{ "cells": [ { "cell_type": "markdown", "id": "1c6056e6-86ea-4c92-9e95-a64b38d6f4c9", "metadata": {}, "source": [ "# Build a self-supervised learning pipeline\n", "\n", "This notebook will explain how to use the selfeeg library to build a self-supervised learning pipeline. \n", "\n", "To summarize, typical steps include:\n", "\n", "1. [PRETRAINING](#Pretraining-Phase):\n", " 1. [define the pretraining dataloaders](#Define-Dataloaders)\n", " 2. [define the augmenter](#Define-the-data-augmenter)\n", " 3. [define the pretraining model](#Define-pretraining-model-and-other-training-objects) (and optional training elements)\n", " 4. [pretrain the model](#Pretrain-the-model)\n", "2. [FINE-TUNING](#Fine-tuning-Phase)\n", " 1. [define the fine-tuning dataloaders](#Define-fine-tuning-dataloaders)\n", " 2. [define the fine-tuning model](#Define-fine-tuning-model-and-other-training-objects) (and optional training elements)\n", " 3. [transfer the pretrained encoder's weights](#Define-fine-tuning-model-and-other-training-objects)\n", " 4. [fine-tune the model](#Fine-tuning)\n", "3. [FINAL EVALUATION](#Evaluate-fine-tuned-model)\n", "\n", "To better understand how the **dataloading** and **augmentations** module work, check the respective introductory notebooks" ] }, { "cell_type": "markdown", "id": "2813d14e-e23b-4cec-8490-6e92ed1bc02a", "metadata": {}, "source": [ "First, let's import all the packages necessary to run this notebook.\n", "\n", "
\n", "WARNING \n", " \n", "to run this notebook you will also need matplotlib, which are not listed in the main dependecies of the selfeeg library. Be sure to install them in your environment.\n", "\n", "
" ] }, { "cell_type": "code", "execution_count": 1, "id": "1732513d-39eb-4ddd-a6de-51e9574e515b", "metadata": {}, "outputs": [], "source": [ "# IMPORT BASE PACKAGES\n", "import os\n", "import random\n", "import pickle\n", "import copy\n", "import sys\n", "sys.path.append('..') # needed if you run this inside the selfeeg/doc folder\n", "\n", "import selfeeg\n", "import selfeeg.augmentation as aug\n", "import selfeeg.dataloading as dl\n", "\n", "# IMPORT CLASSICAL PACKAGES\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "\n", "# Draw figures inline with this notebook\n", "%matplotlib inline\n", "\n", "# IMPORT TORCH\n", "import torch\n", "import torch.nn as nn\n", "import torch.nn.functional as F\n", "from torch.utils.data import DataLoader\n", "\n", "# set seeds for reproducibility\n", "seed = 1234\n", "random.seed( seed )\n", "np.random.seed( seed )\n", "torch.manual_seed( seed )\n", "plt.style.use('seaborn-v0_8-white')\n", "plt.rcParams['figure.figsize'] = (15.0, 6.0)" ] }, { "cell_type": "markdown", "id": "969bdb4c-fc43-48ca-aac0-3a4bbfa898d4", "metadata": {}, "source": [ "## Pretraining Phase" ] }, { "cell_type": "markdown", "id": "475f25ac-e564-45df-9fb2-76edd8665d11", "metadata": {}, "source": [ "### Create simulated data\n", "\n", "Similarly to the dataloading introductory notebook. We will create a dataset with simulated EEG samples.\n", "\n", "EEG files will be stored in the format `\"{dataset_id}_{subject_id}_{session_id}_{trial_id}.pickle\"`. Also, EEG files will contain a dict with keys:\n", "1) data: the **EEG** with **16 Channels** , random legnth **from 512 to 1024**, sampling rate **128 Hz**\n", "2) label: a binary label with **0 = normal EEG** and **1 = abnormal (epileptic) EEG**. Class ratio is 80\\% normal and 20\\% abnormal.\n", "\n", "EEG files will be generated using an order-1 AutoRegressive model. Coefficients were calculated using a channel acquisition from two real EEGs, one for each category. \n", "\n", "The following cell will create a folder with simulated 1000 EEGs coming from:\n", "1) 5 datasets (ID from 1 to 5);\n", "2) 40 subjects per dataset (ID from 1 to 40)\n", "3) 5 session per subject (ID from 1 to 5)" ] }, { "cell_type": "code", "execution_count": 2, "id": "44b27b59-cbb0-46ba-b51e-572f397d3399", "metadata": {}, "outputs": [], "source": [ "classes = selfeeg.utils.create_dataset(p=0.6,return_labels=True)" ] }, { "cell_type": "markdown", "id": "2b6a7253-65e1-4ace-afc7-9c2c8fcadc2c", "metadata": {}, "source": [ "### Define dataloaders\n", "\n", "Let's assume we want to pretrain our model with the first four datasets, and evaluate it on the fifth.\n", "\n", "
\n", "NOTE 1 \n", " \n", "In this phase, it is suggested to put the fine-tuning data in the test set. In this way, it will be easier to extract the subtables with only the fine-tuning samples.\n", "\n", "
\n", "\n", "Since each EEG will have a different length, we will use 2s windows with 10\\% overlap" ] }, { "cell_type": "code", "execution_count": 4, "id": "f862b8c0-b7cc-466c-ac51-5aa35daff630", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "extracting EEG samples: 100%|███████████████████| 1000/1000 [00:00<00:00, 1930.72 files/s]\n", "\n", "Concluded extraction of repository length with the following specific: \n", "\n", "window ==> 2.00 s\n", "overlap ==> 10.00 %\n", "sampling rate ==> 128.00 Hz\n", "-----------------------------\n", "dataset length ==> 3203\n", "\n", "train ratio: 0.70\n", "validation ratio: 0.10\n", "test ratio: 0.20\n", "\n", "train labels ratio: 0.0=0.589, 1.0=0.411, \n", "val labels ratio: 0.0=0.597, 1.0=0.403, \n", "test labels ratio: 0.0=0.620, 1.0=0.380, \n", "\n" ] } ], "source": [ "# define file path, sampling rate, window length, overlap percentage, workers and batch size\n", "eegpath = 'Simulated_EEG'\n", "Chan = 8\n", "freq = 128\n", "window = 2\n", "overlap = 0.1\n", "workers = 0\n", "batchsize = 16\n", "\n", "# define custom loading function\n", "def loadEEG(path, return_label=False):\n", " with open(path, 'rb') as eegfile:\n", " EEG = pickle.load(eegfile)\n", " x = EEG['data']\n", " y = EEG['label']\n", " if return_label:\n", " return x, y\n", " else:\n", " return x\n", "\n", "# calculate dataset length\n", "EEGlen = dl.get_eeg_partition_number(\n", " eegpath, freq, window, overlap, file_format='*.pickle',\n", " load_function=loadEEG, verbose=True\n", ")\n", "\n", "# split dataset\n", "EEGsplit= dl.get_eeg_split_table(\n", " partition_table=EEGlen, val_ratio= 0.1, stratified=True, labels=classes,\n", " test_data_id=[5], split_tolerance=0.001, perseverance=10000\n", ")\n", "\n", "# check split\n", "dl.check_split(EEGlen, EEGsplit, classes)\n", "\n", "# define training dataloader\n", "trainset = dl.EEGDataset(EEGlen, EEGsplit, [freq, window, overlap], load_function=loadEEG)\n", "trainsampler = dl.EEGSampler(trainset, batchsize, workers)\n", "trainloader = DataLoader(\n", " dataset = trainset, batch_size=batchsize, sampler=trainsampler, num_workers=workers)\n", "\n", "# define validation dataloader\n", "valset = dl.EEGDataset(\n", " EEGlen, EEGsplit, [freq, window, overlap], 'validation', load_function=loadEEG)\n", "valloader = DataLoader(dataset = valset, batch_size= batchsize, shuffle=False, num_workers=0)" ] }, { "cell_type": "markdown", "id": "4415f14b-4d59-4628-a2f1-d81e2d3d414f", "metadata": {}, "source": [ "### Define the data augmenter\n", "\n", "Now we need to define an augmenter. To keep things simple, we define an augmenter which combines:\n", "\n", "1. the addition of some noise or channel lost from ``add_band_noise`` or ``masking``\n", "2. the ``warp`` or ``crop_and_resize`` augmentation\n", "3. a final rescale of the range [-500, 500] uV in [-1, 1] with soft clipping with horizontal asintote of 1.5\n", "\n", "This is similar to the augmentation proposed in the augmentation module introductory book" ] }, { "cell_type": "code", "execution_count": 5, "id": "f42f5a53-01b8-4e2f-9b95-9eb0b719030b", "metadata": {}, "outputs": [], "source": [ "# DEFINE AUGMENTER\n", "# First block: noise addition\n", "AUG_band = aug.DynamicSingleAug(\n", " aug.add_band_noise, \n", " discrete_arg={\n", " 'bandwidth': [\"delta\", \"theta\", \"alpha\", \"beta\", (30,49) ], \n", " 'samplerate': freq,\n", " 'noise_range': 0.5\n", " }\n", ")\n", "AUG_mask = aug.DynamicSingleAug(\n", " aug.masking,\n", " discrete_arg = {'mask_number': [1,2,3,4], 'masked_ratio': 0.25}\n", ")\n", "Block1 = aug.RandomAug( AUG_band, AUG_mask, p=[0.7, 0.3])\n", "\n", "# second block: warp or crop and resize\n", "AUG_crop = aug.DynamicSingleAug(\n", " aug.crop_and_resize,\n", " discrete_arg={'batch_equal': False},\n", " range_arg ={'N_cut': [1, 4], 'segments': [10,15]},\n", " range_type =[True, True]\n", ")\n", "AUG_warp = aug.DynamicSingleAug(\n", " aug.warp_signal,\n", " discrete_arg = {'batch_equal': [True, False]},\n", " range_arg= {'segments': [5,10], \n", " 'stretch_strength': [1.75,2.25],\n", " 'squeeze_strength': [0.45,0.55]},\n", " range_type=[True, False, False]\n", ")\n", "Block2 = aug.RandomAug( AUG_crop, AUG_warp)\n", "\n", "# third block: rescale\n", "Block3 = lambda x: selfeeg.utils.scale_range_soft_clip(x, 500, 1.2, 'uV', True)\n", "\n", "# FINAL AUGMENTER: SEQUENCE OF THE THREE RANDOM LISTS\n", "Augmenter = aug.SequentialAug(Block1, Block2, Block3)" ] }, { "cell_type": "code", "execution_count": 6, "id": "c89c01e0-1701-4c02-b908-b523fba80620", "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABdEAAAJOCAYAAABYwk4SAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOydd3wUZf7HP7vpvRd6T+hFEcGOBVFOxV5Bz8JhOXvXu7Oeet5Pz97FgoKAiIigIB2kKh1CIKT3ZLNJNsn2+f0xOzPPzM5sSTbZbPi+Xy9e7PQnuzuz83zm83y+Oo7jOBAEQRAEQRAEQRAEQRAEQRAE4YY+2A0gCIIgCIIgCIIgCIIgCIIgiO4KiegEQRAEQRAEQRAEQRAEQRAEoQGJ6ARBEARBEARBEARBEARBEAShAYnoBEEQBEEQBEEQBEEQBEEQBKEBiegEQRAEQRAEQRAEQRAEQRAEoQGJ6ARBEARBEARBEARBEARBEAShAYnoBEEQBEEQBEEQBEEQBEEQBKEBiegEQRAEQRAEQRAEQRAEQRAEoQGJ6ARBEARBEARBEAThIw6HI9hNIAiCIAiiiyERnSAIIgRxOBz46aef8OCDD+KCCy7AhAkTMGHCBJx//vmYO3cuvvnmG5hMpmA3k+gEli5ditzcXOTm5ga7KUFHeB/efPPNYDclZJg1axZyc3Nx4403+r2t1WrFu+++i+nTp2PcuHGYOHEiLrnkEjQ1NXVCSwmia3nnnXfEa4rdbg92c4huzLZt23DXXXcF5djCNfzJJ58MyvG7C+y9UHFxcbCbQxAEQZwkhAe7AQRBEIR/HDt2DA899BCOHTvmtqy1tRXl5eVYv3493n//fTz33HO46KKLgtBKgiB6Gs8++yx+/PFHcdpsNiMiIgKJiYlBbBVBEETXsXjxYjz77LPo06dPsJtCEARBEEQXQyI6QRBECFFTU4Pbb78dNTU1SE9Px1133YUpU6YgMzMTOp0O1dXV+P333/HJJ5+grq4ODz74ID744AOcc845wW46QRAhjNlsxs8//wwAOOuss/DMM88gLS0NZrM5yC0jCILoOqqqqoLdBAJAfHw8+vfvDwCIiIgIcmsIgiCIkwUS0QmCIEKITz75BDU1NUhKSsLixYvRu3dv2fLk5GTk5ubiggsuwNVXX42mpia8/PLLOOuss6DXU4IXQRDtw2g0ihEXN998MwYPHgwASEpKCmazCIIgiJOQadOmYdq0acFuBkEQBHGSQYoKQRBECLFu3ToAwCWXXOImoLP0798fc+fOBQAUFRXh0KFDXdI+giB6JmxGdFxcXBBbQhAEQRAEQRAE0fWQE50gCCKEqK2tBQDYbDav65577rn48ccfkZKSolkk7dixY/juu++wa9cuVFZWoqWlBfHx8RgwYADOO+883HzzzW5O07KyMlxwwQUAgLVr14LjOHz00UfYvHkzDAYD0tPTMXXqVNx7771IS0sDAPz4449YsGAB8vPzwXEccnNz8de//hUXX3yxZvuPHz+OL7/8Etu2bUNNTQ0iIyMxcOBATJs2DbfccgtiY2N9es9YhGKc8+bNg9FoxNtvv42ysjKkpqZixowZeOKJJwL23qxevRoxMTH4+OOPsXHjRlRVVSEuLg7jxo3DzTff7DFip7y8HPPmzcPWrVtRUVGBxMREnHPOObj33nt9+js3bNiAxYsXY9++fTAajYiLi0NOTg4uvfRSXHPNNapDn2fNmoWdO3fivvvuw9y5czF//nwsW7YMxcXFiIqKwqhRozBnzhxMnjwZAP/5fPjhh9i+fTuMRiOys7Mxbdo03Hfffe36bIS/e9GiRVi/fj3Ky8vhcDjQr18/TJ06FbfffjuSk5M1tzWZTPjss8+wevVqlJWVITo6Gjk5Objuuutw2WWXeTzmwoULsW3bNpSVlaG5uRmxsbHo06cPzjrrLMyaNQtZWVlu27HfpTFjxuDTTz/FmjVrUF5ejsjISIwYMQLXX389ZsyY4bbtO++8g3fffRennHIKFixYgC1btmD+/PnYv38/mpubkZmZifPOOw9z5sxRPbZAZ5wjSnbs2IHZs2fL5rHTa9euRd++ff06t+rq6vD1119j48aNKC4uhsPhQHZ2NiZPnozZs2dj6NChHttx9OhRHDlyBJ988gl27tyJpqYmZGdnY/r06Zg7dy5iY2PhcDjw7bffYunSpSgsLERYWBjGjBmDuXPnit9hf7FYLPj++++xatUq5Ofno6WlBSkpKZgwYQKuvvpqnHvuuarbdeS7ImCz2bBkyRKsXLlSPHZaWhomTpyIWbNmYfz48e36m8xmMxYsWIBff/0Vx44dg8ViQXJyMkaPHo0ZM2ZgxowZmqOYOnrufPXVVxg2bBg++ugjrF27FtXV1UhJScHkyZNx3333iVENGzduxBdffIFDhw7BYrFg0KBBuOGGG3DDDTe47fv8889HeXk5nnvuOVxxxRX48MMPsXLlStTU1CAjIwPjx4/HnXfeiREjRrTr/dqzZw+++eYb7N69G/X19YiJiUFOTg5mzJiheW1lfxeuvPJKvPrqq34f12AwYOHChdi6dSsKCwvR1NSEqKgoZGVlYcqUKZg1axYGDhzotp3wXs+dOxcPPfSQ6r6Fa79wPVJiNBoxf/58rFmzBiUlJYiMjMS4ceNwxx134PTTT8eYMWNgtVrx1Vdf4fTTT3c7dmd8zgIVFRX44osvsHnzZlRWVkKn04m/GbfddhtSUlLctmnv9Xfp0qV46qmnxOny8nLZ38j+7RzH4eeff8aPP/6IQ4cOoampCcnJyRg3bhyuu+46zWuFwKZNmzB//nwcOXIETU1N6N+/P66++mrccsstHrdTo6amBueddx4cDgcefPBB3H333Zrr3nrrrdi+fTsmTZqEr7/+WrasPe/1k08+iR9++AGXXXYZHnroITz33HPYtWsXIiIiMGTIELzzzjvIyMiAw+HA0qVLsWLFCuTl5aGlpQWJiYnIycnBtGnTcM011yAyMlK2b/bzWL16NQYMGOB2/I7cC82dOxcPPPAAFi1ahB9++AEFBQWw2+0YMGAAZsyYgdmzZyM6Otqnz4AgCILoOZCIThAEEUL06dMHJ06cwKpVqzB79mwMHz5cc92hQ4di+fLlmsvfffddvPvuu+A4TjbfaDTCaDRi3759WLx4Mb755htN1/vu3bvx4osvwmQyifMqKirwzTffYPPmzVi6dClefPFFWTFCgBci9uzZg3/+85+4+eab3fY7b948vP7663A4HOI8i8WCAwcO4MCBA1i4cCE+/vhjVbHNF1avXi0TC6qrq2UCbSDem0OHDuH555+H0WgU51mtVmzYsAEbNmzA/fffryqKb9q0CQ888ABaW1vFeWazGUuWLMEvv/ziURBua2vDI488grVr17q1e+fOndi5cye+/fZbfPjhh5pF0dra2jB79mz8+eefsnlbt27F9u3b8e677wIAHnroIVkedmlpKT777DP8+eefmD9/PsLD/bvFWLVqFZ5++mnZ3w0A+fn5yM/Px9KlS/H5558jJyfHbdvi4mJcfvnlKC8vF+eZzWbxb/7jjz/w3HPPuW23ePFiPP/8824PpZqamtDU1IQjR45g0aJFmDdvHkaNGqXa7vLycjzzzDOoqKiQHXvHjh3YsWMHtm7din//+9+af/fbb7+N9957TzavrKwM8+fPx/Lly/H111+rnuedfY60F2/n1oYNG/DYY4+hqalJtl1xcTGKi4uxZMkSPPLII7jjjjs0j7Fy5Uo8/vjjss+tuLgYH330EXbu3InPP/8c99xzD7Zt2ybbbtu2bdi5cyfee+89TJ061a+/q6ioCPfeey+OHz8um19TU4Nff/0Vv/76Ky699FK8+uqriIqKUt1He78rlZWVmDNnDvLz82Xzq6qqsGLFCqxYsQJ33XUXHnnkEeh0Op//psbGRsyePRt5eXmy+bW1tVi/fj3Wr1+PJUuW4KOPPnITiwJx7pw4cQKPPPKI+HAY4L8vP/74IzZv3owlS5Zg0aJF+PDDD2XbHTlyBP/6179QWlqKxx57THXfzc3NuOGGG3D06FFxXllZGcrKyrBy5Uo899xzuP76672/SS6cTif+85//YN68ebL5VqsVu3btwq5du8S2enrw1R42btyIBx980O3aaLPZYDKZUFBQgMWLF+Pdd98NeP2TgoIC3H777bIc8NbWVmzcuBGbNm3C448/7nUfnfU5//zzz3jqqadgsVhk848ePYqjR49i4cKFeO+99zBx4kTNtrX3+uuJpqYm3HfffdixY4dsfm1tLX777Tf89ttvuPzyy/Hyyy+7CcMOhwPPP/88vvvuO9n8/Px8vPLKK1i3bp3b3+uNzMxMTJkyBVu2bMHKlSs1RfSamhrs3LkTAHD55ZfLlnX0vW5sbMSsWbPE3+i2tjYYjUZkZGTAarVizpw5btfr+vp6bNu2Ddu2bcOCBQvw5ZdfIjU11ae/ORD3QjabDXfeeSe2bt0qm5+Xl4e8vDysXLkS8+fPR3x8vE9tIgiCIHoIHEEQBBEyfPTRR1xOTg6Xk5PDjRo1inv44Ye51atXc42NjX7tZ9WqVeJ+/vrXv3Lbtm3jqqqquKqqKm779u3cnDlzxOWPPfaYbNvS0lJx2ZgxY7hJkyZxS5Ys4aqqqrhjx45xDz74oLj84osv5nJycrgnnniCO3LkCGcwGLgNGzZw5513HpeTk8NNnDiRa2trk+1/0aJF4va33HILt2XLFq6+vp6rqKjgFi1axJ155plcTk4Od84553D19fV+/d3CfnNycrgZM2Zwf/zxB1dTU8MtX76cq66uDuh7M27cOG7ixInc119/zZWVlXE1NTXcjz/+yE2aNInLycnhRowYwRUVFcm2P3HiBDdq1CguJyeHmzp1Kvfzzz9z9fX1XElJCff+++9zo0ePlv0NSubOnSsuu//++7k9e/ZwDQ0N3PHjx7n//ve/3MiRI7mcnBxu2rRpXHNzs2zbW265RfxMc3JyuH/84x9cfn4+V1VVxX333Xfisc855xxu/Pjx3PTp07m1a9dy9fX13LFjx7j7779fPPaKFSv8+lz++OMPbvjw4VxOTg53wQUXcD/99BNXXV3NlZWVcV999RU3YcIELicnh7vwwgs5i8Wi+nmOGDGCe+2117hjx45x1dXV3Jo1a7ipU6eKy7dv3y475r59+7jc3FwuJyeHu+qqq7j169dz5eXlXG1tLffnn39yjz/+uLjtTTfdpPldGjduHDdmzBjuvffe44qKirj6+npu7dq13Pnnny+us23bNtm2b7/9trhtTk4Od/3113ObN2/mDAYDV1BQwL344ovitjfffLPbsTtyjgif8w033ODTZ2O32zmTycQdO3ZMPOaGDRs4k8nEmUwmzul0un0WWufWn3/+KX6PpkyZwi1cuJCrqKgQ37PLL79c3Me3334ra8f27dtl152pU6dyq1at4mpqarjDhw+Lf5dw3cnNzeVeffVVrqCggKuvr+dWrFjBnXbaaeJyfzAYDOLnOWrUKO6NN97gjh8/zjU0NHB//vkn9/e//1089kMPPeS2fUe+Ky0tLdz06dO5nJwcbvz48dxHH33EFRYWcg0NDdzevXu5hx9+WNz2o48+8uvv+sc//iHud/78+VxxcTFnMBi4w4cPy77/H374oWy7QJ07Y8aM4caOHct9/vnnXFlZGVdUVMS99NJLbr8fc+bM4fbt28c1NDRwu3bt4i677DLxnK+srJTtWzjnx48fL14HDx8+zNXX13MbNmwQ95mTk8Nt3bpVtq1wXubk5HA2m0227M033xSX3Xfffdzu3bu5hoYGrqSkhJs3bx536qmncjk5OdwVV1zBmc1m2bZVVVXcxRdfzF188cXcf//7X78+o/LycvE6MW3aNO7nn3/mSkpKuPr6eu7AgQPcSy+9JF7XzzvvPPF8VL7Xb7zxhuYxtK4JLS0t3Lnnnsvl5ORwEyZM4L788kuusrKSq6qq4hYsWCD+lmldYzvzc966dav4m3H55Zdza9as4erq6rjq6mru559/5qZNmya2u7CwULZte6+/NpuNM5lM3P/93/+J77dwHbTb7RzHcZzD4RDfz5EjR3Kvv/46l5+fzzU0NHBHjhzhXnzxRfHc+ec//+n2WbDfwXvvvZfbv38/ZzAYuD179sjuPYT7Kl/58ccfxe3y8/NV15k3b574eTU1NQXkvX7iiSfE444aNYr79ttvudraWm7v3r3chg0bOI7juA8//FD8nN9//33u+PHjnMFg4I4fP869+uqr4vb/+Mc/ZPv+/vvvxWXKe6lA3AsJ34/HH3+cO3jwIGcwGLi9e/dyt956q7jvt956y+fPgCAIgugZkIhOEAQRQlitVu7222+XdaRycnK44cOHc5dffjn3wgsvcL/++qtbp0DJ1VdfLYpdrCgp4HA4uCuvvJLLycnhJk+eLFvGCsXDhw/nDhw4IFve1tYm61w/88wzbvvfsGGDuHzXrl3i/KamJlEwveeeeziHw+G2bXl5uShYvPjiix7/TiVsu5WdPYFAvTcjR47kDh065Lb9unXrxHU+++wz2TKh4zdp0iSuoqLCbds1a9bIPnet/b788suqfxv7gOC1116TLWNFyJdeeslt2+eee05cfuaZZ7qJs1arlTv77LO5nJwc7qmnnlI9vhaCWHLeeeepPhhh/+6lS5eK89n34ocffnDbLi8vT1yu/K4Iov/kyZM5o9Go2q577rmHy8nJ4XJzczmTySRbxh577dq1btsePXpUXP7CCy/IlrFCybXXXstZrVa37QVhNjc3V/aedPQc8VdEF2C/20qxjON8O7euuuoq8ftdUlLitry1tZW79tprRQGD/btZEX38+PFu50dVVZUojOTk5HDvv/++2/6/+eYbcbna+aUFK+SsWbNGdZ0XXnhBXEcQhwQ68l156623RAFq7969qsd++eWXuZycHG706NFcTU2Nz3/X6aefzuXk5HDvvfee6vLZs2dzOTk53F/+8hfZ/ECeO2rv54wZM8Tlf/3rX92E4ePHj2ue9+yDsyeffNJt3waDQXyIq/y7tET0oqIiUUTU+s05cOAAN2LECC4nJ4ebN2+e6jrt4bXXXhM/f7VzhuOkzz8nJ4c7duyYbJkwvz0iuvDdGzFihOx3WuDIkSPiQ1e160Jnfc52u1188HTNNde4PbTgOI4zGo3iOnPnzpUt68j1l91+6tSpbtstWbJE3Pevv/7qtpzjOO6LL74Q1zl48KA4v6qqihs7diyXk5PDPfDAA27vh8Ph4O677z5xW39E9NbWVvHB0ptvvqm6jnDv88ADD4jzOvpesyK61nGF34Wnn35adflTTz3F5eTwpgv2905LRA/kvdArr7zitq3ZbBbvdZTXEIIgCKLnQ4VFCYIgQoiIiAh8/PHHePTRR5GYmCjOdzqdyMvLw/z58/H3v/8dkydPxkMPPYSSkhK3fTidTpx33nmYOXMm7rnnHrfhxACg1+vFYbkNDQ2a7TnjjDMwevRo2bzo6GjZvDvvvNNtuwkTJoivq6urxdfLly9HS0sLAD5LUy2Lt3fv3mIu6NKlSzXz3j2Rm5urmh8byPfmrLPOwsiRI93mn3nmmWLUSVlZmTi/ubkZmzZtAsBnTvfq1ctt2wsvvBBnnXWW6vEWLVoEAEhLS8Ojjz6qus706dNx/vnni+uzUSAsc+bMcZvHDtO++uqr3YZVR0REiJ87+5l649ixY2Lkwr333qs6XPvCCy/EaaedhtNPPx1Op9Nt+ZAhQzBz5ky3+bm5uWJOamlpqWzZKaecgmuuuQb33nuvW7a9wKRJkwDw+baNjY2q6wwbNkx8T1lycnLEqB/2c1Zy++23q+ayCpm5HMfJYmq66hxpL1rn1uHDh3Hw4EEAwD333IN+/fq5rRMTEyPG7rS1tWHZsmWqx7jsssvczo+srCxxnxEREbj11lvdttO67njC6XTi+++/B8B/Dy+88ELV9Z544gnxu6uWKw34/13hOE6MdZgxYwbGjRunut/7778f0dHRsFqt+OGHH3z6uwCI0Qz19fWqy5999ll8+eWXbjEbgTp3Bg0apPp+nnLKKeLrO+64wy2iZsiQIUhISADAR1CoER8fj2eeecZtfkpKCv7+978D4CMylFE2anz33XdwOp2IiYnRzBUXMuSF9QNFTk4Orr/+etx1112q5wwgvdeA598kfxHOv0svvVQ1pmP48OGqcWxKAv05b9myRTxPHnnkEdX4pKSkJLG4+vr16zW/J/5ef70hnPunnXYapk2bprrOLbfcIkaICL/bAPDbb7/BbDZDp9Phsccec3s/9Ho9nn32WYSFhfncHoGYmBixBs2qVavclhcXF+PAgQMAgCuuuEKcH8j3evr06arzrVYrAO3r0D333IPPPvsMS5Ys8SmuKlD3QjqdTvVeKCoqClOmTAHg+bedIAiC6JlQJjpBEESIERYWhrvuuguzZs3Cpk2bsGHDBmzfvl3W0bPZbFi5ciV+++03vPzyy7J8S71ej/vuu09z/06nE8ePHxc7BxzHweFwqHbcxo4dq7oPoaBobGysqqDGZkiyGZtCHmdKSgpSU1NFsVDJmDFjAAAtLS3Iy8tzE/K9oVVUrivem8jISCQkJKChoQFtbW3i/F27dolip6dc24suughbtmxxm79r1y4AwNSpU1XFf4Hp06dj3bp1aG5uRl5enlteca9evZCRkeG2HStuqz0cAKTPVegU+wKbg+opp3r+/PmayzwVVUxPT0dxcbHbd0lNZGUpKipCQUGBOK0lRGt9zgD/nlVUVMg+ZyVawqhwDgGQZc931TnSXrTOLaHdADwWFB45ciT69++PkpIS7Ny5E7fffrvbOp6uO4WFhRg0aJBqUVWt644njh49KorAWqIYwJ/X559/PpYsWYJdu3aB4zg3wcff70pBQQHq6uoA8O+r1met0+mQm5uLffv2yWoZeOO0007Dxo0bMX/+fFRXV+PSSy/FmWeeKQrjw4YNw7Bhw9y26+xzh/3ue7rWNDc3a36O5557rmZWMXud2bZtm9fMa+G7O3jwYADQ/BzGjh2L5cuX48SJE2hoaFAttOgvM2fOVH1AKFBZWYnDhw+L04F6YHbixAnxnkLtwY/AxRdfjM8//9zjvgL9ObNZ4zk5OZqfh3DN4zgOe/bsUb3u+Hv99YTJZBI/i5EjR2q2C+Cvz+Xl5bLzdfv27QD4ejZaOd1ZWVkYM2YM9u7d61ObWK644gr88MMPKCoqwsGDB2W/CT/99BMA/jp09tlni/MD9V6Hh4erXksA/gF9fn4+1q9fj9mzZ2PmzJk4++yzxfuQvn37om/fvj7/nYG6F+rbt69mBrsw39NvO0EQBNEzIRGdIAgiRImOjsa0adNEYaeyshI7duzA5s2bsW7dOrS2tsJqteLJJ59E3759Za4vgcbGRmzevBnHjx9HSUkJiouLceLECbcCZpyiwKaAlkgguGO1RAw19ywguYUbGhpU26tGVVWV3wIhW+hQi856bwCIHTt228rKSvF1//79NbcVhBwWk8mE5uZmALx7zxPs8srKSreOo7fPFPD/c/WE4AiOj4+XCRf+4OnzFB5waLnu29rasHnzZuTn56O4uBilpaU4fvy4+H4KBOpz9nV7tvPPuu+76hxpL1qfhVCYMDY2FtnZ2R73MWTIEJSUlMjOCRZv39G4uDjV5f4U3RRg2+DruSWcj+xoIcD/7wo7kuiVV17BK6+84ld7vfHEE09g3759MBqNWLNmDdasWYOwsDCMGTMGZ599Ni666CLk5uZqbt9Z504grjU5KgWI2eMmJSWhsbFRVjBTC+GcO3TokF/nXCBEdAGbzYbff/8deXl5KCoqEt9rpfPc07XGH9iH8moPwgXUfo+UBPpzZkcVCY5gb/h7LdG6/nqivLxc/J358ssv8eWXX/rVLuG11ogDgcGDB7dLRD/99NORnZ2Nqqoq/Pzzz7LfhJ9//hkAP+KFLQoeqPc6Pj5e00F/7733YvPmzSgtLRWLLAsPBs855xxccMEFHh+Us3TFvRDg2287QRAE0TMhEZ0gCKKH0KtXL9G1ZjQa8dprr2Hp0qVwOBz48MMP8fHHH4vrWq1WvPnmm1i4cKGbKBwVFSXGZgiOHi1iYmI8LvdXtDKZTH6t395tPLmTAvXeqA0R9wQrPHl6X4Xh7SysO0zNgcvC7lvNVRYdHe1xe6B9YqQWgsvX23fJE/6+1wAvjHzyySf4/PPPYTQa3fY3YcIEJCYmYuPGjR73wwoO7cHf7bvqHGkvWueW0AYtgZtF+C4ozz/lci0C+f1k3zt/zq3W1lY3Eb27fdZDhgzBihUr8NFHH2HlypWor6+Hw+HA3r17sXfvXrzzzjuYOHEiXnzxRZlYGqhzpzOvNWrXSeWxGxsb3QR/NYJ9zi1evBjvvfeemzip1+sxYsQIDBw4UDWioyOwn6unz8nbOeFtewF/PudAfh4dvX77cgxftxG+i96ub1oPHLyh1+tx2WWX4ZNPPsEvv/yCxx9/HDqdDocPH8aJEycAQDZqUdk+X1HbRi0GRiA9PR3Lli3DJ598guXLl6OiogIcxyEvLw95eXn4+OOPkZOTg5deeklz5IBAIO+FAvndIAiCIHoO9OtAEAQRIqxatQqHDh1CXFwc7r77bo/rJicn45VXXsGxY8dw4MAB7Nu3T7b8kUcewerVqwHwMQHnnXcecnJyMHToUAwePBjh4eF48803vQrFgUbobI8fPz6gubL+EKz3hs0Wbm1t1RQj1aJS2M6ilvAo4E8nsysQPvOuHhb9n//8B/PmzQPAOy0vuOACDB8+HEOGDMGwYcMQGRmJxYsXexUCu5rucI60B+G75iniQEAQYTryYCVQ+HNuseJRINrO7uPTTz+VxSwEioyMDDz77LN4+umnceDAAWzduhW///479uzZA7vdjt27d+O2227DqlWrxAcgoXDueIvrET5LX9zi0dHRMJlMmDFjBt54442AtM9Xvv76a7z00ksA+CiPadOmYcSIERg6dCiGDRuG2NhY/P777x0S0dXiStjvnqdrczDiLIRrYEZGhmq0WbBg37Pnn38eN9xwg1/bC/cA3q4z/sSlKbniiivwySefoKKiAnv27MEpp5wiRrkMGjTILXqnq97r+Ph4PPTQQ3jooYeQl5cnXod27doFi8WC/Px8/PWvf8WKFSvE+hFqhPK9EEEQBBEakIhOEAQRIqxatQq//vorYmNjcccdd3h0UwucdtppOHDggKzTtWfPHlEkvvnmm/HPf/5TddtAFijzld69e+Po0aNeizWp5Q0HgmC+N2zH8MSJE5qxAcoCmQDfAU1MTERTU5Msi1gNdrmnzmhXIbTBZDLBYDBoZpCuXLkShYWFGDZsmMdsal+oqqoSh9qff/75ePfdd1WHmgfjHPBGsM+R9iJ8zq2traiqqvIY6SJ8R7vD95PNJi4oKPAYiyO4OePi4jQLbvoDWzy1sz9vvV6PcePGYdy4cbjnnntgMBjw2muvYdmyZaiursbq1atx5ZVXhsy5o1ZUW6Curk50/WplT7P07t0b+fn5XX7Omc1mvPXWWwD4DO2vv/5a9eGMp/dar9fD6XR6zEpXjiYA5JFixcXFblEX7LKuRrguGAwGtLa2dhsBlL2mtee70qtXL+zbtw+FhYUet+1IMcthw4Zh5MiROHz4MFavXo0JEybgl19+ASAvKCoQjPd6+PDhGD58OO644w60tLTggw8+wCeffIKWlhYsXbrUY92aUL4XIgiCIEID/8NLCYIgiKAgiKqtra1YunSpT9sIQgKb/bhnzx7x9Y033qi6ndPplBWU8jUTtKOceuqpAHiRQ+meZ/noo48wceJEXH755R7FEn8J5ntz2mmnia6v3377TXO9TZs2uc3T6XTie7d+/XqPTrVff/0VAO+88pQb3FWwDws2b96sud68efPw9ttv49tvv+3wMfft2yd+btdff71mVitb9LSrzgFvBPscaS8TJ04UXwuijRoHDx4U85h9zZ/uTHJycsRYFuHcUcNqtWL9+vUAgAkTJgTk2MOHDxfd32vXrtVcr6WlBWeeeSamTp2K//73vz7te//+/bjxxhsxadIkHDt2zG15amoqnn76aXFaqF0QKufOli1bNPOK2ffy3HPP9bov4Zw7dOiQxwz1f/7znzj99NNx9dVXByTOhc2Xv/LKKzVHN3h6r4VttIT2+vp61QezQ4cOFWtUqP3mCGzYsEH7D+gkhM/D4XB4PP5PP/2ECRMmYMaMGdi9e3fAjq/1oCQ1NVW811q3bp3m98/pdGLGjBk4++yz8dhjj4nzhZEmRUVFquckwD9s9qd4sBqCWL527Vrs3bsXFRUV0Ol0uOyyy9zW7ez3ury8HLNmzcIZZ5yhOnIlLi4Ojz76qBjPJFyHtAjleyGCIAgiNCARnSAIIkSYOXOm6G589dVXvQ6t3bhxI9atWwdALgizhbqOHz+uuu27776LoqIicdpms7W32X4xc+ZM0WH/4osvqg4VLykpwbx589Dc3Ayr1eq1CJc/BPO9iY2NxSWXXAIA+Oabb3D06FG3dfbu3SsWAFNy3XXXAeBFES0h7bfffhOFviuvvLJdWeKBZvz48Rg6dCgA4L333lPNKN64cSP2798PgC981lFY4U/rc/7+++/x+++/i9NddQ54I9jnSHsZPXq06Gb94IMPVIU7s9mMF198EQCfoavM5w0GYWFhuPrqqwHwopPWA67XX39dFCqvvfbagBw7PDwcV111FQD+AZNWZMf//vc/1NfXo6KiAsOHD/dp37169cL+/fvR2NiIr7/+WnWdI0eOiK8FZ3KonDulpaX46quv3ObX1dXh3XffBcAXWvTFiS5cW+12O55//nnVIsX79u3DDz/8AKPRiOTk5HbnVrP48l5v3bpV9lBd+V4PGDAAAC+Eq10r3nrrLVWxV6/X45prrgEArFixAgcOHHBbp7S0FF988YX3PyTAXHDBBUhPTwcA/Pe//4XBYHBbx2Aw4O2330Zrayvq6uowYsSIgB1f+FzUBFrhu1JQUIDPPvtMdfuvvvoKBQUFqKmpEX/7AOCiiy4SH9i9+OKLqufNm2++6TWmxBt/+ctfEB4ejpKSEnz00UcA+Iecffv2dVu3s9/rjIwMHD16FPX19fj6669Vv4ulpaXiQylPRdcFQvVeiCAIgggNSEQnCIIIEZKTk/HGG28gIiICbW1tuOOOO3DPPfdg5cqVKCoqQmNjIyoqKrBx40Y8/vjjuPvuu+F0OnHeeefJhumeeeaZopPqxRdfxPLly1FVVYXq6mps3rwZc+fOxXvvvSc7dkc7bb6SkZGB+++/HwBw4MABXHfddVi9ejXq6upQUVGBZcuWYdasWTAajdDpdHjmmWcCOnw+2O/NY489huTkZJjNZsyePRvfffcdampqUF1djW+++QZ33HGHprvt/PPPx/nnnw8A+PLLL/Hggw9i3759aGxsREFBAd588008+OCDAIB+/frh4YcfDli7O8qzzz4LvV6P4uJi3HDDDfjtt99gMBhQXFyML774QmxrTk4OZs6c2eHjnXrqqaLr/91338U333yDsrIy1NXVYdeuXXjiiSdkLlyg684BbwT7HOkI//rXvxAREQGj0YgbbrgBixYtQlVVFQwGAzZs2ICbbroJe/fuBQA88cQTHiNfupK7775bFFsffPBBvPnmmygoKEBjYyP27duHBx98UBRsL774YkyfPj1gx7733nvFWJdHHnkEr732GvLz89HQ0ICDBw/iiSeeEI996qmn4tJLL/VpvxkZGeJDiu+++w5PPfUU9u/fD4PBgNLSUnz//fd49NFHAfCC+4UXXigeI1TOnVdffRWvv/46ioqKYDAYsGbNGtxwww2oqalBRESEZlyXkpEjR4oPotetW4fZs2djy5YtMBgMKCkpwfz583HXXXfBZrMhKipK5i4GePfs9OnTMX36dPzf//2fz+3PyclBRkYGAGDhwoV4//33UVxcDIPBgP379+Oll17CnDlzZKK+8r2+4IILAAC1tbWYM2cO9u3bB4PBgD179uD+++/Hd999pxk9NGfOHGRlZcFms+H222/HN998g6qqKtTW1mLZsmW46aabZA89u+o6ExUVhWeeeQYA72S+5pprxNih6upqrFmzBrNmzRJH4DzyyCM+FTT2leTkZAD8A5kNGzbAaDSKGfw33XQTRo4cCYB/sPb000/j4MGDMBqNOHr0KF599VW8+uqrAPh6ArNmzRL3m5SUhMcffxwAsGPHDtx2223YtWsXjEYj8vLy8Pjjj2P+/Pmaoz98JT09HWeccQYAiEKyWpQL0PnvdWRkJGbPng2Af1B49913Y9euXairq0N5eTl++eUX3HnnneA4DrGxseJDRU+E8r0QQRAE0f2hTHSCIIgQ4qyzzsInn3yCf/3rXyguLsbatWs1h/nrdDpcc801ePbZZxEeLl3uhw0bhrvuugsff/wx6uvr3Tr8AJCQkIBrr70Wn3/+OQA+91TozHc2d955J1pbW/HBBx8gPz8ff//7393WiYiIwL/+9a+AF9oL9nuTlpaGefPmYc6cOaitrXUTeaKiovDggw9qCjH//e9/8eijj2LdunVYtWqVqnN11KhReOuttwLilAwUU6ZMwSuvvIJnn30Wx48fx7333uu2zuDBg/HRRx8FxDGWkpKCJ598Es8//zza2trwwgsvuK0TGRmJO+64Ax988AEAfoj9+PHjO3zsQBDMc6QjjBs3Du+++y4efvhh1NXV4R//+IfbOuHh4Xj44Ydx8803B6GF6iQlJeHzzz/H3LlzUVhYiA8//BAffvih23qXX345nn/++YAeOyUlBZ9//jnuvvtuFBUV4fPPPxevPSxjx47Fu+++KxtN442nn34aJ06cwN69e7F06VLVmLD09HR8+OGH4uiHUDl3Jk6ciPLycnz66af49NNPZcsSEhLwzjvvyFzA3njmmWdgs9mwZMkS7N69G3fccYfbOnFxcXjjjTfcRgPYbDYx57q2ttbnY4aFheHFF1/EfffdB7vdjrfeekvMSBfQ6/WYM2cOvvrqK5jNZtkoKQC47bbbsHbtWhw+fBg7d+4UXboCF110EXJzc0V3Pkt8fDw+/vhj3HbbbWhoaMALL7wg+7z1ej2uv/56scBxR8Vdf7j00kvR1NSEl156CeXl5XjiiSfc1tHpdLj33nvd/uaOctpppyEsLAwOhwN/+9vfAACvvPIKrrrqKkRGRuLjjz/GPffcg/379+P777/H999/77aPgQMH4pNPPnHLGL/22mthMBjw5ptvYvfu3bjllltky0eNGoWhQ4fixx9/7NDfMHPmTDGmJyoqyuODv85+r//2t7/h8OHDWLt2LdavXy8K+yxxcXF45513xIghb4TqvRBBEATR/SERnSAIIsSYMmUKVqxYgTVr1mDz5s04cOAADAYDmpubERcXh+zsbEyZMgWXXXaZZjGwRx55BKNGjcKCBQtw+PBhtLS0IDY2Fv3798fZZ5+Nm266CcnJyVi0aBFMJhNWr14tyzTuTHQ6HR544AFcfPHFmD9/Pnbu3Inq6mo4nU707t0bkydPxuzZs2U574Ek2O/NyJEjsXz5csyfPx9r1qxBaWkpYmJicNppp+Hee+9VHZIvEBcXhw8++ABr167F999/j/3798NoNCIlJQVDhw7FFVdcgUsvvdSnorRdzcyZM3HKKafgiy++wNatW1FZWQm9Xo/BgwfjkksuwS233KKZCdwebrzxRgwcOBBffPEF9u3bh6amJkRHR6NPnz44/fTTccstt2DgwIFYtWoVioqKsGbNmoC44ANBsM+RjnDeeedh9erV+Oqrr7BhwwYx1qV3794444wzcMMNN3TLdg8cOBDLly/HokWL8Msvv+DYsWNobW1FVlYWxo4di2uvvRZTpkzplGMPHjwYy5cvx+LFi/Hrr78iPz8fJpMJ8fHxyM3NxWWXXYarrrrKbxEzISEB33zzDRYvXoxVq1YhPz9f/B3p378/pk6ditmzZ4t5xAKhcO706tUL7777Lt577z2sWbMGDQ0N6NOnD6ZOnYpbb70VWVlZfu0vIiICL7/8MmbOnImFCxdiz549qK2thV6vR79+/XD22Wfj1ltvDfjoialTp+K7777Dp59+it27d6OhoQGRkZHIzs7GKaecgptuugmjRo3CoUOHsHXrVqxZs0b2EDI+Ph4LFy7EV199JRZnDgsLQ05ODq699lpcddVVeP/99zWPP3z4cKxcuRKffPIJ1q1bh8rKSkRHR+OUU07B3Llz4XQ6RRE9KioqoH+7N2644QaceeaZ+PLLL7Ft2zZUVFTAZrMhMzMTEydOxC233IKxY8cG/LhDhw7FG2+8gffeew/FxcWIjo6WZc5nZGRg4cKF+Omnn7BixQocPnwYjY2NiI6OxrBhw3DxxRfjxhtvFEd0KPnb3/6GKVOm4PPPP8e+fftQV1eHXr164dJLL8XcuXPx73//u8N/wwUXXID4+HiYTCZMnTrV7RxX0pnvdUREBN577z38/PPP+PHHH3H48GEYjUZER0ejd+/eOOecc3DrrbciMzPT532G8r0QQRAE0b3RcVrjwgmCIAiCIAiCIEKE888/H+Xl5bjssst8LrJKtJ9NmzbhrrvuAsAXGRWihwiCIAiCIHoi5EQnCIIgCIIgCIIgRP7xj38gPj4e06ZNw4QJE1TXOXjwIAC+MLa/7n6CIAiCIIhQg0R0giAIgiAIgiAIQqSoqAg7d+7EgQMHMH/+fLflBoMBCxYsAACcccYZfuXxEwRBEARBhCJ0t0MQBEEQBEEQBEGIXHbZZQCAXbt24YEHHsCePXtgMBhQWlqKn3/+GTfffDNqamoQGRmJhx9+OMitJQiCIAiC6HzIiU4QBEEQBEEQBEGIXHPNNdi1axeWL1+OX375Bb/88ovbOgkJCfjPf/7TLYsBEwRBEARBBBoS0QmCIAiCIAiCIAgRvV6P119/HZdccgmWLFmC/fv3w2g0Ii4uDr169cLUqVNx/fXXIzs7O9hNJQiCIAiC6BJ0HMdxwW4EQRAEQRAEQRAEQRAEQRAEQXRHyInuI7m5uQCAsLCwILeEIAiCIAiCIAiCIAiCIIiTGYfDAQA4evRokFtyckCFRQmCIAiCIAiCIAiCIAiCIAhCA3Ki+4jgQD98+HCQW0IQBEEQBEEQBEEQBEEQxMnMyJEjg92EkwpyohMEQRAEQRAEQRAEQRAEQRCEBiSiEwRBEARBEARBEARBEARBEIQGJKITBEEQBEEQBEEQBEEQBEEQhAYkohMEQRAEQRAEQRAEQRAEQRCEBiSiEwRBEARBEARBEARBEARBEIQGJKITBEEQBEEQBEEQBEEQBEEQhAYkohMEQRAEQRAEQRAEQRAEQRCEBiSiEwRBEARBEARBEARBEARBEIQGJKITBEEQBEEQBEEQBEEQBEEQhAYkohMEQRAEQRAEQRAEQRAEQRCEBiSiEwRBEARBEARBEARBEARBEIQGJKITBEEQBEEQBEEQBEEQBEEQhAYkohMEQRAEQRAEQRAEQRAEQRCEBiSiEwRBEARBEARBEARBEARBEIQGJKITBEEQBEEQBEEQBEEQBEEQhAYkohMEQRAEQRAEQRAEQRAEQRCEBuHBbkCg2HC0Bp9sPoH9pY2ADhiaGY87zxqMGWN7ydarN1nw9tpjWJtXg9pmC/qkxOC6if1w51mDEB5GzxQIgiAIgiAIgiAIgiAIgiAIiR4hos/bWojnfzqM9PhIXHlKHzg5DqsOVOHeb/9EhXEE7jpnMACgsdWG6z/ejoJaE6aPysaAtDhsPlaLV1flYV+pER/ccmqQ/xKCIAiCIAiCIAiCIAiCIAiiOxHyInp+dTP+vfIIhmXGY8GcyUiPjwIAPHhhDqb/bzNe//Uobjy9P+KjwvG/tfk4XmPCizNHY9bkAQCAxy7Oxb3f/IlVB6vwy8FKTB/dy9PhCIIgCIIgCIIgCIIgCIIgiJOIkM8vmbe1CDYHh39fNUYU0AEgPT4Kj1+ci6tP7Yu6ZgusdicW7ixF76Ro3Dypv7hemF6HZ2aMAAB8s6Oky9tPEARBEARBEATR1ZhtDrRY7NKMw8uBeZcCKx4GHLbgNYwgCIIgCKIbEvJO9HV51chKjMJpA1Pdll13Wj9cd1o/AMCfJQ1oszlwyZBs6PU62Xr9UmMxIC0WOwoNcDg5hCmWEwTRg6jNB357DohNBS55DYiMC3aLCIIgCIIguhSO4zDj7c1osTiw6dFzELnxJWDr//iFxVuBXmOBU28LZhMJgiAIgiC6FSEtohtarKhusuDsYemoaTLj/1bnY21eDZrNNgzPTsA9U4fi4lHZAIDjNSYAwIBUdcGsf2osiutbUWpoxcB0EtUIosey7G6gfDf/OioRmP7v4LaHIAiCIAiii2lss6GgtgUJaIXz2+uBorXyFYq2kohOEARBEATBENJxLlWNZgBAs9mOGe9swbYT9ZgxJhszxvTC8RoT/vb1H/hiayEAoKmNH5KYHBuhuq/EGH5+k5mGLhJEj6U2XxLQAWDfAsBuDV57CIIgCIIggkBNswWDdRVYFvkPRCsFdACoPtj1jSIIgiAIgujGhLQTvdXKZ/jtLTXijCFp+PTWiYiN5P+k4zUmXPneVry88gguHJmFVqsDABAZrv7cICqMn2+xO7ug5QRBBIVDS+XTbQbg2GpgxF+C0x7i5MDaCkTEADqKCiMIgiC6By2Fu7Es8p9I1LWqr1B7FLCZgYjorm0YQRAEQRBENyWknehstvnzl48SBXQAGJoZj9vOHAibg8PKA5WIconnNoe6SG5xzY+NDOvEFhMEETQ4DjiwxH3+vgVd3xbi5IDjgCV3AP/uBcy/CrCYgt0igiAIggAApO//SC6gJ/cHzn5UmuYcQG1e1zeMIAiCIAiimxLSInpiNC+ax0aGYWhmvNvy0X2SAADF9a1ijIsQ66JEmJ8YrR73QhBEN8PpAArWAw1Fvq1fdQCoP+Y+P/9XoKU+oE0LGBV7gT++AFoNwW4J0R4q9wIHXQ9uCtYB39/Jf28JgiAIIsiEt1SKr1sSBgF3bQDGXi9fqepA1zaKIAiCIAiiGxPSInr/1DiE63WwOzlwnPtyu4OfGRMhiezF9epDFksMrYiNDEPv5JhOay9BEAHCbgW+nsn/e/8MoGS7920OqrjQAcBpAw5+H8jWBYbjvwGfXgD89ADw4VmA4USwW0T4S/mf8un8VcDqZ4PTFoIgCIJgCLcYxdeGxBFAXBqQNgQIZ/pClItOEARBEAQhEtIiemS4Hqf0T4HV7sSOQnen5t7SBgDAiF6JGNU7CfFR4dheWA+nU664lxpaUVzfilP6pyBMT5m1BNHt+fUpoHAT/9rWAiy7B7C1aa/PccBBJg89eywQnyVN7/u2c9rZXoylLtcyX/cBTeXAF5f57ronugdV+93nbX8f2PlJ17eFIAiCIBiibI3i65YwfvQu9GFA1khpJXKiEwRBEARBiIS0iA4At0wZAAB46efDaDJLUS1HKpswf3sJUmIjcPHobERHhOHy8b1RamjDvN+LxPUcTg4v/3wEADDLtS+CILoxe74Bdn0qn2coANb/W3ubsl1AY6k0PfY6/p9AxR6gppvkftotwOJbgbYG+fymMpeQXhycdhH+U6kiogPAqieAY791bVsIgiAIQoDjEOtoEidN+gRpWfYY6XXVQagO9yUIgiAIgjgJCfe+Svfm8nG9sTm/Fov/KMO0NzZh+uhsmCx2/Ly/EnanE69ePR7xUfyf+ei0XGzKr8WLKw5jW0EdhmYmYMvxWhwsb8KMsb0wbWSWl6MRBBFUyv8EVjykvmzbu8ComUCfU92XKQuKjroSMDcCv78jzdu3ALjo+YA1td38+gxQ/gczQwfA1YFtLAG+/Atw20oguV8wWkf4isMO1ByWppP6858fwBdrW3wbcMevQNaooDSPIAiCOImxNCEcUo2OJh0jomeNZtZr5E0Iyf27sHEEQRAEQRDdk5B3ogPAf64Zi/9cMxZZiVH4blcpVh+qwumDU7FwzmRcPCpbXC81LhJL7z4D10/sh72ljZi3tRBmmxNPXTIcb143HjodRbkQRLfFVAt8NwtwWKR5uZdKrzkn8OPf+bx0FqcDOPSDNN1/CpDUlxcvs8dK8/d/F/yijweWALuYqI+YVGDOeiBjuDTP6BLSmyq6vn2E79TlA3azND31aeDU26RpazOwdA45/AiCIIguh1MULDeCdaKPla9MkS4EQRAEQRAAeoATHQB0Oh2um9gP10307szMTIzGa9eM9boeQRDdCIcdWPJXPtJEYMTlwHVfAd/fIRUGrTkEbHkDOO9Jab2iLUBLjTQ9+mrp9fibgF9ckRvNlUDhRmDI+Z33d3iiJg9Yfj8zQwdc/QnQewJw60/AF38B6o7yixqKgLUvAld+EIyWEr6gzEPvNRYYcw0fx3NiPT+v+iC/Xq9xXd8+giAI4qSltbEGccx0gzNemmAz0QE+0mX4jC5pF0EQBEEQRHemRzjRCYLo4ez4ACjaLE2n5wIz3wd0OuCS/wCxadKyTf8FqpkYjYNMlItOD4ycKU2PvgbQM88S9y4IeNN9wunkC4naWqR55z4BDL2Qfx2fyQvpqUOk5Sc2dGkTCT9h89DDooD0HCAsAvjLG/L1Dv/Yte0iCIIgTnoa62tk03UcI6JHJQCpg6VptSLZBEEQBEEQJyEkohNEiMFxHHYWGtDYZvO+ck+BFRqjEoEbvuE7eQAQl84L6QJOG/DZRcCXlwG/PQccXi4tG3QuEJ8hTcdnAEMvkqaP/ASYpUJbXUb1QaCaGS495Hzg3Mfl6yRkAafMkqabK4A2Y5c0j2gHrOiQNZIX0AFemGCHyh9aRpEuBEEQRJdialCI6PZY+QpsLnr1wS5oEUEQBEEQRPeHRHSCCDE25Nfiuo+24ZkfTpKMSo7jo04Exl4PpA+TrzP6aiDnEmnaagIKNwFb3gTMRvl6SsbfKL22t8kd711F6Q759PTXAH2Y+3oZI+TTtUc7r01E++E4uYiuzJcdeYX02lAAVB/qmnYRBEEQBABLU61susYeJ1+B/d1qKAqOwYAguhGHKhpR2dgW7GYQBEEQQYZEdIIIMTbk8e6hE7UtXtbsITSW8UUYBTJHuK+j0/ExGXGZ2vvRRwAj/uI+f9A57sfrakp3Sq9jUt0fEghkDpdP1xxWX48ILsZiwNwoTfdSiugz5dOHl3V2iwiCIAhCxNosFRa1cWGotUbKV8geLZ+mh73ESczW43X4yztbcOeXu4PdFIIgiPazZz7w4dnAy72B/wwBvr8LMJa0b1/HfgOeSwIK1nfN8boRJKITRIix/QTf8TG0WIPcki6i5oh8OnOk+nqJvYF7tgGXvwOcehuQNQbQMW7uM+4DYlLct4tOBsJjpOmmio622H9YJ3q/0/mHAmok9QciGLdYbZ76ekRwqVTkx2YrCoemD5UPladIF4IgCKILcbbWi6+NiEOLzSFfIXuMfLrqJBn9SBAKzDYHnvnhADgOKKo7SQxMBEH0PNb8C/jxXsBhBSbdCQw+Fzj4PfDxefyIM3+oOggsvavrjtfNCPe+CkEQ3QVDixVHq5vF1xzHQacluPYUlG5rpRubJS4dOGU2/w8ArC18xy8ihhfV1dDpgMRegOEEP91c2fE2+0NzFe9cFug3SXtdvR7IyAUq/uSnyYnePWGjXHR6IGuU+zojZ0o5s/XH+IdFWRoPiAiCIAgigOjNDeJrI5eAFotCRE/sw5sMhEi8ahLRiZOT99YfR1F9KwCgxeqAw8khTN/D+14EQfQsKvcDW/8H9J8CzF4OhLtGn426CvjuZmDVk8BNC33b15GfeHGcHXXdmcfrhpATnSBCiJ2FknPI6nDCZLEHsTVdBOu2Tuil7ibXIjIO6D8Z6DWOF6C1SOglvfbHiW5uBP78Gijd5fs2StgoF4B3onuCdeLXkBO9W8I60dOGAZGx7uuwueiAvHguQRAEQXQi4Raj+LoB8WhR3k/qdHI3OjnRiZOQY9XN+HBjgWyeyXwS9L0IguhZ7PqE///cJyRBG+CjbgeeDeT/4l0DaaoEFtwIfHcLEBYF9BrfucfrxpCIThAhhBDlIlBvCnykS5PZhr2lxoDvt92wbmu1PPQAYIvLliaaq3zfcMkdwPL7gM8uchfDfYWNctGHA70neF6fdeK31AAt9drrEsGhcp/0WpmHLpCRIy8US7noBEEQRBcRbTOKr41cPNpsvMNWBiui1xwBHCQeEicPTieHp384AJuDwwXDMxEZzssmTWZbkFtGEAThJ8XbeJ1hwJnuywadC4ADCjd73kf5H8DRVcDY6/kIXa2I3UAdrxtDIjpBdCM+2liA6f/bhHKjevX37Sfkgml9J+Si3z3/D8x8byt2Fxm8r9zZOB1A7VFp2tPFugMcbYmXJnyNc2lrAI6vcU1wwJ6v23dwVnzPHqvuWmbJUDxIqD2ivh4RHEw1gIl5EJOtIaIDwKiZ0uvaPBpZQBAEQXQJcY4m8XUDlwAAaLUqRHJWRLebgfrjXdE0gugWLNpdil1FDYiNDMMLM0cjMZpPwW0mJzpBEKGE08H/fif2kbvCBVIH8f/X5XveT9ZI4J7twFUf8xG6nX28bgyJ6ATRjfhqWzHyqprx1bYit2UNLVbkVfF56H1T+EKYgS4ueqSyCVuP80L9gXIPOVddRUMR33ETyPCQh94BqjgmIsZqAsxN2isLKAueFm3x/8A2M1C5V5r2FuUCuLvxle0ggouyqGivcerrARTpQhAEQXQ5bVYHkmASpxt1vJHALRedLYANSHU8CKKHU9tswb9X8vfXD1+Ugz7JMUiIjgAANJMTnSCIUMLSBIDTjsSNTuL/95RxDgCpgz3Xpgv08boxJKITRDehsc0mOtB/+LMcdodTtnynyxk+NDMeOVm8a8jQYgloG+ZvlwpcljWou+G7FKVA3ElO9GpOcZH3xY2uLOppOAE0lvt34Mp9fMVqAU9FRQUSewNRiUw7SETvVlTtk09naxS0BfgHIum50jSJ6ARBdAN+L6jDbfN2osRVTI/oWdQYm5Cgk+7xWsP4Dq1bnZ2M4YA+QpquUjwkJogeyisrj6DJbMeo3om47YyBAIAEcqITBBGKWFv4/8Oj1JeHudzirHExlI4XBEhEJ4huQl6l5H6uabZgy/E62XIhymXy4FSkxvEXn7oAZqI3m21YtkcSgcsaukHnWSkQZ+Sqr9dByhzJsumG6mL1FVmqD7vPK/Iz26vMz6KiAF/si3Wj11IESLeCdaIn9QdiUz2vz7rRaw4Bdcc6p10EQRA+8smmE9hwtBY/7Q/dok+ENoa6Gtm0OZwX0d2Ki4ZHyu+7qsiJTvR8Glqs+HEff+17+coxCA/j5RJRRLeQE50giBAiPJr/36GhGwnzI+PVl3f34wUBEtEJoptwpFIeIbLkjzLZ9A5XUdHTB6UhzSWiBzLOZdmecrRYHQjT6wB0Fyc6I1Qn9weiOudiW2JNkk1v2+tDR1HpRAf8L5DBFhVN7Ask9fFtOzbWpuYIwHHa6xJdC+vU0yoqysLmogPAoWWBbA1BEITfHHbdj9Q2B3a0myatBmDb++2LRSP8ptEgL6BujnCJ6MpMdEA+mqrqQGc2iyC6BasPV8Hh5DCiVyLG90sW5ydECXEu5EQnCCKEiEoEdHrt+BRhfnSi+vLufrwgQCI6QXQThLzzc3MyAACrD1ejsY13OzS22nCkiu/Unj44FWnxgRXROY7D/O0lAIDrJvYF0E1EdNZl3UlRLgBwwpIgmy4oOIY2q0NjbfCidUed6BwnLyrqS5SLAOtEbzPwxSyJ4GNu4mN9BDwVFRXIHAmkDZWm/R3NQBAEEUDqTRZUN/HieZ2pi0T0+VcDvz4FfDEDKP+za455EtPSUCubtkYm8/OVmeiAPBe9pQZoru7ElhFE8Pn5AP+QacaYbNl8inMhCCIkCY8EUgYBjWWAQ2UkjaGQ/z9QI/67+nhBgER0gugmCE706yb2Q25WAqx2J1a4hlLvLDKA44DBGXHITIhGahyfMVUfIBF9V1EDjlY3IyYiDPdfMAwAn9HeFMziOXarvGqzsqBmAKlt41DPSUJ6gq0Oy/Z6yDdvKgcsKk9XjcWAscS3gxqLARPTGfUlykVA+V7UUi56t0BZdM0XJ7pOB/Q9TZqmOBeCIILIkcpm8XWXONGtLUAFI5wXbuz8Y57kWJrkcYHO6GQAKnEuAJClMDCwD4oJoodhbLXid1ec5qVjesmWSYVFSUQnCCLEGHgmH6NSst192YkNAHRAv8mhe7wuhkR0gugGOJwcjlbzHdcRvRJw9al8rIcQ6SLloacBgBjnUh8gl9jXroKiV4zvjV5JMUiJ5W8Uy4PpRjcUAE7mRjWjc0R0juPQ0GpDNSdlV2frGvD5lkJwWjEpai50AV8jXUqVeeh+ONGV7wUVF+0eVCqKrvniRAeA9BzpdXMFYGnWXpcgCKITYaPlusSJblPcZ7QZO/+YJzk2k1xEd0TxxdXdCosCQLzcjYvWOvd1CKKHsPpQNexODsOzEzA4Qx4hKTnRKROdIIgQY8Is/v+1L8jvu46sAIq3ALmX+h4r2x2P18WQiE4Q3YDCuhaYbU7ERIRhQFocZo7vgzC9DntKjCioNWFHoVxETw1gJnptswW/HKwEANwyeQAAoG9KLIAgi+hKYbiTnOhNZjscTg5VXIo4r7e+AcdqTNh8TKOzWHNIPh0eI732NdOVzUMPj5HnjnojPhOIYQpWkojePWDz0GPTgMTevm3HiuiAfAQGQRBEF8KK6F3iRFeK6GZj5x/zJIdrbZDPcBXAblXLRI/LkE+3kIhO9FxWuvpDMxQudIDiXAiCCGH6TQJOuwso2wl8cCaw+h/AkjuARbOBuEzg4peldQs3A+tf4QXvrjheCEIiOkF0A/Jceec52QkI0+uQmRiNc4alAwC+2FqEQxX88smD+I6OkIle32LVdkv7yKLdpbA5OIzvl4zRffjiUn1TeFG4rKG1Q/vuEKwwrNO7C40BosH1IKJeL4nSAyP5qJbPtxaqb8Q60RP7AIPOlqaLNvtW6JMV0fucCoRF+Nxm6HTyhwpsdnx3huN6dhFU1omePZb/nHzBTUSnSBeCIILDYUZEbzLbYbF7qA8SCOxm+TQ50TsdXZtBmoiMR3Q0f89nUstEj00FwPyWkYhO9FAaW23YKkS5jHUX0RPFOBdyohMEEYJc+jpwyetAeBSw4yOgeCsw+mrgjtVA6iBpvaItwMZXgbyfu+Z4IUh4sBtAEITk/BrZS8rlvubUflh/tBbzdxTzeejpcchMjAYApLky0a12J0wWu5jT5y8OJ4dvd/AZ3rNcLnQA6JMsiOjBdKIzQnXqYCAiulMO09DKi+hN4RmAq/+YYDcgTOfEhqO1KKg1YYhiSKdM4M8cCQw8Gzi2mp9uLAUaijz/OFiagWrGze5PlItAxnD+x0hoD8f5LtoGA0Mh8P0dfJ7qJf8Bxl4X7BYFFo4D6o9L01mjfN82dRCgD5fii8iJThBEELDYHTheY5LNqzNZxXuCToGc6F1OhNUoTcSkIi6K7w6qZqLrw3ghvZUfEUlxLkRPZfXhKtgcfJSL230/yIlOEESIo9MBp8/h/3li6lP8P29c+QH/r6PHC0HIiU4Q3QChkNeIXonivAtGZCIpJkI07p4+WHJKx0SGISYiDEDHIl02HK1BubENybERmMG4LiQnehBFdNZd3YlFRY2tvKOkNVoasqzjHJg+kL887jhhkG/gsAF1R6XprJHAwLPk6xR5yUUv/wPgnNJ0e0R09j2xNAFNFf7vo6toawC+uZb/u9sagJ8e5IvJ9STMRsDOnC9J/XzfNiyCf1AkQCI6QRBB4Fi1CXYnh8TocPRO4h9cd3qkCznRuxSL3YE4hzTaALEpiIvk7ydVRXQAiE2XXrfUdmLrCCJ4rDzAR7koC4oKUGFRgiAIAiARnSC6BXkuJzorokdHhOGycdKNnJCHLiDkotdriOhHKpuwPq/G43F3FfG5mJeO6YVolygPSJnoZcYgxbnY2njHskDmyE47lPAQwhIjL541Kp7/2ysbFQ8S6gv4atNs23qNA6KSpHnectGVRUX7dlBEB7pvLrrDxuef1TMRJbYWIG9l8NrUGTRVyqcT1TthmrCRLrUkohME0fWIo+J6JyIjgR/xVtfVIjo50TuV2mYLknXMaAPWiW7ViO5hc9EpzoXogTS22bBFiHLRFNGpsChBEARBIjpBBB1jqxUVjXwnMjc7QbbsmlMlN+vpg+QierorF91gUhfR53y9G3/9YheK67Udv6WuzPPB6XGy+X1Tg+xEr8uXO7UzhnfaoYQ4F0dclmz+gEheTKhsVHTwlUVFM0fyw50HnCHNK/SSi87moacNBeLStNfVIkMhotd2QxGd44CfHwYKN7kvO7Co69vTmTQrRgIk+FhUVIAV0Q0n+IcPBEEQXchh5oF+ejwvoteaOllEt5ETvSupabYgBc3SjNhUxHuKcwHk9ygkohM9kDWHq2FzcMjNSsDQTPcoF4DiXAiCIAgeEtEJIsgIUS59U2LEojUC4/om4eGLcvD0pcORnSTPBJec6O4d3DarA6UGXgA/Vs04jpoqZR1UQSQXnOcCQv6psdUWHMdFjaJQZic60QURHQly50lvPe/Sd3Ois0VFdWFARi7/mi0u2lwhd9KzcBwfayLQHhc6wHdq4zKlaeV71h34/R3gz6/Ulx1fC5h60LDwQDrRnTagobjjbSIIgvADqT5LVzrRlZnojYDTqb4u0WFqmixIUTjRY6P4kYgmTRGdcaJTJjrRA/EW5QIA8S4R3WS1w+n0YJQhCIIgejQkohNEkDmiEuUioNPpcP8FwzDnnCFuy1JdxUXV4lzKjVKnVHCbY/sHwBvDgbfGApX7+PVcy4QMdIGE6Agkx0a47avLYIuK6iOANPe/P1A0uDLRo5Iy+WO5SOf4IlqVRqUTnWlb2lC+4jTgnouu5r4GAGMxnwsu0OeUdrUbAJDJOPTZdnUHjqwA1vxTmtaFAWc9JE1zDuDQD13frs6imRXRdUB8luaqqrAiOiDP3ScIguhkOI6T1WcJmhMdHF/ng+gUaprakAxGRI+V4lxarb5kotfRQw6iR9HYZsPmY7ypY8bYbM31BKMTx/FCOkEQBHFyQiI6QQSZvCptEd0TaR7iXMoapCzzEoPr9bb3+f/NjcDeBWi12lHn2rafwokOSMJ6eTAiXdh87/RhfOHFTqLB9RAiJS5K5kZPtvNuq8pGMzg2mqWaiXPJYhzyWWOA6GRpWisXvWKvfLp3R0R05vi1R7tPx9ZUC/zwNwDM+3bp68D5/wDimQ7K/u+6vGmdBlvYNT7T/+9s+jD5NBUXJQiiC6loNKOxzYZwvQ7DsuIlJ3pni+hKJzpAueidiLHBgAgdk30ew8a5aGWiMyI656DPh+hR/OaKchmWGY+hmQma60WF6xERpgNAkS4EQRAnMySiE0SQEZ1f2do3bmqkueJcDCpOdDbLvNTQCjgdQFO5tEJzhSiOJ0SHIynWXfATIl2CkovO5nsrC2gGGCHOJTk2QhbBEWvlXSltNgca21yRNpZm3kkutm2U9Fqvl7vRizRy0Sv2MNuEA1mj3NfxFTYr3tYCNJa0f1+BZP9CwMo43SbfC5x2B58dP+YaaX75br5Qa0+AdaIn+BnlAgDRifLt6o5pr0sQBBFgjlTwD/SHZsYjKjxMcqJ3dpyLmxMdlIveibQ0KgrOx6QgLtIVU6EZ55Iun6ZcdKIHsf0EP/L04lHaLnSAHx2c4HKjU3FRgiCIkxcS0QkiiNgdThytloZP+4OQiV7nVURvA1rrefeQQKtBMw9dQJjPutq7BEszYGTEYGUBzQDT0MLfCKfGRcpEzLDmKvE9FouL1ioiNrIUWe0DmVx0U7W6EMqK6JkjgIho93V8RZkV311y0fctlF4n9QOmvShNj7lWvu6BJV3Tps6GdaIn+llUVIB1oyu/awRBEJ3IYUW0nOREVy9eHjBUneiNnXvMkxhLk0IAj01FnCsTXbuwaIZ8uqUH1TMhTnqEWMx+qTFe1qTiogRBEASJ6ATRcQ4vB1Y8BJT94X1dBUX1LbDanYiLDEP/VHUxWwsxzkWlsKgyzoVrVhQ9bKkT11HmoQsI87vcia4UD7vIiZ4SGykXP5srkJ3IC9xicVE2ygVwF7GVuehFm+XTHAdU7pWme09oZ6tdCEVNBVgHf7CoOgBUH5Smx17PO9AFeo2T53/v/07dsR9qdNSJDgDpzOdZd8z7++J0AoeWAbs/V3dzEgRB+AhbVBQA0l33GEFxolNcSKdhN9XLZ8SwmegO9YKJShGdiosSPQhpRGqk13UlEZ2c6ARBECcrJKITREeo3Acsms2LWF9dDjSWe9+G4bAryiU3OwF6vc6vbYXCouqZ6JLw3WZzoKlW0a7WepS61lHLQwdYJ3oXi+g1CiG4E0V0juNgdBUWTYmLBBKYoZzmRgx0DQ6oEIqLssU7I+KA5AGKto4EYlKlaaWI3lAod9h1VESPSQbiMpn9F2uu2mXsXSCfHneDfFqnA8ZeJ00bCoCKPzu/XZ2J3Sp35iW2V0RnHi5YGgFTjfa6ALDnK2DxrfxDvIU3dp9MfIIgQg4tJ7rJYkebVSMrOxCoOdEpzqXT4FoN8hmxKWImOgC02lQ+61hlnAs50Ymeg9gP8EVEjxLiXMiJThAEcbJCIjpBdITDP0Isnmg1ARv+7dfmgvNruJ9RLoCUiV7fYpUXvgRQbpR3So01pfKNW+tRbuAzq7070bs4zkXmmtcBKQM77VAtVgesDl54TImNABLkMRzDYlsAAFVCnAvrRM8czuegs7jlom+Ru4nZKBeg4yI6ACT3l14bg5yJ7rADBxZJ030muhfMBNwjXfYv7tx2dTamKvl0QgDiXACgzkuky59fS68L1gFb/9e+4xIEcVJjsthRXM//1o/oxddniY8KR1Q4/xvXqcVFyYneZdgdToRbjPKZMamICtcjzGXkUI10iUkBdMz9Tku9+zoEEaIItaVSVOpDKRGc6E0kohMEQZy0kIhOEB2hYL18eu+37pEfHjiicH75gxDnYrE70cK4xMw2hzj8OjeL7wy31pfJN+YcaDDww3G1RPQ+rvkNrTbtnExvcBzv0vVAqaEVc7/+A3tKGuBqrLQwNlUeBRJgGlw3zlHhesREhLk5iAdG8q7xisY2/m9hnejKKBeBQedIr1tqgVomp1xWVDRCex/+kNxPet1Yqr1eV1CwTu5QU7rQBVIGAv0mS9MHl/ACfKjSpIhLaq8TXRnPU5evva6pBihXREitewko3dW+YxMEcdJytIq/F8lKjEKaq6CoTqcT3eg1nRnpYqfCol1FncmKZDBFv3VhQHQSdDodYiP5ey3V4qJ6PRCbJk2TE53oITicHJpc0Sy+xbnwQruJRHSCIIiTFhLRCcIDX28vxtI/y9QXthrcncWcE1jzL9gcTp+E5zxXnMtIl/PLH2IjwxEdwZ/CbKSL4EKPiwzDmL5JAABbY6Xb9q1GPipCq7BoYnQEEl2OC6Wz3Sdq8oD/jQFe7efRafzf1Ufxy6EqfLq5kJ/RwmRtsp22ToDNQ9fpdG5Z1n3DeWG/0mjmRUtW4M8apb5Tt1z0LdLrir3y7cOj2tt0CaUTPZj54vu+lV7rI4DRV2uvO5Zxo7fUAoUbOq1ZnU5zhXy6vU70hF5AZLw0rVaYViD/V4ijYAQ4B/D97VSUjyAIvzhcIc9DF0iPF4qLdrGITk70TqGm2YxkHSOix6TwEWuAGOnSatGI7mFz0SkTneghNLbZxNvmZD+c6JSJThAEcfJCIjpBaGBsteIfyw7iye8PwKFWaKlwE9xELAA4vgavvPchznh1ncebrIYWK6qa+M5jbrb/TnQASHPlotczxUWFDPO+KbFisVKdqdpt27A2XhDu66EavZSL7meki8MOLL2Td0bbzcCm/6iuZmixYtUBPgqjxOA6hsyJnq6yVeBoYPPQATcRPRNGAOA/pxovRUUFMobLO5uFm/j/nU65iB6IKBdALqLbzcFziLUZgbyV0nTudH4kgRYjrwT0Ug4rDi3rrJZ1PoFyout08kgXZZFdlqOr1OcbS4CfHuwZxVoJgugShPosylFxghO9U4uL2igTvauoabIgRdcszWB+o4XioqpOdEDhRCcRnegZCGaahOhwRIR5l0USRRGdnOgEQRAnKySiE4QG9a6oD6vDiTa1QksnmCgXfYQsL/Kq+o/R1GYRM0bVEKJc+qfGyoo6+UOqS/wV8vwASfDumxKDfi6BPLLNvcOTqmtGUkwEEqO1nRdSLrqfTvQdHwBVB6TpRnU3//d/lImZ5JKIzjrRPYiwAaBBmYMYGQtEJ4nLUxx8WyqMbeCUMT1aTnSdzj0X3ekEDCcAK9N5DZSIntRfPm0MUqTL4WWAgxFaxt3oef24NGDg2dK0MpoklGCd6BFxQFT7HooBANKZSBctJ7rNLL/+5F4KpAySpg8tBfbMb38bCII4qRCKio7sTU70nkx1sxkpbJwLUwg9zhXnojmKkjUHkIhO9BCMzIhUXxDiXMiJThAEcfJCIjpBaNDUJt0gtVpVOhVsHvqAM4AJt4iTo3WFuEy/TczZU+NIleD88j/KRUDIRa83sSK64ESPEZ3o8Xb3IlCpumbNPHQByYnuh4jeUAysVxRYtbUCVvkDBY7j8O1OqRBmY5sNja02PiZHoAvjXESYKI44Cx95Y7E7YS3fL60TlwnEeXDJs+Jwm4HPUu+MoqKA3IkOAMbiwOzXX/YukF7HpgFDL/K+TZ9Tpde1eYC1JfDt6gpYJ3piL3F4fLtgnehNZYDF5L5O4Sb+nBIYcw1wzWdyZ/+qx4H6gva3gyCIkwKHkxMz0cmJ3rOpabLI41xUnOgtave7gPyehzLRiR6CocU1ItWHKBeAjXMhJzpBEMTJConoBKFBIyOit1kVTnRDoVysHDIVOO9pIFwSpR+P+A4tLdqiYE0z777SyiT3BcGJXs840cuZOJd+KbEAOKQ5DW7bpqHJBxFdcKL7GOfCccDKR+UCn0CbvA3bCupRWNeC+KhwMYew1NAij3PxJFQHANGJHsfcPCdkiy/DTNVIj49EJGwIO/6rtE6vsZ53zIroAO9GZ0X0sCggc0R7my2HLSwK8HEeXY3hBFC6XZoefQ0Q7oOrp/d46TXnlI9eCCWaGRE9oZ1RLgLpOfLpehU3ej4T5aIPB4ZeyD+QuOCf0nxbK/DHvI61hSCIHk9hXQvMNidiIsIwMC1OtizD9aCenOg9g5pmC1LAjIiLURHRfclEbzMATo31CCKEEM00cf460UlEJwiCOFkhEZ0gNGhibpBalSI6G6UAAIOn8g7UM+4TZ/XV1SHzyFfa+2/j958U45v7QY00Mc6FzUTnBew+KTHISIhCergZ0Tp3R3yKrtklsmvjd5zLoR+AY6vVl7XK3fDf7ODF3pkTemNwOt9xL6+tk3eoO92JLjhQmJvnRKYoZHMFspOicb5+D8ItTLHGUVeKL1Xz8tOHAfFZ0nTRZrmInj0aCGv/5y4jMk7+PjUGIc5l33fy6XE3+Lad0o3PZsaHEk1MnEtiO4uKCmTkyqeVkS4c5yoq6mLAGVIE0ZS/AxnMw5lQfShBEESXkV/Ni6o52QkI08tH0XSNE11FRCcneqdQ26xwosckiy/jRRHdh0x0zgm0NXRCCwmia/E/zoU/TzyNNCYIgiB6NiSiEyFNYV0L/vdbPsxqmeUdpFEW56LYPxvlEpsGZLucyWfcD3u01NEYcuJrPg9bBeEGTChS0x5SxcKi6nEuOp0O45JUOqgAUnW+ONH9iHNpMwK/PKm9nBHRa5st+PUQX1D0pkkDxNiZumpFgcZOLyyqFufCOImbq9ArMRpXh22S5oXHACMuBwCUGlox/oXVeOGnw/Id63RyN3rRFqCKiYMJVJSLABvpEgwn+oHF0uv0XN//vsQ+cnebMvImFOC4wDrRUwYBujBpui5fvrxqP9BULk3nXCK91uvl731NXsfaQhBEj0cQyPskR7stkzLRrW7LAoZd5f7C3EjFkTsBU2srEnXM+83EucS6MtE1C4uyv9UA5aITPQLBTJPsY5xLPMW5EARBnPSQiE6ENK//mof//XYMX28LfA50k1aci9PBZxILDDqXF68AIDoR5aPmiIviLdVA8VaP+0/siBNdkYlutjlQ4+oQCwJ4brx6FEsqmr1GyfRxieyGFqt6LjzL2ucBU7U0Pfoa+XIm63zxH6WwOzlM6J+Mkb0T0d81hNxYpxTRuygTnY1zSWREUIcVY6MqcZ5+nzRvxGVANJ8buy6vBs1mO1YfrnLf+SBGRDcbASvj/upJInpzFWBgsrfHXOt7JrhOB/QaL01X7g1ky7qGtgb56ImOOtHDI4FUpkho7VH58qOr5NO50+XTmcOl16YqcgsSBOERQ4u2E5N1onOdJWqrOdE5B2Bpdp9PdAzl7wET5yI40TXv9ZTxepSLTvQA/HWiJ4oiOjnRCYIgTlZIRCdCmjxXcc49pYEXimQiOut0r9grz+scfJ5su5I+M+DgGBHxwCL1/btcDInRgYhz4W8CK4y8wyg2MkwskjMkRqUwIfjCov1SPYvoSTER4tDFck9udEMhsPtzZueDgemvyNdxOdGdTg4LXAVFb5rEi7+CE72loVq+DeOS6gwaWlTiXBRO4ouMixChYz7/8TeKL/eX8REvFcY2WO2KEQfKXHSWzhbRu9LBV7ZLPj3wTP+2Z9+L2qPqhTS7M82KBz8ddaIDvJtfQBnnworo6bn8ucaSocjaJzc6QYQ+pTuB/NWdkkMtPExOVckEFpzobTYHWpQj8gKFmhMdoFz0TiBM+Z6qFBY1+ZKJDgCt5EQnQh/pIaKvhUX59UwWe+c9WCQIgiC6NSSiEyGLzeFEST3vshbEzEDC5t3JnDkn1slXHDJVNlmvS8VW52hpxqEfVZ1WzQFwoqcqRHRllAsA9A1vUt8WzaLT3BM+RbpU/CmfvuR1vsOlZ/42l4i++XgdSg1tSIgOx1/G8q5dQUS3NiucTZ1dWNRbnAuAnGpGtEzozY88cHGwnP/eOTmV4qupg/n1lYTHyEXSQJDEiOi2Vpnrv9NhRXRdmNxZ7gtscVFw8tibUKBJIaJ31IkO8Jn6AoYCwGGXjsW69XMvgRusEx0Aao90vD0EQQSPQz8An10EfHst8NUVAR9d4smJHhcVLsZ81HVWLrqaEx2gXPROINyq7UQXPmefMtEBinMhegRibSSfC4vyD5ucHDrvwSJBEATRrSERnQhZiutbYXcVdSxraEO9KbAdvEatOJeCDdLr1CFyFzCAZosdPzoYN66lUbXYppiJHtP+TPQ0MROdH2pdbhREdMlhnqU3qm+rbxKH73pCKi6qHgsDwF207XMKH9XBdrpcHa5vtvPRO1ef0hcxrk6bIKLrFMVHuyrORebAU4igeo7pUI69DtDzbW612nGsRhpuXlyveH90Onmki0D2GCCs/Z+5KorvIIyBjzfSpJQR0bNHA5GeRze4Eejioi11wOHlwKongA/PAl4dAPz0YKc4OAEAzRXy6YA40XOk1w6r9Hnm/yJfT01ET+oHRMZL0+REJ4jQhh19UrQZ+PQioL5Ae30/UY01YxAjXQJ8jyVi1xDRyYkeUDiOQ5RNYTiJdY9z0RTRo5MBPXPvQiI60QPwN84lJiJMLMBMkS4EQRAnJySiEyFLQa089uFAeWDd6E1tUkdCLCxqMQGlO6SVFC50ADCZ7fjVORFmjumQKiJdOI4T99+hOBdXJrrZ5kSr1SEK3X2SJYd5ilPdtRYLC2DzXjBUEtE9rMuK6Do939kC5CJ4az3qTBaszasBANx8uiT8ZiZEITJcj2QwrvnwGCAyzmv72kub1QGzjY9gkRUUisuQF3Zk4MZJUS6HK5rgZEZyFte3uG+gFukS6CgXQEVE76JcdIdNXgy07yT/95HQC4jLlKbbW1y0cBPwwZnA60OARbOAHR8CVQd4IeaPecC2d9u3X2+wTnSdHojP6vg+MxQjFb6eCfz6DLBvgTQvNg3oe5r7tjqdfHtyohNEaGNWjCarPwZ8eiFQvC0guzeoxZoxiMVFO8OJ7rDx+edqkBM9oLRYHUiCIi4txj3OpUUrE12vVxgjKBOdCH38LSyq0+lENzoVFyUIgjg5IRGdCFmUInqgI10a1TLRi38HnIzzYLCKiG6xwYRY/OY8VZqZ/6usQ2ixO2F18AJuR+JcYiPDEBXOn8aGFqsszkUg3ubBLaR0fqvgU5xLGyOix6RIhVbZTPPWehwob4TDyWFoZjyGZSWIi/R6HfqlxCAVTCGxLnKhh+t1cke+PkxVCN3rHAxDrFTwUfnQpkjpRAeAgWe5z+sUEb2ffLqxNPDHUKP6oDzPVk3U9YZOJ39P2lNctNUALJrNt0eLdS8B1Yf937c3WCd6XGZgRhmkD5M/yDGW8A8B2Ad4w6aJoyLcYHPRyYlOEKGNVaVORJsB+OpyYL96zRV/aGjRzkQHgIz4TnSie3qQT070gNJstiEFimKtqk50D6O22Fz0QGWicxx/j3xoGeB0el2dIAIFx3F+O9EBMCI6OdEJgiBORkhEJ0KWghre+Su4pAItoqtmop9YL62gC1ON6xCcCcvYSBeHFTiyXNq3S6DX64C4SA0hzAd0Op1YXLReJqJLkRphLTVSMyIT5TvwYTiu33EujLNJ7kQ3oKCGFwNyspi4CRf9U2ORqmNF9M4tKirmwMZFivnxIonukRzfO85BZaM07PyA6/uWnRgNACgxqLw/KQP5eA2WzhDRoxL4hxcCXeVEL9stn+7XDhEdkL8ndccAS7P2umqsfcE9J1gfAWSOkqYdVuCHvwF2a/vaqAXrRFf53rSL6CTgzAc8r5MzXXsZm4veUtO1GfkEQQQWreuhwwosvQs4/lu7d81xHAxeRKT0BH5+pzjRtaJcAHKiB5hmsx0pOuaBTEQcEB4lTnrNRAdUI/o6zG/PAd9eByy+FVj3YmD2SRA+YLLYYXPwQ0r9EtGjePNTEznRCYIgTkpIRCdCFsGJfvk4PsN6f5kxoPtnnehinEsBI6L3OZUXuxSYXDdVG53jYAQjFjOOMSkPPcJdwPWT1HihuKhFFLpZJzqaq8WXjQnDZNv64kQXomF8dqLHaono9ThWzX9mQzPURfQUVkTv5KKiRqGYkNoQTkWutQ3h+MkxBRVG6T3Y73KizxjLr1ukFuei08kjXSLi5EUjAwkb6dJVInrpTul1bBqQMkh7XU8oi4tW+lFctPwP4I8vpOn0HODWn4AnS4C7t8rF5qr9wKbX29dGLVgnuloh2fZy4b+A+/4ALnwO6DNRviw2DRh6gfa2rBMdAGoo0oUgQhbWiT5yJjD6GvnyfPeaK77SZnPAaufdv9pOdP5BMTnRQ5tmsx3JbJyLwqggxLmYPInorBM9ECL6oWXA1v9J02z+P0F0MkI/IDpCL9Zo8gWKcyEIgji5IRGdCEk4jhNF9MvG9YJeB9Q0W1Dd5MHV5Of+mxgR3Wxz8B0GNl9YJQ8d4AuLArzw+rNjsrSgaAvQWA4AaAxAHrqAUFy0stGM6ia+kyuK6LY2vrCpi6rowfKNfRDR+7lc7fUtVnmBVdl+fHGi1+O4qxDnUCbKRTxOaizS2Ez0LopzUXWfKET0g3GTYUQCqlzfrxaLXfz+/cUlopcZ2uBgQ9IFptzDu6IBYPJc7QiOjsI63o1dFOdSxhQV7TuJf2jQHnqNl0/7movudAI/PwqAed9n/B8w6By+wKlOB1z2ttylv/n/eOE9UHSGE10gfShw1kPAXWuBh4/wf9uU+4DZP/KjD7RgnegA5aITRChjYYTPuHTg6k/lDyyVo3D8QBiRFRmuF53ISgQnem1zgEfxAORE70KazTa5E539XYQU59KqdZ8HKET0Dmai1x0DfrxP0cgK9XUJohPw2A/wQIKr72YiEZ0gCOKkhER0IiSpNVnQbLZDrwNG9EpEjkuU3VdqDMj+TRa7rGhkq9UhFwwBYMCZUIO9qfrBfgazhAMOLgHAOtE7np8sxLkcdDmjYyLCJEdZc5Vs3eO6AfKNfXASJcaEi7nrtVrDuX1xojttqKzhO11qTvQBaXFyJ3ps5zrRPd48K8TQvKy/AAAqjHyH/1BFEziOj3IZ2zcZEWE6WB1OUWSXkT0GeOQocPfvwAX/DOwfwZLMfLbGEj5n1F92fAy8MQr44W7A4aVzYKoFGgql6b4Ttdf1RmIvID5bmvZVRN/zFVDxpzQ9+mpeQGdJyAJmvCFNcw7gh7k+FdX1it0iz4VNCLCIzpLYGzjtTuDil/nvlMd1+wCRjMhOuegEEbqwcS6R8fzDwXimGHMHRPQGV1HR1FiVWDMXlIneM+DjXJjvkkJEZwuLclr3D3HMPV1bg/f7BC2sLcB3swCrIqrI3AhYPUQHEkQAkYqK+ieiJ1ImOkEQxElNjxTRfy+ow6CnfsaDC92FmHqTBf/68SDOem0dcp9dhfP/bwM+3FgAu4OK2YQSQh56v9RYREeEYUwfPlYlULnoypy7VqtDHl2h0wN9TlHdlh0K+weXA0ci4xDevwio2IPUw1/jtfCP8arpWWDVE0DJ9nYXVBIE832l/N/eJyVG6gybqmXr/mFKhYVj3O8+ONF1Oh0yE/lOdE2zhmusVUNEV0SyhFkM0OmAwRlxbrvonxyFZDCRKJ3tRHeJBylqQ9gHnye9ThkEUz9+1EFVI9/hF4qKjumbhDC9TnTrF9epRLoAfMcza5T6skDBxrlYm/0XVpqrgF+eBJrKgH3fAn/M87x+uTIPfZJ/x1Pib3HRVgOfpSoQGQ9Me0l93dFX8QK7QF0+sP7l9rRSjuIhFRIDGOfSEXQ6ICNXmq4lEZ0gQhKnE7AxvytRrgfQ0cnSvI440Vul2iBapCfwv/+UiR7aeI9z4UcicJwHNzrrRAcnN1D4CscBy+/XHiHVXKk+nyACjFBUWTXW0QMU50IQBHFy0+NE9GazDY8t3q9qwmxsteH6j7fjq+3FGNMnCX89cxBiIsLw6qo8/H2Bj85HolsgRGkMcTmax/ZLBiDlVHeUxla5u6BN6UTPHKkZp8A6Ezjo0TjkCmlh9UHg4/Mwbt8LuD58A0Zb9gI7PgQ+vxh4cxTwy1NA6S73nXpAyEQ/Ws07euR56HKRb3tNBAxg2t3qW6al4ESrUetEO2yAhYlhkcW5yDtpqWhGf9eDDyX9YszQ65gTt5MLi0pOdJWb5z6nAjcuBM55DLjle2Sm8O9Zhauw6AFX/v5Y18Ob/mkuEV2tuGhXkawoYNroZ6TL8d94l7bAtncBp4dh3cqHSr3VHyr5DJuLXn+cd6R5Yu3zcvHo3Cc8i9iX/lfudt/5KWDrYPyTsrPfmU50f2EjXSgTnSBCEzYPHQCiXMXBWRdxh5zo3kUk1omu6VBuL0onOivSkhM9oLjHucjvsWIiwqB3+S80i4sqRwi2Jxd916fiqEwAQJjiAQ6J6EQX0dE4F3KiEwRBnJz0OBH9hZ8Oo9yoPjz0f2vzcbzGhBeuGI0PbjkVT14yHMvvOwvTR2Vj1cEq/HKQbtxCBUlE5x3N4/oKTnRjQDp5TYobI7PFApQzsRF9T9PcVlmUqXLg5b4dtLkC2P4+8NmFvJjuI+muTHQhj1smoiuc6JXOJBg4VkT37kQHgMwEV2ExNRFd2YHXinMBkKJrVo1yAYBYu0I07eTCol5vnnMvAc5/Fkgbgt6u4qqVLie68LBmtOt7NzCN/x6qFhftKlgnOuB/cdHja+XTDUXAkeXa68seKo2SHJLthXWiA56Li5buBP74UppOzwUm3+15/7GpfJFOAXsbULZTc3WfaFLkt3YXJzogLy7aWheYInAEQXQtShE90nWdDZCILmSie3KiZ7ic6Fa7022UXodROtETmAed5EQPKCazDclgI/PkIrpOp0NcpBDp4osTHf7notcXyO9v9RHA5e/I12mivhjRNQhxLilxnedE5zgOjyzahwcX7gn8Q0iCIAgiKPQoEf23w9VY/EcZLhyR5bbMandi4c5S9E6Kxs2TJLEpTK/DMzN4seGbHX6KTkTQKKjlxUrBiZ6bnYCIMB2MrTaUNXQ867ixTS6iZ5gL5UOqPYjowk2V4OyqiRoEDDjL/RhcLKpjhqrvZMeHgLlJfZmCVEXnt68rWgSATES3cBFoQpxcRG/xUUT3FOfSqhjOq1VYFEAamjE0U0NsVYp8XuJcSg2tuPjNTXh/w3GP62nhi3gg0CuJf4hQ3WhBk9mGQldsixAj1D+Vf89L6oPoRE9SONH9EdGdDqBgnfv8rW+pZ6s77IqHSh3IQxfwtbho3TFgwY2QFRO99HUgzIdO0LBp8ukTG/xooAqh4kQHyI1OEKGIRelEVxHRzcZ2x8EJD5NTPTgxoyPCkODKy64LdC66m4jOXEPJiR5QLK1NiNQx4niM+2i/WFeki6YTXWlu8HE0o8ihHwAnc3998b+BEQqjCRUXJboIYwed6L48VKxqMuP7P8uwbG+FKNoTBEEQoU2PEdENLVY8ufQAJg1Mxe1nDnRbfrCiEW02ByYPSYNeLy+e1C81FgPSYrGj0CC6eYnuTUGNy4nuEmSjwsMwohc/zHmfK2qjIzS5RHSh4zjMcli+gkb+s9XuhMXOd2Z7JfHu5SazDbh2Hh8NMuU+4OrP8N/chRhn+QTfTVwIPHwEmP6qPA6Dc7oXMtVAiHMRkMe5SCJ6vT4FgA4GJErLfewAZbqcaDVNak50hYjOupsUnbQUnQcRXemK91JY9L31x3G0uhnvrD3uNnLAF4yCA8WHLMSsxGjodIDV4cSm/FpwHNA7KRrprmHuA9N5Eb0omCJ6TDIQlSRNG/2Icyn/U12wqNgDFG1xn197xOeHSj6TkAUkME5utVx0Yynw1Uz593bMtcDgc307RlwakD1Wmu6oiM460SPjgehE7XW7GtaJDlAuOkGEIsrCi2pOdM7pvp6PNPiQiQ5IueiaxcXbizJSS+lEJ+dmwHAoTRMqkXlCcVHliEpphQ7GuZTukF6nDAQm3QVExgLRzL0LOdGJLqK9hUXj/SgseqxaehAqXG9DGsMJYNGtwIqHqAgwQRAnLT1GRH922QG0Wu14/dqxUlFFhuMu0XVAqntBQ4B3klrtTpQGM9OY8Ik2q0OM7BnCRIMIruADASguKjjRs1wO5BwbI0BFJwOpQ1S3YzsevZP5bZvMdiA+k48GufhlYMw1KOSyAOj4Cu+Jvfkoilu+l++sZLtPbU1TdH77JLNxLlImuimCd3a3J85FGM5dq+ZCU+6DFc4jY4EIyRmf6lFE992JXtNkxtI/ywEAbTYHftzrv3PJHyd6RJhezIVdfYh/MDGmr9Tp6++6rpTUtwR3uCYb6eKPE/34b/JpHfPT8Pvb7uuXKmJQOlpUVICNdFE60U21wNcz+cKnAn0nAZe95d8x2KKxFXs6FIUgc6J3Jxc6wF9XohhRn5zoBBF6uDnRXb/fMcny+e28jgkFtlO9PEwWfv8C70RXjBxk61ZwDvc4G6L9KL8jKk70eJeI3mrVENGjEvkIFgF/RHSnU37v0H8KXwQbkD9AJyc60UV0RWHRYzXSNczoTUS3d3OR3dwIfHkFcHgZsPtz4Fffo0cJgiB6Ej1CRF+2pxwrD1ThyUuGY0CaukguOIuTNX4oE2OEoVk01Kq7c6KOvyFJiY2QRZmM65sMIEBOdNeNkRDjMcp5VFrY9zRAr37qmFzbxUaGicMD1ZwKwvdR+N7xG6XK3aOlPorors6t2Dw2zoVxolujMwEoRXSD5+KRLoRMdFUnujLORSl+M9MpaBZHD7jvRynGp6ivB2De70WwOpyIDOM/h4U7/Y9i8ncYZy/Xw4n1eTUAgLGu7xsA9EuNgU7H54jWmYJ4ExwIET1zJDBypjR9bDVQrRiJUbZbeu3hoZLfsMVFDSf4nPb6Ar5A7vwr+YKjAlmjgZsXAZHq13xNWBGdc6o77X2FdcwldjMRXacDMphIF3KiE0To4UsmOtBuEd3Xh8npCfzyLnWiA5SLHkD0nurXuBAy0U0WjftCnU6ei+5PJnpdvnzEW7/Tpdfs7yc50Ykuor2FRRMFEd3iXTM4XiONEjJqxblwHPDdLOCVPsCqJ/xqS5ey6gmgkelb7P2Wvz8nCII4yQh5Eb2q0Yx/LT+EKYPTMGvyAM31Wl1FciLD1f/kKJcYJ0RxEN0XZR66wNh+vDP4YHkTnB2M5RFE7qzEaCSjGYN0zE29pzx01w1VfFS4lJnX5u5UEER6YR2R/kynomw34PB+gxYXGSZ+r6PC9Uhn410YJ7oznq8VIItzAedTJ1VwoteoFhZVZqIrOvdMR61XRAsSlX+zACPGNyIBCAtXXc1ksWP+9mIAwEszRyMyTI9DFU1+jUCw2B1i4SxPWbAsvRL5BwnNrtEGo/tITvSo8DD0dsX3lBi6SXHRRh9F9FYDUP6HND30AuDM++Xr/K4o/MUW5PTwUMlvlMVF518FvHMK8H+5QNUBaX7qYOCWpR4ftGjSfwoQxnzmBevb11ZA7phjnXTdBTYXveYIRSMQRKjhSyY60H4nupCJ7kVE7zInulJEp1z0wGC3YHSrwpih8vsZ5y0THeBj0QT8EdHZKBdALqKzI7lIlCO6CDHW0YcRqSxC380nJ7oszkWjT1f8O3BkOeCw8jWx6gv8ak+XcHApsG+BfJ7DCuz4KDjtIQiCCCIhL6I//v1+OJwc/nONeoyLQJRLZLQ51EVyi2t+bGRY4BtJBBQxD10hog/NiEd0hB4mix0n6jomZAoienZiNMbrFYUr+2mL6IITPT46HIkx2pl5zYITPVohFPefIr22tQJV+722VafTiZEufVNipPPAYZcNtY1M4junMic64FMuupCJbmixuNcNYMTvFsTgj3LFe8840XtFeIhLYtpaz8XDrnGuLtxZgmazHYMz4nDNqX0xfTT/dy3Y5bvzWrhx1uukYZne6OWK5xEYw4joADAgzZWLXhfESKhkprioudE3F1/BOsiKdA69kBezB50jzTuwWMr/bjXIHeGByEMX6D0B0Hm5Bif0AmYt4zPU20NkrLzz3t5cdI7r3k50QD6ypc3gf34tQRDBxZdMdKDdjm3Rie7lYbJQ/6NTnehhke4RI+aOx/Od1DgdwL6FwDsTcYV5uXyZMt8cUia6ZxGdcaL7GAkIQC6iRyXJR0rJRPTKdhfK7YlUNZphtnkfMUr4j+REb3+ci6cIR47jfItzaSiST3e3+L3Gcj4DXY1dnwGW9tXkIAiCCFVCWkSfv70Ym/Jr8fSlI9AvNdbjukKMiyCOKhHjNbRcskS3oaBWKCoqj3EID9NjdG9e2NzfwUgXNhN9gkxE1wF9TtXcTshET2Cd6CpOBSE2SBbnAgD9J8unSxTOHQ1SRRGdOQ9aasCKowkZffm/QNlx8qETlBYfBb0OcHJAvdKJxjjRDc54fL6lUL6cKRCaqvNwo8W0o55LQGWj2W0Vq92Jz1z7/9s5g6HX63DjJN59/eOecs8dPwZ2CKey0LAWgtMc4B9WKJ17goheHMy6CqwTHQAafSguenyt9DoiVnqQc8YD0nynDVg6B1j3MrDhFfn2Hh4q+U1cOl87ICxKfXlMKi+gp2iPOvIJNtLFUOBf9I1AWwPgYM6F7u5EB/iCsARBhA6amegdd6JzHOdzYVFhNFrA48pYJ3p4jErWuzGwxzuZKPsD+Ogc4Ie/uY9My50hL+bpQhLRPYi2bNH39jrR+ylGsLEPoZ02/8T5HkxxfQvOem0d7vpqt/eVCb8w2xziKHV/C4sK/TuHk0Obhwcctc0WsT8JeCgsqjyPDCf8ak+n4nQCy+6WjwrqNV56bWkE/vyqq1tFEAQRVHyzYHZTftrHOyOf/uEAnv7hgNvyZXsrsGxvBa4+pS9uOp13aBbXqwtcJYZWxEaGoTdblJHolmjFuQB8TvXu4gbsL2vEVaf0hdPJYf3RGizbW4Hzh2fgygl9fTqGIHKnxUVisP6YtCBjuGrHQ6CZdaK7nArKBzccx4kRL24ievIA3pEjFCws2QZMucdre1MZJ7rUGPmQ2N59B+Jfl+VgfFQqsIJZ4IM7NUyvQ1p8FGqbLahptiAzkXFlt0qd9wbEY8PRGljsDkSFuxzFjBM9wdmkfRDGEd/AJcBqaHV7OPbTvgpUNpqRkRCFmRP6AAAmD07FoPQ4FNa1YMX+Clx/mkJIVkFw32nVSFAjO0n6m5UudEAqLlpc303iXADAWApkj9Fe3+mU56EPPBsIdwnYQy8AMkcBNYf46aLN/D8Znh8qtYuzH+YL7TZV8OdBcxX/2mEBxlzXcQEdAAZPBda9KE2f2AicMsu/fTRVyKe7uxMdAGry5CMMCILo3rCZ6GFRQJjrN0t5H9IOEd1kscPm4B+0e4s1E4uLd6YTPSKar7HBQnEu7cNm5uuIKJz89VwCws97Akln/011szjXaNwWrcKigCIT3cfRTS118hFs/RSGEeVD6OYKID4DJzs7Cw2wOznsLTUGuykhidPJoa7FguSYSLc4V2FEaphe5z4q2AtxkWGisajZbEdspPr2rAsd8BDnohwR3FCovl4w2PEhULhRmu49Abh1BfDOqVJk6Lb3gUlzpN8ngiCIHk5Ii+jXnNoXkwenuc0va2jD93+WYXh2Ai4elY2RvRMxqncS4qPCsb2wHk4nJ3OflhpaUVzfirOGpiPMR1cqERycTg4natXjXABgbF++Y/lnSQMW7CzBp5tPiKL79hP1vovoLpE7KUqPcXomm67vRI/bNYtO9AhRIFfGuVjsTlhdUSVuN246HR81cXgZP12ynY+N8BBVBEgu6UHpjDvfVC1bR5eQjb/mDAJM8XIR3Yc4F4DPRK1ttrh3ohnHUAOXgBabA78X1GNqLl/IlBXRo+yN/PBivUpkBxMLY+ASUG9oxZnMYo7j8NEm/rP465kDRZFep9Ph+tP64dVVeViws9QnEV3MQfTDfdKbiXMZ09ddRB8oONE1HtR1CUn95NPeHNbVB10jFlwMvVB6rdMBZz0ILL1Le/vMER4fKrWbiBggbQj/rzPoPZ4fTm5xCQwnNvgvojdXyqe7oxM9IZv/fAQhhZzoBBFasE70KOaeRx8mv4a1Q0RvaOF/B6Mj9IjxEmXYaXEuMid6NDnRA4XhhExA5yLi8HbbNHxin4Hfp8wEwtXFLsGJbvI1E91s5Gv3eBPP3PLQJ8mnlQ+hmyqBXuM87/MkIK+KH73ZbLajyWyj0dJeOFTRiG92lKDU0IryhjaUGdtgtTsxKD0Ovz18rqyPL7jCk2MiPMbBqqHT6RAfFY4msx3NZhuyEqNV1ztWLR99qxnnonwY1V2c6LX5wG/PSdMRscBVn/K/RZPnSsuayvjM9HHXB6OVBEEQXU5Ix7lcO7EfHroox+3fNafyQunw7AQ8dFEOLh6VjeiIMFw+vjdKDW2Y93uRuA+Hk8PLP/PCwqwpAXA4Ep1KubENFrsTkWF6uevahSCi7y9rxFNLD6CgtgUJrk5BbbNFNZ9cDWH4XYa5GPFgOnnKG38FbCa6kJmnjHMRnOl6HRCn5l5gc9FbanxyJNx3/lA8dnEurp3IiKgKER3xroJdMSkAmBtGH4fNZiYKxUUVMStMnEsD+E7+6kPSsa1R0rBzHefU7hQz7TAgESWKWJQNR2uRX21CXGQYbj5dfq5efUpfhOt12FtqxJFKD253Yf8tvg1hZ8lm4lzG9kl2W97fJaIr292lxKQAkUzmvTcRnXWhA7z7nGXMtcB5TwNZY/jvj57pwIVF8dEroYg+DBh0tjR9YoP/Gayh4ETX6eRu9Jq84LWFIAj/YTPRIxXGAVZwbofYLBYV9eFhsuBEr2+xdLhwuwyZEz2G/8fGeZETvX0ozBGNV3yJN+3XokUXi3gN1ywAxPubiQ74dg/Jiui6MPcRbGpOdAJ5VdL9bIWxzcOaBAC8/PMRfLujBJuP1eFEXQusdv6+rrCuxW2UqK9RVlp4iuwUEJzofVyj3IUHl250VxF99+fy2MKLXwbSh/KvT/2rvL/x+9tUvN5H5n79B654byssdqp1QBChSkiL6P7y6LRc9E2JwYsrDuPOL3fh1VV5uOK9LfjlUBVmjO2FaSPbWaiO6DKEPPSB6bEID3P/+g5Mi0O2yxHQJzkGz84Ygd+fOl90Ufla9FGMczHuky/wUkTRZOG3i48KFx0jSuFe2HdCdIR6HrdbLvp2r+3tlxqLe6cORRIbD9PMiOg6vVREKixc3vlu8VFEd3Wia5qUTnRGRHcVLV1zuFrsaFfZFPUK1Dpc1la+kKoLA5cgE6MdTg7vb+CHAt90en/53wm+gz9tFH/+LtzpPd/a2I5iQlkJUUiMDkdUuF41zmVAGj8KwNBiFT/jLkenk0e6KHNQlbB56CmD3J3fOh1w3hPA3VuAR48C/6gFniwF7t8LPFEEDJ8RqJZ3PWwuemsdUHPYv+1ZJ7pOD8RlBqRZAYfNRa89Qp0cggglZE50RVFwNhe9HWKzwQ8RKS2eX8fm4GQZvx3Gzojo4S43ZwcfDhBwE+WaI3nhOz4y3GMdGL8z0QHfctHZ+j7Zo+WjKgD+/pQtKt6kGOl1kpJXKT1EK28gEd0bwkiZe84bgm/vOh2bH5+KUb0TAbhHq0gjUtvn7heMUiYfRPTTB/EFk41a107lOdRYBtgDXH+iPZTtlF73Gs8L5wIxycCpt0rT1QeBAqZPEcIcrzHh4UV7xZHvgaTC2IZfDlVhX6kRx6oDv3+CILqGk0pET42LxNK7z8D1E/thb2kj5m0thNnmxFOXDMeb1433ezgX0fV4ykMHAL1eh/l3no4vb5+EjY+dhzvPHoyE6AgMdsWcnKjz/oNlczjFYjPxtX+K8+0R8UB6rsdthZuphGhJRBeiYQQaxTx0DTdQ1mgggollKdnmtc3qjWEy0eMy5REqbCfIVyd6At/BrWGHc3OcbBi5Pi4VCdHhqDNZsMeV4VhiVowYUDueYp6BS0CpS0Q32xy455s/sKuoAZFhevz1zEGq7bvBFePyw55ytFk9P903uNwg/jhQwsP0+O5vU7B47hQkqdx0x0eFI90lNJQEM9IlmRmN4MmJbm4CSpkHNGyUixY6HRCdCKQOAiI9F3Pu9gw5Xz59YoN/2zeWSa/js/iHU90R1one1gCYarTXJQiie8Fmors50RkRvV1xLi4nug+/g1HhYWL8XJ2yuHhHsCuc6IA8F52c6O1DcU/VqOMf/Cd4yX6OFTLR/XGie8tFt1uAij3StDIPHeDvTxOypWllXNpJSG2zBfUtkpBaTk50rwh1qaaPzsYZQ9LRLzUWuVn8w0dltIoY5+JnUVEBySilfq5wHCce8zRBRNeKc1H2izhn+wreBxKbGajcL00PPMs9WnTyPYCeuaZsfbtr2tbJLNxZgqV/luO7XaUB3/euIsl4VhTMGloEQXSIHimiTxmShqJXZ+B/N0xwW5aZGI3XrhmL3c9eiKMvXYLfHj4Xfzt3iFvBEaJ74ikPXWBoZjzOzcmQOdWFrPDCOu8/WGwh0MgqSUQ3powF9J6/J2Jh0ahwUSRvszlgc0hREYJLWTPbMCwc6Mc43kt2qK/nDdaJnqAYZRHHiug+ZqKrFRYzNwKcJFiHx6eLWeirD/Mi/nETMzQb8E1EB+9EbzLbcOvnO/HroWpEhunx9o3jNYv/njU0HX1TYtBktuOXQ547YJIT3b+b5xG9EjG2b7LmcsGNHtQbI9aJ7ukmvHAT4GRu/n0R0XsSqYPlGfL+iuhlu5l9dVJ2eyBgnegA5aITRCihlYkOKBzb/ovoYqyZj7+DgtiuWRyvPdhUnOhsnQ1yorcP9p5KFwYjx9+bJHjJ1BbjXDwWFlXUovImolfuk0dCaMUiJjCRaCSiy6JcABLRfaGZGekrMDSLv266Ffls8X9EKovwQEorJrS+xYqGVht0OmDiAP6BZ4OaiM5x6qM5gh3pUrUfcDJ/m1pNsKQ+fOyjQOFGoOpg57etkxFGW9WZAj8aYEehJKIHtYYWQRAdgpRjIqQQ4lyGZMZ5WVPO4AzfRXThx7N3lAW6WilDuCbZe5EjsbBodITYGQHkTgVBpPdYIIjNRa876nPkigzWiR6fLV/GFPv02gFyIca5sJnoCvE7KiFNjFVZfagaHMfhsFHxd6qK6PI2NHAJMLbacO0H27Cj0ID4qHB8cftpmD5aO3dar9fh0jH88r0lRo9/i8GPLFh/GJDaDYqLsiJ6WwNgaVZfjx12qY/gXSYnEzodMPhcabp4q+/DZ021cjG6O793ytEz9ceD0w6CIPynM53orb470QFp5JahJYDCgrKwKCB/OEBO9PbB3tfFpsIkjK704kT3rbCoMhPdyz2kMpJQGVkowNYVoTgXWZQLQHEu3nA4ObS4vufsiIucTN6Jnq+IzmgQ41zam4kuiOjq54oQ1dEvJRbZSfy1zWxzwmxTjJS1tshH5AgEW0RnjSKAdpzplPvk08VbO6c9XYhw/VN96NFBdspEdHKiE0SoQiI6EVJ4i3PRQnCin6j1wYnuuiE6PbJINr8yYbTXbdnCouFhesS5hsay7nZh/5pxLgDQ73T5dGk73OienOisiM5kmntCKizKOIoUHfe45Cycm5OByDA9CutaUFBrwj6D4jKjKqLL28C54maOVjcjPT4SC+dMxhlD0t23UzA0k/9eFHj5nIWb5+R2OlC0EJzoQb0xYt3VANBQrL7eiY3S6/6T3V2OJwODp0qvba3y/EdPKDsJ3VlET8iWx0PVFwSvLQRB+Af7ENTNid4xEV2INfP1d1AQmzQjCdqDrLCo4ERPluaFghO9pQ7Y+B9g9zzAEaR6KEpYYTs2Xbzv9BbnIpg/Wj1lokfGSw88AO+Z6Oz9a2IfIKmv+npscVEqLIojLif68GxeBCYnumfYbHL2ez4sS+gXmOBgiiIHqrColhP9eA1/7R6WGY/4qHCEu2oRuAmzWudP0EX0XdLr+Gz+3FUjaxQQkypNVx3o3HZ1AZ0lotebLDjOjIgoIic6QYQsJKITIUNjm02MEhnsp4jOOtE5L4X1BCf6KWHHZPPLY0d5PY7ww5vg6ogkxrhn5vnkRO87UV5kqdR7cVEZHAeYGBE93pOIXudTscGMeL7TVNtskd5DhfidlJaJhOgInDGU3//KA1UoqLegiWPys9VEdIUbPjGVb2+/1BgsmXsGRqsU8lRDeLhS4KUYjNFPB56vDEjrBk70tKHy6cq97uu01AMGRkwddE6nNqnbovy7C9b7tl3RFul1WJTXgsNBRaeTF4wlJzpBhA4yJ7qHwqJ2M2DzT2TzJxMdkER0QyCFBZkT3RXVFmpO9FVPAOtfBlY8COxbEOzW8LD3WbFp4j2otzgXnzLRdTp5XR1Poxk5Ti6iKw0iLKwTva3B7+9zT+NoFS/CXjiCvx+uIBHdI0JUZmS4HlHhUv+pb0osoiP0sNqdKDFI9+aBKizapOVEd4mlQ7PiodPpxIeVDS0K0V2rLlWwRfRyxoned6J7HrqATscXCxaoDv04F+F6aQxkdBmAXUX8w+5IV9xsUOtnEQTRIUhEJ0IGIQ89OzFaFpXiC/1SY6HX8SJ3rZeiWILIPYY7Ks4rcPZCA+dduBccCcKQWekmi3Wiu0T0GA83blEJQPYYaVo5HNYbrQZ5lp1SRGcz0e1m3oXrBcGJbrE7pZvGNrmInprOd4IuGskf76ttRbA7OTQgUd42t/YyN5FhUbhv+njMnjIA3889AwPTfY/uGeJ6WFLZaPbYCRSGo7e3oJAW3UJEzxwhH/avNoqBdZgA3VsE7kziM4FM5uFY8e++bceK6H1PkxyU3RX2wQqJ6AQROnjMRE+RT/vp2jb4WRtEEJsCKiz44kT34SF/UDm+RnpdsC547WBhIwDj0pisaN+c6C1Wu2fDSZyPInpDodxp60lET1DE9Z3Eueh2h1OMAzl/BF9nqKbZAqvd6WmzkxpB+ExUfMfD9DrRYMMWF+1oPyDBS2HR/GrBiZ4gO47bSB4tJ3pDYbvaFRBMNfKaSmp56CxZTH+15gjg8PAQLgQQrpcBjS6DVFT0IlfsaVWTGW1WD6N+CILotpCIToQMwhAowVXuD1HhYeibwguchV6iPhrbbAiDA8Oth8V5fzhz0GrzflMgOtFdN3GJKsP9mtrssmWasLmRFXvknU2vDamSTycoM9EV0Sg+5KJHR4SJf1etkIuuEMQzs10iuss5IxRlaYtIllbyVlg0Ng2Th6TjhStGIzPRP3EyOTYS6fH8japW/r3N4RRvetvrQNFCiHOpajK75x52Ffow+Q1vqUpEiUxE1wF9Tu30ZnVbBpwhvS7/A7B7fsgWUnnoAqyI3lDcfSIHCILQxmGTF2RUZqKzYjPgd6SLvyOyhNiDhk7LRFdxojttPj3kDxqtBr7AukBjefDawqKIczH5GOciZKI7OT6/WXtFH4vTlyge4vf3R0SvUl/vJKCwrgVWhxOxkWEY1zcZ0RF6cBxQ2UhudC1MFu3RFjlZvJDNFhc1+vkQUUm8l8KiQp81xxUnIz6EbFOsr9X/aigOnhjtax66AOtEt5vlI11DEOG71GS2we4I3IMrIQ992sgs8WEPOzqCIIjQgUR0ImQ4Jt6QJHhZUx0hF91bcdEmsw2jdEWI5qSb1R3OET49LRbEWcHNIzrR25g4F9GJ7sVNz4roDisvpPsKG+UCeC4sCngvDOVCKi7q6tgzTnQrF46UpGR+vcRoTOifLC5zRDOOOW+FRZVt85PB6Z4jXQQXnU4HJHkaDdAOUmIjxCifoN4YsW6v2jx3cYXN/s4cAUQn4qRFdp5ZgMp9ntcPpTx0AVZE5xzaOfkEQXQflEWhvTrR/RPRhUx0353oLhG9KzPRge6di95QJJ9u6gZZ3hwnv8+KS5fiXLyM4oyJkGIwfC4u2lQBODWEJnYkXEQskOWhtlBib/l0d3gvg8QRV5RLbnYCwvQ69E7mHzBRcVFtPI22EOolsU70hg7GuSR6KCxqaLGKJiLBBZ+sdf3UcqI7bUBTWbva1mFYo41OD/Qa73l95Xkd4rnowkNHjpMiXju8T4sdhyr4B66TBqV2jxpaBEG0GxLRiZBBuPkRbob8xVcRvbHNhsn6w7J5O7gRaPUiolvtTlhcQy0TovibMiGyRRbn4ksmOgD0myyfLtnmeX2WZoWIriwsGqcU0X0rLprhEtGFbHq2o9asT4ROL11Spo2UhHu9N9cSe3xl2/xkSCb/ORfUqIvoNS4XfXJMBMLDAnsJ1Ol0GJDeDSJd+k2ST7OuEqcDKP9Tmj5Zo1wE+ivPMy/RSaGUhy6gzMmnSBeC6P5YFb9hnjLRAb9EdI7jRDHH90x0V6ZvoOJcOM67Ex3o3rnoShG9uZL/jQ0m5kbAyQh7sWlotggCo+f7Tr1ehzhfctHZIoONpcCm19XXY0X0PqcCYR6OT3EuIkfFoqK8waGPIKJTLromzR5GWyid6A4nJ/bL2l9Y1CWiW9yvh4ILvU9yjDi6QzMOy9NI4GDlorMieuYo9we4SjJyAT3zvodwLrrDyaGF6e8H6vfuj+IGODm+1levpJjuEf9JEES7IRGdCBnyqzvmRBfysk94c6K32XG6Pk+cNkX3QhmX4dWJznY44qL4TogglLOFZ4TXHjPRAb7IUvIAadofEV0Z5+KpsCjgU5wLAGQm8E6xmiZBRJfE79ZwefFPIRcdAGKSGNeSmmDfEjgnulhcVONzPlTBd05ys9v3PfLGgNRu4C7oMxEAUwSIjXSpOSIXZ5SC+8lGUl8gsa80rZYhzxJqeegAkDZYPk0iOkF0fywKET2ATvQmsx0OJ595neyjEzPgcS4OxX5C0omuyC3mHHyecDBRjvaTFRb1Xk9IEP08OtHH3QibLkqa3vBvIG+lfJ3f3wVqGEOKpzx0gP9+RzGj4ppOXhE9r5I3DQ133af2TSER3RuiEz3K/Xo2zGW+Ol5jgsPJobHNJpZaSG7niFRPmejHalx56FnSNVscyaO8frLGoijFqFBDEHLRnQ75yGdveegAEB4FpOdK01WhK6K3WOWfp1uGfTvZ5YpymTSQ7+OKIrqBnOgEEYqQiE6EBC0Wu3jzmJPVXic6v90JjZgPgeZWM05jRPS6NP4GotXqoUMB6UYqJiJMdDhLcS7Sk+xm0YnuQ3FUNq+5ZIfvDifWiR6Twt/gsCgz0bWqwysQ4lzE4qxMnIstKlm27tDMeEwamIqEqHCkZzHDdC1NgF15E8lmoiva5idCZr6WE/1QOT+cbnTvJNXlHUXo7FQY/ciwDzQxyXxMiwArDJcpMtL7nuQiOiDPaS3Zrl3ILhTz0AH+GsCeVySiE0T3x82JrhTRk+XTfji2BSEnLjIM0UyEhycCHudiUwiC4S4RPZSd6ADQFORcdKUpIi5dNG94c6IDUhyhx9GX6UPxPP4mn7d0DlB7lP/9XPNPYPUz8uU5070eW+ZGbz5541zyquQieu8kinPxRpOHB0X9UmMRFa6Hxe5EWUOrWDAyITq83SNSEzzEuQhFYYcxI6eTtEbysHEumSP42COBYDjRa/Pkvz2+iOiAPBc9hJ3oJsXnGajiokIe+qRB/MNvKc6FnOidhqU5+CPDiB4LiehESCAMwctIiGp3JfVBLnG1xNDqsVBIcvNRJOqkG1VjJi8ytnkpFCkNl5Vu4AS3ebPMiW6TLfMIK6JbGn2/MWFFMtZlKxAZKw2dBnzORBfiXGqahMKikvONi051W//L2ydhyxPnIz5F4YRnxHc4nfLpADnRC+taRKcdywGXiD6mb+eI6OJ71BxEER2QO8zL/5AKFJUywzSjk9yjPk5G2Oik1jqgXqMoUijmoQuwnzOJ6ATR/fGWiR4RI/8d98OJbnAJ4f7cT6XE8fcsjW021d9Wv7ErfiPDQ9GJXuQ+L9giuvJ+LjYdJg950Upio7zHubRY7JjfNhmf2C+VZlqbgQU3Aj/MBba+Jd/gnMeAfj5EnyUyIvpJ6kRvbLOJpiExzkUwZ1BhUU2aPTwoCtPrxL5BfrWpw0VF2eM0m23gFMYLyYkujXgVjuXmbGYfesVlACmDpOlgONH9LSoqwOaiN1cCLb6Zs7obyhE4bvE77cBsc2BvqREAMGkQ38cdSCJ657LuZeCVvsCHZ4Xsd5Ho3pCIToQE+a489Pa60AGgV2I0osL1sDk4j0MiB5rkBTxbe/ECm7c4F+HpdTzTSRGd6K4ODMdxYpFR30T0M+XTxb973wbgIzsEWEcyiyyn3EcneqJ2YdHweHfxOyYyjHdfuBUyZY5nNgIc81Aj1l2M94e+KbGIDOMdJxWKz9nh5HC4ko9zGdVJTvTMRF4IEHPjgwXrMLeapGHVbNZh39MAPf0MuOWil2rkoodiHrqATETXeEhAEET3wVsmOiB3bfshogtOdF/z0AEgOYZf18nJR9e1G6UTPcKHTPS6Y55HC3U1hiL3ecEuiOkhziXeS2FRAIiL9B7nItxDv2q/EeUpzL2GoQDYv1C+8vRXgfOf9aHhICc6pP5O76Ro0b3chwqLekWIc4nXeFAk9B+P1TR3uKgoIPXvbA5OrIcloOZEFzPRlddOmYieDqSyInoQnOhsHyEqCUgb5tt2rBMdAKpDs7iocmSBIQAjr/aXNcLqcCIjIQoDXTEuQpxLWUMrrHZtYx/RDuwW6UFuzWHg+JrgtofokZB6QoQEQpGWYZntz7HW63VicVFPuejDzfuliYRe4g2Nt8KiQocjgemkJDJOBQCw2J2wulzwPsW5pA4G4qUCnW5OWDXMjfKK7loiOits+/iUVshEVyssGpXoIYbFk4iu7PDFediPD4Qxn3OBIrqnoNYEs82JuMgwDHatE2gy4hXFV4OFMn+0dAefR19/TJpHUS48WaPkApVWcdFQzEMXYHPRmyvc85YJguheeMtEB+S56P6I6IKI5IeIHhmuF0XYgES6aDnRI2IBPSNuCU70/F+B904HPr8YWPdix4/fUexW+b2WQKPKvK5EGefCiOheC9pDEto9OdHLGnj3pANh+H7wS0Byf/eV9OHAVZ8Ck+/2seFQiOhV3edhSReSV+let6d3shQT6AzEKJAeiPQdV+9bCa7wY9Um8frV3qKiABAfGQ6dq/SQYJQCgMZWm2g0GsqI6MlqcVgcJx85EpvO9/sEGor40bpdCetE73OK70abrDHy6RDNRW82yx9yBOK3bmch38+dNDAVOteXJjMhCtERejg5qnUQcMxNgIPpg9uD3B8neiQkohMhgeDMGNYBJzoAUVwtrNUQ0Z1OjHYwhZAGnImYSB/yISHdwKk60V3uc8G9pddJbh+P6HTySJfi3713Kmry5NOZI9XXY4Vtf+Ncmi2AtVXWCY5TRrbIjuUhg12lw9dRxFx0xed80BXlMrJ3IvR6ndt2gcDNrR8s0oYAMYyrv3Sn+zBNX4ZXnwzow+S5j2rFRUM1D11AGdsTDIcTQRC+4y0THWi/iC440f10YgqRLgER0d2c6C4RXaeTu9HNRl6wXvkYX7gTAPYu6PjxO0pjqXwUnUB3cqJHJcGmCxfjCP0pLNri4Z63jHFEl1tigBu+lWc5R8QBNy0Cxl7rX9sTmfo5Dqt6IfoezhEhD72XVGQyOykaeh1gdThRZyJBSI1mL5FFgiv8WE1zQOJc9Hod4iPdc9GP1/KfX6+kaFm0jBTnwoi0liZ5geW4DLkT3d4GmKra3Ua/MTfxmegCvuahA0B8BhDP9ANDNBfdLc6lpeOjrnYW8b/NkwZJfTKdTocBqUKkCxUXDShu906dY5ojTm5IRCdCAmFoXE5W+53oACOiazjRuZpDSAZz8R14JmJdN0neM9Hdh8sKkS2CS6FJvMmL8F3EZUX01nqgLt/z+jWH5dOBjHNxieiNbTaYm2tly+KSM7Q3VEa0eHKid7CwKCDloiud6AfLeYfP6D6dE+UCSO+RyWL3Woy2U9Hp5G700h2KoqI6oM+pXd6sbgsb6VKX7z46I5Tz0AF3EZ1y0Qmie6PMRPcqoht93rWhnU5MsbhoAIQFdyc6k+/O5qK3GYE/vwSMxdK85krAEYA2dIQGjbziYIvosniINJmjXCvqgiXOh0x0VkSvb7EA2WOA674Gkvrxxo1bfwKGXuB/21knOhCykS5L/ijDmsPV7dpWcKIPZ5zoEWF6ZLuiAsm1qo6nTHRAcqIfrzGh3iTUhGh/nAt/LFf0ESOi57v6q6wLHWDiXFqt0mgClSLAMic60LWGh4o/ATBGLX8jC9lc9BB1orsVFu3gA2O7w4k/iviHgacNlPeFhUgXykUPMFaFxqN270QQHYREdKLbY7LYxZvGnA7EuQDAYJe4eqJOPcrAWrBZPmPg2YiN5DsU3gRRk8oNnBTnwi9rFPPQfXChi21QiHVspIQabB56ZDzfqVFDFufimxM9KSYCkeH8ZaOxXt5B0HtykEcnAzrmcsO6i1TyOzvKkEyXE71GKaLzTvTRnZSHDvAPUWIi+O9MTVOwI12YuBZjMZD3szSdMZwvLErwqMXfsIRyHjrg3jGjXHSC6N6wbqqIOPVh9R3NRPfTiZmiFknQXrSc6ID872quBDa9rtiYA0ztEykDhlpRUSD4Ijp7T8VEuURH6BER5r3bJ4yS9CXOBQDqXIIkhl0IPHQQuGcb0LedD+gTFSJ6CBYXrWxsw6OL9+H+BXv8jl5xOjlRhB3BONEBqbgoiejqSCK6ev+qf2osIsP1MNucOODqC3TEic4fS97HA9g8dHl/Vci3d3LM+t1NRFeOVu3jhxMdkOei1+bxI4hCDMGJHuXq67oVgvWTI5XNaLE6kBgdLotoAoCB6VRctFNQOtHVovAIooOQiE50e4Q89MyEKPEmpL14i3NxFkpCGReXCaQNRYxLRDfbnB5viE0WV1EbWSa6MNSPr94uONF9yaUUSc+Vx3J4Ky7KOtEzhmvn2bFitdkIOLy7pnU6nZj53WSoUezPQ0FQvV7+N7A3jsoomQ4WFgWAwenCwxLpc3Y6ORyq4G+cx/TtPPFYp9OJkS61wR52qxSG2e8GRbnI6TsR0IVJ08rioqGchw7wRfvYB2rkRCeI7g3rRNfqBCod2z5iaGmvEz2AcS6+OtGLNqsL5o3lHW9DR9AS0ZsrAKfnkYudiiJjmR0B6QtCnIunwqKlBknIFb5LASGht3w6BJ3oggjeZnOg0c8CvOXGNpgsdkSESbV9BHqfRMVFP99SiMcW7/PrIUSzl+95mF4njlL9s4R/4NiRwqL8saQ+HgDUmyxYm8dfq5Txo1HhYaIpS7x+Kvs/cRlAYh8gjLkuGzRGvHQGrIieMgiI89PUxOaiO23eR053Q4QHHH1dD60aWjs24mmHKw994sBUhClGoPdPFZzoFOcSUCjOhegCSEQnuj2BykMHIBaTrGg0o02Z98hxiCjbJk7qBp4J6HTiTQ8AmO3aHSOTigtCiHNxcny+pJCJ7peIrtf7l4vOOtG1olwA9wKebb5lTwq56C0NChE9xov4LctgZ+NcmONGJwFhHbupBaRM9Npmi9iJKaxvQYvVgegIfacVFRUQIl2C7kTvPYEv7qUGFRWVE5Ugd9GUME70UM9DF0gbIr0mEZ0gujdsYVGt4chsnIul0aeH4YAk4qT6K6LHCU70AESp+OpE16IpyCK6lrjltAMtterLugI2iiwuzatDV4kQ5+KpDhDrRK8PpFkgLkM+ajEEnejsCEh/88uPuKJchmYmuI0a6JN8cjjROY7Df1cfxeI/ynDY9X74gi/fcyEX3Wzjaxl0pLAoe6xmsx2GFitu/nQHiutbkZkQhWkj3etEuY3kUV4nYtP5Gj3JA6R5XeVE5zignBHR2zPakr2HBkIyF114eCgI3A0dfEi4p9QIwD3KBQAGpvF90SIS0QMLxbkQXQCJ6ES355ggoncwygXgb5iEDDy3H63aPISbGUF3wJkAgOhwSUT31KkQC4syTvSocD0iwvgnz01tNjQJ1eP9iXMB5CJ6c4W2A8pUK3c2aBUVBdxjU3yMdBEEYkuT8uavnSI6e9wARLkAvBMly+UGP+HKRReLivZKRLgPQ5o7glSA1exlzU4mMhbIHqu+LNTiSLqCfkwuesUeqaL7xtfk64WsiM7kohsozoUgujWsm0rLic6K6ABgbvRp14J72N9MYCkTPRBOdIXAqOVE1yLYsSkNTEY723YguAK/wonuLStaiTcnuslilz1EabE63E0p7SUsXF6csDkERfRaVkT37zzJcxUVHZHt3t8R4lwqeriI3thmE/tavo7mdDo5mKzeRfQchRkrUHEupQ2tuPnTHcirakZGQhQWzJmMNNeoXRbhemsURigo+11CH4iNdOkqEd1YLBf1/SkqKpA2jI87FKg60PF2dTGCIa6fS0Q3ttn8jmViKTPwDxyHZLibt4RM9FJDGxwdOAahwE1EJyc6EXhIRCe6PfkBKioqoFlcVJk17hLK9HodoiP4U8VTR0EsLMrcwOl0Olkueruc6IAo6ItoRbr4WlQUcC/g6WtxUZc4bTdJ63PQe8/XZkV2rcKiASgqKiAM2zzhiu4R89A7saioQGYC76iraQ6yEx1wj3QB+M8qPafr29Ld6c+8Vw4LULEX2PstsOsTaX5cZug+gGBF9LYG+SgQgiC6F2ycS6TG/Y+biG70adeCCOq3Ez2gcS5+OtFTBsrfh2CK6BwnNzP0U4zsClbUjLUVsDH5urFpYtRgoo9OdMEIopWJLsSJJEaHizVy6lsCeK/DFhcNeRHdv/flqEtEH97L/XwX4lzKenicS4VRMp/UCffQ9QXA4tuAX59RzdlusdrFAbqe+ldDFWasQBUWfW/9cRypbEJ6fBQW3DVZ7H8oEUR7MWebFdFjUviHSIBCRC/0PPo4UCgF7z7tqGsQFg5kDpemQ9iJ3i+FF7gdTk6Wee8vFY3891k4f1l6JUUjIkwHq8OJqqYgm656EiSiE10AiehEt0dwoisdBO1FU0Qv3iq+bNIn8XniLmJdhZY8OdHVCosCUqRLk9kmZaLH+Hnjlj1G3nlk2iqDjXIB/HOiK7P5NMiI5zu6HCPA2aOS+CGInpA50TUKiwbIiQ5IkS4FohOdHxbaFSJ6RneJcwHUs8/7TNTOyj+ZYZ3oALDzY2DF/7N33uGNW2fWP2AXKUpUryNN77bHM+69xXFLd6rjZOMUJ9mUTdkkW5Jssl82m16cTe9O4hSnF/fe7fGM6/QqadQLVSh24vvj4gL3ggAJsImS7u955hmCBElIIoF7zz3veT/E3CEBr/zO4stDp7AiOiAiXQSCaqYQJ7qF5qKZjKyKOLYbi5Y0zkWfic6cV42c6Bf/B1DfrW0vqNt7Akgwixx6o8NCCfx6M0Sg2bBKMhe0sSgdr+qhUS4rGv1oVj4PEzYd13pOhKNq1SDqmFz0xRjnwvRcsht1s3eYjFM3tNdlPda9TOJchqa1n29sLk4WE3/2cuDFPwKPfQu45zNZz6GfcbdTUhtCGlFqJzo1TWVkoLnWg1veeSbWtprPValoPxVRvlvsvCvQot1mRfTErGWTU1Hom823bCjsddhc9OEXKrMAUEKoIa4x4EFAn2Fvk0Qqoy6kddRnzxtcTocq1otc9BKiz0R3CxFdUHqEiiKoamZjSXUVtxRxLoCWi36EbS4qy8AxTZg+4j8FkLQGIDVumhFpvhpNV6+DuokKdSrMRJOYiSpxLnad6A4n0MMIfKYiOuNEr2kEalvNX1OfiW7Tie5gom8kK81ATTPRSx/nAmhO9MNjc5BlGS8oTUW3dlbCiV4lcS6AsRNd75oTEOq7gPoebfuFW/nmd5f8J7DussofV6lgM9EBIaILBNWMpUz0EL9tQUSfiSVBK8dDdkX0ksa56MRAVw4neusWYOu1OoF1AZ3o+ki9zm38RH2hBP6sRu3NtjPRacTA0bGIYYwBdUJ3N9SokRXFNBeVZRmv+fajuOamh8k4mnOiL67GotPRJMaYCkQ7cS6Hx+ZwZCwChwRs6cwW0Wmcy2wsZbrAsRSgcz4AGJ9NAPf8NzAzoO3w1I+yYlDYyCJJ4ps3svQ0+uFh4hyLFdFblM9/Y8CDX77jLKzLUzGd7URn4lPYStzGVfwTKxHpwr5HoJX0CSoENhd9fty4KXQVozWodanXx0JF9JGZGGQZ8LgcplVfPU1URJ83fFxQAKwT3eXTKjwEghIiRHRBVXNIadDTGvSivsiyO8pqGvMxzkxQJw4BEa1R5kDdqdxzaHPRnHEuyoW3VjdR4eJcVCd6ASd0Nhd96phxuTDXVHQztxCQhS/EN3CKWBTRFYGYdaI7rXRwZwXyVJSUHQO8K91uJ/gcaCJ6BH2T85iNpeBxOUrSoDYfrXVEDBirhjiX+m6grou/r5Csw+VCj8GiAwBsvAY478OVPZZSU98DOJjzqBDRBYLqpUxOdCp4Br1aHIdVtMZ4JXaiu3z8eIV1nAPApZ8k1VPVKqI3rKqOY9OP4wJN6rjTaib6quYAPC4HIok0+qeyhZ1+JeN3RYNfFYbsxpawTM0nMTwTw3wijeHpKFDHiOjzE9nZ+VXMkTHeAWkn5ubXT/YBAC7e0Ipmgzxtv8elxikt5Vz0IeZnqx3bRSoCWVJR4PFvc3exwmcuXE6HWqXqcztQ48lTQZuHa3d048MvWY/fvftsbDDIsdejOtHp+ZNrAsyK6IwTHai8iK43XNihTddcdHhxRbrQqvJanwsNgeLiy+j3tKPeZ7q4I5qLlgFWRBdRLoIyIUR0QVVzsMR56IBBnIssA098j9tnrIkXGVURPZkjzoVmops50WPJwjPRgexmhn2P8duyrBPRc+ShA2RCWsPmlFuMc1FE9EBaa2BmyYlu5HxPxnihoAxxLscnInhG6Y6+qT0Id5mbigKaO6UqRHQg23neJUR0U4yc+83rSYzLYo/Acbp4h5MQ0QWC6oV1opu5AgsQ0akg0GAzD508R2mMN5+AXGyZfkonorOsPB9YdzkgOYHT3g6sv4Lczy4Izw4BmRI1tLTL5FFmQwJCPToRfaGc6DoR3d/ERA1aM2+4nA419mLv0EzW47wTXYlzKcKJzsZ3TM0neSc6sKhy0dkoFwAYm7X2e4mn0rj1aeK2fuMZPab70VzlE0s4F50Kj26k8NrBLwIwOM888X3uXGen2oK6xYt1oQOkkucDl64zzUA32h9gRFnWic7OkepXkHMfpRIiOhvn0liEiN6uE9FHFldzUa2q3M1UXhW2aDykVFUYRblQepQGpn3CiV46WF1BiOiCMrHIFQHBUueAkodeSvcwXfUNzydJSfKjN3GNA0fkEOJNvABN3Qq5MtHNcid5J7oS52I3Ex0AOrYBLqYxiT7SZXqAz+jMJ6ID/KDNapyL0jSzAcxFqsZmnAt9v6MP5N6nCDrra+BzO5BMy7jt+WEAlclDB7TIm4lIAsl0piLvmZOtr9Fur73MuGmbgNCjy0X3BIHX/xLwZZdXL0rYXHR9BqZAIKgOZJm/nps1FvXW8WJLNJz3pScVQaAgEV0RFVIZWc2OLZgkIwS6dU3XHE7gTb8FPjkGXPNVzaXOCtVyeuGiAlgnel0n6ZNRDXntOeNcrI87NymZ3HuHZrMeGwgTsae7wa86pu1mf7OMMA31JiOJbBF9EeWi0x48tGLTqhP9zhdHMDWfRHudDxdtaDHdr2sZ5KLTOJcbnX/FitRx7YGGldrtxCzwpDZvU6stvPk/4+uVzHK7UValgFYSTEeT5Bxvlonu8gChFdp2uUX0xDwfnaSPk7FDTQNQx5wLF7MTvcg4Fyqid9ZnNxWlrGwmIvoxIaKXDs6JXv7qc8HyRIjogqrm4Gjpneg1Hic6lVXh8BO/AO76pPpYBhI+lXwb6v18KSVtLGoW55JIZRBPEbFU7zJnM9FnVSd6AXEuLg/fJPKYTkS301SUworWEWtO9OZaDyQJaJBYB7kVEV23zzO/BH7zZv4+K8K/RRwOCaubycXz3n0kqqdSInqj3wOXg0z6iylzLhkbrwFe+zPgsv8CXvndhT6a6qZ1M9NUWAJe9R2gZf2CHlJJYct0Jw4DmSpY5BEIBDzJKCAz302zOBdJ4hdFrTjRI7SpqP3FfJ/bqfaICRfozkumM7jjxWGMTYW1O/VOdID8bPqG5fposoWKTWFFdCru6RtiLsS5lR3HuXyAJ6BFXVhsLAoAGzuoiJ7Did5Yg6YSNBYdntbGSFORBP97BBaXE12Zs5yxiox3rY7/blGiXF53+gq4clRL0lz0pSyiD01HsUY6gfe7/qjd6W8CbriTVHxQHv82aToKxsRkYW61vZdU79D+WJWEE2VjYSDDLET6ddW6DYyQzVW+lAG9SF9MnAvAu9FH7IvosRxV3+Ukk5Exl9AMcY2BYkV0Jc4llMuJTj6HfROR4qu7BAQR5yKoAEJEF1Q1B6kTPUe380JY1RLA+Y7n0PvQv3L3/yT4HtyROT1LCM/XWDTCOLICXn7SR13nM1wmeoH57r3narfH9wNzTCkg21QUAFo3Ii9cs89J8/0YXE4HWv1O1EnMqrndxqIAyTlMMwOTza8EunZYOgarrFE+NwnFDV6JpqIAEfCpQ2t0pgpEdEkCtrwSOO9DQK25y0kAItq85S/ASz8P3HAHsOllC31EpYUt001FF13jNoFgWcCWIwO53VS+kHbbSiY6jXMp0IlJ3ZSTNoWF8HwC37n/MC744n248ean8eRB5tyjd6KboRdYF8rxPcWIWkYieibJRzVUCq5RezMgSbYbiwLApg5iXNk7zIvos7Ekwkqec1eoRhWZiolzGWbiXCbnDZzoi0hEP6LERJ65mox3rSwuHB2P4NHDE5Ak4PWnr8i5b9cSj3PJZGSMTM/jf9w/gldi5lsv/TwQbCNjWEp0Ctj5YwD24lzOWdOE37zrLPzPq04q6bFbQc1EjyQN+hfoRHQ2F73c0XuTuqrEYuJcAD4XffwgX3WUhz/sGsCmT92OPz9T+XP7fDINqmOTxqK6DHubDIZpnIv59W1FYw0kCYgk0rYaEQtyIER0QQUQIrqgapmNJdWyvnwdz+1ynr8P33F/HQ6ZGaSd92H8znklAKBeJ3KrcS4mq+N0AFfjdma5SOoYJ/pMtIg4F4AX0QE+DoV1ogc7s7NSjeBEdGtOdABYGdBd6AuJc2HZei3wmh9afn+rsE4Tt1PC+vbKlXXRSJeqyUUXWCfYBpz9XvMmo4sZNs4FELnoAkE1EtfFaJg50QH+Wm/DiV5InAv7PKvuvOHpGP7zT8/j7M/fiy/cvk8tcffIzPONnOhGZInoC7AImIzx70sdo1ku+QUQ+FkzhGJuKCbOpX8yqjrZAc0BHfK7EfS5tTgXGw009QwzcS5TkQSJTvMy8WnlduGWiGQ6g+NKc8AzFSf6fCJtar6h/Pop4kK/cH2LKpKbsdTjXMYjcVwhP4ozHfvU++K9FwMnv45sbLuOX2R59FtAUvuMWuk3JUkSzlzdhPoCKnGKhS5chucT2YtsehGdHavFwpaNTgBIVIz+GpILfbSfvrGpXTpOZo4lDQw9Z/mpDx8chywD37n/cMWd2TTKxeWQ4HU5mEz0Ip3oOTLRvS6nGvdyXDQXLQ1cJrqIcxGUByGiC6oWGuXSVufNErULIhUHXvg98PNX4t0H3oVaiWlqdcqbgEs/RXLqANTV8G4GtbGoSZzLbJw8z6iUkE5cxmbjqiO6oDgXAOg+jc9Ff+TrUJfNWSe61VgUfSa6xQHLqoBuwmTFie6pBZwGk/btbwFe/X3AWfoB7RqmgmF9WxBelzPH3qWFZmKOChFdUE0IEV0gqH6ynOg5jAR2RXRF/G4sVERnhSAL/Outz+IXj/chmkxjY3sQX7r2ZAS9LnjBPN+qE91Xz0+KF0Konu4H1+xQdaJXQdQMG+eijO/URnk2xp0NAQ/a64jws39YE+P6J4kotKKB5PiqjUWLcFDSRRVAy+tHMxOhNrYPi4H+yXkk0zJq3E6sbamFz02m2Ll+N4lUBrfuzN9QlELjXAaXqIg+FI7hEududXte9uLYWf+t9URweYFzPqA9ITIK7Lq5oGqLhYA6myOJNJKzehFdVyVa6Fhtbgz4zjnA//YA937O2nPYOJfa9tyLtlboOo3fHnjK8lPpnGnf8CyePzFd3HHYZI6Zy0uSZHvBWI/WWDT39a23iZxPj4tc9NIgnOi52f0L4LvnA5/rBL64Bvj9O4Fwn/Xnj7wI/Po64EvryGv88DJgz1+M973ns8B/1Rv/+1yn8XMWCUJEF1Qth0ZKlIc+Pwnc/m/AVzYAt94AHLkPEjMBivZcDLz8m4AkYUYR0U2d6CYi+lyOARx1nVPniEMCAp4CB3ruGmD79dr28PPAvr8BmTQwtl+736qIzmbwpROWnQsrvLoBvBUnuiRlZ/6d+R7gZd/Mzj0tEWtatIvnSRXKQ6e0qCJ6LM+eAkEFCbYDbmZQKZqLCgTVR1wnopfQia42Fi00zkURFiYtZqLvU0TYr73+FNz2wfPx2tNWoN7vhldinm/ViS5JuuzxBRCq9c5oozgXYGGOTR/nAqbpok2BUY10YXLRB6ZoU1EiCjWpjUUTBbtG2caiqljVwsQRsmPbKubwGBFuVrcE4HBIaAoo1Yg5ctHv3juCiUgCrUEvLtnYmvc9qBN9dDaORGrp9TMZDEexQtLE5Z2Z9Rh0tPE77XgrP5d45OuYj5LPZbWL6HU+N5R2SYiGh/kH9fMjfS65VRH9ns8QU5WcAR76sjUHOyuiF5uHDgD1XaQimnJip+WnsueD3+7sL/5YbDCjm8s3sPE7Nokl06RRMoDOHJnoANDbRMbkwoleIoSIbs5dnwb+/M9E8znjHcDqC4nB9PsX8b1ezDixC/jR5cCR+4GNV5Hz8fQA8NvrgSe+l73/8PMAJODCjwMXfoL/d/6HS/uzVRghoguqlgNqHnqRIvqtN5AGNAaTywfTJ+GWlf8NON1IpjOIKCK5viTQ71Yai5rEuahOH4PGTfRiTMu6gj43HHQUVQjnfRhwMo1P7/s8EcLSzEDdSlNRIDtixWKkS4dHJ6LnimphYfOlL/hX4IrPaw6TMrCKiXPZUnERnQyahBNdUFVIUnZzUYFAUF3YyURnRfRYOO9La070wqq/qLBgxYk+F0+pkWaXbGyDpFzv63xu+FBAnAvAi9XTC+BE1080G5U4l5oGvlJwZqBih6Sic6JnMrI6PrXSdJFlk9JcdM+QZq5Qm4pSEV1ZUEmkM5iN544tMWN42khE36DtEBm1F2WxQBweI9/ZNS3ku9oc1BYYzFAbip62Au4cDUUpjQEPfG4HZFmbUywlBqdj6GZE9AG5GeP6MbQnAJz9z9r2zAmsmHoSgL3IooXA4ZBUk1Z8epR5RMqu6A31Ag7mO2tFRB/dBzzzS21bzgADFgRsdhxYbJQLpZvpcTXwtOWnsSL6n58ZrGiTUWqIq/WSvxHXCNYm9LxW43bmraZXneiTwoleErg4FyGiqww9RxIMes4GbnwIeMlngWt/DLzu5ySN4LZP5H+Nv36AJDvccDvwsm8QHefdj5BYu7s+lW0eGH6eGA0u/nfg4n/j/13w0XL8lBVDiOiCquWAEueyrq2IsrL4LJ8bDhAX5qnX4x9n/hxvSX4Cf9/Pd3cHst0M+eJcck1SqCCfUUw6+qgY29R1AKfdoG2Pvgg88AV+H8txLnoR3dpEpdWlWy23EucCkJPtG38DvONe4JL/LKuADgB+j0ttSntar4WM+BKixrlUQ2NRgYCFLRMWcS4CQfVhKxM9pN2OTuWNZVMz0Qt0oof81ImeX1igzrrGgIcTEuprdCK6246IzsSmLITbmxXRPbWaiWChXfLpFL+I4m9EJJFSPw5W8qJZNioi+r5hIyc6EX18bicCyvh4soBIl/lESnV/Akz2MOtEBxaFG/3wqE5EVxYYxk2c6H0T83jo4LilhqIUSZLQuYSbi45OhtEmhdXtAbnVuNniaTcAkiZh9EaeBVD9TnRAO++mZxkR3d+UXZHrdGn9FgBrY7W7/4sI5ywDT+Z+TnwOmGNc8SUT0U/Xbk/3AbMjeZ8SS6bV80FzrRezsRRuf2E4z7NKh94QR6uuwvNJ25U2gzQPPeRTF4/NWKmI6MdEnEvxyDJfyZcrCm+58dQPyP8XfhxwMeO/TdcAK88HDtyee9zS9zgRxbe8CmhnGjMHmoALPwakYsAzv9Luj0yQxuDsvksIIaILqpaDihN9fTEi+uBufkBx4ceBjx4AXvEtbD/npQAkPH18CqMzMTUPvdbrymoO6lPjXIydNjPq6rVRnAt/n92JjCHnfYh3PL1wK/OgxLt4cqF3kEesOdHbXLoLvZU4F4AMEjdcwTsUysx33rwDN7/9DNVVVSmoiJ6rlFcgWBBYEX3qGHEVCASC6qHQTPRMKvu5OiaLzERvVJ3o+Uvcj42TsQJ12lHqalw6J7rFTHSAF6pnB4FMhWMtppg4l4aVvBmgfgEF/qjOBOFvVkUht5M0yrPDZiXOZf/wLDKKC0TvRAeYSJcCmouyLnSAWZhp1Yvo1Z+LrjrRW4nzUW26ajIGpFEV569rwYpGv+E+Rizl5qKJiePc9oDcolaycNSEOGFmfewFANXvRAe0XHTZoH9BFs3rtNv5qgaPPQIcuC37/v48Ijob5QKUJs4FyM5FtxDpQk1HPrcDbz6L9AioZKSL6kTXxbkkmEp1lvv3j+Lp48YRakNhcm7rzJOHDog4l5KSipNmthThRNc4/hipbuk9N/uxVRcCkIGjD+V4/qPk/9UXmjwfwNEHtfuGyeKmENEFggoRS6bxtbsOqA051hYT56IvYzv9naqjq73eh1N7QgCAO/aMmOahA4DfbS0TnZaAsegHdSUR0YNtwOlvN36sYaX1i4Y+g29+wtLT1rCNRT21/IpmlbG2tRbnr2vJv2OJaVWaco3NxPLsKRBUGHaRTU4D4wcW7lgEAkE2hWaiAzlz0VPpjGoYaCi0saiNZmvHFFFgVRM/JqmvccMnFepEZ0T0TAqIjJnvWw5YJzrNQ6ewLvnpCse56E0QgWam4aI7rxtSz8qmALwuB+YTaTVmgIrorOhLm4saOobzQEV0KlbNxFJIpjNAXTffu6PKneiyLKuZ6NSJnu/3squPfE+vOanD1nvRBYylKKLrm9sNyM2mTn70nK3eXJc+AC8Si8qJ7mDjM/VNRSn66D2zBUNZJlEKRpx4mvTNMmNSJ843lkhE79wGSIy73kJz0RGlh1RbnQ/X7uiGJAGPHp5Af4ViTmgkFTXE1bid6uLjlK7yamg6irf/bCf+6SdPIpXO/rvQuKWO+vzXth7lfBqeT2LawuK0IAcJ3UKEENEJmTSpZqnrMtZsaCxdrvkgfYytkKHUdZJYvvGD2n3Dz5P/47PAL19Hmph+rhP4yVXAobsL+zmqCCGiC6oGWZZx54vDuOyrD+Ab95Av4etO686bJZYTVkQP9QK1/EDlii3tAIA7XhhWJ5ZGg7D8cS7mzw16XZxRqeg4F8q5HwTcBu4Vq3noQMGZ6I4Y43iy6kJfZrBO9EIbbgkEZaFtK7898uLCHIdAIDCGdZNLztyZ4TZE9IlIArJMGpyHChxbNdiIczk2Tia0K5v5iWxxmehd/PZMBXPRZTmPiM665Icq65LXj9/8zZiNaRWWdnE5HVjfRkws+4ZmMBNLquNk6oYGoDbQzJX9bcawYjLY0B5Ux8nh+STgcAAt67Udx/bafu1KMhFJYDqahCRpvXioE91MBO5XonH034180N/9wBKMc/HO8QtP/XKLJRHdgxROlo6gbhGI6DQOyxVjTEtmfaXYqsFU1Pxct+fPvNM7wDSpTcyRRqNm6B3upYpz8QSANmY+aiGbneahtwV96G7w47y1xOj1uwq50fVOdEmS1OudvvJqz+AM0hkZs7EUjo5nO8gHlQVCKyJ6wOtCizJnPD4p3OhFYaefzHIiPgNAzh4vUnxK37jYtPlr0MeMXkOSAG+Qf/7Qc+T/x75F/j/1OmDdZaQ65hfXAk/+wNaPUG0IEV1QFQxMzeOGnz6Fd938NAamouis9+Hb123HF15zcuEvKsv8yjebz6bwUkVEf/zIBPqUlW4j0b7GY82JbiSiOxwSaj3a/SVxogNAbStw+juy77eahw4Q9xd7gbEY54J5ZpLur2zW+GKBTqCSaRlTwlkgqCaa1gBOxokgRHSBoLpgM9G9tbn7h2SJ6GHTXQ+OkAnmyqZAVmydVcxEBSOoE10f50Iy0Znnu+3EuehF9ArGpsyNAknGFZlLRE8nLFf3lQT9ewWa1ajBQh26m5RIl71DMxiYJKJtY8CDACPKNyuOa7PYklxQEb0zVKMu6mjNRZlIlyp3otM89O6GGviUylXNiZ79e0mlMxhU4h56bES5kPcg+y+1TPRkOoO6+JC6nXZ6MYaQJREdAE537FtUcS7eBGNGMnWir+W3jXLR00ngns9o204v8Orv8fv0P2F+QGycS7AT8Nj7POaEjXQZ3J3bEQ8tzqW1jsyfXnsa6RVw69MDSGfKb0ZSDXHM+Y1WXk3qKq8OjGhi7d5hXQ8TAENhmolu7drWq5wHjotc9OIQTnRj6O/F5TV+nM4JUzmq5+kChelrePnnO91AfQ9w3a3Adb8ljUxf93PgHXeTMd/tn8iOk1pECBFdUBV8+LfP4r79Y3A7Jbz3ojW4+yMX4qqTOmyXn3JM9wMRpnFL92lZu6xsDmBjexCpjIw/7iYr/HVGcS6KCG7WJXw2Rya6/jWNXr9gzv0gX/IK2BPRAd4BYbGxKJe9aeagWOZ4XA61RHl0VkS6CKoIp5uPdBEiumAZEkum8dU79+N7DxzGk0cnTSvNFgTWTZWvMZYvxG9TJ3omA9z7OeDX1wEDTwMADo6Syf7a1sLdWVQEmpxP5K2yoo3SVmU50Z3wSoyIbsuJ3slvV1JEZ13ogFYCTanr5rcr6ZLXmyD8TUycS6EiOukls2dolmkqyotCNFt/wkJlgh4a59Je59NigtTmosw1anYo5+LQQnNknI9yAYCWWnOH/tB0DOmMDI/LoVYtWoX+/gfCS0tsG5mJYYWkzdlSwW4AknEmOkBiLRnX9OmO/YskzsUNCRnUJBnHplkmuhUR/emf8kLUmTcCqy/mF1f7c0SpsM8tVR46hTWvJeby9jagcS6tQXI9uHxzG+pr3BicjuGRQxZNXkUwF8+eyzeoPUD47zHt2waQRUY9Qzac6IDIRS8ZQkQ3ho6x0ibXaXp/Lud+3teI889/1XeBDz0PrL2U369zG3DWe0gc3wt/yHvo1YoQ0QVVAXVU/OAtp+FjV2xUReui0OevGTjRAc2NTpuDGGai53GiqzlqJgM4dmBXMic6QAZeZ76Lv69ti73X4ER0q050EediBToQpO4KgaBqaGXOE7lKfQWCJcpfnhnEN+89hM/ftg+v+95j2Ppfd+Dqbz6Ez/51T+5c0qnjwG0fBx7/Lql4KwdsJnquPHTAPM5l54+AB78I7Psb8McbAWjuuXVFNGynomkilUHUxFgAEEGCCmC9ukz0Ro8u5sSOE72mgW9EWkmhmm0qCmRng2YJ/BU8NtaJLjkBX4ipkixs3LmxnYjo+4ZnDJuKAmxj0cJF9I56n1rhoDnRdYaQKu7dQZ3orIjeHDSPc6GVr90NNXA47JmFqBN9KBwzzGJerAxNx9AtaXMQKUQaS07NJ0lOvhE956g3T3McQI3TeLdqIuT3oB4ROMGcO81E9No2XpTSR6+k4sADX9C2fSHg/A+TyqXuM7T7B3I0F2Vfs1RRLhS9eS1PLjqdK7UpTnSf24lXbiPn1N9UINLFaNHRLL7swKgmou/LIaJ3WnWiNwknekkQcS7GeOsAyWEe10Lv99WZvwYdaxq9hiyTCspcz2fp2kH+1xsTFhFCRBdUBXSFd2VTCVcMFecVAFKmYtIdmIroFCORW4tzSRm+xlw+J7qPdaKX2Clx7ge1ycb6K/kSWCuwgzerpcecE12I6GbQkkRTJ41AsFCwi22zQ9arUASCJcJRxfHVFapBa9CLdEbGi4Mz+PEjR/HOn+9EPGUiEP/xRuCJ7wK3fxx4+KsAiBj4+X/sVd26RcM50fOJ6CF+OzpFJjQ7f6zdN3EQSERwSJn406zrQvB7nPAoUTC5osqoo64x4MkyJ9R7dL9bO050SeLF6nIL1dMnSHbnza8C/vw+5jgcQP0Kft+FjJphnej+RsDhUDPRC3Xoblac6ANTUdVtSUVcSjFxLmoGcp2PEauUzxTrRAfyulgXksNj2SJ6k9qAN4lUMgkcf1S9ztJGiSsa7EdntAa9cDslpDIyRpbQ2HIwHEW3pDUJdjWthFNZYDDtv9BzlnozKEUhLQJDQIPfgyZJJ7r6TUR0SdI1F9U50Qee4hsrn/8RTehawRjHJo8AcwYNmGMzfMV2qUX0pnWAt5453ty56Oz5gEIjXe56cSSruWepUavKmTl7Q4DGTGnXukxGxqFRJs5liI9zmU+k1P4R1p3oQkQvCcKJbozLQxb9pwdIBJSeScUgoL/usjSv5/dlmRkkUS70+ak4aWpsVgWTUD7ndgwUVYYQ0QULTiKVQURxeNNBdElgV7zbTzbNcNrUEeQyCQ0z0ZWMQzPXFS0BM3OZs8J5SZ3oABkw3fgA8P5dwBtvyZ2dagQ7eLOSiS7LfOMy4UQ3hTaKGV1CEx3BEkFfsSIiXQTLDFoB99ZzevHEv1+KRz9xCb7xhm0Iel148tgk/u33z2fHlcRngb7HtO0HvwzMDOLHjxzF9x48gp89eqw0B2fHie5085Ev0SngxK6sChM5MqY60YuJc5EkSRMWcogax8aVxolN2UJhyK0bS9mdSHEiuk6ofuZXwA8vAx76ir3X1JPJAH98D/C1zcA/PgocvhfIMJPPprVkYsrib+QXBBbKia6M61RnZQGNRQGg3u9GpyIC3buPiG0r9E70IhqLapEHNWhUxSrldUI9fMVBFeeiHx6jcS6aYNPg94CazNO/egPwkyuBm3YAM0OqE91uHjpA+iypzUUnl47gNjY5hRZJczg6GnrVqhdTI0rvOfx23+PlOryS0eB3owk6Ed0sEx3gI130Ivrxx/jtU96g3Wad6ICxC1yfR1zqOBeHA+jazhxDbhGdzpXYiKOtXfVY3RJAIp3BM/3h0h6fDuM4F9oDRDu/9U/NI5bMwO0kX/DhmRh3LaT9DoJel+UqIGoiPCbiXIpDiOjmrDyXRLEYnSeP3A9AAlaclf2Y+vzzyP9HHzR5PrSFzUQE+MGlwC9ebSzaH3+E/E8d6YsQIaILFhx6YXJIhbtlskglgKFntW2TKBeATAiv2Kq50Y2c4jTOJZmWDcsK5/LGuZQpE53i8pLBTyEZ8qyT3IoTfbqf5FhRgm3233OZoMa5iEx0QbUhRHTBMkfLePZDkiR0hmrwim1d+L/rtsPpkPCH3SfwrXt1osWIzumYnAfu/gz2DBJRpGRNpBOMs81KOTIb6RILA7t/nrXL1NggpqNJOCTeMVsIWdEbBlAxwKjCsM6lq+oza1RlBuv4ZoXqqWPAX95PBKN7PgsMPWfvdVlO7ASe/ZXxY9564LLPZN+vd8lPV1JEZ53oJKZPc6IXPu7cqLjRaVyL3omuZaLbMwsk0xmMKe71tnqv1sCPilEOJ9C8TnvC6F7bx14JYsk0+pVzyWrme+VwSGgMeNGCKXiP3k3ujE4Cu36GfmUBb0VjYS48+jcYWELNRefHjvF3hHrQrEQFjZlVOTSuRsLHGIH6Hi3PwZWQkJET3SzOBSBubkr4OJnfUtift2ktUNuqbXftINUyCnfc8Rf800+e5BeGJ3XxMI0lFtEBfv49to+4302gTvTWOt69Ta9X/aWq9DJhzmKci7YYHVS/w3uHtZ9raJp8L9stutABzYk+Ohs3rXoXWCArzkWI6CqnXk/+v+ezQJK5duz9G3D8YWDDVUB9l/FzAWDFmcSN/sKtfNpDZAJ48EvEQHDqW8h9/kZg9UVAfAa4//P86xy+D9j1czKO2/yKkvxoC4EQ0QULDp101te4bWcDmjLyPGlwQDFoKsry0i2aEGzoRPdoQXtGueh0omIe5+IyvF0VsIO3+Aw/QDNCP5HR51YKVIQTXVC11Lbx/RBGhYguWF6cCJNJRJcus/SC9S347CvIItNX7jqAPz/DCKEjL2S/0HO/hnuITCjoJLxoOCe6heiVGqZkfvoE8Pzvs3YZGSQ/R0+jHz53ceHBZjmxLMeUZosrm7MnsUGnXkQv0olOhaF9f+cX+YuJAJnWZfAGWoHtbwHe+Bvgo/uBjVeZHBsr8FcyzoUxQQQUET1eXGNRgFRrsugz0Wmcy2QkgUzGeo+Asdk4ZBlwOSQ0B7xopAsz7GeKjSesUif6sYkIZJmM7envgtJc6+FyvgEAB24vyokOaOesUojo+ZoDV4rM5HH+jlCvOoYeNxtDSxLGGxmn8/HHytenokSE/G4DEd2iE13OaBnC6RTQz2Sd95zNP89by5klQhO7cf/+MX4+MqFzousbJZcCbv4tA4O7DXeLJtJq5QzNRKfQ70lfmaNODJ3oAdpYVFsgPzBCY9FqsYn2jWAiXYYUJ3qHxTx0QMnKV/SHviVUYVJxspzoIhNdZcUZwOnvJD0SvnMucOcngVvfDvz2LWR889LPafsefQi47/NEYKdIEvCyb5LFuZ9eTeLtbv934Lvnkn4xl/8/3lh59VfIue2hrwA/vhK48z+BW95E3OnuGuDan4g4F4GgGKgTvbRRLrqSsTwi+qkrGtTyMeqGYfE4HWo2X9RQRM+diV52J3oxsEIakN+Nrs8cbLWZwb6MoJ+pMdFYVFBtSBLQulnbFk50wTIikcqoYkJXQ/Yg/roze/GO84ig8K+3Poenjys9A4xEdAAfTP4QEjLqJLz4A7SRiQ7wTvQj9/FOdoWJ0QEAwLoi8tApRsKCHupE7zWIc6nVi+huG5noAC+ipxPauGX/bfx+bPScXeK63+F7HgFefhOw4YrcE79K5rWzcE50XZxLEU70TR18ozD994WOmTMyEI5ar8QYZvKPHQ7JuLqBHV/ODOR0sS4Uh0eVKJfWWki6atDmWi86JN2YenA3ohPkc6F39VuFLmQU24Ph5seOYcun78BzA+GiXqcUuGcH+DtCPeqixHiOqKCh+lO1jbnhqm9U1+D38HEuShNgU/QRKxMHyf8jz/PXCb2IDhDnqMLJ0hG4kOIXPlknel13eQQtfVyDSXNRWrHr9ziz5tJURC+3E101xDGLjiGDBeODqogeVCt19jLNRQcVJ3qnDSc6IHLRSwIrojvc2ZFry52rvgRc+SVS/ffE90isytbXAG+/k19EO/Yw8MD/EmMCS+/ZwNtuI9Eue/4M7L6Z9IZ5/S+AM97J79u0BrjxQWDHP5Eqmse/Q77/J72W3N9zJhYzVWaJFSxHVCe6v4TiMnuRDrQAod6cuzscEr702lNw//5RnLc2u6xOkiT43U7MxlNZueiJVAbxFIl4MXP7cJnoVSei637e+QmgrsN8f9aJXtfFT94FHKqIXkDDLYGg7LRtBY49RG6P7iUZwA6xti5Y+gxNRyHLgM/tUBsA6vm3qzbh2MQ87t47ghtvfhoPfuxi+IeNRfRtjsN4peMRHIm/rDQHaCcTHeCvw3J25BwAzE2NAFiPdUXkoVNCluJciBCwysCJ7pF110S7TvT6bn575gRxRx3XxTlEw/Zel4X9GwCAt854Pz16J7osFxa1ZwdZ5g0QASqiF9dYFAA2tms/d1PAA7+Hfy2304GQ343wfBITc3E13iUfI9NUROcNLJPswkyLzqQxfhDorq4MVaOmopTmWg+a9CI6gFPiT+EALkKPwQKTFbobi3eiZzIyvnP/Ycwn0rh//xhO7g4V/FqloDbKLDi5aoBAC1pqye/ONBMdwLHASeA+EX2PlcdRXSJqPE60OJkFOn9T7nFXloiuRIzp89B7DUT07jOAp35I3ldKYJPUpxPRGSd6U4mbilICzaSh4ZTSjPDE04a7jcxoeej6xSjViT5ZvvgiWZbVRXC2h0SjQSY6jXNZ11qLjFL5sG9Y+5sOM70e7NDbFMBzA9NqU25BAXAGBBHlkoUkAWe+i/zLxcX/Rv4Z0bUdePOt1t6vrhN42TfsHeMiQcyWBQtO2Z3oXadZmsBcuL4Fn37ZFridxl8LGumizyqLMM4z8zgXxolezXEuAO9mMoJ1oreKKJdc0Fy/0RmRiS6oQtoYJ3pyXpvkCARLHCo+dYVqsibsFKdDwjffuA2tQS/G5xJ45vgkX7Gx8RpO/P2E+xakY9kOcNtkMkCSmUTbdaKbkJwmjSHXtRUvohtGbzDMxVOq8NVrkIkupXSiWDFOdICI1YfuBmRdpWAxTnR2Mu5wWc9t51zycWu9ZoolNs3H2KiZ6Ln79VhhVXMAPjcZF+ujXCh0ISqXY1jPkE5oajRqVqsX0YuJ5ykTuUT0plovOqXJrPsvcexGfY2bmxvYQc1EDxfuWH1mIIxB5W8wssBj1FgyjabUsHZHqAeQJDUTfTyHEeW4ezXmZOb80feY6b7VQruLObfkinIBAF89iVqgUBGdzUOvbSdCtZ4VfHPR7Y6DvIg+wTjRy5GHTmGrwQeeMozcMctDB4AV1Ik+OV90/NCDB8bw12ezY7aiyTRoGlWtQSY6NfylM7L6nV/fFlQXGfePzCKl9Eyj36uOkL3rGm3CfUw40QsnYXPsJBAUiBDRBQsOLf8MlcqJHhnnxaA8US5WoSK6Ps6FTlJq3E64TAR4WkrrkICAp8pEdDtxLukUMHZA2xYiek6oEz2SSHOLLQJBVSCaiwqWKSeoiJ4nTsHvceH0VaT59pGDL/Li9rrLgfM+pG62SWFcG/lt8Qenb4xlKRPdRER3MZP4+TEAwLrW4uNc6HjNrJEqzUNvDHgM+8wgpXMU2s5E1zW/mh4A9v8je79Sxbl4g9bd5Ppjq0Ski37cphPRizFvOB0SNigRQN0mGd5NATLWyZWRr2eEiXMBGLGKfY1QL+BkFi/Gqq+5KBXUVrdkLxY113rRqc9EB3C+43msbih8zkMXM4bCMaRt5NCz/OO5IfX2yAJHDg5Nx9AtjWl3hHoAaH2FconoM3FgV4Zpvql3aJeTI/cDf3gX8MwttrLYWxzMuSXQZL4jhc1FnzhM3ov9OXvPNj4/NazkRHpORI9N86YpveO9lLDNRSNjQLgvaxf9+YB7uvJ5n4unimrenUhlcOPNT+MDv96txsdQaD8Tp0NCDdMzJKQs7kWTacSSafRNziOeysDrcmBFox89jX74PU4kUhk1wmwoTONc7F3XKpX9vqThRHThRBeUDyGiCxYcWg4cqimRE11fKlYqEd1Nneg6ET2enaGmh8a5BH0lbJ5aKvQieiSHiD55hG/YymYqC7IIeF0IKIsvormooOpo2QSAOR/p+x0IBEuUAZOmokZsU2IO5o7rGqK1bwXOeT9GHZpI8cb0X4D5bOepLfQiuhU3lVGmbvtJQMcp6mYwPQ1JMnbM2oVGdpjFudBM15VmcRVJnfPVrhPd3wQ4mTHj1DHg4N3Z+8XC9l6XhRXRPTYWHvQu+anjxvuVkohOqFUqDNV4giIy0QFgcydpXNtrJqIr2dUTEevjHM2JTv729DM1G08hoUQkwukCmhmBtMqai8qyjGPj5LO+xkBEb6r1ZGeiA6iVYrjYdyDrfqu0Bn1wOyWkMnJBLvJMRsY/ntdEdL2gWC5GZ2J45f89gt88xYuog+Eo34BVEdGtONFnYkk8ldmg3TFxEJgbM92/ZMxPAr+5HnjuN8Cf3g385f1A2prA24RpbSOfEx3gBe6JQ+QfK4D3nGP8PEkikS4KOxwHNBGddaEDQGOZ4lwAUhHOYpCLTiuX2oLZFT8+txPtirheTNPNYxMRRJNpyDJxtbPMML3N2Oq0oNcFlzJvn5pPqE1F17bWwumQ4HBI2NBOrg97lOai9NzWbjMTnTbhPibiXApHiOiCCiFEdMGCE46QQUdDqZzo3MVZAjq3m+5qB7/HWESnq9dBkygXQCtX7bB5Qa0IvnpSqkzJFeeS1VRUONHzQZ00ItJFUHV4/PzEyaRpokCw1KAN+cziKVhOWRECALhYF6zkAFo2IeOqwZdTr1fv9kpJZAafLe7gsrK4C4xz2f5WrudJkzSNnka/WlVXDIZNIBmoCLDSIMoFQPFOdEnixeoX/mDYTLV0TnQbCw9Na8nng3Jip/m+pUI/bvM3Q5blkmSiA8B7LlyDt5+3CtefbdxfqMlCA0g9amNRZVxc53ODekzY/GEu0qXK4lzG5xKYi6eIVmlQ1dJiEucCAGdnjLOhreB0SOgMFZ6Lzka5AJWLc/nrc0N4pj+Mz/19L2JMf6mR8Qk0S0yzzQbyOWsOks9Vrkz02VgKT8m62J/+x0t30GYcfwSIM8e8+2bgljdkNyQ2oF5mnqfvS2UE60SfGwEO3sk/bpSHTmEiXbqlcaSmlcoYNg8dKG+cS/tJfEWJQS66FudiHJu1QukDUIyITgVwIPt7Qxcc9bGskiRpPUAiSa6pKGUT01x0JpZUX6vTZpwLXaQcDEe1hUSBPVgTgp3rtkBgEyGiCxaccFRxoltsRpQXVkRv3QT4LDaDygNtphRN8rEc6oU3xyRlbWstvnPddnzjDaea7rNgSBLvRs8V58I2FYWUnVcpyKI1qOSiCye6oBphc9FHhBNdsDygcS5WRPStXXVwSEBPUpcf6/HjRDiK3ckebv94tMhcdL0IU0gmutMLnHQt1/OkUZotSVNRgIlziRg7L48qcS4rDZqKAgCKzUQH+NiU2eyMWwCly0S3Eqmj7ltLRCNKXwUEPYM4l3gqg2SaREwU60TvafLjk9dsNm2UR+NcJmw0UaeiGXWYOhyMWGXWXDTcxzsNFxjaALCzvgY+d/biVHONjFYpbPjcjTOP2ooA0UPPXXRB0A5/V6Jczl5Nxv5js/GCY2Hs8OIgcWDPxFJcLvX8qK4fC41zUZzoU/NJJNPGouJsLIlnMmuQkZjPeCUiXfqfyL7v0N3AT64CZoezH6OkkwikGRFd35fKCLYaAwB2/0K77a3PXRWsy0UPTTxDbnBOdIlEv5QLlwfoOFnbHsheWKSRQkZxLgCfi14otCEoAJwI60R0aogzmMur/RrmE1pTUaa3yCbFib5vaAZDYXJeq69xZzVhzkdL0IsatxMZubDvtQC6xqJCRBeUjyUhokfiKXzx9n245Cv3Y/1/3oatn74Dr/veY7jjxeyL2MRcHJ/+8ws47wv3YsN/3oZLvnI/vvvAYbUZhKDy0MFyyCg30y6ZDHBil7bdtcN8X5vUmDnR4+YXXpYrT+pQS76qDlZE15cFs7BO9MbVgNume2wZ0qK4KnI5aQSCBaNtq3Z78khVCRQCQbk4YSPOxe9xYX1bEJskJoKgnXxv9g3PIgreORef1znJ7aJ3VBeSib755eQ+RqBpxgzWliDKBbAS55JHRE8W6UQHsmNTjIiG7b8uhYtzsfl7Y+MVTuzK/nlLjX7c5m/CjOJClyTAbyDwlhI1zsWiE12W5aw4F0CrSOWy1Vs2cM/FeOExKKWGLhatMvmct0K3iMMsCATn+7UmkQXQHVKai9p0omcyMm5Tolzeek4vHBKQke0tgBTKnkFNPP7FE9r5NDV5jN9REdEb/B44lfIEs7z92VgKMXgx28j0eKlEc9H+7EgSAMDwc8APX8L3j2IJ98EBRnOoX5H/vVgnOsDPxXrOBBw5vt+dpyIF7fGXjN8M7PwJMPQMfwyFLGTaga0KH34eyPBz6RElUogaj/T0lEBEP8g40U9kOdGVaFaDqvIQU3lF3ezrW42c6LMYnCavW0jluSRJ6FUi0I6LXPTCEHEuggqx6EX0uXgKr/nOo/j2/Yfh9zhx/Vm9uOqkduwdmsGNNz+N/7tPG6BMzyfx+u8/jp8/fhwnddXjbeeuQo3bif+9bR/ef8vuHO8iKCe0bJOWBxfF+AG+vI5tZlIkfpPGomyO2qLFshOdGbiJKBdL0OaiwokuqEo4B5NcdeXyAkGpSaUzGFYEvC4LTnQAOLPThRUOJmdXWXzaPzyDqMyPXRJFO9ELyERvWgtIjJCy45/I/0zerldKYnNzaYb9VFSYT6S5WAbK0fE8megpLT5ChgQ4CzBRmIno7IJCdKpwt2+8QCc6APScpd3OJIHBMs8x2HGbtx5weVRnZa3XVfZePHYbi4bnk2pcARvfYLg4o694rKJcdDW2qNn4cx5KjnLb8o638TscuL3g9y7Uib67n0S51HpduGhDqxo5WO7movFUGodGyXfKIQHP9ofxwgniTHdM9/M7h1aS+x2S+pkwM6LQ5rnRDma+N/QskCijCJmK89/ptZcBte3a9nQf8Mcbjc89+ixyvUBuRMNKPiKKpSdHlAsAuGtwwLFK3VyZPAz87V/4RsxNZcxDp3Ru026nolmLYWPK588szkVtulmiOJdBnROdfo6Mqsrp4t74bBxHxsh3no1zoQa54ZkY9g4RDaLTwgK9EZqILgwtBSFEdEGFWPQi+vceOIx9w7O47swe/PV95+GT12zGF689BXd96EK0Br346l0HcExxCnz9ngM4NDqHz75iK77z5h34xJUb8Zf3nYcrtrTjtheGcfsLQ3neTVAOVCd6KTLR9dmTFRDRtYlKiTLdFwK2nNBMRE9G+Qw90VTUElqci8hEF1QhbVv47ZEXF+Y4BIIKMTIbRyojw+WQTF1ves4LjvB3KHEd+0fmMK9zoieiRU5+9Y1FreR61nUAV3+ZVN+95L+BleeR+3V5uxtqS3MdqvO5VIdoeJ6PdJmLp9RGgL1mmeiMMzsheYld2vZBdBvfv/Va7XYmWXh1DVsRYDdblRXRgfI7Y1kneoCYIqgoVFdklIsV1Ex0i41FqQu9KeCB16Ut/lAzDSfGN64CHMzPwMUKlpF9/wC+dQbwy9eRxrUGHFMXi4w/594IP6+c7LwQAzLznTxwR8GH191YWCY6bSh62aZW+NxONT6j3LnoB0fmkMrICPnduPpksgD2i8dJ092ayIC2ozsA+BvVTdpcdMzEKU9z/zNdTGyJnCaO8HIx9ByQZo5n25uAd9wFNDNVE4O7gFkDXUFffcA2DTXD5VXd+Vn0mjQVZfhN6qLcO5QzD53CNLkGQBY6FCLxFGaVqm6zOJdiRfR4Ko1jjLs7K87FJBMd0Bb3nh2YRiKdQY3byUXBBX1uNbP9/n1ksb3QHmj0XHJMONELQ8S5CCrEohfR//7cECQJ+NgVG7luyu31Prz5rF6kMzLu2z+KRCqDXz/Zj856H647Q7sQOR0S/uNq4qj95RN9Wa8vKC+yLGNamYA1lCITnW1W4qnNLgMtApp3OJ/Ux7mUpnHTgmIlzmX8ACAzJYjCiW4J6vIRcS6CqqRhFeBmXHQiF12wxKFl3J2hGlUIzsdWl84pqSw+7R+eQQz82CUVK1JELyQTHQBOuwF4573AuR/Q7tPl7fbWlCZWRJIk1Z2nj3ShxpXGgAf1ZjF9jBM9IRU49jNyojdvyI7xi4ULe30uzsWmEz3YTs6tlHJmNKdTfCa8nxfRKzE2bbYZ50IFW71gpjrRWRHd6eYzoSvhRO9/EvjdW4Hx/cDBO4DvXQDsvy1rt7wNdGcGuM3jqQbcm2Z6I/U9VnDkEG1kakdEz2RkVUS/6qQOAJrRY6TMRg8a5bK5ow5vPpPMw//8zCBmYknUxxmxOdTDLarRMfS4wRiaNM8ln3PnCt33no32LDUDT/Lb3WeQ477qS/z9Rp9VRkQPI8gtGOTEyLHu9CLWcjJ+//SAunCpJxJP4WeJS/CmxL/jN6mLMCkbnMs6K9Cvq3kD4GK+74PPqDdppW7A4zSt6u5hmm6a5ePn4uh4hMv9PzEVhcxUCuTKRKeVV08cISazta21WdU9G9tJpMvO46SRcKEiek9T8Y77ZY1wogsqxKIX0d927kp89PINhgN1j4v8eJF4Ci8MTiOaTOOsNU1ZJ74VjX70NvnxxNHJijRWEWjMJ9JIKBfDkmSih5mJbsuG3DlxNsnnRF/cIrrOiW5Ugqh3/wgnuiXUOJcyl8oKBAXhcPALYiMvLNyxCAQV4ESYTE6t5KFT2uY14SPtDQF1XUikMjgyFoEMBzJMpncqXmonehG9VHQiui8+Wfhr6WjwG0RvgBUWTaJcAM6JHkcJRfQNV2bnwxfaXLSYOBeAj1nofzIrA7gohp4F7v0c8LOXA1/oBY4+qD2mjOeoQ7cSUYM0zmU6qsW05GJ4JjsPHdDMNJP6rH3WEFPuyLHpE8CvrwPSzDHEpoFb3gDc9WmyaAEi4B7L10B3WhPR494m9E2ncW9mm/Z4JgUcvregw6RO2MFw1PLcdXd/GENKlMsF60nUU1tdZeJc9gxpIvoZqxqxvq0W0WQaNz92HO0yE3ujc1zTBZpxgwWaWDKDlPKzB5p7+EgV1lRVatimosFOoF6pirGS38+I6Icz7dYzvo1E9K4d+PTfD+Ejv3sWX7vLOIOdius7pZPx8dS7cHr824i86c/Ame8hOeWnXg+c9Fprx1AMThffg4dxopstqrG0BL3wuhzIyNlRLFagDUE3K/nlkUQa01Gtikrrb5atRTQq17pBpYKGbSpKobno9Kto1oQ5H5oTXcS52CaV4M/bQkQXlJFFL6Jff/ZK/PPF2RcWWZZx+wuksejG9jo1h6230fgL1dPoRyKVKaphhcA+dPLlcTpUkboo5phmssGO4l+PgXbZnk+kuPtpJnpgqWSiy2lj5xabh+70WCtBFKj5fmalqALBgsNGuoy8WHiGsECwCKBOdKt56ADgGNUWlyZq1wGShMNjJJ4g6HNBYppsZ4oV0Vnx1uktLC+cwmSiAwDmczQOt4kqokf4OBfaEM1UWAQ4J7reyW+Zuq7s+zZeDdSE+PsKEdFTCT6uwW6cC8BHusSnSxdDcvhe4ox+8IvA0QeyF10aiQO+kk70+hq3WtVh1myWhca5tOlFdFrdoM9WZ3PRp46Vr1FrMgr85jogMmr8+CNfB37+CmB2GGNzcUQSaTgkqFEOWUyfUG/Oeolg+lhmC4kwohy8s6BDbQ364HZKSGVky1Esf3+Oj3IBNOFytMxxLqoTvbMOkiThujN7AQDffeAwuiWm30RDL/e8FiXOxchpPcs2z/W4gC6meeVgmZzoskwWxSgrztCc87VtpCcBxdCJrmWiH5U78NjhHH2oWAxE9P66U/CbncQ8RnUOPfT31lbvhd/jRBpOjDedDlz5v8C77gNe8a3yNxWlsLnow88BGbLgRp3otOrACEmSiop0OaTkoZ/UVa8uzLCRLrM54lz0cbNsHjplUzt/X0eoQCc600BVGDttktSNvYSILigji15EN+MXjx/HM/1h9DT6ceGGFsxEc+du1ykuaNrNXlAZwkweulRIJqaeWSa3tLa1+NdjqKFxLjon+hHFidJtY0JedQSa+O15A7caOwFsXl/cxH4ZQUtlJyMJSw4tgaDitDIienQSmBsx31cgWOTQ+APLTvRMmos5OuxYCUBrUraxPQiJmaxlim1ox4qihYi3LH7dtT0yZrxfAYRM4lyOjueJuAA4ETQqFziWCLQADhe/3bXDwIketv/apagG0Df8K1Uu+s6fmD/WugU4410AtL9LXSmqPPPgcEjqooqVSJcRRURvr9OL6LS6QTcX4xy+cnlcxrIM/PWDfMPI9pOBs/6Z3+/4w8CPXoK+YSJ+doZquFx3jhlNRJ9ytaBvch5xeDDQwOR3H7yzoCoFp0NSmxdaiXTJZGTcpvT+opnkAOtEL5+InsnIqhN9SycRmV+1vQs1bicysVk0Ssz3TedEzxWJOKNvnsuK6JNHjOcyxTI9wGedr2D+lpIEtKzXtvVO9MQ8F/FzNNOORw9bXNg0ENG/flCrNBqcNv4MUAd/U8CrxiVNWGwAXHLYXPTEHDBJFhRGLTjRgeJy0akTfV1brfq9OcF8b2Zj5iI6PS9R1udwolM6C3Sid4Zq4HZKSKZlDJn8TQUm6PufiEx0QRlZkiL6354bxH/9dQ9cDglfed0pcDsdqvBJI170eJ3k/rgQuSpKuJRNRdMpfoLIlvWVAKM4l3RGxv5hMjDUX0AXFbrmY4a56GxWsshDt0yD3w23kywQmWUWCgQLimguKlhGUPeZ5YXvySNASpvMPhUlAtS+YSKir28LAowTHckiRXTWiV7sJNDpxpzEvIZZz5MCMMyvBnB8Ik/EBcA50eczBY7/HA6gicnKXn8FifArRZxLVi59ASJ68zp+EaPvcfuvoSedAo48oG0HO4Ad/wS88jvA+3cB73lEdaLTxQzT5q4lRs1Ft9BcdEgRzdp1TnT1M6V3s3efDoAx2jz5g8IP1IxHbwKe+422HWgB3vAr4Ir/AV53M+BlxvjhPkQOPgQAWJXrcz6tRUyOOZrRP0XODTPdF2n7zE8A4eMFHTI9hw1M5T/nPDOgRbmcv04b81PhcriMcS79U/OYi6fgcTmwuoX8vup8brzy1E7ehQ4YxLmYO9FpBIfaPLdzO78DuyBSKtgoFwBYcSa/zTYX1TvRJ49wm0fldjx6eILL5jZFJ6Jn4MAd073q/Hl4OmboXKa/t+Zar+k5u2J0bOO3lVx0Lc7F3IkOkPhdoEARfVS7XtMFdNaJPkfjrwwqd/Q929a1Zl8Pehr9XEW9/txmFadDUn/O46K5qD2yRHThRBeUjyUnot/8+HF84BZy0fzK607B6StJww6vIp6bNaOIK/eXJFJEYBk6UA75CyznZZkfB8AMIIJtxb8mQw0V0ZnGoscnIoglM/C5HbldV9WO3q2mL/mOTfMNkoSIbhlJktRy1OEyl8sKBAUhRHTBMsJ2nIuuT8C9U62Ip9LYP6w50UsqoicYAbeYPHSFcZkR/0oooodMXMNHx5U4F4uZ6PMZT+FVWpd+koxfWjcDF/8Huc8X4vcphYheSEWAJPFu9FKI6CeeJtEwlEv+E3jZN4BtbyIRe0xFJ413WNNSmbFpk43moqZOdJqJrhf56rtJ3j1l71+AcF8RR6vjyAPA3Z/Wth1uIpyHVpDtzS8H3s7HrqRGSHWm6dg/PkfGzgqDchP6J8nnvqZTN4aeGUIhdIesNxd9foAcy9lrmtQoF6AycS40ymVDWxBupyY9XHdmb1EiOo1zUSOL9A0yy9FclI1ycXpJtQIL60SPjPLnn8nD3K4Dji6MzsbViuac1HVxjTn3ZHowBz+++rpT4HQQ57LR72h8lnyXWoIerVpkoUT0lo0kDpQy9AwALc6FVu6aQcXlgUl7Du14Kq0K0pyIznxv1Ex0Qye6ttDr9zgNq9gcDgkblEiXpoCH+47ZpVeI6IWhryATIrqgjCwZET2TkfH//rYHn/zTC3A7Hfi/N23HK7ZpeYl0pXYmahzXQu+vM2goISgfYeX33lAKJ/rsML9dYie6UZzL3iEy0drQFlTzIBcluuZjmNdl9I3qGjmJpqK20FwFolGMoArxN/I9JISILliiyLKsOdFDOURelmFNRE/Bgb3pLuwdmlVF9A3tdYBbm6w5UkWWYJfQiT4ZSWA0wwjxJYxzaQyQcVuYcQ3PxpKqkJPbia6JPTG4C49S3Hg18NFDwHsfA+qUc5jLw/09DHu85KNUzV3ZXPSZASDcb76vFQ7fw2+vucR81zHyM6xtrUxJe2PAXOzUY9ZYtNGfwyl71nu023KmdG70TBq4/RPkNSlXfxno1cXxtG4CahrVTc8UaRBp+jlnolwA4EiiQY1naO5Ypdt3sKBDt+NEp80YaSQGhYroE2WMHHxxkEa58BW7W7vqcVaD7rsW4jPRm4PmjUWzcv/9jUDjam2HcsT+DDAietd2cr5hYZ3oADDGRLowTUUBoLGbZP0/aiUX3eFQKjII92W24dXbu3DJxjZ1MeqEQcNN1onetNBOdJeHN20ozUWpE701jxO90DiXI2MRpJX+JW11XnUBnY3AUeNcDJzojYwTfV1rLYkOMmBjO/l8F5qHTqHVQ2LOaBMR5yKoIEtCRE+kMnjvL3fhhw8fRcjvxi/ecSau2MoLqHQQabaq1zc5D7/HqeZkCSpDWLmQh2pK4ETX5/iWOBOdNhaNciL6EohyAbiJAYBstxrbVBQQTnSbrG4h558jY2JAJKhS2k/Sbh97WDQXFSxJxubiiKcykCQb5daME33EvQJxePDwwTFVsNigi3MpWkTnMtGLc6IfHJnFBOtEL2FjUepEn2REdDrGbgp4cptSmN9RHB5MmxhcLOEwmMqwkS4liXMpcDKelYue240uyzJ2Hps0F7kOMSJ6yyagrtNwt4m5OKbmk5AkYHVzZYSEJot5y9FEWv17ZzcWJa8RSaQRS+pywleeD7Rt1bZ3/YxfcCqUF//Ij3FPvZ5E5BjBZLOH5o8CAFY1myzGTQ9wmzun/MjIpDK6sYMXivWCu1W6G3NkoqcSJE5k/23AY9/GOfv/F99yfwMvnbqF+72xkYNjZYocpHnomzuz50qvWcMI955gVhwTreScjCSyqsk1Jzpzrunaod0+8XRpxzKJCDD0nLbNiNoqrBMdAMaZSBemqSjqurB9HTH7PWY1F/2qL+Fww/n4Teoi/Nb7GnzyamJo6lRE2xMGnwNWRDet9KgkbC760LNAJoNRJUqoXJnotH/J+rYgJEkyzESfy9FYtM7nBtXN1xk0FaWc0k3y/ouN0OpVqriOCRHdHiLORVBBFr2InsnIeO8vd+H2F4exorEGv3/POWqEC8uWznrUel14/OgEMrrMsP7JeRyfmMf2nobF7SZehNAy4FCgDE70YImd6Eqcy3wypd63T8lD39hefMn1guLy8B3ls5zoTFNRTy1Qz5dbCnJDy6mpM0wgqDpWXajdnhnIboglECwB6KS5vc5n2iMnC8aJPltPRLTfPU0Eso56H+r9bsCjCWmudJGxCKwwWGRj0YOjc5gsU5wLFTz7Jubxj+eH8OdnTuAPu4gY2JsrygUAktrvKAa3aZVowdSEtNsliXMpcIzXfjLgYsw5eZqLPtMfxrXffQz/eutz2Q/OTwKDTDzF2ktNX+ewsmDfFapRx67lhmaiT+aJc6EudL/HmRWdEPS51HlYWN9cVJJ4N3psGnj2luIOOp0C7vsfbdvtBy79lPn+zZpA2pUkcTKmgplORO9LEXF4RaOfNCJmxeICnehdZnEuRx8CvroJ+L8zgFveANzxb7hw+k+4xvkEzjj8TeCmHcCzvwYyGUiSpMZolKu5KI1z2WxgOGpMMFE2oR4ukggg5xn6mdCLv1lOdIDPRY+MFrxAYcjgbkBmFnf0eegAcdI7GUf1uIkTvWkNzllDojQfPzKZpU0YccKzEleMvhcfT70Ln3jF6aooTkXhwTxO9MaqENG3abfjM8DUUSYTPV+cC/k5p6NJTOvPDzk4qDQVpQ1BDTPR4wafJQWHQ0K90qDZqKko5ZWnduGzr9iCT1yx0fKxGbFSdaKLOBdbCBFdUEEWvYj+7fsP4e69I+gK1eB3N56DNS3GJzef24mXb+tE/2QUP3n0mHp/OiPjc38nAuH1Z/caPldQPmgZsL7zdUHMjTIbEhAotRM9u7EojXNZ9E50AAgwuehZIjrj0mnZaOz8EphCGykJJ7qgatELMofuMd5PIFjE0EmzUaapIfOTXD8QdyfJv6WTW5qBCrcmGrvTxTrRGQG3yHLkgyOzGIdORC+RM5MKMkfGI3jvL3fhg79+Bj9+hLpz8xw340SPyUU60Y3gnOhh+88vVZyLywN0n6Zt5xHRqcPy6LjBgvuR+/nIkRxRLloeeuXK2ZsUxzBtLDoTS+Kb9xzExV++H1+9c7/a9HCY5qHX+yDpBFOHQ1LjHbOaiwLA1mtJw0/KE98FMkVEkDz7Kz6n+swbc1exMk70RmkWzdIMVjSYLBgx4m1KdmAU5DOpxqnUdRnuawca5zIYjvJNJe/5TO6qk7lh4I83Aj96CTCwU23oWI5c9Im5OIZnYpAkYKPRXInNtg9lG3QcDkk914zN8k75GSMRnXWiA6XNRc9qKnpG9j4OJ2kqTDGLc2lcg5O7Q/B7nJiMJLB/RLdwZ8Dvnx5AMi3jjFWNuOokzSiWW0Qn36PmWo95416bROIpa81QjWCd6ABi/bsRUebVrcHccS5+j0vNyO+3EGFEoU502hCUfm/G5xKIJdOQZRlz6mfJ2NRHF5poZIsRPrcTbzl7pRrhWSh0Ebpvcr7w3/NyJCsTXcS5CMpH9nLbImJ6Polv308GP5s763DLk8ZNZs5c1Yhz1jbjo5dvwIMHxvDff9uDxw6PY21rEA8fGsMLJ2Zw9ckduHxzaRtRCvJDM9FDNSVwos8xTvRAM+As7cebiug0E306mlQn5IYDw8WGv0nrHM+K6LLMZySLKBfb0Ins0fEIMhnZNE9PIFgwWjYCwU5gVnHEHb4HOPu9C3tMAkGJsd9UlO8P0LJ2B/CkNqnd0GYgosuldKIXGecyOod1rBM9kyQOXtapXSCndNfjVad24dhEBG6HA06HBJdTQq3XhXdesCr3kzknukcVw0pGqZ3oxUzGe84Cjj1Ebo/uIceji6ygUDek4aLC4Xu12y4f0HuO6VtWOg8d0OJcBqai+MbdB/Gjh4+of9dv3nsIu/rC+MYbtmF4RqsGMaLB78H4XMI40sbtA057O/DA/5LtiUPAobuB9ZfbP+BUHHjgi9q2tw445wO5n6PLuz67bsK8omVaE8ZH0IiM4ltbQc89dZ1aVFSBTvS2Oh9cDgmpjIyRmRgRU+OzhsLxlFyLDCQ0Scxn+8RO4IeX4p0N1+E9uBojM6WPc6FRLiubAoZRGQgf1243GJvZmmu9GJuNZ+XtG8a5dJwMSE7NMX7iadIYthT0P8Uc60rzBZfm9drflsa5zE/yc6umtfC4HDh9ZSMeODCGRw9P5DRkybKMP+wiC7qvP20FtwClxpOEs68948rCQ3PQq8657TQWTaQy2Hl8Es8PTOO5E9N44cQ0jk/M45w1TfjVO8/K/wJ62rYADheQIeeG2PGnAZyHWq8LAaPPh46exhqMz8XRNzmPrV31efcHyHUQIHEuAFBf44bf48R8Io3BcBSdoRqklEUow88ogP+8ZhOePDqJc9c2Gz5eSrob/HBIRG8Ym4vnbbgqUBBOdEEFWdQi+rMDYVXQvGvPCO7aM2K43/suXotz1jajMeDBH95zDr5y5wHcs28UDx0cx4pGP/7tyo1427mrshwRgvJDV8NDpXCis3EutaVfEKElsVFl1XqfMjDsCtWoZV6LGj8zMGBLviNjQHRS2xZNRW3T3eCHx+lAPJXBiXC0aJeCQFByJIk4G5/5Bdk+9ggRutxi8C5YOtDYA8tOdJ2IXrfyVKxo3Iv+SSUP3cCJ7s0UKaKzbqoinVTHJ+bRKOuEmfmJkojoLqcDX3v9tsKenNKJ6KV2ovtC2u1CnOjsQobk5DLvbaPPRe9/Elj/UsNdI4qIHp5PQpZlbV4iy7yI3ntuzmNaGCc6GcfvG57FPqXp7trWWlx1Ugd+8OARPHxoHNfc9DB29JIFBLOeBGpus5lb9rQbgIe/CqSVxx//dmEi+tM/BaaZRq/nvJ80psyFLu96u3/UZEdwFSzjDm18rY7/2Dz7AkV0p4PkO/dNzmNgioiB6H+Cjxy5+is40XUlzv3mM6hzJfHMpfvgeOTrXDXIlVO/RDvOKkucS64oF0TDZFGPYuBEB4CWoBd7h7Kd6IZxLu4aoG0zMPw82R4skRNdlnknulGUi3rAzGLL1HEgGdVMSpSmtQCAs9c04YEDY3js8Djefp754uPTx6dwbIL0b9P3fetSMtH1TvRYMo1Z5ZzSXOtVY1zsNBZ958934oED2Q2pHz08gUg8ZUn45nB5iRmL/n2GngVwXt6mopSeRj929YUt56LHkmm1QSeNYpEkCV2hGhwcncOJcFRtJipJmmFOz/nrWnD+uhbDx0qNx+VAZ6gGA1NRHJ+YFyK6VVgRXXKSxWaBoEwsahH9gvUtOPa/V9t6TmudD1+49uQyHZHALjTzkJZvFgXbWLQMIjptLCrLQDyVUScJmzoWeR46xc/GuTAiumgqWjROh4TeJj8Ojs7hyHhEiOiC6mQtI6KnoiR6YM3FC3tMAkEJodVj3WYRDHqogxgg18hgO07pHsoW0ZlMdC+KcHOmEpo4CBSViZ5IZTA0HUWPpBOvImNA05qCX7ckJPnGoulyxrnEwvafH5/Rbntrs3KabdF9OiA5tCiWvsdMRXQaKZDKyJiLpzSH7dh+PvIjRx46oDnRaT+WStDJLEyta63FBy5dh6tO6oDTIeGakzvw7pufxpHxCP72HMnANnOiNyqmGlOhL9hGYl2e/RXZPnIf6dtjZ2yamAce/LK2XdMInPnu/M+r6yYLZkki4G105RC/mUz0sLsVUD7yK4ziXOZGgHQScNqfC3U3UBF9HmesagSOP8o8KgFbX4OBIfLZa6yvh+PiTwDb3wzc8R/Anj+pe57mOICRGT5qoxTkairKLWIApiI6zdsf1+XtGzrRARLpooroz5DIn1wxlJkMMDsE1HeZ7zNxmDcUGUW5qAfMLrbIpGKCbSoKqCI6zUV/4sgkUukMXE7j4/y94kK/cmtHlnCtxrlM8yI6de57nA7U+VxqdKodJ/rTx0klz2WbWrG9twEnd4XwgV/vxmQkgaPjEctucI6Oberfxz/xPAAZbRaFYhqH1G9RRD88NoeMDNT5XGhh4mK6GhQRfSqqLqrXel1VY6jsbfKrIrpRrz+BAWwFmafI67ZAkAcRbCxYUNRM9EApMtEZEb3ETUUBoMatrU7PJ9LYO0Sbii6BKBdAl4nODBTZpqIAKcUT2EbLRRfNRQVVyuqLATCDzsMiF12wtLAV5zJ+ENj3d22752xAkrBtRQgAWRxV4zIYV7APicJzmkuY6TkYjiIjA7OOEP9ACZuLFkQmTWJlFGKyBzOxMoro8RkiUNqB/Tt4ixzj+eqAtq3adt/jprvOxTUHMddYU38uXmMuokcTaXWxqJJxLh31NfjWm07F967fgTv+5QK87JROtSHk+rYg/vy+c3El46DtMHWiE0F0MpLjb3aWTvC+//Pkc2WVJ79Pmk5SzvsQ+Tvlw+Hg8q5XpAeM95NlLs4l4tN+7h4jJzpkvprWBjTfmZ7bOBG9bStQ06CKq+pCR3038OrvA05t7rXDcQCjs6V3or84mENEf+ZX/LaZE13Jwc6OcyGLTnX6ZpBsc9H4DJ9FrifcD3zvfOBrm4HvXQCM7DHeb+BJfrs7h4jOOtEBsgjGHoPkVKNrtnTWI+hzYTaeUn9XemLJNP72LFl8es2ObKGf/l3D80m1mgXg89AlSVIjl2ZjKSTT+a9RM7GkGjF10xu3470XrcV565qxVqlwoRUvtmFy0T3JGXRL45ad6HQRSu9Ez2Rk3PnicFYUltZUNMgJ5GxzUbWpqF1XfRnpVZuLil5almGd6CLKRVBmhIguWDAyGVm92BWdiS7LwGx5nehOh6RmH84nUqqIviSaigK8Ez0xp2WWsuXs/ia+qZPAMquVQadoLiqoWvyNQBcz+Tx0r/m+AsEiQ5Zle41FH/46AKapl5KXfNZqcq08qaseXpeyuO7mJ2zJeIHneX0WdxGZ6LTxmi+ky+2NZJfmV5SkLnIA7tLHuejjatjICCvESxepA4CPdBl4ChjdZ7jbXFz7PXBiENvoOdiZLdIxHBmfgywDIb9bbSRYKa45uRMv3dJu2Pcl6HPj29dtx6dfthkXrG/BSzYbm12oWzZn88OOU4De87TtPX8GfvU6a9E90wPAI1/XtmvbgdPfkf95lJaN6s3m2DHjfaJTXFxK3N+h3jaMcwEKjnShVTUDU1Hy3TrxtPagkps/qGRls9UCcHmJI1jhVMfBkse5RBNp1TiyRT9X2vcPEsVDqW0zjYukDmIzET2oF9Gzmos+DUNmh4Gfv1zLLx96Fvj+RcBj3+YXQuNz/IKqpzZ3tGXTWlJ9Qhk/wIvoDSvVqgOnQ1KvKY8eZjLTGe54cRiz8RS6QjU4a1VT1uN1PrcqAA8xbnQ2Dx0gWeD0q2kl0mVI+dyE/G410hQA1iiLc4cLNQV1nsptbpWOos2kMkWPmRP9pnsP4V03P41/+smTXJNdtaloG38tVXPkp6JqBVCt/nO0gPSYLBYIciBEdEEFESK6YMGYiSVBr3NFZ6LHwkCaGVyVQUQHtKy0SDytdlJfOnEuumYp8xPEvXXwLu2+1s2iPKpAVjcrTvRx4UQXVDGsw3H0RWBmaOGORSAoIdNRzVWXV0QP9wHP/Vrb7j0P6CEZuFu76vHbG8/G/13HLDjp8qnn54wdhXkpoROdRs7UN7aBqzCZX2AneooX6kgmeqkbi+oad9rNRWcXM4qI1FFZx2R2Z1LAn98LpLN/5gjjRFdF5GQUOP6IttPaS3KOw6g7dG1LbdVEE1AkScLbzl2Fn99whmkmOhX+c4roAHDRJ8B9rg/dDfzwUlJBYkb/k8D3L+abzV7wUS6OKR8yE9Xhmx/iF1wo07xDPRMk7uHGgEdrXFincxSzcT02oE70gfA8EYvZOChFRKeLh5368x4TSbJFOo7wdIHnLRP2j8wiI5M87lZWJA33A396D7/zRi2jrAABAABJREFUy28yjbNpVpzo2ZnoJnEuLRsBF/OzGonokQng56/MzipPx4E7/g24+ZXAkQeAv30Y+MpGYN/ftH26dgDOHIKry0uEcoreia6L0zpbEdEfO2Isov9+F/lsvHp7l+ECFWDcXJQuOtDfn8MhqYtUpj0HGKggr49eojFRBTvR27YQN77CVsdRtAbtOdEHpqKqWD4YjuI7D5Df7+6+MH7wkPY31ZqK8udxtYIjHFVz482aii4EK9jFMYE1uH4yQkQXlBchogsKI5MGdv4Y+OO7gfu/QFbow33EEW4RWqoa8DjNu9tbZU7X3CdYJhFdiXTZOzSDWDIDn9uhllwtegJ6EX2c/F3nmBLTjddU9piWEMKJLlgU6LN2Dws3uqC6mY4m8asn+vJGEdDJaHOth3PVGfLoTUTspJz/Ye7hM1Y18kK8bsIWiRQoLugFuSIEXOpE724K8qLyQse56EV02ZNVgl80WSL6lPF+ZnBxLiUwSqy9lBfSTzwNPHZT1m6zTBSDGudy/FH+d5YjygUADitjjEo2FS0lqsiXzym76nzgDb/kF5omDgE/uBQ4cEf2fOTZXwM/vZqPcWleD2x/i63jCwdW83eMH8jeSSeiO0LdAIAVbIxUOZzoxx7hH1Sd6LQCR7dwwYjobimN3vgBxJI2YnHysMcoyiWdBH7/dr5XwTnvN+0TAGgi8KhJY9Es8dPpAjq3adv65qLRMPCLVwFjTFylW7eQcvQB4lLf+SMgoasQWvcS02PVDprJRR/bz2eiK3nolHPWEhH9qaOTSKT4mJWRmRgePkiqh16zvdv07ToNmovS7PMmpiJFbdxrxYk+bVDBAC0mqmAnuruGq+g4STrKL7LkoK3OB4/TgVRGVkX+/71tH2LJjPo5+eqdB3BQMbrR/9frnOhcnIvqRC9Bf7YSQUV+q9nvAuic6Ivz+idYPAgRXWCficPAT64C/vYh4NlbgPv/B/j1m4CvnwR8oRf41Ruyc7QNoC6Tol3oQHaWYG3pM9EBqBNv2mhlQ3udmvm46PHrSgQj48BTP9S23X5g2xsre0xLCOrcGJqOYT5RYtedQFAquk4DvEyjKJGLLqhyvv/gYfz7H5/HFV9/CHftGTHdz3KUy9wosOvn2nbHNmDNJbmfo3OiRyOFOtF1Yk0RE0FaBr6i0c/HsC14nAsvosfhLn0mui/Eb9sV0dnGoqWYjEsS8LJv8OfW+/4nK9Zljvk9hOnCAruQKTmA1RflfKvD1IlewTz0UmLmRO+bmMe//eF5DEwxotLGq4G33wWEerX74tMk2uXL64DfvBl47P9IE80/3si7tDtOAa7/E3EN2+CYpBMyjUR0nat81doNcEjA2WsYs4o3yOftFyyiK00lw1HIbMVC83qgtlV9DDBwoutyvbc7DmJ0pojGyDpeHCQxSpvZKJf7Pgf0P6Ftd50GXPKpnK+zpjUAh0Scz88PaNFMdNEpK84F4HPRh58nTZsBslD5q9eR6BZK/Qrgn58ErvlatpjOUtMAnP8R4Ix35TxeADoRfR+QZAQ+nRN9fWsQTQEPosk0nukPc4/9cfcJZGTgtN4GrGw2N22pzUUZEX1MF+cCaI17LYnoymvp+xfQBbqj4xGkLGSrG8Lkom9xHENbrTUtwOmQ1M983+Q8dh6bxF+eHYQkAT992+m4eEMLEukMPvK7ZzEXT+G4ch1cp3Oi074ow9MxdRHX8HO0QFDH/ehsvKQLW0saEeciqCBCRBdYIpnO4ObHjuL4bV8Hvnse0G/SGCk2DRy4DfjRS3M2TwI0l03IX4KV3zndxLlcTnQPucDu6iMTsk3tSyTKBcgW0fseA449pG2f9FrAV0AXdgEAslhEJ4fCjS6oWpwuYPUF2vbh++w1bBMIKsyz/URUmYwk8M6f78R//PF5RBPZn1nLTUUf/zbv/D3/I/ljzHTCSzxaKid64WOMAUU86G7w85VmC+5E12eiV8CJzrpercD+HYptLEqp6wSu+Ly2nU5kxbqwcS5hKnKxInrndtK7IgfUHbqmdXGKCNQpO6VrLPrZv+3BLU/24ReP9/FPaNsMvOt+YOX5/P2RMWDvX4E7/h147Fv8Y1teBbztdqA+u0ljPvYnmpCUmUqWsf3ZO7FOdJcP29avwe5PXY6PX6HLsmfd6AXGubTV+eBySMTh3c80v1Rc6LIsq+e+LBG9rgOo15p57nAcwEgJm4vuUXpHbaFO9AN3Ag9/TdvBVw9c+2PAlVtA7aivwSu2kb/V1+4mixbxVFp1bWfFuQB8f5d0Anj2V8DfP0IMX6yIX9sOvPUvQGgFcNoNwI0PZWeq954LvPoHwIf3AZd+ytrCC9e3QFcVoXOiOxwSzl5D5mDv/eUu/PmZE5BlGbIs4/dPk8/Sa3aYu9AB7bp2ghHR9XEuALNIZUFEHzRxoneFauBzO5BMy+gvNG6EqRRolmbQ6Zi0/FQqMB+fmMdn/koawb7+tBXY2lWP/33NyajzufDcwDQ+duuzan+Illr+b9YaJN+bVEZWz5nV1Fi0we9GQDHuiUgXiwgRXVBBhIgusMQv7nwUq/9xHXqf+DSQZFwgDhMBPD5NsuYO3Gn6muEouYA3lMWJXh4RvUaJc9k3TPPQl0hTUSBbRH/ie/y2ncZLAkO0XHQhoguqGNZ1G50Ehp5ZsEMRCHIhy7La5PvyzeS6/8sn+nDNTQ/hhRN8M0k6EaXxB4ZEw8CTTAVW8wZrMWZ6EX2+QBFdn1eud1TbgIobKxprqktE1znRSSZ6Nce5lNDRve1NOWNd5tg4l2iSuJNH92j76+O2dKQzsjq+WNuyOE0eDYqxhnXKjs7EcN9+EsOiby4JgCwsXP9Haw7hi/8DuPYntnLQWY5OJXBMZqpdjUR0VhCv6wIkCfU17uyMek5EL8yJ7nRI6AzVYKt0FBK7QNV7LnnZWAoRZVGxs95gAXHF6erNUx0HMTJdGsEunZGxb4jMlba0uIE7/xO45fX8Tq/4P6Ch1+DZ2Xzg0nVwOiTcu28Uu/qm1CgXwCTLmhXRAeCvHyTVtVFGrPU3AW/5M9DIRPQ0rwVuuAN4+beASz5JHOpv+wdw8usAt7XIEfI65s1/9SI6AHzk8g1Y0xLA+FwcH/z1M3jLj5/E354bwsHROXhdDlx9cofBC2l0GTjRNRE9O85lwlKci3EmusMhYXWzEulSaC4640QHgNaIcbNlI2jTze89cBjPn5hG0OvCRy4nv++2Oh8+84otAIB/PE+0gfWtwazvntMhqX0ZaPPRaspElySJyX8XkS6WEJnoggoiRHRBXhKxKC554p041/kid/+zrpNx3+W3Qf5EP3DDncDVXwFWnKXtkIoCv34j8NzvDF+XukzqS+1E9wTLdvKkcS60mcmSEtG9QcDJLGiw5czdZwAdJ1f+mJYYq5VIlyOF5ggKBJVAn7l7SOSiC6qTsbk4JiIJOCTgG284FTe//Qy0Br04PBbBq779CG665yCSSrn5iTCZiOaMc3nqB3ykyvkfBhwWhsq6OJdEoU501hDgcGcvbltkLp5SRcisOJcFbyyqc6LLHszEUpBt9NTJizfINa6zH+fCfAZKma2aJ9ZlTp+JzkZOAMDqi3O+/MDUPBKpDDwuR/6KiyqFinzRZFqNMfjD7hPquNvUQet0A1d9iYieV34J2PJqIMiI1G4/8NqfARd+LH9lSQ6OjUdwSGZedzyPE70+h4O4BCI6AKxqDuAMh06E1OWhNwZMekGsOFO92SLNYH7kcPY+BfD8iWlEk2lc5t2LVb+7jPSZkJnojzNuBDa9zPLrrWoO4NWnKm70uw5weeiGsZoNq7IX01gCrWThpXVj9mNON7D9etJ0tiWHGJ6LlvXG97tq+M+lwqrmAP7xwfPxkZesh8flwEMHx/H+W3YDAC7f0o66PHndWpwL21iUfFdYF3aTDSc6zUTv0GfpQ4uLOmRjPjM+F8eDB8bw6KFxPBXtgszIUN6xFyy/DhXRj02Qa/r7L12LFiay5pXbutRFdSA7yoVCxwLUGFdbRXEugLbgX7Dbf7khMtEFFUSI6IK87Pn7t9ALbXCXcPjw35kb8Mq5j+FtfxrF7QcjQM+ZxKn8lj/zrq1MCvjDO4Enf5D1uuF56kQvsYhepigXAPDrBqAbllKciyQB/mbjx4QLvSTQ5qKHRZyLoJpp6OWdUiIXXVCl7FWcjiubA6jxOHH+uhbc/i8X4KVb2pBMy/jKXQfwqm8/gv3Ds/kz0RMR4PHvaNuhHmDra6wdiG7hPhkr8BzPiui1bdYEfANoM7KQ303EF7/OiZ4pMMe2FGRlonuQzsiqW7YkSBJQE9K2o2Hrz00n+TifUjQWZTGKdXnoy5BlWSeiJ7KFVQMHKwuNJVjdHFi0/XqCXheJJwHJRZdlGb/d2a8+PjmfR/xr2QCc+S7gtT8BPrwH+JfngTf9Fnj/LmDLK4s+vmPj8zgkMzEwk0e1vG3KNONEzymiM68zO1RwdNrmzjqc6WB6UYV61PfV8tBNXNTdp3ObvuGdBR2Dnkf39uMLru/jh9J/Q5o6xj940uuAy//b9mt+4NJ1cDkkPHRwHPfuI5UJpjnWkqS68Tl6zwVe9k3g/U9nuaFLiq/euD9X42rT87rX5cT7L12HO//lApy/TjtnX5snygXQRPSh6SgyyoKT6kRnBGarTnRZljGkCPJGFQw0F92qE12WZbz8pofxlh8/iTf98Am89ifP4UiGma+PvGj+ZB3UoQ2QxYd/OmcV97gkSfjcq05S9YWNJnN1OhYwbVC7wNDs9wHRXNQaIs5FUEGEiC7IiZyMoevF72p31DTC88+P4n0f/wIu3EAufs8MhLXH3T7i9Dj1zeyrAP/4KHD0Qe61p5RM9JLHuZSpqSgAzsXRFapBfU31dPIuCUaut5pGYPMrKn8sSxA1zkU40QXVDutG73+S9LsQCKoMGuXCVoU1Bjz47pt34Ouv34b6GjdeODGDa256CAeGyXnX1KF76B5gfkLbPveDxJFoBZ0TPRUrgRM9WPhYhoroK2h0DRvnIqftZ4SXEp0TPeUgY8Cy5qLbcaKzLnTAUpzLTx85ivuVuBFLbHsT0HOOtj38AuKpjOq2BpQ4l9khbR8LlQmHRmke+uJ14UmSpAp9k5EEdvVNcX1krDhomRcjgvL6l5L87yLJZGQcm4jgUIZxEstpYJJxb2fSwCyz+MEK5XrqdK8zZ+MzxLCpPYDTHIwjvvc89aYqohtFuQBA+0nEHa3QMPms8X42WbX7C3i9637+zmAn8IZbgNf8wHZDV4CIp689bQUA4BtKNnrOZpCXfopUSHduJ9Es//I8iWbZ8VbAV4FKYiM3uq6pqBErmwP4+Q1n4PvX78AXrz0ZF6wzMTgxtAW9cEhAMi1jfC6OZDqj9h5rNnKi51mMmo4mEVUqQdrrsxdgaM8Fq070SCKtZqyvba3F6pYA+t2M+G1DRO9hRPT/uGoTPK5sOasl6MWP/ul0vPXsXrxqu/EihH4sUE2NRQFtsaBfxLnkJ5PmxxZCRBeUGSGiC3Jy7J4foiXDlP6e8z6gaQ0aAh5cupF0fT80oruAOl0kS+6c9/P3P3MLtxlWJkwlEaJZJ7rSjb4csE70TR1LyIVOCRhM0LZfby8HUGDKaqajfUlL1wWCUsNm78pp4OhD5vsKBAsEFdE366LVJEnCK0/twl0fugCXbSKu9IQS62IqorMRDABppm0VXSZ6Ol7gpLdUIrpS/q2KDQGdCLOQueg6J7rbRya7Jc9FZ/Pk7YjoCd2YNk9j0UOjc/ivv+7Bx3//nPX3kCSusR5mB7mMZ4A60RkRPdiRtzLh8CjNQ1+8IjoANPq15qK/fYp8L0/tCQHgs9IrzchsDPFUBscknSjH5qLPjZIqXEqu5qV6gb3ASJdt7gHUSWweurZAc4K6ic0qcJxuLj+8e+75go6BZXo+iZURnRh/2tuBf34c2HhVUa/9vkvWwuN0YEb5vhg2FaW0bADefgfwrvtINEuox3zfcmCUi56nmoQiSRIu39KO1522IjtL3wCX06Fmlw+Eo5hQolycDgkhZp6tLVDlPt/SWJjGgAc+d3YMEI1zOTw6Z2k+QxdJ3U4Jd33oAtz7kYtw0fkXaTtMHcturG3ChvYgXn5KJ95x3ipcusl8zr+9pwGfecVWU4e5viqt1ltdxrgV1Iku4lzyk9BV/4k4F0GZESK6wJxUAnU7v6Ft+0LA6e9UN9fkykOTJODy/8fnN/Y9xu2ixbmUwInOxbmUz4nu92gX4iWVh07JcjlJwI63LcihLEV6Gv1wOiTMJ9IYnonlf4JAsFCsPA8AM3Eb3Wu6q0CwUGhOdONF7dY6H37wlh342utPQcjvxikrQubZsmxWuNObVzzl0InoGf2EziqzOtG0QKgTvbtREQnYTHRgYXPRdU50j5f87srqRLfjvNc70fNMxqnTd2w2rsYoWIJ1IcemMT/LV/uE55O8o5kZ297+wjC+cPs+xFN8/Acdjy9mJzpAYogAkvH+t+fI7+DGC4iDdyaWUvscVJqjStPWeP1q/oHxA9pt/WKc1Ux0gG9IaoPumd38HYyIPpgvxgrgIl16Ukcti5lmPHxoHF0Sc445/R3ANV8lESdF0hWqwRvOWKFuV5t7mMMoT92iiF4InUxzURrl0hTwwMFEO9EFqsmIQYNehuEZ8rnpMHChA8DKpgAcEvk+jhk1+9Uxwxjn1EWBti3MHjIwZq25qNMh4ZtvPBX/ec1mSwsMZugX1KstE111oos4l/xkiejCiS4oL0JEF5gy+ejP0JRiSgvPfh9X/raulUxa+yfn1eY/WazUSgoxdRSY1cRuWkrWEChy5TcZ5aMGasuXiV7jZp3oS1FE17nV1r0EaFxlvK/ANh6XQ3UGHhG56IJqxhPgz6UzA+b7CgQLQDyVVvtLbGw3vx5LkoRXndqNp/7jMvz+3WebvyDrzg4022s+6PIADm0CLicKmPSmEry4Xco4F/21PTJW8GsXjc6J7qkpkxO94DgXvRM9tyA9NksEpIyMLDd5TnSLJPEpco6lWebhaBIy60Rn4kj++2978J37D+OHDx1V75NlWYtzaVncAkKj4pb95RN9iCTSWNUcwGWbWtWvJI2pqDTHxsn3qr2liXc1s050/bWyzo6IXpgT3dH3qHo75mshudsKWiZ6DhGdaS7qRAY48XRBx0F5Yu8R1EnMObBxtfnOBfDei9aqER45negLTbNRnEtlRHQqbLNRLgDQWKtVeeRykFMneodJDJDP7VRFXloBkwu6SMotYrdt5neyEelSCvTfiWpbkKGZ6FPzSa5fhsAAIaILKowQ0QXGpJOQHv6Ktu2rJ416GJprPaivcSMj5xAEe87it/sfV2/SQXCoWCc660IHyupEZzPRzRqVLGr0Jd+nvX1hjmMJI3LRBYsGtgxd764TCBaYgyNzSGdk1Ne4Td1yLG6nAy5njmEvm4eeJ3va+A0YN3qyABE9ostDLirORRHR1TgXnRN9IeNcUryI7vMTkbr0TvSQdttWnIs+Ez33WG+ccWGGozaiRnRRHqkwOcfSSIZ0RtZVJhDBNZORMaJUsn3r3kMYmiYi6UQkgeloEpIErG5e3E50Gjnx/AlikHntad1wOR1qLEW+POdycWyCzHVWNgX4qI5xRkSf1rnJc8W5+EL8eaMQJ7osA8c1Ef1o4BRuATBvY1Egq7lo/NjjJjtaORwZBw/u5+/M5cYvgPZ6H95yVi8ALfKiKlkwJ3oM47PZTUUBzYmeSGdyCrP0vJLr2kqbi1rJRaeLpHVshGtoJeBmxM4Ki+j66oxglTUWDfrcalWOcKPnQR/DJuJcBGVGiOgCQ6JP/woNCWbwftZ7s8rwJElSM9FML6Cd20kzJErfE+pNVUQvNhN9Vieil9GJTjPRa9xO9DYtwVVONn6n7STiRBeUlNWKQ+ywcKILqh124qsXBgSCBYaNcimmpFtF70S3CyOGSckCMkzZPHSgYBFdlmX0T+oy0WsaAIkZ8leLiC45EfQTkWbGjovbCpwTPUzERitkxbnkFtGpEx0gjkHL6FzI8jRxITcHvfC6HPAiAYmNoVGc6FPzCaSU2JhoMo3P/4NEIBxWXOhdoRrO8LEYaWTMNQ4JeI3SGJBtOLoQ0DiX1S0BXiAdPwRklIgZVgj31udehJEk/nNQiBN9bD+3APiUrLl7U+mMGh2YM86llnevp44/Yb5vHvaPzMI7r/s5SiyiA8AnrtyI7755B959Uf5GnQtGbRv5DFB89YC/sWxv16UslJwIRzGuZKI31/JGtRqPU62qnsqRiz5Eneg5Fl/YXPR8TBv1QXM4gNZN2vbonryvU0p8bif3+6m2OBdAqyYTInoe9E50Cw3BBYJiECK6IJt0Csn7vqRuyt464Mx3G+5KmxcdMruAevxAxynatpKLnkhpK+BFZ6LrnehlFNFpc5IN7UG15HZJseJ04M1/AC77DHDd7wDH4p6IVSO0ueiRcSGiC6octgy9wKxWgaBc7B0iYmfJotXYKBV9/IkV3JpI5UgVMOFlXcdAwZno43MJRJNp0ruSCiAOB++uX9A4F2aBwV2DOkW40DvRR2di+N4DhzGfKFBcZ0X0TDJ7km1GVpxLHhF9jhXRbYi7ur+vpOSf13qdaPB70C5N6vYnYuuoItp7XA5IEvCXZwfxxJEJ1cyydpHnoQOaWA4AF21oRZviztcaji6siL6qOcBHdaSiwHQfuT11TLvfinhcrIjORLkAwG0zmhg+MhtHRibNHPWxHll0n6He9Aw9bX3RSccD+8fQJU3wd9avMN65CFxOB67Y2m7e46IakCSghfmcNK21FxNmE6NMdKO/O41LmsiRiz40rTSkNYlzAbTYqMNWnOjKImmd3jjH5qKPvFDw565Q2EgXswakC8mKRtFc1BIizkVQYYSILsgi89xvUBftV7els97Dl8UyrGujIvqs4eMA+EiXoWeBRESdLEmSwQXVLhWMc7loQyuuPqkDH7x0XdneY8FZeylw3r9w+ZuC0iHiXASLBrYMPT7D954QCBYYzYleIhG9WCc6M2lzpgtoHJ3lRC/sGkyjXNrrfPC6mIVwdmFgQRuLMr8bl091Juoz0f/f3/fi87ftwy8f7yvsfXwhfttqpIveiZ7H0cbGuUzbcaK7PFzMjnOOLKLUel0I+d1oh+54lbEtdb6vbg7gjWeQXO5P/+VFHBgmx00jFhYzjUyvpNedpgnRqhN9AeJc0hkZxycYEV0f1TG2H7j/C8D+f2j35YpyobCxPoUsVg89p96clv14fK5ZderTKJeO+hquuaQhK7RIF3ciDEwcsn8sAB44MMY3FXV6C1uUXCpsf6t2+9Try/pWrIg+oYro2UY1KqLnWvSzE+diz4muE6pZET06lX0dLDNshUbAU4UiOnWiTwknek70MWxCRBeUGSGiC7LZ82f1puypBc56j+mua1rzONEBXkSX08CJpxFWLtx1Pnfxjm72guv08O6jEtMY8OD/rtuOize2lu09BEsb6kQ/EY6aN+QVCKoBvYtORLoIqgRZlrFvWBHRczQVtUw6CbDRGQVlomuTcVe6yDiXIsYyalPRRj//ALswsJBxLnonuoGInsnIeOQQOcaDuUwaudD//ti/by7YbFXJwWdWG8DHudgUdxkXsjdKDCEBrwv1NW606Z3odbwTvSXoxUcv34D6Gjf2Dc/iNzuJ+WUpONFbg4rzPODBJRu16tKFdKKfmIoimZbhcTmIO1ffNPLvHwHu/x/+vvUvzf/CrBN9dsi+E3d0r3rzmHMlZDjUBUZLeegUprkoAKDffqRLJJ7CU8cm0cmK6PXdpBJmuXLqm4F3PQC8815gxz+V9a2oiD41n0S/4lw2cqJrsUjGi36yLKtOdLPGooAmog9OxxDJ0/hyxqixKMCL6MCC5aLXel35F5oWgO5GGucinOg5yXKiL/7roKC6WcZXNYEZDiYXTzr/wzkncjTO5eh4BKl0xninFbrmon2Pq7mRDf4SlOHNMRPP2raylsoJBMXSXOtB0OeCLGtNqgSCqqROJ6KLSBdBlTAyE8fUfBJOh6RWxBWF3qVcZCa6O1OcE33e24xEurCydlVEb6hSEd3MiR7TBJ0Do7OYUITS4xMFOvD0Y9dCnOieYN4xZcGZ6ADnQq5RRPSg14UGvwdtkt6J3sG9X0vQi8aABx+5nIi5sSQZgy8FJ/rZq5vwgUvX4VtvOhUelzZVzSf+lZMj42RxZVVTgIht/ka+Ye90P/+Ei/4dOO3t+V+YFdHTCb7BcT5kmRPRw7WkaSUV0U+oIrqF5putmxF3MOeM/ietH4fCY4cnkEzLWOVmPrtlyENfVEgS0LkN6NpR9vlpnc+lRpK8OEgqBw3jXJS596RJnMtkJIF4ipxP2urNY4AaAh40Kd/JI3n6PM0YZaIDQOtmfnu0wiJ6gyaiVyPdDTTORTjRcyLiXAQVRojogmwu+y/gZd8EXvMj4LwP59y1K1SDGrcTybSM42ZNL2pbgEam8Uvf46pbJ1RsHjoAzI0y7yUc4oLqRpIkdZKbb9ApECwo+lJ0vUggECwQVCRa3RyAz12C3h16UbmgTHRNgPJkYpDtOkqZTPS9cwH8aXdhi1bUsUazVFVYwa9qMtF9qjORzUR/9JAmJBbcUE0fQ1iIiJ4nyiWZznDC+XQRTvRAXHOih/xutLMiuq+e9BgCMDpLFiGoW/tNZ/RgY7uW274UnOgOh4QPv2Q9zlnDfw9pzIttx38JoHnoK5sZobl5Q/aODhfwyu8AF33cmmhap7vO2lmsnhkE4lrMWqaFNGncM8g70XM2FaU4nJgInaRtn9hl/TgUHjhAziu9TqaKogx56AJjJElSqw7ooppxJjq5z2wxirrQm2u9fCSYAbQiPV8uOl0kzYpw9Teq/R4ALJgTPViFTUUBbTF8YCpqf0yxnGAryCABLgvnPIGgCISILsjG7QN2vBU46dq8A0CHQ8KaVrLalzvS5Wztdv+TmImQC3SoFE50tgS6tnx56AJBqVjdInLRBYuAQCvgYM7RIs5FUCXsKXUeuj4jvKBMdE1c8yGuihiWYcYyI3ID9o8UFmNCs1OznOjswkB0EsgsUJwY50SvYTLRtTiARw9rIvrQTAzxVAHHmuVED1t7HjsZz9NUdGKOF3NtO9GZ3Pva1BQ8SKLW50K9XxfnwuzHOtEB0mDxMy8nkQjdDTVq3vFSpMFPnegLJ6KvamYWKVp0kS6eIHDdrcC2N1l/YdaJDthrLsq40AGgtudkANr5cTCsNIe0IqIDiDVu1DamjtqKlpFlGfcfGIUTadSldHEugoqh/1s3B40y0XM70dWmohZigKgpKKcGADYT3WDe38a40Uf25H3PUnLu2mZcvKEFN5y3qqLvaxXqRJ+LpxC2e31ZTrBOdE9geUdICSqC+IQJimatlQtoD5O1l5gFxsjAr6EkTnSmsWiwzXw/gaBKUJvxCCe6oJpxOPgJvohzEVQJZW0qChToRNfECz/imI3bnPAy0XSjcqjg8m0qovc05YhzkTPWndmlRu9EVxrNUZElnZHxxFFNRJflAvNgS9FYNE+uKhvlAhSSic67kFulMGqVOBfOic6I6DQTvTWoOUzPXN2EP7z3HNz8dl2u9RKDzhkW0olOm8MDALa8Srsd7ARuuB1Yc7G9Fy7GiT7KC45d67YBIPOxeCrNZKJbE9FdTSu1jcQcMD9puq+eYxPz6J+MossZhkNmFr2EiF5R2L+1JGl9BFjyO9HJ56a9Lr+IvtaqE11ZJM3KRAf4SJexfaRHSYUIeF34ydvOUJs0Vxs+t1M914vmojnQi+gCQZkRIrqgaNa1EadOzu7crBMdQP3oTgAlcKJn0nxZsnCiCxYBdBImnOiCqoedAE8PLNxxCAQMVETf2JHbKWwZfQ5xoJDGotrErUZKIBK34Z5OxbljGJUbChKOU+mM6j7NzkRv4bcXKhed/V37QlmZ6C8OTmM2lkLQ58IGZXzZN1nAgrPLw/1NLDcWjbNO9Nwi+vgcL6KzkTSW0LmQ2zGBWq8LoRo32lknOrPfuM6JTtne04BVzUtbPKCZ6AshotP4vVUtzO941QXA9X8CrvwS8J5HgPat9l/Y30QaCVMKdaLXtqOjowt1PhdSGRmHRufUTPQuK41FAdS0ruHvCB+zfCgP7CfRmpe06/42IRHnUknY6J5GvwcuZ7bUk8+JbqeCYU2LhWp05HOiM9+bTBKYOJT3fZcT1I0umovmQIjoggojRHRB0VBX7cFcF9CmtWSgqNA6/QwAIFRTpBM9Mk4cVRSRiS5YBNBJ2LFCG6YJBJVCiOiCKiOWTKuu0M3lcKI7XNkuZiswTvQaxDEXS+XYWQdbUQcS51KI62xoOoZ0RobH5eCcygCyI2oWKhedjeALdqjOxPlEGsl0Ro1yOWt1k5o/XZLmolad6Ak2Ez33Ig11otcpebr2nei8iN4hTZJM9BoXWmHdib5coFE1UxVuLBpLpjGouHOzFirWXAyc+S6S7VwIkqSr+LIjojNO9NZNkCRJrc558ugkZpVzUEe9NSd6XYdORJ86bvlQHjxIzqHntemaKotM9IrCRrAY5aEDmhPdLH6KOtE76q070Y9NRJBKm0eYaZnoBtnjbJwLUPFc9GpnRSPNRRdzRlPYGDYhogsqgBDRBUXDlnJlMib5eZIErDhL3eyZew4A0BAo0onOlD8DAILCiS6ofuji0Vw8JRrFCKobttR8ZhDI2Mx5FghKzIGRWWRkIqiVTEhkM9H9TdYaAuphGovWII7ZmA1BdZYfy4ygAbOxlG1nM23C2d1QA4dD9zPonej6HPhKkIqTPHZKsI1r6DYTTaoi+tmrm9DbRCbDfQU3Fy1AROfiXPKI6IoTnVZkhu2Ku4w4DgBt0hSCXheanXPwSEwlQx3Zbz6RwlycCKN6J/pygEZTzMVTheXkF0jf5DxkmTQfbCpH5jx3nbUY55JJA2P7tW0lEmNzJxHR79lLnOEhvxsBr7Wmid5mXS502JqInsnI2HmMfK+3BGb4B/WZ74Ky0sksmBjloQOsE934GkUz0TssONE762vgczuQTMum5+lkOoP5BPm+GjrRm9eTxWuKENE5aFWZiHPJAedEX/zNtQXVjxDRBUXT2+SH2ylhPpHG0EzMfEcmF70xNYpOjCNUbCb6LO/eQq3IRBdUPzUe0u0+nZGRyOHcEAgWnHpmcp+OL4zwJhAwaHnoQUiFiN1GsE70QvLQAa6xqEvKIDJvo/RaL6LLRPzttykeUxEjK8oF4KoBASxMnMucfszWDpfTgVpF5JuIJPDUUSLGnbO2CT2KA6+vYCd6SLtttbFo3HpjUepEX99GJu2z8RSSdq7p3lrAV69uUid6U0YXL6SI7fT9atxO9Xe2nAj6XHAqi0OVbLJHo1xWNwdKd85hKcSJPnUMSDHnmNZNALQ+EY8fIZ+hTosudACAJ4CwFGLew5qIfmR8DjOxFHxuB1ozo9oDgRauQkdQfroaGBHdxIlOewtMR5OG5ys7TnSHQ8Lq5tx9ntjF4KBRJrrLCzSt07aFiM6xolHEueRFxLkIKowQ0QVF43Y6sFJxCx0cmTXfUZeLfprjABqKzUQXTnTBIsSviOgAMG8nN1cgqDT6UmwR6SJYYPYOkXHGpvYSRbkAfE53IXnoQJZYFI/mGA/p0Ynoo3IIgP3ybepUo5NuDl+Id/sthIiuNz4oYzbqTnzwwBiiyTSaAh6sbw2iV2mOerxgJ3pIu12IEz1PJjp1otNYQ6CQXHRtobJdmkCtz4VQUve30YnoLUFvecTcKsfhkNR5g5mLthzQ+KiyZc7rRXQrFYpsHjqgOdEVET2lVAZbbSpKmfYxx2LRib6rLwwAOLkrBCfrpBdNRStOW50PtAjJTEQP+T1qsZV+MSqTkTFMnegWRHRAq0g3y0WfUc6JQa+2CJZ94Fu027qGucsd4US3gIhzEVQYIaILSkK+CygAoOMUwKld0Hc49hefic65mqTscmWBoApxOx3wKM1+5pNCRBdUMWyZOWC91FwgKBN7VCd6CUX0UjjR3fzELTZvo3H07JB6My67MQ3yWgNT9pxn1Klm6ER3OHg3+kJkojM/JwBVRKeRLre/QBYTzlrdBIdDQm+jFudiGheYCy7OJZx//3SKd/fmKQunTT7b6nxqLnrYbi46E+nSLk2h1utCbWKU30cRWZdzHjqFuminKiqik+/yquYyxQSw19nkvLUmuHoRvWUDAGBdWy1cjFBptakoJRFkFs4tOtF395EFqlN7Q/xCuxDRK47b6UBbHfmbm4noToeEUI3xYtR4JI5kWoYkQX2dfNBFxMNjxtc8urBYZxTlQmFz0af7rVcOLQO6lev5iamoiAA1Q8S5CCqMENEFJWFda+4LKABSrtW1Q9083XEAoWKd6Kyryd8EOIt8PYGgQvi9xI0+H7fRfE4gqDT1OhFdONEFC4gsy9iniOgbO3JHbdiCjSnSN+C0SpYT3Y6IrjnRR+QQACKC2Y1zoU41GoOSBWs0WAgRXR/nogjI1In+tCLGnb2GiP2dIR+cDgmJVAYjszniAs1gG8RaESYTur9ZvjgXxYneXOtFg5KVbTtmhHEht0uTqPW64J7XRHRZcqp/N9aJvlyhv+dJu4sVRaA60Vsq4EQHrEW6sG7dUK9aNeF1OVVjE2Dfie5oWKltTPdb6oOy63gYALC9p0EnovfYem9BaehS/ubNteZGNdqkVy+iUxd6a9ALt9OaTESbmdJFPj0zSoPb3CL6Vn5bv0i0jOkIkeqCeCqjXgMEOkSci6DCCBFdUBLWKAO2gyN5Jo1MLvoGqQ8NzgImRSxsnIuIchEsIvxuIqJHEsKJzpLOyLh7z0hFXWaCHPhCvKtDiOiCBWTf8CxmYim4HBInFBVFJgPMM80uC85E5yduyahxPqwhjEN7BJp72r4Tnca5mIjorDPbiqhcatjYGqdHPR4qrlCT3TmKiO5yOlRB6HghuejszxufAdJ5BO64LoLHYiZ6S9Cr9viZsi2iawuVbZiC3wVgVhNRUzUtgIOMF0ZnNYFrudK4IE50RURvKpM4EyxERGdExtbN3ENslY5dEb22nWkumk5kx2bqmI0lcWCUfG+2tznI94winOgLwnsuWoMrt7bjJZvN+4SZieiDYXKOabeRpU8d7xNzxgIvdaLX1+To46D7DGPkBcvvv9RxOx3oUP4eItLFBCGiCyqMENEFJUGNcxmby11qxOSiOyUZgfFni3vjOabkVTQVFSwi/EpTsPmEcKKz3P7CMN7x8534f38XLpSqQJL4UnMR5yJYAGLJNL5x90G88v8eAQCcsiIEr8uZ51lWXzwMyMxiZoky0VMxG050xqE9KoewWnG82pkwzydSGJ8jgohhnAvANbFEbMZ4n3LCiui1baDBvPWMQ7GtzstlT9Nc9L5CctFZER0AYtO597fhRI8l05hVHJYtQa8ajzBl1yFdp8W5uKQMpMgYMKMtqkRrtLGtcKIzTvRIZRqLzsSS6vdqZbPJ96pYspzoea6zqQQwcVDbVpqKUjYXIaI3dK3j78gT6fJs/zRkGehuqEFLShdDJET0BeHSTW34zpt3qAt7RtBYJH1FB20q2mkxDx3QRPRxExGdZqLXGTUVpdR3A17m+iRy0TlEc9EcZDIizkVQcYSILigJa1pqIUmkjHUilzukczu3KQ0/X9wb6ydkAsEigTYXFY1FefYNE2FHuC2qCDbSRTjRBRXm7j0jeMnXHsDX7j6AeCqDs1c34auvO6V0b6BvsFlwJjovsKXihTnRR+UGbOsOASATZqsZqNS1Xudzod4sKs/L5MjHF0BEN6keZMWVc9Y0c00zaTRNX0FO9BC/nS9nV+9EzzEZp4KRx+lAnc+lNrycLsKJDoC40JnPQ8SjRfBomej2cq6XEo2BAhcrCuSY4kJvCXoRzCUCFkNtKyAxi4L5nOgTh4AMY8BgmzKCd6J32RTRPU2r+DvyNBdV89D1US6AENGrmKZa44qOIbWpqPXPDX2tibmE4fVKc6Ln+P5IEp+LPvKi5fdfDqjNRQttsr2USUUBMJ87IaILKoAQ0QUlwed2qif4nJEutS2IM64aFCOiyzKfrxkUIrpg8aCK6KKxKAcVgqhzRVAFsBPhaeFEF1SOf/3ds3jHz3eifzKK9jofvvWmU/Grd56J3lLGKszrRPSCM9F5ET1jVURPxoDolLo5Ijdgaxdx5EWT6axyezOoyGwa5QIAPkZEz+fKLgcmxgdWXKF56BTqRD9eCic683s2JCvOxXwyzrrCJUli4lzsOtENojwYEXXarX0ehROdcdBWKM5FjXJpLmNEgMPJNZjN60TXu3R1TvStXXXwuByor3Hb/6zUr0AG2iJWPif6LkVE394TIhnqutcSVCdm36PBsOJEt9GQlkbDpDIyZqLZ1bUzMQuNRQE+0mVkj5bvJVCbi9qNeFsWxHW6k4hzEVQAIaILSgYb6ZKL6bqN2kYxInpsGkgxmeq1IhNdsHgIeJQ4F9FYlIO6LISIXkXUMSL63HD+XGGBoAQk0xncuos4G2+8YDXu+ciFuObkTs6lXBKynOgFxrl4dCJ6wqLoq8scHpEb0F7vQ1sdEb/6LU6aj00Qsc80ygXgneixmcqLFKyIzoiGdUxW7tmr+d9/TyOZEPdN2HD2U+yK6FlxLnXG+wFqxEezIlKG/NQhXXhjUQDA5FEgqmX0Tzg0EX1UiOiqYFcpJ/qRMfK5W11OER3gPwf5FqvZPHSHC2jiI1hCfg9+/a6zcMs7z4LTYfN86fJg1t2qbedwosuyjN39YQAGTUVdvsIXJAVlJ19j0XYbcS4+txNBHzmHjxlEusxYcaIDfEVFYhaYPGL5GJY6apyLqNLNRn/dFiK6oAIIEV1QMqiIfng0t4g+EmAGe+MHgGSBq6pzuuw94UQXLCJqqBNdNBbloC6LaSGiVw9snIuc4aIGBIJyQUrDAZdDwsev2IiAN0dTsmLQO9FLFOeCpMXJ7uwItzmKEBr8Hlvl2zOxJH788FEAxIVqCutEzyR5I0K5SSf533Uw24m+orEmy0lP41wKcqL7Qvx2vmaqNuJcVFe4EmVAnZ3TUZviri+ElJMRrAZ38e+jNJpNZ2S1cd9ybizaYCL+lYuKONEBIMS4to89BAw9Z74vK6I3rQVc2dnX23sasLkzx7kgB7FaZuE8hxP96HgE4fkkvC4HiZBhRfT6brXngaD6oCI6bVZMKSTOBcjdXJS60+t8ea7hndv47V0/t3UMSxl6XRQiugEJ3QK7iHMRVAAhogtKBhXRD47O5txvwLtW25DThTcP0XeMF5nogkWE6kQXjUVV4qk0RpQBfSSRRjKdWeAjEgDIzjUVkS6CCkAn9821XjjsuintEJlgNiTA31jY62SJ6BYNArpFqRG5AU21HnXSbKV8+//9bQ8Gp2PoafTjhvNWme+od1ZXsrnoHL9YwFYPnre2Gaf1NuB9F6+Fnh4lziU8n7S/uGo7zsV6Y1F9tIrqRLfb8FKSMO9lxq8DT3MPD2XIzzARiSMjAw4JaKpdviJ6o984y7lcVExE3/Jq7XYmBfzpvaSBqBHsvEkX5VIKpIZebSOHE313XxgAcFJXPTwuBx/nIvLQq5p1reTc9viRSfz5GTKmS2dkDM+Q666dOBcAaFJEeaO+aGomulmvDkrndj7S5emfZgukyxS6qD4YjiEl5kY8WSK6cKILyo8Q0QUlQ41zyeNEP4CV/B2FRrrM6NyQQkQXLCKEEz2bwXCMSxcQkS5VQp1uMpwvr1WgkkxnLDeGFPCMziiu27oyC4asO7qmgeQTF4KLFx0clp3ovCFgVG5Ag9+D7gZr5dv37RvFb3cOQJKAL7/2FPg9Odx+Pp2IXsnmojrHPRvn0lrnw63vOQevP70n62m1XheaFbe37aZq3iDfsNFWJrqUczJOG4u21FIRvfCYkVmmeSim+7jHBlIhAJpo3xjw2o/oWEKoMRQViHORZVkV0Ve3lFmY2Xg1sPEabXvkeeChL2fvl4gAU8e0bVZ0LBH+tjXqbXnmhGmE2y61qWiI3KF3oguqlpO663HjBasBAP9663PY3TeFsdk40hkZTodku3kxdaKPGznRaSZ6vsa8kgSc9R5tOxYGnvmVreNYqrQGvfA4HdxCh0BBiOiCBUCI6IKSQUX0kZm4esE0YvdcA+Zk5uJcqIg+fkC7LTlFAxvBoiLgFSK6Hr1AIiJdqgR9Zq++eZjAkEg8hfO/cB9u+OlTC30oixKa/1z26Ao2E72YDF+Hg3OjO9L2nehR2YMZ+BHyuy3FuUzPJ/GJP5DYhxvOXYUzVuVx0S+kE10fA2Ujgk+NdJnI/l3ct28U+4ZNfg5JAmpC2nY0nPuNEoyI7qnNGUdBRW2aid6guCwLuW6F3S2mjx1NkL9Zxb4PVQ51/MeSGUTLPH4am4tjLp6CQ8rTsLcUSBJwzdeAGuY7/OCXgcHduoPaD4BZmC2DEz3Qtlo7LDljes3fpTjRt/c0EKGd/Y6LOVnV87ErNuKyTa1IpDJ4181Pq4sirUH7C3VNykIn7RXBMm01Ex0ATnodH6n2+HeAjHBeOxySurD+4mAFr9uLgaxMdBHnIig/QkQXlIw6n1sd3B8dMy+/OjYZxV6ZcRsVKqKP7dNuN60xzAQUCKoV/yKOc4kl02URuPWxBUJErxI8fn5iL+JcLLFveBbDMzE8dHAcmYxwo9tFH5dRNlgneqF56BS3liPrTseQtvJ3Z2JORuQG1Nd44HY60K00EjuRI87lM399ESMzcaxuDuCjl2/I/16+en47Pp3/OaUiK4LPejP43ibiLDs+yY8tdx6bxNt++hT++Ze7jJ5GYCNd7MS55IhyAbQGetSJ3lCEE33SYfy5m5VrMBwnwlPFvg9VTq3XBbeTCHzldqPTuUx3gx9eV4EVKnaobQWu/oq2LaeBP74HSDHuXjYPHSiLE11qWMnfYZCLHomnsF9ZvNre20AEdJkRO4UTvepxOiR8/Q2nYmN7EGOzcXz0d88CADpsNBWlNOXMRFec6FZEdLcPOP3t2vbkYeDgnbaPZylyyUbS8Jf2PxEoCCe6YAEQIrqgpNDMQFr+qCeZzmBgKoo9GSZvb/iFwlaZx/Zrt1ssTB4FgirCr8S5RBaZE12WZbzspodx0ZfuK7kLTB9bMBNbfAsMSxZ2QiziXCwxopTcpjJyRaIHlho0E73FZlm5bdhM9EBTca/FONFrEEfEyiIp494cRUiNq6BO9IGpqOEizJ0vDuMPu0/AIQFfeu0pakRYThbUic7EuThcgN/675o60ft0TvTf7iQO2b7JefPYJLa5qJ04F29uN5sa56KI2vWMQzqWtHdtHJeMfxejcgjT80SAGhNOdACAJEnagkWZc9ErlofOsvXVwOZXattje4H7/1fbZvPQXT5AL3iXglAvvx3uy9rl2YEwMjLQWe9DW52Pj3IBhBN9kVDrdeGHbz0NzbUetTK2I2SvqSigNVjWx7nIsqyO5S050QHg9HcATsYY9/j/2T6epcgN562CyyHhiaOT2N2X51q2nBAiumABECK6oKTQzMAjJiL6YDiKVEbGAYlpfJWMAFM2V1VTCWDyiLbdLER0weKCiujz8YURimVZxtfuOoBfP5k9OcrF+FwCB0fnMDWfzHIFFotwolcxrIgu4lwsMTyt5VaOiAxL21QsvqKkTnRGRJfimLOyEMhkopM8dCI0dNT74HRISKQz6u+CMhNL4t//+AIA4J0XrMaOXl0DTTMWNBOdiXqobSPxNxYxinOZT6Tw9+fIaybTMmbNrqWsEz0Wzv1GCRtOdBrnojgwg14XXEoEQnje3rVrGMYi+rDciHA0CVmWhROdgS40FeL6t8PRiQUQ0QHg6q8CASbi55GvA3+4ERjYyYvoLRsK7+GQi2A70hLTW8GguShtKnoqPfeEdeMC4URfNHQ3+PG963fA4yTn5I66Ypzo/HcykkirFVl5M9Epta3ASa/Vto8+WHjV+hKiM1SDV2zrAgB894HDC3w0VURWnIsQ0QXlR4jogpKSz4l+TJkAhet0ovfwc/beaPIwKXOktGy093yBYIHR4lwWxom+uz+Mb9xzEJ/88wu2Or0fGdMGKyMz2WWbxUCzf2mpthDRq4i6Lu32cotziU4BT/+UCBg2YIXz0RJ/V5YDFRHRZbl0megAF+dSgwTmrCySMuLyiNygCoQup0Mtqx/QVenc+eIIxufi6G3y40OXrbd+fPo4l0o60ZnYGgStR7kAQG+T4kRn8uFvf2GYq+SaNMjiBWAzzkWXiW5CJJ5Sr91U1JYkSc3rtivuDsnGiyDDaEA6QxYIaGXGcneiA1p0zmS5nehjCySiB5pIPjpFzgDP/Rr44aXA4Xu1+8sQ5QIAcDgR9TO9UAziXKgT9tQVIXKHfnGdHTMIqp4dvY34xhu24aSuelWotUOTct2a0H0n6Tje43TA57YhO531Xn778e/YPqalyLsvJP0K7twzgsNjc3n2XiawTnRXTXkWFgUCHUJEF5SUlU1URDc+sR9XXB1yyybSDJQyZFNEZ/PQARHnIlh0qE70BRLRH9g/BoA4+AbD1l2ybJXJaIndtdSJvq6VOABnhIhePbCusugkkDBvdrjk+P07gL9+kAgYQ89aftrwjHCiF8N4JZy38Rkgw5xninWiMw4oP2L5RfRkFIhpueQjshbnAkBtJKaPurp7DxGkX3FKJ3xuGxNGlw9wMG7AijrRmUx0G3noANCjiOhD01EkUmTR99an+fgIvXijwjUWtRPnYu5Ep5EFfo8TAa/m2KVxBXZF9P5UyPD+CSXmZXo+yTjRyxxvtAhQnehLMc6FsullJNYiF2VoKqrCRrronOiyLKtO9O3Uic7GuQRaSba1YFFx5Ukd+Ov7z8NJ3fX5d9ZBGyyP66qmaBxVXY0LUo5GzVm0bwVWXahtP/87PhJsmbKuLYjLNrVCloHvP3Ak/xOWA6yILlzoggohRHRBSaFxLsfGjfMpj42TiWBXSwPQzLin7JZpsXnokIDmdXYPVSBYUBa6seiDB8fU23ZiWdgqE33EQDFEE2lVmNjaRSIHhBO9itCXZi+XXPTYDHDobm173z8sP3WIi3MRTnQ7sPEVrQWUlluGdaEDJXWi+6REdpyLflw0yzfbHJUb0MCI6DQXvX9Si7qKJdPq+fslm+2J0ZAkPtKlopnozM8abLP11JZaL/weJzIyceUPTM3jsSMky76tjog3pq5kzokezv4bsFgU0fVRLhTqkJ62GecymKxFUs5eDJn1kM/j1HxCq8yoE070hgBZrJi0+Xu2Qzojq/FBCyKiA8BVXwbe9Ftg7WXGj7duKdtbe1tWq7czk8e4x/om5zERScDjdGBLp3I+YUV0EeWy7GgOkPPSbDzF9YSYidloKqrn7H/WbqcTwFM/KOoYlwrvvnANAOCPu08IgwbAx7kIEV1QIYSILigpKxr9cEjAXDyFMYMO3dSJ3tsUADpO1h4oRkRv6OUmrgLBYiDgXTgn+vR8Es/2h9Xt4xPWXcVsnEspnegnwuQYgj6XKhzZFSIEZURfmq1vIrZU0TdUi4xafio7uRmZFRMdO4Tnk0goMVMttWUUDecn+G0bzS4N0TUW5ZzoRx8CvrQW+Opm4MTT5D6diD6CBjT6GRG9kTYX1c7Rjx2ewHwijfY6n7rgaAu2uWilnOjpFBDRFm4R7LD1dEmStFz0yXn8cdcJyDJwzpombOogP89kxGShiq0uyCSB+UnzN+Im4+ZxLmb55CHa8NLmtWs2nsEIsiNd5j2tAMj3QX3Pcn4fFgmNFWgsOhiOIpHOwONyoLOARoslQZKA9S8F3vx74P27gLPfp0UytW0FVl+Y+/lF4G5aqd52zI9x1We7lCiXLV118LqUxR82ziUkmoouN+pqXGoUI1sVRM0wlpuKsqx9CdDEmOQe/jofZ7RMOW1lI05f2YBEOoMfP2yzp9xShHOi524ILhCUCiGiC0qK1+VEtyKA0SxBlmOKiL6yKQC0n6Q9MDcMzFkXJzgRXeShCxYhCxnn8vChcWQYMx6bM5uPI8z3upTuWuq07G7wo17JlRVO9CpiuTrRs0T0MeP9dMiyzDUWLXX00VKHum4b/G54XGUcqpbcia6J6H7oGos+9GXSxHTmBPC7fwLic3yzTQCjZnEujBP9rr2kpP2yza32yuMpnBN92ny/UhIZBcBcdGrtOdEBbUHh+HgEt+4ii3jX7uhWf1+TEZPrhV7Qm87RTDturbEorZrSC9qFZqLPxlMYlhuzD8dPKg1OhKNZGezLGVqtMVnGxqK06q630Q+no4DvWalpWgO89HPAR/YD73kUeNf9gLMAYdIqDb38NnMt/OPuQQDAGauUz6ws65zoQkRfbkiShKYAbS6qzQ1oLKPlpqIsDgdw7ge07UwS+PWbtUXoZQx1o//yiT4xVxJxLoIFQIjogpJj1lw0nZHViWBvk58X0QHrbvR0Cpg4qG2LPHTBImQh41wePECEwKCPHAOtEMlHMp3hBPdSumup03JFQ4062KZloIIqINgBSMyQYbk60eesiejT0STiKa1hr4hzsYeZ07fkzOtE9KIz0TUR3SfpGotOMo6xcB9wz2eznehMY1FAE45pJnomI6t56Jdtsi9EA+Cd6JUS0XU/p10nOkDETICUsB+fmEfA48QVW9vVhnamTvRQD7+t/05TMmkgyVwLvfmd6M1BD3d/QwELwLIsI2IioqeV7PgDIyRmJqDLYF+uVCITnX7naAVE1eCuAdq2lFdAB4DQSn5byUV/cXAaDx4Yg9Mh4c1nKkJ7LMxXcYg4l2VJU63SXHSuRE50ANj2ZuCUN2rbyQjwi2uBsQMFH+dS4OINrVjfVou5eAq/fCK78e+yQsS5CBaAZSui/3ZnP67+5kPY/KnbseO/78K//Ho3Vy4rKBwzEX1ompRGup0SKY1sK1BEnzpGstEowokuWIRQJ3oyLauN0iqBLMtqnu61O8hEx2qcS9/kPFKMhX20lE70KcaJXiOc6FWH08U3A1w2IrpucmLRiT6sc56L3Ep7jCoLdK3lbqKod6KXMM7Fr49z0X92nvw+sO9v6mYUXsyhxjATfWg6hlQ6g+dPTGN0No6Ax4mz1xR4rD6maVyl4lzmdA3hbGaiA4r5AsCzA0T4v+qkDvg9LjRS96OZoKoX0adMBAd2Ig4AnhyZ6KoTnf98hgqIGYkm08jIwFCWiC7BESRxLgdHyLGVtT/AIoJmz5vm4JcAahhYUW0ieqXQO9GV7813lWaG15zcof1u9OMBIaIvS2iPCDbOdUapxqqrKXDxz+EAXn4TsO6l2n3RSeAXrwaml0lFpAEOh4QbLyBu9J89esywD92ygY3l82cvRgsE5WBZiuj/e9s+fOzW55BIZXD92b04Z20z/vrcEF7+rUfQbyPWQGAMbS56RCeiU6FuBS2NDDTxObvDz1l7g/H9/HazcKILFh/UiQ6QppqV4tDoHIamY/C6HKqI3jdp3AhYD41oom67sdl4yQZuqhO9sUZtQCRE9CqDnRgv2ziXceP9dNCmos2KM2t8Lo50ZhlPcmyiNlEsuxOdmXx56wGXx3xfK5hlosfngKR+fCkDxx9Rt0blBgCS6qwGyM/vcTqQzsgYmo7hbiXK5cINLVoWsV28C9BYVBdbwy3IWaSniXeY0euX5kQ3EVS9QaCGmVibOdHZpqL0eSaMzZL3ys5Ep3Eu1q9d9DOSlYle24qgn3yeqBNd5KETVCd6GeNcBpTK2WUrovubkHYxP3v4OPom5vH350iUCxXwAAgRXQDA2Ik+U6wTHSBVF6/9KbDiTO2+6X4ipEenCn/dRc6VJ5Hr6MiMbsF+ucEaFAItC3ccgmXFshPRXxycxncfOIzTVzbg7x84H/925Sbc9MZT8e3rtmMyksBn/vriQh/iosfMic7loVPaC2guOraP325Zb/sYBYKFxuNywKXkbM4nKzf4eUCJcjljVSPWttZCkkgu+/hc/snokfE59bkAkEhnEC5R808uE12I6NVJPbPouVyd6PFpIJnfVT6iiOibO+vhkICMzOeECnJDq1zKHufCLooEinShA1yTc5+UxFxU+ZvrndgGDMshAOCc6A6HhC6aiz41j7uKjXIB+Ez0SjnRZ5mfX3IWlD3fy4iZPY1+nL6SXIca84noAO9GNxXRdU70XHEuyneZLpJRqEN6Ompd3KW5+WGnbvIf7FAXrOmiUkudENEB7TsyFUmWzYHZz0TMLUskCel67XsjTx3H9x86jIwMXLi+BZs7mfNIuJ9/rshEX5ZQJ3rJMtFZPH7gTb8BWjdr943tAx69qbjXXcT4PS7UKvFeNGJs2ZFO8c3Ci+1rIxBYZNmJ6Dc/RibEH7x0Pdes6qVb2nHW6kbcs2+UawYmsA8V0Y9PRDjnHXWi05JcAHwu+vhBvjmEGWxT0brunG4hgaCaoZEukXjlnOhURL9wPXEydtaTCWLfZP7vHm0quqEtqAoXpcpFp0707oYaVUSfjaWEe7eaYCuHpk+QZmJLHSPBTZ+jbQCNc+kK1ahCsMhFt44al1FJJ3qxeehAVh5nIqacV/VRLlK2i3xEboDbKSGoy7ymzUUfPzyBfcOzcDokXLKxtfBj1DvRK/E9Zp3ota2Aw76LvquhBrS/42u2d8OhbFBBdSLXQrAVEV0f56KMLY2uQeMmmf2FONHp9X/Go/ub1nWqr0cRTnRCo7JYkUhnEClTJV//co9zAeBuWqnejo4ewe92ksVz2tRQZeoY8yR/8bFYgkUJW3lHKToTnaWmAXjz7/mx6OH7in/dRQyt1htdhCL6+Fzcck8uU6KT4JqWCye6oEIsOxH9yWOTcDkkraM4w7lrmiHLwGNHrJVrC4zprK+Bx+VAMi3jhJJzDGjNC3knOpuLLgOje/O/AetEF01FBYuYSjcXjSXTePIoWbG/cD0ZaNCmWVZy0amIvrqlVh24lUIYnIunVNGBFdEBYFY0F60eWHdZMkKaiS1lomHjxotzo3mfSjPQ2+t8aFNyjEUuunVGld9V2TOg2QWRUjiY3LxrNUVFdP1n5pL/yHrqiNyABr8HkiRx93cruei/epKIv6f1NqjZ2wXBOtEzSSBVgc8l68SvLcxF73Y6cNbqJtTXuHHtaVpcRJOVaA823zl83HjhQO/K9wTxu5392Pyp23Hni1pjVFmWTRd5QjXkWMI2YkZm4+QaN+fV/V6CHaiv4f/OrcKJDgCo8Tjhc5MpbDmai87GkuqYZDmL6FJI+95kpo4hnspg24oQzlqtm0NPHNJuN60BdOcwwfKgyaA/xYwyhq8rhYgOAHWdwJZXadsjLwCpxScgl4qWRSqiR+IpvPymh/HSrz+IE+Fo/ieYvpDOoFAKM4RAYIFlJaKnMzKOjkfQEfJxLnRKj+KQPjxa5KrYMsfhkLCqieaia86evE50IH8ueiZDHOsUIaILFjF+L3HjzVcoE/2Jo5OIpzLoqPdhbSspVaffR0siuvJ9Xt0SUMWt0RIIg9SFHvK7EfS54XE5UOMmvxsR6VJF1HXw27P5IyoWNWaOVQu56LSirb3eqzbHLFXVxnJgrFKZ6BHWiV6KOBdecEtGqRNdJ6LveBuw9TXcXaNySK3wYVnRSIR5Grn1ks1FRLkAvBMdqEwu+qwmQiPYYb5fHn7yttPx4McuRldIW6xoVNyP84k0YkmTaykjBiI5z1cgUAziXO4/MIZ4KoPP/WMvUmnSAHwmllKbgTfrnOENASIUheetx4xQJ3rc1wI4mb9/qEeNc6EIJ7pGYxmbi9J4ucaAR41LWJYwi0+1cgR1iODdF67JWujjRfS1FTo4QbXRpDrRte9kSZ3olM5TtdvpBBHSlymlnItVku8/eASD0zHEkhnc8cJw/ieYoR+PCyd6fnb/Avju+cDnOoEvrgF+/07z+Y4RIy8Cv74O+NI68ho/vAzY85fyvV+VsqxE9LlYCrKsOUX00FXSGeF8LJqVzWQiSXPRZVk2zkQP9fITuny56NP9fHMuIaILFjE0zqVSTvQHlSiXC9a1qJMgunjYl6ep8nQ0qQ6MVzUH0FZC9wOdsK5o0AQoOuCeiS7jZjnVhi/Eb1cqT7kQBnYCP7gU+NUbSPRMIZiK6Pmd6MNKhUZbnQ9tdSLOxS6jJnEZJafkTnReRM/QiBDWie5wke/SlV/kJnyPZzarmdos3Q38axYtovt0InolvseciF748XtdziwxJuh1we0k17MJM0GVjXMBsnsdAIaNRWlsy/GJefzlWdJQkS7wBH0u+Nx8LA39+6UysuVGb3OKE73G5wPOfh+5M9ACbHtTVsVB2SszFhH0dzNZhuaiyz4PncIuPgE4q3EOl+vPP+kkH+ciRPRlC11UHOcy0cl5sOhMdJauHfz2iV2le+1FBjUaLKZM9NGZGL7/4BF1+849xYjoOie6yETPzV2fBv78z2Tx6Yx3AKsvBF74PfD9i/jzuBkndgE/uhw4cj+w8Spgx1tJj6zfXg888b3Sv18Vs6xE9IgiVBm50AHA6yT3x5OZih3TUmVVM3G5UhF9dDaOWDIDJ9MkCwDgcABtW7XtfCI6m4cOAC0bS3G4AsGCoMW5VMaJroro6zXhprdR62GQC/pdbg16EfS5SxpRweahU+pqyO9GONGrCL34VgkHa6Hc9jHgxE7gwG3Az18BzI3lf44eUxE9/2sNT5OFofZ6Lc5lsTmFFor5REoVIMvqRE/M84vypSgD1sW5ZOLK67MieqCFjH0CzcDb7wLO+zDuOeVreF5ebexEZ86L61pr0csaEQqh0k70TJpfeCrCiW6EJElac1GzXPQsEd3gu63PRPfUqrEtAPCtew8hnZFVgchogcfndsKrzDGsNt2eU5zotV4XcNmngQ+9CHxoD1DbKjLRc9CoNhcthxNdGZMs4ygXAHwMEoC3boTai0Bl6jggM2NYIaIvW6iIPhlJIKP0kiiLE71hJVDDRAotYxG9ZHEumQxw+F5g4nAJjio3X73rAKLJNNa0kLHMk0cnCz+PCye6dYaeAx75OtBzNnDjQ8BLPgtc+2PgdT8n1Xm3fSL/a/z1AyQ+6YbbgZd9A7ji88C7HwEaVgF3fQqYGSzt+1Uxy0pEpwPbZNpYJI8r99OIBUHhrFaai1Lh7Zjyf3dDDdxO3ceu42Tt9vDzQDJHNhabhw4AzeuLPlaBYKEIUCd6BRqLDoajODg6B4cEnLdWE4t6LTrRj4xpUS6Als06WgJ3repEb8x2ogsRvYrQi29xg7zwaiCT5hdkJw4Cv3gVyTi3g5FbFcgb5xJLptU83XbOiS5EdCtQR1WN21neKAV9g9hSOJh0jUVlKtKzCy+1TAPJxlXAZZ/Gs4FzyaZhnIt2XizahQ4YVJSU+XscGQNkZtxdYCZ6LhrVLF6T6xHbzwEgwp8evRPdU6s60V0OCUfGI/jrs4Pq51Mf5UKhbnTLInqMLBgF6Ge9vhtwkdfQC08iE12DNpQtR5zLwFR2ddyyROdEP7NxNnsfNsoFECL6MoZev9IZGdPRJBKpDKJKxBY1xpQESQK6tmvbg8tXRC+JEz02Dfz0KuDmVwE3bQfu/gwZR5eB/cOz+O3OfgDAF689GZs76pCRgXv25a/wNIQdWzm9akNwgQFP/YD8f+HH1TEGAGDTNcDK84EDt/MiuJ6+x8ncasur+DjmQBNw4cdIf51nflW696tylpWIHvS54ZCAGRNRht4fLGXJ0TJllSK00UaEWh66gYNqxZna7VQMOP6I+QuPM070QCvgz24QKxAsFirZWPShg2SgsW1FCPWMu43GuYzPJXKWn9PvMq0yKWXOs5ETXYjoVYivnt+uVif6dD8pHWQZfh741euJ+9gqZk70PI1F6cKSz+1AfY1b+66IOBdLUEdVa503O3u3lOgXQ8rgRFc/b5wTvRV6aCRFg4GI3hTwoM5HrhWlEdEr7ESf1ZVqB9tL/hZN+QRVby3/9zX6brMiuqcWsbSMGUXgvuG8VQCAm+49qC6GmUUNUfd4zkanDBHluhv0ZYtMPrdT7Q/idEhqDrgAaPRr+fOlhpoKaD+CZYuvjnP8uqaOZu+jF9EbV5f5oATVisflUMfu43NxLh635NpKJyOij+3PXgRdJtDx5Wihc7H5SVKt2feYdt/DXwVueQMR10vM52/bi4wMXLm1HTt6G9UxDdu82xasiB5oEU2Nc3H8MRIn2Htu9mOrLgQgA0cfyvH8R8n/qy80eT6Aow+W7v2qnGUlontcDvQ2BTAYjhm60fsUoXed0nBPUDirFCf64HQUsWSayUM3cHWsuRiQmI/iwbvNX5iNcxF56IJFDs1Ej1QgzuUBgygXgOQU0uZlfTmai9KmomvK4UQ3cH3VCRG9+shyolepiK6f1FP6Hwd+82YgZdG5WGCcy7AisrXX+SBJkvZdEY1FLUHPKWWPrtA3lwyUvrGoOxNDPJUG5pgmvLUGIroi/jb6s4UGSZLwjTeeis+/+iSc2tNQ/DFW+nvM/uxAWUR0S65kNtIlX5yLp1bNV/c4HXj/JWtRX+PG4bEIfvUEea7Z59OuiE4XrwMmVbD09ZprPdlRGssY9W9ejkx0RUTvWe5xLgDQukm7bdTAkb3e+puEuWmZwzYXpeP3oNcFZ6nPXVwuugwMPlPa118kaOPLAuZikXHgZy8HBndnP3bwTuAHlwBjB4o8Qo2HD47j/v1jcDkkfPwKEsd7+RYioj94cAzRQubCrBmiFGO4pUomTc7VdV28K5zSSIwCGM/x96aPNazKfqyuE3D5gPGDpXu/KmdZiegAcOaqRiTSGew8NpX12MOHxiFJwGkrSzBJWeY0BTwI+lyQZeLoyOlEr2kAus/Qtg/dZfyisqwT0UUeumBxQ0X0ggYONjgRjuLuPcQJefGGbAGnR/le9k2a56JTJzqNc1FznmdjkGW5qOMTTvRFgstDBkmUanWiTxzht1nR8PA9wB9vzP8asly0iE6/I/T/8bmEaZycQGNMWWwoe3RFWZzovOhWI8URiaez3VI6VBHdRJi9eEMr3nhGj+Fjtqm4E32I365dACc6kF9EZ52M3iAT2+JB0OfGOxQ3+hElntDMiU7jXKxeu6iIXus1dmrSBpplb7K7yChXJrosyyLOhYUt2x95MTvmgRXRRZTLsofGXE1E4mqFf10p89ApbJwLAJx4uvTvsQigcS7h+SRZsLfK7DDwk6uAESb60N/Mj2EmDgE/vBQ4cGfRx5nOyPjcP/YCAK4/uxcrFbPl5o46dIVqEEtm1IppW3DN4UUeuinxGQAy0dyMoJXGuaoP6GNGryFJJEqH7lOK96tylp2I/trTSC7il+7Yh1hSO9nc8eIwnjg6ics2taGjfpmX75UASZLUXPQjY5HcTnQAWHeZdnviEDBpUDI4O8Q7poQTXbDI8SsZqJEyx7l88+6DSKQzOGdNE05ZEcp6vFdxWx03caJnMrL6HV6txLlQF14yLav5z4UwPZ/ErFIy381MWKmIzpaDCqoAVpCuVif6JNMYyVsHvOVPgIfJSXzxD0D/k7lfIzrF/3wS4xLNJ6IzTUUBoNHvgUtxYRWVW7lMUONcgr48exZJWTLRdSI64ojMTPMNTA0ywaci5DxXkbgOlw9wMKJGub/Hs4wTXXKUZaLbWIiIrl/8jTNOdK+Wh96siBRvPXelGqsD5HKiU3HXXiZ6rZkTXbkWlv37sMigixWlzkQfn0sgmkxDkoDOkJgPom2rdjs5nz0/YxsRChF92dNMneiz8fI0FaXUtvK9LpZpLnp9jRsepdec5fHl7Ajwkyv5iNy6LuCGO4C338lfK+MzwG+vByaPZL+ODf78zAnsHZpB0OfCBy5Zp94vSZLqRr9zz4jZ083JY1AQKCQUk5zLZDHeqYw9UzkqZmm1nulreLXnl+L9qpxlJ6Lv6G3AW87uxa6+MK78xkP4/D/24gO37MZ7f7kLzbVefPLqzQt9iEsGGulyZHwutxMdANa+hN8+ZBDpwrrQASGiCxY9fnf5nehHxuZw664BAMBHX2r8naHNRY+bNBclsUwZuJ2S6hb3uByqcFFMTEW/4kJvrvWgxqOJCMKJXqWwLtaqdaIzzrjG1aTs9w2/5PcZ2Jn7NfRO1XZGSIiMAxlzR/nwNJnItCsOdIdDUt1CBZXcLjPo76jszlvWie4OZOeZF4LeiY44YmFd4ySDOBcaHdIQqEBPHkmq7Pd4jsk6DbQAztI3i6XXoolcgmoD0yQxFc1eDEvonOhzfKxQnc+tZqMDJcxEVxbRaw0y0dnXK3u80SJDdaKXOM6F5qF31PngcS27aXI2rBMdAIaf027H54BZ5vzWtKYyxySoWprUJs8JtadESZuKsnSeqt0+sTxFdEmS1GuR5fHlnf/Ji+KhHuBt/wCa15Lv+zv/P3vnHR/HWef/z2xv2lWXLMuy3Ftsx4nTC3EKSYCEFgIcHBDq5Ti4g99x1DvKcQcHHAfHHeUuQKhHCRBIgISQHqc7sePYcbckW5LVy2p7md8fzzwzz8zOVm2T9H2/Xn55Z3d2dyR7pZnPfOb9fQhYdbn2eDIKPPnteW3n3S+wK9Lec9nqjNkvL9/Mrk67/6URJIu9WlOncylDEWKxwq8iNs6L4vD7HTmU1nlfI6Y9vxzvV+csyb2Dz964BZ+9cQscVgu+/3gfnjo5gRu2LcOvb71YHbJHzB8+gHBP3xTmYklIUo4hPZ3b9MO2jppcOpQRopPOhVjYaE30yoXo//Hno0ilZVy1sR3nZPHpcu9nNic6V7msbPHCZtV+bfBgcD4DEzWVi/5nr18ZQpRtEDRRIxZCE13XjFMO6lddDrgahXWO5n4NY4i+fKd2W06xpnoWRgw6FwBo97t0jxHZqVqIHq6AS9Nq17W83YgjMWtoVxnaUrIsq0EgDyAqjvg5rvTltOJgUZMWfjkoTOeyUr9s/IzrBos2aE10Iby+5ZJV6gDQbPuzfMZIoSeA+ZVYXod50MTfn1/ZQjC0Jnp59xH4PskK8qEz2jbqr8Q6I+gfjO1UaqIvefjPq/G5CjfRAb0XfeZU3qHvixU1RC/kWGzyJPDiHdpy82rglj8CTb3afd4W4K2/0Z+k2PuTee0rHD7Dfr9esCpzZsJ5vU1o9NgxFU5gT3/2fesMElH9cQg10bPj9LMrAbP9G/L7jbo/Ea5mMXsNWWb7UPz55Xi/OmdJhuiSJOHtF/fi3g9djiOfvx5PfeJqfO1NOyhALzOrFHfy7uPsQLUr4IbTZn65KiwWYJ3QRj/5KPvhKDJ2SLvtbqIflsSCR3OiV0bncnBoFnftYy2hD798fdb1+BUi/Vmc6CcVByy/uoTDg8HReQSD3D0q+tABaqLXLfXeRE/Ggel+bZkf1EsS0KpdQqoOv8lGRohu8G/mULpwJ/oyIfTqUAfxUoiejzFV51LpJrowWLQcPnSO0EZ3SzHEps/oHzc00WejSaTSTC3SaDJYtCK4qngyTAzRG5ZV5C2K1rkA+p8TQIbORW2iC/8PA247fvruC/GNN+/A2vYGmKHqXAptosdyN9FvuaQX77i4t3xO/EWC2ESf71wWET5UlEJ0BbtLf+WvOFzUOMS7ZR2IpY04WFR1orsqFaIbvehLs43O95X476ycPP4NQBba3jd8HQh0Z65ntQEXvl9bjs8Bz/8kc70CCMWSGJxmx3rrOzJ/b9qsFly5ke0XFaV0yVDyUS6UFZuDDQSdOQ2kTI6ruaYrl+Whdb1+XZHZIXbFAn9+Od6vzlmSITpRHbgTPZpgP6xX5jtJsVbwoicjQP9j+sfFJnrrBhaKEMQChofooVhlmuhfvY99Zl61bRm2dAWyrsc/m0PTUdPBhyfGWLjAh4pyOsqgqOAHrMYmeqDINh9RJZzCDnCsDgfCTPfrDxCahcvLW4UTScaDf7PX4bga9a8DAKHsjaczM0oTXRei8yY66VzyoQ4WraYTvZyXAQtedDdiiM8YQ3R9G5sHv16HFS57lqJBualVE72hQk10JbiZyBUiiP5cIHcT3dmA8TltsKjI1u4AbtjelfVtuMO80Fkh2mBR8xB9dZsPn7lxCzXRDfATTqm0jNloEqenwviP+47gii8/iPf+8Fmk06UF66cmaahoBqIXXWyi636PSkDzKhBLm1bhZ/FspZvoy84GIGQBS9SLzoewj+UracyNAs//WFtefi7Qe1n29Te/Wj8I/OnvZA4WLoCjo+wYsq3BmaFy4XCly58Onin8pKixzFLOMsRipPcSplEZeDLzsRMPAZCAFRfmeP6l7O+Tj2R5PoAe4fnzfb86h0J0omL0GlqrWX3onDW72KUfnKOCF12W9QMwFvCZK4LgeJTLt8MVaKI/NzCFP780CosEfOia7C10gLUYXHYLUmkZg0ozXOSE0kRf06p3l3WUQVHBm+jGS+OpiV6nOIWTMfXYRBdVLoD+8nLxdnBYH5oZEQO2xp5Mj3WWJno6LaszAjr9ZiE6NdFzkUylVa91VZ3oZW2iaz/LPIghLQ7WtNj0WiFoIXqzz/zgsiK4hM9xJZvo6bT+hJN4QF5GuNpjNpo0PREMgJ3cELWBU4Ymelxoojt86hURrUX+P+QhwXSBTfR8ITphjstuhVcpItzy/adx2ZcexNfvP4q+iTD+dHBEnQVTLKdUnQsNFVURvejBYe1npxiiB1aUZ64EsaDRdC5xzEaVJnqlQnSXX1+OGNxTmfepc3jhIG+h6clvMm8159IP5y4k2hzAee/Wlqf6gCP3Fr19RxSVy/qO7P7ry9e3wmmz4NRkBIfO5Ng3BztB8+jRMf3VhAA50fOx4y/Z3/d/DkgIx/ov3c2KqxteAQSWZ3/+igvY5+3FO4DTwmctNAE88mXmQd/xtvK9X51DITpRMXxOm+4guDdfE93dBHSfry0fu0+7/ez3gLDww5J86MQigB8AhivgRP/Kveyk0+vP6caattyDOyRJUr3oZsNFuRPd2ERv9xfh4QMwPBPBnw6c0QWJp7I40XmIPhtJlNwoIypANTUQpZBxeflq7XbrutzrioghetPKzJ3zOfMQfSIURyIlQ5L0IbA6P4AGi+ZkfC4OWQasFkn1XFcMcZ+iXE50gA0pVXBLcVjEEy7eNqavE5jiIbqniiG6s0papvAEkBZOEjdUJkRv9DjULCCnRkVUuoif8UREH6I7fRifU07mFDnQkzvRpwtoosuyrOlcKEQvGn7C4rmBacgycMnaFrz+HKYm+NI9hxGMFn8SfoB0LplkDBdV2uji71AaKkoAaFF+Xk5Uw4kO6L3og8+x0t0So6DB9dEZ4JnvasutG1iImY+dtwBW4XfgU98qevuOjPAQ3VyBBrBS2WXrmI7lTweyK11CsSRe963H8ZfffRonBwxaEdK55GbF+cB57wFOPw186xLgT/8I3PEu4BdvYwWDa/9FW/fko8CDX2CBN0eSgBv+kxVeb38l8Nu/Ae75BPDtS4Cpk8DLP6+/2rCY91uAUIhOVBTRoZy3iQ4A6wSly8Qx5kwaPQTc+wntfqsD2FjAD36CqHPcFQrRHz8+jsePT8BulfC3VxfmqOxpZp/PgQm9Fz0ST6kuu9WGMJ63H0aC+du10UQKb/qfJ/HeH+3BBf96Py7/0oP4+1/u0w5YszjR0zIQqpAznigBY/hWbwcsk0IT3dOiDcIBMn2t41lCdFk2NNFXsq9bPJDI0kTnJ4hafU7YhSG8HWWYH7AUUNu/Pgcslgoq25Ix/UmgCjXR3YjBHhW1MZkHeZNK6JvtMueKUK2TYcFh/XKFQnSrRRIGTZYQohsvT25ZV3ITPeDmrfiE6rrPRiSRAl8lmxOdyM6bz+/B+g4f3r9rDR7+yBX4ybsvxBdetxWrWr0Yn4vhvx7Mo+0ykEylMazouHooRNcwhugjL7Lfk7oQnYaKEppaKxRPqfo6v7uCP9tEL3pkkrWllxjqYNFcx2LPfFf/u/7Sv8s4oW+KtxXY9gZt+eQjwJkXs69vwhFF55IrRAeAl29mAeyvnjud9QTov/zhJfRPsOPG2THD/gU10fPzii8D138ZsDmBp74D9O8Gzno98K4/6XVcfY8BD38ROPR7/fNXXsQG0fZeChz8LfD8j9hVSG/8MXD+e0p/vwUI7bERFWV1qxdPn5wEAPS2FrBDuvYa4IHPa8uHfg/s+xkbVsC55nP6KdIEsUDxOiujc7nnReagvenc7oyGdza4F53vnHD6lFA94LarDTtOMU307+0+if6JMJw2CxKpNAYmw2qADgBdjfoQ3WW3wmGzIJ5MYyaSQEOlBhMRxSGGb3IKSIQBRwEnSKuFeFBv9Jg3r2INCu5Mn8gyXDQ8qW+lNvawBoa3DZhVFAFZnOjch77M4C8mnUthjFbLhy6qXICKOtHdceEyVoMPHRB0LtUM0c1OhlVizsycoVFWoRAdYN+/yVAck3MFhugzp7Sv+/AftfstNkR6Xoa52OMAitcKcVe3LDMdGf93nQkn8OLQDC5e0wJJ+V7PRdnvfosEuKvlw19EvH/XWrx/lz68ddgs+MdXbcI7b38W33vsJN50Xk/GUPRsDM9EkUrLcNgsRV+BsKjxtrKhwPyk2Jn97CoTcZ4ChegEgAanTd13PzmuHT9UDONw0aHnFnw4VyyqziXbsVgiwlQuHH83sPUN5uuaccGtepf6U98GXv1fBT+9EJ0LAFx7Vie+8qfDGJgM44P/9zxue/t5sApligcPj+KnT2knv1PiFaGOBtJJFYIkARe8l/3Jxa6Psz9mLD8HeOsd5X2/BQg10YmKIu64FtTq6Nymd1be/1lgRBhis/Zq4IK/KuMWEkTtUAeLlrmJzptUm3MMEzWihugGnYuocpEMIYvarg1Gcw6CGZ2N4r8eYOHmF1+/FXs//XJ8/5bz8NdXrMGFq5vxwavWmQ7UIy96HSKGb0D9edEnTmi3jZeX25ysVc4ZzxKiTxtcyfw5YtBqDGEVzigheYffGKKzQGYqnEAsWZlBwosBfjlyxX3oc7mHfc4Lu7av45Fi8Ccnhfdpz1i9JjoX0YmeTuiLCuXE2ESvkBMd0E5CTBTaRE9GWcgvy3rP68pLMJ5g//8cNgsaitSs2K3ac7gXPZpI4ebvPIG33PYU7juonVjgPnSv05bx+5UonV0b2vGy9W1IpGT8y+8PFvw8bdC5u7JXwixEjMNFM9RpFKITTA/Z6tVfFeSvZAmm4yzAIrz+4NIbLsoLTROhuPnVT3t/or968uIPANYi/k06z9IPIN3/y0wfeRZmIgl1v3hdniZ6wG3HbW/fCafNggcPj+Ff//CS+thUKI6P3vECAKi/Xy2VGg5PEAVAITpRUXiI3t7gVIco5sRiYUE5JyUcDHnbgNd8qzJtKYKoAfwzEU+mkcw2DK0EuDKio4ggip/kGpgwhuiskbu6NbNBwJtaiZSMqRz+1y/dexjheApnr2jEq7cvh99lx64N7fiH6zbiZ++9CB/OMvjUr1zeTiF6HeEyhOj15EVPRLSmOGDuaBW96Nma6KLmAdCCNzEAzaNz6TSE6AG3HQ4b2+UqdIbAUoR/b9orHaIHDQ3pCoXoLsTRJAttTTOdS6jGOhegcifDxOGdksX0JEK5aPEWoHNpWqlfnh5ggWBwSLtv/XUYm1NO5vicJYXbAaWNzn8v/vPdB3FY8cLec0A7gUNDRSuDJEn4x1dths0i4c8vjeLhI+Y/r41oejlSuWQgKl3GjzCliwg50QmFFsNVHBVtotuc+v+bSzBEb/GymSCptJz5+y+VBHb/p7bsaQHOeRuKRiwwJqPAnu8X9LRjo+z33rKAq6CTKdu6G/HVm88GAHz3sZP4v6fZ/vg//vZFjAZjWNPmxfuvZCfsbFFxrg2F6ER1oRCdqCiXrG3FxWta8N7LV+dfmSN60UVe/c2KHoARRLXhTXQACCfK107lHkJjGzYXfGbBwGRYbZWn0jKeOMF2UoxDRQHW0uPtv2wuvhdOT+OOPSzY/PQNm4tqd4nDRYk6wdhEjwVrsx1mTJ7QLxt1LoDeiz5xHEibnLzKCNFXsL/FAHTOXOfCrwLpNOhcJEkShj+R0iUbY3Nc51LpJnoFNSPCJcWt0gw8knDSxGQfpuY6F6ByJ8OG92q3WzcU134rkuZCQvRGkxD9yD36+zZch/ESfegc7mefDsfxx/3D+IlwCfojR8bUYdkUoleOte0+vP3iXgDsJEaigKICH3ROPnQTOoUmejqpd+Va7PqrPIglTatP/7vMX8kQHdArXYb3suB4CWGzWtDizbJ/eew+/dWVF/yVTjlXMBuu1//+fOxrwNDevE87fIYVsTa0e9hJdbN9bgOv3LZMLVf9450v4rN3HcDdLwzDapHw1ZvPVk9yuuLCVX40VJSoMgXvtd387SfK844S8Iv3XVSe1yLqHq/Thp++58LinrR6l95bC7Af+utfXt6NI4ga47RZYJHY8MxIPFWWSx5TaVlt0RUToi9vdMMisUFnY8EYWn1OfOSOfXj8+ARsFglXbDDfQWlvcGIyFMfIbAwbDTmULMv43F3sUurX7liOHT1NJq+QHdK51CEZDdYZ8/VqwcRx/bLZ5eWtwn2JMGugBrr164gHHO5mwKlcgirupGfRuYxk0bnw+05PRdSTXEQmvIleeZ2LEKJL1jIPFtUOUFslQzht5kQP1yBEr0YTXZb1B9ldZ5f/PQQKCtHNPutiiN66HmhejbGj7GdAqV5s7kU/ODSL/32Undx716Wr8PNnTmF8Lo4Xh2awrbtRdaJ7KUSvCB+8ah1+8/wgjo3O4bZHT+LWK3K3pU9NsvkFK5rJrZtB5zb98omHtdvNqwELOf0JRlWb6ACw/FzgmdvY7UQYGD8MdGyp7HvWGW0NTozPxTAajEH3lR9/QLttsQPnvbu0N7BYgQtvBe75GFuOB4Efvx545z36KzwNHBkJYp10Gl8d/Qjw9UFg46uAm3+Ud6jpB65ci6Ojc7hr3xC+v7tPvW/7ikbEkiwf8qWmtSdQE52oMgXvtT3TP5l/pQIgEQeRF08zsPISoO9Rtty+Bbj6s7XdJoKoAJIkweuwIRhLIhQrT3NiIhRDKi3DImW2QXLhsFnQ1ejG6akI+ibC+Op9R/Dr5wZhtUj4xpt3YEsWv3q734VDZ4KmAxPvemEYz/ZPwW234qPXbSz6a6EQvQ5xGv4f1JPOxehobTa5AqrFsLM/fsQkWBOa6KL+QQzREyEgHsoYqsoHixp1LoDmRR+l4aJZ0ZzoFR4sGhSc6L72vAd0RZGr5WXSlpqqSRPd+DmuwMmw2UFAdJYuO7v87yFQUIhud7MTGfwkyuk9wOAe7fH11wIAxoPsNdoaSvs3aVSa6F+//yiSaRnbVzTiY9dvxOmpMO49MIKHDo9hW3cjQspQ8QYXheiVIOC24x+u3YCP/Xo/vnTvISxvcuPG7V1Z1+dNdNK5mNC8mp0gTCjKP1m4epJ86IRAi3Ds4bBa4LRVWHzQZRguOrxvyYXo7Q1OvDQMjBlLGicf0W53n8cyllI5791sfsiJB9lyeBz44WtYkM6v2DRy6kn80vFZNMbYfC0cuhs48Gtg600530qSJHz5pm04NRnG3lPT2NYdUIdIs5KFjMb0tBYsUhOdqDJF7bW985JVuOWS3pLf7LuPncQPHu8r+fnEEuK6LwJ3fwiwu4Abv8H+JohFiNthRTCWRLhMw0V5k7PV54TNWtyO68oWD05PRfDxX7+A42MhWCTgP954Nq7fuizrc7h3fSyo33GLxFP4gjIU5q+vWJOhtygETeeytC7NrGuq5VIuhUmhid6wDHBmevwzGjPjx4A1V+rvE0N08RJ1o4ojNJYZos+a61wAoF0JhkeC1ETPBv85wgdlVQyxiV5uTZw9RwBn8l58EGZTVQeLVuFzbLzUu0pN9IlQns9X40rt3//IH/WPrb8OgKYVKrWJ3qQ00ZNpGQ1OG77xph2wWy3YtaEd9x4YwYOHR/HBq9ZpTfRCZgYRJfHG81bgxaEZ/PjJAXz453vhd9lwxQbzz7zWRKcQPQOLFWjfDAw+m/kY+dAJAfHnpt9tr/zQ5Ja1rGWdVgo32YbGL2JMdYFzo8DYIW151eXzexOrHXjjj4EfvQY4/Qy7b/Y0W77lHsBnCLKP3IuPjX0MLslwYvuBzwObX51X7+ayW3H7Lefhd/uG8Iqty2BXjmlbfQ54EYVLEgpWFKITVaaovTa/y47ueZydr/jlPMTiofMs4N331XorCKLieJ02IBgrW4jOm7DFqFw4Pc1e7MYEjo+FIEnAV96wPWdrS3wfYxP99sf7MDwTxfJGN95TzEwEAWqi1yHVcimXwoTgRDfzoQOshepoYJeiApnDRWU5e4huvFx0bgxo6lUXw/EkgkooZhaiZ/usEAxZlrUQveKDRcUmehl96ECeEF2vc0mk0ur/mZZ6dqLPDgO/eS8wdph9fU4f+xw5fcDKi4GLPgBYDYcUog9dsuiHv1UA7oTN2UQH2Gf69NPstqgNdAWAFUw/yJvopTrRG4UTIv/yuq3oaWH/J16maNH2nprGZCiOuRj7ve+jJnrFkCQJn7vxLMxEkrhr3xD+6sd78ON3XYCdvfpGZjiexLiiwqMQPQudW7OE6NREJzTEJrrfXYWfbVYb0LyKXVkIZB8av4jhxQNdoUlsoQPzD9EB9jv/Lb8Evv9KYPQAu2/iGPDDVwM73sqUaG3rgZOPQv7dB+CCybHt1EnguR8UpJZp9Djwtot6dff5nDZ02Q3zmChEJ6pMwT/ZvvWWc7C+o2Feb3bD9i5sXubPvyJBEMQSwW1nHkl+Wfd8GQnyEL34g/+VLdqB47+9bhted053jrUZ7aqiQttxS6dl/PRp5pT926vXwWUvzZXppxC9/rC7AKsDSClBVT010UWdS0uWEzeSxLzoQ8+zZWNjKTSuXa4O6AcpGXfSQ2O6RX4Cy+e0mQ4K7DD5rBAaM5EE4srwv9YSG8AFIw6Gbcj0lM8LexafssUGuBp1d00pPnSLVIXhayLFzjZ46luZB+Sco38C7F7ggvfq7xeb6K3rM67aKDcF6VyA7AMQ116jngjgc0VKbaKfvYLpct56YY/uRPSygBsbOxtw6EwQjx4dw1yM/W6jwaKVxWKR8O9v2I5gNIGHDo/hltufwc/fexE2d2mfg9NTrIXud9mo9JUNcbioCIXohID4+7tqn6WWdUKIfjz3uosQfqXjaLYQ3eYGuneW583cTcBf/hr43nUsEAdYoH7vx3Wr6a4/2PAKoP9xIDrNlh/+ErD9zSXtF0iShNWeKCDuSntain4dgpgPBV/r/5nfHcRvnh/Eqclw/pWzsKbNh5dvKXPjhyAIYgHjdbKAOVKmJjofWtheQhP9VduW4dK1rfjqzdtx83lZ/HYGNEWF1q598sQETk1G0OC04YZtuZvsuaAmep0itljrpYkenQVCQjCa66Be9KIbPepiCx0whOgmOhcB7SoQ8+CNmui54Qd/Abe95BNvBSHLBp1LmfdLsx0Uetsy3OtTIfazrdHjgNVSxalBNhe7/J2T72TY5Incj7/4K/2yLOub6BX2oQNa+3EqnEA6LWdfMVuIrqhcAKiN5FKb6Fdu7MCeT12Nf351Zui4ayP7OfLgoVGEeBOdQvSK47BZ8K23nIudK5sQjCbxtu89jeNjc+rj/PiWWug5MA4X5VCITgjwq4KAaobowtWHE8eBdHmOqRYKms4lS4jecyFgK2M5oaETeNudTJ2Yh4d9r2AamMs+rN05NwI8+a2S377XZcgjqYlOVJmCQ/TRYBT//eAxXPGVh/DW257C7/YNIZ5M538iQRAEkRW34kIt12BRPrSwo4TBfN1NHvz43RcU1EDnmDXRf/bMKQDAjWd3we0oPQyjJnqdIrZY66WJPmloHmXTuQB6L/rMKSAu7IxP9+vXFQM3Twt03RoxtEduHzqghesUopvDf4ZUXOUSntTcqUAFnOhZmugmB3nc313VoaIAuyLDVcTJsPCUdrthGfOZiiejTj+jb7PPDulPMlXYhw4AjYqHPJWWMRvN8TtDHBbMkazA2qvURX5JfKlNdABo8TlNXcBXrGf/Dx4+MoZZ5Xebl0L0quB2WPHdd5yHTcv8GJ+L4Y3feQIHh9j/fTVEp6Gi2WnfDEO/lGmdyv0zlFjQtIo6F1eVQnRxvy4VY/t2S4g2oxN9ekBriQPlUbkYaeoF3vF7YNMNgNt8YOl/Jl+DJzb/E5upcP579aH77q+z/bESWO6Y099BITpRZQoO0Z/8+FX4h+s2Yk2bF7uPj+PvfvY8LvjXP+Ozdx3AoTN1chBNEASxwPAqIXMkUa4meuk6l1Lg7drRYBSyLGMmnMA9B5hv+I0FttmzoQ0WpRC9rqjHJrrx8t1czTjjcFExgM9oogv/h602wCMcKITGdaueUT975iE6vzpkNpos25Uniwl+8Ff5oaJn9MsN5XaiZ2mi+zK1MbyJ3lzNoaIcZxEnwyLCgW7PhcDNPwSu/VftPjmlb72JLXSgKk10p82KBiWMnsildGk0CdF7LlQ/2yFh0HepTfRcnLOyCQ0uG6bCCTxxYgIAOdGrScBtx4/fdT62dPkxPhfHm/7nCTw3MIUBdaholpNgBPMhNxtUaa1r2Uk5glAQTwpXVeciMn7MfL1FiqpzmY1BlmXg5KP6FVa9rDJv3LKGtcw/ehL4yAnglj8CN3wduOzv8cWmz+KryZuxvlPRQdvdwBUf054bmwUe+2pJb7vMZgjRSeeytEmny/enQArea2v3u/BXL1uDv3rZGuw/PYM79pzC3S8M4/bH+/CDx/uwtbsRbzpvBW7c3kWNCoIgiALxqE308upcOrK0YcsNb+olUjKmwgncpVyltGmZH1uXB+b12qRzqVN0TfQ8LuVqodNNSLqBnxlkHGwd1YYeik10b1ummsPbBoRZ8KXzagMYUXQuy7J89hqcNrjtVkQSKYwGo1jZUllH9EKjHO3fghBVLkAFBouah3Cyr83Y4cTgNGu/VvzEgRlFNdEntNu8cdZ7iX4+wrH7WSMN0PvQIVV8qCin2edAMJbEZCiONdmKaQGTK63WX6ve5CoXt92qnuQuJ3arBZeta8Uf9p/BsDpHoYL6IiKDFp8TP33PhXjn7c9gT/8U3nrbU+oVRD2kc8lN51n6E8+kciEM2KwWNHnsmAonqjNYFMgsR0wcA9ZdXZ33rgP4PkQsmcZsNImAeFLb6QeWba/8RnhbAO/FwMqLIcsyfr77PgAJ/UzFs98KPP4NTaX41P8AF9wKBJYX9VatkjBY1N2cOdicWFr8c7lOokjApwu7OqLgJrrI1u4APvvqs/DUJ67Ct996Lq7e1IGXhmbxyd/sx/n/8mf8wx37sKe/tMszCIIglhIe3kQv02BR3uYsRedSCg6bRW2djMxGVZXLG3d2m17KXgxqEz2aYM0Koj6oyya60DoKrGADULPRsga6S9L5c9NpoG+3dr+ZO1m8ZNToROc6lyxNdEmSBKULDRc1wl2epcxzKIqgIUQv92BRh3kIF3O2Ztx3cjwEAFjd5ivvNhSCs8CTYbKsv+SaN74cXqDnIu3+4/ezdQF9E711PWuwVgH+u2hiLkcT3ebM9Liuv169qfnQHfP+HZaNKzbo9Rc+Jw2yrDYBtx0/etf5uHRtK8LxFE6Msc9iN4XouTGeEKMQnTChRTkZXrUmuqcFcAnFnYmj2dddhLjsVjQoVzSNzUb0V4atvKTqIfP4XBxT4QQkCVjbLvz+t9qAK/9RW07FgEe/UvTrN8nT2gKpXAhZBpwNrCRR6h9nA4DCs4Z5faJsVguu3dKJa7d0Yjocx2/3DuFXz53GL/ecxh17TmN1mw9vOm8F3n3Z6vwvRhAEsQTxKA20UBn0DvFkGuNKeFAtnQvAHMaToTgeODSKl4Zn4bBZ8JodxbUKzOA734mUjEgipbb2iRojHqjUixNd1Lm05PChA6wpHFgBzCjqlnHlYOvovcD4YW291VdkPjdXiD6TW+cCsIC4byJMXnQT+Pek4k50o87FODB2vtjNQ7igrQnG/xk8uFvdWoOrEsTPca6TYfE5vUNeVBqtvQo4+TC7PT3APocta/RN9Cr40DktSog+mUvnAjClS3CY3W5apWsxVuOKCO5F53ipiV4TPA4bbnv7Tnzg/57HfQfZyTVyoufBOFyUQnTChFafA8dGqxiiSxK7ynDwWbY8vrRCdIDtOwWjScwOHQaCQ9oDqy6r+rYcGWFN8ZXNnsxB8ZtfDXTtAIaeZ8vH/lz06zekhRP/3syCArEEufD9wBUfLf35D30RePjfCl69pCa6GY0eB95+cS9+9zeX4pGP7MLfXb0eQ9MRfOGPh8r1FgRBEIsOj50Fw+EyhOhjSoPObpXQVEXHLm+O3vYoU2pct6UTjWV4f4/DCpuFNQFJ6VJH1HsTPV+IDjCPq/pc5WBr99e1+6wONgTJSI4QfayAJjUP2ClE1yPLMvb0s+GVa9or3FoWm+iuxtxXLZRClhB9Co0Z9/Em+qpah+i5ToaJKhdAP0BszVX6x47fz8JpcehuFXzonGY1RM9zpcfZb9ZuX/w3OqfzmHIiurWCIXq734UtXdrP0QZqotcMl92Kb77lHLz38tV4ywU9WNNGmq2c9FyoXY1i9wC9l9Z2e4i65O0X9eLSta3YtaGKQ2dFpYu4T7hE4F50a98j+gcqMVQ0DzxEXyeqXDiSBKy9RlueGQTSxR0De+LC1XEUohPlQJa1qykLoKy1vlgyhT8fHMUf9g/j4SNjiCRStRmWRBAEsUDgDbRwGXQuWpPTBYuleoOeOpTm6FSYBd3zHSjKkSQJAbcdE6E4ZiIJLAvQwK+6QOdEn2U7HbUcLBaeBKLT2nIhzbiWdcDxB9jt8WPAwJPAwBPa49vfbD5w0ieE6OFJIJVUL5OdjbLPcGOO5hVvWfPAnWAcHglieCYKp82Ci1ZXeECU6EQv91BRIGuIPioHsF5YDkYTqsKmtxYheqEnw8IGPaM4wKtjCxuYyr+nxx9gV3mIVLGJ3uxln6/JUJ6True+A+g4i91efq7uIbWJXuErInZtaMeBIfZ9pyZ6bbFbLfjEKzbVejMWBq4A8Jd3Agd/y2YJVOJnKLHguX7rMly/dVn+FcuJWKCYHQTiocy5NosY/jurYVjYl3U3A+1bqr4tR0bY4M8NZiE6ADQK+wlyCgieKcqL7ohp+yWyJ3PeDLHE+Fg/YJtnIeayDwMX/XXBq887RJdlGbuPTeDOvYO498AZhGJJSJKEy9a14o07V+DqzWV2TRIEQSwi3A4eos+/iT7KQ/QqD6kT9RUrmt1lDcHUED1MTfS6QQzf0gkgGc06TLEqGBtHzYU00YXGUjwI3PtJ4UEJuPiD5s/TuRdl1tJt6EAylcZcjIXo/hwhuuZEpya6yAOHWHP54jUtmZf+lhsxRPdVYB81ixN9OKk/mOwbZ0NFW32O6l3yLlLoybCMEL1Juy1JwJorgX3/x5ZPPgq0i2GklKl/qCDNXvZ9zNtEB4DunaZ3q070Cg+4vWJDG/7rQfazy+ciVRmxgFi2jf0hiHrCODR+4viS+n/a3uCEhDQ6Jp/R7lx1GWApm3iiYLQmepYrC40DvmdOFR6ip9OwRLQr5GLO5gxVnizLuO/gCM5Z2VTx3+VEHSBeWVkqNif7U+jqpb7P/tMzuHPvIO7aN4TxuRhksInm77t8NW46d4U65ZwgCILIjtfBdS7laKKzg/9qDRXliKH9zeeuKGsLvkEJl0jnUkeI4RvAAriahujH9cuF6FyMbXXu0QSATTfodS8iRn92aAxo6EAwqn1+/TkCMbUpSyeFdDx0iKlxrtxYhUu/g4ITvRItyixtmIGY/mDyxDhraq1urcFQUaDwk2GRHE10gCldeIieCAHP/VB7rHVd1YaKAtrnayKfEz0H1Wqin72iEes7fEikZLpqlyAIYr60GkP0o0srRPc7sUE6DW9yWruzBioXWZbVEH19tia68Yq1mdOFv0F0GpKsFc9mrY0ZIfoDh0bx3h/twY3bu/Cfb95R+GsTRIEUFaIPTIRx595B3Ll3EH3jIcgAXDYrXn32crxhZzcuXkNOIoIgiGLgTfRQbP5NdN5urfZJTO7hs0jATTu786xdHAEK0esPp2GnODYLNNTwqrNJIUS32NjQwHwYD7ZELv277I959QMBuft5Nsr+f3odVtis2Vs/vCk7NY+Qb7ExE05gzwDzoV9RDX+qrolegfeTJKZ0SYTVuxKyFQMRfShbUx86UPjJMGMTXXSiA8CaXQAkAIpLUnSoV9GHDhQxWDQH1Wqi26wW/PFvL0cqLef8mUEQBEEUQPNq6H4XGQsWi5z2BhcuthzQ37nqZVXfjpHZGILRJKwWCauzzZgwa6IXimEe0RT8MO7JvXCaDR51V/rKRmLJUnCI/tpv7sa+U9MA2I+mrcsDeMPOFXj12V3wu2ggDkEQRCnwJnqkDDoX3kSvts7lvN4mLAu4cPWmjrJ7y3mIPhvVN/VlWYZUSw/3UsZpEr7VkvEj2u2mXtVRnpOGroygEwDQe1mGI1mHcYBRaBwAMBvJr3IBRGczhejcJ//I0TGk0jLWtfuwotlchVI2YnNAfE5b9lXI52v4vzUBP0bm9P/maoheq0GGxs9xtpNhYihusWWeRPO2Asu2A8N7M59bRR86IA4Wrf8mOgBYLRKsVZxfQhAEsWixu1nDeWaALY8fre32VJn2BicuEkP0hmWFzQgqM4eVFnpviwdOW5YQ2+4GPK1AmO1DY7qYEH1ctzgm+7HBsMoJZf8qa4hPEPOk4BB976lpNLrtePXZy/HG81Zg0zJ//icRBEEQOVGb6GXQuYwGWRO92jqXFp8TT3z8qoq8dsDNfk2JTXRZlvH27z+D0dkofvs3l2TfSSMqg9E9F5upzXZwxoQQvdW4K50Fi4VpX87s199/yd/lfp6xuaw0YngTPV+pgGsbpsJLOESXZeC+fwKe+g7QeykednwCALCrGioXsYUOVG4onmG46LgcUIeIck6M1WET3QxR5+JpMfemr7nSPESvchOdh+gToXhJJ1plWVab6G3kUSUIglhYtK7VQvSJJRai+6xot7yk3bHqcvPf1xXmxBgrKqxtz6NyC3RrIXoxOhdDE/1MMlMZc3yUbcOathrp8ohFT8HXD37jzTvw1Ceuxmdu3EIBOkEQRJnwOlkAXI4m+pkZJUT3L56ZFGoTXQjR952ewSNHxnDoTFAdzkdUkYwGa7A22wGwNrM4WLRtfeHPNQ6h6tgKrM1zMsjh1Qekc0znwk/y5BsQ2aToXMLxFKKJ+X/mFySPfgV4/D+BVAw4fj98h34FgA1arDjGEL0Sg0WBjOGi43IAY7NaiC7LstpEX1OzJrrhZFh02nw9UediVLlwTD83UtV9tC0+FqLHk+mShnXPxZKIJtIAgNYG8pQTBEEsKMT9uvFj7KT9EqEj2g+/FNHu6L2sJtvRP8GOy3rzFQQaBS/6PEL003H9+6TT2v4VNdGXCE9+Gzi9Z36vcfpZ9joFUnCI/qptXXDYyNlHEARRTrjOpRxNdO5E76iyzqWSmDnR73x+UL09MRfLeA5RYQptsFaDqT42FJFTaBMdyPSiX/K3hbV2RC+6qnNRmuju3Bf4+Zw22K3sPZak0uXQH4AHPq+76+zkXvicNpzXmyWgLSfiUFGgciG6wS0+JgcQjCXVk6VjczHMxZKwSKi8wiYbxs9xLMvnWNS5eLL8G3WfDzgMja+WtZnqlwrjcdjgsrNjlVI+X+OKcsfrsMLjKGpsFEEQBFFrxP26eFAtOiwFfON79XesOL8m29E3oVxl15InwBaHixblRNd0LgnZitNh/Qnv4dkoIokU7FapdvtXRHW552PAsT/P7zWO/Rm49+MFr17yHuLFX7i/oPUkScLuj11Z6tsQBEEsarjOJZpII5WWS/ajRuIp1Rvevgib6DxET6bSuPuFIfXx8aUYRNYaM5dyrRg/rF8upom+8ZXAw//GbnecBWx5bWHP87YB0/3stmGwaD6diyRJaPI4MBqMYTIUR1djeWcI1DWjLwG/fk/G3RdZDuKytS2wV2O4YobOpVIhuv7gccrSCIApt1a2eHFSUbl0N+VwhlaaQmcb6HQuWUJ0m4NdOn74D9p9Vfahc5o9DgzNRDERihd9AF1NHzpBEARRZlrW6JcnjtZ28H0VkYa0Jm7K3gCr8WrLKsGb6CuLCdFjs0B0JlMXaYbQRJ9EA0ZDCd3DXCfT0+ypzn4lUR9M9QEnHp7f84ug5BB9WGk85mJFkweNHho6ShAEkQ2v0HaLJFLwOUv7scx96G67FX7X4mnQGUP03ccn1LYgQE30mmB3swGDaeXqiVo20ccMIXprESH6su3ALfcwl/OW1xU2kBTQe9G5E73AwaIA8zaPBmNLy4sengT+7836oZ4KHdI0blxRJS2TGKLb3JlBcrkwNNFjzlYgDowGYyxEH6+xDx0oook+pd3OpnMBmBddDNGr7EPnNPtYiD4ZKv53A/eht5IPnSAIYuFhDI7HjwK9l9ZmW6qNoLOYajwLrZbqB8jJVBqnJrnOJc9J7EC3fnnmdPEhuuxXT35zyIe+RHnhZ+xPqchyUTMESk5aTn7hlab3R+IpnBwP4T/vP4oDwzO47e07S30LgiCIRY/LboEksZ/d4Xiy5BB9RPHtdvidRQ9Tq2d4s5eH6L8VVC7AElVi1BpJYuEjb6jWtIkuDBX1Ly9eH7HyIvanGLyt2u0542DR/J/fJmW46JL5v5tKAnfcAkyd1O5r3wyMHlQXL7G+BODyym9LUAjRGzoqN3TL4ERPuVuBIDCq/Jw+UQ8hus0FWOyaDinbybBCdC4AsO7l+tdbVRsfa7OXBeATc8V/vqiJThAEsYDxL2cnyJOKG1ycmbOYic0BY9pQ0SHfFrTmWL1SDE5HkEzLcNos6GjIc1W0MUSfPgV0bMn/JsI+ybhJiH5C9aFTiL5kOPsvqv6WZa8ruh1WbO7y47/fcg6u//oj+NI9h/GVN2wv99sQBEEsCiRJgsduRSieQjiWAkpUyHIf+mJSuQBas3c2kkAknsK9B5jT+PL1bXjkyJiulU5UEZcQotdLE72YFvp80DnRxwBZVk/yFNpEB4CppRKiP/EN4MRD2nLTKuAdv0fsP3bAmZgGAPjPPAHgfZXfljnBie7rrNz72A0NrIZ2YFS7YujEWI2HigLsBILLrx2Qmp0MS0S0MAIAPC3ZX69pJfDq/wae+yGw6VXsSo8a0OIt/SQVNdEJgiAWMBYLm8cxsp8tL5UQfXgvIKfVxeOODcg21luWZXzu7oM4MxPFf/3FOSVrRM3oU1UuHljyvW5jj365UC+60ESfgB+ToZhOh3p8jDfRaajokuE136z6W1bsOg+rRcKuDe146PDSGehAEARRCm5F6RJWhs6VgjZUdHGF6KLO5b6XRhCKp9Dd5MY1m5njkHQuNULUYMRmarMNsqxvorcVMVR0PngFnUsqBsSCwmDR/CF6k5etMxlO5FlzkXDgN9ptRwPw5p8BnmYcdAqHeH2PsX/PSiM20UUtT7kxhOh2PwvsR5XG1MlxdpC3qrXGTSnx0mmzk2HhSf1yLp0LAGx/I3DL74ELb53/tpVI8zxCdGqiEwRBLHBEL/r40dptRzU5/axu8UWsybIisPvYBL6/uw9/fPEMjowEy7oZ/cpQ0bw+dICdlLcJx6wzpwt7E53OJYC0DEwI+jZeUqAmOlFJKipLGp+LIxQrPRQiCIJYCnidbLBcOJ4s+TXUEH2RHfwHlLkasWQav3iGtRRec/ZytPlYUDKxVNq89Ua+8K0azA7qHdtVa6IbLpINjalDffMNFgXY4EMgdxP983cfxEfveAFyNYLlShOZ1m5veTXQvhGJVBp/CK7V7g+NAWOHKr8tohO9oZJNdL0T3dXE3mssGEMylcaA4gxdVeumlO5kmFmIPqFfztVErxN4iF7KVUrURCcIgljgtApe9Kk+ILkEjhMGNR/6oNyCewckRBOZGZwsy/jyn7QrOEeD5S0i9Y0r+zaFqOokSa90KaSJnkoAEW1OS9jBTuzzE+ChWBLDM+x4mJroRCUpOURPp2XTP8lUGjPhBH615zTufmEIGzpLdBMQBEEsEdx2FqKH5tVE5070xdVE9zls4FcEPnZsHADwmh1daPFx7y010WtCvvCtGhiHilarie7v0i9P9QlN9AKc6Lwpm2WwaDiexG2PncTPnz2lHgwsaOIh7baD7RPu6Z/Cg/GN+vVOPlrZ7UglgPC4tuzrqNx7uRq121YH/E2s9T4ajOH0VASJFHOGLqv1z2txuGjU5IqSiKGJnsuJXifwFvl4Cb8bqIlOEASxwBGHi8opFqQvdoQQ/ZBlPU5PRfDNBzNVNn9+aRT7Tk2ry6Oz5d3H7FOb6HmGinICK7TbhTTRDSf2Ey52Yn9MvcqPvX+L14FGpbBCEJWgZCf62k/+Ie86MoBbr8h+OQlBEAQBeJVhopFyNNEDiytEt1gk+N12TCvqiy1dfqxtb8AJxXlXyvA4ogzowrcaheiiygUAWqsUores1S9PHMNMhO3rFNREz+NE58MnAfa57mp0m663YNCF6KwZ9PCRMRyTl2PW2gR/SmkV9T0CXPDeym3HnEEvWMkm+sZXAI/9B5AIAdvfjPYAO6AcnY2qB3mrWr35naGVptgmej6dSx3QrgTgpTTseHu91UcH3wRBEAsSk300tFXpSsVaMDvMrsxU6NpyKfAs8O2HT+C153SrrfB0Wsa/Ky10m0VCMi2Xv4muhOi9hehcAEMTvYAQPTSuX/a0AJNaiK750EnlQlSWkpvoywJudDVm/lne5EZvqxdXb+rAD995Pq7dUsGDFIIgiEWAx6E00eehv+I7QotN5wJoXnSAqVwAqE30YCxpesliNr50zyFc8sUH1AF/RInUWxPd3ZSpWakU3jbAKehsxo9iNspO8gQKcaJ7cjubR4Rm0MjsAr/SIp3SD6ZUQvTHjo4DkDDTcaH2WN9jQDqNiiEOFQUq20Tv2AL8zdPAO/8E3PB1NdgdC8ZwYpz7OuvgUuNinegLqIk+VmQ4IMuy+hzSuRAEQSxQWo0h+iL3ogstdADYuPMKXL6+DfFUGv/02xdVLeBdLwzh0JkgGpw2vPE81gAvZxM9lZZxalIbLFoQYhM9OMyuGMyF4EMHAIsy24afAD8+Vkf7V8SipuAm+vd3n8SOniacvaIRALD7Y1cW/WbPD0xh76lp3HLJqqKfSxAEsVjhIXq4iDBYRJblRTtYFNDavZIE3LC9S7nPBrtVQiIlYzIUL7it+4tnT2N8Lobn+qdw3VnLKrbNi556a6K3rmf/QaqBJLGDNOXAJT1+BNHELgCFDRZVm+hZdC5iM2ikzJfalpt0WsbxsTmsbfdBMvv+iy10AHD4MBmK48Uhpg4JbL4SGPojeywyBYweADq3VmZjxaGiQGVDdIA1rJSWFQ92J0JxHFUGeRXkDK00+U6GCe5RSBa9pqZOaW9gvwMnQsw/b7MW1heajSYRT7GTOKRzIQiCWKC4AmwAfEi5+myxDxcVQ3TJAqlrBz53I/Dyrz2CR4+O4w/7z+DaLR342p/Z9+E9l69WCx/lLGoMTTNVncNqwbJAgVdQNgohupwGZoeAppXZ1zc00R2BDgBR9QT4CWqiE1Wi4Cb65+4+iIcPj+VfMQcPHxnDP999cF6vQRAEsdjwOtj5zHCsNJ3LXCyJsOJTb/cvvoN/vrN30eoWdCq6GkmS0OLlXvTClC4TczHVk1vK0DlCQAzfUjEgWYPGtNhEr9ZQUY7g3JTHmXdSkoAGZ/5uAg/RJ0Nx08Gh+iZ6fYfo33r4OK75j0fUg7MMMkJ0L3YfG4csAxs6GuDfdJX+8Up60ecMIXoldS4Gmj0O2BR1y9N9rN29qrUODvKMJ8OM/x9FnYurEbCUfAFr1Wj2OmCR2JdSzOBp/ruhwWmDS5lTQhAEQSxAxOGiE5lu8EXF4LPa7fbNgMOL3lYvbn0Z0wx+7u4D+OET/Tg5HkKz14F3XroKHX6uPSvfPmb/BGuh97R4YC1UVSfqXID8SpeQXsvnbmRliLE5rnOhJjpRHYpyog9MhvH4sfH8K+Z4PkEQBKHHzZvohsGiLw7O4Pf7h/E3u9aq3nQzeJOgwWWDx1HyqIu6ZXOXH48dG8dbL9S3E1p8DpyZjWI8VFiAe/hMUL1dytA5QkAM3wAWwPnaqvf+4Un9kMhqDRXlCJcLW4ODcCMKm9NXkOOa61wSKRlzsSQaDB51UUNxpo5D9HRaxk+e7AcAfOuh43jtjuXoNbarTUL0x15i/26XrmsFmlcDDV1AcIg93vcocNFfV2aDxRBdsjKXZpWwWCS0+pw4MxvFiTHNiV5zxJNh6QSQjAJ2oUEm6lyq+P2aD1blez0ajGEsGCv46iwaKkoQBLFIaFkD9O9mtxdziJ5OA4PPa8vLz1Vv3nrFGty5dxD9E2F8Timx/vUVa+Bz2tCmXLFVzia65kMvUOUC6HUuADBzKvf6Iwe02+4mNDU2AujHWDCKdFrGyXFqohPVoai05TfPn8Zvni9A+p8FGUCNRygRBEHUHTwgDxsGi/7rH17C48cnsKrFi5vPW2H2VABY1CoXAPjwNetx884VWNuu3yniXvRCm+iHRyhELxtOQ4geq3KILrbQgeoNFeUITXQAWCWdwYxrU0FPdTuscNutiCRSmAolMkJ0sX0+WiUn+pGRIP780gjeflFvzhN2Ik+enMDQDNvWeCqNf777IL77jvP0K8XndIuyw4vHlDLGZetaWX1/1WXACz9nK/TtZh51SwWawEHBie5tq8x75KDd79SdFFlTD00ps5NhYogeEUP0+vehc9r9LERnLbtA3vUB7XcC+dAJgiAWOOI+WmgMOPg7YPONtdueSjF+BIhrxzZiiO6yW/GZG7fglu8/AwDo8DvVMhJvoo8FY5Bl2VzHVyR9yryXlYUOFQUAfxdYOqhcBZcvRD8ttO6Xn4s25bh3LBjD0EwE0UQadquE7qYCdTIEUSIFh+ivP6c7/0oEQRBE0bjtmU10WZaxf5B5g/m08Wzw0K1zkYboLrs1I0AHgBZVi1FCEz1IOpd5kRG+zVT3/ccNIXpblXUurfoQfbU0jBPubQU/vdnrwOB0BJPhOHoMrZ3RGjTRv3TPYfz5pRH0j4fxbzcV9nX8+rlBAMAla1vw1IlJ3H9oFA8eGsWuje3aSoYm+lDYisHpCBxWCy5YpTSbe4UQPTYDnHkB6Nox768pA7GJ3lBhH7oJ7ULDucljR6NyRUJNMZ4Mi87ovzeizsW9cEL0Nl/xw0XVJvoiVKIRBEEsKdZdA9z3T1DD2d99AFh+TqY+xIx4CBg5yOaz2Ov8uEpUuQBA907d4q4N7XjVtmW4+4Vh/L+Xb1BVZfyKq3gqjelwAk3e+e+P9Ck6l6Ka6DYnm0/DB79P5wjRI9P6ff/lO/VD28d4E95b8CwUgiiVgkP0r7xheyW3gyAIYsnidWaG6KenIghGWTP95HjI9HkcfjneYvSh54KH6NRErwFOQ7vTbChhJRkThora3ECgp7rv37waYntmtTSMcXfhF/c1ee0YnI5gysTZXIvBoqcU3d7Pnz2FG7Z3MdVKDsLxJP64fxgA8HdXr8d9B0fwP4+cwOfuPoiL17bAaVNa3oYQfc8w+3rPXdmkaqyw6jL9i598tPIhuq96PnQOv3waqBOVC5B5Msz4OV6AOhdAGy5azJUcaohOTXSCIIiFTfsm4NIPAY99lS1Hp4FfvQd4x925r0ILngFufyVTwARWADf/QNfurjvEoaJ2L9C2MWOV/3jj2fjAleuwobNBvc9ps6LJY8dUOIHRYKwsIXr/RAlNdIANF+Uhei4n+tBz+uXunWjzsd/1s9EkDg6z/RfyoRPVgE7TEARB1Bi3I1PncmBIa/Zyz1w2FrvOJRtc51LIkNB0WsYRcqKXDzMNRDUR2yita6s/8NDuZjv+CqstQ/AbtCy54F70SZMQXQzOg9FkhuapEoiN94/9+gWE8gw5/tOBEYTiKaxodmPnyiZ84Mq1aGtw4uR4CN/f3aetaNC5PHmafe50IX1Tr/4kSF+FhosG66eJXhdDRYHMk2HGK0oiU9ptT1Plt6dM8JbdWBE/58mJThAEsYjY9QlgudDMHngceOQr2dePh4CfvlFzqM+cAr53HfDcDyu7nfNB1Jt07TA9QWC3WnQBOqdd9aLPv6yRTsvoV8oYRZcExKsDcoXop/fol5efC7/bBofSOn/6JDvpTz50ohpQiE4QBFFjvCaDRQ8OaaFk/0QY6bSc9fl8unrHEjv4b/EpTfQCdC6D0xGEhO9vIcE7kQMzJ3o1EZvo1fahcwTn5mppGH534SF6s9L6mQrr/x9G4in1ChSbMqS0nIOfzIgmUpiJJAAwT+bpqQi+fO/hnM/51XPsQOd1O7ohSRIaXHZ87DrWgPrG/Ue1gzJDE333KXb/5esM/nyxjd7/OJAq84kDWa55E128UqhumlK5mujJuH55Aelc+Pe6qCb6HIXoBEEQiwarHXj9bYBDCJAf/iIw8GTmuukU8Kt3A8N79fen4kwFc9ffAsk6K98kIvpBm93FNebV35NFaM+ycWY2iniS+ciXBYosdOlC9FNsf80MUV3TvAbwNEOSJPV39jNKiL6aQnSiClCIThAEUWM8JiH6ASFEjyXTOd3IPGRbak30Vl/hOhfuQ1/eyIbNzMWSiCZSuZ5C5CIjfAuar1cJ4iFgZkBbbqtRiN6qD9EDriJ0Llma6PyEmMtuQU8z80qemams0oW/vttuVdV9P3iiD8/2TWZdf7cyHPR15yxX73/tjuXY0dOIUDyFL/zhJXanIUQfi9nQ5LFjS5fh/0/vpdrt+Fym836+hCeBdEJb9rVnX7dCtAs6l9X1onPJcKILobnYQgcWlM6FK1n456kQeOBOITpBEMQioXkV8Kr/0JblNAvLZwb16937SeDwH7Rli6EUsed24PvXA7PDFdvUohneB8jCcUyR2plyNtH5FdMrmjzF+8jFKxET4cx9D4AF62LrXnC/tyq/s4PKFZR1MbSdWPRQiE4QBFFjPIrORVQo8BCdD0zvy+FF5ztA7UssRG/xsh2niQIu2ec+9PN6m9RL/0jpMg/sHkASLhutps5l/Kh+ubXKQ0U5LWvVmz4pimWWwoerZmui80ZQe4NLPSlWTBBYCvwEXWfAhcvWteHmnd2QZeAffvWC6Ymm3+4dRFoGdq5s0rkvLRYJn71xCyQJuHPvEP58cCRD5xKBAxevbYVFadmrdJ2jXx7eV54vjiO20AGgoQZNdFHnUi8Hee4m/edYvJQ6YjiJ4ll4TfSidC5z5EQnCIJYdGx7A7D9L7TlmVPA184CfvhqYO9Pgd1fB576lva4zQ28817g6s8AkhCVDe4BfvWuqm12XgaNepOd5utlQf09WYYmet84U7msLGaoKMc47HXGZLjodD8QHteWha/V+DubmuhENaAQnSAIosYYm+gTczGcmY1CkoDzellwwaeeG5FlWW3QdSy1waJKE308FIec7fI/hUNKE31Dp19tsJPSZR5IEuAULpGtps5l/Ih+uQ6a6ACwPJ3D5WiAD3EyNtG1+QZOdCqXxFa6iS6+JwB88hWb0dbgxImxEL72Z/0JC1mWNZXLOYYDHwDbuhvxjot7AQB/9/O9mJzWGkURyQUZFlxuNrS0dR07cOWUPUQ/o1+ugc6FX+JskYDeYgdvVQqbgzX1OGOHtNvhCf26C6qJrg0Wzfe7AQBSaVk9GdtOTXSCIIjFxSu+pAyEV5DTwImHgDtvBe77J2FFiSlgus9lg0nf+mu9ymzgCSAlXNVWS8Rmtq8T8HcV9XSuAC1HE73koaKAbr4QAGDaJEQXv1ZAp65pa9CGorb6nAgUoVYkiFJZ8CF6IpXGtx8+juu+9gg2fOqP2PSP9+DV//UYfv7MgOn6oVgS//6nw9j1lYew4VN/xKX/9gD+7Z5DiMTpsn6CIGqDRx0syn4O8QnjvS1eVXuQbbjoVDiBeCoNQK8LWArwJno8mcZcnkGIfKjoxs4G9dK/8TK0L5Y0otLFOJCwkowJqg/JytyItaBFH6J3xE12/LPQrOhcpkL6gzF+Qqy9waW2hCrtROchfafSfA947Pj8a84CAHz74eP42589rzaVDgzN4sjIHBw2C165dZnp633iFZtw/qpmzMWSeGj/SfX+uTT7ei41+tABNgyr8yxtefiFeX9dOoLGJnoNBov6XfjodRvx+ddshcueOfyrZogzBcQTVGFDE30BOtFjybR6iXcuJkNxpGV2bpBfJUIQBEEsEpwNwJv+L/+Vi9f+C7DpVdryml3Ayz+vLctp86Z0tZFl4NRT2nL3Tu3S5QJpV692LEMTXTlG7S1LE92kkCKG6FYn0LFVXRSb6HUzb4ZY9CzoED2VlnHL95/BF/94CPFUGm86bwVes2M5hmei+Oiv9uPjv96vWz+WTOGdtz+DbzxwDN1Nbrzz0lVY3ujGtx46jrfc9iRiSQrSCYKoPl4nb6Kzg32uctm8zK82Fk9m0bnwBkGL1wGHbUH/SC8at8OqDmXN5UWPJ9M4Psa0Eus7G9Cq7HCRzmWeOAPa7XxN9OgscOgPmWFmKYi+7OZVrE1bC/xdgF3bYW+Jmp+8N6PJy5oyk9l0Ln6nGmqXoyWUC65z6RCGQV27pRMfvHItLBLw271DuPqrD+NnTw/gjj3s4OaaTR0IeMzbPnarBd98yznoCriQjGg6l5DswupWrzqXIINl24WNegFIp+f5lQkYm+je6jvRAeDWK9bgLy7oyb9iNRGv5Jg4prXsMproCydEd9mtaFBmFBQyXJSfJGrxOor3uRIEQRD1T/tG4K+fAt7xe+Cct+v3YQHgvHcDF/515vPEq7UAYKq/cttYKNMDwKzgde+5sOiX6PCXs4nOrpbuLWXei6sRcAgKFrOTFOJQ0WXbdfv94hyTNaRyIapE4VOw6pA79pzCY8fGcdXGdnzrreeqAdJsdCPe8K0n8H9PD+CG7ctw8Rp26fCPnxzAUycn8b6XrcbHr9+kvs5n7zqA7+/uw4+e6Me7L1tt+l4EQRCVwq0EwZFECum0jIM8RO/yqzsk2Zzop6ciAJbeUFFOi8+J0GQYE6FY1p23E+NzSKZlNLhs6Aq40OLlOhcK0eeFromeI0RPxoHvXQuMHgR8HcCtjwNeE6VHoYwJbdnWGqlcANb6aVnDAl8AgXBfwU/lV1FMGQeL8vkGghN9vgc40UQKDqsl00OuwF+/0/Az5MMv34CrN3fg47/ejwNDs/iYUEwQB4qa0epz4n/ethOnvqN9xsJw4TIzlQtHDNHjc8DkCaB1bfb1i0E8eeNqBOxL8+elKW0btdvpJPu+t23IdKK7m6q7XfOkrcGJYDSJ0WAUa9tzH1hzH3or+dAJgiAWLxYLG2Teeylw/ZeAo38C+h5lDfWd7zRvczeu1C9P10GI3v+4fnnlxUW/BL96eTTItGdSAU320dkoJsNxbOzU9v9lWRaa6CWE6JIEBFYAY8pAemOInozrr07s1rvf9SE6NdGJ6rCg6xZ3v8AmJP/9tRt0DUy/y473vYyF4fe/NKre/6Mn+uC0WfDBK/WXYH/k2g3wOKz46VOFt8gIgiDKhVfRucgyEE2mcGCIqTG2dPmxStkh6Z8MI53OdLvuH2TrblzWkPHYUqDZm99vfpj70DsaIEmSpnMhJ/r8cAoheq4mev9uFqADbMDjvv8r/T1TCWDyuLbcVqOhohzBi+4Jnij4abyJPhWO6z7XvIne4XeqIfqZeYTok6E4rvjyQ3jz/z6ZdR2jzkVkW3cjfvv+S/CpV26CW1GQtHgduHy9iZLFwFnLAzhnmdZWD8FprnLhiCE6AAzvzfseANgPzniY/Z0NcbBoDYaK1jXGzxDXJYk6F2cAsC4szyh3mxcyNI2v00Y+dIIgiKWB3QVsvhF4xZeB89/DtHJmNCwDrMIVj1N9Vdm8nPTv1m7bvUDn9uzrZoH/vosn05iNZNeexZNp/HH/MN55+zO48Av347qvPYq79g2pj4/MxhBNpGG1SFjelOVKw3yIShejzmVkP5ASfo8vP1f3MDXRiVqwoJvorz+nG9u7G00/MDxUDykuxNFgFH0TYZy/qhlep/7L9jhs2NHTiN3HJjA8E8GyQIk/AAiCIErALfhxJ+biOKG0zrd0BdDkscNulRBPpjE8G81QIbxwehoAsL27sVqbW1fwIaHGAY0iPERf39mgPId0LmWh0CZ636P65YO/BS7+QGnvOXmStWU5tWyiA5Bb1oJ3d+zB00AyBtjyB3FNihM9LQMzkYQ6aHRE10Rnr8OHIxbSEjLy6+dO48xsFGdmo4jEU+pVLyLcuS7qXERsVgvefdlqXHdWJ37weB8uX98Ge4HKiw6npsmLwIULV+dQgrRtAix2IK3oRIb3AVtvyv0GyRjwy3cAh/8ArL8euPkHmd//dFrv0fdV34de1xgdsWYhumdhtdABoE1p2VGIThAEQZSMxQI09jDdGVAfOhexid5zAWAtPtJz2a0IuO2YiSQwEoxmKPpmwgl8/f6j+M3zpzEV1s/v+eRv9mNnbxOWBdxqC727yV3wvmEGuUL0jKGihia6T9t3JSc6US0WdBP9NTuWZ7TQOX/cz/yXG5XQ5Ngo83KubDYfeNDTzD50x0fNlQkEQRCVwmKR1CD9uYEpyDI7mG9rcMJmtWBFE/u5ZVS6yLKMF06zJvq2boPbb4nAtRgTOQLxw8JQUUAL3ilEnyeFNtFPGkL008+YDw4qBN5o59S4iZ5o0nQjkpxmKowCsFstqrNZ9KKLTnR+qW08lc44gCkEWZZVhzkADEyGM9ZJp+WsOhcj3U0efPKVm3FZrja5kbjmRF+9vAMNrhxtZpsD6NisLQ/vy//6j/47C9AB4MgfgQc+n7nOc7cDowe0ZVFfQgAOLxAQPO185oCoc/G0VHebygA10QmCIIiy0NSr3a51Ez14Rn9FZgkqF45Y1jDy3cdO4Hu7T2IqnECH34lbr1iDP33ocmzvDmA2msTf/3If0mkZ/UqIvrIUlQuncYV2e24ESAhXYIohuqc1Q6/TGXCh1edEV8CF7qYSBpsSRAks6BA9Gw8eGsXv9w+jwWXDq89m3k5+mUpjlkFYfjc7mJ2NFn+gShAEMV88SkP0mT4WXGzp0gJK1Ys+oQ/RT09FMBmKw2aRsGmZH0uRFl8BOpcRpYnewUL0Nh/pXMpCIU30WBAY3JN5/8HflfaeQ89pty02oH1z9nWrQNDbq79j/GjBz+UqIu5FjyZSmImwfZCOBhccNovq7y/Fi/7i4CwOKSeQAKgHOiIToTiSaRmSVKEAMa69Z3d7AeF75zbt9vC+3IqWMy+yEF3k8W8AfcJlzrNDwH2f1pbtHuDCW/Nvx1JDHC46doj9LTbR3QtnqCiH/38eLSREV06otpETnSAIgjAiBre1dqJn+NAvKfmleFnDbB+T60Lfv2sNdn/0Snz0uo1Y39GA/3jj2XDbrdh9bALf230SfcpQ0VUt8wiwAyv0y+LQVHGoaPfODG+9w2bBnz98Oe750OWwZpn9QxDlpu50Llf++0M4MZa/Dd73xVea3v/UiQm8/6fsIPtfXrtVvUQ6HGchullrHQCcyuUnsWTK9HGCIIhK4nFaMRECnu2bAmAI0VvMh4vyFvrGZQ1w2bO4/BY5LUroMZFF5zIXS6rDV9UmegPpXMqC2ERPRpiv3OhNHngSkE1+rx68E7jor4t/z0EhRG/fBNhrq1+bcvdA19GdKDxEb/I40D8RVlVEvA3rsFnUE/sdfhcmQnGcmY0WfaLsF8/qhzP1T2Q20fmBU6vPWfpluLkQQnQ4CmgpLdsOPP8jdjs6zQZMNfZkrpdKAr99v17tAwCQgTv/Cvir3YCzAfj93+uvktj1SaB5VbFfxeKnbQNw7D52e/wokE4B4Qntcc/CC9GLa6KzzwE10QmCIIgMxCZ6eIIVRJw1mkUlhuhWJ9B1Tskv1e7PfrKZlzB2bWiHTdg/XN3mwydfuQmfuvNFfOnew1iraJXn1UQXdS4Au1q1ZQ07mS9e4WlQuXAaPQ7T+wmiUtRdiN6TRbdSCH86cAYf/NnziCbS+Pj1G3Hj9i71MaeNBUyJlHmrKZZKA2B+dIIgiGrjsbOfPbw1vXmZpmdZ1cp+Lp4c14dg3Ie+bYn60AFNzZJN53JE+X52+J3qThZv906HE0ik0pUJD5cCLkOoG50FvAbtw8lHzJ976inWEvZ3mT9uRjqtV3zM48ChXMykHBiWm7FMUlq748cKfq7aRFd0LqNKkNfhd6r+8w6/EweHgdEim+jRRAq/3cuaPNtXNGLfqWn0T2YWFPhQUX5Jb9kpOkQ/W788vM88RH/iG/rBow6fpo6ZHgDu/Tiw9hrg8O+1dbrOoRZ6NsQmejLKvocLXOeiNdHzf3ZI50IQBEFkpUmvEMFUP9B5Vm22RQzRu3eyAaklkq2JPh2OY1jZP9zQmXmy4C0X9OCBQ6N44NAoDg6zokJvaxmb6DNKCcToQ19uHqITRLWpu8T49lvOL+l5tz16Av/6h5cAAJ979Ra87aJe3eNc4zIbMde1cN0Ld5QSBEFUE4+Tnejj9gKxic7P7ht1LvvUoaJL04cOiE508ya6OlS0Q9sJbPI4YLVISKVlTMzF0ZlloCKRB6chRI/NZIbo4lBRZ4Ctwzn4O+DCvyr8/SaO6VvFy2sfos9Gkoiml2GZVQkci2yiA8BkiO2X8AGf/KAGgPp/88xM5kmiPx04g0//7gA+fcNmXHfWMv1jB0cwG01ieaMbbz5vBQvRTZroZwr0oZeELOuc6HBkDoHPoGMLIFkAmRUbMLwP2HSDfp3xo8CDX9CWXQHg3Q8AP36ddpn18z8GDvxWW8diA278BmBZmlfs5MU4oHf0JSAyrS0vQJ0L/xwVpHPhswgoRCcIgiCMiE10gHnRaxGihyf1M17m4UMHtAKF8Yqtl4bZsdOKZrfpLBtJkvDF12/FdV97VL2acl5N9IZlgGTVrlx95Mtsv0MsS0Cqi/1+ggAWgRNdlmV85ncH8PnfvwS71YJvvuWcjAAdANa2s4M3s4NIABhQGlrr2mt0aQ5BEEsa7kQHAJ/TprsqZ5XiRB+YCCOVZil7Oi3jxUEWKC7lJjp3ok+EzIMS41BRgA1y5S1gUrrMA5fh5I3Rix6Z1jfHz30b21HmHLyzuPcTfehAfTTRIwmckIWvafxobo+3QLOXHZioTfTZzFa42hIyadP+4tnTGJ6J4sO/2KcOT+f8UlG5vP6c5epMBbPBoiPqe1YgRE9EAAjfi0Ka6A6PPtA1DhdNp5jGJSV8bq/9AtC6FnjttwEIPsy45oPHJX9Xu9bYQsA4oPfUU9D92y1gnct0OJFT1RhNpDAbZUWaNh+dUCUIgiAMGIZZ1syLPvCkfnmeIXq2JvpLSrt8Y2d2jWB7gwtffN1WAIDLbkF30zz0ilYb0LpOW57qA372ZmD317X7WtdnHncQRI1Y8CH6p393ALc/3ocmjx0/fc+FGW0sToffhd4WD/aemlb96JxwPInnB6bR2+KhSzkJgqgJokpq07IGWIThKF2NbjisFsRTaQzPML/3ifE5zMWScNktWNdeQMNzkcJD9MlQXD3BIMJD9A2GHcFWH3nR501GE90Qovc/rjWKAWDVy4BNN2rLA08Cs8OFv5/oQ7e5mBO9xsxGDSF6dFrvks4Bn9nCWzwjwcwmOg+3R2b0BziyLGPvKTY/IRxP4f0/eQ6ROAsKh6YjeOzYOADgpnNXYKUy7GlwKoJEKq17Ha5zqUgTPW7QxxQSogPMi84xhuhP/68S8CqsuQo4+y/Y7ZUXA5d8MPP1WtYBl3+ksPdeqribAF+HtjzwhP7xBRiiN3rssFvZ79FcQ6T57wBxFgFBEARBqLgbAVejtjzVV5vt6BcGp0tWoLs0gwOnI4sT/dAZtj+fbxbPy7d04ttvPQf/+7adqjq5ZF7xZcDTqr8vKez7ZvGhE0QtWNAh+p3PD+KHT/SjwWXDz993Ec5d2ZRz/ZvPW4FIIoWv3HtEd/+X7z2McDyFvzRpsBMEQVQDsYm+pUt/pt1qkbCimZ3h71O86PtOMS3GWV0B3cCXpUazosRIy8zhJyLLsuqY39Chv8qIu9RzhStEHsyc6CKiysViA3ouBLa8RlhBBl66q/D3E5vondsyh5jWgNlIAidkg9d9vDClC/+/OxXiTXQlRBea6J0BdtvYRD89FcH4XBx2q4RWnxOHR4L4p9++CAD41Z7TkGXgwtXN6GnxoKPBBYfNgmRaxtB0RPc6XOfSUQmlUVzfji8pRJ8bAYJn2O3QOPDA54XX8wE3fA2QhPb5rk8CHYbG+Y3/OS9n6JJB9KIPGq76WIA6F0mS0ObLP1xU9aH7tFkEBEEQBKFD9KJPlaGJLstsv2boeeDwPcDE8fzPEX3oXWcDzvmVqMQmuixcRcl1LptMfOhGrjtrGS5b1zav7QAArLoc+MAe4KK/YccMRpafO//3IIgysWArF6m0jC/dcwgAsHmZH79/YRi/R2ajbdOyBrWd/q5LV+GP+8/ge7tP4uDwDHb0NOG5/ik8dXIS5/c2460XmgyvIgiCqAJiE31zV+aZ/94WL46PhdA3EcKl61ppqKiCzWpBo8eO6XACE6E4WnxaADk2F8NkKA6LBKzr0O9otlETff7ka6KfFEL0rnMAZwOw4gLWeJ0bYfcf/C1wwXvzv1cqAZzZry3XiRdxNprEcdlwBdzEUWDlRXmfqzbRDYNFxSY6v210oj9/ahoA2//56PUb8dbbnsIv95zGeauaccdzpwEAbziXDWqyWCSsbPbg6Ogc+ifCOm/lSCWd6BlN9AIP9pZt0y8P7wMaOpkjU1S0XPXpzKGjNidw0/eAn9zEhmNe9U/zvtx5ydC6QRsEnDbMD1qAg0UBNih0aCaaczAvDRUlCIIg8tLUq10dV2oTPTQB3PsJYHAPMHMaSArFBskKvOqrwLnvMH9uLKi/Oq8M+za8tBFLpjEbTSLgtiOZSuOIUkDK10QvO+5G4Np/Yd+Dez4OHLuP3W91Auuuqe62EEQOFmyI3j8RwpByGfJTJyfx1MlJ0/Vec3aXGqI7bVb89D0X4Ot/Poo/7B/GcwPTWBZw4f271uDWK9bO/zIUgiCIEvEKTfTNJjst3GvcN86CqX2nWRN9+wryw7V4HZgOJzA+F9MNED1yhjVhe1u8cNn1P99blcBkvIChc0QWcjXRQxPAiBB6r7qM/W2xMqXLM//Llvt3A8ERoEFQSZgxelB/WWcd+NAB1kQfkluRlBywycpVDYU20b1ZmugNYhOdhdsToRgSqTTsylUnzw8wlcuOniZcvKYVH7p6Pf79viP4+K/3I5WW4XPacP3WTvV1VrYoIbrBi67qXCrSRC9R59K5Vb88/ALQshZ45jbtvraNwM53mj+/bQPwN88CyVjm/1EiO2IT3cgC1LkAQFuDC8BMzuGiY3MUohMEQRB5EL3o0wOsSV7M1UuJKPB/bwROP2P+uJwC7vpbIDLF5rgYX/vU09rgTQBYeUnh750Fl90Kv8uG2WgSo7NRBNx29E2EEEum4XFYdfO5qkrrOuCtdwDHH2B/1l+XWZogiBqyYEP01W0+9H3xlUU/r8Flx6detRmfetXmCmwVQRBEaXCdi90q6YJgjhqiT4QQT6ZxcJiGinJafE4cHwthwqBm4U4/s+9nCw0WnT8OHyBZNO+52ETvf0y/bu9l2u0tr9FCdMjAobuA896d+72Meok6aaLPRBJIw4KgtwdNc8fYnYVckgstRJ8I6Zvo4pDPZo8DdquERErGWDCGrkamddqrNNHPXtEIAHj/rrV4um8Sjx5lLvRXbl2mu7qlp5kPJ9aC7UhcG6hYkcGipepcXAGgeTUweYItD+8FRg8AaWGezdWfZYOosmFzsj9E4eQK0RegzgXQgvGCdC4UohMEQRDZaOrVbicjwNxo/gIIR5bZUPRsAbrInz/DZutc88/6IF1UuUBiisQy0O53YTY6h9FgDOs6GlSVy4ZO/XyumrDmSvaHIOqMpSvSJQiCqCM8ThYIrWtvgMOW+aN5VQsP0cM4fCaIeDINv8uG3pYatQTqiFaffkAj52nlCqWt3ZltfW2waP060SdDcZ2jsO6QJKZo4YghuqhysdiZxoXTcxHgbdeWD9yZ/71EH7rTDzSvKXpzK8FslGkvIv7V2p0TxTnRg9EkQrEkpsLstcQmusUi6ZyVABBLpnBgkH2vd/Q0qut97Y1no9PvgiQBb75A39jhw0X7JrQmOvehu5UmUtkpVecC6L3oJx4CDvxGW155KbD+2nltGmFC20bz++3eBeuU55+lnE10wYlOEARBEKaITnSgOKXLw18CXrxDW/a0sLb5K74CvPlnwGV/r1//8W8Av/0bICWUB8QQvWMLGwheBvhwUb6P+ZJS0trYSVfyEUQ2FmwTnSAIYjGxto0FTJeuazV9nIdgAxNh7D3FVA7buhtpEBqAFi/bAZwQWuWJVBqPH58AAFxm8j1VdS512kTfd2oar/nmbrz9ol585sYttd6c7DgDQJSphXQ6F3Go6IrzAYdwssdiBTbdADz7Xbbcv5s9N5d6Y/B57XbX2YClPjoAsxF2gJP0rwCGlDuDIwU91++2wyKxobjcP+lQHP8i7X4nBqcj6gHOwaFZxFNpNHsdukttW3xO/O4Dl2B4OortSkOd0yP8/OCoPvSAqzI/RxJ6dUzBTXSAheg8ODc22q/5XHGXUBOF4W0DXI1AdFp//wJVuQCa75Wa6ARBEMS8aOzVL0/3Az0XmK6q48VfAw/9q7ZsdQBv/Il+ds6G69nslz98BIBSntn7Y9ZcX3UZ0H0+MPistn4ZZ73wogY/2XzoDNsf3bws/1BRgliq1MdRKEEQxBLn6s0deOjvr8A/XGt+SX1XoxsOqwXxVBr3HDgDANhm0rBeirQoTfRxoYn+/MA05mJJNHsdOKvLrInOdS712UQ/MDQLWQb2KQNk6xYx+OZN9OAIMHZIu19UuXDEJrGcZq7HbMTDzInOqRMfOqA10a1eIWiMzQDpVJZnaFgtEhqVNjo/aGlrcGYE2nzo54jiTBdVLsZ12xtcGQE6AKxUwvaBybB6dQMP0XkLqeyUqnMB9E10kS2vA7rPLX2biOxIknkbvUxtt1rA2+VjwRyDRcmJThAEQeSjcQUAYZ+rkCb66T3Anbfq77vhP82Hz5//HuD1twEWoeM6fpjNg/nNe4GUcLxSzhA9WxO92kNFCWIBQSE6QRBEndDb6oXNav5j2WqR1DbpE0rDmnzojBZfZhP9kSNjAIBL17aaOv14uDIZiiGVrj9lSlAJZ6cVxUfd4hR2snkTXWyhA9pQUZEVF0B3MDLweOY6nDP79cOU6sSHDrDBogBg9xraurydn4cmpXV+WAnRzQJt7ivn+pXnB6YBADtMwvJsdDd5YJGASCKlNm/VoaKV8KEDpQ8WBYBOkxDdYgeu+qf5bRORm7b1mfd5Wqq/HWWiXfm/TU10giAIYl7YnIC/S1ue6s+9fmQa+NmbgaRwEvfSDwNnvzn7c7bexPQuNnfu1+6pTBN9OhzHsLJvuKGTmugEkQ0K0QmCIBYI3H/OM9/tK6iJDgCtfECj0Cp/5CgL0S9f32b6nGavA5Ki0pgK118bfS7GNCFGz3vdITrRB58FvncdcO8ntftsLqD7vMznuRuZ05HT/0T29xgyDBWtkya6LMvqYE5HgyFojEwV9Bp8uChv/vCDGZEOv96JrjbRFR96IThsFnUoaf8k06zwUL4jUIUQ3eoErPbs6xrxtgD+bv19578HaF5Vnm0jzDFroi9gnYs6WHQuZjpfQpZl9RJ2cqITBEEQORGHi+Zrou//JTAn6P023QBc+Y/532PdNcBfPw5c+iFg5SVsP1qk56LCB5oWAC9vjM5G1asiu5vc8LuK2GcjiCUGOdEJgiAWCL0tWpOzrcFZuQbpAkNtoiuB82Qojv2DrAl8eRbHvM1qQZPHgclQHONzMXXQaL0QVMLZmUgCyVQ66xUKNUfUuURngAFDGL7iAtbeMaPnImDkRXZ7cA+QjJmvOyiE6N42INCduU4NCMVT6lUMnoDh/1mBIXqTonM5rDjR202b6NqltuNzMQxMhiFJMNW25GJliwenpyLoGw/hvN5mzYlejSZ6MS10TtfZwOxpdtsZAC7/SFk2i8hBq4lOzL1wQ3Su7UqkZEyFE+pJK85sNIl4Mg2AmugEQRBEHhpXsjk+AHOi56LvMe22pwV47XcKn+fTvBq4+jPsdjLOrsg89RRrtZ/ztqI3OxdiE50XOjaRyoUgclKnR+UEQRCEkd5WLYja3h2goaIKqhNd0bk8enQMsgxs7GxQL+c3fZ4SqIwH66/tzV3bADAdqWOlS8va3I9vvSn7Y6ITMhXTh+UiYhO965y6GSrJVS4OqyVT5xKZLug1eKjHtT0dJv9fRSf6XkXlsqbNV3RLqKeZ/fwY4E30iutcBCe6w1f88y/+IAvPJQvwqq8u6Eb0gqHNJERfwDoXp82qDuo1U7rw+xpcNrjs1qpuG0EQBLHAEJvoM6dZwG2GLOtD9N5LSysTAIDNwWbBXPTXwGUfBrzm5aBSEYsaaohOKheCyAk10QmCIBYIq4QQnXzoGjwMD0aTiCVTeOTIOADgZVlULpxWnxNHR+fU8L2emFOa6AAwHY5nbco/fGQMAxMh/OVFvVXaMgMX/BUwcQwYOcgOEJwNgNMHOBrY4KPtOdyPRqfjwOOZw5Yi0+z1OfXkQ1dOdPjdNkgew/DFQpvohmasWRuWnwgamYmqKpdifOiclYoOqn+Cheh8UGlVdC6lHDz2XAB8+AAAif2fIipPoBuwe4GE8G+3wE9etDc4MR1OYDQYzXC8kg+dIAiCKJimlcKCDMycAlrWZK43dhgIj2vLvSazgeoE3kSPJtJ4po/tu1ITnSByQyE6QRDEAoGHYACwrZt86By/yw6bRUIyLWNiLo5H8/jQOa1KcFKPIXpQCNEnQ9mb6P/vF3sxPhfH5evbsLKlxJbLfHA3Aq+/rbTn+pcBTauAqZNsuf9x4LL/p19neK9+uU586AAwo7TH/S474GrUPxidLug1mj36EN2sic5bQsFYEruPs4OyHT1NGevlg89U6J8MI52W61/nAuid+0TlkSQ2XHToee2+BdxEB1hAfmRkzryJPkc+dIIgCKJAxCY6wLzoZiF636P65d5LK7VF88btsKLBZUMwmsTJcbbftpFCdILICelcCIIgFghdATe6Ai40uGzYsaL4EG2xYrFIqhZj97FxjAZjcNut2Nmb+3vUqmpg6k/nEoxpwXm24aKxZErd9lGTgGhBsFJoo596Gkin9I8bFS911URnJzoa3HbA6WfaEU6JTfR2k0Zsg8sOr4OpJp5XdC5nl9BE5zqX/okQJkJxJNMyy0wr1cLV6VxqcIKHKA3jcFH3wv5dI/pejVATnSAIgiiYxpX65WxedKMP3Wxodx0h7nu67Vb0NHtyrE0QBDXRCYIgFggWi4Rf3noxEsk0Ah6ami7S4nNiNBjDnXsHAQAXrm6G05bbccsVKfXYRBd1LlNh8xBdDNe5U7vSzEQSuO/gCCQAXqcVbocNXocVq1q96oDXoui5CNj7E3Y7NssGjS7brj0u+tADPWV3Qc4H7kQPuO1sWJSrEYhMsgcLDNGbvfrPsVkTnd9/QmkIeRxWrO8oXm/SozTRp8MJHFUGmbb6nLBXamitrolOOpYFQ+t6/fIi0LkAwOgshegEQRDEPPB1ADYXG/AJsCa6EVnWho8CrIVeJ7N8stHhd+H4GNtn29DZAKulvreXIGoNhegEQRALiOWN7lpvQl3CW+WPH58AkF/lAmiX8NdjiB4sIESfmBND9Oq06b9x/1Hc9tjJjPvddit2f+xK9YqAgllp8KL3P6GF6LKsb6Iv31Hk1lYW1YnuUnal3I1CiD5d0Gs0e7Xwzm6V0JTl5JgYom9dHoCthODb57Sh1efA+FwcT/ex7ayYygUoj86FqD4rztdu21yZzbsFBg/Ix0x+zlOIThAEQRSMxQI09gDjR9jylEkTffwIEBrTluvYh84Rm+iblpFGjyDyQToXgiAIYsHDh4vKMlsuJERvUXUudR6iZ9G5iNs9E6lOE31wOgIAWNvuw86VTdi8zA+HzYJIIoUjSru5KJpXA952bXngce32S78DZge15eU7S9zqyjAbYf9GfrcSfIvai0Kb6IITvb3BBSlLW4l70YHSfOgcfonu0ycnldelEJ0wsPIS4JK/ZZefX/9vC76J3qY20aMZj5ETnSAIgigK0Ytu1kRfQD50jrgvSENFCSI/1EQnCIIgFjyiSmR5oxurW/OHdqrOJVhfTvRYMoV4Kq0uZxssKjbRs7XVy81cjAXH79+1Bq/d0Q0AuPk7T+Dpk5OledklibXRD97JlvufYGdC5DTwwOe19axO4KzXz3Pryws/ceF3KSG6OFy0YCe61jzP1YbtCGgHODt6GrOul4/eFi+eG5jGcwNs+zoDFQwPyYm+MJEk4JrPsT+LAGqiEwRBEGVDvDrLzIm+wHzogP534MZOCtEJAIf+AOz+GjD6Epv5tPJiYNcngc6zCn+NqX7gwX9lJ5bCk0DLWuD8dwPnviNz3T23A3f9bfbX+sgJwFs/g+4pRCcIgiAWPLxVDgCXr2/N2ugVaVV2GidCMciyXNBzqoHYQgdy6FxCWihULSc63zafUwt/23M0PQtCDNFDo8DEceD009rlsgBw/nuAwPLSXl9BlmUcGJrFmjYf3I7cvvxCUHUubq5zERri0emCXsPntMFulZBIyaZDRTkdDUKIXsJQUQ73okcT7CQN6VyIxQ4fLDpGTnSCIAhivjQJIXpkCojOAK4AW5ZlfYi+AHzoANAu7AtuJJ0L8cx3gd9/mM2iOudt7Jhm/6+A4w8C7/g90H1u/teY6ge+ew0Lz896PeBrBw7dzYLy8aPAtf+iX//Mfvb3BbdqnycRe33pbClEJwiCIBY8rYJb+vJ1+VUugKaASaRkzEaSdTOsteAQXXSiV0nnwpvoPqe2+6CGVKU00QE2XFTk5EPA7q9ryw4fcOmHSnttgTv3DuJDP9+HtgYn/mbXWrzp/BV5h8/mQjdYFChJ5yJJEpo8DowGYznVKp1KE315o1t3sFMsK5UQnVMxnUsyDqSE/7cUohM1ggfkwVgSkXhKPYGWSsuYDFGIThAEQRSBqHMBWFi4bBu7vQB96ACwookFlCtbPNrVlcTSJDgC3PsJoHkN8N4HtUD73FuA718P3PVB4K8ey39y6J6PA3MjwF/8Elj/cnbfrk8AP7gBeOK/ga03AV3CrKsz+9kcnpd/HrDWf0RNTnSCIAhiwcOb6FaLhIvXthb0HJfdigZlKKTZpf61IhjVB+LZneja/TNVaqKHzEJ0xdddks4FADq2AE7h8tEHPg9MD2jLF70f8Bb2b5qLoyNMLzIWjOHTvzuAXV9+CD97egAJQZ1TDNpgUR6iN2oPRqY1QX8e+DDWXE30S9e14ooNbfjbq9aVsqkqPc36MLszUKEQPRHSL1OITtQIv8sGp40d7ogn+iZCMaRlwCIBLV4K0QmCIIgCMA7bFpUuC9CHDgBnr2jEp2/YjK/evL3Wm0LUmud/CCSjbDaO2Ajv3glsfQMw8iJw6uncrzFzGjj8B2DFhVqADrA2+dWfASADz35fu1+WgZEDTH20AAJ0gJroBEEQxCJga3cADU4brtzUrjWDC6DN50QwmsT4XAxr230V3MLCmVOa6FzzMVnAYNHpSJWc6Fzn4hKb6DxEL1HnYrECKy4Ajt3HlsUWt7uJhehlgDf6z1/VjP6JEIZmovjYr/fjq/cdwbKACz6XDV6HDT6XDRetbsEbdq7I+Xo5B4umYkAiAjg8Js/Uw08AdeQItP0uO26/5fy8r5UPYxO9YjqXuDFEr4/PFrH0kCQJbQ1OnJ6KYDQYVZVGPFBv9jphtdT/5fYEQRBEHdBkCNHF4aJ9u7XbC8SHDrDfk7dcsqrWm0HUA/1PsL9XvyzzsVUvA/b+BDj5CNBzQfbXGHgSgGz+GisuZI3zk49o902eYHOUivGt1xgK0QmCIIgFT3uDC3s//fL8Kxpo9TlxYjykC6RrzawSVC9vdKNvIozZaBLJVBo2q/7isWo70dNpGXPx7DqXURPncMGsvEgL0UUu/ZC5G68EuP7mhu1deMO53fjJUwP45oPHMBqMZbTof/P8IK7e1IEmr8PspQCIg0VNnOgAOxlQQIj+/ivWotPvxrVbOov4akqjxeuA12FFKJ4CkDu4nxcZITo10YnasSzgwumpCJ4bmMLO3mYAWoie6woQgiAIgtDhCrD9PV74OHofcP57AatD70NfecmC8KEThI7xI4DFBgRMikTNq7R18r0GADSZnJix2oBANwvOk3HA5gDOvMAekyzAL28BBp5gV/S2bwQufD+w7Q0lfzmVgnQuBEEQxKLAapGKbhTyFvB4qSqSCsB1LiuaPer+t5nzfKLKOpdwIqUaShpcZdS5AOxgw4ivEzjvPaW/pgHeRG/2OOCyW/GuS1fh0Y/uws/eeyFue9tOfO2NZ+PzrzkLTR47ZBnonwznfD2uc1GvfHA16lcocLjoxWtb8e83by/qCopSkSQJK1tYoO1xWNHgrFCXIj6nX6YQnaghrzunGwDwjQeOqSdMaagoQRAEURIdQmP25MPAj18PnH4WCI1q9y8QHzpB6IjOMMWmxWRmFC81RWfyvwaQWS4SX0dOA7EgWx5WQvTnfshmCmy7Gdj0KjaA9NfvBv70j8V/HRWGmugEQRDEkqXVxwIU0S9ea/jwzoDbjoDbjulwAlOhuLqtACDLsi5ED8aSSKTSsFsrd26c+9CtFkl1DANak3MmkkA0kYLLXsKwzq4dgNXJNCicl32koCZ3oUwoWpxmoV3ucdhw4eoW3Xq/2zuEp/sm0T8RwtkrGk1fK52W1X8nU50LUPBw0WqzssWDg8Oz6PS7IFWqJUU6F6KOuHnnCvzkqX68ODiLL91zCF+6abs6B4NCdIIgCKIodn0S+OGrtX3WvkeBH96oX2eB+NCJJcA3dgITR/Ov95kZVoLxdZg/blWOn5J59J28SGPLcjWv1Wl4HRkI9ABXfAzY8RZtvak+4LvXAo//J7D+OqDXpHBVI6iJThAEQSxZtBC9nproLJxtcNnQ5GE7IEYvejCWRNwwEHPGpK1eie3yOW268DXgtsNhMrivKGxOYIXg/G5cCex4W8nbasakSYhuxopmFtyfytFED8aSma18cbAowC5FrEO4E7qjUj50gHQuRF1htUj47I2sOfiLZ0/j+YEpaqITBEEQpbHyIuAvf6PXDSaEfcYF5EMnlgBNvUDLuvx/AOYrT2UplvH78xVjbMrxRSrLcSk/+cSPDa7+DPCh/foAnW/3ro+z2y/8PPd7VhlqohMEQRBLltYGRedSR010rnNpcNnR5LHjJDQVCYe30L0OK6wWCbPRJKbDCV1bvdzw5rXPoACRJAltPicGpyMYDcbUELpodn0C+NlbgHQKeM03szcYSiCZSqsnGfKF6Hz4Zv9E9hB9Vnktl90Cp01p3i+QJvo5PWw7z1rur9ybUIhO1BnnrmzC685Zjl8/N4jP/O4AupvY57ytgj8zCYIgiEVK7yXAO+9lKpfZQf1jKy8BLNRVJeqEt95R+LruJqZUkeVMpz/XtLjyHD/w46Fs2pfoDACJaWPysfxc9rc4wLcOoBCdIAiCWLLUYxOdh9UNTpsa+E4ZnOcTyva2KNs/G01iJlLZEwFzQkPeSLufhehjwTyX+OVi5cXA/zvMBtqU+eBjOpJQm+NNntzu8R7lJMBAjia6NlRUeC2jE71OQ/SXb+7Anz/8MvS2lE+Vk0GGE510LkTt+dj1G/GnAyPYd3oGLw0zFyc10QmCIIiSaN8EvOs+4Cc3AaMHtfvJh04sVFrXs5NCM6eBRsNw0cmT7O+2DXleY4N+fZFUkr126zp2rJdOA2f2AbE5YJXJ5yauHIvZ3cV9HRWGTpERBEEQS5Z6DNFnuTYlh86FN+dbfA40KqHwdIWHi/Jw32syjJJ70ec1XBRg7fMKtHemlO9fwG2HLY83nutOcoXoGUNFAcDuAmzCTl6Bg0WrjSRJWNvuy/t9mBfURCfqkPYGF/7uanbJMtdhUYhOEARBlExgOXDLH4G117DllrXA2X9R220iiFLh3vGTj2Q+duIh9nfPRblfY+XFACTz1xh4grnQey5ky5IE/ODVwA9uAIIjmev372Z/80Z6nUAhOkEQBLFkaavDEF1zotvRxJvohhB9IsS2t9XnRKMStFcrRDfqXAAWTgHA6Gz9fB9F+FDRljwqF0Brop+ZjSKaSJmuMxsxDBXliEqXOm2iVwUxRJcsmh+RIGrM2y/uxdp27coICtEJgiCIeeFuZMqMDx0E3v8M4KSr74gFyrY3ssGfj3wZCE9q95/eA7x4B9C5VQvAsxFYDqy5Euh/DHjpbu3+RAS4/7Ps9nnvZn9LErDlNQBk4L5/ZM10zpn9wGP/wa5m3fGX5fjqygbpXAiCIIglS6OXhaDRRBrRRAouu7XGWwTMqU50rYmeqXNhoXCrz4FQjAW90xUeLMq3y2emc1Gb6PPQuVQQ3uRvKiBEb/E64HVYEYqncHoqogvcOLyJ7jd+L9yNQHCI3V7SIbqgc3H4Mr2KBFEj7FYLPnPDFrz1u0/BImk/uwiCIAhiXgSW13oLCGJ+NPYAV36KBdrfuhjY8jogNgvsvwOw2IEbvq5ff/gF4NDv2fPEwaDXfwn47tXAL94GbHkt4O9i600eBy7+ILBsu7buVf/EGucv/JxpkVa9DJgdYuvLKeCm7wP+ZdX5+guEQnSCIAhiyeJz2GCRgLTMhkXWQ4iuNtGdNjQrIX/mYFHFie51wm5lge50uMJOdMHVbqTdXyadS4XgIXq+oaIA052saPbg0JkgTk2GzUN07kTP2USfLnl7FzxiE51ULkSdcem6Vnzp9dtgtUhocOWekUAQBEEQBLFkuOSDLPR+4r+AZ78LOBuANbuAXZ9gTXSRM/uBh78IrLxUH6K3rgXefT/wwD8Dx+8HkjGgZQ1w4zcyW+XeVrbuo18BXroLeOo77D3XXQNc9v+A5edU/msuEgrRCYIgiCWLRQlRZiIJzEQSaPfXXjuh07kU4ES3KCXfyutcWOPd3Ile3zqXySJ0LgCwsoWF6P0TIdPHZ80GiwKkc+FQiE7UOTeftyL/SgRBEARBEEuNrTexP/nY8RZ9eC7SsgZ4w+2FvZ+7EXj559mfBQCF6ARBEMSSJuBmITpXdNQatfHtsiGa5DoXY4iuNNF9TqRldl/FdS4xRediEqK3lWuwaIUoRucCaF70gcmI6ePTahPd8L1wNWq363SwaFXQ6VwoRCcIgiAIgiAIYuFDITpBEASxpAkoSo6ZCofQhZBKy9oAT5cNSSUhzxwsqjjRvQ4kkmwIS8V1LlEt3DfCdS4ToRiSqTRs1vqaW15sE72nhQW/A5PmTfSjIywkXtlsCIjdjdptaqIzHDRgiyAIgiAIgiCIhU99HeUSBEEQRJXhbeLZSLLGW6K10AEWVnOH92w0iURKm1g+ITTRGz3VOQmghvsmTfQWrxMWCZBlLeCvJ3iTn+tx8qE10cMZj8myjBeHZgAAZy0P6B8UdS7RGSCdKmFrFwGkcyEIgiAIgiAIYpFBITpBEASxpKmnJjoPqh02C5w2KwJuOySD8zyZSmNKud3ic6BRCYYr70Rn22bmRLdaJLT6FKVLHXrRJxSHfLOvQCe6EKLLsqx7rH8ijGA0CYfNgnUdhpa12EQHWJC+FKEQnSAIgiAIgiCIRQaF6ARBEMSShg+HrIcQPah42RuUoNpqkdSQn7epJ5W/LRJrVvMmesV1LoJmxgyudBkNRiu6HaVQrM6lq9ENiwREE2mMGTzv+wdZML5pmR92o7ZGbKIDS1fpQk50giAIgiAIgiAWGRSiEwRBEEsaHlLP1kWInukdb1aa5tyLrraqvQ5YLRIa+fZHk0gKypdyozrRTZroANDe4AJQf8NFZVlWTzwUqnNx2CzoanQDAPoNShceom9d7s98ojhYFFj8w0VTSWB4X6a2hpzoBEEQBEEQBEEsMihEJwiCIJY0/nrSuaghul29r0lpT/MmOg/RW7ys+c1PAgAsSK/YtuVrojfUp84lFE8hrgxfbSlQ5wIIXvQJQ4h+mofogYznLKkmejoF/M8VwHcuB267CkgLJ3BI50IQBEEQBEEQxCKDQnSCIAhiSeNXm9y1D9H5NojDO3l7ejLEHpsI8aGi7H6b1aK2wyupdFGd6I48IXqd6Vx4g99ps8Bttxb8vJUtLEQXm+jiUNGtyxszn5QRok8Xta0LioEngJH97PbQ88CpJ9ntdApICCceKEQnCIIgCIIgCGIRQCE6QRAEsaSpp8GiZjqXJo/eiT7Om+jKIE8ACHAveoW+hkQqjWginbFtIm3++tS5TAg+dIlPaS2AFUoT/ZQQouccKgpkDhZdzE304Bn98vhR9ndC39wnnQtBEARBEARBEIsBCtEJgiCIJY1fCYVnIpVToRQKb3uLOpdmr96JPj6nNNGFIZm8rT4TrkyIHopp3xtvVic6b6LXV4g+qTT3m4tQuQDAymbWoO6f0NQkOYeKAoAzAEAI6hdzE31uVL88eZz9LapcAGqiEwRBEARBEASxKKAQnSAIgljS1NdgUbYNuia6EpZPqk50Fgq3CqFwo9pEr4zOhTfkXXaLeXgMLUQfm60vnQvX4BQ6VJSjOtEnI+p9L+YaKgoAFgvgElzpi7mJHhrTL0+eYH9TiE4QBEEQBEEQxCKEQnSCIAhiSVNfIXqmzqXZo2+iT5jpXJSvYbpSTfS4MlQ0SwsdANoVncvYXAyyLFdkO0qBN9HF5n4h9ChO9PG5mNrE3z+YY6goR/SiR6eLes8FhTFEn+Ah+pz+ftK5EARBEARBEASxCKAQnSAIgljS8MGiwVgSqXRtw985kxCdt8wnlYB8XAnTW4UQvVH1plcmROfblStEb1O2J5GSK7YdpaA20YsM0QNuu3py4tRUGLIsqyH6WYWG6EutiZ5OA3GjE52a6ARBEARBEARBLHwoRCcIgiCWNH7BP851KrViVg2rM53o0wadS4uoc3FzJ3qFdC5KE9uXZagoADhsFnUI6miwfpQupTbRAWCl0kbvnwjrhoqu72jI/iRxuOhidqIbQ/RkBAgOk86FIAiCIAiCIIhFCYXoBEEQxJLGYbPAbbcCAGZqrHTJ6UQ36FxavZlN9OkKbX8hTXQAaG9gSpfR2foZLsq/b8U20QFgheJFPzUZzj9UlLNUm+gAGy5KOheCIAiCIAiCIBYhFKITBEEQSx7Ni56s6XbMxbI70YPRJGYiCUQSKQD6JnrFneixAkN0Pwv2R4PVDdH3nZrG3S8MmT7GQ/SSmujNWhM971BRzlII0WUZmDMJ0SeOUxOdIAiCIAiCIIhFSe6jYYIgCIJYAgTcdpyZjdZBEz0zRPe77ZAkllseG2UtX5fdAo/Dqq7TpATtFWuiFxiitzXwEL16OhdZlvFXP96D4Zko1rT5sGmZPuTmIXqz0NwvlB4lRB+YDCORSgPIM1QUAFyN2u1yDRaNzgKH7gbaNgLLzynPa86HeIjpW4xMHgcCPfr7KEQnCIIgCIIgCGIRQE10giAIYsnjd7NweLbGTnRN56I50a0WCY1K0/y4EqK3eJ2QJEldh+tcKuZEj+Z3ogO10bmMzMYwPMNC+4NDsxmPayG6PeOxfPSoTvRQYUNFAX0TPRkFEiZhc7H85n3AnbcC/7sLGNo7/9ebL2YqFwCYPJmpc7F7Kr89BEEQBEEQBEEQFYZCdIIgCGLJw3UotWyiy7JsqnMBNJ/30dEgAKDVp1eT8BB9qkI6F75d3rxOdNb2HquizoVrVgDg6Kg+wE2k0uqw1vk00fsKHSoK6AeLAvMfLpqIAof/oC0f+PX8Xq8cZAvRjToXuwewWM3XJQiCIAiCIAiCWEAsOp3LsdE5vPI/H8XW5QHccevFGY+HYkl8++HjuPuFYQxNR9DW4MQN27vwwSvXwe2gAz2CIIiliL8OQvRYMo1ESgaQqU1p9jhwAiEcGVGa6D59IBxws1B9NppAKi3DapFQTrgTvaFgJ3p5dC6np8J443eexF9c0IP371prus6LQ1qIfkw5ycCZUpr5kqSdKCmGZQE37FZJ/XfZ1NmQe6gooG+iA8yL7l9W9HurzJzWL0+fKv21ykW2EH3qJBAT/g1I5UIQBEEQBEEQxCJhUTXRk6k0/t8v9iKWTJs+Hkum8M7bn8E3HjiG7iY33nnpKixvdONbDx3HW257ErFkqspbTBAEQdQDfhcfLFq7EJ2rZCQJ8Dr0YXWj4jznTnRjE50HxLKsKWHKSbDQwaJc51KmJvqe/ikMTkfw4yf7s67z4qCmcDlmaKJzlUuTx1HSiQWrRUJ3k6YjyatyAcxD9PkwYwjNpwfm93rlIFuInowC40e0ZQrRCYIgCIIgCIJYJCyqEP2/HzyOfadnsj7+4ycH8NTJSbzvZavxo3ddgI9etxE/f99FuOWSXjw3MI0fPZH9IJ0gCIJYvNSDzkX1jjttsBgCX+7zHpxmfm1jE91hs8CrXE01XQGly5zqRM/d5uY6l9HZGGRZnvf78u/J8ExU/dqNHBCa6AOTYUQT2gnxyTnuQ3dkPK9QuNIFALZ1FxCii4NFgfkPFzU20Y2hei2YyxKiA8CZF7TbDl/lt4UgCIIgCIIgCKIKLJoQ/cXBGfzXg0dx9aaOrOv86Ik+OG0WfPDKdbr7P3LtBngcVvz0qTpodxEEQRBVh+tcuD+7FvCg2kyZ0mQIgVtMQmHeVp+uwImAObWJnlt7xnUukURKfU453hdgrXQj43NsqChr71uRloETY5qTe1LRuTR7yhOi16aJbgjR50bKM6x0PohNdMnwfyI8od2mJjpBEARBEARBEIuERRGix5IpfPgXe9Hd5MFHr9tgus5oMIq+iTC2r2jMGIzmcdiwo6cRJ8ZDGJ6p8YEpQRAEUXXqqYneYNL2bjKEwK2+zCGZfLjotBIcl5OQGqLnbqJ7HDZV+VIOpcuccFJjT99kxuMHhpjKZVWrF5uW+QFow1cBTecynyb6yhYWohc0VBQwGSxaZp0LkBmsVxsxRG9ZC9jc5utRiE4QBEEQBEEQxCJhUYToX7n3MI6NzuHLN22Dy27ekuOe1JVCo0ykp5kd6B0fDZk+ThAEQSxe6iNEZ+/d4Mpsohub1C0+syZ65b4GVTVjsm1GRKXLfNE10Qcyw+gXB5nK5ayuANa2M3XIccGLrjrR5xGi89fdtjyQf6goANjdgM2lLUemS35vAOYheq296GKI7msHmlebr0chOkEQBEEQBEEQi4QFH6I/fXIS333sJN592Wrs7G3Out5shB2I85DBiN/NgoHZCgxkIwiCIOobvxIOB2sZoseyB9WZOheTJrqbrTMVKn8Tfa7AwaIA0MZD9GB03u8bFJroLw0H1UY856DSRN/S5VfD7qMmIbqZ/qZQLl/Xhi+9fhv+7aZthT9JVLqUW+cC1FeI7m0DWrKF6OREJwiCIAiCIAhicZD/aLjKXPnvD+l8ptno++IrEYol8fe/3IdVrV58+Jr1OdcPx9mBt8Nmft7AqbTLYsmU6eMEQRDE4iVQwRZ3oeTSufDBopxWkyY6/xrK7USXZbmoEL3dz1rYY+XQucS0ryWVlrHv1DQuXtuq3veiMlT0rOUBJNNskKkYok+UQedisUi4+bwVxT3J1QgEh9nt+QwWTaeBmcHM++spRPe165v3ItREJwiCIAiCIAhikVB3IXpPFt2KGZ///UsYnI7gV7denFXjwnHa2OOJlGz6eCyVBsB8rgRBEMTSwu/ig0UTkGUZkiRVfRty6VwaDToXs1C40c2d6OUN0WPJNFJKQF2UzqUsIbp2AjyeTOPZ/ik1RJ+JJNA/EQbAmujhODsJ3jceQiKVht1qUVv58wnRS6JcTfTQGJAy+T6aKV6qRSoJhAU/vbcV8GUZ6k4hOkEQBEEQBEEQi4S6S4xvv+X8gtZ76PAo/u/pAfz1FWtw9orGvOtzjctsloYe172YhRcEQRDE4oY70RMpGZFEqiYnVPkQzQaTtrfoRG/y2GEzcXNXyonOG/KSBHjynLAGRCf6/HUu/HtywapmPHp0HHv6tUCaq1y6m9xo9DgQcMvwOqwIxVPonwhhbXtDWQaLlkS5QvRsA0Rr2UQPTwAQCgnethxOdNK5EARBEARBEASxOFiwifFd+9hl0t986Di++dDxjMef7Z9C78d+jwtWNePn77tIdaXy1pqRgUmmkFnX3lChLSYIgiDqFY/DCptFQjItYyaSqEmIrulcMt/b77bDIgFpGWjxZfrQAc2JPh0urxNdVbk4bLBY8jf02/3la6JzT/wVG9rx6NFxPDcwhXRahsUi4cCQNlQUACRJwtp2H/adnsHRkbkah+iN2u35DBbN1jivZYguqlwAwNsONK8xX5ea6ARBEARBEARBLBIWbIj+8i0d6G5yZ9w/G03g+7v70BVw4Q07V6jrdPhd6G3xYO+paYTjSV1AEo4n8fzANHpbPOpANIIgCGLpIEkS/G47JkNxzEaSWBao/jYEY1znkulEt1okNHocmAzFsw7JbKyQE523wb0F+NABoL2B+bHLonNR3vu83iZ4HFYEo0kcGQ1iY6cfLw5yH7pfXX9tewML0UfncJ0sYyq8SJvowTNAMgbYarDPEhrVL3vbgIZOwO4FEoaZNvbCFX0EQRAEQRAEQRD1zIIN0a/d0olrt3Rm3H9qMsxC9EY3PmQYNnrzeSvwpXsO4yv3HsE/3bBZvf/L9x5GOJ7CX17UW+nNJgiCIOqUgBKi12q4KG+iZxve2ehh29earYmuKF9myuxEV5voBerOeGBdjkY8f+8mjwNnr2jE48cnsKd/ioXois5ly3LtjMe6DnbV2bHROcxGk+oclKqH6K5G7XZ0hg0ItZgPNs9JVve5zAL2liwN8EoSGtcve1uZ66d5NTCyX/8Y6VwIgiAIgiAIglgkLNgQvRTedekq/HH/GXxv90kcHJ7Bjp4mPNc/hadOTuL83ma89cKeWm8iQRAEUSP8SkicbXZGpcmlcwGYF/0EQmjx5W6iT1VK51JgE93jYN70iDLos1RSaVkdFupz2nDuyiYWovdN4bU7luP42BwATecCAGvbWGh7dHROHSrqcVjzDh8vO6LOBTIQm9G30wtF10SXoHORTw/UKEQ36Fx87ezv5lUmITrpXAiCIAiCIAiCWByUUItauDhtVvz0PRfg3ZeuwsBEGN997CTOzEbx/l1r8L1bzoPTVuWDbIIgCKJu8LsrM5izUILR7DoXAGhS2tQt3mxOdG3702nZdJ1SmFM1M4WF6G4lRA8nUpDl0reDh/cAU8mcu5KF0HsGpvDS8CxkGejwO3UaNt5EPz42h/E5ppOpegsdyAzMS1W6iE30zrOyP1ZN5gSdi92jBeVmgT6F6ARBEARBEARBLBIWXRN9RbMHfV98ZdbHG1x2fOpVm/GpV23Oug5BEASx9AjUOETnoXG2sPqSNS14+PAYzl/VbPo4PwmQltlATv71zHu7uBO9wGGrbqX1LctALJkuuQXOvx9OmwUOmwU7epogSWxA+MOHWRtabKEDQHeTB06bBbFkGvtOM2d6bUL0Rv1yqcNFxSZ693nA6EtAWjm5UKvhoqLOxduq3TYbLko6F4IgCIIgCIIgFglLqolOEARBENngIfRstD51Lu+4ZBVe/Oy1uGhNi+njLrtVDbDL6UWfiylKlUKb6EJoPh+ly5zh+xFw27G+vQEA8NOnWYC8pcuve47VImGNonR56sQEgAXcRI+HgPCEtty4Egh0a8s1C9EFnYu3XbtNTXSCIAiCIAiCIBYxFKITBEEQBGrbRE+m0qr/O5vOBQActty/trkXfTpSPi8617kU6kS3WS1wWNl2hhPzCNFN3vccRekyPse+PnGoKGdtOwvRn+mbBMBc8lVHHCwKlBaizwzqlwPdQKMwu6VmIbqgc/G2abebV2euSyE6QRAEQRAEQRCLBArRCYIgCAKAXwmvZyPJPGuWn1BMC5sLDavNaFQC4+lyNtHzNOTNcJdhuChv5osN+J0r9Q3vs0xC9HVKiD6lfA/qookenS7+NYzO88YeICCG6DVyomfTufg6MvUtFKITBEEQBEEQBLFIoBCdIAiCIFDbJjpXyHD/d6nw4aLTZfwagoqb3FtEuO8pQ4jOnejiSYVzhRC9yWNHV8CV8Tw+XJTT7KtFEz0AQNKWS2qin9YvG5vowSEgWb4rDgpClvU6F5+gc5EkoHmVfn1yohMEQRAEQRAEsUigEJ0gCIIgoIXoszUI0TUf+vyGgao6l3D5wtWQSZidD+5Fj8xH58Kb6E7te7KyxYNWJRQ/a3kAkiRlPG+t4k3n1ETnYrECLsHXXspgUbGJbrGxprcYostpYHYw83mVJBYEklFtWdS5APrhohY7YKvB954gCIIgCIIgCKICUIhOEARBEAD8bhYS12KwKG9d+4tQppihhejlHCxaus4lHC9djWP2vpIkqW30LV2ZKheABe02ixau10TnAuiVLvNtovu7WDDfuMKwTpWVLmILHcgM0TvP0m77uyq/PQRBEARBEARBEFWCQnSCIAiCQG11LkEluPfNM0QPuCvnRC+piV4OJ7rhfT98zQbcdG433nlpr+nz7FYLVrVqLu6ahegewRduDJ8LQQzRuQtdbKID1R8uKvrQgcwQ/Zx3AN3nA9524JrPVm2zCIIgCIIgCIIgKs38jtYJgiAIYpGgDRatpc6lTE30SPl0LqU40dXBovPQuZgNFgWADZ0N+Mobtud87roOH46OzgGoYYgu+sJLCtGFlnmgm/3d0AVIVkBWvq9VD9FH9cvGEN3XBrz7vuptD0EQBEEQBEEQRJWgJjpBEARBQGuih+IpJFLpir3PcwNT+OZDx3TvwYPqBuc8nei8TV/GJnopTnSPqnOZz2DRRNHvy1nbpg20bPE6S96GeeEVmuhzRYbo6TQwI/jOeYhutQH+5dr903WmcyEIgiAIgiAIglikUBOdIAiCIKBvgc9GEmjxVSZ8/dxdB7H31DS8DhvefnEvgPLpXBqVIZrTZWzTz5XQkuc6l+h8BouW4GLnrO1gw0WtFmne7f6SEQPm0Bggy4DJIFRT5kaAtPBvyEN0gCldZpQGei11LpIF8DRX9/0JgiAIgiAIgiBqBDXRCYIgCAKAzWpRW8+z0dIHYuZjaDoCAPj+7pNIp2UA5de5TIXLo3NJpWWElDZ5UU50B1t3Pk30bE70QtjS5QcALG90w2IpMLguN15B55KKAbFg4c8VfeiAfqCo6EWvdog+J+hcPC1s2ClBEARBEARBEMQSgJroBEEQBKEQcNsxF0tWbLhoOi1jIsQC7r6JMO4/NIprNncIbe/56Vy4kqZcXvdQXDuZUIwTvTw6l9JD9DVtPnz7reeiu8ld8vvPG1HnArA2ustf2HNnDJqWgBiiC7dnB4FUkmleqoGocyGVC0EQBEEQBEEQSwhqohMEQRCEgr/MIbSRqXAcKaV9DgDfe+wkAE3n0lBCYCzCQ/SZSAKyLOdZOz/ch263SnDaCt9lKIvOJctg0UK57qxOnLU8UPL7zxtxsChQ3HBRY4guetDFJrqcAoJDxW9bqYg6FwrRCYIgCIIgCIJYQlCIThAEQRAKfiWwrVQTfWwuBoCFzFaLhCdOTODA0EzZdC48RE+kZETmEWBz5gSlilSozxuAW22il67FmSvTsNWaYQyZRRVKPkSdi7sJcGqDUnUhOlBdpUtI+BooRCcIgiAIgiAIYglBITpBEARBKIhN7kowFmQh+soWD64/qxMA8L3H+hCMlUfn4nFYYVMc4OX4Gvh2FdsG5030eelc5tlErzne+TTRhRBdVLmYLU8bWuuVhHQuBEEQBEEQBEEsUShEJwiCIAgFVecSrWyI3tbgxLsuXQUAuGvfEE5NhgHMPzCWJEk9ETAdnv/XwINsr6O47eJO9FJ1Lum0jLl46U70usDdBEjC4M1SdS7G0Ny/HJCE3bdqNdFTCSAypS37KEQnCIIgCIIgCGLpQCE6QRAEQShUq4ne5nNiR08TdvQ0Ip5KY3gmCmD+OhegvF9DKFaaZsY9z8Gi4UQKXOleju9JTbBY9MNFiwnRxXZ5oFv/mM0BNHQJ61YpRBd96AA10QmCIAiCIAiCWFJQiE4QBEEQCgF1sGjpLu9c8BC9tcEJAGobneMvQ2DsL2OIrupcimyDc51LqV523oC3WYobaFp3iEFzoU70WBCITmvLjSsy1xHvm6lWiG44CUAhOkEQBEEQBEEQS4gFfGRKEARBEOWFh9izFWqij89pTXQAuG5LJ7oCLvVxXxmGaJazia55yYvbLo+if4mU2ESfiyWU9y1uoGndIQbNxiZ3NmYG9cvGJjqgHy5atSY6hegEQRAEQRAEQSxdKEQnCIIgCIWAp8JO9DnNiQ4ANqsFb7+4V328nDqXcpwImFOb6NY8a+pxO9juRak6l2C0NI1M3aEL0QtsootDRYFMJzqgD9FnTgPp0ge4FgyF6ARBEARBEARBLGEoRCcIgiAIBb/LvMV9ZCSIz951AMF5huviYFHOm87vQVfAhXXtPnUg53yohBO9eJ2L0kQvVeeivu/8m/k1xdeu3S7UiW7Us5g10cVgPZ0EgsPFb1uxZITorebrEQRBEARBEARBLEIWeMWLIAiCIMpHtgD6Y796Ac8NTKOn2YNbLlll9tSCMAvRA247/vThl8FhtZRFXVLOED1YYpjNTwaUrHPhTfQiw/u6QwyaozNAMs4Gg+ZCbKJbHYC3PXMdY7AePGMetpcTMUS3ewGHt7LvRxAEQRAEQRAEUUdQE50gCIIgFMxUKAMTYTw3MA0AGJyKlPza8WQaU2H2utyJzvE5bXCUaYBmZZzoRTbRHdpgUVmWi35fNbxfTDoXoLA2uhii+5cDFpP/F65G/XJstuhNK5o5Ydt9pHIhCIIgCIIgCGJpQSE6QRAEQSj4eYgeTarh72/3aoMez8xGS37tiRBrodsskhp0V4KyhuglO9HZ+qm0jHgqXfz7RkvTyNQdxhZ5ISH67JB227/cfB2XX78cC5qvN/AU8M2LgO+/AgiO5H/vXIjbTj50giAIgiAIgiCWGBSiEwRBEIQCD6BTaRmhOGtR3ymE6CPzCNHHg3EAQKvPCYtl/tqWbPgrEqIXF/q77VroXorSZW7RNNEN3vBCQnRxHZ+JygUAnA365Wwh+oOfB0YPAv27gYf/Lf97F7pdFKITBEEQBEEQBLHEoBCdIAiCIBScNgscVvarcSaSwIGhWRwfC6mPz6eJPjbHniv60CtBPehc7FYL7FZ2oqCU4aI8RF/wTnRjCF5IiD4nNMZ9HebrGEP0aBady8QJ7Xb/7vzvnQsK0QmCIAiCIAiCWMJQiE4QBEEQCpIkaU3ucAJ3Ps9a6Nu7AwCAkdlYSY5vwHyoaCVo9GR63UtFa6IXH2bzNnq4hCZ6cLHoXDyGJvrcaO71k3EgMqUtZ3OP2z2AJOzCZWuii670scPZ18tHLMiGl6rblaUhTxAEQRAEQRAEsUihEJ0gCIIgBAJuFtxOh+P43T7mp37P5asB6IeDFosaovuq10QvNfDnzCtE58NFl7LOxe4CnAFtOV8T3fh4tia6JOnb6GaDRdNpQ2guA0N7c79/Nk4+CsjCv2PXOaW9DkEQBEEQBEEQxAKFQnSCIAiCEOBN9D8dHMFoMIaA246Xb+5Es9cBADgzU5rSpVpNdB6iJ1JySSoVkfmE2R4He05JOpcoO1Gx4JvogN6LnjdENzTVjYNJRcRw3qxhHp8DYDiJMrgn9/tn49iftdsWO7Dq8tJehyAIgiAIgiAIYoFCITpBEARBCPAQ+o49pwEAr9i6DA6bBR1+F4DSh4uOzbEQvdXnKMNWZsfjsMKmDC6djxc9lkwhnkwDqL7ORXWiL/QmOqBXn+QL0Y26l1zaFF0T3SREN2unlxKiy7I+RO+5EHD6in8dgiAIgiAIgiCIBQyF6ARBEAQh4HexEJ0Hua85uwsA0OlnDfJSh4uOB+MAgLYG13w3MSeSJJVluGgopoXfXkXNUgzz0bloTnR70c+tO8Qm+lylQnSTwNxs2Ojgc7nf34zJE8B0v7a89uriX4MgCIIgCIIgCGKBQyE6QRAEQQjwABoAugIunNfbDADoDLDwu2Sdy1x1dC6A4EUv0d8OAHNKkO22W2GzFr+74OEheiJZ/HsvFic6oFey5G2ijxiem2WwKFBaE332tH5AaCGILXSAQnSCIAiCIAiCIJYkFKITBEEQhIAYot949nJYFDXKvHUuVXKiA5rXfT5N9PkG2a556Fy0JvpiCNGFIDw0xgZ+ZkMM2d3NgDVHEz9fiG7WRAeKb6Mfu1+77esEOrYU93yCIAiCIAiCIIhFAIXoBEEQBCHgd2vB7Wt2dKm3O5UQvRSdSzieVEPpqjbRyxCiN5QYZHtK1LnIsrzInOhCiC6ngOh09nXFJrqvI/fruvza7UKb6EBxXvREFOh7VFteexUgSYU/nyAIgiAIgiAIYpFAITpBEARBCLT6WMi9sbMBGzu1oLJjHjoX7kN3260l+cWLpTwhOnuut8ohejSRRiotA1iETXQgt9JFdKb7cqhcgPxO9HKE6ANPAImwtrz2qsKfSxAEQRAEQRAEsYhYBEenBEEQBFE+XrF1GQ4OzeI1O5br7u+ch85lbI49p7XBAakKTV4eos/OI0SfjcyvDc51LpFEcSF6UAnvJUkL4hc0xhB9bhRo22C+bjFNdKehiS7L+pZ4Np3L0HNMKWMpoEdxXFC5QAJW78r/HIIgCIIgCIIgiEUINdEJgiAIQsBlt+JTr9qMs5YHdPfzEH0qnEC0yGB4TGmit/kqr3IBytNEH1cGobaUuM08AC/WiT4n+NCrccKh4oiDRYHcTfTQqHY7b4guNNHlNBAP6R/P1kSPzgCTJ3K/Nkf0oS8/F/A0F/Y8giAIgiAIgiCIRQaF6ARBEARRAI0eOxw29mtzdDZW1HPH5qo3VBQoV4jOgv9Wn6Ok57vtpelc5utirzu8rfrlbCF6IsoCbvV5RehcgEwvupknnVOI0mVmEBg9qC2vvTr/cwiCIAiCIAiCWLgc+gPw3ZcDX1gBfHEl8H9vBs68WPrrnXkR+FwrsOcH1XvPCkIhOkEQBEEUgCRJJQ8XHQvWJkSfLkMTvbXEJrrbwULwYnUuahN9MQwVBQBXALAKJyKyhejG+4tpogOZobmoc2lcCdhc2nIhIfrxB/TL5EMnCIIgCIIgiMXLM98FfvZmYHYYOOdtwKZXAccfBG67GjhdxFwlzswg8PO3AOkcx6Tlfs8KQyE6QRAEQRTIvEN0nyvPmuXBX0adS6kKmlJ1LsGYpnNZFEiSvlU+N2q+nvH+vINF/frljCa6EKJ7moFl27XlQkL0Y3/Wbrsaga5z8j+HIAiCIAiCIIiFR3AEuPcTQPMa4NbHgGv/BXj1fwPvuBuQU8BdH2QzmAql/wkWhE/1Ve89qwCF6ARBEARRIB0BFoKPLpAmejlC9NaG+elcivXHa010e0nvW5eIIXpo3HwdcagoUNxgUSDTgS420Z1+5jTnnHkBSMazv3YqCZx4SFteswuwLpKTGgRBEARBEARB6Hn+h0AyClzyt+xKWk73TmDrG4CRF4FTT+d/negs8JtbgdtfwY5Pei6u/HtWEQrRCYIgCKJAOv0sBD8zU2SIrqpRSguki4WH6LPzCdGD3Ileqs6FN9GTRT1v0TnRAUOInqWJbrzfOJDUSIbOZTb7sssQoqfibKc0G0PPAdFpbXkNqVwIgiAIgiAIYtHS/wT7e/XLMh9bpdx38pH8rzPVB+z7KZundOtuYNXllX/PKrKIjlAJgiAIorJ0lKhzGa92E92jNdFlWYYkSUU9X5ZlTITYNrdUWeeihuiLxYkOAD4hEM/mRBd1LpIlcyCpkbyDRcUmegBYbtCxDO7JvI8jqlwA8qETBEEQBEEQxGJm/AhgsQGBFZmPNa/S1slHwzLgXX8GVpxXvfesIovoCJUgCIIgKkunonMZKSJEl2VZbaJXW+eSSMmIJFLwOIr7dT8TSSCRYv65Fm91dS7B6CJzogP6QHyugBDd0wJYrLlfs5jBoi4/0LQKcDcBkSl23+Bz2V/7zH7tdusGwN+Ve1sIgiAIgiAIgli4RGeYAtLsGISrVqIz+V/H15Z/tlO537OKLKIjVIIgCIKoLKUMFp2NJhFPpgGUrkYpFq/DCqtFQiotYyaSKDpE5z70BpcNLnueMDcL7pKb6ExB41tMTXRRzZIIAfEQ4PDq1xGd6Pl86AB7vmQBZPZ/Sxeiy7Khid7ABpx2nQMcv5/dl2u4aGRaux3ozr8tBEEQBEEQBEHUF9/YCUwczb/eZ2aA+Fz2YxCrUqpKFnc1dl5q8Z7zZBEdoRIEQRBEZeE6l5HZWIYmRZZl/ODxPmzuCuD8Vc3q/XyoqH8egXSxSJKEgNuOyVAcM5EElgXcRT1/TPGht80j9OdN9EixIfqibKIb2hih8cwQXdS8GNc3Q5JYOM7bGWJonowCacFFz4eQLj9XC9HHj7DnikN8OGLjw+xxgiAIgiAIgiDqm6bewte1udjcJDP4/Q7fvDep5u85TxbRESpBEARBVBYeoseTaUyFE2gWVCePH5/AZ+46iGavA0994irYrWx291iVfegcNUQPFz9cdFwdhFr6NvP2e6RIncvidKIbQ/QxoGml/r5im+gAC8d54C3qW6KGIaMuIURXkYGhveaDfMShohSiEwRBEARBEMTC4613FL6uu4kdo8gyK+uI8OMNfkxRLmrxnvNkURyh7j01jf964Cie7Z9CPJlGb4sXb7mwB286rwdWi/4fIhRL4tsPH8fdLwxjaDqCtgYnbtjehQ9euU699JwgCIIgzHDYLGjxOjARiuPMTFQXoj95YgIAMBmK47Fj49i1gSk8xsoQSJeC360NFy0WNURvKM2HDmg6l2RaRjyZhsNmKeh5mhPdXvJ71x0ZTXQTL7roSi/UIyh60UWdS8wQovMmeudZ+vunB8xfl5roBEEQBEEQBLF0aF0PzA4CM6eBRsOgz8mT7O+2DQv/PedJYUe0dcw9Lw7jpm89jqdOTOK6LZ24eecKzEYT+ORvXsTn7jqgWzeWTOGdtz+DbzxwDN1Nbrzz0lVY3ujGtx46jrfc9iRiyeLacgRBEMTSQ1O66P1sT5+cVG/ftXdIvT1ewyY6MM8QvQw6F6C4Njpvoi8uJ7ohFBeHiAJAPAzEhRC84CZ6lhDd2ETn63la9PeLjXNOKsn8hBwK0QmCIAiCIAhicdN7Cfv75COZj514iP3dc9HCf895sqBD9LFgDP9wxwto9jrwuw9cii++fhs+c+MW/OlDl2N9hw8/eKIfx8e0A8EfPzmAp05O4n0vW40fvesCfPS6jfj5+y7CLZf04rmBafzoif4afjUEQRDEQqAzkDlcNJZMYe+paXX5TwdHEFWCY95EX1AhuuJEn0+I7rBZYFOuBivGi66G6IvJie5p1S8bm+ghQ6guDiLNRaFNdB6E292AVfg3FQeI5nsuQRAEQRAEQRCLk21vZMcJj3wZCGvlMJzeA7x4B9C5Fei5cOG/5zxZ0CH6L549hdloEh9/xUasatUGdHkcNnzk2o14484V6mXhAPCjJ/rgtFnwwSvX6V7nI9dugMdhxU+fynJZM0EQBEEo8Cb6mRktRH9xcAaxZBotXgeWN7oxF0viwUMsGK2dE52F0LM1aqIDWhs9HE/mWVODDxZdVE50qw1wa8NmM0J0YzPdV2iILjgCC9G5AIC7Ubtt1kQ33ieuTxAEQRAEQRDE4qOxB7jyU8DUSeBbFwP3fAL47d8At78SsNiBG76uX3/4BeDBLwDP/6R671kHLOgQ/YFDo7BbJVy7pTPjsWs2d+DfbtqGs1c0AgBGg1H0TYSxfUUjvIZ2m8dhw46eRpwYD2F4JlKNTScIgiAWKJ0mOpenFJXLeb3NeNX2ZQCA3+1jShc1RK+yE708OpfSneiA5kUvRucSXIxNdEAfjJctRBeb6AUMFgUAV6N226yJLvrQAWqiEwRBEARBEMRS4JIPAq//LtDQCTz7XeDIPcCaXcC77gWWn6tf98x+4OEvAnt/Wr33rAMW7BGqLMs4NDyL7iYPkmkZn7/7IP6wfxjjoThWtXjxjkt68ebze9T1j40yrcvKZo/p6/U0e7EbEzg+GsKygLsqXwNBEASx8OgMsDBc1Lk8w0P0Vc24YFUzvvPwCTxwaBTBaKKGTfT5hOiKzmWe2+zhIXqBOpdYMoV4Mg1gkTnRAeZFHzvEbhtD87kR/XJJTvRZ89tAkU10Y4jemLkOQRAEQRAEQRCLj603sT/52PEW9icfuz7O/pTjPeuABXuEGowlEYqnIAF4/Tcfx3Qkgas3dUCWZdx3cAQf//V+HB+dw6detRkAMBthzbZGj9309fz8svdo8WEDQRAEsXQw6lxSaRnP9k8BAM7vbcaWLj9Wt3lxYiyE+w6OqE70+apRiqXUEF2WZbWJPt/2vMteXBM9FNPW8zoW7C6KOeJw0dC4/jGxmS5Z9eqXXBh1LrIMSJJe7SJZmQudk6+JbryPmugEQRAEQRAEQRD1F6Jf+e8P4cRYKO96T378KgDAifEQ1nf4cO/fXY5mL7vs/MPXRPHabz6O2x47ieu3duLclc2qj9VhMzfYOK3s/liy8EvOCYIgiKUHHyzKdS6HzswiGE3C57Rh07IGSJKEG7Z14ev3H8Wde4cwGWKt7vYF0kSfiyURU9rg8w3+eRM9XGATnfvQvQ4rrMpQ0kWDLkTP0UT3tgGWAm17YhNdTgOJMODw6nUuLj8L1jlFN9EpRCcIgiAIgiAIgqi7EL0ni27FiHh8+fHrN6kBOgC0+1344FVr8dFf7cdv9w7h3JXNcNrYgXwiJZu+XizFAgPPYmu+EQRBEGWFO9GnwglEEylV5XLOyibYlBOyN57NQvRHjrCGsSRB93uqGgTc7P2KDdG5ysXrsKpO81Lhv1ML1bkEY2xbF53KBQB8QogengRSSTZwFNDrXcT18iG6zgHWQHd4gZgQhDsN6xTrRDc+nyAIgiAIgiAIYglSd0ept99yfkHrRYVLw7d1Z7akzlrO7uufCAPQNC6zWcIErntpWIwH7gRBEETZCLjtcNosiCXTGJ2N4Zk+rnJpUtdZ0+bDli4/DgyxRnCL16EG7NXcTgCYUX6/FYo6VLQMzflidS7B6CIdKgrom+iQgfAE0KC4z3UheoE+dEDfRAdYiN7QmdlEFzE20bkCRr1PCNFtLsDuKnx7CIIgCIIgCIIgFinVPaIvIy67FV3KJfVm7fKkcp9bOYBf2/7/27vz8Cjre/3j90z2lbCGxWDYElYXICKbIARBlLoclwIVUcbXmakAACw3SURBVFBaQZGqPegRrZ7Tiod6lEKPWxXrQdFfqdRigyxl07KICIKCaNkRCGt2ksny/P4YZpLJzCSTzDMkk3m/rosrmWcf53HyzZ3PfL7xkipD9eqOnLO3kOnWJsHjegAAJMlisThbupzMK9YXhy5OKprq2sd63JXtnd9f6n7oktSsyh+PDcPzp7A8OZNvXg/3+rZziY/2PH9JUItr4/q4ah/0qu1dqm9Xk+ohuiM8rzqxaE2V6BVlkq1aC72qITqTigIAAACApCAO0SXp2s4tJUmf/XDabd3OozmSpB7t7L88JidGK7VlrHYezXH2R3cospVpx5EcpbaMVetL3LMWABB8HJOLbj1wVqfzSxQZZtWVKUku21QN0RviZ4ujEt1WXqHi0gqf93NWosf7337G8YfsCzbfquELSi5+KqzJV6LLNUR3qUSvS4hevZ2LI0TP975N1Up0yb0vukuITj90AAAAAJCCPESfeO3lslikV9b8oJO5xc7lx3Mu6A/r/qXIcKtu79vBufyujBRdKC3X71Z+73KceSv3qchWrnsGpl6qSwcABDFHX/RPdp2QJF2Z0szZusShQ1KM+l9ub/HSugEq0atOzlmXvuinL/ZEN6MS3dFT3ed2LiVNuJ1LQrU2LSd32b+WFNgnBHWoU4juoZ2LVHM7l+rV5dX7ohOiAwAAAICboP4ttd/lzTVzZDe9suYHjX5lo266op0kacXuEzpfVKrnftJLKVUmKp0ypJNW7D6pt/95UHtO5Orqjs311eHz2nrwnK5JbaGfXduxoZ4KACCIONq57Mu2h5bVW7k4TB3aWTuOfqWhaa0u2bU5WCwWNYuJ0LlCm3IvlDqvuTaVlegN2c4lqIcnnjVLkZI6SjlH7I+/y5IGz5QKsl23q1NPdA8Ti0o1t3OptRK9ymNCdAAAAACQFOQhuiQ9mpmmXu2b6Y+fHdDHO36UxWJRz/aJ+vmwzhrR3fUX0ajwML3/wADNX/ODsnaf0FdHctSuWbSmX99FvxjeVVHhYV7OAgBAJUc7F4eMTp5D9DG922rff4655JOKOjhC9Jwim8/7OHuim9CCxtHOpdjHSvSCEnvFfJOsRLdYpO43S1v+1/746FZ7G5fCai3pqrd9qQmV6AAAAABwSTSJ31JH9UzWqJ6+VW4lREfo6Zt76umbewb4qgAATVXbKiG61WL/ZJQ3DRWgS1Lixb7odWnn4qhEb21GT/R6VqInNMVKdElKH1sZosuQ9q1wrwyvSyV6ZJwki/1Ykr0CvbxUKrtQuU31oJ2e6AAAAABQZ0HdEx0AgIbQtllllXaPdolKjI5owKvxrlm9QnTze6L7GqI36Z7oktRxoBRT5Q8u+7JcJxWV6tYT3WJxbddSkuc6qajk3s6FSnQAAAAAqDNCdAAA6qhqOxdv/dAbg/qF6Ob3RPe5nUtT7okuSWHhUtqNlY/3r5POHax8bI1wD7lrU7XSvCTfNQSX3IPwiBgprMqnDKpWopfZXCc5rV61DgAAAAAhihAdAIA6apNQGaIP8NIPvTFoFmMPo/N8DNGLbGXOqnFzeqKHXzyurz3RHe1cGmdlvym6j638vrxE+mZp5eO41pK1jkOz6iF61UlFJfdKdIvFNaivWolefV8q0QEAAABAEiE6AAB1FhluVUZqc7WKj9LALi0b+nK8qmsl+pl8eyuX6Air4iL9n2zb0c7lQl1D9KbazkWSuoyQwqtMTFuQXfl9XVq5OFSdOLQk30M7l2o90SXXCvOqlei1VbEDAAAAQIhqwr+lAgAQOIunDpCtrKJRV03XNUQ/fbGVS8u4KFksFr/P72jncoF2LpUi4+xB+r4s93X1CdGrhuTFefZ/VUVXq0SXvFeiV++PTogOAAAAAJKoRAcAoF6iwsMadYAu1aMS3dEP3YRWLpIUE+GYWLTMp+2b/MSiDt1v8rzc3xDdl3YuUg2V6Dmu2xGiAwAAAIAkQnQAAJqs+oboreMja9nSN3Vu51IcIiF62hjJ4mEIFmdCiO5PJbpbO5ckAQAAAAAI0QEAaLIS69kTvVW8OZXodWnnUlZe4dwuoSm3c5GkuFZSyrXuy+OT636sqpXmJXnuleiR9EQHAAAAAH8RogMA0ERVVqL71k7F2c7FpBDd0c6ltNxQaXlFjdueK7IH+FZLCFSiS55busS3rvtxoqpPLFolRI9MkKwehnrVK9ENw/49IToAAAAAeESIDgBAE+UI0fMulMpwBKU1qAzRzW3nItVejX4qrzLADw8LgeFJ97Huy+pViV6l0twol/KzKx97auUiuVaiV5RKpUX276uG6BFxUljj7vkPAAAAAJdKCPyWCgBAaHKE6LbyChWX1lwJLpk/sWhkmFVhVouk2vuin8ovliS1STTn3I1ei85Sm56uy/ztiS5JeT9WWeclRK/e69zRF71qiE4VOgAAAAA4EaIDANBExUeFO0NsX/qinykwtye6xWJxtnSpNUS/WImenBBtyrmDgktLF4uU4GcluiTlHqv83pdKdKmyLzohOgAAAAB4RIgOAEATZbFYlHhxkk6fQvR8c3uiS5UtXYpqCdGzL4boIVOJLkkZD1RWn181oX7BtVsl+nHv6xy8VqLnVNmGEB0AAAAAHEJg5i4AAEJXs5gInS8qrTVELy4tV36JfQLS1maG6I5K9NKaJzfNdrRzCaVK9IRkaebXUs5hqXX3+h2jethdXlL5vbd2LlSiAwAAAECdEKIDANCEOfqi1xainy20t3KJDLMqMca84UFspKOdS8092Z3tXBJDKESXpMhYqU2P+u/vrdpc8t7OhZ7oAAAAAFAntHMBAKAJS/QxRHe0cmkZHymLxWLa+SvbudRcie6cWNSkSU1DRk0hOpXoAAAAAGAKQnQAAJowXyvRzxSY3w9dqtrOpbae6PYQPeQq0f1Vn0r0iFjJGlH52FMlevWgHQAAAABCGCE6AABNWN1D9EhTz1/ZzsV7iF5eYehMgb2dTHIoTSxqhog4SV4+OeCtEt1icQ3Ji3Ok0mKprLhyGZXoAAAAAOBET3QAAJqwpFh7iP7+1iOylVXotqs7KL2te/WyI8Q2uxI9OsLRzsV7iH62sETlFYasFqmlyedv8qxWezV6SZ77Om8humTvi1542v79hRz3/QnRAQAAAMCJSnQAAJqwkT2S1SwmQmcKSvTahv0a/cpGjZ3/md7550GVlVdO9nn6Yk/0Vib3JHdWotfQzsUxqWir+CiFWc3rxx4yvIXl3tq5SO6V6I6WLs59CdEBAAAAwIEQHQCAJqxvx+ba+tRI/e/EvhrVM1kRYRbtOZGnXy/foxnv71BJmT3cDlRP9NhI+4feamrn4pxUlFYu9eOtL3ptlegOF3Jc+6FLhOgAAAAAUAUhOgAATVx0RJjG9mmnNyf11xdPZWrOzT0VGWbVp9+e1NQ/fakiW1nAeqJH+zCxaPbFSvTkBCYVrRevIXoNk45Wr0QnRAcAAAAArwjRAQAIIc3jIjVlSCe9PTlDMRFh+uyHM5r01hc6dv6CJKm16ZXotfdEd7RzaZNIiF4v3sLymtq5uFWi51RbT4gOAAAAAA6E6AAAhKAh3Vpp8dRrlBAdri8Pn3eG6AHriW4r87pNtqOdi8nnDhleK9FrCMLdKtFzXNdXDdkBAAAAIMQRogMAEKL6Xd5CHzx4rVrGVbZwMbsnui/tXE7l2UP0ZCrR68dbxbmvlejlNik/23V9Tf3UAQAAACDEEKIDABDCerVvpg+nDVRqy1hdmZKkpJgIU4/vUzuX/Is90ZlYtH48Bd7hMVJYDa9l1Up0Sco5XPl9ZIIUFm7KpQEAAABAU8BvSAAAhLiubeL1j8eGyyLJarWYeuwYRyV6DSF6dp6jnQuV6PXiqZ1LTZOKSu7tWs5XCdHphw4AAAAALgjRAQCAwkwOzx1iImtu51JeYeg0lej+8RSY19TKRXKvRD9/qMq+hOgAAAAAUBXtXAAAQMDERtr/Xu+tEv1sYYkqDMlqkVqa3I89ZHisRK8lRK9eiV5wsso6QnQAAAAAqIoQHQAABExMLROLnsqzV6G3io8KWDV8k2dGJbrLvoToAAAAAFAVIToAAAiYmFomFj2Vb++HnpxIP/R681R1Xtee6FXVFLADAAAAQAgiRAcAAAHj7InuJUTPvliJ3iaBVi715jFEr6WaPDJOsnqZGodKdAAAAABwQYgOAAACJvZiOxdbeYXKyivc1mfn2SvR21CJXn/1aedisXivRidEBwAAAAAXhOgAACBgHJXokue+6KfyqUT3W30mFpW8t20hRAcAAAAAF4ToAAAgYKLCrbJcnC/UY4ieR090v9WnEl2iEh0AAAAAfESIDgAAAsZisThbunjqi+7oiZ6cSCV6vUXGS7K4LqttYlGJSnQAAAAA8BEhOgAACKiYSPsElkUeQvRT+Rd7oidQiV5vVqt7aO5LOxcq0QEAAADAJ4ToAAAgoGIi7cON6u1cyisMnc6nEt0U1UN0X9q5eK1E97IcAAAAAEIUIToAAAio2Ah7JXr1di5nC0tUYUhWi9QynhDdL26V6D5Uk1OJDgAAAAA+IUQHAAABFR1p74levZ3LqYv90FvFRynManHbD3XgFqLTEx0AAAAAzEKIDgAAAso5sWi1di7ZefZ+6MmJ9EP3W/Ue6L60c/FYiW7xrZ86AAAAAIQQQnQAABBQsRcr0S/YylyWn7rYD71NAq1c/FafiUU9VaJHJdonKgUAAAAAOPFbEgAACKhoZ4juuRK9DZXo/qsaolvDpYiY2veJae6+jFYuAAAAAOCGEB0AAASUo51LkVs7F3slenIileh+i2tV+X1sK8niQ495T+1cCNEBAAAAwA0hOgAACKgYL5Xop/MvVqInUInut953VAbgGVN928dTOxdCdAAAAABwE97QFwAAAJo2byE6legmattbemSnVJInNU/1bR9PleiegnUAAAAACHGE6AAAIKBiI+zDjbOFNpflpy5WoifTE90csS3s/3wVGWfvn15RZcJXKtEBAAAAwA3tXAAAQEBlpNonsPxk13EdOlMoSSqvMHQ6316J3iaBSvQGYbG4V6MTogMAAACAmyZRif7htiN6d/Nh/etUgcKtFvXu0Ey/GN5Fw9PbuG1bWFKm1zbs1ye7Tuh4zgW1TojSuCvb65ER3ZwfNwcAAOYZ1LWVhqW11obvT+u3WXv1xqT+OltQogpDslqklvGE6A0mJkkqOlP5mBAdAAAAANwEfSX6b7P26t//sltnCkr004wU/eSqDtpzPE+TF23T//vyqMu2JWXluv+dbVqw9l+6rHmM7h/SSR2SYvTq+v2a+MctKikr93IWAADgj6dv6qEwq0Wr9mRr0/4zOnWxCr1VfJTCrJYGvroQRiU6AAAAANQqqEP0w2cL9cfPDuiy5jFa9egwPXdLb71wex/97eEhiosM0/PL97hMYrZ4yxFtPXhO04Z11v9NGaB/H9NdH04bqPsGp+qrIzn6v82HG/DZAADQdHVLTtDPBnSUJP3nJ3t1POeCJPqhN7jqE4kSogMAAACAm6AO0b8+lqsKQxrdq62axUY4l3dqFaeBXVqpoKRM+7Lzncv/b/MhRYVb9ciIbi7HeWJ0umIjw/T+1iOX7NoBAAg1j2amKTE6XHtP5OnVDfslScmJtHJpUFSiAwAAAECtgjpEbxEbKUk6dr7IbV12XrEkqWWcfZtT+cU6dLZIV6YkKS7KtRV8bGS4ru6YpANnCnUi90KArxoAgNDUPC5SMzPTJEk7juRIklonUIneoNwq0ZM8bQUAAAAAIS2oQ/RrO7dQtzbxWrUnW29s3K+cIptO55fo+eV7tPvHXI3t01YpLWIlSf86VSBJuvzi4+o6toiTJO0/VXhpLh4AgBB0z7WXq3OrOOdjKtEbGJXoAAAAAFCr8No3abzCw6xa8uC1euqj3fpt1nf6bdZ3znWTBl6up2/q6Xycd6FMkpRUpe1LVYkx9v8UecWlAbxiAABCW2S4Vf9xUw9N+dOXkqQ2VKI3LHqiAwAAAECtGl2IPuKl9TpwuvZq8ENzb5JhGHpj4wGt23dKnVvH6bpurXXBVq61+07pgy+OKjkxWtOv7ypJKrLZQ/TIcM/F91Fh9uUlZeUe1wMAAHOM6N5Go3oma83ebF2VktTQlxPaWnSp/D4yQYpr3XDXAgAAAACNVKML0Tt6abfiyR8/O6g3Nh7Q2D5t9crdVzsD8twLpZr09heat3KfOrWK09g+7RQVHiZJKi03PB6rpLxCkr0/OgAACByLxaJXJ/bVuSIblegNrdsoqd9k6eg2acgsKYLXAwAAAACqa3SJ8Tv3XePztku+OCJJeu4nvV0qzJvFROjZcT11+/9u0pIvjmhsn3bONi55Fzy3a3G0e0mIbnT/SQAAaHLCw6wE6I1BWIQ0bn5DXwUAAAAANGpBPbHosZwLSoqNUOsE90nJ0pMTJEk/5lyQJHVtEy9JOny2yOOxjpyzt5Dp1iYhEJcKAAAAAAAAAAhCQR2it46PUt6FUp3OL3Fbd/CMPRRPvljllpwYrdSWsdp5NMfZH92hyFamHUdylNoy1mMgDwAAAAAAAAAITUEdot9yVXtVGNJ//X2Pyi72NJfsofh//X2PJOnWq9s7l9+VkaILpeX63crvXY4zb+U+FdnKdc/A1Ety3QAAAAAAAACA4BDUDcAfHtFNWw6c1cc7j+ubH3M1PL2NikvLtX7faf2Yc0E39WmnO/ulOLefMqSTVuw+qbf/eVB7TuTq6o7N9dXh89p68JyuSW2hn13bsQGfDQAAAAAAAACgsbEYhmE09EX4o6SsXH/87KD+tvO4Dp4tVJjForTkeN2d0VHjr0mRxWJx2T6/uFTz1/ygrN0ndKbQpnbNonXzFe30i+FdFR/l/W8KPXv2lCTt2bMnoM8HAAAAAAAAAGpCVnlpBX2IfqlwYwIAAAAAAABoDEzPKr/Lkv75inRqr2SxSpcPkq7/D6lt7/od7+Q30hvDpZtekvrd675++zvS8pne93/igBTXsn7nDoCgbucCAAAAAAAAAPDDtrekv/9SatZR6jtJKs6Rdv9F2r9Omvx36bJ+dTte7o/ShxOlilLv25zcbf864BdSdDP39RExdTtngBGiAwAAAAAAAEAoys+WVj4ltegiPbiuMtDud5+06EZp+SPSzz+XqrXM9urwZmnp/VL+8Zq3O7lbCo+WbvgvKazxR9TWhr4AAAAAAAAAAEAD2PGuVFYsDZ7pWhF+WX+pz51S9jfS0S9qP05xnrTsF9I7Y6WSPKnjIO/bGoaU/a3UuntQBOgSIToAAAAAAAAAhKbDm+1fOw9zX9fp4rKDG2s/zvlD0tfvS10zpV/8U+p0nfdtzx2QbAX177feAIIj6gcAAAAAAAAAmOvM95I1XGqW4r6uRafKbWqT0E6askZKyah925O77F8tVunP90lHNksXcqQ23aVrp0tX3Onz5V8qVKIDAAAAAAAAQCgqzpWiEiVrmPs6R3uX4tzajxPf2rcAXZJOXAzRv3pXKjwtXXGX1ONm6cwP0kdTpVVzfDvOJUQlOgAAAAAAAAA0FQv6S2d/qH27X+fa26rEJ3teHxZp/1pWbN61SZIMqVlHafhs6eqJlYvPH5LeGi1t+r2UNkZKHWzyeeuPEB0AAAAAAAAAmormqb5vGx4tlds8r3Msj4z3+5JcZP7a/q+65qnS9U9Ky2dKuz4kRAcAAAAAAAAABMDPlvq+bUxze0sVw5AsFtd1jjYu0YnmXVttOvSzfz1/6NKd0wf0RAcAAAAAAACAUNQqzV5xnnvMfd25g/avrdPNO19FhXR8h3TwM8/rbUX2rxEx5p3TBIToAAAAAAAAABCKHC1TDm50X3dgvf1rx4Hmnc9ikf50i/SncVJ+tvv6w/+0f3VUpDcShOgAAAAAAAAAEIquuFsKi5I2zpOKzlUuP7Zd+map1LaP1PFa885nsUi9bpVkSKvn2CvTHU7ulj5/2d6D/ep7zDunCeiJDgAAAAAAAAChKKmjNOJpe6D96iCp1+1SSZ60e6lkjZDGzXfd/sQu6bu/2/e7emL9zjnyGXvF+a4PpVN7pE7DpLzj9uMa5dIdi6TEdv4/NxMRovuovLxcktSzZ88GvhIAAAAAAAAAocyRVZpi8CNSYntp80Lpy7ekqASpy/XS9U/ZK9GrOrlb2jBXunxI/UP0uFbS1H9In/1O2rtc2vq6/ZzdRklDH5M69PX/OZnMYhiG0dAXEQzS0+0N9MPCwhr4SgAAAAAAAACEMkeIvm/fvga+ktBAiA4AAAAAAAAAgBdMLAoAAAAAAAAAgBeE6AAAAAAAAAAAeMHEok3UX/7yFy1evFiHDh1SdHS0Bg8erFmzZqlDhw4+7b9v3z79/ve/186dO1VUVKS0tDRNmTJFN9xwQ0DOh+Dh72u9cuVKLV68WN9++61sNps6dOigMWPGaNq0aYqNjXXZ9uWXX9Zrr73m8TixsbHasWOH388HjYc/99ahQ4c0evRor+vnz5+vMWPGmHY+BJf6vtazZ8/WsmXLaj1+1R6EvG+FrlmzZmn79u3auHGjz/sw3oIv6nNvMd6CL+p6bzHegi/qcl8x1kJtCgsL9frrr2vVqlX68ccfFRERoZ49e2ry5MnKzMz06RiMt2AWQvQm6He/+53efPNNde3aVRMmTNCJEyeUlZWlzz//XH/+85+VkpJS4/67d+/WpEmTJEnjxo1TdHS0VqxYoYcfflhPP/207rnnHlPPh+Dh72u9cOFCLViwQElJSRozZozi4+O1detWvfbaa/rss8+0ePFil1/s9u7dK4vFooceekgWi8XlWBEREQF5jmgY/t5be/fulSSNHDlSPXr0cFvftWtXU8+H4OHPa52Zmel1sLx582Zt375dAwYMcFnO+1Zoeu2115SVlaXk5GSf92G8BV/U595ivAVf1OfeYryF2tT1vmKshZoUFBRowoQJ2rdvn3r16qUJEyYoPz9fq1at0vTp0/XLX/5S06ZNq/EYjLdgKgNNyp49e4y0tDRj/PjxRklJiXP56tWrjbS0NGPatGm1HuOWW24xevXqZezdu9e57OzZs0ZmZqbRp08f4+TJk6aeD8HB39f6wIEDRo8ePYxhw4YZp0+fdi4vLy83Zs+ebaSlpRnz58932WfIkCFGZmamuU8EjY4Z7yP/8z//Y6SlpRnbtm27JOdDcAjUa33kyBGjb9++xpAhQ4yzZ8+6rON9K7QUFxcbc+bMMdLS0oy0tDRj6NChPu/LeAs1qe+9xXgLtfHnfYvxFrzx577yhLEWDMMwXn75ZSMtLc145plnjIqKCufykydPGoMHDzZ69OhhHDp0qMZjMN6CmeiJ3sS89957kqQZM2YoMjLSuTwzM1PXXHON1q9fr+zsbK/7b9++XXv37tWYMWPUvXt35/IWLVrooYceUklJicvHrfw9H4KHv6/1qlWrVF5ervvvv1+tWrVyLrdarXrkkUckSevWrXMuP3funE6dOuWxygVNixnvI46KlKrvW4E8H4JDIF5rwzD0xBNPqKCgQM8//7xatGjhXMf7VmhZu3atbrzxRn344YcaNmxYnfZlvIWa+HNvMd5CTfy5tyTGW/DM3/uqOsZacFixYoUsFosee+wxl08dJCcna/z48SovL9eGDRu87s94C2YjRG9ivvzyS4WHh6t///5u6wYOHCjDMLRlyxav+2/fvt25raf9Jbns7+/5EDz8fa379u2rRx55RIMHD3Zb5/gBVVhY6Fzm+Lhoenq6v5eORs6M95G9e/fqsssuU3x8/CU5H4JDIF7rv/3tb9qxY4dGjhyp66+/3mUd71uhZenSpSosLNSzzz6r119/vU77Mt5CTfy5txhvoSb+3FsS4y145u99VR1jLThMmjRJjz76qBITE93WefqZVh3jLZiNnuhNSHl5uQ4dOqQOHTq4/NXMoWPHjpKkAwcOeD3G/v37XbatKjk5WVFRUc79zTgfgoMZr3VGRoYyMjI8rvv0008luQ6GHAOkwsJCTZs2Tbt27VJxcbF69eqladOmaejQofV+Pmg8zLi3HBUp/fr109y5c/WPf/xDJ0+eVPv27TVu3Dg9+OCDzmPzvhU6AvFa22w2vfTSS7JarXr88cfd1vO+FVruvfde/fd//7dPYVJ1jLdQE3/uLcZbqIk/9xbjLXjjz31VHWMtVDVx4kSPyw3D0OrVqyXV/AcVxlswG5XoTUhBQYEMw1CzZs08rk9ISJAk5efnez1GXl6eJHk8hsViUXx8vHN/M86H4BDI1/r48eNasGCBJGnChAnO5Y4B0qJFiyRJt99+u4YOHaqdO3fqgQcecH7UCsHNjHtrz549kuyVBuvXr9eIESN02223yWazacGCBbrvvvtks9lMOx+CQyBe648//ljZ2dkaNWqUOnfu7Lae963QMmDAgHoHBoy3UBN/7i1vGG9B8u/eYrwFb8x8z2KsBV+8//77+vrrr5WSklLjH04Yb8FsVKI3IUVFRZLk8a9mVZeXlJT4dYycnBzTzofgEKjX+tSpU7r//vt1/vx53X333S4fswoPD1eHDh30/PPPa8iQIc7l3377rSZOnKjf/va3Gjp0qMe/KiN4mHFvFRQUKDU1Vddee63mzJmj8PBw57GnT5+uTZs26Y033tCMGTN43wohgXit3333XUnSAw884HE971vwFeMtXEqMt2AGxlu4FBhroTZZWVn6zW9+o/DwcM2dO1cRERFet2W8BbNRid6EREVFSZJKS0s9rndUBsTGxvp1DMf+ZpwPwSEQr/WBAwc0fvx4HTx4UCNHjtScOXNc1r/44otau3aty+BIknr16qV7771XZWVlysrKqsvTQCNkxr01ZswYrVy5Us8995zzFzrHPs8++6wkafny5aadD8HB7Nf666+/1vfff6/evXurT58+HrfhfQu+YryFS4XxFszCeAuBxlgLtXn//ff12GOPSbLfC556l1fFeAtmI0RvQuLj42W1Wr1+vMSx3PExFE8cH11xfOylKsMwVFBQ4NzfjPMhOJj9Wm/ZskU//elPdezYMY0bN07z58+v8S/I1TkGVUePHvV5HzROgX4fSU1NVWJiovNe4X0rdJj9Wq9atUqSNHbs2HpdD+9bqIrxFi4Fxlu4VBhvwQyMteBNRUWF5s6dq+eee04RERGaP3++br755lr3Y7wFsxGiNyGRkZHq2LGjjh8/7vGvZ0eOHJEkde3a1esxunTpIsnzD57s7GyVlJQ4tzHjfAgOZr7Wy5cv19SpU5Wbm6upU6dq3rx5br/Q2Ww27dq1Szt37vR4jAsXLkiSoqOj6/hM0NiYcW8dPHhQmzdv9jgze0VFhUpKSpyVBbxvhQ6zX+u1a9dKkm688UaP63nfQl0w3kKgMd6C2RhvIdAYa8ETm82mmTNnatGiRUpKStLbb7+tUaNG+bQv4y2YjRC9icnIyFBpaam++uort3WbN2+WxWJR3759a9xfsleuVLdp0yZJUr9+/Uw7H4KHGa91VlaWfvWrX6m8vFy//vWv9cQTT8hisbhtV1hYqLvuuktTpkzx+ANs27ZtkuT1Y34ILv7eWy+++KImT56sDRs2uK3btWuXSkpKdMUVV5h2PgQPs17r3NxcHThwQCkpKWrfvr3HbXjfQl0w3kIgMd5CIDDeQiAx1oInFRUVmjlzplatWqXLLrtMS5YsqbWFS1WMt2A2QvQm5t/+7d8kSS+//LKKi4udy9esWaMvvvhCI0aMUNu2bb3u37dvX3Xu3FmffPKJdu3a5Vx+7tw5vfrqq4qKitIdd9xh2vkQPPx9rQ8ePKinnnpKFRUVeuGFFzR+/Hiv2zZv3lyDBg1SQUGBFi5c6LJu06ZNWrp0qdq2basxY8b4+azQGPh7b910002SpIULF6qgoMC5/Pz583r++eclSffee69p50PwMOu1/uabbyRJV155pddteN9CXTDeQqAw3kKgMN5CIDHWgievv/661q5dq/bt2+v9999X586d67Q/4y2YzWIYhtHQFwFzPf/883rvvfeUmpqqkSNHKjs7WytWrFDz5s31wQcfKCUlRZK0detWffHFF+rRo4cyMzOd+3/55Ze6//77ZbFYdPPNNys+Pl5ZWVk6deqUnnnmGU2cOLFe50Pw8+feeuyxx/TJJ5+oQ4cOuu222zwePz4+Xvfdd58k6dChQ5owYYLOnj2r/v3764orrtDhw4e1bt06RUdH66233uKvwE2IP/eWYRh65JFHtGrVKrVp00ajRo2SzWbT+vXrdfr0aU2ePFlPPvlkvc6H4Ofvz0TJPonRc889p4cfflgzZszwei7et0Jbenq6kpOTtXHjRpfljLfgr7rcW4y3UBd1ubcYb8FXdf15KDHWgrvc3FwNHz5cRUVFGjlypHr06OFxu/79+2vgwIGMt3BJEKI3QYZh6L333tOHH36oQ4cOKSkpSQMGDNDMmTNd/odfsGCBFi5cqNtuu01z5851Ocbu3bv1+9//3vkxlm7dumnKlCkee0/5ej4EP3/urSFDhuj06dM1Hr/6YCs7O1t/+MMftGHDBp05c0ZJSUkaNGiQpk+frtTU1IA8RzQMf9+3KioqtGTJEi1dulT79+9XWFiYunfvrp/97GfOyqn6nA/Bz4yfiQsXLtSCBQv0m9/8xqVaxRPet0KXt9CA8Rb8VZd7i/EW6qKu71uMt+CL+vw8ZKyF6j7//HNNmTKl1u1+/vOfa9asWYy3cEkQogMAAAAAAAAA4AU90QEAAAAAAAAA8IIQHQAAAAAAAAAALwjRAQAAAAAAAADwghAdAAAAAAAAAAAvCNEBAAAAAAAAAPCCEB0AAAAAAAAAAC8I0QEAAAAAAAAA8IIQHQAAAAAAAAAALwjRAQAAAAAAAADwghAdAAAAIefo0aNatmyZy7IRI0YoPT1dZWVlDXRV3s2ePVvp6enOf++8847P+65Zs8Zl3/HjxwfuQgEAAIAmKLyhLwAAAAC4lL777jvdeeedGj16tG677Tbn8kmTJik/P19Wa+OtM7nllluUkpKiq666yud9OnXqpBkzZigvL0/vvvtu4C4OAAAAaKII0QEAABBScnNzZbPZ3JZPnjz50l9MHd16660aNGhQnfbp0qWLHn74YR07dowQHQAAAKiHxltmAwAAAAAAAABAAyNEBwAAQMiYPXu2Jk2aJElavny50tPT9dFHH0ly74m+detWZ//xFStW6Pbbb9cVV1yhIUOG6MUXX5TNZtOJEyc0a9YsZWRkaMCAAZo2bZoOHz7sdt6TJ0/qmWee0bBhw9S7d29dd911mjNnjrKzs/16PsXFxXrppZc0btw4XXXVVerfv7/uueceZWVl+XVcAAAAAJVo5wIAAICQkZmZKUlatmyZ0tLSdMMNN6hHjx417vPxxx/rhx9+0OjRo5WRkaFPP/1Ub7/9ts6dO6dNmzapbdu2uuOOO/Ttt99q/fr1OnLkiJYvX67wcPtQ+4cfftCkSZN0/vx5DR8+XF26dNGRI0e0dOlSrV27VosXL1anTp3q9XweffRRrVu3TkOGDNF1112ngoICrVy5UrNmzVJ+fr7uvvvueh0XAAAAQCVCdAAAAISMzMxMJSQkaNmyZUpPT9fDDz9c6z579uzRq6++qhEjRkiS7rrrLo0dO1Z//etfNW7cOM2bN08Wi0WGYWjixInavn27du/erauvvlqS9Ktf/Uo5OTl64403dN111zmPu2HDBj344IN68skn9cEHH9T5uXz//fdat26dfvKTn2jevHnO5VOnTtWNN96ot99+mxAdAAAAMAHtXAAAAIAadOnSxRmgOx4nJSVJsgfWFotFkmSxWJzB+Y8//ihJ2rVrl/bs2aMbbrjBJUCXpGHDhmnw4MHasWOH9u/fX+frMgxDknTgwAGdO3fOuTwlJUUrVqzQxx9/XOdjAgAAAHBHJToAAABQg9TUVLdlcXFxysnJUceOHV2WR0dHS5JsNpskaffu3ZKks2fPasGCBW7HKSoqkiR9++236tKlS52uKz09XRkZGdq2bZuGDRumfv36adCgQbruuuvUvXv3Oh0LAAAAgHeE6AAAAEANYmNjva6LjIyscd+8vDxJ0rZt27Rt2zav2+Xm5tbr2t58800tWrRIn3zyiTZv3qzNmzfrpZdeUmpqqp566ikNGzasXscFAAAAUIkQHQAAAAiQuLg4SdLjjz+uBx54wPTjx8TE6KGHHtJDDz2kkydPasuWLVqzZo1Wr16t6dOn69NPP9Vll11m+nkBAACAUEJPdAAAAIQURw/zS6Fnz56S7L3RPXnvvfe0cOFCHTt2rM7H3rlzp+bOnaudO3dKktq2batbb71VCxcu1O23367S0lLt2LGj3tcOAAAAwI4QHQAAACElPNz+YcyysrKAn6tv377q3LmzVq9erdWrV7us2759u1544QW98847at68eZ2PXVBQoEWLFmnBggWqqKhwLjcMwzmxaUpKin9PAAAAAADtXAAAABBa2rVrJ0nauHGjXnzxRY0cOVL9+/cPyLmsVqvmzZun++67TzNmzNCQIUOUnp6uEydOaPXq1TIMQy+88IKz7UtdDBo0SMOHD9f69es1btw4DRo0SGFhYdqyZYv27t2r0aNH66qrrjL/SQEAAAAhhkp0AAAAhJR27drpscceU3R0tBYvXqxNmzYF9Hy9e/fWRx99pDvvvFP/+te/9O6772r79u0aPny4lixZolGjRtXruFarVa+88ooef/xxhYWFadmyZfrggw9ksVj05JNP6qWXXjL5mQAAAAChyWIYhtHQFwEAAADAu9mzZ2vZsmVatGiRBg0aVK9jHDt2TCNHjlTfvn21ZMkSk68QAAAAaLqoRAcAAAAAAAAAwAt6ogMAAABB4q9//au2b9+uoUOH+tzvfP/+/crKylJeXl5gLw4AAABoogjRAQAAgCDx8ccfS5ISEhJ8DtEPHjyohQsXBvCqAAAAgKaNnugAAAAAAAAAAHhBT3QAAAAAAAAAALwgRAcAAAAAAAAAwAtCdAAAAAAAAAAAvCBEBwAAAAAAAADAC0J0AAAAAAAAAAC8IEQHAAAAAAAAAMALQnQAAAAAAAAAALwgRAcAAAAAAAAAwAtCdAAAAAAAAAAAvPj/y3noNx+64QYAAAAASUVORK5CYII=", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# visualize a random data augmentation\n", "Sample = trainset.__getitem__(random.randint(0,len(trainset)))\n", "t = np.linspace(0, Sample.shape[1]-1, Sample.shape[1])/freq\n", "SampleAug = Augmenter(Sample)\n", "RandChan= random.randint(0,Chan-1)\n", "\n", "fig, ax1 = plt.subplots()\n", "color = 'tab:blue'\n", "ax1.set_xlabel('time [s]', fontsize=15)\n", "ax1.set_ylabel('[uV]', color=color, fontsize=15)\n", "ax1.plot(t, Sample[RandChan,:], color=color)\n", "ax1.tick_params(axis='y', labelcolor=color, labelsize=15)\n", "ax1.tick_params(axis='x', labelsize=15)\n", "ax2 = ax1.twinx()\n", "color = 'tab:orange'\n", "ax2.set_ylabel('[ ]', color=color, fontsize=15) # we already handled the x-label with ax1\n", "ax2.plot(t, SampleAug[RandChan,:],color=color, linewidth=2.5)\n", "ax2.tick_params(axis='y', labelsize=15, labelcolor=color)\n", "plt.title('Same random channel from one sample: augmented version', fontsize=20)\n", "fig.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "a93fd97f-e732-4a2e-82be-eaac4c0db58a", "metadata": {}, "source": [ "### Define pretraining model and other training objects\n", "\n", "Now we need to define the pretraining model. To do that, one must:\n", "\n", "1. instantiate an nn.Module defining the encoder (backbone)\n", "2. instantiate the right SSL module, giving the encoder and the network head's spec.\n", "\n", "For now, let's use a simple EEGNet with default parameters, and SimCLR as the SSL algorithm.\n", "\n", "
\n", "NOTE 1 \n", " \n", "each model in the models module have an extra class with only the encoder with the name modelnameEncoder (e.g., EEGNetEncoder). This will make model creation much easier. \n", "\n", "
\n", "\n", "
\n", "NOTE 2 \n", " \n", "each model in the ssl module can accept a list or a nn.Module to create the network head. In case of a list, the head will be a sequence of dense layer with input and output size equal to the values of the list. Batchnorm and activation are based on the original works.\n", "\n", "
\n", "\n", "
\n", "WARNING \n", " \n", "Remember to check if the encoder output size matches the head input size. All modules in the ssl class doesn't check that.\n", "\n", "
" ] }, { "cell_type": "code", "execution_count": 7, "id": "144a6e7e-98dd-4c4c-8884-295c7b523103", "metadata": {}, "outputs": [], "source": [ "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n", "\n", "# Encoder\n", "NNencoder= selfeeg.models.ShallowNetEncoder(Chans=Chan, F=8)\n", "# It's suggested to copy the random initialization for embedding analysis\n", "NNencoder2= copy.deepcopy(NNencoder)\n", "\n", "# SSL model\n", "head_size=[ 88, 64, 64]\n", "SelfMdl = selfeeg.ssl.SimCLR(\n", " encoder=NNencoder, projection_head=head_size).to(device=device)\n", "\n", "# loss (fit method has a default loss based on the SSL algorithm\n", "loss=selfeeg.losses.simclr_loss\n", "loss_arg={'temperature': 0.5}\n", "\n", "# earlystopper\n", "earlystop = selfeeg.ssl.EarlyStopping(\n", " patience=25, min_delta=1e-05, record_best_weights=True)\n", "# optimizer\n", "optimizer = torch.optim.Adam(SelfMdl.parameters(), lr=1e-3)\n", "# lr scheduler\n", "scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.97)" ] }, { "cell_type": "markdown", "id": "8e211933-5c74-4acc-8cf3-99730f6c8fcf", "metadata": {}, "source": [ "### Pretrain the model\n", "\n", "Each SSL algorithm has an already implemented fit method, similar to scikitlearn or Keras. Of course it's not complete as the fit of bigger framewoks, but it certainly save you lots of lines of code and help you monitorate the training." ] }, { "cell_type": "code", "execution_count": 8, "id": "228a4a30-365f-4713-a6db-5617edfaabdb", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "epoch [1/5]\n", " val 20/20: 100%|███████████████| 161/161 [ 8.59 Batch/s, train_loss=4.72287, val_loss=4.97601] \n", "epoch [2/5]\n", " val 20/20: 100%|███████████████| 161/161 [ 8.28 Batch/s, train_loss=4.46746, val_loss=4.84657] \n", "epoch [3/5]\n", " val 20/20: 100%|███████████████| 161/161 [ 1.97 Batch/s, train_loss=4.47779, val_loss=4.97520] \n", "epoch [4/5]\n", " val 20/20: 100%|███████████████| 161/161 [ 2.03 Batch/s, train_loss=4.34311, val_loss=4.78815] \n", "epoch [5/5]\n", " val 20/20: 100%|███████████████| 161/161 [ 3.59 Batch/s, train_loss=4.34365, val_loss=4.78576] \n" ] } ], "source": [ "loss_info = SelfMdl.fit(\n", " train_dataloader = trainloader,\n", " augmenter = Augmenter,\n", " epochs = 5,\n", " optimizer = optimizer,\n", " loss_func = loss,\n", " loss_args = loss_arg,\n", " lr_scheduler = scheduler,\n", " EarlyStopper = earlystop,\n", " validation_dataloader = valloader,\n", " verbose = True,\n", " device = device,\n", " return_loss_info = True\n", ")" ] }, { "cell_type": "markdown", "id": "08381a75-fbe4-4153-b482-1934eaa2d2bb", "metadata": {}, "source": [ "## Fine-tuning Phase\n", "\n", "Now that the encoder is pretrained, let's perform fine-tuning. To do that, we need to recreate the right dataloaders and models. After that, the **selfeeg** library provides a **fine-tuning** function that is similar to the fit method." ] }, { "cell_type": "markdown", "id": "02e8dc32-4b9a-45cd-9615-77f2ff00c3a8", "metadata": {}, "source": [ "### Define fine-tuning dataloaders\n", "\n", "This phase is basically the same as the previous one. The only differences are:\n", "\n", "1. We are using the fine-tuning data\n", "2. We are creating a test set for evaluation\n", "3. We need to extract a label from each sample.\n", "\n", "The used classes and methods are the same already used from the dataloading module" ] }, { "cell_type": "code", "execution_count": 9, "id": "576a2c11-b6e7-4898-91bb-c7a19d096129", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "train ratio: 0.72\n", "validation ratio: 0.08\n", "test ratio: 0.20\n", "\n", "train labels ratio: 0.0=0.620, 1.0=0.380, \n", "val labels ratio: 0.0=0.627, 1.0=0.373, \n", "test labels ratio: 0.0=0.617, 1.0=0.383, \n", "\n" ] } ], "source": [ "# Extract only the samples for fine-tuning\n", "filesFT= EEGsplit.loc[EEGsplit['split_set']==2, 'file_name'].values\n", "EEGlenFT= EEGlen.loc[EEGlen['file_name'].isin(filesFT)]\n", "EEGlenFT = EEGlenFT.reset_index().drop(columns=['index'])\n", "labels = classes[ EEGsplit[EEGsplit['split_set']==2].index.tolist()]\n", "\n", "# split the fine-tuning data in train-test-validation\n", "EEGsplitFT = dl.get_eeg_split_table(\n", " partition_table=EEGlenFT,\n", " test_ratio = 0.2,\n", " val_ratio= 0.1,\n", " val_ratio_on_all_data=False,\n", " stratified=True,\n", " labels=labels,\n", " split_tolerance=0.001,\n", " perseverance=10000\n", ")\n", "\n", "# TRAINING DATALOADER\n", "trainsetFT = dl.EEGDataset(\n", " EEGlenFT, EEGsplitFT, [freq, window, overlap], 'train', supervised=True,\n", " label_on_load=True, load_function=loadEEG, optional_load_fun_args=[True]\n", ")\n", "trainsamplerFT = dl.EEGSampler(trainsetFT, batchsize, workers)\n", "trainloaderFT = DataLoader(\n", " dataset = trainsetFT, batch_size= batchsize, sampler=trainsamplerFT, num_workers=workers)\n", "\n", "# VALIDATION DATALOADER\n", "valsetFT = dl.EEGDataset(\n", " EEGlenFT, EEGsplitFT, [freq, window, overlap], 'validation', supervised=True,\n", " label_on_load=True, load_function=loadEEG, optional_load_fun_args=[True]\n", ")\n", "valloaderFT = DataLoader(\n", " dataset=valsetFT, batch_size=batchsize, num_workers=workers, shuffle=False)\n", "\n", "#TEST DATALOADER\n", "testsetFT = dl.EEGDataset(\n", " EEGlenFT, EEGsplitFT, [freq, window, overlap], 'test', supervised=True,\n", " label_on_load=True, load_function=loadEEG, optional_load_fun_args=[True]\n", ")\n", "testloaderFT = DataLoader(dataset = testsetFT, batch_size= batchsize, shuffle=False)\n", "\n", "dl.check_split(EEGlenFT, EEGsplitFT, labels)" ] }, { "cell_type": "markdown", "id": "2abb4ed0-f149-4a62-9d1d-ba32c0895c9f", "metadata": {}, "source": [ "### Define fine-tuning model and other training objects\n", "\n", "Remember that in this phase you need to transfer the pretrained encoder" ] }, { "cell_type": "code", "execution_count": 10, "id": "5bb89f11-7ade-4722-b3b5-78f24cc34665", "metadata": {}, "outputs": [], "source": [ "FinalMdl = selfeeg.models.ShallowNet(\n", " nb_classes = 2, Chans = Chan, Samples = int(freq*window), F=8)\n", "\n", "# Transfer the pretrained backbone and move the final model to the right device\n", "SelfMdl.train() \n", "SelfMdl.to(device='cpu')\n", "\n", "FinalMdl.encoder = SelfMdl.get_encoder()\n", "FinalMdl.train()\n", "FinalMdl.to(device=device)\n", "\n", "# DEFINE LOSS\n", "def loss_fineTuning(yhat, ytrue):\n", " ytrue = ytrue + 0.\n", " yhat = torch.squeeze(yhat)\n", " return F.binary_cross_entropy_with_logits(yhat, ytrue)\n", "\n", "# DEFINE EARLYSTOPPER\n", "earlystopFT = selfeeg.ssl.EarlyStopping(\n", " patience=10, min_delta=1e-03, record_best_weights=True)\n", "\n", "# DEFINE OPTIMIZER \n", "optimizerFT = torch.optim.Adam(FinalMdl.parameters(), lr=1e-3)\n", "schedulerFT = torch.optim.lr_scheduler.ExponentialLR(optimizerFT, gamma=0.97)" ] }, { "cell_type": "markdown", "id": "069787cc-1a7a-4698-b777-fdd5918c3c33", "metadata": {}, "source": [ "### Fine-tuning\n", "\n", "Fine-tuning can be easily performed with the fine-tuning method.\n", "\n", "
\n", "NOTE 1 \n", " \n", "it is better to first pretrain only the new head, and then update all model's weights. However, EEGNet is small and this is a simple example, so we directly fine-tune all the network.\n", "\n", "
" ] }, { "cell_type": "code", "execution_count": 11, "id": "cd036c7a-3de5-4898-a8b7-71869429f488", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "epoch [1/10]\n", " val 4/4: 100%|███████████████| 33/33 [101.80 Batch/s, train_loss=0.74088, val_loss=0.52754] \n", "epoch [2/10]\n", " val 4/4: 100%|███████████████| 33/33 [107.47 Batch/s, train_loss=0.51391, val_loss=0.42154] \n", "epoch [3/10]\n", " val 4/4: 100%|███████████████| 33/33 [112.56 Batch/s, train_loss=0.41865, val_loss=0.26677] \n", "epoch [4/10]\n", " val 4/4: 100%|███████████████| 33/33 [106.99 Batch/s, train_loss=0.26489, val_loss=0.18554] \n", "epoch [5/10]\n", " val 4/4: 100%|███████████████| 33/33 [113.98 Batch/s, train_loss=0.18971, val_loss=0.11007] \n", "epoch [6/10]\n", " val 4/4: 100%|███████████████| 33/33 [115.99 Batch/s, train_loss=0.12763, val_loss=0.07227] \n", "epoch [7/10]\n", " val 4/4: 100%|███████████████| 33/33 [108.40 Batch/s, train_loss=0.14895, val_loss=0.05631] \n", "epoch [8/10]\n", " val 4/4: 100%|███████████████| 33/33 [109.08 Batch/s, train_loss=0.09240, val_loss=0.04787] \n", "epoch [9/10]\n", " val 4/4: 100%|███████████████| 33/33 [114.48 Batch/s, train_loss=0.11739, val_loss=0.03897] \n", "epoch [10/10]\n", " val 4/4: 100%|███████████████| 33/33 [114.82 Batch/s, train_loss=0.08307, val_loss=0.03281] \n" ] } ], "source": [ "finetuning_loss=selfeeg.ssl.fine_tune(\n", " model = FinalMdl,\n", " train_dataloader = trainloaderFT,\n", " epochs = 10,\n", " optimizer = optimizerFT,\n", " loss_func = loss_fineTuning, \n", " lr_scheduler = schedulerFT,\n", " EarlyStopper = earlystopFT,\n", " validation_dataloader = valloaderFT,\n", " verbose = True,\n", " device = device,\n", " return_loss_info = True\n", ")" ] }, { "cell_type": "markdown", "id": "93c66d5a-e56b-4bc3-9bee-0943880ed454", "metadata": {}, "source": [ "## Evaluate fine-tuned model\n", "\n", "Now you can evaluate your model in whatever method you prefer. Here is a simple example with the classification report from sklearn." ] }, { "cell_type": "code", "execution_count": 12, "id": "076b9201-f10d-460d-8722-50984f66b778", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Results of trivial Example\n", "\n", " precision recall f1-score support\n", "\n", " 0.0 1.00 1.00 1.00 58\n", " 1.0 1.00 1.00 1.00 70\n", "\n", " accuracy 1.00 128\n", " macro avg 1.00 1.00 1.00 128\n", "weighted avg 1.00 1.00 1.00 128\n", "\n" ] } ], "source": [ "from sklearn.metrics import classification_report\n", "nb_classes=2\n", "FinalMdl.eval()\n", "ytrue=torch.zeros(len(testloaderFT.dataset))\n", "ypred=torch.zeros_like(ytrue)\n", "cnt=0\n", "for i, (X, Y) in enumerate(testloaderFT):\n", " X=X.to(device=device)\n", " ytrue[cnt:cnt+X.shape[0]]= Y \n", " with torch.no_grad():\n", " yhat = torch.sigmoid(FinalMdl(X)).to(device='cpu')\n", " ypred[cnt:cnt+X.shape[0]] = torch.squeeze(yhat) \n", " cnt += X.shape[0]\n", "\n", "print('Results of trivial Example\\n')\n", "print(classification_report(ytrue,ypred>0.5))" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.11.6" } }, "nbformat": 4, "nbformat_minor": 5 }