Hugging Face's logo

Datasets:
hirundo-io
/
RoboShaul

Modalities:
Audio
Formats:
soundfolder
Size:
1K - 10K
Libraries:
Datasets
License:
Dataset card Data Studio Files
xet
Community
1
tomer-hirundo commited on
Commit
85e9e8e
·
1 Parent(s): bc642ed

removed the scripts

Browse files
Files changed (2) hide show
  1. corruption.ipynb +0 -125
  2. plot_roc.ipynb +0 -100
corruption.ipynb DELETED
@@ -1,125 +0,0 @@
1
- {
2
- "cells": [
3
- {
4
- "cell_type": "code",
5
- "execution_count": 2,
6
- "metadata": {},
7
- "outputs": [
8
- {
9
- "name": "stdout",
10
- "output_type": "stream",
11
- "text": [
12
- "Mislabelled metadata CSV saved to: mislabelled_meta.csv\n"
13
- ]
14
- }
15
- ],
16
- "source": [
17
- "import pandas as pd\n",
18
- "import random\n",
19
- "\n",
20
- "# Hebrew alphabet (22 letters)\n",
21
- "hebrew_alphabet = [\n",
22
- " \"א\",\n",
23
- " \"ב\",\n",
24
- " \"ג\",\n",
25
- " \"ד\",\n",
26
- " \"ה\",\n",
27
- " \"ו\",\n",
28
- " \"ז\",\n",
29
- " \"ח\",\n",
30
- " \"ט\",\n",
31
- " \"י\",\n",
32
- " \"כ\",\n",
33
- " \"ל\",\n",
34
- " \"מ\",\n",
35
- " \"נ\",\n",
36
- " \"ס\",\n",
37
- " \"ע\",\n",
38
- " \"פ\",\n",
39
- " \"צ\",\n",
40
- " \"ק\",\n",
41
- " \"ר\",\n",
42
- " \"ש\",\n",
43
- " \"ת\",\n",
44
- "]\n",
45
- "\n",
46
- "\n",
47
- "# Function to generate random Hebrew tokens\n",
48
- "def generate_random_hebrew_token():\n",
49
- " length = random.randint(2, 4) # Token length between 2 and 4\n",
50
- " return \"\".join(random.choices(hebrew_alphabet, k=length))\n",
51
- "\n",
52
- "\n",
53
- "# Load the dataset\n",
54
- "file_path = \"meta.csv\"\n",
55
- "meta_df = pd.read_csv(file_path)\n",
56
- "\n",
57
- "# Parameters\n",
58
- "N = 300 # Number of rows to corrupt\n",
59
- "\n",
60
- "# Randomly select N rows for corruption\n",
61
- "corruption_indices = random.sample(range(len(meta_df)), N)\n",
62
- "\n",
63
- "# Initialize corruption tracking columns\n",
64
- "meta_df[\"is_corrupted\"] = False\n",
65
- "meta_df[\"corruption_location\"] = None\n",
66
- "\n",
67
- "# Corrupt the selected rows\n",
68
- "for idx in corruption_indices:\n",
69
- " original_text = meta_df.at[idx, \"transcript\"]\n",
70
- " num_tokens_to_add = random.choice([1, 2]) # Randomly decide 1 or 2 tokens to add\n",
71
- " tokens_to_add = [generate_random_hebrew_token() for _ in range(num_tokens_to_add)]\n",
72
- "\n",
73
- " # Decide random locations for corruption\n",
74
- " insert_positions = sorted(\n",
75
- " random.sample(range(len(original_text) + 1), num_tokens_to_add)\n",
76
- " )\n",
77
- "\n",
78
- " # Corrupt the text\n",
79
- " corrupted_text = list(original_text)\n",
80
- " for pos, token in zip(insert_positions, tokens_to_add):\n",
81
- " corrupted_text.insert(pos, f\" {token} \")\n",
82
- " corrupted_text = \"\".join(corrupted_text)\n",
83
- "\n",
84
- " # Update the dataframe\n",
85
- " meta_df.at[idx, \"transcript\"] = corrupted_text\n",
86
- " meta_df.at[idx, \"is_corrupted\"] = True\n",
87
- " meta_df.at[idx, \"corruption_location\"] = str(insert_positions)\n",
88
- "\n",
89
- "# Save the mislabelled version\n",
90
- "output_path = \"mislabelled_meta.csv\"\n",
91
- "meta_df.to_csv(output_path, index=False)\n",
92
- "\n",
93
- "print(f\"Mislabelled metadata CSV saved to: {output_path}\")"
94
- ]
95
- },
96
- {
97
- "cell_type": "code",
98
- "execution_count": null,
99
- "metadata": {},
100
- "outputs": [],
101
- "source": []
102
- }
103
- ],
104
- "metadata": {
105
- "kernelspec": {
106
- "display_name": "Python (pixi-dev)",
107
- "language": "python",
108
- "name": "pixi-dev"
109
- },
110
- "language_info": {
111
- "codemirror_mode": {
112
- "name": "ipython",
113
- "version": 3
114
- },
115
- "file_extension": ".py",
116
- "mimetype": "text/x-python",
117
- "name": "python",
118
- "nbconvert_exporter": "python",
119
- "pygments_lexer": "ipython3",
120
- "version": "3.12.3"
121
- }
122
- },
123
- "nbformat": 4,
124
- "nbformat_minor": 2
125
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
plot_roc.ipynb DELETED
@@ -1,100 +0,0 @@
1
- {
2
- "cells": [
3
- {
4
- "cell_type": "code",
5
- "execution_count": 25,
6
- "metadata": {},
7
- "outputs": [],
8
- "source": [
9
- "import pandas as pd\n",
10
- "from sklearn.metrics import roc_curve, auc\n",
11
- "import matplotlib.pyplot as plt\n",
12
- "\n",
13
- "# Load the mislabelled meta file and the results CSV\n",
14
- "mislabelled_meta_path = \"mislabelled_meta.csv\"\n",
15
- "results_path = \"optimization/suspects dataframe.csv\"\n",
16
- "\n",
17
- "meta_df = pd.read_csv(mislabelled_meta_path)\n",
18
- "results_df = pd.read_csv(results_path)\n",
19
- "\n",
20
- "# Extract corrupted files from the mislabelled meta file\n",
21
- "corrupted_files = meta_df[meta_df[\"is_corrupted\"]][\"audio_segment_id\"].unique()\n",
22
- "\n",
23
- "# Flag suspects in the results dataframe as true or false positives\n",
24
- "results_df[\"is_corrupted\"] = (\n",
25
- " results_df[\"audio_segment_id\"].isin(corrupted_files).astype(int)\n",
26
- ")\n",
27
- "\n",
28
- "# Extract true labels (is_corrupted) and scores (suspect_score)\n",
29
- "y_true = results_df[\"is_corrupted\"]\n",
30
- "y_scores = results_df[\"suspect_score\"]\n",
31
- "\n",
32
- "# Calculate ROC curve and AUC\n",
33
- "fpr, tpr, thresholds = roc_curve(y_true, y_scores)\n",
34
- "roc_auc = auc(fpr, tpr)"
35
- ]
36
- },
37
- {
38
- "cell_type": "code",
39
- "execution_count": 26,
40
- "metadata": {},
41
- "outputs": [
42
- {
43
- "data": {
44
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAArMAAAIjCAYAAAAQgZNYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/GU6VOAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB/XElEQVR4nO3dd1hT1/8H8HcChCXgYImg4NY6wAHFPXDPrwPcuEddxYmj4rZuW0etVsVBnVVrnXXhwM1wi5M6QXGwlJnz+8OSnymgBIFL4P16Hh7Nybn3vpOTwIfDuTcyIYQAEREREZEWkksdgIiIiIgoq1jMEhEREZHWYjFLRERERFqLxSwRERERaS0Ws0RERESktVjMEhEREZHWYjFLRERERFqLxSwRERERaS0Ws0RERESktVjMEhEREZHWYjFLRJQOX19fyGQy1Zeuri5KlCiBvn374tmzZ+luI4TA5s2b0aBBAxQuXBhGRkaoWrUqZs6cibi4uAyPtWfPHrRq1Qrm5uZQKBSwsbGBu7s7Tpw4kams8fHxWLp0KVxcXGBmZgYDAwOUL18eI0aMwN27d7P0+ImItIVMCCGkDkFElNf4+vqiX79+mDlzJhwcHBAfH48LFy7A19cX9vb2uHHjBgwMDFT9U1JS0KNHD+zYsQP169dHp06dYGRkhDNnzuD3339H5cqVcezYMVhZWam2EUKgf//+8PX1hZOTE7p06QJra2u8ePECe/bsQWBgIAICAlCnTp0Mc0ZGRqJly5YIDAxE27Zt4ebmhkKFCiE0NBTbtm1DeHg4EhMTc/S5IiKSlCAiojQ2bNggAIjLly+rtU+cOFEAENu3b1drnzt3rgAgxo0bl2Zf+/btE3K5XLRs2VKtfeHChQKA+P7774VSqUyz3aZNm8TFixc/m7NNmzZCLpeLXbt2pbkvPj5ejB079rPbZ1ZSUpJISEjIln0REWUnLjMgItJA/fr1AQAPHjxQtX348AELFy5E+fLlMW/evDTbtGvXDp6enjh8+DAuXLig2mbevHmoWLEiFi1aBJlMlma73r17w9nZOcMsFy9exIEDBzBgwAB07tw5zf36+vpYtGiR6najRo3QqFGjNP369u0Le3t71e2wsDDIZDIsWrQIy5YtQ5kyZaCvr4/g4GDo6upixowZafYRGhoKmUyGFStWqNrevXuH77//HnZ2dtDX10fZsmUxf/58KJXKDB8TEZGmWMwSEWkgLCwMAFCkSBFV29mzZ/H27Vv06NEDurq66W7Xp08fAMD+/ftV27x58wY9evSAjo5OlrLs27cPwMeiNyds2LABy5cvx+DBg7F48WIUL14cDRs2xI4dO9L03b59O3R0dNC1a1cAwPv379GwYUNs2bIFffr0wc8//4y6deti0qRJGDNmTI7kJaKCKf3vukREBACIiopCZGQk4uPjcfHiRcyYMQP6+vpo27atqs+tW7cAANWrV89wP6n33b59W+3fqlWrZjlbduzjc54+fYr79+/DwsJC1ebh4YEhQ4bgxo0bqFKliqp9+/btaNiwoWpN8JIlS/DgwQMEBwejXLlyAIAhQ4bAxsYGCxcuxNixY2FnZ5cjuYmoYOHMLBHRZ7i5ucHCwgJ2dnbo0qULjI2NsW/fPtja2qr6xMTEAABMTEwy3E/qfdHR0Wr/fm6bL8mOfXxO586d1QpZAOjUqRN0dXWxfft2VduNGzdw69YteHh4qNp27tyJ+vXro0iRIoiMjFR9ubm5ISUlBadPn86RzERU8HBmlojoM1auXIny5csjKioK69evx+nTp6Gvr6/WJ7WYTC1q0/PfgtfU1PSL23zJp/soXLhwlveTEQcHhzRt5ubmaNq0KXbs2IFZs2YB+Dgrq6uri06dOqn63bt3D9euXUtTDKd6+fJltuclooKJxSwR0Wc4OzujVq1aAICOHTuiXr166NGjB0JDQ1GoUCEAQKVKlQAA165dQ8eOHdPdz7Vr1wAAlStXBgBUrFgRAHD9+vUMt/mST/eRemLa58hkMoh0rsaYkpKSbn9DQ8N027t164Z+/fohJCQEjo6O2LFjB5o2bQpzc3NVH6VSiWbNmmHChAnp7qN8+fJfzEtElBlcZkBElEk6OjqYN28enj9/rnbWfr169VC4cGH8/vvvGRaGmzZtAgDVWtt69eqhSJEi2Lp1a4bbfEm7du0AAFu2bMlU/yJFiuDdu3dp2v/55x+NjtuxY0coFAps374dISEhuHv3Lrp166bWp0yZMoiNjYWbm1u6XyVLltTomEREGWExS0SkgUaNGsHZ2RnLli1DfHw8AMDIyAjjxo1DaGgopkyZkmabAwcOwNfXFy1atMC3336r2mbixIm4ffs2Jk6cmO6M6ZYtW3Dp0qUMs7i6uqJly5b47bffsHfv3jT3JyYmYty4carbZcqUwZ07d/Dq1StV29WrVxEQEJDpxw8AhQsXRosWLbBjxw5s27YNCoUizeyyu7s7zp8/jyNHjqTZ/t27d0hOTtbomEREGeEngBERpSP1E8AuX76sWmaQateuXejatSt++eUXDB06FMDHP9V7eHjgjz/+QIMGDdC5c2cYGhri7Nmz2LJlCypVqoTjx4+rfQKYUqlE3759sXnzZtSoUUP1CWDh4eHYu3cvLl26hHPnzsHV1TXDnK9evULz5s1x9epVtGvXDk2bNoWxsTHu3buHbdu24cWLF0hISADw8eoHVapUQfXq1TFgwAC8fPkSq1evhpWVFaKjo1WXHQsLC4ODgwMWLlyoVgx/ys/PD7169YKJiQkaNWqkukxYqvfv36N+/fq4du0a+vbti5o1ayIuLg7Xr1/Hrl27EBYWprYsgYgoy6T9zAYiorwpo08AE0KIlJQUUaZMGVGmTBmRnJys1r5hwwZRt25dYWpqKgwMDMQ333wjZsyYIWJjYzM81q5du0Tz5s1F0aJFha6urihevLjw8PAQ/v7+mcr6/v17sWjRIlG7dm1RqFAhoVAoRLly5cTIkSPF/fv31fpu2bJFlC5dWigUCuHo6CiOHDkiPD09RalSpVR9Hj16JACIhQsXZnjM6OhoYWhoKACILVu2pNsnJiZGTJo0SZQtW1YoFAphbm4u6tSpIxYtWiQSExMz9diIiL6EM7NEREREpLW4ZpaIiIiItBaLWSIiIiLSWixmiYiIiEhrsZglIiIiIq3FYpaIiIiItBaLWSIiIiLSWrpSB8htSqUSz58/h4mJCWQymdRxiIiIiOg/hBCIiYmBjY0N5PLPz70WuGL2+fPnsLOzkzoGEREREX3BkydPYGtr+9k+Ba6YNTExAfDxyTE1Nc3x4ymVSrx69QoWFhZf/M2C8iaOofbjGGo/jqF24/hpv9wew+joaNjZ2anqts8pcMVs6tICU1PTXCtm4+PjYWpqyjewluIYaj+OofbjGGo3jp/2k2oMM7MklK8oIiIiItJaLGaJiIiISGuxmCUiIiIircViloiIiIi0FotZIiIiItJaLGaJiIiISGuxmCUiIiIircViloiIiIi0FotZIiIiItJaLGaJiIiISGuxmCUiIiIircViloiIiIi0FotZIiIiItJaLGaJiIiISGtJWsyePn0a7dq1g42NDWQyGfbu3fvFbfz9/VGjRg3o6+ujbNmy8PX1zfGcRERERJQ3SVrMxsXFoXr16li5cmWm+j969Aht2rRB48aNERISgu+//x4DBw7EkSNHcjgpEREREeVFulIevFWrVmjVqlWm+69evRoODg5YvHgxAKBSpUo4e/Ysli5dihYtWuRUTCIiIiLKoyQtZjV1/vx5uLm5qbW1aNEC33//fYbbJCQkICEhQXU7OjoaAKBUKqFUKnMk56eUSiWEEFAqlTh4/QWWHruHuITkHD8uZR+Bj+Mol8shkzoMZQnHUPtxDLUbx0+7Jb2PxpO/16NEs/4oYWWBfSPq5vgxNanRtKqYDQ8Ph5WVlVqblZUVoqOj8eHDBxgaGqbZZt68eZgxY0aa9levXiE+Pj7HsqZSKpWIioqCEAILD9/BP29z/phERERE2SH+6S1E7lsAkZwInYpNoGtQCC9fvszx48bExGS6r1YVs1kxadIkjBkzRnU7OjoadnZ2sLCwgKmpaY4fX6lUQiaTwcLCAgkptwAAchlgaaKf48em7MEZBe3HMdR+HEPtxvHTTikJHxCyexaMLEuhjMdk6JoUg5WpISwtLXP82AYGBpnuq1XFrLW1NSIiItTaIiIiYGpqmu6sLADo6+tDXz9t4SiXyyGX5875bzKZ7N9jfXwLW5oY4MLkprlybPp6SqUSL1++hKWlZa69Zih7cQy1H8dQu3H8tMurV69gYGAAExMTXO10Bt988w3kcnmujqEmx9CqV5SrqyuOHz+u1nb06FG4urpKlIiIiIgo/zh16hQcHR3h7e0NAKhevTp0dfP23KekxWxsbCxCQkIQEhIC4OOlt0JCQvD48WMAH5cI9OnTR9V/6NChePjwISZMmIA7d+5g1apV2LFjB7y8vKSIT0RERJQvpKSkYNasWWjSpAnKly+PKVOmSB0p0yQtta9cuYLGjRurbqeubfX09ISvry9evHihKmwBwMHBAQcOHICXlxd++ukn2Nra4rfffuNluYiIiIiyKDk5Ga1bt8axY8cwbdo0/PDDD9DR0ZE6VqZJWsw2atQIQogM70/v070aNWqE4ODgHExFREREVHDo6uqiSZMm8Pb2RpMmTaSOo7G8vQiCiIiIiLJdSkoKZs6ciaJFi2L06NGqNbLaSKtOACMiIiKir/P8+XM0bdoUs2fPzpVr7uc0zswSERERFRBHjhxBr169oFAocPLkSTRo0EDqSF+NM7NEREREBYAQAkuXLkWtWrUQEhKSLwpZgDOzRERERPnakydP8Pz5c7i4uGDnzp0wNjbOVx9ekX8eCRERERGpOXDgABwdHTFq1CgIIWBiYpKvClmAxSwRERFRvpOUlIRx48ahbdu2qFu3Lg4ePAiZTCZ1rBzBZQZERERE+Uzv3r3xxx9/YMmSJfj+++/zbSELcGaWiIiIKN9ISEgAAIwbNw4BAQHw8vLK14UswJlZIiIiIq2XkJCAiRMn4vLlyzh16hRq1aoldaRcw2KWiIiISIs9fPgQ7u7uuH79OhYvXgwdHR2pI+UqFrNEREREWmr37t3o168fLCwscO7cOdSsWVPqSLmOa2aJiIiItNTbt2/RsmVLBAYGFshCFmAxS0RERKRV7t27h4ULFwIA+vfvj23btsHMzEziVNJhMUtERESkJbZu3YoaNWrgt99+Q3R0NGQyWb6/WsGXsJglIiIiyuM+fPiAQYMGoUePHujQoQOuXLkCU1NTqWPlCTwBjIiIiCiPW7x4Mfz8/LBu3Tr069evwM/Gfoozs0RERER51IMHDwB8/BCEwMBA9O/fn4Xsf7CYJSIiIspj4uLi0LdvX1SpUgVPnz6FgYEBKlWqJHWsPInLDIiIiIjykBs3bsDd3R3//PMPfv31V9ja2kodKU/jzCwRERFRHrF//344OztDR0cHgYGB6NOnj9SR8jwWs0RERER5RPXq1TFkyBBcunQJFStWlDqOVmAxS0RERCShq1evokWLFnj37h3s7OywdOlSGBoaSh1La7CYJSIiIpKAEAKrV6+Gi4sLXr58iXfv3kkdSSuxmCUiIiLKZVFRUejWrRuGDRuGAQMG4Pz587C3t5c6llbi1QyIiIiIctm1a9dw9OhR7NixA127dpU6jlbjzCwRERFRLhBC4I8//oBSqUT9+vURFhbGQjYbsJglIiIiymFv375F586d0aVLFxw7dgwAYGpqKnGq/IHLDIiIiIhy0MWLF9GtWze8e/cOe/bsQfPmzaWOlK+wmCUiIiLKIUFBQahXrx5q1qyJkydP8iSvHMBlBkRERETZLD4+HgDg5OSEX3/9FadPn2Yhm0NYzBIRERFlo3PnzqFChQo4dOgQZDIZ+vfvD4VCIXWsfIvFLBEREVE2UCqVmD9/Pho0aAA7OztUqVJF6kgFAotZIiIioq/0+vVrtGnTBt7e3pgwYQL8/f1hZ2cndawCgSeAEREREX0lhUKB2NhYHD58GC1atJA6ToHCYpaIiIgoC1JSUrBw4UJ06dIFZcuWxenTpyGTyaSOVeCwmCUiIiLSUEREBHr27IkTJ06gWLFiKFu2LAtZibCYJSIiItLA8ePH0bNnTwDA0aNH0bRpU4kTFWw8AYyIiIgok969e4fOnTujSpUqCAkJYSGbB3BmloiIiOgLnj9/DlNTUxQuXBhnzpxB5cqVoaOjI3UsAmdmiYiIiD7ryJEjcHR0xNSpUwEAVatWZSGbh7CYJSIiIkpHcnIyJk+ejJYtW6JGjRqYMmWK1JEoHVxmQERERPQfiYmJcHNzw7lz5zBv3jxMmDABcjnnAPMiFrNERERE/6FQKNC6dWvMmzcPdevWlToOfQaLWSIiIiIASUlJmDx5MkqWLImRI0fC29tb6kiUCZwvJyIiogLvn3/+QYMGDbBs2TIolUqp45AGODNLREREBdrevXvRr18/mJmZ4cyZM/j222+ljkQa4MwsERERFVhCCKxYsQKNGjVCcHAwC1ktxJlZIiIiKnAePnyIyMhIODs7Y+/evTA2NoZMJpM6FmUBZ2aJiIioQNm1axecnJwwbtw4CCFQqFAhFrJajMUsERERFQjx8fEYPnw4unbtihYtWuCvv/5iEZsPcJkBERERFQjdunXD4cOHsWrVKgwdOpSFbD7BmVkiIiLK1+Lj4wEAU6ZMwYULFzBs2DAWsvkIZ2aJiIgoX/rw4QNGjx6NO3fu4OTJk6hdu7bUkSgHsJglIiKifOfOnTtwd3fHvXv3sGLFCsjl/GN0fsWRJSIionzFz88PNWvWRFJSEi5fvowBAwZwWUE+xmKWiIiI8pXY2Fh07doVV65cQZUqVaSOQzmMxSwRERFpvZs3b2LBggUAgMGDB8PX1xfGxsYSp6LcwGKWiIiItJYQAuvXr0ft2rWxefNmxMXFcUlBAcNiloiIiLRSbGwsevfujQEDBqBnz564ePEiZ2MLIF7NgIiIiLTSggUL8Oeff8LPzw89evSQOg5JhDOzREREpDWEELh79y4AYNKkSQgKCmIhW8CxmCUiIiKtEB0dje7du8PJyQnh4eEwNDREuXLlpI5FEuMyAyIiIsrzgoKC4O7ujlevXmHDhg2wtraWOhLlEZyZJSIiojztjz/+gKurKwoXLqwqaolSsZglIiKiPK1mzZr4/vvvERAQgDJlykgdh/IYFrNERESU51y6dAlubm6IioqCvb095s+fD319faljUR7EYpaIiIjyDCEEli5dinr16iEmJgaxsbFSR6I8jsUsERER5Qlv3rxBhw4dMGbMGIwaNQpnzpxBiRIlpI5FeRyvZkBERER5wtWrV3H+/Hn89ddfaNu2rdRxSEtwZpaIiIgko1QqsX37diiVSjRu3BiPHj1iIUsaYTFLREREknj16hXatm2Lbt264fTp0wCAQoUKSZyKtA2XGRAREVGuO3PmDLp164bExEQcOnQIjRo1kjoSaSnJZ2ZXrlwJe3t7GBgYwMXFBZcuXfps/2XLlqFChQowNDSEnZ0dvLy8EB8fn0tpiYiI6GtduHABjRo1QtmyZRESEoKWLVtKHYm0mKTF7Pbt2zFmzBj4+PggKCgI1atXR4sWLfDy5ct0+//+++/w9vaGj48Pbt++jXXr1mH79u2YPHlyLicnIiIiTX348AEA4OzsjA0bNuD48eO8WgF9NUmL2SVLlmDQoEHo168fKleujNWrV8PIyAjr169Pt/+5c+dQt25d9OjRA/b29mjevDm6d+/+xdlcIiIiktbZs2dRvnx5/P3335DL5ejTpw90dbnakb6eZK+ixMREBAYGYtKkSao2uVwONzc3nD9/Pt1t6tSpgy1btuDSpUtwdnbGw4cPcfDgQfTu3TvD4yQkJCAhIUF1Ozo6GsDHsyeVSmU2PZqMKZVKCCH+PZb4t1XkyrEpe6iPIWkjjqH24xhqr5SUFMyaNQuzZ89Go0aNUKVKFY6jFsrt96Amx5GsmI2MjERKSgqsrKzU2q2srHDnzp10t+nRowciIyNRr149CCGQnJyMoUOHfnaZwbx58zBjxow07a9evcqVtbZKpRJRUVEQQiDl34FJUSozXEpBec+nYyiXS77MnLKAY6j9OIbaKTIyEkOHDsX58+cxYsQIjB8/HnK5nD8DtVBuvwdjYmIy3Ver5vf9/f0xd+5crFq1Ci4uLrh//z5Gjx6NWbNm4Ycffkh3m0mTJmHMmDGq29HR0bCzs4OFhQVMTU1zPLNSqYRMJoOFhQV0/h18HbkclpaWOX5syh6fjiF/iGonjqH24xhqJwMDA+jo6ODo0aOoVKkSx0+L5fZ70MDAINN9JStmzc3NoaOjg4iICLX2iIgIWFtbp7vNDz/8gN69e2PgwIEAgKpVqyIuLg6DBw/GlClT0n1y9fX1oa+vn6ZdLpfn2htKJpP9eyxZagvfzFomdQw5btqLY6j9OIbaITk5GXPmzEHv3r1RunRp+Pv7QwiBly9fcvy0XG6+BzU5hmSvKIVCgZo1a+L48eOqNqVSiePHj8PV1TXdbd6/f5/mweno6AAAhBDpbUJERES55OnTp2jSpAlmzZqFgIAAAB8LIKKcJOkygzFjxsDT0xO1atWCs7Mzli1bhri4OPTr1w8A0KdPH5QoUQLz5s0DALRr1w5LliyBk5OTapnBDz/8gHbt2qmKWiIiIsp9Bw8eRJ8+fWBoaAh/f3/Uq1dP6khUQEhazHp4eODVq1eYNm0awsPD4ejoiMOHD6tOCnv8+LHaTOzUqVMhk8kwdepUPHv2DBYWFmjXrh3mzJkj1UMgIiIq8F6/fg0PDw80bNgQGzduRLFixaSORAWI5CeAjRgxAiNGjEj3Pn9/f7Xburq68PHxgY+PTy4kIyIios95/PgxihUrhmLFiuHChQuoVKkS18RSruMrjoiIiDS2b98+ODo6qi5/+c0337CQJUnwVUdERESZlpiYCC8vL3To0AENGzZU+/AjIilIvsyAiIiItEN8fDwaNmyI4OBg/PTTTxg5ciSvVkCSYzFLREREmWJgYID//e9/WLlyJWrVqiV1HCIALGaJiIjoM+Lj4zF+/HhUqlQJ3333Hby9vaWORKSGa2aJiIgoXffv30edOnWwdu1a6Opy/ovyJhazRERElMb27dtRo0YNxMbG4sKFCxg8eLDUkYjSxWKWiIiI1CiVSqxevRpt27ZFYGAgHB0dpY5ElCH+zYCIiIgAAHfu3EFMTAxq166N/fv3w8jIiFcroDyPM7NERESEzZs3o1atWpgyZQoAwNjYmIUsaQUWs0RERAXY+/fv0b9/f/Tp0wedO3fG7t27pY5EpBEuMyAiIirAOnfujNOnT2PDhg3o27ev1HGINMZiloiIqIARQuDDhw8wMjLCjBkzUKhQIVSuXFnqWERZwmKWiIioAImNjcWwYcPw5MkTnDhxAs7OzlJHIvoqLGaJiIgKiGvXrsHd3R1Pnz7Fr7/+Crmcp86Q9uOrmIiIqABYv349XFxcYGBggKCgIPTs2VPqSETZgsUsERFRARAfH4++ffvi/PnzKF++vNRxiLINlxkQERHlU0FBQTh27BgmTJiA7777Tuo4RDmCM7NERET5jBACK1asgKurK3bs2IEPHz5IHYkox7CYJSIiykfevXuHrl27YuTIkRgyZAgCAgJgaGgodSyiHMNlBkRERPnIvHnzcOzYMfzxxx/o1KmT1HGIchxnZomIiLScEAK3b98GAPj4+CAkJISFLBUYLGaJiIi02Js3b9CxY0c4OzsjMjISRkZGsLe3lzoWUa7hMgMiIiItdf78eXTr1g0xMTH4/fffYW5uLnUkolzHmVkiIiIt5OfnhwYNGqBEiRIICQlBu3btpI5EJAkWs0RERFrI1dUV3t7eOHXqFEqWLCl1HCLJsJglIiLSEmfOnEHjxo0RHR2N0qVLY9asWdDT05M6FpGkWMwSERHlcUqlEnPnzkXjxo2RkpKC9+/fSx2JKM9gMUtERJSHRUREoGXLlpg6dSomTZqEEydOwNraWupYRHkGr2ZARESUh127dg3Xr1/HkSNH0KxZM6njEOU5nJklIiLKY1JSUrBlyxYolUo0a9YMDx48YCFLlAEWs0RERHnIixcv4ObmBk9PT1y8eBEAYGRkJHEqoryLxSwREVEecfToUTg6OuLOnTs4fvw4XF1dpY5ElOexmCUiIsoDTp8+jRYtWsDR0RFXr15Fo0aNpI5EpBVYzBIREUkoLi4OAFCvXj1s2bIFhw4dgqWlpcSpiLQHi1kiIiKJHDx4EA4ODjhx4gTkcjl69OgBuZw/mok0wXcMERFRLktKSsKECRPQpk0b1K5dG9WqVZM6EpHW4nVmiYiIctHz58/RpUsXXL58GQsWLMDYsWM5G0v0FVjMEhER5SJjY2MYGBjg9OnTvFoBUTbgr4JEREQ5LDExEZMmTUJYWBjMzMxw4sQJFrJE2YQzs0RERDno0aNH8PDwQEhICJycnGBvby91JKJ8hTOzREREOWT37t1wcnJCZGQkAgIC4O7uLnUkonyHxSwREVEOiIiIQO/eveHm5oagoCDUrl1b6khE+RKXGRAREWWjBw8ewMbGBlZWVggMDESFChUgk8mkjkWUb3FmloiIKJts27YNTk5OmDNnDgCgYsWKLGSJchiLWSIioq/04cMHDB06FN27d0ebNm0wYcIEqSMRFRhcZkBERPQV3r9/jzp16iA0NBRr1qzBwIEDORtLlItYzBIREWWREAJGRkbo1q0bWrduzY+lJZIAlxkQERFp6P379+jfvz/Wrl0LAPD29mYhSyQRFrNEREQauHnzJmrXro1t27bB0NBQ6jhEBR6LWSIiokwQQmDDhg2q68VeuXIFvXv3ljgVEbGYJSIiygSlUon169eje/fuuHz5MipXrix1JCICTwAjIiL6rGvXriEhIQG1a9fGkSNHYGRkJHUkIvoEZ2aJiIjSIYTAmjVr4OzsjFmzZgEAC1miPIjFLBER0X9ER0ejR48eGDJkCPr27Yvt27dLHYmIMsBlBkRERP/Rvn17BAUFYdu2bfDw8JA6DhF9BotZIiIifFxWEBcXh0KFCmH+/PkoVqwYypYtK3UsIvoCFrNERFTgvXv3DgMHDsTbt29x7NgxuLi4SB2JiDLpq4rZ+Ph4GBgYZFcWIiKiXHf58mV4eHjgzZs3WL9+PWQymdSRiEgDGp8AplQqMWvWLJQoUQKFChXCw4cPAQA//PAD1q1bl+0BiYiIcsqKFStQt25dmJubIzg4GJ06dZI6EhFpSONidvbs2fD19cWCBQugUChU7VWqVMFvv/2WreGIiIhykhACI0aMwNmzZ+Hg4CB1HCLKAo2L2U2bNmHNmjXo2bMndHR0VO3Vq1fHnTt3sjUcERFRdrtw4QLmz58PABg5ciSWLFmiNjlDRNpF42L22bNn6Z7dqVQqkZSUlC2hiIiIsptSqcTChQtRv3597Nu3D4mJiVJHIqJsoHExW7lyZZw5cyZN+65du+Dk5JQtoYiIiLJTZGQk2rVrhwkTJmDMmDHw9/fnbCxRPqHx1QymTZsGT09PPHv2DEqlErt370ZoaCg2bdqE/fv350RGIiKirzJ37lxcvHgRBw4cQOvWraWOQ0TZSOOZ2Q4dOuCvv/7CsWPHYGxsjGnTpuH27dv466+/0KxZs5zISEREpDGlUokbN24AAGbOnImQkBAWskT5UJauM1u/fn0cPXo0u7MQERFli5cvX6J37964cOECwsLCUKRIERQqVEjqWESUAzSemS1dujRev36dpv3du3coXbp0toQiIiLKqpMnT6J69eoICQnBrl27UKRIEakjEVEO0riYDQsLQ0pKSpr2hIQEPHv2LFtCERERZcX69evh5uaGSpUqISQkhMvfiAqATC8z2Ldvn+r/R44cgZmZmep2SkoKjh8/Dnt7+2wNR0RElBlCCMhkMjRo0AAzZ86Et7e32rXQiSj/ynQx27FjRwCATCaDp6en2n16enqwt7fH4sWLszUcERHRlxw9ehSzZs3CgQMHULZsWUyZMkXqSESUizK9zECpVEKpVKJkyZJ4+fKl6rZSqURCQgJCQ0PRtm3bnMxKRESkkpycjKlTp6JFixYwNDREQkKC1JGISAIaX83g0aNHOZGDiIgo0549e4bu3bvj3LlzmDNnDiZOnAi5XOPTQIgoH8jSpbni4uJw6tQpPH78OM3HAY4aNUqjfa1cuRILFy5EeHg4qlevjuXLl8PZ2TnD/u/evcOUKVOwe/duvHnzBqVKlcKyZct47UAiogLk+vXrCAsLg7+/P+rVqyd1HCKSkMbFbHBwMFq3bo33798jLi4ORYsWRWRkJIyMjGBpaalRMbt9+3aMGTMGq1evhouLC5YtW4YWLVogNDQUlpaWafonJiaiWbNmsLS0xK5du1CiRAn8888/KFy4sKYPg4iItExSUhJ8fX3Rr18/tGzZEnfv3oWBgYHUsYhIYhr/TcbLywvt2rXD27dvYWhoiAsXLuCff/5BzZo1sWjRIo32tWTJEgwaNAj9+vVD5cqVsXr1ahgZGWH9+vXp9l+/fj3evHmDvXv3om7durC3t0fDhg1RvXp1TR8GERFpkcePH6Nz584YMmQIAgMDAYCFLBEByMLMbEhICH799VfI5XLo6OggISEBpUuXxoIFC+Dp6YlOnTplaj+JiYkIDAzEpEmTVG1yuRxubm44f/58utvs27cPrq6uGD58OP78809YWFigR48emDhxYoaXYElISFA7KSA6OhrA/5/QltOUSiWEEP8eS/zbKnLl2JQ91MeQtBHHULv99ddf6NevH4yMjHDixAnUqFGDY6ll+B7Ufrk9hpocR+NiVk9PT7XI3tLSEo8fP0alSpVgZmaGJ0+eZHo/kZGRSElJgZWVlVq7lZUV7ty5k+42Dx8+xIkTJ9CzZ08cPHgQ9+/fx3fffYekpCT4+Piku828efMwY8aMNO2vXr1CfHx8pvNmlVKpRFRUFIQQSPl3YFKUSrx8+TLHj03Z49Mx5Akm2oljqL1Onz4NDw8PtGjRAj4+PihVqhS/f2ohvge1X26PYUxMTKb7alzMOjk54fLlyyhXrhwaNmyIadOmITIyEps3b0aVKlU03Z1GlEolLC0tsWbNGujo6KBmzZp49uwZFi5cmGExO2nSJIwZM0Z1Ozo6GnZ2drCwsICpqWmO5k3NLJPJYGFhAZ1/B19HLk93TTDlTZ+OIb8JayeOofaJiYmBiYkJOnXqhO3bt+N///sfIiMjOYZaiu9B7ZfbY6jJMiKNi9m5c+eqquU5c+agT58+GDZsGMqVK4d169Zlej/m5ubQ0dFBRESEWntERASsra3T3aZ48eLQ09NTW1JQqVIlhIeHIzExEQqFIs02+vr60NfXT9Mul8tz7Q0lk8n+PZYstYVvZi2TOoYcN+3FMdQeu3fvxqBBg7B79240bNgQ7u7uqh+kHEPtxfHTfrk5hpocQ+M0tWrVQuPGjQF8XGZw+PBhREdHIzAwEI6Ojpnej0KhQM2aNXH8+HFVm1KpxPHjx+Hq6pruNnXr1sX9+/fV1lHcvXsXxYsXT7eQJSIi7ZGQkICRI0eic+fOaNy4MU/uJaJMybbSOigoSONPABszZgzWrl2LjRs34vbt2xg2bBji4uLQr18/AECfPn3UThAbNmwY3rx5g9GjR+Pu3bs4cOAA5s6di+HDh2fXwyAiIgk8fvwYderUwZo1a7By5Urs3LmTl10kokzRaJnBkSNHcPToUSgUCgwcOBClS5fGnTt34O3tjb/++gstWrTQ6OAeHh549eoVpk2bhvDwcDg6OuLw4cOqk8IeP36sNs1sZ2eHI0eOwMvLC9WqVUOJEiUwevRoTJw4UaPjEhFR3mJqaopixYrhwoULcHJykjoOEWmRTBez69atw6BBg1C0aFG8ffsWv/32G5YsWYKRI0fCw8MDN27cQKVKlTQOMGLECIwYMSLd+/z9/dO0ubq64sKFCxofh4iI8pYPHz5g6tSpGD16NEqWLIm///5b6khEpIUyvczgp59+wvz58xEZGYkdO3YgMjISq1atwvXr17F69eosFbJERFQwhYaG4ttvv8WqVasQHBwsdRwi0mKZLmYfPHiArl27AgA6deoEXV1dLFy4ELa2tjkWjoiI8h8/Pz/UrFkTCQkJuHjxIjp06CB1JCLSYpkuZj98+AAjIyMAHy/NoK+vj+LFi+dYMCIiyn+ePXuGQYMGoVOnTrhy5QqqVasmdSQi0nIanQD222+/oVChQgCA5ORk+Pr6wtzcXK3PqFGjsi8dERHlC6GhoShVqhRKlCiBa9euoUyZMpDJZF/ekIjoCzJdzJYsWRJr165V3ba2tsbmzZvV+shkMhazRESkxtfXF9999x28vb0xbdo0lC1bVupIRJSPZLqYDQsLy8EYRESU38TGxmL48OHYtGkT+vfvj3HjxkkdiYjyIY0/zpaIiOhLYmJi4OzsjCdPnmDz5s3o1auX1JGIKJ9iMUtERNlGCAEAMDExQb9+/dC+fXtUrFhR4lRElJ9l28fZEhFRwRYdHY0ePXpg/fr1AIAJEyawkCWiHMdiloiIvlpwcDBq1qyJAwcOwNTUVOo4RFSAsJglIqIsE0Jg1apV+Pbbb2FqaoqgoCDVB+wQEeWGLBWzDx48wNSpU9G9e3e8fPkSAHDo0CHcvHkzW8MREVHelpKSAj8/PwwePBjnzp3jZbeIKNdpXMyeOnUKVatWxcWLF7F7927ExsYCAK5evQofH59sD0hERHnP5cuXceXKFejq6uL48eNYvnw59PX1pY5FRAWQxsWst7c3Zs+ejaNHj0KhUKjamzRpggsXLmRrOCIiyluEEFi2bBnq1q2L+fPnAwAMDAwkTkVEBZnGxez169fxv//9L027paUlIiMjsyUUERHlPW/evMH//vc/eHl5YcSIEfDz85M6EhGR5teZLVy4MF68eAEHBwe19uDgYJQoUSLbghERUd4hhECbNm0QGhqKP//8E+3bt5c6EhERgCwUs926dcPEiROxc+dOyGQyKJVKBAQEYNy4cejTp09OZCQiIokolUrExcXBxMQEy5Ytg7W1NUqVKiV1LCIiFY2XGcydOxcVK1aEnZ0dYmNjUblyZTRo0AB16tTB1KlTcyIjERFJIDIyEu3bt0eXLl0ghICLiwsLWSLKczSemVUoFFi7di1++OEH3LhxA7GxsXByckK5cuVyIh8REUng7Nmz6NatG+Lj47Fp0ybIZDKpIxERpUvjYvbs2bOoV68eSpYsiZIlS+ZEJiIiktCiRYvg7e0NV1dXbN26Fba2tlJHIiLKkMbLDJo0aQIHBwdMnjwZt27dyolMREQkIT09PUycOBEnT55kIUtEeZ7Gxezz588xduxYnDp1ClWqVIGjoyMWLlyIp0+f5kQ+IiLKBf7+/qrrxo4ePRpz5syBrq7Gf7wjIsp1Ghez5ubmGDFiBAICAvDgwQN07doVGzduhL29PZo0aZITGYmIKIekpKRg5syZaNq0Kf7++28kJSVJHYmISCMaF7OfcnBwgLe3N3788UdUrVoVp06dyq5cRESUw8LDw9G8eXNMnz4dP/zwA/7++2/o6elJHYuISCNZ/htSQEAA/Pz8sGvXLsTHx6NDhw6YN29edmYjIqIcNHv2bNy6dQvHjh3jX9aISGtpPDM7adIkODg4oEmTJnj8+DF++uknhIeHY/PmzWjZsmVOZCQiomySnJyMa9euAQB+/PFHhISEsJAlIq2m8czs6dOnMX78eLi7u8Pc3DwnMhERUQ549uwZevTogevXryMsLAympqYoVKiQ1LGIiL6KxsVsQEBATuQgIqIcdPjwYfTu3RsKhQJ79+6Fqamp1JGIiLJFporZffv2oVWrVtDT08O+ffs+27d9+/bZEoyIiLLHqlWrMHz4cLRq1QobN26EhYWF1JGIiLJNporZjh07Ijw8HJaWlujYsWOG/WQyGVJSUrIrGxERfQUhBGQyGZo1a4ZFixbBy8sLcvlXXcSGiCjPydR3NaVSCUtLS9X/M/piIUtElDf89ddfqF+/PmJjY1GuXDmMHTuWhSwR5Usaf2fbtGkTEhIS0rQnJiZi06ZN2RKKiIiyJjExEWPHjkX79u1RrFgxJCcnSx2JiChHaVzM9uvXD1FRUWnaY2Ji0K9fv2wJRUREmnv06BHq16+P5cuXY8mSJdi7dy8KFy4sdSwiohyl8dUMUtdg/dfTp09hZmaWLaGIiEhzN2/eRGRkJM6ePQtnZ2ep4xAR5YpMF7NOTk6QyWSQyWRo2rQpdHX/f9OUlBQ8evSIH5pARJTLEhISsGXLFvTv3x9t27ZFs2bNoK+vL3UsIqJck+liNvUqBiEhIWjRooXahbYVCgXs7e3RuXPnbA9IRETpe/DgATw8PHD9+nU4OzujatWqLGSJqMDJdDHr4+MDALC3t4eHhwcMDAxyLBQREX3ejh07MHDgQFhaWuL8+fOoWrWq1JGIiCSh8Qlgnp6eLGSJiCR04MABeHh4oHXr1ggKCkKNGjWkjkREJJlMzcwWLVoUd+/ehbm5OYoUKZLuCWCp3rx5k23hiIjo/0VFRcHMzAwtW7bEnj170KFDh89+PyYiKggyVcwuXboUJiYmqv/zmycRUe7y8/PD8OHDcfDgQdSpU+ezn8ZIRFSQZKqY9fT0VP2/b9++OZWFiIj+4/379xg1ahTWrVuHXr16oVq1alJHIiLKUzReMxsUFITr16+rbv/555/o2LEjJk+ejMTExGwNR0RUkD18+BAuLi74/fffsW7dOmzatEntSjJERJSFYnbIkCG4e/cugI/faD08PGBkZISdO3diwoQJ2R6QiKigKly4MGxtbXH58mX079+fS7yIiNKhcTF79+5dODo6AgB27tyJhg0b4vfff4evry/++OOP7M5HRFSgxMXFYeTIkXj69CmKFi2KQ4cO4ZtvvpE6FhFRnqVxMSuEgFKpBAAcO3YMrVu3BgDY2dkhMjIye9MRERUg169fR61atbBhwwbcuHFD6jhERFpB42K2Vq1amD17NjZv3oxTp06hTZs2AIBHjx7Bysoq2wMSEeV3Qgj89ttvcHZ2hp6eHq5cucKPByciyiSNi9lly5YhKCgII0aMwJQpU1C2bFkAwK5du1CnTp1sD0hElN89fvwYI0eOhKenJy5evIiKFStKHYmISGtk+uNsU1WrVk3tagapFi5cCB0dnWwJRURUENy4cQPlypVDqVKlcOvWLTg4OEgdiYhI62g8M5sqMDAQW7ZswZYtWxAUFAQDAwPo6ellZzYionxJCIFffvkFtWrVwtKlSwGAhSwRURZpPDP78uVLeHh44NSpUyhcuDAA4N27d2jcuDG2bdsGCwuL7M5IRJRvREVFYdCgQdi5cyeGDx+O77//XupIRERaTeOZ2ZEjRyI2NhY3b97Emzdv8ObNG9y4cQPR0dEYNWpUTmQkIsoX3r17hxo1auDvv//Grl27sGLFChgYGEgdi4hIq2k8M3v48GEcO3YMlSpVUrVVrlwZK1euRPPmzbM1HBFRfiCEgEwmQ+HChTF06FB07twZpUuXljoWEVG+oPHMrFKpTHdtrJ6enur6s0RE9NHbt2/RqVMn+Pr6AgDGjx/PQpaIKBtpXMw2adIEo0ePxvPnz1Vtz549g5eXF5o2bZqt4YiItNnFixfh5OQEf39/FC1aVOo4RET5ksbF7IoVKxAdHQ17e3uUKVMGZcqUgYODA6Kjo7F8+fKcyEhEpFWEEFi8eDHq1auH4sWLIyQkBO3bt5c6FhFRvqTxmlk7OzsEBQXh+PHjuH37NgCgUqVKcHNzy/ZwRETaKDk5Gbt27YKXlxfmzJnDyxYSEeUgjYrZ7du3Y9++fUhMTETTpk0xcuTInMpFRKR1AgICYGBggJo1a+LUqVNQKBRSRyIiyvcyvczgl19+Qffu3XHlyhXcu3cPw4cPx/jx43MyGxGRVlAqlfjxxx/RsGFDLFu2DABYyBIR5ZJMF7MrVqyAj48PQkNDERISgo0bN2LVqlU5mY2IKM97+fIlWrdujcmTJ2PixInYsGGD1JGIiAqUTBezDx8+hKenp+p2jx49kJycjBcvXuRIMCKivE4IgVatWiEoKAiHDx/GnDlzoKur8akIRET0FTL9XTchIQHGxsaq23K5HAqFAh8+fMiRYEREeVVKSgri4uJgamqKVatWwc7ODjY2NlLHIiIqkDSaQvjhhx9gZGSkup2YmIg5c+bAzMxM1bZkyZLsS0dElMeEh4ejV69e0NfXx4EDB+Di4iJ1JCKiAi3TxWyDBg0QGhqq1lanTh08fPhQdVsmk2VfMiKiPOb48ePo2bMnZDIZ/Pz8pI5DRETQoJj19/fPwRhERHnbrFmz4OPjg6ZNm2LLli2wsrKSOhIRESELnwBGRFQQGRsbY9asWThy5AgLWSKiPISn3RIRZeDIkSO4evUqJkyYgDFjxkgdh4iI0sGZWSKi/0hOTsakSZPQsmVLnDp1CikpKVJHIiKiDLCYJSL6xJMnT9CoUSMsXLgQ8+fPx19//QUdHR2pYxERUQa4zICI6BOzZ8/G48ePcfr0adSpU0fqOERE9AVZmpk9c+YMevXqBVdXVzx79gwAsHnzZpw9ezZbwxER5YakpCSEhIQAABYuXIjg4GAWskREWkLjYvaPP/5AixYtYGhoiODgYCQkJAAAoqKiMHfu3GwPSESUk8LCwlC/fn00a9ZM9alexYoVkzoWERFlksbF7OzZs7F69WqsXbsWenp6qva6desiKCgoW8MREeWkvXv3wsnJCREREThw4IDaR3YTEZF20LiYDQ0NRYMGDdK0m5mZ4d27d9mRiYgoxy1duhT/+9//0LhxYwQHB8PZ2VnqSERElAUaF7PW1ta4f/9+mvazZ8+idOnSWQqxcuVK2Nvbw8DAAC4uLrh06VKmttu2bRtkMhk6duyYpeMSUcEjhAAAtGrVCsuXL8cff/yBwoULSxuKiIiyTONidtCgQRg9ejQuXrwImUyG58+fw8/PD+PGjcOwYcM0DrB9+3aMGTMGPj4+CAoKQvXq1dGiRQu8fPnys9uFhYVh3LhxqF+/vsbHJKKCadeuXahXrx7ev3+PihUrYsSIEZDJZFLHIiKir6BxMevt7Y0ePXqgadOmiI2NRYMGDTBw4EAMGTIEI0eO1DjAkiVLMGjQIPTr1w+VK1fG6tWrYWRkhPXr12e4TUpKCnr27IkZM2ZkeTaYiAqO+Ph4TJo0CR4eHrC1teWHIBAR5SMaX2dWJpNhypQpGD9+PO7fv4/Y2FhUrlwZhQoV0vjgiYmJCAwMxKRJk1Rtcrkcbm5uOH/+fIbbzZw5E5aWlhgwYADOnDnz2WMkJCSorrgAANHR0QAApVIJpVKpcWZNKZVKCCH+PZb4t1XkyrEpe6iPIWmbe/fuwcPDA3fu3MHKlSsxZMgQyGQyjqeW4ftQu3H8tF9uj6Emx8nyhyYoFApUrlw5q5sDACIjI5GSkgIrKyu1disrK9y5cyfdbc6ePYt169aprgn5JfPmzcOMGTPStL969Qrx8fEaZ9aUUqlEVFQUhBBI+XdgUpTKLy6joLzj0zGUy/mhedrm/PnziI2NxdatW+Hi4oJXr15JHYmygO9D7cbx0365PYYxMTGZ7qtxMdu4cePPrjE7ceKEprvMtJiYGPTu3Rtr166Fubl5praZNGkSxowZo7odHR0NOzs7WFhYwNTUNKeiqiiVSshkMlhYWEDn38HXkcthaWmZ48em7PHpGPKbsHZ4//49Nm/ejMGDB6NPnz5wd3dHVFQUx1CL8X2o3Th+2i+3x9DAwCDTfTUuZh0dHdVup35yzo0bN+Dp6anRvszNzaGjo4OIiAi19oiICFhbW6fp/+DBA4SFhaFdu3aqttRpaF1dXYSGhqJMmTJq2+jr60NfXz/NvuRyea69oWQy2b/HSv0lQMY3s5ZJHUOOW953+/ZtuLu748GDB2jUqBEqVaoEAwMDREdHcwy1HN+H2o3jp/1ycww1OYbGxezSpUvTbZ8+fTpiY2M12pdCoUDNmjVx/Phx1eW1lEoljh8/jhEjRqTpX7FiRVy/fl2tberUqYiJicFPP/0EOzs7jY5PRPnLpk2bMGzYMJQqVQqXLl1CpUqVpI5EREQ5LMtrZv+rV69ecHZ2xqJFizTabsyYMfD09EStWrXg7OyMZcuWIS4uDv369QMA9OnTByVKlMC8efNgYGCAKlWqqG2fen3I/7YTUcGyZ88eeHp6om/fvlixYgU/zYuIqIDItmL2/PnzGq1vSOXh4YFXr15h2rRpCA8Ph6OjIw4fPqw6Kezx48f8kwQRZejt27coUqQI2rdvj/3796NNmzZSRyIiolykcTHbqVMntdtCCLx48QJXrlzBDz/8kKUQI0aMSHdZAQD4+/t/dltfX98sHZOItJsQAuvXr4eXlxeOHj0KFxcXFrJERAWQxsWsmZmZ2m25XI4KFSpg5syZaN68ebYFIyLKSExMDIYNGwY/Pz8MGjQI1apVkzoSERFJRKNiNiUlBf369UPVqlVRpEiRnMpERJSh0NBQtG/fHs+fP8fvv/+O7t27Sx2JiIgkpNFiVB0dHTRv3hzv3r3LoThERJ9nbm6O8uXLIzAwkIUsERFpVswCH68a8PDhw5zIQkSUrqioKAwZMgTPnz9HsWLF8Ndff6F8+fJSxyIiojxA42J29uzZGDduHPbv348XL14gOjpa7YuIKDsFBgaiRo0a2LZtW4Yfc01ERAVXpovZmTNnIi4uDq1bt8bVq1fRvn172NraokiRIihSpAgKFy7MdbRElG2EEFi+fDnq1KmDIkWKICgoCE2aNJE6FhER5TGZPgFsxowZGDp0KE6ePJmTeYiIAHz8+Orx48dj2LBhmD9/frofS01ERJTpYlYIAQBo2LBhjoUhIgoODsY333yDsmXLIjQ0FKVKlZI6EhER5WEarZmVyWQ5lYOICjghBJYsWQJnZ2esWLECAFjIEhHRF2l0ndny5ct/saB98+bNVwUiooLn9evX6Nu3L/bv349x48Zh5MiRUkciIiItoVExO2PGjDSfAEZE9DUiIyNRo0YNxMXFYf/+/fxIWiIi0ohGxWy3bt1gaWmZU1mIqAARQkAmk8Hc3ByjRo2Ch4cH7OzspI5FRERaJtNrZrleloiyy6tXr9C6dWts3rwZADBu3DgWskRElCWZLmZTr2ZARPQ1Tp8+DUdHRwQGBsLKykrqOEREpOUyXcwqlUouMSCiLEtJScHs2bPRuHFjlCtXDiEhIWjevLnUsYiISMtp/HG2RERZkZKSgv3792Pq1Kk4duwYbGxspI5ERET5gEYngBERaer48eMoWrQonJyccObMGejp6UkdiYiI8hHOzBJRjkhJSYGPjw+aNWuGVatWAQALWSIiynacmSWibPf8+XP07NkTp0+fxowZMzB58mSpIxERUT7FYpaIspVSqUTz5s3x9u1bnDhxAg0bNpQ6EhER5WMsZokoWyQnJ+P9+/cwNTXFunXrULp0aVhYWEgdi4iI8jmumSWir/b06VM0btwYffr0AQC4uLiwkCUiolzBYpaIvsqBAwfg6OiIsLAwjB8/Xuo4RERUwLCYJaIsmzx5Mtq2bQtXV1eEhISgbt26UkciIqIChsUsEWVZ0aJFsWjRIuzbtw/FihWTOg4RERVAPAGMiDTy559/IjQ0FBMmTMC4ceOkjkNERAUcZ2aJKFMSExPx/fffo2PHjrh48SKUSqXUkYiIiDgzS0Rf9vDhQ3h4eODq1av46aefMHLkSMhkMqljERERsZgloi+bNWsW3rx5g3PnzqFWrVpSxyEiIlJhMUtE6YqPj8ft27fh5OSEZcuWAQDMzMykDUVERPQfLGaJKI179+7B3d0dERERePjwIYtYIiLKs3gCGBGp2bp1K2rUqIG4uDgcPHgQBgYGUkciIiLKEItZIlKZO3cuevTogfbt2yMwMBCOjo5SRyIiIvosLjMgIgghIJPJ0LFjR1haWmLAgAG8WgEREWkFzswSFXCbNm1C3bp18eHDB1SuXBkDBw5kIUtERFqDxSxRARUXF4d+/frB09MT5cuX54cgEBGRVuIyA6IC6ObNm3B3d0dYWBh8fX3h6ekpdSQiIqIsYTFLVADduXMHcrkcly9fRuXKlaWOQ0RElGVcZkBUQMTGxuKXX36BEAKdO3dGUFAQC1kiItJ6LGaJCoCrV6+iZs2amDBhAh4+fAgA0NPTkzgVERHR12MxS5SPCSHw66+/wsXFBYaGhggMDESZMmWkjkVERJRtWMwS5WPbtm3D0KFDMWDAAFy4cAHly5eXOhIREVG24glgRPnQ69evUaxYMXTt2hXFihVD8+bNpY5ERESUIzgzS5SPCCGwYsUKlCpVCoGBgdDV1WUhS0RE+RpnZonyiXfv3mHAgAHYvXs3Ro0ahSpVqkgdiYiIKMexmCXKB27cuIF27drh3bt32LNnDzp27Ch1JCIiolzBZQZE+YCVlRUcHR0RHBzMQpaIiAoUFrNEWurNmzfo378/wsPDYWFhgT179sDe3l7qWERERLmKxSyRFjp37hwcHR3x559/4v79+1LHISIikgyLWSItolQqsWDBAjRo0AAlS5ZESEgI6tWrJ3UsIiIiybCYJdIid+/exbRp0zB+/HicPHkSdnZ2UkciIiKSFK9mQKQFLl++jOrVq6NixYq4f/8+bG1tpY5ERESUJ3BmligPUyqVmDNnDr799lusXbsWAFjIEhERfYIzs0R5VEREBHr37o1jx45h6tSpGDJkiNSRiIiI8hwWs0R5UEREBBwdHSGEwN9//w03NzepIxEREeVJXGZAlIcolUoAHz8EYfz48QgJCWEhS0RE9BksZonyiBcvXsDNzQ1bt24FAIwZMwbW1tYSpyIiIsrbWMwS5QF///03qlevjjt37sDGxkbqOERERFqDxSyRhJKTkzFlyhS0bNkSNWrUQEhICBo2bCh1LCIiIq3BYpZIQikpKTh27Bjmzp2LgwcPwtLSUupIREREWoVXMyCSwMGDB1GiRAlUr14dAQEB0NXlW5GIiCgrODNLlIuSkpIwYcIEtGnTBr/99hsAsJAlIiL6CvwpSpRLHj9+jG7duuHy5ctYtGgRvLy8pI5ERESk9VjMEuWClJQUNGvWDAkJCThz5gy+/fZbqSMRERHlCyxmiXJQYmIi4uPjYWpqik2bNqF8+fIoUqSI1LGIiIjyDa6ZJcohjx49Qr169TBgwAAAgIuLCwtZIiKibMZiligH7N69G05OTnj9+jUmTpwodRwiIqJ8i8UsUTbz8vJC586d0axZMwQFBaFWrVpSRyIiIsq3WMwSZbPixYtj5cqV2LFjB8zMzKSOQ0RElK/xBDCibLBt2zY8ffoU48aNw4QJE6SOQ0REVGBwZpboK3z48AFDhgxB9+7dcfXqVQghpI5ERERUoHBmliiLQkND4e7ujrt372Lt2rUYMGAAZDKZ1LGIiIgKFBazRFk0a9YsJCYm4tKlS6hatarUcYiIiAqkPLHMYOXKlbC3t4eBgQFcXFxw6dKlDPuuXbsW9evXR5EiRVCkSBG4ubl9tj9RdoqLi0NQUBCAj6/by5cvs5AlIiKSkOTF7Pbt2zFmzBj4+PggKCgI1atXR4sWLfDy5ct0+/v7+6N79+44efIkzp8/Dzs7OzRv3hzPnj3L5eRU0Ny8eRPOzs7o2LEjEhMTYWZmhkKFCkkdi4iIqECTvJhdsmQJBg0ahH79+qFy5cpYvXo1jIyMsH79+nT7+/n54bvvvoOjoyMqVqyI3377DUqlEsePH8/l5FRQCCGwbds2uLi4QCaT4fDhw1AoFFLHIiIiIki8ZjYxMRGBgYGYNGmSqk0ul8PNzQ3nz5/P1D7ev3+PpKQkFC1aNN37ExISkJCQoLodHR0NAFAqlVAqlV+RPnOUSiWEEP8eK/VMd5Erx6bs4ePjgzlz5qBfv374+eefYWRkxPHTMurvQ9JGHEPtxvHTfrk9hpocR9JiNjIyEikpKbCyslJrt7Kywp07dzK1j4kTJ8LGxgZubm7p3j9v3jzMmDEjTfurV68QHx+veWgNKZVKREVFQQiBlH8HJkWpzHAZBeUdSqUScrkcTZo0gampKXr16oXY2FjExsZKHY009On7UC6X/A9SlAUcQ+3G8dN+uT2GMTExme6r1Vcz+PHHH7Ft2zb4+/vDwMAg3T6TJk3CmDFjVLejo6NhZ2cHCwsLmJqa5nhGpVIJmUwGCwsL6Pw7+DpyOSwtLXP82JQ1QgisWbMGmzZtwvHjx9GgQQNUqlQJFhYW/CaspT59H3IMtRPHULtx/LRfbo9hRnVdeiQtZs3NzaGjo4OIiAi19oiICFhbW39220WLFuHHH3/EsWPHUK1atQz76evrQ19fP027XC7PtTeUTCb791ip1yCV8c2cR0VHR2Pw4MHYvn07hg4dqnqdpI4hx017cQy1H8dQu3H8tF9ujqEmx5D0FaVQKFCzZk21k7dST+ZydXXNcLsFCxZg1qxZOHz4MGrVqpUbUakACAoKQs2aNXHw4EFs374dv/zyi0a/GRIREVHuk3yZwZgxY+Dp6YlatWrB2dkZy5YtQ1xcHPr16wcA6NOnD0qUKIF58+YBAObPn49p06bh999/h729PcLDwwEAhQoV4mWS6Kvcu3cPpqamOHToEMqWLSt1HCIiIsoEyef6PTw8sGjRIkybNg2Ojo4ICQnB4cOHVSeFPX78GC9evFD1/+WXX5CYmIguXbqgePHiqq9FixZJ9RBIi7179w4rVqyAEAIeHh64ePEiC1kiIiItIvnMLACMGDECI0aMSPc+f39/tdthYWE5H4gKhMuXL8PDwwNv3rxBu3btUKpUKejq5om3BBEREWWS5DOzRLlNCIFly5ahbt26MDc3R3BwMEqVKiV1LCIiIsoCFrNU4GzcuBFeXl4YOXIkzp49CwcHB6kjERERURbxb6pUYLx69QoWFhbo2bMnSpUqhcaNG0sdiYiIiL4SZ2Yp31MqlVi4cCHs7e1x9epV6OnpsZAlIiLKJzgzS/laZGQkPD09cfDgQUyYMAGVK1eWOhIRERFlIxazlG8FBwejXbt2SEhIwMGDB9GqVSupIxEREVE24zIDyrdsbGzg6uqKkJAQFrJERET5FItZylciIiLQq1cvREREwMrKCjt37kSJEiWkjkVEREQ5hMsMKN84efIkevToAaVSiX/++Uf1KXJERESUf3FmlrReSkoKZsyYATc3N1SqVAkhISFwdnaWOhYRERHlAhazpPVu3bqFH3/8ET4+Pjh69CiKFy8udSQiIiLKJVxmQFrr3LlzqF27NqpWrYqHDx+yiCUiIiqAODNLWic5ORlTp05FvXr14OvrCwAsZImIiAoozsySVnn69Cl69OiBgIAAzJ49GwMGDJA6EhEREUmIxSxpjadPn8LR0REGBgbw9/dH/fr1pY5EREREEuMyA8rzlEolAKBEiRKYMmUKQkJCWMgSERERABazlMc9fvwY9evXx44dOyCTyeDl5QVzc3OpYxEREVEewWKW8qx9+/bB0dERT58+hZ2dndRxiIiIKA9iMUt5TmJiIsaMGYMOHTqgfv36CA4Ohqurq9SxiIiIKA9iMUt5jlKpREBAAJYtW4a9e/eiaNGiUkciIiKiPIpXM6A8Y/fu3ShXrhyqVq2Kc+fOQUdHR+pIRERElMdxZpYkFx8fj5EjR6Jz587YvHkzALCQJSIiokzhzCxJ6v79+3B3d8fNmzexcuVKDBs2TOpIREREpEVYzJJkkpKS0KxZM+jq6uLChQtwcnKSOhIRERFpGRazlOs+fPiApKQkmJqaYvv27ahYsSJMTU2ljkVERERaiGtmKVeFhobi22+/xdChQwEAzs7OLGSJiIgoy1jMUq7ZsmULatasiYSEBHh7e0sdh4iIiPIBFrOU44QQGDx4MHr37o1OnTrhypUrqFatmtSxiIiIKB/gmlnKcTKZDGXKlMH69evRt29fyGQyqSMRERFRPsFilnKEEAK+vr54/fo1xo0bh4kTJ0odiYiIiPIhLjOgbBcbGwtPT0/0798f9+7dgxBC6khERESUT3FmlrLV9evX4e7ujidPnmDz5s3o1auX1JGIiIgoH2MxS9lq9uzZUCgUCAwMRIUKFaSOQ0RERPkci1n6atHR0bh79y5q1aqFX3/9Ffr6+jA0NJQ6FhERERUALGbpqwQHB8Pd3R0pKSm4e/cuChcuLHUkIiIiKkB4AhhliRACK1euxLfffgtTU1P8/fff0NXl70ZERESUu1jMUpZMmDABI0aMwJAhQ3Du3DmULVtW6khERERUAHEqjTSSkpICHR0d9OrVC66urujUqZPUkYiIiKgA48wsZYoQAsuWLUO9evWQkJCA6tWrs5AlIiIiybGYpS968+YNOnbsCC8vL7i6uvLjaImIiCjP4DID+qzz58+jW7duiImJwZ9//on27dtLHYmIiIhIhcUsfVZYWBhKlCiBbdu2oWTJklLHISIiIlLDZQaURmRkJH7++WcAQPfu3XHmzBkWskRERJQnsZglNWfOnIGjoyNmzpyJ58+fAwB0dHQkTkVERESUPhazBABQKpWYO3cuGjdujNKlSyMkJAQ2NjZSxyIiIiL6LBazBABYu3Ytpk6dCm9vb5w4cQK2trZSRyIiIiL6Ip4AVsCFh4fD2toa/fr1wzfffIN69epJHYmIiIgo0zgzW0ClpKRgxowZcHBwwM2bN6FQKFjIEhERkdbhzGwB9OLFC/Tq1Qv+/v7w8fFBxYoVpY5ERERElCUsZguYixcvon379pDL5Th+/DgaNWokdSQiIiKiLOMygwKmVKlScHNzw9WrV1nIEhERkdZjMVsAPHv2DB4eHnj16hWsra3h5+cHS0tLqWMRERERfTUWs/ncoUOH4OjoiICAADx79kzqOERERETZisVsPpWUlISJEyeidevWcHZ2RkhICBwdHaWORURERJStWMzmUzdv3sTy5cuxYMEC/PXXXzA3N5c6EhEREVG249UM8hl/f3/Uq1cPjo6OePToEaysrKSORERERJRjODObTyQmJmLs2LFo3Lgxfv/9dwBgIUtERET5Hmdm84FHjx6hW7duCA4OxtKlS9G7d2+pIxERERHlChazWi4sLAxOTk4oWrQoAgICULt2bakjEREREeUaLjPQUikpKQA+fgjCjBkzEBQUxEKWiIiIChwWs1ro/v37cHZ2xh9//AGZTIbRo0ejcOHCUsciIiIiynUsZrXM9u3bUaNGDURHR6N06dJSxyEiIiKSFItZLREfH4+hQ4eiW7duaNOmDQIDA+Hk5CR1LCIiIiJJ8QQwLSGEQHBwMNasWYOBAwdCJpNJHYmIiIhIcixm8zg/Pz84OTmhcuXKOH/+PORyTqYTERERpWIxm0e9f/8eI0eOxPr16zFt2jTMmDGDhSzlG0IIJCcnq67KkdOUSiWSkpIQHx/P95GW4hhqN46f9suJMdTT04OOjs5X74fFbB5069YtuLu74+HDh1i/fj369u0rdSSibJOYmIgXL17g/fv3uXZMIQSUSiViYmK4REdLcQy1G8dP++XEGMpkMtja2qJQoUJftR8Ws3lMYmIimjdvDjMzM1y5cgWVK1eWOhJRtlEqlXj06BF0dHRgY2MDhUKRKz/YUmeCdXV1+YNUS3EMtRvHT/tl9xgKIfDq1Ss8ffoU5cqV+6oZWhazeURsbCyUSiVMTU2xd+9eVK5cGUZGRlLHIspWiYmJUCqVsLOzy9XXN3+Qaj+OoXbj+Gm/nBhDCwsLhIWFISkp6auKWS5cyQOuXbuG2rVrY8SIEQCAWrVqsZClfI1r5oiIKLuKYv5EkZAQAmvWrIGLiwsUCgUmT54sdSQiIiIircJiViJCCPTu3RtDhgyBp6cnLly4gIoVK0odi4iIiEirsJiViEwmQ5UqVbB161asXr0ahoaGUkciIspRr1+/hqWlJcLCwqSOkm94e3tj5MiROX6c6dOnw9HRMcePkxF7e3ssW7ZMsuPnlrCwMMhkMoSEhEgdRavkiWJ25cqVsLe3h4GBAVxcXHDp0qXP9t+5cycqVqwIAwMDVK1aFQcPHsylpF9HCIGYoAN4cXYXgI/fhLp16yZxKiLKjL59+0Imk0Emk0FPTw8ODg6YMGEC4uPj0/Tdv38/GjZsCBMTExgZGaF27drw9fVNd79//PEHGjVqBDMzMxQqVAjVqlXDzJkz8ebNm8/mOXnyJFq3bo1ixYrByMgIlStXxtixY/Hs2bPseLg5Ys6cOejQoQPs7e3T3NeiRQvo6Ojg8uXLae5r1KgRvv/++zTtvr6+KFy4sFpbdHQ0pkyZovoZYW1tDTc3N+zevRtCiGx6JGn5+/ujRo0a0NfXR9myZTMc70/t2LEDjo6OMDIyQqlSpbBw4cI0fVauXIlKlSrB0NAQFSpUwKZNm9TuHzduHDZu3IiHDx9qlNfX1xcymQyVKlVKc9/OnTshk8nUxmncuHE4fvy4Rsegz+vbty86duyo1mZnZ4cXL16gSpUquZrlzZs3GDlyJCpUqABDQ0OULFkSo0aNQlRUlFq/EydOoG7dujAxMYG1tTUmTpyI5ORktT6ZeV1nOyGxbdu2CYVCIdavXy9u3rwpBg0aJAoXLiwiIiLS7R8QECB0dHTEggULxK1bt8TUqVOFnp6euH79eqaOFxUVJQCIqKio7HwYGUpJSREvXrwQr1+/FkWrNBAAhHWdTrlybMoeqWOYkpIidRSt9+HDB3Hr1i3x4cOHXD2uUqkUiYmJQqlUZnkfnp6eomXLluLFixfi8ePHYs+ePcLU1FRMmDBBrd/PP/8s5HK5mDRpkrh586a4d++eWLRokdDX1xdjx45V6zt58mSho6Mjxo0bJwICAsSjR4/E33//LTp16iSWLVuWYZbVq1cLuVwu+vXrJ06ePCkePXokTp06JQYMGCC8vLyy/BgTEhKyvO2XxMXFCVNTU3H+/Pk09/3zzz+iUKFCYtSoUWLo0KFp7m/YsKEYNWpUmjHcsGGDMDMzU91++/at+Oabb4Stra3w9fUVN2/eFKGhoWLNmjWiTJky4u3btznx0MTDhw+FkZGRGDNmjLh165ZYvny50NHREYcPH85wm4MHDwpdXV3xyy+/iAcPHoj9+/eL4sWLi+XLl6v6rFq1SpiYmIht27aJBw8eiK1bt4pChQqJffv2qe2rS5cuYty4cRpl3rBhgzA2NhZFixYV586dU7uvRYsWomTJkqJUqVIa7fNzvvY9WKpUKbF06dJsy/MlGb0XEhMTs+0Ynp6eokOHDtm2v69x/fp10alTJ7Fv3z5x//59cfz4cVGuXDnRuXNnVZ/g4GChUCjE9OnTxb1794S/v7+oWLGi2ve1zLyuP/W5nwma1GuSF7POzs5i+PDhqtspKSnCxsZGzJs3L93+7u7uok2bNmptLi4uYsiQIZk6nhTF7KFDh4SDg4PQMTAW5h0nCZc5x3Ll2JQ9WMxmH20vZv/7g6dTp07CyclJdfvx48dCT09PjBkzJs32P//8swAgLly4IIQQ4uLFiwJAhkVrRoXXkydPhEKhEN9///1nt/Px8RHVq1dXu2/p0qVqBUrqY5o9e7YoXry4sLe3F5MmTRLOzs5p9lutWjUxY8YM1e21a9eKihUrCn19fVGhQgWxcuXKdPOk2rlzp7CwsEj3vunTp4tu3bqJ27dvCzMzM/H+/Xu1+zNbzA4bNkwYGxuLZ8+epTlGTEyMSEpK+mzGrJowYYL45ptv1No8PDxEixYtMtyme/fuokuXLmptP//8s7C1tVU9RldX1zRF6pgxY0TdunXV2jZu3ChsbW01ypz63I0YMUIMHDhQ1f7kyROhr68vvL291V4r/309nTx5UtSuXVsYGRkJMzMzUadOHREWFiaEECIkJEQ0atRIFCpUSJiYmIgaNWqIS5cuqcbvzJkzol69esLAwEDY2tqKkSNHitjYWNW+IyIiRNu2bYWBgYGwt7cXW7ZsyVQxu27dOlG5cmWhUCiEtbW1Wm3xzz//iPbt2wtjY2NhYmIiunbtKsLDw9M8vrVr1wp7e3shk8mEEEIAEKtWrRLt2rUTRkZGwsfHJ83rTggh9uzZIz6dH0zd3+rVq4Wtra0wNDQUXbt2Fe/evVPdD0DtK/UXUwAiODhYtS9/f39Ru3Zt1eOaOHGi2mu5YcOGYuTIkWL8+PGiSJEiwsrKSvj4+Hz2ucqMHTt2CIVCoTqWt7e3qFmzptp7cN++fcLAwEBER0cLITL3uv5UdhWzkl5nNjExEYGBgZg0aZKqTS6Xw83NDefPn093m/Pnz2PMmDFqbS1atMDevXvT7Z+QkICEhATV7ejoaAAfL96uVCq/8hF8WfsVAQhYPRtJiQoU7/sTdMysAYhcOTZlD6VSqfrkE/o6qc9l6leq9ivO4lVMYo4eW0BAhv+/DIyFiQL7RtTTfD//5r5x4wbOnTuHUqVKqdp27tyJpKQkjB07Ns2ftAcPHozJkyfj999/h7OzM7Zs2YJChQph2LBh6f7528zMLN32HTt2IDExEePHj//sdqn3fdonvbbjx4/D1NQUf//9t6pt3rx5uH//PsqUKQMAuHnzJq5du4Zdu3ZBCAE/Pz9MmzYNy5cvh5OTE4KDgzF48GAYGRnB09Mz3eft9OnTqFmzZprMQghs2LABK1asQIUKFVC2bFns3LkTvXv3TrOP/+b/9F+lUolt27ahR48eKF68eJrjGBsbp3nsqc6cOYPWrVunmzvV6tWr0bNnz3TvO3/+PJo2baq27+bNm8PLyyvDpQ0JCQkwMjJSu9/AwABPnz5FWFgY7O3tkZCQAH19/TR9Ll26hMTEROjp6QEAateujadPn+LRo0fpLuFIT+o++/Xrh8aNG2PZsmUwMjLChg0b0LJlS1haWqr1+/Tf5ORkdOzYEQMHDsTvv/+OxMRE1fJAIQR69uwJJycnrFq1Cjo6OggJCYGenh6EEHjw4AFatmyJWbNmYd26dXj16hVGjhyJESNGYP369QA+/vn9+fPnOHHiBPT09DB69Gi8fPkyzfeNT/3yyy8YO3Ys5s2bh1atWiEqKgoBAQGq10aHDh1QqFAh+Pv7Izk5GSNGjICHhwdOnjypyn3//n388ccf+OOPP6Cjo6M61vTp0zFv3jwsXboUurq6OHHihNpzktHzdP/+fezYsQP79u1DdHQ0Bg4ciO+++w5btmzB2LFjcfv2bURHR6sed9GiRfH8+XPV9kIIPHv2DK1bt4anpyc2btyIO3fuYPDgwdDX18f06dNVx9+4cSO8vLxw4cIFnD9/Hv369UOdOnXQrFkz1TiHhYWpHm9mvHv3DqampqrnIiEhAQYGBmqP08DAAPHx8bhy5QoaNWqUqdf1p1IfZ3o1mSY/cyUtZiMjI5GSkgIrKyu1disrK9y5cyfdbcLDw9PtHx4enm7/efPmYcaMGWnaX716le5at+wWEfUBJi1GQaargEzn4zcefR0ZXr58mePHpuyhVCoRFRUFIQSvj/qVkpKSoFQqkZycrLbO6mVMAiKiEz6zZfYTEGnWen2OUqnE/v37YWJiguTkZCQkJEAul2PZsmWq/YSGhsLMzAwWFhZp9i2Xy+Hg4IC7d+8iOTkZ9+7dg4ODA2QymUY57t69C1NT03SP8d+8qYXHp20AVG1KpRLGxsb45ZdfoFAoVP2qVauGLVu2YMqUKQCAzZs3w9nZGfb29khOTsb06dMxf/58tG/fHsDHdX43btzAr7/+mmHBFxYWBmtr6zSZjx07hvfv36Np06ZITk5G9+7dsW7dOnTv3l3VJ/WHXUpKCoD/vzblp4/n5cuXePv2LcqXL6/R8wkAjo6O6a7V/ZSVlVWG+33x4gWaN2+udr+5uTmio6MRExOT7gm+bm5uGDduHHr16oVGjRrh/v37WLx4MQDg6dOnsLW1hZubG9atW4d27drByckJQUFBWLduHZKSkhAeHo7ixYsDgKrwfPjwIWxtbTP1mFOfu6pVq8LBwQHbt29Hr169sHHjRixYsACPHj0CoP5aSX09vXnzBlFRUWjVqhVKlSoFAChXrpyq/+PHj+Hl5YWyZcsCABwcHCCEQEpKCubOnYvu3burrqvu4OCAJUuWoGnTpvj555/x+PFjHDp0COfOnUOtWrUAfPxFolq1aqrvHemZM2cOvv/+ewwfPlzV5uTkhOTkZBw7dgzXr1/H3bt3YWdnBwBYt24dHB0dceHCBdSqVQtKpRKJiYlYt24dLCws1B57t27d1H65+u/7CIDqtfnp8xUfH49169ahRIkSAIClS5eiQ4cO+PHHH2FtbQ19fX3o6enB3NxctZ/U7VO/R65YsQK2trZYtmwZZDIZypYti2nTpmHy5MmYPHky5HI5hBCoWrWq6v3q4OCAFStW4OjRo2jcuDGAj6+R/37f/ZzIyEjMnj0bAwYMUG3TtGlT/PTTT9iyZQu6du2K8PBwzJw5E8DH12xycnKmXtefSk5OhlKpxOvXr1W/nKWKiYnJVFagAHwC2KRJk9RmcqOjo2FnZwcLCwuYmprm+PGtzAwB2ccfZDIAxvq68HIrp/rmQ3mfUqmETCaDhYUFi9mvFB8fj5iYGOjq6kJX9/+//Via6KvNmuaE9GZmP83wJXK5HI0bN8aqVasQFxeHZcuWQUdHB+7u7qo+qUVWRvtNPYEs9f5P/6+JzGwnl8vT9Et9/aa2yeVyVK1aNc2HtPTs2RMbNmyAj48PhBDYsWMHvLy8oKuri7i4ODx48ABDhgzBsGHDVNskJyfDzMwsw1wJCQkwNDRMc/+mTZvg7u6umvHp2bMnvL298c8//6hmhmUyGeRyOXR0dNR+4H36eFI/PUgul2v8nJqYmHzVpRFT83163NQ8/32tpxoyZAgePXqEjh07IikpCaamphg1ahRmzJgBPT096OrqwsfHBy9fvkS9evUghICVlRX69OmDhQsXQqH4/9eviYkJgI/PcWYf+6fPXf/+/bF582Y4ODggLi4O7dq1w4oVK1T3p/ZPfT1ZWlqib9++aNOmDZo1a4amTZvC3d1dVVx7eXlh6NCh2Lp1K5o2bYquXbuqxvL69eu4du0atm7dqsqS+svKkydPcO/ePejq6sLZ2VmVsUqVKihcuHCGY/vy5Us8f/4czZo1S/f+1CLWwcFB1VatWjUULlwYd+/exbfffgu5XI5SpUqpHsOnateu/dn3EaA+3ql9SpYsqSr2AaBevXpQKpV48OABbG1tIZfL0zym1P+nvm7u3r2LOnXqqL3u69evj9jYWISHh6NkyZKQyWSoVq2a2n5sbGwQGRmpaps/f36ax5WR6OhodOzYEZUrV8bMmTNV+2jdujXmzZuHESNGoF+/ftDX18fUqVNx9uxZ1Ws2M6/rT+nq6kIul6NYsWKq7wGp/nv7cyQtZs3NzaGjo4OIiAi19oiICFhbW6e7jbW1tUb99fX1oa+vn6Y99UWU0/aNqIuXL1/C0tKShZAWS/1hxTH8Oqk/EFO/Uv01sn6OHjd1RulrP4bR2NhYNQO1fv16VK9eHevXr8eAAQMAABUqVEBUVBRevHgBGxsbtW0TExPx4MEDNG7cGDKZDOXLl8fZs2eRnJycZkbic1KP8enMXHpS/zT46eNNnWH5tM3Y2DjNc9KjRw94e3sjODgYHz58wJMnT9CtWzfIZDLExcUBANauXQsXF5c0x8zo+TU3N8e7d+/U7n/z5g327NmDpKQkrF69WtWekpKCDRs2YM6cOQAAU1NTREdHq7ZN/TcqKgpmZmaQyWSwtLRE4cKFERoaqvEYnzlzBq1atfpsn8/NOltbW+Ply5dqx3358iVMTU0z/DRHmUyGBQsWYN68eQgPD4eFhYXqagFlypSBTCZT/dl/zZo1iIiIQPHixbFmzRqYmJjA0tJSdby3b98CgFrbl3z6XPbq1QsTJ07EjBkz0Lt3b+jp6aV5rv/774YNGzBq1CgcPnwYO3bswA8//ICjR4/i22+/xYwZM9CzZ08cOHAAhw4dwvTp07F161a0a9cOsbGxGDJkCEaNGpUmU8mSJXHv3j3Vcf77WNJrA6B6jjO6/7/Z09unTCZL970AAIUKFVJrz8x7K71jftqWXvvX9FEoFGn6/DdjZsTExKBVq1YwMTHBnj171P5iI4SAl5cXxo0bh/DwcBQpUgRhYWGYPHmy6jWbmdf1p1K3Se/nqyY/byX9yaxQKFCzZk21y30olUocP34crq6u6W7j6uqa5vIgR48ezbA/EVFOkMvlmDx5MqZOnYoPHz4AADp37gw9PT3Vn9U+tXr1asTFxan+fN6jRw/ExsZi1apV6e7/3bt36bZ36dIFCoUCCxYs+Ox2FhYWCA8PV1u7ltlrV9ra2qJhw4bw8/ODn58fmjVrpvprkpWVFWxsbPDw4UOULVtW7evTma//cnJywq1bt9Ta/Pz8YGtri6tXryIkJET1tXjxYvj6+qr+dFuhQgUEBQWl2WdQUBDKly8P4ON4dOvWDX5+fqp1h5+KjY3N8E+stWrVUjt+el+pSyrS8zU/l3R0dFCiRAkoFAps3boVrq6uqj9zp9LT04OtrS10dHSwbds2tG3bVu0H/Y0bN6Cnp4dvvvnmi8dLT9GiRdG+fXucOnUK/fv3z/R2Tk5OmDRpEs6dO4cqVarg999/V91Xvnx5eHl54e+//0anTp1UlyqrUaMGbt26lea1U7ZsWSgUClSsWBHJyckIDAxU7Ss0NDTD9wPwcWba3t4+w0uHVapUCU+ePMGTJ09Ubbdu3cK7d+9QuXLlTD/eVBYWFoiJiVH9Ygek/956/Pix2mvxwoULkMvlqFChAoCPNVDqazwjlSpVwvnz59XexwEBATAxMcn0kpLMio6ORvPmzaFQKLBv374MZ0ZlMhlsbGxgaGiIrVu3ws7ODjVq1FDrk5nXdbb64iliOWzbtm1CX19f+Pr6ilu3bonBgweLwoULq84y7N27t/D29lb1DwgIELq6umLRokXi9u3bwsfHRysuzcUz4bUXxzD75LerGSQlJYkSJUqIhQsXqtqWLl0q5HK5mDx5srh9+7a4f/++WLx4cbqX5powYYLQ0dER48ePF+fOnRNhYWHi2LFjokuXLp+9NNfKlSuFTCYT/fv3F/7+/iIsLEycPXtWDB48WHUlhVu3bgmZTCZ+/PFHcf/+fbFixQpRpEiRdK9mkJ61a9cKGxsbYW5uLjZv3pzmPkNDQ/HTTz+J0NBQce3aNbF+/XqxePHiDDNfu3ZN6Orqijdv3qjaqlevLiZOnJim77t374RCoRD79+8XQgjx4MEDYWBgIIYPHy5CQkLEnTt3xOLFi4Wurq44dOiQarvXr1+LihUrCltbW7Fx40Zx8+ZNcffuXbFu3TpRtmzZHL801/jx48Xt27fFypUr01yaa/ny5aJJkyaq269evRK//PKLuH37tggODhajRo0SBgYG4uLFi6o+oaGhYvPmzeLu3bvi4sWLwsPDQxQtWlQ8evRI7fg+Pj5q+86M/56R//79exEZGam6/d8rX3x6NYOHDx8Kb29v1Wv2yJEjolixYmLVqlXi/fv3Yvjw4eLkyZOq12WZMmXE+PHjRWJioggJCRGGhoZi+PDhIjg4WNy9e1fs3btX7coDLVu2FE5OTuLChQviypUrol69esLQ0PCzVzPw9fUVBgYG4qeffhJ3794VgYGB4ueffxZCfHz/Ozo6ivr164vAwEBx8eJFUbNmTdGwYcN0H9+nAIg9e/aotb1+/VoYGxuLUaNGifv37ws/Pz9hY2OT5moGxsbGws3NTYSEhIjTp0+L8uXLi27duqn6zJkzR5QsWVLcuXNHvHr1SiQmJqa5msHTp0+FkZGRGD58uLh9+7bYu3evMDc3V7taQcOGDcXo0aPVMnbo0EF4enqqbnt7e4vevXtn+PxFRUUJFxcXUbVqVXH//n3x4sUL1VdycrLqeZw3b564evWquHHjhpg5c6bQ09NTe34y87r+VL65NJcQH9/kJUuWFAqFQjg7O6suXSPEx0H6dECE+Hi5iPLlywuFQiG++eYbceDAgUwfi8UsaYpjmH3yWzErhBDz5s0TFhYWapcW+vPPP0X9+vWFsbGxMDAwEDVr1hTr169Pd7/bt28XDRo0ECYmJsLY2FhUq1ZNzJw584uF19GjR0WLFi1EkSJFhIGBgahYsaIYN26ceP78uarPL7/8Iuzs7ISxsbHo06ePmDNnTqaL2bdv3wp9fX1hZGQkYmJi0tzv5+cnHB0dhUKhEEWKFBENGjQQu3fv/mxmZ2dnsXr1aiGEEFeuXBEAxKVLl9Lt26pVK/G///1PdfvixYvCzc1NWFhYCDMzM+Hi4pKmyBDiYyHs7e0typUrJxQKhbCyshJubm5iz549XzX+X3Ly5EnV81G6dGmxYcMGtft9fHzUnvtXr16Jb7/9VhgbGwsjIyPRtGlTtZ99Qnz8hcTR0VEYGhoKU1NT0aFDB3Hnzp00x65QoYLYunWrWlupUqU+e3mm9C4v9anPFbPh4eGiY8eOonjx4kKhUIhSpUqJadOmiZSUFJGQkCC6desm7OzshEKhEDY2NmLEiBHi/fv3qvfgpUuXRLNmzUShQoVUr/k5c+aojvXixQvRpk0boa+vL0qWLCk2bdqUqUtzrV69WlSoUEHo6emJ4sWLi5EjR6ruy+yluf4rvWJWiI+X4ipbtqwwNDQUbdu2FWvWrEn30lyrVq0SNjY2wsDAQHTp0kXtl7mXL1+qngd85aW5vlTMenp6qhXv/3Xy5Mk0lwpL/Ur95UmpVIpGjRoJMzMzYWBgIFxcXMTBgwfV9pOZ1/WnsquYlQmRgx+JkgdFR0fDzMwMUVFRuXICmFKp5JpZLccxzD7x8fF49OgRHBwcNFrc/7VENq2Zpa9z4MABjB8/Hjdu3ND4vcQxTN+hQ4cwduxYXLt2TXVyzfv371GsWDEcOnQIjRo1kjbgvwra+E2fPh179+7NVx9LmxNj+LmfCZrUa/n+agZERJQ3tGnTBvfu3cOzZ89Ul0iirxMXF4cNGzaonSV+8uRJNGnSJM8UskQ5jcUsERHlmu+//17qCPlKly5d0rS1adMGbdq0kSANkTT4N1MiIiKibDR9+vR8tcQgr2MxS0RERERai8UsEeW6AnbeKRERpSO7fhawmCWiXJP6SVfv37+XOAkREUktMTERwP9/HHBW8QQwIso1Ojo6KFy4MF6+fAng48dQ5sZlegraZYHyI46hduP4ab/sHkOlUolXr17ByMhI7WocWcFilohylbW1NQCoCtrcIISAUqmEXC7nD1ItxTHUbhw/7ZcTYyiXy1GyZMmv3h+LWSLKVTKZDMWLF4elpSWSkpJy5ZhKpRKvX79GsWLF+MEXWopjqN04ftovJ8ZQoVBky75YzBKRJHR0dL56nVRmKZVK6OnpwcDAgD9ItRTHULtx/LRfXh7DvJWGiIiIiEgDLGaJiIiISGuxmCUiIiIirVXg1symXqA3Ojo6V46nVCoRExOTJ9eYUOZwDLUfx1D7cQy1G8dP++X2GKbWaZn5YIUCV8zGxMQAAOzs7CROQkRERESfExMTAzMzs8/2kYkC9rmSSqUSz58/h4mJSa5c6y46Ohp2dnZ48uQJTE1Nc/x4lP04htqPY6j9OIbajeOn/XJ7DIUQiImJgY2NzRdnggvczKxcLoetrW2uH9fU1JRvYC3HMdR+HEPtxzHUbhw/7ZebY/ilGdlUXLhCRERERFqLxSwRERERaS0WszlMX18fPj4+0NfXlzoKZRHHUPtxDLUfx1C7cfy0X14ewwJ3AhgRERER5R+cmSUiIiIircViloiIiIi0FotZIiIiItJaLGaJiIiISGuxmM0GK1euhL29PQwMDODi4oJLly59tv/OnTtRsWJFGBgYoGrVqjh48GAuJaWMaDKGa9euRf369VGkSBEUKVIEbm5uXxxzynmavg9Tbdu2DTKZDB07dszZgPRFmo7hu3fvMHz4cBQvXhz6+vooX748v59KSNPxW7ZsGSpUqABDQ0PY2dnBy8sL8fHxuZSW/uv06dNo164dbGxsIJPJsHfv3i9u4+/vjxo1akBfXx9ly5aFr69vjudMl6Cvsm3bNqFQKMT69evFzZs3xaBBg0ThwoVFREREuv0DAgKEjo6OWLBggbh165aYOnWq0NPTE9evX8/l5JRK0zHs0aOHWLlypQgODha3b98Wffv2FWZmZuLp06e5nJxSaTqGqR49eiRKlCgh6tevLzp06JA7YSldmo5hQkKCqFWrlmjdurU4e/asePTokfD39xchISG5nJyE0Hz8/Pz8hL6+vvDz8xOPHj0SR44cEcWLFxdeXl65nJxSHTx4UEyZMkXs3r1bABB79uz5bP+HDx8KIyMjMWbMGHHr1i2xfPlyoaOjIw4fPpw7gT/BYvYrOTs7i+HDh6tup6SkCBsbGzFv3rx0+7u7u4s2bdqotbm4uIghQ4bkaE7KmKZj+F/JycnCxMREbNy4Maci0hdkZQyTk5NFnTp1xG+//SY8PT1ZzEpM0zH85ZdfROnSpUViYmJuRaTP0HT8hg8fLpo0aaLWNmbMGFG3bt0czUmZk5lidsKECeKbb75Ra/Pw8BAtWrTIwWTp4zKDr5CYmIjAwEC4ubmp2uRyOdzc3HD+/Pl0tzl//rxafwBo0aJFhv0pZ2VlDP/r/fv3SEpKQtGiRXMqJn1GVsdw5syZsLS0xIABA3IjJn1GVsZw3759cHV1xfDhw2FlZYUqVapg7ty5SElJya3Y9K+sjF+dOnUQGBioWorw8OFDHDx4EK1bt86VzPT18lI9o5vrR8xHIiMjkZKSAisrK7V2Kysr3LlzJ91twsPD0+0fHh6eYzkpY1kZw/+aOHEibGxs0rypKXdkZQzPnj2LdevWISQkJBcS0pdkZQwfPnyIEydOoGfPnjh48CDu37+P7777DklJSfDx8cmN2PSvrIxfjx49EBkZiXr16kEIgeTkZAwdOhSTJ0/OjciUDTKqZ6Kjo/HhwwcYGhrmWhbOzBJ9hR9//BHbtm3Dnj17YGBgIHUcyoSYmBj07t0ba9euhbm5udRxKIuUSiUsLS2xZs0a1KxZEx4eHpgyZQpWr14tdTTKBH9/f8ydOxerVq1CUFAQdu/ejQMHDmDWrFlSRyMtxJnZr2Bubg4dHR1ERESotUdERMDa2jrdbaytrTXqTzkrK2OYatGiRfjxxx9x7NgxVKtWLSdj0mdoOoYPHjxAWFgY2rVrp2pTKpUAAF1dXYSGhqJMmTI5G5rUZOV9WLx4cejp6UFHR0fVVqlSJYSHhyMxMREKhSJHM9P/y8r4/fDDD+jduzcGDhwIAKhatSri4uIwePBgTJkyBXI559ryuozqGVNT01ydlQU4M/tVFAoFatasiePHj6valEoljh8/DldX13S3cXV1VesPAEePHs2wP+WsrIwhACxYsACzZs3C4cOHUatWrdyIShnQdAwrVqyI69evIyQkRPXVvn17NG7cGCEhIbCzs8vN+ISsvQ/r1q2L+/fvq34RAYC7d++iePHiLGRzWVbG7/3792kK1tRfTIQQOReWsk2eqmdy/ZSzfGbbtm1CX19f+Pr6ilu3bonBgweLwoULi/DwcCGEEL179xbe3t6q/gEBAUJXV1csWrRI3L59W/j4+PDSXBLTdAx//PFHoVAoxK5du8SLFy9UXzExMVI9hAJP0zH8L17NQHqajuHjx4+FiYmJGDFihAgNDRX79+8XlpaWYvbs2VI9hAJN0/Hz8fERJiYmYuvWreLhw4fi77//FmXKlBHu7u5SPYQCLyYmRgQHB4vg4GABQCxZskQEBweLf/75RwghhLe3t+jdu7eqf+qlucaPHy9u374tVq5cyUtzabPly5eLkiVLCoVCIZydncWFCxdU9zVs2FB4enqq9d+xY4coX768UCgU4ptvvhEHDhzI5cT0X5qMYalSpQSANF8+Pj65H5xUNH0fforFbN6g6RieO3dOuLi4CH19fVG6dGkxZ84ckZycnMupKZUm45eUlCSmT58uypQpIwwMDISdnZ347rvvxNu3b3M/OAkhhDh58mS6P9tSx83T01M0bNgwzTaOjo5CoVCI0qVLiw0bNuR6biGEkAnB+XwiIiIi0k5cM0tEREREWovFLBERERFpLRazRERERKS1WMwSERERkdZiMUtEREREWovFLBERERFpLRazRERERKS1WMwSERERkdZiMUtEBMDX1xeFCxeWOkaWyWQy7N2797N9+vbti44dO+ZKHiKi3MJilojyjb59+0Imk6X5un//vtTR4Ovrq8ojl8tha2uLfv364eXLl9my/xcvXqBVq1YAgLCwMMhkMoSEhKj1+emnn+Dr65stx8vI9OnTVY9TR0cHdnZ2GDx4MN68eaPRflh4E1Fm6UodgIgoO7Vs2RIbNmxQa7OwsJAojTpTU1OEhoZCqVTi6tWr6NevH54/f44jR4589b6tra2/2MfMzOyrj5MZ33zzDY4dO4aUlBTcvn0b/fv3R1RUFLZv354rxyeigoUzs0SUr+jr68Pa2lrtS0dHB0uWLEHVqlVhbGwMOzs7fPfdd4iNjc1wP1evXkXjxo1hYmICU1NT1KxZE1euXFHdf/bsWdSvXx+Ghoaws7PDqFGjEBcX99lsMpkM1tbWsLGxQatWrTBq1CgcO3YMHz58gFKpxMyZM2Frawt9fX04Ojri8OHDqm0TExMxYsQIFC9eHAYGBihVqhTmzZuntu/UZQYODg4AACcnJ8hkMjRq1AiA+mznmjVrYGNjA6VSqZaxQ4cO6N+/v+r2n3/+iRo1asDAwAClS5fGjBkzkJyc/NnHqaurC2tra5QoUQJubm7o2rUrjh49qro/JSUFAwYMgIODAwwNDVGhQgX89NNPqvunT5+OjRs34s8//1TN8vr7+wMAnjx5And3dxQuXBhFixZFhw4dEBYW9tk8RJS/sZglogJBLpfj559/xs2bN7Fx40acOHECEyZMyLB/z549YWtri8uXLyMwMBDe3t7Q09MDADx48AAtW7ZE586dce3aNWzfvh1nz57FiBEjNMpkaGgIpVKJ5ORk/PTTT1i8eDEWLVqEa9euoUWLFmjfvj3u3bsHAPj555+xb98+7NixA6GhofDz84O9vX26+7106RIA4NixY3jx4gV2796dpk/Xrl3x+vVrnDx5UtX25s0bHD58GD179gQAnDlzBn369MHo0aNx69Yt/Prrr/D19cWcOXMy/RjDwsJw5MgRKBQKVZtSqYStrS127tyJW7duYdq0aZg8eTJ27NgBABg3bhzc3d3RsmVLvHjxAi9evECdOnWQlJSEFi1awMTEBGfOnEFAQAAKFSqEli1bIjExMdOZiCifEURE+YSnp6fQ0dERxsbGqq8uXbqk23fnzp2iWLFiqtsbNmwQZmZmqtsmJibC19c33W0HDBggBg8erNZ25swZIZfLxYcPH9Ld5r/7v3v3rihfvryoVauWEEIIGxsbMWfOHLVtateuLb777jshhBAjR44UTZo0EUqlMt39AxB79uwRQgjx6NEjAUAEBwer9fH09BQdOnRQ3e7QoYPo37+/6vavv/4qbGxsREpKihBCiKZNm4q5c+eq7WPz5s2iePHi6WYQQggfHx8hl8uFsbGxMDAwEAAEALFkyZIMtxFCiOHDh4vOnTtnmDX12BUqVFB7DhISEoShoaE4cuTIZ/dPRPkX18wSUb7SuHFj/PLLL6rbxsbGAD7OUs6bNw937txBdHQ0kpOTER8fj/fv38PIyCjNfsaMGYOBAwdi8+bNqj+VlylTBsDHJQjXrl2Dn5+fqr8QAkqlEo8ePUKlSpXSzRYVFYVChQpBqVQiPj4e9erVw2+//Ybo6Gg8f/4cdevWVetft25dXL16FcDHJQLNmjVDhQoV0LJlS7Rt2xbNmzf/queqZ8+eGDRoEFatWgV9fX34+fmhW7dukMvlqscZEBCgNhObkpLy2ecNACpUqIB9+/YhPj4eW7ZsQUhICEaOHKnWZ+XKlVi/fj0eP36MDx8+IDExEY6Ojp/Ne/XqVdy/fx8mJiZq7fHx8Xjw4EEWngEiyg9YzBJRvmJsbIyyZcuqtYWFhaFt27YYNmwY5syZg6JFi+Ls2bMYMGAAEhMT0y3Kpk+fjh49euDAgQM4dOgQfHx8sG3bNvzvf/9DbGwshgwZglGjRqXZrmTJkhlmMzExQVBQEORyOYoXLw5DQ0MAQHR09BcfV40aNfDo0SMcOnQIx44dg7u7O9zc3LBr164vbpuRdu3aQQiBAwcOoHbt2jhz5gyWLl2quj82NhYzZsxAp06d0mxrYGCQ4X4VCoVqDH788Ue0adMGM2bMwKxZswAA27Ztw7hx47B48WK4urrCxMQECxcuxMWLFz+bNzY2FjVr1lT7JSJVXjnJj4hyH4tZIsr3AgMDoVQqsXjxYtWsY+r6zM8pX748ypcvDy8vL3Tv3h0bNmzA//73P9SoUQO3bt1KUzR/iVwuT3cbU1NT2NjYICAgAA0bNlS1BwQEwNnZWa2fh4cHPDw80KVLF7Rs2RJv3rxB0aJF1faXuj41JSXls3kMDAzQqVMn+Pn54f79+6hQoQJq1Kihur9GjRoIDQ3V+HH+19SpU9GkSRMMGzZM9Tjr1KmD7777TtXnvzOrCoUiTf4aNWpg+/btsLS0hKmp6VdlIqL8gyeAEVG+V7ZsWSQlJWH58uV4+PAhNm/ejNWrV2fY/8OHDxgxYgT8/f3xzz//ICAgAJcvX1YtH5g4cSLOnTuHESNGICQkBPfu3cOff/6p8Qlgnxo/fjzmz5+P7du3IzQ0FN7e3ggJCcHo0aMBAEuWLMHWrVtx584d3L17Fzt37oS1tXW6H/RgaWkJQ0NDHD58GBEREYiKisrwuD179sSBAwewfv161YlfqaZNm4ZNmzZhxowZuHnzJm7fvo1t27Zh6tSpGj02V1dXVKtWDXPnzgUAlCtXDleuXMGRI0dw9+5d/PDDD7h8+bLaNvb29rh27RpCQ0MRGRmJpKQk9OzZE+bm5ujQoQPOnDmDR48ewd/fH6NGjcLTp081ykRE+QeLWSLK96pXr44lS5Zg/vz5qFKlCvz8/NQua/VfOjo6eP36Nfr06YPy5cvD3d0drVq1wowZMwAA1apVw6lTp3D37l3Ur18fTk5OmDZtGmxsbLKccdSoURgzZgzGjh2LqlWr4vDhw9i3bx/KlSsH4OMShQULFqBWrVqoXbs2wsLCcPDgQdVM86d0dXXx888/49dff4WNjQ06dOiQ4XGbNGmCokWLIjQ0FD169FC7r0WLFti/fz/+/vtv1K5dG99++y2WLl2KUqVKafz4vLy88Ntvv+HJkycYMmQIOnXqBA8PD7i4uOD169dqs7QAMGjQIFSoUAG1atWChYUFAgICYGRkhNOnT6NkyZLo1KkTKlWqhAEDBiA+Pp4ztUQFmEwIIaQOQURERESUFZyZJSIiIiKtxWKWiIiIiLQWi1kiIiIi0losZomIiIhIa7GYJSIiIiKtxWKWiIiIiLQWi1kiIiIi0losZomIiIhIa7GYJSIiIiKtxWKWiIiIiLQWi1kiIiIi0lr/Bw/fICa5v0rRAAAAAElFTkSuQmCC",
45
- "text/plain": [
46
- "<Figure size 800x600 with 1 Axes>"
47
- ]
48
- },
49
- "metadata": {},
50
- "output_type": "display_data"
51
- }
52
- ],
53
- "source": [
54
- "# Plot the ROC curve\n",
55
- "plt.figure(figsize=(8, 6))\n",
56
- "plt.plot(\n",
57
- " fpr,\n",
58
- " tpr,\n",
59
- " label=f\"ROC Curve (AUC = {roc_auc:.2f}), Missed corruption: {len(corrupted_files) - y_true.sum()}\",\n",
60
- " lw=2,\n",
61
- ")\n",
62
- "plt.plot([0, 1], [0, 1], \"k--\", lw=1) # Diagonal line\n",
63
- "plt.xlabel(\"False Positive Rate\")\n",
64
- "plt.ylabel(\"True Positive Rate\")\n",
65
- "plt.title(\"ROC Curve\")\n",
66
- "plt.legend(loc=\"lower right\")\n",
67
- "plt.grid(alpha=0.3)\n",
68
- "plt.show()"
69
- ]
70
- },
71
- {
72
- "cell_type": "code",
73
- "execution_count": null,
74
- "metadata": {},
75
- "outputs": [],
76
- "source": []
77
- }
78
- ],
79
- "metadata": {
80
- "kernelspec": {
81
- "display_name": "Python (pixi-dev)",
82
- "language": "python",
83
- "name": "pixi-dev"
84
- },
85
- "language_info": {
86
- "codemirror_mode": {
87
- "name": "ipython",
88
- "version": 3
89
- },
90
- "file_extension": ".py",
91
- "mimetype": "text/x-python",
92
- "name": "python",
93
- "nbconvert_exporter": "python",
94
- "pygments_lexer": "ipython3",
95
- "version": "3.12.3"
96
- }
97
- },
98
- "nbformat": 4,
99
- "nbformat_minor": 2
100
- }