDTiaTUVtbS0tLi/BIwO2b3mw2s3v3bnbt2sXx48eZmJgQ/dyknCmpXZMUn5dct7Cxrbqk0Gp9fT0KhYJ4PM7MzMxnsrjz+TypVIpLly6xuLjIvn37yGazdHZ20t7ejtlsZmhoiIsXLwoRzWJCr9fT2dlJY2MjZWVluN1uZmZmWFxc3BJtuKR7VyaTkc/nhazAZjPkpKrk8vJympubOXr0KB6PB6vVyvLy8qcactJzXl9fz86dO2lqaqKiooLh4WEmJiZ45513yGQyGI1GvF7vOs3qwfj9fmKxGNeuXSOVSlFZWUkgEOD06dOMjo4Kjxz8c+GGwWCgpaWFr371q8IzDbdP/AsLC1y9epULFy4wNjZW0KLYUqj87rX3QQVoCoWCbdu2iQ3f7/dz+fLlTSU9olKpKCsro7q6mqamJhwOByaTCZfLhcvl4p133sHpdDI7O/upe1ahH26z2Szz8/OsrKwQiUTIZDJs27YNjUZDRUUFjY2N1NTUEA6HsVgs/OIXvyg6Q06lUmEwGITihNPpZGZmRhjZN2/evOO9KpWKuro6ysvLef7552ltbaW/v5+6ujqam5uJRCL85je/ETnyT5J1MeQcDgd1dXV3hAqkySwtLbG8vExdXR1dXV0kEgm8Xi8ajYbR0VE++OADHA4Hra2t9PT00NDQgEajIZlM4nQ6cblcG2LYSOXXZWVlT8xDmMvl8Hg8/OAHP6Cvrw+ZTEZNTQ1NTU0cPXqUmpoafvSjHzExMfFEvm+tkSRnampqePHFF6mrqyOVSnH9+nV++ctfbqpF/GFIv0FZWRkAw8PDnDx5clN5XGQyGS0tLdTW1vLHf/zHNDU1YTQamZqa4uLFi49UzFJTU0NHRwdHjx7l2WefRSaTMTIywne/+10hHivl6BSCkZPJZMjlcrz//vucPn2aY8eOkc1mmZqauqdIoaysDIvFwv79+2lraxMVivDPKQbHjx/n448/xuVyFXznj8XFRcLh8B1pJHq9nr6+PrLZLLdu3brj/UqlkhdffJHBwcFNWakqk8lob2/n9ddfp729nZaWFn784x/zwx/+UBj8Lper6EOMd5NKpXA6nVy9ehWdTserr75KS0uLqN7s6OhApVKhVqs3eqiPjc1mo6+vj2AwyLlz55idncXn8933AJ7JZMhmsywsLLC8vIzb7aa6uhqVSkVrayuDg4PU1taye/duzp49+8RTD9YlCcNisVBRUYFCoSCXy5FKpQiHwywvLzM3N8fCwgJGo1EkgxuNRlEYoVKpKC8vp62tDYfDIUqdV3dE2IgHQ+pQ8Sgl949KPp8nmUwyNjaGQqFgbGwMrVZLV1cXTU1N5HI5PvjgAxYWFopiQZCM3fLycurq6rBYLEQiEVwulwgvbAXUajUVFRVotVoAgsHgptKPk6Qn6urqhD6k3W4nlUrh8/lYXFx8JCkFi8VCZ2cnLS0t1NfXMzY2xtzcHNeuXWN2dnbdSvkfFalizel0IpfL8fv9AEIodDUmkwm73U5HRwetra2o1Wry+TzRaBS3283ExASTk5PMzs4WbG7UaqSK/NXzVCgUYg2/G7lcTk1NDXV1des5zHVBJpOhVCopLy9n27ZtNDU1UV9fj8fj4dKlS5s2jQJuOx/i8Ther5fx8XFcLpdQbEin05SXl1NeXi6klwo5XHw/ZDKZKN4JhUIP3HelyKDkaQuFQkSjUebm5kSXKqmjkxSVfJKsS47cnj17ePrpp9Hr9UQiEa5fv8758+f527/9WyKRCNFolIqKCvR6PYAw0uC2cOIzzzzDN7/5TbEIxONx/H4/x44dY2RkZMNO52uVq5fP55mfn+ev//qveeWVV2hpaRGNiy9dukRZWRknT54seENIq9XyzDPPsH37dhobG1lZWeHMmTOcPn2ay5cvb/pqVYmKigp2794tevFuNurq6qirq+Pb3/4227Zto66uDp/Pxy9+8QsuXLjARx999Eihov7+fr797W8LXcW/+qu/4tSpU0xNTRV81VsulxNr1uoIgRRuloS/n3/+eex2O3q9HpfLxUcffcTZs2d57733RH7wZt30NytqtZrq6mq6urrYt28fsViMxcVFgsHgI6kXbAb8fj/Xrl2jqqoKhULBvn37RPRMr9fT1taGQqF4pLByoXD9+nWR857P51leXn7ksUvpUk6nE4fDQT6fR6/XU11dLTQUnyTrElqVKlYl2RGPx4PT6RSJ7pKHTnK/yuVyobPT3NxMe3s7NTU16PV6stksMzMzzM3NsbS0RDAY3JAbI5/PEwwGCQaDmEymJ/75yWQSl8vFzMwMIyMj4jdoaWkhHA5z48YNMpkMiUTiiX/3k0KhUFBZWYndbhfizy6XS1Q9bRUkD/JmW9Alz3lXVxdtbW2icm12dpaFhQVu3Lgh5HikDi4SuVxOFEYYDAYcDgfd3d04HA6Wl5dZWVlhampKiP8Ww293twEmk8mwWCyi+KO1tRWbzYZWq2V6epqZmRlRtehyuUR4phiwWCyUl5ffETKThI4lz+RWQRKtdjgcGI1GUqmUKFpQKpUiN3Yzk81mRfh4dHRUpAbJZDI0Gg09PT3I5XLm5uaK5reIxWJi3X7U/rcS0gHOZDIJB5X0GWvxjK9LaHW16ncikRAhk0gkIi6qVAkCt8v1v/CFL9DV1cXRo0dFNUgmkyEWi/G//tf/4vTp04yNjW1YuCWdTnPp0iXS6TQHDhx44p8v9ds7ceIETqeTN998k1dffZUjR46wY8cOxsbGGB0dZXp6umAXf6VSSXt7O62trcjlcoLBIDdu3NgSBQ6r8fl8XLlyBavVSkdHx0YP54kgl8vp6uri0KFDvPTSS+zYsUN0XvjP//k/Mz4+zrVr14QnTtJIk4jH4+j1ep599ln6+/v5xje+IYQ3f/azn/GTn/yEiYkJQqFQ0Sz8q5G0xAYHBzl06BBHjx6lu7sblUrF0tIS/+W//BcmJia4fPkyyWSy4JPb72bv3r089dRT1NTUiL9bWVnh7//+77eciLHFYuHFF19kYGBARJ0kHcXy8nLhrNjMSGkGIyMjrKysMDAwQHNzMxqNBofDwb/9t/+WEydO8MknnxTNb5HJZD5zKFjqn3vgwAGx/0kdrNZCgmRdDLnVIUiNRkNDQwOLi4sYjUZxk0vGXm9vL7W1tezZs4eGhgZsNhs6nQ6ZTCZOsVIsPpVKbegin06n1/SmzOfzhMNh5ufnGR4epqGhgdraWmw2G3v37qWsrAyn01mQHgupLLuhoQG73U40GsXr9TI/P7+pkvwfhVgsxsLCgmjhVIyGiYRMJqOiooLe3l62b9/Onj17qK+vF0LfarWazs5OkRMmzVdq5QOIyl2tVsuuXbtoamrCarWSTCbx+/3IZDKhqC7lmxXbbyZVMDY1NbFjxw6RIyzlBUveRklnr9iQWs2tXtulHN9iy4P6vEit1iQ5GZ1Oh9VqZe/evRgMBs6cOcPKygpLS0tFea0fh2QyKQTxJaksKXq12nGz2ZFCyg6HA7PZTCqVwuv1isPpk2bdFSdNJhMHDhwgHo/z3nvv4fP5SKVSoh3GN77xDfr6+hgYGLinH+XJkyd59913uXz5ckF4dSTdn7W8OaXw7a9//Wt8Ph9vvPEG3d3d/MEf/AHXrl0TlXKFdKKXum7U1NQwMDCAw+FgYWGBmZkZrl+/XhA9BNeTYDDIrVu32L17d9HnzCgUCtra2vjX//pfCxFMiVwuh8Fg4Ld/+7fvCUNoNJo7Nv5kMolCoaCqqkqEXAOBAHNzc5SVldHf38/KygrJZJJQKFR0G6Ber6e2tpZdu3bxpS99Cbh9wr916xa3bt3i2rVr+P3+optXiXuROlqUl5cDtyNKZrOZN954g2g0yn/7b/+N69ev8+tf/3rTX+9UKkUkEhERNqm48ebNm0xOTm4JQ04mkwkB/MbGRqxWqxAJl7RgnzRrbshls1mOHTtGKBTimWeeEZ6a9vZ2XnnlFbxeL36/n+7ubmpqanjqqadETlU6ncbr9TI6OsqFCxc4f/48IyMjBZHkn8vl8Pl8+P1+8vk8RqOR9vZ2ZmZm1sRQkZIuDx06JH7DpqYmvvzlLzM8PMyJEycK6iHR6XSii0cikWB6epqFhQXC4XDRuNafFJJHzuPxEI1Gi9ZjYTQa2bdvHwMDA3R3d9+h5A7/nBditVrvMValtlsSer1evH/15zc0NKDVamlpaUGr1TI1NcWvfvWrognXSSG17u5uXn75ZWHoStdf6v5SiF70z0o2mxW9bwtpDVovPB4Pf/d3f0dVVRXNzc0YjUaMRiP9/f3Y7Xa+8IUv0NbWhlwuZ35+nitXrhTl76TT6dBoNCJC5vV67xGoN5vNOBwOHA4H5eXlKJVKstksFRUVojXjZkeKWthsNqHUEY/H7/BSPmnW3JDL5XIcP36cxcVF+vv7Rc/VtrY2XnrpJXw+H4FAgBdeeIG2trY7BH6l6p/jx4/z/e9/n1gsVjCep3w+j9/vF6dqg8FAW1sb0WiUycnJJ/59i4uLeL1ehoeHMZvN7N27l/r6el599VXMZjMff/xxQS0O9zPkpNJ0+PSK30Kay+clHo+zuLiIz+crekPu6aefpr+/n46OjnuEXyXD7GEL9v2uu3StDQYDBoOBuro6crkcWq2WsbExPvnkk6Iy5CT9qd/+7d8WXhqpJdOZM2eYnp4u+CrcxyGbzbK4uPhATc/VgtibEY/Hwz/8wz8IMWC73U5lZSV6vR6z2czTTz9NJBIhEolw9epVrl27VnSeudVFi9I9HY1GRXWmhNlspr6+HofDgdVqBRAi4Xd3MtmsyOVyrFar6D2byWSEIbdW131dDLnR0VHC4TDj4+Pkcjmam5spKytj27ZtpFIpUqkUDoeDXC5HMBgkFApx/vx55ufnOXPmDFNTUwW3AUpK7FarlWw2i8ViYffu3fj9fs6ePfvEv0+qUD19+jQ+n4+uri5sNhu9vb34fD56enpwu9243e4n/t2Pi0wmo7GxkY6ODtRqtVjcrVYrg4OD9Pb20tvbi9FovEcoMhKJMDs7y9DQECdOnCjKThZ3o9frqaiowG63YzKZnqj24HqgVCrp6+ujo6ODl156icrKSlGBnkwm8Xq9hMNhPB4PkUiEW7duEQgEmJ2dvUeK4/d///cZHBwUOXHHjx9nZmaG06dP3/HefD7PysoK4XC4qKogKyoqeP3119m+fTsOh0PoxS0vLzM5Ocn09DROp7PoNvKHkclkGB8fZ2Zm5o5r2NPTQ09PD3v27KG5uZmWlpYNHOXaE4vFcDqd+Hw+ZmdniUajNDQ08Ad/8AdUVlYyODgo0g8knbVCRqq67e/vZ//+/TQ2NmK324UI/tzcHHNzc/z0pz9FLpdjsVg4ePAg+/btu+NaS+HVQp/vk+LuLiahUIhz584xPT29Zt+5Ljlyfr8fhUKBx+PBZrORy+XQ6XR3CEfKZDJyuRyhUIjl5WUuX77M5OQkH3/8MdFotODCcVJo1ePxCO+B1FJoLcrNJQmLqakpstksgUAAi8VCVVUVtbW11NXVkUgkCsaQs1qtVFRUoFQqxbUzGo00NTWxe/dunn76aSwWC1qtVhjoMpmMQCAgtAHPnz9PIpEouGv/uEiJ73q9HrVafd8WRoWMXC6nsbGR9vZ22tvbMRgMpFIpotEooVCIhYUFUcji9/s5efIkS0tL3Lx5UxgskrfuwIED9PT0YDKZiMViXL9+naGhId56661NYdzodDr6+vpoaWnBaDQKyQGpX6wkLLqZkBqEJxIJTCYTGo0GgObmZpEj2N3dvcGjXHsymQzhcFgUc0l9hg8dOoRGo6Gurg6n04lOpyOTyRS8YSOXy9FoNDQ1NfH000+LvqFSpGVpaYmRkRHOnDmDSqWisrKSvr4+nnrqKeG1k5B6DksHm2Iv+noQ0jpXU1Mj2nfGYjGmpqbWVAR+TQ05pVKJUqkUlajAA3XPYrEYsViMn/5/9t4suM0rvdN/sO8bARAkCJLgTnHRRlnUYlle1Lacjp047u64e3qSTiczXTPTNTWTmkrNRSoXU1PJTZLKpLpSnfRW6U56MkkqXtW23Jbd3iRLsjaK4r4AJECCABeQBLEv/wv9v2NSu2xKBCQ8VS5JJkWdg+8757znXX7vv/0bw8PDHD9+nJWVFVZWVorSIyO1ocnn86yurgo9LanHntQIeTMpFAr4/X6Wlpb42c9+RkdHB1/72teorq7m2Wef5e233y7K9l3SBmc2mzl8+LDIrwoEAiwtLTE2NkYqlcJoNFJZWcnevXvJ5/NEo1E+/fRThoaGtnoKX4hMJiPEXu+le/1eoVKpOHz4MJ2dnSSTSWZnZzl79iz9/f2cPn2aaDRKPB4X+lmrq6vX6SVJFw6v14vT6cTn8zE+Ps6//Mu/MDs7W3KfyZ3i9/sZHx/n9ddf59SpU5vemqcY0Ov1vPDCCzzzzDP87u/+rtivzWYzJpNJhNgeNmKxGJlMhldeeYWRkRH+63/9r9TU1LBjxw58Ph8jIyNbPcRbYrFY6OjoYN++fTz++OOiYEm6iNbU1KBWq/nWt74lOjA1NDRsKGACRNeX7du385WvfAW/38/w8LDofvAgYTabsdvt9Pb20t3dTT6fZ2ZmhjfffJOZmZl79u/eE0NOKk23Wq2YTCZaW1vxeDyYzWZxWysUCkIUVIqb5/N5wuEwwWCQUChU1GK3hUJB5D3kcrkN4n8ajWZDEvdmIn0mo6Oj6HQ6crkcOp2O6urqeyJMvBlIrYiktmbRaJSZmRkmJiYIh8OMjIyI23wulxNtvaQ8k1Inn8+L6uZivJTcCfl8nmQyic/nIxwOc+XKFfr6+rh06dItPeZSflRlZSVtbW0iAToQCDA+Pk4wGCyp0OndsrKywszMDKFQqCi85fcC6fkCeL3e676eTqdJJpNoNJqizpGSDJT1edpfhFwuRzqdJhqNEo1GRahSpVKVhFderVbjdDpxOBzXedjgs5Zsra2tQjfN6XSKMx4Q8iOSyH9nZ6fY0+fn50XbK8lDJ10ES3WfNBqNVFRUiOKOeDzO8vKyaF5wr7gnhlxVVRVNTU08++yz9PT00NzcjMViQaVSoVQqUSgUJBIJVlZWMBqNGAwGEWa9VwbQvUahUKDVasVc7uU80uk0n376qVCC12g0VFVVFa0hl0gkGBoaIhgMMjk5iUajQaPRiGbSkgq6TqfjwIEDPP/882Sz2ZIMQ94ISV9NevdLjWQyyY9//GMREpL6g6bT6dv2/JVkhZ577jm+/e1vU1FRQTwe54c//CHnz59naWmpqHJfvyjSxVT6b25uTkiNPKz4fD6i0SgdHR33pD3RZmEwGFAqlaytrZHL5TbFSyyXy2loaKChoYFUKsX8/DyDg4Ml4Zk1Go20t7cLI/1GmEwment7gc8ubevJZDLE43HGx8fR6XT8wR/8AYVCgUwmQygUYmFhgb6+PhYWFkilUoTDYV555ZUb9isuBaTcOLvdTqFQYHR0lNHRUUKh0D0t1NxUQ06pVIo8qEceeYT29nbq6+tFW5qlpSWSyaQQB5yfn6e2thaPxyMWkdfrJR6Pc+HCBfF9xUw6nWZ8fJxsNktzczNms5m6ujrW1tbuqQW+Xg0+l8sRi8WKNpdMoVBgNpvFwlUoFCgUCtbW1sRild4dpVIpKn3Wf72UUalUmEwmdDqdEA+tqakhHA6XxPzy+TyRSASlUikKHG4n6qxWq8VNvr6+Ho/Hg1KpZHR0lEgkgt/vZ35+/oEKqapUKtRqtXjOuVyOtbU1oZX5IDE/P8/ExASXLl0iGo1iMpnI5/NCDFaqaASYmpoiHo9TV1dX1IZcU1MTdrud6elpYrEYoVDoC3mGpPZsDQ0NeDweVlZWxOW1FN6HVCpFKBQiGo2SzWaRy+XXGWpSHp3kxZSK8hYXF4XKxOrqKhcvXsTlcokCAKlln8lkIpPJEI1GyWQyRCIRpqamWFlZIRqNiiKJ1dVVkslkUe8Xcrkcl8uF1+tFpVKRSCQYHh7G5/Pdc73ZTTXkzGYz27Zt4/nnn+f3fu/3NiR3p1IpLl68SDAY5NSpUywsLBAMBvnSl77EkSNHRBXmV77yFaanpxkbG2NycpIrV64UdVLk0tIS//AP/8DevXtpbGyksbGR5557TmiH3Q9WV1cZGRlhfn7+vvx7d4vBYGDXrl0kEgl+9atf3fB5qtVq2tvbaWpqQq/Xk0wmmZiYKBnZiVthNBpF7ojFYqG3t5dsNssbb7xxTxNgNwupgvRusFgsHDlyhM7OTo4cOUIqlWJ8fJy/+7u/4+zZswQCgaJOnbhb5HK5kGaorKzEYDCQTCaZn58XFYwPEmfOnOHcuXNcvnwZl8vFzp07yWQy9PX1MTs7y9DQkFjnhUIBk8nEgQMHqKqq2uKR3xi5XM5LL73Evn37OH78OOPj47z22mtf6B3t6Oigra2N5557jsrKSs6ePcvIyEjJaGnOzc3xxhtv4HA4OHr0qNCRW8+eb/e2AAAgAElEQVS1YWipyvydd95hbGyMdDrN8vIy77//Prt37+bxxx/HaDSi1+uFoVtbWyuKIWKxGD09PczOzjIwMMDCwgKRSIS+vj5mZmaEt7TYkAzT7u5uHnvsMdRqNeFwmJ///OeiQPFesqmGnCTBIR3GkobYJ598QjAY5Pz58ywsLDA+Pk48HicajXLmzBlisRjf/OY3cTqdaLVarFYrPT096HQ6xsbG7lmj2c0gnU7j9/vxeDykUiksFotwRxuNRhKJxKaOXboBtba20tLSAlw1Ji9dusTs7Oym/TtfFCk0IS1QlUqFQqG4oREntXN69NFH8Xq9+Hw+JiYm8Pl8D0Q7Lymksr4A5tr2Rg8Kcrkci8WC2+1m+/bt1NXVodPpCAQCjI6OMj09LYREHyQkNXeLxYLD4cBgMIiWY8FgcNMLn7YaKfdJ0oaU9mipFd21YaRkMsmVK1fEhc1sNrNnz557mgB+NxQKBaHvVllZiV6vJx6Ps7S0xOzsLNFo9JbhUKkiU6/X43K5qK6u5tChQ7S0tKDVaolGo5w+fZrBwUHRiL3YkTzKgUCA8+fP09TUhMvl2rB3Sd8jRdqGhoYYGhqir6+PUChEJpMRX/P7/bz22mu43W68Xq+QlpJabUqpV1K+t8lkIhaLsbq6islkYmxsjIsXLxbd5V4ul1NXVydC6Ha7XeT6hcNhotHoPXdGbaohV1lZybPPPovX6xV9EmOxGD/72c84c+YMExMT191EpqamePvtt+nt7WX37t1oNBpsNhtf+tKXsFgsvPPOO0XdjzCVSjE8PExNTQ2JRIKKigpsNpvoH7nZRqhSqUSv17Nv3z46OjooFAqEQiE++OCDosq7kPrQrjfk1lcyScjlcux2O42Njbz44ovIZDL6+vq4ePEi/f39WzDyzWdtbY2pqSnC4TCxWOy6Ip8HCSm80NzczKOPPorJZEKtVuPz+URV9YOYLyYZsA6HQzSSj0ajzM3NFWUl+WZQKBQIBoMAt60sz2azfPTRR6ysrNDQ0CC8PBMTE/djqLelUChw4sQJ+vr6+JM/+ROqq6tpb29nenqaDz/8kNHR0Zvur5LkhMFgoKqqioMHD7J//34OHz5MfX09wWBQePhmZmaKRtT+duRyOeLxOKOjo/zyl79EJpNhMpk2dGlJp9NEIhHC4TCTk5O8+eabHD9+/IapPsPDw/yf//N/2LFjB4cPHxZh05WVFdLpNCaTierqar72ta+h0WjYvn27+Lsul4uLFy/i8/mKypCT8gK3b9/O0aNH6e7uprKyUkgxBYPB+xIp21RD7tqk+6mpKXw+Hz6fj7m5uVsmNfv9fvr7+/F6vajVarxerxC+DQaD+P3+zRzqpiFp34VCIQYHB3G73bjdbnp7e8nn87z++utMT09/YQ9ERUUFZrOZRx55BI/Hw9GjR3E4HKyurrKwsMDCwkLR3Prz+bzQgltcXKSiokKEVi9cuCBU4G02GxaLha9//eu0trZiNBqZmpril7/8JQMDA1s9jU1Dkh+RckaK9VKyGSgUCuGRqKqqIpFIMDY2Jm7qD5qGWpk7I5/PMz09jcViKdriltXVVQqFAu+++y719fXs2rWLzs5OmpqahJrC0tISsVhsQ76WRqMRLak8Hg8ulwuXy0UsFuPChQu89dZbjI+Pb+hsU0pMTU1x4sQJZmZmRD67RqPB5XKRSCTo7+8nGo0SDofx+XyigO1a0uk0oVCIbDYr8g9zuZxwdkjRuOnpaSGaXl1djcfjETJU8Xj8fk//lmi1WioqKti2bRsHDx7E4XCQzWaFPNP9Mto31ZBTqVTYbDb0er1QMx8ZGWFubu627sVQKMTk5CTV1dXodDpcLhcej4f6+nri8XhRG3Jra2ssLCwwOTmJXq+noaGBjo4OtFotZ8+eFaLBUjjiZkgemms9NZLAbk1NDYcOHWLbtm3s2bMHhUKBz+cTIqPF4q4vFApMT08LLTibzUZraytLS0t0dHSQzWaJRCLYbDbcbjdPP/20aM8mhduLNd/v85DNZkWblnud9LrVyOVyKioqcDgc2Gw2kskkwWCQYDBIIBAomne0zP1FihzY7faivchI3RbOnTvH0tISvb29VFdX4/V6SSQSxGIxpqammJ+fF1qJhUJBtGe02+3U1tYK4+TixYvCSzU5OcnCwkJJvv9SxyCfz4fJZEKr1aLVatm2bRvxeJzTp0+ztrZ22zQY6WK/uLh4Qw09pVKJVqtlbGxMePy6urrYuXMnAwMDTE5OFo2zQkKj0WC32/F6vXR3d1MoFIjH41y5coXLly/ftxSSTTXkEokEU1NT1NTUYDAYUCgUaDQaGhsbUalUImZ+LVJXBynsJKHT6airqysJ/aVAIMBPfvITnnvuOWpra6mqqsLpdPJHf/RHTE9P89ZbbzE3N3dd2yIJmUyGw+HAaDTi8XiETIWUd7F3716hiC+TyXjnnXeYmZnh+PHjwmgqJtbW1pibm+OVV16hq6uLF154gaamJn7nd36H6elpAoEAHR0dVFZWUlVVRTQa5c0332RwcJBgMFgy4YcyG1Gr1ezZs4f29nbRIFqn09Ha2soTTzxBf38/c3NzWz3MMveZXC7HxMQESqWyaD1ycHWco6OjIgrkcDhobW3F6/WKqIFUgV0oFFhcXKRQKAjP1FtvvSX2Nym8JlXtFtsefbesrKwQj8dRKBTI5XLC4bDoMrQZzzSXy5FIJDYUCS4sLAgFi0QiUVSGnFwup6amhqeffpqmpiYAIpEIkUiEjz76qHQ9cpLFbbVaKRQKqNVqDAYDdXV1Itn9Zoac0WjcoLEleS1KRXtrdXWVwcFBOjo6mJ2dxeVyYbFY2L17N3V1dUxPT2M2m2/qkZHJZFRXV2OxWGhubhauZY1Gg16vZ/v27bS3txOJRFhaWmJwcJDx8XFOnjxZdO5muOqFisfjDA0NiX5zarWatrY2nE6n2BgrKirw+/2Ew2EuX74s+uqW+qZ3IySBUKlg5UHNkZOEwFOplGixplKpsNvt1/XWfZDI5XKi9VIp7Fn3G0lSopjXdqFQYHl5meXlZWZnZ7FarUQiEaLRKBqNhurqaioqKkSbqmw2SzqdZmVlhWAwyKVLlxgbG2NiYoKFhYWi3Js/L5lMZsP5vdlpElLrrvXh51gsVlRFfBJS3rfNZhNC57lcTsinBAKB+3ph3VRDbmVlhYGBARQKBZWVlbS0tNDa2sqTTz4pXvibhZXMZjM6nU6EZSXL/NixYyVxg0+n08zPz/P6669z+fJlfu3Xfo2enh46OjpoaWnhv/23/0Y2m72lgKqkryPp8kiaPCsrK/j9fs6fP88HH3zA9PS0kDQoZhXsZDLJiRMnGBgYIB6P09XVxZe+9CU8Hg+NjY2EQiGmpqb4m7/5GyGcmEgkinY+X5SZmRnOnTuHxWJhx44dvPbaa1s9pE0nnU5z8eJFFhcXyWazjI2Ncfz4cfx+P9PT0w/UwbaeXC7HzMwMNpuNK1euCGmFMqXNysoKly9fZnR0lOPHjwthb0nvMh6PCwkOqUIznU6LdnVlHky0Wi2NjY088sgjvPDCC2SzWWZnZ/n+97/PiRMnmJqauq/j2fTQ6sTEBDqdDovFgtPpFL9KHqabGTGSdyIWi5FIJESlTyQSKYkEUcnwkqQVrly5gkqlQqPR4HQ6RXsyg8Fww78r5dpJt558Pk8qlWJlZUW0sZJ61M3OzrKwsFD0G0WhUGBlZQWFQsGVK1eAq8mrer0evV7P1NQUc3NzQjRxaWnpgTXiABYXFxkfH2fbtm1UVVXhdruFMPaDMm9JgiKbzaLVaoUBFw6Hi6rabLOR1rCUwP0g50F+EaR2fYlEAp1Oh9PpZOfOnQSDwaLUVMzn8yKkt7S0JKoUJWHcB01Gp8ztUSgUGI1Gmpubqa+vp6KiQoTSfT4ffr//vqcGyW6TfH9Xu5HUisjtdlNfX8/evXtpbW3l6aefFiX5tzLkstksp0+fZnJykp///OcEg0EGBwdvWyRwJxQKhbuKY93t3Ncjl8tRqVSoVCra29upqqpi//79WK1WHA7HdSE1qUvDmTNnxGEn9egLh8NMT08LXTbJyLubz+Nu5v5F5n2LnylC6yqVSrQvkg4+qQJssw+/rZ73tUhFAH/6p3/KE088wb/+678yMDDAj370o029rNzPd/1GrFdu/7zv7Odlq565QqEQDdH//M//XLQc/Mu//Ev+6q/+arP+mZuy1c/8TvF4PHzve9+jqamJzs5O0cLpf/7P/8nf/u3ffq6fWWzr/H5RKs/8XrBVz1wmk2GxWOjo6OB//+//TW1tLc3NzbzxxhscO3ZMCCHfK2427031yEmH8uLiInK5HK1Wy+LiIplMRvQeuxkymYxcLsfQ0BChUAi/38/S0lLRVjjdCsmbJpVbJxIJtFotRqMRs9m8wZCTXPKZTIaRkRGhAJ/NZonFYiwvL5e8ZEOhUBCaQsWUrHq/SSQSLC4uiiowSXfsQcunkrwUD1vBimS0ptNpNBoNWq0Ws9lMZWXlDUVyH0ZSqRSDg4PkcjmcTifpdJpYLPbAdb4o82CiUqloamqipaUFt9uN0WgkHo8zNzfH+Pj4lgnYb6ohJxGNRolGo4yPjwPwox/96K7+/rVtP0qVQqFAIBAgEAjcsS7agzDvMjcmkUiQTCbp6+vDYDDgdDqpra194Ay5hxFJdiAWi7G0tCRkCTweD93d3eVq3f+fWCzGG2+8QXt7OxaLhXA4zOjo6AMrmlzmwUKr1fL000/T3d2N1+sV/WGHh4f55JNPtsxRcU8MOYn1vfYedsqfQRm4+h6cP3+e+fl5DAaDyBcqU9oUCgWSyaTwolutVlQqFUaj8YGv1r0bMpkM09PTJJPJDRqcUoeIMmWKGYVCQVVVFXa7nUwmw+TkJCdOnBBSI1uV63xPDbkyZcpcz5kzZzhz5sxWD6PMJiJV2sdiMVZWVkgmk6hUKkwmE3a7/bpm4w8r2WyWqakppqamOHfu3FYPp0yZu0Iy5BwOB6lUipGREf7v//2/BAKB61qS3U/KhlyZMmXKbDJSn8pwOMzExERJVN6XKVPm1qyurvI3f/M3mEwmjEYjoVBIaJ9uJWVDrkyZMmU2kVQqRTweZ2Vlhfn5eUKhEMlkcquHVaZMmS9IKpXi/fff3+phXMemyo8UM+VS7TujPO/Sp/yu3xn3Yt5arZa6ujrR1WZubo5IJEIqlbqnuo/lZ35nPKzzhod37g/DvMuG3E14WOdennfpU37X74yHdd7w8M79YZ03PLxzfxjmfUtDrkyZMmXKlClTpkzxIt/qAZQpU6ZMmTJlypT5fJQNuTJlypQpU6ZMmRKlbMiVKVOmTJkyZcqUKGVDrkyZMmXKlClTpkQpG3JlypQpU6ZMmTIlStmQK1OmTJkyZcqUKVHKhlyZMmXKlClTpkyJUjbkypQpU6ZMmTJlSpSyIVemTJkyZcqUKVOiKG/1xYehtcXNeFjnXp536VN+1++Mh3Xe8PDO/WGdNzy8c38Y5l32yJUpU6ZMmTJlypQoZUOuTJkyZcqUKVOmRCkbcmXKlClTpkyZMiVK2ZArU6ZMmTJlypQpUcqGXJkyZcqUKVOmTIlyy6rVMmW2AplMtuH36/+cy+W2YkhlNgnpWV77XNeTz+cpFEq70EwmkyGX3/ye/CDMEdgwx0Kh8EDM6Yty7budz+e3cDRl7hXr9zLp161a12VDrkxRYbFY2LZtmzggnnzySXbv3g3A8vIyf/zHf0wwGNzKIZa5C+RyOUqlEo1Gg0ajYffu3dTW1tLa2orFYsFut6NQKICrRnomk+H111/n5MmThMNhEonEFs/g7lCpVNjtdnbs2MELL7yAwWDAYDCIr8tkMpaWlviXf/kXpqenuXLlSskaPwqFgm9/+9vU1dUxPT1NMBjk+PHjZLPZrR7aliAZ70eOHOGpp55CqVSSyWT4wQ9+wPT0NKlUaquHWGYTkMvlWCwWenp6cDqdeL1e3G43brebv/u7v+P06dOsrq7eV6dD2ZC7x2i1WpRKJSqVSvy/TCZDIpF4YG7lnxeVSoVCoUChUCCTySgUCjgcDtra2oQh99hjj3HkyBEAIpEIf/VXf0UkEiGdTm/l0DcFaf5arVbc4jOZDKlUimw2W9I3eZlMJgw4yZjR6/V0dnbS3t5OT08Pdrsdl8uFUqlEJpMJQ87n8zE8PMzS0lJRG3JyuRy5XC6enUKhQK/X43a76ejo4OjRo1gsFsxmM/CZpyYcDjMwMADAyMhIST5ruVyOWq1m9+7ddHV10d/fj16v55133nloDTm5XI5KpaKlpYWnnnoKtVpNKpXitddeY35+nnQ6/VDv96WOZKgbDAbsdjvbtm2jrq6Orq4umpubaWho4KOPPsLv9yOXy687o/L5PLlcjmw2u+lGXtmQu4coFAq+/vWv09nZSW9vr7iZf/jhh/z1X/81i4uLLC0tbfEot47e3l6amppobGxEp9NRKBRwuVwcOnRIGHZ2ux24ughUKhXPPPMMHo+HX/7ylyVtzCmVSg4cOEBjYyO//du/jcViQafTcfr0aV555RUGBwfx+XxbPczPhVKpxGKx0N3dTUtLC4cOHcLtduNyubBYLGg0GuRyOblcjrGxMWQyGVarFb1ej9lspra2lu3btxMOh1lZWdnq6dyU6upqHA6HGHt7eztVVVXs37+fyspKXC6X8DYC4hC3WCx85zvf4dy5c0xPTzM/P08oFNqqaXwuqqurqa6uprOzk46ODhYWFgiHwzcNlz8MmM1m6uvraWtro729HZlMRjKZZN++fej1ej744IOyV67I0Gg0Yo3m83mSyeRNv9dsNmO32/nWt75FQ0MDBw4cQKfTodPp0Gg0AHz3u9/lpZde4sKFC9ftXXNzc0xMTNDX18f4+PimzmPLDDmFQoFGo8Fms2G1Wq+7qeTzeTKZDMvLy6RSKeLxeEl5sKQbq9vtprGxke7uboxGIwDRaJTOzk7GxsaIx+NiTplMpmTm93lQKBSo1WrhhautraW5uZlt27YJI9dut1NfXy8MufUHg1KppLGxkUQiIb6/FI05yXvT0NBAR0cH27dvx2KxoNfrWVxcpKqqCr/fv9XDvGPUajUqlUqEUe12O3a7nc7OTpqbm2lra8Nut1NRUUEsFmNlZYVEIkEikWBiYgKNRoPH46GyshKr1YrT6aShoQGz2YxarS66daFSqdBoNNTW1lJfX09FRQUmk4n29nZcLhft7e3odDrhhZfGLr3TSqWSmpoaZmZmMBqNLC8vb+V0Phc6nQ6LxYLBYECr1ZJOp0kmk0X1nO43khda8rRLnmaXy3WdUV+mOFgfNcjlcoRCIeEdlzxwFosFo9GI0+nE6XTS2dmJ1+vF4/Fc90xrampwOp2kUimi0SjJZBK5XI5Op8PhcGAwGIhEIszOzpJKpTbNM7dlhpzNZqO5uZl/9+/+HS+99NKGRNlCoUAqlcLv9/Pyyy8zNDTE2bNnicVit7SYiwmj0YjZbMZkMqHX64HPNvTe3l5++MMf8pOf/ISf/vSnIpQmPdwHFYvFQmNjIyaTCbPZzOHDh+no6KCjo0MYZlK4SkIy5ORyOVqtlmeeeYaGhgYuXbpEIBBgYmJiS+byRVAqlRiNRr761a+ya9cuHA4HCoWCQqGAwWCgtraWK1eubPUw75i6ujrcbjcGgwGr1cpXvvIV3G43LS0t5HI51tbWCIVC9PX1ceLECcbGxlhbWyOZTOLz+aiqquLLX/4y+/fvp7a2ln379tHR0cH4+DjpdJqpqamiMtjdbjdtbW184xvf4NChQ5hMpg0XFJVKdUvPlEwmQ61Wo1arUSqVJXnAm81mKisrUavVpNNpTp8+zZUrVx7qYqR4PM709DR+v5/p6WlcLhcajYZt27aRz+c5duwYiUTioTZ2iw2TyYTVauXQoUNkMhn+3//7fyKdQ61Wo9frefHFFzl48CCNjY3YbDbq6+s3ePLWIxnxbW1trKyscPnyZUwmE7t27RIXOb1eT6FQYGBgYNMicvfVkJM8Mna7HY/Hw65du2hqasLhcFxnyGUyGfL5PLt27UKv15PP5xkaGmJqaqokFoLT6cTj8VBTU7PhoIarD9tqtbJt2zaeeOIJ1tbWiMfjnDp1isXFRdbW1kpijneKQqHAZDJRV1fHI488gtVqxWw209jYSGVlJXq9Ho1Gc8s5S+9HKpUikUiUlFF/LVVVVbjdbpxOJyaTSRiukgdSoVDcsuKxWFCr1eh0Ojo6Oujs7ESv12M0GsVGNzIywurqKuFwmLm5OZEbNjMzQyqVIpVKsbi4iFqtJhQKiVCEVqsFEO9FsYXr7HY7HR0deDwe7HY7Op0OpXLjVnqjCEOhUBD5oDKZDK1WS01NDbFY7H4Of1OQDjlpX0un02Qyma0e1paSy+VIJBIkk0kSiQS5XA65XI7b7SYajQoP3YO0t1+L5JE0Go1oNBqMRiNarRa73Y5SqdywTqR8sZGREZaWllhbW7vvuaLpdJpEIkEkErkuV1WtVmMymfB4PLS0tFBdXS3mk8/niUQiJBIJVlZWWFtbY21tDbi69qU/Dw0NYTAYiMfjVFdX4/V6hdd+amqq9Aw5mUyGRqPB4XDw+OOP09nZyZe+9CWqqqpuWLauUqmora3lpZdeIhgMcujQIX7wgx8wMzNDNpst+sXQ3d3NgQMHhJfhRofR0aNHeeaZZ5ibmyMSifCnf/qnDA4OigToBwWdTkdzczOHDh3iv/yX/yIMOfjMfX2jd2D9nwuFAtlslomJCYaHhxkeHhYLp9To7e2lp6eH6upq4Yks9vf5RlitVjweD7/927/N888/LzxRqVSK4eFh/uzP/ozZ2Vmmp6dZW1sTBT6wcb6JRILJyUnm5+cBREhSq9WKQpBioqWlhRdffJHGxkaRLnEzCoUC+XyedDpNLpfDYDCI+dhsNg4fPkyhUODixYv3Y+ibhtFoxG63byjietjJZDJkMhlWVlZYXV0lm82iUCjYuXOnMACK7V3ebAwGAxaLRaQZdHV1UVVVxaOPPorZbBb7PiCMnf/1v/4Xp0+fZmRk5L5fzqU89YWFBWBjqo7RaMTtdtPV1cXevXs3yIysrq5y4cIFpqenuXjxIqOjo4yNjQGfnVWZTIalpSVUKhUVFRU8++yzfOc736GhoQG73c6lS5eYnp7elHncc0NOJpOxfft2nE4ntbW1OBwOdu7cKfKBLBYLhUKB+fl5YrEY8XicdDpNNBrFaDTS1NSEWq3G6/VSV1dHTU0Nc3NzRV3NBp8ZKLfSy5K+x2QyUSgU6OnpQafT4fP5HihDzm6389WvfpVt27Zhs9mEB0My3q79fK69ta5fQAaDAaPRiNFoJJfLlZRXTqfTYTAY6OjooKenB5PJtOHr+XyexcVFhoaGSqIIpqqqir179+J2u0UBQyKR4K233mJoaIjx8XGWl5dZXV0lk8ncNOyWTCYJBAJEo1FkMhnLy8ssLy8zOzvL3NxcUa4FaW3fjKWlJXw+H/F4nHg8jlarRaPR0NnZKYw/tVpNZWXlde9BKeB0OmlpaRHRknA4zPz8/G0vJBqNhpaWFmEI+v1++vv779Oot4ZbnQGlhnSe63Q6hoaGRCqQ2+1m+/btOBwOHA4HHo8Hq9VKdXU1ZrMZm82GRqPZkDeq0+mQy+Xs3r0bhUJBIBDYkv1civRce+4olUp0Oh2Tk5N88MEH+Hw+YrEY2WyWeDyO3+8nGo0SCASYn5/fsGfncjny+bzw8i0vLzM4OMhrr73G7t27aWpqoqamhqmpKRYWFr5wSsI9N+QUCgWPPvoo3d3d9Pb2UlFRgdvtxu/3c/LkSRwOBwAzMzNi415dXWVsbAy32y1egubmZvFfLBYrekPubpBua4cOHcLpdHLs2DHi8fhWD2vTqKqq4j/9p/+E0WjcYLhdu8Hd6Pfrf1UoFJjNZqxWK1arlUwmU1KGnMlkorKykj179vDYY49d582QDsQLFy4QDoe3aJR3Tm1tLU8++SS1tbUiXyQej/OP//iPDA8PMzIyckehEqnoIRKJCJ01KdcoEAjc62ncNbc7lAuFAuFwmA8//JBwOEwkEhEpFl6vVxhyGo1GHHSlRm1trSjgSiQSBINBZmdnb/u8dTodBw8exO12097ezttvv/3AG3IPEgqFggMHDlBdXc38/Ly4fHV2dvIf/sN/oKamRoQgpUrO9axPn5KKpA4ePEh1dTW/+MUvWFxcvN9TArhhWoBCocBgMHDlyhWmpqY4fvw4gUBgQ4HinZDL5VhdXeXixYv4/X7q6ur4tV/7NRoaGggEAqysrBSnIedyuaitraWhoYHq6mqOHDkiqtLkcjmBQIDh4WFOnTrF6dOnhbbS6uoqiURCVKs6HA6Wl5fp6Oigt7eXQCBAKBQqiYKAcDjM2NgYgUAAtVot8uQeFqqrq4VoYmdn5wYPHHzmdSsUCiIfLJ/PC+PuRt+bSqU4efIkw8PDJZmPo9frcTqd6HQ6kSslEQ6H+fjjjzl16hShUKikLiqpVIqVlRVOnjzJ2NgYQ0NDzM3N3fFmZzQahahmoVAgGAzS39/P6urqPR753aFWq7Fardjtdsxm83X5e1LIpa+vj/7+ft59911x2D3xxBM4nc4N73cpemkkEefGxka8Xi9TU1PMzMyI/ft2z9xkMvHrv/7rGI1GEUovUxo0NzeLy1tDQwPNzc0iFOlyuUTRml6vR6VSUSgUiMViYk3E4/ENF+89e/bQ1taG2+0Grq6vYmJhYYG+vj5RgS7ZHp83DUatVmOxWJDL5SSTSZEnvBlpNffEkLPZbLS0tLBv3z7a2tro7OzEYrGIZNBQKITf72doaEgogksieetvdFarlWQySTKZpK6ujsXFRZFEWuysd7lWVFRQUVHxUBhycrkchUJBZWUlbrebJ554goaGhhvOXTLWpJxHyZC72fdmMg2PBT4AACAASURBVBnh6SnGcNvt0Ov1OBwOtFrtdfpiy8vLnD59mqGhIaLR6BaO8s6RJILW1taIRqOcPXuWy5cvEwwG7yqBX6/X09jYiNPppFAoiLBksRmzUpGS2WzGaDReV+CQy+VYWVlhYGCAvr4++vr6WF5eZmVlhd7e3g2FDhKlJgSs1+upqqqiqqoKp9PJwMCACKHfyfPSaDT09PQgl8v56KOP7sOI7z8PUihVQiaTUVNTQ1dXF93d3bS2toqOOxLrLyeSuHc0GiUSiXD27FmxFqTPR/LK2mw2MpnMdetpq1ldXd3Uy6RkyCkUCmHXbJZI9KZ+clLOV0NDA08//TTd3d3U1dWRy+UIh8McO3aMYDBIX18fkUiEiYmJDVbptRNKp9MsLi4yMzPD6OgoXV1dNDU18eMf/5ihoaHNHPqm4/P5mJ+fZ//+/Wi1Wmprax+KxOCdO3fS2trKV7/6VRobG3E4HCIn7kbh0kwmwxtvvEEkEgEQ8hsVFRVUVVVhMpnQ6XTi50ejUebn54VSeikgSVJIbZvq6uo2fD2bzbK4uMinn37K1NTUFo3y7hkcHORnP/uZuLH29fWJy9adoFAoRPX2v//3/56WlhYAkR9XbGFzvV5PQ0ODiC6sX89ra2v827/9G8PDwxw7doylpSUikYho2bVnzx6eeuopzGaz2OvW1tbw+Xwi0boUWH8Zyefz9Pf3c+HChaIzurcCjUaDxWLB6XTicDhEaLHUe9A6HA7cbjeHDh0S6VFwvT4iIAyUjz/+mPHxcT755BMikQg+n49MJkM2m6W6uprm5mZSqZSo/gwGgyUZZbkb9uzZw3e/+11yuRwnT55kZGSEQCCwKfPeVENOrVZjs9lwu914vV5xiAeDQebm5rh8+TJ+v1+4WW9n7UoCi3K5nHw+j8lkEsrwxc7a2hqpVIpwOEw4HCaVSt1Ue+ZBQBKDdblcNDc3s2PHDpqamm64geVyOVGosLq6yuDgoMiFslgsovrLZrMJDT7J+Eun06TT6U0VU7zXSLpxkrba+krHfD4vStil8FSpsLS0tCEPLhQK3ZXxJZfLMRqNoi2bzWYTYdqFhYWi29ilFlw6nQ61Wi0OMOkZ9vf3Mzg4yPDwsLhpS2KjlZWVVFdXi5ATXM3LKbVnrlQqMRgMwnsiXazuZi1mMpmSkNe5W5RKJXq9Hr1ev+EzisfjWyKt8UWRRO0dDofowFNfX49Wq71OKkzyzMdiMWKxGAMDAwwMDHDmzBkWFhZYWVkRf8dkMm2ISsRiMaLRaMns57difUEHIOwXpVKJ1+ulu7ub8+fPMzo6ysLCwqblwm+qIdfZ2cn/+B//Q6j1T01NMT4+Ljxok5OTIjZ8JzcUl8vFCy+8wK5du3jiiSf48Y9/zOuvv75pJbv3mnw+z8cff8zs7KyoUvF4PA+c2x2uVqbW1NTQ2dlJW1ubED2UNq/1EiOhUIi5uTl+9atfMTw8zIkTJ4RXwmq10t7ezuOPP05bW5t4T/L5vNDiKrUN0eVysXfvXnp6emhra9uw0BOJBKdPn+bs2bNMTk6WRP6nxOrqqtiICoXCXW/EarWa1tZW2tvb8Xq9LCwscOrUKU6dOsWZM2eKrj1XNpsV0hLxeByNRoNSqWRxcZHp6WnefPNNfD7fhnCJyWQS+aJSlZ7EwsIC77zzTkmJWut0OpHn+XlIp9MMDw+LlnwPElqtVnRxkPLBM5kM7777LleuXGFtba2kjBW73c7OnTs5evQoX/va10RRniQzVCgUSCQS+Hw++vv7efXVV0WqlOTIkFKh1j/r1tZWvvnNb9LW1gbA5OQkg4ODJbX33Yz9+/fT0NAAXDXqvF4vVquV+vp60W7yV7/6Ff/8z/8solCbwaYYcpK4ZUVFBS0tLSK3LRAIMDIywvj4OH6/n+Xl5Tt+kVUqFWazWSRYSjIlUvlvKSDJqmi1WpaXl7Hb7SWb5Hw7DAYDNTU1uN1u3G73DRNXs9msWPgDAwNcvnyZ8fFxIpGI8EpIniu73S4kLSQKhQLxeLyk1NFlMhl6vZ7q6mqsVut13uRMJsPExATT09Ml5WWEz4zrz4PU3aC+vl7Il6yurjIyMiIq14vts5B0oaLRKNFoFIPBgEqlYnx8nPHxcRYWFjZoG0o9ZOvr6zEYDBveZUlbbmVlpehCyLdCq9VitVpRq9XCG3M3KQ75fJ7V1VXhpZbEn6+VprjbysBiQPK8rO/WUSgUiEajLC0tFd37fDPkcjl6vZ7Kyko6OjqEEwI+y4PL5/OsrKywuLhIf38/fX19DA0NCeWJG6FUKjGZTDgcDqqqqkS0pRQ9cnK5XHjbKyoq0Ol0aLVadu3aRX19PXDVhqmrq8NsNuPxeEilUuLz2WxZpU0x5NRqNQ0NDbS0tNDa2srMzAwnT57klVde4ezZs/h8vrvqVqBQKLDZbDQ1NfHlL39ZiKZK/c5KJY+oUCgwNTXF2toa8/PzooPFg4jH4+Gxxx7jscceo7u7W4QV1h9eKysrjI2N8Q//8A/89Kc/FSHW9Qu4tbWVv/iLv0Cv12/Iq5PC6+Pj40xOTpaEV04mk4kcqe3bt+Nyua77ntXVVV555RXGx8dLYk6bhVKpxGaz8Ru/8Rs0NjaiUCgYGRnh7//+7/H7/UWZ/7i8vMylS5eor6/n0qVLIlfse9/7HhcuXNggnSAd6t3d3XzlK1/ZkBdZKBREovOt9PWKEZvNJvQg4Wpo9W50sHK5HIuLi5hMJtra2vj000+Bq17rqqoq4KrXbmhoqKQM3FsRj8eJxWIls/drNBqhdfkf/+N/FBJh68Op6XSaTz/9lIGBAb73ve8JY/VWe5jNZqO3t5ddu3ZRV1eHRqMRkktSwWOpoNFoqK+vF3nPLS0tNDc3i1CqxHot2TNnzvD6668zNDS06fvbphhySqVSNI7O5XLMz88zNDREMBgkEoncdWWGWq2msbGRurq6Dc2nTSYTTqezJHLkJPL5vHAtl8pCvhuMRqMIqXZ1deFwODYYYIVCgdXVVa5cucLc3JxQwL5ZcrRCoRA5SOu9l9Ln90W8QPcbtVqNy+Wirq6OlpYW7Hb7hvlkMhkSiYSQ3XlYkMvlNDU10dzcLLxxktc+FAoVdceOXC6H3+/nvffew2QyoVKpRL7LemPGarXS2NgowsbSZRSueqZDoRCzs7MsLi6WlGak1WqloaEBjUYjLqi3y5GT9MIsFotoa2az2TCZTHR3d/P1r3+dyspKLBaL8NBLKvmlhF6vp66uDovFAiA8V1NTUyUl8m4ymXjyySfZtm0bdrtdeM6y2SzZbFaoMXzwwQfCE30rNQnpQutyuTh8+DDt7e2o1WpSqRTJZJKpqamivbzdDL1eT1dXF52dnTQ3N1NVVYXRaNywv+fzeWKxmPBwarVacfnbbDbFkFOpVKK4YW1tjfHxcd577z0GBwdv6ma9FXq9Xnh21Gq1sHAdDgfNzc188sknmzHsMpuAw+Hg6NGjPP744xw9evQ6g7VQKDA7O8v3v/99AoEAY2NjLC8v3/bnSgthfR/SUjOEdTodnZ2d9PT0cOjQoetC6rFYTJTkl1K4+IuiVCo5fPiwqEJfXl7m3Xff5dNPP8Xn82318G7L+fPnOX/+/IZN+1pqa2v5rd/6LR599FF27Nix4WupVIorV65w+fLlkjrg4aq4d29vr2hrJBkpt8JkMmE0Guno6MDr9dLS0oLL5cLhcPAbv/EbPP/888DVS++Pf/xjzp8/Lzx1pYTdbueRRx7B7XZvkOE4f/48586dKxlDxeFw8N//+38XnjhpLqlUilgsJs72n/70p3fUgUYqampra+O73/2u6KsdDoeZnZ3l/PnznDlzpqQuszabjV//9V+ntbWVnTt3iv+/fi/IZDLMzMygVqtxu93YbDZ27tzJxx9/vOnj2RRDLp1OMzY2hlarZXh4GJ/PRygU+lw3Taky9UaNwxOJBNFotKQ2Prhakv3mm2/i9/uFknsptuVZj1KppLKykra2NpHgea04aiKR4N1332VoaIjLly+ztLTE8vLyHYdMpOd/rWxJqaDRaKirqxNCsOvJ5/PMzc0xMzPzUHnkNBoNJpOJrq4uOjo6mJ2dZWJigjfffJPR0dGtHt5dcSMDTqlUYjabqa2tpaurC5fLtcHgk+SU3nrrLYaHh0vGu2w0Gqmrq6O6uhq4WrF8bUslqarXaDRSUVFBbW0tHo+H2tparFYrTqcTq9WKy+USHsqlpSWCwSDDw8NMTk7y8ccfMz09XZLrQa1Wi3ypa3OhS/GStl6YXSaTMTo6yujoKO+//z7j4+N39Iy0Wi02m43nn3+enTt3blBtGBwc5PTp0+J5l8pakLhWL1ASM5c600SjUS5evIjVauXAgQMYDAZcLhd2ux2TyUQikdg0W2bTDLnJyUm0Wi2jo6P4/X7m5+fvOsdB6jtqNptvWJ6+trbG0tJS0ckS3A6p/2QwGOTZZ58V87xVD9ZiR6VS4Xa7aW5uZu/evVitVmDj4pfmPTAwQH9//x29tNLiKGX1ewmNRoPH4xG6S/DZhi4ZclKLllI8uD4POp0Oi8VCV1cXzc3NjI2N0d/fz/Hjxx+IqjUpzcTj8dDV1SXWBXxW/OTz+XjrrbcIh8Mlc3iZTCY6OjpEnmc0GmVmZkb0qISrc5cMtpaWFvbv309vby/Nzc0b8oPXr+nFxUUGBwd5+eWX+fDDDz/XuVEsaDQa0UsaPosilKIRd+3YC4UCIyMjfPTRR3z44Yd31DpPJpOh0+lwuVy8+OKL1NfXb1AvGBwc5O233yYYDJbk/rc+1Uc6ryQP9enTpwkGg3z44Ye43W7kcjldXV0cPHiQiooKzGaz0NXbDDbFkMvlciwvL6NSqdi/fz82mw2r1UooFGJpaYnZ2VkSiQQqlQqFQiFK9n0+HyqVSjSnVSqVon9mTU3NhtssQDAY5MKFC0UnS3A7JKNGqsSS2lKtN1iUSiWtra1YLBZ+//d/n8HBQX7xi18U7UZvMpk4cuQIbW1t191Kstksr732GiMjI1y6dIlQKHTbeRgMBvbt28fevXuva9FVKBQ4e/YsY2Njd9UxYKtRq9XU1NSIsnMJKdn95Zdf5sKFCyWVI/VFkMlkvPTSSxw8eFAkBkuemEwmU7Tv+p0gl8ux2+3U19fz7W9/m5aWFhwOx4bq7Uwmw6lTp+jv72d5ebmkDdfKykoKhQLf/OY3mZubw2KxYDQaaWpqEl45h8OB3W5ndHSUU6dOMTMzQy6Xw+PxUFVVxZ49e5ibm+P06dNMTEywsLBQMuHHO2FiYkLkk5VSFW48HufMmTNUVVVRUVHBpUuX+Oijj0QqwJ0KWCuVSvbt20dHRwfNzc3Y7XbkcjnhcBifz0dfXx/Dw8NFnRN7MyKRCD/72c8wm82iSKdQKOD3+1laWhIXktXVVQKBAK+++ipyuZynn35aOEASicSmXVo2xZCTDia5XE5NTY1o3TMzM8P8/Dw6nY61tTXRm1D6z+fzIZfLUalUQp8mHo+jUCioqKjAYrGIhFFJ/X5mZqYkb2xSouj6xXyt58lut6NWq+np6aFQKPDWW28V7eGm0Whob2+ntrb2OomQfD7PpUuXOHfuHIFA4LYVWzKZTPw8r9d7nReuUCgwNzcn1MFLAUnM2mazbUh0h6sXn1QqRX9/P5cvX36gDq9rkbq9KJVKVCoVPT09HDlyBIPBINrYzc3NCY3AUkTaw5xOJ42NjRw6dAin0ymSxNcXtoyNjTE6OloyrQYlpAuaJDci6cnt2bOHeDwuLu/Nzc1iz5b2gkAgwODgICMjI+RyOTo6OkilUvT09LC2tsbMzAxLS0sl6ZUBRFtBySkh7YcLCwtMT0+ztra2aa2Y7gfpdJrR0VFWV1dxu91cuHCBY8eOMTc3d0f5zXA1zC7pRG7btg2Hw4HBYBByLKOjo8I+KEVisRjnz58X+5rEjd7jbDbLlStXOHDggEgtqaio2NTespsqCLy8vMzly5dxOp0888wzQhcrmUySy+WQy+WEQiFeffVVYYWn02mhL6ZWq9m7dy/t7e309vYKC350dJSzZ8+KkvRiNW7uBKmf7Preouu9T9LXin3RGwwGDhw4QGVl5YaOFZLrvK+vj1OnTt1QEHI9MpkMm82Gx+PhwIEDNDU1CW+l9HWFQsGhQ4dobGzk7//+7z9XAc39RCaTYTKZcLlcdHd3b/DIyWQy/H4/09PTRCKRkpIluBvWG7J1dXXs3r2b3bt3c/DgQQwGA8ePH2dkZIR/+qd/uiv5imJCoVCgUCiora2lqqqK//yf/zNerxev1ys2d+kS+t577zEwMMDrr79OIBAoOW/cwsIC77//vqgqb2lpobq6mh07diCTyVhdXWV5eZlXX32VhYUFcYGLxWJcvnxZdP2Qwo5SHpkkZzI5ObnFM/z8mM1mHnnkEQ4dOsSePXuEAS8J4paaxEwkEuGv//qvUalUqNVqlpeX77rTSltbG/X19bz44ou0trai1WrJZrOsra1x9uxZvv/97xedELakh5vP52+7PqW+ytfmyd0oVCr1mJc+v3uRLrSphlwymcTv92MwGLBarRsOZElAMB6Pb4gNS4rw0ovucrlwu91YrVbhyVhZWWF8fJzFxcXrvFqlyrWVnaUyJ5lMhtFoxGq1YrPZRO9IyWCPx+OixdLtmr8bDAaRRyYlUldUVNzwRVepVCUjOyOXy7HZbCIX4loV/KWlJeFZLqUN/k6QyWSYzWbUarVoTdXW1kZ3dze7du2ioqKCQqHA5OQkQ0NDzMzMlGxoWalUioKWpqYmOjo6qK6uRqvVXtc8XFK/D4VCt10XEtIhIbUDk3oTb4VXWhJD9vv9XL58mXw+TzKZxGaziVSZ+fl5+vv7mZ+fx+/3s7q6SiwWw+fziTkbjUZisZjwWkj9SUtlbV+LJPjd1NQkijrgqoNiaWmJubm5kvO4ZzIZ/H7/5/q7kmeytraWtrY2amtrcTqdwNWQ7fz8vChwKqYUKbVajUajobq6mlwuJ95R6aJ9s1aTd8q9ls3aVENuamqKH/3oRzz//PNCzVyv1ws173/6p39ieHiYN99884abmVKpZPv27ezYsUMIygIEAgHee+89fD5fSYZVHyTUajXPPPMMXV1dQutNSuIOBAJ8+OGH9Pf339Fta9++fbS3t3PgwAGqq6vZvXu30Ni5Nkfu/fffZ2BgoCT6Uup0Oo4ePUpXVxdGo3GDYn0+n+fKlSucPn26qDayzUKv1/P888/j8Xjo7u7G7XbT0dGBWq1GrVaTTCZZWFjgxIkT9PX1lfR6ttlsVFZW8kd/9Efs2bMHk8mEQqHYcBFJJpMsLy/zq1/9il/84hd3FT5Uq9Xo9XoaGxtRqVTodDr8fv+WejKkvKZrw4jSQZVOp8XFXHrfb5XQrdPpqKys/Nwtv7YSKZm/pqaGr371q9TV1aFSqVhcXGRxcZFXXnmF9957b1NbMRU7UteGl156iUOHDlFVVSWiTuFwmJdffpmTJ08yNzdXNM4LhUJBc3Mz9fX1fOc73yGfz+Pz+fjkk0/45S9/KULjm8W9mPeme+RmZ2cZHBzk1KlTmEwmoRmztrbGpUuXmJqaIhqN3nRDk6pWFQqFCLlKjZlL9eb+ICGTyXA6naKXoEQ0GmVsbIyJiYnrDG6VSiU8rJIgok6nY+/evXi9XpqamqioqECr1QoxYellj8fjpFIpJicnGRkZKYnbrVwup7KyEqfTucErLYVaAoFAyfVVXY8UXjObzdhsNnGow1Uv665du6iurqapqQmHw7GhajeXy6HRaGhsbBQVj/F4/I69VMWA5HXwer20tbVRU1OzYY7w2WYt5QhWVFRQVVV1x2FVpVIpqj2lTikymYx0Or2lhlw6nf5CazCXyxGJRIhGoyI3+Frjt5RQKBSix6qU0y21XpO6HZSaXNbnwWAw4HQ6aWpqorGxEa/XK1JK1tbWmJ2dZXR0lP7+foLBYFGlR0leVavVSl1dnRDwXVtbIxKJMDU1RSQS2ZReuffKeN1UQy4ej4vqo/fff39DVwapp2QqlbrpQ5Q2PIfDgVwuJ5FI4Pf78fl8+Hy+kj34HiQUCoXIA1rfwWF8fJzXXnuNsbGx6zQETSYT+/fvp7Ozk6efflroS0nhomtz4tb/Knn6Tpw4wblz50rCI6dUKvF6vWJTkJBCqidPnuSdd94pqs3sbtDr9TQ0NLBr1y6R2C+p2avVarq6ujaEF6/9u1qtlj/8wz8kEAjwox/9iLGxMc6cOVMyn4der6eiooLf+q3f4tlnnxW9FW+EWq3GaDRy5MgR3G43P/jBD+5IukGn0/GNb3yDzs5OnnzySVHsJZfL74mg6P0ikUhw5swZkZJRykhFWiaTifr6eoxGI3A1FUjqanSnxQGlTn19Pc8995xQHjCbzWi1WlZXV4lEIvz85z9nYGCAV199tegK1qQ8bZfLhdfrxWQy0dnZSWtrK729vbz22mucOXOGoaGhoo2ibKohB59JbSwtLRGLxVAoFMLFnk6nb7hZy2QyGhsbRdN1Kb9udXWVixcvivYdpbLR34xkMilaz1RVVYlk6VLj2gRPgLq6Op566ikaGhoIhULMz8+TSqUoFArY7XYee+wxPB4PHo8Hm80mwrLSz4PPwqnrPXJS8ce1PVmLFa1Wi8lkwu12X+e1XFhYYHR0lGg0WlLvskwmw2AwCE2o6upqkdS9vLws8qekFm0ajYZsNsvc3Bzz8/PC+5jL5dixYwdutxuj0YjH4+Hxxx+nubkZm82G3+9nYGBgq6d7W+rr69m/fz/btm3D6XSKy+r691j6vUKhQKPR0NzcjF6vJxgMEgwGmZ2dven7XFFRIfrzer1etFotqVQKlUp1Q33NUkPKl7506RIqlYrm5uYNenulyI3EYR8GDAYD27Zto7u7m/3799PY2IjJZEKtVpPP54Wu7Pnz55mamiKTyRTtZ7PesQBgsVioq6vj4MGDuFwu5HI5gUCAUChUdF7WTTfk4GoI6U69Z9IHJ1W1NTU1ieTIxcVF0Q6k2Kz4z0MsFuPMmTMUCgW6u7s3VHuWEjcq1JB6rU5MTBAKhVhYWLjOkFMqlTes1L3Rz4XSqeBdj9FoxGaz0dDQgMfj2XDwzszMcO7cuTtqa1NMSNI4drud/fv3097ezm/+5m9y4cIF3n//ffGc9+7dS21tLXD1XR8YGODixYu8/PLLxGIx1tbW+MM//EMOHz4swoZVVVUsLi6ya9cujh07VhKGXFdXF7/3e79HQ0OD2KtudpArlUqR+9vZ2UlVVRWhUIiPPvropvmB3d3dVFdXc+DAAeHludG/UcpIe3tPTw9PPfWU+BxLmfU5vQ8LNpuNZ599lt27d/Pcc89teD9TqRTnzp3j8uXLvPfeeyWlAQpX+wpbLBbq6+tFT+xz586Josti4p4YcneD1WqloqKC/fv3c/DgQSwWC4VCgXQ6zfLyMhMTE4TD4a0e5qawsrLCxx9/jFar5ctf/rLItSkVKisrr9O/ubYtl5QDJ1X/wFUvlZQHc23brWt/L/15aGiIkydPMjQ0xOTkJP39/aysrBS1V04ul1NfX09zczM6ne46Iz0WiwkZhlJBo9Gg1+vZt28fXq+XJ598knQ6zbFjx9DpdOzfv1+ES202G8vLy3z00Uf4/X7effddZmdnxU08k8nw8ssvc/78eX7zN39TVLaZzWY6OztZXFxkenqakZERpqamiq5C3eFwsHPnTnp6evB4PBv0Aa+93NwISWdTqva+0bssk8moqKjAaDRuqORMJpMEAoEHJlSXSCSYmpqioaEBQIgph8PhktGTk8vlWCwWkVYgEQqFuHjxYknlfd4tUhpUY2MjTz31FDU1NcBn774kHvz2228zNjZW9GlR155H66NCUsOC5557jp6eHnp6epidneXs2bMsLy8TiURuKrGlUqkwGAyiwUEymWRxcZHZ2dlNPQe23JAzm82iwq2npwdA6Lisrq4yMzPzwGxe8Xicvv+PvfcKkvM67/SfzjlN94Se6ck5ARgARCJBSjApmKRESpS10tprs5xkeX2xti/sC5dLrrVvtspW1dbuastyKEmWlxJpWVSiJZIgKIBEIsIAmOnJOXWa6TSd0/8C/+9ohgRAgJzBdA/6qZoiiG70nNPf953znjf83hs3aG1tJZ1OiwrNjWxs+1FsSH0S5XL5B04k0oNgsVgwm8139RzcqXfqRk/d3Nwc//Ef/8Hg4CBTU1NbPJPtQSaT4XQ6aWxsFMYr/NJrGYvFRMi5VNBqtZjNZtEb9fjx44yOjvLv//7vHDx4kMOHD1NZWYnZbCYej7O2tsY777zDyMgIP/vZzz7gST9z5gwajYaamhpisRjNzc0YjUaam5tF/85EIiG6gRSL4S4ZWIcPH6arq4vKykqRNiK9fjtjbuPrUqEQ3ArP3m7hv50YdqFQIB6P4/F4ijZH535Jp9Ob5mOxWKirqyupdnWSXqTRaNzkeQ8EAoyNjZVEPu9HRaFQYLfbqaur48CBA0I7T9q7xsbGuHbtGpcuXWJ5eXmHR/vhbHxepdQI6ZmWy+Wo1WqOHTtGOp2mubmZqakpsV5FIpFNgv/SZ8nlcjQaDVarlY6ODux2O8lkUhRvbmXh3o4ZcjqdjoqKCp555hk+/elP09XVJV6Lx+O88847XLlyBa/XW1IejI9KoVDA4/GwvLzMW2+9xcTERNEZc0tLSwQCAfx+P93d3Rw7duwDuW4bb2RJ4R1+KRR8u/dKD00+nycej7O4uMjo6ChjY2MldaqVyWT09PRw4MCBTUb6+vo6Xq8Xt9vNtWvXSuZgIpPJOHnyJHv37uXpp5+muroapVJJc3Mz/+2//TdhyEhJwFLD8wsXLty1Wi+TyfD973+fd999l6WlJTo6Onj66adpbm7mi1/8oujHeuHChaIQf9br9ezbt4+BgQG+8IUvUFlZuanQkzR+MQAAIABJREFU53ZIRru0sVVXV2Myme77d8fjcUZHRxkcHOSll14qaeHcjaTTaQKBgAi39ff3k0gkWF1dLYnUA6nS8dOf/jR9fX0olUqhsuDxeJiamtq1hpzVahWyOx0dHaJjU6FQEL1Yz549Kwofi518Ps/MzAxyuZzr16/jcrloaWm57XulanKn00lLSwsej4eRkRGSySTJZJLLly+LYiaj0cgnP/lJ6uvrOXz4MMFgkB/+8IdcvXpV9CneKnbMkJMaDLe2tgoNJgmpwnVubo54PF40p/LtRPLYBINBZmdn8Xg8RRVWglubUywWIxKJoFKpbrtR3ynEdLu5ZLPZTcUM0mllZmaGxcVFVldXS8qIlzwuTqdzkw5iOp0Wgql+v7/oruudkMlkNDU10dfXR0NDAyaTiUwmI4y5QCCAx+Nhbm6OlZUVLl68yNzcnOideify+TzT09P4/X7q6urIZrMcPXpU9Ors6OjA5/MxOjpKMBjc8QRplUpFbW2tEP6VBHo3IqWDSALAwWCQhYUFcY8rFAry+bwoWJAMwY3e6Y3dXuDW87G+vs709DRjY2Ncv369ZLxVH0YulyMajRKLxUilUlRUVNDe3o7JZEIulxfdIfb9aDQazGYzHR0dNDc3I5fLN8mOrK2tlYRU0v0ik8mwWq2i4KmhoUFIhUkyUVI/4ZWVlZ0e7j2Rz+cJhUJ4vV6mpqbI5/MfkFUCRHqQ0WjEZDLhdDrx+/1YrVaSySSJRIJYLCbSpaxWKwcPHqS+vp49e/Zw6dIlbt68ycrKypZLqe2YIedwOHj00Ufp6OjAZrNtck1HIhH+7d/+jbm5uaJ/oLeKQqFAOBzG6/UyMjJCIBAoyg1foVAIrSCVSvWBMKnkebtd2PT9f15aWmJtbY1gMMja2hr/8R//gcfjYXR0lPX1dUKhUEkZ8TKZjNraWpo2tGiCWx65mZkZ1tbWivKa3g6lUolarRYdKiKRCOFwmMnJSVZXV5mcnGRyclI0gI/FYqyvr4uWe/dCLBbj9ddfZ2RkhFgsxtGjR/n0pz/NkSNH6OjoIJVKce3aNS5fvryjGpIKhQKr1So6VtyuQCmVSnHu3DlWVla4du2aOKlLIRepWrezs5PKykoGBgaEbqJGo0Gn04mCELlcTi6XE+K//+t//S98Ph/hcHjXrIdSMUxjYyNvv/02drudnp4e6urqmJ+fL+q2bXK5nMcee4zOzk4ee+wxqqurUSgUeL1e3n33XQYHB0WO525DoVDw/PPPs3//furq6kSe6MzMDKdOneL111/nnXfeKakDR6FQECobX/3qV6murmbv3r10dHSwd+9e4NbaLglXazQatFqtyBkfGBgQh7ADBw6Ig6dCocBsNpPNZlleXubChQv8y7/8y7YUfeyIISeXyzGbzTQ3N2Oz2TblEkkbuM/nIxgMlszGtxVIi34ikSjq05xUbTs/Py/CIBaL5QOiqHciGAyKxsk+n49AIMDq6iputxufz8fc3FzJXnep1ctGw1XyyJWSd1FCavwt5YFMTEywurrK1NQUs7OzTE1NfeT7NZ/PEw6HUavVuN1u7HY7c3NzonCira2NZDLJ6OiokC/ZCQqFAqlUSkggSbqH0uIt3c/Dw8MsLS3hdrvxer3MzMyQzWaFCHIwGCSdTuNwOERXAK1Wi9FopKKiAq/Xy+rqqvBIzc/Pi7680Wi0aA2bj4LU4svv9zMxMYHBYKC2tlZ4PhOJBIlEoiiNIZlMhsViweFwYDKZ0Ol05HI5wuEwU1NTW57/VCyYzWbMZrM4yEti/6FQiKWlJW7evMni4mJJhpQleaulpSVisRhKpZJ0Or3JUVFdXS0MOb1ez9raGkajUbSYk7y08Mu9XOpBPD4+zszMDIFAYFvG/8ANOUk1ua2tjRdeeGFTQ/FsNsv169cZGRm5a/eHMjtLPp/H5/Pxt3/7tyJP6uTJk/zn//yfP+AxuJ3EyOnTp/nxj3+8yZBLpVJkMhkRat1NhMNhhoaGSqpVj6T7+N3vfpcf/ehHwtMmeUlzuZwoRvi41ysQCPCzn/2M2dlZxsfHOXHiBPv37+ezn/0sTz75JNevXyebze7YwS6ZTHLz5k2sVivr6+vo9Xo0Go0Ip/zgBz9gZGSEV155hVAoJMKpG+/loaEh3G43Z86cEaFVtVqN1WqlsbGRnp4eLl26xMjIiPi90necTCZ33TMhMT09zfe//32cTidHjx7lhRdeoL+/n2984xsiJ7cY567T6TAYDMJ7Go1GmZiY4NVXXy2ZkOL9cujQIQ4cOMCTTz5JW1sbCoWCtbU1zpw5w9mzZ/nmN7+5K2TCwuEwV69e5fr16/zgBz8AfpkTqVQqRTcPl8tFb28vzzzzDO3t7SKvToquBYNBzpw5w8TEBD/96U+3zYiDHTDkDAYD/f39tLe3Y7VahZRFNpslmUzidrsZHh4mmUzumjDChyHF1m/evInb7S7qh6FQKAjvjNRiCRAikLer+oHNxQ3nz59nbGyMlZUVIpHIrvE2FAoFFhYWmJqawmQyCW/N0tLSpsbhpYDkbQqHw8TjcTKZDLlcjkQiseUbq5Rf5vf7GRkZQavV4vP5MJvNpFIpQqHQjj4TkhE5Pj7Oj3/8Y1QqFRqNRnjpLl26xMLCAuFw+I6HT8n43TgPKUFeYmlpaddUpd4rsViM5eVl3n33XVE4I1WzFqtXS3rO5XI5P/rRj9Dr9cTjca5fv74rW0lKwvUul4uuri4sFgtKpZJCoUA0GuXChQuMjIzsGsdLoVAQHrWNZDIZkScnCR5LOXOTk5M4nU6xz0WjUaLRKDdv3mR5eRmv17u934+0ud7uByhs9U9LS0vh61//euH06dOFbDZbyOVyhVwuVwiFQoXp6enC008/XairqysoFIot/b13m+eDmrv087nPfa6wsLBQCIfDhWw2W5idnS2cP3++cOTIkYLZbC7IZLIdm/tH/R0ymawgl8vv6Wer57eT8974o1AoCr/2a79W+Ku/+quCz+crrK2tFd58883CX/zFXxTMZnNBrVYX3by3+17finupWK75ne5xmUz2se/pj/vvd9M1v9/vYifmLV3znVjXHvQ11+l0BYfDUfi7v/u7wszMTCEWi4l9+9y5c9uyZxXjNb/X+2Cr14Z7mfcD9chJmlR1dXWbQqpwy50p5cVFIpFd7Y0bGxvjf//v/41Go0GlUgmv1OLiYsmGUTY8NA8t+XyekZER0dVCqVSyuLi4qUVVmQ+nWO+l7RxXMc73QVKs1/xOSGMtpTF/VNRqtRA+tlgsyGQy1tfXRWFHOp1+KL6H21Es98EDM+Tkcjk6nQ6z2Ux9fT02mw345RcQCoXweDysra2VZLLk/eB2u0uiFVGZ+6NQKDA8PMzw8DCnT5/e6eGUKVOmzMdGKj6SWlZJMis//OEPGR0dLcqClIeNB2bI5fN5YrEYY2NjfPWrX0Wr1aLT6cTroVCIaDSKx+N5UEMqU6ZMmTJlytyFdDotcr4ikQjT09MsLCxw5coVFhYWypGGIuCBhlbT6TRer5cf//jHD/LXlilTpkyZMmU+Atlslng8TiQSYXV1ldnZWSHYX0qV+LsZ2d1iu/9/kt6uoFAo3Lmfzm14WOdennfpU77X742Hdd7w8M79YZ03fPS5KxQKlEolLpcLm81GPB4nHo+ztLS0Y9Xk5Wu+mbIhdwce1rmX5136lO/1e+NhnTc8vHN/WOcND+/cH4Z539WQK1OmTJkyZcqUKVO8yD/8LWXKlClTpkyZMmWKkbIhV6ZMmTJlypQpU6KUDbkyZcqUKVOmTJkSpWzIlSlTpkyZMmXKlChlQ65MmTJlypQpU6ZEKRtyZcqUKVOmTJkyJUrZkCtTpkyZMmXKlClRyoZcmTJlypQpU6ZMiVI25MqUKVOmTJkyZUoU5d1efBhaW9yJh3Xu5XmXPuV7/d54WOcND+/cH9Z5w8M794dh3mWPXJkyZcqUKVOmTIlSNuTKlClTpkyZMmVKlLIhV6ZMmTJlypQpU6KUDbkyZcqUKVOmTJkS5a7FDmXKlNl62traqKqqwmQyoVQqUalU+P1+zp8/Tz6f3+nhlSnzkZDJZKhUKpxOJ319fajValQqFV6vl3A4zPj4OPF4fKeHWWYb0Gq1WK1Went76enpwev1EgqFuHjxIuFweKeHt+spG3JlyjwgZDIZMpmM1tZWOjs7qampQavVotVqGR0d5eLFi7vSkJPJZOK/0p/z+TyFwq4pJisDyOVyNBoNLpeLEydOYDAY0Ov1DA8Ps7S0xOLiYtmQ26VotVqqq6t57LHHeO655xgaGmJ+fh6321025B4AZUOuTJltRi6Xo1Kp2Lt3L3v37uVTn/oU7e3taDQastms8FhIRs5uQKFQoNfrsVqt1NbWUl9fT319Pd3d3ZjNZr75zW8yPT3N9PQ02Wx2p4dbZgswGo0cOnSIw4cP8+yzz6JUKlEqlezZs4dgMCi8clNTU+Vrvsvo6uriD/7gD+jq6qKpqYmZmZnyQe0B8kANOYVCgVwuR6FQoFAoUKvVwC9P7IVCgWw2Sy6XI51Ok8/nd6WHoszDhUKhQKPR0NjYyKFDhxgYGKC5uZlcLkc8HicUCqFQKHZ6mB8ZmUyGXC5Hp9OJeahUKiwWC5WVlbS3t9PR0UFHRwcDAwNYrVZOnTrF2toacnk5TXc3IIXW2traaG9vp62tTbzmdDpZX1+npqYGr9f7UF1zmUwmUig2UigUKBQKJJNJsdeVInK5HLVajdPp5PDhwzgcDiwWC0qlsmzIPUAemCGnUqno6OjAZrPR0tJCe3s7n/jEJ9BqtajVamHEXbp0iYWFBV5//XX8fj8LCwvlG6JMSWMwGGhoaODIkSM899xzmM1m8vk8CwsLzM/P87WvfY2ZmZmS9VI4HA4qKir48pe/TEdHBwBKpRKdTofBYMDhcJDL5cjn85w5c4aJiQmuXbuG1+sll8vt8OjLfFwMBgMvvPACHR0dPP/889hstk2vp1IpYrEYs7OzzM7Olux9fr8oFAqMRiP/9//+X3p6ejbtY8lkklgsxksvvcQbb7yBz+cjmUzu4GjvH7lcTlVVFU8//TRHjhyhqakJuVwuHDHJZLJkDdRS44EYclqtFoPBQEdHB5WVlbS1tdHZ2cnevXvRaDTCM5fJZEgkElitVhYWFjAYDAQCAdLpdEk9/DKZDKPRiFarxWKxfMDzKJFMJkkmk4TDYZLJJLlcrmy07kIMBgONjY04nU7sdjsymYxsNsvi4iKTk5NMTEzg8/lK9tqr1WoMBgNtbW309/cjk8nEwUzKi1tfXxehteHhYdbW1ojH4yU75zK30Ol0WCwWOjo66OrqwuVyodVqN70nmUyyvr5ONBp9aK65ZORUV1fT19dHf3+/iDBls1lCoRB+vx+j0YharS45L6VcLqe6upr6+nr6+vpoampCq9WK6+z1evH5fGQymZ0e6sdG2re1Wi0ymYxUKoVCocBkMiGXy0WRj0qlQq1WI5PJ8Hq9JJNJUqnUA7nft92Qk8lktLW10dLSwl/+5V9SW1uLXq9HqVSKSUsTVSgUPPLII+zbt4/HHnuMK1eu8Nd//dd4vV6Wlpa2e6hbglwuR6lUcvToUfbu3cvnP/95GhsbxQWXkMlkDA8PMzw8zKuvvsrw8DB+v39X3PhlNtPe3s5XvvIV2tvbxd+l02m++c1vcu3aNWZmZkin0zs4wo+HSqVCr9dTWVlJTU0N+XyeYDDI5cuXmZmZ4fLlyywsLLC4uCgOLdlstlzwUOLIZDK6u7tpbW3ls5/9LC6XC6PR+IED69LSEtPT0wSDQWKx2K6/5tKG/lu/9Vs89thjuFwuYcTF43FWVlYYGRnh/PnzuN1uMplMyX0nRqOR3/u936O7u5tf/dVfRafTATAyMsKlS5f413/9V65fv04qldrhkX58JGWBrq4udDodo6OjmEwmnn32WRFRbGxspKGhgbq6OlQqFf/9v/933G434+PjDyR0vq2GnGSpNjY20t7eTmVlJVqtlqWlJTExh8OBw+EQ/0alUol8AqfTSWdnJwDLy8slcbObzWYcDgc9PT3s27ePuro6KisrP3DikslkNDQ0kMvlOHjwIEajkffee49QKFTSi51CoUCpVOJyucTDrVQqMZvNqNVqNBqNeG8oFCKRSLC2tkYqlSKRSJDJZIjFYjs1/C1FJpMJb5XNZhOeinQ6LXLjIpEI2Wy2ZK83QCKRIBQK4fF4WFlZweFwkEwmGR8fZ3p6msnJSXw+n/CuF3M4taKiApvNhs1mE1WXG5/djQfPjeRyOTwezx0jB/l8nlwuRyAQIJVKCc9UqV53aW3v7++nv78fh8OBTqfb9F2trq6yurrKlStXGB8fJxqNlux874eKigqqq6tpaWmhqakJlUpFNpvF4/Hg9Xq5fPkys7OzuN1uVlZWiMfjRf1MvB+pOrmuro66ujqMRiNyuZxMJoPP52Nqaoq1tTUSicROD/VjITmbOjo6cLlctLa2otPpaGhowGg0sn//fiGxU11dTXV1NZWVlSiVSg4fPozFYiGVShEMBvH5fNs71u38cJVKhU6n4/jx4xw8eBCr1cra2hovv/yyCCU+9dRTnDhxYtO/k8lk6PV66uvr+cxnPoNWq+X69eslcYJvaGjg2LFjfPazn+XRRx+9o8u8UCgIK769vR2fz8ff/M3fMDIywvT0dMl65nQ6HUajkc997nM0NzcDt4zbvr4+bDYbTqdTvFda0M6dO4fX62V+fp61tTWmpqaK/jrfC0qlUhgGUggFIBKJEAgEiMfjJZcXczu8Xi9ra2ucPXuW9fV1Tp48STAY5Kc//SlLS0vMzc2VzPXs6+vj8OHDPPbYYzQ2NtLS0rIpVHg7Q07ytPz4xz8mEonc9nOlg8qpU6fweDzMzc0VvVF7N1QqFQaDgRdffJHHH3/8thXXw8PDnDlzhu9///sMDw+X7Fzvl66uLp544gkOHz5MV1cXcrmcSCTCO++8w+XLl/n6178uPNJAyTwbEhqNBpPJRFdXFy0tLSgUCnEAn5yc5OzZs/j9/p0e5sfGYDBgt9v53d/9XT7zmc9gt9tFPv9GKSXgA///R3/0R3g8HtRqNSMjI7z11lvb6pXbFkNOqkzt7u6mubmZnp4e6uvrSafT+P1+Ll++TDqdRi6XYzQayefzdHV1UVVVhUqlEl+IwWCgq6uLQCDAzMwMU1NTLC8vF6VBp9PpqK6upr+/n+PHj1NbW4tcLhdhpI25IdL3o9Vq0ev1mM1mZDIZzz33HAMDAwwODuL3+4WA5vr6+g7P7s6oVCqMRiNOp5OGhgaqq6upqKjg0UcfpaqqCrj14FdXV6PT6VAqleL61tXVodPp0Gq1RKNRgsEgoVCIqakpsegvLCywsrIiTq6lglqtxmQy0draSn19PXa7XXgol5aWmJmZIRAIsL6+visSgqXiDamQQ9roNRpN0T2rt0OhUKBSqWhoaGDPnj00NjZSVVWFRqPZVHF4O0OuUCig0+no7e29oxcil8uRyWSoqKhgbW2NiYkJVlZWuHTpEqlUqmQMeoPBgMVioaenh+bmZmpqapDL5cK7mM1mWV9fx+fzMTg4yIULF/D7/Q+FEWcymaitrWXv3r088sgjIh82FosRCAS4evUqo6OjZDKZkn3mZTIZNpuNqqoqrFYrJpMJAI/Hw8WLF7lx4wYej6dk7ue7oVQqxR5tMBhQqVR3VBfYaMQVCgXkcjkGg4H+/n6y2SynT5/e3rFux4dKcgsHDx7k+PHjHDhwgKqqKnw+H0tLS5w9e5ZMJoNerycajTI/P89v/MZvYDabhTQJ3How9u7dC9w60f70pz8VCZTFtjmYTCa6u7s5cuQITz/9tNi0Y7EYiUQCv98vwi5KpRKNRkNVVRV6vR6TyYTJZOLFF19kfX2d06dPMzIyQiKRwOv1FrUhJxlpjzzyCE8++SRtbW04nU6xCUqnF4mN162+vh6XyyWuMdzyVk1PT4uF7o033uDSpUusr6+XjCEnk8nQaDTYbDb6+vro7OykurpavD4zM8O1a9fweDx39OCUGvl8nunpaRQKBfl8Ho1Gg9Vq3faQwlahVCoxGo20trZy7Ngx8WzC5nv2TuuOWq1mYGDgQ3/PiRMnSCaT3Lx5k6tXr4rcsVLZ+CwWCy0tLTz33HM8+uij1NXVie8kn8+TSqXwer1cvXqVc+fO8dZbb5VsdOF+qaioYN++fRw9epQTJ06gUqkoFApEo1E8Hg9nz55lcXGxZI04icrKSurq6rDb7cKQm5ub49/+7d+4efNmyeSzfxhKpRK9Xi9+7rcgRavVcvjwYTKZjKjm3S62xZAzGo1UVlbS39/PkSNHMJvNJBIJzp49KxIgM5kMuVyOyclJAoEAPT09aLVaent7MRgMAESjUSYnJ0kkEvT19TE4OIjJZCIcDhfdw+BwOPiVX/kVenp6hOcJbp3Ew+Ew3/3ud/H7/UQiESHN0NraSktLCz09PVRWVpJIJMjn8/T09OByuWhubmZ4eJhz584xNzfHysqKCEnvNFJeQFdXF//pP/0n6urqaGhoIBKJMD4+zsWLF4nFYoyPj5NIJISnYqNhZ7Va0ev11NXVYbFYaGtrQ6FQUCgUhKfyU5/6FAMDA9TX1zM+Ps67775b1IYt3DJuH3vsMdrb23n++eepra0VrxUKBS5fvszPf/5zgsHgDo5y65HCh4lEgkgkwuzsLIFAYKeHdU9InqQzZ84Qj8dpaGhAr9czOjpKMpkUYfKOjo57Em6WEqTr6urE9ZdSRuRyOY2NjRgMBoxGI263mwsXLjAxMVH0m6DNZqO9vZ3m5mZcLtemnNdkMsns7CzvvfceL730kpDUKbZD91YjaSi2t7fzwgsv0N3dLSIP6XSal156SXQ6KPWDm0Kh4IknnmD//v0YjUZisRgTExNcvXqVK1eusLa2ttND3DIikQhzc3PMzc0xPz9PXV2dONxJGqDSXr7TbIshp9VqqaioEPlfhUKBUCjE0NAQo6OjZLNZ8SMlgE5NTeF0OmlpaUGv1yOTyUgkEszOzmIymaiursZsNqPRaERMvpgwmUz09vZSV1cncqHg1ik1kUhw4cIFZmZmWF5eFhe/p6dHPPSSwadSqdi3bx8ul4uenh5qamqE8ba+vi4Ek3capVJJTU0NPT09fOYzn0GtVqNUKrl69Spzc3O43W58Ph9nz54lHA7ftk1LbW0tNpuN3t5eampqgFvfo9FoFO7s3t5eVCoVkUgEi8XCtWvXir4YRKVS0dPTQ39/P0ePHhX3gyQ9MDk5ydWrV3d4lFtPNpslnU6TTqdJJBL4fL6Sac+Ty+XI5XKMjIzg9/tpa2tDp9Px9ttvE4lE0Gq1uFyuO+aDvR+1Wo1erycej6NSqcTfV1ZWotFoqKiowG63097ezvnz54Xx6/F4iuL5vhMmk0mkUGwsUisUCqTTaZaXlxkdHeXUqVM7OMoHi0KhwGAw4HK5OHLkCBaLRXimk8kkZ86c4dq1a7sixCyXy+np6WH//v0iJWZqakoUNu0mpEOp1+vF6/XicDjQarXk83kymQzr6+vodLrda8i9n1QqRTgc5s0337xte5ZCocAbb7yB2+3GYDDQ3t5OS0uLCNtdunSJn/zkJyKvqNiMOIBAIMCpU6d4/PHHN8lMVFRUAHDw4EEMBgNer5dsNkssFmNkZIT5+Xlu3ryJyWQSi357ezudnZ184QtfoL6+nt/7vd9j//79jIyM8P/+3/9jcnJyp6YpqKqq4s///M9paWnBYrEwOjrK5cuXOXXqFNevXycej5NOpwmFQnes5AsEAoTDYQKBAGq1mrfeeku09amvr6enp4cjR47Q39/PI488QltbG2+//TZyuRyv11vUxtztuHLlCpcvX2Zqamqnh7JtFAoFcrmceOZLJRwusba2xvr6Ol6vF4VCwdramhA4larx7gVJbshoNGI0GsXfNTY2Ul1dzVNPPSWay/f391NdXY3T6cTpdHLp0qWiSxZXqVTisHry5Enq6+vFa4VCgXg8zuLiIj/4wQ9wu907ONIHj8Vi4Vd+5Vc4cOAADodDRGPeffddRkdHGR8fx+fzlbwRB7eu9draGj6fj+bmZsLhMKdOnWJ4eHinh7ZtDA0NoVarSSaTWK1Wzp07Ry6Xo6amhpaWFg4fPrzTQ3wwhlw+nyedTuPxeO64QC0tLbG+vs7U1BQ6nY7GxkYUCgUWi4VMJsP8/DyhUIh0Ol2UG3g8Hmdqaoru7m6y2SwKhULIT0iimSaTSSRLSx64cDhMKBRCpVKRSCRQKpUEAgGSySRPPfUUVVVVNDU1iaR4aVPYKaSWM9XV1ezdu5eKigpisRhLS0vcvHmT4eFhRkdH7+mzJO+NJDeysLAgXvP5fCQSCerq6kR1r8lkwmKxCGHGYrwPpNZzOp0OjUYjxlkoFPD5fLjd7pIPr9wJqXJLktq4U1VmsV47+OU9+f7wvVSs8HHC+nK5HI/HQ01NDXV1daTTadrb2zEajVitVtrb21lZWRGaksWElNPrdDqpr68X6S9wa30PhUL4fD4mJibweDz3/fnSvWO1WlGpVKRSKbLZrEg3Kdb7RSr2kiQqpP7J8XicmZkZhoaGCAaDu0JPTULq0gK3npeVlRVCodBt3ysZtaUk6P9+/H4/k5OTuFwurFYro6OjQhD4Ttc1k8mQTCYJhUIPJBXogfZavRuSB+fv//7v6erqEkbCxryD73znO7z++uvEYrGiuzECgQBnzpyhrq6O48ePU1lZidlsplAokMlkmJubu2OiayKRIJlMivdOT0+j0Wi4fPkyfX19QkTZZrN9oGffg0ar1fL7v//79PX14XQ6mZmZ4dvf/jY3btzg0qVLW+aBWV5eZnV1lebmZurr66mqqir6fqQymUwUcHzyk5+kqakJhUJBMpkkEomwvLzM7OzsrtHJez9yICYdAAAgAElEQVRSXtjdQo9SFajUvqdYN+jtQKrsXVlZYWpqij179ohDa0dHB93d3ajVat59911mZ2d3eribaGpq4o/+6I/o6+ujqqpqU+J3PB7ne9/7HsPDw1y5cuUj6YdJskV/+Zd/SX9/PxcvXmR2dpZXX32VaDRalHmxKpWKzs5O+vv7+Y3f+A0RfZmenmZsbIxXXnmFixcv7qqDm0wmo6Wlha6uLnFQvROSB1omk5V0a7aRkREhq2I0Gjl+/DidnZ08++yzothjI/l8nomJCebm5vjnf/5n5ubmtn3uD8QqkCQ07rZoS6XrXq8XrVbLlStXcLlc1NfXYzQaaW9vx+FwoFari1JoMJvNEo1GWVpaYmRkRJSYB4NBEWMPBoO3/Q7eLwwqFYJs7AYhlfTvdAcAuVyO0+kUCtapVIq5uTmWl5fveCq7H5RKJSaTCYPBgNVqpaqqCqPRKLyx6+vrO/4d3AmZTEZNTQ3Nzc04HA4hKyOXy5HL5VRUVNDa2ko4HBZJ4rlcjlAoJE5wpYpMJsNisWC32zcZ3FIF+8bkf5PJRDAYJB6Ps7q6Sjqd3lUei7uRyWTIZDKkUikhhC0t8pI+104f1jYiyShUV1fT3t5OTU3NpvFt1A+bnp4mHo9/6KYled8UCgU6nQ6TyYTdbhfFJB0dHUSjUfR6PTdv3sTj8RSltqRMJkOr1WIymaioqBDREkkgW5JTKrbCvI+KVqvFaDSK9Vnas5PJ5KZ0J7lcLrQz+/v7UavVGI3GTXtYKa11qVRKtNpSqVTU1tZSX1+P1WrdVOwDCG3QsbExpqammJmZeSDe9W1fMfL5PLOzs6JVxd2QiiLW19f5zd/8Tfbv3y/aerW3t1NVVYXBYCjK05mU7Hvx4kV8Ph/PPPMMe/bs4Y033mB6elpUW97LQy1JIbS3twvZisHBQX7yk5/g9Xq3eyp3ReqxV1NTIwpS5ufntyyp3WazcfjwYQYGBnjyySdpamqiqqqKH/3oR6JYplj7kioUCk6ePMnBgwdFVaIkReJwOHjuued45plnOHXqlEgMjkQivPHGG6LgpxjndS/I5XL27dvH3r17NyX/Go1G6uvrRQeXL33pS/T29jI0NMTCwgIvv/yyEMgt1bl/FCTdOqknczEil8vR6/X09/dz4MABjh07tqlwAyAYDLK4uMibb755z54HqS+lwWCgu7ubY8eO0d3dTVNTE93d3dhsNj71qU+xtraGw+Hg/PnzfP3rXy+6XtRSFbKURiEZuKlUilAotOuaxkuOlerqakwmk+gTvrS0JHJHpQPbJz/5Sfr7+/nCF76AzWZjbm6OkZERvvnNb7K4uFhyecIymYz29nZaW1v5rd/6LaGB+n4WFxdZWFjgf/7P/8nExASrq6sP5L7dFkPObrfT29uLzWYDbim/Ly8v39NDLln54XCY1dVVFhcXMRgMyOVyamtr6evrIx6PF6VXDm55zpaWlrh8+TKBQICbN2/i9/vv+aGWvBZSOFE65WWzWVKpVFEsDLlcTlxLKd9vq8a1UZbF5XJhsVhEKE6q2i2mxVzCarVitVppamqisbHxA42wpXxJlUpFc3OzONEmEgnkcjk+n4/h4WGWl5eFIHIxzvNuSO3ZlEolDoeDX/3VX0WtVlNfX4/FYsFsNosS/rq6OjQaDUeOHGFqagqPxyM80bsdqcWRJNMkPePxeJxgMFgUxVxSoUZVVZXYlFUq1aZ7OpfLce3aNUZGRkSruY1oNBrhsdJqtaLlkbSu1dTUUF9fT2dnJy6XS7RwlMlkm3T9AoEAXV1d+Hy+otMlfL+iP9zKqRobGyMaje7QqLYHKf9Xyv/emDcq7cdSi8re3l4GBgZwOBwYjUZqa2tJJpPs27cPoCQNOY1Gg06nQ6/X39YTF4/HuXr1Km63W2iEPqh1fFsMudbWVp5//nmamprI5/NMTk4yNDR0z+ETKVdsdXWVGzduYDAY2Lt3L/39/RQKBWZnZ4vugZaIRCJCQ+ujoNVqGRgYYGBggNbWVnHKk6QrdnpzlyrUpDyve5FiuB+sVitHjx6lo6ODpqYmkTyfTCaJx+NFYcjeDpfLRXt7O/v27aO3t/eO75PJZPT394t5ATz77LOsrq5y9epVfvazn/FP//RPQmux1JAM1tbWVv7qr/5KbNhKpRK5XI7f7ycej9PU1CTyCd977z0uXrxINBot2gPaViKXyzGZTDgcDtrb24U8zdraGnNzc0XxHcjlcqqqqujt7eWP//iPxaF84/qTzWZ55ZVXePfdd2+bVmE2m6mqqhJpMZIxv3//fmpqatizZ49IH7mdaLhOpxP9LBcWFjh37lzRrvsbmZqa4vXXX9/x6MlWIxnYkjEvhdUDgYDYD2pra+nv7+fJJ5/k2LFj4t+5XC5xeAX4xS9+sTOT+BhIXR6USuUH8rXX1taYn5/ne9/7Hr/4xS9YX19/oIfSbWvRJem9SXwUA0TKH5LaW1VXVwuxXI/HQzAY3HUneKPRyNNPP01nZ6dofSMZccXQ2iWXyzE9PY3BYGD//v3YbDaOHz/OlStXtkTMVK/Xi3ZtGzGbzVRUVBRV/tD7udd7XOqzKrVykXSnWltbaWpqora2Fp/PtyU5hw+KQqFAOBwWIRaVSoXVaiUWi3Ht2jUCgYDwsEejUUwmE1arlSeeeILGxkaefvpphoeHuXz58g7PZPvR6XQcOXJEGDJw6/vzer2MjY0VRTGMQqGgs7OTrq6uD4RTNxIMBgkEAuRyOVQqFXa7HaPRiMPhYM+ePfT19VFZWSnatWk0GpxOJ0aj8Z6LlxQKBXq9/q7jKCYkT2Y4HC6Ka7lVRCIRVlZWiEajxGIxkRu9cU+SvG8bRf2j0SiDg4Mi2lRXV0dLSwuBQGDXFIL4/X5GR0eF4sSD3qe3ZVeUWlB93CrDTCZDMBgkFouRz+epqqoS4ZnZ2Vmi0eiuM+QMBgMnT54UvVqlMKaUIF0MhtzMzAwGg4FsNktFRQXHjx8nGAxy5syZj/35Uin/xh6dhUIBo9GIzWYr+srV9xeu3A6/34/f76dQKIgQrF6vp6WlhaamJurq6ojH4yVlyAGEQiECgYBICjYYDPh8Pq5cuYLb7WZ0dJSZmRnC4TB2u52WlhY+8YlPUF9fz8mTJ5HJZA+NIXf48GF6eno29Sn1er1MTEwURQ6wQqEQhQd3euYkCSXJeNfpdNTU1OB0Ounq6uLJJ5/kxIkTm7w4H+VAL/Wl3ii0Xizc7nmXek8vLi7u0Ki2h3A4TKFQIBKJbDLkNs7fbDZTU1ODVqsV711ZWeG1116jtraW3t5enE4nzc3NZDKZXWPIBQIBxsbGWFtb2xGJtC015CRPnNQI/eNuuqFQiHPnztHY2EgmkxFNbKWHeqvDejuN2WzGYrF8oLQ/EAgwNTXFxMTEjp/wstmsKKn/7d/+bYxGIwcPHhQ9Za9du8bs7Ox998OVjBmFQoHH48FqtWKxWEQexszMDG63u6Sqnd6PZJS/8sorIrRQUVHBF7/4RRoaGjhw4AC9vb384R/+IT/84Q955513hJ5esZPP5xkaGhJCugBjY2NEIhEh/CxtAFLlJsD58+dxOp10d3czOjqKXq8nnU6XrFTB3ZB00mprazl06JCQZohEIuJEf/PmzZLIrXrvvfcYGRnB6/WK6s3GxkZ+93d/F6fTSVNTE9XV1SKf6qOQSCSYmJjg0qVLvPLKKywvL2/xLD4emUyGkZERVCoVU1NTVFdXU1VVRVdXF88995w4sBWr9un9kkwmkcvl4vmV0og2zs1ut9PZ2YnJZCKfzzM3N8fk5CTnzp2jqqqKtrY2EokETzzxBLFYbJN2aDFTKBQIBAKYzWai0Shms3lTsUN/fz9VVVXU19czMTHB2bNn8fl8eL3eB+J82VJDTion12q1W+KRSyaTzM/Pi81MSjLceMLbLUjJxWazedPcpAqoYDDI6urqDo7wFvl8nsXFRbRaLWtra2i1Wmpra+ns7MTj8bC6uirU8aU2bO9HLpejUChEfoxcLkepVIreq7FYTPS0S6VSxGIxVlZW7rlgZieQEv1vd19KBTxSL9LBwUFOnz4N3GrZJOUBFQoFampqOHr0KENDQwwPDxMMBkvCkCsUCiLBV6PRkEwmuXz58h095ul0mmAwyMzMDBqNhgMHDmCz2VCr1ZuKaXYTkkRLVVUVjY2NVFVVIZPJiMViLC8vi3aFxYJUXXo7FhYWGBwcFP0mzWYzTqeTw4cPU1NTs6m/8PsNuWw2K577uxl56XRaVDjeuHGj6O6JfD6P3+9nZWUFj8cjRJMrKyuFzqbZbBbyQqWO1LFFKm673YFL0juVIiqhUAiv1yv0M2dnZ4X4tdVq3aGZfDTW19cJBoNEIhFsNhtarVas9zU1NdTU1JDL5XA6nSwtLSGTyQgGg6TT6W035rbUkDOZTPT19dHd3U1bW9uWucKXlpZ466232Lt3L83NzVvymcWGSqXi5MmT9Pf3iypdQCzyO+2Jez9+v5+//uu/5uDBg/z+7/8+jY2NfOlLX+Lw4cMsLS3x05/+lOnpaa5evfoBL5okflpZWYnJZKKtrU2o2zscDpqamsQG8sYbb3D27FneeOMNFhYWirblU2NjI/v377+tQKTH42FwcJDh4WGGh4cZHBwUr4XDYb797W9z/PhxTp48iVarRaVS4XA4sNvtJXNihVselFQqxdWrV0U1893I5XIEAgHW19exWq0YjUbUavWu1JSTDrlPP/00PT09VFRUCNmRiYkJXnrppXvuiPIgUKvVPPHEE+zZs+cDFXoAq6urLCwsUFdXh8vl4oUXXqClpYX29va7yqkkEgmuXLmCRqOhra0NrVZ7x16V8Xic69evMzk5WdRV3IFAgO9973scO3aM1tZW7HY7ZrOZr3zlKzz11FN87WtfY3Z2tqSjCbdDkl7ZSCAQYHx8HKfTiclkIhAI4Pf7hSFTKBRwOBz09/fz2muv7cSwPxKFQoGVlRVisRjf+ta36Ovr48UXXxTPhlSw09XVRVNTE62trczNzfE//sf/YHl5mcXFxW29f7fUkFMqlRgMBtH0HLamNYdkzLS1tQG3qkc2GjuljiRF0NLSQktLizBiCoWC8EYVQ97MRpLJJCMjI+j1emZmZsQpXKlUUllZydzcHFqtlvX19Q8YX93d3fT29lJdXY3ZbKajowOj0YjBYECv14vwWjgcZnZ2lps3b7KyslLUDdgtFgtOp1McXjKZjGgxJIlEDw0NcePGjU25b4VCgWQyKYwXSTxY8oaU0j2ez+fJ5/P3HBqUtBdzuZyQLfkwL02pIj3jzc3N4pBbKBSIRqP4fD4mJyeL6v7e2Cf2dhIbBoMBh8OByWRCp9PR398vktzfH4mRPDeZTIZoNIrf7xc5tvl8Xnz2xtZtmUyGeDzO8vLyHYXUi4VkMsnU1BQ1NTUsLy8Lwdzm5mYhtROJRPB6vbsipzudTpPJZDCZTJhMJlQqFdlsllwuJzrYSN2XpDVBqng1GAwYDAZMJlNR5jzeDUmQfmxsDLVaTTAYFNI60rot7V9SillLSwu5XG7b8yW3pdhBar2zVQtyOBxmamqKvXv3Cqs3kUgwOjpatB6a+6Grq4uWlhaeffZZ2tra0Gg0QnJjZGSE7373u0xMTOz0MDeRSqWYnJxkdXWV2dlZnn/+eX7nd34Ho9FIRUUFf/iHfyjCie9fhFUqlchx3Nh7VjIAZDIZc3NzXLx4UXjkij000dbWxrFjx7BareRyOZaXl/H5fJw/f54bN27w2muviVZsGw83ZrOZX//1X6e/v7+kjLYy90dzczNNTU089dRTdHR0oNVq8fl8vPPOO/ziF7/gypUrReWxkYqaTCYTvb29H7g3P/e5z/HMM88Atw4fUk707dJppPSYyclJ1tfXyWQy1NTUAGwy4qQ/53I55ufncbvdnD59WhTQFCvRaFQIwft8Pp544gkef/xxbDYbdrudX/u1X2NoaIhvfetbRXcgv18KhQKLi4s4nU4Rgairq8Pv97O2tib089ra2kTOd1VVlRBFP3bsGAaDQYRoS41kMsnZs2dZXFyks7OT5uZm+vv70Wg0qNVqNBoNcrkcnU6Hy+Xiv/7X/8r58+e5fv36tu5hW2rIpdNpcUFDoZBIXv+4SLpykns2kUiIStbdgM1mo6amBrPZjF6vRyaTkUqlWF1dxev1CpdusSG1JZMW3fPnzwuleilMJgl8bjxtS7kWUmsXqU2TTCajoqICh8MhTuQbvVXFjFSpLVUaLy8vs7CwwI0bN5icnBRSORvvWZ1Oh8VioampCafTuSs9UQ870knd5XLR0dGB1WpFq9WSz+eJRCK43W4WFhZIJBJFtbFJhpwkyqvT6Tat5ZLnQTLw7rYWazQaDAYDFRUVIlJjt9uFuKxEPp/H6/USDocZHBxkfHxcyE8VM4VCQazXY2Nj1NTUUF1dTV9fHxUVFbS0tJDJZESP2lKu1CwUCkxPT4uUGL1eT2VlpWg5F4vF8Pv9uN1uNBoNqVRKdPCQws5SrlwxeaDvh1QqRTAY5Nq1a3i9XlZXVzGZTJjNZtra2nA4HCJn2mq1YjKZUCgU26oDu6WGXCgU4sqVK7S2tuJ2u2ltbcXhcGzlr6BQKDA+Ps7g4GBJJIHfC42NjezZs0e0dIJbp7xr165x8+ZNxsbGivZEmkwmWVxc5NVXXxVNhXU6HX19fTgcDurr6++YMC2FTa5cuUI6naa/v589e/bQ3d39gGextaRSKS5cuMDQ0BAvv/zybWVjpFZnLS0tPPLII0JupszuQqVSodVqeeyxxzhx4gQVFRUinL6wsMArr7yC3+8vOo9zOp0Wea4HDhzAbreLQ+ZG7uUw7XK5cLlc9PX1iXVMKnaQkApczpw5g9vt5tVXXxUerlI5sPv9flZXV4lGo8zNzfEHf/AHVFZW8vjjj9PW1sa5c+eYnJzkypUrRbuefxi5XI4f/ehHXLt2jUcffRSdTse+fftQqVTCOItEInzrW9/itdde47/8l/+C0+nkK1/5CtXV1bhcLs6dO8f3v//9kuvusBGv18s//uM/ivtYal/2p3/6pzz++OPiwCJ1M5EOb9vVJ3zLQ6uSgO1WPHwqlQqLxUJraytHjhzBarUSDodZX18nFouV7MMgodPpMBgMNDU1icRfCamH6erqatHPU2o1JZ3IpNwHo9HI1NTUHb2y4XCYeDwuEkFtNhtOp/NBDn3byOfz5HK5O7Yvk8vlNDc3i+otKWk4nU4LDbndUu12J6QuENJpVcqpKfb7/X6orq6moaGB1tZWkUOZSCS4ceMGw8PDrK6uFqW3XfKOKZVKfvKTn4h+11J+k8Fg+EBRw52um2T8KRQK8Z58Pi9aknm9XoLBIOFwmLffflt07olGoyVjxElIlayjo6OMjY1RUVFBQ0MDarUai8Uicg5L9R4vFAqsr6+zurrK8PAwdrudRx99lOrqauRyOdFolEgkQj6fJ5VKodfrsdvtWK1WzGazEPmfnZ0tac8ksMmDHgqFUCqVRCIREomE8GBbrVZcLhdHjx5lZmZm2wqailcmn1tFDS6XiwMHDvD888+LG0hSzC61h/z9SOKJvb297Nu3T6hhw61S59HRUTwezw6O8N5JpVKbQqDz8/P39e+lPqQbZQt2MwqFQngg7Xa7uPbxeJyVlRW8Xi8+n2/bTnDFgKQ/plKpREL8bjPkmpubeeKJJ+jt7aWxsRG45W1/6623uHnzJj6fr6hCqhL5fJ6FhQXRYaStrY1nnnmGuro6Ghsbqauru2t16och9dMeHx/n/PnzjI+PMz8/z9DQkBAYLlWWl5dZXl7m4sWLKBQKHA4HKpWKiooKLBbLTg/vYxONRikUCpw7d47+/n5efPFFlpeXaWpqYmZmRmh+hsNhrFYrNTU1olNRPB7H7/czMjJS8td5I5FIhGQyKRoYWCwWNBoNlZWVdHV18cILL/DWW2+VliE3Pz/PqVOn0Ov1VFVVcfDgQaxWK4ODg/dchSSTybDb7Xzyk5+kp6cHrVbL+Pg44+PjLCws7Ir2XD09PZw4cYK2tjZRhSudVFdWVjh//nxR6UptJ9lslkAgQCgUIp/PY7fb2bNnD2+99dZOD+2eWFtbY2FhAZPJhFKppLq6mmg0yv79+wmFQqyurmKxWLBYLBw5coTm5mb27dtHZWUlarVa6KpduXKFn//851y+fJlAILArDTm5XE5DQwMNDQ2cOHECh8PBzMwMy8vLIhm+1LHb7cKIO3nyJE6nU3RvmJ+f59y5c8zNzRX9YTSbzeL1ekkmk8RiMaxWK3a7HYfDgc1mY+/evdhstruu55JHTsolm5ycJBKJsLS0xNLSEjMzMwSDQdH6abcg6Y5ls9mi70hzv6RSKd555x1WV1fZt28fer2eAwcO0NbWRjgcxufzkUql2LNnD5WVlSgUCnw+H2+//TbvvfeeaGW1W2hpaaGzs5O2tjbsdvumVpKpVIrl5eVt7dSzLYac1+vl8uXLHDlyRFSZms1mamtrSSQShEKhuz74kkBsRUUF+/fvp6mpCbVajc/nw+124/f7S776BxAtilwul9CjkUq4vV4vbrd7V27ktyOfzxMKhUTlqsVioa2trWROsOFwmOXlZVpaWjCZTNjtdhKJBN3d3aLRt9Rn8POf/zwHDhwQxREymYz19XV8Ph/Xr1/nBz/4Qcn0abxds/MPQy6XU1dXR3t7OwcOHCCfzwvv1G5Z3C0WC319fQwMDHDo0CFkMhn5fB6fz8fc3BxDQ0OiTVsxk8vlCAaDBINBZmdnhayC1HklkUjQ0NBwTwdzKSx35swZ/H4/MzMzRKPRkmtFdzs2GqtScVcymWR9fZ18Pr/rDLlMJsONGzeEekRHRwcHDx4EELlghUIBnU4nciHX1tZ49913cbvdJR9W3YhcLqe+vp5Dhw5RX1+P2Wze9HoqlcLv929rx5ZtMeTi8Tg+n4+1tTWi0Sg6nY6Ghgb+5E/+hOvXr/M3f/M3xOPx2yaA19fXU1NTw/PPP09zczOPPvooKpWKYDDI4OAgb7zxBoFAYDuG/cDQ6XRUVFQIMc2Noopra2t87Wtfw+12F52S+YNErVZjNptvK0ZajLz99ttMTU0Rj8fp6Oigr6+Pvr4+jhw5Iow8h8MhCkC0Wq3QmfP5fIyNjfEv//IvTE9Ps7q6WhLXvqamhsbGRpEHODEx8aGLVWVlJXa7nc997nO0t7eTzWaZnJzkH/7hH0o6+fn9OBwOjh49Sn19vQgpRSIRvvGNb4jcuFI8pGUyGVF1vrq6yne+8x2R2H0n7+LGooZsNiv6Ucbj8ZK4z++ETCbDbDZjNBppbGwkFAoxPT0tijs+9alPMTAwgMViKfrK249CPp9nZWWFv//7v6ezs5PJyUn6+/vp7e0V67bU0mtlZQW3283bb7+N3+/f4ZFvHU1NTTz66KN84hOf4IknnqC6ulq8JgmeT01Ncfbs2W2Nrm2LIScJP0ptpaqrqzEajezbtw+41V9SLpeLqlNJAFWtVuNyucSX43Q6qampER6q5eVl5ufnS75aVafTUVlZSUVFBUajcZMbNplMcvXqVWZnZ4v+tL6dSC28SqWSc2lpiVAoxOjoKEqlkoaGBtGtIpVK4XQ6qaiowGaziVZd8XiceDzO7Ows4+PjvPfee4RCoZKQW4FbOawVFRVCuFhK+JWkgqR+u+/vxehyuWhra6O+vp5QKMTi4uIHhJJLFZlMhkKhwGw209DQIDzK8XictbU1bt68ydDQEIlEoiSfb6mQTTLASqEv7HYhk8lES6qOjg48Hg9+v5+amhra29tpaWmhsbERtVotcrpLPR3o/cRiMYaGhkilUjgcDiwWi4gwKRQKMpmMKNybm5tjcXFx1xi1CoUCu93O3r176enpobW1FfhlZCKbzRIKhQgEAiwtLW1rhGVbDDkpLPTyyy9z48YNvvzlL9PR0YHNZqOvr4+vfvWruN1u3nzzTVG5eejQIZqammhvb8dms9HS0iKqYE6fPs13vvMdcZIt9rySOyGXy1Gr1XR0dPDFL36RQ4cOYbFYNhkrmUyGxcVFvF5vSS70W0WpzT2ZTJJOp/nXf/1X7HY7V69eFS72pqYmjh07Jt772muvMTQ0RDAYJBQKceHCBSKRCKurqyW10K+treF2u/nsZz/L4cOH+fVf/3UymQyDg4MioTkWixEKhYRY7IsvvsgjjzyC2WwmEAjwf/7P/2Fubq6o++jeDzqdjra2Nvbt28eRI0dEQcDFixe5fv06CwsLH2g0XqY0USqVNDY20tfXx5/92Z8xOTnJq6++yv79+xkYGKCxsRGbzSZ0QRcXF0umeO1eyeVyrK+v43a7mZub47333uP1119nYGAAp9NJLBbD4/HwD//wD/h8vl1z76vVaqqrqxkYGOBLX/rSB8KpiUSCSCTChQsXuH79OpFIpHQEgSWkGLnUlmJ5eRm73U5tbS1Go5Guri7kcjl+vx+dToder6e/v5/6+noRdpK8FsvLy4yPjzM2NlZyG937kar07Hb7JuFACSlfRBLKLVM6SN0p/H4/sViM0dFREXKSksQtFgsmk4kbN26Iqq5IJMLs7GxJhtnS6TTRaJRwOEwwGBSe94aGBmw2m/BEhMNh5HI5CoVCeOPn5+dZWVlhYmICv99fkvO/HVqtltbWVurr6zEajWSzWdbX15mbm2N8fHxXVNuX+SVSNEkSxe3t7aW1tVW0K5PL5SwtLTE3NyfypHaDIbORfD5PIpEgkUgwOzsr2nFJa6Hf72d6errk89r1ej1qtZqqqir0ej1Op1Ps4xujaoVCgUAggNfrFdXY2ykGDNssPzI/P4/H4+HUqVP4/X6effZZLBYLBw4cYGBggBdeeAFAXHiFQoFMJiMajfLmm28yNTXFz3/+c1HZVOoLoEqloqqqio6ODp588slNFz+bzfLee+/hdruJx42xTZcAAATYSURBVOMlbbBuBaWqtZRKpUin07z55pub9LO+8Y1viCRoSWpDCjuWqidK6rrx0ksv8e///u8cOnSIxsZGPv/5z1NTU0NLS4u4jlI47nvf+x7f/va3OXv2LIFAQCSD7xZqamr48pe/TGNjI3K5HJ/Px+LiIj//+c85ffp0yaeFlPkl+XyeQCBAIBAgHA5TW1vL7/zO74i9TEof+sd//Edu3LjBuXPnSjakfq/Mz8+ztLTEhQsXxPontZssZeRyOV1dXTQ2NvKVr3yFqqoq0YtYajcpkU6n+clPfoLb7ebll19+IHqI22rISZ650dFR0um0EMfbv38/mv+vvXtpaV2NwgD8WmqjtURKLRmITgUVHPg7/Rv+AEF00IGDItIbor3RFC32klZjDL2EoiTdg00+To8ctx52L0neZ2gFs9qmXeb78i5JmtrI7t6JOhgMYBgGbm5u0Gq10Gq1YJqmLxobNwU6HA5jdXX10x1/lmX5Iug46NyRcv/k1y9wN1LCvQL//v6O6+trJBIJ1Ov1qQ9zx3FQKBSgqip0Xfdk4OtXIpEI1tfXEY/HRS6gruuoVCoibsFP9QadO2ZN13U0m01sb29jd3dXPD4cDmGaJh4fH9FoNDw7X/Qn3PPcjytKkUgEGxsbUBQFiqJAkiQxtcj9zu50OtB1HdVqFbVaDcPhcC6rDTMPBHYcB6lUCpFIBPl8HsfHx9jf358aRwUAxWIRqqqiXq+j0+ng7OwMlmUFqqlxG1m/n+zf8e9N8rTcJpOJSLNPp9Nf/p4fX9dQKARZlsUSurusXqvVcH5+jlarxfPaZxzHQbfbxdraGjKZDA4PD6caOU3T8PT0hHw+P7MgWJofSZLEbNlkMjn1mNvAZrNZ3N3d4fLyEu12e25bRuYy2cFdPtI0Dfl8HicnJ+JON+D3k+DGLhiGgX6/L3Jo/MTdS/Dy8oJSqQRFUZBMJsU+o1wuh3K57Mv/Zn6KA+S9ya+N2p+srKwgFotBlmXIsixuckgkEtjb20OlUlnwEdKsmKaJVCqFYrGITCYjfv76+irCwMmfms0mcrkcDMOAYRgolUpot9siCHpen4VzG9Fl27ZI8s7lcvP6s0vFtm1YloVut4vb21scHR1ha2sLpmlC0zRkMhmUy2XPxE8Q0W+hUAixWAybm5uQZVlckVMUBQcHB7i6ulrsAdLMmKaJi4uLRR8GLUCj0cDp6SlUVcXDw8PCls+Xetaq39i2jcFggPv7ewyHQ8TjccTjcYxGI1iWBVVVA79Hrt/vo1Ao4OPjQ+wvIVp2tm3j+fkZ7XYbnU4HkiRhPB4jm80inU77LnaCKEgcx0G1WoWmaXh7exMh/r1eD5VKRawiLmoPLBu5OZpMJhiPx+LKJH3mNrSDwUAMjidadu7Gd8Mw0O12EQ6HMRqNoKoqisWir0YSEQVRr9cTkSLLZuUPw459c2loMpn8aNNVUGtfdN3RaBQ7OzuwbVsMkv+/6fFeqvtv4nv9e/523aFQCNFoVCS827YN0zRFNuQsY2b4mn9PUOsGglt7EOpmI/cfglo76/Y+vte/J6h1A8GtPah1A8GtPQh1f9nIEREREdHy8sZEciIiIiL6hI0cERERkUexkSMiIiLyKDZyRERERB7FRo6IiIjIo9jIEREREXnUL1OG+ZDK7RcgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 792x792 with 100 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#dataset = 'MNIST'\n",
    "\n",
    "if dataset == 'MNIST':\n",
    "    #C:\\Users\\giann\\Desktop\\ergasia1\\\n",
    "    path = r'data\\mnist'\n",
    "    x_train, y_train, x_test, y_test = load_mnist_data()\n",
    "    plot_mnist(x_train)\n",
    "else:\n",
    "    #C:\\Users\\giann\\Desktop\\ergasia1\\\n",
    "    path = r'data\\cifar-10-batches-py'\n",
    "    #C:\\Users\\giann\\Desktop\\ergasia1\\\n",
    "    labels_dict = cPickle.load(open(r'data\\cifar-10-batches-py/{}'.format('batches.meta'), 'rb'))\n",
    "    labels_names = labels_dict['label_names']\n",
    "    x_train, y_train, x_test, y_test = load_cifar_10_data()\n",
    "\n",
    "    # plot a random sample from the train set\n",
    "\n",
    "    ind = np.random.randint(0, 50000)\n",
    "    plot_cifar(ind)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "784\n"
     ]
    },
    {
     "name": "stdin",
     "output_type": "stream",
     "text": [
      "Type \"tanh\" or \"cos\" or \"softplus\"  tanh\n"
     ]
    }
   ],
   "source": [
    "batch_size = 100\n",
    "epochs = 100\n",
    "lr = 0.0007\n",
    "lamda = 0.01\n",
    "M = 100\n",
    "D = x_train.shape[1]\n",
    "print(D)\n",
    "K = 10\n",
    "\n",
    "\n",
    "inp = input('Type \"tanh\" or \"cos\" or \"softplus\" ')\n",
    "\n",
    "if (inp == tanh):\n",
    "    non_linearity = tanh\n",
    "    derivative_function = derivative_tanh\n",
    "elif (inp == cos):\n",
    "    non_linearity = cos\n",
    "    derivative_function = derivative_cos\n",
    "else:\n",
    "    non_linearity = softplus\n",
    "    derivative_function = derivative_softplus"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.ones((x_train.shape[0], x_train.shape[1] + 1), dtype=float)\n",
    "X_train[:, 1:] = x_train\n",
    "\n",
    "X_test = np.ones((x_test.shape[0], x_test.shape[1] + 1), dtype=float)\n",
    "X_test[:, 1:] = x_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Shuffling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "idx_list = [i for i in range(y_train.shape[0])]\n",
    "shuffle(idx_list)\n",
    "X_train = X_train[idx_list]\n",
    "y_train = y_train[idx_list]\n",
    "total_batches = X_train.shape[0] // batch_size\n",
    "\n",
    "# we must also shuffle the test data\n",
    "idx_list = [i for i in range(y_test.shape[0])]\n",
    "shuffle(idx_list)\n",
    "X_test = X_test[idx_list]\n",
    "y_test = y_test[idx_list]\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient Checking and Training "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 0/100 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Gradient Checking ---\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  1%|          | 1/100 [00:01<01:52,  1.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -86.12038218213468\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  2%|▏         | 2/100 [00:02<01:50,  1.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -35.89438683903743\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  3%|▎         | 3/100 [00:03<01:50,  1.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -32.081628068621555\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|▍         | 4/100 [00:04<01:53,  1.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -30.0023140270282\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  5%|▌         | 5/100 [00:05<01:51,  1.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -28.23374584861484\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  6%|▌         | 6/100 [00:06<01:47,  1.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -26.50908939412528\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  7%|▋         | 7/100 [00:07<01:42,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -24.85318431968152\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  8%|▊         | 8/100 [00:09<01:41,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -23.33353604951832\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  9%|▉         | 9/100 [00:09<01:36,  1.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -21.96735331596027\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10%|█         | 10/100 [00:11<01:34,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -20.741936592704857\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 11%|█         | 11/100 [00:12<01:32,  1.04s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -19.641578691178243\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 12%|█▏        | 12/100 [00:13<01:30,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -18.654239386426294\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 13%|█▎        | 13/100 [00:14<01:28,  1.02s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -17.769439842993577\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 14%|█▍        | 14/100 [00:15<01:29,  1.04s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -16.976293261934867\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15%|█▌        | 15/100 [00:16<01:28,  1.04s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -16.26373370472401\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 16%|█▌        | 16/100 [00:17<01:26,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -15.62154931070619\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 17%|█▋        | 17/100 [00:18<01:25,  1.02s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -15.040817976905588\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 18%|█▊        | 18/100 [00:19<01:24,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -14.513788342610997\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 19%|█▉        | 19/100 [00:20<01:27,  1.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -14.033660121983663\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 20/100 [00:21<01:27,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -13.594443470738247\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 21%|██        | 21/100 [00:22<01:24,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -13.190891311561499\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██▏       | 22/100 [00:23<01:21,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -12.818458062466322\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 23%|██▎       | 23/100 [00:24<01:19,  1.04s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -12.473254222135575\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 24%|██▍       | 24/100 [00:25<01:21,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -12.151988706605332\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25%|██▌       | 25/100 [00:26<01:19,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -11.851900620930008\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 26%|██▌       | 26/100 [00:27<01:19,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -11.57068475016441\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 27%|██▋       | 27/100 [00:29<01:21,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -11.30641600524484\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 28%|██▊       | 28/100 [00:30<01:18,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -11.057477982911422\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 29%|██▉       | 29/100 [00:31<01:17,  1.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -10.822500200610541\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 30/100 [00:32<01:18,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -10.600307190088705\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 31%|███       | 31/100 [00:33<01:18,  1.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -10.38987986867233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 32%|███▏      | 32/100 [00:34<01:16,  1.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -10.19032676048407\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 33%|███▎      | 33/100 [00:35<01:19,  1.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -10.000861729785997\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 34%|███▍      | 34/100 [00:37<01:19,  1.20s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.820785845925014\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 35%|███▌      | 35/100 [00:38<01:15,  1.16s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.649472280895392\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 36%|███▌      | 36/100 [00:39<01:18,  1.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.486353835947314\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 37%|███▋      | 37/100 [00:40<01:16,  1.21s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.330912868953233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 38%|███▊      | 38/100 [00:41<01:13,  1.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.182673376193526\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 39%|███▉      | 39/100 [00:43<01:09,  1.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -9.041194941373009\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 40/100 [00:44<01:06,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.906068229435924\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|████      | 41/100 [00:45<01:04,  1.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.776911681243604\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 42%|████▏     | 42/100 [00:46<01:00,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.653369080348261\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 43%|████▎     | 43/100 [00:47<00:59,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.535107720271396\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 44%|████▍     | 44/100 [00:48<01:00,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.421816975399299\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 45%|████▌     | 45/100 [00:49<00:57,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.31320714008556\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 46%|████▌     | 46/100 [00:50<00:56,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.209008435079893\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 47%|████▋     | 47/100 [00:51<00:56,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.108970094260973\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|████▊     | 48/100 [00:52<00:58,  1.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -8.012859452133643\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 49%|████▉     | 49/100 [00:53<00:58,  1.15s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.920460963292644\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 50/100 [00:55<00:57,  1.15s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.831575101258988\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 51%|█████     | 51/100 [00:56<00:54,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.746017106984025\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|█████▏    | 52/100 [00:57<00:52,  1.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.663615592651324\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 53%|█████▎    | 53/100 [00:58<00:51,  1.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.584211058878418\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 54%|█████▍    | 54/100 [00:59<00:49,  1.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.507654441405454\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 55%|█████▌    | 55/100 [01:00<00:48,  1.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.433805833104421\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|█████▌    | 56/100 [01:01<00:47,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.362533496092692\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 57%|█████▋    | 57/100 [01:02<00:45,  1.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.293713192707117\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 58%|█████▊    | 58/100 [01:03<00:44,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.22722777187348\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 59%|█████▉    | 59/100 [01:04<00:44,  1.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.162966899401622\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 60/100 [01:05<00:42,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.100826828798369\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 61%|██████    | 61/100 [01:06<00:43,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -7.0407101490856725\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 62%|██████▏   | 62/100 [01:08<00:44,  1.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.982525487367889\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 63%|██████▎   | 63/100 [01:09<00:45,  1.24s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.926187170519646\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 64%|██████▍   | 64/100 [01:10<00:44,  1.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.871614861409386\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 65%|██████▌   | 65/100 [01:11<00:40,  1.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.818733186192303\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 66%|██████▌   | 66/100 [01:12<00:38,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.767471366092701\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 67%|██████▋   | 67/100 [01:13<00:36,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.717762863175357\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 68%|██████▊   | 68/100 [01:14<00:35,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.669545046273141\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 69%|██████▉   | 69/100 [01:15<00:33,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.622758880733493\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 70/100 [01:16<00:31,  1.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.577348643779512\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 71%|███████   | 71/100 [01:17<00:30,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.533261665825768\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 72%|███████▏  | 72/100 [01:18<00:28,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.490448096928155\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 73%|███████▎  | 73/100 [01:20<00:27,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.448860696669342\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 74%|███████▍  | 74/100 [01:21<00:26,  1.03s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.408454645219153\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 75%|███████▌  | 75/100 [01:22<00:26,  1.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.369187373072698\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 76%|███████▌  | 76/100 [01:23<00:25,  1.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.331018407019604\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 77%|███████▋  | 77/100 [01:24<00:25,  1.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.293909230147267\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████▊  | 78/100 [01:25<00:24,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.257823154023526\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 79%|███████▉  | 79/100 [01:27<00:25,  1.23s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.222725201541821\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 80/100 [01:28<00:24,  1.21s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.188581999181968\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|████████  | 81/100 [01:29<00:23,  1.24s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.155361677617678\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 82%|████████▏ | 82/100 [01:30<00:21,  1.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.123033779696572\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 83%|████████▎ | 83/100 [01:31<00:19,  1.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.09156917486142\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 84%|████████▍ | 84/100 [01:32<00:17,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.060939979110329\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|████████▌ | 85/100 [01:33<00:16,  1.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.031119479645666\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 86%|████████▌ | 86/100 [01:34<00:14,  1.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -6.002082063462338\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 87%|████████▋ | 87/100 [01:36<00:16,  1.25s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.973803149290531\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 88%|████████▊ | 88/100 [01:37<00:14,  1.19s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.946259122535546\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████▉ | 89/100 [01:38<00:13,  1.19s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.919427273129896\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 90/100 [01:39<00:11,  1.19s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.893285736497767\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 91%|█████████ | 91/100 [01:41<00:10,  1.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.867813438082234\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 92%|█████████▏| 92/100 [01:42<00:09,  1.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.84299004204781\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 93%|█████████▎| 93/100 [01:43<00:08,  1.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.818795904801229\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 94%|█████████▍| 94/100 [01:44<00:06,  1.16s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.795212033847348\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 95%|█████████▌| 95/100 [01:45<00:05,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.772220052229913\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|█████████▌| 96/100 [01:46<00:04,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.74980216844658\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 97%|█████████▋| 97/100 [01:47<00:03,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.727941151352069\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 98%|█████████▊| 98/100 [01:48<00:02,  1.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.706620309252455\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 99%|█████████▉| 99/100 [01:49<00:01,  1.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.685823472208937\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 100/100 [01:51<00:00,  1.11s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch average loss:  -5.665534976534376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('--- Gradient Checking ---')\n",
    "gradient_check()\n",
    "\n",
    "learned_weights_1, learned_weights_2,loss = train()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicting "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also a plot with our faults will be created"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on Test Data  : 0.976\n",
      "Writing to file: \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:20: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x936 with 25 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict()"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}