Non-Intrusive Load-identification Monitoring(NLM
1 Introduction
2 Solution
3 Hardware and Software
4 Prepare for Non-intrusive Load Monitoring running with LLama2 Model
5 Run the Model
6 Final Result
7 Conclusion and Next to do

Created March 27, 2024

Non-Intrusive Load-identification and Monitoring(NLM)

Non-Intrusive Load-identification and Monitoring(NLM) with LLama Chat and Time Series Large Model on AMD Radeon W7900 Pro

Things used in this project

Hardware components

AMD W7900 Pro

Software apps and online services

Ubuntu 22.04 LTS

ROCm Software 6.1.2 installed with amdgpu-install

pytorch 2.3.1 with ROCM6.1

Hand tools and fabrication machines

Dell Worksation 5820

Story

Non-Intrusive Load-identification Monitoring(NLM)

1 Introduction

Nonintrusive load-monitoring (NLM) is smart energy management by obtaining bus load information from a single measurement point and using algorithms to analyze information about individual devices and customer consumption patterns. This can be very helpful in home automation and reduce cost of energy metering and monitoring for smart meter in each power energy consumer point.

The proposed NLM with AMD GPU's Genetic AI use open sources LLama as process engine instead. Such model shall be made scalable to fit into single GPU of large memory larger than 32GB, such as that of 48G by AMD Radeon™ PRO W7900 .

2 Solution

The AMD GPU's Genetic AI is key feature in this proposal and the high performance and large memory is why choising AMD GPU in this proposal.

First, tokenized the energy measuring data sampled continuously, such tokens can be encoded as input to LLama, the LLama shall be trained with AMD GPU to output new tokens, then new tokens can be decoded as rich-information contend, including energy comsumption prediction, electrical load analysis and predict maintenance indexes. Since the energy data is sampled continuously clock-round, there are large amount data generated for this training and the output shall be envolved after each training epoches. The application of Genetic AI is top different from existing solution and it is very useful since more information can be mined out of raw data. It is preferable to deploy this Genetic AI model locally, to make use of such data mining process so as to protect pricacy.

3 Hardware and Software

3.1 First unbox the AMD Radeon PRO W7900 and Dell 5820 workstation

1 / 3

Insert into the PCIex16 slot

Note: AMD Radeon PRO W7900 is full size video card occupy 3 slot place. It can not fit into normal 1U blade server and normal desktop computer. This Dell 5820 is newly ordered just for this AMD Radeon PRO W7900, while the BIOS is not supported. One normal video card such as Nvidia GT630 shall be used first to install necessary software . Then reload the AMD Radeon PRO W7900 and wait silently with blank screen for splash screen reappear.

3.2 Ubuntu 22.04 LTS and ROCm 6.1.2

The versioning problems can be frustrating to kill times again and again. At last, Ubuntu 22.04 LTS with HWE kernel is stable choice.

First install amd-gpu-install, then install video drivers and ROCm with --usercase choice

1 / 2

Then run rocminfo and rocm-smi to validate installation

1 / 3

3.3 Docker pull for pytorch:rocm and docker build for vllm

Safe start from docker image pull with pytorch in rocm .

Build docker image for environment like vllm works well too.

1 / 4

3.4 Prepare large model from huggingface

It works perfect

4 Prepare for Non-intrusive Load Monitoring running with LLama2 Model

4.1 Prepare environment

## Introduction
This is repository for ulm based on Llama2 7B large model running on AMD RADEON W7900 Pro with ROCM software running on ubuntu 22.04.
## Preparation
### 1 Install AMD graphic drivers with rocm
wget https://repo.radeon.com/amdgpu-install/6.1.2/ubuntu/jammy/amdgpu-install_6.1.60102-1_all.deb
sudo apt install ./amdgpu-install_6.1.60102-1_all.deb
sudo amdgpu-install --usecase=graphics,rocm
### 2 Install necessary python packages with ROCM 6.1 supported
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm6.1
Check availablity of ROCM in python 3.1x
import torch
print ( torch.cuda.is_available() )
### 3 Install flask server
pip install flask
### 4 Downlaod Llama-2-7b-chat-hf from huggingface
pip install hugginface-cli
hugginface login
huggingface-cli download --resume-download NousResearch/Llama-2-7b-chat-hf  --local-dir NousResearch/Llama-2-7b-chat-hf
### 5 Start the flask serve and run ULM
flask --app start run

4.2 Llama Chat Page

This page runs with HTTP Get to energize the Llama chat bot

#Llama 7B chat
import torch
from datasets import load_dataset
from transformers import (
AutoModelForCausalLM,
AutoTokenizer,
BitsAndBytesConfig,
TrainingArguments,
pipeline
)
from peft import LoraConfig
from trl import SFTTrainer
base_model_name = "Llama-2-7b-chat-hf"
# Model
model = AutoModelForCausalLM.from_pretrained(base_model_name, device_map="auto" )
model.config.use_cache = False
model.config.pretraining_tp = 1
llama_tokenizer = AutoTokenizer.from_pretrained(base_model_name, trust_remote_code=True)
llama_tokenizer.pad_token = llama_tokenizer.eos_token
llama_tokenizer.padding_side = "right"
def chatbot(input=None):
#query=input
#text_gen = pipeline(task="text-generation", model=base_model_name, tokenizer=llama_tokenizer, max_length=200)
#output = text_gen(f"<s>[INST] {query} [/INST]")
#return output[0]['generated_text']
if input == None:
return None
query  = llama_tokenizer(input, return_tensors='pt')
query  = query.to('cuda:0')
pred = model.generate(**query, max_new_tokens=256, repetition_penalty=1.1)
return llama_tokenizer.decode(pred.cpu()[0])

Here is the output for interactive chat in python

4.3 ULM view and model training

This is how to use the data for deep learning running on ROCM in cuda

>>> import torch

>>> print(torch.cuda.is_available())

then, run

rocm-smi

Then train time-serial model with pytorch framework, refer to notebook file as attachment.

from __future__ import print_function
import argparse
import torch
import torch.nn as nn
import torch.optim as optim
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

opdata=pd.read_csv('data/opdata.csv'); indata=opdata.drop(0);col=indata.columns
plt.plot(indata['PowerOutput.2.KW'].head(3200).tail(1500))

class Sequence(nn.Module):
def __init__(self):
super(Sequence, self).__init__()
self.lstm1 = nn.LSTMCell(1, 51)
self.lstm2 = nn.LSTMCell(51, 51)
self.linear = nn.Linear(51, 1)
def forward(self, input, future = 0):
outputs = []
h_t = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t = torch.zeros(input.size(0), 51, dtype=torch.double)
h_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
for input_t in input.split(1, dim=1):
h_t, c_t = self.lstm1(input_t, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
output = self.linear(h_t2)
outputs += [output]
for i in range(future):# if we should predict the future
h_t, c_t = self.lstm1(output, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
output = self.linear(h_t2)
outputs += [output]
outputs = torch.cat(outputs, dim=1)
return outputs

steps=15
np.random.seed(0)
torch.manual_seed(0)
# load data and make training set
input = torch.from_numpy(data[3:, :-1])
target = torch.from_numpy(data[3:, 1:])
test_input = torch.from_numpy(data[:3, :-1])
test_target = torch.from_numpy(data[:3, 1:])
# build the model
seq = Sequence()
seq.double()
criterion = nn.MSELoss()
# use LBFGS as optimizer since we can load the whole data to train
optimizer = optim.LBFGS(seq.parameters(), lr=0.8)
#begin to train
for i in range(steps):
print('STEP: ', i)
def closure():
optimizer.zero_grad()
out = seq(input)
loss = criterion(out, target)
print('loss:', loss.item())
loss.backward()
return loss
optimizer.step(closure)
# begin to predict, no need to track gradient here
with torch.no_grad():
future = 1000
pred = seq(test_input, future=future)
loss = criterion(pred[:, :-future], test_target)
print('test loss:', loss.item())
y = pred.detach().numpy()
torch.save(seq,'model_seq.pth')

calculation the loss and predict the output values

5 Run the Model

5.1 Start the Flask and run with chat and ask "what is Nonintrusive Load Monitoring" as test run

flask --app start run

1 / 5

5.2 Run with Make ULM view as flask run in app.py and html templates welcomepage.html

flask run

as of code for app.py

import json
import random
import time
from datetime import datetime
from flask import Flask, Response, request
from flask import stream_with_context, jsonify, render_template
import llama_chat as lc
import nlm
from nlm import Sequence
app = Flask(__name__)
random.seed()  # Initialize the random number generator
@app.route('/')
def home(command=None):
#chatback=lc.chatbot(command);nlmoutput=" <p><a href=\"/\">--- Return for Home Page ---</a></p>"
return render_template('welcomepage.html' , firstL=command)
@app.route("/get/chat", methods=["POST", "GET"])
def get_chat():
if request.method == "POST":
ask_chat = request.form["chat_input"]
response = {'chat_input': "ask_chat"}
print(ask_chat)
response = {'ask':ask_chat, 'answer': lc.chatbot(ask_chat)}
return jsonify(response)
elif request.method == "GET":
ask_chat = request.args.get("chat_input")
response = {'chat_input': ask_chat}
print("response")
response = {'ask':ask_chat, 'answer': lc.chatbot(ask_chat)}
return jsonify(response)
else:
response=ask_chat
return response
@app.route('/ulm-data')
def chart_data():
def generate_ulm_data():
while True:
nlm_value=random.random()*1000
nlm_pre=nlm.update_nlm(nlm_value)
json_data = json.dumps(
{'time': datetime.now().strftime('%Y-%m-%d %H:%M:%S'), 'value': nlm_value, 'value_pre': nlm_pre})
yield f"data:{json_data}\n\n"
time.sleep(10)
pass
response = Response(stream_with_context(generate_ulm_data()), mimetype="text/event-stream")
response.headers["Cache-Control"] = "no-cache"
response.headers["X-Accel-Buffering"] = "no"
return response
if __name__ == "__main__":
app.run(host='127.0.0.1', debug=False)
pass

as of code for nlm.py

#NLM Prediction nlm.py
import torch
import torch.nn as nn
import numpy as np
#from model import Sequence
import random
class Sequence(nn.Module):
def __init__(self):
super(Sequence, self).__init__()
self.lstm1 = nn.LSTMCell(1, 51)
self.lstm2 = nn.LSTMCell(51, 51)
self.linear = nn.Linear(51, 1)
def forward(self, input, future = 0):
outputs = []
h_t = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t = torch.zeros(input.size(0), 51, dtype=torch.double)
h_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
for input_t in input.split(1, dim=1):
h_t, c_t = self.lstm1(input_t, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
output = self.linear(h_t2)
outputs += [output]
for i in range(future):# if we should predict the future
h_t, c_t = self.lstm1(output, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
output = self.linear(h_t2)
outputs += [output]
outputs = torch.cat(outputs, dim=1)
return outputs
def update_nlm(input=None):
future=1
pre_input=torch.from_numpy(np.array([[input], ]).astype('float'))
model=torch.load('model_seq.pth')
pred = model(torch.from_numpy(np.array([[input], ]).astype('float')), future=future).detach().numpy();   print(pred[0][-1])
return pred
if __name__ == "__main__":
print("Start")
update_nlm(133)

and template welcomepage.html with chart.js and ajax

<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Nonintrusive Load Monitoring (NLM) Project for Pervasive AI Developer Contest with AMD</title>
<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.3.1/js/bootstrap.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/Chart.js/2.8.0/Chart.min.js"></script>
<link href="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.3.1/css/bootstrap.min.css" rel="stylesheet">
<link href="https://cdnjs.cloudflare.com/ajax/libs/Chart.js/2.8.0/Chart.min.css" rel="stylesheet">
</head>
<body>
<div class="container" >
<div style="justify-content:center;">
<h2 >LLama2 Chat</h2>
<div class="col">
<form id="chat_form" >
<label>Input question here...<input type="text" id="chat_input" /></label>
<p><input type="reset" value="RESET" /> | <input type="submit" value="submit" /></p>
</form>
<lable> Output the result here...<span id="chat_result"></span></lable>
</div>
</div>
<div class="row">
<div class="col-12">
<div class="card">
<div class="card-body">
<canvas id="monitor_canvas"></canvas>
</div>
</div>
</div>
</div>
<div class="row">
...
</div>
</div>
<script>
$(document).ready(function () {
const config = {
type: 'line',
data: {
labels: [],
datasets: [{
label: "Realtime Dataset",
backgroundColor: 'rgb(255, 99, 132)',
borderColor: 'rgb(255, 99, 132)',
data: [],
fill: false,
},
{
label: "Predict Dataset",
backgroundColor: 'rgb(55, 99, 32)',
borderColor: 'rgb(55, 99, 32)',
data: [],
fill: false,
}],
},
options: {
responsive: true,
title: {
display: true,
text: 'Real-Time Non Instrusive Monitoring'
},
tooltips: {
mode: 'index',
intersect: false,
},
hover: {
mode: 'nearest',
intersect: true
},
scales: {
xAxes: [{
display: true,
scaleLabel: {
display: true,
labelString: 'Time'
}
}],
yAxes: [{
display: true,
scaleLabel: {
display: true,
labelString: 'Value'
}
}]
}
}
};
const context = document.getElementById('monitor_canvas').getContext('2d');
const lineChart = new Chart(context, config);
const source = new EventSource("/ulm-data");
source.onmessage = function (event) {
const data = JSON.parse(event.data);
if (config.data.labels.length === 20) {
config.data.labels.shift();
config.data.datasets[0].data.shift();
config.data.datasets[1].data.shift();
}
config.data.labels.push(data.time);
config.data.datasets[0].data.push(data.value);
config.data.datasets[1].data.push(data.value_pre);
lineChart.update();
}
});
</script>
</body>
</html>
<!--
{% if firstL   %}
<h1> {{ firstL  }}!</h1>
<p> {{ secondL  }}!</p>
<p> {{ ThirdL  }}!</p>
<p><a href="/chat/">--- Go for Llama Chat Page ---</a></p>
<p><a href="/ulm/">--- Go for Nonintrusive Load Monitoring (NLM)  View Page ---</a></p>
<p><a href="/">--- Return for Home Page ---</a></p>
{% else %}
<h1>Hello and welcome to Pervasive AI Developer Contest with AMD!</h1>
<p><a href="/chat/">--- Go for Llama Chat Page ---</a></p>
<p><a href="/ulm/">--- Go for Nonintrusive Load Monitoring (NLM)  View Page ---</a></p>
<p><a href="/">--- Return for Home Page ---</a></p>
{% endif %}
-->

6 Final Result

This shows the potential of running large model on AMD W7900 with perfect result.

The performance of ULM can improved further when more data is gathered and trains accordingly.

7 Conclusion and Next to do

The AMD Radeon PRO W7900 is great in performance and highly adaptive. More data shall be fetched to train better large model.

Code

## Introduction
This is repository for ulm based on Llama2 7B large model running on AMD RADEON W7900 Pro with ROCM software running on ubuntu 22.04.

## Preparation
### 1 Install AMD graphic drivers with rocm
wget https://repo.radeon.com/amdgpu-install/6.1.2/ubuntu/jammy/amdgpu-install_6.1.60102-1_all.deb
sudo apt install ./amdgpu-install_6.1.60102-1_all.deb
sudo amdgpu-install --usecase=graphics,rocm

### 2 Install necessary python packages with ROCM 6.1 supported
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm6.1
Check availablity of ROCM in python 3.1x 
import torch
print ( torch.cuda.is_available() )

### 3 Install flask server
pip install flask

### 4 Downlaod Llama-2-7b-chat-hf from huggingface 
pip install hugginface-cli
hugginface login
huggingface-cli download --resume-download NousResearch/Llama-2-7b-chat-hf  --local-dir NousResearch/Llama-2-7b-chat-hf 

### 5 Start the flask serve and run ULM 
flask --app start run

#NLM Prediction nlm.py
import torch
import torch.nn as nn
import numpy as np
#from model import Sequence
import random

class Sequence(nn.Module):
    def __init__(self):
        super(Sequence, self).__init__()
        self.lstm1 = nn.LSTMCell(1, 51)
        self.lstm2 = nn.LSTMCell(51, 51)
        self.linear = nn.Linear(51, 1)

    def forward(self, input, future = 0):
        outputs = []
        h_t = torch.zeros(input.size(0), 51, dtype=torch.double)
        c_t = torch.zeros(input.size(0), 51, dtype=torch.double)
        h_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
        c_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)

        for input_t in input.split(1, dim=1):
            h_t, c_t = self.lstm1(input_t, (h_t, c_t))
            h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
            output = self.linear(h_t2)
            outputs += [output]
        for i in range(future):# if we should predict the future
            h_t, c_t = self.lstm1(output, (h_t, c_t))
            h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
            output = self.linear(h_t2)
            outputs += [output]
        outputs = torch.cat(outputs, dim=1)
        return outputs


def update_nlm(input=None):
        future=1        
        pre_input=torch.from_numpy(np.array([[input], ]).astype('float')) 
        model=torch.load('model_seq.pth')
        pred = model(torch.from_numpy(np.array([[input], ]).astype('float')), future=future).detach().numpy();   print(pred[0][-1])              
        return pred



if __name__ == "__main__":
    print("Start")
    update_nlm(133)

<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Nonintrusive Load Monitoring (NLM) Project for Pervasive AI Developer Contest with AMD</title>
<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.3.1/js/bootstrap.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/Chart.js/2.8.0/Chart.min.js"></script>
    <link href="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.3.1/css/bootstrap.min.css" rel="stylesheet">
    <link href="https://cdnjs.cloudflare.com/ajax/libs/Chart.js/2.8.0/Chart.min.css" rel="stylesheet">
</head>

<body>
<div class="container" >
  <div style="justify-content:center;">
    <h2 >LLama2 Chat</h2>
    <div class="col">
      <form id="chat_form" >
        <label>Input question here...<input type="text" id="chat_input" /></label>
        <p><input type="reset" value="RESET" /> | <input type="submit" value="submit" /></p>
      </form>
      <lable> Output the result here...<span id="chat_result"></span></lable>
    </div>
  </div>


    <div class="row">
        <div class="col-12">
            <div class="card">
                <div class="card-body">
                    <canvas id="monitor_canvas"></canvas>
                </div>
            </div>
        </div>
    </div>

    <div class="row">
      ...

    </div>

</div>


<script>
    $(document).ready(function () {
        const config = {
            type: 'line',
            data: {
                labels: [],
                datasets: [{
                    label: "Realtime Dataset",
                    backgroundColor: 'rgb(255, 99, 132)',
                    borderColor: 'rgb(255, 99, 132)',
                    data: [],
                    fill: false,
                },
                {
                    label: "Predict Dataset",
                    backgroundColor: 'rgb(55, 99, 32)',
                    borderColor: 'rgb(55, 99, 32)',
                    data: [],
                    fill: false,
                }],
            },
            options: {
                responsive: true,
                title: {
                    display: true,
                    text: 'Real-Time Non Instrusive Monitoring'
                },
                tooltips: {
                    mode: 'index',
                    intersect: false,
                },
                hover: {
                    mode: 'nearest',
                    intersect: true
                },
                scales: {
                    xAxes: [{
                        display: true,
                        scaleLabel: {
                            display: true,
                            labelString: 'Time'
                        }
                    }],
                    yAxes: [{
                        display: true,
                        scaleLabel: {
                            display: true,
                            labelString: 'Value'
                        }
                    }]
                }
            }
        };

        const context = document.getElementById('monitor_canvas').getContext('2d');

        const lineChart = new Chart(context, config);      

        const source = new EventSource("/ulm-data");

        source.onmessage = function (event) {
            const data = JSON.parse(event.data);
            if (config.data.labels.length === 20) {
                config.data.labels.shift();
                config.data.datasets[0].data.shift();
                config.data.datasets[1].data.shift();
            }
            config.data.labels.push(data.time);
            config.data.datasets[0].data.push(data.value);  
            config.data.datasets[1].data.push(data.value_pre);     
            lineChart.update();
        }
    });

</script>


</body>
</html>

import json
import random
import time
from datetime import datetime
from flask import Flask, Response, request
from flask import stream_with_context, jsonify, render_template
import llama_chat as lc 
import nlm 
from nlm import Sequence


app = Flask(__name__)
random.seed()  # Initialize the random number generator

@app.route('/')
def home(command=None):
    #chatback=lc.chatbot(command);nlmoutput=" <p><a href=\"/\">--- Return for Home Page ---</a></p>"
    return render_template('welcomepage.html' , firstL=command)

@app.route("/get/chat", methods=["POST", "GET"])
def get_chat():
    if request.method == "POST":
        ask_chat = request.form["chat_input"]
        response = {'chat_input': "ask_chat"}
        print(ask_chat)
        response = {'ask':ask_chat, 'answer': lc.chatbot(ask_chat)}
        return jsonify(response)
    elif request.method == "GET":
        ask_chat = request.args.get("chat_input")
        response = {'chat_input': ask_chat}
        print("response")
        response = {'ask':ask_chat, 'answer': lc.chatbot(ask_chat)}
        return jsonify(response)
    else:    
        response=ask_chat        
        return response
    
@app.route('/ulm-data')
def chart_data():
    def generate_ulm_data():
        while True:
            nlm_value=random.random()*1000
            nlm_pre=nlm.update_nlm(nlm_value)
            json_data = json.dumps(
                {'time': datetime.now().strftime('%Y-%m-%d %H:%M:%S'), 'value': nlm_value, 'value_pre': nlm_pre})
            yield f"data:{json_data}\n\n"
            time.sleep(10)
        pass
    response = Response(stream_with_context(generate_ulm_data()), mimetype="text/event-stream")
    response.headers["Cache-Control"] = "no-cache"
    response.headers["X-Accel-Buffering"] = "no"
    return response


if __name__ == "__main__":
    app.run(host='127.0.0.1', debug=False)
    pass

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import argparse\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "opdata=pd.read_csv('data/opdata.csv'); indata=opdata.drop(0);col=indata.columns\n",
    "plt.plot(indata['PowerOutput.2.KW'].head(3200).tail(1500))\n",
    "#plt.plot(indata['LOAD.kW'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "col=indata.columns\n",
    "n,i,j=2,1000,1000; #n=1,2,12\n",
    "#data=[indata.iloc[a:a+j,1:].values for a in range(i) ]\n",
    "data = np.array([indata[col[n]][a:a+j].values for a in range(1500,1500+i) ]).astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Sequence(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Sequence, self).__init__()\n",
    "        self.lstm1 = nn.LSTMCell(1, 51)\n",
    "        self.lstm2 = nn.LSTMCell(51, 51)\n",
    "        self.linear = nn.Linear(51, 1)\n",
    "\n",
    "    def forward(self, input, future = 0):\n",
    "        outputs = []\n",
    "        h_t = torch.zeros(input.size(0), 51, dtype=torch.double)\n",
    "        c_t = torch.zeros(input.size(0), 51, dtype=torch.double)\n",
    "        h_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)\n",
    "        c_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)\n",
    "\n",
    "        for input_t in input.split(1, dim=1):\n",
    "            h_t, c_t = self.lstm1(input_t, (h_t, c_t))\n",
    "            h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))\n",
    "            output = self.linear(h_t2)\n",
    "            outputs += [output]\n",
    "        for i in range(future):# if we should predict the future\n",
    "            h_t, c_t = self.lstm1(output, (h_t, c_t))\n",
    "            h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))\n",
    "            output = self.linear(h_t2)\n",
    "            outputs += [output]\n",
    "        outputs = torch.cat(outputs, dim=1)\n",
    "        return outputs\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "STEP:  0\n",
      "loss: 4229308.8375472445\n",
      "loss: 4229126.171447406\n",
      "loss: 4228204.362346975\n",
      "loss: 4209039.371515984\n",
      "loss: 4010624.398729459\n",
      "loss: 3617811.335570265\n",
      "loss: 1574380.2090736968\n",
      "loss: 1570786.3501395595\n",
      "loss: 1517590.111720877\n",
      "loss: 1525055.9162346714\n",
      "loss: 1510851.0329699842\n",
      "loss: 1578511.304254132\n",
      "loss: 1480321.467325149\n",
      "loss: 940286.9482451878\n",
      "loss: 919920.5238456377\n",
      "loss: 850661.0110311769\n",
      "loss: 738997.6611439213\n",
      "loss: 727609.270031233\n",
      "loss: 700608.4647058692\n",
      "loss: 681335.6276213053\n",
      "test loss: 689517.9904782836\n",
      "STEP:  1\n",
      "loss: 659880.8706200161\n",
      "loss: 636254.4840585578\n",
      "loss: 628380.0121522277\n",
      "loss: 612716.9001586228\n",
      "loss: 555285.6693866277\n",
      "loss: 552464.3442940136\n",
      "loss: 523870.8941396616\n",
      "loss: 503157.79014876246\n",
      "loss: 462367.9776926957\n",
      "loss: 390644.5763691746\n",
      "loss: 350996.7701292465\n",
      "loss: 323543.8451591094\n",
      "loss: 317626.9889090754\n",
      "loss: 310644.897598034\n",
      "loss: 290408.69087756105\n",
      "loss: 286439.43566547614\n",
      "loss: 274917.7605501571\n",
      "loss: 241912.34213964723\n",
      "loss: 227018.70815192795\n",
      "loss: 222906.88023728254\n",
      "test loss: 246295.2391243019\n",
      "STEP:  2\n",
      "loss: 221021.8834850581\n",
      "loss: 217730.30161722004\n",
      "loss: 213088.0331550284\n",
      "loss: 294039.78003409656\n",
      "loss: 225384.85666872055\n",
      "loss: 3172538.856013644\n",
      "loss: 3204916.3546418617\n",
      "loss: 2425244.443952121\n",
      "loss: 1556225.7607728357\n",
      "loss: 1726817.765184505\n",
      "loss: 1348789.5116432854\n",
      "loss: 1317624.7678301763\n",
      "loss: 1200517.4389796134\n",
      "loss: 1180192.687205082\n",
      "loss: 1134112.580162119\n",
      "loss: 1748354.5198837253\n",
      "loss: 1634408.6653953511\n",
      "loss: 1264640.2731982851\n",
      "loss: 1231424.106930328\n",
      "loss: 1317967.0650237831\n",
      "test loss: 3309976.4543191926\n",
      "STEP:  3\n",
      "loss: 2915747.9386233035\n",
      "loss: 995845.2724107329\n",
      "loss: 971823.32198672\n",
      "loss: 767501.5835903304\n",
      "loss: 660658.2017648214\n",
      "loss: 623356.856248768\n",
      "loss: 571414.0062534651\n",
      "loss: 481831.1077602665\n",
      "loss: 7186074716.310981\n",
      "loss: 577685287.6958594\n",
      "loss: 15882974881764.082\n",
      "loss: 5.3779978745030575e+22\n",
      "loss: 2.123514805987358e+21\n",
      "loss: 3.484340063996101e+20\n",
      "loss: 7.744995241376502e+22\n",
      "loss: 6.284406018017003e+20\n",
      "loss: 5.760430704696887e+20\n",
      "loss: 7.37836446319385e+20\n",
      "loss: 1.288541213513832e+19\n",
      "loss: 5.920244461298217e+18\n",
      "test loss: 8.989174303555019e+19\n",
      "STEP:  4\n",
      "loss: 8.854752114998116e+19\n",
      "loss: 2.0502240833171337e+19\n",
      "loss: 2.5335191618434023e+19\n",
      "loss: 2.518225273505821e+19\n",
      "loss: 9.807924053216302e+19\n",
      "loss: 1.368051841661368e+19\n",
      "loss: 7.24458207381586e+18\n",
      "loss: 4.661239506579395e+18\n",
      "loss: 3.5895448236583715e+19\n",
      "loss: 1.1945160102867442e+19\n",
      "loss: 2.2686438355788914e+21\n",
      "loss: 7.65052719093968e+19\n",
      "loss: 6.430270402310691e+19\n",
      "loss: 2.425677021916226e+18\n",
      "loss: 2.8358231364196512e+17\n",
      "loss: 8.522854030603946e+19\n",
      "loss: 2.1116473471601283e+17\n",
      "loss: 5.962680752531842e+16\n",
      "loss: 7774830994559339.0\n",
      "loss: 5675439424281885.0\n",
      "test loss: 6962306828945597.0\n",
      "STEP:  5\n",
      "loss: 5589776197272147.0\n",
      "loss: 5585503892689164.0\n",
      "loss: 5584355618155153.0\n",
      "loss: 5581723227161267.0\n",
      "loss: 5575596387631142.0\n",
      "loss: 5561393686197859.0\n",
      "loss: 5528565852048158.0\n",
      "loss: 5661855484052179.0\n",
      "loss: 3508648373011442.0\n",
      "loss: 3384593810110273.0\n",
      "loss: 8.451278192788898e+18\n",
      "loss: 1.3205813019811722e+21\n",
      "loss: 2.7871366140421837e+18\n",
      "loss: 6.738058945854244e+18\n",
      "loss: 1.2116661668935768e+19\n",
      "loss: 2.021414010377537e+18\n",
      "loss: 2.2070712090942963e+18\n",
      "loss: 2.836123584533348e+18\n",
      "loss: 7.142107398951318e+18\n",
      "loss: 4.860244652011833e+18\n",
      "test loss: 4.930215511245041e+18\n",
      "STEP:  6\n",
      "loss: 4.782976490524536e+18\n",
      "loss: 3.1957816552215224e+18\n",
      "loss: 2.881394038685149e+18\n",
      "loss: 1.0025746724872874e+22\n",
      "loss: 1.416417561199966e+21\n",
      "loss: 7.708961660557937e+22\n",
      "loss: 5.078093570538268e+22\n",
      "loss: 3.089078368752528e+23\n",
      "loss: 7.115959510209808e+23\n",
      "loss: 1.5491714843949407e+24\n",
      "loss: 1.8228170581712446e+24\n",
      "loss: 6.526293091990615e+22\n",
      "loss: 4.784570932315678e+21\n",
      "loss: 1.8258459283111574e+21\n",
      "loss: 2.5521706184880633e+21\n",
      "loss: 6.949525254789608e+21\n",
      "loss: 1.8322630879114842e+22\n",
      "loss: 8.061883389755882e+22\n",
      "loss: 5.263141278383561e+22\n",
      "loss: 1.5632826028675937e+22\n",
      "test loss: 1.98604101123322e+21\n",
      "STEP:  7\n",
      "loss: 3.4044462860939276e+21\n",
      "loss: 2.9309216783292946e+21\n",
      "loss: 1.2589365426676215e+21\n",
      "loss: 6.427960621116192e+21\n",
      "loss: 3.818063869931953e+22\n",
      "loss: 3.3663732785627953e+22\n",
      "loss: 3.3184750973247727e+21\n",
      "loss: 1.3911328559452258e+21\n",
      "loss: 1.1323187022826459e+22\n",
      "loss: 3.7150372212807917e+21\n",
      "loss: 9.6482295819601e+21\n",
      "loss: 6.00082298284013e+21\n",
      "loss: 5.342801216073593e+21\n",
      "loss: 6.610823739944242e+21\n",
      "loss: 4.408676647101628e+21\n",
      "loss: 1.2469687120334827e+23\n",
      "loss: 5.451426261922335e+21\n",
      "loss: 2.0350783859747368e+21\n",
      "loss: 3.553311447870185e+21\n",
      "loss: 5.868359730099477e+20\n",
      "test loss: 3.85350652786694e+20\n",
      "STEP:  8\n",
      "loss: 3.09067874716307e+20\n",
      "loss: 4.008929834557779e+20\n",
      "loss: 7.0853972220544434e+19\n",
      "loss: 6.266430205358155e+19\n",
      "loss: 1.5729935231416805e+20\n",
      "loss: 6.702154029712895e+19\n",
      "loss: 2.8176719895919587e+19\n",
      "loss: 1.4743308403664101e+19\n",
      "loss: 9.71603284646075e+18\n",
      "loss: 9.304721700176509e+18\n",
      "loss: 9.193982643633251e+18\n",
      "loss: 9.482772606336948e+18\n",
      "loss: 8.205424226290608e+18\n",
      "loss: 7.129908831129315e+18\n",
      "loss: 5.847773931033383e+18\n",
      "loss: 6.720438669440779e+18\n",
      "loss: 6.214317060715695e+18\n",
      "loss: 4.713305327218745e+18\n",
      "loss: 4.526105102340476e+18\n",
      "loss: 4.227960471530717e+18\n",
      "test loss: 4.57156033182246e+18\n",
      "STEP:  9\n",
      "loss: 4.156138089578904e+18\n",
      "loss: 4.117729098220786e+18\n",
      "loss: 2.9575873367270267e+18\n",
      "loss: 2.2332768214009103e+18\n",
      "loss: 2.35734979500617e+18\n",
      "loss: 1.9679096540747016e+18\n",
      "loss: 1.0008509775910537e+19\n",
      "loss: 5.711797621236783e+18\n",
      "loss: 5.3616816807986836e+20\n",
      "loss: 4.86031446099056e+23\n",
      "loss: 6.159679664742802e+22\n",
      "loss: 6.6563462977916926e+28\n",
      "loss: 2.723801845348054e+27\n",
      "loss: 3.522550974070969e+26\n",
      "loss: 1.183826784981614e+27\n",
      "loss: 1.7218259015503517e+28\n",
      "loss: 2.6721410547263592e+28\n",
      "loss: 3.6543916622628e+30\n",
      "loss: 1.0581081788655151e+30\n",
      "loss: 9.120068360735249e+29\n",
      "test loss: 1.1121766235176609e+29\n",
      "STEP:  10\n",
      "loss: 1.752299704568742e+29\n",
      "loss: 6.074040465439501e+28\n",
      "loss: 3.8489203317830915e+28\n",
      "loss: 3.736200756274378e+28\n",
      "loss: 2.2764320521214716e+28\n",
      "loss: 9.795301202119256e+27\n",
      "loss: 1.1160113313096871e+27\n",
      "loss: 8.683076259267755e+26\n",
      "loss: 8.657782093675231e+26\n",
      "loss: 1.8737249406133645e+29\n",
      "loss: 3.5538919666117054e+28\n",
      "loss: 1.253298547925453e+32\n",
      "loss: 7.949485219663823e+30\n",
      "loss: 1.2993674184595233e+30\n",
      "loss: 5.032154509612925e+30\n",
      "loss: 7.087154929845263e+28\n",
      "loss: 1.9067182627807933e+29\n",
      "loss: 2.604928287302854e+29\n",
      "loss: 9.522178800716955e+28\n",
      "loss: 3.6596267305375385e+27\n",
      "test loss: 1.012617280407818e+29\n",
      "STEP:  11\n",
      "loss: 4.6423453268857115e+28\n",
      "loss: 8.472719147487659e+28\n",
      "loss: 1.4227743149183834e+28\n",
      "loss: 1.4169579323736715e+28\n",
      "loss: 1.297619778174661e+28\n",
      "loss: 1.5946735978854383e+28\n",
      "loss: 4.397945384642612e+28\n",
      "loss: 3.488061309356017e+28\n",
      "loss: 6.526537225879723e+27\n",
      "loss: 2.084666656859595e+28\n",
      "loss: 2.5062970914067153e+28\n",
      "loss: 8.890751344267596e+27\n",
      "loss: 8.437256527872097e+27\n",
      "loss: 2.440769481884083e+29\n",
      "loss: 2.1788172031604336e+29\n",
      "loss: 1.88993724004194e+30\n",
      "loss: 1.2395252954134014e+29\n",
      "loss: 2.151599551600394e+29\n",
      "loss: 7.091237144660865e+28\n",
      "loss: 6.48348516346586e+28\n",
      "test loss: 1.3837336315676337e+29\n",
      "STEP:  12\n",
      "loss: 1.2194632710501971e+29\n",
      "loss: 2.5343858525873984e+28\n",
      "loss: 2.754746745399109e+27\n",
      "loss: 1.9536195536962825e+28\n",
      "loss: 1.1490727202039862e+29\n",
      "loss: 2.0193132195555384e+28\n",
      "loss: 5.398194126683178e+27\n",
      "loss: 2.137616397225387e+28\n",
      "loss: 1.2272414511574153e+29\n",
      "loss: 2.28931966467625e+29\n",
      "loss: 6.493070046656572e+27\n",
      "loss: 5.088021398868565e+28\n",
      "loss: 5.501702013266639e+28\n",
      "loss: 1.683903928635051e+28\n",
      "loss: 1.4677131889536392e+29\n",
      "loss: 1.5564846450744639e+29\n",
      "loss: 2.4449090337113042e+28\n",
      "loss: 8.116933757060467e+26\n",
      "loss: 5.816798900305919e+26\n",
      "loss: 3.3976918518103173e+25\n",
      "test loss: 2.0595527717958443e+25\n",
      "STEP:  13\n",
      "loss: 1.5908129945614884e+25\n",
      "loss: 1.5218902402751574e+25\n",
      "loss: 1.518819500239654e+25\n",
      "loss: 1.5183882637309243e+25\n",
      "loss: 1.5177232249371335e+25\n",
      "loss: 1.516068640941223e+25\n",
      "loss: 1.5141231151605883e+25\n",
      "loss: 1.5050835719163522e+25\n",
      "loss: 1.6056646442452418e+25\n",
      "loss: 1.9646269635289595e+25\n",
      "loss: 1.936131503711977e+25\n",
      "loss: 1.877174737366576e+25\n",
      "loss: 1.8125265049446746e+25\n",
      "loss: 2.3529210418665806e+25\n",
      "loss: 3.363968043735889e+27\n",
      "loss: 2.8071074649487495e+27\n",
      "loss: 4.685292305395069e+28\n",
      "loss: 9.014249485842906e+28\n",
      "loss: 3.589841965889721e+30\n",
      "loss: 4.327373992067767e+30\n",
      "test loss: 6.416750302087601e+31\n",
      "STEP:  14\n",
      "loss: 6.942324930510634e+31\n",
      "loss: 5.834094180639929e+30\n",
      "loss: 8.378321112507395e+30\n",
      "loss: 7.274884269084158e+30\n",
      "loss: 1.6314558089154024e+33\n",
      "loss: 2.1263266320356805e+31\n",
      "loss: 5.13748160534097e+30\n",
      "loss: 1.0926326242537545e+31\n",
      "loss: 3.099379664565872e+30\n",
      "loss: 4.0564505938301796e+30\n",
      "loss: 7.180862708031849e+30\n",
      "loss: 3.243408758758547e+31\n",
      "loss: 7.279874596350342e+30\n",
      "loss: 3.5553394784780254e+30\n",
      "loss: 1.639331793420979e+31\n",
      "loss: 3.3455923116582575e+30\n",
      "loss: 2.2780419651089666e+30\n",
      "loss: 1.8298317441151598e+29\n",
      "loss: 1.4123743676282105e+30\n",
      "loss: 2.9968865602356403e+31\n",
      "test loss: 3.250574739307066e+31\n"
     ]
    }
   ],
   "source": [
    "steps=15\n",
    "np.random.seed(0)\n",
    "torch.manual_seed(0)    \n",
    "# load data and make training set\n",
    "input = torch.from_numpy(data[3:, :-1])\n",
    "target = torch.from_numpy(data[3:, 1:])\n",
    "test_input = torch.from_numpy(data[:3, :-1])\n",
    "test_target = torch.from_numpy(data[:3, 1:])\n",
    "    # build the model\n",
    "seq = Sequence()\n",
    "seq.double()\n",
    "criterion = nn.MSELoss()\n",
    "    # use LBFGS as optimizer since we can load the whole data to train\n",
    "optimizer = optim.LBFGS(seq.parameters(), lr=0.8)\n",
    "    #begin to train\n",
    "for i in range(steps):\n",
    "    print('STEP: ', i)\n",
    "    def closure():\n",
    "        optimizer.zero_grad()\n",
    "        out = seq(input)\n",
    "        loss = criterion(out, target)\n",
    "        print('loss:', loss.item())\n",
    "        loss.backward()\n",
    "        return loss\n",
    "    optimizer.step(closure)\n",
    "        # begin to predict, no need to track gradient here\n",
    "    with torch.no_grad():\n",
    "        future = 1000\n",
    "        pred = seq(test_input, future=future)\n",
    "        loss = criterion(pred[:, :-future], test_target)\n",
    "        print('test loss:', loss.item())\n",
    "        y = pred.detach().numpy()\n",
    "torch.save(seq,'model_seq.pth')        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "seq=torch.load('model_seq.pth')\n",
    "test_input = torch.from_numpy(data[:3, :-1]);future=1\n",
    "pred = seq(test_input[:1,:100], future=future).detach()\n",
    "#loss = criterion(pred[:, :-future], test_target);print('test loss:', loss.item())\n",
    "#y = pred.detach().numpy()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAACZcAAAO3CAYAAABVluhXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3wU1frH8e+mkxBK6DUJvbdQQhMFRMSOFOnoT8R2xc69FkTFKyqIvaBIEQEVRVS4KAiI9N6RmlCkBgihpu38/sCsmexuspuyG7Kf9+uVe/fMnjlzZvbMzDr78ByLYRiGAAAAAAAAAAAAAAAAAADIxM/bHQAAAAAAAAAAAAAAAAAAFD4ElwEAAAAAAAAAAAAAAAAA7BBcBgAAAAAAAAAAAAAAAACwQ3AZAAAAAAAAAAAAAAAAAMAOwWUAAAAAAAAAAAAAAAAAADsElwEAAAAAAAAAAAAAAAAA7BBcBgAAAAAAAAAAAAAAAACwQ3AZAAAAAAAAAAAAAAAAAMAOwWUAAAAAAAAAAAAAAAAAADsElwEAAAAArilRUVGyWCyyWCyKiorKtu7o0aNtdS0Wi5YuXeqRPhZW+/fv1zPPPKNWrVqpbNmyCggIMB2f+Ph4b3cRyJMpU6aYxvSUKVO83SVIMgxDP/30kwYMGKBatWopPDzc9Dldf/313u7iNYVjBwAAAAAAPCnA2x0AAAAAAAAFb8KECXrmmWeUnp7u7a4A8CGJiYnq1auXfvvtN293BQAAAAAAALlAcBkAAAAAeFBUVJQOHjyYbR0/Pz+VLFlSpUuXVv369dWqVSv17t1bDRo08FAvUdTMmDFDTz75pLe7AcAH9e3bl8AyAAAAAACAaxjBZQAAAABQyFitVp09e1Znz57VgQMHNG/ePI0ePVpdunTRxx9/rNq1a3u7i3BTfHy8oqOjbeUhQ4Z4bLq+tLQ0Pf3006Zlt912m+6++25VqlRJfn5+tuUVK1b0SJ8ymzJliu69915befLkyRo6dKjH+wEg/82fP1+//vqrrVy8eHE98sgjatOmjcLDw23LS5cu7Y3ueQXXPAAAAAAAcK0huAwAAAAArhG//fabmjVrptmzZ+vmm2/2dndwjVi0aJGOHTtmKw8ePFhTp071Yo8A+Ipp06aZyj/++KNuuOEGL/UGAAAAAAAAuUFwGQAAAAB40bhx49S0aVPTsvT0dJ0+fVobN27UrFmz9Ndff9neu3Tpknr37q2VK1eqSZMmnu7uNWf06NEaPXq0t7vhVatWrTKVhw8f7qWeAPA1ma8/tWvXJrAsnxiG4e0uAAAAAAAAH0JwGQAAAAB4UUxMjK6//nqH7/Xv31///e9/9eyzz+rdd9+1Lb948aKeeuopLVy40EO9xLVsz549pnLDhg291BMAvuTKlSs6dOiQrcy1BwAAAAAA4Nrk5+0OAAAAAACcCwoK0jvvvKMBAwaYli9atEjbt2/3Uq9wLUlMTDSVS5Qo4Z2OAPAp586dM5W59gAAAAAAAFybCC4DAAAAgGvA2LFj5edn/k+4X3/91Uu9wbXkypUrprLFYvFSTwD4Eq49AAAAAAAARQPTYgIAAADANaBq1apq2rSpNm3aZFu2bds2t9s5ffq0Vq9eraNHj+rUqVMqXry4unfvrjp16mS7XkpKilavXq34+HidOnVKVqtV5cqVU+3atRUbGyt/f3+3+5LVzp07tWnTJv3111+SpCpVqig2NlY1a9bMc9t5deLECa1Zs0YnT55UQkKC/Pz8VKpUKdWpU0fNmjVTqVKlvN1FpwzD8HYXfMaFCxe0YsUKHT16VMePH1dISIg6deqkFi1aZLteenq61q9fr3379unkyZNKTk5WuXLlFB0drfbt2ys4ONhDe+BZJ0+e1Pbt27V//34lJiYqLS1NERERqlixotq0aaOKFSt6u4sFriCvrampqdq2bZu2b9+u06dP6+LFiwoODlZ4eLgiIyNVp06dAr2+5ve1JzEx0XZ+JSQkqHjx4ipfvryaN2+e4z0sN3J7v7zWXLhwQcuXL9fhw4d16tQplSxZUi1atFBsbGyOAYFHjhzRypUrdejQIaWnp6tixYpq3769atWqlS99O3TokNavX68TJ07o7NmzKlmypG0b3rg+nD9/Xps2bdLu3buVmJio5ORkhYaGqnTp0oqKilKDBg1UoUKFXLfviXtBXFyc1qxZo7/++kupqamqWLGiWrZsqUaNGuW5bW/y5FhJTU3VypUrtX37diUmJqpEiRKqVq2aOnXqpNKlS+fLNgrye+fu3bu1ZcsWnTp1SufOnVNERIQqV66sDh06KCIiIk/9PnjwoDZt2qQjR44oKSlJFotFYWFhqlSpkmrUqKGGDRsqJCQkT9sAAAAA4CUGAAAAAMBjIiMjDUm2vyVLlri87t13321a99Zbb822/cjISNvylStXGrfddpsRGBhoakOSMWHCBKfb3LZtm9G7d2+jePHidutl/JUqVcp49NFHjRMnTrhxJP7x008/GU2aNHHafmxsrLF06dIc99GRl156KdfHOzk52fjwww+NJk2aGBaLxWn//P39jfbt2xufffaZcfHiRVMbWT9vV/8mT57sxhE0mzx5cq62GRcXZ2sjLi7O9N6QIUPc6kOnTp1M62eVtf3c9tPR/rp77DKv26lTp1zt144dO4x+/foZoaGhdv0dMWKE0/bi4uKM++67z4iIiHC6v6GhocbAgQONAwcOuLVfrkhOTjZtOzQ01EhKSnK7nbi4ONM50rBhQ4f1rFarsWzZMuPRRx816tatm+Nn3bhxY2PKlClGamqqy31xZzwsWbLEVPell15ya7/duRZlVZDX1mPHjhmPPPKIUbp06RyPcdmyZY0+ffoYCxYscGsbzmQ9R1z5y+m8+/33343OnTsbAQEBTtuoVauW8fbbbxvJycku9zW/75fZyc9rnmG4d90aMmSIw/aOHDli3HfffU7HYM2aNY2ffvrJYZubNm0yunXr5vTe2LFjR2Pbtm25OlbJycnGO++8YzRo0MDpMbFYLEbLli2NuXPn5mob7tqwYYNx1113GUFBQTl+XtHR0cajjz5q7Nixw+X2PXEvWLlypdGuXTun7Tds2ND47rvvbPVzuo9n5k7drPJyDy+oseLsu+OVK1eMV155xenn5O/vb/Tt29eIj4936xhk3p+8fu905vz588bo0aON6OjobNu9/vrrjWXLlrnV7/T0dOPTTz/N9rt8xl9QUJDRtm1bY9y4ccalS5dyc5gAAAAAeAnTYgIAAADANSIwMNBUTktLc2m9N998U+3bt9dPP/2k1NRUl9ZJS0vTv/71LzVt2lTffvutLly44LRuYmKiPvjgA9WqVUvz5s1zqX1JslqtGj58uG677TZt3brVab3Vq1frhhtu0FtvveVy23m1atUq1alTR4888oi2bt2abQae9PR0rVixQsOGDdM333zjsT6icPjqq6/UvHlzzZw5U5cuXXJ5vTFjxqhu3br64osvdObMGaf1Ll26pOnTp6tevXqaNGlSfnTZJigoSPfcc49pW7Nnz3a7nWnTppnOkSFDhjis98wzz+i6667TBx98oN27d+fY7rZt2zR06FDdcMMNOnnypNv9KowK+tq6cOFC1atXTx9++KHOnj2bY/2EhAR98803evXVV13eB09JSUnR4MGD1alTJy1evDjbe96+ffv05JNPqlGjRvrzzz9zvc3c3C+vRatWrVLTpk31xRdfOB2D+/fv1+23367333/ftHzSpElq3bq1fv31V6f3xj/++ENt27bVihUr3OrXmjVrVK9ePT3++OPauXOn03qGYWj9+vW64447dPvtt+vixYtubccdY8eOVatWrTRnzhylpKTkWD8uLk4ffPCBZsyY4VL7nrgXvPLKK+rQoYNWrlzptM6OHTt0991367HHHrsmMp56eqz89ddfio2N1ahRo5x+Tunp6fr666/VsmVLU6ZhVxTk986ff/5ZNWvW1OjRoxUXF5dtu0uXLtV1112n4cOHu/TfGWfPnrXVz+67fIaUlBStWrVKTz/9tC1TMQAAAIBrA9NiAgAAAMA14ujRo6ZymTJlclzn008/1ciRI23loKAgRUVFqWTJkjp69Khdm9LVHzB79uypX375xe69ihUrqmLFivLz89ORI0dMAR/nz5/XHXfcoZkzZ6p379459u2hhx7SxIkT7ZZXqFBBVatW1fnz5xUXF6fU1FQZhqFnn31WlSpVyrHdvJo1a5aGDh2q5ORku/cqVaqkihUrymKxKCEhQYcOHSrw/qDwmj9/vgYPHiyr1SpJ8vPzU3R0tCIiInTy5EkdOXLEbp309HT93//9n6ZOnWr3XpkyZVSlShUFBQXp2LFjph9eU1JSdP/99yspKUlPPPFEvu3DkCFD9NFHH9nK06ZN07333utWG9OmTbO99vf318CBAx3Wu3Llit2y0qVLq2LFiipRooSSk5N14sQJHTt2zFRn+fLl6ty5s9atW6dixYq51bfCpKCvrTt27NBtt91md+0KDg62XffT09N17tw5HTx4sFAHTyUnJ+vWW2/VokWL7N6rVKmSqlSpoqSkJNs9IsPevXvVoUMHLVy4UM2bN3drm7m5X16L9u7dqz59+igxMVGSFBAQoJo1a6pEiRI6evSo6bpjGIYef/xxNWvWTB07dtTMmTM1bNgwW+BLWFiYoqKiFBISogMHDpgCGi9cuKC77rpLu3fvdmmqwJ9++kl9+/bV5cuXTcuDgoIUHR2tkiVLKikpSfv27TMFvfz000/q3Lmzfv/993yfbm/SpEn6z3/+Y7c8PDxcUVFRCgsL0+XLl3XmzBkdOXLEraAsT90L/vvf/+qll16yWx4REaHIyEglJycrLi7Odtzff/99lS9f3uX98AZPj5XExER17drVFLhapUoVVapUSVeuXNHevXtN192EhATdfvvt2rFjh0qUKJFj+wX5vXPixIl6+OGHlZ6ebloeGhqqyMhIhYeH68yZMzpw4IDtu0zGeidOnNCcOXOcTo9rGIbuuOMOuyBSi8Vi63dgYKDOnz+vo0eP2q45AAAAAK5R3kmYBgAAAAC+KbfTYl68eNEoVqyYad0XXngh2/bDw8Nt61SqVMmYNGmS3ZR3Bw4csJu6avDgwabtFC9e3HjxxRcdTsW0efNmo1evXqb64eHhxt69e7Pdn5kzZ9pNldOlSxdj/fr1pnpnz541xo8fb5tusGTJkkapUqVs6+T3tJhr1661m/aqRIkSxquvvupwarIzZ84YP/zwg9G/f38jKCjIbjqn5cuXGwsXLjSmT59uarNbt27GwoULnf4dPXo0235m5+jRo6a2sk5T5Gybly9ftrVR0NNiXr582bbdZ555xlT3mWeeyfbYZO6nYXh3WswKFSrYxuX48eONU6dOmeofP37cWLNmjWnZqFGjTG0EBgYaI0aMcDiF2r59+4zhw4ebpscKCAgwVqxY4dY+5qRevXqmKcTcmdJr+fLlpv3p3r2707qPPPKIER4ebtx7773G7NmznY7zI0eOGGPHjjWd65KMxx57LMf+FOZpMQv62tqjRw9T/WbNmhk///yzw6kiU1JSjI0bNxqvvfaa0bhxY6N9+/Zu7bsz69evt52rrl73sl73DcMwnnzySbt7xJ133mls2bLFVO/06dPGm2++aXd/rFmzpnH+/Pls+5of90tX5ec1zzDyNi1mVFSUIV2ddvWdd94xzp49a6q/Zs0ao3HjxqZ1WrZsaezbt892L65Tp47x/fffm8ZWWlqa8fXXX9udt08++WSOx2f79u12n2HHjh2Nn3/+2W7/k5KSjIkTJ9quvxl/Dz74YI7bcceVK1fspj+8++67jfXr1xtWq9WuflJSkvHbb78ZTz31lFGhQgXj+eefz7Z9T9wLVqxYYTe9YosWLYwlS5aY9uHixYvGpEmTjDJlytjarlKlSrb38cw8OS2mJ8ZK1u+ONWrUsB2XRx991Ni3b5+p/oULF4xx48bZTaX77LPP5rj/+f29M7NFixYZfn5+prZvu+02Y+nSpXbTTZ8+fdp44403jPDwcFP9sWPHOm3/m2++MdUNCwsz3nrrLeP48eMO6x86dMiYOnWq0bNnTyMgICDH/1YAAAAAULgQXAYAAAAAHpTb4LKxY8fa/dC+ePHiHNuXZNSvX984duyYS9uZNWuW3Q/0WX9Ec+TNN980rXf77bc7rZuUlGSUL1/eVH/YsGEOf6zNsGHDBrsfvFwJ6HAnuOzKlSu2H90z/ho0aGAcOnQop903DONqQMyff/7p8L28BmvlRW5+9C3o4LLM8hoc5s3gMklGxYoVjV27drm0rRUrVph+6C1btqyxbt26HNebNWuWab0mTZq4tD1Xvf7666Z9evXVV11ed9iwYaZ1Z86c6bTu+vXrjXPnzrncdnx8vBEdHW1ru1ixYsbp06ezXaewBpcV9LU1MTHRCAgIsNWrV6+ecenSJZf3xdUx7I7cXkfWrl1rFxAzatSobNdZv369UaJECdM6I0aMyHadvN4vcyuv1yzDyFtwmXQ1MNZREFOGU6dO2d2n69evb0hXA3myBt5ltnjxYtN65cqVM1JSUpzWT01NNRo1amRa5+WXX872O4FhXL3n1q5d27Texo0bs13HHfPmzTO1PXjwYJfXTU5ONvbv3+/0fU/cC9LT042GDRua9qFHjx7Zfhbx8fF2QWWu3Mc9FVzmqbGS9bujJCM4ONiYN29ettvJGlBbvnz5bI93QX7vPHv2rFGxYkVbu35+fsakSZNybHPHjh1GuXLlbOsFBQU5vSb27t3b1HdH/13izKFDh3IMAAYAAABQuPgJAAAAAFCo/fTTT3rxxRdNy6Kjo9WhQ4cc1w0MDNQ333yjihUr5ljXMAyNHj3aVg4NDdUvv/yimjVr5rjuM888Y5qu7aefftKePXsc1p0xY4ZpyremTZvq448/djrtjiS1aNFCH3/8cY79yIupU6cqPj7eVi5TpowWLVqkatWqubR+lSpVVLdu3QLqHQqrKVOmqF69ei7VfeWVV0xTaM6dO1ctW7bMcb2+ffvqqaeespW3bt3qcLrA3Bo4cKD8/P55RPTll1+6tN6VK1f07bff2solS5bUnXfe6bR+TEyMS1OEZYiMjNRnn31mK1++fFmzZs1yef3CwhPX1ri4ONP0b0OHDnVrClFXx7AnTJgwwTS94K233qqXX34523ViYmLspln+/PPPde7cOZe368798lo3ZcoUNWjQwOn7ZcuW1ZNPPmlatmvXLkVEROibb75ReHi403VvuOEGdevWzVY+deqU1q9f77T+7NmztX37dlt5+PDhGjVqVLbfCaSr99zvvvvOdO0aP358tuu4I+t59vDDD7u8blBQkGrUqOH0fU/cCxYuXKgdO3bYypUqVdKsWbMUGBjotP3IyMhCfY315lgZO3asevTokW2dAQMGqE2bNrbyyZMntWHDBqf1C/J75yeffKLjx4/byq+99pruu+++HNts0KCBpkyZYiunpKTogw8+cFg38zlSr1493XDDDS71W5KqVaum4sWLu1wfAAAAgPcRXAYAAAAAhUx6eroSEhL0yy+/qH///rrjjjuUmppqqjNmzJhsfyDM0LdvXzVq1Mil7f7yyy/6888/beURI0a4FPyQ4YUXXrC9NgxDc+bMcVjviy++MJX/+9//yt/fP8f2BwwYoObNm7vcH3e98847pvKbb76pSpUqFdj2cO3r0KGDbrrpJpfq7tq1S7/88out3LdvX7Vr187lbY0cOVIBAQG28nfffed6R3NQtWpVdenSxVbes2ePVq9eneN6c+fOVWJioq3cp08fhYSE5Fu/JKlLly6m83DlypX52r4neOLaevnyZVPZlftDYZSYmGga2xaLxeUgkL59+yo2NtZWvnjxombMmOHytt25X17L2rVrp+7du+dY7/bbb7db9vDDD7sUfHfHHXeYyps2bXJaN/O9NzQ0VK+//nqO7Wdo3LixaVtz585Venq6y+tnp6DOKU/dC7J+13rhhReyDQrM0KFDh2yDhL3JW2OlSpUqeuSRR1yq27dvX1N548aNTusW1PfO9PR0vf/++7Zy9erVTUGJOenRo4fp+7azMZb5HLlW7zkAAAAAXEdwGQAAAAB40Q033CCLxWL6CwgIULly5dS9e3fNnDnTlMFFkp588kn179/fpfb79evncl/mz59vKg8aNMjldSWpSZMmph+d//jjD7s6Fy5cMGUwqVChgsvBOZI0ZMgQt/rkqiNHjmjXrl22cpkyZTRgwIAC2RaKDnfOr//973+msrvnV5kyZRQTE2MrOzq/8iLruTV16tQc15k2bVq2beSXqKgo2+vsglQKK09cWytXrmwqf/PNN3ZBydeCVatWKSUlxVbu0KGD6tSp4/L6WTPzLFu2zOV13Tmfr2WZM+Flp06dOgoKCjIt69Wrl0vrNm7c2FQ+dOiQw3qnT5/W2rVrbeVbb71VpUuXdmkbGTJnSbtw4UK+XSOynlPTp0/Pl3Y9dS9YunSp7XVgYKDuuecel7cxdOhQt/rkCd4cKz179nQ5eKpZs2am8uHDhx3WK8jvnVu2bNHRo0dt5Xvuucft4K/Mx+rPP/9UQkKCXZ3M58jOnTuvyfszAAAAANcRXAYAAAAA14iwsDB99NFHbk3l07p1a5frZv6BMiwsLFfTpGWeyifzj2YZNmzYYJsKSroaOOBK1rIM119/vdt9ckXWH2c7d+6s4ODgAtkWio7cnl+SXJoCLavq1avbXv/55592gad50bNnT9OUlV9//bUpyCerEydOmLLv1KpVS+3bt3d5e/Hx8XrzzTfVu3dv1a9fX+XKlVNwcLBdsK3FYtGqVats6zn6gbuw88S1NTIyUrVr17aV16xZo+7du19zmd7WrFljKnfu3Nmt9TNn4JPkUga+DO6cz9eyzIFJ2fH391fJkiVt5cDAQJczu5UpU8ZUTkpKclhv+fLlputYXq+LkuPzIzc6d+5s+n4yYcIEPfzwwzpw4ECe2vXEveDgwYOm6cebNGmiiIgIl9vv1KmT230qaN4cK+5sq3z58qays6l5C/J7Z36PMcnxsbrxxhttr9PT09W9e3d9/vnnunTpktvbAwAAAFD4EVwGAAAAAIVYSEiI2rZtq7feekvx8fF66KGHXF63ePHiKlu2rMv1M/9wdPHiRfn5+TkM9Mjub926dbY2zpw5Y7eNuLg4U9ndKcgaNGjgVjCaq/bv328q5+aHOPie6Ohol+tm/WG2fPnybp9f3377rW399PR0pwEbuVGsWDFTRqOzZ8/qp59+clr/q6++Mk0pNnjwYJe2c/DgQd15552qUaOGRo4cqdmzZ9uyomQXzJYh8zSc1wpPXFslafTo0aby4sWL1b59e0VHR+vBBx/UzJkzdfDgwQLZx/yStX9NmjRxa/0aNWqYpv47fPiwS0GY7t4vr2XlypVzuW5oaKjtdUREhMv338zrSfZTTGbIel189tln3T43brnlFlMbzs4Pd1WrVs0uE97HH3+smjVrqmXLlvr3v/+t+fPnu709T9wL8vpdq1SpUqaA1sLAm2Mla8BYdsLCwkxlZ2O/IL93Zj1Wffr0cftYZZ0G1NGxevDBB1WhQgVb+eTJkxo2bJjKlSunO+64Q++8847WrVt3TWbRBAAAAGAvwNsdAAAAAABfNm7cODVt2tS0zN/fX+Hh4SpVqpSioqIUEJC7/3TLnIUoJxcvXlRycnKutuOMo2wNWQNDsmY3yUlgYKDCw8PzPcAk649m7vyQCN/lzjl2+vTpfN/+uXPnTJmF8mrIkCGaNGmSrTx16lTdfffdDutmnjbTYrG4NLXb2rVr1a1bN6eZXFzhSgBaYeKpa6sk9e/fXwcOHNCoUaNMAVXx8fH69NNP9emnn0q6muWsW7duuueee2xTMxcWZ8+eNZVzE/BVpkwZnT9/XtLVwJvz58/neK66cy5f60JCQjy6niSnAX4FdV3ML++9956OHz9uF2i7YcMGbdiwQW+88YYsFouaNm2qm2++WQMGDFDDhg2zbdMT94K8ftfKWMfZlI7e4M2xUhBjvyC/d3rqWEVEROjnn3/W7bffrmPHjtmWX7p0ST/++KN+/PFHSVeDdzt27KiePXuqV69eKlWqVL73DwAAAEDBI3MZAAAAAHhRTEyMunbtavq74YYb1LJlS9WqVSvXgWXS1UAsVxVENiBHP6hduHDBVM6a3cQVWbNC5IeMQIQMxYsXz/dtoOjx9jmWeYrZ/NChQwfVqFHDVl6wYIFOnTplV2/Lli3aunWrrdypUydFRUVl2/bp06fVo0cPux+omzRpokceeUQTJkzQ119/rR9//FG//vqrFi5caPtzN3tVYeKpa2uGF154QatWrVKPHj2cZpk6ePCgPvvsM3Xp0kVNmjTRwoUL872PuZX1HpGb633WdbJe3x1x51xG/ins18WQkBDNnTtXM2bMULNmzRzWMQxDmzdv1uuvv65GjRrp1ltv1b59+5y26Yl9LqzftfKisI8VdxXk905PHquWLVtq+/bteuaZZ5xOvXrhwgX973//07Bhw1S9enWNHj2abGYAAADANYjMZQAAAAAAux8eIyIi9PXXX+f7drL+WHnp0iW327h48WJ+dccm8zRqkv0Ps0BehYaGmqYu+9///pen4FFJqlixYl67ZWKxWDR48GDb9IqpqamaOXOmHnvsMVO9zFnLpKsZz3Ly2muvmbKp1K5dW9OnT1fr1q1zXDc3gRGFhaeurZm1adNG8+bN07Fjx/Trr79q6dKlWrZsmQ4cOGBXd/v27brppps0fvx4PfHEEwXaL1dkDbDIzfU+6zpZr+8oPLKeH48//rjd1IXuyhwgmx8sFov69eunfv36aefOnVq4cKGWLl2q5cuXKyEhwa7+vHnztGzZMs2bN08dO3a0e98T94LC+l0rL66FseKOgvzemfVYjR07VjExMXlqM7uMfBEREXrzzTc1ZswYLV26VEuWLNHvv/+u9evX2wWRnT9/Xi+//LIWLVqkX3/99Zq+vwMAAAC+huAyAAAAAIBKlSqlgIAApaWlSZIuX76srl27Fsh2MnP0w2x2UlNTXcpC466s2RZOnjyZ79u4luR1mrzc/JCdW3npqyf7WbZsWVNAQYsWLQrl9KuDBw/Wyy+/bMuONXXqVFNwWVpammbMmGErh4WFqVevXjm2mzmgKiQkRAsWLHD5h/2s04flF0+Mc09dWx2pVKmShgwZYgv+O3r0qBYvXqwffvhBP/74o+1Hf8Mw9NRTT6lt27aKjY31SN+cKV26tKmcm+ndMq+TMc00Cqes055WqlTJY+dHbjRo0EANGjTQiBEjZBiG/vzzT/3666+aPXu2li9fbqt3/vx59erVS/v377cLmPTEvSCv37Uk9849T9yHr7WxkpOC/N6Z9VhFR0d75FgFBQWpW7du6tatm6Sr97vVq1dr/vz5mjFjho4ePWqru2LFCj399NP66KOPCrxfAAAAAPIH02ICAAAAAGSxWBQZGWkrX7582fQjUH7JGkyyfft2t9bfsWOH0tPT87NLkq5mUcps/fr1+b6Na0lISIipfPnyZbfWdzSVYkHJS1892c/o6GhTObtp07wpOjralG1n48aN2rFjh638yy+/6MSJE7Zyz549c5zO69ChQ6brSffu3V0OLLt8+bLi4uJc7b5b8jJ20tLSdPbs2Rzreera6orKlStr4MCBmj17tvbu3atWrVrZ3jMMQxMmTPBKvzLLfKykq1OwuuPAgQOmAOTq1avnOYgQBedauS46YrFYVL9+fY0YMUJ//PGHli1bZgrqOXnypL788ku79Tyxz3n9rpWYmKjDhw+7XN8T9+Freaw4UpDfOwvLsSpWrJhuuOEGvfXWW4qPj9fzzz9vev/zzz83BVoCAAAAKNwILgMAAAAASJJuuOEGU3nx4sX5vo2YmBj5+f3zn6LLly93K1js999/z/c+SbKbumrx4sVKTk7Ot/Yz77MkW1aowqpEiRKmcuZgopwkJCQoPj7e5fp5PTZ56eu6devc2lZeeOL8yi9Zp7mcNm2aw9eO6jqS9TOpW7euy335448/7KbVyi95GTubN2+2ZSPLSWH87CMjI00Z6CSZMi95S9bMae4eq6z1vZ2JzZFr7X5QkArjuZFbHTt21NixY03LHJ1TntjnyMhIUza0bdu2uZUB0t3vWp64DxelsSIV7PfOwnisAgMDNWbMGHXo0MG2LDU1VWvWrPFirwAAAAC4g+AyAAAAAICkq9mEMvvggw/yfRvFixdXTEyMrXzy5En98ssvLq8/ZcqUfO+TdDWjT+PGjW3l06dP66uvvsq39sPCwkxlT07HmBvFihVTuXLlbOWtW7fKarW6tG7m6Q9dkddjkzXT0aZNm1xe192+5kXW82vixIkFFjSVV71791ZoaKitPH36dFmtViUmJurHH3+0La9WrZrdj9iOZA2eSUlJcbkvBTllVrVq1UxZrQpq7Hji2pobtWrVUoUKFWzl3Eydl99iY2MVFBRkKy9fvtytrDtffPGFqdypU6d861t+udbuBwWpSpUqatSoka28f/9+/e9///Nij/Kmffv2prKjc8pT94LMYz81NVWzZs1yeV13v2vl9j6ckJDgcuBTURsrBfm9s3Xr1qYphhcvXqydO3fmS9t55co5AgAAAKBwIrgMAAAAACBJuvPOO1WrVi1bec2aNfr444/zfTv33Xefqfzcc8+5lL3sq6++0ubNm/O9PxlGjBhhKj/77LM6duxYvrRdokQJ+fv728oFNc1ffmrRooXt9dmzZ7Vw4cIc1zl37pzeeOMNt7YTERFhKrt7bOrWrWsKhFq4cKESExNzXG/dunWaM2eOW9vKi5iYGFMg1uHDh/XCCy94bPvuCA8PV8+ePW3lo0ePatGiRfrmm2905coV2/JBgwbZZWFypGLFiqayqxmy5s+fr7lz57rYa/cVL15cderUsZW3bt2q3bt357jewYMH9emnn7q8HU9dW92VkpJimpIsczCCt5QqVUq9evWylQ3D0NNPP+3SurNnz9aqVats5eLFi6tfv3753se8yus1r6h55plnTOXHH39c586d81Jv8iZroIyjc8pT94Ks37XGjBljmjLWmeXLl+uHH35wa1uZvy9I0jfffOPSeq+88opbU2gWpbEiFdz3zsDAQD3++OO2smEYGj58eKEIaHflHAEAAABQOBFcBgAAAACQJPn7++vVV181LRsxYoQ+++wzt9rZs2ePHnjgAf31118O3x8wYIApK9aWLVv08MMPZ9vmpk2bcqyTV4MGDVLNmjVt5dOnT6tr1646cuSIS+v/9ddfTgNTAgMDTUEsmzdv1v79+/PW4QLWo0cPU3nkyJHZZti5ePGi+vbtq8OHD7u1nYYNG5rKP/74o1s/gPr7++umm26ylS9fvqxnn30223X279+vPn36uDUla3549dVXTcFYb775pl555RW3psU7cuSInnnmmQKf0tPR1JhTp07Nto4z1atXV5UqVWzldevW5Zj5a+3atRo4cKCLvc29rOP88ccfz3ZcnDp1Sj179nQpSCODJ66tM2bM0Ouvv66zZ8+63N6HH35oCuzInFXSm5544gnTeTJ37lyNGTMm23U2b96s+++/37Ts/vvvt5uurzDI6zWvqBkwYIDpmOzZs0c333yzjh496nIbqampmjp1qtvBzdl58cUXNX36dJenvzUMQ+PHjzctc3ZOeeJe0K1bN9WvX99WPnbsmO65555sx9rBgwd1zz33uNyHDF27dlVgYKCt/M033+QYRPz555+7ncWxsI6V3CrI750jRowwZaZcvny5evXq5VYw3sWLF/Xee+9p0qRJdu8lJSVpwIABbmX8jI+P1+zZs21lPz8/NWvWzOX1AQAAAHgXwWUAAAAAAJt77rlHw4cPt5VTU1P1wAMPqEuXLvr555918eJFu3VSU1O1ZcsWvfPOO+rYsaPq1aunzz77zOkPmOHh4ZowYYJp2cSJE9WtWzdt2LDBtDwxMVFvv/22OnbsqKSkJJUsWVKlSpXK+446EBQUpK+//lrBwcG2ZTt37lSjRo302muv6eDBg3brJCYmau7cuerfv79q1KhhypqTVbdu3Wyv09PTdd111+nll1/WnDlztHDhQi1atMj2l18Z0/JiwIABpoxgW7ZsUdeuXe1+SLxy5Yq+//57tWrVSr/88ouCgoJUtWpVl7dTvnx5NW3a1Fbeu3ev2rZtq3fffVc///yz6bgsWrTIlDkrw7Bhw0zlzz77TPfdd5/dD85nzpzRO++8o9atWys+Pt70o64ntG/fXq+99ppp2UsvvaRWrVpp1qxZDgOD0tPTtWvXLk2cOFE33XSToqOjNW7cOIfnYn7q3Lmz6XOcPXu2Vq5caSvHxsaaAiZzMnjwYLvy66+/bsqcJV0NmHjhhRd03XXX6ezZswoJCVFUVFTudsIF//d//2cK8liwYIHuvPNOu6kYz58/r8mTJ6tFixbauHGjIiIi3LoWFfS19eTJk3ruuedUtWpV9erVSzNnznR4zZKuBmQ8+uijeuqpp0zLs55H3tKyZUs98cQTpmUvvviievXqpe3bt5uWnz17VuPGjVP79u1NQRM1a9bMMSDNW/LjmleU+Pv767vvvlPJkiVty1atWqVGjRrppZde0p49exyud+LECf38888aPny4qlSpoqFDh2rXrl351q9t27Zp0KBBqlKlih566CEtWLBAp0+ftqtntVq1fPlydevWzZTxKzQ0VP3793fYtifuBX5+fvr0009NU//Onz9fbdu21dKlS02BbJcuXdLkyZPVsmVL/fXXXwoICDAFBOekbNmyuuuuu2xlq9WqW2+9VVOmTLGbBnnbtm0aOHCghg0bJsMw3LoPF9axklsF+b2zZMmS+vbbb01Bfz/++KMaNmyo8ePH69ChQw7XO3z4sGbPnq2BAweqcuXKGjFihMN/NGC1WjVjxgy1aNFCrVq10htvvKGNGzc6nPY6KSlJkyZNUtu2bU3X6dtuu80usykAAACAQswAAAAAAHhMZGSkIcn2t2TJkgJrPzIyMldtpKSkGH369DH1M+MvICDAqFmzptGqVSujefPmRnR0tBEUFOSwblxcXLbbue+++xyuV7FiRaNly5ZG3bp17dr+8ssv3drHl156ye3j/fXXXxvBwcEO+1alShWjRYsWRkxMjFG9enXDYrGY3p88ebLTdnfv3m2EhIQ4bDfrX3btuKtTp06mtt0xbtw4p8ehVatWRoMGDYxixYqZ3vvkk0/c3ua0adNcOi7Zjatbb73Vrq7FYjHq1KljtG7d2qhVq5bh5+dney8sLMzYsGGDqX6nTp0K7Fhm9sQTTzjcNz8/PyMqKsqIiYkxYmJijJo1axqhoaEO6+b3tcOR//znP04/h48//tittk6fPm1UrVrV4TWlYcOGRuvWrY3o6Gi7c2rixIluHffJkye7fS49+uijDvcxOjraaN26tVG3bl0jMDDQttzf39+YN2+e29fbgry2TpgwwWHdUqVKGfXq1TNiY2ONFi1aGOXKlXNY75577smx/+6Ki4szbWPIkCEur3vlyhWja9euDvtauXJl2z0i8+eS8VemTBljw4YNOW4jP+6XuZXXa547160hQ4a4dA11JLfHKDef/eLFi43SpUs7PAZly5Y1GjVqZLRp08aoX7++03HszhjLyR133OFwG5UqVTIaN25sxMbGGo0bNzaKFy+e62ukJ+4Fr7zyisP1ypQpY7Ro0cJo2LCh3X18zJgxbt/v4uPjHR6L4sWLG02bNjVatGhhlC9f3vTeddddZ0ycONHta3ZBj5XcfHfMkJuxX1DfOw3DMGbOnOn0u2elSpWMpk2b2u5zzo7pSy+9ZNfu2bNnHdYNCgoyatSoYcTExBitW7c2atasafj7+9vVK1eunHHo0CGXjysAAAAA7wsQAAAAAACZBAYG6uuvv1ZMTIxGjx5tmjYtLS1N+/fvz3FKx7Jly6pYsWLZ1smYEu6LL74wLT9+/LiOHz9uWmaxWPTWW29p4MCBeuGFF9zZHbf16dNHlStXVr9+/eymJvrrr7+cTveZkzp16ujLL7/UvffeqwsXLuRHVwvcE088od27d9tN3+foOPj5+entt9/W8OHDNXPmTLe2M2jQIG3fvl1vvfWWW9OCZTZ16lR1797dNEWYYRgOM5lERERozpw5atGiRa62lVdvv/22mjVrphEjRigxMdG23Gq1Kj4+XvHx8dmuHx4eXmAZ/DIbMmSIXn/9dbvlwcHB6tu3r1ttRURE6Mcff9TNN9+sEydO2JanpaVpx44ddvX9/Pw0btw4DRs2TF999ZX7nXfDW2+9pbi4OM2bN8+0PC4uTnFxcaZlISEhmjZtmt10mq7w1LU1s8TERNMYc2TYsGH66KOPXG7TE4KDgzVv3jz93//9n6ZPn2567+jRo06nwatdu7Z+/PFH1atXzxPdzLX8uOYVNTfccIPWrVunfv362U31mJCQoISEhGzXt1gsqlatWkF2UdLV6SWzyy5arFgxTZgwwZSp0BlP3AtefPFFpaWl6dVXXzWNtdOnTzvMxPbYY4/p+eef18KFC3Psf2aRkZGaPXu2evbsaZpC+8KFC9qyZYtd/c6dO+v777/XnDlz3NqOdO2MFVcV1PdO6WrWzNq1a6t///5234dyGsvS1WxxlStXdnl7KSkpOnDgQLZ16tWrp7lz5xaqzwAAAABAzpgWEwAAAADg0LPPPqu4uDg9/fTTql69eo71K1asqIEDB+r777/X0aNHVaFChWzr+/n5adKkSZo7d64aN27stF6bNm20ZMkSuyncClKHDh20d+9ejR8/XnXr1s22blBQkLp27aovv/xS/fr1y7Zur169tGfPHo0dO1Y33XSTqlWrpuLFi5umrSpM/Pz8NHHiRE2ZMiXbMdCxY0etXLlSI0aMyPW23njjDW3evFlPP/20OnTooPLlyyskJMTl9SMiIrRkyRK98MILKl68uMM6AQEBGjhwoLZt26brrrsu133ND4MHD1Z8fLxeffVVl6aXLF26tHr16qVp06bp+PHjatasWYH3sW7dumrTpo3d8ttuu02lS5d2u73mzZtrw4YNGjhwoPz9/R3WsVgsuvHGG7V69Wq7qRELSkhIiH744QeNHz9eZcuWddqvW2+9VZs2bVLv3r3ztL2CuLYOHz5cc+fO1f33369atWrl2GZwcLDuuusu/fHHH5o4caICAgrfvz8NCgrSl19+qaVLl6pz587Z9rFmzZoaP368tm/fXugDyzLk9ZpXFNWsWVNr167Vjz/+qM6dOysoKCjb+v7+/mrbtq1eeeUV7du3T6+++mq+9eWzzz7TF198obvvvjvH7zPS1XvQgw8+qF27drkUWJbBE/eCl19+WX/88Yfatm3rtE79+vX13Xff6d1333W571nddNNNWrt2rbp37+70e03FihX1zjvv6NdffzVNb+muwjRW8kNBfe+UpJiYGO3cuVPTpk1TbGys0/tvhuDgYHXu3Fnjxo3T4cOH9cADD9jVKVWqlDZt2qTRo0erffv2pqk9nWnSpInee+89bd261a1ptQEAAAAUDhaDfx4HAAAAAHDBvn37tHnzZp06dUpnz55VQECASpYsqerVq6t+/fqKiorKU/s7duzQxo0bbRlpqlSpotjYWJcCJQrawYMHtW7dOp08eVJnz55VcHCwIiIiVLduXTVr1kxhYWHe7mKBMwxDmzZt0qZNm5SQkCDDMFStWjW1a9dO0dHR3u6eSUpKiv744w/t3btXp0+fVkhIiGrWrKlOnTrlKijKE/766y/bGDt9+rT8/PxUokQJValSRfXr11fNmjXl51d0/o3gmTNntGzZMh08eFDnz59XWFiYoqOj1a5dO5UvX95r/UpPT9eaNWu0Y8cOJSQkKCAgQJGRkerYsaMqVapUINssiGvrqVOntHPnTh04cEBnzpzRpUuXFBoaqtKlS6tevXpq2rTpNXfdSkxM1PLly3X06FGdPn1aYWFhqlChgpo1a5ZjMAauTZcuXdLq1at1+PBhnT59WpcvX1bx4sVVtmxZ1a1bV/Xr1/fYOI6Li9Pu3bt18OBBnTt3TikpKSpevLjKlSunxo0bq0GDBvkSpFnQ94IDBw5o9erVOnr0qFJTU1WxYkW1bNnSLsj/+uuv1++//24ru/sTwokTJ/T777/r6NGjunjxoiIiItSkSROXgptyozCNlfxQkN87z507ZxsDCQkJSk1NVXh4uMqXL6969eqpbt26bge6Jicna+fOndq3b5+OHTumCxcuyGKxqESJEoqKilKzZs1UpUqVXPcZAAAAgPcRXAYAAAAAAAAAACTlPbgMAAAAAFC0FJ1/8goAAAAAAAAAAAAAAAAAyDcElwEAAAAAAAAAAAAAAAAA7BBcBgAAAAAAAAAAAAAAAACwQ3AZAAAAAAAAAAAAAAAAAMAOwWUAAAAAAAAAAAAAAAAAADsElwEAAAAAAAAAAAAAAAAA7BBcBgAAAAAAAAAAAAAAAACwYzEMw/B2JwAAAAAAAAAAAAAAAAAAhQuZywAAAAAAAAAAAAAAAAAAdgguAwAAAAAAAAAAAAAAAADYIbgMAAAAAAAAAAAAAAAAAGCH4DIAAAAAAAAAAAAAAAAAgB2CywAAAAAAAAAAAAAAAAAAdgguAwAAAAAAAAAAAAAAAADYIbgMAAAAAFDkDBs2TBaLRRaLRb179/Z2d4qsoUOH2o6zxWJRfHx8gWwnPj7etJ2hQ4cWin4BuHYsXbrUdF0YPXp0tvWjoqJsdaOiojzSR181evRo02ezdOlSb3ep0Lr++utNxwp51717d9vxfPLJJ73dHQAAAAAolAguAwAAAAAUKevXr9cXX3whSQoICNBrr73m5R4BAACgMPrvf/9rC9T74IMPtHv3bi/3CAAAAAAKH4LLAAAAAABFyuOPPy6r1SpJGjx4sOrUqZNt/axZrhz9BQUFKTw8XNWqVVNMTIx69eqlF198UfPmzdO5c+c8sVsAAADIZy1atFDPnj0lSampqXrqqae83CMAAAAAKHwILgMAAAAAFBnz58/XihUrJEkWi0XPPvtsvrSbmpqqCxcu6MiRI9q4caO+++47jRkzRrfeeqvKly+v2267TT/99JMMw8iX7QEAIDFlJOAJI0eOtL2eN2+eVq9e7cXeAAAAAEDhQ3AZAAAAAKDIGDVqlO317bffrrp16xb4NlNSUvTzzz/r9ttvV0xMDD9IAgAAXENatWql66+/3lbO/H0SAAAAACAFeLsDAAAAAADkh0WLFmnDhg228kMPPZSrdgYNGqTBgwebllmtViUmJioxMVHHjx/X2rVrtWbNGiUkJJjqbdq0SR06dNC4ceP0+OOP52r7yLspU6ZoypQp3u4GgGtYfHy8t7sAwIMefPBBW2bAhQsXavPmzWrWrJlX+wQAAAAAhQXBZQAAAACAIuHdd9+1vY6MjNSNN96Yq3Zq1Kihrl275ljPMAwtWLBAEyZM0MKFC23L09PT9cQTTyglJSXfpuUEAABAwbnrrrtUpkwZnT59WtLV75WTJ0/2cq8AAAAAoHBgWkwAAAAAwDXv4MGDmj9/vq08YMAA+fkV7H/yWiwW3Xzzzfr111/15ZdfKjQ01PT+v//9b1PQGQAAAAqnoKAg9e3b11aeOXOmEhMTvdchAAAAAChECC4DAAAAAFzzZsyYIavVaiv37NnTo9sfOHCgfvvtNwUFBdmWGYahBx98UCkpKR7tCwAAANx311132V4nJydr9uzZXuwNAAAAABQeTIsJAAAAALjmzZgxw/a6SpUqiomJ8XgfYmNj9dZbb2nEiBG2ZQcOHNCUKVP0wAMPuNxOYmKitm/frt27d+vs2bNKSUlRqVKlVL58ebVq1UqRkZF57uuZM2e0ceNG7du3T+fOnVNaWppCQ0NVtmxZRUdHq2HDhipdunSet3P58mUtW7ZMf/75py5cuKDSpUsrKipKnTp1UlhYWJ7b95QtW7Zo/fr1OnnypIKDg1WxYkW1a9dOUVFR+dJ+YmKiVq5cqWPHjunUqVMKCQlRuXLl1Lx5czVo0CBPbZ8/f16bNm3S7t27lZiYqOTkZIWGhto+iwYNGqhChQr5sh+uiIuL044dO3To0CGdO3dOAQEBioiIUGRkpGJjY1W8ePF83+aFCxe0YsUKHT16VMePH1dISIg6deqkFi1aZLteenq61q9fr3379unkyZNKTk5WuXLlFB0drfbt2ys4ODjf++qu06dP648//tCBAwd05coVVaxYUXXq1FG7du3yPXvj9u3btWvXLh07dkwXLlxQhQoVNHjwYAUGBma73smTJ7V69WodP35cp0+fVvHixVW+fHm1bt1a0dHRee5Xamqqfv/9dx04cEAJCQkKCwtT7dq11bFjR4WHh+e5/bwwDENbt27Vn3/+qVOnTuncuXMKDQ1VxYoVVb9+fTVu3Fj+/v5e69/u3bu1ZcsWW98iIiJUuXJldejQQREREXluPzExUUuXLtWRI0eUlJSkiIgINWrUSLGxsQoIKNqPpa1Wq/bu3asdO3bo6NGjSkpKUnBwsCIiIlSrVi21bt26UFxDClpBnv8F8V3m+uuvV8mSJXXu3DlJV79f3n///XnqJwAAAAAUCQYAAAAAANew+Ph4Q5Ltb+DAgW6tP2TIENP6L730Uq77kp6ebtSoUcPUXvPmzXNcb+PGjcazzz5rNGvWzLBYLKb1s/7VqFHDePfdd41Lly653b9FixYZN954o+Hn55ftNiwWi1GvXj1j5MiRxqFDh5y2l/XYxcXFGYZhGImJicYTTzxhhIWFOWw/KCjIeOihh4xTp0651O+4uDjT+kOGDMm2vrN+ObJkyRKnn/+MGTOMunXrOj1Obdq0Mf744w+X9sGRH3/80bjuuuuMgIAAp9uoXr268fbbbxtXrlxxq+0NGzYYd911lxEUFJTtZy3JiI6ONh599FFjx44dud4XZy5fvmzMnj3b6N+/v1GxYsVs++Hv729069bNWLJkiVvb6NSpk6mdDDt27DD69etnhIaG2m1rxIgRTtuLi4sz7rvvPiMiIsJpX0NDQ42BAwcaBw4cyOWRcY2zsRwXF2f07t3b8Pf3d9i/ypUrG+PGjTPS0tJc2o6z8yA1NdV47733jIYNGzrcztmzZx22l56ebkydOtVo2bJltte0+vXrG5MnTzbS09PdPjaXLl0y/v3vfzv9nIKDg41hw4bZrjPZneuOREZG2upGRka61beDBw8aw4cPN8qXL5/tmC9durQxYMAAY9myZab1s/bVnb+cnD9/3hg9erQRHR2d7bl4/fXX2/XLVYcPHzb69u3r9PpTrlw5Y+zYsUZqaqphGIbx0ksvmd539xrgzLfffmtq193vBxlGjx5tauett95yWC8pKcmYOnWqcddddxmlS5fO9nMKDg42evXqZWzYsMGtvji73jnizr0wK3fPl8wK+vzP7+8yWd1yyy22NgICAoxz58651T8AAAAAKIoILgMAAAAAXNM+/fRT04+Jn332mVvr52dwmWEYxoQJE+x+4Dx48KDT+u+//36uAggaNGhg7N2716U+Wa1W49FHH83VdrI7no5+uN6xY4cpKCO7v5o1a7r0Y7eng8uSk5ONAQMGuLQP/v7+xuTJk3Pch8xOnDhhXH/99W59DnXq1DH27dvnUvuvv/56jj+6O/p7/vnn3doPV8TExORq3D3yyCO2wJOcOAq2mD59eraBdc6Cy1599VWXAvIy/oKCgozPP/88vw6XHUdjefHixUZ4eLhL/WvTpo1x+vTpHLfj6Dw4c+aM0aFDh2zbdxRctmfPHqNp06Zufd6xsbHGyZMnXT4uBw4cMGrXru1S25UrVza2bt3qkeAyq9VqjB492q0xlHncZiio4LKffvopx4C3rH8PPPCAy+eiYRjGvHnzjBIlSrjUdocOHYzExMQCCy67cuWKKcgrLCzMOH/+vFttWK1WU9C4v7+/cfToUYd1y5Qp4/bnZbFYjNdee83l/hT24LKCPP8L6rtMVm+++aZp3e+++87ldQEAAACgqMrf/PgAAAAAAHjYH3/8YSq3bNnSSz25qmfPnnbLlixZ4rT+lStX7JaFh4erbt26at26tVq0aKFq1arZ1dm5c6c6duyokydP5tinUaNG6YMPPrBbHhERoaZNmyo2NlaNGzdWpUqVcmwrO/Hx8erSpYsOHjxoWxYZGalWrVqpfv36dtO/7d+/Xz179lRaWlqetpvfhgwZoq+++spWLl26tJo0aaIWLVqoVKlSprrp6em6//77tW7dOpfa3rt3r2JjY7V06VLTcovFoqioKLVs2VKNGze2mx5yz549atu2rfbs2ZNt+5MmTdJ//vMfWa1W0/Lw8HA1btxYsbGxatq0qapVqyaLxeJSn/PC0fiuXLmybWq8xo0bq2TJknZ1PvzwQ7emk81s/vz5Gjx4sFJSUiRJfn5+qlmzpm1aWUfTEKanp2vo0KF68cUXbetlKFOmjJo0aaKWLVuqSpUqpvdSUlJ0//33a8KECbnqq7t27typ22+/XefPn5d0ddxER0erZcuWDqfMXbNmjW666SZbfVelpaXp9ttv1/Lly23LMs6DJk2aOPzMMrbXrl07bdmyxbTc39/fNhVggwYNFBISYnp/9erVatu2rU6dOpVj344eParOnTtr7969DrfRsmVLVa5c2VS/e/fuOnHiRI5t50Vqaqr69Omj0aNH242hoKAg1axZU61bt1bDhg3triOeMHHiRN15551294zQ0FDVr19frVu3Vq1ateymU504caJ69eolwzBy3MbixYt19913KykpyW4bDRs2VLNmzUzTFC5fvly9e/d2qe3cCA4O1j333GMrX7x4Ud99951bbSxfvlwHDhywlbt16+b0Xpn1emexWFStWjU1adJEsbGxatiwoUJDQ011DMPQ888/r1deecWtfhVGBX3+e+q7TNap1ZctW5an9gAAAACgSPBycBsAAAAAAHlSv359U0YRd6cPzO/MZYZh2E3/l90UfG+99ZYRHBxs9OnTx5g2bZrTLGcJCQnGxx9/bFSuXNnU9u23355tX44cOWIEBgaa1hk+fLixc+dOh/XPnDlj/Pzzz8aDDz5olCxZ0q3MZRnZXYoVK2aMGjXK+Ouvv+zaHjlypN00WR999FG2++DJzGWZM9R0797dWLVqlWG1Wm3109LSjDlz5th9Dq1bt862T4ZhGBcvXjSNV+nqlJSffvqpcebMGVPd1NRU43//+5/RokULU/1mzZo5HeNXrlyxmyLw7rvvNtavX2/ahwxJSUnGb7/9Zjz11FNGhQoVCiRzWcOGDY3q1asbTz31lPHbb785nF7MarUaW7ZsMR566CG7aR6///77HLeRNZNPhQoVDElGyZIljfHjx9tNv3r8+HFjzZo1pmWjRo0ytREYGGiMGDHC4VSh+/btM4YPH24axwEBAcaKFSvcPDo5yzqWM7JpWSwW41//+pfd9WLfvn3Gvffea5e158EHH8x2O1nPg4xjKMno2rWrsWLFCtPUdVar1Vi4cKFpLB47dswuK1aTJk2MmTNn2mWLunz5svH1118bNWvWNNXv0aOHw7GaWffu3U3rBAUFGaNHjzZOnDhhqrd9+3ajZ8+etnpRUVFuXevdzVz2+OOP2x33Zs2aGd99951x4cIFu/p79+413nnnHaNVq1ZG1ke0Z86cMRYuXGgsXLjQGDRokKnNcePG2d5z9OfIokWL7LIZ3nbbbcbSpUvtspKdPn3aeOONN+yy440dOzbb/T979qzdva98+fLG1KlTTdM4p6enGwsXLjQaN27s9LPJr8xlhmEYq1evNrXduXNnt9a///77TevPmjXLad2wsDCjXr16xosvvmisXLnSuHjxol2d9PR0Y9WqVcY999xjatff399Yu3Ztjv0prJnLCvr8L8jvMlmdPHnStJ327du7vC4AAAAAFFUElwEAAAAArllXrlwx/WBeo0YNt9soiOCyW265xe5HfGe2b99uFxSRnTNnzthNNegoACbDhx9+aKo7atQol7d1/vx549ChQ07fz3rsJBkRERE5/kA+ZswYuwCM7HgyuCzj78UXX8x2G7t37zZCQ0NN62zevDnbdR588EG7cZHTFG1Xrlwx7rrrLtN6b7/9tsO68+bNM9UbPHhwtm1nlpycbOzfv9/l+q5atmyZkZaW5nL9X3/91QgODrbtgytBe1mDLSQZFStWNHbt2uXSNlesWGG6jpQtW9ZYt25djuvNmjXLtF6TJk1c2p47HJ1jFovF+PLLL7Nd75133rFbZ/Xq1U7rOzsPHn/8cZf7mjXo64EHHjBSUlKyXefs2bNGu3btTOtlF1D49ddfm+oGBwcbv/32W7bbeP755x3uW34Gly1YsMCu/ccee8wUkJedRYsWOX0vr1NGZg368vPzMyZNmpTjejt27DDKlStnWy8oKMg4duyY0/oPPfSQqZ+RkZHGkSNHnNa/cuWK0blzZ4efTX4GlxmGYdStW9d0LmR3X8vs8uXLRsmSJW3rlixZ0rh8+bLT+u72e+rUqab97tOnT47rFNbgsoI+/wvyu4wjmQO1ixcv7ta6AAAAAFAUMS0mAAAAAOCadejQIdP0f1WrVvVib/5RtmxZU/nYsWNO6zZs2FDly5d3ue3SpUtr5syZpqnLpkyZ4rR+1mkUH374YZe3Vbx4cYdTcmZn0qRJatWqVbZ1Ro4cafqsNm/eXOBT1rnjjjvuyHGKsjp16uhf//qXadn//vc/p/UPHz6szz//3FZu0qSJvv32W7vpL7MKDg7W9OnTFRUVZVv27rvvKj093a5uXj7roKAg1ahRw+X6rurYsaPDaSidufHGG/XMM8/YymvXrtXOnTvd3u6UKVNUr149l+q+8sortuuIn5+f5s6d69L0un379tVTTz1lK2/dulWLFi1yu6/ueuSRRzRw4MBs64wYMcI0HaBhGHr33Xfd2k67du309ttvu1R39erVWrBgga18880365NPPlFgYGC265UqVUrfffedwsPDbcvGjx/vtH7W6Udfe+01de7cOdttjBkzRjfeeGO2dfLq5ZdfNpX79++vd999126KSWe6dOlSEN2SJH3yySc6fvy4rfzaa6/pvvvuy3G9Bg0amO4tKSkpDqcklKTExERTXYvFom+//dZuGtnMgoOD9d1336lcuXI570QeDRkyxPbaMAx9+eWXLq33ww8/6Ny5c7Zy37597aZ0zOz66693q1+DBw82ncvff/+9aXvXCk+c/57+LpN5at0LFy7o9OnTbq0PAAAAAEUNwWUAAAAAgGvW4cOHTeVKlSp5qSdmpUqVMpUvXLiQr+3Xrl1brVu3tpVXrlzptO7ly5dN5Zx+7M2L1q1b684778yxXkBAgHr27GlatmHDhgLqlfv++9//ulSvb9++pvLGjRud1v3www+VlpZmK7/11lsKDg52aTuhoaF64oknbOWDBw9q/fr1dvU8+VkXpKyBU9mNb0c6dOigm266yaW6u3bt0i+//GIr9+3bV+3atXN5WyNHjlRAQICt/N1337ne0VwICQmxC2RyZuzYsabgpu+//16JiYkub+uVV16RxWJxqe4777xjKk+YMMHldStWrKj777/fVl6xYoXDYNNdu3Zp9erVtnKVKlX02GOPubSNt956y6V6ubF27VqtWrXKVg4PD9eHH35YYNtzR3p6ut5//31buXr16qaAyJz06NFDzZs3t5Wdje+ZM2earj+9evXKMchYunqvfP75513uT24NGjTIdC5MmzbNpfWmTp1qKmcOUssvma93aWlpWrduXb5vo6B54vz39P0t6/fJrN83AQAAAMDXEFwGAAAAALhmZc3wkVMWKE/J2o+UlJR830Z0dLTt9aZNm5zWy5x9Q5KmT5+e733JkDXYKjvNmjUzlQvLD7eNGzdWgwYNXKrbqFEjU2BRdvswf/582+uKFSuqa9eubvWrW7dupvIff/xhV8eTn3VByjy2pezHtyP9+vVzuW7WbHODBg1ya1tlypRRTEyMrezoc8lPt9xyiyIiIlyqGxkZqU6dOtnKycnJpuCs7FSoUCHHjGAZrFarKWtR69atVbduXZfWzeDK+F66dKmpfM8997gcYNK0aVO7a05++fXXX03lwYMH2wUYe8uWLVt09OhRW9mdY5Yh82fz559/KiEhwa5O1s9m8ODBLrc/YMAA03W0IFStWtU0nnfv3q01a9Zku87x48e1cOFCW7l27dpuBZ66Kq/XO2/z1Pnv6ftb1u9x12JGOQAAAADITwSXAQAAAACuWZcuXTKVixUr5qWemJ0/f95UdjVD1YkTJ/T+++9rwIABaty4sSpUqKBixYrJYrHY/c2cOdO23qVLl+yyemTIOh3cU089pRdeeME0TVp+cWUqwQxZpwItLD/curMPgYGBpiASZ/tw9uxZbd++3VZu0aKFy9PlZahevbqpvGvXLrs6nTt3Nk1BOWHCBD388MM6cOCAW9sqKGvXrtVzzz2nHj16qEaNGoqIiFBgYKDd2M56vjgKZslO5qx+OckaxODO558h82fz559/yjAMt9twlbvT7mWtv3btWpfWa9mypcuZh7Zt22Ya+3k9hpLj8Z2173k9Fvkl6xjq0aNHgWwnN/J7fEs5fzYWi8UU1JiTsmXLqlGjRm73y11Dhw41lXPKXjZ9+nTT9MPuBMxZrVYtXrxYTz31lLp27arIyEiVKlVK/v7+dte7rIFY7l7vvM1T578nv8tIVzOGZnbx4sUC2Q4AAAAAXCsK9p+FAQAAAADgQQUZ1OEOdzOqJSQk6Nlnn9W0adNMP2a7IzEx0WFwXbt27XTjjTfaMrCkpaXptdde0+uvv6527dqpS5cu6tixo1q3bq3w8PBcbTtD1oCx7ISFhZnKzoLjPM2dfZCu7kdGMICzfdi9e7dpbM6fP9/lwB1nzpw5Y7esWrVquu+++/TZZ5/Zln388cf6+OOPFRMTo65du+q6665TbGysy9mv8sMff/yhRx99VFu3bs3V+u5M5SjZZwLKTtYgBnc//6zS09OVlJSkkiVL5qkdZ9wNwslaPy4uzqX18nIMP/roI3300Ucur++Io/Gdte/uHovGjRvnqU/O7N+/31TOTXBNQcn62fTp0yfPbWb9bKxWqw4ePGgrV69e3e17SePGjbV58+Y89y07d911l8LDw23B37NmzdKECRMUFBTksH7m4DOLxeJyVsO5c+fqiSeecPlcy8rd6523eer89+R3GanwfJ8EAAAAgMKCzGUAAAAAgGtW1swSV65c8VJPzE6ePGkqZ53OKbP9+/erefPmmjx5cq4Dy6SrU945M2PGDMXGxpqWWa1WLV++XC+//LK6du2qiIgItW3bVq+99pri4+Nz1YeQkJBcrScVnh9yC2IfTp8+nes2nXGWJe29997TbbfdZrd8w4YNeuONN3TLLbeobNmyat68uZ577jnt2LEj3/uW2aeffqpOnTrlOrBMyn5sO1KiRAmX63rys8kPZcqUyVN9VwNXCuMxzNr3vB6L/JI5EMZisahs2bIFsp3c8MRnk5SUZLr25eY4F9Rnk1loaKh69+5tK585c0Y///yzw7qbNm3Stm3bbOXrr79ekZGROW7jueee05133pnrwDLJ/eudt3nyGuqp7zKSfbB41oB4AAAAAPA1ZC4DAAAAAFyzMk9JKNlPR+ktGzduNJVr1qzpsF5KSop69OihI0eOmJbXrl1bnTp1Ut26dVWlShWFhYXZpsfM8NZbb+nXX391qT9ly5bVsmXL9Pnnn+vtt9/Wvn377OqkpaVp9erVWr16tUaNGqX+/ftr3LhxqlChgkvbgHMFkYnGarU6XB4SEqK5c+dq1qxZevPNNx1mAzIMQ5s3b9bmzZv1+uuv65ZbbtE777yjWrVq5WsflyxZooceesgUeBIQEKAOHTqoTZs2ioyMVPny5RUSEmI3FWbWKdDcERgY6HJdT342+SFrQG1OsgZEXLhwwaX1CuMxzNr3vB6L/JL5vhMaGur2lLcFyROfTV4/F8lzgTtDhgzRF198YStPmzZNPXv2tKuXdcrMIUOG5Nj21KlT9frrr5uWFStWzJZNq3r16ipbtqyCg4NN2dJOnDihgQMHursrhYYnr6Ge/C6TdVwXVDZKAAAAALhWEFwGAAAAALhmVatWzVQ+duyYl3ryj4MHD+rEiROmZc2bN3dY95NPPtGePXts5QoVKmjKlCnq3r17jtuZNGmSW/0KDAzUQw89pIceekjr16/Xb7/9pqVLl2rlypVKSkoy1bVarZo+fboWLVqkpUuXqm7dum5tC2ZZgy1uuOEGPffcc3lqs3Tp0k7fs1gs6tevn/r166edO3dq4cKFWrp0qZYvX26bwjOzefPmadmyZZo3b546duyYp35l9tRTT5kCy2655RZ98sknqlq1arbreTJzT2hoqGn8/+9//1NAQN4el1WsWDGv3XLq0qVLbtW/ePGiqZzTFL25kXV89+vXT/fdd1+e2nSU7TFrANKlS5fcmv4u67HIL+Hh4Tp79qytT1artdAEmGX9bMaOHauYmJg8tdmwYUNT2dHn4q6C+myy6tixo6Kjo22ZxebPn6+EhARTtrm0tDTNmDHDVg4LC9Pdd9+dbbspKSkaOXKkadl9992nN998M8esbLt373Z3NwoVT53/GTz1Xebo0aOmcvXq1d3bCQAAAAAoYgguAwAAAABcs6pVqyY/Pz9blousGcC8Yfbs2XbLbrjhBod1Z82aZSrPmTNHbdu2dWk7madic1fLli3VsmVLjRw5UlarVVu2bNGCBQv09ddfa8uWLbZ6x48fV69evbRly5ZCEyxxLco6TV5ISIi6du3qkW03aNBADRo00IgRI2QYhv7880/9+uuvmj17tpYvX26rd/78efXq1Uv79+/PlwCkPXv2aNOmTbZyo0aN9P3335sy9jiTl7HtrrJly5oCElq0aKHy5ct7bPvuchQcmJ2sU9ZlzfaYH7KO71KlShXI+M7a94SEBLeCywpi+j5JioiIsAWXGYahhISEQjOGsn420dHR+f7ZlCxZUhaLxRZI6u4YlQrus8nKYrFo8ODBevnllyVJqampmjlzpv71r3/Z6ixYsMA0tfXdd9+d4zVx6dKlpqDybt26uRwAXpDXu8zZTt3lapCgp85/Rwryu0zm4LLw8HBFRETke/8BAAAA4FrCk2EAAAAAwDUrODjYlIni0KFDunLlitf6Y7Va9eGHH5qWtWrVSlWqVHFYd926dbZys2bNXA4sk6QdO3bkvqOZ+Pn5qXnz5vrPf/6jzZs367vvvlOxYsVs72/fvl2//PJLvmzLV0VHR5vKjqby8gSLxaL69etrxIgR+uOPP7Rs2TJTYMDJkyf15Zdf5su2Vq9ebSrff//9LgWWSfk3tl1RWD4bV23fvt2t+tu2bTOVs+5vfvDUMaxRo4ap7O6x2Lp1a352x6Z27dqm8vr16wtkO7nhic/Gz89PkZGRtvLhw4ftMkjlpKA+G0cGDx5sCrrKOgXm1KlTTWVXpsTMer17+OGHXe5PQV7vQkJCTOXLly+7vO6pU6dcqldYrqH5+V3m5MmTtoBRSWrSpEm+9xcAAAAArjUElwEAAAAArmmZp/hKT0/Xzp07vdaXDz74wDbdVoYHH3zQYd3Tp08rLS3NVnZnuqY9e/bor7/+yl0nc9CzZ0899dRTpmWZM1zBfVWrVlWtWrVs5b179+rw4cNe7NFVHTt21NixY03L8uuzzjo1rDvje/HixfnSB1dkzSroyW3nxu+//56n+q1bt87P7tjazDw13sqVKwskyLdVq1amcl6PRX7JOpXs/Pnz87X9rJmWMk81mxNPje/Mn41hGFq2bJnL6yYkJHg0oLRGjRrq0KGDrbx+/Xrt2rVLknT27Fn99NNPtveqV6/uNPNoZoX1eleiRAlTOWs/s5M5+D07njr/3ZWX7zJZg3JbtmyZb/0CAAAAgGsVwWUAAAAAgGta1h/2N2zY4JV+rF69Ws8884xpWZ06dTRo0CCH9bMGCKSkpLi8rY8++sj9Drqhffv2pnJupjmDWffu3U3lDz74wEs9MSuozzq34zs5OVlffPFFvvTBFVk/l4kTJyo1NdVj23fXvHnzXJ5G7+DBg6aAquDgYMXGxuZ7n4KCgtS5c2db+eLFi5o8eXK+b+f66683lWfNmuXyZ7VlyxZt3rw53/sk2Y+hL7/8UomJifnWflhYmKns6nSF0tXAn9KlS9vKixcvLpAA7KyfTdZsYNn56quvTIHWnjB06FBTOSNb2ddff63k5GTb8kGDBrk0tWRur3cnTpzQ999/71Ld3MicUU6Saari7KSmpmrOnDku1fXU+Z8bub2/Zf0eed111+VbnwAAAADgWkVwmY9YtmyZbrvtNlWuXFkWi0U//PCDW+tfuXJFQ4cOVePGjRUQEKA777zTYb3k5GQ9//zzioyMVHBwsKKiojz6UBgAAACA77nppptMZXcypuSX6dOnq0uXLqYflP38/PTpp58qMDDQ4TplypRRQECArbx69WqXfmDfvHlzgQeXZf0BNnNwAnLniSeeMH3e77//vjZu3OjFHl1VUJ91xYoVTWVXM8a8+OKLbmXXyauYmBhTZqLDhw/rhRde8Nj23XXlyhW99NJLLtX997//LavVaivfddddKlWqVIH0K2tg7UsvvaRDhw7l6zYaNGigNm3a2Mp//fWX3nvvvVz1Lz+1aNHClAkrKSlJjzzySL61HxERYSpnzY6ZncDAQD3++OO2smEYGj58eL4HUPbr1880BePs2bNdynyVmJio1157LV/74orevXubsm199dVXslqtdkFxgwcPdqm93F7v/vWvf5mC2fJbixYtTOVvvvnGpfXef/99HT161OXteOL8z43c3t8yB+UGBASoS5cu+dovAAAAALgWEVzmIy5evKimTZvqww8/zNX66enpKlasmB577DF17drVab0+ffrot99+06RJk7R7927NnDnTrVTwAAAAAOCuyMhINWzY0FZesmSJR7ZrGIYWLFigbt26adCgQXbZZMaPH2+XzSUzf39/U6DEsWPHNH78+Gy3uW/fPt1xxx1uBQY88sgj+umnn1yeSi05OdkuYCPz1KPInRo1auj//u//bOXLly/r1ltv1apVq9xqZ/HixXrggQccvvfiiy9q+vTpLmcBMgzDbszl12fdrl07U/mTTz7Rvn37sl3n008/1bhx4/Jl++549dVXTVMPvvnmm3rllVfcmn7wyJEjeuaZZ1yeSi4vPvzwQ3311VfZ1nnvvfc0a9YsW9lisWjEiBEF1qfrrrvOFOh76tQpdevWTX/++afLbVitVv3www8aOXKk0zqZA6Uk6fnnn8/xmv/iiy9q4cKFLvcjN1566SVThqsZM2bo8ccfd3kM/fbbb07fy3x/k6TvvvvOrb6NGDFCFSpUsJWXL1+uXr166dy5cy63cfHiRb333nuaNGmSw/dLly6tIUOG2MqGYahPnz7ZBiglJyerV69eOnXqlMv9yC/h4eG66667bOUjR47o008/NV2P27Ztqzp16rjUXtbr3dixY3PMkvXCCy/o22+/daPX7ouJiVH58uVt5dWrV2vGjBnZrjN//nw999xzbm3HE+e/p77LpKWlacWKFbZy+/btVbJkSZe2CQAAAABFGcFlPuLmm2/WmDFjTA9OMktOTtbTTz+tKlWqKCwsTG3atNHSpUtt74eFhenjjz/WsGHD7P41XoYFCxbo999/1/z589W1a1dFRUWpbdu2dinIAQAAACC/9e/f3/b6r7/+0vr163Pd1oEDB7Ro0SLT36+//qpvv/1Wn332mV599VXddtttKl++vG6++Wa7oIXAwEB9+OGHdkEQjmTNivKf//xHTz/9tE6ePGlanpCQoPHjx6tly5Y6dOiQLBaLy/+QZ8WKFbr99tsVHR2tp59+WkuXLlVSUpJdvdTUVC1YsEDt27c3BchUrFhRt956q0vbQvYmTJig5s2b28rHjh3Tddddp3vvvddp5roLFy5o+fLlev7551WvXj116dJFv/76q8P2t23bpkGDBqlKlSp66KGHtGDBAp0+fdquntVq1fLly9WtWzdTZvPQ0FDTuZQXtWrVUtu2bW3l8+fP67rrrtO3335rt59btmxR37599eCDD8owDNWvXz9f+uCq9u3b22VPeumll9SqVSvNmjVLZ8+etVsnPT1du3bt0sSJE3XTTTcpOjpa48aN08WLFwu0r5GRkTIMQ4MGDdJjjz2mw4cPm94/cOCA/u///s8ukGzYsGEFMiVmZtOmTVO1atVs5d27dysmJkaPP/64tmzZ4jAo5OzZs1q0aJGeeOIJRUVF6a677tKaNWucbuOee+4x/aPH5ORkde/eXS+//LJdkNLOnTvVq1cvjRkzRpIUFRWVxz10rmvXrnryySdNy959913FxMTo+++/dzgu9u3bp3fffVctW7bM9h9yNm/eXGXLlrWVly5dqs6dO+uTTz7R/Pnz7e5XWZUsWVLffvutKYvmjz/+qIYNG2r8+PFOM0wdPnxYs2fP1sCBA1W5cmWNGDHCbrxl9t///tcUyBQfH68WLVroyy+/1JUrV2zLrVarfvvtN7Vu3doWVFeQn40zmYPhJNl9flnfz06nTp1MU1AePnxY7du318KFC03j3jAMrVy5UjfeeKPtmlOQ17vAwEC7KUDvvfdejR8/3m5M7t+/X4899phuv/12JScnq2bNmm5tq6DPf099l1m6dKkp8DK/7okAAAAAcM0z4HMkGXPmzDEtu//++4127doZy5YtM/bt22e89dZbRnBwsLFnzx679YcMGWLccccddssfeugho0uXLsbIkSONypUrG7Vr1zaeeuop49KlSwW0JwAAAABwVXx8vGGxWAxJhiTjP//5j8vrDhkyxLZeXv9atmxprFu3zuVtp6SkGM2aNbNrx8/Pz6hbt67Rpk0bo1atWoa/v7/p/eeee86u33FxcQ630bRpU7v2LRaLUbVqVaNp06ZGbGys0aBBAyMkJMSunr+/v/Hzzz+7fOyc9cGRJUuWmNZ96aWXnNaNi4sz1R0yZEi2bbvTL3f64UhkZKRt3cjIyBzrHz582GjcuLHD8RMWFmbUq1fPaNOmjdG4cWOjatWqpnGd03buuOMOh+1WqlTJaNy4sREbG2s0btzYKF68uMN6H3/8sVv7npMVK1YYgYGBdtspXry40bx5cyMmJsaoUKGC3THYuHGjaVmnTp2y3U6nTp1M9XPriSeecHhc/Pz8jKioKCMmJsaIiYkxatasaYSGhjqsu2TJklxv35GsY3nevHlGeHi46VyuUaOG0apVKyMqKsphn5o3b26cO3cu2+3k9TzIsHXrVqNatWoO+1GyZEmjQYMGRps2bYyGDRsalSpVclgvp8/78OHDpvMu4y8gIMCoXbu20apVK6Nq1aqm96pUqWLMmjXLrX1099xOTU01+vTp43CfgoKCjNq1axtt2rQxGjVqZJQqVcquTnZeeeWVbO89rrQzc+ZMh9f5jGtE06ZNjdatWxt169Y1Spcu7bBeTsds4cKFRnBwsN16oaGhRqNGjYzmzZsbERERpve6detmjBo1qkDPI0fS09ONKlWqONzP4OBg4+zZs26198033zhsq3Tp0kZMTIzRrFkzu32vUKGCsWzZMtOynO5v7l7vzp0753A/g4ODjUaNGhktW7a0O1/q1atnzJkzx+1rQkGe/wX5XSazhx9+2HSMzpw549J6AAAAAFDUBQg+79ChQ5o8ebIOHTqkypUrS5KefvppLViwQJMnT9Z///tfl9o5cOCAli9frpCQEM2ZM0cJCQl6+OGHdfr0aU2ePLkgdwEAAACAj4uMjFSPHj00b948SVenJBszZoxpqruCEhQUpO7du2v48OHq0aOHW+sGBgZq7ty56tq1q/bu3WtbbrVatXv3bofrPPnkkxozZozuvffeXPfZMAwdOXJER44ccVqndOnSmjx5sm655ZZcbwf2qlatqlWrVunBBx/UV199ZcrmcvHiRZemEatevbpb2zx27JiOHTvm9P1ixYppwoQJGj58uFvt5qRdu3b67LPPNGzYMNNUrhcuXNCmTZvs6pcuXVrff/+9KbubJ7399ttq1qyZRowYocTERNtyq9Wq+Ph4xcfHZ7t+eHi4SpUqVaB9bNCggebOnas777xTSUlJMgxDBw4c0IEDBxzWb9Wqlf73v/+pRIkSBdqvDI0bN9aGDRs0aNAg/fLLL6b3zp0759JUjDmN76pVq+q3337TTTfdpP3799uWp6Wlma6jGSpXrqwFCxbkOE1hXgUEBGjWrFmqU6eO3njjDdOYT0lJcdg3Vz333HPas2ePpk+fnus27rnnHtWuXVv9+/fXnj17TO/ldI2Qrk7lnPHc0JmuXbtq9uzZ6t+/v86fP29bfunSJW3fvt2ufocOHfTNN99owoQJbuxJ/vDz89OgQYM0duxYu/fuuOMOt8/l3r1767XXXtMLL7xguq6fPXtWGzZssKtfrVo1zZ8/X8WLF3e77+4oUaKEfvzxR910002mcyA5OdnhZ9K0aVPNmzcvV+PVE+d/Zvn9XSYlJUXffPONrXzPPfeodOnSLvcHAAAAAIoypsWEtm3bpvT0dNWpU0fFixe3/f3++++mh3Q5sVqtslgs+uqrr9S6dWv16NFDb7/9tqZOnarLly8X4B4AAAAAgEzTUB48eNBuusq8CAgIUFhYmKpUqaIWLVro7rvv1osvvqh58+YpISFBc+fOdTuwLEP16tW1bt06PfroowoJCXFaLzY2Vr/88ovGjx8vi8Xicvs//fSTPvjgA91yyy0u/VheuXJlPfPMM9qzZ4/uuOMOl7cD14WFhenLL7/U5s2b1a9fP5c+l3r16mnEiBFauXKlli1b5rDOZ599pi+++EJ33323KlSokGObERERevDBB7Vr1658DyzLMGTIEC1btkzXXXed0zohISG67777tGPHDl1//fUF0g9XDR48WPHx8Xr11VdVp06dHOuXLl1avXr10rRp03T8+HE1a9aswPt4ww03aMuWLerVq5f8/f0d1qlUqZLefPNNrVq1SmXKlCnwPmVWrlw5LViwQMuWLdNtt92msLCwbOtbLBY1b95c//73v7V161ZNmzYtx23UrFlTW7du1ciRI50GfwQHB2vYsGHasmWLGjVqlKt9cZfFYtGrr76qXbt2aejQoTme2xUqVND999+vtWvXZlvP399fX375pZYvX65HHnlErVu3VtmyZRUcHOxW/2JiYrRz505NmzZNsbGxTsdPhuDgYHXu3Fnjxo3T4cOH9cADD+S4jVtvvVU7duxQ7969FRQU5LBO2bJl9frrr2vJkiUqWbKkW/uQn5xNfZl1ympXPffcc5o3b56aNm3qtE6JEiX01FNPadu2bR4bly1atNCGDRvUr18/p595qVKlNGrUKK1evVpVqlTJ9bYK6vz3xHeZH374wRSAl3V6YQAAAADwZRYj8z+lgk+wWCyaM2eO7rzzTknS119/rQEDBmjHjh12DxiKFy+uihUrmpYNHTpUiYmJ+uGHH0zLhwwZohUrVmjfvn22Zbt27VKDBg20Z88e1a5du0D2BwAAAAAytGzZ0pYh5Pbbb9fcuXO93CP3XLhwQX/88Yf27dunc+fOqVixYqpWrZpiY2PdzlbliGEY2rNnj/bu3atDhw4pKSlJ6enpCg8PV8WKFdWkSRPVqVPHIxnf8A+r1aqNGzdqz549SkhIUFJSkkJDQ1WqVCnVrFlTDRo0ULly5dxuNy4uTrt379bBgwd17tw5paSkqHjx4ipXrpwaN26sBg0aKCDAc0nt4+PjtWLFCh07dkzJyckqVaqU6tatq3bt2ik0NNRj/XDHX3/9pXXr1unkyZM6ffq0/Pz8VKJECVWpUkX169dXzZo1C/R8GTp0qKZOnWorx8XFKSoqylZOSEjQH3/8oQMHDig5OVnly5dX3bp11b59+0JzHqempmrt2rWKi4tTQkKCLl68qLCwMJUuXVp16tRRgwYN8hRglJKSot9//10HDhxQQkKCwsLCVLt2bV133XUKDw/Pxz1xX3p6utauXav9+/fr1KlTunTpkooXL64qVaqoYcOGqlevnlvBwvnt3LlzWr16tY4ePaqEhASlpqYqPDxc5cuXV7169VS3bt1sg55zcvbsWS1dulSHDx/W+fPnFRERoUaNGqlt27YevfZ4w65du7RmzRqdPHlSaWlpKlOmjOrXr6/Y2FinQXeekJiYaPtMzp07p5IlS6phw4Zq376928GKriiI87+gvst07txZS5YskSR16dJFixYtcmt9AAAAACjKCC7zQVmDy/bs2aO6detq2bJl6tixY47rOwsumzhxoh5//HGdPHnSltJ97ty56tmzpy5cuKBixYrl964AAAAAgMn8+fNtUx9ZLBbt2rVLdevW9XKvACB3cgouAwDk3fr169WqVStbeeXKlWrbtq0XewQAAAAAhUvh+CeMKHAXLlzQ5s2btXnzZklXH0Zu3rxZhw4dUp06dTRgwAANHjxY33//veLi4rR27Vq9/vrrmjdvnq2NnTt3avPmzTpz5ozOnTtnak+S+vfvrzJlyujee+/Vzp07tWzZMj3zzDO67777CCwDAAAA4BE9evRQ+/btJV3NbPHmm296uUcAAAAozN544w3b6x49ehBYBgAAAABZkLnMRyxdulQ33HCD3fIhQ4ZoypQpSk1N1ZgxYzRt2jT99ddfKlu2rGJjY/Xyyy+rcePGkqSoqCgdPHjQro3MQ+jPP//Uv/71L61YsUJlypRRnz59NGbMGILLAAAAAHjM+vXr1aZNG1mtVgUEBGjHjh2qU6eOt7sFAG4jcxkAFKxNmzYpJiZGhmEoMDBQW7duVb169bzdLQAAAAAoVAguAwAAAAAUOcOGDdPnn38uSerVq5e+/fZbL/cIANxHcBkAFKzu3bvrl19+kSQ98cQTevvtt73cIwAAAAAofAguAwAAAAAAAAohgssAAAAAAADgbX7e7gAAAAAAAAAAAAAAAAAAoPAJ8HYHULCsVquOHj2q8PBwWSwWb3cHAAAAAAAALkpNTTWVz58/r6SkJC/1BgAAAAAAAEWJYRg6f/68KleuLD8/5/nJmBaziDty5IiqVavm7W4AAAAAAAAAAAAAAAAAKGQOHz6sqlWrOn2fzGVFXHh4uKSrA6FEiRJe7g0AAAAAAAAAAAAAAAAAb0tKSlK1atVssUXOEFxWxGVMhVmiRAmCywAAAAAAAAAAAAAAAADYZMQWOeN8wkwAAAAAAAAAAAAAAAAAgM8iuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYIfgMgAAAAAAAAAAAAAAAACAHYLLAAAAAAAAAAAAAAAAAAB2CC4DAAAAAAAAAAAAAAAAANghuAwAAAAAAAAAAAAAAAAAYCfA2x0AAAAAAHjf5t+/1oxfx+t0SqIupl/RhQvlFLe9ny6fr6S01DBZrQX7n4+h4UfVqcUkjQ6uqypPvyJFRRXo9oqqY/s36/OvntLhC0d1If2yrqQGKH5HH51LqKe0lDClpwd7u4s2Jcvs1m1Np+qFkl30S4NX9d43SYo7tlfWixckw/B29wq1oJBEtWj+hV4KClfj4S9KLVt6u0sAAAAAAAAAiiiCywAAAADAx73131v0bOp8KUhX/w50lmb+KKWGebQfe0NPK7Dkc/roBas0fbpHt10U/D73Xd2y9nFdDJIUJimpsjT9N+lsTW93zaG/9t6qnZU3KnjlPI15a7ysRnFJlb3drWtGXHIJnWp3h36//6C0ebO3uwMAAAAAAACgiCK4DAAAAAB82MEdK/TclfmS/98L0v2lnz/xeGCZJCkxSvGRkuLiPL/ta1x6aooe/P1ZXSyZaeHiVwttYJlNYpS+P91FVsPP2z259iRGKb6UpPh4L3cEAAAAAAAAQFFGcBkAAAAA+LA3pg1XWujV173OVVFM8e/1nzO1JUm1a17W+NfjVLpUmoKDrAWy/dNnAnVzz4ZXC2khuhIg6cqVAtlWUfbd1JH6s2SKJKlOUqDeaPqVem+/W2mSgoOt+uz9fYqqfkXBwYYsFu9OOblmfbj+9fTfQW8XKmrHsUFXXwcnKvi+Vpo3O0UlPpkslYnwXicLsRtvb6RzSQGcLwAAAAAAAAA8guAyAAAAAPBRf+1Zr0lBOyRJxVOkj5/8Xdd1+yfT1cTPi+n66xsUaB8SEzMVUovpcoCky5cLdJtFjTU9TWN2fiL9nbXsg9hX9N1vvZWWdrX873/7adCwOt7rYBbJ/pkKZ6NlGIFXX9f5Wd0S96lL/euknp290rdrQVhx6VySpLRiuhwoKTlZslolP7K/AQAAAAAAAMh/BJcBAAAAQCFitRqSYchqtcpqNWQYVlnTrTIyXluvvmdYry6zWv9+X8bV+lZDhtUq6991JclI/7stq/XqupLS09L0zMShSgkvIyVL96a21OfTa2rXrqv96NBB6tSp4Pe3WLFMBR/PxGRNtyotLV1Wq1XpaWmypv/9/9ZM/59qVbo1XdZ0q6x/153xzYvaFlhMuhSqFueKy6/9s/r886tthodLI0Z4d7+yMn3mmadfDbii4Kup1jzdpWuK7fhlnC/S1QAz04EFAAAAAAAAgPxBcBkAAAAAeNBHH0lPj0zR5Yt/p28yLJIyZxyy/P3niSxE222v3s/yzosvShZLwfcgKOjqdgxD/2RiKuKZy6ZNkx55LFkXkgKyfP5+mV67E2A1y/Zqo6Sun/7zzqOPSqVL56m7+S4kJFMhJVNwmcUqP0Nk4MqB7filFlOqv5RukfwvXya4DAAAAAAAAECB4IktAAAAAHjQ669Lly8ESYb/1b9C+J9ld94p3XijZ7ZlsWQKlvGRzGVvvSVdOBdc4J9/jRrSU08VWPO5ZoqBSin+z2u/dPkbkvz9s66CTDJnLpPkE+cMAAAAAAAAAO8hcxkAAAAAeNClS3+/8L8iVdiqsFRJfhbJcnUKS4uU6bUhyci0zJBkkcViv9yS5T1Jf7+W7bXl7+X6+3WAxV/VI6JUtlINSVdnI7ztNmngQM9kLcsQEvJ3srLUYrocoCKfucw2BixpUqWNCks1JH/L1Zx1GZ+nxWr7fC0y/v7/q2PC4me9uroMWSyG/CxS+dCyqlq9ofwD/OXnJ7VuLf3rX1KJEl7ZxWyZM5dlCi6zpMvfKoLLcmA7fkaAlB6gy4FpCivi5wwAAAAAAAAA7yG4DAAAAAA8yGr9+0WpeFXs10bHPiqWKdrINxUrJp09q3+mxUxOvnqgiuj0iLYxEJqgkHvb6MJrurrPQUHe7JbHmDKXpWaaFpPMZS4xH79iuhxwvsgHZAIAAAAAAADwnqL5pB4AAAAACilbYJHFKj9DRTaAyh1202JKV4Otiii7MSD51DgwZy7LFFxG5jKXmI6fj0wlCwAAAAAAAMB7fOfpNQAAAAAUAgSX2bNlYkotplR/Kd2iIp2JieCyTIX0TAXOCZeYMpdlZPsrwucLAAAAAAAAAO/iiS0AAAAAeBDBZfYyZy6TVOQzMTkMLrNYvNUdj7NYpOBgB28wLaZLyFwGAAAAAAAAwJP4FQMAAAAAPIjgMnu2TExGgJQeUOQzMdmNAYvFp4LLpCzZtzIwLaZLTMcutZguB6hIny8AAAAAAAAAvItfMQAAAADAgwgus+drmZgYA1k+8wxkLnOJr50vAAAAAAAAALzL955gAwAAAIAXEVhkz9cyMTEGyFyWF6Zjl1asyGf6AwAAAAAAAOBdvvcEGwAAAAC8iMAie76WiYkxQOayvPC18wUAAAAAAACAd/neE2wAAAAA8CICi+z5WiYmxoCzzGW+ezzc4WuZ/gAAAAAAAAB4F09sAQAAAMBDDCNTgUAaG3NwWdHPxERwmZPMZUyL6RJfO18AAAAAAAAAeJfvPcEGAAAAAC+xBRVJPh1YlJUp0MgHMjERXOYkcxnTYrrEPC1m0c/0BwAAAAAAAMC7fO8JNgAAAAB4CcFljjEtpu+NAcfTYpK5zBVMiwkAAAAAAADAk3zvCTYAAAAAeAnBZY6ZMzEV/Wn+CC5zMi0mmctc4mvnCwAAAAAAAADv8r0n2AAAAADgJQSXOeZrmZgILnOWucx3j4c7fC3THwAAAAAAAADv4oktAAAAAHiIYWQqEEhj42uZmGzjwIfHgMPMZUyL6RJfO18AAAAAAAAAeJfvPcEGAAAAAC8hc5ljvpaJicxlTjKXMS2mS3wt0x8AAAAAAAAA7/K9J9gAAAAA4CUElznma5mYCC4jc1le+Nr5AgAAAAAAAMC7fO8JNgAAAAB4CcFljvlSJibDYFpMicxleeFrmf4AAAAAAAAAeJfvPcEGAAAAAC8huMwxX8rEZAssk3x6DJC5LPd86XwBAAAAAAAA4H2+9wQbAAAAALyE4DLHfCkTE2PgKoeZy3z4eLjDlzL9AQAAAAAAAPA+ntgCAAAAgIcQWOSYL2ViYgxc5TBzGdNiusSXzhcAAAAAAAAA3ud7T7ABAAAAwEsILHLMlzIxMQaucpy5jGkxXeFLmf4AAAAAAAAAeJ/vPcEGAAAAAC8hsMgxX8rExBi4isxluRcUJFksfxeK+PkCAAAAAAAAwPt87wk2AAAAAHgJgUWO+VImJsbAVWQuyz2LJVNwXhHP9AcAAAAAAADA+3zvCTYAAAAAeAmBRY6Rucz3xoDDzGU+fDzcZTt+Rfx8AQAAAAAAAOB9PLEFAAAAAA8hsMgxU6Iqw09Wi7IcrKKDMXBVYKCDhUyL6TLbISri5wsAAAAAAAAA7/O9J9gAAAAA4CXmwCJDFunqHHc+zhRbZfVXukVSerq3ulOgso4BXw0uCwhwsJBpMV1mGzKGv9L9VGTPFwAAAAAAAADe53tPsAEAAADAS8ha5ZgvZy6zSD45BhwGl5G5zGVkLgMAAAAAAADgKb73BBsAAAAAvITgMsdMh6CIZ2JiDFxF5rK8sQ2ZjEx/EgFmAAAAAAAAAAqE7z3BBgAAAAAvIbDIMV/OXOarY4DMZXljl7lMKrLnDAAAAAAAAADv8r0n2AAAAADgJQQWOWY6BBmZmMhcVqQ5zlzmu8fDXbZDlJHpTyqy5wwAAAAAAAAA7+KJLQAAAAB4CIFFjpG5zPfGANNi5g2ZywAAAAAAAAB4iu89wQYAAAAALyGwyDHTIcjIxFREszAxBq5iWsy8sQ2ZjEx/UpE9ZwAAAAAAAAB4l+89wQYAAAAALyGwyDEyl/neGCBzWd6YMpfZAs2K5jkDAAAAAAAAwLt87wk2AAAAAHgJgUWOWSyZChmZmIpoFibGwFVkLssb25Axrh4raxE+ZwAAAAAAAAB4l+89wQYAAAAALyGwyDGLJXOwDJnLfIHjzGW+ezzclTlzmaQifc4AAAAAAAAA8C6e2AIAAACAhxBY5FzmTEzpfiqyWZgYA1cxLWbe2IaM9eqxKsrZ/gAAAAAAAAB4l+89wQYAAAAALyGwyLnMmZiKchYmxsBVTIuZN2QuAwAAAAAAAOApvvcEGwAAAAC8hMAi5zJnYirKWZgYA1eRuSxvMmf6k1Sks/0BAAAAAAAA8C7fe4INAAAAAF5CYJFzZC7zLYGBDhaSucxl/5wvV18U5XMGAAAAAAAAgHf53hNsAAAAAPASAoucy5yJKd1PRTZQhjFwFZnL8sY0ZKyWIp3tDwAAAAAAAIB3+d4TbAAAAADwEgKLnLPLXFZEA2UYA1c53GUfPh7uMsXfFfFsfwAAAAAAAAC8iye2AAAAAOAhBBY5ZzsMVv+rWZiKaKAMY+Aqi8VB9jKmxXSZachkZPsrogGZAAAAAAAAALzL955gAwAAAICXEFjk3D/TYpK5zFfYBZcxLabLzMFlZC4DAAAAAAAAUHB88wk2AAAAAHgBgUXO/TMt5t9ZmIpooAxj4B9kLss90yHKyPZXRAMyAQAAAAAAAHiXbz7BBgAAAAAvILDIObvMZYZx9a+IYQz8g8xluUfmMgAAAAAAAACe4ptPsAEAAADACwgscs4WT5SRhUkqksEyjIF/2AeX+fbxcIcp/i4j2x+ZywAAAAAAAAAUAJ7YAgAAAICHEFjknF3mMqlIBsswBv7BtJi5R+YyAAAAAAAAAJ7im0+wAQAAAMALCCxyzhZPlDm4rAgGyzAG/sG0mLlnzlxGcBkAAAAAAACAguObT7ABAAAAwAsILHLun8xlf0/xJ5G5rIgjc1numYZMxlSyRfB8AQAAAAAAAOB9vvkEGwAAAAC8gMAi58hc5nvIXJZ7ZC4DAAAAAAAA4Cm++QQbAAAAALyAwCLnbIfB6i/DIhlSkczExBj4B5nLcs80ZDKy/RXB8wUAAAAAAACA9/nmE2wAAAAA8AICi5zLnLlMUpHNxORwDFgszqoXafaZy8Q54SIylwEAAAAAAADwFJ7YAgAAAICHEFzmnO0wGFejZopqJibGwD9MwWWWNEliWkwXmYaM1V/pFhXJ8wUAAAAAAACA9/nmE2wAAAAA8AICi5zz6cxlPjoGAgMzFfyuBkYxLaZryFwGAAAAAAAAwFN88wm2C06ePKmff/5Zo0aN0s0336yyZcvKYrHIYrFo6NChBbLNmTNnqlu3bqpYsaJCQkIUGRmpgQMHatWqVQWyPQAAAACeRWCRc7bDYP07c1kRzcTEGPiHOXPZ1QND5jLXmIaM4V9kM/0BAAAAAAAA8L6AnKv4pgoVKnhsW5cvX1avXr00f/580/JDhw7pq6++0syZMzVq1Ci99NJLHusTAAAAgPxHYJFzZC7zPabgMsMiicxlriJzGQAAAAAAAABP8c0n2G6qXr26unXrVmDt33fffbbAshtuuEE//PCD1q5dq0mTJqlmzZqyWq0aPXq0Jk6cWGB9AAAAAFDwCCxy7p/D4CcZKrKZmBgD/zAHl109BhYfPh7uMB0iq3+RzfQHAAAAAAAAwPvIXObEqFGj1KpVK7Vq1UoVKlRQfHy8oqOj8307ixcv1qxZsyRJt912m+bMmSP/v/8JcqtWrXT77bcrJiZGhw4d0siRI9W7d2+VLl063/sBAAAAoOARWOScfSYma5HMxGQ3Bqzy2TFgCi6T5Z//JXNZjshcBgAAAAAAAMBTfPMJtgtefvll3XrrrQU+Pea4ceMkSQEBAfroo49sgWUZypYtqzfeeEOSlJiYqM8//7xA+wMAAACg4BBc5pzDTExFMFiGMfAPR9NiSiK4zAWmIWP4F9lMfwAAAAAAAAC8zzefYBcS58+f12+//SZJ6tq1q6pWreqwXs+ePVWiRAlJ0pw5czzWPwAAAAD5i8Ai5xxmYiqCwTKMgX84ylwmieAyF5C5DAAAAAAAAICn+OYT7EJi3bp1SklJkSR16tTJab2goCDFxsba1klNTfVI/wAAAADkLwKLnHOYiakIBsswBv5B5rLcc5jprwgGYwIAAAAAAADwPt98gl1I7Ny50/a6Xr162dbNeD8tLU179+4t0H4BAAAAKBgEFjlH5jLfY85clukY+OjxcAeZywAAAAAAAAB4Ck9svejIkSO2186mxMxQrVo12+vDhw87rZecnKykpCTTHwAAAIDCgcAi5xxmYiqCwTKMgX+Yg8sysVicvIEMDjP9FcFgTAAAAAAAAADe55tPsAuJ8+fP214XL14827phYWG21xcuXHBa7/XXX1fJkiVtf5mD0gAAAAB4F4FFzpG5zPc4DS5DjshcBgAAAAAAAMBTfPMJdiFx5coV2+ugoKBs6wYHB9teX7582Wm9//znPzp37pztL7ssZwAAAAA8i8Ai5xxmYiqCwTKMgX8QXJZ7DjP9FcFgTAAAAAAAAADex6NcLwoJCbG9TklJybZucnKy7XWxYsWc1gsODjYFogEAAAAoPAgsco7MZb6H4LLcI3MZAAAAAAAAAE/xzSfYhUR4eLjtdXZTXUrSxYsXba9zmkITAAAAQOFEYJFzDjMxFcFgGcbAPwguyz2Hmf6KYDAmAAAAAAAAAO/zzSfYhUTVqlVtr48cOZJt3czTW1arVq3A+gQAAACg4BBY5ByZy3xPYKC3e3DtInMZAAAAAAAAAE/xzSfYhUSDBg1sr//8889s62a8HxAQoNq1axdovwAAAAAUDAKLnHOYiakIBsswBv5B5rLcc5jprwgGYwIAAAAAAADwPt98gl1ItGrVSkFBQZKk33//3Wm9lJQUrV692rZOIP+8GwAAALgmEVjknDm4jMxlvoDgstxzeL4UwWBMAAAAAAAAAN7nm0+wC4nw8HB16dJFkrRo0SKnU2N+//33SkpKkiTdddddHusfAAAAgPxFYJFzpmn+MjIxFcFgGcbAPwguyz3ztJh/Z/orgsGYAAAAAAAAALzPN59ge8iUKVNksVhksVg0evRoh3WefvppSVJaWpoeeeQRpWd5GJyQkKCRI0dKkkqVKqX777+/QPsMAAAAoOAQWOQcmct8j4/udr4gcxkAAAAAAAAAT+HfCTuxfPly7du3z1ZOSEiwvd63b5+mTJliqj906NBcbadz58665557NGvWLP3444+68cYb9fjjj6ty5cratm2bXnvtNR06dEiS9MYbb6h06dK52g4AAAAA7yOwyDmHmZiKYLAMY+AfPrrb+cJhpr8iGIwJAAAAAAAAwPsILnPi888/19SpUx2+t2LFCq1YscK0LLfBZZL0xRdfKCkpSfPnz9eSJUu0ZMkS0/t+fn568cUX9cADD+R6GwAAAAC8j8Ai58hc5nssFm/34NpF5jIAAAAAAAAAnuKbT7ALmWLFimnevHn66quvdOONN6p8+fIKCgpStWrV1L9/fy1fvtzptJoAAAAArh0EFjlnzlxWdINlGAP/ILgs93zlfAEAAAAAAADgfWQuc2LKlCl2U1+6a+jQoW5lNOvfv7/69++fp20CAAAAKLwILHLOdBgypvkrgsEyjIF/EFyWe+bMZX9PI1sEM/0BAAAAAAAA8D7ffIINAAAAAF5AYJFzDjMxFcFgGcbAPwguyz0ylwEAAAAAAADwFN98gg0AAAAAXkBgkXMOMzEVwWAZxsA/fHS384XDTH9FMBgTAAAAAAAAgPfxKBcAAAAAPITAIufIXOZ7yFyWe2QuAwAAAAAAAOApvvkEGwAAAAC8gMAi5xxmYiqCwTKMgX8QXJZ7DjP9FcFgTAAAAAAAAADe55tPsAEAAADACwgsco7MZb6H4LLcI3MZAAAAAAAAAE/xzSfYAAAAAOAFBBY55zATUxEMlmEM/MNHdztfOMz0VwSDMQEAAAAAAAB4H49yAQAAAMBDCCxyjsxlvofMZblH5jIAAAAAAAAAnuKbT7ABAAAAwAsILHLOYSamIhgswxj4h4/udr5wmOmvCAZjAgAAAAAAAPA+HuUCAAAAgIcQWOQcmct8D5nLco/MZQAAAAAAAAA8xTefYAMAAACAFxBY5JzDTExFMFiGMfAPgstyz2GmvyIYjAkAAAAAAADA+3zzCTYAAAAAeAGBRc6Rucz3EFyWe2QuAwAAAAAAAOApvvkEGwAAAAC8gMAi5xxmYiqCwTKMgX8QXJZ7DjP9FcFgTAAAAAAAAADe55tPsAEAAADACwwjU8HHA4uy8pXMZYyBf/jobucLMpcBAAAAAAAA8BQe5QIAAACAh5C1yjmHmZiKYLAMY+AfZC7LPYeZ/opgMCYAAAAAAAAA7/PNJ9gAAAAA4AUEFjnnK5nLGAP/6N07U+GmJzRqqaQRI7zUm2sLmcsAAAAAAAAAeEqAtzsAAAAAAL7CHPthyGKI9E1/c5iJqQgGyzgcAz4aXFahgrRxo7Rkw2GVqxypPmcmSz37eLtb1wSHmf6KYDAmAAAAAAAAAO8juAwAAAAAPISsVc75SiYmxoBZ8+ZS8+bVJD3u7a5cU3zlfAEAAAAAAADgfb77BBsAAAAAPIzAIud8JRMTYwD5wWGmvyJ4vgAAAAAAAADwPp5gAwAAAICHEFjknK9kYmIMID/4yvkCAAAAAAAAwPt4gg0AAAAAHkJgkXO+komJMYD84CuZ/gAAAAAAAAB4H0+wAQAAAMBDCCxyzhwsU3QzMTEGkB985XwBAAAAAAAA4H08wQYAAAAADyGwyDnzNH9FNxMTYwD5wXS+FOFMfwAAAAAAAAC8jyfYAAAAAOAhBBY55yuZmBgDyA++cr4AAAAAAAAA8D6eYAMAAACAhxBY5Jw5c9nfwTJFMBMTYwD5weH5QnAZAAAAAAAAgALAE2wAAAAA8BACi5wzHYaMaf6KYLAMYwD5wZy5rOhOIwsAAAAAAADA+3iCDQAAAAAeQmCRc2QuA1xH5jIAAAAAAAAAnsITbAAAAADwEAKLnHOYiakIBsswBpAfHGb6K4LBmAAAAAAAAAC8jyfYAAAAAOAhBBY5R+YywHVkLgMAAAAAAADgKTzBBgAAAAAPIbDIOYeZmIpgsAxjAPnBYaa/IhiMCQAAAAAAAMD7eIINAAAAAB5CYJFzZC4DXEfmMgAAAAAAAACewhNsAAAAAPAQAoucc5iJqQgGyzAGkB8cZvorgsGYAAAAAAAAALyPJ9gAAAAA4CEEFjlH5jLAdWQuA/D/7N13nCQFnffxb1V3T9ydzRlYwrKSFJAgiIhgOAFBEFFPkHyYRR/QM52PPh7HqQQR707QFY7zTpAgoJg4JQdhEclIDrvE3WXTxO6uev6oDlXV1d3VYbqraj7v12uhu6e7K0zVzOzUd78/AAAAAACATuE32AAAAADQId5gkU2wyCWwiSmBYRnCZWiHwKa/BIYxAQAAAAAAAHQfv8EGAAAAgA4pB4ssyRDBIheay4DwaC4DAAAAAAAA0Cn8BhsAAAAAOqSU/TCcGwSLygKbmBIYliFchnYIbPpLYBgTAAAAAAAAQPfxG2wAAAAA6BDCZdVNlSYmwmVoh6lyvgAAAAAAAADoPn6DDQAAAAAdQrisuqnSxES4DO0Q2PSXwPMFAAAAAAAAQPfxG2wAAAAA6BDCZdVNlSYmwmVoh6lyvgAAAAAAAADoPn6DDQAAAAAdQrisuqnSxES4DO0wVZr+AAAAAAAAAHQfv8EGAAAAgA4hXFbdVGliIlyGdpgq5wsAAAAAAACA7uM32AAAAADQIf5wmSERLCqYKk1M3nCZTbgMTZkqTX8AAAAAAAAAuo/fYAMAAABAh7jDZWbxNsEiSVOniam8SZZk0F6H5kyV8wUAAAAAAABA9/EbbAAAAADoEE+4zC7cJlgkaeo0MTEaFe0wVZr+AAAAAAAAAHQfv8EGAAAAgA4hXFbdVGliIlyGdvCGMZN7vgAAAAAAAADoPn6DDQAAAAAdQrisuqnSxES4DO1SOmwS3PQHAAAAAAAAoPv4DTYAAAAAdAjhsupoLgMaUzpnEny+AAAAAAAAAOg+foMNAAAAAB1CuKw675i/5DYxES5Du5QOmwQ3/QEAAAAAAADoPn6DDQAAAAAdEhguM4xurU6keMNlyW1iIlyGdimPxUzu+QIAAAAAAACg+/gNNgAAAAB0SEW4zDAIlxV4xmImuInJHy4zJMJlaEp5LGZym/4AAAAAAAAAdB+/wQYAAACADqkIlxEqKpmKzWVm8TbHAZpAcxkAAAAAAACATuA32AAAAADQIYTLqvM0lxXDMglsYgocjcpxgCaUm8sIlwEAAAAAAACYPPwGGwAAAAA6hHBZdd7mssKYvwSGZQiXoV1Kh02Cx8gCAAAAAAAA6D5+gw0AAAAAHUK4rLrA5jLCZUBVNJcBAAAAAAAA6AR+gw0AAAAAHUK4rDrPrkhwExPhMrRL6bApNv0l8HwBAAAAAAAA0H38BhsAAAAAOoRwWXU0lwGNqWgukyTbrvZ0AAAAAAAAAGgKv8EGAAAAgA4hXFadZ1ckuIkpMFxmGNWeDlRVOmeKTX9SIs8ZAAAAAAAAAN3FlQwAAAAA6BDCZdV5msum0lhMjgE0qdxcVghjSok8ZwAAAAAAAAB0F7/FBgAAAIAOKQeLbBkSjVUu3l1hyDaUyBF/HANol9KhYxsqnSkJPGcAAAAAAAAAdBfhMgAAAADoEFqrqvOOxSyEZRIYlOEYQLuUD51CGFNK5DkDAAAAAAAAoLv4LTYAAAAAdEg592HLsEVrlYtnV9imLEOuJFZycAygXcrNZYXzRUrkOQMAAAAAAACguwiXAQAAAECHlIJFjESs4C3wSu5YTI4BtEvpnGEsJgAAAAAAAIBJRLgMAAAAADqE1qrqaC4DGuNpLiv+dieB5wwAAAAAAACA7iJcBgAAAAAdQmtVdZ7msmITUwJbmDgG0C7lc8Y5hpJ6zgAAAAAAAADoLsJlAAAAANBxtFb50VwGNMbdXCbJGSWbwHMGAAAAAAAAQHcRLgMAAACADvAUCtFaFai8OwwnKJPAFiaay9AupeYym+YyAAAAAAAAAJOHcBkAAAAAdIA3XGbJtOWbBQl3E1Pim8s4BtAif3NZUs8ZAAAAAAAAAN3Fb7EBAAAAoAO8mQ9GIgZxNzEltYWpfBxwDKA15VxiobksoW1/AAAAAAAAALqLcBkAAAAAdABjMeubWs1lHANoDc1lAAAAAAAAADqBcBkAAAAAdIC3UIjWqiDuJqaktjCVN4ljAK1xN/1JSmzbHwAAAAAAAIDuIlwGAAAAAB1Ac1l9NJcB4dFcBgAAAAAAAKATCJcBAAAAQAfQXFafu4kpqS1MNJehXdxNf5IS2/YHAAAAAAAAoLsIlwEAAABAB9BcVh/NZUB45fMlJYnmMgAAAAAAAACTg3AZAAAAAHQAzWX1uZuYktrCRHMZ2sV0/0bHVmLb/gAAAAAAAAB0F+EyAAAAAOgAmsvqo7kMCM9z6NhGYs8ZAAAAAAAAAN1FuAwAAAAAOsAbLrNk2vJVD6G0O2wjsS1M5XBZ4RggXIYmeb98JLftDwAAAAAAAEB3cSUDAAAAADrAWyjESMQgU6G5rLxJhWOAgCGa5G0uS+45AwAAAAAAAKC7+C02AAAAAHQAYzHrK++O5LYwMRYT7eLJJSa47Q8AAAAAAABAdxEuAwAAAIAO8GY+aC4LUh6LORVamDgG0BqaywAAAAAAAAB0AuEyAAAAAOgAmsvqK4/FTGYLE8cA2sk7UTW5bX8AAAAAAAAAuotwGQAAAAB0AM1l9bmby+wEtjBxDKCdaC4DAAAAAAAA0AmEywAAAACgA2itqq+8OwwnKJOwFiaOAbSTp7ksoW1/AAAAAAAAALqPcBkAAAAAdACtVfV5msukxAVlOAbQTjSXAQAAAAAAAOgEwmUAAAAA0AG0VtVX2h12oblMSlTAjGMA7eQ9dAxnlGyCzhcAAAAAAAAA0UC4DAAAAAA6gNaq+jzNZcVdk6AmJo4BtJN3LCbNZQAAAAAAAAAmB+EyAAAAAOgAb2uVJdOWLx2Ccs4qmc1lntxP8RggXIYmecdiGokcJQsAAAAAAACg+7iSAQAAAAAd4A0WMRIxSLm5rBCUkRIVlgkci0nAEE3yHjqMxQQAAAAAAAAwOfgtNgAAAAB0ACMR6yvtjuKIPylRY/44BtBO3uYyxmICAAAAAAAAmByEywAAAACgAwJbqwgWeZSbmAotTFKimpg4BtBOnuYyxmICAAAAAAAAmCSEywAAAACgA2itqo/mMiA8mssAAAAAAAAAdALhMgAAAADoAFqr6is1MRVbmKRENTFxDKCdPM1lxba/BJ0vAAAAAAAAAKKBcBkAAAAAdACtVfV5msuKf1tNUBMTxwDaieYyAAAAAAAAAJ1AuAwAAAAAOoDWqvrKTUzOfrGlRDUxcQygnTzNZcW2vwSdLwAAAAAAAACigXAZAAAAAHQArVX1uZvLJDlj/hLUxMQxgHaiuQwAAAAAAABAJxAuAwAAAIAO8LZWWTJt+aqHUNoddjKbyzy5n+IxQLgMTfJ++TCc0yZB5wsAAAAAAACAaOBKBgAAAAB0gDdYxEjEIP7msqQ1MQWOxSRgiCbRXAYAAAAAAACgE/gtNgAAAAB0ACMR6yvnrArNZQlrYuIYQDt5com2kbimPwAAAAAAAADRQLgMAAAAADogsLWKYJHHlGwu4xhAk2guAwAAAAAAANAJhMsAAAAAoANoraqvHC4rNJdJiWpi4hhAO3knqhqJa/oDAAAAAAAAEA2EywAAAACgA2itqq8Ulik0l9kJa2LiGEA70VwGAAAAAAAAoBMIlwEAAABAB9BaVV95dzg3rIQ1MXEMoJ08zWW2kbimPwAAAAAAAADRQLgMAAAAADqA1qr6ys1lKed/UqKamDgG0E40lwEAAAAAAADoBMJlAAAAANABtFbV5w3L0FwG1OJpLpPhjJFN0PkCAAAAAAAAIBoIlwEAAABAB3hbqyyZtvzpkCmvYsxfwpqYPJtSPAYIl6FJNJcBAAAAAAAA6ASuZAAAAABABzASsT7v7jCS3VzGMYAWVYQxpUSdLwAAAAAAAACigXAZAAAAAHSAt1CIkYhBvGEZ0wnLJKiJKXAsJu11aBLNZQAAAAAAAAA6gd9iAwAAAEAH0FpVnzcsQ3MZUIs3l1gYI5ug8wUAAAAAAABANBAuAwAAAIAOCGytIljkUdFclrAmJo4BtBPNZQAAAAAAAAA6gXAZAAAAAHQArVX1eXcHzWVALf6mP1tK1PkCAAAAAAAAIBoIlwEAAABAB9BaVV9Fc5mUqCYmjgG0k/98obkMAAAAAAAAwGQgXAYAAAAAHUBrVX3+Jiaay4Dq/E1/dsLOFwAAAAAAAADRQLgMAAAAADqA1qr6PE1MCQzLcAygnbzNZYzFBAAAAAAAADA5CJcBAAAAQAd4W6ssmbb8aaopz9tclrwxf55NKR4DhMvQpKSfLwAAAAAAAACigSsZAAAAANABjESsL+lNTBwDaKekN/0BAAAAAAAAiAbCZQAAAADQAd5CIUYiBkl6E1PgWEza69CkpJ8vAAAAAAAAAKKB32IDAAAAQAfQWlVf0puYOAbQTklv+gMAAAAAAAAQDYTLAAAAAKADAlurCBZ5JL2JiWMA7ZT08wUAAAAAAABANBAuAwAAAIAOoLWqPm9YJnlNTBwDaKekN/0BAAAAAAAAiAbCZQAAAADQAbRW1ecd85e8JiaOAbQTzWUAAAAAAAAAOoFwGQAAAAB0AK1V9Xl3R/KamDgG0E7eMGbymv4AAAAAAAAARAPhMgAAAADoAFqr6vM3l9lSopqYOAbQTjSXAQAAAAAAAOgEwmUAAAAA0AG0VtXnDcsYTlgmQU1MntwPxwBa5AljJrDpDwAAAAAAAEA0EC4DAAAAgA7whsssmbb86ZApr6K5LGFNTIHHAOEyNInmMgAAAAAAAACdwJUMAAAAAOgARiLW590dyWsu4xhAO3nDmIYzRjZB5wsAAAAAAACAaCBcBgAAAAAdwEjE+iqay6RENTEFjkalvQ5NorkMAAAAAAAAQCfwW2wAAAAA6ABaq+rzhmVoLgNq8eYSDWeMbILOFwAAAAAAAADRQLgMAAAAADogsLWKYJFHRXNZwpqYOAbQTjSXAQAAAAAAAOgEwmUAAAAA0AG0VtXn3R00lwG1eMOYhjNGNkHnCwAAAAAAAIBoIFwGAAAAAB1Aa1V9Fc1lUqKamDgG0E40lwEAAAAAAADoBMJlAAAAANABtFbV5w3L0FwG1OIJY8pwxsgm6HwBAAAAAAAAEA2EywAAAACgA2itqq+iuSxhTUyeTeEYQItoLgMAAAAAAADQCYTLAAAAAKADvOEyS6Ytf/XQlOfNWSW8uax4DBAuQ5O8YUzDGSOboPMFAAAAAAAAQDRwJQMAAAAAOoCRiPVVNJdJiWpi4hhAO9FcBgAAAAAAAKATCJcBAAAAQAcwErE+b1gm6c1lHANojb/pz07Y+QIAAAAAAAAgGgiXAQAAAEAH0FpVX0VzWcKamAKPAUajokn+84XmMgAAAAAAAACTgd9iAwAAAEAH0FpVn7+JieYyoDp/058tJep8AQAAAAAAABANhMsAAAAAoANoLqsv6WEZjgG0k7f0jrGYAAAAAAAAACYH4TIAAAAA6ABaq+pL+pg/jgG0kzeMmbzzBQAAAAAAAEA0EC4DAAAAgA6gtao+/1jMpDUxcQygnbxhzOQ1/QEAAAAAAACIBsJlAAAAANABtFbV528us6VENTF5NoVjAC2iuQwAAAAAAABAJxAuAwAAAIAO8IbLLJm2fGkqeMMyhhOWSVATU+AxQLgMTfJ++Uhe0x8AAAAAAACAaOBKBgAAAAB0ACMR66toLktYExPHANqJ5jIAAAAAAAAAnUC4LITnnntOp59+unbYYQcNDg5q9uzZ2muvvfS9731PIyMjLb33JZdcIsMwQv255JJL2rNBAAAAADqOsZj1eXdH0pvLOAbQGm8Y03DGyCbofAEAAAAAAAAQDelur0DU/epXv9Kxxx6rjRs3lh4bGRnRypUrtXLlSv3kJz/R9ddfr2XLlnVxLQEAAABEnbdQiNaqIBXNZVKimpgCm8sYjYom0VwGAAAAAAAAoBMIl9Vw33336cMf/rBGR0c1bdo0feUrX9GBBx6o0dFRXXbZZfrxj3+sxx9/XIceeqhWrlyp6dOnt7S83//+91q8eHHVj2+xxRYtvT8AAACA7qG1qj5vWIbmMqAWby7RcMbIJuh8AQAAAAAAABANhMtqOO200zQ6Oqp0Oq0//OEP2nfffUsfO+igg7T99tvrS1/6kh5//HGdc845+uY3v9nS8pYvX66tt966tZUGAAAAEEmBrVUEizwqmssS1sTEMYB2orkMAAAAAAAAQCcwf6OKu+++W7feeqsk6eSTT/YEy4pOP/107bjjjpKk888/X9lstqPrCAAAACA+aK2qz7s7aC4DavGGMQ1njGyCzhcAAAAAAAAA0UC4rIprrrmmdPvEE08MfI5pmjruuOMkSevXr9eNN97YiVUDAAAAEEO0VtVX0VwmJaqJiWMA7URzGQAAAAAAAIBOIFxWxW233SZJGhwc1B577FH1eQcccEDp9u233z7p6wUAAAAgnmitqs8blklec5kn98MxgBZ5wpgynDGyCTpfAAAAAAAAAEQD4bIqHn30UUnSsmXLlE6nqz5vhx12qHhNs0488UQtXrxYPT09mjt3rvbZZx99/etf1+rVq1t6XwAAAADd5w2XWTJt+dMhU15Fc1nCmpgCjwHCZWgSzWUAAAAAAAAAOoErGQHGxsa0Zs0aSdIWW2xR87mzZs3S4OCgJOmFF15oabk33XSTXnrpJWWzWa1du1Z//vOfdeaZZ2rZsmW68MILQ73H+Pi4Nm7c6PkDAAAAoPsYiVifd3ckr7mMYwDt5A1jGs4Y2QSdLwAAAAAAAACioXol1xS2adOm0u1p06bVff7g4KCGh4e1efPmppa37bbb6gMf+ID23XdfbbnllpKkp59+WldddZWuvPJKjY2N6ROf+IQMw9Cpp55a873OOussfetb32pqPQAAAABMHsZi1lfRXCYlqomp4hggXIYW0FwGAAAAAAAAoBMIlwUYGxsr3e7p6an7/N7eXknS6Ohow8s68sgjdfzxx8vwXVDYa6+99OEPf1i//vWv9YEPfEDZbFZf+MIXdPjhh2vhwoVV3+8rX/mK/s//+T+l+xs3biwF1gAAAAB0jzfzQbAoiDcsQ3MZUIt3qq7hjJFN0PkCAAAAAAAAIBoYixmgr6+vdHtiYqLu88fHxyVJ/f39DS9rxowZFcEyt/e97336xje+IUkaGRnRihUrar5fb2+vhoaGPH8AAAAAdB/NZfVVNJclrIkp8Bgw+Ws5mkNzGQAAAAAAAIBO4LfYAaZPn166HWbU5fDwsKRwIzSbceqpp5YCaDfffPOkLAMAAADA5KK1qj7v7qC5DKjFG8Y0nDGyCTpfAAAAAAAAAEQD4bIAfX19mjNnjiRp1apVNZ/7+uuvl8JlkzV+cv78+aX1Wb169aQsAwAAAMDkormsvormMilRTUwcA2gnmssAAAAAAAAAdALhsip22mknSdKTTz6pXC5X9XmPPfZY6faOO+44aetTa3QmAAAAgOijtao+b1iG5jKgFu9EVcMZI5ug8wUAAAAAAABANBAuq+Jtb3ubJGfk5b333lv1ee4xlfvtt9+krMtrr72mNWvWSJIWL148KcsAAAAAMLlorarP38RkJ6yJiWMA7URzGQAAAAAAAIBOIFxWxRFHHFG6ffHFFwc+x7IsXXrppZKkmTNn6sADD5yUdbnoootkF65CHHDAAZOyDAAAAACTi9aq+vxNTElrLvPmfjgG0BrvGFnDGSOboPMFAAAAAAAAQDQQLqti77331v777y9JWrFihe68886K55xzzjl69NFHJUmnnXaaMpmM5+M33XSTDMOQYRg64YQTKl7/7LPP6r777qu5Hr/+9a/1//7f/5Mk9ff368QTT2xmcwAAAAB0mbe1ypJpy5+mmvIqmsukRDUxBR4DhMvQJJrLAAAAAAAAAHRCutsrEGXnn3++9ttvP42Ojuo973mPvvrVr+rAAw/U6OioLrvsMl100UWSpOXLl+v0009v+P2fffZZHXjggdp333112GGHadddd9X8+fMlSU8//bSuvPJKXXnllaXWsrPPPltLlixp3wYCAAAA6BhGItZX0cSUsOYyjgG0k/fQSd75AgAAAAAAACAaCJfVsPvuu+vyyy/Xscceq40bN+qrX/1qxXOWL1+u66+/XtOnT296OXfeeWdgM1rRwMCAzjvvPJ166qlNLwMAAABAdzESsb6kNzExGhXt5A1jJu98AQAAAAAAABANhMvqOOyww/TAAw/o/PPP1/XXX69Vq1app6dHy5Yt09FHH63PfOYzGhgYaOq999hjD/3sZz/TnXfeqZUrV+qll17SmjVrlMvlNGvWLO2888565zvfqVNOOaXUaAYAQLe98orU0yMZ/es1lhvTQnNIWrNG2mqrbq8aAEQarVX1eaeEGs5YzAQ1MXEMoJ28YczknS8AAAAAAAAAooFwWQhLly7Vueeeq3PPPbeh173jHe8ojbQMMn36dB1zzDE65phjWl1FAAA64sEHpT32kNIZW8bn9tLEwLP685Wz9OYH10jXXSe9733dXkUAiCxaq+rzh2WshI35CzwGvIk6ILSKMGbCzhcAAAAAAAAA0cBvsQEAQGjHHy9ls9LoiKGR3/yTclZOx+z/mnMh87DDur16ABBptFbV5x/zZydszB/HANop6WNkAQAAAAAAAEQD4TIAABDaunWuOxPTJEmv93dnXQAgbmguq8+7O6ZIcxnHAJrkDWMyFhMAAAAAAADA5CBcBgAAQvM3rrj+BwCog9aq+iqay6RENTFxDKCdaC4DAAAAAAAA0AmEywAAQGj+xhVJ4pI4AIRDa1V93rBM8prLvLkfjgG0xhPGlOGMkU3Q+QIAAAAAAAAgGgiXAQCA0GguA4Dmeb+GWjJt+dMhU15Fc1nCmpgCjwHCZWgSzWUAAAAAAAAAOoErGQAAIDSaywCgeYxErM+7O5LXXMYxgHbyhjEN5yezBJ0vAAAAAAAAAKKBcBkAAAiN65UA0DzGYtZX0VwmJaqJiWMA7URzGQAAAAAAAIBOIFwGAABCYywmADTPk/mgtSqQNyxDcxlQi3eqruGMkU3Q+YL2e/S+xzV90d0aXPAX7fmWn3V7dQAAAAAAABAT6W6vAAAAiA/GYgJA82itqq+iuSxhTUyBx4DJv/lCcwKby/LJOV/QfvlsXptf3luStGraa11eGwAAAAAAAMQFv8UGAACh0VwGAM2jtao+7+6guQyoxRvGNJzYf4LOF7Rf/2Cfc8PIqXDEAAAAAAAAAHXRXAYAAEKjuQwAmkdzWX0VzWVS8pvLOAbQpMDmsgSdL2i/bXZYquxETulMWtJ7u706AAAAAAAAiAnCZQAAIDSaywCgebRW1ecNy9BcBtTinahqOGNkE3S+oP3MlCkzxRADAAAAAAAANIbfKAEAgNBoLgOA5tFaVZ+/iclOWBOTd1M4BtAamssAAAAAAAAAdALhMgAAEBrNZQDQPO/XUEumLX/10JTnb2JKdnNZ4RggXIYmecfIGk7sP0HnCwAAAAAAAIBoYCwmAABoEhcvAaARjESsr6K5TEpUExPHANqJ5jI049gPrdDImDR3pqGLLj2p26sDAAAAAACAGCBcBgAAQgtsLuvOqgBA7DAWs76KJqYkN5dxDKBF/qa/pJ0vmBz/fe2HpInp6p39iC7q9soAAAAAAAAgFpjBAgAAQvNfFJcYiwkAYXkKhWitCpT0Jiaay9BOST9fMDkMMytJsqxMl9cEAAAAAAAAcUFzGRBx//nDX+snv1ij19YPKJtPa6elr2vRXENPvmBo1+Wm/vyQ9OPz36JFW87XP33lWt33N1OH7J/Rr27J6YQjp+uTXzxS//G9X+qyX2/QLtvZeu5lSwO90mVXnqh7b3tA/3L2vRoeNbTnLhndeI+tf/vO7lq201L905ev1j0PG3rv2zL67W05ffjgAX3+6x/UivOv08+uWas3bGVrh51+LdmWPnvGFUqlw/1i2ralm26Sfv976W8PPK5nXnpCwxN5WbYtW3bgP7Tv7VuvhUtv06ur9tb8Le7Wuhf31sFv6dNt99nae2fpgScsbbvE1strbfX2SAO9hl5ZZ+ijh/9VDz7/lDa9cpIefvZ1HbRXSr+7w9S3z9hGBx76Fv3z136hP/45rwP2NHXbX/Lad1dT37ngOF37sz/p3y55VrOHbL3nHXfpxfXP6Auf+7kGZ87zrFc2K113nXTzzZb+9tBftXrNKo3nbVm2rYVLb9bI5gXK9G7W9m/6b+06Z0d94fQrlekbaPmYWLVxla54+Ar9bfX9yj71uPKvr9WsbFqfy++pbT75VekNb2h5Gdl8Vtf97Trd8tzNGnnmcWVXrdKLz7xJLz+3j2bOfUwnLblPJ7zpOOkTn+CC6BQT1FxmEi4DgFBoraqvoolJSlQTE8cA2qmi6U9K1PmCyXHacb9SLm9r3pweSdt3e3UAAAAAAAAQA4TLgAj77CmX6IcrTvA89vQj5ds33uT8/10f/p1GxtZq44snSpLuvMN5/O47pRvvWKErrjlZknTLLeXXZo9YoWt+9/dSbjdJ0g1/dB5/90dvkmG8prXPn+B5r7vukO65f4X+50rnvW6SpE98X1r4oOZc+Bl97NMXhtqmc86RvvjF4r3lhT/1/e2vzvo8eu+pkqQH73Eev+N25/9/CnjNH8bPlO77R2nzYknS7bc6jx964nNavu0v9ej9x3re49ZbpVfX/lSX/Pyk0ntc3nuttMMfNX7ukfr2/7vN8/4f/ah05ZWSUwL55sIfx9OPHlG6fc/Ce/U/M36rrX/2NX3olPNCbW81L2x4QTv/+87aNLGp/GC/82fVw4/ois+97CT3WnTM1cfoikeuKD9w3+ek351funvradtoz29+Sru86U3Sfvu1vDzER2BzWVfWBADih9aq+pLexBR4DJgUiqM5ST9fMDnOu/D4bq8CAAAAAAAAYobfYgMRNTE2oQuvfmuo57789Hu18cXggM8Vv/lQ4OPX/PpkKVfZorXuhXdo7fMHBb7mf67y/RJ6/daSpH9//Geh1nN4WPrXfw311Pa49WulYJlbftNSPfrAUYEvcQfLJJW28Z9Tt3sefuCBYrAshL86++3Zlx8L+YLqzr7jbG+wTCrme/TsTEnPPNPyMu5/+X5vsCzbJ936Ve+TNm7hLO/ZZ1teHuKFMgwAaB6tVfVVNDEZStQ3H44BtFNF01/CzhcAAAAAAAAA0UBzGRBRXz/jMmVfP06SlJr+nD74rv/V5b88ufE3mpjevpWyfV8ysk44LZ0P9/IVK6S1awt3Ft8jHX6yFo5buvBdZyuTySid7pFppjyvOfLYvbRpc7iRmw3xb0s12eAxlt/7nuvObhdL+56jt2+ermkvX67f3LBV4GsmJkYbXEmvtSNr9eO//Lh0/7c/k3666j90hXG09P4TNTHzV9LEREvLkKTv3vHd0u0T75OW/PEE/fPwAu+T8j2aSKkty0O8BI3FNLiGCQCh0FxWX0UTk6lENTFxDKCdaC4DAAAAAAAA0AmEy4CIuv72ntLtr596m2bOnKHLf9nFFQqS65ckpaxwyZIbbnDdOfwUaeGDumjbr+uwj7636mumD0mbNreyki0qbKNfcfJkqne98od8Rqn0iC7/wl/0n1dvpd/cEPgSTeTGWlqVO164Q6M5J6D2qbulXZ9cqCv0CeeDl12niU8bbQl73fCUswEzJ1L64W/yOjb7nson5TOEy6YoxmICQPNoraqvoolJSlQTkzf3wzGA1njDZck7XzA5Xn7+FW3asFnjoxPaZe8du706AAAAAAAAiAHCZUBEbRieVrp98CE76ZGHVnVxbaooNZeFu4CxZo3rzryHJUlv3P3var5mILg4rHMK25hyXQi0rHIDmznzCeV7RjRv2NTC7Xevub4TudaCWGtGyjvwja9Kr2q+9/3bEPaybEtrR52N225jSgPZvNZkFkpZ3xNpLoNEcxkANIjWqvqS3sTEMYB28o6RTd75gsmx7Z4va/S1XaXUuOxct9cGAAAAAAAAcWDWfwqAbhgeHSrd3mb5Eg3N6Ovi2lSRdVq9wobLioGsdM8GKeXM0py9YOuar+l+uKywja5rNBs2lK/Z5AfXSZJm55ysbn9w0ZnzVrnxllalGPqSpNmj0iZ5R55mU5Ky/hRYYzaOb5RlOxs3e7PzOVprzKt8Yr5HWbP15SF+aC4DgOZ5g0WWTFv+qq4pzxuWMWQbSlQTU+AxQLgMTfIeOsk7XzA5TLPwd7h8prsrAgAAAAAAgNiguQyIqLGxGc6N1LjmLpqrWbOn1X5BNxSby3Lh/nX8OieHJbPfCUml89L0OYtrvqZWWKsjitvo2sTidkiSNeBsyxzLCf/VbC7Lt9bytW60vOA5I9IGzfC+fxuaxNaOlANsczY54bJ19szKJ9JcNmV5yjBoLgOAhjAWsz5/WMZKWFiGYwDtVBHGlBJ1vmBybL/l37S6f71SZk657LuUzvCrQQAAAAAAANTGb5CAiMqOz5YkmX1rZaYWa+asCIfLQjSXWZb0+uuFO4Vw2ZxxU0adto7uN5fVDpep37kzx3CeF7y+zgXDVsNlnuDXqPSA5ng+3o6wlyfANur0Uq3Nzah8IuGyKYvmMgBoHiMR6/OP+bOlRI354xhAOyV9jCwmx333HNPtVQAAAAAAAEDMMIMFiKj8mBMuy/Q5iaw58wMCPt2Wc2rFUvn6FzA8oyQLbV+zs/XzrV1vLituo6XS1cCgcNnslDOistb6ttxcNlZe8OxRaZ1me98/JWcn5/NNL8M/enNYg8raAeNSCJdNWf6L4q7/AQDq8GY+aK0K4g3L0FwG1OL9dzqMxQQAAAAAAAAwOQiXARG0fs16KTsoSerrWy9JmrNgdvUXFBgDrwQ+bg6+XPU11T4W6vEGxmKuLWeWlB/0jpKspS3NZe3YRkvS5s2SqjSX9Tjhv8D1tQvNZVYu/DoH8DSXjUhrg5rLpJYCX/7Rm/5llFgZwmUAADSI1qr6KprLEtbExDGAdqK5DAAAAAAAAEAnEC4DIujpx14o3R7o2+j8f9qAZNRupOqb9mLg472DLwW/IDWmTO+6wA/1DQa/l+fx0lhM1f0X8oFtX0b95FhbwmVDqwMfrrZf+tyPu8NlGzZI8gblVGxh65/l3A1a35wTosta2fDrHKAY/OqxTA1kK5vLsqnCkMJs88txB9j87WizZrmemO9RNtXashBP3ovizsVLLokDQDi0VtXn3R1TpLmszph4oBpvGNNw/i6QoPMFAAAAAAAAQDTwW2wggp57ptxANn3AacsyU6aUGan5uhnTgpvLqj1u9G5QKjMa+LGhaa/WfzzrzIA0JGmk9rp5Aln9heay9PSar5HaNBZzenC4rNp+GehfJ5mFRq7CNqYtSevXS6rSXDY4z7kbtL6F0ZoTdovNZYWRlXMmUjIU3CqWM9W+5rJR7zIWLnQ9kbGYU5b/orjEWEwACIvWqvoqmsukRDUxcQygnWguQzMOee9PNX+bGzRnq5v08MrHur06AAAAAAAAiAHCZUAEPf/8+tLtGYNjpdtGOjgIVjRnaH3g43NnBD9upiaUTo0FfqzaazyPF1q9sqZKwatqvOGyQnNZz1DN10htai4rjBb1q7aNg/2bykG+gOaywBa26QskVVnfwntMqHbzXD3FVrHZY85VpLWp+RXPaTXwVQywSU5zGeEy+AWVYXBJHADCobmsPm9YJnnNZd7cD8cAWuMtvTOcMbIJOl8wOR5+do5ee/bdWvfCO7T21Y3dXh0AAAAAAADEAOEyIIJefLncAjZ7ennsoJmu3Q62eE7wx6s9LiOvTCY4XBbqvYqNXCmVglfVeAJZhVGSc/orm7f82tJcVqXxrdo2Dg2MlPd1YRtTtmqGy+bMXuLcDVrfYrisheay0eyoRnNOuHDOsHPBaF1qXvkJheBhq4EvT3PZiHcs5qJFricSLpuy/I0rrv8BAOqgtaq+iuayhDUxcQygnWguQzPSqfIxMjo63sU1AQAAAAAAQFyku70CACq9/Fo5hDR/VvnxVHqsZvfVksoiK0nSNouDL1gZhqWedPAvk5cuCnxYi+dJMvKSnSqHpqqEy+68Uzrz3HV6+PlH9Ozdbyt/oNj2NTg3eCEubWkuqxIuq7ZfZgxOyEyPypJK25iqMxZz9pwtqq9vYbTmhNHYhZ7hYenr/5TT7//ysF5d87ykyyRJT22UPizpLxNvLD85n3GW0WDg65JLpIsuf16rXn5SGhnWC5nDJB0mSfriBulafbj0XG9zWSYW4bJf/EL64X+u1nMvPy578+Y6z7a1aOubdNLWt+jjOx4rnXYaF3sDBI7F7MqaAED8eINFlkxb/uqhKc/7rTd5zWWBxwA/b6BJ3jCm4fxklqDzBZPj5l/trfHRZ9Q/2Kf5S/bt9uoAAAAAAAAgBgiXARHz+6tu1aWXn1S6v2BO+TRN1RmLOTQt+OLk1lv1S7LkLys0jLx6MsHhoIXzegIfXzQ/44S1JqaXx2KmVDEWc3hYOvxwW2vWzJb0Nu+b9Beay4YWqp72hMuC99v22w4GPz1lK5UeVU4qbWPeVClA99qavKSU8+S+1yVJc+YtlVS7uSwry2kSCHkR+Wtfk84/Py1p18Ifx2pJv/A/2U5LlqlsypKyWf9HA91+u3TiiZK0VeGP17W++/7msqyp0Mvqhocekj7yEVu2vUTSklCveeHxw3X3Z5dp7+98Qbvvvrt0wAGTu5IxRHMZADSPsZj1+ZuYbClRTUwcA2gnmsvQjC22Dfd3IwAAAAAAAKCIfyYPRMxJX97oub9kcTldlakbLgvOi265ZEhKBYSADFt9VcJlM4aCw2UL5vXLKK5HtvpYzIsuktasqXKhrDgWc3b9X2pP5ljMnXYODreZppROF8aFZvslu7CN69frueeku/9cCJb1rpdSTpfc7EXbSJLSQZ8Cd8PbWPAYUr9XX5UuuqjBxE6DoyrPPLOxt/c2l0V/LOZZZ0m23cTF2g1L9cwsSc880/Z1SgKaywCgeYxErM/fxJTs5jKOAbTG+29WDGeMbILOFwAAAAAAAADRQHMZEDEvPnmo5/4WW8wo3U5XGWFZNDQ9E/j4/AUzVAyBuBlGXn09ucoXSJox1Bv4eH9/WmZ6xBnP6Q5N+ZrLfvhD152jj5auuML1Jt5RkrVM5ljMJVsGj+VMmXY5XCZTyvVqIjUue/0GveMdrif2l+dj9s+qMpNUknKuEN7oaKiNuvhiaXS0cKFxz/+Q3nZW6WP/cK+UvvXL+g99yvuifI8mUmOhAl9PPSX99reFOzOe1ewPv0NfvDejr+xfPsbu/f5c7aG/lO4HhstGohkue+016bLLbEmGNPCaBj62r677pa2B//iJNDRU8fyL/3uRfnzJYudOrk9jaYUOAk5tXAwHgEZ4C4VorQpS0VyWsCYmmsvQTjSXAQAAAAAAAOgEwmVAhDy88jFJO5TuGwOv6u3vKo9D7KkRLuud87CGpgcHwnbefdvAxw/e9w69vK4ykJaa/pyGZvQFvma/t+8gI10Ia7nDZRu9jWvPPVe4seCv2nrxlTqkd4X+ffxkLZj/jF7pW6/5I4aW7/auqttT1HJz2Zsvkqo0vs2cMz3w8aMPmaP7nnQFi7IDmkiNa+z1UT37rOuJs5xmq/e82F/7omB2sNx+NjIizZlTd7VL+0+SdrtYZ977gr71rrQm7JyOf21AO+urunbWCXrxdVdQrYE2seefd93Z+Qod9dJz+tKyD+nqnZ7RPS/eoxPXbaU36z6dtuXVOv+FD2iHHaTFi12viXhz2erVkmUVPidvuFaHrHtK79z13dIHDgx8/m33ue5k+zWalhMERG22U5fBWEwACIfWqvoqmpikRDUxcQygnfxNf0k7XzA5fvlf/6sbb1mlsXFLn/rk3tpt3126vUoAAAAAAACIOMZiAhFyzdXlhEtq6Fld/5+Paf4W5UasnkxwuGzZzlfpVz9er5kzvUkso3+N/s8n/lOLt15U8Zo93/Jf+p+f/70GevOex7fd6WpdfeFTmjVrmvcFmWF98sSLtfOeO8jIFEI3/kYul3zxbdNj6stJ58/6pv70J+mxx5bq1j3/Tfd//K/qH5pdbVeUtNRctte/Se/+x4rmshlLbtePvneN5i6oDHkdedgKnfKF9yuTdoWmcv2aSEn5Ee/+3+4tn9Mf/1O65qO/qr8uud7A/VRN3v1pSY9p/+ekB3a6QA984gHt99uHNfPuG3TvIwNasMD9ovCBL//79+Uks69fN3zsBv3puD/pwjMflP74R53zyMH605+kO++U+tx5QysT6XBZ0PZ5N8DLE2KkuSy8wthRLokDQDi0VtWX9CamwGPA5K/laE7SzxdMjnN/vEoX/OQE/fi/TtLvfvNQt1cHAAAAAAAAMUBzGRAht64sB49OOeJGHfzBEz0f780EB3meeOgoSdJvr7zF8/h3vvhHffFbxxfula88LH/jFbrnro9Jkvr7vBcfnnr4A5Kke27+q+fxr376ap15nrM+peYyKyPl08qaOaeRq8BzPcOwZNpSOmXrwAMlydTb3ucb51hD0+GyuY9Ih37GuZ3xBrru/NU87bj7fhUv+cDhK3TVtSdL8gX5sgPK9kv5YVfYaLvfa+60R3RQ7w7SAe+svz7ZAWVT4579VIsnHFXYh28Y2Epa8Ebnsa231kJJBx0k/fznxRf1KGtKymaben+Zpmb0zdCB2xTavQ46SCmp8HmTNm92v0GPsqlwy+qGattXjSd3luvXaEY0l4VBcxkANITWqvoqmpgMSVZyvtFwDKCdKpr+DNFchroyqfIxMj6er/FMAAAAAAAAwEG4DIiQR56bV7r9d+/asuLj6VTtf4U+e7Z3zGO1MZnpVPkXyINVypxmz53huT99Wnl8ppF2j4zs10RqkyeI4wn2mHmlbEmpVM11r6b5sZiui3S+5rJZc2YoyGBf+TXpim2U8qOucJ+RV8pS+O3KDmgi/Xpz4bIa+7Cnx/2iJpvLQm5Lxj1BNeJjMRs9Bmkua1IxXNbl1QCAuKC5rL6kNzF5N4VjAK1J+vmCyXHMEXM0Y9oK9Wakd77zTd1eHQAAAAAAAMQA4TIgQtZvmlu6/e7D92n49bPnDXnue8dklq9muq9BTBsIvpg1Z8Esz/0ZQ+WgmuEOa2UHnHDZpirhskZDWD4tjcUs8oXLZs+fFfi0aYOucFnGHS4bkGVK2VHXY42G5nL9muhVc2Mxa+zDtoTLQm5Ls8vqhsD911N9+zzNZdl+jaZFc1kohbGYFGQAQCi0VtVX0cQkJaqJiWMA7VTR9Ccl6nzB5Dj5tMN18mndXgsAAAAAAADESfUZYQA6Lp8vBLiMnKbNmNbw62f52saGZgbXfrkvN0wbCP4yMDTLG1Rzt6CVxmJKhXCZPEEczz+Wb7G5rPlwmWsr096QUE9fj4JMGyivY8r9mqyzEqPjrnCZYTmBpRqjFj2K+ylkc1nYfegJfOV6Qwe+vKNLwwUA4xQua/QY9DaX9TvNZYTL6qO5DAAa4g0W1R/bPBUlvYkp8BggXIYmJf18AQAAAAAAABANXMkAIqQULks3N45v5tyZnvv+MZlBpk8LDtyYKe+Xh1mzB0u3DXfwKtdfEZrytka1duG0+bGYLplwga6haeUyx5T7c5BzVmJszDsW02wkNFcYrdlcc1n1fdiesZjhPkfeZWUiHS5rdPs8zWW5Po1mxFjMMGyaywCgEYzFrK+iiclQopqYOAbQThVNfwk7XwAAAAAAAABEA+EyIEKsvJPeMVLjTb0+nfFOuvWPySxyX74a6A83HXfW7HKTmn8sZtYXmor6WMxqpk8vp6cyvrGYkjQ+7vq8NNrIlh1Q3pSs4c2hnt7sWMxsSlI229j7h9yWtPtQyfcoa4ZbVjc0egx6QoxZmstC8zeXcXEcAGpiJGJ9SW9i4hhAOyX9fMHksvKWctlct1cDAAAAAAAAMUC4DIgQK+/UJxlNNpf5+cdkBjFDXsyaOafcgmb4Rkb6G7maCS5V057msnAhob7eKs1lhXDZ2IQrSNVoaK7wHtmRTaGeHnYftqe5LNy2GIZreREfi9noMegdi9mn0bRoLgujGC4rXijn4jgA1ERrVX0VTUxSopqYOAbQTt5wWfLOF0yOT55wsYxUVqm0qY995D+7vToAAAAAAACIAcJlQIRYhbGYZpPNZX7+MZmtmDOv/F5G2t1c1q9sSrJHhksPef6xfIvNZb29Tb3MK2RzmVvaEy5zkkfjWfdYTKux0FxhtObEaLjmsrD70LN/Ggh8ed6/gQBgJtP4srqh0WPQOxaT5rLwDNd/xcVxAKjD+/2J1qogSW9iorkM7VY6fBJ4vmByGKYky/mL3QTFZQAAAAAAAAgh3Dw8AB1h55yES7VwWV+msd/8usdkGj2bZI8679+TLtc69fSEy5jOXjC7dLs3Pa5S/1axkWtsRKVSK08rliXTlr+GIrSmr7X1lMNuSocLCaVS5YX197hayorNZe6RIWa+6nYNDEgj/jxb4T0mRpoZi1l9H7anuSz856inRxoebmxZ3dDo9lU0l2VEc1kYheay0nVyLo4DQE20VtXn+XZtG7INJaqJiWMA7WYYxeMqeecLJseWi/o1MP8vMs2ctljQ3D8CAwAAAAAAwNRCcxkQJYXmslQqOLDz0x/sKqW8gZdTjv2p5/4Ob/qFJGnOVn/yPP69f7zRuZEad96n4OOfP0TmtFWSpKOPWOF5zW57/rckafqiOz1BtWN3Osy1zk66aWJicsZiStIhhzTxovd9vHw7PS4tuUuStNdb/svztCPe52xzauhZHfvxvys9/vfv/VT5SYVtHMu5wmU12rCuvjpgfYr7KRsu6BbFsZie5VmZ+ITLQhyDnuayLM1loRXDZcVr4lwcB4CaaK2qb0o2lzX5jzAAyXX4JPB8weT4ypkf0fArb9aml/bW+Red0O3VAQAAAAAAQAzQXAZERC6bc4XLgpvLdtt3F91z4/16cdVa7bjLUt19+9/09/9wguc59915uH56wdX66MkHeR4//f9+SNtu+7/aYqu52m3f3UqPT5sxTY/f9Zpu+eOvdPynT/S85s+3Hq0VF/xSRx+7v+fxZW/ct3ynME4jO1au6mo2uFTN1VdL994r7bGH8//ddrX085/8Wge/fy/d+Lv7tOe+y/XEo6u0YNFMjY9lNToyoTftdqleWfWYtlq+l5544CYtPWVb/fKya3XS5/7e895XXXOiLv7hdXrXwbupb6CcMFq4dPuKbZywXVc7awSW/u7vpAcflK6/Xvryl4s7pbCfcuHCWGH3oTdc1qusKSmbrXhezfdvIABYWl6+R9mUnCuk+XxLn9/J0OgxWNFclhbNZWEUzgmaywAgHFqr6vPmrAzne0yCmpi8uR+OAbSuPBYzeecLAAAAAAAAgGggXAZExOYNmyXNlFS9uUyS9ty/3Dq2/Ru3q/h430CfPvWPHwh87ZEfe1fg49vtvI2223mbisd7+nr0yS8eWfF4JuO6YzlfRtyNXJ6LZm1oLuvtld76Vue2839TJ592uCTpmE8skiS9YddlFa9bsM0ukqS93n28JOnjZ7y/4jlmqvxebkHbOJ5yXe00rJqBpV12cQJm/veYyAUHB/3C7sNmm8s8799Mc1lxWSosz5PO6r6K/ZdX+OayHM1loRWayyyaywAgFJrL6vM2lxnO95gEhWU4BtBu5UAmYzEBAAAAAAAATA7mbwARsWHdxtLtdDpcAKlb0u5YaqGRa2Ki3PLkbY2yZNqK3bifoG0cN13hJCNfd7s871FsPwsZLgu7D9szFjP856gUuis0l9kKt7xOa3T7MhnXh3N9Gs2I5rJQCs1lhMsAIBRvsCiePyNNNs/usE3ne0yCxvwFHgN8/0QLys1ljMUEAAAAAAAAMDm4kgFExKb1w6XbmVT9sYbdFNhclp8oXcho91jMbghsLnOHy0I0sgXuJyvc5zbsPuztdb+oyXBZk2MxJTmjMSMfLqt/DBqGq70sS3NZaP7mMgBATYzFrM+7OxLeXMYxgDYoBTIZi4mQ/njtHdp+l6u0zY7X6JMnXNzt1QEAAAAAAEAMMBYTiIgNG1zhsnS0w2WBjVwpOWGcwcGmg0tRErSN/uayeoGlwIa3ZprLJmEsZrMBQH+4bCIl9UQ9XBbyGOzvl0ZG5DSXpUVzWRiMxQSAhjASsT7vWEzTCcskqImJYwDtRnMZGvXss2v05MNHSZLuHPyfLq8NAAAAAAAA4oDmMiAiNm8styRFPVwW1MiVLYbL5LuekaDmsqx7TlOTzWXZkM1lYfehP1yWTUnK1l+G5/2baS6zU5JlKmuGW16nNXMM9vcXbuQKzWUTE76UGioU5mGWrpNzcRwAaqK1qj7vWEyay4B6yueM4fxolqDzBZNjYKD8l8hcnl8LAgAAAAAAoD6ay4CI2LSp3JLUk851cU3qC2zkcoXLvK1YlkxbviuF0Re0jd7mMssJLGWqb1dgw1s+XMuXdx/aVfdhe5rLwn+OvMvLaCI1HoOxmOG2zz0Wc7QYDBwbkwYHJ2ENE4LmMgBoiDf8TGtVkIrmsoQ1MdFchnajuQyNOvjIfXTttD+pv69XW227Z7dXBwAAAAAAADFAuAyIiE0bXeGyTIzCZYVGromUCjMFmxtJGDVB2+gfi2k20Vw2EbK5rLQPjbxkKNxYzFzvpI/F9GxTvice4bIGxmJKknJ9yptSzpTShMvqKDSXES4DgFBorarPmwU3nHbMBDUxcQyg3UrnjJ288wWTY+bcmTr87w/q9moAAAAAAAAgRuJVJQQk2PBIOaDTk472KD5vaKpec1kSxmIWttFwbUOIwFJgw1ujzWWGcyPsWMymwmXNjMVscHmd1swxWGouyw1ItjSaVumYRhU0lwFAQ2itqs/fXJa0JqbAYyBmDb+IFprLAAAAAAAAAEw2fosNRMTm4XKjVV9PtMNlVZvLCkEcz/WMBDWXTbgv/IUILAW+hxWula60D81CuCxMc1kDYS/vWK7wAcC4hMuaOQZLzWWSlO/RWFrOWEyUVBRhEC4DgIbQWlWfJ2dlG047ZoKamLy5H44BtK58ziTvfAEAAAAAAAAQDYzFBCJiZKQcOurNRPuCgHc0onMna6pKc5lVGB8Zryxr4Da6m8sMq25gKaj9LNvwWEznCqRpK7DVorfX/aIe5/OQrb+MwM9RiNYMT7jMyiibCre8Tmtm+0rNZZKU7ddoZoLmMp+8P/damIcZ7a9YABAdNJfVNyWbyzgG0AKay9CozRs268Lv/0Zj43ltuWRIx3360G6vEgAAAAAAACKOcBkQEcMj5dRGX2+0LwhUbS4bGZFUZSRhTwKay3xjMesFlgLfww7XXMZYzNY0MxbT01yW69NYegPNZT4V1yr9zWVczASAmmguq8/7o5XhBJgT1MTEMYB2K50zdvLOF0yOl55/RWd880OSpHlb36DjPt3lFQIAAAAAAEDkxatKCEiw0fFyKKOvp8YTIyCokcs9FrPZ4FKUBG6j6W4uqx9YCmo/m2i0uczMy7DltFpMVrisgbGY3m2KSbismbGYuX6NpkVzmU9lcxnhMgBohLe1Knxz6FQytZrLCscA4TK0gOYyNKq3v/yXSMuK19/TAQAAAAAA0B00lwERMTJavtI00BvtC0xVm8sCx2KGDy5FSZjmsnqBpXY1l6WK14doLgutmWPQMxYz16extGgu86m8VlkYi1m6qElTBgDUwkjE+jxZO9twvsck6PsLxwDarXzOJO98weSYOXuG9t//EknSDlsTRgQAAAAAAEB9hMuAiBgbL18E6Ovt4oqEENjI5RqL6QmgJKG5rLCNWffVTsNqrLms2H4WMlxW2ofFRgspsNnEG/bqDR328nyOGmhOiUu4rJnt8zSXZfs1mhHNZT40lwFAaxiJWN+Uai7jGEAbVDSXSc6BxnGFKoZmD+mWW07o9moAAAAAAAAgRpjBAkTE6Hj59mB/tINYQY1c2arNZfEc+RS4jYZ3LGa97Qp8D9ufzgnmHouZKl6EDNFclk1JytYfvdlsu5x3eRllzXDL67RmmtloLquv4tp+obKsdJ2cpgwAqMkbfqa1Koj3RyvD+R6ToO8vNJeh3UrnjG3wMxkAAAAAAACASRGvtAeQYGMT5YtK/f3RPjUDG7kSNhYzaBuzDY7FDGx4YyxmRzRzDHqay3L9Gk2L5jIfmssAoDW0VtVHcxnQmMDmsgSdMwAAAAAAAAC6L9oJFmAKGZson46DA9GeWBvUyFU1XBbTsZie1owqzWX1AkuB+0ktNJfVHYvZZLisgQCgNzAXk3AZzWVtU9lcRrgMABpBa1V9nh93bMMpyUxQCxPHANqtfM4YxVLZRJ0zAAAAAAAAALqPcBkQEeNZd7gsU+OZ3RfUyOUeWeQd+RTP5jLDcG1nYRst95dMw2qsuazQfpYLORaztA/r7D/DcH3ISitnSsrVb0fzfI4aCAB6tyn88jqtmWPQHwaM6rZ1U0VzmQpjMbmQCQCh0FpVX9Kby7ybUjgGYjY+HtFCcxkatealNTIHX5E5+KoWL/tNt1cHAAAAAAAAMRDteiRgChnPloMv0wajHS4LauRyt0p4W7EsmZZiedEsnZayWZW20ZJvLGad7QraT9lGm8sMS2adi47pdOH5+YyypgorHfL9Qy7DvawSK6NsKtzyOq2Z7fNuW0p50/9GqNtcZtvOH4ISABCI1qr6vN+uDc8/YEgCjgG0W+mcsQvni5SocwaTwx5ZIEkaHZ/e5TUBAAAAAABAHBAuAyJiIlcOLk2f3lfjmd0X1MjlvvBXMZIwr9g1l0nOdo6OqryNtncsptlIc1mh/SzX7FjMOssZH1dDbVttGYsZ4XavZsZiej4c4W3rpoqsnT9cJhEuA4AaaC6rL+nNZRwDaDeay9CodCat9IynJUkzBtd1eW0AAAAAAAAQB4TLgIiYyJZTO4PTeru4JvU11lwWz7GYkms7i81ltq+5rE5gyR9WkqSswl3oKTeX1d9/5fUsNIk1Gi5rYCym53PfQFNapzVzDDIWs77K5rLCWEzPYzRlAEA13taq8M2hU4lnd9iG52fMJAg8BgiXoQXlc8ZgVDlCmTl3prLrZxbubdvNVQEAAAAAAEBMcCUDiIhsrpxsGRoa6OKa1BfUyOVulfAEUBoILkVNaTuL2+j+kmlYdQNLhuENfklSzg4XLivtwxD7zx2Cy4UMe3k+Rw0EAOMSwGrmGPRsm51S3hBjMX1CNZfRlAEAVTESsb4p1VzGMYA2oLkMAAAAAAAAwGQjXAZExESunNiaNtTfxTWpL7C5TKrSXBbfVo6K5jJ5x2KmQmyX/z1yDTeX1d9/pRCcO+xVp62g2c9RLMdihty+uATnuqkya1doLvOPxQQABGIkYn3eb9eG52fMJOAYQLuVw2U0lwEAAAAAAACYHPFLewAJlcuXUzszZk3r4prU5x336Kx3EsdilkNbheYy31hMM0Qblr/9LGtM4ljM4phKzxvUeX+p+bGYxTGcCRmL6R9jSrisUuVYTJrLAKARtFbVR3MZ0JhSILPwc5ktJeqcAQAAAAAAANB96fpPAdAJuVxP6fb0GYNdXJP6iuMecznVby6L8VhMf+uYbfuay0IEllpuLmtmLKbkfHLS1b/ENxsAjEu7VzPHYFy2rZsYiwkArfF+iaS1Kog3XGZ4/gFDEtBchnYrHz7lRlkjQecM2m9ibEJvfcflkm1om0XjuuKak7u9SgAAAAAAAIg4wmVAROTyvaXbQ7OHurgm4WQyhdxNvrK5zHPhNAnNZYVttNxlj4YVKrDkbz8LGy4r7cMQ+69iLKZUNxTl+Rw1EACMy1jMZo5Bb7gspbypug1wU01lc1nhIqbnMS5mAkA1tFbVR3MZ0Bh/c5llSGaCzhm0n2VZuvfPH5MkPb7ozi6vDQAAAAAAAOKAsZhAROTzheYyI6e+gb7urkwIFa1eUpWxmJYzPtKM35ebms1lZnPNZVlTdS+QeoNR9fdfeRkZWWahRarOqMpmP0eeAFZxDGfkx2KG2z7GYtZHcxkAtIbWqnDcTUzunzGTgPY6tFvp8HGH/hN0zqD9zBj+3RwAAAAAAADdRXMZEBFWIXykVFZxODX9jVzu5rKkjMX0b6Ml71jMMIElf/tZKbDU01P1NY3uP3dzWXEZPXVCUc2OxYxLcxljMSdHZXMZ4TIAaAStVeGYZuF7+VRpLiPogRYENZcl6ZxB+6Uzaf37d38pSZo/f1qX1wYAAAAAAABxEP0ECzBF2IWLATLiMYYvfHNZfMdiuhvBZEuWv7mskcBSsbkspcbCZSH2X2kZhQBbttFwWQMBQG8AK6NsRjFoLmtiLKadUt7wvxEqd0ehIcMzwoymDACohuaycDKZwvecBIa9OQbQbkH/0KQ3QecM2s9MmfrkF4/s9moAAAAAAAAgRvgn0kBUlMJl8Qhm+Bu53P9CPnHNZZJkpWS5v2QaVqjAUtDFnnoXSBvdf6VQlJ2W7MIyGhqLGT4AGJd2r2aOQcZi1kdzGQC0xttaFd/R4ZOt/HNmjyZSkiYmurk6bRV4DBAuQwtK54udliwjcecMAAAAAAAAgO7jSgYQEYWhOJIRj2BGRXOZayymJ1sS4wun/iCVbXvHYjbWXOYai1kn+NXo/vOH4MKEopr9HMVlLGYz2xeX4Fw3VTSXFSrLvCUs8QjIAkA3MBYznFLBaz7jtL5GsCW1WRwDaDdPIbKVvHMGAAAAAAAAQPcxFhOIikIDkKF4hMvKjVzOjSSOxfQGqTKVYzEt1Q0s+ZvLso02lzUyFrOwntlUvu3taIHLymec7YngxauWx2KGDOlNNTSXAUBrvMEiMRKxisDmMttOxL5iLCbazfN3lnyPJlITNJehrst/8ntZtq2ZswZ08Aff3u3VAQAAAAAAQMQRLgMiwo7ZWMxazWVJGYtZ0Vwmb3OZ2UhzWT58c1nTYzEL68lYzPaMxcyb/jdCZXMZ4TIAaIS/3JHWqmDl5rJCuExyft7wpGjiieYytJunuSyBo2QxOT7yD38nSRpceLc2f7DLKwMAAAAAAIDIi9+cOiDpYjIWs9woMUWay/K+5jLDChVYKr+HKVlGqDBWo/vPv54NL6OBAGBcxmK23lwW3W3rpsrcWGEspvuaOGMxAaCq8tdR5watVcFKYRmr0JIqRbIptRk0l6HdvOGy6DYLI6Jsvv4AAAAAAACgPprLgIgoN5fFI1xWq7nME0AxLKfhq874yCiqbC5zbYPZTGApo2xqou7Fnkb3n389w1xQavZzVLk9kkajd/Gqme3zbJvNWMwgNJcBQGtKwaJCUy2tVcHcYzFzKecfMRgTE9LAQDdXqy1oLkO7VY7FFM1lqGun3S6TJG01f0TSXt1dGQAAAAAAAEQe4TIgKorhMsWj9ad0EcPqkewazWUxHovpbenKyLa9YzFTIQJLlU1fE21vFWumTazZdrlYNpeFPAb9Ib284X8jVOTGCJcBQEPKwaJCuIzWqkDlsZjODx7ZlNSTkCYm77dJmsvQOk9zWfEffyTkfMHkefi+j3R7FQAAAAAAABAj8asSAhKq0FsgI27NZZJkm57msqSMxfSHjTxjMc2804bVSGAp5JiaRvdfYJtYQwE2O3QAsJnt6YZmjkHPhyMcnOumyuYyxmICQCNoLgvH3VwmKVFNTIHNZTFs+EV00FwGAAAAAAAAYLLxW2wgKmI2FtN7ESOT/OayfEBzWYjAUsutYiH2nz8El2skwGY46xI2ABi4rAgGsNrRXBbVbeumamMxJVfnIs1lAFAVzWXhlJuYTMkynbBMBMPszfBmsGkuQ+s8zWUR/scfAAAAAAAAAOKLcBkQEXbMxmJWtHoZKoVKPNkSw3IavmLYyFCxje4vmYbVRGApEyr41ej+aybAVlpGIcwY9nMUl7GYzRyD3s9VKrLb1k2VubHyxfBSexnNZQBQVbm5rLHvv1ONNyzT44RlEtLE5G0uK/yMQrgMLfCfLzSXAQAAAAAAAGg3rmQAUVEIl8VlLKY3ZJRJ5FjMim30jcVsprms4ZGVjY7FLLYVhF2G6dxoaixmcQRnBJsR2jEWM2/63wi1msus4nVxmssAoCrGYoaT5DF/gWMxOQbQgiSfL5g8A/PvV9+cR7TF8l93e1UAAAAAAAAQA+n6TwHQGYWLSjEJl/lbvZI4FrNiG31jMRtuw8pnlEsr/MhKqQNjMQvhspABwLg0lzEWc3JU5MYIlwFAQxiLGU7FmL+IhtmbwVhMtJvnfInwP/5AtIyu3VGyerQuRRARAAAAAAAA9REuAyLCjnNzWX4KNJflM7Lkay4LEViqaC5rpFVMCrX//A1rDbWjFZrLzGaay4otaRG8eNXMMejZNpuxmEEqm8tcYzFLNxiLCQDV0FwWTpKbmGguQ7sl+XzB5DF6NsvO9yidHuv2qgAAAAAAACAGCJcBUVFqAIpHMKNWc5mnuChRzWWuljLDajywZGVCtYo1uv+aadwqLcPIy7QKF7cbHosZ3XYv7z60Qx2DFWMxDTEW04fmMgBoDc1l4fibmCYS1MREcxnaraLpL6L/+APRYo3OLtzap6vrAQAAAAAAgHioPc8NQAc5F5Vi2Vxm1Wous0KNj4yiim30jcVsprksTBir0f0XuIzQYzEL76/ay6i5rAiGy/xjPxseYRrhbeumyqydq7mseJPmMgCoqhzuprmsFn8TU9ZUYpqYaC5Du9FcBgAAAAAAAGCyxS/tASRVsQHIiEcwo1ZzWVLGYlY2l3nHYjYcWMoXRlaGDX5JofZf4KjKBsZipoqHXMPNZeG2pxv84bLGW+YYixmE5jIAaA3NZeF4m5iSFZahuQztluTzBQAAAAAAAEA0EC4DIqI4ctGIyVhM77+Qr9FcFuOxmBXbKF9zWYjAkr/pK29KdiPhsmbHYoZuLitsh2ovo+ayIhjAcofnJDU3FtMUYzF9KnYH4TIAaEgpWERzWU3+MX9JGovp+TZJcxnawD9GNqr/+AMAAAAAAABAfKXrPwVAZxQuKsVkLObUbC5z5XENK1Rgyd/0JUn57HjNL76B+y9TfTn+8Z0Njd50N5eFGItpGM4m5/OKR7is6eay6G5bN1U2l7nGYpZuxCMgCwDd4G8uk0SwKEDFWMwENTF5x2IWcoYxHB+P6GAsJppxxPtWaGzC0MI50iU/P6nbqwMAAAAAAICII1wGREWxuSwm4TJ/oMkyVEqeeBsZrFDjI6OoYht9YzGbaS6TpGydcFmj+88fisqmVDcUVVpG8f0NI/TF7XS6EN4qjuCMYDOCe/skNT7C1GYsZhCaywCgNe7mstIkdMJlFfxNTElqLvNnsGkuQ6v8TX9R/fkc0XLtHz4sZaepd87DuqTbKwMAAAAAAIDIazntcfbZZ+vVV19tx7oAU1sppBGP1p+KVq8EjsWsbC7zjcVsNLCUd5Jmuex4zdd0ZSxmA5+fUmAuwu1e7mY2wy5cuG1wLGZUt62bKpvLCJcBQCPK4bJCuFuKZQB/slU0l5lKTBNT+RhwflgphfyBJtFchqYY8fi9AwAAAAAAAKKh5SsZX/rSl7TlllvqyCOP1K9+9StZXFQGmlMYLxfL5rJ8JpFjMSu20ddcZoYIzQU2l+UaCJeF2H/+9cw2Mxazgc9PKcxmZZyWtAg2I1SE56SGx2LmDQVUdU1tlbvDNRazeJOxmABQlXssJs1l1fmbmJLZXObcMAiXoUX+pr+o/nyOaPn6p6/Vlz/7Xzrzsw93e1UAAAAAAAAQA20Zi5nNZnXdddfpuuuu0/z583XcccfpxBNP1A477NCOtwemhtJYzHgEM6Zkc5k7j2tYoUJz3vfoYHNZ2HBZE+G/crgsuu1eFeE5qe42eopjIrxt3URzGQC0xjMWs/ggwaIK3nBZspqY3MeAxFhMtC7J5wsmz7fPPbbbqwAAAAAAAIAYabm57MEHH9TnP/95zZ07V7Zt65VXXtHZZ5+tnXfeWW9961u1YsUKbd68uR3rCiRbrMdiepvLPNmS4tinGI58qtjGoLGYTTSX5XK1L/Y0uv/8ywgzFrO0jCY+P4FjMSPWVlWxfVLdbTQMd3AuRbgsQEVzme1qLivdiNaxAABRQnNZOBVjMRMUlqG5DO3GWEwAAAAAAAAAk63ltMfOO++sc889V6tXr9bVV1+tww47TKlUSrZt689//rNOPfVULVq0SCeeeKJuueWWdqwzkEyl5rJ4tP74A01Vm8uSMhbTSleMxWy4uSzvvGG2keayEPvPH4LLptSZsZjFEZwVK919zYzF9DzFSitvKnLb1W2V4TKaywCgEeXwM81ltfjH/CVyLCbNZWgT/xjZbIh/aAIAAAAAAAAAjWhblVA6ndYRRxyha6+9VqtWrdJ3v/td7bjjjrJtW8PDw7r00kt14IEHavny5TrrrLP04osvtmvRQEI4F5XiEi7zh6bczWWJHIuZr2wuC9P45R9ZKUm5fO0mgdiNxVT95XVa4FjMEO1scRj52U2MxQSA1tBcFk5Fc5mpxDQx0VyGdqO5DM34650P6Z6b/6oH/vxIt1cFAAAAAAAAMTApc+rmz5+vM844Qw899JDuuusunXrqqRoaGpJt23ryySf19a9/XUuXLtUhhxyiq666Sln+VS0gFYNLMQmXTfnmMsNqYiymc6feWMxG95/3glIm1FjMVprLKsZiSpELYTUbniuFy2zGYgapLHIr/xhRmpDJWEwAqMrdWkVzWXX+JqYkNZe52+skmsvQOn/TXzZB5wsmz5sPXKK937Gb9jo4fn9PBwAAAAAAQOdNSrjMbe+999aPfvQjvfTSS7r00ku1cOFC2batfD6v3//+9/rQhz6kJUuW6Mtf/rJefvnlyV4dIJKsfDlQFpdLS/62LHdzmae4yLBCNXxFUcU2ulqamhqLWWguy9YJlzW6//zLyIYIRZUvbDb++Sm3e2VkmYXGqohdwGp2+9xjMQmXVQosJSsEJWguA4D6aC4LxxsuS1YTE81laLckny+YfHZsfgMBAAAAAACAbupI2uO5557Td77zHX3jG9/QK6+8IqPwy3PbtmXbttasWaPvfe972m677XTeeed1YpWASLFcYYy4jMX0N3JZhkqhkqSMxfRvY8vNZfkmmsuaHYtZI+xl2+7wVePNcu7mMkmRDGE128zmHouZNxRU1TWlBe6OQuiScBkA1EdzWTgVYzETFJZxHwNS4R+WxPAfYSA6GIuJZmyz3Y1asux67bjtym6vCgAAAAAAAGIgXf8pzRkbG9OVV16piy++WDfffHMpSCZJy5cv18knn6yPfexjeuihh7RixQpdddVVGh0d1RlnnKF58+bp2GOPnaxVAyInl82peDrGJVxW0eplSrKSNRazsp2tOLo0LxkK1YgV2FyWDzmysrCshsZiFkfh1Ah7ebI/TYT/SttUCMtlTaknYs1lLY/FtApjMfN55yowF/4lVWsuKwTGS/cZiwkA1ZSDRYVmTYlgUQD/mL8kjcV0HwNS4edJfs5AC/xjZLN1/qEJIElPPfyBbq8CAAAAAAAAYqTt4bI777xTF198sX7xi19o06ZNkpyGsv7+fn3wgx/UKaecov3337/0/IULF+pd73qXnnrqKX3wgx/U/fffr/POO49wGaYUJ1xWFI9gxsCA6874kOx+la6WJaW5rGIbrWK4zLkY2OxYzFwj4bJmmssyqhkuazX8V1qenZbsZDWXVYzFLL5ZetKy2LESeJ3SNiXlaS4DgBAYixlORXOZqcQ0MTEWE+1GcxkAAAAAAACAydaWq+UvvfSSLr30Ul1yySV6/PHHJanUUrb77rvrlFNO0THHHKOhoaGq77HddtvpO9/5jt773veW3gOYKqy8eyxmPMJlS5e67ry+jewBla6WebIlxWaOGLZy+LexNEnYcJJLYUJL3s12LhzmVHvUYqP7zz96s95YzFY/P96mtJRyZj5y4bLy2M/Gts8zFpNwWYWRkaBHC81lxeviNJcBQFWBYzFRwd/ENNGjxDQxBY7FJFyGFvib/rIJavoDAAAAAAAAEA0tXy0/5JBDdMMNN8iyrFKgbMaMGfroRz+qU045Rbvvvnvo99p2220lSSPBV6+BxHI3l8VlLOZ227nuvL6d7CUKbi6L8VjMJUuknh5bExOGs42l8VWFcFmI7fJcK7SdtFLWrh0ua3T/+ZvLsnWaxFptlvMuL6NsKh+5C1itj8V0NZflclJvb5vXMJ5GRwMeLBzXNJcBQH0VzWWEigJ5w2XJamLyN5dJ4jhAS5J8vgAAAAAAAACIhpbDZb/73e9Kt/fff3+dcsopOvroo9XX19fwew0MDOjtb3+7DH65jinGEy6LyVjMRYuknl5LE+Om9Pq2TmtR4FhMO7ZjMVMpaeutpccfl7ONA887Hyg0l4VpxPKGywrNZY2Ey5oZi9lIuKyVsZghl9cNLY/FtFPKm85lXyNi29ZNgdlvwmUAEFpFcxl/7wlUMRYzQWEZzzHAaFS0AWMx0YzFy36j0fHpmjFtrZ599Ihurw4AAAAAAAAiruVw2fz583X88cfrlFNO0fbbb9/Sey1evFg33XRTq6sExI5llQNlcRmLaZrSllvl9NQTPdLr28qyVRkuM5xQTlybyyRp220L4bLsoCZGFzkPGpYMuzDGqImxmNk6YzEbDX95x1TWH4XTanOZfwxnts4Yzk6zbfdYzOabyyQpb0ppwmUlwc1lhbGYpfvx+BoGAN1Qzt/SXFaLf8zfRILG/FW010kcB2iJf4xs1H42RzS9/NKeskfma/OMp7u9KgAAAAAAAIiBlsNlq1atUjrd8tsAU1ouF7+xmJK0dJtCuCzXr9HxRQHhMmdbwjR8RdV225Uv9uUnZjo3zEJgSWpqLGZDzWWGVXf/BTaJZcM2l9V//1DLi1AAy1Oc1eD2VYTLDCmdr/35mkpoLgOA1tBcFk5Fc5mpxDQxlQPwhWNA4jhAS2guQ3P4ByEAAAAAAAAIr+VUGMEyoHX5nCu8EpPmMknaauvyem8a2VbKOFfLShfNTOfjcR2LKTnNZRWMQtuX1GC4rDAWU7XDN55sziSMxfSGrxofi+ltSot4uKzBZrZyuMx5fs6UeiO0bd0W3FxGuAwAwqporSJUFMjfxDSRkjSajCYmmsvQbo22GAOStObpjKR1SmfmdntVAAAAAAAAEAMkw4AIyGXLIS0zRuGypduUQzebNm8nzXxJkru5rBAui/FYzO22C3iwgeaywLGYDTWXNTgWM8QonFbHYnrDbNG7gNXo/nMrPc1OS7YiF5zrtsDmssJxbRevizMWEwCqorksHG+4LFlNTBXHgBTbhl9EQyrl/MnnlbjzBZNn9oLZ3V4FAAAAAAAAxAi/xQYiwHI1/RgxCpct2aKc4hkbCxiLWWguM2PcXLbllgEPGnlnm6S6FwMDx2KqgXDZJDSXtRK+amZ5ndZKeM67bSnlTf8bTm2MxQSA1pSDRY2PpZ5KktzEVHEMSIQM0bLSORPiH5oAAAAAAAAAQKO4mgFEQM41FtMw4hPMmDatfDubG6wMlxl5mVah1yim4bLp0wMeNKzyWMyGwmWFsZh27c+xN/xV/+Kz/wJsrpHmsiYubkd9LGYr2+cJl9mpyG1btwWPxSw0l5XuxycgCwCdxljMcKZEcxljMdFGpXMmYecLAAAAAAAAgGhgLCYQAXnXWExD8QlmDEwrr2suP610taxUXORuZIhpM8fgYMCDZvi2L89mFxqesnWayzzFTyGaxfxNYtmUagaiPO/f6ljMCLYjNLr/3DxPs9LKmROEy1xoLgOA1jAWMxzvyO8e52eNhIRlAsdichygReXmMsJlCOe0Uy/Rhk2WhqaZ+sGPT+j26gAAAAAAACDiCJcBEWC5mn7iNBZzmit4lQtqLisGl6TYNpe529lKjPCBLO+1wkJzmRpoLovgWMxYNZe1NBYzrbzhf8OpLbi5jHAZAIRFc1k43uayjBOWiVCQvRU0l2EylM4Z9xhZ2+bYQlU//Pm7ZG3eQqnpL+gHP+722gAAAAAAACDq4lklBCRMLhvPsZiDnnDZtMCxmKni5sQ0XFatuSzsqEVvc5lzcSfbSLgsRPjL2+5Rv0mslfCV5A9guS5gRUQr4TnvtjEW0y+wuawQmrSL1y4ZiwkAVdFcFs6UGItJcxnaqGIspsTPsAjFtvn6AwAAAAAAgPpoLgMiIJ93j8WMD3erVz6fzOayVEpSelTK9ZcfNKzQgSXPtcJCw1POrt2EFfXmskaX12mthOcqx2IqUtvWTbZNcxkAtKr8JZLmslq8LanRC7K3guYyTIbyWMzCPzSRnHPGczIBZZ84+gZtGrY1fdCQdGK3VwcAAAAAAAARR7gMiAAr7x6LGZ9gxuC08nq7m8tKF04Ny2n4kuJ90axnsy9cNrljMT3ZnOI+rNGS1mjYq9H394v6WMxWti/qwbluGh+vUkpGuAwAQqO5LBxvK2uymsvKPyfTXIb2cTeX5VLOz2XmxIQ0MNDV9UJ0/dtPCZQBAAAAAAAgPMZiAhGQczeXGfEZKTc4UP4SkssNlq6WVYzFjGlrWZHR45sFaIbfro6PxQzR7tHWsZghxnB2WvvGYqaVN/1vOHUFtpZJpeO6XMQSn69hANBpFa1VhIoCmabrW3e+x/lZIyHhMprLMBnKzWVOyixJ5wwAAAAAAACA7iNcBkRAPlsOG8UpXJYyDSmzWZK3uaxiLGbsw2XDvgfyoduwAsdiirGYk6ltYzHtVOS2rZtGRqp8gOYyAAiN5rLwyk1MGae5LEJB9lZUHANAG5TOF8tJmSVplCwAAAAAAACA7iNcBkSAZcczXGYYhlQIXuVzg5XhsoQ0l8nfXGZYkzoWs9HwVyrlWk4+01i4rIkAYNTHYrazuSxq29ZNVZvLCse1XTwGaS4DgKrKwaLGx1JPNe4xf0kai8kxgMlQOl/slGSZiTpnMDnGRsa0ecNmbd6wudurAgAAAAAAgBjgN9lABORz5bCRGadwmQypp7K5rFRclJCLZkavr7lsksdieoqfQu7DUijKStcdU9nM+wcuSwo1hrPTWtm+irGYhhiLWUBzGQC0jrGY4Q0NFW6MzNW6fik/OiIND9d8TRxwDGAyTJ/uujMyV2sGJK1Z063VQQxMX/SSps+cphlLCJcBAAAAAACgvngnPoCEyOXK4ZU4XV8yDEPK1GguS8xYTH9zWfjtCh6L2UBzWch96A6XTfZYzFg1lzV4DHrDZYzFdKvaXEa4DABCYyxmeNttV7gxOlcT2ZlaNSTp6ae7uUptwTGAyVA6XyRp3TI9OVvSk092a3UQI7bN1yAAAAAAAADUR7gMiADLNUbOMOITzHA3l1n5AeUL6ZKkjcU0eppvLgsaizlR53PcTPirr69wIztYdwxOq2MxPQGsCI6qaiU853laBINz3VS9uawwFrN0Pz7tiwDQabRWhbf99q4767ZzwjJPPNGt1WkbjgFMBs/5snZ7PZGQ8wWTZ9HChzS0+HYtWHhft1cFAAAAAAAAMZCu/xQAky22YzENQ3IFr4bzfRrSVGguCz9qMWgs5rhqj1lsJvy1YIH0+uuShudrPC1pfDzc+zcRACwF2SQp16vxVO3ldVr7mssIl7nRXAYAraO1KjxvuGx7PTHnXr0zAWEZjgFMhsrzRYTLUNPzfzus26sAAAAAAACAGKG5DIgAd7jM1f8Tee7mMkkazvVKcmVLGghhRZnR62sua3Es5rhRO1zmyeaE3IcLFxZuTEzXWH7QSVhVCUU18/5uvb2uO/neumG2Tmtl+zzhMjulvClfWm3qqtpcVmjkK03UobkMAKqitSq8ZctcdxI05o9jAJMhqecLAAAAAAAAgGiId+IDSIicK7wSp+Yy0zA94bLN+X7ZtjtcloyxmGavL1XT4ljMsTrhsmaaxUrhMkljYwucG1UCX62OxfQ2l/VpLC1pbCz06ycbYzEnB81lANC68s9ItFbVk9QxfxwDmAxbby2lUoW/RybofAEAAAAAAAAQDYTLgAhwF/0YMQqXGYYhZVxjMXO93lxJQsZimhVjMcNvV+BYzEbCZSH3oTtcZo8sdEJRYcJlTQQAPc1ljMWcMjzNZYZrnxAuA4Am0FpVz7bbuu6sXa7H50h6/PHYN2TSXIbJ0NMjLV1aOJbWba9nZhgaX79WWreuuysGAAAAAAAAIBHS9Z8CYLK5x2KaZnwumPnHYm7O97ccXIoiw99c1sCoxeCxmLXDN602l2nzQo2npPQkNZdVjMXMKLrhsuL+yzQTLkspb/jfcOryNJf1DEvjM5zbVkp6+IO68LW36oKhaXruoq2VveQm5YY3OUEz247TtN/oKwSQq0cRAj7u/UJU++3rBpzLHzcCbgV+3LMy9ZZfZ/GByw9+oPKtGt12w3evxuurrLf381D5eu8+rPFxQ4Ef9zw38HPnXkKdjzfwuQ9699rb578T9FD15ddeN0Nz5z2lHXf+vVIpbxh4lt2n43M7a8ngIunv/15avrzG+0w+TyaK1qq6BgakpUul556TtHofPXbdbZrdv0rTtvullEl7jgvDsGXIdv5vWDIMW2bh/+XHnIbg8sdtmaalBYse0vbL/yTTlEzDkCFDhmFqsT1NJ2Z30aw5S6Tjj5cWL27LdpWOA44BtNkOO0hPPy1pYrrs/3hIc7d5ULP2+JPU01NxvkjF7wuu4694TJb4j1Hv1+LKr83un0EM7/2A72OVh77vZxj3qtVdvlH5fc7/et/6+rff834V+6Lye5ZRfCTEtvnXrWLbfT/fzV/wuN6w0w0yzXx5OZLm2v06PreLFk5fJB1zjC+F25jd9/pvrdkwpP7ecT3+4Aebfh8AAAAAAABMDYTLgAjI5crhlThdXjIMwwmZFAzn+3zBnvAhrCgze3zzADs6FjPcPlywwHVn80KNpaXBKqMqW/0cVYzF7FeEx2I2tn2MxazO01zWs7kcLvvZ76XhBbqu+LH7OrxiACDpd3cfJR1/kDS4xvP4P5u36Yr/lg799relCy+UTj65S2voL9yitSqM00+XPve5wp0X9tPrkl6fhOX8bus/ScccKmW8P8+cad2sP1wg7fHP/yxddZX03ve2vCyayzBZPv956fe/t5XPG9KanbR5zU7aXPdViKrf3HOkdNy7pP71nse/bdyqay+W3vntb0sXX+yEzJrw8JN7K7t+exl9tNsBAAAAAACgvngnPoCEcF9sjHtz2auvup7Qt0EzxiXNmNHxdWsns3fY+8CM57TVBvkqvKq8NnAsZu3msmb2YUVzWVpV28Q879+7QTPGJA0N1Xx/t6iPxfRu38aGjkHGYlbnaS5zjcPV8IKK5wJAx736RumWr1c8PJqRPn6YNK68dNpp0po1AS/uDG9zWTIC+JPts5+Vzj1X6hucmNwFPXuQtPLjFQ+vG5A+dahkj4xIH/+4NNH6epSbyzgG0F7vfrd0xRWGZs+Pzj/6QAte2kO6/UsVDw/3SKceJuXyWekzn5E2bGju/es2lgIAAAAAAABlNJcBEZDPl8NG9cdSRYdhGJ6QybDVp0cecT1h7qPa8TVJO+7Y8XVrJ8PfXLblndr3BUkD/fVf6x7HYpmyJY2ZtZvLSvuwb53Sfa9q29flzLmpwR8uG0urapuY53M071Ht+Kwa+hxVNJfVWFY3VGzfk5L+Ltz2ES6rbu1a152egB6MvX8g7fFjvefZMX3ubkt9b9hZxmmnSb19XDhvE3fbjft/vkdluxMs7pueh+2KB4NfVuW9Al4U+Ho78J1c6+h+keH6eJ3X+9bftv3DqnzbV2VfBa275/WFD/rfv9a6lT8cvG8qPm77xpL539O2g1/vOg6Cf3JwfTxg/d37KHD93SsfuHzX+gWuv+tFAZnq4OOh8jl26T/Bz3jltek689x3Km+Z6rnv0/rFp4c0Z9aoJqysTn/i3/TXzU9o9ZB0yW7Sx+8dls47TzrzzCrvN7kYi9mcL3xBOu20Hr30sqWX1r+middelTUxIVmWbFuybMm2LVl5ybIs5fO2LMtWPl/4Y1nK55zHLNu5nc9LlmXrkb/N1Nk/3EuSNPeef9EVn9pVmcyEhnOjOunhf9Hq8dd09xbSDdtJ73nqeenSS6VTTmlpexiLicl05JHSEUf06bXXbL24bp1G1r4qa2Sk4nyxLdv5Ou35dmm7v3U4X5tdH/d83/B8e/W+ruLjru8zztd7u/Q9v/ytsPhOhf+7vm/YKqx46X7lzzTFty2tn2uf2KXv47bnMc+3KfdrbcP1Pcz2fp+1fftC7p8RbM/PDJU/l3m3zf+jTvE9X3xpSP/y/QNl24b67z1dV356voamjWvcmtBn//Z9PTrynJ6eLf18F+ljD6yXfvhD6WtfU6Puum5c42P3Kp1JSZrd8OsBAAAAAAAwtRAuAyIg5w6XdXE9GlXZXDagNY+6njDvUe34kqS3xDtcZvaOeB/Y4i7tc6ekXRsLl6UsQznVbi7btElatapwZ96jesM6Kb14C2n69JrL8YbLFtRsE3vU/TlqIgDoaS7L9zotacPRaS6r2L67FHr7PGMx7ZTypnxzNqeuO+4o3DDy0oIHpZff7PqoJR34f6X+9TrnKmmXtZK+doF0yDu7sKYApprNltNuNTGR1q33n6izz3Ye//GL+2mvHzvBoX/dXzrpPilzwQXSGWdIs2Z1fD0Zi9k805SWLDa1ZPECSe1tzHxqtfTLX0pr1g3o4RdO1Kc/7Tx+3rZL9KErPyRJ+vbbpXc/JRlnnSWdcIIvjd4YxmJishmGNH++ofnz50ia0+3VQRPWjko/+pE0OtqjPz96sr71LefxHz33Zh1wyQGSpDPfLn30QSl13nlOM+e0aQ0t481ve2O7VxsAAAAAAAAJRpUIEAG2619im3FrLuspN5etWDyuex9wBbHmPqqdXpO0006dX7k2Mnu8rVx9057Wm16R1F8/XOYubEpZzsXDWs1ljz3mutPA/ps3TzLMQmitTnNZKXzVu16pgZe0bJ0S1VxW2j4jJ815wgnPhTwGaS4L9vrr0gMPOLfNRQ9Kfeu9T1jwgNS/XrNH5Byzpintt1+nVxPAFHXGGeXg8w9/KN1/vxPg2XPxnnrvsvdKkp6dKf3r26Tc8Cbpy1+WstmOryfNZdH0ddc01W9+0wn527Z01E5Hace5zs9Hty2VfvYmKf/M09J3v9vSzwY0lwGo58tfLv+95JxznL8j2rb09qVv19uXvl2S9Le50vf3kfLr1jpfyPg7CwAAAAAAACYRzWVABOTzrnCZGaNwmQypd0Pp/l8276aHbn5E0p7OA/OSMRbT7PWOxdzzJSltKVS4zNtc5iTNajWXeUc6PuLsv13r779UShqYuUnD62ZIr+2sF+Yu1u4BzWWbN0vPPVd8/0e1/HUps3CxNGNG3WUUeZrLcr01W9I6LZ93BfRmP6WFoxOaZQ5IW24Z6vWe4NxLu+u+hVdK993X9vWMmxtuKF8Mt7a6sfIJS2+VJL3tecm0Je22a0PHFAC0YtEi6dRTpQsucL4d7bab8/23v1+y9Wsp63wf/4akb6QkXSzp4nFJnf7eZUgaLNymtSoq3vxm6X3vk379a2nNGudHBtOU+vpM5e37pZxznBwn6bi0pG9K+mYLofpssV2IYwBAsKVLpeOOk376U2l42PnrdPH7mqU/Slnna9AZks5IS/p3Sf/e2NelXXb5tR78y0favu4AAAAAAABIJsJlQATkXWMxFaPmMtMwpS3ukgZelUbmS499QBPFDw69oCF7sxaPpqRly7q5mi3rWfC0tPPl0rMHSB86WnOKWbMGw2Vmobls/V9P0H77SZblhMJM0/kzsvE1PfLUiKSlzgvmPaqd/qrQrVvL3/aw7rvurVJ2UO9/abX2+dw6mXMK7WnWuJ5+4X5tyhqS9iq8/yNNNcsFjcX8/gtH6doD/SO/pI3rXtTz655VTtUDdc0yAs4VK5/R2Ng+zp15j2jHNXKuxpjhijrf8x4pnbaVyxnSbV/Vvy88WBfbo0ovvkOGacmQLSccUP7E2rbhnLe2yo/bhuziU0o3nP/bnsecx+3SB+R7rut5ga8xKl9TfMywa3/c9Zj7eZ5lFm5PrNuh/NDSW6Tn9pfH7CclyWnBk2J/zgOIn698Rbr6amn1aue+bUsjI5KUktTYqLCO6H9ds8bUlfGcqPQv/yLdfLMznlxyfkZzjp9M4c8k4BgAUMM3vuGEXl991blf/r6WVju+r738+hx97Qv/pZ6Mqf/73WNafj8AAAAAAAAkG+EyIAIsVyLHjFF5gWEYUu+w9J4zpGsu9X6wEFwytl8u9fR0ZwXbxDQM6eiPqJgrMotjF5sZizk6Q5t//wPdEZi1mue721j4610n3aL7/nd7acR5n7v+Ntv10V5Jewe//xsbC5f5x2KuH9lWX3jxi9KLQc9eXPjTBU2E57bfXjr9dEPf+U7hgZd312jNV0wxQy9I2/xJPc/sVw6SSlLGGYc7UJwyNzDQ6TUDMMUtWuSMwzznHOmuu5ymztHSF3Bb68fWa8PYBuVzE7Jy2UJAtzuB/vTAGi1/4zn61mu7SGed1ZV1gNcb3yg99JD0ve85/9+8uTzx27YtrRtbp03jm5TPTsjK5yTZvlB4Y1J9r2vbN/+L/nXVG6Rz/rUt2wAgWZYulR58UDr7bOmee5wGs+L3NVu21o+9rg1jG5XPjcsqjsRs4Pva2ld21b98/91Sz0b93+9OwgYAAAAAAAAgUQiXARHgGYsZo+aykl3/ywk03fMp6fXtpP61GtjjPJ3xlz7pH/+x22vXMqNYP1b4X2lyacPNZaY0Mley6oXtLGm3/9RRLz6vXXZ/j7TvvqHWc+ZsSzr2vdLvvq/MC3spa/dVf/Lsx7XF0p/pxFsXSBecGur9i/xjMYdHt2jo9R0x41nN336FPn7HbOmfPtXQS7/5TWl8PK+fXDqszeuGJmf9qgpqMbQDHnM9Hvaxuu9ZYzmGpfRWN6nvHf+ofV4Z0aIn0vov92qnnStd/YXrWmHODQBotzlznAaqSoakWYU/UXFbt1cAPltt5YxWrWRKmlv40043tPn9ACTN/PnSdwODX4ak2YU/zemb84jGR+dLitG/bgMAAAAAAEDXEC4L4bnnntMPfvADXX/99XrhhRfU29ur7bbbTh/60If06U9/WgNtamj57W9/q4suukj33HOPXnvtNc2bN0977bWXTj31VB188MFtWQaiyR0uM2L2u11DhmzDlt56rvTWc/V/bjf0nRskc9Y/yLxnszP3MeYM3y/cjSbDZSnLkLKDpfunnCL96EfO6KV8XtrtS0P629CEBnOW1n79NfVmPupLctXWl+6TFv9FOunt+tkV0tGn/VjWSafIsqQvfPUt+rfe+yVJv9/1bO3/3mPVl35WRm9v6JGRRd5wWZ/GrPI2ff3rzp+ief/Uq0290rKNGa38xqqGllNLvX+U39MzU32Zh2T29jW8fX190nnnpXTeeU6wzPn82Mrmcsplc7LzWRmWLaNQM2gYhgzTcI4Tw5Bh+B6TnNuFxyTnmCo+VniTwtIbW9fO+kDhj/SP8yS5L3TRXAYAAADExlEH3alXX/+zejKWpJO7vToAAAAAAACIOMJldfzqV7/Sscceq40bN5YeGxkZ0cqVK7Vy5Ur95Cc/0fXXX69ly5Y1vQzLsnTqqadqxYoVnsdXr16t1atX65prrtEpp5yiCy+8UGaDIQnEQz5fbiuKW3OZYRiyXUmflGznC8v0oUQEyyRXc1lBI81l7lPWtExpohzEmjbN2UWplJTJSGM9I1Imr2lZU73TZjS8nr2pcuprPCUZE+Ol958wNkqZcUnSwoUL1T99ZsPvX2QYzqTTiQlJ+V5N2OUg0YwZ5fCZbVka6Z+QTGnITGnGnOaX2U2mKZmmoUwmI/VnJNHKVfGtKFNoLiuGy2guAwAAACLrv68gUAYAAAAAAIDwSCrVcN999+nDH/6wNm7cqGnTpunMM8/UHXfcoT/+8Y/6h3/4B0nS448/rkMPPVSbNm1qejlf+9rXSsGy3XffXT//+c9199136+c//7l23313SdJPfvITfd1dB4REcbcwmWbMwmWqErwaHKx8ckxVNJcVb7TYXObfRcNpJ2Q4mG/uS3Nv2hUuS0saHy+/d360vNzpzY9PKS2ruKhcrybywds0MbpZxU0ZtDMtLxPRUdGwSHMZAAAAAAAAAAAAACQSzWU1nHbaaRodHVU6ndYf/vAH7bvvvqWPHXTQQdp+++31pS99SY8//rjOOeccffOb32x4GY8//rjOPvtsSdKee+6pW265Rf2FwMpee+2lww8/XAcccIBWrlyp733vezrppJNaaklDNLnHYppxG4tpGJIrD1cqXps2rSvrMxna1lxmG57msspwmfPGA1ZzjW996b7S7bG0pLGx8ntb5aDZQBvCZX190qZNknJ9mrCCt2l4w2vlZRqEy5KkorksXWguyxXu01wGAAAAAAAAAAAAAIlAc1kVd999t2699VZJ0sknn+wJlhWdfvrp2nHHHSVJ559/vrLZbMVz6vn+97+vXM65Gn/BBReUgmVFAwMDuuCCCyRJuVxO5513XsPLQPS5x2JWNAJFXNXmsiSFy6ptY4PNZaZlVm0us/I5jRbyuIHejQAAflJJREFUV4N2c7lf/1hMT3OZPVFe7tDcpt7fs6ziovK9yrqay9yFVSOb1pWXaZTXDfFXORaT5jIAAAAAAAAAAAAASCLCZVVcc801pdsnnnhi4HNM09Rxxx0nSVq/fr1uvPHGhpZh27auvfZaSdIOO+ygffbZJ/B5++yzj97whjdIkq699lrZdrzGJqI+yz0W04jX59ff6lW6l6SxmP5tbDpcZkjZcujGvYtGN7qCWHZPU+tZq7lsROXw6+DM+U29v2dZxUXl+pTLB2/T8MY15cfN8roh/irHYhaay4qHGc1lAAAAQGTNWHK7jL4NMgdf7faqAAAAAAAAIAYIl1Vx2223SZIGBwe1xx57VH3eAQccULp9++23N7SMZ555Ri+++GLF+9RazurVq/Xss882tBxEn+Ueixmzs9I0vCucxOayqtvYxrGY7hGSg0Zz4bLetKu5LC1vc5nhpH56c1Iq09z7e5ZVXFSuV3aVNrZhd3NZinBZktBcBgAAAMRXNjtNGp8he3xGt1cFAAAAAAAAMRCzGEvnPProo5KkZcuWKZ2uPqJuhx12qHhNWI888kjg+7R7OYi+vKu6LDFjMZPUXKYq7WxtHIs5vHFt+XGzuRGStZrLhs28897Z9hxg7uayats0Mry+/HiaJqskqQiXpQvNZbnCfZrLAAAAgMiaNWOVemY/pr5Zf+v2qgAAAAAAACAGqqemprCxsTGtWeOMc9tiiy1qPnfWrFkaHBzU8PCwXnjhhYaWs2rVqtLtesvZcsstS7drLWd8fFzjrraijRs3NrRO6A7LNeo09mMxE9hc5t/GRprLKsZiVmsu21QOlw00GS7rTbmay1LyNpelnHDZQL49meJSc5mdlsaHSo97tmlkfen2QJomqySpHItJcxkAAAAQF6ufOLTbqwAAAAAAAIAYobkswKZNm0q3p4UIyAwW0hSbN2+etOUMuhIbtZZz1llnacaMGaU/7lAaosvKl2/HbSzmlGwua2Uspqvly52/Gd7sHiHZXOtTzbGYact5byvV1HtXLMudfxudXbrp2aaRDaXbgz3JOR4Q8HUq5aTKSuEymssAAAAAAAAAAAAAIBFiFmPpjDHXKLmenp66z+8tpCxGR0cnbTm9riRHreV85Stf0YYNG0p/Gm1TQ3fEeiymv7mseIPmssJr3a9LVW0uG3EHsTLNtT7VGos5UuipbFe4rK/PdccVLvM0l42WmxMHe5JzPCDg61Thfj/NZQAAAAAAAAAAAACQKIzFDNDnSk1MTEzUfX5xDGV/g00tjSzHPeqy1nJ6e3s9QTTEg22Vb5tmzMZiTsHmstI2hgjQuEM4hq+5zBPEGl5ffjzT3L6rGIs56nzdmBjdrFwhUzZoZ5p674plVWku8wTmxsrtjIN909uyXERDtYZFmssAAAAAAAAAAAAAIFkIlwWYPr0cgggz6nJ4eFhSuBGazS6nuIxmloPoy9vlQFkq5s1lpeBVgo7Tqu1szYzFrNJcNjxabi4baHKEZLXmsuH1r5Xf22hPuCyoucw0bfX2lvfV8ET569oA4bJEqRYu6ymO+KW5DAAAAIisQw9eoedfGZRh2Hrg3r/v9uoAAAAAAAAg4giXBejr69OcOXO0du1arVq1quZzX3/99VLwa8stt2xoOVtssUXpdr3luMdbNrocRF8+H+OxmL5WL6OBVq+4qNpc1uBYTMMypWw5lVV1hGRvc8G83rSruSwtqdB4OLxxTfm9jfY0G3qay0bmSJIGevMyjPK3FXe4bLB/RluWi2io9nWqkeAlAAAAgO64+b5dNfzKnpKsus8FAAAAAAAAqnSPYKeddpIkPfnkk8rlclWf99hjj5Vu77jjjk0tw/8+7V4Oos9VXFa1ESiqTMO7wqYtJ1gWtw2pwb+NRpPhMtM3FtOdvxsZdwWx+oaaWs9qzWUjG9eW39vs87+suWV53sbZP4O93q+VIxMj5eUOzmzLchENdU/vBIVLAQAAgOSx6z8FAAAAAAAAKEhO+qPN3va2t0lyxlHee++9VZ938803l27vt99+DS1jm2220eLFiyveJ8gtt9wiSVqyZIm23nrrhpaD6HM3l8UtkxU4MjJBIzGlKqM/e3pCfbLcTzFsszQWM9Ob9XzM2/LVXLisN+VqLkup3Fw2/Hr5vVPtCZf1BhSg+cNlwzl3uGxWW5aLaMjn6zwh6AABAAAAEAmXfX9Y//Xvv9HP/uN33V4VAAAAAAAAxEDMYiydc8QRR5RuX3zxxYHPsSxLl156qSRp5syZOvDAAxtahmEYev/73y/JaSa76667Ap931113lZrL3v/+91cEXRB/lmsSRdw+u4EjI93zHhMgcBtDjv3zjMW0zVJzWU9f1vO84Ynh0u3BgZlNrWfKTCllpCT5xmJudoXL0u0ZVxgYLuvxbVN+tPyxabPbslxEQ6EUrzq+TwEAAACR9b6PvEPHfvIQHfOJQ7q9KgAAAAAAAIgBwmVV7L333tp///0lSStWrNCdd95Z8ZxzzjlHjz76qCTptNNOUyaT8Xz8pptukmEYMgxDJ5xwQuByPv/5zyuVcsIgn/3sZzU6Our5+OjoqD772c9KktLptD7/+c+3slmIqLxVbi5LxeysDGz1SnhzmSE1FS4zXc1l6b4Jz/OGs+WWr4GBGU2tp1Qejekei+luLhtIt2dcYV9AAdpgxrdN+XICaWAazWVJUsgtAgAAAAAAAAAAAAASLmYxls46//zz1d/fr1wup/e85z0666yzdNddd+nGG2/Uxz/+cX3pS1+SJC1fvlynn356U8tYvny5vvjFL0qSVq5cqf3220+XX365Vq5cqcsvv1z77befVq5cKUn64he/qO233749G4dIscvZsviNxfS1ehk0l3mf6x6LKaPUXJbu9QWx3CMkWwhi9aadSjHPWMyRDeX37mnP5yaouWwg7W0uG7HL2zg4NKcty0U01G0uAwAAAAAAAAAAAAAkQrrbKxBlu+++uy6//HIde+yx2rhxo7761a9WPGf58uW6/vrrNX369KaXc+aZZ+rVV1/VT3/6U9133336yEc+UvGck08+Wf/8z//c9DIQbXnXWMzYhcuCmsumJ7y5rMmxmHI3l/V60zkjVrkKqpURkkHNZSNjm8rv3dOez01gc1naW2c17A6XzZjXluUiGjzhMnOi6vMAAAAARM91P/+TXn3Z+XviSZ87TGbcKtQBAAAAAADQUYTL6jjssMP0wAMP6Pzzz9f111+vVatWqaenR8uWLdPRRx+tz3zmMxoYaG3MnGmaWrFihY466ihddNFFuueee7RmzRrNnTtXe+21lz7+8Y/r4IMPbtMWIYosd7jMF2SKuormMil5YzFbaC5zfzpXDh9Uup3KrpE2zJBmOCMwh61yWqeVlq/eVKG5LK1yc9nYxvJ79zUfhPUsJ6C5bDDtDcwNG+XQUf9Q84E5RI8nXJamxgwAAACIk2NPH9Cml5y/n57wmRzhMgAAAAAAANREuCyEpUuX6txzz9W5557b0Ove8Y53yHbPO6zjkEMO0SGHHNLo6iEBYj0WM6i5LGljMYO2sYmxmG7ZiZekgz8j3X67ZBgatsutXwPTmw+XFZvLNvRKa81xzVm9WsPjm8vv3aZwWWBz2dg6z/1hIydJ6s9KZopvN0lCuAwAAABIBitvSZlurwUAAAAAAACijKv9QATkrXK6LHbhsqBWr56e7qzMJAlsZ8uE++17tSK617JL9e5ld2rZqYtlzpqjBzOvlz42OLP5EZL7bbmfHl3zqCbS0i6fkt5/2rb66xYZaVbhvftnNP3ebvvuKxmGLdsub+At1l/0yc8eL7N/QKZh6qEZTmBuMBevNj7UR7gMAAAAiK937fmwnnnpKRmSzNSHu706AAAAAAAAiDjCZUAEeMdidm89mmEa3jScYat6oiqm/Nto2gqdAqy6K7a6Xf+7nfS/elnSy6WH03mpp7/5saLfOvBbuvyhy7Qpu1kvT5cufOOEpPJ4ysHBmU2/t9vOO0unnmrowgvLjz2xcFhPzH2m4rmDuZglJlEX4TIAAAAgvq6+7uRurwIAAAAAAABihCv+QAQkbixm3DaiDv82NhKgq7YrzN0vCnz8qM1bNrJqFRZPX6yff/AyzU1XNpQtHjb1loOOa+n93f71X6X3HuyMvlRqXNr+t4HPO6pnt7YtE9FAuAwAAAAAAAAAAAAApgaay4AIyLuay1Ixqy4LHBmZtHBZ0OjPFprLvqTv6Bv/+Rc9/YEDlf/Mp2RlJ2Tlc+rtG9Qubz2y5fU9dPmhevkra/XU2ic08szjyg9vkpXP6Y37vF9902a2/P5FM2dKv/1NWmvXSutHRzU89q+yXlytvJWTZeVl5XMamjFfO+59aNuWiWg49VTpj38s3Nnn+5KkL99auH/UUd1YJQAAAAAAAAAAAADAJCBcBkSA5WouM2KWy5qKzWWthstMWRrMSm9MLZLe9sE2rGGllJnS8nk7SPN2mJT3d5szR5qjmZL2k5ZN+uIQAR/8oPSDH0j5vDT7HQdp1aZl+vzub5D2e0D63Oe6vXoAAAAAAAAAAAAAgDYhXAZEgHssZuyby5IYLmuhnS3oaaas6h8EYsA0pc9+tnjv2PIHDqe1DAAAAIi6bXa8Rqte2kWStPbp+RqaPdTlNQIAAAAAAECUES4DIsByjcWMW96I5rJ6r618jHAZAAAAAKBb1m+eo9wGp3Y6l13X5bUBAAAAAABA1BEuAyLAEy4LSiNFmL/VK5Hhshba2QiXAQAAAACipK9nWEb/Gkm2pFS3VwcAAAAAAAARR7gMiIC8ayxm3PJG/lavRkZGxkWr7WyG4R19SrgMAAAAANAtLz313m6vAgAAAAAAAGKEZAMQAe7gUSqVgOaymLWv1VPRXCY1tI3+pxIuAwAAAAAAAAAAAAAAcUCyAYgAyyqnj1JmvIJZpuH9MtLIyMi48G9jo81l/qcaKqQJExbCAwAAAAAAAAAAAAAAyZKsBAgQU5ZVvm3E7KxsdWRkHLRjLKbn9TSXAQAAAAAAAAAAAACAGEh3ewUA+MZixqy5LHAsZsJCUxVjMQmXAQAAAABi6pMnXKzHnnNuX3vVURqaPdTdFQIAAAAAAECkES4DIsCy4zsW09/qZUiJC0212lzmfyrhMgAAAABAt/zixq217vkDJUkb168mXAYAAAAAAICaSDYAERDrsZhTsblMorkMAAAAABBLhuz6TwIAAAAAAAAKaC4DIsD9a93YN5clMVzWYnMZ4TIAAAAAQFSc88URPffC/8g0pNnzjuj26gAAAAAAACDiCJcBEWBZrrGYqXgFjgKby/xpqphrdRsZiwkAAAAAiIrjP/O+bq8CAAAAAAAAYoRkAxABlqu6LHZjMVts9YqDVtvZaC4DAAAAAAAAAAAAAABxRLIBiADbdjWXxW0spq/Vy5ASF5oKbC5rIVxmFAehJqzhDQAAAAAAAAAAAAAAJAtjMYEIsKzy7biNxTR9VWtJbC7zb2OjATrGYgIAAAAAouKBPz+iNa9ukCS97d17qKevp8trBAAAAAAAgCgjXAZEgOVqLjPj1lzW4sjIOGh19CdjMQEAAAAAUfGuj6zWa8++W5L05EPPaLudt+nyGgEAAAAAACDKSDYAEeAOl6VSMQuXtTgyMg7aPRaTcBkAAAAAIAosd5U6AAAAAAAAEIDmMiACbLt8OxXz5rJEhstabGdjLCYAAAAAICr22uF5Pdx3jQxJg9Pf0u3VAQAAAAAAQMQRLgMiwPY0l8UrcORv9TKkxIWmaC4DAAAAACTF9b89udurAAAAAAAAgBgh2QBEgGWV00dGzM7KwOYyf5oq5iqay5wHG3i99z7hMgAAAAAAAAAAAAAAEAckG4AIcE3FjH9zWRLHYrbYXOZ/qlH8jCcshAcAAAAAAAAAAAAAAJIlWQkQIKYs11jMdNzCZUHNZUkLl/mbyxiLCQAAAAAAAAAAAAAApgCSDUAE2K5wmT/IFHWmb45nEsNlrW4j4TIAAAAAQFTs89ZLNX3xnzV98Z/1xINPdXt1AAAAAAAAEHHpbq8AAG+4LJWKV7isYiymlLjQVLvHYhIuAwAAAAB0y3OvzNfml94iSRre/FCX1wYAAAAAAABRR7IBiADLLt9Op1PdW5EmTMmxmBLNZQAAAACAWDIMu/6TAAAAAAAAgAKay4AIsCzXWMyY5Y0qmsuSGC6juQwAAAAAkBDPP/puSTmZpilpp26vDgAAAAAAACKOcBkQCe6xmPEKHAU2l/mrumKuormswW2kuQwAAAAAEBXpDL8OBAAAAAAAQHgkG4AIsOxy+iiditlYzBZbveKg1W0kXAYAAAAAAAAAAAAAAOKIZAMQAbYrXGbEK1tW2eolJS40FdjO1sJYTEN28Y1bXDMAAAAAAAAAAAAAAIDJQw8+EAE2zWWR5t/GRgN0NJcBAAAAAKLirK9dpvsfHZYknXvuIVq89aIurxEAAAAAAACijHAZEAHusZipVLwCRxXNZUkMl7XYXEa4DAAAAAAQFf9+xaBWPfERSdJpn7qXcBkAAAAAAABqItkARIC7ucyMWeDINLzrm8TmMtP3pbLRAJ3/qYTLAAAAAADd4v8HUAAAAAAAAEAtNJcBEeBpLkvHK3DU6sjIOKC5DAAAAACQFKcfN66H//ZTmYahbZYf0u3VAQAAAAAAQMQRLgOiwBUuS6dSXVyRxgUGrxL2z6D9ATrCZQAAAACAuDrtax/s9ioAAAAAAAAgRkg2ABHgaS7LxOu0bDV4FQf+AJ3hPBj69YzFBAAAAAAAAAAAAAAAcUSyAYgA2xUuM2PW+lURvEpiuIzmMgAAAAAAAAAAAAAAMAUxFhOIANsVXkpnYjYWcyo2l7UYLjNkB38AAAAAAIBJtu6VdRoZHpUkLdxygdIZfj0IAAAAAACA6pKVAAFiyt1clkrFLFwWNDIyaeGyFgN0jMUEAAAAAETFm99xi7bcbom23G6JbvvDym6vDgAAAAAAACKOZAMQAbZdPhXNVLzarKZicxljMQEAAAAASWBZdrdXAQAAAAAAABFH7z0QAe7msnTMm8sSGS5Ta+1shMsAAAAAAFGxw5brtGnkRhmyNXPWvG6vDgAAAAAAACKOcBkQAbYrvJTKxCtcZhregJSRwHCZfxsZiwkAAAAAiKvf/eGkbq8CAAAAAAAAYoRkAxABnrGYMQscBY7F9Fd1xZy/nc1ocBtpLgMAAAAAAAAAAAAAAHFEsgGIANsu306n49VcVhG8khIXmgoM0DEWEwAAAAAAAAAAAAAAJBzJBiAC3M1l6ZiNxWw1eBUHrQboGIsJAAAAAAAAAAAAAADiiGQDEAG2XQ4vGWa8Rkr6g1eJDJe1ubnMkB38AQAAAAAAJtlR71+hRdv9Tou2+51W3np/t1cHAAAAAAAAEZfu9goA8DWXpeN1WvqDV0YSw2UtBugYiwkAAAAAiIr7n5qpl59+ryTp1Zdu7vLaAAAAAAAAIOpINgARYLtup9MxG4s5BZvLGg3QMRYTAAAAABAVdGgDAAAAAACgEfGqSAISytNclonXaVkRvJISF5qiuQwAAAAAkBR/vGZvbVz3qAzT0HY7vaXbqwMAAAAAAICIi1eKBUgod7jMTMUrcBQYvPKnqWIuMEDXwDYSLgMAAAAARMVWy7bs9ioAAAAAAAAgRkg2AJFQTh/FrbnMNLxfRpI4FrPVbWQsJgAAAAAAAAAAAAAAiCOSDUAE2HZ8w2UVrV4JDJcxFhMAAAAAAAAAAAAAAExF8UqxAAnlGYsZs8BRq8GrOGg1QOd/qiG7cCNZ40MBAAAAANH3s//4jf7ywGuSpM+ffhBjMgEAAAAAAFAT4TIgAuwYj8WsCF5JyQuX0VwGAAAAAEiIM3+0WY89cLwkab99biBcBgAAAAAAgJpINgAR4B6LaabidVpOyeYyiXAZAAAAACD2bLvbawAAAAAAAICoi1dFEpBUpbGYluKW+fQHrxIZLmsxQOd/KuEyAAAAAEC3HH+4pduXrJBpSDvvun+3VwcAAAAAAAARR7gMiIDSWEwjhuEyX/DKSGK4rMUAHc1lAAAAAICo+PK3P9LtVQAAAAAAAECMkGwAIsAuNpcZVndXpAmBwSt/mirmAgN0DWwj4TIAAAAAAAAAAAAAABBHJBuAKCiFy+zurkcTTMP7ZcSQEhea8m8jYzEBAAAAAAAAAAAAAMBUQLIBiAC7mCmLY3OZ0drIyDjwt7M1GqCjuQwAAAAAAAAAAAAAAMQRyQYgEhI2FjNhoalWA3T+cJkhO/gDAAAAAABMst33+m8Z6TEZ6TH9/Me/6/bqAAAAAAAAIOLS3V4BAJJdHIup+I3FrGj1SmK4rMVtZCwmAAAAACAqcvmUlO+TJOVz8ftHbgAAAAAAAOgswmVABJTCZXFsLjNaGxkZB+1uLiNcBgAAAADolsVzRvTk3PtlGNLQzL5urw4AAADw/9u78zC5yjJ/3J/u7FtnJZCQkISEsCtqQBCBIALKIgYURFRWN8bRmfGnjjrq6MyIon6dGccFMRAREdlkERQRIcoui4qEJftGyErSSTpbd9fvjyZFOtWdhe6kernv68qV0/2e89ZT51Q9feqcp94XAGjjFJdBm/BK9VF7LC5LE1M7drCiqZZO/am4DAAAgLbi7nsuKncIAAAAALQjKhugDdg8cllFeywu27pyquGXuz+QXajJ0dl24jmaFhMAAAAAAAAAaI9UNkBbUNhcqFQoaxivhZHLdmB7I5cBAAAAAAAAAO2QygZoAwqFdjwtZlMjeHWwoqmSkcsUlwEAAAAAAAAAnYDKBmgTXnkrtsPissqKJtJIByua2vo57uzIZabFBAAAoK341EemZP/X3Zj9X3dj7r/zkXKHAwAAAEAbp7IB2oBCoeGtWFFhWsy2qLWnxazYPP1pU6O+AQAAwC70x6e65YWn35sXnn5vZrywuNzhAAAAANDGdawKEGinCvXdkiQVlRvLHMnO65TTYiamxQQAAKBd8jUnAAAAAHZG13IHACSFuu5JkorKTWWOZOcZuWz7TIsJAABAW3HV/74+s2f8PpUVlTn6hKPLHQ4AAAAAbZziMmgDNo9cVllZW+ZIdl6TI5d1sOkeS0YuK2SnnqORywAAAGgrDjvqkBx21CHlDgMAAACAdkJlA7QFm4vLurTDaTGNXLb97bfaRRUpvNJRx9pPAAAAAAAAAEDHorIB2oLitJgdZOSyDlY0tfVzbOm0mBXNNQAAAAAAAAAAtCGmxYQyq6+rf3XksspNZY5m53XGkcsqkhaNXFbUwfYTAAAAbd/9dz6Sv/5lQZLk3e89IqPG71PmiAAAAABoyxSXQZnVrKlJ0jdJ0qU9FpcZuWwHtm+moYPtJwAAANq+z/zH9Dz+6AeTJF263JxP/KviMgAAAACap7IByqxmdU1xuUs7nBazsqKJNNLBiqa2fo4VLZwW89WOmqs6AwAAgF2vUCh3BAAAAAC0dUYugzJbV7O+uNylS/srLuuM02IauQwAAID26rRjK1NZeU0qkhz6uvHlDgcAAACANk5xGZTZmuotRi5rj8VlpsXcrmZX7WD7CQAAgLbvK5efl6+UOwgAAAAA2g2VDVBm69ZtKC63x2kxmxy5rINN97j1c6xIduo5GrkMAAAAAAAAAGiPVDZAmdWsfrW4rKuRy9qklo5cprgMAAAAAAAAAGiPVDZAmW05clnXLvVljOS16ZQjl5kWEwAAAAAAAADoBFQ2QJmtX7epuNxhRi7raMVlFU1Mi2nkMgAAANqhU95xVbpWzU/Xqvn54bd+Ve5wAAAAAGjjVDZAmdWs21hc7tqlroyRvDZNjlzWwTT5HFujuKyDFeEBAADQ9q1a0yV1q0embvXIrFmzcfsbAAAAANCpdS13ANDZNR65rB1Oi9kJCqSafI6mxQQAAKAd6t+3LpV9FyYppFcvlwYBAAAA2DZXkKDM1m1RXNata/sbuayyouMXSDX5HE2LCQAAQDt0128v2uKnEWWLAwAAAID2QWUDlNn6Da8WlHXrUihjJK9Np50WcydGbFNcBgAAAAAAAAC0RyoboMw2NCouMy1mW2RaTAAAAAAAAACgM1LZAGW2bn1tcbl7NyOXtUVNPsfWmBazExTmAQAAAAAAAADtl+IyKLMNG18draxb13ZYXNYJCqRaOnKZ4jIAAADaim995Zc56uhrctTR1+Tum/9U7nAAAAAAaOMUl0GZbdz4akFZDyOXtUktHbnM7JcAAAC0FXdOXZdHHvpQHnnoQ/nzn+eXOxwAAAAA2jglD1Bm6zdsOXJZGQN5jYxctiPbt2IwAAAA0AIVFe3vi20AAAAAlE87LGWBjmXDplcv6nbvVsZAXiMjl+3A9h1/FwEAANBOfPUzB+fPE29IRUVFTp/05nKHAwAAAEAbp7gMymzDltNidm9/VUiVFR1/AMQmn6NpMQEAAGiHjj3liBx7yhHlDgMAAACAdkLJA5TZxk2vLvfo1v6KyzrttJg78bw7wS4CAAAAAAAAADogxWVQZptqX13u2aP9vSVNi7kD23f8XQQAAAAAAAAAdECmxYQy27Dp1cqj7u1wWsxOO3KZaTEBAABoh57/64w8P21ekmTCUQdm+OhhZY4IAAAAgLZMyQOU2abaVwuXenRvf29JI5ftwPYdfxcBAADQTnzi03/KGe9/W854/9vyo+/dW+5wAAAAAGjj2l8lC3QwG7cYuaxXT4MJtkVNjly2ExVjissAAABoiwqFckcAAAAAQFunkgXKbFPdqzWePXp0KWMkr00hHf9KdJMjl+1ExZhpMQEAAGgrjnp9t8xdfFOSQg45aI9yhwMAAABAG6e4DMpsU92rRUo922NxWSf4mnOTI5ft1PatFAgAAAC00Ne+84F8rdxBAAAAANBuGE8HymxT7asFZb16dStjJDSnyZHLdmZ7xWUAAAAAAAAAQDukuAzKrHaLaTF7Ki5rkyorWpYqTYsJAAAAAAAAALRHSh6gzGrrXh25rHev7mWMhOa0dFpMAAAAAAAAAID2SHEZlFltXdficnscuayQQrlD2OVaOi0mAAAAtBUf+dBV6T30L+k99C/5zldvKHc4AAAAALRxXbe/CrArbTktZq9ePcoYCc1p6chlhY5ffwcAAEA7sXBpIeuWHpYkeWnx38obDAAAAABtnuIyKLMtRy7r3a/9FZcVOkHlVEtHLusEuwgAAIB2omtlksqNSUUhFeY0AAAAAGA7FJdBmdXVv/o2bI8jl3WKaTFbOHIZAAAAtBW33XnxFj99qGxxAAAAANA++H4ilFndliOX9elVxkhoTktHLgMAAAAAAAAAaI8Ul0GZbVlc1qdf+ysu6xTTYrZw5LJOsIsAAAAAAAAAgA5IcRmU2ZbTYvbu17uMkdAcI5cBAAAAAAAAAJ1R1+2vAuxKdfXdisu9+youa4sqK1pWh2vkMgAAANqKX1z52/zytoVJko9esF/e+Z5jyxwRAAAAAG2Z4jIos/rNxWWVm1LZpdu2V26DCun4lVOmxQQAAKCjuOH2hbntzouTJPsMm6K4DAAAAIBtMi0mlFlh87SYlZvKGwjNMi0mAAAAHUULvz8FAAAAQCdj5DIos/q67g0LXTYmaX/TYhY6wbBcRi4DAACgo/jo+eMzcq8pqayoyNnvPbTc4QAAAADQxikugzKrf2Xksgojl7VZRi4DAACgozj5rGNy8lnHlDsMAAAAANoJ02JCmRXquyVJKrpsLHMkr00hHX9YrpaOXAYAAAAAAAAA0B4pLoMyK9Q3TIvZXkcu6xTTYrZw5LJOsIsAAAAAAAAAgA7ItJhQZoW6hpHLKttpcRnbp7gMAACAtmLF4hVZ/OKyJMmwkUMzYMiA8gYEAAAAQJtm5DIos80jl7XX4rLOMC1mZ3iOAAAAdA4f+9ivctAbx+egN47P5/+/X5U7HAAAAADaOCOXQZn99Y/zs3b1ulRWqvXsqIxcBgAAQFvk8yoAAAAA26O4DMrsdW8+qNwhsIu5WA8AAEBbcfB+PTN1zO9SUVHIfvv2KXc4AAAAALRxissAAAAAoJP4yuXn5SuXlzsKAAAAANoL8/ABLVLoBMNydYbnCAAAAAAAAACwNcVlQIsU0vELr1r6HNWmAQAAAAAAAADtkeIyAAAAAAAAAAAASiguA1qkM0wZ2dLn2Al2EQAAAO3E1z53XYaMvjdDRt+b//rCL8odDgAAAABtXNdyBwDQ0SkuAwAAoK2YMWd9ls89IUkyffZVZY4GAAAAgLbOyGUAAAAA0ElUVJQ7AgAAAADaEyOXAS1SSMcflqulz9HIZQAAALQV3/vBe/K1r81Lkgze8+wyRwMAAABAW6e4DAAAAAA6iapBVakaVFXuMAAAAIAO6iffvS0/uG5NunWry6XvH5TzP3FauUOihUyLCbRIoRMMy9UZniMAAAAAAAAAtNR9Dy3LU4+fl8ce/lCmPryk3OHQChSXAexiatMAAAAAAAAA6Aw2bHp1uWePivIFQqsxLSbQIoV0/MqpzvAcAQAA6Bzuv/OR3HDzsynUJ+8+fVxOPuuYcocEAAAAdCBf/vxROenYW7NhQ12OP+HIcodDK1BcBrRIZ5gysqXPsRPsIgAAANqJa38xLZN/flGSpHrtZMVlAAAAQKt63ZsPyuvefFC5w6AVmRYTYBdTXAYAAEBb5PMqAAAAANtj5DKgRTrDlJGd4TkCAADQObz7tH2zas3kJMl7zxhV5mgAAAAAaOsUl21HTU1N/u///i833nhjZs6cmQ0bNmTkyJE59dRT88lPfjKjRrXsItycOXMyZsyYHVr3/PPPz5QpU1r0eMDu55vgAAAAtBWnvW9iTnvfxHKHAQAAAHRQD//+icyasTg9e3XLcSe9IUOGDSl3SLSQ4rJtmDFjRk455ZRMnz690e+ff/75PP/88/nJT36Sn//85znttNPKFCEAAAAAAAAAALQN//D55/LU4+clSa74zq35yL+8u7wB0WKKy5qxevXqnHrqqcXCsg9/+MN53/vel169euW+++7LZZddlurq6pxzzjl58MEHc9hhh7X4Mf/zP/8zZ5xxRrPtAwcObPFjAAAAAAAAAADArlBbX1lc7t2rexkjobUoLmvGt771rbzwwgtJkssvvzyf+cxnim1HHXVUJk6cmOOOOy41NTX5p3/6p9x///0tfsy99947hxxySIv7gd1pVP9Xp4Z9/UtJhg4tXzC7yOBeg1u0/ejRry6/MU80LFRUtKhPAAAAAAAAAGhrjn/ThtTXX5/ausrst/9+5Q6HVlC5/VU6n02bNuV///d/kyQHHnhgPv3pT5es85a3vCUXX3xxkmTq1Kn585//vFtjhLbikjdekmP2OSZj1vfKDX/aK7n77nKH1OoO3/vwnHPwOdmrrlf++KsByW9/u1Pbf/SjyVvfmuy7byHXnzwlGTYseeCBXRIrAAAAbMulF12dior6VFQk73335HKHAwAAAHQw//PjC/L3p96X5/52dt78tjeUOxxagZHLmnDfffdl1apVSZLzzz8/lZVN1+BdcMEFueKKK5Ikv/rVr3L44YfvthihrejWpVv+eOEfUygUUvH1dNgRua5/z/UpnFVIxVez08+xW7fkT39KCoWKVFR8Lyn8b4fdTwAAALQHDde6CmWOAgAAAIC2T3FZEx7YYkSh4447rtn1JkyYkN69e6empiYPPvjg7ggN2qyKTlAs1dLnWNy8E+wrAAAA2qa99+qZPns2jMA/ck+XBgEAAADYNtNiNmHatGnF5QMOOKDZ9bp27Zpx48YlSZ599tkWP+73vve9jBs3Lj179kz//v1z8MEH52Mf+1iefPLJFvcNAAAAAF/8+rlZ89LhWfPS4fnuFeeXOxwAAAAA2jjFZU1YsGBBkqRPnz4ZMGDANtcdOXJkkmTp0qXZsGFDix73ySefzMyZM7Nhw4ZUV1dn2rRpueKKK/KmN70pH/vYx3ao/83bbvkPAAAAAAAAAAB2tf1fd2N6Dp6WXnv8NQtmLSx3OLQCY983YfXq1UmSvn37bnfdPn36FJfXrFmTHj167PTjDRgwIJMmTcrEiROz3377pWfPnlm0aFF+97vfZfLkyVmzZk2uuOKKrF69Oj//+c+32ddll12Wr371qzsdAwAAAAAAAAAAtMTSlXtkw4qDkiTduy8pczS0hopCoVAodxBtzdixYzNr1qyMHDky8+bN2+a6H/rQh/Kzn/0sSTJ//vyMGDFipx5r48aNqa2tTe/evZtsnz59et7+9rcX47jtttvyrne9q9n+NmzY0GiEs+rq6owcOTKrVq1KVVXVTsUGAAAAAAAAAAA7as9978mShUcmdd2yenlt+vbf/sBOlEd1dXX69++/3Zqidj0tZkVFRYv/TZkypaTfnj17Jmko/NqeLQu5evXqtdPPoXv37s0WliXJfvvtl2uvvbb48/e+971t9tejR49UVVU1+gcAAAAASTL5f27P2IN/lX0P+lW+9rnryh0OAAAA0MEsnnViChv6pVDbU2FZB9Gui8t2lX79+iVpmOZye9auXVtc3pFpNF+LY445Jgcd1DBk4AMPPJD6+vpd8jgAAAAAdGx/n/ZyZk2blNnPTsqfn15X7nAAAAAAaOO6ljuAlnj22Wdb3MewYcNKfjdixIg8+uijWbt2bVauXJkBAwY0u/38+fOTJHvssUd69OjR4niac9BBB2XatGlZv359li9fnj322GOXPRYAAAAAHVOFr5oCAAAAsBPadXHZAQccsEv6Peigg3LzzTcnSZ577rkceeSRTa5XW1ubmTNnJkkOPPDAXRLLZhUVFbu0fwAAAAA6vv/vcyflnSc9mCQZd8DbyxwNAAAAAG1duy4u21Xe+ta3FpenTp3abHHZ448/XpwW8+ijj96lMU2bNi1J0qNHjwwePHiXPhYAAAAAHdPw0cMyfHTpSP4AAAAAreGsMyZn7bqKDKoq5LqbLi53OLQCA+E3YeLEienfv3+S5Kc//WkKhUKT602ZMqW4PGnSpF0Wz4MPPphnnnkmSUPhW2WlwwYAAAAAAAAAQNty6x9Ozt33XJQbfndSuUOhlahSakL37t3zyU9+Mkny7LPP5tvf/nbJOg8//HAmT56cJDnuuONy+OGHN9lXRUVFKioqMnr06Cbbb7311maL15JkxowZef/731/8+dJLL93RpwEAAAAAAAAAALtNob57kqSicmOZI6G1mBazGZ/5zGfyy1/+Mi+88EI++9nPZsaMGXnf+96XXr165b777svXv/711NbWplevXvnv//7v1/w4kyZNyrhx43LmmWfmiCOOyIgRI9KjR48sWrQod999dyZPnpw1a9YkSc4+++yceeaZrfQMAQAAAOhsnnn8udz+q7+kvlDI0UePycRTjyx3SAAAAEAH8vV/+UNWrdqUPr27JBlb7nBoBRWFbQ2b1cnNmDEjp5xySqZPn95ke1VVVX7+85/ntNNOa7aPioqKJMmoUaMyZ86cZtu35+Mf/3i++93vpkePHju0/mbV1dXp379/Vq1alaqqqp3aFgAAAICO5fOf/Fm+8b0PJknecdJV+c3dF5U5IgAAAADKYUdrioxctg3jxo3LU089le9///u58cYbM2PGjGzcuDEjR47MKaeckk996lMZNWpUix7j9ttvz8MPP5xHH300c+fOzbJly7J27dpUVVVl3333zTHHHJOLLroohxxySCs9KwAAAABIfOUUAAAAgO0xclkHZ+QyAAAAADa764ap+eb3ZydJ3vvOfvnEv55V5ogAAAAAKIcdrSlSXNbBKS4DAAAAAAAAAGBXq91Um4fueTK9+nTPkKEDM+bAls0GyK5lWkwAAAAAAAAAAGC3WLF4RY479YgkycCR92fFPMVlHUFluQMAAAAAAAAAAADat7Wr1xWXu3apLWMktCYjlwEAAAAAAAAAAC3SrUfX7LP/bamr65oDRy0pdzi0EiOXAQAAAEAn8Y0vXZ/KPotT2WdJzjh1crnDAQAAADqQEfvunbnPnZEF00/NPb+/sNzh0EqMXAYAAAAAncT6dXUp1OyZJFmzzvdOAQAAANg2xWUAAAAA0EkMGNAjXfvPSpIMrKovczQAAAAAtHUVhUKhUO4g2HWqq6vTv3//rFq1KlVVVeUOBwAAAAAAAAAAKLMdrSky9j0AAAAAAAAAANAid90wNVXDH87AkX/Ke989udzh0EpMiwkAAAAAAAAAALTIggUvZ/WidydJZu41r7zB0GqMXAYAAAAAAAAAALTIunW1xeVu3erKGAmtqaJQKBTKHQS7zo7OjwoAAABAx/ebm/6YL31rXgqFZNLbKvNv33h/uUMCAAAAOoj6uvps3LAxNatrUtmlMgOGDCh3SGzDjtYUmRYTAAAAADqJvzz1Yp547ANJkj49p+TfyhwPAAAA0HFUdqlMz94907N3z3KHQisyLSYAAAAAdBIVW1wN/NOfLkiXfgvSc8gz5QsIAAAAgDbNyGUAAAAA0Emc+8G3ZGD/W5MkH/vMu1O/ZkQ2rBlR3qAAAAAAaLMqCoVCodxBsOvs6PyoAAAAAHQuFRWvLrtCCAAAALTUPb96MD/9xXPp2qUi7zlj35z2vonlDolt2NGaIiOXAQAAAEBnVLkxqe+eXnv8JclhZQ4GAAAAaO9u/NXz+fmNFydJ1m+YrLisg1BcBgAAAACd0LXf/33Wb6hNv349yh0KAAAA0AHU1r263KVL+eKgdSkuAwAAAIBO4qV5izP1d39JfaGQAw8ZmTe+9dByhwQAAAB0EOe+Z/8UClelri45e9LYcodDK1FcBgAAAACdxPXX/Cn//KX3JEmOfMs1efhBxWUAAABA6zhx0tE5cdLR5Q6DVlZZ7gAAAAAAgN3vkSfOyJFvuSZvfss15Q4FAAAAgDbKyGUAAAAA0EkcfPCwHHTY9amrq8zzT5+dRx/+ULlDAgAAAKANU1wGAAAAAJ3EiZOOzjOTkmefeiEHvfHV39duqk3Xbi4VAgAAAK9dzeqarK9Zn+49u6dn756uNXQQpsUEAAAAgE5m3doNxeWKXktTWekyIQAAANAyZ5/ziwzea1D6Deibz33y5+UOh1aiRBAAAAAAOpnBQ/vnDRN+ntr6ihw2bn0qu1xU7pAAAACAdq6+vqK43KVLxTbWpD1RXAYAAAAAncyo8fvkyT+fV+4wAAAAgA5k7IjKDBjxp9TXV2afEX3KHQ6tpKJQKBTKHQS7TnV1dfr3759Vq1alqqqq3OEAAAAAUEa/uPK3ufDzI5NCRU444uHc+ZuLyx0SAAAAAGWwozVFRi4DAAAAgE5i5cp12bD84CTJ32bOzK9+9vvU1RXyrvcdl+49u5c5OgAAAADaGsVlAAAAANBJdO/eJem+OkmyYPrpOfNDDb+f85Z5GTV+nzJGBgAAAEBbVFnuAAAAAACA3ePiT70rhQ39cuMVjzT6/cYNm8oUEQAAAABtmeIyAAAAAOhkli9b2+jnPv16lykSAAAAoKP4yIeuyt773ZmR4+/Ir6+/v9zh0EpMiwkAAAAAnUxFZVLRa1kK9V1z9OG3Zfjo88sdEgAAANDOPflCt7w449Qkyby5t5Q5GlqLkcsAAAAAoJP5yL+8O/U1Q1JYPyAP/ElhGQAAANBy9XWvliF17aYkqaMwchkAAAAAdBJ/efjv+fy/P5ZCIXnbkX3y2a+dU+6QAAAAgA7i9l9OzEsL/5pNG2pz8BvfVu5waCWKywAAAACgk3jqz7Pz299dlCRZtOK6fPZrZQ4IAAAA6DBG7Lt3Ruy7d7nDoJUZgw4AAAAAOonKyori8t+eeH96D/1reu3xtzz6h6fKGBUAAAAAbZXiMgAAAADoJI494dD86z/+LCe87eokybqlr8/6Za9L9aqaMkcGAAAAQFukuAwAAAAAOokxB47KZf/7wew/qvHv6+vryxMQAAAA0GFc+8O78vlP/ixf+pdr89K8xeUOh1bStdwBAAAAAAC716baQnH5fWdOzslnXVzGaAAAAICO4Cv/tz6zpn0wSfKWt/wx79xnzzJHRGtQXAYAAAAAncxHPvymjBh2XTZsrMvJJx5a7nAAAACADqC+/tUJFLt3U5LUUTiSAAAAANBJrFm1Js/9bWYK9YV88IKjM+bAUdvfCAAAAGAHfPCUNXls5FWpq6/IvuPfVu5waCWKywAAAACgk7jjhgfy/o+8I0ly0GHX55mnFJcBAAAAreNr3/lAuUNgF6jc/ioAAAAAQEczY+4bcsapk3PqOyfnLw//vdzhAAAAANAGGbkMAAAAADqJvfcelL3H3Zma9X3z8oLjcvtd+ydJ3vPOO3LYUYeUOToAAAAA2hojlwEAAABAJ3HsKUdkwfRTM2H/mY1+v6m2rkwRAQAAANCWKS4DAAAAgE5mU11FcXn/Q2/I8SceWsZoAAAAgI5g+LjfpKLH6lT0XJW5L8wrdzi0EtNiAgAAAEAnM+HALpm96LbU1Vfma//UP/sdOrbcIQEAAADt3IZNPZON/ZIkXbquKXM0tBbFZQAAAADQyXzrBx/Kt8odBAAAANCh7NF/adYOmpZCKtOz15Byh0MrUVwGAAAAAJ3EA3f/Oadfsj6FQkWOPvT53Pmbi8sdEgAAANBBPPe3s8sdAruA4jIAAAAA6CSWL12dlQveliT5e9/leebx57JpU21G7zciA4YMKG9wAAAAALQ5leUOAAAAAADYPSq6vLo87/kzcsjhB+QNbzkk//ftu8oXFAAAAABtluIyAAAAAOgkTnnPsVn+0oq89ZgpjX5fW1soT0AAAAAAtGmmxQQAAACATqJrt64ZtOegvLSiX/F33Qc9lzGj+5cxKgAAAKAjOOntV2Xh8t7p2qU+f338/eUOh1aiuAwAAAAAOpm6+lcnNLjrqhU54YzTyhgNAAAA0BE88Lc3Zd3S1yddNpQ7FFqR4jIAAAAA6GRmTZuU2k21qVldk979jih3OAAAAEAHUCi88mW2irryBkKrUlwGAAAAAJ3EvBnz85Uv35NCIXnDIVX51BffU+6QAAAAgA7i1h+tzIoVv019XX2SU8odDq1EcRkAAAAAdBLP/nVOpvzioiTJnw66JZ/6YpkDAgAAADqMk886ptwhsAtUljsAAAAAAGD3mzXzbRky6g8ZtM/9+dZXflnucAAAAABogxSXAQAAAEAnceDrR+eDZ1+VoWPuSTYMyPJ5b8vL8ydm1tyacocGAAAAQBtkWkwAAAAA6CT2GTcy1/zyogwc+adGv6+rL5QpIgAAAKCjuPp/78jGjXWp6t8z5374HeUOh1aiuAwAAAAAOpm6ulcvC86dPj9Dh7+/jNEAAAAAHcElX3xD6teMSGXfBTn3w+WOhtaiuAwAAAAAOpnPXTAnc+Y/m42bkhFjLkhll8pyhwQAAAC0d4UuSZKKiroyB0JrUlwGAAAAAJ1EfV19atbU5B//v1PTtVvX9O7Xu9whAQAAAB3EcYf/NtU13VLVe1OSC8sdDq1EcRkAAAAAdBIP3P14jjv1iCTJqP1vy5znzihzRAAAAEBH8Yf7FJR1RIrLAAAAAKATWr5qWD52/tWpqy/kjFP3zWnvm1jukAAAAABoYxSXAQAAAEAnMWBw3/Tf+8Fs3Ng3NWuG54prGkYxW1E9WXEZAAAAACUUlwEAAABAJ/G6Nx+UlQuSrv3npH7NiOLv6+vLGBQAAAAAbVZluQMAAAAAAHavQv2r3zk996zJOe+sfcoYDQAAANARdB84Pd0GzMqwsb8tdyi0IiOXAQAAAEAnM3LvJ7K6Zkb69KrOdTddXO5wAAAAgA5g06oxSaFrqnusLHcotCLFZQAAAADQycx57oxyhwAAAAB0MJW9l6ZQ6JKePVeVOxRakeIyAAAAAOgkZj4zO8ef+fcUCslh45bkjruMWgYAAAC0jro1w15ZGlrWOGhdissAAAAAoJN4eUV15r9wepKkvnBn6uvqU7upNknSvWf3coYGAAAAQBtUWe4AAAAAAIDdo7Li1cuBL86dmC5dK9OjV/ec/M6fly8oAAAAANosxWUAAAAA0Ekc9Mb98vtbH0qPwc8km/oUf19XV1HGqAAAAABoq0yLCQAAAACdRM/ePbP/oaOyYfnexd/1G/ZIRg+rL2NUAAAAQHu3ctnKnDbptnSpLOTgsYX84KoLyx0SrURxGQAAAAB0IqtXrS0uDx3zuyyedVKSI8sXEAAAANDuLV/8ch584PwkybPzfp8flDkeWo/iMgAAAADoRPY7ZN/M+PvsrK5emx49R5c7HAAAAKAD2LhxU3G5osII6R2J4jIAAAAA6CSqV1TnW/95e+oLhYwbMyAXfvL0cocEAAAAdAAj9x2eb37pl9lUW59RI/uXOxxaUUWhUCiUOwh2nerq6vTv3z+rVq1KVVVVucMBAAAAoIyeefy5HHL4AUmSvfb9bRbNfEeZIwIAAACgHHa0psjIZQAAAADQCb28amTGHHhrCoWKvH3C8vzk2ovKHRIAAAAAbYziMgAAAADoJPbce0jeMOHneXbO61JX2zNznnt3kuSxPr8ob2AAAAAAtEmKywAAAACgkxgybEguPLNnPvmFQxv9vr6+okwRAQAAAB3BmlVr8se7n0i3bl0zfOTgHDzhgHKHRCtRXAYAAAAAncjamk3F5ZHjb881/29o9t7n8DJGBAAAALR3f7rnyZx6znFJkn0PuiUzn1Fc1lEoLgMAAACATuSYY8dl0umTs35jRU6Z2D8TTz2y3CEBAAAA7dymTbXF5S6V9WWMhNamuAwAAAAAOpGjT5yQo0+cUO4wAAAAgA5k2PBBGXvwzamvr8yRB1WXOxxakeIyAAAAAOgk5r4wL6NfV5WkInuPfCALpp9a7pAAAACADuDw4w7LjL8fVu4w2AUUlwEAAABAJ1FfV0g2DEiSrFk3IP/1hV+ktrY+4/cbnHM//I7yBgcAAABAm6O4DAAAAAA6iW49uqb7wOeTikIqKzfl3y47N0ky6oBbc+6HyxwcAAAAAG1OZbkDAAAAAAB2jxH77p3Tj30gG18el5fnTyz+vlCoKF9QAAAAALRZissAAAAAoBNZvzFJoWFCgx6Dn8nxE6/OOW9fVd6gAAAAgHbtF1f+Nn32fDJ9hz2WsydNLnc4tCLTYgIAAABAJzJscEV6DX0q9XXd8l//OC2f/sqF5Q4JAAAAaOcWL16TmiXvSJLMW/xcmaOhNSkuAwAAAIBO5MqfXZQriz8dUsZIAAAAgI6ivr6QVNQmha7pUlkodzi0oopCoeCIdmDV1dXp379/Vq1alaqqqnKHAwAAAEAZVa+ozltPvDOFQkXGj1ybm2+7uNwhAQAAAB1IfV196uvr07Wb8a7auh2tKXIkAQAAAKCTWF+zPk8/eW6SZOHSqWWOBgAAAOhoKrtUprJLZbnDoBU5mgAAAADQSWx5cXfDxn6p7LMklb2WZ8T4X5cxKgAAAADaKiOXAQAAAEAnMWDIgBz4+l9m9sIDkySFmqFJkvUbepczLAAAAADaKMVlAAAAANBJdO3WNes29Mj6Za8r/q7bgOkZ2O/lMkYFAAAAtHf33vZQvj/52XSpTN5z2oicc8nJ5Q6JVqK4DAAAAAA6kU11r14SfOieJ3LU29+UZL/yBQQAAAC0e3f+dnp+dcfFSZL1GyfnnEvKHBCtRnEZAAAAAHQid/1s38ybfX/WrtmQQw8/utzhAAAAAB1AbW2huNy1SxkDodUpLgMAAACATqJ2U20ee2h6kmSPPfqlb/++ZY4IAAAA6AjOOfuQFApTUltbyKnvGFvucGhFFYVCobD91Wivqqur079//6xatSpVVVXlDgcAAACAMlqzak36DWgoKOu/94NZucDIZQAAAACd0Y7WFBm5DAAAAAA6oUIhmfDma1NbV5mRQ9fljrsuLndIAAAAALQxissAAAAAoJPo3qN7+uz5eOrqumdgv2V54s/nJoUumbHn4+UODQAAAIA2SHEZAAAAAHQS3Xt2T21tz2xYfkjmrhqbVNQnhS6pL3Qpd2gAAABAO7Zy2cosX/xyuvXomiF7Dk7vfr3LHRKtRHEZAAAAAHQi9XU9kiQVXTbkx9/8fSoqKzJkcN8yRwUAAAC0Z5dccnNuvu3iJMnHL7w6P7jqwjJHRGtRXAYAAAAAnchpx/wxLy1/KN27FXLJP19Q7nAAAACADmDl6sri8uAB3coYCa1NcRkAAAAAdCK33H5xuUMAAAAAOpiD9q3IM3N+m3Ub+uaA8YPKHQ6tqKJQKBTKHQS7TnV1dfr3759Vq1alqqqq3OEAAAAAUEb1dfXpOWRmCoWK7Dn0uSx44bRyhwQAAABAGexoTZGRywAAAACgE9m0cr8kyapeK3LTlHtSs3ZjevbqmrMvOrnMkQEAAADQ1iguAwAAAIBOorJLZSp6Lk8qCunRY3XOvvSwFNbtka79Z+bsi8odHQAAAABtjeIyAAAAAOhE6tcNfmXphFT2WdywWKgsWzwAAAAAtF2KywAAAACgk3rzYXdn7fquGdRvY5Ix5Q4HAAAAaKdGH3hr1qzrlwF9V2bG388qdzi0IsVlAAAAANBJPfzgh8odAgAAANABzF8wIfVrRuTlvgvLHQqtzHj3AAAAAAAAAADAa1a/sSpJ0rV7dZkjobUZuQwAAAAAAAAAAHjNNqzqmYWz52bVyrpyh0IrU1wGAAAAAAAAAAC8Zt17ds+YA0eVOwx2AdNiAgAAAEAnNXTMPekx+Ln03evxcocCAAAAQBtk5DIAAAAA6KRWVY/MxhUHZGOPleUOBQAAAIA2SHEZAAAAAHRS3bqtycYeK9Ole3WSAeUOBwAAAGiHHv3DU/mfHzyZwQO65J0njckpZx9X7pBoRYrLAAAAAKCTWvPShFeWBpQzDAAAAKAd+/Wvn8kvbr44SfLs7KsVl3UwleUOAAAAAAAAAAAAaJ+WrdhUXO7ft76MkbArGLkMAAAAAAAAAAB4Td571gFZv/GqvFxdyNmnjyh3OLSyikKhUCh3EOw61dXV6d+/f1atWpWqqqpyhwMAAAAAAAAAAJTZjtYUGbkMAAAAADqplctW5mMfuTnT5vbKkKoN+cN9F5Y7JAAAAADakMpyBwAAAAAAlEfP3j3zy9vPy9NPvj8P/u2oRm0//b9fZ/i4uzJ83G/y6Y//tGTb0QfemuHjfpPXvem6krYPnD05w8b+NsPG/jb33vZQo7a7b/5Tse2Cc68q2fbgN1yfYWN/m30P+lVJ26c+MqW47bU/vKtR298enVZsO+PUySXbHvmWa4rttZtqG7V97XPXFdv+39dubNS2YvGKYttxx00p6ffkE68qts99YV6jtiu+fWux7fOf/FnJtiPG/zrDxv42bzj85yVt7333q/vwwXseb9R227V/KLZ9+IOl+3D8oTdm2NjfZr9Dbi5p+/gFVxe3vWnKPY3aHv3DU8W2s84o3YdvOuLaDBv72wwfd1dJ2xf/+WfFbf/vG40fd8Gshdlr7N3Za+zdOeFtV5dsO3HilGL7S/MWN2r77/+8qdj25U9fW7LtsLG/zV5j786EI0v377tPm1zc9skHnm7U9suf3F1su/Si0pj2PehX2Wvs3TngdTeUtF1w7lXFbW//xR8atd1/5yPFtnPPKt2Hr3vTL7Lnvr/LyPF3lLR9+uM/zZ77/i577vu7/OS7tzVqm/70zGLbySeWHvOj3/rqttUrqhu1feNL1xfb/usLv2jUtr5mfbHtyLdcU9LvO0++qtj+7FMvNGq7+n/vKLZ96iNTSrbdZ//bs+e+v8vBb7i+pO3975lc3HbrHPGbm/6YPfe9J3vue08+cHbpPjzw9Tdkz33vyegDbitp+8dLphS3veb7dzZq+9uj04ptp76ztN/Dj/xZsX3rHPHVz/682Patr/yyUduyRcuKbW89pjRXvv2Eq4vts5+d26jtB9+8pdj2uX8s3f/Dx/0mQ8fck9e96RclbWedMTlDx9yToWPuyQN3/7lR269+9vti20Xnlb5exh1yc4aOuSdjDy7Nsx/+4FXFbW+46u5GbY/+4ali26TTS/fhYROuy9Ax92SvsXeXtH3+kz8rbvu9yxrniHkz5hfbjj++9P14zLFTMnTM7zN0zO9LcsT/+9qNxbYv/Utpjmjo9/d5YxN59rRTJhe3ffxPf23U9vMf3VVs+9j5pTGNPuC2DB3z+4w/9KaStg+dc1Vx261zxB/ueLjY9t53l+7Dg99wfYaO+X323u/OkrZ//uhPs8foe7PH6HtzxbdvbdT2/F9nFNtOfHtpvEe+5Zpi+9Y54r++8Iti29c+1/hv+vqa9cW2w5vIsyefeFWx/ZnHn2vU9pPv3lZs++SHp5RsO2K/O7PH6Htz4OtL8+z7zpxc3PaeXz3YqO2uG6ZmyOh7M2T0vXn/e0r34fhDb8qQ0fc2mWcvvejq4rY//b9fN2r7y8N/L7a98+TS982bjri22L51jvjyp19tu/zLpTlic9tRR5e+z48//tWYts4R//eNm4ttn7m0dNs99/1dhoz6Q5N59t2nTc6QUX/IkFF/KMkRN//0nmLb+e8rfa5jDrw1Q0b9ock8e9F5VxW3/eVPGr/XH/79E8W2008pPTaHvvH6DBn1hwwd8/uSts/94zXFbf/nvxq/r+bNmF9sO/bYKSXbHv3Wn2bwqPsyeNR9JTniW1/5ZbHtC/9U+hoeMuoPGTzqvibPZ095x1XFbbfOEdf+8K5iW1PnYiPH35HBo+5rMs+e997JxW1vu7Y0R2xua+pc7IDX3ZDBo+7LsLG/LWn71EemFLf94bcaP+7zf52Rwfvcn8H73J+3NZFnDz/yZ8X2rXPEf/7rdcW2r362cS5dX7O+2PamI0pz8NtPuLrYvnWO+PH/u7XY9g9NnIsNG/vbDN7n/ibz7HvePbm47d03/6lR26+vv7/Y9r4zS/fh2INvyaB97m/yfPZj51+dQfvcn0H73J+r/7dxDvnLw38vtp309tJjftiE64rtW+eIL/7zz4ptl32x8ft12aJlxbYjjip9jR577JRi+9Y54nuX3Vxs+5cmPq8NHfP7DNrn/hz4+l+WtJ1+yuQMGjk1g0ZOzR/veqxR201T7im2NXUuNuqA2zJo5NQm8+wF515V3PYXVzZ+nT54z+PFtqbOxQ467JcZNHJqBo+6r6TtM5deU9x2689r82bML7Yd/dbS/XDkW17ddusccfmXf5mBI/+YgSP/2OS52KCRUzNw5B9zSBN59uQTrypu++epf2nUds337yy2NXUutvd+De2jD7y1pO3csyYXt/3Vzxrny3tve6jY9u7Tmv4bOHDkHzN0zD0lbf94yZTitj/45i2N2p596oUMHPmnDBz5pyY/877piGuL7VvniK9+9ucZOPJP6Tf80SyYtbBkWzoOI5cBAAAAQCfVs3fP9Bz4dNYvOzTf/v+eSXJAsW3GrJVZNPMDSZKnR5deEJ87e2KyYUCWD3qupO1vM3vnpVnvSJIserHxjZsFC17OS7PelST566DSm/3T57wpm1bul4pey0ranp5eKPY7a07jm4ArllUX2/7a/daSbafNGZ/Vi4585afGN3yenbGuuO1zMxo/1/XrNhTbNm66v6Tfv80aXmxfvfLZRm0vzFyVl2a9O0ny9xmlF/8Xzj4hqe2VldV/LWl7ela/Yr9LXmp8U2He/Jfz0qyzGh5/z9KbILPmvjl1q/dJZd/Si/t/m1FZ7Hfu3MY39JcuWVVse7p36Y28Z+cenHVL3pBUbippmzZjY3Hb6bOmNGpbu3pdFs86OUlSKPyuZNtn5ozMsrknJEk2rG9coPfCrNVZPOs9DY8/s3QfvjTrpCSVWbP2zyVtT88eVHzc5UsbFyfMnvtyFs96X5Lk73tP2XrTzJn7lhRq9szy/rNK2v4+q3ux34ULGt+8XbxoVbHtb1WlN6FemPu6bFh+cNJ9dUnbtFl1WTL7pCTJzNmNby6uWrmm2PZMl9Kik2lzx2TlgmOTJLWbVjRqe35mTZbMft8ry41f3/V19cV+129ovI+S5Jk5exTbVy5/olHbrDmrsmT2B16JvfSm8Py5xyYbBmRlEzni77N7F/vdOke8uHBllsxuyBHPDC7NETPnvaHZHPHMzEKWzD4xSTJnXmmO2Nz2TI9bS7Z9ft5+zeaI52auL247fXbjfbhxw6ZiW23t/aUxzRlWbF+7unGOmDlndZbMPjNJMq2J1/eiuROT2l5ZvaY0Rzwzu1+Wzmnod+mSxjliwYJVWTqnIUc8M6w0R8yZd3jqVu+T5U3kiGmzuxT7nT+/NEdsbnumb2mOeGH+gc3miOdmbyxuO2P2lEZt69ZuKLZNqyjNEc/Naz5HzJizJkvnvLdhvVml+3DpnBOSVKZmXWmOeHbuoCyd8/YkycvLGr/+5y+sztI5De+baSOnbL1p5s0/MoWaPfNyEzli2pzuxX63zhFLF1dn6ZyGHDFtQGmOmDn/0GZzxHNz6or7Yc68xjlidfXaYtuz3UpzxPPzRzebI2bMWVfcdsac0hyxuW3jptIc8ey8PYrt1Ssb54i581dn2dwzGtabXZojFs4/OtkwINVN5Ihpc3sX+138UuMcsWjRqiyf25Ajnt2jNEfMWfD6ZnPEc7OT5a/0O3d+4xyxcsWaYtuzvW4t2Xb6gnHN5ojpczYUt916H27csKnYVqi/vzSmecOK7VvniNlz12T53Ib38nNzSl/fS+Yfk9T2ytqa0hzx3Lx+WT7vbUlKc8SLL1Zn+bxX+t27NEfMX/jG1K3eJy83kSOen9ul2O+ChY1zxPJlq4ttz1eV5oiZC/dvNke8MGdTcdtZc6c0alu3dsOr/XYpzREvLBiRFfOOT1KaI2bPrym2TW9iHy6fNzFJZTZsKM0Rz88fWNx26xyx4MXqrJh3SsN6c+eWbLtw4REp1OyZ6iZyxPPzehT7fXFRaY5YMa8hRzw3aHHJtnNePLjZHPHC3Ppiv/Pml+aIFfMnNqzXY23JtjMWjmo2R8yct7647cy5pTlic1ttXWmOeGHBkGL71jli/oI1WTH/3Q3rzZtdsu1LC49MNgzImrWlOeKF+b2L/S5d0jhHLF5cnRXzG3LE83uWvobnv3hosznihXnJy6/0O39haY7Y3PZCn5Ul2856cd9mc8SseRuL286eX5ojNrfNzP0l/U5fuFexfescMWfemrw8v+G9PH1uE38DFxyd1PbKunWlOWL6gn55ecFxSZLlyxvniEWLqvPyglf6XdDE6/vFN6Ru9T5Z1USOmD6/S7HfF19snCNeXr4mLy9oeC7TBywt2XbOovHN5ogZ8zcV+527YEqjtnVrNxTbZnTbULLtzBf3LrZvnSPmLKgpvvZnzZ9esu3LC45JUplNm0pzxIwXBxa3Xbmi8ev/xZdWZ+WCU1+JvTQPLHpxQgo1e2ZNEzli+oIexX5feqlxjli2dHVWLmjIEdOHvFiy7bxFBzabI2bMry/2O39h4xyxds26rFxwTJJkZs/qkm1nLdqn2L51jpizYEOx7ctfuipX/fyiku3pGIxcBgAAAACd2JkTH0sqN2XC4aPLHQoAAADQDt335KByh8AuVFEoFArlDoJdp7q6Ov3798+qVatSVVVV7nAAAAAAaIPmzZifocP3SM/ePYu/W7lsZeZMX5Ak2WvvPbLXPns22ubvjz2b+kIh3bp3zYFvGN+obcGshVmxdFWSZNxBo9O7X+9iW83qmsx6ZUqXQUMHZPjoYY22ff6vM7Jpw6ZUVFbk4AkHNGp7ad7iLHup4ZvS+4zbO1WDXr3etb5mfWb8vWHUg6pBfbPPuJGNtp35zOysW7s+SXLQm/ZPZZdXv3e7bNGyvDS/4dvzw0ftmUF7vnpRvHZTbZ57quFb7H369c6YA0c16nfuC/OyemXDCBDjXzc23Xt2b7QPF8xalCQZOnxwho4Y2mjbzdMD9ezVI2MPHtOobcGshVm1ouEb52MPGt3o2KxZtSZzXzk2g4cOLDk2z/91Rmo31aZrt67Z//XjGrW9NG9xli95OUkyar8R6du/b7Ftfc36zJw2J0nSf1C/jNh370bbznxmdtavaxgNYOtjs2zRsixe2DACxN6j98qAIQOKbbWbavP8X2ckSfpW9c6o8fs02nbuC/NS88qx2e+QfdO126sTbqxctjKL5i9Jkuw5fEijY5OkOFVjz549So7NglkLs3pVw7EZs/8+Jftw/qyGb/sP3mNAybF5/q8zUl9fn65du2S/Q8c2antp3uK8vLzh9T1q3IhGr+/1Nesz+/mGERH6D+xX8vqe+czsbNy4KZWVlSXHZtmiZVm+ZGWSZNjIoY1e3xvXbyz226dfr5JjM2/G/Kxb23Bs9jtk30av75XLVhaPzZ57D2l0bOrr6jP97w0jJvTs1b3k2CyYtTBrV69L0rAPt3x9V6+oLh6bwUMHZMiwIY22nf70zNTXF9K9e7eSY/PSvMVZ9XLD63vUfiMaHZua1TXFY9N/YL+S1/fsZ+emtrY2lZWVJe+bZYuW5eVlDcdm2D57Nnp9b1y/MXOnz0/S8F7e+tjMmzE/G9ZtTJKMPWhMyT5cumh5kmTPvfdodGzq6+ozc1pD7unVp2fJsVkwa2Ex94w5YFSj13f1iuosXrj0lX04sOT1Pf3pmUmS7j26lRybJQuWFF/fe48ZVrIPF70y7VP/Qf1Kjs3cF+aldlNdKisrS47NskXLirln2D57Nnp9b1y/MfNnNoxS0qdf75JjM2/G/Gza0DBay9bHZuWylVm+uCH37DFscKN9WLupNnNfaDg2vfr0LDk2L85ZVNyHo8aPLNmHm/sduEf/Rq/vJMUpxLr16FpybJYsWFJ8fe89Zlij13fN6posXtBwbPoP6ldybOa+MC/1dYV06VZZ8vdm2aJlxb8Lw0bt2ejYrK9Zn0VzG45Nn369SnLPvBnzU7epPklKjs3KZSvz8it/W/cYPrjR67t2U20Wzm543/Tq3bOk35fmLc6G9Q05YuTYEY1e39UrqvPyspVJkoFDBpS8vufPbPh70617t5Jjs2TBkqyraTg2e48Z3ujY1KyuydJFDbmn/6CqkmMzb8b8FOoL6dK1S8mxWbZoWdaurkmS7DliaMk+XLygIff069+35NgsmLUw9fX1qaysLOl35bKVqV7Z8PoesufgRq/v2k21eXFuw9/s3n16lbxvXpq3OBs3NuSIEWP2LtmHK1c0HJtBewxsdGzq6+qzYHbD+6Znzx4lx2bJgiVZ/8qxGT5qWMk+XLa4IfcMGNS/0bHZvA+TpGvXriXHZsXiFalZ2/D6Hjp8j0av7/U167PspYZ++1b1KTk2L85ZVNyHW/e7ctnKrKlueH0P2Wtwo2Ozcf3GLHmx4X3Tu0+vkmPz0rzFqa1tyBFbH5vqFdXFYzNoj4Elx+al+Q3vm569epQcmyULlmTjxoaRjoaPGtbo2NSsrsnKV/5mVw3s1+jYbH6uSdK1a5eSY7Ni8Yriec/QvfdodGzW16zPilfOp/pW9Sk5Ni/NW1zch1vnypXLVqZmTcOxGbLX4EbHZuP6jcVj07tvr5Jjs2TBktTW1jU8162OTfWK6uKxGTC4f8k+XPLK35uevXqUHJsVi1dk44ZNxee69T6sfuVvdtXAfo36ra+rL/bbtVuXkmOzctnKrK9p2IdDhg0u2YcrX/mb3bd/n5Jjs3n6wsrKipJjU72iurgPBw0dWLIPNx+b3n17lRybJQuWpL6+oTxi62OzZtWa1LySe6oGVTV6fdduqs2KxQ2fQ7r37F5ybFYsXlGcDnPIsCEl+3DNqjWvPNe+Jftw2Su5smu3riXHZuWyldm4fmOT/a6vWV+conDrfjc/1+b6rV5RnfWv5O9Bew5qdGw2rt+Y6pcb+u3Zu2fJsVmxeEXx9b11v2tWrSn2WzWwqtGxqd1Um5Wv/L3p3qN7ybHZ3O/m57qlmtU1qVlTU3yuWx6b+rr6rFiyotl+Vy5bWTw2g4YOKtmHm49N7769S/bh5mNe2aWy5JivWbWmeGyqBlWV7MPN/fbs3XOn+q1ZXVPch3379y3Zh5uPefee3UuOzcplK1Nf17APtz422+q3vq6+eGy6dutasg+rV1QXj03VwKqSfbi5362fC+3DjtYUKS7r4BSXAQAAAAAAAAAAW9rRmiLTYgIAAAAAAAAAAFBCcRkAAAAAAAAAAAAlFJcBAAAAAAAAAABQQnEZAAAAAAAAAAAAJRSXAQAAAAAAAAAAUEJxGQAAAAAAAAAAACUUlwEAAAAAAAAAAFBCcRkAAAAAAAAAAAAlFJcBAAAAAAAAAABQQnEZAAAAAAAAAAAAJRSXAQAAAAAAAAAAUEJxGQAAAAAAAAAAACUUlwEAAAAAAAAAAFBCcVkz1qxZkz/+8Y/59re/nbPPPjtjxoxJRUVFKioqMnr06F3ymA899FA+8IEPZNSoUenZs2f22muvnHzyyfnFL36xSx4PAAAAAAAAAACgORWFQqFQ7iDaouOPPz73339/k22jRo3KnDlzWvXx/v3f/z3/8R//kfr6+ibbTz311Nx0003p2bPnTvVbXV2d/v37Z9WqVamqqmqNUAEAAAAAAAAAgHZsR2uKjFzWjC1r7gYNGpSTTjopffv23SWPdcUVV+SrX/1q6uvrM3bs2EyePDmPPfZYbr311hx//PFJkjvvvDMXXXTRLnl8AAAAAAAAAACArRm5rBk//vGP069fvxx++OEZN25ckmT06NGZO3duq45ctmLFiuy7775ZtWpV9tlnnzzxxBMZMmRIsb2uri6TJk3KHXfckSS57777MnHixB3u38hlAAAAAAAAAADAloxc1kIf+chHcu655xYLy3aVn/zkJ1m1alWS5Jvf/GajwrIk6dKlS37wgx+kS5cuSZJvfetbuzQeAAAAAAAAAACARHFZ2d16661Jkqqqqpx55plNrjNixIi8/e1vT5Lce++9Wb169e4KDwAAAAAAAAAA6KQUl5XRxo0b89hjjyVJjjrqqHTv3r3ZdY877rgkyYYNG/L444/vlvgAAAAAAAAAAIDOS3FZGb3wwgupq6tLkhxwwAHbXHfL9meffXaXxgUAAAAAAAAAANC13AF0ZgsWLCgujxgxYpvrjhw5srg8f/78ZtfbsGFDNmzYUPy5urq6BRECAAAAAAAAAACdlZHLymj16tXF5b59+25z3T59+hSX16xZ0+x6l112Wfr371/8t2VRGgAAAAAAAAAAwI5SXFZG69evLy537959m+v26NGjuLxu3bpm1/v85z+fVatWFf9ta5QzAAAAAAAAAACA5rTr4rKKiooW/5syZUrZ4u/Zs2dxeePGjdtcd8upLnv16tXsej169EhVVVWjfwAAAAAAAAAAADurXReXtXf9+vUrLm9rqsskWbt2bXF5e1NoAgAAAAAAAAAAtFTXcgfQEs8++2yL+xg2bFgrRPLajBgxori8YMGCba675fSWI0eO3GUxAQAAAAAAAAAAJO28uOyAAw4odwgtMn78+HTp0iV1dXV57rnntrnulu0HHnjgrg4NAAAAAAAAAADo5EyLWUbdu3fPEUcckSR5+OGHs3HjxmbXnTp1apKkR48emTBhwm6JDwAAAAAAAAAA6LwUl5XZu9/97iRJdXV1brnllibXWbBgQX7/+98nSU444YT069dvd4UHAAAAAAAAAAB0UorLdqE5c+akoqIiFRUVmThxYpPrXHLJJenfv3+S5F//9V+zfPnyRu11dXW59NJLU1dXlyT5zGc+s0tjBgAAAAAAAAAASJKu5Q6grZoxY0YeeOCBRr9bs2ZN8f8pU6Y0anvHO96Rvfbaa6cfZ9CgQfnmN7+Zj33sY5k7d27e/OY354tf/GIOPfTQvPjii/nv//7v3HfffUmSc889t9kiteYUCoUkDSOjAQAAAAAAAAAAbK4l2lxb1BzFZc144IEHcuGFFzbZtnz58pK2++677zUVlyXJRz/60bz44ov5j//4j8ycOTMXXXRRyTqnnHJKrrrqqp3ue/Xq1UmSkSNHvqbYAAAAAAAAAACAjmn16tXFWReborisjfjqV7+ak08+Od///vfzpz/9KYsXL86AAQPy+te/PhdeeGHOPffc19Tv8OHDM3/+/PTr1y8VFRWtHDWtobq6OiNHjsz8+fNTVVVV7nCANkaOAJojPwDbIkcAzZEfgG2RI4DmyA/AtsgRQHPkh7atUChk9erVGT58+DbXqyhsb2wzYJeqrq5O//79s2rVKskUKCFHAM2RH4BtkSOA5sgPwLbIEUBz5AdgW+QIoDnyQ8dQWe4AAAAAAAAAAAAAaHsUlwEAAAAAAAAAAFBCcRmUWY8ePfKVr3wlPXr0KHcoQBskRwDNkR+AbZEjgObID8C2yBFAc+QHYFvkCKA58kPHUFEoFArlDgIAAAAAAAAAAIC2xchlAAAAAAAAAAAAlFBcBgAAAAAAAAAAQAnFZQAAAAAAAAAAAJRQXAYAAAAAAAAAAEAJxWVQJnPnzs2nP/3pHHDAAenTp08GDRqUww8/PN/61rdSU1NT7vCAnfD444/na1/7Wk466aSMGDEiPXr0SN++fTN+/PhceOGFeeCBB7bbx5QpU1JRUbFD/6ZMmbLd/mpqanL55Zfn8MMPz6BBg9KnT58ccMAB+fSnP525c+e2wrMGdtSOvrcnTpy43b5+85vfZNKkScVcM2LEiEyaNCm/+c1vdjie2tra/OhHP8oxxxyTPfbYI7169crYsWPz0Y9+NM8880wLnimwsyZOnLjDOWLzv/vvv79RH84hoH1asmRJfv3rX+fLX/5y3vnOd2bIkCHF9+oFF1yw0/21pXOEZcuW5ctf/nJe97rXpaqqKlVVVXnd616XL3/5y1m+fPlOPzfobFojP9TU1OSWW27Jxz/+8Rx++OEZOHBgunXrlsGDB+eoo47Kv//7v+ell17abj87c66yI/7+97/nox/9aMaOHZtevXpljz32yDHHHJMf/ehHqa2t3aE+oLNrjRzRVj9DuGcCLdPS/DBnzpydvkYxevToJvtyDgFtT2vcy9yS6xCdUAHY7W6//fZCVVVVIUmT/8aPH1+YPn16ucMEdsAxxxzT7Ht5y38f+tCHChs2bGi2n6uvvnqH+klSuPrqq7cZ0/Tp0wv77bdfs9tXVVUV7rjjjlbeE0BzdvS9fdxxxzXbR11dXeHiiy/e5vaXXHJJoa6ubpuxLF26tHD44Yc320ePHj0KV155ZSvvAaA5xx133A7niCSFysrKwoIFCxr14RwC2qdtvVfPP//8He6nrZ0jPPLII4W99tqr2X6GDRtWePTRR3f4+UFn1NL88Ne//rXQt2/f7Z4XVFVVFa6//vpt9rUz5yrb8+Mf/7jQvXv3Zrc/4ogjCkuXLt3R3QSdVmucQ7TFzxDumUDLtTQ/zJ49e6euUSQpnHTSSU325RwC2pbWupdZKLgO0Zl1DbBbPfXUUznnnHOybt269O3bN5///Odz/PHHZ926dbn++utz5ZVX5oUXXsipp56axx9/PP369St3yMA2vPjii0mS4cOH573vfW+OOeaY7LPPPqmrq8vDDz+c73znO1m4cGGuueaabNq0Kdddd912+7z77rszfPjwZttHjBjRbNvq1atz6qmnZvr06UmSD3/4w3nf+96XXr165b777stll12W6urqnHPOOXnwwQdz2GGH7dwTBl6zj3/847n00kubbe/Tp0+zbV/84hczefLkJMkb3vCGfPazn83YsWMzc+bMXH755Xnqqafyk5/8JHvssUe+/vWvN9lHXV1dJk2alD//+c9JkjPPPDMf/vCHM2jQoDz66KP5z//8zyxZsiQf/ehHs/fee+ed73xnC54tsCOuvvrqrF27dpvrTJs2Leecc06S5IQTTsjee+/d7LrOIaB92meffXLAAQfkd7/73U5v25bOEebPn5/TTz89S5cuTdeuXfMv//IvOe2005Ikv/71r/P//t//y6JFi3L66afniSee2GZOAhq8lvxQXV2dNWvWJEmOPvronHbaaZkwYUIGDx6cpUuX5pZbbsmVV16Z6urqnHfeeamqqtruuf+ECRNy9dVXv+bncdddd+VjH/tY6uvrs+eee+aLX/xi3vzmN2fFihW58sorc8stt+Sxxx7LpEmTcv/996dLly6v+bGgM2nJOcRmbeEzhHsm0PpeS37Ye++98/TTT293vcsuu6x4n+P888/f5rrOIaBtaM17ma5DdGLlrm6DzmZzZXDXrl0LDz30UEn75ZdfXqyk/cpXvrL7AwR2yqmnnlr45S9/WaitrW2yfenSpYXx48cX39dTp05tcr0tvzE4e/bs1xzPl770pWI/l19+eUn7gw8+WOjatWsh2fYoSUDraenf9eeff774vp0wYUKhpqamUfvatWsLEyZMKJ5fNPdN3smTJxdjufTSS0vap0+fXvyW8Lhx4wqbNm16TfECreuzn/1s8b37s5/9rKTdOQS0T1/+8pcLd9xxR+Gll14qFAqNRwnY0VFH2to5wgc/+MFiPzfccENJ+y9/+cudfo7QGbU0Pzz44IOFs88+u/DMM880u86tt95aqKioKCQpjB07tlBfX9/keptHHWnJ3/6NGzcW9t133+IIRjNmzChZ59JLL93hUZKgs2uNc4i29hnCPRNoHa2RH7antra2MHz48EKSQr9+/Uo+g2zmHALalta6l+k6ROemuAx2o0cffbSYwD760Y82uU5dXV3hwAMPLCQpDBgwoLBx48bdHCXQ2u64447ie/8f//Efm1ynNS7qbNy4sdC/f/9CksKBBx7Y7JCzH/3oR4uP9dhjj72mxwJ2XEsvgH784x8v9vHwww83uc7DDz+8zQ9jhUKheH4xaNCgwtq1a5tc57LLLtvmhzFg96qrqyvsvffehSSFvn37NvnedQ4BHcNrufHTls4RFi1aVKisrCwkKZx88snNxnzyyScXkoZpfhctWrQDzxLYFTeGC4VC4ayzzir2+8QTTzS5TmvcGN7yhs5ll13W5Dpr164tDBw4sJCkcNBBB73mx4LOqFzFZa31GcI9E9h1dsU5xG9/+9tinxdeeGGz6zmHgPZnR+5lug7RuVUG2G1uvfXW4vKFF17Y5DqVlZX50Ic+lCRZuXJl7rvvvt0RGrALHX/88cXlmTNn7rLHue+++7Jq1aokDcNRV1Y2/Wf+ggsuKC7/6le/2mXxAC1XKBRy2223JUkOOOCAHHnkkU2ud+SRR2b//fdPktx2220pFAqN2l944YU8++yzSZKzzz47vXv3brIf+QHalnvvvTcLFy5MkrznPe9p9r3bUs4hoP1pa+cIt99+e+rr65M0f71jy37q6+tz++23N7sesOvtrmsVW14P3TKXbKl37945++yzkzRMCf7CCy/ssniA1tFanyHcM4H25Zprrikub29KzJZyDgG71/Y+H7gOgeIy2I0eeOCBJEmfPn3ypje9qdn1jjvuuOLygw8+uMvjAnatDRs2FJe7dOmyyx5nc45JGueRrU2YMKF4sibHQNs2e/bsvPjii0m2/b7esn3hwoWZM2dOo7YdzQ977bVXxo8fn0R+gLZgy4u2m2+m7ArOIaD9aWvnCDvaj+sd0Hbs7msV+++/f/baa69m15MfoH1prc8Q7plA+7F69epiwdfo0aNz7LHH7tLHcw4Bu9f2Ph+4DoHiMtiNNlfhjhs3Ll27dm12vQMOOKBkG6D9mjp1anH5wAMP3O76F154YYYPH57u3btnyJAhOfLII/Nv//ZvxZFLmjNt2rTi8pZ5ZGtdu3bNuHHjksgxsDvdeOONOeigg9K7d+/069cv++23X84///xtfuN2R9/XW7dv/d5+Lf3Mnz8/a9eu3ea6wK6zZs2a4rfyRo0alYkTJ253G+cQ0Hm0tXOEzf30799/mzd+hg0blqqqqiZjAXavnblW8dxzz+XNb35zBgwYkJ49e2bEiBE544wzcs0112TTpk3NbrdmzZrMnz8/SctyFbDrlPszhHsm0H7cdNNNqampSZJ88IMfTEVFxXa3cQ4B7cf2Ph+4DoHiMthN1q9fn2XLliVJRowYsc11Bw4cmD59+iRJ8eQJaJ/q6+vzjW98o/jz5iGat+X+++/PokWLsmnTpixfvjyPPvpo/uu//ivjxo3LFVdc0ex2CxYsSNLwTb8BAwZs8zFGjhyZJFm6dGmjbyMAu860adPy7LPPZt26dVmzZk1mzJiRa665Jm9729syadKk4nQSW9r8vk62f/6w+X2dlJ4/vJZ+CoVCo+2A3evmm28uXjT5wAc+sEMXbZ1DQOfR1s4RNv+8vT627Mf1Diifv/71r7nzzjuTJIceeuh2i8sWL16cxx57LKtWrcqGDRuycOHC3H777Tn//PNz2GGHNXuTprVyFbDrlPMzhHsm0L68ltHVnUNA+7Aj9zJdh6D5rwEArWr16tXF5b59+253/T59+mTt2rVZs2bNrgwL2MW++93v5rHHHkuSnHnmmdsc3n3ffffNmWeemaOOOqp4ojNr1qzcfPPNuemmm7J+/fp87GMfS0VFRT7ykY+UbL85z+xojtlszZo16dGjx049L2DH9e7dO+9617tywgkn5IADDkjfvn2zdOnSTJ06NT/60Y+yfPny3HrrrTnjjDNyzz33pFu3bsVtd+b8Yev39ZZaqx9g99mZi7bOIaDzaWvnCK8ljzjPgPLYsGFDLrnkktTV1SVJ/uu//qvZdSsrK3PCCSfklFNOyetf//oMHjw4q1evzpNPPpkrrrgizz77bKZNm5bjjz8+jz32WPbZZ59G2/scAm1XW/gM4Z4JtB/z5s0rjmr0lre8pTgiYXOcQ0D7siP3Ml2HQHEZ7Cbr168vLnfv3n2762/+gLVu3bpdFhOwa02dOjX/+q//miQZOnRofvjDHza77qRJk3L++eeXjEpy+OGH55xzzsmvf/3rnHnmmdm0aVP++Z//Oe9617tKhnndnGd2Jsck8gzsagsXLmzyW7wnnnhi/vEf/zHvfOc789RTT2Xq1Kn54Q9/mE9+8pPFdXbm/GFb7+vW6gfYPRYsWJD7778/SXLkkUdm/Pjxza7rHAI6p7Z2jvBa8ogcAuXxiU98Io8//niS5Pzzz8/pp5/e7Lq33HJLk59ljjnmmFx66aX58Ic/nJ/+9KdZvHhx/umf/im33HJLo/V8DoG2qa18hnDPBNqPa6+9NoVCIcmOjVrmHALajx29l+k6BKbFhN2kZ8+exeWNGzdud/3Nw0P36tVrl8UE7DrPPPNMJk2alNra2vTs2TM33nhjhg4d2uz6/fv33+Z0V6eddlq+/OUvJ0lqamoyefLkknU255mdyTGJPAO72ramh9hzzz1z0003FUcr+973vteofWfOH7b1vm6tfoDd49prr019fX2Shpu+2+IcAjqntnaO8FryiBwCu99ll12Wn/zkJ0kaiki+//3vb3P9bX2W6datW37yk59k//33T5L86le/ysKFCxut43MItE1t5TOEeybQfvzsZz9L0lCgcc4552x3fecQ0D7szL1M1yFQXAa7Sb9+/YrLOzLk4tq1a5Ps2FCOQNsye/bsnHTSSXn55ZfTpUuXXH/99Tn22GNb3O9HPvKR4oWfzUNQb2lzntmZHJPIM1Bu++67b0488cQkyYwZM/Liiy8W23bm/GFb7+vW6gfYPXb2ou32OIeAjqetnSO8ljwih8DudcUVV+QLX/hCkuSAAw7IXXfd1Wiamdeia9euufjii4s/b32e4XMItF+74zOEeybQPjz22GN57rnnkiTvete7tlk4tqOcQ0D57ey9TNchUFwGu0nPnj0zePDgJA3T3GzLyy+/XExyI0eO3OWxAa3nxRdfzNvf/va8+OKLqaioyFVXXZUzzjijVfoeOnRoMY9s/U2eJBkxYkSShpOklStXbrOv+fPnJ0n22GOPRsPKAuVx0EEHFZe3fH9vfl8n2z9/2Py+TkrPH15LPxUVFY22A3aPxx9/PNOmTUvSMGLAwIEDW9yncwjoeNraOcLmn7fXx5b9uN4Bu88vfvGLXHrppUmSUaNG5Z577smQIUNape/mPsskyd57711cbkmuAna/3fEZwj0TaB+uueaa4vKOTIm5o5xDQPm8lnuZrkOguAx2o80nSjNmzEhtbW2z623+BkCSHHjggbs8LqB1LFu2LCeeeGJmzZqVpGF6u9b8sJVkm0PWb/lhbMs8srXa2trMnDkziRwDbUVz7+0dfV9v3b71e/u19DNy5MgWj2QA7LwtL9pub0rMneEcAjqWtnaOsLmfVatW5aWXXmq2j0WLFqW6urrJWIBd4/bbb8+HPvSh1NfXZ9iwYbn33ntb9Usk2zrH6NevX/EGTktyFVAeu+MzhHsm0LZt2rQp119/fZKGotN3vOMdrda3cwgoj9d6L9N1CBSXwW701re+NUnDt3meeOKJZtfbcvjXo48+epfHBbTcqlWrcvLJJxdHGvnGN76Rf/iHf2jVx1i6dGmWLVuWJBk+fHhJ++YckzQ9XP1mjz/+ePGbfnIMtA2bc0fS+P09ZsyY4s/bel8nyR//+MckDd/sGz16dKO2Hc0PL730Ul544YUk8gOUw5YXbffYY4+8853vbJV+nUNAx9PWzhF2tB/XO2D3uvfee3P22WentrY2gwcPzj333JOxY8e26mM091lms8354fnnn9/mTR/5AdqW3fUZwj0TaNvuvPPOLF++PEny/ve/P127dm21vp1DwO7XknuZrkOguAx2o3e/+93F5auvvrrJderr64ujFQwYMCDHH3/87ggNaIGampqceuqpefLJJ5MkX/ziF/O5z32u1R/nxz/+cQqFQpLkuOOOK2mfOHFi+vfvnyT56U9/Wlx3a1OmTCkuT5o0qdXjBHbO7Nmzc8899yRJxo4d22jY94qKiuJw1M8991weeeSRJvt45JFHit/iOeOMM0q++Td+/Pjit3JuuOGG1NTUNNmP/ADl9Zvf/CZLly5N0roXbZ1DQMfT1s4R3vWud6WysuEyY3PXO7bsp7KyMu9617uaXQ9ouYceeihnnHFGNmzYkP79++fuu+/OwQcf3KqPUVtbm6uuuqr487HHHluyzpbXQ7fMJVuqqanJDTfckKRhBILx48e3apzAzttdnyHcM4G2bVeNru4cAna/lt7LdB2CFIDd6phjjikkKXTt2rXw0EMPlbRffvnlhSSFJIWvfOUruz9AYKds2LChcNJJJxXft5/61Kd2uo/Zs2cXnnzyyW2uc8cddxS6d+9eSFLo1atXYcGCBU2u96UvfakYy+WXX17S/tBDDxW6du1aSFI47rjjdjpWYOfcfvvthU2bNjXb/tJLLxXe8IY3FN+33/nOd0rWef755wtdunQpJClMmDChUFNT06i9pqamMGHChOL5xQsvvNDkY02ePLn4OP/wD/9Q0j5jxoxCVVVVIUlh3Lhx24wb2DXOOuus4vv0iSee2O76ziGg45g9e3bxPXj++efv0DZt7Rzhgx/8YLGfG2+8saT9hhtu2OnnCLy2/PDUU08VBgwYUEhS6NOnT+GBBx7Y6cf9wx/+UHj55Zebbd+4cWPh/PPPL8Z2+umnN7vevvvuW0hSqKqqKsyYMaNknUsvvbTYz9VXX73TsUJntrM5oi1+hnDPBHaN13IOsaXly5cXc8Ghhx66w9s5h4C2pzXuZRYKrkN0dhWFQjNfJwB2iaeeeipHH3101q1bl759++YLX/hCjj/++Kxbty7XX399fvzjHydpqNp9/PHH069fvzJHDGzLWWedlVtuuSVJ8ra3vS3//d//XVKFv6Xu3buXfHvm/vvvz/HHH5+jjjoqp59+el7/+tdn6NChSZJZs2blpptuyk033VT8BuD3v//9XHrppU32v3r16kyYMKE4VOxHPvKRvO9970uvXr1y33335etf/3rWrFmTXr165aGHHsphhx3W0l0AbMPo0aOzadOmnHXWWTnqqKMyevTo9OrVK8uWLcv999+fK664ojjNxFvf+tb8/ve/T48ePUr6+fznP59vfOMbSZI3vOEN+dznPpexY8dm5syZ+eY3v5mnnnqquN7Xv/71JmOpq6vLcccdlwcffDBJQ/768Ic/nIEDB+axxx7Lf/zHf2TJkiWprKzMr3/961abjg/YMS+//HKGDRuWDRs25JBDDsnTTz+93W2cQ0D79cADD2TGjBnFn5ctW5bPfOYzSRqmabjkkksarX/BBRc02U9bOkeYP39+3vSmN2Xp0qXp2rVrPv3pT+e0005Lkvz617/Od77zndTW1maPPfbIk08+mREjRuzg3oLOpaX5YebMmXnLW96SJUuWJEm++93v5u1vf/s2H3Po0KHFc4gt+7355pvzrne9KxMnTsz++++fqqqqrFmzJk888UR+/OMfF6fTGTp0aB555JGMGTOmyf7vuuuunH766amvr8+ee+6Zf/u3f8sRRxyRl19+OVdeeWVuvvnmJA2fie6///506dJlO3sJOq+W5oi2+BnCPRNoHa31GWOzH/zgB8Xp8r797W/n05/+9A7F4RwC2p7WuJe5mesQnVhZS9ugk7r99tuL1bZN/Rs/fnxh+vTp5Q4T2AHNvY+b+zdq1KiSPu67774d2rZ3796FK664YrsxTZ8+vbDffvs1209VVVXhjjvu2AV7A9jaqFGjduj9fdZZZ23zG311dXWFiy66aJt9XHzxxYW6urptxrN06dLC4Ycf3mwfPXr0KFx55ZWtvBeAHfHDH/5wm9/6b4pzCGi/tvym/o78a05bO0d45JFHCnvttVez/ey1116FRx55ZKf3F3QmLc0PV1999U5fq2hqJKAdjePQQw8tPPPMM9t9Xj/+8Y+Lo5809e+II44oLF26tDV2IXRoLc0RbfUzhHsm0HKt9Rljsze/+c2FJIUuXboUFi1a1OpxOIeA3WdnPx80dS9zM9chOi8jl0GZzJ07N//zP/+TO++8MwsWLEj37t0zbty4vPe9780nPvGJ9O7du9whAjtgW5X9TRk1alTmzJnT6HerV6/O7bffnocffjiPP/54Fi1alGXLlqW2tjYDBw7MwQcfnBNOOCGXXHJJyTeJm7N27dp8//vfz4033pgZM2Zk48aNGTlyZE455ZR86lOfyqhRo3YqbuC1mTp1aqZOnZqHH344s2bNyrJly1JdXZ2+fftm5MiRectb3pLzzz8/Rx111A71d9ddd+XHP/5x/vznP2fZsmUZMmRIDj/88Hz0ox/d4ZHGamtrc+WVV+a6667Ls88+m7Vr12b48OE54YQT8qlPfSoHH3xwS54y8BodffTReeihh9KlS5fMmzcvw4cP3+42ziGg/brgggvy05/+dIfX397lu7Z0jrBs2bL8z//8T2699dbiZ58xY8bkjDPOyD/90z9l8ODBO9QPdFYtzQ9TpkzJhRdeuFOP+ZWvfCX//u//3uh3zz77bO6+++48/PDDmTZtWpYuXZoVK1akR48e2XPPPTNhwoS85z3vyaRJk3Z4lJC///3v+d///d/ce++9efHFF9OnT58ceOCBOe+883LJJZeka9euOxU3dEYtzRFt+TOEeybQMq35GWP69OnFUYve8Y535De/+c0O9+scAtqe1riXuTXXITofxWUAAAAAAAAAAACUqCx3AAAAAAAAAAAAALQ9issAAAAAAAAAAAAoobgMAAAAAAAAAACAEorLAAAAAAAAAAAAKKG4DAAAAAAAAAAAgBKKywAAAAAAAAAAACihuAwAAAAAAAAAAIASissAAAAAAAAAAAAoobgMAAAAAAAAAACAEorLAAAAAAAAAAAAKKG4DAAAAAAAAAAAgBKKywAAAAAAAAAAACihuAwAAAAAAAAAAIASissAAAAAAAAAAAAoobgMAAAAAAAAAACAEorLAAAAAAAAAAAAKKG4DAAAAAAAAAAAgBKKywAAAACgg7nuuutSUVGRioqKXHrppc2uN2/evAwcODAVFRU58MADs27dut0YJQAAAABtXUWhUCiUOwgAAAAAoHWdd955ue6665Ikv/71r3Pqqac2aq+vr8/b3va2TJ06Nd26dcsjjzySN77xjeUIFQAAAIA2yshlAAAAANAB/eAHP8g+++yTJLnooouyZMmSRu2XX355pk6dmiT52te+prAMAAAAgBJGLgMAAACADuqPf/xjjj/++NTX1+e0007LHXfckSR54oknctRRR2XTpk059thjc99996Wy0vdQAQAAAGjMFSMAAAAA6KCOPfbYfO5zn0vSMDXmD3/4w9TU1OS8887Lpk2b0r9//1xzzTUKywAAAABokpHLAAAAAKAD27RpU4466qg88cQT6dWrV0466aTcdtttSZJrr7025513XpkjBAAAAKCtUlwGAAAAAB3c888/nze+8Y2pqakp/u7cc8/NddddV8aoAAAAAGjrjHcPAAAAAB3c/vvvn8985jPFn/fYY4/84Ac/KGNEAAAAALQHissAAAAAoIOrrq7OT3/60+LPy5Yty5NPPlnGiAAAAABoDxSXAQAAAEAH94lPfCJz5sxJkvTr1y+FQiEXXHBBVq5cWda4AAAAAGjbFJcBAAAAQAd244035mc/+1mS5JJLLsl1112XJJk/f34+/vGPlzM0AAAAANq4ikKhUCh3EAAAAABA61u4cGEOPfTQvPzyy9lvv/3y1FNPpU+fPvn4xz+eH/3oR0mSa6+9Nuedd16ZIwUAAACgLVJcBgAAAAAdUKFQyIknnph77703Xbt2zYMPPpgjjjgiSVJTU5M3vvGNef7559O/f//87W9/yz777FPmiAEAAABoa0yLCQAAAAAd0He/+93ce++9SZIvfelLxcKyJOndu3euvfbadOvWLatWrcqHPvSh1NfXlytUAAAAANooxWUAAAAA0ME8/fTT+cIXvpAkOeqoo/LFL36xZJ0JEybkK1/5SpJk6tSp+fa3v71bYwQAAACg7TMtJgAAAAB0IBs2bMjhhx+ep59+On379s1f/vKXjB07tsl16+rqMnHixDzwwAPp3r17Hn300Rx22GG7N2AAAAAA2izFZQAAAAAAAAAAAJQwLSYAAAAAAAAAAAAlFJcBAAAAAAAAAABQQnEZAAAAAAAAAAAAJRSXAQAAAAAAAAAAUEJxGQAAAAAAAAAAACUUlwEAAAAAAAAAAFBCcRkAAAAAAAAAAAAlFJcBAAAAAAAAAABQQnEZAAAAAAAAAAAAJRSXAQAAAAAAAAAAUEJxGQAAAAAAAAAAACUUlwEAAAAAAAAAAFBCcRkAAAAAAAAAAAAlFJcBAAAAAAAAAABQ4v8HXDsNEmexdawAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 3000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# draw the result\n",
    "i=12\n",
    "def draw(yi, color):\n",
    "        plt.plot(np.arange(input.size(1)), yi[:input.size(1)], color, linewidth = 2.0)\n",
    "        plt.plot(np.arange(input.size(1), input.size(1) + future), yi[input.size(1):], color + ':', linewidth = 2.0)\n",
    "plt.figure(figsize=(30,10))\n",
    "plt.title('Predict future values for time sequences\\n(Dashlines are predicted values)', fontsize=30)\n",
    "plt.xlabel('x', fontsize=20)\n",
    "plt.ylabel('y', fontsize=20)\n",
    "plt.xticks(fontsize=20)\n",
    "plt.yticks(fontsize=20)\n",
    "draw(y[0], 'r')\n",
    "draw(y[1], 'g')\n",
    "draw(y[2], 'b')\n",
    "#plt.savefig('predict%d.pdf'%i)\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}

from flask import Flask
from flask import render_template
import llama_chat as lc 
import ulm 



app = Flask(__name__)

@app.route("/")
def hello_world():
    return "<h1>UML project for Pervasive AI Developer Contest with AMD!</h1><h1><a href=\"/chat/\">Go for ULM Llama Chat Page for More...</a></h1>"


@app.route('/chat/')
@app.route('/chat/<command>')
def chatbot(command=None):
    #chatback=chat(command)
    #chat(command)
    #command='help' ;   chatback='green'
    chatback=lc.chatbot(command)
    umloutput=ulm.update_ulm("UML")
    return render_template('welcomepage.html' , firstL=command, secondL=chatback, thirdL=umloutput)

@app.route('/llama_chat/')
@app.route('/llama_chat/<command>')
def llama_chat(command=None):
    #chatback=chat(command)
    #chat(command)
    #command='help' ;   chatback='green'
    chatback=lc.chatbot(command)
    umloutput=ulm.update_ulm("UML")
    return render_template('llam_chat.html' , firstL=command, secondL=chatback )


@app.route('/ulm_view/')
@app.route('/ulm_view/<command>')
def ulm_view(command=None):
    #chatback=chat(command)
    #chat(command)
    #command='help' ;   chatback='green'
    chatback=lc.chatbot(command)
    umloutput=ulm.update_ulm("UML")
    return render_template('ulm_view.html' , firstL=command, secondL=chatback )

#Llama 7B chat
import torch
from datasets import load_dataset
from transformers import (
    AutoModelForCausalLM,
    AutoTokenizer,
    BitsAndBytesConfig,
    TrainingArguments,
    pipeline
)
from peft import LoraConfig
from trl import SFTTrainer


base_model_name = "Llama-2-7b-chat-hf"
# Model
model = AutoModelForCausalLM.from_pretrained(base_model_name, device_map="auto" )
model.config.use_cache = False
model.config.pretraining_tp = 1
llama_tokenizer = AutoTokenizer.from_pretrained(base_model_name, trust_remote_code=True)
llama_tokenizer.pad_token = llama_tokenizer.eos_token
llama_tokenizer.padding_side = "right"


def chatbot(input=None):
        #query=input
        #text_gen = pipeline(task="text-generation", model=base_model_name, tokenizer=llama_tokenizer, max_length=200)
        #output = text_gen(f"<s>[INST] {query} [/INST]")
        #return output[0]['generated_text']
        if input == None:
            return None
        query  = llama_tokenizer(input, return_tensors='pt')
        query  = query.to('cuda:0')
        pred = model.generate(**query, max_new_tokens=256, repetition_penalty=1.1)
        return llama_tokenizer.decode(pred.cpu()[0])

Credits

norths

7 projects • 2 followers

Fun explorer

Non-Intrusive Load-identification and Monitoring(NLM)

Things used in this project

Hardware components

Software apps and online services

Hand tools and fabrication machines

Story

Non-Intrusive Load-identification Monitoring(NLM)

1 Introduction

2 Solution

3 Hardware and Software

4 Prepare for Non-intrusive Load Monitoring running with LLama2 Model

5 Run the Model

6 Final Result

7 Conclusion and Next to do

Schematics

Block Diagram

Program Flow Chart

Code

Readme

ulm.py

welcomepage.html

app.py

nlm training program on Python Notebook

start.py

llama_chat.py

llam_chat.html

ulm_view.html

Credits

norths

Comments

Embed the widget on your own site

Non-Intrusive Load-identification and Monitoring(NLM)

Non-Intrusive Load-identification and Monitoring(NLM)

Things used in this project

Hardware components

Software apps and online services

Hand tools and fabrication machines

Story

Non-Intrusive Load-identification Monitoring(NLM)

1 Introduction

2 Solution

3 Hardware and Software

4 Prepare for Non-intrusive Load Monitoring running with LLama2 Model

5 Run the Model

6 Final Result

7 Conclusion and Next to do

Schematics

Block Diagram

Program Flow Chart

Code

Readme

ulm.py

welcomepage.html

app.py

nlm training program on Python Notebook

start.py

llama_chat.py

llam_chat.html

ulm_view.html

Credits

norths

Comments