From c20f219ae1f4a6e20608fcd8a94cd8e6d4a78d47 Mon Sep 17 00:00:00 2001 From: Anjana Tiha Date: Thu, 14 Mar 2019 21:28:30 -0400 Subject: [PATCH] added new code --- ...from Cell Images using Deep Learning.ipynb | 1305 +++++++++++++---- output/figure/__results___10_1.png | Bin 0 -> 90016 bytes output/figure/__results___30_0.png | Bin 0 -> 42438 bytes 3 files changed, 1025 insertions(+), 280 deletions(-) create mode 100644 output/figure/__results___10_1.png create mode 100644 output/figure/__results___30_0.png diff --git a/code/Malaria Detection from Cell Images using Deep Learning.ipynb b/code/Malaria Detection from Cell Images using Deep Learning.ipynb index 68236fe..4ea0f5a 100644 --- a/code/Malaria Detection from Cell Images using Deep Learning.ipynb +++ b/code/Malaria Detection from Cell Images using Deep Learning.ipynb @@ -1,298 +1,1043 @@ { - "cells": [ - { - "metadata": { - "_uuid": "b1ca1df7f5915973755864390b0003fa660cf60d" - }, - "cell_type": "markdown", - "source": "# Malaria Detection from Cell Images using Deep Learning" - }, - { - "metadata": { - "_uuid": "1139e3ffd2535759da620e22c5cf5a8a09d72b79" - }, - "cell_type": "markdown", - "source": "Malaria is a life-threatening disease caused by parasites that are transmitted to people through the bites of infected female Anopheles mosquitoes. It is preventable and curable.\nMalaria causes symptoms that typically include fever, tiredness, vomiting, and headaches. In severe cases it can cause yellow skin, seizures, coma, or death. \n \n \n \n \n #### Signs and symptoms\n A malaria infection is generally characterized by the following signs and symptoms:\n- Fever\n- Chills\n- Headache\n- Nausea and vomiting\n- Muscle pain and fatigue\n\nOther signs and symptoms may include:\n- Sweating\n- Chest or abdominal pain\n- Cough\n \n\n\n#### Diagnosis: \nMalaria is typically diagnosed by the microscopic examination of blood using blood films, or with antigen-based rapid diagnostic tests.\n\n#### Prevalence:\nThe disease is widespread in the tropical and subtropical regions that exist in a broad band around the equator including much of Sub-Saharan Africa, Asia, and Latin America.\n- In 2016, there were 216 million cases of malaria worldwide resulting in an estimated 445,000 to 731,000 deaths. Approximately 90% of both cases and deaths occurred in Africa.\n- Rates of disease have decreased from 2000 to 2015 by 37%, but increased from 2014, during which there were 198 million cases.\n\n#### Risk Factor:\nMalaria is commonly associated with poverty and has a major negative effect on economic development. In Africa, it is estimated to result in losses of US$12 billion a year due to increased healthcare costs, lost ability to work, and negative effects on tourism.\n\n \n Reference:\n1. [WHO](https://www.who.int/news-room/fact-sheets/detail/malaria) \n2. [Wikipedia](https://en.wikipedia.org/wiki/Malaria)" - }, - { - "metadata": { - "_uuid": "11aa55f3db0dec51c09334c64fae7f56b1a2cde1" - }, - "cell_type": "markdown", - "source": "# 1. Import \n" - }, - { - "metadata": { - "trusted": true, - "_uuid": "ae198668fcf74479c92b2dfbf2dc3f291b599d0e" - }, - "cell_type": "code", - "source": "# System\nimport sys\nimport os\nimport argparse\n\n# Time\nimport time\nimport datetime\n\n# Numerical Data\nimport random\nimport numpy as np \nimport pandas as pd\n\n# Tools\nimport shutil\nfrom glob import glob\nfrom tqdm import tqdm\nimport gc\n\n# NLP\nimport re\n\n# Preprocessing\nfrom sklearn import preprocessing\nfrom sklearn.utils import class_weight as cw\nfrom sklearn.utils import shuffle\n\n# Model Selection\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.model_selection import GridSearchCV\nfrom sklearn.model_selection import cross_val_score\n\n# Machine Learning Models\nfrom sklearn import svm\nfrom sklearn.svm import LinearSVC, SVC\n\n# Evaluation Metrics\nfrom sklearn import metrics \nfrom sklearn.metrics import f1_score, accuracy_score, precision_score, recall_score, confusion_matrix, classification_report, roc_auc_score\n\n\n# Deep Learning - Keras - Preprocessing\nfrom keras.preprocessing.image import ImageDataGenerator\n\n# Deep Learning - Keras - Model\nimport keras\nfrom keras import models\nfrom keras.models import Model\nfrom keras.models import Sequential\n\n# Deep Learning - Keras - Layers\nfrom keras.layers import Convolution1D, concatenate, SpatialDropout1D, GlobalMaxPool1D, GlobalAvgPool1D, Embedding, \\\n Conv2D, SeparableConv1D, Add, BatchNormalization, Activation, GlobalAveragePooling2D, LeakyReLU, Flatten\nfrom keras.layers import Dense, Input, Dropout, MaxPooling2D, Concatenate, GlobalMaxPooling2D, GlobalAveragePooling2D, \\\n Lambda, Multiply, LSTM, Bidirectional, PReLU, MaxPooling1D\nfrom keras.layers.pooling import _GlobalPooling1D\n\n# Deep Learning - Keras - Pretrained Models\nfrom keras.applications.xception import Xception\nfrom keras.applications.resnet50 import ResNet50\nfrom keras.applications.inception_v3 import InceptionV3\nfrom keras.applications.inception_resnet_v2 import InceptionResNetV2\nfrom keras.applications.densenet import DenseNet201\nfrom keras.applications.nasnet import NASNetMobile, NASNetLarge\n\nfrom keras.applications.nasnet import preprocess_input\n\n# Deep Learning - Keras - Model Parameters and Evaluation Metrics\nfrom keras import optimizers\nfrom keras.optimizers import Adam, SGD , RMSprop\nfrom keras.losses import mae, sparse_categorical_crossentropy, binary_crossentropy\n\n# Deep Learning - Keras - Visualisation\nfrom keras.callbacks import ModelCheckpoint, EarlyStopping, TensorBoard, ReduceLROnPlateau\n# from keras.wrappers.scikit_learn import KerasClassifier\nfrom keras import backend as K\n\n# Deep Learning - TensorFlow\nimport tensorflow as tf\n\n# Graph/ Visualization\nimport matplotlib.pyplot as plt\nfrom matplotlib.pyplot import figure\nimport matplotlib.image as mpimg\nimport seaborn as sns\nfrom mlxtend.plotting import plot_confusion_matrix\n\n# Image\nimport cv2\nfrom PIL import Image\nfrom IPython.display import display\n\n# np.random.seed(42)\n\n%matplotlib inline\n\n# Input data\nprint(os.listdir(\"../input\"))", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "43ec897e74117506f17611b58af485a2e013e848" - }, - "cell_type": "markdown", - "source": "# 2. Functions" - }, - { - "metadata": { - "trusted": true, - "_uuid": "cbe4bc8e76957ae5b139e87a19e75e51be7af414" - }, - "cell_type": "code", - "source": "def date_time(x):\n if x==1:\n return 'Timestamp: {:%Y-%m-%d %H:%M:%S}'.format(datetime.datetime.now())\n if x==2: \n return 'Timestamp: {:%Y-%b-%d %H:%M:%S}'.format(datetime.datetime.now())\n if x==3: \n return 'Date now: %s' % datetime.datetime.now()\n if x==4: \n return 'Date today: %s' % datetime.date.today() ", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "cc1f975dabb307ccf1c219b5021dfc4a97a7a117" - }, - "cell_type": "markdown", - "source": "# 3. Input Configuration" - }, - { - "metadata": { - "trusted": true, - "_uuid": "a8a36cda21a37ddf8002c994b682d91a81c59b56" - }, - "cell_type": "code", - "source": "input_directory = r\"../input/cell_images/cell_images\"\noutput_directory = r\"../output/\"\n\ntraining_dir = input_directory\n# testing_dir = input_directory + r\"test\"\n\nif not os.path.exists(output_directory):\n os.mkdir(output_directory)\n \nfigure_directory = \"../output/figures\"\nif not os.path.exists(figure_directory):\n os.mkdir(figure_directory)\n \n \nfile_name_pred_batch = figure_directory+r\"/result\"\nfile_name_pred_sample = figure_directory+r\"/sample\"", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "9e02a5a1b4538b67bf351ec59a9cfd7f832fb0e3" - }, - "cell_type": "markdown", - "source": "# 4. Visualization" - }, - { - "metadata": { - "trusted": true, - "_uuid": "90fed2e4b122755b68312621acd1cbfa6bdd3ddd" - }, - "cell_type": "code", - "source": "def plot_image(file, directory=None, sub=False, aspect=None):\n path = directory + file\n \n img = plt.imread(path)\n \n plt.imshow(img, aspect=aspect)\n# plt.title(file)\n plt.xticks([])\n plt.yticks([])\n \n if sub:\n plt.show()\n \ndef plot_img_dir(directory=training_dir, count=5):\n selected_files = random.sample(os.listdir(directory), count)\n \n ncols = 5\n nrows = count//ncols if count%ncols==0 else count//ncols+1\n \n figsize=(20, ncols*nrows)\n\n ticksize = 14\n titlesize = ticksize + 8\n labelsize = ticksize + 5\n\n\n params = {'figure.figsize' : figsize,\n 'axes.labelsize' : labelsize,\n 'axes.titlesize' : titlesize,\n 'xtick.labelsize': ticksize,\n 'ytick.labelsize': ticksize}\n\n plt.rcParams.update(params)\n \n i=0\n \n for file in selected_files: \n plt.subplot(nrows, ncols, i+1)\n path = directory + file\n plot_image(file, directory, aspect=None)\n\n i=i+1\n \n plt.tight_layout()\n plt.show()\n \ndef plot_img_dir_main(directory=training_dir, count=5):\n labels = os.listdir(directory)\n for label in labels:\n print(label)\n plot_img_dir(directory=directory+\"/\"+label+\"/\", count=count)\n ", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "trusted": true, - "_uuid": "0e6059fc1714fba66dfd160d7a37c9bfd0566594" - }, - "cell_type": "code", - "source": "plot_img_dir_main(directory=training_dir, count=5)", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "08472c5647d039fe88134dd5bbf51b1bacafe700" - }, - "cell_type": "markdown", - "source": "# 5. Preprocess" - }, - { - "metadata": { - "_uuid": "6368be0cd0b5b944d842b963bdfb9a8b22843510" - }, - "cell_type": "markdown", - "source": "#### During Preprocessing, all of the image has been transformed to target size (224, 224) and pixel value has been rescaled to unit value. (224, 224) is the input shape for Pretrained model \"NashNetMobile\". The target class is treated as categorical and both training and validation image set has been re-shuffled. Some of the images has been horizontally and vertically flipped randomly and sheerness and rotation has been changed to introduce heterogeneity. A part of training dataset has been used as validation set. " - }, - { - "metadata": { - "trusted": true, - "_uuid": "5f0a5bf3d06e33bec36a8b9355483d3432d5c63f" - }, - "cell_type": "code", - "source": "def get_data(batch_size=32, target_size=(299, 299), class_mode=\"categorical\", training_dir=training_dir, testing_dir=None):\n print(\"Preprocessing and Generating Data Batches.......\\n\")\n \n rescale = 1.0/255\n\n train_batch_size = batch_size\n test_batch_size = batch_size\n \n train_shuffle = True\n val_shuffle = True\n test_shuffle = False\n \n train_datagen = ImageDataGenerator(\n horizontal_flip=True,\n vertical_flip=True,\n rotation_range=45,\n shear_range=16,\n rescale=rescale,\n validation_split=0.25)\n\n train_generator = train_datagen.flow_from_directory(\n training_dir,\n target_size=target_size, \n class_mode=class_mode, \n batch_size=batch_size, \n shuffle=True, \n seed=42,\n subset='training')\n \n validation_generator = train_datagen.flow_from_directory(\n training_dir, \n target_size=target_size, \n class_mode=class_mode, \n batch_size=1024, \n shuffle=True, \n seed=42,\n subset='validation')\n \n test_datagen = ImageDataGenerator(rescale=rescale)\n \n test_generator = None\n \n \n class_weights = get_weight(train_generator.classes)\n \n steps_per_epoch = len(train_generator)\n validation_steps = len(validation_generator)\n \n print(\"\\nPreprocessing and Data Batch Generation Completed.\\n\")\n \n \n return train_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps\n\n# Calculate Class Weights\ndef get_weight(y):\n class_weight_current = cw.compute_class_weight('balanced', np.unique(y), y)\n return class_weight_current", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "e02af0cac288607e1ae124f8242db9b1c05688c8" - }, - "cell_type": "markdown", - "source": "# 5. Model Function" - }, - { - "metadata": { - "trusted": true, - "_uuid": "868b5798867b18a80f8d5b12957cd7965a0b65ae" - }, - "cell_type": "code", - "source": "def get_model(model_name, input_shape=(96, 96, 3), num_class=2):\n inputs = Input(input_shape)\n \n if model_name == \"Xception\":\n base_model = Xception(include_top=False, input_shape=input_shape)\n elif model_name == \"ResNet50\":\n base_model = ResNet50(include_top=False, input_shape=input_shape)\n elif model_name == \"InceptionV3\":\n # base_model = InceptionV3(include_top=False, input_shape=input_shape)\n base_model = InceptionV3(weights='imagenet', include_top=False, input_shape=input_shape)\n elif model_name == \"InceptionResNetV2\":\n base_model = InceptionResNetV2(include_top=False, input_shape=input_shape)\n if model_name == \"DenseNet201\":\n base_model = DenseNet201(include_top=False, input_shape=input_shape)\n if model_name == \"NASNetMobile\":\n base_model = NASNetMobile(include_top=False, input_shape=input_shape)\n if model_name == \"NASNetLarge\":\n base_model = NASNetLarge(include_top=False, input_shape=input_shape)\n \n# for layer in base_model.layers:\n# layer.trainable = False\n \n# for layer in model.layers[:249]:\n# layer.trainable = False\n# for layer in model.layers[249:]:\n# layer.trainable = True\n \n# x = base_model(inputs)\n# x = GlobalAveragePooling2D()(x)\n# x = BatchNormalization()(x)\n# x = Dropout(0.2)(x)\n# out = Dense(2, activation=\"softmax\")(x)\n# model = Model(inputs, out)\n\n x = base_model(inputs)\n \n output1 = GlobalMaxPooling2D()(x)\n output2 = GlobalAveragePooling2D()(x)\n output3 = Flatten()(x)\n \n outputs = Concatenate(axis=-1)([output1, output2, output3])\n \n outputs = Dropout(0.5)(outputs)\n outputs = BatchNormalization()(outputs)\n \n if num_class>1:\n outputs = Dense(num_class, activation=\"softmax\")(outputs)\n else:\n outputs = Dense(1, activation=\"sigmoid\")(outputs)\n \n model = Model(inputs, outputs)\n \n model.summary()\n \n \n return model\n\n# Custom Convolutional Neural Network \ndef get_conv_model(num_class=2, input_shape=(3,150,150)):\n model = Sequential()\n \n model.add(Conv2D(16, (3, 3), activation='relu', padding=\"same\", input_shape=input_shape))\n model.add(Conv2D(16, (3, 3), padding=\"same\", activation='relu'))\n model.add(MaxPooling2D(pool_size=(2, 2)))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n\n model.add(Conv2D(32, (3, 3), activation='relu', padding=\"same\"))\n model.add(Conv2D(32, (3, 3), padding=\"same\", activation='relu'))\n model.add(MaxPooling2D(pool_size=(2, 2)))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n\n model.add(Conv2D(64, (3, 3), activation='relu', padding=\"same\"))\n model.add(Conv2D(64, (3, 3), padding=\"same\", activation='relu'))\n model.add(MaxPooling2D(pool_size=(2, 2)))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n\n model.add(Conv2D(96, (3, 3), dilation_rate=(2, 2), activation='relu', padding=\"same\"))\n model.add(Conv2D(96, (3, 3), padding=\"valid\", activation='relu'))\n model.add(MaxPooling2D(pool_size=(2, 2)))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n\n model.add(Conv2D(128, (3, 3), dilation_rate=(2, 2), activation='relu', padding=\"same\"))\n model.add(Conv2D(128, (3, 3), padding=\"valid\", activation='relu'))\n model.add(MaxPooling2D(pool_size=(2, 2)))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n\n model.add(Flatten())\n \n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n \n model.add(Dense(256, activation='relu'))\n model.add(Dropout(0.5))\n model.add(BatchNormalization())\n \n model.add(Dense(num_class , activation='softmax'))\n\n print(model.summary())\n \n return model\n", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "3e56fcfe7a60a2759f1952f30f7fb05726b022c5" - }, - "cell_type": "markdown", - "source": "# 6. Output Configuration" - }, - { - "metadata": { - "trusted": true, - "_uuid": "644c92b9756bb9fd1d980302a315d38298aa8472" - }, - "cell_type": "code", - "source": "main_model_dir = output_directory + r\"models/\"\nmain_log_dir = output_directory + r\"logs/\"\n\ntry:\n os.mkdir(main_model_dir)\nexcept:\n print(\"Could not create main model directory\")\n \ntry:\n os.mkdir(main_log_dir)\nexcept:\n print(\"Could not create main log directory\")\n\n\n\nmodel_dir = main_model_dir + time.strftime('%Y-%m-%d %H-%M-%S') + \"/\"\nlog_dir = main_log_dir + time.strftime('%Y-%m-%d %H-%M-%S')\n\n\ntry:\n os.mkdir(model_dir)\nexcept:\n print(\"Could not create model directory\")\n \ntry:\n os.mkdir(log_dir)\nexcept:\n print(\"Could not create log directory\")\n \nmodel_file = model_dir + \"{epoch:02d}-val_acc-{val_acc:.2f}-val_loss-{val_loss:.2f}.hdf5\"", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "d5133db17f6ff324823b131eacc3fb471943adc1" - }, - "cell_type": "markdown", - "source": "## 6.2 Call Back Configuration" - }, - { - "metadata": { - "trusted": true, - "_uuid": "104999913bc3bd3f1181b365fff5cc23f9a56ce7" - }, - "cell_type": "code", - "source": "print(\"Settting Callbacks\")\n\ncheckpoint = ModelCheckpoint(\n model_file, \n monitor='val_acc', \n save_best_only=True)\n\nearly_stopping = EarlyStopping(\n monitor='val_loss',\n patience=2,\n verbose=1,\n restore_best_weights=True)\n\n\nreduce_lr = ReduceLROnPlateau(\n monitor='val_loss',\n factor=0.6,\n patience=1,\n verbose=1)\n\ncallbacks = [reduce_lr, early_stopping, checkpoint]\n\ncallbacks = [checkpoint, reduce_lr, early_stopping]\n\nprint(\"Set Callbacks at \", date_time(1))", - "execution_count": null, - "outputs": [] - }, - { - "metadata": { - "_uuid": "6fbe80b2f63f0b9972c1546b088b667331108aa5" - }, - "cell_type": "markdown", - "source": "# 7. Model" - }, - { - "metadata": { - "trusted": true, - "_uuid": "27dc13b90ff17042f2912f50ef4a8ad288f9914f" - }, - "cell_type": "code", - "source": "print(\"Getting Base Model\", date_time(1))\n\n# input_shape = (96, 96, 3)\ninput_shape = (224, 224, 3)\n\nnum_class = 2\n\n\nmodel = get_model(model_name=\"NASNetMobile\", input_shape=input_shape, num_class=num_class)\n# model = get_conv_model(input_shape=input_shape)\n\nprint(\"Loaded Base Model\", date_time(1))", - "execution_count": null, - "outputs": [] - }, + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "_uuid": "b1ca1df7f5915973755864390b0003fa660cf60d" + }, + "source": [ + "# Malaria Detection from Cell Images using Deep Learning" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "1139e3ffd2535759da620e22c5cf5a8a09d72b79" + }, + "source": [ + "Malaria is a life-threatening disease caused by parasites that are transmitted to people through the bites of infected female Anopheles mosquitoes. It is preventable and curable.\n", + "Malaria causes symptoms that typically include fever, tiredness, vomiting, and headaches. In severe cases it can cause yellow skin, seizures, coma, or death. \n", + " \n", + " \n", + " \n", + " \n", + " #### Signs and symptoms\n", + " A malaria infection is generally characterized by the following signs and symptoms:\n", + "- Fever\n", + "- Chills\n", + "- Headache\n", + "- Nausea and vomiting\n", + "- Muscle pain and fatigue\n", + "\n", + "Other signs and symptoms may include:\n", + "- Sweating\n", + "- Chest or abdominal pain\n", + "- Cough\n", + " \n", + "\n", + "\n", + "#### Diagnosis: \n", + "Malaria is typically diagnosed by the microscopic examination of blood using blood films, or with antigen-based rapid diagnostic tests.\n", + "\n", + "#### Prevalence:\n", + "The disease is widespread in the tropical and subtropical regions that exist in a broad band around the equator including much of Sub-Saharan Africa, Asia, and Latin America.\n", + "- In 2016, there were 216 million cases of malaria worldwide resulting in an estimated 445,000 to 731,000 deaths. Approximately 90% of both cases and deaths occurred in Africa.\n", + "- Rates of disease have decreased from 2000 to 2015 by 37%, but increased from 2014, during which there were 198 million cases.\n", + "\n", + "#### Risk Factor:\n", + "Malaria is commonly associated with poverty and has a major negative effect on economic development. In Africa, it is estimated to result in losses of US$12 billion a year due to increased healthcare costs, lost ability to work, and negative effects on tourism.\n", + "\n", + " \n", + " Reference:\n", + "1. [WHO](https://www.who.int/news-room/fact-sheets/detail/malaria) \n", + "2. [Wikipedia](https://en.wikipedia.org/wiki/Malaria)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "11aa55f3db0dec51c09334c64fae7f56b1a2cde1" + }, + "source": [ + "# 1. Import \n" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "_uuid": "ae198668fcf74479c92b2dfbf2dc3f291b599d0e" + }, + "outputs": [ { - "metadata": { - "trusted": true, - "_uuid": "77a593fffaf1b82c2b82bbfd8d5335bc7995a6bf" - }, - "cell_type": "code", - "source": "loss = 'categorical_crossentropy'\n# loss = 'binary_crossentropy'\nmetrics = ['acc']\n# metrics = [auroc]", - "execution_count": null, - "outputs": [] + "name": "stderr", + "output_type": "stream", + "text": [ + "Using TensorFlow backend.\n" + ] }, { - "metadata": { - "_uuid": "c975187a17a5e9d07ff15484cac2824fcab1fd53" - }, - "cell_type": "markdown", - "source": "# 8. Data" - }, + "name": "stdout", + "output_type": "stream", + "text": [ + "['cell_images']\n" + ] + } + ], + "source": [ + "# System\n", + "import sys\n", + "import os\n", + "import argparse\n", + "\n", + "# Time\n", + "import time\n", + "import datetime\n", + "\n", + "# Numerical Data\n", + "import random\n", + "import numpy as np \n", + "import pandas as pd\n", + "\n", + "# Tools\n", + "import shutil\n", + "from glob import glob\n", + "from tqdm import tqdm\n", + "import gc\n", + "\n", + "# NLP\n", + "import re\n", + "\n", + "# Preprocessing\n", + "from sklearn import preprocessing\n", + "from sklearn.utils import class_weight as cw\n", + "from sklearn.utils import shuffle\n", + "\n", + "# Model Selection\n", + "from sklearn.model_selection import train_test_split\n", + "from sklearn.model_selection import GridSearchCV\n", + "from sklearn.model_selection import cross_val_score\n", + "\n", + "# Machine Learning Models\n", + "from sklearn import svm\n", + "from sklearn.svm import LinearSVC, SVC\n", + "\n", + "# Evaluation Metrics\n", + "from sklearn import metrics \n", + "from sklearn.metrics import f1_score, accuracy_score, precision_score, recall_score, confusion_matrix, classification_report, roc_auc_score\n", + "\n", + "\n", + "# Deep Learning - Keras - Preprocessing\n", + "from keras.preprocessing.image import ImageDataGenerator\n", + "\n", + "# Deep Learning - Keras - Model\n", + "import keras\n", + "from keras import models\n", + "from keras.models import Model\n", + "from keras.models import Sequential\n", + "\n", + "# Deep Learning - Keras - Layers\n", + "from keras.layers import Convolution1D, concatenate, SpatialDropout1D, GlobalMaxPool1D, GlobalAvgPool1D, Embedding, \\\n", + " Conv2D, SeparableConv1D, Add, BatchNormalization, Activation, GlobalAveragePooling2D, LeakyReLU, Flatten\n", + "from keras.layers import Dense, Input, Dropout, MaxPooling2D, Concatenate, GlobalMaxPooling2D, GlobalAveragePooling2D, \\\n", + " Lambda, Multiply, LSTM, Bidirectional, PReLU, MaxPooling1D\n", + "from keras.layers.pooling import _GlobalPooling1D\n", + "\n", + "# Deep Learning - Keras - Pretrained Models\n", + "from keras.applications.xception import Xception\n", + "from keras.applications.resnet50 import ResNet50\n", + "from keras.applications.inception_v3 import InceptionV3\n", + "from keras.applications.inception_resnet_v2 import InceptionResNetV2\n", + "from keras.applications.densenet import DenseNet201\n", + "from keras.applications.nasnet import NASNetMobile, NASNetLarge\n", + "\n", + "from keras.applications.nasnet import preprocess_input\n", + "\n", + "# Deep Learning - Keras - Model Parameters and Evaluation Metrics\n", + "from keras import optimizers\n", + "from keras.optimizers import Adam, SGD , RMSprop\n", + "from keras.losses import mae, sparse_categorical_crossentropy, binary_crossentropy\n", + "\n", + "# Deep Learning - Keras - Visualisation\n", + "from keras.callbacks import ModelCheckpoint, EarlyStopping, TensorBoard, ReduceLROnPlateau\n", + "# from keras.wrappers.scikit_learn import KerasClassifier\n", + "from keras import backend as K\n", + "\n", + "# Deep Learning - TensorFlow\n", + "import tensorflow as tf\n", + "\n", + "# Graph/ Visualization\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.pyplot import figure\n", + "import matplotlib.image as mpimg\n", + "import seaborn as sns\n", + "from mlxtend.plotting import plot_confusion_matrix\n", + "\n", + "# Image\n", + "import cv2\n", + "from PIL import Image\n", + "from IPython.display import display\n", + "\n", + "# np.random.seed(42)\n", + "\n", + "%matplotlib inline\n", + "\n", + "# Input data\n", + "print(os.listdir(\"../input\"))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "43ec897e74117506f17611b58af485a2e013e848" + }, + "source": [ + "# 2. Functions" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "_uuid": "cbe4bc8e76957ae5b139e87a19e75e51be7af414" + }, + "outputs": [], + "source": [ + "def date_time(x):\n", + " if x==1:\n", + " return 'Timestamp: {:%Y-%m-%d %H:%M:%S}'.format(datetime.datetime.now())\n", + " if x==2: \n", + " return 'Timestamp: {:%Y-%b-%d %H:%M:%S}'.format(datetime.datetime.now())\n", + " if x==3: \n", + " return 'Date now: %s' % datetime.datetime.now()\n", + " if x==4: \n", + " return 'Date today: %s' % datetime.date.today() " + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "cc1f975dabb307ccf1c219b5021dfc4a97a7a117" + }, + "source": [ + "# 3. Input Configuration" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "_uuid": "a8a36cda21a37ddf8002c994b682d91a81c59b56" + }, + "outputs": [], + "source": [ + "input_directory = r\"../input/cell_images/cell_images\"\n", + "output_directory = r\"../output/\"\n", + "\n", + "training_dir = input_directory\n", + "# testing_dir = input_directory + r\"test\"\n", + "\n", + "if not os.path.exists(output_directory):\n", + " os.mkdir(output_directory)\n", + " \n", + "figure_directory = \"../output/figures\"\n", + "if not os.path.exists(figure_directory):\n", + " os.mkdir(figure_directory)\n", + " \n", + " \n", + "file_name_pred_batch = figure_directory+r\"/result\"\n", + "file_name_pred_sample = figure_directory+r\"/sample\"" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "9e02a5a1b4538b67bf351ec59a9cfd7f832fb0e3" + }, + "source": [ + "# 4. Visualization" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "_uuid": "90fed2e4b122755b68312621acd1cbfa6bdd3ddd" + }, + "outputs": [], + "source": [ + "def plot_image(file, directory=None, sub=False, aspect=None):\n", + " path = directory + file\n", + " \n", + " img = plt.imread(path)\n", + " \n", + " plt.imshow(img, aspect=aspect)\n", + "# plt.title(file)\n", + " plt.xticks([])\n", + " plt.yticks([])\n", + " \n", + " if sub:\n", + " plt.show()\n", + " \n", + "def plot_img_dir(directory=training_dir, count=5):\n", + " selected_files = random.sample(os.listdir(directory), count)\n", + " \n", + " ncols = 5\n", + " nrows = count//ncols if count%ncols==0 else count//ncols+1\n", + " \n", + " figsize=(20, ncols*nrows)\n", + "\n", + " ticksize = 14\n", + " titlesize = ticksize + 8\n", + " labelsize = ticksize + 5\n", + "\n", + "\n", + " params = {'figure.figsize' : figsize,\n", + " 'axes.labelsize' : labelsize,\n", + " 'axes.titlesize' : titlesize,\n", + " 'xtick.labelsize': ticksize,\n", + " 'ytick.labelsize': ticksize}\n", + "\n", + " plt.rcParams.update(params)\n", + " \n", + " i=0\n", + " \n", + " for file in selected_files: \n", + " plt.subplot(nrows, ncols, i+1)\n", + " path = directory + file\n", + " plot_image(file, directory, aspect=None)\n", + "\n", + " i=i+1\n", + " \n", + " plt.tight_layout()\n", + " plt.show()\n", + " \n", + "def plot_img_dir_main(directory=training_dir, count=5):\n", + " labels = os.listdir(directory)\n", + " for label in labels:\n", + " print(label)\n", + " plot_img_dir(directory=directory+\"/\"+label+\"/\", count=count)\n", + " " + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "_uuid": "0e6059fc1714fba66dfd160d7a37c9bfd0566594" + }, + "outputs": [ { - "metadata": { - "trusted": true, - "_uuid": "6daf75886fbc7d444222398f04637e5cc4aed9e7" - }, - "cell_type": "code", - "source": "# batch_size = 32\nbatch_size = 176\n\nclass_mode = \"categorical\"\n# class_mode = \"binary\"\n\n# target_size=(96, 96)\ntarget_size=(224, 224)\n\ntrain_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps = get_data(batch_size=batch_size, target_size=target_size, class_mode=class_mode)", - "execution_count": null, - "outputs": [] + "name": "stdout", + "output_type": "stream", + "text": [ + "Parasitized\n" + ] }, { - "metadata": { - "_uuid": "cd11c2eb7719f3f866c9e9ee201ca67e556fc73b" - }, - "cell_type": "markdown", - "source": "\n# 9. Training" + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { - "metadata": { - "_uuid": "8e335384316089698ee7e1a2f2cd1d6ad2ed4df9" - }, - "cell_type": "markdown", - "source": "#### Trained model on full tranning dataset for 10 epochs and validated on full validation dataset. Adjusted class weight has been used for trainning. For optimization used Adam optimizer with learning rate of 0.0001. For loss calculation used categorical crossentropy and for model performance evaluation used accuracy metrics. " + "name": "stdout", + "output_type": "stream", + "text": [ + "Uninfected\n" + ] }, { - "metadata": { - "trusted": true, - "_uuid": "5ac153e86d2338beab9cb0a3693f9c61fe7d3f04", - "scrolled": true - }, - "cell_type": "code", - "source": "print(\"Starting Trainning ...\\n\")\n\nstart_time = time.time()\nprint(date_time(1))\n\n# batch_size = 32\n# train_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps = get_data(batch_size=batch_size)\n\nprint(\"\\n\\nCompliling Model ...\\n\")\nlearning_rate = 0.0001\noptimizer = Adam(learning_rate)\n# optimizer = Adam()\n\nmodel.compile(optimizer=optimizer, loss=loss, metrics=metrics)\n\n# steps_per_epoch = 180\n# validation_steps = 40\n\nverbose = 1\nepochs = 10\n\nprint(\"Trainning Model ...\\n\")\nhistory = model.fit_generator(\n train_generator,\n steps_per_epoch=steps_per_epoch,\n epochs=epochs,\n verbose=verbose,\n callbacks=callbacks,\n validation_data=validation_generator,\n validation_steps=validation_steps, \n class_weight=class_weights)\n\nelapsed_time = time.time() - start_time\nelapsed_time = time.strftime(\"%H:%M:%S\", time.gmtime(elapsed_time))\n\nprint(\"\\nElapsed Time: \" + elapsed_time)\nprint(\"Completed Model Trainning\", date_time(1))", - "execution_count": null, - "outputs": [] - }, + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAABZgAAAFJCAYAAAAWgEhdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsvV2TLrlx55eo6jOk/BkcojiitKJG+0EcIUuUluRoHWtJtiWuLh2+2HCE7bDDEfaGfeMra/kmcsS3GZISScn213CsLZLDocQhdz+Ed+f0U4AvqhJIJDIB1NN9zvTp/v8uTp+nCgWg3lBA5h+JkFIiAAAAAAAAAAAAAAAAAOAsy4ddAQAAAAAAAAAAAAAAAACvJjAwAwAAAAAAAAAAAAAAALgKGJgBAAAAAAAAAAAAAAAAXAUMzAAAAAAAAAAAAAAAAACuAgZmAAAAAAAAAAAAAAAAAFcBAzMAAAAAAAAAAAAAAACAq4CBGQAAAAAAAAAAAAAAAMBVwMAMAAAAAAAAAAAAAAAA4CpgYAYAAAAAAAAAAAAAAABwFTdnEocQ0ouqCLgf3vjVjxEddykd/wkU6PgP76j40fv/Zi7v1z82nf43X//lumzFdJm/+rF6g6j7bB5clx+//2+n0oP7I6VkPwAAAAAAAAAAAAAA4FEQUpq3GcPA/PB596++TDHGbpoQQvX3k5/+k6m8f/SdLxIR0Ruf/dww7Y+/+6Xu/t/8zJ9Olfne975CRET8nMrndbbeXJfZMsH9AQMzAAAAAAAAAAAAwOMGITIAAAAAAAAAAAAAAAAAXMWpEBng1YCVyRpLBfyi66DLOls2q7HvUueXcb4AAAAAAAAAAAAAADxFoGAGAAAAAAAAAAAAAAAAcBWIwfzA+eG3v0BEbdxkvm/edm/bDKP01n5dD+8Yb793XjNlnz3GOz+9/R//wZ+Z6cA8iMEMAAAAAAAAAAAA8LiBghkAAAAAAAAAAAAAAADAVSAG8yvGSI2bUnLVzV5eOg+dfqT4DSGU/0c7LefIWTd1Ow7b9HETsZw5zeIKr8ORd1/pDAAAAAAAAAAAAAAAOAcUzAAAAAAAAAAAAAAAAACuAgrmVwSt4GW1blEAH2rkGPP/hVy4SpvzYtWxLkMrfB0Fc0lV9m9eWpWnp65u6uDVRZZzpIlOWczImxKhbAYAAAAAAAAAAAAA4BRQMAMAAAAAAAAAAAAAAAC4iuDF1zUThzCfGJzi3W9/0dweYzS3e6SUKKTy/2vw4h2PYjWfyXP2WK10ts4p57XYKunZMnVZsRFP72W/8dnPTdQcEBGllCALBwAAAAAAAAAAAHjEQMEMAAAAAAAAAAAAAAAA4CoQg/mB4amHQ1q66UosZiohkQdK5tntvoK5PTYs9rFtQltdrMvyYjVbecWtPoYVzToPLy9v+7VKcAAAAAAAAAAAAAAAHjswMD8QOBQG/11o3XdwhAEnXIOFNtJ6abmsZgHBNsPqbzbUGmVuqQ7pMQpT0YSlUHXShmYL3rcdx/Lv5TDKh2X/yzb60QKDZ0J/AAAAeFwgHNjT4rde/xgRyeWKd7gn8MP3/810Xm8cef3oxDE9uG5n6gBePggHBgAAAAAAAzMAAAAAAHiifO9/+W+JyHfG/6MT6y58/8jr1z/zp3evGJW63Vd+AAAAAAAAvChgYH7JlMX8DlUtD2hiOrYeoSK2Wo27XS77/kONy2rirHg+tscYs3KXCeQodVlwoUNpLKpux/GLs+De/pfL2I/dto2IiG7C/ntdd0V2imoAp9THy3Jj5D1YJPD4/8qbjqRB/c3iar48qw6dUd+LXARf6+P3D7/9hUbljIX/AAAAAAAAAAAAAMBTBAZmAAAAAADwJImX7d7yYnHAveWHNSAAAAAAAMArAgzMLxkO0xZUXONWscsxiXlwwcpg3l8PiC6HwjmE4Kp+vYGKF5PYW3jPyp/TrlnJbKumR4MlVmSP6iy3u+e1qHjO6nzWUvn976oWUqS6LnL9wkiI1wwAAAAAAAAAAAAAQLOeGwAAAAAAAAAAAAAAAAAwQzgz/Q4rq9+dH719xGCOSg2sZmgmJ8Yyxzb2tq/rmpW2swpmzUiVq9XIIYRGBa2VyHq7p5o+8zxOxWeWvw+FMv/mcNb5981KFlnJHWKzjf9+8k3EYLbAyuoAgFcN9HUeNj/97peIqN9f0H0Qiv1+gszPy3cVM7PO9F14/YbpvgqVvp1Ow9svcRseE53z+fVP/4lbV3Ad6OsAAAAAAEDBDAAAAAAAAAAAAAAAAOBKEIP5JcMLwDTxjJXYptnPi9CwApjVw8TxhA9FbUxZtcLMqoW1alof11PbeHnq7Tqu86iOGqkrS+ovx0XWZeVjI2/n9KH6nW4vdZ1Z+Lyubd2PfWHZ//Ojtz9PREVBxEDZDAAAANwfun+0CGVx/v6rrsim+jVLao+h3K+yaVTRFot1tMo/qL4Hn4+s72r3kbgOll626VepYyHLBwAAAAAALxIomAEAAAAAAAAAAAAAAABcBRTML5msgImshqXjr1LfxlpJc9nqeHtazSKVs54yeVaRzPGc5+MhL7mmSQeTdvCUymciX7YKZfWbr96RZwxKmc0SID5vVSUOw5k4nyOGc0qpvcahr54GAAAAwD3AQuLjM1x9bfXUpoOsWOb+QGz7KqOZVTwDraqKUiFzP6Q3w2s5+hKl/tbMLtXHOH4HjrPM/RERW5r7bjkWs4rJnGhCgQ0AAAAAAMCVQMEMAAAAAAAAAAAAAAAA4CqgYH7ZRBVHjxUtSonDQpkS95dVMayirY9nUkpunONGJe0qmVez6r6guS3TqwMra/L2WJ/4SMDcUwbrMt14iVzWUqu/WT2dy1j4OixVXQNJFfSRFwTLAAAAwCnee+cLRNSfMbUqLURW7ao+jfy/9/3Xs8Tq41U5vK5DjqFc8tR9gNJByh0Jq/CjbnU/KRpJQ6r7Z0156hoQEW2qB5XPBjOqAAAAGIRwZu4weBV54+MfJyLZFfDWw6r3z+D13Wbz+uH770+XBR4GKY2tXjAwAwAAAAAAAAAAAADwSPje//w/EZEfZpWxhIeeYJDxnPmeiFHza2/+QXc/eDWBgfke+PE7XyKiWpWrFbo59DKvZK4VOE20klFMX46tx9nk9cJF2jqPtnGo40B76LItJXSrADqOaRoeHQfxUHAvnjdNN1Dtyu8pkUpT4yujWE3NZaiy8z3azyEusi51fTx1+Lvf/mL1m/d/8s3POXUCAAAAngZBxVO21Mib+oQ3M6OsT7wXbrgVI1P5jtcHef2R44eZr/UzH5e7hdxPODaHtrIcPznnw32I42TT0g7eFl5W4ui7JI4zzWlWRMUDAAAAnhK5l+DM8Mo2DE5/wsDclFUMU1Wevi0KPEbQ2wQAAAAAAAAAAAAAAABwFVAw3wM6nvKuWq5XCW9iJg88OBwXmPNe17X6zehpDsuyDKcrjMgq6yM910UjvVE6PiCXFB1lr/Zk6biJbVxof0X2kvd6/OVjJldMz7etLrupQ8wZt+qiG6VIGsSiBgAAAJ4627arbEv/qZ2lFOKlm0cM7TEU7W+u9S3O5XTSeMeMZ10RJW89CD4mtX2stOlEXC7nqTpbVFTNOWb0ombAIRQzAAAA8CTRtha9LlbK/a/Sr/FMR6VfUdtOmv6TXg9M28zAowQKZgAAAAAAAAAAAAAAAABXAQXzfcBKkm33ymxbosCx73Ks5N2zs9Kusk3BVr94ahitStZep5462VPkekqfZ+tNVWa8aCnNka/MP9rno+upz4OV2W7dHGQ+nEejEld10HlrhTaHX/RjOXPZqVFa8yGb8gouHPNwgXQIAAAAAAAAAAAAPj/55jeq39q+4c361jOsLpd9Flg4bFBsm2JTFa9F1ZvhxTQ2Ej7WOYc2Pdtc9rr/+OtfM8vzZtHz7zyz/0iPxQIfFjAwAwAAAACAO/Het9WCx8aCfbSp34djN2UndWyO4UX8XIe7sfivFyqMkUOefIx2ToebqtwiHBDH0NjJz+hwamWKaTu4Gy5abDmt85p+vKjf8TsP0Pby3/vm50t+aoHlT/4nfzY8DwAAAAAAACxgYD7Bj9/ZB0954KJUu/Wgoa9MZryYNbPxknurctr1KnGSPfQgyKubtV8rlnVddB7eX4+esrmJJeTUYeQB7J2fV4d4xJHMymV9HVJ97I/e/nydwaF0fuOzn3PPDwAAAAAAAAAAAI+YpO0a9e4l2jaXcLjQ89+UV8ZS+U2sIdHYPJL4t9g7PHsWLzGh8/bX92JxAC88Uafz1iIDDwsYmAEAAAAAwJ0ILPBlxXEqHf/tCCGWVMitPBRRauV90+EcV4MsHjRpFbHcpsdEehBSq5U5v/qYGPT005JHCIczmdOyY749M3E8L/5c19EKdZbUsXnwxiplY8W+fHzO96jTcVF5nLYI9XNgBXaTGwAAAAAAAOeAgfkMHKtY/WX1alH+BlrJVsGOpj16Cl3PYzOjYNa/PQWzuwKo2m+lP6P2tfbPxJL28vViEI3q0HjunFjNVv5NjKBj5LYkL+a0io9E6rwRoxkAAAAAAAAAAHjShJi9xdVfHV7LC7+lVcJ5ba0TKmDPvuOtb9Wk5/WtlNJZe/SzM53tZxwlQBmt4rb/3o6YYOsN7CcPERiYAQAAAADAnWgGOandR83gQw0eLGf51o+nfKZuFiNHONOfNlpv76VtwpZ1QoS5CyDPlEM8+OTf1jU459wHAAAAAADAAwbmE+gBCq+6mUMy5/FTzIvDaEaDEm8gsm31tFJvMGAOUJyx06y6uKdc1niDuFl18SiGs/w7qt/ZuM5BxdSOqR4s985HxyAql+HIM9YxmsPK6Y9k9zCABgAAAAAAAAAAwMvnvbe/tf9H2S60E17bPNa0Vvl4imPPVuKtJaXLZWZneMttntNb/21tZvXxOv26rnb6rZ5FzmHUdDi0d7/x9SqdVnFr1fcnPvumcebgvoCBGQAAAAAAZN5754tENBhoxP4gJsQyhTJFnv7I23hwoBedMSozGNiMtlnHWvvCQFks8+Zt20R4MM8JfoZmIJnyjpIm2GmzJ1stOryzVGne/fqfj+u81osZ/9pn/2TuJAAAAAAAwKPmSRuY3/vrt4hITuesBwqX57fVbx2LVy/QwizLkmPHaEaq4dEAyvVWHZtDar00evWWkTp4ZhDn1ZuV1rMDKO1hOhvD2crrmvqfSRdCGCqyYzPoPJ6dw0HXKJ6PVD/81r+qtnO6T775uam6AQAAAAAAAAAA4OUSL1qtu28PWx1neDkcvJfLZf9L+99Gdcv5agU02wq0Klil85TTi7ZldOwgzVLHjnK5LMbMB9jn4Kmzy2G1TceLM31MEBcxoetrzuY4XtwY4cBeDk/awNwYjtXLwQ9rfqiPhzg6L2xloJy3i+5lqxdJG2p1MPVFLyJIpU43NzfVvtkQEbou3n6v7kT+FAfPiK0Ny16okF5swmvPzwtQX6Zh1Ib61CkmqZ1LVmftv2NY8h6rrCaMCdo/AAAAHxLNN1s6zbMjte7401b3SxYqob3yJy3ag4+ch+GcH4UEu8oRbR1zYuBRzvH43U9c/2z6WKFJVs6Z+zm6v2KVo+Jc69PJ5Yi4zUo4EXK/hw9pFznONTnZzwUAAAAAAI+bJ21gBgAAAAAAAAAAAADgLrBIcNViQCcOsVbp8t+sNHYEcb3Fi8/Uc2btqqKO5n1H3RdbGOgePxn32VMsM/z7cpyDvmbM6Dd4MTxtA7NanU+vXt5MORgol0MUqhBH2uGtHj4bNzCra6lujOTxZwOrn6V33NlQF6Ng8PeBp2AeXYemgRU/vY9Cmwc3ZPVHI2xKsdQ0eHe7RwAAAMC15G9V1IMKGR7s6H/ktKH6ncypOKqPwJ/OnL91zHmFcvNF1dM2jX3XMDXrKtW10TOWrDrpaai6OxKtg5pJrGpabGpnTPHMPD4kLnZfhqeX2uUAAAAAAADw1A3MAAAAAAAAAAAAAAAI3v36N4loLFijjZ3udZiq0XGe6FArjPX21VE2j/BEfXkxZsN3XLKuw2bFbWsTSwahx7ywrJ7qW/9dOL61Oj7X1jnX97759n78TR3a9dfe/Gy3vmCOJ21gdhfMY9n9scgfv8hx6z/sTIwxxyI+O43BVcOknGD/47yvIYTpOMajKRvXKJ03p6EZxUuenVrR2+cFxdfnpVXf/Jvvmc4n/5XlDQLvMzmOtxHHsKp7qBs4HWDxR29/vpzHoSR647NY+A8AAAAA4FUnhGY1DvDI+Mef+JXyw7jbvMleICu1iWcnEcw8WYPJI3/3s19MFgZeFkkvAAQAAA+AJ21gBgAAAAB4zLz3nS/WGwx/t17Mb2OHel43Lhpp64XhmvBhTaAKyiHJPAf2In6yg3ZT1pH2GH+MXco5NphObidUGTt1808rLmH9Vy/OZ9elr8ixmBEppEUdn+x85Xk2tbVjb1SGpu1YvJHzeffrf16Vk4zV2vU5/vqbf9o/GQAeIf/H//Y/ENHx7qS2ffQWQpf75O/RIufesRa9uKtERB//3T+aKguAx4b+/uZ3ZasX1m32O/B7dblciKiNC5zLc/oKrFwehUTV+QZaq/0sCizv+VhY2JxbGqiom/5ALbg72y/S4sBG0az6XkmVo/MJm70f3A0YmAXxsr9o+YU7xLh54JBjER4v9KE6DfwwH3+WZWleFFbHatwHmsuiPGqp8uu9kKOYwrNqYd1A9eIo8z7vPL0y9G+vLEslzsyqxd141qpML1B8lCMsDh85KEtP4eBnJ1IduJ+fuRzQX035oCWU+2eWCAAAAAAAAHiV6c0gldu8BbA0vf1yn1XerBEbAAAAIIKBGQAAAADg0ZONB8nYdpA1JofDPTvWDcNDTMppzerXnLQ1hGgWtdkyZuiV2GNUDmoj32vUKLOqo5fJmVBlgbUPHBtRnw5vkOepV4AfiBiIimI9HcfGVa14XwJPiqJrEQEATxGpuptZ7H22zZk1NMv0Vrl3qQMAjxXdB9HKYsYTz3lxgRs17rE/h+RU6RtVr54FoWZsxbq6lFItZvOU2b12ZBiHekDQIUGbwED9GWt6e9Z8ZpFnLR5c9D3J3aBaxT0byhbMAQMztQ1AfvFjvT/kaZi14tV6Eb0XbqQuHqWbYRRr+Syjl7tXpndMo/Bd7Gs6o9QeXdPZ63G2I0c0bpzdON86XjSP7fJHQH1mZAdwqpaPB8QlBC+C35KxEK8gENHf/cMv7qMq4J5BXEIAAAAAAAAAeLnAwAwAAODJ8bdHLMRrCSHQx38bMQnBw0cvUJxiG0+5/D6OYcVMVjAbjnSl5tE+2miUw0qg7Fx1FuGtynEcznq//r+V1kvXS3tfXLvIce+YGcViL39+NrLzO6nrn33craDAi4VoqiCHtQTg1eXnf/tVIqpDBs7SE+tIZtusPINBvZcyvadU7pXx87/5Szc/ue1XfucP3XMA4CHw7tffJqI2PCfHVNahMss7tamceHvdBxrFFeYvIscLjlyuqk8IS5U+5fCpVG3Pv5xym/UxBv2Hlzlroa2DF9PZC2t6cfKp8/faRS10/Mk3vlVt565vVkSrZ2JLe/kcrpW3v/77n/ZO+UnwpA3MaWOl8v6XF7Xh8U3c6oYkqPmGXkzfEELzMjdlO0rd0fSE2SDovbI13nl4ZZ2J0TWrFh7VwWoMR4HedbpRGTqWc6Nsp5KPG0DfuS9l9qk9MEt6UEc6n7KHP0jvfntfuElPg33js58jAEAf3Z5X+4z2TC80RtPLtwMAAAAAAAAAAI+bJ21gBgAAAAB4DPzk218gIqKbZe/a8QrleqVzGfd4U4qdGFnJyn8Px6tQFruhoZT62Yq3KzTNVV16DvGRctlyCHmxfntx9s44ynX+o0WCrbqNHO2STa3uHpY2njLnxg7rsKn8uU7VMcnc5y32vO880hxqqnA50twcaRbjfnDYxeOYn7y9P6vs2N+O2JC/8eY/b44F4KFzVuQTQiBrsXCtjrPylkKVs/FPLWT7NhN71YofrdMA8JDhuLu5L6TeRS9msha06e2M7g/wd65RLKvjve+u954/f/7cLFcfp/sjrTLbVjhb6Hefr+WZ+O9W+plwrNZ2r681UpHna/OsbXPP1JP7vTmYJ7RHRPTEDczNlEyOhq6mhGoV6cxL4A3A3N9lZGDun5ka6qU9M4ixmHnJRjGHz9Zt9kW38hgplj3OdNS8evoDXjuofWnUdQlald350Jy8nwCA/d2bnb1B5C/yAQAAAICnS8ozokIOHbSwAYOC6K8nTkXLehho+MgQaNuO6d4kZlQe3hkr5IV2aN2sqwhvdNQtRYqkFuIMoenLBCozd2UfaDEcU3qMa9UNAADA0+RJG5gBAAC8mvzs+1/rxGZtVYRlXzK22b9HsV3f/8Fb1XGWh5z3ffx3/2j63AC4C5s2Eqj98jnPaSIrWY7tWZXSOUYpNu7D5dKLR3qXeIE6fp4Xg9TCUiWfXeBXHj8KoUZElBZbbeO3eZTjZbPBKDrtW3Ud17pOU+IGFY+b41LmY5U4Y/+x/8mO8nz5jv/4gm8AAADg3igCw/r7tunQqErhy/tZkewpj91Qm8nOx+vreI6cUWhPnU973rWyutd/cc/F6cc0eal8/FCh9Tl5fSmt+tbXYBbOn2f6efksR+eEZ3uxL/HmxjahhnPVeLQ8aQMzK5TL3/pF1orR3HlXeFMjztXFV+j28u797sUv7uUxyrs3MPMMLd50Uq/sURxoabiZmdJh5a3xPhT5HDrq6Vl1dP6d1PmtukFW5xITZdUz6YEgFAMAAAAAAADcF7/4268R0d5X12OMjba2b09i9iuV/nw+disJn3MIo4N4uZRp55cy9X1d9z4/G0IqZ9mR7yXe7sfJWVdsqCExtuAFX2WdLNdgVlyLsfFSDEbZt2iMfazFAEu2pSzt7EspwREPAACPgCdtYAYAAPBqsixLxxNO1W+ZZl3qYyxk+JmsJDCcSZ6Ty3KEAfCiyY7Ly6EoPbYf4W2zWifHZCbKloLyDHNa5cgUDvbs2uzEQJ7lrIO4l4e1raf4HeXjqYd75ehje2XmtsVQied7d+K6XON0X5SCKy59UcD+o65/kya1xyQVQ5ofNDaCQcIMAADgPnnva29Xv5s1GI7PDztv+PvkzXTSs5A0ervO5/aY7ZNFZMmeCZW/l1stfLNitPcYzUxitCJbKrlHZYdVXStH++appL2YzqwQ5nL1PTob/927N3q/Li+G+hpyvTicUI6AyvU5Anu//93v7JvV7Lmo+0Ii74//3j/pnsOrxJMwMP/kO18iojLI4oe3lfdrxfLxUB17U97POdvK5ZSS6ZmVaAn9rHJ5NjazlffsYHBW0dwryzt2ti7eoFBu9waQZ8/3LEEYnXR9mdG1ap4HFd81lIT7n6Wc75LVETt6GuwPj4WedN3e+OznunUCAAAAAAAA1EYJbaxZF+lATvlfXvySTTRLEoYbaVzYaoPGtolF+4S4ORwGi3g58k21kjnnd2zLBpLnt7m+SS28mYIcI7XjFeEeyz+inMWrDGFBrgGak7RjLzleO7tODgAAgFeDJ2FgBgAA8LhYlkUM7qL6u2NOa70Yg6lmYMODIDGAMsLXuM64LPEjBOQCAAAAXkGkQVQr3sK+fN++n7seKVG61MKYLaVinBYzRHIM0Nt9482y5P7JwrK4lHIojUVFtNjS1ghrtmVrlHjLsuSDc19nCSXmqFiD3FRs5lCH5VpspGLVxiLo4ZB/Zs/HWCgQhmXwovHi+J5d50HH//XiBHsCN08ZfW2o0t6aD738dPqs1jWui7d2hAzBIxkJBJt2VKX3VOKje+XNaNX3grff3u4OOFYk8zXIDkQVk1kfl/dT/zwD1dfNulfhZm22veo8CQNzbhCO2FNFwbzvL1P2tAq1H+M2BA5EzuW0XmWPUexeT5U7wyjP0XFn1bdy27XqaK8OXuNqTU+ZVU97ddBTRPTxllfea7jOxoNuP1aqkReGsvaa2HkCAAB4NXnvu/vMKx0qofm2iM9VLJ2a+phjc572uBnfT/aJHP2ehXimTDFI3AfN94s/ccfv9UNYU+Bsv0RuG02lleT7s9b7rB7jKNzFSkZ/ZrKPJ/sIIyOPGRakWwpV4VRyPup5yq5BHlQe46v33vmiLLzK4x+9+aejkgEAAAAAwIfIkzAwAwAAeFzc3t4aCuYaOQ0zO3MuRXkzckpZC9LINDOGKZ46+vd/9eV2nzb4HH9f/9Qfd/MFAAAAwBheqI/I/rY3/QD5bddh62T6mLLT7gjTSTGWPkZiVW9KlFhVzKq/GLJKOTue4kIpxxk98k1E2/93LODHoTeOMqWqmlV4YSV6tnC+XGGidHM4wnjTKtaV2Mo1Ccr5FbdYnGjyImhH4XFYWEJW9GXBjbhm7EiyFmoPIVSLKjJY+A9cy0is5imPdboPG1etm3WQ9v62bav/bqw+YEd/dviLsQ+/uzdK0TtwNXuCwZGyWCuHvbjYntJ5dG91LGlPLDBSn+u/+VugjtPr/lhiRS+m96vMkzAwc5D0qD/Om05py/bPIqc+5bz0ezhoxEbTLrzjesxOyfDK6m33pkK8qLpZnUOvMRtdw5H6uHfNdYMoO35W3bwGrcRl4zKX6neKzcNa6hOcwPtQMoNHTNqi296UdzRQUIO48m4l0dfy2puSb17PYS1TV7VSVLOsK10+eF5v487NsuTFH3K9w+PrZIDryN+K43d+1lP9vVrko5cVy0p9HNkIUu+X5YhMju31lD2ZbjRY0wOA3vlpYieN2y8wZhUt/S6EXfaJKZl3ma016mNY+5r+jaESHh3Tq0u+V4PprfvG/c/GCxYFtYNaVbRsd806x9Ack665iQAAAAAA4EPjSRiYAQAAAAAAAAA8baSClpQTRS6cLR0t2RlD7ORLlC61YCmllJ0lOTZRWigcEt+Y14AIWfS0HLLe7fmFliMdxwNdtxuiD6g69nJ4u5dloeXZfuzCIRvXhT7y7CN7ulgW+YtK5RRTKgvfixihN9Q6grLjW8RxDiqUUUycBxWPE/s3g1BBEzvYA0XDEcVAIAPO8O5b3yKiViW7bfUztm2X6rhZMZsnDpvFU/FmwcpB89w74rSMUseOhGYj4V3vXRyFMPUQQYywAAAgAElEQVTK9Bzds6pdr26esMELi6qv+bNnz4ioxFQenZenYOYnQca3r+oZ1e/QtnsPRSl/nzw6A/NPvvMX+3+OD/flEikd038iTz1yxKDNQ5Pq7Ux5aI+iDOVrUMfyHi3N9/LmB96T9PcaB+9BHSl7dV5n0CpgXeZsniMlsw6+LtOMGoVRnrMveIzRrac+X/3xsPIiKlNQckOoGqBAN7mDmv/ma3qkzXNcjvPiuh3/+fER09Otg7pOv/kZxDsED5ccT5aK2rJMXRXpEsfc53eqbS/Ku9QOrnRcdGYmRMa+v86zbiJVnhhTgQP9TWB0/0Q+T6xupk11cvJUbR7kl4PK9GxlMIj2dD+57czMI2+g1mO232Cm0538bbzQiqaXRk+H1N9P2RfiS5sXxXEGmz3a6zY/IPEGWnKbnPUhf1v3q7kurIxnIxPbl8QtWLOBqy6nGKGK0SyXk/NBwwjuF/v7b//m9GWsd7wjly0bfYNsP1Vg+T09bxPvVUzVtrRRnm0Snx8GsssHFG95ysDR1sRj8b605MWhjiWAKKVI4fixvsaLTAValDE3hXYcIxdOttq+fI4hNFNEilFd9JW47VhCuRY8Ng+RaOX2orQxmIkJAACvPo/OwAwAAAAAAAAAAAAAHi9Z8b/qkHi12Oty0Y7MviDPc3gURXQdFnMWTw08cqg0Kl5ne94/KSTUQkkpjvPOVcdK9srw/uoyvRjIOsaxjr2s042uhefY52dIw+fpCT95Jgqny763Q9Sw6XodAkI5U4YeoSPt0RmY8wMnFMTJia17Fs+za/3WJY1eZv3iedMmZlTIZ+L79Zh9WXtljZROozjP3pSEXppZrvWUz0znOv2R0OcX6u3d+JdOwPwmFmVWR9v5gMfJz//PbxAR0cd/+599yDWp+fvvflW0D/77wm2lbjPTJt+PWh1nt4tUpVF7j796esuS92XF6OarDa3OYtP2LyUtd0x0mp++89W9zJvyHfjEp//QLRc8PrJCn9VyUT3beQpVOYY7rE2fJ6r+g3jMo1Kfzk6DlHjfP9mBHw0CrL6CHnjMzDbyFNIbK2XVtMYZZr7DTBYyirTLoH9w5lucr8mJfl0v/9EMLwvdT9UDvXzfRV6xOee6rV2O4YiMu5zP8fjzk7e/UJWTjnb/N97858M6g6eL1abI/2/q+dXpiPb2OLfJHDFiS3lxP8qi3VW00awglu9t+abzgnfbbSn/cnsomA9l9O0Ht5T+/ZH50Yg9f76v6xBjpNd+aTd6fPSjHyUioptfWnPbdhH58mLD68ptaTm/GzED7KLWewkhNG1ECKFWYpNcRyaxCDs3gSHF/CMdM0siEekoAym1izKHEOj9H7yV9/M2LIQMAAAPl0dnYAYAAAAAAAAAAAAAT4drnbijeL6jEF9nQ496YUWnfw/SjcRu7DyaEep5CuOzQkLPQa2VyTqfTYd+G5yDFihoRxkrlq1QQVY+mnJ+nK9Tf1LPjLGYu17w/THw6AzMWsEcYyzxrZqHRN/Q/os82m6lGTVynrJ5tswZNe2IUSMxOq4Xy8xLO6v4HTXyVn1HeZ398PTSex8b71qOlOy9unrTRqKe8rP2PxSeuh6xzh4nOnbwQyFGErFI630hlFjyrLDkv0yqjqunbnE8QKkgDjm2aFsX/x0XsUpDPR1MHuO9x0ct6/SbVDerd3rhenM9It7LJ0p+pjbV+c0C5vbd4QWnolLVkZoFcLkYMZgPuNOvpz9ayr8z0zu979cZVbKXZw890ym3B7E9n9FASZIHIUut4ON2xppiqqdm9q7b8L2/g7K812fzcrXU6KO/3Rl2hkK0LbQ/S62dKwiAQYgUqDZmxG3Lz9V6fNvzFPCLUNIej1jYiNaLUvuLR7Y8m0m857u6eCHRFgilbzq2s3EhXVLuv2+HSvmy3eZFotLzvcAP/t0HR70DcVMe075gVaSQFdnp5hgPv3aTVcVs3FjWlbbICxQei1/drBTUzIKwZ76n41kgMYp2la9TuTaL0a7l/x/1vQmBIs8OW0KTPq9DsywUVvSBAADgVeLRGZgBAAAAAAAAADxtUgzVwqZEtTGTlWxFoCQWS4/CKaT8GZaPJ6VUjNLsSxEHJhnGy3D+s9H72bPdYLz7zg9lXNrruR1W5Xi50Ha7b7v9YDdI0xpoWTiWKIejuNB6/H8Je75p2fSKw5XhmKl+c0ivknVX5Vc5/HNYm+JcK9fiSL8YBuYQsiG/OLCaosAT4N2vvUNEagHd493NoVqUo1KrVm9ujlBMykE/cpbPivzOChOvDaeZfw/KGYn6PDWx5fC9JpyWVYeR49sN96nKH6muR45vHX4xL8KsQn1d1DOm97fK6zrk0pIXfT3KNa5DPK73e9/65r7v8ElynX71039gnuND5pU1MP/kO3+x/4dXpOUOwkWvnt5Tl9UUZdD+V8flGipK8orA40bBiyk4Wk3ce1Hk357ySP4+ozqZqZOs/6ghGV3LM/EfvTiNev/oeP17Rtkzm7euo77fzXWK6p6IeGlt/TjPWpmVlaurUuoPGu+z9x88bEJ8mPczbaWNpVQ/oyHId+R43nnQJj7u+l1YjqlKchA3o7JsH3lr5oBSHdJS6qJUpuJ1LeVzC5IHUSknzG0XN/9HkstygYL5qcJxPlPded0OtVnI8bvFeg38HmnVfw6J23bUzfia1PYHet8Fb5/OewZLXTuayWSpq0fH1k3Okd/xbibVD7t5dlMdK/NbnO+p7Mt5M5lmlL5ePy1S26/1+iw690Wt8SHrq/sqer8sx1tsx7rvpd08foe6v5Fns4iBbf4f1/eZGhSH9jwAAAAAAMCHxytrYOYB/KY6xU3sk9ib7mi7o8t+7iQHuTv/1gaPFOz4MFW9J40Foxgw3hTVlNJ0/BgvL8/oefYcesfMGjG9c0mpXQzirmXp45v8hcd+dnrvrPF6ZNyV5zuaJpp4YRE9gONBGT+//FyfMCSAh8H7f/MVIhLPUWOgDc39/Nn3v7rvO9JUxtfcbJZYUk04lzz9PuY0vbZm9E7Ejcr0f7XAWEyxfefzuXE+ok3mbTxN3fQGcbV9Yw4loy3JfsO6fVwWURm1AJsVT6vcK1mw2nZT12elNe979xtvqQyL97xpq8Q58jYsEggAAOAx8/4P3rL72crBEWjJC/ZyaKE8Boyl/xRi+90uymSihazxRisgyv0eoVoO1TFEKy20Hiv/ra8dzhpas5otLrVD5fm/+/d0+8EehuODD/awGZe40fPne0fio//BR/aEzxKFZf9/Wg9H5e1FOISOem4x90EsZ5q8rqMxqcyjCpUh+qpZQchjaXlthJqZVl0W0c+Phf8Wobz8+O/+EYHHj3zGWGnLyuTb29vqN6NVqpxO58nj5vLM2rYWZlaYN2unaOwr+qdT3ki85233xnBWeBttXxvVhfFtcPVvzp9nb7jOdXUvc2gj5xr2VNkyH30+vF0rmHl/nmWiyin52orvMpZt29iSx6tvh3l8UaUBAAAAAAAAAAAAAAAAvBReWQWzjpmlPRhSkTarAtbCtqEStknPXhCiGC9q39y009GUUs9LZnmvRl4lZhSWYwZvauRoKqh3/jPewms9iV7dR78TtXVyw2k4CmcvLMUwpEoSIVzUbZwNS+LlnV8VZxYAeHhsz/l5yluO31b4B614L9vLtGhW4pZEHGKlfReLZz+q1XD1dGeLIlKJJfyFmB1AZIe/yEexgphCURxpFT/XNIT2+Y9lXzm3+sheWyLfj9JeFV3UsScf3yoi5Lmqdy3572r0FOEx5etXouZI9cGr7wl/rLz7zueJiGgxVp9m5Vwzw+gQYbByLsaiyuDFmpL+SHCenX7CSPkyE+7iTLitmZlQo1k1PdXO6Nt75hpY565DkXEf1LpeWnkzM1todqaYpTRiRud1prypb8uZPsTMTDPdx+ZF/1Sb+v5ffzWnyYJT7xofH4Bf/8yfjusIXknybD3xDm4X/UymomjLfYIl7yuzu8rzo9XK8rdsb4KMVUxHqKLcTuxjw2VZKBntSjMmuAnE/bsl7W3Os8Rtz0fokjge8/H39paeH+Pj5WbP42Z9lhcXzu0UBeK1B3P3LyUqs+G4nWzDjOXj9418xZpz6ZFSkp2xuiJU+p8xpqJ65mRLCQW5pXI9weNG232sfdqGwNsZrWJtv5kqTl3ebn/rdD9A5+vNDJ8OSWksgElU2qVRSFSvj6Svj1bxmlVxwmbxN7fM9rb7XrpsXUdPSaxDcV07S1+fh97uxVRmpbK+h/qZ0/XmZy1GfT2OegujI89c4b4Nh0CLl1d3/IYWGQAAAAAAAAAAAAAAAMBVPHgF87vf/jIRCU/FVq/62MRc1nE7qwWu1AqU2bN0pGnUN3VdtBBspKqZ4WzsGg8z1tXJ+MBeHc5sH6lnR3l6Sh9LfdQou5z4QMzs+XrqY+mh1/VZQ+3xcgRk7vmM6iQVFUU0Wqsk+IGVqy/vdVPnp9xKRUU6/9yCDxf2ahaHca2yl/Te/5hjBx/b8uJgl6aNtVX3h4dZFRGNGQa9ennefkluf9k7LOL1eTMBrDwt7/WM17/dtlBRKtfqRVkmxzzPXv1FplHvnlqQ8SIVrSzi4Xw672t6oAs7AkVS761onPMMAlYl83vISTndVp6RrE7TjfxBT+nbSzNi9D27Ji+LGUXu7LfVan8Yb7Vyibd4tNWG6bbTiwVoqZ5nVdxn6LXNnhJwZnZcr65ajXRG8Jjvuy4/qd97Zao66X475mc9bhZjbJBSahbupI0oHf2oJa9LshPSUr6hVZtT9x1kWTlW80a0xfr9jpe2jV6WhZ4t9bsbQ6RArOI7FG0rES1HrM9lz+fm+BtpoV86RvGX2/r7QER5Ucxnr71G4eZoe7jvuB5rPBDRhfsvS6Al8nt6jMWonHfd31TjLVGuHv947QEvjlzyjaWd4UwCUVT9qyBim8ZQ3yfw6vN3X/36/p9Uf1esb+7sGHo0Hinjgnrc49kDvFk63rdQ90u8ta5mz0erbBmvj+TN4NKz/3u4Y8GlPidvweBctlqzRtdRn5PXXzozY24mnS6f661V3qOZ+KPyy6zWtr+fryHVdfj7d96u6rCue14f+/3PTJX9YfDgDcwAAAAAAAAAAAAR0c//5i/z/y2HMS8iHGMk2pRxKobs0A1qUT6KqRhMg1hYTxmdlxCKk14sAiydfkS1gUQuGhi1kVYYhRb22i9FJLId9bwJu8E5hUA3r9ULXUnn4vra/vejH32NtmXfv104RIaIiMaRKpYiGsjGIQpEeqWxRJTUgoMcxoNkaiEKyCEGRVatMW3JR9ch2rJ3ff+zivM9woDEbaP3j4Wsc/iCY6G3X/kdLHAMAAAvk4dvYFaxrHTcNo82XuiMl0N7FdTKlCqftiyRk/IueV4szxMyqwiyPPW8eupsvL9ZL9Csl2amvrpsZhwfqZyLV+/RNZyNS9jzBGqPHMfPKcoadV+N+s/UvVevUiGVd+MZ3etalEN22T0VKLhf3nv7K9Xvth0QsYOPAYOMa8zKGyH9OP7sfy+XS6NEa9uBQByrNStJEsc3XvJAZclpjvakBCwsdVTtXYxx2PYtyyJiVamBFpV9+d0Re/d8Spri8O69+3W85FSprLVXvyhrSl58qjf5eN1eWW3tRnWbkAeYotxNKXny90NcF1btlPdUHnNcB/aCH/mwt5uI6Mdff6sqw1Rnr+q3kKZ/4tMYpL1MrG/7/h9Sv1vDiv7+zHxPRt+eM9//mbR3UU738vfy8RRDvfx1WqmsOdu3s9L2+ipePjPXbRQD2Yojf35WVVs3T01lbeOZWE3/WsaP529ftFVfS7SU5f56BAAAAB44SrmsvyPV2GagJD5dtFOmpwAeKYZH5Yy+v14/Y1YBrRXXI7vIbP2tY2f7jnlyAo9zbmuFsK4r/9b7R7aqszaeWRvY7OzcMzP7Gnsd94PyEE7b87SD7+Hx8A3MAADwyMh9omQvVpBSojU7DupFKmaNK960omraKHc+DgMnh3pJKbofU8t4YU2r9hZi4HpZRnB5Ht6Cob1z0r9718qaZtfrEJxZoKs3hW/m/lnT60v4jOOeCd+CXshP1tU7t1KtVO6VCpcTKJ3qaAIAAAAvG/5KxS1lJxxPbU4pFcduLOl4cVV2yJUQYyX8VYnA0Ib9irE4h1dOGFP278kp4bk/IwxKlmHD6l883z44KsFT0o/815JHCY1Xjr35yGv5MO4jPONQGRRznyLlvmYyw6xZfRh9juy4t8JnVHmJ4/SCnHU6uVDyYYzjsra2fxSoLJbL54CeCwAAfDg8fAOziu2ilTnNlKVs7i/KUu5kbNvtvsdR/Q09UMrrsoiOh6xjSinHr9IxW7w4wbOD+FEcoGVZmlVTmVkP1TXpPLWk52XzPJJnr0OvvrNxefRxnqfvRsRZzh2cY99zfZ66bKcKo7qVcyhTx5pjczwf3nFsTqqDfNSZk60cx4znx/E5rbbRDwAAwP2RVZsc+1wMnPN6E4efKGUrgFJviDnHHLdTz1LRyO/OSA1S8m7zGvWdTMPBFQ6L0TEv2glinc+obydnDzCz/SGr7F5dZuo7OmaUxlJX62N6feiRgtx6DlWI2mwk5Fk00TjPuPTVQ2x8evedz5v1ICL65B/8mbsPAADAy6Nrm7nDt07ixQ324O87C2b0901//y1bjVXfUZ9tVmE9e5yuT68v5YmGPFvaqK4eI1ucJcS5Bk8trpXSI2W1xnsWe/a7pv+jTJ4sFztr1/owefgGZgAAeGTE57Uatf0gEV3IDhcj0/aMAW6HwlhgIeetHXryMGNh1VEnTKbxPrLeeXjpe8rjmTR6u7WvV59eGdbvGUPQjBGp6ZyI69OE7MlxFIgi6WfL6uwV9dW+7di3SqUzAAAA8DCw+jkxpuJgEwIkVrfmxYFjEqG9Sn5ERCGlopbtrEZ5uVzKTDEdJq9iyc4/+a1mw4U0WJj9tjx7iRfjuzn+prLr6Iqt6yL6E/uJ3d5eaH3G+496hqUYULhOTv3N/hWH9TIWTZaLV8l6yLxCCM2inZZKfD+clcnHdRIzsy7HAoqvvfZaYxB6FYwwAADwGHkwBuYfvvMlIiqLGWTjQlZj7n+jM8gvHxLV2UixGcT7niRbdVs6HcSVEKntj9hohVHNWQ8PY8XvGXlP7qoG6hlfvLxG5+fdzzOGG6/es+fD19IzmkkPU6PGWdRzO3neug4z9Wzy2Oq6pMW+tvn3oQLiqX2kTpc75T/5zv5OxlDyeOOznxvWEwAAwBitQub4sURE8cJ9oGIkISrK5tJHEg2441RaS6D+Pe+JadA9PLXLfSuJR/2KGQeX/r7fRQFiqZKZa5xtmjOKb6s/5/X1rjnXniJstv49J157LNdZrqyu+vM8OS/fhjbecjpGNdnwpQvKk8za+sAg9XB5/wdv5f9nw2Qs47uFZ2/E29x2cru6hIUWHY5sae+1nHGY+/hizYRsEj36yAuFPOvkVsQRXY+68L5FhKhYV/99rdalkE9uqp/VyOKDJREbb3PTFGL2CudIbCHRxu8XcVlrrpM876CnDVD7HiZh4M3nIJzT7avVtgMylEYYvXZcLo/B5NU5zvv2g+e0HAv+ZcfCMRvzF9//S7FOIDsO2vN6/VN/PKgIuE9+/LVvEZG4DyouPzOjTs6Ci0EbfraP4tkmuH0YffcbO5Lqu3jrFnh9K09MM7LvjPKd6R95dYjOJfX6aVqsFG7Y6cUCmcMZeDjvtMNo1I86a+Maqce9e+j91cz1l9RMdNKKZg6PVF/Ln769v0N6/ZxPfOafmmW+TB6MgRkAAB4i/++Xv0xEtirCm2Ij92mDUEqJbj+4HGl8g8WW6sGz5UQqH5/6A7Ysi6k41pRORTtdy5vSLDsLo06SxOtgeen1Pp0mxugaumZUwzPppaLGOs7rYFTPSNP7KtdV35kZg10uSy6Ay/efk4dyXfgJ4/jK+r6mtSxW2HSWt6LgevebX6mPy0WlPGjmAesnPv2fN/UGAAAAAAAAAPB4eXAGZu0dCkF5ndm7y3Fvttow4C12ZOVx1tPjpSvGBFlWnfas18zziMwogUfq2bt5Wfr7PbWQ5loPkkzvnd+1KhR+pkbX3LrGS1IGJqfezMggd0ah5eGp+tkKVepgl/3GP0UcQgAAeFEU5U6tTiaianEp+VcvRxGS4UBSawTEzvdkpMKwHE2zay70+j13UYv2vtFeP6D3LfXUSXrWmyx3FLPRcjaOwgX1zsO7XlY5d5mldo26elTHGXiB2SDXmlCq/TxT7Ni9Ja36obI4Cv/Uyiy1Xdb7vhX44P7w7kweA+bY9IHS8X/Zp2+d8kE8E7oNXag8Ke34Rs4Uyc8Rz0LZEm3ptq67McvAG6vl/4sT1k0Xq/29Z1i3W2ZZVVsj3pkX/A5UbYReP4Zkm7Ll9O2s4/J/Oas5JRXyRLYFvTE53vsPhbPfCyt98+1Wv9fBvW3bfj0+JrWfqv2eAlijw+F4+WpG6c9+83U9dZ8kLx5u9Jeiuv4Li64uLJiaU/T2xEZERclc3mP+fdxTJx+P0b3Uz4C+tyMl9Shf77d1rH4O3WPzjJz9Z6Qyg+ah8OAMzAAA8KDIX9zSMc+dGuJNrWHC6uTvlAUbU/I/fEGNMzgm4D63j8r/yQ4dNOtQs0ipVbXqDpLs+KekOxbWx7X+G0KZuh+o5OnWyXQycXl6ytQi0uQjjvPxFei931xWyS9U/5d/y7mmPGWv5FU6Btk1maeX2p1Wfd7etvwc5nNdGpW57iytaS3HB/ueSfg0qmuVFwa93sAEAAAAAAAAAODV5cEYmLPRQhljYqqneWclg7FQlSQbVSh0lCb8Nzj767pRY8TRiteSjx7Ue0YeT2VzraJixiN4Vi3s5d2bvn0tnhKm56nShrzZusx6T8+cd/6r9lsxL8/WwZuGvyz1axxTvaqttlHNKNQAAAAAAAD4sKn6p6n0+XnMyLM2LpdYYiRncXNsHNRExdm6uProPrJfTmINHx2fNYTgzO7Q48OiPrNUzdY4qBfyS6bvzXa9D2bzstTc9pgqB/eiRaiZd0SYNs4jBhGOTuURiCKNnffg5dKM2fWCQBNom8HsjOtrlcOzM99njx/FE/bK8exVmmE9DnFIbkeN6+PVTdtePJW2d2881XcTxk8dxxHYeX9uKa5UxHszynS6We6jXU067v4gVncUC9o+FB6MgRkAAB4yFzUNiKgMUKxOS/4o8SKMx/but0dOW1xvqjzrjxZ/GLVzgDvbKauCex2jlD9KR1Uvh9OMAj1bntWJj3VlolBd5856dqyUslITmqU+JoRQVmBX+8zp2KlOs3dHg6z+QIGsZN/UfsRb11DZpi9fT0lcTU2N9bPRU6nHRgk9DgVkhuzhjpvR17A65LmDGNSq9sGP313k82Vgh6mm5+AFV8vqNnLKc922nHIeD/bXUQjGDtQRnhP/TF69d5+ZGUxeUwf93l4TU14Piqypil4IEet8vEGPNl5Zx5wJUTYaPM7sO/OManoD/nZa6Onsh+Va237ynS+ax/wGFjl+Kfz8b/6ya/LN+8RMq3T0XeIl5ZlR5fsemnUq5O3Xz+CSyn55VH5GcwVCTiDfaX735Xs4Y/TdaMt55x6d2SaU8HbWu2i1A+37FXMpVXtgvGRnvxHT77SRjSUUaqbxO8Zx3Q7JWXgcBiUv/hlKHssV3zwAAAA+D8bAHMSHg0h8INTApawyqgwqomOw7+eMy4e19fQMPFGpTjZS155hpLy+1jtlGSS8Mr1OwFllc68srwxvoDUaZMp8R2Xcl4eyDXHQnp+lWJhh9lr3VNPN85ifHT5G10k9Y8o4+BSMRP/3n3+BiPr3idsSNhDebvU9JmrjofI1X4ThMqb2uRq9i3JAYqUNzgwOeTYjj7tEP+PrujaGE21MX5bFNa706j8TC7U6J8cII8+rtPu1gW1Zlno19mrvXPlnDFZnDCyWCmlRv7dUQnx4Xv2eIijGSGW+Tqj+5llAS2rasZJn60zI7cda7uPlUr8bP33nq+X3omb6hLqOr//ef6YvDQAAAAAAeMX5f776NSIqcfZHylFvfH1N/1rPXNCjj9Fscz3G0ccxZ9ZkICoOaBYunZ1J7NXLs6nk66HWMQgDp7eVt+XAsbYzvL9xFA3O+ebmxkzXxEi+6BkOdn1HjnGvHL5Xo3qctQVWY1hHyZ7rSHUd9XHZ3nOk/+k7b1d158kcr//+Z0/V8S48GAMzAAAAAMBTIWX13f47iM4lq5vjpgY+SoRsDQRi7rcqh0vHUcnoGN05J1HOSOGr91f1dwY+FloccMa57w1GrbrlbTzbRA98eQCzlPK8svVCMGfqYjmdRgNfeYzOjwevvYVfXEe1Mxiz6qLvT0+RrQebeTDK11YOuoIK/0V1+C9ewOlGnF8eOPNCaLyBB5FkDDZ5nTbH8HGNiATcjeZe0NI8qzHGSrlMdNwrXvxRGDz0wnhrCPlZ27K6eSBKEcIlIqKbsORZPr2ZC5aC2HIMS2VuT1RS/84PeD4H27Ci2z07/2gtGjuoh8Rr96y6W2twxKjbIWEUy9++JFTs5WwC3/fIRjR+hmITIiPElBcPuzJCCgAAAIcHY2DOHyK1mrrVySASH/MsL6ZjO3fuy1TH0vk954nSv3nKKtPz9FiDmN6xI4+eV5Y1KBh5+UZl6DqeYXaqpKfU9rxo1nnrfTr+j1f/+1Buj/JutuvnYTl3bWOMrsdyOPB0BorFO/f0elfeNVuSeN+PtNlLeUR6yquWU2lvFh74HoOdLQRX8bp3om2PfDWlsZnSmWW6zWq+Pe+z90xaz7AcHOnp4tYAqueB9mYCSCW0p2aWA7KZQb7XTsQY86gkx/fvGIl6ZZ1RT/TytMpw27uqGvWqwfyN613Hbdum46MRWR55/5yDDItxU+dV6pjKj1JiUy4AAAAAAHhc5PXR1WzCHHNZdQW9sYfpaM7jM5WVGC/JY3UByVgAACAASURBVC7K5uDbg5RTI9R97OLw3//eLHv/+nbbnbqsvtWMlMc6f8/x6zm0sriAxxScX3VW43HVjONf22/0TG59H/U10c5mvV3XUSumM8dsSi26ECP16nfzVzm3WWnvrYmmn0c9tp0dV/bGzZzns2fPqrzzKbMkOb9b9X4mP2cTjsP75sEYmAEAAAAAngraIC+N8XoAkTvrznRAorbjPaMsHk1f7DkCRlMSZXm6nJmFkO/D6e3VtXdeMw5kb0A443Q74+QeDQBlnfQ91GpqK6+pQTzZ92U0CO7VeybtLLVDVe+ryzHDIG112hTsASV4eTTP1L5x/8HOzS1mi8pSbl5jbAohZEW7DA3F5HBoVWHHf4WzVD85iyEkkAIC+W5ZMxv0uSZq11Ow0vXbCXlclklQNJ5lU7jTCXTec7yfFTYRzYcFyOtY8HFO8uaZYSXzknLYOlnfItroFg8AAOAkD8bAPD2dsVE09/NJKVC7kFON2+EddJpnFHU9tdiZvEcf8TPKu9my9P6ZjoSr4B3Um/d7HiCNHEifvdbe/lE+1v5h3Bzvmg06NNdcc08hqY9sp1hzer6mj7+3ldXBSagwiWPe1uefnzUemMTYXPO82FzkpDGv19U8T4GIrz5f8/b9tQYBnIaI75FWmlrPyZnBsXze+Dj2OlvPIqfhqdjyfdDeamtmiI6xJfPVyltuG2R9POPIJXtryzRIvuUyTID3bvcGYcxqTEVd+VnhsWyMtMXa6y6PGbbNSQyq2uGtqKPt6ZextN02SsyOaPalVtFR0vA5JlpYlcx/uZFZicLNcd6cz/L42xcAAABjfvb9r9qzeqjuW0vDJ88i27bSB8i9VzHDi0MmLLQ2YVBklyPob7yYPdYz9MYQREiWfnrLmNobi80YmkdjnMrh2OnbyP4X7zb7jWLswMPq3vhInoM9NuWQN+XY9lnwz5+o9OUTJSprOh7/OUR+S1pKP4yfK9HHZOfDL77/VjO7dDP6R69/6o+bOgEAACg8GAMzAAAAAMBj491vfZ6IiNKldqTSxp4IViUXg4AejnvT8qwBveeg7IXQGTmHLcOEniLZq5s+1lsg5b5Dp5xRC9OgDmeUudY+Pf1xRiE7uj6Wgnm0IKh1/0eq7ZnpnHp77zya89pqZy8R5amrWvSRF7yx6qyPIbUIjjqvPb+mtkceS/UXAADAdTTfPSf4tRfyzXSITH7fdZ9Fhxzwvq/eN7qIadoF0mUdR7PKdHpX4ORQ4pP7Ipn7wsvTO1cvTKQWDc3WdbaPOMqvEYjpeupnaZubaeHlr7dbv70FDb3nQfe94/A98NfheFG8dAPzu9/+cvV7OA0z1o2CHgDosCKs5Cs3aRmqxNwXu3mY7BfXyn/UWc9lGKo2K513faxpp/p8zg6Wzip/rX2zquhGAajihFofFN6u93kxj0b17w2+rHSWN37UKDD5fms1oLoVZ9TybiOZY0jpwV9dl0bZcP/fpZfGv/5Xf9FsC+o92Bd8yXuPfeV3aWp4X3sdtQpcX7PuNEBTpVEjn+85g0t9HjLbnjHDe7+I5hZu0mmtDoVuj/gdX9fVbRdk7Hy9UJXVXsdyafe/xkPshQKwyp+ZCWJts357bbulBLcMgjMd57ZcTksUwsCYJVTSuh1KKTVxzywDkW5j2GSTUsrWm1xH1d7+7Htfyf9fjzb8Y//xf0oAAACeJua3deNYlyVcUAmfcIwdiPKHzPrG5v5yKuvVRP3dFUXL49q+h+inGf0S2dfgfgzH1LQdg853enI86LGn9/t/sh7aESPTVXXmz7lQEMt8vDx0ef06132o/Dem3NfjRRaJSv8lj21EkU3/SyiTK8NY59Let/MTAAAeK1AwAwAeDdGYzmZ1zqc8oarjnGRYCqdfbOXbGA9j22HWnt8QQuNhP0NKQeRtD06iEepDdsI9g6g0RmujsdyujdbsAJLnrD3flmHVUx7sgyF/kUBdx54SoLfPc5Ztru6tjgLRUwAOnZ9BGM0dA7GlRqzPxw5xkQdo8eIqTlOwn+mW1jC+1yPkUR8/DkGFpdnvta10eDSwQYSnXmfHeZ3Mupcar82w0ni/La5xSHtKEOuZHymXZ/LX23t1m6m/hgXlnLuO5mIZhbxy7qtO3vt3rcNLb9exYb1rXk+j979b1rEWM+cTqD53ncZyjIbjJuaUN3XdZGihfIy60RxWyVsYFQAAwBz8ndBjmtHMl9E3lqgVAvZEaTN5j9J7deT0PM4Zjd+8MaDXF3f74YPzuM9vl3dNPKHkSL09ytcbg54VTepyvGvrCR9LzPdjvHtPKrxKJHXc9xlBldzeCOG4z7TY/bmfffc7RES03NTn+iuf+vS9nJPkpRuYP/nmn5jbf/T2F4mI6HKswFlezjouLAftzx1PZ5V7qeKcVeo2D3GTzi7L6uS29bB/ey8Ke7n1YhHeQyaNL9cMLi1mX+plWSrloZfGymNmMCOPl8dl1ZszQPLis3p58outG0edvzXQ0oyeueZaLnbD1stDN4bulOOlP0gPyjjUyBgBAAAAAAC4R6xZOjHGMgspO+YDES/GyFOWY8pxluUKcJbjPETVDw4pz7wJahwpJ+T0RAMxCiNSnqreHhuI6CYca0cc+25jGatlh0YqimA6lLlLCM04txI3N7N4y/XktSjCUuZ0yXGCZTRrjCuGEEFSxhGU/xZ7TNnXCi0iJWqnyOvxt0QKEvL9CGWc1sxg3oq94NnNMY7bSpq8LsR61HMJFIN/rnA0AQDAHFAwAwAeDSGERp+kHUU9dXFIHSdQKp3Q1vnU4nbKY8o9cB4bzUwfnCnL2tZ6l9lpRaRjTJYBFxFl9ZgeOBVnUbk0tZNjHwDUg5d1fVbVay+zdorIhf30In+9a6SnuEqH1owKcuZaa4WBpfLOaYrsmE/RLcval8/LSGoNeHtqQE9lKs6MShw5VY/sD5c7j59ySux61CmvpCiuS35GuB5HkqW+50TFgfzY4FNMl/o3L47Iz+8iYr3OtDEz+4nuNjDuHevF2GPOtGvWe9oLZzPLmXNvVEXq2fWlBP3yrlFce/TyatohNUOkF7f7mmvcMzzNHqvr0TfsjMvxRBg6JJhFMfg98hkVL5if/+AtIhLPlpHGVqzvLCHQdlExUmMboiKE4MZxnSkv561FOU6YCEsV6SkRz5DzTeLcxCLRnqJvT1CfQ4qizyO+uZbh1Hq+LTHTzCwTa+F1K10l1An1uxlCcQ5wDjOzRnrl5rLyreNnshjOqapTkMkoBNHqoD24Cv1ueCpjjSemsvIeHevVZVQHD52vtw7FqB4jMR13ifOT3bkWd6GXn15gXLfrYVVtHu/Q6x5E7kcd77gaSwT17uVxHWe41BdDjtt75zLbT5mZ8UUkZ7rVzrURvfJZVDtan0M/L7EMJuq68/PHv5vnft+jZx/fJw/GwOypZXXn4YzRZbS/6eDm8bNX5vzDOdtpn33gR1MCesaSs5yZwqrLvstA0Dp+pmPDeB+K0YDWUjv00luef+/j4XUMm/0nbtnofnj1Fnvq7fZuAAAAAAAAAAAAAACGPBgDM8ckZLLxbGDsGhsF22M8w9uiF16b9HpYBsyzRl7PS+HVtXfeM4q9XtmeR6Nn7J410o7qous0cw9GaTyDrGf09eIG9QzMnje0USQ6+xPNX6+zz5ZnYC7KnmDtftD86//9S/aOlBo1a/dZ41iowtvnOQd6WM9g793VSqlcHSP8itU2jFQyIRS1D4d+0R52ucjeGceQrBcfp1cDllMYtVqhPj4cx9dp94UY9//rmGY5NlWgvIKyXgzWcnr12gnv/C2FUk/9d+bZmaqHuMbe+Vhtg6VC6q3w7DvEyqKX5bjQHs9zl9VikykSLUqVnJY67BOR/Gbt297/3lv775t9+7Nnz+g//I/+GQEAAHg14Ha919eQYxY5EyorY4XINH8n8jcltN/31C7sPqpf820Pbb6ynsxafTup/J1QjKQq733bRrEoZ0X88XWiz5/rFuR5lW++1Y+ZVTDrPCSeYtMqUx8Xk1+WhalmFn2Wssnof+muVSrq97T46aBavjt6oehRfOQzdg9vHQBvjDQal+h8vO2juuu+OI9j9LlvpNsgVb6q/1mR5QitPjavsVoLwauDt6ZOzkdt13GGPbtKb3xGRCVk0h3tbtemv498vGs3im/tXpvjG5kjLq36ed+3PwkFMwAAAADAq8hPv707vfL6EWKGt57+VgY5HCPnGHjpgJrUOr6jnoEi03KnU00v19MqLftHMzWzWURkbHDwBmvy+DMzonr5jfBmD1mDJM+JpBf5m/G+XuNcmjXmjJgVRUiaEBJqOyP3jwQA5uJ7k7O5qrya8/ACYPnk6bMcg3dpn6moB9Cd8wAAAAAAAD4PzsDMnUEeVOnu3aynK+fX6WjruC1aRerl5XnVPJXqDJ7Xa5RXz7s2Ug17g4SRgteq8zWDoZk6MaM4Rda20eB05JnUdbRUkV6cpdnrkfPUCvDO4dc8633YE3bkc2UuLxPPaGCpTBpiyu86p+yp8hn5bHqD5LPe5F76a/K26nimTOkVPeOt1V5QqwwrHihvYwOcXNz05saOvZxjM7/2rFsnfR6WR7inrpZpLeTipr22dmaRUFd5HFo1lvdcyvrOfIMspZH3e3x8rY5PIZrp9t9FfUZUh2lrnhv2wvcX5AYAAPBAsNRVM30ddgLEbaOVVF86Ul6oj3MKiXJcT56FFlLpw4bOt8Uat+R+XbRngq7qe22dX0qJ2pVA2jLMmUPOMb24nL2ZTFVMf+U8SaF/T/Rxuh762vXWvZAz6qrrtLTn4yk3dZ20EzBQO26S9ynoLtEiZofxvmAoSZeQb8zIRvDU+fE33iaiMmtSzyxkRgrlkS3lLg7ns3mNxmGe7cazBZQJff1+PeOtc3Gt85PfkRkHv8Y7xlPd6nfeU912ncydchk9PGDb3khx7eGVP3Lgzz4jd1Gde0rnpBexVXXSdcnxsF+gE/3BGZgBAAAAAF4luh1ztYuH/c1Ay3J0eXl2Ot09J4Te7w1gtJMlK7M7+eqFOa0yRw5t2eE96ySu8uErxwMqLSiYUA03znbjbpxxLo8GOj3ltzdIsQxL3kCoZ6DyuEYVfY0iu3fubp2sfFS4vRT0QKoe7O751fku8Ri85cWFulV/lPziB1+rfluGRsvJGkIQBtbjGd1SWQBW5VcZC/PeQGnjHI58q8XrDIMk1yO278OI3sBfnokWJGy3F7MeHOopl28YeuXmfGzqv5czxp96BoBsI5QTm2wDexMyj+y2ptd2WXWx7kXv9lRtlFPmzLF8nLajhBRFiLqQ/2xUh3NYUlsPAAAAfR6MgdnzaLF3MmbPdd8TkAcR/HdpY2nltM6AYKTYHQ0KrvGueYpMzWgQI1V5Om+P2fP3jpOdTC+20shDNVtXq/zZAdI1ZYwYKau9weHZutzlmdKjsOuVzg+Hnodx+PyGvtfOe2Y9tYpVHyLfW1vXsa6/pa71yuzVeR8o2LGbegpa+XfUJkkViz5XaYzyypdxFnX567rmgY5l2PLOW2+3ytfbLe7zHTllcMqDu7uV31Ne955nnWY0QORjm+eYDgNjIooqPloK9fO0plUsrGs/85iqDgAAAADwMNEOYR3y6+x41+vHnp2hJ/OctUXo42brzgplHr+EZe8Lb0rFy/TGiWfKLf3rvGH/7aQ7w+gYPdb0hAeek9y7z6Nr740/s2OY6+Gt8XYS31Zm12cUa9zOq74mOg8vuoH3vOi1tsr+Fz+2ejAGZgAAsPjkL/8yffO//m+I45SGEChe7MUSa5kJ/20+sUTOogU9FUY6JDXSQOwhja96AbwqnfIBzHQmrPJ7xtMZx5gVh9MzMMuPqDYA1iqe2unE0+Z4sQuZnhWPnOa1116jbeN7zIsSUlXHjeypmFZ9JL1pnpZToWfEnnLCqc5HTz3I10Geh6f47HVAe+fYe3Z6aXvOwpTs651SalSfvPhEdkYsxYGwspDIuJ8//9uv7mnKg0BERL/8Ci3+V9qp41rq/db7F/02SqYjmu8sWs/gaIHaXnvhDSB6sXi9Osv9swuQTKyp1R1UeG2cWp9SOELOMRp49PDqZg2+dP7eojoz6kidl+U07OXX2y7pDvTz6+J9u8bnQbqdNOrUlBx0G9lW7bFjDWC73xL5rTDCIeRvkrEgr1aSLhTo9oiLlEVGKRVDQef1uR+BT7svxljegaMet7e3JUa+WPxWG1yqMFCyevmRVAYRdR66fyjbAq9f0PyfLxp/Y1PK7VkUb0DQ36YQqnpxvpt+r0S9rX6V2beizcxDpjtzP/PXk+sm89BlhIU4XFeS/cVc3JHHJZa6rOefLQAAeIq8dAPzj9/ZF8LJH8ijzc8xew6yFT/4Six5fPFc1FQGAv19ueIDZmEZIHpqr/ugZyy6q3emN+CTvy0Dh1f2aPvIGGcdPzKAzRpkRga4Xsetp2CdSed58JrpvNVhalpXr7dNpYO+ZI9m3xvIb9GP3v583s77Pvnm57plAQAAAAAAAAB4GmgRycjh7NkSRupdy0HrKYNHAgtvv2eD0OVs2XFz2B1UenbCPn/+vPrdEzfJ/SO7RFTj+WXS/jOr3O7VUcPnNopDzc8HPy+zSmVPODS0t+mwP064Ji+/oV2rI76R+c+Ie9gGxtdS28SyMEelGwk3qLFjvXivORTMAICHTdoX+QoilqJ2LFVxQpNS1yl1pXRmlU6IKIz/50x/kWocuSidJMaieuCPqUyjnRZ533acz7pmp4IOESGVwxrrA9xTl3mdJuvDZ4XcGCmpicoHsKdG1PXets3tyCUxtcf7qJpq8U7nqNeBmlFFd9XETrk3NzfNvdWdCXkNPCdcSqlRG1rP2mgamqx/b8EenXb/vypXLN6j+3a5Fvm6lueQZybwAo/5Pd9KGI78n/XDWfnvH773lb34cHSmj/vBAwg6lPfx0iqL8/k4eaet3Ld8vZ3Od7ezf6PetexcPH76RzZYIWq8QRK3h1Ydvc52Tymd65AGA9PYXtFmIJYbcqqOJSptCi86la+XfnYNhblWIZbpmU2V8nnoQWiPXrgbjVaSuwvCTAgd9DdNtt+63jNK6dFAmgfKVZun0jTvQJ4NIfILdZ1auK0U+Sz1MXGr6zLQuTxKzGdEqVtTjOW9Mo7l67dSaJTL8lnNz6mYWbaoly+kRfbOcrqsQk35xT4tsrHaVtPYpRctiyk/txw/OcZEaa3btrSUOoX8TNn9iF4fxOpvWcaQWSFTboPFldVNqaVIJqKsdM4vn9HPkPfa6mNa18BqH5pt3rno885CtdD2w1LZn9/vlHiipPi2xObeAQAA6POhGZh7AwMrncfIQxC3raQ5W7cmz/6CMyM1aw/d+R6tnjpj2PGO1ZwZdHiMvH6zec9eO8u4NZv3aKBzbZ2sPEa//XvSdqjuqrT30INQPaijF1g2AAAAAAC4X3ihvl/51B92073/vbeI6Hw/fdu2bpxJ6aR3x0diDBiF8/gm1CpISmI8QXLMcigFj79binV4hxPnIzaYDrDGSLpGc7yYLvVCcVH8XyoMm/GSUbeUypzDnnPbCotl9eFH472Zcax5rCg/5k2+0jSlRBTGIZSsfdYzZjkFZpWX1XXKkUSO6xracfn7f/2VyolORPTxT/1xt4zHxLtvv01ERLcfqJnnh4OcF79sQkJt6tlSDit9nzblMGVWkc6LSaudnFpdW96X+twWdV9bEUsd1iW3Cfo3t2uHX1Y733U9Xnvttf2c9SzupX72cvuR/9T7db164iF5XL4++jyodSAzjdhFfwv4P9q+xWWoe6SdlVa7LPPN56rEY54oKJ8H2eV4ymNNmRW+53dZjnuZq3u0Q3yPDKcdUe2016I1XRf9LnmKZ83Q7nXoR97/3neqfOJR6Y//zqfNfGeAghkA8ErQG/zUHek6hMjscfux/qKg8iOr1dCsUpaNvRW7ktEfCd35uVwutD67qdJ4dZbbrMHBGeWu7JR5jjbdkSEyQryENuZeD92Bseo/o9RxleHGvt5zcabuPXrncblcmvP1YtoStZ05/r0si6uutgf7fifOum4zA1Edg5nVZ9bVC1Q/K1u8de9bEfGG7Iy+iyP0PuBONL/VrGAuhpC2np7atef1PuuslIwGFT2agZOB9170Fr0dGS6s53P2/TszY2Emrfce9d6nU9dYlaMHIKP6Eg1i+w+ucW+h11EeMh99T898a/X24AxiLXrtc3M/1DvHA/IzSlEAAAAAADDHh2Zgzh1cY8EH+bfpHDp9Xu/4lFpPtHfsKC+eou95OeT2UZ6MV6fR4imeN9pK79V3tk6jOizL0hg9ZgddIzXxjDd9dsA6692e3d8rc7R/5CkrtgphxOGXQU0rG3lcu1YMMuIlGQvgQcEMAAAAAPBqMGsot5yvxSBf8uLcsnMjpqwczmO6lMrCd7wtiOUxVb4U1uLUOgoIFBqF30Kr41DY68IOP+nUshwpktLnVXmKPm/lQOS+dyxOOFYrcy6Xy4W240qxGm1J5Zgo1c1rHXbMUmZSaldNsZzCkp7zSebRGy/mZ2ErISKySi5RWXxWHRfEZcrXUOQnZ0z2wqtY55q3URR5H/mm5DqPtm3LITwscoikEIhD6ITAYbpiud+HeGVJRaX9FJ1RWvVa2g2qfo8Yqer1diuushbn6GMZS+gj082OcUchqzZSdVaq2hw2SNXbU1qTevd0fb14x14Mao1nx+q1DfpazjrzR7GZz9rocr7q+bAWrZfHhVWpgdWpDm1jui1T7fjImW4JiLz7qp8HrVTOv1Wb21xL/ZyPRBT3YPKBghkA8EqRUmps5j1VL2WD+fGro3qytllOAf0RYCwnU69z1HPEjMLkWMj8Zj7e3kfNqtOMgljum3HQ6A5HLIny921R+ZjKUKeOZzmbn5e+d4/ltdbXvac89ZxTloJ5StnXqXfvvK2OvOfk2p+RXEp1fOnoyNzY0MEduZJWL16ydGJ6v0iyclxM5yaiHA825pigxrukHYedXtxZJ63+/8yxvbRXKXN1x99pw4jmByU97pJ2xml9ZvA5k8bNb2nbtlwHJ9v8ioiPoVeDnqjgmnPV9O6zV46u20ypM+1Tc5+zIew498W4BoZTHQAAAAAAnOelG5jj5ZaIKC8QwGFq3KmVHaPQnpA90HUcHE59eX6bp6838YDI7kxum56ey3v6AyLunG7bdloFO+vRGR1nGSR6RjCZTqef9YSllFqjgZFm5ry8sq38vPMwvdfGeXgeIb3f8nx63kJ9PiOlsne9OM5TUZaUe8iLS23b7hmNZN8frbwv5688YKS8eFHf96fppQcAAAAAeKpUTms21lvOJ6lobcRTxkJtQvVY+v++81OKCmSdFpVutqtq1Unu6/XVeXG6LdUxaHX9rBBdz4zQZyMl8UgZ59E7znPUs/o8x3alMr6TyjstnJB5Ba3MlvuNMq169OseqEj2fCd89ZxoB3+Vdzt2zOe6VA9eyXfsU3u8sKKbBQG8efI5HSlSNT0BhCf2GSl9R2V4dRs7Oev0OkauJzjS6lrdiFYL2Rv1mHG+VrXkeqxsTzhsDh1bUi+UIZFva/JENq0C3hYojOxK+lrqa9TG3VY2oOALe2T6fD5H8o3fg4E4Q9feOh9tS/KeZ/fdcV4h/W317Hmt8OfuTnYomAEArxQppUZxdKbj3evYW44LzbIs+QPGziutQJXqVCs+sS7fism8pfPqv15a72Nv1V8yo2w8o96Tx12ccnvXr6fSnokBa9W5d44jZ6GVp5WmmfJrhBayOlRcvnaEyWfPWwAihDa8jdcJ9Oo9cuD13pmUUu74JKUgZJZ1oSVxx1VfhyPNsjSdtBnVJAAAPDZ+/v2/JKLDcNb5ZnN7zIv9EbXTc/dZJkrhbsxkSik18eYXMo5NiW6CEnMYIoXyO4hvvfjubTlhlZfGEuXM9gc9gySJ8+Lcq+/k8ffm5obiYWS+HP21S4rtgls3a/5ey29lz9hqMdvv6/X1RvnGrf6uyoUc2ZZb5bfUhrFt2+iGz5EXKBPWj1WuZUJ1ft5sDdLpZPnimN71sfovxTjDRpWlhO3kGUok3jHZXw+2oQYAAMCHYGBuPxDnPpicftaju6523K5eGTyQ5wW89P4Zle7IqzRS8PYG87069Jj1zJ3NR27z7sessnmWmbp6KuORYnvmGo+u5axi28MaNIzO2VNoN0abRedTn2/rWY1X36f7wvNs6v0zRsM63bgDrp+LnmHLivmnjX/SeMpYRtSRKr5Xb2m8nPXM9/KTWPnNHNdT6cy059ZAdtSuWlQDPHeAWcrspRm1FbIM3f5YoS1G9ZHbrMGWV1YP63iZj3eNZR15xoWVZua76dWz9871wku8SHKdkprhwlXkAWpZUjqHz2ijad4vdxnsjo61nhN97EwbM2somqnLGcfWmUU+z3Cf13zGQNI79+zE4fzUKVpOu7v0C+/yzMzkMXufZd45lE6+CCrsDittRZZeu1ue6z2Pn/7VX8hSc5rf+xf/Y7d+AADwqvHe2+8QUflefPDBB9V+veYSk9vrw1nBmoLcRVJ2Dv2NHvUneuKWUX9Vc+bbYu5X4eHK92UszDhTr9l8rvnOjvZbMYNnytS4qm1nHDi6Nl5oQC3G8Ww+ud6DLpzuN+lZ7SP7T8o2lyMf43Lpc9dl6d9NDG+qj9N1ieq51M69FzGagoIZAAAAAAAAAF4hzgoiiiGEB5qhzAjjwWoUztcssy3H5FkpIVAW/fBgn0I26pdhq1RE+4PtymmtnHEh3Z9TSNaMyJjVk8qCe1aJUnHLxgOehfPs2bOcHy/yt67r9KJbTT3vQSFrqbpdh+/xNwsiLpswxBhOLTaIpPK7MeikVlQh9zNVNIrQ1tMSF4TjugciSipuhfncO/tz/vo6RfF+LOVd+7BFNwAA8JB56QbmxuMR67+NSsOZM+M27tpKb0yRnlGC9LbPdOh6H21UZQAAIABJREFU0+J7ed7ntN9rOyajVVI1M8pE7xhPTeR5tOS1H3WYzv4exfux6CkfZ/IaKWcsT+3IY9lee/te6P2MDpcQlnaa/YdB9Qw41zeE0E67M96ttmPZlqdPWXo9Pe8iz3pY17Iy+u3tbd7GZTcdaytEhBODq67jeFDZa996CuIzx+vZHr36yBbOaydlmUV5314z/S5Y11GX1VMia6wV6XsDpZ4aUCvZd+Xv/n8dakV76InKNeZ8+LmSSmhd12ux2mVPLbCnGQ+2ZhTMIceTP657KAqZIhrY6/Faeo2IiP7t//W1akX33/kv//thXe5CvjbcZ7lwxepptPJJ5PrxYLgZxPK1MVSVTRrFNQrTGWWxFwtuppz7qlM5aFiFIaM+R69uOo/ethlDytlyZR69vkVzD088M2dU4Wf7NTN5zZhrRmrrXl1yHtS2//oB8759v/7p/2KilgAA8OrjfSu1KtLtGzfK16COt/uto9lGVr9at9ne2kvXKpb18bmPPkjP5XtjnahmEHvXQK8Zpevp3QNdX2scYpVnXSfvmi5ZMOx9n7Ud5wgndKwjxeOfUmn+6zuM9mTswOI++XGNnBmcWvWblc65/83OQtvWousxUky79TbWGvCO9ca4nm3MU4V7Pax8/1Uc6nNWQBsomAEAAAAAAADgFUGuBzGzfgKRGLsf6VdHjZkH7VsxRDQDa2FQiKL41hizijJ8Y5EcJC9qMWprfLwLCmyDiyzfw3U2JSe/lHdXv4lU2JkjHvFys+Z9dxEDeMaDkYNFGo0sg4R1/j3jiFU/a0p2U6+Y8oMnjYpn79nsefeOZZZ9IxERbQPxShHE5S35yV+akIMAAABeqIH5x9/9EhHt1vr8ITvEPkW1sKmj9MdKezBUcqWAtugpVoj8RY+8WCbeb6uT4Hk+ZhUtZ5WjUl1531N4rlGLMCP11V06DN4CZZ6Hz6vTyAN0po7XKnw8b6KkOV/VcZtVvpU61L+Tuk7LA1Aw7+UXT5xWs/ZUdmdVT756XuZRH8O/g5gaynXatljt29XN9b1q4iOFVp3rnY+1rzdg6J1/b/bCjIogn8/+Y6+TU0f9f3l87xx7Su7etTrzDPfa8mvaZPlX3iNPNdC7R5b6Qx9vKbf1X+ta985Nn4dUQqzr+LjcbqlvfLhp0xa4rkQ8uuMq5pWil4XoyunH18BGhajvq/59hezWmkFwzfHeszrzXDEzi0Ge+Tbex7255h1mzhpORnjnfN/9LuZMW+m242qm4Ew5urxR2XL7TDllLHDeQGil02MGPWg4c39y2hdzS6+m9y0K0mBsfLODegYusTU0SiNhVlullI1pOV2w24LGmFfsdk3fiap7V54rW2nO+YlvmLoE8qd1284aNXvfySBmUZXJtoF4SVqpNsvKSvF86m+Ede2s+s6+i71+oIVlHLZmSPXqaV7LJRSlZzWT0K8Ld/EDicgso+9Xiauxb5MKyKTGVEHWQzxrsdxbomM8FKosjrL4dPZ83v/OX5R7fFT+E//k1Z7xoNXA4XD6xIta1yfbX+p2RMeqXZYyi5OoVZQyI5uNhafMHfWJNGfHz6SezZFdqFXh1um8a6Hz4+1a0eyd78gW5c3KlZQ+d6r+agXvrC3CW5x8dN/btkYd55TD6Xk2qC4vXyunvvo8vLBHOr/mWqb2+nj3Qb8j3vhtNBvYW/9F231yOx18W8As96GCBgAAAAAAAAAAAAAAAPAEeaEK5spTnj1cthelWOWVh/xsmYYnwVPieB4dTwl7xvM164HReLGHZqdOWdtGHjztxRkpfXuewGtVtB4znsyRx3/Ws+ipCHvn69VrpI72uEZdEwbKbf0WaYVyCLXnzLou9xkb/DytgloraZieR1B6S3ueyVH8r3VdG2+iFR/Li5m1bZvR7rXvqueBDiG490Oes1f/3jU6o6CUdTQqYpanj/fivNoq5f0vX89t27pxlfU5nFUu6WNmFBC9tkh7pOX5a4+6zE8rFXTcZnkdzrST8ny8b4alVNLXYX/WzsevL3WO5XFZ6m9Qng8QAhEvpqRX7d4PmC53ht96/WP0/f/1v8u/01aua36vD0V1YuXJpuogboVWQnrfGouZZ1uX4x1zlXqzU6deXc6W56kxido4hWc4o+Ie3Y+ZPsBMHe9yX86oxpv8T9xTfWzv23CXc24VoPN1M/fHepsSmjUzD4iIFj6GYzry5LArnreXgqw7t8NWf1z16WOMeVy23Rb1Ya//QeJ542+Y7Dec6RdLLEVwkg+BmtVl9Sk8Ja9WopnHmOWXPKznVSuOtxTzd+lGqPisdjhfuxf0SM2OP4ftn24v91gS+4/OTGHZL5hpH/Z+w9aks9rr2ZauaYuFhi6qPkqK4v4f/Y4lhGo9hyOTkk4sjJlrxZcmiDAkD7TZ8Hjj9dfpe//yX2blPfdH+X3k7lWevcUbjj7P5ihQ9bohy1IrXUd96pnts+N4pom/O1km0zxjg9jFTT0Wu54jFe/IXjWy93g2NX083yI9S5iIaL3p9z1m7RzeWjmzfanWHlAr40ftz0g5780+9GxkvfGTVT73Z2XLpke8Z/seN2GwZgrP4OB2b7Gfv5z8VOlOne4hDwAAeIHwAEgYWI2BDuMtHCE/Dmyc08a7EIKYDmYbJp4/f950TqxOgl64wDL+Mta0l5KEO32lXstalzcyilt1HRlfR/SOX8W+5nN3TE1MVBZ5yHXK8ex8Y4Z1f3qOOH3/5YJ6o86YFWpC5uMtlCHr2HQuOuErPGOyRNdRLiip85Vlt9MVW6O09T55Tgh57tqJIn9zVqWcWB1PSzG086K+QXWC9n3HpiPmZjYGXVIeFJ9doBYAAAAAAAAAwP3wYhXMYiCs1QSaYhipf3vpxAZVlvZcxGZA7hkV9F/PEzBSxPQ8yZ6HxjNw6ONmPDojL8opNYhBV+0wOMa7F7rsszF5vHr18mRcL5bjEe2VPasCmvZ2iu3eMeP7q4x/Kj7cpuKg6zouIcBwAwB4Mvzw/X9Dv/bpPzH3/eTtLxCR9c0+r2DWyJb7LqrgUZqZb5kX2+9MOdfEdDcdR5ZqfVCPUX/nzPnM9J1m1Tc9ZhyDTG/WgtvvDG1ey0CtY93/kWrMq4eX314R/3qd6Qc290pla6qQYn2MXhNBr7D+YWONTeS1kYpl+Xfbtqw6DJcj7RbNWYuWk5KHV7Ksmb6hdKrPqFt3tXCL5dxsjnfGI1b5vf62qaQ1HLk9xaRULVpK7B5nnvkz9PK1nMVyxl0w3lGtpAwh5E+fbnckVmxn77rfB9Y4qRnrLEueCTCatdm0Y4sYo93Y6tiHjr5GWsyRdHuSlc62CMF6X3v7ZxWoZ74D+hhWVet0fpmDsfrAftS8b44NQM9MtJ7N3nEju5ZG789q85Wr2bbx3Dboa6hntnCeWojF6Ub3W5fLzNpPvDjSnhCo7ZPY/Ruz72Dkr89Pp9fl7mn791efm6fA18+Z397bz2WTzx2AghkA8ODRnfbVWdAhxtg0yKsaKMlOLKeVIQq8j1dPlWp2VJ2BV8/w0Pvgyg7GTKdj1qE0W77GGnDlfXIQ10y1MgxIhsHijNGi1yG95j7INGfq4U31mi13ZlBpHe91hGbvfW+AMOrAyutrdfDajnZ9XExbY4AVFRN513XVHSgAAHgKWH2B3D+Jocz24LZ2O9rIC+WF1nnMGy8pD26LIalsWxcR9oudT7lPVcJGWGEBrAE2LxRmOZKqb41ekDKGYvgX+S/qOyRnJlnMGJ9TCGXhWFl35eiKMRbnQ/5AUeNIDEYZI0xDtXEuvf6A9Zz0jk0p0eX57VS6xpAlDczKECkN7PLa6Rl4ltGXqHQLrP5pzwEmsRymTd8kBEpq6jilsuAkP+RpI1pW1a/cxLkd4Wd+/t2vUFRVkfeEQ0v82mde7cUAAQBAcy8G5nf/6stEZKgZ2HkZgjsYbD4OR6zDpBr+jBMDSnse2u3zjAbpozytj9/Zskf7e16pUQzlUZmeQcbzPsoOyMhbpn/Pxse26uClPaOUkOm97VY+s/fJK3t0nXTHy6pPTuOoa0oZ/fvfrL6rY98NjgcAgKdC6eewhUYZ3tkwIvoq3Bdq7OebHeJG/n/229Gr6zVG+Jl+hPeds75f15CvtapTr8+i6zlalb0q5w59v1G5Fr1+wmwfwTKseI6qrFYWmzkOoHfuvWdopOyZdeLpOjXlbLYCzqwbG4iKXPv0+WSDV97xYa5BAQAAL5CU9jZWf++43T1+67bz9nZ3RuhQb7z/RsVc3m7rmMy5veV81fdJz5iwnAlyTRZZdnuK+zGsENZOmNJn4e/r1k2ny9HqXl0eb/f6Rjqk4lm7k9fPkeEfrf0xHeKqhfOpj7Pguupz8L6t3jXTx2ms+239jkk9h7nqR7lU39PLpW9Hi+q5XJz4xl4/i68dvx+6v/JsqZ9ZC90X0XlrNn5p1nbmkXVcoyJX53QfQMEMAHglqJQHzkBzXVd36rBM2yhWjEF6L2at98GUvy11hk7Xq4dGT+uRx40+4BZW2jMGIU8Vo//vVemMYcQqw1TRnDgnebxXhoV8Ds50pPS9tmJx63ysulvGL8/gdrazYHXSR+9TSqm72KJeVKVRDZF4jtTlSOI66GnrZDw/AADwVLDEDiml7NzSi3GlLZbpuVtJ3yMbeagM1KWgh9vlFEq73hoeVvH/ts+Ut3G9jT6e5YDo9QXd8+mcb9knZlAZ9bbENRYzDqMR1wiWTEX4yfJm664NcEREYeUFpYpBZua6z4almL3HVn9cnp+5rkT2S7XXXYpvkiF2y/3xINIp1bvVtwIAgMfGvRiYvUF1ntkUiRYqK8ZbyMGmnVetShipTmXdzg6+76pc9jp9PWbPZ2b/yGgyqtusivgadJ6NJ9OJmeMpjKy8Rsd65zcyblnHnVU/e3nq31b8Ju8a6WO0cUr3YXiAkaexP6sTeCu8AgAAAAAAAAB4vHgGcN7+kY98hIja2Lp5rDqI7z9LT8E8Gq97v/X2EkLFFgaNxt+6PlpQkRXQjo3LY/Y8vHCJZ+kJonqzp3ple3abGbuOhWdH0hSbiHZCtmEzrXoUdW//HnjXWsfT5mdiMeyYo+dzZJ+K5NlhvfdD7dfJju3/8Ffv7HVey3vy2//VvzDz1EDBDAB48DQfFOdDZzXYizp227ams2ApRvVCBrLDwMdrxWfvoz6j4KjSOOrc/W//Qz2LpwC+Np+yIVYx+bw8vfItJ511bWc6UWcdaPIYqVCyOlm6M9YLbTOjPPaOmTmv2fQz6n15jDctrOeotTqnlvOXiCisvto7xzxMgXi6W9kn6l2V9WLVzF6nOiu1yX++wnE+WXWn8rDuxcgxWuV/5fs7gzWo8+rWW4DOo1v3QYd+5rk/0zaPBh73xcxAxXtf9bfrrnXotQv6tzUb49py9H6LM9e9uZd8eXhAFdp7uegYxE7dHxohBI4IkgfRcUuNclmqlrcjDGEYXFJLVdq7zzK2Ml+ucs/n2q3qGdHxllNrKJDomTJ3odcO63qfEfl4+fSOu+Y7P9PmjQxVMt1UPkKpGzptknVtR/lbu2dmi1XbzHzb5y6k+rmT64lUdVfPYgq2yInjoAcjDwAAeKzcq4HZ915sOXQyd25K/Ni+54CU8nk0sNGfkZTa6buM93HTHwBtRGpi1nYMIaNOtDeA8VZjnTGWeHnqgYi+Lp561jOiyXtwrULbq/MonTzf2Rg1I4Wzh3f9rPqO8pz1dlrn4BmxRs8Iv0MlH90xU9PE1Aqsy9LvYAIAAAAAgBdP1Q8Ouq9YHOhJjVG2rSystgqj+qyhz+rf5n5oZVTOZuHjbyecVCrJZB+WDdal/72VY9R1kPWY5RonutXfP2swnh3rvChD5Kgvb41RrXHLjJPvGsPxTJ16zi6rLvIcKseSIYJIVIfGq/Li/wcSsU65kCI4iB3vjVX3h8D+LHM7wGO/vt3h2bNn1f7GJuLYDDwawZByVOnyZ/Ic2X02Nd7VY/BZx653jfRxWvCkGdt4tODDrqfOz3uH9MKlZ/Cc0j3ntEw/cm5f21bycVmVrsQpbb4DMdTka6ptNDo+eP5WqrVX5DE6r1yFgcghbfX9bBYrVfnq23251HGsuVeRa5XbhGX6erwQA3P5zX/FKqxH450vVr6xh3S86SPURl0dYmNkeLXiXHohAGYf6pmHfXSsV8aobK8RkPmO8u7Vu1fXkQKlx6je2tB6pkHRL+/Z8xylmzlvfY1mr80ZFY9HGSjsvzey1QMlb3U/9f57qNOLoHKiTKgUc6dHbV+WpSh6DCfM6F56HW75d1kW19FhKTh6+3Ts5X1QSbm+1vFn71mv3ZnJS1/jYBzXc0ydKbfnwNOcdWz12syZdkPfs9G9Ht0v+TyOypTlyudi7IAqWIP0UQddvg/aObmrvO2O9v/P3rsu3dIk50FV1e/7mXuAQINmRjOSMJfikGdGM/okkDQamRuAcPCDAIODCDAXgLGEZEthI4UkUARchwmDJE4xOvgumL26ix+rsyorK0/Va+293/3tfH7sd6/uOmRXV1dlZT6VhccjWNDMh2nJ7Wddf19o7Xdhcbi6uPfk8eBDL2Q/lHGEq8fSvbwGCS6PZ+zWdICr47Mmo5bW8x48hyBSPEvHwmV5F5NcnuFdumvusNr2LRmCAoFAIBAIBN4yIkRGIBD45KB5RCVHCRhoX19exF0IHDNDMxhIW8G9h7NRcM9B2QFc+AVPmdJvXDYuzxur/p6I5E9FlE0zvJj1LELrD5aB25KDK3vFUK+l8RipNKwaKqUyNVYYxwSiwAZy0WHJ5OvOrv67MzVAgDn9W3GGBQKBwPtCqefcnHLa03wSPYTN2IHAc/7OeT4Yt5QyzzvbfK2UMoUNG+psNOSUKCMZo8ly/j0qcqIe7WILVdAOiqu1OfVwqasH2nIM0tWQPpaTyuNcwvqDNl9fcfo/w1HJOm+QLMd0Im9i00n1359/DnPB5T3IgXpX5vkWVg87iCemYk19RyfUhZz1TLnt+8gv6KypWyuvEjYzPqz4Lekr93bn2kMORVjJmmQaR1obwhii67d0rVOFUIAaVnVBWie9rhGNzv+klGQmNCUIvXzxOvy2QWI8GVhx1OL02vrEy9aW7kt1SYQc15qTKVfMT+rJ5J3R8WUiI9TxulSv1U7aO68Cg5k+izf+NegAbbl0/qet1TJfjiS7dZ3DezEwdwFQA03j1vlCSED1lfAT+P6KMUOSV3phtCyrDu0FXFGCVrCSz8vakYwH3EBWhPcsPvbEBjvI77Guggb16X2JVVT2Ly3bq7ThdKuDuQQ68HgURKlu+i3NoUHuf9sEuJFhgD7nUWF3RCAQCHzW6AxmGI/JYapMnFC6IKPQ5hGq2zzCgvXgmbrUyrZ1dR5+gDEL5a0wdR/R6VYYzI/oDVfyrjC+V96/9DwrzjBPLGmPTHTLcM5jm8O+Imz3oQtEum04GMyBQCAQCAQCPgSDORAIfHKQwtzgA/joQh/HuaaeTC4WEq0De6I9hhRt4W45zkopjWlBnQ93lvWYj3vmR5xr1j1JbpzXWpRzMuLnWFncawah+R6ER+HTj3nwuxodp6WUdBy8YwiHcupx7RiHnHFKt8bs4d7PFbYTZu1zceqlEFLtRGQmDBVnpHzEQHrvT/D/szwhXSAQCHyVMYQiQqzflFLKR226QzutvnQmc4VYii93Jt19zOfPZ8HIObPzA/1/znliLh4J6QOUDXbUxlTsZOU8hzETHBAWm1iCNW9q6TFWHX0eJ+LqfUmeK3PtpC9lJPNFn+YjztBaK+tg9Di8pnIS3z6DjjM5CIV400p5Y/xw+GbOcg8IdZdSYg44/piYdMpjXINIjEyqE7Zyzu+6OezIJzzFOQZOE1rPsXIhPPr9UfasFNpQdIQaxDWJ4Lb6bcpj1BohYL5uO709+r1WJ73PrXdxuavsWSr7tGbeeaY9S4bl5CDVeljfKc19hYYQHNaCTuKq1EZc6MF7Oj4/fVUvLy9DfkpQ9BIxhzLdKR2YGrc9SEG0yfMaeejW8IT9Q6nrJfk6NC5X2kIlsUmkQ+0kvC8m0RVgw441kFlGLsuggdutKZ7OPKv0f+4dTvIKclu4wuCxPv5VwNZDiXXM9THpFGWIT0zPmXhkogvDTSAQCPRdOqA6toVJ6sYUfB3/n8aOl+Y/7h69ru1ykeaGK84aLY00r3uwxPwVyuDmXUnHeAYDnNMBPAs5a0HmceJZixeu/EnHS3NZV9pB0h2v6BBXHJlauxU4kwKelX6P7AIa9C1y4DH5jv6fP/otLET773f+/j8UZQ8EAoFAIBD4nBAM5kAg8EkAL0QhtA5dZG/b5vJ+0lA83MKYerK5RTlcAwcB9v5Ji2AP86KUko4WH2r2embicaWn1kpy07o8DGRJ/iF2o0Jt0QwH0nvgjDPw1xMvkTU8tHxzOs1QJhmH7gzm0SHGH7KXhvy4TsmjvmJ44sD1a48zjzvJ2PLWc05czAYBxrgU/mrbtu5wzmPZ6cggSDpIrNF+eGA4wAKBwFcTf/WHv3v/T+njLybBADurjat4Fxc5JD2jWGfAYM45M7tUeCICNy95CQotHSEXJXSvDfH70RjL1jy/NWboOR8JYnBz7Mq8gdOvOtSs9nqcxag7HDVGL8VBGZHOd33FcafJjH/vxMVoOchY+Q7SFvmY+nhKqWmy8LeiPO1arZ3A0/odYjpXpKsDSQ5upbmut4HStxCckJjEFrFPZFsK3bfpksSxx+mjWv2PQCLUAaw1I12Lvb6OsZbxDlodevmWnDS95QCvwlraU7dV58ra1wOrDUTynFGuh9DHXbfIC1IcbrZ8ehi6ILUkm7S7qLUVWVtlkq9mugZP5O/6u1syMP/8138q/ck/+s/ab8kAQdnH27al/Ri3WPQXks60wGAmwzvp9DtzKMBYnp9Jan001sJaUjYGZepJA6E1oON0EgtYyvsM5QbXz5XtfS+WcdAzCIr3yf+o0qIZuSTZvIO+VaY1MHN1YwMN/j1P6vpBDFMdZCAKA04gEAjcgcdF/Ps4YOzGS9MRkpFfuo+vrbBEH1mIecZ6S2dawZXnurJwkXRGrh5JthVDjTZnX4FHH7HycAt8zcmqlqXIooWMEheBij5u6ZM4z7TddtpRUEUZ3yLaM6dZD845N0MsNjA3h3Medb6CHIZ67yktBRxYhtc1Xucu1ilbXrJLNWP9khqf0/jO2+K7onvNaA79U2f2P2Ic9fYZrX2WDk8m6bQ1iOUU9xjYJaO7Vq+WxoLXYE5lx05tr2y0tbk+xh2gXfJL2xVRU7djcGNSk+/W+zhsPe+Ha3Vnz/swlAYCgcBbQDCYA4HA20bOKZWc6q0rdzssKgpRJFOaQoM0Tx1yyVGvcovZVEv6gnieuxLYf0vebFAma62iAuxZQO/7nmqCQxhtJZQ6GUop06nG3GJC9agS0IXMcS/0vDmnkZwm3CKBW0BwSj99RsnhhNPMrOLeHts2xp26Qaw5zPw6GbgHab99r6mU/r7v16A/dCZ5IX0Uvytrwb7vu8sppS1KadtqB4pSzzvuxzSNxxh6L4c3JOF3B/8HB3JblEHQvpxTbgyQ8x1B3y3b8GzhDAsEAoFAIBBYh0VSo7q55CwEUOdS07fJzk9gz0uMac2Z4oWXdWrp3Va59LolT/+tO7FmuUb5JHksx6L2Dr3P7HWaWXVL9Ui/KVNY2lkqnas0+eGr/rySvLQdaH005vi409W2E3B1UUj5mlMWdlRsZK2297VWSqmd1dD6gyoNj4cMzPQBZ6PL/Xqt80FBtAyxUYQOs+JxttgV1Ehheak5QwnN5x30JPktVomXQcCVsTogW0Yc/H+raOkjWh1g8DWrjaR08MF19gQMRvCB6bJrda7KuMKmkSZ3v2zGhKp45j8Kak2wzT6lu1zYkIv/agM2zi9OvtUeW7gxhWtLa8zT6uDgmfhxe1gKj5aGZWhQBaXOh5/g/JZS8Ch7w/OuuXt8GAu5r9Ctbdx7tcYa6ZpUv9YHPeM57o/WfIDL4NrKmqPx9wQLBjyu5cyPU9jQ3r6bQhSv01u0bRuaZ+cFgnQw5/uENZdh9LFVL+vKeHAFHHvrGVhl3z2zfCnNM5h1V9roWYw+a3GnYRqPjrmPruhc3rq9zEScdmVc0sZAqby+wIbr8lgx6QeQFiuFz/tsVLSwQXtnJmdkqAHHO+z2PN6hEBl/654Z9KVUujMSHyrUhic41OtkLaeUUt6AvTkbiNw4cg+jRfTttB9tXOTmZHCIvuSSXogj/T73jHotBqTD8wdg3/dlvZsbW7i+yM2ZVF+XDqTm1riSwQI74jXjm6XDc3ojfXapbPqsnExSOwEw0YD71sFdD07+mmeSCBeyDOtBjUCMYrS3upBuKxmvzkoGec/M92s4b313ptv6VncYexFhxA6XEAgEAp8mgsEcCAQCgUAgEAgEAoFAIPCZwevQa78JsYYyRdvOtLZDNA33J0LDNu5opAZ4Ljaz5CRaZQ53mcddnVSWiUhJyqX3KUGlTAxlPeyrhe6v051/1JlC5d2FM3802VYdjSbD3dvvyG969tC7d+/GcoVdlOB4nR09vDNOIrRxDrWU5r6gkQJn5+X4jNb7lYhAPf14/UPEN18zMFfeOyuxirdzsLjdZhZVK5LkoS8AHpXe755KiaUHf2dmg8UssZih0ovUWEHWh0VhxQO2ytXSUhm9HVaTHQYn+jiUhWAx/axO72FvUngHrWnCmZ4F/b+RMa4NiF6WFScbDStAtxnN7JvxLx1soeaWHga2JsRz2GeXkU/ZHKEi7iENxj7E9V/6jbffiNUwjwN9DJLaHOeVxkWJ4cI9t5VGGxek7UBYLin8AQ6jQJVJjlmzCYxyTn6chob28Cg1AMyUoWVzjBhOobT6tcTo4Z6LqwMzh6S+IuWj8ltprWtSfvzbw6hZYYiOzCH++fEzSjuWcs5t3OqxDO32DwQCgU9B+bfAAAAgAElEQVQFf/vrX0v/6z/6L1NK8yIWA8+hoIs0A8GtDqzOlNCYi8qwdJB2/ZDnKQmTHsCcvIfXeHS9l1JKmeoviQ+7NcnPzPmczoV1H21OXWkzCRKrWFuDaaxhSW5aLr0uXdPqt9YmAK095fxyu3NlNP2gHpPuyvUF3IbcPe674A4hlu5hmbxs+La+qr3eH//x70xyfON7v+EqLxAIBN4igsEcCAQCgUAgcAHdOCIvqqkBRVoIa04Lj3GB1ue9rqXVjA9SHliIt635CjSHh9e4sQKvY8iqd5WAwN2TyBqcTBY7hqvTw4qSnIwWGcD7HBI0h5uEFUKFWE8er98vnn/OOITl9WVOM/1+vtP9//jxX6ef+u6vDtf+6o/+6Vn5bCi+x5y/X4NuYBkum8GswnuvjYiD+1Q/NNDX5tjfNzsIu5FOc8Im1H8oqShXvs/QfnOga/h5KJPQ+pbbNcPA6+m/3HfNOdMlx/xKHdJ36h37LeKRJBv3TjQDL80Df7n+QftCqrxBV3PUa8+H52Uq5WBMbte6z4RzBEFYmXTkVDP/XmqtQ6ibM/nHQU2DIwm+FWCCNpzjY4vXekIm2tzLeX39Yihv+o7oWToL84jELKbxbqnDoJUNxBvoYy+voyxtYDvJNy1f6wBDeda8RuXrz9pSqOVJ5VoODis0rBR2UpOBQnPwaPelcrz6xURMrOPfnbDSKWt77mcwR5z1k/6Z9jrch34O5dN+TuWnfYB9pjYSCTIQ1BZeEGSC3PffcL7Qy5nuhbTB1H3alAxtvrf08yjJ45KBmb5kum2B8wp2UBYd2VqR+RffO2qLxjTcB8wdej4kSYrTJLIajTo4rC6AVj3i2uLGUsJXBwcr6D1OU4+xDHiPdPsFtC1lWlgyYdm8i88ri0ecjlVSyHPSw+asxaH3Omag0O/M8pyLfWpaYNHnHPN/LJ2HIiNFrU7fMGLqwKRCxhD8PLRv90pqagN6Hrfd4JZYmWi1iZ++I9zXtAnF821zBgiKaccI6negSFXpWfPRWEZ9feg3qnjHSM8462X2cHJw1+A3PmTPs7jk3qdm7LtiZJGeiz4TvcYthOD+Sjx3Tfnv9ffvs1Z+nNLeg1QnV1cpufXDnDUzbyAQCAQCgUAgEAgE3heCwRwIBAKBQCDgQHdQ339LDpM6mLp54367qzDoVhwzNI9n267kgJCYrSlh1gNlQYx5uFA7nBPHep7EHOxoPY/kMNbaxHLccQc5S07CFWeRx8HjCV9jOe0aKwY5L4/byO6hZV1x8LWyGZlXHPlSn/Swt1q5NNwcifOZUmcHWQxIfP9Ajvl1t6Af/dm6s7Jxmxh2Z62yA728bJ05d9Il931Pjc3cPMW93frDZfzjfqXybaaNJaULf08jOA89jmB8rbNb63SP/p9e495z+/+Ui5fDqlMjLmlySqxeWr813nDft9YmVrtL+dacxLOMWjmYHOZhMA+sYmWs0hzY9zBf45iRc546Bu73EB6UK3v4yRAbaDzZD4GaxkMvQWZgtQIjEw4GpfO9NQ/Qvm/NZdLco/U/Koukv1jz2jwe+b4rS9b2WyClWHO3/F34CHVy2FWbnHMVkm5EIc1XlOUN4aC8h3pbbSGls8rLpK/t+ygPZYMDPLqu1U9X9U16X2oDinksvf+9P6OvvZYNzMdxTApCPv++lrEz7O8gNlhtVHXI2geB8UUAlb01NikzHfxL0CZXaWFBX/aKconL805Y+Po2xZ4iA3smr6bSD4Jsd5hYqdcZgi+ERd7eN/mIWkzfWlMS4vu2PgJl08mTLAL65bZ6Z2UdFMBFRiSFNMCvlFcrTasPpNJHT2XCfbfLQ1mjRrxuWDy1/s9vjSnkPn4HtL99aOD3XUoZFlgj+u/pgAkoC/2fm3T7Iu4ci8i7tFitUC5l6VuLHSrHIxO9Z+GRc27s5K5UnW1V+rgObdrG7mackHcRUBkkeBQQiV3shfSOhMTD34NZ3HHvSKpDWsRJsnmMLRq4fNLz4ro8h1BohqGuaIGxChZam6g8tuuptjGKLlwgx+040uvLuMUZphL6HtZ6RyAQCAQCgUAgEAgEnoFgMAcCgUAgEAisAJzdBxjax+vYRk8drtP1E5pjwGIycGk97GfLcc6xjiTmssYS9jIyHoWXQedh5tLfnJPQy0DCaaSwY5q80v0VZ5AHV96H5WjT6tHaj7LRLPYzTuP9y8p9/u0HijLPh0gdz+N8zShA3shMv6zdOQzpDuTU1OINc+MTHqcgx8joI/XfA44RiZl3caBD9mhINiwffr6T0IIPp57CMNaZLZ0yPx7RvPiQYO+362FBY5mwnBK5hNbhcaQ/i3zAQWO4rTjLV+rF84XGLnykDi9rkdbA7VpJKfU4r/C+mEOja0Vxlgv04zeIs89K4UMppPlRym+F2ZTKW9mBJYX708gf7H1DNq/sNJ0fNJTcWC7EwqUSr+pXz9QRvGOKlZ++QylULu0XU3x96foi6VLsr1M6cp8QbuiuBHpfk0ECbUMggsJl+vswv2WD7NhiO/vJhksG5npu0ZL6oTyAa5Py+Ht+caRRyM/5oziGhFixkFij0sRvfYAeBV+q85jqHstcVSC49N6ypHawPjLu+aXOZylyXflslbD3tWvWRGLJ/8yF7rMXzc9Al4n04+m7pYuNj/88datpSz1eN/1e8DcOE3BNs4IMkBQhzYAwbPMUtnphw4uleHDwLng8irY4XinsWtwennFRGiv4Be1cjqeNvEouvnZ1caYqTrTMnCeWPKSnB4xIMpnjYZ1jOF9VBKXtjNKCWCuLyjr3Az4d97smHIsavrUmdEur6RNp6Ftvb+wNBAKBR8DNBTnndJy7Prtji08H95pTCO2S5MZ9zriflaF1WNDSxS5jfKPPhuvk7uH82vZeyUjLHayE8wy/kc7LSc0ZqbGcml7jNZh6dRmt7TS9VqvTU/+KHjLKhnV1W9/l2ovrT7jdNSMi18eH+hz1c7j3f8agSo1Pbat530HYyjC+k0AgEPhUEAzmQCDwtpFz2ratxwBKqZ0ET086PY6jrWq6sY84Pep8yjgAhyKRjG5YCdQ88tQL6zUe07IAmtGWW9DQZ6QLJo1lyCnh2gKGM7RbDivOwM2VrV33vAfJGcAxmKinu2AZmUWL5cnXFnaehQu3WJXS4b/awmrFwI6vSc6UjAzt/dDW1PJAfs7onlIaGGwQrqelKd3I0d8V1DvLWkpJH+qYvyZj6zujVSXjsEl1+LNWPsEjC1DPtyx908+S4ZnwOPc9kPKsfM9Sng8JLzkAX19hpEjw1CPJql3XDKUYK0w3D6x+MN37eK88EAgEngK6Iwli3rZxl9lBgMHtMMCwyF6Wg4CbM9j42EoeaddVyw/lgh5ECVlCuZquMMhF7q/qC/Ph8z7IZCJa/vo8aeWR1oIAi2FMiTtSPvoOAVI8brq2sQhPMllq7EuUwCbp0lyM+PasiZcR1heFpJ9lHsk+YOzdYP1U6LPxbOr2rLATY6H7LRmYczo9w8Ii9ikLn8sljNDYdqtyej+ilcXH3DmIEciQVTMsWYO8JFPvyOsDJz1ggxq3pLYDJjc1elntpzEdrIW51LbWhKEa+Aqtk8bd9S9Y+HT+wOqakYmTJRltnfJj8YADgUAgEAgEAtcw6mD44L/TCITSSQv4kvpi+KjkrI006nysLq46ILtOrOrW8H9GD8/HfI2Tj3MgtzVHneXEC3n8/xUjD3b2Yv2fWydwu4W0urg1icc5bLFxVx3MKekHZlqM4J4G2hjuYVb5nM/rzMRGOu556JZ0LO/q4a7cO8blce+i5SlIZvoczHKrP6NOuggEAoFPBcFgDgQCbxo53xXG26mE3pVR3iOJ0Z0es1KsnUJL44ty7FZpuyX2jtNFg+VsoTJe8baxZcMCjMSnKqXM7GrE9pVYBZbiPT0H92xCmhW2rlQGyJ8SPRxzvHcch3liN37XwK7lZJEcX546sEyak0vz7nsWaFKoKM4ggZ9Hcszh/kyfkaah9aWEDr88uud8YiikXg4cZgo7EnKevf8vLy8zJePJ6Ivc9YXgI446LysIp7myWJX6Mn6XFguIymHds2SR3ulKe66waT1tbLFdNBmkNraMRFxaro2lcZkNcXDhvUiyeOKciswuw8im5eViuIptvTDHtHbjniec7oFA4CsEjcCFwemQWnmSfupZE3HlWY4VTkZrXhN1GCeD2Utwo/PGtMPUSQi05OHmRjZ9HdOvQJJl9X0ew9p+XiNI6QDT9WOUw9LnPIec6/Lzzw3yU30ZHKkFV9PWTGvrC6u/0X7gTZ8S33YrWDIw//lf/k36+e//h/P1//EfD79bBx76QPe4pzQ/RMtrxUadHpZ6SkfgINpWJ7MGzZWG1hjGKaVUMvVwWwO6Xs9gl3pQ8ZUUdAnYYz9NJMLzU8UdIL0jThbv+5AmGmtxJ/3mvNfSIq2n02WVJgHP807XgU0tvLZpcJ0mKN1oFggEAoFAIBB4Hv72T38t/S//9T8YdGqshzU9G8WxbemQoaDlgWXTeWBeRfdut+5c5owUnvVPqbwDU3MMsMZ6Ak7vxc/FyYvZq+36qQvv+z6tUVSnhKA8a85ZjS2L20RzBmNwTl4vNGOV5ZDU2MqcTL0/9XAKGhkB19mczMrzcUQGvJ7SngcbjIAgwPVT7X1yfZFrJ3of7k3PVhMqg9w6jibzj//4d/rzn9e+/t0fTs8aCAQCbxHBYA4EAm8af/HXf5P+vR/+xnDtX/7WP7n/B8ISnSupvKV0EMZyJop9zjkV4VRXzAadnCGMN15aLGCFFhRvfE92bvW68xlmhTJnj+PoXk7m2VK6L7LaCbon0zOXUfndj3dttdf0XOToshaYHlYflk1Lyy2+PLHVtAUn/KYnEUO5mGWuMe9E5xkDzmtM6+W88NRzjtOYYYYQ25w+/7DYJs+mLZZwX6fXOCY3kJBfX19TSind9p+ceRhWAPTd9ji9jj2NOwNa+PR8pG0j/eDM/43vjGPD+wZlgVJ26IopQGJScGlWGL9ets5KGu1718ZIqzyN9fIIe2Kl3SxoZTzC/LnyTj0soWc8s1aW1G89hi2LVXQlryYbxKyE0BDwoQ7pwCcP5Z+XPW1ea1376A1s92CEjbTQDHjH3ueFMy0+xBzC7m0Jj/l9rruXVdJPfgIhNU79gDEIZ7xVf+/kHjqH1YrJLSgdGQfy3ueo9lzMPIzbme6usox6Tbbcy4ZUeB7Udq013SjtLTMcLk3l02TSYopa36al01A5OL2CGm45A7dEXoH/Y/Zgm58RE0/SmbDhtH1zKbUzVJoaUF5cOhXbrinNoSeQzHAN696cvJ4dFwkbog1Zex/E5RL9DR2iXvex72ylNGdQRvrOoRxY/kzQ90zn8xdDF5ZISbRPWzHzrW+Ecw5xTgVcN3fejlYnXT+263SucM61Ld2T5gp5Dcnr8jRf79PrfctaG2pOKk+5kh5Jn1G6n0g6Wr/EVKbppXl/ckJBaCdyDsvW+ti5rjnHw/0Y16SajFKdFuY2J22307GPL+cRouFTDcyeiRHQ75FFBmE6r3ZMCm6B7F3oeBR5qRxrAIXvAooGw5K8dZ9+EPTDconIwrOQ1PNn9Dx8Z/QuJqxg71dlXMm78kG1gdDZVzzfBpc+aR77xlgmg567DYlMEFnivL45ygoEAoFAIBAIBAKBQCAQCHyeCAZzIBD45NDZANw9/jc2vDev9gKbE4AZN5KjBDNnuW2As+Ng9L7n3BlMvDyE+UBiVQ35mvPujmPRA6o52jSHkuWgwwzklbIBLy8vosfZK6MEzbOMy/WwgyXHK3ePluOVV4L38Byp3m3bxHjjg+wnY+eo46njXJlUpsGZKjAF7rG0x2vZcOIGAoHAp4Za8n12p+ys4+gEDDR2tx05cGo82oU0cZgwG/PoLMWmGwAbEDOOT+Sc03GG1QDdZhjL80yO6XraHHoAwB26JxGDXMz1rB9K7SFLbIjBrcXCzEZdXDovccnD/qu1suc3cPqQpmNp1zDTF8tEWYbcnM/qo8O9+Rkt/duDq/oT114eXRDSd5a+XQfXXsP31Jje5cpxLA9B2sVJd3pIzGSN3c6WSyDp6tx7leLyUpnoOkzrvxpAC27PQOqTnsHzLFfgZSpL+SgxzVOH9Wz0t/Xt0V0TANiBQN+lVH9bL2x0pw3f766+C4ltTOWVGPvc976yPvXINF8fWd97JecvtbUW/63C9RU8xcAsdezBCNKY4zy1vWFxDe35iOdtXXqnWmFiS/mlLc90m4TUdt44XRR44sfXJDmvXYePnEsz1rn64VrbVjh54C/d1iOl95TtRXvOwr9H77YVa5LTmlEavLyYZZplfWsM5nxuMzvO+SjvWFkb09bGyD6NsWjbnkcR5yYJyzCIx4AZPoV12sbTgizKExweP3qf4/ue1f8971xTGC2FR1t8aHUBcHx9yEfDN3ALMM7g7xnzr8joVayk+q2dMBY4Iz6n/EvzMt5a3EKuKGPVdGgkStMOvKi3KZ+k7OIqpjI/0pgE5yXAGRNt944rL/luq35fw8qC4Ep5dDcV/t6keKae8UDWF2U8siizxupHypCuSVjRKb2LOW4c7c5Fu37Pu9Pk4NJ6sLLYkxa4HsfiJKNDVvNgpHsFZjmBQCAQCAQCnxuCwRwIBAKBQCAQCAQCHxGSs29wlh7dgdCM7qeL60jdhZ1R3pRSqntu9vWK4kBCDElODnBSHKj+FosXpQebfE41bdvotMCHC0r1pDQ6BfwOoBao//6rVrRbq6dadZB0J8acXmPISmV5nPBSXs0Rver0stqBc2pThzfnuNbwKOnnKuNwNR1HpsHt7wXn9Gu/0/zu2l9G5lqrz3v9ILA8QAYARqZE0pNIB/S5KKmDc1rj/NZYsfoNcfensUw484XiKtmqE4XWnP9WOimeMK3XIrldgZesRGWR+gl3lhC+TsuR+p+X879KsmryGgx96fBS+J7gPjfWinUKpDKaznqvLd05ObZvu/DnaOAzi1bxVAOzxpT1Ds/2YMHHHpY6rDQBc3W1Gha3TKzAK4udjqR31PUIC4iXgatTZ15bz6mxHK2yrjJ3AXSg1tip3kms/x3fWyfhSayucauILv86e2ism/wuM1voUSblszENru20utqjuLdFFfTJc0DN/KEyErj+YA369y399zL3fWRscouovhCD56ntgLMWjL/a/Q8/T1cGR5ZhaYcgjgsw+jzeiSqlmcmI80o7R0opk4Jp1YN/44NmpPfJMfWugvtWHxlLpcVMSuNzaOMQTsvd47Yac3JIbbzv+/B/Lk1K8wJkXGDp73hYfJN7VMnBZV8b9d4fprkN/19oX/oZX+mjmnHEa8Dg0qwwiz15n6F/WCzblfI9Y90VZu4jMnnqxt82LZu2yyO67EoeD9vZ+/6vGKxUIxu9x72fxI+ptJ5v/uKPTNkeRXtn1Ohbc1sUHjAW55zyuUWrsv1gNJLW2sfztPe5sxmsGcPYwYTDaKmQbghDl2agpHXAb+6dz313DpPEre8GPW0qlYfWN72GaayDaP3zal20nNW8Hpmk9bKm+6yPLXiNaJdxZQzWxh8rHb2H0+CQetr6s6L1eYY1cTNWQFsfPRUc7v26vVn9JhAIBFbwXgzM3ATaB2vZcp9St6p7DXjSdRyXR1qM0AnCG0JDyq8ZoK56eJTa2XIt5eYawPDaShTLlcTWjCIecJO9ZNyz2pyWJcnm8fhJdUmGry7jWM78PfgN66sOlPl59HfxPhwtgUAgEAgEAoFAIBD4OKgppSPVZkAHbIWYiBrp5XRw7eAEGtfepYwMTrpGfffu3b38LMTUBeda6WQUCVKYri+++GL4TRnLtE6Q6eXlLvvr62tKKaXbZGcgDl7wGwj2JEyoSamTiCT2dq9nfGaJJU6vUwf0XB7Yxk52+gtPZtOINLRuC5SAIjm+JfsXDfXV0kM5kC7R/6ShPC+hUWrD/kDjO23JTlbYu0Mg1Zz3b6dDCcf3f8m8/WrqP0Lb11MmILRBuEhIv8FfCFVImPuJOIgbC/z8FveEvjOn+S5CZAQCgU8OMCnWG4x0aKIhxKVa54G5DdrkUIE6xC4eB3Yrxje9JzqLCsPYIWyrWvnDW2h5mvOjT56gWMxbqWgMWMxoXmF3e5wv1MliKSeWE2pgvgp1WUwTCyuMbklG6Tk1BjPgdruJyo/GXJTq4+rIeT4zAL8jiZ2MFRRaNj6so8IBgEl4RyU3hTBDh8xQF5RypASLlnztfIJng44PU7xbRgsDVmJz7U3fts1gftxh7APtE/idUuWU/uaUYCnW7woecXZ6vhGvc1rLuyLLI2VwZeFwBin51gEaA9RTp3Z95Xm4g2rpglTeCSCP01MeYpC4XxzTHsfY5z+Ukx3m/ZwzE96itt1Y7UC/pO8OLeS56nH0RTkMtSm1gRbvVmn6DtO2eOzr44I8Lt/qkQoYk8hz4WfdMPMz07bXD/kDDGkUvUTXWa5/69a3fYWlrOmA3D2N7LUKrEPh8jRjTW6TIfzN6P8ytOei92k6S/debQOqC10tdxp/hjNuxvAuuDysPwTJJxAIfGp4qoG5nTafRyt9SVtKzVPFn/jZUEdFcprEpgx0gJ4nVulE01alsYCX7uss5ZGpup1nj2ZyQmPHuJ3dUuBBAW51M892VcmYZZMMa5wiROMYjWUeB1Y4uhI5KyuSdP0dzO+RX+RaXisvPAZG0QPXDlqjz0+8oaTdqMxc2Uel/ZL/xnpcIDAQtc45oKBQErS+//MPf3tMfOb92e/9ZgoEAoFAIBAIBAKBwKeDaa1MwiWmyq+dx1Wt7BCk1w9q76DhGYlcnjW7FN9ZWptTprAV09g6NJw6Juga/zDsMauxda/uBm/1kd3wLUzjQrmPyjjZP8D5KdjaqC1OOlxc6jeSTUjqG2Y5RF6JRPUIuYHKSvNRcs8kg1Bua7tC7ErnQ8GuhoLbyOnvCgZzIBD45NC2KbXA9H0Sh2G0xWIGdiRiwgA8LAluUrOcFLUiJvQ2sxKtSQSzQrWDACQnmP4cuDyZdWQ5aLj6sVInsXM5dvEVp88qu7jXC4cbHaKDy1svvaaxiz1laW1NgRVdjdlnxWHl2hH3PUkBx+VKcpdSph7Wi4P8L5Njjfs9XSv2O38f6M9Kvs8DlN3blKcdRpLpwotXbHEaSenlvhlLEdefR7+uxUuXvl/YpqehEtGGb6cxNXmZuMWgNX5r221lB/F6X/O8U7r4tOLX4t9cvHu6R9RD9qcLZMk4oMlG57i2qEGiAfGEspI1ZvvkTFdi/VP0xR1570yWo7EK+XpXF/GXkft4CmSN43Z/jtvt1rarFrR1nZ57gs/voASPdNRWRuOYVuawMWFOkOaCex5os4xKn9PRcnF9O+7KoKf0h5jSU/l6/We9aOG/J/7b5lES7SjcPCfB038sEpFHv8HnUHBprPFf07/w2ESNdtu2iSQprqpVo8kVYH2S1isxoz3v6cpOO+1xuXUE1qlApK31+5SgL/71//xP7+nS/Vn/zn/0D1TZAoFA4GPhKQZmj5FAWvgsTzzThFHJ33HA5hZfErxMX4/hZU6z9rzWok/zwEgeMAqJ2U2f3jP5c8qdJq+cfx0iq/tiG2rKNv3d+kzht1JKBkApnWYEEr151MvMKMYaJCWre+fkMAO7FGsoEAgEAoFAIOBDzSnVktK+pyONMUNvP3nXdPuXUx3b931y0KQ062sAvD7ADgVqpNy2rTOhUF5On+V0w0rsyYVeEFBRjC4aLehIadqpWWs/1BnVhksc5MKyj/KNvyXnkNfA7HFCe1FrFcNUce3PAafRDKbYiI7lpfUOBz6StbjX8S8Z7CVbAZfmTCDWIeZRZKL/5wzXVvkFMX2PKXYW3MvNntHyomoGcS/2HS9yznenwT46F6e1L8RjRY6GIZ2xZqbf4JHGfHAGl9eGgcvEYxeWkdYpMY6t9bXkcJXGFZF00V71KDfnbODqo05WLQSfBq+dy5PXY79ISY7B3OL9QndIfFtete1Y973f1eCoTWly6FNY9ryhnx+6bJYzjBIPrPFu+k6IjTLTZ+red7VcjGAwBwKBTw4/9yu/zl7/l7/9W22AhMMZSotej5TsptCM+fHQSScjnYUDSgrkvcZi4ZxUk1KGYkhrZfuYiWQCZM57uDoZS+mvssE0ho4UqxPHiOwsnHeoBJ2dkvNmKi0ctLQr72rbNlGBxP3BU5+0XYtzXnFKjcReyzlPyvyRIFxRbt9fV/DHwyfG+sbfFRZnJaevf/c/EJ/xY4AuquiCgdM5ryj9K/3O2ppnLQRW6wN4Hakppb4Fc7kWu36pzqH+i+PPigwWVpi41sJVw8QOh3WRYnCi7+6RvrLSJto4uyITzSNJyM2DNG8gEAgEAoFAYA1PNTA3y/k+eifqsadx+1SHZOG/bpQAZbkvdLzMZalumk5jBzQpDlqnvuBLdHEqyCbJBPWtKMYi29a5pZpnMfgXmjidhGVvtqNMqw6A9i4sz9TctjyDQ65bL2+4JrStthUY55/bb5Qt4zRPMha+L+ScU8onq5rEpm7Gl7r3a9RrrRzyNxovjTGj6GE0PN+ptIUYb1kUmeX7bo4lHFsFx5JaYW1whyFJXmfuW77C+MHtoHltZ695auVKbdS/nccPzLnCMPAYdbixY8XwgvuzxMbAxmOJQZFSSts21nsgmeb283nYA4FA4HNDrXVwIsN6at/3HvbrpMHt+95Opucd4cQwX8rEHtxKTi+M44HTg6hT7UwwXss5lTLPTaLDvqJylTAWkk5yEJly5Rmnnvly1ZniWV9KeS1wTjPVccfoJ1Y7cOVxOgDHmoQ5H7LCmS4rTiePs53rQ14dzNKxOP2Ptom1xtP6OC37/p/7n7v+OupiOOzX8J7oWi4/3wmWs35g6F2++9/t5RxzJjvHCBdzM83EEGtNr41TtA7pm7Heb7+u23uk+rldIVgu6eYNFXYAACAASURBVBlXbWJe0oL1nNMaEKcxwgtJ6wcvOYeyaX1rcGnOs9uS/p6ISTQudZtbzvxE/nzQ+pL6t7UjKqj1NqGfWnMVlcl633TtLKV7BL59S4FAIBAIBAKBQCAQCAQCgUAgEAgQPIXBLHsHuheAxsft1ndnbC7qpRa8OY+wWb2eSx8oa40v28uKlp63s+O6rF7GXPMOjeSBVC+wjvvz8NstJY8lze9hh0syWLA8jt78+P9t273A+l5lbvf3PfnIzD7fn2t8Pr+nttclyU7TWizpD42X15K28wCcdz85v5Hb/dAtfPhNQyadP9XWh2vVv9mUZG90SSUVONQo8acZc9dwv6JMU1yn1Zcx22j2hEOemXzURpNqxxvU5OCYPStx1bh0GvMZ2gg84YB938Wx5y4jfaYxf607Iu3PY9w8PvMsJyyrBzhPOyCOCf8Bv+EwtdttPGBOO5jLE6sLM3poPtxXbzt5/2f+/VZTLuO9KTZe2nsGONepQNnSePh2wLL6hDTWGLzCXJfmVq0cjQF3RSbt+78KjmF4Ze5e0RksttBK+Z681sFRnEzSfYlZycmy8lZW2JKer9Oro2sM0ytwvcuzyhYL8gn9+Bpqqvu79PKypXzGLd7LfUzfttc2P787z744ckrp3f0qzAH56DrMAdHBYKwvta+h2jvk9da+06R/e8CkzChdA1rjwRTZl4KzztSfeO/MTbyOqOOuKHygndZHyvAb9Tlg4eJ3C4wyMhfCQcD3dJ156O2r2niC9To6l3Kxje/r52O6RqGtTzA0neD+l65l+jVu99Lq2Inr4vQ3bn6jOgPWLzF7sNC629Rcp3BBuK2558FysDo482zcPNW+s57q/qxl67pkqxPpp1vvG211AukS3weuIud728NnSnf99XSn7MA4pet9Uq7VN6R+as2t+L4UHo8eMGzFNu7vjrdNUEjrG69tSsKqDUtag3vnLukda7LR9yEdNCodGAwH/kprEEvvu2q7EZnNbe46+5Kw5hftNgLj+Ype6YW5LiOi0jEE4oPCkpfa/UBU2uY5T9GZRUQM5kAgEAgEAoFAIBD4SMjpbhjBxrd2KNWGDHywRkzIS3wukg9kIMEO45TAcMwYachiEnufsX8VDuvTSSbXDV+cAfWZhjQJVx0+XmOqlNZjTOIM17idOOOL5MyV6hjrGo2+HPhD/mRY7YBl1xymnBF9RY4r6TjZrT6JiQ4HMVgm/O6S/O4GOZv/5e062gOBQADjvTCYASvsBprHKvsqC3WlbI05h8Fdp56Z6mwLawJpHqNGIbjXg3l3qx4w+pyWJ85TlnXdq4xKE7qlpHjllcp6HKBU6DLRRYT2DBLrGx9e9wxcYVq+FWxbSfVs8+PlZPkc8C2CN1WTOycPz8v0UpfUYpnTNFx+Cswc1RRpqT9xB791xX1evGFmBK1nhck2soPtb97jkZa81ziPx4vP5bfGesx6z9lmT1GZJVm0vFQOa/HLpfH0ES+k/J7+IMkr1lHJb2U3xVtBG6vpKdCMkUTqy5TJpdajvE+L6eH5lgHa7hRpPn5ER/C841VjBv690u89/fgKO0VibHlko+/OYxSzyuLqtvqqlqeV3+kyPY1QvtbWUsxFrR9b7DcunTfP//2Hv9XzoLjz3/n7/3CSIxAIBD4NZNa4v8rEPMhfS+e1dHtpHNaM/tLZNc1RJ9g5rtoLJFlpfeDsAdD7NJ1Xb5faSNYPLjgpSKztdlCz0LaNEUzXJVQSYV6XdnZQprS5DidYtS9Kf2l5Xp1zZR1D0+7kvYFzCt4FnKOQjGfvbZuG33ui78Be/1sIBnMgEAgEAoFAIBAIfCzknNJWUt6PiVVcSmnMZezEqvRaUhbgJU8HhtVae4Qi5uCfcfEPh3udC/0DhTlAp0PRiEaWQ8vjtKT/l8A5t6Wyr6SxUCsfXkwj8DyDDIONL9TwgQkLdrlzXs7wuGI8WSFnaczhK23mkUkrl0snvWOujBWnez5qgoAYFYUW4wxYtdYrpsJAIBD4IHiqgZkOpDAm3ic8KRdMimfaxHvQWpnGkOqZsKQJRIoXo5XNwZrMpTye++0v8SB5ZGzeJmCFwXWhDq+M+DqNYWqV4W2nlfb0yu9hw3jK5cpo/XmBtcjV1etUFOapDl1GWodE2G3tw5TR4llNscY/LkpJ6TglLscZu/bl9Gjvp6euVtGJi5Vz7h73/5TmvoFPgefK8XinJVYqpwxLMbA0cEryht423V5L5eDq0xY0HsaaVI9034pNiNt6hYmHy9PK9njEV1h32hgsMRRwO9BF7OriXKqLyz+8a2HXC1dmJsHBcil9AUssE22ee3zd/3S05zl/ryz3PPOC931xC3GrPq2fSnmvLOjxe3vEiGKNHZ42WKlHgvZdesc1nGalbaV6PHlb+0FZuFyaZlH/48rwtK005ml9k/7dOOXlhbCpWnHnfxhi/lz3og7wBsenQCAQ8KDWmvZ9T3XXz1CadNDMj8tUN5LmxFV7ADdHUKcYXRM1WZ1xgunanerUE1vXgJRukudMd6u34Tp9Lmvel+ZR73lJLl2EOhDJ78PQJcRyBb1IeiZL56Jt3NoAGNkH3/9W7XHeGMy0npV7ls1gji0+Xqf99ygQbzoNzzDbsmDdBvKJok8IBnMgEPjK4Eh7d2ydWzzagHqmyceR0jin94V3Tp3hc5CRNPMHsEA+LzSjH55caPw8+3CWWRnyysFN7HmafNNZ9umQW1xUc0ahK8Yr/Kz0cD+PgZdTSqQ0uDyYv3s+2ailPZfHeLxiNNq2rfUVz3u/ytJaMoahk2N7daciT/paKahsJibhIzIHAoHAp4Zaco+pDNuD0542UE1u93vv3r1rcZHb2L+lVLZtuNYIAVvfBl93MGT0epufj3HM3n+Pm+BrPdBc2fPO47U+Z3jG90cNOlpazeHXdERGb3tkXpL0D01+T72YOLBKaDl/DXk10sUKuDZe0XOk8ixjDk1vOfi1+5xcnNMJyzQ5nrG8xNmOiSlDGVStC3UoEAi8cTzFwCx5evZ9XoBLk4F38SqVo3k8LLaIdP/qpOoxnEgle+uWvHSrDEYur9czxitIY1mPsJ/eFySjnIdVJt3LPeggXHGXpcnS202OyduNo3CfL9svw3wv9xfLlvUXf/RbRN6Uvv+f/FdqfYFAIPCpoRvDr2NlzrFYPiKr0lE/dw07LqS0HxpevVC7d0Wv8xp0rparQXI6rbK+rPqn2OEXyl3px9QJZu0cxKAhIo7EtL2hG2tnoM+sZ5LXaXx6CCWntNfmiMOy7SfD8HaDv7d2/7W8ppRSesl9WdfYSSieYuuvW38P8/NVZAhD4x3pK0dCB5Uh3bfJDo78Qw8pwLUvPRsNg3OOc/e1V4MNsbSs0Zg812l985YRUypHuqc5ya+utSRHN2W8ceu8K/VxoGVwcWM1IsAjkAgWq8/DGprR/aPHurmnH3Z9jnnLy5bg3A9uR9TXv/ejJdm8yPl89vZbd+zTsVzaSSjNYRabmKaX6sdlUGcajI20jul8LCITjT1r9QfaFvCbxlIW1+9UDqJ/0XRSfuub6O00puMOArVAn8XaQWvZ2KSzFigklroUT7vV04oby/V+66u2sVWn1wrEOWvSd/h+3r6P4zakp+eASTLmnN0OrmAwBwKBrwx+7stZAftXv3MezHMOiqXghaSy2CS3uJRUuWmDcZkPkFth6eSc+3Y1mEAcRg5ONipjD0skL1y0ctp1tGishU/DlYNl9hhNtEWgZyLX0lwx8PXfOfVeMf6tdTzQw1s/d/2Rd4RBDTueBa8lW7849tU0bLeylLijKUOwDasIfwOBQCAQCAQCgUAg8DbxVAYzoDOY50U2QDK+bOnczm6ckL3KytRADTCS3LKhYSzHW68G6bTMzsJIg8ya0YGyPqRYkateeS4d3bIu1SH9XWFNXfViS55Lr5cKswbQ3UuyAK48S/P+VZ/8NJ/sTZ4NkxKDueTRgw14C6y3QCAQ+FCQmCUcPMwcq55HWFweXWVFliuOiSuON1rWM1h0XLmSrndFVi2NhxltMX8H5pwgvyq3cEvrZ1L7WGwtT/m6A21Mw5Uxx9TUWXj3H6TchT3oT9V1cr6HPDpqymeMxO0VdPwtvbvd1yY30PlQn4D/19RZcInR7Whb8vos7ks9HAZ43Xn251wepzNzfYfvT2vfNmZy9TzX2Kik5KF8+v/VsrnvlmMr42t0XYT/cuuyZ/RJ/G1pO2UfYUxraTC7UQv9xREOhJRDHZxjHf/fKvcq+xE/CQ0X5hqr3gNyzun1iy/Su/STlFKaQq61NttHxigNaTaF5KFj8Vkf9GfK8m3hDMk30q6f4x+2YdC+QRm5lq5A51fK8AX7CcT7h9+3U5a06ewHS++h/fHllTfL0TaVCCNz/VD+OXfc+LnTM1dbhCBrRz3NJ7GnLeLQdB121qSxnxXyzUu6gTY2p9TfeYu1TOrdyXchvSPa1zzknj5uwDd2fgtUF0LhrdhnPPv1T87QWGUbZW06A6mv0anOuf/Yq3taDQZzIBD4LNAG6lxEJ0OpWMHnnVz3+9cVvjWWLEp/wESS+4ntQpl44TErEPZCiVu4aLIdZOLknkkrR8OKjFxa6YAS7yJRmvj1RRZWLNYPwuQUNsnpyrUD23+E39w9T9sMRp0kO5OnfO109F5Oe0enX7L9LuPvQCAQCAQCgUAgEAi8TbwXA/O7d+9SSv1QpHtsEJ5pC/GG2mIVPAWG10fykNH027a52S+Sd8HrRcGy1dO7JS3TV9gbrIxN9kT+1h53DYwUza6kP3e7b9h9NJnBu0I9k1IZ1vukxiHNOCexuelvyVhhnUKrsmQ23oglecas996/i1631yDm7VPUC/36eo/hV85T2Bs7pvTYXDTOTyIGoZV4ioFAIPCpgTL2EtI1UkLzhoOd6r2ekuwo4VgQlkGeY+NJu1A0dqP0G92YLq2wgc3ySZmPlMGl1ZjAjzjQHmGq0Z12Hha0lsar72q4kldiPavlGvrkPcn13Y9Wv/7mL76fGKitvnRnDlW0a6wCs+5Iqez3MeYFPG/oIN6mr+Wur210zMB6ZIFxKqVKVirDrrUzVONxHM2ReECYLeRI5b6Ntg4rZeobGiu01jq9a+vbw8/I9S363rkxUC3XkH3lPpVJq09jTq8wa7XnwG3Cse40BuMVNrdUv1XelbmD5sW/r7Kvpb7okR2Sl1ImAwGW6ZFnXUce3r1Yt8iSHb8pUZdw5qP1tuvAFkZtrhEvuLo2cgDqy8sLWw5d04ox/QW9TNLBrB02YH/B4xn3PFQOGZTpzLNttTlcsq149KQV2S1dxLKvWOV77RNwwL3UX2n99LfEIvaMo5KtExjIt2ZY1fPN+YGgNo5D1rdJf9/HLd+YFAzmQCDwlUbeTsUi9YE2vzsH5XwyLyGGbCoptW02dOCW65gmrCoP+F5IzpVa55OpuTSy0teVOEkmbvGjLQYKTMjn3yPP+TQFxruwk9JIhjW8vZKbMK8sFLWFrma00BYOUltzizta7r7vomLPYcXwxbUR258rvde/mXYCeh7+9LRbbguHHnN5ZDB/mEVWIBAIfDz87z/+q/Rv/t1fGa796z/+vZTSaRj5W1+klFJ6OdWWYwgHceKozYhCjSkp9flwz3gemkkYbQw+b+HQgW3sRvmwsZnOdaycCu7zjj8t/ouN6HQRzaZnMMo6O/AAnBHJNF6g+r36g2jcMpwkmiycYxIbGbExhjOsrOq3HgM3/r+me11x0E4qOnoGLdwQV66mN+P0Oc/BdkYnynkN3i8Kz9AMjSm1QzoDgUDgU8FTDMyWd+Lu4eFjkdDrdBZ4lAmJJ2ZrobxalzahHuRWWVSWVmXAE9omxASynr8Zi6i+anncBs8nTNSjl8SSf/K2GJ4fzji27+OpmFJd0m+LhYPrpqzwZOS1ZPLkk9LMhkg9PS1ve+FPq8W/V42kYRAKBAKBQCAQCAQCgbeJP/vLH6dvfv+70/U///1/kVKithy0RszjOh02j0hkClgVtrUncYDRXTmQDq4D29gDa73fUYe/dNlsEURovF2PE0qTV7Kz0Hqs55PW7JZjaJWAcgUS6Uhj+HIyt/SJdxRJJCvLvrQKyW4l9QXuHVvsfdisJL6v0wkFtqkeSxxsa7r9qLTd6GM9X//uD9h8HgSDORAIfKUBSgywAG63I9XtZONUsiUHcXLKtG10njREZoPC1sD/10KnaArAVDq5gB0wtC447AGzlADYIWdtyee2GDYmMRLoQExbnMbDIMb1c5N3JmWzbSUoQdp75K5rLB5JZlyGh8GsOXdoWvzs0jY/DyTWDiePdG9W6ND1TNOMv8d2vF/72i/8+8vP8aEA8c8r+dbhoI/2JMjRe7vdnZ9bHsebjeySOJg+0eo1wizha16FnSvHo2x7FwBwKE9KaDyhZZD+wMEKf6V9S57nsRhp2j2pHo0BRxfU3CKU3pMWtBjSNl8Nx86Xd6W9PIxAiw3Itdv0XFtbBaWU0LbR1AkS7Xk2SHq+l5PU8lKQY/2FEBzaTmx+sf8hsb3cWcs57en4yf0QLghVkbcyfU/36+NqsY1VaWshN2BRXtPcx2qt6d3xLrUE6R46jRp89n1HYxkJd4hQj9wPJ6Jh8DY+pEUmZxdgLuiwg6uVl4e/98d/mWTqdXHGJGgHfoyb++oc0oD/buTvG5fL5dUOubP0J36O7ffod3UcBxrDe5i++f2UlMi5C5o+gOvHYzl97m3bXKF/aH2QXjMGzvaVPOnpOXXykKVLSfpxSuhMEqTj0vGoptQIWS/bRotoBqO7UXddrwsEAoGPiacymOcJtk8yzeE1DcZ0wtbrsLwu3ALkqleCwlp8cxNsXziS6y3d+ffgZfR6VgbZFowEbBkOYwtOry0iOpOZKupQ1jixer1yuMyr79darNYitWOdThn3GJq0+56FjFiHIYvkzeteacMAh4qbFvrnn29/9/3GJnwEudwXSvutP3cmyl5bUe41HZVfpB9oXOMWSWfhw/WUEh/THBZt8O5a14KBci5LNQ4dKF9KKZU8LfL7u+vve9/5fqcZR/B1yehVa50WRe005u3lrHsO7YAXXd0x0GOBU3lo/ZyBRTPscs9L00oeZS49Z0iixiQun2YYarLAO27f3rk4ySWl00gEaSSDHIeadWOZ1G74nZfyOjxH3rocR4EF4z1f32EDBo6jP1K2DWKBQCAQCAQCgfeHSVduJgJBr2xxPQSdkSTfiGFdcqZK+j0uQ9LzOedZSlgf52WV2K60nkNoE8uh6n5GCDsIDucv+DjFmsODA+fw0eTWoDl2NHAOLk7GVTuYZCek4X9aH6FLK6Me3uE210PXfNr5GbS/7oSV3WwGQn+WIhdIxJ6pjTJxhD0BwWAOBAKBQCAQcKApgAdxrB+jMnzsM3uXskV3UFDBeaI4Lx5R/C3H7fAcguOVq3eF4SmSBBzrIc2Zb6WV6n8WPAzwR+q02MIeljDN8yi8CzANtB94Fl0eRz74T9tWUSob3Eeec7pbSfpuPgbaM5ecttf7ko1jgXKY2i8jwwLjzKNjmVVeKoiZmvrCd3pPVe+TnGFCG/f0sWibrnHtxLFqufIAnLO51mOS3QI2omjObo197AUn2+oYgZ/b005SWwO458eOdq4u6ojHBAaNDIHTcHGseZLUmBfLZOc9nwf91g5Ah3m/z4XCvPzxh6FAIBBYwnsxMPdJY743TV7HPDlxf9frnpVvL8vUw9Tl0mGZvQyylodOIHSiIp4VSanMGR+7oUNq6y3xz+vZprWukMOkD9678S6nbEn1e7eRib8Js44ygzUZLKXNy2SWoC+owZAh9wnur1QHvYu36E3f5XPWq+8Vs0Ejpxt4B2Fbe+uGcmiJ1b7d25P37mplasxRT53HfkzsAM6bSp+1FHu8xLJLBxFiZV6L4S55eNmyaxp+vzuYQ4eUBcuKQebqvKNvOZUXThK+2F46u7vOLO+U7gyOFkfroveZyqH1eW4RBycTl/aJyQvnSk6xTlU+bDIQCAQ+Z/xbf/fL6dq//pN/nlKSdQmJxZfykY5zzfVvf+fX1Hr/6g9/957FsG61ei/qgpyBGRspV50ZOc9Ouns69qpY5jwl1ZSSTyapzLt8sx6+uqa4Ml9K61rrGmcwvjsA9PKlcr3PjXUQusartU76Lacz4zbkdCPW6XBRdnyPM2w3Q3Pu+jAN66K4isQ77xNTO7YdyXKYpiGfsDsWnO/wBqm+KTkKuPBQls1Iut7L0tfDNP303l+Y8CZMOgrvenwCEBlOuXcjlMtcD/RN+JbG0IWajYX+1pyAHpkkNq609pXeQXHa6Gj9lnxee6T3ubVyXl/HXaAwwNJxTt7tSsqEebA9K2kbuNruP9+YEwzmQCAQCAQCAQe6wn7/7VHLOGZTSvJCCt97xPiuGTW8eTU86ji16pUcR5rjWXMoPSobJ+cj5Wh5Lbm1+sUFH8NM9S7QtWueg4esxZnmAJTK9TjtIP5wbjLKMvQLUB54zD4BT3ogEAgEAoHAG8BTDMyzFwsUxlM5O6rs7Tz/UM+EzKAaD5J4hHnmXRh4PL9XZfCWDfFYeshW4jlq+dPkqvce/NQUfPI+pQUNll1a7EjvR35/vEwaelm+5xSfx5Wbf0erjHWtLC695EW/X/CV1UHjHOleaO2JrjA5PjR+9su/N137V7/zP6SUUmNetsc4UjogfhhhzGKs9EtgSXOLZW7bHvx9tG09zH2tz6x4fz07DDgDBmWm4PKksYRjOWkGJw2edJcMOUz5LFtGGA/bfHjM+VtsaicTSpKtjVlJ84orDAv0t3Ej2iFZ53OkitjVfHmlyPcCgUAgMGIYJzPoFygBzJGwBEOs3qvHhXHzNicTx0jG4Thm/fW6k0piOms6m0efs1iFjziXvM4+Tl/yOs+49Fy92lru2dB0NG79xLFVtRAdEkuZ1s89K/7bQusoda2wv+n9QV87/2I2bZdlKFx8rveJn/2VX2Kv/9nv3XdQFKZNU1IO5RUGH5qOOxAc/17R2+GAZckpCXYDWpd0ADqkAz38EPrEVTat1yHutblQ0N2O0D4rbWo9C33/0jrXG/rK+96996fICWouDvyByJa8UozplMYduSnJuwOk69N5SLLwrAzf/vL5B6sHgzkQCAQCgUDAAWpEh5AfsGWRWxjDgoke8kEVbM6w71W+MSznkodZStNahgn822MAaffAab7gtNNko7JYZazkWXHgcGmuLkJX04h5HEa2FSe4tRj21GNtj9bK50IDTc9RiMGAO4AKtkkL26UDgUAgEAgEAj68FwPz5IGrncE8KX8mY9WKq0uZb7KHwMKql0RSxD0LMen+I55xSU7629p2KJWjtU9jawoxcizZejl6evwM1JOzuu6qxbfY49rJu3i1FmsSU/9Kv5XqmMqmz03S5bYdtBtK4N63v/8jt1xvGdt2LjrrfQiEHQKYzQkGpNZ+u93/scGopVOYy1o/mvt3r982tMwHw3DxrmbPq8xv0sYk9gATYZyxmEaQxjpReEu5x7cWJfMzhiQ5VuAxXmmMpH6asc1a4pgWHGNgmmPgQDkYBwyjWTcEEjnQ/dsZx+1lezmvQf6cIIRYk4Nw6PJW2G8kEAgEAjOOeus/TnYwZtK1cCCQHs3P3jmtjddobPbkzDnPjgwmY58PBB2B6KmlkvtSeXVm63I6E71/Fat5PfqPVe7qPMmtSaU6/P3j/pfGx+Xq5fpdrT2OMtZbqNNo2zZWv5RYkbyseSrD6hMWvE4uSW/LOc9nL0n1nHn+8k9+937xrOrv/Mf/hUvWZ6M9q3EYKECyb1jORYkBi3/TGLW2zWX8u587TLmY3inN75den6N286BtQsuTmNNS+pZv48/BeebaRUpr2Yyk31J6i7Agln/+tcaE1t9Iv52fg4TIahEDdLlpv6bQ0t3evUsppfTyMh7kS8ctKW51LlRG8g3SJ5zSPx9PMTBbhtnKGJglzB/Jmiz9BGGQ6cakWVvMXvlQvSyQFaOmpzxchlSXZWCmJ0hLTBMNqwZzSEcVjnlDX3deVGN2Pt7jh/MMxRRj7i/yVi+a9hAOXLDaXDJU0S1DJRf3t/KpILcDHc8+3SaUPbWBmRqUL77rch4CAd/PcetbYaTJ4jjmQ/qa7IpS3P5q95j8/TuVn4P7ljUHmfXtY+Wb+46kuJ6cTAXGAmKo5RbFmQk7MaMvfts7SfQb5RdtkoytPY6uKDT1BboYKU9ra26hyCkhk4Ow0O9dbw9qYObS9oNCiNz5SP1QEXgpY18rZUPPO1X/5tEN7v45muZdWehruoDnm9PK1uDt7+8TKw5409GK2uBZ87gX3oUUvuYNdYYxtVMWrjtk1MoXD0ByGPWkRSBXnnSo7JhoNGpOabk2oOsOWv8H7h+BQCAQCAQCnyoiREYgEAgEAoFAIBAIvHFwYXgAP/V3nxNLkToWcynpSGOcyMFJozD5NCYnVyeX3guJpa05ZjWnO5btEcealySk7XR6pM5l5rrSXucvNj9mBlssbO7sB4/DiXsX+DclQ23b1sJSUVlxOswWtpmwMgkArhzHIToTSympbJTFylY1le9lzb5vSI5R8R22nW33v7d35863k7UpjRGUOME5W6X3NDsl+fSU0EPjQEsM4u1lY69b451EzmmxnUkfpjGTWzmkPAmrjmptDKVtAqDfmEXmtGSRHNf0ei5jmwGgLYFwBOchHY7+xMktjQn0HUE7SM/P1UvP1qFpoUx4/3Advh1oA4lIMTPyR5nfB55qYJ4ZAtAwV4+V6ONRayTST72MaO6al95Pr0udTlNcvGwkqw6PkrAiJ5tPKPcK62WFReSpE6c/jjE4PQ39QBksUpmPwGLjSOkBV/rvw21MfmqhGL6qgCaCiQbHZYQwOweZkF/yhvLz3yOexCXFhFVIGYXKUs65evE4ab3XUgpzrStLwErVZFgZC7lJVfterGfUc76BjgAAIABJREFUxtuhjY8x3w1N/lL703JTSintVOHL7ZATAPSZHX1kPexFe7CpfEmZ3PfdHLdxOWzbQEgMOgwwC9ipz6Q+XNBnxZjznTGGc2kxiuupD5RtrOunf+HXxHLfInrb6nP6gZTttmi48bGXtfldSqPpNlIZHmhj0pXyJLQ+u5BW+r0yp1+R3Ta26GPHM3SOS898jDod9+TSYUYrMkmLZY2NbF3nZJQMF8N4T7YMjyaN1MbfivLQse1bX/6mKdPHxIdg33MbBPtGoVlHrKlfo1lrlcct6Tt5plEVGz3xPembsvSvdVnm55Lq4LZ5P2P80HQzLp127S67vB3dO0dRQwzWSQHHcbBzkjYvaUZqLpTFIJtjrV3rHMJtkEko3yp3LEPJ+7wlbCAQCLwXBIM5EAgEAoFAIBAIBAKBQCDAo5EJwVgPFFHK4j2TQZi0Fn7xfr0Q5jINkUkdi8DefH19RaKcDn3DMUqdGVAX/IWyqXMC8tH0UB8tl+4useJM0/vvSCzeFlpRqE/iAdF6ZCfteP84gATRCVW9LHhPI0Fi39+pddBnp0znF8JAbu+fnLHT2va8DmcNlKozkffWH0eHH/3d84/vxGIuAyZmNelrgHdnG+cXdH1yzo9MdWAcz2x/ykyGw4x5L9S3vvxl9vr7wFMMzHQrwZbn2I4Smw9gvUCLwduvy4f8rbJHpYDhEsNT8+BKAbsB0vYLCovVt5Kn3T9GWacBzPG8VnxmU4Y8fvB9cOM/WO7aQX57WeDSYMCxcCAdxzTkYLVhhyO2oIDVPD1O6ih7mxzaYNgH3fe5jeJjYG4z5r2cA3qGSWiDyawnkfoSZqdwLA0uPZXrSszz9pu5tspMozHOe1/GtfAopZgHLtRaxRjU9P8eebXf+BrHatKYzNqcwx0ck9JIMKG5tbLpmIMZPdCPuO2f01Y14VAKTlZcljZPSyzxUgoaM9J0LyXa/9fHuEAgEAjcgeeH/J51szbfMPMlhsaGPRS2sHe3Pz5TpTAUTo01y93PObPzLZVP0929cxg352sH8eJr27ZdZlNz5VlraQ3j89o6A5aDq5ca67DejA0m2rvAeaUD2bhy8fvH9+jZPZbNAetjlMHMsa9hrVdrnoxKHDX5LnsaylDO4g4EAoE3gWAwBwKBQCAQCDiwkwVnczIS1oW2Pdhydmq4Ypj3GBeshbTHqKTJJjrHBRm5vCsOdSq3p61XSQhXYRnDrHvS/el9PGEvtWcr+oqTm3P0pcTHAnyGw532L2iSwZgIh/waB0cHAoHA5w4tjIcGsItDrk3QSyRyzcTSTPI8L/1ushxUmhGSLkAJlf3A+FFGWg4l3FCnFp27KWmN5m/z58GzbWn9EpmJEuo4eaU5usuuOzknshFpcotQKemv7foxk7fwffpbagvrEGFLTqmdpBCIXOgeiQgG96UY3SgYNyv7xyD0PMXATD+QkqGh/GWIH4egIMsLG3nRYDFXpU5nLaK0BZHE4qP3pYFUYtHSgRZ7ccWFnDTwEgaz1B5SOwyeWyGNFA+WMnfnCcZeULZ7ObFpVz323nemyiL8pmWupmPzGpP89G1N7UHamOkn72uh/bHQt5XAO71f38ue9p18D8ykNY8BtPyZsdL6dkbfReYnIu47xpAW+O1+mmPE0bz7vouMD+lZuHKw3NyYoD2DVrY0zlpMGXrvkYmVY75wY7I27mksZWnMxbsKJOUHg26jy+jgEWtcx+/Kw5qn8pRSUi08E6wU+Celtrw4Y5l/tUaUQCAQ+DD4xi/96IPVBZsJ930+sCyVnCpzxs4016Q9ZTgL4NzRsh/jzpahWGbeTMiI0U1DaJ6HM1jqPtUvQXNceXRv6TwJTr/ndDavg2llfUDrhXQeJjZmAWvl3fUK0EfgHjzXfFBfzvMa4v6O5z7ArRW5tSunD0pr3Zxzup20X9jdtdX+/sBog416uKwm0zbrYJzjDRt5JpnaDvbehq3+fExrCVoHucFfDwQCgY+MYDAHAoFAIBAIOMAtlFNCxgaGxUDDEFlx+Tz1P8LqxHklZgzNq5XryWs96xW2KgfLyasZWCxnmsbiXXmHKwwZi4mlyrnQlivvXaqPOpiulMEZ+aS2payelHiH8Hhhvv4x2D1vHZQZlXNuBt2h/8I7OpDDWelLzyAraN/jln19WXv/Hqe6lo6DZ5zgylttL8nx7jGY33+PzM5a5cOcMThDr2eseuZ9qb6c8+TRxnJSVirGEKLDqJum40KuJDQ8TWSFMn932nO9NbRDQNuF8b71HFKoSYkZ6snj1Suk+xIbtelygj5IZZ7JFzzpjpI9pPKaPBMpz6dbULmos4Vj10KZc9jHMV40/TyOY7yw5ZE1DaSuju6sGp/hGO5DPaXJoYfglULR0rZd0Wu5dLQciLkMctKYzDgtxFiWyJVUX28M9MyzpJts+cPH1XmKgXlin7aJKbXfUiNlMc6nb1KdXzR9cX3wkcrWPLs4n3eQZz2agrzeBc2KJ176OKTnphLQwcNi0HEsPwmzLPrz9XXSyILDdV6FZ+FpwTIUSO/fayTwpFOVKuZvne7r20IspsWniG9/+evs9f/tt//7PgnBPp5zAIeft+NIZRsXVbRN79nOgX8nccRTV3aBxZNIfm1Bxt2bJsqXbT5I4WUc7o/jmPov/m0ZCfDCgyolx5FSznB4xv0ACGpke3nZmP6NxwW+XphfsfKR6Z6rJmO/7okjri1atJ0L0rh+L4/WO8dXlpg3nEweIyFsGau1NsZYQ+EVkHtGYBcT1nHqTDa4tucdKmsk5Xa+xJm25pTyyW7+5vc+HOsuEAgEAoFAIBAIBAIfFsFgDgQCgROllJTTbCxNiRgBF5hZjxrnlxhr6LrkecflSh5dbUslZ/y8woy0mC5W+2Fn2gpT7ur78ITm4H/z3vl7Ot2rzBmBJ9aLWf9Za2PG8I6nu2TQRs1lghyMY7sdGTMsaGgNcFKWr9wBoY35cf7u23zBGzqOGxjUeQvgnBiSY4NzFNE80m/uG/Cy9bhyrTEC8weAeNC3sp9pHfV5nCw07TPyUHBOJmkM9DgCaTtiZovk0Grl17HslPoB20flD0fmnn3l8FcqA/1NTzh/Rtm4PAp68GlKKZU6kkxeXl8H2W6nYwz3zY/B7nlL+Ov/6feH37XWNr7l4Ru+o+D+5CA/PMJg5ogw2lxckLMZf28a69k7Rz1C+PCWa42tK/Vzz2rlnas7Eu/s5xmYK+SwTiCydVycDuunOJ1Grpr6ItKROb2XI1hwT2aNV5SgdaR5vm73MnoGPD4xdT1KsHoEP/vLP2Cv/9k/+xcpJb+ufdAD5iVSVKOqzqzh62TE8bo0p9EYxXSupjq8NdfTvxp7XpO/yddigzYBzgRp/AvyGMOTxebnQOd7eggrkKzaM2xjPklXk+JWT9eNcUTsTwI8xB/8G8LvtDBQLRzUHZS1TvsWrmMIL5j6s+7Qz+gjwqHp5MY3vs9/ox8STzUwzxNjv97XtmuTr5ddLOV7BKudEUParkDTSlsPJBZx62zCQpUzLEkDW5vcyMcpKW/SIMAF3KeQPlg6OFsDhAarL1iLb+4k6ZTmtuYWgY/IvZKOy0O3LXrz9b/3660vTpPIx1NkAoFAIBAIBAKBQCAQCAQCbxvBYA4EAoETOecpplb3VM9OHRe7GH6f/6Pe3UfgYbJ4mMjaNclpxHmzNXk8LKEVRxJmpYCIXVbsvqcMA9/7o4A4Ylxz0FhhY32aA4gPn4EdihbDgmdKovqBMbrQtlwfr4TB3A+oRJXl0UlXSklfMQIzAjAKxv61N8Ydaq/mxRuZDRwLlUJiZV2B9s1b9aywp9p9nAyINk/oD542uEI+kL5Vjelo5eHqXGGwS47+eszPYTFuOHm8/UubGyRHu4ct7pmXpHveOWyoh7ntJZN8VTGFUZrmpXsb9TBJ5/dQS5trGuP5qI3hPNSh9EnuXo8e1okUntjex8HHZOX6udafHhljvHm4b4E+I7fTy8sQlZiAcG/l25vv89/vnTQ0PwMXTpB7nlU9lRtnNdZ3QXMwx2CGeKkHozZJcyUnC83T1wNo7J3KRWUz6sPUdhf634dAewZjLqk0tlpnJQ5/K/Qdpiy6a8a7e8Yzb3O/KbHuhexwbUQ1KHcHVu8dx7tbOgu458tC/vM+JbhZu5JMJjTcPsb+CPLelMNXxbNEaJmkv1LZKFFyNbTq9K53nhzIkSBxPpqOXrf6UGNUw8GhTPxqfF0bv6HNDjKuVlJ2z3OMf8na4i3gvTKY8cvNTTGBAR2+Kp0BeigThgZua4xX/lXm8oqSYS0apLKtwQSnt5R96/mkbfMeeD/cLtuKQsPX055LGES82xyuLLBXt7VIAONlBx1MFDnFMqU7tE/xsmKG9Oe68AoEAoFAIBD4nNDWb2CkTPiEm/u1HemGzVlW5kOdkrEO442/sjNHc95IpgqPLm6t6bxleI0+VxxCFrCRFhurrTWnR05ch44xRM/dSE4P6KqJrkXw/9k13kWDKi6rH0TW3xMYk0vJ7VwoTiZvHUmRlzOmr66vuPcZCAQCbwnBYA4EAgFASS1mTG5MxNGrPnhDqefWwaY4MEuBsEITUq6Fc+vOZIIzAC/2CGNrxdOtMXk0pZm7dtnT7khveZzv7+o5DjNugeZhFVjvagT0LShHZ1nBbxpaqbx8gSVl83cH0mxMaGlLRYwfGvYJlVfGez/z/d9gnu3Twb/79a+lP/1v//POkLghB6ARN1tdDNNryrfl/V5WFsKPLEqvxNNdgXdckdJa5WrP7jG+SPWujGuSIYkygjDa6e6ZsLyY+rzvlzPeXOkbEzuPGvsUxjctQ/tuLBbZcI+ePdCKs+exqwasQCAQ+NwxzWOEnQuwyF7SXMTNj0XII83LdO44Dh9DmP6VwpW2+ZrEbqb3AS3d+XtvbG3CaIZ0RZ4nuev0txTHGOSAHbYFOfjgGfbmnBnLWD1nBTt5rmB6JifZk+opWix/rbxJ/qL3Nekd4HCrtE5K/Jv6H2Ro/eHtOZzeC4O5NeZ2ek8PdGBEHQegg2zfzU0bbJ/bUKaX+fs+YHl3LS9sSnJ8X5qPdjxvh9XkFQcegelL00kD9srCsxoR5q3BUfVqOxdK3skN/tKtCfu+L7PXrYnTy6q+gqnOaZC1DVix8AoEAoFAIBD4aiOX2n1hx3N0P5fOXLsTxesY4tZGHmf0s9aKPiaw7LBPSThQDunp2rpIc5RgY5pnnXavc17/edrKKp+TSXP6c89oPeuKoajWHrYDbBVWXTivVs/BrFO58lVnGsOIDtZyIBD4VBAM5kAgEDixbVvTr9sJ9Wd8pwqKYWWcMdRJk5Ca7mAAa9ceXiwZThuuHA8TG4MLy+NRhj0OJI3NQE/n1RhzV2UEcPH6PNCYBlb9+Bmlcuw2o3VAO5xtthXkLSdll9JixeXm7E1nvl4eeM+ps+pTxZ/9+K/TN77zQ/be//UH/ySlNDs28wbjg8fgAFn97NdHnHxXjCsaO9XrCGbHNervVB7LYhVpsBzmV9qAK/+ZGIwqdfxbnxhd7+oYKOWRGNgaIUB6l5pslHnElkGuVZrG0d8CgUAgsAaqfx+dWnj/s/uZySnN4z/HmKXzvJcISP9KOguNEwx/b7ebS3a6bqBzF507aRvS3xswjM/HvFlhZoWQqVQeysqF58PXaB6JAUzTS7qklzhKUdptWIzw8ljl0jjXFN4+Qs+nkfL1ikE3wYWMedvOUXgWSEs4uN/6wZes7G8B78XA3F/C9bxSB5Q+kkeU5GcrlZbH3ZNW+ugB0nYLzbNtfdQ03wpD2dpq4IUUjL2XC4NzSvPhXX7vNZdvZeEpTRbeAVSSscuqeLanzON9iw0NpfTy+An3Z7/8kSrzVxE51wQ7g5qh6IC+3Q1zltECM0boll3vdiJt0rb69L7v05anbvwe+65WFwY3ntAxgitTel7M4JC+G04mnMYzdlNFTsPK2O3BFRYMvrfvFbWjrKzBNTjwZK9rjBdJGb6H6JDGiZ5fM/AHAoFA4O3i3/nOr0/X/uZPf09Mz5Ga4XCiVOt0eJmGYY5iHNsY3HqQ6q377SdTPm5ewtekg7QkrKwVcPqr91fmVc2hndJocPI6ODX5WrujNuTaR2Npc/VwuqC1ZtcckpyRrP0f5fO2CVdHqx+tmaf1M9qcPK/NfHW1a+/ByRkIBALPQDCYA4FAIBAIBAKBQCAwOWZrrfNB2qWmmsGA1g14wAfbsP1LcWK3/yMWl9c4TdNt2+YyTq8akdVwBkJeLg9uT8/OiFUClPTsXtKAB9qOO2t3hOfdXJGJA353WrtyhC1u94UX3Dtu5Z1F4TCYnMOEMza/L1Lcs7CywyglmRkrsY1XwnFa7FipLS0WLC1HqpeSbiz2L5WHxun1PjslJVrkD8xUTqmTVLjnk9jWFlOZ1m0yfAnyecZRe1OOsZmTQ2prL9GJpm9s9jrG46ZtTlnkhYRd5WTeSZu8bC9Dnn1/v2eiPAMfxMB8b7DHBsTrA6rtpfUOICsyeMukgeBpAHZpkKPpcId/9DAeabuAxi6mDADLwyyl876ruyx0UBAeSIDXW8/1AxqX2arD63Hv1ykzfQ4GL8lLr0sMZipjV/I/X684Hqtau1EFATGYW9sf84TVvslFtgtA7huzYizlxfk7g3m8jv+vLRwopAUSzacpUh4WrzQGehnYVA5pEcHJrskjpZPk0O5pym6/Rv+eIVyOY14Mb6+9PCoiNBvagtWU2jNtRUn7uHGO8cBEOkNCfOsHPxSf8asIOo4e5Hpry9L7Geygqzv9puFlwO8+917RSywF2QOP0eHKvCaVS8ejq6x/K88ji3JtgbtSrqTvsDtAyDVpUeeR5REjjQerC0Zv2lVDRUppYs+2HUnnTozMMSkhTezACAQCgUAgEHgIwWAOBAKBQCAQCAQCgUAnsZwutJetpBthpB35loD78Hr+59hyOm53R1k9TodZLT10pOI4LhXSd5bXgJEEmlJO6WjByol3VIDHQX8/jBwc07oDT3PI9f/3g8G30/lbSpnYgynNzMVaK4QYZUMmUgcaDv0Az3AcNcEB61y6sYyRPDM6gLqDjd7Hji+QHZOVNGcaJihJRCZahsf5hNNwRId2/2QD4vMTgBiyp9reUws7xziiBsJEI6PkmdCTcJuP52HkgljN8LfWdmhgkzfnuzPsjTGZIU5sbWIScgekgz6exnjGU6uidkwppReGJHIj34vkgLXIbxsZb64yl+lvSr4DUHKgRQ5o39TZM2772WYCO5iCEgdX5KXfLpWVPov1rF5HdCtfGNQzjM0HX4/EsKbEToihfBxkLiF/u7xAjhhpJttJuqnnnAS7e7ZX2idhDOx9Dnp0l3l8D8dxf9/AMM8vMunwreApBmbaaVocyB13wrGR4G8pIx1/moQ2MuQcY0ebqfkz+1buxBZbwWLpJOF+TWVihY5pm7xnG+WzfTYI0AQfMvw9q5C2OeABl7ahyBwSPl4IoD4F7FfiPdM00sFUPZ4nyCDHZ8V/OYZim5QbZYUf7KbBhChEtM62zWGm+ra/EzuVlCHBz8iZ+5SuDNtt3tt0TZbPCXkrqeSxX8BOlNYXc25Ew3atH37e/iMd2nC/zbNycb+HLTF0EreYw1CnyHQ8r+/vbo1a6VHqJUVIgjXuePJw+bhFhtT3K3oPK+w6bQHpAR4rrYMwOLm5suh4BnNnSmWa9zgltb3jMo9/feECbQ11bhODum//+rzHj0mZzhBz/WxfpF/k9g2Tvnf+7O+437KUck6Z1+4N1S6wXz2MWc+3JfZ/2iQOEukKQ9bDyPbWt7J7gZNFLK/tgOlpwKigjW2SHBrLXZJtpV2suYBre+s9ePoXRRtn0VgEujfsxoDuRMsavhtD1w8EAoFAIBAI+BAM5kAgEAgEAoFAIBAITE7nG3KaVmTEp44i7ABvrsmcm2NbdXag/3udJFfCqEA+jY2oOc+6c2cuX3JCcaw+TnaNuMM5EDlHDcdqhtfHMnjJM1Bwz6uFYpyJX3qbaCQMmkeqQ5OD/l+7RoFZ1WN7zmxOTibNCUf72NCmpacDItPXv/cbprwfE5MTzyBX1NOx2kgTQnsA6QvyYTaul3ks3YdXJbFzaXqLMWwRaWgbrR6M7SHfafXTcqTfK3no9+FxHuP0dCzxOq5pP7tKKqBkMOl5upxA/ji/d263DVNOkwd+Ysd4c3IDsWDcSfLNX/yB+gxvEU81MM8f5v36vYGA6Xd2SKB3AwEBBqKDZ5e2OhoJT2cy0+tcntWPQQKXf2O2cqzUAewL7rRmDA8zxfrYpnQ7H99ZqlNTkqyBQkovAQ9o08df+bZeHbykOrlrTQYjj5fZTLdg4HqsNpIHcV32bSOHAHzGzMScazrSuG2xbG0gSymldHt3SxXo7dBUZOPEkWob44C1CIy0Lc2s+t4/ukL1/727n4YO72XLfbLitkpiaMr0cO8YGasH6nfSJMsp/vQgCtxfpYMP9n0X4zPjxY+k4GElU1vsSOVwiwDunjR+4GfUxvlKDmOANi6ltHaDLYJQNu4fbWvcee/d+dwtDnwpiY4XwC7OOadE2n1ixJfadtBA6PfWZnlv8m4vZ5sQlvPnhm/94O+x1//89/9xSgn1HbyQPv+WMvYX6HW9v11n5qbU3620CPIssi3GNJdWqmeFXaumJXOStoiS5lqPQcIr66OY5J4WMH6dyKPjamVaiz9PPVL5K6z3lf5syYrRd+hBWfe0P/Pl2zbWvEW0dwavY8fXzvdac4/vD/NJ9c3D1tfl6b/eL5QzsErpZDkOt5GSlifVz30XkMxjTJa/I58BiX67WM9Jaf6uuffJ7b7hvmXN6G8ZmDnZpXJxO2kopUznUOD3hP9yZ3oAhuenhkT4U0qzZ3S9G7VNmt9/IBAIfCoIBnMgEAgEAoFAIBAIBAKBQGAJ1LdeC7HWC5yERhoxwno2UsUHtLhLsljO1JaO+pJJ+QcNA3uCOjAk9u4hyGc5VGisZc2hLbGuKcHGqtty8ntlp+k3g1U+kRHBscMQeu5/gbxEyXjn855eIpCSMu8pfGRW8sxb4a5+UniqgXlu3P6SMvnKOpuK5M3jfRqYfAJ4zQXvbD52s7N6mSJeDB75B8u2BjcPU1b6qC1GAefJ5uBmSyrwsoskJkBKqQd8N+qwrnf2XrtApWiTRmdP2ywCrU7rYAJ8j4ZntvrIlI5OEhscMDHGaP0ckUttwxYE6K8wG8HWlS2leuiHB+B7dIrI6N4c5xy+vdzy9UkR/baUmgXmIP5/Ub4vDhMbdoElmHMWx3fMmpmUA0YJ1Rh1K89x9Z5nHG5yo7nKipOP22GjY0SnG3cGMvnG6z3DUBegjTkFxZQvpM3LrcUU/vYv/YhpgQCgjafn6xnCLRuMz97H5fIlxif3HUvnJazoIY/oQ5xskj4hbSNMCT2Hobt4mNn0+qOL1NU5fsgLfYR8//i9cbtCcLlX2Oic7F69SEsjjdEcC5FuaQZojEAJbVcMKvMgcq/qooE7mk57/i6lJDiy7EA7Tfdzh1eB3VA5JzgoribYDTmrldyYhHdstk1ihk5SyT08v15Zu3ju4fHaswODYxrXWlm9H67Briasw2iycc8K/9/3vYXIwEY6+p1u2+YeH6XvystW9n7n3G7VkVU91+tdB+vsdPmehJa3X5jmORz6AroqK29pN3XFIBAIBN4ggsEcCAQCJ+5hC5pf8vw7Gn9qTe2UWBr+ojnIahmUdij7nueYjPyckr6yLVBb0NPfnoUTV+akKAtyS/c4Q4l0wjKux/McmhwUK4YT/P8VIx1OuyLLyoFqEBII19EX2ZpssPg7UyEGBYTB6E622J4ZCAQCgUAgEJAxO1tkhyP+TWMst/KIo0Jbg0gObLpGmpnASc0nlS+RIMF5LJFGJp0/ERTfOqOVV3Q5qRwWqUX6nZJOBtDgIQnh3xbjd34XI7N9JyEDOykHyiHvrIWmOeVL4zqqkH4H7mxpbUxJgNAp4HL7i9Lt3ds0pPmU8WYMzL0jgVeTb17K+JM+Wp8Xdo3VY32QOF09eLkkefZE5addkff2ch+kd2CkgI/+tdDtAnxberzoNM/M2LzmmeWNSzD4ibmIbONg56kL8u9JbxurrLm/gkzjQNkmx4QPThmZQJYB7CBdqTQDEy/jo8yuQCAQ+NwwxRNHcxud5Q6i7+Rz610+ZOeNNM5nPIc2lnwa0h4XNtpZczknI1W0ubm1xRQ/F2aQhjJ2MSRGtvSbu+Z1rGlYdU5J9+j70RjfU15mkS3WI7w7zYEm6Y/a+5e222IZpTaWmPkpJSZGKSmXMgZTShnOd4HzXvK4KKaLywCPr/3Cr7rS/eUf/E5KKSF285FSgRPlzv5Qj/Z/+nV42PT0Pu4rnRn9XL3Vs3aT2PjS2uteznVZ6BpOYjfzux7m7+wKSxeXIcFbBjeOa9e4ucFLnNDmMY0lLx1ASI1lOO1gAD3HodtJCMiov8K5KtCPflLfobyxDgsEAp8u3oyBORAIBD42tm1LOY8nxLbTt8916e3Y22KJHk7aPPHI6HO00BpwWF/fmij9xQoxqzgXflG+yq4FNGU6zwYBbSskzY/zrBhjOAOIx8DllVW65ilHysc58wDcQkirv3nGqfEtpXSDE47zeNLxFBYD5YNVN95e2mSFVwyHe25IrjNfbqGnimuxGAgEAoHPEOecUcGAmfJ06PE9vMZoJLTm4xaZDIgRQnKJqfgoNDZd1038zL8+j+r6jBQSR8sjyT6WNes23MG/2i4qrS6v0Vl7P5LRWYPmLMPgyGlNJtLXpHSafonrBQdyKaWzF8+ycc4eEqbL3WT/BPWub37/u+z1v/jnfzD8pu3eDgInRnUa25ZjOlv9z+7HtJ3XnFzA/ZvXMHo9IqO5OeXGbFJ6L0dPckxbjuWhLsHRTwkF1rpUc5hxsGI8H3kcM3s9iU1PcTvGg9NfCu84BFAB+0bYAAAgAElEQVR5a1ubSU70Mf03vvd9tfyvCp5qYKYePfhbUn9Z0Lc744FsM2gv4vxpDLIWS0Xy0nqwchL7JANJsxPK/TSJCoMPVUqsic0DaRJuvyG4OCgnDn3NYwTS8nHeYFY24XdK1rQw56Uffa8byuYVYcuwpsF6f1JdmalLQo93DoMxmTiaxs6Xc/V7CQQCgc8VbdxsLKY+ju77qRPBFAvMN2PeS8lmlN7weD0N6aB32M6Z6TkUWAttzogkHVpD82jz6yO4UsaVtvHUI+qBDjbnigzWAo9LK5XFGXIsrOgSmg4lyc8yBtse5TTdCwQCgUAgEAi8X7wXBvO07S/N21PaAQZCXsqok7YcWYa5ehzLBjPe++v30Oac00GNk4YnZWqzSdmn4R/GBSkul8oteaQnz5njoDl8H1+X5Le9+9cM05ynW+obXN5BAqPN0xRaA6er5K+3Dv3UU7o1dGRA0Dr0sDLWwrFM2/w+XwNzzr1fZdjmecx9+yBeFyD+QswwenIwBmbFgnGoGZiOkaUK6fFfLMssf5fR6v8aC0D7HjUDAHdPG9s0lrJnLPGwW6zxTLqmlWnJimX2GIokQ1MermmyHuRvD7XTu9JoeMHl0T7/rR/8kJU5EAgEAgEAZVze55x5mz/sAIM72Kmmze3cXIjr1ubuq44piUnLyamtOzDTrTvYenkamxuvQ4sQZ5WrF9d/DPrkrEdJzEP6jOiXKCfWeTidjdMxPc6r4zhcOpulD+q66ZyO01Mtx1yLy3vaN7BM/ZC//vugoRbLzGD+KhB+WlvBeR/nM8HKUzrsG0aI4viOLXvAs0DZqzRU5WrITSpeDxsq2RR4u420hmis8DPbjenfuHyt3SybkWXvkWwtq7LQciQnsvjuyWHmZcG2d2Y4b8CfUf6+IwLWVWNIr88F4doPBAKBQCAQCAQCgUAgEAgEAoHAJTyVwSwxznKtzXBPnTt0qxt1JFieDIvJfDAMZupdl7wjUjwZizW94kWVyrI8QyvbCSkkOakHzuv915h8j8qk5bfKpH3LgnSApCZbfw+jB1Zigvbf68ztq22aiPxX2vpzQS05VTh4qzGXT7bxfjtTzYeOQJO1re+lNK/8luX+N40pwChJ3avKffvTO2oxc3OrUxo7NRaGdo+LEw2Yx3FfHdBeK6cTa2MflcfKr7GjpXz4N83vqVfDDuW0TJ29NMX9qj0e3cSmgAgMpbZDZlqRzft+TsYlt5jL2/Z5edifCbU/tqH4TAPf69HjsqeU0o6DlxJSj2f3ljX/enSUZ8Bi5HGycPN1S0tDiDnEX5nH3nd7TPUs6HOSbNqBV97DsLnypbFQy+vZpWHlYXUmIAxJ4eqYeqBvNNbbNo6bWdteFFhGfze9n1TS5umoiRDpzL7DMZg987QXHEtaY9LybGl77KXA6bR1jcagxboAVx8t4z5WjOFkSinOZxxSsM/R65D1KvoN412ZnE6Jy5O+f+saB23nmDU+4jGXy8vNbVO/gG+jlCGMVkopHfVA79r1OJ8EKFO8rRPO/vTycjdF1Z1n/7b3UkYGdEpp2tkg2UrEMp2yr8Jrn5Dqa/d3sn4j7FirfKon0L0lex377IqtBa63iATG93TV9iDZcKTy2qHWbe7QbR90J8eRyKHFjYHcah7y0xC40ibwz822Eof8BQKBQCAQCAQCgUDADXBgNsPH9tK3HJ8Go7rnFn4JzB17u4eMHuW+JN1SbmykeruXdS75z//30A+VhCMsSTecagbcnh7/lg1+x5ESPctlzg/Py4UB4AyarfSU0t3pC22lObTxvX4dZCtT3n3fWQM7OK778/TyuNBtnPH3IKHe9n2fDF3SgYKccZwzsHNGqj3xBq3zB/ynyzCVMDvpxvbsaWhohP3eGc5yzzYsuYUghL956+LsZKt+PZo9sYXbCwQCgU8RTzEwW16Bfd/TBoM1mSj3AybRc9TlQ/FMdUm/AVc8BfQ5vMxXyVPkqUPyyDePnsCE7ZMkTLy94SQlhD4Xd4Iw/tsmUBL/i+pOmnLg9bJJ0BhZU1vSeFYGm4vWIT0Dl05jhXqeZ/KA1VEhyyQ9fl4J4nNTJsjUbvPJvJ8v+m4LYIFu59u43c7vpxb0rYNnHfKDQl8bAxqGkcZu2EqvY59jLjc4dhJc2dmgMTgwA0MaD3E5nvGYyoTrgDGGG7O8DCftnsXI8fR5jV3NPZOURjvFnKbFDD0pDbc4a9e2Xg4wlimLJKPTzfv/r7E1AqO+k1IabAeU5dfGDxg+mLkKvg0xPiHVndK1b+UKvCw9DGle9bASp2uKDtKuEabPShtIeVbKUMcWgYWuMeg8aaW8njSrsSM52bS01lzVWOmoqGnXhlEvXzc5o+Is8//9o99FFY1t+jPf+9FUbiAQCATWYc1VTdc5eN2W6kDAmE0ppX/jiy/YOqz5RtpJNevp/HlX3R7Cr1OkOZH+3cgZSPNORd6uAfmOY2flppB0mXOpkOpB5mEGVEeQnHaSbclqg1V9i+oOFmNaWst1p9ppf9tehvQAK6oBxbd/8ZdU+b/qCAZzIBAInBhZKPCf+5/X11MJencwk/w5ke59gpMOzcST1u02Kk5U2eCAGRWSYbMePuPvNOEzWz6lcD3YQAoKoNcph8vhyuacOVTJfH19NbdUYaYLldVrXH7EsJNzFrePYTml/APbhrCCcF/Bh4GmlNKRZ4djb5Oxrm//UhhTAoFAILAOzgHS5q9zqjxSJxPVTHSWraa8j3PsfhxdtwLdgqm71ppKgXnw1B3I/SvP4s0rp+edyjSdxwH0yDNA27BO6KofUOhy9KWZmITrwL8n53mtk8GKIxZwbGV88B+uS9PVNLY0lo0j4mjvAK8FKKt52zbGqLX+PoP8EwgEPjU8xcAMJybSk27zy+mVwHGQIR5kBTYg9YCct+sY2xSQM+/pkDwj5eW1GRQoC4JefyGTMWV2dGMS78WA7WAll0bEXo2lTI0gUjoKaJd7mn2Ql0Iy2ABovskzWGRZihDXmpPz/vv+lxpPJE9Z2wKVe3/rhchK3SCjYjDj6uTaa/Zc8V5HUHe7Ejw+bytv42XmFD9a90FUb1Dg4T1B1S8vEO8WygZD23zidyAQCAT8+NaXvyne+4vf++9SSqnFHARSZTttmskD8xQ4KKY4cce8+4HqNIAr7N0VA8vKrh+pfE73eYR53XZaNX10/UxrS2Z8j+bhZJGYL94dJ9x1zmBlMYu59/JMI8aVPpSL/a6l5xv0ozMUQ4s1L+g13/zFH5r1BQKBQOAxlNdz7n13/9P0FHL2FqyDj9omzHt+IEec5W2vr61sLvRKSrP9hs6F1M7R7Dp0V2mjx6bxL8lH67HJI2dx5zPSXevtuaCYVhwvyGQXOPNT8g/Vs6g8RZmyp1jGzYE4thW1kU3lCHpd10VIDGQhfbcfUb1qbKNGADubFPpZI+u8ELsiOUOg6drETtP0qTIysT1Esc8BwWAOBAKBE5jVAfF6+mTRGTK326lUnLaQDKwcWCQftbGBDxx0Ld0n9qYwkYkNjEm11jZ5N4XllIJjikxMjJKn8EOwBSohBYg6zyB2IjbyUEWE25pPlSOcnjKGNaMRNkxIzG+uDvr82mEsAK0cLg0nv7TNDuehhkFVaQLnUFMej6YcQV+BmJMHZpu/QB+7/+m+tzyro8V+D4FAIBAIWPipX/jV6drf/Ok/Syl1feg49kZsKKArnHuzj1tN+3HXe0BF2QrP2mxkILR07bFqz7oqP/d75nYuHwduDh91kFkP8eg/K/XRcjG84cy456YhATC0A57xX9AHcXot78i6HkNNYT0V5+Ocm5pDy8NClu5xeh0tbXBygSGq5G6fbL5lZDw9L37te78m1h8IBAKfIpYMzD//0z+V/ui/+U/FwZibBFIatzLD5NW8PsR4kSvvIUBXhl8Wo0GLK2d7R0ZPlKWEAGOUHjSg5hEmPQ9zxluHVwbpeS0GDd68Vs0TSX3P0fIDW5pk57ZXSe/Jakttm5iER5RFDKvveWRItO+TGNul6O8bjFo9wTrTKxAIBAI8SBTYtL2eqhdYVGAuM86gSGmep7ndLfS3V+9YTSPl0Qwh0jzN1WvNj1fm4Wc/+wor2auDeGTQWM8SO4nKoZWrtfH72LLNvsumo/D6/aQfJua7mJ7VPisl8Bz03a0lVXh3py+0v+8jHUdp/7//3dLtJ++GdCm/pBcSgqDWOqwJ7nn9IQWu6P5j+vkeNtLi8uH/2DHv+Q49cowyzQ5wLtQXXLfK4w72w/Vw8VWlmKtUPs81yYDOG33l98g5+LV5ijuPJBljYCNm4F2+JxuztQlejz1/GA0EAoE3gWAwBwKBwAkcK65tMSKK9lHQ6dfToRRnQbn9MymR29YXINtGtuC08CWprc4aExgU1IROXZcWKDWnQsLRcFu2JAYwZpzQw8awgUhaBHDKO2eA0Q7Hs4xjnPOQM9xITGhpgcGVczUNro8zRJkGwVKaUZK+PpV1g51MJOYlOJxW4mUHAoFAIBAIBAIraIx2QkCDM2isUAoeZjoNGQa7QSmbnsbJnv6Sci1ioUVUtNJJjulWL3GgAajDyutMpuVD+9D7HKSwInmDOq459iSZLTJka7PGzIA25MuhJDz/u+PXkIlsGeXOx/mcccnATD2X0hbg/sHOHtP2UVDvM4nhWwn7krxn1yJf8tLSTiWxnenzSixpzvBCy/B+UKue9ZFdAn/5gdFqszldIfdbTS2dNPhRSM/TjGhp7DM0TIAHV1jfOJ8kG56YpDRSmfQ37ZPZMPZoxrD5QI1RXohJ+HO/LMcIDQQCgcD7QXNq0Bh2QIwCHer2ruWB3V10QQR5tbA0FI8wcrkyVpl+OI+kZ2m6zgoTV3xWiNOnpLP0Le6E9aXdRlQkZYedlPeR+rh6V3VVrh5Jf9VYz9pp9RSUsSwxmlPqDFmWXsrIEXg+2rtt+u7W+9wZ4xTGwi1v7b0e73oc1OM0WtTz4OSSj3SchoSc+lrrIO+zlCyugYZ0SXbwYnDxRLlvxNq5cGXMlO5ZuyP69zfvcrGgPRvnHOe+ee7cIs7oQp3vOBwGLldag2PUWpN2jIyHBIHTeRnMTTZUVIFAtjmlg8z1BbebI+58IBAIfIoIBnMgEPj/2zu3bUuSqzzPiFy7pDF8zZ2hdaAbdUt42I9hY5Cg1Y2alrp1Fn4igyQMAlnWEMjYHvbgPYwkZBtsGi7NA+CunRm+WDnjMOOUa9Wuql27vu8mV2ZGREbmykPEjD/mhB3vMxWy6MDZec3FgI1eNnUr4sup7dqI3FxIo/VmVDf4m+hX9+TKwbm8Qa6dg96U5Xyb7a0HydTFVmWtPnyzSNyzEeUWI4PTpUaQ1gDVbHS57Ei1DQ4tA87IgDUyOrWMJrPzds4NXQKsYqeyW2cKeSfI/I8+HSMJns8/3vjiV6tjAQAAAAA8Kz75m7/Z3P6TP/nB/utY/0Nbx/mgkraObTvbDsxbIeSMXp+oEodNhGmzdD2SOnfcx+n1XY7u7w3aXCRcOCiE7FH34+aCu3y5+LZiONbHl+lb7uWKcquAiuP8H/+NN4f1fVm5ysDcU6v2bq7ckKE3TgwqZRTLOgzY8+dsFc5HbujeA1Z1/CdKkZZPqjxdy6hxVLl8jUrlknQtrEI51V/3H1cNZYVO0o5fwt63r8cRA1jP2DSjZ4zqvYRa22ze2celurc6+fL8h1XSe/AUVfIwFf4y7H+k96SLkWqD+OhHUKMe7xbmLb8/9bdp0LgtRp3VqTfp+7iv57MG9mO4/P6OjQ97LNnrnHzU5YEDz8s9xybiT+336bZtVYOstW7fiflzapU/9uPcmgmQ5++9b5WWoXekSBm9n23djrzDR8dvqW8UHZxoGdiD3g9aHxeKQQsREa/T0fLrE4NMos67j1gFc3w2dH9s2KZ7fQ1nNfPNcm6m2WCZ2oZ6/DipnmfxE1rEezXeVnv5k0Z+qwy73gpIat8nSmt7iqQ+GcDajt/3o2syHdQaTAGN75n4fq0DfClWzXeJavGawbxRO3SkCpzRCvg6O96sjqN3uBZn39fFvWTega994WuH6wB3w8c++/7htB/8+I9iP9A5HWB3cqOq5mwa/bq3s5z6OJZQuQLLg9S13oNJcZp9v7f62z5SCefrI8NML98obYtr+4n5O7f1fLX6xCGUrtBm7azRAHr+XF7StrL7Ru2q8hq0r2cIoXk+R0QDRZ2z907yL74buZYlM2id03/iXWaNAsDLCQpmAAAAAAAAeGZsLogsqkQ7L09uke20G4d3A/O6Bgm7e43tNhMFGSOyz4zDufGzNRMsGjbXcpBlZITNy+2JT+xAyDWM8s7q2TLcXuoSqFXGaFu+bBniR8fP14/W/RJGAh2bzv6+Xdfm4GdLnKCDtE/yv79M9AbE43LitlQkSdbsoHZvQFsFN7asWKa51WaCytkA8FG67qZcu9zeO+jSetjrM3r+es/hTD1tt9d11eW+wyiOrcC0fvZMhfz4+Ha7nUl6cqf28eMAkhV4QYuLDcytF3DvgYt/fvERVPWWjkpLmUfsDdB+uI/UT9ePKj838yCH7boXxqxxMspzFx8mVQHUx7Zlt19oM1/U106jbx1zM1VaJgqsUWCvoaplUGbvvI6+NEdldK9Zx/fWVcr8apr9XFEBc9L/rerDqDOXEM6NlE2VyLpHOxYhiCqXg/GDt2238d5Un+PutD+ze2fK+SAhBvnb0yxar6Rwjve0icye+xes7luX6jx6N/YaNPl6yz9hj9a7+YiK0T7P+XtqpuhpKYiP1LH1Dq/eD67hm88sgyTFcnL+mo6l38F43Oj2QtXKPjtvo1LWd4UPsREaXPvdDQAAAAAAAPA0QcEMAAAA8LSIiodQLHVAZVPXYCGpa25uboo0UR0lOo38vL4sNzFPdFlgDr8e8HGog/vG7dwwyFVXVNBQ13Tdnpk8OXFbNU59uZsLdTFi07XSdpU2jcNWw/YHBok1TS9w9IieoKOoU2dwsZVn9r+MBt16CrRWnqOD8IVK0N4TUQSyXz9RoUo27X7iWgnuF7k7L8VtQfyj/d7a1c2nNT3D6+6CYF3XqGZWdbM/naJQJcc+c9u2xfLc1r9nRm5lRNr3ek/ccg3Oua7o50heXdrzbw2+r+taqwGlLb6xz2vL9Vle75Z4oFWuuttRpWlLyJDfM4VbtY5oyuYd0Sy35UpEGu9hlMsXYV0radwa6w5MaX4nzX0dxTsmSLJdv1SVexT7nCk916p2vRL8WReLOzP3VL02jE3Xe1e1np1ee+Wo+67eehTGdJ6f0YyRc/7QTNdLr8LGeA2knS++00wx14gtX0YuMjD/9H9/IK9/vvYp9NMffVtEBv7/LlD03oW/uV7+Wd7qQZNxXXrHtL9b3IVSOT/OXZQ3e+H20o/22Qc5S3HeHsr9T/L/96bZzF60vbIPq5GfgJkC81AZnbzxOrR6xdCkbUTZ92nQPVmzRrU2fFWBmlSmbivzx6B965bK1ECAGuwv+o/0IjFI4O6vvpiepxXe17fan3xsAGu2qFxOH8+RYaAVeDCvx0gdnCugh0atQQNlZOCwaSx5eaP3QFed3Ohc2PB73vVV4uWx7DOerUfF8r7qy/96c0GCaeykRs+5Rm+8i49RAAAAAAAAeL6gYAYAAAB4SiR3Rn2FlchZfRdRwa1Rb4WOz7zidxWgd5UZM7XQkcHdeLTVum2aDwaNOOru6gij85gN8J6y/ydXQubL0WBZT8U0Op9enVrBtWZuvlp0B6YvuLbXuCESc00uEUv0XeqJyAG1PtwfctVqfIa8kzUqkvdn5iRplHeftOGDk2XT//tchgtJ8ZreoSFOetB4zIs7ybI/zy5YcVSqn9/S/RQFMRps3veePfvuzBSJxXuir3LMVYPq9nCkzM0fgZGrsKTgXLPzTcrEXnD38ljj4IZW6Zgfv/W+yr9lIzdpLUV0fk2W+B+XikjnXBXEtvitgbRdFtR7V2YsS0izifZB/c1t4vRTsG/7xJcI6Hcp1bs++RgUEZHNBi83gjSReWBjq4a235lrZwfMBGyz9lTvW20Vyb3j6bqdmdU7Xk+sZxXere/mfDbQhe0OX14jPwkoPWuD2uDLtr31yTffGtcHngp3amDuPVB5BFslvRQkphGxt2ld1qWdlPyDZcuwnaCemm42HXCmmL2ES1WyrWMdVe72jjVTDbfS1x2lfV9Vxvj/673sr7m2vRf4JR22Fq3GnV1v+YstypBj1751nN7/kxSOw+rDgNF1zu+fTa+xBqeJec6v1NWtyZ9uMMYRn167OjOovjeDbOqnWxtda6aONlN6gsSCRERkW7MpVJ33Wuvd2DrfI8aP1rvEdgpaU7F7x2/VZzRV39Z5VFZLndzaV+W/QGVdHtM0ALMAFPb96Bd9ltP/qj6XXbzJzuunhQcdAABeXLz3VV8shNA0aqgxJB/osd/s9cNV/KYzwvb2xRbidMq4LWuX6WyzvB8Sy40DFlIFHdOp1Udotb9qY+64/XK0r2n7Ifm21rpeV+dc4ZpidC6j2WdHafWpe3E5ZnVq1SHNoC3rni/3lWpbawA3/l0uxZbqGQQBAF5mUDADAAAAPCV+5bfbbkx+9v3fFxGp3HGJZOrMvaO7mIGOOOCy1QYIO8roOgO7Iv2B5p4yJs8zU8eMBmOP1iPnEnXtJQPnM0ONLm0E+jyv9e04Ku+IMWaW5kmEDK0BtNk1vUbRPFJDOWvAsUtJisGYxpdpdPunvvSNw3UDAIDnT1AhhS73Jod+Nayw1bfaG2s5mPL48eNiXb/Zve/zpaK+WRtkplCeKaatYGzd2rEjjtbDDpT0BqmOXIdee28LWzNdj6PCupheP/u+TNcT9MH94E4MzD2lb65O6/7xsb3oinVl28yN2JwSd5lKzaZRegpWO3WhN5Ug91l6yUN718zU3jad0nvhHDlO+s/Hqu7oh7ZzPWb/QatjdFTdfkTt+KTMOtq9OvbytaZzbSbNEpXLbQVzdKAPU5rT59KWuKw+3vseFTN476PKOVSGoKTYUaVqUvDk0wLNR1wbYeE8NbRZs/gOXprTFMs0vjqPvBFkG2R2veXDOT/mkalis8ZRq/7xnA+qzUczNXoNk9bz2goSdsjQZp/5Iq35j8uUxX+0LJdPJQcAALivbBKigWKR9I07GozLsjxa0mwvbYRtIQXy03gZmYJ53Y1UsSG1OQna99R2WEjlxVlZjSoV32VtX2X7R4M7paK5304a9YPyfS1XRa26tq5tqw2V53ta7ZBL+2WrtNp65315zRZjO9iyP6U4l6hSPm97vK5RMa/35LIs4k/H+skAAC8jKJgBAHY+8etfrrb9n//6PRGxAyqlYVY7HKlBH2LAv2hg3Ne3dRMXncWd85+qzpSL/sZsezuENGJcGXijMdjJcnN+vaeOkhpTfUyrgQtv9zSLVrXh/86ybVsVsbnFyH/qkU5k5V5GOwjZ8S9RULrM36M16Maja52zTsuNjVDsg6jvwlWvuwbpW9IxV7FuQLII9N6MwO/3hY4znDeft33qC/j4e2g4825wLk233W73Z6ozPlgpP6X/rMYO96Az3BMKHFHkHuHIYFKvji33OrM8swHla1TQl6QZ1an3Tptty493xB3Qkbx3YSCxxqZrjnPk3rSq59wHLgAAvDjY9nvlj9+kXxq+ere976F9EOvPPfrPbgQpz4+t2GP2hIT2GzdTBPfEOI8ePRoex/atbP118KPlQjDfbtsevTZIq41iA77bdqb6UJ5941OA8nL7JWLGvUb7Uq+5FOv1vwjPgzsxML/xVnt62k9/9G0REdmcE/3jq9dDVOntN5D6ptWXgQ4e68O5qSGlVAVGlVfDsOCNwjOpB8uHwm4P0cbQNqLYm37k7+pIJ0qkfhnMyrMvsbPyUPPaF5s+lO1AM7O6tM7TkjrI5YPf86t1tNN3RFU96tzk9IxSMwV3q9xeHpu3On9VWGjd/TjfKmt9LJ06fdLt5/vz9Xe+WZ80AAAAAMA94WOfff95VyHyN//hD0RE5LQtsumMs1ymvJXGIVmz/kDsF4TUedRskg3cDY6f9y187BNL5Zk572e0+kSjwaeR8jiPlzQ6xtFjjtTaPRX0qO8c+2C+PfttpOYOVm7emCXXOv5yc4p9Lf3zPv7Fr1bHBgCABApmAIABrUZrbyAgppEtUzdLzCdy9iPWdR+xpdHpOICsgzSS1NLW1cymb/I4BrZJbzQ3hFoSmc4nqRePqO0uCdLX2neNElHJg6v0Rvlb00vzgbKR+4y8HBGRsF/cvAso8d7YijQa9tF7H6daZqWnEpbyXJyJF+M9UzAfMtG33P7/395mz5oOrmeucUQy9zih/XyLSBWMSsQOqqb9M5XuaBr1jKflGuySZ6JX/9FAseWa6eBHzn1U3jWD7Uev9yXv4Gvq2EzryuWRKxmVTjoAb6alj3yLA1zK6ZHOqHLRH6xkbv+ie411bxfcSEyXKwNdw1VgiKKbLMjzaHZIpr1S4+gRoU1O3kZqvf9a4qFbFW/p4V1hi9XcXUP5bNbC0brnAfoq1xeD91frWCGEKuBfXostK2NbXJHudOMz5SXvm7vk1Td/s7n9v3//ByKSfC7H56iRNv4n+h/pDhVnmbyu4wM5qaaPtXdi+s18i1T4FZ+9UjHcUxSnGZal4FEVzrpuFcvaF4yBwSdtFZt+2+utojcNNO6yIOKnyj2sER9WAcfL/bZ/rO/AXh9TqplJ5hrZ/tl+/NfefLt5zvB8eCYG5nVdqwZhutHs9Gkp9vdUm7p9NR/yzT5U2Ue0ZwxRrMq0amhPplaMfF/1Gga98zzqi3ikqu2t98qydZp1tvL1o1Mzr6V1rY8ojPO8M1X4UWYj9a1jWdwy/n/tenDpfHM/YCISfYE9rc48AAAAAAAAAABADxTMAAAD2oMF1n2MGSjzPvn8WstplefpfXu6fZmCxuwDTd5V6paQB5QBqXUAACAASURBVP2Lo95GbbKP/OYBbGJsQa1rVqw9NQ3S6bLK9aZnFse1gzf57yfwGdqiNcXSRrbcsrRplL2UB+cK5pjPDkL6kNXf1Nml6aQqN0r3gyrBnVShsLOgm3HmpSYxamfvl6uuEbwYvPE7fXdGP/nDfysiIlHwFm+bXQ2y6jsg5ekF9lTS+nyQdLTeGyi/5l49kufoYPfRfbM8veuXK20uOeejfhpbEeaPihNax7NpW+7NZr4YR+czmwFSfCtcuU/MUv0p53l8VFudl6+/i/sveIqoEi9vN8QZaOnz7E57unWLweLc7Xnps/ZdFFRtmTuI2PYL0cXj1mhr+UG7aSQoEhm/G/I2a2v/yPd+SxEc65EtfXaOeb6KJcUDicdQl5ZaXkhtsFbd3D5bzImTxuS8lGox55CXk7koPJ1KP7PBB/G7arRWasLTIH6nwmbW6+9kTzjWnYljFcOdb/Dsm+2NQjrW3dTDtrV631cbj8b6ju75kt62dvkzl58pw/5sLWU+VUrn21qB3kVEtnAsmGhXcOntf3UsfzcD3AueiYG5mKZjllYJrx3z1EEvg2hV/pFsJNdQps/32QdTt1vn8N2AVVk0YZHKFpDOsTH1qEerQ9HafnQa63nZ/cpexDVTNO3L7bJ6H5+Wkv+eTem8tNN7F53k6fn7dvp4rzrzERSXfInrdPp4H5dlAAAAAAAAAAAAPCueiYG5pZqIxjDrEsME1lP/NjH9xHiYlIGZEq+T1m6vfOFM0ktojxyJ1CM9lp5qxAYinBlLW6415hHKJ+4bLnQlkac/Ono2U/IcrVuOHd0+auQ+eqxL1FJVWjtCZ9WQ5pjOpBfnomFZ/bZ++t2vdesDd8cr/+pL1ba//W9/XKy7hp/dODimAz7q4iQkhctmXPw4n5J64493W9PgWzD3h75hY3mZ8tZbn2JWSZZxWtL7x6aP9cie3eVUvueayhojWCsCs7p5UNP63ZC+Ba5zLkVa82691WCuzsdnSvHxukZtdty3nOp3Wx6E85xLr7+qX5JyJtUt/f7Me98SAAAAeL688rn3DqX7+x9/T0REbkWSxFnb7uJkMUH+8kCB2+3eRtg2STPh9iK2lK9oA3X6Mq12UjFrIMtXu1+s223e+xTQLqPVTxn1r1YJRb48uysUl3W5h2aEbC7mbaVv9i9Lfdjehi77W957DZ8Rt33yna9M6wN3i/an7J1lxYEiIrfr433bvmGr2+n50t631jey9pd69gRZsykNkmxVlp6COQrFzHOmNqPb7bbYbwWPvXeALa/rGrUUg8vSyXdEQGi399zMzmwptUJ5bD+q1m0nC+4FuMgAALiQ+eBDiA2eKgBgCHHqpL8t8+eGajv1KubxLjaiUiNoX9cZHLdbMugam6nf6kZa9SF3rjtVLD/XbbCvV3beHgtp3nSVL6bpLPPBtGp2SmsqeKchlm+z57z5bCaMXmsdwHMhzjRodHHS0kz9JjAVHMHf7PfbbpBYdaAq3n/nhQ7yiEh0jbPdmllYV7iSsB2sUUe+9/yPgkxd4lpi9r59VrN3RmKJ0fnNOrmtMo66xGgN6l0jCuitH6HK41213XeOYwfVi2vcKAcAAAAA7i/PxMC8bVvVmIyjOEYMvKlPKuOTNDY6Xdnp2WJUS9OgjiNgXlbrsyYq+6TY3gvMp1QN+16b17nczWYnSdlZsEao2UhQz/hz5JgzBbPtWM7cdeTX6ag/v17HataRGHXAjnJNR/eusP9vdNOi10H3G2WmW5KfVjpbAAAAAADPGB0MOS1xsF8H90IIlftEtyV/xy72+Vx0Zqgt+txlZBx830Lqm3b6ZCJlX6vVX2z1uVoqTysaaB0jhHBgpmxJPG6uZm4cv9k/02sbZ/xth/pBUb/gnWzG7aZffGZXyI5/OnY+8PT49DvvNLf/5E9+UG+0s4N1FmGMJ9O2pcTs1t6xmvt6sw/zMVWtRffroHvy4bzbuqS8P29vb4t8uq6keBu1YChf3/J4MJLiGlR+lPfytH758223eTNztVYgT56dZODYV0O+Gs/plz//9rgcuNegYAYAuJJLBiuajeiTTpc6r6aBI5FF9PfeqHfauQhp0pV+4GNbIKmcrVehWFWdIiVBlti1SQ13kbM7CG/82av7jdwtxrbt7ib8qXk97HlbqqB6DUa+4I90Air1oM86SWbMLVcsi9hAgFrXxmCYVdrpAKfbskZTuQQYYd1spefA3EDZLXt8oLa1tbz3dWbmaKrk7Di9GBNHsedzzcC0nd0wekcdrU+ep6Vqzo+b/565Dhtda1tWS9E8uy4jIULvGrTyHP3ujZTSdrsqmHPV+7Y8G2U6wCWEvA2lQeF0Bv26JuOwsoVk8IpRwFIgLTsTSyS1xWQLMbicGrt86D9XIQTZtB2nftaciEYjjAGQxceGZ2zhbCn+S15+63m336dR4DWbV2SfgRYn2ui7rOE2zok4U6dctKaXwnufCdeSizI959NHyojKuTHdLdk7yKs4jXcPAMC1PFUDc2tEVUkN4XJdjF8l/cjG/eadf0lHo9rXGcE5qrq124vjmEPNGu+zBntvvVXHZFC6Tqk7O88j13hWxsw/0qxuR+h1/K4p6+ixWqP0R3A6jV5bqToqGRtDyYcaFioAAAAAAAB46Dhf99ndZm0hE7vFwc55sjlZ24tVROtyPIBqfS+nmfPl4NKSxcE5ry/NdcVHccteXasONj6le7aWoWggHqOseyrDCIUaMaRax0qDVKX3ABtjCF5MUDADAFzIkQGh7sdX6oEPO23q3NgoGwJJ2ZFUFirB3WKaXVG8nJK6OZTl5OqaOEVTXbXsjR1/lo3sO3WwozFAZ04p9y8cjyflsti/t5Xsvvz6VddadOrXlpQn9lqF7JrH09C0cVQzNcqk7R4pH6SK10rVRNnlcGYa2q++900BeBJ+5Z2vN7f/9I9+T0QyVWx2IwZ9L+jrQQfR7RTGUD+nShpo7w/AHlXI5vuvGbBWbKCbUWeoNwX2CL3OVy+QTitNz7VYnn8W5LlV3ux4rfJ61zY/fi84kK1jnk6/V1r+41uTtlVnWyftJ2sVd7Xyq18keDHcb37psweDAf75ORh08D621YqAf/qu3fedtlPj+QvZFLfUdlGszS2EEPcXimgzQyUXQRXfEOsSYELrPT7zzR+r0HhPtYKr2XKWm1O7jKXcFv2++xQQWn984p0vX3CWAABwCU/VwPzGW9+otv3Vn31XRNKHwgay8qH8uOQ+pM7bS780M7WxSN+fsaZYD3QeWse4Ju2sQ3bp1MZW56AWao87KLMpmJdMQz/aGTqqYB5dn179L5l+OjqW5ZLr0HNz3SvDji7aDqcjSCoAAAAAAAC8RLQGaCsbSFzd+9ChbVuItiUNiHxwEMTapJzpu9d1NoPH0W1L2xZiXXop6k6nspUspesXG0w31aNt37K2hjW6ScxcVVX+mUs7nPpkPjqYn8QLu13PqK4vsbXB/QUFMwDAhRyZXmQbAmlfnabpT1P9+GqDQ3TwbZWgA3HqD1hdC63aKAhpKlclpEsNG+fsdKvaT6n92OdqwlGD4hp1Y8ut0mxQqlVmvhy5IZpPFwtZQ9IqI3PXT2UDDOBpUQ8q12mSIr89zTOn956KPkBNupxLAz2N0jbfgQdVzqP9s0HvJ6UXcHn0Du29G2fBu/LlSDl4iThgpi4c5ekqu2OQ7WyTKbea0UKnEh4YTg0vW3JxF7JnxW91QPX4POqrYE2z2VJbLnsHbPZ949NHYa3fJ/Ed0djW4txOnAueindPdFJdq5ALDU7myzrmTdPKsuVW1nlpiHCci7PLrDDHOSe30nlXAQDAnfPMDcyzxn4dDKX8YNj0cRyko7oNoe9jWUzek0l3yRTOGTOD1EzJ26tTazqRf0L/NVZNa8+hpSru7bP+hnr/8yzYV+u/mHWi6g5zu/N1VH09oqp/5xj22sZRQzNF314357N6uLrRCAAAAAAAz5/Yv5GQgtHlAytJD3De55fMwLz3SzKDrcbsK/pbxsAcQkguw/K+USiXLuu3lfVtGGDNrN984F6JAfPOhUtaOZeYkmvffZXgyn6hk2zQXu3MIVR2AJcZmKObNJfcYEQ3uWpg9k7S0P/dDCzCsyeOuWQil9PySESSEOfSgWNrH9Ll7e1Zpauz6lVBrOmjmteV5dTll/vjY3jQJmQHcY/aKY7Ogg/xmVHFt+7Jg3YGs2wPmPf8POvytbffbtYBHiYomAEArmTkCqfXUNCABud9ZSRu5XQ6ye3th3uafd9JGzq+agypUkMFcCGkqOXJld/eUAl5Q8iqk7dqawy0qgqQU5rKte4+nxdXD0qNlMO6TPu0EaVlr1kACM1n3SelsuoBxlR3DZoRfVjnHai9NWWnleVi86RKNgM7fst8Z5fT1wCeGsZ3bX7vJj/t+/O46rO7ow/TBZ2w1j191O3WTBl3NO1d5B0JBnqzJHrpRnkvUTAfuT42TyqrNdtkfDybTiTvWNvjjab+tgfI+3WV5G9fO6mLKZ9XJwAAAMALz3MzMPcUrWmEpEx3s3fgw1qOLqnvFhV16shWHK90m2hQK23I6kiUpo1Lo4r2phMwMiblhBBSFM04kGuMTGlHmdcExEqjtDFB8WPdfVJHstlRtn6pI6LXuiwr1T/mNPkPKGet+lkNQPZ84/XZiu0uGphKo5adnr5tWxpNN5HG7P+1dEbVYiAwaXSGsnSxbvpfNHpC9tawI3oWVRaEmK6MFKvuj0435/y32+NYD71/4fnx8X/9fnP73/3F9+PvkeFh5ppBJBktGz4uGunLZzsEnzry2WwOW5x9V+QuHkYzFhoVKZa+8Yw0jTPRSFY+u8tNCnbjl9II7139TPX87Odp07nVRvCsRsXpeJ9+f/qLdUwBAAAAAMsv/sYXq21/++ffExHT38hnZmp/ZMv7EDbIn8SlGaOXJWRuwVy/PedCY5tkgoSMXLms61XTSdtazlWBB1vBS0VEwsm25bZGP9tV/Sl/ytybZX1DbU+uosGqz+l/8c0v1ScFLxz6n9/c3MRtIRPV5Mse89nhGsC2PRhvbVb+pi0uyd3s5evRxmOUyekeHw9G34ayvovxMX2JeywRiUKkVB8pzk+ktjnoes89WE+NDS8XWKoAAAAAXgBef/eb3X1/9cff3n/tnaB9zW9lJ6QcZErGhZwjnYOZynak+K2O3zjebJpna/D/SJpZHsvoGvRnqtQK5ms6Wv1rMC/zEh/M1u2cs8ti8FOn7GvQn3IGSTT65AN/an/at73+Xv8+BgAAAIAXk2duYB41wnO6DX71E6MSz63T+FbVsvfxt/W1E0eATGCmWMddTXp06mZ+br1OTdeX9MEpmj1/wkUaaV+TI6pJkXagmVG+lj8uW9ZRf86q4J358smvY4q6eqwTd00HdLbdKvKra6rporLS1NUo9m3d3vgC6skXgSNTqPPf9lbK75+RT67KH1dzVH7fprdU8jWR8lqlSTAzEBr1z+ve8/d+2EhkDRNaZe+K32X16/fsYgwd8Xy8z94PtfLZlmnVBo6YfQAAAHAHjPpNeVspugjzQdQDTfRulqmWo+JZM24h/g4+tYd8VHrqMrX/fKaCVJdirbZs2ZazfbO0Z7PnlftMzoNc+7qN22ynmXRFGteP3xPqouAF5uZR7VIw3O5tdh30NIrebd+vSl13KtPFcvQ+3p81vcdOi+2Tn/M9evRRERG5NbPItb8RVda7nH+t7BSy10fd7NXxtEr2e9+YZ6r7XvtkVvUf7V+3RT31wdVZ1DFQZmOGhZ2Zviy7knl/BnXWrK734nnBywEKZgAAAAAAAICnxLI0jGQN92nFIP9Su8YQORtrnZni7k6ZAXfNXGT4aFlOS2vECkGWA87QRwKE3HCeC4JaRiYbIC03HLcCisWYG5kywme+4oMRPNjYGgAA8Gx4bgbm3sitjjja754NapVGX8rInGmEZR9JWSUb4Tx/bNV/c/R3u/uTqZSt+jHTj+RBpWuuPrTn21P62kaFVeL1lMut62j9PfeOdZQjCu7cN1crz0yZXZUzUVnnDbHUYLPXrC0/7KmNZ9NxR6rM3vbkC9aM1PtQLPVIUU2/F2OVAHC/aXUUWhxR2Y/8WVVTkrUxvj8CyzpSIOs7JsRnJmx6X5bvSpFMcWIUMOdj999z1j90Wu8rsXUUfdu2IuDfeV8t30kD6pq/qI44F7Ioznsul94Tafq3prE+5A/4nQe4J6hxIX96RZIqJ/XLs+c22H2a4tiMIJFjbYSjM8GKOsT2QKn6Gecpj1fncWZZU89s0LT1+2DmJuQSnrbSp/UPpFuiM7tqQEyjKq9Yhl77kKVFzQQAANfx6m+91d338x/+UETq74sXtRmd17UtpH0o61/4NpT9jVAFvy3zL6fSztCbEZ/6FlKs2wZKz4bQmr0tkhTHiRjVvVi3/UlbTqxfYwZBbzbta2+/XaUFUFAwAwAAAAAAADwl/umvv3so3d/95z+Jv61RKB/IDyYYmc+C4uXT99Ow2m4kylwEhMJdQGmQyjkyeOa9H+bNj+Ub7i2sL3hxIq+8+eXu8QAA4P5xbwzM3eAwJmJm4T9KRILxfxMj2cZyU9p1T7vGPG21bFeRYtaPKFx7UTYt9Qe77VfqiA/mns/Uo1yuwMmv+VipPPMzZK/b6PytglmV6QkdydMyNDps+/yOXNsZSVEapZVFDZLSe1fY67F2JfOio6mofF5IPvZr7x1K93d/8f39V3nP6fJjv1ZHH79r/vo//mH2firr8am3v3xnx/nZD75brBdBxRq+l0XOKgPzCNlXuzhx8pl38U0OENHnJ84E0+/PvtzVOFumxE2+OTWN7uk7IE8q4a1Yt8aJfCZWTwUzIn2Lx8405/5Cj+WZpcmLbJ1jvr3VHuzFlGgdr9cOSWmub6fk2LrZWRxbQ/esNqJV2+d6E8UZWrU6K05frxRXAAAA11Mpe2Nf234/danf4M7seFNOWnaOE7/XpXJY7SNaP1VIx7hLVTnrvh5r3FxXasHxZd9Xb3xAP8nMW4Cce2NgBgB4WbBB7Z4H/mYRF0p/gBow406PcyobJVvmMzAFvCgDsN48WtIUtpvShRGNHAAAAHiolGIcO6jUTy9Zan+TOUuywqh1S9Pus/LUvlYE4xu0uewg2SZbDN43MlqFEGSV0sVlCCENUlmFAQAAvDA8cwPz65//+kXp/8efluo3jXDbU8LGZYwesMVIAluM4rnn3Uea1Cfn4k5FugrjZ3j02VP/znUR16lqevmaARZUXSRmedBA482ZJX+sfeNTrHen6J56yV5r+x/1lOD5Oei1tvfEtK6mbnZ9VI7dt8aGWqlYjmV60zCsGlHnxevvfG1YdwAAgBaf+kK7ffWTP/59Ecn8/2Uql81OiVZVcpgrWlJk8nImxIhLBoiuUeseja2Qf8NnyuLW9O6UxLZF7f5sZkacKVKnsevJD3TUOO3r8/OwMTEs+TXRGWDRQGUDZR1Qh1d+Hfd2+qff+1anBgBwF4xmsbbS6XINIb2H8oSNGQbjHpV5d0r2jpKyX2QqJFL5t+3P+gAYEYNY7qaX0PnIzmZN25lJ27rbh5al2K75bpabIt9yY+KCmePojhTYUverPWoys3oShHNmz9AZSXEQ6gLlcjVDG2AACmYAgGdMauQ8vw/12VdeuW17CkHtlqXd4BGpg6GmutX5iXcJAAAAAAAAcD+59wbm4Evjgo/qlPP6trWNF9HHYHAiOrK1b7K+euyIVM/oUY8ElfiGkuQ4pS/Dnoj6Gv/KvRGtbk2MCsnpWFecseRshsq38KzeszpaWop1q3620VtrrG+ldrTXa6bi21HKSjW1L+Oo5WKO+RwNjQAA8HDRSOmprZP2qUtAWY0f3Wom03HffOkb2pqi3Y6vcIkf5dEso5mv52t8L4/8Ktu2Sc+X46jerfPsnceo/pe29fL6WwXzkRgjsYyTuRae9gzAk3A0GOBR/v6/nIMG/uJvfGmY7oP/9L1i/WgMn9H7KYQgr3zu/UPlANwFSVFvbCZmBpCq6+NMaFcqlT/8x38UEZFHjx7t6c75tE3lYrPpXM7JlbGeRAU26jZG7Pc2TlU6L8zjY+1Vsd69eFa+3W6INopJ2yM02k2vvfWWAFzKvTcwA8BLzz+IyAfPuxJ3yS/9y6cfxG/GswgkKCLy6ptffSbHAdh55XlXAAAAAAAA4GXj3huYrSrE+n5ZjQ+npePbViT5wVOscjlGy9Z069h3j2XL0i2d0dxrA1WNoo3n5Xnvq/PqHbPv1/miqhU++1wcjGtfa1s3i/VzZLcruY/rXvTYHtcqm4a+KK0PQm+vdTnq6KLfJa3L3bsmeCiEEH7hedcBAOBFRRWmqlYuXOFofAD9xK4mloH6Ym58HmOsC/2greb73lQYl9/ykbug5Id4K9LUxaYN1hdyfXz9Vqfjqao6+Xq27S1tb/TrmMqXvfx8m7Zjb4s62EgZ7SaG5rV1qhPbGBYVvm77bMbPtf7RMVBXTJo7lZaiHP+R0o8jflMB7hfbxJOy5VLlcr5+NC/AXRHtHfG7VU4H3/TbtH/K9RP/4f97fM4Xyu+e8uGHH57LMzN9TqeynZT6/trHb/f1tS1mm0hWYR2fK2tL0FgMerzK13Q5S83avVJMhz21OV6ajS0AT8S9NzADAAAAAAAAwGV0g9fbdDEI2NhlkQ0kWkzZN0YzPAACALxc3HsDs1W+2ojXYVmL/euutpGolF3iyM5pj+55e3tWkVQjTzuzCNw2nSWE0B3B7fnN2ba1WO/5E7b5Zv4GR/XMamXKONYQOeI3cNao6fkXtOdvz1f9H7WYqb3T9tVsOaaEzrfbe8gqmH/1va916wkAAPC0ee3tvpuan//wu+cfqpBdzt/s28d7GyoaC+rvoTfqWmfaUvlMo5jnAmlMah+d15Mqx36ri1xF3rot4swy/U7ljNtvefl2vaXITkpsrUuZ5lgE93J7q4a99qCdpVe0Ybwqp9QnpYkqv5/yG+9/q1tHAACA50W0GdjZO/YbmAJ6iUjmS1m/33uTJW437Yebm5ui+KRo3tPvh19O5Xc0+Xw2M8QqhbKNkzGzoZTL2p5jbCxmFpNaWq6J3wAw4t4bmAEAAAAAAB4wDy7eBNwPPv65L99pOrgXEG8CAO4l997AHEdZjHLZRvBWomplOW/fHq9RgaEjV6qyWW6M0iUqUPYRJB3aia5tjo3oeO8lSKlITqpoM/1oL3IxkUt7/pN7kcvz89d9PX/HMVqqXoc9yqluT76B2se2fgNzxXY6dpk3/U3G35/YY7d9V48U2n1/X3qvlGVEVZTJb4uuj6RRZn2m1NJRwbMfJ1lMBFkAAAAAgAHEmwAAuJxPfu7N5va/+fGPRaRvJ7A2pse3Z5/L6mP50aNHIpJsG7p+u9uJPrIrl70uvbFN7etBZ+H783Fe+8JvX3B2AC8e997ADAAAAAB3Txysdjr4fcbtwWi89acpWdBlHWBVVw/WPcXSCjZnpmwaN1r50OzcDUXNLMBUa/9seuiR414T4G6UtjfIrrQC+VXlmem3Xhquy0yWLRtIP+9msBwAAAAAjvHCGJit+nSmcM0VtakjM/MLXEYSj9HTayc3U6zPvV6HxUYK9XuHLNzWvguPHTdMfQfPOkCz/KPO1uz/6V373v83KzenVmyXquht9zak6mPbPesdI6qyJXXAU6RVV5RJFHUAAAAAAACA50NtM9gHx/dYU36fwX27Jzs9OvtY1lnVH/0nHy1y+z2W10eWs/lsebT3/Zfd5rToumbYF75t2wB4qNx7A/Mbb33jovQ///EfiEgyNvpMQaMKHbUObmtbtRINjdYlxv7CiC+IjssM51yl+rjUgfrmynR+oiKpDNaNY03dinSMvEeMpUcNqj0VTjQCRy8c7UA9oRGAMf5fnXGAOD4wqZt1O/Ivvv67kxwAAAAvLtWg7t4ROtnt2UfX6Yd6HwePnad1/17r17YRGNCFsv3hpe3uSyQLGmiDIl+gKD7ibmsUsHm0v1VuS4hQBWduqI8HBRfL1BYyx83S1G0i3b4Hxc4DOdsGkv6XGgzogqoCAAAAwMvNvTcwAwAAAAAAAAAAXIwOnOoguSqOd8Fh2IWEfveVvGxnM5mNXfWRj56VzZvbZ0pH5bK6HDuXE8dtoysyZjfDy8GDNTBHRWwmXvFWybuUgeZUsayPv/qiixjljapAUpDAtH+mWJ65m7AqY+tYwhvhTq7Kmbmh6AUIvFSFPAq4F5VG6gEk+gK0PgK9FnpeTBRD+fbZVJOqDFXt9Op/UOkMAADwEHj1ra80t//8h98+/zCzikSy77mZ9umWRXLytkj8XX22td3V+PJquVvZrqhmNh3xjXzAjVXVHhu04yo3XsXZXMaRmWd2+XhvFfrMz3JImfdFGcA4dW4z9bb62t6vz2fe+9YVZwAAAAAA8IANzAAAAAAAAAAA8PKgQkBdulAOjuu6P+2D4/tgtg6I39yczPKRiKTBWs0XFcyVS9L9cBrLCQUbvCQ8OAOzDfQWZI2/16303+utOsWqhq0SeOKbXZ3GO+dSnMCOQnnm93im0Fn1ANbHnmQvMOvX2GzviX5iHbwr0lv/fkqp6tmXB3U8R4L32WOMto3y9hRMNgBhKzo7AADAS4Mvv68+5N/5/XutguWefDdz4Ovs1NS4w377s7gZsT2nvp1LRoGGq3XrI7mhRtac8XR6AR3yRNVx6ryqzvadNlSL7rH35RLbL1ka85/VRex5JSnN/amv6AYAAAAAuIQHZ2AGAAAAAAAAAICXDysY08HeKMpbz4PYp5O6DtWB8H1w+3Qq0lsXo24fq12WtnAtBsvd3YeVTsQAHi4PzsA8ivrtfJQVi4jIZpQ03rV9F6cCdH9bfVymVb/O6new7UMwJT+mIpkpnEeR0a9RCR+pUyutrZdVg9uX9Eyx3VId2xf+dvD87PKfvf/Ng2cHAADw1m0aCgAACDVJREFU8OnFKDj/NN/pUCayna8ijVH4um0wY0hnna2dtklro21jGXVyivfQP2yvrbDmcTl6zaLazXG9by5glmDambquS1Ue53ijYDZhQ+TUamf6cdwLAAAAAICjPDgDMwAAAAAAAAAAvHxYIVoagz3/uDXpl6hwPo+O60D1uu6Dy/vI9KObGxER2fZB56iUXspRXW8HdRnIhZeEB2dgtiqMkSojKpp1PZSq2mUpXzCVCtfOdcjkIprHKnWjmmYv64CQ5Vy3gy+lEEJXDd3z+3zpMWx5rTJm6ume/2O7/6iPZpF6KkzvfFHqAAAA9PmVt75+cZ6f/+g75x/6uW5EtLGxEMSobCsfzVmBtj3QbheUnpTrz32oq9bxVVy1kbJMvXbFyC/0JbPIZu0W1xB+B1+2Oz/1zjeGxwMAAAAAuEsenIEZAAAAAAAAAABePj75uTcvSv/Xf/YjERHZwu6DeVcSnpabMqH6Vt73f/yzv/Uk1QR4cDw4A3NLxXpUsbql0OEiIrKcVF2rO3Zfv+r83eTPxSma5qiKdqY2Pkp+vj0/x09KN0q7HFf1xOkqRs00Kjsnvy7xWpk09l6wx7IKdgAAAAAAAAAAALiMB2dgBgAAAIBnSzWonUXS8+rTYdMB732g3gYEzqPv7f4MNdK7RniPw8kDDxlO2oP2w/qHtputZtDCdKB9rQzCl9OLY2jFAGZnfpRqBD34tSgj/x1kErAaAAAACsKuTF73Nsdp/+Yvp7bLVN/yVQUAD8/A/Kk3v/bEZfzsR38gIiKriT6u/u2cjdQdO0ZrXHemA2SFyKelrdyNvoi142H9QO+O5nW7rYN3qVulZVn/xj1/yD2l80x9HUKY+nfude620L5AVm3cU4QXZe//zz9//980jwUAAAAAAAAAAAB3y4MzMAMAAADAs2UUvDe5pmq7/WoPcKs6SAfB5y7DeoPZGgW+5Xasq24u4wVKqBxxSSOKYJ3GapysYnkYjLozwL9JqQAHAACA67HuOxcj5lNubs4+mbVdAQAlGJgH9Pwk9zojxfTQUOaxzFTCwe2K3VR6cYyef+XcB3PPD/JMXTzzhxzC2kjfTludv5n2WXXNOnWLdXHqeN9naV1RNgAAAAAAAAAAADwbMDADAAAAwBPx6ptfvdPy/uePvisiIq+98/XDeX7+w+82t3s78J75W+65B6vXa7VSPYbeGOjuqLZt2tff+UYzHQAAwFPiH0Tkg+ddifvAJz/35vOuAsB955UjiTAwt1CVrJynRvQUwL31XME8TJORArmc01XTMvbAN0mz21ZXiyS/xtY3cea1uFV0VlaZru4vtf0s5yVbJXaaFmvV352pqaYzqNnDruR23lcqbqaKAgAAAAAAAIwJIfzC864DADwsMDADAAAAwP2io/w9lGcbuyHLR9Z7A/fWiZeJl9wsv7Vu96lv6l5QZAAAAACAFxEMzA3eeOvZT1P8y+9/u1iv/CBXfp93lbJ2VHxS8a6rdl5iac2yeyR/xxqUxzfrlKuww6bHL6PiRAWyKWMzPbW4XdXIJr0z6bz38unfOT5tFgAAAAAAAAAAAO4eDMwAAAAAcK+4Rtm7j01LULdaoeciLPlg7imJnXF1pu7GjtR1VPcooiY+MQAAAAA8IDAw3xOWU9nB0Q7NptM8gzfbzwrgreo8He/o9ALaOOuUWVRtXPpRzovVPM6lTptIUi5Hf8ma3vhPTr6nNV+nwxc7gs1TAgAAAAAAAAAAgGcIBmYAAAAAuFe8+uZXL87z2ue/dnGe//Xjf7cf7ysX5wUAAAAAgDMYmO8Jdopm5aPYl0rloCpiKdXGW9jymZ9SZOpstlNIVYWctrcLyOsclddVjJxyamnc3Zm2an01V4ru8SkBAAAAAAAAAADAM8SaIgEAAAAAAAAAAAAADoGC+Z7w+m9//aL0f/nvf09ERJbdaXGufLa+la8JlCOS+2Q2fpBNOudqjXNPgdyrk+7/zO9866q6AgAAAFzKtW0kAAAAAABIoGAGAAAAAAAAAAAAgKtAwfyCooob9Vmsy23bav/NHbVwb92qjq0auYWmtT6Ue8fo5QcAAAAAAAAAAIAXBwzMLyj+pgzEt8oqIiJuceJ8aSAOMjbedoP4me22nMKIvO9aw1rkVYNzTKZ1sgZlZqgCAAAAAAAAAAC8cGBgBgAAAEj8g4h88LwrAc+GX/6trzzvKsCLzSvPuwIAAAAA9wF3iWsC5xx+DO45P/vT78Tfs/+25/KipzI+Eghn5lZD93/67W9Oy4IXnxAC2nQAAAAAAAAAgAcMQf4AAAAAAAAAAAAA4CpwkfGAscH67PaZgvmSco+omwEAAAAAAAAAAOBhgYIZAAAAAAAAAAAAAK4CBfMD5hL/2teUe035KJ0BAAAAAAAAAAAeDiiYAQAAAAAAAAAAAOAq3CUqVOfc05HEwr3iZ3/6HREReeOtbzznmsCLTggByToAAAAAAAAAwAMGBTMAAAAAAAAAAAAAXAUGZgAAAAAAAAAAAAC4CgzMAAAAAAAAAAAAAHAVGJgBAAAAAAAAAAAA4CowMAMAAAAAAAAAAADAVbgQwvHEzv1fEfng6VUHAB4Qr4QQfuF5VwIAAAAAAAAAAJ4eFxmYAQAAAAAAAAAAAAAUXGQAAAAAAAAAAAAAwFVgYAYAAAAAAAAAAACAq8DADAAAAAAAAAAAAABXgYEZAAAAAAAAAAAAAK4CAzMAAAAAAAAAAAAAXAUGZgAAAAAAAAAAAAC4CgzMAAAAAAAAAAAAAHAVGJgBAAAAAAAAAAAA4CowMAMAAAAAAAAAAADAVfx/mnCzK2eriVMAAAAASUVORK5CYII=\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_img_dir_main(directory=training_dir, count=5)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "08472c5647d039fe88134dd5bbf51b1bacafe700" + }, + "source": [ + "# 5. Preprocess" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "6368be0cd0b5b944d842b963bdfb9a8b22843510" + }, + "source": [ + "#### During Preprocessing, all of the image has been transformed to target size (224, 224) and pixel value has been rescaled to unit value. (224, 224) is the input shape for Pretrained model \"NashNetMobile\". The target class is treated as categorical and both training and validation image set has been re-shuffled. Some of the images has been horizontally and vertically flipped randomly and sheerness and rotation has been changed to introduce heterogeneity. A part of training dataset has been used as validation set. " + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "_uuid": "5f0a5bf3d06e33bec36a8b9355483d3432d5c63f" + }, + "outputs": [], + "source": [ + "def get_data(batch_size=32, target_size=(299, 299), class_mode=\"categorical\", training_dir=training_dir, testing_dir=None):\n", + " print(\"Preprocessing and Generating Data Batches.......\\n\")\n", + " \n", + " rescale = 1.0/255\n", + "\n", + " train_batch_size = batch_size\n", + " test_batch_size = batch_size\n", + " \n", + " train_shuffle = True\n", + " val_shuffle = True\n", + " test_shuffle = False\n", + " \n", + " train_datagen = ImageDataGenerator(\n", + " horizontal_flip=True,\n", + " vertical_flip=True,\n", + " rotation_range=45,\n", + " shear_range=16,\n", + " rescale=rescale,\n", + " validation_split=0.25)\n", + "\n", + " train_generator = train_datagen.flow_from_directory(\n", + " training_dir,\n", + " target_size=target_size, \n", + " class_mode=class_mode, \n", + " batch_size=batch_size, \n", + " shuffle=True, \n", + " seed=42,\n", + " subset='training')\n", + " \n", + " validation_generator = train_datagen.flow_from_directory(\n", + " training_dir, \n", + " target_size=target_size, \n", + " class_mode=class_mode, \n", + " batch_size=1024, \n", + " shuffle=True, \n", + " seed=42,\n", + " subset='validation')\n", + " \n", + " test_datagen = ImageDataGenerator(rescale=rescale)\n", + " \n", + " test_generator = None\n", + " \n", + " \n", + " class_weights = get_weight(train_generator.classes)\n", + " \n", + " steps_per_epoch = len(train_generator)\n", + " validation_steps = len(validation_generator)\n", + " \n", + " print(\"\\nPreprocessing and Data Batch Generation Completed.\\n\")\n", + " \n", + " \n", + " return train_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps\n", + "\n", + "# Calculate Class Weights\n", + "def get_weight(y):\n", + " class_weight_current = cw.compute_class_weight('balanced', np.unique(y), y)\n", + " return class_weight_current" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "e02af0cac288607e1ae124f8242db9b1c05688c8" + }, + "source": [ + "# 5. Model Function" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": { + "_uuid": "868b5798867b18a80f8d5b12957cd7965a0b65ae" + }, + "outputs": [], + "source": [ + "def get_model(model_name, input_shape=(96, 96, 3), num_class=2):\n", + " inputs = Input(input_shape)\n", + " \n", + " if model_name == \"Xception\":\n", + " base_model = Xception(include_top=False, input_shape=input_shape)\n", + " elif model_name == \"ResNet50\":\n", + " base_model = ResNet50(include_top=False, input_shape=input_shape)\n", + " elif model_name == \"InceptionV3\":\n", + " # base_model = InceptionV3(include_top=False, input_shape=input_shape)\n", + " base_model = InceptionV3(weights='imagenet', include_top=False, input_shape=input_shape)\n", + " elif model_name == \"InceptionResNetV2\":\n", + " base_model = InceptionResNetV2(include_top=False, input_shape=input_shape)\n", + " if model_name == \"DenseNet201\":\n", + " base_model = DenseNet201(include_top=False, input_shape=input_shape)\n", + " if model_name == \"NASNetMobile\":\n", + " base_model = NASNetMobile(include_top=False, input_shape=input_shape)\n", + " if model_name == \"NASNetLarge\":\n", + " base_model = NASNetLarge(include_top=False, input_shape=input_shape)\n", + " \n", + "# for layer in base_model.layers:\n", + "# layer.trainable = False\n", + " \n", + "# for layer in model.layers[:249]:\n", + "# layer.trainable = False\n", + "# for layer in model.layers[249:]:\n", + "# layer.trainable = True\n", + " \n", + "# x = base_model(inputs)\n", + "# x = GlobalAveragePooling2D()(x)\n", + "# x = BatchNormalization()(x)\n", + "# x = Dropout(0.2)(x)\n", + "# out = Dense(2, activation=\"softmax\")(x)\n", + "# model = Model(inputs, out)\n", + "\n", + " x = base_model(inputs)\n", + " \n", + " output1 = GlobalMaxPooling2D()(x)\n", + " output2 = GlobalAveragePooling2D()(x)\n", + " output3 = Flatten()(x)\n", + " \n", + " outputs = Concatenate(axis=-1)([output1, output2, output3])\n", + " \n", + " outputs = Dropout(0.5)(outputs)\n", + " outputs = BatchNormalization()(outputs)\n", + " \n", + " if num_class>1:\n", + " outputs = Dense(num_class, activation=\"softmax\")(outputs)\n", + " else:\n", + " outputs = Dense(1, activation=\"sigmoid\")(outputs)\n", + " \n", + " model = Model(inputs, outputs)\n", + " \n", + " model.summary()\n", + " \n", + " \n", + " return model\n", + "\n", + "# Custom Convolutional Neural Network \n", + "def get_conv_model(num_class=2, input_shape=(3,150,150)):\n", + " model = Sequential()\n", + " \n", + " model.add(Conv2D(16, (3, 3), activation='relu', padding=\"same\", input_shape=input_shape))\n", + " model.add(Conv2D(16, (3, 3), padding=\"same\", activation='relu'))\n", + " model.add(MaxPooling2D(pool_size=(2, 2)))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + "\n", + " model.add(Conv2D(32, (3, 3), activation='relu', padding=\"same\"))\n", + " model.add(Conv2D(32, (3, 3), padding=\"same\", activation='relu'))\n", + " model.add(MaxPooling2D(pool_size=(2, 2)))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + "\n", + " model.add(Conv2D(64, (3, 3), activation='relu', padding=\"same\"))\n", + " model.add(Conv2D(64, (3, 3), padding=\"same\", activation='relu'))\n", + " model.add(MaxPooling2D(pool_size=(2, 2)))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + "\n", + " model.add(Conv2D(96, (3, 3), dilation_rate=(2, 2), activation='relu', padding=\"same\"))\n", + " model.add(Conv2D(96, (3, 3), padding=\"valid\", activation='relu'))\n", + " model.add(MaxPooling2D(pool_size=(2, 2)))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + "\n", + " model.add(Conv2D(128, (3, 3), dilation_rate=(2, 2), activation='relu', padding=\"same\"))\n", + " model.add(Conv2D(128, (3, 3), padding=\"valid\", activation='relu'))\n", + " model.add(MaxPooling2D(pool_size=(2, 2)))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + "\n", + " model.add(Flatten())\n", + " \n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + " \n", + " model.add(Dense(256, activation='relu'))\n", + " model.add(Dropout(0.5))\n", + " model.add(BatchNormalization())\n", + " \n", + " model.add(Dense(num_class , activation='softmax'))\n", + "\n", + " print(model.summary())\n", + " \n", + " return model\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "3e56fcfe7a60a2759f1952f30f7fb05726b022c5" + }, + "source": [ + "# 6. Output Configuration" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "_uuid": "644c92b9756bb9fd1d980302a315d38298aa8472" + }, + "outputs": [], + "source": [ + "main_model_dir = output_directory + r\"models/\"\n", + "main_log_dir = output_directory + r\"logs/\"\n", + "\n", + "try:\n", + " os.mkdir(main_model_dir)\n", + "except:\n", + " print(\"Could not create main model directory\")\n", + " \n", + "try:\n", + " os.mkdir(main_log_dir)\n", + "except:\n", + " print(\"Could not create main log directory\")\n", + "\n", + "\n", + "\n", + "model_dir = main_model_dir + time.strftime('%Y-%m-%d %H-%M-%S') + \"/\"\n", + "log_dir = main_log_dir + time.strftime('%Y-%m-%d %H-%M-%S')\n", + "\n", + "\n", + "try:\n", + " os.mkdir(model_dir)\n", + "except:\n", + " print(\"Could not create model directory\")\n", + " \n", + "try:\n", + " os.mkdir(log_dir)\n", + "except:\n", + " print(\"Could not create log directory\")\n", + " \n", + "model_file = model_dir + \"{epoch:02d}-val_acc-{val_acc:.2f}-val_loss-{val_loss:.2f}.hdf5\"" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "d5133db17f6ff324823b131eacc3fb471943adc1" + }, + "source": [ + "## 6.2 Call Back Configuration" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": { + "_uuid": "104999913bc3bd3f1181b365fff5cc23f9a56ce7" + }, + "outputs": [ { - "metadata": { - "_uuid": "36b7d58a68aa0603ac98b4f3fdcad627b918642d" - }, - "cell_type": "markdown", - "source": "# 10. Model Performance \nModel Performance Visualization over the Epochs" - }, + "name": "stdout", + "output_type": "stream", + "text": [ + "Settting Callbacks\n", + "Set Callbacks at Timestamp: 2019-03-14 23:34:34\n" + ] + } + ], + "source": [ + "print(\"Settting Callbacks\")\n", + "\n", + "checkpoint = ModelCheckpoint(\n", + " model_file, \n", + " monitor='val_acc', \n", + " save_best_only=True)\n", + "\n", + "early_stopping = EarlyStopping(\n", + " monitor='val_loss',\n", + " patience=2,\n", + " verbose=1,\n", + " restore_best_weights=True)\n", + "\n", + "\n", + "reduce_lr = ReduceLROnPlateau(\n", + " monitor='val_loss',\n", + " factor=0.6,\n", + " patience=1,\n", + " verbose=1)\n", + "\n", + "callbacks = [reduce_lr, early_stopping, checkpoint]\n", + "\n", + "callbacks = [checkpoint, reduce_lr, early_stopping]\n", + "\n", + "print(\"Set Callbacks at \", date_time(1))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "6fbe80b2f63f0b9972c1546b088b667331108aa5" + }, + "source": [ + "# 7. Model" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "_uuid": "27dc13b90ff17042f2912f50ef4a8ad288f9914f" + }, + "outputs": [ { - "metadata": { - "trusted": true, - "_uuid": "361892313801cd84bc66a62e5347c20c6c86e6fe" - }, - "cell_type": "code", - "source": "def plot_performance(history=None, figure_directory=None):\n xlabel = 'Epoch'\n legends = ['Training', 'Validation']\n\n ylim_pad = [0.005, 0.005]\n# ylim_pad = [0, 0]\n\n\n plt.figure(figsize=(20, 5))\n\n # Plot training & validation Accuracy values\n\n y1 = history.history['acc']\n y2 = history.history['val_acc']\n\n min_y = min(min(y1), min(y2))-ylim_pad[0]\n max_y = max(max(y1), max(y2))+ylim_pad[0]\n\n\n plt.subplot(121)\n\n plt.plot(y1)\n plt.plot(y2)\n\n plt.title('Model Accuracy\\n'+date_time(1), fontsize=17)\n plt.xlabel(xlabel, fontsize=15)\n plt.ylabel('Accuracy', fontsize=15)\n plt.ylim(min_y, max_y)\n plt.legend(legends, loc='upper left')\n plt.grid()\n\n\n # Plot training & validation loss values\n\n y1 = history.history['loss']\n y2 = history.history['val_loss']\n\n min_y = min(min(y1), min(y2))-ylim_pad[1]\n max_y = max(max(y1), max(y2))+ylim_pad[1]\n\n\n plt.subplot(122)\n\n plt.plot(y1)\n plt.plot(y2)\n\n plt.title('Model Loss\\n'+date_time(1), fontsize=17)\n plt.xlabel(xlabel, fontsize=15)\n plt.ylabel('Loss', fontsize=15)\n plt.ylim(min_y, max_y)\n plt.legend(legends, loc='upper left')\n plt.grid()\n if figure_directory:\n plt.savefig(figure_directory+\"/history\")\n\n plt.show()", - "execution_count": null, - "outputs": [] - }, + "name": "stdout", + "output_type": "stream", + "text": [ + "Getting Base Model Timestamp: 2019-03-14 23:34:34\n", + "Downloading data from https://github.com/titu1994/Keras-NASNet/releases/download/v1.2/NASNet-mobile-no-top.h5\n", + "19996672/19993432 [==============================] - 0s 0us/step\n", + "__________________________________________________________________________________________________\n", + "Layer (type) Output Shape Param # Connected to \n", + "==================================================================================================\n", + "input_1 (InputLayer) (None, 224, 224, 3) 0 \n", + "__________________________________________________________________________________________________\n", + "NASNet (Model) (None, 7, 7, 1056) 4269716 input_1[0][0] \n", + "__________________________________________________________________________________________________\n", + "global_max_pooling2d_1 (GlobalM (None, 1056) 0 NASNet[1][0] \n", + "__________________________________________________________________________________________________\n", + "global_average_pooling2d_1 (Glo (None, 1056) 0 NASNet[1][0] \n", + "__________________________________________________________________________________________________\n", + "flatten_1 (Flatten) (None, 51744) 0 NASNet[1][0] \n", + "__________________________________________________________________________________________________\n", + "concatenate_5 (Concatenate) (None, 53856) 0 global_max_pooling2d_1[0][0] \n", + " global_average_pooling2d_1[0][0] \n", + " flatten_1[0][0] \n", + "__________________________________________________________________________________________________\n", + "dropout_1 (Dropout) (None, 53856) 0 concatenate_5[0][0] \n", + "__________________________________________________________________________________________________\n", + "batch_normalization_1 (BatchNor (None, 53856) 215424 dropout_1[0][0] \n", + "__________________________________________________________________________________________________\n", + "dense_1 (Dense) (None, 2) 107714 batch_normalization_1[0][0] \n", + "==================================================================================================\n", + "Total params: 4,592,854\n", + "Trainable params: 4,448,404\n", + "Non-trainable params: 144,450\n", + "__________________________________________________________________________________________________\n", + "Loaded Base Model Timestamp: 2019-03-14 23:35:45\n" + ] + } + ], + "source": [ + "print(\"Getting Base Model\", date_time(1))\n", + "\n", + "# input_shape = (96, 96, 3)\n", + "input_shape = (224, 224, 3)\n", + "\n", + "num_class = 2\n", + "\n", + "\n", + "model = get_model(model_name=\"NASNetMobile\", input_shape=input_shape, num_class=num_class)\n", + "# model = get_conv_model(input_shape=input_shape)\n", + "\n", + "print(\"Loaded Base Model\", date_time(1))" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "_uuid": "77a593fffaf1b82c2b82bbfd8d5335bc7995a6bf" + }, + "outputs": [], + "source": [ + "loss = 'categorical_crossentropy'\n", + "# loss = 'binary_crossentropy'\n", + "metrics = ['acc']\n", + "# metrics = [auroc]" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "c975187a17a5e9d07ff15484cac2824fcab1fd53" + }, + "source": [ + "# 8. Data" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": { + "_uuid": "6daf75886fbc7d444222398f04637e5cc4aed9e7" + }, + "outputs": [ { - "metadata": { - "trusted": true, - "_uuid": "84f9360eb9554faaeaae16fbd5e9994b9c550c76" - }, - "cell_type": "code", - "source": "plot_performance(history=history)", - "execution_count": null, - "outputs": [] - }, + "name": "stdout", + "output_type": "stream", + "text": [ + "Preprocessing and Generating Data Batches.......\n", + "\n", + "Found 20670 images belonging to 2 classes.\n", + "Found 6888 images belonging to 2 classes.\n", + "\n", + "Preprocessing and Data Batch Generation Completed.\n", + "\n" + ] + } + ], + "source": [ + "# batch_size = 32\n", + "batch_size = 176\n", + "\n", + "class_mode = \"categorical\"\n", + "# class_mode = \"binary\"\n", + "\n", + "# target_size=(96, 96)\n", + "target_size=(224, 224)\n", + "\n", + "train_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps = get_data(batch_size=batch_size, target_size=target_size, class_mode=class_mode)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "cd11c2eb7719f3f866c9e9ee201ca67e556fc73b" + }, + "source": [ + "\n", + "# 9. Training" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "8e335384316089698ee7e1a2f2cd1d6ad2ed4df9" + }, + "source": [ + "#### Trained model on full tranning dataset for 10 epochs and validated on full validation dataset. Adjusted class weight has been used for trainning. For optimization used Adam optimizer with learning rate of 0.0001. For loss calculation used categorical crossentropy and for model performance evaluation used accuracy metrics. " + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": { + "_uuid": "5ac153e86d2338beab9cb0a3693f9c61fe7d3f04", + "scrolled": true + }, + "outputs": [ { - "metadata": { - "trusted": true, - "_uuid": "22c09d411cfaed347995ec8308be13f2d848efe8" - }, - "cell_type": "code", - "source": "", - "execution_count": null, - "outputs": [] + "name": "stdout", + "output_type": "stream", + "text": [ + "Starting Trainning ...\n", + "\n", + "Timestamp: 2019-03-14 23:35:48\n", + "\n", + "\n", + "Compliling Model ...\n", + "\n", + "Trainning Model ...\n", + "\n", + "Epoch 1/10\n", + "118/118 [==============================] - 472s 4s/step - loss: 0.2387 - acc: 0.9138 - val_loss: 0.1632 - val_acc: 0.9505\n", + "Epoch 2/10\n", + "118/118 [==============================] - 347s 3s/step - loss: 0.1461 - acc: 0.9533 - val_loss: 0.1761 - val_acc: 0.9469\n", + "\n", + "Epoch 00002: ReduceLROnPlateau reducing learning rate to 5.999999848427251e-05.\n", + "Epoch 3/10\n", + "118/118 [==============================] - 355s 3s/step - loss: 0.1282 - acc: 0.9583 - val_loss: 0.1615 - val_acc: 0.9527\n", + "Epoch 4/10\n", + "118/118 [==============================] - 360s 3s/step - loss: 0.1166 - acc: 0.9614 - val_loss: 0.1505 - val_acc: 0.9535\n", + "Epoch 5/10\n", + "118/118 [==============================] - 357s 3s/step - loss: 0.1082 - acc: 0.9627 - val_loss: 0.1372 - val_acc: 0.9554\n", + "Epoch 6/10\n", + "118/118 [==============================] - 357s 3s/step - loss: 0.1074 - acc: 0.9626 - val_loss: 0.1467 - val_acc: 0.9570\n", + "\n", + "Epoch 00006: ReduceLROnPlateau reducing learning rate to 3.599999909056351e-05.\n", + "Epoch 7/10\n", + "118/118 [==============================] - 356s 3s/step - loss: 0.1026 - acc: 0.9647 - val_loss: 0.1385 - val_acc: 0.9546\n", + "\n", + "Epoch 00007: ReduceLROnPlateau reducing learning rate to 2.1599998581223188e-05.\n", + "Restoring model weights from the end of the best epoch\n", + "Epoch 00007: early stopping\n", + "\n", + "Elapsed Time: 00:44:54\n", + "Completed Model Trainning Timestamp: 2019-03-15 00:20:43\n" + ] } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3", - "language": "python", - "name": "python3" - }, - "language_info": { - "name": "python", - "version": "3.6.6", - "mimetype": "text/x-python", - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "pygments_lexer": "ipython3", - "nbconvert_exporter": "python", - "file_extension": ".py" + ], + "source": [ + "print(\"Starting Trainning ...\\n\")\n", + "\n", + "start_time = time.time()\n", + "print(date_time(1))\n", + "\n", + "# batch_size = 32\n", + "# train_generator, validation_generator, test_generator, class_weights, steps_per_epoch, validation_steps = get_data(batch_size=batch_size)\n", + "\n", + "print(\"\\n\\nCompliling Model ...\\n\")\n", + "learning_rate = 0.0001\n", + "optimizer = Adam(learning_rate)\n", + "# optimizer = Adam()\n", + "\n", + "model.compile(optimizer=optimizer, loss=loss, metrics=metrics)\n", + "\n", + "# steps_per_epoch = 180\n", + "# validation_steps = 40\n", + "\n", + "verbose = 1\n", + "epochs = 10\n", + "\n", + "print(\"Trainning Model ...\\n\")\n", + "history = model.fit_generator(\n", + " train_generator,\n", + " steps_per_epoch=steps_per_epoch,\n", + " epochs=epochs,\n", + " verbose=verbose,\n", + " callbacks=callbacks,\n", + " validation_data=validation_generator,\n", + " validation_steps=validation_steps, \n", + " class_weight=class_weights)\n", + "\n", + "elapsed_time = time.time() - start_time\n", + "elapsed_time = time.strftime(\"%H:%M:%S\", time.gmtime(elapsed_time))\n", + "\n", + "print(\"\\nElapsed Time: \" + elapsed_time)\n", + "print(\"Completed Model Trainning\", date_time(1))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "_uuid": "36b7d58a68aa0603ac98b4f3fdcad627b918642d" + }, + "source": [ + "# 10. Model Performance \n", + "Model Performance Visualization over the Epochs" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "_uuid": "361892313801cd84bc66a62e5347c20c6c86e6fe" + }, + "outputs": [], + "source": [ + "def plot_performance(history=None, figure_directory=None):\n", + " xlabel = 'Epoch'\n", + " legends = ['Training', 'Validation']\n", + "\n", + " ylim_pad = [0.005, 0.005]\n", + "# ylim_pad = [0, 0]\n", + "\n", + "\n", + " plt.figure(figsize=(20, 5))\n", + "\n", + " # Plot training & validation Accuracy values\n", + "\n", + " y1 = history.history['acc']\n", + " y2 = history.history['val_acc']\n", + "\n", + " min_y = min(min(y1), min(y2))-ylim_pad[0]\n", + " max_y = max(max(y1), max(y2))+ylim_pad[0]\n", + "\n", + "\n", + " plt.subplot(121)\n", + "\n", + " plt.plot(y1)\n", + " plt.plot(y2)\n", + "\n", + " plt.title('Model Accuracy\\n'+date_time(1), fontsize=17)\n", + " plt.xlabel(xlabel, fontsize=15)\n", + " plt.ylabel('Accuracy', fontsize=15)\n", + " plt.ylim(min_y, max_y)\n", + " plt.legend(legends, loc='upper left')\n", + " plt.grid()\n", + "\n", + "\n", + " # Plot training & validation loss values\n", + "\n", + " y1 = history.history['loss']\n", + " y2 = history.history['val_loss']\n", + "\n", + " min_y = min(min(y1), min(y2))-ylim_pad[1]\n", + " max_y = max(max(y1), max(y2))+ylim_pad[1]\n", + "\n", + "\n", + " plt.subplot(122)\n", + "\n", + " plt.plot(y1)\n", + " plt.plot(y2)\n", + "\n", + " plt.title('Model Loss\\n'+date_time(1), fontsize=17)\n", + " plt.xlabel(xlabel, fontsize=15)\n", + " plt.ylabel('Loss', fontsize=15)\n", + " plt.ylim(min_y, max_y)\n", + " plt.legend(legends, loc='upper left')\n", + " plt.grid()\n", + " if figure_directory:\n", + " plt.savefig(figure_directory+\"/history\")\n", + "\n", + " plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": { + "_uuid": "84f9360eb9554faaeaae16fbd5e9994b9c550c76" + }, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" } + ], + "source": [ + "plot_performance(history=history)" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "_uuid": "22c09d411cfaed347995ec8308be13f2d848efe8" + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" }, - "nbformat": 4, - "nbformat_minor": 1 -} \ No newline at end of file + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.6.6" + } + }, + "nbformat": 4, + "nbformat_minor": 1 +} diff --git a/output/figure/__results___10_1.png b/output/figure/__results___10_1.png new file mode 100644 index 0000000000000000000000000000000000000000..bc277ee56e86f50ed1555ca7a9e2e8e7e26ab5ec GIT binary patch literal 90016 zcmZ5{1zeL|`@f0;5~4^+hjd9vgMfsj#ApUccXvrj_dt+FYIHXUNDHGzGbBgHXa?+$ z&-1+h_xZi!vm4GC_wI9@tH0NEB0eh1;$o3vJ$Uc{_k-MfwFeKLBtCfX&;k?ven#YR z+r<5^$1ai|G%)X9zL=)r_v5FIayl*#9$@4Dc|V+3^1r#C6m*r=c2##Ucl9uKHhW-a z?CNOi;A(62iQ3)F*~QAio|>EE{)dg)($&>bh?Dcb|G?qkY{40pJ%#T%F}+jwE332v0UeYsSRcqJ8)ILGzo$ zs2|}Ebt2v!-)0h6MJAUqWz%5N&gRKw_Vh>zjq-=nAMbg!2otN60{|%9Qt)-!oDXF0 z3dOORc2LpiUs?e^s{~AT90op@#H0!N&*k{5q3gqcjcGy_9R%VP|2^~lDMsSc|K33p z!n^X0``_bu268!u|No8mpG<_H4gPy9neIrP_Ma#Ixl``tv)3>Gy$q}FK+}JZX_P8r zA7>R>1!z8@vGd=CHhy_tc~`)19JEhyp`skVw78%|d<7#Dw}W z!2b7=M1WiKLH}pPNm3Q;15r{_9dw@ZpBzvkhJ^hp!B`eF``IQUbB6Z+riPM18-Z zV0R0W+RCZU7Ss(Zi!rW4v8y_MfwO0+T74s6pZp{pd{oVb)U-93YpfTk82nvn3!Z3Z z`vTAZH{%hXa%ulPmXu4{4!?FuFO;PtydBK;Z=RJA@7`MT#K;~fmILNd83C4pg1pkk zDxKamx-f*VUC|yanY5J_73_K?D)@kOC-1AvYnFufY8DBk}?{S47x(z-6$aW0H9 zxSvAZc--w0>*N0|k^{A2XoJr{q#lH_9jc+NS@Z*{(PRyT__IGGs)Vq;*7i8MI@R%n0{f^`vAVC>|Lx46YBTbm@jk|zXozlq-DCK3Yf?KODp3Se*B|z z&M$^*+O}x_5%;W@asJd-i~-jh_Y(^vI-pYP`J!VFVN}+%nhslqC*1Mi9l)lulykj! zehDJ>n3b=BDa*_!ZXvHJ=D4zuJ~VUltA*afHiv>CFRTI<-64iN>!%)(I~H@mcUEmb zZ)VEd%$bb)5IGJOXVoG-CM28(l=2&=yh_HfwPx8@Ck({E2^Mv%%76_Q_gA#Vw*LKJ z#EAYYq?k1H_v`yhh}Nr5+EM%J*PSgXLV;ZZA+}sAiQ*_?luf-#bbU%GyWn{gYA+k8 zY4i^pUEt$*uMdkR@D3@Lo-k4k;yXeDBxeiMHCR&ByBlwm`-!cP`_Nx{G=YFtb$JQBhI3S+0Q8lQ2tdv7PXiTkD5Gd!kD=dFh0YM!u!(Ayb2E@F#}EElEv1r!ou z*Mu(EN!M1{6>%@xvgQ?8VIehuX!-*d*~4ujhC@U8EE%XBh3p737U2QZT)+Q9tjd2y zV1Ul(0OMa&ggi%l^{R?Al>ik97hCiSRxm|Lvyl<@f4f{VkPZ+gUg^TY0J~@6fYdr9 z;Nn6GvxGfiUMUTUb(@vCAH1r419At%dX_hJbcmr|cJ^Kc`QxK7Iz6(U))^KYme=X| z-(}xTS0~Z^BM0im|KnBHS6C9IbxA|t=n*uBTgv#7RN9oB67oeeCHFEq<5$U?;SM-h z05*oibJm@vTCxm{9^aBF2)Y;!BE+wm&V664Nxt2>&;qpPNzM;lGR*+!d6=?#8M(pf zran*3@UOo;%Q2sqHlRH0;w^Gku||f=7gNtrr#NvEE`g1nN!NkdyqSM@44=Ha=R_OM z0Ka<{#M=03TFfBZ`m0ss@^raAy_TbIEiOZRwg~W57PGn4PqAsaV3yC0V4S#M0?V?|Frf9iO5aNh*up zDIZgg?yPXEe}s#8kmyHdwuR<&K^67$sB}5Nd~Do0tl1N6&@P+P0Bd-PM)XaXB^mA- zvsob0FI{PfL*A?Wfx;k#g0O#p_QIgMe`KcPviaV`(Kb!kko`{Yy;;8MTDIK0_Yfg% znCr;P(`}u==-O{=&h^KX;$=mXC32kEI`sS4X|L*)^ay3f8U~ZTYc@=c{7pwZF`nMe z|B&P#xi(ktCDvHga!3g{QlL@J_DzImb-4zu` zs2HfZ{L`tkINJ2gRRkv7uUqiABV7W}p0Cm?UyZ~1Q{l#w;7s0`T~>tQTmuT4SUlz| zUxC6KE=@V&vPnB=?(I%IaaT6#fKdvwL05YLKiAe{-Pmy?nB2B>p*>IEt3lAnaT8z8%jeM{R#Ux@g~>VwLToMc3#GIY8?o8 zV$fe2c3d4Iy>1r;D-Ur)M7udV)-0AxmXYHs2fUoljeEjcCVz?Wc9;M3TH#LZM$V;u z z$jbTPF=6q;_v7TlFHlLRP-4#O7wDb6;EF)UCh-e}XtBr_r=qj!97XD^Cz2r5{be0Q zf~P`4M++CfEJZulegX;HK5+Daxo&AlLO(vwk#Ks8nit7mvGV&?dO z4g1%HC-cPOZm~Cx$1e+5%)KNsYi$3Dk8LTuHR40CJ=!v3qvzC+|F)0)c4thoS zh`bV9h@0N0h5b-ZDLK0&(YFnVhLP|4rQDFk6ocle&klx|G;FoJ61zgCGlU6~k9@DZ zUi?w%c(K)Mz^R+w+k@|t-LnVj#yIZ-k72l{`T?05z|$%G8f zuUhPdj96ycij_t7nI6>*??H-8&?Z?z$703v*!&&hCDmxARgus)n=vbGF8CX&oT90? z6$Z?s&?kOi=k4a9m2wiI@ zt!BtZbDm^+{Fg`5pcbZ0-jazIR*<(xUYn8YxH|?NDeHmi5MKHIBeD#+<(+j>+zv>W z10zVC-mFk{8=uud>-#k|$9ZaVSGsxYZMRm^AD)|L?7yzix0xx0?%7l1V*+N0H0Nn9 zYYzr{;1W205&$o!wyS;49h8tKK|n>GO}d}FOe(mVI`~7M(U^@P(lx<8T`l{DDT_^mf!?LG#r}5Gl0@oh+%k9%iNAn!5DI4{#{nSS9g4n-||d7)&Ck3D|Dg?Y4c0 zPVydII3FLNA?Wnu0bv(+)f)^wV;*o$j|8@g(_Tr)N!K&%;mBYo?~Z8EUWA<=2jhyG z2K*qNqxj|S!))&5YtoGrHe`l~O&SxO=mtNmQjlY!spxaBgPNYgjZskSuJls<+DX+| zE6zQS*KQ0^h>Lr|%WRg!bY|2ZW(1H;dYrVilPu>Zv6BWBoxX1!2lHC4Jno=zj@63J;%S92=el*`MZ${5ag|ugXszk{mezyi8T*#YUYGD60>18=e763Na_ONKF z?DfCB-Xi5{9VOotHEUT#k9P7jE8Q7nyQT6kC3rCyV(*~112x>{{H2^Fj`!(L@Hh~a zZzfl}h*?rfY_YMdIqxOZj-KqMh#;Ud?x1GiU`ZKbj0o}cop{pwsAclc)(`NkwRR%K z2^VN-6Ay1nQAbqEEVmA*C=}PdF1NxicE5Dncs*-b&~5zoyt|22GRmX$4YH(*!GaZhIf9IJ&bYt4pjZ*_GNc~*dp;xy3o^h&ECL0^p}+6e+X(p97DVw6%in-aqvc)KM|JbrCW%dC3zk z*~-SwmBwC&=j_sIGjHp@RprIDr1a z4C0afEfqo(SWAQ9L;MbMd(ZovJJ;BUs-oN&Op|oR8*8dPx6!wA$1N8cz!ICyPlm`o zpt z1GaruIp*vQ<=p>da9LFMEbvkjuc_{*UVFMjC-^$Qk|n_;BZ)~OUs8MG6dm)u4(VRt zvr-eY!hGpEKvz~!b0y-;3U`$J!^zeZTdwB&o@Pa}rl&aj-q>y`drZpn>TCh!56RC$ z?XrWPi40Q5-|g(^!fO=Yum>c%%tzMj{GQXNl$CDA;P_Y=tnX+S%d-A=PT(~+_Li&c zi;Ld$FC$S&hue%6VR}CaE^z7<^mw=9SAid|TjH1~s=v>(9vw{R?218O@W^twYSLnkvu62LsO7uT)jR=+h<;ck= zyoCk;I)~lo@GiS0mR<*i1d*=1qO^=R2iehK3OLr_6ndeS?GK%a@HPb#u&Z8e*7Vds zM6Yf){>dKM`D+N3-^odfLrhAJhu;x@=~0M?O(M-CRfr{DcW-&J5p%j(SaWf_6ee`< zF{^)vry{16@tp4i3vJymODkm_WK{;h>a~&zh2oe4 z3fAbbrWe3o?MKC?=f1RSr;b0>IDE{`G(U>B>|c=`7(_91)XypO_k3H0MHcnHn~1Wq zdow`qKWd7^LTKm5nQrHRFL(_0H8el4ADR>x&4deF60HqO8qF8QLWZaOIsVyornV~! zL(T>>+=Ud{#zn)12t!nU<+tPZyO_F_@gHZY0iM8b7)VoCnR=x!P+VrL*QWrl)Vfu7 ze6Jo=PzDTmIedgobyJ-7jAq^bs?@=kJ0w2tp+@53<;>B!g_l$?tA-E$0hvg>H9PkER3}(QvbhK!*K;Tw=B}yWpFJz6QMyzE` z_1%ql=LwZxn~scgv=84O1a4iB23sji8rSv{wT_wfl>H?u(VF7k^9UztGTH^oX|Rh% zOX=K2J2xUiFR$&mj!a%5N~E36I1x!RwvUr#%hL{CKBqK=`T$k=&7(%OK5Z!DtqYn`AJkaUr0 zYX2_bYrE`$wdv!;Iqi4`6)M6Ze4vnQR%gYhmZtnE3V=!+xYB26UA5j;xBcaLF#bbA zC>^}SOy}KPd$%ic#J`ZG^qZKC8I6Ykka>vSkD2&-Anc~cROQJu+z}ro3E$_xJ4-!n zV$(gn@V}D!8{CZdfnZ@T^%-WA7ilb!+V=Fjx^}5#bm=y`C}gkaOiYr2B^>&mOPY>+#v13mRv>QnKbqoN}pXHRW}GiO`?xb)@R|6xuab`e<0- zO&{_mA~1l{-;Ly}II*GZpVkscHX}rAf2K`W_-wWE^n_+8iqMQQx>Vid6uR(K|^VM8J z4@G;#|MGDs@8NdFCgaR}5bfj`tz^R(*o1zJo zSwWZ1tJraxE?4bg@=S5NZlvRR>sigdhPd72_hWqLN;4|1k7ByT?k(RB55_Y;7~9-7Mb~TDNUQ>EDDfE3UkbDUexv zxLc65d=30c=igg;((95|bQ)vzx^-`jZm`mC|4-NeyNwI*5qNh07h;O@?Z%>6Z}w=A=AzhMXEKD+y5*RP(QD43co(4uj-j^hn#=x0$q)c?h@@wUU z3mX8)h4nmY-nujWtd_hi5{DNr686I8a-{F@v0`FqX`j$!lU$`RVM6EZwHzxyiKGQG zwjGy4H1kFqI}iCEHAfQ*$FtpKt{czEE7W!H%ng4>1uLxhIkPoU6d0(1=SxLmuL=87 zu|NZs8|NCm#Tui8eoc6dlf%<*GqqUFgjn^@HEERl7vxPj4OR9HkT~+k>(^d0h7<9T z(k#zybnQBpTHOb~IO$yNB^hY4^L!7+a(CJ&Kb}?t8I*mNm*#GM;ML%2T=8*Qm1c^E z>C_8S^j?C_eBJ7FC&Vm$g7^C+Yb}O*k=RZ_lzNS;M0J_+(M>x)K=Laz19ogu`?!gk zq?v4|;lVEimQ~;9>jxuoUAB4Ct>zR9Qmsn&u{{572`1fgOX=<}zbA@(B4%dEQl1rF z0Ty-bGMHXsZFh`@Z!bJ7SilZSb8*X?^*DI)o9pi`Z}RPkZ$GEr64>r~+C8w9_3Vi8 zaZ_<<*iC5oX=%8NT)uRE0len#Jk5>6g-hS3NuD1Q{#99_WM?^er1jZy0AD0H5p1Gr zfV0lwwQ2Ti*0qF7-YqqZX~1R1n*j1VrN}U~abKq1IhYCid*3plGzK*rL1XNR$Zoa% zM`@0=v}dnT>yQq7$znf^!64b|cjd9x+JrD7BSgQ83M&Rx>p6_i5j4F()U5`7wm#Rz3$`JEK6@!bwMqgc zUycH+-VRC%v1*#OV=b%NZr3MT=-*Uz(aO_!3{XLC`!h_y2+=>J`P149@C3k+3tfpOW{iz5n<8`V00$hFUcw^ZG&3G;VBJ-G5!mxUN$|D^CT?4t?= z?p$6>2|K{vrjkPeX(4_zzCf7K=GevzfeN=M+^Y2{tT>9MpXWr3Ry&%bc>~_7=!gha zO?<n~SQdnO`ju;@|H%ssq@@2_oWb;}D}E$szowQBYI@#tXc^9ltCpoOUOOxT zc|PTi$I@S4rrgA$sfi9S@*pC+kp9Z_WPXJ|ju90_Js>8|$K6Nplj+OBmj)&_{k#X( zz`SGSP^=H(0{a@Y%Y+bD>bVmWJ<-?H?8_GjkkYee_?tiiDnk;^!+ES|POn!EQ~Wib zcPLJ0wy$%ki!kVD?)}R#M%GFbiby*FW9SqGOwvLh&@BGf% z(v2auBxRn!Q?5??SJH$;{H4g|WTalBa$E$r&6nGCAmYA^CV}ejrw^$3P7^dxdpT-& zJ7NzMzZ@!pY}>aAM@a%_k}uZ9AictD^7&27LsYgLbyYhL9H!ClIL+!u2hUNYk#A`B zzU<7sc)dK+q0M?JDd%>4wK!^pIdFJcl9ovfJw@;Ee@#@t`rum$%J-RB>beo0XL@6q{+hJuXP=phy~jM7g)aer zeCEAY|G8WQ&2-$}dvx}(m%)RjhqpgJ3W?ur3TB$VLA!}A+ zrc2gxA~jWBUk9a|WM^>W{eIP_VY|)@k_c~vWvm`E zkwkMW{%kk0FHW-9ITOEeO-L!$SiTJma6Y~gA<8PH6%t@TGg?0J~{iP51ofvw<&C288wWGhD_jD ztKj}#>Kua9RXlbmg7?nY2nk<2=-xdqVMEouyUvpbxpJeR?9$JhA?uLGi?!@a3J}6< z@sYuIH;}^*32YtP8=}kZA2)4%3Q1YT2d>ui3GsZ%Hz8(c{?;b%cZzknM86|EcE(d2 zG+9Ns2M71vZfkZodR~#Z-G0pMVOY+s&m!azK6Ap{VWj}9cc(1jUQ78tot@ZwK+{fq zELS1-tSUwFEP8$!5~*#-)+ao9PNB<^cPvaKc^!LS(NQ2YCkoBzcsb`&Hq>5Yd!nB@ zzQB;Gdj4d%!EVvhaMHyp-?dLGBsheF+bNGbp@))O>rnG1jnOVG||A0Yp@f4#g4++wxlo4XXqG z(+e=!jB{GHu<727*e=Mor$JP~9{H)y-kuDx{dSAj!y$i_a`qx0hjjL0?(p3wc%7)U zNMSbaYQdA+923l)Fe}Jng5%r20ys>s!FTzm1?VPc1+2IDA75Zkc9be)RYg9)2#sq% zzG7d~ZF%(GRKjKWVnA(}a_!pE zCUV_oSnVM(>b=XrEDgHUP|hqEe6hR5w-{?M-2aP*LN4ofU)j+r=7*HnNW!YA1D~%y zHkzMle@-kb;;ryS6LR)-VEIh^Y%(=f?#9Hczx@VroAsz!9&PG_=o7~`EH%^ah7wG0 zhJ}5X_ZYibd5-CY@(FBk7Eqh7qelC4KlP}}bw)>h7Rpn-FqVgdS_k<*iI-=-Xn=GC z&Rp0)hx#o18<%ZJ)_L^=xxX|K__}>Ah2&r>=Q-)8_6Cpr(kq+_mC0c$B)Kd`ZB_PL zZX4)6os?eZ4$^H%4emE5N`22R$-jc}XD5F(3#76^w`3S?zB|l1#$~3n7sr(gt(-s% z{Z{b8hWQU&%t4^GVdp&ECIO66b@bBb!@t&3u%G@?z~x;?KJ#`|RSxhLJ%BV8#gfrZ zZpa)#y3S_m&1Q-bQDIKF^?@WsrpG*))8aM`9xgcFT1e)(F*2t~PyBCpvzaHd3z==V zR+%~M&npaX*?IgBdzX!!7awj73;EY}CIw*+Wc+c8eBjtSE3FwwVuIi;XTaNs>@!3} z?GIzfxf+zmRSbZ!-s8sNajq1LWtJMEKQ6>GI~WHER)FPKdHgQXFN_T)nw*6-!o3j*)NMoMuwUN z`DfEb@4l*X=1&oov53$Nl2;T{Kqpo^N9;5opRVfJW47A_&3M8T+)j86KxT78(g9j~ z0ezmsHI^3T_!{!4TCS7iWRcWB(?esT`VDX}YxUZ9ev+GVyH@#+fYfNKpv0-#b=pSJ zgB1`&_yE*;6t{J`F~yr1`6F9iOw-CMSP;m_4H|Cv>7(BEL_u&&-E0kfqRA)(IUQ0J z-^ChPMXK(kux5Ok)SNRBbZouwO@Vwd(EQs(-wTG=vjf>SM=0mB#{w(Wc!0APxWV3- zgQKEgs2kx}aIok|UB+Ga&M#YOM|1iE#(TNo2>jCyAvo0V<@wl{lh|wjEVfW zEai**+{Hd@Xg_G+@qG%Qvsg{@# zkQ8{XKV7Q0!nSjxLktlGm@PKc&svcFN{XwaJ#16do=Gr?ZqXi&ivWv;cvp?8HLUS8 zkhxWgC|&I(E|fjGud}YN9L+Zlr@8q|^MikyA5R|jHHwq;9_OUcq!ZO++ zY{X~CrSyQ+9jp$k<&bCHX5<$KV?m!N4a4I^#HVj=i9CR%Y^yBb71w2G75LCVjYvV( zFuI8y>O}fd4l;%hCCIZ>HdEY|yN%cAE4fjFX#u@9GWyS00e4!u_FS9Ok)L>hoHIJN zIWl+7U90g|w2JU$N;w+7nc@1oN^9ez5`P%eu37WNRlAKccsX|e;mLJ*BdY}_u!SN=4rW*!~nw;bb7UdR469PKs+mX`%wwp(~R zBLH<#%J&kUU!&o3?B}mal~EBjSZcL$egc{>KU&0JxrLr%tkV}$s}D6)Fe}M^V5j{F z{HFG}YVvH2LfNm-B9>)L+IOQbw%0o`Raod(3h~=>GxJRI_~f#QvCN-ZHRo?X>3a@T zmU0UbG#K428M^;Y;)Nk(OpTgz27AxAm2hWts zO(gf#6Xc!0xInizK28&&I~vW4YRa^HPH{OHOm#6$ zgmV(&h_u4o6${~cI13-6r!m)wX7K^IIQ*8yG~qM&kqA4k_W z=nH$WV_>rrx;biQ&3E|v!&+ON zz8(eo>DlY`s4#D_#PMr;H%i-nCN4XisBK;U54npLPtLV5#xyCAM2s)0 zoDt5GYXAOBsQ5v-e(fV9&ur%~Qr%l|ALtIT@VI_nuRw?7Cq@?R4bPQWh3RZ($xRa- zKMrBSEWg&*X=y3^r3qVJPkAe*z*R0CrXXGNOf2bQa*eM=l$dG|=C8F=8I0!EwyuG+ z_xYJWiP!cJjXa!$!u6PKc=K~uzBA#3kyi8Pk(PxH(BzaQiD#ht-c^K{^D`RL;Mb!+ zv{{$3q&yrJ&V1fpl3S$(W!|X+0@CXP8rZ_~~0_gR6ADp?`VhmhV@K4@024BI^DL+0zl>K)TX z>dWD6KvT|*P?<)JhEp6LbehD=E;uBLcw`K%ZZY?JuTgE4(PKV?=O32nwoS*~R_HRC zQiS3>U(R>DoES98)#JZv;!~mbcBxk;rL}!he$@(?L zAo9sj3~m64^5ppK$CR7=!x9%~Y3@S9>?9Oyb247Wf;vOFsKT%-L8wnn7AVPDp%wS* zWRIo!#IseMnbXEBrLAwq){*2m28jrwxdh_B3gSEzKtEWBGhwg&vOd;P#ll{PZ zBM0IvGA9;wGdhDON)SDYXM8v8#JtnhyvRWZ`G$GZNqu**UoNggNnePi%b?6Yds{)- zXlFM>#1q2?yK-bt7RY+{V%qN$5|PqNR7)z)Mp;7)@uon_?v@BwRl}zcZB$RIn-@Bu;(b zn3{iJB=dwn>zkqtRSW&d6EQDJ7T%IYO;C58?z*5^|8-u2b_>Fnd}ZQA@NWa=M4tt^ z3aX?i1Gvb$GoF<*o_drK2Yc3m! zcQy<2Gu9gV69d%Kbd~Jqez3&2A;!-$uJK4W>01^#L*u?r{&vc8kdtN1%3A35M2jciWak%OjlKj{% zHksgHokqD1#I2Z}Z1wT0O77qxt5#Qbn0`6hG=foaBQ#0*t2r6CTyAfH$h_|nO8hJF z`rFgpH=dIc^wsIN60WSRu3_?2)kfJQC1i z!^ECA=;E0skzsTSojo>Oe3?`i?V68GsDLF(f0e*}x`_FyX#wl8US8fQwv6zL;16wE z&I(Jpk?9n}mk8{{Lv!lpmMies$09EDZ?j-Oq1$d@z2t=kH{Vm_-<|o9F(~=Gjmc%brxmRayg>jj#UBgaYEs+yQe1?P5f(l1#)l66t<0; z?Rp)^v1vj|nXepSg-(2h%C|}0oTaIhJp^T~tDNFqz>iXe8V6}VE;q|hkep|MY?rkD z_vPA^?a8A1Hh#15YyaQ8(z|36nh{~NBR6`CsLFB1iexGOauEbvs=Geof9MW_5nY|*Zk~F~v^X+2Y|XYuKQyJ*;p~e9ho3m$ zGX(FLr+%|5((1%fcds4d<>HvA-8*M#g&k+rK%-`A&Eeb4W?9lB>et07Ux@tJ^127r z@685(Kb7$zsTY#7#PYc1A}cs+=i2hJj!J_dvxHs_`QRfa0Z#+PC> zQV_9@F{>O0cPL#`TD21&#w`*z(=1q(6ByPn4?I6Q(84X#4=u%|LtVWjbUbVy0+~S{ zy-d@^*i8UaCUV>P6m;OC9X3I=w?=+54b7MIN7lGoDs_^o2I?8Fe=T~$4s-KWHqtes z^a!-6A)TSQrRwP`7yRxJI(O z76u2MTlh3go}FvWh~4=gk{NQxb;8l8CH=Q{Ywb!nKHI0cPr8n&%nv&8 zk4|l$o}6$4YJGEXV}Jjqdge?;;c_%&7E0l!z{M@bCG>2v)a4)#Mn<9HG=p-sJfq@L z@)P41sT82oM`~ow?eC^U(;by(IQj{cn^Vb9MgM)sAzosdNk#5}wZ_Ehvvi?<5xRRr z)7SjEff6E4j#z1aw1Uh{$Nn5X%n<)8L>+xb1%o{f5fpBY!z9sMS*f0+f$N5W^Q-*X zBaY3Rqobu%#Xed$Z^Z+FZ0!|@~3xz78obTY;u}K<$7VG zZu?gU2-8ymX{n2-2QLb>FhSxPwrhAyNH(_6r7bsEZJkz@w`4t44zzPu(C(<(oe?c^!vh9lb#gQmD^ zLD|o)2R^I5{ff4sU7ev<#?wu}a9q#kR;u%-co|1w8?ef2b2A+a-i!xkmk=Kos~uQo z8`s6~%v&^vB)lr833&tAV)S+lWH0&!ibG=L<%ghlz4+ZpuK%~>hxS-r`q z_S-F3V(yU($OFf%8;n&<&I((dwdZ`uCUQ@D4P)H(t&W1XpH``R z?3;h44TB-D;{vYYH(Nh0bBZiPNr3Z|@2w z{Ua&LUnPS;LI-cng7#GvtW}i!O1hy6%aL#p;o55QEbQ%_rY1R4xG{hRTO{ zRxY0zzbSD@@z6>G%OeLYClCj0t?dM(e{5#Q_1(zd$BhP5WpMJ#*V^a4C#_|fYAA&t z-zSIp8ismR~XB$TR9NL`{1r^P8)7Ku-SMrO6`uRJIw`W!UZq1-@O11KwF?Zc7 zMSvEYrf&}cR!y=qme^%a8CgD++#S!cblVrK1oSWP)!#ZT^Z&tRL|DpXq%fV_jOJiR z!SvWnuj_iM=CNTLR4$q5;B6?CGURiRx||EW?QYM!jfOTl>3faO3k}UNaS6<2bh=dlAu`ZIhqX%% zK=vui+h}$>(?gVjsAOX{`p-C~+IKk~o`>MuT(QQy(Qw{>R}Jpb$Y$3RG+&#vr3RHO*Ls9P)XhYlErfVHF`TLPjmKlWflogS}Ey&t1ROtG1Ds=Tn3&5c>x*u2(Sj9$I8HJI@?z+A<3L_w%V@6y)tnlAQJnI_9;W-I zti>0l#TbCc^tfnNW3Th&uV;sU_^T-!^x}OXmH++%EiCor@3N%QI=*#6T-qER zdc*dYy*5CuY_QTsm^W)kJZ#xZd4#<=PTU}ys2}0s78JIfSSq4w*9tn6#&Nq__$Ycf z8bo^~W}`>Y=v=|Ciu3Vu8i2;lJ5}3v{0MjV+qa^*YRlqS=@u_mU}-5mwhP5c@vd^S z|KReK#vUGxr3uwCa?#_|8$sB6d0+q;_sn9Ra>~ugcr{dvon|hv@ilqPdMJH=UPmGT zb>{aI8DBb<^sr6rtzqV{XMOTg=&E0M{}{W$ow&QXQbEI1U~es>8=%d>#Y-G=yh@n{ zB*mbMdbNVRG%A(K=NX=gI@u4+e_MHx0&PGkII>&$}i=_&YH9u zR27;$GTUv!HK`#82DIng^#sD75@TJNw7x3Or61^kO7h=VmIvq^FZjrWYNR-Kdf@sX zRiQD{L*E@^HRs*R*}hFihbFj$!%5ffz66|nfgfpc=Qxcb#jiMF7AXmT1JzMO0%hn) zC&zvb3&)hQpbmC^EiXauZDSG6Y}Mc!)w3gmQF8MA!&pOXUIwbu+-MWDSqrbdyIRT@ zRt%@Fw%10T;%s+$lA3W`duhf3`ilmmW=i!6t#)Gf!;RH?x^#`&kJG2We!R_k2UFI~ zdgKuTi#~_x>cPG$lyY>NZ0*jG^`A_Q0f8_@oG6$_??Oie$9|hnb!F?b@o}?s7Wr--kHj*(^;}Gx(cX?BWlFWZ*0Uki zf<6s5$I% zxrl{f)`FKs9+pz8y9vWZI*7X!&@_LbZvc ztDdgr(7r{#n%%6)K(6ET#(nMuG@r1uelxgevwA{MLdeS#a$!6ge!W%fq}GaIR5pCl zZ5vmdGhn7M)bg=mmopPL9k$?%QYAKM7fUBT*g76R0{U5oUR?i@9iET!sN)<%j-F~P zy*$;ir*o#p3n7#yt?mdv$Ufg<)RYzys-@{$fcI(l&6Zg&29&&GD3aHM;R?c@#v!r0 zh88PYy=#BPb!c7LqSoAk4B}#b4fP|rz;cuTKkVtoe5Bu2@mD7#7EvrLw_U3VTNq>v zZ(Q-}K$YTyuz0FXRh&?w1*BTR zH<^-y!_|T%D^+?MVkL+~Pu(~`G-Xdsa2 zZ&!dxLwPR(eX;x%z+!NOfKk%+)6o?^Z^I34-4xXTfGGB^9S9;@Xy9`UoxsyH&hXwx zVz{G!eh2-=W|OK@drUd-gF-jyeM!OHCn3!SzvZ5m6#wLpd<;9*4%Y3zMXAoHqy6^Z z1gIn|ws#&|g!NW3G%?aVHW*YV-fii#cv<*+*e%JWHG-&G_WL7G4NH@;|MUW!2*aWI z_s1dCnX+r`V#H!TX}=Q5I*(rj{Af*D+ap7qMzYHo~UVVgCMGD8O) zGf9cIce2d9wBTe^&g~bte9257{4X~k8By`<&gZdsev9moVM)Xs@v6gl*3XvpYaxyb zcll-9tBOHn8Fsf>s&0pgfF2`6YNDB2)4l=(tk&-n_AGgL-v^iq7HL2KpsZqyFX=9p z_E3f3!?$tG5|~o{=ug&9nXDIv4 z#ek!%;q(vAl`8y*ed5mQ35S^5RJOg2IY?)=*Fv0+<4sdsom9&qPi!g9>D)TMLr>hB z^e?GJP>fTJABb4J+O4v>c5dCARNdxN_xDZt4?0+OH9lY=C1N6m1RSah7O6j{V5*FBx9sbH z;&iLI?urSmr{H_?uLR@(MqE^TH*eAN%EVfwP)at<(YAl=`8}9blJ+H}m1#Fx;2x{`R_wy5S7>oiUXzVO83x>k+eX=l53TthPt( z$CHYOKA7!*?r%d9ivB7V{DHV`xMoY!w2i;0s;zzaJa@@!2%=n1EWlvHTuP4rN7Yxw zMcH;=zls9VDxH$j-7p{_ASvD5-7yRyDj^NhF(^oP4PDYLIl#b-baxCe#5d3Lf8T@e z8Qf%E(d0HvK!~2frTG8~&$(ldpd_UDk;bHf(#Tjm&g>}4!LU*?&{6-GU?K|p(4Rwt z47Qemjiq*Qut<;%Mpl#J=9owe8Q>Q%)>fMWBH9OpovLSE}$@z zx+MEyGr8Z);*1sq1+e@7urRa$hsD2P!HoJIq3o-qrIZ`0v-y+MsqsVqdj7+?ZUOlH z&uwYknKKAr^q^#OG)5##I=T4!th684EDPNA7d)q24N?u6xnufq_vsZ;#h=u0<&@+- zYd@FNZ6KHfH?@3Y{yP^E1Z7Y4BSf($6ZF+t8>!453qHC!1>{9j(A@X@qXQe$&C?)ywc2eG!@u z1te01JjXU#BtYBRS(ht6N9HTyk$C~YWCnN9oZd^I-a|n>xk)UFK`fu)$@vmwe6q{e zwKn9H`4}wit^upf4WxRB`Hzt8I+zD_t(li%W^6p`v;a}RX|xZwq@oTremJN+Kl^=o z)im8__^rbksE=^dNpBO7rI&<%d4XbblhR>A=$oOg?4u9&7L8T`F=|)ZIi?%}6LW)Z ztWD{noQ12CnSFwz74cGYx*!8P=e-5pN?Dpi5nTa>a;c6Ey_DnxZ?RtN0;dJ|gPG_) z`Bi!cU(-}zE%^*3XXtUWerUt=sQoFar^UcothPz+agu(~o%qhP&g-hbw;tzT1tccA z+=KS(w=H;!ZsU=`?c&wj!!-E+=Qh$#tt8PkcipgzKlC(+e^o_X$=L2=vdjyYuUok> z=v^Q&aO(?pIv(DJXN|XoP`5GWw$onS6glE^uHq_#HH2zqzH|QXXUDIY=>&}5F_>>V zH-V|SqgF)k#ee7@$`p~$jB9+KvFApJk#w~reIWhX7j!OD{d`&X@2u6GG$EC2DBt#3 z(B8I}&r&5B=1N{vQk|-PmFwauB^%+tQ=-@e4Iz_?wVO9&Qm0IFp*my@Q=Dk0sWLKs zE6RPWJvKO|)Sp^HeeY$765Tlmv6q7wMZ-y(XupeqaJgL5n#KXn**;f9pw0|Yp?51h5k#^EJx!i_i4((1 z#c2bg@7{R!i8s57^Cnm9d5ov^2Om4d*c1&o=6tbVB!dhDx5)#o@Hv4w2~GP-`2ISq0;vd}TaegY8rA2bsz&btl1&$tGBxI=|@c#>{#?{udB zL1tl4mcM^aJ!2T-V43BD*rtEho4}P~Irt!hGBKKbbj6RaBkjSA_*ji+OBGb13I}~T zyl?w>onV$Q)%x}k!C~&cC8t3G@Pdo$dGrKR4x^cGjcO~Qv8mJ8OCTE$BSKykS(Y8x+A=5fA)A`H29#z0+(Mf%NQ>a%AU#z zsiGFn%kHqHzd7Nta!Sy?tRpd}H^(%+7ZvfUq$PLb4yBoBfm|JozkBNe@+_&OsC9yjt&#R zK0`ku#o2}l-GSpZ-sPP!q!$#{D4Bleu933!HFFNQ`G;elRyGJg?v@eqk;Jnnt_k=L&EoySVtRY%&IdVc4BS3 zK!KfoFN3Nf3&%!lS2H`fz;EkMsVXv`xjSas6<2^|=MK{&$2J$93CQLOX-zMg`$_lg zZwJ(uo%06J0u=T|-C@x&ZqRnmtf|{sak6T<2;XXG-o=R52Wc3OJt+32x6=!<_>k{5 z{H?4>O^f=6UD*Hi!WjNDkUM1{*I2#s;*+|YUE;gl=Ejyhb!ikcU!S%p|F4j6h^u8L%*J)c&30Rs3f#(X z7lkTF{duOs@%RCQ-%vgN)KoqI^-0@@=E9;~}7;EDn~efnyOlc!tv=yQq8jyK%~1B|O4UaknJA ztMGTj^m!34XxAW3<9W%H9Wb`O^=|aP&jLLKx!xm;5k_bZ%2+?m~;#6a- zJ&FT2^!86U(ZTa&ll=fYZ^Tb25XC+*`cS5a#W2ke!f<%*iee#)pfXdJ^^2vm4&ML; zoe;eiBg{IpoB~_XGLLroDG@Dh_pU9dqsV;lZ$GIGzxD^kJbd~T=ZJvG{NmfW!eEa( zhNDiis_2tx_YUxSet$Nfag~vxipmt{_rK#s6AA8A^Nh9~DNEC5)>Qo+ftdTRM{D*w ziaSoRbxz`A%`cA1|D4ajbAic)uu{Lbltkds?J>3tA z@|g(?v|D7c#peq=ySdU!kmZfpW<>uk2Z*I_)_M9L&t&AW8&dIOT~#oKMH`i*rEVgB zYD*hqbXC?U{yB+LXom7R2;sFWqze5liF@7~-ZOMVEi`N3seoB@icMn=M>8w~ik0%T|nBNR~8O2~g zx;>_wZs6bUbt#`vxLv6F)?5`QkL|^Ge{KalP9@9#?B-gO>_^N?*i(2JV-TES4-wDL z=8Z|H`k%Ye0GoK!+H`r2iRRfOg~EM(ELguT@#YOSf$8MM9n2(zEjR49V_v_6`W#C9W%KxjICf}>92&l(>ig`1XHTHIfDgL9<2i;sPoJl#{JEs;0 z+Kp%2hkf|J!CWC=z0I7Eej_6Myqa9d&2?z%A-};~1p|}P;j7KFS`Dhcg~@YB*qT1>t=H;pi-(D|K#`~l@R(?Sr+@| z{tSBTk*5~-!qKW7f9-hl)G;qo_p!ish96>w(aqUyf^AW_sQnm-hQx)%J2yOsn|0i& z2ej`Q!0ZcU#qQprXX_oe6Z|#EJ)Pt0!gak$d%p(GU6^0fS`kbZtpnPS35#2(Z;C3p z((2!$$YjteNwYS7l{ysuyL4Jd?F9Dmql}!xfk}|jOrjATtxJ0W12>wknagz%yvTkY z#ok)zB4stiFMX6dAg|J67k(+uRq5>@G0+z9e#6lGT4Qh)K&~=Osy{UUea@B>5eX4Jq$~Dc}wG9k~BkWAja%Iq+MA3o^oUo{_3}jMUjWN;SnQX8L8LeMR{yQEa0)J~>v~<5uofKLj1HoCk*l2VWifN0%p!s&bAtzIQkMkZdbo|Wc+t_fetb}dQB=0_5PB9K%Qgsw}HC)f_pdd5#YvZYh~^c1eDeMl(Xk`i?a zxuZrhpoP|F-X#6>>i{iX;^WI=%6>Zv2epX4>p@RliOZ_~nHu85Br z!1QgtQ8Bo@2*1j`A&WN@S-w{6Qg8jTo9zwMM*1PI2v9LIHXQ5i z59|;5t2Fn%YrqRpe$eT3W^_SFUWX5$;HHGVI{m@n(ay)NCi=td`{ZCLt|m#ev@@Jp z`}vp822dgy+uC1YF%3tXfTSnS0P*Wj29^vAr24dXkYE^tNg?xI%Mu5S4V+!`GxH@z zJ&BRYZHSTQY3^>}Ziltr{o<3w+jZY_U{5~YHj)tg-W!Bf7FVK?nvFzQ{~cnWqV(hO!vv0`YFtyT15Z~<-NUZ-X1u4%u()rAn)bf8w7d2{B?C&KOYY^}h0Z8rY>qLsWIoFJ zuwGmzeKxO26J}=h*O>-QfIQ9?Fapo=Mm|5j=%j*n?mjT>nGGM=Pw-t!fsWLDp0Tw-$Br27(q&p(BF+x08zk( zYKW_`kjL_VOvxMmUsp{e2u6+2>H!aK0X@@c=y6;)z1gGIbPqyuSrbD>qZ?|R!_y_D z2Q=~|I=&wIu6r3zrtTwW~Oj&J}IVORu=79`rMR%TA~9>QMwg zmJIek0~ihD=!xmvyOtcVv?Y>KcU~x8a zeEqT2(|7Yw!;6*p@&(0xgj8YtKq1a$dywM#WexE#mopoINx)1nXL!BeZEoN=j&mV2 z2Bxer?etUoT&~0}+<=-5s?OP$_0#aI^t~2jX1))z-%F+-x~cU^;I)jZ zzs!5`Z>sbPJzW+Q?0E=dA&tJDcIf=oVQ~3{WhbT^uU)JmE(zEcynm4nLQNIwO%mx| zP_RhE4?0lrWLoY~g&*)GXJ=>k0s}3LG0$EGmaVFsUH!fs_`C)5tKQuklOlyQ1{JaF zIt(C!x%!`oX5Se87j{vKP*0aw?BtWHnaf?+ef1^&`sGA#6J_%PZ=+5=Bfic4LX^y^ z()x-x-R{R9$-*KYKaE86Tk?}BNJhl+(;BB0yo8Lay}vpdd7cUCE;^@7?ZuWmS1uxb zi@`Y?Mw=nOdPE&oeHG3{Kg514k2p1{un&=c5xZ-MwB)f_Jq_bPjvhQv#3Xg|~k&c1z*_>Z)& zr<(f4++K74U1%+?-1eH?b74zP@YzuO?a+#`DVDr#>c5tdvifT7v5}w*X*FO^=k<%B#fmb^uNL)ek7td3;({E<2qpt#Z?AaAZ&g3zcPt823Wxd_MgdH)Z&7 ziUSM$_R7$6-mUW(3vnBrsjpn$Zud`Lk-cheS7EQ+*Ll0NH?3!J(rZ8Ks6e~h8+$cV ze#Ot75jlC91EzA1=un({n_WY=c!VN5z6iHw_MZyvX)SEpDS`R7f1GkHT-?_fsU}Yb z`R)si~qcK4Z25?+%Ev9I+gJSIC;zQ~23;*cs-%!H;l+bzuE>j(+= z+E=5Y$A!@O;~}~rLegD}%rJCI)nYqo66BW5k$&D{x_MdD7a1<&R{&1Dbr+t6OdCK_ zN~S%DjH2zF|MYm{rmeFTbexz4{j}xfCz;Z$T+R|-EK_OcuL22|Jf%sxqg)5B>rC7k z?2Ebh9-T;SySDz{GISP-Qeik?Uw-)d)Mxsqe)*__|q*3Pc^UWfU8!Gf) zKXNU9CfzAra?=(eKbMJG`kgUYIvWueZ?0J+^S%`N;X?dblkPF(gwq)AnL}V`d~yGz zc>ba4{eY(C)r9Y_`uR3@vrGa?nl@Vrr{MkEHs6N1UCuQ(FOy%SesDribL79LM^Z$` zldMT1S|V=i3U{FH@Us77=0WKkk;%zRlwa!s%p!ENL~GLPa*d)#vRAXZ+Q%nNk0(fj z!f?>tZhI@i8t21~Wt+qqF>KQ#j6Cl^)XXR$2?(C)AW4cJyOj*<0l;x`#R5lZx1zLw zKA0O0&1v+rxHI}CY3l~?bwRb6e*FWj(}AISKp>Z*9n@U8ItMHRi;l!iGcz|QvrI@s z81pW^Qas4@TFd`s)=}p6yNdVWW709+KRZ@J9HAwfLozfVL1U&=p4xe^nJ0^zI)^4bzAnPeT6xU=q4 z2igLx3LU5FYe2DEFbP7AIL(+AfAgrdGRfHVnMHNvp9nw#$UJ7e_ucsDOmG0{8y znyhjR4ZHEsr}Ze;#A-aHbg2lueAO6X`_9R80|i%}yp6b8(no*WVpG#O+|pH~r1_~a zF+&OQ*!eqVqnRP^?iJbnHfKIB>B{Y&BF-Gen;`|d@nEkZobTrP?Fl6jyQx&Y)Lf6P zONUT~Ty4P;1=|TXPKb!6;kD{Jl37!>sWOyc2K9t;|Gin>OR|xQb2^UhDXA})l51#6pJZuauX+Z#CNvfiUa>Q5geNq25`Ge$NQZhmKl!KCq-qZzL8 z#FMqgh<&nOwCNy*Rhs7|trGm;R3k^Tw4YF^R(W1;d6QZ8FA^*-!>M6rO@v2I)N4c6 z)0zr{HfQ%8;bfud%`07j3wF_?6j({e#qFRH#*D%A9D*onl}J#rg9I zzuttJ$cC3b9=~2f1-v%sn&%?}X8+Vj03P{MyRv=$Lkd-7DGK%VC4pq`Kz%NyH=31l zu6I*V^{#EgMwgU=MQW>#^V%tU1G#>bNbz`>;jqM53_@vS9nAPH5@c~YJ^v)Zuh8bP2N~HyFQ5h>>1_&szO&Fhb@@P$M4=!(EsQ8x*9DVgtE9hy(k8R`tVtkWTK^02;Y84+a<%rC zw{6QalK#PT`xf{{%J0I;%992TXY`Ezg62EGBsKN`iLxa(wbVg=Ya{qxrx7Uy4~9@X z8KOQV&Pb`l0}fPTZ@E;jrjyKn&6(HM^I2e&9r~v8ivJDvYxrYt*G=#Fye_5SE$VSR zlnRXx*Z%Nrqz`v4jq7m#+MwGW0L;+?2CgbJ`{-#!n-^KPER>SaZaxq>b@{aXl>I&Y zG_XW`IEnYh3gixg!N%&0qyMIz^*hULJT@)mx=C=w9CP83%Q`)dv}|r8wC6TZel9$M zDTz0}eLhT~<)&_p^pzeNp(ru-ncCkwzA^Kwy2ploZUIk*a{0yECY&EZ%9e4`1_ae^ z7q{4E+0=+}u21wG+RxW7PtK6Pn|3ZG6i2`A(u{YZN1j%|=Fhig8Tl6Sq)c;*uWB_o z*eBCU^1=8AP$=5P{FW|wj?2~6iA%o>;6q;XuX*SFR9OTm_83(-m zm)yr%hBijq_hH+8=jN=%5$PdZH1w{m^_ip4a?g)1U5at)>P*4K?$JNGng}bd)jY|O*?45roElNf+^HkanX26tbv4o>!7j(EdecP#bb6N1sb zaNA?U>_Bw(w0u&%TSBr^;#;c!K-#g-p@YHEtVD(PCeU^tKO#VHR1;mamTC_1iCXu< z!To~zx+`9rrf4B%)B=tHg0QYyr<8=75%Ibl-*W?GxjB{nd^QG6&nY)4;ZOU)7*cp1 zptNsfj2ywuPA@DZa4CHV6i7Kc%dcLQq%HHSbe?tMh-D=mE558NvPqT^%w8135k8 zAFGh}w?DlDus5Lc)q}%&5_B`;h@OkPT2Rc9%)&YEz{Y7PY(T zCi^I~pu8x4#@6hyS0is`e3`iO_5ndTvX1=^RCvTJeLJr&i(QIQGujZBy^#2lv77Su z`0nnFxo1HE2H>>U+sIkEx#o?=DxslN^R#0kKRKP1`<}+b&ofNUU*8%J9ZTR;DJQcn zz#PdV<}Yf^xuG_v;=@2=417fi+I34$qI^Cj$Tr`>dx_+-6*_DgGCm-=BOX!*0qlK0 zx$)7uiV2w%>uKaUbu299JQBzX5RWgL4*u<| z&<`_7fBr<1QZL~}DfhB%qL0dfq_MxE+S}@YJMV_w2r(~D&_aho@ANMKuF-j=JI$Es z!n7ilbsA0Q%h{>Q9x0CHod;UQ$>0q$9ov%MDdjKQpk?0U6I7AWp&~Aue^c-Xszqw8 zlERnOK?{9^P&NX$fquA;UT5v(HfuQqcix#lRiB4e(g*H{|7L~#+j(DYw zkEcaDm|%}hKjzr{cqf94(~WCf(mQm_dP>j}&-zlbQQ=`Mj+Mm}H`I)_fZ7&VTvO$x zCUN;U3%_oXe<6gRfP>Y=xI>8!-K~HKLmoNgN)3qSwu>sSRfFHQ5f;I)-Z!PL;Y}ku zyTF7Vl3h%qF|KVNn=;P=3Y>=wZ2JuH%`%M~jcm_S;tQ0!RGwkzT~9qsmC)%m<)yd1 z{{Fnxvf0xAF2+fMOhJt$NBsU(vbldSVqG};Rn@vc;7p4?Tp*r0=x|(8A;CO|$>xzCpnUGFZvggPt48NnA46bc0AsMY#04!=FqTNfh$}wb=-FO?6Q)W=EO`_3I{8g zu7lU=o&k5~P~&}=341cM*x86ev@A`-eFZI)$nqgrTr+9hh>tR{HQH3N_-tibeYV2cP9$LOxXPXrSr5f)^3{cYhy;A=K zd`0{V!{^;xDQIod(nvu8p9F_KV6=c(mOZKSTE}M@$bUI?N9-p)aSRurKX+DfQ^70+`F#hm5SL7L(FOe8uh>C;hra7@xlMGxOL?u}1mX z`fuNx7Ug!^Z!r{ymv0h^NheHqbJRuna+xD4CtuFU*gD1gJ~;egovYfc?rUiou8suM zq;9|Mx&aLER&HYz-x!Y^y=1PoeMd(TCC7Xgb=%q=za8xL0HnRVuO6M!OBiiTat?-P zOJlJSe$bA;e2@Ka65J5|cf5?Fi)?zl59G@c=|qJS$-=R(sBY=MI_a2|aCYF!^+bK< zmX3#|l34Ll`(d^Z+-T~0RC~4^qbzZ8w!&ZqIQH&yE_l*TYF8t>mTDTa``q*VK)4an2V~|4 z`!hm0A-8&Hgd@9|2thqJW}H&vhN}7^oSN6toL|bSfCNq_J-_t!(JeFCDuft-CcCMp zKTQq^9d@J__=e@_1G%*`l9^|N>53SFu!o_Q(aLr$o%*>xBNyDef3s&g*6r;v`?VkL z^e-cFo|;X%LdBnE)?4Wzop_LBBOXdsLZG5zJn>xhUIC?wiweoP!Suc`!I2A1L!Iqn z;9tHLjhTE!<^#g6(;J3rHYuCqfJgU*w+iUUZDH)R)O0*ZGMRlA^>%kLx=9ds&<|@P z7^_w+3=?0l;YHQtcKlO`U~TK`ac7LhhE4=rh(ebsFu--*?C}|j7%p~!`m}&aXk%?b|sjTUsNU=E{C`_R8=#n|| zFfVmYYd8rZv-7#L7wiZcu+iv<-42V5hi&X0Ek2&oNw~f>F-^X9apwQ%uVGkGX2x?~ z2BDv1w5I6b1&0B3dL+G$+r?{0?3vtE-X)!NORv?4wR6wP^Am%Ic`;*!M!mPS;¨db`0L?HJPkhVf(!dj7ehyj}V@JBK?* zOcOeKGM6IJ>$OtvVYN;?7YUxTplSGu=C8ix`0Ba;-tClXnCEL}f4hBUZ$Q1-&UL+I zFHMDXy!(nu<(Cd;p~FqT+R2@FO(TgZMIQLfJeMbEVAnUux^;tE(JISdHI_DYOHq7fOrR1x4+We zRH6P$)L3g$&HT($t0Bh7u-cT`1N-vtw>$2LJ@45rb4|NJt)();%CH!~pXa#9?Ze+wr# z@l0~Wy|B4&40_5sYT^dE$9Gql-Z@|q6us2RFbf2GI$%AvEx=1rESsK)A?Cx{20BLE1PXOlG|cP@38fZ{FMmIOPsnbf6l!pDefm=6cMZ(l9|32 z($5PQ5GY7MURkZ67u<-1Z&Lr~ED9E3jQ&Hdv66|B)3fFxN1hFsadc&8PdC-zb#KS% z3+8Xol{ScIw5T~rsCZ*)DtY6h%du`H+9niq+wPJKHvmcDmHzYyB>2F>38OW8JhrvU8T!$A-+rdGC%Jkmqj2cp4*1M23EPZUa^Ms->nz!srcC zRHo}^hvw_|OV7@|UuO$}<+8H!2E)TXS4x>vJ0#YbncmLCzaJ1sL+S|E;#3moCPd7%WGvM^O>2saK~WbUY(Zq zfO*Fe)$)3&`?W-w&0$9j7o$B-6|{P{5yge!JMy~P5)_wATgi_clsypSq2O6qC8N;w z2CXKL@107;U6ltSeo6uk=;A0F3keI4Qv|4%IC|@a!Szsd#rWS8f=&Nx7%hx_pXs;M zH|-CX-Y>ZZQr&CgfR!(Xd;ksdKg0>R+W1{h_g-A<9I~1MY->GseHN0g!F%|>pizked8^c=3+)_1=rym0GmT_F2rm#kI-`h2snB}Y{-)V0u6_bI2<_frueN6&G>|r?6+PKzG8hba-Vo9>%X)$WGE(feaQb~QN8=%JneL#* z$Mo@CsqwwRHxD{bt9$_4{hjl`40w~Hjj+nY-B8kaX`8|sUSQDtc}~^e#pEBwLeEae z-mtrORP$r}y+akahe(a-qe-@*ClZex#sh5(b5o64!@9`GTT0^qgpM=42nXlHQzG*S z@2!b^-8`kBo-;AHk&XJaG6HsMqA+U}5Dx^o%A4}@7rbq}3F`0Ce+(!3&BP8IE~vfr zRoV6Gs|MkYij-d;xpO(N9@{w&D3!Z_m3hB@tQ6CP7i2_k6v7r8Ka1C1N;QHpxV{%q zGWiegG^!y>E3=BJRqiht++?J!*psOB#13eGS$9*1djLnffvAAB0}b~k|ykNjnL zM>m@X!Hf#21@I6_Fh_ESWP5`)z^qmL_ckTT1R;eAa~t?wQhO>Qm6?E}@UV`CPbN^; zEPyADH(lzOR!h)q_4F<+x$4HJ5pvs&9 z#Xe;U1i!MWumgb!cqT@>G%Y+&s`@1-BM^pNo|%*?SKbn8+POSM5>V(m1jF-xCzpF$ z9V@x;dzYNOc2Ht;pekvE=JS;JvUFI_&ITyjda`SH2Mwh%*_6WM%u&X z)QR7f{di8oTh<+K5YPGC2CP4sKVRP+Q0CT- zixie;x6W0}Q$u34?2<;^erjyGM1sJ!Z9Qi!0R+;q}ZJhzi8PLL*H3?s&{Zf~-r2>o}86C3E+u zvdbO5FI7VTfh8oPu%UWs_vUh8)Gt~cXJm59G1M@cnlk?=;#NEjX`v~1dsEU-%D0r# zDB!)9SxQ70qB&gh`rEInh>PkPg2DQ*&-rP-zY_e6B|T;JqKs2lOMU_`R%++lhR)l{ zo&8Nnj%Ts}P`3B6JkSz7fOJsNh1&OisUKZnJC;FDtZ%z)cbiRRs8vJQU^rbtw4XbR zJ;B7%*|dD%p)}7!R&XAmM@56{AXj?A;P+K|D&!NCef(+Dpzwk2=d(SrP186U-+}j% zJ?a`8DFcTS6)?S(;&NvgM_B9}(6+ml)_-2UW$|nHk&fjkQ&pwPoj1Cd z{{3b4;vm@qGB4|Qd(~T?4hEDixIkGRj%W(1oNDs z4-TZ;^v7rumqZakV#TQmB70Xo+V>ZbrYT0SvTAR?X42l(thJi8qZK-#cho{U)O-%H z2YO@cHh;x!p7(E*@cS|$O7GgUM9rjBzO68DQHQmFzO!okT@kNaJLl+0*J6L4H&Qni z**=!Gx%ZuY^9i~uRWTkGRim-Uy{eVKxiPBWtuT7TJBpwni9d zJy8kk?ZLMGGdlBhsN1n#i#Dg-RoRS^w1mcxul)6`i7r7JXR#VJh`Jhq3yiB4@XE}> zQha2=ppso|6|O2v`ZAsM)1@a%s$x>+y{dVG;&4wk4o`x~^`L{}lQ!R9JB<9Hgavrx zX0)_rHMR4}9oL|JmRPLN%9R#876u342jr;#UJEHytu^FJ=pzz+Cy@-TT)Xt$Yw3cb zV-Hd43cQVHa>ahNXkFVN|T~_JM{+_jc82-v_5%~LtRRO;5nmN7Z%r;XN-fada#GB!-{1w2F>aqzc z)MA_zX{B=795o7{*8OC0siS(ZDEwJ3hPsUFj8w4Dj$;fvFs8E|pa8+VI(s?o$U#GE z4NFDGu_!v63kHsZwqgB^L#A(e3*t+!YSITh~3`SKr7Yb2!+PcsR#Tt4V zM2wOOAX(2L6WO%Lq{bhYuyUB0#MUuaVwahzp<`^mZpbM1+tfes_gNL$zD&AVLGfC-pTJoD9@n( zzZA=_bD={q!|xoN6td^puqr|)SE0duZwg*KR6h40FdZIpfBkTJdstQbEEH*9DG`-NdyedK|cOc}ZJ7_gUx}_5FbA z(@(WO0CgcDtTWfOC#|vb1*-`~-b_K1ZKIXeZcgpl9EQIhZCK-a;*@o!lLR*C*hs#Oc2v*0Tix4yo_n30&7Dw8(MGUq z{M*CUGCch-qCC?O@!C7U ztH|*PK$c?6w`JUfSli+B>uzmKgOl^o_{STj#S1`PJ{#+c(*~_0&=9%3S3~=Ri^D{3 zA5ITB4P6l%-i6&%8|RCd(i}uR*6=R}CS?rsxvEp3mfId;J3He9oo@gA7{_$-5siM4 zpI-t^FNJb!qcnDRc3B48Gu4mpTi+=CCt!hcT#s}o=1ZDaLEe_Z_40MDc|x&wY#g99 z`*MfdI3{r_rn>R4JsnGhjcwmfQbaL?tLbSchKmsEtF%NcAfs#YY#Vky-!&E`om_E|TFl&2nmNoCl+nnqCJDffTQ!}}qO`z@1J4z)0fQuqF|Cm6v9 zWLc%4bEoa<;2nh2;D%vV+>w&Z7`T)M1Ecj8bAYSoQVT3DBEfB#m30Ym^=B>W+@6CN znjJOJy&*4i8F`-;lW2R%iA~H?HHyd@PSV#*3C7lC1_KI}bhXa7lK9l;F~B}AYe{*> z_U#~#j*Zo6(0%1;R{6IE=3G>B2i3cBtQ)(4!4FrRg%(qO)Ji#v^)I2+x>6|^`HFD& zBvzwfA(|uoR->ugr6JKOzI^ z?)C=#v!cjWgDLp&+y@?R6@>x0uPA6N=kJ#nqBi>)?GnOP&_<}|aumOROCB}EHT0lwmMh__4Tt6{smCZkc=+U1 z;*BQoMnFjiSU|uN#vRw-8iTeHyZ~LCmN^-mO--SISl_RjMW>qkC)NP_Aw>&vmtx1V z1^hV%v4b2!P3M6P=NyX&*jAMZMs!+Eu?TnoH?K%Sv= zYq=V$_x9TwQBgb8BqKAb4OQvnNG8V($0{F^ycr3f1(OfT@(mxCej_MFUSSN}y4J8^ zW{unuYy*W!;`j1b$Q5y(zdYbJq#rZGez~k6NA4%1#mbELUabmG=KbruKb36LCsc+| zBYH+}4Aqe=3lI5PKiefT=l^j5p0UBtjI@~;N5l$3iZ-bp2mRx5W|lgVT>Y0M+w8r0 zkE}e=b#gGfmzOc)y`$!v6sThe>vk?hWPtTH_3x|#1rnO-+vTxz%b=jBiZ`AItui&~ zWU6i11SWK9gZvq^sR^8gx*Be-!)`_D^qW#?VVlERtMbarm38#mi~q!iy(r^9zvx=b zh(RTH#M5)4!%eJ$q5Uu>-hbl*VzgYy@cpLX7jHjqFgky=A_*WUsP)w-snKuu_e{z1 zqNDh=#30#WUw@GFZCOj)fzwcs(h+OV5r(#nLzyL8o38-h2wzOJ_V7j)-C3m7`*v4G zNIp)x30x12m@67v>gm1c#P}{s$YkW!=@%h7S6!1QPIiEp0uC(j-pNfhq^_3> z9U1qrD2&n!Pu;CIzRz*|Y0tXzsxF{&L$K27P{LIk*v$xT{7lSv@e@U)_1EF%gwnRB zOdV^#G&ORoY5;@Wa>q#Z%uKnOL0`vr+LBhk&Ou!8+u2z3sn;kxbJUCW;_t6DO_Jfu z?vGnU;fG@UQvsvT++S}IRxn>*a^}wqf1^Dob3BUGM5x*+Ml!_KHE4D2&CDh>rh?Ot zuehobpI?|p9VH8^yn7#X zbr57M-^Gc$0*?T#VSXWzP#LC`MlAo?Wq}C9V%_*>kqdA73vu93EyCM&VfKQvi1Lm@ z|A_f%7AzIs?4l2zJes&z1y5HxkLP-x8PTaYrrxY4*cqldQ4jF(R5LpJD@{m2$T^ng zo$l1n+O5gGE5?IryGD*kz>F=q(BUJwn|sYwUP4%4F&Rq&PcY$?gs8gy`F{8EKOJqn z2(DopzjnrwH$Ck$o11--dyMA%4{a}cb{Hr=zZd4ssMijf3tGExNaLM%YU_Ncv-zBf zFH)z#y6K|EUb9eZXcsMM82PF=*YZ|eluhHpG&X2lfms#HJ0X&?SlB)lW1DcHqei#u zh|2p^au`2NwSRLxsPMMpwzx%`y*(&Sf{9}H&G7& zW`}nj^T{`JnUq=sPZ`?tGIiH%@2GGISVVpTGNWMW-^ zQN|+#C^|OAa+az%aPt0w^T?kSY+uJuO-@9Gqg{BUR+Jlba3Tiq$tCvk#2&(UmsQ$2 z6bIvnh*D4%z>ZpHg*BgHz9pq9`BKlhA{fjo%Gl=I&W2mnY8$_ux>fvKmCKjNWurHtKGRS^blMSs(*Qpu95wA z!vq^wPI?`C1!Mf#gmhJ*N;Ia>rP-PjCWmQH5d@^-`ze7gY6B5s6*SxD9tr``R7WN( zzFl$k{YytCE1`PCN1A$3a6HklOJ&`Ew_*qSjql4M~35!ndh=6=PY9q9?#uq18U`5%EqD8=MubU z3%u>A*~WEmFMYbYqiw&wKH$i+v7NMXv3Kl3q}A%aw^A?<3TSI+Dm^{jVJjoC{uy`?b{y-fR^z&75QJT`JBw2JG@)w4E6sIh>gwd`w zJjMOx{1PxaH4;G&)UJA_JQx2`$wYZX%&?_wH+XB8|KcS)q$QGX|NB;Emu0}=3FAdS z_2mF(6E>Cqa@%nxHk*se@8kO67YyFf998$!if?vP!#EnanjnO@HjPUoyYi>6ldd$i?`BrJ@( zT(DxZ`rn@Ub;v(J?kAp6hyM(|8=xr<(i2g{zM8O4mgy_ zFQ}QCQ6tk#B7#^9Z70MdEMi_XO~Dx^V4At{Mp`lmi2~I*x>Fii!diUqe6%N^WI6Sn zHGVfTZ*Pp5hoSUFDT7gD88SC>i361@8%1VhqnSJ|T^%W3z6F1`?yXsFu;rsPx((|2 zcHIA%D8FiO9aw3?J|Ucu8cu6>yl0nJ@m5fyb&ZV?PvVhv;D#PTM#7TUNck`~U3o`srn{9?|~!xn2Dft|&>cHO+i-fXA75Fi+= z(a*acuEP`%&rm<}8<|VMQ*`0xYL!Ydw4uA2iDl0jxhZ5^wZy_fHg?-ZJ}>P1`51lW zi65n~7a2Hz?|aBMe~a2w=xq}D0jJ_={vwen&)>76$FW?8{ko}p3+#$%3+E+Me|_bX z1il!qGwjw{wI^QIC&@9L4&6$ZPtHYyGGk>`lmdtEa$u*POPX*l8!M0faCmCeQwM`W-XL1*GkXIqezdc z;m@czXtUfA0A(Lgs7M%z>2ciL*bpmLLD6H9uUm@JwIJzCxwD! z)&R6qr7AhOe@Q78=Cnx367%sZcbUAOR_PVNY!*uI<1KQS{#d@6Vgn8zia5 z`J%{_v+v*G#G)uO#RT+N?PV+z$!L##~rxQ8^k&h{MucIZ((Rhz-|%!|?dg>N+BZC3gz!l1G`~ zfrn6Kx%7lr*fu*yRq3#zoK+g%A${fBybav4D?Aak8qSu*vps0JnW|*zt9!Az9s&+! zwy2*~8kOyoy06{Ku*pUa$n5yruoX2OUF8yY@M-rNAN!hJxl>+yQC zHq?>pL$&t?by{>>lFd=cR>SmDIb>RZ`W4PdI9PBdzMMqFZd@ljabwmtH8T}j19tC- zYiG?yy=Oo?{=k#MTV1gJA4!h-Znu9zoAh6yefB;rbKiFyhmo%2qih-nz5ZR}R7VF2 zW~*f)yE;}vzA%d7?p{8E2R@&Q0v}}xX)UN_(&#I*vz8|xo@kICRt0m1q7`8T^Y+De zl=sT=m4CFMKIKU5^drdN6L=mR#wwP8)K4`?U=aynw$i4rSOF?%-;wJ-=9YXT`})W6 zVMQ8#we#Z`-A!PSHQi49j{nbM*Wb`$0!PMTa@)PN6qeP9y53Um$t(Q9K}sLN8GN1L zFM`{X0lxy%+HPndZrovX#IdR+qe=TPh#sR_j*;p$K2s4Bn#=8@4?M)D0(=DM3$^cSA)NCPC20V&MfqKO(^EQHF?RJxYql}DfBqEl z4;jL7k;VnSxrQlWZ8!6}HXpBn^SN+2V0frpa^}iVKtWJVBPBNd-Q6mYKhQlpJA|cOo!`HHqAOu6df$>k<6|+mGxULnWToN~kmsk-rZW{ex^Tl)B@GCMcjmNH zj?Az5QcV+@=#`^4px!uz^2n$G+&N9+XVvipE{8q`K7VnI4pzJFNKZihBk!NX!9U)5 zq9x*4n|uC!`;*P{f=a)Heq$=W@$)qOh(vKvRq@JERk+F0G^L+^ek)XGo)K2%QFEK$=Xuv#vQAvVlm{`9k?0=P9m>Ofr#ULCQTQmxI5rgbZ!a zNy*Par9VUy4@7C|>MKK=mWYGN(;j2NwwHHAUSnG?y#ad@9-sdscJgyS-nDe+cVEuG z)ayn2wxWlATNTf8shvG2Kl}z)iR?PH$9oZh zFIC|x@GRqds|b^rzf0+K>&O)enko%%G{^FC%Y1(Lwu4aci70DwA>!2SL!OH$w6<_S zew2953oKhN&DrZg%{JRxb&qcQ#S(a5ahZTlI)_#u8*>={Tdo>01=`)TBf(g8ndp$& zS120^dY*r>A|LwHyh%<-4FxE8xrSW0a>uiTg}JkMg_v2|&hx*~U1m9SEq|1DA5BZ~ zg2rSHe6t@^@nY3RKaS~CLzl?oyBnhJ9(v45Q_GY}@*fpy7k|?O;R5xyPo^TqXyH#| zNcm;0#V&WU?Cm#UNIdC&fl%Sz{x|32@D@CTf^q5U*$rW$;DzFvqYv8)Hu`UP*QQ8a zFw9eL2?h(#v@beG2&w0EP%%6=m+D(q`vu+D0*19fTChJ?mAPn|q&kS2r9ASwxl2d3m2-rl;+FJ{U(cufNEP_Kg^`*UJqDrV`4O7kbQgF7M)j z2PT|MiZhi$YU8qmB8<4m>fV-{{29pmcG^ADp^owQU- zR5NZBBc!2N{n-aJ;2COP6YWajB=dZf#;ne#ombR2Jvf9K1F_BPHfC2dX3pc5;cRy4PSr+rt#fXx`5kRx z>QuT*7}OoPqwe9fq=gEogae>FXL0CF(uA~8{gA7A@2_%5OmHvtu)Kr=la3o|`_Z8~BakWSzQxV_f?$!v5 zPONK{9u>z&sO%hB4#Wr;xNCG(*J=7}4mp7#Yz<#(kZKYb<~N`RcF=RrlVE846YQXl zILLjM*iZ83NWa#}o^(CYi>zbq1&rXuA) zBF5=(W0Z^p(z~(ps@LkgtLEX0OG0b9{!(M_gx0auWX!H`ywr$SJ*swI?e_fjeip30 zqg|da-+jh@f+;)aBJ%CDFC$PhujdCbPakmst|Fd<1OrF!2lt4gf^-}fWo z>RtSV=P`#*)gZ+?hVw$Wzp8k9^vn1)vhShcyB}Cyrj?7GI!r!gUywW|9G`bOZpF)o zvBCo4qsSLw-4~!A!CZg89NcSZbwjm8_ck*AIE!9DN%jFB!Im^g$OU8<0`|dCd?kJRA1)$Ui1rJ{XMG6zwUFPZUk7b!K*0 zX5oL4zB9azb;MMwwD?>rgm!{M)f7sAEREY<$A@aK! zN|a`2Ei&YTc2Wp&SpGEiJXz7Tn=2nq9TXKcjHrD}1=av#7Ajz9=H{H|m z=YD2odmBE?H1t1o(nxdy=GMv=&fvmNcKG|2c44X&RJMrKTr8u`CVWN=hDRFTKQ}=S zUvVWl@$_9j(09X419c0UYttMkxyk|;O+((O*LO8XUARqe7gvueJEu2%%2mbB(kK_~ zg4Mk)%Bc(nFvV+XL$7l+Jcd?Mo$ZA!L2VukJ0y>@I`4Nrg;-w+@Ef+=!;NHi`7b4* zMA9#_Wc$}uNtYz8Mx8jo2TDolneH;5XB9ts0?TF1slmMSo`#Rz&iumKBtoDP5C1mJ zX|t=!K>|ySa*P$Ak)Lb4w2y<^BgiDd-TGn{<>a6hR%Z5K_-vp=Qh0s@AJ1K20x>ZgUPn~2~W4?&(VXmYpg@G3BrXu z{sv1E{|)9igc~J3oT@3Kp6KjEHD1e~f$z~ghgJsm3P;x&AqnIPn3K^HI(rMc3ktz& zF;GuDL(U?#m5 zREwE6f==2L+Kv+q0=-rFVeQl?i|cxX&Y9C(}D4-{9P;Qx_o??aWfQ&^Hn$* z^DBI3qqe`3-qK=~0(lh)naCA&brh)E4G5)-yELL(33gLpzp7ijqIt@Z1A#Cs;lrJ9yCY&|67#rGy9-oHAZlkYg>(uXPc@V_# zEg*-XBd-9SIm^(m@+{Ch2-(I=N#RdWS8t^Z>;O|IBe0)3R)J9N@HMV zbs;=E^Kzmx?&*z(4XO4Dz3H2c4-$azfp&Q5dDfA39j{>s)x?BA|Fbhb;dKiLz0diz zO~F9xG%zmNZnmQwh{mBC@%4*M0iLg2a*7@a_RLeATDvrby*6I5G0C}gVL5?NjFFgG zbZP18I)Y&giMiO5x>_F#fy}Q0mfA&7Ynd;fK7}>GR2rFXqsU z=|8=Zts0Z%Fu3u5EX=h}iH+SLuTa0$S}t)O8lhtvXciR4n41UOeH}+(+9=t~ZDfH1 zYqIh$Ag^AUhMG&#e0Uk>%)2N8yODzV)00uSvdpLWUd0aNBF-wM73_iApQQf#kSBO= zOLpnObkuGy0XLLx=J#BoizatWGqj^NJ|D=}TI~Ad69nM{{O7?m+3&%kjsXFN1BN8X zu8hkG%YMtvQC$@1e^Sp%?ltMYQCjGBmP;K8on2pc=G-akWSkY`kbkM*Vby|70w?Ag z3}8AJ-w`n_+@gg(#*GqJ3XkAY?*rJ|+&=VFzSo}TVaN`2OYR6R!|C}+@^{3_?pMI9 z*IN(e0j*ize<1AEsuM(w~-C9I4iCmT!ai;+3dX7Y2;SsIQBm<>nBPvoO>nV7BxQ}6}K zd>o`7^cvsX&1J3{j7)tg5!)5kFMPHf%$U)6A48ve=l88|`wZ|L?GrsjR{0oDZyr~0 z{hTE zRTU)8{2iG3kQC19rkXm5N43DncA;DFn+5_zXKb_s;p|0Cw8X-tn!9M5b@!?|lhB|d zGm^9Vv2bmCr(QMKaa3JIDn`w+CbRpxm{Rf`T*Hzjf_p=#N8%IT6wVQ~-*-=wraQfG z|E7~`2}dfF{(E-%2lguK1`x5amrU~ovgmY_WTer6DeI4^&>#R7NneXeQg0#?Aav>`p{y>y9DFdFEa!(ec&J}9UiKjt}tuGQ+BY84@q}9@< z$W;g0pd#S1>Mr$MW?HR=srrmlGT^Z{7f7!_H&>XEed&;eO+nHn_J@m?X@Z=aivd=<#M=x zEGAtRdXy)1e7I0{@{b=TD!W{`+4NhSLSsI%Th5|JJ!xN}5j=SM5m*%Bm3PGRQv#_j zV|&g=~ zHI-VaeXgj861CEc&8J%}n!0jyu&T$O40Bq(N)bhy~7lqXnE_#=NxO7oIm*93s^UgY}-ISu8WO5|wh1Sr<=Z3GY@bdNpsI?t= zRG$LIBiqlsOnNWnA{&`xaPkRi=<*Vv>`E9Ib)VVE~X|?Ng&T^V0PV1>r zBk!j(dLc}5e>J}blEmei3)_|HbB;_5uY0VrKC^}*lD(OicNm}vX_AbjkImH)aumf< zrwo;7R^iBV?idy~!ULJ%+U!gowrVt|Jzzy&C-tav=xaXcS~oS}-t0FN9_aWp$h5KRH*feFu5>D!;sVJB^`!*VX(zRnZ|b0&wO}r-;`E{4R7JRs8sj zyF7QV34-`|-izEc4E2O{enM3y=uf~LJ&6N^aHpl&H7MSYl&)CY$t>IGKTk(mJHG_%0^X}dEZTrNNw$880Ol;THn`%HwE3Ylzx?}aT zyxDog-KGou;xlg*!ANwW=G$1)2A!ia&5Sj4xMRS9m)F?Bk4#}D^;I&`)la*){kKad z5D)d9`D_Jmw+KYjLS+ZYEhFTBIkzifi-I%ZA4=&zrS^X=Q?FETQfMq^Aa9QLuTd>? zVD8U1D*Hv(?T`hRhVe3#au^p z_c$>RL5K-dII1wi$h!L#3xm2o&ghuUBY(*lpncmcv6S?w&R8E?>$^3?nu|beFk^`@ z{_>v5rIZUX;mcyf(w?~CH>GH|m^i;Cz%vfdx69E#8}@h1`G#?vxRUP)IKFf`Bu0?b zb1H2tgO`GOSn}elZ?88jVs3;D=OFce?dD9{nzXv($A133cOQFS-G944EEVP8SS(Dk zdvl|CdeP$cigP+WL`&;jLI~q*8*@^5ci^y!Hz+JCqT*$}GN!YO$wr`7wu_j}&37+s zYph4ERib4?RN+tM9n@qIiT;z`bEly899{MfV>+yG29FP^-OREJTjVI)@9)R63Oq~u zlj_ZO$$}Mtlqq&Gp76{VlA_6vW_adh!gGAnK5VnU$P5)M_N#<^Aa+Farfbd4h_boL zur*3&8{e~{{2r5&_OkonE4ZK^l369LkV&Sy0OfS_LjnJ(zw7dc zy9I;}ihqF^#oqg5{k9Ydn`^dfB-F?Pg7Eq;7)HS6Z@o3`O-W8#I+ zsmz-v&X*arRU}Kk>7LKP5G8?9Bk0*Ztrsl zM!4Q^jSoiW6#w)K@WG^wuh0lsiDPm9^im;mRL6YjLr;g>CRUp)VM?rf(X{ESCg?;U|VRpxQO>JBhAC^3*yUlfZ^H@`ePa5 zf_DX%N;RFS52U@?dA+pBb1alt1}kghXTG({nRI9*5~grhV`o0P&Y zCn<_l6~?aNisJWwLZ$sZ^BJtrTGfWO$Z;Q`&f2iiOUsU~OvX^7ZW=OGh9py$y^&7R zB_zlOJi!j4w1i;JvQl{mJUlv0K8yM134)ovi2BIjpIEPM@&AajujgP8$fOf3j*uz{}Q8D7lzAovkJ{5I5JA0k(;^1WFPB`Ug zFdDy6Vqx1bRpF~n>}~GigbyaVpQ|?Bgcqxy`AEJF;BwVOW`y;>#x#(m4eLVqER)n`%-m z!(QbX{}(&r*1b;h(iMr79igsgwOJmJO-Rk+&dyUyAfp{DtqnE#7(Ls4!7bzo-#3-& z7ezU-xZ!)NTn6?L#a;jCrxhZ=dB@~5$!vudsy;>Vw~i8Z-_|YCr@Kvi4jD#r_FEy- zDc)j>oM!ic!H0b+!X7x3hwgcZnj6_?0e|CMg+pD_)md(Ab$jmi7Ef9^k3lCHB$YT@{nS`Wu4 z?7SI8lo`Isni|4s1k4Dl^zTaaPp~LM#f{_^Qvxsl3#W*jOf{` zRH^#)X7wQXqtEd1Eche_4W_9$+P!k`=MaY)HAaS%Vzo(h za;{@Q*%@?eY>LVi8lqly5b|Ib!6E8Jus*#HeSbqbkrY3XY@>T!*U`blvUm(CnGN4 z&!@k7p~x{Hy0@>UrAhON%1Lv0JLba?M33P=L5RL7ZjA)UrcyTPLo1X|aE9}0waHYU z3V8huGGOo*BimRQ1S$Qu{Le}-e?i6Zu%5YFLdy5~CT%rBcB*eD%)1S=rY20M73Tp) z6%>+uqy`zZ?)G%XxNJKcpO)Q=XCnOwW&1&qi-i~BiOb2t_&4+y##D5X;Sp8_cwX@V zmi$GM=)b87X4-^f-ol4=l4m30a-+I-*@-T7J(7Yy6_VnCYF`N)(Z&yWvBDYQITz8` zFl*M;w@T_aqT3W~bYxOtqDhwy-)xnfn8j9e1QJ&8jn&uKLcht3%geqs8rhBd8&If< z@xId6>BrU=^9Z%aplp|`MzHXj$r0@@Zi5N)ET6N#U|I3HufM3fNig5XUum=CebiY1<$k)eh8)-QrrwE==nz%>4?M#WRsL^RJ3i*SYqX zjUt+PLVP(1kd%^YnV811&a=FaG=wqr!(9Gd_U{rls#*RY3PA*s7{2Pk>dmrZPX;d- zZ%aI&s4U0QbT}eWaJ$@Ld$|AlQAMXiis|7Ufp&#&!Bp<0Qo4=KDa|A?6`*>ORm@uEXluusA#_JOn3T*wzsoy@Io z8rd$PhwUlh(kKDBa+UC)pRNDZXnx<88Yc5>B?1ym*ChQ^ruD?$DRk2$7DuTq`UBcL z2sbR`;BUu~q+3ZI(*WutEiR;rv)4UzOE}cA`s(g4vsIrVHsoH8SWM8v54Ob9jo5Q0 zUl2=Y+X}y>6x=D<$G!m%p+_n1faf>zAJR9TfoHOKQw^3Xeb)@g@S`Kf&IdCI&(3kE z-OI%7#~g7(k6mpaf@TS|WJe2t?O(!W_hUE}keVM`>y3a~^FGs`g5n#$kRX|mHlWrk z7i0S+s_y>LK{NJBU|vk+PZ&F^{R1U1f^ND&q>T(>2W3wYq-YIID>F*U4>8Jpq>VRx zIT!!Twn}%PCh|G;?MX7z3ymg*Iz|y12wAznQ)I;y#O#=O6<(Y5At0woP#=E38hDakqOgDDvt0W7zm)nkJJ-nG*7M zKg+%KcPV6^TC((ok?!2G$E3-g{4n|^Y`h@MYL>X?B~h$ZVM}E`3x(VxXs_ACaxCN# z74n^wls(8rC))m+6^rBCmem4NS+O*qXHOB_w9b4sr3P;@)tI_V8Bo5g=p zL{#nnGo`O#KlVfJnp)D_MR8GTD zmdl#$5BKnxn0`+?-=zs?$1eoFJw99|TW;LaCft`qv8$Iv(#r{+6qXuPy>#w zPb_5OnG@dTJv|rVt~ZID-4XK_=bcA;6n`=|Zz}eMV7|t4nw2xJH;uptHp*mydY)Sv zbVa^8Nf3+arFBx3D!BZLxV5%tYQbmD^V7Q99(Mbdy75PlkfT;8%`xYKN~6X+Gk-CQ zAd ztz9Mr%mJmN>{f~P-inaChg7nTM*J65i*40grZmh4`e@q$G$qdaS_x^s_688I5XRvN zqeGf;3gI|vw4Y^S*2no#cI>{?ke^Y%kS_FElEkM>d$)e=%A)GyWh!!2rmZ*jUy407 zycqC}wV+V|205j(t?v)#V@uG%y7zo8jelCLIAW_rL!`p+7^e}%)}cRyBrC|bpDZI0 z9UBxRXm6;gECtZ(g{UuP#{7rcR;EJtPRg%R$8(BkfVWUE{uT7{4|iToV2X!{bHD!R zEl<+vg8UQnfLH85>oRsckF61WA7=bf1?s2IBzgsyJcO3ZTX@d}x@9(x<8nfg!bFRs z`)7F%ab|jSC>*V)*sxGPpo8>d>C6`j&C-6{bY61MqblsfB4a04$oV?Ma7}aM48b-{6xu?!&I-~$HrsRYGJ4zK zlzHLV7T(W3VCfw5B8tw;w)N&-8Yuuq`M#qq9v$K(_ZpC!!j2yJ8DPHBS~SykjE(Ai z&|ad5uS>seLc+K?+HHQGitDaK&X;^Tf4Y?{iz%??e+DG4vVcI9vW{ZJ(!;SNZj1l*Z_Fd8O(Gq?m+s7(u3$FUq-=e}>x3~T? zmd{fy0jUvk=wu1)ac=3F($>oeCp!ztC?!o3j(!hD&jI&&frDa(*u-|-4AXrSmicsq zm(#x}nH7s&Gj>L;R}8Peln#OZ6jZjn0sJxxON9iz5(MxeQ_^_d4!@cr8s95NdQ>kX zfQ~RNEXfdDOj$n5hZT&)5U;*Xh8)4btWaBavDb6Hp~U9sQW9OWnnR-+5A$~_8eX|f`%@nqbHqQ0jHOEJGw;t*6!kDSj$M^%6|O;`p9(+sq!{s&BP{RA@OM-H+5z-gMms8qeA*R|``h;Ad$a!}?v- z$w2nIkDU>vOB3-CuwA%ANv-|ZAWB8$rlCtCt(ZFt*NbqUuIu?$vgb}=m46MP zHkSu(5DD-$-}*v}l~U)z>bB7V2vu4NI_>Uz#$$iATKAMEzV3?Op63q(7*}%8skn6Q zhd*N`l1zsw67kfyyku{fUBX>siS?6l#gie}h0fkXkdo_nvih72r#V;Bsx7WKYIkJJ z^#NkBuga9R&vw;>sQ68+=W&D(*26e$++k%~?Xf5~C1%R`Z1 zSfqU(X&2cal%h}*x^Lf5 zugVeYCx%wfPrT0Q*y(LQ4Fxqqn`LeTlQ}??!x0c6=>VYGxXWYM);8xLL)@^9VZeoX z#0&h?0qf0jdgtYjtvGuYDv8sZgXDAK#hJNwV-@bdo}@i|#?8gGE*i2cH87&xyDu-I zibEgO0;0P(ril(@=hC}`tBJ|RD_@X*3zOxOS2nTUfNbK9e1(WGHAgOpLNez23 z&BxBbcSdl7cze=XjrF)xyO}dF54tz|Ga1^A^=y>S28I2%Gdy|$Ptc=}DyXcYsa{5L ztSR}-ftjP-h~LWqgbC@y!yYLpW*x)A`CDxBXoG_E@Cy|+$@%yvOUuKq%BVEwV5Os9^XqWb$2@Bd`Mf2e`5f2Bf4 zWVxNMF{0CQGfFVPH2>>5B7Ns~9pI6=r@(XUbfC=C8;<^8-!55gl`TGX@45yNPjqRS1oVDKufL%~cjmlENv3 zzueisjUShUkN8&`hp66*a=Jyx0fPi`V=^-Ym2QMzK#FPi39W)Gp)}=1#&tB#CGc^W zOMEdh`y}NrEwLw>&+)zwEI=(37E=cQqjZ4ZzVRSZDKwjsa>*EVg~*N?mU+az3xJ_x z4!`xqJBp34GAZQm0PD;=%=QkWQ95^%v5u$slqjt&%RdRjhiS!;8wiGRk4U;D zh54I724+||?IA}^U4yTSIRI@1GtP-{ptk^4|Lm0eLf2D+>hB3 zmNv_jb$eM;qyKsV<`uBFSmv}Y2+0ze&fWp?^&?pOHut=k$`9mN_0cYk;0~LbIx=4~_H4;kukE z4_6l;#$+FFd$onVD zt!Dx|vN7n~B7rMJ+qH1IJtDnh@*V}+^aomVc`hz1<7Rflmf@$PK-`@b&9qd+fd~Fi zr?(t>W{~c?P3ee$*y#EpVCu{agJ5?2*@Q}=By9QICoUSRJr>p;%e;0eb=haW{uU9B z%;|@j?5*0F^Zw^M^ItcOsUvivi#|!W zTdKkTC1BS4caV+PGn|-s5A6F()ZL8Eyheu6uvRTz&oFXJ6fvH$ja1hQp0&}X>u*#B zTi1^z_(xZJUot)pW8e@~^42q7ZhB~<8uCseJWPqKunddo1K7LPGeZ zr0Uk0mcCv%@g%YEX)JALlD$vPp*XPeo+K^NpBs(!g{FOdPH6 z4=5R2RG;+!HKD{8-L6RL5^#;Ax8`K%`?8O&Iv5nPeAryQG_$}mux0&ZZ6rhhP1$bq zWy54T@1E_!NF#Is)KXO4J`_>Jv8}Nk(#;W)+;rKUxUc+nz=aCtXP+q?0)6d=bJIvd zAT*~${ay*HOq-n+Wm$qVy;BpyQ8wRj|HSL6+WuJFJo)2$pO-t;e%wm>tkOxak)sUQw7vc`0ZAXgxh&oRI;ZAb` z(18q`)f;`sEYs!ILnDN#s>B)UradVkMM?i~g$IGuD2x9yfRsCuqf1l|YGGKi=-|?Z z3nR&_@G3sd7G2QtEL(SV4LY&wm~KAaIitvKlzz!KLh*HzOG=cDzG2&-tZ;9KQf7j& z5ZkM+jua9DeN8r~880mu{aC~V#8{?%ZiM+G^79AY-jWD;n$eymqWQ)tZhyfuNw%%A zpjmE)fCM+#!nKccycqpwPB#fGNj$xh);VL86_+&(&A{TIWri&WC>&`(vDoD=f^S9n zOddt_GWkmtz5SXb{};pk`|l;Y{x8ZO-Kf;gFg!!4pwY5^v^la)GKSiD+tQ4HMe5?u za#UnO%$rCWiafQ-nSxt0nsiCVL2(xr|5Ap_Vx}?|ip=;j{>|cq z7m=(l51yjO0pb>H64F9tm5SdJkB6<@VbRkCMOBWU=vN^AkW6D*@L34@wpChq-CZ7V zH0P;W$DiuH4VTGR94Fx5Tjnr6xzW1NrRn2qM`~i%V+jiwn-UpY^_(z4!FL-Sc$+x` zhIndprl)^x)%-IcWDdLw_s_1urV@G;#hW!u!iD*B|HkbfeYg5J=-~ZqlXGgIZJ5fi z0vDhwyQLD-G!1Wjik%r@^(*EM)4b=%Kh}Ew|4{XoL2b5Ov~XKmpt!pfcPQH8?(XjH z?$#D}cZX7n1a~d&4#Az^kfH&Cob);8oo~KB$)C*JcjnrAU0c>(+gv;H{h?N1)iK8^ zze20(kU@g1_2^HnCjxSgY*z%?1VX^1vS2Y^ys#krIW%&$#+SFo;na#FIcX+xI$=y* z8VDzNqg%S?s69PvNxs+3QUy`}-OY9)k+jfMIjZr6i@w5weqyWFX*MSx?i4XF z%lq4$kd`$*3){hfx?{^EY_f9jE@CD(!_D7@Mn)A?ei_T8Kx6m+Yis`BjEC&qKPqLs zH$49}Wtm~_$!OH%Wd@);AXvX$^ui+_>2|^Dmn}VnEX=nf_WLP83vg67pee zovl&|asHM30(~j;I8t6SM@ROJuV}nDT+L;T;M6NM_JlUwH1#ePs-4IWEtYxB1T5DA zOF=8GfKTX4r!+A++j%&6%DJ=-#!CL#)ELlj*d43`k{V0;lcB4XJj8+oOR&DJkknJj z>+Mp%F#KE3fdZRT!O1T#h;Hdc7;*=0P7OMRU9Q0%(ugI)R!zIYwd{K;K^S=AfaH!l zw`yZ#4jd>S+n&W0x_PzER=dLx^Qe4TwSVlMoKo}v*-8>c5A(1(rV>FSW$*yR{3|Ot zw#(1lM5ooaMTct@?s2+ox+E5Av}(b+2w=|Ie68i)&(rQ8ap1dr_Bz2t)7+LF)vP+{ zn});$8sg9Hx*~X5f)|NdJ2{aZh}DoJiy@)(l5wE3K%kO@#MW9jHDxi|%aJ`LrJj*c zsqfy`etu%?9WxE%VB9Dtlz2|56?zvhjy{a*R2@Dl5q$5Rrp@K=qt5y0d;tS0~H47@j4b z&3G&9?bWDa;U{q2(uP1x>YlQBcfrtksIpnkfXg*@yc4K_^5-TaMHA)mQ197*elcu? z#i_r(E(T+Z0ZUo^M7iDAi`&NP^XFq-56evy{egKA>>5}ISsQ`4cB=e5XZ#a_%a3W; z5~o++{rTKHc*AScE|YzhQ&OuFGw8tZNu%r|D}2_N^JiuGAz~qSWvk*rgRh#FD~#%u zQ?|q19Tv9Bj&z>Lm(lNdv9HogaS#}L?Z(jdE8UPDG%rfoA2+n9mtq#b)>z)oV#S|> ze`*aW%)c{TAi)I>kPJZjaETY4c1dy$3l53w4UJ=|V4 z;O)1}2IvpZr?_}CZiV)DsIq_Sr($k)(HWg0lXC~tEt74 zuncBg^Bee^R>Fyz^st}eB0P3c zb@-`W12j6`7PZ!%IjGUuy$(!5kZ2j4ekuhHG0?#t54*XI-6gHk~EMRZwv28{flXzBklS?7-v7umlJWUjMuHx@75 zwF+~DY*s`tTJ+dOO*!!@I9I~o-l)gG=WJe3D9B$e>(1&)m61_pvzwVrT?gY3IJd)V zhfGK~sq|KHm~7Ij5JYy!Sh;h=ToK{QBd1O_BZ?Esl@dNW4AqwuX-*-oj0fm?hRYG% zC>s*&!6;;-M+kD$RZ}Y>vCq8h%rtkD%Kt*f{C_Wp`PT|}wEn}u%$W2xHT4fMs#rw1 zk)Q({)Kje(%=z|&+V?G_08SDz068w_hxbvwYP1lJDur&<%+|DiE4lB}SWjL*JIAj_ z+@dbtT|kbw$ey6lGxhLHtA0@_J|29*enih01?GH7wgJ~qJVwJF&XoEYS(BG^h&@NI zZB-!y98s2`0F4gBwpl*<2&DI-mg#>daAN8V<>gh)s_H|YkC1&C(7TFovp5pBWbyTq zx-#?|`~vU9zdQKL`~T?UZh)A#+HTG_SC&_4n-b6GzZA^*ON@RP%nvkn8;VDvZq^JK zu%p+1AX*bhV6D3P3=lFQb^ECsm{9vpFa0#>(j&BPdAwuDuHj0k@8|rL$zYkFc_Us4 zm4cVQYQ;csVJ-k#Z1JYKo_gMflg4V-`9NRz6*&alNorzjus7mnnGWvM052eb7ZmjD z_6gYr6r!Ozt^vscyjROWf_W!v+T{fiw=GO;oHI zz!2o39{bh2;S+Y-eRV~U#^LjbPp{>~yov66(12|P^mZg3Q-O29wx(XW=*wncT8`A=HS|z&+wE7~tRXG!tJmzkPm3gqHH%Qn_Pq=TC>-d0hsW zgPFI~>d)%cc=H;E`$*Z#=5I{bDIgC}i9vp|5$JOSKX;;DWE1|PDu5$^*2p6RvM4bx zr}D-9|JV9CCI;#0$5LDi0Uo>+Kg$ikmPk?gvo;F^8!^LXB#=~u<{j~AW470lowKV{ zc#?JbSR`H+*6~*T;i4r^wS`7p+kX8r{vG4A^*MVujaBLlv2{A&Zg;A7d3Qeg0wz`Y zzkz@%pDv+f*SYV9IVdwb*(-hlGFX6#JA?bM{T7TlpQBKo$4>lU=egbht#@y+8U68^ zqzD>+uWxMNLRPfN^{GR%m~U*Yng$M&3_GIP?5E(AS7%KxNtzWeLnjofJM3j%iMO7F zC@CNJp+BhZL9#YmIm&%dWuYj9M8T*_Y4~eij?5ZP{WUHvYBd>^Hw_mJ+>ErN5ruE< z>2$4126@QZn>V+HA!FhcUxi2K${He+4ZK732Qr&*oGq$8b1(Z=9OU!z<@u^ z4gAKL;$s+f|G6`+XZR|z)>aN;{qt59+yU$TVE09pNv$!e&>j-B?drQ((lAS_Z6>0o@d@M$FaZV5VA0>YVKv< z_uw0aw)#`*P7+_Q8)X~^5)|~cs&FLFTVZq=W6{7UM zo=ggx)g)iYHh+fgkfjaBZM`I27T=yB=Dg4+3!kJ8B|l71QR=Z33Arq#r=cGc&wim8 z!_+efV@7bbkF^M314q!;*xY8mAe4@obZK#@_IrvlJ!_yE#NUwJZC&{@L&*y~3vV?y zb%)NgofWH2v{bw8IDY**;^l2mr`(=xldiyx{iiI3_?(7T8{6sUhoU_R2`Eo<(4q%f z$tiXUX-;Uzg}=2H{~-2-4-4Cs&XonRar}hErU-FX*CETo@hc=k)%q{x=LGP}R|DZT zA!N2HigoYTE%2PSaeyc+PI~GN77Bgfa@vab^}7Y#x zWpIAd4FlwyM24igI_fhB7v^TK&kX#vU(wk`232E-7}J7?Xpyr(Y(D{`NqQE$%vqn!t$NRMvHcj2`hcc!O$7_FmqZT_wJK;u~Fa7c5n z)$P#R?in?4V2NzWH(Z}rY|=cGBBAV>++E%K19Dt|h_aspj&Vpiq*y=yV%}Po)b^ws zd3eXmnOkCMzZ)gM{wu`*qH_8w*=&UW`!a z{;@8Yd!g~7RnL_x$n98K#BfesLP=t{@*8udUS6etL(+JBUTFTDgVUD zHsI+`?7=*bH3dlK&Z^QjC*xH&P&7kUICoq$Q3eisZda1n*>$COVIH@?F~DE-odind zY#Jx{eG%Cp2a4P0ZCFmB7r)$XKbKo&@7-4ZAsqd`iJlPOnLpR6lk6Y!5cOY6Eu zE4ix?*0tqDEn1u!7qDBDK2^i~_H3(r7to=FIx!=n3)I@hf={yg)@rp`)WXeH%$cu{ z`hb^nD(cOIXMOtRP{d^mPUCe*7;ip}6X%3r=r*3L$7|wNP8uBM>>p1$DNyJKJhIM> zB=mf0xeLb74`1l&e0ejb_k*<1mw_J|Tv_@H#Q+Zr-J91EKGY! zWF2HvJ=vSfU*R;+`30sj4oXGnNxk5g(ut*^Fkf4KQ*aw|&(<-sRzEf5*s0*`B%4Y+ zPI&k3tARdJn&%?n=rZ|)S2$dq9(>;{pf>Z)ibQImqxrFUP>uWAw$=~eK|qB-WB*Vw zCh6zLuodm`SG;-E#ln$*J-ut%S;^kx^AZh35 z7e8w&U*rB_llZFRuxpQL{h%@9=H<)g;^~(*m*k`KYZF0!!34|BI8+jRd*y-%$-g{X z?r7up_leOb5*MLkhpIIE#cv!RHWr~txxd;@QA5BK4FPWpqI@d+K*`Pwdzb}M^VF89 zgR4((9ow?Kifu~*<$iE5^8ZAwoy_26zkl$57ra@t)4VGfOIu@%Gok&IUsislEBH=c zWqSTBY!K2YKw=qAwh}elFmVAj>RM7Wf|Xn4oI1CV*2c@ytpO-h$m_F`QcP^Vx>C?& zVE6Gxz@~-OJKP5at(%*;9mvpD@_~OFV@5I<-OKRr0Ns;Mp zZ9}QTt0}kgZsEAEvdPkxVjSCWX{sO{gU zg`lz3hcF4mn}>{E?W+_Ujh;5&;)4&~Bx9_b<*Ba?$y(k7i`mKsSE;8~Qj0-S#&c6T zGhUVzu5M`XOCtxXYESxf+&$|!L-5i-*!jpqqbJ(U6Gzae)E%V!t@hr#_hLvz zLed3e2A@?EzOl+gbL;?2%}XsUvNJX6PWj?gY5n5IAaw+*a&Q80*R){mvdM+OvAMVL z2R(z~!M``zFufOsH3f)0R(RMB=3$n*Mk=lmYvbT^?y$t&lm@3A%v5$Ggsk5`4rc@^_wm9k! zhA*7_`eJLJ(x}OEK7+>8ttH@_o$b8xq-W0Dyc+ZqAk}Y+Vr>C2q`G$0#T&)vfY+q8 zdUCAemL9+2sRQ?pWOA0)3Ok`n&B+kvIMAD#*#5B4EhAD7PW>1z(#&dn7`y4=;_@8m z-cCaHj2Q75lr9tJ7%s$ly{i)_Pi)=TeEd-0D60RPj4-09xv$?p@-gNMs`N@9Oa|=5 zfyIl<%Ev6d%&Y0(O&pBXqR&EJKa+F%@-;-YYx0OS?qXKs?9?izei>(QiNae2i*j;`P(kS~d#CaCn{AMuz=%ZMfGlpTb|5@3He^1TJ{GX2L84G472FKBd?4d`yjVPi~Fm*M-z z+m-I)FO`vsd!U8-%@xNuU*eI;UsWoVX*ZYQ*{bnok>}~KAZKxZ*NQOJ3G&e}?J+>i z{exa(=~>AVp^A}UZ==EZ zp1jz%m-#3yrqotKpO~Mj+0nfx>^A#&)kn{%h>zu{WTs8ND#lD+hgTl6ce79SaJ&1Vt18l}jFCS`b z8Tx$wto6v^brFdn)E5dO<10AfBVs@(0a0Z)Kl1l~)+5vRj6E){*`thGsotZS_){XZ z&w^(hR9#i-_dhN`@WAjjA{AlQ#e*POSt73PLHX46h>vb?3DR4*C)!v0*V*7=TnMUZ znPU&|XZ%g;A4-SK`g1j7`Wirx>A|o53dcB7z-PUCBcm!GzdaG>rMc#vFIvemp92&# zo38*7H72nSzcc5q`fAbGFGgbp9|n-VAlZ{NU741JKyEpes`yBFVRipeKs4W+M<`51H9*;Qeo z+`YO4IV18P)&w%v{G?iTBfsI!cXmiI;X_3ZK5{?GSdH&)EcHfQaq9uAAv?WT*Pg-T z0WZwcKJVh!qAB?y?<3GPnGLM4W0I-1SSvGnS*i7_fjU}Rp#wsA*F&tKcetUN9uYID zpN$UfU0D#gi>f1PiUf#t(5~k+`-E;oIsPAs=HFL9=eBq?KfK0yw*X)B=Gn1U`%psH z-SoP@bkJVl_#YEYKR2^Mdh@jDdn~xtBhKO%>XYh=VzPaIV+7$5Wua&0aO7#@O-HW0 zkRRK3%I}+DYZ`kp_u^0J%KoSrIsI_nPrD^%B7aAEixJ?mkLAgtdijUHPPfr#U z_w-r&eZ%ou+vA-oahcis6j&`wIqrbaH8fRaQMDB602y3e7ET`?{s%rVCLB2vZ8ql+ zy`s7q^{_wFvD}lTu8q>>YqT{1kF@-s`BlzH@SGewS^}o--wyTEd63+t}7Vx>2AF%a%VZiEkLN? zayuuPTPRN~;E>=VC?4mqTmy$y!>QLgShy{1M(f>GuvrXBuxU)ofH;$ZI$=+vbF9xi zI>N582f2$ob|5hTQz`j=+L3+;35o%{)F{y`(if1y{LCS~8z zll>b09T}8?84O*8`p6U<=$Y+@db$RyIBt}0Ev?hqu+rQk0trC{RV3+{^6Fo09G%8I zb-Bl;>fVMVH3znr<`S;3eTuP2Gle^G7IZ}TTFW`NR_tC_9akFz2qex z{2tJr^lxE@5d0a=kRsVzKpQ}L?1;F8pl0kwVo#jSc<}nIm`adv!qTc|zK2pLG_wF9 zYln79)Z1Is*zI*OMkoI37fFbreQtuX@{a=ZO_r`){FT8#eqnT1$?!|hs0WZcV7KN7 zKV6NY?wW+bscFm&EXGM^hB|2B^S*Y~^~d2d*K(N~VG(5ymqn@oiFmww{AHnN)P`1c zCx2;u+*q0PEt51#n)Imp-gEcD2c`MrrZGYAFUo)6y+>#;m3l7 z*z^*|k}&bDq6fhk-=B_MFzQt{*g~Cverp)keV>wgz5;tg!E!8YHspehxEN&{(0McK^ zh9OB7ha2v0pM4xNSC-~9^Fd`T+S(=W>7)`KV(SwJ`jJprv1B^e^vG9#QGm#y@lF$R zhq{fRc1&~D`n=KQL6~?Jte>lp^zNca zcd5Vo$FlTHRMtpKQ3s1PUVheXC9)`G=Vw>n?7=xRJz*v*n5KKyw@I@mYqg`)%R&*3 zylUX*6DMSY_nHoaX#!Ctt zSEK2yMF0<(Y0%=s?|EUf1qb#@{Xbqan@c)Et@85xY(G82WybH*ULpi+K|7@I)#$J+ zm7F373VHb!bhGsWI(Y#<4Yfp#pSSc5im~{W~S?NhKLujMZ zT4aTuJz+&BqWy=S^Y)}0h1pwQRTCO&QMGy(ZxZQZYCwi2JFR-?1Z_RFAVuW<6 zNB`5ublq)vN!TiLhD|QwLS9C0avdS5)2Wxh*5wZ4I9Du2W*)%1S@``oBw7G(<30u+Ui6 z@pU9*+l$+6QCoNP6H*rz1X}x5Imbu_4QQs|LJD4V^OpZJ_)v&g5{cpTOt)gaCVY$b zj)#WwH}>_y+WsTIM=rW*r49#H=2z3L7RA!q+Pu7v3lFw-&a`TV*YSXCN&5g3=4fi@lPP9C|-!UE#fI$j3P@{tGZQw#zK(R=PIySuP+23A5i~M=FZ6oRvA~k>aTTb0eCj~j zGvq8^4rEk(A;ZQjWfXyvP&?*MOZ4YxP_njVMjl?QtOQv#Ju zoo7>bxGC2j0poC^fZip@lnW%0T{eCy5L^Bpmg>AQq8CnB0H;5*T6Akvu}3e= z!dmzrG*iw|ej{*|+wFYoqS&@98-6IC!|zF;Wwd-Mk=_ zVb1asg@e1F8LR@GUY^7v^E}nYe!5nBvTTy8Duc4*(s)AS=yo2j3uAPg`@t!}I}EU~ zr`rf)Y|PoiYk5i}{*D&=k*)EXcW}ObMAvS_sG@Y~{MC|VK&VWJ72GGN#!wSxhOM=m z7OnSXdxxI$Dc@(Kkf?x2M>zW{Vk?K5$_1L5E&Rk474)HL*Fc7PUU2v!lyhWsy@6QNnla9-WBcw3as~y=>36CG%MD2Z=UzWQ*f(jN=yo;k{X<=N2qEBuiQiPi_ z5b$`N4~JGT0-B5P$2Xbg(Ru{QCt3QhWc<3?j~VQ%0!Y|b|4E&hlYql0bpvi@#n!O6 zZ)OoD#;OKNZ#u?_o(6EDUTf1VS;C^nPQG0mUR3eKd|}Uf5aQ50yz;E8Ps83!BYpcN z6xD)U_gzx1_n#Dq*3h7XZgZ_6a0a#7o1%Mse1tN9@Y;_&C`0hEAPP^qS;1)*vYv7T zjX+fi2u6dax2?q9cg~Z0{ehxs^X6@1@@{%8z}(5pVk<3UNP>?z^yPt=7`ESOzvB!{YW1O$G2AP z7{EBLf4*dbA}$jpLp<-}n}44;p_aCv>w{Y^^lkUAL%8EjNr+tXXT78wT|%S&DS`fX z+3?ruWXucD3Aej!qE}DPy)9~`6|#Cmv?6+=^6M(eu(ODvW1QH$DB!jNAd?RMRUT$e0{iwyD4wKgN8hFbA|Cd+yc_I8L@wIT7EHx z+eDUxtu*}4`f|juv~a^-jTl$lb>O!Q)S6J>SKmovD_mMUq(h_A*O14_ZOE*`PX!F@ zWAZHnCQqyM=yR19IbSS`coEp-zD`U|D&C^Zx)prCtq5B=>?FOBl*xxD>e6>2*Q;hA z7{9VRVu!{OSxo_w$skhIUt6Sc5smqr%yWJO2Fq%1I(u*eyyKg-4S_ zdzi+VN-rl<$gE6 z(RQSFdsWj^QBe>$mD>P$d1dtimvywaKLXk-fWd&1jSl(6p030879%EQB8$4IEDSN3 z1b+6!ry}JAQE<;IbjxK5q{Xf}Dp!UoY?Aoh)Hf!l|7s3hD~${t{gj54{UBu6eTv4@ zT_>BJaL49oik8zqnn;^(*v$8-*Z3vPeX0%r$>~E%eD1(NQY#<%{U0*8qmwWU=s}}? z|5T81b&xT*pZVm4v2{t&Qa~W;A%--8w*lx-3$Xhe&N(D2R{PyOnPA9_1sQ63h#8U0lh`s(~hK`!~=|v3-dz!%n}nZyH|Nw`FcFm9?nU6YVZW{aOo$c>jaQ;TatGuzyjMWFNWy*;`5*m{(ai> z+{DIv_@fwi0bphD-0NjrS(9GkyQAxSgJMzQIOn)ly!s`+AasX=LDW(~cl<>YE8_zG zd_JZl2^#SlBDgVm^?D8!M0+m47hiF*@@_&-E@2n&bAO%5A6`VbtB?EwG45&s*M{Bq zD|M#QJwN*VjVx#$bgbNzK38;OG^d{{$O%E_^NZUJT5E3eWN49vOc6yKPl7%b1y`}N z1JHfef}$4-`uFw&1l7|iiKTHW-7leC_jN77prd1hfjj~k+W-DiPh?Tn?t4<% z-s9dtHfh@0exIKJG;AfrjMpuqGilB8UtW%cjC%0)u;M@Fmdrg?2YYU$RB3pz9_PhS zxbG1`-j=^X8UPm`s3w148-Vgb>mcXw`To+JoBS~Ot&zHnKB-Zh$NHm?5(kxMsoMmq z2TXZ(dQWB~hoo~zFI!6|Zy(J-m?~M5i265wjMgx_@PU`PaPV|?{*SpC?w?8(8BkVi zva}^*fA8zt``E&I19b^<&3BO4mM)JBwF3Dy8Gnhz3RA+Q1cRC+=OG%Bu4choo@!@b zE}NUV+bhy8{NF-etmQSAhQ9fN_=ptyzI#(>)KGd?WF`sHh>M7rONw~%n3>jrR-qY; zwD45hG#Ig|?mN#>wPicXgGaTru=}qHJu{N6P_!xE=?56ONceAGHP;>Z{P6E`1fiIC z$?YtHcxS!|seX+Z)m@m1*Chw4RQ<8)(V1f}sU}fR(KJuN7-vjgTNPaUu0q9cX76)& zB?CEY>H|le*4U{fU7$R#MR0(UAFx6l>J)wA@1Apwq5QC~45}kT91Z05e7rKt!T+T)xbbK_{>?__xHPAfd zrNxU>B@}ZAR86n5Nsr~{m1omVS9*8Vif_PXCmhn0)e&H_eOv(l1ojoT(w0Tc1^DU- zXJs$+DrWic@_OENzl&7y-}L%UxmjV})GzD}ruoI3T%!69@B-eoQjI6~wHk<1>u%`f+r zf*T~Ay@!=E!5=LJCqWp-C?9^^!1)ET$eS|H`;WT)IQw}W$2 zJq`0r?VyYQQ^nEknoF>Nvtm;0bS3x43L37=Vo{U?n!N&WP2Cghh_S_BFE4;9A=@sX zMe7yk6f>7tiAtU3>RTAHh?FxI%}{wBEU_3~(G_5bVw*;&&75XmxlWZQcn)I-)Xq&b zSmM+ZUGQRayBsRnrKx;M+eVZH!jg| z&nHZ(-QRt$y0wn_jN}s+c>-j8&X`F9^)0BC_b1!j4I0Nl{>t)gX#_{9v`#!_xx6)E zdIHUs`1Bh7PwLOzPfxpn)1h=wVU3;~1f*#by zVz1=Y#$uhQZ}zJlaXw5=8c8Ia(2Sm7U=WV77Y1Ut#I&+XA<1HojaK*sPq+7e0%A)A zVj@&le%I^>j5U?6`(WSn(f>RNSjW=KfYz+aV(O{*P)mqyuB_L z%ApV*^=3+E>4r?!;90Kr1^Pay`&I39MEel04@|Xq=d5o50@7Ous2I6@=J+eWaI`qa zJhwjv(co25{TM$BcHCv0*gDyE$f6YVt3~^r<{t-`1k_yy{K(XE8yWtRP*NtQK^D4v z!ZClr(5k&95Oh{etprpfbMKOcNakPSaC}PjR!mVFB@v#xTizC2F6`BC;4eP$1kEWn z@M3DC*v+VXHRkXsD^CYJxpQ-&sn~{m?ve@&a$v&g^)(D9K4>|=`Og{tU0_0OU*JYV z6C=L!eB6Tc+;Si03VypiSp$1rKTl=!ki3%5O2;yk%1JG7ce%WWzOm{FXeCds)S*6q zm;M^?NvCQf`R)?ki<@ldjbZxWgWYbFn+g!hm1GndP$vFSAoJC2}0LmNf@{l78OT83H`LV#4lD`Iwx_t`?sF zKoFK*bZWXC@pzFT)#>A*t(2L7eviJb)Z!`C*x%m{YB=Ir2I}X0Cq14e_F3?E+NoF5 zB}wC4M8sQ28i+Z^znW)0Y@y1tUGfo&w3X!xfDHTdIjVnm*F&5{!+?1 z3htmwTs;wD%J&P(e_a-s^_&z0@WDVy}4`o8QS#u(V1;!8rJH5N8!Z9Ko8FJXP#a zs=(Mh)_H7&F1}68&493$P}h|~@(@ZL!3Va|H@{fNG8a(OPO@7Jx=A7O6wp-!K0j1| zTYOW~&kcI;1e9NO@mk^L*J`?45=o8gf#{KeU7bFeq)bb2apaNI>5r0W-5ZtMU%>I= zix{v8jT{ckih)sv=LS?&9g)=M*2ayk^Y7ckT=zp2V|kq+f_B%GZrP(Umf}rK^{wOm5x=YaI@{a|IBnX$m&MFa z^_>qmEkoc;TeK*^&(yg4zQ055q01AUsHY3#UHkC%33Qil;xl7fuIH4VNnkF!*4Pi! z45R)bu`vvU)OK>w$dVIz@8@cJAWDO)QngMxDw_*(POLR4BL~c+h9j{v?)|z|od)Wb9doN)d+O2uND9A8`i|r&)>WGCfPsL)MKu)J;emT+qIqX*z9-+j0MhOY z-fq(!o|w0>6k_wD>RxJdsdhqcyh!p2==`$EV}Hd}sW6omUQab$02Gzbo)aKv0tQq% zt|7@5ksB*l<_*IN-TztkY@7Fg*ep+a0ICl#Qz(wt)Z_ofW8z}5Jv^oz16X-;>QPSR0OLJ9N^xrzEMtPExS@j}!y!<~2w=C{#NT&ju*3Id31kU=@zA^0dn_v!!% zykt>H4*Gg~wB(5~5tJdBq0^OU^>reFhGt)I2b|dIH*%+O1of)Vq5zJpI?6P{X+oh0 z46%Ha-*1$*-ahf{p4~|ub=CTPG(EHUwaopS&!X5Ms)!I*@$Z8zYIja_m0d@5y4=pr z-yCfsCDz-KJ~y4qLjPEzPiH*&-ND)F0$a*o@?$H?0nIT83@MgBVG=2=W!7)F!CAk9 zXu7K!8BopDwWCfXahlQ}#s*dvIn%1;$B@VEes_?6awEs9u{Ee+e(_c2g6RMNVdB7#Z=cc+W}a=D@JVK{lq65>6!)jN{gXGqw}J za}_(BIr6D0r;H)z^j84>;NAkqzFZgC)`lXuWQ6??quX6~ZhvDGd#Ke*Do}k=cxQ}b z_oXi>=!@AVOq+=8YF9)O&30Rxfi}v$|NzQ4#+!H2f&Qe5k^((P~J@{I*AVX-Cq> zbSo&^>%--PUbU$UTn`eiTqI)8Xi%@A>BqO*Wr;l+Uze23n8&#VGeDYCf=&^#5+|c# zH&KRFAB}-3z=iFZMUJ|;MuQQxG0NaJ%VV-M$lsHSSag=@QyEK*V6L64gHPxbp#}rN zPp7Yc*6(^=`h)j#$JY=%vzI^W%Ow=#rn?rWc@4PC%9*2z?Dqhie|ZUbcmi!0DnBUF z_*=YTU~Aato@*Nb1GJslUa=R#uHqmyl!zQG~*0X>A)## zwXvuS_PnAG+mQ%ga4sIQ7l;0rYEqc5cYL=H8|mKzEeS#1yCqH4x_{j^eir!Z!fB=o zE?1FW2THH!H&Oh^Y>8!&SvRMH(n{9v`{SZ0d^t2xN&aInelqo?)YXPBEySiFXWcXM zBdyahyMm&Y2s=tVih~ZE@#T8`Ikv%Al6@sA$L*M>#pkM9lvtAfv1^Pq{Qq$QIE~{` zievYeG?6gH5P2oAg3n?sBnZ~X!*GvtwyGh38Ae%`S>e0izpXq1Ek(Z4&mA+om%i?O zV%=~`bKdt~V_LJ(b$NEw+{^^q1hGHsRS;t`3o`-$I!S`19xuUeGaVC=UIo6loM5ff z?&mbN^Vbp2OnOT=RaeuUG4Pt`2n^z244g%N=tLWO~%OmwP*H% z7EIka>DH3ID3@rx*1DY%)iu?{6<_=k3tdp$x41ZbAD4~WFmAx*2{N%Q6E=!8OjY0`k*0k?5%?bf4R z#4o{%PQ~T zy00(Eop6|C<_%-#K3ivNSx4K-d{pxC2UP%lPK|M5ENH)Px=q6-s_(Ef^#U)Qf3cof z6GST^iQ4L-vt1X;(jFO#K3?&qwbm`^DIJwDMThP%B=zA%XX?slEqzSAqf|DF+CZe$ z=rSvqBTfPa=REp%fl98|5GPRbD>@>a0T%k(b>67gl%Muo*8~m29y76o+e11Wua16_v4)>n ztFI1nPUVJaj%mZNtirXM1QJrztxR zN)A8oPBzupFzo;^f*AsgJJy%2n1U~sZk&5=y}%wLZ!R~#tE%Eur}Hh3(E)0~H5s*e zm%KCA1h;Y{1@oAD2PzM5xg7TVz=3Mi;`v<=Zr?@{rt4-$%_ocTulbE`MM|qT_N3x| z084TvHYjZ!)bJr65{h*x5c1V^K6Y)$clXtm)JWfw9Kiez2@SYhmlkwn@DYzSu=$s4 z%~n1x#OV{E@aMkpw0H08j4s1>9nMUj(^_ppZDeUFRJ|%M)0gMShZzJn-=O`r+CwfQvWA*mBWn&>FDd8-^dp& z8-fHrZ^a0?uSQGv?DOU%WfNi`FIG+M@**eL3I#U2A%$upBX%w02NIcT&ix*d5^iDC zfL|3*G}m6}A|pj;qP(a!_u*Bzm#(qB-3I1|vLP9S*JRN*iY#l4%DEz?Ma_=B*Cud$ zUsjRV)Ix+3@ODq0f4JQ6q(P~W_hf|l3x>A5|IXEw`EwrBmkV2H}Ti} zc&l_ZYA(UH&A~X-9uO4i5$9ZSwGaZQx3a$)4$bhh)71MfoObiS*bw3PcSEG2UJaxD zE{e8AK{8#%?1XIG0Hy-FD2N>pMJvtZR7Uojk?sH5Y%~+lxr*V+a#e(#J5^UTaMWTT zEE?3`Pn(K}HZ+f)HM!2bL%J!Svezqk@ry~vAK1wz3~?x;iW0t8#A)N?=R@?0xK#Jz_T; zpJz1I0!1g7cC6d_DqKoGI;2FdZ)YW+Qw~LY#?cGKJ@HLjb(HOt9>72R4v* zX<7;&S!+h0NfE{~-uNG?=Lr)mW>9kF5H$4r0zl8dCA{q0$ip?#WPg>S)9>)XWVjxU zJq5QHuF77>s}O23PIyONjQKsw6T7qWJcb?x$k*J(@IU>!PVwV(fRtiEb=_e>L&+)E zz!<>nFdBzsT-C*nYSuYeL{h9%d;XIgPqz%`Gbo@J@o=ST{V zn80++AN>63C*JhCg7uZY^sgT!@Ont6nSki7_2F%NY!38zun}?St@yVp+k-!X1x?lC ze$g_3S>N}QhMtW@9n3L=;ylwOWn2Pkhw(t$|7C9%Y%s@!bsu2GSlbdp!cNoNufo-I zmBgd^P{HDKLUy~|W&!H&p`X5PrEb2TT9gYK$Jm5Jy#QP#&_fzK+4DCei)8uXiQ@9+ z#dS@HZiZ&8E`y%#UF!h2l;Zz~sIQEQvi-hRQBb6$q>*lrZjkQo77%F=keH!Fq*G!* zQo3W1p+h=`kQ};)6owFnA>YyG|6A{ud#zb(*1A7j=RRleefD*Z-sGBB=`vPq$Q(9F z66N>Gv+6G6WJ+P2VK-dIrVLaHM>Yz@7oURd_Weo@0QBXHVh+4meAlQ7d>N$OsgpM1^K6kYtz{l&YlIr|9;E^H%d~uTi z)c!*ZEjGq@EaAbo=IuQO87~;-&;*&$Owj`>m9IJ4Us7&`tTdi-E77U{v7C4kLH{ZF z;$5}5BJYM!Pv93ehu1={!z=sTb zen)KGz}y%l{kIpFUrFD#yOVap_NDi3`aruKq~;IlJ>{(5Uo_1}+B!F5#p(y_ z2E_raF_zhhXly^nKcunGKDOpESecKda=PPa`q{gcPeJoBW}$(LEneJllx(1uT3pGq zHJ?(P25af+L=g#Mi8KcCsrHL)eVG!N$UKMos6%wQu73OWy)&J@4S-Hv!y|S;cw0zKw1VwN-_LM6 z$Z0=c1F>{duw~9LLH%Im=C>2aw-?R=r4B@o8iwh5O-53^sGE&+&f8}O(nP+v%6ujp zAHGi7@xn%_eAqXJr=TQ4@ z6Q!)Z-hkqh>{s_87*^ARnuZVjS;<$01T{cbXRo~b>|3cD%6zGyP0|^wtV-Q5>gAou zI%7+k#&%4@TcU#vrxop6KNa)WKIYHEH-YyjulHJgZS&a~i^sS1_R0d`#${5@(&HDR zm#q0RTHme5KgX+f0+kBmeSBDZbr7*Sx+aqWJd;vMzJfF_B-HX3XyNX3k2fTsfdhZ` zT9jKAJVAjQ0L_N{CU2>2VJ8S;y)LA7yEI-!FIc@XSv*_25jJUxz9?;TH z1Y<}i9E-YyjEsu7+qzMoEhROy_c=-BtKE~3`0j1qC#Nr-ER?n_8?qzR@aq78vJ$fQ z#>3NimzEY71ri|H=CGOCJ+{Ou?Dx}ll`Z`z4<|TMb?ox=#Nrn`+K?(wXfb{qAeYO( z48gxoF8RB&-}Ond2ubjH9}Mz-G3KmP;cU4X`5^>ge-=L!Vu2yP% zH-=B**AHe3?iDVCgoDk6bu%eV4)U^gZXHz_$1rOCQYzl35aNjpnbwkk=?h9;n(ccv zPp&)Oe&&(XpRybcb>7s^XU_Fku+ZHR6x?z03r3xPvk%iOgx&bqdktn%Xl9R2!Ls4%q0w`y7ky;>M5#-jC8r_0|H!#+Ls9 zUb|=G#}{GOBV17dFnEeRn0qI(DU?bk7}1KYImonM!06(@oK6vlmmF>U(#!;*ajQ$>=rP2a9XR!I|6lhx=^`cG7`bs-0ZiRjCqr5)bVjkCSQCvoR$j+-(M|}^LA?_Tl zRySHl=XEJAo^n!9I`nE88yRu9D$%xS*>rqHJiTJ%<9}#!{1!85bCWoAaPky%NT5wM zg`{Aqd;HK#>@Jf|VOX8r1l+jT^896(e@*sSItBmHOo2ZQ$amOW@|PVed;F%Z%}BV!bk@)d%iBIH zv|_pg?R3PPf5{Xa`=g3gGdE@NUCbp;pHl6Ov9NTVCtg>3;F(lQh*}SwR2TiHm%_IP zY?N(t8+R!1^48{BC~G2oGd%@490BXO_RM?L1e1XtP(Tj@78~3wqEX1nc%ei{z_09} zciM-_1v9Nd78(lwq6QsDOYSry|Ao-sAA&BMHEv`YvQfN6rhLQ;9vaEQ82zk@Gjr+t z5SDk~G`p*g^g^zy-8}N7gx|Ttyl)4E<2~d(kU@OqD@SGgRsI2cfQ_I9sz+NxHX`)X z`QkruNYbAPRqp%*Rs(ICNekGG&O`~9M78oGcQHnK<#@P{k z`e#mq$nV5d*t{Q|h`FHwPo)+c+^4regDN?^GFzeGd#uN@DBS`U%>5|OcnLF=O2z}E z9ekdIv?gHtY`HD>3I6+6D7aQ6obKNz ziQjpC)2oz+i+s58i@y$rJm)AiwOyL8dw-&55nlQL7m9!1b{CsmB3Gl6)jG`S?5Ei^ z%F%Vrs0&l-+7-=ZQO z26y~9mw$9p0919`zr1@J$_DDY7OR%jHXqpKBfMpg7RZP;hawEKUwkktW3KvS3x|@L zZXVM&nl>Z#q+L@He{Mc{<}%N%kHX@{EWX2qGc@DrXda+=C;SyT#<3%yJ^NPujNa&t?AE zsepr==YaV_$$S9+cYi=o?3B3=Q@skl**bp`i*b8Z(B1yoPrSX`8o<9SmBWQT4>UZ> zVU$!me}TMnT^xD%oGVu+=KL;rL`5SLKT?&vkP!xhjWxGVaVw^gb*eCLrVx-g3w??T z#U#vE;H=UU+y2NNXJfe`ag;2A?;w{Xm838N3Nd++UX&E*?*Uy-!s3}gu*0uuri9uvnU_a z4&huqC3#E+SgEi!{^_pa0;`0(t0R8Cjlv{@CQcCbK2_=rDNxC)z}I~OUJr!mm7 z37Qq`yRGyRaZ*OUJvL#_i+}pz0xq#m5Js~oo}DJ4N3S@Cyw9xZ^e)ZX5-B^H556|u zOW9L2I7jg>98j8^u~E4eIAiKpWV@aVRrEM1j$ty4wz5=Gejc-F# zgSnulY@c}Ddn_=a@yX{BK{tNm2y;!#P^5QXSCQt4aN?&t-4=HH?K5pb^kQ;z^yx|j zlI|WiV2;^_iXjPKn-2&wIZ_*S=8(yEJVafAhG(4Iw@hc2{2vU(DULO>GmG8TP^)p+ zLeI*JB1uorlK3EEwn_^o3+KIf&d!r^xXtl(St*4KtG@cT+XE?@Y*LsRd?pg$tpzw> z<|uv7oC(ySUf4uJ!`s0c@jF+mWp4jzzEa$_HGk{mRHbm)%YV45nHoRnZ@h7@E0?Ah zj)b{3M>pE;I#P3lD-giT)jax&a{#udDqY|~^;YXojN&wSNxj^iR4mV3>DJkF0ez8- zlWaS^a{es~(sF!H8(i#%n)t4blD{{MS9W{)4*YZj_{D!Ecqm!#PnNnA7edeRSkrQ4 z_R}(xVsVp}tASu#loU5<1&$2+ z$0l)4(v@g_6U80AjspH>IG3fCH8FHh+fiIgH@xx!(JyjBI8G~>Hw-6DqRscrR2{gm z0$crPc>m;mk-$OPU=LiJw-YiQjRJ{aqG|G;r>{A6g?O6^D8$6tQ}X-EJCmM?yvrmT zb$|23?Q4z~G47|xk0Bq*P3Edub7BM664PSw)z<^G8h&m7HoEW`$5{Yt-J#^OM%Db` z-FZb<9%D33wM$)WsfGi3%^ics26Pp5Vk=q< zwJ=8z?op6oPJXhhSMQV=vOMSAw{5g}FJ6Lkq}=rAWJO@%+PJ}zrDaFQ@%#yJ&kXw} zAP`~EcdK}9Z;F;@SsTw4A=N58xhozMKQNKWD6h#M;izF>_unSypgVmBIK~*FMNL!g zUm@;}U>CuaA(lqf$V+S7Tk_v)z4%+7Nd?HCrk+<1t~PV^3;&=Hg6az8R4FQ7&82~U zTH59bHia@PFrlwlBFWb;Xqm-fkFX$)&LV3s#)yQ4}h zNg!>6#uh$OqV&bQl_B)F?Q{(*>Dk}Bpjg(EwWYdrY4uG= z59;=d)wRqHJdx{uQ1*!HxQPqmSkc9!_7OP(oz1v^7^j62?FT$vGj;U%E*Q4Rgb4z% zYahIdMz1q3?W5KNe>yyvGhOA|Ef{VJthpNa)i(Iy4qt;UI@<4Bp`cX+Z_=56YV}uY z2GN`dn}ELSv;nK2r+=n@I@U@47TCO_M}1UiSz41+IOmgUJ!~mK4}>|=`EBjT=3*v# zQf)9Z|0v`Ai8}C6CMd*m(XD9eLew-DIg1>Q-iv3?t&$Nq2&lay=1jY3321>0O?W1k zM2vt53KnstC_Qdd!n7VdG{M)Ho>GNTIQ+&XpbbPM;M88!t{gBrgt|9*j^xZSraD(v z?*YupWSQk+XIepNcp{C%5stNS21OA8j4n!dTyc(ec#@7t%SCw0^AE4)zXQ{23BUAL zTAdFQhzEGwlT)f(eep6avh{Z1hRnuw1QcIS7Q4F}n8MwSt=JO1W@^nF>*%uHmx02) z$~YB6zyfdN;atLV?n8$@tMSy7j$7eIN-1#OG|5~6l=gfVNVh+JmJ`lJdE@v^Y+;ep z4&PC$#f%Jag>KQ}yJRcAmX-q0SRjA?$?>&ka>J2dyUKs@qGA2PcmtP#g-{5kry5gh zpPz|IAT|cZoD5ajcs$N`L1>A&7Zp{P=4iuA^gE8=IB*xOq!r?amsNI#5RZ6Ej`P6$ z#(9KH7^iE%w#SE^c%e`K&+0qcS9xD^&dzjgHd^ zY)W+qZq84sPM=~ZDiFcoBTYnwxLAy+2Xle z4Hc@c#4X;Az#~OsV2v7_w=-Q|WN*n_oGb6=*C<&klIZd}dyOs`x8sVXa52ph6MFn8 zV%9JNOf=&Qv{j|)U$UD9?kQ4ADH+%rgN-MHL=FFomuDG=k&$810cs*RMsxTna0B<|LG=I3om>?YC(sC$Zp?-HP`(2S#y>P z^NPQf|G;-6sDQ?hEknzbgX9+UHk+A11D0`?C$~z{NX`43BVGe6fCh501=-ws>p4qV z)x1wF{t)%!y@qa~hR@5@cE!u+MRsVxfQ-*RUr0t)L<(x6K*_PQR%1+jmH8Sf6kv3A zn-G9u=3NLX_Fm>pbIFy&pQPy)QxZQKL~ZR@jrP-=RfvbJqg8l zqP#TU?eL6}as@;s?q@jZ9X`Vt*XyvdME-{tO0fQBtSz=AeC%|L$24+hFtml3b@u0Y z67H9={!gcoBrDe+N1To;#UcW9!|xu%d5@^!GRy!%z}14UmvymB78O(EXWh|mSxq%t z+=bo`lpR8vZ<}Ym1#J`7JsMmqFq00vJT;pS&@rCFu&u|YZ4A3;q_^j}g3nZ2rO{)2o~7ZpEF(=I1&OV)xI*ljxoG1<%V51*cfX)NS7Z#g;{o~S0a5I3YShaWxu>>90DRVs7ztkY z$f($L_Y%>=y0BoWtE-pOwo8KOcs`!wX)CI}Uy;5&yrtEzQ-txBQLPna3`h8QUa5Hx zVR`)7T%mcjE1YFxvM%@5;AA$YIgg@RmvJD81nTYmUSeBy$d&cwlO;2s6o}6&1MOF4~DGlgs3FKtWYs@$q&-&Q!*m1WSHTnz!6~j{dg5` zXMgCh-tREozPfTi@sIMgoU`;>EFQ#VF2DP8eht=SDkPTlrHnUDETW3-2c(N$XQ*Mg zll<_RtMnn6?p=i=cl@_j!3{<)QArp=Lc?t6AZBx?u}?RAXF&5P-J8=}BB-LEl3ddE zDn#Waq*=Cwx<%4}|No5SbszpwtViJFRpn&gVUB~cxni6Gg&3CNU+I%?wFRnv0*d5{ zLNK9kig9K2XYEBaJF3FG)m;-yg)SYc*8=+{cfH-%vd?7RaPo&A{QUK@-F8E;jCf8} zh|Gs2n*q!5uhV(p%Ev^tTK@9LPvFP9Wj#Vg{VHDMPExVTYJs}npgpyFbcHUB6u z!AQFzZ};&Y+fy!ZbYCgoJfA5<0;WjPQ&7I2^T|;^aQQ4&PMfWnQ`kBt$lV!Ur#JO= zCrz|BRMMUPQ<6~WG%sD=lGm{d*)zC(ns9<%xR$BONGGK0cA48izq)FJ-f<3 z^CNYByD=_lF|Ue}q)b;W@)=;VcV)}lqj?E9RP^`Q(MbFiA z=7)V8N@ZBgiC-2;ftWL3bs0?^mZ_46l7*&L1OI+>Qo%^+djqe2Bhi*B=v88tpUef<#TMNVm~M_VF8ly%*8QR;-t?x?x~ls1EK?G*QI zk|vZ+v6c11M4EPER#A6QfBziM1Q3^d%~d|tUQW@@8~rJxjl5K8CF@tRLPOGKv!LrC zt9wDEf0c4YS|m~ZDV^2>J+2KtxRdktv6IaJOAlr!PO|RU3a=UIfm-*e%SwsJ`-#sS zx`jY(2YuctdTFc?qf8gA#m=;z%v7C3D6?mWGqEi*IC%QCg2tg_!Y|9WfA*Fh`!-sjlH5umrFxV-|1_tSN- zYwgXfZ=`ANE!b#S!7;FS620XbpR^U|F^p`sN`C0|3iWF8VS+0Y>dp9iSlsHt@y|#; zhS|I#S@K*v{gZ9{3P=@_{&^Z6Q2#n=A2`Y3f%p^6TpYutLe=*53_WirgE=h#3~%|} z2@GX_{x~xY`y*am4Ogk{d!zEw53-yeH;LeQ+wN>})ipK9^B<-DvNy7&UnXDj7=0b; zg8fDua{TozXkjliaT{&ZcnnQVMtZ`c!IVMi05goIRnH_Lihh4|b*G_oWm6K_0;$|I z7|VHr2+l)_bSju?AFmZe zt0`X`a_v~|?D~ookT8GJWX9cO7*X}o>C->-Bq8}tKYM*q}ld=LR^JDJnDmXz%1!e4T~ZWryLtyXmLEk^nOs zPCIq{ytDjAqV<{Pj+eg{^tfuNKV)d`U%hazob~W8ml;)tsn&$Rz8fhX%qeXKmD|wR zTearEOo2Wh6+8nwWY&mqy@*Gb1cPq8tn^{`CuyDa_ZT(!;oV}Bjo^mN@au;!ED_|5 zD|wK!_%-og(N6-2aTu67&uVQ@yn^n7IoJ|+Ie(m8F7SUCU+xLjAB3(1{+_XI_xg28 z#=GWxJ>ZotuU}gee9qq~kTP!Z1!D;7Yg3r6O61jSOp<}vo!aGFHQ46=~-yJtmYc3u6BFn45|ShB{pnsw@E=nOz{H0VpTCH zZOW47&*bb#R$xjAG@?6M;udOc(AQ?SoYC;trHt^j+miYe{NE8$CmE)S|JwKeJ#t8? z(4DJ9mN$G~cK302oSgkNi&8=Ybn+@`7guJnLx>XQb&9tR@+ag@pF+JYTEW%v;kJ8a zk!@Nh(FINeNL+oT9t_Cv_aaAL8_9tWihB>d1KuA}z}89DsD_mq<{BDkL~N3(0Vwg# z{_|d2$RuZ2LTT=%#7Eh9Qd3!fQZ75Q>W-hUaRHYYIIek;C4S~ZC4zZ5nBjxJWa$g% z*1b_YbmR7?c7E&{d6TE*KRun1FA(;YK3P=m1}ih9yes)5i8L&J4Vm|*feGszrve6- zFW@Kg3gr|FHdP3Z(?8FkD#LQ#8#>Gc4aVIOeW`S2o+qzo-Br#Fqq*|04Bl@L%a{(j zd`3VS^Nuz$v>P-g(+>1lC0-84!Z~si?06UO6#ZUhspXf1|7c?Tvs83gV(Fso?5U|% z#yOV^1bJ7sB|9SPQk$YHpU&*`L*kPsXb8-|wsx89U%kRg21;r=aoDtu;#Iuk(rXVp z1e#pggk5NwjknDhy?Nq^u&CNMFAB7s!DvS=b%7x&pR%VG$*W%LEdYwd9(RzhE9Rt8 z@=FJd>+rf#|1gryI5_hs;qDWNDN!BRwENsFz=;Cd`73;Yo8qj+f(xXjRJ9iG&Gz|m<1?nK z_P+}Pf5nmd^}p=}iHDX!*Jo&bB5bjIK`&t#+&T7Djx(P*zT z+W?u4qX0eP6;3&ibzBgwKguSVI?z3);(QX~8)#7I6UVDwG(OpJbFqU50#VU*(&qoH zWp?D6f`wbQH7@*E1-i9*{7$(w1w&_U899@GtOn{+IHKMBo}ERc<R_HaCCpCi*;cAm2*jmz?V6}PfLP&9IhV`6 zw?|t}T=Vplwfz$(eytANvm$<&A42ZjBGiX7TBYKE zad-=L_xyoqVgP|7Yi@NRZ+(r`xLxeh9yW!V%x88C{xWw0nwDKB(%EVVRIw6Ac+lTf z`0W>xjmzR;L4BM$+HxBU6A$Gl2|SwmsP1>R6(7G8*^(>rzc*?r;1;gPusRBA&r-7! z9(&CF7XH3!z&V6^o_tAK;V0Gd8`&UO$w6S?+?~ir&Ygc^`5ck2dny4i4Jn$X)r`kI^JE}>Hi$2q;R_;pR3uD7MiE853T0pum8qLW=B3mJ6* zUD-A*FKS3n>j$R~PDCA6Um~EdKjdGkddG7F#k^sfeWr)%sC-p)%(fByHJFpP&7RlL zE5Z+$9*}$^WE_6Gf%>yKj@aYoFLqu`C}(mlwDj%cQ+&7d$9J4vt7y{e3V3=P$$Mm^ z-7?I;s}`6z+We?yrD6XtB0cB=?zV=SiVl^a@@?eQO?zRj#S>RVc*GWvP+IIyBi}?V zsZnVQS-CAx-CTS-qvPQpu*GP>CPrA~5eufxff%h`7O_~)?oi_~(rz*2`l(sm+uv^- zj5W9LWyyy<2YXBE=d7d&78WN~u#WjMgzvRChdIV@aKK-Cs%-T`d>%G}s*OonZeWF8r1VOffQJ9f%> z&z^*M4HDnU^5)*~BWmkBksMuZ6%onja6Eg)4y3)g33eu|X}jyPohJfM?|e_im~z12 zYSOIIlMkU<79W7&C?c2*sb=Nrrwu6N+p? zgb6q(P^>P-G#+-MX+#8eKY)dP!-RW)q z6eBoZOBXfRoC2RSFVkOr^`4{oI^a(>kg!7?P@OVZrKmrSUV7Q!#=vegz0vHz8zh+Ke(1}ifk@A|YD!(j!6m7@KHbz;JwoXh zeFcuvue{8S-?KK$vGS(v*+?C4UzEUeJnOvgPs{RRE4~9JuXJ9zepbpeC~-!39LzW7 zzzpTYrKL0%_$I$P%{;04fZN9(GvJUy3ynml+NCcIPyCaOy;Hd!qz=-}8!|sWDL2b67wX+Y`&1uV z(82ip8bRsYnRJ2o&-Mpsv}D8ldq0#;@ayN%Im97lV^{4&lOWT4ld1b?=k4|mUTWHs zRWE}z(N}{q2s?8(X>0MouTQT&$!wnCjKi719dcK=ID@7ZyjSt(j{!<4Ln6=o#!Gn~ ziPe1JnjRLRZy=A0a&c{puOj~j@uZSDEh8Bnm|5F1VXNsO?hUDn#S`&7EK_D~ECTZO zh%>txZ1mIVaKK53;U~mSO{MafFJrxjt&XeIpCjD$`_{dvmt!t+O{uoqM}SQ44c>dk z8wRg629E)9rM%3;Q_%Yx>abs}=_pIe!aVD(}3R<6U+_mR!jmxKt$RLF3Ns%z{w< zHl@xlc=VPf%LlE(9e!P!kKdfjcuIZ)-LjtrWW<%g&9@Fpt^z9h_n70RpeZpN2VYra zqr-TJ+#T3qf#f9LxtD(wnsRqGXV6~?z(xhsnk^-GaJ>&BPf5fXP`p)R6Om40JP9Mb zuMTs|Sc6PnDnRyr@?KEKcQT}vSDRso&kar3Oex`)K4}A+2i_e3$Gp;x5HojDN27ZkNV5$yf-4!%PHJ6EcpsQFW`WT{0o__p`ZUh*8~PQ8k-0c z86!Tfc7=Lgh1hf3)0@YDgC<8(>L>FvXVS;rLQnCh+H7n?p1t)RX^bvWmgBVrzUqQFOr0-#!n^|6CHE=9Is`B!8iHn%hI2pR8DoqOIxpZUF zSjWmV-j0t+>L6px0?&OmJ@qucsFyfSHaeZkzhT6!?&HV+(kJ@5g~~2_0QQ>P2(NGX ztvEuKJ#LmM-EdLEms;%kexp6jqZ>!s8^wB4RSa@ULmpqCVs(Adi+uBc>E0Mv;&jYCbu3j5O6s+5#Irie;#gfhHy_Yx zunBhF!bRmF^L2vP_(Ls~P07zTH_ytbSN9$2H`cBoE5W_lW-+{NB7ye>9+76{_EG$+ zErv;!8!S7No`gIKePx`l}~I z)xCouQU6hEnSaz8!N2f0VD)w%ePt|vm#s0O_iz|X{FjDlcY``cs3F*v)}EqqYHeg- zMU-sR#q7p^XL48xM`lI5k^5H0!uf+5^QhCI9$Rs7*`aCfAwOY`JY}65?u9X+acQ9< zA{zasW~Lxn+i*?nbE1nFCp?@uw9HrjL25#~gJPgQ5X%e!qJJHY zxW`M_0_BrH#=@1ZObf0pGu-|(RkgLAj}@2bA+zN+O5q$mAhEsYda-(4AzR3p()_3U zrjASUdCm@;Lou}}Zl;-9?DfWdDobz!-@Uq0c4+)_Q<{MpQ6`^L(<5JRgZ`3=l+SgR zmdaM39f9oVTTdoDLt@9#{l&Tdvb{ua zUc#ksRDq}P_Wtgt?s+?uVvkT*bUTnR)Yyl**fVjcqCfo<*|2v~l&!X+-Pin?Wu~~d zr(a0t&SD_s^`Te;<`4EzJrFTVn#%C$RSwiU=K5ViWxahhtt9)Th!N#{k`8+--dcMe z+boQ7P$@gBxIJ~?SC$7&_Lh2Sa}aJF%B z%G(7*Ic>i7tLHRm`o`eVJ}p3l)l~T1`iB}3)Z~H(5t!;5h~7%+Pv;liZe3eNB2`FY z)0hl^O3H8FmQhhE{=8i5a&P|)33+kmTv3zI&v^)OXN{#(`le~~1!sJJKR-~1 znK5C!(T_3Q*WoiJcLxh3L=V9*dbXs`PFNwKVj1<6s?IJOl3NAeAovCZcdv6jFCPPG^_#}9Y(sn`~$VW5Pta|WyjlJ%l zU5q@=%4zMd$li2#ZoWanAr=!uF@%l#wt@{U!dYM1{7O(jX-<6~Lkv7DrFVpQI)Tkh zWgXg0p}57w6}m?3Ms8b)R#h&T=a3Yxeodb~;Jb5pvjH}vHeTN(c#{5KI{!l8IgSFu zI<%{|{if$I23r{^uXk*-f3 zWb=%uJmWfM;G0?=iSUqm|5`5xcPvxj?<%X0g}DD(N}@lVhBBT)5O~AEY{R4ISs_6tw$@e^9n&)MDvOya zb+fSlYn5N(u?$uy z@o`U@&V*598$VGaB-r&a=vFVtv2KYPu<~5+_NRX+Ld7!P$!67s(#KDrD6goNgE_Oc z`>K>%8pMK+rPB_@I6qX5RphFfZLbzH`gZq@ApQFl$M9k|Y*_j_F%;7g*~X?d><=ne*Vg8ed^feP*=+z@r)PV%V+Kns^K2!%h-=JsaSk(Yqj6GT2d7_URUf=va#CXCuek70zrs8t83R?F3 zfO%|JVP7$hOz5`WOVNZkf6o`mlwm~cieez)!r;9rck%xbjjg=otAn?NSpF4c2L)YQ zIGYa%XlnVYyH8F7{jJSS^K%S=u8vV&iaqvJf;WDM(NnZPV60V_QFz#Tl-!Q8{MB?+ zp1k7v1dR|c8C736S00VsNI$3Ge^q1rAy4P+*oOr~5F9$z4DNqUVN#CXy5wuGBmJ4( z6l(t%`g&@(W`?Xg-9g1QU0)e=YO06yrj+i8AFoWE(iLc^J&HiC2sEsB^qoJEvkvO0 zX}fi6l-Lr1!Tc|a=d}ACd>d-NWgMk9U$uHR*4rjJ zU)!1pq6v90HSwNhGrp&*XP8gb+iLNVnfB}yXRb_UmCUghENr*Q? zJmjy8@W&9GZT^Efh0di*t|Eyyv5Dz}ymvew8{UPV&S@JPHoi4~b!@AOTTIwF<2I3) zZ>&l7l5DAP!*cQyHaz+y?6VnNi&J!lx9H$MSj*5~UI9&o}(>hbT@P;~w5$D1GXz1y$*SHO-Zp00_A2)Sl z!Ov|e4?wZ)46=l-^Li5(vO9hV4mk&{C>KE5bI{krq?1U)Qt2x7DhD7a&!yg+@Q9s# zsOx)MADCGsEsEgPfr7-%=H<({w*~+DhDT%|LIneAAsa7Mnu$TigKj?I-q;o=M_vg? zf3n~+`|7na;!vAT$KIPC-GlWu1^I-2!xU8c59H6cB8N@&6rf43|StK*f1>N@P$0#s086#c*tF7JzFK-R6~53<7j^ZOk8}%0~50xI4j`{nw%58eBu} z%RF>#>RCVeXIc7N#W$7`I#Ex&ZyP%=iZ&XpJUAcXuuzf}c$CUM(|NBGkviFk?3GgL z-nXV<6^O!Q99+bv6v@9iB%dok7UFX(o2mhSC*QKALSHAIvkM*JN;tkWi-no+&F^gH zUZMOh_Hi;>kh%M;i#e3 z$nu>RkV=sKfkp3)sl)J*VF}>@h<@}layh=BEnM@cer9KmodL8tc>e2xR;7X?e{}1c zwN$CFmdVZLt4t(o?s!n}pDN$qBhPedS8i8+Mx`bcVQL~;k3s}LGdal6^0{zN&wQh5 zt!#XKBs%NtD85LHF584?y+kqKvSd{xh3C(7gO zCGHa>AJV2;0JeQW8J8?S_`)&}`Eurb^?JcwbmzjC+sEa5XTGHW?3oqLQ{%9ZF0Gxb zWNL_*4!}5~9FKfA0J0=ig~Vf-9-cDvn59IxsRe$zR%>VX9;6v&*s}ed%m)qrjcPmH z-Wm1$XCxu;q$LQojrK6-POj3qdQeoP&@c8uR&xh>>?1lyN{dSyTHj3+A6^e`pcM=U zAT&bfSWnU`ma&F|D>iZ;LVOP9MEH*ly))~)lAl$G5_>uSqa6Pc)fJ-b=3l?KkB*%e zZx<;FqijgEMG8FHQ*#ANL%mpHTgGiW+5!Bh#^-=}4@tBKHUbSm5Kqq}jJWk2AC>Fs z{QmwhrSikI>_%^-sjy~j9ViaON+3ib3;xPqI!v8S@MH?I3Rl1FqekX5nm1L&B z7ey?}>v(w)HCPn+qb^T26VU^*h{wTGouORrp;WG(;yc58FE=Jqkl0xCuW#h%CyZ?;*j8 zvStq6EX95IXZ0wl)7WBxiqlTX56r6(fR!{pa(IX4SZgJvXoVP>f|zgVk?jC7@`4dv zujqexxAK{%h+AZ)2I`8mLoykh=r)K%f#rGRkOCjr|`y=rLSQbcZiT)^51{q-2 zkWKa_I#rtw2hBf;+hGf&I}S(CTV5u+kyKWLC3vK=X<7E$CI51Tf83J56M?!c`6Gf{ zy}PYJ-1{w`inRPxM)xNR`(r)(LtYH7}Vd3`)Z0rL7nvg+F!ItyVWSIq|e zW!)!UOr)%5@>RMTyGiP1D7N0ZK4+#`0ntB?+21(ezf7}*g1Zjo%N=US14vnn%L~o(yOmDg5~S-vp|?OtE8>eWgh{c&Uq^896Bk@JnMHHQI1GhIiq--gU6FFvq&nC9R_3mO`r}aPjI*9B-&Lz7+Vb0?l>7C@}N z=*Xol^K@Sjtfbr`wF2&pGI!p@cm03a2_N)A4#1JEFkV|ke!}_0lgHibcp#Kl-9^l+ zIT+_TaYM|D{mH1Nzo1A9ebP#ol=pkiGcETA!FMI-;c1$@nu(=OKWCgIyL_~NSJ?p_ zU3x!$*Ud805koH*(k(ZkpKuU3SbU<7%s(H(i*hh6uT-?G`kZC$zZVa(VTt-|a)B=$ zw~5W$?`i%J)T?@{F$fR%Q#EdMaQEk+6P{He&e_~O7Co9dKvhBg#HGToreH!>sr>v4 z#8m(Uavq?8PU+vPv8WP%1+&+Zj5VQsfPHeQrh-Nf7$G!E=w+4mU+ z|9lg>We$%XCgtj{!O@O16*UqvJ6FrWG#ECe{=&Zg!p2S6a%jp*qI2smha)k*=6v#Z+EB zJ88#yzjO&33`laMpJcCogL!8QsLjAPGkWhwPl*?&fovI_0XLY_QO>a}W`NW8pr2k= zsAukdc4l!Bg(x0ffxJO`+v?kt(bQUqpy&_PfOYiCcs|oQ3$nzEa zfvw4DDA}T42J_T#wGXpDsedkC?uNW{5zFKMu<^Sru2?^XiQ^XH#UlJoBgEK-9RmhI zS?KyH30Ny6XP(YPAH&`5nl?@puqwZ;Ti%v#Gn@Z&=9=2jrg^Y$zo@@|OusIJiKBNaoB8|Ec`(beQ};Q=t;>AU~B+OOhq zmWumj4#zCWO-_$_oc(_`2(Hg4e^Xj23~(9z6UTS{PRDuXL-fGd@``x1k@M_!ea7n{ z*8Z-I>k&Y^iu6{FntA!J+;3KuJwmbCc_mZlQST zUSrukk6S0We9C%!-%2$n)w;%V8$^i!YC+7o_1eSroJEr#GR_Djb%RIy5SIZ>H>N7t zJLRri3`^Tvm)4VQzfI!Cz6`hyj6Yklq+}WSFnl9`E*HKztlkRASD@di>(c^&e zW=ynEg(`g){usnV{>D?GZWvcpteQh0of*p`oDolBPO9hAQ=#|!cu_%8RUzDkC2qfQ zCC)K36H>JzPafo*|MO%hGn_*B8fkz-LB z5|AFLSm{TDGcVmkmx>}~?w+?zob?Jg9=xg7z)q$D)5{%(nobJ~3{PP;;;j6iORAey zv&Z8sZN$1I%&fu+^OW!gsh`1?8B0IFr0>EvDin#bc7; z`8fy0?yX%zYjFq&cgdIKCwkw1<~tndfDC&Eu&SMIon=gsbZ6qQasA>AVrZi?PNWz5xTpRPQS;oDEDNd?GpF9kNWI?3@}eQ`=fDFEz8#>?&Z(oTcl%ZL zB25>}Y{I=WSr&g1QSwm!mVSm-1Wao6Jq}CrH67*#^d-LN&o-4y6&*1mZz;3 zMMad}n+OuBKmer#P?RPh9i?|LAT>k<5mZE!-VqQYO?og?DS}81Rq0I-2;?eF5<-Lk z`JF_)`hEE$Pfng>&+N{e*>`8>o$b69tJP2}sz?77Uebn0C}f~6nhqyk@}0 z$SOs`JSfVI;u+n`f_b-2cCdxxy6bPQTTfxqQX}1g9fzID`9&U_(hNBuLq^l%JBlx0 z!w8{wFXXlU>vVb);hbAu+xF*p^bxwMjwKXlUd4 zSXZl_%4Mx@rKPqCk{2`v__sY0wOe zw*&O~Y1Kc5>wOs4Pi!{pw~&48P+YGBYGD3~Ykp?;wie1V}_ySB*8%!%v9TG(z|?9-wKRCNXmtP`w* z?<4hP>0R(94BcAATr+YJzHl2nkgh9HYOZj)9B28Q{XOV!-=tehTmk+E1cGEm@Ep*v-G$N(OJ3b&Tc z*8Uj#h+r0?s!=FBb1*QMPCYaqGo@ zb4KiyQ4{?z#{tCe<}7rON*gC{(mraT{LcMGqSoVWMEnN3vrI&AfvL6YV${~o8Kc`W zt}M;T$6(5y^{xKlH8X>;$PMwd{hP|-KdDtzol!ukqvDzXd0YJs;Mv>jUu_fkJJ~Ic zk1uuL8@c}#4fWnM6ru&k9_=d-6zjiX9-DQQCK){QG4qvNs^Jk8hj!+$s%QVSi>kG8 z>Ls+?tM8I`{UNz|CJjBUez4B5@;J%;^v}l*H{?FoJbEDTUZ6OgVSX)5Y{^z3M z@KILugCzskK6L6Vm{+U^QFO$NC=;!6>IzFq1bg>R!MK_`FLeWLkmp|m#+D7euZLyM zvbREHPX*akH!1jA@$*nHQUQ=Get0tn5!AqvzPBR~(hDpQBSN+A9=>w0z|hQfa3#eT z`gMA2gt2s_ziL(|36ZkoHFhi{zI@l6zPBV|Iy`k+<}r=olNmN&|1{O;u=QH8A7~Zz z6y)5e59=RoL(YHRm3wM4?RRK`^!9X{J|?RcPs>-$dZ=77tW;`8)#r$R%S196#ChnZ zE&I@!fd%2gd;D|`4068PBsq1o%YD4L{$g9LBFb)Ce^+L(HAszryS2WP_sF%0W6#|S z1J@jMqQLnd^1X+F6(&i5OTM*w$|c7nJ2+99eq0bMoiq+5d9IWlK3jf!sr`xT=i(Yr zNy81XMQ#zF_Eu>iLo_@fQO&|WkjrqXVMlpMifF&U;o<7SrvoI|)@x5~;?~&CfvbU2 z($3a~IXJmu1B~@T=+Uf;Y|gxkpW>gOvhuf7bA^=3CLHXmx`*sC=bIo~8%{US?vA+~ z)~c&4*rh45l9+538c#z}=+9MW{ACx$tMroSk3yY-vd_?WBtZQm&(3E0H&)OV1}ruu zLo+;RI0Ri@D&|a^JawI@_5X6us`svM#_h16gZtJO@8Sv`HDLRV!t>(SK0jKv;+1)H zud8jqv&h*Z9oCvgOJ?cAGxo1Dj@!zWk#t2yqY9?*5Cz)LXgL7iVxBv8@YKe?=w-2b z^M;94%1SJ|@*!RJSUa0|sa;zxL6TpZj-3$KY8kqw@XvD%t3M(-~OsS`UjN&5oi z#SKMlk55cdbROMRH@X`l(5P=yt2y=qxwn|as&FqaEbLp)sn);_Z9GjKWEy-qwuuYQ zRc0UQCEziArJK3x2ivF>t@YD@&oPY7i^OoM_>D@87>IX3Vh7C{ zyvvYgCmbQe*-o4t58rL2_?}F(-n5gFIW4@LSIkCUMXaF;H1~y@aOu+q*;?Jh3n}As z#bXaPG9CYE7Cm6-saehHB)*T#%rJ}vRnK0IX1X9Y%&A9@wTgdPOjxdjC?=sSQm3!` z>+rKLJgan_l>aL{WV}SbQx+4wWPu)T8(_3d>vf_EQ zt7noHO7myJ(M*OmXoUNqJZ z(MM@^<6~NeoEcP%*RLs0&NnHgKQmmNgd!fGyd4y?dt`E)f@ei!W}QRQQMMv>$aXkm z1>Y5XIdhOaC!vrt(kyx}+9!ApSf2zs?M0B$g$SKpGM*u{;zzkcYHpAF(IsS#286qJ zg3c*!M<1Cr5E7wy+_+kK zPBhVxEb8VG5ohykTi?!hElEVP=9@SQS|#$?qQn`NP+p`bd*B^Q2w5gO-*%8115OfD zd0pf>s8^1uNwt~;m{JKhF}oLQ%}&w z?F6&Nti24cy~Icg%v&JEy7gb~qn=O$i8wA0?bG;IV#4m=mmr}^HOX6Cj|=EDXlmTd zM2IY>g_E$U36bN>TySSTA0yNU+46nQ4nIpr>k^$l>9si(q}c~GvBWcDm;$+37M$d! zAiq1^Bde0FaWB1fYVW{&8AndSblcP%Rqc!8~=Q46!`*s?v)f410m4 z?dltc&U1hC%?oc~bWxWdTVMK2Y)tfJnyhsJGDB^}yZQskg@bpnb?I!lVE5-`A$0-~0cYwawMAsZ4|lsZ?S8cnnv2^3wU9D_xR2WQ zB+6MY2)kwE#+?_KL;?^p8m7(WqiAj4A-v@Z>NrM@F3+RSHagst%%@{|nO8HZ;qgRd z_-*qZ)DO=-&K`-)qgk23#b5H7(`Q`OaK&}maHbh*MR^na-vgF*24C-Wxi7Zis#J=7 z(IpnmGOugNrdg+tMF>NsnQm@z zEe`wGD#ZY-=fmZ0#=orvu6nqudR zPMo^|$m}eK$0Fm3m3U*nfGJO0r;RWLH?74#@6(wTkSd$l=e)S_C?DOl-7kK^m5Ec!@L67A4&fDL!| z*me~_x8eqTh)P@B!3@u6gF2Z5z~8-2q;SG0=AKps^WwgPJDd6Y3AU9E6C|iR>_rb_ z9vnPWOq!McuM`tS)v?sL66Mo^jjcfD56$n67dP5(HcB{->oZBx&dFQaAjja2Ee5s= zWOntpJCEvHII}DGZ8v&LOO3K^;JhQ9g;rA0EXR?}5Cp`)%8@7i+?tZ9RmD6xXy|>- zl=?8NaQ|G)?Ijx(D)M6Br%2sTTh*yYEF2BwZOuq$WgC3QrLSFQ9l127X7=;s6RHB^ zi6?(0wZ?-60vhx`h?m$H^XWkr@czi5AW+Zj?X05Uw3-R&i|=!dEn-u;98>Rg6dLgx z^I^#d4Sh_$By!OG_yPRK>5QD*0Ddw#TYQf zb$RIry9RK|2F1!1k^p?XeJl4_Cm~NZ?xMSBL5e>brMA7>c^+KbaEs;NPR95jALJI&hKNo+I{Fl-o)hYgf(k}I>mRH3(B?x--7xUT&u zJYzf%$3!-bj3oI?TNT92n##76-~1V~1L~L<{1oBmFdTOXk<6AwtR0R!d5GZs;*^-5s+k5F z`ftD~-bBPqIO+cLg1sz_@9&Ffz^w1zarP!6QoQB0zytZSCScPniiRL_xBTL*nVvfb zEs_kW8tcX%Vf$Sd;-`=qMhW`-@~QvaW7RtcDdEq?2BqG?f(vlTxOy+kzdT<(>+e)> z+>!65{@W~g>8I0Ep=FxfUH_U>hZ%Dk$o%z}4kYvsQs}Pz70TGo(RZ9GlaYbZSyyd* zIxf}Xl9u8^zQyGD{}<#0s8#a6oCW_`@~je(W3Yh{D6K;2z9q*a2%p+k*-7dQiGu*m zf0?p>E^owydaw;KB@I3$4j zPE=8WxA?|G+V#ILs9#g&8O9qDiN?B;J@o==Wyyt1-4zY7_z9;+``1k1ETMrLcDFK21z2UKoyRAS3Y@8Si>mHV} z{Il2IS=$Rf{wBH&HwPohlEXi0l#^?DIJ4j86;c(F@c#(Cr;CSYla$AOh|BQn0k3z1 z{wvq|2cf3p#4-1DUQ$29x3x499&(eSH3$_ zlD2!_r%e4ghkY)#zfXywnIHW3lHtFex`^f<}8x`&&BaLC4=(6 zH`TwgRQc~OR-XSqe^8*9F+&Po_w_Zqy?sL}W<$Q9lt+kDI5=bGISA5GotrA0KrpxE z+ZbRnfY2aET4h2x`GwIXx4Y!)$@Ib#A&qj4lB6&Hw^uH-go6LegV0qfa3!$dL$*!;izU3d_q7F|^Ss_QKndMrIjhWVWy3C3_Eoe2B(B9ekb)?)r=+g@uKHEic&K}*K$Hqfzulr;8^pvqS%imjX-^1wcLwXz+?UJD z53@YQB81M|r$8HHF_X{lWjXw|zI-V!U*Pq7PspScfe3EHnZi~LDcz^LtiWX@;u9xf z+f&DL#(#_(NKB0t&ce_BV4+{K_+@Mwb~P3;HX#BZiM<(;*im~!;EAeiT)-U<-${>z z<(u<-q^`X_dKPd)RqHHS#nMk|FFM869gU%vGa*<~c&ZD!8SbTC{HE3MCOSw3SSj;qxfUd11+d2gACu*7z z4DRO_e6AgktRoIeJ#qx8-sN*=-S@3I0^B@N_G??O$BrbAcMGTG)s4L45>0asvrnv9 zMtllIqVs3Kld)(xo?YL~asWr-QoXh57Z3bJ{!&Gk<8qO@{yM`T`p>c$Z*o>i0_gA~(mp-|j&F#+ejv2DOtPq5EfITCjsaR|6 zV*WwQAK0*Igjo%?vK;o5rGu~{uwkT|FdQ|)-G$TjSitvPek52mo={^KE=Pl~(8cc4 z3S$k;QLxx8U-KrYj15>Hj}9w34{(*_f|J-NB{@&IYdd~9oS2YwDXqxmNZH6msuRg- zV*vdy9J!5DeNH}DaQ1_N0=M>A?w3eh=a{#eM)^5X+C}9%(dA{8E^EB&4={JijY7l^ z@;@^M&%F)2f4_NT?vSL4B4}$#E^KNRC2@;mL4Qg!Du_jLR0Tg_m_beWK|?FH`;ge{ zI_R($vSlh!*iDi+5gWkS#E>0H3`gqG!4XbP+BP?tDc5pkQFtbGZ9u?uqf#Q3Sp~7F z+&g6XZ*#vn2y0-$&v30;b=(SZ_zbdP>aB{gI{q7CvB~gMJJsLO;)$BqeD=0&3hOXN z?+1a4UzpfyNjuEWi4BRFt-y9pai83gq8T#*?Yhrc&{#Jgnya~ixdxa9#coJ_ravo# zu0`~TbvV6m5XmOSLNbjU_LC z_8|zu-_BenzPgn}O5RdRB zaJ>g5PnCY*U_>M$f9N>m5P0HFK@Gw3CbU^2V&+1wt*K&s$+GvC@N;lX zh2YW$3G}VZa=dnL46?U|x24bKCQU)`HWUma%qZvnTLwdAD7AUi{t9)Vhs>KtZ9M6^U+HIf**^mmZC8?M6MjYy) ztj(1anb>MWwLCNOw$8NwBM5-@^Lz^=rJTorhJ-6eTy{pthpC3P~()@(rQ9Ge>kQZ{n7Hkx16~rX>W`o_o zipqZV5-i3Pgnx+RA6!io)FXsBX@rD{1rgs^XGQMkrz|EcGb1|)-m@bb|Nr`b@C4JIG~%?yIz%Vku(9zJbo&auR|l*?-2uFMav8Zg#Tx#NcDWH@$tY zY;iLWl@;(_Rc6OnUfx$4)bvk1lfwwvjb-UoPq zK^Lg@Ja1nm;fUz~f}o6wnlj89ToE=zdf@D!m|>U*T`&jhtS+QgfWcviZEcX>AVTbu zVK}tK0eoWz-YDOOwzClMvXj&V>EJ20)5b3F8F7$Hj*uWJuhMQj()$EPx2G@ejwFKzC@q5O`k zOYQSytO7jXg@Ez9Mr95NQ%mN{dCY%ude8TgznMaQ*g);u|V8+fNI+TipV z;R)`Uj4Ro`17;6p*B%B=LuL8hdwpvlJx{UsAsC&T$GEhaGmhx_P0x4;@4gH`v0m?O zv_Al4?q)ST{_L-K*=z##l(=@0x8ZvU#@@q8z(b1?)ZdZlL4YODvT* zItotRk#;{v(R3)*+{cSM(~&15xXiB3&2yuq=rlrRW}B;V>ugv6)sehKZHP5Omc>}T(3x;Fj+16x0PP2-?ARnpf}WE$BQq846G!3%r%5E(iU5O>&Y_w z^T;v~7nDgy!YMO?Xr&@KU2p(vX3qT1MNEKeW$CpqL!oi8+IwmszbEu}jqJcK@?`2_ zaCuM@5}`-fH*EuUfzA|moP?j1vSFQulirxnJIW1W_R%ZmZyl5c-+R68c6=D2o%g5`Qgbczcr@E!&BB2sLy!3fRk< zThH5#iGya zjL)kpB%gh-buY_^S+ACQDV+;ClUr5$3%`v$taGrHErrW58}~Y4LmqYyJx zw>Mb=tp@czXc?`LRZ(rc9|H7PEd8+bJJ z8}U-ycZhB?3V%M!D|Zb4UH(~GJZ>MbSO=m4Tcg`7b@Q<%mS3<6$PXzcfyQdx=}GLk zt1cG_#R8{csd28Lsr1BXfwlTZ0}En`#EW5&A?U#?V8$gv{s z7%6K*!COhv)!0>(_fPn~4Jth9_`DrRFD=o=KGWB$A(c|->wi)% zMX&8wPKUQ|vLH>fyQ3%gS6=+G+&)hnb$+W1t#day^_mTAeIO5BML2!|0`l}le&|#@ zo+o=SC7l=@?pg3KqLwNkFkB85roG=aeM4YOsWT)E7La`##44BNcQ>&9R2w8cC|TE> zeWm!r7xrSTffRpK%AmFj;m92rpCq0QabTaNaeyfv$(_yN@4l_fJts2l!Yy`c*F`pP zNqIj4l6(3d^O%mOw@_VF%T2)4E@G)d(CGI9TMLznI=+swK;#`;g-zL=Mc-q?iolz> zdVp$TY*I1=tzWF9`4SdvR@4qhuH|3XU;wUI?w`EuH zC3nz`RstrLKy1|ZL(&%6u%eMQ8H5ePZbfI!G1kg@9Di$Ed;B`wDPe-;DZzdel*Jyr z8p>B~_CdngXw=9~-^i(nLV{IpX4UHKnw7e6Ue*mMvz;ZqBP7n++&an%lt<#Gf0>d1!T;cZI);C~o#S{QD+p-Wgc&*rrE^$R+8?nHdk zuN|$>o1;Z+{qt#VIRSifZ@(GJ0;G>*&bUz+Xqtj)_9h8>{u8QXcLSWpNbyV2erJBP zvo^BLv%pbtA)Nf9fCsW=vBNVwero+&Xz&&pN((&?VX;&DahlghFJ}Rc4NiS3d`$$1 zH>0A|VLSvnBt`UPK>Q9-pHXD^{pRw}&Zz~zGA~Gasmt~xpMMWq0&R-ZyNp)#Cc61D z*#n+@BRJt97!;0TFvq?oz7oN(J0$d(>BgS8)Wgx#D^jnDbP(IIXALy0UF1yh=i>rF$&!P+52Aer$U|2au77rvd@blEm63Hq~C zC`~x6Kdb;BHkb8kOtKFDCSxadR*s#5B0A*uU~}o!gbeto!$==73UQ7~ZiPX z((nLh4IvC63~Qj=#lMC@ucjqWg?{2v><9fI$Dxe~vcuE&c~DvXF4Ya3M(x(~*yCMw zJv{Z(%t8B$GoqIx_G>ZYnH^hsXXo_KFN?6o4vug=K0zudfTEK-Hm<-wMv&w@e5 zzj=gRR@&%ov0UCe!NSQkmI?y$EJpHR3vHj7T~rA-H@>)?_1)BpbM<7NowFTgLpUwA z$cg9TAr&2n#f}^Q2P_ff8G4=t7T}Wrls&2s*#N8iSrM>t^z%$nULd}@{2w;zrKmRF z`LtckDWJ7|(;}f|?=cUytWp4&hxlj|Pa@KX%7ECP?@jGeCo8rixM6uN)SO+UMxw^$ zWjFfc4}ZMG8@)Cf8x#G39=Sw&+A~X3A=VHk2Oq7jn)zIo0*6LwYNo&YnW?SS$ zHrtvvQO5yu2VFD78IrQ9`*L?K1QPAp3_VjegAqq`)U8oqXl7jA@3@?u6V#6W@%a2| zW&lWm${gRiPjO{IXe_pjo|H%!+^GBCJ2XPipU%G3e-yyFqtz3;4>6E`TzIzBBS2sy z2qyszML9kd6c`XXQ}%o*&vAWg>cm=rt*c!S#br=#yhoMvgtw2K`4H~{sLUAOX7BdT zwF*%?FUTn3Cclv;BkG3zhE=@-$S+%FmjvM|{vZF?R|S$@Yjxy@T_s?z|G*?xnAoxG z6N57E`bV)f22uwmA+jyj0za@7Nk@oPtUzb3+vN^|Tv)L*4wmZTWC6dVJUKUdc8(t2 zHpRH^m4=}2TvC#3%97u_wb?h@)EVBsKpm|j8)%+2Xr1M53z@SAa-VWOC33aPY@u*P zI*GuIg117k(D>%E*C)X@@%4kpDNrYZ>o~T}C(JiW+dS>%$c6+dcQsHGGaI)pVQD;> zblYVpUkjE%=FHrgxdY*~TZdWS4QoItF8ZVqDuahfKQD%t*c z`kYb1n5&y9@k;2>0R#upN8Zzm(CCq)#ps}DK+@utu=E#KR93J^R$YIAD#7U>(ZHlx zFk)K_^MrC)KZpwzgl6p>L2D5)N3pH%?rX+SQY1>~kPvYH#TvB>CUh1pe?TDiF81c^ zv?<9v)rk-5HTgmV#hRsw90P1e0NZ{)==V(s+d~#!2+O|kbz-sZ+0#|edvc#E1Xgdp zOcC_Gp)Z34lT&&ww*|BI5gI3=YXqlVoaI|sOxecIA6;KxPKBNXk{FMvz@&4(WW&_I z?7cDCjjkR+??1Dqk^Ki_c8^b`fzxTcNwCaVY+v6$SUaq3?~siuGc7Lsb4t^F6{8nhvsVp0)99DwSF+znD@PQ{hPnv-_i+6>12uQ? zANCwV#sG0KEu0Bz>(cIDbox(ftJx|KFqI0hsFk2m4dUq>v9Wn<$m_&k9p#}V|7?Oa8`P~NL? z;=m}Js%e5Qi_Okoo7%Mhj`F`l7D-n+aUIh$ulgHb7Y>6lTT{fL6~Z3Ks_XO&~O!^D#5vb;LH?!kF)E%_8C7S_cCb zVp|R6iBfDQLD`^`K}-vSkNdDjaAD+wa@bE8`Jpyos}LBrYE$0EZFZjk6ghxLVL@Zk z`PHnZA<4QyNiX~=hI)}V7U+3^PszC*uyp%o_GtE>XmPP);kWcv$OlaC>rzY*Dr5zKohrqDg>uS z{ew!2QrE;)?0uz6TLrRQ7=HQi7*Zy&kEy9IwX`U} zlN-K+gM1ucJP-^DBgj@CW_4K*9xxv86h|J|cp~}ymVTXr@Us8unOid+cWK3^dT@s_2o4oec zv%MDcnwy)4tJl`n_ybu-#r`!1t)N?TXj=SXtD^Q&OIP>%h%%2I645_ATv$^hvZ#K0 zWc|k7wz1U}_#kN(xyW%t0E@!j`LxTlOClTkwO;=u-<{8H&&Gm-husr3vEO{Rr^#1{ zn?=*B(Gg}^+t_HCnj*^#;j@p5xeRaL3oshTakE=HtPuYLrPL{ZrCqZ1RTJ386)@8) za*_lCnC6hy>tQ_Qub06`*^q=tr%yN^uyRN*<2CP`e8qhzjh2o2o_9$pAc5vCEwOSb ze__X?d5{j&U_3uzn(1a+1Ej}_HEl2cY>=?YW_5OU_Kt{(in5ffb`lHx3L~G5$KR`8 zC1dqQg;cz)23+TlC9;7L)4kTG=qSZ{h7 z8_zq^{m;;F1)~WGw5J4`H~6%Z+#K8#%z-wBAItCvfKn6c^sAbx)3ToN28s{!Mo z?YdvHRXZ<86{$fxIn#|h5m3-fvk!>Q%wGsN9m~UC0~i#c+}lUeUc1JNftOC3o16Uu z1Ix1xl{T|AcFbiU>yfWJj{Uqfh_H#>OP&23$9osf6&fXF)szr zH^husNpp6qDk0&I(q9D6HRZd4)kfR5(MkzL?HH~2bv{Gn|%|-az zOlV@Z<#zek{LhUQLz}uJ0O1;y@!&I)cE(&@md)1`>s;2Jd|3%o03TfmsO`SaUhR!r z{qcg_34;N?A6zco}~iw3D2ZgABzd#*+B;@r6O7huwIiru0rt zuW(OK57r~2+7Ekyr%hODnPPDR$c$3WYkykTS5i|esCaxa(KS9kPHQ6ta{B^U zxKcAG=fhVmLu8LzvRcSEO;%&NvW9tv8bSQ$AKhGj5}13GJ^c+Z#^Rv_w;mdO*XwgM=-uXPFfwZ(g-3_}iMMMy zo8s+=e4JblKPKlO;$s*O;X2CpYdpT$d4Fwti(6=7Xu-9MG;Xgifo*Bq?h&{*;BK<1tRC_LNGVcJlutI97Hw)o#(fU=s5lYb2W1n4zlsp;vEdAB|-ef>yC z{LN+C4Lgr=n;0;_=w)FsP{sY!U39i`UnEWfgIk&3u~?$<8M%|sVQweU8X&-vYa&-81*Ntipl+^>Y2 zzw{xy?s6sY&Mu!_z`!Wc(U5qjHPL6)U#n-ficEAJ=XD3-V?Kc|Iq0dai?uS6ku_S9 zk1~=pEHoIyUyL$9laY#hkM*@x8r7eN3W>0%CROv6IBY4aYkcrZ1}4Xm+x=QV#AWeG zMboa7?B&deSVhxL?VJG|OxkvRT$#U{3`8;?AIVHHrhu#*758$w?vV{P}cxDFCdz)Y8oSWhxBQkYD+`cQ(Tv8^tQAidD z`aXv6%gmp z^{eq#Xk_;I7rl4;GoLm=Kbe-6~+T0gse(7Ha1ob zssb%3Q)WTg!0%fB6biTj{O)kb>^ez3Q%or6+W+wI(4g5{=;h0ow>bQ3ZOv_~yw3e} z!3)>*=Sw=b30-|HV`BsWY8ldAj|JK=!T{%Zax!gIP^{QS$pgS`X-RDke^aJ{RQQr5 z-5Y%tp;;c82D_TEUx_mo?AiXKU++oBrEtgyyu4^_M!fdiEw|`f1U!E9jidfw$UREQ z*bH}vKyHJNrpz&QahCgCph&-h0F7|S$Yd+B1*B|mJJveQ^#ZVj9N>b@Z_;coVb;Y- z7)7FJ`1GnQd+q}0GE={^p6LSq{UdLDYU~r?*XJe1Yq55%EZg9m%(;Fy$ut7d)h{qO7hnC+>zfP zE>1e5cD6RF)+>35nuRtK{3k;rDmK;{QiSpS>mWiJL3h_TRx5VnA3X?)xg2VeN$za$ zL|Yd<{MOPO0j!+zLEs)u7{K5EJmD;wJ^J;%YHdMd=aZ=!VYF7G_gP03ipy!2iBX01 zZp?1$rE?aTzv>hVzuO$|bt0F; z2zo90aJ_r?&T+9Bn#!jyHpgl=Q*Gma5NpQ4&W;;ArXd|ysG9D6IKTx0Bj}4XVPx(1 zBblGOxw&PcfynoBVqCP!#*2o@bH|B!nOgGjYf0_-+v1vE?|T!~2{i^?)Fk2%%2uzD z_wEuDCw(f2&M73{ETVtpa!t;&sMoPO_$C5?I4${q84?f@(xr~abs8=4!5A7e^%&!9= zosryIux!XIEz~xnRGcWeLBde*Q{%z6eiiix{5*+HHWR<(8Th~a^-x>DLvuV8#J(3+ zmwhNJE7RD?ghB_JeU8pH_1gvkrbODlyU;W+GQuS(3GIocJ38oNt26BTsmL~rOa7!C zbAIHr*K%s`%@QBW?_~QvJ9}717$F=u+Z_Kx<@MoA!l!zAw1DICMmZ|Z&FT}%1{Xp&rzkH!)CsI^~F_= zlSfAHdp&L;+B8Ir%rZ?HH@v9vGv^577V(bNvlV`9N-D}vPQ;aTgg4h!5kWf!KfPlY zfY8BzAtuU`=O$b1Llr5?nH@T{95$vjX0>CJ3mi8hxVem4Y`d)Bb7KyoW_`KpHuNRt z6apCUc`{=QnYAna_V(uktk%QS%KdNw{}4ksSSaAa(Mf`V){R%LHFEHUjP^X@2$N2bm7ZdU-EFsygAuwtPz z@t^?A_s2)KM2R^gov92*-WeTrJ44;E%c`lJwFwdVhTl13+UHh zw#8yC;KHPI=wNAOGh$hhSvepqe^GLClZ6ziHHE@7Ru%S0p})AijcU|13Cq9#XI7dG#0QlFQ1MJ8(OIxHiw^()`nLp+vV-V!s{t zv|+b8fL>_A?E6%mYT6^aW_NCueA!qNmpym{L0^1}sKr3x`aseCPYoI^1#!-Y^gd9Y zU$%upk4k-#c;xv0d{Qt4^KS64FvX3{U6yf)s(OoxaKbkk8Q;DQZj2zv3e$CI86@f( zPVjcyy}@d}+It z2uxg}I)Dm;d_m;viAlnN!jdR0jfs>L(FN_g^5%#OZgz;UPICW<3a0>HDnOP&E-O{N z96H{kcE59}DJezn7B_Kl(DShTULvPD2K}=!904wPrM<4GPD92(y|)g~AQA%=K-bs~ z16GS{5v5Rmcr)urRxC4-qC53{TCM~0YqBe}VDuS#b^gQ52%AD-?w3K;FXbCm+%0il zj3&=C)TM=P6C_bmtt@Z=)`+cqGX(IM&n303FO$A{Xbzoc)~4z!W_xk7bAA)*@IdW5 z#^?Mpe=7^Iai{~-SbJ=Uk3ybaIGoa7GSNdv%CkDv70U1EU*|{P~$D~`b?__6AS#@ke zX@jAO&%v>kh4j`lg|9MLpi)=s6u_Fp(0MMkFeCq498E>vXzF_JSnb> zpAtVC$pDoQttXaM%?RpAms0bo#W%oa|LGU7$R{>MvUe77cIx>st|xE2b2_- zl&FL}%>ZVh5XVTf2Ue1UJGiJ@9oh{~3Qwklg33!!MgyscdoI;FIc z)H8GA$YMKtBCmu#tBJW%?Q`CvJEK0!f$T0D%%x{#Au@!lEEmedZ{;Ve-VXs4it^^R z;hImwxnPrNDLRa4!Hg*@RIe1(>3iRYs*!Q`TR#t^4VT;)&e5*39v~oN%4A9y`-dXB z)WA{@Bu!e6mw{b?&37O)y=r08N|r^gMnD-{7kuN6!eas6q5 zV1U()^w9w(6u))QodV4!nkVzeHK1*lwF47MqGhACueE^=g{ctzkePN9aP64n>1nQ- z@V-bDl?3iI)(4yt_ARQ!MJL6V>Q?a+wglE6M&7SD7l@v%5ey#4xccco_E@uNiL}{~Gr!9jfkyEZ;#MZKxEh zu?^fjImT`?x=5tdxy@xBv9cFc5bU}k7ij!jI}sXqkMS#FwXdI`9;4@9bf6Ou+pkbf z?Mq?r=y1;4efrE#M57PRq0G`eZGUGMn;16c zv6UUAv~3yFkp)@$VDK&<8o(++W5AeluHqen3SIfy9{m)itXBAt)BT#_60!wyFwocT zJ;RiFiz@>PS^Mvw!1fhlMPk*-JNhz8RHrUzhm=IUOKk_b7d?>^E4aDo!Di4Ip0iS4 z^3}-9DNo(YmmFdMl?HZ8{kf@j@2B#$GCm$#MFSjOKGA3*e5n6=duZ>kF>HJ&$%q~f zfJJ;B!UVAoiRGTRi1M7xT?3tod;x1!Kj*)Bv$UD%ArdsDB^HN)e1<{L*SGsM>&H8e zaRzNPXKH+x7~-=}&?_2mT(;LHcpQYsP6WL;eK_Peu)dL(?+8B<8B-bNK}|@ENnX8r z)tEdU#&7F6FO)IQaqRv7k3>dBK(*MW!!rRkVld`>seb3VbC}{u@#pF zBgWe6%XXm6P+99lxsMBq#3f;Sxqh5NQ2MELD8J#U|C?&0o0qgXTs=+OcW&4m?Ed{}K+R zFCDKj=PWTDz(J0S<+9SVev^fz>F;B?B6V@hMBz7t)F?$I6Ot7CV~Vm;QqDr0(%@%j zeD5z$5EhBpy^?)t06-^cI@G0oD=w+{W{k|WChH$rDfr5s@o~?kBlFeAfhxpeVlYtP zP!Q@fzi*-Z^=M96WdZNeTYesVR8|vJJr$;>V^>9CSM~GJj8i4tqw8^?e}`59Czq#F zm#+}#m=zEu|2@rHReg%-qD+xW2M`)TMr6$77)i&mgLbponL65=8X3g)BNQ95A;SDf zwOtaN_D}6D7G}(%_WKt#c`R7)Y;#k|4AuT*O3F892T7X=b)qkM3_NQ6BfoxauxCEJ zKQi(_0Q8cG8)*E^g$rdsv$5xije(x0f>dFBTsbURpp$s=U3q*$?0Woi)>6yhs*B6Y zhZm=@UJxJ^zi9D(lDemdk_GvZ!d^Av(Gy{4+Y#4yP+8J)XZ~2O^!qO6cdny_b%KdI zH!-BHlCDkAD}liHuS_9kb?=y9{ZE&(tj2bEZ+%RG+ z6I{1qmqQA4=-r>B-+qKq`sxr&H~dIswxn0Gj-C4mXk|wko2wf@RnYUhtdgaVJ-h3T z$(R#|_tFBo({1BM@#z8Vb)H;}=0VTH5Rd5US?RHxBilMp++JCGbU%(hvmTevr8OX7 z-pJx3D?QQyNn^9Qwvakv@Ai`O&@l)yGBPuH=~h=1@SN9pIxuXh41WD$*El8$2&hEo z&@5T_#{+yr=6~iR#6_&khl}q&5dP@#fS?JFjc<|m(R~4|qtMgzWj3Xw;;4WZS&o(s0|HZRVPzMoNPjSiT%Bb)VE* zCCZi;RfLGkE_P6{P3(IQ%#{%Y|=7JR}}ZFShIZHYiN@1wy-feypohbnTX`!DYYEf)Xp(8>8yS?$LP|< z*T2~N9`gwK#6#{OUfojM$Own+N6J?~X!xe0tbz(3YN4jRSME)ObmxczX!e^CuU2jZuuO67sikuxxX`_VSU~04;<`- z1jAO+=!Bc|x7(3G9PHr&EW$C7W=rj&u2N+t1OTPx-%(lk&-}sKf-KM-Y)Pk z!ApAy)z6TE5F#Y80tSnhG|yNM2Ufg!I{hp8hSJj0ylnAWA|& z4Qj?Li_sDuy>yz_C29w9I!s%BpZj)!f6-q03vJ8iSw_4-p~W?hC2hbh`ELEq65COn z&#IE@f7GfA7T`lIzWou|eum5+VtmiLc3w}3e$2`0&pPli77KcE#6@b=jr!v)t|Kp1 znupEn_53B#>s^j=cjED>XP;Z`UtUzl#xr&qdO5HCtP^uT+5UOp91kqME&r`yLFV|R zcok+f`*zb;$W;vB-kxH&|*&`Q+JN&-X%r!~)4Of)$hk0`K%;l+Qm5553K*HCv=n6p{(X4fV9qY#Pk(S119-J+-Kf;X}twowB3#WANEV;bI^+jHIy5ihz|e^t(C~N3==!3Lgj?y{&z+*gn$j6itijy zr3bz0j3I4uB?E|kwfUfHe0ItRu-O1%?rZw`;L6NIT}S(?q5Ibwo0YG=nt!AS*2Fl| z71$^^V0S|~w*xL-DZTY0tLJA*$`Ele9lMLpbtK+CyW{Ck&rudyFmXYb$yH9N<3DQu zR&bEl%W=$)vs~VH9!k!wAvnF zKKq5aLTVT5c2?IF|5X*CP6v zrTD|d;d_4X&0^L?me!e5oKBQWd*4qeR$Z~5y(Jpg6-!lD3%PIgW|2|Cq$N^0${;z;~K?<1U>~hv>*Jw;{!)H^0#b-_}?0o#u`Q-Gy->8lx z!7GA?!az6j>%`U$$uF}siHc?EYiXdEF^W)VB`T7^t|a^G9}O+2xyL?!DwnBum%rw5)ILb8v3T`8>lQeGrMnA1MYL25$7I-F{C zff)#HfzP6O7^*fmnY_2gOrEsOa^>r}@ZjbOa%JBlqZQX01PFhzMz1Z%Wx9Gnj*d;@^9jQBnP^0FV=;u+WG$ z*B7Or_0#t|KWTgFPbOuUBW8o{Oz~%Qe@i}OfWe~e#v}f`6ZN~|H~pEU?kCo{8&~CJ zG<2m+D83-3CpfdDoh}nf(=(Uhd-e?=Ed?n#a$FlyL_{CqFtc8yLu+}StkRo+am-1( zlY_<1*sx2NnhH6kp4BHxy#@kiWa%DcOK+1Dfq#Ik@KN}bV4 zNwh(OkLIeemP=4$ECC@PvH%GI&blUJuNGB=kVFd;tebOmqGaeZS7zsu{;p$#*~Q0u zKSf4ImZ*JUCes^r~PkjE4Wa^haK6tA^1bonRo*ou6Ree!6T zx)~f7Og)_V@RIvJMa1@Ay4z8vi-X#m@Q0RWdrRqM5#%|c5wk_ramuAdT!dtcx&)%| zV?W-O9O%LK{9gC97cSqEkWN zBQhmEU>?cHD6aHNPB-NrTkHG|h$ZSCUXo3h8NDG4MEs;sE;d(Ii5s$u7O_vOk*Ynf zy^xxQ=dC5X;VK)c+h*;u1hV?g$I*kPPEeE)t@^#sUBN4Wq2d4KkT zC2-{sh;41dH@1QERV0rk+=JFN5w?0iUuh?a@KpW@jv$aT(=n=>+?m09r#pNKmzPW8 z5rK`pr|L{nf40KoAi=Szr?at2nQX3>5^Ae@G-w19ihHR}Bg| zUrW(Fr}^QO&W;1;Bp}Gb3cFMOFc_V_7Ddt=k2X5J9v@FmN@-uVf;6C_t|UFRvGXDC zh@(8dMfl!WH`Y3r&%qncom8-h1_F^q+t1}-V+A9Mhz$`+;lrz->6Zl6-5m5%5!o+n z?SH}k@Yu3SeVT)r4DF?tqv>ol05R>BQa`#%Q1UIm0GfzkyY|g5_pa0fO|vP&#w2VM z?)PaR(12({qVluyn}>D)e~Odc^FlsY%FFj>4#hfufKE-J3uEndIphGB77f=lw zOO`zP@e{xdO90k5BFHCGumK;)&b#@kTB)tI<&DW!zhw`^4k|NRS}#I7?q8BGO9KV_ zY2|aZNBy0DM7Z3@?D_JF=H&`2ikM4==FW_N);q7SzuQ-UvJr(ln7RPw&7rgEuR%c- zP0m=t6%!qGhYrcJl9N6K+2fOyyzQj+JA@2-!@KQgh~*7Xa-}J*`}c_MuXnywaY=s*XAGDnyiu&~pmq{3v05SwiJPj<}Ch z6p5|NAFm<>tvtxQ1dK-X$>z-HCIO4E1PZv_CNIKPbJAm>t!8Aa?otw;!MUZgq@}Rn z)F7*qVm?${L{XwTOKQ@Ld%;hXmJml`rz}_D#)&wfR#ZLyB%LYt3(b6R^p(DiTF zkbN1=6Wzq=0QgmTs9BB-u4hHG7sR)+M5bz;$NNUh;Plb>)YB90K~SII+cq~A%?E|n z7euYc-4Z~4(T8Zo_GbbzDRSerETdVKnXzEiV1}oRRh0EJlBeGC4Pzl}IIF?rAvYdK znOP^UpTq&4?%05)_CNmpcTEcmLop0F*}b7WT|RbSDRqWlr;BsuPs-vRPka=mB)gJQ zyMFNDQ_mos={>U}&gdv5>3_SpnA)pR$I2k-e#uki6q9)D9Ks4hPb@QRR#CW5o^=|O z1sg+jCcCeX5mn-k#r7v^s3^Yw4w-#^%F z)gAz;28W~;yEm3x+t_SzfUy1WSYS>uVp(vj^B#k!QGPZ+2|)iJjqK5O)+M;RglTyD zsQv5Hw<#04?Ve9K=KtFZUEPFGKCrw%V|wywBcZ%gg} za29b?T<>cOeN`sP0{M9~XHuGdPYskGzf8EG(EjdspPdbywUw-uPagTc+q({EP`}^? zsbZ61J4a9d5plI9(Y3*r?viE0#_)bdv;_}YIa!&dr=}qc;is#+=`jtS9UKbl1$|A* zw-GRuQioDpn$P}hE0gJL<-_v`Ym)ZjstLW0$dPmFaYD_D=pn+j&KkJCu>FmsiUP3p z`UAL>A>|ZXbvUpMC48_agW%nI0mtKgy-ntiRQx8c))ZcCuMsT*bdUu!*@u6T3=de6 zGpp|vRmm1(y(==qTf7iGuG?WA=vHcd)+3FA*$kzWcRirK8tP7dn3RZZo8E36-`mGWqkTBVD~7>{itE&oEeS_K1qd|n3UYS z0YPkP>Nf+p+?!<$8ll5Yu4}*OmqcVjgVd`ptK}{7LDa z+|BOe4u-DD(0aPPeXd-4GPSwJ~tx6Q^_7+!r^wslb~V(}`n=t*j# z^0Q?ju6#gY9QPBIwJNDLqRE^|2$sVZ$A*X8Z>IvMZDsj2br)WSXmFZsWl0;z?0v>L z*}J#7pVQ82Hm5$=&+*;oMX!egjR&`G@jbTc;!CL3&{8O_^u33RT=~8xcd|9YzD(!l zQ&S{fY`L_{n_SXdZ?RR+3!0$1hjRN0rwgM_eQH-7c^*pZU6G3@&!_5+3$3N zuMMzfK+lK%EQSUc-EXAuke=hpmr4kKt3D3^K1vd8Dh6L~mwC!$&Lh|=Km29X=+|QC zP1#w(E51WXK`D=ajxY=S&7V$$CWGV{7j+kLDG7v`FSx^XZ%wYj z&%RkRND~?My;ao_N;QT0&YN}>Ui!%1WjbI})2gIJ}~7&G7z*{DrZCW0_eTKous1s@qQ8k`3Lg#yh@U zL-!i%C#EzeAF?PNYVWGu9DdtsD%5p@J5%!|N%>TY$pXe7o|$T9O_OQOChi>G?|+Fm zaPG1ek;eUK@ez7SEwm-pSsy9vbkLPA5JmWP{46-Q%+LN2^#RWM7}iUt{w0F+5361* zKLu`!0oNwUXXc`gZ!wnEYy0^ZMo@Dz^CV^$>oav8Qm5@`!QC51QfH69E#S!N=S~fW z^6#_>gT`6NtcB_n&iVI!!k~$Q8WPzwj1Bd`_-FU-|50e!9R9hbtJ)H!Ezs-v;Z`E? z6C)QgH&&N7>OQ0WeO!z3!X*hJHoxPwJ@4KXtCP29k-SLR+8+%mW@}!$zvHblu$n@k zXW#Q+vtQ^oXYxi5Zkgc8?J8kpBp&mq9|HUsa)j(Je zW?lTCv3NCGuQZl^LQ4qtPunE*NWwj*R=21^fza?Tw~3i3wb-}(lVw4>UU zS2JxwbNQU~Sk(JE=hL9xIJDQaRk7Ar;6VmyFmgNXbL;RDCq%qi!y!)@UDFivGA+vB zjqYeN4BdA}`K(1=%hu_4c*=QgP`yv)+g6`HWoUfO0nDYxO<|7|R(;=oE09an(=T1^ z38wi|&KJPU{RR}4I;CsL*>Z}g#XceGvZpb2uS-%dinlGG(0Jba{(aVDWzc$1xn z)ux!AP!;};)4)FqUBfDmF9i|G9XkZ0A|U73@WKmo^zKwY(vNc0n2>2UaqFOMYoBs@ zo-|{@awcMZ`n5r!b*s{52h4Hz6q?(TG-096jTVq3s z>qRAIYr06*&khQ*;#~jz8BNDrk zJzc9h@3jKmihc~Hcy!ULN+DW`=Jhj%%na9jFYTWCN@swtY%_ll!h_~ zrB$VY=A|bX%lZizIJ6oJ=$&v5BPK7!s;vVDoI8dFg2B@ul|LefsiI9PMkV}2b8O>l?H?~;%_&itw63&9EB z@LBUrlYaFYZ9<10R+PNM{RnwY02G@78L5~I_=fFF%gpW-ug9LQVL=B zOh9(&kDTC3gIyDmz3Bcu9|e_?Q;P^7cF_7QhZ*PTvC5VPPsqQXKA?c7q|z0`z>uNs zKBIOTDD8gyB|T7!Pq%V6b$;~w$|`;VWlw?i)C4y()}^?=q&M*rbSmt7L8Q~@={)Qu z&GH%44i`~#mNAOmpXGAl1NIYRBKRSE9#>{!FTHQ{Q-!beMv5@OKf3_@lW5h7hywLG zxSm@1kwU$_#~glLM{=?r{lzu& zCzj+6fM_QjUk_c##~CG1UkbOffeA3i&8V)ByIfCIsRU@%+2@`>uv+d3PKK8@jS{jz z@5x*u_x9SkZyD|JAg$ln49P*k`27oNZ=u%R>T_fs9Cm4-K`*jlwBq_uU`wmm;pMe$ zvoSOOe` zG*5or0r6-CorV$-g8kOKgn46q;+*4gMKYn@^`3DiU6k&Ps;6~$D+@XKUAa9C1$`l; zdAEJ#%pj;_SW|Dz?J*ENN1G=RM>tG_S#OZ##{Ff~E_C?0(~54ZOP_|66|d)r~B(en^+C8=^t~ih9 z78;pT{(L*TGa9XcfBpZ1$W{1Mh3OU18bY4E>Qkki@xP_#{C52cho+06N-*j$_5$hA z=u>oaU>7Y&trv}SDe*CK9TXo$Ds8&I$biov$saBetHm53AyN6&NpQ0agq9euIV3lE z?CmW37vb$kVDto=BmK6DHqN6tv!`lY_@f((<0SdAvUwi0U?ZIa-y@8Ot`V{oRXc;V^DXidBk<^ z(|wog$vq5j=0aku?u!^8@S)D{h(P3GxV9CirS2TA` z(xu_IF>rmTSR?xB6(SPqOf||M9uFZ5y2jPAOhxeVN@hbn!R0P1mKQUiP?w>PQtH_aQ76w8AkR~q5Pj+OQ1QwPhw?M_RZ}V>z&MwuMaj8&s z5tNPy${te)U7BqT%oj+2ZszvZc0KW2>kh#LxpH4f$<qkEUFG|$}C)FH~o-}5REJ}n4;$NKeZ=3BGziy1LwWCk>J0N^Tu<^veY+fq%1 za>I0?v-o>9r(Xjs0_%U(O(X4b>~R#7lo%u>FFfddwY-lui>tg}RX3FfH~7?i!ZlJt zHK_3l9!@&aRebq}i$(91{sOqu$d0PwHggWh{fLvsngikjml19zIZ*pouHbSD3aHrdxTtFq1nf0d7x znwF>6_VN-NcHK(mab3HgV>jM&WxyZtCodWHnQqgJ0_&shk`l;Ez# zPzj*D9@md>(h5R;iJ=wc0&>)w5crbnCPhIKBUbz@JpYAnzIs5~@!x4VfAIwPY4J}j$(^nU&Y}sYfcm~o zk{bunaPr4Y6RZ$@g5!waPjO5f0F%T7>Xy0{Y{-c8eh1bU9v3*!kI;!-2?60p&sUGf zGZzT)-rDn3e`VJ=m{V(>z~=duDWqtuH>{WQ9T8A^WsPgOv=12v8a}yQM%0YGO#y}s zTF=6*!_};_bNTqJMzhc@_Em1*gkZ5gq#pW5I24v)hH2m8|K>Oxy15ZC2?#&uycZ|` za_VW(MB2KnqA8Uo4p#~EP8jRdG3p9=w zX(GNFTjQ8@<*%>G6n5^FT!t&V_5u+wju@)x(SY!{sBqoTfzESf&VlqTczvJfXq)IL zLNsw6UOz0F=xwm2)kIE1-ZjkXK3h>*wR<4!U9j~u^swdC7Fn~mV9i2F5I0EmC)tv9 zEir6ed&x2CcY^sMk!T^WTun5@0a=pu)P*ULm}^a0`#O^2Ub|K0RNxc-*^~aA z+~Ybm4O~Ho``rLj@SFd@soIisF%_ODr+}Q(q>16vZDKe~;piu{ep#Da71kZwVExoT z!kTj_h(yT|?5R+iXTITxA73!yR^3|xAcM!XLMxn)$Nq4U`)5GQtVoJ{ zMynkXPWBXiXNT&)oQd|=5x;jkpMYad$xGj=7jF6exa^qxc=pxyTEo}c6Mbjm)&&QDM^JJRt zi7PZXMYovP@<~$l1(&~X7JInqP)JKuGCB>!EEyX(Qhl?y%lg%=6gM*9uyqc36}Y*O zVYNCypVUtgGP67}QfkswD5ZGN9{pgIvj2c1J!@5hDATlHzY*-p!Zt|Epf)P=S*zfN<>M6>Fa^-$l^T#^JYd%BTn%vrc4_iu@4sJ(^Vfg>> z|N2OKTL2xA$I2h^ACi z*f0M6@mSt%c_XM)dM7+jn;!8a^$t01s#+*-^I>?_^PcFfADs0+td0;#GlXHOPL2yYDW}e3L3IIk$yqy4BEjHfn8vX&9jbb*qH9|vo!Za zrweTaS#K?+OovU6Esx*HY<2i!{zx37pAPC0IQ>m^+CUYpNey$c9ShT}`+gH8W4u@e zG|$B=+ESs_Sa+g^%HqTYbLCU8(%?-A$~eqsXVtB*w$9!TjrTpcdLUqVd#e46*(*l81zmW~O9 z>8h+q5ht!DE1ig+we3!g`MT2oPyQby~2B^IHw4->y&+q4^aThA>6J>biEv;JBi~H(U z_+y@Jz4Q;(()StIPX{isWS`3Gto!fmB|HiJ>!X@c41~{{DVML(%WU%_)7mqY}kvP6>t9N0ycDoPYmnHyAoS zb~Q&@$F>*k3P3*R@)k;fx6`sAYL(XnhlkzeoN`vr=kGgdm@XW;$&KE|GTbKmr{Q?Q z>!pRA?l0YePjM45EV<`=S`7^jWfgIz_nxLl`c72SCNytay!740ozVz6VbboykM53<6}gT=5bt!c;kjdm|tRTh50XzO&U#}wMj{>JKdhr>%gVm zh`A>i`a}3MA!TFjC8sC(^o9TSrD4;u_Kb4I#F%zH61Z855de&%nttpZlb*0rU-8SGbw4`$6Vc2YfyPte^Cc0wS;!XFXygTf09|y6l&)qu zNF6YK-}*y)x>we}W0SFf9dH>-WZ`)Rl#Fd{-NcI_5sB$IU;WA1j5d{X)9SgSaV`8> ze*U#N*=xJNAUc1P>=vN88SwIoxQgRI`659r@G<=}7seH*l2Z9$CzT!!X~4|u6+iPQ z6G5@OYHNihjw^euQ zPY-ZP@C>VoU+e9nL#oKaoO@}VjMz|>+@lq1h&Ltlu>U+uV>5wQf)&aEaRF;e`yX*A za7p5|H0&vgeEL|wIzx`#hUX-`Bz^#HFxD)Cu2kC!ok1ubmxNqiN>_st%sLrcK)91ugskjV?~ zaaKX+hQL9nOZ!C7dSd ztbO;oH!NJ}r^a&=X+358g2n>AtCBniM$)YXTiMrqWIxl2rJQWmJ+1HULtEWq4pL@2 zw?7#N^D1)kJZ6le_-H0tZJctGmF6EIi}3u88e@*Raoy1d13FTn;z1vZr@JSJZKU_Y z^Gc^fL9vRDLhNIp=GMOJ*L&4PzvYQkZJuwr;i?TGt zQZI}GtL^vd&s;@YS@Y}Qcmqy1XQjRP-daTz1&2xEsbE~b7JR7KKVm=CZDDM;gfgqY zLqi@i*b3Zv^$!0UmJiYQpQ9+`k1efp+*wwL5YS4qAnvTCUx-IUCi-6}=6QIUI_>PL z<;vJDSbI0XoEAEx&d-;EeVVM<&FWs|Wj^&zO%mg%L*CXWj=`(#NT1Jq&5N?|y`v`>-W&w)T_s{B(7*%?vy&N6&w0DD7M^R|sxsoqG26gA7`N8aAy4AKBoD#8;npe8;!@&)7YkPsi(ENo=M# z8#t$K?U>7tm~_?Z>2^quNxy5;hTayLmhvV3)(jr~1m3&lk8 zb=A$^#!h}mU+}6Tx>*gLQ?EHp*L~K@gr^_pbyO+m0Zu<{=WPyzV>vXH+o)-NV}5>V z$1dZF*oTkJ_3cLr?ou||Wa0S_X^IlU;wkjee^r90HeMVR>qwfcZ_R%9esgg@JxlMf zZf2_VDgTOVpt?pjLGUe3P;tOhlCx0pr9ZY4^&Q+E?EIW`+e}$_-BfLv$dA7XH-=>V zrt~7@)VqX!g;xm%pvocG-L zjN+Pv`P}-waMxW{_E_xAM(7V+;uBqGWuxF=?v{q!2j!CcwVkZ_v0gZ|#bFAKfCtjj zHvvQ&jm!mU`*oFyhb2}SHy7)MuI=3EPE*V+O@AjR9W=>-IRB93baeJsi<8aBNE^r1 zQmJ`SHZ3`=i(F=$EQ6c6o01%Js&{vz-zdGtV3+OfIa$sXZLHTBD4iJ-sP5Okmd#?( z65y?*rV8M0(j7-LvH2{=Bxt{uCezSp$(Ts2_^|xxvFe-aa>>a^NAuUIkdr30q0R{# z*PY!0Q^+Jw0i!f11WFm65Sn?jVAE=NgxL{XaP8S`=`SJWJD6^4V%Y@QEJT10$!2Hf}lcaK&O3TnHW$T_nHj{67q)C@)I-d3%fYhkN2u zD+<;_V~zTQLcCF|xux2ss~!UOcrp%$cJdHQi{@9esm2Iz_h*StQi{U|wqaIOF5ej# zT$SVs&o3yF2pjAm`aQPgFgxsN97yJ=#+$Vle&~I@zaT2r2`<)SA~%$G);-u*F(d&r z{q@*3!9I%ps%ljdhcZB@#t{Uk;MqN6CVeJ2od>NNEA21j% zGN?hnP!NI|;>N;*V@YI-3rJ;Ot&cm3g5K^sEL9tlKV2Dr-_Xah&z>2h)!jC@ryuTZ zgd>45d0U0}!e<8QC5CeTReK~f`*3#Id6Sw@1=~4LIq={KpJQI8lj3l#t-|fAV-Iq& zT#MbK?o(V^wPXIEstaG5Bizp9&BRkqe|(C@&ah6AKQ3%8_c!zl>0C3Dn!*%R3s*Fx z0A8iXpx4pl&;YRTS}TIm`A9vSE%%RrESJxJZCb!*j|-+XrQM>Pfj}(Kr7=(+FRh_E^fT9UF!N#`?bZ5_SA7^^lz3)fs%Q-l78!?YyV71x<| zX#yLmH{FzdzwJA$U(=qVUM?^xu*c*n*-{!b075+r$N<1}JO!uTi|h|g;sN(jlkD2G zaII3$v09F)vJ_nI^^boiXs?S2t)fR~1r6!2HZ;t@P-RzD?2vRsF*h1tfIn>P1VQ9u z_Zv4A8md0q_K0r5CND_nU9w2 z9;jvp;HN9oPi7p-iZBIEodIgiW(<~u!D{*|&J!UgP9}sHg@Nv6re{5|x0Z2p1MUs)x9|Ng(;6tB$PMOntk$`U zgsFEgWnQKJNC>eapLVUiUTj`}(h};&Y3-kJ7t+z3xFr7Iq4ie}>K?jK8D&aobBE=v zG-=H#H(PxEx>InSR}P2p7FwEGOE-No*VT9t*j=`_j60EyI-H4T*h z&Xlg!=W;xe`$|gWy66Hw`+kpmVH_na2$acas#(vx2LxoT^J`oNMy+oTd0z5<5^Jd* z^uxq;f2#9*=Ua3B(ec{7N{Me$+HvuSXA~s)R-l^VOE75zp$8P;V=*fTXW-228ks`U zpY+5&P#6&@d}T8(WOe5zwbAJLf6^l?`&4v;HxQEE+Lr#>l4EYc>Qsd964sQ+H%4i{ zH|X@Ge_CLTN+?EV9;a~d0y?183&rQDe!_ z%PSCeKzzZA{)m7UC5lQ2?Y-jOCYvSUKND_hAM!<*Whd?^_J~!yBJEP8;c?Q-A=RwZ zH5Fg+4|eGNG4i=hv7n8+wISSuq(7%wUBmSrW|xc)x*On{GlA=&45DBMtq%MWI4i`q zxR=3KpDOCd8RFTrU#}#=V-#l=WpTcb3IqO=;UE~gx%A>9*L&NThYP!19Q92m@jLf| zxp9J68rs_hvKJz#5kN*9_XXhx_8>G7nA(>(SsIxPi~vd8J8G_g%dNTn=FV2^f7n%? z^t6z!iSUf%?)?zLy+|@`{!6^UE1Z{vUs}7`_zPlBhayk47&P3qUD`okUh(nN5lEwN znFws(*_Qm`sg293y%jmEuUp|t@84WR4=E6S0H9#H2#OAqjy>1Gg&Gxg^5%!AFg}%E z%MZoR^)4^?u8u8lf@R-U`dt?h4c(Z5e!nD#um0Gfhc`Dze=%Y_`i0ts<7Ik*mGu+! z$gunAx4{I(q_G#XXNc@4q-B^M`4J1T)_fpmOV}v~(gnMl;#-)4U<7E;+ z-p!Kn%sDP0y1LXQ8U#pk4)^Att#%t7p;B@Z^3+|~1Z*i}!5cmcpyAqd3E@sJ9sdQAP*Sq&zfIuq8C<6TKK5i|j06=Or=w#!Kw4@) zZW7$DM+&1%CRXQ}hGQS|`16_c=NLO>7F_Uz>AcgBwN7*<(}awWq=%hK^^z?6cn@y>rAi@)r zVCYNZ#ea9PwH>>y`w}8`oBArC^L09WRb;$c=;8qbUuUPBY!pD^HVB|Eq{q;NLuZMbr_2Jcz5B^Wz5@Zn&G8eQz znYb)aXJ*jOd^Sv+3JpGPL4gtdoCpt}H~aX3>`2?sP_+r4A&qNT{<{d>%$w<^vISHH z5BOSY{AAgh{k=u;ot>%so2>dDD%z1#V|QWtiZ2m$PWZp|El}S&;^ZRHkSAv?NOsro zWg`*g`aGS`3e>3$=nZc`BMknwB3piNLeWgyWRbz9Xm02#-p$RJB@!sjmC_fRg1idH z0C<4b3l*6apE?J$v#O`dD8R6=(CRcT^szEga-Xev6c}?oV$eWm{ccTbAj`N7Ju0rk37m$EKWXZtJYsuvP3XC(9uWFXP1&=%*B2i4^N4 z@v?1Ee?yHeU`T3**ni|?e>sBKmp|1s2K`h`>9B|^BmU_Q<-c+Y18LcR)N^sb)wXI2 zq}f;auMq&gYM z8E6G6z%RwDQjK2yLFy!pq4S^pLY8n1`-j7N_$yK&ReeZ6ya^7UaalK`ES5eQuhc+1 zK)!l?P70j;Eam8^eY;k`o|T`?&H?YJV__g2Ps3vLM}O6f*ZeD_a5^B)pCHt2(6fih zBa8mr;;zpQ<)2gSev(K9&KIER8Vnruz&d6Gv(Pa;W{~5F(2HyRuRod>dNU?a>*FzV zO%O?-5n!HT#07*ygqrH6 z*xL@lGAqAi;7`8&9=a)qn$$y0DkByUhH`-AmU8WFoP+ECh3hZCy|F+~xGRf};o)JX zs_jHd$NykuS?d_7u5bcXoA#^^9J~0pfvjOQCWuh4b$enqB;WB&iRy~yXLZP5|IiNg zVENOQ(B?q+*o(Xw)tJ-F=O7SLU=m>EB&;UiCg4Vymm_{?psp3W>*B$2NO9T~93LlB zl2BIEeW04lve7Vm?qW~(pA+c5j#e~obhAO`DM3VXtA+1{cotb|mhEcH1mQ;xfXH8# zAW~07_g_Qbx|B99EO;Odc&w~@!!A2}6i=ASvWHGM5r3SG_8vM0!5VRpt*lk{SeY_M z6Ido-G(L_GLm}ERm^gd!&E66sj46!C#LOI%mgaZ9wesj6i7%C0>UOuxXYgss26-zg z-QQ@TPnzWEM_mz6cTf8tCjb7&M1>x)MsN22rLXIsQ}7+l6D6qmRUK6S=0NN94h4q~ z`mz)rVe+$T=$8(xW&s{Hp_MSXtlL2CXZ*LV^}qalW!?=xUqhB2!1fJyg;)Sb4fhfP z0YhZfoAw-0bRFEdZ~MPIwclSE=(u8#848LWK zY=QmHD*x3RWUX~%-#dW<;ifN?8iX2fmszh3F!pQzR5x6dEo*{oen*JIhX}|!*QfzC zxFSJeN`p&V9tQuCtfpeu${P3I!aW8;)E%v68tHe7Y0g4n>Vvwh9lMj^!)@0U#|+p3 znjfJLE3bB9@kDugez!2RJ3qCB>sn0@0yru&uuDS1m0Vq|d-2=im#-v`Wk5J`DsedR z{B#1iuVNocK&vTi=V%OhaYH~Wd7^zBRnh-^aOT&{!+%R zNf1IH%voH?H2u>3dZzb6d=dG6DFdcu+=x%|@{?PE@#g4-f#oFMN?5j-2yU%znZ3{= za#y-p7hR-GAz7}ibx2)?lIHsghTW;m#HEz(a9t^6Z1~wzj-@l2?IKtYI$wZ(Ko!Ct zOl774Ngi_u7(?iR_7e#|;pPeJMcQG~Goq!v-;=KJ zsd_;>(u27hv=Frk-c|`-JI<>l|2w06p?hc5j9hvZo7oAGVef}xnoH2fvg z-a5ZO-AFs0z8rbU2B;YZ&8XiJZzjs&z3v|tt%*FH24$W-EqRV?MTlL zbVK+;ocqm`G?dTWa5y}fU1WkN41F^TR9}88h>q1pF9aifpqu9WHbj${-_a8=7E+;O z8Oq~{UW!ao)Wd`#1U@(ffP4R3lf%2+{DjZ1qv~p zDVR4ke>{4GyFHs?>^vIJeBO7)UhCuNeXTs(NM4Cl!^0#Y`6SU_U4HZ>J+5X|7-z7e7V%fMTsyUQ-XioKgbuKPGThJcQ zu>C~m?Upcj>2{RUzyE^cL>d#hQZN^)vGGg&2#)+X3!uJj`($J0Lp;%p+2_QJ<_{FbaqEba(#)2i=t^VYGjF z2)|d=GEl*Ra2WywQ#qaGu%~O&O&@D%s(cL%L@pL)&PECn)8&TuBrJyqo1ed%yT1rI z#nO7oV7+#G#v-Qa`4GY;lI$fOYgv%xf=-jA-za}X-^Q5pA{YSP_hv^q10{c#wNRKq9HA6lU&|PS(i)niT;tH(H{I2`E-^Y4JvU2d+uPBK6q$Y9+J8 z{tqQC8hDF#tJtSWt*bk6JcCTwAs3z1jWfro7dj2t{Pe*eaZ9C-YJN$Cou9@lW`LQv1P!loEu{OX7P~0h_4fWG$=QV_Nj3 z6?hP>P4pzE!(xYl<;A_u?FfzdYWrpRAVgo?DAPcW%zs^vPrs8t6;=%Zo0ST4Js*x4UODn91ng zJyAV;)W-a5_s$BmJe_dZ-el^YFu#K3n*PzwB^+Ajy>PuKUo`x5{6?1Flr~px!U*(? zx7r4dTVE+&nzA7`Ceu4?@JFCeuiOM2?$i9s?;`;Ph~VB4XXM88+k~g+^j6w|-$2C- z;6haN*4jinZ*5_6qiuQve4brhEvnirh}^~?A|&*r*LnsdLKG%+_e;> zSp!%IjpX4B{{W>6(g`Xr!C=Df($p3)r;YyfI?L(c9(MaZW=z184SpuImf3FUi>PFFFH`QnQr!rmo-%?uz5bPo7+ z!5qdUZ>L@{B08?cx|sZb+*4%|%<%u^2|W}!$_PDf)tpcrZQeS^e#GC!?5b`6BuPX{ zbFqP4$nIWqum^RuJ|Mccp`Kt2w~^0scJ5f#TV|bO7b~e5dGSN~fp{zTk^V_ksVX@& zCT+(tObZl4zAitNH*f{E{Fy_eX@mXRUM$WDng+PLl&M!E!@AvJPffo$NIshx6~V0^ zp%<-KEj*g-w(#|r)v|7$y?7oE}{{c3wXGx>h> z|5lJ7q66xcjo0Db7=E`+Ae4#jn%B}1EI-T-=z3aqe>75H)IpjlyduNgd)|I-LfpvS zH>)*-U&8`=l$d7;Mu;=Tf>a&|aW1eB|iK)f@S2pS7*}1BQj|SPZt^!wZGkQBj93JgQ z^0Jhyg%JrDEC&UeU2GSGB<)fLRWQ|hqTSFvZ`>OOf5`L(VA62`=_ZxmyK4VCoj{6a z2QgFIHX>~YWD|nBZ@t$p*L7|~j^bJv%;ZAbPVWM_Gemz*)LQI$Q^zRdHdK*jrL`&@}ldxmYz_h25lJA*K#w}anE$0O-j(m1Zh20Of7=Tn(C zGjAvuu`I%tW*9y-mg+`cPwHL9+Qc6mld)u?AHdlixb)e-yHerenAQek!0gKfm4eex z+R10iV*HQwNV3cc6KJ_MOYMiF_tiPWxr!PW_p$f$6d`Rsd`Xvk?w8{_GP7 z-E^9D3V0fXQ#(`rM?ppYtM>PHY|SM zT|8-mq<$I2k|qNrRA!2M2B?gua$&u`^6U6D3lK(2d$6qSdF*>te8oQ^WI}$kXN+dP z9}D`C1)ko!)Gn){9V+oUtYD-42~)V9?p-V;_eTPcYgYa#hV?+-fdynQ&=xrRZg3mj~JvdDH9pLUc; z{ibPx1dsmzSE;L`CC>I=R#zN#h_0+Gg+RJtyac>_Z)7|KLZD>V{$&VaJMIe7@}=#0 z_;7Xk+T2^hUbkrnuIp12SKDxTz-F+A{jnp3X8%&Y*gGZ$LSG2898I{FDbj@%&#l7;j4E0 zxEt*vT?%<$0`xglKt&q8NkQ@T_Kl<&52*VRdx}IIiT&`VHB1_yT;eW@XM9G7=J^YF zPlP_{89FIl3ph^1I0Y=%rf8Acs8zh3l^L6@S$Bo2vZMV z9KEUgs!88i9tsS;5Bbg{fKFNV1Isj&s=`cP$f)ciB|!qGvc;u`F^z6Q zkg@)6U*@Z(^k}Kw#-qtIsiS};`94Y1WPHtLlQvwUB&3#g#DX@8NilY4i{af^TnCT$ zE|pN@&fTUn#J#5a@2Tn1Rq=n%a!-1apSWf>|FTSCD-0OHf4yfwt*LI5p&M` zhYhMd$WiG63GBV8EuMokJ|66fGM|F;A7=Wlb)TP?tNJ+3xfJ|Tt{{?+Wm3H7BAI*Q zd-ySTdn$Xw$KmfiuB%*J(emn{2PxT)_HU2$nJca z?|GM}W(AQPSTCCr$qyK(nMj2>L3*vnl|ZxGw5K81d;RX<`E$`8_1K6A&65ZNclY@a z=0e!8Mh8;SRX3Co95#95tDKX@sOH9Xg6G-^JJOd2FPHt6D19yu5epgq)8}R7AG#610S#ZXQY!HrDi_8 zkE5gn2cMx^&ZE6pV%(7I^tG_Yvc2ZtL4_3Qa+)O#^M&*+gQIr065p!Ze3~P;m~&z% zHeux#8;_m4OJ2?@)YR##?l8Ao8~2#|K>9QIZkM9=sFHaz%Bu23P1Ig9cS_d^$(8l`MXb-ULWY-t-x@lvNH1)??!%zb;!Xz-lkyP$Jht){|bbJ`pxLVSvxh<;jj zp>IwWxW-Lypa0u;M$1x?%hf`BrNuFLu8iM^ zv|~e<@HORDPTLBTG^AsJ=?UFjH`C59fo5V3i5TG1;rbMkEv+_HJ6p|Tw-DhUhmkgR zZ~gQ!Ez}Dw-Uc^YEB=nz&NYv@HVX5NiP}|SwFANjVDB0+`H$G=<_D9hYY*@!I9?+( z`?luOScRc|n))?MKit4&B}e=wg{afDo>x6zt`u^#-m`>yD~f$kxvfd!#Jr97h1%op zv8B@A>6zb`L5R?R$4|t!m*Rh`N=7wPda!QabH=aZ#*|9bsB`$T`R0p3vxlJIXY&Tv z-mHHHx34dYWv43mU2u3i&2w;NWqm%e(0Fz_x5}|EF6CrgogQT)9UStK94_1#cw+Yv zK01#0lUz_FlbDssyAqrqRzpyOQ$t$AeRUBPCL)+UG%bHdjAo2Ejn9GIi&641%~VH3&T+zNPdhE``I&>zb_q0xr{MPw+iP2Rp`mRAbduLw z-GxP}&z+>w)-nuhrqmwgb52(INTRVu7ruF_Uor)MY?jBy-6P_x_)CpqI64|rI4i!S z$H8BOfeS+l&8IH7PZ2JvWno<@9$w+(68mVQMP;GmYDFgqFJfTwPn?@rYy}SX+xt?g)Q#am-0E zp$fft#I4*5MKM1D>o;r+Y=D^?^`@-@qF$a{Z4BDR!wsC?h_Kyw4GVgd7MeDq%Iuc= z3s)DnoyP190}mxK&I87OaaBW_?J*uiCC|M5yo?Y_Oip2{@+Lg zmdDw^f$)@8#_Z(?^W@YqEOatknGha^<>P$0`)A1Jie>g)D}0ZCR-JJ^6d|*CTYn4v z*WTzve}||sh8w!{W@r*_TQiq`lpVMG?j|%Ba`Pl)q3amxwx6&HJ6(3Eq7_XX@*PQ4 z279NEBh_EoE~eA#$z#E#GK8+Gnx8`|hh7@##X2p<4mE$%>wI)D9wl|SaU^GQzl>?d zp3ISPoASx4t#*-Y1{_lFZuc5;P+j~{cz+$XSu&q3+yzMav)K*eG|hjt`5`z1DN!zcm_T}CS6Q*ucpdwXhM z|00S)N#o*Q(9~5Ke|0i4=rQGVDb&Qc!Qa>-#SZ`PLf@=&D|*T4TH#q#eV9Wg@cJyi z({WqhT)ecXa#X*x^S>(RJ1W%d#$$vfa%=QRmZuG9FHy^LDF)LFwP-0>Imk*77~DGl z(vg85=@`4<>qUTMi}Gn`lIU+f`= zjgu@H2z-S_LImP8Ou;(@JfvrGZxLdze|Ue!a9qtw(Gj9N6=>z3tD2YneepOo{GruK zrzzXaFrtTK=Rq0OF~5n#b~GA^6WJ4`zo{o1LbH_!Oazs)!$AsqgUuB-<2)H6G0+`N z{rS&J65V=DC2^d&w5V|aGj<|7(owCYDVsK?%!vNt{M*6^ z*b{a~sM6mEBfDw*1~L!A7Z@CvQo=)Yq)IZ1vOJX;$@zrC4baWIhns85uQn!jg9|v% zH~P{SSFb!`!Pfr8o7&)^j?s^uVlIlSgI_3fG?c2J%Svu1F2-h(FFqG7?h>~8>#Fp) z)^tem39F;9qneVHYqStOQ#I9Tabs^INlm8o!j^+Dhpwuw;&xO8tHrh1-)m=>=bOC; z`sbR?kHi_~hvGH7>2tMW^@ZjN7e|-ZTc0_fjGBU{^nmcVk-Jx$LV#e z1Gr>yn)*4N?NnvxS=zL-$-@KxGu2JQmxCQqVmupLm73}%Z{>8l36+2U;5)I3*EnU_ z?rFWR|HtNx*JPy7z(L8sO(@A|ePt1sxOm}6oxbnSfc6dM_&<$(2{@E*`}Ww9WoYWB zg;0ZEjby19OQ?)AgUZR=d&7vi=@ zl379vedM*Ue$$_p7l9pV(ynk%PEY86yBb~Mu{F}AM~;@ZZ$3O;Dtjikn5g^iNou`A zK;#+g*z;w&3}5V0dTFI+hd-y!GNJ93lAn5J>Vue|hJ=<~N~Y57y2!59 zg!4h4PITMtQeyru{O?bPf7m?z{yBkbW8R~h^(QlaFuaHS`5K^@u|56o0q#TOrhn4E zFrII3FE+I?ZuWUV;``6X-F1Psy1Cf4lV$Xx%Ma$H5N)i)MZA@_&reHLww{nuRM;(w z&Uukvup}5oTH8B^r{5=cw$~Ni6<%M?h>MFeD!Qo?;J0+#Yr0<+*CUhzPzD;K;sp89 z$?Ya&czLsZzk%})>xOP~#;~u~LiUbdbnt%p^#@psPuH6rpwx4Q>@W@Fv|HIyA1=R! z(>{?e=kNv~MVA4E3!X35XMAyrE7)b7&;V0{B+e{%%SPeOU;4ER>giHUZdVY>K4o4A zsvFDj9tybdib>h6ule{pi3d-c)L$_7Ei35YL=J8`_ayjB_Xhl8 zoYT)Q?{<&HQ{+_GT3lUK$Hwz5{avJN_+Wo9^~_vv&JpcE0w^?mOcB}9XE4T{0Anh3}KLFJ|L+%VC z)`+14?K(L`?ru7n`hZ(+Pgh%R&-DAM>Di_>ztNerbyaU=ZEd?F9@nz$!9ci71RT__%m<}&P>;re!F=VrYb~Rd4IPLUQ z@fJIcABVwj23}CO<&^7%WHvnN`#CaL2S9h>iEyixZ!&&M_vHX|+cteNEjv*YvGL-R z$Jqxbdctk-P9lpY0u*0s|93i?;sys(_w=3SEh+T2M9!+%|6&a(qgt3B&uK_r4`(ak zP5cyRBZ7+9+D}P2p6!KFaFfrDMaFk)VIJ$7S}4wy)C8<|DZ1iuP^wsMy-scAbdA+` zicD*~USm;^BL)-ZdfXFT_vm5z#Uii^q-tK!-E3%)T-UF#Mv5+Hl5i%YCsSp^M$9L) ztY)iSg)9o6Gyo-#qmeLHVmGzk&N2-N%@H3BMSHe79;wR$cVOETNBgKNjRIiWzK8J% z^cH#WT2?JB{U(4ZOoo6v4<49`^|e;^5JyS}Y%?eHRfYhXYFy%_)Y4x%|D+<}mC{)M zwRGX8R==l?1}%rLItXd2$8u$8*2RLcMq)?aY#PX9C&jU4Klv@7(1AHAqBmVKl8V!V zR9kDfC#lMtYi%A%M%~pf#+*v?zr|+nvbl&;&TWTP6_H!H)G*^`lN9+Itgx$0tGgek8l5oJ@8u|Mg|+8q^n(7N2e~| zns;b}-0|{Utbcdt*HI6Y7CKMM;kSDi*H=C$dVRjwkWrG#j}hE=49WqLpf*UbqTHt8 z^7gkugq~sl7+|iEdZYzF)PJ)$Nmbu75G9Kl+x0hQNsjCNE`#On2X+`A6d{)l$nG2U z{VZ;f^M%x~{N&7nN;dQ(t0>!jgLiiLu^7-2$$tW4mE>isz}3%h{2|)34oR&CjaT>_ z+g##|Z*+UkVDltE2cFQ^aedfEhXao3fD11i6t%K)q)mSrXaGsNxS`u9^W(wOKLM%3 zqNJ%waR|_L?{>376csCc?rgOO2J7#vJcmNaQLHDBJ>OCxisT zMK8uOHzO?ez|^i*{3>{*F7aVH6B#=Kbd*oq+Q)d=v>- zx?cF)Sj+)Ij4pPY5b#!3_jcWLOm|X4h{7zNNez3}_$2jJmS~+)M9?8FU6z|;-L46P zOJ^*1a_#qhy>%U=30ho^Dt6IS7Yc-p*^1I>NM(_|a5y+aAIC`_Nh_U8w(DTJ4w61} zKH1VZ*AXiSnC6_f%8!%A6#fg|IX_aY^w!WK^d*RLV&R(urcftk)K&*ZAYvY}MGjjc?;{fu1bja9b+sK3u(UPalmjcGw=WbAtgqWVUzkpEac(0bA?=Ro z$)3(os$`a=;M=#KA?sf{WyyOO^L?j(A+RWV(teC8)=IhF63S6*rh+VWYixNTVY%8Y z00}5@7b#UKI!Krd#$HXNoYEWn!fMnA_{NbhM7^UDTF{7a^WJxqFlcD@Jo}T~`V!=I z)QB#8nf;W_<;ewCPWmLsILdaZ-?Udw)r9fNn4v-9b>v7C5^^=e256Ox<`=hh_>lXLc{5J<;qEW7v+mw*POe15;ZiR_DMEuS@j@AUfw?zNv zmV+L2D50gtb>dNtNu#E3MH#-K0KYld8Tp>CVBulrmU^a{)l}3a!@CZymUq!PW%>qj z!#$%$4(J)54jW}IJLT1=Rmb5%pBwqwI4jUYfQO?c+pXvRHC@}x$MFEfeCV<8_LA$3 zHiZ!Xp`o8s7=RIlxk~xDrvwXfff13z$ZuEcuW|pfZL|62&5xf~?<^1Y3~Af<)K0H@ zdKXqK(mSRmua|;Y{Zl?9%cNn;PRqn;w=BXc=D|pdH zo67a!&~Yx^Ux&;0v@`PMZCeh&Et?g|_dUH=B?x0&M(ZPewQPBH!c&ulTi}nt5Szz2 zGp-31i$`^g-;D}bew%4T7FHBaz3jzk-cOCg#7$lPU}oB|_68OzVD5RHDKqI!l9no;Oed%UX+?SzHD+BuV;6pZT)`NT3slO-&lsxB3HM= z^ZO9@Tya24OIjMeqRz^_FduU6@2dswM%M7h+8_ogb4z!xVa{+Avh6E5jPix zXGtVoUCgw9(GFxzmR?1}(k1Bqq~2)k8|7hV(L8=|eIqJU4acs{`a0rKAnYSI16s}= zkWFhtP)}{#^hV67(^2#z^g;wx82TQ1B{IcOAJ`j9`Nnqh5MXmyCXppoleK`ey7stgbTHOXJXc~72# zHF<&ESR{8sX}bc$;yIjMZUTubsO~`@~6-KVPVIvl4(~t6}lyxckldB z2GD>t(_iUNDO?QmTz^%OQn}YaSY@H8dvO7Y7H%IlloXUf$<{91l$vSQ)oP zLrNFSp=L1H>|>HHe1!jxL!yfsEULMUDpIc>gMH+O*{j#D5|7aA?-iw!NU<=P1FxCi zjg_YHPQCyb#g}ArSAI==MwsA)iA))N|1wu53j0PDK1r-$jU)2W8W2?3^}nkujmVdO;haDIlm_%Kf42(aXImFdx;qMA}(sxJtILT+95jW#Vl@Jee3i(U< z&I5*;D&gY=Z@jM?#=#E1BTG|;%{$OFXkxS}!s4E7FEBf|_8|nP3YgVQQODx_0LvH`*2L9h&4;}ds@eArj?dslsrfqn2GjU)1SC;(F*mv95fQ6m3 z-F^P^ggD5tvewO3**_>q=Fs@vF+>pu{y|1E2N;|vJSUW-5Bms%@oLmpX4!}vz+_b6 z_@l56brf*23c}KR)$7wlMq*(3XW?=h_0upJk*Ts^eUzUxT&QWar=3)$3jad!Gy^`C z-`_i{ivq@t*ev)+hM({v`Y)|dCl zv$5w{nQW~MyDBq=17hL=%50Ari{CDsFwKxi(ON90ERqR)G!4Sgk;XH660Fr6@htJY zKUwfXuoJdN^kAm!e*5JQP*?`p1`op0?D@OEnY8vSvur|5iqXgE952fodmQ?LbO+(D zc#ODN=PuL#qSw(Djo(p|nix~BDS>=9j*nYDJj1TEXB@{S!+z>Q{uoXYuL6NvBj&DP zepD&1D6=tY$en40Az5qc^^$Zo1XZ9Q%8Cio1@5=>j&N|i&;c^ zo>!xlHf>?O4ZDfluuluR)&`&&MXXJ_nfS_)ESe%C-6S6sc(g`6SVpXXM@ubaJ?4z{ zwm!ro?|y6`-@Bt((&|xx#+1nb=mkVxde62tt*n+;qw%p5VT`tA%)T4SBE~1WGO#mZ z0Q5h*ZKQr0{)J16T_wg-LaiQ4=cGfwCAfyMM59pt(HHs39cWRsLfY)pP2&}PdZ6(h z>hUIqrm^&nBJ()z;aT=m6a}NN!Q$+vT1M2u4^&NDxh%YyVsj8?$TGn+K_$N_2umX= zGqti#5;TbdBGy(I2lWywzBH2j<3YvKi@?d*o$D)zp^qL)aG_(X;ae4i%(EOBYf%v1 zWy+0W3AjmZ+9-LjSe>jt*LH4EC0EZ8hfZbWejpG{MAN@%J`@CV5kvNvw&&zUX zMe)w@cB+ADFv*Ef`b;xes6#Q~_-ocV;g!0ICxZi8H&)i}?Onrfcc|3*^1*c2;pAMqnK~Bnw9PU;(^lS6x58ae?LUaLng0 z4lpn8gITg|Lya_DlR8VE`Hn3LoaN4ZO(mDG4EWB3kYe6%Ro(gSFAN`raWB`v z)ct=3+{m= zF$6a6hhpdpyL5z;sJnRm96s_KCX-?Ke1o)RfO39c>I;Xf^iN@_u=yxbZ~RVHO*pBU zsY(Zo0B+6mce<*AH3s`JgrtTyppsXXdl{CrYlKp;sA)|w=`?~`(~%e;KBr4p1%-#E z84?}7J$kKCzZVCSKDc&%PF`Ahr|VMy%M9}urg$EPY~G%=#72A?n3;yUqgtIf5*gbq za|N3S9}%ZVoN&9VyEtsfwDS~u&p&)IhfQ+N-=MTbMm7P@`$onZR<{X*k3c6<$TQ)j zRhAt{d@R-kPx0A*3v_UNxPK|rC>CQcT$c?rC?C{A8Xee_?iU$6#6$PqSTh+;ZVSKn zVo?ouURtKc|KMEm7KhH48ptMLabaMBrG+PE-|{W{-03J~74R($hcapQXXj4FH=juF zoz_Gsf9@fYdUqNpdp>ZD;5L3#0(MnYvAoC zWp>|wR7T7}{;)4)8me3^{P}MF^a(^9Ow&4?qd9W0mS z8fOXM^b@6iexbT8hC6c06vbdEU*222hXINtpt?<)#!ggPzc(r+CD z!>JX!Wb8oJ+X-{{tws>w*9P literal 0 HcmV?d00001