{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"# Time Series Data Mining with Python\n",
"\n",
"Feng Li\n",
"\n",
"School of Statistics and Mathematics\n",
"\n",
"Central University of Finance and Economics\n",
"\n",
"[feng.li@cufe.edu.cn](mailto:feng.li@cufe.edu.cn)\n",
"\n",
"[https://feng.li/python](https://feng.li/python)\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Textbooks\n",
"\n",
"- We will stick fairly close to course materials but the following books can *supplement* your learning\n",
" - Hyndman, Rob J and George Athanasopoulos, *Forecasting: Principles and Practice*. \n",
" - [English Free online](http://otexts.com/fpp3)\n",
" - [Chinese Translation](https://otexts.com/fppcn/)\n",
" - Tsay, Ruey, (2010), *Analysis of Financial Time Series*, Available from major retailers."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Software Tools\n",
"\n",
"- In this course we use Python. If you do not have Python experience, watch my teaching videso at https://feng.li/python.\n",
"- Lecture slides have embedded Python code and can be used interactively.\n",
"- Tutorials are built on Jupyter Notebooks.\n",
"- The emphasis is on forecasting in *practice* as well as the theory of forecasting."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Time Series and Forecasting \n",
"\n",
"- The focus of this course is **forecasting** (what a suprise).\n",
"- In this course we mostly consider numerical variables \n",
" - Example: electricity demand\n",
"- Will mostly use the history of the variable we want to forecast or **time series** itself.\n",
" - Make a forecast $\\hat{y}_{T+h}$ targeting $y_{T+h}$\n",
" - Use $y_1,y_2,\\dots,y_T$\n",
"- Here $h$ is the forecast horizon."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Forecasting vs Prediction\n",
"\n",
"- In other courses you have made predictions $\\hat{y}=f(\\mathbf{x})$, where $\\mathbf{x}$ are some independent variables.\n",
"- A forecast is just a prediction in the future\n",
"- Any method can be used if $f(\\mathbf{x})$ includes information available when the forecast is made\n",
"- Unlike general regression methods, in time series forecasting there are techniques to exploit the temporal structure or patterns in the data."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## What can we easily forecast?\n",
"\n",
"In general it is easier to forecast a variable when:\n",
"\n",
"1. We understand the factors that contribute to it;\n",
"2. When more data are available;\n",
"3. When the future is similar to the past;\n",
"4. When the forecasts do not affect the quantity we are trying to forecast.\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Your turn\n",
"\n",
"- In small groups have a think about:\n",
" - Variables that we need to forecast,\n",
" - How forecasting these variables can improve business decisions,\n",
" - Whether it is easy to forecast these variable.\n",
"- Discuss in groups for 10 minutes and then we will discuss together."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Electricity Demand\n",
"\n",
"- Forecasts at different **horizons** play an important role in the operation of electricity markets.\n",
"- Long term forecasts (10 to 30 years horizon) \n",
" - Used by by policy makers or investors to decide how to invest in infrastructure.\n",
"- Short term forecasts (up to 24-hour horizon) \n",
" - Used for operations, e.g. scheduling which generators produce electricity.\n",
"- Consider the short term forecasting problem to revise exponential smoothing models covered last semester.\n",
"- Data sourced from the [Australian Energy Market Operator (AEMO)](https://aemo.com.au/energy-systems/electricity/national-electricity-market-nem/data-nem/aggregated-data) and on Canvas.\n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Electricity Data"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"scrolled": true,
"slideshow": {
"slide_type": "fragment"
}
},
"outputs": [
{
"data": {
"text/html": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"pred = fit.forecast(48)\n",
"fig, ax = plt.subplots(figsize=(25, 6))\n",
"ax.plot(train.index[-96:], train.values[-96:],color='black')\n",
"ax.plot(train.index[-96:], fitted[-96:], color='orange')\n",
"ax.plot(test.index, pred,color='orangered')\n",
"ax.plot(test.index, test.values, color='grey')\n",
"plt.title(\"Simple Exponential Smoothing (Black/grey = Actual, Orange = Fitted/Forecast)\");"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Holt-Winters exponential smoothing\n",
"\n",
"- Corrects for trend and seasonality. Additive form is:\n",
"\n",
"$$\\begin{aligned}\\hat{y}_{t+h}&=l_t+hb_t+s_{t+h-m(k+1)}\\\\l_{t}&=\\alpha y_t+(1-\\alpha)(l_{t-1}+b_{t-1})\\\\b_t&=\\beta(l_t-l_{t-1})+(1-\\beta)b_{t-1}\\\\s_t&=\\gamma(y_t-l_{t-1}-b_{t-1})+(1-\\gamma)s_{t-m}\\end{aligned}$$\n",
"\n",
"- where $0\\leq\\alpha\\leq 1$, $0\\leq\\beta\\leq 1$, $0\\leq\\gamma\\leq 1$ and $l_0$ and $b_0$ and $s_0,s_{-1},s_{-2},...$ need to be estimated.\n",
"- Note $m$ is the seasonal period and $k$ is the integer part of $(h-1)/m$.\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Seasonal period\n",
"\n",
"- Think of *seasonal period* as the number of observations required for a pattern to repeat itself\n",
"- For example monthly data has a seasonal period of 12, for a pattern repeating every year\n",
"- In our example the period repeats every day. \n",
"- With half hourly data that implies a seasonal period of 48.\n",
"- An argument could also be made for a seasonal period of $48\\times 7=336$ or $48\\times 365=17520$."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"### Fit model (Holt-Winters)\n",
"\n",
"The following code fits the model"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"scrolled": false,
"slideshow": {
"slide_type": "fragment"
}
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"
ExponentialSmoothing Model Results
\n",
"
\n",
"
Dep. Variable:
endog
No. Observations:
1392
\n",
"
\n",
"
\n",
"
Model:
ExponentialSmoothing
SSE
7666658.978
\n",
"
\n",
"
\n",
"
Optimized:
True
AIC
12094.541
\n",
"
\n",
"
\n",
"
Trend:
Additive
BIC
12366.943
\n",
"
\n",
"
\n",
"
Seasonal:
Additive
AICC
12098.984
\n",
"
\n",
"
\n",
"
Seasonal Periods:
48
Date:
Fri, 17 Jun 2022
\n",
"
\n",
"
\n",
"
Box-Cox:
False
Time:
23:09:39
\n",
"
\n",
"
\n",
"
Box-Cox Coeff.:
None
\n",
"
\n",
"
\n",
"
\n",
"
\n",
"
coeff
code
optimized
\n",
"
\n",
"
\n",
"
smoothing_level
0.9381307
alpha
True
\n",
"
\n",
"
\n",
"
smoothing_trend
0.9112998
beta
True
\n",
"
\n",
"
\n",
"
smoothing_seasonal
0.0247337
gamma
True
\n",
"
\n",
"
\n",
"
initial_level
7214.7871
l.0
True
\n",
"
\n",
"
\n",
"
initial_trend
-29.817169
b.0
True
\n",
"
\n",
"
\n",
"
initial_seasons.0
93.239648
s.0
True
\n",
"
\n",
"
\n",
"
initial_seasons.1
-45.832722
s.1
True
\n",
"
\n",
"
\n",
"
initial_seasons.2
-242.02690
s.2
True
\n",
"
\n",
"
\n",
"
initial_seasons.3
-407.34103
s.3
True
\n",
"
\n",
"
\n",
"
initial_seasons.4
-599.31886
s.4
True
\n",
"
\n",
"
\n",
"
initial_seasons.5
-724.76436
s.5
True
\n",
"
\n",
"
\n",
"
initial_seasons.6
-769.41277
s.6
True
\n",
"
\n",
"
\n",
"
initial_seasons.7
-792.96722
s.7
True
\n",
"
\n",
"
\n",
"
initial_seasons.8
-781.48527
s.8
True
\n",
"
\n",
"
\n",
"
initial_seasons.9
-771.21668
s.9
True
\n",
"
\n",
"
\n",
"
initial_seasons.10
-676.51721
s.10
True
\n",
"
\n",
"
\n",
"
initial_seasons.11
-553.77344
s.11
True
\n",
"
\n",
"
\n",
"
initial_seasons.12
-276.17953
s.12
True
\n",
"
\n",
"
\n",
"
initial_seasons.13
-100.69164
s.13
True
\n",
"
\n",
"
\n",
"
initial_seasons.14
-75.433699
s.14
True
\n",
"
\n",
"
\n",
"
initial_seasons.15
-119.39831
s.15
True
\n",
"
\n",
"
\n",
"
initial_seasons.16
-219.45105
s.16
True
\n",
"
\n",
"
\n",
"
initial_seasons.17
-356.78664
s.17
True
\n",
"
\n",
"
\n",
"
initial_seasons.18
-444.39352
s.18
True
\n",
"
\n",
"
\n",
"
initial_seasons.19
-562.57459
s.19
True
\n",
"
\n",
"
\n",
"
initial_seasons.20
-612.12994
s.20
True
\n",
"
\n",
"
\n",
"
initial_seasons.21
-668.48214
s.21
True
\n",
"
\n",
"
\n",
"
initial_seasons.22
-720.11188
s.22
True
\n",
"
\n",
"
\n",
"
initial_seasons.23
-732.48385
s.23
True
\n",
"
\n",
"
\n",
"
initial_seasons.24
-716.98479
s.24
True
\n",
"
\n",
"
\n",
"
initial_seasons.25
-682.49037
s.25
True
\n",
"
\n",
"
\n",
"
initial_seasons.26
-619.63748
s.26
True
\n",
"
\n",
"
\n",
"
initial_seasons.27
-522.08684
s.27
True
\n",
"
\n",
"
\n",
"
initial_seasons.28
-358.79810
s.28
True
\n",
"
\n",
"
\n",
"
initial_seasons.29
-164.10064
s.29
True
\n",
"
\n",
"
\n",
"
initial_seasons.30
80.519978
s.30
True
\n",
"
\n",
"
\n",
"
initial_seasons.31
374.56130
s.31
True
\n",
"
\n",
"
\n",
"
initial_seasons.32
642.67196
s.32
True
\n",
"
\n",
"
\n",
"
initial_seasons.33
958.01110
s.33
True
\n",
"
\n",
"
\n",
"
initial_seasons.34
1225.5055
s.34
True
\n",
"
\n",
"
\n",
"
initial_seasons.35
1465.0462
s.35
True
\n",
"
\n",
"
\n",
"
initial_seasons.36
1531.8976
s.36
True
\n",
"
\n",
"
\n",
"
initial_seasons.37
1404.5379
s.37
True
\n",
"
\n",
"
\n",
"
initial_seasons.38
1165.0485
s.38
True
\n",
"
\n",
"
\n",
"
initial_seasons.39
979.30432
s.39
True
\n",
"
\n",
"
\n",
"
initial_seasons.40
850.12616
s.40
True
\n",
"
\n",
"
\n",
"
initial_seasons.41
737.25041
s.41
True
\n",
"
\n",
"
\n",
"
initial_seasons.42
650.60912
s.42
True
\n",
"
\n",
"
\n",
"
initial_seasons.43
540.71120
s.43
True
\n",
"
\n",
"
\n",
"
initial_seasons.44
549.93113
s.44
True
\n",
"
\n",
"
\n",
"
initial_seasons.45
434.20512
s.45
True
\n",
"
\n",
"
\n",
"
initial_seasons.46
340.90809
s.46
True
\n",
"
\n",
"
\n",
"
initial_seasons.47
245.39567
s.47
True
\n",
"
\n",
"
"
],
"text/plain": [
"\n",
"\"\"\"\n",
" ExponentialSmoothing Model Results \n",
"================================================================================\n",
"Dep. Variable: endog No. Observations: 1392\n",
"Model: ExponentialSmoothing SSE 7666658.978\n",
"Optimized: True AIC 12094.541\n",
"Trend: Additive BIC 12366.943\n",
"Seasonal: Additive AICC 12098.984\n",
"Seasonal Periods: 48 Date: Fri, 17 Jun 2022\n",
"Box-Cox: False Time: 23:09:39\n",
"Box-Cox Coeff.: None \n",
"=================================================================================\n",
" coeff code optimized \n",
"---------------------------------------------------------------------------------\n",
"smoothing_level 0.9381307 alpha True\n",
"smoothing_trend 0.9112998 beta True\n",
"smoothing_seasonal 0.0247337 gamma True\n",
"initial_level 7214.7871 l.0 True\n",
"initial_trend -29.817169 b.0 True\n",
"initial_seasons.0 93.239648 s.0 True\n",
"initial_seasons.1 -45.832722 s.1 True\n",
"initial_seasons.2 -242.02690 s.2 True\n",
"initial_seasons.3 -407.34103 s.3 True\n",
"initial_seasons.4 -599.31886 s.4 True\n",
"initial_seasons.5 -724.76436 s.5 True\n",
"initial_seasons.6 -769.41277 s.6 True\n",
"initial_seasons.7 -792.96722 s.7 True\n",
"initial_seasons.8 -781.48527 s.8 True\n",
"initial_seasons.9 -771.21668 s.9 True\n",
"initial_seasons.10 -676.51721 s.10 True\n",
"initial_seasons.11 -553.77344 s.11 True\n",
"initial_seasons.12 -276.17953 s.12 True\n",
"initial_seasons.13 -100.69164 s.13 True\n",
"initial_seasons.14 -75.433699 s.14 True\n",
"initial_seasons.15 -119.39831 s.15 True\n",
"initial_seasons.16 -219.45105 s.16 True\n",
"initial_seasons.17 -356.78664 s.17 True\n",
"initial_seasons.18 -444.39352 s.18 True\n",
"initial_seasons.19 -562.57459 s.19 True\n",
"initial_seasons.20 -612.12994 s.20 True\n",
"initial_seasons.21 -668.48214 s.21 True\n",
"initial_seasons.22 -720.11188 s.22 True\n",
"initial_seasons.23 -732.48385 s.23 True\n",
"initial_seasons.24 -716.98479 s.24 True\n",
"initial_seasons.25 -682.49037 s.25 True\n",
"initial_seasons.26 -619.63748 s.26 True\n",
"initial_seasons.27 -522.08684 s.27 True\n",
"initial_seasons.28 -358.79810 s.28 True\n",
"initial_seasons.29 -164.10064 s.29 True\n",
"initial_seasons.30 80.519978 s.30 True\n",
"initial_seasons.31 374.56130 s.31 True\n",
"initial_seasons.32 642.67196 s.32 True\n",
"initial_seasons.33 958.01110 s.33 True\n",
"initial_seasons.34 1225.5055 s.34 True\n",
"initial_seasons.35 1465.0462 s.35 True\n",
"initial_seasons.36 1531.8976 s.36 True\n",
"initial_seasons.37 1404.5379 s.37 True\n",
"initial_seasons.38 1165.0485 s.38 True\n",
"initial_seasons.39 979.30432 s.39 True\n",
"initial_seasons.40 850.12616 s.40 True\n",
"initial_seasons.41 737.25041 s.41 True\n",
"initial_seasons.42 650.60912 s.42 True\n",
"initial_seasons.43 540.71120 s.43 True\n",
"initial_seasons.44 549.93113 s.44 True\n",
"initial_seasons.45 434.20512 s.45 True\n",
"initial_seasons.46 340.90809 s.46 True\n",
"initial_seasons.47 245.39567 s.47 True\n",
"---------------------------------------------------------------------------------\n",
"\"\"\""
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"model = ExponentialSmoothing(np.asarray(train['TOTALDEMAND']),trend='add',seasonal='add',seasonal_periods=48)\n",
"fit = model.fit()\n",
"fit.summary()\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"### Plot (Holt-Winters)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"scrolled": true,
"slideshow": {
"slide_type": "fragment"
}
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAABaEAAAF1CAYAAADx3iZ5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOzdd1yVdf/H8deXDYogAiqi4p64NQea2jA1G1aaDa37Lisrm3fl3d7dLUtbtn7tbaWWI0pFSUURJyqKCwFlKsoQGd/fH9ehSM91BDyD8Xk+Hj5Q+HBdnwPngOdzvtf7q7TWCCGEEEIIIYQQQgghhBCO4ObqBoQQQgghhBBCCCGEEELUXzKEFkIIIYQQQgghhBBCCOEwMoQWQgghhBBCCCGEEEII4TAyhBZCCCGEEEIIIYQQQgjhMDKEFkIIIYQQQgghhBBCCOEwMoQWQgghhBBCCCGEEEII4TAyhBZCCCEaAKXU9Uqp3xx07E+UUs854th1nVKqjVIqXynlXoXakUqpVGf0ZU9KqaeUUl/Y+HiiUmqkg84dopRKUkr5VLHeZq9VPEaEUkorpTzO5TjCNvk6nxvLz532Djz+OT+W6gKlVC+l1BpX9yGEEELUBzKEFkIIIeoJpVSUUmqNUipPKZWrlPpTKTUQQGv9pdb6Ylf3eDrLkKnAMjCp+POQq/uqKaXUAaXUhRX/1lqnaK0ba63L7HDsy5VSm5VSx5VS2UqpP5RSEed63Gr2UO1Buda6h9Z6pYNaegT4P631SQCl1Eql1EnL/ShPKbVKKRXpoHObUkpdp5T6ytnntSel1E2Wx+ekanyOVkp1dGRfNs7trZR6USmVopQqUkrtUUr9RymlXNGPs1h+5hSd9jM0zPJzZ5+l5owXCk//WWWnXq5TSn1V6QWEyj1tsee5HOX0+7DWeitwTCk1wYVtCSGEEPWCDKGFEEKIekAp1QT4BZgLBAGtgKeBYlf2VUW9LQOTij8vu7qh2sYyFPkMeAAIANoB7wDlruzLlZRS3sA04PTVmHdprRsDzYCVwOdObg1gHLC4up9Uy1b9TgNyLW/rgu+BCzC+9v7AjcB04E1rxcpQX54LTTjtZ2i6i/o4/X4fWKmn3tU9WC16PHwJ3ObqJoQQQoi6rr78x0sIIYRo6DoDaK2/1lqXaa2LtNa/WVZxVaxqjK0otqz2mmFZLXhCKfWsUqqDUmqtZaXtd0opL0vtSKVUqlLqv5YVuAeUUtebNaKUutSyYveYZWV2r5rcIKXUYqXUa5X+/a1S6uNKt+dPpdRcy4rXXUqpCyrVhimlFlpWhCcrpW6t9LGnLLfvM8ttT1RKDTjtc+crpbKUUvuVUjOr8rlKqc+BNsCiihXd6rRIAaXUzUqpnZbP3aeUqupgow+wX2v9hzac0FrP11qnVOrre6XUF5Zjb1NKdVZKzVJKZSqlDiml/loJf5avj7dS6g2lVLrlzxuW9zUClgBhlVdcWj7Ny8bX868Vl1X42vdTSm2yfOx7y/fcLOrlPOCY1trqymytdSnwDdDd7ItqOccR9feq6R6VPuarlHpNKXXQ8vFYpZSvlWNcZbmNPS3/dgMuApZa/j3VcowcpdTjVr4eP1i+b8eBm5RSAUqpj5RSh5VSaUqp55RS7pbvQa6qtLJbKRWqjFWwIWa3sSaUUm2B8zGGuGOUUs0rfcxdGT8L9lq+TxuVUq2VUqssJVss943J6rSfO5bP/2ulqVJqvOX7fdxyH32qhv1eAFwMXKW13q61LtVarwNuAO6sdL6VSqnnlVJ/AoVAe1uPSfX3z74HLI+jw0qpmyt9vJlSapGl/w2W71Xln7NdlVLRlu9bkqrGqvJzVfF1VkpNB64HHrJ8XxYpKz+rLJ8zWBk/s48ppbaoSjE6Sql2SqkYy9cpGgg+7Xz/uN/b6OtsP5ur9Hio9Dm3Vvr+7VBK9bO8/5FK99EdSqkrK31OR8ttyVPG77RvLe8/4z5s+fdK4AJlvPAlhBBCiBqSIbQQQghRP+wGypRSnyqlxiqlmlbhcy4B+gODgYeA9zGGFa2BnsCUSrUtMIYOrTBWRr6vlOpy+gEtA4CPMVaNNQPmAQtr+OT9X8CNSqnRyhh6DwTuqfTx84B9lr6eBH5USgVZPvY1kAqEAVcDL6hKQ2rgMowBZSCwEHjL0r8bsAjYYrmtFwD3KqXGnO1ztdY3Ain8vSrR2oruTOBSoAlwMzC7YmhyFglAV6XUbKXUKKVUYys1EzBW/TYFNgHLMP6v1wp4BuN7UcHW1+dRjPtEH6A3MAh4TGtdAIwF0q2suLT6NTFh9rX3An4CPsFYzf81cKXVIxgigSSzD1qOdz2wzsYxlgCdgFCMr/GXlT72KsbjY6iln4c4beW5ZSD5P+BCrfV2y7sHAfu01tlKqe4YK9avB1pirGJvdVoPlwM/YHw9vgQ+BUqBjkBfjOHqLVrrYoyv2w2VPncK8LvWOsvK7Y+yDBPN/kTZ+LpMBeK11vOBnZb+K9xvOe84jPvxv4BCrfUIy8crrmz41sbxKxRYzhUIjAfuUEpdUYXPO91FQJzW+lDld2qt4zDu55Uf+xUrpP2Bg5z9MdmCv79v/wbervTz9W3LbWiB8XPxr1XjynjRJhr4CuP+NQV4R1V6oaMypdQ7Nr5XW6v9Ffn7a/A+xv3qZcv3ZYK1n1VKqVbAr8BzGPf3B4H56u8XOL4CNmL8vH2WM1fI/3W/P0tLZ/vZXKXHA4BS6hrgKYz7UBOMny05luPsBYZjfO+eBr5QSrW0fOxZ4DeMn5XhGFcQYXYf1lqnASXAGb/zhBBCCFF1MoQWQggh6gGt9XEgCtDAB0CWZbVZcxuf9j+t9XGtdSKwHfhNa71Pa52HMZzre1r941rrYq11DMawwtqqvluBeVrrOMuK7E8xIkEG2+gj4bSByxjLbToC3I4xhHgTmKq1PlHp8zKBN7TWJZZhQRIwXinV2vK1eFhrfVJrvRn4EGP4VCFWa73YktX8OcawFYxBd4jW+hmt9SlLpuoHwLVV+Nyz0lr/qrXea1nNHIMxCBlehc/bB4zEGIR9B2QrI+e18jB6tdZ6mWUF8PdACPCS1roEY3gZoZQKrMLX53rgGa11pmW4+TT//NpZU52viVntYMADmGP5nv4IrLdxnEDghJX3z1FKHQPygbss/Vultf7Ysqq8GGOY1duy8tINY7h6j9Y6zXJfXmOpq3Av8B9gpNY6udL7x/N3JMHVwCKtdazW+hTwBMZjtLK1WuuftdblGIO0scC9WusCrXUmMJu/73+fAtepv2MkbsQkbsRyzkAbf2KtfZ7FVIyhI5a3lQeOt2C8KJFkuR9v0VrnnHGEKtBar9Rab9Nal1uu2vgaYwV2dQUDh00+dph/rtr9RGudaFktXVKFx2QJxuOhRGu9GON+1cWyGvcq4EmtdaHWegfG96fCpcABrfX/Wc6VAMzHuE+cQWs9w8b36mxXk/xc6efnz2epNXMDsNjy2CzXWkcD8cA4pVQbjJ+NFb8DVmG8WFdZ5ft9hexKfT1YxZ/N1Xk83IIxXN9g+f4la60PAmitv9dap1tuy7fAHoxBORjf07ZAmKUPW4+FCicwfuYIIYQQooZkCC2EEELUE1rrnVrrm7TW4RgrmcOAN2x8SkalvxdZ+XflAedRbayErXDQcvzTtQUeqDxUxlhZba22Qr/TBi7LKn3sF8AdSLIyKEjTWlce6FX0FAbk6n8OrA/yzxWoRyr9vRDwUUZkRluMuInK/f8XaF6Fzz0ryyr1dZZL0Y9hrCYNPsunAaC1Xqe1nqS1DsEYko3AWLVc4fTvX7b+e0PEIsvbxpz96xNm+Xflj9n6/kH1viZmtWGc+T39x8rW0xzFWM16upla60DAB2MQ+IOyEgmjjFiJlyyX7B8HDlg+FGz544OxmtLMf4C39ZlxIJVzccMq3watdSF/r9SsUPk2tgU8gcOV7n/zMFbSVqzsLQDOV0p1xVgdutBGj9WmlBqGkTn+jeVdXwGRSqk+ln+3xvbXpTrnOk8ptUIZ0Td5GC86VenxcJpsjJXm1rS0fLzCP+5TVXhM5lhe2KlQiPE4CsF40aTy8U7/Xp532s+S6zFWTdvbFZV+fl5Rw2O0Ba45rd8ojK9fGNZ/B1RmLQc9uFJfr1K1n81Vfjxg476ojBiczZU+ryd/f18fAhSwXhmRQP+ydozT+APHqlAnhBBCCBMyhBZCCCHqIa31LoxYg552OmRTy+XlFdoA1ja/OgQ8f9pQ2U9r/XUNz/s8RhxAS6XUlNM+1koppaz0lA4EKaX8T/tYWhXOdwgje7ly//5a63FV7Pf0Va5/sUSSzMeIeWhuGZQuxhiGVIvWegPwIzX7/p7t65OOMfyp/LGK77Xp7bODw5z5PW1to34rlix0aywrIFcDyRiX8J/uOoxL/y/EuGQ/wvJ+hTG0PAl0sHH+i4HHlFJXVbxDKdUCY2iXYHnXYYzL/Ss+7osRU/OPViv9/RDGlQOVh3dNtNaVIxw+xVi1eiPwg9b6pLXmlFLD1d/Z3db+mK3An2b5GmxWSh0B4izvn1qpR1tfl8oKAL9KPZ0+gP0KY4jeWmsdALxHDR4PwO8YA99/3F+UUoMw7kPLK71bV/r4uTwmszBiIsIrva/y+Q8BMaf9LGmstb7D2sGUUu/Z+F4lVqEfW6w9bk9/3yHg89P6baS1fgnjfmztd0BF76ff781U5WdzdR4PVu+Lysg0/wDjSohmlu/rdizfV631Ea31rVrrMIzoqHeUJTfcGmVk33thI/5HCCGEEGcnQ2ghhBCiHlDGBlgPKKXCLf9ujZFBaisPt7qeVkp5WYZXl2JEPpzuA+B2ywpHpZRqpIzNx6ytWLVJKTUCI6N1quXPXGXkllYIBWYqpTyVkQ3aDeNy8kPAGuBFpZSPZRXsv/ln3q+Z9cBxpdTDytiYzl0p1VMpNbCKbWcA7U0+5gV4YxleKaXGYn04egZl5PveqpQKtfy7K0b+abW/v1X4+nyNMVwNUUoFY0RIfFHp9jVTSgVU97xVsBYoA+5SSnkopS7n78vnrVkPBJ52n/gHpdQQjI0JrQ3x/DEGXDkYg9IXKj5giQL4GHhdGRupuSulhqh/ZpsnYuSqv62UuszyvnHA0kqruX8AJiilhiojo/ppbAw4tdaHMeIgXlNKNVFKuSljw9DKERWfY2Rl3wB8ZuNYq/Xf2d3W/qy28vXywYjZmY6RCV7x527gesuK9Q+BZ5VSnSyP8V5KqYrB+un3/y1AD6VUH8uxnzrtlP4YK2NPWgbG15ndHmVsWrfS5Lb+DvyBkWHcw/L9Goxxn35Xa73H5LA1fkxarjL4EXhKKeVneUxOrVTyC9BZKXWj5WeUp1JqoFKqm8nxbrfxvbKaI10N1n4unf6+LzDuq2MsXz8fZWzMGG6JuIjn798BURgZ9BVOv99bVd2fzVV4PHwIPKiU6m+5L3a0DKAbYQyzs+Cv7Pa/XrBTSl1T8bsS44oKjfGzx+xrNRJYrv8ZxyOEEEKIapIhtBBCCFE/nMDYqC9OKVWAMZzcDjxgp+MfwXiyno4xMLjdstr6H7TW8Ri50G9Z6pOBm85y7C2nrfp7QynVBGPAdpc2MnljgY+A/1Pqr5WycRibymVjrJi+Wv+dTTsFY2VrOsZmd09qI+PUJstgaQLG4G2/5dgfYqyUrYoXMQa4x5RSD5527BPATIxM56MYA7eqRikcwxg6b1NK5QNLMW6Xtc0Pq8LW1+c5jIHTVmAbxurG5yy3YRfGkHqf5TaeLaajyrSRmTwRYyh1DGPI+gvGoNis/hP+uVEfwFsV9yWMge1jWuslVg7xGUYUQBqwgzMH+g9i3P4NQC7GBoT/+L+z1noLxgsyH1gGmP+IJNBG3vrdGNEWhzEep5lmt8liKsZwdAfG/eQHKkVNWOI/EjAGZ2cMks/RFRjRLZ9ZVose0UY2+0cYsTiXAK9j3Id/A45bPuZr+fyngE8t941JWuvdGJti/o6RyXt6pM4M4Bml1AmMFzu+s9Fba+BPGx+/CliB8djIxxiqfoTx9bfqHB+TYKy0DcD4+fg5xmOjuNKxL8bIL0631PwPY+jtbB8B3dU/M6P/8bPKMiC+HCN+KAtjlfF/+Ps+fx3G75hcjI1gK78AYi2Kw0x1fzabPh601t9j/Oz/CuOx9TMQpI187tcwXtjKwNjEtPJ9ZyDG78p8jO/3PVrr/ZaPPUWl+7DlfddjrNIXQgghxDlQZ3nBWgghhBANnFJqJPCFNrKmawWl1E3ALVrrKFf3IhxHKRUHvKe1/j+Tj4dgDGL7aq2LrNU4i2WV8BGggzY297RW0xhjwN6p0tCrJuf6GEjXWj9W02PUNUqpzcAFuoabIDqDUup/QAut9bSzFtcTVbnf12VKqUjgfa31EFf3IoQQQtR1VdpERwghhBBCCEezXGafhLEC/XqgF8bKVqu01llAV+d0d1ZBwOOnD+KUUhMwoiIURvbwNv7eBLHalFIRGCvG+9b0GHWR1rqPq3s4nSWCwwvjezoQYxX/LS5tyvms3u/rC631NkAG0EIIIYQdSByHEEIIIYSoLbpg5AjnYUTJXG3Jha31tNaZWut3rXzocv7eMLMTcO3ZsnPNKKWexYjZeeVcVlILu/HHyIUuwIj0eA1Y4NKOnMzG/V4IIYQQ4h8kjkMIIYQQQgghhBBCCCGEw8hKaCGEEEIIIYQQQgghhBAOI0NoIYQQQgghhBBCCCGEEA5T6zcmDA4O1hEREa5uQwghhBBCCCGEEEIIIYQNGzduzNZah5z+/lo/hI6IiCA+Pt7VbQghhBBCCCGEEEIIIYSwQSl10Nr7JY5DCCGEEEIIIYQQQgghhMPIEFoIIYQQQgghhBBCCCGEw8gQWgghhBBCCCGEEEIIIYTDyBBaCCGEEEIIIYQQQgghhMPIEFoIIYQQQgghhBBCCCGEw8gQWgghhBBCCCGEEEIIIYTDyBBaCCGEEEIIIYQQQgghhMPIEFoIIYQQQgghhBBCCCGEw8gQWgghhBBCCCGEEEIIIYTDyBBaCCGEEEIIIYQQQgghhMPIEFoIIYQQQgghhBBCCCGEw8gQWgghhBBCCCGqIC0tjUOHDtkuKkiBU0ed05AQQgghRB0hQ2ghhBBCCCGEsCEtLY0ZM2bQrl07+vfvT3p6+plFhWkQdyssbA/RUVBa4PxGhRBCCCFqKRlCCyGEEEIIIYQVR44c4d5776VDhw58+OGH3HDDDRQUFHD99ddTVlZmFJ06CpsehkUdYf+n0GYy5O2E+Ltd27wQQgghRC3i4eoGhBBCCCGEEKI2yc7O5uWXX+att97i1KlTTJs2jccff5yIiAhGjhzJtGnTeOm5J3j0an/Y8T8oyYOIG6DX09C4Hfh3gO3PQuhIaD/V1TdHCCGEEMLlZAgthBBCCCGEEMDRo0d57bXXePPNN/9a8fzEE0/QqVOnv2qm3nAdZUnvcXGzF2ALEDYeer8ATXv9faCeT0LmathwBzQbCAHdnH9jhBBCCCFqEYnjEEIIIYQQQjR4v/32G+3ateP5559n3LhxJCYm8vnnn/9jAE3WWvi1Bzf3WEtmgS9XvBVERreP/jmABnBzh6FfgkcjiL0GSgude2OEEEIIIWoZGUILIYQQQgghGrz33nsPPz8/tmzZwrfffku3blZWL29+CEpPwIif8Ri7jmWbCrnxxhspLy8/s9YvDIZ+AXk7JB9aCCGEEA2eDKGFEEIIIYQQDV5CQgIjRoygV69e1gvKy+DoJmh9NYRfTmSvXsyZM4fo6Gheeukl65/T8mLo8Sjs+xj2f+645oUQQgghajkZQgshhBBCCCEatJycHA4ePEj//v3Ni07sgdICCPq75pZbbmHKlCk8/vjjrF692vrnRT4JoSNg/e2Qt9POnQshhBBC1A0yhBZCCCGEEEI0aJs2bQKgX79+5kVHE4y3QX/XKKV47733aN++PVOmTCE7O/vMz3PzgKFfW/KhJ0k+tBBCCCEaJBlCCyGEqFVKSkrOXqStZG8KIYQQNbRx40YA+vbta16UmwDuPtDkn1nRTZo04bvvviMrK4ubbrrpLPnQibBxpj1bF0IIIYSoE2QILYQQolY4cuQIkydPJjAwkMWLF1svKkyFlZfCz+FwdItzGxRCCFFvJSQkEBERQVBQkHlR7kYI7GWsbD5N3759ef311/n111+ZPXu29c9veTH0mAV7P4L9X9ipcyGEEEKIukGG0EIIIVxKa83HH39Mt27d+Pnnn2nZsiVXXnklv/76a+UiSP4Afu0BGSuMf/8xGnI3ua5xIYQQ9UZCQoLtKA5dbsRxBJlnRs+YMYOJEyfyyCOPsG7dOutFkU9DyHDYcDvk7TrHroUQQggh6g4ZQgshhHCZPXv2cMEFF/Dvf/+bXr16sXXrVjZs2EBkZCRXXnklixYtgvwDsPwiWD/dePI/fhtc/Cd4NDYG0Tnxrr4ZQggh6rC8vDySk5NtD6Hz90PJcWhqXqOU4qOPPqJly5b85z//sV7k5gHDvgblCdueOrfGhRBCCCHqEBlCCyGEcLqSkhJeeuklevXqRUJCAvPmzWPFihV06dKFpk2b8vvvv9O3T2+i37mC0oXdIGc9DHwPRv8Ojdsbfy6MAa9AWH4hZK939U0SQghRR23evBmA/v3NVzlb25TQmsDAQGbMmEFsbCx79+61XuTXCiKuh9Sf4dSxavcrhBBCCFEXyRBaCCGEU23YsIGBAwcya9Ysxo0bx44dO5g+fTpubn//Sgp0z+bPZ7yYc2M5f2wrZpnHa9DpNlCVfm01jjAG0d7NYMVFkLXW+TdGCCFEnZeQYAyYbW9KuBHcPCGgx1mPd8MNN6CU4rPPPjMvan8TlBfDwW+r2a0QQgghRN0kQ2ghhBBOUVZWxgMPPMDgwYPJzMzkxx9/ZP78+YSFhf1dpMth5+uwuBceJxIp7PU2T64YyKWTZ/DTTz+dedBGbYxBtE9zWHExZMY67wYJIYSoFzZu3EirVq1o3ry5eVFuAgREgrv3WY8XHh7OhRdeyGeffUZ5ebn1oqD+ENAd9n9aw66FEEIIIeoWGUILIYRwiiVLlvD6669z0003sXPnTq688soziw58DZsegBYXwvgd+PWcwW+/RTNw4EAmTZrE/Pnzz/wcv3C4YKVxefPKSyBzlaNvihBCiHrk7JsSasumhLajOCq76aabOHDgAKtWmfxOUgraTYPstXB8dzU7FkIIIYSoe2QILYQQwinWrl2Lu7s7c+fOJSAgwHpR5krwagojFoCfsUK6SZMmLF26lEGDBjF58mS+//77Mz/PL8wyiG4DK8ZCxgpH3QwhhBD1SEFBAbt27bI9hC5MgeKcag2hr7jiCvz9/fn0UxsrnSNuMGKmZDW0EEIIIRoAGUILIYRwinXr1tG7d2/8/PzMi3LioNlgY4VYJRWD6MGDBzNlyhS+/dZKhqZvC7hwpbFp4crxcOR3+94AIYQQ9c6WLVvQWtseQudaNiVsWvUhtJ+fH5MmTeKHH36goKDApCgMWlwM+z834qiEEEIIIeoxGUILIYRwuLKyMjZs2MB5551nXlRyAo5th2DrNf7+/ixZsoQhQ4Ywbdo0cnJyzizyCYULlkPjDrDmeigvsdMtEEIIUR9VbErYv39/86LcBFDuENirWseeNm0a+fn5/Pjjj+ZF7aZB4SG5gkcIIYQQ9Z4MoYUQQjjczp07OXHiBIMHDzYvyo0HtLES2oS/vz9z586luLiYb775xnqRTwj0eRFOZkL6knNrXAghRL22ceNGQkND/7lJ7umOJhibCHr4VuvYUVFRtG/f3nYkR/jl4BkA+z6p1rGFEEIIIeoaGUILIYRwuLi4OADbK6Gz1xlvgwfZPFafPn3o1auX7Sf1LceAdwjs/6y6rQohhGhAKjYlVKfFQP1Fa8jdWK0ojgpKKaZOncry5cs5dOiQ9SIPX2g7GQ79aFwRJIQQQghRT8kQWgghhMOtW7eOpk2b0qlTJ/Oi7HXQpIuxMeFZTJs2jQ0bNrBz507rBW6eEHE9pC2C4twadi2EEKI+O3nyJImJibbzoIsOw8kMCLIR12HD1KlT0Vrz+eefmxe1mwZlhZDyQ43OIYQQQghRF1RpCK2UukcptV0plaiUutfyviClVLRSao/lbdNK9bOUUslKqSSl1JhK7++vlNpm+dgcZbrkQAghRH0SFxfHoEGDcHMz+bWjtWVTQhsrpSu57rrrcHd357PPbKx0bj8Nyk/BQZPYDiGEEA3atm3bKCsrsz2EPmrZlDCo+iuhAdq1a8eIESP49NNP0VpbLwoeAv6dYL+NK3yEEEIIIeq4sw6hlVI9gVuBQUBv4FKlVCfgEeAPrXUn4A/Lv1FKdQeuBXoAlwDvKKXcLYd7F5gOdLL8ucSut0YIIUStc+LECbZv3247D7rgoLHSLNhGTSUtWrRgzJgxfPHFF5SVlVkvatrH2ERKntQLIYSwosqbEqIgsHeNzzNt2jR27979VzTVGZQyVkNnxkD+/hqfRwghhBCiNqvKSuhuwDqtdaHWuhSIAa4ELgcqntl/Clxh+fvlwDda62Kt9X4gGRiklGoJNNFar9XGMoDPKn2OEEKIeio+Ph6tte086BzLE/MqroQG40l9amoqK1asMC9qNw1y1kPeriofVwghRMOwceNGmjZtStu2bc2LcjcaUVGejWt8nquvvhpfX1/bexm0uxFQspeBEEIIIeqtqgyhtwMjlFLNlFJ+wDigNdBca30YwPI21FLfCqi880aq5X2tLH8//f1nUEpNV0rFK6Xis7KyqnN7hBBC1DLr1hkbDg4aZGPDwex14O4LgZFVPu5ll11GQECA7UiOiOtBuctqaCGEEGc466aEYMRx1DAPukKTJk2YOHEi33zzDSdPnrRe1KgNNB8F+z4FXX5O5xNCCCGEqI3OOoTWWu8E/gdEA0uBLUCpjU+x9r84beP91s75vtZ6gNZ6QEhIyNlaFEIIUYutW7eOTp060axZM/OinDgIGmBsKFhFPj4+TJ48mfnz53PixAnrRb7NoeUlsP9zKDeJ7RBCCNHgnDp1im3bttnOgz6ZCYWp0LRmedCVTZs2jWPHjrFo0SLzovY3QcF+yIo95/MJIYRomEz3HxCiFvCoSpHW+iPgIwCl1AsYq5gzlFIttdaHLVEbmZbyVIyV0hXCgXTL+8OtvF8IIUQ9pbUmLi6Oiy++2Lyo7JSRudnl7moff+rUqbz//vv8+OOPTJs2zXpR+2kQ+ytkLIeWF1X7HEIIIeqfHTt2cOrUKdtD6NxNxtsabkpY2ejRo2nVqhWffPIJ11xzjfWi1hNhwwxjNXToiHM+pxBCiPrlxIkTZGRkkJ+fT0FBwT/eVvy9qKiIcePGMWDAAFe3K8QZqhLHgVIq1PK2DTAR+BpYCFQ8458GLLD8fSFwrVLKWynVDmMDwvWWyI4TSqnByrjmbWqlzxFCCFEPHTx4kIyMDNt50Me2QHkxNKvapoSVDR06lI4dO9rO2Ww1ATwDJWdTCCHEXyo2JbQ5hD660XjbtO85n8/d3Z2pU6eybNkyjhw5Yr3IoxG0uRpSvofSgnM+pxBCiPpDa81HH33El19+yYIFC/j999/ZuHEjqamplJaW0qxZM7p3706zZs1Yt26drIgWtVKVVkID85VSzYAS4E6t9VGl1EvAd0qpfwMpwDUAWutEpdR3wA6M2I47tdYV10DfAXwC+AJLLH+EEELUU3FxxoaDgwfbGDBnG5nRBFd9U8IKSimmTp3KE088wcGDB61vLuXuA20nG5EcJe+Ap3+1zyOEEKJ+2bhxI/7+/nTs2NG8KDcBGncErwC7nHPatGm8+OKLfPnllzzwwAPWi9rdBPs+gUM/Qbsb7HJeIYQQdV96ejp5eXmMHj2aHj160LhxY7y8vM6o27x5MwsWLCAlJcX2xrtCuECVVkJrrYdrrbtrrXtrrf+wvC9Ha32B1rqT5W1upfrntdYdtNZdtNZLKr0/Xmvd0/Kxu7S8NCOEEPXaunXr8PHxoVevXuZFOXHg2wr8ws1rbLjxxhsB+OKLL8yL2k2DskJI+aFG5xBCCFG/JCQk0LdvX9zcbDwdyk2wSxRHhS5dunDeeefx6aefmq9QCx0OjSJkQ10hhBD/kJSUhFKKAQMGEBQUZHUADdC9e3e8vb3/uuJHiNqkSkNoIYQQoibi4uLo378/np42NhzMXlejVdAVIiIiOP/8820/qQ8eDP6d5Em9EEIISktL2bJli+0ojlNHjU0C7TiEBmM19LZt29i8ebP1AuUG7abCkT+g4JBdzy2EEKLu2r17N23atMHX19dmnZeXF5GRkezYsYOioiIndSdE1cgQWgghhEOcOnWKhIQE21EcJ7Mgf2+N8qArmzZtGnv27Pkr/uMMShmroTNjIH//OZ1LCCFE3ZaUlERRUdFZNiW0rCBrat8h9OTJk/Hy8rK9l0G7qYCGAzau8BFCCNFgHDt2jIyMDLp06VKl+n79+lFaWsq2bdsc3JkQ1SNDaCGEEA6xZcsWiouLbW9KmLPeeHsOK6EBrrrqKnx9fc/ypP5GQBnZ0EIIIRqsKm1KWDGEtvNK6KCgIC677DK++uorSkpKrBf5d4CQ4UY2tKQXCiFEg5eUlARA586dq1TfsmVLWrZsSUJCgmxQKGoVGUILIYRwiHXrjA0Hz7opoXKHoP7ndK4mTZowceJEvvnmG4qLi60XNWoDzUfB/s/kSb0QQjRgCQkJ+Pr62l5RdjQBGrUF72Z2P/+0adPIyspi8eLF5kXtp8GJ3ZC91u7nF0IIUbfs3r2b4OBgmjWr+u+kfv36kZGRQXp6ugM7E6J6ZAgthBDCIeLi4mjZsiXh4TY2HMyJg8BI8Gh0zuebOnUqx44dY9GiReZF7aYZ8R9Zf57z+YQQQtRNGzdupE+fPnh4eJgX5SbYPYqjwpgxYwgPD+eJJ54wXw3d5hrwCoJN/wFd7pA+hBBC1H4nT57kwIEDVY7iqNCzZ088PT1lg0JRq8gQWgghhEOsW7eOwYMHo5SyXqDLjSH0OeZBV7jgggto1aqV7UiO1hONgff+z+xyTiGEEHVLeXk5mzZtsh3FUXLcWIVs5yiOCp6ensydO5etW7fyyiuvmBQ1gX6zIXsN7HnXIX0IIYSo/ZKTkykvL6/2ENrHx4cePXqwfft2Tp065aDuhKgeGUILIYSwu+zsbPbu3Ws7D/p4kvFEP9g+Q2h3d3duuOEGlixZQmZmpvUiz8bQ+ipI+RZKZbdoIYRoaJKTk8nPz7c9hD662Xh7jlFRtlxxxRVcffXVPPPMM39lfZ6h3Y3Q4iLY/AgUHHJYL0IIIWqv3bt34+fnR6tWrar9uf369ePUqVMkJiY6oDMhqk+G0EIIIewuLi4OqEIeNECzc9uUsLKpU6dSVlbGV199ZV7Ubpox/E5dYLfzCiGEqBuqtSmhg+I4KsydOxc/Pz9uueUWysutRG4oBYPmGVcObZgh+xkIIUQDU1ZWxp49e+jcuTNubtUf34WHhxMcHCyRHKLWkCG0EEIIu4uLi8PNzY3+/W2sIsuJA89AaFK1XZ6ronv37gwYMIDPPrMRt9F8JPi1gf02YjuEEELUSwkJCXh5edG9e3fzotwE8A0D3+YO7aVFixa89tprxMbGMm/ePOtFjdtB7+cg/Rc4+K1D+xFCCFG7pKSkcPLkyWpHcVRQStGvXz9SU1PNrxQVwolkCC2EEMLu1q1bR2RkJI0bNzYvyl4HzQaBsu+voqlTp7Jp0ya2bdtmvUC5GZc4H/kNCmW3aCGEaEg2btxIr1698PLyMi86utHhq6Ar3HTTTVx44YU8/PDDpKamWi/qPBOCBsLGmVCc45S+hBBCuF5SUhLu7u60b9++xsfo3bs37u7ushpa1AoyhBZCCGFX5eXlrF+/3nYedGkB5G2zWx50ZVOmTMHT09P2auh2U43Lmw98affzCyGEqJ201iQkJNiO4igtgOO7HJoHXZlSinnz5lFWVsaMGTPQ1iI33NzhvA/h1FFIeMApfQkhhHAtrTVJSUm0b9/e9gunZ+Hn50fXrl3ZunUrpaWlduxQiOqTIbQQQgi7SkpKIi8vz3YedE68MQS2Yx50heDgYC699FI+/fRTiopMNh9s0hmCh0LyPCiX/4wJIURDcODAAY4dO3aWTQm3Gr+fgpyzEhqgffv2PPvssyxatIjvvvvOelHTXtD9YSNK6vBvTutNCCGEa2RlZXHs2LEaR3FU1q9fP4qKiti5c6cdOhOi5mQILYQQwq4qNiW0uRI6x7IpYbD9h9AA9957L1lZWXz88cfmRd0egPy9cGi+Q3oQQghRu1RpU8KjlsuVnTiEBpg5cyYDBgzg7rvvJifHJHKj52Pg3xnW32as2BZCCFFvJSUlAdC587nvn9OuXTsCAwMlkkO4nAyhhRBC2NW6deto0qQJXbt2NS/KjoPGHcG7mUN6GD58OEOHDuWVV16hpKTEelH4FdCkKyS+CNYufxZCCFGvJCQk4O7uTmRkpHlR7kbwDgHfVs5rDPDw8OCjjz7i6NGj3H///daL3H3gvA+g4ABsfcKp/QkhhHCu3bt3ExYWhr+//zkfq2KDwgMHDpCbm2uH7oSoGRlCCyGEsKu4uDgGDRqEm5vJrxitjZXQDsiDrqCUYtasWRw8eJBvv/3WpMjNuLT52BY4vMxhvQghhKgdNm7cSI8ePfDx8TEvyk0w8qCVcl5jFr169eLhhx/ms88+Y9kyk99LoSOg4+2Q9AbkbHBqf0IIIZwjPz+f1NRUu0RxVOjTpw9KKVkNLVxKhtBCCCHspqCggK1bt9rOgy5MhaLDDsmDrmzcuHH07NmTl156ifLycutFba8Dv3DY8aJDexFCCOFaFZsS9u9vY8PBspOQl+j0KI7KHnvsMbp06cJtt91Gfn6+9aI+L4FPC4j7N5SbXO0jhBCiztq9ezeAXYfQ/v7+dO7cmS1btlBWVma34wpRHTKEFkIIYTfx8fGUl5efZVPCijxox62EBnBzc+ORRx4hMTGRX375xXqRuxd0fRAyV0HWGof2I4QQwnXS0tLIysqynQd9bDvoUmjquiG0j48PH374IQcPHuSxxx6zXuQVAAPfgWPbYOcrzm1QCCGEw+3evZuAgABCQ0Ptety+ffuSn5/Pnj177HpcIapKhtBCCCHspmJTwkGDBpkXZccZuZaBvRzez+TJk4mIiODFF19Em+U+d7zFyKbe8ZLD+xFCCOEaVdqUMHej8daFK6EBoqKimDFjBnPmzGHDBpPIjfDLoc01sO0ZOLHXuQ0KIYRwmJKSEvbu3UuXLl1Qdo6G6tSpE/7+/hLJIVxGhtBCCCHsZt26dXTo0IGQkBDzopx1xiozdy+H9+Ph4cF//vMf1q1bx6pVq0yKGkHnmZC2yFgFJ4QQot75888/8fDwoHfv3uZFOXHgFQSNIpzWl5kXX3yRpk2b8r///c+8qN8bgDbyoYUQQtQL+/bto7S01K5RHBXc3Nzo06cPycnJ5OXl2f34QpyNDKGFEELYhdaadevWcd55NrKey0uMlWYOjuKo7OabbyY0NJQXX7SR+9z5LmMYvcPGk30hhBB1VnR0NEOGDKFRo0bWC7SGI9HQfLRLNiU8XZMmTbj11lv56aefSElJsV7kFwZtJsO+T+CUDBOEEKI+SEpKwtvbm7Zt2zrk+P369UNrTXx8vEOOL4QtMoQWQghhF6mpqRw+fNh2HvSxrcbGTw7elLAyX19f7r33XpYtW8amTZusF3kHQcfb4ODXkL/fab0JIYRwvKysLDZt2sRFF11kXnQ8ydg4t6WNGiebMWMGAO+88455UZe7oTTfGEQLIYSo07TW7N69m44dO+Lu7u6QcwQGBtKtWzfi4+MpLi52yDmEMCNDaCGEEHZRkQdtcyV0tlHjzJXQYDyRb9KkCS+9ZCP3uet9oNxg52vOa0wIIYTD/fHHHwC2h9BHoo23LWrPELpNmzZceeWVfPDBBxQWFlovajYQgofA7rmgy53boBBCCLtKS0ujoKDAIVEclQ0dOpSTJ0+aL9ARwkFkCC2EEMIu1q1bh7e3N3369DEvyl4HPi3Ar7XT+gIICAhgxowZ/PDDD+a7QfuFQ7upsO8jKMpwan9CCCEcJzo6moCAAAYMGGBedCQaGneAxu2c11gVzJw5k9zcXL766ivzos4zIX8vpC9xXmNCCCHsLikpCaUUHTt2dOh5wsPDadOmDevWraOsrMyh5xKiMhlCCyGEsIu4uDj69u2Ll5eNDQdz1hmroF2Qt3nPPffg6enJK6+8Yl7U7T9QVgy75zivMSGEEA6jtSY6OprRo0fj4eFhvai8BDJW1qpV0BWGDx9O7969mTNnDlpr60VtrgLfMEiS311CCFGX7d69m7Zt2+Lr6+vwcw0dOpS8vDx27Njh8HMJUUGG0EIIIc5ZSUkJ8fHxtvOgi3PgxB6n5kFX1qJFC/71r3/x6aefkp6ebr2oSRdofRXsfhtKjju3QSGEEHa3e/duDh06ZDuKIzsOSk/UqjzoCkopZs6cybZt24iJibFe5OYJne6AI79B3k7nNiiEEMIujh49SmZmpsOjOCp07tyZ4OBg1qxZY/4ipxB2JkNoIYQQ52zLli2cPHnS9hDaRXnQlf3nP/+hrKyM119/3byoxyNQkgd73nNeY0IIIRwiOtrIej5rHrRyg+ajndRV9UyZMoVmzZoxd+5c86KOt4GbN+x+y3mNCSGEsJukpCTAGA47g1KKIUOGcOTIEfbvl43ZhXPIEFoIIcQ5i42NBWDYsGHmRVmxoDyg2SAndXWmdu3aMXnyZObNm0dubq71oqD+xiXZu2ZD2UnnNiiEEMKuoqOjiYiIoEOHDuZFR6IhaCB4BTqtr+rw9fXl1ltv5eeff+bgwYPWi3xCIGIK7P8UTh1zan9CCCHOXVJSEiEhIQQFBTntnL169aJx48asWbPGaecUDZsMoYUQQpyz1atXExERQXh4uHlRVqwx4PXwc15jVjzyyCPk5+fz9ttvmxf1mAUnj8C+T5zWlxBCCPsqKSlhxYoVXHTRRSizvQhO5UHO+lqZB13ZHXfcgVKKd955x7yo891QWgD7/s95jQkhhDhn+fn5HDx4kG7dujn1vB4eHgwaNIi9e/dy5MgRp55bNEwyhBZCCHFOtNbExsYyfPhw86KyYuNJfkiU8xozERkZyaWXXsqbb75JQUGB9aLQkcaK7Z2vQHmpU/sTQghhH+vXr+fEiRO2ozgyVoAuq5V50JW1adOGK6+8kg8++IDCwkLrRUH9jN+zu9+C8jLnNiiEEKLGduzYgdaaHj16OP3cAwYMwNPTk7Vr1zr93KLhkSG0EEKIc5KcnExmZiZRUTYGzLnxUF4MoTYG1U70yCOPkJOTw4cffmi9QCnoPgvy90HKd85tTgghhF1ER0ejlGL0aBtZz0eiwaMRNHPdfgVVNXPmTI4ePcqXX35pXtRlpvG7K32x8xoTQghxThITEwkJCSE0NNTp5/b19aVfv35s376dvLw8p59fNCwyhBZCCHFOKvKgbQ6hs4wagoc6oaOzGzZsGCNHjuTZZ58lKyvLelH4ZRAYCZsfgZITzm1QCCHEOYuOjqZ///40a9bMvOhItHH1i7uX0/qqqaioKPr06cOcOXPQWlsvCr8C/MJh9xyn9iaEEKJmjh8/TkpKiktWQVcYPHgwWmvWrVvnsh5EwyBDaCGEEOckNjaWoKAgunbtal6UGQtNuhobJ9USb7/9NsePH+fee++1XqDcYOA8KEyFLY86tTchhBDnJi8vj7i4ONtRHAUH4cSeWp8HXUEpxcyZM9m+fTsrV660XuTmCZ1mwJHfIW+HU/sTQghRfTt2GD+re/bs6bIeAgMD6dmzJwkJCZw8KRuzC8eRIbQQQohzsnr1aqKionBzM/mVossh+89akQddWffu3Xn00Uf56quvWLzY5LLlkCHQ+S4jXzNLctKEEKKuWLlyJWVlZbaH0IejjbctLnROU3YwZcoUgoODmTPHxkrnDreCmzckzXVeY0IIIWokMTGRFi1a2L5qxwmGDh3KqVOniI+Pd2kfon6TIbQQQogay8jIYM+ePbajOPJ2wKmjEFI78qAre+SRR+jevTu33347J06YRG70ft64tHn9LVB2yrkNCiGEqJHo6Gj8/PwYOtRGDNSR38E3DAK6O6+xc+Tj48P06dNZuHAhBw4cMCkKhojrYf9nxu9fIYQQtdKxY8dITU11aRRHhRYtWtC+fXvi4uIoLZWN2YVjyBBaCCFEjf35559AFfOgQ2vXSmgAb29vPvzwQ1JTU3n0UZPIDU9/GPiuMUzf8ZJzGxRCCFEj0dHRjBgxAm9vb+sFuhwy/jBWQSvl3ObO0R133IFSinfeece8qMvdUFYIez92XmNCCCGqJTExEaBWDKHBWA2dn5/Ptm3bXN2KqKdkCC2EEKLGYmNj8fHxoX///uZFWbHg2xIatXNeY9UwZMgQ7rrrLt566y3WrjWJ3Gg1HtpOgcTnJGNTCCFquZSUFHbv3m07iuPoZijOrjN50JWFh4czceJEPvjgAwoKCqwXNe0DoSOMOKnyMqf2J4QQomoSExMJCwujadOmrm4FgPbt29O8eXPWrl1rvgGuEOdAhtBCCCFqbPXq1Zx33nl4eXmZF2WuNvKga/FKs+eff57w8HBuueUWiouLrRf1fwM8/CHuVmMFnRBCiFopOtrIerY5hD5S9/KgK5s5cybHjh3jyy+/NC/qPBMKDkD6L07rSwghRNXk5uZy+PDhWrMKGowNcIcOHUpWVhZ79uxxdTuiHpIhtBBCiBrJz89n06ZNtqM4ClKgMKVW5kFX5u/vz3vvvceOHTt46SWTyA2fUOg3G7LXwJ73nNugEEKIKvv9999p0aIFPXv2NC86HA2BkeDbwnmN2dGwYcPo27cvc+bMMV+tFn45+LWGJBubGAohhHCJ2hbFUaFHjx40adKENWvWuLoVUQ9VaQitlLpPKZWolNqulPpaKeWjlHpKKZWmlNps+TOuUv0spVSyUipJKTWm0vv7K6W2WT42R6lavCxOCCGETXFxcZSVlZ0lD9rIjCak9uVBn27cuHFcd911PP/88+zYYRK50e5G49LtzY9AwSHnNiiEEOKsysvL+f3337nwwgsxfapRWmRERdXBKI4KSilmzpxJYmIiv/xistLZzQM63wkZy//+fSyEEKJWSExMJDw8nICAAFe38g/u7u4MHjyYgwcPkp6e7up2RD1z1iG0UqoVMBMYoLXuCbgD11o+PFtr3cfyZ7Glvrvl4z2AS4B3lFLulvp3gelAJ8ufS+x5Y4QQQjhPbGwsSimGDBliXpQVa0RYBEY6r7Fz8MYbb9CkSRNuueUWysqsZGgqBYPmgS6DDTNAstKEEKJW2bJlC9nZ2bajOLJWQ3lxnR5CA1x33XV069aNu+++2zwbutOd4NcG1t8GZaec26AQQgirsrOzycjIqHWroCv07dsXT09P4uPjXd2KqGeqGsfhAfgqpTwAP8DWyyGXA99orYu11vuBZGCQUqol0ERrvVYb14x9BlxR89aFEEK4UmxsLL169bL96n3WaggeYqzGqgNCQkKYPXs2a9eu5d1337Ve1Lgd9HrWyNhM+d65DQohhLCpIg/6wgttZD0fiQY3L2PjvjrMy8uL999/n4MHD/Lkk09aL/JsDAPfhrxE2PWqcxsUQghhVW2N4qjg4+NDjx492L59u/l+OULUwFmH0FrrNOBVIAU4DORprX+zfPgupdRWpdTHSqmK7TxbAZWvUU61vK+V5e+nv/8MSqnpSql4pVR8VlZWtW6QEEIIxystLWXt2rUMH24j6/nUUTi2HUJrdx706W644QbGjBnDrFmzSElJsV7UZSYEDYCNd0NxrnMbFEIIYSo6OpoePXoQFhZmXnQ4GkKGgYef8xpzkKioKKZPn87s2bNJSEiwXtTqUmh9NWx7Bk4kO7dBIYQQZ0hMTKRt27b4+/u7uhVTAwYMoKSkhK1bt7q6FVGPVCWOoynG6uZ2QBjQSCl1A0a0RgegD8Zw+rWKT7FyGG3j/We+U+v3tdYDtNYDQkJCztaiEEIIJ9u8eTMFBQVnyYNeC+g6kQddmVKK9957j/Lycu644w7rGz65ecB5H0JxDmx60PlNCiGEOENRURGrV6+2HcVRlAHHttT5KI7KXnrpJUJCQpg+fTqlpaXWi/q/Ce7esP52iZISQggXyszMJCsrq9augq4QFhZGixYt2Lhxo/kGuEJUU1XiOC4E9muts7TWJcCPwFCtdYbWukxrXQ58AAyy1KcCrSt9fjhGfEeq5e+nv18IIUQdExsbC8CwYcPMi7JiQXlAs0HmNbVUREQEzz//PIsXL+brr7+2XtS0N3R7CPb9Hxz5w7kNCiGEOENsbCzFxcW2h9AZlp/X9WgI3bRpU9588002btzIW2+9Zb3ILwx6v2jc/gNfOLdBIYQQf9m+fTtKKbp16+bqVmxSStG/f38yMjJIS0tzdTuinqjKEDoFGKyU8lPGFtMXADstGc8VrgS2W/6+ELhWKeWtlGqHsQHheq31YeCEUmqw5ThTgQV2uyVCCCGcJjY2loiICMLDw82LslZDUP86e7nz3XffzcCBA7nvvvs4evSo9aKej0PjjrDhDig76dwGhRBC/EN0dDSenp6cf/755kVHosErCJr2dV5jTjBp0iTGjh3LY489Zh4l1el2aDYYEu6Hk9nObVAIIQRaa3bs2EFERASNGzd2dTtnFRkZiaenJxs3bnR1K6KeqEomdBzwA5AAbLN8zvvAy0qpbUqprcAo4D5LfSLwHbADWArcqbUusxzuDuBDjM0K9wJL7HprhBBCOJzWmtjYWNt50GUnIWd9ncuDrszd3Z158+aRnZ3Nf//7X+tFHr4w8B04sQd2/M+5DQohhPiH6Ohohg4dSqNGjawXaG3kQbe4ANzcnducgymleOedd9Bac9ddd1m/dFq5wXnvw6ljsPkhp/cohBANXUZGBjk5ObU+iqOCt7c3kZGRbN++nZMnZcGNOHdVWQmN1vpJrXVXrXVPrfWNWutiy9tIrXUvrfVllpXOFfXPa607aK27aK2XVHp/vOUYHbTWd2kJlhFCiDonOTmZjIwM23nQuRuh/FSdy4M+Xd++fZk5cybz5s1j3bp11otaXgRtp0DiC3B8t3MbFEIIARgZm5s3b7YdxXF8FxSl1asojsoiIiJ45plnWLRoET/++KP1osBI6PagESWVsdKp/QkhRENXV6I4Kuvfvz+lpaWyQaGwiyoNoYUQQogKFXnQtjclNGoIHuqEjhzrmWeeISwsjNtvv918w6d+r4O7L2yYIRs+CSGEC/zxh5H1bHMIfSTaeFtPh9AA99xzD3379uXuu+8mLy/PelHPJ6Bxe1h/m0RJCSGEk2itSUxMpH379vj51Z24wrCwMFq2bCkbFAq7kCG0EEKIaomNjSUoKIiuXbuaF2WuhiZdwSfEeY05iL+/P2+++SZbtmxh7ty51ot8W0Afy4ZPB002MhRCCOEw0dHRNG3alP79+5sXHY42cvwbRzitL2fz8PDg/fffJyMjg1mzZpkU+cLAd+HEbkh80bkNCiFEA5Wens6xY8fqTBRHZf379yczM5PU1FRXtyLqOBlCCyGEqJbY2FiioqJwczP5FaLLIetPCKm7edCnmzhxIuPHj+fxxx/n0KFD1os6TIdmgyDhPjhlspGhEEIIu9NaEx0dzejRo3F3N8l6Li+BzJVGhFI9N2DAAO6++27ee+891q5da72o5cXQ9jrY8SLk7XRug0II0QAlJibi5uZmeyFPLRUZGYmXl5dsUCjOmQyhhRBCVFlmZia7d++2HcWRtwNKjtX5POjKlFLMnTuX8vJy7r33XutFbu4w8D0ozoYtjzq1PyGEaMiSkpJITU21HcWRvQ5K8+t1FEdlzz77LK1atWL69OmUlJRYL+o/GzwaG7Ecuty5DQohRANSEcXRsWNHfH19Xd1OtXl5eREZGUliYiJFRUWubkfUYTKEFkIIUWXVyoMOrT9DaIB27drxxBNP8OOPP/LLL79YLwrqC51nwp73IDvOuQ0KIUQDtXz5cgAuvPBC86KM5aDcoPkoJ3XlWv7+/rzzzjts376dV1991XqRTyj0fQWyVsPej53boBBCNCCpqakcP36c7t27u7qVGhswYAClpaVs2bLF1a2IOkyG0EIIIaosNjYWHx8f25mbmavBtyU0aue8xpzk/vvvp3v37tx1110UFhZaL+r1DPiGGSvLyk02MhRCCGE3q1atolWrVrRv3968KDMGAvuAV6Cz2nK5CRMmcNVVV/HMM89w8OBB60Xt/wWhI2DzQxIlJYQQDpKYmIi7u3udjOKo0KJFC1q1aiUbFIpzIkNoIYQQVRYbG8t5552Hl5eXeVFWrJEHrZTzGnMSLy8v3n33XQ4ePMgzzzxjvcjTHwbMgWNbYLfJRoZCCCHsQmtNTEwM559/Psrs905ZMWSvhdDzndtcLTB79my01jz99NPWC5SC/nPh1DHY8YpTexNCiIYiOTmZdu3a4e3t7epWzkn//v3Jzs423yNHiLOQIbQQQogqKSgoICEhwXYUR0EKFKbUqzzo040YMYKbb76Z1157je3bt1svCr8SwsbD1sehQP6TJoQQjpKcnMyRI0cYMWKEeVFuPJSdNFb8NjCtW7dmxowZfPrpp+zatct6UdNe0PZaSHoTio44t0EhhKjnjh49Sk5ODh06dHB1K+esR48eeHt7ywaFosZkCC2EEKJK1q1bR1lZ2VnyoP803tbjITTAyy+/TJMmTbjjjjsoL7eymZNSMGCusdFTwr1O708IIRqKmJgYAM4/38Yq50yjhtDhTuio9pk1axZ+fn488cQT5kWRT0N5MSQ+77zGhBCiAdi7dy8AHTt2dHEn567yBoWm0YRC2CBDaCGEEFUSGxuLUoohQ4aYF2WtBg9/COzlvMZcIDg4mFdeeYXY2Fg++eQT60WN20HPJ+DQj5BmspGhEHVUUVER27dv5+eff+bVV1/ltttu4+KLL2bBggV/F5WXwPHdxv1/52tGTvqKS2TTTmFXMTExhIaG0qVLF/OijBgIjATvZs5rrBYJCQnhvvvu4/vvvychIcF6UZNO0OHfkDwP8g84tT8hhKjPkpOTCQwMpFmz+vE7aMCAAZSVlckGhaJGVG0PFB8wYICOj493dRtCCNHgXXTRRWRlZbF582bzosW9jE35Ri11Wl+uUl5ezsiRI0lMTCQpKYng4GArRSWwpC+U5sP4RPBo5PxGhbCD5cuX880337Bnzx6Sk5NJTU39x8eDg4MZ0r6IKwZ5c/PVw1DHkyB/H+hKm3N6B0NpITQfCSN/de4NEPVW27ZtGTRoEN9//731gvJS+CEQ2t0EA99yZmu1Sl5eHu3bt+e8885j8eLF1osKU2FhR4iYAoP/z7kNCuFAxcXFbNy4kd27dzNhwoR6MwwUtV9ZWRkvv/wykZGRXHrppa5ux24++ugjioqKuPPOO833YxANmlJqo9Z6wOnvl5XQQgghzqq0tJS1a9cyfLiNS5lPHYVj2+t9FEcFNzc33n33XY4fP86zzz5rUuQJA9+FgoOQNMe5DQphJxkZGVx22WV8//33FBcXM3r0aJ555hm++uorNmzYQG5uLll71/DzzCKu659LYcZ2Y9Vp94dg8Cdw8Vq4KgeuyoJuD0D6Esjf7+qbJeqBAwcOkJKSYjuKIzcBSgugecPblLCygIAAHnnkEZYsWcLq1autF/mFQ+c7Yf9nkLfTuQ0K4QAFBQUsX76cN954g+joaFJSUli6tP4vlBC1x6FDhzh16lS9iOKorH///uTk5HDw4EFXtyLqGA9XNyCEEKL227JlCwUFBWfJg14L6AYzhAZjc44bb7yRDz74gP/+9780b978zKLQ4dByLOx6HbrMlNXQos554YUXOHnyJJs2baJTp07Wi/68E+XhQ69ZHgyIGsxXX31lva7DrUbmbPL70OdFxzUtGoRq5UGHNLxNCU935513Mnv2bP773/+yatUq66vXuj9iPD63Pg7Df3B+k0LYwdGjR1mzZg2bN2+mtLSUbt26MWzYMFJSUvjtt9/Ys2eP+e8zIexo7969uLm50a5dO1e3Ylc9evRg6dKlbNy4kYiICFe3I+oQWQkthBDirCpWTQ0bNsy8KGu1sfK32SAndVU7PPLIIxQXFzN79mzzop6PQXG28cReiDrkwIEDvPvuu/z73/82f8J+dAsc/BrV9V7GTbyZH374gczMTOu1jVpD2KWw9yMoO+W4xkWDsGrVKoKCgujRo4d5UWYMNOkCvlZeJGxg/Pz8ePzxx4mNjTVfDeoTAl0fgEPzIUciEUXdcuTIEebPn8/cuXPZtGkTkZGR3HnnnUyaNIlWrVoxaNAgmjVrxrJlyygrK3N1u6IBSE5OpnXr1nh7e7u6Fbvy9PSkd+/e7Ny5k9zcXFe3I+oQGUILIYQ4q9jYWCIiIggPDzcvyoqFpv3Bw895jdUCnTt3ZvLkybz99tvm/wkLGQrNR8HOV6DspHMbFOIcPPnkk7i7u/PEE0+YF215DDwDodt/uO222ygpKTHfsBOg0x1QnGVs2inEOYiJiWH48OG4uZk8pSkvM343hTbsKI7K/v3vf9OuXTseffRRysvLrRd1ux+8gmDrY85tTogaKikp4ZtvvmHevHns3r2bIUOGcM8993DZZZf9Y88Od3d3xowZQ05ODnFxskmucKz8/HyOHDlChw4dXN2KQwwdOhQPDw9+/PFHeVFHVJkMoYUQQtiktSY2NtZ2HnRZMeRsgNCGE8VR2X//+1/y8/OZM8dG7nOPx6DoMOyTzZ5E3bB9+3Y+//xz7r77blq1amW9KOtPSP8Fuj8MXoF069aN888/n3nz5pkPuFpeDI3aQfJ7jmte1HtpaWns3bvXdhTHsa1QkidD6Eq8vLx4+umn2bRpE/Pnz7de5NkEesyCw8sgI8a5DQpRAxs2bCApKYnzzz+f++67j4suugh/f3+rtZ06daJTp07ExMSQn5/v5E5FQ7J3716AepcHXSEgIIBLL72UtLQ0Vq1a5ep2RB0hQ2ghhBA27d69m4yMDNtD6Jz1UF7coPKgK+vZsydXXnklb775JsePH7de1HwUBA+BxJegvMS5DQpRA4899hj+/v48/PDD1gu0hi3/BZ8W0OXuv959++23s2/fPn7//Xfrn6fcoNNtRkxC3g4HdC4agoonvCNG2Mh6rsiDDpU86Mquu+46evToweOPP05paan1ok53gm+Y8RjX2rkNClENxcXF/Pnnn7Rv356RI0fi4+Nz1s8ZM2YMpaWl/PHHH07oUDRUe/fupVGjRrRo0cLVrThMz5496d27N6tXr+bQoUOubkfUATKEFkIIYVOVNn7KWAkoCLExqK7nHn30UY4dO8bbb79tvUApYzV0YQrs/8K5zQlRTevWrWPBggU89NBDNGvWzHrR4d8gc5WReV5pw80rr7ySkJAQ3nvPxkrn9jcbGfJ75tm5c9FQxMTE4O/vT58+fcyLMldB4/bgZyNKqgFyd3fnueeeIykpic8//9x6kYcv9HwCstdA+mLnNihENcTFxVFYWMjo0aOr/DnNmjXjvPPOY/PmzaSnpzuwO9FQaa3Zu3cvHTp0sL4JbD0yduxYAgIC+PHHHykuLnZ1O6KWkyG0EEIIm2JiYmjRooXtXcQzYyCwF3gHOa+xWqZ///6MHTuW119/nYKCAutFYWOhaV9IfMHIKhWiFtJaM2vWLEJDQ7nnnnvMiowVko0ioMOt//iQt7c3//rXv1i4cCFpaWnWP98nFFpfDfs/hVKTx4sQNqxatYqoqCjc3d2tF+hyyFolURwmLr/8cgYOHMhTTz1lPjTo8C9jiL/lUePrKUQtU1RUxJo1a+jSpYt5bJSJESNG0KhRI5YuXYqW1f7Czg4fPkxhYWG9zYOuzNvbm4kTJ5KXl8fixfKipbBNhtBCCCFMaa2JiYnh/PPPN38Vv+yUsVJKnujz2GOPkZ2dzfvvv2+9QClj1Wh+MqR869zmhKii6OhoVq5cyeOPP07jxo2tFx2aD0cTIPJpcPc648PTp0+nrKyMjz76yPxEne4w8noPymNBVE9mZiY7d+60fYVO3g4ozpHfTSaUUrzwwgukpKQwb57JFQlunhD5DBzbAinfO7dBIapgzZo1FBcXM2rUqGp/ro+PDxdccAGHDh1i27ZtDuhONGTJyckADWIIDdC6dWtGjBjB1q1b2b59u6vbEbWYDKGFEEKY2rdvH2lpabaf6OfGQ1kRNJcn+kOHDmXUqFG88sornDx50npR+BUQ0AMSn5eVZaLWKS8vZ9asWURERDB9+nSTolLY+jgEdIeI662WtG/fnjFjxvDBBx+YZ86GRBmPhT3v2ql70VBU5EHb/N0kedBndcEFFzBq1Cief/558w3a2l4LAT2Nx7zsZyBqkfz8fOLi4ujZsyfNmzev0TH69OlDy5Yt+f333zl16pSdOxQN2d69e2nZsiWNGjU6e3E9MWLECMLDw/nll1/Iy8tzdTuilpIhtBBCCFMVedBV2vgpRJ7og7Ea+vDhw/zf//2f9QLlBj0eNVbppf7s1N6EOJv58+eTkJDAM888g5fXmSucAdj/ORzfBb2eAzeTKASMDQpTU1PNL81UCjrebryQlRNvh+5FQ7Fq1Sr8/Pzo37+/eVHmKvBrbUTGCKuUUjz//PNkZmYyZ84c60Vu7tD7eTixB/Z96twGhbAhNjaW0tJSRo4cWeNjKKUYO3YsJ06cIDY21n7NiQbt5MmTHDp0iI4dO7q6Fadyc3Nj4sSJaK356aefKC+XxTbiTDKEFkIIYSomJobg4GC6d+9uXpQZY6xm9Al2XmO12KhRoxgyZAgvvfQSJSUmq8baTAL/TrD9OSNbV4haoLS0lMcee4wePXpw3XXXWS8qK4ZtT0HQAGNVvw2XXnopYWFhtjcobHcjuPtBso0aIU4TExPD0KFD8fT0tF6gtfG7KfR848UOYWrIkCFMmDCBl19+maNHj1ovajUBmp0H25+GMpOrfIRwory8POLj4+ndu7f55rlV1Lp1ayIjI1mzZo35Y0CIati/fz9a6wYTxVFZ06ZNGTt2LAcPHmTNmjWubkfUQjKEFkIIYSomJoYRI0aY50GXl0LWn5K5WYlSiscee4yUlBS++OIL60Vu7tB9FhzdBOlLnNugECY++eQTdu/ezQsvvGC+2Vvy+1CYAr1fOOtwz8PDg1tvvZWlS5eyf/9+60VeARBxHRz4Gk4dO7cbIBqE3Nxctm3bZjuK48RuOJkhURxV9Nxzz3H8+HFeffVV6wVKGauhC1Nhj0l+tBBOtGrVKrTWtn8OVMOFF16Im5sb0dHRdjmeaNiSk5Px9vYmPDzc1a24RO/evenRowcrVqwgPT3d1e2IWkaG0EIIIaw6ePAgBw8ePEsedAKU5ssQ+jRjx46lX79+vPDCC5SVlVkvancD+LWB7c/KamjhckVFRTz11FN/rYq0qiQfEp+D5qOgxYVVOu4tt9yCUsp8s06ATrdDWaER8yHEWcTGxqK1PktMlJEZLb+bqqZXr15MnjyZN998k8zMTOtFLS4wHvs7XoDSAuc2KEQlubm5bN68mf79+xMYGGiXYzZp0oSoqCh27txp/qKpEFWgtWbv3r20a9fO/AX9ek4pxfjx42ncuDE//vij5K2Lf5AhtBBCCKsq8qBl46fqq1gNnZyczHfffWe9yM0TejwCOesgY4VzGxTiNG+//TZpaWm8+OKL5lc+7J4DJzOrtAq6Qnh4OBMmTOCjjz4yfxIS1B+CBhobFMoLMuIsYmJi8Pb2ZtCgQeZFmTHg08KIPRJV8tRTT1FUVMSLL75oXtTreeNnQJJJfrQQThATE4ObmxvDhw+363GHDBlCYGAgS5culSxbUWPZ2dnk5eU1uDzo0/n6+nLllVeSk5PDsmXLXN2OqEVkCC2EEMKqmJgYmjZtSmRkpHlR5kpo0gV8Wzitr7ri8ssvp0ePHjz//PPmT2ba3wy+LY3V0EK4SF5eHi+++CKXXHKJ+YtOxbmw42VodRkED67W8W+//XaysrL46aefzIs63Q7Hd0LW6modWzQ8q1atYvDgwfj4+Fgv+CsPeoTkQVdDly5dmDZtGu+++y6pqanWi0KGQNh442eBxOcIF8jMzGTr1q0MGjQIf39/ux7b09OTUaNGkZmZSUpKil2PLRqOvXv3AjTIPOjTRUREMGzYMBISEtixY4er2xG1hIerGxBCwNatW3nllVdwd3fH19f3rz8+Pj5//T3cL4MhzZMIbR4G7t7g5m1561Xp794QPBSayMofce5iYmIYPnw4bm4mr1eWl0FWLLS91rmN1RFubm48+uijXHfddfz8889MnDjxzCJ3H+j2ECTcB5mxEBrl/EZFg/fGG2+Qm5vLCy+8YF608xUoOQ69n6v28S+++GIiIiJ47733mDx5svWittdCwgPGami5skKYOH78OAkJCTz66KPmRQX7jexiieKotieeeIIvvviC5557znxD0V7PwtJ+sOt16PWMcxsUDd7KlSvx8vJi2LBhDjl+ly5dcHd3Z/fu3URERDjkHKJ+S05OJjg42G5RMXXdqFGjOHDgAAsXLiQsLEy+LkKG0EK42qlTp5g8eTKpqakEBQVRVFT015+KLNnGPrDrFWiSBTrfH1VeDOUmlzUH9IDx2514C0R9lJaWxt69e5kxY4Z50bHNxlAqdKSz2qpzJk2axJNPPslzzz3HlVdeaT3moOOtkPgCJD4PobJJoXC+X375hfPPP5++ffuaFx34ClpNgEAbV0aYcHNz47bbbmPWrFns2rWLrl27nlnk4Qftp8Ged4zL/X1Cq30eUf/9+eeflJeXnyUmSvKgayoiIoLp06czb948HnroIdq3b39mUVBfaH017JoNne8GnxDnNyoapPT0dHbu3MmIESPw8/NzyDm8vb2JiIhg9+7dXHzxxQ45h6i/SkpKOHjwIP3793d1K7WGu7s7V111FfPmzWP+/PncdNNNDTYrWxgkjkMIF3vjjTfYtWsX3377LQcPHiQzM5MTJ05QWlrKqVOnyMvL40j0HbQKgpHPwa9+X8G1xTClHCYXwzXHYWIWXJEGvV+EvEQ4lujqmyXquCrlQWdU5EHLE30z7u7uzJo1i02bNrF06VLrRR6NoOv9cHgp5G50boOiwSsuLmbLli2cd9555kUnM6Ew5ZxWKN988814enoyb94886KOt0F5Cez9uMbnEfXbqlWr8PDwYPBgG5EwmTHgHQwB3Z3XWD3y6KOP4unpyVNPPWVe1OsZYzPRHf9zWl9CrFixAh8fH4YMGeLQ83Tu3JmcnBxycnIceh5R/xw8eJDS0lLn50GfLIT0vbA9FlZ9D4vehZx05/ZgQ9OmTZkwYQKpqamsWCH74DR0MoQWwoUOHTrE008/zRVXXMG4cePO+LinpydNdCqNDn1AWcRNJGUH/r3JmVLg7gWe/uATDH5h0P4mUG6QYrIRmhBVFBMTQ5MmTejTp495UWYMNO5o3PeEqeuvv55WrVrx6quvmhd1ugM8/GHna85rTAhg27ZtlJSUMHDgQPOinA3G22Y2as6iefPmTJw4kU8++YSioiLrRQHdjCsrkueBlk2hxJliYmIYOHAgjRo1Mi/KiIGQ4ZIHXUMtW7bkrrvu4osvvjDP8AzoBhE3wJ63obD2DDpE/ZWSkkJycjLDhg0zz4O3k86dOwOQlJTk0POI+ic5ORkPDw/atm3ruJMs+z948Tr4zyj4V1e4IgAuawQ3dYT7h8Nzk2DuDHjtX7Vqs+cePXrQr18//vzzz79ys0XDJENoIVzo/vvvR2vN7NmzrRdoDfF3g0dj3Pu9zJVXXsmCBQsoLi62Xu/bwliVmvJdrfqlI+qemJgYoqKizC+X0uXGBmLNZRX02Xh5eXHPPfewfPlyNm3aZFIUYMRypHwHBbIZjnCeDRuMAbPNIXRuPKCgqY24jiq4/fbbOXbsGPPnzzcv6nQ7FByADFkpI/6psLCQDRs22L5Cp+CQkQktV+ick4cffpjGjRvzxBNPmBdFPgnlpZBY/Zx4IaprxYoVNGrUiEGDBjn8XIGBgYSGhrJ7926Hn0vUL3v37qVt27Z4eno65gRHM+GN6bB5OZSWQLtIuGga3PwCPPh/8MJSeG+L8e/4ZRD3q2P6qKFLLrmEkJAQfvrpJ/Lz813djnARGUIL4SK//fYbP/zwA48++qj5xhcp30PGcmMjKJ8QJk2axPHjx/ntt9/MD9xmEhzfBXmSCy1q5siRIyQlJdl+on9sG5w6Kk/0q+jWW2+lcePGvPaajZXOXe4x3ia96ZymhMAYQoeEhNCmTRvzopwNxspHT/9zOtf5559Ps2bNWL58uXlR2HhQ7pCx8pzOJeqftWvXUlpayogRNmJhKvKg5QXSc9KsWTPuu+8+5s+fT0JCgvWixu2hwy2Q/AHk73dug6JBKSgo4MCBAwwaNAgvLy+nnLNz586kpKSYX7kjxGny8vLIzs6mQ4cOjjvJH59DWSm8vBxmx8Lj38Odc2DKLLj4JhgwBtr3gmsehNZd4b374JTJ4jUX8PT05Oqrr6a4uJiffvoJLYvmGiQZQgvhAsXFxdx999106tSJBx980HpRST4k3A9N+0DH2wG44IILaNq06d+RHNa0nmhEchyUSA5RM6tWGU/ibW/8JHnQ1REYGMitt97KN998w6FDh6wXNWoDbSYbT+hP5Tm3QdFgxcfHM2DAAOubZoJxVU3uBgiqeRRHBaUUw4YNIzY21rzIs7Gx4jrLRo1okGJiYnBzc2PYsGHmRZkx4BkAAdXfQFP80/3330/Tpk15/PHHzYt6PgZuHrDtaec1Jhqcw4cPA9C6dWunnbNLly5orUlOTnbaOUXdVnFfcVgetNaw7GPoNhjanmXPAw9PuOMNSE+Gn+c4pp8aCg0N5ZJLLmHfvn38+eefrm5HuIAMoYVwgddee43du3czd+5cvL29rRclPgdFaTDgbXAzIhE8PT2ZOHEiCxYs4OTJk9Y/zycUmo+GlG8lkkPUSExMDI0aNaJfv37mRZkx0CjCGJyKKrnnHmOl85tv2ljp3O0BKD0Bez9wUleiISsoKCAxMdF2FEfhIWNjwqABdjlnVFQUe/bsISMjw7woJApy4qDslF3OKeqHVatW0a9fP5o0aWJelGnJg3YziZISVRYQEMDDDz/M4sWLWbNmjfUiv1bQaQYc+Bzydjm3QdFgpKWlARAW5rw9SMLCwvDz85NIDlFle/fupUmTJgQHBzvmBLvWw8EdMOZfVasfMAYGT4CvnoXcI47pqYb69etH9+7dWb58ufniHFFvyRBaCCc7ePAgzz33HFdddRVjxoyxXnQ8CXa9Du2mQcjQf3zommuu4cSJEyxbtsz8JG0mwYk9cGyLHTsXDUVMTAzDhg0zzzPT5cYlz7IKulratm3LNddcw/vvv09enslK56B+0HwU7HpDBnDC4TZt2kR5eXkV8qA5p00JK4uKigKwvfolJArKiuCoSQyAaHBOnjzJunXrbEdxFB2BE7slisOO7rrrLpo3b86jjz5qftl094fB3Re2Penc5kSDkZ6eTnBwsPnCHQdwc3Ojc+fO7Nmzh7KyMqedV9RNZWVl7Nu3j44dO5pfWXauln0M3n5w/uSqf85tr8Gpk/DxLMf0VENKKSZMmEBAQADz5883X1wn6iUZQgvhZPfddx9KqbNvRujuB33+d8aHR48eTVBQkO1IjvArjUxNieQQ1ZSdnU1iYqLtKI68HVCcLUPoGnjggQc4ceIEH374oXlR1weNqyBS5PErHKtiU8IBA2yscs7ZAMoDmva2yzn79++Pj4+P7UiOEEvcgkRyCIv169dTXFx8lpgoSx50iI1BtaiWRo0a8eijj7Jy5Ur++OMP60U+odDlXuN31tHNzmxPNABaa9LT0526CrpC586dKS4uJiVFNowWtqWlpVFcXOy4KI6iAlj5NYy4BhrZuBrodK06wcT74LdPIGmDY3qrIR8fH6666ipOnDjBwoULJR+6AZEhtBBOtGTJEn766SeeeOIJ81yzQz/CkWjo9Qz4Nj/jwxWRHAsXLjTfLMMnGJpfYDwhkB/oohqqlQctq82qbcCAAYwcOZI33niDkpIS60Vhl0BAd9j5qjx+hUNt2LCB8PBwWrRoYV6UswECI8Hdxy7n9PLy4rzzzrM9hPZtAY07yhBa/GXVqlUopRg+fLh5UWYMeDQ2rigRdjN9+nTatGljezV0twfBMxC22MiPFqIGTpw4QX5+vkuG0B06dMDd3V0iOcRZ7dmzB6UU7dq1c8wJYudD4YmqR3FUNuVRaNoc3pkJ5eX27+0chIeHM3r0aHbu3MnGjRtd3Y5wkioNoZVS9ymlEpVS25VSXyulfJRSQUqpaKXUHsvbppXqZymlkpVSSUqpMZXe318ptc3ysTnKYdcqCFH7nDx5krvvvpsuXbpw3333WS8qLYSE+yCwl5GxZ2LSpEnk5+fbjuRoOxny98rlzKJaYmJi8PX1tX15fmYM+IVDIwf9R6uee+CBB0hNTeX777+3XqDcoOv9RpxOxnLnNicalPj4eNuPda2NOA47RXFUiIqKIiEhgYKCAvOi0OHGEFpeiBEYv5siIyNp2rSpeVHmKmMVvZuH8xprALy9vXniiSdYv349ixYtsl7kFQjd/wPpv0D2Oqf2J+q3ijzoVq1aOf3cXl5eREREyBBanFVSUhIRERH4+NjnBfszLPsYwjpCpI0XYs00agL/fgl2roPlX9q/t3M0dOhQOnTowJIlS+SqgwbirENopVQrYCYwQGvdE3AHrgUeAf7QWncC/rD8G6VUd8vHewCXAO8opSp2B3kXmA50svy5xK63Roha7JVXXmHv3r28/fbbeHl5WS9KfMHYBGrAWzafRI0aNYpmzZqdJZLjCuMSaonkENUQExPDkCFDzO+jWhtD6NCRIK8j1si4cePo2rUrr776qvmqsojrwae5sRpaCAc4duwYe/bssT2EPpEMJXkQZP8hdFlZGXFxceZFIVFQnGPskSAatJKSEtasWWP7Cp2T2ZC3XWKiHGTq1Kl07NiR6dOn89RTT7F3794zizrPNKI5Vl0Bmx+RjQqFXaSnp+Pm5kbz5mdeHeoMnTt3Jjc3l+zsbJecX9R+ubm5ZGVl0aVLF8ecIC0ZtsYYq6Br+tzrwqnQZSB8+DAU5du3v3OklOKqq64iMDCQb7/9lmPHjrm6JeFgVY3j8AB8lVIegB+QDlwOfGr5+KfAFZa/Xw58o7Uu1lrvB5KBQUqplkATrfVabTzr/qzS5whRr+3fv58XXniByZMnc8EFF1gvOr4Hdr4CETcYK8Bs8PDw4KqrrrIdyeEdBC0ukkgOUWVHjx5l69attp/oH0+Ck5nyRP8cuLm58cADD7Bp0yZWrFhhvcjdBzrfDYeXwrHtzm1QNAjx8caGg7Y3JbTkB9p5JfSQIUNQSp0lF9rYwFAiOcS6desoLCxk5MiR5kWZK423oZIH7Qienp58/fXXREZG8swzz9CxY0eioqL44IMP/h4YeDaGEQuMF612vgq/doNlQ2DPPDh1zJXtizosPT2d0NBQ882yHaxz584AshpamNq1y3jBzWFD6N8+ATc3uGhqzY/h5gYz5kDuYfj6Bbu1Zi++vr5MmTKF8vJyvv76a4qLi13dknCgsw6htdZpwKtACnAYyNNa/wY011ofttQcBkItn9IKOFTpEKmW97Wy/P309wtR791zzz24u7vz2muvWS/QGjbeA27e0PflKh3zmmuuoaCggCVLlpgXtZ0EBQeMy6mFOIvVq1ejta5aHrQMoc/JDTfcQGhoqPnPBIBOtxsblO563XmNiQajYlPC/v37mxflxBsviAR0t+u5AwIC6NWrl+0htH8n8A6RIbRg4cKFeHl5cdFFF5kXpS4EryBodp7zGmtgBgwYQHR0NCkpKbz00kvk5uYyffp0WrRoweTJk/n1118pCegPIxfBFanQ91UoPQEbbocfW8CfUyB9GZSXufqmiDrClZsSVggMDKR58+YyhBamkpKSaN68OYGBgfY/eFkZRH8CAy6B4HMcnXUbDBfeCPNfg8P77NKePQUHB3PNNdeQlZXF/PnzKa9l+dXCfqoSx9EUY3VzOyAMaKSUusHWp1h5n7bxfmvnnK6UildKxWdlZZ2tRSFqtZUrV7Jo0SKeeuop8zyz9CVweAn0ehp8W1bpuCNHjiQ4ONg8VxaMSA43Tzj4bfUbFw1OTEwM3t7enHeejSfxmTHGfdTfQbs/NxA+Pj7cddddLF68mMTEROtF3s2g/c1w4AsoOuzcBkW9Fx8fT8eOHW1n7OZugKZ9jd8jdhYVFcXatWspLS21XqCUsRo6a7Xdzy3qDq01CxYsYNSoUfj7+1svKi+F9F8hbLzkQTtBeHg4Dz/8MImJicTHxzN9+nSWL1/OpZdeSnh4OO+//76xuWi3B2DcNrgkHjrcAoeXwcpLYEFbOPSzq2+GqAOOHj3KyZMnXTqEBmM1dEpKivnVp6LBKigo4NChQ3Tt2tUxJ9j4G2Sn1WxDQmv+/RK4e8K8B+xzPDtr3749Y8eOZc+ePfz++++ubkc4SFXiOC4E9muts7TWJcCPwFAgwxKxgeVtpqU+FWhd6fPDMeI7Ui1/P/39Z9Bav6+1HqC1HhASElKd2yNErfPbb7/h4eHBjBnmGw2S+qOxgqfzXVU+bkUkx6JFiygsLLRe5BUILcZIJIeokpiYGM477zzzTTX+yoM+X/Kg7eCOO+7A19eX11+3sdK5633GgCVprvMaEw3Chg0bbEdxlJdCboLd86ArREVFkZ+fz9atW82LQqIgfx8UWv3vomgAkpKS2LNnD5dddpl5UfYaOJUL4Zc7rzGBUor+/fszZ84c0tLSWLBgAZ06deKuu+76OzNaKQjqDwPfgisPQ9T3Rlzc+lvg1FHX3gBR67lyU8LKOnfujNaaPXv2uLQPUfvs3r0brbXjojiWfQwBwTB4gn2O1ywMpjwKa36GhNo55B04cCADBw5k7dq1JCQkuLod4QBVGUKnAIOVUn5KKQVcAOwEFgLTLDXTgAWWvy8ErlVKeSul2mFsQLjeEtlxQik12HKcqZU+R4h6Ky4ujl69euHn52delLPeuIS0mqvNJk2aVLVIjsJDkGNjAyjR4OXl5bFp0ybbURz5e6EoXaI47CQ4OJibbrqJL774giNHjlgv8u8ArSfCnnehpHZtJCLqroyMDA4dOmR7CH18J5QVQrMBDukhKsrIfK5SLnT2nw7pQdR+CxcuBLA9hE5dCG5e0PJiJ3UlTufl5cVll13Gd999h5eXF4888siZRe7e0OZqGPI5FOfC9ued36ioU9LT0/Hw8MDVi9JatWpFo0aNJJJDnCEpKYmAgABatGhh/4PnZcPaBXDBjeBpsmF8TVx1H7RsD+/eA2UmV6O52CWXXEKHDh349ddfOXDggKvbEXZWlUzoOOAHIAHYZvmc94GXgIuUUnuAiyz/RmudCHwH7ACWAndqrSvCv+4APsTYrHAvYGNyJkTdV1ZWxoYNG2zHG5TkQ14iBFc/x3DEiBGEhoby3XffmRe1usx4cnbQRo1o8P7880/Ky8ttD6EzVhpvZQhtN/fddx8lJSW89dZb5kXdHoSSY7Dv/5zWl6jfKvKgBwywMWDOsewl4KCV0OHh4bRt29b2EDqor5GLnim50A3VggUL6NevH+Hh4dYLtIbUBdB8NHiaxHUIpwkLC+Ohhx7ihx9+MH9sN+0NHf4Fu+fAiWTnNijqlPT0dFq0aIG7u7tL+1BK0alTJ5KTkykrk0xzYSgpKWHv3r107twZ5YgrRP/4AkpL7BfFUcHLB6a/Bgd3wPKv7HtsO3Fzc+Pqq68mKCiI7777jtzcXFe3JOyoKiuh0Vo/qbXuqrXuqbW+UWtdrLXO0VpfoLXuZHmbW6n+ea11B611F631kkrvj7cco4PW+i6tJR9A1G+7du3ixIkTtofQufGgy6HZoGofvyKS45dffqGgoMB6kVcAhI21RHJIwL+wLiYmBk9PT4YMGWJelBkDPqHQxEG5Zw1Qp06duPzyy3n33XfNH8PBgyF4KOyabUQkCHGO4uPjcXNzo1+/fuZFuRvAwx+adHZYH1FRUcTGxmL630E3T+P+L5sTNkiZmZmsXbvW9iro47sgP1miOGqRBx54gFatWvHAAw+YbyzV61ljgcTmh53bnKgzysvLOXz4sMvzoCt06dKF4uJiUlJSXN2KqCX27t1LaWmpY/KgtTaiOLoMgnY97X/8oZcbq6F//9z+x7YTHx8fpkyZAsDXX3/NyZMnXdyRsJcqDaGFEDUTF2dEYNgcQlfEZNRgCA1GJEdhYSGLFy82L2ozCYrSIHttjc4h6r+YmBgGDhxoHhtTkQcdMkLyoO3swQcfJDc3l08++cS8qNuDULAfUn9yWl+i/tqwYQPdu3enUaNG5kU5G4wsV+W4/ypGRUVx+PBh9u/fb14UEgXHNkPJCYf1IWqnX375Ba01l19uY8CcZsR10OpS5zQlzqpRo0Y8//zzrF+/nm+++cZ6kW9L6P4IHPoRMlc5t0FRJ2RnZ1NSUlJrhtDt27fH3d2dpKQkV7ciaomkpCS8vb1p27at/Q++ZyPs32b/VdAVlIJR18GW5ZBTezc/DwoKYtKkSeTm5vLDDz+Yv7Ap6hQZQgvhQHFxcQQEBNC5s42VZNlx0LgjeDer0TmGDx9O8+bN+f77782LWk0AN2+J5BBW5efnEx8fbzuKo+CAkS0uURx2N3ToUAYPHszrr79unnvW6jLj50TiC3B0q1zVIGpMa332TQnLiuHYFmjmmCiOClXOhdbl8iJqA7Rw4ULatGlD7969zYtSFxgvlviZxHUIl7jxxhvp27cvs2bNoqioyHpR1/uN71vC/fI7TZyhtmxKWMHLy4t27dr9tRGdaNjKy8vZvXs3nTt3dkxczNKPjdiMUdfa/9gVRl8P5eWw0uTFwloiIiKC8ePHs3fvXv744w9XtyPsQIbQQjhQXFwcgwYNws3NxkMtJ65GedAV3N3dzx7J4ekPYePg0PfyH31xhjVr1lBWVmZ7CJ0ZY7xtPtIpPTUkSin++9//sm/fPtq1a0fbtm258cYb+fDDD9mzZ4/xZMfNHSKfhKObYUlv+DEUVl8Du9+BvF3GSnUhqiAlJYWsrCzbedDHtkF5icOH0N27dycwMND2EDp4sLEaWyI5GpSioiJ+++03LrvsMvOszaIMyF4HrSSKo7Zxc3PjtddeIyUlhTfeeMN6kYcf9H4JcjfCgS+d2p+o/dLT0/H29qZZs5ot0nGEzp07c/ToUXJyclzdinCxQ4cOUVhYSJcuXex/8OIiWPEVDL8aGgXY//gV2nSFjv1gee3/+duvXz/69evH2rVrOXy49q7cFlUjQ2ghHKSgoIBt27bZjuIoTIWidGhW8yE0GJEcRUVF/Prrr+ZFbSdD0WF5Ii/OEBMTg7u7O0OHDjUvyowxVusHdHdeYw3IhAkT2LFjB3PnzmXQoEEsW7aMW2+9lc6dOxMeHs6UKVN4b1k+Kf1iYfCnEHap8QJW/J3wazf4KQz+vB6SPzQe50KYiI83Nhy0uRI619i4kCAbg2o7cHNzY9iwYbaH0J7+ENhHfnc1ML///jtFRUW286DTfwU0hNuoES4zatQoLrvsMl588UUyMjKsF0VMMTY/3TwLSgud26Co1dLT02nZsqVjNnyroYorWyWSQyQlJeHm5kbHjh3tf/DYH6EgDy75t/2PfbrR1xvRH4dq/336oosuolGjRvzyyy8Sy1HHyRBaCAfZuHEj5eXltofQ2RV50Oc2hI6KiqJFixZ8952NuI2w8eDuK5Ec4gwrVqygf//++Pv7Wy/QGjJWWvKg5deGo3Tr1o277rqL77//noyMDHbs2MG7777LiBEjiImJ4Y477qBznwtILOoPQz6Byw/ChGQY9AE0HwUZy2H9rfBrTziZ6eqbI2qpDRs24OnpSa9evcyLcuKNF50aRTi8n6ioKHbu3El2drZ5UUiUseK1vMTh/YjaYeHChTRp0sT2FTqpC6BRWwi0cV8WLvXyyy9TVFTEk08+ab1AuUG/1419S3a+5tzmRK1VWlrKkSNHak0edIWAgABatGjB7t27Xd2KcCGtNUlJSbRv3x5vb2/7n2DZx8amgZEj7H/s04261siHrgOroX18fBgzZgzp6el/LagQdZNME4RwkCpvSujmBU1t5B1Wgbu7O1dffTW//vor+fn51os8GxuD6EM/QHnZOZ1P1B+JiYmsXbvW9sZPmauMTOiwcU7rq6FTStGtWzduv/12vv76a9LS0ti+fTu+vr7MnDnTiOhQCvw7QMdbYNhXcGU6XPQnlBw3VpUJYcWGDRvo3bu37SdOuRuM1YlOWIFWkQu9Zs0a86LQKCgrgtxNDu9HuF55eTmLFi1i7NixeHl5WS8qLYQj0UZefi1aKSn+qUuXLtxxxx188MEHbN++3XpRaBS0vhp2vASF6c5tUNRKGRkZlJeX15o86Mo6d+78VxSDaJiys7PJzc11TBTH4f2weTlcfDPYivO0l2Zh0Ge0MYSuA9F+PXr0oEOHDvzxxx+cOCEbVtdVMoQWwkHi4uJo164dISEh5kU566FpH3A/91dRJ02axMmTJ88SyTEJTmZA1upzPp+oH9544w18fHyYPn26eVHSbGNVZMT1zmtM/INSih49evDss8+yfPly5s+fb60IQoZC1/tg38eQvd75jYparby8nPj4eNt50KUFkJfo8DzoCgMGDMDLy+vsmxOCRHI0EOvXrycjI8N2FMeRP4wXJiSKo9Z78sknadKkCf/5z3/Mi/q8BLoUtj7uvMZErZWebrwYUdtWQoMxhNZak5yc7OpWhIvs2rULwDFD6OhPjP/PXzTN/sc2M/p6OLwPdsU575w1pJRi3LhxlJWVsWzZMle3I2pIhtBCOEhcXJztVdDlZZAbf85RHBWGDRtGy5YtqxDJ4QcpEskhICsri88//5xp06YRHBxsvehEMqQuhI53gIevcxsUZ7j99tuJjIzkgQceMF+F0/Nx8G1p5EXLRqSikuTkZI4fP247D/roZuN+4+A86Ao+Pj4MHDjQ9hDatyU07iAvoDYQCxcuxN3dnbFjx5oXpS0AzyZGTJSo1Zo1a8Zjjz3G0qVLzYcG/h2gy0zY93/GzyDRoKWnp+Pn50dAgAM3ZauhsLAwGjduLJEcDVhSUhJhYWHmMYY1VVYKSz+G/mMgtLV9j21L1ETw9K4TkRwAQUFBjBgxgsTERHkxqI6SIbQQDpCenk5qaqrtIXReorHizE5DaDc3Ny6//HKWLVtGcXGx9SIPP2h5MaT9UicuuRGO9e6771JcXMy9995rXpT0Jrh5Quc7ndaXMOfh4cHcuXNJSUnhf//7n/UiT3/o84rxItfej53boKjVNmwwNhy0OYTOsWxK6KSV0GBEcsTHx1NUVGReFBJlrISW31313oIFCzj//PNp2rSp9QJdDmmLoOVYcDeJ6xC1yl133UX79u154IEHKC0ttV7U41HwDoKE++Vx3sClp6cTFhZWqzYlrKCUolOnTiQnJ5vfl0W9deLECdLS0ujatav9D75+CWSnwjgbV6c6QqMAGDwBVn4LpXVj742hQ4fSrFkzfv31V0pK6kbP4m8yhBbCAaqcBw0QbJ8hNMD48eMpKChg1apV5kVh46HwEOSZZPOJBuHkyZO8/fbbjBs3zvw/UqeOGkPMtlPAt4VzGxSmzj//fK699lr+97//sX//futFEdcZQ7sts4zvoxAYQ2g/Pz+6detmXpSzAXxbGauPnSQqKoqSkpK/huRWhURBcTackNVn9VlycjI7duywHcWRs97YfDXcxl4Golbx9vbmf//7H4mJiXz8scmLo16BEPk0ZKwwXmQQDdKpU6fIysqqlVEcFXr06EFxcTFJSUmubkU4WcX33CFRHIvfh6AWMPhS+x/7bEZfD3lZkPC7889dAx4eHlx66aUcO3bM9txD1EoyhBbCAeLi4vD09KRv377mRTlxRs5u4w52O+/o0aPx8fGxnQtdsblcmo0aUe99/fXXZGZmct9995kXJX8AZYVGxrCoVV555RXc3d25//77rRcoBQPeglO5sPUJ5zYnaq0NGzbQt29fPDw8zItyN0Az50RxVBg6dCiA5EILFi0yho82h9CpC0B5QNglTupK2MNVV13FsGHDePzxx803lOo4HZp0hU0PQtkp5zYoaoXDhw+jta6VmxJWaNeuHf7+/mzZssXVrQgnS0pKomnTprb3fKqJzEOwYTGM+Td4eNr32FUxcCw0DqwzkRwAERER9O7dmzVr1pCVleXqdkQ1yBBaCAeIi4ujd+/e+Pj4mBdlx0HQILvu6u7n58eoUaNsD6H9wqBpX0iXIXRDpbVm9uzZREZGcsEFF1gvKi+B3XOh+Who2tu5DYqzCg8P57HHHuPnn3/mt99+s17UtLeR5b3nHTgqT5QautLSUjZt2mQ7iuPUMTixB4KcF8UBRr5fjx49bA+hm3QxXriVIXS9tmDBAiIjI2nXrp15UdpCCB0BXiZxHaJWUkrx+uuvk5mZyU033WR9aODmCX1fNX4OrbsZTmY7v1HhUrV5U8IKbm5u9O7dm+TkZPLz813djnCS4uJi9u/fT5cuXewfFbP0IyOGaOwt9j1uVXl5w/BrYM3PUFTgmh5q4KKLLsLb25tffvkFfZYYp/LycrZt28ann34qWdIuJkNoIeysrKyM+Ph421EcJSeMTGg7RnFUGD9+PMnJyezZs8e8KGw8ZK+B4ly7n1/UfsuXL2fbtm3cd9995v+JSvkBClNlFXQtdv/999OxY0dmzpzJqVMmK8Z6PQNeQRB/t2RsNnA7duygqKjI9hA6d6Px1ol50BWioqJYs2YNZWVl1guUMlZDZ8oQur7KyckhNjbW9iroE8mQt0OiOOqoQYMG8cILL7Bw4UK6du3KBx98QHn5aRvoho0zNthN+Q5+7Qr7PpHfXw1Ieno6TZo0oXHjxq5uxabevXujtWbr1q2ubkU4SXJyMmVlZfbPgy4rhaUfGhsStoiw77Gr44Lr4WQBrF3guh6qqVGjRlx44YWkpKSwefNmqzVlZWVs3ryZd955hx9//JGUlBR++uknCgrqzrC9vpEhtBB2tmPHDvLz820PoXPjAW23TQkrGz9+PIDt1dCtxhsb+xw22aVc1Guvv/46oaGhTJkyxXqB1rBrNvh3/ju+RdQ63t7evPnmmyQlJTFnzhyToiDo/QJkrYaDXzu3QVGrVGtTwqD+Tujon6KiosjLyyMxMdG8KGQ45CdD0RHnNSacZvHixZSVlXH55TYGzKkLjbetJjinKWF3s2bNYvPmzfTs2ZPp06cTFRX1z1gDpYwXUMcmgH8XY0X0HyMhb6fLehbOk56eXqujOCoEBwfTqlUrtmzZctYVmKJ+SEpKws/Pj9atW9v3wOuXQHaa8zckPF3P4RDSuk5FcgD07duXNm3aEB0dTWFh4V/vLysrY+PGjbz11lssWLAADw8PrrnmGqZPn05xcXGVVk8Lx5AhtBB2VqVNCbMtmxI6YLVZREQE3bt3tz2EDhoI3sESydEA7dq1i8WLF3PnnXeax8VkrzFyYbveC0p+TdRm48aN49JLL+Xpp5/m8OHD1ova/wuCBsCm/xhXYYgGacOGDQQEBNChg419CHLjoXF7I/bCyaKijMxnyYVuuBYuXEjLli3p39/GiyBpCyEwEhrbiOsQtV6PHj1YuXIln3zyCXv27KF///7cf//9/8yKDoyEi1bDoA/g2DZY0hu2PAqlRa5rXDhUUVERubm5tGzpvI1xz0Xv3r3JzMzkyBF5YbS+KysrY8+ePXTu3Bk3Nzs/N1o8D4JaumZDwsrc3GD0dRC/DI7VnYxlpRTjx4+nuLiY6OhoSktLWb9+PXPmzOGXX37Bz8+Pa6+9lttuu43u3bvTvHlzRo8eza5du+RKBheR6YIQdhYXF0fTpk3p1KmTeVHOemjc0WFP9MePH09MTIz5xi9u7tByLBxeCuUmlz6LeumNN97A29ub22+/3bxo1+tG1ma7qc5rTNTY7NmzOXXqFA8//LD1Ajd3Y5PConTY/pxzmxO1Rnx8PAMGDLD95Clng9PzoCu0bduWVq1a2R5CN+0L7r4yhK6HiouLWbp0KRMmTDC/jxbnGN/7VhLFUR8opZg2bRpJSUn8+9//Zvbs2XTr1o0ffvjh79Vpyg063gKXJkHbKZD4AizuCelLXdu8cIiKPOi6sBIaoGfPnri7u8sGhQ3AwYMHOXnyJF26dLHvgTNTYMMSuMRFGxKebvT1xmxg1Xeu7qRaQkNDGTJkCJs3b+aNN95gyZIlBAQEcP3113PLLbeckeM9ePBg2rRpw5IlS8jLy3Nh5w2TDKGFsLO4uDgGDRpke8OCnDiH5EFXGDduHCUlJfz+++/mRa3GG0/ocuIc1oeoXXJycvjss8+48cYbCQ0NtV6Uvw9Sf4aOt4NHI6f2J2qmY8eOPPjgg3z++eesWbPGelHwedD+ZkiaDceTnNugcLni4mK2bt1qO4rjZCYUprgkDxqMgVRUVJTtIbS7lxFjJUPoemfFihXk5+fbjuJIXwK6DMJtZEaLOicoKIh58+axdu1agoODueaaaxg3bhxpaWl/F/mEwJBP4YLlxuaFK8dC7GQoTHdd48Lu6sKmhJX5+vrSpUsXtm3bZr6fgagXkpKS8PDwoH379vY9sKs3JDxdu0jjTx2L5AA4//zzadmyJSEhIUydOpWbb76Zjh07Wp3JuLm5ccUVV1BeXs6CBQsklsPJZAgthB3l5+eTmJhoO4qjMNVYkeiAPOgKw4YNIyAgwHYkR8sxoNwlkqMBmTdvHkVFRdx7773mRUlzADfofKez2hJ28N///pfw8HDuuusu8ydCvV8Edz+InymbPDUwW7ZsoaSk5Cx50PHG26ABzmnKiqioKA4dOkRKSop5UUgUHN0k0TL1zMKFC/Hz82P06NHmRakLwLelSzLLheMNHjyY+Ph43njjDVatWsXkyZPP/H3WfBSM3QK9njXuDxnLXdOscIj09HSCgoLM4+Jqod69e1NYWEhycrKrWxEOorUmKSmJDh064OXlZb8Dl5UaQ+gBl0DztvY77rkafT3sWAuH97m6k2rx9PRk+vTpTJs2jXbt2tleEAg0bdqUMWPGsH///r/2TRHOIUNoIewoPj6e8vLyKuZBO24I7enpycUXX8zixYvNX9nzCoSQYZAmQ+iG4NSpU7z11ltcfPHF9OjRw6QoD/Z+BG0ng1/duBRSGBo1asSrr77Kpk2b+PDDD60X+TaHyKfhyG/Gk3fRYFT853rAABsD5twNgIKgfs5pyooq50LrcrmKpx7RWrNw4ULGjBljPnwqKzYixFpdJnsV1GMeHh7cc889zJs3jz///JPXXnvtzCJ3b+j5GEzYAxHXO79J4TBpaWl1JoqjQocOHWjUqJFEctRjR44cIS8vz/5RHOsX144NCU83yrJx/fKvXNuHE/Tr14+OHTsSHR1NTk6Oq9tpMOR/cULYUcWmhIMGDTIvyokDNy9o2tuhvYwfP57Dhw+zadMm86Kw8XBsi7E6W9Rr3377LYcPH+b+++83L9r7EZTmQ9f7nNeYsJtJkyYxcuRIHnroIdatW2e9qPOdENAT1t8KR+UJU0MRHx9PaGio7R3dczZAQDfw9HdeY6eJjIzE39//LEPoIcYQMlMiOeqLhIQE0tLSbEdxZKw0fj+1kiiOhuD6669n4sSJPP7442zbts16UaPWcJaVbqLuOHHiBCdOnKgzURwV3N3diYyMJCkpicLCQle3Ixxgx44dKKXo3LmzfQ/8ay3ZkPB0oW0gcoQRyVHPr5xUSnHZZZfh4eHBTz/9RHl5uatbahBkCC2EHcXFxdGhQweCg4PNi3LiLJsreTu0l7Fjx6KUsh3JETbeeJu+2KG9CNfSWvP666/TvXt3Lr74YutF5aWQ9CaEjpBLnesopRSfffYZoaGhXHTRRcTExJxZ5OYBw380Nnf7Y5QxeBT13oYNGxg4cKD5pYlaGyuhXRjFAcaT+aFDh9oeQns2gcBekLXaeY0Jh1q4cCFubm6MGzfOvChtgbFPQQsbcR2i3lBK8d577xEYGMiNN97IqVOnXN2ScLC6lgddWe/evSkvL2f79u2ubkXYmdaabdu2/bXi3W4qb0jo7mG/49rL6Ovh0C5ItrGYrZ7w9/dn/PjxpKWl2f7/p7AbGUILYUcVmxKaKi81cjeb2aixk9DQUAYOHGh7CB3QHfzaSCRHPRcTE8PmzZu57777zIdQqT8Zm5J1tbFSWtR6rVu3JiYmhtatWzN27Fh+++23M4uadIILV4FnICy/ELJMNjMU9UJ+fj47d+60nQddmGpsTBjkmk0JKxs+fDjbt2/n6NGj5kUhwyF7HZSXOK8x4TALFixg6NChhISEWC/QGlIXGntZuNedrFhxbkJCQvjggw/YsmULzzzzjKvbEQ6Wnp6OUooWLVq4upVqa9GiBc2bN5dIjnooNTWVvLw8evbsad8DL7FE59WWDQlPN/xq8PCskxsU1kTPnj3p0aMHMTExHDlyxNXt1HsyhBbCTtLS0khPT7edB52XCGWFDs2Drmz8+PGsX7+erKws6wVKQavxcOR3KDvplJ6E882ePZvg4GCuv95GduLO16FxBwirZZeEiWoLCwtj5cqVdO7cmQkTJrBo0aIzixpHwEWrwKc5rLjYuNRd1EsJCQmUl5dXIQ8aaOb6IXRUVBRaa1avtrHSOXS48bs0Z73zGhMOkZqaypYtW7jsMhsxG0c3Q1EatJrgtL5E7XDZZZdx88038+KLL5rHTIl6IT09nZCQEPtu/OZEvXv3Jj093fw5l6iTtm3bhoeHB127drXfQWvrhoSVNQmCgeNgxddgtuF5PTNu3Dj8/Pz46aefKC0tdXU79ZoMoYWwk4o8aJtD6IqNlIKdN4TWWrN06VLzorDxxpP5DCuX7os6b8+ePSxatIgZM2bg6+trvSh7HeSsgy73gJu7cxsUDhEaGsry5cvp3bs3EydO5Pvvvz+zyC8cLoyBRm1h5Vg4bGXVtKjz4uPjAWyvhM7ZAMrD4XsVVMWQIUNo0qQJCxcuNC9qcTG4ecoGm/VAxdUal1xyiXnR4WXG25Y2akS99cYbbxAeHs7UqVMlc7ee0lqTlpZWJ6M4KkRGRqKUktXQ9UhZWRmJiYl06dIFb287xmjG/Qo56TD+Nvsd0xEuuAFyD0O8jTlCPeLn58dll11GZmYmK1ascHU79ZoMoYWwk7i4ODw9PenTp495UXYceDczVpw6Qd++fWnRooXtSI7mo4zLW9MlkqM+mjNnDp6ensyYMcO8aNds8AyA9jc7rzHhcEFBQURHR3Peeedx7bXX8sUXX5xZ5NsSLlgJTbpCzARItbJqWtRpW7ZsISwsjNDQUPOio5sgoEetiDrw8vJi/PjxLFy4kDKz1TdeARA6Cg79VO83zanvli1bRlhYmO1LnQ8vg8De4Fv3LtMX565JkyZ88skn7Nmzh4cfftjV7QgHOHbsGEVFRbRq1crVrdRY48aN6dSpE1u3bpXNzeqJffv2UVhYaP8ojsXvQ7MwOG+8fY9rb0Mvh+Bw+OFVV3fiNJ06daJfv36sWbOGL7/8kpUrV5KcnExRUZGrW6tXZAgthJ3ExcXRp08ffHxsPInPWQ9Bg5y2m7ebmxtjx45l2bJl5peVePhB89HGEFqezNcrhYWFfPbZZ0yaNInmzZtbLzqZCYd+hA7/Bs/Gzm1QOFxAQABLly5l5MiRTJ06lQ8++ODMIp8QGP2HMeRZPRFSfnB+o8Jhtm/ffvYnUMe2Q2CkcxqqgiuvvJKsrCzWrLGRV976SshPhrwdzmtM2FVZWRnR0dFcfPHF5vsVlORD9p9GHrRosEaNGsU999zDW2+9xe+//+7qdoSd1eVNCSvr3bs3J06cYP/+/a5uRdjB9u3b8fHxoWPHjvY7aMbB2r0hYWUenjDxXtiyEnbHu7obpxkzZgwDBw7k+PHjxMTE8OWXX/Lyyy8zd+5cfvrpJ9avX096err5QglxVjKEFsIOysrKiI+Ptx3FUXLCyIR2UhRHhfHjx3Ps2DHbT+bDxkP+Pjie5LzGhMP99NNPHD/+/+zdd1RU59bH8e+hN7EAil2wIPZeY+8aI1hj7zVqTHljYopJbno1UaOxYI9GE8UuNuzG3kCKCnYpSpcO5/3jiNE4Z2xwpvB81nKRO26G7V0KM8/Z57eTGD16tHpR5EqQs5VDaMEsOTk5sWXLFrp27cq4ceP49ddfnyyyLQHtdylLUw8PgMjCsYjE3OXk5HDx4kX9h9CZCUrebrF8nvR5CV27dsXW1pYNGzaoF5V9kCF801+TnoT8d/LkSeLj4+ncubN6UXSgsoBSHEIXel9//TXVq1dn5MiRJCQkGLodIR/dvn0bS0tL9YEJE1GtWjXs7OxEJIcZyMrKIiQkhBo1amBllY+HxTsWKx+7msj7rm5jwcEZ1hWeaWgbGxu6d+/OxIkTef/99xk2bBjt27fHzc2NiIgItm/fzsKFC/nuu+84duwYshjie27iEFoQ8kFwcDD379/XfwgddxKQNVtKmKdTp05YW1vrj+Qo++B2IBHJYVb8/Pzw9PSkdevWugtkGSIWg0szKFpD2+YETdnb27NhwwZ8fX158803+eKLL5580WRTFNoFQMk2cHQoXJpnmGaFfBMZGUl6err+Q+jEYOVj0ZraNPUMihQpQseOHfH391d/ce9QRvl5elPPQbVg1AICApAkiU6dOqkX3QkASwdwa6ldY4JRsre3Z/ny5dy5c4c333zT0O0I+ej27du4u7tjaWnae0msrKyoVasWISEhZGRkGLod4SWEhYWRlZVF7dr5eJdY3kLCxt2MdyHhfzk6K9nVB9fBncI34W9ra4uHhwetWrXi9ddf5+2332batGn07duXChUqsGPHDpYsWcLdu3cN3apJEYfQgpAPnmkp4d0HSwld9CyHKgDOzs60atVK/yG0Y0XlAEIcQpuNyMhI9u7dy8iRI7GwUPlWf++Ecit75VHaNicYhK2tLX/++SeDBw/m448/ZsCAAaSkpDxeZO0EbbZCmW5wYhIcGwc54o2UqQoKCgKgZk09B8wPD6GNZxIawMfHh8jISM6fP69eVN4X4k7B/RvaNSbkm4CAABo2bIirq6t60Z2AB7sr8nEplGCyGjduzIcffsjy5cv13ykhmAxZlomKisLd3Twy3+vWrUt2djbBwcGGbkV4CUFBQRQpUoQKFSrk35PmLSTsPi7/nlMLvg8W16//2dCdGJwkSRQtWpSaNWsyaNAgfHx8iI2NZf78+Rw6dEjkwT8jcQgtCPng2LFjlChRQn9m1L1j4FRFWUyosR49ehAcHMy1a9fUi8r0gJiDkJmoXWNCgVm6dCmSJDF8+HD1ogg/sLSHigO0a0wwKGtra1asWMG3337L33//TdOmTQkPD3+8yMoeWm+CGh/AlYWwuzWk3jRMw8JLyXsTXKOGnjsdEoLAygkc8/GNVj547bXXkCQJf39/9aJyPsrHmxu1aEnIRwkJCRw7dowuXfTEbKREKLnfIopDeMRHH31EgwYNGD9+PDExMYZuR3hJKSkppKen61+ea0LKli2Li4uLiOQwYWlpaVy6dIlatWqpD/K8iIAlUKK08S8k/C/XstBukDLFnXTP0N0YDUmSqFu3Lm+88QbVqlVjz549LFq0iKioKEO3ZvTEIbQg5INjx47RpEkT9cU6sqwcQmucB52nRw/lh91TIznkbIjapVFXQkHJyclhyZIldO7cmfLly+suyk6Fa6uhQj+wdta2QcGgJEnivffeIyAggOjoaBo3bsymTZseL7KwhHpfQau/lWn57Q0gep9B+hVeXFBQEJUqVcLJSc/S0cQgJY5HMq6XhCVLlqRly5b6px2dvcC5uojkMEF79+4lJydH/yH0nQDloziEFh6RdzE1JSWFgIAAQ7cjvKS8CwnmcgiddzB1/fp14uPjDd2O8AIuXrxIbm5u/kZxpCbDyR3Qup/xLyTUpe+7kJEKm0VU3385OTnRv39/+vXrR1JSEgsXLmTv3r1kZ2cbujWjZVzvOATBBCUnJxMcHKw/iiP1JqTd0TwPOk+1atWoXLmy/kNo1xZgXUxEcpiBvXv3cuPGDUaN0hOzcWM9ZCWBp4jiKKw6duzIqVOnqFKlCr169WLmzJlP3kZWvjd0Oa4sLtzbEUJ/Vi6qCSYhODhYfx40KHEcRhbFkcfX15dz584RGaknh7CcL8Tsh4w47RoTXlpAQABFihShWbNm6kV3AsCxEhSpqllfgmmoUaMGERERDB061NCtCC8p7xDazc3NwJ3kn7p16wKIaWgTdeHCBVxcXPI3IubYVsjKgFZ98+85teRRCxp1hY2zITPd0N0YpRo1avDGG29Qu3ZtDh48yIIFC7h5U9xJqos4hDZWIT/AUT230QtG4+TJk8iyrP8Q+l5eHrRhDqElSaJHjx7s3buX1NRU3UUWVsq00e1tIIs8I1Pm5+dHiRIl6NWrl3pRhB84VYaSKksLhUKhYsWKHDp0iBEjRvD555/Ts2fPJyd3inorB9Fle8Lpt+HIYMi+b5iGhWeWlZVFaGio/jzo9FhIjzGqpYSP8vHxAXh6JIecIy6gmhBZlgkICKB9+/ZYW1vrLsrNgqi9yusStbvMhELNXDKEC7uYmBgcHR1xdHQ0dCv5xtnZGU9PT86fP6++XFcwSomJiVy7do3atWur3+H8Ig6uU6I4aprwkt1+/wcJMbB7haE7MVr29vb4+PgwaNAgMjIyWLJkCQkJCYZuy+iIQ2hjlZ0Gkcsher+hOxGeIm8pYZMmTdSL7h0DCxsoXlejrp7Uo0cP0tPT2bdvn3pR2R7KgUTcKc36EvJXXFwcGzZsYPDgwdjaqixySr4C0YHgOVK8uRewt7fHz8+PefPmsWvXLho1avTkMjhrZyWao+5XcG0N7Gyu/D0SjNbly5fJysrSPwmdt5SwmHFOQnt6elKnTh39h9AujcC+DNwQkRymIjw8nGvXrumP4rh7FLKTRRSHIJi52NhYs4nieJS3tzfx8fHExsYauhXhOeTt0sjXKI60FDi+DV7pDfmZMa21eu2gSgP460cQC/j0qlq1KpMmTaJ3794UK1bM0O0YHRP+V2DmvN8Fhwpw6k3IzTF0N4Iex44do0qVKri46Fk4eO84FK9v0O3ubdq0wcHBQX8kR+mugAS3xESZqVq9ejUZGRn6ozgilgISeIq7LQSFJElMmDCB/fv3k56eTrNmzdi48T/L3iQLqPkBtN2uRAztaKQsMxWMUlBQEID+SegEpcZY4zhAmYY+dOiQ+ht5yUKZhr6zQ8m6F4xeXo7vU/OgJUso1V6jrgRB0Josy8TGxppVFEceLy8vAEJDQw3cifA8Lly4QNmyZSlRokT+PemJ7UqEhalGceSRJGUa+mYY/LPZ0N0YPVtbW/2vwQsxcQhtrKzsof73kHBOuW1eMEqyLHPs2DH9URy52XDvJLjomZTWgK2tLR07dmTr1q3qt4bZuSmRIeK2ZpPl5+dH/fr1qVevnu6C3ByIXKpMlzmU07I1wQQ0b96cU6dOUatWLYYMGcKVKzqmnct0ga6nwK4kHOoP6Xe1b1R4qqCgICwsLKhevbp6UWKQsgvAvrRmfT0vX19fcnNz2bxZzxuecj6QkwZRuzXrS3hxO3fupEqVKnh6eqoX3QkA12ZgU1S7xgRB0FRiYiKZmZlmOQldpEgRypYtS1hYmKFbEZ5RbGwsUVFRT9+l8bwOrINiJaFWq/x9XkNo3RdKVYR13xu6E8GEiUNoY1ahH7i1gnMfQmaiobsRdLh58yZ37tzRfwidGAw5qQbLg35Ujx49uHbtGhcvXlQvKtsD4k5CWpR2jQn54uzZs5w+fVr/FHTUbmWKtbJYSCjo5u7uzt9//42VlRVDhgzRvd3ZyQNeWQeZcXB8jFhWaISCg4OpXLky9vb26kWJwUoUhxHH8tStW5eKFSuyYYOeuI2SbcC6KNwUkRzGLiMjg8DAQP1T0OmxEHca3EUUhyCYM3NcSvgoLy8vbt++TXJysqFbEZ7BhQsXkCQpf6dX01OVpYSv9AZLy/x7XkOxtILeb0HwYbh41NDdCCZKHEIbM0mChrMg4y4Ef2HobgQd8nJTGzRooF4Ud1r56NJYg4706969O4D+SI4yPZSPt7dr0JGQn5YsWYKNjQ2DBg1SL4rwA1sXKPuado0JJqd8+fL8/vvv/PPPP3zxhcrPn+J1oN43cHMjXFmobYPCUwUFBemf5pFlJY7DiKM4QImK8fX1ZdeuXaSkpOgusrSBsq/Crc3K3UeC0Tp8+DCpqan6D6GjdgGyyIMWBDOXdwhtjpPQ8G8kh5iGNn6yLBMUFISHhwdFihTJvyc+uQMyUk0/iuNRXUdDkeLw1w+G7kQwUeIQ2tiVaKAsDwv7BZIuGbob4T/ycr703u6cFKosJXSqrFFX6sqVK0eDBg1Yv369elHxeuBQHm78rVlfwsvLyMhg5cqV+Pr6queYZdyDm/5QaYhB88kF09C/f3+GDRvG//73P44cOaK7yOtNcO8Ep96CJPEmy1ikp6dz+fJl/YfQaXcgKwGKGn9enY+PDxkZGezYsUO9qJyP8j0u9rBmfQnPLyAgACsrK9q2batedCcAbEpAiYaa9SUIgvZiY2MpUqQIdnZ2hm6lQLi5uVG8eHFxCG0Cbt26RXx8fMFEcRR1hTpt8vd5DcneCV6dCIc3wC1xPiU8v6ceQkuS5CVJ0tlHfiVJkjRNkqRPJUm69cjj3R/5nA8kSbosSVKYJEldHnm8oSRJFx783q+SZMT3fxqTul+ChR2cedfQnQj/ERYWhouLi/6lhEmhUKQKWBjHLTgDBgzg2LFjREZG6i6QJKg4QHkTmBGnbXPCC9u0aRNxcXGMHDlSvejqH5CbqVzYEoRnMHv2bCpWrMiQIUNISkp6skCygGZLlT0GRwZDTqbmPQpPCgsLIycnR/8tpYkPlhIWM+5JaICWLVvi6uqKv7+/elHprmBhKyI5jFxAQAAtW7ZUnzSTZbizU7m4ZSSvmwRBKBgxMTFmOwUNyp08Xl5eREZGkpGRYeh2BD0uXLiApaUl3t7e+fekGWlwbAu08FViLMxJrylgZQ1//2ToTgQT9NRDaFmWw2RZrifLcj2gIZAK5L3C/znv92RZ3gYgSVIN4HWgJtAV+E2SpLxXkfOAcUDVB7+65ucfxmzZu0OtD+HWJrizy9DdCI8ICwvTPwUNkBwGzk+p0VD//v0BWLt2rXpRhQEgZ4s38ybEz8+PcuXK0bFjR/WiCD9lsqx4Xe0aE0yas7MzK1as4Nq1a0ydOlV3kUMZaLIQ4k7BhZnaNijoFBwcDKB/oifhwSG0CUxCW1lZ0bNnT7Zs2UJWVpbuImsn5eDypr/IKDdSUVFRnDt3Tn8UR8IFSI8SURyCYOZyc3O5e/eu2eZB56levTo5OTlcvnzZ0K0IKnJzcwkODqZatWr5O5V/aiekpZhXFEeeEu7QcRjsXArxMYbuRjAxzxvH0QG4IsvyNT01vYA1sixnyLIcCVwGmkiSVBpwlmX5qCzLMrAc8HmRpgslr2ng5Amn3xJ5h0YkLCzsYd6XTrlZkHwFiuip0VilSpVo1qwZa9asUS8q0VCJD7mmp0YwGjdu3CAgIIARI0Zgqbb0Iu4MxJ8FT7GQUHg+LVu25KOPPmLZsmXqF6/K+0LlsXDxW4jep2l/wpOCgoKwsrKiatWq6kWJwWBXEuxM4wDA19eXxMRE9u3bp15U3gfuX1O+1wlGZ9cuZZBC7yH0nQDlY+nOGnQkCIKhxMfHk52dbdaT0KDs2LC3txeRHEYsMjKS+/fvU7t27fx94oN/QZESUK9d/j6vsejzNmSmw+a5hu5EMDHPewj9OrD6kf89WZKk85Ik+UmSVPzBY2WBG4/U3HzwWNkH//3fx58gSdI4SZJOSpJ0MjY29jlbNFOWtlD/B+VN4+XfDd2NACQkJBAdHa3/EDolQpkodjaeQ2hQIjnOnj2r/oIoL5Ijei+kRWvbnPDcli9fjizLjBgxQr0owk+5Vb3SQM36EszHxx9/TNOmTRk/fjw3btzQXdTwZyV66OgwyIzXtkHhMcHBwXh5eWFjY6NelGj8Swkf1bFjRxwcHPRHcpTtqUTE3NRTIxhMQEAAbm5u1KtXT73oToDy99JB51sEQXgusrgrwmjlvcc390NoCwsLqlWrxqVLl8jJyTF0O4IOQUFB2Nra6r9w/7wyM+DoJmjho8RWmKMK3tCsJ2yaC/d1RPYJgopnPoSWJMkGeA1Y9+CheUBloB5wB/gxr1THp8t6Hn/yQVleIMtyI1mWG5n7LTrPpZwPlGoP5z8RWb1GIO8AV/9SwgeHvEYUxwHQr18/JEnizz//VC+q+DrIuWJBoZHLzc3Fz8+Ptm3bUrmyyvLLnHS4ugrK9wab4rprBEEPKysrVq5cSVZWFsOGDdP9RsrKEVqsUhbeHZ8gIhEMKCgoSH8etJyrXNQ2gSiOPPb29nTt2hV/f39yc3N1F9mVBNeW4hDaCOXm5rJz5046d+6MhYXK24/s+xB7UERxCPkiNjaWBQsWcPfuXUO3IugQE6Pcwl8Y3ut7eXmRnp7O9evXDd2K8B9ZWVlcvHgRb29vrKzyMbf59C5ITTLPKI5HDfkEkuNgyQxDdyKYkOeZhO4GnJZlORpAluVoWZZzZFnOBRYCTR7U3QTKP/J55YDbDx4vp+Nx4VlJEjScpWyzv/CZobsp9PIOofVOQieFKh+NbBK6bNmytG7dmjVr1qhPiRStBUVriEgOI3fw4EEiIiIYNUpPzMbNjcpkamURxSG8uCpVqvDrr7+yb98+fvzxR91FLo2hzudwfS1ErtC2QQGA+/fvExkZqT8P+v515cDPBJYSPsrX15fbt29z8uRJ9aJyPpBwXrkTSTAaZ8+eJTY2ls6d9cRsRO9XlueKQ2ghH+Tk5JCcnMySJUu4fVu85TQ2sbGxFCtWTP8dO2aicuXKWFlZERoaauhWhP8IDg4mMzOzYKI4nIpB/Q75+7zGploj8JkKm3+D4COG7kYwEc9zCD2QR6I4HmQ85/EFHmy4YRPwuiRJtpIkeaAsIDwuy/IdIFmSpGaSJEnAMGDjS3VfGBWrDZXHwaW5kBhi6G4KtdDQUKysrPD09FQvSgoDu1JgU0yzvp7VgAEDCAkJISgoSHeBJCkLCmMPQepN3TWCwfn5+VGkSBH69OmjXnTFDxwqKHdSCMJLGDlyJH369OGjjz7i9OnTuou834OSreHkG+Ig0ABCQkKQZVn/JHRi3lJC0zqE7tGjB1ZWVmzYoGdpbnkf5eMNfy1aEp5RQICS9az3EPpOAFjaQ8lWGnUlmDN3d3dGjhyJtbU1y5Yt4+rVq4ZuSXhETEyM2Udx5LGxscHT05OwsDAREWNEZFnm6NGjlCxZEg8Pj/x74qxMOLoRmvcCa/O/yMKIL8CtPPw8RokhEYSneKZDaEmSHIBOwPpHHv5OkqQLkiSdB9oBbwHIshwMrAUuAjuAN2RZzrtvdyKwCGVZ4RVge378IQqdOp+DlROcftvQnRRqYWFheHp6Ym2tJ+cpOczopqDz9OnTB0tLS/0LCisOAGS4vk69RjCYpKQk1q1bx8CBA3FwcNBddP8aRO0Cz5FKVqogvARJkvj9999xc3Nj8ODBpKamPllkYQnNl4NkCUeGiGW6Gsu7sKh3EjoxWPlYtIYGHeWf4sWL07ZtW/250E6eUKyOiOQwMgEBAdStWxd3d3f1oqgAKNkGLO20a0wwXxHncfl5GKP696Zo0aKsXLlSLIczEjk5Ody9e7dQRHHk8fLyIjEx8WEMiWB4ERERxMTE0KxZM5QZyXxyZg+kJJh/FEceeyeYOg+uh8Cf3xi6G8EEPNOJhCzLqbIsu8iynPjIY0NlWa4ty3IdWZZfezDpnPd7X8qyXFmWZS9Zlrc/8vhJWZZrPfi9ybK4FKgqLi6OEydO6P5NOzeoPRPu7IBb27RtTHgoLCxMfx40KHEcRYzzELpkyZJ06NBBfySHsxcUrwfX9GRHCwazZs0a0tLS9EdxRCwDZPAcoVVbgplzcXFh2bJlhIaGMnXqVN1FjhWh8Ty4e1TZYyBoJjg4GFtbW/WMeICEIHAoZ5R36TyNj48PoaGh+m9rLuej3MWTLt7sG4Pk5GSOHDlCly56YjbuX1PuHhNRHEJ+kSQ4sR3nLbMZMWIE7u7u/Pnnn5w7d87QnRV6cXFx5ObmFppJaIBq1aoBiEgOI3L06FGcnJwKJorDwRkadMrf5zVmTbpDu4Gw+ku4dtHQ3QhGTozFGalhw4bRvXt3bt5UiUGo+oZyQHhiImTc07Y5gZycHC5duqQ/DzrjnvLLyJYSPmrAgAFERERw6tQp9aKKr8O9Y5ASqV1jwlPl5OTw448/UqdOHZo0aaK7KDsVLv0G7p3BqZKm/QnmrWPHjsyYMYPFixezaNEi3UWVBkLlMXDxayWXXNBEUFAQ3t7eWFpaqhclBpnUUsJH9erVC0D/NHR5X0CGW5s16UnQb9++fWRlZek/hL6jxHVQWk9chyA8D4/a0HkEbPwVh6QYhg4dSqVKlfD39+eff/4xdHeFWt40cGE6hHZycqJcuXJiGt9IxMTEcOXKFRo3bpy/Cwmzs+CIPzR/DWxs8+95TcGEWWBfBH4eC2oLpAUBcQhttH788UfS09Pp378/WVlZTxZY2kDzlZAeBUeGKpvuBc1cvXqVzMzMpywlfPAiw0jjOEBZ8mRtba0/kqNCf+Xj9bXaNCU8kzVr1hAeHs4nn3yifgvZ5d8hPRpqfaRtc0Kh8Pnnn9OpUycmT56sviiu0Wwo0QiODoOkcG0bLKSCg4P1R3Hk5ig7JUwsDzpPuXLlaNy4sf5c6GJ1lWl8kQttFAICAnBwcKBly5bqRXcClOl8Z2/tGhPM37DPlSiyZR9ja2vLoEGD8Pb2JiAggMDAQJHPayAxMTFIkoSrq6uhW9GUl5cXd+7cISkpydCtFHpHjx7FysqKRo0a5e8Tnw2E5LjCE8XxqOIlYfxPcPEIbP3d0N0IRkwcQhspLy8vFi9ezNGjR3nvvfd0F7k0goa/wJ3tEPy1tg0WcnlXsfXGcSQ9uN3KiCehixcvTteuXVm7di25alcsnTzApSlc03NQLWgqJyeHzz//nNq1a+Pr66u7KDsVLn4LpTqIJU9CgbC0tOSPP/6gVKlS9OnTh7t37+oosoNWf4GFNRzsDVkp2jdaiCQmJnLjxg39SwlTrkBuBhQzzUNoUC6gHj9+XP1uMUlSIjmidkFWsqa9CU8KCAigbdu22NqqTIXlZkPUHiWKIz9zOQXBrRz4ToM9K+HyGaysrOjbty/16tXjwIEDbN++XRxEG0BsbCwlSpTI3wlUE5D3vlFMQxtWSkoKFy5coF69euo7dV7Uwb+UjOSGhfSunk7DoH5HWDwdYlVeowmFnjiENmL9+/dn6tSpzJo1i7/++kt3UZXxUGkwXPhEeQEvaCLvxcNTJ6EtbMCxkjZNvaABAwZw48YNjh49ql5UcQDEn/13ulswqLwp6JkzZ2JhofJtPG8KuvZMbZsTChVXV1f+/vtvoqOjGThwIDk5OU8WOVaElmsgKQSOjQHxhr/AXLyo5PA921JC04zjAOjduzcA69evVy8q30c5bL+1VaOuBF0iIiK4fPmy/iiOe8cgK1HkQQsF4/X3wdkFFipDPRYWFrz22ms0b96cEydOsGHDBnEQrbGYmJhCtZQwj6urKy4uLuIQ2sCOHz9OTk4OzZs3z98nzsmGIxugWU+wtc/f5zYVkgRvzlf+v5jzhnjNL+gkDqGN3Pfff0+zZs0YNWoU4eE6bmWWJGg8X5m2PTIIUm9p32QhFBYWRokSJfTfRpYUCkWqgIWeXE4j8Nprr2FnZ/cMkRySWFBoBMQUtGBsGjVqxNy5c9m9ezeffKKyhNC9I9T5Aq7/CWG/aNtgIRIUFASgfxI6QamhaA0NOioYXl5e1K5dW/0CPYBrC7BzhxvrtGtMeMLOnTsBnp4HLVko3ycEIb85FoXBH8OZ3XBS+fsoSRKdOnWiTZs2XLhwgZCQEAM3WXhkZ2cTFxdXqPKgH+Xl5UVkZCTp6emGbqVQyszM5OTJk1SvXp0SJUrk75Of3w+JdwtnFMejylSGoZ/B0U1w8G9DdyMYIXEIbeRsbGxYu3YtNjY29O3bl9TU1CeLrJ3glb8g+z4cfh1ydWRIC/kqNDRU/xQ0QHKYUUdx5ClSpAg9evRg3bp1uqcYARzKKoeZ19aIK5oGJqagBWM0evRoxowZw1dffcXGjSpLCGu8r0QknHkXYg5q2l9hERwcjKOjIxUrVlQvSgwCRw+wctSusQLQt29fDh06xJ07d3QXWFgq09C3t4kYGAMKCAigYsWKVKtWTb3oTgCUaAI2xbVrTChcXp0IpT1h0Xvw4LWuJEm0bt0aFxcX9u/fL6ahNXL37l1kWS6Uk9CgHELn5uZy+fJlQ7dSKJ07d460tLT8n4IGJYrD1gEadc3/5zY1fd6CKvXhtymQHG/obgQjIw6hTUD58uVZtWoVQUFBTJw4UfeLpKLe0GQRxB6CczO0b7KQCQsL058HnZsFyVegiPEuJXzU66+/TnR0NPv371cvqjBAuZ0+MUi7xoTH5E1B16lTR0xBC0Zn9uzZNGrUiGHDhqnfudNsKTh5wqF+kHpb8x7NXVBQEDVq1FC/QAVKHIcJ50Hn6devH7Is8/ffeqZsKvSDnHTlIFrQXGJiIrt27aJbt27qC3RTb8K9E1BGvGkXCpC1DYz8CiLOwd5VDx+2sLCgTZs2xMTEPIwzEgpWTEwMQKGdhC5XrhwODg4iksMAcnNz+eeffyhbtizly5fP3yfPyYFD66HZq2CXzznTpsjSCt5aBAmxsGi6obsRjIw4hDYRXbp04ZNPPmH58uUsXrxYd1Gl16HqGxDyA9zQszVeeClJSUlERUXpn4ROiQA5G5xN4xC6e/fuODk5PSWSo69yu6xYUGgwzzQFfWm+mIIWDMLOzo6//voLa2trevfuTUqKjulTm6LQar2yLO5QP8jJ1L5RMxYUFKQ/DzonU8n2L2r6h9De3t7UrFlTfySH2ytgVwqui0gOQ1i2bBn3799nzJgx6kWX5isfPYZr05RQeLXuB16NYelHkJH28OGaNWvi6urK/v371Zd0C/kmNjYWCwsLXFxcDN2KQVhYWFCtWjUuXbqkfgeqUCDCw8OJi4ujefPm6hdGX1TQQUiIEVEcj6raAHq/BdsXKlElgvCAOIQ2IR9//DGdO3dm8uTJnD59WndRgx+hRGP4Z4QyiSvku2deSggmEccB4ODgwGuvvcbff/9NVpZKnItdSSjVXsmFFrcsau7RKWgfHx/dRdmpECKmoAXDqVixImvWrCEkJIQxY8bovnOnWC1ouhjuHlGiOYR8cffuXaKjo/XnQSdfUi6QmvBSwkf17duXAwcOEBUVpbvAwhLK94bbW5XIMkEzubm5zJ07l2bNmtGwYUPdRTkZcGUhlO0JTpU07U8ohCwsYMx3EHsDNs5+5GFlGjo2NlZMQ2sgJiYGFxcXLC2Ne2dOQapevToZGRlcu3bN0K0UKkePHqVo0aJ4e3vn/5PvW6MsI2zcLf+f25QN+0yJQvppDNxPMnQ3gpEQh9AmxNLSkpUrV+Lm5kbfvn2Jj9eRr2NpC63WgWQJh/pCdtqTNcJLCQ0NBZ71ENo0JqFBieSIi4tj9+7d6kUVX4eUKxB3SrvGBOB5pqBjoPanmvYmCI/q2LEjX375JX/++Se//vqr7qJKr4PXNAifDZGrdNcIzyU4OBhA/yR0XpySGcRxwL+RHOvXr1cvqtAPctLg9nbtGhPYvXs34eHhTJ48Wb3o+l/Kz6xqemoEIT/VbQtNe8DqryDp3sOHa9asiZubm5iG1kBsbGyhjeLI4+npiZWV1cP3lELBu3XrFtevX6dZs2b6I8teRGY67P8TWvYGe6f8fW5TZ+cA7yyBqEj4caQYZBMAcQhtctzc3Fi7di03btxgxIgRuqfMHCtC8xUQfxZOTdW8R3MXFhaGpaUllStXVi9KClVuAbYppllfL6tz584UK1aMP//8U72onC9IVnBdT42Q755rCtq9I5R8RdP+BOG/pk+fjo+PD++88w7Hjh3TXVT/O3BrBcfHQpKODGnhueQdQuudhE4IUi5Sm9AFUn1q1qyJt7c369bpidtwaw22biKSQ2Nz5syhZMmS9O2r59bk8DnK30X3Dto1JgijvoG0ZPjjy4cPSZJE27ZtuXv3LkFBYvdJQcnMzCQ+Pr7QLiXMY21tTeXKlQkLCxMLMTVy9OhRbG1tqV+/fv4/+T9bICUBOg7L/+c2B3VaK3ehHFoPa78zdDeCERCH0CaoefPm/PDDD2zatImffvpJd1HZHlBzBlxZJKbM8llYWBienp7Y2NioFyWHmdybfFtbW3x9fdmwYQPp6ekqRSWgdJcHkRxiUkQrzzUFXUtkQQuGJ0kSy5Ytw83NjY8//lh3kYU1tHyQMR/8lXbNmamgoCCKFi1K2bJl1YsSg6FIFbC0066xApYXyREdHa27IC+S49YW5WKdUOAiIyPZsmUL48aNw9bWVnfRvZNw7x9ll4kk3o4IGvKoBZ1Hwua5cCfy4cPe3t6UKlWKAwcOiGnoAnL37l2g8C4lfJSXl9fDPUNCwUpISODixYs0bNhQ/WfSy9i9HEqUhvrigqqqPm9BmwGwZAac2WPobgQDE6/6TNTUqVPp3r07X375Jffvq+Qc1v4MSjSCCzMhVyw+yC+hoaH6ozhAieMoYlqH0KBEciQlJbFjxw71oooDIPUG3P1Hu8YKMTEFLZgqZ2dn3n77bXbt2sWJEyd0FzmUgSrj4OpKSLmqaX/mJjg4mFq1aulftpMYZBZLCR/Vr18/cnNz2bBBz0LmCv0gJxXu6PnZJuSbefPmYWFhwfjx49WLLs0FKyfwFAsJBQMY9plygWrpRw8fkiSJNm3acO/ePS5cuGDA5sxXTEwMIA6hAapVqwb8u2tIKDj//PMPkiTRtGnT/H/y+Bg4sR06DIFCnHP+VJIEby+C8t7w5QCIFnnohZk4hDZRkiQxY8YM4uPjWbZsme4iCyvw/j8lw/f2Fm0bNFM5OTlcunSJ6tX1LBzMuAcZd01mKeGj2rdvj6ur61MiOXqBhS1cW6NdY4WYmIIWTNmECRMoXrw4X32lZ9LZ+11lEjLke+0aMzOyLBMUFKQ/iiM7TXk9YCZLCfPUqlULLy8v/ZEcJduArauI5NBAamoqixYtwtfXl3LlyukuSr8LV1eDxzCwdta2QUEAcC0Lvd+GwD8g/N89J9WrV8fd3V1MQxeQmJgYLC0tKV68uKFbMThHR0cqVqzIuXPnxN+1ApSens6ZM2eoWbMmzs4F8PNm3xrIyRZRHM/C3glmrofsLPi8j5KlLRRK4hDahLVo0YLGjRvzyy+/qP/wKt8bHCpA6M/aNmemrl+/TkZGhtktJcxjZWVF37592bRpk/qEvbWzEvdyfZ2YsC9gzzYFfV9MQQtGq0iRIkydOhV/f/+HmcVPcCgHHsPhymJIu6Ntg2YiOjqauLg4/UsJk0KVGCUzWUqYR5Ik+vbty759+4iNjdVdZGGl7DS4tVksbC5ga9asIT4+Xv9CwojFkJsB1d7QrjFB+K/+70FRV1j4fw+XZeVNQ8fFxXH+/HkDN2h+YmNjcXNzy//FcCaqWbNmJCQkqL8+El7aqVOnyMzMpHnz5gXzBXYvhyr1lZgf4enKVYPpK+DSKZgjlhIXVuIngAmTJIm33nqL8PBwtm9X2fpuYQVeUyBmP8Sd0bZBM5R3y5T+Q+gHm45NcBIalEiO1NRUNm3apF5UYQCkR0HUbu0aK4TEFLRgDqZMmYKjoyNff/21elGN6SBniQumLyhvkZbeSejEB8u2zCyOA54jkiP7PtwJ0K6xQkaWZWbPnk2tWrVo3bq17qLcHAj/DUq1h6I1tG1QEB7l6AyDPoZzgXDxyMOHvby8KF26NAcOHCAnRwxb5KeYmJhCv5TwUV5eXri6unL48GGxoLAAZGdnc/z4cSpVqkTp0qXz/wtcDVYOU8UU9PNp/hoM/BB2LIZtCw3djWAA4hDaxPXt25eyZcvy88963rhXHgNWjuLNfT4IDVUOmJ86CW1hA46VtGkqn73yyit4eXnx8ccfk5amMjFWtic4lIcz70JulrYNFhLPPgX9nZiCFoyai4sLEyZMYPXq1UREROguKlJFubh1aR5kxGnboBnIO4TWOwmdGKwsgyxSRaOutFOnTh2qVq2qP5KjVFuwdRGRHAXo6NGjnD17lsmTJ6tnk9/eAqnXoZqYgBKMQNfR4FQMNvzy8CFJkmjbti3x8fGcO3fOcL2ZmfT0dJKSkkQe9CMkSaJly5ZER0dz+fJlQ7djdg4fPkxSUhKtWrUqmC+wZ4WSLd9uYME8vzkb9hk06gJzJ0PocUN3I2hMHEKbOGtrayZPnsyePXvUbxuzKQaeo+D6GnGr80sKCwujePHi+q/iJ4cpb/ItTHM5gaWlJb/99htXrlxRn1y0sodGs5XJutBZmvZXWKxevfoZpqDniSlowSS8/fbbWFlZ8d1336kX1ZwB2SkQPlu7xsxEcHAwrq6u+t/cJwQpd+hYWGvXmEYkSaJfv34EBgZy9+5d3UUW1lDOR4nkyBE5hAVhzpw5FC1alMGDB6sXhc9RLmKX7aldY4Kgxt4Ruo+DQ+sh5vrDh6tWrUqZMmU4ePCgmIbOJ3lxSWIS+nG1a9fG2dmZQ4cOGboVs3Lv3j0OHjxIzZo18fT0zP8vkJMDe1ZC465QvFT+P7+5s7SE91dBiTLwv76QoBKnJpglcQhtBsaNG4eDgwOzZs1SL/J6E3KzIXyuZn2Zo7CwMLy8vNQnfECJ4zDRKI487du3Z8iQIXzzzTcPp7+fUK6X8ibywqdw/7ruGuGFZGVl8emnn1K3bl31KeisZLj4Lbh3ElPQgtErU6YMI0eOZMmSJdy+fVt3UbFayveVsF+Uv9/CMwsKCtI/BQ3KJLSZLSV8VN++fcnJycHf31+9qHw/yE6GOzs166uwiIqK4q+//mLkyJE4OTnpLkoMUWK8qk5Q4uIEwRi89iCbfNO/75HypqETEhI4e/asYfoyMzExMQBiEvo/LC0tadGiBdevX+f6dfF+Kj/IsszWrVuxsrKiS5cuBfNFzgXC3VsiiuNlOLsoiwoTY+Gr15UFj0KhIA6hzUCJEiUYPnw4q1atIjo6WndRkcpQ7jW4PF8s5XkJeYfQqnKzIPkKFDG9pYT/9cMPP+Do6MikSZPUc8oaPZhYPDVVu8YKgaVLl3LlyhW++OIL9SnosF8g4y7U+ULb5gThBb333nvk5OTw448/qhfVmAGZ8XD5d+0aM3GyLBMcHKw/DzorGe5fNcs86Dz16tWjcuXK+iM53NuDTXERyVEAFi5cSFZWFpMmTVIvuvSbEldWeYx2jQnC05SsAK/0hu0LIe3fpdxVqlShbNmyYho6n8TExGBtbU2xYsUM3YrRqV+/Pvb29hw+fNjQrZiFCxcuEBkZSYcOHShSpEjBfJHdy8GxqJJvLLy4KvVh6nw4uxcWf2DobgSNiENoM/Hmm2+SmZnJ/Pnz1Yuqvw0Z9+DqCu0aMyNJSUncvn2b6tX1TDmnRIKcDc6mfwhdqlQpvvnmGwIDA1m5cqXuIseKUHsm3NwIN/UsMhSeWXp6Op9//jnNmjWjR48euosy4yHkB2Vq1LWJtg0Kwgvy9PRk4MCBzJ8/n3v37ukucm2iZJyH/CgiE57RjRs3SE5Ofkoe9EXlYzHzPYTOi+TYs2eP+t+vh5EcmyAnQ9P+zFlWVhbz58+na9euVK1aVaUoCSKWKtnvdmISUjAyvm9CcryS8fqAJEm0a9eOxMREzpwRy91fVmxsLG5ubvrvJi2kbGxsaNq0KeHh4Q8nxoUXk5aWRkBAAGXLlqVhw4YF9EVS4ODf0GYA2NgVzNcoTDoPh16T4a8fwF9E8hUG4hDaTHh5edG9e3d+++030tNV3ri7tYLiDZQMX7GB97mFh4cDT1tK+CC6wsTjOPKMHTuWZs2a8c477xAXp7IsrPpbyi3eJ6coi/KEl7JgwQJu3rzJF198of5C/eL3yhv62p9r25wgvKT333+f1NRUfv31V/WimjMgPQoilmjXmAkLDg4G0D8JnajUmHMcB/wbybFx40b1ogr9lO+fUbu0a8zM+fv7c/v2bSZP1rNsMHKFkvkuFhIKxqhGC6jaUFlQmJv78GFPT0/Kly8vpqHzQUxMjIji0KNJkyZYW1uLaeiXtHv3btLS0nj11VfV7yZ9WYfWQ0YqdBJRHPlmwixo4QPz3oT9aw3djVDAxCG0GXnrrbeIiYlh9erVugskSTkwTAqBOwHaNmcGwsLCgKcdQis15jAJDWBhYcH8+fOJi4vjgw9UbpGxsIbG85Rt90H/07ZBM3P//n2++uor2rZtS/v27XUXpccoURwVB0DxOto2KAgvqWbNmvj4+PDrr7+SnKyS+1yyLbg2VzLPc7M07c8UBQUFAU85hE4IAkt7cPTQqCvDaNCgAR4eHvojOUp1AOtiIpIjH82ZMwdPT0+6du2qu0CWlYWEJRqLu3cE4yRJ4DsNboTC6V2PPCzRunVrkpKSHn6vFZ5famoq9+/fF0sJ9bC3t6dhw4ZcuHCBhIQEQ7djkq5fv87p06dp1qwZ7u7uBfeFdi+H0p7KxSshf1hawgd/QM1X4LuhcGavoTsSCpA4hDYjHTp0oHbt2vz888/qGb4V+oN9aQj9WdvmzEBoaCgWFhZUrlxZvSgpFOxKgU0xzfoqaHXr1mXatGksWLCAI0eO6C4q2Qo8Rym30CeIF+kvau7cuURHR+ufgg7+GnIzoPZn2jYnCPlkxowZJCQkMG/ePN0FkqRMQ9+/BldVLqoKDwUHB1OmTBlKlCihXpQYDEVrgIWldo0ZQF4kx+7du9Xv3rG0UaKMbm6EnExtGzRD58+f58CBA0yaNAlLS5W/X9F7lddHYgpaMGZt+kMJd2Ua+hGVK1emZMmSHD58WP39laCXWEr4bJo3b44kServtwRVOTk5bNmyhaJFi9K2bduC+0IxN5T84o7DlNerQv6xtYfPN0HZavCZD1wWMUjmShxCmxFJkpg2bRoXLlxg716Vq0eWNsqbgKidkBCsbYMmLiwsDE9PT2xtbdWLksPMZgr6UZ9++inly5dnwoQJZGWpTCbW+xasneHERJBzddcIqhITE/n222/p1q0bLVu21F2UehMuzQOP4eBcTdsGBSGfNG7cmE6dOvHTTz+RlqayKLdMDyhWBy5+Lb6fPEVQUJD+KWiAxCCzj+LI07dvX7Kzs9m0Sc+eggr9ICsRonZr15iZmjt3Lvb29owcOVK9KHwO2LpCxf7aNSYIz8vaBl6dBCe2w/XQhw9LkkTLli2JjY3l8uXLBmzQdMXGxgLiEPppnJ2dqVOnDmfOnOH+fRFx+DyOHDlCbGws3bt3x8bGpuC+0N5Vyt09HYcW3NcozJyKwVc7wKk4fNgN7kQYuiOhAIhDaDMzaNAg3NzcmDVrlnpRlfHKbblhemqEJ4SFhemP4gAljqOI+R1COzk5MXv2bC5cuMAvv/yiu8jOFep/D7GHIGKZtg2agVmzZhEXF8f//qcn0iTof0Au1PpYs74EoSDMmDGD6OholixRyX3Om4ZOCoUbG7RtzoTk5uZy8eJF/UsJM+Mh7TYUNd+lhI9q1KgRlSpV0h/J4d4RrIvCDRHJ8TLi4+NZuXIlgwcPVp/Ev39dWQRZeSxYigVOgpHrMV45jN74+HKsmjVr4uzsLPJ6X1BMTAy2trYUKVLE0K0YvZYtW5Kdnc2xY8cM3YrJiI+P58CBA3h7e1OtWgEO6ciyEsVR6xUljkMoGK5l4asAyMmGDzpDfLShOxLymTiENjN2dnZMnDiRLVu2PFyk9wRbF/AYpiyJSRcbeJ9Fbm4u4eHh+g+hM+5Bxl2zWUr4X7169eK1115j5syZXL9+XXeR5whwawln/0/5/0N4Jvfu3ePHH3+kd+/e6puck6/AFT+oPA6cKmnanyDktzZt2tC8eXO+++479bsryveFIlUh+EuxTFdFZGQkaWlpT8mDLhxLCfNIkkTfvn3ZtWuXeq6mpS2UfQ1u+ItIjpewdOlSUlNTeeONN9SLLs1XPladoE1TgvAyipeEdoNg51JIjn/4sKWlJc2aNePatWvcunXLcP2ZqLylhKpRc8JDrq6ueHt7c+LECTIyMgzdjtGTZZmtW7diYWGhvpcgv1w6BddDlCgOoWBVqA7/2wr37sCH3SFVZY+MYJLEIbQZmjRpEjY2NuoTqwBe05Rc2bw3B4Je169fJz09nerV9Rwwm9lSQl1+/fVXAKZMmaK7QLJQlhRmJsLZ6Rp2Ztq+//57UlJS+Pzzz9WLLnymLIGs9aF2jQlCAZEkiQ8//JBr167xxx9/6C6ysIQa70P8GbFMV0Xeoiy9k9CJD3L6ixWOSWhQIjmysrKeIZIjQckrFp5bbm4uv/32Gy1btqRevXq6i3Iy4MpC5cDfsYKm/QnCC/N9EzJSYcfixx5u0KABdnZ2Iq/3OcmyTGxsrFhK+BxatmxJeno6p06dMnQrRi84OJgrV67Qvn17nJ2dC/aL7VoO1rbQul/Bfh1B4d0UPv4LIs7BZ70hSwwNmAtxCG2GSpUqxaBBg1i6dKn6Yp6i1aF0N7j0m/ImQdArLEw5YNY7CV0IDqErVqzIZ599xqZNm9i4caPuomK1ofrbcGUxxIrbFp8mKiqKX3/9lUGDBqlPMyZehKsrlTx3+9LaNigIBaR79+7UqVOHr7/+mpycHN1FlYaAQ3kI/krb5kzE8ePHsbS01D8JHXcKrIsp/z8WEk2aNKFChQr6IzlKdwarInBdRHK8iMDAQC5fvsyECXomnG+sV+4QqzpRu8YE4WVVrgd12sCmOcrt4A/Y2trSqFEjLl68qP7+SnhCSkoKaWlpIg/6OZQtWxYPDw+OHj1Kdnb20z+hkEpPT2fHjh2ULl2axo0bF+wXy8qEfauhRS8lt1jQRpNu8PZiOLMbfhgBuWJPjDkQh9Bmatq0aaSmprJw4UL1oupvQXo0XFutXWMm6tkOoUPBwgYcK2nTlIG8+eab1K5dmylTppCSkqK7qPYn4FABjk8QFzme4uuvvyYzM5NPP/1Uvej8J2DlBN7vadaXIBQ0SZKYMWMGYWFhbN26VXeRpQ14/x/EHoS7x7Vt0AQEBgbSuHFjnJyc1IuiA6Fk60K1xT0vkmPnzp3ExKjEjlnaQrlecONvyFL5WSaoWrBgASVKlKBv377qRZcXgJOnksEtCKbEdxpEX4Mjjw9cNG3aFEtLSzEN/RzylhKKSejn88orr5CSksL58+cN3YrR2r17N6mpqfTs2RMLiwI+1jq5AxLviigOQ+g8HMZ8C4GrYfH7hu5GyAfiENpM1a1bl3bt2jF79mz1vE33jsqiotCfRd7mU4SFhVG0aFH9V/GTw6BIFbCw0q4xA7C2tmb+/PncvHmTrl276p4GsXKERnOU28D3vwpZIsdJl+vXrzN//nxGjhxJlSpVdBfFnVYOSaq/pSx/FAQz0qdPH0qXLs2CBQvUizyHg6UDXNFTUwilpKRw4sQJ2rVrp150/wakXIFSemrM1NixY8nKymL27NnqRdUmQ1aiEhkhPLOYmBg2bNjA8OHDsbNTWTaYFAYx+5SFhJJ4uyGYmGY9wd0DNjwebejk5ETdunU5d+4c9+/fN1BzpiXvQqCYhH4+Hh4elC5dmsOHD5Mrpj+fcOTIEU6dOkXTpk0pXVqDu0R3LYdiJaFh54L/WsKT+v0f9JwE675/4uKgYHrEq0Iz9tZbb3Hr1i3++usv3QWSBNWnQcJ5ZVJKUBUaGkr16tX1L9RICjXbpYT/1aJFC9asWcOJEydo1aoVN27ceLKoXE9otkT5u7WnnViCqcMXX3wBwMcff6xedP5jsCmuRJwIgpmxsrJi5MiRbN++nZs3b+ousnaGiq/DtTXigtYjDh06RHZ2tv5D6Lyf7YXwELp69er4+PgwZ84ckpNV/t64NoWSbSD0J7Gg8DksWbKErKwsxo4dq150eQFIVuA5UrvGBCG/WFrCa5Mh6CBcOv3YbzVv3pzs7GyOHxd35zyLmJgYHBwccHR0NHQrJkWSJF555RXi4uIICQkxdDtG5Z9//mHXrl3UrFmTTp06FfwXTIqDY5uVpaVW1gX/9YQnSRKM/wmqNlRiOaKuGroj4SWIQ2gz1qNHD6pUqfJwmZxOlQaDrRuE66kRCAsL0x/FkZsFyVegiPnmQf9X//792bFjBzdu3KBFixYEBwc/WeQ5AlpvVDKNd7aElAjN+zRWly9fxs/Pj/Hjx1OhgsrCptgjcHsb1JgONkW1bVAQNDJ69Ghyc3Px8/NTL6oyDrLvi/ioRwQGBmJtbU3Lli3Vi2ICwdZFyeovhKZPn05CQoL+aLIa0yH1JlxTWZApPCY3N5cFCxbQunVrvL29dRflpEPEUijvC/alNO1PEPJNt9Fg7/TENLSrqyvVq1fn+PHjZGaKi1dPc+PGDdzd3fUP8gg6Va9eHRcXFw4cOCCmoR84duwYAQEBeHt74+vrW/AxHAAH/1IyoTsOLfivJaizsYWP1iq50F8OEIsKTZg4hDZjFhYWTJ48mX/++YeTJ0/qLrK0gypj4dZmSLmqaX+mIjk5mVu3buk/hE6JBDnbrJcS6tKuXTsOHDhAdnY2r7zyCocP61hEWLYHtN8DmXGwswXEndG+USP02WefYWNjw4wZM9SLzn8EdqWUW8YFwUx5enrSsWNHFi9erL6g0KWJcpB6WcQm5AkMDKRp06Y4ODioF0UHKpO+hTQOoWnTprRt25affvpJ/bCodFcoVgcufgeyeJP/NHv37iUiIoLx48erF13/W/mZX2Wcdo0JQn5zLAqdR8D+NRAX9dhvtWjRgvT0dM6cEa9p9YmNjeXu3bv630MJqiwsLGjbti0xMTG6h30KmRMnTrBjxw6qV69Onz59sLS01OYLH1gHZapAlfrafD1BXWlPeMcPwo6D3weG7kZ4QYXzXUkhMmLECBwdHZkzZ456UZUJgASX5mnWlykJDw8HlKvRqpJClY+FJI7jUfXq1ePo0aOULFmSjh074u/v/2SRW3PodEhZ3Li7DUTt1bxPY3L69GlWrVrFlClTcHd31110e7tygFRzhpKxLQhmbOzYsVy/fp1du3bpLpAkJVs27qS4kAUkJiZy6tQp/VEcKZFw/xqUaq9dY0Zo+vTp3Lp1i1WrVukukCRl6WtSCNzaom1zJuj333/HxcWFPn36qBdd/h2cKhf6v3uCGfCZqkzbbZn/2MPly5enQoUKHD16VEyo6hEaqrw/0vseStCrZs2alCpVin379hXqv2unTp1i27ZtVKtWjb59+2p3AJ14F84FQut+hWrBs1Fr1Qd6TYa/fxL50CZKHEKbuaJFizJs2DDWrFnzcDvxExzLQzkfuLIIstM07c8UhIWFAei/ip+k1BS2Seg8lSpV4tChQ9SpU4c+ffroXjJW1Bs6HwHHCrCvG1xbq32jRuDGjRv07NmTMmXK8N577+kuSrwIhweBszdU0TNtJghmolevXri6uuqPTfAYoty9I5bIPbw1V38e9IOLfYUwD/pRXbp0oW7dunz33Xfqb+ArDgDHSnDxW017MzVRUVH4+/szYsQIbG1tdRclXoTYg8oUdCGdwBfMSNmq0LQHbJkHGY+/R2rRogWJiYliQlWP0NBQypYti7Ozs6FbMVmSJNGuXTvi4uI4e/asodsxiNOnT7NlyxaqVq1Kv379tDuABji8AXJzlENowXiM/QGqNBD50CZKvDosBCZPnkxGRgaLFi1SL6o2Rbl1UuRtPiEsLAwLCwuqVKmiXpQcpsQm2BTTrC9j4+bmxt69e+nSpQvjx4/ns88+Q5blx4scykGng8qt9Ydfh7DZhmnWQBISEujWrRspKSls374dFxeXJ4tSb0NgN+Wwre02sFR5oy8IZsTW1pbhw4ezadMmoqOjdRfZFIfyfeHqKiUfuhALDAzE1taW5s2bqxdFByo/l5xVcnsLCUmSmD59OqGhoWzatEl3kYUVVH8H7h6BmEPaNmhClixZQnZ29tMXElpYKzshBMEc9HsPEmJg++Pvo6pVq4arqytHjhx58vWuQGJiIrdv31bPjheeWbVq1ShbtuzDCMTC5OzZs2zevJkqVarQv39/rKystG3gwDooUxkq19P26wr6PZoP/dXrIh/axIhD6EKgRo0adOjQgXnz5qn/4CrZGorWgvDZIF5IPSYsLIxKlSqpT/2AEsdRSKegH+Xo6MjGjRsZMWIEn376KRMmTHgy49WmOLTbCeVeg1NT4dxHheLvXEZGBr6+voSHh7NhwwZq19axKCwrCfZ1Vy4Itd0GTpU071MQDGXMmDFkZ2ezdOlS9aIqY5V/J9fXadaXMQoMDKRFixbY2dnpLpDlB3nQbcXto0C/fv3w8PDg22+/VT8sqjwKbF3FNLSK3NxcFi5cSNu2bdXvDMtOg4hlUK432JXUtkFBKCh1WkOtVrD2O8jMePiwJEm0aNGCqKgoIiLE4u3/CgkJAUQUR37Im4ZOTEzk9OnThm5HM+fOnWPjxo14enoa5gA68S6c3QutRBSHUSpTGd5eDKHHRD60iXnqIbQkSV6SJJ195FeSJEnTJEkqIUnSLkmSLj34WPyRz/lAkqTLkiSFSZLU5ZHHG0qSdOHB7/0qiTW5mpk8eTI3btxg40aV3BxJAq8pEH8WYnUslyvEQkNDn/4CKikMiohDaABra2v8/Pz44IMPWLBgAZ999tmTRVb28MpfUHkMBH+p5EeasdzcXEaOHMm+fftYsmQJ7dvryMnMyYSDfSAxGFr9DSXE8guhcKlevTqtW7dm4cKF6rEJbq2UC36FeEFhXFwc586d0x/FkXwJ0m4X+iiOPFZWVrz77rv8888/HDx4UKXIQbkr7PYWSAjStkETsHv3biIjI5+ykHAdZCVAVREjJZiZwR/D3Zuwa9ljD9euXRsnJyeOHDlioMaMV2hoKCVLltR915/w3Dw9PalYsSIHDx4kKyvL0O0UuLwDaA8PD15//XWsra21b+LIRhHFYexa9/03H/qoyt1ugtF56iG0LMthsizXk2W5HtAQSAU2AO8De2RZrgrsefC/kSSpBvA6UBPoCvwmSVJecM88YBxQ9cGvrvn6pxFU9ezZk4oVK+pfUFhpMFgXg3A9NYVMbm4u4eHh+vOgM+5Bxt1CuZRQjSRJfPnll4wcOZL//e9/bN269ckiCyto8juU7gan3oS7x7VvVCMzZsxg9erVfP311wwePPjJAlmGY2Mgajc0XQilO2vfpCAYgbFjx3LlyhX27dunuyBvQeHdI5BQOHM49+/fjyzLT8mDDlQ+ikPoh0aOHImbmxvffPONelG1N8DSAS5+p11jJuL333/H1dUVX19f9aLLv0ORasoEviCYkwYdwasJrPkasv89ALSysqJp06ZERERw584dAzZoXO7fv8/169fFFHQ+kiSJ9u3bk5KSwvHj5vueKTMzk82bN+Pv70/FihUZOHCgYQ6gAQ6uA3cPqCIGg4zao/nQ0dcM3Y3wDJ43jqMDcEWW5WtALyDvcvAywOfBf/cC1siynCHLciRwGWgiSVJpwFmW5aOyci/k8kc+RyhglpaWTJo0iX379nHhwgXdRVaOyu2oN/5WcmkFbt68SVpamlhK+AIkSWLu3LnUq1ePoUOHEhkZqaPIAlqsAPvScKgvpN/VvtECNnfuXL799lsmTpzI9OnTdRed/wiuroA6/xM5mkKh1qdPH4oVK/aUBYXDlMzZQrqgMDAwEAcHB5o0aaJeFB0I9mWhSFXtGjNy9vb2vPnmm2zfvp3z58/rLrJ1USJfrq2G+9e1bdCI3blz52HUlmo0WUKQcnGoyjhx27JgfiQJBn0E0Vdh7x+P/VajRo2wsbER09CPCAsLQ5ZlkQedzypUqECVKlU4fPgwGRkZT/8EExMVFcXChQs5ffo0LVq0YMiQIYY7gE6KgzN7lClo8TPNuOXlQ+fkwJcDHs+Hzs2FnGzlscx0SE99YsmsoL3nPYR+HcjbXFdKluU7AA8+5oW/lQVuPPI5Nx88VvbBf//3cUEjo0ePxs7Ojrlz56oXVXsD5By4PF+7xoxYWJhywCwOoV+Mvb09f//9N7Is07dvX9LT058ssnVR4ifSY+DIIOW2JzOxceNGpk6dSs+ePfn111/RmUB0aT4Ef6VMd9b8UPsmBcGI2NvbM3ToUNavX8/duyoXpezcoJwvRK6AHB3fU8xcYGAgLVu2xMbGRneBLENMoDIFLd44PWbSpEk4OTnx7bd6cp+rv618DP1Jm6ZMgJ+fHzk5OYwbN0696PICsLABj+HaNSYIWmr2KnjWhTVfKYcdD9jZ2VG/fn0uXrzI/fuFe2luntDQUIoVK0apUqUM3YrZadeuHWlpaRw9etTQreQbWZb5559/WLRoEenp6QwdOpROnTphaWn59E8uKEf8lcNLEcVhGspUhrcXKfnQPR2gqyV0lpSP3ayhhy28ag+vOUIvJ9j0m6E7LtSe+RBakiQb4DXgaduAdL3jkfU8rutrjZMk6aQkSSdjY2OftUXhKVxcXBg0aBArVqwgPj5ed5GTJ5TpodxSmWN+V1ifV2hoKPCUpRpJocobL8dK2jRlYjw9PVm+fDmnT59mypQpuotKNIRGcyBqFwTpyJA2Qf/88w8DBw6kUaNGrF69WvcyjZub4eQbyr+5xr+JAyNBQInkyMzMZMWKFepFVcYqCzxvrNeuMSMQGxtLUFCQ/iiOxIvKRT0RxfGE4sWLM27cOP7880+uXr2qu8ixAlQapOSOZ9zTtD9jlLeQsH379lStqjJZn50KkcuhfF+wc9W2QUHQSt409M1wOPD42+GGDRuSm5vL2bNnDdObEUlPTyciIgJvb2/dwxfCSylTpgzVq1fnn3/+ITU11dDtvLSUlBT++OMPAgICqFKlChMnTsTT09PQbSn/xktVgqoNDd2J8Kxa94MPVsOA6fD6DBjyCQz9FIZ9DiO/hFFfw5hvlUWzC95RvpcLBvE8k9DdgNOyLEc/+N/RDyI2ePAx5sHjN4Hyj3xeOeD2g8fL6Xj8CbIsL5BluZEsy43c3Nyeo0XhaSZPnkxqaipLlixRL6o2WXkDe/0v7RozUmFhYTg7O+u/kp8cBkWqKBnHgk49e/ZkxowZLFq0CD8/P91FlUeD50gI+h/c0pEhbUIuXbpEz549KVOmDJs3b8bR0fHJorvH4PAAKN4AXvlT/P0RhAdq165N06ZNWbhwIUp6lw6l2isXTQvZgsK8rGyRB/3i3nrrLSwsLPjxxx/Vi7zfg5xUCNdz51ghsXPnTq5du/aUhYRrIStRieIQBHP2Sm+o4A2rv1Ru837Azc2NChUqcPr0afWfW4XEpUuXyMnJyd88aFmGP7+F1V/BqV2QrDJMVUi0a9eOjIwMk4+AuXz5MvPnz+fq1at0796dAQMG4ODgYOi2lL9fZ3aLKA5T1O515cB5xP9g2GcwdCYM+RgGzoDX34f+7ykH1bb28MPIx+5qEbTzPIfQA/k3igNgE5B3z91wYOMjj78uSZKtJEkeKAsIjz+I7EiWJKmZpFwWHfbI5wgaqV+/Pi1btmTu3LnkPvLi6TGlOymLZcJna9ucEQoLC8PLy0v/lfykMLGU8Bl8/vnndOjQgTfeeIMzZ848WSBJ0GguFK8PR4ZAio4MaSOXlpbG2rVr6dpV2bm6Y8cOSpYs+XhRVjJELIP9rypZ2G23KnnsgiA8NHbsWEJCQtTfYEkWUHkMxOyDpMIzyRAYGIiTkxMNG+qZzIkJBMeK4OShXWMmpFy5cgwZMoTFixejerddsZpQ5lXldVC26U+avYzff/8dNzc3fHx81Isu/a68DirZWrO+BMEgLCxg4IdwNQiOPP42tkGDBsTFxXHtWuFejBUaGoqTkxPly5d/evGz2rUMFr8PSz6EDzpDnxIwoip8PQj+/hmCDkFa4YlCKVmyJLVr1+bYsWOkpKQYup3nlp2dzY4dO1i1ahWOjo6MHTuWxo0bG8/k/NGNIorDnLmUhom/wsUj4P+robsplJ7pEFqSJAegE/Dofa/fAJ0kSbr04Pe+AZBlORhYC1wEdgBvyLKcd4lhIrAIZVnhFWB7PvwZhOc0ZcoUIiIi2L5d5f9+yUKZhr53DO6d0LY5IxMaGqr/Sn5uFiRfhiIiD/ppLC0tWb16Na6urvTt21d3JIyVPbR6MIF/sA9kG//iAFmWOXToEOPGjaN06dIMGDCArKwstmzZQpUqVZSi3By4s1M5XF/vDv+MAJsS0HYH2JXU+/yCUBgNGDAAJycn/QsKPUeAZAlXFmnWl6Ht3buXVq1aqS/qkXMhep+Ygn6K//u//yM9PZ3Zs/VcbK8xHTLuwhWVu3cKgdu3b7N582ZGjhypnkEefx7u/SMWEgqFR9sBUKYK/PGFMqH7QI0aNbC1teX06dMGbM6wsrKyuHTp0tMHeJ7Hvdsw/y3lFvq/7sE3u2DkV+BRGy4chN/fhrdbga8zTKgHhwpHTFfbtm3Jycnh4MGDhm7lmSUmJhIYGMjs2bM5duwYTZo0YcyYMU8O7BjagXVQqiJUa2ToToSC0mEwNOsJS2bArUuG7qbQeaZDaFmWU2VZdpFlOfGRx+7JstxBluWqDz7GPfJ7X8qyXFmWZS9Zlrc/8vhJWZZrPfi9yXJhv1/JQHr37k3p0qX1v/nyHA5WThA+R7vGjExycjI3b97Uv5QwJRLkbLGU8Bm5ubmxbt06bty4wbBhw3RP4zt5QosVEH8GTqlkSBuBiIgIPvvsM6pUqUKrVq34448/6NWrF7t37yYyMpKmTZtCwgU483+wsQIEdlFiRjyGQqfD8GooOKvkawpCIefk5MSgQYNYu3YtCQkJuovsS0PZnhCxFHIyddeYkdu3bxMWFqY/iiPhgpKVXVIcQuvj7e1Nr169mDNnjvoUWclXwLUFhP6gXHAuhPIWEo4dO1a96PLvYGErFhIKhYelFbz+AVw+DSf+HeixtramTp06XLx40Syyel9EREQEWVlZeHt7588TyjL8MgGyMuAdP3AuAQ06wsAPYOZ6+OMGrLkDn21SJtSR4fM+MPsNyDTvxcUlSpSgXr16nDp1isTExKd/goHk5OQQEhLCqlWrmDVrFgcOHKBkyZIMGTKEbt26qV9UN5SUBDi9C1r1FRdWzZkkwZvzwcZOxHIYgAghLYSsra2ZMGECM2fOJDw8nGrVqukoclbeUFxZCPW/L5TTmrt37wagefPm6kV52ZslxJXSZ9WsWTN++uknpkyZwjfffMOMGTOeLCr7KtT8CIK/ANfmSl50AUtPT+eXX37h2rVrWFtbY2Njg7W19WP/7WCTQ13b/dy+eoFrN25jnQMze3tQu05vatSqg62dE0hBymbe6+sg/ixIVlCmO3gMg7I9wNKuwP8sgmAOxo4dy4IFC/jjjz+YNGmS7qLKY+GmP9zaBBX6atqf1kQedP6aPn06/v7+LFy4kLfeekt3UY3pcKAXXFsLHoO1bdDAcnJyWLhwIR07dvz3rp7/yr4PV1dChX5gW0LbBgXBkDoOhZWfwar/QeNuDw+rGjZsyIkTJzh//jzNmjUzcJPaCw0Nxc7OjkqVKuXPE+79A/7ZDON/grIq34dKuEPznsqvQR8pk41//QjBh2DGn1DBfCMT27Rpw/nz59m/fz+vvfaaodt5TFxcHKdPn+bcuXOkpKRQpEgRWrduTf369SlWrJih21N3ZCNkZ0ErEcVh9lzKwMRf4PvhsGkO+L5p6I4KDcnYh5EbNWoknzx50tBtmJ2oqCgqVKjAxIkT+eWXX3QXJYbCVm+o+yXU1HFQaOaGDRvG1q1biY6OxspK5XpNYFdIvgI9w8XV0ucgyzJDhgxhzZo1BAQE0LFjxyeLcnNgXzeIOQCdD0OJgttOfO7cOYYMGUJQUBCurq5kZWU99gugZTVYPhEquUJqlgW21hZYWYIkZ+t+UpcmUGkoVHwd7FwLrHdBMFeyLNOwYUNkWeb06dO6b+3NzYFNHuDsDe0DtG9SQ2PHjmXdunXcu3cPS0tL3UX7X4PEEHhN3Fr4LNq2bcvly5eJiIjQHTch58K22krsS7dzhern/LZt2+jRowdr166lXz+VN+NXFsOxMdDxoDI5LgiFyeZ5MHsSfLsb6nd4+PCiRYvIzMxk4sSJxpNxq4Hc3Fx++OEHqlatiq+v78s/YVwUjK0J5b3gx4Og9nNPl+PblIOl9FSY8ht0Nt87NbZv386JEyeYNGkSrq6Gf78RERHBoUOHiIyMRJIkqlatSoMGDahatSoWFs+zjsxAPn4VIi/AiquF6md+oSXL8ElPOLsX5p9Xv9glvBBJkk7JsvzEtKYJfCcQCoK7uzv9+/dnyZIlJCcn6y4qWh3cO8KleZCrctBmpvIyfXv27Kl+AJ2ZCNF7obyP+CH1nCRJYsGCBXh7ezNgwABWrlz5ZDSHhSW0+EOZwt/3Klz787HsvfyQk5PDN998Q+PGjbl79y7btm0jNjaWhIQE7t+/T2ZmJrlZaWSffIeDMyUqVqpEcrMtOI7IxnpIFtLALBiYC69nwYA06JcEfeOgbzx0OQZek8UBtCC8IEmSGDt2LGfPnuXUqVO6iywslTslonaZ5DLT5xEYGEibNm3UD6Bzc5SLdmIK+pm9//773Lp1iz/++EN3gWQB3tOVmJPbhWuNyS+//ELJkiXp1auX7gJZViLbitYAt5baNicIxqDLSGWS7o8vHnu4QYMGxMbGcvPmTQM1ZhjXrl0jLS1N/y6dZyXLygF/+n142+/5DqABmnSHeWehehP4YQR8NwxSVd7vmrhWrVphY2ODn58f4eGGW9QcFxfHmjVrWLFiBffu3aNt27ZMmzaNgQMH4uXlZRoH0PcT4dROEcVRmEgSvPk7WNnAj6NAV1SokO9M4LuBUFAmT55McnIyK1asUC+qNgVSbyq3OxciBw4cID4+Xv82+NtblZzIcvlwtb8QcnR0xN/fHw8PD4YOHUqLFi04duzY40V2rtB2K9i7w+HXYXdriMufhS9XrlyhdevWfPDBB/j4+BAUFES3bt0eL4o/h7SzCZZhPyJVHoNlj/MUrdLj8ckWSQILKyVmw7oI2BQHm2L50qMgFHaDBg3CwcHhKQsKRyn/Dq8s1q4xjd24cYMrV67oj+KIPwNZieIQ+jl06dKFunXr8u233+reUQBQaSA4lIeL32jbnAEdPnyYnTt38u6776ovJLzpr0ROeU8Xb9aFwsnGDvr9H5zbB0GHHj5cq1YtbGxs1C+emqmQkBCsrKzU43uex/61cHgDDP/8xeM0XMvCN7th2GewdxW80RAun3n53oyMk5MTo0ePxtnZmdWrV7N9+3ays7UbHsvIyGD37t389ttvREZG0qFDB6ZMmUKbNm1wdnbWrI98cXSTEsXRWkRxFCquZWHiLAg6CJvmGrqbQkEcQhdiTZs2pVGjRsyZMwfVWJYyPcCxUqFbULhhwwbs7e3p3LmzetGNDWDnDq6FL/Mtv1SpUoXjx4+zdOlSrl+/TrNmzRg6dCi3bt36t6hYbehyEposgKQw2NEIjo2FtOgX+pqyLLNgwQLq1q1LcHAwK1eu5M8//8TFxeXfotwcCP4GAhpDegy02QJNFyiHzIIgaKZo0aL079+fP/74Q/2uHcfyULorRPiZ7QK5wEAl6/nZ8qDbFnxDZkKSJN5//31CQ0PZuHGj7iILa/B+F2IPQuxhbRs0kJkzZ1KyZEn1LHY5Fy7MhCLVoNIgbZsTBGPSfRwUdXtsGtrGxoZatWoRHBxMerp5L8fLI8syoaGhVK1a9eUXzSXEwtzJ4NUY+rz9cs9laQlDPoHv9ipT1W82A//Z+X5npaG5ubkxZswYmjZtyvHjx1m0aBGxsbEF+jVlWebcuXPMmTOHw4cPU6tWLSZPnswrr7yifhexsTuwDtzKQ/Wmhu5E0Fqn4codFIvfh9tXDN2N2ROH0IWYJElMmTKFkJAQ9uzZo7vIwhKqToKY/crtqIVAbm4u/v7+dO3aFQcHB91F2WlwZzuU66Xcriu8MAsLC4YPH054eDgzZsxg3bp1VKtWjc8///zf7eIWllBlrJK9Xf0tiFgKm6tCyA+Qk/nMXysqKoqePXsyfvx4mjZtyoULFxg8ePDjk83JV2BPGzj3AZTtBd2DlIWCgiAYxPjx40lJSdF/106VCZB2x2zv2gkMDMTFxYXatWurF0UHgnN1sC+tXWNmoG/fvnh6evLNN9+oX5CvPBpsXeDit9o2ZwAHDhxgz549TJ8+HUdHR91FN/5WXhPWnqncCSQIhZWdA/R9B04GQOjxhw83bNiQ7OxsLlwoHO+dbt26RXJycv5EccydDKlJ8M4SsMyn7y912sD8c9CgE/w2VTmINjNWVlZ07dqVQYMGkZyczIIFCzh16pT6z7WXcPPmTRYvXoy/vz9FixZl9OjR+Pj4UKSICQ/r3E+EUwHwSh9xd09hJEkwbQFYWcNPo0UsRwETp2eFXP/+/XFxcdF/q3Pl0WBhA5cXadeYAZ06dYpbt27pX6oRvUfZCi+iOPKNk5MTX375JSEhIXTv3p2ZM2dSvXp11qxZ8+8LKJti0OBH6BEEJVvDmf+DbbXg5ma9Uw2pqaksWbKEWrVqsWfPHn755Rd27dpF+fLl/y3KTITQX2B7XUgIguYr4JW1ItNZEAysadOmNG7cmNmzZ6tHJpTpDk6eEParts1pJC8PWjVTMTdLmdQt1V7bxsyAlZUV//d//8fx48fZt2+fSpEjVJsKtzYrPx/M2MyZM3F3d2fChAm6C3Jz4MKnyjLQCgM07U0QjFLPSVCkOKz+8uFDZcqUwd3dvcAOAY1NSEgIFhYWVKtW7eWe6NB6JYpj8CdQqWb+NJenqCt8tgmavgoL34Vw84xLqVq1KhMmTKBChQps2bKFdevWkZaWli/PnZKSgr+/P4sXLyYxMREfHx9Gjx5NuXLl8uX5DeroZsjKFFEchZlrWZjwM5zfD1vmGbobsyYOoQs5Ozs7Bg8ejL+/P3FxcbqLbEsoh61XV0JOhrYNGsCGDRuwtLSkRw890683NoC1s8jeLAAeHh6sW7eOffv24eLiwsCBA2nUqBHffPMNQUFByot5Zy9ouwXablMm0Q+8BjtbwMXvITEEZJnc3Fz279/PqFGjKFWqFKNGjcLDw4PTp08zdepU5TAnN1tZNnV4IGxwh9PTwKUZdL8AHkPElXBBMAKSJDF16lRCQ0PZvXu37iILS6g2GWIP5VtuvLGIjIzk2rVr+qM44k5Bdor4mfSCRowYQalSpfjmGz25z9XeUA6jL36nXWMaCwwMZN++fXzwwQfqd4JdXwuJF6H2p8q/O0Eo7ByKQK8pSp7s9dCHDzdo0IDo6Ghu375twOYKXl4Uh4eHB3Z2di/+REn34NeJUKU+9H8v/xp8lIUF/N9SKFYKvhoA95MK5usYWJEiRRgyZAgdO3YkLCyM+fPnc+3atRd+PlmWOXnyJHPmzCEoKIiWLVsyefJk6tat+/jdpKbs4DrlENJbxGwWap1HQKOusPA9uGm4RZ/mThxCC4waNYrMzEz17fAAlUdBZhzc2qRdYwayYcMG2rZtS4kSJXQX5GYr/z+U6QGWKgt7hJfWpk0bTp48yaJFi5AkiQ8++IDatWvj4eHB5MmT2bFjB+kl2ikHxg1/gZw0OPsebK3BvWUuLJlSnM/faIv/+rX069ePffv2cezYMby9vSH+HJx+B/zLwb7ucGcneI6Gzseg/S4lY1YQBKPRr18/SpUqxezZem6h9RwJlg4Qbl632T5bHvRe5WPJtgXfkBmys7PjrbfeYufOnZw+rXIRw9YFKo+Da3/A/Rd/M2+sZFlm5syZlClThnHjxukuys1WpqCL1oIKfTXtTxCMWs83wNoWNsx6+FDt2rWxtrZW/55iJmJjY4mLi3v5KI7f3oTkOCWGw+olc6X1cXaBGash6irMGmd2+dB5JEmiZcuWjB49GisrK5YtW8b69euJjn6+nTrR0dH4+fmxdetWypQpw8SJE+nYsSO2trYF1LkB3E9SInVa9VUuVAiFV14sh609fPwqJKkMaQovRfwrE6hbty4NGjTAz89PvahUB2U7/BU9NWYgNDSU0NBQ/VEcsYch4y6UF1EcBc3S0pLRo0dz8uRJbt26xcKFC6lbty5+fn5069YNFxcXfHr3Y9FBBxbdnkyfFY2Y6AfHLsYztGkKe2bAvQWW+I1KoU35a1iE/Qzb6sL2espBlWsLaLUBfO9A4zng2kRMPwuCEbK1tWX8+PFs3bqVy5cv6y6yKQaew+Hqakgv2IU8Wtq7dy8lS5akRo0a6kXRgcoSVxEf9MImTJiAs7Mz336rJ/fZ+23l7puQH7VrTCN79uzh4MGDzJgxQ32a8dpqSA6HOp+JfRiC8KjiJaHjUNi1TFmsh3Jxq2bNmly4cIGMDPO9kzQkJATg5Q6hj26Cvatg4IdQuW4+daZHrVdg+Oew/0/Ybt5xk3kXFps1a/ZwKnrVqlVcvXpVb1RMVlYWu3fvZsGCBcTFxeHj48PQoUMfX+RuLo5tgawMEcUhKEqWh0/9IeYafN5biWkR8pV4BSkAyjT0mTNnOHPmjO4CC0vwGA53AuD+DW2b05C/vz8AvXr1Ui+6uQEsbKF0N22aEgDlRdSYMWPYuHEj9+7dY9u2bYwYMYLTp08zduxYxo4dS3BkMhU6fEXtKdexGZQErf2RKvaHmIPwz3A48y5Y2kOjucrBc+v1UN5HTLQLggmYMGEClpaWzJ07V72o2mTIzYArevYcmBBZlgkMDKRt27bqt7zmZCgXR0uKKI6XUbRoUSZNmsRff/3FpUuXdBc5lINKQ+HKIrO60CHLMp988gnlypVjzJgxuotys+HCZ1C8HpTz0bI9QTANfd6GzPTHskQbNGhAVlYWQUHmmyUfEhJChQoVcHJyerEnyM2FedPAozYMnJGvvek14P1/FxVGmvcCSVtbWzp37sy0adNo164dt2/fZtmyZSxevJiQkJAnDqMvXbrEb7/9xuHDh6lTpw5vvPGGeUVv/NeBvCiO5obuRDAWtV5R7so4vx9+Hmu2d0wYijiEFgAYOHAgtra2LFmyRL3IcwQgQ+RyrdrS3IYNG2jcuLH6ggVZhpv+4N4JrF/wxZbw0uzt7enWrRtz587l2rVrnDt3jlOnThESEsIHH3ygLBy0coRyvaDpQvC9BV1Pwath0OUfqDZJubVaEASTUbp0afr374+fnx/Jycm6i4rWUL4/h/+mLOszcZcuXeL27dv6ozjuHVfiiEQe9Et78803sba25ocfflAv8v4/yEk3qyWYAQEBHD16lI8++kj9FuvIFZByBWqLKWhB0KmCNzTpDpvmKofRQLly5XBzczPbSI74+Hiio6Nfbgr6/H6IioTXPwBrDYdCLCxg+gpwLApf9Ie0+9p9bQOxt7endevWTJs2je7du5OamsratWuZO3cup0+fJiEhgb/++os//vgDKysrhg8fTq9evdR3BJiD1GQ4sR1e6SOiOITHtR8Ewz6D3cth9VeG7sasiH9pAgAlSpTA19eXlStXkp6erruoSGUlbzLCD+RcTfvTwq1btzh+/Lj+KI74s0oWpIjiMBqSJFGnTh0aNGigfoVesoASDcD5Jbd2C4JgUFOnTiUpKYnly/VcDPWaCmm3lAWyJu7Z8qADAQlKtdGmKTPm7u7OyJEjWbp0KXfu3NFdVLS68hogfA5kqVwMMSF5WdAVK1Zk5MiRuotyMiHocyjRCMr21LZBQTAlfd6BhBjYsxJQXqM2bNiQ27dvExUVZeDm8l++RHHsXAoOztDCJ196ei7FS8H7q+BmGMydrP3XNxBra2saN27M5MmT6dOnD9bW1mzevJlffvmF0NBQ2rVrx/jx46lUqZKhWy14+/9UojjaDDB0J4IxGvwxdBgCSz+CfX8auhuzIQ6hhYdGjhxJfHw8mzbpWT5YeRSkRCjxBmYmL4pD7yH0zQ3KgaZ4EyYIgqC5pk2b0qRJE2bPnk1ursrF0DLdwakyhJv+pGpgYCClS5emWjU9F9CiA6F4fbAprl1jZuzdd98lOzubn3/+Wb3IezpkJcDlBZr1VVC2bdvG8ePH+eijj7CxUZlCjFwK969Cnc/F3gRB0KdeO6hcD/7+SYmZAOrUqYOlpSWnTp0ybG8F4MKFC5QpU4bixV/w509qMhz8C9q+riwCM4T6HWDQR8ph+C7zvdtXFwsLC2rVqsW4ceMYMmQILVu2ZOLEibRu3RorKytDt1fwZFm5c8GjNtQQURyCDpIEby2CWq3g++EQfMTQHZkFcQgtPNShQwfKly+vP5KjfB+wKgIRempMlL+/P9WrV9d/Nf/GBnB7BezctGtMEARBeGjq1KmEhYWxa9cu3QWSBVR7Q8lJjjPdW6BlWWbfvn20a9dO/S6P7DS4e1REceSjypUr079/f+bNm0d8fLzuItcmUKo9hP6kZHKbqLwsaA8PD4YPH667KCcDgr4Al2ZQuqu2DQqCqZEkZRr6egicDACUCIQaNWpw4cIFsrJMPyYqz507d4iKiqJevXov/iSH/oaMVOik8v1HK0M+gdqtYfYkuB5q2F4MQJIkKleuTMeOHc1z8aCai0fhylno+Ya4wCqos7GFTzeAW3n41AfuRBi6I5MnDqGFhywtLRkxYgQBAQHcuKGyfNDKASoNhOvrICtJ2wYLUHx8PPv27cPHx0e9KPkyJAZBORHFIQiCYCj9+vXD3d2dX3/VM+nsOVLJhQ+frV1j+SwkJITo6Gj9URx3jyqLGMUhdL6aPn06KSkpzJs3T72oxvuQdhuurtSusXy2adMmTp8+zSeffIK1tbXuoiuLIfWGmIIWhGfVpj+4lIG/f3z4UMOGDcnIyCA4ONiAjeWvM2fOYGlpSa1atV78SXYuhbJVDT+FamkFH/wBNnbw1QDISDNsP4I2Ns1VomA6DDZ0J4Kxc3aBL7ZCbg581ANSEgzdkUkTh9DCY0aMGIEsy/rzNj1HQU4qXFurXWMFbMuWLWRnZ+uP4sjLFxVb4QVBEAzGxsaGCRMmsG3bNi5duqRSVAw8hsPVPyA9RtP+8svWrVsB6NSpk3rRnR1gYQ0lW2nUVeFQr149unXrxqxZs0hLUzmMcO8IxRvAxe+UNyUmJjc3l5kzZ1KlShWGDBmiuygnHYK/VO4Ac++obYOCYKqsbcBnKpzZo0xZAhUqVMDFxYUTJ04gy7Jh+8sH2dnZXLhwAW9vb+ztXzBG406EspSw8wjjuMDlWhbeWw4R5+H3tw3djVDQ4qPh4Drl75+9k6G7EUxBuWowcz3cuQL/6wvZ5nNni9bEIbTwGE9PT9q1a4efn5963qZLEyhaQ1lQaCY2bNhA2bJladSokXrRTX8ld9OpklZtCYIgCDqMHz8ea2tr5s6dq15UbTLkZsLlhdo1lo/8/f2pX78+FStW1F0gy8rF0VLtwdpZ2+YKgffff5/Y2Fj1iDJJgprvQ3K48vrAxPj7+3Pu3Dk++eQT9ezPywuUaW8xBS0Iz6f7OLBzVLKhUeIOmjZtyu3bt7l+/bqBm3t5oaGhpKenU79+/Rd/kl3Lle8rHYbmX2Mvq0l36Pd/sGW+chFBMF/bFymHiD0nGboTwZTUaaNkRJ/ZA79OVF6LC89NHEILTxg1ahQREREcPKiyfFCSlFud7x6FxBBtmysAqamp7NixAx8fHywsVP5JpEUpf14xBS0IgmBw7u7u9O/fHz8/P5KTk3UXFfUG985w6TfINa1phaioKI4ePao/IirxIqRcFhFRBaRVq1Y0b96c77//nuzsbN1F5XpDkapw8RuTeiOSNwXt5eXFwIEDdRdlp0Lw11CyrYh7EYTnVaQ4dB0Ngavh7i1AucPC3t6eo0ePGri5l3f27FmKFi2Kh4fHiz1Bbi7sXg71OkDJ8vnb3Msa/jmUqgTzpkGOyvd+wbTlZCsXGup3hPJehu5GMDWdhinLTHcshm2mOehiaOIQWnhC7969cXZ2xs9Pz6RzpaEgWZrFgsJdu3aRlpam/83+zY2ADOXFm31BEARjMHXqVJKTk1m2bJl6kddUZZIzL07JRGzevBlZlp/yc8lf+VjuNS1aKnQkSeL999/n6tWrrF2rEj9mYQne70HcSYjeq22DL2HJkiUEBQXpn4IO+QHSo6DOZ9o2JwjmwncayLmwcQ4A1tbWNGrUiLCwMO7du2fY3l5CYmIiV65coV69eupLc58m6CBERSpRCMbGxg7G/whXg2DrAkN3IxSEo5vh7k147Q1DdyKYqmGfQaMuMO9NJcJHeC7iEFp4goODA6+//jp//fUXSUkqywftS0HZVyFyuclNmP3Xhg0bKF68OG3atFEvurkBnCpD0ZdYviEIgiDkmyZNmtC0aVNmz56tHh9VppvyvTtczxJDI+Tv74+Hhwe1a9dWL7rpDy7NwL60Zn0VNq+++io1a9Zk+vTpXLt2TXeRx1CwLwPHx0PqbW0bfAELFy5k3LhxtGrVigEDBjxZIMsQ9CVcmAkV+kHJ1to3KQjmoLQHtOwNW+dDWgqg/NyytLQ06Wnos2fPAspk9wvbuRQcikBLIx3uaekLddvBso8hKc7Q3Qj5bfNccCsPzV41dCeCqbKwgP9bDk7F4csBD7/HC89GHEILOo0aNYrU1FT16R9QFhSmR8Pt7do1ls+ys7PZvHkzr776qvpm+MxEZcKpvK/IRBQEQTAiU6dOJTw8nJ07d+oukCyUbOjYwxB3StvmXlBycjJ79uzBx8dHfcrs/g1l+ra8j6a9FTYWFhasXLmSlJQU2rdvz61bt54ssrSFVn8rr4f2doT0WO0bfQayLPPVV18xbtw4unbtyo4dO7C0tPxPUS6cfgvOfwSVBkOLVYZpVhDMRZ+3ISUBApQ7R52cnKhTpw7nzp3j/v37hu3tBciyzLlz5/Dw8KBYsWIv9iRpKXBgHbTuD3YO+dpfvpEkmDgL7ifAik8N3IyQr66HKnm+PcaDpcqdQILwLIqXhA/+gFvhMGeyobsxKeIQWtCpSZMm1KhRQ38kR5luYFfKpBcUHjx4kLi4OHx99VyJv71VmfYWuZuCIAhGpW/fvri7u/Prr3omnT1HgpUjhM3WrrGXEBAQQEZGxjNERCF+LmmgXr167Nixg9jYWDp06EB0dPSTRa7NoM0WuB8JgZ0hM177RvXIzc3l7bff5sMPP2TIkCH4+/vj4PCfw5/cLDg6HMJ+Aa83oflysFC5OC8IwrOp0Vz5tWEW5OQA0Lx5c7Kzszl58qRhe3sB165dIz4+/uWmoA+th/T7xhnF8SjPOtBjAmz+Da4GG7obIb9smQdW1tBtjKE7EcxB3bYw+BPYtQx26okHFB4jDqEFnSRJYtSoURw9epSQEJXlgxbW4DEMbm1RFveZoA0bNmBnZ0fnzp3Vi276K4ftrs0060sQBEF4OhsbGyZOnMj27dsJDw9XKSoKHiPg2mpIj9G0vxfh7++Pq6srLVq0UC+66Q/O3uBcTbO+CrOmTZuydetWbty4QadOnXTnuZZqA638lYWRgV0hSyXOTGNZWVkMHz6cWbNm8eabb7Js2bIn7/zKToUDPnB1JdT9Ehr8rNxFIAjCy+vzDtyJgCP+ALi5uVG1alVOnDihvvTUSJ05cwZbW1u8vb1f/El2LoUylaFmy3zrq8AM/xwcnWH+Wya1fFZQkZai/P1r1Q+KlzJ0N4K5GPSRchg9e5IyaS88lXiFKagaMmQIVlZWLFmiZ/mg50iQc5Q3LiZGlmX8/f3p0qULjo6Ouoty0pW4kXK9xBsyQRAEIzRu3Disra2ZM2eOelG1yZCbCZeNe4t1VlYWW7ZsoWfPnuoL4zLjIWYflPPRsrVCr1WrVmzatInw8HC6dOlCQkLCk0VlusAra5Xol32vKoe7BpSamoqvry8rV67kiy++4Oeff8bC4j+vZTLiYG8nuLMDmvwONWeI6DFByE8tfKC0J/z948OHmjdvzv379zl/3nQWWqWnp3Px4kVq1aqlHmH4NNHX4FwgdBxuGt9nnF1g6Gdwehcc3WToboSXtWcVpCaJhYRC/rK0hPdXKfFCX/aHjDRDd2T0xKmaoKpUqVK8+uqrLF++nKwsleWDRb2VxUhX/EzuCvGpU6e4ceOG/iiOqN2QnSJueRYEQTBS7u7uDBgwgCVLlpCYmKi7qGh1KN0FLs2FnExtG3wO+/fvJzExUX8Ux60tysVfcQituQ4dOrB+/XrOnz9P9+7dSU5OfrKoXC8lS/nuYWW6OCdd8z4B4uPj6dy5M9u2bWP+/Pl8+OGHT2aMp96C3W2UfPGWa6HKOIP0KghmzdISfKfBxaPKL6BSpUq4u7tz9OhRZBN5/xQcHEx2djb169d/8SfZtVz52GlY/jSlhVcnQMUa8PvbkJlh6G6EFyXLykLCyvWUiBxByE8uZeC9FRB5QblzQtBLHEILeo0cOZLo6Gh27NihXlR5FCSFwL3j2jWWD/z9/bG0tOTVV/Vsxr2xAaydoVR77RoTBEEQnsu0adNISUlh4UI9k85e0yDtDlxbo1lfzysvq7dTp07qRTf9wb4MuDTSrC/hX927d+fPP//k+PHj9OzZk9RUHdPOFQdA08UQtQsO9VfyljV0+/Zt2rRpw4kTJ1i7di3jx49/sijpEuxqCfevQtvtUKGPpj0KQqHSZSQ4FYNV/wNZRpIkmjdvzt27d7l06ZKhu3smZ86coWTJkpQpU+bFnkCWYfcyqNceSlXM3+YKkpU1TJilRKpsmGXoboQXFXRIOSB87Q3TmMIXTE/jrjBgOmz9Hfb9aehujJo4hBb06tatG6VKldK/oLDiALC0N6kFhbIss2HDBlq3bo2Li4vuotwcuLUJyvQASxttGxQEQRCeWcOGDWnbti2//PKL+p07pbtA0VoQ+qNR3rnzaESUvb297qLsNLi9Q5mCFhFRBuPr68uKFSs4cOAAvr6+pKfrmHb2HAGN5sKtzXBkMORqk/164cIFWrZsSUREBFu3bqVv375PFsUeUQ6gs+9Dh0BwFxfaBaFA2TvBoI/hxHY4sA6AmjVr4uzszNGjRw3c3NPFxMRw69Yt6tWr9+QdFc8q+DDcvmL8Cwl1adgJmr8Gf3wB9+4YuhvhRWyaq1wIajfI0J0I5mz4/5RJ+1ljle93gk7iHYygl7W1NcOGDWPLli26N8KDMilcoR9cXa28oTEBZ86c4eLFi/Tr10+9KGonZNyFCjrewAmCIAhG5Z133uHmzZusW7dOd4EkQfW3IeE8RO/RtrlncOrUKW7duqU/iiNqN+SkiigOIzBw4ED8/PzYuXMn/fv3JzNTR8xLtUlQ/3u4vg6OjQY5t8D6yc7O5quvvqJhw4akpqYSGBhIx44dHy/KSYcz02F3K7Bygk6HxES9IGjFdypUawRzp0DSPSwtLWnSpAlXr17lzh3jPtg8e/YsFhYW1KlT58WfZOdS5TC+Ze9860tT436E7ExYMsPQnQjP694dOPQ3dB6p5PYKQkGxsoYPVoOlFXw5QET4qBCH0MJTjRw5kuzsbJYvX65e5DkKspPh+t/aNfYS/Pz8sLOzY+DAgepFV/zA1hXK6InrEARBEIxC9+7dqV69Oj/88IN6xmalQWDnDiE/aNvcM8iLiOrRo4d60U1/sC4KJdto1pegbsSIEfz2229s3ryZVq1asXPnzif/7nm/C7U/g8jlsLstRO/L9z5CQkJo0aIFH374IT4+PgQFBdG4cePHi+6dgO0NIOQ7qDwGup8FZ69870UQBBWWVvDWIkiOgwXvAspdPDY2NkY9DZ2Tk8P58+epVq2a+iL3p0m7DwfWQut+YP+Cz2FoZatA77eUw/SwE4buRnge2xdCTjb0nGjoToTCoFRFeGcJXDoFi6cbuhujJA6hhafy9vamZcuWLF68WP2NfcnW4FQFrizStrkXkJ6ezqpVq+jduzfFihVTKboLtzZCpSEiikMQBMEEWFhY8Pbbb3PmzBn27dunu8jSFqpNhjsBkBCkaX9P4+/v/5SIqGwREWWEJk6cyMqVK7lz5w5dunShRYsW7Nix4/HXS7U+hsbzIeUy7GmXb4fROTk5fP/999SvX5+IiAj+/PNP1q5di5ub2yNFGXDuI9jZXBkWaLsDmvyu3MUmCIK2KteF/u8pB5mnd2NnZ0eDBg0ICgpSX6xrYJcuXeL+/fsvt5Dw8AZITYZOI/KtL4MY+CGUcIffphplrJegQ3aWktHbqAuUrWroboTCokUv8H0TtsyD6GuG7sboiENo4ZmMGTOGsLAwDh8+rLtAkqDKGIg9CElh2jb3nPz9/UlISGDUqFHqRVdXKYuEKuupEQRBEIzK0KFDcXNz48cff1QvqjoBLB0g9CftGnuKS5cuERwcrD+K4+4RJSKqvK9mfQnPZvDgwVy6dIn58+dz+/ZtunXrRrNmzdi2bZtyGC1JUHU89LwCDX+F5PAHh9FtIDrwhQ4zwsPDadWqFe+99x7dunUjODiY/v37P14UdxoCGkPwl+AxHLoHQZku+fSnFgThhQz+GMpVg1njIO0+TZs2BeDYsWMGbky3s2fP4uTkRJUqVV78SXYtA3cPqPVK/jVmCA5FYNTXEPIP7P3D0N0Iz+LIRrh3G3q+YehOhMJm9Lcw+4RpLWLViDiEFp5Jv379KFKkCIsW6Zl09hgOkiVcWaxdYy9gyZIlVKxYkXbt2ukukGWIWAwlGkGx2to2JwiCILwwOzs73njjDbZu3UpISIjuIlsX8BwJV1dCmnHkcG7cuBGAXr16qRfd8AcLW2XBomB0bG1tGT9+PJcuXeL3338nOjqaHj160KRJE7Zs2aIcRlvZg9cUeC3iwWH0JdjTHva0febD6NzcXGbNmkXdunUJDQ1l5cqVrF+/nlKlSv1blJMJ5z+FgKaQcQ/abIVmi8GmaEH98QVBeFY2djBtIURFwoqZFCtWjBo1anD69GkyMowrPzQlJYXw8HDq1q2LhcULHhvEXIeze6DTcHjR5zAmHYcp2d4L3oHroYbuRniazXOVQ8Am3Q3diVDY2NiC50vk6JsxM/hJIGjB0dGRQYMGsW7dOvXbxezdoWxPiFymvAEyQtevX2fXrl2MGDFC/cVU/GlIuCCmoAVBEEzQpEmTsLOz46ef9Ew6V5+mxFuEz9GsL338/f2pX78+FSuqTEvIspIH7d4RrIto2pvwfGxsbBg3bhzh4eEsXLiQu3fv0rNnTxo3boy/vz+5ublgaffIYfRsSL7872H0nV2qh9HHjh2jbdu2vPXWW3To0IGgoCAGDx6MJElKgSwrh9k7m0LQZ0oGeo8gKCvefAuCUanTGl6dAOt/hrATtGjRgoyMDE6fPm3ozh5zr+j3egAAKMBJREFU7tw5ZFmmXr16L/4kAUuU700dh+VbXwZlYQHvLlX+++1XIPS4QdsR9Lh0Gs7tg1cngqWlobsRBOEBcQgtPLMxY8aQmprKmjVr1Isqj4H0GLi9RbvGnsOyZcuQZZkRI0aoF13xU94gVtSztFAQBEEwSm5ubgwfPpwVK1YQHR2tu6hIFSjnA5fmQfZ9Tfv7r+joaI4cOaI/iiPhPNyPVHoWTIKNjQ1jxowhPDycxYsXEx8fj6+vL7Vr12bFihVkZWU9OIyeDK9d+fcwOrCzEqFxYz3IuciyzJ49e+jQoQPNmjUjODiYJUuWsHnzZsqUKaN8MTlXmZTf2Uw5zE6Phtb+0HwZ2BQ35P8NgiCoGf0NFHeHn8dQpqQbFStW5NixY+Tk5Bi6MwBkWebs2bOUL18eV1fX53+C3FxY+Tms/Awad4PSHvnfpKFUqgk/HwbHovBeezi509AdCbqs+BSciikXfARBMBriEFp4Zg0bNqRu3br6IzlKdwH7snDZ+BYU5ubmsmTJEjp06EClSpV0F2WnwdU/oHwfsCmmZXuCIAhCPnnrrbfIyMhg7ty56kXe70JmPEQs1awvXTZv3owsy/oPoW/6A5Jyt5FgUqytrRk1ahRhYWGsWrUKCwsLhg0bRtWqVZk7dy5paWmPHEZHQJOFkJkIB/uQvKYiX4yqQtcuHbl48SI//PADV69eZcSIEcr0c24WRCyHbbXhoC+kx0LjecrzlNMT7SIIguE5FoWp8yDiPKz7nubNm5OYmMjFixcN3RkAN2/e5O7duy+2kDAhFj7sBstnQvvB8NG6/G/Q0MpUVg6iy1aFT16FwNWG7kh4VPgp+Gcz9HlH+bcmCILREIfQwjOTJIkxY8Zw8uRJzp49q7vIwkrJ2ryzA+7f0LS/p9m/fz+RkZH6FxLe9IesBPAUURyCIAimysvLi9dee43ffvuN1NRU3UVuLcClmbKgMNdwk2f+/v54eHhQu7aeHQQ3/cGtJdiXUq8RjJqVlRWDBg3i3LlzbNq0iTJlyjB58mQqVarEt99+S1JSEljakl1pBGtSPuVd//JcuXqTjztFELe0BNd3/x/vvDmRIkWKKBfMw+bA5qrwz4N9HC1WQc/wB4s37Qz9xxUE4Vk0fw1a94OVn1PNXsbNzY29e/cqd0oY2OHDh7G1taVGjRrP94nBR2BSfTi/H978Hd5bDvaOBdOkoZVwhx/2QY0W8PUg2PCroTsS8qz4FIoUB5+phu5EEIT/EIfQwnMZPHgwtra2LF6sZ/lg5VGADBFLNOvrWfj5+VG0aFF8fX3ViyL8wLESlGqrVVuCIAhCAXjnnXe4d+8ey5cvVy/yfgdSIuDWRu0ae0RycjK7d+/Gx8fn31zf/0q5CvFnRRSHmbCwsKBnz54cPnyYffv2Ua9ePd5//30qVKjAG2+8QfXq1Rk4eAjbg4sQVH45Oa02U6RUDazPvQMbK8HxCbCxIpyaotx51mYzdDun5D9bWBn6jycIwvOa9CvYOSDNGke3rl1JSEjg8OHDBm3p+vXrhIWF0bJlS2xtbZ/tk2QZ/voJ3m0D1rYw6wj0GAdqP9vMhWNR+GoHtPSFeW/C0o+facmsUIDCT8KxLQ+moJ0N3Y0gCP8hDqGF51K8eHH69u3LypUrlVtIdXHyUJYnXVls0OmyRyUmJvL3338zcOBA7O3tdRelXIWoPcoktyT+aQiCIJiyVq1a0ahRI3766SdlGZwu5XzB0QNCftS2uQcCAgLIyMh4higORLyCmZEkiTZt2hAQEMDJkyfp1KkT8+bNo1ixYqxfv54LFy4wZMhQLMu/Cp0OQseD4NIYLv8OJRpCx/3Q6RCUfdX8D3kEwZyVcIdxP0LQQTwu7qJmzZocPnyY+Ph4g7QjyzK7d+/GycmJpk2bPtsnpSTA531gwTvQ9FWYewqqNijQPo2KjZ0SOdJtLPzxBfwyAYwk27tQWvEpFCkBvaYYuhNBEHR4ppM2SZKKSZL0lyRJoZIkhUiS1FySpE8lSbolSdLZB7+6P1L/gSRJlyVJCpMkqcsjjzeUJOnCg9/7VVId+xGM2ZgxY0hISGD9+vXqRZXHQOp1iN6jXWN6/Pnnn6SlpemP4ohcpnz0HK5NU4IgCEKBkSSJd955h0uXLrF582bdRRaWUP0tuHsEYo9q2yBKFIerqystWrRQL7rpD0VrKcsUBbPUsGFD1q1bR3JyMidOnMDX1xcLi/+8RC/5CrTdCgPSod12KNlaHD4LgrnoPALqd4BF79G5UR0kSSIgIMAgrYSFhXHjxg3atm2LjY3N0z/h8hl4o6GSvzvuR5i5XlkGV9hYWsK032Hgh7BtAXzZHzLTDd1V4RN6HI5thb5iCloQjNWzjnv+AuyQZbk6UBcIefD4z7Is13vwaxuAJEk1gNeBmkBX4DdJkiwf1M8DxgFVH/zqmj9/DEFLbdq0oXLlyvoXFJbzAZsScMU4FhT6+flRq1YtGjVqpLtAzlXiQ9w7gmNFbZsTBEEQCkTfvn2pUKECP/6oZ9LZcyRYF4NQbaehs7Ky2LJlCz179sTKSiVGIf0uxB6E8npipASz4ejoqB7LksfyGW+NFwTBdEgSTFsAOVk4v9+GNqUdCQsL49KlS5q2kZuby549e3BxcXm2hYTbF8ObzSErQ8lG7vt24b44Jkkw8guY+AscWg8ze0FOtqG7KlxWfiamoAXByD31EFqSJGegNbAYQJblTFmWE/R8Si9gjSzLGbIsRwKXgSaSJJUGnGVZPirLsgwsB3xesn/BACRJYvTo0ezbt0/9xZGlLXgMUya40mM17e+/goODOXbsGKNGjVJ/cxe9F+5fEwsJBUEQzIiVlRXTpk3j4MGDHD9+XHeRtZOyzO3mBiUfWiP79+8nMTFRfxTH7S3KRVKRBy0IgmDeSnvCD/uhZEWarXkHl+wUtm/cQHa2doeYZ8+e5e7du3To0OHJuzEelZkOP4+Fn8dArVbw2xmo2VKzPo2e71TlosKpnbDyc0N3U3iEHIPj26Dvu+BQxNDdCIKg4lkmoT2BWGCJJElnJElaJElS3orbyZIknZckyU+SpOIPHisL3Hjk828+eKzsg//+7+NPkCRpnCRJJyVJOhkba9gDTEG34cOHY2lpiZ+fn3pR5dGQmwWRK7RrTIclS5ZgZWXFkCFD1Iuu+IFNcSjvo1lfgiAIQsEbPXo0zs7O+qehq00ByRJCZ2nWl7+/Pw4ODnTq1Em96MYGcCgPxZ9hIk0QBEEwbV6NYdYRLKcvp9udg8TfT+PIV1PhTmSBf+msrCz27dtHuXLlqF69unphzHV4uxVsXwQDZyhL+Yq5FXh/Jqf7WOgyUsmIPr3b0N0UDis/A2cX6DXZ0J0IgqDHsxxCWwENgHmyLNcH7gPvo0RrVAbqAXeAvHd3ukZNZT2PP/mgLC+QZbmRLMuN3NzEDzVjVKZMGXr06MHSpUvJysrSXVSsFrg0UyI5DLQlOCsrixUrVtCzZ09U/y5lxsON9VBpMFjaadugIAiCUKCcnZ0ZP348f/31F1evXtVd5FAGKg6CCD/IiCvwnmRZxt/fny5duqgvy82+D1E7lSnownx7syAIQmFiYQHtB1F5zl68HeFgdgkSJjUFvxmQmlxgX/bYsWMkJyfTsWNH9TtHz+xR8p9vhsHMDTDySyULWdBt0myo4A3fDIZ7dwzdjXkL+QdObId+/yemoAXByD3LIfRN4KYsy8ce/O+/gAayLEfLspwjy3IusBBo8kh9+Uc+vxxw+8Hj5XQ8LpioMWPGEBUVxbZt29SLqoyBpBC4q/3CJ4Bt27YRExOjfyHh1dWQmyGiOARBEMzU1KlTsbCwYNasWepF1d9WDn4vzy/wfpYuXcqtW7fo3bu3elHYr5CTDuX11AiCIAjmyc6BLmOnIdnas7PBcFjzNYyqBjv88j1nODU1lUOHDlGtWjUqVtSxG0eWYe138EFnKFYSZp+Alj752oNZsneEj9ZBWgp8Mwhycgzdkfla8SkUdYXX3jB0J4IgPMVTD6FlWY4CbkiS5PXgoQ7AxQcZz3l8gaAH/70JeF2SJFtJkjxQFhAel2X5DpAsSVIzSbm8OgzYmF9/EEF73bp1o3Tp0voXFFYYAFZOcGWxdo09ws/PD3d3d7p21bMDM8IPiteDEuJ2Z0EQBHNUrlw5Bg0axG+//caBAwd0FxWvA2V6QNAXEHeqwHoJCwtj8uTJtG3bloEDB+ouuvsPnP8YKvSDkm0KrBdBEATBeBUtWpRWrVsTkuvElQ+2QKlK8NNo6F8Kvh8BRzdBRtpLf52DBw+SmZlJhw4dnvzN1GT4Xz9YNB1e6QO/HoPyXk/WCbpVrAFT5sK5fbBK5EMXiItH4WSAMgVt72TobgRBeIpnmYQGmAKskiTpPEr8xlfAd5IkXXjwWDvgLQBZloOBtcBFYAfwhizLeZf9JgKLUJYVXgG259OfQzAAKysrRo4cybZt27h165buImsnqPg6XFsDWUma9hcVFcXWrVsZPnw4VlZWuovizymHDWIKWhAEwazNmjULDw8PevfuTUSEygLCZn5g6wb7X4PU/L9ZKyMjg9dffx17e3tWrlyJpa7bmDMT4PBAcCgHTRaIKA5BEIRCrHnz5pQoUYLtwZFk/3AAPt8MTXvA0Y0wsxf0c4PP+8LeP+B+4nM/f0JCAidOnKBu3bqULFny8d+8HgpTmsARfxj3A3z4pzjkexGdR0Cn4bDqf0qkiZC/8qage4opaEEwBZJsoKzeZ9WoUSP55MmThm5DUHHlyhWqVKnCF198wYcffqi76O4x2NlMeTNdZaxmvX3//fe89957hIaG4uWlcsX+1P+3d+fhURXpHse/b9gSkCUEJCDIMjKyKERUQCAqSQA3BrxXmBFk0IugjsLgVcdlRkQdr4yOGy74iKIiMOA4yiguAQK4AEFBFgniJEggrAIS1mAkqfvHaTSSzgbd6U7z+zzPebpzuk5VnTz9ptJv16kzFjInwdXboFZcpfVNREQqX2ZmJt26dSM+Pp6lS5dSv3794oX2roF5PaFeO0j5GKrXDlj7Y8eO5ZlnnuHdd9+lf//+xQs4B4t/Bzn/gpRPofFFAWtbRESqpqysLKZPn05ycjK9evXydh790Ztdu/gdWPIOfL8DqteAhCTodhXUqg0/HPa2I4d/fn7s5/w8KCxkdlQrMqwht+Wvoj4/eOPQsW19OtSM8ZLPCb1D+juo8vIOwegL4cD3MGkVNIwPdY8iQ8YSuL0n3PgYDL4r1L0RkSLMbIVz7oJi+5WElpOVnJzMxo0bycrKIirKz+R65+CDTt4H+X7Lir8eBM45OnToQMOGDVm8eLH/QgU/wOwzoEky9JpVKf0SEZHQWrhwIX379iU5OZk5c+b4v1Jmy3vwyQA48xroOROsvBeOlWzOnDn079+fMWPG8Mwzz/gvlPUyfD4SOv8fdLz3pNsUEZHIMGvWLDZs2MCtt95a/AvUwkJYv8xLSC9+G7ZtKF5BrRgvMX1sqxnNzpoNeLHhpfTI20Cfw98A5l19c2xr2AxueRoaNy9en1RcdoaXiG7XHSbM000dA+GevrBhFUzd6K3BLSJhQ0loCZp//OMfDBkyhPnz5/tfSwxg/TPw5Vi4fLW37maQLV26lB49evDyyy8zYsQI/4U2vwWfDYJLP4Jm/YLeJxERCQ+TJ09m1KhRjB49mokTJ/ov9PXfYeVdcM4D0Gn8SbW3detWOnfuTPPmzUlPTyc6Orp4oX3r4KMLoFEPSJobkMS3iIhEhtzcXJ5//nnOPvtsrrnmmpILOgffbfaSyEUSzviZKDRjxgxycnIYM2YMMTExQey9/CT1VXjif+C6B+D340Pdm6otYzHc3gtGPg6D7gx1b0TkOCUlofUJR07a1VdfTWxsbOk3KGx9HUTVrLQbFL7yyivUrl2bwYMHl1xowyvempvxKZXSJxERCQ8jR47k9ttv59lnn2XSpEn+C7W7A9rcAGsfhOyZJ9xWQUEBw4YNIy8vj1mzZvlPQB/N85bhqH4a9HhDCWgREfmFBg0a0KtXLzIyMli+fDklTiQzgyYt4fQzvXVyo2v7TUBnZ2eTmZlJYmKiEtCVqe/1kPJ77yaFKxeEujdVl3Mw9QFocDpcdUuoeyMiFaBPOXLSoqOjGTZsGG+//TabNm3yX6hWHLT4L8h+Aw5vCWp/Zs+ezZQpUxg+fDh169b1XyhrMmz/CM66CaJ0KZSIyKnm8ccf58orr2T06NHMnz+/eAEzuHASNE6EZTfA7s9PqJ0JEyawcOFCnnvuuZLvT7DyDsj9Ci6aCjFNT6gdERGJbD179qRly5a8//77TJ06lT179pxQPc455s2bR7169ejatWuAeymlMoPRz0Pzs2HCEG8tb6m4RbO8mzxe+2ctwyFSxWg5DgmIzMxMLrzwQmJjY1mwYAGtW7cuXmjvapiXCNGNIXkh1Dkz4P344osvuOSSS+jUqRMLFy70/83+9rmw6AqI7wOXvAdRftYDFRGRiLd//3569uzJli1bSE9P958kPrILUrtCwRG47AvvCppyWrJkCRdffDGDBg1ixowZmFnxQjlvw6f/7c287vL3kzgbERGJdM45VqxYwfz58zl69CiJiYn06tWLauVYX9g5R1ZWFkuXLmXjxo0MGDCAhISE4Hdaitu4FsZ0hfjWMHYydOwR6h5VHft2w43tIb4NPL1Ea2uLhCmtCS1Bt2LFCvr27Uvt2rVZsGABbdu2LV5o9zJY2A9qxnqJ6NNaBaz97OxsunfvTkxMDOnp6TRp0qR4ob1rYF4vOK0N9PkUapQwU1pERE4J2dnZdO3alfr165Oenk5cXFzxQrlrYW4PqHuWN3ZUL3vWTW5uLgkJCURFRbFy5criN5ICOLQZPujsq3cxVKsZgDMSEZFId+DAAVJTU8nIyKBRo0b079+fM8/0P8EnPz+f1atXs2zZMvbs2UPdunXp1q0bPXr08P/lqFSO5XPhyRGwewtcPhJGTIB6DUPdq/D36FD49J/w/JfQ+pxQ90ZESqAktFSK1atXk5KSQo0aNUhLS6N9+/bFC32/Ahb08da+TF4IdX910u3m5ubSs2dPtm7dytKlS/23e3gbzO0GrhD6LavQbDYREYlcixcvJikpiR49epCamkrNmn6SwVs/gE/6Q/OB0Oufpa7b7Jzjt7/9Le+88w6fffYZ3bp1K16o8CikXepdJXT5Si8RLSIiUgGZmZm8//777Nu3jy5dupCSkvLTlaD79u3j888/58svv+TIkSM0a9aM7t2706FDh3LNnJZKkHcQ3hgPbz8NdRvCTU9A8nXesh1SXPocGNcfho2HYQ+EujciUgoloaXSrFu3juTkZAoKCkhLS+Pcc88tXmjvKliQAlG1vER0vV+fcHs//vgjV1xxBYsWLSI1NZWkpCQ/hQ7C/IvhQKY3iy024YTbExGRyDNt2jSGDRvGwIEDmTRpEvHx8cULrX8KvvxfaDUUujzlLS/lx+TJkxk1ahQTJkzg7rvv9t/gmnGw9mG4aBq0HhrAMxERkVNJfn4+ixYtIj09ndq1a5OYmEhOTg7r1q0DoH379nTv3p3mzZtr5nO42rAaJt4MX6dD50th9CQ4s12oexVeDu2DkR3htFh4fgXU0NVjIuFMSWipVP/5z39ISkoiLy+PefPm0aVLl+KFcr+CtGSwapC8AOr7mb1cBuccN954I1OmTOG1115j+PDhxQsVHoVPBno3IrzkPWh2ecVPSEREIt4TTzzBvffeS3R0NA888ACjR4/+5axo5+Cr8ZDxf97VPJ3/6rvBrXdvgU2bNvHQQw/x+uuv07t3b1JTU4mKOm7G9L51Xh2b/wltrofur1bW6YmISATbsWMH7733Htu2baNWrVp06dKFrl270qBBg1B3TcqjsBA+fBleuRuOHILBd8O190EtP/c4OhU9czN8OBmeXgrtdENNkXCnJLRUum+//ZakpCT27dtHamqq/7sv71sHaUmAg6Q0aFCxdZ0effRR7rvvPu6//34eeuih4gWcg+W3QeYLcOEkaHvziZ2MiIicEjIzMxk7diwffPAB7dq1Y+LEifTp0+eXhfatg+VjYGcaNOjM7lbjGf/CfF566SWioqK45ZZbGDduHLGxsT8fsz8T1j4I2TO8NaXPHgsd74Pq+nApIiKBUVhYyNatW2nSpIn/paUk/O39DibfCfPfgKZt4PaXIaF3qHsVWqsXwV294Zo7YJRu4ixSFSgJLSGxadMmkpKS2LVrFx9++CE9e/YsXmj/N14iujAfkuZDbOdy1T1z5kyuvfZahgwZwrRp0/xfXvb1k7DyDmh/F5z32EmejYiInCrmzJnD2LFj2bBhA1dffTVPPvkkrVq1+rmAc+zPeJWC5WOJrXmA6YuN1dWGMuZPj9K8eZF7DhzMhrUPwcapEFUTfj3aG5OiG1X2KYmIiEhVsXIBPHsLbM2EoeNg6P1wKq7lfeQw3NzJe/7iGoiuHdr+iEi5lJSELvmuOiIB0LJlSz755BOaNm1Kv379WLRoUfFC9c6GlI+hWrSXjP5+RZn1Ll68mOuvv57ExESmTJniPwG9+V+w8k5ocQ0kTDj5kxERkVPGVVddxdq1a3nkkUdITU2lffv2jB8/nry8PHJzc7l/3DjOuOiPnPmHg7y74VyG9KrBY71m0/zATCjIh8Nb4PNb4L223uznX4+G32yE8/6mBLSIiIiU7rwkb+3j5GEw7UG4pw98vyPUvap8U8fBtg3ejHAloEWqPM2ElkqxY8cOkpOTWb9+PSkpKVx33XUMHDiQunXr/lzo4EZI6+19cI/v6934qcVA77LlIr766it69+5Nw4YNWbp0KXFxcb9szDnY9Sks7AcNErz1pnW5s4iInKCcnBzuuusuZs2aRYsWLThw4AC5ubkMGjSIBx98kPbt28OBLFhxO2ybA3VaQd52oBB+NdJbdqP2GaE+DREREamK5r4Gz/4BateDe6bDecmh7lHlWP85jL0ILh8Jf3wx1L0RkQrQchwScnv27OHJJ59kxowZZGdnExMTw4ABAxg6dCj9+vWjRo0a3of2b56F7OlweDNUr8ORRpez7Lu2zFy4i/kLFpGVlUVcXBzp6emcddZZXuWHNsPOhbBzgfd4OAdOawN90yG6cWhPXEREIsLHH3/MX/7yF+Li4hg/fjwJCQnFC22d4924sH4HOOd+qNOy0vspIiIiESY7A/46CHLWnxrLc/yYD7eeDwf3wuQMqFM/1D0SkQpQElrChnOOJUuWMH36dN5880327NlDXFwcgwcPZsiQIXTs2JFPP/mY7C+m0/SHNFJ+vZfYOrBzn7FsRysONh7IxZddS/PqmT8nng9+61VeqxGcfik06Q1nXgPRp4f0XEVEREREREROWt4hb0b0/KnQuTfcOwMaxoe6V8HxxoPwxnh4eA50uzLUvRGRClISWsJSfn4+c+fOZfr06fz73/8mLy/vp9diYmLo1asXfZIuZmDXmvyq+jKits3xbmB4TI36cPol0CTJSzw3OAdMS52LiIiIiIhIBIr05Tk2roVbu8DFg+GeaaHujYicACWhJewdPHiQ2bNns2nTJhITE+nWrRu1atX6ZaH8vZDzNuTnQpNLvTWfoyL4MiQRERERERGRoiJ1eY5tG2Bcf9i3G15eB/V1M2eRqkhJaBERERERERGRSFB0eY7z+3qzoqty0jZ9DvztOoiKgvv/BQm9Q90jETlBJSWhtW6BiIiIiIiIiEhVElMH7noNxr4Eaz6GW86Dr9ND3auKKyiAqQ94M6CbtoHnVigBLRKhlIQWEREREREREalqzOCKkfD0EqheA/43Ed6ZCGF+xftP9n8P918F0x6CvtfDU4uhaetQ90pEgkRJaBERERERERGRqqptF3h+BXS9Aib9ER75LRzaH+pelS5rJdx6PqxKgzEvwh1ToFZMqHslIkGkJLSIiIiIiIiISFVWNxbGz4Yb/wafvQ2jL4SNX4W6V/7NfR3G9oCCo/DEp3DVTd6sbhGJaEpCi4iIiIiIiIhUdWYw+E/wWBoc3g9jusG8qeGzPEf+DzDxD/D366H9Rd7s7fbdQt0rEakkSkKLiIiIiIiIiESKTpfACyuhXTd4fDiMOhdmToDvNoemP9/lwLSHYUQ7mDPJS5RPmAuxp4emPyISEubC5RuxElxwwQVu+fLloe6GiIiIiIiIiEjVUXAUPpoC86dCxmJv37kXQ/J1kHiNt4RHsPyYD0vfhY9egRWp3mzshCS45g5v7WoRiVhmtsI5d0Gx/UpCi4iIiIiIiIhEsO0bYeEMSJsGOeuhRk3oeiUkDYVuV0LN6MC0k53hJZ7T3oB9u6HRGdD3Buh3AzRtE5g2RCSsKQktIiIiIiIiInIqcw4yv4QF02HRP+D7HRBdB85o6yWJ41tDvO+xaRto0vKXCerCQjjwvXdc7k7vce8O2LsT1n4KX6dD9RrQ/Tdw2Qg4vy9Uqxa68xWRSqcktIiIiIiIiIiIeAoKYNUCSH8Ptm+A7d/CzmzIP/LLcnHNoH5j2L/bSzYXHC1eV41a0OJsSBkOKcOgQeNKOQURCT8lJaGrh6IzIiIiIiIiIiISQtWqwfl9vO2YwkJvZvOOjV5S+tjj/t3QtgvExkNsE++xYfzPj7XrgVnozkVEwp6S0CIiIiIiIiIiAlFR3sznuGbQsWeoeyMiESQq1B0QERERERERERERkcilJLSIiIiIiIiIiIiIBI2S0CIiIiIiIiIiIiISNEpCi4iIiIiIiIiIiEjQKAktIiIiIiIiIiIiIkGjJLSIiIiIiIiIiIiIBI2S0CIiIiIiIiIiIiISNEpCi4iIiIiIiIiIiEjQlCsJbWYNzOwtM1tvZl+b2UVm1tDM5plZpu8xtkj5e80sy8y+MbN+Rfafb2Zf+V6baGYWjJMSERERERERERERkfBQ3pnQzwAfOefaAZ2Br4F7gDTnXFsgzfczZtYB+B3QEbgMeMHMqvnqmQSMAtr6tssCdB4iIiIiIiIiIiIiEobKTEKbWT3gYuAVAOdcvnMuFxgAvO4r9jow0Pd8ADDTOfeDc24jkAV0NbOmQD3n3FLnnAOmFjlGRERERERERERERCJQeWZCtwF2Aa+a2Uoze9nM6gBNnHPbAXyPp/vKnwHkFDl+i2/fGb7nx+8XERERERERERERkQhVniR0daALMMk5dx5wCN/SGyXwt86zK2V/8QrMRpnZcjNbvmvXrnJ0UURERERERERERETCUfVylNkCbHHOLfP9/BZeEnqnmTV1zm33LbXxXZHyLYoc3xzY5tvf3M/+YpxzLwEvAZjZLjPbVM7zqUoaAbtD3QmRCKKYEgksxZRIYCmmRAJLMSUSOIonkcA61WOqpb+dZSahnXM7zCzHzM52zn0DJAPrfNtwYILv8d++Q94FZpjZk0AzvBsQfu6cKzCzA2bWHVgG/B54thztNy7z1KogM1vunLsg1P0QiRSKKZHAUkyJBJZiSiSwFFMigaN4EgksxZR/5ZkJDTAamG5mNYFvgRvwlvJ408xGAJuBQQDOuQwzexMvSX0UuNU5V+Cr5xbgNSAG+NC3iYiIiIiIiIiIiEiEKlcS2jm3CvCXwU8uofwjwCN+9i8HzqlA/0RERERERERERESkCivPjQklOF4KdQdEIoxiSiSwFFMigaWYEgksxZRI4CieRAJLMeWHOedC3QcRERERERERERERiVCaCS0iIiIiIiIiIiIiQaMkdDmY2WVm9o2ZZZnZPX5ev9PMnJk1KuH4x81svZmtMbN3zKyBb38NM3vdzL4ys6/N7N4Sjm9tZsvMLNPMZvluEIl5Jvr6tcbMugTwtEWCJgxi6jZf279ow8yG+upcY2ZLzKxzgE5ZJGiCGE81zexVXzytNrNLSzheY5RElDCIKY1RElECEFMP+973q8xsrpk1K/Lavb56vzGzfiUcr3FKIkoYxJTGKYkoJcWUmY03s62+WFllZleUcHxDM5vnG2fmmVmsb3+cmS00s4Nm9lwp7Z8y45SS0GUws2rA88DlQAfgWjPrUOT1FkAfYHMp1cwDznHOdQL+AxxLjA0CajnnzgXOB24ys1Z+jv8b8JRzri2wFxjh23850Na3jQImncg5ilSmMImpxUAKsOm4/RuBS3z1PozWcZIwF+R4Ggngi6c+wBNm5u//Bo1REjHCJKY0RknECFBMPe6c6+ScSwDmAON8x3YAfgd0BC4DXvC1dzyNUxIxwiSmNE5JxCgrpvDGjwTf9kEJ1dwDpPnGmTTfzwBHgPuBO8voxikzTikJXbauQJZz7lvnXD4wExhQ5PWngD8BJS6u7Zyb65w76vsxHWh+7CWgjplVB2KAfGB/0WPNzIAk4C3frteBgb7nA4CpzpMONDCzpid0liKVJ6Qx5Tt+pXMu28/+Jc65vX7qFQlXwYynDnj/ROGc+w7IBS4oeqzGKIlAIY0p32saoySSBCKmiv4vV6dI2QHATOfcD865jUCWr72faJySCBTSmPIdr3FKIklZMVUeA/DGFygyzjjnDjnnPsNLRvt1qo1TSkKX7Qwgp8jPW3z7MLPfAFudc6srUN//AB/6nr8FHAK2431T+Xfn3PfHlY8Dcot8mPmp/dL6JhLGQh1T5TWiSL0i4SqY8bQaGGBm1c2sNd7VBS2OK68xSiJNqGOqvDRGSVURkJgys0fMLAcYim/WZml1F6FxSiJNqGOqvDROSVVR1vv+Nt9SGFOOLbPhRxPn3HYA3+PpFWj/lBqnlIQum/nZ58ysNvBnfv6DXXZFZn8GjgLTfbu6AgVAM6A1cIeZtSlP++V4TSRchTqmylNvb7x/nO6u6LEilSyY8TQF7x+d5cDTwBLf62W2X47XRMJVqGOqPPVqjJKqJCAx5Zz7s3OuBV483VZa3eVpvwLHi4SbUMdU2R3UOCVVS2nv+0nAr4AEvIluT1Ry+xE3TikJXbYt/HKWSnNgG94bsTWw2syyffu/NLN48246s8rMflovxsyGA1cBQ51zx940Q4CPnHM/+i7LXEzxyzJ34025r35c+6X1TSSchTqmSmVmnYCXgQHOuT0ndIYilSdo8eScO+qcu923/tkAoAGQeVz7GqMk0oQ6pkqlMUqqoIDEVBEzgP8uo+6iNE5JpAl1TJVK45RUQSW+751zO51zBc65QmAyvuVp/MTUzmPLZPgev6tA+6fWOOWc01bKBlQHvsX7g14T71LKjn7KZQONSqjjMmAd0Pi4/XcDr+J9u1HHV6aTn+P/CfzO9/xF4A++51fiXeJiQHfg81D/vrRpK2sLh5gqqQ3gTLy1z3qE+vekTVt5tiDHU22gju95H+CTEo7XGKUtYrZwiKmS2tAYpa0qbgGKqbZFno8G3vI97+irr5av/m+Ban6O1zilLWK2cIipktrQOKWtKm6lxRTQtEi52/HWTPdXx+PAPb7n9wCPHff69cBzpfThlBmnjmXapQTOuaNmdhuQClQDpjjnMipYzXN4f8jneWuOk+6cuxnvDpyvAmvx3lSvOufW+Dn+bmCmmf0VWAm84tv/AXAF3h/6w8ANFeyXSKULh5gyszF4N+yIB9aY2QfOuRvxLl+Lw7sTNMBR51yFZlKLVKYgx9PpQKqZFQJbgWElHK8xSiJGOMSUxiiJJAGKqQlmdjZQCGwCbvbVnWFmb+J96XMUuNU5V+DneI1TEjHCIaY0TkkkKSOmHjOzBLwlMLKBm0qoZgLwppmNwLs31aBjL/iuTKgH1DSzgUBf59y6444/ZcYp82XXRUREREREREREREQCTmtCi4iIiIiIiIiIiEjQKAktIiIiIiIiIiIiIkGjJLSIiIiIiIiIiIiIBI2S0CIiIiIiIiIiIiISNEpCi4iIiIiIiIiIiEjQKAktIiIiIiIiIiIiIkGjJLSIiIiIiIiIiIiIBI2S0CIiIiIiIiIiIiISNP8PITnJBtgapvMAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"pred = fit.forecast(48)\n",
"fig, ax = plt.subplots(figsize=(25, 6))\n",
"ax.plot(train.index[-96:], train.values[-96:],color='black')\n",
"ax.plot(train.index[-96:], fitted[-96:], color='orange')\n",
"ax.plot(test.index, pred,color='orangered')\n",
"ax.plot(test.index, test.values, color='grey')\n",
"plt.title(\"Simple Exponential Smoothing (Black/grey = Actual, Orange = Fitted/Forecast)\");"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Evaluating Forecasts\n",
"\n",
"- In the previous example we had test and training data.\n",
"- We fit the model once and made 1-step ahead to 48 step-ahead forecasts.\n",
"- We could evaluate the accuracy of these forecasts using mean square error, mean absolute error or other metrics.\n",
"- Is this the end of the story?\n",
"- Since methods can have very different performance 1-step ahead compared to 48-steps ahead, does it make sense to average across these horizons?"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Your turn\n",
"\n",
"- Consider a case where only 1-step ahead forecasting is needed.\n",
"- Would the above evaluation be the best strategy?\n",
"- If not, what would be a better strategy (*Hint: think of leave one out validation*)?\n",
"- Discuss for 5 minutes (and don't cheat by looking at the next slide)."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Rolling or Window\n",
"\n",
"- A common way to evaluate forecasts is via a rolling window.\n",
" 1. Fit the model from $y_1,...,y_{T_{\\textrm{train}}}$ and make forecast $y_{T_{\\textrm{train}}+1}$ \n",
" 2. Fit the model from $y_2,...,y_{T_{\\textrm{train}}+1}$ and make forecast $y_{T_{\\textrm{train}}+2}$\n",
" 3. Fit the model from $y_3,...,y_{T_{\\textrm{train}}+2}$ and make forecast $y_{T_{\\textrm{train}}+3}$\n",
" 4. and so on...\n",
"- This gives a series of one-step ahead forecasts to evaluate one-step ahead forecast accuracy.\n",
"- We can compute mean square error or mean absolute error over the rolling window."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Expanding Window\n",
"\n",
"- A alternative way to evaluate forecasts is via a expanding window.\n",
" 1. Fit the model from $y_1,...,y_{T_{\\textrm{train}}}$ and make forecast $\\hat{y}_{T_{\\textrm{train}}+1}$ \n",
" 2. Fit the model from $y_{\\color{red}{1}},...,y_{T_{\\textrm{train}}+1}$ and make forecast $y_{T_{\\textrm{train}}+2}$\n",
" 3. Fit the model from $y_{\\color{red}{1}},...,y_{T_{\\textrm{train}}+2}$ and make forecast $\\hat{y}_{T_{\\textrm{train}}+3}$\n",
" 4. and so on...\n",
"- Which do you think is better and why?"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Rolling vs Expanding\n",
"\n",
"- Expanding window mimics the process of using all available data whenever forecasts are made.\n",
" - This is common in practice.\n",
"- Rolling windows use the same amount of training data\n",
" - Rolling windows make sense if model parameters are unstable over time.\n",
"- Common Sense: *Evaluate forecasts in the way you anticipate you will be using forecasts!*\n",
"- Also not, although the expanding/rolling windows described here use one-step ahead forecasts, we can also use multistep ahead forecasts too."
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Think probabilistically\n",
"\n",
"- We often think of forecasts as single number or *point forecasts*, for example \n",
" - Electricity demand at 3PM tomorrow will be 8000 MWh\n",
"- Sometimes decisions may depend of the variance of a forecast, i.e.\n",
"\n",
"$$Var(y_{T+h}|\\mathcal{F}_{T})$$\n",
"\n",
"- Other decisions may depend on quantiles, i.e.\n",
"\n",
"$$q:\\textrm{Pr}(y_{T+h}\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"cell_style": "split",
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"- Rain probability in weather apps.\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "subslide"
}
},
"source": [
"## Confidence vs Prediction Interval\n",
"\n",
"- A *confidence interval* describes the uncertainty around my forecast due to parameter uncertainty\n",
" - e.g. in a regression model uncertainty due to $\\beta$\n",
"- A *prediction interval* also accounts for the uncertainty since the future outcome is unknown\n",
" - e.g. in a regression model it will reflect the uncetainty in $\\epsilon$\n",
"- Prediction intervals are wider than confidence intervals"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Probabilistic Forecasts\n",
"\n",
"- A probabilistic forecast is a density (or probability mass function)\n",
"\n",
"$$f(y_{T+h}|\\mathcal{F}_T)$$\n",
"\n",
"- From this we can get forecast variances, quantiles, prediction intervals\n",
"- We can also get point forecasts either as\n",
" - The expected value $E(y_{T+h}|\\mathcal{F}_T)$\n",
" - The median $q:(y_{T+h}
![\"weather\"](\"figures/weatherf.png\")
\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "subslide"
}
},
"source": [
"## Confidence vs Prediction Interval\n",
"\n",
"- A *confidence interval* describes the uncertainty around my forecast due to parameter uncertainty\n",
" - e.g. in a regression model uncertainty due to $\\beta$\n",
"- A *prediction interval* also accounts for the uncertainty since the future outcome is unknown\n",
" - e.g. in a regression model it will reflect the uncetainty in $\\epsilon$\n",
"- Prediction intervals are wider than confidence intervals"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"## Probabilistic Forecasts\n",
"\n",
"- A probabilistic forecast is a density (or probability mass function)\n",
"\n",
"$$f(y_{T+h}|\\mathcal{F}_T)$$\n",
"\n",
"- From this we can get forecast variances, quantiles, prediction intervals\n",
"- We can also get point forecasts either as\n",
" - The expected value $E(y_{T+h}|\\mathcal{F}_T)$\n",
" - The median $q:(y_{T+h}