{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "uTxi8m8GEfHL" }, "source": [ "# Konwolucyjne sieci neuronowe\n", "Autor: Anna Dawid\n", "\n", "Korekta: Artur Kalinowskid\n", "\n", "Dzisiaj poznamy prawdopodobnie najpopularniejszy model sieci neuronowej zaprojektowany do klasyfikacji obrazków wszelkiego rodzaju, a bardziej precyzyjnie: do wszelkich danych, w których istotna jest zależność przestrzenna. Zbudujemy przykładową sieć konwolucyjną (convolutional neural network, CNN), pobawicie się Dropoutem oraz ostatnim sposobem na zwiększanie generalizacji modelu, który poznamy na naszym kursie. Nazywa się on 'data augmentation', polega na generowaniu dodatkowych danych z naszego zestawu danych, bazując na prostych transformacjach - rotacjach, przerzucaniu, podkreślaniu krawędzi, itd. Będziemy znów pracować na zbiorze MNIST, czyli zbiorze czarno-białych obrazków z ręcznie napisanymi cyframi." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Sieci neuronowe ze splotem (konwolucyjne)\n", "\n", "Dzisiaj poznamy prawdopodobnie najpopularniejszy model sieci neuronowej zaprojektowany do klasyfikacji obrazków wszelkiego rodzaju, a bardziej precyzyjnie: do wszelkich danych, w których istotna jest zależność przestrzenna. Zbudujemy przykładową sieć ze splotem (convolutional neural network, CNN) \n", "\n", "Dodatkoweo użyjemy tecdhniki wzbogacania zbioru danych - `data augmentation`\n", "Technika polega na generowaniu dodatkowych danych z istniejącego zestawu, bazując na prostych transformacjach - rotacjach, przerzucaniu, podkreślaniu krawędzi, itd. \n", "\n", "Dziś nadal będziemy pracować na zbiorze MNIST, czyli zbiorze czarno-białych obrazków z ręcznie napisanymi cyframi." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Przygotowanie środowiska programistycznego\n", "\n", "By zapewnić powtarzalność wyników ustawiany ziarno generatora liczb losowych:\n", "```\n", "seed = 128\n", "rng = np.random.RandomState(seed)\n", "```" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Requirement already satisfied: np_utils in /Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages (0.6.0)\n", "Requirement already satisfied: numpy>=1.0 in /Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages (from np_utils) (1.18.5)\n", "\u001b[33mWARNING: You are using pip version 20.2.4; however, version 21.3.1 is available.\n", "You should consider upgrading via the '/Library/Frameworks/Python.framework/Versions/3.8/bin/python3.8 -m pip install --upgrade pip' command.\u001b[0m\n" ] } ], "source": [ "!pip3 install np_utils" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import sys, os\n", "\n", "from termcolor import colored\n", "import numpy as np\n", "import pandas as pd\n", "import tensorflow as tf\n", "from matplotlib import pyplot as plt\n", "\n", "from tensorflow import keras\n", "from tensorflow.keras.datasets import mnist\n", "# from tensorflow.keras.utils import np_utils\n", "\n", "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n", "from tensorflow.keras.layers import * \n", "import tensorflow as tf\n", "\n", "seed = 128\n", "rng = np.random.RandomState(seed)\n", "\n", "\n", "def testModelOnMyDigits(model): \n", " #Code created by M.Fila\n", " if not os.path.isdir(\"colab_freehands\"):\n", " !git clone https://github.com/m-fila/colab_freehands.git\n", "\n", " from colab_freehands.canvas import Canvas \n", " canvas = Canvas(line_width=2)\n", " example = (\n", " canvas.to_array(size=(20, 20), margin=(4, 4), dtype=np.float32, weighted=True) / 255\n", " )\n", " predictions = model(np.expand_dims(example, (0, -1)))\n", " plt.imshow(example, cmap=\"gray\")\n", " plt.show()\n", " print(\n", " \"Predicted class: {} ({:.0f}%)\".format(\n", " np.argmax(predictions), np.max(predictions) * 100\n", " )\n", " )" ] }, { "cell_type": "markdown", "metadata": { "id": "D1bbElA8FMVw" }, "source": [ "## Import danych MNIST\n", "\n", "Proszę:\n", "\n", "* wczytać dane korzysatając z funkcji ```keras.datasets.mnist.load_data()```\n", "* wypisać na ekran kształ danych uczących i testowych. Ile jest przykładów uczących i testowych? Jaką rozdzielczośc mają analizowane rysunki?\n", "* korzystając z funkcji ```matplotlib.pyplot.imshow()``` narysować przykład numer 0 z danych uczących i wypisać jego etykietę" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[34mTraining data features:\u001b[0m (60000, 28, 28)\n", "\u001b[34mTraining data labels:\u001b[0m (60000,)\n", "\u001b[34mTest data features:\u001b[0m (10000, 28, 28)\n", "\u001b[34mTest data labels:\u001b[0m (10000,)\n", "\u001b[34mTrainig data example number 0:\u001b[0m label: 5\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "(X_train, Y_train), (X_test, Y_test) = mnist.load_data()\n", "\n", "print(colored(\"Training data features:\", \"blue\"), X_train.shape)\n", "print(colored(\"Training data labels:\", \"blue\"),Y_train.shape)\n", "\n", "print(colored(\"Test data features:\", \"blue\"),X_test.shape)\n", "print(colored(\"Test data labels:\", \"blue\"),Y_test.shape)\n", "\n", "print(colored(\"Trainig data example number 0:\", \"blue\"), \"label:\",Y_train[0])\n", "plt.imshow(X_train[0], cmap='gray');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Wstępne przygotowanie danych (ang. preprocessing)\n", "\n", "\n", "Sieci ze splotem przyjmują na wejściu kolorowe obrazy. Kształt pojedynczego przykładu to: `(x pixels, y pixels, channels)`, gdzie:\n", "- x, y pixels - liczba pikspeli w obu kierunkach\n", "- channels - liczba kolorów, czyli \"kanałów\"\n", "\n", "Dane które analizujemy mają postać dwuwymiarowych macierzy o kształcie `(x pixels, y pixels)`. Proszę:\n", "\n", "\n", "* zmienić kształ danych wejścionych na `(x pixels, y pixels, channels)`. W naszym przypku mamy jeden kolor, czyli kanał. \n", "* wypisać kształ macierzy po spłaszeniu. Czy wymiar jest zgodny z oczekiwaniem?\n", "* znormalizować wartości danych do zakresu **[0,1]** korzysatając z funkcji ```numpy.amax(...)```\n", "\n", "Oczekiwany rezultat to:\n", "\n", "```\n", "Training data shape before reshaping: (60000, 28, 28)\n", "Training data shape after flattening: (60000, 28, 28, 1)\n", "```\n", "\n", "**Wskazówka**: Nowy kształ macierzy można uzyskać dodając 1 do starego:\n", "```\n", "X_train.shape+(-1,)\n", "```" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Training data shape before reshaping: (60000, 28, 28)\n", "Training data shape after reshaping: (60000, 28, 28, 1)\n", "Maximum value in training data: 255.0\n" ] } ], "source": [ "print(\"Training data shape before reshaping:\",X_train.shape)\n", "\n", "X_train = X_train.reshape(X_train.shape+(-1,))\n", "X_test = X_test.reshape(X_test.shape+(-1,))\n", "\n", "print(\"Training data shape after reshaping:\",X_train.shape)\n", "\n", "maxValue = float(np.amax(X_train))\n", "print(\"Maximum value in training data:\",maxValue)\n", "\n", "X_train = X_train.astype('float32')/maxValue\n", "X_test = X_test.astype('float32')/maxValue" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Zmiana reprezentacji etykiet\n", "\n", "Etykiety zawierają numer klasy - cyfry. Łatwiejszą do analizy postacią jest reprezentacja za pomocą słowa bitowego o długości równej licznie klas.\n", "W takim słowie wszystkie bity, oprócz jednego - wsazującego na daną klasę mają wartość **0**:\n", "\n", "```\n", "Original label encoding: 5 shape: (60000,)\n", "One hot label encoding: [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.] shape: (60000, 10)\n", "\n", "```\n", "\n", "Takie kodowanie nazywa się \"one hot encoding\".\n", "\n", "Proszę:\n", "\n", "* korzystając z funkcji ```tensorflow.one_hot(...)``` zamienić etykiety na reprezentację \"one hot encoding\"\n", "* wypisać na ekran orginalne i nowe kodowanie etykiety dla przykładu 0 ze zbioru uczącego" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Original label encoding: 5 shape: (60000,)\n", "One hot label encoding for training data: tf.Tensor([0. 0. 0. 0. 0. 1. 0. 0. 0. 0.], shape=(10,), dtype=float32) shape: (60000, 10)\n", "One hot label encoding for test data: tf.Tensor([0. 0. 0. 0. 0. 0. 0. 1. 0. 0.], shape=(10,), dtype=float32) shape: (10000, 10)\n" ] } ], "source": [ "print(\"Original label encoding:\",Y_train[0], \"shape:\", Y_train.shape)\n", "depth = 10\n", "\n", "Y_train = tf.one_hot(Y_train, depth)\n", "Y_test = tf.one_hot(Y_test, depth)\n", "\n", "print(\"One hot label encoding for training data:\",Y_train[0], \"shape:\", Y_train.shape)\n", "print(\"One hot label encoding for test data:\",Y_test[0], \"shape:\", Y_test.shape)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Stworzenie architektury sieci konwolucyjnej\n", "\n", "W dzisejszym modelu bedzimy używali nowych rodzajów warstw:\n", "\n", "* warstwy konwolucyjnej - klasy typu `tf.keras.layers.Conv2D` itd.\n", "\n", "Warstwa konwolucyjna dokonuje splotu na danych wejściowych (zielony rysunek) z filtrem (czerwone wartości w żółtym kwadracie). Jako wynik zwraca nowe dane (różowy rysunek). Żółta macierz przesuwająca się po zielonym obrazku (z czerwonymi elementami, które się nie zmieniają) to filtr lub jądro. \n", "Przesuwa się on po danych wejściowych w sposób ustalony przez hiperparametry takie jak\n", "- size - rozmiar samego filtra, tutaj: (3,3)\n", "- stride - krok, tutaj gifie = (1,1)\n", "- padding - co zrobić kiedy filtr wykracza poza dane, czyli gdyby w tym przykładzie filtr był 4x4: można ograniczyć ruch tylko do ruchów bez wychodzenia poza ramkę (padding = valid, domyślny), można zrobić ramkę powielającą piksele sąsiadujące (padding = same), można zrobić ramkę z zer, itd.\n", "\n", "\n", "Wielkości modyfikowane w ramach treningu to wartości wag filtra posostałe wielkości to hiperparametry.\n", "W sytuacji kiedy kanałów jest więcej niż jeden wartości dla wszyskich kanałów są domyślnie sumowane dla danego piksela. Zwykle każda wartwa splotu zawiera wiele filtrów, czyli wiele wariantów wartości wag." ] }, { "cell_type": "markdown", "metadata": { "id": "4R0z_b5pOoE6" }, "source": [ "\n", "![convolution.gif](data:image/gif;base64,R0lGODlhDgKAAfd6AP/AAILQUJCOperq6lJwKrBMIBpULTh3PJ48AF+zTWR7AE9NAAA0AAA1MOO2AN/f3wAADmy/UC1UI7hnAKilAC0AAAAiAIQhAMWeAIjQUHkiAPexsVtDThAQEEcAAGMjAPv7+3KeO2lgDvWjAPu9AJAwADRjMAAARyAgIEVYGdlsAPCOaIqnLYKKAMOOAMyWnmaHMGa7T3XJUBoAAAAxZkVni+PAANLS0riJAIu1P/WyAONyAMiSAAA1LT4+PkxMTLu7u4ijMPLy8rl1AI+3OP+6AFU1AP2xqZLJQcWlAGKzTfzAAIarOG5wF0klAO3AAK5LAHh8HXmLAKoAAC5SAGhIRwAALIlnAAAAAF4AAGm7TwBDMZLGQF9LAN23AJaaAP2ojKmpqQBBJ8d3AP2xsbZYANWAAOW6AL2yAJ+fn39/f+Tk5JLQT/j4+IirOABcOGl1k1xbAFanS3uHIJLCPXIAAGaXOwAlKxpUMuGkAP///4WeKvugg5RpX+7u7mRUB8pmOf7+/vKTanNzc6VFAKiKAPeadzBPfnTGUJLNQ5kiAGJqGYjQTMVnAI9NAC1fi/KpAICRI9exAPuomNFwRemxsfX19YPJUPWdAC8vL8y3ALCZq5DQUHarRNypAFtBAPm6ANmYmmzCULyYAPfAAMSIiAAkWv+2AL+jsJHNUNKeAMCaANJyAABNAG+3TPKvsV+RPZkAAIvQUDmGQ/uiAK50AMmosUaWR8lUAMM6AOiIAMvLy+uYAKtfAC15QJLQSbUiAOyLX8+cp2VlZfqpAP39/fLAAFNtAFdBAIWJo5WVldnZ2ZLQUP+9APG9AJHLS5gxAGx8IDwkAI3QUBpjOY3JTeWcl7KyspLNRqVqALSsAMPDw3y8TJLQTN27AIqKioKxQgAAGYysMv2wmvWoqVlZWn7LUJmSqQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH5BARkAAAAIf8LTkVUU0NBUEUyLjADAQAAACwAAAAADgKAAQAI/gD1CBxIsKDBgwgTKlzIsKHDhxAjSpxIsaLFixgzatzIsaPHjyBDihxJsqTJkyhTqlzJsqXLlzBjypxJs6bNmzhz6tzJs6fPn0CDCh1KtKjRo0iTKl3KtKnTp1CjSp1KtarVq1izat3KtavXr2DDih1LtqzZs2jTql3Ltq3bt3Djyp1Lt67du3jz6t3Lt6/fv4ADCx5MuLDhw4gTK17MuLHjx5AjS55MubLly5gza97MubPnz6BDix5NurTp06hTq17NurXr17Bjy55Nu7bt27hz697Nu7fv38CDCx9OvLjx48iTK1/OvLnz59CjS59Ovbr169iza9/Ovbt3iCiw/ogfT768+fPo06tfz769+/fw48ufT79++mHfe9rfz7+///8ABiggevnxJB4ACCao4IIMNujggwAcCOGEFFIoYYUYZoigeMx06OGHIIYo4ogkMsNhiSimmKJ4Be50oYYwQvhijDQuOGONOJ6o4o48dqhjj0CWyGKLOd2II41GHgljkkpm+GOQUIL4ZJRUDknkTUw2WWGWWk7IZZcPTkkllGKOCaSVV9b0JZgNrsmmjVi8uSQWZppZZp07opnmTG7KuWGcfmLYp5934rkinYaSicWeNg0qp6NvQspmoYmSSGmlIurJKEySgtlpl59qeSmmUiJKap6LbsonoIFayGqr/jK+CquDo57qo6m2oqipqi2F2qSvSgJ7ZK22EnvqrryuFCokOOCAQY3LNvsskrLCyKyz0OLKIxvgwEDAt7BwoyiU3HoLrrhBIptsSp8OMZ4FT1CLo7viwStvjfRiYW+MxnbohjTnGWDOmdqq+G/AA/eo7ronScpLFuQxYMy9MD4c8cT8VluhxeNJTC2QQahnRcKogiwyySouzHBJjjbTi3keZwyjyzBjPOfML5cX85w90lHBeGLYEQIB5L3BCY/G+gy00ESPZzTSqa7M0qDN1HHezhpSbXXNFFdY9dU2Z10wiWyIMN4BHxLxMxYQxAD1tmaLh7aHaovXNtRS96rx/oOneCDeMU9AIR7WTu7tYN9/Bz542IVriDgWgAuOBeGCjj0iEmtvIQuIkYw3y9s7Yi6e5px7jnfeyhreYBEeMJAEgggs3jWFrLsOu+wyZ1j76wDEPjnjlfPIwnhyhCg6FltMU/KOw4tXPIjHJ48q6qnDSIKCvlO+peoNXp9g9sBvbz32uPO8YxPihYNIiN18kv76KVsuIvpYqM+++/XDfyj1KoUK/uww+l/uaiRA86WoG0YQTwMyIKIovG95B0wgFhbYwAfGj3/s4t6ECii2I3GwcTj6YPBU1L7RbS5Ec7DgoXZUQuSdEEQpzN/0MNgwDUJIhOILYfk6qMPffYyE/vjzxYjEkb4IQBBFLRSiiIhYPyNekIYm8d8OQUjAKVbOg1bc3o4S4TcsKDFETMTC81aoIi6K54sgCuMYdRU1KI5Eij4cII1w6CosxtGAKOLC2tD4oTDe4ogl0uMZhziePz7RjW+04YPo6CVFOoiRsbKj9ly1I2x0kY8eCmMCAEkiSw5yiePZ5CERGRI4TrKRkgxfHXt4ylixEH/UOFqImhcOGXByRC2MpYhoactRkvIjplQlKlkpzEgS84cHxF8DXvghB8owfq9UIDM95Ez7+fKXHQkmADUEyTA5skHdpJX8QNQNZIjnDgGYnwrZyEJzYgGd6nzm/rAJEm3KMUbh/mzTNxmUTwbVqmzpE0WIfuGE0TEQmioCaP0ECiKCGnSG9PSIPW92zHsGMIuU3NEeChkiIpjuliPaqHgMmbaPXjOiGZkoD6t4x5XOEaON5JHSsGCFdHqIE38YjxJAp6KZ1vRDONXp6VDKEUhdrxkAKMHg4oVUl2LoqEldKgCaSsUMQVWpk2MqHpHYhfHg4RIdegb9XMjTA3ZVPF8N61hJB1GiaoRqGmBPC6pKoWbEdT1z7R4o4KQhu8q1cD0iwgzII4EUDDZ9WlDYODt6WPEUtrHhSOxQ3ZrSbz5OPQpo0ClwEYtp/clxXcQsg4iRi1jEQhE4SNCgLpuezDJos51V/lC/RIqeNSKUR7Q9j23nSdnK9vWu6pECg3Rh2lisQrbf9Ot6hKugU0whFgiAgmlH8dkMKTe4wy3ucVW7WBHlIAXmMQG6FBuk74Z3vJPtrUWElSAdLEFBOgDAKU47AeMil03uhS8AdhCLbCCIuAggRYT2OaH8tle+9LUvd6P0DHB0ohPcqEaU+uWhBj84whNuo3orwl4A0KK/7wWACmKRhGaYwRgf3m51tfThbIR4xEkorSQQ1IznemHAYGrxi0lsYhQr+E+5um2QLaXhDU+kw6cARn9FbFoHJAgTP8Zxl5K85BHHwgFTUES8EFTaGyNZyf61spMRBGUVD3jI7EQz/pGNfJEOA6AISi7umMkcZTfDubhXdm6AE9RlKXfpznJWUJmRq2YiF3pEKmPzQtz85ufGwhsLGjR331QER0O60giwGWedzOhKmxbSgq5zd4NM4UolWtEJYbSVY1GLSIv6TatutXOhYbM+q7q4rQ61mUudKF4b6tSoPoibR6yIPIA5xACQ9Iq1RGxjg7jPCHruGfzM7NM628VPFvWhMzXqXAE72AXpMJQVMW3nxoK6CPqwZ6mtpHGX+7mr4GyuieuI94r7tO8+d4LUTehtl8rfpQK3RDps4jmfYgzOaAYueiFdQpRhAgImuBkMjvAPx6IXZTDttCRO8YQvvOEPj3i3/os18mMVWeCLJnCMQOHo4tKa3YEibnFdMOlZsRzPsXi5r/G08zp9G+V6YDRFZ4UgHRh42UT3Z8lJ1XM7nRzoCBG6U5N+X6rrE+AfavqYfo7yAXn962APu9jHTh6XDIPsaE/7f/CzFbW7/e1wj/t/XCL3utu9PFyROl2tjnS+nxnrtwL8reiOBTIY/vCIT7ziF8/4xpOB69l8ehQlPxW9X9HvVcc8kAWv9SpRniTicbzoR096yBf184nMu8o9tXpQtV5US8dU5zNMeNLb/vaIN/1GdP9W1EPF8jnUPMz5Pvtxcd73pSw87pc/et5XlvCqF37fif/6X8Xe1NfvNfLr/ql85nt/8c7HSPjbvP2mAH+Vwj9/TI8veBOVX6Ld/778DT/+9b4/8tGXvvqNmf7s/9r/PHd/pzd/BPh4Arh7B9h7+dd/0jd8Vld86QKAPpeAlVWA81d/HEaB4qeBRbF/3tSAHihO7QeBBFN7Fvh9GEgRKXhkHEgUIahPIFh9wSKBTjeCLZiBJ4iCN6iCO8iCCwgjeVALVzCEhTBjFqUh17JuU4eE0pItY5IDdmAHnZBh5NIt30IA4aIoJih/k1AKXvgCG7B8KzhwPUiGP4ghmPAB50EFoLZ3GZIv+zJ0MQKHW7ZVPNIN0TAed3BQBBMkB2MeAlOCejN/47A14iEA/mJYhhExhouoiD7hKLywHm0YfBrCMTDFf5UIMR1TTN4EJUywNiZkfDwSMukxMoq1hd4HCOWBiLjHiBDhig8BiwZiWaHFAIXAA9lAHq2AbOiHITSjM5woTjjDNRkDJNiwCOaxTKLYU6AYNENTNLJ0UjUkf3wAioeYiNCnN2c4IWUwHq6FIJAQWtrghg/yNcQoh3VliBeDTCriSck4TSAlIgqFBXPTIXXDNm7TVv0Tf7h3BIRgHqx4e7LYEAPJEAVZJI5UBJrIAM6wILrgjeR4OF0UOZf4gRnyOBTZUoDFI4JEj4yQUxMEj0KGItEzTZ0jHp+jjxn0fcEgHlbQB2sT/pC2d5AKQZOp5oj64UiRKB5fwCAKKR6t4D29SDutwzv9xFe6U5S3o5EjFDoVIAauwAzlJE3LiCLNI0bGkznKI42T532FKB6PYAjjIZOlh5MOYZNRZ5Yu4kiNMB6aAE5ueXlWRT5MSYkVIpS9U5Ei2CMS1iEtpIwRyCP0Y00f0kKEyUaoeHuUIB4nYAuCMJbYqI2SqRV9Ag1SxSCsEJd2eVF1OZSc2UqdGCV/KZJkFEEKxIfNtE5CopYGyY+kxwddBAdk8JjX2IqsmXLZSJmOhENmoJmeyU16CYMVZYd3qEykmWbJFIoopJprNoj9+I9YYAqVMJuQaZu5mTrbuEh//uMgmDAeaCCXL9WZw8RSoLmXUDKaVUkiSURITaSSKBF6uLcC43EOhkebWECWzXebNamfN5mdDWKOx8Cd49GTmwmc4omJ+BScSkcl6BmYZXRJ7ImVXMkyrul443AB4nEIYUidtSmQ/JmW14kVboKRDjIC48Fcv5khR5l54VmeVyeaxpmeI9KRmNQhfuSe06iYLokKh2ef+Cl6aClsHyqk/skgACqg4kEB4JmgB2qR5BmML3qeMeqgKeKONcoMmoSjXWl7YBCbiOejkTk1Q2oQQUoTfeI7VMCLT+ab49miUKp0qcSOQNKgfQhE4qFLs5Q+vcRbqXN7BSAeNDCdPTqf/mGKnZOZFWeaRZkpHm9ZoCqqoEj5pHLaI3RKXnYakhUkT4jpnKMnluJRA9YQCqIqDIspHnAQqq8wk2MabqtKEGW6Ko+qLzbQIG3JqEv6mW8aqW46qcVJlVSKRO4ETyEymPqzqWJqe/L5HsmgqiGaQUXKIN0oHt/JIBiaVbdqoC4Kp8MpNgw6pXWaIgoVDgz1IQ6FPKi5monZeMnqHstals36nq0aE32yk1iAoglCDBB5rbGarbrKpPzab1Lqq9+aIrRFUnRjUny6j7bHB1XQsA7rsKFlBVXAAZvArIe6j8+6IJc1qwoiOVigpPqKIStac5JajDCqnAOLIj5lUx0S/lTisVNaSqEFaJ/0aZ0X66xt903QiQVUMG1TNQGbGJEPclVSRVUhO7RTFVXWarRaBCSpwAypgEDi0QMM9LSWSkJnhQVpxQxiNR5sNaGgV6HLB6Y2K6bvWhWDQgzWuAB/YI3j6CAksFcsWlfAlR555agOcl12C7dyu2A8gg0FtR4Gi5wqIliEZVjjEVmnyKnyR7YeeraTl7EMQq/nQaCjVVqnlVrTxzeh1VqaxVnrtlqdix7feK+Yi1p+uyMdqR4pOZIEqx67Zax9SoDr+qOO96quGq8CgbvySmCQ4AjmEQdG+FrPFV3Ttbl5W7foYa//pV1zOyF6u7waW7zSpW9//ldJgaseg7ua5QVe5SFeEZiuzGcIa1OxZWuoZpuzNQIKeaAKquAJDQkARye/+7VkAABgIodfvBhf86UI9bVrMghf+1u//nW/0CVyNshgDgZhfRm+jCt/R0AOqVqoGHuzVtFhOoYgMCZjNGZjDkgjGcxkJXZiKfa8OBLCGxwLRlhjseBlNLh1L+x54puD+Qm5FCq5FwlmTHZlWVaHthbACEJlYdZk2QbAYCLEO4xlWsZnLXy9WEeCi3usNEzBOJu+uskmgNZkeiZgXNbEdhZnRFzEJowjWZxnU7BnXezCCXx8MzzFjce7u6u7QSfHhEdpltZomZYgm/bBNeJpj+Zq/kaMxXeMaZp2ZU4McFCcXivpxo9rwTWEwxoSa/I1BS+XxnxMI5Ksa2OMI5I8a7XmxTFMhWz8wIx8u3IMx3MMyRjSbMcGbQAgbZcMI6wMYmJMstVWbK3cxNEWC9OWyPGoZqgMn6Vcw44cuerbJe42ycYlb81bb7GcIclsbujmYbFly+2Gb8ocb6zWzPYWyjJ6aMEstsOceOFsw1LBcQlycM5gcRinccgbIwWXzgincAwXCw4Hcc/sixMnz+tsWu1czb7sutsWzuNMzFaMvogKxBMic6ZFc+/cJTeHZzqn0BDC0LHg0IfsbwGdsItc0G98yiCtyk1idLz4gtqadCQN/rBP7M0OLMUezXjlXMyVR9FH+IA0TZzgzNKC6NIvrXgxfdAiendCPdREjXaEV9RIPXaql9RM3dRO7R5H/dRSzR8i/Sg3vYQ2vcaA99MIXcFAfRUm3a9+F9Yqjcg6HcVdXcVpDa9VHSlXLbREt9GEa9bmnHoyHbZtPSlvfbRxfdZlRdd3bddfzdbHzID6t9dNu9V+HbN4HdgiwdVBHYOSfdhavdJ1/dgh7dhPQdbWPNaInVGKbYOXnXyjzX15zXqT3X+VDdiD/ciabdqFrXmc/dCzItfcO8qtbcy5fcOxjXmznc+BYtuGhttr7dq73dhXTCNBOIRXUITbVCHLTYTD/ovVGJKETkglUCiFohwl2T2FWvjawJTZxx0VkJKGaziJePsg5m0ebIiOGkKHvEpCeXhO5/rLJYKHeljfYCvYxa3b/T3TFSOJ1O0glHse6N2mGWKJTWqePfKJ5PG1V9vg1gjhikzY443ZpW1+tNgxt5iL47GLcK2xtdjhuqimCFpXObOOJvu3yFgegJmyVdri5PHiaO3V/43cN07e3xStWFC64Tgeb5ve0Jqv4CiOIa4g5giM8T0i7ujix3nblTS6Alvjam3jVl7lYJ2QCxm/CfKQ4lG6CA4hPzk5XP5fRJ6iEpk4kvOvqasiHXkAHznlf51HawPnIEnjFW7cOY7h/uDNFPM6oD6piUHJ1wpCr5abIGM+6EK+ILuzlGy+eU4JlVLpTnhu39DzlFE5lZja0ntO2n2Of70NIbWKBY26IL5D6oSeIKNe6nSJ6ou+IHg5spC+Iw1cqXOOIrXurVRu4Z0O2xe+2Y5kmdaKmWx64g8i7AxQhwqyqK6O5hUi6xndq5sO49Je6fvN57/u672u4fsaPr1pq6/u6N5e7E66qyseJLZu6eqp63nu31fO6++Otru5nQ3SndKa6ksZoPXunfj+SJBa1pTK7uqOSwJ/7Z6e7eGd4X6uSEe67zyJ7w3PIPaOBYdu7N2+5Mk57RE+pwXP0fCO5R8f8vKOISTa/iAmKh7MW+4QUvIMcvL12u9wueBRiu4dP9cBL+eMzd/x7u4gP/JeszX6LvHjAbLOjuRAj6QfC/P89O9tzvE4L9DRpPHtztsID38KvxSJyrMmnmzkLowiKx5p6vDNHubYmqsAX+1PPtwBK/U5j+3bbvWfrhRZrz3MzuoWH/N03/XCWbI4fanWXpo0//Qer+c7T/Vvj/XzLqu0qvcnTSG+YwEcuyCrrvSmzvSz7vRsD/V+n/aIJt6Hf3pXL/eOxOPTuiDVmuz9TvoNcvrKfvfPbvnRHvV/b/No/91VD+q3D/qh/iD0mvL4+uVHjiC9P1pnTvYXf+6YP/tQHviZP/g8/i/yz0/4WX6RoRX5COKxRF/0zVX9C4L9we/vMr+gJ0tWv4r5FN72B//5CBj6STEoO9uzNAa0kOr+4wH/Pxu0lE+0S9v3JPK0UStBANEjA7NUzAweRJjQIBYsCh0mLJiqmxGGAgk+xMhQz0aOHT1+BBlSpB6NI02ePFkS5UqWG1W2hDnyZUyaNW3KxAJA506ePQEQq8CQ4YI/QYVq85l0J0OlSoEKxULUKEOkSkmA6sm0qc9mGqB+Fdpia9aGGBVicwIW7C2zCRm2TYhW7Ve2cBdiuZmXJF69Nmf2bfkX8ErBgw0f3jvWJ6+5DL80PYUrFgaeWhXvZNz4cVJiuWLF/lKEY2nOyzpPeWgsVEHSyJMrl7XLjMvUxrNiM3t7e3ZqhrZjF0YcEnjwj8OJdzR+3CVf5c39ki6tE5IjsHEkNdX1Odaq19F3Tq9+nfWUWAigfB6l0/LlrryxSPGZ/TP30beZyeVd127u2PhT64crOeUEPI5A4gwMDkHnFuRoPe9AyUMVVTxxRicdluBJBwBOAW2C7brzbicIJaTQQgx30nCHWLLRKTsESAHAwRB9ujDDDTv8sD77dgwQNh5/VEjBw4Q0jMjBjAQMSQYLhG5Gn2hZ8UQVYkmiGTOMgZI+9Zp0kicos5GSSs/Ea4Y8L2LkskudvgyzyiuzfA1IOd3y/nFOIJXUC8+89LyJz+eWBBQlGdU8BZgVAZgyFgd2wiTHLdX0qdBDE3VgCkWe2MmzMwftUlIWKWXUUTTttJM/Un/0s6ZUaVo1plZhejXQ5yD1qQhDtVs0VC3RpLUnW7VT9JQpXsw0lk3ThPRXXHlqdFdTT93xWWhvi5WlagljTta9tG0wW2655VTNIsiLxZuemgWx153G/czccREwZifJFg23S3bLPVdUaafdr05++/1222+vFdTbQAkOOCVkIU00llrydXZhNRt+WFho4tVJU17V3YliiOP81759Q87IYEARNgllnAJWOWHhJHZySkXyMBRMXdNVV2aao9RYJ/LO/tiYY51rPhEAdHUkud+kfzN5yZZfZrlpBp922aN6Q2xUEaCFjSW9NV0bjWOjQduavFUkexiA7BzB8Grvsi67652gpAzppc0a+W6EqC5O6gX5ttpv5wCvOjGOrcx1wzGcaQaXXs4jpIwJYHQ7OsR3OmVxKGPppYzP6q68tMtNW7zxx2OJfPJR9cbbX9b3Frw5wruNuvbCD4aZY1DI1e7ioMVOSr7PXAgb+KR2BzYW3/N+nXnWZ1/OdnBjH5D624vM3XifQldXhxqL17405/Ue/27oBZ5eem3Pjzr86LgPH37ty1+a/qTPx9/6A/VPkP/r83RPAAU4QAIW0IAHRGAC/hW4QAaqJWoNhGAEJTjBxvzPaRTEYAY1uEEOdjA1D/RgCEU4QQtOzXWvuwsKH2I/krEwZC78V/7UJysZlnBwJ0QhDPmlw2nxEFo+PFUN0zcw/w2piDZ0FQ6bp8TnMZF8TjQfFOt3xCNRMUlW7IsQkbg/FToEiKT6YqmkeL8xthCLAJwh7tK4Rex1MUhlfCEcYyjHHdKxh2fcEx77pMc/EZGNTHIjnQIJu0EeJIxzOqSctEhDPqqqkax65B+RY8cfUjKIlgQjJsVYyLus8YKeNKEk+8dJ3GgSkaZUJCrvpEpURRJWrgwMLK0lS2yJEjGJXCUpcdlKXbKSR4tUox+F/mnLKvbSmJzc5S99GS1aFgyUf2umwojZRh6xARwwIEA2YcENXlbzmtkkwDa7CaQc2MEOncilnMp5znTuyJrY1CY3lfnMG9JTdtFEIgzdIA21GMAczOTRPvv5T5HZsRvREModBqJMIB00oQsF6I4EChZ/MtOe1bsok6ZZTPsEITVWICjTduTRxoC0oD9iAm22IItx2ielUFlpS+1C0rmYlFr4XNkw14dTkbCPiPahw1TEYIcQEAAqb+DETYEqVKIaVShIVap9sLEIsDSApQyVKlW/YlWZtiWoQhlqUY+aVKZldH9m7d9Gr3gbNohAKAdACBGMAoEYRNUubX1rXOda/leR3gYbqKnqVSPaH8BuVbAnjQ1eGQLXg8iVIXS9KVpvyVOo6VStz7kNEowSU4REQii+UVpsNMsQzh7Es72xq112g4UDMOIPDOEqVmOz2ta+FguxHSxcRouF0hrktFgAbY8ka8ThYu+yALwNC4QiB4XsdgvT6KtdlMsQ5ibEudANrWgrIAZXMKMbyIDtYVOr2+1297vh7SpGpouF6iLkumW1bDCHON/jOvI2TWBIOBChkG58Ir/7zS5c8IsF/fLXvwQGcI92VI2D9Be9so0Ngw3i4NuKN7oC/q+BMwyw+J6MsiABZn1jGZuJwBaiCInChhVslxLf9sQHSTGCL7wj/grjFrE/qrGFA2yWFjfgxQaJcYE5TF9GFreKIsYsiw/c24PMQcVtoR+FmWwQJ8t4x/bJcXpJfGAbjxcjUtZxlYUs3A5/ssyhRDIkt8wQXzxEHPmNwIy/fOA2O+TNBI7zlW+TZQjTmMs6XjFcKFxnhdw5HHkmM5Hlu1MjI5l+iQAsoRNyZ/bK+SGQZrObl2vpNVdYy0p+cG7NgmksSBohlG4vlD/ct0ZncdWBS7OaVWsUUx+E0gBq3WxprWmG4LpkcuKzqLH8508nZLW1Nsit4atoD7cajbF+JWEzbWehJIDTZ4k0r7FgbT13usvXFjSx+4yRv0670NVeNqPPDE1n/h83ygemBlkTMt1wyADcCKFwvB1Cb3t3G9SeHveexS3sh+Rb3gjhd7qLvO56MhzaffO2joOcYFVH3CETvzePB35jHgWb4xoPNYqf3Lp277HkfXw4tkgMXizcIQAOGfCYK85ilrsc5iP/dUM37mWBh5znDjlvy1+ukJhTnOQOxyjSNZpyld/VrQQWhUJ+kRbe/niFUlRsOKKekKmT1upepKTHf/7vb/s7IVnfOkK6XnWFL3rhzDYz0xV2mz0IxddE+OzYFVL3XisE76jN+Jx9HviC75zwzOA7Fu6e97Y3W+lnfXzK7fdVLFhh6AbhhG2xoAS9J4TylkdI5oXC+cNr/pi3gJ45jpeM+lzH5vOXZ4boGUL6Iasb7mi+vdwnuecuCAUPlzDIMwZ8+s7ju/cM+X3whz/lnNsnIi22SEE+bpfnUwQL0Sd4wY+PheQzQ/hCYf7VT+7I8UNS93O/DRFmABUJpGD9+dXC9M2ifva7XyjhiH/xEeKfxvi6+Xbhv7nwv6uzD/oTivZ7PwLLv8aLO9tzwLc7v5fZkcRTi1QzO4egQLCwwECbNfcIrtTrQN74wNaju9TYQFUrvyRKwVeKQJzgkRxIAbAwAXmSv7aAQRmkQf07iABUiwEEO6miuv/QQYS4wa+YwXmKvMlawVhqwZ6yo2cAh07oBG6QsICz/gsolEIqbKdBSqbsgwssnMIqtKgkJC4yNK4mBLFlqsEcUsMhNKNeWsJZisNaQkOIQ6Y2LL04OqZCCrEGhMA/rMO92EMuxMMLfCI4NMMjS8QrCsTAGcRA6sI1XCJEzL2Gq8R7asTdu8NHdKNIdMM4mkNnWkRXy8Ru4cQu8sQ8nCNKfEC3c0XHK8VUNMQoOkUVksVEa0VYzEU/DMRbBEFILEQOBEZWBERdLEZerENfJEFCrEU2JMZXRMZoxL1MVMb/G8ZNxEY+fLVJCkVpGsX6qkYCzEZmHEdg7MaU2Uba+UZ3G6F2dMd3hEcMAqF4pMd6hIpSXA571Md95Ed3DBgU/ujHgBShDsBHQSzHTgzGX0TIZyzIhmzIcPzBg0TFhFxGc3TIi3RIiHyjZpxEibTFdMTIkEwYjRQkj3RGk1wikVTJXqRIa1xIlGwihlzJmXw4kiQkmDxEnIwimuRJubNJQ2pJcdRGjmyinjRKaPvJFNLJKSLKJzrKpxSxpCylpqTFpSQjqMRKtZLKrQzKiBzKO8zKsLQlHXoncBInK4yNdUKnYoMLtdxCtvqmeGLLhyhLuURCscTLfJKjifqKipJEHkMohlCouSy4wGy5r1NIuOBLqPDLT2SGxRSKxmTAvKTMkYQjmlILm1JFZngp8GM9l7SLziStzxTK2MBMsNDM/llEiNP8itTExcqETctcKrBqqrH6xKkKLMLcP62CirITRq9iKrF6qoNLTIygPCwIK6diCKiazNh0ThNiq6djLb16LL5STYMot9xES4fITsPSTWZQrOlsrL0awvBkLINwLCyArOZ8zvYEJO0azYT4rREETYygLdcavOuUDaOorfz8TYxwLvGazyEMUPlkvNpzzwS9p+TaNOvaLOz6TwAtL+9iOd8szofQLO6iUP+8UIVYrw18r+v80OZ6UPZU0BNFrtgoOtOTuYqEiyoUuzyEUcO7zhVNCApr0fq8OSvDtwPLUfFD0SCdLBKzPh+7OJwrTT/jUBftOYDzwhstUqvD/jjV7DEpRVKwE9Is5ahwi0+FEDOj01EuddK/BLkx1UEwc4gvJTw09dIrfSMthdM86jRkYwZDQ7QObYsY1c8bpdEINT06tdM1pTNeOzQTjdNDTcPYIDVAbVA/FVMLZVKL206EWFRtO0GvbItKpTbqMlRE9dTl0DVzmzS728wJ61M8LVNIDVOEODZt88GShItW3VTF69RPRVT6yU5GZQhuc9Q8PdVIJTvSxFSzyFVt41VUxc5sm9VjPTpbddZE/Td989D86rdeTVVh3UidW9JVbTB4I06DSDgq9dZ9o9ZafVYtfbf8nFJrFTwzLVXv+lVuNVV1dVNY9VV65VEUPNd9/gVVmhNM2DsIG2XXwtvWJB22gh3WL6s5gDUIgUXWDRW6Hf1RLOXXfaUftJM6qtsCxLRXJXXXPe1RhM1WuMBYrtNYjr1Jkn06rctYrzPXij1R+0m8xQM8kO1Wke3Yg/3YgUU8Uk2IvwOuIZxZvzvQ14RZT508owC9g5C9zXPMmyW+SS3T8HtYZni90NM82uPZq2XarI2so3VW++mG7eu+7+tSkK2+ihgI6XPMtL2+tX3SGyVb4PO+5cPWqSSxuVU+z/xasP1UFjJAhkDA+1tA/eTBtXDDw6ULNwxcLBhc+CPTuEpAx7U/yO1bv71VOMrAr7hUeT2I1UoN+jRYswDd/tr4xM2Fis4dXbNA3UZ9Wcx1ThgqQqg4wriNiyDsv8TFXQF0zNkVitqN3ITwXYYA3suFXTjlITDUwu98Q6tMCOUVw+BVCOhdpePNXKpkSufVQ7C03jjlSuwlI/A1o+713q4cWXL8yvQlRPJFXvPN2YkU3+3VRvZFV/dNWfTF32ucX/oV0u/V3lX83zri3/61X6CMXwBWX4sc4JgtYKVM4Jd84IlcYBT13wj+yAMW4AlW0ArOXwjuYAnW4ATlYP394Avm3hBuT4FU4RVmYQRCYfcEyBaW4RmmYSwgyBfG4RzW4R3m4R724R8GYkAJhCAm4iLWgzBgCB+omh9giG0w/uInDo422IU0UIY02AYhYDplYAgUqJpMYIhvgOIw7osHiGGwSIOH02Is4GKX8WIsAGMxhmOaUAP3KIZYS+M1Tpg2fuM45mOUGIav+AFlUIM2ZohlsOMt7uIv7uNFHolrgAo1GGKOsIQ/xoIHOGQ1TmQ3ZuRN9ohAgIowCIllQAEsTrM7zuQ95uRNdmSGKIeYaINtUIZv+IZrGICQKIZlWAYQ2Ag/CANZ3oY65ggQwGVS/ohAeIBl8IOPeOVYnuVaBglT1gNhXgZi7uRjTmaPEIJrkGVluAFgDolA2IVYVgZL1gM9TuVULmNqPglLIOSv2IWP+AaGUIZA8AGwuAaO/kgDofDmjhgERJbkdoaKd/YIaM5nhthnjuhnTOYIPyhjqFCGkEDir+gAS2BiTT5nRgYBociElngA3njojkjjDmgMJ9YDIBAKgQ4cLFDijejo1PhojoBmk2YIlJ4kleaIZUiNVh5o90Dli47jNRCKM16JNoAKFLiGBwCCdr4BkDbja2jojSgGoVjpjgDqmd4IohYKo0ZqpWZqhY7qJP6IqsYCgcbqL/aDV4YKIKBqqPCBKQaLnvZpMY7oSmaJchAKNfCIglbojQhpLNDpjahnhnDmisYCXeZnoYhku2YIvO4IvcZjPYBmPSBsw0ZoxN4ISsYCQ+aIr7bhjiBsUOYI/sK26LiO4zSma5Qo643+CMJ2ZsgWimHwCD8QinvWgxsQCrWuadjWg9QGCdaGaX+u7dtOad22hLv+CNsW7KvW6I9Q7NEmbTE2bXI+CeTObJBoaSygbdfu7I8Q6dGW6o64bnKmbs32iOvO7sjeC5vmiPDeCLGm7JomZ7Feau5W5OeG4+heiXg2aJDg7D1OY9X2CHPeiIQubI7AbI7Qbyw4aKgWCv8Gbj0gcMo+8I3Qayvehgu/8FU+7W0QivfmCAG37zDGb5Qw7W9u8N/e644AcbHObuPma6Ew8frW7sdm8W7BAsbWbvcg5wRf8HKW8RB/4ut+aZMo8ZDobhxH7w///vHl1gOcZohrznGROHIUf+z0Vm0nxwIoN23eIGfT9vCNAHEgN2IhyOqVKPJiPvEXT3Eld+4EbwM9oOQbVvOm8WQZT3I3h3OGkPMcXwNL8PM///M3n/NIDvAfF/MihorWHokE//JoDmoqB4kwJ3OGuOdHR/AOB4mMZgihnnFsnu305nQ90GtCb2ShEPRCd+5DJ+IEB3D+3ggOb2KQ0HDpTnIwN/QY9oHrPnU9gHUsIGmPmHVI9whc1/W1Zoj5Ngmx/nWOmPRUV3UgLmu/Dol4NuSy3vONqPOXqHUfT/W5ZnKOsHY0h3Fh7whvZ4hW14PifnCR4Oxrj3a4fvYfTnAb/u7mq55rnSbsTKBsSyjjUN/2MN9tsMBtz9bofe/3rq7ygP+KgTdwofiBg16DchjywMYC3W7yr4D3ePdh0W4MzW72JB6Edr72fzd0btfnj/h4lQ55qBj5dbd1qFjwaMeCQP4Gwt5zsWbl7oaKjNf4Hp7j1EDl61YLFCD1Tv8IgOd1qPiBkBB6sCD6nV7zV1f6kBgA3pjqjTB3oSgH/eb5nudhPyBwoeiAb9j1jQABzBZ7hsfncweJwA51bIcK6f6IsweLDlB7Cmd7cWcIufeIYvj5r8iE7O6IXdB5hlDrggZtHl6DXeB7r99tIfADIehxj7CEAbD8RueIQCj6jtB8/pHofJSo/Msfic//5s0PiTaw/AGwBNPvCEtYgwGgZtLfYXWvbse3faNk79vX/ZkU68bf/d8vyN4H/uF3SOEn/uPPROOHalw27DUQZGUg76sGAl+e/I0Qgm1Ig29QBiBQ59gGgiq+hjXwg/Eff5TXZu2vd+RX/5BQfj3Qb2UoBoDGAlIO+1j/CL0G/C/nd97wZoZWiyEHCD0CBxIsaPAgwoQKFzJs6PAhxIgSJ1KsaPEixowaN3LsSHANlpAPCCoLafKkyTYoUJpcVvAHS5QFLcVkGUjgspohy3ns6fMn0KBChxItavQoRJAiSaLscK1kzG/XOphEUXAlFhRh1gwI/nNyG8FMJtUIsfTtZBplAtuc/OanzbaTQJDSrWv3Lt68evduVIpl5ECoVIcN9MOyzUCsWIQQLIcC8ECaIX8MZBsyDUEgKQcOa0mwmMkOfEeTLm36NOrUBf1C1gMVC2GCg07eHJhzKcPZWQc+MFmsoEnAkrGoMXjD5ADVypczb+78uUHWTLGILtg7pKWCoHEv1G1VoOHdwHH7BXEwOPT06tezb99ReuCQmaL7LhgI/cFAlgYMEOIj5Hd6XIeFeQOBgF8aJqWxDYMNXoOfexFKOCGF6sEnEFQBDuTXbwTdx91AYSiGUoCWwdaYSdm5phNKrVX4IowxynjUhSuK91F9/h5CqIcQLN6oB0whOXWNYpRh6OOOMyq5JJNNJrVjhvSF1OFAH/410HYAhrEMV0FqSJVOVL62hiVlmmkmYk6quSabTdYY5Wo5VglhXCGpRVCCNz4oJErF4WlSbW0KOiihML4JoJRYUCmQlZDliYVBeQZ4FhZh6NEGfyrGGdINhXr6KajPHfqjQBzaB6FXpOoh1o3/YZGJpggNp2Gotdp6q12j0qqHqTpy9xpYaykW4GuhdYCCD24R1Nlki/Jazp24UiQEf9Vay5+02Raqa6LONlrYSeVcw2xVBIGp0yCVofSDMt8ESZ22EmWpU6zx2rvkgK3BieOU2u2oG0tgBhjI/rkodHCuggMN4KMP9z60sI+MDbXLN8o6fHFRPYa0xp+vGgQxFoEyahLHBO0plx7yDfSuHx5qhiiWasSUyTUYN+RXBz7ovLMPmYgMFFZ+2jw0UM4y+vNnCQWCtED7DRArlSbuctBrRmOaKdNEb4pFyXax+o3WYcs4INOUVic2RTXSxWq0aLvtnl+Wbj3f2xKprZAQ11SszA1GEyTENml8owwQEmPZH5jlWOKHH4aD8MAy9QokxDIPiFzMMsuYV4zeyrjIo96D91036RRZmdWWDwDxLtelPwliQn6MaFLbAz2KUiYF6nE7Sx3ueba5IYE9EKXKhCen7DHV7jrzCb3M/mLNzd+c5EG31cQTQazHZDuLHUJFd0FfT4dwv3pYHxP20qv/sasslWP4+gndXZCJWLgFF8qJaclVqiEFu8wwAEacQRAwPqpalfCmQ53JUOU39btfnUIyl/hRkFFCeMANbvAA+FUQIX5xiULIBUKBZAl4jnHRcIw0ELZRDWbhS6ABqeIzgRRIhJ8JTQdzqMPpsSg2wxHaQI4TkuR0x4UCEZ9B9hUWGB4JUDMZi3GQs8MpUnFDPvpOec4Du4N454X2a+EBkdhELBCRXwTS4pWqqMYc+qUmsXnUghrEoJN9Tg/64Y9/jIjALyZRj0dkoo12tTsFyXGO1FsjIpnnl07Z/seASHKRiGpCKzFOR5CUVKIjfVTHRHLybfPLJIsgozGdTBKQBrQkIDE5Rk12spWl++QYyXSmM6VJD/PSCpcG4CUvDq8gqtxjLwMJRq7NkpauPKYnD8m9kGStIBGsnaR4CUbwLZGPTRTko5qJzG2KDZa8MgkjF/KoSOmRhX2E1xNT6cdvcoqb7iSdN2fVkFTtilWlPNE5W6cukwTzl/J8J0C7qUzOmOQHzloDtBwZrEsNy4tYSJPh2pimAanzgHogl0FXk9CAclRa3rzUutr1rrMdDwviIpceixUkKqHEB7OzpjAPUj8ssMtdOOwoTmuVL4aArCYNG4gATyKwghyIsCW6a+NJhkGpfq5TID2NyU9zKtVCjbJrCimGzFhCM4OczCRzUVlBBoAwWg1gdnPJk9wEkidqGgSrM4veVONKKL8t5GpP06bToJaQNjBOd/TDo4eQtjSH2NUS2pQrYhOr2MUytrGOfSxkIyvZyVK2spa9LGYzq9nNcraznv0saEMr2tGStrSmPS1qU6va1bK2ta59LWxjK9vZ0ra2tr0tbnOr293ytre+/S1wgyvcogQEACH5BAVkANgALAAAAAAOAoABhwAAAAAADhoAAAAAGS0AAAAiAAA0ADwkAAAALAAlKwA1LQAgPwA1MC0gIEcAAF4AAHIAAEklAFU1AGMjAHkiAFstIABNAC5SAABBJxpULQBDMQBcOBpUMi1UIxpjOTRjMDh3PE9NAFdBAFtBAF9LAFxbAEVYGWRUB1NtAGlgDmR7AGJqGW5wF3h8HVJwKmx8IAAARwAxZi15QABMeTBPfltDTllZWmhIR09OaWVlZWZle31sf5kAAKoAAIQhAJkiAIg3AJAwAJgxAJ48ALUiAMM6AI9NAKVFAK5LAKtfALZYAIlnAKVqALhnAK50ALl1AJBUN7BMIMlUAMVnANlsAMd3ANJyAONyAMpmOYxaS5RpX7ZuX4dgY51ze6Z0d9FwRcF/eHmLAF+RPXuHIGaHMGaXO3KeOzmGQ0aWR1anS1+zTXarRGKzTW+3TGa7T2m7T3y8TGy/UGzCUHTGUHXJUH7LUIKKAJaaAKiKALiJALyYAKilALSsAL2yAIWeKoqnLYijMIysMoarOIirOI+3OIu1P8OOANWAAMiSAMCaAMWeANKeAOiIAOuYAPWdAMWlANypAMy3ANexAN23AN27AOGkAPWjAPuiAPKpAPqpAOO2AOW6APWyAP+2APG9APm6AP+6APu9AP+9AJLCPePAAO3AAPLAAPfAAP/AAIKxQtyJceiHZfCOaPKTavead43JTZLGQJLJQZLNRpHLS4jQTJLQSZLQTIPJUJHNUILQUIvQUI3QUJLQUEVni7aGi4WJo5COpZmSqbCZq7+jsMSIiNqMhcyWntmYmuWcl/ugg/2ojPuomM+cp9ymq8mosditseylpvWoqf2xqeatsfKvsemxsfexsf2xsQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAj+ALEJHEiwoMGDCBMqXMiwocOHECNKnEixosWLGDNq3Mixo8ePIEOKHEmypMmTKFOqXMmypcuXMGPKnEmzps2bOHPq3Mmzp8+fQIMKHUq0qNGjSJMqXcq0qdOnUKNKnUq1qtWrWLNq3cq1q9evYMOKHUu2rNmzaNOqXcu2rdu3cOPKnUu3rt27ePPq3cu3r9+/gAMLHky4sOHDiBMrXsy4sePHkCNLnky5suXLmDNr3sy5s+fPoEOLHk26tOnTqFOrXs26tevXsGPLnk27tu3buHPr3s27t+/fwIMLH068uPHjyJMrX868ufPn0KNLn069uvXr2LNr3869u3eIOb7+ix9PfjOv8+jTq1/Pvr17XmJRyZ9Pv779+/jzoyrf9r3//wDGp9+ABBLIH1sAJqhgegIW6OCD8h241oIUBhgWhBgWKKFaFXboXoMZhmjfhml5aKJ6IIqo4n4knnXii/BduOKMLboIo4kpzphhjWbdiKOMOorIY1k+ephjkA8OSdaJhZRRxhpGAolhJU4sYSUekoSo5Fge2vLCQAno0uGRAzoywUEXUILhlmJ1KAgBBGkgZoVk5teIQgBMAiGbYVEoywoGMTAnhXXe14kDAxmAByJMEGTBnnx+taAsiAY66IKF2qfEQCrQh0mlAPCRZKReLQgLnNiAQMsJAgk6ppT+A4LygEAGeFIfI5yOSipXC8ZCAAZt8GKLCK1eqmCm9N0p0B32ySqQBaE4uGtXFL6Cni0jFPsqWA9OIRAAkdw3xLfhajjtViZiqy2dsOonBK2l3GcFudKeq1W62WLjKrvcOjguNgaYct8h9JprL1b4rktou/n9iwJ+jnzbR70HX5WwvsYmiKx8okAg0MP3RbwsxRVXdfG+C/dL4KEf42fJQGGQXPJUJ2dsocoDdtxyyAPtIfPMUdW87VcP/nsBxAUbCDRVQvNLtL+0CmzfvNiA+/PSTjWd8tMF/lsAKfd5W3W5SmMddJf5oowpw/htWvXE9vkA79VmL6X12jgPqCz+NjHXl0mudNed1MVyDu3VgyxXDTZ9SPSsq+BNVYgLL7jYIoFACog5Od5cF3jEQBdswnETiUptMORMTRoBQ2gci1KELYXyyUEQZoKqQCGccHuokKKeuoKnMnSG6yWh0okU2CjCIkqZFMEDNj/kURCGex/ELH7H86D8fL5HrqAsqy/UusbF48oDD4nAblInPWAzBBLn60FQhpgYYVAJWeLHyPno09f970Xamkg6Ab0m9E99JLkCNpggn/0N4RQDEdEnKrGIRUDCVqjgRH002Ake/MCA6ePe/5QSQKeJRBSHMMUlDrg8kjgvf6LoAQ8mEUEkrZCB8qECDx6BQhWysIX+IzRKCQUYkvk4goUn6cEP4iUf51GihkHqBBF4wEAd8kATRvxhEAc3ROKJhD5HDCEQQ8K+B87HeTQUCJLkA4op8g+LWRTjGLcolC56sYhxFCFJQNGDIZhOCleEIpL4eD41gVGLdBSiHW/2xTwicIA9EAIGUYFGQQbJijxwQn3C6L9EFmWRjGwkKjj5SJE4cT4y3IQlZ6TDH1RiijiUDylL6cmfgPI/JdHDfFa4vTl+JBGA1CQq9me/VaroiD8QHft4oEv58LKTtazjLd9jElFIIQnwO4ISmgBBk6yQB0lQwvmUZ0wRoRCOxquCJ6yJTR5ok5u+jOZOpknNknyifQP+4YEQTHGSYRLEEOpbY37uyb/z7TOe8swJPT/0OlqahBOcoJ1A15RQoCy0PRubKDQrytGOevSjD7koezKqUYcq5hooTalKV8rSlrr0pdeAjEjXQ9KSSgamOM2pTmU6UwaxraT3ualOh0rUlPK0p+epqUaFWtSm4vSoSFXqRJnq1KqyFKo9lapAqWrVrsb0MUhFj1bXyFWvVhWrMx0rkspq1qaiVaRqDRJb20rUt140rjqaK113CtYK2SIVZHCBYMUAB8N1ZUWYyEMeeklRyFQVGWDQAhe40AVjUOOsfV3QIA5wkAzUgYhcUdETBlIAJjb2MU1VxQQAwNrWAmABwnD+ql3dAwiFIOCzdzxshhoxq9JpKTJFxYJrh8va2BZ1tuwZxe0wUAYzuIAgG+Dc4SAkiiQYJGC/dexQpSGE1trAF6qAgmtjUI3jZhZAKRgICNJDCFQFwA25De2DdHZd0/UOtdv9AQB00AxqXPYaq2gtApxhXscoyFcCKVx6xjCQ4ZEvbwNKHApK0TiA2fdxjikqMpjR0ii0NhgFboyC/jCQNKwHwdhQMC5/eh9QOMAAj5DPv7C7I+C29Qut/UWIGaMgFghkAHNYjy1I8OMgh1K3D4rWfGZ84cAlhq7CBQAChrHjxSTIcgpLTwuK/ODOZYjJ2cWvV5XxANbC4BlVVsz+lfOl4gVz+cjyFRGYa6zdri4DCK3thVvP+x51YUMG7QnEj+PQ5enKOWphznBRpVEMXxADDFlgbQMAQN49G/g/tqCUQADNHkELxMRw3sqK5nxaRRN1GWUm7gysIVs+uyd4fw70QMa3YghDiNT3NfVQleEA4rIWB2hOc2ISpOlYd3ogaig0kkOEaww3pqnSAIMXpp2FA7gWBgSuq6vb42cPtIfE2BgAHZQdZ2Yjms5i7qoqCNDaGfyXr5fGdNpstuVwG7nWXsZQs52MGLq2QsBUHipy0zMsgSQgF+zxsb3JLWoV7ft0uvaqh1kLDG3H+z/pDbcchBy+NtfT1kU7d6n+n01XHLNWxwLfdnv8MOv1EKLB8W34oS2caJK3deIDMC68RZwg5QoEAQhPD6sEwoaYa2XUIs+1zXO6DC9Mo6WuEHC2d87jKxNZIBy4xXlmofAU2wyjLMZPtESBiiDMjexKZwxRWSHlHTRDparoNWtpwOqUX/w/hBAAQTpgAr3/+A3SJYnsXEJL6lKAIXYQ+yfqo1eWsr21CajBDeTO2gELGzEUAjdCQG10jzTvedFTifGQR875Ia5SCumU35znwTzoMd04TQa7fW1mndud5wtKhQkM8oHCgvYj7HMf/Jj5OvMd0PQOEsXhF9K3+SzzffEz6ZNPvYUKEHcBO3i6pXH+T6FZpGINa4CDtQwLEgXi0IHdLAkBP4hENa5Rg/TR4BWo2EAemLHxL40GMo7Bf2h0deC3ZBIvxDEylEYl0UMrJEblFCI3NB869AgDiAoxNEMI1W97hVl3l1YloUSmVUkm4UjLs0ZSRH9WpAkceEYUiH8X6FIACErq10engIIGOBKHpIDuN0hudD6aUEYx2EQpaGMreHmHEVZJtUd99EeBVDwguIAiQkg8oCZ85EfzAUhYpIJBuFItuEgveFA+OIOiNErthw0ThUmaxD5cSEk/WGdXSHVWRoQCSIHykUr9JEthKFCt9Er0h0aoxAOiY4VriFJZaEclAUyZVH/FlEv+u6Q9erRGyKRMMqQHwWSI0meBf5hTgdhF3nQ+4TROFcgR7JRN25R+a3ROzqdO37SJitiJhlGJbKhmblg+/zSJHHFPBKFP/HSDQFV//ANQsriKrPhUKrdQKQFRhFdKuWgfEMV4QPiLLBiM9IRXNLKMzHhVIFWN1niNZxNW0Lgi2HgTRBgjIHeMG9WNMvGN26gi5FgT5hh24piONLGO4SiOveiOJgGP+SaP40iPLWGPhoaPI6KPMMGPy+aP+QiQKiGQ5UaQr2eQK4GQMqeQC8mQKdEhfxVYg+V7v3d0KpJYiyUkEskSFbJZnYVbnYcVokVappV2H1kSFFJbCXFbgZf+kA7CWwRBYyO3kiypID4nEMzlXNAVkw9ZINVVXzWHkySxIBmXKuzlXvDFcBqZfB5DlOhmlDkJICjmcQwmEA4Wak9ZIBJGYUnnbFQpEiNWYuqRaaiiAbvglFkBIS4GYzIWlvw2lhzRY29GcFcHZGxpkhCiZHFJc1NJlyFxZZeDMexRb3rJlW3pcHIJcYIJEmuWYDaTlYmJb/34ZY1ZNo85mADiZ5y2Hp42AISmmHw5czapkpu5EcRWKZ+pHp6GDZznH3H1cJqZmh7Rc6jSmunxmrQmm+yoH7Q5ILbJmf9RbLqJHq+ZbKR5FUgHmDc5nBkRmdjgbewBbuK2l8zJmM7+iZrQaRHSqTbogZj35pvxSCDBqR/deZudSSzYcHAJd5fkeY8ht51imZ7emSAZNwAbpx610HFfR1O/2TCZKZz2qZoJwnIC0Zvn8XJaWZLZaZpN5pgFWhE6iSpApx5Dhw1Fh51W0ZynWZ8TKhEKMmQDkXVb13Vq6aAdiiFjV3Zn95whGhELknd713cDMQCAp6IcMY8gMXgGYXiIB6IxGlILonkHEZvLuRGjlzyq2BGfBz3Sg3xeiXoJoXr1kT2MNaQXQSF/tXsF0XsmVETGF0IoEXzQR3xMaB/KxxDNNx/7cz5ypKX3WSHeB37iFyUksX4gFJEhYX7190BpiiHwNx/+HORBe8qncuoQDukRCBiGLsQDMFSAgfogDZhDO9SocZqoE7GoHbGEJnGCXTipDjKCVaSDINikmloQnKqknpqnMCiDouogbVRQ6ASGmZqqELGqGlGDiOoRUYiEmhCrsipDT7hJiISrRKqNJMGrPHoRZmg6HiiGdsg/wnSqyJqrr7isdGiDJaGHcciHwkogdwhLzNqsyKqrGdFMqPBM5loRhChMxCSlM9KIxvOIiZil16qo2ToSn+hOoXgSpyhOihiu+kGK8tEJ6tSv7xSD+aqvykoStJhP+9RPuDIQvEiwGUJQBXVQDdsQ6Lqj7boRxPijEFmQHauq+3qZEHmyLNv+si77sjAbszI7szRbszZ7szibszq7szzbsz77s0AbtEI7tERbtEZ7tEibtEq7tEzbtE77tFAbtVI7tVRbtVZ7tVibtVq7tVzbtV77tWAbtmI7tmRbtmZ7tmibtmq7tmzbtm77tnAbt3I7t3Rbt3Z7t3ibt3q7t3zbt377t4AbuII7uIRbuIZ7uIibuIq7uIzbuI77uJAbuZI7uZRbuZZ7uZibuZq7uZzbuZ77uaAbuqI7uqRbuqZ7uqibuqq7uqzbuq77urAbu7I7u7Rbu7Z7u7ibu7q7u7zbu777u8AbvMI7vMRbvMZ7vMibvMq7vMzbvM77vNAbvdI7vdRbvdZce73Ym73au73c273e+73gG77iO77kW77me77om77qu77s277u+77wG7/yO7/0W7/2e7/4m7/6u7/827/++78AHMACPMAEXMAGfMAInMAKvMAM3MAO/MAQHMGlERAAIfkEBWQAxgAsAAAAAA4CgAGHAAAAAAAOGgAAAAAZLQAAACIAADQAPCQAAAAsACUrADUtADUwICAgRwAAXgAAcgAASSUAVTUAYyMAeSIAAE0ALlIAAEEnGlQtAEMxAFw4GlQyLVQjGmM5NGMwOHc8T00AV0EAW0EAX0sAXFsARVgZZFQHU20AaWAOZHsAYmoZbnAXeHwdUnAqbHwgAABHACRaADFmLXlAME9+W0NOaEhHmQAAqgAAhCEAmSIAkDAAmDEAnjwAtSIAwzoAj00ApUUArksAq18AtlgAiWcApWoAuGcArnQAuXUAsEwgyVQAxWcA2WwAx3cA0nIA43IAymY5lGlf0XBFeYsAX5E9e4cgZocwZpc7cp47OYZDRpZHVqdLX7NNdqtEYrNNb7dMZrtPabtPfLxMbL9QbMJQdMZQdclQfstQgooAlpoAqIoAuIkAvJgAqKUAtKwAvbIAgJEjhZ4qiqctiKMwjKwyhqs4iKs4j7c4i7U/w44A1YAAyJIAwJoAxZ4A0p4A6IgA65gA9Z0AxaUA3KkAzLcA17EA3bcA3bsA4aQA9aMA+6IA8qkA+qkA47YA5boA9bIA/7YA8b0A+boA/7oA+70A/70AksI948AA7cAA8sAA98AA/8AAgrFC7Itf8I5o8pNq95p3jclNksZAkslBks1GkctLiNBMktBJktBMg8lQkc1QgtBQi9BQjdBQktBQLV+LRWeLaXWThYmjkI6lmZKpsJmrv6OwxIiIzJae2Zia5ZyX+6CD/aiM+6iY/bCaz5ynyaix9aip/bGp8q+x6bGx97Gx/bGxAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACP4AjQkcSLCgwYMIEypcyLChw4cQI0qcSLGixYsYM2rcyLGjx48gQ4ocSbKkyZMoU6pcybKly5cwY8qcSbOmzZs4c+rcybOnz59AgwodSrSo0aNIkypdyrSp06dQo0qdSrWq1atYs2rdyrWr169gw4odS7as2bNo06pdy7at27dw48qdS7eu3bt48+rdy7ev37+AAwseTLiw4cOIEytezLix48eQI0ueTLmy5cuYM2vezLmz58+gQ4seTbq06dOoU6tezbq169ewY8ueTbu27du4c+vezbu379/AgwsfTry48ePIkytfzry58+fQo0ufTr269evYs2vfzr279+8iGf6AH08+d6vz6NOrX8++vftWYjXJn0+/vv37+PNrKt/2vf//AMan34AEEsgfWwAmqGB6Ahbo4IPyHbjWghQGGBaEGBYooVoVduhegxmGaN+GaXloonogiqjifiSedeKL8F244owtugijiSnOmGGNZt2Io4w6ishjWT56mGOQDw5J1ol3WGEFF0YCieEhRgxhZRqEhKjkWB6e0sJACazS4ZEDAiLBQRUYguGWYnVIBwEEYSBmhWTm94dCAKiZJJtgUThKCgYtMCeFdd73SAMDGZCGHkQQRAGEfPap4CiIBjrogoXaJ8RAKNCnSKUAtLFnpF0tGAqcxnhQSgkCCTqmlP4DSuKAQAZAUp8fnI5K6lYLikKABV60cgoIrV6qYKb03SkQGvbJKhAFkzi4a6kLgoLeKSEU+ypYDyohEACD3LfDt+FqOC2vXWZrjKt0wqqfDrRecl8T5Ep7rlYmYqttu9w6OK4xBmByXx71mnsvVvmqyy6h7ub3rwn4AfKtG/YefFXC+zLcL4GUPCAQxPdJvGzFFleF8brGJoisfId+jB8iA0lBcslTnbwwpg3f17HLIQ/Exsw0R2VzyhZuTOC/FURcsIFBUzX0tl89+G/A8y49YNNOp5sxzkYP+G8BltznrTHgAo11U0/zG7WDm5JNsX03xGv22UulrfHaBSprjP7M9S2S69x0J3WynFB79WDLZIdNHxA+6xo4UxWm0koqp0QgkAJiSs413gX+MFAFjchHSRGJCgz440YtOAoEDGVxLEoRtjRJJAdBuAiqAn1QAu6hQoo62gqeyhAWr5ekySNJGMMHiygt0kMNxuCgRkEY6n0Qs/ghX8Py8/0OfIKrt178SJrgWkMNe8Ru0iM2GLMDEOevQVCGivhg0AhZ4ufH+ejT5z3kRVKbSB4RvSL0T30kcYIxiCCf/e0gEwMRUSQO0Yc+CMJWmnBEfTT4iBrgwIDp697/lBJAAYaEEnnARCIOyDySPC9/lLBBDQoRQSStkIHyWUINAoFCFbKwhf4jTF0JNyeS+QCChSexAQ7kJZ/n0VAgSDoeD2rAQB3WgBFG/GEQBTfE8YWEPkcMIRBDwr4HzseJNUSSJKbIPyxmUYxj3KJQuuhFkIARiSWRhA12YDpNJOGKaVSjDGugpzf6T45CpON/jGdIBA7QBjroIxqhGEVNWLEGRqhPGA+JSKIosmjkk88mHSkSJ85Hho0IpI50iINDTBGHotRiJz35Sf+UZA3zWSH3TrKHP2ayfDWwnypXdEQchI59NcClfHTJyVkGpZa2LAklkhAE+P1ACEWAoElWWIMgCOF8yxumilDoxuMxARLTrGYNrpnNODqTJ9B8j0ki0b6B1CCSJ/4pH0HwoL5K4oee/DtfJN35Tp3E80OwI2hJHOGI2vnTdwX9yUHbs7KHijCiGM2oRjcKkYmyp6IWVShiikHSkpr0pChNqUpXWgzIeHQ9ILWoZFhK05ra1KUvZVDOQjqiyNj0p0A1KU5zep6YPnSmQU1qTYdKVKP6E6lKjWpKmZpTp1YSqlLNKkmp+lKrRhGrWpUqVz3qVSSBNaxKHetEyxqks6I1qGo9KFt15Na3/jSuAGrSk6LUtQdRyUpDwJKWfKpVXtjisLcgRlLx+h4vgYloi9xpfsyEpkKebqRZ7cUDAMBZAMhisY9xE+6MQTgTcqV6eLIs0yCj1Sd0lrOfhf5raFUHqILcrI6nPVylALaoRg3kUY5rTFZ1QYDXeha0jlHdbglyW5VJVlN/k8+nviWqyx5GqsH4gXGPK9vkBg9VqmIVygrXlQc5C2AYnA+uBNIp6xpGqpzgLAKgUFzuApWx6vEVsIRFrPGadisP0hv26HNeaLm3MFHVLGdd8YnOxva+s1WQtc6jL//ezXAOGlvZ7POvDa/2MVGNAmdd8AtPOBi5jbEbEcvrIHgBjIn1oRfZyvVhxyhVFw3gLCyKYWLYopgxKsYtgP1Fqz7Sh2AzPjBhkppdzr5gGDw+cXdTrDULrzi3BXqY0tym5MEktROdnQVJe2zfu0a4QhVuLv4oMcwxjxkDZPYRmTEGfDXC/rQXN+CsDBQbZR9PGchVVnNk+5q9SsG5PjATCN9qLFygihgACKhFScn8YDN7t0Nphmw0Ca0zNx8ajI0zGIh/uoscA2DHk5YyhC+NZoVpWp7PpQ/StuzhOrP2p0jgLAygnGrOivnPiwmyczktriJXLcmitrFNG8zZV+QCF9D2xaNP/WxhWJrKmHY1ebF8NIGATWxW009dUQrm7Zqbs7G4NqCzvTUha+VBbQPA2+oTtxd3WTA/Lfe5t5vum56ZQpne9pDzFjP7+I29wWXMT3VBg4Y73OGmhjQNZkALdQe7yqW9MIsLhDgAKG4+jBPIz/7uHRi0kvnXq8a2giRHOcsZA3OT0zi3CeQ5gYBOdKQzNskBY3JVW1wxqmPdQlw3bJPMziWkNIbtRqs73lX3PrOrz7iBSukfXzxBwlsI8Yo+EudBT3oqOV7ywjk/CFnPIHSmj9c9qIaLjlqr+q60v1n9n/AN3d0dYZ/74JdM2JnvgGWHUP3ulz/7IPN98Uv6YtD6ifpWHNhAJ2pRS6JAHDpQmyUh4AfxqPQQTbCCF5SPBumjQSdQsYE1MOPUgRoMYFjb6pGXvEleKDoZPlGaKVyhGMUpohvOR4eBoL0mYjhDkV7XrlHF7ydNokQYT5KRmhgl86L4iFda8nyMaP4Zi/6/euSrVPmKzPweM7H924cylrunpBrZiP0ykr+J3Lez91O+7qaSRI98nM8fGWGSO6a/81GkR+ejJvjXR/tnfO81f5CXGJI3eSPBPgMFf+ZXRI00fZV0SZkEgZIUf7emgD/HgA04e8UnH6iUT+h3URfoQa50eqZEgjUQOt3ngSUFfnRUEr2ESagnTP23TNuDQFFUTMckQ2vgSzmoeIohgx+IGA0YIyTBTd4ETgioEelkTdiEeVFETvPxCOfkhN/Ug1G4ZEi4VP/WVcazXgLBT1+IEfREEPeECbznT/vDP2iYhvgWhjRFg12UEgyFdHHEU/bBUFInf3Y4VWNIVrHmh/4xiIQctYiM2IhSsYRzRSOOaBOQeIg8NYmUGIKWGFKYWBOVSGx+6HadCBOfyDmheB+jOBOlyGaniIqpGBOruHGt2FOv+BKxOHOzKIq1uBK3OHC52Ey7yIuaCIqnGIwu4SGbUAUssIxTEAYC924qoghqoAbcsyPGyBIVUgcHcBAXYAYy54sYcgQDUQAwBlHXmBIUIgcKgQDeiHdYESJ/MCulM1jniI4KUgm4YwFWcAUsQBAZcGXgyDFBYBBUY431iBIKcgonMBAekB52gCoB8AXueBUQsjMEaWQJd5AioSC+IhAZdx5vMBBbt2aySCCIYwKXEHIFuSYaeRIKEgcDof4F69GRpMUKXFeSsdIABhAI8jE1GLlzLUkRCqICAjEAZLAeFWaUN4mLAxIt8+GT9BiUI5EgldNu57ECRXmUJMmUUqNzLCmVUwkgFfaR6EEFWbmUAQkhUGmQYBkSVKkuMdAec1CUYoCW0Cgia/mVbQkS4FMpcckecykQMrmVadmVAPOTybaXHYF1qPKX6xGYxkB0g2aKGJKX5qiYiwkglCIQjqkekLkFdpkVK2KZGYmZGPGWAsEB7QGTxjAAZRCa76gipAmUprkQqGll6IGVramVk8mKGTKbiVmbpymW/ZUAqsAeRLmbsEmRsumVlymcw/kfClmUY7AepiB0ZIlQxP7oNc5ZmtA5EQoCBwMhmehhByI5kVYxmt1Jm99pEPeIKghwnOkhXsbQBehZFep5mFHZnhaRkCIwEBqACudBCslJWq+GIps4H9FCCZqQA3LDoM/Jn+CpIHYgAASxASRgoUUJBgD5RXTYEUdnEBU5AQxxBt4poQ5BIeKZEIN5nxkhdsrzoRjhddEzPYHHccuFEO1VH9pTjTKKogRRIXdAAgbRAc74jR6hT+cTQiihd4jXd2+oMyS6EIumXvwDR0Danx1CCpvABVwQBhP2Xx6heSCkiyFReaj3QFGaIaM3HxzkQWVqplmaosM4Ej2ke3L6EcJHfLcXRb6XQzt0p7I0p/4S0YtJeoJG6BHaJ4FrCiHVd3pWVE7S96Nzaqgc4X95mnfjV36NCiFrFFDlFH2DSqgdVaceiqiUGhEFqH+ApH6CNICaNKqkSqeyRxKYmqgcoYGc6qpBgoGxiqWzWqq1aquoehItqAklyKszwkorCEuiCqzBSqv2NxLKpAnMlKoRcYO/tD86CIA6AoTHI4S55IW4Gq1BaqonRE1UmE0nwYVQ2KkOgoUsc05TuE5ViK0oaqkcsYb2hE/9Z4bGgIbwuiIAFVADZa7SWlUJ5RJ7KKK/GIgIa5voypWzGLEWe7EYm7Eau7Ec27Ee+7EgG7IiO7IkW7Ime7Iom7Iqu7Is2/6yLvuyMBuzMjuzNFuzNnuzOJuzOruzPNuzPvuzQBu0Qju0RFu0Rnu0SJu0Sru0TNu0Tvu0UBu1Uju1VFu1Vnu1WJu1Wru1XNu1Xvu1YBu2Yju2ZFu2Znu2aJu2aru2bNu2bvu2cBu3cju3dFu3dnu3eJu3eru3fNu3fvu3gBu4gju4hFu4hnu4iJu4iru4jNu4jvu4kBu5kju5lFu5lnu5mJu5mru5nNu5nvu5oBu6oju6pFu6pnu6qJu6qru6rNu6rvu6sBu7sju7tFu7tnu7uJu7uru7vNu7vvu7wBu8wju8xFu8xnu8yJu8yru8zNu8zvu80Bu90ju91Fu91jt7vdibvdq7vdzbvd77veAbvuI7vuRbvuZ7vuibvuq7vuzbvu77vvAbv/I7v/Rbv/Z7v/ibv/q7vy0bEAAh+QQFZADFACwAAAAADgKAAYcAAAAAAA4aAAAAABktAAAAIgAANAA8JAAAACwAJSsANS0ANTAvLy9HAABeAAByAABJJQBVNQBjIwB5IgAATQAuUgAAQScaVC0AQzEAXDgaVDItVCMaYzk0YzA4dzxPTQBXQQBbQQBfSwBcWwBFWBlkVAdTbQBpYA5kewBiahlucBd4fB1ScCpsfCAAAEcAJFoAMWYteUAATHktR2swT35lZWWZAACqAACEIQCZIgCQMACYMQCePAC1IgDDOgCPTQClRQCuSwCrXwC2WACJZwClagC4ZwCudAC5dQDJVADFZwDZbADHdwDScgDjcgDKZjmUaV+HYGOdc3vBf3h5iwBfkT17hyBmhzBmlztynjs5hkNGlkdWp0tfs012q0Ris01vt0xmu09pu098vExsv1BswlB0xlB1yVB+y1CCigCWmgCoigC4iQC8mACopQC0rAC9sgCAkSOFniqKpy2IozCMrDKGqziIqziPtziLtT/DjgDVgADIkgDAmgDFngDSngDoiADrmAD1nQDFpQDcqQDMtwDXsQDdtwDduwDhpAD1owD7ogDyqQD6qQDjtgDlugD1sgD/tgDxvQD5ugD/ugD7vQD/vQCSwj3jwADtwADywAD3wAD/wACCsULsi1/ciXHoh2Xwjmj3mneNyU2SxkCSyUGSzUaRy0uI0EyS0EyS0E+DyVCRzVCC0FCL0FCN0FCS0FBFZ4tpdZO2houFiaOQjqWZkqmwmau/o7DajIXMlp77oIP9qIz9sJrcpqvspab9sanyr7HpsbH3sbH9sbEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAI/gCLCRxIsKDBgwgTKlzIsKHDhxAjSpxIsaLFixgzatzIsaPHjyBDihxJsqTJkyhTqlzJsqXLlzBjypxJs6bNmzhz6tzJs6fPn0CDCh1KtKjRo0iTKl3KtKnTp1CjSp1KtarVq1izat3KtavXr2DDih1LtqzZs2jTql3Ltq3bt3Djyp1Lt67du3jz6t3Lt6/fv4ADCx5MuLDhw4gTK17MuLHjx5AjS55MubLly5gza97MubPnz6BDix5NurTp06hTq17NurXr17Bjy55Nu7bt27hz697Nu7fv38CDCx9OvLjx48iTK1/OvLnz59CjS59Ovbr169iza9/Ovbv37yJr/oAfT768QVjo06tfz769+/ewBMKfT7++/Pr486c3z1m/f/vF/Cfge/cNaOB+/Gl24ILxBciggQU++F+CCkooYIQW5odhhgBSeBmH/m0IInwijuieh5iZiF+JKq7HYosIolgZjPO9SKONMMpoGY0kOshjezi2qOOMPwLpY5HqBanikJQhyZ6SJkI5IpOTGZhKJ1hg4cWAUqrXyRUshFnFGBceqd+XYbIwZplUQjZgKi0MlIArZf53BwQHXYBGiGbWd8cBee6pX5uRCWgHAQRhQOeEfc43h0IICLpio/DRAamkHRLamH+mpGDQAovyqd8liApkARZZsEBQBoNS6h6p/gOdmuqqrWq6aX6mNHAQqHXmd8JAHqiHR6kBhKGhq+39KlCw6Q0rULHH2sqYfqSU6gEqJQjEK6P5lVKqouvFMZAWx3b7bajoiSsQuZNKu5h+3loABiypgKAtuuXi96hAXLDnrUAYvDJpfvsW0+96/xYTcLvuJubfKOmlEsK9vdangkADmMGexBhrDGB+FxeT8cYTi+wxfQ2/+2bJ24paXyoRUMzeCh0P/HLMxbSsHs0mM5yyYVayjK/N9HGs8NCwWFHzxy+XDC57Svec6c+FBS2zy0WXHIN7dWBcBitMZy3Q1u11LTIZPlM9mNU5I40yshHnOjbXAx38Nn5yF0M2/ntmG5y22oGxrXO+9FU7d9kDbUH0fIbrTbdAik8NeOArX90q3ro6jrhAXywOX957r9d3F39P7pfgbtcId8Qlc+DevgOc4fl7RrveHuyyS2766ZW3XfF8Rg+OHs8jhw280O0Rf7Lqu1MuYPCp94hfvQIl0Ep7IRd/98v2FmM99ku/3TxgqP8+n7IDlLExnkfP7l4q6Ku/XirsPy3++Lw/jzy3+MmROHt4GBfh6OM/yAFQgKXDH17YZr/L4QdWxUDA9dSTLc4NcD4QlKB6XlHBYnQugQq0y3/AxgqYCUQBdAKbA18mgoFoYBXoOUXI2ue+97VQIC+M4QwXBsIQ0kU//qZgn0IiZzz64EEABNkACZCIMTGssGhHTOISBzIAJ0bLh3qhVqkUwq4iElAhdqvh+yyVkDDqDot1AaIQE0LE7eknDyQwSAfIhLX8dCKOBZkjo9CYFye5aHXuMUUnvOCFMUDMfPk5xSALecg98vEufkwSIJc0ySg9coGRRE+XOLTJDF0Sk5nspIVEKaFPQjKTDUJlKlFpShGyspJTgiWIWpnGV6qSlA+i5Q9tyctQ6lIut5QlJ4XpyV/GJZjIfKUx35JMXwZzmW5pZiRxySBoWvOa2LQVJ7bJzW5685vgDKc4OSGQcZrznOiUiTT9mM2FoPOd8OxmOeNJz3Oqs5fT/mxnQurJT3HOs58A5eY9nalMfRokoAglZzESGtCB5vOZBi0IQwH6z4nS06HsJOYoI3pQi9azoh59J0adRM0FcbSjIYUnSFNqzpEiqaQHOqlEWSrShdI0nTFZJ0llOpCb4tSn9swpPg+kQj7x1KZA9SdSkwpOlxYJprAoaq142s9EHIEIWF2DIfi50ngygg1s8EM/nfoyLGmJS8S8EpjChAU6GpWq9BSEBA5SAURcdKn0RMJACpAJrgr1P3CSU/QIJMw/BeqtMqVnIBQCALuqFK/vDIQDCGIATfgVJobaIg3rWB8yIiRSU00sPCORuWIYYA18KAJBKBDPro7TEkIw/khlL/sSTnmqIMLzInwgWAxZqWogrLqiaN85hIGgoJuMyBwA3vDYeFriAQeZ7Uf/Wp+84XawJ4KlsorBLPQ4qxjQ6mE24UmJyZpWEt4EhHGbO9rMmSATQRCIdO+KWXMtC1uWuyB8EsbD9KirGF2833DPuViBqOGb5RUIBSpR03hSogEGGMQ2eSBfy063vviJ17xgsT/OwqdgZoQFfwV2RoPCUwkCAUAhwEnhYqi4wfFkMDdbPN/WUrc+p2BdfsXInuwtDz1G0x7z4IrOHci3r99sQopX/FOA0tjC9K1t7xZAYg+/D2e5VZ54sQnPJ4NzD0uGcT+9fGEp62/Hum2P/tEamJ6oCblHR+2yQEwQTkGkGA5i5ieZo+yS8vFPbJrjG8bQVmJ9vvO5c67zQA7c5DFXmLZ9njJ2jYS5wwmaX1u+5jtJm2hwKmIgVMhzPfdsYwwDtsNPhE/jQqeevrURzkQ2J6KLQWdw2lkgbhA1PUn9WFP7B3qIDGTmWJ2e0WXamnIuRgUU7WImBzWgvK6pr/UD7D8fTyC2Yw/ujg3NZNeYm0putq7jGW2cTjs/1baymlGdHi0Xup3JLgAmwIlicTdaz48uc6TP7Dtrw4d63pvgenzM7WXCs7guxvM3cXDkcXsbyqU286nR7EbgxW99AJt0g+IcWVB/sxHrdfg7/ssd1HNPj908Xk8Bi/FqWARwXcI9aXtTPO9uxhfX7HW0aSHea4n/2mka/2N+MihwWHBwIB9893jhCYSBVOAR27SEEQby7XG6dpwMtgQndNBwrUd83/khoQmLgcKoqlvNNyxGDmEhw4H0V8Ayh2cjNPuBEmxxuTZ27gQYkoZwVmIS3kzJNg+iRoa0nLDojuJAlMhEkVmx4MZUrEIYDc5IJMEGYhUoZMfJ6YUc15uN8IENbJADNmi+JJywfDHEWhAtMiTAqqO2Zw8S4iEP+J2M+IFBRrDVcAJi9Dbogzw3L05L7H0hoe5mJG5gAx4EYfRtGDzq1Tt64StUSkE0vH7p/nNHObo15hzt5yQS8Yc/EAK9nICEN9UfCdIbIfjD96j6u6l+J9igCNv8PQ82cf2RRKIYOfB+1nd9OoUfikRIhoRWHMdQi3B/3LQENjAIlrAHmtCAA6hQFtWA+LdNEDgIotd7lsB8h9B/IjGBFQh/0qdRpaSCuXRUxMdPkdADDgiBNuAI3CQIKCh9FhWDMzh6jnADOYBknCB6I2gSN5iDA1GAT+WCHkUJMgh8NniEF3h1/OSEwFeDy7d/3ESEJAgS3YSDA5iEQ7VTCzhRlMB8NuBYUhh/TYiGdnWGPABll2eDqLeGKaiEP+KCL9hPNGgDR+BNOBh9mhdSffiHy7cD/lDGhXW4TWB4enjII3roURCYA4kggxvIiJjHhhM1iZXogFy4TcwHdUaIiWEIVWgFUbcXUDiYA1C3fDYgiJzQgJmngxO1iq3IfG1weX/ICb/3AykoErAoi7/4iDfChBM1gVGYekwgCZaQBELwfEAwBEbAf1RIT8jITZGwjA1oA0IwBKOXeSbRjM9oA9E4jWJIUL5UhjQ1CWgIfIiIgUn1e8CnB6c3EuxIEDaAiOf4UAWVikw1iP8ICfNXj3VIeGP4Usb4j99UjUBFVpDIgtWkhxI5kRRZGwrZVHuYVA5ZjKjYEsTwkSAZkiI5kiRZkiZJDMBxkQuZkQ15YzvVkSxx/pIyOZM0mZIqqYk32YUsQYw5EhM0+ZNAGZI2mZPwSJQb2ZMwuRJBuZQ1+RtESYtG6ZIImZQqwZRWaZJDmZMM6VNHKSQQaVI+eZViKZRO+ZRbeVNdSUlUmRJj2ZYoWZZR+ZQ6uRI86ZUw4ZZjmZU3eZY0lZaWtJYogZdiqZcqyZcs5ZexBJgnIZhXSZjvZFVYRQRa5VdVdVVZ1Xt81hIHkgdZsiXBNj1m5ZmO5BJW+QtTAAVREAVSoAvDYJWOaU5yRVdqKHLiFJsGUVeZuZPPEyfVE3SS9Ca86T2+GSMvsZSfIAEAkJzKCQAvgAtM+ZriVGAI0Vh5F0/SeRDU2XNg/pcfh5Iow6lJxNSdbldloeWRQPkEy5meADAAt7CU0Fl5pXVaqbVa1blp8YlaqjUQrKWdmglEt0UQuVVx1fWfA0FlbFKcPwkMO6CcDBAFUAABywkDwhCU7/lNCFcMn7dNyZVizEWbFhpyGqpcHWpuPjcf1gWg37lKlfYp5Al+SpmgOQAAM6ALIAkM6KmctkChcHlOCXZe6QWiz/ZOPWoA6MdN6iUQGVpyJapq1oJf/XZ27ZEw11JBAWp75vmTuTALI6mgylkLOuob8CSdlMdNPbpgHtpNYopg5mWmJLqd86Fh9NI9VSo95iIv9IJySmcSeHmjACALX9obJxZm39Ri/i92b+ZUb4XqTYTqbC0llfDRSOmWavSRY0CGp3AXk27Jp14KlBXaTUZmWkLYTeGWqI0KT59qAKEKboKqpG6KbpYae5J2oKSZqV36p7zxcF+2qqU6cvnmTWBmb6zan/w2p4gXq6N5pWLZCziQnAjgnJy6o+bUYrX2TbcGAAoXpOgkrcxmrdK2pCdHcbA6rCkqE27pCcopoba6G4cGXbTGbMUwplbHklHHrtMKiIvWra36rU8qqa4KrrA2q8m6rMlJC+4JreLUefXaTZ8mEMmHreaEsOG0sMXQsI1qck3jr8UqrrKKrFbJpzIQDAULpug0awl7gwORa4ZafPTqrigb/qy6qbH+Rm0s06KQBxJjCQrK2azPabDi1GLLZmu6Gq/J9rPUGrRKZbGARqzZ5WeIxbFBKQoEUKs7K7LZ2quiarThxJflNqqMGq9Ie237un36qrRGcpdWuQsNoJx+6po8G04tJm/0hrUYGW81502I2rVH660XG7Yp92+v+q9OS5O84ADKGQuD2bbghHDcunANl7LipLjX2k0MB6r4KqwTx7dptrdk+yRhGZS88ADKSQMg25iI+03SSbHbBHJImnME5nGgB6QVq7dJO64qGKkuWpVBqayh25qHS7Xn1HkAULfbdHPF0LIO+7DKJbycQLzGG7v5urds1re0A3Qbi6k//tkLOhC6o9u7gMp0Tgd1nCB1VMdzQuu9AvF0UTd1Vuu8losfYoczZSdVmfse73tCKdS01juTwAAEyukCvvALABzAAJyuuhFPc0cQdXd3I/pNfxd48spNBzwQCTwQeMefI/F3Bok3a4QQh7e0uLLBB9HBZYugNCkK6nnCAECwP9mp3nSdBgGvECx6pGd6UNlxCQHD2Hh5sziXHxF6NgCAbNB6+dE4XCS27kHECQF7gJu/MhkKKKyea9uUvot7ulcQvBdOruh80FfDVLx7mPlN8piDPNwRy1cMWvyKBFF4CyHCnKvGQ3S7bPmTg/vEy7mpUty99TR+5Xd+2zSQfcwJ/va3gfpHjQ/8TXpsfujnx+mXeu4nxibhBMUgyM23CfuYUYqpp4w5tXicUBr4gBH4gVEngkXJyQ7IgRFoghY4fCQByuErgpX8kv0YuJkskyy8aZbICTT4g0G4hTZQhIX8u7ecy3b4iyIBhEIoeojwylMZy0w8yzNZy0L6hD6Yhfy3TYr4yzwqzTX4hY5MEtTMy4egzEt4ySXhzAScGyF1hqP3hjcQh9w0h6PMUOqchoDYzSMBh3JYg+Kch18ZU51rzs9cuvxUiKl3A+9ozfRsmOhE0Nw8hd5s0InYy/v8kORMEgC9wgJNT5x4y5/ICaEYzwi10aVMiqo8Eh0dihPN/pHM/KIXTcsZDU+2WNCvqIv5ZwO+CNIAFdOuGIyZSJAg0Qc0zYs2ncYHuYRm29InCc3ndI3blI2SsI3d+I1cHFBMrYzM6IzQKI2ETBJQ7Y2ZmNJIudK4i9RY+dIJFcY2QI9TnVLseIX5aFlGeKQCodZg7ZUVPRJkndRmzVAC6cBymRICmcH8mI4knNcjqdQppdAhhZiz1M8Q8s+GTZYVOdmUXdlE8dfYvNiZ7VGWjRWY/dlm2dlWAdpxGdqiTRWkrZWbbVGnXRWpvZerPVGtjdqh/dqFOdtSYdsXqdicjdtRodsKydus7dtQAdz/KNyyTdxPYdxMhdwMpdzLzU+Q/nmZlCndlimZX8y69fRVYTVW0N0U9WSbBYGb9QnTc3Wbs+m476RXAsFXkPbdSCF5CZGdZ9rCjJXeu6pY5rW+lQvfRhFPnWda+Emf2o28VDfg+lneIxtbBVF1berf/31wQLqhLrbA+Y1OF5qkFF7B6h1Os9bg5PvgEH7ZQmpeRPqjq1vf2zSkRZp/sFu+9jln8MXfIj7iQhGm9+pNZSpjxztOaarja8rjF86jECZhnEBy7GvjN/5Od8ticruSgQqs3bSoKs5NQo7kXqvkJF5kjetNXFvlp5qq2/TlHV61O/feWg4UuPpNv0qqMH5O5dbmeJu1sX3kNO6yae4TyVay/ox4Z1WurUCbcFXu5Geub3n+E+vaaUVrYCpOsu6Kw3PrZHee5IfOEzHerp7muj2Oxe4VsZo+5Ls26Vle6Xo+sisb6MXb6Ke+6Kle5nAu6nlL6j0xtNs655Fu5kRrr1L+5qMG63Qu66Vu5g5O5pvutjRO7KBObr6OkcA+67xaDHD7TU3u6lPe3sq7TdNe7M/u4Hje7DYh4YLOuJRL7dwEueA0uag66IO67Cvp7TuB4wz7cS+uVPA+sfKe4uTes+zuwO6uEzPnYsrLvAV+sMnrTQKv7oq67yXd7zcRT013vuArvvtumA+vbBGvvoWe792UdVvXdbnJ8DNhwHRndxRs/uHJLk4RLBATzKHi1MA4udTHpxB9198gLxPyjRCQnno6/PKte8Ph5MMzDJCXrhBJmsM9Tcw1H/L0lHteLE5hDIs4bU5Mb8XZjY3Md8aCaJjGxxCoK9Rb7NNJHxPiR36I/Mf0x8gB2M1jv8eJvH6AXMqDHPX9pMjs18g7HPbfPlGdbMoSSIGpLPSkfIkdyMoh2MtyX097j8un7Pf2jPc2P1E8iH/CTNKAj1CRr/g1aMzgfPj0dPmTzwmNCPaO3xJNqM3JCPpqb1FWCIXfjNC+XPpXePqBuPCj7xLp7Ib17NC3v86cgM/vvM3ODU7znN6hj/S1zxKECHy7OMycX08E/n2IEf36kqj8uU/7x78S00+Jt8z8we9NIo1/J20Doljn3/T9DW391y94tUh6t/iK3CSMla+K7C/TueiHNX3T3X+ErEj/73/0Y5z+AFFM4ECCBQ0eRFiM00KGDR0+fGhpjyOGkZhIspRESBAbQIYY2cRJIESSJRlKpLjQoqRFNmwIGeLSz8KRJm06RFnxYsaNHT+GTBhU6FCiRY0eRZpU6VKmTZ0+hRpV6tSiN63anHTD5dYdmkQqvBoWIqCtNvQwrClWLaesZW10/UpV7ly6de3exZtX716Ea/2WTPtXLCRIDgMLFsxX8WLGjR0/howX8eSvlBEftiw28mbOnT1//gYdWvRo0qVNn0adWvVq1q1dv4YdW/Zs2rVt38adW/du3r19/wYeXPhw4sWNH0eeXPly5s2dP4ceXfp06tWtX8eeXft27t29fwcfXvx48uXNn0efXv169u3dv4cfX/58+vXt38efX/9+/v39/wcwQAEHJLBAAw9EMEEFF2SwQQcfhDBCCSeksEILL8QwQw035LBDDz8EMUQRRySxRBNPRDFFFVdksUUXX4QxRhlnpLFGG2/EMUcdd+SxRx9/BDJIIYckskgjj0QySSWXZLJJJ5+EMkopp6SySiuvxDJLLbfksksvvwQzTDHHJLNMM89EM00112SzTTffhDNOOeekorNOO+/EM0899+SzTz//BDRQQQcltFBDD0U0UUUXZbRRRx+FNFJJJ6W0UksvxTRTTTfltFNPPwU1VFFHJbVUU09FNVVVV2W1VVdfhTVWWWeltVZbb8U1V1135bVXX38FNlhhhyW2WGOPRTZZZZdltllnn4U2WmmnpbZaa6/FNlttt+W2W2+/BTdcccclt1xzz0U3XXXXZbddd9+FN1555xU1IAAh+QQFZACrACwAAAAADgKAAYcAAAAAAA4aAAAAABktAAAAIgAANAA8JAAAACwAJSsANTBHAABeAAByAABJJQBVNQB5IgAAQScaVC0AQzEAXDgaVDItVCMaYzk0YzA4dzxXQQBbQQBfSwBFWBlkVAdTbQBpYA5kewBucBd4fB1ScCoAAEcAJFoAMWYteUAwT35bQ05oSEeZAACqAACEIQCZIgCQMACYMQCePAC1IgDDOgCPTQClRQCuSwCrXwC2WAClagC4ZwCudAC5dQCwTCDJVADZbADHdwDjcgDKZjmUaV/RcEV5iwBfkT17hyBmhzBmlztynjs5hkNGlkdWp0tfs012q0Ris01mu09pu098vExsv1BswlB0xlB1yVB+y1CCigC4iQC8mACAkSOFniqKpy2IozCMrDKIqziPtziLtT/DjgDVgADAmgDFngDoiADrmAD1nQDFpQDXsQDdtwDduwDhpAD7ogDyqQD6qQDjtgDlugD1sgD/tgDxvQD5ugD/ugD7vQD/vQCSwj3twADywAD3wAD/wACCsULsi1/wjmjyk2r3mneNyU2SyUGSzUOSzUaRy0uS0EmS0E+DyVCC0FCI0FCN0FCS0FAtX4tFZ4tpdZOFiaOQjqWZkqmwmau/o7DEiIjMlp7ZmJrlnJf7oIP9qIz7qJj9sJrPnKfJqLH1qKn9sanyr7HpsbH3sbH9sbEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAI/gBXCRxIsKDBgwgTKlzIsKHDhxAjSpxIsaLFixgzatzIsaPHjyBDihxJsqTJkyhTqlzJsqXLlzBjypxJs6bNmzhz6tzJs6fPn0CDCh1KtKjRo0iTKl3KtKnTp1CjSp1KtarVq1izat3KtavXr2DDih1LtqzZs2jTql3Ltq3bt3Djyp1Lt67du3jz6t3Lt6/fv4ADCx5MuLDhw4gTK17MuLHjx5AjS55MubLly5gza97MubPnz6BDix5NurTp06hTq17NurXr17Bjy55Nu7bt27hz697Nu7fv38CDCx9OvLjx48iTK1/OvLnz59CjS59Ovbr169iza9/Ovbv37+DD/osfT768+fPo06tfz769+/fw48ufT7++/fv48+vfz7+///8ABijggAQWaOCBCCao4IIMNujggxBGKOGEFFZo4YUYZqjhhhx26OGHIIYo4ogklmjiiSimqOKKLLbo4oswxijjjDTWaOONOOao44489ujjjx9JIuSQRBZp5JFIJimJWIE06eSTUEYp5ZRUBgIkREpmqeWWTFbp5ZdfXvnQlmSWSWSXYKapZpNiOmTmm1yGteacYLbZEJx4JokmnXxGaSdDeQZa5J59FmrlnwoJquiSchrqKKKJLhoooY7SCWlCkk7aaKV9XopQpnlSyqmanh6EZyOCJEHCqkdQgaeo/mrKscUWaHRaakFwinHAQRJk8SasYPYwUAF+8HkrrmaCoRACvpYJbJVqMECQAX8Ye+xAZvZBwEARKLEECQRRYOazUvKBg0HUWnvtKmU2AsJAGRA5xrarBCCFs5uCyUcDB6Vr6brskokIvRNEUmQXAzGBL1hr3rGAQB/4cYNA/s4JcMBbfjGQE0ZqK1DBZJILpR4LGMBGkzJQXO2/627ZSCMiCDTAFUY2soHMNMfJ8Jp7PJnyKhWvebHLjTwgkAKQHIkEziHnS+fPQZMKMJmM3LzKBEkbufQqMze9c59Qr2zx1C5bjQKSYchchddfGRq2ute6rMjDq5x9ZNoCcaxz/tuFvs1y3Ft6XDfaAzXBtlduqwz3sXLTbbeReK/yxOFdJQ602EKTneXLVl+ApMZcY0E5V5ZHnebQW1Z9dNZFjsD03oj3rfjfjJctUAKPHBkz1zlrKXKVfo/dssvvcm2FkYw48DHrWf5OZfCZD7+lF4UbOUbCC/MN9uzCA765JIIjkDuRHgwURfaxb3/54rcS3QgHA1XgiJCL7H418807DWbPfAQCA8XF6l/0vLelMQiAIBbowAFlNoVxoYRNLdlDHg6yJj5AgCFakJr0yES9hOgNfSIJxB1+sIpaqWQONGDBKl6whYI0jG4KCYGURsiCWjkJdWYiQwcMggFX/aok/oFIwypYwIIzQNAkd2jBKmRwAyJygSAVvOBCjBClNBCxiE/C4ZsWIQgoQIEKhwhVSe6wwh1g8YgkEcIqdNAkK8oAEAMZVZPsACU63oEFLzCjEW+oOVA5kCR8MMMf4nDGQ5UkhW1oEh9awAI3xHFUhGRjk4DAAjYEcpCFNCQB/Ti6jzhpDYU8SQteUKwmpdCRApHjHWbAAjZSkgV0+GQmtchJ2IHkSaDcoyZFksQ3OumUjxyVHlh5xVjKUpe7bF8tQejJY/KRJHpogQzE9gNYBlOYjGTBG6CUyyz2cZn5GwkuQznGFsRAbMBMpRwD8UoW8ICbs/wmOPUExCZ1E40j/jmlkxhZh2tWipIvgAMrJWnPeG5wnvQkiZMIacOTnKGa7wwiC2oAxVGB8gV1ECEjubDQGnrzoAg9kkn48AMcNNEGOdgBHE1CSBbgIAdENKE6ORVIY4owCHggqUlZgFKVJrNUIQ3nSPKgxIGw4JwnCSJBynDEdU6JqFck4jl/6qmgKikl+EyJHexAQacOsHZWFan+vFqli5n1rGg1SViR5DyyUlUxqoirXOdK17ra9a54VQVk1irWr7nVS5LJq2AHS9i98nVQY/2rnyJD2MY6dq6GPeyQ2krWwD72soONrGQZ5VfFTsmymA2tXTUrWcp6FbSiTW1cSXtY0zoVtaoV/i1r+eradcI2tpid7VprK8fb4vaxug0rb0fl2982NrhKQpWqWOXDH3Z2TrKila0gE9tQbOK6nEjFZZGLJF3xqlnM3EqhhCUQYrGPMaoVRQMAwF4AXGK7j3mTshLCrD9qb03RmhbmNPgY1Q6hvex9L3DjWybBraJb3wqXfdOXJnOha7+nY6xoP0EAALsXvo5pV/FWEa8hzUsg9gqvVirIrwefdzGiNYUNLHzhAWdYYATDH8IEorBOjlhNDoOYxLj31caIdhDsRQARKtxix3KXSKBbxQeHNLCPGcyWpFMTyUyGMh7z1zGhVS97J2GI9grYyAROrpB2NwDRFclmr/Nd/mKp1DMnQe/KPsZsEdhbAlIUwssYboyWXla01RGJz67jnY2zUjoI10nCj/3EAthbCVXcOcB5ZgyZ0Hy1Jxdpa12Dsnhltz7axdmxKmavCVDhaDy7WM+pMxvhjDdorBT6xHB9LCHai4m4PrrIxw1zlhLhuFUrudVXebWn0etYUbiAvSnQbqkhfWpJ77nJg7tb9TR9Y/WZ7tDUdeycAYAATcj11l/O9Yu1NDeBPK5IkTsftQnN6WuHCdGDBcWiAdDob5sazONOLqU9dyTQlRnYVhF297DcWB+w9wSktjd7a93sxbhMdatA2pECnWk1P3dNb45wtgfbZfZSwhOdCPko/rZNb5CfQtyo3lwjNHC78RWJzL0TKoOfZmWN95ews2axztlrCZQ7e8/FG8DxipS85QG8KgLvMbEHm/Ods7jnhdV1ljq4CsMV6Xo0FjG7rW3od29csJ9YgdjHPvZ5c3sFKsiEzx1OpvC5XEjlE4i61+3qdncdsPAW7a0Zju+U7xl+ApEf/ewHsqNThU788x8AAyFAOC89teCONNvJZEAEKnAgA2ig1j8iQZdkdRVRxOCUJAil4j428g1XzJuofpAl050jKFQhC1UiQhLK1J9eyvFCZAil2ONxC8+8uWqbHu6o55tMOuRhcze/kSQusYkseGJShUjEPVa0wVJUCBWf/pREFjDRiZ9HcWwNQWS1pz4xeeKiF8EoxjSusY3eX+kYy0hO0MuRjk+ioxBaCX9fmv6xplAKJyd5qrdZQmISiKRIjIRKJXFJhKRLuFcokeQklMQGCch4C/hWiWFcsiV1rVUSo1RKgZBO9RQI92RIqjRQ7EREdBCCv9RIGogYHBhaR4ZQ5eRLptRIJjFOEDhTnDJMUUUHvQQIL+gGMXgYM5hbHkhb0CRN1GRNJWiC9edU0URE2xRN0+Qk1RRL/5eEo7WEu0USSTRVOciA4uRMTbVO7fROY4hOMNiFXkhXNThPCAiDTcJPSVVQPWh/owJQAsV/+nSHLJBRcBiHcjWH/uBUEg/lTvBHUSbBUU3CUGgkRxeVUd3HBRDViOEXa4aYWWAoXCXRUi8VU0e4ETp1Uiklf3JUU05yBzglijDlUaVYGJ1ofH5XWkAkRAPBVLOYEURFEEf1BxHoVVZ0RbzYi4NRi554fB94ElvleT/lWfmHf8FHcMqIV4i4TMPFKYXYiWn1jeDYGgZ4gGsmjcgIJOPIWfdljqUnT0x4cex4jj+SjttYKbTUjOsYj5sIKfRYjtJ4j++Yj/oIkGEIj+xIkKBokOaIkFZVj4/ijnuWKqtCAq3yKv7oJdFlQ8NWVW/iXQbRK84lkGlCXqtgXht5KfK1LOD1egFHJ/k1EO6G/ncgpSQGhmDgMhDiwnwtqSYOVhAxWVYQeSTuAi/yQi8hZnhTQWL9cndAOZNIAm2FNyQztgo1xpJIh2N0EzET02kDp0wZszEdQzCWJnOVI2UlczKBkHHYtklKQmZm9mdW829WeXg84zM1t5ZgpW9GE3H4IyQUF3NXdZHAc5dex5ZIAnFYozRpRpZRxnWwxo+2E22QIzNDZ3EimSZqWZh5qSS8Zm6+5noJdZlgkpkyaZhHIjjnRiSRY3WWOXNzQppNaZpnVm6SiW4DMTlzmZR295iIQjSd8zky85aMuWmOeZKQqSUQJ3FG8pdIKRVJ53gcuWdWgzu6s5iBqZBfAptU/sKQfxZ0lTkkRXc/zRkVz2lzm6kkVMeaHoY94wkV5YmXXqklblckcbcKc9eaZVmcXQlUkwZ4qyB4klA/A5GY7fkUiMd4igc0AbSf0blnlTcQCbRAXKN5BVoR+0gSnWcQobcQGWSe8fk9rGcQoDmcnjRCJSSPGBF7K9RC1wcmuhdDM1RNGomiPQInyVcQPQQnO0h9Z4QSzvd90deiX2JBDLF9T1KMBiWbWaJ+XxRGFimG9LeH7idJbqSK91dHIoRHevRRShpUIyVID1iNInGBi6SDPugoEzhJleSASXqeARlCeiimIeGCZTiMfLJK/PdKNnWCNMoj/XiGcXqhzSdN/kRYp2fqKEBYTDzIpW5akHAaqH0aEVj4hHRgp31ShdoET8jEnSEVhSeIRObkhgzoVGuoqYz6ofj4qJ9ah0YoiP10qIbihyqIhoJqJ38KqJIYqRGxiBFlRY4Iq31SiRoVfR01o5xqgw1YUqioUicBi6RoqXPCik3iijmlrDyVirqaI7cqEr9oVEi1o0uVhooFVVE1VcdKhw+UrRbxjBqqj+3olAkpmv8YjvRar/Z6r/iar/q6r/zar/76rwAbsAI7sARbsAZ7sAibsAq7sAzbsA77sBAbsRI7sRRbsRZ7sRibsRq7sRzbsR77sSAbsiI7siRbsiZ7siibsiq7sizbtbIu+7IwG7MyO7M0W7M2e7M4m7M6u7M827M++7NAG7RCO7REW7RGe7RIm7RKu7RM27RO+7RQG7VSO7VUW7VWe7VYm7Vau7Vc27Ve+7VgG7ZiO7ZkW7Zme7Zom7Zqu7Zs27Zu+7ZwG7dyO7d0W7d2e7d4m7d6u7d827d++7eAG7iCO7iEW7iGe7iIm7iKu7iM27iO+7iQG7mSO7mUW7mWe7mYm7mau7mc27me+7mgG7qiO7oNEhAAIfkEBWQA2AAsAAAAAA4CgAGHAAAAAAAOGgAAAAAZLQAAACIAADQAPCQAAAAsACUrADUtACA/ADUwLSAgRwAAXgAAcgAASSUAVTUAYyMAeSIAWy0gAE0ALlIAAEEnGlQtAEMxAFw4GlQyLVQjGmM5NGMwOHc8T00AV0EAW0EAX0sAXFsARVgZZFQHU20AaWAOZHsAYmoZbnAXeHwdUnAqbHwgAABHADFmLXlAAEx5ME9+W0NOWVlaaEhHT05pZmV7fWx/mQAAqgAAhCEAmSIAiDcAkDAAmDEAnjwAtSIAwzoAj00ApUUArksAq18AtlgAiWcApWoAuGcArnQAuXUAkFQ3sEwgyVQAxWcA2WwAx3cA0nIA43IAymY5jFpLlGlftm5fh2BjnXN7pnR30XBFwX94eYsAX5E9e4cgZocwZpc7cp47OYZDRpZHVqdLX7NNdqtEYrNNb7dMZrtPabtPfLxMbL9QbMJQdMZQdclQfstQgooAlpoAqIoAuIkAvJgAqKUAtKwAvbIAgJEjhZ4qiqctiKMwjKwyhqs4iKs4j7c4i7U/w44A1YAAyJIAwJoAxZ4A0p4A6IgA65gA9Z0AxaUA3KkAzLcA17EA3bcA3bsA4aQA9aMA+6IA8qkA+qkA47YA5boA9bIA/7YA8b0A+boA/7oA+70A/70AksI948AA7cAA8sAA98AA/8AAgrFC3Ilx6Idl8I5o8pNq95p3jclNksZAkslBks1GkctLiNBMktBJktBMg8lQkc1QgtBQi9BQjdBQktBQRWeLtoaLhYmjkI6lmZKpsJmrv6OwxIiI2oyFzJae2Zia5ZyX+6CD/aiM+6iYz5yn3Karyaix2K2x7KWm9aip/bGp5q2x8q+x6bGx97Gx/bGxAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACP4AsQkcSLCgwYMIEypcyLChw4cQI0qcSLGixYsYM2rcyLGjx48gQ4ocSbKkyZMoU6pcybKly5cwY8qcSbOmzZs4c+rcybOnz59AgwodSrSo0aNIkypdyrSp06dQo0qdSrWq1atYs2rdyrWr169gw4odS7as2bNo06pdy7at27dw48qdS7eu3bt48+rdy7ev37+AAwseTLiw4cOIEytezLix48eQI0ueTLmy5cuYM2vezLmz58+gQ4seTbq06dOoU6tezbq169ewY8ueTbu27du4c+vezbu379/AgwsfTry48ePIkytfzry58+fQo0ufTr269evYs2vfzr279+/gw/6LH0++vPnz6NOrX8++vfv38OPLn0+/vv37+PPr38+/v///AAYo4IAEFmjggQgmqOCCDDbo4IMQRijhhBRWaOGFGGao4YYcdujhhyCGKOKIJJZo4okopqjiiiy26OKLMMYo44w01mjjjTjmqOOOPPbo449P8SLkkEQWaeSRSCbJi1ioNOnkk1BGKeWUVKICJERKZqnllkxW6eWXX1750JZklklkl2CmqWaTYjpk5ptchrXmnGC22RCceCaJJp18RmknQ3kGWuSefRZq5Z8KCaroknIa6iiiiS4aKKGO0glpQpJO2milfV6KUKZ5Usqpmp4eJGghZJChRqibzllJE/5KxHqHJHyWalCetrwwUAK64Cmql45McNAFlMxpa0F4CkIAQRr0CuevVDaiEADFknrsQG/KsoJBDDj7JrRSduLAQAbcgcgSBFmw5rXYlinLuNx6aya4USYxkApPYgIvAHtYy66ZsCyLDQi0nCBQt7626iUoDwhkgCdQMnKvv9eaGQsBGLDBiy0iHCxvmfQ+Ka1AdkTJsEAWhJImuwK9+cqQtozgccJgqSmFQABEIqUQOOtcJ8uBxjzzswpXGYTDpUhZRc8rA42rzNggTHTNafKMjQGmSHkI0z//+/TQ3xZNpdUoTOkIznw07TWeQkf9MZkhNykKBAKVLeXZJKtd8f7XbtP8VZri1j2lJQOBofexQUMtddhUfzm34HcPpMfhtiYO9rxiT2n1BWZzHabTbCv+dpyNf2k11kp77iXL2Fje99R/Vy1QAaRIeTM2OVNequuLY166l/binnaUPSCtu6e8j65l3E2OjI3hUGYy8fGXut6s316pGTjutT95hOQUI/4mLrzgYosEAinQK/m+x56mEQNdsIncTJCbNfWQmilLBAydATJKbGpJKD5xkDVlQmACCcEJEMivdYFuSwFjiBn+VxJUdCIK2FDEoVCSCSLsABs+wENB5uS8g5RsShfcgQadxDr98W8h/oNbBSW2gx0kIoAm6QQPsCGEI9QwD/4EoRMmimCQEtBqSoyooQ2f1EJQwU4knQAhE5aIQ5JYARtLaFIShXCKgfTpE5VYxCIgATFUcAJKZ+zEDnwwxRuy8IFOlCFJRHEIU1yCihskiQePKAoe7GASXhzVHbPYpCns4BF0tCMe8yi+OFJwJE5yBB5PwgMfJK1JHgSkQEZlwSHsIIuG3IEmIrnIJjpSjiJ5kiTdyMiQ6JCLTspkIEcFCk8qcZSkZGUrd3dKVIZElZMsCSh4IIT7oSIKopwlLf24g2rlkolw7KWeSAJMVuaQB0Ewpiw3yUlUhHIHTYDSKqG5NmkqySTPrOJIMukkP25CmZUypA8q4UlCNmmcb/4spzmRVJI8OOmOKzxJIpAZTlQkkYjwNJQkfTA/He7An00CKDn3ts9pzjEKSPChEZLAhC6a5I47QEISaqjBhBaKjri0IBU8IQqMapSjp9gl8ipq0ZF8YocD2UE2T2JQghgCh92c0k2VWMNsyrR6NOUnAI9aEk5wooBBdaA+kzok5kVVnazLqla3GhKqHsmqV5XMNcZK1rKa9axoTatarwEZrxoJrFEV61rnSle6ttWtZ8rcVaUk17r69a9jvStehQTXoPYVsIhdq2AHW9huHjaxkDXrYvHaWE4+NrKYnaxbKzuqy2IWspr1Kmc55dnPIja0VB1tpUpr2r+iNqmqff5UZFr72ddmKRVjcIFuw/AG7HXFUJjAAx5WaKnZJhYZX8jCFrbABWNQA7SPgdMgDnCQDNCBce6jkxMGUoBLGsu4f1XFBABA3vICYAHCSKxtjwQIhSDguo/MHp0a0TD71Qq8db2CefdL3vQCdr1FGgUCMUCGMriAIBton3zVJAokGAR1xYXMX6URhPLawBdfeIJ5Y1CN/0a3TCkYCAiIRAiBBaAN8f0tg+n2YGNK9TET9gEActCMsq6ivAhwhocdU6aLCeR6ROrDQCboSxUDDl4oKMX3rubi8DUGsMhgBlqhUN5g7LgxZfrDQNBgJB9jQwO7KDJX1gQKBxjgEU06Xf6T8YcY2nqhvL+4MmPKxAKBDEAORmrbncW8lTmpzElqvq+EW6tfACBgGHJeDJnOd7khtcDOeCZddukU6AjD2LTKeAB5YfCMRCtm0VADcpHEAGk+a8VQlf7uoDG7jB+Utxen/bCW2iYDJAXCznAwdVZQ7bA1d+3SEy6GL4jxBSyQtwEA4HCseTzrdwmk1ke6tUC4LOkF8ynVL3YMYpehaf7OwBrqlXWWIogNaBtJ2tiI4fL0CiZsO5kxiFWGA/hLXhx02tOJIZOzy23rgaRB11jhNZMFDWy/SuMLXUg4Fg5gXhjo2LXiVlLbPIAkLWNjAHMA+FUEDmFVFzyyqiBAef5n8Fy/AlhIbeudo0tdbSNfu9cE13ZrW4FjRJs84knimEASkIsj1fnikV737+bkbjYfhrbXoDJ5gQFxZm8pxBePg5Fq8UJRZ2m0Rf+1zFv7ZvLG+eZO15IfBqLuIRFiyCkec6Gy/rlVm1bpA/BvXU8uJAELBAE9J5LBBLKGtPd57TC39NbpuowuTAOtrsDxw+eO85yTYCAcuIWQZvHzL4dZ41ahk8pEgQogGI/z2X6yX1lhaB3UmKyqmDd5aQBusGO5TIQQAEE6YALZ29kNCibJAF2CVWysSRQUYEgdpjRAKLH2rKQvbwJqcAPVkzfH+EbMm9qbEGr7/SMd/GAIVf5iQQyWNIjag5dC8BU9D64RD/n8+FqTIXJ6b1rurp+zmWyRChMY5AO9xS4Ud9jDHwKQhlQEfmkCfAwBPU7iUP33UL23GIC1DFpQAfy1ADpweMv2enAyC6mgBmrwBrPAKlaERVq0A1x0ElHERsHke5x0Rk9yRlbwSSEIS8enVtGADMdQg9CQWY2XWiWxR3LjR5pUEol0R9bETZwySE5iSI/Ag6jQR3/EVG2GdNAVdpRVEpXkXdtUQfd0gpzUCfXkTTWkCVUYS00Yg1CYVnTXSyXxSjGFSX+ETum0Qd1US0SlCWoohpPghEdXhhUof4PFKCMxTMXkJMikCW6YhUOIgv6cNEw1VCyAaEyDiIeGoYd7qGh96If7Z1Rs+IOQ9IYmZSjfFE46hImowE5kKIlldYanZBLs1CTuxFOGmE/dJE/05IKriAruBImFYYpNZ4GMRRIDBU4hiFCFKFF5xEkL1VB+lAcEFYwLqBi6GH+U2IcfVUMiRVK4mBEtlVE7sFEd1Yl8glIHuFIgVY0q1IyJ8YyMJ4WbNUM+ZY4dcVMEoVOm4I3dlERK9FPumIfoqFg5CFso4VS8t0t7JSVOZXz4tY9ohYqOFFuGUorPyFUQGZGtUYmWaG0DSSWm1ItDd5F+Ek2ixW4ceY1XQpEMWSgZOYUbGZLp10gaOWkqOVEsif6SLvmS+WgnJAmSHHmS65iSL6mTH8mTKumTW4JbuuUCvOVbakcnrxIrSjArMVc5bzJd1QVf17drcxIsw+JMRmeTZkJ9CPFeuZeUaVJCBkEtocdLZGJ3AkFgBoZgYfl3R0Yu5oIuA6Eu7zZTW2ILUDcwJGZiKIZ5VbEmwYMN5Nck+oIz/bKVbdJjAmN1vCBkAkFkLSeWC1NfDxMx06N1ULklFocN1jckXgZmgEkVauI8J/QkJ4MNKaOYYkJnLEckehZ0V4eTTnI7uRMlVnObbUdRWcJor1MkjwZ0ozkVanI0V+NdT7I0uOMzuxmTORdqo0NqwjmZcGk6gQclW7OcrP45knkJNeZWJNI2ALlGnacmO9hgN1GCNwAwPM25mVmyb99JJOj2medEm0vIYugpTgNxmqvjkUdCbvE5JOhWdvUJlE+yPfn5JIQjEAbYn1OFJPDZbwLxb+RplY6Dn50jEJOjmWjZm1BDcUdicRg3nFKhJpuTobrpoLwpcaJzJMG5ZxUacObZccmpOlUilHnWMdjAcz73mrNpoE9iNbRjOzaKkf6ZZ1A3AFJXJFT3Y8rzVvaJCsGznlJSPMe5nUBSJmMnEATKC2cXmVUpo2DiPA3aJNIjEIXZnh2aJWqJd0Wyd9jQdyQaFeGHM93jJEuGDRvKoXg5a48nEJE3eZXnmP4FOpNUAj8CIT/0Y19Y+iNmEnuzV3sDMQC4F6YeUZMhsXsGYUAIhA0KxECJ2ag+8iZbihD0KXTUdEEZJJIakX0gJEICOJYKwZ9QkkLExaqO+iaFYH8FgX9P9EsA6EYooUM85EMPFatpMkRFdERSYo+ldKRZgoEa+AYvg5QfUYJttJIicUWEtEUehYhzAkZiREZNooJOkkZrlK3auqY7ORJBeIJ6tAN85IP0OCdGWEiH9K66hKM0VYj4hKsVEYaZWK9rwoUuGEop9a8A2yM3uYmviKkYUYcDS4ScIoe3VE0w6Z4y+UucKEzE5IjJRLEVy0xaiQoKy68VhYUPu7ASEf6K2tSGIhtPSlRQb8iyO9KwDvuvJ1GLrRiznrhGs2hPJvusD6qDIwFRqECMArWMBrUDwgiujnKMFpSM/1SO69qnGxsS2fhS3TiNITVS5UiwDHYIKdUJK7W12whTNqsjOCsS8JhTO4VOEjMQ+Ci2jjJURGVUKLtPKQGxHwGQm0qTHVm0/gikFymRiJu4iru4jNu4jvu4kBu5kju5lFu5lnu5mJu5mru5nNu5nvu5oBu6oju6pFu6pnu6qJu6qru6rNu6rvu6sBu7sju7tFu7tnu7uJu7uru7vNu7vvu7wBu8wju8xFu8xnu8yJu8yru8zNu8zvu80Bu90ju91Fu91pV7vdibvdq7vdzbvd77veAbvuI7vuRbvuZ7vuibvuq7vuzbvu77vvAbv/I7v/Rbv/Z7v/ibv/q7v/zbv/77vwAcwAI8wARcwAZ8wAicwAq8wAzcwA78wBAcwRI8wRRcwRZ8wRicwRq8wRzcwR78wSAcwiI8wiRcwiZ8wiicwiq8wizcwi78wjAcwzI8wzRcwzZ8w94REAAh+QQFZADEACwAAAAADgKAAYcAAAAAAA4aAAAAABktAAAAIgAANAA8JAAAACwAJSsANS0ANTAvLy9HAABeAAByAABJJQBVNQBjIwB5IgAATQAuUgAAQScaVC0AQzEAXDgaVDItVCMaYzk0YzA4dzxPTQBXQQBbQQBfSwBcWwBFWBlkVAdTbQBpYA5kewBiahlucBd4fB1ScCpsfCAAAEcAJFoAMWYteUAATHktR2swT36ZAACqAACEIQCZIgCQMACYMQCePAC1IgDDOgCPTQClRQCuSwCrXwC2WACJZwClagC4ZwCudAC5dQDJVADFZwDZbADHdwDScgDjcgDKZjmUaV+HYGOdc3vBf3h5iwBfkT17hyBmhzBmlztynjs5hkNGlkdWp0tfs012q0Ris01vt0xmu09pu098vExsv1BswlB0xlB1yVB+y1CCigCWmgCoigC4iQC8mACopQC0rAC9sgCAkSOFniqKpy2IozCMrDKGqziIqziPtziLtT/DjgDVgADIkgDAmgDFngDSngDoiADrmAD1nQDFpQDcqQDMtwDXsQDdtwDduwDhpAD1owD7ogDyqQD6qQDjtgDlugD1sgD/tgDxvQD5ugD/ugD7vQD/vQCSwj3jwADtwADywAD3wAD/wACCsULsi1/ciXHoh2Xwjmj3mneNyU2SxkCSyUGSzUaRy0uI0EyS0EmS0EyDyVCRzVCC0FCL0FCN0FCS0FBFZ4tpdZO2houFiaOQjqWZkqmwmau/o7DajIXMlp77oIP9qIz9sJrcpqvspab9sanyr7HpsbH3sbH9sbEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAI/gCJCRxIsKDBgwgTKlzIsKHDhxAjSpxIsaLFixgzatzIsaPHjyBDihxJsqTJkyhTqlzJsqXLlzBjypxJs6bNmzhz6tzJs6fPn0CDCh1KtKjRo0iTKl3KtKnTp1CjSp1KtarVq1izat3KtavXr2DDih1LtqzZs2jTql3Ltq3bt3Djyp1Lt67du3jz6t3Lt6/fv4ADCx5MuLDhw4gTK17MuLHjx5AjS55MubLly5gza97MubPnz6BDix5NurTp06hTq17NurXr17Bjy55Nu7bt27hz697Nu7fv38CDCx9OvLjx48iTK1/OvLnz59CjS59Ovbr169iza9/Ovbv37+DD/osfT768+fPo06tfz769+/fw48ufT7++/fv48+vfz7+///8ABijggAQWaOCBCCao4IIMNujggxBGKOGEFFZo4YUYZqjhhhx26OGHIIYo4ogklmjiiSimqOKKLLbo4oswxijjjDTWaOONOOao44489ujjj3+9IuSQRBZp5JFIJvmKQEo26eSTTD4p5ZRDAgkUlVhCSUyWXCYZZZdgVmmlT2GWueSWZoL5ZZpZjkkmm1yuCeeUcs6ppZs72YllnXoqyWefSOLJE6BS/klokYYeKqagOSnaZKKOQqooozo56iealh4p6aGUNpqppph+SuSmhHaKk6hGkgqoqn2aelOY/nhccUUXXbJqZKyz1hrqlKlwYgULwFIhRpyu1tRlKi0MlEArxB6brEDLNoulHQccdMEZexZLE5d1EEAQBsy2uauU3X4bbrZYzqEQAtjSqe1MWJaSgkELnEulrfLSa6+7VFrirUAWXIEFCwRlcO+7MlFZSgMH1SutlAs3vG+h4yp5wkAeEHnHvwGA4S7CMFE5yr8enFKCQA6LKzLJJqM8sZZTkvIvuEXCMVAWH4PsEpUyW/DFK6mA4PLDT/b8c9BDqyylHANtYaTMAmHgCsU670ylKEOmEkLS6F6d9dbEpNz1kyoINEAZRmptNtp3Vs0SmGqH/bKTthIZt9gHS5lK/gRcE7nC2lS7vRLcYOPN77GFz/1oxUfGTbORVQDetuApEd734VzerfileoMdA5J0mD1G4JSjZLncRGOpeepJRkzM50eGLpDTk5du0umGU4z75l4ybuTIAsFupOzEaEG67SXtzjqviS9/pOvCF0k8F8cjP5LySmfefPZNxs0BkkwTM4AZ1VsfEvZjZ7k690quX+Tf4rNNt/m3I3657vajzn6SSBOTACtHKlv8ykc/j6Avb/nLHcyelIqLiY8MRkIFBKLGu0UVUCQHxJz6tpe+JsVhIMYr0h1ulrMLguR0j+sglRxXQVBNyV8CQQAAh+SKkwnECyU0oQGxtIpXrGJv/gJRALN6iEAq9fCHfCOGEH2oQv6JYCAaUIWQTCFAYkgthzrkiMImuJAQLhBiXFSIF+fHqzsIgCAbIMEZzRaGg2XxIytjCM7wJyXgLWSOX2SguhJCOyy+MSNbZMgYFxfILmpQSpwggUE6MKxs/bEjqEKU7yxVtylyogtdEAPWiPVILUZSTJ8EZSjP1MmNjPJMp6wknErpyVCqkk2vTBMrTTnKWJrJlmWapUZSOclI9XJSusQIL4dZy2BehJiu/CWnjGkRZH4Sl2FiZjOL6cxISrMi1UQVNNV0zYlkU1Tb1FU3x0lO+GzinOhMpzrXyc52unMTAnmnPOdJz4RRs5ge/qKnPveZznjy85/ztGcyeZlPgBq0nf48qELRKdBnKrNUH1qoROFJjIkutKHWfOiqImrRgya0o//EqDY12iqOgjSkFT0pSkN2z2QWVKX7/ChMAxqTb34KRDONaUpzSlOWDhSfHeJpPXcqVHeKFJwk1RNOi/pOmTJVnUe9aVLttFSAIsIIQ8iqGgphUKfq86pZHcJWu1pTMOGKVs5jICdkhVZxvXSfgZDAQSpwCJQCNK5zrSs/o+okZCmrhamaqpD8Ci3ASvKt9ASEQgCgV3169Z2KTQhj91pWLJVrICksIpUuG7WpuTWo+4QEwwRiADXsgQgEoQBl+SnagZT2tKml/qxPITavgiiQjAqrLUEW4FlHglafQhgICtK5iNECwA065WdwBTJcdBZXIMfV6Wyd5DrbGnZUJK3ubnvrxt/OcxIOIG0k1PkH4SZXn+AVL3nN69jKPsmOJbPhbQkZM5bJ97oW5NA+I0uMNKwzvcSgACXau9+B+FedABZwe6fbJKMBTWj6a2KSHOy+7upXn0mALiHYuQMNExjDHl5nh4kBgA0PlcFN2iTQOKjZJ5nia/ernYb2qQPSYoKdTAgxTWlsYxzrWJ587SuLD7nCIRPwQvscsQEywU49/LipRJWnkpm8TieT2MQ9fUkG6bjBGM8Pse8csQnaGQjovuHD9BQz/pnNvGAtJxC/Qnplhf2IIX1W4gECGTM7yyyQA++YnnfO85r73Garac/L9D10hC28IX22lhh6XmciBjIFNMvz0ZFW56QFUukTu1nR8+Xcmznp3XcGGtKDJkYbLG1qPKN6zwNZtacN3eVFE5l5KOMunZGszxFXYNAlZnWYBfJrWF+50C3Zch5VZ2QZZyjJpKWyOnN87KFCmxhL9nG1s0xrZiNa1KCGMylLPWxiFOAS7MzwtoEc5XKfO91PNqp7vW1rLtM71L0DszuXC4Azr/MGPbY2cNn874BzO9mj3h+u6+3sOuuTv51OJyPYK/DEUnqdE2eudD9da3wHSrCDbfaX/sntzkcDAN3pBEKsz0tPk6McnSoXiKxnjXBFZ/bWnaMgqS+szx8MpAKOOGcliuBaaf+55z8P+iaGXvSNd/tJRwSiEocoYSVFPYlLJOKuLcRPRvxLIB8owdejy3J6ep0gYR87cttJCUlAFSXnPEghxYhzJJUijAkZ5KX0DVmF+HmdkEBCDfrQz3bLk78H+bvEe1CDGuBgDQw1ySYCTwzCFySOd6z7keyoEDyOnOf7XIQPDDICrrbzD42vAR8KD1DRk970gLdBDXYAhMazIe4l2UR5G796irLq7oLUvJGAb8gjc92gkkCEH/wwiPFu4hHqhD4kHF8E1bMe+cpnvvOh/p9O6DehBkQ4J+p3oAnfjwQSxMBB9Xvve5tmqqoTVQT40amEGgiiEnrIhPzZT1GQyj/851R/gsB4plcJsmcI5icS+Kd/1od7IAdLDyhLJjVRkMAD81d/NdAI6BQIDYh7IFWBF9h4jWADOHBj58R4CCh558SB7DcQ7kdJEzhRk2CBqaeBG9iB/QdSM5h6GQgJNkB+6ISCCQgS6cSCkReBt4SEuRSDMih7NdBYK4iDj6VQk+CEdVWFOyBtgqeBuXeDLRhOaQWDfKdQGFgDRqBORhh5KlWGZ+iDOiBtQtiFUfiFShhNdchNY2hQ9YcDiGCBADiH19dRe9iH8yeE5yR7/kGngpuQhu3XUg6Vh//EgTgQdD5YA7d3TvJHeGrYUZJIibLHBoJ3hrpXAz7ggCJxiZuQiab4gr4EifyEfzY4eUsQCZWABEFQez8gBEVQflMIULCITpAwi/JXA0EgBI2niSZRi7dYA7m4iy7oiBnlihMlCU6Yem+Yg0WFeqmXB0c4EtRIEDXwhs/4Uy5Fck/lgef4CNzXjV0od9A4UtLIU73IVEHmSwRVTviYj+ZxjghlePxYj8B0jy4xDARZkAZ5kAiZkAq5kMMAHPzITvNYVAC5TALZEgx5kRiZkQ75kG/HkW+HYlJVkSyRkSRZkga5kR6Jjik5hG/zjkgFEyYZ/pMa+RsrqZIpOZEQJZIrIZM8uZAoeZP+eI44uVE6qRI9eZQH+ZMeGZFCNZQlVZQpgZRS2ZA0WZNMKY/zNlJQiRJTiZRKyZFXmVNOqVR3KE4v0ZVH+ZUPGZYzNZZUVZY7Z5FoyZNqSU9gpVWwt1oGdZdilZdOV3Nccla60iW+AiwsICxxOZIy6QtS8ARQAAVRkAvCQJdVyU94ZRB0ZVeRKFeYCYU015Lq8yz+I251Qy3W0i5bdxIx6QkSAACu+ZoA8AK3IJN16U6IZxCTVXYWJ1meyW4g2SScZUWkmVR7hBDswmiKSZJOAJvMCQADYAsxWZvs9GjYZlqoNRCqpZuX/jZa1Qlb2ClbHPck+WJdYWgkMAQwAkMwA2EwqWkSJfkLOvCaDAAFTwABsAkDwWCS0rlOy0UMzXVOz0Via1dx9NSf/7kJAUp2nzk4U6JdA+FxLjQlDkQMGTMkGyMQHdOeJfGeOAAAM5ALBfkLy/mataCflfld4YVtzodO5aVxwuZOAGYAKyp+FHdwoPle9vVtvVNfOkckNiMQnrc4MFmSuCALCAmfr0kLJuobBUZoCJaiCkagh2dg/wWlA7agKsEzBOAzD6ajHzcl4UMMfTQkUGNFupZvZzmVIwoAsbCkvbFP6hZsIhZv/QindDpicmqjDDolKjZn9qYkAnQ2aQM2/oLacFGppknqprzBY9hmgtNGpxAZlOtUY42qbXnqm+FZZF4aoX2VRAoEP4X6eXIplWuqpCW5n+k0ZU0GqesUlqpaZaz6kZm6cBAaWDknnEcSOQNkqFwplbxwA66JALN5qicqZYJmbP32ou2kZsjqb3qapQlXdUUSN9FDJLIzAKPDq6oplZ3wmvipqLthZ66WaUVIpVLqTqdGrhtors9aOdHaYq0zWtU6JMQzpnuXpkf5q685C9FZrCU3Wup6TptGDBHHbo4GsO00sAULZb8pZJtqqzgaPKADQsa3rUe5pjIADP3KpIA2rqk2cwbbsce6Tnymash2o/c2nBAjrxMr/hDUo63ueZSf8JrCSpv+6k6+BmxYFrJpRmw6e7J72nEqy0Bg8z1HEj7jU7Exy5OhQACJarMc27PYZnToRG2XCmXXlm3rZLU7y7Czequ1Kklg+zKgKj9Ciq8xqQsN8JptSplRO08j9m7rFKddK2/XJrfqRLdAC63hVp5ZA2H/E0CSI6rJaZK74ACvCQtpebPtxG/Omk4AV6nnum8Ep06RawCOiqlPR6tD21cONAAQVCQS1KOEu5MxuQsP8Jo0oLE9iaroBHEYV6M8O0+wq04Z559/ibKaynC49SQfJBB6N0JAqqEkEZP66po0MJmL+7bbCV0vd04xZ7LKOp3G9byb/hC9IKu5gElvN/enSXKeMkQkNTQQOKS0G1qSvJADqsu6y/um++RzxKZ0TBdt2vlO8EsMQCd0REe/e+uuRuRDUpd10jqtTyQQUTRFVXRF5lu8JPkLP/CaLtALvjDBFDzB4KobXfd1xJB2A6Gg69R2HblPZzcQHAxdA4qlINF27ghGwee9SJIKZoRGajQQA9BGxDsSJRkKzbnDAMCvJOm66XSbBaF46MQIjOd4kGeTu4kQRAyMgqeJpjgSRlwD6bcGlzclnJcQQQpueqQQ9lq6RkmSoMDDzdm2M8m89OR6BVF67VSJtGd7SpzGo7fGfplO2iiFJeGDxPDGlkgQc5d3/sJ3JIm0SI2EnKabkYdLxrBpqmfsvgCVfMvXfOe0jpO8Cd8HgOPHi5L6TpCsfZXcfZNHfXhMEk1ADJg8e5owjg7VJaZwSZmkYr41kHMJtY4cf/MXgPZHgEJ3gNhoy38ogAu4f4VHErq8dAeoyhm1lRY7y8SKxgoFguGHgSNYgkFYAym4yazlh5sgzV44zCNBgo6KgsislUAly8zczLUsgzQogj64A+kUh9jMTztYg0U4yiLRzuV3gtY8zkilzEt7zo28qCdVhY13hT+ohRnYyxZF0E+IhvYcEliI0I3AzyFZzqMK0BgJxPvEhpNnA9eoz9e8hqknit3MjiDhhnD4/oQU/X5w+VnmjNEXqdH0NIjazHiNhYgKLVE0fcuAaNIgYYibgIgrDYP+fL4wHdOMa1Cd2NGWGIriR4pxrFBLXYmoqIo+7RF84NSjOMdgqHBSNaRHzZAyLU+/eE7BGAnDWIzHGNUHVdaySIu2iIu6qMkkkdbGOHiNSI6PiLZhnZBjfVB3XAPcyNYnRY08GI5MJnktKhCDPdT2aNGF29cI+dcKpY4hvJIpoY4rvMqQfciSPdlJ/VRsCVNuOSddHcsX/dknqY+s3dp/UZM5vZbxjJWujRCwPdoqhdsnVdsJcduzLZa/3Za8fRC+XdxWOdwGYdyYHdykjdwFodxA6dvO/j0Q0L2UzJ3b003dx13dYJndug1S391R4W1R3j3eE2XeEoXeF1XeB8WXYwVQ470Ia7AGUKyZByXf9O1R7L2ZeWXfAHUEA1EAmVu/+gTgAiHgZOXd/yTEBJGb0wtZKcq//r1fET61CZ7drMWdr3WdApGdD85OlRAEBqG1Ez5PIT7iVNu/rq1cNZqgJzy7Ju5qBUHieimuB0HjuTvc+xSjM6p7sou1oQWwmBBzOP7h6oRpQy7h4DndTdpfVdrhV3p09DQJDWAAgnBOr1ri70TlVo7lSp7jvG2n64ZOeFq3dcpPUb4JWV7j+pTmaw7mtc2omGupZh6pC/XmRl5uRY7C/ivea19+TlZ2tXarUHg+uVmb4nze2tcWsItYuTD+T4Uu5QAV6dqL3Db+aiTLro/OT5QO5AfV6fKm4C2HsOyksHmOs39u6H5u4fC93zE+smi4cqoutXsu6ZCe6u3a51JbbJk+5oP+6bi+6Yd+4Uy+6nvOtae+rMHu6QYF6v0o6lKLt+mkt7MOt8v+681+7c+O4QNHYo+LTpc74Ng+6dpu54Re7q3q6rR7cbb74+N+66yu5cPe6tA+Ty6nTthL4MaO6LbO6eguq8WOdPGrv02X7Og0YJWwCTnQYwnP5m2+dArP8CvF7fo0wmAndh384sLeTpUwAQyBBmzndoEIaB6//hAgr+KKzk8MrulH/sQjb+/cmRAHek5TjMSbeLAMMfNm7fJXHeb/pMYEwcanl3qoGNvoWvIKsbBuXHuWSNggjvQJsbCjCMc9H+fYF8nbF32hrH54ffMWRcnPZ8m3nMlGf1BgL32izH/lfd3u9H/0Z3/BjMf+x9MCWMwGaM2VoN6YSPdwn3/CXPW6ToHazM097fT/BM3bLILgXM11pfeTN/giWNJR7NwDvc4JfYNF7/ibMM/s/IP5vAk2XfYGxfmXX/gsqeMDbYUOrfZsv+WqH9HotIWiP/qqX8+sX+8WxdG2//ITxdEoXc0hfVK6L/mn7/OC6HiE+IeLKPcgtdPh/gfUOK35zr/7gJ/yFjXVn4hOVm34lul4ntjUZvjUpTj73T+JTF3VXT/5lt5Rbn3WyijXu0j++9T+wth4at31jk//cL2MzUjXFM9TACHJRg2CBHVk2kSM2CaGDR0+hBhR4qY/BWvkaahw4kaOEAVarHEwoUKSJU2eRJlS5UqWLV2+hBlT5kyaNW3exJlT506eJzv+BBpRY1CiER89eji06NKJPZ0+hRpV6lSqVa1etclUq9CFW70yVPpVK1ayZc2eRZtW7Vq2bd2+hRtX7ly6cQHUxZtX716+ff3+BRxY8GDChQ0fRpxY8WLGjR0/hhxZ8mTKlS1fxpxZ82bOnT1//gYdWvRo0qVNn0adWvVq1q1dv4YdW/Zs2rVt38adW/du3r19/wYeXPhw4sWNH0eeXPly5s2dP4ceXfp06tWtX8eeXft27t29fwcfXvx48uXNn0efXv169u3dv4cfX/58+vXt38efX/9+/v39/wcwQAEHJLBAAw9EMEEFF2SwQQcfhDBCCSeksEILL8QwQw035LBDDz8EMUQRRySxRBNPRDFFFVdksUUXX4QxRhlnpLFGG2/EMUcdd+SxRx9/BDJIIYckskgjj0QySSWXZLJJJ5+EMkopp6SySiuvxDJLLbfksksvvwQzTDHHJLNMM89EM00112SzTTffhDNOOeekE7NOO+/EM0899+SzTz//BLSsgAAAIfkEBWQA0gAsAAAAAA4CgAGHAAAAAAAOGgAAAAAZLQAAACIAADQAPCQAAAAsACUrACA/ADUwICAgLSAgRwAAXgAAcgAASSUAVTUAYyMAeSIAWy0gAE0ALlIAAEEnGlQtAEMxAFw4GlQyLVQjGmM5NGMwOHc8T00AV0EAW0EAX0sAXFsARVgZZFQHU20AaWAOZHsAbnAXeHwdUnAqAABHADFmLXlAAEx5ME9+W0NOWVlaaEhHT05pZmV7fWx/mQAAqgAAhCEAmSIAiDcAkDAAmDEAnjwAtSIAwzoAj00ApUUArksAq18AtlgAiWcApWoAuGcArnQAuXUAkFQ3sEwgyVQAxWcA2WwAx3cA0nIA43IAymY5jFpLlGlftm5fh2BjnXN7pnR30XBFwX94eYsAX5E9e4cgZocwZpc7cp47OYZDRpZHVqdLX7NNdqtEYrNNZrtPabtPfLxMbL9QbMJQdMZQdclQfstQgooAlpoAqIoAuIkAvJgAqKUAtKwAvbIAgJEjhZ4qiqctiKMwjKwyiKs4j7c4w44A1YAAyJIAwJoAxZ4A0p4A6IgA65gA9Z0AxaUA3KkAzLcA17EA3bcA3bsA4aQA9aMA+6IA8qkA+qkA47YA5boA9bIA/7YA8b0A+boA/7oA+70A/70AksI948AA7cAA8sAA98AA/8AAgrFC3Ilx6Idl8I5o8pNq95p3ksZAkslBks1Dks1GkctLktBJktBMktBPg8lQgtBQiNBQjdBQktBQRWeLtoaLhYmjkI6lmZKpsJmrv6OwxIiI2oyFzJae2Zia5ZyX+6CD/aiM+6iYz5yn3Karyaix2K2x7KWm9aip/bGp5q2x8q+x6bGx97Gx/bGxAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACP4ApQkcSLCgwYMIEypcyLChw4cQI0qcSLGixYsYM2rcyLGjx48gQ4ocSbKkyZMoU6pcybKly5cwY8qcSbOmzZs4c+rcybOnz59AgwodSrSo0aNIkypdyrSp06dQo0qdSrWq1atYs2rdyrWr169gw4odS7as2bNo06pdy7at27dw48qdS7eu3bt48+rdy7ev37+AAwseTLiw4cOIEytezLix48eQI0ueTLmy5cuYM2vezLmz58+gQ4seTbq06dOoU6tezbq169ewY8ueTbu27du4c+vezbu379/AgwsfTry48ePIkytfzry58+fQo0ufTr269evYs2vfzr279+/gw/6LH0++vPnz6NOrX8++vfv38OPLn0+/vv37+PPr38+/v///AAYo4IAEFmjggQgmqOCCDDbo4IMQRijhhBRWaOGFGGao4YYcdujhhyCGKOKIJJZo4okopqjiiiy26OKLMMYo44w01mjjjTjmqOOOPPbo449ABinkkEQWaeSRSCap5JJMNunkk1BGKeWUVFZp5ZVYZqnlllx26eWXYIYp5phklmnmmWimqeaabLbp5ptwxinnnHTWaeedeOap55589unnn4AGKuighBZqqFK2JKrooow26uijkNoi1iiUVmrppZhmqummo0AZ6aeghjopp6SWWqqnoaaqqqKjmurqq/6UorrqrJ+2CuutnMpK666N2orrr5fqyuuwkoYF7LHBPknsssWCheyznSrL7LC+QnursNOuWq21r2Kbbarbcmuqt6GSEkYL6H7Bxq7hujpJHXUUAiy5kf5xwEEZxDFru6UyMVABoPxK76N9KISAvqryuykiDxBkQCgCS5uqJwQMhIEYY7RA0AbaGgtrJ0YY9HDETq6awkAgLApIxdIEsEbCHrvaCQQHjYzrwIyuwrIGtTCqx0BkwOzsq5g4IBAKoBQhkM3XSgwqHwOZ0ajOAvEMbsymbuKAAYpQCsTSEN/s9KcrCDTAG43CMoLZaIuKtamcWPq1NEzDirOisEgg0P4CtDjKAttXDw3s3HV3Ozakalfdd6NgAO624L8SHnbTJYeauDQwPOqH2W0E/hWykpPcZKqsGI255lF77hXoYIvOZKqqsJy5o5sLVIbqXbFO9+R2H/5oK6bP3mjt0qSBO1e6F+7q3Yle7sGjUEszABzHb5U874ZXDiosIuzdc6N/S982qAqTGrrY2n/KvUAJzOJo2eJXr9X1ri+ZKiwnS+9Go69EULX7j/vcsc5HudGlag8DuR2jAAE0oQlwcK1DnwFDRTGBIACAijrBQIwnv6zQT4Kvux8JBsIBWSTKFfCThtU6iBVcxa0To/DB0gIGw96lD1SAEABBOmACHZrtZf4OLEmsWsIJTRzkYxRgiByyN8FUFSwhUuuYEDHxBGnISyWUEEIOpMGDOhQEVkVjiAoyRcUcyKtSzFsULEhhAoN8YF37EuIhpJGDHBBiiCbBhA6kAYQi1NEOBEEiQ7yAqUPU0Y6WSmOjXEEKNKCBDa7gVR65qARE4pEkVJBGEihlSCCIYiDiGsUlLjVKTOSAB5W8Ixp99y3ylaQTggiFJCwZrZJosRGU6oQOcuAIUHJrlpukVBRyoAhYypKWtWxiK1m4kUolgpYn0QEPAkYpLfZSIOLCRBBysMlh5qASzkSmIpeZKCGGU5XJFIkePVkpa/qSW5vY5iHBec5EspKcj/4yJ6WeiU6TbEIHQJjcE775TnjuMgePuBQ/7XlDfEZKn6NY6CXVqYMfTM6d2AylN3OwBIWK854OZZRJ6jlRkVizUru0REGhNUweQGKbwdznRxsaUkeNlFKzPONJCDHQjo7CkEMIJLeeyQNLjEKPObBDpXLKUGXW1KavfIIR/EiEIyjhkyaZZQ6McIQ6XjGj1oIlPY8qhUx0QqpUtaoo0mm/pz60JJrY40ByYNGT/JQggcBjKDUV10PW0aJsVZJb32rXwJ7kEpc44l5t6NTBsuptiyXVoSZLWeo4Np+QjeymJBONznr2s6ANrWhHS9poQOayUIWcZnMVmdK69rWwPf4takWa2dViirOwza1uPSvb2T5WtbbNFG53S1zX9ta3zXpgcDU13OI6N7TH9W35bNvc51rXtI9B7m+Vu9zbtva64I3ubKe72uqCt7jiRS15NWve8+42vZddb2Tb697cwhdU5kJXC9TFrtq6C15nBGFjnCuMLlwhC1nQAjCc4dz7Qspe+EJYEHMHLH8JBGD1YwxxSzEBAHj4wwBQAC/Qm91VPREhB5Mid13FMIdhb3nfzW0VQExjD4/4vSWeGMukcbGMbUzFq/tYyAqivHHF+LXM+MGHaYCLLjQBxC94Bo4dYzKUqYxlAVDDhJH3qpnV7MVGhoxumcEDANzgGJ81xf6HEYCMKTdGVVRT4eIU9TOBBI2ZVgGj6ZCmtN1leDG7FYYxROuED+/CzYxRVfSkEcVFxXmFrgQup7TGNa9FsICOcS8XPpwLRC9GVfA7W9rWFr8ABvlVcasUARmb6fPOGABs9rRi7qc3afDNb46L9IphtWomtvq6xHiAh12QDFkn5n6khvSiGldqXZ86cpdm9YCvW4wefPgWxHWwGkktPEbVbgCdMzWFIehnAWt4t8z4BS580QUreLgBAIhytnMcKVgATyDdXhTxGl0r/5qq1zAW826LIewaxwAaJKZyqGKHb9TZDs9V+SCmp61bYjigxh62QbGNjZhU3ft0tBvIGf4gThWJS/vcYy6FFrSwBSscAMQuaLNutY03Uj/PUdGbHsmnYnJfU3y30EB4Z0tBgA/HgMH2pXe9SX1r8OW635J+FcDD/Bj3omLNvZi50hHXPWm0731PJ+yupR5tn6McvIX2sC60rvBQ5W8A+2NU/xS3c6n0POBV1zSn2f7mAyawUQy085atN8Cy4/3XaPfwAG4c261DqoLSuCCjNCgQDoqby+Qu8qmO7NpibKEZok3FmmXe+LZvb4QCKeEJU6iBOTt73KgeBQxlSDcaTvzsrj0FrHGAZs+W4uIeloHQS9/3VOVwhz0cCLiBTJIiuqSk0hDkQpaYqSImS+Cw1f2HE/4wgxoA38Ox5nuiV4VAKMaxJFncYhdVctQqfnWlpArjQsZ4qfSfsg6rxP5rh1F0jA+b8UlneqnCRm4ER8wHEnrER36UVChxV3WkSkIlM0m0EIRkKUjVR38EfYCmW8WABRVQYwqAA6A3bwJoObawRo5ECtRSEpkUTJ2EVSWBCZQETWBlLaNkKaNEBdzESTnATvTlWssgDMEwhMpgXTRXUyZxS7m0S9f0SrE0S/1Ug88CTJUyTIqghLLHhIalGPX1XEcYUiYhTdQ0ChilTxJVS9kEU6PgTZUghu3ES1uYGF3YYI4XXySxTmtVTbx0UzIVhdEnLvHkV5WAh2/oCHGIGP5zmHDFh1wl8U8BVSkDVQl8GFE0+IeAeFAJ5YgC9U2HeBiJSIKLKF13WFEXtYcQdYbw9ywb1VF6BFh6aIg/+Imi9YUOlYRwSCkpVViU6IcadUovtYMnhYs5YFSxKIugRYv4VBI8xVE8GFQmoVQ4ZUaXJC5EZVRIZQc91YwayIXGSHzjp11ZVUdc5VWduBFnNVU5UFVXlYrAIlaVggllpVXjKI3lWBjd6I2fpl3JJRJ3NRB5VY8ZEVcEQVehwI57ZUiH9I8AORj3+FrISE4pgVjPF1jdRSmIdX1515Ck9ZDLJF+LVYwNWVkiOZILoY/7CHsV2VQhBI7+tlzjBJEtGf5cL9mRMUldIOVWHrlXM9lKORlKO/ktPSkuP5ktQcktQzktRWktR8ksSQktP5lf6WKAB0h4uAIJS4AEWEkHuGRuSUIrEGYQ+XJ+Y8cpiTABB3EBCXVyXWliBiNhl0eVLKYQAJCWh9dWsLNjPaYxA8Exgzc/RGM6dEMHg5AEBGEBaokkVSYQKaMoKyMQWdaXHvQqRzAQ9Ecpk2A6AIAHZreWofJormcLdSYNd/aWfpk1DUM3mXApcyQQlbl5NAU9qZMzO/M9r4d5pYIIAzEHmLIJp2kBqeaajfUooTY+NRd2kLJeUCAQAMAImTI3y1mXgmU5tdZ0jBI+okaakWkqP/4wQ5kyBcrJnFS3kuqTbLS5bMaJWVHXnIY3CoLwndDJmePZcCEnPeFWm3BpPkejKYmgnHnwnogZKh+Xb4qyb3UXFTJDM9KAAvqZm/55JHcpn8Pzd9jZQqYifwqaKZEwEBUInOJZb8HjcNIwchN6FQean5mynwJxBw1qJMgmEDfXKDlHPSOaZ64yNxewoNLwnOFpl/Fpa5+ZKNZJnGL3bPhZbpfinTkKnhzKo/XWdV/XKMNZoFBBdtJQAJ+QKcmZpCtaJPfzdnG3KHMnZ1L6FJLJn5myA9y5o9EZKuUnDQqkMg00pk7xKrgpEBtaKZRAmZv5nxTEMpK3KJRXPJBJof4ViplXail9Jg0quqVEoiqwgHrSoHq2gEID0XqDSqKvQgQDcQFGJXtKMBCaJ1w3uUA+JBA8VKrLd6kdsY0l4XwGASuUsGPSEAInsGOZeZgOSn4KwW8zqhHtZ0ULqRHpx0VeFIFxmRC6qSllFGDB+iO0QoAF8Ua0MlJz9ICs6hEJiIEMaJCbMgnOSBAlsJWZgpAzFZz1doKNhAakUJ5Tia0z6IeYpEk86EncyimaAAmGYAiLkJqiREpHdUqppJJMOl5OeEzwOhJYqEumaInQQoXCREzGBIUCu6aMaIaVSBJu+Ir1CkZqyIYkda1cqo+neLEUxU4aK4XPEojzZCln2P6sPmKSI3uwIaGJkEhQKJuymOhR6OSyPQKzJMGyJBsSrViKTbhXq6izEwufovizfZh/JBGMo5CLN3ssLfWLMbWLScunFcu0o8BUPDsRy+hTQGWsz1KNR7VL0Ni19AiyjSqyUYWO6giDJCGPXSWNG9tlgjBW8GhWaJWOavW1O+KzJCGQc1VX1IpXehVcfeVXgLWUy5ISbBsSEvmqKeldr+lYTfksJLm5QcEAnPu5oBu6oju6pFu6pnu6qJu6qru6rNu6rvu6sBu7sju7tFu7tnu7uJu7uru7vNu7vvu7wBu8wju8xFu8xnu8yJu8yru8zNu8zvu80Bu90ju91Fu91lV7vdibvdq7vdzbvd77veAbvuI7vuRbvuZ7vuibvuq7vuzbvu77vvAbv/I7v/Rbv/Z7v/ibv/q7v/zbv/77vwAcwAI8wARcwAZ8wAicwAq8wAxsEwEBACH5BAX0AcQALAAAAAAOAoABhwAAAAAADhoAAAAAGS0AAAAiAAA0ADwkAAAALAAlKwA1LQA1MEcAAF4AAHIAAEklAFU1AGMjAHkiAABNAC5SAABBJxpULQBDMQBcOBpUMi1UIxpjOTRjMDh3PE9NAFdBAFtBAF9LAFxbAEVYGWRUB1NtAGlgDmR7AGJqGW5wF3h8HVJwKmx8IAAARwAkWgAxZi15QFtDTmhIR5kAAKoAAIQhAJkiAJAwAJgxAJ48ALUiAMM6AI9NAKVFAK5LAKtfALZYAIlnAKVqALhnAK50ALl1ALBMIMlUAMVnANlsAMd3ANJyAONyAMpmOZRpX9FwRXmLAF+RPXuHIGaHMGaXO3KeOzmGQ0aWR1anS1+zTXarRGKzTW+3TGa7T2m7T3y8TGy/UGzCUHTGUHXJUH7LUIKKAJaaAKiKALiJALyYAKilALSsAL2yAICRI4WeKoqnLYijMIysMoarOIirOI+3OIu1P8OOANWAAMiSAMCaAMWeANKeAOiIAOuYAPWdAMWlANypAMy3ANexAN23AN27AOGkAPWjAPuiAPKpAPqpAOO2AOW6APWyAP+2APG9APm6AP+6APu9AP+9AJLCPePAAO3AAPLAAPfAAP/AAIKxQuyLX/COaPKTavead43JTZLGQJLJQZLNRpHLS4jQTJLQSZLQTIPJUJHNUILQUIvQUI3QUJLQUC1fi0Vni2l1k4WJo5COpZmSqbCZq7+jsMSIiMyWntmYmuWcl/ugg/2ojPuomP2wms+cp8mosfWoqf2xqfKvsemxsfexsf2xsQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAj+AIkJHEiwoMGDCBMqXMiwocOHECNKnEixosWLGDNq3Mixo8ePIEOKHEmypMmTKFOqXMmypcuXMGPKnEmzps2bOHPq3Mmzp8+fQIMKHUq0qNGjSJMqXcq0qdOnUKNKnUq1qtWrWLNq3cq1q9evYMOKHUu2rNmzaNOqXcu2rdu3cOPKnUu3rt27ePPq3cu3r9+/gAMLHky4sOHDiBMrXsy4sePHkCNLnky5suXLmDNr3sy5s+fPoEOLHk26tOnTqFOrXs26tevXsGPLnk27tu3buHPr3s27t+/fwIMLH068uPHjyJMrX868ufPn0KNLn069uvXr2LNr3869u/fv4MP+ix9Pvrz58+jTq1/Pvr379/Djy59Pv779+/jz69/Pv7///wAGKOCABBZo4IEIJqjgggw26OCDEEYo4YQUVmjhhRhmqOGGHHbo4YcghijiiCSWaOKJKKao4oostujiizDGKOOMNNZo44045qjjjjz26OOPQAYp5JBEFmnkkUgmqeSSTDbp5JNQRinllFRWaeWVWGap5ZZcdunll2CGKeaYZJZp5plopqnmmmy26eabcMYp55x01mnnnXjmqeeefPbp55+ABirooIQWaqheqySq6KKMNuroo5CuIhYmlFZq6aWYZqrppphAGemnoIY6Kaekllqqp6Gmqqqio5rq6qv+lKK66qyftgrrrZzKSuuujdqK66+X6srrsJKGBeyxwT5J7LLFgoXss50qy+ywvkJ7q7DTrlqtta9im22q23JrqrehZjLFCuhG8cWu4bqKCBpo6AEsuZHOccBBFpAxa7ulFjFQAZX8Su+jcCiEgL6q8rtpHw0QZIAlAkub6iQEDFQBFVWsQBAG2hoLqyQ/GPRwxE6uasJAHSxKR8XEBNBFwh67KokDB42M68CMgsLyBakw2sZAVsDs7KuNMCBQCZX4IJDN10oM6hsDYdGozgJdoAq4MZsKCQMG/EFpDktDfLPTn6Yg0ABiNFoKCGenLWrWpkZiKdjEMA0rzoqWAoH+QAv03KgKbWM9NLB0290t2ZCuXbXfjEoR+NuD/1q42E2XHKrixMDwaBxngyH4V8hOTnKTl4didOabR/25V6GHPTqTqX7CsuaOci7QFat31XrdlN+N+KOmC0R7o7YTk0XuXO1uuKt4J4r5Bo9CTcwAYyC/lfK9H245qJj37Sjg07sNqsKkij729p+W8oFACaDiqNnhW68V9q8vqerJ04fRKCkPLC5/VvQ7H+lS5YaB4I5RdACa0EB3LPNVboChophAEOC+RZFgIFtYIOsa6DoBwi5VpQjBQDJgikSJAn7EsJoGdXcruUkCEzdYWsBe6Dv0gYoOAiCIBkaQw7N5oWP+J4lVSyLxiIN8TAIMKYP2IJiqgiVEakAkCSYacQRiyEslidjBDIhhAzQUBFZFY8gJMkXFGcirUs1bVCkyMQKDcGBd+yoJJvhAjBnMIA9CNEkjaECMHPjAjmkgyBEZAgVM8cGOd7RUGhslikxoQQtfEAWv9MjFISQyjyRhAjGEQKlD5uASAxEXJhhxKVI2YgY2sCQe0fi7b42vJJK4gyUOcclolUSLgqCUJGgwg0GEklu05CSlkjCDP8RylrW0JRNd+T+OVMoPtTwJDWwQMEpp0ZcCEVcjdDADThJzBop4ZjIXycxEyVGcq1SmSPb4yUpd85fcggQ3ERlOdCqyleX+fNQ5KQXNdJoEEjTIAeWOAE54xpOXMyDEpfp5TxvmM1L7xARDMblOGuCAcu/Mpii/OQMiLHSc+Hwoo0xiT4qK5JqV4uUiDAotYtqgENwUJj9B6lCROqokaagULc94kjwQ1KNznAEPBMktaNpgEVPkZU4ptdOGLtOmN4XlEX7wxx4AYQigNAktZ/ADINjxihq1VizrOUUlOEISU63qVS+hTvtBFaIleQQfBzKDiwaRjgOxQx5FqSm5ItKOF22rkt4K1yAK9iSMYIQR+VrDpxKWVXBjLKkORdnKUuex+oysZDclmWF49rOgDa1oR0va0g4DMpiNauQ2m6vImPa1sI3+LWpTO1LNshZTnY2tbnf72dnSFrKrvW2mcsvb4r7Wt79tFgOFy1nXGve5pEXub8l3W+JC97qelS5tqcta62IXutpNLXc3693vGje8mB2vZMtrXt6i97HqZSx726vb937KXOhagbrYZdtSFYIIQQjwGXLpwcZ8Vxe0SHAthFFc+z7KXvhC2AqTdys/ROAgFFBoYx2D3V04AAAgBgAsGvyYWTkRIQeL4gZf1QeFAEDDzHMudJsQYhCP2L0lVpUEBXKxjG1MxSx0VRiXdgY8CIEgE9iwgaGLCwLUWMQkdoyqSoE/YqRMUSsTiMsmfL1XAWEgY6wUIk4HgDUskcPP/UX+D54MZRxLOVVUSyHjEvUzgQStmVd5FSQaVjdHXAqvxAjzuGRcXE2AGAFOcHKbd+vgRUmPGFBcVJxVCLnllqrFAjEDpvYskAnIbdCQea6HQcyKToT4xozOcajghza1sS1+lV5xqZAgEAAEIlN0s3WMQ23cJ4C4Bb3gxKmj3BgQ7o0Y3vvb414Z3E3hQIaZWkKtbw3qxxgXFwwAsSuGIWwbE5sxIHw1zxzlOFgz29Ll6+Cl7jDtXVubt2oGsQuCwe1hu7nYl3v18BhluwF4LtZBLhXdSqApP9SaDe5G8243EeJYeLbbi66vqj8VPNTVTnUAp3CpZna0gg9E09VWeGz+d1EDUjO43t6+N7hDJTvhpe52eLaKkE9H8EwZYiCFDPmSdetrACBgFp+FOKol/mZQVXzfiyre8TLe5Y3TjBg1x5TBBaKGhO8ctrnINgC2HXR7p7ro6Xs19BwlPerFvCqvohsFPE4MXeucMbs1AohfQO+ug9jhKl9MuPk2Z0WBr9VMn5+rHGgpabed2qcitGlNDeJW3MIWkOdFz7f+eGAQHd/cWx8x2ve+Zdeqv5iiWwEokSlaH97qcI8tw9nMehC/4vIr5x7+BqA/RvHPf4EHoKu+3HaEY6oG0H77YnS7+taz+fWynfinCghzBCrw7FRhMc4xlQgwn/nqpsWFDLb+z33ua93nMoiBLGCv94mxjIKMuqBAMgj9qRCNzKS3lNKofv3UYxfieP865rknQoGQ0IQoNG7tJxWw0gMDQQFIhQmSMAQDsTzNxWv353XkpxirgkM6xEMDMQA/xGUhQUQuYVLEACuJwDIC4QEkQIJltilElCwQeF1C923lpyrMhxCRNoAWkUVb1EUqMUVVBFYsdWkKAXLUp0WohAas1ILQVXxDl3xglyp10EYF8Ua0okd85EeAhBJzVEeXRFSwgghDVRAiQGCYskczYIUzsFTzxVudoGjjl3cUuCuN9Ehf4AmTVBKaJEyelFUl0QiVFE1hdSuPUAh7sAeA4GejVEr+mMAE3dRJZchWaQhvvmB5MPiGyaVcJzUDBLZLvXQSx0RL/vSHzxJMlUJMf4BLusRLg3BYikFf4KV821US01RNmJBR+zRRtqRNMYUJ36QIsehOvaSKicGKz9VoNrWHAcVW1rSJEWWLP4gs8vRXisBOyDiLv/iIwhhdriheJAFQAlUpBKUIJFVStyhKAGVHCsWNAwVOwIgY1ziJiVGJ5kQSexRYyYhNUiSOzYgsHOVR84hR1ah47Xhc2Zhet/SLlKJShiVRfhiC4uJSMLWIKHWQM4BU1hiQoEWMIlUSPtVRjPiF4YgJTaVM4mJUSEWGafBTHQmCw2eRTLh/01USW9X+VV+1jhyBVlQ1A1aFVfn4K2NVKY1gVjHpVWakkqvIkrCFkQ9FUoBGDHpFkxshVwRRV5awk6J0SIjUlE5ZGEZ5lAMJXyiRWB8oWMxVKYnFgu+2laWFlPkUX3xVkUZpWXAZlwsBj5YYcGNplh9UiWwpSuRUTnspLn3JTH/JLYHpSoNpLYX5LYcJLYmZLYv5LI05LY+JLJHJLJN5LJW5LJc5LyEFKnVABVSgBXXYbKbyXwEWBANWP4O1K6XAAgORAH1ng1CBKxaGYTCGekmyK3JAgnLGX6S5MC52m4lXU5ESCihgEMkWR79JRqdTN0V2ZAORZPWXm6pScQWRnECmcab+wnuBZiljVmtmhptIsiotZ2WjoH7YyYG6pzV8ZgCGWCmAJmjD6ViQojMVwAWroD5855voximYRgxCWCmcRgyeJp5HMit06Dyvlp6y+RSvYnpudym5hniT1ZmR0j2xmXtY8SrPVjeyWHjtJnzjySsYyp+ylm68kynsdnoieqAkuqAZem4nyikDx3YA4HvzmZe0UqJTCHqWwnFQx3YAaqBGMiw8qpz9qSlDFnWXcnMCkXM56lasCaMmapebAqRMailTRwxV16JF+qL72aPLKaECsXaZMnURWqHEmW9hiqQzuimEVymGl6attabcQ6VimqSaInrxdykQSqF1Sp9simz+Mfp5Y2opvHejmQJ8HkqkRWKkeOqmVgqcT0p91ueoRAKpuCep2kkqQwYAfUop88el0zmiq3IKq3AKeiMQCtAzqJqdTWcqBlimCbiADZg9gaqjRtc/C3FAGuoRRFkSHmgQIsibJoiC4YmpQ0KevIkQd/arG8GDVpSVHIGDXORFXOgq/3kQAXopZXRGwfqlqRIKvKoQviqjIZGFdrRKKLFHffRHZ5it7uKRAxGGm2KVNCWohEVJqbSQmbRJjPhJVMkpgTiIhUgppGQppoRKquRUuvqKJNGJ/joSpqiAqDiwsCKKw1RMEptO1BokdPmRzGgSvViPGEs0ubiL4vixQBL+svc4U58oj8foi/ZIjvNkR2SlkB6bmcSyjBMbEujojQUFitBSjgn1UTtrocX4sjp7hCPRjzR7srCyjx+1VOH6qPAosj8bEhGJCQhJtPqISg8pUwoJriz7Iy47ElYbkj2FkkFFryOJSiWpVDo1lE77sNoYsWmFk2t1EkE5k1JrKj1JKT95Vnubk46otBkZV3MlEFJ5VwTRlIF7LH71V4HFs9SChWfLEWBJrHc5XIqblD4qXHJZugICAKabuqq7uqzbuq77urAbu7I7u7Rbu7Z7u7ibu7q7u7zbu777u8AbvMI7vMRbvMZ7vMibvMq7vMzbvM77vNAbvdI7vdRbvdZie73Ym73au73c273e+73gG77iO77kW77me77om77qu77s277u+77wG7/yO7/0W7/2e7/4m7/6u7/827/++78AHMACPMAEXMAGfMAInMAKvMAM3MAO/MAQHMESPMEUXMFCEhAAOw==)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* warstw łączących - klasy typu `tf.keras.layers.MaxPooling2D`, `AveragePooling2D` itd.\n", "\n", "Warstwa łącząca dokonuje redukcji rozmiaru danych wejściowych (zielony rysunek) z filtrem (fioletowy kwadrat). Jako wynik zwraca nowe dane (żółty rysunek). Rozmiar filtra i parametry jego ruchu: `size`, `stride` i `padding` są hiperparametrami modelu. Dane pokryte przez filtr są redukowane do\n", "jednej wartości. Najczęściej spotykane algorytmy redukcji to:\n", "\n", "* maxpool - wybierany jest element o maksymalnej wartości\n", "* average pool - obliczana jest średnia elementów\n", "\n", "W sytuacji kiedy kanałów jest więcej niż jeden operacja \"pooling\" jest wykonywana dla każdego kanału oddzielnie." ] }, { "cell_type": "markdown", "metadata": { "id": "QJWTyyXfQ52u" }, "source": [ "\n", "\n", "![maxpool_animation.gif](data:image/gif;base64,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)" ] }, { "cell_type": "markdown", "metadata": { "id": "9T_hJjRYR0CD" }, "source": [ "### Definicja architektury sieci ze splotem (CNN)\n", "\n", "Proszę zdefiniować sieć neuronową o archiketurze w pełni połączonej (ang. fully connected). Sieć powinna mieć:\n", "* warstwę wejściową\n", "* warstwy ukryte złozone z par:\n", " * warstwa `Conv2D` o parametrach: `filters=nFilters, kernel_size=kernel_size, activation=tf.nn.relu`\n", " * wartwa `MaxPool` o parametrach `pool_size=pool_size`\n", "\n", "* warstwę spłaszającą \n", "* warstwę gęstą o parametrach `nNeurons, activation=tf.nn.relu`\n", "* wartę wyjściową z funkcją aktywacji ```sofmax``` \n", "\n", "Proszę:\n", "\n", "* dokończyć funkcję ```getModel(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, inputShape, outputWidth)``` która tworzy model o zdefiniowanej powyżej architekturze\n", "* w funkcji skompilowac model tak by:\n", " * jako algorytmu minimalizacji używał `adam`\n", " * jako funkcji straty używał `categorical_crossentropy`\n", " * jako metryki używał `accuracy`\n", "* obliczyć samodzielnie liczbę parametrów pierwszej warstwy \"MaxPool\" i porównać ją z wynikiem działania funkcji ```model.summary(...)```\n", "* porównać liczbę parametrów w obencym modelu i modelu z poprzednich zajęć\n", "* obejrzeć rysunek przedstawiający architekturę modelu, uzysdkany przy pomocy funkcji ```tf.keras.utils.plot_model(...)```" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[34mNumber of parameters for the first maxPool layer:\u001b[0m 320\n", "Model: \"model\"\n", "_________________________________________________________________\n", " Layer (type) Output Shape Param # \n", "=================================================================\n", " input_1 (InputLayer) [(None, 28, 28, 1)] 0 \n", " \n", " conv2d (Conv2D) (None, 26, 26, 32) 320 \n", " \n", " max_pooling2d (MaxPooling2D (None, 13, 13, 32) 0 \n", " ) \n", " \n", " flatten (Flatten) (None, 5408) 0 \n", " \n", " dense (Dense) (None, 128) 692352 \n", " \n", " dense_1 (Dense) (None, 10) 1290 \n", " \n", "=================================================================\n", "Total params: 693,962\n", "Trainable params: 693,962\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def getModel(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, inputShape, outputWidth):\n", " \n", " inputs = tf.keras.Input(shape=inputShape)\n", " x = inputs\n", " for iHidden in range(nHiddenLayers): \n", " x = tf.keras.layers.Conv2D(filters=nFilters, kernel_size=kernel_size, activation=tf.nn.relu)(x)\n", " x = tf.keras.layers.MaxPooling2D(pool_size=pool_size)(x)\n", " \n", " x = tf.keras.layers.Flatten()(x)\n", " x = tf.keras.layers.Dense(nNeurons, activation=tf.nn.relu)(x)\n", " outputs = tf.keras.layers.Dense(outputWidth, activation=tf.nn.softmax)(x)\n", " model = tf.keras.Model(inputs=inputs, outputs=outputs)\n", " model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n", " return model\n", "\n", "nFilters = 32\n", "kernel_size = 3\n", "pool_size = (2,2)\n", "nNeurons = 128 \n", "nHiddenLayers = 1 \n", "inputShape = (28, 28, 1)\n", "outputWidth = 10\n", "\n", "model_basic = getModel(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, inputShape, outputWidth)\n", "\n", "print(colored(\"Number of parameters for the first maxPool layer:\",\"blue\"),(kernel_size**2+1)*nFilters) \n", "\n", "model_basic.summary()\n", "tf.keras.utils.plot_model(model_basic, 'ML_model.png', show_shapes=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Trenowanie modelu\n", "\n", "Proszę:\n", "\n", "* przeprowadzić trening dla `10` epok z rozmiarem paczki wynoszącym `128`\n", "\n", "**Uwaga**: proszę pełączyć śrdowisko wykonawcze by używało karty graficznej \"GPU\": z menu na górze:\n", "```\n", "Środowisko wykonawcze -> Zmień typ środowiska wykonawczego\n", "```\n", "Oczekiwany efekt:\n", "```\n", "Epoch 1/15\n", "329/329 [==============================] - 4s 5ms/step - loss: 0.2923 - accuracy: 0.9166 - val_loss: 0.1296 - val_accuracy: 0.9632\n", "\n", "Epoch 15/15\n", "329/329 [==============================] - 1s 4ms/step - loss: 0.0033 - accuracy: 0.9992 - val_loss: 0.0650 - val_accuracy: 0.9852\n", "```\n", "\n", "Proszę:\n", "* porównać wartości metryki na danych uczących i walidacyjnych. Jaki wniosek wynika z tego porównania?\n", "* porównać metryki obecnego modelu z modelem z siecią gęstą. Jaki się sprawuje obecny model? \n", "\n", "**Wskazówka**: w czasie treningu proszę użyć parametru ```validation_split=0.3``` by uzyskać automatyczny podział danych uczących na uczące i walidacyjne." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/15\n", "329/329 [==============================] - 14s 41ms/step - loss: 0.2638 - accuracy: 0.9239 - val_loss: 0.1173 - val_accuracy: 0.9642\n", "Epoch 2/15\n", "329/329 [==============================] - 15s 44ms/step - loss: 0.0812 - accuracy: 0.9755 - val_loss: 0.0757 - val_accuracy: 0.9778\n", "Epoch 3/15\n", "329/329 [==============================] - 12s 38ms/step - loss: 0.0534 - accuracy: 0.9839 - val_loss: 0.0677 - val_accuracy: 0.9806\n", "Epoch 4/15\n", "329/329 [==============================] - 14s 44ms/step - loss: 0.0382 - accuracy: 0.9891 - val_loss: 0.0608 - val_accuracy: 0.9806\n", "Epoch 5/15\n", "329/329 [==============================] - 14s 44ms/step - loss: 0.0302 - accuracy: 0.9913 - val_loss: 0.0622 - val_accuracy: 0.9822\n", "Epoch 6/15\n", "329/329 [==============================] - 13s 38ms/step - loss: 0.0231 - accuracy: 0.9929 - val_loss: 0.0620 - val_accuracy: 0.9824\n", "Epoch 7/15\n", "329/329 [==============================] - 12s 37ms/step - loss: 0.0162 - accuracy: 0.9954 - val_loss: 0.0609 - val_accuracy: 0.9833\n", "Epoch 8/15\n", "329/329 [==============================] - 13s 40ms/step - loss: 0.0129 - accuracy: 0.9962 - val_loss: 0.0619 - val_accuracy: 0.9828\n", "Epoch 9/15\n", "329/329 [==============================] - 14s 42ms/step - loss: 0.0106 - accuracy: 0.9969 - val_loss: 0.0647 - val_accuracy: 0.9832\n", "Epoch 10/15\n", "329/329 [==============================] - 12s 38ms/step - loss: 0.0068 - accuracy: 0.9983 - val_loss: 0.0612 - val_accuracy: 0.9848\n", "Epoch 11/15\n", "329/329 [==============================] - 14s 41ms/step - loss: 0.0045 - accuracy: 0.9990 - val_loss: 0.0665 - val_accuracy: 0.9838\n", "Epoch 12/15\n", "329/329 [==============================] - 18s 55ms/step - loss: 0.0046 - accuracy: 0.9989 - val_loss: 0.0714 - val_accuracy: 0.9826\n", "Epoch 13/15\n", "329/329 [==============================] - 15s 44ms/step - loss: 0.0059 - accuracy: 0.9981 - val_loss: 0.0741 - val_accuracy: 0.9827\n", "Epoch 14/15\n", "329/329 [==============================] - 12s 37ms/step - loss: 0.0031 - accuracy: 0.9994 - val_loss: 0.0684 - val_accuracy: 0.9854\n", "Epoch 15/15\n", "329/329 [==============================] - 13s 40ms/step - loss: 0.0042 - accuracy: 0.9989 - val_loss: 0.0922 - val_accuracy: 0.9789\n", "CPU times: user 7min 35s, sys: 30.4 s, total: 8min 6s\n", "Wall time: 3min 26s\n" ] } ], "source": [ "%%time \n", "\n", "epochs = 15\n", "batch_size = 128\n", "\n", "model_basic_fit = model_basic.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size, validation_split=0.3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Analiza historii treningu\n", "\n", "Proszę uzupełnić funkcję ```plotTrainingHistory(model)``` tak by tworzyła wykresy:\n", "\n", "* na jednym wykresie wartości metryki w funkcji numeru epoki obliczone dla danych uczących i treningowych\n", "* na drugim wykresie funkcji straty w zależności od numeru epoki obliczone dla danych uczących i treningowych\n", "\n", "Wartości potrzebnych parametrów są dostępne w obiekcie ```Model.history```" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])\n", "[0.923880934715271, 0.9754762053489685, 0.9839047789573669, 0.9891190528869629, 0.9912857413291931, 0.9929285645484924, 0.9953809380531311, 0.9961904883384705, 0.9968809485435486, 0.9983095526695251, 0.9989761710166931, 0.9989047646522522, 0.998119056224823, 0.999404788017273, 0.9989285469055176]\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "print(model_basic_fit.history.keys())\n", "history = model_basic_fit.history\n", "print(history['accuracy'])\n", "\n", "def plotTrainingHistory(model):\n", "\n", " fig, axes= plt.subplots(1,2,figsize=(10,5))\n", " history = model.history\n", " axes[0].plot(history['accuracy'])\n", " axes[0].plot(history['val_accuracy'])\n", " axes[0].set_ylabel('accuracy')\n", " axes[0].set_xlabel('epoch')\n", " axes[0].legend(['train', 'validation'], loc='upper left')\n", "\n", " axes[1].plot(history['loss'])\n", " axes[1].plot(history['val_loss'])\n", " axes[1].set_ylabel('loss function')\n", " axes[1].set_xlabel('epoch')\n", " axes[1].legend(['train', 'validation'], loc='upper left')\n", " \n", " \n", "plotTrainingHistory(model_basic_fit) " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Analiza klasyfikacji\n", "\n", "Proszę dokończyć funkcję `printScores(model, X, Y)` która wypisuje na erkan:\n", "* raport klasyfikacji\n", "* macierz pomyłek\n", "\n", "i użyć tej funkcji do wypisania wydajności modelu `model_basic`.\n", "\n", "**Wskazówka**: proszę użyć funkcji `printScores(model, X, Y)` zdefiniowanej na innych ćwiczeniach.\n", "\n", "**Uwaga**: funkcje typu `sklearn.metrics.classification_report(...)` jako dane wejściowe przyjumują numery klas. Proszę użyć funkcji `numpy.argmax(...)` by z wektora \"one_hot\" uzyskać numer klasy o największym pradwopodobieństwie." ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 0.98 0.99 0.99 980\n", " 1 0.98 1.00 0.99 1135\n", " 2 0.99 0.98 0.98 1032\n", " 3 0.99 0.98 0.98 1010\n", " 4 0.98 0.99 0.98 982\n", " 5 0.98 0.98 0.98 892\n", " 6 0.99 0.97 0.98 958\n", " 7 0.97 0.99 0.98 1028\n", " 8 0.99 0.94 0.97 974\n", " 9 0.96 0.98 0.97 1009\n", "\n", " accuracy 0.98 10000\n", " macro avg 0.98 0.98 0.98 10000\n", "weighted avg 0.98 0.98 0.98 10000\n", "\n", "Confusion matrix:\n", "[[ 974 1 0 0 0 1 1 2 0 1]\n", " [ 0 1134 1 0 0 0 0 0 0 0]\n", " [ 2 9 1007 1 1 0 0 11 1 0]\n", " [ 1 1 2 988 0 7 0 3 2 6]\n", " [ 0 1 0 0 973 0 0 0 0 8]\n", " [ 2 0 1 8 0 872 3 0 0 6]\n", " [ 6 2 0 0 11 4 932 0 3 0]\n", " [ 0 2 5 0 0 0 0 1020 1 0]\n", " [ 7 5 3 2 4 5 1 6 919 22]\n", " [ 1 5 0 0 5 1 0 5 0 992]]\n" ] } ], "source": [ "def printScores(model, X, Y):\n", " Y_pred = np.argmax(model.predict(X), axis=1)\n", " # użyj classification_report() żeby policzyć najpopularniejsze miary \n", " print(\"Classification report:\")\n", " print(classification_report(Y,Y_pred))\n", " # wypisz macierz pomyłek \n", " print(\"Confusion matrix:\")\n", " print(confusion_matrix(Y, Y_pred))\n", " \n", "printScores(model_basic, X_test, np.argmax(Y_test, axis=1)) " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Ilustracja błędnie sklasyfikowanych przykładów." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "Y_pred = model_basic.predict(X_test).argmax(axis=1)\n", "missclassified_indices = Y_pred != np.argmax(Y_test, axis=1)\n", "missclassified_data = X_test[missclassified_indices,:,:,:]\n", "missclassified_labels = tf.boolean_mask(Y_test,missclassified_indices)\n", "\n", "def plotExamples(model, data, labels):\n", " fig, axes= plt.subplots(3,3,figsize=(10,10))\n", " for k, ax in zip(range(axes.size), axes.reshape(-1)):\n", " ax.imshow(data[k,:,:,0].reshape(28,28), cmap='gray', interpolation='none')\n", " Y_pred = model.predict(data[k:k+1,:,:,:])\n", " Y_pred = np.argmax(Y_pred, axis=1)[0] \n", " Y_true = np.argmax(labels[k]) \n", " ax.set_title(\"Predicted={}, True={}\".format(Y_pred, Y_true))\n", " \n", "\n", "plotExamples(model_basic, missclassified_data, missclassified_labels)" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "## Model z warstwą \"dropout\" oraz regularyzacją L2\n", "\n", "Proszę:\n", "\n", "* napisać funkcję ```getModelWithDropout(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, inputShape, outputWidth)```\n", "która zawiera warstwę `tf.keras.layers.Dropout()` po warstwie `MaxPool2D`. Dodatkowo parametry modelu powinny mieć regularyzację L2.\n", "* stworzyć nowy model `model_with_reg = getModelWithRegularisation(...)`\n", "* wytrenować model\n", "* wypisać jego metryki i sprawdzić działanie na **tych** samych przykładach, które były błędnie sklasyfikowane przez poprzedni model.\n", "\n", "Jak wyglądają krzywe metryk dla zbiorów uczącego i walidacyjnego w funkcji numeru epoki? \n", "\n", "Czy z ich zachowania można wyciągnąć jakiś wniosek odnośnie czasu treningu?" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"model_1\"\n", "_________________________________________________________________\n", " Layer (type) Output Shape Param # \n", "=================================================================\n", " input_2 (InputLayer) [(None, 28, 28, 1)] 0 \n", " \n", " conv2d_1 (Conv2D) (None, 26, 26, 32) 320 \n", " \n", " max_pooling2d_1 (MaxPooling (None, 13, 13, 32) 0 \n", " 2D) \n", " \n", " dropout (Dropout) (None, 13, 13, 32) 0 \n", " \n", " flatten_1 (Flatten) (None, 5408) 0 \n", " \n", " dense_2 (Dense) (None, 128) 692352 \n", " \n", " dense_3 (Dense) (None, 10) 1290 \n", " \n", "=================================================================\n", "Total params: 693,962\n", "Trainable params: 693,962\n", "Non-trainable params: 0\n", "_________________________________________________________________\n", "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 0.99 0.99 0.99 980\n", " 1 0.99 0.99 0.99 1135\n", " 2 0.98 0.98 0.98 1032\n", " 3 0.98 0.98 0.98 1010\n", " 4 0.99 0.98 0.98 982\n", " 5 0.94 0.99 0.97 892\n", " 6 0.97 0.98 0.98 958\n", " 7 0.97 0.97 0.97 1028\n", " 8 0.99 0.95 0.97 974\n", " 9 0.97 0.96 0.97 1009\n", "\n", " accuracy 0.98 10000\n", " macro avg 0.98 0.98 0.98 10000\n", "weighted avg 0.98 0.98 0.98 10000\n", "\n", "Confusion matrix:\n", "[[ 966 0 1 0 0 2 7 2 2 0]\n", " [ 0 1128 2 1 0 2 2 0 0 0]\n", " [ 0 3 1016 4 2 0 0 6 1 0]\n", " [ 0 0 3 989 0 12 0 4 1 1]\n", " [ 0 0 0 0 960 0 5 1 1 15]\n", " [ 1 0 0 4 0 881 4 1 1 0]\n", " [ 3 2 0 0 2 8 943 0 0 0]\n", " [ 1 4 16 4 1 1 0 996 0 5]\n", " [ 1 0 1 7 3 15 8 5 921 13]\n", " [ 2 4 0 4 6 12 1 7 0 973]]\n", "CPU times: user 9min 10s, sys: 39.2 s, total: 9min 49s\n", "Wall time: 3min 21s\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "%%time \n", "\n", "def getModelWithRegularisation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth):\n", " \n", " inputs = tf.keras.Input(shape=inputShape)\n", " x = inputs\n", " for iHidden in range(nHiddenLayers): \n", " x = tf.keras.layers.Conv2D(filters=nFilters, \n", " kernel_size=kernel_size, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.relu)(x)\n", " x = tf.keras.layers.MaxPooling2D(pool_size=pool_size)(x)\n", " x = tf.keras.layers.Dropout(dropout_rate)(x)\n", " \n", " x = tf.keras.layers.Flatten()(x)\n", " x = tf.keras.layers.Dense(nNeurons, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.relu)(x)\n", " outputs = tf.keras.layers.Dense(outputWidth, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.softmax)(x)\n", " model = tf.keras.Model(inputs=inputs, outputs=outputs)\n", " model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n", " return model\n", "\n", "dropout_rate = 0.2\n", "l2_lambda = 0.001\n", "\n", "model_with_reg = getModelWithRegularisation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth)\n", "model_with_reg.summary()\n", "model_with_reg_fit = model_with_reg.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size, validation_split=0.3, verbose=0)\n", "plotTrainingHistory(model_with_reg_fit) \n", "printScores(model_with_reg, X_test, np.argmax(Y_test, axis=1)) \n", "plotExamples(model_with_reg, missclassified_data, missclassified_labels)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Krzywa dokładności dla danych walidacyjnych nie odbiega od krzywej dla danych treningowych po 15 epokach. To wskazuje, że trening można przedłużyć.\n", "Proszę wytrenować model przez kolejne 20 epok i przeprowadzić na nim sprawdziany jak poprzednio." ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 0.98 0.99 0.99 980\n", " 1 0.99 0.99 0.99 1135\n", " 2 0.98 0.98 0.98 1032\n", " 3 0.96 0.99 0.98 1010\n", " 4 0.99 0.99 0.99 982\n", " 5 1.00 0.96 0.97 892\n", " 6 0.99 0.97 0.98 958\n", " 7 0.97 0.98 0.98 1028\n", " 8 0.97 0.98 0.98 974\n", " 9 0.99 0.97 0.98 1009\n", "\n", " accuracy 0.98 10000\n", " macro avg 0.98 0.98 0.98 10000\n", "weighted avg 0.98 0.98 0.98 10000\n", "\n", "Confusion matrix:\n", "[[ 969 0 1 0 0 0 3 1 3 3]\n", " [ 0 1128 5 0 0 0 1 0 1 0]\n", " [ 3 2 1014 1 1 0 0 9 2 0]\n", " [ 0 0 1 1003 0 0 0 5 1 0]\n", " [ 0 1 1 0 972 0 1 0 2 5]\n", " [ 1 0 1 25 0 852 4 1 7 1]\n", " [ 6 3 1 1 4 3 934 0 6 0]\n", " [ 1 3 12 1 0 0 0 1009 1 1]\n", " [ 1 0 2 4 2 1 0 6 954 4]\n", " [ 3 6 1 5 4 0 0 7 3 980]]\n", "CPU times: user 12min 33s, sys: 57.4 s, total: 13min 30s\n", "Wall time: 4min 11s\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "%%time\n", "epochs = 20\n", "model_with_reg_fit = model_with_reg.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size, validation_split=0.3, verbose=0)\n", "plotTrainingHistory(model_with_reg_fit) \n", "printScores(model_with_reg, X_test, np.argmax(Y_test, axis=1)) \n", "plotExamples(model_with_reg, missclassified_data, missclassified_labels)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Generacja dodatkowych danych (ang. data augmentation)\n", "\n", "Data augmentation polega na generowaniu dodatkowych danych z naszego zestawu danych, bazując na prostych transformacjach - rotacjach, przerzucaniu, podkreślaniu krawędzi, itd. To muszą być transformacje, o których wiemy, że nie powinny wpływać na informacje zawarte w naszych danych. \n", "Możemy w ten sposób np. nauczyć maszynę, że tożsamośc obiektów na rysunkach nie zależy od ich orientacji. \n", "\n", "Dodatkowe dane mogą być uzyskane na różne sposoby. Jednym z nich jest dodanie do modelu warstwy która będzie losowo modyfikowała podany przykład.\n", "Tak warstwa jest aktywna **tylko** w czasie treningu. Lista dostępny modyfikacji w pakiecie TensorFlow jest \n", "[tutaj](https://www.tensorflow.org/guide/keras/preprocessing_layers).\n", "\n", "[Tutaj](https://machinelearningmastery.com/image-augmentation-deep-learning-keras/) \n", "znajdują się przykłady zastosowania różnych transformacji.\n", "\n", "\n", "Proszę:\n", "* narysować pierwszy przykład ze zbioru uczącego po losowym obrocie w zakresie $\\pm$0.05 korzystając z warstwy `tf.keras.layers.RandomRotation()`" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# from tf.keras.layers import *\n", "image = RandomRotation( (-0.05, 0.05) )(X_test[0,:,:,:]) \n", "plt.imshow(image, cmap=plt.get_cmap('gray'));" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Trening z użyciem \"data augmentation\"\n", "\n", "Proszę:\n", "\n", "* napisać funkcję, `getModelWithAugmentation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth)`\n", " która tworzy model z warstwą \"data augmentation\", oraz pozostałymi warstwami takimi jak tworzone przez funkcję `getModelWithRegularisation(...)`\n", "* stworzyć nowy model `model_with_aug = getModelWithAugmentation(...)`\n", "* wytrenować model\n", "* wypisać jego metryki i sprawdzić działanie na **tych** samych przykładach, które były błędnie sklasyfikowane przez poprzedni model.\n", "\n", "A następnie wytrenować model, wypisać jego metryki i sprawdzić działanie na **tych** samych przykładach, które były błędnie sklasyfikowane przez model `model_basic`. \n", "\n", "**Czy zastosowana metoda wzbogacania danych poprawia wydajność modelu?**" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"model_2\"\n", "_________________________________________________________________\n", " Layer (type) Output Shape Param # \n", "=================================================================\n", " input_3 (InputLayer) [(None, 28, 28, 1)] 0 \n", " \n", " random_rotation_1 (RandomRo (None, 28, 28, 1) 0 \n", " tation) \n", " \n", " conv2d_2 (Conv2D) (None, 26, 26, 32) 320 \n", " \n", " max_pooling2d_2 (MaxPooling (None, 13, 13, 32) 0 \n", " 2D) \n", " \n", " dropout_1 (Dropout) (None, 13, 13, 32) 0 \n", " \n", " flatten_2 (Flatten) (None, 5408) 0 \n", " \n", " dense_4 (Dense) (None, 128) 692352 \n", " \n", " dense_5 (Dense) (None, 10) 1290 \n", " \n", "=================================================================\n", "Total params: 693,962\n", "Trainable params: 693,962\n", "Non-trainable params: 0\n", "_________________________________________________________________\n", "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 0.97 0.99 0.98 980\n", " 1 0.99 1.00 0.99 1135\n", " 2 0.99 0.97 0.98 1032\n", " 3 0.98 0.99 0.98 1010\n", " 4 0.99 0.99 0.99 982\n", " 5 0.98 0.99 0.98 892\n", " 6 0.99 0.98 0.99 958\n", " 7 0.97 0.97 0.97 1028\n", " 8 0.99 0.97 0.98 974\n", " 9 0.97 0.98 0.98 1009\n", "\n", " accuracy 0.98 10000\n", " macro avg 0.98 0.98 0.98 10000\n", "weighted avg 0.98 0.98 0.98 10000\n", "\n", "Confusion matrix:\n", "[[ 973 0 1 0 0 1 2 2 1 0]\n", " [ 0 1131 1 2 0 0 1 0 0 0]\n", " [ 7 4 1002 4 1 0 0 10 4 0]\n", " [ 0 0 1 996 0 5 0 6 1 1]\n", " [ 0 0 0 0 976 0 0 0 0 6]\n", " [ 2 0 0 7 0 879 3 0 1 0]\n", " [ 8 2 0 1 3 6 937 0 1 0]\n", " [ 1 5 6 0 3 0 0 1000 3 10]\n", " [ 5 0 2 1 2 6 0 4 943 11]\n", " [ 4 3 0 2 4 2 0 4 0 990]]\n", "CPU times: user 13min 6s, sys: 56.6 s, total: 14min 2s\n", "Wall time: 4min 33s\n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmcAAAE9CAYAAABOT8UdAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAABcd0lEQVR4nO3deXiU5bn48e89k0z2PSEsISFAEFCQJSAuCGpVtK37XluXVlqrR3vanlM9PaftsfZo1bbWX+2Cu7Zq1dZKW6xaixu4EARU9jUQ1kASyL7evz/eN8kQAiRhJpOZuT/X9V7zzrvNM4k83nm2W1QVY4wxxhgzMHhCXQBjjDHGGNPJgjNjjDHGmAHEgjNjjDHGmAHEgjNjjDHGmAHEgjNjjDHGmAHEgjNjjDHGmAEkJtQFCJTs7GwdMWJEqIthjOlHS5cu3auqOaEuRyBYHWZMdDlS/RUxwdmIESMoKSkJdTGMMf1IREpDXYZAsTrMmOhypPrLujWNMcYYYwYQC86MMcYYYwYQC86MMcYYYwaQiBlz1p3m5mbKyspoaGgIdVEiRnx8PHl5ecTGxoa6KMZEPKvDAsvqLxMuIjo4KysrIyUlhREjRiAioS5O2FNV9u3bR1lZGYWFhaEujjFBJSJzgF8CXuBRVb23y/nrgfuB7e6hX6nqo+65VuBT9/hWVb2gL2WwOixwrP4y4SSig7OGhgar1AJIRMjKyqK8vDzURTEmqETECzwMnA2UAUtEZL6qrupy6R9V9dZuHlGvqpOOtRxWhwWO1V8mnET8mDOr1ALLfp4mSkwHNqjqJlVtAp4HLgxFQezfXODYz9KEi6AGZyIyR0TWisgGEbmjm/MFIvKmiHwiIm+JSJ7fuftEZKWIrBaRhyRM/1VVVVXx61//utf3nX/++VRVVQW+QMaYnhgGbPN7X+Ye6+pSt/56SUSG+x2PF5ESEflARC4KZkGDyeovY0IjaMGZX7fAecB44GoRGd/lsgeAp1V1InAXcI977ynAqcBE4ARgGjArWGUNpsNVbi0tLUe8b8GCBaSnpwepVMaYAPgrMMKtv94AnvI7V6CqxcA1wIMiMqq7B4jIXDeIKxmI3W1WfxkTGsEcc9bRLQAgIu3dAv5jNsYD33b3FwJ/cfcViAd8gACxwO4gljVo7rjjDjZu3MikSZOIjY0lPj6ejIwM1qxZw7p167jooovYtm0bDQ0N3H777cydOxfoXC28pqaG8847j9NOO43FixczbNgwXnnlFRISEkL8zYw5jIb9sL/M2WrLIT4dkrIhMdt5jU+Dgd8Qvh3wbwnLo3PgPwCqus/v7aPAfX7ntruvm0TkLWAysLHrh6jqPGAeQHFxsfakYA3NrdQ2tpCR5MMT5J+j1V/GhEYwg7PuugVO6nLNCuASnBlRFwMpIpKlqu+LyEJgJ05w9itVXR3EsgbNvffey2effcby5ct56623+PznP89nn33WMVvo8ccfJzMzk/r6eqZNm8all15KVlbWQc9Yv349zz33HI888ghXXHEFf/rTn7j22mtD8XVMtGtrhZrdULUN9rdvbiBW5e437j/yMzyxkJjlBGr+QVv76+izID2/f77P4S0BikSkECcouwqnFayDiAxR1Z3u2wuA1e7xDKBOVRtFJBunF+A+AqS2sYXtVfWkxsfiiQlucGb1lzGhEerZmt8FfuVOSX8HpxJsFZHRwDicv1YB3hCRmar6rv/NIjIXmAuQn3/kyvx//7qSVTsOBLTw44em8sMvHt+re6ZPn37QNO6HHnqIl19+GYBt27axfv36Qyq3wsJCJk2aBMDUqVPZsmXLMZXbRIi6CtizGsrXdG6+ZBj7BTjuPEjMDMzn1JTD6vmw6i9Q+j60NR98Pj4d0oY7AdWIUyEtz3mfNhySspyWtNp9ULcXave6r+WdxypLneNN1c7zrv5jyIMzVW0RkVuB13CW0nhcVVeKyF1AiarOB24TkQuAFqACuN69fRzwOxFpwxk6cm83szx7rb0Oa21TGppbSfB5j7nlrLd1mNVfxvSPYAZnPekW2IHTcoaIJAOXqmqViNwEfKCqNe65V4GTgXe73N/rLoFQS0pK6th/6623+Oc//8n7779PYmIis2fP7naxybi4uI59r9dLfX19v5TVDBC1+9zgazWUr+0MyGr9xij5UiBnDFRshrULQLww4jQY90UnWEsd0rvPrNnjBGQr/wKli0DbIGs0nPR1yBzpBE9pec4WlxKY79ncAHX7ICE9MM87Rqq6AFjQ5dgP/PbvBO7s5r7FwISgFUzaP6dzv79Y/WVM/whmcNaTboFsoEJV23AqucfdU1uBm0TkHpzqZxbw4LEUprctXIGSkpJCdXV1t+f2799PRkYGiYmJrFmzhg8++KCfS2eOSWMN1FdAU6271TjH2vf9j7e/b20GbYW2Nve11e+1rcv71s5xW+18KZBzHIw5F3LGQc5YGDQWUoc547hUYedyWP1XWDUfFnzX2fKmO4HauC9C5mEW4Kze7baQveIXkBXBzO/A8RfDoPHBHSsWGw9p3U2INNBZhzU0t7JudzX5mYmkJ/qC+plWfxkTGkELznrYLTAbuEdEFKdb8xb39peAM3FW2FbgH6r612CVNZiysrI49dRTOeGEE0hISCA3N7fj3Jw5c/jtb3/LuHHjOO6445gxY0YIS2p6rK4C3r4PljwCbUeetQZAbBL4ksCXCF6f06rl8YJ4nFdPTOcxjxc8ftfkHu8EYYPGOoFYexB2OCIwdLKznfUDp6Vt9XwnWHvjf5xt8AQYd4ETqCVkdgZkW94DFLLHwMzvwvEXBT8gM73m9Ti/j5a24HcWWP1lTGiIalj0Bh5VcXGxlpSUHHRs9erVjBs3LkQlilxR+3NtbYaSJ+Ct/3PGUU2+Foaf5AZeye5r+777PjYRPANkrefKUljzNydQ2/oBzt89ruzjnGBs/EUwaFzYBGQistRdsiLs9bQOa1Pls+37yU2NJzc1vj+LGBGitv4yA86R6q9QTwgwJjys/ye89l+wdy0Ung7n3gODTwh1qXonowBOvsXZqnfD2r87rYBjP+8EZCYseETweqRfWs6MMaFhwZkxR1K+Fl77Pmx4wxkIf9VzzkzIMGlZOqyUXCi+MdSlMH0U4/HQ2mrBmTGRyoIz00nbnEHrDQegsdoZH5WYBfEpzvioaFJXAW/dA0sec7ooz/kJTJ8LMcEdgG1MTzgtZ22hLoYxJkgsOIt2rS3QeMAZQ9VY7cwQRJzxUs21ULnfGbCekOmsmxUb4St7tzbDkkedwKyx2mldmn2nsziqMQNEjEdobrXgzJhIZcFZtFGFlgYnGGs44ARg4AZg6RCX6qxb5fE6LWmN1c7aU7XlULvHGeDeWAP1VcFZj6ql0VlpvmoLVG5xBrFXlTr7LU3OLMRhUyCvGHJPAG9sYD63uR42LoQ3fgD71sPIM+Dc/4PcrulgjQk9r0eob7ZuTWMilQVn0aKxBuornaCsfYX32ERIGQxxaU6LWNdxVOJx8iDGpzktSvWVTqBWXwE/O85Z3HTytVA4q2czEltbnPvbV4o/sKMz8GoPwg7s4KBZhF4fpBc4g9nFA+tfhxXPOudi4mHIiTBsaueWMeLw48FUoXqXE3ztdbd962HvOicgRJ2FVq95AYrOCf9xZSZixXiF1jZFVRH779SYiGPB2QCTnJxMTU0NO3bs4LbbbuOll1465JrZs2fzwAMPUFx8+BUEHnzwQebOnUtiYiK0NHL++XN49uF7SB80zGkdi0/rXauTNxaSB0FSDuxpgslfhk9fhM9ectL0nHi1szhq3T4n8Kotd4OwfZ379ZXdPFggdagTgBXOcoKw9AInyMoogOTBBwd+qlC1FbaXwPaPoawESh6HD37tnE/M7gzUMkc6gV97ALZ3Q2eKIHCC06zRzgKtk77kzFg87vzAtcYZEyRej9CmSpuCdwDFZkGpv4Dzzz+fZ599lvT09GAV3ZgBxYKzAWro0KHdVmw99eCDD3Lttdc6lVvDfhY88/+cBUVj4o5+85GIOK1Zn38AzrnbSRW07Pfwzv10tHiJxxmjlpTtBHO5xx+c2Lp9P2UIpA/vXZlEnKAtowBOuNQ51toMe1Y5gdr2pc62/vXO8qTmQXYRTLraWWA1a7TzmjrUWsdMWIpx/2BpbWvD6/GGuDSHCmj9BSxYsOAodxgTWSw4C7I77riD4cOHc8stTvKDH/3oR8TExLBw4UIqKytpbm7m7rvv5sILLzzovi1btvCFL3yBzz77jPr6em644QZWrFjB2LFjD8pNd/PNN7NkyRLq6+u57LLL+N///V8eeughduzYwRlnnEF2djYLX3yEETO+QMnHK8jOjuPnP/85jz/uZMr62te+xre+9S22bNnCeeedx2mnncbixYsZNmwYr7zyCgkJR5gAEBsPJ1zibNW7nHFoSdmQkOGMWesv3line3PIiTDtq86xhv2wf7sTxPmSjny/MWEmxi9LQDDnDw+I+mvhQkaMGEFJSQnZ2dmBq7+MGchUNSK2qVOnalerVq065Fh/+/jjj/X000/veD9u3DjdunWr7t+/X1VVy8vLddSoUdrW1qaqqklJSaqqunnzZj3++ONVVfVnP/uZ3nDDDaqqumLFCvV6vbpkyRJVVd23b5+qqra0tOisWbN0xYoVqqpaUFCg5eXlqq0tqtuXacHwPC0vL9eSkhI94YQTtKamRqurq3X8+PH68ccf6+bNm9Xr9eqyZctUVfXyyy/XZ555ptvvNBB+rsaoquKkggt5/ROIrTd1WE1Ds67YVqkH6pt69wPrpZDXX67291Z/mUhypPorelrOXr0Ddn0a2GcOngDn3XvESyZPnsyePXvYsWMH5eXlZGRkMHjwYP793/+dd955B4/Hw/bt29m9ezeDBw/u9hnvvPMOt912GwATJ05k4sSJHedeeOEF5s2bR0tLCzt37mTVqlUHnaexGtCOdcree+89Lr74YpKSnNakSy65hHfffZcLLriAwsJCJk2aBMDUqVPZsmVL334uxvSQqlJV18zu6gZ27W9gz4FGTivKZmi6tXgcwq8OS1BlZFMrcbGeY0sPdpQ6LOT1VxdWf5loET3BWQhdfvnlvPTSS+zatYsrr7ySP/zhD5SXl7N06VJiY2MZMWIEDQ0NvX7u5s2beeCBB1iyZAkZGRlcf/31hz6n8YCTRLsHY6vi4jrHfnm93oO6H4zprYbmVsoq69h9oJHdBxo6XvdU++0faKSpy3pdv712igVnR9Hxz7kfVtMIaf3VC1Z/mUgSPcHZUVq4gunKK6/kpptuYu/evbz99tu88MILDBo0iNjYWBYuXEhpaekR7z/99NN59tlnOfPMM/nss8/45JNPADhw4ABJSUmkpaWxe/duXn31VWbPng1ASkoK1QcOkJ1Y56xb5po5cybXX389d9xxB6rKyy+/zDPPPBO07276T11TC3VNrWQnH+Okj15QVfZUN7KxvIaN5bVsKq9hU3ktm/bWsL2ynq7pH1PiYhiUGkduajzTRmQ6+ynxDE6LJzc1jkEplsz7sPzrMFU2b99PjvuzC6aQ1V/V1WRnH7z4s9VfJlpET3AWQscffzzV1dUMGzaMIUOG8KUvfYkvfvGLTJgwgeLiYsaOHXvE+2+++WZuuOEGxo0bx7hx45g6dSoAJ554IpMnT2bs2LEMHz6cU089teOeuXPnMmfOHIZmp7Hwn//oOD5lyhSuv/56pk+fDjgDaidPnmxdAEHS2NLKZ9v3U7Klkk/K9jMxL42rT8onNT5wy3Xsr2/miUWbeey9zVQ3tJCV5GNMbgrHDU5xX5MZk5tCSh8/s6mlraO1a0dVfUfwtckNxmqbWjuuTYj1UpidxIl56Vw8OY/C7ESGpCWQmxrPoJQ4kuKsygkEEcHr8fRLCqeQ1l9Dh7Jw4cKO41Z/mWghzpi08FdcXKwlJSUHHVu9ejXjxo0LUYkGgOpdUL0zsCvpYz/XI9lf18zHWytZsqWCki2VLC+roqnF+R/o4NR4dh1oICUuhi/NKODGU0cw6BhaifbXN/P4e5t5fJETlJ0zPpfphZls2FPD2t3VrNtVfVDgNCw9gTG5yYwZnMJxbvCWmeRjT3u3Y3Ujew40dOmCbKSitumgzxWBoWkJjMxJYlROMiNzkhiZ7bwOTo3H4+m/5UlEZKmqHn7BrDDS2zps7a5q4mM9FGTZbOTesPrLDBRHqr/sz9hI1rDfWWjVFlU9qg17anht5S68HiEpLobkOC9JvhiS42JIcjdn3znu8Qiqyvaqekq2dAZj6/ZUo+osdXDCsDS+MqOA4hGZFI/IIDs5jk/KqvjdO5uY985GHn9vM5dMGcZNp49kVE5yj8u6v66ZxxZt5on3NlPd2MK5x+dy21lFHD807aDr2tqc8q3bXc3a3dWs3eVs723YS3Nr93+UeQRyUpxux7yMRKYWZJCb6nY5psYzJC2eEVlJxMcOvLW1ok2MR2jp2m9sjIkIFpxFqtYWaK5zVtg33VJV3t+0j8fe3cyba/b06t5En5dYr4f99U4qrOS4GKYUZPCFiUMoHpHJpOHpJPgODWAm5qXz8DVT2LK3lkff28SLJWX8sWQb54zP5euzRjElP+Own7m/rpnH3tvEE4u2UN3YwpzjB3PbWUWMH5ra7fUejzA8M5HhmYmcNS6343hzaxul+2pZu6uGqvomct1xXrmpcWQlx+Htx5Yv03cxXqGxxZKfGxOJLDiLVI0HnNf47v/HHc2aW9v42yc7ePTdzazccYCsJB/f+lwRXzqpgESfl9rGFmoaW6htbHVfW6htaunYr2lspbaxhYbmVo4bnEJxQSbHDU7pVVAzIjuJuy+awLc+N4anFm/h6fdLeW3lbqYXZvKNWSM547hBHTkTq+qaeOy9zTzpBmXnneAEZeOG9O13G+v1MHpQCqMHpRz9YjNgeT1Cy2FaQI0x4S3igzON1sTADQfAE+N0awZQOI9R3F/XzHNLtvLkoi3sOtDA6EHJ3HvJBC6aPOygbrqkuBgG9VOZspPj+M45x/GNWaN4fsk2Hnt3Ezc+WcJxuSl8bWYhWyvqeGLRFmoaWzh/wmD+7cy+B2UmPB2uDovxWPLz3grn+stEl4gOzuLj49m3bx9ZWVkDv/JSDVyeR1Wn5Sw+LaC5I1WVffv2ER8fXksdbN1Xx+OLNvNCyTbqmlo5dXQW91wygVljcvp18PqRJMXF8NXTCvnKyQX8dcUOfvf2Jv7jJWfJgfMnOC1lYwdbUBZtjlSHeT0eFKVVlZiBXr8NAOFaf5noFNHBWV5eHmVlZZSXl4e6KI62NmhrhraWzq21xTkGkDLYae06Vi2NULMbkpphV92xP89PfHw8eXl5AX1mMLS2Kcu3VfLou5s7Bvp/8cShfPW0wkMGzg8ksV4Pl0zJ4+LJw/hocwWZST6Kcq37MVodqQ6ra2qhorYZqYojxnsMWQKiSLjUX8ZEdHAWGxtLYWFh/39wYzWsWQAVG2HfRqjY5GwNVX4XCaQNh8xCZ1v2e5g+F+bcc+yf/8YP4f1fwX9uclrPIkxDcyu79jew60DDQa+7DzSw033dU91Ia5uSGh/D12eN4rqTRwR9sc5AEhFOGpkV6mKYEDtSHbZw7R5uem4Jf/7mKUw4wkQSY0z4iejgLGTeeQAWPejks0zLg8xRcMKlkDkSskY5r+kFEOsXLDTVwcdPw+w7jj2gWv865J8c1oFZU0sbm/fWsm53Net3V7Nudw1b9tWy60ADVXXNh1yfHBdDbmocQ9ISGDUqm8FpcYzISuL8CUNs4VMTkTITfQBUdlmHzhgT/uz/WsGw5V3ImwbX/x1iephK5+RvwqcvOAHaKf/W98+u2gZ7VsE5d/f9Gf2oubWNLXtrWbe7xgnE9riB2N7ajjWcPAIjspIYmZNE8YgMBqfGMzgtwX111uTq6+r3xhyOiMwBfgl4gUdV9d4u568H7ge2u4d+paqPuueuA/7bPX63qj4V6PJlJjnBWddFgo0x4c+Cs0BrrIGdK+DU23semAEMnQwFp8EHv4WTbgZvH381619zXovO7dv9fXSgoZk9Bxo6lpnoWILCb+mJg5elaGXX/no2763tWBBVBAoyEynKTeHc43MZk5tC0aAURubYoqemf4mIF3gYOBsoA5aIyHxVXdXl0j+q6q1d7s0EfggU46QmX+reWxnIMma4wVllnQVnxkQaC84CrWyJM9C/4JTe33vyLfD81bD6FacbtC/WvQ4ZIyC7qG/390Brm7J+TzXLtlaxbGsly7ZWsaG8hiPNUvcIJPnaV9v3khwXQ767OOqY3GSKBqUwelCyBWFmoJgObFDVTQAi8jxwIdA1OOvOucAbqlrh3vsGMAd4LpAFTPJ58Xk9VNQe2s1vjAlvFpwFWuliZ6zZ8JN6f++YOc74tMW/guMv6f0yGM31sPkdmPLlgC6hsbemkeVbq1i2zQnEVmyr6sjZmJEYy+T8DC44cSj5WYkd6Y460x45gVhCrHfgL2diTKdhwDa/92VAd/+oLxWR04F1wL+r6rbD3Dss0AUUETKSYm3MmTERyIKzQCtdDENOhLg+LH/g8Thjz/7+Hdj6ARSc3Lv7t7wHLfXH3KV5oKGZv63YyYeb97FsaxVbK5zlOGI8wrghqVw6NY/J+elMHp5BQVaiBV0mWv0VeE5VG0Xk68BTwJm9eYCIzAXmAuTn5/e6ABmJPiqsW9OYiGPBWSC1NDrdmtNv6vszTrwa/nW3sxRGb4Ozda85GQFGnNbrj1VVVpTt59kPS/nrip3UN7cyODWeKQXpfHlGAZPz0zlhWJp1O5posR0Y7vc+j86B/wCo6j6/t48C9/ndO7vLvW919yGqOg+YB1BcXNzr5eszk3zWcmZMBLLgLJC2fwytjX0bb9bOlwTFN8K7P3fWRssc2bP7VJ3JAIWzDl6i4yiqG5p5ZfkOnv1wK6t2HiDR5+XCSUO55qR8Jual9+07GBP+lgBFIlKIE2xdBVzjf4GIDFHVne7bC4DV7v5rwP+JSPviY+cAdwajkBlJPlbvPBCMRxtjQsiCs0AqXeS85veyxauraTfBooecmZvn33f06wHK10LVVjjt33t0+SdlVTz30VZeWb6DuqZWxg9J5e6LTuDCSUNtWQoT9VS1RURuxQm0vMDjqrpSRO4CSlR1PnCbiFwAtAAVwPXuvRUi8mOcAA/grvbJAYGWmWgtZ8ZEIgvOAql0MQwaD4mZx/ac1CEw4TIna8AZd0JCD1b/7lhC45zDXlLT2ML85Tt49qNSPtt+gIRYL188cQjXnFTAiXlpNnbMGD+qugBY0OXYD/z27+QwLWKq+jjweFALiNNyVlXfTGub4h0geWKNMcfOgrNAaW2BbR/CiVcF5nkn3wIrnoOlT/asNWzd65B7gpORwE99Uyvvb9rLG6t2M3/5DmqbWhk7OIUfX3g8F04eRqq1khkTtjITY1GF/fXNHYvSGmPCnwVngbLrE2iqObbxZv4GT3DGj334O5hxC8QcoeKtr4Kt7zsL3wJb99XxrzW7Wbi2nPc37aOppY1En5fzJwzhmpPymTw83VrJjIkAGR1ZAhotODMmglhwFiili53X/AAFZwAn3wrPXg6r/gITrzjsZc3r/0mstvJE+Rie+dlbbCqvBWBkdhLXnlTAGWNzmF6YSVyMzbQ0JpJ0pnCyhWiNiSQWnAVK6WJnZmXqkMA9c/TnIHuMs6zGhMsPWlh25/56Fq4pZ+HaPXx+41PMIpmfrkxl+shEvjyjgDOOG8SI7KTAlcUYM+BkJFp+TWMikQVngdDWBlsXw9jPB/a5Hg/M+Cb87VvOTFB3/bJnPijlh698RptCXlocv4j9hJqhs/n42nNJ9Nmv1JhokWn5NY2JSJ5QFyAilK+B+kooODXwzz7xKkjMgvcfRlV56M31/M9fPuOM4wbx+r+fzrvXppHcUsXg4gstMDMmyljLmTGRyf5vHgjt65sFajKAv9gEmPY19O37+NWL/+DnH7dx6ZQ8fnrpBGK8HvjX604uz9GfC/xnG2MGtASfl4RYr611ZkyEsZazQChdDKnDIL0gKI9vnnIjrRJD+ieP8rXTCrn/solOYAaw/nXIm3bsa6sZY8JSZpLl1zQm0ljL2bFSdYKzwpkHDdgPlPqmVm7+81bOaz6Vq3zvEnNGLtK+2GT1Lti5HM78n4B/rjEmPFh+TWMij7WcHauKTVCzKyhdmvvrmrn2sQ95Z105KWfcRmxbI7L0ic4L1r/hvI45N+CfbYwJDxlJPirqbCkNYyKJBWfHqn19swBPBth9oIErfvc+n5bt5+FrpnD+WWfBqDPho3nQ0uhctP41pzs194SAfrYxJnxkJsZay5kxESaowZmIzBGRtSKyQUTu6OZ8gYi8KSKfiMhbIpLndy5fRF4XkdUiskpERgSzrH1WutiZTZk9JmCP3LK3lst+u5iyyjqeuGEa501w1047+Rao2Q2f/RlammDjW1B0dlC6U40x4SHDujWNiThBC85ExAs8DJwHjAeuFpHxXS57AHhaVScCdwH3+J17GrhfVccB04E9wSrrMSld5HRpBihAWrljP5f99n1qG1t5bu4MTh2d3Xly1FmQMw7ef9hZV62pGoqsS9OYaJaZ6KO6sYWmlrZQF8UYEyDBbDmbDmxQ1U2q2gQ8D1zY5ZrxwL/c/YXt590gLkZV3wBQ1RpVrQtiWftmfxlUlQasS/PDTfu46ncf4PMKL3z9ZCbmpR98gYjTerb7U/jnj8Drg5GzAvLZxpjw1J5fs8pmbBoTMYIZnA0Dtvm9L3OP+VsBXOLuXwykiEgWMAaoEpE/i8gyEbnfbYkbWErfd14DMBngjVW7+crjHzEoNY6Xbj6F0YOSu79wwuWQlAM7ljkZA3yWosmYaNaRX9OCM2MiRqgnBHwXmCUiy4BZwHagFWeJj5nu+WnASOD6rjeLyFwRKRGRkvLy8n4rdIfSRRCXeswD8l9Zvp1v/H4pYwen8OI3TmFoesLhL46Nh2k3OfvWpWlM1LMsAcZEnmCuc7YdGO73Ps891kFVd+C2nIlIMnCpqlaJSBmwXFU3uef+AswAHuty/zxgHkBxcbEG52scQeliyJ8Bnr436r20tIz/eGkFJxVm8uh100iO68Gv5KSvQ90+mHhFnz/XGBMZOvJr1tpyGsZEimC2nC0BikSkUER8wFXAfP8LRCRbRNrLcCfwuN+96SKS474/E1gVxLL2Xk057F17TF2af1yylf94aQWnjsrmieun9ywwA0hIh/Pvs6wAxhgykmIB69Y0JpIELThT1RbgVuA1YDXwgqquFJG7ROQC97LZwFoRWQfkAj9x723F6dJ8U0Q+BQR4JFhl7ZOt7ePN+jYZ4A8flvK9P33K6UU5PHpdMQm+gTekzhgz8LV3a9pyGsZEjqCmb1LVBcCCLsd+4Lf/EvDSYe59A5gYzPIdk9LFEJMAQyb1+tan39/CD15ZyZljB/HrL00hPtYCM2NM38R6PaTEx9iYM2MiiOXW7KvSRTB8GsT4enXbY+9t5sd/W8XZ43N5+Jop+GJCPSfDGBPuMpN8VFq3pjERwyKDvmjYD7s+7XWX5u/e3siP/7aK804YzK+/ZIGZMSYwMhJ91nJmTASx6KAvtn4IaK8mAzy8cAP3vLqGL0wcwkNXTybWaz96Y0xgZCZZcGZMJLEIoS9KF4EnFoYV9+jyX/5zPfe/tpYLJw3lwSsnWWBmjAmojETLr2lMJLExZ31RuhiGTQFf4hEvU1V+8cY6HvrXBi6dksd9l03E67Ek5caYwMpMirWlNIyJINaE01tNdbDj46N2aaoq9722lof+tYEri4dzvwVmxoQVEZkjImtFZIOI3HGE6y4VERWRYvf9CBGpF5Hl7vbbYJc1I8lHQ3Mb9U2twf4oY0w/iM6Ws1Z3JW1vbO/vLVsCbS1HnAygqtzz6hrmvbOJL52Uz48vPAGPBWbGhA03l+/DwNk4eYGXiMh8VV3V5boU4Hbgwy6P2Kiqk/qjrACZiZ35NYf5jpD+zRgTFqKv5aytDV64Dl7+BrT14a/M0sUgHhg+/bCXPPbeZua9s4nrTi7g7ossMDMmDE0HNqjqJlVtAp4HLuzmuh8DPwUa+rNwXXWmcLKuTWMiQfQFZx4P5BXDZy/BX293grXeKF0EgydAfFq3p/fWNPLgP9dz1thB/OiC4xGxwMyYMDQM2Ob3vsw91kFEpgDDVfXv3dxfKCLLRORtEZkZxHICncGZzdg0JjJEZ7fmzG9Dcx28cz/EJsJ5P4WeBFEtTU63ZvGNh73kF2+so6G5lf/6/DgLzIyJUG5O4J8D13dzeieQr6r7RGQq8BcROV5VD3TznLnAXID8/Pw+lyejveXMJgUYExGir+Ws3Rnfh5NvhY9+B2/+L6ge/Z4dy6Cl4bCTAdbtrua5j7Zy7YwCRuUkB7jAxph+tB0Y7vc+zz3WLgU4AXhLRLYAM4D5IlKsqo2qug9AVZcCG4Ex3X2Iqs5T1WJVLc7JyelzYTvGnFnLmTERITpbzsBpKTvnbqcF7b1fQGwSzPqPI99Tush5zT+529P/t2A1SXEx3HZWUYALa4zpZ0uAIhEpxAnKrgKuaT+pqvuB7Pb3IvIW8F1VLRGRHKBCVVtFZCRQBGwKZmFTE2LxiI05MyZSRG9wBk6Adv7PoLkeFt4NsQlwyq2Hv750MeSMhaTsQ069s66ct9aW8/3zx3WM/zDGhCdVbRGRW4HXAC/wuKquFJG7gBJVnX+E208H7hKRZqAN+IaqVgSzvF6PkJ7os7XOjIkQ0R2cgTNB4IJfOQHa6993ArRpXz30urZW2PoBTLz8kFOtbcr/LVhNfmYiXzmloB8KbYwJNlVdACzocuwHh7l2tt/+n4A/BbVw3chIjKWytrm/P9YYEwQWnAF4Y+CSR5zxZH//tjNJYNLVB1+z61Noqu52fbMXS7axZlc1v/7SFOJivP1UaGOM6WT5NY2JHNE7IaCrGB9c/hSMnA2vfBNWvnzw+dLFzmuX8WY1jS088Po6igsyOO+Ewf1TVmOM6SIj0WezNY2JEBac+YuNh6ueheEnwZ++Bmv/0XmudBGkF0DaQUsd8bu3N7K3ppHv29IZxpgQspYzYyKHBWdd+ZLgmhechWZf+DJsXOgss1G6+JAuzR1V9Tzy7iYuOHEok/MzQlRgY4xx1jqrrGtCe7IskDFmQLPgrDvxqXDtnyF7DDx/DSx9EuorDlnf7IHX1tKm8J9zjgtNOY0xxpWZ6KO5ValpbAl1UYwxx8iCs8NJzIQv/wVSh8HfvuUc8wvOPimr4s/LtvPV0wrJy0gMSRGNMaZdR5YAm7FpTNiz4OxIknPguvnOWLP0fMgcCYCqcvffV5OV5OObs0eFuJDGGAOZSbEAttaZMRHAltI4mtSh8I13obGmI//m66t289HmCn580QmkxMeGuIDGGOPM1gTLEmBMJLDgrCfi05wNaGpp454Fqxk9KJmrpw0/yo3GGNM/2jOT7LPgzJiwZ92avfT7D0rZsq+O758/jhiv/fiMMQND55gzC86MCXcWXfRCVV0Tv3xzPTOLspl9XE6oi2OMMR1S4mKI8YiNOTMmAlhw1gv/718bONDQzH+dbwvOGmMGFhFx1jqzljNjwp4FZz20ZW8tT7+/hSumDmfckNRQF8cYYw6RZVkCjIkIFpz10L2vriHW6+E754wJdVGMMaZbll/TmMhgwVkPfLS5gn+s3MU3Zo1iUGp8qItjjDHdsvyaxkQGC8564PmPtpKRGMtNM0eGuijGGHNYGUmxVNZZhgBjwp0FZz1QXtNIflYSCT5vqItijDGHlZnoo6quidY2S35uTDiz4KwHquqayUi0TADGmIEtI8lHm8KBems9MyacWXDWAxW1TWS6qVGMMWagas8SYGudGRPeLDjrgaq6JtItODPGDHCWX9OYyGDB2VE0trRS29RKZpJ1axpjBraOljMLzowJaxacHUWVO/PJWs6MMQNdR35N69Y0JqzFhLoAA117JZdhwZkxYUlEcoCbgBH41XmqemOoyhQs7WNjK2ptQoAx4cyCs6No7x7IsG5NY8LVK8C7wD+B1hCXJagSfF7iYz3WcmZMmLPg7CjauzWt5cyYsJWoqt8LdSH6S2aiZQkwJtzZmLOjaK/k2gfaGmPCzt9E5Pze3iQic0RkrYhsEJE7jnDdpSKiIlLsd+xO9761InJuXwveFxlJPputaUyYs+DsKKrc7oF0W4TWmHB1O06A1iAi1e524Eg3iIgXeBg4DxgPXC0i47u5LsV9/od+x8YDVwHHA3OAX7vP6xeZST5b58yYMGfB2VFU1jWT5PMSF2Opm4wJR6qaoqoeVY1391NUNfUot00HNqjqJlVtAp4HLuzmuh8DPwUa/I5dCDyvqo2quhnY4D6vX2QkWsuZMeHOgrOjqKy1BWiNCXcicoGIPOBuX+jBLcOAbX7vy9xj/s+cAgxX1b/39t5gykzysc+CM2PCWlCDs6ON2RCRAhF5U0Q+EZG3RCSvy/lUESkTkV8Fs5xHUlnXZDM1jQljInIvTtfjKne7XUTuOcZneoCfA985xufMFZESESkpLy8/lkd1yEj0Ud3QQnNrW0CeZ4zpf0ELzno4ZuMB4GlVnQjcBXStMH8MvBOsMvZERV2zzdQ0JrydD5ytqo+r6uM448A+f5R7tgPD/d7nucfapQAnAG+JyBZgBjDfnRRwtHs7qOo8VS1W1eKcnJxefKXDa89mYstpGBO+gtly1pMxG+OBf7n7C/3Pi8hUIBd4PYhlPKqquiYLzowJf+l++2k9uH4JUCQihSLiwxngP7/9pKruV9VsVR2hqiOAD4ALVLXEve4qEYkTkUKgCPgoQN/jqDqyBNhCtMaErR4FZyLyZxH5vNuU31M9GXexArjE3b8YSBGRLPdzfgZ8txefFxQVtU22jIYx4e0eYJmIPCkiTwFLgZ8c6QZVbQFuBV4DVgMvqOpKEblLRC44yr0rgRdwulD/Adyiqv22+K3l1zQm/PV0EdpfAzcAD4nIi8ATqro2AJ//XeBXInI9TvfldpwVvL8JLFDVMhE57M0iMheYC5Cfnx+A4hyspbWN6oYWW0bDmDCmqs+JyFvANPfQ91R1Vw/uWwAs6HLsB4e5dnaX9z/hKAFgsGRafk1jwl6PgjNV/SfwTxFJA65297cBjwC/V9Xu2s+POu5CVXfgtpyJSDJwqapWicjJwEwR+SaQDPhEpEZV7+hy/zxgHkBxcbH25Lv0RlW9ZQcwJlyJyFhVXePOqgSn9R5gqIgMVdWPQ1W2YOrMr2nBmTHhqsfpm0QkC7gW+DKwDPgDcBpwHTC7m1s6xmzgBGVXAdd0eWY2UKGqbcCdwOMAqvolv2uuB4q7Bmb9obIjr6YFZ8aEoW/jtKz/rJtzCpzZv8XpH+1L/9haZ8aErx4FZyLyMnAc8AzwRVXd6Z76o4iUdHePqraISPuYDS/wePuYDaBEVefjBHX3iIjidGveckzfJsAqO/JqWremMeFGVee6u+epqv8isYhIfAiK1C98MR5S4mIsS4AxYaynLWcPqerC7k6oanF3x91zRxyzoaovAS8d6YNV9UngyR6WM6DauwWsW9OYsLYYmNKDYxHD8msaE956GpyNF5FlqloFICIZwNWq+uuglWwAaM+rad2axoQfERmMM0M8QUQmA+2zi1KBxJAVrB9kJPmoqLOlNIwJVz0Nzm5S1Yfb36hqpYjchDOLM2JZt6YxYe1c4HqcyUg/ozM4OwD8V4jK1C8yE2PZW2MtZ8aEq54GZ14REVVV6Fj9P+KbkyrrmoiL8ZAQa0nPjQk3qvoU8JSIXKqqfwp1efpTRpKPdbtrQl0MY0wf9XRR2X/gDP4/S0TOAp5zj0W0ylonO8CR1lozxgx4U0Ukvf2NiGSIyN0hLE/QZSb6bJ0zY8JYT4Oz7+GkV7rZ3d4E/jNYhRoonKTnEd9AaEykO699vCw4wzJw8m1GrIwkH3VNrTQ091tiAmNMAPV0Edo24DfuFjUq65ptvJkx4c8rInGq2gggIglAXIjLFFT+WQKGpCWEuDTGmN7q6TpnRTj56cYDHesDqerIIJVrQKisa2LckNRQF8MYc2z+ALwpIk+4728AngpheYIuwy9LgAVnxoSfnk4IeAL4IfAL4Aycyq03SdDDkjPmzFrOjAlnqvpTEfkEOMs99GNVfS2UZQq2jpazWltOw5hw1NPgLEFV33RnbJYCPxKRpUC3SYAjQWubsr++2RagNSYCqOqrwKuhLkd/yUxy/qi0LAHGhKeeBmeNIuIB1rspmbbjJCSPWAfqm2lTyw5gTLgTkUuAnwKDcNY6E0BVNWLHLGRYfk1jwlpPuyZvx1lR+zZgKk4C9OuCVaiBoLIjO4B1axoT5u4DLlDVNFVNVdWUSA7MANISYhHpTEFnjAkvR205cxecvVJVvwvU4Iw3i3gdwZm1nBkT7nar6upQF6I/xXg9pCXEWnBmTJg6anCmqq0iclp/FGYgaR9Ia8GZMWGvRET+CPwFaGw/qKp/DlmJ+kFmos/GnBkTpno65myZiMwHXgRq2w9GcuVmLWfGRIxUoA44x++YAhFbf4EzY9PGnBkTnnoanMUD+4Az/Y5FdOVmY86MiQyqGhVDMbrKSPKxraIu1MUwxvRBTzMERF3lVlnXTIxHSI7rafxqjBmI3MVntetxVb0xBMXpN5mJPj4pqwp1MYwxfdDTDAFRV7lV1jp5NS3puTFh729++/HAxcCOEJWl32Qk+aisbUZVrR4zJsz0tFko6iq3yjrLDmBMJFDVP/m/F5HngPdCVJx+k5kUS1NrG7VNrdYDYEyY6Wm3ZtRVbpV1zaTbZABjIlERzoK0Ec1/IVoLzowJL339FxvxlVtlbROjciI6CYIxUUFEqjl4WMYu4HshKk6/ac+vWVHbxPDMxBCXxhjTGz0dcxZ1lVtlXbPN1DQmjInIqaq6CMhR1YZQl6e/ZbQHZ7bWmTFhp6fdminBLshAoqpU1TXZGmfGhLeHcNLNLQamhLgs/S7T8msaE7Z6lFtTRC4WkTS/9+kiclHQShVi1Y0ttLSpBWfGhLdmEZkH5InIQ123o90sInNEZK2IbBCRO7o5/w0R+VRElovIeyIy3j0+QkTq3ePLReS3QfhuR5Xh161pjAkvPR1z9kNVfbn9japWicgPcdKhRJyq9tRNSRacGRPGvgB8DjgXWNqbG92cwg8DZwNlwBIRma+qq/wue1ZVf+tefwHwc2COe26jqk46tuIfm9T4GLwe6VhQ2xgTPnoanHXXwhax038qOlI32ZgzY8KVqu4FnheR1aq6ope3Twc2qOomABF5HrgQ6AjOVPWA3/VJdLMWZCiJCBmJPircPzaNMeGjR92aOImDfy4io9zt5/TyL9Fw0v6Xpi2lYUz460NgBjAM2Ob3vsw9dhARuUVENgL3Abf5nSoUkWUi8raIzOzD5wdEZlKsjTkzJgz1NDj7N6AJ+CPwPNAA3BKsQoVae2WWad2axpgjUNWHVXUUzuz1/3YP7wTyVXUy8G3gWRFJ7e5+EZkrIiUiUlJeXh7w8mUk+my2pjFhqKezNWuBQwbERqrKOnfMmXVrGhOttgPD/d7nuccO53ngNwCq2gg0uvtL3Za1MUBJ15tUdR4wD6C4uDjg3aKZST427KkJ9GONMUHW09mab4hIut/7DBF5LWilCrHK2iY8AqnxFpwZE+5E5HYRSRXHYyLysYicc5TblgBFIlIoIj7gKmB+l+cW+b39PLDePZ7jTihAREbiLNq9KVDfpzcyknw2IcCYMNTTbs1sVa1qf6OqlURwhoDKuibSE314PJYs2JgIcKM7eP8cIAP4MnDvkW5Q1RbgVuA1YDXwgqquFJG73JmZALeKyEoRWY7TfXmde/x04BP3+EvAN1S1IsDfqUcyE31U1jXT1jag5ioYY46ipzMu20QkX1W3grOODwNsZlIgVdU1k25dmsZEiva/ss4HnnGDrKP+5aWqC4AFXY79wG//9sPc9yfgT92d628ZST5a25TqhhbSrE4zJmz0NDj7PvCeiLyNU9HNBOYGrVQhVlHb1LG6tjEm7C0VkdeBQuBOEUkB2kJcpn6R6aag21fbaMGZMWGkpxMC/iEixTgB2TKcxWfrg1iukKqsayIvwxIFGxMhvgpMAjapap2IZAI3hLZI/aM9y4mNOzMmvPQ08fnXgNtxZiwtB2YA7wNnBq1kIVRZ18TEvLSjX2iMCQcnA8tVtVZErsXJs/nLEJepX2QlxQHYQrTGhJmeTgi4HZgGlKrqGcBkoCpYhQolVaWyrtnyahoTOX4D1InIicB3gI3A06EtUv/IcLs1bSFaY8JLT4OzBlVtABCROFVdAxwXvGKFTn1zK00tbZZX05jI0aKqipN+6Veq+jCQEuIy9Yv2hbRtIVpjwktPJwSUueuc/QV4Q0QqgdJgFSqUKmotr6YxEaZaRO7EWUJjpoh4gKj4B54Q6yUuxmMtZ8aEmZ5OCLjY3f2RiCwE0oB/BK1UIVTlZgewvJrGRIwrgWtw1jvbJSL5wP0hLlO/EBEyk3wdf3QaY8JDT1vOOqjq28EoyEBRYXk1jYkobkD2B2CaiHwB+EhVo2LMGTgzNm22pjHhpadjzqJGeyVm3ZrGRAYRuQL4CLgcuAL4UEQuC22p+o+1nBkTfnrdchbpqjqSnlvLmTER4vvANFXdA07uS+CfOKmVIl5Gko/tVRG7LKUxEclazrpo/wszLcFazoyJEJ72wMy1jyiq+zITY63lzJgwYy1nXVTVNZEaH0OMN2rqbmMi3T9E5DXgOff9lXTJmRnJMpJ87K9vpqW1zeo1Y8JEUP+lisgcEVkrIhtE5I5uzheIyJsi8omIvCUiee7xSSLyvoisdM9dGcxy+quoa7bJAMZEEFX9D2AeMNHd5qnq90Jbqv7TXp9V1VuWAGPCRdBazkTECzwMnA2UAUtEZL6qrvK77AHgaVV9SkTOBO7BWYuoDviKqq4XkaE4iYtfU9WqYJW3XVVdky2jYUyEUdU/AX8KdTlCoSO/Zm0T2clxIS6NMaYngtlyNh3YoKqbVLUJeB5nhW5/44F/ufsL28+r6jpVXe/u7wD2ADlBLGuHitomazkzJgKISLWIHOhmqxaRA6EuX38ZnpkIwNLSyhCXxBjTU8EMzoYB2/zel7nH/K0ALnH3LwZSRCTL/wIRmQ74cPLhBV1VXTPptoyGMWFPVVNUNbWbLUVVU0Ndvv5yYl4axw9N5dH3NtPWpqEujjGmB0I9OvS7wCwRWQbMArYDre0nRWQI8Axwg6q2db1ZROaKSImIlJSXlwekQJV1TbaMhjEmYogIc08fyYY9NSxcu+foNxhjQi6Ywdl2YLjf+zz3WAdV3aGql6jqZJy1iGgfVyYiqcDfge+r6gfdfYCqzlPVYlUtzsk59l7PhuZW6pparVvTGBNRzp8whKFp8cx7Z1Ooi2KM6YFgBmdLgCIRKRQRH3AVMN//AhHJdpMQA9wJPO4e9wEv40wW6LeFIjvzalq3pjEmcsR6Pdx4WiEfbq5gxbaqUBfHGHMUQQvOVLUFuBV4DVgNvKCqK0XkLhG5wL1sNrBWRNYBucBP3ONXAKcD14vIcnebFKyytuvIq2ndmsaYCHPV9HxS4mN45F1rPTNmoAvqIrSquoAuiz2q6g/89l+imxQqqvp74PfBLFt3qty8mraUhjEm0iTHxXDNSfk88s4mtlXUdcziNMYMPKGeEDCgVLrdmjbmzBgTiW44pRCPCI+9tznURTHGHIEFZ34q3JazDBtzZoyJQIPT4rlg0lBeKNnW0VNgjBl4LDjzU1Vr3ZrGmMg29/SR1DW18ocPt4a6KMaYw7DgzE9FXRPJcTH4YuzHYoyJTGMHp3L6mByeXLyFxpbWo99gjOl3FoX4sewAxph2IjJHRNaKyAYRuaOb898QkU/d2eTvich4v3N3uvetFZFz+7fkRzd35kjKqxt5ZdmOUBfFGNMNC878VNZZXk1jDIiIF3gYOA8nB/DV/sGX61lVnaCqk4D7gJ+7947HWdfxeGAO8Gv3eQPGqaOzGDcklXnvbrKUTsYMQBac+amsbbLxZsYYgOnABlXdpKpNwPPAhf4XqKp/8vQkoD3KuRB4XlUbVXUzsMF93oDhpHQqZMOeGt5eF5jUd8aYwLHgzE9lXbPN1DTGAAwDtvm9L3OPHUREbhGRjTgtZ7f15t4+a6yGqmMfzP+FiUMZYimdjBmQLDjzU1lrSc+NMT2nqg+r6ijge8B/9/Z+EZkrIiUiUlJe3sMWrE/+CA9OhD9cDmv+Dq0tvf1YwE3pdGoh72/ax6dl+/v0DGNMcFhw5mpubaO6scWCM2MMwHZguN/7PPfY4TwPXNTbe1V1nqoWq2pxTk5Oz0pWdC6c/l3Y+Qk8fw08eAL86yd9ak27avpwUuJimGcpnYwZUCw4c1W6CzJmJlm3pjGGJUCRiBSKiA9ngP98/wtEpMjv7eeB9e7+fOAqEYkTkUKgCPgoYCVLHw5n/jf8+0q48g+QewK8c3+fWtNS4mO5+qR8Fny6k20VdQErojHm2Fhw5qpyUzfZhABjjKq2ALcCrwGrgRdUdaWI3CUiF7iX3SoiK0VkOfBt4Dr33pXAC8Aq4B/ALaoa+AXFvDEw7gtw7Utw+4o+t6Zdf8oIBHhi0ZaAF9EY0zdBTXweTipr21M3WXBmjAFVXQAs6HLsB377tx/h3p8APwle6brIKHBa02bdAev+AUufdFrT3rkfis6GaTfBmHO6vXVoegIXnDiU55ds5fazikizSVHGhJy1nLnauzUzrFvTGBOuDtea9uzlsGXRYW/72kwnpdOzH1lKJ2MGAgvOXJVut6a1nBljIkJ7a9q/LQWvzxmLdhjjh6YysyibJxZttpROxgwAFpy5Kqxb0xgTieKSYcRMWP/aES+7aeZI9lQ3Mn+5pXQyJtQsOHNV1TURH+shwTegsqwYY8yxG3Mu7NsA+zYe9pKZRdmMHZzCI+9uQtVSOhkTShacuSrrmsm0VjNjTCQqcicDrDt865mT0mkk63ZbSidjQs2CM5fl1TTGRKzMQsgec9SuzS9MHMrgVEvpZEyoWXDmqqxrspmaxpjIVXSOM2Ozsfqwl/hiPNxw6ggWb9zHR5sr+rFwxhh/Fpy5nKTn1nJmjIlQY86FtmbY9NYRL7v6pHyGZybw1aeWsHxbVb8UzRhzMAvOXJV1lvTcGBPB8k+GuNQjjjsDSI2P5fm5J5OR6OPLj37Ix1sr+6mAxph2FpwBrW3K/vpmMpIsODPGRChvLIw6A9a/AUeZjTksPYHn584gM9nHVx77iKWl1sVpTH+y4AzYX9+MKmRY2hJjTCQrOhdqdsHOFUe9dGh6An+cezI5KXF85bGPKNliAZox/cWCM/xSN1m3pjEmkhWd7byuf71Hlw9Oi+f5uTPITY3nK49/ZJMEjOknFpzhl/TcujWNMZEseRAMnXLUcWf+clOdAG1IWjzXP/ERH2zaF8QCGmPAgjPAP6+mdWsaYyLcmHNh+1Ko3dvjWwalxvPc3BkMS0/ghieWsHhjz+81xvSeBWf4tZxZt6YxJtIVnQOoMzGgFwalxPPsTTMYnpnAjU8uYdEGC9CMCRYLzvAbc2bdmsaYSDdkEiQNOmq2gO7kpMTx7E0zGJGVxI1PLuHd9ZbmyZhgsOAMp1vT5/WQZEnPjTGRzuNxWs82/Atam3t9e3ayE6AVZifx1adKLA+nMUFgwRnteTVjEZFQF8UYY4JvzDnQuB+2fdin2zOTfDx30wxG5yRz09MlvLV2T4ALaEx0s+AMyw5gjIkyI88AT2yvZm12lZHk49mbTqJoUDJzn17KG6t2B7CAxkQ3C86wpOfGmCgTnwoFJ/d4vbPDSU/08ezXZjBuSArf+P1SXijZFqACGhPdLDjDkp4bY6JQ0blQvgYqS4/pMWmJsTx70wxOGZXFf770Cb95ayN6lPRQxpgjs+AMqKprspmaxpjoMmaO83qMrWcASXExPHbdNL544lB++o813P331bS1WYBmTF9FfXCmqm7LmXVrGmOiSPZoyBx5TOPO/PliPPzyyklcf8oIHntvM995cQXNrW0BebYx0SYm1AUItQMNLbS2qXVrGmOiT9G5sPQJaKoDX+IxP87jEX74xfHkpMRx/2trqaht4jfXTiHRF/X/qzGmV6K+5cyyAxhjotaYc6ClATa/E7BHigi3nDGaey+ZwLvry7nmkQ876lljTM9YcNaRHcC6NY0xnURkjoisFZENInJHN+e/LSKrROQTEXlTRAr8zrWKyHJ3m9+/Je+FglMhNqlP2QKO5qrp+fzm2qms2nmAy367mO1V9QH/DGMiVdQHZ1UdSc+t5cwY4xARL/AwcB4wHrhaRMZ3uWwZUKyqE4GXgPv8ztWr6iR3u6BfCt0XMXEw6gxn3FkQZliee/xgnrlxOnuqG7nsN4tZv7s64J9hTCSK+uCswro1jTGHmg5sUNVNqtoEPA9c6H+Bqi5U1Tr37QdAXj+XMTCKzoED22H3yqA8/qSRWbzw9ZNpaVMu++37LC2tDMrnGBNJoj446+jWtODMGNNpGOC/omqZe+xwvgq86vc+XkRKROQDEbkoCOULnKJznNcgdG22GzcklT/ffAoZibFc++hiSv/4XVj5ctA+z5hwF/VTaCrrmvB6hJT4qP9RGGP6QESuBYqBWX6HC1R1u4iMBP4lIp+q6sZu7p0LzAXIz8/vl/IeInUIDJ4I616Hmd8J2scMz0zkpZtP4e+//g8KVj9F7boX+P3O0eRmZTI0PYGh6fEMTo0nxhv1bQbGBDc4E5E5wC8BL/Coqt7b5XwB8DiQA1QA16pqmXvuOuC/3UvvVtWnglHGyrpm0hNi8Xgs6bkxpsN2YLjf+zz32EFE5HPA94FZqtrYflxVt7uvm0TkLWAycEhwpqrzgHkAxcXFoVu1dcy58O7PoK4CEjOD9jHZe97nK/XPsCl+PCMbVrF74e+4p/W8jvMegcGp8W6w5mzDMhIYlh7PCcPSGJQSH7SyGTOQBC048xtQezZOl8ASEZmvqqv8LnsAeFpVnxKRM4F7gC+LSCbwQ5y/RhVY6t4b8MEKlbWWHcAYc4glQJGIFOIEZVcB1/hfICKTgd8Bc1R1j9/xDKBOVRtFJBs4lYMnCww8RefCO/fDhjdh4uXB+Yz9ZfDSjUj2GEZ+7Q149kr+p+KfXHPNXeyobmVHVT07quopc1+Xb6vi1c920tzqxKyxXuGSyXl8fdZIRuYkB6eMxgwQwWw56xhQCyAi7QNq/YOz8cC33f2FwF/c/XOBN1S1wr33DWAO8FygC1lZ12TZAYwxB1HVFhG5FXgNp+X/cVVdKSJ3ASWqOh+4H0gGXhQRgK3uzMxxwO9EpA1nXO+9Xf4oHXiGTYHELGfcWTCCs5ZGeOE65/WKZyAuGWZ+G/n9JYze+TdGT/lKt7e1til7axopq6zjL8t28ELJNl5Yuo3zThjMzbNGMyEvLfBlNWYACGZw1t2A2pO6XLMCuASn6/NiIEVEsg5z7yGDcQMxXqOqrpn8zGNfGdsYE1lUdQGwoMuxH/jtf+4w9y0GJgS3dAHm8cLos53grK3VeR9Ir/0XbC+By5+CnDHOsVFnwpAT4b0HYdKXuv1Mr0fITY0nNzWeqQWZ3HZWEU8s2swz75ey4NNdzCzK5ubZozh5ZBZugGxMRAj1yMvvArNEZBnOYNrtQGtPb1bVeaparKrFOTk5fSpARW2TzdQ0xpgx50B9JZQtCexzV/wRljwKJ98Kx1/UeVzEmYBQsRFWvdKjR+WkxPGfc8ay6M4z+d6csazeWc01j3zIRb9ezGsrd1mydRMxghmcHXVAraruUNVLVHUyzqBaVLWqJ/cGgqpSVddMumUHMMZEu1FngXgDlggdgF2fwV9vdzIRfO5/Dz0/9ouQPQbe/XmvFsFNjY/l5tmjeO97Z3D3RSdQWdvE159ZyjkPvsNLS8ss4boJe8EMzjoG1IqID2dA7UFpTEQkW0Tay3AnzsxNcMZ5nCMiGe7g2nPcYwFV29RKU2sbmdZyZoyJdgnpkD8D1r8emOc17IcXvgzxaXDZE+DtZhSNxwOnfgt2fwrr3+j1R8THerl2RgH/+s4sfnnVJGI8wndfXMGs+xby27c3smFPDRqEzAfGBFvQgjNVbQHaB9SuBl5oH1ArIu3pTGYDa0VkHZAL/MS9twL4MU6AtwS4q31yQCBZ0nNjjPFTdA7s/syZWXksVOEv34SqrXD5k5CSe/hrJ14BacPhvZ/3+eNivB4unDSMV2+fyRPXT2NYRgL3vrqGz/38bU6/fyE/eOUzFq7ZQ31Tj0fNGBNSQV3nrAcDal/CyUnX3b2P09mSFhQdeTVtKQ1jjHHWO/vnD53Ws+Ib+/6cRQ/Cmr/BufdAwclHvtYbC6fcBq/+B5QuhoJT+vyxIsIZYwdxxthBlFXW8dbact5au4cXS8p4+v1S4mI8nDwqizOOG8QZxw0iP8smg5mBKaqXxa/oSN1kY86MMYacsZCW72QL6GtwtultePMuOP5imHFzz+6ZfC28/VNnIdxjCM785WUkcu2MAq6dUUBDcysfba5g4do9vLW2nB/OX8kPWcnInKSOQG1aYQZxMQGepWpMH0V1cFblBmfp1q1pjDHODMox58LyP0BzA8T2ckX+/dvhpRshazRc8P+c5/WELxFO/qYT1O1c4SyxEUDxsV5OH5PD6WNy+OEXYfPeWt5au4eFa8t55oNSHntvM74YD2Nykxk3OJWxQ1IZNziFcUNSrWfFhERUB2cV7pizTPvHZ4wxjjHnwpJH4PlrYOQsyJsOQydBbMKR72tpghevg5YGuPL3EJfSu8+d9jVnzbN3fw5XBCVbX4fC7CQKswu54dRC6ppaeH/jPj7YtI81u6pZuLacF5d2jrnLTY1j7OBUxg1JZdyQFMYOTmVkThKxlgPUBFFUB2eVdc2IQFqCdWsaYwwAI2fD9K87C9JufNM55olxkqMPnw5505zXtOEHt4y9/n1njbTLn4Sc43r/ufFpboD2C9i7HrKLAvFtjirRF8NZ43I5a1znpIXy6kbW7DrAmp3VrN51gNU7q1m8cVNHKimf10NhdhIp8TEk+LzEx3pJiPWS2L7v85Lovsa7xxN9MQxJi2d4ZiIZibG2aK45oqgOzqrqmkhLiMVrSc+NMcbhjYXz7wPug5o9TsC17SPndelT8OFvneuSB8PwaU7LmrbBR/Ngxi3OWLO+mvFN+ODXzoSCCx8OxLfpk5yUOHJScphZ1Lm4eXNrG5vKa1m98wCrdx1gU3ktdU0t1Da2UF7dSENzK/XNrdQ1tdLQ3NoRyHUnyedleGYieRmJDM9MYHhGIsMzO/eT4qL6f82GKA/OLDuAMcYcQfIgGPt5ZwNobYbdK/0Cto9g9V+dc/knw9ndLDTbq8/LgSnXQcljMOsOSB9+9Hv6SazXw3GDUzhucAoXHZpN8BDNrW3UN7fS0OQEbdUNLeyoqmdbZT3bKuooq6xjW0Udizbspb754CU+MpN8DM9M5JzxuVwzPd/GvUWhqA7OquqaSbeZmsYY0zPeWGf82dBJMP0m51jNHmcQf16xc/5YnfJvTnD2/q/gvJ8e+/NCJNbrIdbrITW+82dywrBDE7WrKvtqm9hWUXdQ4LZudw33v7aWh95cz8WTh3H9qSMYOzi1P7+CCaGoDs4qapsYktbL2UjGGGM6JQ+CorMD97z04TDxSqcL9fT/gKTswD17ABIRspPjyE6OY3J+xkHn1u2u5olFW3h5WRnPL9nGKaOyuOHUQs4cO8iG40S4qJ5uUlXXZMtoGGPMQHPqt5xZnx/8JtQlCakxuSncc8kEPrjzLL43Zyxb9tZy09MlnPHAWzz67iYONDSHuogmSKK75ayuiUxLem6MMQNLzhgY90X46BE49TZnJmcUS0/0cfPsUdw0s5DXV+3miUWbufvvq/nFG+u4bGoe150ygpE5yYfcV9vojHPbXlXPzv0NHfs7qurZUdVAclwMk/PTmZKfweT8dAqzkwIyi1RVKa9upLaplRFZiTYztQ+iNjhraG6lobnNWs6MMWYgmvltWD0fljzm7BtivB7OnzCE8ycM4dOy/TyxeDPPfbSNp94vZfZxOeRnJroBmBOI7a8/uGXNIzA4NZ6h6QmcODydqrom5i/fwR8+3Ao42XIm52cwJT+dyfkZnDg8neQjzBxVVXbub2D9nhrW765mw56ajv0DDS0AZCf7mF6YyYyRWcwYmUXRoGQL1nogaoOzyjpbgNYYYwasoZNh1FnO0hozbj76IrhRZkJeGj+/YhJ3njeOP3xYyrMfbmXZ1iqGpMUzLD2B4oIMhqYnMDTdeT8kPYHclDhiuiye29qmbCyv4ePSSj7eWsnHW6v415o9gBPMjclNYUpBBpOHp5OZ5OsMwPbUsHFPDTWNLR3PykzyMXpQMl88cShFg5KJi/WyZHMFH2zax4JPd3Vcc1JhJicVZjJjVBZjBqXgsfFzh4ja4Kw9O4Dl1TTGmAFq5rfhyc/Dst93zg41B8lJieNbnxvDtz43pk/3ez3CmNwUxuSmcNX0fAD21zezfFsVH5dWsmxbFX9dsYNn3dY1cLImjB6UzGVT8xg9KJmiQcmMHpRMVnLcIc+/eno+qkpZZT3vb9rHh5ucYO3Vz5xgLT0x1g3WsjhxeDrgLEPS1NLW8drk99rc/tqqtLQqSXFeUhNiSUuIJTXefU2IITUhlmRfTNgGflEbnFXVOc291q1pjDEDVMGpMPwkWPRLmHp9YJbqMEeVlhDLrDE5zBrjLMLb5rauHWhoYfSg5F5n1RERd5HdRK4odtau21ZRx4ebK/hw0z4+2LyP11buDvj38Aik+AVsaQmxDEqJZ3BaPEPS4hmcGs+QtAQGp8WTleQbUIFc1AZnllfTGGMGOBGY+R149gr45AWY/KVQlygqeTxCUW4vc6UeRXuwdtnUPAC2V9WzZucBPB4hzuvBF+OsE+eLcbeux7wevB6hprGFA/XN7K9v5kBDMwfqmzlQ38KBBvdYx7kWquqaWLKlgt0HGg7J4BDrFXJT3aAtLaEjeBuRnUjRoBSGpSf0a/AWtcFZlTvmzBahNcaYAazoHBg8AebfCmv+DtO+CiPPAE9UrwQVcYalJzAsvffjCtPcLs3e5JJoa3MW/t21v4Gd++vZdaCBnfsbOt5/WlbF6ysbaGxp67gn0edltNt9WzQohTG5zmteRnCCtqgNzirdbk1L32SMMQOYCHzpT05Oz4+fhrV/h8xRTpA26RpIyDj6M46mtRmqd0FiFvgSj/15ZkDzeMTNnxrHhLzul2lRVSrrmtm8t4Z1u2tYv7uG9XuqWbxhH3/+eHvHdfGxno6ArSg3mSn5GcwYmXXMZYza4KyitomUuBhivfbXlzHGDGgpufC5H8LsO2DVK7DkUXjtv+DNH8OES2HaTU5KqZ5qroftS6F0MZQucvKENtc553zJkJTjZD5IHgRJ7a85kJx78L4FchFLRMhM8pGZlMnUgsyDzu2vb3Zmre6uZv2eGtbtruaDTft4edl2Pj9xiAVnx6Kqrol0W4DWGGPCR0wcTLzC2XZ+4gRpn77ozOYcVgzTvgbHXwyxXdLyNVbDtg/dYGyxE5i1NgECg0+AKV+BnLFQXwm15U6+0No9sHcDbFkE9RXdlydjBOSeALnHd75mFFqXa4RLS4hlakEGUwsObrU90NBMXWPrYe7qnagNzirqmsm0Lk1jjAlPQybCBQ/B2XfBiuedQO0v33Ba1KZ82QnW2gOynStAW0G8zvppJ33DmQmaPwMS0o/+Wa3NULvXCdhqyqFmNxzYDntWwe6VsHYBqDs+KTYJcscfHLDlHh/1WQ6iQWp87EGJ7o9F1AZnVXVNNt7MGGPCXUI6zPgGnPR12Py2E6Qt/pUTjHnjIK/YWS+t4FTImwZxh6Y5OipvLKQOcbbuNNVB+RrY/ZkTrO1eCSv/Akuf7LwmNc+5PzHbSeaelN39fmL2oS1/JupEbXBWWdfEqG5ykRljDICIzAF+CXiBR1X13i7nvw18DWgByoEbVbXUPXcd8N/upXer6lP9VvBoJQIjZzvbgR2wvwwGT+yfQMeXCMOmOFs7Vaccu1fC7k9hzxqn5W1/GexYBnV7oa3lMM9LcQK1zJEwaBwMGg+Dxjpdr76k4H8fE3LRG5zVNtsyGsaYbomIF3gYOBsoA5aIyHxVXeV32TKgWFXrRORm4D7gShHJBH4IFAMKLHXvrezfbxHFUoc6WyiJQNowZxtzzqHnVaFhP9Ttc8a51e51Arbavc6xmt2wdz0sWQQtDe0PhYwCJ1jLGesGbeMgu8gZj2ciRlQGZ00tbdQ0ttiYM2PM4UwHNqjqJgAReR64EOgIzlR1od/1HwDXuvvnAm+oaoV77xvAHOC5fii3CRciTpdsQjpkjTr8dW2tULnFGd+2Z3Xntv71zpY38TrPGDnbWV5kyCTn+SZsRWVw1rEArWUHMMZ0bxiwze99GXDSEa7/KvDqEe4d1t1NIjIXmAuQn5/f17KaSOZxA6+sUTDui53HW5pg3wYod4O1XZ/C0qfgo3kw6HgnSJt4hbP0hwk7URmcdS5Aa92axphjIyLX4nRhzurtvao6D5gHUFxcrEe53JhOMT53Vuj4zmP1lfDZn2H5s/D69+GNHzgZFiZdA2PmOPeYsBCVwVlHXk3r1jTGdG87HJQRJs89dhAR+RzwfWCWqjb63Tu7y71vBaWUxvhLyHAyJ0z7KpSvdYK0Fc/DulchIRMmXO52e55o3Z4DXFQGZ515NS04M8Z0awlQJCKFOMHWVcA1/heIyGTgd8AcVd3jd+o14P9EpH2FynOAO4NfZGP85BwHZ/8vnPk/sOktWP4HZ2mPj37X2e1ZcArEJjiTCWLi3S3OWYLEO0DDA1Wo2AQb/wXbP3YmXAwa72xZo5xlTyLAAP3pB1d7t2amjTkzxnRDVVtE5FacQMsLPK6qK0XkLqBEVecD9wPJwIvitEJsVdULVLVCRH6ME+AB3NU+OcCYfueNgaLPOVt9JXz2p85uzyPxxHQGa+2vnhgnOMLtgW/f1/YeeT34vC/ZmZwwbAoMm+osytuXrtWGA7D5Hdj4Jmx4E6pKneNJOVBX4axpB+CJhewxTldvxxIk4yAtP+yyNkRpcNbechYZEbYxJvBUdQGwoMuxH/jtf+4I9z4OPB680hnTBwkZToqraV+D8nVQsdFZpqOl8Siv7n5rs9sdKn7dotJ5DA4+X1fhzCpd8axzzutz1p4bNrVzyxx5aODU1gY7l8GGfzktZGUfOTNTfclQeDqc8m8w+izn3uYG2LfemRSxe6XzuvVDJ61XO1+yu/TIOGch4oJTnVa2Ady1G53BWW0TCbFe4mO9oS6KMcYY0/9yxjhbsKnC/m1OPtPtS52uyGXPON2r4KS1Guq2rKUOdRLRb1zYmc90yIlwym1OMJY3/dCWt9h4GDzB2fw1HHCyNnQsQbIK1vzd+WxwWt3yT3a6dgtOcVr1PAMnJojK4Kyirsm6NI0xxphgE4H0fGc7/mLnWGsL7F3rBGtlJU7A9t4vnO7JpEEw5lwYdSaMPAOSc/r2ufGpMHy6s7VTdZYfKV0Epe/D1sWwer5zLs69vuAUyD/F6YoN4cK+URmcVdVZdgBjjDEmJLwxnQnhp3zFOdZUC9W7nK7KYHU3ijjZFLKLYOr1zrH9ZU6gVroItr4Pb97lljHOac1LG9Y5SSIm3mm5i4l3umg7xuTFdR5Lyzs4jVcfRWVwVmktZ8YYY8zA4Us6cqaEYEnLg4mXOxtA7T4nSNv6Pmz9wGnd6xh71+S8tjUf/nnjL4Qrnj7mYkVlcPbglZNoabP1Ho0xxhjjJykLxn3B2Q6nrQ1auwRsre5rgBLTR2VwVpAVmB+eMcYYY6KMxwOeBGeNuGB9RNCebIwxxhhjes2CM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAcSCM2OMMcaYAURUIyPHpIiUA6W9uCUb2Buk4gxU0fado+37QvR95wJVzQl1IQKhl3VYtP2ewb5zNIi273vY+itigrPeEpESVS0OdTn6U7R952j7vhCd3zkaRePv2b5z5Iu273sk1q1pjDHGGDOAWHBmjDHGGDOARHNwNi/UBQiBaPvO0fZ9ITq/czSKxt+zfefIF23f97CidsyZMcYYY8xAFM0tZ8YYY4wxA07UBWciMkdE1orIBhG5I9Tl6Q8iskVEPhWR5SJSEuryBIOIPC4ie0TkM79jmSLyhoisd18zQlnGQDvMd/6RiGx3f9fLReT8UJbRBJ7VYVaHRQKrv44sqoIzEfECDwPnAeOBq0VkfGhL1W/OUNVJETxN+UlgTpdjdwBvqmoR8Kb7PpI8yaHfGeAX7u96kqou6OcymSCyOszqsP4uVBA9idVfhxVVwRkwHdigqptUtQl4HrgwxGUyAaCq7wAVXQ5fCDzl7j8FXNSfZQq2w3xnE9msDotQ0VaHWf11ZNEWnA0Dtvm9L3OPRToFXheRpSIyN9SF6Ue5qrrT3d8F5IayMP3oVhH5xO02iJhuEANYHWZ1WOSz+ovoC86i1WmqOgWnK+QWETk91AXqb+pMS46Gqcm/AUYBk4CdwM9CWhpjAsPqsOiow6z+ckVbcLYdGO73Ps89FtFUdbv7ugd4GadrJBrsFpEhAO7rnhCXJ+hUdbeqtqpqG/AI0fO7jhZWh1kdFrGs/uoUbcHZEqBIRApFxAdcBcwPcZmCSkSSRCSlfR84B/jsyHdFjPnAde7+dcArISxLv2ivyF0XEz2/62hhdZjVYRHL6q9OMaEuQH9S1RYRuRV4DfACj6vqyhAXK9hygZdFBJzf97Oq+o/QFinwROQ5YDaQLSJlwA+Be4EXROSrQClwRehKGHiH+c6zRWQSTvfHFuDroSqfCTyrw6wOC10JA8vqryOzDAHGGGOMMQNItHVrGmOMMcYMaBacGWOMMcYMIBacGWOMMcYMIBacGWOMMcYMIBacGWOMMcYMIBacmaggIrNF5G+hLocxxvSW1V/Rx4IzY4wxxpgBxIIzM6CIyLUi8pGILBeR34mIV0RqROQXIrJSRN4UkRz32kki8oGbJPfl9iS5IjJaRP4pIitE5GMRGeU+PllEXhKRNSLyB3FXtTTGmECw+ssEigVnZsAQkXHAlcCpqjoJaAW+BCQBJap6PPA2zkrSAE8D31PVicCnfsf/ADysqicCp+Ak0AWYDHwLGA+MBE4N8lcyxkQJq79MIEVV+iYz4J0FTAWWuH8UJuAk+m0D/uhe83vgzyKSBqSr6tvu8aeAF90cfMNU9WUAVW0AcJ/3kaqWue+XAyOA94L+rYwx0cDqLxMwFpyZgUSAp1T1zoMOivxPl+v6mnOs0W+/Ffvv3xgTOFZ/mYCxbk0zkLwJXCYigwBEJFNECnD+O73MveYa4D1V3Q9UishM9/iXgbdVtRooE5GL3GfEiUhif34JY0xUsvrLBIxF3mbAUNVVIvLfwOsi4gGagVuAWmC6e24PzrgOgOuA37qV1ybgBvf4l4Hfichd7jMu78evYYyJQlZ/mUAS1b62sBrTP0SkRlWTQ10OY4zpLau/TF9Yt6YxxhhjzABiLWfGGGOMMQOItZwZY4wxxgwgFpwZY4wxxgwgFpwZY4wxxgwgFpwZY4wxxgwgFpwZY4wxxgwgFpwZY4wxxgwg/x+4KVZBYIovSAAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "%%time \n", "\n", "def getModelWithAugmentation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth):\n", " \n", " inputs = tf.keras.Input(shape=inputShape)\n", " x = tf.keras.layers.RandomRotation((-0.05, 0.05))(inputs)\n", " for iHidden in range(nHiddenLayers): \n", " x = tf.keras.layers.Conv2D(filters=nFilters, \n", " kernel_size=kernel_size, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.relu)(x)\n", " x = tf.keras.layers.MaxPooling2D(pool_size=pool_size)(x)\n", " x = tf.keras.layers.Dropout(dropout_rate)(x)\n", " \n", " x = tf.keras.layers.Flatten()(x)\n", " x = tf.keras.layers.Dense(nNeurons, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.relu)(x)\n", " outputs = tf.keras.layers.Dense(outputWidth, \n", " kernel_regularizer = tf.keras.regularizers.l2(l2_lambda),\n", " activation=tf.nn.softmax)(x)\n", " model = tf.keras.Model(inputs=inputs, outputs=outputs)\n", " model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n", " return model\n", "\n", "model_with_aug = getModelWithAugmentation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth)\n", "model_with_aug.summary()\n", "model_with_aug_fit = model_with_aug.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size, validation_split=0.3, verbose=0)\n", "plotTrainingHistory(model_with_aug_fit) \n", "printScores(model_with_aug, X_test, np.argmax(Y_test, axis=1)) \n", "plotExamples(model_with_aug, missclassified_data, missclassified_labels)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Techniki przedstawione powyżej dają różne efekty w zależności od zagadnienia. Nie zawsze są one konieczne.\n", "\n", "Proszę:\n", "\n", "* wyskazać model, który będzie miał najmniejszą liczbę błednie zidentyfikowanych cyfr w zbiorze testowym." ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Misidentified examples count:\n", "\u001b[34mBasic:\t\t\u001b[0m 189\n", "\u001b[34mRegularised:\t\u001b[0m 185\n", "\u001b[34mAugmented:\t\u001b[0m 173\n" ] } ], "source": [ "def misIdCount(model, X, Y):\n", " Y_pred = model.predict(X).argmax(axis=1)\n", " missclassified_indices = Y_pred != np.argmax(Y, axis=1)\n", " return np.count_nonzero(missclassified_indices)\n", " \n", "print(\"Misidentified examples count:\")\n", "print(colored(\"Basic:\\t\\t\",\"blue\"),misIdCount(model_basic, X_test, Y_test))\n", "print(colored(\"Regularised:\\t\",\"blue\"),misIdCount(model_with_reg, X_test, Y_test))\n", "print(colored(\"Augmented:\\t\",\"blue\"),misIdCount(model_with_aug, X_test, Y_test))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Proszę:\n", "\n", "* wytrenować model, który będzie miał mniej niż 100 błędnie zidentyfikowanych przykładów w zbiorze testowym" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 0.99 1.00 0.99 980\n", " 1 1.00 1.00 1.00 1135\n", " 2 0.98 0.99 0.98 1032\n", " 3 0.99 0.99 0.99 1010\n", " 4 0.99 0.99 0.99 982\n", " 5 0.99 0.99 0.99 892\n", " 6 0.99 0.99 0.99 958\n", " 7 0.98 0.99 0.99 1028\n", " 8 1.00 0.98 0.99 974\n", " 9 0.99 0.98 0.98 1009\n", "\n", " accuracy 0.99 10000\n", " macro avg 0.99 0.99 0.99 10000\n", "weighted avg 0.99 0.99 0.99 10000\n", "\n", "Confusion matrix:\n", "[[ 976 0 2 0 0 0 1 1 0 0]\n", " [ 1 1130 3 0 0 0 0 1 0 0]\n", " [ 0 0 1024 0 0 0 0 7 1 0]\n", " [ 0 1 1 1004 0 1 0 0 1 2]\n", " [ 0 0 2 0 974 0 3 0 1 2]\n", " [ 1 0 2 6 0 880 1 1 0 1]\n", " [ 3 2 4 0 1 1 947 0 0 0]\n", " [ 0 1 5 0 0 0 0 1018 0 4]\n", " [ 6 0 3 2 0 1 3 6 950 3]\n", " [ 0 1 2 1 8 4 0 4 0 989]]\n", "\u001b[34mFinal:\t\u001b[0m 108\n", "CPU times: user 13min 46s, sys: 41.6 s, total: 14min 28s\n", "Wall time: 4min 29s\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "%%time \n", "\n", "nFilters = 32\n", "kernel_size = 8\n", "pool_size = (2,2)\n", "nNeurons = 128 \n", "nHiddenLayers = 1 \n", "inputShape = (28, 28, 1)\n", "outputWidth = 10\n", "\n", "dropout_rate = 0.1\n", "l2_lambda = 5E-4\n", "\n", "epochs = 20\n", "batch_size = 64\n", "\n", "#model_final = getModelWithRegularisation(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, dropout_rate, l2_lambda, inputShape, outputWidth)\n", "model_final = getModel(nFilters, kernel_size, pool_size, nNeurons, nHiddenLayers, inputShape, outputWidth)\n", "\n", "model_final_fit = model_final.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size, validation_split=0.3, verbose=0)\n", "plotTrainingHistory(model_final_fit) \n", "printScores(model_final, X_test, np.argmax(Y_test, axis=1)) \n", "plotExamples(model_final, missclassified_data, missclassified_labels)\n", "print(colored(\"Final:\\t\",\"blue\"),misIdCount(model_final, X_test, Y_test))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**Czy postawienie problemu jak w poprzedniej kmórce jest prawidłowe?**" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "colab": { "authorship_tag": "ABX9TyNmNw1lMlOB1SGZYL5DdQYT", "collapsed_sections": [], "name": "11M_CNN_data_augmentation.ipynb", "provenance": [] }, "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": "3.8.6" } }, "nbformat": 4, "nbformat_minor": 4 }