diff --git a/core/gcns/gsaint_main.py b/core/gcns/gsaint_main.py index 0f0e3a715e..54863db9a8 100644 --- a/core/gcns/gsaint_main.py +++ b/core/gcns/gsaint_main.py @@ -101,4 +101,4 @@ def get_loader_RW(data, batch_size, walk_length, num_steps, sample_coverage): # best_auc_metric, result_all_run = trainer.result_statistic() - print(f"best_auc_metric: {best_auc_metric}, result_all_run: {result_all_run}") \ No newline at end of file + # print(f"best_auc_metric: {best_auc_metric}, result_all_run: {result_all_run}") \ No newline at end of file diff --git a/core/gcns/heart_main.py b/core/gcns/heart_main.py index dad44828ef..68d835f36a 100644 --- a/core/gcns/heart_main.py +++ b/core/gcns/heart_main.py @@ -17,7 +17,7 @@ from graphgps.network.heart_gnn import (GCN, GAT, SAGE, mlp_score) from data_utils.load import load_data_lp from utils import Logger, save_emb, get_root_dir, get_logger, config_device, set_cfg, get_git_repo_root_path -from trainer_heart import train, test, test_edge +from core.graphgps.train.trainer_heart import train, test, test_edge def get_config_dir(): diff --git a/core/gcns/text_plot.py b/core/gcns/text_plot.py index a054b92d3f..06b3eeb72b 100644 --- a/core/gcns/text_plot.py +++ b/core/gcns/text_plot.py @@ -1,31 +1,32 @@ import matplotlib.pyplot as plt import numpy as np -# 生成一些随机数据 -data1 = np.random.randn(1000) # 第一组数据 -data2 = np.random.rand(1000) * 100 # 第二组数据 +# 生成示例数据 +data = np.load('/hkfs/work/workspace/scratch/cc7738-benchmark_tag/TAPE_chen/core/gcns/data_seal.npz') +data1 = data['pos_pred'] +data2 = data['neg_pred'] -# 创建一个新的图形 -fig, ax1 = plt.subplots() -# 绘制第一个直方图 (左轴) -color1 = 'blue' -ax1.hist(data1, bins=30, alpha=0.7, color=color1, edgecolor='black') -ax1.set_xlabel('Value') -ax1.set_ylabel('Frequency (Data 1)', color=color1) -ax1.tick_params(axis='y', labelcolor=color1) +# 创建图形和主轴 +fig, ax = plt.subplots(figsize=(10, 6)) -# 创建一个共享x轴的第二个y轴 -ax2 = ax1.twinx() +# 创建左侧直方图 +ax_left = ax.twiny() # 创建一个共享y轴的次坐标轴 +ax_left.hist(data1, bins=30, orientation='horizontal', color='blue', edgecolor='black', alpha=0.7) +ax_left.invert_xaxis() # 反转x轴，使柱状图朝向中心 +ax_left.set_xlabel('Count (Left Histogram)') +ax_left.set_ylabel('Frequency') +ax_left.set_title('Histograms Facing Each Other') -# 绘制第二个直方图 (右轴) -color2 = 'red' -ax2.hist(data2, bins=30, alpha=0.7, color=color2, edgecolor='black') -ax2.set_ylabel('Frequency (Data 2)', color=color2) -ax2.tick_params(axis='y', labelcolor=color2) +# 创建右侧直方图 +ax.hist(data2, bins=30, orientation='horizontal', color='green', edgecolor='black', alpha=0.7) +ax.set_xlabel('Count (Right Histogram)') +ax.yaxis.tick_right() # y轴刻度标签显示在右侧 +ax.yaxis.set_label_position("right") + +# 调整布局 +plt.tight_layout() -# 添加标题 -plt.title('Dual-Axis Histogram') # 显示图形 -plt.savefig('hah') +plt.savefig('/hkfs/work/workspace/scratch/cc7738-benchmark_tag/TAPE_chen/core/gcns/hah.png') diff --git a/core/gcns/trainer_heart.py b/core/gcns/trainer_heart.py deleted file mode 100644 index d068a86bc5..0000000000 --- a/core/gcns/trainer_heart.py +++ /dev/null @@ -1,227 +0,0 @@ -import os, sys -sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))) -from utils import * -# from logger import Logger - -import torch -import torch.nn.functional as F -import torch_geometric -from torch.utils.data import DataLoader -from torch_sparse import SparseTensor -from torch_geometric.utils import negative_sampling -from torch_geometric.graphgym.config import cfg -from sklearn.metrics import * -from torch_sparse import SparseTensor -from heuristic.eval import get_metric_score -from torch.utils.data import DataLoader - - -def train(model, - score_func, - train_pos, - data, - emb, - optimizer, - batch_size, - pos_train_weight, - device): - - model.train() - score_func.train() - - # train_pos = train_pos.transpose(1, 0) - total_loss = total_examples = 0 - - # pos_train_edge = split_edge['train']['edge'].to(data.x.device) - - if emb == None: - x = data.x - emb_update = 0 - else: - x = emb.weight - emb_update = 1 - - train_pos = train_pos.t() - for perm in DataLoader(range(train_pos.size(0)), batch_size, - shuffle=True): - optimizer.zero_grad() - num_nodes = x.size(0) - - ######################### remove loss edges from the aggregation - mask = torch.ones(train_pos.size(0), dtype=torch.bool).to(train_pos.device) - mask[perm] = 0 - train_edge_mask = train_pos[mask].transpose(1,0) - train_edge_mask = torch.cat((train_edge_mask, train_edge_mask[[1,0]]),dim=1) - - # visualize - if pos_train_weight != None: - pos_train_weight = pos_train_weight.to(mask.device) - edge_weight_mask = pos_train_weight[mask] - edge_weight_mask = torch.cat((edge_weight_mask, edge_weight_mask), dim=0).to(torch.float) - else: - edge_weight_mask = torch.ones(train_edge_mask.size(1)).to(torch.float).to(train_pos.device) - - # masked adjacency matrix - adj = SparseTensor.from_edge_index(train_edge_mask, edge_weight_mask, [num_nodes, num_nodes]).to(train_pos.device) - - ################## - # print(adj) - x = x.to(device) - adj = adj.to(device) - h = model(x, adj) - - edge = train_pos[perm].t() - pos_out = score_func(h[edge[0]], h[edge[1]]) - pos_loss = -torch.log(pos_out + 1e-15).mean() - - row, col, _ = adj.coo() - edge_index = torch.stack([col, row], dim=0) - edge = negative_sampling(edge_index, num_nodes=x.size(0), - num_neg_samples=perm.size(0), method='dense') - - neg_out = score_func(h[edge[0]], h[edge[1]]) - neg_loss = -torch.log(1 - neg_out + 1e-15).mean() - - loss = pos_loss + neg_loss - loss.backward() - - if emb_update == 1: torch.nn.utils.clip_grad_norm_(x, 1.0) - torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0) - torch.nn.utils.clip_grad_norm_(score_func.parameters(), 1.0) - - optimizer.step() - - num_examples = perm.size(0) - total_loss += loss.item() * num_examples - total_examples += num_examples - - return total_loss / total_examples - - -def train_opt(model, score_func, train_pos, x, optimizer, batch_size): - model.train() - score_func.train() - - total_loss = total_examples = 0 - - for perm in DataLoader(train_pos, batch_size, shuffle=True): - optimizer.zero_grad() - - num_nodes = x.size(0) - - mask = torch.ones(train_pos.size(0), dtype=torch.bool).to(train_pos.device) - mask[perm] = 0 - - train_edge_mask = train_pos[mask].transpose(1, 0) - train_edge_mask = torch.cat((train_edge_mask, train_edge_mask[[1, 0]]), dim=1) - edge_weight_mask = torch.ones(train_edge_mask.size(1)).to(torch.float).to(train_pos.device) - - adj = SparseTensor.from_edge_index(train_edge_mask, edge_weight_mask, [num_nodes, num_nodes]).to(train_pos.device) - - h = model(x, adj) - - edge_index = train_pos[perm].t() - - pos_out = score_func(h[edge_index[0]], h[edge_index[1]]) - pos_loss = torch.nn.BCEWithLogitsLoss()(pos_out, torch.ones_like(pos_out)) - - neg_edge_index = torch_geometric.nn.negative_sampling(edge_index, num_nodes=num_nodes) - neg_out = score_func(h[neg_edge_index[0]], h[neg_edge_index[1]]) - neg_loss = torch.nn.BCEWithLogitsLoss()(neg_out, torch.zeros_like(neg_out)) - - loss = pos_loss + neg_loss - loss.backward() - - torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0) - torch.nn.utils.clip_grad_norm_(score_func.parameters(), 1.0) - - optimizer.step() - - num_examples = pos_out.size(0) - total_loss += loss.item() * num_examples - total_examples += num_examples - - return total_loss / total_examples - - - -@torch.no_grad() -def test(model, - score_func, - data, - evaluation_edges, - emb, - evaluator_hit, - evaluator_mrr, - batch_size, - data_name, - use_valedges_as_input, - device): - - model.eval() - score_func.eval() - - # adj_t = adj_t.transpose(1,0) - train_val_edge, pos_valid_edge, neg_valid_edge, pos_test_edge, neg_test_edge = evaluation_edges - - if emb == None: x = data.x - else: x = emb.weight - - x = x.to(device) - data = data.to(device) - h = model(x, data.edge_index.to(x.device)) - # print(h[0][:10]) - train_val_edge = train_val_edge.to(x.device) - pos_valid_edge = pos_valid_edge.to(x.device) - neg_valid_edge = neg_valid_edge.to(x.device) - pos_test_edge = pos_test_edge.to(x.device) - neg_test_edge = neg_test_edge.to(x.device) - - pos_train_pred = test_edge(score_func, train_val_edge, h, batch_size) - - neg_valid_pred = test_edge(score_func, neg_valid_edge, h, batch_size) - - pos_valid_pred = test_edge(score_func, pos_valid_edge, h, batch_size) - - if use_valedges_as_input: - print('use_val_in_edge') - h = model(x, data.edge_index.to(x.device)) - - pos_test_pred = test_edge(score_func, pos_test_edge, h, batch_size) - - neg_test_pred = test_edge(score_func, neg_test_edge, h, batch_size) - - pos_train_pred = torch.flatten(pos_train_pred) - neg_valid_pred, pos_valid_pred = torch.flatten(neg_valid_pred), torch.flatten(pos_valid_pred) - pos_test_pred, neg_test_pred = torch.flatten(pos_test_pred), torch.flatten(neg_test_pred) - - - print('train valid_pos valid_neg test_pos test_neg', pos_train_pred.size(), pos_valid_pred.size(), neg_valid_pred.size(), pos_test_pred.size(), neg_test_pred.size()) - result_train = get_metric_score(evaluator_hit, evaluator_mrr, pos_train_pred, neg_valid_pred) - result_valid = get_metric_score(evaluator_hit, evaluator_mrr, pos_valid_pred, pos_valid_pred) - result_test = get_metric_score(evaluator_hit, evaluator_mrr, pos_test_pred, neg_test_pred) - score_emb = [pos_valid_pred.cpu(),neg_valid_pred.cpu(), pos_test_pred.cpu(), neg_test_pred.cpu(), h.cpu()] - - result= {} - for k, val in result_train.items(): - result.update({k: (result_train[k], result_valid[k], result_test[k])}) - return result, score_emb - - - - -@torch.no_grad() -def test_edge(score_func, input_data, h, batch_size): - input_data = input_data.transpose(1, 0) - # print(input_data.size()) - with torch.no_grad(): - preds = [] - for perm in DataLoader(range(input_data.size(0)), batch_size): - - edge = input_data[perm] - preds += [score_func(h[edge[0]], h[edge[1]]).cpu()] - - pred_all = torch.cat(preds, dim=0) - - return pred_all - diff --git a/core/gcns/universal_tune.py b/core/gcns/universal_tune.py index fb4ff9d508..f904c7ad75 100644 --- a/core/gcns/universal_tune.py +++ b/core/gcns/universal_tune.py @@ -71,7 +71,7 @@ def parse_args() -> argparse.Namespace: 'GraphSage': 'core/yamls/cora/gcns/graphsage.yaml' } -def project_main(): +def project_main(): # sourcery skip: avoid-builtin-shadow # sourcery skip: avoid-builtin-shadow # process params args = parse_args() diff --git a/core/graphgps/visualization/data_gsaint.npz b/core/graphgps/visualization/data_gsaint.npz new file mode 100644 index 0000000000..e03508a9ea Binary files /dev/null and b/core/graphgps/visualization/data_gsaint.npz differ diff --git a/core/graphgps/visualization/data_seal.npz b/core/graphgps/visualization/data_seal.npz new file mode 100644 index 0000000000..681040ad89 Binary files /dev/null and b/core/graphgps/visualization/data_seal.npz differ diff --git a/core/graphgps/visualization/pos_neg_plot-2.ipynb b/core/graphgps/visualization/pos_neg_plot-2.ipynb new file mode 100644 index 0000000000..f730691d51 --- /dev/null +++ b/core/graphgps/visualization/pos_neg_plot-2.ipynb @@ -0,0 +1,357 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 87, + "metadata": { + "ExecuteTime": { + "end_time": "2024-05-19T12:26:56.763838Z", + "start_time": "2024-05-19T12:26:54.793802Z" + } + }, + "outputs": [], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "from mpltools import style\n", + "from mpltools import layout\n" + ] + }, + { + "cell_type": "code", + "execution_count": 88, + "metadata": {}, + "outputs": [], + "source": [ + "def pos_neg_hist(y_pos: np.array, \n", + " y_neg: np.array):\n", + " fig, ax = plt.subplots(figsize=(10, 6))\n", + " \n", + " # left histogram for negative samples\n", + " ax_left = ax.twiny() # shared y axis\n", + " ax_left.hist(y_pos, bins=50, orientation='horizontal', color='#1f77b4', alpha=0.7, density=True)\n", + " ax_left.invert_xaxis() # invert y axis\n", + " ax_left.yaxis.set_label_position(\"left\")\n", + " ax_left.set_ylabel('y_pred')\n", + " \n", + " hard_thres = (max(y_pos)+min(y_neg))/2\n", + " plt.axhline(y=hard_thres, color='red', linestyle='--', \n", + " label=f'Hard Threshold: {hard_thres.item()}')\n", + " \n", + " ax_right = ax.twinx() # 共享x轴\n", + " # right histogram for positive samples \n", + " ax_right.hist(y_neg, bins=50, orientation='horizontal', color='#ff7f0e', alpha=0.7, density=True)\n", + " ax_right.yaxis.set_label_position(\"right\")\n", + " ax_right.set_ylabel('y_pred')\n", + "\n", + " ax_left.set_xlabel('y_true')\n", + " ax_right.set_xlabel('y_true')\n", + " ax.set_ylabel('y_pred')\n", + "\n", + " ax_right.grid(False)\n", + " plt.title('Scatter Plot of Predictions vs True Values with Hard Threshold')\n", + " plt.legend()\n", + " plt.tight_layout()\n", + " plt.show()\n", + "\n", + " # plt.savefig('/hkfs/work/workspace/scratch/cc7738-benchmark_tag/TAPE_chen/core/gcns/hah.png')\n", + " return " + ] + }, + { + "cell_type": "code", + "execution_count": 89, + "metadata": { + "ExecuteTime": { + "end_time": "2024-05-19T12:26:57.351535Z", + "start_time": "2024-05-19T12:26:56.765122Z" + } + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "No artists with labels found to put in legend. Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# 生成示例数据\n", + "data = np.load('/hkfs/work/workspace/scratch/cc7738-benchmark_tag/TAPE_chen/core/gcns/data_seal.npz')\n", + "y_pos = data['pos_pred']\n", + "y_neg = data['neg_pred']\n", + "\n", + "norm = lambda x : (x - x.min()) / (x.max() - x.min())\n", + "norm1 = lambda x : (x - x.mean()) / x.std()\n", + "\n", + "pos_neg_hist(y_pos, y_neg)" + ] + }, + { + "cell_type": "code", + "execution_count": 90, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "No artists with labels found to put in legend. Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "y_pos = norm1(y_pos)\n", + "y_neg = norm1(y_neg)\n", + "\n", + "pos_neg_hist(y_pos, y_neg)" + ] + }, + { + "cell_type": "code", + "execution_count": 92, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "No artists with labels found to put in legend. Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n" + ] + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA94AAAJNCAYAAADH6K1yAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAABunUlEQVR4nO3deXRTdf7/8VfaphtQ2gKlZWspBdyQRQaRfVFEQCyLiOAooqjIF386AjOiDuCAijqKI8o4yrDosImCZSsiggOCgqJgQWVXlrKWULaWLvf3R6eR0KS0oTdJ0+fjHI/k5nPvfd/Pvb3JK3ezGIZhCAAAAAAAmCLA2wUAAAAAAODPCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gDK3P79+2WxWDRkyBBvl1Iqn332mdq0aaPIyEhZLBYlJyd7uyS3rV27VhaLRePHj3cY3qlTJ1ksFtPmO3PmTFksFs2cOdO0eaDiMXu79XUWi0WdOnUq1ThDhgyRxWLR/v37TanJV5m9rYwfP14Wi0Vr1641bR5lxROfxe7Mg88JVFQEb1QoeXl5eu+999SxY0dFR0fLarUqJiZGN954ox5++GGlpKR4tJ7ivkz5Wni1WCwO/wUGBqp69erq0qWL5syZY8o8PfnhvH//ft11113at2+fhg4dqnHjxmngwIHFjlMYbi/9Lzg4WHXq1NHdd9+tjRs3ml63p7kK9Cidwm27NP/5orZt28pisWj58uVXbNu4cWNZLBZ9//33HqjM/3kyABZur8V9HhXuG0r7A4G3FAb0kv7nK5/FAMqvIG8XAHhKXl6eevXqpdTUVEVGRqpnz56qU6eOLl68qO3bt2vOnDn6+eef1bt3b2+X6tPGjRsnScrJydHPP/+sTz/9VGvWrNG3336r119/3cvVue/zzz9XVlaW/v73v2vQoEGlGjc+Pt7+pezcuXP6+uuvtXDhQn3yySdauHCh+vTpY0LF7pk9e7bOnz9v2vT79Omj1q1bKy4uzrR5+INmzZrZ/5YK7d+/X7NmzXLYnnzdsGHDtGHDBr3//vvq0aOHy3Zffvmldu7cqZYtW6p58+YerNA//PTTTwoPD/d2GX5lyJAhRX4kWLx4sbZu3aq77rpLzZo1c3jv8tcAUFoEb1QYc+fOVWpqqpo2baovv/xSVatWdXj//Pnz+uabb7xUXflx+ZHO1atX67bbbtOUKVP0xBNPKCEhwSt1Xa3Dhw9LkmrVqlXqcRMSEor0y7hx4/TCCy/o6aef9qngXa9ePVOnX7Vq1SJ/WyiqWbNmRb7Ir127VrNmzXK6Pfmqe+65R08++aSWLl2qo0ePqmbNmk7bvf/++5KkRx55xJPl+Y1rrrnG2yX4HWc/bu3fv19bt25VcnJyufnxC0D5wanmqDA2bNggqeDD1lkwCA8PV+fOnZ2OO3/+fHXt2lXR0dEKDQ1VQkKC7r33Xn377bf2NqdPn9arr76qLl26qE6dOgoODlaNGjXUu3fvIqccF562JxUcCbr0dLbx48dr/Pjxql+/viRp1qxZDu9fftr1ypUr1aNHD1WvXl0hISFq0KCBRo8eLZvNVmQ5EhISlJCQoMzMTP3pT39SQkKCrFbrVX3J79q1q6655hoZhqHNmzdfsX16erpGjBihhIQEex/17dtX3333nUO7Tp066cEHH5QkPfjggw59UNJrFhcsWKAOHTqoatWqCgsLU5MmTfTSSy8pOzvb3qbw9MjCo4+dO3e2z+dqTuEcMWKEJGnfvn06fvy4fZksFosuXryoF154QY0bN1ZISIjDF7yDBw/q//7v/5SYmKiQkBBVq1ZNvXv3dtm3R48e1UMPPaSaNWsqLCxMzZo106xZs1zWVdz1j5999pnuvPNOxcTEKCQkRHXr1tVdd92lzz//XFLB307h38iECRMc1klhXxV3ecB3332nfv362acfHx+vxx9/XOnp6UXaXnp96rvvvqsmTZooNDRUNWvW1COPPKLTp08XGWfbtm269957lZCQoJCQENWoUUMtWrTQk08+qZycHJd9Iklff/21LBZLsT+SXHvttQoJCVFGRoYkyTAMzZo1S23atFGNGjUUGhqqunXr6vbbb9f8+fOLnV9pXHrZyc6dO3XPPfcoJiZGAQEB9n4v/Nt2prhTkn/++WcNGTJEdevWVXBwsGrWrKlBgwbpl19+KVFtYWFhuu+++5STk+PykhCbzaaFCxeqcuXKuvfeeyUVbCf9+vVTYmKiwsLCFBERobZt2+rDDz8s0XwLp1HcpSiuTnvOzc3VO++8o9atWysiIkLh4eFq3ry5pk6dqvz8/CLtU1JS1LVrV8XFxSkkJES1atVSx44d9c4771yxxpUrV8pisejZZ591GL5mzRr7386BAwcc3rvnnntksVi0d+9el8uSkJCgCRMmSHLcZ7n62y7p35AZSvPZWKhweY8cOaKHH35YtWvXVmBgoMO6njdvnm666SaFhYUpJiZGf/zjH+0/oHrKwoUL1apVK4WHhys6OloDBw7UoUOHirQr633/mTNn9Le//U033HCDIiIiVKVKFTVo0ED33HNPkc/SQvv379fAgQNVvXp1hYaGqmXLllq6dKnTttnZ2Xr55ZfVpEkThYeHKyIiQu3bt9eCBQtK1T+7d+/W3XffraioKFWqVElt2rTRsmXLSjUNwJ9wxBsVRrVq1SRJO3fuLPE4hmHowQcf1KxZs1S9enX17dtXNWrU0MGDB7VmzRo1btxYLVu2lFRwKuCzzz6rDh06qGfPnoqKitJvv/2mlJQUrVixQkuWLFH37t0l/X6a6YQJE4qcVlr45cpms+nNN99U06ZNHW7ydelRsgkTJmj8+PGKjo5Wr169FBMTo23btum1117T8uXLtXHjRkVERDgs08WLF9WlSxdlZGSoW7duioiIsId8dxmGIUlXvA513759ateunQ4fPqwuXbro3nvv1YEDB/TRRx9p2bJl+vjjj9WrVy9JBcErMjJSn376aZHT/iIjI69Y09ixY/XSSy+pevXqGjRokCpXrqwVK1Zo7NixWrlypT777DMFBwcrISFB48aN09q1a/Xll1/qgQcesAeYqzl6X9gnUtF+6devnzZv3qw77rhDycnJiomJkSRt2bJF3bp1U0ZGhm6//Xb17dtXJ06c0OLFi9WuXTstWrTI4XTeEydOqE2bNtq7d6/atWundu3aKT09XY899pi6detWqnoLj9BXrlxZycnJqlu3rg4fPqwNGzboww8/1K233mrfDmfNmqWOHTsWCQLFWbp0qfr16yfDMNS/f3/Fx8fru+++07Rp0/Tpp59q/fr1TrfDMWPGaOXKlbrzzjvVrVs3rVmzRu+99552796tL774wt5u27Ztuvnmm2WxWNS7d2/Vr19fmZmZ2r17t9555x1NnDhRVqvVZX2tW7dW48aNtXz5cp08edK+vyi0adMm/fzzz+rXr5+io6MlSc8++6xeeukl1a9fXwMGDFDVqlWVnp6uzZs366OPPtI999xzhV4vnT179ujmm29Wo0aNNHjwYF24cKHI33dppKamqm/fvsrJydGdd96ppKQkHTx4UJ988omWLVumNWvWqEWLFleczrBhw/T2229r+vTp+vOf/1zk/Q8//FBZWVkaNmyYKleuLEkaPny4rr/+enXo0EFxcXE6efKkli9frj/+8Y/65Zdf9Le//c3t5SpO4bKuXLlSjRs31qBBgxQaGqo1a9Zo5MiR+uabb/TBBx/Y2//rX//So48+qtjYWN15552qXr26jh07pm3btmnGjBl6/PHHi51f+/btFRwcrNWrV2vSpEn24atXr3b4d+FngGEYWrNmjRISEpSYmOhyuk8++aQWL15cZJ/lTEn/hsxSms/GS2VkZKh169aqXLmy+vbtq4CAAPsZFW+88Yb+9Kc/KTIyUvfff78iIyO1cuVKtWnTxmNn3LzzzjtKSUlR79691bFjR33zzTeaP3++tm7dqh9++EEhISFFximLfb9hGOrevbs2bNigW265RQ8//LCCgoLs30vat2+vm266yWG+v/76q1q1aqXExET98Y9/VEZGhubPn2//YfXSgw4XL17U7bffri+//FLXXHONRowYofPnz2vhwoW655579MMPP+jFF1+8Yv/s2rVLt9xyi06ePKk77rhDzZo10+7du5WcnKw77rjjaroeKL8MoILYsmWLYbVaDYvFYtx3333Gxx9/bOzfv7/Ycd59911DkvGHP/zBsNlsDu/l5uYahw8ftr+22WzG8ePHi0zjwIEDRlxcnHHNNdcUeU+S0bFjR6fz3rdvnyHJeOCBB5y+/8UXXxiSjFtuucU4deqUw3szZswwJBlPPvmkw/D4+HhDktG1a1fj7NmzTqfriiTD2S5j1apVhsViMSwWi70/XdXerVs3Q5IxceJEh+FfffWVERgYaERHRxtnzpwpshwzZswoVa0bNmwwJBl169Y10tPT7cNzcnKMXr16GZKMSZMmOYwzbtw4Q5KxZs2aEs9nzZo1Ltfh888/b0gyEhMT7cM6duxoSDKaNGlSZFvJyckxGjRoYISEhBhr1651eO/QoUNGrVq1jNjYWCMrK8s+fNiwYU7X8+bNm42goCBDkjFu3DiH9wpruNTKlSsNSUb9+vWNgwcPFlmWAwcOFFnmy6dbyNk6O3PmjBEdHW0EBAQY//3vfx3av/zyy4Yk47bbbnMY/sADD9jX4a+//mofnpOTY7Rv396QZHzzzTf24X/6058MScbixYuL1JSRkWHk5eU5rfdSL774oiHJeOutt4q89/jjjxuSjJSUFPuw6Ohoo3bt2sa5c+eKtHe2L7gSV9tT4d+TJOOZZ55xOm58fLwRHx/v9D1n23ZGRoYRGRlpVKtWzdi+fbtD+x9//NGoVKmS0bx58xLX3qpVK5d/P02bNjUkGZs3b7YP2717d5F22dnZRpcuXYygoKAi26Gz7fZK+wdnfVnYF//3f/9n5Obm2ofn5uYaQ4cOLbINtWjRwggODjaOHj1aZPolXcft27c3AgMDHT5DWrdubTRv3tyoVq2acd9999mH//DDD4YkY+jQoSVeFlf7rNL+DRWnsK+bNm1qjBs3zul/hfO7vE53PxslGX/84x+NnJwch/f27dtnWK1WIyoqyti3b599eF5entG3b1+Xn1UlUbgMxX3mFPZ7lSpVjG3btjm8d++99xqSjPnz5zsML8t9/7Zt2wxJRnJycpHa8vLyjIyMDPvrS/cd48ePd2ibmppqSDLuuOMOh+GF+8E77rjDoe+PHj1q/w7x1VdfFZnH5Z/3t912myHJmDJlisPwxYsX22sq7Wc7UN4RvFGhzJ8/34iNjbXv9CUZ0dHRRnJyssMX6kI33HCDIcnYsmXLVc135MiRhiSHLz+GcXXBOzk52ZBkpKWlOX2/WbNmRo0aNRyGFX5o/vDDD6VehsL+KvySNXbsWKNfv35GYGCgIcl46qmniq39wIEDhiSjXr16xsWLF4tM/7777jMkGbNmzbIPczd4P/zww4Yk49133y3y3i+//GIEBAQY9evXdxh+NcE7Pj7e3i+jRo2yf6kNCAgwPvnkE3v7wi9fzsJh4ZeRUaNGOZ3XlClTDEnGsmXLDMMwjIsXLxrh4eFGlSpVivwoZBi/f4EsSfAu/DHi0lqvtMylCd4ffvihIcm49957i7TPyckxEhISivx9FNb/3nvvFRnn3//+d5GAXBi8V65cecVlcOXAgQNGQECA0bJlS4fh2dnZRnR0tBETE+PwRTQ6OtpISEhw+DHkalwpeNesWdPlvEobvAu3p6lTpzod58knnzQkFQnlrrz//vuGJGPw4MEOwzdt2mRIKnGI//jjj4vsBwyjbIJ3Xl6eER0dbcTGxhYJc4ZhGKdOnTIsFotx991324e1aNHCCA8PdwgzpTV+/HhDkvHpp58ahmEYmZmZRlBQkDFmzBijf//+Rq1atext//73vxuSjP/85z/FLothlDx4l/RvqDiFfV2S/1x9pjlT3Gejqx88Jk6caEgy/vrXvxZ5b8+ePUZAQIBHgvezzz5b5L3CH8Sffvpph+Flue8vDN7O9qeXK9x3xMfHO/zQVKhevXpGtWrVHIYlJSUZFovF+Omnn4q0L/w7f/DBB4vMw9nnff369Z3Ot7A/CN6oaDjVHBXKgAED1KdPH61Zs0br16/X999/r/Xr12vx4sVavHix7r//fvt1g+fOnVNaWppq1qxZ4rvwfvXVV3rzzTe1ceNGHTt2TBcvXnR4/9ChQ2V2c6uNGzfKarXqo48+0kcffVTk/YsXL+r48eNFTpsNDQ3VjTfe6PZ8C68rtFgsioyMVPv27fXQQw/pvvvuK3a8wkcItW/f3ukpv126dNGHH36o77//Xvfff7/b9UkFp+0VTvNyjRo1Up06dbRv3z6dPn26TE5L/PXXX+39EhQUZL9u/emnn1abNm2KtG/VqlWRYYXXOv76669Or7nftWuXpILTNnv06KGff/5Z58+fV/v27Z0uQ6dOnYq91vtShdc3OzvdsywUtz6CgoLUoUMH7d+/X99//32Rv4/CSzkuVbduXUnSqVOn7MPuuecevfnmm0pOTlb//v116623qm3btmrQoEGJ66xTp466du2qVatWaceOHbruuuskSUuWLFFGRoaeeuopBQX9/rE5ePBgvfXWW7ruuus0YMAAdezYUbfccotpp7o2bdrU6emr7ijc3rZu3ep0eyu8JOenn36y90NxBg4cqKeeekoff/yx3nrrLUVFRUmS3nvvPUkFp6Nf6rffftPkyZO1evVq/fbbb7pw4YLD+86uk71aO3fuVEZGhho2bKiJEyc6bRMWFqaffvrJ/nrw4MF6+umndd1112ngwIHq2LGj2rZtqxo1apR4vl26dNH48eO1evVq9e7dW19++aVyc3PVtWtXJSQkaOHChfrpp5907bXX2k/9dva34q6S/g2VxAMPPODymvq1a9e6vE+KO5+NCQkJ9lOxL1W4P+nYsWOR9xITE1W3bl39+uuvJVmcq+JOv5bFvv+6665Ts2bNNHfuXP3666+666671K5dO7Vs2VLBwcFO59usWTMFBgY6rffS6+zPnDmj3bt3q3bt2k5v6Fe4XV7pkYCF77dr187pfDt16qQvv/yy2GkA/ojgjQrHarWqW7du9mtg8/Ly9PHHH2vo0KGaPXu2+vTpo+TkZPvNyWrXrl2i6S5atEj9+/dXaGiobrvtNjVo0ECVKlWy3wDpyy+/dLip19U6efKkcnNz7YHPlbNnzzoE75iYmKt6JrBxybXLpVF4Ix9Xj5kqHO7spnBmzOu3336TzWYrk5DUsWPHUt2ILTY2tsiwkydPSpLTH1EudfbsWUm/L6Oru0g7m4crNptNUVFRCgsLK/E4pXE1697Z9fyF4TcvL88+rFWrVlq3bp0mTZqkhQsX2q/Tbdy4scaNG2e/qdeVDBkyRKtWrdKsWbM0efJkSbL/gPHAAw84tH3jjTeUmJioGTNm6OWXX9bLL7+soKAg9ejRQ3//+9+VlJRUonmWVGnW6ZUUbm+FwdiVwu3tSipVqqRBgwbp3Xff1YcffqiRI0fq3LlzmjdvnipVqqTBgwfb2+7du1etWrXSqVOn1L59e3Xr1k1Vq1ZVYGCg/ZFqZbmvLFS4zLt27Sp2v3npMv/pT39S9erV9c477+gf//iHpkyZIovFoo4dO+rVV191Gr4u17p1a1WqVMl+Xffq1asVHBysdu3a2a/NXr16tRo2bKj//ve/uu6668p0XZf0b8gs7n42uuqDkuz7PBG83enXstj3BwYG6osvvtALL7yghQsX2u+rUKVKFT3wwAN66aWX7PdSKK7WwnovvaFgWX1Ol+XnE+BPCN6o8AIDAzVgwAD9+OOPmjhxor744gslJyfbP6hKeuTl+eefV3BwsL799ltde+21Du89+uijZf7rbtWqVZWfn2+/w3JJXU3ovhqFAffIkSNO3y+8s3VZBOFL5+XsiGdZzssdztZBYS2ffvppiZ4lX9j+6NGjTt931c/OREZG6uTJk7pw4YIp4dtT6/6WW27R0qVLlZ2dre+++06pqal66623NGjQINWoUUO33nrrFafRp08fRURE6MMPP9SLL76okydPasWKFWratKmaNm3q0DYwMFBPPvmknnzySR07dkzr16/XvHnz9NFHH2n79u3avn17mR2hlor/2w0ICChyFLGQsy/JhX29devWqzoD5lKPPPKI3n33Xb3//vsaOXKk5s2bpzNnzmjo0KEON4F7/fXXdfLkSc2YMaPII5vmzp1b4jM1AgIKHsySm5tb5L3ilrlPnz765JNPSrhU0v3336/7779fNptNGzZs0KJFi/Tvf/9bt99+u37++ecrHv22Wq1q166dVq5cqSNHjmj16tW65ZZbFB4ebj8D5/PPP1eLFi105syZMj3a7Qvc/Wx0tb1fuu+7/vrri7xfmn2fp5XFvl+SoqKi9MYbb+iNN97Q7t279eWXX+rdd9/V1KlTZbPZHG4QWBplta8uy88nwJ/wODHgf6pUqSLp9yO6lSpV0g033KCjR49e8bQqqeCxGdddd12RLxb5+flav36903ECAgJc/jJeeHqWq/dbt26tU6dOafv27VeszRcUnq6/fv16p1+U16xZI0kOd1G+Uh9caV7OjkLv3r1bBw8eVP369Ut0d3RPad26tSRp3bp1JWp/zTXXKDw8XD/88IPTxwKV5gh869atZRiGUlNTr9jWnXVS3PrIzc21L3NJ7qBdEiEhIWrTpo1eeOEF/eMf/5BU8KW2JMLCwjRgwAAdPnxYn3/+uebMmaPc3NwiR7svFxMTo759+2rBggXq0qWL9uzZo7S0tKtelpKKiorS0aNHnT427dLHHhYq7fZWEi1atNBNN92kbdu2adOmTS6f3b17925JBXd4vlxpfqAsPJ398sdxSc6X+ZprrlFkZKS+/vrrKz5ezpnIyEj16NFD7733noYMGaKMjAz997//LdG4Xbt2lVTww0JaWpr9tVRw+u7atWu1atUqh7ZX4u7+0dPc+WwsTuF+wtm2snfvXqfbgy+72r/FpKQkPfTQQ/ryyy9VuXLlEu/rnCl8LNmhQ4fsp7hfytnntDOXft472z6v5lGdQHlG8EaFMXfuXK1atcrpc1qPHDliP+WyQ4cO9uFPPPGEpIJf5S8PN/n5+Q7PH05ISNCuXbscniNqGIbGjx+vHTt2OK2pWrVqLr8kREVFyWKx6LfffnP6/lNPPSWp4NpJZ88uPXfunL7++mun43pDnTp1dNttt2n//v2aMmWKw3vffPON5syZo6ioKIfnKBeeIu+qD1wZOnSoJGnixIn2Z2hLBV9QR40apfz8fD300ENuLok57rrrLjVo0EBvv/22li9f7rTNxo0bdf78eUkFR9EGDx6sM2fOFLku8Ntvv9V//vOfEs975MiRkqSnn37a6Rkelw5zZ50kJycrOjpac+fOLbJNTpkyRfv27dOtt956Vfc/2LBhQ5HrhKXfj7iEh4eXeFqFR2Fnz56t2bNnKygoyOFUaangObdfffVVkXFzcnLsZ6GUZp5Xq1WrVsrNzdWMGTMchs+cOdNpnQ8++KAiIyM1YcIEbdq0qcj7+fn5bn05LryWe9SoUfr6669144036uabb3ZoU3h69eXTX7lypT2sl0TLli0VEBCgOXPm2P8upILHUI0ZM6ZI+6CgII0cOVLp6el64oknnG4v6enpDvvrNWvWOL285tixY5JKvo4Lj2K//PLLMgyjSPA+ffq03nnnHQUEBDh99rgz7u4fPc2dz8biDB48WFarVW+99Zb2799vH56fn6/Ro0c7/Yz3ZaXd9+/bt8/hGe+FTp06pezs7Ks+a2no0KEyDEOjR492CM0nTpywP+av8DPWlcLP+3379mnq1KkO73366adc340Ki1PNUWF88803evPNNxUbG6t27drZnxm8b98+LVu2TBcuXNBdd92l/v3728d5+OGHtW7dOn3wwQdq2LCh7rrrLtWoUUOHDx/WF198oaFDh9pDz1NPPaXHHntMzZs3V79+/WS1WvXVV19px44duvPOO7VkyZIiNXXt2lXz5s3TnXfeqRYtWshqtapDhw7q0KGDKleurJtvvlnr1q3T4MGD1ahRIwUGBqp379668cYb1bVrV7388st65pln1LBhQ/Xo0UP169fX2bNn9euvv+rLL79Uu3btSnQU01P++c9/qm3btho9erQ+++wztWzZ0v4c74CAAM2YMcN+5oEk++mYU6ZM0cmTJ+3XhY0cObLYU93atGmjMWPG6JVXXtENN9yg/v37q1KlSlqxYoXS0tLUrl07jR492vTlLQ2r1apPPvlEt99+u3r27Kk2bdqoWbNmCg8P14EDB7R582bt3btX6enp9i/7L774olavXq0pU6bo22+/tT/He/78+erRo4dSUlJKNO9u3brpueee08SJE3Xttdfan+N99OhRrV+/Xq1bt7bfUKlx48aqXbu25s2bJ6vVqvj4eFksFv3xj39UfHy80+lXrlxZ//73v3X33XerY8eOuvvuu1WvXj199913+uyzzxQbG6t33333qvrvlVde0RdffKH27durfv36qly5srZv364VK1YoKiqqyFHX4rRt21ZJSUn66KOP7M99vvwmTxcuXFC7du2UlJSkm266SfHx8crKytKqVav0008/qXfv3kWO8Jlp5MiRmjFjhoYPH67Vq1erbt26+uGHH7Rx40b16tVLS5cudWhfrVo1LVy4UH369FHr1q3VtWtXXX/99bJYLDpw4IA2btyokydPKisrq1R1DBo0SKNGjbIfvbv8pmqS9Pjjj2vGjBm6++671b9/f9WqVUtpaWlKTU3VgAEDNH/+/BLNKy4uToMHD9YHH3ygZs2aqWfPnsrMzNTy5cvVoUMHp2cqPf/889q6dav++c9/asmSJerSpYtq166tY8eOadeuXfrqq680adIk+w3l+vTpo8qVK6t169ZKSEiQYRhat26dNm/erJtuuqlEly9IBUcAo6KidOzYMVWpUsXhJluFIfzYsWNq2bJlic/E6dy5swICAvTMM88oLS3NfgbAc889V6LxPcWdz8biJCQk6OWXX9bTTz+t5s2b65577lHVqlW1cuVK2Ww23Xjjjdq2bZtJS1P2Srvv37p1q/r27as//OEPuvbaa1WrVi0dP35cn376qXJycuzXfLtr1KhRWrFihT799FM1bdpUPXr00Pnz5/XRRx/p2LFjGjNmjNq1a3fF6bz99tu65ZZb9OSTT+qzzz5T06ZNtXv3bi1atMit9Q74Ba/dTx3wsN9++82YOnWqkZycbDRq1MioUqWKYbVajdjYWOOOO+4wPvjgA5fP+v3www+NDh06GBEREUZISIiRkJBgDBo0yPjuu+8c2s2YMcNo2rSpER4eblSrVs1ITk42tm3b5vKxL0ePHjXuvfdeIyYmxv4IlEsf07Rr1y6jV69eRnR0tGGxWJw+fmPdunXG3XffbcTFxRlWq9WoXr260bRpU+Opp55yeG6uYRT/yKErUSmejVrco9AOHjxoPPbYY0a9evUMq9VqVKtWzbjrrruMTZs2OZ3WihUrjNatWxuVKlWy13Dps1uLM3fuXKNt27ZG5cqVjZCQEOO6664zJk6caFy4cKFI27J+jrczzh6JdLmjR48af/7zn43rr7/eCAsLMypVqmQkJSUZ/fr1Mz744IMij0FKT083HnzwQaN69epGaGio0bRpU2PGjBkuH/tVXA3Lli0zbr/9diMqKsoIDg426tSpYyQnJxurV692aLdp0yajS5cuRkREhH27LOy34h7xtGnTJiM5OdmoXr26YbVajbp16xqPPfaYcejQoSJtCx/r42xdO1u2lStXGkOGDDGuvfZaIyIiwggPDzcaNWpkjBw50v58+dL429/+Zt/eFi5cWOT9ixcvGpMnTza6d+9u1K1b1wgJCTGqV69u3Hzzzca0adOM7OzsUs/zSo8Tc/VowULr1q0z2rdvb4SFhRlVqlQxevToYWzdurXYbXvfvn3GiBEjjKSkJCMkJMSoUqWK0bhxY+O+++4zFi1aVOplMIzfH+cXFhZmnDp1ymmbr776yujcubMRGRlpVK5c2Wjbtq2xaNGiUm+3WVlZxqhRo4zatWsbVqvVaNCggfHiiy8aOTk5Lv828/PzjdmzZxtdunQxoqKiDKvVatSqVcto27atMWnSJOO3336zt502bZqRnJxs1K9f3wgLCzOioqKMZs2aGZMnTzYyMzNL1S+Fz5ju0aNHkfcaNWpkSDLGjBnjdFxXy/LBBx8YTZs2NUJDQ4vso0v7N1Scwr/r4rbB4vaHpf1sLMl+dc6cOUbz5s3tf3uDBw82Dh06VKL9rCuleZyYq78nZ/1Ulvv+AwcOGM8884zRpk0bo2bNmkZwcLBRu3Zto3v37sby5ctLVM+V6rpw4YIxadIk4/rrrzdCQ0Ptf6Nz5swp8TIbRsF3mH79+hlVq1Y1wsPDjdatWxtLly51+1GhQHlnMQw3b1EMAAAAAACuiGu8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAPAxGRkZWrBggfbv3+/tUgAAQBkgeAMA4GNOnTqlhQsXErwBAPATBG8AAMq57Oxsb5cAAACKQfAGAKCMpKWlacCAAdq0aVOR99avX68BAwZo586dxU5j+/bteuaZZyRJ77zzjgYMGKABAwZo7dq1kqTx48fr6aef1t69ezVu3Djdd999mjt3riRpwIABWrBgQZFpjhgxQm+//bbDsHPnzmnmzJkaPny4Bg0apJEjR2rx4sXKz893Z9EBAEAxgrxdAAAA/uL6669XtWrVtG7dOrVq1crhvXXr1qlmzZpq1KhRsdOoXbu2PUDfeuutuuaaayRJjRs3trc5c+aMXnzxRbVp00bt27dX1apVS1Vndna2xo8fr4yMDN16662qXr26fvnlF82dO1c2m01Dhgwp1fQAAEDxCN4AAJQRi8Wi9u3ba9myZTp//rzCw8MlSZmZmdq2bZv69OlzxWlERkaqefPmWrBggRo1aqQOHToUaWOz2TRs2DDddtttbtW5dOlSHTlyRK+88ori4uIkSbfddpuio6OVkpKiXr16qXr16m5NGwAAFMWp5gAAlKGOHTsqJydHX3/9tX3Yhg0blJeX5zREu8Nqtapz585uj//111/r2muvVaVKlZSZmWn/r0mTJsrPz9dPP/1UJnUCAIACHPEGAKAM1a5dWw0aNNC6devUpUsXSQWnmTds2FCxsbFlMo/o6GgFBbn/EZ6enq5ff/1VDz/8sNP3T58+7fa0AQBAUQRvAADKWMeOHTVjxgydPHlSOTk52rVrl4YOHVpm0w8ODi5V+8tvmGYYhm688Ub17t3baftatWq5XRsAACiK4A0AQBlr06aNZs2apa+++koXL15UYGCg2rRpU+LxLRaLW/OtVKmSzp8/7zAsNzdXp06dchhWs2ZNZWVl6cYbb3RrPgAAoHS4xhsAgDIWERGh5s2ba926dVq3bp2aNWumiIiIEo8fEhIiqeCRX6URGxtb5Prszz//vMgR71tuuUU7d+7UDz/8UGQa586dU15eXqnmCwAAiscRbwAATNChQwe9/vrrkqR77rmnVOPWrFlTlSpV0qpVqxQWFqaQkBA1bNhQMTExxY7XpUsXvffee3rttdd044036tdff9XWrVtVpUoVh3a9e/fWt99+q8mTJ6tjx45KTExUdna2fvvtN3399dd6++23S/VDAQAAKB5HvAEAMEHLli1VqVIlhYeHq2XLlqUaNygoSCNGjFBAQIDee+89vfnmm9qxY8cVx+vatavuuusu/fTTT/rggw907NgxPffccwoNDXVoFxISogkTJujOO+/Ujh07NHPmTC1evFhHjhzRgAED7I9BAwAAZcNiGIbh7SIAAPA3eXl5evTRR3XTTTdp+PDh3i4HAAB4EUe8AQAwwebNm5WZmamOHTt6uxQAAOBlXOMNAEAZ2rVrl3799Vd9/PHHql+/vq677jr7e7m5uTp79myx44eHh5f6cWEAAMC3EbwBAChDn332mdatW6eEhAQ9/vjjDu/98ssvmjBhQrHjP/744+rUqZOJFQIAAE/jGm8AADzk7Nmz2rt3b7Ft6tatq6ioKA9VBAAAPIHgDQAAAACAibi5GgAAAAAAJuIa73Ji0aJF2rRpkw4dOqTg4GA1atRI9913n2rVquVynLVr1+qdd95xGGa1WvWf//zH7HLhp9zZDiVp48aNmj9/vo4fP67Y2FgNHjxYLVq08FDV8Dc7duxQSkqK9u3bp1OnTmnUqFFq1aqVy/bbt293el31v/71L0VGRppYKfxZabdDqWBbnD17tg4cOKBq1aqpX79+XM+Pq5aamqolS5bIZrMpPj5eQ4cOVVJSktO2fDeEGUqzDUoV93shwbuc2LFjh26//XY1aNBAeXl5mjt3riZOnKjXX39doaGhLscLCwvTm2++6cFK4c/c2Q5/+eUXvfnmmxo0aJBatGih9evX69VXX9XkyZNVr149Dy8B/EF2drYSEhLUpUsXvfbaayUeb8qUKQoPD7e/joiIMKM8VBCl3Q6PHTuml19+WbfddptGjhyptLQ0/fOf/1RkZKSaNWtmfsHwSxs2bNDs2bM1bNgwNWzYUMuWLdOkSZM0ZcoUVa1a1ek4fDdEWSrtNliRvxcSvMuJZ5991uH1iBEj9PDDD2vv3r0Oj6q5nMVi4YgOyow72+Hy5cvVrFkz9e7dW5I0cOBA/fjjj0pNTdUjjzxies3wP82bN1fz5s1LPV7VqlVVqVIlEypCRVTa7fCzzz5TTEyM7r//fklSnTp19PPPP2vZsmUEb7ht6dKl6tq1qzp37ixJGjZsmLZs2aI1a9YoOTnZ6Th8N0RZKu02WJG/FxK8y6nz589LkipXrlxsu6ysLD3++OMyDEP169fXvffeq7p163qiRFQAJdkOd+7cqV69ejkMa9q0qTZv3mxqbcDlxowZo5ycHNWtW1d33323rrnmGm+XhApk165datKkicOwpk2baubMmd4pCOVebm6u9u7d6xBuAgIC1KRJE+3cudPleHw3RFlxZxusyN8LublaOZSfn6+ZM2eqcePGxZ6SUatWLQ0fPlxjxozRyJEjlZ+fr+eee04nT570YLXwVyXdDm02W5FTjapWrSqbzWZyhUCBqKgoDRs2TE8//bSefvppVatWTRMmTLjiY72AsuRqX3jhwgVdvHjRS1WhPMvMzFR+fn6Ro9eRkZEuP2P5boiy5M42WJG/F3LEuxyaPn26Dhw4oBdeeKHYdo0aNVKjRo0cXj/11FNatWqVBg4caHaZ8HMl3Q4Bb6tVq5bDDQAbN26so0ePatmyZRo5cqQXKwMAz+K7IeA9HPEuZ6ZPn64tW7Zo3LhxqlatWqnGDQoKUv369XXkyBGTqkNFUZrtMDIyUqdPn3YYdvr0aa4vg1clJSWxL4RHudoXhoWFKTg42EtVoTyLiIhQQEBAkSOFNputxJ+xfDfE1XBnG6zI3wsJ3uWEYRiaPn26Nm3apL/+9a+KiYkp9TTy8/P122+/KSoqyoQKURG4sx02atRIP/74o8Owbdu2qWHDhmaVCVzR/v372RfCoxo2bOh0X3jp0UegNIKCgpSYmKi0tDT7sPz8fKWlpZV4u+K7Ia6GO9tgRf5eSPAuJ6ZPn65169bp//2//6ewsDDZbDbZbDaH68KmTp2qOXPm2F8vXLhQW7du1dGjR7V371794x//0PHjx9W1a1dvLAL8gDvbYY8ePbR161YtWbJEhw4d0oIFC7Rnzx51797dG4sAP5CVlaX9+/dr//79kgoe07R//36dOHFCkjRnzhxNnTrV3n7ZsmXavHmzjhw5ot9++00zZ85UWlqabr/9dm+UDz9R2u2wW7duOnbsmD788EMdOnRIK1eu1MaNG9WzZ09vlA8/0atXL61evVpr167VwYMH9f777ys7O9v+fHi+G8Jspd0GK/L3Qq7xLic+++wzSdL48eMdhj/++OP2DfvEiROyWCz2986ePat3331XNptNlSpVUmJioiZOnKg6dep4qmz4GXe2w8aNG+uJJ57QvHnzNHfuXMXFxWn06NF+/6xGmGfPnj2aMGGC/fXs2bMlSR07dtSIESN06tQpe/iRCu66Onv2bGVkZCgkJETx8fF6/vnndcMNN3i8dviP0m6HMTEx+stf/qJZs2Zp+fLlqlatmh577DEeJYar0qZNG2VmZmrBggWy2WxKSEjQ2LFj7aft8t0QZivtNliRvxdaDMMwvF0EAAAAAAD+ilPNAQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHB24/k5ORowYIFysnJ8XYpqMDYDuFtbIPwBWyH8AVsh/A2tsHfEbz9SE5OjhYuXMiGDa9iO4S3sQ3CF7AdwhewHcLb2AZ/R/AGAAAAAMBEBG8AAAAAAExE8PYjAQEBatu2rQICWK3wHrZDeBvbIHwB2yF8AdshvC0gIEAPPvigt8vwCRbDMAxvFwEAAAAA8E85OTmyWq3eLsOrCN6XuHDhQrm/8D8iIkKZmZneLgNX4O/ryd+XrzyqaOukoi2vr6H/C9APnkNfu0bfmIN+LRmr1aqwsDBvl+ETgrxdgC/JycnR2bNnvV2G2ywWiyIiInTu3Dnxe4rv8vf15O/LVx5VtHVS0ZbX19D/BegHz6GvXaNvzEG/llzlypUJ3v/DBR8AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAibi5GgAAAACgiNzcXJ0/f/6K7cLDwxUURLQsDr0DAAAAAHCQm5urc+fOqUqVKgoIcH2idH5+vs6cOaNKlSoRvovBqeYAAAAAAAfnz5+/YuiWpICAAFWpUqVER8YrMoI3AAAAAKCIK4Xu0raryOghAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAAA4MAzD1PYVDcEbAAAAAOAgKChI586du2KgNgxD586dU1BQkIcqK5/oHQAAAACAg0qVKik7O1tnzpy5YtuQkBCFhIR4oKryy6eC944dO5SSkqJ9+/bp1KlTGjVqlFq1alXsONu3b9fs2bN14MABVatWTf369VOnTp08UzAAAAAA+ClvBGp/zYQ+dap5dna2EhIS9NBDD5Wo/bFjx/Tyyy/r+uuv1yuvvKKePXvqn//8p3744QdzCwUAAAAAlDl/zYQ+dcS7efPmat68eYnbf/bZZ4qJidH9998vSapTp45+/vlnLVu2TM2aNXM5Xk5OjnJycuyvAwMD7b/kWCwW94r3AYW1l+dlqAj8fT35+/KVRxVtnVS05fU19H8B+sFz6GvX6Btz0K+ld+HCBYdrxa1Wq6xWq9O2nsqEnuZTwbu0du3apSZNmjgMa9q0qWbOnFnseIsWLdLChQvtr5OTkzVo0CBFREQoIiLCjFI9KjY21tsloAT8fT35+/KVRxVtnVS05fU19H8B+sFz6GvX6BtzVKR+fWjmZrfGa1U/Wo92jND48eO1b98++/D+/ftrwIABZVKbu5nQ08p18LbZbKpatarDsKpVq+rChQu6ePGigoODnY7Xp08f9erVy/46MDBQkpSZmalz586ZV7DJLBaLYmNjdeTIEW7n78P8fT35+/KVRxVtnVS05fU19H8B+sFz6GvX6BtzVMR+zcrOcmu83NxcSdL48eOLHPEuK+5mQk8r18HbXcWd2uAPfzyGYfjFcvg7f19P/r585VFFWycVbXl9Df1fgH7wHPraNfrGHBWqX69yMcPCwsqmjnLMp26uVlqRkZE6ffq0w7DTp08rLCzMZ37ZAAAAAACYo7xkwnIdvBs2bKgff/zRYdi2bdvUqFEjL1UEAAAAAPCU8pIJfSp4Z2Vlaf/+/dq/f7+kglvD79+/XydOnJAkzZkzR1OnTrW379atm44dO6YPP/xQhw4d0sqVK7Vx40b17NnTG+UDAAAAAK6Cv2ZCn7rGe8+ePZowYYL99ezZsyVJHTt21IgRI3Tq1Cl7h0tSTEyM/vKXv2jWrFlavny5qlWrpscee8ynbhsPAAAAACgZf82EFqPC3BHgyjIzM3X27Flvl+E2i8WiuLg4paenV5wbPZRD/r6e/H35yqOKtk4q2vL6Gvq/AP3gOfS1a/SNOSpiv45O2ePWeG0axmhE18ZlXE355FOnmgMAAAAA4G986lRzAAAAAEDpuXtUGp7BEW8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwERB3i4AAAAAAPzF6JQ93i4BPogj3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjneAMAAADweT7zfGyLFBpyUFnZWZLh7WJQXnDEGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwUZC3CwAAAAD80eiUPd4twCKFhhxUVnaWZHi3FKCi44g3AAAAAAAmIngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGCiIG8XAAAAAJSl0Sl7vF0CADjgiDcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACYieAMAAAAAYKIgbxcAAAAA3zU6ZY+3SyieRQoNOais7CzJ8HYxAOAcR7wBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARz/EGAADwMp9/VjYA4KpwxBsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMFGQtwsAAACoiEan7PF2CQAAD+GINwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgoiBvF+BLwr58QcHHdymj+zRvlwIAAFwYnbLH2yVcmUUKDTmorOwsyfB2MQAAb+OINwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAin7y5WmpqqpYsWSKbzab4+HgNHTpUSUlJLtsvW7ZMn332mU6cOKGIiAjdfPPNGjRokIKDgz1YNQAAAADgavljHvS5I94bNmzQ7Nmz1b9/f02ePFnx8fGaNGmSTp8+7bT9+vXrNWfOHN19991644039Nhjj2njxo2aO3euhysHAAAAAFwNf82DPnfEe+nSperatas6d+4sSRo2bJi2bNmiNWvWKDk5uUj7X375RY0bN1a7du0kSTExMWrbtq127drlch45OTnKycmxvw4MDFRISIj9tcViKaOl8azCustr/RWFv68nf1++8qiirZOKtry+xiP9Xy5WreX3/1t4npi56GvX6Btz0K+ldeHCBRnG731ltVpltVqdtvVEHvQGnwreubm52rt3r0OHBgQEqEmTJtq5c6fTcRo3bqx169Zp9+7dSkpK0tGjR/X999+rffv2LuezaNEiLVy40P46OTlZgwYNkjUoSNbQUMXFxZXZMnlDbGyst0tACfj7evL35SuPKto6iY2N1UMzN3u7jArqgKlTDw0JNXX6ZSn0kh/2YS762jX6xhz065UFBRXEzfHjx2vfvn324f3799eAAQOKtPdUHvQGnwremZmZys/PV2RkpMPwyMhIHT582Ok47dq1U2Zmpp5//nlJUl5enm677Tb17dvX5Xz69OmjXr162V8HBgZKknJyc5WXlaVT6elXuSTeYbFYFBsbqyNHjjj8ogTf4u/ryd+XrzyqaOvk0uXNys7ydjkVkEWhISHKys6W5P/bm2v0g+fQ167RN+agX0sqNzdXUkHwvvyItzOeyoPe4FPB2x3bt2/XokWL9PDDD6thw4Y6cuSIZsyYoYULF6p///5Oxynu1AZJ5f6LqWEY5X4ZKgJ/X0/+vnzlUUVbJ4Zh8H3IG+ynXVbw/qcfPIe+do2+MQf9WmphYWGmTdudPOgNPhW8IyIiFBAQIJvN5jDcZrMV+dWj0Pz589WhQwd17dpVklSvXj1lZWXpX//6l/r27auAAJ+7fxwAAAAA4DL+nAd9o4r/CQoKUmJiotLS0uzD8vPzlZaWpkaNGjkdJzs7u8gNXHylcwEAAAAAJePPedCnjnhLUq9evfT2228rMTFRSUlJWr58ubKzs9WpUydJ0tSpUxUdHa1BgwZJkm666SYtW7ZM9evXt59aMH/+fN10000+2eEAAAAAAOf8NQ/6XPBu06aNMjMztWDBAtlsNiUkJGjs2LH2UwtOnDjh8ItGv379ZLFYNG/ePGVkZCgiIkI33XST7r33Xi8tAQAAAADAHf6aBy1GRbrTzhXkLBmlvOO7lNF9mrdLcYvFYlFcXJzS09Mr1A2Uyht/X0/+vnzlUUVbJ5cu76hPd3u7nIrHUvC4r6zsrIp90yH6wXPoa9foG3PQryXWpmGMRnRt7O0yfILvHHsHAAAAAMAP+dyp5t5WXo92AwCKerV3A9OmPTplj2nTBgAA/oUj3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiYK8XQAA+JvRKXu8XYIjixQaclBZ2VmS4e1iPKCiLS8AAPB5HPEGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8L5MdOpwb5cAAAAAAPAjBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAEwU5O0CfFHhs7wzuk/zciUAfMXolD3eLgEAAADlFEe8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATBXm7AF+U0X2at0soYnTKnis3skihIQeVlZ0lGebXBDf5+3ry9+UDAAAASokj3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiYK8XQBK5tXeDa7YxmKxKC4uTunp6TIMwwNVXb3RKXu8XQIAAAAAmIoj3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiYK8XYAvik4drozu07xdRoXwau8Gbo87OmVPGVYCAAAAAObgiDcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACbiOd4uRKcOLzKMZ3v7lqt5Brg3WSwWxcXFKT09XYZheLucMmfm8vHsdgAAAJRHHPEGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAEwU5O0CypPo1OH2f2d0n+bFSoCK6dXeDbxdgs8ZnbLH2yUAAADgCjjiDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJjIJ2+ulpqaqiVLlshmsyk+Pl5Dhw5VUlKSy/bnzp3T3LlztWnTJp09e1Y1atTQAw88oBYtWniwagAAAADA1fLHPOhzwXvDhg2aPXu2hg0bpoYNG2rZsmWaNGmSpkyZoqpVqxZpn5ubq4kTJyoiIkJ/+tOfFB0drRMnTig8PNwL1QMAAAAA3OWvedDngvfSpUvVtWtXde7cWZI0bNgwbdmyRWvWrFFycnKR9l988YXOnj2rv/3tbwoKKlicmJgYT5YMAAAAACgD/poHfSp45+bmau/evQ4dGhAQoCZNmmjnzp1Ox/nuu+/UsGFDTZ8+Xd9++60iIiLUtm1bJScnKyDA+SXsOTk5ysnJsb8ODAxUSEhIqWq1WCylau8JhTX5Ym34nb+vJ39fPp9Tom62/P5/i2FiMb6ioi2vr6H/C9APnkNfu0bfmIN+La0LFy7IMH7vK6vVKqvVWqSdp/KgN/hU8M7MzFR+fr4iIyMdhkdGRurw4cNOxzl69KiOHz+udu3a6ZlnntGRI0f0/vvvKy8vT3fffbfTcRYtWqSFCxfaXycnJ2vQoEGyBgXJGhpaolrj4uJKtlBeEBsb6+0SUAL+vp78ffl8xX8e9dy+6KGZmz02r7IQWsofVFG26P8C9IPn0Neu0TfmoF+vrPAI9Pjx47Vv3z778P79+2vAgAFF2nsqD3qDTwVvdxiGoYiICD366KMKCAhQYmKiMjIylJKS4rKj+/Tpo169etlfBwYGSpJycnOVl5VVovmeSk+/+uLLmMViUWxsrI4cOeLwixJ8i7+vJ39fvvKorNZJVnbJ9o/eZ1FoSIiysrMlsQ16Hv1fgH7wHPraNfrGHPRrSeXm5koqCN6XH/EuK+7kQW/wqeAdERGhgIAA2Ww2h+E2m63Irx6FIiMjFRQU5HAaQe3atWWz2ZSbm2v/leVSrk5tKA1fDhSGYfh0fSjg7+vJ35evPLrqdVJeVqf9tD+j/NTsT+j/AvSD59DXrtE35qBfSy0sLKxE7TyVB73Bd056V8GpCImJiUpLS7MPy8/PV1pamho1auR0nMaNG+vIkSPKz8+3D0tPT1dUVJTPdDIAAAAAoHj+nAd9KnhLUq9evbR69WqtXbtWBw8e1Pvvv6/s7Gx16tRJkjR16lTNmTPH3r5bt246e/asZs6cqcOHD2vLli1atGiRbr/9di8tAQAAAADAHf6aB33nJ4D/adOmjTIzM7VgwQLZbDYlJCRo7Nix9lMLTpw44XC35OrVq+vZZ5/VrFmzNHr0aEVHR+uOO+5weqt5AAAAAIDv8tc8aDG4CNMuZ8ko5R3fVaK2Gd2nmVxN6VksFsXFxSk9PZ1ra32Yv68nf1++8qis1snolD1lWJWJLFJoSGjBzeDYBD2P/i9AP3gOfe0afWMO+rXE2jSM0Yiujb1dhk/wuVPNAQAAAADwJz53qnl5EZ063CePegMo/8r8yLJFCg05yC/zAAAAXsIRbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADAREHeLqC8yeg+zdslAPBzr/ZuUKbTs1gsiouLU3p6ugzDKNNp+6KKtry+ZHTKHm+XAACAT+KINwAAAAAAJirxEe+FCxe6NYP+/fu7NR4AAAAAAGbwdL4tcfD+6KOP3JoBwRsAAAAA4Es8nW9LHLznz5/v8DojI0MvvfSS6tatq549e6pWrVqSpEOHDmn58uU6ePCg/vKXv7hVFAAAAAAAZvF0vnX7Gu/3339fcXFxeuKJJ9SgQQOFhYUpLCxMSUlJeuKJJ1SzZk1Nnz7d7cIAAAAAAPAEs/Ot28F7+/btuuGGG1y+36RJE6Wlpbk7eQAAAAAAPMLsfOt28LZardq5c6fL93/55RdZrVZ3Jw8AAAAAgEeYnW/dfo53u3bttGLFCoWHh+uOO+5QzZo1JUlHjx7VihUrtH79et1xxx1uFwYAAAAAgCeYnW/dDt733Xefzpw5o5UrV2rlypUKCCg4eJ6fny9Jatu2re677z63C/NV0anDHV5ndJ/mpUoAoGIYnbKndCNYpNCQg8rKzpIMc2oCAAD+xex863bwDgoK0siRI9W7d29t2bJFJ06ckCTVqFFDzZo1U0JCgttFAQAAAADgKWbnW7eDd6H4+HjFx8df7WQAAAAAAPAqs/LtVQfvnTt3avv27Tp9+rRuv/12xcXFKTs7W4cOHVKtWrUUGhpaFnUCAAAAAGAqs/Kt28E7NzdXU6ZM0ebNm+3DWrZsqbi4OFksFk2aNEk9e/ZU37593Z0FAAAAAACmMzvfuv04sXnz5um7777TsGHDNGXKFIf3goOD1bp1a4eiAQAAAADwRWbnW7eD91dffaVu3brp1ltvVeXKlYu8X7t2bR07dsztwgAAAAAA8ASz863bwTszM1P16tVzPeGAAGVnZ7s7eQAAAAAAPMLsfOt28K5WrZoOHTrk8v1ffvlFsbGx7k4eAAAAAACPMDvfun1ztXbt2mnp0qVq3bq14uLiHN77/PPPtXHjRg0aNMjtwsqL6NThyug+zdtleM3olD3eLqH8sUihIQeVlZ0lGd4uxgT+vnzlEesEAACgWGbnW7eDd9++fbVr1y6NGzdOtWvXliTNmjVLZ8+eVUZGhpo3b65evXq5XRgAAAAAAJ5gdr51O3gHBQVp7NixWrdunb7++mvl5+crNzdX8fHxGjhwoDp06CCLxeJ2YQAAAAAAeILZ+dat4H3x4kXNnTtX119/vTp06KAOHTq4XQAAAAAAAN7iiXzr1s3VgoOD9fnnn+v06dNlXQ8AAAAAAB7jiXzr9l3NExMTdeDAgbKsBQAAAAAAjzM737p9jfcDDzygl156SXXr1lWnTp0UGBhYlnV5R36elJdXdLjFIgVc8hvFZW0s58/b/21YLFJYmNP3Llek7YULkuHilsMWi4wrtLVYLNK5c9L58w7T1YULsriariQjPNy9tllZCsm+4LJtdsjvNVhzshWQn++6bXBoQT9LCsq5qMB8J+vBjbYXrSEy/rfuAnNzFJSXW0Ztg2UEBJa+bV6ugvJyFCxDhpM7TOcEWZUfGPR729wcl9O9tG1AXq6sxbTNDbIqz422lvw8BedcdN02MEh5QVbHthY5XT7HtvkKznH9HMTStM0LCFSuNbjghWEo5GJWmbTNDwhQjjXE/rq4bb1UbS0Bygl2r23wxSyXf5+GxaKLwaHO2162Ti5va72YrQCjmL/PS/+WS9P2Sn/3Zu0jQn5ftnK5jyjh373P7iMu2d5yA5zsI1xN19/2EZf0Q77Fx/cRV2jr8/sIi6RL+sHvv0eUZh9Rwu8cPvU9okRtfWQfkX3B5dNC/Op7xBXaFvd3b811vU59jdn51mIYxaSsYowaNUpnzpyRzWaT1WpVdHS0goODHSdusejVV18tk0I9YlI3KW19kcF51asrp3lz++uQL76Q5dLwPe/3DT77llt0cuFC++uaTZooMCPD6ewuNm2qE8uX21/H3Hyzgg4edNo2p1EjHV+zxv66RufOsu7c6bRtbp06OvbNN/bX1Xv0UPDWrU7b5kVH6+iPP9pfV+vfXyEbNzptmx8WpiO7d9tfR//xjwr94gunbSVp8LT/2v/9xHt/1c1b1rpsO3TKSvsH7KOzXlSHr1Ndtn3slRSdqRIpSRoy93Xd9t/FLtv+v4nzdaJaweMA7v34HfX6fJ7LtmOen6VDtepLkvou/bf6LZvpsu3zf35XexOulST1/GyuBi1y/Ui5iU+9qZ8aFWw/t639REPmT3HZ9tXHJ+uHJrdIkjpsXKFHZ7/ksu2bD0/Qpps6S5JafbdG/+/9cS7bvnv/M/rvLXdIkpr9uFGj3/mzy7Yz73lSqzr1lSRdu/N7PffG/3PZdk6f4VrW7V5JUuL+n/S3yY+6bPtxzyH6pNdQSVLtw/v0yt8ecNl26a0DNbff45Kk6ifT9eZz97hsu6pDsmbe+ydJUpUzNv1zTG+Xbf/burvefWCspIIPqn8/ebvLtt+06KR/DHvB/vo/w11f6/P9Da312ohX7K+n/79uCnXxwb2jYTNN+tM/7K+njb5TEWedn9a0J/4a/fUv/7K/nvLsANXIOOK07cG4BP35r7Ptrye/cL/qpO932vZ4dKyenLTA/vqFlx9Rg19/dto2s3JVDX91if31s68/oet2/eC0bVZwqB568zP761Fvj1HztK+dtpVM3Ee8mqKc6rHKys7SkDnsIyT2Eewj/H8fMfyfXypTFsngewT7CA/sIyxShAxNe6yjy7bsIwqcf/Ahhf/7fafv+Rqz863bR7wrV66sKlWqqFatWu5Own8MvOTo8q/eK8PXvNq7gf3fUUsrFdt2Us9E+9H0yC+qSK4/hzW+e4Lyq1WTJFX9uqr0X9dtx94ar7y6dSVJEVsjpc9dtx3Vua5yGxfUXGVXtLTMddsnOtRRTrOCtpUOVZMWuW77WJtautimoG14Rg1pvuu2Q2+OVfatBW3DLtSQZrtu+8eWNXX3nQVtQy07pGL2aQOa1dCd/1sfIeH7pHdct02+sYa6/a9t8Iaj0huu2/a8vpo6/a+t9Ycz0mTXbbs1jtYt/2sb9Euu9DfXbTslRarF/9oGHgiWnnPdtk39qrr+f20DTp6Uxrhu27JuFSX9r63l/HnpSddtb4yr5LANa7jrttfGhDu0Df6T6zteNqge5tC20rOuf02tGxni+Hc0KUhy/juealYJdmhb441gKd1526jwIL3au4FGp+xxOW8AAICKxOx86/YRb3+U8+lTyju2q+gbVzjV/FIZXd706qnmsbGxSj9yxKGtmaeaW4o5RaxUbcPC7Kd9KTvb8YyCq2kbGvr7urt4UZZc16dylaptSIhUePpJadrm5CggN1exsbE6cuSILv/zM4KDpaAge1tLjutTuRza5ubKctH1qTyG1SpZraVvm5cnS7br07OMoCCp8JfA/7Ut3A4vXz6Htvn5smS5PpWrVG0DA6WQ/51GZRgFfxtl0TYgQAr9/TSqYv+WS9PWC5ejFFkn/2tbGLx9/jRSlf5U89DQMGVlZynoop+fRuqjp5qHhIQqm1PN7f3AqeYu2pbhqeaWKpHKupgtGZxqfvmp5oXbIaeaO+fWPsIihQaHyDhj41RzFf933/qaOD1yexOXNVYkBO9L5CwZpbzjToJ3KWR0d32qkNksFovi4uKUnp5eJNDBd/j7evL35SuPXK0Tvz3ibZFCQ0KV5eSLJjyA/i9AP3gOfe0afWMO+rXE2jSM0Yiujb1dhk9w+1RzScrJydHq1av1/fff69ixY5KkmJgYNW/eXF26dClyTjwAAAAAAL7IzHzrdvA+efKkJk6cqMOHDysyMlKxsbGSpP379+uHH35Qamqqnn/+eVX737W4AAAAAAD4IrPzrdvBe/r06Tp+/LieeuoptW7d2uG9jRs36u2339b06dM1ZkwxdzoCAAAAAMDLzM63bgfvH3/8UT179ixSlCTdcsst2rdvn1asWOHu5AEAAAAA8Aiz823AlZs4FxYWpqpVq7p8PzIyUmGX3lkbAAAAAAAfZHa+dTt4d+rUSWvXrlW2k0cNZWVlac2aNerSpYvbhQEAAAAA4Alm51u3TzVPSEjQ999/ryeffFIdO3a0X3x+5MgRffnll6pcubLq1aunb775xmG8m2++2e1ifZ03HyUGoCi/fVxWaVmk0JCDPPYEAADABbPzrdvB+80337T/e9GiRUXez8jIcGhTaP78+e7OEgAAAACAMmd2vnU7eI8bN87dUQEAAAAA8Blm51u3g/d1111XqvYXL17Uxo0bZbPZFBkZ6e5sAQAAAAAoU2bnW7dvrlZa58+f1zvvvKODBw96apYAAAAAAJS50uZbjwVvAAAAAAAqIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACZy+67mKCo6dbgkKaP7NC9XAnjW6JQ9v7+wSKEhB5WVnSUZ3qsJAAAA8BUc8QYAAAAAwERuB+9du3aVZR0AAAAAAHiF2fnW7VPNn3vuOcXGxqp9+/Zq3769atasWWz7ypUra9y4cUpISHB3lgAAAAAAlDmz863bwXvkyJFat26dPv74Y3300Udq1KiR2rdvrzZt2qhy5cpFZxQUpOuuu87d2QEAAAAAYAqz863bwbtdu3Zq166dMjMztWHDBq1fv17Tp0/XrFmz1LRpU3Xo0EEtW7ZUUBD3bwMAAAAA+C6z8+1Vp+KIiAh1795d3bt315EjR7R+/XqtX79eb7zxhsLDw9W6dWt17NhR11xzzdXOCgAAAAAA05iVb8v0cHRwcLBCQkJktVolSRaLRd9++62++OILJSYmasSIEapTp05ZzhIAAAAAgDJXlvn2qoP3hQsX9PXXX2v9+vXasWOHLBaLmjVrpv79++umm25SQECANm3apNmzZ+udd97Riy++eLWz9HmFz/MuDs/6hj95tXcD+78tFovi4uKUnp4uw+BB3r6goq2Tira8vsaM/h+dsqdMpgMAwJWYlW/dDt6bN2/WunXrtGXLFuXk5KhBgwZ64IEH1LZtW1WpUsWhbevWrXX27FlNnz7d3dkBAAAAAGAKs/Ot28H7tddeU7Vq1dSzZ0917NhRtWrVKrZ9QkKC2rdv7+7sAAAAAAAwhdn51u3g/de//lXXX399idsnJSUpKSnJ3dkBAAAAAGAKs/NtgDtFSSpVUQAAAAAA+Cqz863bwRsAAAAAAFwZwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAE7l9V3OUTEb3ad4uAQCAcu3V3g28XUKpWSwWxcXFKT09XYZheLscv+bpvh6dssf0eQDwPxzxBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAEzkkzdXS01N1ZIlS2Sz2RQfH6+hQ4cqKSnpiuN99dVXevPNN9WyZUuNGTPGA5UCAAAAAMqSP+ZBnzvivWHDBs2ePVv9+/fX5MmTFR8fr0mTJun06dPFjnfs2DF98MEHuvbaaz1UKQAAAACgLPlrHvS54L106VJ17dpVnTt3Vp06dTRs2DAFBwdrzZo1LsfJz8/XW2+9pQEDBigmJsaD1QIAAAAAyoq/5kGfOtU8NzdXe/fuVXJysn1YQECAmjRpop07d7ocb+HChYqIiFCXLl30008/XXE+OTk5ysnJsb8ODAxUSEjIVdXuisViMWW6xc3Lk/NE6fn7evL35SuPKto6qWjL62vo/wL0g+d4vK/L1Sq1/P5/C8+TLzv0a2lduHBBhvF7X1mtVlmt1iLtPJUHvcGngndmZqby8/MVGRnpMDwyMlKHDx92Os7PP/+sL774Qq+88kqJ57No0SItXLjQ/jo5OVmDBg2SNShI1tBQt2p3atB8xZXd1EosNjbWC3NFafn7evL35SuPKto6qWjL62vo/wL0g+d4qq//86h53+4emrnZlOmGmnSAqaKjX68sKKggbo4fP1779u2zD+/fv78GDBhQpL2n8qA3+FTwLq0LFy7orbfe0qOPPqqIiIgSj9enTx/16tXL/jowMFCSlJObq7ysrDKr71R6eplNqyQsFotiY2N15MgRh1+U4Fv8fT35+/KVRxVtnVS05fU19H8B+sFz/Kmvs7LL7ntoAYtCQ0KUlZ0tqXz3jW+hX0sqNzdXUkHwvvyId1lwNw96g08F74iICAUEBMhmszkMt9lsRX71kKSjR4/q+PHjmjx5sn1Y4QodOHCgpkyZ4vTXT1enNpQ1b+38DcMo9x88FYG/ryd/X77yqKKtk4q2vL6G/i9AP3iOX/R1WZdvPw3aIB+WJfq11MLCwkrUzlN50Bt8KngHBQUpMTFRaWlpatWqlaSCC+XT0tLUvXv3Iu1r1aql1157zWHYvHnzlJWVpSFDhqh69eoeqRsAAAAAcHX8OQ/6VPCWpF69euntt99WYmKikpKStHz5cmVnZ6tTp06SpKlTpyo6OlqDBg1ScHCw6tWr5zB+pUqVJKnIcAAAAACAb/PXPOhzwbtNmzbKzMzUggULZLPZlJCQoLFjx9pPLThx4gR3CAUAAAAAP+SvedDngrckde/e3empBFLBhfnFGTFihAkVAQAAAAA8wR/zYIC3CwAAAAAAwJ8RvAEAAAAAMJFPnmruL6JTh5e4bUb3aSZWAgAAAADwFo54AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIp7j7QU8sxsAAMD/jU7Z4+0SAPgIjngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACYK8nYBFVF06nBJUkb3aV6uBAAAAGZ5tXcDj81rdMoej80LQOlxxBsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMFGQtwuoyKJTh9v/ndF9mhcrAQAAQHn2au8GRYZZLBbFxcUpPT1dhmF4oarya3TKHm+XAD/DEW8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwERB3i4ABaJTh7s1Xkb3aWVcCQAAAOB7Rqfs8XYJgNs44g0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIl4jnc5V+T536GhisrKuurp8nxwAAAA+JJXezdweM1zvVGecMQbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADBRkLcLgG+KTh0uScroPs3LlQAAAADS6JQ93i4BcBtHvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAEwV5uwD4tujU4d4u4Yoyuk/zdgkAAAAw2au9G5Sq/eiUPSZVApQeR7wBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARz/FGsXhGNgAAAHwBz+VGecYRbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAEzkk3c1T01N1ZIlS2Sz2RQfH6+hQ4cqKSnJadvPP/9c//3vf3XgwAFJUmJiou69916X7QEAAAAAvssf86DPHfHesGGDZs+erf79+2vy5MmKj4/XpEmTdPr0aaftd+zYobZt22rcuHGaOHGiqlWrpokTJyojI8PDlQMAAAAAroa/5kGfO+K9dOlSde3aVZ07d5YkDRs2TFu2bNGaNWuUnJxcpP0TTzzh8Pqxxx7TN998ox9//FEdO3Z0Oo+cnBzl5OTYXwcGBiokJKTsFsJbLJbf/28YZTRJS5lMB78r7FN/7Vt/X77yqKKtk4q2vL6G/i9AP3gOfe2a3/WNzyyG5ff/W8rmO7e/u3DhgoxL8onVapXVanXa1hN50Bt8Knjn5uZq7969Dh0aEBCgJk2aaOfOnSWaRnZ2tnJzc1W5cmWXbRYtWqSFCxfaXycnJ2vQoEGyBgXJGhrqdv2+ItTdHxEGzS8yKO4qa4FrsbGx3i7BVP6+fOVRRVsnFW15fQ39X4B+8Bz62jV/6ZvQkIPeLsGB29+5K5CgoIK4OX78eO3bt88+vH///howYECR9p7Kg97gU8E7MzNT+fn5ioyMdBgeGRmpw4cPl2ga//nPfxQdHa0mTZq4bNOnTx/16tXL/jowMFCSlJObq7ysrNIX7issFoWGhCgrO9utI96n0tNNKAqXs1gsio2N1ZEjRxx++fMX/r585VFFWycVbXl9Df1fgH7wHPraNX/rm6xsX/mefsl3bpX/fjVTbm6upILgffkRb2c8lQe9waeC99VavHixvvrqK40fP17BwcEu2xV3akO5Vrgxu7lj9YcdcnliGIZf97m/L195VNHWSUVbXl9D/xegHzyHvnbNb/rGVxbBfnq54Ts1+biwsDCPzKekedAbfOrmahEREQoICJDNZnMYbrPZivzqcbmUlBQtXrxYzz33nOLj480rEgAAAABQ5vw5D/pU8A4KClJiYqLS0tLsw/Lz85WWlqZGjRq5HO/TTz/Vxx9/rLFjx6pBgwaeKBUAAAAAUIb8OQ/6VPCWpF69emn16tVau3atDh48qPfff1/Z2dnq1KmTJGnq1KmaM2eOvf3ixYs1f/58DR8+XDExMbLZbLLZbMoqz9dqAwAAAEAF5K950Oeu8W7Tpo0yMzO1YMEC2Ww2JSQkaOzYsfZTC06cOOHwSIRVq1YpNzdXr7/+usN0XN0pDwAAAADgm/w1D/pc8Jak7t27q3v37k7fGz9+vMPrt99+2wMVAQAAAAA8wR/zoM+dag4AAAAAgD/xySPe8I7o1OHeLsF0Gd2nebsEAAAAuOHV3r5x0yyLxaK4uDilp6c7fUzb6JQ9XqgKvo4j3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiYK8XQDgSdGpw71dQoHQUEVlZZXZ5DK6TyuzaQEAAMB9r/Zu4Pa4o1P2lGEl8CUc8QYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQ8xxvwA5c/n5znegMAALiv2OdpW6TQkIPKys6SDM/VhPKNI94AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgImCvF0AgLIXnTrcuwWEhioqK8u7NcBRRVsnZbi8Gd2nlcl0AMBMo1P2FB1okUJDDiorO0syPF8TgN9xxBsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMFGQtwsAAMCXRacO93YJ5U9oqKKysrxdhffRD55DX2t6sPPhocGhysr3z77J6D7N7XFHp+wpw0qAK+OINwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgoiBvFwAAAACUFzuOnvN2CaUSEJCl/Pw8b5dhjln3uz3q01c564CAQJ/p179HjfN2CSgBjngDAAAAAGAigjcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmIngDAAAAAGAinuMNAAAASVJG92nFvj86ZY+HKikFixQaEKqsnCzJ8MD8ojwwj7JikUJDQpWV7aG+qSjoV7iBI94AAAAAAJiI4A0AAAAAgIkI3gAAAAAAmIjgDQAAAACAiQjeAAAAAACYiOANAAAAAICJCN4AAAAAAJiI4A0AAAAAgImCvF0A/EtG92neLsHnWSwWxcXF6VR6ugzD8HY5Zc7fl688qmjrpKItr68xo/9Hp+wpk+l4lEUKDQhVVk6WVJ42w/LY1wBQDnDEGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADARwRsAAAAAABMRvAEAAAAAMBHBGwAAAAAAExG8AQAAAAAwUZC3C/BFGd2nebsEt1gsFsXFxelUeroMw/B2OQAAlIlXezfwdgmlVviZnM5nsunoa9d8sW9Gp+zxdgmAV3DEGwAAAAAAExG8AQAAAAAwEcEbAAAAAAATEbwBAAAAADCRT95cLTU1VUuWLJHNZlN8fLyGDh2qpKQkl+03btyo+fPn6/jx44qNjdXgwYPVokULD1YMAAAAACgL/pgHfe6I94YNGzR79mz1799fkydPVnx8vCZNmqTTp087bf/LL7/ozTffVJcuXTR58mT94Q9/0KuvvqrffvvNw5UDAAAAAK6Gv+ZBnzvivXTpUnXt2lWdO3eWJA0bNkxbtmzRmjVrlJycXKT98uXL1axZM/Xu3VuSNHDgQP34449KTU3VI4884nQeOTk5ysnJsb8ODAxUSEiI/bXFYinDJfKcwrrLa/0Vhb+vJ39fvvKooq2Tira8vob+L0A/eA597ZpP9o0PleI+y+//t/jGY9p83YULFxweaWe1WmW1Wp229UQe9AafCt65ubnau3evQ4cGBASoSZMm2rlzp9Nxdu7cqV69ejkMa9q0qTZv3uxyPosWLdLChQvtr5OTkzVo0CBZ73xNVklxV7UU3hcbG+vtElAC/r6e/H35yqOKtk4q2vL6Gvq/AP3gOfS1a77UN/95tLx/04Y7xo8fr3379tlf9+/fXwMGDCjSzlN50Bt8KnhnZmYqPz9fkZGRDsMjIyN1+PBhp+PYbDZVrVrVYVjVqlVls9lczqdPnz4OKycwMNA+/3PnzrlXvA+wWCyKjY3VkSNHHH5Rgm/x9/Xk78tXHlW0dVLRltfX0P8F6AfPoa9do2/MQb+WXKVKlRQREaHx48cXOeLtjKfyoDf4VPD2lOJObfCHPx7DMPxiOfydv68nf1++8qiirZOKtry+hv4vQD94Dn3tGn1jDvq15MLCwrxdgtf51M3VIiIiFBAQUOTXCZvNVuRXj0KRkZFFLrQ/ffq0y/YAAAAAAN/jz3nQp4J3UFCQEhMTlZaWZh+Wn5+vtLQ0NWrUyOk4jRo10o8//ugwbNu2bWrYsKGptQIAAAAAyo4/50GfCt6S1KtXL61evVpr167VwYMH9f777ys7O1udOnWSJE2dOlVz5syxt+/Ro4e2bt2qJUuW6NChQ1qwYIH27Nmj7t27e2kJAAAAAADu8Nc86HPXeLdp00aZmZlasGCBbDabEhISNHbsWPupAidOnHB4JELjxo31xBNPaN68eZo7d67i4uI0evRo1atXz0tLAAAAAABwh7/mQYvBHQHsMjMzdfbsWW+X4TaLxaK4uDilp6dzowcf5u/ryd+XrzyqaOukoi2vr6H/C9APnkNfu0bfmIN+LbnKlSsrIiLC22X4BJ871RwAAAAAAH9C8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAEwU5O0CfInValXlypW9XcZVq1SpkrdLQAn4+3ry9+UrjyraOqloy+tr6P8C9IPn0Neu0TfmoF+vzGq1ersEn2ExDMPwdhHelpOTw0YBAAAAADAFp5qrIHj/8MMPys7O9nYpqACys7M1Z84ctjd4DNscPIntDZ7GNgdPY5uDOwje//Piiy8qLy/P22WgAsjLy9PixYvZ3uAxbHPwJLY3eBrbHDyNbQ7uIHgDAAAAAGAigjcAAAAAACYieKvgbnv9+/fnBmvwCLY3eBrbHDyJ7Q2exjYHT2Obgzu4qzkAAAAAACbiiDcAAAAAACYieAMAAAAAYCKCNwAAAAAAJiJ4AwAAAABgIoI3AAAAAAAmCvJ2Ad6WmpqqJUuWyGazKT4+XkOHDlVSUpK3y4IfWrRokTZt2qRDhw4pODhYjRo10n333adatWp5uzRUAIsXL9acOXPUo0cPDRkyxNvlwE9lZGToww8/1A8//KDs7GzFxsbq8ccfV4MGDbxdGvxQfn6+FixYoHXr1slmsyk6OlodO3ZUv379ZLFYvF0eyrkdO3YoJSVF+/bt06lTpzRq1Ci1atXK/r5hGFqwYIFWr16tc+fO6ZprrtHDDz+suLg4L1YNX1ahj3hv2LBBs2fPVv/+/TV58mTFx8dr0qRJOn36tLdLgx/asWOHbr/9dk2aNEnPPfec8vLyNHHiRGVlZXm7NPi53bt3a9WqVYqPj/d2KfBjZ8+e1fPPP6+goCCNHTtWb7zxhu6//35VqlTJ26XBTy1evFirVq3SQw89pDfeeEODBw9WSkqKVqxY4e3S4Aeys7OVkJCghx56yOn7n376qVasWKFhw4bpxRdfVEhIiCZNmqSLFy96uFKUFxU6eC9dulRdu3ZV586dVadOHQ0bNkzBwcFas2aNt0uDH3r22WfVqVMn1a1bVwkJCRoxYoROnDihvXv3ers0+LGsrCy99dZbevTRRwlAMNWnn36qatWq6fHHH1dSUpJiYmLUtGlTxcbGers0+KmdO3eqZcuWatGihWJiYtS6dWvdeOON2r17t7dLgx9o3ry5Bg4c6HCUu5BhGFq+fLn69u2rP/zhD4qPj9f//d//6dSpU9q8ebMXqkV5UGGDd25urvbu3asmTZrYhwUEBKhJkybauXOnFytDRXH+/HlJUuXKlb1cCfzZ+++/r+bNm+vGG2/0dinwc99++60SExP1+uuv6+GHH9aYMWP0+eefe7ss+LFGjRopLS1Nhw8fliTt379fv/zyi5o3b+7lyuDvjh07JpvN5vDZGh4erqSkJHIEXKqw13hnZmYqPz9fkZGRDsMjIyPtO3DALPn5+Zo5c6YaN26sevXqebsc+KmvvvpK+/bt00svveTtUlABHDt2TKtWrVLPnj3Vp08f7dmzRzNmzFBQUJA6derk7fLgh5KTk3XhwgU99dRTCggIUH5+vgYOHKj27dt7uzT4OZvNJkmqWrWqw/CqVava3wMuV2GDN+BN06dP14EDB/TCCy94uxT4qRMnTmjmzJl67rnnFBwc7O1yUAHk5+erQYMGGjRokCSpfv36+u2337Rq1SqCN0yxceNGrV+/Xk888YTq1q2r/fv3a+bMmYqKimKbA+BzKmzwjoiIUEBAQJFfpWw2W5Gj4EBZmj59urZs2aIJEyaoWrVq3i4Hfmrv3r06ffq0/vznP9uH5efn66efflJqaqrmzJmjgIAKe7URTBAVFaU6deo4DKtTp46++eYbL1UEf/fhhx/qrrvuUtu2bSVJ9erV0/Hjx7V48WKCN0xVmBVOnz6tqKgo+/DTp08rISHBO0XB51XY4B0UFKTExESlpaXZb5qQn5+vtLQ0de/e3cvVwR8ZhqF///vf2rRpk8aPH6+YmBhvlwQ/1qRJE7322msOw6ZNm6ZatWrprrvuInSjzDVu3LjIpVqHDx9WjRo1vFQR/F12dnaRfVlAQIAMw/BSRagoYmJiFBkZqR9//NEetM+fP6/du3erW7du3i0OPqvCBm9J6tWrl95++20lJiYqKSlJy5cvV3Z2Nr+SwhTTp0/X+vXrNWbMGIWFhdnPtggPD+dUYJS5sLCwIvcPCAkJUZUqVbivAEzRs2dPPf/88/rkk0/Upk0b7d69W6tXr9Yjjzzi7dLgp2666SZ98sknql69uurUqaP9+/dr6dKl6ty5s7dLgx/IysrSkSNH7K+PHTum/fv3q3Llyqpevbp69OihTz75RHFxcYqJidG8efMUFRWlP/zhD16sGr7MYlTwnwVTU1OVkpIim82mhIQEPfjgg2rYsKG3y4IfGjBggNPhjz/+OD/2wCPGjx+vhIQEDRkyxNulwE999913mjNnjo4cOaKYmBj17NlTt956q7fLgp+6cOGC5s+fr02bNun06dOKjo5W27Zt1b9/fwUFVehjSygD27dv14QJE4oM79ixo0aMGCHDMLRgwQJ9/vnnOn/+vK655ho99NBDqlWrlheqRXlQ4YM3AAAAAABm4iI/AAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATETwBgAAAADARARvAAAAAABMRPAGAAAAAMBEBG8AAAAAAExE8AYAAAAAwEQEbwAAAAAATPT/AWZadXIDHLodAAAAAElFTkSuQmCC", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "y_pos = norm(y_pos)\n", + "y_neg = norm(y_neg)\n", + "\n", + "pos_neg_hist(y_pos, y_neg)" + ] + }, + { + "cell_type": "code", + "execution_count": 93, + "metadata": { + "ExecuteTime": { + "end_time": "2024-05-19T12:28:04.840079Z", + "start_time": "2024-05-19T12:28:04.606559Z" + } + }, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 93, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAArUAAAIoCAYAAAB3WBPcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAABy1ElEQVR4nO3dd3gU1eLG8XeTbCAJhBBKAqEkVOkgKDUEqUrvIB0U5YI0UbxiARQEFAEVUC7wAwTpXaQJ0kKRIkVA6Z1QY6ghdX5/5GYvSwqbkLKr38/z8Bx25szMmdnd2TezZ86aDMMwBAAAADgwp8xuAAAAAPCsCLUAAABweIRaAAAAODxCLQAAABweoRYAAAAOj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1yHDnz5+XyWRSjx49MrspKbJx40bVqFFDXl5eMplMatmyZWY3KdW2bt0qk8mkESNGWE2vU6eOTCZTum139uzZMplMmj17drptA/886f26tXcmk0l16tRJ0TI9evSQyWTS+fPn06VN9iq9XysjRoyQyWTS1q1b020baSUjPotTs41n+Zwg1D4mJiZG06dPV1BQkLy9vWU2m5U3b16VL19er7/+ulavXp2h7UnuRGVvwdBkMln9c3Z2Vu7cuVW3bl3Nnz8/XbaZkQHp/PnzatGihc6dO6devXpp+PDh6tixY7LLxAfHx/+5urqqQIECateunXbv3p3u7c5oSYVlpEz8azsl/+xRzZo1ZTKZtHbt2qfWLVmypEwmkw4ePJgBLfv7y8hwFf96Te7zKP7ckNLwnVniw6+t/+zls/ifziWzG2AvYmJi1LRpU61fv15eXl5q0qSJChQooMjISB07dkzz58/Xn3/+qebNm2d2U+3a8OHDJUlRUVH6888/tWrVKm3ZskX79+/XhAkTMrl1qbdp0yY9evRIX375pTp16pSiZQsXLmw54T148EB79uzR0qVLtXz5ci1dulStWrVKhxanzvfff6+HDx+m2/pbtWqlatWqKV++fOm2jb+DihUrWt5L8c6fP685c+ZYvZ7sXe/evbVr1y7NmDFDjRs3TrLetm3bdPLkSVWpUkWVKlXKwBb+Pfzxxx9yd3fP7Gb8rfTo0SNBAF+5cqUOHz6sFi1aqGLFilbznnyMzEGo/a8FCxZo/fr1qlChgrZt26YcOXJYzX/48KF+/fXXTGqd43jyCt3mzZvVoEEDTZo0SQMGDJC/v3+mtOtZXb16VZKUP3/+FC/r7++f4LgMHz5cn3zyiYYMGWJXobZQoULpuv4cOXIkeG8hoYoVKyb4kNy6davmzJmT6OvJXnXo0EGDBg3SmjVrdP36dfn4+CRab8aMGZKkN954IyOb97fx3HPPZXYT/nYS+8Px/PnzOnz4sFq2bOkwf1j+09D94L927dolKe6FnNiHrru7u1566aVEl120aJHq1asnb29vZc2aVf7+/nr11Ve1f/9+S507d+7oiy++UN26dVWgQAG5uroqT548at68eYKvoeO/ypHirmA8/hXHiBEjNGLECAUEBEiS5syZYzX/ya/iN2zYoMaNGyt37tzKkiWLihYtqnfffVdhYWEJ9sPf31/+/v66e/eu3n77bfn7+8tsNj/TB2i9evX03HPPyTAM7du376n1Q0JC1K9fP/n7+1uOUevWrXXgwAGrenXq1FHPnj0lST179rQ6Brb2EVu8eLFq166tHDlyyM3NTeXKldOYMWMUERFhqRP/lVn8VbOXXnrJsp1n+VqvX79+kqRz587p5s2bln0ymUyKjIzUJ598opIlSypLlixWJ8/Lly/rrbfeUpEiRZQlSxblypVLzZs3T/LYXr9+Xa+99pp8fHzk5uamihUras6cOUm2K7n+Zhs3blSzZs2UN29eZcmSRQULFlSLFi20adMmSXHvnfj3yMiRI62ek/hjlVyXkQMHDqhNmzaW9RcuXFh9+/ZVSEhIgrqP9wecNm2aypUrp6xZs8rHx0dvvPGG7ty5k2CZI0eO6NVXX5W/v7+yZMmiPHny6Pnnn9egQYMUFRWV5DGRpD179shkMiX7B0ipUqWUJUsWhYaGSpIMw9CcOXNUo0YN5cmTR1mzZlXBggXVqFEjLVq0KNntpcTjXZFOnjypDh06KG/evHJycrIc9/j3dmKS+5r6zz//VI8ePVSwYEG5urrKx8dHnTp10okTJ2xqm5ubm7p06aKoqKgkuwmFhYVp6dKlypYtm1599VVJca+TNm3aqEiRInJzc5Onp6dq1qypefPm2bTd+HUk1z0pqa/Co6OjNXXqVFWrVk2enp5yd3dXpUqVNHnyZMXGxiaov3r1atWrV0/58uVTlixZlD9/fgUFBWnq1KlPbeOGDRtkMpn0wQcfWE3fsmWL5b1z6dIlq3kdOnSQyWTS2bNnk9wXf39/jRw5UpL1OSup97at76H0kJLPxnjx+3vt2jW9/vrr8vPzk7Ozs9VzvXDhQlWuXFlubm7Kmzevunbtark4kVGWLl2qF198Ue7u7vL29lbHjh115cqVBPXS+tx/7949ffrppypbtqw8PT2VPXt2FS1aVB06dEjwWRrv/Pnz6tixo3Lnzq2sWbOqSpUqWrNmTaJ1IyIiNHbsWJUrV07u7u7y9PRUYGCgFi9enKLjc/r0abVr1045c+aUh4eHatSooZ9++ilF63gSV2r/K1euXJKkkydP2ryMYRjq2bOn5syZo9y5c6t169bKkyePLl++rC1btqhkyZKqUqWKpLivhz744APVrl1bTZo0Uc6cOXXx4kWtXr1a69at048//qiXX35Z0v++ehw5cmSCrxrjT1xhYWH66quvVKFCBasblh6/ujNy5EiNGDFC3t7eatq0qfLmzasjR45o/PjxWrt2rXbv3i1PT0+rfYqMjFTdunUVGhqqhg0bytPT0xKgU8swDEl6ar+/c+fOqVatWrp69arq1q2rV199VZcuXdKSJUv0008/admyZWratKmkuFDj5eWlVatWJfgqyMvL66ltGjZsmMaMGaPcuXOrU6dOypYtm9atW6dhw4Zpw4YN2rhxo1xdXeXv76/hw4dr69at2rZtm7p3724JB89y1Tn+mEgJj0ubNm20b98+vfLKK2rZsqXy5s0rSfrtt9/UsGFDhYaGqlGjRmrdurVu3bqllStXqlatWlqxYoXVV7y3bt1SjRo1dPbsWdWqVUu1atVSSEiI+vTpo4YNG6aovfFXlrNly6aWLVuqYMGCunr1qnbt2qV58+apfv36ltfhnDlzFBQUlOBDNjlr1qxRmzZtZBiG2rZtq8KFC+vAgQP69ttvtWrVKgUHByf6Ohw6dKg2bNigZs2aqWHDhtqyZYumT5+u06dP65dffrHUO3LkiKpWrSqTyaTmzZsrICBAd+/e1enTpzV16lSNGjVKZrM5yfZVq1ZNJUuW1Nq1a3X79m3L+SLe3r179eeff6pNmzby9vaWJH3wwQcaM2aMAgIC1L59e+XIkUMhISHat2+flixZog4dOjzlqKfMmTNnVLVqVZUoUUKdO3dWeHh4gvd3Sqxfv16tW7dWVFSUmjVrpmLFiuny5ctavny5fvrpJ23ZskXPP//8U9fTu3dvTZkyRTNnztR7772XYP68efP06NEj9e7dW9myZZMk/etf/1KZMmVUu3Zt5cuXT7dv39batWvVtWtXnThxQp9++mmq9ys58fu6YcMGlSxZUp06dVLWrFm1ZcsW9e/fX7/++qvmzp1rqf+f//xHb775pnx9fdWsWTPlzp1bN27c0JEjRzRr1iz17ds32e0FBgbK1dVVmzdv1ujRoy3TN2/ebPX/+M8AwzC0ZcsW+fv7q0iRIkmud9CgQVq5cmWCc1ZibH0PpZeUfDY+LjQ0VNWqVVO2bNnUunVrOTk5Wb4JmDhxot5++215eXmpW7du8vLy0oYNG1SjRo0M+6Zo6tSpWr16tZo3b66goCD9+uuvWrRokQ4fPqxDhw4pS5YsCZZJi3O/YRh6+eWXtWvXLlWvXl2vv/66XFxcLLkkMDBQlStXttruhQsX9OKLL6pIkSLq2rWrQkNDtWjRIstFi8cv6EVGRqpRo0batm2bnnvuOfXr108PHz7U0qVL1aFDBx06dEifffbZU4/PqVOnVL16dd2+fVuvvPKKKlasqNOnT6tly5Z65ZVXUn/gDRiGYRi//fabYTabDZPJZHTp0sVYtmyZcf78+WSXmTZtmiHJeOGFF4ywsDCredHR0cbVq1ctj8PCwoybN28mWMelS5eMfPnyGc8991yCeZKMoKCgRLd97tw5Q5LRvXv3ROf/8ssvhiSjevXqxl9//WU1b9asWYYkY9CgQVbTCxcubEgy6tWrZ9y/fz/R9SZFkpHYy+nnn382TCaTYTKZLMczqbY3bNjQkGSMGjXKavrOnTsNZ2dnw9vb27h3716C/Zg1a1aK2rpr1y5DklGwYEEjJCTEMj0qKspo2rSpIckYPXq01TLDhw83JBlbtmyxeTtbtmxJ8jn86KOPDElGkSJFLNOCgoIMSUa5cuUSvFaioqKMokWLGlmyZDG2bt1qNe/KlStG/vz5DV9fX+PRo0eW6b179070ed63b5/h4uJiSDKGDx9uNS++DY/bsGGDIckICAgwLl++nGBfLl26lGCfn1xvvMSes3v37hne3t6Gk5OTsX37dqv6Y8eONSQZDRo0sJrevXt3y3N44cIFy/SoqCgjMDDQkGT8+uuvlulvv/22IclYuXJlgjaFhoYaMTExibb3cZ999pkhyfjmm28SzOvbt68hyVi9erVlmre3t+Hn52c8ePAgQf3EzgVPk9TrKf79JMl4//33E122cOHCRuHChROdl9hrOzQ01PDy8jJy5cplHDt2zKr+77//bnh4eBiVKlWyue0vvvhiku+fChUqGJKMffv2WaadPn06Qb2IiAijbt26houLS4LXYWKv26edHxI7lvHH4q233jKio6Mt06Ojo41evXoleA09//zzhqurq3H9+vUE67f1OQ4MDDScnZ2tPkOqVatmVKpUyciVK5fRpUsXy/RDhw4ZkoxevXrZvC9JnbNS+h5KTvyxrlChgjF8+PBE/8Vv78l2pvazUZLRtWtXIyoqymreuXPnDLPZbOTMmdM4d+6cZXpMTIzRunXrJD+rbBG/D8l95sQf9+zZsxtHjhyxmvfqq68akoxFixZZTU/Lc/+RI0cMSUbLli0TtC0mJsYIDQ21PH783DFixAiruuvXrzckGa+88orV9Pjz4CuvvGJ17K9fv27JEDt37kywjSc/7xs0aGBIMiZNmmQ1feXKlZY2pfSz3TAMg1D7mEWLFhm+vr6WAyrJ8Pb2Nlq2bGn1YRWvbNmyhiTjt99+e6bt9u/f35BkdWIxjGcLtS1btjQkGUePHk10fsWKFY08efJYTYt/QR46dCjF+xB/vOJPYMOGDTPatGljODs7G5KMwYMHJ9v2S5cuGZKMQoUKGZGRkQnW36VLF0OSMWfOHMu01Iba119/3ZBkTJs2LcG8EydOGE5OTkZAQIDV9GcJtYULF7Ycl3feecfygeHk5GQsX77cUj/+xJZY8Ip/o7/zzjuJbmvSpEmGJOOnn34yDMMwIiMjDXd3dyN79uwJ/uAyjP+dnG0JtfFB//G2Pm2fUxJq582bZ0gyXn311QT1o6KiDH9//wTvj/j2T58+PcEy//d//5cgfMaH2g0bNjx1H5Jy6dIlw8nJyahSpYrV9IiICMPb29vImzev1Une29vb8Pf3t/pD41k8LdT6+Pgkua2Uhtr419PkyZMTXWbQoEGGpASBNykzZswwJBmdO3e2mr53715Dks0BedmyZQnOA4aRNqE2JibG8Pb2Nnx9fRMEJcMwjL/++sswmUxGu3btLNOef/55w93d3SoopNSIESMMScaqVasMwzCMu3fvGi4uLsbQoUONtm3bGvnz57fU/fLLLw1Jxg8//JDsvhiG7aHW1vdQcuKPtS3/kvpMS0xyn41J/TExatQoQ5Lx8ccfJ5h35swZw8nJKUNC7QcffJBgXvzFpiFDhlhNT8tzf3yoTex8+qT4c0fhwoWt/oiLV6hQISNXrlxW04oVK2aYTCbjjz/+SFA//n3es2fPBNtI7PM+ICAg0e3GH4/UhFq6Hzymffv2atWqlbZs2aLg4GAdPHhQwcHBWrlypVauXKlu3bpZ+mk9ePBAR48elY+Pj8136+7cuVNfffWVdu/erRs3bigyMtJq/pUrV9LsRp3du3fLbDZryZIlWrJkSYL5kZGRunnzZoKvUrNmzary5cunervx/bhMJpO8vLwUGBio1157TV26dEl2ufhhfAIDAxP9Grhu3bqaN2+eDh48qG7duqW6fVLcVznx63xSiRIlVKBAAZ07d0537txJk6+qLly4YDkuLi4uln7CQ4YMUY0aNRLUf/HFFxNMi+9bduHChUT7OJ86dUpS3Fd5jRs31p9//qmHDx8qMDAw0X2oU6dOsn1rHxffnzSxrwDTQnLPh4uLi2rXrq3z58/r4MGDCd4f8d17HlewYEFJ0l9//WWZ1qFDB3311Vdq2bKl2rZtq/r166tmzZoqWrSoze0sUKCA6tWrp59//lnHjx9X6dKlJUk//vijQkNDNXjwYLm4/O+U2rlzZ33zzTcqXbq02rdvr6CgIFWvXj3dvv6sUKFCol9ppkb86+3w4cOJvt7iu2n98ccfluOQnI4dO2rw4MFatmyZvvnmG+XMmVOSNH36dElxXRQed/HiRY0bN06bN2/WxYsXFR4ebjU/sX6Jz+rkyZMKDQ1V8eLFNWrUqETruLm56Y8//rA87ty5s4YMGaLSpUurY8eOCgoKUs2aNZUnTx6bt1u3bl2NGDFCmzdvVvPmzbVt2zZFR0erXr168vf319KlS/XHH3+oVKlSlu4Aib1XUsvW95AtunfvnmQf5q1btyZ5X0pqPhv9/f0tX88/Lv58EhQUlGBekSJFVLBgQV24cMGW3XkmqTmuaXHuL126tCpWrKgFCxbowoULatGihWrVqqUqVarI1dU10e1WrFhRzs7Oibb38X7N9+7d0+nTp+Xn55fozYnxr8unDcsXP79WrVqJbrdOnTratm1bsutICqH2CWazWQ0bNrT0OYyJidGyZcvUq1cvff/992rVqpVatmxpudHKz8/PpvWuWLFCbdu2VdasWdWgQQMVLVpUHh4elps5tm3bZnWD0rO6ffu2oqOjLWEqKffv37cKtXnz5n2mMS+Nx/qKpkT8TQlJDfUUPz2xG9zSY1sXL15UWFhYmgSQoKCgFN1U5uvrm2Da7du3JSnRP1Aed//+fUn/28ek7jZPbBtJCQsLU86cOeXm5mbzMinxLM99Yv2n44NlTEyMZdqLL76oHTt2aPTo0Vq6dKmlX2TJkiU1fPhwyw1KT9OjRw/9/PPPmjNnjsaNGydJlj8OunfvblV34sSJKlKkiGbNmqWxY8dq7NixcnFxUePGjfXll1+qWLFiNm3TVil5Tp8m/vUWHzqTEv96exoPDw916tRJ06ZN07x589S/f389ePBACxculIeHhzp37mype/bsWb344ov666+/FBgYqIYNGypHjhxydna2DGuWlufKePH7fOrUqWTPm4/v89tvv63cuXNr6tSp+vrrrzVp0iSZTCYFBQXpiy++SDTYPKlatWry8PCw9KPdvHmzXF1dVatWLUtf2M2bN6t48eLavn27SpcunabPta3vofSS2s/GpI6BLee+jAi1qTmuaXHud3Z21i+//KJPPvlES5cutfRjz549u7p3764xY8ZY+q4n19b49j5+c2RafU6n5efTkwi1T+Hs7Kz27dvr999/16hRo/TLL7+oZcuWlheBrVcMPvroI7m6umr//v0qVaqU1bw333wz1X+VJCVHjhyKjY213Iltq8waxD0+PF67di3R+fF3wKdFyHx8W4ldqUvLbaVGYs9BfFtWrVpl01jJ8fWvX7+e6PykjnNivLy8dPv2bYWHh6dLsM2o57569epas2aNIiIidODAAa1fv17ffPONOnXqpDx58qh+/fpPXUerVq3k6empefPm6bPPPtPt27e1bt06VahQQRUqVLCq6+zsrEGDBmnQoEG6ceOGgoODtXDhQi1ZskTHjh3TsWPH0uzKqpT8e9fJySnB1a94iX0AxR/rw4cPP9M3N4974403NG3aNM2YMUP9+/fXwoULde/ePfXq1cvqhrYJEybo9u3bmjVrVoJhkxYsWGDzNwxOTnGD+0RHRyeYl9w+t2rVSsuXL7dxr6Ru3bqpW7duCgsL065du7RixQr93//9nxo1aqQ///zzqVdtzWazatWqpQ0bNujatWvavHmzqlevLnd3d8s3R5s2bdLzzz+ve/fupelVWnuQ2s/GpF7vj5/7ypQpk2B+Ss59GS0tzv2SlDNnTk2cOFETJ07U6dOntW3bNk2bNk2TJ09WWFiY1c2OKZFW5+q0/Hx6EkN62Sh79uyS/ncl0sPDQ2XLltX169dt+gWc06dPq3Tp0gnetLGxsQoODk50GScnpyT/oou/ZJ/U/GrVqumvv/7SsWPHnto2exDfhSM4ODjRD6EtW7ZIktXd1k87Bk/bVmJXT0+fPq3Lly8rICDAplEUMkq1atUkSTt27LCp/nPPPSd3d3cdOnQo0aF5UnLluFq1ajIMQ+vXr39q3dQ8J8k9H9HR0ZZ9tuVOe1tkyZJFNWrU0CeffKKvv/5aUtwHhi3c3NzUvn17Xb16VZs2bdL8+fMVHR2d4Crtk/LmzavWrVtr8eLFqlu3rs6cOaOjR48+877YKmfOnLp+/XqiQ5c9PvRgvJS+3mzx/PPPq3Llyjpy5Ij27t2b5Ni0p0+flhR3J/iTUvLHf3wXhyeHxJIS3+fnnntOXl5e2rNnz1OHeEuMl5eXGjdurOnTp6tHjx4KDQ3V9u3bbVq2Xr16kuJC+9GjRy2PpbivdLdu3aqff/7Zqu7TpPb8mNFS89mYnPjzRGKvlbNnzyb6erBnz/peLFasmF577TVt27ZN2bJls/lcl5j4ocGuXLli6fbwuMQ+pxPz+Od9Yq/PZxkuk1D7XwsWLNDPP/+c6DiE165ds3wNV7t2bcv0AQMGSIr7a/LJ4BAbG2s1vqa/v79OnTplNU6eYRgaMWKEjh8/nmibcuXKleQbMGfOnDKZTLp48WKi8wcPHiwprq9aYmPzxf+ylb0oUKCAGjRooPPnz2vSpElW83799VfNnz9fOXPmtBonNL7bRFLHICm9evWSJI0aNcoyRqwUd/J/5513FBsbq9deey2Ve5I+WrRooaJFi2rKlClJ/uTo7t27Lb8GZjab1blzZ927dy9BP6z9+/frhx9+sHnb/fv3lyQNGTIk0W8mHp+WmuekZcuW8vb21oIFCxK8JidNmqRz586pfv36z9TffNeuXQn6ZUr/u1KQkl9jir96+P333+v777+Xi4uL1dfnUtw4jjt37kywbFRUlOXbk4z8BagXX3xR0dHRmjVrltX02bNnJ9rOnj17ysvLSyNHjtTevXsTzI+NjU3VB09839l33nlHe/bsUfny5VW1alWrOvFfuT+5/g0bNliCsC2qVKkiJycnzZ8/3+pX8kJDQzV06NAE9V1cXNS/f3+FhIRowIABib5eQkJCrM7XW7ZsSbTL1Y0bNyTZ/hzHX30dO3asDMNIEGrv3LmjqVOnysnJyeafmU3t+TGjpeazMTmdO3eW2WzWN998YzVmeWxsrN59991EP+PtWUrP/efOnbMawzjeX3/9pYiIiGf+tq1Xr14yDEPvvvuuVSC9deuWZai9+M/YpMR/3p87d06TJ0+2mrdq1apn+uaa7gf/9euvv+qrr76Sr6+vatWqZRkT89y5c/rpp58UHh6uFi1aqG3btpZlXn/9de3YsUNz585V8eLF1aJFC+XJk0dXr17VL7/8ol69elkCxeDBg9WnTx9VqlRJbdq0kdls1s6dO3X8+HE1a9ZMP/74Y4I21atXTwsXLlSzZs30/PPPy2w2q3bt2qpdu7ayZcumqlWraseOHercubNKlCghZ2dnNW/eXOXLl1e9evU0duxYvf/++ypevLgaN26sgIAA3b9/XxcuXNC2bdtUq1Ytm66+ZZTvvvtONWvW1LvvvquNGzeqSpUqlnFqnZycNGvWLMsVc0mWr+gmTZqk27dvW/rh9O/fP9mvP2rUqKGhQ4fq888/V9myZdW2bVt5eHho3bp1Onr0qGrVqqV333033fc3Jcxms5YvX65GjRqpSZMmqlGjhipWrCh3d3ddunRJ+/bt09mzZxUSEmL5IP3ss8+0efNmTZo0Sfv377eMU7to0SI1btxYq1evtmnbDRs21IcffqhRo0apVKlSlnFqr1+/ruDgYFWrVs1yc0jJkiXl5+enhQsXymw2q3DhwjKZTOratasKFy6c6PqzZcum//u//1O7du0UFBSkdu3aqVChQjpw4IA2btwoX19fTZs27ZmO3+eff65ffvlFgYGBCggIULZs2XTs2DGtW7dOOXPmTNEvWdWsWVPFihXTkiVLLOOaPnnDSnh4uGrVqqVixYqpcuXKKly4sB49eqSff/5Zf/zxh5o3b57gylR66t+/v2bNmqV//etf2rx5swoWLKhDhw5p9+7datq0aYJB1nPlymX5Cedq1aqpXr16KlOmjOXHAHbv3q3bt2/r0aNHKWpHp06d9M4771iuOj15g5gk9e3bV7NmzVK7du3Utm1b5c+fX0ePHtX69evVvn17m3+4Il++fOrcubPmzp2rihUrqkmTJrp7967Wrl2r2rVrJ/oN20cffaTDhw/ru+++048//qi6devKz89PN27c0KlTp7Rz506NHj3acnNcq1atlC1bNlWrVk3+/v4yDEM7duzQvn37VLlyZZu6tEhxV65y5sypGzduKHv27FY3DMUH3Bs3bqhKlSo2f4P00ksvycnJSe+//76OHj1quXL94Ycf2rR8RknNZ2Ny/P39NXbsWA0ZMkSVKlVShw4dlCNHDm3YsEFhYWEqX768jhw5kk57k/ZSeu4/fPiwWrdurRdeeEGlSpVS/vz5dfPmTa1atUpRUVGJjhWdEu+8847WrVunVatWqUKFCmrcuLEePnyoJUuW6MaNGxo6dKhq1ar11PVMmTJF1atX16BBg7Rx40ZVqFBBp0+f1ooVK1L1vFukeLyEv6mLFy8akydPNlq2bGmUKFHCyJ49u2E2mw1fX1/jlVdeMebOnZvkWJbz5s0zateubXh6ehpZsmQx/P39jU6dOhkHDhywqjdr1iyjQoUKhru7u5ErVy6jZcuWxpEjR5IceuX69evGq6++auTNm9cyDMnjQyWdOnXKaNq0qeHt7W2YTKZEh8DYsWOH0a5dOyNfvnyG2Ww2cufObVSoUMEYPHiw1biQhpH8sD9PoxSM/ZfccGSXL182+vTpYxQqVMgwm81Grly5jBYtWhh79+5NdF3r1q0zqlWrZnh4eFja8PjYhMlZsGCBUbNmTSNbtmxGlixZjNKlSxujRo0ywsPDE9RN63FqE5PYsERPun79uvHee+8ZZcqUMdzc3AwPDw+jWLFiRps2bYy5c+cmGIooJCTE6Nmzp5E7d24ja9asRoUKFYxZs2YlOfRWcm346aefjEaNGhk5c+Y0XF1djQIFChgtW7Y0Nm/ebFVv7969Rt26dQ1PT0/L6zL+uCU3zNLevXuNli1bGrlz5zbMZrNRsGBBo0+fPsaVK1cS1I0fWiex5zqxfduwYYPRo0cPo1SpUoanp6fh7u5ulChRwujfv/9Tx6NOzKeffmp5vS1dujTB/MjISGPcuHHGyy+/bBQsWNDIkiWLkTt3bqNq1arGt99+a0RERKR4m08b0iup4f3i7dixwwgMDDTc3NyM7NmzG40bNzYOHz6c7Gv73LlzRr9+/YxixYoZWbJkMbJnz26ULFnS6NKli7FixYoU74Nh/G9IPTc3twRjaMfbuXOn8dJLLxleXl5GtmzZjJo1axorVqxI8ev20aNHxjvvvGP4+fkZZrPZKFq0qPHZZ58ZUVFRSb43Y2Njje+//96oW7eukTNnTsNsNhv58+c3atasaYwePdq4ePGipe63335rtGzZ0ggICDDc3NyMnDlzGhUrVjTGjRtn3L17N0XHJX4M1caNGyeYV6JECUOSMXTo0ESXTWpf5s6da1SoUMHImjVrgnN0St9DyYl/Xyf3GkzufJjSz0Zbzqvz5883KlWqZHnvde7c2bhy5YpN59mkpGRIr6TeT4kdp7Q891+6dMl4//33jRo1ahg+Pj6Gq6ur4efnZ7z88svG2rVrbWrP09oVHh5ujB492ihTpoyRNWtWy3t0/vz5Nu+zYcRlmDZt2hg5cuQw3N3djWrVqhlr1qxJ9XCdhmEYJsNI5e3qAAAAgJ2gTy0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4RFqAQAA4PAItQAAAHB4hFoAAAA4PH4mV3G/iRwdHZ0h28qTJ49u3ryZIdsCAABIDxmZZ1xcXCw/9ZxsvQxoi92Ljo5WVFRUum/HZDJZtscPuQEAAEdkr3mG7gcAAABweIRaAAAAODxCLQAAABweoRYAAAAOjxvFAMAORUREKCIiIrObAQCJCg8PV2RkZJqtL0uWLMqSJcszrYNQCwB25sGDBzKZTMqePbvlLmMAsCdmsznNRo4yDEPh4eF68OCBPDw8Ur0euh8AgJ2Jjo6Wu7s7gRbAP4LJZJK7u/sz/2YAoRYA7AxhFsA/0bOe+wi1AAAAcHiEWgAAADg8Qi0AwOEMGjRIvXr1sjxu27atPv7442daZ1qsA+njyy+/VIMGDTK7GbBzhFoAQJoYNGiQ/Pz85OfnJ39/f9WsWVMTJ0585ps/bDF9+nQNHTrUprq7du2Sn5+f7ty5k+p12LvHn4vE/lWtWjVD2vHdd9+pdOnSevToUYJ54eHhKlmypGbOnJkhbcHfH6EWAJBmXnrpJR08eFDBwcF644039OWXX+rbb79NtG5ajnGZM2dOZcuWLdPXYS8++eQTHTx40PJPkiZMmGB5vHbtWqv6aflcPK5t27Z6+PCh1q1bl2DemjVrFBUVpdatW6fLtvHPQ6gFAKQZV1dX5c2bVwUKFFD37t0VGBiojRs3Svpfl4GvvvpKzz//vGrXri1JunLlit58802VKlVKZcqUUc+ePXXp0iXLOmNiYjRixAjL/FGjRskwDKvtPtl1ICIiQqNHj1aVKlUUEBCgmjVrasGCBbp06ZLatWsnSSpdurT8/Pw0aNCgRNcRFhamAQMGqHTp0ipatKi6dOmis2fPWuYvWrRIpUqV0tatWxUUFKTixYurc+fOun79uqXOrl271KRJExUrVkylSpVSixYtdPny5USPXfPmzTV69Girabdv31bhwoW1Z88eSdLs2bNVs2ZNFSlSRBUqVFDv3r0TXZenp6fy5s1r+SdJOXLksDxu3LixJk6cqAEDBqhkyZIaOnRoolewjx49Kj8/P6vnY+/evWrVqpWKFi2qKlWq6KOPPtLDhw8TbUfu3LnVoEEDLVy4MMG8RYsWqVGjRsqZM6dGjx6tWrVqqWjRoqpevbo+//zzZMdATayrSK9evSzPpRT3Gvjkk09UuXJlFStWTE2bNtWuXbss8y9fvqzu3burdOnSKlasmF566SVt3rw5yW3C/hFqAQDpJmvWrFbhJDg4WGfOnNGCBQs0Z84cRUVFqXPnzsqWLZuWL1+ulStXysPDQ507d7ZcPZw2bZqWLFmiL7/8UitXrlRYWJjWr1+f7HYHDhyolStX6tNPP9XWrVs1duxYubu7K3/+/Jo+fbokafv27Tp48KA++eSTRNcxePBgHTlyRLNmzdLq1atlGIa6du1qtT/h4eH67rvv9PXXX2v58uW6cuWKPv30U0lx4w2/9tprqlatmjZt2qTVq1erc+fOSQ5b1Lp1a61atcoqsK9evVo+Pj6qWrWqDh8+rI8//ljvvvuutm/frh9++EHVqlWz4VlI3LRp01S6dGlt2LDBKgwm5/z58+rcubMaN26sn3/+Wd9++6327t2rDz74IMllOnbsqJ07d1qF+QsXLmjPnj169dVXJUkeHh6aOHGitm7dqpEjR2r+/PmW5ym1PvzwQx04cEBTp07Vpk2b1LRpU6s/TIYNG6bIyEgtW7ZMmzdv1rBhw55p4H9kPn5RDACQ5gzD0I4dO7Rt2zb17NnTMt3d3V3jx4+Xq6urJGnZsmWKjY3V+PHjLWFvwoQJKlWqlHbv3q2goCDNmDFDb731lho3bixJGjt2rLZu3Zrkts+cOaMff/xRCxYssFwNLly4sGW+l5eXpLiriDly5Eh0HWfPntXGjRu1cuVKvfDCC5Kkb775Ri+88ILWr1+vZs2aSZKioqI0duxY+fv7S5J69OihSZMmSZLu3bunu3fvqn79+pb5xYsXT7LdzZo10/Dhw7V3715Ln9cVK1aoZcuWMplMunLlitzd3VW/fn1ly5ZNBQoUUNmyZZNc39PUrFlTffr0sTy+evXqU5eZPHmyWrVqZblCXKRIEX366adq06aNxowZo6xZsyZYpk6dOvLx8dGiRYs0ZMgQSdLixYuVP39+1apVS5KsQnXBggV19uxZrVq1Sn379k3Vvl25ckWLFi3S3r175evrK0nq06ePtmzZokWLFun999/X1atX1bhxY5UqVUqS9WsEjolQCwB/Y9euOWnq1Gzq2/e+fH1j0317mzZtUvHixRUdHa3Y2Fi1bNnSEmQk6bnnnrMEWkk6fvy4zp8/rxIlSlitJyIiQufPn1elSpV0/fp1VapUyTLPxcVFFSpUSNAFId6xY8fk7Oys6tWrp3o/Tp8+LRcXFz3//POWad7e3ipatKhOnz5tmebm5mYJrJLk4+OjW7duSYrro9u+fXt17txZgYGBCgwMVLNmzeTj45PoNnPlyqXatWtr+fLlqlq1qi5evKgDBw5o3LhxkqTatWurQIECql69uurUqaOXXnpJr7zyitzc3FK1j+XLl0/xMsePH9cff/yhFStWWKYZhqHY2FhdunQp0dDu7Oysdu3aafHixXr77bdlGIaWLFmiDh06yMkp7gvjVatW6f/+7/904cIFPXjwQDExMc/Uv/mPP/5QTEyMAgMDraZHRkYqZ86ckuK6K7z//vvatm2bAgMD1bhxY5UuXTrV20TmI9RmoBUrsmjIEOnLL7OoZcuEd4ICQFqbOjWbZs6MCweffHI33bdXo0YNjRkzRq6urvLx8ZGLi/XHjLu7u9XjBw8eqHz58vrmm28SrCtXrlypakNiVwvTi9lstnpsMpmswvbEiRP12muvacuWLVq9erU+//xzLViwQJUrV050fa1bt9ZHH32kUaNGacWKFSpVqpTlSmK2bNm0fv167dq1S9u3b9f48eP15Zdfau3atUlecU7Ok89FfMB8vP1Pjlzx4MEDdenSxWo4tXh+fn5Jbqtjx46aPHmygoODZRiGrl69qg4dOkiS9u/fr/79+2vIkCGqU6eOsmfPrlWrVuk///lPkutLrAvH491CHjx4IGdnZ61bt07Ozs5W9eK7GHTq1ElBQUHavHmztm/frsmTJ+vjjz9OdN/gGOhTm4EGD/bSo0dxJQBkhL597+u11+6rb9/7GbI9d3d3BQQEyM/PL0GgTUy5cuV07tw55c6dWwEBAVb/PD095enpKR8fH8sd/FJc0Dpy5EiS6yxVqpRiY2O1e/fuROfHB9GYmJgk11GsWDFFR0frt99+s0wLDQ3VmTNnku1CkJiyZcuqf//+Wr16tUqWLKmVK1cmWbdRo0aKiIjQli1btHLlSrVq1cpqvouLi2rXrq0PP/xQmzZt0uXLl7Vz584UtScp8X9E3LhxwzLt2LFjVnXKlSunkydPJniuAgICrK7AP8nf31/VqlXTokWLtGjRIgUGBqpAgQKS4kJtgQIFNHDgQFWoUEFFihTRlStXntrWx2/Ii4mJ0YkTJyyPy5Ytq5iYGN2+fTtBO+NvnJPigni3bt00Y8YMvfnmm5o/f74NRwr2ilCbgSZODFPWrHElAGQEX99YffLJ3QzpepAarVu3Vs6cOdWzZ0/9+uuvunjxonbt2qWPPvrI0sfztdde0+TJk7V+/XqdPn1aw4YN0927SV91LliwoNq1a6chQ4Zo/fr1lnWuXr1aklSgQAGZTCZt2rRJt2/f1oMHDxKso0iRImrUqJGGDh2qvXv36tixYxowYIB8fX3VqFEjm/bt4sWLGjNmjPbv36/Lly9r27ZtOnfunIoVK5bkMu7u7nr55Zf1xRdf6NSpU2rZsqVl3s8//6yZM2fq6NGjunz5spYsWaLY2FgVLVrUpvY8jb+/v/Lnz68vv/xSZ8+e1aZNmzRt2jSrOn379tX+/fv1wQcf6OjRozp79qw2bNiQ7I1i8V599VWtW7dO69evV8eOHS3T40PsqlWrdP78ec2cOTPRIcAeV7NmTW3evFmbNm3S6dOn9f7771u9JooWLarWrVtr4MCBWrt2rS5evKiDBw/qm2++0aZNmyRJH3/8sbZu3aqLFy/q999/186dO5N9bmD/CLUZqFixGJUsGVcCAOL6pC5fvlx+fn56/fXXVadOHb3zzjuKiIhQ9uzZJUlvvvmm2rRpo0GDBql58+by8PDQyy+/nOx6x4wZoyZNmmjYsGEKCgrSu+++q/DwcElSvnz5NGTIEI0ZM0YVKlRIMpBNmDBB5cqVU/fu3dW8eXMZhqG5c+cm6HKQ3L6dPn1ab7zxhgIDAzV06FD16NFDXbt2TXa5Vq1a6fjx46patarVV/o5cuTQunXr1KFDBwUFBWnu3LmaMmWKSpYsaVN7nsZsNmvq1Kk6c+aMGjRooKlTpyb4MYrSpUtr2bJlOnv2rFq3bq1GjRrpiy++SLKf8OMaN24sV1dXZc2a1er5a9iwoXr37q0PPvhADRs21P79+586GkPHjh3Vrl07DRw4UG3atFGhQoVUo0YNqzoTJkxQ27Zt9cknn6h27dp67bXXdPjwYcsxjY2N1QcffKA6deqoc+fOKlKkiD777DMbjxbskclIqqf9P8jNmzeTHQ8vrTRokEfHj5tVunSUfv75ZrpvD4Bjunv3rjw9PTO7GQCQJLPZnObZKalzn9lsVp48eZ66PFdqM1C/fveUNWtcCQAAgLTjsKMfxMbGavHixdqxY4fCwsLk7e2toKAgtWnTJsmBrTPblCnZ9ehRXMnoBwAAAGnHYUPtypUr9fPPP6tfv34qUKCAzp49q6lTp8rd3d0yQLe9mTgxTEOH5tHnn4dldlMAAAD+Vhw21J48eVJVqlSxDIydN29eBQcHWw2KbW/KlYvWoUNSSEi06MkMAACQdhw21JYoUUKbN2/W1atXlT9/fp0/f14nTpxQt27dklwmKirKqlOzyWSy/BJLRnRZ+Oord40bJ733nrsGDnyY7tsDAABwJM+Sxxx29IPY2FgtWLBAq1evlpOTk2JjY9WxY8cEA1U/bvHixVq6dKnlcUBAgOXnBzPC48+TYx51ABnh7NmzluGsAOCf4t69eypSpEiql3fYK7W7d+9WcHCwBgwYoIIFC+r8+fOaPXu2cubMqTp16iS6TKtWrdS0aVPL4/i/Bm7evJngpwDTQ9OmnlqzxkNNmz5QSEj6/1wlAMcUGRmZIcMMAkBqpceQXpGRkQoJCUkw3cXFxaYhvRw21M6bN08tWrRQzZo1JUmFChXSzZs3tXLlyiRDrdlsTnLQ7Iy4YL1hg5ulNIw76b49AAAAR/Isecxhx6mNiIiQk5N1852cnDIknKaWh0esVQkAAIC04bChtnLlylq+fLl+++033bhxQ3v37tWaNWv0wgsvZHbTknTvnpNVCQB4Nrt27ZKfn5/u3En9t19ffvmlGjRokIatss2gQYPUq1evZ1rHokWLVKpUqWTrZNb+ARnNYdNVr169VK1aNc2YMUODBw/W3Llz1aBBA3Xs2DGzm5aknDljrUoA+DtJKqSlRfBMjfjtJvdv165dGdomR/HDDz+oVatWKl26tEqXLq0OHTro4MGDNi+/b98+FSpUKNEwPXv2bFWtWlVFihRR06ZNrdZ76dKlJJ+rH3/80VLvo48+0ssvv6yAgIBEt/Hll18muo5ixYpZ6pw4cUK9e/dW1apV5efnp+nTpye7T5MnT5afn58+/vhjq+k3btxQ//79VbFiRRUrVkyNGjXSTz/9ZJmf3Ovw0KFDlnpbt25V06ZNVaJECZUrV069e/fWpUuXrLa1a9cuNWrUSAEBAapZs6YWLVpkNT8mJkaff/65qlWrpqJFi6pGjRqaOHGi5VvsqKgojR49WvXq1VOxYsX0/PPPa8CAAbp27ZrVczBkyBCrdYwfP16RkZFWxzdv3rzJHt9FixYlmP8sN4HZwmH71Lq5ualHjx7q0aNHZjfFZq6u1iUAwDaRkZFyTeHJs0qVKlaB6eOPP9b9+/c1YcIEyzQvLy/t3r07Q9rjSHbv3q0WLVqoSpUqypo1q6ZMmaJOnTrpl19+Ub58+ZJd9s6dOxo4cKBq1aqlmzdvWs1btWqVRo4cqbFjx6pSpUqaMWOGOnfurO3btyt37tzKnz9/gvD8ww8/6Ntvv1XdunWtpnfs2FG//fab/vjjjwRt6NOnj7p27Wo1rUOHDqpQoYLlcXh4uAoVKqSmTZtqxIgRye7ToUOHNG/evESvig8cOFB3797VrFmz5O3trRUrVqhPnz5at26dypYtm+B1KElffPGFgoODLe25ePGievXqpd69e2vy5Mm6d++eRowYoddff10bNmyw1OnWrZu6du2qyZMnKzg4WO+++658fHws9xJNmTJF33//vSZNmqSSJUvq8OHDevvtt+Xp6anXXntN4eHh+v333zVw4ECVLl1ad+7c0fDhw9WzZ0+tW7dOknT69GnFxsZq3Lhx8vf314kTJ/Tuu+/q4cOHlkDfp08f9ezZ0+om+yePryRlz55d27dvtzxO7+FTHTbUOqIHD6xLAEgJ08Okx7c2nJykrFltq2sySf8dozupuoa7eypb+XShoaH68MMP9euvvyosLEz+/v7q37+/WrZsaanTtm1blSxZUs7Ozlq+fLmee+45LV26VJs3b9bw4cMVEhKiSpUqqV27dklux9XVVXnz5rU8zpo1qyIjI62mPW7p0qX64osvdOfOHb300kv64osvlC1btmTb8+eff2rUqFH69ddf5e7urtq1a2vkyJHy9vaWJK1Zs0YTJ07U+fPnlTVrVpUtW1azZs2S+2PH97vvvtO0adMUGRmpFi1aaOTIkZabmsPCwvTxxx9r06ZNioiIUPXq1fXJJ58ke8Vr8uTJmj59usLDw9WsWTPlypXr6U9KIut43Pjx47V27VoFBwcne8wl6d///rdatmwpZ2dnrV+/3mre9OnT1alTJ3Xo0EGSNHbsWG3evFkLFy7UW2+9JWdn5wTPz7p169SsWTN5eHhYpn366aeSpNu3bycaaj08PKzqHzt2TCdPntTYsWMt0ypWrKiKFStKkj777LMk9+fBgwd666239Pnnn+vrr79OMH///v0aM2aMKlWqJCnuG4vp06fryJEjKlu2bILXYVRUlDZs2KCePXtaQt6RI0cUExOj9957z3K/0JtvvqlevXopKipKZrNZc+fOVaFChTR8+HBJUvHixbV3715Nnz7dEmr379+vRo0aqX79+pKkggULatWqVZYrwp6enlq4cKFV+0eNGqUmTZroypUr8vPz00svvaSXXnrJMr9w4cI6c+aMvv/+e0uo9fDwkJeXl2X0g8SOrxQXYpN6v6UHh+1+4Iju3HG2KgEgJfIVL57kP+/eva3q+pQvn2TdXE9cwcpbtWqCOukpIiJC5cuX15w5c/TLL7+oc+fOGjBgQIKrWUuWLJGrq6tWrlypsWPH6sqVK+rdu7caNGigDRs2qFOnThozZkyatOnChQvasGGD5syZozlz5mjPnj0Jgt2T7blz547at2+vMmXKaN26dfrhhx9069Ytvfnmm5Kk69evq1+/furQoYO2bt2qpUuX6pVXXrG6oXnXrl06f/68lixZokmTJmnx4sVavHixZf7gwYN15MgRzZo1S6tXr5ZhGOratWuSQymtXr1aEyZM0L///W+tXbtWefPm1Zw5c6zqxH8d/uRX28kJDw9XdHS0vLy8kq23aNEiXbx4UW+//XaCeZGRkTpy5IgCAwMt05ycnFSrVi0dOHAg0fUdOXJEx44de+auhQsWLFCRIkVUtWrVFC87bNgw1atXT7Vr1050fpUqVbR69Wr99ddfio2N1apVqyx/gCRm48aN+uuvvyzBXpLKly8vJycnLVq0SDExMbp7966WLVumwMBAyx84Bw4cUK1atazWVadOHatjV6VKFQUHB+vMmTOS4sLm3r17rULqk+7evSuTySRPT89k6yT33Cd1fB88eKAXX3xRVapUUc+ePXXixIkk15EWuFKbgXLkiNGdOy7KkSMms5sCAOli06ZNKv5EKI6Ntb6PIF++fOrTp4/lca9evbR161b9+OOPlqtdUtwP5Hz44YeWx2PGjFHhwoUtV6qKFSumP//8U1OmTHnmdsfGxmrixImWK7Nt2rRRcHCwVZ0n2zNp0iSVLVtW77//vmXal19+qRdeeEFnzpzRw4cPFR0drcaNG6tAgQKSlODr6xw5cmj06NFydnZWsWLFVK9ePQUHB6tz5846e/asNm7cqJUrV1pugv7mm2/0wgsvaP369WrWrFmC/ZgxY4Y6duyoV199VZL03nvvaceOHYqIiLDUcXNzU9GiReXiYnsEGD16tHx8fKwC6ZPOnj2rzz77TMuXL0903aGhoYqJiVHu3LmtpufJk8cSwp60YMECFS9e/JluAn/06JFWrFihfv36pXjZVatW6ejRo1Z9ZJ/03Xff6V//+pfKli0rFxcXubm5aebMmQoICEi0/sKFC1WnTh3lz5/fMq1QoUKaP3+++vTpo/fee08xMTGqXLmy5s6da6lz48aNBGO15s6dW/fu3VN4eLjc3Nz01ltv6f79+woKCpKzs7Pl6m/r1q0TbcujR4/02WefqWXLlkn+4Mu5c+c0a9YsffTRR0muI7HjW7RoUX355ZcqVaqU7t27p++++04tWrTQL7/8YrXvaYlQm4E8PKQ7d+JKAEipkFOnkpxnPDHE4fUjR5Ku+0S/thu//vpsDXtMjRo1Elw9PXjwoPr37295HBMTo6+//lpr1qzRtWvXFBkZqcjISMvPlscrX7681ePTp09bhV4pbiSctFCwYEFLoJWkvHnz6vbt28m25/jx49q1a1eCEC/FXfkNCgpSrVq1VK9ePQUFBSkoKEhNmjSxuuJVokQJOTv/79s7Hx8fy9fpp0+flouLi55//nnLfG9vbxUtWlSnT59OdD9Onz6doC9p5cqVrW6Iq1SpklU/x6eZPHmyVq9erSVLlijrY11cHhcTE6O33npLQ4YMUdGiRW1ed3LCw8O1cuVKDRw48JnWs379et2/f/+p3SaedOXKFX388cdasGBBkvstxfWPvXv3rhYuXChvb29t2LBBffr00fLlyxP8EXP16lVt3bpV3333ndX0Gzdu6N1331W7du3UsmVL3b9/X+PHj9cbb7yhhQsX2twX9ccff9Ty5cs1ZcoUlShRQseOHdPw4cPl4+Oj9u3bW9WNiopSnz59ZBhGkt94hISEqEuXLmratKk6d+6caJ2kjm+VKlVUpUoVq8d16tTRvHnzNHToUJv2J6UItRnKeKIEANulpJ9retV9Gnd39wRXqJ78haBvv/1WM2fO1MiRI/Xcc8/J3d1dw4cPT/CV+pMhNz09eWXRZDIluML8ZHsePnyoBg0aaNiwYQnW5+PjI2dnZy1cuFD79+/Xtm3bNGvWLI0bN05r1qxRoUKFJCnRHwSyp/HWv/vuO02ZMkULFy5U6dKlk6x3//59HT58WEePHrVczY6NjZVhGJarkC+++KKcnZ1169Ytq2Vv3ryZ6K9F/fTTTwoPD09xGH3S/PnzVb9+fZt+kepxv//+u27duqWXX37ZMi0mJkZ79uzR7Nmzde7cOV26dEmzZs3SL7/8opIlS0qSypQpo19//VWzZ8/WuHHjrNa5aNEi5cyZUw0bNrSaPnv2bHl6elp9E/D111/rhRde0G+//abKlSsrb968CW68u3XrlrJnz255bX766ad666231KJFC0lx3wxcvnxZkydPtgq18YH28uXLWrx4caJXaa9du6Z27dqpcuXK+vzzz5M8TrYeX7PZrDJlyuj8+fPJ1nsW9KnNQH/9ZbIqAeCfaN++fWrUqJHatGmjMmXKqHDhwjp79uxTlytWrJjVEEiS9Ntvv6VTK5+ubNmyOnHihAoWLKiAgACrf/E3gplMJr3wwgt65513tGHDBpnNZstd5k9TrFgxRUdHW+1jaGiozpw5k+jV4fhlnuybnNpjNHXqVE2aNEnz5s1LcFf7k7Jnz67Nmzdr48aNln9du3ZV0aJFtXHjRj3//PNydXVV+fLlrbp1xMbGKjg4ONEr7gsXLlSDBg1SdaNbvIsXL2rXrl2p6pNbq1atBPtUoUIFtWrVShs3bpSzs7PCw8MlKcGPQTk7Oyf448QwDC1evFht27ZN8MdMeHh4ouuQ/td9p3Llytq5c6dVne3bt1sdu/Dw8ARXdZ2dna3+QIsPtOfOndOiRYssNzU+LiQkRG3btlX58uU1ceLEBG2Ll5LjGxMToz///DNdbxwj1Gag8HBnqxIA/okCAgK0fft27du3T6dOndJ7772X4OpdYrp166Zz587p008/1enTp7VixQqrm6oyWo8ePRQWFqa+ffvq0KFDOn/+vLZu3arBgwcrJiZGv/32m77++msdPnxYV65c0dq1axUaGppkIH1SkSJF1KhRIw0dOlR79+7VsWPHNGDAAPn6+qpRo0aJLvPaa69p0aJFWrRokc6cOaPx48fr5MmTVnUOHjyo2rVrJ7iC/rgpU6boiy++0JdffqmCBQvqxo0bunHjhh48NnzPmDFjNGDAAElxoe65556z+pc7d25lyZLFcjVeknr37q358+dr8eLFOnXqlP79738rPDzc6qYpKa4f5549e9SpU6dE23fu3DkdPXpUN27c0KNHj3T06FEdPXrUaixVKS4Y+/j4JBgOTIq7cS1+uaioKF27dk1Hjx7VuXPnJEnZsmVLsE/u7u7KmTOnnnvuOUlxf0T4+/vrvffe08GDB3X+/Hl999132r59e4LnKDg4WBcvXkx0n+rVq6dDhw5p4sSJOnv2rH7//Xe9/fbbKlCggMqWLStJ6tq1qy5cuKBRo0bp9OnTmj17tn788Uf1fuwm0QYNGujrr7/Wpk2bdOnSJa1bt07/+c9/9Morr0iKC7RvvPGGDh8+rG+++UYxMTGW5zb+2MUHWj8/P3300Ue6ffu2pc6Tkju+EydO1LZt23ThwgX9/vvv6t+/v65cuZLkc5oW6H6QoWIUd8i5UQzAP9fAgQN18eJFde7cWW5uburcubMaNWqke/fuJbucn5+f/vOf/2jEiBGaNWuWKlasqH//+9+J3mmfEXx9fbVy5Up99tln6tSpkyIiIlSgQAHVqVNHTk5Oyp49u3799VfNmDFD9+/ftwzcn1gASMqECRP08ccfq3v37oqMjFS1atU0d+7cRLstSFKLFi0swSciIkKNGzdWt27dtHXrVkud8PBwnTlzxmqM0Sd9//33ioyM1BtvvGE1/e2339aQIUMkxY3ucPXqVZv3Jb59oaGhGj9+vG7evKkyZcpo3rx5Cb66XrhwofLly6egoKBE1/Puu+9ajS8cHyD37NmjggULSoq7wrl48WK1a9fOqt9yvOvXr1sFz++++07fffedqlevrqVLl9q0P/FDbY0ZM0Y9evTQgwcP5O/vr0mTJqlevXoJ9qlKlSpWP1AQr1atWpoyZYqmTp2qqVOnys3NTZUrV9YPP/xg6VpQqFAhff/99xoxYoRmzpypfPny6YsvvrAM5yXFDc/1+eefa9iwYbp9+7Z8fHzUpUsXDR48WFJcl4KNGzdKUoIuEEuWLFGNGjW0fft2nT9/XufPn7fqEyvF9TOO97TjGxYWpnfffVc3b95Ujhw5VK5cOa1atUolSpSw6dimhsmwp847meTmzZtJDo+Slvz88kkySTJ05UrSfyED+Ge7e/dussPrAEBmM5vNaZ6dkjr3mc1mm/pE0/0gA2XJEmtVAgAAIG0QajNQ1qyGVQkAAIC0QajNQPyiGAAAQPog1AIAAMDhEWoBAADg8Ai1AGCHnvw1KwD4O0uLcx6hFgDsjLu7u+7du0ewBfCPEBsbq3v37ll+pCO1+PEFALAzLi4u8vDw0P379zO7KQCQKFdX1wS/4PYsPDw85OLybLGUUAsAdsjFxYUfYABgl0wmk/Lly6eQkBDZ02940f0AAAAANrt2zUkDB8aV9sS+WgMAAAC7NmVKNn39dVxpT+h+AAAAAJv163dfHh4e6tHDvvr9c6UWAAAANvP1jdVXX8WV9oRQCwAAAIdHqAUAAIDDI9QCAADA4RFqAQAA4PAItQAAAHB4hFoAAAA4PEItAAAAHB6hFgAAAA6PUAsAAACHR6gFAACAwyPUAgAAwOERagEAAGCza9ecNHBgXGlP7Ks1AAAAsGtTpmTT11/HlfbEJbMbAAAAAMfRr999eXh4qEeP+5ndFCtcqQUAAIDNfH1j9dVXcaU9IdQCAADA4RFqAQAA4PAItQAAAHB4hFoAAAA4PEItAAAAHB6hFgAAAA6PUAsAAACHR6gFAACAwyPUAgAAwOERagEAAODwCLUAAABweIRaAAAAODxCLQAAABweoRYAAAAOj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAABsdu2akwYOjCvtiX21BgAAAHZtypRs+vrruNKeEGoBAABgs/btH6pChbjSnhBqAQAAYLPFi911+HBcaU9cMrsBzyI0NFTz5s3ToUOHFBERIV9fX/Xt21dFixbN7KYBAAD8LfXrd18eHh7q0eN+ZjfFisOG2vv37+ujjz5SmTJlNGzYMHl6eiokJEQeHh6Z3TQAAIC/LV/fWH31lRQSEivDyOzW/I/DhtpVq1YpV65c6tu3r2Va3rx5M7FFAAAAyCwOG2r379+vChUqaMKECTp+/Li8vb3VsGFD1a9fP8lloqKiFBUVZXlsMpnk5uZm+X9GyujtAQAApIX4DGNvWcZhQ+2NGzf0888/q0mTJmrVqpXOnDmjWbNmycXFRXXq1El0mRUrVmjp0qWWxwEBARo3bpzy5MmTQa2OZ1K+fPkyeJsAAABpx9fXN7ObYMVhQ21sbKyKFi2qTp06SYoLqBcvXtTPP/+cZKht1aqVmjZtankc/xfGzZs3FR0dne5tlnwlmSQZCgm5lgHbAwAASFsmk0m+vr66du2ajAzoVOvi4mLTBUiHDbU5c+ZUgQIFrKYVKFBAv/76a5LLmM1mmc3mROdlxJOSmdsDAABIS4Zh2FWecdhxakuWLKmrV69aTbt69WomdCUAAABAZnPYUNukSROdOnVKy5cv17Vr1xQcHKzNmzerUaNGmd00AAAAZDCTYU/XjVPowIEDmj9/vq5du6a8efOqSZMmyY5+kJSbN29ajYqQXvz88im+T+2VKyHpvj0AAIC0ZjLF3fAeEhKSId0PzGbz37tPrSRVrlxZlStXzuxmAAAAIJM5bPcDAAAAZLxr15w0cGBcaU/sqzUAAACwa1OmZNPXX8eV9sShux8AAAAgY/Xrd18eHh7q0eN+ZjfFCldqAQAA4PAItQAAALAZ3Q8AAADg8Oh+AAAAAIfn6xurr76KK+0JoRYAAAA2+/13F1WsGFfaE0ItAAAAbDZ4sJcOH44r7QmhFgAAADabODFMFSrElfbEvq4bAwAAwK6VKxetQ4ekkJBoGUZmt+Z/uFILAAAAm/EzuQAAAHB4X3wRN07tF1/Y1zi1hFoAAADYzDBMVqW9oE8tAAAAbNaz5wP98Ye7evZ8kNlNscKVWgAAANhs1iwPHT4cV9oTQi0AAABsFh5uXdoLQi0AAABs5uZmXdoL+tQCAADAZkOH3lPevO7q0eNeZjfFCldqAQAA4PAItQAAALAZ49QCAADA4T18aLIq7QWhFgAAADYLDzdZlfaCUAsAAACbnT/vYlXaC0ItAAAAbDZw4D1lzRpX2hNCLQAAAGw2ZUp2PXoUV9oTQi0AAABsNnFimCpUiCvtCaEWAAAADo9QCwAAAJsNHuylw4fjSntCqAUAAIDNPvzwrnLnjivtCaEWAAAANlu0yE23bsWV9oRQCwAAAJsdP262Ku0FoRYAAAA2mzw5bvSDyZPDMrspVgi1AAAAcHiEWgAAANiM0Q8AAADg8Bj9AAAAAA5v8+asunUrrrQnhFoAAADYrH37h6pQIa60J4RaAAAA2GzqVA8dPhxX2hNCLQAAAGx28KCrVWkvCLUAAABIAeOJ0j4QagEAAGCzUqWirUp7QagFAACAzXLmNKxKe+GS2Q0AAACA4xg69J7y5nVXjx73MrspVrhSCwAAAIdHqAUAAIDNpkzJpq+/jivtCd0PAAAAYLN+/e7Lw8NDPXrcz+ymWOFKLQAAAGzm6xurr76KK+0JoRYAAAA2+/13F1WsGFfaE0ItAAAAbDZ4sJcOH44r7QmhFgAAADb78MO7yp07rrQnhFoAAADYbPZsd926FVfaE0ItAAAAbLZ7t6tVaS8ItQAAALBZ9eqRVqW9sK/b1gAAAGDXhgy5r+vX3TRkCOPUAgAAwEHNmuWhw4fjSntCqAUAAIDNwsOtS3tBqAUAAIDNwsNNVqW9INQCAADAZqdOuViV9oJQCwAAAJuVKBFtVdoLQi0AAABsZhjWpb3424TalStXqn379po9e3ZmNwUAAOBv6+hRF6vSXvwtQu3p06f1888/q3DhwpndFAAAgL+1e/ecrEp7YV8ROxUePXqkb775Rm+++aaWL1+ebN2oqChFRUVZHptMJrm5uVn+n5EyensAAABp4d//vqePPvLSv/99z67yjMOH2hkzZqhSpUoqX778U0PtihUrtHTpUsvjgIAAjRs3Tnny5EnvZj7BpHz58mXwNgEAAJ7dDz9IsbHSDz94adgwr8xujoVDh9qdO3fq3LlzGjNmjE31W7VqpaZNm1oex/91cfPmTUVHZ8QdfL6STJIMhYRcy4DtAQAApK3ixb10/LibihcPV0hIWLpvz8XFxaYLkA4bam/duqXZs2frww8/lKurq03LmM1mmc3mROcZGXwLX0ZvDwAAIC106PBQO3e6qUOHh3aVZ0yGPbUmBfbu3avx48fLyel/nZRjY2NlMplkMpk0f/58q3nJuXnzplVf2/Ti55dP8Vdqr1wJSfftAQAApLWXXsqjkyfNKlEiSlu23Ez37ZnN5r/3ldpy5cpp/PjxVtO+/fZb5c+fXy1atLA50AIAAMB2vr7ROnnSLF9f+/rxBYcNtW5ubipUqJDVtCxZsih79uwJpgMAACBtXLvmYlXaCy5nAgAAwGYDB95T1qxxpT1x2D61aYk+tQAAALYJDMyjs2fNKlIkSjt22E+fWq7UAgAAwGbR0Sar0l4QagEAAGCzSpUirUp7YV89fAEAAGDXhg+/p8KF3dWjh331qeVKLQAAAGx286aTtm2LK+2JfbUGAAAAdq1/fy8dPhxX2hNCLQAAAGzm7x9tVdoLQi0AAABs9uiRdWkvCLUAAACw2ZEjrlalvSDUAgAAwGZTp4Ypd+640p4QagEAAODwCLUAAACw2b/+5aVbt+JKe0KoBQAAgM1Kl46yKu0FoRYAAAA2u37d2aq0F4RaAAAA2CxPnmir0l4QagEAAGAzhvQCAADA34DpidI+EGoBAABgs5w5Y6xKe0GoBQAAgM3c3GKtSntBqAUAAIDNzpxxtSrtBaEWAAAANsuRI8aqtBeEWgAAANisWrUoq9JeEGoBAABgM8OwLu0FoRYAAAA2++03s1VpLwi1AAAAsNmtW85Wpb0g1AIAAMDhudhacenSpanaQNu2bVO1HAAAAOyRobhfE7OvTrU2h9olS5akagOEWgAAgL8PFxcpOjqutCc2N2fRokVWj0NDQzVmzBgVLFhQTZo0Uf78+SVJV65c0dq1a3X58mX9+9//TtvWAgAAIFNFR5usSnuR6j61M2bMUL58+TRgwAAVLVpUbm5ucnNzU7FixTRgwAD5+Pho5syZadlWAAAAZDJXV8OqtBepDrXHjh1T2bJlk5xfrlw5HT16NLWrBwAAgB1ycjKsSnuR6lBrNpt18uTJJOefOHFCZrN9jV8GAACAZ+PmZliV9iLVXXxr1aqldevWyd3dXa+88op8fHwkSdevX9e6desUHBysV155Jc0aCgAAgMx3546zVWkvUh1qu3Tponv37mnDhg3asGGDnJziLvrGxsZKkmrWrKkuXbqkTSsBAABgF0ymuCG94kr7kepQ6+Liov79+6t58+b67bffdOvWLUlSnjx5VLFiRfn7+6dVGwEAAGAnunS5rzlzPNWly/3MbooVk2EY9hWzM8HNmzcVFRWV7tvx88un+MGKr1wJSfftAQAApLWCBX0VG+skJ6dYXbp0Ld23ZzablSdPnqfWe+Zhc0+ePKljx47pzp07atSokfLly6eIiAhduXJF+fPnV9asWZ91EwAAALATsbEmq9JepDrURkdHa9KkSdq3b59lWpUqVZQvXz6ZTCaNHj1aTZo0UevWrdOkoQAAAEBSUj2k18KFC3XgwAH17t1bkyZNsprn6uqqatWqWQVeAAAAOL769R9alfYi1aF2586datiwoerXr69s2bIlmO/n56cbN248U+MAAABgXzZvdrMq7UWqQ+3du3dVqFChpFfs5KSIiIjUrh4AAAB2yDBMVqW9SHWozZUrl65cuZLk/BMnTsjX1ze1qwcAAABslupQW6tWLW3atCnRn8rdtGmTdu/erdq1az9T4wAAAABbpHr0g9atW+vUqVMaPny4/Pz8JElz5szR/fv3FRoaqkqVKqlp06Zp1lAAAAAgKc/0i2LDhg3Tjh07tGfPHsXGxio6OlqFCxdWx44dVbt2bZlM9tXXAgAAAM8q7mdy40r7kapQGxkZqQULFqhMmTKqXbs23QwAAAD+IcqVi9Tvv2dVuXKRmd0UK6nqU+vq6qpNmzbpzp07ad0eAAAA2LEzZ1ytSnuR6hvFihQpokuXLqVlWwAAAGD3jCdK+5DqUNu9e3ft3LlTmzdvVkxMTFq2CQAAAHbK2zvGqrQXJsMwUhWz33nnHd27d09hYWEym83y9vaWq6v1ZWiTyaQvvvgiTRqanm5euqToRH4ownBykrJmtTw2PUz65+AMk0lyc0u2brHivpJMipVJZ66E/a9ueLiU1NNgMsl4fL0pqKvwcJmSeXoNd/fU1X30SKbY2LSp6+Ymxd9QGBEhUzJ/IKWobtasktN//2aLjJQpOjpt6mbJIjk7p7xuVJRMUVFJ13V1lVxcUl43OlqmyKT7NBlms2Q2p7xuTIxMyfx4iuHiIsW/31NSNzZWpkeP0qaus7OUJct/Hxhx7420qJuS930anyOSrMs5Iu4B54iU1+Uc8d8HnCNSUzep9318nnkod125EpJsXUubn+EcYTabladgwSTr/2/BVBo+fLgxYsSIp/5zCJUqGUbcU2z1L7xuXePKlSuWfzFubonWMyTjUfXqVnWjvb2TrLtXVazqRhUokGTdyBIlrOpGliiRZN2oAgWs6kZUqJBk3Whvb6u6j6pXT7JujJubVd3wunWTrGtIVnUfNmmSbN2rp05Z6j5o1y7ZuiFHjljq3u/ePdm61/bssdS916dPsnWv//KLpe7dt99Otu6Nn36y1A378MNk695cssRS96/Ro5Ote2vOHEvd0AkTkq17+7vvLHVvf/ddsnVDJ0yw1L01Z06ydf8aPdpS9+aSJcnWDfvwQ0vdGz/9lGzdu2+/bal7/Zdfkq17r08fS91re/YkW/d+9+6WuiFHjiRb90G7dpa6V0+dSrbuwyZNrF7DydVNr3NERIUKnCP++49zRNw/zhFx/zhHxP2zh3PEDeU2fH3DM+YcUamSTXEu1UN6jRgxIrWLAgAAwMHdumVfN4qluvvB3wndD1JYl68WU16XrxbjHvDVYurqco6Ie8A5IuV1OUf89wHniNTUTep9X61aHt267SL33Fl0+PCNZOta2pwB3Q+eKdRGRUVp8+bNOnjwoG7ciNupvHnzqlKlSqpbt26CPrb26ubNm4pK5gSRVvz88il+sGJLHxQAAAAHUqZMXoWFucjLK1rHjt1I9+2ZzWblyZPnqfVS3f3g9u3bGjVqlK5evSovLy/5+vpKks6fP69Dhw5p/fr1+uijj5QrV67UbgIAAAB2JizM2aq0F6kOtTNnztTNmzc1ePBgVatWzWre7t27NWXKFM2cOVNDhw595kYCAADAPphMhgzDJJPJvnqwpjrU/v7772rSpEmCQCtJ1atX17lz57Ru3bpnahwAAADsi2GYrEp7keofX3Bzc1OOHDmSnO/l5SW3xzsbAwAAwOG5u8dYlfYi1aG2Tp062rp1qyISuaPx0aNH2rJli+rWrftMjQMAAIC9MT1R2odUdz/w9/fXwYMHNWjQIAUFBVluFLt27Zq2bdumbNmyqVChQvr111+tlqtateqztRgAAACZxtXV0MOHcaU9SfWQXh06dEjVBhctWpSq5dITQ3oBAADYJqPzTLoP6TV8+PDULpomVqxYob179+rKlStydXVViRIl1KVLF+XPnz9T2wUAAPB3liWLoYgIk7Jksa8rtakOtaVLl05R/cjISO3evVthYWHy8vJK7WYtjh8/rkaNGqlo0aKKiYnRggULNGrUKE2YMEFZH/v1DgAAAKQdL69YXb/uJC+vpH8VLDOk+kaxlHr48KGmTp2qy5cvp8n6PvjgA9WpU0cFCxaUv7+/+vXrp1u3buns2bNpsn4AAAAk5OxsWJX2ItVXau3Nw//+RnK2bNmSrBMVFWXVd9ZkMlmGHTOZMvYOvozeHgAAQFp4+NBkKe0pz/wtQm1sbKxmz56tkiVLqlChQknWW7FihZYuXWp5HBAQoHHjxtnU+ThtmZQvX74M3iYAAMCzCwuLL13sKs/8LULtzJkzdenSJX3yySfJ1mvVqpWaNm1qeRz/18XNmzcVHR2drm2M46v4uwVDQq5lwPYAAADSWsbmGRcXl/Qd/cBezJw5U7/99ptGjhypXLlyJVvXbDbLbDYnOi+VI5ulWkZvDwAAIK3ZU57JsBvF0pphGJo5c6b27t2rjz/+WHnz5s3sJgEAACCTOGyonTlzpnbs2KGBAwfKzc1NYWFhCgsLU2RkZGY3DQAAABks1d0PTp06peLFi6dlW1Jk48aNkqQRI0ZYTe/bt6/q1KmT8Q0CAABApnmmn8n19fVVYGCgAgMD5ePjk2z96OhonTx5Uv7+/nJ3d09VY9MLP5MLAABgG3v9mdxUh9rg4GDt2LFDR44cUWxsrEqUKKHAwEDVqFEj2bFi7RGhFgAAwDZ/u1Ab7+7du9q1a5eCg4N16tQpubi4qEKFCqpdu7aqVKkiFxf7H2CBUAsAAGAbPz9fxd2WFasrV9J/SK8MC7WPu3btmoKDgxUcHKyQkBC5u7urWrVqCgoK0nPPPZdWm0lzhFoAAADb2OuV2jS9jOrq6qosWbJYxoI1mUzav3+/fvnlFxUpUkT9+vVTgQIF0nKTAAAAwLOH2vDwcO3Zs0fBwcE6fvy4TCaTKlasqLZt26py5cpycnLS3r179f3332vq1Kn67LPP0qLdAAAAgEWqQ+2+ffu0Y8cO/fbbb4qKilLRokXVvXt31axZU9mzZ7eqW61aNd2/f18zZ8585gYDAAAAT0p1qB0/frxy5cqlJk2aKCgoSPnz50+2vr+/vwIDA1O7OQAAANgBX99HunbNTb6+jzK7KVZSfaPYsWPHVKZMmbRuT6bgRjEAAADb2OuNYqn+mdy/S6AFAACA40t1qAUAAADsBaEWAAAANnN1jbEq7QWhFgAAADaLjHS2Ku0FoRYAAAAOj1ALAAAAh0eoBQAAQAoYT5T2gVALAACAFDA9UdoHQi0AAAAcHqEWAAAADo9QCwAAAJu5usZalfaCUAsAAACbRUY6WZX2wr5aAwAAAKQCoRYAAAAOj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDNXF1jrUp7QagFAACAzSIjnaxKe2FfrQEAAABSgVALAAAAh0eoBQAAgMMj1AIAACAFjCdK+0CoBQAAgM1MJuvSXhBqAQAAkAKmJ0r7QKgFAACAzQzDurQXhFoAAAA4PEItAAAAUoAbxWCnLwIAAADb0acWdvoiAAAAcHSEWgAAADg8Qi0AAAAcHqE2Q9GnFgAAID0QajMUfWoBAADSA6E2A2XJEmtVAgAAIG0QajNQRISTVQkAAIC0QboCAACAwyPUAgAAwOERagEAAODwCLUAAABweITaDMU4tQAAAOmBUJuhGKcWAAAgPRBqAQAA4PAItQAAAHB4hFoAAAA4PEItAAAAHB6hNgOZTLFWJQAAANKGS2Y34FmtX79eP/74o8LCwlS4cGH16tVLxYoVy+xmJcownKxKAAAApA2HTle7du3S999/r7Zt22rcuHEqXLiwRo8erTt37mR20wAAAJCBHDrUrlmzRvXq1dNLL72kAgUKqHfv3nJ1ddWWLVsyu2kAAAB/S87OsValvXDY7gfR0dE6e/asWrZsaZnm5OSkcuXK6eTJk4kuExUVpaioKMtjk8kkNzc3y//Tm5NTjGJjXeTkFJMh2wMAAEhrMTFOltKe8ozDhtq7d+8qNjZWXl5eVtO9vLx09erVRJdZsWKFli5dankcEBCgcePGKU+ePOnZVIvY2PjSRfny5cuQbQIAAKQPk13lGYcNtanRqlUrNW3a1PI4/q+LmzdvKjo6OgNakEdxhzxaISE3M2B7AAAAac1XkkmSoZCQa+m+NRcXF5suQDpsqPX09JSTk5PCwsKspoeFhSW4ehvPbDbLbDYnOs8wjDRuYWKcLWXGbA8AACD92FOecdgbxVxcXFSkSBEdPXrUMi02NlZHjx5ViRIlMrFlAAAAyGgOe6VWkpo2baopU6aoSJEiKlasmNauXauIiAjVqVMns5sGAACADOTQobZGjRq6e/euFi9erLCwMPn7+2vYsGFJdj8AAADA35NDh1pJevnll/Xyyy9ndjMAAACQiRy2Ty0AAAAQj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4RFqAQAA4PAItQAAAHB4hFoAAAA4PEItAAAAHB6hFgAAAA6PUAsAAACHR6gFAACAwyPUAgAAwOERagEAAODwCLUAAABweIRaAAAAODxCLQAAABweoTZDGU+UAAAASAuE2gxleqIEAABAWiDUZiiu1AIAAKQHQm2G4kotAABAeiDUAgAAwOERagEAAODwCLUAAABweIRaAAAAODxCLQAAABweoRYAAAAOj1CbgTw8YqxKAAAApA1CbQYKD3e2KgEAAJA2CLUZKDbWugQAAEDaINRmIDe3WKsSAAAAaYNQm4HCw52sSgAAAKQN0lUG8vKKsSoBAACQNgi1Gchsti4BAACQNgi1GSg01MmqBAAAQNogXWUgX99YqxIAAABpg1Cbgbp0uW9VAgAAIG0QajPQxImeViUAAADSBqE2A1Wt+siqBAAAQNog1GagHTvcrEoAAACkDUItAAAAHB6hFgAAAA6PUAsAAACHR6jNQCVKPLIqAQAAkDYItRno5MmsViUAAADSBqE2A7m4xFiVAAAASBuE2gwUHe1sVQIAACBtEGozkI/PI6sSAAAAaYNQm4GuX89qVQIAACBtEGoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADs8lsxuQGjdu3NCyZct09OhRhYWFydvbW4GBgWrdurVcXBxylwAAAPAMHDIBXr16VYZh6I033pCvr68uXbqkadOm6dGjR+rWrVtmNy9Jzs4xiolxkbMzvygGAACQlhwy1FasWFEVK1a0PPbx8dHVq1e1ceNGuw61MTHOViUAAADShkOG2sQ8fPhQ2bJlS7ZOVFSUoqKiLI9NJpPc3Nws/89IGb09AACAtGZPeeZvEWqvXbumdevWqWvXrsnWW7FihZYuXWp5HBAQoHHjxilPnjzp3cQnmJQvX74M3iYAAEBasq88YzIMw8jsRsT74YcftGrVqmTrTJw4UX5+fpbHoaGhGj58uMqUKaM+ffoku2xSV2pv3ryp6OjoZ2u8DfLn95VkkmTo6tVr6b49AACAtJbRecbFxcWmC5B2daW2WbNmqlOnTrJ1fHx8LP8PDQ3VyJEjVbJkSb3xxhtPXb/ZbJbZbE50XkZk+9y5o3Xrllm5c0dnyPYAAADSkz3lGbsKtZ6envL09LSpbnygDQgIUN++feXkZP9D7hYoEBdqCxRI/6vCAAAA/yR2FWptFRoaqhEjRihPnjzq1q2b7t69a5nn5eWVeQ17ikOHslqVAAAASBsOGWqPHDmia9eu6dq1awn60S5evDiTWgUAAIDM4pChtk6dOk/tewsAAIB/DvvviAoAAAA8BaEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4RFqAQAA4PAItQAAAHB4hFoAAAA4PEItAAAAHB6hFgAAAA6PUAsAAACHR6gFAACAwyPUAgAAwOERagEAAODwCLUAAABweIRaAAAAODxCLQAAABweoRYAAAAOj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4RFqAQAA4PAItQAAAHB4hNoM5OERbVUCAAAgbRBqM1BkpLNVCQAAgLRBqM1AkyaFKWvWuBIAAABph1CbgVq1ilB4eFwJAACAtEOoBQAAgMMj1AIAAMBm2bNHW5X2glALAAAAm3l6mqxKe0GoBQAAgM1cXQ2r0l4QagEAAGCz0aPvKnfuuNKeEGoBAABgs0WL3HXrVlxpTwi1AAAAsNkff7hYlfaCUAsAAACbffNNmCpUiCvtiX1FbAAAANi1cuWideiQFBISLcOO7hXjSi0AAABsdu2akwYOjCvtiX21BgAAAHZtypRs+vrruNKe0P0AAAAANuvX7748PDzUo8f9zG6KFa7UAgAAwGa+vrH66qu40p4QagEAAODwCLUAAABweIRaAAAAODxCLQAAABweoRYAAAAOj1ALAAAAh0eoBQAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADs/hQ21UVJTeffddtW/fXufPn8/s5gAAACATOHyonTdvnry9vTO7GQAAAMhEDh1qDx48qCNHjqhr166Z3RQAAABkIpfMbkBqhYWFadq0aXr33Xfl6upq0zJRUVGKioqyPDaZTHJzc7P8P73FbyMjtgUAAJAe7DXPOGSoNQxDU6dOVYMGDVS0aFHduHHDpuVWrFihpUuXWh6XKFFCo0aNUp48edKrqYny9fXN0O0BAACkNXvLM3YVan/44QetWrUq2ToTJ07U4cOHFR4erlatWqVo/a1atVLTpk0tj52cHLr3BQAAAP7LZBiGkdmNiHf37l3du3cv2To+Pj6aMGGCDhw4YHXZOzY2Vk5OTqpVq5beeuut9G5qqoSHh2vEiBEaMWKEpdsDAACAI7HXPGNXV2o9PT3l6en51Hq9evVSx44dLY//+usvjR49WoMGDVLx4sXTs4nPxDAMnTt3Tnb0dwQAAECK2GuesatQa6vcuXNbPc6aNaukuL4duXLlyowmAQAAIBPRqRQAAAAOzyGv1D4pb968Wrx4cWY346nMZrPatm0rs9mc2U0BAABIFXvNM3Z1oxgAAACQGnQ/AAAAgMMj1AIAAMDhEWoBAADg8Ai1AAAAcHiEWgAAADg8Qi0AAAAc3t9inNrMNmLECBUsWFCStH37drm4uKhBgwbq0KGDTCaT7t+/r9mzZ+vAgQOKiopS6dKl1bNnT+XLl0+SdPPmTc2cOVMnTpxQdHS08uTJoy5duuj555/PzN0CAAD/ECNGjFChQoXk6uqqzZs3W7JM+/btJUkPHjzQ3LlztW/fPkVHR6tIkSLq3r27/P39LetYtmyZ1q1bp8jISNWoUUPZs2fXoUOH9MUXX2TIPnClNo1s27ZNzs7OGjNmjHr06KGffvpJmzdvliRNnTpVZ86c0dChQzVq1CgZhqExY8YoOjpakjRz5kxFR0dr5MiRGj9+vDp37mz56V8AAICMsG3bNmXJkkWfffaZunTpomXLlunIkSOSpAkTJujOnTsaNmyYxo4dq4CAAH366ae6f/++JGnHjh1avny5OnfurLFjxyp37tzauHFjhrafK7VpJFeuXOrevbtMJpPy58+vixcv6qefflKZMmW0f/9+ffrppypZsqQkacCAAfrXv/6lffv2qXr16rp165aqVq2qQoUKSZJ8fHwyc1cAAMA/UOHChdWuXTtJUr58+bR+/Xr9/vvvcnV11enTpzVjxgzLr4h169ZN+/bt0549e1S/fn2tX79edevW1UsvvSRJatu2rQ4fPqxHjx5lWPsJtWmkePHiMplMlsclSpTQmjVrdPnyZTk7O6t48eKWedmzZ1f+/Pl15coVSdIrr7yiGTNm6MiRIypXrpyqVq2qwoULZ/g+AACAf674i2vxcubMqTt37uj8+fN69OiRevXqZTU/MjJS165dkyRdvXpVDRs2tJpfrFgxHT16NH0b/RhCrR2oV6+eKlSooN9++01HjhzRihUr1K1bN73yyiuZ3TQAAPAP4eKSMBYahqFHjx4pZ86cGjFiRIL57u7uGdAy29CnNo2cPn3a6vGpU6fk6+urAgUKKCYmRqdOnbLMu3fvnq5evaoCBQpYpuXOnVsNGzbUO++8o2bNmln64wIAAGSmIkWKKCwsTE5OTvL19bX65+npKUnKnz+/zpw5Y7Xck4/TG6E2jdy6dUtz5szR1atXFRwcrHXr1qlx48bKly+fqlSpomnTpunPP//U+fPn9c0338jb21tVqlSRJM2ePVuHDh3SjRs3dPbsWR07dkx+fn6ZvEcAAABSuXLlVKJECX3xxRc6fPiwbty4oRMnTmjBggWW4Pryyy/rl19+0datWxUSEqJly5bpwoULVl0z0xvdD9JI7dq1FRkZqffff19OTk5q3Lix6tevL0nq27evZs+erbFjxyo6OlqlSpXS+++/b7nMHxsbq5kzZyo0NFRubm6qWLGiunfvnpm7AwAAIEkymUx6//33tWDBAk2dOlV3796Vl5eXSpUqpRw5ckiSAgMDdf36dc2dO1dRUVGqXr266tSpk+Cb7HRtp2EYRoZt7W9qxIgR8vf3V48ePTK7KQAAAHbh008/lZeXl/r3758h26P7AQAAAJ5JRESE1qxZo0uXLunKlStavHixfv/9dwUFBWVYG+h+AAAAgGdiMpl08OBBLV++XFFRUcqfP7+GDBmi8uXLZ1wb6H4AAAAAR0f3AwAAADg8Qi0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4RFqAcDBhIaGavHixTp//nxmNwUA7AahFgAczF9//aWlS5cSagHgMYRaAPibi4iIyOwmAEC6I9QCQAY5evSo2rdvr7179yaYFxwcrPbt2+vkyZPJruPYsWN6//33JUlTp05V+/bt1b59e23dulWSNGLECA0ZMkRnz57V8OHD1aVLFy1YsECS1L59ey1evDjBOvv166cpU6ZYTXvw4IFmz56tf/3rX+rUqZP69++vlStXKjY2NjW7DgDpziWzGwAA/xRlypRRrly5tGPHDr344otW83bs2CEfHx+VKFEi2XX4+flZwmn9+vX13HPPSZJKlixpqXPv3j199tlnqlGjhgIDA5UjR44UtTMiIkIjRoxQaGio6tevr9y5c+vEiRNasGCBwsLC1KNHjxStDwAyAqEWADKIyWRSYGCgfvrpJz18+FDu7u6SpLt37+rIkSNq1arVU9fh5eWlSpUqafHixSpRooRq166doE5YWJh69+6tBg0apKqda9as0bVr1/T5558rX758kqQGDRrI29tbq1evVtOmTZU7d+5UrRsA0gvdDwAgAwUFBSkqKkp79uyxTNu1a5diYmISDaipYTab9dJLL6V6+T179qhUqVLy8PDQ3bt3Lf/KlSun2NhY/fHHH2nSTgBIS1ypBYAM5Ofnp6JFi2rHjh2qW7eupLiuB8WLF5evr2+abMPb21suLqk/vYeEhOjChQt6/fXXE51/586dVK8bANILoRYAMlhQUJBmzZql27dvKyoqSqdOnVKvXr3SbP2urq4pqv/kzV+GYah8+fJq3rx5ovXz58+f6rYBQHoh1AJABqtRo4bmzJmjnTt3KjIyUs7OzqpRo4bNy5tMplRt18PDQw8fPrSaFh0drb/++stqmo+Pjx49eqTy5cunajsAkBnoUwsAGczT01OVKlXSjh07tGPHDlWsWFGenp42L58lSxZJccNupYSvr2+C/rCbNm1KcKW2evXqOnnypA4dOpRgHQ8ePFBMTEyKtgsAGYErtQCQCWrXrq0JEyZIkjp06JCiZX18fOTh4aGff/5Zbm5uypIli4oXL668efMmu1zdunU1ffp0jR8/XuXLl9eFCxd0+PBhZc+e3ape8+bNtX//fo0bN05BQUEqUqSIIiIidPHiRe3Zs0dTpkxJUQgHgIzAlVoAyARVqlSRh4eH3N3dVaVKlRQt6+Lion79+snJyUnTp0/XV199pePHjz91uXr16qlFixb6448/NHfuXN24cUMffvihsmbNalUvS5YsGjlypJo1a6bjx49r9uzZWrlypa5du6b27dtbhiIDAHtiMgzDyOxGAMA/TUxMjN58801VrlxZ//rXvzK7OQDg8LhSCwCZYN++fbp7966CgoIyuykA8LdAn1oAyECnTp3ShQsXtGzZMgUEBKh06dKWedHR0bp//36yy7u7u6d4yC4A+Ccg1AJABtq4caN27Nghf39/9e3b12reiRMnNHLkyGSX79u3r+rUqZOOLQQAx0SfWgCwE/fv39fZs2eTrVOwYEHlzJkzg1oEAI6DUAsAAACHx41iAAAAcHiEWgAAADg8Qi0AAAAcHqEWAAAADo9QCwAAAIdHqAUAAIDDI9QCAADA4f0/sLo5qDJdvU4AAAAASUVORK5CYII=", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "data = np.load('./data_seal.npz')\n", + "data1 = data['pos_pred']\n", + "data2 = data['neg_pred']\n", + "\n", + "y_pred = np.concatenate([data1, data2])\n", + "y_true = np.array(['pos'] * len(data1) + ['neg'] * len(data2))\n", + "hard_thres = (max(y_pred)+min(y_pred))/2\n", + "plt.figure(figsize=(8, 6))\n", + "plt.scatter(y_true, y_pred, color='blue', s=1, label='Predictions vs True Values')\n", + "plt.axhline(y=hard_thres, color='red', linestyle='--', label=f'Hard Threshold: {hard_thres.item()}')\n", + "plt.xlabel('y_true')\n", + "plt.ylabel('y_pred')\n", + "plt.title('Scatter Plot of Predictions vs True Values with Hard Threshold')\n", + "plt.legend()\n", + "plt.grid(True)\n", + "plt.show" + ] + }, + { + "cell_type": "code", + "execution_count": 98, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "data = np.load('./data_seal.npz')\n", + "data1 = data['pos_pred']\n", + "data2 = data['neg_pred']\n", + "\n", + "y_pred = np.concatenate([data1, data2])\n", + "y_true = np.array(['pos'] * len(data1) + ['neg'] * len(data2))\n", + "hard_thres = (max(y_pred)+min(y_pred))/2\n", + "\n", + "# 创建主图和左右轴\n", + "fig, ax = plt.subplots(figsize=(10, 8))\n", + "\n", + "ax_hist_top = ax.twiny()\n", + "ax_hist_bottom = ax.twiny()\n", + "# 绘制散点图\n", + "ax.scatter(y_true, y_pred, color='blue', s=1, label='Predictions vs True Values')\n", + "\n", + "# 绘制阈值线\n", + "ax.axhline(y=hard_thres, color='red', linestyle='--', label=f'Hard Threshold: {hard_thres.item()}')\n", + "\n", + "# 设置标签和标题\n", + "ax.set_xlabel('y_true')\n", + "ax.set_ylabel('y_pred')\n", + "ax.set_title('Scatter Plot of Predictions vs True Values with Hard Threshold')\n", + "ax.legend()\n", + "\n", + "# 绘制顶部和底部的直方图\n", + "# 调整直方图范围和高度\n", + "hist_range = (min(y_pred), max(y_pred)) # 范围\n", + "hist_bins = 30 # 柱数\n", + "density = True # 归一化\n", + "\n", + "ax_hist_top.hist(data1, bins=hist_bins, range=hist_range, color='#1f77b4', alpha=0.7, density=density)\n", + "ax_hist_bottom.hist(data2, bins=hist_bins, range=hist_range, color='#ff7f0e', alpha=0.7, density=density)\n", + "\n", + "# 调整布局\n", + "plt.tight_layout()\n", + "plt.show()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 99, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA90AAAMWCAYAAADs4eXxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAACU8ElEQVR4nOzdeZyN9fvH8fc5sy8GYxm7sUT2texGVApl7IUKpZQUKX1bUUSLLCF95YuSPVsLypo1iYiUPfveYBiznfv3xzTn5zaLccwx5z5ez8fD43Lu+3Pu+7rP+Zz7zHXuz33fNsMwDAEAAAAAgGxnz+kEAAAAAADwVhTdAAAAAAC4CUU3AAAAAABuQtENAAAAAICbUHQDAAAAAOAmFN0AAAAAALgJRTcAAAAAAG5C0Q0AAAAAgJtQdAMAAAAA4CYU3QAs7eDBg7LZbOrWrVtOp3JDfvjhB9WvX1958uSRzWZTdHR0TqfkslWrVslms2nQoEGm6U2aNJHNZnPbeqdMmSKbzaYpU6a4bR24/bi733o6m82mJk2a3NBzunXrJpvNpoMHD7olJ0/l7r4yaNAg2Ww2rVq1ym3ryC634rvYlXXwPQFPQdENuElycrImTpyoqKgohYeHy8/PTwULFlTVqlX11FNPadGiRbc0n8z+kPK0wtVms5n++fj4KH/+/GratKmmT5/ulnXeyi/mgwcPqnXr1jpw4IB69OihgQMH6pFHHsn0OamF7dX//P39VaxYMXXo0EEbNmxwe963WkbFPG5Mat++kX+eqEGDBrLZbPr++++v27Z8+fKy2WzaunXrLcjM+93K4i+1v2b2fZS6b7jRHwdySmpxntV/nvJdDCD7+OZ0AoA3Sk5OVqtWrbRkyRLlyZNHLVu2VLFixZSQkKCdO3dq+vTp+vPPP/Xwww/ndKoebeDAgZKkxMRE/fnnn1q4cKFWrlypzZs36+OPP87h7Fy3bNkyXblyRSNGjFDnzp1v6LklS5Z0/kF26dIlbdy4UXPnztW8efM0d+5ctWnTxg0Zu+aLL77Q5cuX3bb8Nm3aqG7duipcuLDb1uENqlev7vwspTp48KCmTp1q6k+ermfPnlq/fr0+//xztWjRIsN2q1ev1u7du1W7dm3VqFHjFmboHXbt2qXg4OCcTsOrdOvWLc0PBAsWLNC2bdvUunVrVa9e3TTv2scArI+iG3CDGTNmaMmSJapWrZpWr16t3Llzm+ZfvnxZP//8cw5lZx3XHuFcvny57rvvPo0aNUovvPCCIiMjcySvm3Xs2DFJUpEiRW74uZGRkWlel4EDB+qdd95R//79ParoLlGihFuXnzt37jSfLaRVvXr1NH/Er1q1SlOnTk23P3mqTp06qW/fvvr222918uRJRUREpNvu888/lyQ9/fTTtzI9r3HnnXfmdApeJ70ftg4ePKht27YpOjraMj98AXAdw8sBN1i/fr2klC/a9IqC4OBg3XPPPek+d9asWWrWrJnCw8MVGBioyMhIPfroo9q8ebOzzfnz5/Xhhx+qadOmKlasmPz9/VWgQAE9/PDDaYYZpw7Vk1KOAF09hG3QoEEaNGiQSpUqJUmaOnWqaf61Q62XLl2qFi1aKH/+/AoICFCZMmX0yiuvKCYmJs12REZGKjIyUhcuXNBLL72kyMhI+fn53dQf+M2aNdOdd94pwzD0yy+/XLf98ePH1bt3b0VGRjpfo7Zt2+rXX381tWvSpIm6d+8uSerevbvpNcjqOYqzZ89W48aNlTt3bgUFBalKlSoaNmyY4uPjnW1Sh0SmHnW85557nOu5mWGbvXv3liQdOHBAp0+fdm6TzWZTQkKC3nnnHZUvX14BAQGmP+6OHDmi559/XqVLl1ZAQIDy5cunhx9+OMPX9uTJk3ryyScVERGhoKAgVa9eXVOnTs0wr8zOd/zhhx/00EMPqWDBggoICFDx4sXVunVrLVu2TFLKZyf1MzJ48GDTe5L6WmV2SsCvv/6qdu3aOZdfsmRJPffcczp+/Hiatlefj/rZZ5+pSpUqCgwMVEREhJ5++mmdP38+zXO2b9+uRx99VJGRkQoICFCBAgVUs2ZN9e3bV4mJiRm+JpK0ceNG2Wy2TH8gqVChggICAnTu3DlJkmEYmjp1qurXr68CBQooMDBQxYsXV/PmzTVr1qxM13cjrj7VZPfu3erUqZMKFiwou93ufN1TP9vpyWwY8p9//qlu3bqpePHi8vf3V0REhDp37qy//vorS7kFBQWpa9euSkxMzPA0kJiYGM2dO1ehoaF69NFHJaX0k3bt2ql06dIKCgpSWFiYGjRooGnTpmVpvanLyOz0k4yGOiclJWn8+PGqW7euwsLCFBwcrBo1amjs2LFyOBxp2i9atEjNmjVT4cKFFRAQoCJFiigqKkrjx4+/bo5Lly6VzWbTG2+8YZq+cuVK52fn8OHDpnmdOnWSzWbT/v37M9yWyMhIDR48WJJ5n5XRZzurnyF3uJHvxlSp23vixAk99dRTKlq0qHx8fEzv9cyZM1WrVi0FBQWpYMGCeuyxx5w/nt4qc+fO1d13363g4GCFh4frkUce0dGjR9O0y+59/8WLF/Xuu++qcuXKCgsLU65cuVSmTBl16tQpzXdpqoMHD+qRRx5R/vz5FRgYqNq1a+vbb79Nt218fLyGDx+uKlWqKDg4WGFhYWrUqJFmz559Q6/P3r171aFDB+XNm1chISGqX7++vvvuuxtaBuBOHOkG3CBfvnySpN27d2f5OYZhqHv37po6dary58+vtm3bqkCBAjpy5IhWrlyp8uXLq3bt2pJShv+98cYbaty4sVq2bKm8efPq0KFDWrRokRYvXqxvvvlGDzzwgKT/H1o6ePDgNENJU/+wiomJ0ejRo1WtWjXTBb2uPjo2ePBgDRo0SOHh4WrVqpUKFiyo7du366OPPtL333+vDRs2KCwszLRNCQkJatq0qc6dO6f7779fYWFhzgLfVYZhSNJ1zzs9cOCAGjZsqGPHjqlp06Z69NFHdfjwYc2ZM0ffffedvv76a7Vq1UpSStGVJ08eLVy4MM1Qvzx58lw3p9dff13Dhg1T/vz51blzZ4WGhmrx4sV6/fXXtXTpUv3www/y9/dXZGSkBg4cqFWrVmn16tV64oknnMXLzRy1T31NpLSvS7t27fTLL7/owQcfVHR0tAoWLChJ2rJli+6//36dO3dOzZs3V9u2bXXmzBktWLBADRs21Pz5801DeM+cOaP69etr//79atiwoRo2bKjjx4+rV69euv/++28o39Qj86GhoYqOjlbx4sV17NgxrV+/XtOmTdO9997r7IdTp05VVFRUmiIgM99++63atWsnwzDUvn17lSxZUr/++qs+/fRTLVy4UGvXrk23Hw4YMEBLly7VQw89pPvvv18rV67UxIkTtXfvXq1YscLZbvv27apTp45sNpsefvhhlSpVShcuXNDevXs1fvx4DRkyRH5+fhnmV7duXZUvX17ff/+9zp4969xfpNq0aZP+/PNPtWvXTuHh4ZKkN954Q8OGDVOpUqXUsWNH5c6dW8ePH9cvv/yiOXPmqFOnTtd51W/Mvn37VKdOHZUrV05dunRRXFxcms/3jViyZInatm2rxMREPfTQQypbtqyOHDmiefPm6bvvvtPKlStVs2bN6y6nZ8+eGjdunCZNmqRXX301zfxp06bpypUr6tmzp0JDQyVJzz77rCpVqqTGjRurcOHCOnv2rL7//ns99thj+uuvv/Tuu++6vF2ZSd3WpUuXqnz58urcubMCAwO1cuVK9enTRz///LO+/PJLZ/v//ve/euaZZ1SoUCE99NBDyp8/v06dOqXt27dr8uTJeu655zJdX6NGjeTv76/ly5dr6NChzunLly83/T/1O8AwDK1cuVKRkZEqXbp0hsvt27evFixYkGaflZ6sfobc5Ua+G6927tw51a1bV6GhoWrbtq3sdrtzJMXIkSP10ksvKU+ePHr88ceVJ08eLV26VPXr179lI23Gjx+vRYsW6eGHH1ZUVJR+/vlnzZo1S9u2bdNvv/2mgICANM/Jjn2/YRh64IEHtH79etWrV09PPfWUfH19nX+XNGrUSLVq1TKt9++//9bdd9+t0qVL67HHHtO5c+c0a9Ys54+qVx9wSEhIUPPmzbV69Wrdeeed6t27ty5fvqy5c+eqU6dO+u233/Tee+9d9/XZs2eP6tWrp7Nnz+rBBx9U9erVtXfvXkVHR+vBBx+8mZceyD4GgGy3ZcsWw8/Pz7DZbEbXrl2Nr7/+2jh48GCmz/nss88MScZdd91lxMTEmOYlJSUZx44dcz6OiYkxTp8+nWYZhw8fNgoXLmzceeedaeZJMqKiotJd94EDBwxJxhNPPJHu/BUrVhiSjHr16hn//POPad7kyZMNSUbfvn1N00uWLGlIMpo1a2bExsamu9yMSDLS2z39+OOPhs1mM2w2m/P1zCj3+++/35BkDBkyxDR93bp1ho+PjxEeHm5cvHgxzXZMnjz5hnJdv369IckoXry4cfz4cef0xMREo1WrVoYkY+jQoabnDBw40JBkrFy5MsvrWblyZYbv4VtvvWVIMkqXLu2cFhUVZUgyqlSpkqavJCYmGmXKlDECAgKMVatWmeYdPXrUKFKkiFGoUCHjypUrzuk9e/ZM933+5ZdfDF9fX0OSMXDgQNO81ByutnTpUkOSUapUKePIkSNptuXw4cNptvna5aZK7z27ePGiER4ebtjtduOnn34ytR8+fLghybjvvvtM05944gnne/j33387pycmJhqNGjUyJBk///yzc/pLL71kSDIWLFiQJqdz584ZycnJ6eZ7tffee8+QZHzyySdp5j333HOGJGPRokXOaeHh4UbRokWNS5cupWmf3r7gejLqT6mfJ0nGa6+9lu5zS5YsaZQsWTLdeen17XPnzhl58uQx8uXLZ+zcudPU/vfffzdCQkKMGjVqZDn3u+++O8PPT7Vq1QxJxi+//OKctnfv3jTt4uPjjaZNmxq+vr5p+mF6/fZ6+4f0XsvU1+L55583kpKSnNOTkpKMHj16pOlDNWvWNPz9/Y2TJ0+mWX5W3+NGjRoZPj4+pu+QunXrGjVq1DDy5ctndO3a1Tn9t99+MyQZPXr0yPK2ZLTPutHPUGZSX+tq1aoZAwcOTPdf6vquzdPV70ZJxmOPPWYkJiaa5h04cMDw8/Mz8ubNaxw4cMA5PTk52Wjbtm2G31VZkboNmX3npL7uuXLlMrZv326a9+ijjxqSjFmzZpmmZ+e+f/v27YYkIzo6Ok1uycnJxrlz55yPr953DBo0yNR2yZIlhiTjwQcfNE1P3Q8++OCDptf+5MmTzr8h1q1bl2Yd137f33fffYYkY9SoUabpCxYscOZ0o9/tQHaj6AbcZNasWUahQoWcO3xJRnh4uBEdHW36YzpV5cqVDUnGli1bbmq9ffr0MSSZ/vAxjJsruqOjow1Jxo4dO9KdX716daNAgQKmaalfmL/99tsNb0Pq65X6B9brr79utGvXzvDx8TEkGf369cs098OHDxuSjBIlShgJCQlplt+1a1dDkjF16lTnNFeL7qeeesqQZHz22Wdp5v3111+G3W43SpUqZZp+M0V3yZIlna/Lyy+/7PyD1m63G/PmzXO2T/3DK73CMPUPkZdffjnddY0aNcqQZHz33XeGYRhGQkKCERwcbOTKlSvND0KG8f9/PGal6E79IeLqXK+3zTdSdE+bNs2QZDz66KNp2icmJhqRkZFpPh+p+U+cODHNc/73v/+lKY5Ti+6lS5dedxsycvjwYcNutxu1a9c2TY+PjzfCw8ONggULmv4IDQ8PNyIjI00/hNyM6xXdERERGa7rRovu1P40duzYdJ/Tt29fQ1Kagjwjn3/+uSHJ6NKli2n6pk2bDElZLuC//vrrNPsBw8ieojs5OdkIDw83ChUqlKaQMwzD+OeffwybzWZ06NDBOa1mzZpGcHCwqZC5UYMGDTIkGQsXLjQMwzAuXLhg+Pr6GgMGDDDat29vFClSxNl2xIgRhiTjq6++ynRbDCPrRXdWP0OZSX2ts/Ivo++09GT23ZjRjx1DhgwxJBlvv/12mnn79u0z7Hb7LSm633jjjTTzUn8M79+/v2l6du77U4vu9Pan10rdd5QsWdL0I1OqEiVKGPny5TNNK1u2rGGz2Yxdu3alaZ/6Oe/evXuadaT3fV+qVKl015v6elB0I6cxvBxwk44dO6pNmzZauXKl1q5dq61bt2rt2rVasGCBFixYoMcff9x5nuClS5e0Y8cORUREZPlqu+vWrdPo0aO1YcMGnTp1SgkJCab5R48ezbYLWW3YsEF+fn6aM2eO5syZk2Z+QkKCTp8+nWaobGBgoKpWreryelPPI7TZbMqTJ48aNWqkJ598Ul27ds30eam3CWrUqFG6w3ybNm2qadOmaevWrXr88cddzk9KGaqXusxrlStXTsWKFdOBAwd0/vz5bBmK+PfffztfF19fX+d56v3791f9+vXTtL/77rvTTEs9t/Hvv/9O9xz7PXv2SEoZqtmiRQv9+eefunz5sho1apTuNjRp0iTTc7uvlno+c3pDPLNDZu+Hr6+vGjdurIMHD2rr1q1pPh+pp29crXjx4pKkf/75xzmtU6dOGj16tKKjo9W+fXvde++9atCggcqUKZPlPIsVK6ZmzZrpxx9/1B9//KGKFStKkr755hudO3dO/fr1k6/v/39Fd+nSRZ988okqVqyojh07KioqSvXq1XPb8NZq1aqlO2TVFan9bdu2ben2t9TTcHbt2uV8HTLzyCOPqF+/fvr666/1ySefKG/evJKkiRMnSkoZgn61Q4cO6f3339fy5ct16NAhxcXFmeand17szdq9e7fOnTunO+64Q0OGDEm3TVBQkHbt2uV83KVLF/Xv318VK1bUI488oqioKDVo0EAFChTI8nqbNm2qQYMGafny5Xr44Ye1evVqJSUlqVmzZoqMjNTcuXO1a9cuVahQwTncO73Piquy+hnKiieeeCLDc+hXrVqV4XVRXPlujIyMdA6/vlrq/iQqKirNvNKlS6t48eL6+++/s7I5N8WV1zU79v0VK1ZU9erVNWPGDP39999q3bq1GjZsqNq1a8vf3z/d9VavXl0+Pj7p5nv1efUXL17U3r17VbRo0XQv3pfaL69327/U+Q0bNkx3vU2aNNHq1aszXQZwK1B0A27k5+en+++/33nOa3Jysr7++mv16NFDX3zxhdq0aaPo6GjnhciKFi2apeXOnz9f7du3V2BgoO677z6VKVNGISEhzosdrV692nQBr5t19uxZJSUlOYu9jMTGxpqK7oIFC97UPX+Nq85VvhGpF+3J6FZSqdPTuwCcO9Z16NAhxcTEZEuBFBUVdUMXXStUqFCaaWfPnpWkdH9AuVpsbKyk/9/GjK4Wnd46MhITE6O8efMqKCgoy8+5ETfz3qd3/n5q4ZucnOycdvfdd2vNmjUaOnSo5s6d6zwvt3z58ho4cKDzAl7X061bN/3444+aOnWq3n//fUly/njxxBNPmNqOHDlSpUuX1uTJkzV8+HANHz5cvr6+atGihUaMGKGyZctmaZ1ZdSPv6fWk9rfUojgjqf3tekJCQtS5c2d99tlnmjZtmvr06aNLly5p5syZCgkJUZcuXZxt9+/fr7vvvlv//POPGjVqpPvvv1+5c+eWj4+P87Zp2bmvTJW6zXv27Ml0v3n1Nr/00kvKnz+/xo8frzFjxmjUqFGy2WyKiorShx9+mG7hda26desqJCTEeR738uXL5e/vr4YNGzrPxV6+fLnuuOMO/fTTT6pYsWK2vtdZ/Qy5i6vfjRm9BlnZ992KotuV1zU79v0+Pj5asWKF3nnnHc2dO9d5HYVcuXLpiSee0LBhw5zXTsgs19R8r754YHZ9T2fn9xPgThTdwC3k4+Ojjh076vfff9eQIUO0YsUKRUdHO7+ksnrE5a233pK/v782b96sChUqmOY988wz2f6rbu7cueVwOJxXUs6qmym4b0ZqcXvixIl056dewTo7iuCr15Xekc7sXJcr0nsPUnNZuHBhlu4Vn9r+5MmT6c7P6HVOT548eXT27FnFxcW5pfC+Ve99vXr19O233yo+Pl6//vqrlixZok8++USdO3dWgQIFdO+99153GW3atFFYWJimTZum9957T2fPntXixYtVrVo1VatWzdTWx8dHffv2Vd++fXXq1CmtXbtWM2fO1Jw5c7Rz507t3Lkz245MS5l/du12e5qjh6nS+wM59bXetm3bTY18udrTTz+tzz77TJ9//rn69OmjmTNn6uLFi+rRo4fpgm8ff/yxzp49q8mTJ6e5LdOMGTOyPELDbk+52UtSUlKaeZltc5s2bTRv3rwsbpX0+OOP6/HHH1dMTIzWr1+v+fPn63//+5+aN2+uP//887pHvf38/NSwYUMtXbpUJ06c0PLly1WvXj0FBwc7R94sW7ZMNWvW1MWLF7P1KLcncPW7MaP+fvW+r1KlSmnm38i+71bLjn2/JOXNm1cjR47UyJEjtXfvXq1evVqfffaZxo4dq5iYGNPFAG9Edu2rs/P7CXAnbhkG5IBcuXJJ+v8juSEhIapcubJOnjx53aFUUsqtMSpWrJjmjwqHw6G1a9em+xy73Z7hL+KpQ7Iyml+3bl39888/2rlz53Vz8wSpQ/TXrl2b7h/JK1eulCTT1ZKv9xpcb13pHX3eu3evjhw5olKlSmXpKui3St26dSVJa9asyVL7O++8U8HBwfrtt9/SvfXPjRx5r1u3rgzD0JIlS67b1pX3JLP3IykpybnNWblSdlYEBASofv36eueddzRmzBhJKX/QZkVQUJA6duyoY8eOadmyZZo+fbqSkpLSHOW+VsGCBdW2bVvNnj1bTZs21b59+7Rjx46b3pasyps3r06ePJnurdGuvrVhqhvtb1lRs2ZN1apVS9u3b9emTZsyvDf33r17JaVcyflaN/LjZOoQ9mtvuSWlv8133nmn8uTJo40bN173FnLpyZMnj1q0aKGJEyeqW7duOnfunH766acsPbdZs2aSUn5U2LFjh/OxlDJkd9WqVfrxxx9Nba/H1f3jrebKd2NmUvcT6fWV/fv3p9sfPNnNfhbLli2rJ598UqtXr1ZoaGiW93XpSb312NGjR53D2q+W3vd0eq7+vk+vf97M7TiB7ETRDbjBjBkz9OOPP6Z7H9YTJ044h1k2btzYOf2FF16QlPJr/LWFjcPhMN1fODIyUnv27DHdJ9QwDA0aNEh//PFHujnly5cvwz8Q8ubNK5vNpkOHDqU7v1+/fpJSzpVM796kly5d0saNG9N9bk4oVqyY7rvvPh08eFCjRo0yzfv55581ffp05c2b13Sf5NRh8Rm9Bhnp0aOHJGnIkCHOe2RLKX+cvvzyy3I4HHryySdd3BL3aN26tcqUKaNx48bp+++/T7fNhg0bdPnyZUkpR8+6dOmiixcvpjkPcPPmzfrqq6+yvO4+ffpIkvr375/uyI6rp7nynkRHRys8PFwzZsxI0ydHjRqlAwcO6N57772p6x2sX78+zXnB0v8faQkODs7yslKPvn7xxRf64osv5OvraxoeLaXcx3bdunVpnpuYmOgcfXIj67xZd999t5KSkjR58mTT9ClTpqSbZ/fu3ZUnTx4NHjxYmzZtSjPf4XC49Idx6rnbL7/8sjZu3KiqVauqTp06pjapQ6qvXf7SpUudhXpW1K5dW3a7XdOnT3d+LqSUW00NGDAgTXtfX1/16dNHx48f1wsvvJBufzl+/Lhpf71y5cp0T6k5deqUpKy/x6lHr4cPHy7DMNIU3efPn9f48eNlt9vTvbd4elzdP95qrnw3ZqZLly7y8/PTJ598ooMHDzqnOxwOvfLKK+l+x3uyG933HzhwwHQP91T//POP4uPjb3q0Uo8ePWQYhl555RVTwXzmzBnnrfxSv2Mzkvp9f+DAAY0dO9Y0b+HChZzPDY/B8HLADX7++WeNHj1ahQoVUsOGDZ33BD5w4IC+++47xcXFqXXr1mrfvr3zOU899ZTWrFmjL7/8UnfccYdat26tAgUK6NixY1qxYoV69OjhLHj69eunXr16qUaNGmrXrp38/Py0bt06/fHHH3rooYf0zTffpMmpWbNmmjlzph566CHVrFlTfn5+aty4sRo3bqzQ0FDVqVNHa9asUZcuXVSuXDn5+Pjo4YcfVtWqVdWsWTMNHz5cr732mu644w61aNFCpUqVUmxsrP7++2+tXr1aDRs2zNLRy1tlwoQJatCggV555RX98MMPql27tvM+3Xa7XZMnT3aOOJDkHII5atQonT171nkeWJ8+fTId3la/fn0NGDBAH3zwgSpXrqz27dsrJCREixcv1o4dO9SwYUO98sorbt/eG+Hn56d58+apefPmatmyperXr6/q1asrODhYhw8f1i+//KL9+/fr+PHjzj/033vvPS1fvlyjRo3S5s2bnffpnjVrllq0aKFFixZlad3333+/3nzzTQ0ZMkQVKlRw3qf75MmTWrt2rerWreu8eFL58uVVtGhRzZw5U35+fipZsqRsNpsee+wxlSxZMt3lh4aG6n//+586dOigqKgodejQQSVKlNCvv/6qH374QYUKFdJnn312U6/fBx98oBUrVqhRo0YqVaqUQkNDtXPnTi1evFh58+ZNc7Q1Mw0aNFDZsmU1Z84c532dr72gU1xcnBo2bKiyZcuqVq1aKlmypK5cuaIff/xRu3bt0sMPP5zmyJ479enTR5MnT9azzz6r5cuXq3jx4vrtt9+0YcMGtWrVSt9++62pfb58+TR37ly1adNGdevWVbNmzVSpUiXZbDYdPnxYGzZs0NmzZ3XlypUbyqNz5856+eWXnUftrr2AmiQ999xzmjx5sjp06KD27durSJEi2rFjh5YsWaKOHTtq1qxZWVpX4cKF1aVLF3355ZeqXr26WrZsqQsXLuj7779X48aN0x2h9NZbb2nbtm2aMGGCvvnmGzVt2lRFixbVqVOntGfPHq1bt05Dhw51XjyuTZs2Cg0NVd26dRUZGSnDMLRmzRr98ssvqlWrVpZOWZBSjvzlzZtXp06dUq5cuUwX1EotwE+dOqXatWtneQTOPffcI7vdrtdee007duxwHvl/8803s/T8W8WV78bMREZGavjw4erfv79q1KihTp06KXfu3Fq6dKliYmJUtWpVbd++3U1bk/1udN+/bds2tW3bVnfddZcqVKigIkWK6PTp01q4cKESExOd53i76uWXX9bixYu1cOFCVatWTS1atNDly5c1Z84cnTp1SgMGDFDDhg2vu5xx48apXr166tu3r3744QdVq1ZNe/fu1fz581163wG3yLHrpgNe7NChQ8bYsWON6Ohoo1y5ckauXLkMPz8/o1ChQsaDDz5ofPnllxney3fatGlG48aNjbCwMCMgIMCIjIw0OnfubPz666+mdpMnTzaqVatmBAcHG/ny5TOio6ON7du3Z3hrl5MnTxqPPvqoUbBgQedtTq6+FdOePXuMVq1aGeHh4YbNZkv3Fhtr1qwxOnToYBQuXNjw8/Mz8ufPb1SrVs3o16+f6b64hpH5bYWuRzdw79PMbnd25MgRo1evXkaJEiUMPz8/I1++fEbr1q2NTZs2pbusxYsXG3Xr1jVCQkKcOVx9b9bMzJgxw2jQoIERGhpqBAQEGBUrVjSGDBlixMXFpWmb3ffpTk96tz261smTJ41XX33VqFSpkhEUFGSEhIQYZcuWNdq1a2d8+eWXaW51dPz4caN79+5G/vz5jcDAQKNatWrG5MmTM7y1V2Y5fPfdd0bz5s2NvHnzGv7+/kaxYsWM6OhoY/ny5aZ2mzZtMpo2bWqEhYU5+2Xq65bZbZw2bdpkREdHG/nz5zf8/PyM4sWLG7169TKOHj2apm3qrXvSe6/T27alS5ca3bp1MypUqGCEhYUZwcHBRrly5Yw+ffo47x9/I959911nf5s7d26a+QkJCcb7779vPPDAA0bx4sWNgIAAI3/+/EadOnWMTz/91IiPj7/hdV7vlmEZ3T4w1Zo1a4xGjRoZQUFBRq5cuYwWLVoY27Zty7RvHzhwwOjdu7dRtmxZIyAgwMiVK5dRvnx5o2vXrsb8+fNveBsM4/9v2RcUFGT8888/6bZZt26dcc899xh58uQxQkNDjQYNGhjz58+/4X575coV4+WXXzaKFi1q+Pn5GWXKlDHee+89IzExMcPPpsPhML744gujadOmRt68eQ0/Pz+jSJEiRoMGDYyhQ4cahw4dcrb99NNPjejoaKNUqVJGUFCQkTdvXqN69erG+++/b1y4cOGGXpfUe0i3aNEizbxy5coZkowBAwak+9yMtuXLL780qlWrZgQGBqbZR9/oZygzqZ/rzPpgZvvDG/1uzMp+dfr06UaNGjWcn70uXboYR48ezdJ+NiM3csuwjD5P6b1O2bnvP3z4sPHaa68Z9evXNyIiIgx/f3+jaNGixgMPPGB8//33WcrnennFxcUZQ4cONSpVqmQEBgY6P6PTp0/P8jYbRsrfMO3atTNy585tBAcHG3Xr1jW+/fZbl28HCmQ3m2G4eHlgAAAAAACQKc7pBgAAAADATSi6AQAAAABwE4puAAAAAADchKIbAAAAAAA3oegGAAAAAMBNKLoBAAAAAHATim4AAAAAANyEohsAAAAAADeh6AYAADmqY8eOGjhwYE6nAQCAW1B0AwAAAADgJhTdAAAAAAC4CUU3AABwyYULF3I6BQAAPJ5vTicAAABuzDfffKMvv/xSrVu3VpcuXUzzJk2apKVLl+rpp5/Wvffem+lyvvvuO02dOlUPPvig9u3bpz179sgwDOXOnVuvvPKKypUr52zbrVs3JSQkqFevXpo0aZLi4uIUGRmpDz74QJcvX9b777+v3bt3Kzk5WXa7XaVKldJbb72l4OBg5zIuX76sQYMG6e+//3au59VXX83eFwcAAA/DkW4AACymZcuWstvt+umnn9LM27Bhg3x8fK5bcF9t2bJlOnDggGrVqqU77rhD58+f18CBAxUbG2tql5ycrHHjxqlAgQJq1KiRqlevrqSkJD3//PPatWuXihcvrkaNGqlYsWLat2+f+vbta3r+gAEDdPDgQeXPn19169aV3W7nAmoAAK/HkW4AACzGbrerbNmy2r17t86cOaP8+fNLko4ePaoLFy6oQoUKN7S8pKQkffrppwoPD5ckTZ06Vd99950mTJigl19+2dnOMAw1bNhQL7zwgnPaxIkTFRsbqyeffFLNmzd3Th8zZozWrl2rH3/8Uffdd5/Wr1+vU6dOqXjx4hoxYoSzXe/evXX69GmXXgcAAKyAI90AAFhQ27ZtJUmzZs1yTkv9f/v27W9oWXfccYez4Jakxx57TDabTbt27UrT9umnnzY9/uWXX+Tr66vKlSvr6NGjzn9RUVGSpI0bN0qSVqxYIUnq2rWr6fkdOnS4oVwBALAajnQDAGBBNWvWVEBAgDZv3uyc9ttvvykwMFBVqlS5oWUVLVrU9NhutyswMFBxcXGm6T4+PgoMDDRNi42NVVJSkvr165fuslMvtnb27FlJUqVKlUzzbzRXAACshqIbAACLqlGjhjZu3Ki9e/cqLi5OV65cUd26dd22PpvNlu50Pz+/DI9YlyxZ0m35AABgBRTdAABYVKdOnbRx40bNmTNHCQkJzmk36ujRo6bHDodDV65cUWho6HWfGxwcrIsXL+rhhx+W3Z7xWWv58uXT0aNHtXPnTtWoUcM5/ffff7/hfAEAsBLO6QYAwKKKFi2qPHnyaMeOHfrzzz+VO3fuNEPFs2LPnj06d+6c8/GXX34pwzCydEG22rVryzAMjRs3Ls282NhYnT9/XpLUtGlTSdK0adNMbebMmXPD+QIAYCUc6QYAwMIaNmyob7/9VpLUoEEDl5bh6+urPn36qHr16oqJidGePXvk4+OT5qJp6Xnqqae0efNmrVmzRrt27VLZsmXlcDh07NgxHTlyRL1791ZUVJTq16+vadOm6fDhw+rdu7fKli2rP//8M81tyQAA8DYc6QYAwMI6duzo/P+NXrU8VbNmzVSqVClt3rxZe/bsUVhYmAYPHqywsLDrPtfX11djx451FuwbN27Upk2bdOrUKVWqVEmVK1d2tv3ggw9UsmRJnTlzRhs2bJDD4dDgwYNdyhkAAKuwGYZh5HQSAADANQkJCXrsscdUqFAhjR49+oae+91332nq1Klq1aqVHn/8cTdlCADA7Y0j3QAAWNjMmTNlGIaaN2+e06kAAIB0cE43AAAWtGLFCu3cuVPr1q2Tv7+/WrRo4Zx35coVnTx5MtPn58uXz90pAgAAUXQDAGBJ8+fP18mTJ+Xv768+ffqY5i1fvlxTp07N9PlNmzZV8eLF3ZkiAAAQ53QDAOB1Tp48qQ0bNmTapnr16oqMjLw1CQEAcBuj6AYAAAAAwE0YXp6Oy5cvq2fPnkpMTFTfvn1Vv379nE4JHmzXrl2aMGGCTp48KYfDIR8fH5UrV06vvfaaAgMDczo9eJiPP/5YmzZtksPhUEBAgHr06KF77rknp9OCBxsyZIj++usvxcfHy2azKXfu3Hr22WdVo0aNnE4NFjJ06FBt27ZNkZGR+uCDD3I6HXi4ffv26eOPP9bp06clSX5+fnr66acVFRWVw5nBUyUlJWnw4MHas2eP8+/hO++8U2+99Zbsdq7dzSuQjoEDByooKCin04BF/PHHHzIMQ506ddLAgQP18MMP688//9SgQYNyOjV4mMmTJ2vjxo2KiorSgAEDlC9fPn366ac6fPhwTqcGD3bgwAE1aNBAr7zyil588UU5HA69//77On/+fE6nBotYtWqVfv/9d/n5+eV0KrCAkydP6o033pCPj4+effZZDRo0SO3bt1d4eHhOpwYPNmzYMO3evVtt27bV4MGD9eCDD2rnzp36+OOPczo1j8CR7mvMnTtXR48e1UsvvcQvwciSdu3aqV27ds7HlSpV0sGDB7Vt27YczAqeaMWKFSpevLieffZZSSnn1Hbp0kVffPGF3njjjRzODp5q0qRJpsclS5ZUv379tG7dOtMVy4H0xMTE6LPPPtMjjzyiBQsW5HQ6sIDRo0crICBAY8aMcU6rWLFiDmYEK/j7779VqFAhdezYUZJUoUIFrVu3TgcOHMjhzDwDR7qvcujQIc2ZM0fPPPOMQkJCcjodWNjly5fl68tvWvh/V65cUXx8vGrWrOmc5uvrqwIFCujvv//OwcxgNefOnZPELb+QNW+++aaKFy+u6OjonE4FFnHw4EFFRESoV69e6tSpk7p27apPPvkkp9OChytZsqROnjyp3377TZK0fv16xcTEcCrUv6gK/uVwODR48GBVqlRJUVFR2rVrV06nBIv6/fff9ddff6lRo0Y5nQo8yLFjxyRJERERpum5cuXSkSNHciIlWFBSUpJGjx6toKAg1alTJ6fTgYf73//+p3/++UfDhw/P6VRgIUlJSTp48KDKlSunrl27auvWrVqzZo38/PzUq1evnE4PHuq1117T22+/rffee885rWrVqnrqqadyMCvP4fVF9+uvv669e/dm2uY///mPVqxYoaSkJL3++uu3KDN4uqz2nauPXO7bt0/vvfeeChcunOa+uQBwswYMGKDY2FgNGTIkp1OBh9u9e7eWLl2qvn37KjQ0NKfTgcUEBgY69zMNGzbU33//rfXr11N0I0NffPGF9u/frwceeECVKlXStm3btGzZMk2YMIF+o9ug6O7du7dOnjyZaZuKFStqwoQJiouLU+fOnU3zRo0apenTp2vs2LHuTBMeKKt9J9X+/fv15ptvKl++fBoxYoS704PFFClSRJLS9KmLFy9y4UZkSf/+/XXs2DG9/fbbKlu2bE6nAw/3yy+/yDAMjRw5UiNHjnROP3jwoDp27Kjp06dzGhTSZbfblTdvXtO0YsWKcdFPZOqHH37Q3XffrR49ekiS6tSpoxMnTuinn36i6NZtUHQXLVpURYsWvW67l19+Wf/884/z8bFjxzRjxgy1atWKW4bdprLad6SUI9xvvvmm8ubNq1GjRvGHDNIIDAxUQECAtm7dqi5dukhKGcJ3+vRpVa1aNYezgydzOBx65ZVXdOTIEb3++utc0AhZ8uCDD6b5cWbs2LHKnTu3HnvsMb6nkKHw8HDFxMSYph07dkwBAQE5kxAsweFwpLk1mN1ul2EYOZSRZ2GP+69y5cqZHqee0122bFmOKCBTqQV3YGCg+vfvr6NHjzrnlSxZMgczg6dp2rSpFi9erAkTJuiuu+7StGnTZBiGHnvssZxODR7s5Zdf1pEjR9StWzflyZPHeeG9fPnyMWwYGQoPD09z3v+nn36q4OBgrgeATHXs2FHjx4/Xu+++q+joaG3atEkHDx7U/fffn9OpwYNFRERow4YNKlKkiKpUqaItW7bo999/V5kyZXI6NY9A0Q3cpB9//FHJycm6dOmSXnvtNdO82bNn51BW8ETdu3fXP//8o1WrVmnFihUKCAhQr169VKJEiZxODR4s9UJ7U6ZMMU1v2rQpQ/YAZLsmTZro5MmTWrRokd599135+vqqUaNGXBALmRo8eLCGDx+uefPmae7cufLx8VG5cuW4Jeq/bAbH/AEAAAAAcAvu0w0AAAAAgJtQdAMAAAAA4CYU3QAAAAAAuAlFNwAAAAAAbkLRDQAAAACAm1B0AwAAAADgJhTdAAAAAAC4yW1fdF++fFkDBw7U5cuXczoVWAx9B66g38BV9B24ir4DV9Bv4Cr6Tlq3fdEdFxenXbt2KS4uLqdTgcXQd+AK+g1cRd+Bq+g7cAX9Bq6i76R12xfdAAAAAAC4C0U3AAAAAABuctsX3QEBASpXrpx8fX1zOhVYjK+vr0qVKkXfwQ2h38BV9B24ir4DV9Bv4Cr6Tlo2wzCMnE4CAAAAAABvxM8P/7pw4UJOp5CpsLAwj8/xdsT7AlfQb+Aq+g5cRd+BK+g3cFV29B0r9D8/Pz8FBQVdtx1F978uX74sh8OR02mky2azKSwsTJcuXRIDEzyH3W5XWFiYR/cdeB76DVxF34Gr6DtwBf0GrsqOvmOV+ic0NDRLRfdtf043AAAAAADuQtENAAAAAICbUHQDAAAAAOAmFN0AAAAAALgJF1IDAABAlsXHxys+Pj6n04Cbpd5jOTY2VklJSTmcDawku/pOXFycEhISsistlwUEBCggIOCmlkHRDQAAgCy5dOmSbDabcuXKJZvNltPpwI3s9pQBsaGhoVy9HDcku/qOn5+fEhMTsystlxiGobi4OF26dEkhISEuL4fh5QAAAMiSpKQkBQcHU3ADuC3YbDYFBwff9GgPim4AAABkCcU2gNvRze77KLoBAAAAAHATim4AAAAgm/Tt21c9evRwPm7fvr3efvvtm1pmdiwD7jFixAjdd999OZ0GPBxFNwAAALxa3759VbRoURUtWlSRkZFq0KCBRo4ceUuuyj1x4kQNGDAgS23Xr1+vokWL6vz58y4vw9Nd/V6k969OnTq3JI8JEyaoYsWKunLlSpp5cXFxKl++vCZNmnRLcoH3o+gGAACA17vnnnu0detWrV27Vk8//bRGjBihTz/9NN222Xmborx58yo0NDTHl+Ep3nnnHW3dutX5T5I+/vhj5+Pvv//e1N5dt4xq3769Ll++rMWLF6eZ9+233yoxMVFt27Z1y7px+6HoBgAAgNfz9/dXwYIFVaxYMT3xxBNq1KiRfvjhB0n/PyR89OjRqlmzpho3bixJOnr0qJ555hlVqFBBlSpVUvfu3XX48GHnMpOTkzVo0CDn/CFDhsgwDNN6rx0aHh8fr6FDh6p27doqVaqUGjRooBkzZujw4cPq0KGDJKlixYoqWrSo+vbtm+4yYmJi9MILL6hixYoqU6aMunbtqv379zvnz5o1SxUqVNCqVasUFRWlO+64Q126dNHJkyedbdavX6+WLVuqbNmyqlChglq3bq0jR46k+9o9/PDDGjp0qGna2bNnVbJkSW3cuFGSNGXKFDVo0EClS5dWtWrV1LNnz3SXFRYWpoIFCzr/SVLu3Lmdj1u0aKGRI0fqhRdeUPny5TVgwIB0RwDs2LFDRYsWNb0fmzZtUps2bVSmTBnVrl1bb731li5fvpxuHvnz59d9992nmTNnppk3a9YsNW/eXHnz5tXQoUPVsGFDlSlTRvXq1dMHH3yQ6W2s0jsVoEePHs73UkrpA++8845q1aqlsmXLqlWrVlq/fr1z/pEjR/TEE0+oYsWKKlu2rO655x4tX748w3XC81F0AwAA4LYTGBhoKp7Wrl2rffv2acaMGZo6daoSExPVpUsXhYaGat68eVqwYIFCQkLUpUsX59HXzz77THPmzNGIESO0YMECxcTEaMmSJZmu98UXX9SCBQv07rvvatWqVRo+fLiCg4NVpEgRTZw4UZL0008/aevWrXrnnXfSXUa/fv20fft2TZ48WYsWLZJhGHrsscdM2xMXF6cJEyZozJgxmjdvno4ePap3331XUsqt35588knVrVtXy5Yt06JFi9SlS5cMr9Dctm1bLVy40PSDwqJFixQREaE6depo27Ztevvtt/XKK6/op59+0ldffaW6detm4V1I32effaaKFStq6dKlpmI1MwcPHlSXLl3UokUL/fjjj/r000+1adMmvfHGGxk+55FHHtG6detMPzb8/fff2rhxox599FFJUkhIiEaOHKlVq1Zp8ODBmj59uvN9ctWbb76pX3/9VePHj9eyZcvUqlUr0w8nr7/+uhISEvT1119r+fLlev3112/qHtHIeb45nQAAAABwqxiGoTVr1mj16tXq3r27c3pwcLA++ugj+fv7S5K+/vprORwOffTRR85i9OOPP1aFChW0YcMGRUVF6fPPP9fzzz+vFi1aSJKGDx+uVatWZbjuffv26ZtvvtGMGTOcR9NLlizpnJ8nTx5JKUdhc+fOne4y9u/frx9++EELFizQXXfdJUn65JNPdNddd2nJkiV66KGHJEmJiYkaPny4IiMjJUndunXTqFGjJEkXL17UhQsXdO+99zrn33HHHRnm/dBDD2ngwIHatGmT85zr+fPnKzo6WjabTUePHlVwcLDuvfdehYaGqlixYqpcuXKGy7ueBg0aqFevXs7Hx44du+5zxo4dqzZt2jiPsJcuXVrvvvuu2rVrp2HDhikwMDDNc5o0aaKIiAjNmjVL/fv3lyTNnj1bRYoUUcOGDSXJVPQXL15c+/fv18KFC/Xcc8+5tG1Hjx7VrFmztGnTJhUqVEiS1KtXL61cuVKzZs3Sa6+9pmPHjqlFixaqUKGCJHMfgTVRdAMAAOCWO3HCrvHjQ/Xcc7EqVMjh9vUtW7ZMd9xxh5KSkuRwOBQdHe0stCTpzjvvdBbckvTHH3/o4MGDKleunGk58fHxOnjwoGrUqKGTJ0+qRo0aznm+vr6qVq1amiHmqXbu3CkfHx/Vq1fP5e3Yu3evfH19VbNmTee08PBwlSlTRnv37nVOCwoKchbUkhQREaEzZ85ISjlHvGPHjurSpYsaNWqkRo0a6aGHHlJERES668yXL58aN26sefPmqU6dOjp06JB+/fVXvf/++5Kkxo0bq1ixYqpXr56aNGmie+65Rw8++KCCgoJc2saqVave8HP++OMP7dq1S/Pnz3dOMwxDDodDhw8fTvdHBR8fH3Xo0EGzZ8/WSy+9JMMwNGfOHHXq1El2e8qA4IULF+p///uf/v77b126dEnJyck3dX79rl27lJycrEaNGpmmJyQkKG/evJJShqO/9tprWr16tRo1aqQWLVqoYsWKLq8TOY+iGwAAALfc+PGhmjQppXh5550Lbl9f/fr1NWzYMPn7+ysiIkK+vuY/g4ODg02PL126pKpVq+qTTz5Js6x8+fK5lEN6R1vdxc/Pz/TYZrOZfgwYOXKknnzySa1cuVKLFi3SBx98oBkzZqhWrVrpLq9t27Z66623NGTIEM2fP18VKlRwHokNDQ3VkiVLtH79ev3000/66KOPNGLECH3//fcZHrHPzLXvRWoBfHX+1155/tKlS+ratavpdm2pihYtmuG6HnnkEY0dO1Zr166VYRg6duyYOnXqJEnavHmz+vTpo/79+6tJkybKlSuXFi5cqP/+978ZLi+9IfpXD/u/dOmSfHx8tHjxYvn4+JjapQ4h79y5s6KiorR8+XL99NNPGjt2rN5+++10tw3WwDndAAAAuOWeey5WTz4Zq+eei70l6wsODlapUqVUtGjRNAV3eqpUqaIDBw4of/78KlWqlOlfWFiYwsLCFBER4bwCt5RSCG7fvj3DZVaoUEEOh0MbNmxId35qoZycnJzhMsqWLaukpCRt2bLFOe3cuXPat29fpkPE01O5cmX16dNHixYtUvny5bVgwYIM2zZv3lzx8fFauXKlFixYoDZt2pjm+/r6qnHjxnrzzTe1bNkyHTlyROvWrbuhfDKS+iPHqVOnnNN27txpalOlShXt3r07zXtVqlQp0wiGa0VGRqpu3bqaNWuWZs2apUaNGqlYsWKSUoruYsWK6cUXX1S1atVUunRpHT169Lq5Xn3BuuTkZP3111/Ox5UrV1ZycrLOnj2bJs/UC8tJKT8UPP744/r888/1zDPPaPr06Vl4peCpKLoBAABwyxUq5NA771y4JUPLXdG2bVvlzZtX3bt3188//6xDhw5p/fr1euutt5znGD/55JMaO3aslixZor179+r111/XhQsZH7UvXry4OnTooP79+2vJkiXOZS5atEiSVKxYMdlsNi1btkxnz57VpUuX0iyjdOnSat68uQYMGKBNmzZp586deuGFF1SoUCE1b948S9t26NAhDRs2TJs3b9aRI0e0evVqHThwQGXLls3wOcHBwXrggQf04Ycfas+ePYqOjnbO+/HHHzVp0iTt2LFDR44c0Zw5c+RwOFSmTJks5XM9kZGRKlKkiEaMGKH9+/dr2bJl+uyzz0xtnnvuOW3evFlvvPGGduzYof3792vp0qWZXkgt1aOPPqrFixdryZIleuSRR5zTU4vshQsX6uDBg5o0aVK6txi7WoMGDbR8+XItW7ZMe/fu1WuvvWbqE2XKlFHbtm314osv6vvvv9ehQ4e0detWffLJJ1q2bJkk6e2339aqVat06NAh/f7771q3bl2m7w08H0U3AAAAcI2goCDNmzdPRYsW1VNPPaUmTZro5ZdfVnx8vHLlyiVJeuaZZ9SuXTv17dtXDz/8sEJCQvTAAw9kutxhw4apZcuWev311xUVFaVXXnlFcXFxkqTChQurf//+GjZsmKpVq5Zhwfjxxx+rSpUqeuKJJ/Twww/LMAx9+eWXaYaUZ7Zte/fu1dNPP61GjRppwIAB6tatmx577LFMn9emTRv98ccfqlOnjmnIdu7cubV48WJ16tRJUVFR+vLLLzVu3DiVL18+S/lcj5+fn8aPH699+/bpvvvu0/jx4zVgwABTm4oVK+rrr7/W/v371bZtWzVv3lwffvhhhuepX61Fixby9/dXYGCg6f27//771bNnT73xxhu6//77tXnz5uteTf2RRx5Rhw4d9OKLL6pdu3YqUaKE6tevb2rz8ccfq3379nrnnXfUuHFjPfnkk9q2bZvzNXU4HHrjjTfUpEkTdenSRaVLl9Z7772XxVcLnshmZHSlh9vMiRMn5HB45i+tNptNhQsX1vHjxzO8MAduPbvdrkKFCnl034Hnod/AVfQduCo7+86FCxcUFhaWTZnBk7HPgauyq+/4+fllek/0WymjfV9oaGiW9okc6QYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HotoDff/dV9eopEQAAAABgHRTdFvDss3m0bVtKBAAAAABYB0W3BaTeJYy7hQEAAACAtVB0W8CECTGqVi0lAgAAAACsg5OELaBKlST99pt0/HgSR7sBAAAAwEI40m0BXEgNAADAWtavX6+iRYvq/PnzLi9jxIgRuu+++7Ixq6zp27ev+vfvf1PLmDVrlipUqJBpm5zaPuBWo+i2gH79Ui6k1q9fnpxOBQAAwHL69u2rHj16pJmeHYWxK1LXm9m/9evX39KcrOKrr75SmzZtVLFiRVWsWFGdOnXS1q1bs/z8X375RSVKlEi32J8yZYrq1Kmj0qVLq1WrVqblHj58OMP36ptvvnG2e+utt/TAAw+oVKlS6a5jxIgR6S6jbNmyzjZ//fWXevbsqTp16qho0aKaOHFipts0duxYFS1aVG+//bZp+qlTp9SnTx9Vr15dZcuWVfPmzfXdd98552fWD3/77Tdnu1WrVqlVq1YqV66cqlSpop49e+rw4cOmda1fv17NmzdXqVKlFB0dnSbH5ORkffDBB6pbt67KlCmj+vXra+TIkTL+HcabmJiooUOHqlmzZipbtqxq1qyp3r1768SJE6b3oH///qZlfPTRR0pISLih13fWrFlp5pcuXTrT1/hmcejUAkaOjNGAAQX0wQcxOZ0KAAAArpKQkCB/f/8bek7t2rVNBd3bb7+t2NhYffzxx85pefLk0YYNG25JPlayYcMGtW7dWrVr11ZgYKDGjRunzp07a8WKFSpcuHCmzz1//rxefPFFNWzYUKdPnzbNW7hwoQYPHqzhw4erRo0a+vzzz9WlSxf99NNPyp8/v4oUKZKmuP/qq6/06aefqmnTpqbpjzzyiLZs2aJdu3alyaFXr1567LHHTNM6deqkatWqOR/HxcWpRIkSatWqlQYNGpTpNv3222+aNm1auqMKXnzxRV24cEGTJ09WeHi45s+fr169emnx4sWqXLlymn4oSR9++KHWrl3rzOfQoUPq0aOHevbsqbFjx+rixYsaNGiQnnrqKS1dutTZ5vHHH9djjz2msWPHOgv29evXq27dupKkcePG6YsvvtCoUaNUvnx5bdu2TS+99JLCwsL05JNPKi4uTr///rtefPFFVaxYUefPn9egQYPUvXt3LV68WJK0d+9eORwOvf/++4qMjNRff/2lV155RZcvX3b+4JCV11eScuXKpZ9++sn52GazZfo63yyKbgAAAEDSuXPn9Oabb+rnn39WTEyMIiMj1adPH9ORu/bt26t8+fLy8fHRvHnzdOedd2ru3Llavny5Bg4cqOPHj6tGjRrq0KFDhuvx9/dXwYIFnY8DAwOVkJBgmna1uXPn6sMPP9T58+d1zz336MMPP1RoaGim+fz5558aMmSIfv75ZwUHB6tx48YaPHiwwsPDJUnffvutRo4cqYMHDyowMFCVK1fW5MmTFRwcbFp38+bNdejQIbVu3VqDBw+Wn5+fJCkmJkZvv/22li1bpvj4eNWrV0/vvPNOpkcMx44dq4kTJyouLk4PPfSQ8uXLl/kbksEyrvbRRx/p+++/19q1azN9zSXpP//5j6Kjo+Xj46MlS5aY5k2cOFGdO3dWp06dJEnDhw/X8uXLNXPmTD3//PPy8fFJ8/4sXrxYDz30kEJCQpzT3n33XUnS2bNn0y26Q0JCTO137typ3bt3a/jw4c5p1atXV/Xq1SVJ7733Xobbc+nSJT3//PP64IMPNGbMmDTzN2/erGHDhqlGjRqSUkZ8TJw4Udu3b1flypXT9MPExEQtXbpU3bt3dxah27dvV3Jysl599VXZ7SmDpJ955hn16NFDiYmJ8vPz05dffqkSJUpo4MCBkqTy5ctLkqZPn+4sujdv3qzmzZvr3nvvlSQVL15cCxcudBboYWFhmjlzpin/YcOGqXnz5jp69KiKFi2qe+65R/fcc49zfsmSJbVv3z598cUXzqI7K6+vlFJkZ/R5cweGl1sAw8sBAIAns12+nOE/XbmS9bZxcVlq6y7x8fGqWrWqpk6dqhUrVqhLly564YUX0hwNnDNnjvz9/bVgwQINHz5cR48eVc+ePXXfffdp6dKl6ty5s4YNG5YtOf39999aunSppk6dqqlTp2rjxo1pCs9r8zl//rw6duyoSpUqafHixfrqq6905swZPfPMM5KkkydPqnfv3urUqZNWrVqluXPn6sEHH3QO9ZVSiiRJ+uyzzzRq1CjNnj1bs2fPds7v16+ftm/frsmTJ2vRokUyDEOPPfaYEhMT092ORYsW6eOPP9Z//vMfff/99ypYsKCmTp1qapM63PnaocuZiYuLU1JSkvLkyZNpu1mzZunQoUN66aWX0sxLSEjQ9u3b1ahRI+c0u92uhg0b6tdff013edu3b9fOnTv1yCOPZDnX9MyYMUOlS5dWnTp1bvi5r7/+upo1a6bGjRunO7927dpatGiR/vnnHzkcDi1cuND5A0l6fvjhB/3zzz/OHx4kqWrVqrLb7Zo1a5aSk5N14cIFff3112rUqJHzB5hff/1VDRs2TLO833//3ZTL2rVrtW/fPkkpxfCmTZtMRfS1Lly4IJvNprCwsEzbZPbeZ/T6Xrp0SXfffbdq166t7t2766+//spwGdnBske6HQ6HZs+erTVr1igmJkbh4eGKiopSu3bt3D484FZjeDkAAPBkhe+4I8N5V5o21bkvv3Q+jqhaVfZriutU8fXq6ezcuc7HBevUkc+5c2naHTt69IZzXLZsme64Jk+Hw2F6XLhwYfXq1cv5uEePHlq1apW++eYb59FCSSpVqpTefPNN5+Nhw4apZMmSziN9ZcuW1Z9//qlx48bdcJ7XcjgcGjlypPPIdrt27bR27VpTm2vzGTVqlCpXrqzXXnvNOW3EiBG66667tG/fPl2+fFlJSUlq0aKFihUrJklphifnypVLkhQZGakSJUqoWbNmWrt2rbp06aL9+/frhx9+0IIFC3TXXXdJkj755BPdddddWrJkiR566KE02/H555/rkUce0aOPPipJevXVV7VmzRrFx8c72wQFBalMmTLy9c16iTJ06FBFRESYCuZr7d+/X++9957mzZuX7rLPnTun5ORk5c+f3zS9QIECziLxWjNmzNAdd9zh3H5XXLlyRfPnz1fv3r1v+LkLFy7Ujh07TOdoX2vChAl69tlnVblyZfn6+iooKEiTJk1SqVKl0m0/c+ZMNWnSREWKFHFOK1GihKZPn65evXrp1VdfVXJysmrVqqUvr/pMnzp1SgUKFEizvNjYWMXFxSkoKEjPP/+8YmNjFRUVJR8fH+fR87Zt26aby5UrV/Tuu+8qOjra2RevdeDAAU2ePFlvvfVWhstI7/UtU6aMRowYoQoVKujixYuaMGGCWrdurRUrVpi2PTtZtuhesGCBfvzxR/Xu3VvFihXT/v37NX78eAUHB6tFixY5nV624pZhAAAAN6d+/fppjj5v3bpVffr0cT5OTk7WmDFj9O233+rEiRNKSEhQQkKCgoKCTM+rWrWq6fHevXtNRbkk1apVK1vyLl68uLPglqSCBQvq7Nmzmebzxx9/aP369Wl+ZJBSjpxHRUWpYcOGatasmaKiohQVFaWWLVuajhheO0w8IiLCOVx679698vX1Vc2aNZ3zw8PDVaZMGe3duzfd7di7d2+ac21r1aplumBcjRo1TOfZXs/YsWO1aNEizZkzR4GBgem2SU5O1vPPP6/+/furTJkyWV52ZuLi4rRgwQK9+OKLN7WcJUuWKDY29rrD4q919OhRvf3225oxY0aG2y2lnJ994cIFzZw5U+Hh4Vq6dKl69eqlefPmpfmR5dixY1q1apUmTJhgmn7q1Cm98sor6tChg6KjoxUbG6uPPvpITz/9tGbOnJnlg53ffPON5s2bp3HjxqlcuXLauXOnBg4cqIiICHXs2NHUNjExUb169ZJhGBmOGDl+/Li6du2qVq1aqUuXLum2yej1rV27tmrXrm163KRJE02bNk0DBgzI0vbcKMsW3bt371bt2rWdH/aCBQtq7dq1GX7QAQAA4B7H9+zJcJ5hN5/NeHL79ozbXvMH/Kmff765xK4SHByc5gjf8ePHTY8//fRTTZo0SYMHD9add96p4OBgDRw4MM2Q6WuLcHe69siszWZLc4T+2nwuX76s++67T6+//nqa5UVERMjHx0czZ87U5s2btXr1ak2ePFnvv/++vv32W5UoUUKSnEOHr2Z40NGfCRMmaNy4cZo5c6YqVqyYYbvY2Fht27ZNO3bscI4GcDgcMgzDeRT37rvvlo+Pj86cOWN67unTp9M9gvvdd98pLi7uhovla02fPl333ntvuuvIzO+//64zZ87ogQcecE5LTk7Wxo0bNWXKFB04cECHDx/W5MmTtWLFCuc51pUqVdLPP/+sKVOm6P333zctc9asWcqbN6/uv/9+0/QpU6YoLCzMNJJizJgxuuuuu7RlyxbVqlVLBQsWTHNhOkkKDQ119s13331Xzz//vFq3bi0pZWTFkSNHNHbsWFPRnVpwHzlyRPPnz0/3KPeJEyfUoUMH1apVSx988EGGr1NWX18/Pz9VqlRJBw8ezLTdzbBs0V2uXDktX75cx44dU5EiRXTw4EH99ddfevzxxzN8TmJiommn6ePjo4CAAEkpOzBPHZb+++9+evBB6YMP/FSlSvrnyeDWS+0vntx34HnoN3AVfQeuuhV9x7jm4ls50TY7/PLLL2revLnatWsnKaU4279/v8qVK5fp88qWLasff/zRNG3Lli1uy/N6KleurO+//17FixfPcKi2zWbTXXfdpbvuukv9+vXT3XffrcWLFzvP+85M2bJllZSUpC1btjiHV587d0779u1L9+h66nO2bt1qKlRdfY3Gjx+vMWPG6KuvvkpzVepr5cqVS8uXLzdNmzp1qtatW6f//ve/KlGihPz9/VW1alWtXbvWWcg6HA6tXbtW3bt3T7PMmTNn6r777nPpQnCpDh06pPXr12vy5Mk3/NyGDRum2aaXXnpJZcqUUe/eveXj46O4f0/hsF/zo5ePj0+aH08Mw9Ds2bPVvn37ND+2xMXFpbsM6f9Pz6hVq5ZWrFiRJs8qVaqYlnPt/sfHx8f0A1JqwX3gwAHNmTNH4eHhaX7wOn78uDp06KCqVatq5MiRaXJLdSOvb3Jysv788880V6G/1s3sPy1bdEdHRysuLk79+vWT3W6Xw+HQI488kun5HPPnz9fcq84Tio6OVufOnSWl/OrnqerUkQ4flp56Kr8OHcrpbHAtT+478Fz0G7iKvgNXZUffiYuLS/cIqKez2+2y2Wxpck8tSP38/OTn56cyZcro22+/1datW5UnTx5NmDBBZ86cUfny5Z3Ptdls8vHxMS2rR48e+u9//6uhQ4eqa9eu2rZtm/OiY6nLdiW/9Kb7+PiYpqWXT8+ePTVjxgw9//zzev7555UnTx4dOHBACxYs0MiRI/Xbb79pzZo1atKkifLnz68tW7bo3LlzqlChgvz8/JzrTc3f4XCY1lu+fHk98MADevXVV51XUh8yZIgKFy6sVq1ayc/Pz/TaSilXvH7hhRdUs2ZN3X333Zo7d652796tkiVLOtts2bJFzz//vL7++usMb/81ZswYffjhh/r0009VqlQpnfv3nP+QkBDnMPwhQ4bo+PHjznPqry7+pJQRsoGBgabpzz77rPr06aMaNWqoZs2a+uyzzxQXF6cuXbqYXtv9+/dr48aNmjFjRrrv6/79+3Xp0iWdOXNG8fHx+vPPPyWlXNH76lu5zZkzRxEREWrevLmziE2VkJDgvLBXYmKiTp06pT///FMhISEqXbq08ubNq7x585qeExISovz58zu3qUKFCipVqpT+85//aNCgQcqbN68WL16sn376SV999ZUp959++sl5269rt6l58+aaOHGiRo8erbZt2yo2NlZDhw5V8eLFVaNGDfn5+al79+6aMmWK3nvvPXXu3Flbt25V79699dhjjzmX17x5c33yyScqWbKkypcvr99//10TJ07Uo48+Kj8/P2fBvX37dk2bNk02m00nT56UJOXNm1f+/v7OgrtYsWJ65513dP78eWee1+7fMnt9P/roI9WqVUulSpXShQsXNG7cOB09ejTd7U/l7+9/3VvSZcayRfeGDRu0du1avfDCCypevLgOHjyoKVOmKG/evGrSpEm6z2nTpo1atWrlfHz1G3Dy5Mk0Q3U8xYkTEZLsOnHCoePHT+Z0OviX3W5XRESER/cdeB76DVxF34GrsrPvJCQkZHh1ak+WOpz42tyTkpIk/f9oyD59+ujgwYPq1KmTgoKC1KVLFzVv3lwXL150PtcwDCUnJ5uWFRERof/+978aNGiQJk2apOrVq+s///mPXnrppTQjLW8kv/SmJycnm6all0++fPk0f/58vffee+rYsaPi4+NVrFgxNWnSRMnJyQoKCtL69ev12WefKTY2VkWLFtXbb7+txo0bKzEx0bne1NfG4XCkWe+IESP09ttvq2vXrkpISFDdunX1xRdfOJ9z9WsrSS1bttS+ffs0ePBgxcfHq0WLFnr88ce1atUqZ5uLFy9q7969iouLy/A1mzJlihISEvTkk0+apr/00kvq37+/pJSjoUeOHMlwGem9ri1bttSpU6f0/vvv6/Tp06pUqZKmTZumvHnzmtpNmzZNhQsXVoMGDdJdfr9+/Uz3V2/WrJkkaePGjSpevLhz/TNmzFCHDh3kcDjSfC6PHDnifJ6UcmR//PjxqlevnukA4tXS6wdffPGFhg0bpq5du+rSpUuKjIzUqFGjFBUVlWabateurcjIyDTbVLduXY0bN07jx4/X2LFjFRQUpFq1amnatGny9fVVYmKiihQpoqlTp2rQoEGaOHGiqlevrt69e6tOnTrO5b3zzjv64IMPNGDAAJ09e1YRERHq0qWL+vXrp8TERB0+fNh5G7drjzjPmTNH9evX1/Lly3XgwAEdOHAgzQiHo1ddXPF6r++5c+f00ksv6fTp08qdO7eqVKmihQsXqnTp0hn2mYSEhDSno0gpP3ZkdnX1VDbDk07OuAHPPvusWrdubTqX4euvv9aaNWs0atSoG17eiRMnPPaPmAULAtW/f7hGjDin6Ogr138Cbgm73a5ChQp5dN+B56HfwFX0HbgqO/vOhQsXsvQHJqyPfQ5clV19J/UIuCfIaN8XGhqapX2iZe/THR8fn2YMv91u96gLPAAAAAAAbm+WLbpr1aqlefPmacuWLTp16pQ2bdqkb7/99qbuleep+vXLoytXUiIAAAAAwDose053jx49NGvWLH3++ec6f/68wsPDdd9996l9+/Y5nVq2e/TRWE2dGqZHH43N6VQAAAAAADfAskV3UFCQunXrpm7duuV0Km43bVouZ3zvPQpvAAAAALAKyw4vv70Y10QAAAAAgBVQdFvAyy9fMEUAAAAAgDVQdFvAN9+EmCIAAEBO4fZRAG4n2bHPo+i2gN69LyowMCUCAADklODgYF28eJHCG8BtweFw6OLFiwoODr6p5Vj2Qmq3k4EDw3TlSkqMjr6S0+kAAIDblK+vr0JCQhQby4VdvZ2vb0qZEBsbq6SkpBzOBlaSXX3H399fCQkJ2ZWWy0JCQpzb5CqKbgs4c8bHFAEAAHKKr6+vwsLCcjoNuJndnjIgNjQ0lJENuCHZ0XdsNpsKFy6s48ePyzCsfzFphpdbQKtWl0wRAAAAAGANFN0W8MMPwaYIAAAAALAGim4LGDkyRoGBKREAAAAAYB0U3RZQtmyyypdPiQAAAAAA66DotoBnn82jbdtSIgAAAADAOii6LeD4cZspAgAAAACsgaLbAuLifEwRAAAAAGANFN0WUL36FVMEAAAAAFgDRbcFbNsWYIoAAAAAAGug6LYAw7CZIgAAAADAGii6LaBRozhTBAAAAABYA0W3BRw96meKAAAAAABroOi2gAsXzBEAAAAAYA0U3RZw7pyvKQIAAAAArIGi2wIMwxwBAAAAANZA0W0BFN0AAAAAYE0U3QAAAAAAuAlFNwAAAAAAbkLRDQAAAACAm1B0AwAAAADgJhTdAAAAAAC4CUW3JRjXRAAAAACAFVB0W4LtmggAAAAAsAKKbgAAAACAxzhxwq4XX0yJ3sA7tgIAAAAA4BXGjQvVmDEp0Rv45nQCAAAAAACk6t07ViEhIerWLTanU8kWHOkGAAAAAHiMQoUcGj06JXoDim4AAAAAANyEohsAAAAAADeh6AYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HoBgAAAADATSi6AQAAAABwE4puAAAAAADchKIbAAAAAAA3oegGAAAAAMBNKLoBAAAAAHATim4AAAAAANyEohsAAAAAADeh6AYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HoBgAAAAB4jBMn7HrxxZToDbxjKwAAAAAAXmHcuFCNGZMSvYFvTicAAAAAAECq3r1jFRISom7dYnM6lWzBkW4AAAAAgMcoVMih0aNTojeg6AYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HoBgAAAADATSi6AQAAAABwE4puAAAAAADchKIbAAAAAAA3oegGAAAAAMBNKLoBAAAAAHATim4AAAAAANyEohsAAAAAADeh6AYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HoBgAAAADATSi6AQAAAABwE4puAAAAAADchKIbAAAAAAA3oegGAAAAAMBNKLoBAAAAAHATim4AAAAAANyEohsAAAAAADeh6AYAAAAAwE0ougEAAAAAcBOKbgAAAAAA3ISiGwAAAAAAN6HoBgAAAADATSi6AQAAAABwE4puAAAAAIDHOHHCrhdfTInewDu2AgAAAADgFcaNC9WYMSnRG1B0AwAAAAA8RseOl1WtWkr0BhTdAAAAAACPMXt2sLZtS4newDenEwAAAAAAIFXv3rEKCQlRt26xOZ1KtuBINwAAAADAYxQq5NDo0SnRG1j6SPe5c+c0bdo0/fbbb4qPj1ehQoX03HPPqUyZMjmdGgAAAAAA1i26Y2Nj9dZbb6lSpUp6/fXXFRYWpuPHjyskJCSnUwMAAAAAQJKFi+6FCxcqX758eu6555zTChYsmIMZAQAAAABgZtmie/PmzapWrZo+/vhj/fHHHwoPD9f999+ve++9N8PnJCYmKjEx0fnYx8dHAQEBkiSbzSabzeb2vG+WFXK8XaS+F1bpO/AM9Bu4ir4DV9F34Ar6DVyVHX3n6mV4A5thGEZOJ+GKLl26SJJatmypevXqad++fZo8ebJ69uypJk2apPuc2bNna+7cuc7H0dHR6ty5861I96Zc3des+W4BAAAAwO3Jske6HQ6HypQp4yyaS5UqpUOHDunHH3/MsOhu06aNWrVq5Xzs4+Pj/P/JkyflcHjq1fEKSbJJMnT8+ImcTgb/stvtioiI8PC+A09Dv4Gr6DtwFX0HrqDfwFXZ0XdsNpsKFSqkEydOyJOPEYeEhCgsLOy67SxbdOfNm1fFihUzTStWrJh+/vnnDJ/j5+cnPz+/dOcZhuHRb2gqK+R4u0h9L6zSd+AZ6DdwFX0HrqLvwBX0G7gqO/uOt/Q/y96nu3z58jp27Jhp2rFjx1SgQIEcyggAAAAAADPLFt0tW7bUnj17NG/ePJ04cUJr167V8uXL1bx585xODQAAAAAASRYeXl62bFm9/PLLmj59ur7++msVLFhQTzzxhBo1apTTqQEAAAAAIMnCRbck1apVS7Vq1crpNAAAAAAASJdlh5cDAAAAAODpKLoBAAAAAHATim4AAAAAANyEohsAAAAAADeh6AYAAAAAeIwTJ+x68cWU6A28YysAAAAAAF5h3LhQjRmTEr2BpW8ZBgAAAADwLr17xyokJETdusXmdCrZgiPdAAAAAACPUaiQQ6NHp0RvQNENAAAAAICbUHQDAAAAADwGF1IDAAAAAMBNuJAaAAAAAABuwoXUAAAAAABwEy6kBgAAAAAAsoSiGwAAAADgMbiQGgAAAAAAbvLhhykXUvvwQ++4kBpFNwAAAADAYxiGzRStjquXAwAAAAA8xoABF1WwYLC6dbuY06lkC450AwAAAAA8BlcvBwAAAADATbiQGgAAAAAAbjJuXMqF1MaN40JqAAAAAABkq44dL6tatZToDSi6AQAAAAAeY/bsYG3blhK9AVcvBwAAAAB4jN69YxUSEqJu3WJzOpVswZFuAAAAAADchKIbAAAAAOAxuJAaAAAAAABuwoXUAAAAAABwkylTUi6kNmWKd1xIjaIbAAAAAOAxDMNmilbH1csBAAAAAB5jwICLKlgwWN26XczpVLIFR7oBAAAAAB6jUCGHRo9Oid6AohsAAAAA4DFOnLDrxRdTojfwjq0AAAAAAHiFDz9MuWXYhx9yyzAAAAAAALIVF1IDAAAAAMBNune/pF27gtW9+6WcTiVbcKQbAAAAAOAxJk8O0bZtKdEbUHQDAAAAADxGXJw5Wh1FNwAAAADAYwQFmaPVcU43AAAAAMBjDBhwUQULBqtbt4s5nUq24Eg3AAAAAMBjnD5t1+rVKdEbeMdWAAAAAAC8Qr9+ebRtW0r0BhTdAAAAAACPMXJkjKpVS4negKIbAAAAAOAxChRwKCoqJXoDim4AAAAAgMf48MNQjRmTEr0BRTcAAAAAwGMYhs0UrY5bhgEAAAAAPEb37pe0a1ewune/lNOpZAuOdAMAAAAAPMbkySHati0legOKbgAAAACAx4iLM0ero+gGAAAAAHiMoCBztDrO6QYAAAAAeIwBAy6qYMFgdet2MadTyRYc6QYAAAAAeIxChRwaPTolegOKbgAAAACAxzhxwq4XX0yJ3sA7tgIAAAAA4BU+/DBUY8akRG9A0Q0AAAAA8BiGYTNFq+NCagAAAAAAj9G9+yXt2hWs7t0v5XQq2YIj3QAAAAAAjzF7drC2bUuJ3oAj3QAAAAAAj9G7d6xCQkLUrVtsTqeSLTjSDQAAAACAm3CkGwAAAADgMQYPzqWFC6W//86l8eNjcjqdm8aRbgAAAACAx9i+3c8UrY6iGwAAAADgMYoXTzJFq6PoBgAAAAB4jBMnfE3R6ii6AQAAAAAe45NPYlStWkr0BhTdAAAAAACPUaCAQ1FRKdEbUHQDAAAAADzGuHGhGjMmJXoD7xgkDwAAAADwCr17xyokJETdusXmdCrZgiPdAAAAAAC4CUU3AAAAAMBjMLwcAAAAAAA3YXg5AAAAAABuUqiQQ6NHp0RvQNENAAAAAICbUHQDAAAAAOAmFN0AAAAAAI9x4oRdL76YEr2Bd2wFAAAAAMArfPBBLo0ZkxK9AUU3AAAAAMBj2GyGKVodRTcAAAAAwGN063ZZ1aqlRG9A0Q0AAAAA8BhTpgRr27aU6A0ougEAAAAAHuPsWbspWp13bAUAAAAAwCvs3etrilZH0Q0AAAAA8BhVqiSaotV5x08HAAAAAACvMHDgRZUsGaxu3S7mdCrZgiPdAAAAAACPUaiQQ6NHp0RvQNENAAAAAPAYJ07Y9eKLKdEbeMdWAAAAAAC8wgcf5NKYMSnRG1B0AwAAAAA8RlycOVodRTcAAAAAwGPExdlM0eoougEAAAAAHmPPHl9TtDqKbgAAAACAx6haNdEUrc47fjoAAAAAAHiF5567pIMHg/Xcc5dyOpVswZFuAAAAAIDHmD07WNu2pURvQNENAAAAAPAYzZpdUf78KdEbeE3RvWDBAnXs2FFTpkzJ6VQAAAAAAC76z3/CdOZMSvQGXlF07927Vz/++KNKliyZ06kAAAAAAG7CP//YTdHqLL8VV65c0SeffKJnnnlGISEhOZ0OAAAAAOAmDB9+XoGBKdEbWP7q5Z9//rlq1KihqlWrat68eZm2TUxMVGLi/1923sfHRwEBAZIkm80mm83zb75uhRxvF6nvhVX6DjwD/Qauou/AVfQduIJ+A1dlR99ZuzZQV66kxDZt4rMzvRxh6aJ73bp1OnDggIYNG5al9vPnz9fcuXOdj6Ojo9W5c2dJUkREhFtyzF42FS5cOKeTwDWs0Xfgaeg3cBV9B66i78AV9Bu46mb6jmGkxmAVLmz9K5hbtug+c+aMpkyZojfffFP+/v5Zek6bNm3UqlUr52MfHx/n/0+ePCmHw5HteWaPQpJskgwdP34ip5PBv+x2uyIiIjy878DT0G/gKvoOXEXfgSvoN3BVdvQdmy23pGDZbJd1/LjnDjEPCQlRWNj1L/Zm2aJ7//79On/+vF599VXnNIfDoV27dmnJkiWaPn267HbzKet+fn7y8/NLd3mGYchI/UnFg1khx9tF6nthlb4Dz0C/gavoO3AVfQeuoN/AVdnRd7p1u6Q//ghWt26XvKL/WbborlKlij766CPTtE8//VRFihRR69at0xTcAAAAAADPN3lyiLZtS4kjRsTkdDo3zbJFd1BQkEqUKGGaFhAQoFy5cqWZDgAAAACwBpvNMEWr43AwAAAAAMBjPPTQFeXPnxK9gWWPdKdn0KBBOZ0CAAAAAOAmDBwYpjNnUuKqVadzOp2bxpFuAAAAAIDHqFgx0RStzquOdAMAAAAArG3gwIsqWTJY3bpdzOlUsgVHugEAAAAAHuP0abtWr06J3sA7tgIAAAAA4BWefTaPtm1Lid6AohsAAAAA4EFs10Rro+gGAAAAAHiMIUPOK3/+lOgNKLoBAAAAAB5j+fJAnTmTEr0BVy8HAAAAAHiM3r1jFRISom7dYnM6lWzBkW4AAAAAgMcoVMih0aNTojeg6AYAAAAAeIzff/dV9eop0RtQdAMAAAAAPEafPim3DOvTJ09Op5ItKLoBAAAAAB4jMjLJFK2OohsAAAAA4DHCww1TtDqKbgAAAACAx2jU6IoCA1OiN6DoBgAAAAB4jDffzK0rV1KiN6DoBgAAAAB4jHHjYpQ/f0r0Bt5xDXYAAAAAgFdo0iRBp09Lx48nyPCC07o50g0AAAAA8BjcpxsAAAAAADfp1SvlPt29euXJ6VSyBUU3AAAAAMBjJCebo9VRdAMAAAAAPEaFCkmmaHUU3QAAAAAAj5E3r2GKVkfRDQAAAADwGFWqxMtuT4negKIbAAAAAOAxBg/OLYcjJXoDim4AAAAAgMcYOPC87PaU6A0ougEAAAAAHmP16kA5HCnRG1B0AwAAAAA8xsaN/qZodRTdAAAAAACPUbdugilaHUU3AAAAAMBjREVdkd2eEr0BRTcAAAAAwGO8/XYeORwp0RtQdAMAAAAAPIhxTbQ2im4AAAAAgMd45pmLpmh1FN0AAAAAAI/x1Vehpmh1FN0AAAAAAI9Rp06CKVqdb04nAAAAAABAqv79Y3XyZJD694/N6VSyBUe6AQAAAAAeY8SIUG3blhK9AUU3AAAAAMBjrFsXYIpWR9ENAAAAAPAYiYmGKVodRTcAAAAAwGMkJ9tN0eq8YysAAAAAAF7hjjsSTNHqKLoBAAAAAB5jzx4/U7Q6im4AAAAAgMfw8zNHq6PoBgAAAAB4jMjIRFO0OopuAAAAAIDH+Osvf1O0OopuAAAAAIDHyJs32RStjqIbAAAAAOAxOKcbAAAAAAA3iYuzmaLVUXQDAAAAADzGq69ekN2eEr0BRTcAAAAAwGOMGZNLDkdK9AYU3QAAAAAAj3HqlI8pWp1vVhvOnTvXpRW0b9/epecBAAAAAG4/devGa8OGQNWtG5/TqWSLLBfdc+bMcWkFFN0AAAAAgKzq0+eS9uwJVJ8+l3I6lWyR5aJ71qxZpsfnzp3TsGHDVLx4cbVs2VJFihSRJB09elTff/+9jhw5ov/85z/Zmy0AAAAAwKu9+mqYzpxJiRs3ns7pdG6ay+d0f/755ypcuLBeeOEFlSlTRkFBQQoKClLZsmX1wgsvKCIiQpMmTcrOXAEAAAAAXu70aR9TtDqXi+6dO3eqcuXKGc6vUqWKduzY4eriAQAAAAC3oYQEmylanctFt5+fn3bv3p3h/L/++kt+fn6uLh4AAAAAcBsyDMMUrS7L53Rfq2HDhlq8eLGCg4P14IMPKiIiQpJ08uRJLV68WGvXrtWDDz6YbYkCAAAAALxfQIChK1dSojdwueju2rWrLl68qKVLl2rp0qWy21MOmjscDklSgwYN1LVr1+zJEgAAAABwW6hUKVG//uqjSpUSczqVbOFy0e3r66s+ffro4Ycf1pYtW3TmzBlJUoECBVS9enVFRkZmV44AAAAAgNvEH3/4maLVuVx0pypZsqRKliyZHbkAAAAAAG5zhmEzRau76aJ79+7d2rlzp86fP6/mzZurcOHCio+P19GjR1WkSBEFBgZmR54AAAAAgNuAw2GYotW5XHQnJSVp1KhR+uWXX5zTateurcKFC8tms2no0KFq2bKl2rZtmy2JAgAAAAC8X3Ky3RStzuWtmDlzpn799Vf17NlTo0aNMs3z9/dX3bp1TQU5AAAAAADXExl5xRStzuWie926dbr//vt17733KjQ0NM38okWL6tSpUzeVHAAAAADg9rJvX6ApWp3LRfeFCxdUokSJjBdstys+Pt7VxQMAAAAAbkP/3o3aGa3O5c3Ily+fjh49muH8v/76S4UKFXJ18QAAAACA25Cfn8MUrc7lorthw4ZatmyZdu/enWbesmXLtGHDBjVu3PimkgMAAAAA3F4iIhymaHUuX728bdu22rNnjwYOHKiiRYtKkqZOnarY2FidO3dONWrUUKtWrbItUQAAAACA93v11Yvq3z9cr756MadTyRYuF92+vr56/fXXtWbNGm3cuFEOh0NJSUkqWbKkHnnkETVu3Fg2m3fczBwAAAAAcGsMHRqmK1dSYnS09a9g7lLRnZCQoBkzZqhSpUpq3Lgxw8gBAAAAANni9Gm7KVqdS1vh7++vZcuW6fz589mdDwAAAADgNjZo0HnZ7SnRG7j800Hp0qV1+PDh7MwFAAAAAHCbW706UA5HSvQGLhfdTzzxhNatW6fly5crOTk5O3MCAAAAANymVq3yN0Wrc/lCauPHj5fdbtd///tfTZ48WeHh4fL3N78oNptNH3744U0neSvYLl+WzWG+JL1ht0uBgaY2GTFsNikoyLW2cXGSYWSQmPlidNdra1y1XMXFyZZRW0lGcLBrba9cSfNaudw2KOj/tzE+XrZMfsC5obaBgZL939+UEhJkS0rKnrYBAZKPj7OtlH7fSdM2MVG2xMSMl+vvL/n63njbpCTZ/s0j3bZ+fpKf3423TU6WLT4+47a+vlLq5/1G2jocsl3J+GIYN9TWx0cKCPj3gZHy2ciOtjfyub/JfURGn40b3UcYLrZlH+H+fUSW22blc5/62UhMzPwzxz4ipa0X7COy3PZ6f0eEhGS5LfuIf9tacR+RzX9H2OzXHJtjH5G1thbcR2T33xFp+k4mbZ05X/O5902Ik69sUoKRZls9bh+RFYaLBg4caAwaNOi6/yyjSBHDSOlizn9xTZsaR48edf5LDgpK0yb135V69Uxtk8LDM2wbX62aqW1isWIZtk0oV86QHP8+dBgJ5cpl2DaxWDHTcuOrVcuwbVJ4uKntlXr1MmybHBRkahvXtGmGbQ3J1PZyy5aZtj22Z4+z7aUOHTJte3z7dmfb2CeeyLTtiY0bnW0v9uqVaduTK1Y421546aVM25767jtn25j338+w7xiScXrOHGfbf4YOzXS5Z6ZOdbY99/HHmbY9O2GCs+3ZCRMybXvu44+dbc9MnZpp23+GDnW2PT1nTqZtY95809n21HffZdr2wksvOdueXLEi07YXe/Vytj2xcWOmbWOfeMLZ9vj27Zm2vdShg7PtsT17Mm17uWVLUx/OrK0r+4jjx48bhmEYSZUqZdj2RvcRV7dlH+FB+4g338y07Y3uI1L7zrnPPsu0LfuIlH9W3Ue44++I1L5z/Phx9hFevI/I9r8jihRx9hv2Ed69j8j2vyOu6TvevI84f/58lkpNl490Dxo0yNWnAgAAAABwW7AZhmHkdBKe4OT+/XJ46PDyImVLS7JJMnRs7/5M2zIs7N+2t2BYmD0pSYVKlEi371zbluHlDAtL/dzb7XYVKlQow35zdVvnchlenvLAYvuI7B46avf3V6FChXTi8GEZDC/32n2ES22v87m3hYSk9J0TJ2RcusQ+IittLbiPyO6/I+x2uyJKl9aJEydSvq/YR2StrQX3Edn9d0SavpNJW2fO13zuuzwSpk2/BOnuu+I0fXpMpm1zch8RGhqqsLCwDNulcvlItyQlJiZq+fLl2rp1q06dOiVJKliwoGrUqKGmTZumOcfbkxnBwTIyecNS29zI8rLc9urOm41tFRSkLP+iciNtAwPd0zYgwD1t/f1TvmDc0FbKWt+Rn1/Kl1FW3EhbX9+UL5jsbuvjk/U+fCNt7Xb3tLXZ3NNWbvzcZ6XfpLZlH5HCgvuILLe90X1E6h/i18M+IoVF9xFZbnudz/3VV4dhH+FCWyvuI7Lh7wjj2vNV2UfceFtZYx/hatuMPvdp+k4mbdMVGCj/vCG6rCD557XLCM74xx6P2EdkgctF99mzZzVkyBAdO3ZMefLkUaFChSRJBw8e1G+//aYlS5borbfeUr58+bItWQAAAACAd9u2zc8Urc7lonvSpEk6ffq0+vXrp7p165rmbdiwQePGjdOkSZM0YMCAm04SAAAAAHB7OH3abopW53LR/fvvv6tly5ZpCm5Jqlevng4cOKDFixffVHIAAAAAgNuLv7905cr/n6JvdS7/dBAUFKTcuXNnOD9PnjwKupFzAgAAAAAAtz273WGKVudy0d2kSROtWrVK8elcbfDKlStauXKlmjZtelPJAQAAAABuL5cv+5ii1bk8vDwyMlJbt25V3759FRUV5byQ2okTJ7R69WqFhoaqRIkS+vnnn03Pq1Onzs1lDAAAAADwWq1aXdK334aqVatLOZ1KtnC56B49erTz//Pnz08z/9y5c6Y2qWbNmuXqKgEAAAAAXm779gBTtDqXi+6BAwdmZx4AAAAAAOjMGbspWp3LRXfFihVvqH1CQoI2bNigmJgY5cmTx9XVAgAAAAC8WHCwocuXU6I3uGU/HVy+fFnjx4/XkSNHbtUqAQAAAAAWc+WKzRStzjuO1wMAAAAAvEKVKommaHUU3QAAAAAAj/H33z6maHUU3QAAAAAAj3Hhgo8pWp3LF1LLafPnz9emTZt09OhR+fv7q1y5curatauKFCmS06kBAAAAAFxUs2a8fvopSDVrxud0KtnCskX3H3/8oebNm6tMmTJKTk7WjBkzNGTIEH388ccKDAzM6fQAAAAAAC5ILee8payzbNH9xhtvmB737t1bTz31lPbv33/DtzMDAAAAAHiGvXt9TdHqXN6KPXv26I477sjOXG7K5cuXJUmhoaEZtklMTFRi4v9fAc/Hx0cBAQGSJJvNJpvN8y9Jb4Ucbxep74VV+g48A/0GrqLvwFX0HbiCfgNXZUffCQ5OluSn4OBkr+h/NsMwXLrjeKdOnVSoUCE1atRIjRo1UkRERKbtk5KStHv3bkVGRio4ONilZDPicDj0wQcf6NKlS3r33XczbDd79mzNnTvX+Tg6OlqdO3fO1lzc4ep+5tq7BQAAAADW4G31j8tF99q1a7VmzRpt375dDodD5cqVU6NGjVS/fv1Mjza7w8SJE/Xbb7/pnXfeUb58+TJsl9mR7pMnT8rhcLg9V1cUKVJIkk2SoWPHTuR0OviX3W5XRESER/cdeB76DVxF34Gr6DtwBf0GrsqOvmOV+ickJERhYWHXbefy8PKGDRuqYcOGunDhgtavX6+1a9dq0qRJmjp1qqpVq6bGjRurdu3a8vV17zj8SZMmacuWLRo8eHCmBbck+fn5yc/PL915hmHIxd8fbikr5Hi7SH0vrNJ34BnoN3AVfQeuou/AFfQbuCq7+4439L+brojDwsL0wAMP6IEHHtCJEye0du1arV27ViNHjlRwcLDq1q2rqKgo3XnnndmRr5NhGPrf//6nTZs2adCgQSpYsGC2Lh8AAAAAcOvZbCnDyr3gdG5J2Xz1cn9/fwUEBDiPJttsNm3evFkrVqxQ6dKl1bt3bxUrVixb1jVp0iStXbtWAwYMUFBQkGJiYiRJwcHB8vf3z5Z1AAAAAABurdy5kxUT46vcuZNzOpVscdNFd1xcnDZu3Ki1a9fqjz/+kM1mU/Xq1dW+fXvVqlVLdrtdmzZt0hdffKHx48frvffey4689cMPP0iSBg0aZJr+3HPPqUmTJtmyDgAAAADArRUTYzdFq3O56P7ll1+0Zs0abdmyRYmJiSpTpoyeeOIJNWjQQLly5TK1rVu3rmJjYzVp0qSbTjjV7Nmzs21ZAAAAAABPYbsmWpvLRfdHH32kfPnyqWXLloqKilKRIkUybR8ZGalGjRq5ujoAAAAAACzH5aL77bffVqVKlbLcvmzZsipbtqyrqwMAAAAA3AYCAw1duWJTYKD1r1wuSS4Pkr+RghsAAAAAgKwoWzbeFK3OO85MBwAAAAB4hT//DDBFq6PoBgAAAAB4jOBghylaHUU3AAAAAMBjXLjgY4pWR9ENAAAAAICbUHQDAAAAADxGcHCyKVodRTcAAAAAwGNcvuxjilZH0Q0AAAAAgJtQdAMAAAAAPEb+/MmmaHUU3QAAAAAAj1GxYqIpWh1FNwAAAADAY+za5WeKVkfRDQAAAADwGKdP+5ii1VF0AwAAAADgJhTdAAAAAAC4CUU3AAAAAMBj+Psbpmh1FN2WkHxNBAAAAADvFBzsMEWro+i2BJ9rIgAAAAB4p5gYH1O0OopuAAAAAIDH8PU1TNHqKLoBAAAAAB4jOdlmilZH0W0BQUFXTBEAAAAAvJVhmKPVUXRbQFxcoCkCAAAAAKyBohsAAAAA4DECA5NN0eoougEAAAAAHuPKFR9TtDqKbgAAAAAA3ISiGwAAAADgMfz9DVO0OopuAAAAAIDHyJs32RStjqIbAAAAAOAxfHxspmh1FN0AAAAAAI+RkGCOVkfRDQAAAADwGOfO+Zii1VF0AwAAAAA8hq+vwxStjqIbAAAAAOAxDMNmilZH0Q0AAAAA8BiJiTZTtDqKbgAAAACAxwgOTjZFq6PoBgAAAAB4jMuXfUzR6ii6AQAAAABwE4puAAAAAADchKIbAAAAAAA3oei2AH//ZFMEAAAAAFgDRbcFJCT4mCIAAAAAwBoougEAAAAAcBOKbgAAAACAx/DxMUzR6ii6AQAAAAAeI1cuhylaHUU3AAAAAMBjxMT4mKLVUXQDAAAAAOAmFN0AAAAAAI9hsxmmaHUU3QAAAAAAj5E7t8MUrY6iGwAAAADgMWJi7KZodd6xFQAAAAAAL2G7JlobRTcAAAAAAG5C0Q0AAAAA8CDGNdHafHM6ASB8ybO3dH3nHvj0lq4PAAAAwI1geDkAAAAAAMgCim4AAAAAgMfw8zNM0eoougEAAAAAHuPOOxNM0eooui0gV64rpggAAAAA3uqPP/xN0eq4kJoFXLwYaIrudqsvbAYAAAAAqXx9peTklOgNONINAAAAAPAg3nXLMIpuCwgMdJgiAAAAAHirpCSbKVodRbcFXLliN0UAAAAA8FbJyTZTtDqqOAAAAACAx/DzSzZFq6PotoCAgGRTBAAAAABvlZjoY4pWR9FtAalX7fOWq/cBAAAAwO2CotsCLl3yMUUAAAAA8FZ+foYpWh1FNwAAAADAYyQm2kzR6ii6AQAAAABwE4puAAAAAADchKLbEoxrIgAAAAB4J1/fZFO0OopuC7DZbKYIAAAAAN4qKcnHFK2OotsC8uVLMkUAAAAAgDVQdFsAR7oBAAAAwJooui3g9GkfUwQAAAAAWANFNwAAAAAAbkLRDQAAAACAm1B0AwAAAADgJhTdAAAAAAC4CUU3AAAAAMBjlCyZYIpWR9FtAal3CuOOYQAAAAC83d9/+5ui1VF0W8Djj180RQAAAACANVB0W8C0aaGmCAAAAADey7gmWhtFtwUkJ9tMEQAAAAC8l+2aaG0U3QAAAAAAD+K4Jlqbb04nANxq4UuezZ4F2WxSQIDyxMdLRsZDX8498Gn2rA8AAAC4LdividZG0Q0AAICb9sqifbd0fR8+XOaWrg8AXOUdPx14Pe+6kAAAAAAA3C440m0J3nUhAQCAdXD00rpS37sgfx993qOQ3vx+v+ISknM4KwC4/VB0W4CPT7KSk33l48MXpRVl2znkWcD544D3o3ACUtzKH4T4MQjAzaDotoCICEPHjqVEAACQPW71UXxYFyM+gFvNUMooX++ofyi6LYHh5QCA2wOFMECRD3gbim4LuHDBHAFPcSuHzksMn4dnulV/HKeelwsAgPfzroOOli+6lyxZom+++UYxMTEqWbKkevToobJly+Z0WtkqNtbHFIHbFUU+AADZL70fD73lAnwcxbemXLmSdPGin3LlSsrpVLKFpYvu9evX64svvlDPnj11xx136LvvvtPQoUM1atQo5c6dO6fTs6w/Tl7K6RTcqmJESE6n4Da3uigFAADwZAzVt6bLl31N0eosvRXffvutmjVrpnvuuUeS1LNnT23ZskUrV65UdHR0ziaXjXLnTtb5877Kndu6vzIC8CycNwsAQPajyM8eyckpF1JLidZn2aI7KSlJ+/fvNxXXdrtdVapU0e7du3MuMTeIj5cWPdJJdh9D4Uvi3b6+E25fQ87KriP5PjabqpUI0F8nLynZ8Iwdwq0+iu/toyI09fFsX6SPzaZCJcJ18tC5HO03/XNszXCVj80mTQ+X9GJOpwIAgJtxTrdHuHDhghwOh/LkyWOanidPHh07dizd5yQmJioxMdH52MfHRwEBAZIkm80mm80z39QrV3zkXzhGkvTHSY52e46rdwaeUXR7fRHsFTyv38AqUvpOn1ODlGw4cjgXWImPzS5Nz6s+p/6h71jEiPCBOZ2CTN9XnvknMiS98o3njVwL8vP993oABxSX6No52QXaHdXpr++SJI+t0W6EZYtuV8yfP19z5851Po6Ojlbnzp0lSRERETmV1nVVrCg9MHKpKlaUdu7M6WxwrWo5nQAsiX4DV9F34Cr6jnV8ldMJXOXzHvVyOgVY1M30HVsv5/9UuHDhbMknJ1m26A4LC5PdbldMTIxpekxMTJqj36natGmjVq1aOR/7+Pz/1cBPnjwph8Mzf/39449Ckmz64w9Dx497++Bv67Db7YqIiPDovgPPQ7+Bq+g7cBV9B66g38BV2dN3UuofybPrn5CQEIWFhV23nWWLbl9fX5UuXVo7duzQ3XffLUlyOBzasWOHHnjggXSf4+fnJz8/v3TnGYYhw0POy82MFXK8XaS+F1bpO/AM9Bu4ir4DV9F34Ar6DVyV3X3HG/qfZYtuSWrVqpXGjRun0qVLq2zZsvr+++8VHx+vJk2a5HRqAAAAAABYu+iuX7++Lly4oNmzZysmJkaRkZF6/fXXMxxeDgAAAADArWTpoluSHnjggQyHkwMAAAAArCblPt3ecqcXe04ngOsLDEw2RQAAAADwVna7zRStjqLbAq5c8TFFAAAAAPBWdrthilZH0Q0AAAAA8BhJSTZTtDqKbgvIlSvZFAEAAAAA1kDRbQGxsT6mCAAAAACwBopuSzCuiQAAAAAAK6DotgCbzWaKAAAAAOC9vOugI0W3BTgc5ggAAAAA3st2TbQ2im4AAAAAANyEohsAAAAAADeh6LYE7zqnAQAAAABuFxTdluBd5zQAAAAAwO2CohsAAAAAADeh6AYAAAAAwE0oui2Bc7oBAAAAwIooui2Bc7oBAAAAwIooui2gUaM4UwQAAAAAWANFtwWsWRNkigAAAAAAa6DoBgAAAADATSi6AQAAAAAexLsuJE3RbQE2m2GKAAAAAOC9vOtC0hTdFhAY6DBFAAAAAIA1UHRbQFycjykCAAAAAKyBotsCbDZzBAAAAABYA0W3BYSEJJsiAAAAAMAaKLotIDbWxxQBAAAAANZA0Q0AAAAAgJtQdAMAAAAA4CYU3RbAhdQAAAAAwJooui3AMMwRAAAAAGANFN0AAAAAALgJRTcAAAAAAG5C0Q0AAAAAgJtQdFtArlzJpggAAAAAsAaKbgvgQmoAAAAAYE0U3RYQG+tjigAAAAAAa6DoBgAAAADATSi6LSAkJNkUAQAAAADWQNFtAXFxPqYIAAAAALAGim4LMP69gprBldQAAAAAwFIoui0gf36HKQIAAAAArIGi2wIuXzZMEQAAAABgDRTdFnDpkq8pAgAAAACsgaIbAAAAAAA3oei2AF/fZFMEAAAAAFgDRbcFJCX5mCIAAAAAwBoougEAAAAAcBOKbgvw80s2RQAAAACANVB0W0Bioo8pAgAAAACsgaLbAqpXv2KKAAAAAABroOi2gN9+CzRFAAAAAIA1UHRbQKtWl0wRAAAAAGANFN0WsHRpsCkCAAAAAKyBotsCQkIcpggAAAAA+L/27j1Gyvre4/hnFpZFUASkLhfvFqwaEFtSIymCCqaaarW1a+IFqOlN1J6ktk3oaSLEVm01mp5GcoyaYqwhIeKlwWBsbbTYhhStgLciakArIrW4XAV2lzl/cNzTEQ+FLY+zz/p6JeSb2Zln9jfZ3x+893lmthxEdwls3NhQMwEAACgHFVcCDQ2VmgkAAEA5iO4SOPfcrTUTAACAchDdJbBwYf+aCQAAQDmI7hKYNm1zzQQAAKAcRHcJ/OEPB9VMAAAAykF0l8DOnZWaCQAAQDmI7hKoVqs1EwAAgHIQ3aVQ+dAEAACgDER3CfTtW62ZAAAAlIPoLoGvfW1rzQQAAKAcRHcJ3H77gJoJAABAOYjuErj99tb07bt7AgAAUB6iuwQuumhH3n9/9wQAAKA8RHcJfOtbh6RS2T0BAAAoD9FdAgsX9q+ZAAAAlIPoBgAAgIKIbgAAACiI6AYAAICCiG4AAAAoiOgugS99aWvNBAAAoBxEdwksWtSvZgIAAFAOorsEqtVKzQQAAKAcRHcJNDV11EwAAADKQXSXwPvv96qZAAAAlIPoLoGBAztqJgAAAOUgukvgy1/eVjMBAAAoB9FdAvfee0jNBAAAoBxENwAAABREdAMAAEBBRDcAAAAURHQDAABAQUQ3AAAAFER0AwAAQEFENwAAABREdAMAAEBBRDcAAAAUpHe9F9AV69evz4IFC/LCCy+ktbU1gwcPzoQJE/KVr3wlvXuX8iXt1cEHd2TLlt45+OCOei8FAACA/VDKQl27dm2q1Wq+9a1vZejQoXnzzTdz5513Zvv27Zk6dWq9l3fAbdnSq2YCAABQDqWM7rFjx2bs2LGdt5ubm7N27do8/vjjPTK6K5WkWt09AQAAKI8e857ubdu25eCDD673MgpRrdZOAAAAyqGUZ7o/bN26dVm0aFGuuOKKvT6ura0tbW1tnbd79eqVpqamJEmlUkmlBKeSy7DGT4oPfhZl2Tt0D/YNXWXv0FX2Dl1h39BVB3rv9IT9V6lWu8/50/vvvz+PPPLIXh9z++23Z8SIEZ23N2zYkOuvvz4nn3xyvvOd7+z12Pnz5+eBBx7ovH3hhRfm0ksv/fcW/TH4533WfX5aAAAAB15P659uFd2bNm3K5s2b9/qY5ubmzk8o37BhQ2bPnp2RI0dmxowZaWjY+9XyezvT/c4772TXrl3/5isoxvDhQ5NUklSzdu26ei+H/9XQ0JDm5uZuvXfofuwbusreoavsHbrCvqGrDsTeKUv/9O/fPwMGDPiXj+tWl5cPGDBgnxad/F9wH3vssfsU3EnS2NiYxsbGj7yvWq2mG/3+ocbw4e1Zu7Yxw4e3d9s1fhJ98LPoznuH7se+oavsHbrK3qEr7Bu66kDvnZ6w/0r5QWobNmzIrFmzMmTIkEydOjWbNm1Ka2trWltb6720Qvznf25O3767JwAAAOXRrc5076sVK1Zk3bp1Wbdu3R7v454/f36dVlWcH/5wQLZv3z0vvHB7vZcDAADAPipldE+aNCmTJk2q9zI+Nlu39qqZAAAAlEMpLy8HAACAMhDdJdC/f0fNBAAAoBxEdwm4vBwAAKCcRDcAAAAURHQDAABAQUQ3AAAAFER0AwAAQEFENwAAABREdAMAAEBBRDcAAAAURHQDAABAQUQ3AAAAFER0AwAAQEFENwAAABREdAMAAEBBRHcpdHxoAgAAUAaiuxR6fWgCAABQBqIbAAAACiK6AQAAoCCiGwAAAAoiugEAAKAgohsAAAAKIroBAACgIKIbAAAACiK6AQAAoCCiGwAAAAoiugEAAKAgohsAAAAKIroBAACgIKIbAAAACiK6AQAAoCCiGwAAAAoiugEAAKAgohsAAAAKIroBAACgIKIbAAAACiK6AQAAoCCiGwAAAAoiugEAAKAgohsAAAAKIroBAACgIKIbAAAACiK6AQAAoCCiGwAAAAoiukugUqnWTAAAAMpBdJfAoYfuqpkAAACUg+gugc2bG2omAAAA5aDiSmDQoI6aCQAAQDmI7hJoaqrUTAAAAMpBdJdAnz7VmgkAAEA5iO4S+O//bs0pp+yeAAAAlEfvei+Af2306PYsW5a8/XZ7qk52AwAAlIYz3QAAAFAQ0Q0AAAAFEd0AAAB0GyecsL1mlp3oBgAAoNtoa2usmWUnugEAAOg2Pv3ptppZdqIbAACAbuOgg2pn2YluAAAAuo1+/XbVzLIT3QAAAHQb06dvyymn7J49gegGAACg25gz5+AsX7579gSiGwAAgG7j5Zd718yyE90AAAB0G7/8ZWtOOWX37Al6xq8OAAAA6BFGj27PsmXJ22+3p1qt92r+fc50AwAAQEFENwAAABREdAMAANBtrFvXkP/4j92zJ+gZrwIAAIAe4Y47Ds5//dfu2RP4IDUAAAC6jauv3pL+/ftn+vQt9V7KAeFMNwAAAN3G0KG78otf7J49gegGAACAgohuAAAAKIjoBgAAgIKIbgAAACiI6AYAAICCiG4AAAAoiOgGAACAgohuAAAAKIjoBgAAgIKIbgAAACiI6AYAAICCiG4AAAAoiOgGAACAgohuAAAAKIjoBgAAgIKIbgAAACiI6AYAAICCiG4AAAAoiOgGAACAgohuAAAAKIjoBgAAgIKUPrrb2trygx/8IC0tLVm9enW9lwMAAACdSh/dv/71rzN48OB6LwMAAAD2UOrofu6557JixYpcccUV9V4KAAAA7KG00d3a2po777wz11xzTfr06VPv5QAAAMAeetd7AV1RrVYzZ86cTJkyJccff3zWr1+/T8e1tbWlra2t83avXr3S1NSUJKlUKqlUKoWs99/1wbq66/o+qf755+Jnw76yb+gqe4eusnfoCvuGrjoQe6en9U+3iu77778/jzzyyF4fc/vtt2f58uV5//33c9FFF+3X8z/00EN54IEHOm9feumlufDCC5Mkzc3N+73ej9vQoUPrvQQ+Qhn2Dt2PfUNX2Tt0lb1DV9g3dNWB2Ds9pX8q1Wq1Wu9FfGDTpk3ZvHnzXh/T3Nyc2267Lc8++2zNbz527dqVhoaGfOELX8g111zzkcd++Ex3Y2NjGhsbD8ziAQAA4EO6VXTvq3fffTfbtm3rvP3ee+/lpz/9ab73ve9l5MiROeyww+q4ugNvx44dWbBgQb761a92Xg4PAADQE/W0/inlB6kNGTIkRx11VOe/YcOGJdl9+UFPC+4k6ejoyMMPP5yOjo56LwUAAKBQPa1/ShndAAAAUAbd6oPUuurwww/P/Pnz670MAAAAqOFMdwk0Njbm4osv9qFvAABAj9fT+qeUH6QGAAAAZeBMNwAAABREdAMAAEBBRDcAAAAURHQDAABAQUQ3AAAAFKRH/J3uspo1a1aOPPLIJMkf/vCH9O7dO1OmTMkll1ySSqWSLVu2ZO7cuXn22WfT1taWk046KV//+tczbNiwJMnf//733HPPPVm5cmXa29vzqU99Kpdffnk++9nP1vNlAQAA7GHWrFk56qij0qdPnzzxxBOd/dPS0pIk2bp1a+67774sXbo07e3tOe644zJt2rQcc8wxnc+xYMGCLFq0KDt37sz48eNzyCGHZNmyZbnlllvq9Kr+NWe66+ypp55Kr169ctNNN2X69Ol59NFH88QTTyRJ5syZk9deey0//OEP85Of/CTVajU33XRT2tvbkyT33HNP2tvbM3v27Nx666257LLL0rdv33q+HAAAgP/XU089laamptx44425/PLLs2DBgqxYsSJJctttt2Xjxo350Y9+lJtvvjnHHntsbrjhhmzZsiVJsnjx4jz44IO57LLLcvPNN2fIkCF5/PHH6/ly9okz3XV22GGHZdq0aalUKhk+fHjeeOONPProozn55JPzzDPP5IYbbsgJJ5yQJPnud7+bq666KkuXLs3pp5+ed999N6eddlqOOuqoJElzc3M9XwoAAMBeHX300fna176WJBk2bFgee+yxPP/88+nTp09effXV3H333WlsbEySTJ06NUuXLs2SJUsyefLkPPbYYznrrLNy5plnJkkuvvjiLF++PNu3b6/b69kXorvORo4cmUql0nl71KhRWbhwYf72t7+lV69eGTlyZOd9hxxySIYPH5633norSXLuuefm7rvvzooVKzJ69OicdtppOfrooz/21wAAALAvPjhh+IFBgwZl48aNWb16dbZv354rr7yy5v6dO3dm3bp1SZK1a9fmnHPOqbn/05/+dF544YViF/1vEt0ldvbZZ+eUU07JX/7yl6xYsSIPPfRQpk6dmnPPPbfeSwMAANhD7957Jmi1Ws327dszaNCgzJo1a4/7+/Xr9zGsrDje011nr776as3tVatWZejQoTniiCPS0dGRVatWdd63efPmrF27NkcccUTn14YMGZJzzjkn3//+93P++ed3vh8cAACgLI477ri0tramoaEhQ4cOrfk3YMCAJMnw4cPz2muv1Rz34dvdkeius3fffTf33ntv1q5dm6effjqLFi3Keeedl2HDhmXcuHG5884789e//jWrV6/OL3/5ywwePDjjxo1LksydOzfLli3L+vXr8/rrr+fFF1/MiBEj6vyKAAAA9s/o0aMzatSo3HLLLVm+fHnWr1+flStXZt68eZ1h/cUvfjG///3v8+STT+btt9/OggULsmbNmpq363ZHLi+vszPOOCM7d+7MzJkz09DQkPPOOy+TJ09OksyYMSNz587NzTffnPb29px44omZOXNm5yUZu3btyj333JMNGzbkoIMOytixYzNt2rR6vhwAAID9VqlUMnPmzMybNy9z5szJpk2bMnDgwJx44ok59NBDkyQTJkzIO++8k/vuuy9tbW05/fTTM2nSpD2uHu5uKtVqtVrvRXxSzZo1K8ccc0ymT59e76UAAACUzg033JCBAwfm2muvrfdS/l8uLwcAAKDb27FjRxYuXJg333wzb731VubPn5/nn38+EydOrPfS9srl5QAAAHR7lUolzz33XB588MG0tbVl+PDhue666zJmzJh6L22vXF4OAAAABXF5OQAAABREdAMAAEBBRDcAAAAURHQDAABAQUQ3AAAAFER0AwAAQEFENwB8gm3YsCHz58/P6tWr670UAOiRRDcAfIK99957eeCBB0Q3ABREdAMA+2zHjh31XgIAlIroBoASeuGFF9LS0pI///nPe9z39NNPp6WlJa+88spen+PFF1/MzJkzkyRz5sxJS0tLWlpa8uSTTyZJZs2aleuuuy6vv/56rr/++lx++eWZN29ekqSlpSXz58/f4zmvvvrq3HHHHTVf27p1a+bOnZurrroql156aa699to8/PDD2bVrV1deOgCUSu96LwAA2H8nn3xyDjvssCxevDif//zna+5bvHhxmpubM2rUqL0+x4gRIzrjefLkyfnMZz6TJDnhhBM6H7N58+bceOONGT9+fCZMmJBDDz10v9a5Y8eOzJo1Kxs2bMjkyZMzZMiQrFy5MvPmzUtra2umT5++X88HAGUjugGghCqVSiZMmJBHH30027ZtS79+/ZIkmzZtyooVK3LRRRf9y+cYOHBgTj311MyfPz+jRo3KGWecscdjWltb881vfjNTpkzp0joXLlyYdevW5ec//3mGDRuWJJkyZUoGDx6c3/zmN/nSl76UIUOGdOm5AaAMXF4OACU1ceLEtLW1ZcmSJZ1f+9Of/pSOjo6PDOiuaGxszJlnntnl45csWZITTzwx/fv3z6ZNmzr/jR49Ort27crLL798QNYJAN2VM90AUFIjRozI8ccfn8WLF+ess85KsvvS8pEjR2bo0KEH5HsMHjw4vXt3/b8Lb7/9dtasWZNvfOMbH3n/xo0bu/zcAFAGohsASmzixIn51a9+lX/84x9pa2vLqlWrcuWVVx6w5+/Tp89+Pf7DH45WrVYzZsyYXHDBBR/5+OHDh3d5bQBQBqIbAEps/Pjxuffee/PHP/4xO3fuTK9evTJ+/Ph9Pr5SqXTp+/bv3z/btm2r+Vp7e3vee++9mq81Nzdn+/btGTNmTJe+DwCUnfd0A0CJDRgwIKeeemoWL16cxYsXZ+zYsRkwYMA+H9/U1JRk95/12h9Dhw7d4/3Yv/vd7/Y403366afnlVdeybJly/Z4jq1bt6ajo2O/vi8AlI0z3QBQcmeccUZuu+22JMkll1yyX8c2Nzenf//++e1vf5uDDjooTU1NGTlyZA4//PC9HnfWWWflrrvuyq233poxY8ZkzZo1Wb58eQ455JCax11wwQV55pln8rOf/SwTJ07Mcccdlx07duSNN97IkiVLcscdd+zXLwkAoGyc6QaAkhs3blz69++ffv36Zdy4cft1bO/evXP11VenoaEhd911V37xi1/kpZde+pfHnX322fnyl7+cl19+Offdd1/Wr1+fH//4x+nbt2/N45qamjJ79uycf/75eemllzJ37tw8/PDDWbduXVpaWjr/1BkA9FSVarVarfciAICu6+joyLe//e187nOfy1VXXVXv5QAA/8SZbgAouaVLl2bTpk2ZOHFivZcCAHyI93QDQEmtWrUqa9asyYIFC3LsscfmpJNO6ryvvb09W7Zs2evx/fr12+8/CQYA7B/RDQAl9fjjj2fx4sU55phjMmPGjJr7Vq5cmdmzZ+/1+BkzZmTSpEkFrhAA8J5uAOiBtmzZktdff32vjznyyCMzaNCgj2lFAPDJJLoBAACgID5IDQAAAAoiugEAAKAgohsAAAAKIroBAACgIKIbAAAACiK6AQAAoCCiGwAAAAoiugEAAKAg/wPvmVWs2c4/kQAAAABJRU5ErkJggg==", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "data = np.load('./data_seal.npz')\n", + "data1 = data['pos_pred']\n", + "data2 = data['neg_pred']\n", + "\n", + "y_pred = np.concatenate([data1, data2])\n", + "y_true = np.array(['pos'] * len(data1) + ['neg'] * len(data2))\n", + "hard_thres = (max(y_pred) + min(y_pred)) / 2\n", + "\n", + "# 创建主图和左右轴\n", + "fig, ax = plt.subplots(figsize=(10, 8))\n", + "ax_hist_top = ax.twiny()\n", + "ax_hist_bottom = ax.twiny()\n", + "\n", + "# 绘制散点图\n", + "ax.scatter(y_true, y_pred, color='blue', s=1, label='Predictions vs True Values')\n", + "\n", + "# 绘制阈值线\n", + "ax.axhline(y=hard_thres, color='red', linestyle='--', label=f'Hard Threshold: {hard_thres.item()}')\n", + "\n", + "# 设置标签和标题\n", + "ax.set_xlabel('y_true')\n", + "ax.set_ylabel('y_pred')\n", + "ax.set_title('Scatter Plot of Predictions vs True Values with Hard Threshold')\n", + "ax.legend()\n", + "\n", + "# 绘制顶部和底部的直方图\n", + "# 调整直方图范围和高度\n", + "hist_range = (min(y_pred), max(y_pred)) # 范围\n", + "hist_bins = 30 # 柱数\n", + "density = True # 归一化\n", + "\n", + "# 绘制超过和未超过硬阈值的直方图，分别使用不同的颜色\n", + "ax_hist_top.hist(data1, bins=hist_bins, range=hist_range, color='#1f77b4', alpha=0.7, density=density, label='Over Hard Threshold')\n", + "ax_hist_bottom.hist(data2, bins=hist_bins, range=hist_range, color='#ff7f0e', alpha=0.7, density=density, label='Below Hard Threshold')\n", + "\n", + "# 设置顶部和底部直方图的标签\n", + "ax_hist_top.set_xlabel('y_pred')\n", + "ax_hist_bottom.set_xlabel('y_pred')\n", + "\n", + "# 调整布局\n", + "plt.tight_layout()\n", + "plt.show()\n" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.8.18" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +}