{
"cells": [
{
"cell_type": "markdown",
"id": "bf1b03a9-4aeb-43dd-870f-b8630151a247",
"metadata": {},
"source": [
"# Stock Prediction Deep Learning Model with Sentiment Analysis\n",
"\n",
"##### I will be testing the LSTM only and then the LSTM with Sentiment Analysis."
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "eb97a1b2-770f-4e5e-b944-bc44aa6e9243",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Num GPUs Available: 1\n"
]
}
],
"source": [
"# Last updated: 10/18/2023.\n",
"# Copyright 2023 Shane Khalid. All Rights Reserved.\n",
"#\n",
"# Licensed under the Apache License, Version 2.0 (the \"License\");\n",
"# you may not use this file except in compliance with the License.\n",
"# You may obtain a copy of the License at\n",
"#\n",
"# http://www.apache.org/licenses/LICENSE-2.0\n",
"#\n",
"# Unless required by applicable law or agreed to in writing, software\n",
"# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
"# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
"# See the License for the specific language governing permissions and\n",
"# limitations under the License.\n",
"# ==============================================================================\n",
"\n",
"\n",
"import math \n",
"import time\n",
"import numpy as np\n",
"import pandas as pd\n",
"import seaborn as sns\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.dates as mdates\n",
"import tensorflow as tf\n",
"from keras.models import Sequential\n",
"from keras.callbacks import EarlyStopping\n",
"from keras.layers import LSTM, Dense, Dropout\n",
"from datetime import date, timedelta, datetime\n",
"from pandas.plotting import register_matplotlib_converters\n",
"from sklearn.preprocessing import RobustScaler, MinMaxScaler\n",
"from sklearn.metrics import mean_absolute_error, mean_squared_error\n",
"from tensorflow.keras import utils\n",
"# Always use the GPU\n",
"print(\"Num GPUs Available: \", len(tf.config.experimental.list_physical_devices('GPU')))\n",
"import os\n",
"import csv\n",
"import json\n",
"import datetime\n",
"import pandas as pd\n",
"from tqdm import tqdm\n",
"import snscrape.modules.twitter as sntwitter\n",
"from datetime import datetime"
]
},
{
"cell_type": "markdown",
"id": "a830f253-822d-400d-953c-3b01ab057f66",
"metadata": {},
"source": [
"### Loading Dataset\n",
"\n",
"##### This time I will be importing (scraping) from Yahoo Finance. "
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "dae01ef7-9acd-4671-987c-9dc582af0bbc",
"metadata": {},
"outputs": [],
"source": [
"# Downloading hostorical stock price data from 1st Jan 2010 to today\n",
"stockname = 'GOOG'\n",
"interval = '1d' \n",
"date_today = date.today()\n",
"date_start = datetime.strptime('2010-01-01', \"%Y-%m-%d\")\n",
"\n",
"period1 = int(time.mktime(date_start.timetuple()))\n",
"period2 = int(time.mktime(date_today.timetuple()))\n",
"query_string = f'https://query1.finance.yahoo.com/v7/finance/download/{stockname}?period1={period1}&period2={period2}&interval={interval}&events=history&includeAdjustedClose=true'\n",
"\n",
"# Saving data to CSV file\n",
"stocks_data = pd.read_csv(query_string)\n",
"stocks_data.to_csv(stockname + '.csv')\n",
"\n",
"# Loading data into dataframe\n",
"df = pd.read_csv(stockname + '.csv',parse_dates = True,index_col=['Date'])\n",
"df = df.drop(['Unnamed: 0'],axis=1)"
]
},
{
"cell_type": "markdown",
"id": "8d2dbb84-4fbc-46b7-a401-5172a4ab0504",
"metadata": {},
"source": [
"### EDA"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "f4412171-e5b0-4c7f-9053-b8bdad4983b2",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" Open | \n",
" High | \n",
" Low | \n",
" Close | \n",
" Adj Close | \n",
" Volume | \n",
"
\n",
" \n",
" | Date | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
"
\n",
" \n",
" \n",
" \n",
" | 2010-01-04 | \n",
" 15.615220 | \n",
" 15.678981 | \n",
" 15.547723 | \n",
" 15.610239 | \n",
" 15.610239 | \n",
" 78541293 | \n",
"
\n",
" \n",
" | 2010-01-05 | \n",
" 15.620949 | \n",
" 15.637387 | \n",
" 15.480475 | \n",
" 15.541497 | \n",
" 15.541497 | \n",
" 120638494 | \n",
"
\n",
" \n",
" | 2010-01-06 | \n",
" 15.588072 | \n",
" 15.588072 | \n",
" 15.102393 | \n",
" 15.149715 | \n",
" 15.149715 | \n",
" 159744526 | \n",
"
\n",
" \n",
" | 2010-01-07 | \n",
" 15.178109 | \n",
" 15.193053 | \n",
" 14.760922 | \n",
" 14.797037 | \n",
" 14.797037 | \n",
" 257533695 | \n",
"
\n",
" \n",
" | 2010-01-08 | \n",
" 14.744733 | \n",
" 15.024933 | \n",
" 14.672753 | \n",
" 14.994298 | \n",
" 14.994298 | \n",
" 189680313 | \n",
"
\n",
" \n",
"
\n",
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Plotting line charts\n",
"df_plot = df.copy()\n",
"list_length = df_plot.shape[1]\n",
"ncols = 2\n",
"nrows = int(round(list_length / ncols, 0))\n",
"fig, ax = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, figsize=(14, 7))\n",
"fig.subplots_adjust(hspace=0.5, wspace=0.5)\n",
"colors = ['blue', 'red', 'green', 'pink', 'orange', 'purple']\n",
"for i in range(0, list_length):\n",
" ax = plt.subplot(nrows,ncols,i+1)\n",
" sns.lineplot(data = df_plot.iloc[:, i], ax=ax, color= colors[i])\n",
" ax.set_title(df_plot.columns[i])\n",
" ax.tick_params(axis=\"x\", rotation=30, labelsize=10, length=0)\n",
" ax.xaxis.set_major_locator(mdates.AutoDateLocator())\n",
"\n",
"fig.tight_layout()\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"id": "847927f0-10a2-4b70-8a06-7ea57090509c",
"metadata": {},
"source": [
"### Preprocessing Data"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "217b1fdb-52f6-4673-b78d-b93e64d34544",
"metadata": {},
"outputs": [],
"source": [
"# Creating n samples, sequence_length time steps per samples\n",
"def splitData(sequence_length, data, index_Close):\n",
" x, y = [], []\n",
" data_len = data.shape[0]\n",
" for i in range(sequence_length, data_len):\n",
" #contains sequence_length values 0-sequence_length * columns\n",
" x.append(data[i-sequence_length:i,:]) \n",
" #contains the prediction values for validation, for single-step prediction\n",
" y.append(data[i, index_Close]) \n",
" # Convert the x & y to numpy arrays\n",
" x = np.array(x)\n",
" y = np.array(y)\n",
" return x, y\n",
"\n",
"def processData(df, FEATURES):\n",
" # Indexing the batches\n",
" train_df = df.sort_values(by=['Date']).copy()\n",
" # Saving a copy of the dates' index, before we need to reset it to numbers\n",
" date_index = train_df.index\n",
" # Reset the index, so we can convert the date-index to a number-index\n",
" train_df = train_df.reset_index(drop=True).copy()\n",
" # Create the dataset with features and filter the data to the list of FEATURES\n",
" data = pd.DataFrame(train_df)\n",
" data_filtered = data[FEATURES]\n",
"\n",
" # Adding a prediction column (target variable) and setting dummy values to prepare the data for scaling\n",
" data_filtered_ext = data_filtered.copy()\n",
" data_filtered_ext['Prediction'] = data_filtered_ext['Close']\n",
" # Number of rows in the data\n",
" nrows = data_filtered.shape[0]\n",
" # Convert data to numpy values\n",
" np_data_unscaled = np.array(data_filtered)\n",
" np_data = np.reshape(np_data_unscaled, (nrows, -1))\n",
"\n",
" # Transform the data by scaling each feature to a range between 0 and 1\n",
" scaler = MinMaxScaler()\n",
" np_data_scaled = scaler.fit_transform(np_data_unscaled)\n",
" # Creating a separate scaler that works on a single column for scaling predictions\n",
" scaler_pred = MinMaxScaler()\n",
" df_Close = pd.DataFrame(data_filtered_ext['Close'])\n",
" np_Close_scaled = scaler_pred.fit_transform(df_Close)\n",
"\n",
" # sequence length: this is the timeframe used to make a single prediction\n",
" sequence_length = 50\n",
" # Prediction Index\n",
" index_Close = data.columns.get_loc(\"Close\")\n",
" # Split the training data into train and test datasets with 80:20 split \n",
" train_data_len = math.ceil(np_data_scaled.shape[0] * 0.8)\n",
" # Creating the training and test data\n",
" train_data = np_data_scaled[0:train_data_len, :]\n",
" test_data = np_data_scaled[train_data_len - sequence_length:, :]\n",
"\n",
" # Generate training data and test data\n",
" x_train, y_train = splitData(sequence_length, train_data, index_Close)\n",
" x_test, y_test = splitData(sequence_length, test_data, index_Close)\n",
"\n",
" return x_train, y_train, x_test, y_test, data_filtered, \\\n",
" date_index, scaler, scaler_pred, train_data_len, sequence_length"
]
},
{
"cell_type": "markdown",
"id": "60add952-ca35-487f-bc5b-3fbe5558c4c8",
"metadata": {},
"source": [
"### Train LSTM model\n",
"\n",
"##### Same as usual. Adam optimizer. "
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "88c6193d-4946-4f19-ae55-cf5bfe6707f0",
"metadata": {},
"outputs": [],
"source": [
"# Setting up LSTM model architecture\n",
"\n",
"def getModel(x_train):\n",
" # Creating model with n_neurons = inputshape Timestamps, each with x_train.shape[2] variables\n",
" model = Sequential()\n",
" n_neurons = x_train.shape[1] * x_train.shape[2]\n",
" model.add(LSTM(n_neurons, return_sequences=True, input_shape=(x_train.shape[1], x_train.shape[2]))) \n",
" model.add(LSTM(n_neurons, return_sequences=False))\n",
" model.add(Dense(5))\n",
" model.add(Dense(1))\n",
"\n",
" # Compiling the model\n",
" model.compile(optimizer='adam', loss='mse')\n",
" return model"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "e0dd7500-3179-4691-aee6-7d51ae1cdda1",
"metadata": {},
"outputs": [],
"source": [
"def plotLossCurve(history, color, title):\n",
" # Plot the Loss Curve\n",
" fig, ax = plt.subplots(figsize=(7, 6), sharex=True)\n",
" plt.plot(history.history[\"loss\"],color=color)\n",
" plt.title(title)\n",
" plt.ylabel(\"Loss\")\n",
" plt.xlabel(\"Epoch\")\n",
" plt.xticks(rotation=45)\n",
" plt.legend([\"Train\", \"Test\"], loc=\"upper left\")\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"id": "445d8322-15b5-4a58-be0f-323fd901a9a7",
"metadata": {},
"source": [
"### Evaluate Model Performance"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "a8ae2151-be7f-4572-9c80-2f45fe25979b",
"metadata": {},
"outputs": [],
"source": [
"def eval(model, x_test, y_test, scaler_pred):\n",
" # Get the predicted values\n",
" y_pred_scaled = model.predict(x_test)\n",
" # Unscale the predicted values\n",
" y_pred = scaler_pred.inverse_transform(y_pred_scaled)\n",
" y_test_unscaled = scaler_pred.inverse_transform(y_test.reshape(-1, 1))\n",
"\n",
" # Mean Absolute Error (MAE)\n",
" MAE = mean_absolute_error(y_test_unscaled, y_pred)\n",
" print(f'Median Absolute Error (MAE): {np.round(MAE, 2)}')\n",
"\n",
" # Mean Absolute Percentage Error (MAPE)\n",
" MAPE = np.mean((np.abs(np.subtract(y_test_unscaled, y_pred)/ y_test_unscaled))) * 100\n",
" print(f'Mean Absolute Percentage Error (MAPE): {np.round(MAPE, 2)} %')\n",
"\n",
" # Median Absolute Percentage Error (MDAPE)\n",
" MDAPE = np.median((np.abs(np.subtract(y_test_unscaled, y_pred)/ y_test_unscaled)) ) * 100\n",
" print(f'Median Absolute Percentage Error (MDAPE): {np.round(MDAPE, 2)} %')\n",
"\n",
" return MAE, MAPE, MDAPE, y_pred"
]
},
{
"cell_type": "markdown",
"id": "c9dee881-880e-4851-af06-275de1c88089",
"metadata": {},
"source": [
"### Visualize"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "2fd57b78-c270-49cd-ae74-56caa4477e18",
"metadata": {},
"outputs": [],
"source": [
"def visualizePreds(data_filtered, train_data_len, y_pred, date_index, color):\n",
" # The date from which on the date is displayed\n",
" display_start_date = pd.Timestamp('today') - timedelta(days=500)\n",
" # Add the date column\n",
" data_filtered_sub = data_filtered.copy()\n",
" data_filtered_sub['Date'] = date_index\n",
"\n",
" # Add the difference between the valid and predicted prices\n",
" train = data_filtered_sub[:train_data_len + 1]\n",
" valid = data_filtered_sub[train_data_len:]\n",
" \n",
" valid.insert(1, \"Prediction\", y_pred.ravel(), True)\n",
" valid.insert(1, \"Difference\", valid[\"Prediction\"] - valid[\"Close\"], True)\n",
" # Zoom in to a closer timeframe\n",
" valid = valid[valid['Date'] > display_start_date]\n",
" train = train[train['Date'] > display_start_date]\n",
" \n",
" # Visualize the data\n",
" fig, ax1 = plt.subplots(figsize=(10, 7), sharex=True)\n",
" xt = train['Date']; yt = train[[\"Close\"]]\n",
" xv = valid['Date']; yv = valid[[\"Close\", \"Prediction\"]]\n",
" plt.title(\"Predictions vs Actual Values\", fontsize=20)\n",
" plt.ylabel(stockname, fontsize=18)\n",
" plt.plot(xt, yt, color=\"green\", linewidth=2.0)\n",
" plt.plot(xv, yv[\"Prediction\"], color=color, linewidth=2.0)\n",
" plt.plot(xv, yv[\"Close\"], color=\"black\", linewidth=2.0)\n",
" plt.legend([\"Train\", \"Test Predictions\", \"Actual Values\"], loc=\"upper left\")\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"id": "b217a05e-7ac9-4160-947a-3d4dcf779ce7",
"metadata": {},
"source": [
"### Predict tomorrow's price"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "e13ad85b-663e-462f-90ad-b54bed486051",
"metadata": {},
"outputs": [],
"source": [
"def predictFuturePrice(model, df, sequence_length, feats, scaler, scaler_pred):\n",
" df_temp = df[-sequence_length:]\n",
" new_df = df_temp.filter(feats)\n",
" N = sequence_length\n",
" # Get the last N day closing price values and scale the data to be values between 0 and 1\n",
" last_N_days = new_df[-sequence_length:].values\n",
" last_N_days_scaled = scaler.transform(last_N_days)\n",
"\n",
" # Create an empty list and Append past N days\n",
" X_test_new = []\n",
" X_test_new.append(last_N_days_scaled)\n",
" # Convert the X_test data set to a numpy array and reshape the data\n",
" pred_price_scaled = model.predict(np.array(X_test_new))\n",
" pred_price_unscaled = scaler_pred.inverse_transform(pred_price_scaled.reshape(-1, 1))\n",
" # Print last price and predicted price for the next day\n",
" price_today = np.round(new_df['Close'][-1], 2)\n",
" predicted_price = np.round(pred_price_unscaled.ravel()[0], 2)\n",
" change_percent = np.round(100 - (price_today * 100)/predicted_price, 2)\n",
"\n",
" plus = '+'; minus = ''\n",
" print(f'The close price for {stockname} at {date_today} was {price_today}')\n",
" print(f'The predicted close price is {predicted_price} ({plus if change_percent > 0 else minus}{change_percent}%)')"
]
},
{
"cell_type": "markdown",
"id": "f75d1329-a3b7-4630-948d-1fbf2358b983",
"metadata": {},
"source": [
"### LSTM without Sentiment Analysis"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "22157680-4bf2-43c9-bd82-1f895063a270",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"2023-10-19 17:27:27.723503: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:27.723609: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:27.723648: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:28.951549: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:28.951619: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:28.951627: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1977] Could not identify NUMA node of platform GPU id 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
"2023-10-19 17:27:28.951697: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:880] could not open file to read NUMA node: /sys/bus/pci/devices/0000:43:00.0/numa_node\n",
"Your kernel may have been built without NUMA support.\n",
"2023-10-19 17:27:28.951724: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1886] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 21080 MB memory: -> device: 0, name: NVIDIA GeForce RTX 3090 Ti, pci bus id: 0000:43:00.0, compute capability: 8.6\n",
"2023-10-19 17:31:22.443601: I tensorflow/tsl/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n",
"Num GPUs Available: 1\n",
"Defaulting to user installation because normal site-packages is not writeable\n",
"Requirement already satisfied: pydot in ./.local/lib/python3.10/site-packages (1.4.2)\n",
"Requirement already satisfied: pyparsing>=2.1.4 in /usr/lib/python3/dist-packages (from pydot) (2.4.7)\n",
"Defaulting to user installation because normal site-packages is not writeable\n",
"Requirement already satisfied: graphviz in ./.local/lib/python3.10/site-packages (0.20.1)\n"
]
}
],
"source": [
"# List of considered features\n",
"features1 = ['High', 'Low', 'Open', 'Close', 'Volume'] \n",
"# Split and process dataset\n",
"x_train1, y_train1, x_test1, y_test1, data_filtered1, date_index1, \\\n",
"scaler1, scaler_pred1, train_data_len1, sequence_length1 = processData(df, features1)\n",
"# Get Model\n",
"model1 = getModel(x_train1)\n",
"#Visualizing Model Architecture\n",
"import tensorflow as tf\n",
"tf.keras.utils.plot_model(model1, show_shapes=True)\n",
"print(\"Num GPUs Available: \", len(tf.config.experimental.list_physical_devices('GPU')))\n",
"\n",
"!pip3 install pydot\n",
"!pip3 install graphviz"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "19ce2cc3-a825-42d7-a5a0-07fb82cb7aea",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Model: \"sequential\"\n",
"_________________________________________________________________\n",
" Layer (type) Output Shape Param # \n",
"=================================================================\n",
" lstm (LSTM) (None, 50, 250) 256000 \n",
" \n",
" lstm_1 (LSTM) (None, 250) 501000 \n",
" \n",
" dense (Dense) (None, 5) 1255 \n",
" \n",
" dense_1 (Dense) (None, 1) 6 \n",
" \n",
"=================================================================\n",
"Total params: 758261 (2.89 MB)\n",
"Trainable params: 758261 (2.89 MB)\n",
"Non-trainable params: 0 (0.00 Byte)\n",
"_________________________________________________________________\n"
]
}
],
"source": [
"model1.summary()"
]
},
{
"cell_type": "code",
"execution_count": 33,
"id": "adf10162-a17a-4225-bd2b-8b9b31f1057b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/300\n",
"6/6 [==============================] - 0s 36ms/step - loss: 5.1939e-05 - val_loss: 5.8007e-04\n",
"Epoch 2/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.1459e-05 - val_loss: 6.0817e-04\n",
"Epoch 3/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.6616e-05 - val_loss: 5.9028e-04\n",
"Epoch 4/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.9010e-05 - val_loss: 8.4292e-04\n",
"Epoch 5/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 6.1818e-05 - val_loss: 5.7464e-04\n",
"Epoch 6/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.3029e-05 - val_loss: 5.6822e-04\n",
"Epoch 7/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.1655e-05 - val_loss: 5.7920e-04\n",
"Epoch 8/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.0274e-05 - val_loss: 5.6693e-04\n",
"Epoch 9/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.1249e-05 - val_loss: 5.5994e-04\n",
"Epoch 10/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.0582e-05 - val_loss: 5.6805e-04\n",
"Epoch 11/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.1368e-05 - val_loss: 5.7401e-04\n",
"Epoch 12/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.1662e-05 - val_loss: 5.5698e-04\n",
"Epoch 13/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.4154e-05 - val_loss: 5.5627e-04\n",
"Epoch 14/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.2049e-05 - val_loss: 6.3676e-04\n",
"Epoch 15/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.1783e-05 - val_loss: 5.7822e-04\n",
"Epoch 16/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.9757e-05 - val_loss: 5.7506e-04\n",
"Epoch 17/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.9588e-05 - val_loss: 5.7315e-04\n",
"Epoch 18/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.0603e-05 - val_loss: 5.9778e-04\n",
"Epoch 19/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.1515e-05 - val_loss: 5.5057e-04\n",
"Epoch 20/300\n",
"6/6 [==============================] - 0s 37ms/step - loss: 5.0911e-05 - val_loss: 5.9532e-04\n",
"Epoch 21/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.0137e-05 - val_loss: 5.9072e-04\n",
"Epoch 22/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.8706e-05 - val_loss: 5.4667e-04\n",
"Epoch 23/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.8973e-05 - val_loss: 5.4473e-04\n",
"Epoch 24/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.8683e-05 - val_loss: 5.7136e-04\n",
"Epoch 25/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.0623e-05 - val_loss: 6.3984e-04\n",
"Epoch 26/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 6.0414e-05 - val_loss: 8.6026e-04\n",
"Epoch 27/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 6.5069e-05 - val_loss: 6.3403e-04\n",
"Epoch 28/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 6.5269e-05 - val_loss: 8.0473e-04\n",
"Epoch 29/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 6.5151e-05 - val_loss: 6.2235e-04\n",
"Epoch 30/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 6.0876e-05 - val_loss: 8.0697e-04\n",
"Epoch 31/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 6.1552e-05 - val_loss: 6.3418e-04\n",
"Epoch 32/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 5.7528e-05 - val_loss: 5.2875e-04\n",
"Epoch 33/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.5271e-05 - val_loss: 6.4730e-04\n",
"Epoch 34/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.1992e-05 - val_loss: 7.8030e-04\n",
"Epoch 35/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.5727e-05 - val_loss: 5.3203e-04\n",
"Epoch 36/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.5673e-05 - val_loss: 6.3177e-04\n",
"Epoch 37/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.1613e-05 - val_loss: 5.2338e-04\n",
"Epoch 38/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.0783e-05 - val_loss: 5.8298e-04\n",
"Epoch 39/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.7435e-05 - val_loss: 5.2410e-04\n",
"Epoch 40/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.7786e-05 - val_loss: 5.3591e-04\n",
"Epoch 41/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.7503e-05 - val_loss: 5.4940e-04\n",
"Epoch 42/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.6452e-05 - val_loss: 5.4002e-04\n",
"Epoch 43/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 4.6442e-05 - val_loss: 5.1866e-04\n",
"Epoch 44/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.6059e-05 - val_loss: 5.1688e-04\n",
"Epoch 45/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5808e-05 - val_loss: 5.2306e-04\n",
"Epoch 46/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.6525e-05 - val_loss: 5.1967e-04\n",
"Epoch 47/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.8092e-05 - val_loss: 5.1659e-04\n",
"Epoch 48/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.7516e-05 - val_loss: 5.2063e-04\n",
"Epoch 49/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5809e-05 - val_loss: 5.1318e-04\n",
"Epoch 50/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.7254e-05 - val_loss: 5.1308e-04\n",
"Epoch 51/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5251e-05 - val_loss: 5.5543e-04\n",
"Epoch 52/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5905e-05 - val_loss: 5.1017e-04\n",
"Epoch 53/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.5034e-05 - val_loss: 5.6716e-04\n",
"Epoch 54/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5452e-05 - val_loss: 5.2440e-04\n",
"Epoch 55/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.7777e-05 - val_loss: 6.3141e-04\n",
"Epoch 56/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.1305e-05 - val_loss: 5.5474e-04\n",
"Epoch 57/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.7679e-05 - val_loss: 5.0443e-04\n",
"Epoch 58/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5841e-05 - val_loss: 5.0486e-04\n",
"Epoch 59/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.4836e-05 - val_loss: 5.0822e-04\n",
"Epoch 60/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.4979e-05 - val_loss: 5.2020e-04\n",
"Epoch 61/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.4259e-05 - val_loss: 5.0977e-04\n",
"Epoch 62/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.4635e-05 - val_loss: 7.4517e-04\n",
"Epoch 63/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.9458e-05 - val_loss: 5.0636e-04\n",
"Epoch 64/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.9610e-05 - val_loss: 5.0729e-04\n",
"Epoch 65/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.8559e-05 - val_loss: 6.3253e-04\n",
"Epoch 66/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.7167e-05 - val_loss: 5.4755e-04\n",
"Epoch 67/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.7096e-05 - val_loss: 4.9066e-04\n",
"Epoch 68/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.4310e-05 - val_loss: 4.9044e-04\n",
"Epoch 69/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.4141e-05 - val_loss: 5.4068e-04\n",
"Epoch 70/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.9005e-05 - val_loss: 6.0567e-04\n",
"Epoch 71/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.9279e-05 - val_loss: 5.2227e-04\n",
"Epoch 72/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.7606e-05 - val_loss: 6.8591e-04\n",
"Epoch 73/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.4288e-05 - val_loss: 5.7212e-04\n",
"Epoch 74/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 5.2279e-05 - val_loss: 4.8281e-04\n",
"Epoch 75/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.0421e-05 - val_loss: 5.0018e-04\n",
"Epoch 76/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.0324e-05 - val_loss: 6.8538e-04\n",
"Epoch 77/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.7253e-05 - val_loss: 6.2901e-04\n",
"Epoch 78/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.2506e-05 - val_loss: 5.3627e-04\n",
"Epoch 79/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.8117e-05 - val_loss: 5.5424e-04\n",
"Epoch 80/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.5993e-05 - val_loss: 5.8062e-04\n",
"Epoch 81/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.8207e-05 - val_loss: 5.1133e-04\n",
"Epoch 82/300\n",
"6/6 [==============================] - 0s 37ms/step - loss: 4.9103e-05 - val_loss: 4.7516e-04\n",
"Epoch 83/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.9031e-05 - val_loss: 7.5069e-04\n",
"Epoch 84/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.7567e-05 - val_loss: 4.7432e-04\n",
"Epoch 85/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.3508e-05 - val_loss: 4.7671e-04\n",
"Epoch 86/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3062e-05 - val_loss: 4.6603e-04\n",
"Epoch 87/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.3116e-05 - val_loss: 4.6714e-04\n",
"Epoch 88/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3769e-05 - val_loss: 4.9306e-04\n",
"Epoch 89/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3359e-05 - val_loss: 5.5865e-04\n",
"Epoch 90/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.3120e-05 - val_loss: 5.0523e-04\n",
"Epoch 91/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3626e-05 - val_loss: 4.6584e-04\n",
"Epoch 92/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3846e-05 - val_loss: 5.3466e-04\n",
"Epoch 93/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.0501e-05 - val_loss: 5.3762e-04\n",
"Epoch 94/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.2296e-05 - val_loss: 5.5350e-04\n",
"Epoch 95/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.9425e-05 - val_loss: 4.9249e-04\n",
"Epoch 96/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.7276e-05 - val_loss: 5.4912e-04\n",
"Epoch 97/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1909e-05 - val_loss: 4.5487e-04\n",
"Epoch 98/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.2255e-05 - val_loss: 4.6518e-04\n",
"Epoch 99/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.3957e-05 - val_loss: 4.6023e-04\n",
"Epoch 100/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.2253e-05 - val_loss: 5.0644e-04\n",
"Epoch 101/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0570e-05 - val_loss: 4.7234e-04\n",
"Epoch 102/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.0526e-05 - val_loss: 4.8120e-04\n",
"Epoch 103/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0771e-05 - val_loss: 4.4871e-04\n",
"Epoch 104/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1719e-05 - val_loss: 4.4550e-04\n",
"Epoch 105/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.2153e-05 - val_loss: 5.8857e-04\n",
"Epoch 106/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.3016e-05 - val_loss: 5.0342e-04\n",
"Epoch 107/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.7574e-05 - val_loss: 4.7303e-04\n",
"Epoch 108/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.3402e-05 - val_loss: 5.7601e-04\n",
"Epoch 109/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.5834e-05 - val_loss: 4.3906e-04\n",
"Epoch 110/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5280e-05 - val_loss: 4.5233e-04\n",
"Epoch 111/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 4.0639e-05 - val_loss: 4.3581e-04\n",
"Epoch 112/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1012e-05 - val_loss: 4.4967e-04\n",
"Epoch 113/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.1113e-05 - val_loss: 5.2586e-04\n",
"Epoch 114/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.0591e-05 - val_loss: 4.3176e-04\n",
"Epoch 115/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.0358e-05 - val_loss: 4.7936e-04\n",
"Epoch 116/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.9503e-05 - val_loss: 4.9555e-04\n",
"Epoch 117/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.9956e-05 - val_loss: 4.3167e-04\n",
"Epoch 118/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.9008e-05 - val_loss: 4.3674e-04\n",
"Epoch 119/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.9271e-05 - val_loss: 5.8607e-04\n",
"Epoch 120/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 5.1134e-05 - val_loss: 4.3971e-04\n",
"Epoch 121/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 5.1691e-05 - val_loss: 4.6761e-04\n",
"Epoch 122/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.8591e-05 - val_loss: 5.3190e-04\n",
"Epoch 123/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1040e-05 - val_loss: 4.2018e-04\n",
"Epoch 124/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.9335e-05 - val_loss: 4.8495e-04\n",
"Epoch 125/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1402e-05 - val_loss: 4.5981e-04\n",
"Epoch 126/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1858e-05 - val_loss: 5.0781e-04\n",
"Epoch 127/300\n",
"6/6 [==============================] - 0s 38ms/step - loss: 3.8837e-05 - val_loss: 4.3776e-04\n",
"Epoch 128/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0212e-05 - val_loss: 4.1703e-04\n",
"Epoch 129/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.2124e-05 - val_loss: 5.8645e-04\n",
"Epoch 130/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.6044e-05 - val_loss: 4.1222e-04\n",
"Epoch 131/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.2733e-05 - val_loss: 4.2163e-04\n",
"Epoch 132/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.9469e-05 - val_loss: 4.2418e-04\n",
"Epoch 133/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7501e-05 - val_loss: 4.3677e-04\n",
"Epoch 134/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.8176e-05 - val_loss: 4.2952e-04\n",
"Epoch 135/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0369e-05 - val_loss: 6.6701e-04\n",
"Epoch 136/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.9337e-05 - val_loss: 4.0733e-04\n",
"Epoch 137/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.3170e-05 - val_loss: 4.0414e-04\n",
"Epoch 138/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.1523e-05 - val_loss: 5.1315e-04\n",
"Epoch 139/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0638e-05 - val_loss: 4.1847e-04\n",
"Epoch 140/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0677e-05 - val_loss: 4.3206e-04\n",
"Epoch 141/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0680e-05 - val_loss: 5.6356e-04\n",
"Epoch 142/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.2306e-05 - val_loss: 4.1415e-04\n",
"Epoch 143/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.0676e-05 - val_loss: 3.9718e-04\n",
"Epoch 144/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7425e-05 - val_loss: 4.8362e-04\n",
"Epoch 145/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0463e-05 - val_loss: 4.2217e-04\n",
"Epoch 146/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6883e-05 - val_loss: 3.9380e-04\n",
"Epoch 147/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6913e-05 - val_loss: 3.9412e-04\n",
"Epoch 148/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0054e-05 - val_loss: 4.6880e-04\n",
"Epoch 149/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7090e-05 - val_loss: 3.9191e-04\n",
"Epoch 150/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7152e-05 - val_loss: 3.9188e-04\n",
"Epoch 151/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5870e-05 - val_loss: 4.0480e-04\n",
"Epoch 152/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.6648e-05 - val_loss: 3.9267e-04\n",
"Epoch 153/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.7164e-05 - val_loss: 5.5292e-04\n",
"Epoch 154/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0360e-05 - val_loss: 3.8849e-04\n",
"Epoch 155/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.8028e-05 - val_loss: 4.0467e-04\n",
"Epoch 156/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.8752e-05 - val_loss: 4.3394e-04\n",
"Epoch 157/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6166e-05 - val_loss: 4.1157e-04\n",
"Epoch 158/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7406e-05 - val_loss: 3.9664e-04\n",
"Epoch 159/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5835e-05 - val_loss: 4.7533e-04\n",
"Epoch 160/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7883e-05 - val_loss: 3.7862e-04\n",
"Epoch 161/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7002e-05 - val_loss: 4.1774e-04\n",
"Epoch 162/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.0066e-05 - val_loss: 3.9038e-04\n",
"Epoch 163/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7831e-05 - val_loss: 3.7863e-04\n",
"Epoch 164/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6573e-05 - val_loss: 3.8238e-04\n",
"Epoch 165/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7520e-05 - val_loss: 3.7538e-04\n",
"Epoch 166/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5161e-05 - val_loss: 4.0763e-04\n",
"Epoch 167/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7949e-05 - val_loss: 3.7453e-04\n",
"Epoch 168/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5391e-05 - val_loss: 3.7064e-04\n",
"Epoch 169/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7920e-05 - val_loss: 3.6871e-04\n",
"Epoch 170/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6607e-05 - val_loss: 4.5369e-04\n",
"Epoch 171/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6203e-05 - val_loss: 3.6751e-04\n",
"Epoch 172/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5052e-05 - val_loss: 4.9162e-04\n",
"Epoch 173/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7034e-05 - val_loss: 3.8094e-04\n",
"Epoch 174/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6693e-05 - val_loss: 4.2844e-04\n",
"Epoch 175/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6451e-05 - val_loss: 3.9608e-04\n",
"Epoch 176/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.4949e-05 - val_loss: 3.6128e-04\n",
"Epoch 177/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.4564e-05 - val_loss: 3.7150e-04\n",
"Epoch 178/300\n",
"6/6 [==============================] - 0s 37ms/step - loss: 3.5198e-05 - val_loss: 3.7639e-04\n",
"Epoch 179/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.6648e-05 - val_loss: 3.8561e-04\n",
"Epoch 180/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7191e-05 - val_loss: 4.2380e-04\n",
"Epoch 181/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.8103e-05 - val_loss: 3.6462e-04\n",
"Epoch 182/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 3.5416e-05 - val_loss: 6.5856e-04\n",
"Epoch 183/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.5402e-05 - val_loss: 3.8154e-04\n",
"Epoch 184/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5820e-05 - val_loss: 6.0560e-04\n",
"Epoch 185/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.0189e-05 - val_loss: 4.0743e-04\n",
"Epoch 186/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.9509e-05 - val_loss: 3.7497e-04\n",
"Epoch 187/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.7833e-05 - val_loss: 3.7615e-04\n",
"Epoch 188/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5893e-05 - val_loss: 3.5360e-04\n",
"Epoch 189/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5397e-05 - val_loss: 3.4571e-04\n",
"Epoch 190/300\n",
"6/6 [==============================] - 0s 31ms/step - loss: 3.4754e-05 - val_loss: 3.7442e-04\n",
"Epoch 191/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.3364e-05 - val_loss: 3.7269e-04\n",
"Epoch 192/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.3811e-05 - val_loss: 3.9378e-04\n",
"Epoch 193/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5349e-05 - val_loss: 3.4427e-04\n",
"Epoch 194/300\n",
"6/6 [==============================] - 0s 30ms/step - loss: 3.8081e-05 - val_loss: 4.5326e-04\n",
"Epoch 195/300\n",
"6/6 [==============================] - 0s 31ms/step - loss: 3.5818e-05 - val_loss: 3.8512e-04\n",
"Epoch 196/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.6307e-05 - val_loss: 3.8736e-04\n",
"Epoch 197/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4238e-05 - val_loss: 3.4454e-04\n",
"Epoch 198/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.4051e-05 - val_loss: 3.4162e-04\n",
"Epoch 199/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5975e-05 - val_loss: 3.7907e-04\n",
"Epoch 200/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4723e-05 - val_loss: 3.5411e-04\n",
"Epoch 201/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5441e-05 - val_loss: 3.4065e-04\n",
"Epoch 202/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.2994e-05 - val_loss: 3.3443e-04\n",
"Epoch 203/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.2386e-05 - val_loss: 3.3433e-04\n",
"Epoch 204/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.2532e-05 - val_loss: 4.9409e-04\n",
"Epoch 205/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.6390e-05 - val_loss: 3.5703e-04\n",
"Epoch 206/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7373e-05 - val_loss: 3.7756e-04\n",
"Epoch 207/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.8520e-05 - val_loss: 3.7989e-04\n",
"Epoch 208/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.7669e-05 - val_loss: 3.5140e-04\n",
"Epoch 209/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4343e-05 - val_loss: 4.0296e-04\n",
"Epoch 210/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4826e-05 - val_loss: 3.3830e-04\n",
"Epoch 211/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4411e-05 - val_loss: 3.2794e-04\n",
"Epoch 212/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.3735e-05 - val_loss: 3.2816e-04\n",
"Epoch 213/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.2009e-05 - val_loss: 3.8783e-04\n",
"Epoch 214/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.2901e-05 - val_loss: 3.4561e-04\n",
"Epoch 215/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.2039e-05 - val_loss: 3.2541e-04\n",
"Epoch 216/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1737e-05 - val_loss: 3.3265e-04\n",
"Epoch 217/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1533e-05 - val_loss: 3.2900e-04\n",
"Epoch 218/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1495e-05 - val_loss: 3.7068e-04\n",
"Epoch 219/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1388e-05 - val_loss: 3.3259e-04\n",
"Epoch 220/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1156e-05 - val_loss: 4.0736e-04\n",
"Epoch 221/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1986e-05 - val_loss: 3.2558e-04\n",
"Epoch 222/300\n",
"6/6 [==============================] - 0s 35ms/step - loss: 3.2770e-05 - val_loss: 4.9332e-04\n",
"Epoch 223/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5997e-05 - val_loss: 3.2944e-04\n",
"Epoch 224/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.3435e-05 - val_loss: 4.8294e-04\n",
"Epoch 225/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4517e-05 - val_loss: 3.3350e-04\n",
"Epoch 226/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1024e-05 - val_loss: 3.2739e-04\n",
"Epoch 227/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0829e-05 - val_loss: 3.1635e-04\n",
"Epoch 228/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1500e-05 - val_loss: 3.6727e-04\n",
"Epoch 229/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1012e-05 - val_loss: 3.1509e-04\n",
"Epoch 230/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0797e-05 - val_loss: 3.9436e-04\n",
"Epoch 231/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1639e-05 - val_loss: 3.3696e-04\n",
"Epoch 232/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0828e-05 - val_loss: 3.5289e-04\n",
"Epoch 233/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1682e-05 - val_loss: 3.9576e-04\n",
"Epoch 234/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4716e-05 - val_loss: 3.7956e-04\n",
"Epoch 235/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.9644e-05 - val_loss: 5.6274e-04\n",
"Epoch 236/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.5305e-05 - val_loss: 3.1149e-04\n",
"Epoch 237/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1806e-05 - val_loss: 3.6977e-04\n",
"Epoch 238/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.2046e-05 - val_loss: 3.1771e-04\n",
"Epoch 239/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1964e-05 - val_loss: 3.1637e-04\n",
"Epoch 240/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0958e-05 - val_loss: 3.5127e-04\n",
"Epoch 241/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.2261e-05 - val_loss: 3.0755e-04\n",
"Epoch 242/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4718e-05 - val_loss: 4.4235e-04\n",
"Epoch 243/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4675e-05 - val_loss: 3.1811e-04\n",
"Epoch 244/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.3075e-05 - val_loss: 3.8481e-04\n",
"Epoch 245/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0910e-05 - val_loss: 3.2136e-04\n",
"Epoch 246/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.3528e-05 - val_loss: 4.6475e-04\n",
"Epoch 247/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.6748e-05 - val_loss: 3.0626e-04\n",
"Epoch 248/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 3.2193e-05 - val_loss: 3.0778e-04\n",
"Epoch 249/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0542e-05 - val_loss: 3.7856e-04\n",
"Epoch 250/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.1912e-05 - val_loss: 3.0847e-04\n",
"Epoch 251/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5287e-05 - val_loss: 3.2022e-04\n",
"Epoch 252/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1589e-05 - val_loss: 3.1819e-04\n",
"Epoch 253/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9823e-05 - val_loss: 3.0381e-04\n",
"Epoch 254/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0014e-05 - val_loss: 3.6404e-04\n",
"Epoch 255/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.2922e-05 - val_loss: 3.0589e-04\n",
"Epoch 256/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 2.9711e-05 - val_loss: 4.4349e-04\n",
"Epoch 257/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.3159e-05 - val_loss: 3.6379e-04\n",
"Epoch 258/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.3526e-05 - val_loss: 3.0306e-04\n",
"Epoch 259/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0376e-05 - val_loss: 3.5862e-04\n",
"Epoch 260/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9825e-05 - val_loss: 3.1904e-04\n",
"Epoch 261/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1879e-05 - val_loss: 3.4188e-04\n",
"Epoch 262/300\n",
"6/6 [==============================] - 0s 37ms/step - loss: 3.3935e-05 - val_loss: 3.5108e-04\n",
"Epoch 263/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0101e-05 - val_loss: 3.0878e-04\n",
"Epoch 264/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 2.9251e-05 - val_loss: 2.9988e-04\n",
"Epoch 265/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 2.9135e-05 - val_loss: 3.1429e-04\n",
"Epoch 266/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1148e-05 - val_loss: 3.2367e-04\n",
"Epoch 267/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0476e-05 - val_loss: 2.9820e-04\n",
"Epoch 268/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 2.9238e-05 - val_loss: 3.2258e-04\n",
"Epoch 269/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9202e-05 - val_loss: 2.9815e-04\n",
"Epoch 270/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1455e-05 - val_loss: 3.2140e-04\n",
"Epoch 271/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9394e-05 - val_loss: 3.6962e-04\n",
"Epoch 272/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0026e-05 - val_loss: 3.0022e-04\n",
"Epoch 273/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0705e-05 - val_loss: 4.9474e-04\n",
"Epoch 274/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.4969e-05 - val_loss: 3.0065e-04\n",
"Epoch 275/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1780e-05 - val_loss: 3.2616e-04\n",
"Epoch 276/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0128e-05 - val_loss: 3.0402e-04\n",
"Epoch 277/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.4030e-05 - val_loss: 6.9602e-04\n",
"Epoch 278/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 4.2202e-05 - val_loss: 3.0068e-04\n",
"Epoch 279/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0910e-05 - val_loss: 3.0203e-04\n",
"Epoch 280/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.0341e-05 - val_loss: 3.4467e-04\n",
"Epoch 281/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.5805e-05 - val_loss: 2.9584e-04\n",
"Epoch 282/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1648e-05 - val_loss: 3.5150e-04\n",
"Epoch 283/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.4300e-05 - val_loss: 3.0483e-04\n",
"Epoch 284/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.8822e-05 - val_loss: 3.5656e-04\n",
"Epoch 285/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.6199e-05 - val_loss: 6.8535e-04\n",
"Epoch 286/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 4.1642e-05 - val_loss: 3.7541e-04\n",
"Epoch 287/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.8214e-05 - val_loss: 4.0116e-04\n",
"Epoch 288/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.2742e-05 - val_loss: 2.9232e-04\n",
"Epoch 289/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9566e-05 - val_loss: 3.0349e-04\n",
"Epoch 290/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 2.8411e-05 - val_loss: 3.5827e-04\n",
"Epoch 291/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9510e-05 - val_loss: 3.1645e-04\n",
"Epoch 292/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 3.2016e-05 - val_loss: 4.0342e-04\n",
"Epoch 293/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.0703e-05 - val_loss: 2.9043e-04\n",
"Epoch 294/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.1408e-05 - val_loss: 2.9512e-04\n",
"Epoch 295/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.8574e-05 - val_loss: 3.0567e-04\n",
"Epoch 296/300\n",
"6/6 [==============================] - 0s 35ms/step - loss: 2.8477e-05 - val_loss: 3.2119e-04\n",
"Epoch 297/300\n",
"6/6 [==============================] - 0s 29ms/step - loss: 2.9123e-05 - val_loss: 3.8052e-04\n",
"Epoch 298/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 3.2265e-05 - val_loss: 3.0611e-04\n",
"Epoch 299/300\n",
"6/6 [==============================] - 0s 27ms/step - loss: 2.9846e-05 - val_loss: 3.6055e-04\n",
"Epoch 300/300\n",
"6/6 [==============================] - 0s 28ms/step - loss: 2.9634e-05 - val_loss: 2.9422e-04\n"
]
}
],
"source": [
"# Training the model\n",
"epochs = 300\n",
"history1 = model1.fit(x_train1, y_train1, batch_size=512, epochs=epochs, validation_data=(x_test1, y_test1))"
]
},
{
"cell_type": "code",
"execution_count": 34,
"id": "dd1998bc-5ad2-4db7-83f0-156b52b79d78",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Plotting Loss Curve\n",
"plotLossCurve(history1,\"red\", \"Model 1 Loss\") "
]
},
{
"cell_type": "markdown",
"id": "55fc1c73-7b1d-46a0-8d90-112328196d23",
"metadata": {},
"source": [
"#### Hmm.. My loss is still pretty high. Maybe I'll try it with more epochs later"
]
},
{
"cell_type": "code",
"execution_count": 35,
"id": "7559641e-9e23-4877-b153-d9b169467ddc",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"22/22 [==============================] - 1s 18ms/step\n",
"Median Absolute Error (MAE): 1.79\n",
"Mean Absolute Percentage Error (MAPE): 1.55 %\n",
"Median Absolute Percentage Error (MDAPE): 1.12 %\n"
]
}
],
"source": [
"# Evaluating model\n",
"MAE1, MAPE1, MDAPE1, y_pred1= eval(model1, x_test1, y_test1, scaler_pred1)"
]
},
{
"cell_type": "code",
"execution_count": 36,
"id": "0ac5630b-239f-4c36-9789-f2463a772a16",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA2oAAAJlCAYAAACrCgFqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3iUZdbH8e+kF1IIBJLQe5UmIogKKgoi2DsqKPa21rWsi113rbgC7qsiyiq2XVSsCFIFpEjvLfSEkpDeM8/7x2SemScz6T38PteVy6fPnRCTnDnnPrfNMAwDERERERERqTd86noAIiIiIiIiYqVATUREREREpJ5RoCYiIiIiIlLPKFATERERERGpZxSoiYiIiIiI1DMK1EREREREROoZBWoiIiIiIiL1jAI1ERERERGRekaBmoiIiIiISD2jQE1ExM2iRYuw2WzYbDYWLVrkcX7ChAnYbDbat29f62NzN3z4cGw2G8OHD6/TcYjUhH379pn/H3788cd1PZwqqS8/M0Sk4VGgJiKV4h7QFP8ICQmhXbt2XH755cyaNYuCgoK6Hq5IvbB48WLL/yvLly+v6yE1OiNHjjS/vnv37q3Qve+//75579tvv11DIxQRKR8FaiJS7bKzszlw4ADfffcd48aN46yzziIxMbGuh1Xv6Z33xu+TTz6x7M+cObPGX/NU+7665ZZbzO1PP/20Qvf+5z//AcDPz48bb7yxWsclIlJRCtREpMruueceNm3aZH6sWLGCd9991/zDcPXq1Vx22WUYhlG3A60GH3/8MYZhsG/fvjodx6JFizAMw2t5ptRP2dnZ/Pe//wWgSZMmAHz11Vfk5ubW5bAanSuuuIKwsDCgYoFafHw8y5YtA+Ciiy6iZcuWNTI+EZHyUqAmIlXWokULevfubX4MHjyY+++/n7Vr19K5c2cAVq1axQ8//FDHIxWpO9988w3p6ekA/Otf/wLg5MmTfP/993U5rEYnJCSEq666CoBdu3bxxx9/lOu+Tz/91HwzyT0rJyJSVxSoiUiNadq0KU899ZS5/8svv9ThaETqlrPMsU+fPtx6661069bNclyqT2XKH53XRUREcNlll9XIuEREKkKBmojUqEGDBpnb+/fvN7eLd1e02+189NFHnHfeebRs2RIfHx8mTJjg8by1a9dy9913061bN5o0aUJoaCjdunXjnnvuYefOnWWOJzs7m1deeYW+ffsSGhpKs2bNGDp0KB988AF2u73M+8s73yc9PZ0333yT888/n5iYGAICAggPD6d///488MADZokVwHPPPYfNZjPnL+3fv99rkxZ35e36+Pvvv3PzzTfTvn17goKCiIyMpH///jzzzDMcP368xPu8db/86quvuOCCC4iOjiY4OJhu3brx17/+leTk5FLHsHPnTh544AF69+5NWFgYAQEBxMXF0a9fP2677Ta+/PLLCpf/nX/++dhsNtq0aVNmSW1OTg4RERHYbDauvfZaj/MLFizghhtuoEOHDgQHB5vNcAYPHsxjjz3GggULKjQ2bxISEpg/fz4AN910k+W/v/zyS6n/Fu5q+vvKeey5554rdRxlff8lJCQwbdo0rr76arp06UJoaCiBgYG0atWKyy67jC+//LJc/79V1vDhw2nbti0AX375Jfn5+aVev3LlSvPnxzXXXENQUBAAdrudBQsW8NhjjzF06FCaN2+Ov78/kZGR9OvXj8cee4wDBw5UepwV6W7Zvn17bDab15+L7qryMzIlJYWXX36ZIUOG0LRpU/z9/YmOjqZnz55cccUVvPfeexw9erSCn6WIVJohIlIJCxcuNAADMJ599tkSr9u+fbt53ahRo7ze//PPPxsjRoww950f48ePN68vLCw0Hn74YcNms3lc5/zw8/Mz/u///q/EsSQkJBg9evQo8f6RI0cac+fONfcXLlzo8Yzx48cbgNGuXbsSX2fevHlG8+bNS3wd54fTs88+W+a1xX9cDxs2zACMYcOGeR1DYWGhcd9995X6vIiICOPXX3/1er/7v89vv/1m3HTTTSU+p3PnzkZCQoLX53z11VdGQEBAmZ/bpk2bSvx6evPhhx+a9y5atKjUa7/++mvz2m+//dZy7qGHHipzbM2aNavQ2Lx5/fXXDcDw8fExDh06ZBiGYezdu9f8fn7nnXfKfEZtfF+V5/9pwyj9+6+goMDw8fEp83UvvPBCIz093evz4+PjzetmzJhR5tfGm6efftp8xnfffVfqte7/ryxZssQ8Xp6vYUhIiDF79uwSn13az4yKfJ7t2rXz+Lnorqo/I7du3WrExcWV+fm+++67pY5TRKqPHyIiNWjTpk3mdlxcnNdrnnjiCTZu3Mill17KhAkTaNeuHUePHiUtLc285oEHHmDatGkAnHvuuUyYMIGOHTsSEhLChg0bmDx5Mlu2bOGuu+4iJiaGSy+91PIaBQUFjBkzhm3btgGOZgH33HMPbdq04cCBA0ybNo25c+eWmR0qy8KFC7n44ospKCjA19eXm2++mcsuu4y2bduSk5PD1q1b+fnnny3zku69916uvvpqnnnmGb777jvi4uKYO3dulcbx5JNPMnXqVAA6dOjAE088wYABA8jMzGTOnDlMmTKF1NRUxowZw6pVq+jbt2+Jz/r73//O8uXLufzyy7nlllvMf5+pU6fy448/snv3bh5++GE+//xzy31Hjx7l1ltvJS8vjxYtWnD//fczePBgmjdvTnZ2Nrt372bx4sV8++23Ff78rrrqKu677z5yc3P57LPPGDZsWInXzpo1C3CU4l588cXm8R9++IHJkycDjnLEe+65hx49ehAREUFKSgpbtmxh/vz5rFq1qsLjK87ZTXD48OG0atUKcPy7nHXWWSxbtoyZM2fy4IMPlnh/ffm+Kg+jKMN5/vnnc/HFF3PaaacRHR1Neno6e/fu5YMPPmDFihXMmzeP++67z6MTZnW55ZZbeOWVVwDH17/4zwSn/Px8vvzyS8Dxb3L22Web5woKCoiNjeWKK65gyJAhdOzYkaCgIA4ePMjy5cuZNm0aGRkZ3Hjjjaxdu5YePXrUyOdSHlX9GXnzzTdz5MgR/P39ueOOO7j44ouJiYnBbrdz6NAh/vjjD7755pu6+NRETl11HSmKSMNUnoxafn6+MXjwYPO6mTNner0fMJ555pkSX+vXX381r/vwww+9XpOdnW2cf/755jvX+fn5lvNTpkwxn3HnnXd6fcZtt91mGVNFM2rZ2dnmO9IhISFe73c6cOBAhZ5dXGkZjY0bN5oZjd69exsnT570uObnn382rxk0aJDH+eL/Pi+99JLHNXa73bjooovMd+qPHTtmOT99+vRyZcyysrKMrKysMj/n4q688koDMJo2bWrk5uZ6vSYlJcUIDAz0+u9+8803m1/vkjI7hmEYSUlJFR6bu3Xr1plfh48++shy7r333jPPbdmyxev9tfl9Vdb/006lff/Z7XZj165dpd4/adIkAzBsNpuxc+dOj/PVkVEzDMMYNGiQARhBQUFGSkqK12u+++4787UmTZrkMY68vLwSn3/w4EGjVatWBmDcdNNNXq+pjYxaVX9G7tmzp1wZM7vdbiQnJ5c6ThGpPpqjJiLVLjMzk8WLF3PhhReaHdfatWvndX4QQNeuXUudE/OPf/wDcGRRJk6c6PWaoKAgpkyZAjjm4ixcuNBy3vlOc8uWLUtcyPadd94hOjq65E+sDDNnzuTIkSMAvPLKK6XOH2vTpk2lX6cs7733njn/58MPPyQyMtLjmlGjRnHbbbcBjo6cq1evLvF5p59+Ok8//bTHcZvNxiOPPAI4Mg8rVqywnHeunde0aVN69+5d4vODg4MJDg4u/ZPyYty4cYCjc+LPP//s9Zr//ve/5vw35/XFxzdgwACzXb43UVFRFR6bO2ezkODgYLMbodO1115LQECA5Tpv99eH76vystlsZrfXkkyaNInmzZtjGAZz5sypsbE4m4rk5OTw9ddfe73Gme10v96pffv2+Pv7l/j81q1b8/jjjwMwZ86cOluCpKo/I93XuTz33HNLfB2bzUbTpk2rY8giUg4K1ESkyp5//nlLc4ImTZowfPhwswlFixYt+PbbbwkMDPR6/3XXXYevr6/Xc2lpaeZzrr766lLH0aNHD5o3bw5gCRoSEhLYunUr4PjDOCQkxOv9TZo0KTGYLA/n8gOhoaHccccdlX5OVTmbVvTq1YszzzyzxOvcx+i8x5sbb7zRo+mE0+mnn25u792713IuNjYWcARS3333XdkDr6BLLrmEiIgIwFXeWJzzeNu2bTnnnHO8jm/JkiXs2bOn2scHjgDWOYaxY8cSHh5uOR8VFcXo0aMB+Oyzz7w22Kgv31eVZbfbOXLkCDt27GDz5s1s3ryZbdu20bp1awA2bNhQY699ww03mIGwe0DmlJKSYpaLnnXWWXTq1KnU56WlpREfH8+WLVvMz8X588R5rrZVx89I5/8LQJlNTUSk9ihQE5Ea06FDBx5//HE2bdpEv379SryuT58+JZ5bt26d+cfrDTfc4LVrnfvHiRMnAOs7xO7z5M4444xSx+zepbKi1q1bBziCl5KCwZqWm5vLrl27AEoN0gD69+9vZgs2b95c4nXdu3cv8Zx7tsm5RpjTpZdeambzrrjiCs4//3zefvtt/vzzTwoLC0sdW3kEBgaaf5h+//33Hq9/5MgR8w9Y5/eOO2f2JCkpid69e3P99dczY8YMdu/eXeWxOc2dO9fskufs8lic8/ihQ4c8MsFQP76vKsowDD799FPOO+88mjRpQqtWrejevTunnXaa+bF+/XoA8//ZmhAVFcUll1wCwNKlSy2dZwG+/vprM+N68803e33G/v37eeCBB2jfvj0RERF07NiR3r17m5/HnXfeaV5bk59LSarjZ2SHDh3MNzLefvttevXqxaRJk1iwYAFZWVm1/jmJiIMCNRGpsnvuuYdNmzaxadMmNm/ezO7du0lJSWHv3r289tprtGjRotT7SyulOXbsWKXG5P7HhXuDkLLG0rJly0q9Hrj+SHN/d7q2nTx50twu63P19/enWbNmAKU2USktOPDxcf0aKR58NWvWjDlz5tCqVSsMw2DhwoU88sgjDBw4kKioKK688soqL4LuLGfMzs5m9uzZlnNffPGF+Qds8bJHgAsuuIApU6YQHBxMTk4OX375JbfddhtdunShdevW3H333VXO9jjLGZs1a8aoUaO8XjNmzBgzoPVW/lgfvq8qIicnh0suuYSbb76ZRYsWkZ2dXer1ZZ2vKmdAbhgGn332meWcM8sWGBjIdddd53Hvzz//TM+ePZkyZYpHkOdNTX8u3lTHz0iAzz//nCFDhgCwdetWXnzxRS644AIiIyM599xz+fe//01OTk6Vxysi5aeujyJSZS1atCh1DlJZSip7BOsf///3f//HWWedVa5nlhT8lVTC1xjVh8/1nHPOYffu3fzvf//jp59+YsmSJRw6dIi0tDS++eYbvvnmG0aOHMns2bMrlS0aNmwYrVq14vDhw8yaNYvx48eb55wlh87Mhzf33Xcf11xzDbNmzWLevHksW7aM1NRUDh8+zP/93//x/vvv8/TTT/PSSy9VeGypqanm/KukpCSzBK80s2fPZtq0aYSGhlb49eqLl19+2ZwzOGzYMO677z4GDBhATEwMwcHBZnB/7rnnsnTp0hqf13XJJZfQrFkzkpKS+M9//mPOt9y3bx+///474AiWi//MOHHiBDfeeCNZWVk0adKExx57jJEjR9KpUyciIiLMf88FCxZwwQUXANTJHLXq+hnZqlUrli9fzm+//cbs2bNZvHgxW7duJT8/n6VLl7J06VLeeOMNfvrpJ7p27Vqtn4OIeKdATUTqNWfGBxyZncoEhO5/kJS1WGtVFnNt3rw5hw4dIiEhodLPqKqKfK4FBQUkJSUBVW+YUZqgoCDGjRtnZrXi4+P58ccfeffdd9m5cydz587lb3/7W4lNXkrj4+PDDTfcwBtvvMFvv/3G0aNHadmyJTt37uTPP/8EvGfT3LVo0YKHHnqIhx56CLvdzvr16/nmm2+YMmWKuQDwGWecwWWXXVahsX311VcVzkBkZGQwe/ZsSxlebX5f2Ww2DMMoczHqzMxMr8cNw+DDDz8EHEH6ggULLFlXd1VdCqO8/P39uf7665k6dSrbt29nzZo1DBw4kE8//dQMrIo3EQFHI5qUlBQAvvnmG0aMGOH1+VX5PNy/NpX9mlfHz0h3F1xwgRl4JiUlMX/+fN5//30WLFjAnj17uO6668xyXBGpWSp9FJF6rV+/fmZmaNmyZZV6hns2pbTuhuU5X5oBAwYAsGbNmkrN66iODFhgYCBdunQBYOXKlaVeu27dOvLz8wGq/MddRXTo0IH777+f1atXmw0lvvrqq0o/zxmIFRYWmuthOUvcbDYbN9xwQ7mf5ePjw4ABA3jxxRf57bffzOOVGZ+zjDE2NpbPP/+8zA/n16J4+WNtfl+FhYUB1hLa4gzDKHEeX3Jysjn36ZprrikxSMvIyGDHjh3lHldVuQdin376KeAqe4yOjrasr+e0ZcsWwPEmRklBGjj+XSrL+fWG0r/mycnJ5psqxVXHz8iSNGvWjOuuu47ffvvNXHdt/fr15jxYEalZCtREpF6Ljo5m8ODBgKOU7fjx4xV+RlxcnLkQ7ddff13iPJLMzMwqBQxjx44FHHM/3n///QrfHxQUBGA2N6gs5x+VW7ZsKXWxZmfmw/2e2hQeHm42d6lKE4Z+/frRs2dPwBWgORffPuecc2jbtm2lnjtgwAAzQ1nR8cXHx5t/NF911VVcf/31ZX44W/cvWLCAw4cPm8+qze+rDh06AKUHHz///LOZaSquoKDA3C4pAwSO7z33a2vaoEGDzKY4n3/+OcuXL2fnzp0AXH/99V5b8DvHl5OTU2K2Kysry2s3yfJq2rSpOT+xtK/5F198UWJZZXX8jCwPZ5YN6qZpisipSIGaiNR7zzzzDOBoQ3311VeX+EciOP4YnTp1qkfJ2T333AM4Op09+uijXu99+OGHKz0xHxzd+1q1agXA3/72NxYvXlzitYcOHfI45mwWcezYMY8OhhVxzz33mJmMO++8k7S0NI9rfv31V6ZPnw44/ogtqxtmZcydO7fUcr3U1FQzkHQGCJXlzKqtWrWKzz//3HzHv7Syxy+//LLU5g9r1qwxsxwVHd/MmTPNP6zLapnu5LzObrebWR+o3e+rYcOGAY5srLfsTGJiIg888ECJ90dHR5uBx+eff+41OFy9ejV///vfSx1HTXCWkx47dszSqdFb2SNgZqazsrK8voFTWFjI7bffbq5xV1nOdcu+++47r8tE7Nixo8yvV1V/Rq5fv97swumNYRjmEh42m4327duXOh4RqSZ1scq2iDR8CxcuNAADMJ599tkq3b9w4cIyr//LX/5iXh8TE2M899xzxvz5841169YZv//+u/Hxxx8bEydONJo2bWoARnp6uuX+/Px8o3///uYzRo0aZXz77bfGn3/+aXz77bfGRRddZADGwIEDSx3X+PHjDcBo166d13EuWLDA8PPzMwDDz8/PuPXWW43vvvvO+PPPP43ly5cbH330kXH11VcbAQEBHvfOmzfPfO0bb7zRWLFihbFr1y7zw92wYcMMwBg2bJjXcTz++OPmszp16mS8//77xurVq41FixYZjz76qOHv728ARkBAgLFu3TqP+yvy71PS98H48eMNf39/Y/To0cbkyZON+fPnG2vXrjUWL15sTJ061ejRo4d579tvv13qa5QlPj7esNlsBmBERkaan1tycnKJ97Rr186IjIw0xo8fb0yfPt1YunSpsXbtWmPevHnGs88+a0RFRRmA4evra6xevbpC4+nUqZMBGC1atDAKCwvLdU9hYaERGxtrAEavXr0s52rr+2rz5s3m6zRt2tR4++23jdWrVxvLli0zXnvtNSMmJsZo1qyZ0aVLlxK//+677z7z9QYOHGjMmjXLWL16tTF//nzjkUceMYKCgozmzZsbXbt2LfEZ8fHx5jNmzJhRrq9fWfbv329+jzg/evbsWeL1Bw8eNAIDAw3ACAoKMp544glj/vz5xurVq42PP/7YOP300w3AGDp0aJV+Zvzyyy/m/a1btzY+/PBD488//zQWL15sTJo0yYiIiDA6d+5sREdHG4Axfvx4r8+pys/IGTNmGIBxxhlnGC+88ILxww8/GGvWrDFWrFhhzJo1y7jwwgvNZ1922WUV+KqLSFUoUBORSqntQM1utxvPP/+8+UdkaR+hoaFGVlaWxzMOHz5sdOvWrcT7LrroImPu3LlV+qPLMBx/eDn/GCrto7jCwkJj8ODB5bq+rECtsLDQuPfee0t9/YiICGPu3Lle76+uQK2srwFg3H333eUOZkrj/gdzef6gbNeuXZljCwwMrHCg8Pvvv5v333XXXRW61/3fbM2aNZZztfF9ZRiG8dZbb5V4bVRUlLFkyZJSv/9SUlKMfv36lfqMxYsXl/qMmgjUDMMwzjvvPMtYXn311VKv/+ijjwwfH58SP5frrrvOmD9/fpV/Zjz44IMlvkbbtm2NrVu3mt+vJQVqVfkZ6QzUyvo466yzjBMnTpTnSy0i1UCljyLSINhsNiZNmsTOnTv561//aq7F5evrS1hYGD179mTcuHF88sknJCQkEBwc7PGMuLg41q1bx0svvUTv3r0JDg4mMjKSwYMHM23aNH7++edytVAvy8iRI9m7dy+vvPIKZ511Fs2aNcPX15fw8HAGDBjAQw895HXumI+PD7/++ivPPPMMffv2pUmTJpVuMOLj48PUqVNZsmQJ48aNo23btgQGBhIeHk6/fv14+umn2bVrFxdddFFVP90Svf3223z66afcdtttDBw4kFatWhEQEEBwcDBdu3Zl/PjxLF26lPfee6/EphMVUbzMsaxujwsXLuSdd97hqquu4rTTTiM6Oho/Pz/Cw8Pp378/jz32GFu3bmXChAkVGod7MxDnvLPycr++eFOR2vq+evjhh/nll18YOXIkTZs2JTAwkA4dOnDfffexbt06c2HkkkRERLBs2TJefPFFTjvtNIKCgmjSpAk9evTgscceY8OGDWa5X21zX77Bx8enxEXInW699VaWLl3K5ZdfTnR0NP7+/sTGxjJq1Ci+/PJLvvjii1KXFymvd955h1mzZnHuuecSHh5OcHAw3bp148knn2Tt2rXmHNvSVOVn5A033MBPP/3Eww8/zNlnn02HDh0ICQkhICCA1q1bc+mll/LZZ5+xdOlSS5dJEalZNsOog0U/REREREREpETKqImIiIiIiNQzCtRERERERETqGQVqIiIiIiIi9YwCNRERERERkXpGgZqIiIiIiEg9o0BNRERERESknvGr6wE0JHa7nSNHjhAWFlbptY1ERERERKThMwyD9PR04uLiqmU90OIUqFXAkSNHaNOmTV0PQ0RERERE6omDBw/SunXran+uArUKCAsLAxz/GOHh4XU8GhERERERqStpaWm0adPGjBGqmwK1CnCWO4aHhytQExERERGRGpsSpWYiIiIiIiIi9YwCNRERERERkXpGgZqIiIiIiEg9ozlq1cxut5OXl1fXw5BTgL+/P76+vnU9DBERERGpAQrUqlFeXh7x8fHY7fa6HoqcIiIjI4mJidG6fiIiIiKNjAK1amIYBgkJCfj6+tKmTZsaWfROxMkwDLKysjh27BgAsbGxdTwiEREREalOCtSqSUFBAVlZWcTFxRESElLXw5FTQHBwMADHjh2jRYsWKoMUERERaUSU9qkmhYWFAAQEBNTxSORU4nxTID8/v45HIiIiIiLVSYFaNdNcIalN+n4TERERaZwUqImIiIiIiNQzCtSk2rVv357JkyfX9TBERERERBosBWqnMJvNVurHc889V6nnrl69mjvvvLN6BysiIiIicgpR18dTWEJCgrn95ZdfMmnSJHbs2GEea9KkibltGAaFhYX4+ZX9LRMdHV29AxUREREROcUoo3YKi4mJMT8iIiKw2Wzm/vbt2wkLC+Pnn3/m9NNPJzAwkN9//509e/Zw2WWX0bJlS5o0acIZZ5zB/PnzLc8tXvpos9n48MMPueKKKwgJCaFLly7MmTOnlj9bEREREZGGQ4GalOrJJ5/kH//4B9u2baNPnz5kZGQwevRofvvtN9atW8eoUaMYO3YsBw4cKPU5zz//PNdeey0bN25k9OjRjBs3juTk5Fr6LEREREREGhaVPtagge8PJDEjsdZfN6ZJDGvuXFMtz3rhhRe48MILzf2oqCj69u1r7r/44ot88803zJkzh/vvv7/E50yYMIEbbrgBgFdeeYV//etfrFq1ilGjRlXLOEVEREREGhMFajUoMSORw+mH63oYVTJw4EDLfkZGBs899xw//vgjCQkJFBQUkJ2dXWZGrU+fPuZ2aGgo4eHhHDt2rEbGLCIiIiLS0ClQq0ExTWIa/OuGhoZa9h977DHmzZvHG2+8QefOnQkODubqq68mLy+v1Of4+/tb9m02G3a7vdrGKSIiIiLSmChQq0HVVX5YnyxbtowJEyZwxRVXAI4M2759++p2UCIiIiIijYyaiUiFdOnShdmzZ7N+/Xo2bNjAjTfeqMyYiIiIiEg1U6AmFfLWW2/RtGlTzjrrLMaOHcvIkSMZMGBAXQ9LRERERBoyw4Dd78O2N8FeWNejqRdshmEYdT2IhiItLY2IiAhSU1MJDw+3nMvJySE+Pp4OHToQFBRURyOUU42+70RERKRROL4M5p3t2D79Hej2YN2OpxxKiw2qgzJqIiIiIiJSZ7Kzs/nfJ29yKKnowJ9/qdPx1BcK1EREREREpM489dRTXP34N5z/ChQ6Wx/kpdTlkOoFBWoiIiIiIlJn3nnnHQB2JcLh5KKDR36puwHVEwrURERERESkXjjoLH88/F2djqM+UKAmIiIiIiL1wgFnoJYwF4xTewkoBWoiIiIiIlInijeg33+iaCPvJKRsrP0B1SMK1EREREREpE6kp6db9s1ADeDoQnNz+fLlfPXphxT8NgYy4mtpdHXLr64HICIiIiIip6aTJ09a9g+cDADyHDtHF0H3h9m6dSvDhg2joKCA6XfAbRlbYMRiCG1b6+OtTcqoiYiIiIhInSgeqO0/GQyBzRw7xxaDvZBHHn6YgoICAN7+GbDnQWFOLY+09ilQk1OGzWbj22+/BWDfvn3YbDbWr19f6edVxzNERERETmXJycmW/f1HszCihzl28lMxTq5j4SJXCWR4kwBHNi28a20Os04oUDuF2Wy2Uj+ee+65Kj3bGRSVdwwREREMHTqUBQsWVPp1y6tNmzYkJCTQu3fvcl0/YcIELr/88io9Q0RERESsimfUMrLySQkebO6v+e0/5OXlm/v5/rEQ1rnWxleXFKidwhISEsyPyZMnEx4ebjn22GOP1co4ZsyYQUJCAsuWLaN58+aMGTOGvXv3er02Pz/f6/GK8vX1JSYmBj+/yk/TrI5niIiIiJzKigdqAPtzOprbX83+3nIu4URmjY+pvmgQgdqSJUsYO3YscXFxZWZq7r77bmw2G5MnT7YcT05OZty4cYSHhxMZGcnEiRPJyMio2YHXczExMeZHREQENpvNcuyLL76gR48eBAUF0b17d6ZNm2bem5eXx/33309sbCxBQUG0a9eOV199FYD27dsDcMUVV2Cz2cz9kkRGRhITE0Pv3r157733yM7OZt68eYAj4/bee+9x6aWXEhoayssvvwzAd999x4ABAwgKCqJjx448//zzZu0ywK5duzj33HMJCgqiZ8+e5vOcvJUtbtmyhTFjxhAeHk5YWBjnnHMOe/bs4bnnnuOTTz7hu+++M7N/ixYt8vqMxYsXM2jQIAIDA4mNjeXJJ5+0jGv48OE8+OCD/PWvfyUqKoqYmBhL5tIwDJ577jnatm1LYGAgcXFxPPjgg2X+W4qIiIg0RMVLHwEOJPtBUAsAflm+33Iu8XgKdvupsb5ag0gFZGZm0rdvX2677TauvPLKEq/75ptv+OOPP4iLi/M4N27cOBISEpg3bx75+fnceuut3HnnncyaNasmh95gffbZZ0yaNIkpU6bQv39/1q1bxx133EFoaCjjx4/nX//6F3PmzOGrr76ibdu2HDx4kIMHDwKwevVqWrRowYwZMxg1ahS+vr7lft3g4GDAEQg6Pffcc/zjH/9g8uTJ+Pn5sXTpUm655Rb+9a9/mcHUnXfeCcCzzz6L3W7nyiuvpGXLlqxcuZLU1FQeeuihUl/38OHDnHvuuQwfPpwFCxYQHh7OsmXLKCgo4LHHHmPbtm2kpaUxY8YMAKKiojhy5IjHM0aPHs2ECROYOXMm27dv54477iAoKMgSjH3yySc88sgjrFy5khUrVjBhwgSGDh3KhRdeyP/+9z/efvttvvjiC3r16kViYiIbNmwo99dPREREpCHxmlE7cAD6Dydx41dsPmgNygoKCkhKSiI6OrqWRlh3GkSgdvHFF3PxxReXes3hw4d54IEHmDt3Lpdcconl3LZt2/jll19YvXo1AwcOBODdd99l9OjRvPHGG14Du2rxy0DITqyZZ5cmOAZGranSI5599lnefPNNMzDu0KEDW7du5f/+7/8YP348Bw4coEuXLpx99tnYbDbatWtn3uv8H8eZKSuvrKwsnnnmGXx9fRk2bJh5/MYbb+TWW28192+77TaefPJJxo8fD0DHjh158cUX+etf/8qzzz7L/Pnz2b59O3PnzjX/bV955ZVSv4emTp1KREQEX3zxBf7+/gB07eqapBocHExubm6pn8+0adNo06YNU6ZMwWaz0b17d44cOcITTzzBpEmT8PFxJLD79OnDs88+C0CXLl2YMmUKv/32GxdeeCEHDhwgJiaGESNG4O/vT9u2bRk0aFC5v4YiIiIiDYm3QO3IkSMw6jwWbPnK6z1HjhxRoNZQ2O12br75Zh5//HF69erlcX7FihVERkaaQRrAiBEj8PHxYeXKlVxxxRVen5ubm0tubq65n5aWVrGBZSdC9uGK3VMPZGZmsmfPHiZOnMgdd9xhHi8oKCAiIgJwNNe48MIL6datG6NGjWLMmDFcdNFFlXq9G264AV9fX7Kzs4mOjmb69On06dPHPO/+7wawYcMGli1bZpZBAhQWFpKTk0NWVhbbtm2jTZs2lgB8yJAhpY5h/fr1nHPOOWaQVhnbtm1jyJAh2Gw289jQoUPJyMjg0KFDtG3rWOvD/XMDiI2N5dixYwBcc801TJ48mY4dOzJq1ChGjx7N2LFjNQ9OREREGqWTyUkexxISEqDlBOZvcR07qwss3+U637dv31oaYd1pFH/9/fOf/8TPz6/EuTyJiYm0aNHCcszPz4+oqCgSE0vOeL366qs8//zzlR9YcPmzSdWqiq/rnLv3wQcfcOaZZ1rOOcsYBwwYQHx8PD///DPz58/n2muvZcSIEfz3v/+t8Ou9/fbbjBgxgoiICK/vjoSGhnqM7/nnn/daBhsUFFTh1wdXyWVtKB4M2mw2s9a6TZs27Nixg/nz5zNv3jzuvfdeXn/9dRYvXlylIFJERESkPko+tt/j2JEjRzCadGH+Fh/ATnAAXDXIGqidChp8oPbnn3/yzjvvsHbtWksmozo89dRTPPLII+Z+Wloabdq0Kf8Dqlh+WFdatmxJXFwce/fuZdy4cSVeFx4eznXXXcd1113H1VdfzahRo0hOTiYqKgp/f38KCwvL9XoxMTF07lz+NqsDBgxgx44dJd7To0cPDh48SEJCArGxsQD88ccfpT6zT58+fPLJJ+Tn53sNiAICAsr8fHr06MH//vc/DMMwvxeXLVtGWFgYrVu3Ls+nBjiCxrFjxzJ27Fjuu+8+unfvzqZNmxgwYEC5nyEiIiLSEJw84Zjz72NzJFLy8gtISEhg1+7dHDzheCP77K7Qwe29fAVqDcTSpUs5duyYWVYGjjK4Rx99lMmTJ7Nv3z5iYmLM0jKngoICkpOTS51zFBgYSGBgYI2NvT57/vnnefDBB4mIiGDUqFHk5uayZs0aTp48ySOPPMJbb71FbGws/fv3x8fHh6+//pqYmBgiIyMBR+fH3377jaFDhxIYGEjTpk2rbWyTJk1izJgxtG3blquvvhofHx82bNjA5s2beemllxgxYgRdu3Zl/PjxvP7666SlpfG3v/2t1Gfef//9vPvuu1x//fU89dRTRERE8McffzBo0CC6detG+/btmTt3Ljt27KBZs2ZmCai7e++9l8mTJ/PAAw9w//33s2PHDp599lkeeeQRc35aWT7++GMKCws588wzCQkJ4dNPPyU4ONgyB1BERESksXDOUWsaCmFRLdl34DAJCQnMnz/fvGZEb4iNdN1TvKFbY9Ug2vOX5uabb2bjxo2sX7/e/IiLi+Pxxx9n7ty5gGN+UkpKCn/++ad534IFC7Db7R6lfeJw++238+GHHzJjxgxOO+00hg0bxscff0yHDh0ACAsL47XXXmPgwIGcccYZ7Nu3j59++skMSN58803mzZtHmzZt6N+/f7WObeTIkfzwww/8+uuvnHHGGQwePJi3337bDGZ8fHz45ptvyM7OZtCgQdx+++2W+WzeNGvWjAULFpCRkcGwYcM4/fTT+eCDD8zs2h133EG3bt0YOHAg0dHRLFu2zOMZrVq14qeffmLVqlX07duXu+++m4kTJ/LMM8+U+3OLjIzkgw8+YOjQofTp04f58+fz/fff06xZswp8hURERETqt5UrV9K5c2f2HskCICrMl9hWjsRLUlISP/30k3lt8UDtVMmo2QzDMOp6EGXJyMhg9+7dAPTv35+33nqL8847j6ioKEsmzal9+/Y89NBDlpbsF198MUePHuXf//632Z5/4MCBFWrPn5aWRkREBKmpqYSHh1vO5eTkEB8fT4cOHSo9T0qkovR9JyIiIhWWsQ/8wyCwbt4IzsjI4LTTTmPfvn3msUE9mtK6x3nMnj0bcJRBFhQUENXEh+Pv2ckvhKAJjmuHDBnC8uXLa3/gxZQWG1SHBpFRW7NmDf379zczM4888gj9+/dn0qRJ5X7GZ599Rvfu3bngggsYPXo0Z599Nu+//35NDVlEREREpP45Mhe+7wQ/dIf0PXUyhKefftoSpAHkGwGWjt0FBQUAXHBmB3x8IDC4Cc2aRQGnTuljg5ijNnz4cCqS+Cv+Dw+OBYq1uLWIiIiInNIS5oJhh9wTsPpeOO8XqOaGfKVZunQp7777rsfx0NBQswmcuxFX3g/98qD5WcTG3kNSUjIJCQmW5m2NVYMI1EREREREpGoWLlzIJ2/M4YGz4PQOQOKvsP9zVh/vQuL+LVzcPRm/jtdBSKsaef2srCxuu+02c//Jiefy8f+WkJgCN199If4tvQRqoy6Fjh0BGD16NH369CE2Npb8/HwCAgJqZJz1hQI1EREREZFTwI033khiYiLfL4Gk/3Mcmz3tHq56PQ2A4T3gi7++Rctx28G/SbW//rPPPmv2nRgyZAgv3d6Lh09fwuGT0P/2W/llzUnL9e3bt6djUZAGjrWTTyUNYo6aiIiIiIhUXnp6OomJiQAkZ8CxVNiVCLdOSTOvWbQNrnvtMGx6rtpff+XKlbz11luAYwmsjz76CN+8BFpEQP/2QHCcZY4awIgRI6p9HA2JAjURERERkUbu4MGDlv1fN/ty9Ts+pGVbr9t4ANgxudobjTz33HPY7XZzu3v37pDtbApig+AYjzlqp3qgptJHEREREZHGKi8FDn7DwZ3W/MzdH9nJzHE06+sRB3mFsOcoZOUBRiEkrYKwTtU2jF27dgEQHh7OY4895jiYddjx36AW4ONPs2bNzLb8AOeff361vX5DpIyaiIiIiEhjteZBWHkbB+Y/bjnsDNJCg3z471+gZdEyYLn5UGgH8lOrdRgpKSkANG/eHD8/P7AXQo6jFJNgR/MSHx8fBg4cCMCwYcOIjo6u1jE0NMqoiYiIiIg0RoYd9v0HgINHjnu95P13XqRn5KuEBGaYx7JyISw/zev1lRqGYZiBWmRkpONg7jFH5g4g2DU37euvv+b777/nsssuq7bXb6iUUZN6zWaz8e2339boawwfPpyHHnqoRl9DREREpNIOfgN7PwZ7fsXuS9tpbh444Xn63nvv5cY7n4ZLthDS5jzzeFYe1ZpRy8rKorDQEZSZgVq226LVbssBtG7dmnvuucejscipSIGaALBixQp8fX255JJLKnxv+/btmTx5cvUPqgxjx45l1KhRXs8tXboUm83Gxo0ba3lUIiIiItXo+ApYeiX8cSssuAhyvGfGvDqxwtw8kOQ6fG53uPfK7mYXRkLbEhLR0jyflQvkVV+g5symAURERBS9yGHXBcEKyrxRoCYATJ8+nQceeIAlS5Zw5MiRsm+oByZOnMi8efM4dOiQx7kZM2YwcOBA+vTpUwcjExEREakmib+6to8tgrlnwMn15bs36Q9z82BRoBYeDIv/DlP/dgmBgYHm+ZCQEHPbkVGrvtJH90CtrIyauChQEzIyMvjyyy+55557uOSSS/j44489rvn+++8544wzCAoKonnz5lxxxRWAo2xw//79PPzww9hsNmw2G+Bou9qvXz/LMyZPnkz79u3N/dWrV3PhhRfSvHlzIiIiGDZsGGvXri33uMeMGUN0dLTHeDMyMvj666+ZOHEiSUlJ3HDDDbRq1YqQkBBOO+00Pv/881Kf663cMjIy0vI6Bw8e5NprryUyMpKoqCguu+wy9u3bZ55ftGgRgwYNIjQ0lMjISIYOHcr+/fvL/bmJiIiIAI7ui+4y98OvQ2H/V2Xemn14GQeTwG6Hg8mOY22bFZ0MjLJcGxoaam5n5VKtpY9eAzVl1MqkQE346quv6N69O926deOmm27io48+wjAM8/yPP/7IFVdcwejRo1m3bh2//fYbgwYNAmD27Nm0bt2aF154gYSEBBISEsr9uunp6YwfP57ff/+dP/74gy5dujB69GjS09PLdb+fnx+33HILH3/8sWW8X3/9NYWFhdxwww3k5ORw+umn8+OPP7J582buvPNObr75ZlatWlXKk0uXn5/PyJEjCQsLY+nSpSxbtowmTZowatQo8vLyKCgo4PLLL2fYsGFs3LiRFStWcOedd5pBrIiIiEi5GIYrUAtoCs0cf39RmAXLroP1Tzu6J3qxed1KOt62hbYPwgcLHd0cAdo4A7UAa6DmnlHLzKVaM2qpqa6gzwzU3MoyadKx2l6rMVHXxxo0cOBAcwX42hQTE8OaNWvKff306dO56aabABg1ahSpqaksXryY4cOHA/Dyyy9z/fXX8/zzz5v39O3bF4CoqCh8fX0JCwsjJiamQuMsvjbG+++/T2RkJIsXL2bMmDHlesZtt93G66+/bhnvjBkzuOqqq4iIiCAiIsK1VgfwwAMPMHfuXL766isz2KyoL7/8ErvdzocffmgGXzNmzCAyMpJFixYxcOBAUlNTGTNmDJ06OdYf6dGjR6VeS0RERE5hmfsgt6gLSLPBcO5sWHU3xH/iOLb1Vcg+DEM+sdy2d+9eLrp4NIkpjv3Jv7jOmRm1gKaWezxLH2twjlpeKhxb7DgQ2h7CulbbazUmCtRqUGJiIocPHy77wjq0Y8cOVq1axTfffAM4slTXXXcd06dPNwOf9evXc8cdd1T7ax89epRnnnmGRYsWcezYMQoLC8nKyuLAgQPlfkb37t0566yz+Oijjxg+fDi7d+9m6dKlvPDCCwAUFhbyyiuv8NVXX3H48GHy8vLIzc21/DCqqA0bNrB7927CwsIsx3NyctizZw8XXXQREyZMYOTIkVx44YWMGDGCa6+9ltjY2Eq/poiIiJyCkla7tpudAb5BMHgGNO0P6x51tLePnwm9J5mLUx85coQRI0aQcDTZvHW723Sw8mTUarz0MeEXMByLWtPqUlDVkVcK1GpQRTNMdfG606dPp6CgwNIC1TAMAgMDmTJlChEREQQHB1d4DD4+PpZyRHCUDLobP348SUlJvPPOO7Rr147AwECGDBlCXl5ehV5r4sSJPPDAA0ydOpUZM2bQqVMnhg0bBsDrr7/OO++8w+TJkznttNMIDQ3loYceKvU1bDZbqWPPyMjg9NNP57PPPvO417kw44wZM3jwwQf55Zdf+PLLL3nmmWeYN28egwcPrtDnJiIiIqeub77+jAdegRvPgtf+r6gSyGaD7n+BnKOOjBrAyXUQ1omkpCQuvPBC4uPjS3xm+TNqNdhM5PDXrpOtx1bb6zQ2CtRqUEXKD+tCQUEBM2fO5M033+Siiy6ynLv88sv5/PPPufvuu+nTpw+//fYbt956q9fnBAQEmGtjOEVHR5OYmIhhGGZ54Pr16y3XLFu2jGnTpjF69GjA0aDjxAkvi3yU4dprr+Uvf/kLs2bNYubMmdxzzz3may5btozLLrvMLO202+3s3LmTnj17lvi86Ohoy1y7Xbt2kZWVZe4PGDCAL7/8khYtWhAeHl7ic/r370///v156qmnGDJkCLNmzVKgJiIiIuVSWFjI1U/MwW6H13+E+zJiaOd+QTO3KRwn15MdfQkXX3wxW7duBaBDNGTnY5Y/OrVtXrRRnmYihlEt2S5LoBYeBkd+cuz4h0P0uVV+fmOlZiKnsB9++IGTJ08yceJEevfubfm46qqrmD59OgDPPvssn3/+Oc8++yzbtm1j06ZN/POf/zSf0759e5YsWcLhw4fNQGv48OEcP36c1157jT179jB16lR+/vlny+t36dKF//znP2zbto2VK1cybty4SmXvmjRpwnXXXcdTTz1FQkICEyZMsLzGvHnzWL58Odu2beOuu+7i6NGjpT7v/PPPZ8qUKaxbt441a9Zw99134+/vb54fN24czZs357LLLmPp0qXEx8ezaNEiHnzwQQ4dOkR8fDxPPfUUK1asYP/+/fz666/s2rVL89RERESk3H755Rfsdtf+nF+XWy+I6u/aPrmOWbNmsXq1o1QytkUE85+Grl6KrNqUI6OWmYtjcW17bhU+Axf3ZiIR9r2Qd9KxEzsKfAOq5TUaIwVqp7Dp06czYsQI18KDbq666irWrFnDxo0bGT58OF9//TVz5syhX79+nH/++ZauiS+88AL79u2jU6dOZulfjx49mDZtGlOnTqVv376sWrXK0tTD+fonT55kwIAB3HzzzTz44IO0aNGiUp/LxIkTOXnyJCNHjrSUcT7zzDMMGDCAkSNHMnz4cGJiYrj88stLfdabb75JmzZtOOecc7jxxht57LHHLD+8QkJCWLJkCW3btuXKK6+kR48eTJw4kZycHMLDwwkJCWH79u1cddVVdO3alTvvvJP77ruPu+66q1Kfm4iIiJx6/u/f/7bsf/fdd9YLQtqCf6Rj++R6SyXXJ0/2pWMLiLPGYths0KopgA38rX//eZQ+QrUtem3JqGW71naj1aXV8vzGymYUn4wjJUpLSyMiIoLU1FSPkrecnBzi4+Pp0KEDQUFBdTRCOdXo+05ERKTxOX78ODExMdjdUmp+fn4sX76c7t27uxqazT/PsQg2cM7UM/l9+UoATn4cSaR/Co9+HsBbP7jm5cdEQsJUHE1Jrsu2vObChQvNjtxPjoVXrwfG7IDwqndkHDlyJL/+6li4++RnHYgkHmy+cOUxjxLMhqS02KA6aI6aiIiIiEg9sn37dkuQBo7eAs6lhaKjo+nYsSMdI1OY0A8uPA22bN0CQKvYFkT6HwMgrm03YJP5DLORSHArj9f0mlGrpoYizoyazWYj3B7vqOmLPrtBB2m1QaWPIiIiIiL1iPtSRTeehdkkzen48eOsXLmSz+fu4PK3YWcCnEzJAKBXB1dmJ66zdc3YtnFFgVGHWzxe06OZCFRbi37nHLWw0EB8nNGHyh7LpIyaiIiIiEg94h6oXXkG3PvIcyzYamPv3r3mx+HDhzEMg+w8+Mat0XivGFdJY2z3C4Dp5n6bgePh6kkQEOnxmh7NRKDaAjVnRi0yxG3GVSu15S+LAjURERERkXrEPVBr2wzOGHYRQ6+yLvHzj3/8g6eeegqAH9a5jvduXrTEUFgX4jqdbrmnbdu2XoM0qJ3Sx8igoggwvDuEd6mWZzdmKn2sZurNIrVJ328iIiKNjyVQa46jw2Mx7du3N7eX7XQd79WqaG5bzEXExsZa7mnTpk2Jr2kJ1JwZtSp2fVy9ejXdu3cnN9fxwEjnS6jssVwUqFUTX19fAPLy8sq4UqT6OBfidl/nTURERBo2Z6AW6A/REX4Q7LkgWocOHbze29PZJyR2JGFhYa4OkRRl1EpQExm1KVOmsGPHDnPfFaip7LE8VPpYTfz8/AgJCeH48eP4+/vj46MYWGqOYRhkZWVx7NgxIiMjzTcKREREpOFzBmptosCnSVuwef5d6Z5RM49FQ1gwYPODlsMBiIuLM4Ol0jJqfn5+BAQEkJeX5xaoVS2j5mzJ72Q3cCy03XxIlZ57qlCgVk1sNhuxsbHEx8ezf//+uh6OnCIiIyOJifF8l01EREQaptTUVNLSHJmsts2B0HZer2vRogVBQUHk5OSYxwY6k2zNzgB/RyZt4MCB7Nixg3bt2tGiRYtSXzskJMQRqFVTM5FOnTqRmJho7keHA80Gg4/eYC4PBWrVKCAggC5duqj8UWqFv7+/MmkiIiKNTPFGIgS39nqdzWajffv2bN++3Tw2sGPRRsvzzGNvv/02Z5xxBhdeeGGZFV8hISGkpKS4dX2sWumjsy0/gJ8vXDcYaD645BvEQoFaNfPx8SEoKKiuhyEiIiIiDZBHoFZCl0ZwzFOzBGpdm0BgMHS5xzwWHR3NX/7yl3K9tnOeWnWVPjoDtdBgf3a/kU9MJNDszCo981SiQE1EREREpJ7w6Pjo16TEa4vPUxtw31Zo1gJ8Ayv12mag5syoFWRW6jlOzhLOVlG+xETmOw42H1TKHeJOgZqIiIiISD3hkVHzLzlQa9WqlWW/aYuSm4WUR2hoKAA5+WC3g09hThl3lMxut5uBWoS5flo3RzMRKRe1JhQRERERqSeSkpLM7RbhlJpRCwysXOasJO4t+rPzgMLsSj8rMzPTXO81PKho3VeVPVaIAjURERERkXoiM9NVbhgWDPiFlXjt9ddfT0BAAACff/55lV/bPVDLzKVSgVpOTg6PP/44f/vb38xjEc7Hhveo4ghPLSp9FBERERGpJzIyMszt0EBKLX1s3bo169atIyEhgfPPP7/Kr+2x6HUlSh+nT5/OG2+8YTkWHux8Ae8dLMU7BWoiIiIiIvWEe6DWJJBSSx8BevbsSc+ePavltS2BWiUzan/++afHMTOjFlK1OXSnGpU+ioiIiIjUE85AzWaD4ADKDNSqk7OZCDgzahUP1Hbs2OFxLEIZtUpRoCYiIiIiUk8456iFBICPD7UaqHlm1HKgqCFIeXkL1Fylj608zknJFKiJiIiIiNQTzoxak6CiA3UUqGU611Kz53q/2IukpCRL10qniBAgMBp8gzxvkhIpUBMRERERqSc8ArVSmolUN6+BWgXKH71l06Aoo6ayxwpToCYiIiIiUk84Sx9DnUuk1WJGLSzMtRRAhrPhY0HVA7WIENRIpBIUqImIiIiI1AMFBQXk5DgipCbOQM03pOQbqpl7oJbmjM/s5W/Rr4xa9VKgJiIiIiJSD7gvdt0kCEeQ5uNba68fHh5ubqdXIqO2fft2r8cjFKhVigI1EREREZF6wD1QK2ux65rgNaNWbXPUVPpYUQrURERERETqActi10HU6vw0KCGjVs5AraCggD179ng955ijpoxaRSlQExERERGpB+o6UHPPqKWbGbXyzVGLj48nPz/f67nQQKBJxyqO7tSjQE1EREREpB6wzFELpE4DtYqWPrqXPUZFhlrO+fT5O4S2rfL4TjUK1ERERETk1GEYsOEZWDgaMuLrejQW7hm10DoI1KpS+ugeqJ3dN8Z6ss8LVR3aKUmBmoiIiIicOhLnwZaXIeFnx3/rEY/Sx1puJhIcHIyPjyM8SKtg6aN7oHZuz4DqHtopSYGaiIiIiJw6Dnzl2t4zvdZfPicnhy3rl2NsfgUSF1jO1fUcNZvNZpY/ViWjNqRdWnUP7ZSkQE1EREREGr3s7GxWrVrFp//7je1Hig4GxZR6T3UzDIOzzjqL3v2H8uarf4MFF0B2gnneoz1/LQdq4Cp/TC/HHLVff/2VF154geTkZDNQa9myJb2aHzGv6dOnT42NtbHzq+sBiIiIiIjUhN27d/PPf/6TlStXsnXrVgoLCwGIagK734KmwbZaHc/BgwdZt24dAI/PgscuAQ59B13uBuo+owauhiJlNRNJSkpi5MiRAGzbto2jR48C0K1TGyKCj/LF/fDDrvY8885XXu+XsilQExEREZFG6fbbb2fx4sUex5MzYG08XNDkKNgLwKd2/iT22r4+dau56TlHLczz+hrmDNQyc6HQDr4lzFFbu3atuf3FF1+Y292aJgJw3RC47s7boFu3Ghxt46bSRxERERFplHbv3g045l716dOHM3tEmOf2HAMMO+Qcq7XxuAdippOugKc+lT4CZORQYkbN6+cCdIs45NqJ6FGdQzvlKFATERERkUYpLc3R1KJbt25s+PkfvHhZqnlud2LRRvYRL3fWDK/BTdIaKMz1OF8X66iBl0WvC7wHakeOeP+6dYst2ogeCnGXVPPoTi0qfRQRERGRRsdut5Oeng5AeHgYrH2Uzi1d5/c4E2l1GKgZBtjsuXByHTQfXC/mqHmspWb3XvqYkJDg9Xi3G36EPufW+tICjZEyaiIiIiLS6LgHPeF+aZC2jTbNwM/X0UBk99Gik9mHa21MzsDRKTuvaOPECsBL6WMdBDvuGbW0CmbU/P396dD3IgVp1USBmoiIiIg0Ou5BUVjBXgD8fKFDu9YA7DnqyGiRVUpGLTsRdn8Afz4CJ9dXeUzFM2opWUUbRc+uDxk1S+ljKXPUvAVqnTp1ws9PBXvVRV9JEREREWl0nPPTAMIDi7otth9Hp65J7Np7kMxcOJYGLUsqfcxOhB+6QX7RcxLnwiVbqjSm4oFaahbENQVSNlvO22wQHAD4hVbp9SrDUvqYjRmoGYbBf/7zHwoKCpgwYYLXQK2bOjxWKwVqIiIiItLoWAK1YMA3GPq+SufOr5nHdyeWEqgd/90VpIGjjX5+RpXK+oqXPpoZtbStYC80Sx+bBDqCNfwjK/1aleVR+ljUnn/69OnccccdAMTExChQqwUqfRQRERGRRict1dXhMTwY6PFXCG1Dp06dzON7juG1mciRI0f46wvv8flysNvdH7qtSmPyllEDHMFQZrx5PjSo6HhABLXNW+ljRkYGzzzzjHl8w4YNJCUledyrQK16KVATERERkUYnbcfX5nZ4eDj0fByAzp07m8d3H8VrM5Gnn36a16cv4MapMPR5OOaM+dwWpwbYtGkTEyZMYO7cueUaU4lz1ABSNpvnmwQWHfOv/UDNvfQxraj08Y033uDo0aPm8e0rv/V6rwK16qVATUREREQal5wTpG/92NwN63GzOd+rbdu25vGEk0BuEtgLLLevXr3a3P5jN8xYUrSTap2jdtttt/HJJ59w0003YRhGmcPyyKjZOpjb+Sc2mKWRTYIA3yDwDaS2FV9HLeF4Bq+//rrlmq2bVnnc5+Njo3v37jU+vlOJAjURERERaVxOriMt0xV8hbcd4touvk4YWOaiFRYWsmfPHsvj9p8o2nDLqB04cIA1a9YAcOLECa+lgMV5zFGjtbn91Tc/k5/vaHrSJYY6mZ8Gnl+f52YdJSsry3LNVrck5NgBcHoHeOmv42nWrFltDfOUoGYiIiIiItK42PMcZXtF3IMPj2YZAPmpEBgFwMGDB8nNzbU87kiKH1BgCdS+//57yzWHf32A5ldPg4CmJQ7LI6NW2BR8/DEK83ntPxvM4/dfSJ3MTwPr1+eP3bBuv6PBSUBAAHl5joXfsty+PNcMgpvPAUY9WJvDPCUooyYiIiIijYs9v1yBmiuj5mo8snPnTo/HHUkrKkHMjIeNz0LeSebMmWO55tC6L+DnAZBzvMRhecxRy7RDeHd+3QQb9zkGc2YnOKc7dTI/Daxfnz/jXc1Unnzodq/Xxzrj0qDoGh7ZqUeBmoiIiIg0LvY8VxCGNVALCAggMNAReKU7g7m80gO1hBS3+WebXyD1twksXLjQcs2hZCBzH6x7DDIPQsKvHnPfipc+pmbkQ0RvXvvBdeyvY+uuNT9ARIRngNiqVRxPDFpEoL/n9THOywMVqFU3BWoiIiIi0rgUy6i5Z4nc971l1Hbs2OHxuMTkHEub/rm//GjOJ3M6fNJ58Tz4ZQAsHAk7Jluu8ciopeew5lAkC4p6lHRp15zLTi86WUelj6Ghodxyyy2WYy/d1JyQ3K1Ehnhe3yIcR/avDhqfNHYK1ERERESkcTFKLn0EV6BmmaNWxD2jdkZHx38LCuycaD7ePD7nz0KPlzyUXLSRnQC5Rd1HjvxkucZjjlp6Fq9/utHcf+zqOHydf53XUUYN4OOPP+bjv/akRxxMOBdu7u0YY2SoNXSw2aBZGMqm1RAFaiIiIiLSuJQyR81938yoeSl9jAiBvq5O/hyJeQhGLCW/AH5c5zgWFBRknj/kreljymbLbvFAbe36zfz3pxWAIzN1y+mHXCfrKKMGYLPZGD+6E1tfhxl34Qgebb5ExnS1XNesSdG5oBZ1Ms7GToGaiIiIiDQubnPU/P19zTlpTs6MWm4+5BVgZtRycnLYv38/AF1jIM6tgWNCQgJEn8Wy/VHmQtWXjhlNcIANcCt9dJd7HHKOOYZkt3sEatnZ2diLair/MgqCjGTXyTrMqAHgG2zdH/AWkdHtLIeinRWlaiRSIxSoiYiIiEjj4pZRC28SjM1ms5wuvqizcx21PXv2mAtXd42F2EjXPUeOHAGbD3O2tTGPXXZ2LK2jHNcfcouxLIoWyS6+Fpm70CAf7hlR7GAddX00uS8z0GkidH2AyMhIyyUtnEMMbo1UPwVqIiIiItK4uAVqYaGeHTA8Fr0uyqi5z0/zllEzDIM5Kxzt93194OK262hVdE1atiPoO3LSus4YKY5ArXg2zd0dl3WjaWixgwGRpX6KNa7LvdC0P3S6AwZOA5vNoyOkmVGL6FH74zsFKFATERERkcbFbcHr8DDPQM0zo+YZqHWLhbgoVybuyJEjbNu2jT37jgAwrAc0zV5O6yjXc6cvgtYPQOdHINM5/60oo1a8Nb+Tj48Pf3vgWs8TdZ1Ra9oHLl4LZ74PvgEAnhk1Z7wbrkCtJjSIQG3JkiWMHTuWuLg4bDYb3377reX8c889R/fu3QkNDaVp06aMGDGClStXWq5JTk5m3LhxhIeHExkZycSJE0t9Z0NEREREGqa83Bxyi7rnh4cVT1VZA7W0bMxmIu6t+bvGQmzX8839Xbt2MWvWLHP/0gGO/7oHag9/CoYBCSnw31VFB1MdDUVK+rvzX//6F807nOl5oq4zal4UD9SizUCte62P5VTQIAK1zMxM+vbty9SpU72e79q1K1OmTGHTpk38/vvvtG/fnosuuojjx10rw48bN44tW7Ywb948fvjhB5YsWcKdd95ZW5+CiIiIiNSSdLegqKxAraTSxy4x0PKMe835bfPnz+fll182z4/1Eqi5y3d28E/dAoZhCdQu6AWDe0XxyCOPcO+990JEb88H1HVGzQuvGTW/MAiOrZPxNHZ+dT2A8rj44ou5+OKLSzx/4403Wvbfeustpk+fzsaNG7ngggvYtm0bv/zyC6tXr2bgwIEAvPvuu4wePZo33niDuLi4Gh2/iIiIiNSetPRMczusSROP8yXPUdsOOOamNYmMgbZjadmyJYmJiZb7e/XqScfWCZB3kg4ldKa3G0UbeSchJ9ESqJ3TDZ594loY9KbjQEhr8A83m5oADSejFt7dsaCaVLsGkVGriLy8PN5//30iIiLo27cvACtWrCAyMtIM0gBGjBiBj4+PR4mku9zcXNLS0iwfIiIiIlK/uQdq4eGegZq3OWonT57k+HHHYmhdY4COE8DHn5tvvhmAtm3b0qdPH9q3b88bb7wJrS8HHJk3b5LdKx1TNlvmqDUJwrr2mM0GEb2sD2goGbUmHepkLKeCRhOo/fDDDzRp0oSgoCDefvtt5s2bR/PmzQFITEykRQvr2x1+fn5ERUV5vEPi7tVXXyUiIsL8aNOmTYnXioiIiEj9cCAhxdwOdwvKnLzNUdvl3vExFuh4GwCvvfYax48fJz4+ng0bNhAfH8+oUaOgzdUAtG9etOhzMcmZbgdTt1gyamHBeC4SbSl/tIG/57jrmkdGLQwIbFYnYzkVNJpA7bzzzmP9+vUsX76cUaNGce2113Ls2LEqPfOpp54iNTXV/Dh48GA1jVZEREREasLRo0e55x+Lzf2e3bt4XONR+liQzs4/vzePdevSCcJd9zVv3hwfn2J/NsdcAP4R+PtBBy/rPSflux1M3WwJ1JoE4iVQc8uo+YeDrf79me51HbUQraFWU+rfd0AlhYaG0rlzZwYPHsz06dPx8/Nj+vTpAMTExHgEbQUFBSQnJxMTU0K+GggMDCQ8PNzyISIiIiL11z//+U8OH3eUPp7dDW66/nKPazxKH4EdK782j3Ud6HmPB99A6P4IAF06eDbTSM5zK11M2cKJEydcrx8MBBYL1CLdMmr1cH4a4LGOWlRkGHRSc76a0mgCteLsdju5uY7VBocMGUJKSgp//vmneX7BggXY7XbOPNNLO1QRERERaZB2795tbs+4EwKCytGeH9i5w6308axbyvdip02CK4/RddA1HqeS0/MhtJ1jJ3ULq1evNs/1bEXppY/1cH4aeGbUfMZug6DmdTOYU0CD6PqYkZFh+Z8uPj6e9evXExUVRbNmzXj55Ze59NJLiY2N5cSJE0ydOpXDhw9zzTWO/2l69OjBqFGjuOOOO/j3v/9Nfn4+999/P9dff706PoqIiIg0Inl5eeZ28zDAx9/jGo/2/MDOorYFvr42OnSpwALOQdF06drV43BSah5E9IPM/djz0vnjj+WAo1NixxZ4BmpBLaDZYEj6A1oML//r16Lg4GDrgZBWdTOQU0SDyKitWbOG/v37079/fwAeeeQR+vfvz6RJk/D19WX79u1cddVVdO3albFjx5KUlMTSpUvp1ctV6/vZZ5/RvXt3LrjgAkaPHs3ZZ5/N+++/X1efkoiIiIjUAPdALdAf8AnwuKb4HDXDcAVqHdu3wd/fM7grTZcunvPgkjMKzXlnOxMhJcXRPXxIZ7D5+EJAU+sNNhuc/ytc+DsMeKtCr19bbDYbL7zwAi1atGDmzJl1PZxGr0Fk1IYPH45hGCWenz17dpnPiIqKsqwmLyIiIiKNj3PqC0CAH2ArI6OWDUdOQlbRbV27n1bh1/QaqCWfxAjviQ1Ysct1fEgXIDDae7MQ/zCIHlrh169Nf//733nmmWfMhcCl5jSIjJqIiIiISHk4M2o+tqK2+V5KH4OCgvD19QUcc9R2JrjOdfVSxliWtm3behzLzc0lK6Az4CVQK1722MAoSKsdCtREREREpNFwBmqBzvjMS6Bms9nM8sf0HNh2xHWuW7duFX5NX19fZsyYwdlnn23pjJhcEE1+Afy2peg6HxjYgQYfqEntUKAmIiIiIo2Gs/QxwDnBx8scNXCVP+5KhPs+dh2vTEYNYMKECSxdupRrr73WPJaUmsXbC5qxt2iVqHO6QWgQnq35RbxQoCYiIiIijYYzo2YGajZfr9f5+Xlv1VDZQM2pWbNm5vbatWt57osUwFGK+ca4ohPKqEk5KFATERERkUbDLH30w1H2WMJ8qs6dO3s9XtWlm6Kiosztxx9/nOzcQgDuuxBO71B0QoGalEOD6PooIiIiIlIeltLHEsoeAV588UUCAgJo164dH330EdnZ2Zx33nlVbpThnlFLTk4GIDYSXnRfE1uBmpSDAjURERERaTQspY9eWvM7DRo0iO+//x6A22+/nS+++ILbbrutyq/vnlFzmnwzRIS4HQhuXeXXkcZPgZqIiIhIQ5W0Gg7/AJ1uh9A2dT2aesGj9LEc+vXrR79+/arl9d0zagAjL7qQa86c5zoQ2Qdizq+W15LGTXPURERERBoiww5zB8HmF2DVHXU9mnrDklErZ6BWndwzakFBQUyd9h622AsdB0I7wHm/1Mm4pOFRoCYiIiLSwGRmZnLVFWO5YQrk5AEJc+t6SPVCQUEBdrsdKHuOWk3p2rUrnTp1AuDVV191bA/9HIZ+CRevg+DYWh+TNEwqfRQRERFpYJ566ilmf/cTAGd3hfsuquMB1RPObBrUXUbN39+fdevWcfToUVdnycBm0O7a0m8UKUYZNREREZEG5pNPPjG3l++qw4HUM+6BWkXmqFW3sLCwEtv/i5SXAjURERGRBiYtLc3cjmtahwOpZ5yt+aHsro8i9Z0CNREREZEGLCq0aKMwr9TrTgWepY+1P0dNpLooUBMRERFpQAzDsOznFhRtFKTX/mDqGUvpoz/qrigNmgI1ERERkQbk+PHjlv1sZ2ySr0DNo/RRgZo0YArURERERBqQ+Ph4y35OftFGY8yo5aXA/PNgwUVQkFn25e6lj74oUJMGTYGaiIiISAOyd+9ey36jzqjtmQ7HFkHiPNj/ZZmXe5Y+ao6aNFwK1EREREQakFMqUDux3LWdsqnMy+vDOmoi1UWBmoiIiEgDUjxQa7Slj4YBJ1a49lO3lHmL2vNLY6JATURERKQBOWUyalkHITvBtV+OQE1z1KQxUaAmIiIi0kCkp6ezcuVKy7HsxppRO/GHdT/7COSdLPUWzVGTxkSBmoiIiEgDMXv2bLKzsy3HchprRs297NEppfSsmtrzS2OiQE1ERESkNuSnQUZ82deV4j//+Y/HsVMmowZllj+qmYg0JgrURERERGpaXip83wXmdIRD31fqEYcPH2bBggUAdGoJTYIcxxvlHLXCXDi51rFtc/tztQKBWqCaiUgDp0BNREREpKbt/Rhyjjm2V91e8nX2QsfaYdvfcWy7mTVrFoZhAHDTUAgpmn5ldn1sTIHayXVgLwq6Yka6jqduLfU2j9JHX81Rk4ZLgZqIiIhIDdq/fz83/GUaU38tOuAM2IrLToSFF8LK22HtQ7BrquW0e9njTUMhuCgGMTNq5Sh9dM841Wvu89NaXwp+TRzb2YdLvc2j9FEZNWnAFKiJiIiI1KAXXniBL37dyYMzITEFwOZ5UeIC+LkfHF3oOrZzChh2ADZu3MimTY4Fnwd3hs4xEBQcApS/9PGbb74hMjKSyy+/3MzM1TeFhYWsWLGCpD2LXQebD4GgFo7tkoLcIp5dHxWoScOlQE1ERESkBv3xh6Mpht2AQ8mAb5BjMWdwlDdueh4WjICco9Yb03dB4m+ANZt289mO/wY3aQaUM6OWsZcrr7yS7OxsvvvuO3bt2lXVT6tGvPvuu5x11ll0v/F7dicCfqEQ0QsCiwK1vGSw55d4v9ZRk8ZEgZqIiIhIDcnKymL79u3m/slMoDAb8lMh+ygFv10Em54DigK3mIvgjGmuB+z+N4WFhcyaNQsAP1+4bjDg409wRAwA+YVQaMfRVdKbwhz4dajlUHJycrV8ftXt66+/BuBEmp0rJ0NmyOng4+fKqAHknijxfs/2/JqjJg2XAjURERGRGrJx40bsdru5n5JVtJF5gIdv7EvYJQt46RscnQ37vATn/QydboeAKMd1x5excOFCjhw5AsDovtAsDIgbTVBwE/O5OXmUXPqYug0jO9FyKCEhoXo+wWpUWFjI+vXrzf1NB+H2KccdZZrugVop5Y8qfZTGRIGaiIiISA1Zu3atZf9kpuO/2Sd28s53R8nJh7//F/7v0EPQ+2+OgM3HH8I6Oy7MOcZ/Zn5i3u8se6TdDQQHB5vHs/MpufQxfRdJGdZD9TFQ27VrF1lZWZZjX8zdxuTJk4tl1I6X+Aw1E5HGRIGaiIiISA1Zt26dZd8ZqO3ctAz3fh73PfUOv/76q+tAcCsAcvMNZn/zDQARITbG9Mcxb6vVWEuglpMHFGSAtyYhGbvZX6xaMGHtJ7DtDTjwNSStdmSp6rjBiHtQe2531/HHH3+chetTXQdKyah5lj4qUJOGS4GaiIiISA1Z++cay74zUNu+1RrAFRYWcs0117B582bHgeA4wFH+l5HhuGlMf4OgAKDVZeAXQlBQkHl/dj6ODpGF1owUAOm7OFA8UNuzCtY9Dr9fC3MHweyW8FUorLqrzgI290DtybHwt2siAcfX5rpH/8MJZ8KwnKWPjmYimqMmDZcCNREREZEakJeX5wq8iqQ4A7Udu81jbVuGApCWlsaYMWM4evQohDgCtbX7XPcO6li00f5GAGvpY2kt+tN3eWbUUrwMuDAbdr8PqVtK+axqjnv2cUAHeP6Bixk1ahQAx5PSmL266KTmqMkpQoGaiIiISA3YumULefkFlmNmRm2vqxX///45itNPPx1wLI596aWXkkVzANbGu+4d0AFHk5GYCwFroJbj7FhfkUCt51PQ91XofDdE9HadTNtR3k+x2hiGYWbU4ppCywjwbXEWTz31lHnNpoNFG7kqfZRTgwI1ERERkRqwbtGnHsdOZvsBsP2wI4Dz9YE+PdozZ84c2rRpA8CqVau45Yn/YLfDn/sc99ls0K8d0PZq8HWU81lKH8211Iq16M9Pg5xjHEiyHk5IATrfAb2ehEHvQa+/uU5m7Ka2JScnk5KSAkDv1kUHmw+mV69e5jWbnYFaeUsfFahJA6dATURERKQGrF3yX49jKdn+2O2wo6jpYqeWEBAcQVxcHD/88ANhYWEA/O/H3/nyD9h4wHFdt1hoEgS0u8F8ltfSxzy3phsA6Y6gq3hG7VgqFAbGuQ6EdXK7Z0+5P8fq4t7tMdz5aYW0oVmzZsTGxgKOjJphUP7SR62jJg2cAjURERGR6pa6jXVbD5i7gYGBgKP08WCSK7DqHgv4O9ZD69OnD//+97/Nez5aDHlFlZMD2uPoBBl9jnnee6B20jKMP5fP5S8z4c94y2HsBhw74bbotXM5AKiTjFpOTo65HeRMgvmFANC7t6MsMykDjqVRsYya2vNLA6ZATURERKSaFW59i/X7HdvtW0URF+fIXp3MLGS72xJm3eMAvzBzf+jQoeb2fLc+JAPaA22uAB9f85j3OWop5rGXX36ZgWOf5l9zvY/RspZaQFPHB9RJRs1roOZrDdSgqPxRc9TkFKFATURERKQ65Zxg9x8zySyKGQacMZSmTR1BUEpGPvFucUbnloBfE3O/devW+Pn5eTyyf3sgdqTlmNc5akUZtRMnTvDcc8+VOkyPRa+bFGXVsg5CYa7nDTXIPVALDgB8As2g1BKoHQIKMh0fXmiOmjQmCtREREREqtOe91m7xxUw9D99kBmoFRYaHHSrOGweBvi7Mmq+vr60b9/e45G9WwMthluOeS99TAFg9uzZFBQ46iavPAM+vRem3Qpv3Oi6f+3atRQWFroOmPPUDMgoVitZw7Kzs83tIH8ci3oX8cioAeQc9/ocZ6Dm7+towKI5atKQKVATERERqU7xn7Jun2t3wIABZqAGWDJqkSFYMmoAHTt2tOw3D4MWsa3NuWxOXksfizJqX3zxhXnub5fBuBGx3DMCusS47p80aRKxsbFMnDiR77//nmy/dq6TtTxPzVL6GIA5Pw2gZ8+e5vaWw0UbxxZ7fY6z9DHAmZRURk0aMAVqIiIiItUp66Bloer+/ftbAzW3ZFBkKB4BWPFArVdroNmZHi/jvfQxhYSEBBYtWgQ4Siv7twfCukBgc87pDjGRrmccP36cjz76iEsvvZTm573N47Pgqz8gpt+NPP300+X9jKvMY46aW0atSZMmZufHvc4gd9d7Xp/jzKgpUJPGQIGaiIiISHUpzMHIz2BdUSORmJgYYmNjiYyMNC/Z59Yq35FRC7M8wjNQ83EsTF2MpfTRLaP23//+F8MwALh+SFEJYFgXOOcbmsZ2Z/u39zNz5kyuuuoqQkNdAVFWdh5v/AjXvQtHT6Tz6quvkpparN1/DXEvfQwOwGwk4uT8mhxNhaxcIGklJP/p8RxnoBbojM/U9VEaMAVqIiIiItXEnnWUAycgOcOx379/fwBLRi0xxXV9ZChllj72Gvk0hHfxeC1L6aMzo5afwpdffmkev25w0UaTztDibBizjYhh73LzzTfz3//+l+PHj/P9998zePBgvFm/fn2Jn2t1Ki2jBtChQwdze58zI7lvlsdzzNJHZ3NMX81Rk4ZLgZpIVRTmwfKbYelVHmvXiIjIqSUtLY2eA4bT/iHXsQEDBgDWQM1deDBllj72PuNCr/daSh8LHLV+Bw4dZ9myZQD06tyC3m2KLgjzDPTAEeyNGTOG0aNHez3/55+eWaua4BGoFcuouQdqZvlj1mGK8yh9VEZNGjDP/q8iUn77PnV8gOMd0SGf1O14RESkzixcuJAdu/dZjnnLqDmFB4OvD+BrzR55ZNR69fL6epbSx8IAoICvFh81j11/XhxQFNWUEKg5lRRI1lmg5lfy18Sc45ef5vGcvDxHRi3Q+Yyg6OoeqkitUUZNpCoOz3Ftx8+ErEN1NxYREalTHuuSUXpGLTIER+bIbRFrgIiICMt+s2bNvL6eJVArcGSOvvw9wzx23RC3i83W+96VK1DLPAiJv4FhL/VZleExR82v5IyaK1DznD+Xl+t4ToAf0P1R8A3yuEakoVCgJlJZhgFJq63Htv6zbsYiIiJ17sSJE5b9phGh5ppoUVFRHtc7Oj6GeRwHeO+992jfvj3Tp08v8fXcSx9zCvzYnQhr9jqaiAzo35cuoXscJ4PjPDJUHmNxa3bibufOnaSnp0P2UfjldFgwAra+VuqzKqMic9TiTxQFtsUCNSM3mbx8RxAZ4O8PPR6v9nGK1CaVPopUVvpuyD5iPXbkl7oZi4iI1Lnjx62LML/2zO3YbDYA4uLiPK73toaa0913383dd99d6utZuz768NVK17nrzouDgg2OndhRZY69pIyaYRisX7+ecwI/g9yizy/pjzKfV1FlzVFr1aoV/v7+5Ofns/eY42taPFDLT3f9Tg4Mbeox90+koVFGTaQSCgoK+HLG69zxAczd6HYi91iJ94iISOPmHqjtehNuv+0Wc79Fixb4+Fj/7CotUCsP90AtPRu+WOE6d22XVUVbNuj51zKfVVKgBrBhxc+w5wPXgbyUCo60bO6BmqP00ZpR8/X1pV07x4Lc8cfsGAYec9TyMpPM7YAAdXuUhk+BmkgFZGRk8K9//YsuXbpw/UMf8OEiuOwtSEhxlmGkQWFOqc8QEZHGyT1Qiw4HApub+35+frRs2dJyfWmlj+UREBBAixYtAFix+TibDjqOD+ndjPbhRUFL22sgvFuZzyqp9BEgfs3n1nlpXuaGVZX7HDVvpY/gKn9Mz7Y7lj/IT7eMK9cSqAVW+xhFapsCNZFy+uijj2jbti1/+ctf2Ldvn3k8Nx/eXe6qnSfnuOfNIiLS6DkDNX/forb7boEaOMr33EUEU6WMms1m45lnnvE4fv+5roCFXn8r17NKy6jFx++zHsgrX6BmGAZTp07l3XffxW4vvQFJWaWPgDnfD+BAEoDhCNaKpCS5mrm4ZxtFGioFaiLlUFhYyAMPPMDJk6610i46zfHLGOC9OQdId74ZqPJHEZFT0rFjjp//0eFg8wvx6FxYfJ5aZCjgU7USvXvvvZeBAwea+w9cBDcOLdppNRaa9inXc4KCgizNSdyZXRbxPjesJD/88AP3338/Dz74IL/8Uvoc7rKaiYC1IUtqFh5j2bh5m7ndo0vbco1RpD5ToCZSDqmpqWRlOX4rdO/cho2vwtwnYdwwxy/hlPQ8vl1TdHGOAjURkVONYRhm18cW4Xhdv8sjUAsBjMIqva6vry9ffPEFo849jUcuhrducjtZzmyaOZ4Syh/jjwPNBkHTfo4D+amOzsdlWLHCNWlu48aNpVxZdnt+gLAwV5loujOuc5untn7zLnO7X++uZY5PpL5ToCZSDqmprnfs+salc1rRG3VXXH+XeXy/syuzAjURkVNOamoq+fn5AESHAYHlDdQKqvzanTp14ufPXuLNm8DPuSRbzAhofmaFnlO8/DG0aJpXahac7PgcBEQ6DhiFUJhFWbZu3Wpue1tjzl15MmqWQM0Z17ll1DZs229u9z3N+yLhIg2JAjWRcnAP1CL8UhwbTQfQvOeV5vETzjVGFaiJiJxySmsk4uS19NFe9UANcAVRThXMpoFnoNa/vWs7Pr0l+LstxF2OeWrugdqRI0dKubJYoBaA1zlqXjNqbuNYv90RDDYJgo5dym6gIlLfKVATKYeUlBRzO8I5P3nAGzSPbmEeT3LOZ9YcNRGRU45noFY7pY+moBjXdvRQaDGswo8oXvo4oL1rOz4+3hqo5aeU+qycnBz27Nlj7ids/QkOzSn1egAfW9H8by8ZtfDwcHM7rVhGLSUlhf0JjjLIPm3AJ7Dk5igiDYUCNZFycM+oRYbimKDd8jyaN3e9Y3rCGagpoyYicspxNhIBZ+ljOTJq1VT6CEBYF+h8F0QNhDP+DUULbVeER0atnWs7Pj7emrUrLaOWtoNdf3xm6fSYcCILNj1f4i3OOWpBAUVDL/ccNcc4NmzYYJ7r1w5rUCnSQPnV9QBEGoLU4wfM7YgQG/T7J+B499HHxwe73a7SRxGRU5hHRi3IM1Ar3p4/MhTocIvHdZVis8Ggf1fpEe6BWnAAdHOLK+Pj4+Eit8+ppM6PGXth7iC2LrEuRn3kJBjpuygpfHRm1IL8iw74lneOmuN11q1bZ57r1w4IUKAmDZ8yaiLlkLr9S3M7osP5ENEDAB8fH7NdsDJqIiKnLkugFgYEeJbeNWvWzLIfOeAR6HhrTQ+t3NxLH5uGQge36k3P0scSArXt70B+GlsPWw/n5ENqajrkpXi9zSNQKyOjVrz08eeffzbPndGRKq1PJ1JfKFATKUvaTlL2Lzd3I3qNt5x2lj8mZRS9T6g5aiIipxz3QK1FOODvGajZipUjhp/1T/CpP8VN7hm1pqHQMsK1cLSj9LGMQC0/DfbOAPAI1MCRVSNzn9fXdpY+BjuXlStjjpp7M5GTJ0+yYMECANo1h76dw8CmP3Gl4dN3sUhZEuaSmuVaLyayRXvLaWeglp5tkJuPI6NWjvVlRESk8fAoffSSUQO44YYbAOjfvz9+fvUnSANroBbZJADbaX+nffv2AOzbtw/DzxUoeZ2jtmcGFDjKS7Z5afKYkAJk7PP62p6lj2XMUXPLqH3//fcUFDjm+l15BtiKd8AUaaAUqImUJT+VVLflYiIirHXv7qUsSRmAPc+yAKeIiDR+iYmJ5naLcDzb5ReZNm0aM2fO5Pvvv6+dgVWApfSx80XQ5wU6dOgAOAKpxJNuHSqLd3007LDzXcepAtjpZdm0hBQgc7/nCYoFajY/8A3wuMZ76WMa//vf/8zjV52BGolIo6FATaQs+emkZrt2iwdq7p0fXS36jyMiIqeO/fsdAUiTIIgIocSMWmRkJDfffLNHY5H6wJJRKwranIEaQPyRDNfFxTNqR36CDEc7/j0FZ5BfFNOZGTJKLn0sKCigsNBxQ3AAXuenAfj5+ZmlmM7Sx4y0ZObOnQtATCQM6YIaiUijoUBNpCwF6aRkunaLrzPjtUX/5pegMK/mxyYiInXObrdz4ICjO3D75kXt5UsI1Oqzzp074+Pj+NOwWzfHgtHWQC3F3J7x7VpeffVV8vKKftfteMc8t7XwAnP7HLd1px0ZtX0er+ucnwZFgZ2X+WlOzqyaM1D7aflBcnNzAbhiIPj4oIyaNBr1qzhapD5yy6jZbDZL6QVYSx/NFv3xn0BoO+hT8poxIiLSOBw9etQMWNo537trgPOkWrduzeeff86mTZu4//77gWKB2sHj0A3+jIfbXlkBrKCwsJCUo7sJT5jP42MgOKoT2/a6Aq0LesO8zY5tR0bNrfQx5zjY88nJcaXdgvzxOj/NKSwsjGPHjpGeYwMMZv9+wjx31RlFG/7hXu8VaWgUqImUpSDDnKMWFtbEfLfRyT2jdu2/4KlLbbx8rYHtwNcK1ERETgHOskcoCtT8QsHHv+Qb6rFrr72Wa6+91ty3BmpHoRvMWua6/u9//7u5PWctfDP9JrZu22Yeu6CX69ojKUBGPBz5BXZNgyM/gs2XnB7/Na8pb0YtLdsgJw9+XOMoeYlqGsG53YvKMZVRk0ZCgZpIGYy8NDOjVnx+GlgDNYBX5xhceQYM7LgNso9CcMvaGKaIiNSRffv2mdvtmtMgyx5LYgnU9h0AnwCimngv7f8zHgZePw1/f0cjEF8f6NPWMXcsMQU2HwR7bgo+iy523WTYyTm00NwNDqDUjJqzRX9+AfywDjKKSiAvGzkEf79fHDsK1KSR0Bw1kVK8//77NL9yMYkpjv3ISM9fvsUDNXD8MgLg2OKaG5yIiNQLHhm1RhSoRUZGmnOz9+7dC/4RZJUyBfvYseMcPuxYRK1z2+YE+MGgjo5zKVmwK9HznuxMV6fk8mbUAD5e4jp+5bC2rh01E5FGQoGaSCnuuusuktNd7Yi9ZdTc56h5OLaoBkYlIiL1SWMO1MCVVTt48CD5tnCOe1mBpn00XDBskOVYzz6DILwbZ/Z0/Z5cuQeIuRD6vGQey8mqXKD24/qiY0FwYfD7rouUUZNGokEEakuWLGHs2LHExcVhs9n49ttvzXP5+fk88cQTnHbaaYSGhhIXF8ctt9zCkSPWlRaTk5MZN24c4eHhREZGMnHiRDIyMhCpiPKUPgKkZNscG8qoiYg0eh6Bmn9knY2lJjgDNbvdzsHUYI6ne14zsJM/v8xbykMPPWQeG3b+RTBmO4Nu+dw8tsp+M5z/K7S92jyWk+V6YHAA4G9t2uXOWfrobkx/CHSfEqhATRqJBhGoZWZm0rdvX6ZOnepxLisri7Vr1/L3v/+dtWvXMnv2bHbs2MGll15quW7cuHFs2bKFefPm8cMPP7BkyRLuvPPO2voUpJHwFqgVb9cPkGoUrY+TuhVyjtXwqEREpC45A7VAf2gZQaPNqAHEn/DzmlE7rUdn/PwDePvtt1mwYAHTp0/nnnvuAeCMQYOw2RxvYK5cW9RoxK+JeW92tuuNc0dGzXWuuOKdlwGuvKC79YC6Pkoj0SCaiVx88cVcfPHFXs9FREQwb948y7EpU6YwaNAgDhw4QNu2bdm2bRu//PILq1evZuDAgQC8++67jB49mjfeeIO4uLga/xxOZbt27WLMmDF06NCBH374AT+/BvFtR0FBgccxb0GZj48P/fr1Y/369eaxFCMGOOTYObYY2l5TM4MUEZE6ZRiGGai1bVa0jlejDtR8vWbUThswxNw+77zzOO+888z9iIgIunfvzrZt29iwYQM5OTkEuWXNcrKKBWqlZNSKB2oBAQFc/Pgy2PQA7J/l6LYZNbAin55IvdUgMmoVlZqais1mM/+oXrFiBZGRkWaQBjBixAh8fHxYuXJlHY3y1HHrrbeyc+dO5s6dy+zZs8GeD5tfhjUPQn79LT/1VhrrLaMG8L///c+SoU3Jd/slfXRRdQ9NRETqiZMnT5q/L1xrqDXiQO2Y3XtG7awrS33GmWeeCTimrGzYsAF8XfPQcrKzzO2gACqUUevfvz+h4VFw1qcwYgmM3gwhegNeGodGF6jl5OTwxBNPcMMNN5h1zImJibRo0cJynZ+fH1FRUSQmemk/VCQ3N5e0tDTLh1TcsmWuBVd27twJax+Fjc/Azndh78d1N7AyePv3DgwM9Hptx44defbZZ839lJxAsPk6dtRQRESk0Tp+/Li5HRtZtNEAF7sujXugtutwDiczPa/p2Hdkqc847bTTzO3t27eDj6/Zhj8nxxWoBfsDfuWfozZ48GDHhs0GLc6B8K6ljkOkIWlUgVp+fj7XXnsthmHw3nvvVfl5r776KhEREeZHmzZtqmGUp5bCwkLLfvLexY4AzSkzvpZHVH7eAjX3CePFuWfbUtMzXaUXmqcmItJouf+uiHAu/9XIMmrt27c3t9dsO2o5Z7PBX6+Kxse39GkN3bu75pHt2LHDseHvyJxlZ2eb5ypa+ujM1Ik0Ro0mUHMGafv372fevHmWd1xiYmI4dsz6h3JBQQHJycnExMSU+MynnnqK1NRU8+PgwYMlXive7dy507K/e8Nv1gvyUmpvMCXJ2Avb3oDMA65jhXmk/XaTx6VXXllyaUdISIg5/y4lJQVaDnedVPdHEZFGyT1QCwsq2mhkgVpwcLD591L8oWTz+K3DIP1D+Oe9/ct8hnugtn37dsdGUeYsJyfHPFfR0kcFatKYNYpAzRmk7dq1i/nz53usazVkyBBSUlL4888/zWMLFizAbreX+j94YGAg4eHhlg+pmLVr11r29xw1rBfkp9TeYEqy9BpY9zgsvhSMovHFzyQ9YYPlsr/ffmaJTW0Ay7zIlJQUaDHcdVLz1EREGiX3QC08uGijkQVqYC1/dIoOg9AgIKhlmfe3a9fOnD7gCtQcAVlOTq55XXAFM2rexiXSWDSIQC0jI4P169ebXfXi4+NZv349Bw4cID8/n6uvvpo1a9bw2WefUVhYSGJiIomJieTl5QHQo0cPRo0axR133MGqVatYtmwZ999/P9dff706PtYwj0DtGNibDXUdqOuMWmEOnFzn2E7ZAAm/Fm1vIs1VicEbN8ILD4zE19e31MdZArXooZqnJiLSyKWnu1ogmoFaI1tHDUoI1JzvXweXHaj5+vrSpUsXAHbv3u3orFwUkJ3McE2TCAum1DlqAQEBln1n23+RxqhBBGpr1qyhf//+9O/vSK0/8sgj9O/fn0mTJnH48GHmzJnDoUOH6NevH7GxsebH8uXLzWd89tlndO/enQsuuIDRo0dz9tln8/7775f0klJNigdquflwpPN7rgCmrgO1jL2AW5Zvx9uO/9pzLIFaeDClvsPn5JynlpqaiuHXBKJOd5zQPDURkUbpVM+oAeXKqIGr/DE/P5/4+Hgzo3Y01XVNywjMuWveDBw4kNNPPx1/f39+/PHHcr2uSEPVIBa0Gj58OIZhlHi+tHNOUVFRzJo1qzqHJeWwadMmj2O7DybROiAScpPqvvQxfZd1P2GuI6jKPGgJ1Mp6h8/JmVGz2+1kZGQQ1mI4JK1ynNR6aiIijY7XOWqBzbxf3ICVmlErZ6DWrVs3c3v79u10aer4vXrMrXdXywhK/X3r6+vLqlWryMrKokmTkgM6kcagQWTUpGEyDMNRAljMnj17XGUhdZ1RKx6oAWyfDBm7Sa9ERs19QWyPhiKapyYi0uh4ZNQCosDX+1IuDVl1ZtSgqPNjsYyarw9EhVLm71sfHx8FaXJKUKAmNSY/P9+jPT84atPNNWbyU1wNPOqAPXUXufnFDu77D2TEe5Y+ltKFyskjULPMU1PnRxGRxsYSqIUAwSV3k27Iunb1XJ+sohk1j86PRZkzZ6AWHQ4+PpTr963IqUCBmtSYzEzXipid3H6G79mzx1W/b9ihIKOWRwZ5eXm89tprRF/0EbH3wbbDQIfxjpOFOWAUVGqOmkeg5h8OTTo5DmQdqq7hi4hIPeHRTCSocQZqrVu35umnn8bf3x+A5mEQ55yKV85AzT3Y27FjB/g3wTBcpY8tnYGfX2g1jVqkYVOgJjUmKyvL3O7ZylHSAEUZNfeOWHVQ/vjXv/6VJ554guT0Ak5mwodLguC058Dm+l/CMkctiHLNUbMsep1a9BahM8AryKjT7KGIiFQ/j9LH4Ni6G0wNe/nll9mzZw9THujKgqchwNnpILB5ue4PDw83u207M2opWZBfVHzTIhxHNs2mP09FQIGalCA+Pp5t27ZV6RnuGbWIYGjX3NFCd/fu3Rj+roCmLhqKrFy50rK/YJsPNGkPra8wj6W71t+sfEYNXCUcRiHYcz3uERGRBsowSDu2x9wNC6LRZtSc2rRpw33X9uO0tm4Hfcrfm85Z/njixAmS0m2eHR9V9ihiUqDWGBTmOEoIq8nevXvp2rUrvXr1YunSpZV+jntGLTQQOsc6yiXS09M5keU20boOMmpmtqvI+j1ZnDhxAro/bB7z7PpYiTlqFLsvv/bLPEVEpIZsepa0xC2Ao2okOIBGnVEzVaGrpXvnxx0H0znm0Zq/7DdFRU4VCtQauuMr4H/R8MvpUJhXLY9cuXIlBQUFGIbBzJkzK/0c94xaSCB0igsy93cfKXBdWA8CNYBFixZB87Mg6gzAFagF+ReVd1Q1owZ1Mh9PRERqQMZe2PpPs/oiPBhsNhp9Rg0An4CyrymBpaHIvmSOurXmN0sfRQRoIOuoSTE5x+HECsf6XFtedhw7uR4Oz4G2V1f98Tmumr/58+dX+jnFM2qtm4cBjp/IexJyGNKq6GQdlD56C9SmTJlCbm4u3eMepluTf5FWsB1IcWTToFy/PLzPUVOgJiLS6Kx7HOx55pt65mLXjbTro4Xh2dG5vCxrqcWfoK3rTwVl1ESKUaDWEG1/E7b+0/N43slqeXxurmse1b59+9i7dy8dO3aEE3/AytshbjT0f63M51gyagHQuU0kcBiA3YfSwRmo1XJGraCgwBzbaW1g2xEoKITFixezeLFnC33zl285JjcroyYicgo4uggOzgbwEqidAqWP7oFaBRt/WNZSi08gKNh1zpFRU6Am4qTSx4ao2Znej+d7Zokqwz1QA7es2q9DIHULbHsdsg6X+ZziGbVObVw17XsOpbgurOVAzb1DV1xTuPuC0q8PDwY631muZzdt2tTcPn78uGNDgZqISKNxMukExpq/AJBfANlFsw7CnNX9p0LpY4cJru3+b1Xo1jZt2hAc7IjOtu8+qGYiIqVQoNYQlRSoZR2plse7lz5CCeWP5Qg4is9R69jOtc7K7v3HXBfWcumje9ljRDC8Ox72rJ7NnDlzeO211xgxYoTl+vA2Z8Gg/yvXs+Pi4ggKcvy23r59u+OgmomIiDQKr7/+OlHNo7l80kYMA9KD+5rnzIxaQFPvNzcmzQfBkE+h/xvQ5Z4K3erj42Oup7Yn/hCHkl3nVPooYqVArSEKiYOQ1p7Hs6snUCueUVuwYAH2gnzrRQVZlMU9UAsNhOAmTWnVylHvuGd/guvCWs6oWQK1EMd/O3btxdixY3n88cd55513LNeHRUSV+9m+vr5mWcfu3bsdX0svGbUtW7ZYMnsiIlL/ffjB+wDMWQtz/oT0js+Y51xl8rY6GFkd6DAOejwKvhVvLOL8PVlYWMiKXa7j0Sp9FLFQoNZQecuqZZddjlgexTNqSUlJbFg1z3pRQSZlKV76iF8YnTt3BuD4iWTSnKeraW5deXkL1AiINI/16NHDcr0zQ1ZePXv2BBy/gHbu3OnRTOTjjz+md+/e9OzZk7y86unUKSIiNSs3N5fde1xrpv3t23BO+nY298ODvd0l3rjPUztZ9OdE01Bnh2WVPoo4KVBrqLwGajWTUQOY/8t31gPlCNSKlz7iH0anTp3MY3uc1Y91XPoIgNsC3DabjV69epn7ZgljObnfu3XrVo+M2q233grA4cOH+e2n/1bo2SIiUjd27NiB3W6Y+1vi0/j3v/9t7itQKz/3zo9O7ZoXbSijJmJSoNZQNRvkeSz7CBiG5/EK8hqoLfzdeqCwEhk1/3Azowaw+2jRt19dlz76BoNvoOWaSy65xNx2lmuWlzOjBo4Sx9LmqPn8cRMc+blCzxcRkdq3detWj2Pvv/++uR0WDJz1WS2OqOFyz6g5DexQtKE5aiImBWoN0J49e/jqt/1MXwSzlkF6UWtgCnOqJTvlXvro4+P4Flm6Zie57tPUKppRCwD8rBm13SeKSgrrOlBzK3t0evrpp+nQoQMhISE899xzFXp+WRk1d4F+Buz/skLPFxGR2uceqDUNdcxDM9zeHA0f9Cy0v7HWx9UQOZuJuDvdGaip66OISYFaA/TLL79w3bjx3P4BjJsGD850O1kNnR/dM2pDhgwBIDunwDLht1Jz1PzDLBm1PceLlvHLPVGl8VaUR+mjf6THNREREezatYuEhAQGDx5coed37NiRwEBHhm7Lli1lL3iddbBCzxcRkdq3ZcsWc3vaneEe58ObeWnyJV6FhobSpk0by7GBHYs2/D2/tiKnKgVqDVBYmLUsYNk+t1bA1TBPzT2jNnr0aHN7/ma3iyozR614Ru1oUWesgvRydZGsLuXJqIGjg2N4eMV/Ybh3fty1axd5dreOWMUCtdwCIPNAhV9DRERqlzOjFuQP15wdwVVXXWU5X5nfF6ey4uWPp3UMh5A20PK8OhqRSP2jQK0BGjRoEG+++aa5n5zm1jmwGgI194xaVQI1b3PUIiIiaN7cMWN4T6LbuHOOVnq8FeURqHnJqFWVpfNjvNuaccUDtXwcGbVqmFsoIiI1Izc3l127HGUl3ePA1z+IF1980ZweAJ5vokrpWre2ZiADrzkEl8ZrjpqIGwVqDVD37t155JFHOOOMMwA4mZaN3V50spoDtQ4dOtCzh6OWfPVeSHXGXpWZo1b0w9eZVTt0LJtsZ6yWnVjlcZdXeTNqVWGZp7bTlTEz8tIt1+UVAPZcyD1e7WMQEZHqsWfPHgoLCwHoEQf4BtKjRw8mTJhgXtOuXbu6GVwDFRzsapPZvHlzx98IPr51OCKR+keBWgMWFeVYiNlut5PiDKCqufQxKCiIEWf3c7yOAYucc6krkFHz9wV/P8xAzX2emplsyqn+QC0zM5Ovv/6axETrsz3mqNVAoGbp/Lh9t7mdlWld5Dq3oGhD5Y8iIvXWyZOu9T5jIwEfxzzkyZMn8+ijj/Luu+9afu5L2e677z5sRYuDf/jhh3U8GpH6ya+iNxw8eJBt27YRHBzMOeecYzl39dVXk5ycXOK9b7zxBgMGDKj4KMWrZs2amdvJGRDVBMg6VOXnumfUAgICuGBQW/71gWN//ha4bCDlas/vzKiFOjvf+3kGaruPQs/W1Ejp4+23384XX3xBr1692LBhA76+jnfqaqP00ZJR27YNeodCQSZpacUCNWcnzawD0GxgtY9DRESqzr2UPzgAc0mXsLAw3njjjToaVcPWs2dPVqxYQUpKChdddFFdD0ekXqpwoHbTTTfx+++/849//MMjUFu2bBnHjh2ztKt1stlsPProoyxcuLDyoxUL90AtKdNGZwzIOlzl5zoDtYCAAGw2G2d2DTLP7XVmwCqQUQtxBmpFnZysDUWKNmqg9HHlypWAo1PXnDlzuOKKKwBXoBYaCH6+1EhGrWPHjgQEBJCXl+foFHZDE0eglu6lmQgooyYiUo9lZ2eb2yGBmBk1qZozzzyzrocgUq9VKFBbv349S5cupU2bNjz66KMlXjd+/HiPYz/++CNLlixh8+bN9O7du+IjFQ/O0keApNymQDJkVz2j5ix9DApyBGjhNlczjCxnsq0Cc9RCAwEff/MdyPbt25vXHHImYGug9PHYMde433nnHY9ALSKk6GQNBGp+fn50796djRs3Ojo/Gm0JAFLTrF+3PAVqIiL1nntGLSQABWoiUisqFKjNnj0bgLvuusvS6ai4GTNmeBx77bXXePLJJ/niiy946aWXKjhM8cZS+pgXiSNQSwR7viMwqiRnRs25FlhQwSFsNkdjwsxKBGrOxa6d4uLizO0jzrL/ag7UsrOzLc1MFi9ezLo1f9A/ZDGpJx2NOyKc85hroPQRHGUdGzdupKCggF3H/OkVAWkZ1mUILKWPIiJSL3mWPgaVfLGISDWpUDORFStWYLPZGDVqVIVfyJnN+OOPPyp8r3hnyajlOBdVNiA7oUrPdWbUnIGaLfugI9ii/IFaQUEBeXmOlo7Oxa6dYmNjzW0zUKvm0sfjxz27KL7z/G0Urn2SjCzHuGoyowbF5qkV9XhJzci3XKPSRxGR+s9S+ug2R01EpCZVKFDbtm0bNpuNfv36VfiFOnfuTEBAANu3b6/wveKdZY5altu7e1kHq/RcZ0bNWfpI5n6zIUhWnnOR6tIDNUuZSCDg18S1HxJCZGQkAAmpRa14qzmj5i1Q+/yXbexyexkzUKuhjJp7oLblgCMiS8u2XuPKqFXt30xERGqOx+80lT6KSC2oUOljcnIyERERJZY9XnbZZZaOeu5sNhvh4eGWFrdSNZbSxyy3f8oqdn60lD7mpUB+mtkQpLwZNY/Frv1CLefj4uJISUnhyEk7hgG27ERHbWVRq96qcg/UQkJCyMrKIq8A/vm96xqz9LGGMmrurZq3HnRkKYsHauYctZxEKMzVu7QiIvWQe0YtWBk1EaklFV5HzVnO5s2///1vPv/88xLPZ2dne+0IKZVjKX3McPu6VjFQszQTydwPuFrsm4FaGe353eeHhQYCviGW887yx6xcwxG82HMh33uQXxnugdp9d9+Ob9F3+sdLXNe0dn75aihQ69SpEwEBjprRLfsc3R5TrVPUXKWPUC1LK4iISPVTMxERqQsVCtSaNWtGVlYW6enpFX6htLQ0MjMzLVkgqRpL6WOa29ynKvzBX1hYSGFhIVCUUSuaO2WWPuYaGAYVK30MAPysgZrXhiLVOE/NPVAbEHuCq87wvGbcUBwlmQFNq+113fn5+dGtWzcAdh5IJ7+glNJHUEMREZF6yqP0URk1EakFFQrUunbtCsDSpUsr/EJLljhSGc4/XKXqIiIizEWck9NyXCeqEKi5L3btCNQcGTVnMxGA7DygMAfshSU+x5JRC8Ijo1bpzo97P4GVd5S5Xpx7oBad8TMPXWw9P7B/bwacdzMM+aRKHTLL4pynVlBoZ1eil4yaLcK1o4YiIiL1kqX00R/wUddHEal5FQrUhg8fjmEYTJkypcIvNGXKFGw2G8OHD6/wveKdzWajaVNHNijppFuWswqBmrPsEYpKH7OspY/gXv5YLOpwU6mMWs4xSpWdCH9MgD0fwsqJpV5qCdSCTjK4Mwzq4cqc3XXvX+CsmdDmytJfs4os89QOe5mj5hPp2lGgJiJSLymjJiJ1oUKB2q233oq/vz9z587lvffeK/d906ZN49dff8Xf358JEyZUdIxSCmf5Y1JSMgS1cByswqLXJWXU3AO18ix6XdYcNUugllK0kZ9W+uDS3DqGJswt9VJLoBbm6FHy1huvERERwdChQ7nxxhtLf61qYun8eMhL6SPhrh2VPoqI1Ese66hpjpqI1IIKBWqtW7fm3nvvxTAM7r//fh544AESEkpesyshIcG8zmazce+999K6desqD1pcnA1F0tPTyfdv5TiYnQD2glLuKpl7Rs1S+ugto1bOQM3Rnr8cGbWymonkJRfbL7mDqHug1jwMaHkBQ0ffTkpKCr///jshISEl3ludLBm1xCBSPQI1t26YyqiJiNRLWkdNROpChdrzA7z22musXbuWpUuXMm3aNN5//30GDBhAv379zKAhOTmZ9evXs3btWgoKCjAMg2HDhvHaa69V+ydwqrO06C+IpiWAUQg5RyGkVYWf555Rc3R9LGomEhoCON5RLE+g5h7AR4dRYtdHgISUoo0yMmqFGYe5/h1YthO+uB/OTVoDsRd6vdYZqEWGgL8f0OupUp9dUzp37oy/vz/5+fmsOxKJf751Hl6uEQq+wVCYrbXURETqKa2jJiJ1ocKBmr+/P7/++iv33XcfM2bMID8/n1WrVrFq1SqPaw3DwGazMXHiRKZMmYKfX4VfTspg6fyYF+UI1MAxT62KgVpggJ/Z4CO0SQTOQC3LuUJDKYFafHy8ud0hGo+MmnugVt6M2tSPf+C/Rd9m7/0G516/usRA7dgxx3y36HAcXR1bnl/qs2uKn58fAwYMYOXKleza59ksJS+/AELbQtoOR+ljNa4lJyIi1cMZqPnYwN8XZdREpFZUeB01cJTEffjhh2zYsIG7776bzp07YxiG5aNz587cfffdbNiwgQ8++MBRRifVzn0ttZN5bh0EK9lQxNJMxNfVOz40zBUQlmcttX379pnb7aPxyKgFBQWZYz9sBmolZ9QKCwt5+t0F5v78zUCS55sD4Ag209Icz4oOAwKa1Wnw88wzz5R4Ljc3F0LaOnYKMkst5xQRkbrhLH0MCSz6daKujyJSC6qU4urduzfTpk0DoKCggORkxxyiqKgoZc9qSVhYmLmdYQ8D36KdSgZqloyajytoCwlvbm6Xp/TRmVEL8IPYSDwyagAtWrQgOTmZJGfDyhIyart37+bRRx8lM9s1725IF+DESq8ZqBMnTpjbZkatDl1yySVccMEF/Pbbbx7ncnNzIdTVcISsAxAY5XGdiIjUHWdGzVyqRhk1EakFlcqoeePn50eLFi1o0aKFgrRa1KRJE3M7Pd+tMUUlOz9amonYXDX5oRFmUaVb10fv7fkNwzADtXbNwccHj4waQGRkpGPcOVBQCOR5D9TGjBnDnDlzLMfyCnCUZXrJQLk3EmlRDwI1m83G1KlTLQ1UnBwZtTauA2ooIiJS7zgDtWBnoKY5aiJSC6otULPb7Zw4cYITJ05gt9ur67FSBktGLT/YdaIaMmpBNrcW+01dQUZZGbWkpCSz62OH6KKDXjJqzjXgoKhtvZfSR8Mw2Llzp+c4nVWZXhbJdi+7jGtKnQdq4FjofdeuXcyYMYMnnnjCPJ6Xl+cqfQQFaiIi9ZBZ+qiMmojUoioFalu3buX++++nV69eBAQE0LJlS1q2bIm/vz89e/bk/vvvZ8uWLdU1VvHCPaOWkefvOlEdpY+4AqeQSFdjkrICNfdGIu2dgVopGTWAk5l4LX3My8vDMAwAzuzsdtxZBZntGaht3rzZ3O7ZinoRqAGEhIQwYcIE/vGPf5jNVBylj26BmtZSExGpd8zSR2d8poyaiNSCSgVqubm53HHHHfTp04f33nuPbdu2YbfbLc1Etm/fznvvvUefPn24/fbbLSV1Un0spY9ZuRBQNL+pGpqJBBquwCk0qr25nVVGMxH3jFZpGTX3QC0lC68ZNfe1ayKCwa9oDl6uM1DzklFzD9R6t6beBGrunM11LM1EQC36RUTqmfz8fAoLCwG30kdl1ESkFlR4MllOTg7nn38+K1euxDAMfHx8zHXUmjd3NJw4ceIE69atY926dRiGwYwZM9i8eTMLFy4kODi4jFeQirCUPmZkQEhrx8LQ2YfBsIOtYrG4pfTRKJr/5R9JaES0edzMqJXQpdFrRq2sQK2EjJplkdFAR3OSgsLSM2rOLK6/L3SJAQIiPa6pawEBjt/2jkDNbRF4lT6KiNQrljXUzEBNXR9FpOZVOFC7++67+eOPPwC47bbb+Pvf/067du28Xrtv3z5efPFFZsyYwerVq7nrrruYOXNm1UYsFpbSR2eglrIR7PmQcxyCW5ZytydL6aOR4tgIbUtIiCvQMgO1XFd3RXdeM2pllD6mZOFY9NmeDz6uEk73X5DB/hDo58jolZRRy8/PZ8eOHQB0iy1a7LoeZ9Ty8vLALxiCWkDOMZU+iojUMx6LXYNKH0WkVlQo3bJ27Vr+85//4OPjw4cffsiHH35YYpAG0L59e6ZPn84HH3wAwGeffcbatWurNmKxsJQ+pqdbszOV6PxoWUfNz1HqQWg7QkNdHSXNBa9zjuONx2LXUGYzkZPOKspiWbps90AtwJFRA7eMWs5Ry/W7du0iP9/RaaS3s5liPQ7UzMDYWf6YfcQRrIqISL3gXtkR7HwfUaWPIlILKhSozZw5E8MwuOWWW7jtttvKfd/EiRO55ZZbMAxDGbVq5lH6GOwWqFVinpolo+bMtxYL1MyMWs4xr8/Ys2cPACFBvo51zKB8GTWwlj/mJpE9/zJz12ugVqz00X1+Wi9n/5N6HKjZ7XYKCgpcDUUMuyNYExGRekEZNRGpKxUK1BYuXIjNZuPBBx+s8Av95S9/MZ8h1cej9DHUbU2uSgRqlmYizncOQ9tZSx/ziqKlXM+MWkFBgZlR6xwX7FqLujxz1MCaUds3i+xkV3YuOMAVPJbUnt/SSKQeZ9Scc9RAa6mJiNRnWcUqOwBl1ESkVlQoUDt06BD+/v7069evwi/Ur18/AgICOHSoct0IxbtSSx+rmFELcgvULKWPBUUnvARqBw4ccGSIgM6xrmAEX88mMl4zau6LXp9cT7ZbFWCIe0atsCgCLBao7d2719zu7lz6rR4Gas6MGmgtNRGR+szS1EoLXotILapQoJaZmWn5g72iQkNDLe9MSdUFBASY2ZlqL310BmohbQkKCsJWlB7LzCvqkZ+fBoW5lvt3795tbneOKbrON8hr98ky56ilbCA7z7UbHOAak6uZyHGwF5jXuGcEmzh/j9bDro/ugZrWUhMRqb+8dn1UoCYitaBCgVrz5s1JSUmxvLtUXtnZ2aSkpNCsWbMK3yulc2bVHF0fXQtTV7n00W2Oms1mM8sfM3NtrhuKZdWsgVrRhpf5aVDGHDV7AaRsdq3ZRtEctaKgNC/fwLEOtmEZg7ORCBR1fMQG/hFeX78ueQRqWktNRKRe8ih9tPmBj2/dDUhEThkVCtR69OgBwM8//1zhF/rxxx8xDIPu3btX+F4pnbOhSHp6OviHuQKT6ih99Al0tI4HM5vqHjwVbyjibCQC0LmF3bHhZX4aQESEK4ByzVErCtTSd4I911L6GBxgDXDyC51jcHV+dA/UAvxwfC0quJZcbfCYoxaq0kcRkfqo+Hqemp8mIrWlQn/Bjho1CsMwmDRpkiXzUpacnBwmTZqEzWZj9OjRFR6klM6SUQPXPLXsQxSlncrNo5lIaFsz0HEGapk5hW43lJJRiy4KmkrIqPn7+5vPdGXUikofT25wfArupY/+EBDoepa3zo+WjJov9XJ+GnjJqAW1cK0fV4kAW0REaoZH6aPKHkWkllQoULvjjjuIiopi27ZtXHrppSQnJ5d5T3JyMmPHjmX79u1ERkYyceLESg9WvHMGapmZmdjtdlegVpgDeWX/G7nzyKi5leSZpY85BW43eA/UAgMDaRVR9C5kCRk1cM1TO1k8o5ayEcBzjlqQ61neOj/m5bluaCiBWl5eniMY9o90HHBfokBEROqUR+mjMmoiUksqFKiFh4fzzjvvYBgGv/32Gz179uSll15i69atHtdu2bKFF154gR49erBgwQJsNhvvvPOOpdxNqof7WmqZmZlV6vyYm+Vq5uHIqLkWNDdLH7PzXIk6t9JHu91ulj526tQJH0rPqIFrnlpKFqzZCz2u+D9uv/12jOT1gDVQCwmEgCBXl8tGlVEDV8lqsUW/RUSk7nh0fVRGTURqiV/Zl1iNGzeOxMREnnjiCY4dO8azzz7Ls88+S0BAgCs7cvKkmdkwDAMfHx9effVVbrrppuodvQCea6mFFe/82LRvuZ+Vc8K1DlmgH14DNXAEUCGBWDJqhw8fNoOOzh3bA0UBfCkZNWeglp0Hr/8A2/elsH36dB7sFU2flp6lj4Eh4ea+q/OjZ6Dm6wM+PtTbQM1jjhqAf9Hnlp/mKFm12bzcKSIitcljwWvfoLobjIicUirVZeHRRx9l3rx5DBgwAMMwMAyD3NxcEhMTSUxMJDc31zzev39/fv31Vx5//PHqHrsUqa611AzDYNteR+AVEgghwf7Q5grzvPui12apolugZpmf1tFtAedyZNQA/tznOr5hp+O5WcVKHwOCXBlZbxk15xsE/s6GXPWwNT+UlFErCtSMAkfZanXI3A9/PgyJ86vneSIipxj3jFqwMmoiUosqnFFzOu+881izZg2rVq1iwYIFbNmyhaSkJACaNWtGz57/3959h0lVnn0c/87szs72yhbaIk2agESKICgKYhcjGsVYE2NDjfGNSTSWiCYaYoslscTYULEXYiIogqggTZDe67INtvednTnvH2fmlCnb2N0pe3+ui8sz55yZOXtYl/nt/Tz3M5xp06Yxfvz4DrtY4Z9x6GN1dTXktC+o7dy6kfxStVHIqUMgalYB2PXlFAYPHqxtP/cF/OUyTEMfTUGtv2GZgFbMUQPYozdvZKO7Q30daUAZ4J6jlqCfr89R8+36qAW1EGzND37mqIEe1ECdpxbtu0h4my07Hyo2w+4X4NJKsLb7f3khhOiWfJqJyBw1IUQXOeZPbePHj5cwFmTeQx9NFbW61ge1Jf97V9s+Y8pPTCEN4I477uAf//gHjY2NPPU53DoDemUEqKj1ywbPLyFbWVEz2uQJatG90IKaPYqYOENFjTigzu/QxxjPd3aIDlFptqIG6vDHuByOiaKoIQ3AWQdN1SFbYRRCiFDl00xEKmpCiC4SegtMiTbrqKGPX321RNuedtrJPsf79evHLbfcAqhzx974lsBDH3Oz9Ce2Yo6aNy2oWXpo++KSs7HH6sGrwZLuPsnP0EdPULPqc8FCSbNz1KBjGopU7zE/djb4P08IIURAPs1EpKImhOgix1xRc7lcrFu3js2bN5uGPo4YMYKxY8ditUoW7Gw+Qx9tKRCdAE01UHuoxeeXlZXxpz/9iQ/+twqAtAQYPW6q33Mvu+wynnrqKQDyy/Ab1Gw2G31zkmCH+0AzFbWsrCy/+/PLoKTaQp1TD2bxfSYTs0cPOI1RacBhcJSrc7qiYn2HPoboP6jNdn2EjglqJavNj10S1IQQoq18molIRU0I0UXaHdRKSkqYN28eL774IpWV/j9UJicnc8MNN3DXXXfRo0cPv+eIY+cz9NFigZh0Nag1trwm1xNPPMHTTz+tPZ46DKJSh/s917i8QkUt4KgCxYWCRQtq/fv3JxpDKGimojZu3LiAxzaV9KW2Tn+duEnPErPyWe1xgyVVP7m+CBL6+Q59DNF/UFueo9b+oNbU1MSuXbvY9Ml8nHlw6QSIjkIqakII0Q6yjpoQIljaFdS+++47fvazn1FYWIiiLajlq6Kigscee4z58+ezYMECpkyZ0u4LFYEZK2pVVVXqRpS7EYWzzs8zzFav1isvsTb4w4UWSBrk91xTUKsDUMBRRWFJrfaP2aBBg8Cp/8PWXEVtzJgx2O1200LbHhuLepiGnMQmZpgDjsUQbOoKIaGfb9fHEB362Ko5am3Q2NjIgw8+yP/+9z+2bt1qup97i+GPFyEVNSGEaAffro+hOfdZCBF52hzUNmzYwNlnn60urIza/fHiiy/mxBNP1KpmR48eZf369Xz44YcsW7aMgoICzj33XJYvX86YMWM69isQvhU1aFNQ8yxYHh0Fta+AJWlgwN8Y+lTUABwV7N59QNs/aNAgtZrn0UxFLSYmhrFjx/Ldd9/5HNuUF6X9A2m327Farea5XegB1dP5MVyGPnb0HLWFCxfyl7/8xe+xbfnuDQlqQgjRZp5fQtpt6hqdofrvihAi8rQpqLlcLmbPnk1NTQ25ubm88cYbfqtkQ4YM4ZRTTuHWW29l+fLlXH311Rw8eJArrriCLVu2yLy1DtZsUHM1gssJ1ig/z4TKykry8tSGI+MHuNdYTh4S8L3i4+OJiorC6XTqQa2x3NxIZNAgaCzXn2RLbfb6J02a5D+o7a2krs4FQFyc+vWYKmqmoKY2FAnHoY/+g1rLQ1aN9uzRG4ccd9xxDBvYk/8tWQkY1puToY9CCNFmnqAWZ3PvCNF/V4QQkadNientt99mx44dZGRk8PXXX7dqKOOpp57K0qVLycjIYOfOnbz99tvtvljhn9+hj8Y1uFyBF0/evn27tj3cs/SZsWukF4vFolXVKjzFOkeFb1AzBo0WWsJPmjTJ9HiAu7/I5p2HtMqtJ6iZK2oJ+pPq1GG4TU1qKgmnoY8dMUettLRU2/73v//NK/eepj3W1pvrqEW0hRCiG/GM7Ij3/NiOCc31OYUQkadNQe3jjz/GYrHwwAMP0K9fv1Y/r3///tx3330oisJHH33U5osUzTNW1Coq3AEpyhDUmgIPf/QMewQY7sln0Qn+T3bTgppp6GNzFbXm/1GbOHGitp2RCKNy1e2amlry89Vxe/Hx6vBJU8BRDEMq6wu1ahqE/tDHju766Om4CpCeno69/BvtsVZRk6GPQgjRZp6KWrzn9372zOBdjBCiW2lTUFu7di0Al19+eZvfaPbs2QCsW7euzc8VzcvJ0RdGLigoUDeMQc1ZBy4H5H0K1ftMz922bZu2PayXe6OZ5h9gDmqKAjTqQS0qKkoN8Y5y/QktVNSys7N54P77GdrLwtu3wsi+vuf4q6ipC1671XkFtRBfRy3WsB7cvn3uv5MOqqilJ8cSU6E3iGmQoY9CCNFu2tBHCWpCiC7WpqBWXFxMUlJSu1rtZ2ZmkpycTHFxcZufK5qXkpKiDX88dMi9bpp3UNv2OCyfCYsmmIbAmSpqnqGPrayoOZxQ7wDFMEetX79+aphqw9BHgD89+CDb/jmQM0c2H9RMlSinTT/Bq6IWow19DM2K2ogRI0hPVxfs/vDDD9m0aVOHBbUM13bsUfq90IY+SkVNCCHaxOVyUV+v/pupVdRiJagJIbpGm4JaVFSUNgeoPZqamoiK8t/UQhybPn3UcYt5eXnqkgneQe3Hu9XthiNwdKV2yBPUEhNi6Zvh3tlMl0bw7fx4tDhfW0tv0CB3W/82NBPRxKtJcVRrK2pNLnW9OID6In2uF4aKWogOfYyPj+fuu9W/E0VRuPfee81Bral9Qx/tdjtxFd8QZQWrRT3W6HSfJBU1IYRoE09IA8McNamoCSG6SJuCWk5ODnV1dRw4cKDlk70cPHiQ2tpa0zA90XH69lXTTV1dnVpd8Q5qRopLO9cz7G7ogGy14yO0uqIGalDbvcerNT/oQc0S1eLraRKOU18jB2JjzQHLb0WtoQHi3N9PdYU4jEEtxJuJAMyZM4fevdVw+umnn7Jy9Q/69bazopaeno6laAmgtpIGqagJIUR7+ayhBlJRE0J0mTYFNU/Th9dee63Nb/TKK68AcPLJJ7f5uaJlnooaqFU1U9dHn6Cmllh27NihLVg+3NNqEVo9Rw3Uzo+79+drj7Wg5hn6aEtBT4AtGHI7pIwg6vgbGT58hOmQ34paYyPEuoOasxZHXbl2LNTb84P6Nd1///3a43vuuQcl2t3Bs7Ft7fk9QS0jPRXKNgBgd4//lPb8QgjRPp75aSDNRIQQXa9NQe3SSy9FURTmzZvXpqYga9euZd68eVgsFi699NI2X6RomaeiBu55aqaKmldb9iZ1rTXT/LQB6frxNg593H1Qn3eoB7Vy9b+tmJ+mSf8JnLcZxj/PqFGjTIf8dX1saGjQgxrQWKUHxlDv+uhx3XXXafds2bJlfLHF/UmgqZLGxkYWLVrEkSNHmn2Nuro67be+6YbiZUyM+rU3SNdHIYRoF2NQi4tB/UVmC/9GCiFER2lTUDv//PMZP348tbW1TJ8+nfnz57f4nDfeeIMZM2ZQX1/P2LFjueCCC9p9sSIwn4qa19BHl8twsntYnanjYz9Da/i2Dn08qDeyGDhwoNoK0jP0sYXW/IGMHDnS9DhgRS1OD2qO6kJtOxyGPgLYbDYefvhh7fE988vVTpqOSv70wAOcffbZTJo0CafT6ff5DzzwADNmzNAep8fpHyrsserfo6yjJoQQ7WMc+hgfgwx7FEJ0qTYFNVAXvc7KyqKiooJrrrmGwYMH87vf/Y63336bxYsXs3jxYt5++23uuusuBg0axLXXXkt5eTk9evRgwYIF7brI5cuXc8EFF9CrVy8sFgsff/yx6fiHH37IjBkzyMjIwGKxsGHDBp/XqK+vZ86cOWRkZJCYmMisWbMoKipq1/WEouYqajf87ll63ASfeoqg7qBmqqj1M4Sztg59zFMX2bZYLAwYMACctdrwyjZV1AwCBTXfilq29thRo/99hsPQR49LL72UE088EYB1u+v4YDXgcvDIo48CsHv3br2bp8GqVauYO3cu3377rbYvw+ZeT81qJyZOHUYp66gJIUT7mIY+2pFhj0KILtXmoNa/f3+WLVvG8OHDURSFPXv28Pjjj3PllVdyzjnncM4553DllVfyxBNPsHfvXhRFYejQoSxbtoz+/fu36yJramoYPXo0zz33XMDjkydP5q9//WvA1/jNb37DwoULee+99/j666/Jz8/n4osvbtf1hKJAFbWCMnhpwVLKamDmE+4TvIKa3W6nf3a0/mJtrajl12nXEBsb276Oj168hz62OEcNaKzWg1q4VNQArFYrf/nLX7THD33sXp/O4Ghhns/zNm/e7LMv3a4OayVzMna7ulabrKMmhBDt4zP0UYKaEKILRbd8iq+hQ4eydu1aXnrpJf75z3+yffv2gOfdfPPN/OpXvzIt8NtWngAYyFVXXQXA/v37/R6vqKjg5Zdf5q233uKMM84A1OYmw4YN4/vvv4+IBieBKmpV/ka7OdT5T561z4YMGUKUYmg40kJFLTlZbyO//wiUVqnVM7+t+dtZUcvOziYzM1Obn+UvqJm6PgKOGn0uV7jMUfM4++yzGT9+PKtXr2bjQfjG63+poqW/hhM+gMTjtH0JCb6BOj3RvZEzHbv9XUAqakII0V4y9FEIEUztCmoAsbGx3Hbbbdx2220UFhayZcsWbS2njIwMhg8fTs+ePTvsQo/FunXrcDgcTJ8+Xds3dOhQcnNzWblyZUQEteTkZJKSkqiqqnJX1NRgXNfo52RHJbt379bWxBs+fLg6XNGjDRW1dfv0/T4dH6Hdc9RAHf741VdfAf6HPnpX1By1R/W3DaOhj6AOG73hhhtYvXo1APd/YD5etO8H+LQ/ZJ0K/a+Gvpdoa9cZZRiCWkzMx4Aa1BQFLFJRE0KINpGhj0KIYGp3UDPKyckJ6fXRCgsLiYmJITU11bQ/OzubwsJC/09Crdg0NOgfbv19MA4lffv2ZevWrRw6dAjFGouFwBU10/y04cOh6Tv9eBu6Pq7br+/vyIoaqMMfPUHN0/WxuYpaY40e1PQ5arZ2v39Xu+yyy7jj9jlU1zbw9TbzsUJP9i1erv5ZeyuVPw71eY30RCAmDdLGmEKtwwkxUlETQog2MQ19tCEVNSFEl+qQoObx/fffq1WOAE4++WTTB+1Q98gjj/Dggw8G+zJarU+fPmzdupX6+npKqxxkAJV1fk50VJk7Pg4bBk1f6Mfb0ExE6yiIn9b80O45agDjxo3Ttj3VWZ+KWkyGuqi24sRRp3eftEWhzk9r7RpuISAxMZHLL5/Nv/79qs+xIttESC6Fyh3qDmc9lQUbfM7LSASyzwBrlDnUOiBGKmpCCNEmpqGPUlETQnSxNge1v//977zzzjtMnDiRxx9/3HTspz/9KcXFxQGeCQ8//DB3331326/yGOXk5NDY2Eh5ebmpqlZUVNRsJfDuu+/mzjvv1B5XVlaa5oKFGtM8tcJKMoAqf0GtyU9F7UCN+sASDVHNh2ljUDPyX1Fr/9DHn/3sZ6xZswan08mFF16ovpx3Rc0apf7DWV+Io9Y7qIXHsEejGWef6z+oNeXCed9B6VrY9zrsm09FbbnPeemJQM6ZgFeHzCZIckl7fiFEmFEUWP9bOLoSxr8IqSd06dv7LHgtQU0I0YXa1PWxqqqKBx54gDVr1nD99df7PUdRlIB//vrXv5p+O9VVTjrpJGw2G0uWLNH27dixg4MHDzJx4sSAz7Pb7SQnJ5v+hDJT58cidaycd0XNs0aXJ6hFR0erAcszR60VC3kmJiZitfp+6wwcOFDdMM1RS2319XuLjo7mySef5Omnn9ZCh82mD2XUqrfu4Y+Oev19Y6JpMXCGokDfj0VFRWp1MGMcjH0GBt/ot1qann0cHHcF4NUhswnp+iiECAvFxcVaIymOroTtT6j/3flMl1+LT9dHGfoohOhCbaqoLVy4kMrKSi666CJ1uJwfFouFvXv3+uy/9dZb+e9//8sHH3zAlVde2aaLrK6u1joUAuzbt48NGzaQnp5Obm4upaWlHDx4kPz8fEANYaDPnUtJSeGXv/wld955J+np6SQnJ3PbbbcxceLEiGgk4mGqqOWXQI7vHLV6B8TUV7Bjh3qvBg0apH6gb/IEteYbiYD6d5ycnEx5ebm2r2d2ht6FsIPmqAV675iYGBobG/X5g+6GIo0OfVHocK2o9enThz59+qgNYQx81vyz9/Ab1DLO+RBs6vpppoqaA+n6KIQIedu3b2fUqFFYLBY2bdrE8VXv6wfrA4/Y6Sy+XR+zA58shBAdrE0Vtc8//xyLxaK1ww+kX79+Pn/mzJmDoigsXry4zRe5du1axowZw5gxYwC48847GTNmDPfffz8An376KWPGjOG8884D4PLLL2fMmDE8//zz2ms8+eSTnH/++cyaNYtTTz2VnJwcPvzwwzZfSygzVdQK1A6c3h/maxpg3+FyLeQMHz5cPeB0D31sYX6ah/fwx0FZTshfBIrLPEetg4Ma6AHEp6LWpJ8TrkEN/FfVfIJaTAYVtT6nEZczRts2zeeTipoQIgwsXLgQh8NBY2MjH3/0ERwytMB1VHX59fhU1OJ6d/k1CCG6rzZV1NavXw/AlClT2vxGp5xyCgA//PBDm587depUFO8VgA2uvfZarr322mZfIzY2lueeey7gotmRwFxRK4YxvnPUquth6yH9HzstqLWhogZqY5gDBw5ojwell8OysyHhOLW5h8cxtOcPxDOkz7uiZiiohe3QR1CbqLz33numfaWlpTgcDn3opz3DFMInDoYbLhxoeo5pPl8TUlETIpCmOth0P0QlwMgHwqoJUaQxjsjZuuEb6HtQP+jo+s7LpopaUkbY/rsihAhPbaqoHT58GLvdTo8ePfweby5MJSUlkZycTEFBQduuULSaqaKWry474K+itjVPLz0NGzZMrYI53Se2sqI2d+5c0+O+6Z432A/Ve/QDXVFRi1cDamOEVNRGjx7td7+pUY8hqKXEw4o/wbUXmifZ+wx9lIqaEP7lfQTbHoPND8Lh/wT7aro1Y1DbsnGd+WBT1we12ppqbTs+tVeXv78QontrU1Crrq7W5yH58d1335m6CXqz2WxUV1cHPC6OjbHhyaE8dQ6ad1Crrodt+fpjdbFrw0mtaCYCcPzxxzNnzhzt8bgL7oGeZwOG30RHxUF0xzdg8amoJR8PmCtqtmjU9vxhaNq0aUyePJmYmBiys/X5EKbhjzEZVHiCWpx7nz3L9Do+zUSkoiaEf3WGH4qla4J3HRGqtLSUxx57jDVrWr63poraniJcLsPBIFTUaqv0bsJxqX2aOVMIITpem4Y+JicnU1FREfC41vUvgPLy8oCt3UXH8KyllpeXh2KxUVXvMB2vaYCth9Vti8XCkCFDoMkw7r+VQx8BHnvsMeLj44mJieGcqx+EqCio3g97/w2FX0L/q9T2+R3Mp6KW5CeoRQFR4VlRi4qKYvny5dTU1DBv3jweeughwCuoGZqJJHuCWqw5qHm355eKmhC+1q1bx3MPL+DaE+DUYUDFlmBfUsS55557eOGFF0hJSeHAgQMBPwc4nU7279+vPa5tUDhYAsd5Gi0GY+hjTbm2HZ+W2+XvL4To3tpUUcvOzsbpdJoWS26trVu34nQ6TRUC0fE889Tq6+spqY31qahV1cE2d1Dr378/cXFx+vw0aPXQR1Dn/c2bN4+HH36YqCh3IEs8DkbNhRkrYPDNx/CVBOapFOlDH/tAVKxp6GNMNGE79BHUEJ2YmBiwouawJFLrKSgGCGq+FTVZR00IbxMmTOCVj9dx2sPuHRLUOtz27dsBqKio4KOPPgp4Xl5eHk1NTaZ9W4wNcJtqwOWkK9VW67+cjk/v36XvLYQQbQpqnlb2H3/8cZvfyPPDOZLa4Yci0zy18hif9vzbC9SqGhgbidToJ7ShohYsnkqRNvTRYoWkwb4VtTAd+mhkXJB95cqV2nZVjR6uUzzZ2mshVpmjJkTLnE6vD/5Vu+X/lQ5WU6P/G7NgwYKA5/lb2mdLHpBqmLfb1EGdH2vzYc/L0FDa/GnuOWoWC8Sk9OuY9xZCiFZqU1C74IILUBSFJ5980rddeDMKCgp46qmnsFgsXHDBBW2+SNF6ps6PpdE+FbXVhj4femt+Q0WtlXPUgslTKXI6nfqHrKTjI2boo9Fpp52mzQt9+eWXtTmglZX6EKDWDH1sdCJz1IRoQWMToDihckewLyWiGIPal19+qS9m7cVfUNtakg1Jg/UdbWzRX1/vZySBowq+mAyrroeVzS83VFen/vsYHwOWBBn6KIToWm0KajNnzuT444+npKSEc845x9SePZADBw5w7rnnUlJSwpAhQ7jooovae62iFcwVNYtPe/41hn8HtUXLjUMfw6CiZhzSd9ddd7F06VIcsQMjauijR48ePbj77rsBNZjeddddAKa5olpFrZmhj1JRE6JlntEGMvyxYxmDmtPp9Fl+xMNvRa3ADjZDU6o2zFP75S9/SXJyMr/+9a/NldMNv4eafep2/n+bfY3aOjXoxdvROgwLIURXaVNQs1gsvPbaa9hsNn788UdGjhzJrbfeyqJFiygqKsLhcOBwOCgqKmLRokXMmTOHkSNH8uOPP2K323n11Vc76csQHqaKWolv18dDJfq236GPbZijFixZWXogefLJJznjjDPInPYMj32mnxMpQx9BXeDd8/f63//+l8WLF/uvqNlbaCaiNKlLMQgh/Kr2FF8kqHUoY1CDwMMfjUHN7l4yctu+o7iikvSTWhnUHA4Hr7zyCg6Hg6effppZs2apw+WPfg+7/mk+2c/SQvn5+YwYMYJ9BeovMtXFrnu26r2FEKKjtCmogTrxesGCBcTHx1NdXc0///lPzj33XHr16kVsbCyxsbH06tWLc889l+eff15r6b9gwQLGjx/fGV+DMDBW1PYUNpmGA3rTKmphNvTxgQceYOrUqVit+rdvhVfp0BZNRAx9BIiLi+ORRx7RHv/f//0fZWVl2mM9qGWYnufTTASkqiZEM7QfI+6g1tTUxPLly2VZmWPkHdS++eYbDh065HPevn1qlctiganDPM+t5eBRwy+YWhnUampqTGu7fvLJJ2pYW3uf78kO327Wzz77rGm5oXh7NFjb1ChbCCGOWZuDGsBFF13E2rVrufjiiwF1oWt/fwBmzZrF2rVrmTlzZsddtQjIWFHbejBwl78+PWJIsrtTXJg1Exk+fDhLly7lyJEjvPnmm0yZMsXnnJgwXvDan9mzZzNu3DgANm/ezFNPPaUd09ZR8/oQ4dNMBGSemhDNqPb871G+GYA//OEPnHbaaUybNi14FxXmXC6X33li77zzjs++khJ1yEdaAowx9O3YcsDwb1QrF732DocAn332GZf+8UvTMHkA6n3nzH344YemxzsLvJ8khBCdr92/HhoyZAjvv/8+hYWFLF26lK1bt2o/ZDMyMhg+fDinn366qWud6HxJSUkkJydTWVnJ1gOBfws8vGcjLD4ZTl1orqiFwdBHj/T0dK644gqampr45ptvTMfCecFrf6xWK0888YQWSpcuXaod0ypqXqSiJkTzXC7zUGBt6GP1Hmiq4/HHHwdg9erVOBwObDZbF19h+Kut1f996devnza3fcGCBfz21quhcjtkTgaLVZt7mxwHIwxrS2/dV855nsaPbaioefTp04eSkhLq6upY+AP87Gl4984EYnCf03AE0BuWVFdXs2fPHtPrZaZGzi/+hBDh45jr+Dk5OcyePbsjrkV0kL59+7JlyxZczUxHGtYbtbPZ4gmQdZp+IAwqat5SU1N99kVK10ejyZMnM2vWLD744APT/pQA2dpnjhrIWmpCGDgcDtNjLaihqAHC61wJam1nDEyjRo2iR48erFu3jnXr1rHr9YkMTtwLw+6CMfO0ubcpXkFty56jcAxB7bzzzuNnM0Zw3uW3U++AT9bBZf/qy3vXbSc6CqgvNj13xYoVPuu53XjxyNZ/0UII0UHaNfRRhDbjPLVABh/nrnQ2lkHex/qBMKqoeaSkpPjsi4mwiprHX//6V58Pi4Eqan6HPkpFTQhNY2Oj6XG18fcYXg1FvEOdaB1jYEpISODyyy/XHi/4wt08ZNvfaKgq0tbGTI6DIT3VBmYAW3YX6C/YjqCWkJDAGan/YeFvIdb94/Pjpdv5ZJ37hAbz0Mdly5Zp2//6Ffz4CPzp9gtb9b5CCNGRJKhFIOM8tUAGnvcc9DrP90AkVdQiaI6ax8CBA7n99ttN+5Lj8GkkAgGGPsocNSE0PhU14/8eXkHNO9SJ1vEOTJdddpn2+O0VesPFys3/1vanxEN8rJUBAwYAsG1Xnj5CpJXrqJnel1IoXMz0E+CfN+k/K3886N5oJqiddyKMygWLdHwUQgSBBLUI1JqK2sAhI+HUT9QhJ0bG9WrCRHcZ+uhx7733kpGhf9hITYqBKR/7nOd36KNU1ITQSEWt83kHtb59+zJl8ikAbMuHje6wVLH5Fe089ZdPmYwYMUJ9jdo6DnqWlmlPRa3yO2179PQ52naBp3muoZlIdXU1a9asAWBo/x7kpLoPxMl8eyFE15OgFoFaqqhZrVb69esH1igYMw9OfhVisyHnTEgb3exzQ5G/oY+RtI6at9TUVJ599lmsVisTJ57MoFuKIWuyz3lSUROieX6DmqcSL0GtQ3gHNYDLzxqm7VuwUv1vZfEubV9KPBCbo6/1CWw97N5oT1BzuF874Th6jr1e219Q7t4wVNSM89OmjjGsTRkrQU0I0fUkqEWglipq/fr1M32IZ8A18NN8OGMxWMLvWyI5OVmby+ChzlGLzIoawOWXX05VVRXLl3+DJcY3qIJU1IRoid+glnoCAE0Ve5s9V7SOv6B26ehiotz/1CxYE4uiQIWh+XByHBDXU6uoAWzJc2+0FNRcTVCylpoqfW20BM+PwhF/JDO7l7YGpxbUDBU147DHqSMNUwGkoiaECILw+1QuWtRSRW3gwIG+O8MwoHlYrVaSkpJM+yJpwetA4uPjiY4O3LjVGNSkoiaEL5+g1mCBVLW7X51XLpOKWvv4BDVHNZm1XzBdzcPsL6xn1W6orNOfkxIHxOX4D2otraO28ipYNI6aTc/p7+v5UZgznaioKLKzswH/FTVjUDttqKF1st1QXRNCiC4Svp/ORUDeFbXTTjvN9NhvUAtzxnlqFgvqb2sjdOhjaxmrptL1UQhffoNaihoOJKgBB9+DNbdA1e52v4RPUDu8EJx1zJ6on/P2SqgwBrV4ILYnQ4YM0UZLbM13j5porqJ2+DM4sEB93yP68gpaUHM3XerZU20MUlwJThdaUDPNTxs6lJy4Uv15Ud373xMhRHBIUItA3tWlRx55xPQ4Nze3Ky+nSxiDWoynyBTBQx9bw//QR1lHTQgP36Bm1YJabWPz50a8+qOw4krY9U9YPBFK17X8HD98gpo7SF00FqLc4x+X7YjxM/Qxh/j4eK3z49bDqJ0fAwQ1Z30Ff7j9SmY/C0UVUGP4nVSCHbDaIDoR0IOa0wVHq4DaPNj+d1Ys/1Kfn3baaVBfqL6AzE8TQgSJBLUIddNNNwHwi1/8gokTJ5qOeYZ9RBJjQxGtjXOED31siTQTEaJ5vgteS0VNU7EZXO6b0HAUvjwdipa1+WVMQS0GKPgcgJSMnowYoY5/3HLIQaE+pUxrJgJoDUVq6hW186O/9vyuJl68ezJ//bCcBSvhpa/8BLWYdHW4BXpQA0Pnxx/uYNn7f9b2T50yAZzuMp8ENSFEkEhQi1D/+Mc/yMvL46WXXvI5dtxxx3X9BXUyY0XN4XRvdPOhj34XvJagJoTGt6IGxPeF6CSfoNZYX0e3Urnd/LipCpaeDXmftOllTEGtbr0e/nJ/xtixYwFwOhWWG97O00wEMM1T23oYtaLmWXwNQFEo/eoG7v3XZm3XwRJzUEuMxbTWpCmolevnLft2rbZ92vjj9QPSSEQIESQS1CKUxWKhd+/eWnert956C6vVypgxY5g6dWpwL64T+GvR392HPvqtqMkcNSE0frs+WiyQMtxn6KNj+WyoPUy3UblD304eqv7X1QDfzIKD77f6ZUxBreIb/UDuZVpQA/jW8HYpcWhVLJ+GIkoT1Bn+HrY/wQN/e4XSan1XSbW/ilrzQa26Hta4G30OHTqUnBSn/gJSURNCBIkEtW5i9uzZ5Ofns2bNGqKiooJ9OR3O36LX3b2iFhUVpf1da3PUmmoCP0GIbsY7qFXVuSs1qSf4Dn2sPQI//KaLriy4XC4Xi5d8y+ZD7h2nL4bjfq5uK07Y8IdWv5YxqMVXq406SOgHPU5m3Lhxfp+THI9WUTOupbYt371x8D33f99n86e/5Z9fmp9/tMpPULOna49zcvTg5QlqK3ZCkzubTZ06FeoK9ReQipoQIkgkqHUj2dnZERnSIEBQ6+Zz1EAf/qgNfaw5ELyLESLE+FbU3EEtZQS1XsXnxibMVaYI9v7773PW79cy6m74ZEMcxPeBia9r8/eo2Q+Kq9nX8DDPUXM/J/dnYLEwcuRIbDabz3NS0nuBTW38MWjQIG3/3mL3xv63oGIryoor+fUb7s6NBkerra2uqH1/dCTlqeezbJt+/tSf9IQNv9N3xEbevG4hRHiQoCYigt+hj1j87OtePMMfGz2jeKr3Bj5ZiG7GN6i5P/GnjKDOq3eIw0m3qdJ/uVht+KEocNHf6ti0ebO61ma8e41OxQmNZc28gq62eJO2rbXJ73c5oP4iadSoUT7PSZ42X99OTiYjQw1Z+46673/pWth4Hx+tauCrLequ/v37awHMGNSsFrDbCDhH7T9LNzHg58v5bIP+/qc6/qT/UismHbJPb9XXKoQQHU2CmogIfitqrm7WTtsPT0Wtut5CZS1QvSe4FyRECPEOavWNitqePT7Xp6LmaKLbBLWKEvNcvEcffVTdsGfqO+uP0KKag9QU6U0+EuxA4iBIG6PtO+mkk0xPiYuLw9bbHIz69+8PwKGjDhqboLQann35Q2Y9pZ/zxBNP0KtXLwBKqpzqfEP3e1osqIHLzTj0EaCsvJKNB9XtzGTomequrGZOhnN+UCuKQggRBBLURETwqaglDoD0k/yf3I14Kmr5ZQq9b4PN2/eDq6n5JwnRTfhbG62mpgZiUny7PkZ4UKuvr+fCCy/kzDPPZM8e8wLX69evVzdis/SdDcW0qHyzeQhiLGo1zaKPdjjxxBNNT0lOTvZ5GU9QUxSFg0dh1lNw22v68enTpzFz5kx69OgBqF0kD5ca3hNMFTW73a6tz+ZtRG/AEgWjHoJpy9T5dEIIESQS1ERE8Alq524Ea7T/k7sRY4v+6np44r9OqD3UzDOE6D78rY1WXV0NthT/Qx8jeN7rP//5TxYuXMiXX37Juk3mIdK7du2ioaEBYttYUavaoQU1uw2irEC/y0yneAc1f8PYPUENYNPhKL42tPJPiLXy978/jcVi0YIaQJWhoqZegB7UAF577TXOO+88n/ca0T8ZzvwWTrgXrJE5p1sIET4kqImIkJiYaN4RnRCcCwkxxhb9AEcqkXlqQrj5q6hVV1dDVBy1jeZ/HhubAItv44tIsXr16oDHmpqa2Llzp3noY0MrglqlHtQS7MCQOyD1BNMpI0eOND32V1EzVr8+/jFZW0YtKxm2vXWh1hnSGNQ8tKBmGPoIMHnyZBYuXOjznBHn/QV6nNzCFyaEEF1DgpqICAkJEsz8qaysND1OjEXmqQnhFjCoWSzUOc3VM7WiFrlDH/3O8zXYsmWLeehjaypqxqCW1htOetLnlMTERFNzD38/y40VtQ9XVmnbf5oFfQfra7E1G9S8KmqgrjfqXdEbccJIn/OEECJYJKiJiODvt7AC8vLyTI+LpaImhCZgUANqHeZQ5nBCJP+T6b9zrm7Lli1eFbVWzFGr2kGNZwhiQlLA04wVs8LCQp/jxqBWXavPsR3TD0gZpj1uvqLmG9QARo8ebXpsXGBbCCGCLXL/1RHdynHHHceUKVMAeO6554J8NaGruAKoivCK2o5n4MOesPvFYF+JCHHNBbW6JvMc18YmwNXgc36ksFh8lzOZOEwfUq5W1NowR81RiVJboFfUmhn1YAxi+/bt8zmem5vrc31WC4zKBZL1oOZp42+kV9TSfY6Bb1Dz9xpCCBEsEtRERLBYLCxdupSDBw9yyy23BPtyQsZ9991nelxcCVTtCs7FdJV1t0N9Iay+MdhXIkKcv6BWVFQEQK3DHNQcTsAZuUHNuDC1x/gTehMbq7ZN9K2oNR/Utq39nAn3g8s9n6y5oHbKKado2+PGjfM5brfb6dPH3CJ/SE+ItwNJ+oLYzVbUomJ9jgHa/DYhhAhFEtRExIiKiqJv377BvoyQ8uCDD7Jq1SoGDx4MwNEqcJb+GLHDH48UF3GgFVNnhAD/QW3JkiUA1DnMHf8cTsBV3xWXFRT+glpOz95akNm9ezf1TdF64GkmqJWWlnL+5bexxvBjxl/FzuMXv/gFkyZNIjMzk7///e9+zxk0aJDp8ZjjgIwJYNUbvDQb1AIYM2YMU6ZMISoqirfeeqv5k4UQootJUBMiglksFsaPH68FNZeiLhbLnleCe2GdID8/n9x+xzHgN7A6wkd3io7hrz3/559/jtPppLbRHCwamwBn9wpqmT37aXO2XC4X23fs0Ktq9f7nqCmKwtVXX83eg+bj/kKUR0xMDN999x0FBQU+C2B7/P73vzdV5U6eciac/GqL79FSULNarXz99deUlJQwe/bs5k8WQoguJkFNiG4gK0vv1lZcCex7FVzOoF1PZ/jDH/5AfX09LgVufTXYVyPCgbGiNjhH/W9paSmrV6/2WfA60oc+eubmGWX2GmxqrmEa/thwFBSXz3P+97//8dlnn/nsP+OMM1q8hqiowOuWnXXWWezevZvf/OY33HTTTfzivo8gZajpnIBz1I6/rdn3tVgsLTZTEUKIYJAVgYXoBkxBrQJG1OZB4WLodU4Qr6pjHTqkL+Rd71soEcKHMahddBL8zZ0v/vvf/1LXqJjPbaLbDX3M7DsCa7r++9wtW7bAme6fJYoTGstNTTqampr47W9/qz1+5zaYcWIsu09Yxknjxh/zNebk5PDEE08EPB4TE0NycrJpWZKE4b+EnwR+jhBChDKpqAnRDfhU1AD2vByci+kkZWVl2nZ6YjMnCuFmDGozDSPu/vvf/1LbYK4WOZqI6Iqa36DWZygnnKAvUN1SQ5GXX36Zbdu2AXDyILh0AqT2GsHY8ROanaPWkbyHPybkjAKr/E5aCBGeJKgJ0Q2Yglqdez2jw5+2btHaMFFaWqptp8v656IVjEGtXw/4yTB1/OMPP/zAocIK07nq0MfuVVHLys4hNzdXmxu2efNmrxb9+jy0yspK7r//fu3x4z8HiwVI0YNeV7jkkktMj5vrNimEEKFOgpoQ3YAxqB2xuj84uRyw740gXVHHMwa1NM9nM0Xxf7IQmIOaLRrOnXK89riiqs58bjdsJpKUlITVatU6P+7bt49al2Eul6GiNm/ePIqL1eB26bnjmOS5lSldu4D0ww8/zOWXX649HjVqVJe+vxBCdCQJakJ0A6aKWlOufmDvyxETZowfNJPj3BsumawmAjMGtZhoOPfUwGtqqe35u9fQR89wRU9DEUVR2JZn+H+qoQSAvLw8Hn/8cQBsNhuP3GhYRDq1aytqNpuN+fPn89577/Hpp5/6XZdNCCHChQQ1IboBU1Arb4RM9wKzFVuhZBUcXQ3fXQEFi4N0hR1Li55KUzAvQ4Q4U1CLgvGjB5Cenu73XG3oY4T8YsObd1CbdXp/bdvU+XGfYUioO6j98Y9/pL5erTbedtttDEw6rJ/TxUMfQe0eeckll3DBBRd0+XsLIURHkqAmRDeQmanPKykuLoYBv9Qel/zwHIfemwYH3oaVVwfj8o6Z4vXhucmz8oBU1EQzjOuoxURDlC2Ws88+2++5jU0ASkSGf5fLRV2dOtTzuEx4aw7868GLtOOmhiJ7j+pPbCwhLy+PN95Qh1CnpaXxxz/+ESo2qcdtyRDfp9OvXwghIpUENSG6gZiYGFJTUwHYunUrZUnTWbUvjiuehZxz53PcnGoWbwTqi6ChtNnXCkXea0A5JKiJVjBW1KKjAKuNc8891++52vdUBM5Tq62t1bYHZMHsSZDaQw9YporaTkO1rOEoe/bs0X5RcvXVV5MeUwm1eerxtDHujiJCCCHaQ4KaEN2Ep1JQVlbG0JFjOfneOt5eqVafXAr890f3idV7g3eR7eRpYuChVdQisPohOo4nqMVEu/OENYazzjrLbyv5Rs+3UgQGNeOwxwS7e8Out7nv06cPycnJAGzebvj50FBCeXm59jA7OxsKv9CPZ0/rjMsVQohuQ4KaEN3EvHnzSExUFxjzDjYARzzrq1Xv6cKr6hg+Qc2zBJZU1EQzjEENAKuNHj16MGHCBJ9z9Spt5DUUaSmoWSwWrfPjgYN5VHuyqldQS01NNQe1njM654KFEKKbkKAmRDfRt29f5s2bpz0eNWoUT9+Yoz3Wg1okVdQkqInAfINaDADnnXeez7mRPPTRGNQS/QQ1MA9/3FrkXv+i0SuoJSdD4ZfqA1sqpI/thKsVQojuQ4KaEN3IzTffzDfffMP333/Phg0buPX232KLUo8VR1BFTa9+yNBHEZgnqHn+H/AENX/dAvWhjxFeUYt1b8Sag5qpoUiBe/0L74pa9FFoLFMf5JwB1iiEEEK0nwQ1IbqZyZMnM2HCBCwWC5ZBvyAzRf0wdaTKfUKoVtQqd8C638CRFT6HAlbUZOijaIa/oY8Ao0eP5h//+Ae/Ol0/1+EJaq7Iq6gZm/H4G/oIXg1FDrsDWGMp5WVl2v7Upm36E3LO7PDrFEKI7kaCmhDdWUwaWbkjAXXoo6IAVSFaUVv3a9jxFHw32+dQwDlq0kxENMPTnt976COo1ecXX3yB9ET3ud1k6GOCHTWwRieZzjEGtc2H3P9fKS7Ky/T/91Lrf9CfIEFNCCGOmQQ1Ibo5zxprDidU1AK1h8DZ2PyTWqvmEDTVQvkmWHE1HP6s/a9VsUX9b+1B9TUNfIY+atUPqaiJwLSKmjb00WY+YdANxCRmq+d2l6GPdtRqmlfny549e5KWlgbA1gN12v7y0iPadppjvftF+kPSwM67YCGE6CaiWz5FCBHJsrKytO0jVZCaoEDNfkg+/the+MC78N3lkDgQUNS5b4c/gYuPQFRMi083URSoN4SxhiMQ3U97uG3bNtPp0vVRtEagZiJGNpu6L6IrakVbtG01qGX6nGOxWOjfvz9lZWXkl9ThdEGUFcpKS7RzUmPdabanVNOEEKIjSEVNiG7OU1EDKK5wb3TEPLUtf0ENaLv1BiWOSqjY3PbXcpSDy1Dlq9d/i79r1y5+/PFH0+kRtY6asx52/VPvpic6jE9Qs9h8zrHZ1H0R3Z5/1zvatlZR86Nnz54AOJ0KR91zWsvz1f+f7TFRxHpybo605RdCiI4gQU2Ibs4Y1PSGIh0wT62xxP/+kjVtf626IvPjBj2ovfPOO3iLqIrarhdgzS2w9Byo2Brsq4kYTqcTl0v9RtGCmp9Kb0yMui9iF7xWXNQc1f9/by6o9erVS9vOd/cQKa9W/x9LjXf/T2exqh0fhRBCHDMJakJ0c8ahj1qL/mNsKKIoCmXWAEMnS9sR1Oq9glp980EtouaoVbqHdSpNsP+t4F5LBPFU06CNFbUICGrr1q3j4YcfJi8vD2r2U2MoEramogZQUK7+t9w9XTQ1TlE30sdCTFrHX7QQQnRDEtSE6OZMFTVPUKs5tqGPl1xyCT1++hW/f9vdSdKoZG3bX7C+0PzYXVHbunUrmzerQ68mjupNtLspRER1fWzSW6dz4B0/N1S0hzGoea+jZuRTUQvzoY9Op5OZM2dy3333MXHiRPZv/NI3qNUX+32ud0XN5YJKd1+R1Hj3Aen2KIQQHUaCmhDdnHnoo/tHwjFU1Kqrq/nwww9xuWDefyD7FrjpZbQ5LVRs9una2CLvipo7qBmraZdN60u0+/Ijah01Y1Cr3g1lG4J2KZHEb0XNu+sjekXN6XJn5DCvqBUUFHD48GEA8vLymHbZH9hl+D1Igh3Imeb3ud4Vtco6/fcGqQnuAxLUhBCiw0hQE6KbMw19rHV/2qre2+7KTWVlpenxkUp44St4ds2J6g7F2faw4Wfoo6IoWlCzWCxcOjnVT0UtAoKao9r8+OC7wbmOCONZQw1a6vqohzeHk7CvqB04cMD0eG9eGf8z9OJJ6HcG9Lvc73ONQS2/TB/2CO6KWnQC9JjYkZcrhBDdmgQ1Ibo5U0Wtxv1B1VnrG45ayTuoeew9atcftLWhiJ+K2saNG9mxYwcAU6ZMoVdyrVZR0zv0RdjQR5Dhjx2ktRU1z9BHcA9/DPOKmndQ85ZwxlsQk+r3mHHoY0E5lOvLr6lBLeu0ti+9IYQQIiAJakJ0c8nJydqHUa09P7S7Rb8xqE0wrHlb3mD4ANfWhiLeXR/rj5iHPV52GTSUYHN/4I7YoY8ANfugdF1wriWC+A9qraioRVBQ+9ujD9Mn3Xw8ISGBQLKzs7G4F8L+ZB2M+aN+LC0RGHxzR16qEEJ0exLUhOjmLBaLVlU7UmEINu1s0W8MalOG6vvLqxX9g3CbK2rmZiJKfbEW1KxWK7NmzYKGo75z1CJh6KN3UAMZ/tgBTEFNayYSeI4auLuJOiNn6ONpP+nJl3dDVrL6uGfPnsTHxwd4JkRHR5uGShuljrsXep/fodcqhBDdnQQ1IYT24etIWQ0uz/yuDqioZSZBnF39MVNeUQFpJ6oHqnZCY3nrX9Rr6OO6bUXs3ate3+mnn052VhY0lGhz1CJy6KM9AyzuL/DguzL88Ri1tqLmM/TRFTkVtX7JpQzpBSsfhLtvOJP3338fq7X5jwXGeWpGqZl9OvQ6hRBCSFATQgA5OTkANDU5KfXMO2ln50djUEuOg7REtSJRVlYG6eP0E1s7fE9RfILaO9/qk2Muu+wyaKoCpUlrsx4JC147HA614YWnmUhcH8iZrm7XHICS1cG7uFBVuQMWnQyrb2wxyPpfRy3K5zzfoY+RUVGLi4sj07IPgAFZ8Je5f2TSpEktPt84T80oNTW1w65RCCGESoKaEEILagCF5e6Nqp3tei1jUEuKg9QktSJRXl4OGYag1trhj44K3l3RyEMfQbW7mPG+O6NER0dz8cUXQ8NR9bEnqIX50MfCwkJyc3Pp168fy7e4g4EtEXJ/pp8kwx99rb0NSlbB7hch/3/Nnuqzjpo1Btzzr4wiqZmIoihaUMvNzcVSsVE/mDaqVa8RFxfnd78ENSGE6HgS1IQQZGdna9uFDe7QVr6xXUMHvStqqUmxgLq+WlPKGP3EVjYU2bh2OZc9A/e/D88uhqIK2K8uo8Ypo7LJyMiAhhIAP+uohefQx88++4zCwkIKCgo47WFYuQuIToQ+F4HFXf45+C4oruZeptupO/iF/uDod82eW1ur95aPjcHv/DTw154/fIPa0aNHta+7X79+6v/jAPF9ISatVa9hDLhGzTUhEUII0T4S1IQQpopaUVOuuuGsg8rtbX6tqqoqbVsNavpv4CtcOWrggFZX1J5+7kVt++53YNMh/diYrMNq5cQT1LznqIVpRc0YIgBufBmUqASwp0PPGe6T8uDo90G4utB03XXXkfRL+Icnq9UV+D+xqQbKt3A4L0/b1TMViPJfKYqkoY+m+Wm9MsDh/qVKauuqaQBz5swBICoqytR4pH///h1zkUIIITQS1IQQ5qGP9T30A+1oA+9TUUvWP8yVV1ZB+knqg9pDvm33/di+c7e2HRsDGw/qx0blAqt+pXWojJQ5ag0N5jCw6RD8sMddyTAOfzz8aRdeVejKz8/n1VdfxemCfy1z76zL9z3R5YBF4+G/J3D4x/e13X3SgTj/TTIiaeijKahlGv75b0NQmzFjBkuXLuX7779n9+7dXHXVVTz77LP07t27Iy9VCCEEEN3yKUKISGcKatWGIUyl62DANW16LZ+glpKoPS4rK1PnqRV/7X79Nc229K6srGTVhl3aY6fLyvoD+nC/kX2BusOw8T7APPRRUcASpkMfvYMawKv/O8BJ1wE9z9Z3Fi7puosKYUuW6PehyLMWYM1+3xPLN0LFVgDy9m3WdqtBzX+TDN+hj+FbUcvP18Nr39Q6/UDa6Fa/hsViYerUqdrj119/vSMuTQghhB9SURNCmOeolRsOBKioVVRUqB0J/fDp+piSpD0uLy83d35sYfjj0qVLaWrSg5mjycWnG9XXs1qtDO/nDoGNZYA+9BHApRC2Qx/r632rNm8t3qUGuLhsSB2p7ixdBw2lXXx1occY1I5Uuhs+Vu3xraiWrNU28wr11d17p9G6oBbmFbWaGr1bahKGtQnbUFETQgjRdSSoCSHMc9SOlkOCe75J2QZwOU3nrlq1ipycHAYPHmz64OfhW1FL1h63tfPjl19+6ef11TlwgwYNIn7iE6ZjNkNQczQRts1EjBW1Punqf0sr6nnppZfUB9nuNv0oULysS68t1CiKYgpqDidU1AJKEw1Ht3H++edz2mmnUVJSYvrFw+GjakUpJhp6JAFx/ofumYY+OgnroFZXp1fR4pvcY4itdkgaHKQrEkII0RwJakIIUlJSsNvtAHzxxRdc/nQTPx4AnLVqu3ODDz74gPr6eg4cOMA333zj81qeoBYTDXabuW13eXk5JBynLt4MULq22fWutmzZEvDYqFGjYOD1+tpimCtqTS4iYo7a3Rfq+x9++GE1HOdM03cW+obZ7mTXrl3kGRqDABxx97P5z0dv8tlnn7F8+XL+9a9/qd9vbnkl6i8geqeB1QrER35FzdikJq7psLqRegJYZRaEEEKEIglqQggsFoupqvbOV4e46y33g53PmM4tKyvTtsvLy31eyxPUkt1N9FLT0sznWyz68MeGI1B7kEB2794d8NjIkSPV15rwL62TZLThJ1qTk7Ad+mgMapMGwyz37SoqKuLvf/87ZJ2qt+n3zPfrpozVNI9i96jGfbv1eWjrf1gL5ZsAqG2AUvc64r3dFctAQx99K2q1fs8LB6aKWoz7FyQy7FEIIUJWWAS15cuXc8EFF9CrVy8sFgsff/yx6biiKNx///307NmTuLg4pk+fzq5du0znlJaW8vOf/5zk5GRSU1P55S9/SXV1dRd+FUKENuM8NYAvPJ9xD76ntoJ3M4az5oKae/k0UtMytGNayEv/if6Ecv3DtFFdXR2HDqm9+If4ach3+umnqxsJ/WDi6xDfl+iUAdpxtfFD+A99tNvgoUvBalUXY543bx6lVQ5IdA9PrT0cjEsMGf6Gx3oqanmH9C6HP65fC4r6/XBY/12DNrQ00NBH34pand/zwoGxohbvyZ8S1IQQImSFRVCrqalh9OjRPPfcc36Pz5s3j6effprnn3+eVatWkZCQwFlnnWWakP/zn/+cLVu28MUXX/Cf//yH5cuXc8MNN3TVlyBEyDNW1DwcTYDihJ3PavuMFbWyPUvUdakMfCtqert/LdglDtKf4G6t723fvn3a9kleSzRde+21TJkyRd/R96dw0UFsPfTudeFcUTP+7LLbYFhvuOZSdbhjRUUF8+bNA7v7vjoqwnaI57FyOp0sXbrUZ3+xe5qksWHIzj0HqHOvcHDY0H+lTwsVNZ+uj03hW1EzDX30BLU2dHwUQgjRtcIiqJ1zzjk8/PDD/PSnP/U5pigKTz31FPfeey8zZ85k1KhRvP766+Tn52uVt23btvH555/zr3/9iwkTJjB58mSeeeYZFixYYGpXLER3lpmZ6bPvcIX7Q+ruF7VAZqqobX8fNj+sPW5oaKCxUf007AlqaelZ+vme5yYN1N+kyn9QMw57HJQNvzpX/SA9ePBgdfifH9HR+lybSJmjZnd/SX+663ptGN7TTz9NQXWi4QklXXl5IWPDhg3aLw6SEvX1+jxDH/OK9VETLpfCZvdi6XmGoNY7DbBYIVb/PjXyWUfN1QCKy++5oc409NHu3kgZGZyLEUII0aKwCGrN2bdvH4WFhUyfrjcUSElJYcKECaxcuRKAlStXkpqaytixY7Vzpk+fjtVqZdWqVT6v6dHQ0EBlZaXpjxCRyrshA8CBqDPUjcYy2Keul2QKarXA1ke1x1VVVdq2VlHL0IdU6hU1Q1ALUFHzDmpP/d8UFi1axA8//EBycrLf55iCWoQMfYx1Z+Xc4wZyyy23AOoH7ofePGR4wtGuvLyQYZyfdtHZk7RtbejjEXP160f3dMg874pabE7Ahho+FTUI24YipoqaDbWKGNsj8BOEEEIEVdgHtcJCdS0Y7/k12dnZ2rHCwkKyssy/LY2OjiY9PV07x59HHnmElJQU7U/fvn07+OqFCB1XXnmlz74DllP0Bzv+DorLFNTKvLrze7fmB0hJ9xPU4npClHsSW2uCWg7Ep/ZmxowZJCYm+j0f/HyoDtOhj95z1ACITuSee+7Rvv6XPt3J3mLPEySozT5/vLZdXKkG9YLSRtP5Px4ELFG+c9QCDHsEPxU1CNvhj8aKWlwMkDQo8MlCCCGCLuyDWme6++67qaio0P54GhsIEYkuueQS7rzzTk444QRt34GjQNZU9UHlDpT8z80VNU9Qc7fY9xfUomNTtHChPddihUR344/qfaC4+M9//sMrr7xCU5P6adi7okas+Zcx/vhW1CIgqHm+JFsimZmZ/OY3vwGgqcnFR55l6LphUGtoaNCWh+jTpw+njNabgRypVBdud3mNUPzxAJA5xTRHrXcLQS0uLk7b9sxxC9eGIp6KWnQU2KKB6KTmnyCEECKowj6oeRogFBUVmfYXFRVpx3JyciguLjYdb2pqorS01G8DBQ+73U5ycrLpjxCRKjY2lscff5zXX39d23fgwAEYeof2uHr9Yzid+gLY5Z7Cgjso+AtqRCdoa6kZG5FoDUVcDaxfuYgLLriAX/ziF7z77ruAHtRS4t0LErc1qLkI+6GPFothbTj3EgRTp07VzvMM8euOc9RWrlypVYimTZtGkr1Bqz4WV5qHN3psPARKr3Mp0nuMkJ0CxPvv+AgQH6/PfavVglp4V9S0jo/RgavTQgghgi/sg1r//v3JyckxDYGprKxk1apVTJw4EYCJEydSXl7OunXrtHO++uorXC4XEyZM6PJrFiKU5ebmatsHDx6EXudr1a/yfeYOe9rQR3f7ft+gZoGoONLca6mVlpbqw68M89Qef+IJbfvmm2+msbFRDYmo1TSLhTYHtZAf+uhshM1/hu9+7v5zBXw3GzbcQ32dGgTsNqv6tQNEJwBo9xIM978bVtSMP/OnTZuGpamSLPfv0gIFtYpaOFA/iOIq9Z++1Hh1YXaShwZ8n0isqGkdH20S1IQQIpSFRVCrrq5mw4YNbNiwAVAbiGzYsIGDBw9isVi44447ePjhh/n000/ZtGkTV199Nb169eKiiy4CYNiwYZx99tn86le/YvXq1Xz33XfceuutXH755fTqFXjIixDdUXp6OgkJaig4cOAAWKNgyK8B3zlpWkWtVh0W7BPUouPBYmHEiBGAWimaN2+eeoKh82Nl2RFtOykpif379+Nyj1sb5MlnrQhqxjlqQRn6uP9t+KgX/PjHVpw7HzbeCwfecv95Gw4sgK2P0FCtjhDQ5qdZ7WBVH0hQU3kHNRyVZLpH8h2tgkOGoJabqX9f/LjfQVGFOlQ3OwU1AB/nOz/Tw29FrSk8g5pU1IQQIryERVBbu3YtY8aMYcyYMQDceeedjBkzhvvvvx+A3/3ud9x2223ccMMNjBs3jurqaj7//HNiY2O113jzzTcZOnQo06ZN49xzz2Xy5Mm8+OKLQfl6hAhlFouFfv36AWpFTVEUGHQjDLhWn5PmVl7rnp7mp6KWFIdWBbr//vu1atejjz7K/v37TRW1qnJ96HJyQozv/LSY9GarHh4+c9SULhz66GyEFVdAXQFs+QvUH2n+/JJmOs7Wqx+oY23ucpqh8iFBTf0+W716NQBDhw5Vf+HmqNAqak4XbDJMKT5vtB7YV63dQFWdGtSykoF+l4M9nUD8V9TCc+ijp6KmteaXoCaEECEtLILa1KlTURTF58+rr74KqB8s586dS2FhIfX19Xz55Zccf/zxptdIT0/nrbfeoqqqioqKCv7973832z1OiO7ME9Tq6+tZv349RNnh5FcoH/KY6bwmJ9Q2oAU1YxfV9AQgSg1qw4YN47bbbtNe87e//S0kDdbOrSzVn5dkPcrulfo8uUE5Fpj0JkTrH5gDCeo6aoc/MT8u/LL58yt36NvnbFD/uDU0qtdtj1YDhfEDdVJSElar+qO7uwa1TZs2aXMltTl7jkoyDdOIf9DXS+e8E/XtL774QtvOTgGG/6HZ9zJV1Dw9XsJw6KOiKFpFLU4qakIIERbCIqgJIbrW4MF6iJo8eTLPP/88iqJQ5vRdc6msBm3o47Zt27T9Q3qhVdQAHnjgAW0ZjQ8++IAlq/ZBqrrYbqmhUudsqGL3mne1x4NO+w30OrtV1+3Tnr8Lg9r3n8zj1a/Vag4ABYuaf4InqMVmQdpoSB0FVvUTdEOjGkLsNt+gZrVa9eYs3TSoGdcD0xZqb6xQm864bT2sb586VA8nxrnKWcPOa7FFvami5vl2CsOg1tjYqA0n1oc+JgR+ghBCiKCToCaE8HHnnXcybNgwQJ3XcvPNNzNz5kx27drlc255LVpFzRPUoqNgYBamD4IpKSk8+qi+OPZtt9+OY8BtuFyY2qWX1cDuQkV7POj0u1p93cEa+vj9sk+ZfMdarnsRHvEU1goXa8sWGLlcLpx1pVDvriJ6hnRaLGpoA+o9QS3KN6iBPvyxvNY9NLKbBbX6en3BaW2Iu6NCreK6eRanjrNbSIqDke5lMBXD30n2AH3ttUD8VtTCcB0102LXUlETQoiwIEFNCOGjX79+rF27ljlz5mj7Fi5cyJ///Gefcz1Bzel0snPnTgAGZ7vXabKlmM69+uqrtU6r27Zt49nPyigo1z9UgzuouaesJSQk+Cxm35xgrKPmcDj4xY23a5W0uR+581ldAXx3GWyaC/sXQOUOqqurOeGEE8juM4Ctee4XSBqiv1isulxIg/uytWYiXvdRD2qKulaYBDVwVJLmp0CUHq8Gs9H9LD7HsrKyWnyvSOn6aFzsOl66PgohRFiQoCaE8Cs+Pp5nn32WhQsX6sPL/CirAery2Ld3r7b+1zDPslTxfUznWq1WnnnmGSzunvN/mvtnNsSbK2bltbDviHp80KBB2rmtEYx11F5++WW27TygPXY4YZWnF8rB92DTA7BiNvxnKItf/z3btm2jpLSCpz0jI5ONQS2bJie43EWfWC2omddw9AQ1lwuq6oGmanDW010ErKj5yR2e8DZ6iO9aaa35JYDNZiMqSl3MTu/6KBU1IYQQnU+CmhCiWeeffz7Lly8PeLy8BnDWs3Wt3qRhmGfVC6+gBjBu3DiuvfZaQO3e9+InO0zHFQWanGpSGTSo+flD3nzmqHXBOmrfffedz74FP/YCfAPmwS16S/lFm9yVN2M3y9hsrZoGxoqaOah55qiBcZ5a91n0uk0VNXcWGT16jM+x1lTULBaLVlWLuIqaBDUhhAhpEtSEEC0aMmRIwDUHPWupbVv5trZvuFZR6+v3OVdccYW2/emnnwZ837YGtWAMfSwvL/fZ9+73CsrPauDczTDlQ4hWu1wcysvXztl/xD3E06ui1mAoAto9X06AoY/QPRuKGENHbGwsOBugqdpvRc2zb9TYyT7HWjus1jNPrTaMg5qxoqa355dmIkIIEcokqAkhWmSxWJgyZYrfY2W16rCwbRtXa/sCDX30mDRpEjExMX6PGR1zUOuCZiLGoDb9BPW/BQUFbN+1H1JHQN+fQoramOVQUZXpuYs2RUHCcfqOuBzqW1FR8x/UWli3LYIYK2pxcXFaYxa/FTX3vuTMgfTv3990rDUVNe09MFTUwnDoozHcytBHIYQIDxLUhBCtMnmyb0UCoDxKXbNwW576KdZigSE93QcDVNTi4+M5+eSTW3zPYwlqXdWe3xPU4mJgxkh9v2m4qLthyCGv0YmLtsSBVb9mn6GPWkWtFUGtsawdVx+efIY+1hUAAYKaJ4vE9WT06NGm5yUlJfk+wQ+tohbG66iZKmoS1IQQIixIUBNCtErAoGYdiKLANve6Vf2yY/WhVQEqagBnnHFGi+95LHPUumrBa09QS42HKYZRjKaglqLOQ/MOaks319HY2Kjv8Br6qDUTiWnN0MdSugvfoKYOKU2N9z1XC29xOaaglpWV1epGNVpFTVtHLUIqatL1UQghQpoENSFEq4wcOdLv/pIaK/muEWr3QWBYjjt4RCf5BAwj76D229/+1vQ4NjY24Ly4QII59DEtAX7SX6++LF++XF+zK2kITU4oKDc/t6bOaW5GEpvTqmYi/itq3Tio1apBLToKkhPjTOdqa6vFZpuCWluWffD8nTY5wdFE5FTUomSOmhBChDIJakKIVomKiuKDDz5gxowZ/O9//9P2HzlyhG2uGdrjYT3dC4ol+B/26DFx4kQmTZqEzWbj3//+t89QyIEDB2K1tu1HVFc3E2lqaqK6uhpQqzkxMXYmTpwIQF5eHgcOuNv2Jw8hv0xvu2/XC38sWrRIfxAXoJlIdOCg5mnm0q2DWn2B9jg9zXyv0hNRf2kQncCJJ56o7W/LLwF81lJrCr+g5lNRi4oDa1TwLkgIIUSLJKgJIVrt4osvZtGiRZx99tlaWCguLmZbmf6hV2skEhd42COooerbb7+lpKSE6667zhQ+oO3DHsFfe36nuwd+56ioqNC2UxMAezqnnnqqtk8b/pg0iDxDjrroJH178eLF+gNbKg1OPWy2rZlIM0Gt/gismQO7X2zuywkbgYY+AqSlpZvOTU8E4tSFxPv3788tt9xCbm4ut99+e6vfz1NRA3fnxzAc+ujT9VE6PgohRMiToCaEaBdPx7yioiK27tij7dda87dQUQO1m6SnoUNHBDWfBa+hU4c/Gjs+psYDMRmcdtpp2r6lS5eqG1GxHKrRFw0/qT+MOU7dXr9+PUVFReoDi4V6UrXztIpaa+aoNddM5Mc/wq5/wOoboWxDi19XqAs09BEgPcO8OHt6AhCboz1+7rnnOHDgQKvmSHoYK2pqUAu/ipppwWsb0khECCHCgAQ1IUS7eOb4VFdX88MPP2j7W1tR89bhQc3p3ujE4Y8+Qc2ezsknn6wtwvzVV19p89QOVffQzu2bAWf/RK/SfPzxx9p2gyVV2+6wOWp7XtK3CxYFPi9MBBz6aLWTlmFuuZ+WAMT15FgYK2rq0Mfwq6iZFry2I0FNCCHCgAQ1IUS7GNegWrduHQDZaTF6l71WVNSMwj6oJQAx6djtdk455RQADh48yN69ewE41DBAO7dvOlwyTW8R+dprr2nbDVb9PsQGCGopKSlax8KyGnfnwtbOUWsoafmcEBdw6GNcT5/vo/RETBW19jBV1BoI+4pafAwS1IQQIgxIUBNCtIsxqDmdaioaNmwYYAFLNGSdFuCZ/iUlJZmah3TIHDXo2qGP9gzA3NHyq6++AuBQub7Ad98MGDMsR+ukuXLlSnbs2AF4VdSiASw+H6qtVispKepwyLJa9z1rbo5aVKy+HWlBzWbRv6a4XqSn63PUoqyQHIc2R629fCpqYRjUfJqJSGt+IYQIeRLUhBDtYgxqHsNGT4LzNsMFOyCpbUHLarWSmpoKqIGrb9+2VeQgwBy1rhz6GKOGBGNQ88xTO3ToEABWC/RKA0u/y7n22mu18zxVtQb06pndBtiSwOL7o9oTSEpbM0ctJkPfbjjawlcV+oxBza4Yvm6vilpagroAe4dW1AI0E2loaOBPf/oTTz31lL4sQwjxbSYiQU0IIUKdBDUhRLv4W4dq2LBhkDIcEgf4eUbLxo8fD8Bpp51GVFTbW4eHwtBHgLFjx2pNUjzz1DxBrWd2BtGnvArHXcHPf/5z7et8/fXXcTqd1Cv6B2h7ND7DHj169FDnvJVWOXG6UMODs97vudgNnRDri9v8dYYaT1CLiYnBWl+oH/CqqKV7bmVHV9T8tOd/+eWXefDBB/nNb37DZ599dkzv1xlMFTUb0vVRCCHCgAQ1IUS7+K2oDRt2TK/55ptv8tZbb/H222+36/mhMvQxOjpaa9NfVFTEhg0btM6OffoNggHXgDWa7Oxszj33XAAOHz7MkiVLaMAQ1GyAzf+i4Z6gpiit6PzoMtyD2kNt+yJDkCeoea+hRlwvc0XNk686vKJW57PswyOPPKJtv/DCC8f0fp1BKmpCCBF+JKgJIdrFX1AbPnz4Mb1meno6s2fP1kJIW3V2Re2zzz7jxRdfxOFQX7OsTA9GxqGPAKeffrq2/cYbb2jb3kM6jcMfX331VRoUPRTE2mixogZwtMq9EWieWlONvl2X3+kLgXc2U1AztOYnrqf/ilqsb/W3LXwqaijgajCd46mgAlRVlHbq+n3t4TNHTYKaEEKEPAlqQoh28Q5qycnJ9Ox5bG3Qj5X3HLWqOliz9get2cmx2LVrFxdccAE33ngjc+bMAaD8aJ523LPgtYdxntqbb76pbXsHtfPPP5+MDLUS99FHH1Fc1qgds7c1qAWqqDkNQQ0Fag/7Py9MmIJanTGo+Rn6GBULsb6/VGgLn66P4NNQJDFRDz5Vh1bAwkGw/e/H9L4dybSOmgQ1IYQICxLUhBDt4j1HbdiwYVrL+GAxBrXGJhh/P4yfNptHH330mF9706ZNWpOIl156iffffZfy3fqaZGkJgF1fbHn06NHaMLziYn1emHdQi4mJ4YorrgDUAPL6/Le0Y+octeaHPoIxqLWiogZQc8D/eWGiuaGPI0aMYMiQIVgsFn76s2tgxiqwRgd4pdYxVdQ8xUivtdSMw26r6oHqvfDDHVD24zG9d0fxBDW7Te2GKV0fhRAi9ElQE0K0S3JyMjExesv5Y52f1hGMH5YPl8J2d7HllVdeOebXNg5zBPjVjb9iR57evCOl/3RIGaE9tlqtpuGPHv66WRqHPx49qndltNsIWPlodVBzOX2bjNQe9Pua4aK5oY82m41NmzZx+PBhZt3+KqSNOub3a01Fzfj9UVhuGPm499Vjfv+OUF1dDbjXUAOpqAkhRBiQoCaEaBeLxWIa/hgKQc1YUSs3FDz27NnD/v379R3V+6GmbU01vINaeXklu9X+ICTERWM7c7G7F7zOOPzRw19QGzNmjLammpE9GnA1+uyHNsxR87fmV034BjVFUfwPfbTaIUatYNpstg4dhus7Rw2f+2oM2FX1UFrtfrD/zZCYE1haqn5vZHjymXR9FEKIkCdBTQjRbqEc1Cq88smSJUvUjdL18OkA9U/V7la/tjGoGd8HIDUlySekAa2uqFksFlNVzSPWBjRV+eyHNsxR8x72CGHd+dHhcGhDUE1DH+N6+f076Ag+XR/BNPTR5XJRUmJeSHzfEfdGwxHI/1+nXFdrOZ1OrUOp1mDFfmzz9oQQQnQ+CWpCiHYzVi1CIagZhz5WeK1J/OWXX6ob318HKGrb/q3zWv3axqD22GOPYbXqoSA1Nc3fUxg2bJhpLl90dLTf9ecAZs2a5bPPbgMclX7Pb/XQR6efoObwH/7CgbF7Yaw9BhrcASmu8xrZ+K2oHflG21deXo7L5TI9Z7/VENI7a/ijswHqClo8raysTAu3WkWtE++XEEKIjiFBTQjRbr/61a9ITEzk0ksvZeDAgcG+HHNFzSuoLVmyRP0wXblN36m0vhukcc208847jweu7KM9Tu/hf50ui8ViGv7Yu3fvgAt55+bm+oQ4ezSQ+zO/57c2qLkaq8j3LrT5C29hwjPsEdwVR4+4Xp32nsagplXUNs2FOnWxbeOwR499DUP09dvy/wP1vuccE5cD/jcaPuoF+xc0e6qx2pfRQYuACyGE6HwS1IQQ7TZz5kzKysp49913g97xEcxBrd5rWtCRI0fYvHmzec6XvfXrtRkramlpafzxvBquPAUSYi3cfMutAZ9nDGp9+vQJeJ7FYmHcuHGmffah18PAX/o9Py0tTbvnWlDz03Z/1lW30/tWmPuhYadXx8JwYg5qhrXKOrFCZBz6WGfrp240VcGP9wABglp+NfS/Sn3gcsCB9i3iHlDpD1C5Q91eMbvZU01BLQmwRLfpe18IIURwSFATQhwT7/lawdTStWjz1Dza0OTBtLh1rIOoplLeuAUqPjmd2bMDf1CeNm2aFqiGDh3a7HuMHz/e9Ng+fh5YbX7PjY6O1tr/H61xV+mqdpjOaWho4OP/LgfggQ8Mi4D7m7cWJoxBLc5mqIh2VUXNdpy+ZMLeV+Doav9B7VAx9L9G39HRwx+9q6eB1tBDbyQC7opabDZY5J9/IYQIdfKTWggRMYxz1Pz5crFXUwdHRatf2xPUkpOTiarZpe2PSh/e7PP69+/PCy+8wBVXXME999zT7Lk+FTW7vdnzPcMfj1a5q5n1xabOj8YP6ADfbHdvOCOkohZlCNqdGNRMFbUGJ4x8UD+47jaOHin2ec6evfsgdQSkj1V3lP0A5Zs65Hp27NjBidNv5Op/GpYBKPgi4PnGilp6AjI/TQghwoQENSFExGipovb1N9/iaDLsCNCowx9PUEtLSzPPc0tuvkoG6ly+N998kwEDBjR73tixY02PWxvUKmqa9K+rcrt23LsT4cfr3BuRMvQxqkE/0Inhw2azad9btbW1cPwtkOIO6CWrObpzsc9z9u3bR1NTk1dV7bUOuZ7777+fH7cd4o1vYd0+985D7wc832foowQ1IYQICxLUhBARw2r1/yPNM0SwpqaOVXsMB1pZUVMUxRzUKvQwRHLHdbs0NggBAjYe8Xd+iWfdLkOI9Alqa90VmAgZ+hhrNSzk3YkVNYCkpCQADh8+jGKJhpOe1o4d2fKetp2ZrP63qamJAwcOwHGz9eGr++eDy/ibgrarq6vjs88+0x7vca/lx8H3YOk5ehdMA59mIhLUhBAiLEhQE0JEDIvF4nf443nnnadtf7nZcKCVFbWamhq1OkL7Kmpt8eqrr5Kens59993X4rl+Oz82U1E7WAI7C4icoY9Ww9cR37lBbeLEiQAUFRWxadMmyJkGfdUlFY4aVjs4eZC+vXv3brBnQO8L1R31RVCw6JiuY/HixdTU6EH7oPGvuOBz2Dff5zk+QS1WgpoQQoQDCWpCiIjib/jjhRdeqG0v2WI40MqKmnfHRy0M2ZI7vDpxzTXXcPToUebOndviucagdsSTOSsCV9TAHSoipaJmcX8dVjvYUjv1fc8++2xt+/PPP1c3fvI4rphsig15f6IhqO3a5Z7L2IFNRT744APT4wNHgdSR+g6vhjLgp5mIVNSEECIshE67NiGE6AD+gtrQoUMZPHgwu3bt4vvdUF0PibG0uqJmCmopSVBzQH2QPBQ6YVmC1i51YFxw/FB5DNDYbEUN3OuAuRrVIXjW8PsnwBzU3KWsuF6d8vdgZAxqH374ISkpKSxevJglS+qpMOT98YblBHfv3q1u9DobYrPUZi+HP1UbvtjT23wNjY2NfPrpp6Z9B0qscPoX8JF7XbTqfT7PkzlqQggRnqSiJoSIKP6CWnx8PNOnTQPUFvXLPVmmlRU142LXaQkK4G6114Hz09pj0CC9fLO7LEPdqNkHDnXCmr+gVuPpv+Gs6+zL6xTmoOYe+tjJwx5BvdeeZjCrVq3ipptu4sMPP6TCkNLSEmBYb/05WlCz2qDfz9VtVyMcaH6B6kCWLFliej+Ag6VRagiMci8hUL3X53me7wNbFCTYkaAmhBBhQoKaECKi+JujlpCQwLSJg7XH2jy1phpwOX3O92ZaQ81uaGDRwfPT2soU1ErcLeQVF2x5GPBtzw9Q6wlqYTr80bzgtWej84OHxWLh3HPP9dmfkZGhLWR+5Yxceo68WFt3TRv6CDDgWn374LvtugbvYY8AB464qKquRknor+6o2a9+Dxh4glpGkrvwKEFNCCHCQviNexFCiGYkJSVRXGxe1yohIYHTR0RhsahdD03z1JoqISat2dc0DX20GzpHpAS3ota/f38sFguKorD7aAJYY9SKzbbHoN/lLVTUwrOhiCmoxbg3Ornjo8e9997Lzp07aWho4Mwzz+Sss85izJgxWK1WysvLSU1NxWKxMGjQaDZu3Ki16I+Ojoa0UWpjkYYSqDnY5vduamri448/BiA+Po6RPdUOphU1TpKTk/nplBw+vAn1778uH+L7aM/1BPaMRACLuuC1EEKIkCcVNSFERMnIyPDZFx8fT3rTFoa4Cwm7Cg0HWzFPzRTUbIYqVZArana7ndzcXAB278uDEe4FtRUnrPoVJSVHfZ5T2+jeiKSKWhcMfQTIzs5m0aJFLFu2jD/+8Y+MHTuWqKgoLBYLaWlp2txCT6XT4XCwd69hKGKM+3uz0TdAt2T58uVa8D53+ikM9fqSP/qmkApP9jYMf6yvr1fXfsMd1GLS9OUChBBChDQJakKIiOK9FllsjHs9spLVpLhHB9Y1gsszOqyx5XlqpqBmdS9cZbVB4sAAz+g6nlBQVlZGac4N+kLMpWspKdjtc75WUQvTRa+DNfSxLYwLly9btkw/EONuIOKoBJejTa9pHPY46+xx9Ovhe47WfdLQUMSnNX8L1WMhhBChQ4KaECKieFfUEuyo1aOKzcTb9f31ns/Jba2oWQ+rG0mDQ6Jromme2r5DMP4lQK3slBwt8jk/IueodVFFrbWmT5+ubX/55Zf6Abvhe7OxvNWv53K5+Oijj9SXsNs5b/Jx/oOa53cOhoqaMailJwK2lFa/rxBCiOCSoCaEiCjeFbX4GKBkNShONbS5aZWlVnR+NAW1OHfCC/KwRw9TUNu9GzInweCbURQorfY9P6LmqHmCWhfNUWutn/zkJ6SmpgJqp0aXp3wbY2jJ39D64Y8rV66koKAAgLPOOouk6Er6+Onu76+i5rOGWowENSGECBcS1IQQEcVvRa1wCeAObW5aZakVFbXKSv2c5DjPRnAbiXj4BDWAEx+hUulJk7uhZWayfr4+Ry2SglpoDX2MiorijDPOANSgtGHDBvWAce20Rt+OnIG8//772vasWbOgrpAT+oLVa+m4I54+NzV6Rc30S4YEpKImhBBhRIKaECKieFfUjEEtIVbf35aKWmNjo7athYMQrKhp7eBtyZT2v0/bn2vIrjWRNPQxBoiKBVtq0K4nEL/DH2MMfxFtqKgtXboUUAPgBRdcAGUb6JUG82+BAf1ztfOKaxPVjfJN2rITPkEtJrVtX4gQQoigkaAmhIgofitqpasBiI/V55RplaVWVNSMQS3G8xJBbs3vMXDgQKxW9Uf55s2btf0lMXpDi76GW1Ib5kMfa2r0gBlnQx32aLEEfkKQ+A1q7ayoHT6szovs06cPaSnJULIKgNnTevPmm/ri2cWNOeqGoxLK1gNSURNCiHAmQU0IEVH8VtTcCwDHp+Ro+2vbUFFraGjQtrWgljTkWC6zw8TFxTF0qFrd27JlixYqSwxzk/oa8kG4V9RMXQyTCLlhjx6DBg3Slk745ptv1EpgOypqDoeDo0fVZRZ69uwJFZuhyT35sMdEMrOytHOL6wwhrOgrAMrLy7VdEtSEECK8SFATQkQUvxU1z3aaPkyspg1z1Dzhx2qB6Cggvi/YEo/1UjvMmDFjAPVD/ZYt6mre+fn52nFTRS3M56gdOXIEALvN/XcbYo1EPCwWi1ZVq6+vZ8WKFe2qqBUV6Z07c3Jy4OhK/WDmJLIMQe1ItWF9tCJ1uKSxopYajwx9FEKIMCJBTQgRUXy6PhqCWnz6AG1bCyytWEfNE9S0alqIzE/z8AQ1gPXr17NmzRruvPNObd/gc+YRHa1evN71MTwrap7qUmaSe8RjiK2hZuQz/NHUnr91Qc3T7RHcFbUjK/SDPSaSmJhIbKw6+bK4tFqvMB75BlwOGfoohBBhTIKaECKiNFtRy9SHK+pDAFuuqHmGPtpDrJGIhzGovf7660ybNk0b8jZp0iTOu/wOEhISAOM6auFXUVMURQtqPZLcO9PHBH5CkHk6P4I7qLWjPX9hYaG23bNnT72iZrVD2hgsFotWVSsuLoas09XjTTVQskaGPgohRBiToCaEiCgxMTGmx1pQi80hPrW3tl8LLO2pqMVmH+NVdqwTTzxR2/7666+pqlL7tE+dOpVFixZhs9mIj48HwnsdtcrKShwOdR07LahlnRa8C2pBdnY2o0aNAmDt2rWU1hianrSjopaTEQ/V7iUY0k+CKPWbOzMzE1Crja6sqfqTi74yVdRS4pF11IQQIoxIUBNCRDQtqGWMJ95dVQLjXC0/q0J78VTUYqLcO6LiAp8cBOnp6fTr18+0b8aMGXz22WckJqpz6bSKmvZ1h9/QR081DdShj8T3hYTjgnY9reEZ/qgoCku/WQsW9zdRKytqpqGP8eX6gcxJ2qanouZyuSi1GSqMRUu1oJYU655fGYJLGQghhPBPgpoQIqLFeQpsPSZoYQWgxuEex+isa/E1PBU1behjdGgFNYBx48Zp2xdccAGffPKJVkUDPajVhOPQR2c9lG3gaN4ObVePJCBraki25jcyzVNbskQf/tjKippx6GOO7ZB+oMdEbdPYUKS4Jg4S3KH9yHfa0Mc0z7e+DH0UQoiwEd3yKUIIEb4cTe6NjPHEF0dp+2sd0YCjVYHFZ+hjiFXUAO69914OHz7MySefzKOPPuozBNQT2hoc4HRBVIhW1BZ+/CGrvv6IOT8dRM/oPep6YJXbQHFy5Ed9xfIeSUB26A579JgyZQo2mw2Hw6HOUzsvAxqOtK+iFrUdXO4HgYLakSMMzz4d9r4KrgbKytT3SfUENRn6KIQQYUOCmhAiotU53BvpY0lI2Kntr210h7ZWVNS0ZiIhHNRGjx6ttoAPwFhNrG2ApBCco1aYf4hLLr2ExiaFt96EL++BAXoG4Wh5vbatBrXTu/4i2ygxMZGJEyeyfPlydu/ezf7Sn3BcNOqQW2cjRMU0+3xPRc1isZDl/BEsqMM9DevHeeaogXv5grFnwN5XqWuEhgb1fwCpqAkhRPiRoY9CiIhW34japTEm1TQUsEYLas0HFkVR/FTU4gM/IUQZv/baRkJy6OPqJW/R2KQAsO8ITJkL2/KjIHUUJA7kqGE6Yebx0yFxQIBXCi3G4Y9LNjv1A60Y/uipqPXISMVmcQfVHpNM5xgrakVFRVqALTMUTdMSgKhYrQGJEEKI0CdBTQgRca655hpte+JgoM9FgHdYcf/4a6Gi5nQ6URQ1PGhBLQTnqLXEND+vgZBsJrJ+9VLT4/wyOPWvaazv+SpM+4ojNfp97zH+d118de1nmqe23tBltIWgpiiKVlHrmWH4njMMewT3QthuhYWFEN8HkgZTbvgrTo1HqmlCCBFmZOijECLiPPHEE9TW1tIzM5lZd10EPc8E/IQVAJcDXE1g9f/j0DPsEQzNREJw6GNLTCG1gZBrz19fX8+GH7dojwcN6MvuvYc4evQop59+Ov/97385mnIh8A4APbJyArxS6Bk3bhxJSUlUVVWxZF0RrmvAagVq8yBleMDnlZWVadXcniku/UCmuaLWq1cvbTs/P1/dyD6DsrW7tP2yhpoQQoQfqagJISJOeno67777Ln9/7l9Y+pyvDfcyV9QM3QKbqap5PihD6Lbnb41QrqjddNNNxMXF8fE3eQAkxcHq1Ws55ZRTAKioqODMM89kwfsLtecY52WFuujoaE4/XR2OeKSsjk2e5o1l65t9nmkNtQR1bTyi4tWhoAb+g9rpvkMfY1Lbc/lCCCGCRIKaEKLbMAc1RT/QzHwtY0UtlLs+tsRvRU1RAj+hiyiKwgsvvGDaN3pAEmkZWSxatIgzz1SrobW1tdTW6n9PGRkZXXqdx8o0/HGze6P0h2afY+r4mOhOXRnjfaq/6enp2O3qLyO0oJY1lXLDt7UMfRRCiPAjQU0I0W1ERUVpH2hr6g0hpZUVNX0dtfBrJuJTUVNcULU7eBfkVlfne+9PHNEfUK954cKFXHTRRabjKSkp2Gw2n+eFMmNQ+2CNVc3IpeuafY5xDbWeqe4Nr/lpoHaE9FTVtKAWl02ZSx8eKkMfhRAi/EhQE0J0K57KUm2DYc5PM/O1TEMfI6Wi5vmSDr0XnIsxqKys9Nk36sSTtG273c5rr71GUlJSV15Whxs6dCiDBg0CYOUuF88vAar3QGN5wOeYhj6muje85qd5eIJaSUmJVgUusw7RjmtdH4UQQoQNCWpCiG7FU1mqqTe0SW+momZqJhLGQc1vI5UD7wTnYgwqKip89p16+rmmx8nJyaZOnv6eE+osFgvPPfec9viut+BIJc3OUzMNfUx1b2Sc7Pdcf/PUymNGavvUoY/Jbb5uIYQQwSNBTQjRrWgVtfomfWczc9QisqIW3U/dKN8IFduDdEUq74ra27fCkBFjfM674447tO2f//znnX1ZnWLGjBlcd911gBqWv9xMs/PUjEMfc1KBpOMhtoffc/0FtSMl5dq+tASg1zntvXQhhBBBIEFNCNGteCpLpqDW2q6P0YAlOmAr/1BmqqjF6ZUWDr4bhKvRGYPaPTPh8on4nUs1cOBAPvjgA2644Qbmzp3bhVfYsS677DJt+7udwN5XwNng91yfipqf+Wke3kHt+++/55131IppdHQ0vS77Hnqff0zXLoQQomtJUBNCdCueylKjw0mTZ/RjW4Y+hmEjETAHtdqYQfqBIAc14zDGZE+hMsZ/04uLL76YF154gQEDBnTBlXWOiRMnYrWq//R+uwOo2AKbH/Z7rqeilhir/gk0Pw3MQW3Dhg1ccsklOBwOAG6//XZS+03omC9ACCFEl5GgJoToVnza1EPbhj6G4bBHMH/dNY5o6OH+0F+xBcq3BHhW5zNW1FLiUNcJs4ZXR8e2SE5OZuRItaK56RBU1gLb/goNJT7neipqOZ7c2kxFrXfv3tr2o48+yuHDhwGYMmUKjz76aMdcvBBCiC4lQU0I0a2YKkueDNbaipqNsA1qxq+7qqoK+ulD8IJZVTMGteQ4usWizJMnTwbApcD3uwGXA/I+MZ1TV1enVRt7pqI2AkkeHvA1jRU1l0vtaJqTk8M777wTdksZCCGEUElQE0J0K6bKUjeqqBk/yH/yySfU9TgPsKg7Dr4btMWvfYY+doO1vk455RRt+9ud7o2D5qUSfNZQy5gA1qiAr2n8+wV1Xtp7771Hz549j/VyhRBCBIkENSFEt2KqqHmCWluaiYRpUMvJyWHWrFmAGgJeeGMhZE1RD1Zuh/JNQbku09DHeLpFUJswQZ8vtqXA/f1U+CU0lGr7Cw7t1bZzUoHsM5p9zaSkJBITE7XHf/vb37TKnRBCiPAkQU0I0a34rag1s+B1pDQTAbj//vu17Xnz5uHsfYl+MEjDH7vj0Mc+ffpo2wW1qeqG0mQa/liw6iltu2dOFhx/a4uve+ONN2KxWLj55pv59a9/3VGXK4QQIkgkqAkhuhVTM5FWzFGLlIoawKhRo7jgggsAtVHFqsLjwOL+Z+DAO50z/LGx+cWpu+PQx5iYGDIyMgAoKDcc8Ax/rD9K4bZF2u6eJ/8GbIm05LHHHqOiooJ//OMfWCyWDrxiIYQQwSBBTQjRrfgd+tgN5qh5XHzxxdr2f75YCVmnqQ+qd0PZBnA1wfKfwv/GQPW+Y3uzDffA+6mwJnA1yGfoYzeoqAHa3LGColKUuL7qzqIvobEMdj5LQalDOzenv+8C4IEkJSV16HUKIYQIHglqQohuxf/QxzasoxbmQe2cc87Rqi3/+c9/oK9h+OPhhbDvNcj7Eh0ZegAAHhBJREFUWA1t6+86tjfb+oj6313P+W0/D+aglhRLt6iogR7UGhoaKE9Tq5y4HHDwfVw7nmHlbt9zhRBCdC8RE9Sqqqq444476NevH3FxcUyaNIk1a9ZoxxVF4f7776dnz57ExcUxffp0du3aFcQrFkIEg7Gi1q6uj9HhHdSys7MZP348AJs2beKAU6/WKHkL2bb6U+o8X3Lex+1/I5fT/DjvU7+neYJaXAzYogm42HWkMYavApuh6cemP/HAW6V85V7arkePHhx//PFdfHVCCCFCQcQEteuvv54vvviCN954g02bNjFjxgymT5+uLfo5b948nn76aZ5//nlWrVpFQkICZ511FvX19UG+ciFEVzIODavy/O/fpnXUwreZiMf555+vbX+2dD2kjgbg6bfXMvyKTxl7LzhdQExa+9+k0auCduhDv6d55qileG6rLbX97xlGTEGtLhPi1eGPC77K5+GP1f1Wq4X58+cTGxsbhCsUQggRbBER1Orq6vjggw+YN28ep556KoMGDeJPf/oTgwYN4p///CeKovDUU09x7733MnPmTEaNGsXrr79Ofn4+H3/8cbAvXwjRhYwtzKu1oNZ95qiBV1D77DPorT6+4w1139bDsK8YiD6G+U51hebHhYvBUeVzmqeiluy5rd1s6CNAQWEh9L2EDfvhuhf0cx77yx8566yzuv7ihBBChISICGpNTU04nU6f3zrGxcXx7bffsm/fPgoLC5k+fbp2LCUlhQkTJrBy5cqAr9vQ0EBlZaXpjxAivPkPas10fazR17aKiSIigtro0aPp3bs3AEuWLKEmdTp5XgWwo1WojS3aq77I/NjVCMXLTbsURdF+rqZ4bms3aSZiXKA6Pz+fhpyLuPp5qHf3ELluWiJ3/G5ukK5OCCFEKIiIoJaUlMTEiRN56KGHyM/Px+l0Mn/+fFauXElBQQGFhepvdrOzs03Py87O1o7588gjj5CSkqL96du3b6d+HUKIzud36GOgOWrOBhp2vaE9tNsI+zlqABaLhfPOOw9QfyH11Y8VfLEj1XTOqj0w951ytm5c1743qffzs9UrvNXkr8PlcgHdvKJWUMDc5z5n0yH18ehc+OcDl0qLfSGE6OYiIqgBvPHGGyiKQu/evbHb7Tz99NPMnj0bq7X9X+Ldd99NRUWF9ufQoUMdeMVCiGAwVdQaotSNQBW1svU01ldrDyNl6COYhz/+57P/8sXuLNPxO96ABz6A82dehNPp9H56s1wuFy++/gkPfQSOJsMBr86PlRue0ra1oNYNm4l88cUXPPrXvwJgi4LXbgJ73zODdWlCCCFCRMQEtYEDB/L1119TXV3NoUOHWL16NQ6HgwEDBpCTkwNAUZH5t7lFRUXaMX/sdjvJycmmP0KI8GYKao0tBLXq/TQagobanj/8m4kATJs2TRsuvnDhQhavKfJ73r79eSxbtqxNr/3ss89y458+4P734aWlhgONpabzKqv1Ri3duZnIli1btMri/Vf2Y/Tki6HvrGBdmhBCiBARMUHNIyEhgZ49e1JWVsaiRYuYOXMm/fv3JycnhyVLlmjnVVZWsmrVKiZOnBjEqxVCdDXz0Ef3j8BAQx+rdtBgCGqRVFGLj4/njDPOANShdyWlFQHPnT9/fqtfd+vWrfz+97/XHq8wroLiVVGrqNOH9nW3ilp8fLzPL//Gjh3LH/61G6Z8AFExQboyIYQQoSJigtqiRYv4/PPP2bdvH1988QWnn346Q4cO5brrrsNisXDHHXfw8MMP8+mnn7Jp0yauvvpqevXqxUUXXRTsSxdCdCHjOmrVLS14XbnDVFGLhHXUjIzDH5vzwQcfUFtbC7V58O1l8L+fQPkWn/MaGxu56qqrTMuebDaOGPeuqJUf1ba1oBadSHdhrKrZ7XZee+01oqOjg3hFQgghQknE/ItQUVHB3XffTV5eHunp6cyaNYs///nP2Gw2AH73u99RU1PDDTfcQHl5OZMnT+bzzz+X9WmE6Gaio6OJjY2lvr6+5fb8VTtpcOgP1XXUIieoeRqKtKSqqoqF//o/Luv5Njjclbd1t8O0Jabz5s6dyw8//GDat70AmpwQHYXvHLUKPbhpQx8tEfP7wxb16dOHHTt2APDQQw8xfPjwIF+REEKIUBIx/yL+7Gc/Y8+ePTQ0NFBQUMCzzz5LSoo+hMZisTB37lwKCwupr6/nyy+/5Pjjjw/iFQshgsUz/LGqTlF3uBzgajKfpCj+K2oRFNRyc3NNj88880xiYvwPuZv/6vN6SAMo+gpK1mgPV6xYwSOPPAKoYXhkrjr/r8EBe46ovzDzrqjtPqBX1FIjY+pfm9x2222kpqZy5ZVXcueddwb7coQQQoSYiAlqQgjRWp6GItX1Ln2n9/DHugIURzWNhoaH6jpqkZUo7rvvPm3717/+NZmZmX7P+3wjHKkE0n6i7dvx+b1s2rSJqqoqrrrqKq0hxp8euJ9Z4/R7u6XIPdzUUFFTFIXXvyzWHp86lIi7ty2ZOXMmJSUlvPHGG0RFRQX7coQQQoQYCWpCiG5HC2p1hhTmFdTWfbuQgb+BrwxTsSJlHTWj3/72t1x//fU89NBDnHvuuWRlmdv0n+4ejdfkhHeP3gAzVkBsDjvyYfhVixk9ejTp6ens3bsXgEmTJvH7X1/HiN6K9hqbD7urdIaK2urVq9l6SB1XOmUIHD/iJJjyfid+paHpWJaQEUIIEdkiZo6aEEK0lmfoY12DS58/Zej8WF9fz9izbjI9x2KBKCsRNfQRIDk5mZdeekl77B3U/nABLN2qbs//bCNz7rVD5mSWffY+LgVAoalJHR+akJDA6y89SXTBp5zQV3+NzYfc1TVnHTTVQXQczz7ztHb8F+cPhLPXdsaXJ4QQQoQt+VWeEKLbMa6lVuOn8+NDDz3k8xx7tBrWsET2EDXvoY/Tbv6QUaNGAfD999+zZ88eSMil3E//laeuS2Lghgmwdg6Dst1z+oAtB/X10mgsZf78+cx/8y0AkmLhkmmDO+VrEUIIIcKZBDUhRLdjWvTaq/Ojw+HgmWee8XlOTDRgz4T4vj7HIol3RS2q30/5+c9/rj1+8803IT6X0mrz8y47GX55cqH2ODoK+mep/8QcLNbb9a9f8y2/+tWvtMeP/xwSU7M78ksQQgghIoIENSFEt2Ne9Nq94R76uGbNGqqqqnyeExObCNOXQZS9C64weIzdcj1mz56NxaIuTj1//nyU+FzKavTjr98Eb80BizUaMibAsLvgtIVkDzwZgOpaB7UNUFIFF191h7bO2vVT4VdnAPaMzv6yhBBCiLAjc9SEEN2O34qaO6h99dVXfp9jj0+BlMhf58rhcPjs69u3L1OnTmXp0qXs2rWLNTuqTEHttGFg7TcLJr4G0fqC4lnZr2nbBeVwyyuw/5BadRt/4mCevXaXelCCmhBCCOFDKmpCiG7H/9BHNXksWbLEzzMIuL5YpMnI0ENT3776MM8rr7xS257/0TeUGoJaeiLQY5IppAFkZ+tDGm95BRZvUrczMzN5/+lfqV00AWIkqAkhhBDeJKgJIbodY1AzDn2sq6tjxYoVAPTPiWFYL/053SWo/fKXv6RXr17YbDYWLFig7Z81axZ2uzrsc8H7n1BcqQ6FjI6CBDuQeoLPaxnnu3lCWlSUlXfffZe+GXr7fuw9Ov4LEUIIIcKcBDUhRLdjnKNmbCby3Xff0djYCMC0kbH0SdefU13t1T0jQiUlJbF3714KCwuZNGmStj8lJYULL7wQgCNHjrD5kBq00hLc3TBTRvi8lndjEoBLph3P1KlTTYtfy9BHIYQQwpcENSFEtxNojppxftoZI6CvIT8UFxd30dUFn91uJz093We/sfujR7pntGNcL59jxqGPHgNyPItfS1ATQgghmiPNRIQQ3Y7/oY81pvlpZwxtYMs+/Tn+mmx0N+PHj/fZl5YARMW7y2pm/ipqx8Vsgy9Ph5r9+k6ZoyaEEEL4kIqaEKLb8Tf0saK8lLVr1wIwYsRwspMayEoOxtWFruzsbGw2m2lfWgJgS/J7vr+g1i/DAcXLzEFNKmpCCCGEDwlqQohux9/Qx69X78LlcgEwbepkALJ9lxTr1qxWK3369DHtS08EBt3g93y/Qa2P13DIxEERvzadEEII0R4y9FEI0e34G/r41Wp9nOMZk08ClwQ1f3Jzc9m3T79XaTnHw/A/+D03OTkZu91OQ0OD/vxf7AVLJZSug+q90OvcTr9mIYQQIhxJUBNCdDv+hj4uWXMIUKtGp508DFbAxMHQNzuBQ0U1zJ07NxiXGnKMa6sBpA2/DKLj/Z5rsVjIysri0CH13mZmZhIfHw/EQ+/zOvtShRBCiLAmQU0I0e2Yhj42QFEFbN5bAcBJJ51Eqjt32G2w4tUr2aTM5MwzzwzGpYYc76DmrzukUVpamhbUvJ8rhBBCiMAkqAkhuh3T0Mc6WLpVPzZt2jRwVGiP+/TuTZ+R53Tl5YW03Nxc0+O0tLRmzzd2yzTedyGEEEI0T5qJCCG6nYSEBG27uh6+2qIfO+OMM8BRqe+wSetHI5+hjy0Etbq6Om07Li6uU65JCCGEiEQS1IQQ3Y7VatXCWnUDfOWuqMXExHDKKadIUGuGd0WtpaGPtbW12rYENSGEEKL1JKgJIbolzzC87fmwp0jdN3HiRLXZhQS1gNpaUbvnnnu07V/84hedck1CCCFEJJI5akKIbikpKYmioiKcLn3fCSecoG5IUAsoJcW8ZkFLQe3GG2/k6NGjpKWlcf7553fmpQkhhBARRYKaEKJb8tfYIjnZHcokqAVksVhMj1sKarGxsTz00EOdeUlCCCFERJKhj0KIbkmCWvvdfffdAEydOlXmnQkhhBCdRIKaEKJbMi56re3zhDcJas3685//zMaNG1m8eHGwL0UIIYSIWDL0UQjRLfmtqCXGqhumoJbic153Z7FYGDlyZLAvQwghhIhoUlETQnRL/oJaUrxN3dCCmgWiE3zOE0IIIYTobBLUhBDdkt+hj56g1uQOarZk8GqeIYQQQgjRFSSoCSG6Jb9DHxOi1I3GcvW/Mj9NCCGEEEEiQU0I0S35HfoYFwWlP0C9ewXshNwuviohhBBCCJUENSFEt+Rv6GNyvBV2Pa/v6H91F16REEIIIYROgpoQolvyW1GLqoT9b6oPopOg3xVdfFVCCCGEECoJakKIbsk7qFkskJD/Gjhr1R39rwKbb5gTQgghhOgKEtSEEN2S99DHRDtYjyzTdwy+qWsvSAghhBDCQIKaEKJb8q6oJccZHmSeAqmyoLMQQgghgkeCmhCiW/IOaknGoDZIqmlCCCGECC4JakKIbsl76KNWUbNnQO4lXX9BQgghhBAGEtSEEN2ST0Ut1r0x4DqIivV9ghBCCCFEF5KgJoTolgIOfRx0Y9dfjBBCCCGEFwlqQohuKS4uzvQ4OQ7IOROSBgXngoQQQgghDCSoCSG6JYvFYnqclJIJP3kySFcjhBBCCGEmQU0IIYDkkddD6ohgX4YQQgghBCBBTQghAN8ukEIIIYQQwSRBTQghkKAmhBBCiNAiQU0IIQCbzRbsSxBCCCGE0EhQE0IIwGqVH4dCCCGECB3yyUQIIfDtAimEEEIIEUwS1IQQAkhNTQ32JQghhBBCaCSoCSG6rcWLFxMTE8OIESOYOXNmsC9HCCGEEEITHewLEEKIYDnzzDMpKioiKSmJqKioYF+OEEIIIYRGgpoQoluTIY9CCCGECEUy9FEIIYQQQgghQowENSGEEEIIIYQIMRLUhBBCCCGEECLESFATQgghhBBCiBAjQU0IIYQQQgghQowENSGEEEIIIYQIMRLUhBBCCCGEECLESFATQgghhBBCiBAjQU0IIYQQQgghQowENSGEEEIIIYQIMRLUhBBCCCGEECLESFATQgghhBBCiBAjQU0IIYQQQgghQowENSGEEEIIIYQIMRLUhBBCCCGEECLESFATQgghhBBCiBAjQU0IIYQQQgghQowENSGEEEIIIYQIMRLUhBBCCCGEECLESFATQgghhBBCiBATHewLCCeKogBQWVkZ5CsRQgghhBBCBJMnE3gyQkeToNYGVVVVAPTt2zfIVyKEEEIIIYQIBVVVVaSkpHT461qUzoqAEcjlcpGfn09SUhIWiyXYlxMSKisr6du3L4cOHSI5OTnYlxNx5P52DrmvnUPua8eS+9k55L52DrmvnUvub+c41vuqKApVVVX06tULq7XjZ5RJRa0NrFYrffr0CfZlhKTk5GT5wdGJ5P52DrmvnUPua8eS+9k55L52DrmvnUvub+c4lvvaGZU0D2kmIoQQQgghhBAhRoKaEEIIIYQQQoQYCWrimNjtdh544AHsdnuwLyUiyf3tHHJfO4fc144l97NzyH3tHHJfO5fc384R6vdVmokIIYQQQgghRIiRipoQQgghhBBChBgJakIIIYQQQggRYiSoCSGEEEIIIUSIkaAmhBBCCCGEECFGgloEeuSRRxg3bhxJSUlkZWVx0UUXsWPHDtM59fX1zJkzh4yMDBITE5k1axZFRUXa8R9//JHZs2fTt29f4uLiGDZsGH//+99Nr/Hhhx9y5plnkpmZSXJyMhMnTmTRokUtXp+iKNx///307NmTuLg4pk+fzq5du7Tjy5Ytw2Kx+P2zZs2aY7w7xy7c7y/ADz/8wJlnnklqaioZGRnccMMNVFdXH8NdOXahfl8//PBDZsyYQUZGBhaLhQ0bNvic8+KLLzJ16lSSk5OxWCyUl5e36150tK66t99++y2nnHIKGRkZxMXFMXToUJ588skWr68137N//vOfmTRpEvHx8aSmprb/ZhyjSLiXF154Ibm5ucTGxtKzZ0+uuuoq8vPzj+GuHLtIuK/HHXecz79Zjz766DHclY4R7vc2lD8ThPu9hdD8PAChf2+77DOBIiLOWWedpbzyyivK5s2blQ0bNijnnnuukpubq1RXV2vn3HTTTUrfvn2VJUuWKGvXrlVOPvlkZdKkSdrxl19+Wbn99tuVZcuWKXv27FHeeOMNJS4uTnnmmWe0c379618rf/3rX5XVq1crO3fuVO6++27FZrMpP/zwQ7PX9+ijjyopKSnKxx9/rPz444/KhRdeqPTv31+pq6tTFEVRGhoalIKCAtOf66+/Xunfv7/icrk6+G61Xbjf38OHDytpaWnKTTfdpGzfvl1ZvXq1MmnSJGXWrFkdfKfaJtTv6+uvv648+OCDyksvvaQAyvr1633OefLJJ5VHHnlEeeSRRxRAKSsrO+b70hG66t7+8MMPyltvvaVs3rxZ2bdvn/LGG28o8fHxygsvvNDs9bX0PasoinL//fcrTzzxhHLnnXcqKSkpHXdz2igS7uUTTzyhrFy5Utm/f7/y3XffKRMnTlQmTpzYgXep7SLhvvbr10+ZO3eu6d8u4/UHS7jf21D+TBDu9zZUPw8oSujf2676TCBBrRsoLi5WAOXrr79WFEVRysvLFZvNprz33nvaOdu2bVMAZeXKlQFf55ZbblFOP/30Zt9r+PDhyoMPPhjwuMvlUnJycpS//e1v2r7y8nLFbrcrb7/9tt/nNDY2KpmZmcrcuXObfe9gCbf7+8ILLyhZWVmK0+nUztm4caMCKLt27Wr+i+1CoXRfjfbt2xfwh7LH0qVLQyqoeevKe/vTn/5UufLKKwMeb+vPhFdeeSWoQc1bON9Lj08++USxWCxKY2Njs+/flcLxvvbr10958sknW/rSgi4c761RKH8mCLd7Gy6fBxQltO6tUWd/JpChj91ARUUFAOnp6QCsW7cOh8PB9OnTtXOGDh1Kbm4uK1eubPZ1PK/hj8vloqqqqtlz9u3bR2Fhoem9U1JSmDBhQsD3/vTTTykpKeG6664L+LrBFG73t6GhgZiYGKxW/X//uLg4QB0CECpC6b5Gmq66t+vXr2fFihWcdtppAc9pz8+EUBLu97K0tJQ333yTSZMmYbPZAr52VwvX+/roo4+SkZHBmDFj+Nvf/kZTU1PzX2gQhOu99QjlzwThdm/D5fMAhNa97UoS1CKcy+Xijjvu4JRTTuGEE04AoLCwkJiYGJ95HtnZ2RQWFvp9nRUrVvDOO+9www03BHyvxx57jOrqan72s58FPMfz+tnZ2a1+75dffpmzzjqLPn36BHzdYAnH+3vGGWdQWFjI3/72NxobGykrK+MPf/gDAAUFBc1/wV0k1O5rJOmKe9unTx/sdjtjx45lzpw5XH/99QGvpz0/E0JFON/L3//+9yQkJJCRkcHBgwf55JNPWvx6u0q43tfbb7+dBQsWsHTpUm688Ub+8pe/8Lvf/a5VX3NXCdd7axSqnwnC8d6Gw+cBCL1725UkqEW4OXPmsHnzZhYsWNDu19i8eTMzZ87kgQceYMaMGX7Peeutt3jwwQd59913ycrKAuDNN98kMTFR+/PNN9+0+b3z8vJYtGgRv/zlL9t9/Z0pHO/viBEjeO2113j88ceJj48nJyeH/v37k52dbfqtWjCF430NF11xb7/55hvWrl3L888/z1NPPcXbb78NRN69Ded7edddd7F+/XoWL15MVFQUV199NYqitPvr6Ejhel/vvPNOpk6dyqhRo7jpppt4/PHHeeaZZ2hoaGj319HRwvXeeoTyZ4JwvLfh8HkAwvPedpg2D5YUYWPOnDlKnz59lL1795r2L1myxO9Y2dzcXOWJJ54w7duyZYuSlZWl3HPPPQHf5+2331bi4uKU//znP6b9lZWVyq5du7Q/tbW1yp49e/yO5T311FOV22+/3ee1586dq2RmZobU3AmPSLi/hYWFSlVVlVJdXa1YrVbl3XffbcVX3rlC8b4ahfMcta66t0YPPfSQcvzxxyuK0jHfs6EyRy0S7qXHoUOHFEBZsWJFq66jM0XSfd28ebMCKNu3b2/VdXS2SLi3ofqZIBLubSh+HlCU0Ly3Rp39mUCCWgRyuVzKnDlzlF69eik7d+70Oe6ZgPn+++9r+7Zv3+4zAXPz5s1KVlaWctdddwV8r7feekuJjY1VPv7441ZfW05OjvLYY49p+yoqKvxOHHa5XEr//v2V//u//2vVa3eVSLm/Ri+//LISHx8f1GARyvfVKByDWlfeW28PPvig0q9fv2avrS3fs8EOapF0Lz0OHDigAMrSpUtbfS0dLRLv6/z58xWr1aqUlpa2+lo6Q6Tc21D8TBAp99YoFD4PKEpo31sjCWqizW6++WYlJSVFWbZsmamdrfG3ADfddJOSm5urfPXVV8ratWt92jNv2rRJyczMVK688krTaxQXF2vnvPnmm0p0dLTy3HPPmc4pLy9v9voeffRRJTU1Vfnkk0+UjRs3KjNnzvRpc6woivLll18qgLJt27YOujMdIxLu7zPPPKOsW7dO2bFjh/Lss88qcXFxyt///vcOvEttF+r3taSkRFm/fr3y2WefKYCyYMECZf369UpBQYF2TkFBgbJ+/XqtXe/y5cuV9evXKyUlJR14p9quq+7ts88+q3z66afKzp07lZ07dyr/+te/lKSkJOWPf/xjs9fXmu/ZAwcOKOvXr1cefPBBJTExUVm/fr2yfv16paqqqgPvVMvC/V5+//33yjPPPKOsX79e2b9/v7JkyRJl0qRJysCBA5X6+voOvlutF+73dcWKFcqTTz6pbNiwQdmzZ48yf/58JTMzU7n66qs7+E61XbjfW49Q/EwQCfc2FD8PKEro39uu+kwgQS0CAX7/vPLKK9o5dXV1yi233KKkpaUp8fHxyk9/+lPTN9cDDzzg9zWMv2E47bTT/J5zzTXXNHt9LpdLue+++5Ts7GzFbrcr06ZNU3bs2OFz3uzZs03rYYSKSLi/V111lZKenq7ExMQoo0aNUl5//fWOuDXHJNTv6yuvvOL3eQ888ECL72/8GoKhq+7t008/rYwYMUKJj49XkpOTlTFjxij/+Mc/TK2f/WnN9+w111zj9/27ugoU7vdy48aNyumnn66kp6crdrtdOe6445SbbrpJycvL67B71B7hfl/XrVunTJgwQUlJSVFiY2OVYcOGKX/5y1+CGn49wv3eeoTiZ4JIuLeh+HlAUUL/3nbVZwKL+2YIIYQQQgghhAgRodPSRQghhBBCCCEEIEFNCCGEEEIIIUKOBDUhhBBCCCGECDES1IQQQgghhBAixEhQE0IIIYQQQogQI0FNCCGEEEIIIUKMBDUhhBBCCCGECDES1IQQQgghhBAixEhQE0IIIYQQQogQI0FNCCGEEEIIIUKMBDUhhBBCCCGECDES1IQQQgghhBAixPw/8qBmhKr4ld8AAAAASUVORK5CYII=",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Visualising Predictions\n",
"visualizePreds(data_filtered1, train_data_len1, y_pred1, date_index1, \"orange\")"
]
},
{
"cell_type": "markdown",
"id": "da42add1-1f67-43d3-9e7c-9d0dc2e7b495",
"metadata": {},
"source": [
"#### Where is my Train data? Fix this later.\n",
"\n",
"### Predict tomorrow's price"
]
},
{
"cell_type": "code",
"execution_count": 37,
"id": "8becd22c-a2a2-4e81-ae80-43bd98c1a60e",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"1/1 [==============================] - 0s 28ms/step\n",
"The close price for GOOG at 2023-10-19 was 139.28\n",
"The predicted close price is 140.5500030517578 (+0.9%)\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/tmp/ipykernel_15950/3096130203.py:16: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`\n",
" price_today = np.round(new_df['Close'][-1], 2)\n"
]
}
],
"source": [
"predictFuturePrice(model1, df, sequence_length1, features1, scaler1, scaler_pred1)"
]
},
{
"cell_type": "markdown",
"id": "9f3d7414-56b0-4063-92a7-401dc622bd8c",
"metadata": {},
"source": [
"### Accurate prediction within 0.9%!\n",
"### Now let's add Sentiment Analysis"
]
},
{
"cell_type": "markdown",
"id": "7f445d28-9f12-4347-a4ca-8510b883aaa9",
"metadata": {},
"source": [
"## BERT Sentiment Analysis\n",
"\n",
"##### Going to use tweet data"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "463ece8c-b7e5-471f-9af6-727a1338f89f",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import csv\n",
"import json\n",
"import datetime\n",
"import pandas as pd\n",
"from tqdm import tqdm\n",
"import snscrape.modules.twitter as sntwitter\n",
"\n",
"path = \"./GOOG_tweets.json\"\n",
"\n",
"if os.path.isfile(stockname + '_tweets.json'):\n",
" print(\"File already exits...no need to scrape it again\")\n",
" print(\"To scrape again, delete : \" + stockname + '_tweets.json')\n",
"else:\n",
" maxTweets = 3\n",
" start_date = date_start.date()\n",
" period = (date_today - start_date).days\n",
"\n",
" tweets = dict()\n",
" for i in tqdm(range(period)):\n",
" dayTweets=list()\n",
" start_interval = start_date\n",
" end_interval = start_interval + datetime.timedelta(days=1)\n",
" try: \n",
" for i,tweet in enumerate(sntwitter.TwitterSearchScraper('#Apple + OR @Apple + since:' + str(start_interval) + ' until:' + str(end_interval) +' -filter:links -filter:replies lang:\"en\" ').get_items()):\n",
" if i > maxTweets :\n",
" break \n",
" dayTweets.append(tweet.content)\n",
" key = start_date.strftime('%d/%m/%Y')\n",
" tweets[key] = dayTweets\n",
" start_date += datetime.timedelta(days=1) \n",
" except Exception as e:\n",
" print(i,e)\n",
" pass\n",
"\n",
" with open( stockname + '_tweets.json', 'w') as fp:\n",
" json.dump(tweets, fp)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "72ccf9fe-9fac-475a-8eb8-a42f63e14150",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" Open | \n",
" High | \n",
" Low | \n",
" Close | \n",
" Adj Close | \n",
" Volume | \n",
" Unnamed: 0 | \n",
" NEG | \n",
" NEU | \n",
" POS | \n",
"
\n",
" \n",
" | Date | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
"
\n",
" \n",
" \n",
" \n",
"
\n",
"