├── image.jpg
├── README.md
└── mel spectrogram.ipynb
/image.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/OmarMedhat22/Sound-Classification-Mel-Spectrogram/HEAD/image.jpg
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # Sound-Classification-Mel-Spectrogram
2 |
3 | 
4 |
5 | ## Definition
6 |
7 | This project classifies sound signals from different environmental classes in the ESC-10 dataset. the above photo summarizes the model steps:
8 | 1. The model read all the signals of different classes and assign a label number to each class.
9 | 2. The Mel Spectrogram are extracted from the time domain.
10 | 3. Full Convolutional Neural Network(CNN) is defined and used to classify 10 different classes of ESC-10 dataset.
11 | ## Dependencies
12 | ### To run this project you will need to:
13 | 1. download the ESC-10 dataset from this link: [ESC-10](https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/YDEPUT)
14 |
15 | 2. change the directory name that contains the dataset to the name in the notebook file in these three lines:
16 | ```html
17 | #here my directory name is "dataset"
18 |
19 | data, samplerate = librosa.load("dataset/dog/1-30344-A.wav", sr=44000)
20 |
21 | for filepath in glob.iglob('dataset/*'):
22 |
23 | for j in glob.iglob('dataset/'+i+'/*'):
24 |
25 | ```
26 | ### install this libraries:
27 | - numpy
28 | - keras
29 | - matplotlib
30 | - librosa
31 | - pylab
32 | - glob
33 | - tensorflow
34 |
--------------------------------------------------------------------------------
/mel spectrogram.ipynb:
--------------------------------------------------------------------------------
1 | {
2 | "cells": [
3 | {
4 | "cell_type": "code",
5 | "execution_count": 1,
6 | "metadata": {},
7 | "outputs": [
8 | {
9 | "name": "stderr",
10 | "output_type": "stream",
11 | "text": [
12 | "Using TensorFlow backend.\n"
13 | ]
14 | }
15 | ],
16 | "source": [
17 | "import numpy as np\n",
18 | "import matplotlib.pyplot as plt\n",
19 | "import librosa.display\n",
20 | "import pylab\n",
21 | "from sklearn.model_selection import train_test_split\n",
22 | "from sklearn.metrics import accuracy_score\n",
23 | "import librosa \n",
24 | "import glob\n",
25 | "import tensorflow as tf\n",
26 | "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
27 | "from tensorflow.keras.models import Sequential\n",
28 | "from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten\n",
29 | "from tensorflow.keras.layers import Conv2D, MaxPooling2D\n",
30 | "from keras.utils import to_categorical\n",
31 | "from keras import optimizers\n",
32 | "import keras\n",
33 | "\n",
34 | "%matplotlib inline\n"
35 | ]
36 | },
37 | {
38 | "cell_type": "code",
39 | "execution_count": 2,
40 | "metadata": {},
41 | "outputs": [],
42 | "source": [
43 | "audio_data = []\n",
44 | "labels = []\n",
45 | "sampling_rate = []\n",
46 | "file_names = []\n"
47 | ]
48 | },
49 | {
50 | "cell_type": "code",
51 | "execution_count": 3,
52 | "metadata": {},
53 | "outputs": [
54 | {
55 | "name": "stdout",
56 | "output_type": "stream",
57 | "text": [
58 | "(220000,)\n",
59 | "44000\n"
60 | ]
61 | }
62 | ],
63 | "source": [
64 | "#samplerate, data = wavfile.read(\"dataset/dog/1-30344-A.wav\")\n",
65 | "data, samplerate = librosa.load(\"dataset/dog/1-30344-A.wav\", sr=44000) # Downsample 44.1kHz to 8kHz\n",
66 | "\n",
67 | "print(data.shape)\n",
68 | "print(samplerate)\n"
69 | ]
70 | },
71 | {
72 | "cell_type": "code",
73 | "execution_count": 4,
74 | "metadata": {},
75 | "outputs": [
76 | {
77 | "name": "stdout",
78 | "output_type": "stream",
79 | "text": [
80 | "5.0\n",
81 | "2.2727272727272726e-05\n",
82 | "176000.0\n"
83 | ]
84 | }
85 | ],
86 | "source": [
87 | "time_sec = (len(data)/samplerate)\n",
88 | "step = time_sec/len(data)\n",
89 | "print(time_sec)\n",
90 | "print(step)\n",
91 | "i=0\n",
92 | "time_divion=[]\n",
93 | "while i<=time_sec-step:\n",
94 | " \n",
95 | " time_divion.append(i)\n",
96 | " i=i+step\n",
97 | "# the fourth second step\n",
98 | "four_sec_step_number = (4*len(time_divion))/time_sec\n",
99 | "print(four_sec_step_number)"
100 | ]
101 | },
102 | {
103 | "cell_type": "code",
104 | "execution_count": 5,
105 | "metadata": {},
106 | "outputs": [
107 | {
108 | "name": "stdout",
109 | "output_type": "stream",
110 | "text": [
111 | "['004 - Baby cry', '005 - Clock tick', '006 - Person sneeze', '007 - Helicopter', '008 - Chainsaw', '009 - Rooster', '010 - Fire crackling', 'dog', 'rain', 'sea']\n",
112 | "the class = 004 - Baby cry, the label = 0\n",
113 | "the class = 005 - Clock tick, the label = 1\n",
114 | "the class = 006 - Person sneeze, the label = 2\n",
115 | "the class = 007 - Helicopter, the label = 3\n",
116 | "the class = 008 - Chainsaw, the label = 4\n",
117 | "the class = 009 - Rooster, the label = 5\n",
118 | "the class = 010 - Fire crackling, the label = 6\n",
119 | "the class = dog, the label = 7\n",
120 | "the class = rain, the label = 8\n",
121 | "the class = sea, the label = 9\n",
122 | "0\n"
123 | ]
124 | }
125 | ],
126 | "source": [
127 | "classes = []\n",
128 | "label_number=0\n",
129 | "audio_data = []\n",
130 | "labels = []\n",
131 | "sampling_rate = []\n",
132 | "file_names = []\n",
133 | "data = []\n",
134 | "noisy_removed=[]\n",
135 | "noise=[]\n",
136 | "for filepath in glob.iglob('dataset/*'):\n",
137 | " \n",
138 | " #print(filepath[9:])\n",
139 | " \n",
140 | " #print(filepath)\n",
141 | " classes.append(filepath[8:])\n",
142 | "\n",
143 | "print(classes)\n",
144 | "\n",
145 | "for i in classes:\n",
146 | " print(\"the class = \"+i+\", the label = \"+str(label_number))\n",
147 | " for j in glob.iglob('dataset/'+i+'/*'):\n",
148 | " #samplerate, data = wavfile.read(j)\n",
149 | " y, s = librosa.load(j, sr=44000) # Downsample 44.1kHz to 8kHz\n",
150 | " #reduced_noise = nr.reduce_noise(audio_clip=y, noise_clip=y, verbose=False)\n",
151 | " #print(s)\n",
152 | " #print(j)\n",
153 | " data.append([y,label_number])\n",
154 | " #noise.append(y)\n",
155 | " #labels.append(label_number)\n",
156 | " \n",
157 | " label_number = label_number + 1\n",
158 | "print(len(labels))\n",
159 | "#print(labels)"
160 | ]
161 | },
162 | {
163 | "cell_type": "code",
164 | "execution_count": 6,
165 | "metadata": {},
166 | "outputs": [],
167 | "source": [
168 | "import random\n",
169 | "\n",
170 | "random.shuffle(data)\n",
171 | "audio_data=[]\n",
172 | "labels=[]\n",
173 | "for i,j in data:\n",
174 | " audio_data.append(i)\n",
175 | " labels.append(j)\n",
176 | " \n"
177 | ]
178 | },
179 | {
180 | "cell_type": "code",
181 | "execution_count": 7,
182 | "metadata": {},
183 | "outputs": [
184 | {
185 | "name": "stdout",
186 | "output_type": "stream",
187 | "text": [
188 | "1\n"
189 | ]
190 | }
191 | ],
192 | "source": [
193 | "print(labels[0])"
194 | ]
195 | },
196 | {
197 | "cell_type": "code",
198 | "execution_count": 8,
199 | "metadata": {},
200 | "outputs": [
201 | {
202 | "data": {
203 | "text/plain": [
204 | "Text(0.5,1,'signal in real time')"
205 | ]
206 | },
207 | "execution_count": 8,
208 | "metadata": {},
209 | "output_type": "execute_result"
210 | },
211 | {
212 | "data": {
213 | "image/png": "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\n",
214 | "text/plain": [
215 | ""
216 | ]
217 | },
218 | "metadata": {
219 | "needs_background": "light"
220 | },
221 | "output_type": "display_data"
222 | }
223 | ],
224 | "source": [
225 | "#save_path='dog_time.jpg'\n",
226 | "plt.plot(time_divion[0:192000],audio_data[164][0:192000])\n",
227 | "#plt.show()\n",
228 | "plt.title('signal in real time')\n",
229 | "#pylab.savefig(save_path, bbox_inches=None, pad_inches=0)\n",
230 | "#pylab.close()\n"
231 | ]
232 | },
233 | {
234 | "cell_type": "code",
235 | "execution_count": 9,
236 | "metadata": {},
237 | "outputs": [
238 | {
239 | "name": "stdout",
240 | "output_type": "stream",
241 | "text": [
242 | "(128, 430)\n",
243 | "7\n"
244 | ]
245 | }
246 | ],
247 | "source": [
248 | "example = random.randint(0,400)\n",
249 | "mel_feat = librosa.feature.melspectrogram(y=audio_data[164],sr=44000)\n",
250 | "power = librosa.power_to_db(mel_feat,ref=np.max)\n",
251 | "print(power.shape)\n",
252 | "print(labels[example])"
253 | ]
254 | },
255 | {
256 | "cell_type": "code",
257 | "execution_count": 10,
258 | "metadata": {},
259 | "outputs": [
260 | {
261 | "data": {
262 | "text/plain": [
263 | "113"
264 | ]
265 | },
266 | "execution_count": 10,
267 | "metadata": {},
268 | "output_type": "execute_result"
269 | }
270 | ],
271 | "source": [
272 | "example"
273 | ]
274 | },
275 | {
276 | "cell_type": "code",
277 | "execution_count": 11,
278 | "metadata": {},
279 | "outputs": [
280 | {
281 | "data": {
282 | "text/plain": [
283 | "Text(0.5,1,'Mel Spectrogram')"
284 | ]
285 | },
286 | "execution_count": 11,
287 | "metadata": {},
288 | "output_type": "execute_result"
289 | },
290 | {
291 | "data": {
292 | "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAD9CAYAAACC7q1lAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzsvXnQLet11vdb79Dde+9vf9MZ7tWVribbsWxkWbHM4CCDTEwMhEQuUmGIA5gYKBIHF5VA2UASlFAxxEWFSipAYsAKOAGbYAJOBQewg8EE5EEgGeNCRuOVdO8990zftIfufof8sd7u/en43qNzrCOfI6mfqq/O2VOPb7/9rGc9a7XknJkwYcKECb/4MI97AyZMmDDhCxXTBDxhwoQJjwnTBDxhwoQJjwnTBDxhwoQJjwnTBDxhwoQJjwnTBDxhwoQJjwnTBPw5DBF5h4h84nFvx4QJE35hmCbgxwQR+aiIdCJy9Z733yciWURe/wjW8c6yvDMRuSUiP/Iolvtp1plF5Is/m+uYMOHzBdME/HjxEeC3DS9E5CuA2aNYcJkE/zLwnwMHwBuAPwukR7H8z2C73JO8vAkTfjExTcCPF98L/I5Lr38nOmmOEJFaRP6UiDwnIjdE5H8WkQeZpN8KfCTn/CNZcZ5z/oGc83Nlue8Skb8uIt8vIuci8k9F5CsvrfcZEfkBEbkpIh8RkW+79JkVkT8iIh8qv32viDwrIv+wfOX9InIhIr9lkElE5NtF5EXg3WUZv0dEPigid0TkB0XkmUvL/7dE5AMicioif1ZE/oGI/O7y2TeLyP8nIn9aRO4A7xKRLxKR/1dEbhem/7+LyOGl5X1URP6QiPy0iKxE5C+KyFMi8kNl+39YRI4e7JRNmPDoME3AjxfvAfZF5MtExAK/Bfjf7vnOfwf8a+iE+sXAq4H/6gGW/U+BN5WJ6utEZO9lvvNO4P8AjoG/AvxNEfEiYoD/C3h/Wd+/CfwBEfmG8rv/DGXuvwHYB/4jYJ1z/lXl86/MOe/lnL+/vH66rON1wO8VkV8D/AngNwOvAj4GfB9AkWT+OvCHgSvAB4B/457t/uXAh4HrwH8LSFneM8CXAc8C77rnN/8e8GvRY/nvAD8E/BHgKnodfBsTJvxiI+c8/T2GP+CjwNcD/wU6efw64O8BDsjA69GJZQV80aXffQ3KbAHeAXziPuv4FcBfA24CW+B/BfbKZ+8C3nPpuwZ4AfhadIJ77p5l/WHg3eX/HwDe+QrrzMAXX3r9DqADmkvv/UXguy693gP6ss+/A/gnlz4T4OPA7y6vv/nebXuZbfhG4J/dc6y/6dLrHwD+3KXXvx/4m497TEx/X3h/k372+PG9wD9ENdq/fM9n14A58F4RGd4TwD7IgnPO70FZJiLyS4HvB/4oOpmCTmzDd1NxVDyDTqLPiMjJpcVZ4MfK/58FPvQg21BwM+e8vfT6GZShD+u+EJHbKNt+5p7tyi/j9Pj45Rcich34H9GbxxK9mdy95zc3Lv1/8zKvXy5CmDDhs4ppAn7MyDl/TEQ+gobz33LPx7fQyeGX5Jw/+Rmu5ydF5G8Ab7709rPDf4rs8BrgeSCgLPtLXmFxHwe+CPiZB139Pa+fR+WIYd0LVG74JMrCX3PpM7n8+hWW9yfKe2/JOd8WkW8E/qcH3LYJEx4bJg34ycC3AL8m57y6/GbOOQF/HvjTheUhIq++pMW+IkTk7SXRNfzuTcC/i+rOA94mIr+pOAn+ANCWz38COCuJs1lJur25sGiAvwD8cRH5ElG8RUSulM9uAG/8NJv3V4DfJSJvFZEa+E7gx3POHwX+b+ArROQby3Z9K6oh3w9L4AI4EZFXA3/o03x/woQnAtME/AQg5/yhnPNPvcLH3w58EHiPiJwBPwx86QMs9gSdcP+5iFwA/w/wfwLfdek7fwtN/N0Ffjvwm3LOfc45oomqt6JWuVvopHtQfvffo9ry3wXOUE13cGa8C/hLInIiIr/5Ffb3R4D/EtViX0DZ9G8tn90C/v2ynbeBLwd+Cr05vBL+a+CrgFN0Av8b9z0yEyY8IZCcp4bsX4gQkXehybL/8HFvy/1QpJFPoEm0v/+4t2fChEeJiQFPeOIgIt8gIodFnvgjaOLxPZ/mZxMmfM5hmoAnPIn4GtRlcQuVQr4x57x5vJs04fMZInIsIn9PRP5V+ffTFuaIyI+KyFeX/39URP65aOn/PxeRdz7QeicJYsKECV8oEJF3AN+cc/7me97/LuBOzvlPish3AEc552//NMv6UeAP5px/SkQ+Cnx1zvmWiHwp8Hdzzq+73+9hYsATJkyYAFoV+pfK//8SWszzKSiOoO8rJe3fzyv3bdnn5/vQXxYP5QMW8RNdnjBhwn3xNQf/CQD/+OR/kE/z1fviG77hl+Xbt08f6Lvvfe/P/Qu02nPAd+ecv/shVvdUzvkFgJzzC4N98x78x2jJ/VtE5C1cKiYq+PvFt/5GSgHUp8NUiDFhwoQnErdvn/LjP/G/PNB3nf26bc75q1/pcxH5caBGKx6PReR95aNvzzn/nQfcpF+FVlySc/5pEfnpez7/uiJBfBHwIyLyoznni/tu9wOueMKECRN+cZGB9Gi6p+acfzm8sgYM3BCRVxX2+yrgpfts1adb14dE5AbqYf+J+3130oAnTJjwSLE1G7bmUZhWMoTwYH+fOX4QbQdL+fdvvcx3/iHwTQAi8mbgLS+3oCJfvAHt8ndfTAx4woQJjxQXn9LD6TNABn7xXFp/EvhrIvItwHNoNea9+HPAu4v08D5+Prv9+yISAQ98R875xr0LuBfTBDxhwoQnFPmRSRDjEnP+UeBHX+b922jf6/v9dkMpmX+Zz17/C9meaQKeMGHCI0XI92vb8ZB4xBPwk4ZpAp4wYcKTiUeYhHtSMU3AEyZMeELx6CWIJw3TBDxhwoRHCiv+0SwoZyQ+EofDE4tpAp4wYcKTi4kBT5gwYcJjQAbS53f3g2kCnjBhwhOKSQOeMGHChIeCz/WjWdAXgAviiShFdu4Q5w5/Qb+1dvGKn82b147LfxAYo93lKn/1F7Qtn4sQ7t+wqvIv1xRK4d0xs/reBxY/OKzdZ3/xII+3ezCIOPQJRi+/LucO8e74ZX73cJeBMY9ogvk8xTwvmeflI1hShhge7O9zFBMDnjBhwpOJSQP+VPzOp/8ognDWBxbO8lLb4sVQGcPSWz62WXHVzXhq7vAGPnLeURvD1cbRWFiFjLbLhI9cbGhzwIvlbcdzughnfeZj6w1P1w1Pzw1328w6JGKGLia2KRDIfMnenE+uO/5V/hhv9W9kExIHlcMZqK2w6jN3u56nZ542gisEJyTYxMSV2nK7jRjgr978Tr7s6DfzZeaNHFaOD63PePP+AR883/CcPMdb3BdjBWKGmTWElPmX3UtkEtc5QhBO8waPo6XjWX/AzBlutz3/gp/hWn6W17tjKmP4eHvOa5slIWX+6s3vxNp93r73zXzlwQFtzPzsxSlX7YIP5I9ywW0Mho/e/Tu89ujX8tXmbQDUxnDa9+w5x61+ixeLRXj1vOZfrk7ZyJYj9ulyz9sOD3juImCNcN73AHzVlRkfX0VWfeKwtmxD5sfDz/IV8qWcx55bcof9tEQwvKZecNwY3n96ylwq3s9P8Hb3K7kIgdpYrjQOKzBzwo115IVuza+8uk9jMze3cGMTebG74Nl6yfv6j/Il8iznseOmuc3t/DHeZn45X35Yk8n89J0te06HY2UM+5VhGzNdzDROeHHTYRAuUsexa/BGeL67IJC4bve40jhqK+QMlYU+QRszIcFHNxe8ablkbnWM6fuZhJ5TIxCzrqtNiZgz1xrPea/hrxfhZtfS54hB+KK9OSddoov63ZmznPQd16qamDN3+o7n5DkuuM0vNb+ULiW8GI5qx1kX+eHt3+Ra/SaO0nVebY4QgZ9LH+dV+WlyznzAvJ+P3/0Rvvbg93NhLrjNJ7iWX8tVjjhwnjYlTuKGRirmxvGe9GN8ff0OQsr8TPg4X+Gf5SPtCVuz4Z/dfffLXstX97+KZ+1X8s/uvpvj5VdwzX8p87zkLs/zZfkrsGK4nS6ocFTieUlu8Sb3ajYhsuctd7uOHzr5UwB88fE7+RXuzZz0PXfTmj2ah5qEXhmTBjxhwoQJjw+f5xPwQz0T7g++4b/J3mhUEDPcbTOHlWANnPewKNN5YwUrykIymXWAhRO8geNKmccn13Bjnbg+MzgD3qgaeWOTaKyw8LDqy/Ic3Nxk5k54cdPzqrnHG9iETMpwtTHEnNkv/u8uCaugr43AOux0zhubxGElbCPsV3DWgRW43Ub2K8sm6Da1EZ5fBQ5ry6pP7FeGhdPlnnUJZ4SQMqsQ2feOLiVEhMZKOT6ZTUhU1nC1sWxC5pObLc/OG0KC97Yf40q6wmvqBa/ZM6x6+MSq50rjuL0NbFPgal3zUtsyN47GGp6aW6zAKoC/JN0uPKwDXPSZVUgYYB0jX35Yc2ub6JMeh8oKbz3K3Gr1x3c7PY8vrCNWoLGGo1roIlgDTzdw2gvvv7PhsPI01nDSafTzhqXl1jZzUOm57lJmE+CXHGacZF5qDbdLS4AuQp/RKKjPxAz7lXDWZb70QDACHzjVc7lwggjMHdzeZpZeqK3QxoyI7sc26JjYBtgUlnul0eX0CSqj0U4CYrl+9yuhMpkbpUvizMFLZaxl4KBS9nzWZ2ZWaByctJmZ0+10hvGcP7sQ/tVpIpG56CMLZxHRxl0Lb6itfu+lTeCg0mgwA3MnbELmZ1cnXLd7xJy52ngOK+GjFz13wpZaHDEnbpibvLV6LSFlrAjnIXBUeeZOcGXs9wkWXvjI+ZavujKjTXncpyF62MbER/MNTrjBKt3mjfKvM88zXt3MyRlO+h4rwpXac9ZFTkLLa2ZznIGb255XzSuqsr4uZWLSsX0eAvveETOc9B2vXzQ4I9xY6zUD8Oc/+cc+oydifPWbX59/8gf+6AN917zp9773fg3Zn1RMDHjChAlPKPIvZjvKx4KHmoC7CNsIh5WwtMp+2yhsorLNuRP6BI3NHFWZFzYGb4TrTeash8pkFi4Rs7Dvhb5RtnBYZWLOGMDMzchkFl7Z0sIJUh5/9+amorGZbRSebqBNsPSZbYRVUDZ2rU5Y0f/XFoRMbaGNcH0mXKnhdgszC1RgRRDRu/bR3FJb/WwdLPteOKotKSv7ShmOG6saYhb2vMEa4c5WmYc3Qp+Ulc2cYe4MVpRlf/Fyxq1tpDLC2+rX6baJsC5M7krjqIzwxn3PRe9wRqitMubGCZXRbTisdDuGdcUMTuBqIzS95aCS8dTuF4YakmrhlckYEfZ9ZhVUHH/dnqVLsOeUDepnsI5CYzNvPmrYBGW5AIeVYW4zr9sDyOPyn4+wdInKJG61Bm/02DZW938TMwuv+1EZJUcxwyYKc6f7MHO6ToCtVzbc2ExtdWwB2DIuGqfHRdDjIgi10+8bUYtPm2AT9b2UNVqIScfU0huemcNLW/19bZVtAyxcJpZICkDQfWi8cNrBXmWIKWMQEhlvNPKpjBBLhHi1cWxjZmkNXVLteeGFN+8dMnPCSZvZr6RcA46FW+CNKKtOz3CtMWVcJ643nv1KoxOA6zPDSZc5qoS417DnMm0HB5UhlCigscIz3nKlezXOvIacM9uYqUyJ1IC9qmZeNPOULd40hKTH+3rj8aLnZeEFEwALtTXsR8NBJVz0sPQNCyfMnRCz4+nZZ0R8d8g8qmbrTyweagLer3aDs09QG6h8prE64GqTOao0RO6STqx6YsGJXqgpayh44IWlh3XIdEnDUwGqnAlZeNUscdLphOAl47yGzHOb6TNY0fUOk4pe6FnD5xLCCkPop+/FpIPyyEe8GE56KevNZbLQCbsxetEd14IVoUt6QbdRB2JjoYs6IXij4fpLGx3YSy/0SfC1XvxdzDiBo1on58ZaNgH2ijzSJz1e+2WyiVknWIo9zCA0NePF3djMKggLl9lNDSAG/KXjYERliosgtKmcLyt4yWW5eh4BQhZCt2MaGV3+gc+0CfokBJO52gjHdUWXMiHDgdPjNhznkIWQwWa9uM86vYk60eWlLNSG8Xg2Vif7nDNt1AlbE2IaQh9Vuj1zp8doG3V9ADOrk+PMCnOXyw1SJ2ZgvGHq+nQLawOdEazdjZ/a6vJVvsrc7XR/NlFYR50gYxpu9LvlbiNsE4iARTjrItdnjph1HOzXKnMZ0TEiIuNNAXSsLb1KEntOikSn59IbS58sIev4vlKbMZFc2Z3ENy9X7yC9vaqQlPNet7+LOmb2PKPElLImJ+dOx/VgNLja6M0MlIj0SfdRtx32nWDYJTi9ERZOSVdIQsgZbzJPz4TwyGTbPLkgJkyYMOGx4AugEOOhJuCZVTYHemd3hbEMiSGAhU1sgqExmW0SxBR2bJWRbZOwTUKfYWEzycrIaoZQeukz1+uIF0uflREpK1X2uZDMaS/YwqproxJERsZEjRdwJpe7s7La3givqpKeV5QBaoJBWYQVZVGJT2UqlVH2MrPKwqwoqzjpdL3XK+E1C8/c6Q17VsLiFJRp1IW5doUx4xhZf2N3IXqfIORMzFIkkExld0nPLikDXLhc2HMeZYk+6TYuCisdmKA3eYwuvIF1NOP5G74XorJjZzRS2UYBBGc0sZizSjjnvW7zIBdoJCR4o8zzwCvbdDLsm763CgNrynRJijykrNULRDRZawtj7bMmFSsjHFeZPZdYB8OsyANzl4HheCq775My4cqobJhgHAuV0XX1wFGVy7HUfW5MorEGbzRZvOfgrIdG4OmZRjnnMRcZSc/r0uvyNJGox22/MmMUlQtL3vdC9nru77Y6LmKJ8GKGPS+0RR7JlV4HQ6TSFWlpGBfXK5WEtlHXvQmaIFTJSI91m3TdMWvEcaWGm9th2wf5Sfdjz2XWsRxPm7FGz1NIyob7UboZrr3MJu4igCG5KWScEZoiVUCmf0QKBPB5PwE/EZVwEyZMmPDzUZJwD/L3EBCRN4nIPxGRVkT+4D2f/ToR+YCIfFBEvuMBlvV6EfmZ8v93iMipiLxPRH5aRH64PKDzFfFQDNgJhKwsQdLutWqOw900FT0r02d978CnwqoUbRJmVpnmvOiVKvXslgGUZIaykIVVS1KXoM9S2NxOw41ZuAhw4NUGlRC6pJqnyGCV0r+FU13ZCyR2tqcBManWNXeqV6+LjjokWPatMrlYrFWVUdY5s8rC0yXpSqAwS6i9Fiisws6yVxllkY3Vf60oW7dOf+3L5yHn8TcDFq5okXH3vbnVbWlEGevcKcvUKCOX46tMdxUMVvIYzaQMtduxSlc06caCy3DS7fZ16VSLrowey3U0VEbPJ2gy7jULTSItnLK5jDK4mVWdHDJGhshGX4eyvuvNbntX0VAbjU4ExmTcUaXX3pCoWwVh6TVqsJLHsdNYxijrsrY5RG+NyZz3eq63UTCSsZLJ6DhdBR0TJ10mZ80FhBKVqIUSTDnnA2N1Bi6iRh+XZUxBcx+CbnMbd9FKY1Qnj1mPUVtsX43VPMOyJFAXTiMvQTXcxirjN6LHYiGqdwMlMaysuS56rhW11TUM1j+NVpder7cu6XcGXbyyytRqo4zbCiRzOWLLPNXomLjdfbri9odARsOzR487wLcB33j5TdFM/J8Bfi3wCeAnReQHc84/+xDL/rGc828sy/sTwLcCf+yVvvxQE3BlYVbkhspoiG8zHPnEzVa9kNtkOPRJ/bdRL9jKZNokxCxcrSKnvRmryyqTud0ZDnzGk8lWv3fWG7ZlAq3LxdKUEDaVi3RuE8Juwl5HQ5+FQ5PGC69NwqFPeiFFTTA0JrEVS1ekkOFCTCU09UYvUic66flygaas+z6zOiHuFzNun3KRL/R3oHJNyIzh9pC8mDspSaWMoBd7PUyaVieNRCbEXULKlCyXLRPpOmhCyKAXpClSzLZ4ZX2ZfIaEYmMzoTfl4tEb4ybuJJzBuaL7pyHntTpzVEXudpZV1AvvqKJcpHreNlGojSZfDEXOyIwy1dIlGiOsSsKyTxoWx6w3vzbpzSGgN009BoNcoZN5ZQbpwmDRi/y0l/E46PnUdVq/m+lC0iTYUZU56fXGNhwXlXF0ewapwoqOLZXV9NhsoyY3Z1bwXs+ZJrP0RrIJur3dPVGyhvA7F07MIJJZlJuWnsciB5idFDRsh97Ii6Rm4LjW9d7tZEwQ+/Jb9eVSXDf6e8NOetlGo1JRFs57lYSs6HYMN6XK6HugrppBSgEdzzEP7pvMSbeTunIuske5FQxH/7B6VImz/FmRIHLOLwEvici/fc9Hvwz4YM75wwAi8n3AO4FPmYBF5G3A9wBr4B+93DpES36XwAfvty2TBDFhwoQnE0MS7kH+Hg1eDXz80utPlPfuxbuBb8s5f83LfPa1IvI+9NH2X49O1K+Ih2LAVjTJlrPBmUxjMrXJVCZRGcPCJbbRYAxYSSxs4qiKnPYWAzQ20RS2erc34928LsmfwMDKMjM7pMrU3nSzVe+pwJhEMYN8YPU7S5dZlnBqkIWM6HmsTCJlS18CJDswsGJjyxkigrN61x8YgEGTbtoLIjMvjMaKjCGnNzu54m6rbHcTlU1UxQ43sI+h8jBZlR0GBjqs0xvdh5N+SG4JQqax6hc2AlfqXMJoKSF/kWncIBkomyUODH0XOsastqpe5FKYmgurhaXNeFFp4Kw3bKKMyTIpoX1t4CIYrlSRhYusgqWx+v1FOc8LlznpHLEkbKoSsp71yiCVMwmVSXjUstiV82aE0cqk4TAl0skcV5E+2cIaNdwNCXJhbUNSdgitUxbOe2Hh0pg8sqLntTZSqialWPJUekHgIgxyws7ytyjUfrC/7c9ykX+U1fdJrWVqB9QIQdAoKmVh36cxytBxN/ihM3OrY3Dpdaw4ySwqZfl9gi4PLLVIHaZIS0V+WFi9Ftsk43Hrk1ap9iXpd+Bh4ZKyZ4FVlLGi8vKxmVmNUDK7isvGZs6DYel3VsCYVeoakuDDsRiu60eCB7ehXRWRn7r0+rtzzt/9kGt7OfXkUzZARA6Aw5zzPyhvfS/w6y995bIE8e3AdwG/75VWONnQJkyY8IQiQ35gdnvrlUqRReRbgd9TXv6GnPPzr7CMTwDPXnr9GuDe7xZB8IHwg8AP3O8LDz0B1yYTS6LFycAwlf3NrNaLJ1QLzsB5b2lMwouMmq7YyAtbO94pF04PspVMKgkzK2rgD+XOb0VYlzvsvledq8+CyZoAKgVLFALDrCTaDHpXf+NyxXnYJ+bBmpWgtzxVR/ok3OkNhp09LWZhHXUdKauWqP0jdjfJMPRYKJpxznBUQxtVSxwShKZYdLalJ8JgGZtb1ckHS9cqKKM/LT6eqiSe1B6VR1Z4udigLVqjLdppzjurWZtg6XSflPWpDdAWBrPn1VaYMrzUGoxk5lYTpjFqAqrPO1rgjWqrxuSi9wuxdzQ20QbV2mujDHqbDOdB6LOe94ZdYmhICgl67jLKbkNJRgqw7zInvWHPZVZRqyMBFjZSGctxFTjtfbG7FSYqeayiYxyTifOgWvXSJWZW2EZl9pWlFO7oua2LlpyAV8+0CCWUYgs9H4MGrAxzGEug9jsxuk9qxSsJz1IEtFc0067Y5Y4K429sHsfdeS9cs3qcrFFW2xg9ziKwZ3f675CrOK4iz2+cnrek+6nsW3Myy1Jss466jG0UxCrjtqLjcRNFC24yVKLnOItGkouScJZiE0yixyyhyx6PDXqtthHyI3om56NqR5lz/jNocu3T4SeBLxGRNwCfBH4r8B/cs6yT4nR4e875HwHfdJ/lvR340P1W+FATcGM0YRSznuSEhvaNTdxqLXc6i5PhhDHKDdtkeKruSeikGKLhuEqjyyGURF6fRE9wEuaaLyNmDZ+dwHGVSlXTruT1xa3DyVDGqsvYcz//rtlFM/42I+QsXKkSbdIKn6SR/Ji0OunVgwo6aBPq/ljHvGtu43Z+yT6BOE0IpeLlvVJpgrEpx8pKqTMyOsm0Q4a+SC5DgiYXeWRwItRml8V3ohPC3O/kmHWU8j29adVmlwDss2AYJkXhehXZRjPKNEP2/LAksKoie+y7SCyyTioTzUlJnlYm00hxZ5BHJ0UoNxeKVLFwmZutek7L4aWx+r2u7FtmF7JaUW/qvtOJvC4TqBXhpNcEq95wdBuHSaRLug5hd7OQcqxCgmt1LI4IvakYEn02uCIxHHhNKMMueddn/bexfEpZszeMLgFQSQWUFGyjJpCtaBUh5Zx7Ua1qcIlsk54TQ5F9XOasV6nDoOThvIeF3RGA4eZnZXdHFOBWa8fjH4vUkLLup1bz6Y1xGCsxq6skZ9j3iblkToM6WM56GStcBwluHaWcW73Brcr1UpnBV66vh2vCG0YZ5DPGZ8kFISJPAz8F7ANJRP4A8OU55zMR+U+Bv4Neqt+Tc/4XL7OI3wV8j4isy3cvY9CABTgFfvf9tmWSICZMmPCE4rPmgngRlRde7rO/DfztT/P79wJfeemtd5X3fxQ4eJhteagJuB49qeqfVeuRssnjKlKZREbG6iiAbTR4SayipY3CG/Y2OGM5611poJK5VkdShovgtOop6jISGhYduMTCwrW65wMXDY3JY2i/79Noe5rbRLbaIGh43SWVJ857T2MyV+qeTbQaNiftC1GZxDrWJVGRuNtrwm9brFpQmskEZRDbqBYvK5llCbdssVFtR/uYMgG14GkoOrASk0qrxAw1ynKqYq0bkjdDGOoL49VqPWXnXjTZMtj35jazijvpZuk02WMwoy/3PFyyMJUwdfi/fq6SkAH2XSSjbK4qkoIvCZ662KeWLrKOZrShWdHtuFZ3LH3Phy8WY9XUXrGODdJOiMr6NEoIeJNZBU2s+XJez4MZv2MFnmliSSYq2x8YtheVKKyo3dGJHvfjqscJnPaWulRyVSZxHuzoAR9aZA4VmKlEBJWFY59Yul1fiL5EA13SRkYXfZEFnCa3+hJ+g74+9AkRTTQfV4FPbjzHVcAby3nQhLWJ2uPBDlGN3ckcsTQl2ib1uoMmDgeJRYpVzghUwKEPpOy43Wn16JFPNBZmNjG3Ghlukyk2zkxf/OptqTLdc5ovPl7IAAAgAElEQVRIPwuGo0qPdZfUW6/yj46Fud1FC0PyeG6VNafy3nyUgR4Bpl4QEyZMmPAYkHmYJNznJB5qAvaF7WwKw2ysWnu0Mm2oM0+kbNiWRFBjtY4/p6EDVSBlmLvERTBjQka7Y+nBrorWvO4My9K+cuEi3qRiptdKo8YkLtBlXATDHC3gqEtCUDUp1SsXLlAZpavrsl5leInaJg595DxYrUord/h9nwsTKFVVeVf5Nmhkg4XNu0E3G+xpmT2XOQ+qy63CrirLWrUXhWJNGhhDX5jY0N1qE7R/wVCgMBSL+EtVhIdemehQSddn1eVBtb7GpkIiLEunbOxO5wDtg+CMJg33XR4tgI1NnPVuZOkHRZ/vkvC6eVf+b1RXLus48JHaJOYucNC0VOsZ+74k/SSrzipacFGPrRvVikZhvY3JLH0svS9s0d0TQuY8WOY2caXq8KIHKJbxkoLmByqTuNM5DENfkMTSx1KMENlz+v91NByU99ukkd3Sq04cSqWlE2iDYekiXbLMbGITNfq75lPZB+F2qfZc2EzVKCPc94mFi2NSLGTVzuc2sQqWpUssi8a+KJFIU6rN+jxYMfW3T9eBO70drwNfzs/tzrGKeg7mNnNQ9Zz0GkF2pegp5MEiZrTAJysjjlmIQc8NZRwNeYEhwdulnQYOl3uj7KKHwU43FB81Fq6U5OKjQZ4Y8IQJEyY8NkwT8A5qDUs4McxdYmH11hcLC9A+EYnapDEbDPqQxat1yzY6Tjp9pHcozGP4/dIFYlb9NmehMpE9J5e6WAW20XJcxVH76rM6HhqToFJG3NSpMDxlqfs+4EphQWNTsaUNbA9OekcVE7XNOKP70ycp7DzRo9rfrU64XiujPfS6vG1UJ8HAPtUVohHCvGT7KzOUeqZStKK65U3RbdyzWdlSNIAwsxpBVEabffdFAwWNyGbFEL+wiaVTXXYdhSOfuN1pNnsdDbloe+aSXmolc1T1RW/V46ZsMLGKhotgePWs427nuFL3zG2kS4Y7nS+arGrq26gRxIEPnPZOmZyNbJPhqGmZ1b3qjjkzLz1oj6tYLFqGfR8I2dBG4UrVEbJhFSxv3FvTJ8M6Wg4/ZcxBb9LI7BcucFB1PLeaE7Nw6COHvue0dxz6MPYoOa46YhY+fDEvkVakKQxQI6BcIqtMYxI3Ws+rZx0hCSe946BEF8tLrhpfSqNjMGOHuW1UnTtGtf0NOnnKQmPjqGOfBzvaJbUvhpDGMvs0FkwMjg0nOzae845V2hJJuBLlAbywqREyR1UqfTbgpFe9uS+5iFyutWFOG0qWQa+TPjvqodwftUYunTpqhrxGzEJjEzYJ+J0lblaiuCE/8EjwWXJBPEl4aAbsilSwdIGmiPspGfZcpI1mbG4+N5GQhNpGzjczrs43bIMjRMM2WpbeMBt+n7WfgVbUUarmIksXyAh3Os8zexfc2cx29p/gqEziwAfWwRafok6MSx9Lbb9ul8oiUiSUNIablUnMSj+JLpnSsCSP/Rt8mSxjFp6qIwnIGOYlgbcoSZm5zZxnGe3ZTZEoqpL8GQbv0iXudLZ4P3Vw+2K3shLps3DgI5uo68gYmtJoaGH1Am1sYhMttc3suTD2XggZrpUwXGWFXYhpi793kIDmNjEvF8kwtQx2r5mNBC/F0iQsfWAdLUsXS1JRl5MyHNUdoFLKNmkizNmI95Gl77nd1hjgoIo0RhOxw/lrTGJhMwdlkryxrcbJYBPtOCEdVdrwwRvDwgX2q47b24aL3o99Mg58pE9mbEi054JOBFXPeVfkiixso+V60/HSthpvBANZUBIQuQgacw9e6sYk5k5v3MONXUrvioxa/So3TEx5XM6+7znvdVlLp9Wgtzo7jpdBftNzkTjtbZEYKH73PPbtSJdkgWuznqt1y3lYcOT1nGSEq3WPkLnTea0YLcd56fTGeBEGWXDnhz4PhtfNW86D42rdcbvT6eBKOebDNbG0mXUwWs1XblhrhjEqn9L/Y7gxPxLkSYKYMGHChMeHz/N+wA81AT/dtNqVSrSptDdJQ+XouNV69lxk4QMhGW61Fc8u1hrqtglnEw2BLQ6iHR+NM7NqefImsXCBTXSj8X/mAttgmdnI/nxL27uSoDEQNKkjZJa+x4gbE3rX6pbKJmoTWQVPyrCNdmzXqL0NdgmnxgZOu6q0lIyc9J4oMrL5gSb2yRQ5g5G9bIsFyop2N9tzyhAvwq7P0dCKU/sT6MIOvbIbQavGBGVvN1vH0iWVVkqS6ax31DZzzfXlmOv7WhxgmBUp6MCHUQ5aznue38xY2EhtNZF11nu20XFcd0UOsKOV8MhHQoaz3nNcd5x2nj4ZUtbwPiOE3jK3cTT9NzZwrUmsgw6jLllOtw3WZGobEclcGwpwTKLJMjIzDYWH46KsceYCC5c5rltOurqMgzS2G43ZEJIpIboZmdatdrB4JU6Sv1SQoNu09BpNWcmcFka8cIM0sCsCAUNdorJnmo6TXvfrZuu0+KR8tiySkVrTNIELaE+MEiHs+Z4+z1hesvxdrSK3Osu+i+y5xN3ejp3knqp7+iyc9Y6hqdvCRRaoTNaUCMVJJmZNNItkDsq6F67npKvY9xp5ngWnFX5Jx6GXzF4duNU6rtY9IQkLp3LPtbqjsZEDn2hL4cX2UgL3stwXy3U/I4/jrktGe3qIdhw0O7XkM8f0UM4JEyZMeAzITBLEZdQ24koyY+l7rMmcdRWNjbxx2dKVEkcribmNnHUVlVHt1NpETMLdbU1ld70fttFytdnywmbG3EacJLbJYkVtZ84nQjaEaLFFw7VZDf8pw7X5hk1hKkvfF/uN4azzNNaMSZDGqi6selnHna6CLCxcz8wH1sHR2KIlSobCMuc2lv65lmMXWAWrCYsk3Okc+1718OFQDn1gXUkAngfL9TqMWuvzG09VjO9zszsOoFpmYxPnvWVddOAhcZNKAoRkxgTNwvfU1vDS1rP0cWQuS79jyt4k2rjrBbwOtryfS1/lyDZqUm44H678f+l7llXPi2t92qP20Iha1FKYsDOqrx/WLV20bIKjC7oMPS5wHhxtiZxqm1i4npvbBpGMs7uEbRcNbbTUVq1ofTJkhKO6pUt2HIdL3/PCpuGZ+RYoDLqMxzaZMarZb1piNrRFz527QZu0JbkWS/mzMvtlrd+vjBbuPNV0bKPhwGtktCn9pDNaFHQWHHc6Xdb1ui+PNTKaGA1ujPJ0HBVmbPP4sILzYDgqib6hj68WlJTS+vKboT+FdpQznHaeAx8I5bito2UV/FjY5AxcqbqSx0h0yY7FUY1J2usBjZhmLrAJDmsyh75nbTQ6rb0WyGgeRxPrdzo/JjEbk9hzmljdRkuXDLWLGMls4qPqcpunJNxlVDYRU+asr7RW3QcO6pYQDVK8i9toOWpaGhtJCJWNnPUeYxLWGK7PN9Q+EEuCaOF61sFx4HuWvuf59VzDVRs57zxzF6hMxJbJqk+G097jZBcC7VU9IRtmPrDqPSar77dPhspGLvqK/XqLEfVkLsoEdaet2UbLXt2NYW1dJJGLoBflukgXw7pC1oy0XiAqgaQsoxRxXOlk25RBG7IZm9EbgWdm2jxgHQzrYKjHBisqaSTg2fmWF7c1Iev6ZlbD2NNew+mhwq8r4fhxFeiSTjRzF6ls4ryvSl8OU56yoJntWjTJtY2WpuzrkHCsjMoVOaMTpe/Ls8/0ePUlnF36nm2ZjOzYQ8GQEParTltZdprougjqoW2TKVJRHieGlIUb67neBJJhFXy5CTjWUX/nJHHSVqPH/LzXRyUfVz0vbhp1wFAqMsuYUjnHcNHWbILDlX0DuNJsCcnw/HrOng8snO7L3KmEkaLQJzverHSC03PbJlP6IOj47JLhsHi2T0tlp573zNwF1tHwVNPhjbotuqTuIf08jonjobdFm4S5TcSsY25djvGBD7TREBHmLo5EQc9XpI56GS+M3kyH4wvQuMiqz3ij4+Og6nVZWVj6jtpF7rYVdJUmqotnvi7j/bSrNBFX5J6qJC230XK3cyWRbkq7V5X6HplqOzHgCRMmTHiMmCbgHbbFP3oRtKrqvPNcmWmyJRYGenPbcHW+obKRNji6aHnd/jk5q2d21fsxKaPPn7LEbAq78SMra0o4sw6Ope+pm4Bdp2KpilxptqWJtGEx39IEtfE0LnDWVuM2r3pPYwMxGQ2XoxCTUNnEU7MNziRi8Z5qWC5cn69Z9o42Wg6qjnVwrIKjT2ZkqQbGHgYHJVzb92EMJfdc5Lx37LuAM2oPaooksI4Dk43cbJWpSmFBQ/DWlPDzwEcMcK3Zsgpeu3dVHTGrna9NO9aibUEjhsx+1XHjbI9n55uRsRiBNhpOe8vc6bbUNlEV5pQRLnrHYd1RmUjMBkNm4Xtmvse5xK3zOa48+65PhqPZhhANbbI0NnBSjr0RypOkNZkYslbPDZ+plU141WLFeae/qYwmcc+6iqULGMksfBitZOvgilVPo4irdTeyuZgNlQ0cVx3nvcebxKp3ozXysO6wJvHhs+UYrVz0ruyn7ktfooXz4Eq3NWWBR1Uo0tLgdc2cBzd6eof2opAJGA5sN0YVfTLEnMfK0G00HM97rtRb+qS2yqd8z3nvMSJcqVsq40fZbPi3sZpca1yksXqM9Lw6QpLxetLHHGUWVc+i6fjInUP2q44u6uOR1sFyrdlyp611PJuew6qjjXZM4A7H+rR3bKLl0OvYSmUc6ZhJzI3KDYdVN46fyiSuNtuHmVZeGZMNbcKECRMeH/I0Ae9girZ04ANLr3f5s7ZiWfUsqpZtYauNC3gfCWvVhrtgOVs3qkUmoe2d6m3Jsuc6LvpqTAYMTxW2JuGy6o2Vi/g6UllNzDVO2dmqtyyrnhQFb1VTPJhv6QvbiNmUm6iuN5aCj+H9Nlqu76+I0XBcdRw1LS+s5phWT/omupHpCVqBFXo3MqE+CdfqXnXXXhmIK8xHSvctTZwxJsQqk9j3PZvoWPgwJsBOek9VrHir4EoBQOBW0T+30WkVotXowaOaWxu1Yu+g6jjvq5JwMcxd4Gqtiam+PP04ZuG4bqmtMsrjuqVNllXZp5TV8mRENcpV75i7gCQhJkO7daMGC7BvEqZo2Avfc2M9Z+GUnS5cPz48NZRjP3ex2AFN6S9skTKm2ig7HTbrgzGXvht150Gb9yU5d7mPxxA9EXRcDTrzaUkQNy6yDg5TCih0/Pacdnq8Dqtu3KcuaRJK+2QMBSdbbm1nPDPf0EXDXtXzc6f7NCZxVH57q/Xaf6I8aslZtWUNPSxqq5FbKIUV571nE9VKV9nEeqtFQxllvCGrJa6NWgxR2cTMqT6/jY79quOkrRkeWQVwPNtyYzUHdMznJLTRlC5qkVXQHIFIHo/1CxcLaqvX1EnvmdvIqkQA2lOl186BJWpYRT1enag+Pbd6LQIcVi1ttIT8CB81+XluQ3uoI1XZROUiV0voLpLZRMedrYYzMQkLHzA2s209L67m3N3W4+QYoqFLVpNzTkPlykUWxd/aWE2K9MmwDY5ckkeVjfStnnBtg5mJSThsWma+ZxucZsJnLV2vkoYpLgFbZIyE4EzicLZV94CLWJN56WxBV6qfzspkl7KUR53HMRNvioPCyfBI+8yVSsM3K2msALrWbItDQCultKWmjN5pLUG1LFyPk8Rx3ZZQPY5VdwdVz3HdlueI6QVzq/WsgiNkg4iei2a4IdnIrW1TEoKaeJl5bXp0t63YlqTbRbDjhNYnYRU8F70vE7ylS5arte5TSJpQa6PlpK1Zd57bm0abtkTDzAW6aNi2Opm3wbLne46alnnV4cbkjF7Yr9+7GC/iYf2bKHTBjglZZxKVjRzX7SiL3GlrumRYl0lBb206Sc1dwJnEKvhSoZdwJnGlaVlWPYd1R0Y4aStiEl7czJgVN8w66I1m5gL7dUfjdn7ape/ZK2XWjY18crUo4z8igpKIuAv7L4Itz/IL7JUkWWXj6OMGTQyGIjHdaj3r4EpDeMNF70dP80Up7baSS3N4y52uIiTDJrjxRnxrO2NZ9Vxptuz5nsOmxdtYKj7VcXS6qfWGVm4Irlxjw/UoAtfnei13hYQsilNk3/fjjalPwtVmyyra0T9eGZVChht7ZRK3tg1dsuT8iObNjPYseJC/z1FMEsSECROeTHwBaMAPHSv0JeS1peHIwvU0LhJKkqdxgbN1w91NQ2MjV2ZbtaiZHSMyZFal0soXBrMulTtHTcuzeytlMyURdXfT6LoLyx6SBduwu3+EaGg7x83VHFOYYJ8MlYvs+Z4re2tC0iY186Zj5nsOC8tsg1qIVsGN1p6ch0fA7KqB2qiWsiEp0iXDee9p41ABlYqtKLH0PW0y1CZx4Hu8ZI6rjk3phXGrrTntKtpo2UTD0gWe3VtxXLcA2kdhsGxJ5rhSNr+Jhi4azjuvlp8sbKKjtolVcKN1CSjWNWV6h1XHvLQP3US1h7XRjEmqvhzTTVS2uax6ttGNydCb24Z5qSYLpQrOmcS86ZhX6qWel8rFLiqrnTv1V2+jsrfDumPmtHdHzMLVKpCLLLQsFV2b3rEKni4Nj9rJY9+KPd8zs4FNNKyC43Zbc7et9Jgkw8IHLnrPeRlbm7LuurDxhQucl4TqUIl30lZcdJ5cZIHhuLkitY0WvgynbU1MQih2sjYZNlF92drAyY4eWG/j2Ahde4gMVWxaUbYtDwoYjntTGPO2XBNW0hjKd8X7e967MYIZMHin+2ioaz0Hm+CI2YyR4Fmn++jK+Oyj4aSrVHoqUpcpUZVaMROHdcsmuuKpjiPTDdmM198qeM6DQ8isgo7BlOHmtuHmtnnYqeXlkfKD/X2OYmLAEyZMeGLx+Z6Ee2gGHMsd0IjqsDMXOKzbkeXe3TY4GzloWmyxdW16Ndg3Tosz2mSLLSpzVoohaqO65TZoUuis2yV7EkK7VU140G9nPmAlcWs9I5YigIHVJKQw9Mx5V6mOW54s2ydNDK67ik3vuLa3Zn++5ajZ8trlhVq7olWtrbB8ffJxLE3WlY2d9Z42abvIXJ6S2Cdlt6vgaZMlZViXyMCZzI2tdgdbem312CfDJlr2CsvsCnvaL1VMQGleH2iKJcpL5rz3YyFEU6oHvdEOZ9pU3Iz6+bPzTUm8RfZ8ICS1cV2ttX1jm7TKb+h85YrWvQmOTUlqDeuhbHttEpvgqFzkdN1wYzVnU/p0hGRoe8dJW9NGy1nvR7355qbh5rYuRQfaGH1ddG0DypyzJo7WwXLWV3izq7TaBstJV42JTmWfZve6REQZ4fa2Hgsd7nZutHMtXRiLCgYb3jo4LnpfKikNm+h4ft2MhQWvX16MfSu0V4lWgi1dYGaHikntDDYrWuuqq6ht4mpp6zmMEiuZfd9zte64Wncc+H4s5rlTKty20bIOjn3fcVD1Y1WbdgvUBF1Iwt0SReUy3nOGRdWX8aOFNI0N2tC+99oYvhTOHFad5hSiHcedIXNlti05hZnaTX1HZRPPrzW550RZ/VmxjB5V/WgrrIwWLt3t9e8zxlCI8YgZsIh8k4j8dPn7xyLylZc++3Ui8gER+aCIfMcDLOv1IvIz5f/vKE9Mfl9Z9g+LyPX7/f4RpisnTJgw4RHjsyNBfAT41TnntwB/HPhuABGx6OPrfz3w5cBvE5Evf8hl/1jO+a1l2T8JfOv9vvxQE3BMouy0rbhZ+gN00bLqPadtPWpnbe9Yd1712iR00bDuPK5oq6tiQ5u5oAZ/0Uztpne8uJ5zq61ZlLLiQcs7Wzf0UbthGclcdH78PCa1ioVoeWpvRWW1dNlKVmbttPQ5Z9Wo717MWQendpx5x2LRcnfb6H70jtqqM+NOW7EuDonz0n82ZeGk9yxKA/mhm1hX3BJ9YZKb4Mbm5Te2NWe9ZRMtJ71qkKCseNdRynLaVWyjIyYZnQkZLVa53dY4k3Em8/R8UzLwpWdt6SQGjLX+oKzVlkfirIMWJYSiC/pSyz/s09xGFqXHc3WpMf3NTTN+f171IxNso2Xdec6KywLUHpWycDjfUpfeCSHD0qtmfqdznPVawDAUJgznL6GaYsiG8+BYR1M6dvXs+46mlF8POCudxAAtd3ZDqWxk6TvmRftvbGS/9AveRnVqCKorq6adWHi1ZA02snWwHFf96AgA3f9PrBtutZ5F1VPbxJ7XrmL7XiOYQ9+X/s6RxoWR9aYs7DmNQPoyVlLWYodBd4XB0qWFIEPpd3upt8e69MmoTRx12aFUv3GBW8XRI5LxNo0Wv5DV9tdFw0lb0ZW+KoLaPaui3SaEVT9ozcryL1s4w/joeWXjw3nXHMmukGhm89ig/TNCzhDTg/091GLzP8453y0v38PuCcm/DPhgzvnDOecO+D7gnff+XkTeJiLvF5F/witMsCIiwBK4+3KfD3ioCXhIvPWlD0Eqlqo+GfZ8x6rX3g3baJmVMG8THEdNy3nvWXdaoTRMvmddxVlbs6x6rJQ2jXXHU82WlzaN+mbLoHxpPWcTHHfampNtzXmvtqxhEDjRAbeYd3TR0gWtSZ+V9pi5VA2BDpicVU7pe0MM2j7zfAwBXTmIO0+v2p52MsDNViuRTkq1FWiV0a1SCXZYteMz84aG3/s+cK00MT/pPW3U3hPDJK+tPjUk7pNwpfTUGOxblYns+268aGYusKzUd/vSpqEykYOq0xaOoudKx3BpXF514+Q6PLdusEoJepNZBzs2vtGnEzsuSlVgzhrqzsvz+VLWRNtx3VK7yEXvSajdcPA8OyltM6MZ7XZDQtGI7rNWGGoDnhc3DUsXylMlMidFQtoWuWEVLUsfOCr9L0DD/03YJajutDWNDeVZbHm8IQGc93reznvHSVdxq605aSs20bEOdqzGvNN5NtGxTYaXNjP60uTHCNxcz2ij4YXNrFRrWs6D47xUj22jZRscJ73nTufIMDa2MagsdavVXhkfX81YR6NWspKk9UUGOinVboPv2aA3uTZZDuuWdbA8v5mxDo67m2acWNtgicnQRYuI3qz6JLy4bcYJ/7BueWaxZhsc29LzYhMcq94xK3a2bblpNcU7fFSFUnXZc1Asaje2Neclgd4lbUy07wL77jNvyp7RZ3I+yB9wVUR+6tLf733A1XwL8EPl/68GPn7ps0+U9+7Fu4Fvyzl/zct89rUi8j7gOeDrge+538qnJNyECROeTDxcM55bOeevfpjFi8jXoRPw24e3XmErLv/mADjMOf+D8tb3opLFgB/LOf/G8t1vB74L+H2vtA0P3Qvibud51WyjIXKwY3KmK2b/lIVl1VO5gC2JkbM2Uw1G8MKUbm+b0jdB6+WbKnCw2PDhW0fEbLg+23J7W9NYDSVrMzx/Ti1Ig/3lmfmWmQ/c2TRUOXF6PsNK0iRUMrxULGxVqZjqoy19FBxWem6dLTDCmFyzkkdG2ycp7EZbOBqB096OzdLvdI6l03ab29IRanjO3SBFACxKS8uhfZ8rT5deuJ5V8NQ2MbOBfZ9KRzB9TNJZV40JlfPej5ao58732C/tFw15ZDC1jWyjMsFcWN6+18IWZ7T4YOitsAqOW51jZjV01o5gkXWvdr6hY9fTsy0XJYHjTMKwqxKsbcvMB25vGl226zFkTkthzqo8NsoZ7elw0vvxKcEHPozWsLMiJR3XHdtYnqcG4zYNlrSL8uipxgbWoR7tc+tiNdv3HV2yLFxAhNEWZ2TXHU+tfWo5XLiAJG2s7kl4A8+tKw585HrTlQo+ZXc9Qm30OXoZXWeXDMsmMHeRu51nUzqR3WlrYlZ7nxFRVmsyoVjPhme++fLEYm8yH7lYjJES6OOPnm5USrmxrcdEb7utOPBh7Og39E4Z2P/z5zMakzisO0wpmhjG9qHvxwgjZm2ZqZ3m8ljI4o3KOF0yLIrMELPaJLvRdleSxaUadpBJQH87yFSPBI/ABSEi3wr8nvLyN+ScnxeRtwB/Afj1Oefb5bNPAM9e+ulrgOfvXRz3TMr3wQ8CP3C/L0xJuAkTJjyxeAgJ4pWXkfOfKYmxt5bJ97XA3wB+e8755y599SeBLxGRN4hIBfxWdBK9vKwT4FREBtb8TfdZ9duBD91v2x6+F4TkMaExlNwOdqPhkS3nnR+ZUmOD9gTuKrwxXHSGg6obizAA2mQ5ayv6UqQwaJdaSOBxQW06B5UmPfpkeHa+5qSr6KLhzqZByk13OBfzqufuttEHPNrE0d6a7cmSLloqG8dS1726025owbL0meGJuYMRft/2Y7KjS5bzXlmYQRnyNpqxC1RjlBXvOeGsV3vQJgop25KsdaXUWXXL067ClQdJguOsl5LcM+yVUuKXNg2zkiAbCgi0ObijyjsL2kn//7P3Jj+27Vl+1+f327/dny66273MrKzOlOWSyiALJgwsBsaCAY0HFhNskGECDJlihAfgPwAGRip5wMQSEgODhWUhWcgzjGRhGdyks5yZr7nvxo3unLO7X8tg/c65L4vMrLrkK9VzKpYUei/ixjmxI84+e6+1vl1JyOT3m2amLV1+PeT4vphk5/grq4GEeM9eVp7p7AIXcxK17H4frJEduvHsnbwmT7MkWp+aktoE6tKj5ySeHAkel5rXqwEfT+5fMqnYLDYZQ8HixOt3CAVNkI71dqnPe+kHKwKKF40778AlfNNzb8vc3cpO3kehb317faCtHe/2PUqJQOHk/xDy7hriBzAV2cOLn4fNwhqJaHrTTkIlDJqXzYKLpx28BKI+OZERW4Ty9n4pz54MISluszzbKPBKvmf08nuaqAkJ6uwlffq9Dl5ofL0J2UND5amlYA6ab2Xz+cEb8TQOmkcnDnuv2onGBO7y6x+y8+Cjq84+xm0GDeMiUuV/tl+ffTNOknuAgzO8bMUroyoit1NDb+xZKHXq0JUS8LfQicevALEggpmvpRKcLee+3vovgCvgvxOsDJ9S+hMpJa+U+k+BvwkUwG+nlAQsN4MAACAASURBVP7BT3j8fwD8tlJqzN/71TrtgBXwBPyFn3UgH3UBXrIV3apyKAW1DtmIPXG31JRaFFchKVkLHGs5wZqFLgriXBSJlBRKBXzW+J98DeQiLRe/00l/Slu+rBeeMjc4Irr8XbaKvGgWBleeAbtCJ462otbyPU0RWKwowlISVkZjAjZotv1MVXt+eFhxOJkEBYWL5dmgZlWKKc4plaItYk5REC+AvvDMeZRtcmJyk8hJt2Lg3uUsuMEX5+fRCoooN7XRyw1JbAnT+Xtk7ObMDji9Cb67Gr7ypimpdeTJyUrk/dyc1xUuanySHLODJ/smhDPIdVk55qBZoqZDzNePrqQrIrvK8vmYWRAZUf987M4Zc4WOHOb6vPbospH3Mb9Oj7bidTdydNU5OUF4uI5PRxmVTxzSq2yx2RQCJN7ODT4qVHGKeE+8aOcMRqWzCdGJe1tloKgq4nnFctPMZyMlTQLN+UYa8w1BK+Fdn5g5r1sBl9qccrLk9caY/SZOpuWnlZPk2Dn+0X7F69ZybyWdZAgFS1CZVSIrsVNiy8FXrE0QS8evKM1O6zwQ5sTJFP2qWc4c500pytPHpT4zaAZfcnAlfebGx/y6L1Fzk3McXVTMWYGoFdkcX1YGt3NDXUTurJjxHFyVwURJDbFRGqze+HNa84nh1BiPm5rsUwIvm4Wu+rouwOkPRIiRUvoL/JQLY0rpbwB/4/d4/P8J/NZXvvRf5q//bWD7McfyDMI913M91ze3/vn12fl91UddgHe1dJqFjqwri8sa/6oQi8VX/UgEfChY9zPrTMm6n1pxOqsc7waJHGpLz94Kb9Rk34c1Ljt0GV62U45AUhSq4Fdf3/OPP7+WhX+2wCt15KqdqYwX1zVdYHI3oYOoflolYOCqX5hcKVQdJZlXGoMxEaVkvTJHTacl0fbkUOWigFmn5NfOSKd5ylXrdeB2EZP5bSm2kz57LdQ6MQVZaRglCcpTUFxU0rnMQZ5zzLS+b3fS4T05Q1eclFuSw3eKrtmVDqPTmddcZG5wb9y5O16XjvulzlxSGX2l85MRtVCJbeX4fGros6/FmF23rmuLUZG2EqDnBeKHccqCW5fiMgZwXCr6yvGrVw/cH1v2Vkbem/XAjx42rEt/9tDoS5kSykw709lt7ip3qXjOU8mmtrmrFibtEgREnDKoV5eeLo+860IAxKeppsrj+5zXKqf1kotyLly3I/djS4enz+brWgmIeYqVAuncb5qZh1BxVS/iwKfE76BQiZt2Foe17CUB8OubgcGJjegSFX0RKNQH+liA88ptCorecOb27peaN+1MU3iWUNCXjs8GAeXEx0Lc1PrSSe6eM2f3shNl8GQvWWtZC5w8Hj4fu3NE0Spn/H0+tjmvT6aSJ1dwkc+vKWienCSMn7pdef4PIPWJOz56w+jNeVNglICzj08bAP7ox12LfnL9YiuRnzvg53qu5/qGVvrF94L4qAtwXzm27UJVeY5jffYe1Qi1aPZC1q9MIIQPrknX7QTAejVn3wa5+7/oJN6nMZ4hE+4rHegMbNuZ90OXjby9GLLrHGSYqV6nvZ/Ke+Yi+1Gc0pNjUhyWipDAewno3PUTi5XOtyoC41hRFJGbZsYnzd1cc1FLGOQJ9PKJcxKxUdIFa5VYYqLNCq1dKTSdlBS70nH0Bb0JgES6NEVgZSRocW38j3XRv745ADk+yXyIyEkohlBAECrbiaB/6ly0gspYCisgZZVBlU1tebM58tl+TaFc9qFosFGzUe6clPuiXrhdKnrjucrE+lO3mpJQC7c57bgygTKLbgZn2DYL0RuMjhwm2QO/WR95P3aMc8WmFmHOIYtvdvXC7SznTELxspmZg4gATt4d5w7RGZrC0xo4upKE/N2rIlIVFg28Wg8c5hqTw1991JQpsutnCWGtLMfcnTYmMHnDVZE4ug9ioJPA4USdFEe06qwKU3CO/WmLU2yT4A+nKWNXW2x2qDv54hZKOmaAF9nd7m32wCh14nUGF09KyROV6+AEcJ28ycpAQ0iBSst7IGVsZVVb5kNx7kZPu9/BlWxqy+QMhdZnT+sl6nM80ukx75YaoyKv25m6yFTQTBkEwV1O9L6TN0dvPBcZCAeZNLaVpcrnhcrYxalr/lrqeQXxoY5LxbpZGKfqfIJd9RILf8VMn5fvx6USVVopUmOlErdDJ7zPyjK5UqwmdcSYiPeaVW2zzWW2raw8vXO8HTpe9SNKc35TryrL01wzWJGFqvzmnb1BZ2CqUAlThDMoUxRi/v44tPnYPaOV41x8wbaRC827qRFLxczSeNUs5zETBAzrjGRgbbPh+2UliPppnD+90TSwNsJg8CnnzqnEtrLc2YpGy+rm5cURpRKf3W3z2A8HKxdjl1kfRqfzyahIGSyrsL44G68MrqQuAj6P0oLKa9bGcY2YfZ8UTDZKmoiNmqtWEHatErejrItOSdJt6ahCQVs7XNCsTje/QhIzJr/iRTfyyW6PUolNWOgay7TIuue6m3l7XPHpccV1Lcq+u6WhN/EMzIWkucg/6/6wYpUR+EJF7uZaYtGzyTnIjeHt0J0z8M6cXgVrtZyTtjf1Qsg3Rasi81JymQ3j348dNhZc1MuPqegqLbzaE8/2YMuzlP2qnTnakskZNrU9NwJzVjJWOnJnK3blB4Man3RmNUSxLS1kVXBiQLTZVP9k8nNiFdX55iyyac/eVbSFp06BxuicIlKczdfRcgO4m5pz07JkhkmTRM5+MqO6qZezmrXIQOzJcrMpAl9M9RkAb43cjBOyGntZLLzoRQr/995f5tWWJG2cbEZ/Zbv/mMvKT68E6ecX1H2j63kF8VzP9VzfyDpJkX+R66MuwNt2ZnIl67yGmB7WAKwae9aer1u5u3Yry+PQUuhIXXu2YUED24uJdZiZpxLnChZrCLn7OFHJAJyT8fZFN9GWDl1Kd9wYT1kEAWp8gdGR0oj5TtdanC8Y5pPhiDzmqp1xTsxjQLo3YyJtcqxXM9Yafviwpc22jz8YWjoT+VY3YjPdSewGodKeR1udlVo3zUyVrQlPPOg5m9BcVg6fZJVwVc9Aw9EXPNqam9qe44Pe3q8pdPpKhyeULgEqS4oiiBE55tyR+KjpjUwXNhZnmpJEFsWz78ZFK2ufh0U6yV03Z58MlU3Tw3macUGzrhyr2uKjxELVxtM1lqZ1XAKfP63Z1QuHueayXti2M8elPr+G1hdYV7BZzRzvK5ra0c/u3EXG3LXOQdYMmwyUreoFXSR29UJjwtnY/apZMhArFMKusjnhusrUPQEr6yJQqMh05uCKKdSqFm+QE9f1ZEUqE4OhMZ7LTqa4lBQuiJH5yV71op3xmTsur7m8Bvul4uBKPlkNgEwspzzDMRS8bGYKnbidxMgeOP/7+0W65F21CGhaW/7Rww6QDrQtPTaDjDI1uvO5URaS4r3L70Wbx/8Q1Dmiag4F192Ei5JU/jg2jD6dj0NUqJI/WGVwtVCy3HJRc/AFL2sngHJWdK5Kl9OfZbKLSZ3VjC/a+TxB+Kjo2+VjLis/vYS6/Qtdzx3wcz3Xc31j6xc8k/PjLsCDrWhLR1kGdJFoS3cGQtpMkfFec7kZAQHXrDeUpdzVD3PNcKhZMv0MoK48RYjcj2025k6YIvKUgZ2+dNJdHw2miBwXoToplbjeDDhnsNZQFJGUEikDBKsMftS543mxmbk7dJgiim9FDmX0XrqdywwUMte8bu3ZiPrgKjGg9oa18SSUdAiNRfEB7Lu3H+JsNlkJVBeBFCTdty0941BktzXpTEISe0+rc7Jxs7BdTxyt+D7sl+qs9tq7EoV0doVOLLHgshsYbMWuWnifyfQVAR9Ntt4MsiPOu8R1aXmaala1ZdvMQq2yskd+sTvyeGxJXlGVgWszMi8lm83M+4eeLx7XYvZdL/TtQlPL79jUEopamoDWiXKOLN7gDwWN8RgjVqObeuFuatFJBAeReFY71pXluNTUOU7HZrP+637Enqw7i8DnhxVdZXPMlKUpNLWR3fBJeVcWgW0bCFEzuZK68rhZoqlAunybFZB96Vm3C9NSSpRWDpDNWCBVEXiYGgqduJ9roTzqwKZe+PSwktMlh4qa7OtwU9uzw9vtXJ070lMKc6kjb9pJnjPbrh5zejPARTvzNNf4pFlXljZ3nEUGZ0PUxKxGtXkHfHrPtKWT7l5LKnm0YvFal5mKmXfirTfczjUXlbxnuxxMepqM1ov4k8geO+KiwQbwSXG/1Kwql8VFEsxpwwfAb0wFm6uvqQPmeQXxXM/1XM/1h1PPK4gfr4uVdLbHocYUkaZ2tI2lKKSTrPEcpprKe0KQzrIynrZzPDyUdJXFOtnNqiSdltaRqPRZvhiiFnlj1FxvBqG1VYGYpY8qE+dvtgNaJ2KMdL3l8bFFKyHX17nzsq7gMAtBv1rFM/OiKgK3Q8d3rp5wXrr2lCS2/lvro6DS2aFrVy1UJtDNTY4IL/mlfqLNx3g7N+wqy6YU17bbuabPktw5fDDWdlGzLsU8uzcuf+5kJ5oN5K2XnXihEnOQgNDRG6a8ax69oc77btyHUNKTV8UpDl0rMb3ftTNfHnt2zcInq+EskklJUVfyGq3bhXkpGSdhetxsj/hQUFceUwTmSfbym9pSGc96I4wJpcDUETsVrBpLt7a4peB6NxCj4jjUVDpyOIqv80PehcYszV7VlpgUD3PDpl6ojadvF7RKTE7Olaep4c31nmkueRyas+ggRHWeIE7BnVURMIWIYMoyECK8HVtU3h0XeVI5MUXa0rNfah6HNnsnC71x08283/dMToItbSy4qGb60nHMe3XrizOl6zELXprC88VUc1m5c4zSEhWFku64KQKhFJP2iKJAWEInxklv/JkJcZq6TPaGPu2FxRhds6osizdsmxkbCmKSiUGrRJ0xEufkNXwcGla1UPeMiUQnE+LrdmL0hohilXfMIWqMkhildWnxSWc6oOdxqbnOIbbvxlac4ErP/SLUwpt2lunIGe7e9oC4kf+8FcPv/T3/PNdHXYDLMuakaHkR1xcL89GIL0TjCV5zsZ7odxa/ZKVWkfBO09SOca7oGovWiX5rSRHsZDgcBdhzrsBmSlCXR9y2cwSvKUxks5nRh4QxkRA042RkBJ4MVSmcYO8L+tXC/UNP11iuq4FuZbFHza6f+PJpRVFL+vLbxzVXq5G2cRyfKvrGMmYQZ/ZiRL2qF5SCt0NHpQO/dXVgcoZj5otWJ4P5wlPrwK9sDgx5hbCtl2yLqWhLx2sduZsaBl+eL+y3Y8uLXt4ApwvFqrYMXt7s68phfKI2gV/77h1+0Xz+bsuum3kcm6x6SrzsR8ZMXatLz6abeTi2fLI74JyorVaVFf5xEQQo28zEoDhMNdu18KOtNegiEaOiaeU19V642zEpOUYT0UXCzZoYFGUZUBqqNnuDtInlR3JqvRtaXm2OdF6zeINRiWNOZXBBc9HMtLVjsZJPdgLKbCjYtjOHoxgAaQVX9UzbOOrGM40l+7Hh6GR90JaOtrUch5phrrjaDvyLu5F5KlmsnB/9eqF/aJm9JLasK8tgS7776gFnC+rg5cJ1ChNwcnwhalyCdW0ZbcnD0mQucWLKRlQnc3sXxWxq9IZt6c9UOclHhJAKSkSxFqLiwTcfvDiKwBfHnpjEq8FkvvYJiEtKcXSyXlIKnuZGfE2KgAuag+1pCp/zBcWU3UfN7dDRGs9w6IULXy/MvuCqnc+rOBclabsvHb1xrGvL0VZs24UiryNOtqeFTpRBc9lOmGx+3+QVT1t67gZJy/nuz3dtyh3w12ht+Q2s5xXEcz3Xc30j65mG9ru/uQoUZaJd75mPJft7cd0yRaDIGVBlFfCLolpFvAsss2FzOTO9q2T0bTwpKtyssYuh6RzGROrGo5R0t1pH+o0lOE3VBWKIQkNzibiXNcR+bOgqAeia2hECtL1lmUq0TmzXEyEofCiYhpKmleffNgt1JSDFOs44X/A0NOc7eJG7gZBkNXAa83e1ZcrKuKOruGhmEXE4AQdvZxmRX5iRTW3Zrie+fFhhQ8FlN7HqFw7Hhr50JJfdqmwhCbcZTJm9ofOyovnq87uo2XYz1VqihmpzGkelu9qtJiHL58mkKCLbq4kU1blDLZ2o5CStWv5fawEtd/2EUommdnivKUs5600dUTqxrSauqsgXn20Zx4quswyHUsQ26yW72yW81cSgqfhAbat14DDV9JVlv4hCTkQMYtB/Asf2S00fHRfric/vN5ApjfuxoTaeQkfaMjDNJd3aYkzkejdgbXE+b6rGY5aS0WrqzqOLRPCaYaplUhpLXBDLSbJQByB4TQjS6c/Z76EyMpHsVhNfPKzF8yCpbIx/ykULZyHIycUvJFESbpuFL4491810prVJPJVEV61KS2MCL+qR41Kd1ydfTi2/unsCRNDUZuFGzC5+S9DMoeZ1P2B9wU0/8jQ3Irgwgb5d8sqvQ6lEWyq6vNa5aCf2c40NMuUd54rWeIkTCgW7RkC6F5uBeSnPMWFFTo5eV5bZG67bkbKUNYcPBat64f3Q0ZZCa3u1O/z/uhj9f0udz+lf1HrugJ/ruZ7rm1npuQP+sQpeYypP0Saq4NkfGlb9QlkF3t+taGqHKQPHfUMxRJrOsdouBKdoW8tqPWMXoYy1lx77heGwb6QjrQOFidTrgJ810cMyG7zTrK5FPjvcl6xXC+3GYR4jISiKItGuhRbjF43WCWc/mEPvR3EA+87uASbY7UZ07tbtbKgrT1157vcdj0PD1Xrk7iCy6cel5robSUkxJvGInVyJ0ZHBlmfgJ2X5Zcg70pAUizXcbAesNTyOzRn4unl1xE4F+70c1+PDjjZT7W7WA3XtcU7Tto4vH1asm4V1sxCCRjegl0TXWEl0znu+41jjo2bbT5Rl5HCs2d83LE48BYLV1EbI/bXxvPrlA7qE8Z2hWTk2O8t0q3GLoaoDKSqq1mO6iH8wKJ1QWqhgzhWkqGhaR2EiSkOzDcQA3oKpIstoWG1mVM5beRrFMP+mHyl0YrQlXeXFt9ZJgKoNmhArTBG47CYex4b3+55CJ7489iKmieLn8f0fXbGpF8pSAMmUFN3WEZxivZkpTeDpoaWuPPNcnoUBzhdc7EaOh/osHOobyzyXObg0nVOuF2foG0vTOdaj5dXuwHFsOLiSbWUzWFayqy0pO8aFpNnVM7MvmLNbWqETu3Y8e1Gvm+UMFI5LyWA/yPJ90LzqxjMouG1nyEDk3VPHi83MVRC6ZFFIVxyDTGqCrUSqOtBsPWs3U3eeh/cdMYOVVy8H1tPM01PLdjsxjVmKbw2FjmzWM9NcMi+SYL4tPcbI1PGiH3g3yI65zHhLiJp1I3/bi3aW9OdOdsZfRyUghl/sDvijIolMGbh/1+OOGm81Fxcj01QSs/rFe42pIse54vapp2wibsmKHicm2GUV6LaOoobty5m+X1jfLJhGgB0B5gq615FubdFFYno0xCDrjar1zEfDPJc4Z9A6UdSJooqYOrJ+Y+l3Vi7mjefl5UGYAdkIvt15pqFkHkua3vH2fs2XD6sze0BnUGtXWy6ymqks5aKwyonDtQ6iMMux8CEq2tqx6Wb6xlIbz7yULIthcYZ1s2SAUDM8VeeTd9UvfHdzOANPbWdpN462l1VASGKjWJWBqvL4A8x7Q9vJDWfVWNat8HJfv3zCOcM0C/d1zmBiqYX37KPmejdwcSm+Gv6oOOwb5mPJD//vLcFrprlkmQ2mCsSgsHuNqSNlE0HLmF9XHl1EUlS014H2ZWR8b3BHTfcikCKsXjm0gWUUDjhA0zi224mmEXDpNPKXpYzNfTaZmZx4LSxR+MsXq5HvXj9ysR1ZNbJ6eLk9sNuNjEtJiFo4v/sSXSTsbGh7S4qK/bFB6UTTOhmVVwvBa+pabrqLkxvwcanEurR17DajKD2NAHL7x5ausfisTNtUltqIMU+R2Rcuai7bmSWIauxHYyeJEjrybmz5Z/dbPn9aczc3jLY8r4DaUixaCx0pVOJiNVEWkdFWcmNdxMDncd8SouYw1RgTKEthsMSgmL2hzbzztnfMU8nbH6wFOA2K69cDIQoPOgaFLuT8NlVkczkLy6gMmdsfWa1kRTVmfvgw1YzZYN8GeZ6HY3deGe6uRqrKM1hJPV8WwzhWjGPFz10JUlS/r49/Xut5BfFcz/Vc39j6RVfCfVQH3H0SuXpzZBkNT08tzorD0jIbmtZRlYFlFLDiZjucx4eUFBevJ1YvLc3KE4PCD6BL6G88fs4qplUkBUXdeVSpsJPBlLKaUAq6F56ijtSdZ7We5c47VDx92ZASmCaiDIxPAsQB1L3n5as9074kJcXj24a68TSd4/hUs21n3tzsKcvModWJy/Uo/hMq0TSOeSmpKvFEOK0EyjJwOzV8MXTsuplpKYlRs96IwmzdC1/WRU1VBjaZPztOFcNQi+ubK86RTl1jmcaK4VE65KZ1fHL1xCFbON4fOo7vK2LQeCerlq63wtc1Ee8KqsqzWi00rWOzFhBvtxv5pVcP/MpvPrD5jmU41Hz/H17y9//xS6rKU9YeHzXDIHQvYySSqbkQapmbNSnKa3X9ySAKq6CpWo8/Ko6fGqqVZzxUTLcaOxvsk+bxrmVeSupW/o6LNRwPDcdBujofRKk2zBXzUrLrJ2xOMC5U4qKZiYnM4wVri/PI7UPB3f2KtpRO/+EotCc3F5gysEwlSif6dsEUQVZnRWD/1KB1YlkMd4cOgP2xoa8sTSNUOF2kPGJnlWYRGfMxTk4y/RrjedFNwu/N3ObJGeoi8rBI2vftLFS1m3am1JG9LdlUloe5YbAVx7litCVtY/P0Jd4TUwZdh6ysXNdW3k+7I1UR+OH9ltunFcYIQOxDweINdS0dqc52rD4UPD20qCJRFoGuE572eJTO9uF9x3Qo2d1MtJ3l8sWIqSLVytM2lrZ0rLolm9XL+/83PnnPpptljeIK2sbhlgJj5GdWJtD3lqryVNXXY2OW8lrv9/r4mFJK/VtKqf9LKfX3lFJ/9ysBmyil/pxS6p/kjz/3+3iuP6mU+p/z//95pdRtft5/oJT6H5VS3c96/HMq8nM913N9Y+sPaAXxvwG/lVL648B/iMTTo5S6BP4i8K8A/zLwF5VSFx/53H8tpy//MSS39c/+rG/+qAvw8GlBWDTHoeZpbmRvmUMM56nEe40uEpebEa0TuoysXsh+SpcQJkWKED0oA/svKsb3huAUKSiWp4Kn9w1KgzJKns8I+HdS8CQv//WuoO0tm+3Mcax5/LIjBljuNO3aUXXSxdipoKgj1hZoHWn7D4GB3hfsLkfaraPrLBfbkeNQY61h1csuDGC3GRmmWu702Uui0JFX3ch31scMIhU8Di2fvd0RoqZbW9pWnMWKQuhjst8sebsXetq8SADnxW6kbjyby5mykg4TwJSi729qx+ubPdaaM91M63QWqJgyMk8ldeMxdcRUke7C0rcLZRMpm8Dw1uAeFVpLQvQf+eQ9Ve0pysTVdhB/jqAIQaEUFD0UtYCl1S6RPNijuG4djjVFHamuE921+PA2nUNp6C4swSmaWo47eM3N5ZHNdhKhSO3oagEdH+b67E7WtI6r9cj1bpDg01p8kqs6YK1BZVpb20qnenV5pMyCiXWzSFd73+OddMHrzZz3lPJaF0VitV4wVWC1XrLbWWLx0vVqLbQ0+RtoRlsyzrLHnJxhn0G1qgg0taPKYZxP2Wj96HL8Vg62FDVjcTak32S12U030mRz9dp4hqk+x2UZE6mN+KpcriaqIgiopSPjWDG57PRWRPZjQ5uTs9ftwjBUfPHllmmWgIGmdtS1OOe1rcNaAX6HLLoRsLdAlxJokKJMkMvB8PnDhs1aXAJPe2pdJAoj759+JcBb0znBTIyIh7b9hHNfX08n5/rv7+PjnjcdUzo/qudD8NG/DvytlNJ9SukB+FvAn/7dj1dK/Wml1D9USv0d4N/9ST9DKWXycz/8rGP5qB1wvQ7Yo4zUL6ojzcrjF02z8uzvG3SRaFYe0yXsQROdZp6guQg8fV4TgmL3asbNBcnLKgEgOA35grJ9MREWDRrWl0sGV2QsXh7kAq4NhKAouwhEikcB/naM7O8bQtBsdoIWl3VgOcivWTceOxs2rxeSh3LweFeg7Qe0tSwDdeVp19lsppdVxeQMft9RG48GxrkiJE1IkW9/+wk7FqQE01hhXcGwr4hRY31B2woouOlm5kVktgLueao6ULWBZZBjVDrhbUFhIkUZWa9mmpUAj3XtMVVgPFaUZeBwrHn57QPTU0lde4oy4pcPbwBjRCrcrDz12pOCwlrDejPTXnoOX8raobuwVI3HLYbNtyzJgz+ALqBo4fBpeU4eAeg7y+F9IyDlShFDouwCd5+v2OwmTBUpSlHK1WvPcjAUpawWDtnsu28XXgCV8RxDwfFQC6vBiLLwMNXctEcANtuZcaioK8/DoaWrHTFoikL4qxfbUf5eRaQwHxD4ZTYURSLkG9rDQ8+LqyPLbHh1ecBaI8BxEVA64ZzBOVhclnjrmNO0OQcGxHzzezf0hKjOKdat8Xz/0FNn2XGfpcMgqdW/tN2LApAirzIEiDVFENOjbqZqPDsv5+0wyc0pBM3D1PBqe2RnJt68eELpxHCQ1VRfOdrOMgw1ry+f8Jml8rhvaWvH4X3D47ElJsUnrx6ZhgqTed4xKtByzgHoStSMm9rinKYsPf1qwZQBtxiGQ83Fq5HooryHo+LuvhdZe1SMQ8vVdviYS8rvUer82v0+6lop9Xe/8vlfSSn9lZ/6zEr9O8B/DbwA/s385U+AH33l2z7NX/vq4xrgvwf+NeB7wF/7XU/9Z/NK4zXwj4G//rMO+nkF8VzP9VzfzPq4Dvh9SulPfOXjp158AVJK/1NK6TeAfxv4S/nLP2mX8bv7698Afiel9E9yF/0//K5/AFldUgAAIABJREFU/2t5tfEK+PvAf/6zjuPjWBAaTBvp04JbCqaDrB2aJN1iaQKH+5p2sXhX0Gwcw1NDSo527RieKsaHkqenljdXe+o6Yg+aZuuZHg3Ba5aDoWoDx+8pxmONc4Z+tVCuInFRoBO6kO7Oz7LyWK8WprlkfJDxqywDdjasrhamp1Jy37TQ3AyBsIBpoN9YiloAPrGlzPaKS4l/0FRVEDVeFdi2CyqPqSl5tE602a/CzQKMlXUQzqVO7F7NMrI/ZKNqV9CvBbyz9gNPNgbFfDSEoHn/2Zar7cDmtWV5+nBvfLwTY/vTGL6+EPNya825uxdqmACVX35vRb8SD4vuwsnfKR/nCWByR03dOoZDfQYxlE7Yx+z3UEV0JZ2wKQPDoZYJwhr63tL0jvlBU4WILiE6ePlrR9xeUV0m7D3YRXxCYtB89pmAZm1r+eJuQ2UCkzNcXgxstjPOFvzo/Zb4qLhoZiYna4cTr9h7ialKSTEuJduLCTsJBSsl+f2bTtRudeMxVYQBhrHCGBnJqyJQlJF+vbB/ajFFYLV2xCB7xKKQjrdvs21jk9dVB9huJx4eO7rGYq2hM57bqeGyXmhLjw+a39jtGVzJLkf+VFVgu54kDGCuKAuhm01AVURMEbjajtS95//5/g3OFZgifuBvh4IwK677kX69COe3C5hVws6BpnekKCup07rHZNrfdj3l8yLx4urAMNSyIlpb7m57Lq8GobI5qLfyGD8o5mPJqptpe0fVB+a9oewiprWUkydYzeFBzoV5EmA7BiUKVuVQ6kNH/fNW4oO68+cppdR/AvxH+dN/I6X0+flnpPS/K6V+VSl1jXS8f/IrD/0W8Ld/yqH9zEopJaXUXwf+M+C/+Wnf99wBP9dzPdc3tr4OFkRK6b/NwNgfTyl9rpT6NaVkoaaU+peACrgD/ibwp5RSFxl8+1P5a1+tfwj8slLqV/Pn/97P+NH/KvBPf9axfZwXxBriCO8+7Sh0Oi/jQybUz3NJv1rOf5DpsSIlKJvI0/vm7D1QV/7chZariN3LrjDtFd4J4b1sIpurhRjENe3hs5bVdsEtIr6YF4k0ajt3JpArnWgrR9EmhtuK5WBY5twh5h1z8Jrj5w27KxEkmAb8qKhq8a2426/51T96R/Lw8EVLmROVu85StZ7Hu+5MWWpqASdOIguA3dUkdLooXa/SiarxxCAULZ99B+xsZHd+19KvFvoMwBUmEhbOpHldJNrGnQUWphJRiSqSuFQVkapLHO5r+sbiR9hejBRlYj6WjA+yv11mI6GUVwN2FgXiqz9yRJsFOxUCppTx3P1Gp5j3Yl3Z3zimIWHKIFaRSQQyuklEB0T5O9onzfGxopk9zdbTrS0xKMaxYtuL78bJpvR07oB0rzGKNeLL1UDTuPM51NxE/GcSa+W9qP12b2aKGp6+J0KM+31Ht7izJ8k8yevebyzsZVpq147jUJ+FQSkqPAWtccxTTdO688dp330COfeLTHUnOl3bOGpfsKstV6uRFBWHuaYtHYsvsgeDAI1VI6KJQkeOtmJVWa77kdV64bN3W5m8poKLZmG9kugskLH6pDirG0/3KjC+LfB5ClQq8fC+k/OykcmgagNuLpgnQ9348+c6g3LLKCKJVS9/+25tWQ4GU4micRkMZe1pVolyI+rHrvJn8FxpmI+GJYOiU1bMta3j6SBTmrXF1+rfEL/G5/pK/Rng31dKOWAC/mxeJ9wrpf4S8H/k7/uvUkr3X31gSmlWSv3HwP+ilHoP/B3gN7/yLacdsEY66j//sw7koy7A85dZsdU4jmNNMZU/NhabnJRRdpLR5UaN0gVmleiXLAddZPy2R0N16YmjjEmmAfe+YHMjfNn6WwXT78gFKCnoVpZyFQlOLqI3vzRgdgp3L2nBYcljtJGLgnOi7NpeTMyDoMenN1SIiuA0RRnZf1GxemFJeyUXcz0QJkVwArZ8+n4LwK6dqRbD7mrk7lYksqYUs5r9U8vF9UhRJ8KiGG5Lyjqc1UCmTjy9L7N/sYzS1hqKWcbN6cGwtTMhaNrO8v6zFRc3I+UqsryVl8i5gv3Y8J1vPzDtSwHOannjLYOm31qK/EbSBspV4vioSQlWO4tSnmmucIuhu7B0W8f4rjiP7CCvgzIJ08P774t8erHCxQ5BpMpGx/PN7+KTCRVhfpILXlkFmtbTXgn7YNhXfPqw5Tf/2JcQhUWhjVxcBNU3bMyEdwXOF/zamzseH1vataDrMWiI0Gw8XeVJXpoAVSqOP9BsthPN7JgmAeiGsWK3m4izrHSUFnWYnYWbvtnMaJ04HmuaxnG/7yhNOPOL56U8S5t1TuxWOnGzHtjczNjPhDHhfEFlAi/bA9Nc0bUW4wy7DPy+eHMEnShXsP+04vEoSraLdsKYKNLso7A31peyWnh8W+MeNS8ujtSNrLj2Tw3rzcwyG9pBFJzGRJZJ2ConlWFRJvraUl0m1H3I51xEl4mm9bijrMf2j62khARhvwQnDc/62544Q/stR3KJ5Z3C7TMovZH3d1EnlIOq8ayVAHhdY9m+mEhejmeYarROVPUHptHPU6f1xtddKaW/DPzln/Jvvw389u/x+P8V2QX/7q//VeCvfsyxPCvhnuu5nusbW39AHfA3pj7qAqyU3DWNiVxdDsyTdBbGRIanimUxXL0cpLtdedrXEfMQ+fJ70tGdRpwTfUy3GqLYTCoD6+uZ6SlzHdcO03Pmn8agKFro29wJXWiSlzFJlTDcifqp7jy6TOzeSCftR03TO+wkP9vUiXGMOFsQvKbbOnQJzcrRvIL4fXWmlDWd4xO9Z39suLoRZZ+34j1Q6MA0luc0YDcXLKOMgULbKUR9NTZsnaJbWfEV6Czea57GBu81V5dHprGiW4v9ZkpgTGB/37Bhpqylo7lZHXm47Tg81FSVgCOHfcNqLf4GZSPdL1rWOmhY7RZ0mdCV8KFX/SJm6iXEBQFZxpLgNe3OiqrxYDCNFzOhV0ehFxpEaaWF9qVdYvtiYnkqKMpElWluuoLD24owyZhsrWFwhvtPO7pe+OLaiNXo413HqpEVhdYJ5woO+0Y8EB6r83k1vhO6U7XyQmt80syTKMAuX4xMgxZjmrzaqteesgncvl2xDAbnNE3rsZPwg5sL+V6AvhGOtvMF01RxcTPiF3n+EwilNLjF40ZRfCklx7qfa5oMlr17XPFid6TqAmu/sAyyRji8N+KT0liGuRIubRG5XAtP/jjUTPuseMsqufW8sFotOFuIjWgT8S5xuK1lgomalzcH3FKc7TaVhuAUIF4qpo64WfPlF2tevj6IJ0cdzuq0sgo83bXiPbJ2+ANMj4byEPFLgbOSzFJXnia4M6Wyrj3TXHL5asCNBf5YEZ1mGcTEybmA88LP/7rq2Y7yuZ7ruZ7rD6ESzx3wj1V1mXD7hNIetxS0vdCYYqbwrNYLx8eazc2MWYtyrmwi682M6SJFZdEVqAJSgHCI+ANn28LoNN2FIzrFMe/bUlToQhy57r/fsH09447SOUcH5YVi/zuFdBo7ixuLnNAL0UrXblYJ01iKGuYHoReVtYgwilZ2yNUuEa0IN3SRWEYDCOgSkoAQKSqmseL6+kjwGrtI7pZzBdNYsrmUrnuZKgoT6VaWqg4cHptsXA7tzmKPhpUV/wFTRRiRztrLrq5pHXUfUEVCF4plkG69acQCsqwDD+9FYn4ySK/WkflBwL3NJ5a4SPcQrKJoI83GEZ0W1aFHdpRbWIeFx9uWcpLpoNl6dAM3r47iCdE63n8pO+/rT46UDyIOCVnxZLqAHzVFFdl/UVNW4ojmRsNmO/PrWSAhYhrN/r7hOFdcbUfGseLufsWqk2icVjm67Hj2dGi52I28f79i1S2YOhKcYpkNU1aohayODEGfO/TxocJUYqZ+ONZUZWAcKhYrp/rr1V7+ztUHdzDv5Zzwi8bZghg1X96u2Y6ipktJ5Wkh8u5xxa6buVqPoiQrxNUMMkC1GO6PLZtGaIAnsNHkyClrDZudmOVfNiJaiEHx4upA9Sj2mtXK05SeT7+3E0pdGYRGphLbfqbq5dinUWiXJ4C7XM1nEHU6lLS1o1xFdCF0tXDQtK09hyRUa4kYMzuFLj2HtwKaOyfvoaZ3mCZixki3XXj3+UqsYzfw+GUtBvZHw+3jShzkKs+yGOz8NfV16bkDfq7neq7n+kOrX3A/9o+/AIdFMw9ZgNE6CXdsxYjdtJH0qHCjJoXEPJbYOVG3jjAplIbhVhDzqgqsXlqCFfP1dBB/2s12kp3j2vP4ZYf3movrkelQ4nwBEcoucnhb0awcyw8162977KOn2oHOyLBSEJ1CmSR74JuIH2A8VrS90JyajaPoBVWfP5Mu2E4ixwVPUSfKbmC7aJ4eWpraSTTMbM40tO7C8v1/ekWZ/R6ufnmmqBfGh1JYA1VgezUxH0sGW6Hf9izW8PL1Ab9ollEM6pfZUJaRsgu4sZBd3grGd+KjPAzVOXZnGQ3bC4kRerzvRESwUhx/0Ij+P0iabPSZOnSnhWlghW3Q2iycWVnsWND14pLWsxCGAjN+8HEuTKTvLDEqHt52XF8dzzvU5sJj1jDdS2e9vllQRnwFHt7XYtyexQzTUPE0NGJ4Hj+wUS62o+yWnwqut4N4NJjIbjOJub+taDPdLzjF0yBG9m3pz/vjkydEDGLwHqNi1S9nKa8xkaZ13D/0oEXG7sfyg6F5GUgJvNPMc4n1ErUzzSWrnXTfT+8ajJFk7c1u4v6ux4aCq25AKTgcaw5DIx4YJmBMZFpKqhwOUBSJqnGMx0p8mBcxit++mDAN3P+oEU/ezJ4AePlqLwIkB8dDw3olXhZP7xr6jaV0cq5rnQhBUb8pGP++5uTE2/UiKz95bLedOKuZJBFfy5OmfR3xj+n8O0i8lz4HrbpR4p2WQfyZ2052xlXlqWsRXVyux3Mq9vb1zO0PVh97WfmJlVBnjOUXtT7qt3N7eTEXKzzDYRCDmuA1yyRKNF1ETJ2wGbQwZaDsZHy0o/gSbC5nNt+xuKMo2Uyd6F4ELq5HmouIaSJ+0mebvPoisn0jkeVPXzaMDyXrVxZvNavvRKKVG4O9zybivVCd7FSgC6i2Efsoaq3rX58l1+4yCH1mgPCUsFPBfKfpLixuryhXQinb3zbEoOj7hWblhU9pDXfHjuMgJkCVCVxuRoyJuL1iuKsIPgNqVR6dF4MPYqpdFgE7ydqiyhSwuvHUvefwviHF041DLgrLYri4HkXzf6gxpRiil10kRLk4H3+oz6O+3WuIiuNe/DlO3On1pWSsOSdm6MnL89drLzTCOp1N8bur04WzzIkjMuovszxXtfLYvSZZaC8lf21+MvhBVhxayYVxmkrWLxd2n8w0Riw9qyLgnWa7nc4rlVUlfhliJC6/c/Cal9uDcE73JcOhpi091ov1pluEv2xM5HHf8sPbnRgmZbDMlJGyCpgqsN83rLqZpy8ayjLyOLQsOedtdzNlGmXk4mbk1Zs9IYqi0c0Fx0e5pI1jxbabsbMYza/bhbf3a94/9rSN425q2I8NfbtIOoUJFIWM+d5r9o8t+7FhmCsq4yUtZtGERRLHh7mirAJur9h/WuGWgsfPm6z8DNS9p14L1/nw2EhAQaasWWtwt0HO0ZeRpnf84O0F85NBmUS99kxjxTyVrK5kPVLUkeMPRa1oD3Ip2I9iYh+jwk4Fdja8f7cSlWgj54EysPnE8rgXS9qq9ngrKryHz1pujx23x5/pwvj7rj8IO8pvUj2vIJ7ruZ7rG1vx61E1f2Proy7A7z9bYYyAam4pzgqgCz3Sdo5m6xkfSuxYnO+ipspOS0bG4uo6EWcgZrPtCPV1yqbfCT/KWPh433H5YqR5GSGCbjX1W+kSi0poVvU6EI5yR/ZWA1F+jmBh1L1klakAT+9a+vVCkcGc/RcV7dqhCumMCxPPa4buwhKdHE+3smeLQ52dtQod2TbzOaZm1ViOQ83l1UCK8lzd2jPvM02vTlSV53oz0HaOaSzZ7xu22+ncnZaNdG1tb4lB8/SulQyx3SzdSpl4vMsROxeeuCjuvhBNf9UHihaqSehp9UVEN1DvfXZDc+zvG6EnRY1zhra12UMhEayAaNNTeTbMd0fpiDZXQqtaXyy42cuYPhlRun0iIGpcZL00T1l16D4o115858jyVDBkt7PDVKORjjAlRbsWoKdvl7N4xYeCrhO1XYwarSP7Q8N+qfnk6okQFT+823LTjzS1Y5qFthZRlJUoGo2RiKrpUNJsHPohnalU8StJ0KdILeC8Cuq3ltdXe0AEBzFqbFaoyVpDMcwVuxwhdLmRHLdds2C92JI2xrNZzxQmMk1ZkGMC60LimE6rGV0kUlA4p+kbK6DqhcNUsl5Is5GVXgaC/SyWoE3rKMqY/UQiL74rPhzBKfxnCtNG3lzuqVaymvMD9OuFaSjxs6Yo09lbYroXsLzQiZvtEZ/XOPt9w6tfPtBtXVbCSeSTH2Uze319FJrcohiPlUzFU80f/Rfefex16CdWegbhnuu5nuu5/vAq/kSDsl+c+qgd8En6WHby36ryXG0GLr89nbuIU5JxDEILM1VkuK9oXka5236hme80YRIakbea6UvN/fcbHr+UaJkU4fqTo5DeHwQki0ukv7JU20j9RnN8W+bOQDpYU0XKVcR0EVWJ+bsdi/OOrW4801ARFtg/tuJlbDVFLwKP4DVN7SSl92gIi4CJICCHWwwpwuZmPict7zYjbeMY54rNeuawFzN57zRFLccUPXz+oy3/7PaCtpOuJUSdo38033t7SciOaLoUYr13GucKvNfc3fYcjjXDY8Xj2FCaILvzUiThZReo32g++0cbfvg7Fzw9tdK99wX9lRWjfKtZXywUVeRb109cvRGf3aIX/1elEvvbJp8Rice7FjsZ1t9y6DJRtYFql5hGAV9j1MQg0nTTfzg/upXF1NKVtTvLardwuK2xs2G9m+nXCy+ujl+R+yaOjxVulOd8vOsY5opVpm7tH9sc3BnZ7UY2tbjALd6wbRYuLkd2b2baxrJdT/zS9SMqRw4ti+H+XUdhItHpc9QQwDQJZdA5cXYzJp79RWJULIP54BCnEqvNzG43sd1OzEvJD+52cq4vhsvNSNV4HvcdRsuEAdC14qkridLyoXWSeKlRfq/jU42dCspdosiUy0JHju8rxmMlaQ9JUa4ERDSN+D1bK1PTMhqWIcujc1CuaQWDqXawfbMwPVbsf1jy8FlLkSe6E+3TdBFdJwFcdY7fysEKIB4augRdi69IexnoNlboenP2NvlK9/v01GLDh0Tyr6P+IAzZv0n1UR3w9sVEdGK0XtYBd5Q/dsi+KsOd8C1PFxrTSZ6XOoKuNbrORtBBcXxf0W0dfoH6IvJ4Z34Mze5eKt7/A7G9U9qDlvErHsEcBSk3TTa9GTXNVWR50Nm4p2Z9sTBPhjho1ruZFBXrazFxufmlI2FSDIcGfrBQdoHuWkCl+UEM4INT3N/1XFyO2fAmUG4S47uC1y+fKLvI7acrLq5Hbp96SQtoXObuiiG9nbL5TlJctZMAZJM569sf9y2b2lIUAhjZQXilMcgNKDrNdCgpK+GyrmpL3UguXljU2Xz8+D3Y7UZi0JgqYPea9z+SC6qkGCfKVcQdtTBPJkm9ePiecLi3L+ezB8f4INHwZe0JEywH4SDPR4X3RU7lDZgG3n/ZUFQjuk6kCUwrnh7KgDuKL4d4FshFWUzaE5v1LH+LRZgfy2gEtKo8TeOIUXF7v+Lm8piZDZrdlUUXA2UTufRiwK6LRJgURebynnIAX3z3iNIw3RmqLjAfTyb0EgxgTOTprqUsZR0xz5L5Z4pAlZWHRVZVvvtizXo1Y8pI3QdebfaoHyU2m5lxrOTxo3ii9O1CW0uidVFE3j2suWSgNIHjWDMuJW3pWVWWxRrWq4Xb+xXBjyzWcJhr3rx4Eu+HbDI1zhU7N2ZgVt5nde1xtmB9sxAWxe3bFat+Yftty3wn74X5VmMnUXS2K09hIg+3HW1rMWWke+GxT+Ltcr/v8rpHsb5Y2N83uMyNXh6EAVSawPsfiKJxHCR04LhUfHf9wPpCDLhebg7cfrri7Q83APzax1+PfqwS/y97bxIqW57n933+w5ljuHHjju+9fPkyszKrqrNa7ZZaEgIJemFsSWCMjaHRStpYG9neGdsbC2QwXnojMBhkLbyQwAaDJGwjLwQWlmQN7u5qdVfl+CrzTXeMG8OZz///9+J37s3sVHV1vu6Uurq4Pwh4L27EiXNOnPif3/AdFEP42UZB3Lcg7uM+7uOnNv4wZ7dfJ17r9uJauau+ejHj+SuB/Fjr6XaW4rAnm/XM57UIl6fCzLr+PGP6oCP0AZPB0Aq8xTmxIdIWhgqWD3YUsw47EUhaaMMd7MnOwRaicRBPPJurhPzIiWTiypI9Crha4GbFE1Fmi6f+Dou5XmXMHnX4XqBX8cGIW91rSKYObaBZGdrKMpvXrFc51S5mOhEsaTzxYpnUCizLJpLZ985wc5Xz8Gg9SjnKBeMGzWefL2hbS91EWC0i8enSY2PBiGoTGLyoX0WxQxvPdpNSlxFlGdPtLH2jWW8zqlKGOLNpg03Et8s7NbZfRKWsGzHHUeq5upwwmbQUxa0CnaFZGZqdZfUspW8E5tV2ln4wrM9SdlcJ7dayHUXwTSSqaCaSTDCb9eJCXSc0TUS/E6U0ydrBFgEdCc4bDUbcjliX0lZKFmIrZSLJVm9ucl7dTO/Eu9NsuIONuUGT2IGuFWumvZOG3WXM0Bm6ytyVyE0ZsbsRRtZNmVE3EdtNKtjXAaLUoZNwJ1J+/XkqA9KiI0kG8kknGFqvuFgX7J2Ik3TXGqnwGsPhyU70QozHZHJOZjPR6NBKIHAhwLoRL8G2s/RjGZ7YATP6zSXxwLxoyLKOgwNpAWV7HXnSYyOPVoHYOF5ezJgfN+JpOP46q1VMUgwMlSZKRZkuLXqihQysj9/YMn+zwx5HeKfobhR9o9lsUpGq9NI2zHMRk79tlSVLsRZK7cDqJhcIXyzMxaJo8V6YggcPhbFXtxFnl1PyomOxX3GyvyWeShXWN5p6HZHnHc7rbwS/e0tF/jqPP6xxnwHfx33cx09thJ/xIdzrCbKnosd7dLTBJsJMurVEuRURN9bTNZa4cAy9OP/aRYtbe8xUC9jfOskgIsAHhlp0EOoqJjuuUIkitIFi2hIXTlTTgOqlZGRKgaulx5jOB5TWmCLQXAgULXhNsxK21fHpFj/IoC57EFCxRsUaM2ZFzdrSd4bZaYvZBdrS3gmux4lDW+mDJlN5/eywAa8Igzj3TmaNCL2P2giuFZ2BR6crgfMExcGypN5G+FYykalp2N6kTLOW2aNOHIvHjFQngemsuSO0FHVLmvXU1cj8qywX1xOW85LtLuXhuxuiSUd5EXF2MZVtThviyUAYFN4Jk6naxAyDDPfqJsYaz96yom8NZ1dTBqeJrSM2jigSh+Nh9FfcXKdMFy1pLv3ZEBSry5yDox0mGfUEZqLN3DaWZCN6wDaHB6drtjcpm2eQzXra0rAuxdF3loiSW98bug6yXCB/xX47moQKPAoE2uedZrsRjVzvFfmsG4XEDY/3xew1zh2u0+jIY5JRK3rQNHVEPulYr7I7jYbbDNPowHJaER0oqpWimHXs1gmL0xq0mJPGB4FhC2jI5j27q0R0rUfI2ywR0lA/2mRtbjIiI0OrtrKERo1EGTFdLfKO5EBRbFvSucw+hkEzTQaihcJVgeFKsb8oRfFMBapNTLHXMV/WmCygMgOIIqArwbc9SskAPJkO7KlazFCTwO5S9EnSTPRE+o1iaBX5wYCJRnOC3N9d467TTE4H1p/HRJ0jSQeeX884me84u5yODsiW6NpxcTEliQaiSKqKg2j3e1yOvhLhHgf826Ir5QIPHnFkQBZd18tFqpRHKUUUZLFKZz1JPuDWMrzKcdgJ+FYuiCjv0Emguko4/F5LflOJbf024FvBSOpYhkzxxGMziPcgmjR8/sGc49MtfaVI37Wo3BO3Pb7lTn5PFgHBEYsAUEBpTfuZOFTcvm7xrRZXysS7OBQM8DRqKS/iOxGeahWR5EJPxgTQ0PWGzXosseOBJBNspxtE7F3rIC4Bo+PEy6ezO9cDHxRaiQznZi00V6MFIxpnjrayZMc90dqRzmSQZpKA22hOj9fsbnHEO0W8F0hnPdNGqK16HEQPvSLORqZd1qNby/4bNZuXIuyitBzzycEGY6V10XuN7S0ma1g9S5ktW5J0EFqzEipquhBH5Vs3k2ji6W++EOZ37bgI2FuqrOB4i8OO2AtrsGkjppmciygS4RelZaFtt9JS6TqLsZ5mJ04O+0cV211yd2311eg+sT+Ah+RABof9SnQ59Tj8m745kG0HtIFq5yjrhJmtyR8F0rYjvhhIDz2+/kK0P0kHXCv+drc49LAS3LMtwKz93f5OHg7cfD8XEX8jLMu6lMWorSyrm5y9WS1uEwuh0cfpQBiCLIAZtGej196kx1WBzfOYJL91tRDEQ5LJkPe2xaOsIp45fCf71G8UyZ4neSel+s0Gc+t+7LiTNU3mMmy2k4DNA9GxYf3CUuy3mERaN+1OfyFpetDdfd7j/oaqioVlaD1J2mBjz3xa3wkGmSig9Dej4BBQuPsh3H3cx33cxx9M/KxnwK91e0kXDkZ+vB+1HWwuw5VuZE61lSV4JSI7vcYkYhGULkbn1UpEckJQUmrVUvL1l3Kmmwt9ByfbrRPCALtNimsVyaHi8sOUVx9PePTummQpMpXNxy1+I95VrhV2VVyIAy4edCKf26/AlzK8GzpNXcdM3vL4BnQkgyQ7Bby6E2kHyJYDwSs216lk6L3sd5r0HDwsmc4amibCO021FUm/oROMsh79u7JZz/HDLUk+MD9omC9q9g5rdCZiQ8W0ZVK0TKYN21VCuUtoLjQ3m5xWOs2SAAAgAElEQVT1eSrHuRb8p7YitJIverarhKtPU7qdZXFaE+8JDMxkEE89yUKGRyBMpqFUzE5b8kWHTT1xJoM7E4W7wUma9rgWFo8alAlMjvrfxkjqNuKo3O9Et8H3kiFtLxKiWaDcCIYVxraUEfxrv5PhZ77o7v6ulAiTl5uYZicOvFHqmey1GOPJ9iT7d16kLOPIkc4HbDrqVBx66muL7xW+g9WzFKUC21Uin9eCMiNTb6WZ7TcUWStuyxeB609S+tbgWynjh0FTl9FdhYAHnSpUaiQzH8XWs3k/DhY17ZUmtsIgTObiul3MOrHfOup54ztrYXBGAZ1LKybKHfWZOHH7BrKioyxj2srSrUejABNodpakcOSLXgaiJ5poJjC/9tnAzfP0rpWSHEsGO7ySNllXW0wiw9FsX4abeOgaK1jxHHYfBPYeS9XSrjT1tSGeiQCQK0WMp762bJ7FOKfFcGGQ1uLtNZFk4sjctBHKjvjk6pvJ7QLqaz1+L6GU+uNKKaeU+o++9NxfVEp9OD7+4tfYxi8rpf7u+O+/pJS6UEr9qlLqXyql/hel1E8UxfjZzu/v4z7u4w9tCAri6z1eN5RSBvGF+z+/9Nw+8FeBPwn8CeCvju7IrxN/e3Rffh/ogF/5SS9+rQW4vIjkju0V21WCNp6bFynJdCCeeOGYWwF5m9izu4mpVtHYR2XMHgV+NjttGbbhTgfCO2GvJQtPduoJTjHbbyTbHIW8L35TCBaLg0p6ugPYXMD4/XXAZGGECgWGRphVaBF/H2rN5iKlPRe+vFJw8t4OXwfMVEmfKwOl1ahJAclRwCRehjaHHfODhqef7NPsrMB3MpFjjAuHtaJB0LYWG4mmQD2acqazAZ0EkmPu1Ma8U1TriOalp9rGwhzcF0WwphV36aaM8IhCVbOz4nZbDGPm7zCF9FznRwJb8r1iKOU8OvGHxPeMtjriZGsy+XyQ4zWJZIvxfmA+rXn4cI2NHN3OoiPJpr0bdTuCiOMrDenCc3VR4L0iPjHYpSEteq6eZvS9RRupPvITd2fcGuWi1WFzycbrOibNBvb2KnFf3hfBfj/I+cnyjusXOVUlTMOLbSEEgMuYfqfYVOndoLDZWZoryShtJhBEGUBB9UxU8uKZp9rEVE1M3xgxgh0Mu23C9lVMs7Ycvldz8O3m7jtDQ/MKNt/34zE4IZpUmurcslrnXJ/nLA4q6d2vDUpDVwsssFkJ89OmUomETq7ZvjLY1OOdZKImCuR5N0o/CsMsfQjFfkdbGtqtwRaB5rlk5smhIn0ScfCdlqHRdDcKlWrQ0J7LgG36oKOvNOVFzNBAuUu4eZWSTgaUhtBB/nDUajECHVVKWHU2hehAjZWcIp30TPbEQPTB4Ya9Zc1mk6JsGC3GhBl69WLyjSqU/WuEof2nwP8KfFm44t8F/n4I4TqEsAL+PvBnv/pGpdSfVUr9QCn1D4H/8MdtXCllgQJY/aSduM+A7+M+7uOnNsLXfAAHSql/9qXHX/6dtqmUegj8B8D/8JU/PQQ+/9L/n43Pffm9KfA/Av8e8GeAk69s41eUUr8KPAf2gb/zk47vtRo16UQUvrQOzJYt9SZistehI7h+Jq0OYzy5dQyNmCFq4++mq3iBsukETK7YfabJjBglomHYyWuGLdjJOH1OgVeSge09aIjmCt+N0KRC01/Lv3Ui7203wodvyohhtCZ3taKtI+anDfGBBqsYrgbqC03xOKBTjS8dKlO4tVCBtYHNU+kF9lUY4UOa4+X2zm4+TXp2zyxRKlnerkrEyHA+CAVbCTFDx3L8m08Mn50teLC/GUW0Nd3OjqpxopkRJY7ppCWb93Sl4bTYYCLRV54dNtgc2iuBOiktPfds3mMngWEnfda2sux/b6A7l3M/lBAlkvXoCKpzK1TxEeJ18HaNKyEt+ruMRykht9gpXHyQCdTqtKVZGbKlo9+J/boygfaFbDeaQNb3uMGJottZzPy0EePHkTxhYukv5rnoVJjIs71JmcyEJr55KedgBNKgtaeqRCP5cFrSjTZAfWPYtAmrV7ko1BmPGQXnNxcieK5UYDaRXndfafxI3y4yIRkoC5NJy9Br7C1ioBFqfTTnjtDhekUydaxeiv4tvqOtI2zkmE9r0SROPdUqpusMJmpHyJ9ocLhe8eyTPY5PNuA97VbgdcrKNetKmT/sHXU0Z5o4G1EvM0vU94BcT+2N9NxtPhA/SNj+fx3xLNwZaboPevJTqS5UK7OXWxpxtRLT0c11io4EiREljuk7AbeSa7tv1B0JRueikw3QjeL8k5Me30oFZWJPHDl8Lyaqq13O4d6OPO+IkuF1lpXfMUKA4etnt5chhF/6mq/974H/IoTglPpt2/9xH/bVBsd3gE9DCB8CKKX+Z+DLi/3fDiH8J0o2/NeB/xz4736nHXmtBTiMmMpbSEsIcH0uPPK+N0wnDZOjDldDuRGBjij3lK8sxcmAXRrKjwI0kIyi2avPUpbfaggDlDcxtpShkJShmvxoYP+oHMtfcLtwdzHesbUSWSyUlvJTG8imPVrLxTK0ChuJdxmXUhoHD/HMc/3DRH6Uhw5KYbspJX/PFz3lVUyUDuxuYvYeiB9d6nrOV1Mmk/YO6rMuM3qniYyjvIqJs4Fk7tmdR3eaBG7QHE5LhkFTTFuiBIbejHoXUgbentcQ4OxiyunJBpMF2mtxmm6uhPmW7YnzbDrr0bFAkHQUSCYBk3R054ifi5ZFND4I+DqgU0W2HKS9EEsJ5CvBQ2/WKUXR3ekqDJea3HXsHdd0pWHzUiBg4ZUiXYj7tG8h2lMEJzfM28WsqS19bykvYibHHTMtLiHGemwsQv2mE/ZhmvW8Op9x0Jc0bcTkqEMbOPtkQtvbu3Pig/rC2dkETmY7trVA6iamo24ilocl61XG8rDk6qIg3/VsrhLKOmGxV5HOenLTY+dw87FAvW7prt4rtq/iUQrS0VeadOFFMtQOd7jgmxcpceLYbVMO39xx8yKl3QpUbl0nLB/XRJMGO1M0L+U6nhaNfL8ahs6QLhx2pmgvAuVFIsNaHcjmPfFS0Z5D9UNpR8zf85z9WkLfi0zn0IA5b8mOA2ZuMMWArwbsQjOsYCilfdG+FHhkFDua2tLVhuleg00hX3TibvNcbmbVLmYYDPtHJV1tiKtbZufoTl3F7D5M8EGxPCzpK3HNuD7P6Z1hknRcb3LmRUPdCDTzdZunP3bN+QZaGUqpvwL8x+N//zzwS8DfGhffA+DPK6UGJOP95S+99RHwD37cbv1unxlCCEqpv4O0Or6ZBfg+7uM+7uPfVAS+GU+4EMJfR7LR23jr9h9Kqb8J/N0Qwv82DuH+2y8N3v4d4L/6yuZ+ALyllHonhPAx8Bd+wkf/aeDjn7Rvr7UANzux0Lm+KnBe8+jdNfpCuPNaSZleXUVMH/QYIyLjIUCci80KOIbRakapnnTpiScNvpFsbHba3gH7ywvJls4+mpCmIhOZtx3aePJTz/kHM7GOaQVm8/yTPd78+RvMmPUp6+nqQBggOxbZvesfxhSRSPTVTcT+SSmZciRtEZ0p1p8mLN9qaFeyH0kxEM89ce1GGJXIAR4ttmLP0432LYPhcL7jclOI+lgyUF5YmjriYluQ2oHlXCZGt3d1P2pdpFlPVxuqKsaMuhFXLybiHLyNacrAthQLGoD52x3tuSJ4UR9DQ72N2HunQxlFt1Ykx5LxDlsZpNVningW6DfSLticxcyfdPhGsksTeWbz5k6bIc5ErvNWyF4b8fary5jLqwlvf3tD/SMZzHH9xc/k1qVZqXHQ1kRkdY936k4o/emLGQ/2N+JmnIoi2cGivFOU250La2++qGjrCBA4X9uKfU5TRRjj2T+q2HM1Zy+nGONxbSTDJCM6Cnt7tTALF614smUOvGJ7HZNWosHQlNFd9qlUIFtIReFaGVyGayfkiF7hHdgI4sRJ+8R4cYQeKxfnFfOsvSPH1C+EmGEijys1zS5iuNHk047oyKAzgy07pqbl+kVOmvUCIdyP0Gth+RW5QCCLaYt3I7Qxguq5In8IvvR3g+WoFanWofHgFHYSMDsZ2HqvSPKBvjG0a6n+TCJyk8mpJ69rLj/K5DtfZ6Sz7TiAVdjI4yvFZZmjVWBSiX3TrZ1Wobu7czT0hqqO+abi3yQOOIRwrZT6b4B/Oj7110II1195TTP2l/+eUuoS+IfA9770kl9RSv1ppLh8Bvyln/SZ9xnwfdzHffyUxu8d4/t1I4Twl77y/78B/I3f5T3/B9IL/urzfxP4m6/z+a8nyD5SKbO0F53ZBCZHPWUtvaEocWKIGEOx35IsPXiBxJQ3Me1KBKonx3J3VxriB4Z+pzCFRkdwfjYVGMxoAVT3lnzWkeUi9h3PhA56+nCNNoHyOsEk8OhbN5QvrGQuGxEbz+Y98ZEiOo7QE8P8UYvNoSxj5oua9FTcknUk7+nOA3snDWhIDz2mkKFhv5HhTTTx9J3BGDGSdE5zuS5oWzty4b1o9sYD61V+l+kaJbY1xX4nrsq9oa0j+k6otKIJK/Td1S6n68QyyMaSucTpwMFyR77oSWc93aXAADefxQIRs0I1vfk4pr8OQqstA9214uYso10b4oknmiuSI4FQaSNmpkMzkjZmnnQu+gu3QvDF4UCW95TbBO/UF2aZQbH9UGi6yVLgh81aYGuzeUOx3+GcaDxoFWDMLofBsNpmPHlwLcae8cDl1YQPXyxpG0vX2DubpKaN2G3SO4PNr8oS7spkhBMGqk6y2EkuurSTvU6GiIOmWdlRfU/dQSGTTGjPyXQgn8nQKHhFOhuIjxR2qkTvI0DxQAxUtQW8orq0JPnA3mFN24kgehQ7rq7FJdl5NVo6SVVi80D+OLD/qGLx3YFiLu7BbuVw17IfvldcbAs2m1Rq7iCV5K3OtrLqrlrK9geifUW69Kw+jNh8aqi3QoqIlwpfygykfqno1prJieiH5MueahPTdzLEjI4MOpIhY38tOipp1lOXEXHk2FykbM+SUYBfURQds6TjdLHl1fWUm5sMpQJp0VMcdnS1XNOTk57OmW9EmP1fJw74pyVeKwO+uc7J6p6qjmkGS/iB4vCNHUXW0nWWpoyEFVMK1nP3XLj8Ngls1pZp2qKNTOpv/eCGCycoCT8iJYBXL2bkaXeHH11d5iTxwHaTcvR4Ny46gvE0O4/OFCoOnH804VhvRwGXkf1lFb5y9Jee9kYTTzzL45J4PzCsYdjJIpm9pXFbh4rUXeNJxQp3HohmAb+SH0U27bG5aCC8WE85yCvqPuJwf0dTR3eaEKttRubkGGZpy/UmZ3osGOZqE+OcIit60oWj/kj49ZNpS1F0XK1z8iAlY5b3/8r3oIw4KFytc+ZPOnShibeOzU2GO9Ok257Jmx61gcm8JX8U0LlBJYb+VSdDy16YT1Huqc80k+9oVKRZvZSbwuk7W3QK2niyvL/znfO+ha3shytFJF8pKbNXz1KcGz3jsp5i2pLMHTcvUowJrOuE/UlNXDievtjnaLbDBUURyWttEhhaxWad3Q3b4szhnaIpIyazhngmzLe6iXC9pikNy6kItKeJu2OPBa+oqphiIl5z2V5DfKJx60C/C2wvUxaPGyCQHAV81bE7j9CxnO/NlQjaq0gw4wDVOmK9zZjkLftPGh5ONlz/SG4S+4uSrrXEiQjmawM2cthDWYjaNTQrJYvZYU99Iddit9NoCw/2NygleiGPD7aAxbWaamU5+JaheFBTfRiPjteCGjLWk0wHhlpL6f88EM/Endp7RTbvuf6RuECHQVh+w2AElXIhfontWr7vupSh+W3SkE9Fx+JiPSWLBmI7MJ/WZNOe622O0YHNOhM9k7UwFcsmZv1bGXuT+nWWlZ8Y7g+x1OTXifsWxH3cx338VEb4Q57dfp14rQV4Nm9oG8v+sqRrLJudZAnWerz37MqE+aOW6lwgU8l0ENzqSpPEg5TKRUBnUva0Ky18+l7RngXareXkwYbdTcLeQ1Hjeuv9G9afy1AmG9XGTCKc/8UjEdCunslA4nB/dydfOKxhd5XQV6JjYCKBrZlM3I51rHjxg4IHv1DitgFfSmvDJKLcVjy4hdopXC2WPlEuegzbVULbWazyTCcte7Zm6DXbOmE5L9EWsmi4GzI1bYTzgkOFUVWsGKjLSKB0o1MuCOsuiwbyvOPs1YzDwy029rx8PmfflzR1xN5xxWqbcV1nTH7YkWY9k5PRrboTjOnuR5pk6mhLQ38zEMeB0EmJ0WxE2c0WArvbrROydY0uDFpr1mVKsenIMtE6iHPJfs8/KcjSnrqPmLoGZeVYvFMy6DSCeQXIT71YGt3I0BJgOa3ET+8VpHZg77gmuh6dkHuDiQaqnQwiqyrm8NGO8jpGKcne7KCprzXZ/kAy3YplUxWRFx2bdYpbayaFiOhvdymTUZA+WXiuf5SysA2uluHU/tsN9jDCbAbaM7m+tRYWWL0Vxmc/GNCwWuUsD0SY3Tk5nup8LOUjT11FTOYt2vRStXlFW2mi1NOfSytrvZrIcYythLJOeDS7EcdtHHUtn5mlHdVnXzgS921geFbTXmn2DiWz7HaaybfAFsN4PXncRqyI6gvpKnatoT0T9mNSDAyN5vnNjBDg9Hs72nMxRrCxp28FP31rMVYULW1lRwfwgAuKTSNtxr43zHP53d2UGZOkEwy29jSDfM/p8K9Wbb/XuNcDvo/7uI/7+AOK+wz4S6GN2JUMnSGfdTinGBp9N3ybL2uaK83VdUG4EofX+bwWBo0TaxS7tIQhkDzWhE96UUdzimQv4HvJXrOiu1PjsscR+a4j3pesFg03z1OBL3UwfWOgu1Z0VyN8aj/QXgpkKMl6vNNcXk04WO5I54M4JD8wqGnMwdUOsxehYkfzubjSxkeKuHQ0F+L8GqWeZiPDD5NBkQ3wChYPa+bXNdme6EG8+kFBkcrw6TYLdE4T4Th9e0N5EdM3+g7wXuzv6FtDuY5l0Jh3dzZBxw+32DwwfXDD9dMU23mOjjbsNqmodd3EnJ5sOGhLbOQ4P59RHN6I63QVEe0F1h/FNFVgstcyVJr6B6LTMXnP4H8kgyLfiRuytaII168cVV1Q9+Kk3N14drsEU3nypmNvWeMHqTSG3uBqqRKUhvpM+skhKHY3MeUHiiQdSCcDk2nLZ2d7PFxu2J9VbHYps1xMQPN5T7Hf8enHS6Zpy2yvluHqRlTltrsUo/0dcy6fdNgCuhvRjEjHqmjXJiwm1d2Are97onGI6Ud1sjCIdU76MDCsPPWHA20pFUNXG9JZTzQBHbVEM5lh9Jei6Xx9VYj1z3FDu5W+6dCJyWm30Tx7sccbj24IHrHMmnRjL3qADUymLTZy1GVM11kmeUO30eTLnjDAYb4ThcFayCnTSUvyyGKynuZCtI1tKnBN0wbc9gtGZLOzomp3GhGGHrQn2jjWV5kQalpN3xnePLihrGJ0phk6Je+ZBNqXYsV1fSUVTtda9p80Qu7Rgct1QRH3HJ7saHYRSSEEkalviaxYbF2tc2LjOFrsaNtvSgntm8EB/zTHa52pahcTRZ7pg4760jA/auhKw9AZsllPcgTdlefoeCvCHlbEYIadYvLYo6cjq6kTQZdXn82YTRumJx3VKzOWs4ZsKYsrwPY3nOAdG6EbKw2TvZZoFmiuNDoWkZNmZcgP5H35ty1+N6DOPevzmCQamDz64gZw9S80s9Oa5JGhedpjp4IEqK8sw4eB1fWUvb2aZDq6NJeWdi3HGQLEI3OtOJS/bz+X48pSWQyurwrmcykXb8VNjPXEhSNKG7bXCb5XzB52VOeWtBXK9v6DivIq5vo858Ev1ritp25kKh1nA3vHFTYXNxAdQ/U8Jl/2HOsNykC9iUizHh1/gRvVUaAanZ5t5unPYXbq6LYaHUOSDmR70h5K347YfRjz5ME1USrokIOHJauXGdn+F+zH4cKw/+2O8jNFXcYcfKclDIHt57IQJCM+VJhfgoB4dLDmZpOzaWPeOr3GDZryWphoQ2nYn5Z3A6Ao90x0RxgUp082dDtpH7GFZO5QiSJ9Q7P7IIyC5YaTgw2Tk17o7oeGwrcMK/n5KgvBe8qrmGwmdNr4cYz/tMP24oBtRgt2V0NyrPBloDgciN+IOGKHnQtj0LUwfTdQPQU3iDRn3wr6IXgwiWe2lNbH0AfimWz35oOcurcMzvDk7StsKggUHcl2dCy/leJ4oDyzlGXCza+3FIfw4sWMo+WOofakS4/JoLsW4SplIX3TE+qAb7xs00D+jqatRCYSYLIUwfWpa8Fr8hOHK6G6lEG5iQJpIjctpYQxmU4V9SZwuLfDRiJk1HaWKBnI9jq6c0EEGetJ7cCzzZQ3965pz765Bfh+CHcf93Ef9/EHFPctiC/FdCE4y36jqHYx+UlNCE4chPNOss+fz+k+qXCtwMp0BBCwT6aEdYNvPGYesf3+wGJRiXbAQlOkju5aEbxDGbFYGUrENXkLzcoyeTxgFhEgjCG5u3uiI4Mp5I5u5xq9zAjVjqFRLB41xCeG0GmGlac/G8hHomGoPek7CSo27P6fjsmjgWELRSPpt+818dyTeRG7UVYYVtVLLVjl3LM9S8hnHTauaSsrljSpyEq6QWPjcSjYa6ZHcjffXEG1ickRZl8cD6SzgfTtiGi/J3wA/bVAhyaF+KN1tWXvUU9/HYgPFK4MxMmAbxXxzGMXmqyS/R62UJcJy/ca8DCbdoROMqwwyLeeLDx2aXCttFbWP4opth0HM/EgSzP5Hkyu2Lc1ysr3ES8VWddTv9BM3gb9WYdZWFQRk21L7BziQayBbuvH2/bNYmwXDDeB7atYRMeTgFvJsFIbyZbrG8mkTRYoL8QROJoFqo24YEf7juHKkR3LsYZacNC+h/TbGX7V4ksvOhX76u6cKd0Rn2j0PGF41Ug19XOG9qloQ/gG4kcxwQVCL/vqd6MFV6QYmoDvxVNQRwOTpXxGVniiRFQHw6BIT8V26/PPJkwfrQiDsAInaUfbWdKHmjAElB2FbRJpLahWLKqK4wFz5Sm3Cfo68O4v7widZ1h5ghPIZvaOQc9Fm8OvW/wQoJEhbvLI4NaO6SPH5UcZxV53J0gPYDIZkOoEtpuUh3+8ov4RLN9qsIcWtx5ozxXJqbA8k6wXC6xlzXzSsn0Vk+x1ZHlPthyoLi0+KBapZP6Lk+p1lpWfGD/j6+99Bnwf93EfP51xa0v/sxyvtQArG0iPoXwqgsv9jUB6Dt+r6W6gOY9YLDvMwmBgpKlAGALurELFcgdWs4TiycCwGui2Ig/ZXiqiWcAUClCoWKFiT/Ygwb1q0NGA0gp31UtGohErnQLMgwl6VeOrjuZ5IJ82Y795IH5gQYv0YPsUXK/JlgP20OC3Dr/uCUOH0mKDhIb9b3eYw4SLfxhIlu0oJA/tpRGBcS12P0OlKEZVsrjwZKeirWCmA2e/nhHHg/TCNQzdaI+Ua/bfaBgayZaUBjNKSw5XQmKZHHSiyhVJf22zztg/KcfMK9C8hBA02d5As7aY3qOUJwyiiOY7xr6sfG9mqnFrj9aS0QYvGhHhTIxLwyBwu88+XfDun1rja8m+daawbxSET8TlNn1iaZ8N6Bjio4CeRuRvO9zNgGqEHGAPYnw5gPf0OzlGgH6nmP68JfSe/pkM6xh1J9L5QH1jxXJn6bn8KJHhZgHqKhDPJWNTKpAfDYReerUmg/Sxwa4GyheWrPCEbQdaYZYRwfeoVONaT2wVdh5waw9evidlQC8zzEVPvxIJSkLAbwaUUQQdwAqbcB5qwqC4ucpIVyU2Bx0rlA0oo8gfBrZPDeXGEkKHTQNWe7bPopFFqDj5+ZLymWG4kmtHjzKt9ToiX0jGrTPFcCO942SvZncWoxJF+0yMR5WV4eB86vHrFn1UoGxHv5bf59Bo2mciGu9aiGMng8dMHL77RqOzwLAN2CPN0Vs7uldiRKDG1WD3mUXpQPUDMZPtGsvBWxVDJeLxxbKj3wk8sd8JLDCdbQlBXMFN8s2Nzu4z4Pu4j/u4jz+IuCdi/PbYXKRMB+kD+yCmnPH+QHcD6ammvNb4bYc5iOmejpPg6jbTcZjEkbyVQOdQsSJ6GGEbh7IKkwXpv/UeXzl0bmiee+yBw+3EzLBfiZ5qtC9Zgo68ZC4hoCYx/a4ne1Phrnv01EjmsXXoQuPWjmgCxRGUn2jASZbXeXQq2bQrobyKWb6jCGVPFCt0pjBJuJt2myKQDNI/ffWPM7GimX1xx7dzjUo1e8fVKIauWF9lZHlH6INksanDalg/i5ksWxLfiwh8YahfBqbfM3Q7sXmyU5hq6ck1nwdA9qevxAg1PxqIDg27DzTJ3N+JyatE0bwCkwXcpegmuxZ8HwSGpz3l54p0MR5/LHbxKlZQQ/QooX/WMjwvufpIKMYnf7QheWDwjSe0AV9LL7a+FH3baC66wKENDCVUq5i9t4Qq7WtP82l3Zw1kao9rNdlbGr91AtdCsvXlk1qMPueGfNHjW8n8Zg87fA/JewU6qwlDAKvRmWf6jsMsEtyqQ6caXUSYzuNLRzSTnuxtH5zSY+dSjbnnO8IAXWkwmcOPmTyjxoYuLPOjGh1B/IYmPanAKOgCKlG4q4CZgDlJKZqaQgss0s5hf1GSFA6dBPYPBnafjBrWC42KpTLqroUyHO0r+hs5d8k7Kcpq+mcV05OO7a9p4plkg0pBnDmaz8HmA7bbgof4SMwEiifQX0PyOGK46snfDfQvHTpVRAfQfO4wM4u+6RlWnmhfs/1YUTySGYpKHJNHQb6zspM5TASmUPheKqdmbcmWA9GB9LJDKxVJszIkUzfOfX7/ISiIb2ZbP63xWmI8h99tyd5URLnj4I2S+S8obj4VppKKFVHscFXArTrhqk8U0V4gmgm0xxSA1fhybCMkBp0bhouB+Eij9lJC4zEzuS/E+7D6514EdlayOA6VRir40O0AACAASURBVGmFnSmSA4WdgooMKrE0u0g0DyKF3zmKB45hG0ArdKEFV7kUF1wdjZ5XO0V3qcjeMUQLxd67PaH3qMQQ547mOVw+L9isM3Fq6KVdsP2h/BhMFIgfWDk2D8Pa43cO12rsVNoq6YhHtkfxF35shWLvScvmIiU79uJ0ey5QL78TR4lb8RilFemJCNaXN7INbaQ0DE4kCbNTGcTFR5ro1KIXCUMnx5yeQPr+hPQNQ7SncGuH7wLZociFqkgxfeIo9lt85TFzw/CqlTJ2Hdh72BBFcqMMQ8CXARUpdGGJDg3ZgfzozIHIEPbrQHCKyYEsvvWPAnYZkTyOyB+KhONum3B1UXD1LzTdlbzeZNCdyw0hDHJ9wOh2ggzI8NB9UsoNoAy4VY9ZxPSXgf5FCw6GlWM4q9GZQWea+HGCr8Ztjhox0maRNoqZKqbvBuzSYJYpemLByPdZ//A2kdB0LxzNKzknvkf2oZfzR+fEQWQh3m0As7fl+9QRdJee7NiTv63wrae/UQxbaQm0WysCUN/OUFbB4OmeVrLPo8N2+t2C9K2Y8sKyvsqIDwLpH1vga083mgxExwbzsMDOFKqIMQvxY6wvNPVLBVpRbyJCJ78pUyi6c8H1+2r0bKwD9lFGaGURLi9irn+U4tvA6nlGGJCb/pGh/ATas0D1UnP1WX7HQlWWu3bG7y8U/ms+/rDGfQviPu7jPn5q46sqeD9r8XpDuEihpxE2baWUQxSZzj6fcuR2gMHkCqUV0bGh/lSsb5JH8jHD1UD/VGxdTKFh8BAZ0D3VU0jrDdHjgjCMwxSE7OCd2OoUtsNORYXrs9+Yc/JgI+XlRQlWM33QUX0oGYdkExpb9ehvHdL94+eEtSeexqSTAV0o+suAzccs42rAzA2h8ZQ/cCQLAbyXFxGn36u4+SgiLhx6HrH7zLPdpGRZR3Ik50FnmtAFdp9Z+s6QFR0oyTLSiTCubv4FxJNBysillG1p3qMz2RdlkTZCC30X4dcKtQnkBwPJmzH73+3AAShWH0bMHnZEx5bdDzyT9zRYjTIKt+pRnWf2XfG+ix4k+HVLaDwEKF+MpWICvhNdjviBQUcOlWpUYhi2DjuXKsbtRHZQJZZh1WEPLX7nqH4g2WH+XiyVzarDvj0nm3bU3y8ZGo3/PNDsItKbHjRUz8Wz71ZXZHLU4VsZ5sZvZ4Rthy890SKIclshZXr+rmG47MS5+ktXrdIKlVqUbbELg688OhYihZkFQhcI9UD0MEIvM8r/d8PmIuX4l3rQCl/KcNfXfhxytbj1ILoYE42eetzKEQbJGpUN6BTaM1Els3NwW49vO9qtIdGintZdKpQR1bmrjwra3vL42ze0zxzJA032jiY0Hh05EcSfRfhVy+ZTQ3Iu5yo91bjSi7FB5yA2TN/oiC8d+PG6Hyuq4ATS1v6mCKlv/klN/jBgH+Wki1JgobUiXziGK4j3wJVBjAiMFxH5/UB/o9BZI47lRwmzxw1mqlHWcPRzDcMWYa+2Du8Mqpfv8/C4wi4Mw5XDHn8zPYh7Jtx93Md93McfYNwP4b4U/bWH61aIEnO4/KeabNpzNN2iI9Ct9NXMwqKf7JNU5+h0hJ5lFlN4Pv0nUx6+s0ZZj8oDZpliW8ewFShQd12SviF30ug0QhceFWtUZglPR3qvHs0GZwGTK5qPO6IDhZkKLCg6iXEXDUorsApdd1x9kDI/bgidI94PbD+22MSTHQRCH+guFTrz6Jklt47dJxqbBBY/79HzjGJVES0EBpcdQPGgpD6TY6s/HQiDotlFQtG8MOK2PPayt78lxpv5kUCJQoDuKhAtIHs/ZfNParo6ou81WdEzO4bpYYspoHxhiU+MQLscuErcj23kqc4ts2UgO/X4RuFLGSy6FpIThYq1ZE7O43cDw0Zo0bP3FX4rmW15HVPsdwwXAimKgeGyE+LFGqIDsEuNTj39K3lezxNCU5O9Y9n+hiO0jlA79NQSBg9akTyJ8B+LtkJ80GEWMW4t1NjJKUTbgayXaiB5x+ArBz5QfezJ3tLQebrrQHIgBpN6kaIKeU1oHcoojFao2BCqnvS7BWqeYrRm+MEVNAG3GtXCEgOxIdw0YkTZRviqpb8RO6uhGinBtWLyXo5bbzCHGe6sIgyBfqNkWPjCkRxpzHHO6pMe1zv23jV0TwXyZ6KAWzlUDHEaWD+NmZ507FHhes3m5XiuVx6lPTpV2IXAMDEaPdUkRcfVK8mYH/i1mB78fIQ7b3E76Tmnp1A90+jLnvShIplrfOnZ/hCm3zbjwNkDCndWYyYyQHM7gamBJ3lkoHREjwt0Wsr+F5LV+3Ik7Gg1Qu0UoQuY/Yh+3RHN5LrMDwY+/q19Hhyvyd5ShEYcr5tPpDye/P7Xp595GNprDeF0Itz34OQL2v9uR7L0ZG9ZkkeW+Z9MMQuL2wzQDZh5hEpECFzFBpVqjo63xEea0MmPym/HMvYXJ8RLRRgUofaYQmGOCvnyK4e/atm9tJi5RSWKo7dLTC4iPJfPC7rzgH1zipkb9MM50dtzunNPqD3+bMvyWw2mAHfZEb09YfZtz+VZgTkUSc1oL8jkOxsl9Rae/HsJ/YWj/pcVrlXULxW7Z5bq3KIiRfEtaf7fTn3TSU/+nYSDdxual3D9zxX9y57JXks8GejWWvCfkeguRI8L/LYjnnnmTzqiyMtAa5nd6RJky4H+zNGdSzFWXxqSY/HZmz5xqFjjK7BvTFAxMmG34EpP84kwBtU0wZdyKU/et+jDHPsoFwFyE8i+nQgD6wkwBMzU4DtInox+bNMIexgRBmjPFfX3S9zWgw8ke7Jf0t4I+Fc7GDzKKPrK0Fxp8NB8LC2lKBVG11BCtK9p1xpfOepnci7Th3IuQx9ITzX2QYZ9Y0JopFXQf9bQPu3x5cDu+z3lr1d0zzr8qiGUHf5iR/ABnSlQcqNz1z1+1TKcd0QTz+HJjvIzxVDJtRzN5KZuskDzaxu6axg+L+Wce0hOIXRfLAXuZcX0uCU/GuTGCOzOI1yv6G6kRWIfpOx9qyc6teSPA+lcXheCyLOGAfRU5C596ehftAyvWuKDwFWZkUQD2+uEV59OCbW8N34coxMwe5bsWLDUoQ2oIgIDk7c8oZMBn8kQk3UNwQeSx5bkvZzJe4rs/Qy3cnSXiuFZRXuuRIslVjKoHC3dhhc1rgq0zxy3O28yaQEqHYgeRjw4XhMXjtB4XCn48b7S9NVrLS0/NkIA57/e43VCKfXLSqm1UupXx8d//aW//Vml1A+VUh8ppf7Lr7GtJ0qp3/gx2/11pdT/pZQ6+knv//2fpfu4j/u4j39N4b/m4/cQ/3cI4d8aH38NQCllEPfkPwf8HPAXlFI/93vc7h9BzD3/yk968et5wh1HZMee5K2E6CTGHmeYqRamWWZxq5bdD7xklZ1DzRKIDbvv97QflKAVxdvQvhixlj7IXXbV469quqvA5I9mmP0IVwaCD6hI01169CJm+sYgLKeZxeQipxgvFY/+jAznwk0jEKvna4LzRAeSIbcfN+ipobtW+CYQmgG39hR5R9h22JNUSrCTFAbPzQ8NQwN+3XH20QTXKmwhOgXFg4H5+0HK18JiFhY7BVuIPGGoBRddbWLibGD9LGFznWISiCYe74RxpCKNu2rwaxmwAcwedTRXmuZf7sQdN4PoSDzz4iPNsA0iKr9nJWM34K56yVgGj11GuK0n2lPYwwg7l2xm+LxEp0o+Ryte/r0Ov27J5lIq9i9azMJI5rhMUEVE8sjg1wM61ZS/0eF3Dp2K3KfvFdGTjPpjR3SgpT2QatQ8BSuX1HDekS7cHbtQoElKvPamiuSRZFz5Q8ksg4dQ9pKpe4Gtdeee4WXN8KMd7ccN7qrBLo140D33TH4hJn83YvtCjmt4usOdNfTXoyP0Rrzk9NTId7WMhD13CtPvGSbvW9Jv54BUdvGRZiiVsNwmhvRJRHCyLawifUtgcFiBP/YbxWf/KKe+NszfHuiaW7W/8Qej5HvGS8tuctShjfgNeieDX2WVwMU66Nfy3u9+74LFYcXyvYaHP7fFrR1mX6oRZRHVs63oeehMgw+CfU61sOgOAqbQmEXE9W9YQgfN0wG/blGZReWyrfSNseo4FYU3vcywS030KMe1SlidxQjxsgq3EcahyaB45AidZ/JY2hl6aom/NcE7yB968oe///HZH4An3J8APgohfBJC6IC/Bfz7X32RUuqPKaV+TSn1j/gdFlillAKmwOonfeB9Bnwf93EfP7URvuYDOFBK/bMvPf7y77LpPzUuov+7Uur98bmHwOdfes2z8bmvxv8E/GchhD/1Y/72Z5RSvwp8Bvzb/C4Oy6+1APvKSU+pFUfX4AP9pWf9awG/7XFrh009epFBbPCrBgZP/gTs0qCMQueG9O2I+EnO8LxEZxozFYH07FuRKDvVkm25F6JBEJ8Y+s8azNzQPevBaFwlfT49t+iDAjPV+O2o37sdoBkwh5J5Jm/GMATydzT20BLqAXscMT1uZUCjZYBHKz29/V9S5O9GVE/h5Ds7iu9aXD3aCVnwlTDUQuuENNHA5mVC8OJKG03g4BcH8keBg18caDtLfWVRRvQMVKSofrMTbQyF9BHHLDGeeOyewsy1VBdDEDbb1hMdiDqbmqfyNwftpWL7LOLyH/SoeYpOFWbPogvp2fpG9CdUpFB5hD6dcfhzLb5yJCea/T8mlY1ZJDKM/KSmezpWEk3AbR3JkQiY60yjM0X6pkallvyPpJiHE0LnMA8mqGmK3wwML2viX1iSvj+5c7H2nWiCBA9mGePLsdc9t+jckCw8uw9kyNSde1QRi45vLxl/8lgy4lsSShgUepGhjybism2V9CvPIHu/IH5/TrSvyd6y9C8H9MM5+iDHzA32zQkqMbhVT6h7yheiYqcXibA8e9DTGLca6C6lLxo6j7vuMccZofYMa2grseeKC0cYAnuPW/AKnSqU1bRn4NaiR9FdShXz/NO5mMiOruC3ZBMzVfheqqPoQJM9CuhEY5dWstwQ5Lr2gJderJlrzBsTgdLFwvbUkehx6IlBH09YvNdLz9kp2mfSaw4rGWa7rccs7N0KFiqB5nVPK9ITqM6t9MqbL5aJeDIQnUYMW/5/9t4tRrLsOtP79t7nGicuGRkZeaus7Ozq6qq+sFlsNilREj3QYAzIYwMewLAB2QMbGMA2YGjgF7/4xcA8zzx65IeBIT/Mi2HAxngeJA9s2ZIsjXgRRTb7Ut1d1VnZ2XnPjIyM64lz29sP61R2i6B6usyWpsnJRQRIRkXG5cQ5O9Ze61/fj+4EOOvIj0Qip1ohfhP8u238u+1nWVr+8jXn82fAF865b3zq9k8+42n/HHjOOfcA+O+Bf1bf/9MmOv5Cfq2U6gBLzrk/rO/6pz/x+KcliNvIQv0PP+vzPdMCXE0s2Zl0RVWosENpoDW3S07+LCQ91YSbGpeV2PMZdlTKYrHVwuUW3U/QW0vobkQ1WGC6AaYfo9eb6F4Dvd2VxW2rhcvBrDVQ3RjvzpJM6cwtwaaHaofSuY40xWFB8WgkTSEEkJJfWKrzBcV+SrATotohNrPoboTuyBbbrCZoAzgnsJ+GoTguZHqqn6BaIUFHmgrVsCRYdtJkuRT9JMDoLTh/KyAbajy/ktLESBYZO5Myi11Y1p+bEHVK8cpbdqhIY2JHflCgm0amjkL5KmaDAO9WgssE+j18KIu/1/cpB6IRdqOFHBtfEd9WNNcLOju56KoB1Y0/+dIs0qTRSBloUWC6PsWlQ3cDVBJA6JE+nONvNaSB1DPMjjy8voe3HonioPmJW7TZbOLGGW5WoIxm9uOM2XdH8txLnqg/8orycEY514SbRizTUyfHc5CjjCL/YCqaZU/hdTTJHWn0PtWYx9/qoVvmenpQafB2OjTue4SrDpcWENRTiErhb/jELxjwNO5SkIjZfkmwE2IPR1QHEygd5Ufyw66bhvmjAj+SJlN1ssCWUExFU226noyzFzL95qzDlZbpvhZtul/RvFeXUzyFvxnQeMlnsmcYf3+BDh3ZMWQHFcGKgIAqJ3jMwb58f66w2HHJ4hiidcFrVhMZ9V4cOtIP66THyo+RK+XcMi05X9w4kx+E84IqheBOA9XwsQtLdTDGf32N1q81CTfqa3gmx8qV8sPrcivqipkl38uklBELsClZLykyD79p0bFh/NgQLMPsoYCSqpMFwYM+warGDjKyHw4wHQ15JbefMZ4C2T/P7bNCKfVbn2q4bTrnxs65KYBz7ncBXym1gmS8tz/1p1vA0U8+HZ9fnPHPgb/xWQ+4KUHcxE3cxJc2vogasHPutz/VcDtSSq3XNVqUUr+ErIMDpGn2olLqeaVUAPwmsoh++rmugJFS6tv1XX/3M17628CHn/XenkkH7HUNOrQQeqhctoP+skZ3fDa6BXbqMGsJarOLe/dYsId7Y/xf2UbtT7CD+TVnQQUa1fBl21Na3CynPErx1kUWNj326L1mcFmJqyzBVsD8g5xGV34ztC/SoMlJwPIbDjWvKC5lWmk+DFhKcsoJqLMMD2luqCTAZSXl/gzTlekolMJ/sYOdZPihQTUD7ECyJ1dJcybpyOcPmmJr027nYAWj2NwuBSG5rtBrCYGdUQ4qLt4LSTo5VaFINqUZFXulHMNejJ1NcTnkRyXR/QblUYoKFUk/x45kDr/5ekB8uqCqmQfB3ViOVVZhj2uN6pVsR5WvWDycoX1Qq23s7gX+eoBailCtCJcN8TcM9mKGyy3x1ztUB5OaxVHgLymIffw1H3zD0usVdl6hEoVuGqpRibfdhsqKvnopQll33XSLNjRulqPXW7i8ojqakB1DvOnQ6y10ZSk/nuIKd/185x822PjVQqAaIKWgjsHrgluUuHJea5ktuh8Tf6MJSYRq5ehpdb1lnh9qgpFolMNtHztIUb5GdwPiDU9KYfVVWlxavJYS7fLCEiyDtxpQnmWAonlXGA+Ld6boEBo7WrChgYDY3azA+I7FWBCSbqGwmbhsq8jDjjLaL4NZSyh2J+SFcC6wjumxR7uxkIasX6GbBpX4aOtIvuqjGj7Kn13LuapC4UWWcugIbhlQoCMEgO9r7KykPMnIryDsK/xYUQ0WoOX1iuMKFQ5R7RDd9Ig3Aty8QDUD0DnVVYnNHOFLCeXhnPETn3BU4kVO3uMtn6WkAASyVBYGpSuSV3x0r0GxO5KvrWkoBwVe11BNKsr3ZgB8ah/2/ys+Vd/9ouM/BP4rpVQJpMBvOuccUCql/j7wLxAayO845975KX//94DfUUrN68d+Op7WgBUwAv7zz3ojN5NwN3ETN/HljL8iHKVz7h8D//gv+bffBX73X/H3PwAefOquf1Df/wdA51neyzOVIFzu8O60pVm1EJyeCupmSOJDBXaQwmAiKupapOcOLlENQ76fiYh+UmDWm6hOjAoMdpDhJuJaqxo+5f6Y5kYJzVAeE0rdN35OGmbKaEbHEdXE0vslsKMSZx1eWxEswdILOSpU6KBmBQSG8EGXcn+Mm+aYjoduBeheTHGUU+yOyB6l2Ekh9UGtqC5ypqcBtlJUM6mT+cuKqFngrwfoRBM/aMpgiJFjk/14LJmShqI01+7I3mpQC/2F1HX1xyn+doLX95ieBkz/POXygwDT9vA7Ct0JqSZSQyvHDn/Nkxn8oxTVDpk/qrCZcB2i+zHBvRamH6JDIV+RxKiGT3VZoFZa0Ioli2wF6Bf6lINKstWGZC7Bg57UGYcp2X7B4pE4FqtAo/tNaWLF0nhTq21cUeGuFmRvT6gOxrT+1jLeKyu1DM2g+i3swpJ8I5ETJytRvsH0Qvw7LfKjEuUr+i/OKU8LqlGBd6dDeV6hOwFmu4O+tybyqsRHaRl+cLMcdzmVDDdQ5HsZ5d6UqGcJtz3JbEsnU5OFDIpU59JwKgcVLhesaDV1pI9KqongTFXDl8ZVIjwP3TJUqSKrh0hc5nCprQeHKrKZJzwRH07fjmvTTweRRzWqcKnFTXP8rYYMkzQdxQgaKyVxUlBlkCxnXP1YDo/Z7sh1cD5Hd4P62EOeelS5xusZsUaqHN669DBcYTFbLbzNiGJuML1A6G6dQIhuDUOwE0qG/O6Y/KCgPK7xar4h2BHCmb8d43LJxpPlnLBj8RJpCsrUoaO4dKjQ0L2T4W01UJFHdTRBRxp7PJYpvliagOUI8okWo4UvINzn/M/Pa9xkwDdxEzfxpYynOuBf5HimBTg7scTdDHyD6XuowFCdzrEjyZgwIjUqd68wGwmeb6RmOZLx0fnQp9ErMP0YNpZgukCtt1GTnPRJSfy8h5vm6E6Aiqrrrj6Rj/I0al6gAulyr/0NAAFXm1UZBsl3U/KRpv3rHbI3h1SpwiSO7IM5wR2LjrTUv6oFxd6M4NUu4Rs97PkM3azQKw0G/+ec5TWFijWt9Rxb1fXjAHAQrjjy/Ryvbyg/nlGO60GCQv776fhvb2MmtkkrCldnwotzTdCyOKsgkFrw8oM55cDhJRnzR/WXsmnxVgLcrJYDWOl+m45XS5DktWzuMNZhhwt0K8CVkGw7GM1kmGE7wZ2NUUmAd2cJPIM9GJJfKcwoxy0sLodqf4T3XAuAMBBZ1FNrHzdKMS1h/9rBHLd3JV34ZZ9yrgg8BaEPaY4bLVBbXShlGEd1YjicSe19kqHrQQ0dIiS6Tn1sPYUrLfMLD3SK388xQL63INiJ0L2Y6jxl9KcpyY7srEw/JPxqggo97JuX6F5MOZhgQoMd5pQji6tKqgzi54xIvzzFbBwAOY1bMuSjY8kIXVXD3ZuK9MOK+AVTg+IFDJ8+sZjljHzgaG3mzE49shmEoWTzNnVc/cGMqvRZul+huzHF45E8L5J4OlvzIlJFsOSwhWX+dootMpqveqJE2F9Q5Zo8NWLY2rKYboP04Vy4wvuZUPTu+FRHU3QsPOZ8r0I3EAloK8btDbCTgtmRYelbMhC1eGsClLjzEf6dJsFLbez5HOWJcUHQU2QXMgKenVj0uGR2Lp+H0mJ6vihMpjkq9tBrTdwkww5S2SWUMr4ePtMS9Nnxiw5kf6YFONoRf7XyIMXfblCdzqkmlnK4oJwqwlUoJ5bgjpQN3DhD3+rghnOi+4Y48rAjabCowQS0xp5PMfdXKd89oRqW+DuhuGpstnDzHBV5gIbQlwtYKwg89K0OzHPyh0OC13rY06mATbCwtYLZH6MTS3EBJoHyaCHA7KWS4lIuAh4OCX5pE5jJhFDk0VwtyB8Xtb8VdZMmZPZ2TrjqsPVvTX5SMT0T3zalHUlST4rdiaCowGgWH2a4zEEk7gnNBwEUFf6Gw40z7KTAZQ5/w6calyyeaOIVYU+YpNaJKtk+Bytykbh8Rvxyg2qwQAWa8miBzSG6v0qwKOUHJhV5IKXFzgpM6GGHKcXBAn87Jn5eGlveRixyvdOSwJOt+lM3Et2Rixbr8G41cKXFjnJML8LO56Ip9nPK84JgNYPNZcgvYGUJpjNZCJJQFtdZjupEFO8NxTeu72NalUDjD1KB6GtFY6VEeXJso/WK6Js9yCuRm1lHvGbBKnDg0hLdiXCzXKBIdiL63e0u6nKGzSaUI4hf8CkOC/KJxgSW5lJOY0eAQhQVKvYp98ZMn2iWvhWi+k2ixYWUMkJFfmLxu5Z07BPUOMzwXoLXSXHWsTgUHofxHHFZUU4t5nYLmiE2FcSoSWQ779/yyd+uaD/wUc2A0e/nOKdIljOGP7Aka45gVZEeQGuroJyA1ze4tCBYN1DKauRvGC6+r2lvFJhESlXOCRIVT0PkM39XUKnxSp3E5DK1WFxIE/L893M6z+V4fR8V+zAWLbE2AlUyo1zcyikEvpSWVCNZiMuBwOfjho8dLuTaKQuMr3CLCm/nmcqgnxk3POCbuImbuIl/DfFvAg/42Who3QgVGPx7bdmKZI7wwTLh3YaQpE6E4FUepNh62sYeXKG6DdnO5pUMQNxdhV5LGmoNH4oSL3R4m5FI3HxNuXtF+eGY/M0BbpRij4TvUB2MqXYvsR9fUR3U5KrHQ1xaYnoRft9AGIjVTD/6C/Buf0VjZ47pWUCeGmmwDGeolmRilJbgti+3LR9b1BNkgTTUdGLQUe0VZmDp+ZygWZJs1lmIhWqQycRXWuK1ZFKtOK/EPuc8oxwUqMSnOMjQ3RC7cBSnBeWVI16u8PvmGjru9X2COxH+doRuyFc13wMCI2L9UuRqpqGgqGlbowXEIS6vyHfnUpqYyNBEMZampOnHeLcSGdjQEH6ljYp99GoTnXg4C8X+nPLjKeW+lDOUJ267BAb/9TXsJCe6H+HfbUHkw9ElqptAEsOtdcrHV7iTEeXACo1tmKLbHqbjY8clOvHQrZD8UjF/VGCHKf6akW3sCxHusm66zXJBKvZj4VvcaYOCalTiZvknmTEyMYmT3YUOdU3ds3h9+f6UhubrId69Hlgn3nZ5hcscrRcd6rVtyEu811apRpLZelKZIW4XeOsRlZzWmFuJSLuel6ar8hSugnANquMZ1eOBODd3DOW4ti7SUoIAqE7ndG7n7J518RrQfUOsk87eiog3HF7Pw1WCGLULaSjqbnDdpGtvZKQDj/xSCHhPeSIuLWC2wGsIjN00FMXBHFdavHs9gnWDt52w8ssQvNRGhZJhn/44FupdSyZJZ6ce+Zm4jrvColqB+CzWEjkdQnk4IzsoxYcv8VANg8udXKMH42dbif6S+GtmQfy1x00GfBM3cRNfznC/+CWIZxtFPk/F9qYTk741RXc8CDzUVpfwuQDlOSGDbcXonoxEVlclbiJwdJUEkhlbizscYk8nIqaPA5JvtdD31nCTjMF36lrXix0xspxkQldbapAdVGIZ02ug+w2CJXlvKvGl3vngFu79A6qzjPxxSrBuCF9duiZi6UTRuZ3jBRabg4o83GiBf68tQxoDGQww602ar8jASL43R3siy6pqx+DwVGdCKgAAIABJREFUpQTli+zN325QzZCsJXNiNdOLMR0PE9Zmjy0tBLVrepU4L5uuQQcKr6Xw+sJOXgw0puNRnBZCV9tdoBJfpHBNi/INXs8TfkPpULEG38NOCrF46jbRd1cJXmqJvZMVN+b4BR/ViVC9BLXSpNofyS5kSXYodrQQLkXd9PPvdYUV4GlUL8GEoJabUAl3Fuvk+1sUkISSCQN4BrPRwI0WhC8l8hpJgO5KE66aOfTdPs464hcM0W2DyyrSJ5aLJw10J5LG3TiDwGDWYvLdGdlegZvleLca6MRQHqdUp6nscrTs0Nxght4RWZy5lWA2EnTiES9XwqjNK9xM2LuUjsWHmZwbDQMnl7KDcLU1TycUZi8QPafRtzrEd32K/RmU9Wh74snj+g2i+w1022P22LHYKwm3Pezc4q8abKGYv18SrQv7Y/CWj44VG60pJlGopQjda5C0ZAcFEKw40mON7gRSJw8MeqON7oQE64ZkvSRYArORyGBJLtm/O50wO/co9maoUGHqzwDCnlDdBvq5rpjZehq13GB5a45qhcJ9LhzGs2KyEIrN1dMdpoqFN60j8NbiOrNXwm9ZaVCOHekTS/rki6GhVe7z3X5e45kWYG+rBYsSWgIMNztd3DSD2QLVDsV1dztGb7Sl5FBazJKH8o1s9YapOBjsX4qWeFRz+57blH3EXJpHK79Rz9DsrKLvrKBaIXpnGbQm3PEJXutJ9/V8Lot9J5Bt7lhwlWQl3k6T4G6MagXY0QKXygnhbTQI7jRobIuzhD2fCfB6UaJaEfNDzeJJLsqIlUQAQz2Pxo5sIScnAaopE0W6oWvtaCVNPSuNG5tZqnOxTfe7CtOLwFMEKxpvu011nmHWI1mQrMN/0MesR1TDCjsrrx2j/a2I/Ki81mQGdyLCF2OZ1NMKOysEQt+PYTiTBtkohyyHMBD9dBKgWyEYWfjL3Svs3iXuYkpxWspCOstQ/Sa635RGTDeiypAF1ZeLt9obilvx5ZTizXN0Szri5VHdlQw8CAOIQpzvi263lP2hm+XYk2ndLa+kZDJJZarOU3hbLXQnJJsZWksLXFZKYygwVMczyoM5rgbEK62wo0y8xzZimSbTiuCl9nWHHoC8RIXCpNAbAoYJ2ha90ULFPv52A2+nTfxqQjmopBxjHXaSU759LsChVigNttVQGmBKnCvKCVSDBeVHU6pBTnW+EPbFaYqKPRq3HMVc167fGrOREKw4zo6aVHNHfmIxnnz21Z0p2bmj+niKm2S0XqnLXr4WUPokEP11ElAezOX1D2eCDH1tieCba2BqB+3KXT+29VxJfvmU2yJ8EzecU17kUFbY47HouevyUuOXupCV6PtrqFAxHsUUY/HKq4YFdlpe+wp6PXEeV70Ev43gOXc6qEZwXQr5ouKvkAf8pYibEsRN3MRNfCnD4XC/4DWIZ1qAXVaKhAzZgrjBjOo8lemv0MgEkM6hW+DySmyIujH0WmitZBLtspAptaTeQlsH4wkYjT0aSdbRb2GyEkYz3FldzG814GIklkGtGJVXcguM6E3LCv9r4v7hSosbLsiOKrFmAYox6GFOuFV90kBcVdhJQXZQEe3kMMqYTwOi3oLqsgCdEtxJruVYajlheX6GG9eYQS2TV+VZRrAstDbd0NCU5o/LLdXcYepmkdlu4dICsyaZ71MwvD4eU+yLTK6YWuI16i2l0N+ebt3d1aL+HgR/qFvC0rCjXBpBiS8Z9g8PAMmi9au3oNvG/t6PMf2YYn8mMqG1Dv5IdhzVyRRllOAOPXWNOyQOUa2Q4tFIGpCJZJg60ZJVhj5+MIJOQzLgOMS1mqiDY3TTSEM1kR1I+lFF8rUEtyjJzsAsT6Vhlluqkym6E7L8a+LlZ0cZekW22+RDqrIgeqNbnwcxavdCPl83rslvY8gr8t0ZwU4DdzVHrbRIf/+Q8LkAtxgT7IQsHmXY44lkif0mbjDDjjLCr7RFc57IlOFkGLL+q5U0jdebZG8O8XdikT/2Guj9jPK8YnbuES+XhJua+VtzwnWNCgxmJUAdFthRgbktzmj+ix1umzGg0JEj8Qq8nSXK0RXHu2221yYCTG8FlEcl+UVO8xsJzcCQvzsmuNOQkh/S8FORhysqVOhTHV1AKUS9Yn+GtxmjjCJ+URraqhNhj0QPXgwhaMVo60QvvH8pMsGulIzUJMVsJGx8ZYpZCaguck7fjlm9P6cYldKETnxxQH58iQ6V7A5KK5yOZkDjlc+mkz1L/Dw32D5P3GTAN3ETN/GljV/w9ffZasDlxzMRes9zvNWQ/PEM75U+uhfjsorsjJoYlVMdTMjenUpT43yMm+UC7fYV3nYb1Q4Fuq0VlOLcqwIjDntpjtpcun5NtbkkGbAVeReTVCRpFzmqFeJGKSQRzHOqt47QG21c7oheCNGhwl8PCPvi1Gu2OwKGb4kcx+UCV88OSoZvaVZeXJCPNN6WvN7TZpBKAvCERSGuwQrd8fFfWUb58nwqUDJAsqikdrzRkJn7aSXZ6iQT6ttByvC7JXaYCzO4roX7y4pwTUnWjfAlXC5cAeUb4Rs4R35SyWTaqEC3RFblLFTnGS4tUKFh9Bbou32YSdZsbjVROytiBHo2heMrEdDHMl13/h3F4tBBKTVbfzOE8zGq1yR40MdsJJhbTfROT4T5oxR3MZHa/mAKozl8eISaTOFyJPXzE3nt6jyjcS9AxT7VMCd+0UPf7aNfXMXc72NeXJHs7+4q3murApMfLeD2mrz3fgTrXdhawR1dSWMuURCHEIdU45JiTwhcaq1FdTyDSYq/ogUUHhiqYY6zcPzHhvSh1FJBskk7SLn8wwV2b4Dfdqx+Nac4LVCepjqakF9JE8qdT+TzpCITbN+pxLbIQrCqMBsNaXD1E8KOmFRWxzMhkOnaXbhwwrqOZUcZ3m3QX59iugF6LUG3Qmwuxpaq20Btdimmsquz4xJ7cMXiY0f+eIryDeSlGM9mFnMrIbi/hIo8TE/4I1iHu5yjVyUTDzc1pDlEPu5iUp9nlWSzG23QWrgdXQ/dT7ALR6eXUqVgn3pzZhXZBzO8rRbe7QSz3ZId2nBGsTtFv3pLdl4/Y/xrsCT6a49n84T72qpoeBOZSAtfW4JFIYvUwkrjohNRHsrFENyWRg2tGDvKpMjf8qCboF5Yh7UO6mvPS8MkCXB5Rfb2hME/u5SLa2UJ/2+/IifZw49hsyceW3GISkK8F2QbrOK6+95ricNAEuH9xivo15/DrEUyTefArDev9cnenSXcopTuNxC/1iRqlnjroagfkkC2+lmJijyyHw5EvRCb2tHWl5N7kuGtCexmcSiKA514mI4Rh4q1FrrjU55l2FGO3mwTfGuTsFUy3YVgp0F+ZrnaD9Ftj2rkKI/n6NWEcixwFbSCXovZB5ZqWFKlsgU1t+rySCsWCEzlSN+aURxk4sK7siR/e3YpxygOqQYL9PM9sRxvhei1FvNHFa21DK/pUA2DfqGPHckYtN2/hG5C+VGtWKkbVU+B27oVSkP1eAyrSzhfjosdzFGxptgdiedZJhpi0w0wO125atIcsgKaEemTEoYzmCzkx846GE3lhzowuI/Ose8cCJAoCXCFw354jv3wHO92E/9OUwDjo1qoW1o5djUy0/RCoh2P/kupAMT3h/LjnVuyvYLuGxo7zPD74o3nb0W4oiL9sCJccaKwmOW4rKTxIKGcydWjV5vo2JAdiy62ePOc6mCM1xPFjNJyjKqTKfN9pIxTO5W4RYnqxtLg7dSj256mXGjivqXaG8JgQrAMxcFCYPdrLZJfaYvrc6cBvTbpsSI7U5QfTeUH/uMp1UWO3loC55i/X8p1WKth7OGI6vFAyklrrRp65OFOJ+AbsreupLQTeAR3GjS/HuNKGB1HnH8QY6cl/poH233USrN+3wJh8rZiuJzI7WcNB5Vzn+v28xo3JYibuImb+FLGvwkwnmfDUZ5OKL9/AJcTXFpSPBpBvw1xyHQX8YZCpF5mI5G5/E4MXfHgUrEvyupJChcjObpXE+zDI1xaYIcZg/2YsjDCMwh8aESQFah2BGmGvr0E68vQiqVc0QglEz0bQ5qhPC3PX5RwNGD2owXlscztV0cT+fdApn/ShwvMZovwm6u4hdgOqYYv6L+iksbiaCGZigeM5ui1Jv6WZNXVsODw9ywEBu92k+iWZv69Eaf/UuDd5e4V2Q/OP8Edtnzs+QxmC+IXxGFadSL8JUXvtQK90sDf8IQh0AyJvtbBpaXAwtOMxh2Ff38JHTrmb83J3p5QvHuJO74CBNYTrGr8DZ9gVUNVfTIhN8tx7x9huqFknp6UitxwTrBUN90s6BVxCfZeW8VOMrGdmuf497qiob2cYjZbUj6qdx7l8VzKKf0u6uQckgj90jqmG7I4FHmUXVjwxKcM38OdjKj2hiID3D2n8VKIm0kJha0euteAk6G8hlaol7fQr93GuyOjaf69tsDfa82sHaTSwG1FMj1W73R0J5JMrxtj1pviPxhoQVk+vqpNBUTH6krHYr8iP3NiS5WEeA0n1kTDhbynvEKttohflCy8OhjjCou/5CivHK502ElFOZTSlgoNNEPxvzOSaYb3GjgnGEyXFqKv7bcEatRt0rjl8NZ8itNSJghjmbbDaMnwncO73ZTvdzglm3koT6yuiH38e105LmXF4PuKxn0PN8koDubMPrACaO83rqVpdlJQnmVSinjnDK9ncMOU4s1z1N1VVDfGWVh5NSNuyNRmOShhMsddzmCWUb43gNG81tWnn+xEfsZw7vPdfl7jJgO+iZu4iS9t2F/wNtyzNeGOU8pRXZuykhm4/QEUJWdnbY732pLVaiWZL8iU1GAszIfNZfRmGzdaMP39M+x7J7iDSyEs7U5AK9a/mdNay6R+OZmBMbDSwY1S8j/el8xXaxhMyN6dQBhQPB4Jme2DCxkAGc5lqmlRUhXS+DAdj/yohNCTZoNSxK/GuKyEsqI6z67JUOZ2E3V7RYhrvqG6yBk+ljoepYWiojyYix2QsVQ1clF3Q1ylaK0sKM/F4BPqWl/Dp7rIWby/IH/nEhUawte6uEmG8mRYQ9UoSK/vYw9HMrgSe2Ch2r3Ee7ABax3ilxuEWwbdECwjRcX8xGC2O5iNBvnHBfmZFSylUuCLHEz1Eoh9yr2RNOs6ESr2Ce40MC2N6SiZcrucQjOSiTRPUT06l11GJQMDdBOqkynV3hVohf9gDe+bW5AXsmuZLcBaVC8h2hBE5/SJFtnXYA69NuXBXHCQd1evAfhumkud+WAgGXpaUB1NsOdzOa/KCnV7BbWxhNpahn4bc2dZTuRuhNpcwp5OsHPI90Ri52ydIrXEqVuFQnoL7sZSr480Zi0iP6ow/QhXKdKRjz2eYA9H+GuCDdWtQHZy3RiKCpdVYljaCjCvbQoi8n4D/24LnWjKaW1iGhjcWHZR0S3JtNGKfCwTpfZ8Lju0szH2gzNoNeQ929rQIJcGLpp6es+iQo/qcEa1e0n1/hlL9wtpcnpKPmss/9vlFb1f1WC0DOX4ijLTlLtyTMvdK2kM5w5/qyGQ91vSrNMrCd5WA3c4pDoYE25qTMenuSWISpNoyh8eU+xOqI4m4CnsYE51Ov9CO2O/6BnwjSnnTdzETXwp4ykN7a9iEk4p9eu1S/I7Sqk//NT9/45S6n2l1GOl1H/7OZ5nRyn19qeec1Q/74+VUv+XUmr1s/7+mRZgs+ThKrFQMXd7n3SrWzE7r1+x9fKY6nSOm2Rk3z2DvIRuIrKyOiN2aUF5mhLdElPBYm+KTSuRvXQCdL9BuGng0SHuwxN49wkEHuWTCa5EMuonxxB4RA+WoKxHUrshgx8ZyXJW5FdcaUXnV6SDa+72hOx0OZesuqjQay3cOBOA/JJkia60KKPBM7iRMHR1y9D/lqMaltiLWS0rM9hMsbyzwHttlfKDIVffKQjXhCHsb8lsv9f3hQ8ceqhQyZixr7CjHDuYUx6KDY0diURN9xuY1YTxDwvKsxyXVeSXCrOaSN30hwfke3OqodTbVWBw84LO1z3KD4a1saUSmdPRELICdzDAns+whyNUr4lbCE/CTTLhPLRC0Apvp43a7qFaYoxKEmIe3CJ9VFLujTCriSgvBjI4UQ4roV61GjWn2cd97wOpwVsHq0ugofEgofmcXCbpOymk9fiyczCcUn48Iz1W6O1l1FJ0fb7J0IePubcC51eShT85k3PAGNzehXBGkkA4JYMpeqONv2ZIL00tUVzgLJRvnmBPp6imPFa/vIlNHXpnGf3qFvEDOWeqQtHaKlDt8NpAFqNRsY/bPZdjjDCaVUOMNPE0wbqhGma4eSFjwVauLjsU9c/VdwrQcPljMRQIuxY3WaBXk2sp3OLDHIwRm6vEJxtpqU2nJabrY3q+mJVmJboXSi2+lrapplyLqhVBI8S/vwSpGHCqQK4L0/ZI1sW+y2y28O71qIYlOBlq0SsN1NayvP9ZXg9ZGcxmi+zIys7WV3Rfl52FzayYBGiFbvkyYt8NRXHhfQG5nYPKus91e5ZQSi0B/wPw7zvnXgX+o/p+A/w28LeBV4D/WCn1yjO+6/+3dl/+KuKy/Fuf9eBnOkqqE+FKKD8Y4s6n11tYhlOCN9ZkK7vWwC6seJMFHqz3AHDzHMYzyo8mmG6AigzeG1sE394WyM24ZPrDGrqT+OQPr6ShcjyGkyHFlSO/VMIniAPZoqcFOCsnZVoSNqThRFbIlN7R5BO5XBLhrQQUxwX0pbRAXqLvrGA2myK/eW+GatbTeZcyOfT4zWXG7zjsRCRqg+8r0KB8TToWDagbpVQTS3NLXt/r+TLZdjajGhYsdgvsxRzTi2h8s43yFItDy8WfOuFJ3FnBbLbkx+tqAb0mcd/K62jEl+5kKqWMUHgbpmvwVgJhOKw1wdPM9yH/QOQ/pqMFyD6Y40YL9PayNKRORqhAYYe1nGuWY0cLFh9XkBYwrRfHpyd1HNK478n3OkyFRbA3loW7hqdTlPJjOxwLxvKDAe5kBPvnXL0vACCz5KF6CeG6xu0PqKaisS73xDUiaFlxPtmSkkJ1MKYaFfLDNJiJBns8I31rSvGDY+zDI/ELNBo6CVcPDaM/GEMrxuws0fmaodwdUxxl6E6ITS3z9zJUJ8a724UkJtiRhiOzVEpBgSFes3hdma4sjkuRz6UF9Jri8j2ylB8MCO41UcuJAINOR5heRPrEku8uuHpHE646lK8EyK8V7RctxYUlz2vXl0g8+PKHI6a7iLSrq+BsKG4T3ZjGlqMaVqK7XlRSZmn4uKsFuhtLEzW3lCORwLncSfNrOMWVlmqwoPp4SjXMBP/qG8I3euICPkyhlMSnGDr5cdIKRnMWH1muvpNx9VB073aYks+EJ6ECBaUlP3N4q6Ho6jXi3uJkYS4/moq34s8YkgG7z3V7xvhPgP/NObcP4Jw7q+//JeCxc27XOZcD/zPwd37yj5VSbyil3lRK/Sl/yQJb2963gOFnvZGbEsRN3MRNfGnjGWrAK0qpP/vU7b/8jKe9B3SVUn+glPqBUuo/q++/BXz8qccd1Pf9ZPxPwH/tnPuVn/JvT23p94F/G/idz/p8zyZDGy3wVxRmTQYr3DCVLWgngWkmDIZpjndnifbrvmRFo6lI1dKC6vEAb0O2q+mHhfy7pzFf36a8cijlUDsrkhH2fOg0UIGhfG+ANtD8eizErVhcp+wkx35weu2v1no9QCUyHUXooWoYOZ2GbJMDwUIyzyh3x5IhJxE4oUiFDzoivenEsNyGJGLnziXJZiUTTFnF0t0Cb6eD7jUI4hKbQXUwwet7grcMlDTyOrV0qC5QmdVEpr02u5i1BmWqaW/k18QwO0zrZouCwZTw25sEb6yhujFR30rDarMjk2hrLbytlgyYzHPs0ZhiP6X5qkdwJ8G0xV+MUKDnKvIkS22GqOdXcblj8WiBK6QhNP1hRuOlENWJyL93gpssRCp2NoGjgZRNVpq4wspUVQ0Fn3wkjtBMUslQTy7x77awM0u5P6M6mZL0cqqDsZQ58hLdEema19NUpyneVovg9VVpABoDx1e4rKQalXi3EszLa5KpX0ywe5dUqeLP/3iV8kjQqM46SDOMZ7GVguMr2OqjmgGm6+NvRVBa/O0GrlTY86kMkaSZuDj7HnZvIGWGnZW6eSXfSfigI6+RV5AV6J0e0Td76EQQne5iKtv7RoBKAvy2I7jt01wv6uETh1kJ5FhMLOVM0dueQ2UF8xj7YCHui4eeWQ1xwxnVIMMO5vhf7eOt+VSDBd6rqzK9pxXFSY4dppSHM3QvJnohYL6v8F7qkb09IfvhgMkfT6gmFu/BurA/TkcioyvlO3RZJcS/r20TvRhSXYgH4OJPzmh8JaSxXtHoFte0wKVflUEUFQiY32uB3uldUwPt3HL2biRDOZnDZj97Z8x9zuy3zoAvnHPf+NTtn3zGU3vAG8C/B/wG8N8ppe4BPw1i8Rc+iFKqAyw5557Wjf/pTzz+aQniNrJQ/8PP+ow3MrSbuImb+NLGF6FwUEr9FvBf1P/330Uy2wvn3AyYKaX+CHhQ33/7U3+6BRz95NPx+REV/xz4Xz/rAc9WglAK/yt91HoHFRrsTOx1mKTXxokX3wXyiuIgkzriZA7TehiidKj7m+jtZZKvNrD7l0LuuhgTvdrEb0q9jDjErDel6N8KMbcSvJ6IzNk7w318IUX+srZqaYVSO448gbyfjUVONakzzMCD4YTqZFHXdydcPfLRiQ9Zjgo9sd4pKo7fblA9usA9OYWqItrx0LHGdD1pwuUOV1QyYus7ohcCqollsVdiCyiHMo5sZ5b5owLT9Qi3PaqzGdlRJaPb3QaNW5ZgJ0R3AoofHElmc5CR72dSYy1Kked9NGF+rCmOcuzRSCReiwJXWTkejQC13CB8dUmOSWAESN6veQp12OMx1eOBNNvutMW+ZyGvEW9Idu/SguxCidTMOtR2D3otLr6vcZeSbaluLHXIcUbruZLwtY7UfA8uoSO0s/wKvG3hWUQvBChfKGrzf3nJ4lGGWk7QnUBYCWcz3GCGt9OWzlVXnsO0zHXmSRJR7k7QO8sMThO2eyO89QiUovxwDHlJ6wXH0relLspkLrKryBPZVkuMXhv3Pcbfz8jeuoKDC8nuzsZCONtow6LAf7lLcWlxA+kH6G5ENcixe5fYx+cADH8g38vwjxYQeRD6uKIivJdA6GESRbZfoBNpNHtbDfwNn+QVHx3Kx/TuLqF6CTpRsltq+EJp68TChdhaqm2IRAqI1lQHMg4e3JNhlOJCRuEJDNncgzjAdBT+mk/zQSCM5NEc+m05Zx8OZSDjWIZ+VGAgClDtELMe1UYBUs81HUP8oInLatD6wRx8cYoO1rVQ505GMtgEmF7A6isL9EqCDpVQ0n7GcEDp7Oe6febzOPfbdVb6NefcEfC/I6UCTynVAH4ZeIg0zV5USj2vlAqA30QW0U8/1xUwUkp9u77r737GS38b+PCz3tuzlSAKW8/oz9G9Bma9id5awg3n4inWa9D/mz4uK/HWag6EJ1+yzYSjQJZDrw2Rh/K0WKMP56huA69vZDs7mol6YpIKU6DbEJ1sJ6bcG6NWWuhOhNlsofviuFANFoLAbAYCAZrWioI7y7LN3L3A22lSXjlYFHRfLuVvzifYk8n1Cbi6Iw7JLMpr5oVNLXq7S/DaMsUYcTSeObxm3ZzrGsJ1jTbIjHxt4W0LjZ1WmM2WNC7GmurhKcWb59fuyunDBbphKE8W+Bs+roJyZIV/MVmQnTqinkUn0vxweSXNl6sF1f4EjLga0IoEXu+LeoPQQ7+yAZ4gClUzoDgusT/aR22vyLbU05T7M8xyXfrIK9nGe0pm+UMfFgUrb1SQV+J1djjFDhfYtMK/00JphffqqsDPBzIh6bcB36ASWVSqQS5lHg+Cddm657sLXAmHfxIw/m4qDaB5jruYoPsJer3J/O1UmrepAOxJIlZuTVn95RqyXlR4m7Loei/1UM+vQuThDgSxOPtRyuzdAjuYC8thUeHFFtPSTP9EFCNuLNt9PC2LfSsmWDeotRasL6NaEcWVu57Ic6MFSb9k9MhDGytNuKzAzQuK/RnVyYJy7Lg6igTAlAuc3y4s1WWBjsVtxU0yKCuyY8iOLPPvj6mO56KYeaogiAPUc33BkRYl1cSK6gKoThbYQjF606KSgNX/oAPDmUCiVpuiKPnKDuXBRI7HNMd/fVVKJVtSwnOzHB4fixmBp7GTDP9OE72W4L2xJclSJ4TK4m0nqG6C90Ib3TC4RUn6zuz6eVTkCcBKi2v1FzU/4T7nf57pOZ17CPwfwI+B7wH/o3PubedcCfx94F8gC/L/4px756c8xd8Dfrtuwv3kyN+/VcvQ3gT+U+C/+az3clOCuImbuIkvbfxVuV045/4R8I9+yv2/C/zuv+Jvf4CULJ7GP6jv/wOg8yzv49lckTshbjBj8ebVJ9M6swxCIZwR+bW3WIZ5bVOmj7ot3MkV4a9uop/rStNs75TZ98a1P5itywkVlE5kZ0WFOxjiBrVd0DSTKa6skNftJLCxBE35JSfwBHUZeVImueWLZKbbwJ5OZM+nFWq9Q/SauDHrnvipuUlGeV5I+WKSE9wRgpnqJTDPBI1owZ1PcWmB1xAc3/zMcPRBi/O3AsphhXd3SZo2WlyUVezhNy02leOkO6GwH0KhbUWvd9H9hOiOL55iiUFvtom+1sG0lEyBnc5JvtEi2ImE03B7CbXVkwZjw5fZf0/Dehd3IWUDAo/s7TFkJSiNHaa4nDrLQeR3VSVg880lVKioLgvsqLzOvJSG8vElVBV2/1JKCGmBW1jRg4LsSF7cgO0+OIedSBPW3FkmeFnsaZ5yN1Sk0est/FVDObAUe1MWQ43X1ay/MiNoW+zF7JMJym4T1QwJlkHd24RNaTy6/QG20EJTm+fY0wmqJyUL8lI+2KfPAAAgAElEQVQaa7HsvNR6m8ZXI4IlR3mSScZpIX7OiDb5K3ENNbcUeynVBxfivr03QPcbMM9hMKbavxJWxEyacGq9Q7ATEjVFcqgCI3+bloz3fIaPfFwJjWZO9iTDdD+x6ClGkqm5rKQ8nItE0ELQU1S5pppYsFZ2Ivui4eZqKiRBrTAt2bG4cQYKoluKqFtJqenp9QeUuyMpPQQ+phdJSamoZEfTitDdGKUVeiWheDRC7YjtV3Wy+MTnT0sJQT2/ilptyzHOBNGpVxKKo5z45QbVMMPOCuwklx3AwRhnnVynP2P8FcrQvjRxkwHfxE3cxJc0fvEtiZ6NBXE4x5WW8K4I2F1RYetGjzsZkX33TIYJ5pVktNbB7onUB1e7UlOLPMq9EeGKQoUe/nZcO9Xm2KxumJxOKA/nqK1lzGsb19Qugvr3YrbA7Q9wpxPcoqR8byB1qERkZ7pVZx2LgtmPFrjdc6kZDqY1WF3Yr6YX4mYF/rbA18dvWdysIPvBuQx6XM4oj1Jhw9b8W6XB5RY/siz3ZnS3F+i4HmzwtEwndWPmP15gEtCxEgpbJ2JyIBmKv51AIIQqnfhU5ylX72hhP2ws4T3XAqMxm00ZvmgGFJdOsjxn5bNuLmEe3KL84RHu0Qlqaxn76Izi3QuCrYDitICLEXqtheloMYwMQG8vy7BFIRlN+pFjfijWSm5e4DcsztZku7rWnx/k0lxpGmFQ9CR7Ji+vvxMVeaheE6YZOEf5wUBq7KOFcHk9g9lIQMPVhwGtrxiR3YE0BJXCPj6nfPccRnPcZMHVh4Ec8DTD7g9J35xiAukZ2IsZdpST/eAcN8kpd0fMfu9QzqV5gX0ykHr0TozX9zn/rkZFYl5ZjSx6qyv0tNjD36in7YwmfbjAns/J3xnizsYoX+O/IoMyACQRdlIQbkDUtdhZwfxDkQkq5Wh0c2yh8BuWYqrRt5cwG4nYXr0QYtpezeVVKE8TbYC3GRNvODFCLa3ssN7NqXYvZff0NCO3oEKFaod4txO8nTbh86GYxD4+wx6OsDNLNbHSFB8IvL6aiT2W3b2AVoP8faHnEfmYlQCSSKhsy8JCOfl/HMUffQT750It1EpuvpFMGwju1rLKQFMcFthxicvFkGF+KLcvIm4y4E+F/8u3YLkDh+fiDPHuOd5LK2Ad+ZvnoMFlTpCB1knHXiv4yh2YpbC6gpqlePGohuoodDMUYExWCCqwGaK2V9BVJQ25JBLNJ+D/za/hLwrcyRVqOaH481P8V+NrZ1fiEL1hKN+7wFvvYPcvMaEjfzwjeOodt6hB46lFd8BstQUu8/BUuujTHJM46V7f6uDsBNUMsKczQRq2NHZmiV/06w57LI2XoxR1usC71UC1Qpq/1sFeCKQHZNsZNUsWTxzR84C3uJ5kys8sVemLqmGeoZYaojENpSxCaWWMepiiSiufNc1hZQmzJq/H+Ri93sJOr6SJdVlgT6foN57H8z0IPOzDYznu9Xdj9y+Jn6t97a7ApOX1Aht1qutjFb5UK1ISHzfKZZu/2cd+5330y1JqUvNcGq4gAJkadO8sqKc/XqFH9Ou3CA+GMql1llEMASqMp8n2C4532+x4lwB071lp2rabUp4J51SnC3HEnlfobkg1SmV73g9xh6kA6oHqqsQs1YtdJ2Lt7/hyvOei46aSxqId5nivb0h5a5IR3xUcKSCgGovoyK1gVN3FBBUZmFcEtz2K44L4eUN5krH0VQ9XGigd832F37S48ynVYIHyEFeOzBG8dkuuAevwboseuJzMBdQ+WjD4kSFuWTEQOBaQVfFkQnHlCLd9qsMZulUjMjshdijwp2CnQfB6h2r/ShrSV1PUSguvn1KeF/ihgYsRwWu1frfW2btHxwJr8gWd2Xv+CpSS6cBWCLMaS1oO0RttgVIBxXlF+FLM7EeOTs/iL3nob90lGb737CvRTwkHVFRfyHN9WeOmBHETN3ETX9L4+c5uP0882z7BM3A5AkRXOnmkZKs+zzEtg9etM6DSSgmiFYuXV1EKQOdH72PfO8GeTcVP7GyMPRxBM8Kez6RBNZzD1RR3OMR+fCUebx0P/8UOLgwhDqg+nuJOxuhEM/l9yZbU5pI05+Y5Zk00ihQV4ZbBW/Ml+81L8ndEouT1fPK9BeXeCLt7IVuoRSmTYbu15dCybDsFBOQo98fM98G/38GsNxm9I9Iwsx6hfMnC8sfSiKCboJKAyfeFuUBe0bjviXtxU2DmKgmwC4uJwVp1DbO2TwaUH8/I3rrinf97ielbhWRHEwH4uNKKJvN7j6VZ2G0KVjMOxIU38ATcfntJJhGzAkJfyhJpBldT2V3UEjvTjwhfEMmdir1PNLSBR/aWjLK7ieAddSTWNnx0LD5pj07gdCQNvNCnOpuhWiH6hT6udLLD0Ari4FrCWOyL3Gp+qHAOvHWRiGUjw60Xx5z+mS/6bY0AmT4+E0fizQ7e7QT91dt4r/RFGpeBbskkWvIrbTmPAoNuaFQSkD5c1OejJX9c8wm0wh6PKR6P5AqYpNiPr3BpKTZSTYPpyfGavVdhH59dN4qzH49RRqEahvzjAhWASy2q1r2ajQQVaRZTmYycfD/FpfV3/P+x9ybBlWZnet5zzvnHO+ICuAASiURmZWVl1sAim6xuqtXsgVLL4ZYjZIV3lr1yhO2NHN45wt55ba+1bSvCC3vjhTocUsgtd0tqsQc2ySJZQ1aOhUQiMV1cXNzxH885Xnw3UewOiq5sstVFBr4KRCWQSODiDj++833v+7yZp56yfB6OcONcYoweXxBuGEwqsKLeGxXJNaj3J5R7BW5csDhQLIbh5fdWnRg3s7hxcXn8JwngxgbmrU3QisUfnMA8xw4rWcAGmuqjcxkbaQVVjbq1jhss5OPdBu75BfGv9NEtQ32ydBtO5SQ7/l6N3RuhuzEqUAQ9jT1aXILmX572TC/A9H763u5qCXdVV3VVV/U3WO6vTYj2xahXugC7jw4vf8upJKD79SUOcV5SDSw6gfCN7mczSsAPpqjTC+FBxBH+6bl0LO0EvyhlKbRc9ojUqUbNCtRai/rJRGRSU4de96jzEXgv6bNpiOkmNNO5ePqzUmbOt7dQeUn2fz8lXNXoNXHJ+dECN8oJthLsyQLdjYjfbVA9GONL6TAausBXns57iSztAkP4hsj63JmAptOd5UwqCdHGAdL1m7UYN62wGUTdGPfoFDuqSTYloBLnQSni3RA/K0V0DwS3OrjTOStFDm653HTSLYU7Mbcn5yjjCTcD4RRkNearOxBH1N8+JrqBpBffWsfviUPQnUxlKdbvwMlYZuTrPeKvrMLGqkDTByPU9JjoK338vEBtdOTnmheQV7jBAjXOMU19aXRRkSwZ/bzED+a4cY2b1KggQ7cMWgvXg8Nz6bSdlznrtEIj8003LtFtQz2oaN01uLkS8H09Il0vCXqazWs1duQI73VFPtWMcY8HgohcVKhmKvFUaw0ShNJHGsH5HN2VTp6gQG2v0HD+8mQT7iS4cYkdnRN+qQ9FLTNO59HX2oKuXFSodkz58YzZacjKPUu1nxH3JhAYkm9sCPT9ROa60VcFKO/2ztFvXYOxGHnaWznhumb8iSYua3SqCK43xXE5mlOcKcyaxXRD6WZHOa6yeKvQ3UAivCKDuW5QoaEdLvCVk9s5L8FoguvL5O6wRkU19mhOsDKE/gr2+AXJbUniDm815fmxP0Q35TRELdJMmgm6l2DuNEW+lwRwbxf1Ykx4uyUmlt1VeDGnseFQzVQSmhcW0xcTzMqvLU83kcF/sC/36c+kPF5dXYCv6qqu6qr+g9fLEcQvcr0yC8IOc1S/hV9U6G4iv7mHBdGthHC3iWpElPdHLP74XCLGG9HlDI6z5fx4Xgnp7NqKCMZvbl2yJADploHw3XWxuQZCHANE+gRiAGgK9FwlgTBjt9cgTeD0Ah2Cbgfi8//SNn5SSER4N2HyQFMfy20rL6A4EeauL4Thqrtie2U8x53OpeNwnupIYuyz74wp3z8lWXNiz64dblTiM0e0LR3x7GMLCkzbSKgmoDdb1IMKN61Z7IMrHO5sgWqGJPcaEtgJ6GZIuC1x9I0vxaR3RK7la8/kPnKfOScR4ADNGGYSe1N/Kt0vQPWtfdw4F8//bPlz3N+HvSNhYOyu4heianBPz/CHF3I/tmLMTod6WIk5wGjMTpt6UJN/d4QbF+RPCqZ7RtjChUO/sy2qiV5HHr/SotdS3FRMBG7/nE//RUj+zKETjW5q9FaL7EgsyESGoKvJnnnMvQ3p2iKDen1TIqic2IFVK5J9wpLudtm5jrPLkFB7OMXsrmAfCOZVRUb4JUmAbhiUVriTKfP7Nb60SyuyEb51EqB6DfKRpnv7Jd85IP/zc4o/H+BOpui7m9J9aoQ7/XjI9EOLPxzhl12+r0URFCZO7OuVx40LOVEFmmRLoprseSVql1lNuKqYX0Rkj+qlVVuszm4gVnI3qXGHEygt2ffG+NoJf/msxOy00YkWgt3BgOJsaZ92Xh73eX556vKHF0x/UFH98FRs/3EA7RT34GR5aqk+o+DUDn82xdxo8elHK2LWKGqC2yvL00iE6jVEweP8MnhA3n768ljqz/X281qvBmRvRYJY7DSXesUavdHE7LbFseYECG3aBhUgLz7nBBxT1rjnF5KkkQZwrS/HY+vgYiJff62J7qWyTBvNJVHh1jVM20hWVVmB1p8lcews0z52+qLTdR5OZckQ/+Z14UQEAl63o4rFg5p6f4IJHbppsE/PSW8aWr/VI+5agi1ZLtR7E8hL/PGYxSMrcJKtFtGNkHpQC5KzbTBthZsJtPv0hxGzfY2biqY53fG4zFOdWIpTRf5EEILFmci+wpYgCc2tHn4qul7VTXCnc+qj5YJsVAr4xsmiCa1o7TrhIzhJcGZRygWq3UC/tU34SxuyOJmVmF4oUJftNTg4k/uqFYtjajSHdgPViPCjDCqLWmtSn2S4ZyPoNoi+0if+6qoc1a+tABD0jdwX65rGeo1uBpIn9vYb8LW7cDaC0Uz0p4NMfgE2AyhqtnYnNL6cUA8ruRBGhqDhqPcncpFsGFq/1saP5pfyO9YEMqRWEkhjWaY2E0nEGMkFhTSUnyEK4PYWbmFlIbnRXDILFrJoG+UC5Ad8VpO+ptH9Fm5cCeBoXlK/mENe0bzlMVsJ5ZlDr6VEtxNUBPYkByXPwfjdHnaYM/qBovMbLVQ7Qa01mT5SqACme8uMvwTKM0k14dYG3Nwi2E759MkaZj3CZxXh7S66YWiulAQtTz7SjL9dsvikoDoUVKTZTLCTWhoYwI1LBg/Ty8ZHb4hc0A0XNL/euYQz5U8r/GBGcEceQ19aikXAYl9BLU46/3QgbrhhDuOZyO8aEfrdHbmgGs3r743xtUe/3pfvP68lKWRR4sc5Kg7kYn9tRd5+yvKAU+5zvf281tUI4qqu6qq+sPWLvoR7NRraMi6Iqkbf6cuxPgkEMN1L8M6TfyS/PeeDUI56hxPUSkOOOFPpNFQjhOfHywTjTPgQWuGeSQIrraWvfX8Ae0eoSDo65hKj4rMK/+AQpnORZJUV+k4fTi9wnxxLh9Rr46cF9uk51Z++kKXWimexB91vdgjuruHLpcsoCQnaSHxNw+AKh310Rvl4TvqaFjRmIGSzei7HyfLYLjkPMP3YsfZ6TtS0nD2SLsX0E0zPEG4aDg5WmA7Ea18sAnFBGaGe+VlBPbT44/Hlos+sRbJYChT1s5m4rPotFg9qdNugek3JeRsKBtQ+OJXxzsVMEI3Plku4RijLyZfd7zIx2n10CEmIf3JM/clQHttaloQq0hLDgxyvaadylD0ZY17KDKMAs5miguWpqB1DmkJRCnkrr+XksWRTCGYxIb0ny5lgO0VvS4J2tKGpTq0sX0NBjGbfG+MmNfZ0LnLGlxlz1sK68DbcgRz3fV4vOR7LY+jggsGHsUDW+x10v0nQlogjtCJ6s4250SJ7UmH66WfP7XlJ+ckUX3h8URPeWUFpRfKaSNz0nQ2id9cwvRDOp9iDydL9qCgLg5+VIqk0mrBhiXZCZrOYsOWIboRMz2MxiRwO4ewCOypoRJL552clPqtwC0s2CRkfxJR5gAk92oBuKtxATCjBjRYqNjR+bQ2daNZuiUmp+vhMIp3asRiEjMYNMygtda7xeS2jjGZI+TSje7OULndeoK93Jc+utARv9/EHQ/RaKjLMwGAPp9jjGcFOA902+POZGDmagbxOl4EGJKEwSMYLefupy+M+538/r3XVAV/VVV3VF7KEavnze3H9PPVKF+CX8OXqT55jNpNL8b2KQ/zFAmYl5UST/uY6vegc0pBqb0H85YYsxrqRhP/1xA7ri/pynqt2VlFtCXQEoJ2KFz6YC/2qlhiZ+ukFvvDopsYUJeqtG7jvPBFL7HAptJ/nMt/sJtj9hSx8jHSu3gG3t+B0hOknAo9//xSdauqRLJyC9QiXWVSABHYCbpRh546gKSwH/3iOzcBmnqjjCTdDoKI+Mri5x3RqTC9GNSPu/PIFdubR7QDvFLphCACVGKpHY3SqyO/PiXdD1NYy0TkR08bhBw22fyVHDec035YlE1mBPZDImTAUmZg7mYnxIjCEd1ZgZx3OxtQfn1GdO9Kvr6Je26D8/ScCl98b4UuHudYQy2ka4A7H+NJJ5NBGD06la/TWUT+dCk/5+QxqKzbYVHgebK9CVcJHe9JFNyLc8wvqkcNMcuyJ3HfhZojPavRWC3c0RTVCdMNgUuFMVwcFUBBtaEw/RW20JUF5vYv9eIB/PiO4saB+PiO42Sa7nxEtppi1BJcJRN1NS1a2Kqp9iNpjWUAuS6+l0hUCcV/YBn6aQ6DwpSX+u7tUf/ZcZIulcHnLg5I4yWBaoOIA3W/ih3MWTxzpjRKXOZyLGP1ZSXOzJiwGRMvx543fgfIpEAd0+hluqGTeH2nKQ8u116egQ+y4wp+VLF5o0k4F2hN2oJ7C8bM2W0wJe3IqsScLqL3cjtrLTiYOcHMP3Qb2By/wC4sdlZhOgNrq0rgls3gCTf1wRHRd9goNXcFmF5TGbIv0k+0+6sE+9d6YwGg5KZWO6tSS7nTQi4ry4wm6qajOPY2bGpSieDgn7OforZbcp/z4fJ9XK4+l+qm/yhe5XmkEUe3NscOc8szjxiXlXi4pDd7LmMF7On9L8tTMO9dQWknyrFZyJNUKO6okZVYr1Hob1W/JkmSW44dzcVo1Urh7C5D0YdyS/D+ao7sRKtWyyJsvH+gkgPMpvrJynHp6DnsnAnOPFeWxw3RDxk8j4q6TF3UsEd9uWoKCyQNN2DfYYUn2qMSOrbibrKM+mFLtZ0z2I8LtCLXRJtgMSd5ICFcUs9MI78B0Db118fRPP7SyHHKe8J1VWeLEhnSllNSEtQQVaoKthHriJU+um0iSx1hun+43WN2a43OHm5botQb5pyWUFrPTJn5nheyBKCxUVxIi/IGAxpnLhcYVjvi1GHY3oJmKuiCWxYzZbMji89aqLLxKS3CrI8qO43PUdm/JUkgJ312XBRhQP73g9LuhJOS2Yjg8Rz36VPTGm10B7HdiTEOAMy8z5HzpBGrj/eUvYdWMCLYTyZur5RnpsqVGNQpw7+/JUs/9iAvPgr/IibY0wdt91PYK2ckSpRkbkjdiwu2I+ukEX9SSKHJUY5/PuPi+RxlFeFfQnj6vl8CbJoSitbbDnNEfy0IuOxNHWPHxTMBTzViWpEB55igvIIpqOnes5ANOLfUUyoNKHue2lsRgYPZ+QXFgwWiireXtNYrixFOPochCkpsapSDoGcIerK7NcRYwENxqYceW/IWn+vic6qjGdDXubMH0UOD59Ukli7JEo9ox/nz22ahlrS15jjMri9R3esK3eHQsqoeXCggg2GnLEr2sCe6sEq7Lfa83W9gM3NxTL/QS45lLEIFWSyhWLNmMP2V5/NUS7qqu6qqu6m+q3BWM57NaHGka2lItNFHmcBWM/qQiTAa0v9ECowVR98FzSUwd55Ip9WyAv8hRjRA796hBjkmnl50wGoo/PSb+yiqz7yxo/6dAVeFyh9YCkvbljPn3ZiTXNS5z2JEl6mSodCKdWhigQlkEuNkZ7uEFeFCp5vHDFd5sD+neAhUo3MMT6ZoDLYSzbU14dAFadK3pHQG7qySgejwlvN0CtyBpVeADqCz1SUX0ZkJw3TB7X6Pfz+ncsjReN0w/8UzGKc3DGaYXy/dZa1DvjWncjdDdhOrBBWY9wk2F82CWTcr8O1PS1wL8mcQqvQSHV0+n6DWHzZQcC48m+EDjqiU4u5tiH51x8K2E/s1zdHhO8o1NTDeQpdpwgh/NmX9/TjENgIDo+Qw9zDDvbEmW3HYXshLdjbH3TySxeqcjFLbNttDmEo3uJWx+06ESQROiFZyNRfM9nosMbpxLIvDemHroOH7S4uY/EL5B/sIRryuCXXGG2YXHNBXRnVSYDXvS1buHpxKrNJyiW9K1p30lJ6BeStBfHuCKirAlXVB1kGPamvLUYWKwz2c0rjvpoLshjfUCEGypYoTSCnOtKe/XI4pDRzk3xO0anMLWWqKXDqYi6Usi8BCteGwmo4zkmsDsow2Pyxz5OODiIuXmzgJfeFxWY2KYDCJWWvKaQCtsIXB/E4NpwvxZxMZmRBLlTD+0NK5D6zYC3w819fOZjKsOAhpvKtDSrc4/sSil5YTYVASbKfVhhstFVz59Ymi/tiCo3TJBPJal+MEIlrplN10QbOSo8xnV4wnR338TPZriHg9Eax8oqocXBNdS4h1D9szSfNPw/N/GrF+bE617TDfADnKqfTkhJP/FX+2i9Fn5X/gZ8M8G2nlVV3VVV/UzLg84bz/X26uUUup/WOa2fV8p9aFSyiqlVpd/9ztKqQdKqcdKqf/xc3ytW0qpD5d//qZSarz8uj9USv0rpdTGT/r3rxZJFEB+oknWHMWZR4cQJk7ITydz6uNCQgGXfnw3LtFrKfaFLMdUr4HNwOy2qR6MxTteO8LuclYcBbJoUgrORqhAURzU+MMLVKBpvBljrrckiLFG5FFpLDEupxf48xmMZti5x1celHS8d+4ORb4VKjCg+y1wnvxBjtrsQhLS/o0uut9At0NUL8XNpasJ+kItU4nBxB57UYsbKlHUzyUuZvvOFBM4ilNh5qabjvXrc3zmyO4vsMcz7P5YTBrL7sesi7GiOrWYVMI8/aIiSPgMKL8MyvSLivJM4RcV7b8jC0l9YwXVCGm+FaBS6Ziro5qdb+TkowAdASstzE5HghWjAKUVjbcTOncsq38nxRfi7qN26LubwhDYH+NLi7m1IhDvo6nMa/MKwgBfenztLpkg7sUYN8pgo4c7uIAowJ4sqAeVdHdrCc7C9jvCNzbXmoQtuHgcUj+fy2NpIbgt7kfVTUi+uiILr80WtBvQa0nU1Kqi/GSKijSq15QopsGM+pMh1UyLM2v5jA5XFeGtFN0y2Lmc3tBK5JFLp2D+/ojyqTBCik9m+KLGVQoTSoCo2UxYuWfxRY3PnVDlhlPszBNeC0lej4T/3BDJnW4Zgn6ICR03fmlGPajQ3QA7h8lRxNqdHJsp3MIyua+I+0KfM00I1iNu/lYubr/YMJskhLdal6+96lDg57oVsPbrEoBbX3h86WneUXTuWOoDuW8AgtttzFpMcLNN77ebBLc71E8nVI8n8vM7R/EoE9deKRD+2Uc1829PCO90ZE8yzy/NHHZq5THSCrPdlpNZqFm/NifehPBWC/PVHYLNFNM1mO5ny8+/evnPyUJ7tS7Ze/+/vkxJBv4n4N9478+VUgb4J8DfB94G/pFS6u1XvNF/tPzaX0ZSlv/xT/rkVxpBvEx/jr/UIS4tdpgT9DzhWYlea1Eez1CTGr0biQ5TK8hrcbEBrLaJt4dgHdUYpn+YsfpbCcFuE19a6g8HhG+v45MYdXyO6adELsPnNfrWKu7JEG20qCcCxPk0XeAenaK3BB3ps4p6ptCxiFiixBNtadzcSrz8WoIbzFBaEe9ICrPdv8BsNKkeTzj+uMn6jQnjo4TVW3LkWzw6B6eIr4lOFudxpef4kxbd3oLWXUN9bEhXrVicA9ChpBcMnjVZr+UX0MnzNhs7UxrRFNWMqPbnzAch7R1ZnBQPcnSMZIAB1YlldhpSlbDxKwZ7LLbi+vmc8EvrUFrK5xXHT2J66wvaX03whSVdqwi2YtEFayUXy9UO9qMjzFub6L7oal8mM7v9c7lYvr62XKRU+CBncl/R2ikJGzn1wYJQK9FAP85AZ2Rnhtau4Dg5HKLXxSFpejFKF5KKvNWBj0QT6obyfxXB2teRcQ4Qrmv8+UK0yG55gW9F+LMC7TzVwwtUpAhutkR73gjl4lBUAjMfzWlclwWvWRPQjLnWFIVGoAkXI4oLsMMCrQ3l3gJXQDXTuBrCeUXQhvrZjLDjCTclu654lFFcaJLRjGqmMGsz7LBkdhrSsBXlRBN1SuLdkP0/a3Lr7xayfAxnmF7A9H1PO6rRITS6FSpQJDfFLZn0aqoLqM4LwlUlqobKkn37nPi1mI17C+yRJn/hiVYrdKyYPfR0f0uUHPXeAtOEeuKJGqLwmT5RnJy12dqY0H4jlzj72qGvdwW4FClGH4b0b3mY5oQbhhd/lrL15gxfQuO2ws1BGY0/OKf+dCpjvCRAGVmwVfsZ4VaEm9fYsSN9PUBvNMUJVzvBo46KV7ms/KQrzn8IFcQ/Av6P5Z+/Djz23j8FUEr9n8A/BD7+0X+glHoP+F1gAfy7H/dFlVIKaAOPf9I3vxpBXNVVXdUXsl7qgH/WHfDLUko1gN8B/q/lh64Dz3/kUw6WH/vL9b8B/733/m//mL/7DaXU94F94O8hF+p/b73SBdiEnuYtUCsSg6NTgy+8JA07iUsxazEsSsncirSko+a1JLmejdHtkPLhFB1B72uSoKya4jYK76yI+4v+OrsAACAASURBVCqM4O4uqt+SRN21BgQGX4okbPaxaHTdsxFUNcV+JS6frRVUGtL8WoPktYh4J8COLPOn4HKoh8t05GVX6B1UH57hZjX2cEqwHrF5d442MFvEVBOFSjTNr7VIbkj3B5JOHPRD1q/PaN0G04tprklnNNkTuRta8rFWN+aCLVxTNBslrpLonZcJxFp7qgvF+Q8NQRvp7IF832Jaiu7tGq29yPGA+vkcFSmqj8/IPpAly+5vV8IiaAgQJdqR/9vHQ5H/vXEdzif4csnkGC6oH1+Q/Moq8Zc7uGmFub0q8rBAUZ3UVHtzuu+CbhvcuKAae+yepGHbZYMTNa0sxCKDnxfyM93bRX3tNurlGGVeEG0bVCw8jOq4pDxTolledqi+9uidlc9+xrNSRhAbLdjsStxQ4cFICrRqRtjDKdkfD/BFTXy3gUo1diqPT3Uizjo3yiAw6GZA+22N6UV03nC4AqLtgNa7IZ33IsKuAHd0y4BbugKB4X6DxjWH96ACjy8s1YWn965j9CIlSCX3zeeWTifn4nsiIWz+Uop30HnDYWeeYM0QrniC1ztQe/LBUk8/0ziL3O5IeAvhuiZ/VFzq0uNrMka7eBQyGLRR7QSAcCsiWAuJtgPKY4udy/d742+PidtWcu+6Mfa8ktdUaQl2O3RvCmPDDReoSLN5Z4YrILweyoJ6AdXTMeQ14Zf7zPY15eMZZi0iv5+hX2bS3elJxuBGEzIZN/m9s0sN8M+mPN7bz/UGrCulvvMjb//t5/gG/wD4lvf+fPn+j5Mu/wUcm1KqC6x47//N8kP/+1/6/JcjiBvIhfp/+Uk34EqGdlVXdVVf2HoFm/GZ9/6Xf9xfKKX+MfDfLN/9T7z3h8s//+d8Nn4A6Xhv/Mj7O8Ahf7EUf+mi/BPq9/isu/6x9UodcPpGIGGAnw6xh1PK/ZJgpwHL0EW91RZc4PaqEM2WAZIvfh+qA3GxvVyAhKsiJQJgq0d5X4I6Vb8FUYjf2hT0pPO4aYEfzqhOLfXQEiSOcLchHd+sQCmk+xrNKD8QJ52b1vjSEWyGlFmAaUqESn20kBmrViK6n0rXp9oR9VmJK8C0FL2VBdGqpx5aVDcV6tetFvVwGVa5RGcGd1Yo9xZEW4IvjBs1RREw3EupTmqyaQgOTj9OaK4WhC1HOdFUj6dMnwU0t2qcVQShzMXDzZD5Y4+rFK70qFiRNCuqw1KMDpEA6rMDxfgkEYNHN0Ft91h8d0q+V6PXG3i77Jq7DUgTij/YJ9hpU33vBN2OCW51II2EYFU6WAjL4OzbQvKK3l2lHtSUh5ZivyZ5TeKTlFYku4Z6Aem9mOBWl2pvBnlN/fhCknitRbdj6oOM2R9dyPy9YYh3A8KtCGW8fK8/EkddcK2BO5Hnk5tWKKOoD6ZLzkct1LTXO2IWqB2Lb0tMUtBWy2VvgW6HhNdD5p/Ic7F8nF3S+ey5yB19YTE3WvLcSwNhkwDBtiyNdS8h3DT43GHnkKQyS3clVAvD4pElaCIMES1L6CBxTB5oVu5ZVr4eiLvO+cvk43BTDEjR7Sbq1gbB3R6T8wSTQpkF4BQ6AT8vqJfuRu/E5HH8pMXh+03Koed01GK1N78kwOl+k3pQUQ9rpsfL00agIFBE20YMINlyt/B4QX2wWEZxBRIJFWh0N0ZH8lp0s5piX3YrurXsy9opcdeSnYhcE2Dy1AgBbVoI9XBpzikPa+ww5+hfVEwfwPTBq1xZ/n3l8djP9fYTv4r3/+Tl0u3lxXfZyf4W8M9+5FP/HHhDKfWaUipCLtC/95e+1gUwVkr9+vJD/+VP+Na/Djz5SbftqgO+qqu6qi9kecD7vzYd8H8G/D/e+/nl9/O+Vkr9d8C/RKJuftd7/9GP+bf/FfC7SqnF8nN/tF7OgBUwBv7rn3QjXo0FERvhADuJc68HQ/Ghn87FYtpOsXvHBKGRmVYqXvTr/9EF1QEylxxkAlifOgKjscMKczYhfmcFGhEMcsgL1PMXwrHdH6MCjVprkdyrcNNSbMhL+LYvapJvbOBnBWysECWhcIc3U3wurN71b0YUD0ui6wFqtcH8Tya07krXEPSNWFxrh8tBGdDdgJW3LKBBO+zeSAJDpwWuAnc8pR7WQvdKQs6fJvR2ckxboUNP/40F+UCjEwhCRzE1aO1pvBFKF2lEkZGuSlRNrB1JIIzi6qgi2QQVKoLNBsXjBc1bmnoMfgF27tAxtN426Ec56BR7MMV4T7ylhWtwvpDHI43g6AKaKdHbHWhEolhoxaAUbl86STv2BFNRF/R/26O2+tBrU/zBBdk0pLlaLFUJJcOHCb3dXChd17v4wYzoK33oNtD1CZxNJIw1MoR32qhns0uGyMuuU0dgz0qqKsHXFtVvUf/wlKAvcq7gZov5t2e03lralbdXLmPhaYQkt7zMh2uBk9upJWgL+H80aLDZnnH+PGHzWo3PliyQZogd5NijOS5zlKc5JoV6lAuoP1HodkQ1sPJ+DL2veLCOZNcQ5Y7iCHRTs9iD1ooMwk0Mia7FLNGKROZlxFwUXJdTIMtTH2djaCZc+/oFxQHETWESC8tjTnUqnVy0rsiPIAgcda15vrfC1vqE1vUatdHB7Ia4jw6J3lrBjXPWNy1qtYs/X4hZ5T0hzfmJzJJVIOyU82+VpCs10bZBdxN8ZTFrIX5hKU48cV/2HMKollmuSaG9K7FjYU8RbUt378aFENMqi5tX6BiCt/tc68/IPvoZzYG9x/q/HhWE9/6fAv/0x3z8nwP//P/n334X+MqPfOh/Xn78XwPdV7kdr6aCcB68p3oh2W3hZkj1aIzZ7QqIOSvJn1vmfzwi/7OBHPGKClJBU5rrLYK3+0Rv9agmImsJtmL8OIO1FowXgrycTKCq5AnrPGqrKwDxJKAe1pdHR7UmyRxohbreg8kctlZFaxtoAUY3Q9Rak/jdrlz8Ak1yXVIYVKjJn1vqgSzF4i91KMfL9IVIo5sBpicJvW5aUQ8q8lFAPawIt2PJ8to7Z+1uLsfIyhNtKMLdlLgrwPW4XRM2rHzORgudaKKdCN0yJG9IAm89BV/ymYZ1pyHLzEDjCnmUvIVgTRN0FDqSXUHzl8SJFtzq4i9y0aHeWUWtLvkblcUez0SqdzKn/uQMn9XUHw6gqNCbLdy8phjLOMjn9Wd5XvOMi2FKa61AhwjgvRuQNOQCGm0bWMhYhEBDM0Vf71I/PpckkaOpAHbOl4vL7hKTqCH5pS6mF7B2K8NsNkT37WTZFOx2BCDz5QR/PBGp4Us9qlbULxaCsmxEqEjjpiXBugCdzO1VTmZN+SXpNItHVtyUa+IMM9ca6K6wJ+LdALU8udsM7MLjS4u3cjGOrofotVSSLkqPWYtJ3wiYP4XZhQQTJK/HxHcb6BDqk4rs/QnTZ4HoZZ0sTBcfLLDPZ2TvTyTxev8c3Y3JLuSiXI8sdlyT71thQHQUZi0ibHnC0JKmFe20wNYa09aCWk1j0e5mlSxAHWAFJBWui/TMDRYsPqkkLMF50Iq4aSlnIsUsH1zgZyV6vUE9cqSvh7jMk51oSTofZfjDEeFmiLnVE+leLq+L8v6IYk/GcG4wE61/hjy+2yvE24Z4+4urA/4i1dUI4qqu6qq+kPXXPIL4QtQrdcB2WOJyR3YuGEWUQqdyfPelpX44JF5XNL7exVuWqamFkMwMuJO5HMmcB+3lN3gS4EY5zAvK757ijmdwPICTEdX7pyyeuGWcTEb9bEq4HeMzJx13UV3SwBhMYGtVsJXOC1GsFmqbey5OOl867PFMiFDjHLwnbIFZE7ebX1TkM/mdlH26BH2Pa9zUoiLN2aOE02FLHEqBZv7tiSytQlnoBVsxpiPxO/loKcEDml9uEN2T46Jeb+IrJ9Ey0wqVBjTea4vHf98SvdmW3L0kYPZtIWH52hO/nkgn1gspzsSRp9qxuPrW2qhGSHVSScZbI5JlZLeBCjXu4Ql6LUX3YsqDinq0JK6lsXSgIN3qXiYStfePKP7gGZ1uRrylidY19amMfjrvKKJ3V6Vb19KJ28dDAY2HgUgTl/etWm+RfrWNihTl0zneCgBfNQTirhPkuXOxEArazEocVCRZfvVJBlvrAmFfPm90y2APpJvUGy3qQX15Owg0b742IFxVbL6TETQ91Zks7oLdpiA0N9tiLGiG6FjjLcyH0tHmj3JcCeVQsKX2xZzGjhdA/6DAzWraX41Y2c6J1uUoXp9kuAp0Kl18ulLLstBDMfAEbTh8P6VeaPIPpiw+zPC1I4gcSnl0JK+hxlcl8iq42UL3m4Sbhv6vQZxWrN7I6Fwr0c0A/+kp7rt76J4sX+U2VCy+K5mJpisOw9lDT9CQPLhgPcItLNG6p3UHfF4TvbMqeMonE+pc3JfVDKrC8PhfpmA05ceTy5w9FYiBxg4rgrWQsKdknNgQ56pJwR1NLjt83f1ZJCP/bJZwX+S66oCv6qqu6gtbv+gd8CtdgOf7ChM6ul8JkAWVws2X3agWm66KZDabvGMpH85E0vTWCnp3FazAz6uxJ71psJMaXTpZ7sHlQoCDM0gj7MLTfDvCjQQovfi4pNHXhPdWKD8aEbVjsRMbA70YBuPL26G32lTvn6MTRTWyBNbjC4+53YTIoNKY7A8Pid+QRQ9ZhRsWdG86VBDSeDuhPsnIjyGfRazfTdj81Yz10QU+T3CjkmgVgkZJdQbRlsaOSgDhIHccLpO5rl/K8VSgccMFbmoJbrWgtJecXzToSHgZwe4anIyJN7wwcp2kOVcHOcFKQtCQ7/Py5KFCg5uWVBcK/8GY6HZK9NUNGM1w0xqViFlCdxOSewo7kJmqOxihb/dp3pZFcPzuCr6yuCxDR4rWXYXpSVDj9HnA2nsdmUHWlmpwjtlV0gGVVhjA7VS4vtMSs7siUPxpTr5vl/fLMhDzaE5wd41graDam8spCjh/ENFcKWkenMsy7loD5gs4n0poZlGjYiPLv+EcrxRBPyB/IhB39ISk71CRxk4d80GId4rGVgcaMVqfy75BK9yoEIbDuaO1Kfenq0AvR5dmRxZZ1I7y/gWE8riiFcm9FDcqmH5Qc3Tc5drWBB3LcjQ7M0RHc1SoaLwZ4StH73xB65cbzL6zYHDc4npnitIaHXpUajBr0snqjZbIBne3UCcfoG+t0dHnch/XywuRVnKS2GjDNMdnlYDxC1kY6vUmqufx+1PqhSIYl+j20iBVO+rTEt0V1jLOU43ktajSEBND9/WadnaBzwzFSGMGOQxyyjNFMTd07zrsRU1xCso4kmCOSgOR2fUa1I/PCXY7f6WL0V8uj8f6n9/E489TrzSC6LwXkd5QlyAWXwmQ2k1LwTrudsn3atzjAX5WEl4Lid7pyXE/kujr8N4K6VsNAZ5rRXFgUT1Jv7DDivkezP/duYDJLRLbvXR4JTfkhePGOeG9Lm6cy5NXK8gKym8f4g8vsKMauz+mXgpMxs9lSRP0Qwi0HGHPptgC6sMMe1rgxgXmWoNwU2A8KjJgIZ+FtLckut1stainkkbrCkd5pijHWhZSWhICPv7WGrOn4ArF7EWALZZLkRMB99jzSsYxo5xyb4Ed1Sz++Jzy1JMNA/wkh/We8A9CJUkbGrzzmKYsCE1bYVYCFt++QLVimOboXsLZUXN5VJQFWf5nQ1QioxfVa0gSSCKOr8m35rhRAWEgWt77I/L3L/AL4SnYmb/cyru5oywC/GghmW8HIx5/uHZ5gdJrDdHsOicgodrJhW6eM/vWGKXBVYp6Dm4sLyg/ziAOCPrhUt+sWftKjTIe+3zG/u95gaU/PcbujaAnL3DVTT5zW1mH2e0S7wS4uWPy54Us1KYOFUCzX7H6nsc+GeL3l9l34xyMFuhSbFAKkreaVDPwtSLoIiqC2uGnhaSBzKE48cz3oNrLlsoLx2Ia0U4LTo471AsYH8RUhaE6tQKVt476pMLEwrfwXrF1c8rR/RY6FOxlPbCUT+e4cYE7neEPL8AYec4XFfr6Z+MXEOyo2Wrh9kfi9IsDSacB7EW9HM3VvDheQccenznqwxyf1dhxRXaiscNC7v9KQFD5vsVXsgQM73Ql0/GxJ2o7cX9qaLwdkS0iXOYoTiB9TTMZyFjMZ8tggzRCt0OqBxdUDy5+muvSZXnvPtfbz2tdjSCu6qqu6otZ3r+0Gf/C1qtdgGvJC/O5dDEq1CQ3DH5hcYO5SF12DHZUXeIF6TVxe2PU8QVqpSG/aTfbBLUTZ1LTUT0Yo0KFbhtaXSAIcSdTlAEUuMFC3EdrMdVehukvcYSZxL7QbcLGGuE4ww0WzJ8rGtesePx3m6ymmegeS0v1cCROH+dp3I3EzdeXcYovrSQaqxw3rzG9gDCuURH4UUZ9lPHsUY873QvBcvYdvvbUA4huJySJJn1cMzxr0estaN+s8SXYUYWKlGTOObAFlJ86WjcV9cSjYxjupVz7TRnnqNNzFgeK+SSi/6WC4FaH6sFYpD7BArMWo691SNSFdEaNCMYZu98ssCNFvT9HpVp0yWupOBXbKXTa+FlJdVKR9EG3Y6hqgZGH4rD02fKxjSDailBrTabfKgkjixtW+NKjEs29v5XhpynT7+Z0vtlF765i75+gmyH1yBK2LWq8QBmPacJiFNBIpftXgUJtdfH75wI5H8zIH2QkbySowFKPHZ2VDAiEwHVUozdmUEiHN/tYXpTNOwo1zkHLAjS4KIh2QupBhZ0L5Fx348+wnqWlOqqI0kr0yN2I5L0eNGPS10t0M1xK8RJYSHQUWUW8bdC9BDfKUWlM8alI87rXck73WmzfGBM0IK0qlPLENyPctGZ+v2Z2kRBGlnhekfQdNoPrX5uz2Nc0tixH91usX58RbiqK/Yr4pkKdDqlfLOBoIS7BUU51YoluJ/j9ES53zB97Wm8b/LxGaQiaSNjBMEd3I+585RzT1pfwe7eQE2dj12P6CSoOmN2vaf1qm7gZwXqXZKMDRUXQ1ZiGx6zHgphNNSoJ6L8xJ9hJsZmMcfpfkdFjcWCJtoFKnju6/bOQoEn9PEvMPk9ddcBXdVVX9QUt/3M9Xvg89Uoz4MXDUoT8pwXlC+kiVKAwmynj79Xs//6SK+AR6Vllqf70hRgZBgv86RQ/yrCPh8KcLR3hdsz80OAKLwDwhQjny6fimHOZFzNFJBSz6J0u6t429tGA6mCBP1/AdAGjCfWzKdOPHemmY3GkKc6UeN7bIcXDOfVhjss8blwLvwAop2LGsKMae5xjetElr6IeytLo6H6L/MGCxYHixq0RGAhbjuHjBN1UmLaiPs6Z3685XaR02jl1pZntB8uZpMfXnnyoCdY0xTSgKgzZoUIZ6X4/Ha5QHxec/GuP/cELTOjpf2kpsctkdlfNNLOHHntaSIcWCB3MPh7y7J/JjNBsJUtCmcb0linKaQhZAS9OJLKnqbFzmH9YYj864vxbJb4Sytnev20w/kACRlUnxs8LppOEs1FTHIggRpVAkd1fYCKHPZjgHp0y/dCSfbTANLXMBUcZk0FCea5IOhXhqmLy2ODmTihlm23qT4ZkH2UETagPc8qJRkeQ9KzMLYe5OOSKGjcumH5oCRqOxi1hMhT7lQS9Oo+rNG5cE73VJVyXKCG/kMdw8cGC6rCQ08feAtNPsUNhgrj9EboZQhoy/aCm+uGA6vFYFrPLQE2Vhkw/8Sw+kcdkchBRzTStdsH0POb8WULYchSLgPH3La5w6NBTlgHNfoWbWsoLMVqoQOGsEoJZN+fsRQtfe2EyzGvcRy84+qBBde6p9uZUJ5b7H/aZ/7Bg7w8TymPLxbDB4kGNzxxmpyOGJguDH0S4ccnshcDz402FbmrCuz1UpMTN101wpzOyaSgp0XEoobanE9yLMbodYLYSMTIFCp87VL8l+4haYrEufignGd0MiHfETVp/PGD6frmMGftZZCJfzYD/QiXXNHYkR3O3qPGlJ39uSV8vSNctkJEdKOKeQ41LTCMkeL2DuraC++QYe5KhewJj8VmNbgZi4VQV508Tmt2SoOEwo0yOz4lGxQZ7ujwipyHqzjVAbK35J5ZgXJH2JnA8lrj6QKFTRbLmBL7zbIpuBZRjgW+XeUCjVxKtQP6k5PBwhdXOguRajcvc5cXX9FN07Rh8P0IrUSOk1xy+hOpsCXl/rljsK1pvyqIrWnHc6Ioes7Ve4h0UUwG1zz7VxG2LbkZ471gsIrTxJJuOOKl59/YJo6cxUVwzf+iI17zchnaMn5c0byHLuMJjFx730YUs5RgDcO2tGdURmLYFLWMP3TSoKsOOasJ7wMmU2YuAZr/maK/D9XsTqhcVUao4+Shl496C9Y0ZSoMdOdx8io41jUbA9Tfm6H6P7ME5JhV1h9LgrJJfmlru8/jNlui6EZB3USiCmVhq482C9s0aVyLuNq2YPlE0dyQd+PxBRHuzoLwQoH09svhSHIBulFOPLEEC5USTvB7ixhX5yNBeF1Skq2G6Z1h9N7oE59thibGOeEtTnklOXLizTGzJvGAtn1RoA/GOkVRvJwtA7zzZk4rWb6zKz6PBFhqlHEmrIp+FNHolwcJSVwZXwmIRsX59jltAPgkoqoCgrZh8apavIYedOsAQ9hTF1NPp5kubdoGdeRjmpEmEDqG8EKPg3TcG2FKTxBUmhf6NGXWusTNH2O+gtWL+bUdrpSB/DiZwmJUAt7Ci6GinlMfngKLen+BzR6MLOI0/F32+G+X4Zfq0vtZB9x3Vx2fYsSc0RlJOakiua2JnUemyWbGOelDx6P1V5lXIe28O/+pXpL9QHvcLroK4GkFc1VVd1Re2fp67289Tr3QBLo4d3ir0vKacaBjBbJLgbM7FsMk8j7j5+ghbwPkPItrPZ6SvG4xWHP+RYu11h7kWoJoR+XfPCfuAVrS/ZCi/49h/3uPNb4xQqw3CW4lIZWpH2G9CM8afzWBwAdOc7KOMfBaxsllSfDjBFTA6Sums5gw+jNn8VYsyiskPasBiQk/S80S2ZDEMWYwUUVJjlBwTp08szgbEFwJkD7oQbMSs9uYoBUd7HTrtnCCyVIWhpUt+eNzn1mLKrY5AbSZHMYNZg8PBGr8ZP8fWmqRT4yzMpzGt3Yz805LzUZcosByfdejsDmltVEyPI9K2LHHCjmSS1QcZvs5k+aYh3g0pj+VIbDSUp47sYU1j3RJtGYbfN8zmMRvbM0AxfRgyW8R02xmt+ZhibLC1YT4IMFqkWqapOXuWMJg2WTnPePp8jVA7NmZTujdLyrGj0SsxvZD68QUXJynWadavz5kPI9KVkuoc4o6MYnwp4xwVS3bYzX9YUT0u2fvBCsmnFc0di6+h/uSMeuT49EWfu8kZzioanZKgDdmRZnjYYOvNGWfPGoSHVsD1VYhzGqU80aOMZ096tNOC5nxBPQXQKC18Ap87sh/M5DQwr4Sl0YZ6DPZRTrS+zPsziskgIcsjdtIJYduJPM8B1JSzgOqh5BN6B2nfCcfEK9KOZNAVWUhda7yT0cJ8GFEUAcY4GklJceYvU5YHH8Q02iVKeZ5/v8XGzpTpWULxJKeaaSbnKdZqsiqgvVWiDUwHMXFaMZ/G5EVIPihJ+o5yBtVMER9dUD+dEKUKE3tmZxGNXom98NipJ3g3xZ9OKCcaEzuyTx1VbnBOES9qZt+aE3UELq+UJ58Z1swFvnK4uUc3wD09W3bHiuy5RwWe5o2IybdE6zmfJBS1IdRO8vN+JuXhagl3VVd1VVf1N1D+qgP+C3VxktJsF1STEO/BWo3Wjsd7a6w2MvI6YDqIMYFjkUXEixpzWFN+MmEwXuX8+01emw6J1hRnnzZYmWWkr4m0KcsaPJs1uTMeok7n6MqKPCoNRULVTFDTnPI7x5heyOGzLq1GgZ1DPVeMzhrsjzq4M0U3KumPh1Rjz/mwzcmsye7qmHhRE4SWPAvR2pGGHqMd1mqmk4SiDJidRPTSjDC0bOicpOOpc02vt6C5VXOxH7P6VonSmn6S47wSt5qGogzopTndpODF8QpGO1brBVkW8eh8hfDhgL3BCgDNsKYRlZw9SghDx796ssN/fG8fpT1+7IkDh+kavvf/rtOOSjqNnHhY411EZ7dEpwrTVcyHmuJC42tL3PBMZ4kAzytFloc8vujyXjsj2tDYzFGXhoOTFQLjmOxHxM2aKKqJjCWfBaRhRSspGU9TelGF0p7Jfoj72EoH2KgYnjeZnUUMx012eyX5OMA+qAmanvGjmjDWhC1HGAWodgqPp6x0FrJYKwEH5bFjdJTy2vVz5hcRJnBo7UljRz6X5WE9/f/aO5cYyc6rjv+++37Vu7vn4Z4ZzziWrZi8SBRAChCBlHiRfcQGwiabsAxiA4pEskAiSGyyyQaQIqFsWIAiAVmGKAoo2DiZseMZu8cz0z39qK6qe2/VfX4PFrc9GiFn4iEJE5n7k2pRVd991K2qc893zvn+p/uN5aXPdFgglY1jK07ShOU64vvzMb++vSS82+B4Gsc35AsP/r3BjTT5vMsrNKVNlgWcv5qjaoG/Zaju604K0u6SRZc+kCFcQfVqyeIgIoobvEhh2Zpb3x/j2YrJpAA0wjbsvTljNiiI4oa78xHbScH91QCNwC8DQldy/mKGlpAeh/iBJE1DHFtjF11M3HMVsrTQWnD0ZsLROubSNMWyNC8dz3iabjFDVbn8+HBGpWw+dvU+srFY3XaJxw2blYfzcsribsj282Wnb3EicCJI33CIZi2kBfWNHMsW1LnDahURhQ22rZGZoSldZNMttilKj03j0n5nw3I9YPfiiiL3CO61RDug1gatLCygvpFT5AGjcxXzezHnRmuKyqO82yWE3Z/ROL3dE+69TO8B9/T0/JLy3i9DezwPeHPWlqQI2J6u2TsesRWVVMqhkQ6J1yBld0cfjUrWa5+ycGlbm9M64FxUcOvmFjsnayxLc/3mDpcXKVoL0tJn6jUs7wQkWY03XdMsBIPfTkBrq79EQgAADKNJREFUKLuWLSevdXFeYzoPr9lzGI5LJlsFdeMQh12LonpuOLo7ZGtnjVQ2xgiCRFKtHZSyWFceYdwyGpXI1iYeN+QLnyTqyoyUsshv2yhpscpCrnw4xUjQulO9whNsxQXjaUG5coi2JElSU1cOWgvGnsIYOF4OuHxpybOA4youz1LKysW2DGHYEE0a5vsJv7l7SNt0mXJRG4RoESvN+eEa3+s891sHM66dW3ZlXtpQ7XfZeMvt4n5KWQwHFZYHxbI7xjOjDMs2yKWmLlxO04hCOrhaYzsa1QqktEj8hiCRBLnE9yXaCNZ3LMKp4pW757iYrFHaYjooCP0W29EMwpo6d1guI8rW5eJOSlF4yMxiogqC+xk6bynuW7i+IvzVMe3NlINXYibbBVHSMJ8n+J6kyl12djKMNLw1H3NxlDPfT7BtTasslBK4jkJrwcWtjLp2+JgWhG7L4XzI09cWFMtuifx3b+7y/tmCOGpQrUAIg1Q2h3sDkrjGGzecHiQMixIhJMNtjRVblG8ZytyjrLuYrlp21xPAdbvFH/mBix9JfFcynFYsTyKuXlhQVw6h280kQr/FdRWWq9mkAUpbKGlx4ekMo2F5GKGUxezimnLlMdopKZYeu66ibR3Gs4KPtKeUC5ti7REnNYOiZdsp8aJuptK2NpZvcD3F3dfHnJYBs6JEWIa2tbtl/NIiP/Kxr68QDjSlTV05Dz7TaRp1ZXKDmqpwuzi77n4L37t3nokn2WVFK21it2F1p4tFv7U/ZRxV3Lgx5aOXDilXDucma6rKZTIuWK86JbSfXRHCYH5Bguy/LDyWAR7H3dR8FJd4oaKQDmXrsh0VbG/nFGuPrAjYma05Pk2YDguSnYYf/ugcE78m8ZtOqc/AS/d3eHaUEUQt2Spka7hh1Dos8ogwbpC5wRjRdTuWGrMsOXnJRWkLJzZU0mZTBnzohUOEA0bD5fenWIFAFQY0PPVsBsC2yvFHCnckaCvNeFYwtU3X8y2HKgNvyzAOOznCtzsiyAK8MQTzrqQLbRhuV4BAWHDhuTVWIEjfFDgDQexLYiRGdgLq5alDoyxsT5MMKgaXJMJXyFWXcLHsrk5zcKZt4Ab6gcEA8KaGCzsF9XF3/d938ZR4W6JyHmgXeFNDfWqR7DQs90P8WCJsCJKWprGZbBcP9ue4CqlsziUbTjYRtqvZZB5t6zDZLsAyLMuQsnWZDrveY6qGa+NVpxPgSvbmEy5PUjYbH9+TLJcRaeXjWJr5aUIc1jh+Z6yyH3QiN7K18UIFiY89dRmOS8q8+8O/mQ65NsqYTdcYIzi+7jMJKprWplE2V66uGGTVg+Rg6LYko5osDzg/zXFcxdBUrE99btzf4jc+us8z64x17TGZdp99/3jMld0lqrUIRpJmJSgal/M7607knK5u2k0kN/dG+I7CPQs5eIlkZsAYgbAMQSLxtwS7zgongHDd4MeKaEvizyWytQkHLZar8aaQyJp86TO80K3SM4VibIqz7w7CKwpdCfytlnredKLsAzi/vabNYDopMRJmRcF4VuKODcJu8UKFrgXFxqNRNrHbUi0tokCz/WyB2nRSluky4ubLCU9fOyW5IPFWinv3xjz93JJo0nB6P2b/cMwkKXA9ReS3HGYJH92Zo7TF/cMhtujkOQ8WQ3a3UmzLMBhW2KddyCvwW+rGIYlrsizgIE8AuPi/t0sP0XvAPT09PU8A03lW72EeywBHSUO5cRluV+TzgCujjFUZELqGdBmRV343Dd141K2DF0h0C622uHZuibAM+8cjTlcBrjBktYe3DFmWIYnXMBkXRFHTTfvvRaRlwOS0BCRS+liiu/M2mUWlHLLW7abaTvclRecVqjCUcxvbNWclNS5te9aTzFIsTmOmsw22e9Z+RVmkaYi9v6Gtna7xrrTwfIXrK4TfeeKbWwZ/ojFvC21HnZdXzw1NbVPeVzSlQzhqEQ6UC4ey8LgwyclOAlppo29bxNsNTWaRpSG2pRnqisUyppYOly8tyVYhSgsCv6UpJclTEtVaFLmH1gI312xSj3wTMIgrhm5XPL86CDjOEiZPlWT7HkEiKSqPqG1IlxFTU6BVdx3iqEEqG9vXREmD49e0lUW2iPBshe9IsnWAyQUztWY0KWlruyvuFxrb1qAsLMsQ+C1KW1TS5mY65OOjQ9rWIluFBEHbtTRqQLUCvZ8ijxvApm1tmtYmdiTrxsNONcNhRVW7xGFN2zos1wGXGqg2blcaN8k5TWNkY3G0iXFsjacFizxiENZETieEr7SF5yiWiwipbBK/Qasu2eZOBfVbFpHX0qwErtSdDKWrOb4d49maTdsdr9h45FmAVBaNsiGDrfEGy22wfSjmDsYIjvYHXLiWkaYhjbLxAokTweqNLiRykibYtmES1mwObMDGaIFYSHxXs96ziHYUVdalrazc4CWS+X7CaFKyyX206mZG5ZHFOgtoWhtLGLKqm+6X0qEuXcyBxPE0i+OYMGwoW4fEr0mPQ5K6Jl8FD/7Ptg9JUlPULqd5xEjWxEnNpKkQAmxL8/pyzHZQMakLSulQFB6R12C7GscyjEYlrq8ImhZ/IDleDh4sRvp5YN51B/h3z1lH5G8Al+ls4FeNMX9z9t4fAH96NvQrxpi/+yn7+iTwRWPMZ4QQnwP+Etiny0G+Cvy+Mab4Sds/Xk+4np6env9T9Lt8PBZfAG4YYz4EfBL4KyGEJ4SYAl8Cfg34OPAlIcTkMff9TWPMh40xLwAN8NlHDX4sDzhbhZysI1xXc/1wi2vjlJMqYGe4IS99PFuhZGfTlRGczhPuZANiV9I0NkHYclBEtEYQWJqjqrsb3ytCriTgrRWrImCSlHzvcIfYUdxIh7wwTmmUza085uNmzquLCa0WnNQux69f4oPjjElSomRDUXjM84inZlmX/Msjam0T5ZLpZkMlHfYPx5z1KSRvXCJH4qRdadDJJuJ6mrDjS1xL85HiqCuMr1ySoiZfB5yWIdfOL3htf4ursxWb0ufecsQkqLh77OHZnXccOJIobHj9eMZWWHJ7MSI4VriWfhDP3H9zyGtpwsBVOAcKz1YUTec53V2M2C0zHEdzuBoQuy2WZUg3AcebiMBvubM34cZyxMWwYhZWqFLQNA539sYclp2XPS8ivvfKDh/ZWtBqi3kaY1uaxWHMPI+4tnvKS3sXuBBvWDcu8zIkbx1aI/igpbEtzSvzGc+PU7TpPM7QbdFGUNQub6RDDiuX3bBh73jCla0VR2nIUFnUjUPgtyyyiOW/KpT2qaXD7Txh5tcEdpesLFuX1VGAELAsAxa1T6UtqszB9RSnVYCfd8tSf7y/zb0i4PKkS+C+kQ242DrUyua7N3fxLc3uYM28iJDaotWdB+tammQ3Izqv0Kriv26exxIwCSrioGGeR4SuJHIkygjuLoe8msW8eHWfxDZcv7/N9eWYaydrRkHNXjpk7DXczBOubGLOJ2s8W1FuXPyRYpFFvJEOGTiSw/0d4qNu4c9WVGIJw7a/5sZ3pygjsA8Mo7Di5ZMZ7xuu8ReS11cjnIUhsDWXhhnpacj3D3Z432DNURky8hq0EWykgy0ML+/vMHRbZmHFvAzZakt+cDLlA5OU+4sI98QgMFwe5rz6o22ksYgc2bVGEiAsw2btI7XF/VVM1roc1Q6FsvBOFALDfjZg3TocrWOkFtw+nLCRLlthyaioKKTDTvjz8oB/YVUQBhgIIQSQAAtAAp8Gvm2MWQAIIb4NvAj8/cMbCyFeBP4amAP/+U4HEEI4QAwsH3Uiwpifv4vf09PT87MihPhnYOtdDg+A6qHnXzfGfP0n7HcA/CPwPDAAPmuM+ZYQ4otAYIz5ytm4PwNKY8xXH9o2AG4CvwPcAr4JRO8QgrgAvA580jxC1LhPwvX09PxSYox58Re0608DL9MZ0WeAbwshvgO8k4Tb//RQnwf2jDE3AYQQ3wA+/9D73zTG/NGZd/014I+Bv/hJJ9LHgHt6et7TCCG+IIR4+exxEfhD4B9Mxy1gj86w3gMuPbTpLnDwDrv8qWED04UW/gn4rUeN6w1wT0/PexpjzNfOEmMfNsYcAHeA3wUQQpwDngPeBP4F+JQQYnKWfPvU2WsP8xpwVQjxzNnz33vEoT8BvPGoc+tDED09Pf/f+DLwt0KIH9KFHf7EGDMHEEJ8GfiPs3F//nZC7m2MMZUQ4vPAt4QQc+DfgF95aMhnhRCfoHNu7wGfe9SJ9Em4np6enidEH4Lo6enpeUL0Brinp6fnCdEb4J6enp4nRG+Ae3p6ep4QvQHu6enpeUL0Brinp6fnCdEb4J6enp4nxH8DWeNmKPnc7YIAAAAASUVORK5CYII=\n",
293 | "text/plain": [
294 | ""
295 | ]
296 | },
297 | "metadata": {
298 | "needs_background": "light"
299 | },
300 | "output_type": "display_data"
301 | }
302 | ],
303 | "source": [
304 | "\n",
305 | "save_path = 'dog.jpg'\n",
306 | "\n",
307 | "plt.axis('off') # no axis\n",
308 | "librosa.display.specshow(power)\n",
309 | "plt.colorbar(format='%+2.0f dB')\n",
310 | "plt.title('Mel Spectrogram')\n",
311 | "#pylab.savefig(save_path, bbox_inches=None, pad_inches=0)\n",
312 | "#pylab.close()"
313 | ]
314 | },
315 | {
316 | "cell_type": "code",
317 | "execution_count": 12,
318 | "metadata": {},
319 | "outputs": [
320 | {
321 | "data": {
322 | "text/plain": [
323 | "(128, 430)"
324 | ]
325 | },
326 | "execution_count": 12,
327 | "metadata": {},
328 | "output_type": "execute_result"
329 | }
330 | ],
331 | "source": [
332 | "power.shape"
333 | ]
334 | },
335 | {
336 | "cell_type": "code",
337 | "execution_count": 13,
338 | "metadata": {},
339 | "outputs": [
340 | {
341 | "name": "stdout",
342 | "output_type": "stream",
343 | "text": [
344 | "0\n",
345 | "100\n",
346 | "200\n",
347 | "300\n"
348 | ]
349 | },
350 | {
351 | "data": {
352 | "text/plain": [
353 | "(400, 54656, 1)"
354 | ]
355 | },
356 | "execution_count": 13,
357 | "metadata": {},
358 | "output_type": "execute_result"
359 | }
360 | ],
361 | "source": [
362 | "melspectrogram = []\n",
363 | "for i in range (0,400):\n",
364 | " if i%100 == 0:\n",
365 | " print(i)\n",
366 | " mel_feat = librosa.feature.melspectrogram(y=audio_data[i],sr=44000)\n",
367 | " power = librosa.power_to_db(mel_feat)\n",
368 | "\n",
369 | " power=power.reshape(-1,1)\n",
370 | " \n",
371 | "\n",
372 | " melspectrogram.append(power[:54656])\n",
373 | " if power.shape[0]<54656:\n",
374 | " print(i)\n",
375 | " print(power.shape)\n",
376 | " \n",
377 | "melspectrogram =np.array(melspectrogram) \n",
378 | "melspectrogram.shape\n",
379 | " "
380 | ]
381 | },
382 | {
383 | "cell_type": "code",
384 | "execution_count": null,
385 | "metadata": {},
386 | "outputs": [],
387 | "source": []
388 | },
389 | {
390 | "cell_type": "code",
391 | "execution_count": null,
392 | "metadata": {},
393 | "outputs": [],
394 | "source": []
395 | },
396 | {
397 | "cell_type": "code",
398 | "execution_count": 14,
399 | "metadata": {},
400 | "outputs": [
401 | {
402 | "name": "stdout",
403 | "output_type": "stream",
404 | "text": [
405 | "36.726677\n",
406 | "1.0000001\n"
407 | ]
408 | }
409 | ],
410 | "source": [
411 | "from sklearn.preprocessing import MinMaxScaler\n",
412 | "scaler = MinMaxScaler(feature_range=(0, 1))\n",
413 | "melspectrogram=melspectrogram.reshape(400,-1)\n",
414 | "scaler.fit(melspectrogram)\n",
415 | "normalized_melspectrogram = scaler.transform(melspectrogram)\n",
416 | "\n",
417 | "print(np.amax(melspectrogram))\n",
418 | "print(np.amax(normalized_melspectrogram))\n",
419 | "\n"
420 | ]
421 | },
422 | {
423 | "cell_type": "code",
424 | "execution_count": 15,
425 | "metadata": {},
426 | "outputs": [
427 | {
428 | "data": {
429 | "text/plain": [
430 | "DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
431 | " max_features=None, max_leaf_nodes=None,\n",
432 | " min_impurity_decrease=0.0, min_impurity_split=None,\n",
433 | " min_samples_leaf=1, min_samples_split=2,\n",
434 | " min_weight_fraction_leaf=0.0, presort=False, random_state=None,\n",
435 | " splitter='best')"
436 | ]
437 | },
438 | "execution_count": 15,
439 | "metadata": {},
440 | "output_type": "execute_result"
441 | }
442 | ],
443 | "source": [
444 | "from sklearn import tree\n",
445 | "X_train, X_test, y_train, y_test = train_test_split(normalized_melspectrogram, labels, test_size=0.20, random_state=1150)\n",
446 | "clf = tree.DecisionTreeClassifier()\n",
447 | "clf.fit(X_train, y_train)"
448 | ]
449 | },
450 | {
451 | "cell_type": "code",
452 | "execution_count": 16,
453 | "metadata": {},
454 | "outputs": [
455 | {
456 | "name": "stdout",
457 | "output_type": "stream",
458 | "text": [
459 | "0.5125\n"
460 | ]
461 | }
462 | ],
463 | "source": [
464 | "y_pred = clf.predict(X_test)\n",
465 | "print(accuracy_score(y_test, y_pred))"
466 | ]
467 | },
468 | {
469 | "cell_type": "code",
470 | "execution_count": 17,
471 | "metadata": {},
472 | "outputs": [
473 | {
474 | "data": {
475 | "text/plain": [
476 | "(400, 128, 427, 1)"
477 | ]
478 | },
479 | "execution_count": 17,
480 | "metadata": {},
481 | "output_type": "execute_result"
482 | }
483 | ],
484 | "source": [
485 | "features_convolution = np.reshape(normalized_melspectrogram,(400,128, -1,1))\n",
486 | "#melspectrogram=melspectrogram.reshape(400,128, -1)\n",
487 | "features_convolution.shape\n"
488 | ]
489 | },
490 | {
491 | "cell_type": "code",
492 | "execution_count": 18,
493 | "metadata": {},
494 | "outputs": [],
495 | "source": [
496 | "y=keras.utils.to_categorical(labels, num_classes=10, dtype='float32')\n"
497 | ]
498 | },
499 | {
500 | "cell_type": "code",
501 | "execution_count": null,
502 | "metadata": {},
503 | "outputs": [],
504 | "source": []
505 | },
506 | {
507 | "cell_type": "code",
508 | "execution_count": 19,
509 | "metadata": {},
510 | "outputs": [],
511 | "source": [
512 | "model = Sequential()\n",
513 | "\n",
514 | "model.add(Conv2D(16, (3, 3), input_shape=features_convolution.shape[1:]))\n",
515 | "model.add(Activation('relu'))\n",
516 | "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
517 | "#'''\n",
518 | "#model.add(Dropout(0.2))\n",
519 | "\n",
520 | "model.add(Conv2D(32, (3, 3)))\n",
521 | "model.add(Activation('relu'))\n",
522 | "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
523 | "\n",
524 | "#model.add(Dropout(0.2))\n",
525 | "\n",
526 | "#'''\n",
527 | "#'''\n",
528 | "model.add(Conv2D(64, (3, 3)))\n",
529 | "model.add(Activation('relu'))\n",
530 | "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
531 | "#'''\n",
532 | "\n",
533 | "\n",
534 | "model.add(Flatten()) # this converts our 3D feature maps to 1D feature vectors\n",
535 | "\n",
536 | "#model.add(Dense(1000))#input_shape=features.shape[1:]\n",
537 | "model.add(Dense(64))#input_shape=features.shape[1:]\n",
538 | "\n",
539 | "model.add(Dense(10))\n",
540 | "model.add(Activation('softmax'))\n",
541 | "sgd = optimizers.SGD(lr=0.0000001, decay=1e-6, momentum=0.9, nesterov=True)\n",
542 | "\n",
543 | "model.compile(loss='categorical_crossentropy',\n",
544 | " optimizer='adam',\n",
545 | " metrics=['accuracy'])"
546 | ]
547 | },
548 | {
549 | "cell_type": "code",
550 | "execution_count": 20,
551 | "metadata": {},
552 | "outputs": [
553 | {
554 | "name": "stdout",
555 | "output_type": "stream",
556 | "text": [
557 | "Train on 320 samples, validate on 80 samples\n",
558 | "Epoch 1/30\n",
559 | "320/320 [==============================] - 32s 101ms/step - loss: 2.2985 - acc: 0.2031 - val_loss: 1.7388 - val_acc: 0.3250\n",
560 | "Epoch 2/30\n",
561 | "320/320 [==============================] - 26s 81ms/step - loss: 1.3187 - acc: 0.5156 - val_loss: 1.5482 - val_acc: 0.4500\n",
562 | "Epoch 3/30\n",
563 | "320/320 [==============================] - 26s 82ms/step - loss: 0.9481 - acc: 0.6344 - val_loss: 1.1561 - val_acc: 0.5125\n",
564 | "Epoch 4/30\n",
565 | "320/320 [==============================] - 27s 84ms/step - loss: 0.7318 - acc: 0.7281 - val_loss: 1.0862 - val_acc: 0.5750\n",
566 | "Epoch 5/30\n",
567 | "320/320 [==============================] - 28s 86ms/step - loss: 0.5510 - acc: 0.7969 - val_loss: 1.0911 - val_acc: 0.6125\n",
568 | "Epoch 6/30\n",
569 | "320/320 [==============================] - 28s 88ms/step - loss: 0.4617 - acc: 0.8438 - val_loss: 0.9110 - val_acc: 0.6625\n",
570 | "Epoch 7/30\n",
571 | "320/320 [==============================] - 29s 90ms/step - loss: 0.3902 - acc: 0.8687 - val_loss: 1.0375 - val_acc: 0.6375\n",
572 | "Epoch 8/30\n",
573 | "320/320 [==============================] - 29s 90ms/step - loss: 0.4344 - acc: 0.8375 - val_loss: 1.4428 - val_acc: 0.5125\n",
574 | "Epoch 9/30\n",
575 | "320/320 [==============================] - 29s 91ms/step - loss: 0.4131 - acc: 0.8625 - val_loss: 0.9190 - val_acc: 0.6500\n",
576 | "Epoch 10/30\n",
577 | "320/320 [==============================] - 28s 88ms/step - loss: 0.2367 - acc: 0.9281 - val_loss: 0.9850 - val_acc: 0.7375\n",
578 | "Epoch 11/30\n",
579 | "320/320 [==============================] - 29s 91ms/step - loss: 0.1863 - acc: 0.9500 - val_loss: 0.9003 - val_acc: 0.7375\n",
580 | "Epoch 12/30\n",
581 | "320/320 [==============================] - 28s 89ms/step - loss: 0.1662 - acc: 0.9594 - val_loss: 1.0980 - val_acc: 0.6750\n",
582 | "Epoch 13/30\n",
583 | "320/320 [==============================] - 29s 90ms/step - loss: 0.2715 - acc: 0.9219 - val_loss: 1.2641 - val_acc: 0.6875\n",
584 | "Epoch 14/30\n",
585 | "320/320 [==============================] - 29s 90ms/step - loss: 0.1068 - acc: 0.9656 - val_loss: 1.1858 - val_acc: 0.7375\n",
586 | "Epoch 15/30\n",
587 | "320/320 [==============================] - 29s 91ms/step - loss: 0.0652 - acc: 0.9875 - val_loss: 1.0963 - val_acc: 0.8000\n",
588 | "Epoch 16/30\n",
589 | "320/320 [==============================] - 28s 88ms/step - loss: 0.0379 - acc: 0.9906 - val_loss: 1.1565 - val_acc: 0.7500\n",
590 | "Epoch 17/30\n",
591 | "320/320 [==============================] - 29s 90ms/step - loss: 0.0357 - acc: 0.9937 - val_loss: 0.9965 - val_acc: 0.7625\n",
592 | "Epoch 18/30\n",
593 | "320/320 [==============================] - 31s 96ms/step - loss: 0.0283 - acc: 0.9906 - val_loss: 1.0034 - val_acc: 0.7750\n",
594 | "Epoch 19/30\n",
595 | "320/320 [==============================] - 30s 93ms/step - loss: 0.0114 - acc: 1.0000 - val_loss: 1.0554 - val_acc: 0.7375\n",
596 | "Epoch 20/30\n",
597 | "320/320 [==============================] - 30s 95ms/step - loss: 0.0078 - acc: 1.0000 - val_loss: 1.1256 - val_acc: 0.7500\n",
598 | "Epoch 21/30\n",
599 | "320/320 [==============================] - 32s 99ms/step - loss: 0.0053 - acc: 1.0000 - val_loss: 1.1333 - val_acc: 0.7875\n",
600 | "Epoch 22/30\n",
601 | "320/320 [==============================] - 29s 91ms/step - loss: 0.0038 - acc: 1.0000 - val_loss: 1.1276 - val_acc: 0.8125\n",
602 | "Epoch 23/30\n",
603 | "320/320 [==============================] - 30s 93ms/step - loss: 0.0031 - acc: 1.0000 - val_loss: 1.1474 - val_acc: 0.8000\n",
604 | "Epoch 24/30\n",
605 | "320/320 [==============================] - 30s 93ms/step - loss: 0.0026 - acc: 1.0000 - val_loss: 1.1681 - val_acc: 0.7750\n",
606 | "Epoch 25/30\n",
607 | "320/320 [==============================] - 29s 91ms/step - loss: 0.0023 - acc: 1.0000 - val_loss: 1.1910 - val_acc: 0.7750\n",
608 | "Epoch 26/30\n",
609 | "320/320 [==============================] - 29s 90ms/step - loss: 0.0019 - acc: 1.0000 - val_loss: 1.1861 - val_acc: 0.7875\n",
610 | "Epoch 27/30\n",
611 | "320/320 [==============================] - 26s 83ms/step - loss: 0.0018 - acc: 1.0000 - val_loss: 1.1995 - val_acc: 0.8000\n",
612 | "Epoch 28/30\n",
613 | "320/320 [==============================] - 26s 80ms/step - loss: 0.0015 - acc: 1.0000 - val_loss: 1.2172 - val_acc: 0.7875\n",
614 | "Epoch 29/30\n",
615 | "320/320 [==============================] - 26s 80ms/step - loss: 0.0014 - acc: 1.0000 - val_loss: 1.2272 - val_acc: 0.7875\n",
616 | "Epoch 30/30\n",
617 | "320/320 [==============================] - 26s 83ms/step - loss: 0.0013 - acc: 1.0000 - val_loss: 1.2500 - val_acc: 0.7625\n"
618 | ]
619 | }
620 | ],
621 | "source": [
622 | "history = model.fit(features_convolution, y,batch_size=8, epochs=30,validation_split=0.2)"
623 | ]
624 | },
625 | {
626 | "cell_type": "code",
627 | "execution_count": null,
628 | "metadata": {},
629 | "outputs": [],
630 | "source": []
631 | },
632 | {
633 | "cell_type": "code",
634 | "execution_count": 21,
635 | "metadata": {},
636 | "outputs": [
637 | {
638 | "data": {
639 | "image/png": "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\n",
640 | "text/plain": [
641 | ""
642 | ]
643 | },
644 | "metadata": {
645 | "needs_background": "light"
646 | },
647 | "output_type": "display_data"
648 | },
649 | {
650 | "data": {
651 | "text/plain": [
652 | ""
653 | ]
654 | },
655 | "metadata": {},
656 | "output_type": "display_data"
657 | }
658 | ],
659 | "source": [
660 | "import matplotlib.pyplot as plt\n",
661 | "acc = history.history['acc']\n",
662 | "val_acc = history.history['val_acc']\n",
663 | "loss = history.history['loss']\n",
664 | "val_loss = history.history['val_loss']\n",
665 | "\n",
666 | "epochs = range(len(acc))\n",
667 | "\n",
668 | "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
669 | "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
670 | "plt.title('Training and validation accuracy')\n",
671 | "plt.legend(loc=0)\n",
672 | "plt.figure()\n",
673 | "\n",
674 | "\n",
675 | "plt.show()"
676 | ]
677 | },
678 | {
679 | "cell_type": "code",
680 | "execution_count": null,
681 | "metadata": {},
682 | "outputs": [],
683 | "source": []
684 | }
685 | ],
686 | "metadata": {
687 | "kernelspec": {
688 | "display_name": "Python 3",
689 | "language": "python",
690 | "name": "python3"
691 | },
692 | "language_info": {
693 | "codemirror_mode": {
694 | "name": "ipython",
695 | "version": 3
696 | },
697 | "file_extension": ".py",
698 | "mimetype": "text/x-python",
699 | "name": "python",
700 | "nbconvert_exporter": "python",
701 | "pygments_lexer": "ipython3",
702 | "version": "3.6.6"
703 | }
704 | },
705 | "nbformat": 4,
706 | "nbformat_minor": 2
707 | }
708 |
--------------------------------------------------------------------------------