{"metadata":{"kernelspec":{"name":"python3","display_name":"Python 3","language":"python"},"language_info":{"name":"python","version":"3.11.11","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":103436,"databundleVersionId":12476472,"sourceType":"competition"},{"sourceId":11973501,"sourceType":"datasetVersion","datasetId":7529571},{"sourceId":11973518,"sourceType":"datasetVersion","datasetId":7529585}],"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"id":"55c85d2c","cell_type":"markdown","source":"<a id=\"import\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">Step 1 | Import Libraries</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"c21cec14","cell_type":"code","source":"import os\nimport pandas as pd\nimport numpy as np\nimport random\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nfrom tqdm.notebook import tqdm\nimport math\nimport re\nimport warnings\nfrom itertools import combinations\nfrom datetime import datetime\nwarnings.filterwarnings('ignore')\n\nimport scipy\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.preprocessing import LabelEncoder, OneHotEncoder\nfrom sklearn.feature_selection import SelectFromModel\nfrom sklearn.model_selection import train_test_split, StratifiedKFold, KFold, cross_val_score, TunedThresholdClassifierCV, RepeatedStratifiedKFold\nimport itertools\nimport optuna\nfrom catboost import CatBoostClassifier, Pool\nfrom sklearn.metrics import  f1_score, make_scorer, accuracy_score, precision_score, recall_score, precision_recall_curve, auc, classification_report, roc_auc_score, roc_curve\nfrom lightgbm import LGBMClassifier, early_stopping,log_evaluation\nfrom sklearn.ensemble import VotingClassifier\nimport joblib\n\n","metadata":{},"outputs":[],"execution_count":null},{"id":"8e225cf8","cell_type":"code","source":"# Set the resolution of the plotted figures\nplt.rcParams['figure.dpi'] = 120\n\n# Configure Seaborn plot styles: Set background color and use dark grid\nsns.set_theme(rc={'axes.facecolor': '#F3E8FF'}, style='darkgrid')","metadata":{},"outputs":[],"execution_count":null},{"id":"bca49663","cell_type":"code","source":"#Set display max columns\npd.set_option('display.max_columns', None)","metadata":{},"outputs":[],"execution_count":null},{"id":"639b8140","cell_type":"markdown","source":"<a id=\"read\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">Step 2 | Read Dataset</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"94f6ea96","cell_type":"code","source":"#Import delay delivery datasets\ndata_4_6 = pd.read_parquet(r\"D:\\Python_code\\DS_108_BTVN\\data_4_6_full_clean.parquet\")\ndata_7_9 = pd.read_parquet(r\"D:\\Python_code\\DS_108_BTVN\\data_7_9_full_clean.parquet\")\ndata_4_9 = pd.concat([data_4_6, data_7_9], ignore_index=True)\ndata_4_9 = data_4_9.drop_duplicates().reset_index(drop=True)\ndata_10 = pd.read_csv(r\"D:\\Python_code\\DS_108_BTVN\\Kaggle\\PILOT_10.csv\")\ndata_4_6.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"3ec3a825","cell_type":"code","source":"data_7_9.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"59203837","cell_type":"code","source":"data_4_9.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"62812446","cell_type":"code","source":"data_10.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"3153fa0e","cell_type":"code","source":"#Kiểm tra xem số lượng các cột trong 2 bộ dữ liệu có giống nhau hay không\nprint(\"Số lượng cột trong data_4_9 khác data_10 là:\")\ndata_10.columns.difference(data_4_9.columns)","metadata":{},"outputs":[],"execution_count":null},{"id":"cda275b2","cell_type":"code","source":"# #Loại hết các cột có ở data_10 nhưng không có ở data_4_9\n# columns_to_drop = set(data_10.columns) - set(data_4_9.columns)\n# data_10 = data_10.drop(columns=columns_to_drop)\n# #Kiểm tra lại các cột có trong data_10\n# data_10.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"120f1f39","cell_type":"code","source":"# Kiểm tra giá trị thiếu\ndata_10.isnull().sum()","metadata":{},"outputs":[],"execution_count":null},{"id":"6b85f460","cell_type":"code","source":"#Kiểm tra xem số lượng các cột trong 2 bộ dữ liệu có giống nhau hay không\nprint(\"Số lượng cột trong data_7_9 khác data_10 là:\")\ndata_7_9.columns.difference(data_10.columns)","metadata":{},"outputs":[],"execution_count":null},{"id":"96ab5a59-0112-4640-98df-f3894475540a","cell_type":"code","source":"#Chuyển đổi kiểu dữ liệu của cột 'Order date' thành datetime\ndata_10['Order date'] = pd.to_datetime(data_10['Order date'])\ndata_10['VSD'] = pd.to_datetime(data_10['VSD'])\ndata_10","metadata":{},"outputs":[],"execution_count":null},{"id":"3577c37c-1a84-499d-ae02-9bf69e7717b5","cell_type":"code","source":"# data_10 = data_10.sample(frac = 1).drop_duplicates().reset_index()\ndata_10 = data_10.applymap(lambda x: x.strip() if isinstance(x, str) else x) #Loại bỏ khoảng trắng thừa trong toàn bộ DataFrame\n# data_10 = data_10.sort_values(by=['Order date']).reset_index(drop=True) #Sắp xếp dữ liệu theo thứ tự thời gian\n# data_10.replace('', np.nan, inplace=True) #Thay thế các giá trị trống bằng NaN\nindex_10 = data_10['ID']\ndata_10","metadata":{},"outputs":[],"execution_count":null},{"id":"6510862b-2765-49e1-ac56-1b05d200ec0b","cell_type":"code","source":"#Kiểm tra giá trị trùng lặp trong cột 'SPECIAL DIV' \ncnt = (data_10['SPECIAL DIV'] != data_10['SPECIAL_DIV']).sum()\nprint(f\"Số giá trị khác nhau giữa 2 cột 'SPECIAL DIV' và 'SPECIAL_DIV' là: {cnt}\")","metadata":{},"outputs":[],"execution_count":null},{"id":"0e080c88-d738-4a5e-a187-1e81835c8533","cell_type":"code","source":"#Loại bỏ 1 cột SPECIAL DIV vì bị trùng lặp thuộc tính (Các giá trị ở 2 cột này giống nhau)\ndata_10 = data_10.drop(columns=['SPECIAL DIV'])","metadata":{},"outputs":[],"execution_count":null},{"id":"244407e8-c16f-4b0a-a964-19d35ada3674","cell_type":"code","source":"#Trích ra các thuộc tính chuẩn theo bộ dữ liệu data_delay_7_9 để phân tích\nfull_features = ['Order date', 'SUBSIDIARY_CD', 'GLOBAL_NO', 'CLASSIFY_CD', 'CUST_CD',\n       'BRAND_CD', 'INNER_CD', 'SUPPLIER_CD', 'Sales order line number',\n       'Stock class', 'Consider count hodiday Saturday', 'SO QTY',\n       'OTHER AREA SHIP DIV', 'ALLOCATION QTY', 'SUPPLIER INV AMOUNT',\n       'PACKING RANK', 'PRODUCT_CD', 'PRODUCT ATTRIBUTION', 'SPECIAL_DIV',\n       'LOGICAL PLANT', 'PURCHASE AMOUNT', 'VSD', 'DIRECT SHIP FLG',\n       'DELI_DIV', 'Ship Mode', 'SHIP DECISION NO', 'PACK QTY',\n       'WEIGHT PER PIECE', 'SUPPLIER_DIV', 'SO_DAY_OF_MONTH', 'SO_DAY_OF_WEEK',\n       'REASON_CD', 'SOUF_RCV_NO', 'QTUF_RCV_NO', 'SO_TIME']","metadata":{},"outputs":[],"execution_count":null},{"id":"a9472d5c-59d6-4c71-81a3-158845165b35","cell_type":"code","source":"#Tiến hành trích lọc và kết hợp dữ liệu\nfull_filter_data = data_10[full_features].copy() #Lọc ra các thuộc tính cần thiết\n\n#Chúng ta sẽ thay thế toàn bộ khoảng trắng với giá trị NaN (np.nan)\ncols_to_replace = full_filter_data.columns\nfull_filter_data[cols_to_replace].replace('', np.nan, inplace=True) #Thay thế các giá trị trống bằng NaN\n\nfull_filter_data","metadata":{},"outputs":[],"execution_count":null},{"id":"0f0efc5a-5f7a-4583-9316-17e10e07bc58","cell_type":"code","source":"# Chuyển các cột sau sang string\ndata_columns_to_string = ['GLOBAL_NO','CLASSIFY_CD', 'CUST_CD', 'SHIP DECISION NO', 'REASON_CD', 'LOGICAL PLANT', 'LOGICAL PLANT', 'DELI_DIV', 'SHIP DECISION NO', 'SUPPLIER_DIV', 'QTUF_RCV_NO']\nfor col in data_columns_to_string:\n    full_filter_data[col] = full_filter_data[col].astype(str)\n\n# Thay thế tất cả các giá trị 'nan' (chuỗi) thành NaN (giá trị null của NumPy)\nfull_filter_data.replace(['nan', ''], np.nan, inplace=True)\n\n# Chuyển các cột sau sang float(sau khi xử lí missing value có thể cân nhắc chuyển về int)\nfull_filter_data['Consider count hodiday Saturday'] = full_filter_data['Consider count hodiday Saturday'].astype(float)\n\n# Chuyển các cột sau sang bool vì mang giá trị nhị phân\nbool_columns = ['Stock class', 'PRODUCT ATTRIBUTION', 'SPECIAL_DIV', 'DIRECT SHIP FLG','OTHER AREA SHIP DIV']\nfor col in bool_columns:\n    full_filter_data[col] = full_filter_data[col].astype(bool)\n# Chuẩn hóa các cột có kiểu string (object) về dạng chữ thường (lowercase)\nstring_columns = full_filter_data.select_dtypes(include=['object']).columns\nfor col in string_columns:\n    full_filter_data[col] = full_filter_data[col].str.lower()\nfull_filter_data.replace('nan', np.nan, inplace=True)\nfull_filter_data.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"b6a208f0","cell_type":"markdown","source":"<a id=\"preprocessing\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">Step 3 | Data Preprocessing</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"d0135fab-380a-430f-9a73-862b1f862718","cell_type":"code","source":"data_10_clean = full_filter_data.copy()","metadata":{},"outputs":[],"execution_count":null},{"id":"54f358b1","cell_type":"markdown","source":"<a id=\"removal\"></a>\n## <b><span style='color:darkorange'>Step 3.1 |</span><span style='color:#8502d1'> Irrelevant Feature Removal</span></b>","metadata":{}},{"id":"2c3f1915","cell_type":"markdown","source":"<div style=\"border-radius:10px; padding: 15px; background-color: #e2c9ff; font-size:110%; text-align:left\">\n\nSau quá trình EDA ở mục 4, ta sẽ tiến hành loại bỏ các thuộc tính sau đây, dựa vào mức độ ảnh hưởng thấp đến biến mục tiêu (<=0.3), thuộc tính không phù hợp để dự đoán vì chỉ có giá trị sau khi biết đơn hàng bị delay, cũng như các các thuộc tính (ngoài biến mục tiêu)có tính tương quan cao với nhau nhằm loại bỏ trường hợp đa cộng tuyến:\n\n+ 'SUBSIDIARY_CD', 'GLOBAL_NO', 'SHIP DECISION NO', 'REASON_CD', 'ALLOCATION QTY', 'SOUF_RCV_NO', 'QTUF_RCV_NO', 'PURCHASE AMOUNT', 'OTHER AREA SHIP DIV','BRAND_CD', 'PRODUCT ATTRIBUTION','SPECIAL_DIV', 'DIRECT SHIP FLG','SUPPLIER_DIV', 'PACK QTY', 'Sales order line number', 'LOGICAL PLANT', 'SO_TIME', 'SO_DAY_OF_WEEK', 'Stock class', 'PACKING RANK'","metadata":{}},{"id":"1eb24ec4","cell_type":"code","source":"data_10_clean.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"c757d3ab","cell_type":"code","source":"data_10_clean.columns","metadata":{},"outputs":[],"execution_count":null},{"id":"3bbd56f6","cell_type":"code","source":"# Tiến hành loại bỏ 1 số thuộc tính liên quan tới mã định danh chứa nhiều giá trị categorical và một số thuộc tính có mức độ ảnh hưởng <0.03\ndrop_cols = ['SUBSIDIARY_CD', 'GLOBAL_NO', 'SHIP DECISION NO', 'REASON_CD', 'ALLOCATION QTY', 'SOUF_RCV_NO', 'QTUF_RCV_NO', 'PURCHASE AMOUNT', 'OTHER AREA SHIP DIV','BRAND_CD', 'PRODUCT ATTRIBUTION','SPECIAL_DIV', 'DIRECT SHIP FLG','SUPPLIER_DIV', 'PACK QTY', 'Sales order line number', 'LOGICAL PLANT', 'SO_TIME', 'SO_DAY_OF_WEEK', 'Stock class', 'PACKING RANK']\nlen(drop_cols)","metadata":{},"outputs":[],"execution_count":null},{"id":"4918f735","cell_type":"code","source":"#Drop các cột không cần thiết\ndata_10_clean.drop(drop_cols, axis=1, inplace=True)\n# Kiểm tra lại các cột\ndata_10_clean.columns","metadata":{},"outputs":[],"execution_count":null},{"id":"8b8bf2e2","cell_type":"markdown","source":"## <b><span style='color:darkorange'>Step 3.2 |</span><span style='color:#8502d1'> Missing Value Treatment</span></b>","metadata":{}},{"id":"156d396d","cell_type":"code","source":"#Kiểm tra missing value\ndata_10_clean.isnull().sum()","metadata":{},"outputs":[],"execution_count":null},{"id":"99067d58","cell_type":"code","source":"#Tiến hành thay thế các giá trị null trong Ship Mode bằng giá trị 'UNK'\ndata_10_clean['Ship Mode'].fillna('UNK', inplace=True)\ndata_10_clean.isnull().sum()","metadata":{},"outputs":[],"execution_count":null},{"id":"5166c214","cell_type":"code","source":"#Nhận thấy Consider count hodiday Saturday và SUPPLIER_DIV đều có giá trị null chiếm rất ít, ta tiến hành loại bỏ các dòng này\ndata_10_clean.dropna(subset=['Consider count hodiday Saturday'], inplace=True)\ndata_10_clean.reset_index(drop=True, inplace=True)\ndata_10_clean.isnull().sum() ","metadata":{},"outputs":[],"execution_count":null},{"id":"24e6e8d4","cell_type":"code","source":"data_10_clean.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"56a7d056","cell_type":"markdown","source":"## <b><span style='color:darkorange'>Step 3.3|</span><span style='color:#8502d1'>Features Engineering</span></b>","metadata":{}},{"id":"cf180ea5","cell_type":"code","source":"# Tiến hành tạo thêm các thuộc tính mới cho các bộ dữ liệu\nlist_new_features = ['SO_MONTH','day_range', 'is_VSD_weekend', 'WEIGHT']    \ndef feature_engineering(data):\n    #Tạo thêm thuộc tính 'SO_MONTH' = tháng của 'Order date'\n    data['SO_MONTH'] = data['Order date'].dt.month\n    #Tạo thêm thuộc tính day_range = VSD - Order date (Khoảng cách giữa ngày nhận đơn và ngày dự kiến giao hàng cho khách)\n    data['day_range'] = (data['VSD'] - data['Order date']).dt.days\n\n    #Tạo thêm thuộc tính 'is_VSD_weekend' = 1 nếu VSD là thứ 7, ngược lại = 0 vì nhận thấy có sự khác biệt giữa lượng đơn hàng bị delay và không bị delay trong tuần và cuối tuần\n    #Thuộc tính này thay cho các thuộc tính thời \n    data['is_weekend_VSD'] = data['VSD'].dt.dayofweek.apply(lambda x: 1 if x in [5, 6] else 0)\n\n    #Tạo thêm thuộc tính WEIGHT = SO QTY * WEIGHT PER PIECE (Trọng lượng đơn hàng)\n    data['WEIGHT'] = data['SO QTY'] * data['WEIGHT PER PIECE']\n\n    #Tạo thêm thuộc tính Day_of_month_VSD, và MONTH_VSD cho thuộc tính VSD\n    data['Day_of_month_VSD'] = data['VSD'].dt.day\n    data['MONTH_VSD'] = data['VSD'].dt.month\n    \n    #Xóa thuộc tính 'Order date', 'VSD' sau khi đã tạo thêm thuộc tính 'day_range', 'is_VSD_weekend'\n    data.drop(columns=['Order date', 'VSD'], inplace=True)\n    \n    #Quy chuẩn datetime cho các thuộc tính 'Order date', 'VSD' về dạng %m-%Y\n    # data['Order date'] = data['Order date'].dt.strftime('%m-%Y')\n    # data['VSD'] = data['VSD'].dt.strftime('%m-%Y')\n    return data\n\n# Tạo thêm các thuộc tính mới\ndata_10_clean = feature_engineering(data_10_clean).copy()\ndata_10_clean.head(10)","metadata":{},"outputs":[],"execution_count":null},{"id":"0935bfd1","cell_type":"code","source":"#Các thuộc tính mới được tạo ra\nlist_new_features","metadata":{},"outputs":[],"execution_count":null},{"id":"5d4df655","cell_type":"code","source":"data_10_clean.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"6bbe37fa","cell_type":"markdown","source":"<a id=\"eda\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">EXPORT FINAL CLEAN DATA</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"9a388065","cell_type":"code","source":"data_10_clean.to_parquet('data_10_clean.parquet')","metadata":{},"outputs":[],"execution_count":null},{"id":"c3e41e74-fba2-40e5-a3a5-74ec4e27308f","cell_type":"code","source":"data_10_clean = pd.read_parquet('data_10_clean.parquet')\nindex_10 = data_10['ID']\ndata_10_clean.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"eba73778-25b8-469c-8685-e141d6a36b9a","cell_type":"code","source":"data_10_clean","metadata":{},"outputs":[],"execution_count":null},{"id":"9ee72040","cell_type":"markdown","source":"<a id=\"eda\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">TESTING DATA</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"8d3df050-97ba-4591-9042-452a9b2eee9c","cell_type":"markdown","source":"Sử dụng phương pháp ensemble learning cùng soft voting \n","metadata":{}},{"id":"d325a710","cell_type":"code","source":"def objective_catboost(trial, X_train, y_train, X_eval, y_eval, cat_features):\n    params = {\n        'iterations': trial.suggest_int('iterations', 100, 700),\n        'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.2),\n        'depth': trial.suggest_int('depth', 3, 7),\n        'l2_leaf_reg': trial.suggest_int('l2_leaf_reg', 1, 10),\n        'random_strength': trial.suggest_float('random_strength', 1, 15),\n        'loss_function': 'Logloss',\n        'verbose': 0,\n        'task_type': 'GPU',  \n        'random_seed': 42\n    }\n    model = CatBoostClassifier(**params)\n    model.fit(\n        X_train, y_train,\n        eval_set=(X_eval, y_eval),\n        cat_features=cat_features,\n        early_stopping_rounds=30\n    )\n    preds = model.predict(X_eval)\n    return f1_score(y_eval, preds, average='macro')\n\ndef objective_lgbm(trial, X_train, y_train, X_eval, y_eval):\n    params = {\n        'n_estimators': trial.suggest_int('n_estimators', 100, 700),\n        'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.2),\n        'num_leaves': trial.suggest_int('num_leaves', 20, 100),\n        'max_depth': trial.suggest_int('max_depth', 3, 7),\n        'min_child_samples': trial.suggest_int('min_child_samples', 5, 20),\n        'subsample': trial.suggest_float('subsample', 0.5, 1.0),\n        'colsample_bytree': trial.suggest_float('colsample_bytree', 0.5, 1),\n        'reg_alpha': trial.suggest_float('reg_alpha', 0.0, 0.1),\n        'reg_lambda': trial.suggest_float('reg_lambda', 0.0, 1),\n        'boosting_type': 'gbdt',\n        'objective': 'binary',\n        'metric': 'binary_logloss',  \n        'random_state': 42\n    }\n\n    model = LGBMClassifier(**params)\n    model.fit(\n        X_train, y_train,\n        eval_set=[(X_eval, y_eval)],\n        callbacks=[early_stopping(stopping_rounds=30), log_evaluation(period=0)]\n    )\n    preds = model.predict(X_eval)\n    return f1_score(y_eval, preds, average='macro')  # Tính F1 macro\n","metadata":{},"outputs":[],"execution_count":null},{"id":"3ed9eff6","cell_type":"code","source":"def train_and_ensemble_val(data_train, data_test, index, delay_not_delay_ratio):\n    X = data_train.drop(columns=['label'])\n    y = data_train['label']\n    X_test = data_test.copy()\n\n    # Undersample\n    n = delay_not_delay_ratio\n    delay_idx = y[y == 1].index\n    not_delay_idx = y[y == 0].sample(n=len(delay_idx)*n, random_state=42).index\n    undersample_idx = delay_idx.union(not_delay_idx)\n    X = X.loc[undersample_idx]\n    y = y.loc[undersample_idx]\n\n    # Categorical features\n    cat_features = X.select_dtypes(include=['object', 'category']).columns.tolist()\n    X[cat_features] = X[cat_features].astype('category')\n    X_test[cat_features] = X_test[cat_features].astype('category')\n\n    # Tách train/eval để dùng eval_set\n    X_train, X_eval, y_train, y_eval = train_test_split(X, y, test_size=0.2, stratify=y, random_state=42)\n    # LightGBM + eval_set\n    study_lgb = optuna.create_study(direction='maximize')\n    study_lgb.optimize(lambda trial: objective_lgbm(trial, X_train, y_train, X_eval, y_eval), n_trials=35)\n    best_lgb = LGBMClassifier(\n        **study_lgb.best_params,\n        callbacks=[early_stopping(stopping_rounds=50),log_evaluation(period=0)],\n        eval_metric='f1'\n    )\n    best_lgb.fit(X_train, y_train, eval_set=[(X_eval, y_eval)], callbacks=None)\n    joblib.dump(best_lgb, 'lgbm_model.pkl')\n    # CatBoost + eval_set\n    study_cb = optuna.create_study(direction='maximize')\n    study_cb.optimize(lambda trial: objective_catboost(trial, X_train, y_train, X_eval, y_eval, cat_features), n_trials=50)\n    best_cb = CatBoostClassifier(\n        **study_cb.best_params,\n        cat_features=cat_features,\n        early_stopping_rounds=50,\n        eval_metric='F1',\n        verbose=0\n    )\n    best_cb.fit(X_train, y_train, eval_set=(X_eval, y_eval))\n    joblib.dump(best_cb, 'catboost_model.pkl')\n    # Ensemble\n    ensemble = VotingClassifier(estimators=[\n        ('catboost', best_cb),\n        ('lgbm', best_lgb)\n    ], voting='soft')\n    ensemble.fit(X_train, y_train)\n\n    # Predict & export\n    preds_proba = ensemble.predict_proba(X_test)[:, 1]\n    preds = (preds_proba >= 0.5).astype(int)\n    submission = pd.DataFrame({\n        'ID': index,\n        'label': preds\n    })\n    submission.to_csv('ensemble_submission.csv', index=False)\n    joblib.dump(ensemble, 'ensemble_model.pkl')\n","metadata":{},"outputs":[],"execution_count":null},{"id":"28907cb3","cell_type":"code","source":"train_and_ensemble_val(data_7_9, data_10_clean, index_10, delay_not_delay_ratio=10)","metadata":{},"outputs":[],"execution_count":null},{"id":"dfaaadd2","cell_type":"code","source":"#load model catboost\npath_cb = r'D:\\Python_code\\DS_108_BTVN\\Kaggle\\best_model\\best\\catboost_model.pkl'\ncatboost_model = joblib.load(path_cb)\n#load model lgbm\npath_lgbm = r'D:\\Python_code\\DS_108_BTVN\\Kaggle\\best_model\\best\\lgbm_model.pkl'\nlgbm_model = joblib.load(path_lgbm)\n#load model ensemble\npath_ensemble = r'D:\\Python_code\\DS_108_BTVN\\Kaggle\\best_model\\best\\ensemble_model.pkl'\nensemble_model = joblib.load(path_ensemble)\n# Lấy ra các tham số của mô hình CatBoost\ncatboost_params = catboost_model.get_params()\n# In ra các tham số của mô hình CatBoost\nprint(\"Các tham số của mô hình CatBoost:\")\nprint(catboost_params)\n# Lấy ra các tham số của mô hình LGBM\nlgbm_params = lgbm_model.get_params()\n# In ra các tham số của mô hình LGBM\nprint(\"Các tham số của mô hình LGBM:\")\nprint(lgbm_params)\n# Lấy ra các tham số của mô hình Ensemble\nensemble_params = ensemble_model.get_params()\n# In ra các tham số của mô hình Ensemble \nprint(\"Các tham số của mô hình Ensemble:\")\nprint(ensemble_params)\n","metadata":{},"outputs":[],"execution_count":null},{"id":"7a6f313d","cell_type":"markdown","source":"Kết thúc phần code được dùng để sử dụng trên cuộc thi Kaggle DS108, từ phần EXPERIMENT trở đi là các thử nghiệm khác nhau ","metadata":{}},{"id":"53c9def9","cell_type":"markdown","source":"<a id=\"eda\"></a>\n# <p style=\"background-color:#8502d1; font-family:calibri; color:white; font-size:150%; text-align:center; border-radius:15px 50px;\">EXPERIMENT</p>\n\n⬆️ [Tabel of Contents](#contents_tabel)","metadata":{}},{"id":"9f9e5aa8","cell_type":"markdown","source":"<div style=\"border-radius:10px; padding: 15px; background-color: #e2c9ff; font-size:110%; text-align:left\">\n\n## <h2 align=\"left\"><font color=#8502d1>TRAINING PIPELINE</font></h2>\n    \n","metadata":{}},{"id":"d0337159","cell_type":"markdown","source":"![image.png](attachment:image.png)","metadata":{},"attachments":{"image.png":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAB+8AAAVSCAYAAADHJ7k2AAAgAElEQVR4Aezd57Mc133/+csAJpAESRDMJJjFHMCcCYJgzhQzCQYxkyAJRjCDCSRlyz/Lli1bthV+li3bsmXLlrMlW/bTfbxV+2gfeKt2/4B9sLW1Vb31Hampwdzuud0zPXNPd79QhZq5PTOnz3l/P/fO93w/fU4vLCwsZP5jQAM0QAM0QAM00BINvLXgHwIIIIAAAggg0AyB32tJ/qNuo3ZFAzRAAzRAAzTw382kP1pBAAEEEGgDAV/8vvhpgAZogAZogAbaogHmfRuyS31EAAEEEECgHQSY93LgtuTA+kmrNEADNEADzPt25Jd6iQACCDRCwBe/L34aoAEaoAEaoIG2aIB530j6pxEEEEAAAQQQWFhYYN7LgduSA+snrdIADdAADTDvpa8IIIBAjwj44vfFTwM0QAM0QAM00BYNMO97lKQaKgIIIIAAAjMmwLyXA7clB9ZPWqUBGqABGmDezzgx1DwCCCCQEgFf/L74aYAGaIAGaIAG2qIB5n1KWaS+IIAAAggg0G4CzHs5cFtyYP2kVRqgARqgAeZ9u/NOvUcAAQRqEfDF74ufBmiABmiABmigLRpg3tdK87wZAQQQQAABBMYQYN7LgduSA+snrdIADdAADTDvxyR1XkIAAQS6RsAXvy9+GqABGqABGqCBtmiAed+1TNR4EEAAAQQQWD4CzHs5cFtyYP2kVRqgARqgAeb98uWMzowAAgjMnYAvfl/8NEADNEADNEADbdEA837uqaITIoAAAggg0FkCzHs5cFtyYP2kVRqgARqgAeZ9Z1NSA0MAAQQWExj7xX/9jc9mX3nif/iPAQ3QAA3QAA3QwMw1sGbNUWPzkoWFBeb94lzOEQQQQAABBBCYjMBY8/7oY86cee6j3qLeRAM0QAM0QAM0EBq47Ir7lqqHMO8ny/d8CgEEEGglgbFfCvHF8elX/8t/DGiABmiABmiABmaugSOOPHlsXsK8b2WuqdMIIIAAAgikSmCseX/a6VfMPPdRb1FvogEaoAEaoAEaCA18+e43l6qHMO9TzSj1CwEEEJgBgbFfCsx7yYMEkgZogAZogAbmpQHm/QwyPU0igAACCCCAQBkB5r2LU12gQQM0QAM0QANJaIB5X5auOY4AAgj0kwDzXoKSRIIyL2PIeZiQNEADNJCuBpj3/UxGjRoBBBBAAIFlIsC8Vw9RD6EBGqABGqCBJDTAvF+mbNBpEUAAgUQJMO8lKEkkKMy0dM00sREbGqCBeWmAeZ9otqhbCCCAAAIIdJMA8149RD2EBmiABmiABpLQAPO+m8mmUSGAAAKTEmDeS1CSSFDmZQw5DxOSBmiABtLVAPN+0nTO5xBAAAEEEEBgAgLMe/UQ9RAaoAEaoAEaSEIDzPsJMjkfQQABBDpMgHkvQUkiQWGmpWumiY3Y0AANzEsDzPsOZ5yGhgACCCCAQHoEmPfqIeohNEADNEADNJCEBpj36SWKeoQAAggsJwHmvQQliQRlXsaQ8zAhaYAGaCBdDTDvlzMldG4EEEAAAQR6R4B5rx6iHkIDNEADNEADSWiAed+7PNSAEUAAgbEEmPcSlCQSFGZaumaa2IgNDdDAvDTAvB+bs3kRAQQQQAABBJolwLxXD1EPoQEaoAEaoIEkNMC8bzbJ0xoCCCDQdgLMewlKEgnKvIwh52FC0gAN0EC6GmDetz2t1H8EEEAAAQRaRYB5rx6iHkIDNEADNEADSWiAed+qHFJnEUAAgZkTYN5LUJJIUJhp6ZppYiM2NEAD89IA837meZ8TIIAAAggggMCvCDDv1UPUQ2iABmiABmggCQ0w73+VoHmGAAIIILCwwLyXoCSRoMzLGHIeJiQN0AANpKsB5r3UFAEEEEAAAQTmSIB5rx6iHkIDNEADNEADSWiAeT/HDNCpEEAAgRYQYN5LUJJIUJhp6ZppYiM2NEAD89IA874FmaMuIoAAAggg0B0CzHv1EPUQGqABGqABGkhCA8z77iSYRoIAAgg0QYB5L0FJIkGZlzHkPExIGqABGkhXA8z7JlI7bSCAAAIIIIBARQLMe/UQ9RAaoAEaoAEaSEIDzPuK2Zu3IYAAAj0hwLyXoCSRoDDT0jXTxEZsaIAG5qUB5n1Psk/DRAABBBBAIA0CzHv1EPUQGqABGqABGkhCA8z7NJJDvUAAAQRSIcC8l6AkkaDMyxhyHiYkDdAADaSrAeZ9KumhfiCAAAIIINALAsx79RD1EBqgARqgARpIQgPM+17kngaJAAIIVCbAvJegJJGgMNPSNdPERmxogAbmpQHmfeX8zRsRQAABBBBAYHoCzHv1EPUQGqABGqABGkhCA8z76RM7LSCAAAJdIsC8l6AkkaDMyxhyHiYkDdAADaSrAeZ9l1JMY0EAAQQQQCB5Asx79RD1EBqgARqgARpIQgPM++TzRh1EAAEE5kqAeS9BSSJBYaala6aJjdjQAA3MSwPM+7nmgE6GAAIIIIBA3wkw79VD1ENogAZogAZoIAkNMO/7npYaPwIIILAjAea9BCWJBGVexpDzMCFpgAZoIF0NMO93TNL8hAACCCCAAAIzJcC8Vw9RD6EBGqABGqCBJDTAvJ9pzqdxBBBAoHUEmPcSlCQSFGZaumaa2IgNDdDAvDTAvG9dHqnDCCCAAAIItJkA8149RD2EBmiABmiABpLQAPO+zSmlviOAAALNE2DeS1CSSFDmZQw5DxOSBmiABtLVAPO++URPiwgggAACCCBQSoB5rx6iHkIDNEADNEADSWiAeV+ar3kBAQQQ6CUB5r0EJYkEhZmWrpkmNmJDAzQwLw0w73uZixo0AggggAACy0WAea8eoh5CAzRAAzRAA0logHm/XOmg8yKAAAJpEmDeS1CSSFDmZQw5DxOSBmiABtLVAPM+zWRRrxBAAAEEEOgoAea9eoh6CA3QAA3QAA0koQHmfUezTcNCAAEEJiTAvJegJJGgMNPSNdPERmxogAbmpQHm/YTZnI8hgAACCCCAwCQEmPfqIeohNEADNEADNJCEBpj3k6RyPoMAAgh0lwDzXoKSRIIyL2PIeZiQNEADNJCuBpj33U04jQwBBBBAAIEECTDv1UPUQ2iABmiABmggCQ0w7xPMFHUJAQQQWEYCzHsJShIJCjMtXTNNbMSGBmhgXhpg3i9jRujUCCCAAAII9I8A8149RD2EBmiABmiABpLQAPO+f4moESOAAALjCDDvJShJJCjzMoachwlJAzRAA+lqgHk/LmXzGgIIIIAAAgg0TIB5rx6iHkIDNEADNEADSWiAed9wlqc5BBBAoOUEmPcSlCQSFGZaumaa2IgNDdDAvDTAvG95Vqn7CCCAAAIItIsA8149RD2EBmiABmiABpLQAPO+XUmk3iKAAAKzJsC8l6AkkaDMyxhyHiYkDdAADaSrAeb9rNM+7SOAAAIIIIDAEAHmvXqIeggN0AAN0AANJKEB5v1QhuYpAggggMAC816CkkSCwkxL10wTG7GhARqYlwaY9zJTBBBAAAEEEJgjAea9eoh6CA3QAA3QAA0koQHm/RwzQKdCAAEEWkCAeS9BSSJBmZcx5DxMSBqgARpIVwPM+xZkjrqIAAIIIIBAdwgw79VD1ENogAZogAZoIAkNMO+7k2AaCQIIINAEAea9BCWJBIWZlq6ZJjZiQwM0MC8NMO+bSO20gQACCCCAAAIVCTDv1UPUQ2iABmiABmggCQ0w7ytmb96GAAII9IQA816CkkSCMi9jyHmYkDRAAzSQrgaY9z3JPg0TAQQQQACBNAgw79VD1ENogAZogAZoIAkNMO/TSA71AgEEEEiFAPNegpJEgsJMS9dMExuxoQEamJcGmPeppIf6gQACCCCAQC8IMO/VQ9RDaIAGaIAGaCAJDTDve5F7GiQCCCBQmQDzXoKSRIIyL2PIeZiQNEADNJCuBpj3lfM3b0QAAQQQQACB6Qkw79VD1ENogAZogAZoIAkNMO+nT+y0gAACCHSJAPNegpJEgsJMS9dMExuxoQEamJcGmPddSjGNBQEEEEAAgeQJMO/VQ9RDaIAGaIAGaCAJDTDvk88bdRABBBCYKwHmvQQliQRlXsaQ8zAhaYAGaCBdDTDv55oDOhkCCCCAAAJ9J8C8Vw9RD6EBGqABGqCBJDTAvO97Wmr8CCCAwI4EmPcSlCQSFGZaumaa2IgNDdDAvDTAvN8xSfMTAggggAACCMyUAPNePUQ9hAZogAZogAaS0ADzfqY5n8YRQACB1hFg3ktQkkhQ5mUMOQ8TkgZogAbS1QDzvnV5pA4jgAACCCDQZgLMe/UQ9RAaoAEaoAEaSEIDzPs2p5T6jgACCDRPgHkvQUkiQWGmpWumiY3Y0AANzEsDzPvmEz0tIoAAAggggEApAea9eoh6CA3QAA3QAA0koQHmfWm+5gUEEECglwSY9xKUJBKUeRlDzsOEpAEaoIF0NcC872UuatAIIIAAAggsFwHmvXqIeggN0AAN0AANJKEB5v1ypYPOiwACCKRJgHkvQUkiQWGmpWumiY3Y0AANzEsDzPs0k0W9QgABBBBAoKMEmPfqIeohNEADNEADNJCEBpj3Hc02DQsBBBCYkADzXoKSRIIyL2PIeZiQNEADNJCuBpj3E2ZzPoYAAggggAACkxBg3quHqIfQAA3QAA3QQBIaYN5Pksr5DAIIINBdAsx7CUoSCQozLV0zTWzEhgZoYF4aYN53N+E0MgQQQAABBBIkwLxXD1EPoQEaoAEaoIEkNMC8TzBT1CUEEEBgGQkw7yUoSSQo8zKGnIcJSQM0QAPpaoB5v4wZoVMjgAACCCDQPwLMe/UQ9RAaoAEaoAEaSEIDzPv+JaJGjAACCIwjwLyXoCSRoDDT0jXTxEZsaIAG5qUB5v24lM1rCCCAAAIIINAwAea9eoh6CA3QAA3QAA0koQHmfcNZnuYQQACBlhNg3ktQkkhQ5mUMOQ8TkgZogAbS1QDzvuVZpe4jgAACCCDQLgLMe/UQ9RAaoAEaoAEaSEIDzPt2JZF6iwACCMyaAPNegpJEgsJMS9dMExuxoQEamJcGmPezTvu0jwACCCCAAAJDBJj36iHqITRAAzRAAzSQhAaY90MZmqcIIIAAAgvMewlKEgnKvIwh52FC0gAN0EC6GmDey0wRQAABBBBAYI4EmPfqIeohNEADNEADNJCEBpj3c8wAnQoBBBBoAQHmvQQliQSFmZaumSY2YkMDNDAvDTDvW5A56iICCCCAAALdIcC8Vw9RD6EBGqABGqCBJDTAvO9OgmkkCCCAQBMEmPcSlCQSlHkZQ87DhKQBGqCBdDXAvG8itdMGAggggAACCFQkwLxXD1EPoQEaoAEaoIEkNMC8r5i9eRsCCCDQEwLMewlKEgkKMy1dM01sxIYGaGBeGmDe9yT7NEwEEEAAAQTSIMC8Vw9RD6EBGqABGqCBJDTAvE8jOdQLBBBAIBUCzHsJShIJyryMIedhQtIADdBAuhpg3qeSHuoHAggggAACvSDAvFcPUQ+hARqgARqggSQ0wLzvRe5pkAgggEBlAsx7CUoSCQozLV0zTWzEhgZoYF4aYN5Xzt+8EQEEEEAAAQSmJ8C8Vw9RD6EBGqABGqCBJDTAvJ8+sdMCAggg0CUCzHsJShIJyryMIedhQtIADdBAuhpg3ncpxTQWBBBAAAEEkifAvFcPUQ+hARqgARqggSQ0wLxPPm/UQQQQQGCuBJj3EpQkEhRmWrpmmtiIDQ3QwLw0wLyfaw7oZAgggAACCPSdAPNePUQ9hAZogAZogAaS0ADzvu9pqfEjgAACOxJg3ktQkkhQ5mUMOQ8TkgZogAbS1QDzfsckzU8IIIAAAgggMFMCzHv1EPUQGqABGqABGkhCA8z7meZ8GkcAAQRaR4B5L0FJIkFhpqVrpomN2NAADcxLA8z71uWROowAAggggECbCTDv1UPUQ2iABmiABmggCQ0w79ucUuo7Aggg0DwB5r0EJYkEZV7GkPMwIWmABmggXQ0w75tP9LSIAAIIIIAAAqUEmPfqIeohNEADNEADNJCEBpj3pfmaFxBAAIFeEmDeS1CSSFCYaemaaWIjNjRAA/PSAPO+l7moQSOAAAIIILBcBJj36iHqITRAAzRAAzSQhAaY98uVDjovAgggkCYB5r0EJYkEZV7GkPMwIWmABmggXQ0w79NMFvUKAQQQQACBjhJg3quHqIfQAA3QAA3QQBIaYN53NNs0LAQQQGBCAsx7CUoSCQozLV0zTWzEhgZoYF4aYN5PmM35GAIIIIAAAghMQoB5rx6iHkIDNEADNEADSWiAeT9JKuczCCCAQHcJMO8lKEkkKPMyhpyHCUkDNEAD6WqAed/dhNPIEEAAAQQQSJAA8149RD2EBmiABmiABpLQAPM+wUxRlxBAAIFlJMC8l6AkkaAw09I108RGbGiABualAeb9MmaETo0AAggggED/CDDv1UPUQ2iABmiABmggCQ0w7/uXiBoxAgggMI4A816CkkSCMi9jyHmYkDRAAzSQrgaY9+NSNq8hgAACCCCAQMMEmPfqIeohNEADNEADNJCEBpj3DWd5mkMAAQRaToB5L0FJIkFhpqVrpomN2NAADcxLA8z7lmeVuo8AAggggEC7CDDv1UPUQ2iABmiABmggCQ0w79uVROotAgggMGsCzHsJShIJyryMIedhQtIADdBAuhpg3s867dM+AggggAACCAwRYN6rh6iH0AAN0AAN0EASGmDeD2VoniKAAAIILDDvJShJJCjMtHTNNLERGxqggXlpgHkvM0UAAQQQQACBORJg3quHqIfQAA3QAA3QQBIaYN7PMQN0KgQQQKAFBJj3EpQkEpR5GUPOw4SkARqggXQ1wLxvQeaoiwgggAACCHSHAPNePUQ9hAZogAZogAaS0ADzvjsJppEggAACTRBg3ktQkkhQmGnpmmliIzY0QAPz0gDzvonUThsIIIAAAgggUJEA8149RD2EBmiABmiABpLQAPO+YvbmbQgggEBPCDDvJShJJCjzMoachwlJAzRAA+lqgHnfk+zTMBFAAAEEEEiDAPNePUQ9hAZogAZogAaS0ADzPo3kUC8QQACBVAgw7yUoSSQozLR0zTSxERsaoIF5aYB5n0p6qB8IIIAAAgj0ggDzXj1EPYQGaIAGaIAGktAA874XuadBIoAAApUJMO8lKEkkKPMyhpyHCUkDNEAD6WqAeV85f/NGBBBAAAEEEJieAPNePUQ9hAZogAZogAaS0ADzfvrETgsIIIBAlwgw7yUoSSQozLR0zTSxERsaoIF5aYB536UU01gQQAABBBBIngDzXj1EPYQGaIAGaIAGktAA8z75vFEHEUAAgbkSYN5LUJJIUOZlDDkPE5IGaIAG0tUA836uOaCTIYAAAggg0HcCzHv1EPUQGqABGqABGkhCA8z7vqelxo8AAgjsSIB5P+cE5alnvpHd/+CHE/1/5LGvZk8/+7vZSy9/L9v61l9ln3z28ySSC0ZYukaY2IgNDdBAmzTAvN8xSfMTAggggAACCMyUAPN+zvWQNuWlTfT1sSd+Y6Laz6OPfy37yhP/I3viqa9nL2z5TvbmO3+dffzpv6v/0CsN0AANdFgDzPuZ5nwaRwABBFpHgHk/5y/9E790wVjmCwsLlV/fbbc9sqOPOTO79PJ7sk2PfJZ9tP1nkrg5x7OJCb02mKs0QAM08AsNMO9bl0fqMAIIIIAAAm0mwLw3f55pDeXQw06oXN+pUgvae+8DsmOOPTM774Kbs1tue2lg7G//3KIOcynzaRqggS5ogHnf5pRS3xFAAIHmCYydSMSVvl348ktpDE2a96OTuz333Cdbf9VD2bvbfiJuihA0kIAGIvG+4soHFv3f8sofi08C8Unpu0FfmPfNp3haRAABBBBAAIElCDDv5eQznZc1bd6P1oDi5z323Ds786wN2YObPs4+/OTfZjqelOcssTNl0dz7yvUP9pZJyvFKqW+vb/2LQu3cevsrtOM7Yq4aYN4vkbV5GQEEEOgZAeb9nBORWZr3+URu5cpV2b33b5trgpFS4q0vrrhNRQNfOumiwr+xDz38qd/POf/tTUUT+jH+75OV9z3LQg0XAQQQQACB5SXAvJeTz3ReNg/zPq8DxePue6zMLr70y9nWt38003GlOKf5+NOfFc69d9ppp96xSDE+Kffpyad/u1A7a9eeRju+I+aqAeb98iaFzo4AAgikRqAwQcmTfyvvx5sMkySf8zDv8/jFVcfbP//PuSYakzDxmeZ1hmkaTJn3acTB70N74sC8Ty1N1B8EEEAAAQQ6TYB5z5iZab1k3uZ9XgtasWL3bMPGR7OPtv90puNLaZ7FvG/PnC8l3URfmPe0k4ommfedzjkNDgEEEKhNgHk/58nqPM37mLhduf6h3kzWUkm29EPin2uAeU8LuRY8VtMC8752HucDCCCAAAIIIDA5Aeb9nOshfcuJl8u8z038A9ccmb245bu9qAkx76vNt/r2O1hlvMx72qmik3m8h3k/eULnkwgggEAXCTDv5zxZLTPvDzjgsOy9D/5h7P+tb/3VYOL1yGNfzdZv2JQdfMixY+MXE7bYImzTI5/1YrI2j0TKOST1dTTAvKeXOnrx3v/KmPddTDWNCQEEEEAAgWQJMO/nXA/pW75bZt4ftfa0sbWfqA3Ffbhf2/rn2eYXv51FDeiOL7+RXXrZ3dlxx6/LYmV9btAv9bhixR7Z/Q9+2PmaEPPe3HvSvy/Me9qZVDtNf455n2y+qGMIIIDAshAYm/DbNr/5BKbMvD/wwCNqT6ZiS/xHH/9aFsb/uAnbfvsd3Kvt0ppOnrTX/O9BX5gy72mnL1pvapzM+2XJBZ0UAQQQQACBvhJg3jPva9dh6uS9Zeb90cecOdV5P9r+s+yxJ34jW3fO9dluu+0xth4UtaKdd945u++BbVOds864l+O9zHtz70l1x7ynnUm10/TnmPd9TUeNGwEEECgmMDbJZ943n8A0ad7nScL7H/xjduxxZ4+N5e13vtbpiVrOwmPzmsV0cqbM+8nZ0V0/2THvi5M1RxFAAAEEEEBgJgSY98z7mdZJZmXeD8+V3nn/77LLr7w/22WXFWNrQjvvvMvg3t7Dn+3Sc+Z9P+ePTWiYeU87TeioiTaY9zPJ9TSKAAIItJbA2OSeed98AjML8z4ShHe3/X22evXhpfE86qhTZzopbSJJ0UbzesN0eZky75eXP/23jz/zvrX5pI4jgAACCCDQRgLMe+b9TOsk8zDv8znPSy9/Lzvk0ONKa0KxAn/ffQ/M3n7vb2c65rw/835k3rdv7jdvjZSdj3lPO2XamPdx5n0bU0l9RgABBGZHYGxiz7xvPoGZlXkfCcUDD31UGs+ddtqps5O0eSdTztf870VXmTLvaaWr2p7VuJj3s0v4tIwAAggggAACiwgw75n3MzWy52neR36+7aN/zk444bzSulAY+Gevu3amY57VPGGpdpn35t5LaaTsdeY97ZRpY97HmfeL8jQHEEAAgV4TGJvUM++bT2Bmad5v//zn2f77H1Ia002PfNrIJO3jT/89i//zTmKmPd8nn/08i3vDTdtO0ec//ORfZ9Ju0bmaOhYsPvnsP2bW74+2/3RmbU/LIIoa8fsybTtLfb4N5n1od5Y6WIpR3ddn8bv2wcf/km3//D9nroe6Y+3j+5n3vc5JDR4BBBBAAIF5E2DeM+9nOgeYt3kf84eYL8XOi2HUF/2PhR3Pbv7WTMcd/QgzfZ7zzJTN+9RraB98HDWB2ddH6s5vg9uHn/zbzLXKvG++9l031t7/ixgw7+edBjofAgggkDaBwmQ+T/CZ980nMLM07yPZOe/8m0pjetMtL1ROesPMevTxr2XXXPdkdtbZ12SHHX5itnLlqmzXXX91H7Xddtsz22+/g7OYkJ551obsxps3Zy+/+v3K56ianMUk7PkX/nDR/xe3fLf0XC+/9ifZDTc+m518yiWD2wkM3/8t+n3IIcdmr7z2p6WfH+1bTGQ2v/jt7JbbtmQXXHhbduxxZw+2nNt9972+4B1s9tnngOygg9ZmEef1GzZlTzz19Zlc6BCci5hE3Eb7Hj+/+c5fZ7fe/nJ2yqmX/ZLHrl/0O3gceeTJ2SWX3Z0998IfFH6+qM3hY89t/lZ2xZUPZMcety7be+/9sygIxN+RFSt2H1xQEib27Xe+NvfdH0IHofszztww2EJwOF7Rv1133W0Qr9NOv3Lwvjfe/OFE4w8WMbEcjcnao0//gnP+dzUeb7zp+UXvHf3s8M+vb/2LifuVx+nNt3+U3XHn69m6c67PDj/iS9lee+27Q9922WXXgTZOOvmi7Nrrn8xi28X8s008vvrGnxWO+b0P/qHwPKHl2E0k/qaFobvnnvt80d+4X2P87dl47eOFny3qb8Tnkce+Ovi9jDHuf8ChO7QZcdlpp50Hx9asOSo74cTzs/MuuHnwOx/6js8XtetYs9+TzPu0k0a9QwABBBBAoGMEmPfM+5nm+Mth3sf8ZOtbf5XtsefeX8yfhuei8fzkUy5tZNxxUfwjj/1ads21T2RnnnV1dtjhJ2R77bVj3Sjm4DF3i5rSWWdvzG6+5cXsldd/MPX533n/73aYXz7z/O8VjjdqE8Nz6yrPy+ao4+Z+O9bQNo6toQWn4HVT1NBe+5OpWYzrV/5aLF54/MnfzK7e+Ngv60JHDOa+MQce1kfMu1evPiI77vhzsnPPu3HQx6ef+2Y2y8UZccHJg5s+GdSkjjn2rGzVqjWDuflwv/bZd3V29DFnZpddfm/22ONfm+rCkNff/OEOmog6yfC58udxG4oqehl+j7pBs/WBXL99eWTedyzLNBwEEEBgSgKFCUqeqDDvm086Zm3eh0Gbx2/08aoND4+dFIQhHIl8XKUd5tjo56v+HJOyu+55u7GVzTHxLDp3GOWjCVxMKo4/4dzC94+2sZRRHRO2O+/aOrgAYNzEd7Td0YALLMAAACAASURBVJ/DzL5y/UPZu9t+sqi/o/2v+vORR51SOMannv2dHc7x1jt/k51z7g3Z8MULo/0b/TkuTHjplf+5Qztl/QozNCaeo22U/RwXOFx62d2D1QBlbU57PCZLd93zVq1+5f2NiX3o54mnfqvS+If7GhdI5O00/Xj2umtq9yf6FqvJH37088FFFaOT8ip9jN/l+x/8sJFV6aeedkUhn/se2LbD2N7/8J+yK9c/OLbYlPe9Cpe4p+Mll96VrVy5X+H587aWelyxYo8sxnDv/dvmcvX/sLb69Jx5P2VW5+MIIIAAAgggUIcA8555v8NcpOm8e7nM+xjHbXe8Wjr/2XnnnbNJLxCPWsFVVz8yWAAwTd0oFhDce//7E9eNxtW+lprbLfV6XEReRQtbXv1+tmHjo1PX0OLi+rvvfWdiFuP6GvPr9Vc9lO27ak2pHpbiEa9HTSkWZXz57reybR/+UyU+4/oVrwW/uEBgt932qN23vfc+YHDRSFxAstR5Rl+PW0dUGfMk74lFE6Pn83PztfWuMmXe10nhvBcBBBDoPoGxCQvzvvkEY9bm/X0PfFAa01gxXpTgxKrSsknlJMlq/pkwdauawEX9yo9VMe9jdf6FF93+xYrvvA/jHsu2igvzN1bs1zG8x50nfy0uAIiLGvJxTfNYxbx/4KGPKxmgef+GH2NF+qipOtzfmCBNM+GJK5jjSvnhNpt4HjsdxKrp4bFM+jx2k3h3299X7mNq5v2LL38vW7v2tEZYxNXvr23988osimJZxbx/7InfGOxoUTVmsStI0bnyY7Hbw/CK/artLvW+J5/+xtjz5uf3WP87lHnf/cTTCBFAAAEEEEiIAPOeeT/TvL6szhIriGc9V4i6xurVh5fOB6++5iu1+hALBWIHw6XmSnVfDxO/zq6IObflNO+fff73s0MPPb5xFnHx/JZX/rhWXHIeRY9xIX4seqkbk6XeH3XHovNVPfb+B/+YXXjRbYtW1y913qLX46KEhx/9aq3+TFPLKurD8DHmff0aQFXd9OF9zPuEMkRdQQABBBIgMDaJY943n3TM2ryP+9oPJ47Dz8859/rChDau8B1+X5PPY7VqrM6eJslayryP+3PFdu11+x0TrqJ+xT216rZV5/0XXXzn1FdUL2Xex0R2kpXWw+OIVehFk7J33v/JRKvah9uO5zHZrWOOF8UqPxa3NYjbFOTb9Y+ea9Kfo9hR9VYQKZn3YVoP3+Ji0vEPfy622Z9kR4I8RkuZ97FbQt2VG7HVYN7+6GPcxmG4/00+Z943/92Yx495n0BmqAsIIIAAAgj0hwDznnlfOp/I89NpHpfTvI9+x60Ey+ZBkXfXGdu4lfxl56h6PLbWjy3d6/RnOc37WH1edWx13xer0ONWBHVYFL03bmvZdH0kH0tRnaioD0XH4naUcQu7vK0mHmOcG2tcjMK8n918vijmjlXnzbzvTwJqpAgggEAVAmMTJuZ99S/YqsnIrM37cV/0l11xX+EEYJbmfSTisYL9yad/u/DcVbiNM+/DtI1V8pMk/HFPtKLzz9q8j75eevk9hecu6k/RsXHmfVxdPa1xn/OMe9YPX/kdF0qUFSDyz9R5POPMq6biEGw++ew/slglX+e8dd4bV6pXuQ9dKuZ9bGFYZ3x13hsX48TuBkWaXOrYOPM+LvCpa9xHvyPuRee97oanazOoc7ED877578Y8jsz7Kqmb9yCAAAIIIIBAQwSY98z7wvlEnptO+1g2d57Hyvvo+9vv/bi0NhCGZ9xirOoYZ2nex9wudv8rW2BR1MeumvfBImpoTz0z+W5vcWF8nXn+L/ivqPyZSc37qAtOskV+1bHE7QGKtDJ6jHk/u/n8KGs/12M9rqb/y9+D/24o/9EMAggggEALCIxNzpj39b5kqyQlszbvN4wx7sLQKupjkXkfk4W493mYzLHVe0yiXn3jz7JYdR0GbkwCY2uz2Jr98ivuy/bf/5CxWor7vsdni86/1LFx5n3ZleR77bUqiy2142rjGN+mRz4b3Kv6mmufyE488fzBZOjp575Z2J8y8/7gg48Z3I/r5ltfGlyMEPfninvKB4/3PviHwbbiYWxef8MzWWwzvtRVzg9u+qTw/EvxiNfLzPuI1W677bkoFlE0CFM3OMRFC7Gl+uNPfn1wJf5RR5266P3Dk6PY1SDOGfdPD7N9+LV4HobuKademt12xyvZY49/LXtxy3ezuCVB7AJx8aVfziL2o58Z/vnRr/z6xByiX+vOuX5s+1EEOOPMDYP+bX7xj7Ktb/8oi23/33r3x4O+RhHipJMvGhuvAw88YhDjcbH54ON/yeJ3bPh/2Rb+6865bof3DX+m6Hnct37cufPX4rPDbEefx70N4z518XvxzHPfzN5484cDFnELg7hI45773s1iNXv8/o9+Nv85tqGfZFvDMvP+5ltezGJVf95+/hh9jW3/429a7CTw0MOfDv7eRLzOu+DmbNWqNYO45mPPH+PvVFx0krcz+rjf/odkcSFTbK336hs/yOKWG/lnY2vJd7f9ZKCLuP/ilesfzNYefXoWfRluh3nf/HdjHgPmfQsyR11EAAEEEECgOwSY98z7L+YCeT7a5ONym/cxlrif+vBcZvh5nV0Si8z7uAD6uOPXDeZXd9/77qAOELdbizlVXjeKC+FjgcFll9+brdrvoNK+RL9iC/S4R3uVGMQK7uG58zXXPVnYdtRlht9X5flSF+8Xrbz/RQ1t3aCGFjWoqInE3HSURV5Di3npcCxGn8cigkl2Koy5fdTDRtvLf457xV98yZ2D+XWM86PtP/2CdyyMiHPG7S8jZlFDivjussuuO7Q3iXn/3At/MKgd5f0oeozbK4YB/+jjXxvUHIJd1G6CYxy75LK7s332Xb1DX0bbufOurV+Mp0xHUV8Z1sEFF95a2GbEaPh9VZ5Hf8vO6/js6ghdYcu8706CaSQIIIBAEwQKE5Q8+WHeN59YzNq8jyu48/iNPsYEoyihyc37WK0dq9hj0hX3oCp6b9mx7Z//PHvo4e1jJ2Pnnn9TrTbzc5WZ92Eax//hccZk8J773stiRX7++aLHWCVddpX5sHl/8CHHDpL1V17/wdj2is4R263HhQLD/Rt+vu++B2YffvKvtduNc5WZ9ytX7rfD+Q46aO1gklPUv+FjMfkqMv3z/obRn+skPxZ6Oe/8mwYm+HBbo89DS+OubD7s8BMmYhDnueGm53YYb963eIyV3HFhSVxoMtqnop+ff+EPS7lGe0vdX72ozTDKh/uUPw8juuj90xyLCei4HRfigoEwq6ucIwoucUFG3t/RxzBYl/odGz1PmXm/cuWOhYUwys+/4JYldRXnj4tQRs8Tmhztb/wcxZS4UCAM+tHPLPVzFEBiEpUby8z75r8b8xjkjIti+MtjbzWR/GgDAQQQQAABBBBYWFhg3jPva88N8ry1ymMK5n1cuFyWW2+89vHK48/N+5ivnXra5Vlc7LytotGes4o53OA+7GMM2DCW8/fXeYyLsovGGeZ9nXaqvDc372P+HfPmuAh+khpaLOiIi9KL+h3HYl5cpT/D77lqw8OF7UXcrr7mK1nUu4bfX+V5jC3iHQt8ol91zfuoyUT9q2ychx1+4mAhSJW+xEUhMY6ynfviQv669bvYEaCob7GYoEqfvGd29YG+sWXey00RQAABBIYJFCYoedLCvG8+AZmleR+GW1kCGzGNFdFFic99D2zLwlhf6urios+OHourpMvuPx9X677+5g8L+zDazvDPZeZ9rtP88UsnXTixGT58vpjMxGrbMENjtfnwa3Wfx+evvf6p0t+zm27ePFH7ZeZ9ziIeTzjx/MHV7lX7HAZ9mX6OP+HcHVbQRyzjIomqbcfFHbGie7h/w8+LTNil2o6V4mVbncdV6pPcqiGuOj/t9CtL+1l3y/h5mfdxdXyMeZhp/jy2pYuJ9lI8R18P7V5+5f2FbUbbsU3h6GfG/Vxm3uf9jMfd91iZTWOMRzFo9AKWvP26BYaysTz17O8Mdiwoe93x6b43mffDKZrnCCCAAAIIIDBjAsx75n2tOU3dXD8F8z6M5XxONPoYc/SqY4qt2GOFcqyCrvqZsvfF/DVqLqP9iZ9j57xYbFH22bLj8zTvY34dF43XNYmL+h7GeOzcWMQi6i5vvPWXtVjEAo6itmIHyaLz1z0WOx4M31qxyudPO/2Kwj5FPy+86Lbso+2/2g2vSnvxnkce+7XSBShRd63aTryPeT/dHL4Oa+8dz7qCef//Liws/K/+J8vgmBnnrZpHAIGeEShNoCKJYt6P/1KdJOmYpXl/9rprSuMZW2OFgTpJn+t+JrYPL5ukxrZbddurYt4H1+Htr+ueY9bvv3L9Q4WxOejgo2vziL4uZd7HRHiSCVBMnIomeqPHYneGusxigl60PXq0HSvk67YXFxSM9it+DgN4kosB8vPHVnFlhYSjjz6jVj/nZd5fdPGdhSzi6vrYHj4f2ySP55xbfFuC2NKwjsaWMu9jF43YWnCSPuafiUJCkSYinvl7PDb/vdYkU+Z9z7JQw0UAAQQQQGB5CTDvmfcznSeU1UXmdc/7yNPL5kgxb5pnP0bnDLHwI+ohRfO3WAAx+v6lfp6neb9UX+q+Hlutx5bxRSyu3vhYZRZxW8eiNlavPmLqhSl1x5S//7EnfqOwT9HPqD/l75vk8cFNH5fe/rDsNplF52Hep10jKIpZV49VMO9Lf5+KfvcdW5g3r5OWN611dgQQ6BqBsX/EmPfNJzCzMu/vuPP1sbG88KLbp0qK6yZGkSgX3fM97htft62lzPswhKtujV733E29P0zO1asPL4zRlle/X5vJOPM+zOvXt/5F7TZjrHFVd1HchhO+SbaPzznGPcSH28qfx3jy91R5LJtcRXtx+4YqbYx7T2iu7DYCL2z5TuX252HeR1/LdiC49vonK/e1jEdsS3fAAYcVxu3e+7dVbn8p8z5ugVDWh6rH456NuaaGH2NbvapteF/z33t1mDLvu5ZmGg8CCCCAAAJJE2DeM+9nOk9IwbyPWsnw3Gj4+f77HzLT8S81D3j8yd8s7FvMCZb67OjrbTbvYyyxw1tRLSZM/dGxlv28+cU/KuQ5rUledr4qx8tqV7EwIhZOVGlj3HvOu+DmwjGfdfbGym2X1Zdsm7+8tYFxce/qa8z7uZvthX8/hr8naz5n3ied9uscAu0jMPaPFPO++USlafM+tomOK3FjhW3ZF0rch6uJLfHrJkdHHXXqoj5FX+IK6zptLWXe33r7K7Xaq3PuJt97403PL+IRMYvkrO55yiZA0d76DZtqtzd8/iOPPLmwn9F2bKu/9e0fTdx+3Fe+SKex6jq0PNyPcc9PP2N9YTvHHX9O5TbGtR+vXXLZ3YXniHsGLvXZ/PV5mPexm0UR0/0POLTWyvi8z0WPt9/5WuE5Tjr5ososxpn3cVFPnfgX9TGOxVaORSxuuW1L5X6Wte1489+HRUyZ9+1LJPUYAQQQQACBFhNg3jPvZzpPSMG8jx0Yy+pFe+y590zHX5Tvjx4rWnEe28XXvTd728374FJUi4nYxcr8UW5FPz/6+NcK58NXbXi40ueL2pzm2HMv/EFhf+IihXhtmrbzz0a9MPQyWgeI2y/EQoT8feMemffzmeuPi4HXfhED5j3zvsU5ta4jgMAMCCxKcIYTHuZ98wlMU+b9W+/+OIvV9mX3sxqO4wUXTrcV1aRJ1MZrvlKor9g2q06b48z7WB1ddSJT55yzeO/Lr36/kMf5F9xSi0f0rcy8j0nLm1OY69F2mWkdmgrTfBo2UTgIo35Yn/nzqjsQxD3hdtllRWEbMemapn/Dn417+eV9G36sc6uDeZj3++13cGE/b7vj1cZYfLT9p4PbEQxziOcRy3htmFvZ83Hm/U03b67URlnb+fG4r/1oH+PnOhdc5G15bP77rwpT5v0MMj1NIoAAAgjMg8DbCwsLf+9/qxgcurCwwLxn3jcyDynLc1Mw76NvZbvKxQ5uZX2f1/Gy3fmefvZ3a/WtC+Z9LMwpms8+/uTXK7F48ulvFH4+brE5r3gOn6fs9n5RJxl+37TPTzv9ysJxP/zo55XOw7xfnrn/tHHv4ueZ98z7eUxanAMBBNpDoDDByZNF5n3zCUyZeR+TqTBGx/0/9bTLs2OOPat0C+s8bsOPa9YclW2rudK9qQQottQe7kv+vO5W3uPM+7PXXVspGW9qTNO08/Gn/57FzgM5h/zxsMNPrD2GMvP+qLWn1W5rdEzjbsEwyb3uR9uP8eZjH36MLfNG31v0c5lBu8++q7O4OKDoM5MeK9ouPq4Sf++Df6h0nlmb9y9u+W4hy9gh4Z33/65SH6uyiVX2w/HKnz/z3DcrnafMvI+VBO+8/5NKbSzV1yee+q3CPu67ak1rLvJZaoxdf515354EUk8RQAABBHYg8KM8N/LYmsLr8cz75usdXc9V644vFfN+5cpVhfOk+HvV9By6LqM779pa2Lebbnmh1hyxC+b9Pfe9V8ji+hueqcSibMFI3FoxFgDVjc2074/bMhR9J8aOedO2Pfz5WLhQdJ64MGT4fWXPmfe+C8q0Me/jzPvW5JCFf3MWFhZsm7/D9MgPCCAwLYGyPzaD48z75hOYMvO+KNGc9tiq/Q7KYvXwvJON/HxxdXDRGGJld/6eKo/jzPvYzrtKG6m8Z6+9Fk+aV61aU3sMZeZ9EyuMy+4bHrF8beuf1+7rKPuTT7m0UBcPbvqkUtsXXHhr4efPPf+mSp8f7c+4n+MK9SINh0k87nP5a7M272PFelH/1h59eqX+5f2s8njNdU8WnqvqbSvKzPs6Oxks1c/YhSMuXChicsKJ51e+6GKp83i9+e/GnCnzftq0zucRQAABBJaJAPN+oXUFV+a9VfeNz5nynDZ/TMW8L6pDxJwpLkxfbvM+VkcXzd/WX/VQrfh0wbyPGmwRi8suv7cSi7gV3Z577lPYRmzJ/9Y7f1OpnVy/0zxG7ahoLHHs3W3NXLyf9++ll79XeK4vnXRhpfEy72c3v89j5LEaY+Z963LJ0b89zPtlmog5LQJdJTD6R2aHn5n31b5c6yQh8zLv437zYXrX6VvRe2MC9MJL385ixXXcJyu2eD/r7GuyU069LDv+hHPH/j/iiJN20FOeuK875/pa/Rpn3sc91Iv6Pctjr2/9i2zTI59lsYNAbAN2zrnXD3ZMWIpHvF603XvsulC3v2Xm/Zfvnv4K5rKt1las2L12P4vGddbZGwt1UbXvRfeBC201tfX6cJ8vv/L+wr7ee//7lVjM2rxfd851hf2LCxyGx9HE84hP/js8/HjV1Y9UOleZed/07hnHHb+usJ/R5yhabdj4aPbGmz+s1OcmuGmj3vco876r6aZxIYAAAp0nwLxn3ssvXQywSAOpmPe77VZ867q4L3gT85WPtv8s2/zit7O77nk7W79h0y/rRhuzU069dGzNKGokhx1+QuH87cKL6t3+MRXzvryGtjSLw4/4UiGLc8+7sXKcyuotMR/effe9ssuvuC97+bU/qdzepPp44KGPCscSOyZO2mbZ597d9veF5zr00OMrnYt5X2/OXhYHx6fnyLxn3nd+xmSACCBQi0BhgpMbM8z76b94R5OXWZv3sR1amJiffPYflZLU0f7Fz7HNfky64irVuId6roemHmMCV3TesmPjzPt53O8+rkSPldZhiI7bbm4aPrGlftn4i46XmfcPPfxprXaK2o4LIorG0tQk67zzbyps/9bbX67U95hwFvUvdB99b/L/xZd+ufBct9y2pVJfZ23el10gc8mldzXKIZiWrfK/6OI7KrEoM+83Xvt4pc8XabXo2GNP/EZhzEY1c9BBa7PgFDtNfPDxvzbah6J+OVbt+5R5XyuH82YEEEAAgXQIMO+Z9/JJ5v0iDaRg3kdtqOj2fTE/ilXak85T3v/gH7MwmmKHs7Ldz0bnYHV+PvOsq2v1bTnN+1nX0GIuXTVOcWu92FFhKdarVx+exQUSDz28PXt/Brfa3HjNVwr7EBdrNFmzibaee+EPCjW+334HV+LGvK82V6+qQe+bnCfznnmfztRGTxBAIAUChclUnuQx7yf/wi1LVmZh3sfV0scety6Le5V/+MnkJlQYWLE1WawEzzUwi8dYGVvGp+h4mXkfE9Dtn/9nrbaK2h937P4HP8wOPPCImfIIxnGl8Lh+jL5WZt5X3c59tL3hnze/+EeF4w2zc/h9kz6P3RuKdFXFEI8CQdFn530sJqJVxj9r836ffQ5Ydh5Viypl5v3Nt75UiWUV3vl7zjn3hlpcotgUF0LExRpx+4amt/HL++Vx6e9U5n0KqaE+IIAAAghMQIB5z7xvPKeVOy6dO6bOKAXz/s13/rp0brR69RG1dRsLKOJ2fStWFK/mb2puftLJF9fq23KY91FDi/uql+1s0BSLE044rxaL6FOdc++8887ZIYcel11w4W3Zvfdvy95+78e1zlf0e1hW96nTr2nfW3WXS+Z9+//WFmmwjceY98z7CeYgPoIAAh0mMDahY943n8CUmfcx8Vlq2/W4ovmMM68abEEW23mHWR9bk9VdtV2UwMTVuTFxmzY5rvL5Y487u9ZEoMy832PPvWu1UzTusmMxCQvWVcbTxHveeb/ePb/KzPunnv2dqZmUmfeHHHLs1G0H77JJXBXz/pXX/nRuMRkX10svv6cSi1mb92UrKMb1venXqt5Hrsy8v+ue6W/1MPp7HNs2xq09Jh1rcA0TOXYbeOvd6QsXo/3zc/l3K/O+wxmnoSGAAALdJsC8Z95Xmh/IA8vzwC6yScG8f27zt0rnRXVrM7HKef/9Dyltb9L5V9HnYi5dRxPzNu9f2PKdLFavF/W96WPHHX9OLRaxe+S5JTseVu1bmPlxq8g33vrLWufOYzbNfLxqH6u8r0q9lHnfr7/LuUZTfGTeM++7PV0yOgQQqEtgbKLJvG8+gSkz72N193IlDrHF1O57rByrhXFJcVxlHNutjf4v29687gSxzLyPVcezYBbG/VFHnToxj7iv/SiL/Ocys5V5v5BVMe/jIpNxWpzXa5dcdncl7c3SvA+Del7jHXeeE088vxKLMvP+vge2Vfp83d/12JXjxpuez/aY4m9bjDtW5a875/rsta1/PpN+1h1X19/PvK+bxnk/AggggEAiBJj3zHu5om3zF2kgBfN+nBl09rprF/W5bL7x9LO/O9UK81gFnddFhh/Ldn5M2bx/dvO3BvePHzdPHvda3RpaXfM+j2HEfu+995+qbhA1rNNOvyJ76eXvVdZKnD/qBOMYzOu1qJvkPMoemffN177LWDs+nvW4v9fz+p1xnqkuIDgpkXmJbiCAQEcIjE2mmPfjv1QnSTpSM+/DNI57mY/7co4rq9edc1122x2vZnE/6S2vfj9774N/WHLL+qee+UZhu6mb92eetaGw3zmjuCjhpJMvyq657sls0yOfDnY/iG3FqkwKyrY5Z95XM+/jQpM8Dsv5mIJ5/8HH/5IEi1TN+/zvc2yBf9XVj2Sr9jtoKl677LJrdv0Nzyz5dy8/r8fJvj+Z9x3JLg0DAQQQ6B8B5j3zfkmDSH44WX7YZm4pmPdxa7CyuXPMb6rwfeudv8lWrlxV2k60f8ABh2XnnHt9dvudr2VRS3z5l3Wjpdp/9PGvFbabqnn/9nt/m+299/jb101aQyszkSc174N91A2uvf6pqW8HGSb+FVc+kH3y2X9U0kzcLrNMd/M8XqVOV8Z97drTKo11KY17vX9/+yeNeQXz/v9ZWFj4mf9zYfC/TfC3innfvzmgESMwUwJjkynmffMJRmrm/bnn3ViqgbgH9KNf+fWJzaonnvp6Ydspm/ePPPZrhX2OL+y4Ojzuzx33eJs0Edtrr+IJL/O+mnkfV3uXJU8bNj6aXXfD03P5//iTv1lJA7NceR+T5jIWF15021w4BO/7HvigEot5r7wf/R2NrQOfeOq3svVXPTTYWSNW1JfxG3f8zLOuzqKt0fb93Mz3JfN+pjmfxhFAAAEEZkeAec+8lx9aeb9IAymY9wcfcmzpvKdqzS/mQGVzpDA4o53Y+WySOdHDj361sO1UzfvYla2MxZFHnpzFxQiTsog6Q1Hb05j3wzF55vnfy67e+Fh2zLFnZrvuuqLwXEXnHz52wgnnZXGLguF2i56X1UKiHjivuk2c55PPlp67M++bmcsX6cCxemwrmPf/PbtUVssjBJ4d/ttX8TnzfgSiHxFAYDoCY5O1qom8L+PqX8Ypmfdvvv2jwZbQRV9AMTmrcoXquNg/uOnjQn2lbN4fffQZhX3e/4BDs1ff+LMlJyjjeMQEbueddy5sn3lfzbx//c0fFvILDW955Y+nis+42E36WtmE9aGHP22kr7vuulshjwce+riR9icdd9Hnltu8H+1T3B4jiiMbrn4kO2rtaaV/C4v+Pl65/qHk+I6Or60/M++nS+p8GgEEEEBg2Qgw75n38kPm/SINLLd5H/cr32mnnQrnjHEx8/sf/OOiPo/OI6IOUnb7v3POvSGrck/x0TaHf7773ncL+5eieR88y2o6Z519TSVTe3jso88feOijQhZNmffD5/to+0+z2C1z47WPZ1Gji9s/Fs19i45F3IfbKnp++hnrC9uLXT2L3r+cx5j31WvayxmnPpybeb9seXzRiZn3RVQcQwCBuRIoTKby5Ix533wCk5J5f8ONzxbG/8ADj8g+/ORfp06o7/jyG4Xtp2rev771LwontjFRjfu7TZsoxsQ4/90afWTeVzPv4wrvssJBXLE/bYya/vyszfvVq48o1FT8bjc9lmnbS828Hx3P+x/+UxYXVVx40e1LbgkZW+hPezHP6Pn9/IvvW+b9XHNAJ0MAAQQQaI4A8555n1z+Lb9svp5Tl+lym/cbr/lK4Xwx6hGx+rrKeK4uaePgg4+ZesFHnP/mW14s7GOKAC8LQwAAIABJREFU5n2s5B6t5cTPa9YclX34yb9V4jmOedxyoKj9WZj3o/2Ii9tjJ8pLL7s723fVmsJ+5H2LC0Keff73x4432snfP/xYVXej/Zvlz8z75f9bOcv4tqlt5n1ziXkDLTHvG4CoCQQQmI5AYTKVJ1bM++YTmJTM+zJjMSZPTSQ3cT+sXEvDj6ma92VJ0tHHVJvULsXs+Rf+sJBHsGHeVzPvg3HZvctTNKzLfseaWnkfW9YN/27lz889/6ZGfoeX0nSd11M374fHEruOROGk7DYXwfmii+9MjvHwGNr6nHk/XVLn0wgggAACy0agrnn//y0sLPzv/i8rg7ULCwu/l+fPRY+nnX6FfM9q+qk0sJzmfaysHmfCVp0/R/2m6Pfjjjtfn4pNPl+5+NIvF7afonl/4onnF/b1ltteaoTF5VfcV9j+PMz7PB7xGLfou/f+98fqJ1bWD39m9HkwKdLN3nvvP/FtBUbP0dTPzPvma99NxaZv7ZTVpYd+l2ybP79Un3k/P9bOhAACJQQKk6n8S4F533wCk5J5v+agtYXxf23rn49NwqsmT3FFba6l4cdUzfv1GzYV9jeurq465nHvu/nW4slLsGHeVzfvTzn10sI4HX/CuY3EaVwM6742a/O+7AKZVavWJDchbpN5n8f5ldd/kK1cuV+h3mKHkvx9Hpv7rmTel2RrDiOAAAIIpE6grnn/f6Q+oJ70j3nPnJ9pTr+c5v2NNz1fOI+J+kNskf7Wuz+uNPb99ju4sJ233vmbSp9faq50xBEnFbafonkfc8Dh2lb+PG7vt9Q4q7y+9ujTC9uft3mf9/XNd/46i1tI5uMcftxjz70HJn/+3tHHp5/7ZuHnoo1Y2DL6/uX8mXnf3Hx+OePYhXMz75PKPpn3SYVDZxDoJ4HSZCoSKuZ98wlMSub9Xnvtuyj+sf3VtPe6j4Qptt1fsWKPRe2HrlI178+74ObC/t5z37uNTCxi5cbwZGf4OfO+unl//Q3PFHKMAsS2j/65kVg1lfSffMolhX19cNMnjfRz0yOfFrYf2nrhpW83co6mWLTRvI+x33r7y6WMt334T0kxbipWy9kO876fyahRI4AAAh0gwLxvZxCZ98z7mebzy2Xev/rGD7Lddiuux8RcMe7PXjXn33XX3RbNh+I2Yts//3nlNsrOFbcui7aGayP587rm/Sef/bzwNojRXhN9jTHsscfKRX2NGtrHn/771Cxi2/oi1tH/5TLvY8wPbvp40ZjzGL3y2p+WjjtuI7Bixe6Fn41bMZRpYjmOx20y8zENPx551ClJ9XM52Dhn877AOKbM+6SSSeZ9UuHQGQT6SaAwQcmTFeZ981/SqZj3ZRObMEDHJRJVXyu7V1doK1XzvsxcbGKL87iqPdjmv1ujj8z76ub9y6/9SSnHG2/e3Ih+q+p8qfeddfbGwr7ec997jfRz3EUycTHKUv2b5+tlv1/3PbAtqX6OMonf3dHf1/xn971v/juSed/PZNSoEUAAgQ4QYN63M4jMe+b9TOciy2HexwXGhx56fOkcZuedd8m2vPLHlcb9wcf/UtjO7nusrPT50bnV6M833by5sP2Yb9U176PtsgUkMW8ePXfdn8Ogz+eBw4+77tpMDe3W218pbD/OtZzmfSzuiTEOjzl/Hqb3OI5luybGbg5NLBoad+46r7245buF44vfozrteG/z9YG+MWXeJ5VMMu+TCofOINBPAoUJSp6IMe+bTzxSMe8jASq7Cvb9D/5xqgQ1rmo++JBjS7WVqnl/9rprC/vcxL3crrr6kcK289815n118z60e/gRXyrkuffeB2RxxXoqCf4FF95a2M+YmDfVx1g1keto+DFWMDR1C4wm+tpW8z7GXrYaZMur328sjk0w7kIbzPt+JqNGjQACCHSAAPO+nUFk3jPvZ5rPz9u8f3fb32dHH1N8+8J8rnjRxXdUHvMvFn3svGi++YsdG39auZ2ieUrcV3316uJt6KOvk5j3++xzwKK+RltNbfFfZmJPuyNb1NAOKrmtZfR/Oc37iN2q/Q4q5PrEU18fq4H7Hvig8HMxpltu2zL2s0WamdWx17f+RWE/99v/kGT6OKuxa7f5uv80TJn3SSWTzPukwqEzCPSTQGGCEolU/GfeN/8lnpJ5X3bvskce++pUCeq4e6uFrlI17y+9/J7C34dzzr1hKh6xUrzsQon8d415X2/ydtc9bxXGKnheeNFtU8VrmkR79LPrN2wq7Odll9/bWB+fef73Cs8RLI4/4dyx96Eb7e8sf26reR+FmPz3dPQxtnmcJbM+ts2872cyatQIIIBABwgw79sZROY9836m+fw8zftnN38rO3DNkaVzl5jLrF59eFZ3DrNy5arCNp946remYrfx2scL283nXJOY9wcffExhm089+ztT9TWfl+27ak1h+49+5denav/6G58tbDdnsZzm/fbP/7P0FgxR68rZFD3GbgWrSpjttdeqLEzzos/N+1jM+eOClJx3/rjzzjtnH3/6syT6OG8mzte8H1CFKfM+qWSSeZ9UOHQGgX4SWJSc5ElKPDLvm/+yTsm8D2NvON758zDZqiQVRe95Yct3Su/Tlbefqnl/2x2vFvLYY8+9s7iCvWi8Sx37aPtPsyOPPLmw3ZxHPDLv65n3cZX+uCvTI5ZLxWaS1yOece+2qp+NLeGH45w/b/reaSefcknheeJ8l1x6V+X+Vh1XvC9WB9RZYTAP8/69D/6h8bE+8NBHhWz33HOfxs9Vh39X38u872cyatQIIIBABwgw79sZROY9836mOf08zPutb/3V4AL2MBrz+WbRYywo2PziH9Ue79q1pxW2e/a6a2q3lc9hntv8rbG3FYz+T2Len3nWhsK+Xnv9UxP3Ne9zPB573LrC9k8/Y/3E7b/w0rdLt6XP41jVvJ/FfDgufMj7MfwYt1+osuvhHV9+o/Dz0dYhhx6Xxa0Zhhk39bwui1hlPzy+/PmTT//2TPrX1Di103zdfjmZMu+TSiaZ90mFQ2cQ6CeBwuQkT1KY980nASmZ92VX98YVp5se+bR2gvr8C3+YlV2VnWsqHlM178u2yoo+rzvnuiyuOK6TxMUkpGxyN8wjnjPv65n3EYf4+1R0dXTw3GmnnbMNGx8dGMx1Ylb23pj4XXPtE1lsy1/1/oDR1suvfr/wb2z079nnf7+Wnsr6Fsdfef0HpVfDB49zzr2+1kUH484VFy/cdscrgy0OH9z0ceUxzMO8j2LNl066MFtq+75x4xt+LX6H15RsX7junOsrj324Tc/Hf68y7/uZjBo1Aggg0AECzPt2BpF5z7yfaU4/K/M+LqKOudipp12+pAke88HY7n3S1eFXb3ysdE772BO/UZvfM899M4sLoaNf4/5PYt5fd8PThW3G9vxxIf60c7G4CKCoz1GXePjRz2u3HxcxxAr0ojaHj1U17y+57O7smGPPHPSlbv2qiE2snD+q5OKNE088v9J449YL4xa0hIEf9Yyi809yLOpEJ5x4flZ3vn7KqZcWxuG00ydf3DRJ/31mfL2g63yY90klk8z7pMKhMwj0k0BhcpInicz75pOGlMz7V17704HJmcd7+HHFij2yWHFaJTGKZPzW21/Odt9j5Q562mWXFTv8nLefqnkfYy2bXEffL7zo9soTvqef++bgKuJ8zPEYE7qye2cz7+ub9xGvslsd5NyPOfasbNIrpWN7tIcf/WoWt03Ybbc9v9ByHfM+JswHHHDYF5/N+xWPcaFLXIVed9vCst/JcVe0x/kOPuTYQYEnVsyXtVF2PD7z1DPfyC6+9MvZ3nvv/8V4UjPvzzjzVystDjv8xOzGmzdnW9/+Ue3xBofX3/xhNs5I9v3Y/PdjcB/H/Je/P2/1M10zagQQQACBxAkw7xMPUEn3mPfM+4nmCmXzptHjZfWFyHlf2/rnY/+/+sYPslggEfOwuHd4zG3OO/+m7PAjvpQttcp+eN4ZRvmjj39t4nHG7oplF81HDWjTI59VajuM4Jtu3rzoovOyutEk5n30dXjsw8/DhH78yd+c6rZyW179fmkNbbfd9hjMt0c1UPRz7CR4y20vZbvvvtcO/S1jUce8z8ccOxVee/2T2atv/Fml+Iz28+33fjy4DV/e3ujj3fe+W7nd0PLoWIfbCx1df8MztXb2G+5vtB8XbsSFAHm7dc37WKCQf3b08ayzN2axQ8LwOT2fTT2g71yZ9yXZ2vIcZt4vD3dnRQCBIQKlyUkkK8yJ5pORlMz7SIrOOvuasRqIVaxhYG776J8XJaoxCYj72x9yyLGL2ojJZEwURpPe+Dll8/6hh7cX9jkfx4EHHpHdctuWLLamG00q4+r32CY9VhgXTW4vvezubJ99Dihsn3k/mXkfk95xW8bncYsrveOeek8+/Y1CLUc7b77z14OiRlxNf9rpV2Zxu4T888OPdcz70Eis2B/+/Ojz0EoY/FHAiVtZFP2vus3fUhczxLnXrDkqW3/VQ4OVF0W3g4gLDt5+728HFz1Egejsdddm++y7unAMKZv3Oefgu/bo07P1GzYNiilvvPnDRb+7+e9y/B7GriMx0S8rnES7p5x6WWkbeVseJ/v+ZN4PZWieIoAAAgi0iQDzvk3R+lVfmffM+5nm9WXmfT5XmfXjEUecNLhAYNq5SazwH9fXk0+5dDC/LNpGPVZVx66PRbe9i8UNZfPlScz7GOdS84nYhSAubI/bARTNveNYXDBRxqxsa/6cz0knX5Q98thXC7eDz2tocf78/fnjuBraJOZ93m48xgUfl195/6BeFUZ32ar82HEwLvS44MLbsljQM9zG8POYX8cinjJGRcdjl4Zxc+xof489VmbnX3BLds9972Yxby9aeBC1yc0vfju7866t2UUX31m6AKeueR+LKuICjOFxjj6Piwzidzp2NyjTTt3aXhErxyarJXSBG/P+VwlaAs+Y9wkEQRcQ6DuBsYkJ8775hCE18z62iq+yZVlMJPbb7+Ds0EOPHxiN4xL5SHDjqteY8Iwmu/FzyuZ9TGJOOOG8wn6PjiVWIB908NGD/3vtte/Yz8Tk5qPtP2Pef/W/BpOxUZbxc1wUMUmyHdu4R7GgqM2yYzFh33ffA7NV+x009grwos/XNe/DII/zFLVV9VjV+wmGfmN1fNV2431xr7q4HUDc4y0my3U+2wbzvmg8USSK39+4kCEKGatXH54t9Tuct7PvqjVZrEKYRKs+s/R36lLFtoWFBSvv+565Gj8CCCCQJgHmfZpxWapXzHvm/Uzz+uUy72NuE7sj1jVYy+YrL7/2J5XmzTG3HNSNDjsh2/+AQ7MVK3YfO7+8+daXBmZxPtcafpzUvH/s8a+Vro4fbn/c83G7UMaOCWUX+g+3ubiGNp5FXOAQuwYOt5E/n9a8z9vJH6NvsRNgLE75xXz4iGzlyv0Kz51/Jn+M+uGk29zHhfJL1RLz8+SPcb5Y7BD9DX3lx5d6rGveh/avvuYrldsvO/+bE+78V/a75/jSNYQuMWLeL5W2zfV15v1ccTsZAggUERibmDDvm08SUjPvI8mJrcPKtnMvS0jHHb/s8nsHE+A2mvfBI642DlNv3BjrvBYT9nyFs5X3zZv3EbMoSlx+xX1TT9KrxLWueR/9e+rZ31nyKu5x565q3se54n9swbfrrrs1puGyvrXVvC8bz1LHo7gR2+nnnD02/x3JvC9K1RxDAAEEEGgBAeZ9C4JU0EXmPfN+prn9vM37MM5vuOm5ibcfHze/iR0Zw/Rdas5U9fWrrn5kwD5Wehd9ZlLzPsawYeOjhW0Wnafo2DjzPtqPWmkdI7noHMPHopYR7c7LvB8+d53nsQDixZe/N9XvzLObvzW4sKPOeSd57yTmfdSVymq2VfvAvG++RjDu71LXXmPeF2Rqy3eIeb987J0ZAQR+SWBsQsu8bz7pKEsEwxRazqTj0a/8+uBK1qoJadH7YjXz8Orptpr3EYdIuI8++oyxvx9FDEaPnXb6FTtsl8a8n415n//uPPHUbxXexmE0LpP8HGb4GWdelb277ScT/a6+8tqfZnGfvUnOXde8Dx4vvfy9wS4Xk5xvqc9E0Sb+lsUKjJz9Uo9xO4miduNWE0t9turrS21hWHT+Ksdi6/24xcg77/9dY32tOqa+vY95Lz9FAAEEEGgpAeZ9OwPHvGfezzS/n4d5v2rVmsE250889fXGVtqXzUFi5XSVnRvHzbFiNf6dX37jC+6zMO+j/2HAxw5z4/pS9tpS5n20/8hjv5bttdeqidrPzxs1hltvf+ULFtOa93Grxrztph/jdoVvvPWXX/S1TCNVjsctJ2PL+yYXEQ2PN3bIvO+BDybqaxj4ccvFpXaNGD7f8HPmffN19Cqa6sp7mPdJJZPM+6TCoTMI9JPA2MSOed980pGqeR+Jzta3f5SdceaG2ldT77TTzlnc12vUyGuzeR884j7okbRX3U57OGE/cM2Rg8niaALJvJ+tef+LuP08u+e+97Kjjjp17N+34XiVPY8t3U448fzs9jtfy97/4B8nmvyNamDzi3+UXXLZ3dmRR52S7bbbnpX6OIl5n583iiHxdyd+T8vGWeV4rCyI2z/ccOOz2Zvv/HVtFvMw7+PWFHEvveDbxO4ZMeZY8fH0s79be7w5f4/1vkeZ9/1MRo0aAQQQ6AAB5n07g8i8Z97PNM9v0rwPM3H16iOyk06+OLvsivuyu+99Z1l2BYtbL8bcru78Mt5/yqmXZXHf9+E50qzM+zjHx5/+e/bQw9uzs87emB100NrKq+WrmPfRfpjZp5+xfiIWEce4wH+YxbTmfdSwoo0rrnwgO+TQ47K4CL3KXL/sPRGzY49bl8Vin+F+NvU8tt+/8KLbK92SoayP+fGY/19y6V3Zcy/8QSN9ff/DfxrUgeKihf33P6QyR+Z9vfl/U1rqSjvM+6SSSeZ9UuHQGQT6SWBsAsK8bz7p2PzitwdbbAXb4f9PP/fNRhLMJhKWV9/4QbZ+w6bByvNddllRqJG42jruXX/NdU9m8f6i84bZOTzG/PnzL/xh4fuL2ohjH23/aWE7Tz79jVrtlLW/1PEPPv6XwdXQMdEsu7I6dh2IyVHcbzxuQxD3Hi9qN/qccxh+jEll0fvLjj23+VuF7TRhMG/76J8L237m+d+r1ceyvm959fuF7cfFI2WfmfR43I/uplteyE47/crB1mzjtvmLq94jhjH5jm32YhV/mMGTnrvK57Z//vPBhP+lV/5n9uzzv1/IJXQSq+irtDfuPW+9++Psjjtfz85ed2225qC1Wdnvdkx+4wr4mPyG5q9c/2AWWyTG78G49pd67cUt3y0c3yzvH//2e3+bbXrks8Hfs5h0R8EmflfzCf7oY5j1Bx9ybBYXS8QuIvH5pcbl9Wa/J5n3/UxGjRoBBBDoAAHmfTuDyLxn3s803x83xxuuBxQ9j7pJ/I96y7RzsVnMWV5+9fuDueLataeVrqCOhRBxv/brbng6C9O/qB+xs13R+F946duF7y9qo+qxqC1FjSDajlvbFZ03jtWdB4YJvf6qhwYXu5fNs4NFGOHXXl9eQ4tbOBb1KRYAVB3j8PuivTDer974WHbqaZcPdikct6I8zPqoA8Sinhtv3pxtfeuvJjrvcB+qPP/wk38bzNsvvuTOwSKM3Xffq3TOHnP42G0idBXG/133vDVYiFTlPNO8J2ptceFJ/E5GragoTnFs1jWkacbgs83WLmbBk3mfVDLJvE8qHDqDQD8JjE2IIvGYxZeRNtNPGPIYxZW7Ya7Ffb4jSY1JSd+3jv5F0v6DAY8wXWNCE1tr5cw8pqvv0HPEK+IWeg5TPO5hHpPaMNL7FLsYb6ygj4spgsULW74zKGQEi7oXk7SJW4z7nfd/8kXRJsYeBZw49vGns71Yo02clquvzPt+JqNGjQACCHSAAPO+nUFk3jPvezUHnFWOH/PHuFg8n2fHBQeT3m5uVn2cV7vFNbTJbr03qz5HbAYXMWz5zqAWEKZ01PlSMp7f3fb3gwtXYr4e/6OPcVHFh5/8q99Zf7c7qwHmfVLJJPM+qXDoDAL9JMC8l/R0NumZ1URHu+ma82IjNjTQbg0w7/uZjBo1Aggg0AECzPt2BpF5rx6iHkIDNEADNEADSWiAeZ9UMsm8TyocOoNAPwkw7yUoSSQoDLd2G27iJ340QANNaIB5389k1KgRQACBDhBg3rcziMx79RD1EBqgARqgARpIQgPM+6SSSeZ9UuHQGQT6SYB5L0FJIkFpwvTRBvOQBmiABtqtAeZ9P5NRo0YAAQQ6QIB5384gMu/VQ9RDaIAGaIAGaCAJDTDvk0ommfdJhUNnEOgnAea9BCWJBIXh1m7DTfzEjwZooAkNMO/7mYwaNQIIINABAsz7dgaRea8eoh5CAzRAAzRAA0logHmfVDLJvE8qHDqDQD8JMO8lKEkkKE2YPtpgHtIADdBAuzXAvO9nMmrUCCCAQAcIMO/bGUTmvXqIeggN0AAN0AANJKEB5n1SySTzPqlw6AwC/STAvJegJJGgMNzabbiJn/jRAA00oQHmfT+TUaNGAAEEOkCAed/OIDLv1UPUQ2iABmiABmggCQ0w75NKJpn3SYVDZxDoJwHmvQQliQSlCdNHG8xDGqABGmi3Bpj3/UxGjRoBBBDoAAHmfTuDyLxXD1EPoQEaoAEaoIEkNMC8TyqZZN4nFQ6dQaCfBJj3EpQkEhSGW7sNN/ETPxqggSY0wLzvZzJq1AgggEAHCDDv2xlE5r16iHoIDdAADdAADSShAeZ9Uskk8z6pcOgMAv0kwLyXoCSRoDRh+miDeUgDNEAD7dYA876fyahRI4AAAh0gwLxvZxCZ9+oh6iE0QAM0QAM0kIQGmPdJJZPM+6TCoTMI9JMA816CkkSCwnBrt+EmfuJHAzTQhAaY9/1MRo0aAQQQ6AAB5n07g8i8Vw9RD6EBGqABGqCBJDTAvE8qmWTeJxUOnUGgnwSY9xKUJBKUJkwfbTAPaYAGaKDdGmDe9zMZNWoEEECgAwSY9+0MIvNePUQ9hAZogAZogAaS0ADzPqlkknmfVDh0BoF+EmDeS1CSSFAYbu023MRP/GiABprQAPO+n8moUSOAAAIdIMC8b2cQmffqIeohNEADNEADNJCEBpj3SSWTzPukwqEzCPSTAPNegpJEgtKE6aMN5iEN0AANtFsDzPt+JqNGjQACCHSAAPO+nUFk3quHqIfQAA3QAA3QQBIaYN4nlUwy75MKh84g0E8CzHsJShIJCsOt3Yab+IkfDdBAExpg3vczGTVqBBBAoAMEmPftDCLzXj1EPYQGaIAGaIAGktAA8z6pZJJ5n1Q4dAaBfhJg3ktQkkhQmjB9tME8pAEaoIF2a4B5389k1KgRQACBDhBg3rcziMx79RD1EBqgARqgARpIQgPM+6SSSeZ9UuHQGQT6SYB5L0FJIkFhuLXbcBM/8aMBGmhCA8z7fiajRo0AAgh0gADzvp1BZN6rh6iH0AAN0AAN0EASGmDeJ5VMMu+TCofOINBPAsx7CUoSCUoTpo82mIc0QAM00G4NMO/7mYwaNQIIINABAsz7dgaRea8eoh5CAzRAAzRAA0logHmfVDLJvE8qHDqDQD8JMO8lKEkkKAy3dhtu4id+NEADTWiAed/PZNSoEUAAgQ4QYN63M4jMe/UQ9RAaoAEaoAEaSEIDzPukkknmfVLh0BkE+kmAeS9BSSJBacL00QbzkAZogAbarQHmfT+TUaNGAAEEOkCAed/OIDLv1UPUQ2iABmiABmggCQ0w75NKJpn3SYVDZxDoJwHmvQQliQSF4dZuw038xI8GaKAJDTDv+5mMGjUCCCDQAQLM+3YGkXmvHqIeQgM0QAM0QANJaIB5n1QyybxPKhw6g0A/CTDvJShJJChNmD7aYB7SAA3QQLs1wLzvZzJq1AgggEAHCDDv2xlE5r16iHoIDdAADdAADSShAeZ9Uskk8z6pcOgMAv0kwLyXoCSRoDDc2m24iZ/40QANNKEB5n0/k1GjRgABBDpAgHnfziAy79VD1ENogAZogAZoIAkNMO+TSiaZ90mFQ2cQ6CcB5r0EJYkEpQnTRxvMQxqgARpotwaY9/1MRo0aAQQQ6AAB5n07g8i8Vw9RD6EBGqABGqCBJDTAvE8qmWTeJxUOnUGgnwSY9xKUJBIUhlu7DTfxEz8aoIEmNMC872cyatQIIIBABwgw79sZROa9eoh6CA3QAA3QAA0koQHmfVLJJPM+qXDoDAL9JMC8l6AkkaA0Yfpog3lIAzRAA+3WAPO+n8moUSOAAAIdIMC8b2cQmffqIeohNEADNEADNJCEBpj3SSWTzPukwqEzCPSTAPNegpJEgsJwa7fhJn7iRwM00IQGmPf9TEaNGgEEEOgAAeZ9O4PIvFcPUQ+hARqgARqggSQ0wLxPKplk3icVDp1BoJ8EmPcSlCQSlCZMH20wD2mABmig3Rpg3vczGTVqBBBAoAMEmPftDCLzXj1EPYQGaIAGaIAGktAA8z6pZJJ5n1Q4dAaBfhJg3ktQkkhQGG7tNtzET/xogAaa0ADzvp/JqFEjgAACHSDAvG9nEJn36iHqITRAAzRAAzSQhAaY90klk8z7pMKhMwj0kwDzXoKSRILShOmjDeYhDdAADbRbA8z7fiajRo0AAgh0gADzvp1BZN6rh6iH0AAN0AAN0EASGmDeJ5VMMu+TCofOINBPAsx7CUoSCQrDrd2Gm/iJHw3QQBMaYN73Mxk1agQQQKADBJj37Qwi8149RD2EBmiABmiABpLQAPM+qWSSeZ9UOHQGgX4SYN5LUJJIUJowfbTBPKQBGqCBdmuAed/PZNSoEUAAgQ4QYN63M4jMe/UQ9RAaoAEaoAEaSEIDzPukkknmfVLh0BkE+kmAeS9BSSJBYbi123ATP/GjARpoQgPM+34mo0aNAAIIdIAA876dQWTeq4eoh9AADdDqjWKFAAAgAElEQVQADdBAEhpg3ieVTDLvkwqHziDQTwLMewlKEglKE6aPNpiHNEADNNBuDTDv+5mMGjUCCCDQAQLM+3YGkXmvHqIeQgM0QAM0QANJaIB5n1QyybxPKhw6g0A/CTDvJShJJCgMt3YbbuInfjRAA01ogHnfz2TUqBFAAIEOEGDetzOIzHv1EPUQGqABGqABGkhCA8z7pJJJ5n1S4dAZBPpJgHkvQUkiQWnC9NEG85AGaIAG2q0B5n0/k1GjRgABBDpAgHnfziAy79VD1ENogAZogAZoIAkNMO+TSiaZ90mFQ2cQ6CcB5r0EJYkEheHWbsNN/MSPBmigCQ0w7/uZjBo1Aggg0AECzPt2BpF5rx6iHkIDNEADNEADSWiAeZ9UMsm8TyocOoNAPwkw7yUoSSQoTZg+2mAe0gAN0EC7NcC872cyatQIIIBABwgw79sZROa9eoh6CA3QAA3QAA0koQHmfVLJJPM+qXDoDAL9JMC8l6AkkaAw3NptuImf+NEADTShAeZ9P5NRo0YAAQQ6QIB5384gMu/VQ9RDaIAGaIAGaCAJDTDvk0ommfdJhUNnEOgnAea9BCWJBKUJ00cbzEMaoAEaaLcGmPf9TEaNGgEEEOgAAeZ9O4PIvFcPUQ+hARqgARqggSQ0wLxPKplk3icVDp1BoJ8EmPcSlCQSFIZbuw038RM/GqCBJjTAvO9nMmrUCCCAQAcIMO/bGUTmvXqIeggN0AAN0AANJKEB5n1SySTzPqlw6AwC/STAvJegJJGgNGH6aIN5SAM0QAPt1gDzvp/JqFEjgAACHSDAvG9nEJn36iHqITRAAzRAAzSQhAaY90klk8z7pMKhMwj0kwDzXoKSRILCcGu34SZ+4kcDNNCEBpj3/UxGjRoBBBDoAAHmfTuDyLxXD1EPoQEaoAEaoIEkNMC8TyqZZN4nFQ6dQaCfBJj3EpQkEpQmTB9tMA9pgAZooN0aYN73Mxk1agQQQKADBJj37Qwi8149RD2EBmiABmiABpLQAPM+qWSSeZ9UOHQGgX4SYN5LUJJIUBhu7TbcxE/8aIAGmtAA876fyahRI4AAAh0gwLxvZxCZ9+oh6iE0QAM0QAM0kIQGmPdJJZPM+6TCoTMI9JMA816CkkSC0oTpow3mIQ3QAA20WwPM+34mo0aNAAIIdIAA876dQWTeq4eoh9AADdAADdBAEhpg3ieVTDLvkwqHziDQTwLMewlKEgkKw63dhpv4iR8N0EATGmDe9zMZNWoEEECgAwSY9+0MIvNePUQ9hAZogAZogAaS0ADzPqlkknmfVDh0BoF+EmDeS1CSSFCaMH20wTykARqggXZrgHnfz2TUqBFAAIEOEGDetzOIzHv1EPUQGqABGqABGkhCA8z7pJJJ5n1S4dAZBPpJgHkvQUkiQWG4tdtwEz/xowEaaEIDzPt+JqNGjQACCHSAAPO+nUFk3quHqIfQAA3QAA3QQBIaYN4nlUwy75MKh84g0E8CzHsJShIJShOmjzaYhzRAAzTQbg0w7/uZjBo1Aggg0AECzPt2BpF5rx6iHkIDNEADNEADSWiAeZ9UMsm8TyocOoNAPwkw7yUoSSQoDLd2G27iJ340QANNaIB5389k1KgRQACBDhBg3rcziMx79RD1EBqgARqgARpIQgPM+6SSSeZ9UuHQGQT6SYB5L0FJIkFpwvTRBvOQBmiABtqtAeZ9P5NRo0YAAQQ6QIB5384gMu/VQ9RDaIAGaIAGaCAJDTDvk0ommfdJhUNnEOgnAea9BCWJBIXh1m7DTfzEjwZooAkNMO/7mYwaNQIIINABAsz7dgaRea8eoh5CAzRAAzRAA0logHmfVDLJvE8qHDqDQD8JMO8lKEkkKE2YPtpgHtIADdBAuzXAvO9nMmrUCCCAQAcIMO/bGUTmvXqIeggN0AAN0AANJKEB5n1SySTzPqlw6AwC/STAvJegJJGgMNzabbiJn/jRAA00oQHmfT+TUaNGAAEEOkCAed/OIDLv1UPUQ2iABmiABmggCQ0w75NKJpn3SYVDZxDoJwHmvQQliQSlCdNHG8xDGqABGmi3Bpj3/UxGjRoBBBDoAAHmfTuDyLxXD1EPoQEaoAEaoIEkNMC8TyqZZN4nFQ6dQaCfBJj3EpQkEhSGW7sNN/ETPxqggSY0wLzvZzJq1AgggEAHCDDv2xlE5r16iHoIDdAADdAADSShAeZ9Uskk8z6pcOgMAv0kwLyXoCSRoDRh+miDeUgDNEAD7dYA876fyahRI4AAAh0gwLxvZxCZ9+oh6iE0QAM0QAM0kIQGmPdJJZPM+6TCoTMI9JMA816CkkSCwnBrt+EmfuJHAzTQhAaY9/1MRo0aAQQQ6AAB5n07g8i8Vw9RD6EBGqABGqCBJDTAvE8qmWTeJxUOnUGgnwSY9xKUJBKUJkwfbTAPaYAGaKDdGmDe9zMZNWoEEECgAwSY9+0MIvNePUQ9hAZogAZogAaS0ADzPqlkknmfVDh0BoF+EmDeS1CSSFAYbu023MRP/GiABprQAPO+n8moUSOAAAIdIMC8b2cQmffqIeohNEADNEADNJCEBpj3SSWTzPukwqEzCPSTAPNegpJEgtKE6aMN5iEN0AANtFsDzPt+JqNGjQACCHSAAPO+nUFk3quHqIfQAA3QAA3QQBIaYN4nlUwy75MKh84g0E8CzHsJShIJCsOt3Yab+IkfDdBAExpg3vczGTVqBBBAoAMEmPftDCLzXj1EPYQGaIAGaIAGktAA8z6pZJJ5n1Q4dAaBfhJg3ktQkkhQmjB9tME8pAEaoIF2a4B5389k1KgRQACBDhBg3rcziMx79RD1EBqgARqgARpIQgPM+6SSSeZ9UuHQGQT6SYB5L0FJIkFhuLXbcBM/8aMBGmhCA8z7fiajRo0AAgh0gADzvp1BZN6rh6iH0AAN0AAN0EASGmDeJ5VMMu+TCofOINBPAsx7CUoSCUoTpo82mIc0QAM00G4NMO/7mYwaNQIIINABAsz7dgaRea8eoh5CAzRAAzRAA0logHmfVDLJvE8qHDqDQD8JMO8lKEkkKAy3dhtu4id+NEADTWiAed/PZNSoEUAAgQ4QYN63M4jMe/UQ9RAaoAEaoAEaSEIDzPukkknmfVLh0BkE+kmAeS9BSSJBacL00QbzkAZogAbarQHmfT+TUaNGAAEEOkCAed/OIDLv1UPUQ2iABmiABmggCQ0w75NKJpn3SYVDZxDoJwHmvQQliQSF4dZuw038xI8GaKAJDTDv+5mMGjUCCCDQAQLM+3YGkXmvHqIeQgM0QAM0QANJaIB5n1QyybxPKhw6g0A/CTDvJShJJChNmD7aYB7SAA3QQLs1wLzvZzJq1AgggEAHCDDv2xlE5r16iHoIDdAADdAADSShAeZ9Uskk8z6pcOgMAv0kwLyXoCSRoDDc2m24iZ/40QANNKEB5n0/k1GjRgABBDpAgHnfziAy79VD1ENogAZogAZoIAkNMO+TSiaZ90mFQ2cQ6CcB5r0EJYkEpQnTRxvMQxqgARpotwaY9/1MRo0aAQQQ6AAB5n07g8i8Vw9RD6EBGqABGqCBJDTAvE8qmWTeJxUOnUGgnwSY9xKUJBIUhlu7DTfxEz8aoIEmNMC872cyatQIIIBABwgw79sZROa9eoh6CA3QAA3QAA0koQHmfVLJJPM+qXDoDAL9JMC8l6AkkaA0Yfpog3lIAzRAA+3WAPO+n8moUSOAAAIdIMC8b2cQmffqIeohNEADNEADNJCEBpj3SSWTzPukwqEzCPSTAPNegpJEgsJwa7fhJn7iRwM00IQGmPf9TEaNGgEEEOgAAeZ9O4PIvFcPUQ+hARqgARqggSQ0wLxPKplk3icVDp1BoJ8EmPcSlCQSlCZMH20wD2mABmig3Rpg3vczGTVqBBBAoAMEmPftDCLzXj1EPYQGaIAGaIAGktAA8z6pZJJ5n1Q4dAaBfhIYa95fceUD2d33vuM/BjRAAzRAAzRAAzPXwAEHHDY2L1lYWHirn+maUSOAAAIIJE6AeZ94gEq6N9a8j4sK1UPUg2iABmiABmiABuahgfMuuHmpesh/l+QzDjdPgHnfPFMtIoBATQJLfSl4fWEBAwxogAZogAZoIA0NMO9rJnrejgACCCAwFwLM+7lgbvwkY837hTRyHzmoONAADdAADdAADYQGmPeNp4KlDTLvS9F4AQEE5kXAl78vfxqgARqgARqggbZogHk/rwzReRBAAAEE6hBg3tehlc57mfdy4LbkwPpJqzRAAzRAA8z7+eWQzPv5sXYmBBAoIeCL3xc/DdAADdAADdBAWzTAvC9J6BxGAAEEEFhWAsz7ZcU/8cmZ93LgtuTA+kmrNEADNEADzPuJU77aH2Te10bmAwgg0DQBX/y++GmABmiABmiABtqiAeZ905mg9hBAAAEEmiDAvG+C4vzbYN7LgduSA+snrdIADdAADTDv55crMu/nx9qZEECghIAvfl/8NEADNEADNEADbdEA874koXMYAQQQQGBZCTDvlxX/xCdn3suB25ID6yet0gAN0AANMO8nTvlqf5B5XxuZDyCAQNMEfPH74qcBGqABGqABGmiLBpj3TWeC2kMAAQQQaIIA874JivNvg3kvB25LDqyftEoDNEADNMC8n1+uyLyfH2tnQgCBEgK++H3x0wAN0AAN0AANtEUDzPuShM5hBBBAAIFlJcC8X1b8E5+ceS8HbksOrJ+0SgM0QAM0wLyfOOWr/UHmfW1kPoAAAk0T+D8XFhb8x4AGaGASDfxfC/UT5//b3xx/c2mABqbQwAtNJ0LaQwABBBBAoAECzPsGIC5DE59PkZNMMn/yGfNuGqABGqABGqCBSTXwvyxDrtTXUzLv+xp540YAAQQQQKADBPaewLzf3oFxGwICCCCAAAIIIIAAAsMEmPfDNDxHAAEEEEAAAQQQQKC9BJj37Y2dniOAAAIIINB7Asz73ksAAAQQQAABBBBAAIGFhQXmPRkggAACCCCAAAIIINANAsz7bsTRKBBAAAEEEOglAeZ9L8Nu0AgggAACCCCAAAIjBJj3I0D8iAACCCCAAAIIIIBASwkw71saON1GAAEEEEAAgYUF5j0VIIAAAggggAACCCBg5T0NIIAAAggggAACCCDQFQLM+65E0jgQQAABBBDoIQHmfQ+DbsgIIIAAAggggAACiwhYeb8IiQMIIIAAAggggAACCLSSAPO+lWHTaQQQQAABBBAIAsx7OkAAAQQQQAABBBBAwMp7GkAAAQQQQAABBBBAoCsEmPddiaRxIIAAAggg0EMCzPseBt2QEUAAAQQQQAABBBYRsPJ+ERIHEEAAAQQQQAABBBBoJQHmfSvDptMIIIAAAgggEASY93SAAAIIIIAAAggggICV9zSAAAIIIIAAAggggEBXCDDvuxJJ40AAAQQQQKCHBJj3PQy6ISOAAAIIIIAAAggsImDl/SIkDiCAAAIIIIAAAggg0EoCzPtWhk2nEUAAAQQQQCAIMO/pAAEEEEAAAQQQQAABK+9pAAEEEEAAAQQQQACBrhBg3nclksaBAAIIIIBADwkw73sYdENGAAEEEEAAAQQQWETAyvtFSBxAAAEEEEAAAQQQQKCVBJj3rQybTiOAAAIIIIBAEGDe0wECCCCAAAIIIIAAAlbe0wACCCCAAAIIIIAAAl0hwLzvSiSNAwEEEEAAgR4SYN73MOiGjAACCCCAAAIIILCIgJX3i5A4gAACCCCAAAIIIIBAKwkw71sZNp1GAAEEEEAAgSDAvKcDBBBAAAEEEEAAAQSsvKcBBBBAAAEEEEAAAQS6QoB535VIGgcCCCCAAAI9JMC872HQDRkBBBBAAAEEEEBgEQEr7xchcQABBBBAAAEEEEAAgVYSYN63Mmw6jQACCCCAAAJBgHlPBwgggAACCCCAAAIIWHlPAwgggAACCCCAAAIIdIUA874rkTQOBBBAAAEEekiAed/DoBsyAggggAACCCCAwCICVt4vQuIAAggggAACCCCAAAKtJMC8b2XYdBoBBBBAAAEEggDzng4QQAABBBBAAAEEELDyngYQQAABBBBAAAEEEOgKAeZ9VyJpHAgggAACCPSQAPO+h0E3ZAQQQAABBBBAAIFFBKy8X4TEAQQQQAABBBBAAAEEWkmAed/KsOk0AggggAACCAQB5j0dIIAAAggggAACCCBg5T0NIIAAAggggAACCCDQFQLM+65E0jgQQAABBBDoIQHmfQ+DbsgIIIAAAggggAACiwhYeb8IiQMIIIAAAggggAACCLSSAPO+lWHTaQQQQAABBBAIAsx7OkAAAQQQQAABBBBAwMp7GkAAAQQQQAABBBBAoCsEmPddiaRxIIAAAggg0EMCzPseBt2QEUAAAQQQQAABBBYRsPJ+ERIHEEAAAQQQQAABBBBoJQHmfSvDptMIIIAAAgggEASY93SAAAIIIIAAAggggICV9zSAAAIIIIAAAggggEBXCDDvuxJJ40AAAQQQQKCHBJj3PQy6ISOAAAIIIIAAAggsImDl/SIkDiCAAAIIIIAAAggg0EoCzPtWhk2nEUAAAQQQQCAIMO/pAAEEEEAAAQQQQAABK+9pAAEEEEAAAQQQQACBrhBg3nclksaBAAIIIIBADwkw73sYdENGAAEEEEAAAQQQWETAyvtFSBxAAAEEEEAAAQQQQKCVBJj3rQybTiOAAAIIIIBAEGDe0wECCCCAAAIIIIAAAlbe0wACCCCAAAIIIIAAAl0hwLzvSiSNAwEEEEAAgR4SYN73MOiGjAACCCCAAAIIILCIgJX3i5A4gAACCCCAAAIIIIBAKwkw71sZNp1GAAEEEEAAgSDAvKcDBBBAAAEEEEAAAQSsvKcBBBBAAAEEEEAAAQS6QoB535VIGgcCCCCAAAI9JMC872HQDRkBBBBAAAEEEEBgEQEr7xchcQABBBBAAAEEEEAAgVYSYN63Mmw6jQACCCCAAAJBgHlPBwgggAACCCCAAAIIWHlPAwgggAACCCCAAAIIdIUA874rkTQOBBBAAAEEekiAed/DoBsyAggggAACCCCAwCICVt4vQuIAAggggAACCCCAAAKtJMC8b2XYdBoBBBBAAAEEggDzng4QQAABBBBAAAEEELDyngYQQAABBBBAAAEEEOgKAeZ9VyJpHAgggAACCPSQAPO+h0E3ZAQQQAABBBBAAIFFBKy8X4TEAQQQQAABBBBAAAEEWkmAed/KsOk0AggggAACCAQB5j0dIIAAAggggAACCCBg5T0NIIAAAggggAACCCDQFQLM+65E0jgQQAABBBDoIQHmfQ+DbsgIIIAAAggggAACiwhYeb8IiQMIIIAAAggggAACCLSSAPO+lWHTaQQQQAABBBAIAsx7OkAAAQQQQAABBBBAwMp7GkAAAQQQQAABBBBAoCsEmPddiaRxIIAAAggg0EMCzPseBt2QEUAAAQQQQAABBBYRsPJ+ERIHEEAAAQQQQAABBBBoJQHmfSvDptMIIIAAAgggEASY93SAAAIIIIAAAggggICV9zSAAAIIIIAAAggggEBXCDDvuxJJ40AAAQQQQKCHBJj3PQy6ISOAAAIIIIAAAggsImDl/SIkDiCAAAIIIIAAAggg0EoCzPtWhk2nEUAAAQQQQCAIMO/pAAEEEEAAAQQQQAABK+9pAAEEEEAAAQQQQACBrhBg3nclksaBAAIIIIBADwkw73sYdENGAAEEEEAAAQQQWETAyvtFSBxAAAEEEEAAAQQQQKCVBJj3rQybTiOAAAIIIIBAEGDe0wECCCCAAAIIIIAAAlbe0wACCCCAAAIIIIAAAl0hwLzvSiSNAwEEEEAAgR4SYN73MOiGjAACCCCAAAIIILCIgJX3i5A4gAACCCCAAAIIIIBAKwkw71sZNp1GAAEEEEAAgSDAvKcDBBBAAAEEEEAAAQSsvKcBBBBAAAEEEEAAAQS6QoB535VIGgcCCCCAAAI9JMC872HQDRkBBBBAAAEEEEBgEQEr7xchcQABBBBAAAEEEEAAgVYSYN63Mmw6jQACCCCAAAJBgHlPBwgggAACCCCAAAIIWHlPAwgggAACCCCAAAIIdIUA874rkTQOBBBAAAEEekiAed/DoBsyAggggAACCCCAwCICVt4vQuIAAggggAACCCCAAAKtJMC8b2XYdBoBBBBAAAEEggDzng4QQAABBBBAAAEEELDyngYQQAABBBBAAAEEEOgKAeZ9VyJpHAgggAACCPSQAPO+h0E3ZAQQQAABBBBAAIFFBP5/9t4EfI7jLPf9bEu2ZcuLvO+75EXyLkuyLe+7La+yvG/yvsu7ZXmTZTt2NiCQYAJJSHICIYHkZgMCIYQQLvfhOXnuPZwL59zDORzgLuHChbBDIIn75h2nfVozVdPVPd0zVd0/Pc//mVFPT3fVr96p/vp7q6qZeT+ChA0QgAAEIAABCEAAAhBIkgDmfZLNRqEhAAEIQAACEBABzHt0AAEIQAACEIAABCAAAWbeowEIQAACEIAABCAAAQh0hQDmfVdaknpAAAIQgAAEekgA876HjU6VIQABCEAAAhCAAARGCDDzfgQJGyAAAQhAAAIQgAAEIJAkAcz7JJuNQkMAAhCAAAQgIAKY9+gAAhCAAAQgAAEIQAACzLxHAxCAAAQgAAEIQAACEOgKAcz7rrQk9YAABCAAAQj0kADmfQ8bnSpDAAIQgAAEIAABCIwQYOb9CBI2QAACEIAABCAAAQhAIEkCmPdJNhuFhgAEIAABCEBABDDv0QEEIAABCEAAAhCAAASYeY8GIAABCEAAAhCAAAQg0BUCmPddaUnqAQEIQAACEOghAcz7HjY6VYYABCAAAQhAAAIQGCHAzPsRJGyAAAQgAAEIQAACEIBAkgQw75NsNgoNAQhAAAIQgIAIYN6jAwhAAAIQgAAEIAABCDDzHg1AAAIQgAAEIAABCECgKwQw77vSktQDAhCAAAQg0EMCmPc9bHSqDAEIQAACEIAABCAwQoCZ9yNI2AABCEAAAhCAAAQgAIEkCWDeJ9lsFBoCEIAABCAAARHAvEcHEIAABCAAAQhAAAIQYOY9GoAABCAAAQhAAAIQgEBXCGDed6UlqQcEIAABCECghwQw73vY6FQZAhCAAAQgAAEIQGCEADPvR5CwAQIQgAAEIAABCEAAAkkSwLxPstkoNAQgAAEIQAACIoB5jw4gAAEIQAACEIAABCDAzHs0AAEIQAACEIAABCAAga4QwLzvSktSDwhAAAIQgEAPCWDe97DRqTIEIAABCEAAAhCAwAgBZt6PIGEDBCAAAQhAAAIQgAAEkiSAeZ9ks1FoCEAAAhCAAAREAPMeHUAAAhCAAAQgAAEIQICZ92gAAhCAAAQgAAEIQAACXSGAed+VlqQeEIAABCAAgR4SwLzvYaNTZQhAAAIQgAAEIACBEQLMvB9BwgYIQAACEIAABCAAAQgkSQDzPslmo9AQgAAEIAABCIgA5j06gAAEIAABCEAAAhCAADPv0QAEIAABCEAAAhCAAAS6QgDzvistST0gAAEIQAACPSSAed/DRqfKEIAABCAAAQhAAAIjBJh5P4KEDRCAAAQgAAEIQAACEEiSAOZ9ks1GoSEAAQhAAAIQEAHMe3QAAQhAAAIQgAAEIAABZt6jAQhAAAIQgAAEIAABCHSFAOZ9V1qSekAAAhCAAAR6SADzvoeNTpUhAAEIQAACEIAABEYIMPN+BAkbIAABCEAAAhCAAAQgkCQBzPskm41CQwACEIAABCAgApj36AACEIAABCAAAQhAAALMvEcDEIAABCAAAQhAAAIQ6AoBzPuutCT1gAAEIAABCPSQAOZ9DxudKkMAAhCAAAQgAAEIjBBg5v0IEjZAAAIQgAAEIAABCEAgSQKY90k2G4WGAAQgAAEIQEAEMO/RAQQgAAEIQAACEIAABJh5jwYgAAEIQAACEIAABCDQFQKY911pSeoBAQhAAAIQ6CEBzPseNjpVhgAEIAABCEAAAhAYIcDM+xEkbIAABCAAAQhAAAIQgECSBDDvk2w2Cg0BCEAAAhCAgAhg3qMDCEAAAhCAAAQgAAEIMPMeDUAAAhCAAAQgAAEIQKArBDDvu9KS1AMCEIAABCDQQwKY9z1sdKoMAQhAAAIQgAAEIDBCgJn3I0jYAAEIQAACEIAABCAAgSQJYN4n2WwUGgIQgAAEIAABEcC8RwcQgAAEIAABCEAAAhBg5j0agAAEIAABCEAAAhCAQFcIYN53pSWpBwQgAAEIQKCHBDDve9joVBkCEIAABCAAAQhAYIQAM+9HkLABAhCAAAQgAAEIQAACSRLAvE+y2Sg0BCAAAQhAAAIigHmPDiAAAQhAAAIQgAAEIMDMezQAAQhAAAIQgAAEIACBrhDAvO9KS1IPCEAAAhCAQA8JYN73sNGpMgQgAAEIQAACEIDACAFm3o8gYQMEIAABCEAAAhCAAASSJIB5n2SzUWgIQAACEIAABEQA8x4dQAACEIAABCAAAQhAgJn3aAACEIAABCAAAQhAAAJdIYB535WWpB4QgAAEIACBHhLAvO9ho1NlCEAAAhCAAAQgAIERAsy8H0HCBghAAAIQgAAEIAABCCRJAPM+yWaj0BCAAAQgAAEIiADmPTqAAAQgAAEIQAACEIAAM+/RAAQgAAEIQAACEIAABLpCAPO+Ky1JPSAAAQhAAAI9JIB538NGp8oQgAAEIAABCEAAAiMEmHk/goQNEIAABCAAAQhAAAIQSJIA5n2SzUahIQABCEAAAhAQAcx7dAABCEAAAhCAAAQgAAFm3qMBCEAAAhCAAAQgAAEIdIUA5n1XWpJ6QAACEIAABHpIAPO+h41OlSEAAQhAAAIQgAAERggw834ECRsgAAEIQAACEIAABCCQJAHM+ySbjUJDAAIQgAAEICACmPfoAAIQgAAEIAABCEAAAsy8RwMQgAAEIAABCEAAAhDoCgHM+/I7PBEAACAASURBVK60JPWAAAQgAAEI9JAA5n0PG50qQwACEIAABCAAAQiMEGDm/QgSNkAAAhCAAAQgAAEIQCBJApj3STYbhYYABCAAAQhAQAQw79EBBCAAAQhAAAIQgAAEmHmPBiAAAQhAAAIQgAAEINAVApj3XWlJ6gEBCEAAAhDoIQHM+x42OlWGAAQgAAEIQAACEBghwMz7ESRsgAAEIAABCEAAAhCAQJIEMO+TbDYKDQEIQAACEICACGDeowMIQAACEIAABCAAAQgw8x4NQAACEIAABCAAAQhAoCsEMO+70pLUAwIQgAAEINBDApj3PWx0qgwBCEAAAhCAAAQgMEKAmfcjSNgAAQhAAAIQgAAEIACBJAlg3ifZbBQaAhCAAAQgAAERwLxHBxCAAAQgAAEIQAACEGDmPRqAAAQgAAEIQAACEIBAVwhg3nelJakHBCAAAQhAoIcEMO972OhUGQIQgAAEIAABCEBghAAz70eQsAECEIAABCAAAQhAAAJJEsC8T7LZKDQEIAABCEAAAiKAeY8OIAABCEAAAhCAAAQgwMx7NAABCEAAAhCAAAQgAIGuEMC870pLUg8IQAACEIBADwlg3vew0akyBCAAAQhAAAIQgMAIAWbejyBhAwQgAAEIQAACEIAABJIkgHmfZLNRaAhAAAIQgAAERADzHh1AAAIQgAAEIAABCECAmfdoAAIQgAAEIAABCEAAAl0hgHnflZakHhCAAAQgAIEeEsC872GjU2UIQAACEIAABCAAgRECzLwfQcIGCEAAAhCAAAQgAAEIJEkA8z7JZqPQEIAABCAAAQiIAOY9OoAABCAAAQhAAAIQgAAz79EABCAAAQhAAAIQgAAEukIA874rLUk9IAABCEAAAj0kgHnfw0anyhCAAAQgAAEIQAACIwSYeT+ChA0QgAAEIAABCEAAAhBIkgDmfZLNRqEhAAEIQAACEBABzHt0AAEIQAACEIAABCAAAWbeowEIQAACEIAABCAAAQh0hQDmfVdaknpAAAIQgAAEekgA876HjU6VIQABCEAAAhCAAARGCDDzfgQJGyAAAQhAAAIQgAAEIJAkAcz7JJuNQkMAAhCAAAQgIAKY9+gAAhCAAAQgAAEIQAACzLxHAxCAAAQgAAEIQAACEOgKAcz7rrQk9YAABCAAAQj0kADmfQ8bnSpDAAIQgAAEIAABCIwQYOb9CBI2QAACEIAABCAAAQhAIEkCmPdJNhuFhgAEIAABCEBABDDv0QEEIAABCEAAAhCAAASYeY8GIAABCEAAAhCAAAQg0BUCmPddaUnqAQEIQAACEOghAcz7HjY6VYYABCAAAQhAAAIQGCHAzPsRJGyAAAQgAAEIQAACEIBAkgQw75NsNgoNAQhAAAIQSIPAEjM7t8W/S80sq/j38y2WJ6/rtmk0D6WEAAQgAAEIQAACEOgIAcz7jjQk1YAABCAAAQhAAAIQ6D0BzPveSwAAEIAABCAAgfYInFPRWK9qxMe4/zfMbIv2kHJkCEAAAhCAAAQgAAEIjBDAvB9BwgYIQAACEIAABCAAAQgkSQDzPslmo9AQgAAEIACBdAh8sUcG/vfM7Lh0moaSQgACEIAABCAAAQh0hADmfUcakmpAAAIQgAAEIAABCPSeAOZ97yUAAAhAAAIQgEC7BA41s+/0xMB/o12UHB0CEIAABCAAAQhAAAJOApj3TixshAAEIAABCEAAAhCAQKMEtOLqgpb/nqiRS1/Wcpl2bJQiB4MABCAAAQhAYOYE3l0j4IhxSfxxZfobM9t95qQpAAQgAAEIQAACEIBAHwlg3vex1akzBCAAAQhAAAIQgMAsCHytB7nu4Tz4C7MAzTkhAAEIQAACEGiPwA5m9q2OBzWPtIePI0MAAhCAAAQgAAEIQGAsAcz7sXj4EAIQgAAEIAABCEAAAo0R0GNT9fjUYYO7q///MzOb1xg9DgQBCEAAAhCAQDQE7uhwQPOfzWxuNKQpCAQgAAEIQAACEIBA3whg3vetxakvBCAAAQhAAAIQgMAsCfxUh3Pdw4MQrpklaM4NAQhAAAIQgEB7BLY0s292NKi5qD1sHBkCEIAABCAAAQhAAAKlBDDvSxGxAwQgAAEIQAACEIAABBojoMen6jGqw0Z31/7/22a2RWPUOBAEIAABCEAAAtERONnM3uxYUPOl6ChTIAhAAAIQgAAEIACBmAjs88NHSH3bzNr6+7eKMfb3WyxLXsdzYmoEygIBCEAAAhCAAAQgAIGGCTxaMQZPzdjXPcPShplxOAhAAAIQgAAEIiTwyQ4FNUqSHh4hY4oEAQhAAAIQgAAEIBAXgVc6FAOHJB2/HBd+SgMBCEAAAhCAAAQgAIHGCcwxsz/ocJz/wcaJcUAIQAACEIAABKIksJ+Z/WNHgpr3REmYQkEAAhCAAAQgAAEIxEZgOzP7s47EwGXm/XfNbElsDUB5IAABCEAAAhCAAAQg0AIBPU61LD5O8fO/NbM9WuDFISEAAQhAAAIQiJTAxg4ENX9hZjtHypdiQQACEIAABCAAAQjER+CmDsTAIYnH98WHnhJBAAIQgAAEIAABCECgNQJ6rGpInJzSPo+1RosDQwACEIAABCAQJYF5ZvaniQc1d0VJlkJBAAIQgAAEIAABCMRKYAsz+0biMXBZwvGvzWzXWBuAckEAAhCAAAQgAAEIQKAFAoeZ2Xc6FOf/VzPbpgVOHBICEIAABCAAgcgJXJdwQPO/mdlWkfOleBCAAAQgAAEIQAAC8RE4wcy+n3AcXGbe3xsfckoEAQhAAAIQgAAEIACB1gno8aplsXIqn1/cOi1OAAEIQAACEIBAtAS+nmhQc0a0RCkYBCAAAQhAAAIQgEDsBD6aaAxclmz8fQa4xi49ygcBCEAAAhCAAAQg0BKBHczszzsQ5/96S3w4LAQgAAEIQAACiRA43sy+l1hQ8wuJsKWYEIAABCAAAQhAAAJxEtjTzP4usRi4zLjX5+fFiZtSQQACEIAABCAAAQhAYCoE9JjVkLg51n2+a2aLp0KKk0AAAhCAAAQgEDWBn0koqPlnMzsoapoUDgIQgAAEIAABCEAgBQLrE4qBQ5KLv5QCdMoIAQhAAAIQgAAEIACBFglsaWb/PuE4/70tsuHQEIAABCAAAQgkRGAPM/vbRIKajQlxpagQgAAEIAABCEAAAvES2NrM/iiRGLjMvP+OmR0WL2pKBgEIQAACEIAABCAAgakROMXM3kwwzv9LM9t5apQ4EQQgAAEIQAAC0RN4IoGA5v8ys+2jJ0kBIQABCEAAAhCAAARSIXBlAjFwmXGvz19JBTjlhAAEIAABCEAAAhCAwBQI6LGrIXF0TPvcPQUunAICEIAABCAAgYQIaObR/xF5UHN9QjwpKgQgAAEIQAACEIBAGgR+LfIYuCyh+OdmtmMaqCklBCAAAQhAAAIQgAAEpkJgPzP7p4Ti/P9gZltNhQwngQAEIAABCEAgKQKrIg5oftfMtkiKJoWFAAQgAAEIQAACEEiBwFFm9t2I4+Ay8/7mFCBTRghAAAIQgAAEIAABCEyZwEsJxfhnTpkNp4MABCAAAQhAICECvxJhUPN9MzspIYYUFQIQgAAEIAABCEAgLQLvjzAGLjPt9fn/wgDXtIRGaSEAAQhAAAIQgAAEpkZgnpn9aQJx/qenRoQTQQACEIAABCCQJIEjzOzfIgtqPpQkSQoNAQhAAAIQgAAEIJAKgQVm9leRxcBl5v2bZrY8FcCUEwIQgAAEIAABCEAAAjMgcEPkMf6/mNlBM+DCKSEAAQhAAAIQSIzAj0YU1Py9me2dGD+KCwEIQAACEIAABCCQHoGHIoqBy4x7ff7R9BBTYghAAAIQgAAEIAABCEyVgB7D+tsRx/mbpkqDk0EAAhCAAAQgkCyBnc3sLyMJap5MliIFhwAEIAABCEAAAhBIicBWZvYfI4mBy8z7fzCzfVKCS1khAAEIQAACEIAABCAwIwLHm5key1oWY0/78//bzLafERNOCwEIQAACEIBAggTujSCg+W9mtk2C7CgyBCAAAQhAAAIQgECaBM6OIAYOSRo+kyZeSg0BCEAAAhCAAAQgAIGZEPhwhHH+jTMhwUkhAAEIQAACEEiWwJZm9s0ZBzWXJkuPgkMAAhCAAAQgAAEIpErgczOOgcvM+z82s21ThUu5IQABCEAAAhCAAAQgMAMCe5jZ30YU5/+umWlJf/5BAAIQgAAEIACBSgRWmtmbMwpqfqNSSdkZAhCAAAQgAAEIQAACzRA4xMy+M6MYuMy41+dXNlNNjgIBCEAAAhCAAAQgAIFeEXgqkhhfS/gv6xV5KgsBCEAAAhCAQKMEfnEGQc13zWxJo7XgYBCAAAQgAAEIQAACEAgn8PoMYuAQ4/6r4VVgTwhAAAIQgAAEIAABCECgQGBrM/svEcT5HymUibcQgAAEIAABCECgMoH9zeyfphzU/FjlUvIFCEAAAhCAAAQgAAEINEdgBzP71pRj4DLz/ntmdkxzVeRIEIAABCAAAQhAAAIQ6B2By2Yc4/+9me3dO+pUGAIQgAAEIACBxgm8PMWg5q/NbNfGa8ABIQABCEAAAhCAAAQgUI3A2inGwGXGvT5/f7XiszcEIAABCEAAAhCAAAQg4CDwqzOM8592lIdNEIAABCAAAQhAoDKB7czsz6YU1NxXuXR8AQIQgAAEIAABCEAAAs0T2NLMfm9KMXCZef/tH5j3uzVfRY4IAQhAAAIQgAAEIACB3hE40sz+bQZx/h+b2Ta9o02FIQABCEAAAhBojcBNUwho/sDM5rRWAw4MAQhAAAIQgAAEIACBagRONrM3pxAHl5n3D1crNntDAAIQgAAEIAABCEAAAmMIvG8GMf7lY8rDRxCAAAQgAAEIQKAygS3M7BstBzXnVS4VX4AABCAAAQhAAAIQgEC7BD7RcgxcZtz/JzOb224VOToEIAABCEAAAhCAAAR6RWCBmf1/U4zzv9orulQWAhCAAAQgAIGpETjBzL7fUlDzS1OrBSeCAAQgAAEIQAACEIBAOIF9zewfW4qBy4x7fX5heFHZEwIQgAAEIAABCEAAAhAIJHD/lGL875nZ0YFlYjcIQAACEIAABCBQmcDPthDUfMfMDqtcEr4AAQhAAAIQgAAEIACB6RB4voUYOMS4//x0qsdZIAABCEAAAhCAAAQg0DsCW5nZ708hzv/x3pGlwhCAAAQgAAEITJXAnmb2dw0HNa9MtQacDAIQgAAEIAABCEAAAtUIbGtmf9JwDFxm3v+rmS2qVkz2hgAEIAABCEAAAhCAAAQqEDir5Rj/22a2a4XysCsEIAABCEAAAhCoRWB9g0HNn5vZjrVKwZcgAAEIQAACEIAABCAwPQJrGoyBy4x7ff7O6VWNM0EAAhCAAAQgAAEIQKC3BD7TYpz/QG+pUnEIQAACEIAABKZKYGsz+6OGgpqbp1pyTgYBCEAAAhCAAAQgAIH6BL7WUAxcZt7/hZntVL+YfBMCEIAABCAAAQhAAAIQCCRwsJn9Swtx/h+a2ZzAMrAbBCAAAQhAAAIQmJjAFQ0ENN80sy0nLgkHgAAEIAABCEAAAhCAwHQIHGtm32sgDi4z72+fTnU4CwQgAAEIQAACEIAABCBgZq+2EOOfD1kIQAACEIAABCAwbQK/MUFQ86aZrZh2gTkfBCAAAQhAAAIQgAAEJiTwUxPEwGWmvT7/XxngOmEL8XUIQAACEIAABCAAAQhUIzDfzP6fBuP8z1Y7PXtDAAIQgAAEIACBZggcZWbfrRnUfKyZInAUCEAAAhCAAAQgAAEITJXA7mb2NzVj4DLzXgNcT5tqbTgZBCAAAQhAAAIQgAAEICACtzQU4/+rmS0EKQQgAAEIQAACEJgVgffXCGr+wcz2mVWBOS8EIAABCEAAAhCAAAQmJPBYjRi4zLjX5z83Ybn4OgQgAAEIQAACEIAABCBQj8AWZvY7DcT576h3er4FAQhAAAIQgAAEmiGwwMz+qmJQs6GZU3MUCEAAAhCAAAQgAAEIzITAHDP7g4oxcJl5/89mduBMasNJIQABCEAAAhCAAAQgAAERONHMvj9BnP//mtmOoIQABCAAAQhAAAKzJvBQhYDmv5vZtrMuMOeHAAQgAAEIQAACEIDAhATOqxADlxn3+vyFCcvD1yEAAQhAAAIQgAAEIACByQl8fII4/7bJT88RIAABCEAAAhCAwOQEqsw8umry03EECEAAAhCAAAQgAAEIREHglydI7BUN/f/TzLaLokYUAgIQgAAEIAABCEAAAv0msKeZ/V2NOP+bZrZlv9FRewhAAAIQgAAEYiJwdkBA85sxFZiyQAACEIAABCAAAQhAYEICh5nZdwLi4KJR73p/zYTl4OsQgAAEIAABCEAAAhCAQHME9NhXV9zu2/amma1s7vQcCQIQgAAEIAABCDRD4HNjgprvmdkxzZyGo0AAAhCAAAQgAAEIQCAaAu8dEwP7knvF7f+zmW0RTW0oCAQgAAEIQAACEIAABCCwtZn9UYU4X0vt8w8CEIAABCAAAQhER+CQMTOPPhBdaSkQBCAAAQhAAAIQgAAEJiewk5n9RYXEXtG4/76ZnTB5ETgCBCAAAQhAAAIQgAAEINAwAT3+tRi7+97/k5kd0PC5ORwEIAABCEAAAhBojMDrjqDm22a2W2Nn4EAQgAAEIAABCEAAAhCIi8DdjhjYl9wrbv/puKpBaSAAAQhAAAIQgAAEIACBAoGvBsT5zxX25y0EIAABCEAAAhCIjsAOZvatoaDm4ehKSYEgAAEIQAACEIAABCDQHIEtzezfD8XARZPe9f7vzWyv5orAkSAAAQhAAAIQgAAEIACBhgnoMbB6HKwrnte2PzGzeQ2fk8NBAAIQgAAEIACBxgmsLQQ0/8nM5jZ+Bg4IAQhAAAIQgAAEIACBuAicamZvFuJgX4Iv3/54XMWnNBCAAAQgAAEIQAACEICAg8BPjonxVzv2ZxMEIAABCEAAAhCIjoBmHv3eD4OaC6MrHQWCAAQgAAEIQAACEIBAOwQ+NSaxl5v2ev2vZrZNO0XgqBCAAAQgAAEIQAACEIBAgwR2MbO/csT53zCzLRo8D4eCAAQgAAEIQAACrRI42cw+1+oZODgEIAABCEAAAhCAAATiInCgmf2zI7FXNO71flVcxaY0EIAABCAAAQhAAAIQgMAYAuuGYnwtpX/smP35CAIQgMDEBO4yM/5ggAbQQNMaeIi+hb4VDaCBFjRw3MSRDweAAAQgAAEItEdg01Bib9i4/9X2Ts2RIQABCEAAAhCAAAQgAIEWCMwxs/+9EOe/0cI5OCQEIACBzQgMJxP4vxkMYIAG0AAaQANoIE4NbNgsiuE/EIAABCAAgbgIzDOzPy0k9orxxHfNbHFcxaU0EIAABCAAAQhAAAIQgEAAgXN+GOP/jZntHrA/u0AAAhCYiEAxmcD7OI0K2oV2QQNoAA2gATTwlgYw7ycK+/gyBCAAAQhMgcANHvP+R6Zwbk4BAQhAAAIQgAAEIAABCLRD4Atm9kg7h+aoEIAABDYngBmAIYQG0AAaQANoAA2kogHM+83jOP4HAQhAAALxEdjCzH57yMD/azPbJb6iUiIIQAACEIAABCAAAQhAIJDA/mY2N3BfdoMABCAwEYFUkvWUE2MJDaABNIAG0AAawLyfKOzjyy0RONTMfoI/GKABNFDQwC+Y2ZsFA/9rhc/oL+gz0QAamLUGFLvwDwJ9InA+12HiNDSABtAAGkADaCAhDfB890JCBVMIUwgNoAE0gAbQQNwawLzvU4otnbqeRjzJPQUaQANoAA2gATSQkAYUu/APAn0ioPtI7vVhgAbQABpAA2gADaSiAQIXgjc0gAbQABpAA2ggGQ1g3vcpxZZOXTHvuflL5eaPcqJVNIAG0AAakAYw79OJMylpMwQw7+n7uP6hATSABtAAGkhJA8kk61OCSlnpBNAAGkADaAANtKMBzPtmklccpVkCmPft/N7pR+GKBtAAGkADaKAdDWDeNxsLcrT4CWDet9OX0EfDFQ2gATSABtBAOxoYb97vsMMu2YIFe/EHAzSABtAAGkADaKB1DWy11dyygA/zPv7EWB9LiHnfzo1KWX/A53BHA2gADaABNFBPA5j3fYxY+11nzPt6fQV9LNzQABpAA2gADcxGA+PN+zvu+tHstXd9gz8YoAE0gAbQABpAA61rYL/9jywLCDHv+510i7X2mPezuZEp6y/4nHZBA2gADaABNODWAOZ9rFEl5WqLAOa9uy+gj4QLGkADaAANoIE4NYB5z+AEBmegATSABtAAGohDA5j3beWqOG7LBDDv47zR4QaUdkEDaAANoAE04NYA5n3LwSGHj44A5r27L6CPhAsaQANoAA2ggTg1gHmPYROHYUM70A5oAA2gATSAeR9dkosChRHAvI/zRocbUNoFDaABNIAG0IBbA5j3YTEee3WHAOa9uy+gj4QLGkADaAANoIE4NYB5j1mEWYQG0AAaQANoIA4NYN53JzvWs5qUmvdHHHlKduRRK/mDARpAA2gADaABNNCqBhRzWHkCEvO+Z8Eq1bWx5v228+a3+rvkPoD7IDSABtAAGkADaCDXwIEHLgmJ1zHvMWziMGxoB9oBDaABNIAGMO9JqyVKoNS8f/kdv5nRx9HHoQE0gAbQABpAA21rQDEH5n2iESXFbpPAWPNe96Ft/zY5Pv0/GkADaAANoAE0IA3ccdePhsTrmPf8YPjBoAE0gAbQABqIQwOY923mqzh2iwQw798VRx9CX047oAE0gAbQQN81gHnfYsTHoVMmgHlPvM4ADTSABtAAGkADUWgA8x4hRiHEvt84U3+SR2gADaCBcA1g3qecD+t12THvibuJu9EAGkADaAANRKEBzPtex6RU3k8A854+Ooo+mvxQeH4IVrBCA2igqxrAvCcoIShBA2gADaABNJCUBjDv/dkmPomaAOY9fW1SfW1Xb4CpF8kdNIAG0MA3Msz7qGNGCjc7Apj3xOvE62gADaABNIAGotAA5j1CjEKI3DyTQEEDaAANoIFQDWDezy6bxZknIoB5T9xN3I0G0AAaQANoIAoNYN5PFNPx5e4SwLynj46ijw7NjbAfeTQ0gAbQQHc1gHlPUEJQggbQABpAA2ggKQ1g3nc3W9bxmmHe09cm1deSBOhuEoC2pW3RABrAvO941En16hLAvCdeJ15HA2gADaABNBCFBjDvEWIUQiR5QPIADaABNIAGQjWAeV83F8X3ZkwA8564m7gbDaABNIAG0EAUGsC8n3FUyOljJYB5Tx8dRR8dmhthP/JoaAANoIHuagDznqCEoAQNoAE0gAbQQFIawLyPNddFuUoIYN7T1ybV15IE6G4SgLalbdEAGsC8L4na+LivBDDvideJ19EAGkADaAANRKEBzHuEGIUQSR6QPEADaAANoIFQDWDe9zWXlny9Me+Ju4m70QAaQANoAA1EoQHM++TjSirQDgHMe/roKPro0NwI+5FHQwNoAA10VwOY9wQlBCVoAA2gATSABpLSAOZ9O5kqjto6Acx7+tqk+lqSAN1NAtC2tC0aQAOY963HfZwgTQKY98TrxOtoAA2gATSABqLQAOY9QoxCiCQPSB6gATSABtBAqAYw79PMhFFqw7wn7ibuRgNoAA2gATQQhQYw74lMIeAkgHlPHx1FHx2aG2E/8mhoAA2gge5qAPOeoISgBA2gATSABtBAUhrAvHcmmtgYPwHMe/rapPpakgDdTQLQtrQtGkADmPfxB46UcCYEMO+J14nX0QAaQANoAA1EoQHMe4QYhRBJHpA8QANoAA2ggVANYN7PJJHFSScngHlP3E3cjQbQABpAA2ggCg1g3k8e2HGEThLAvKePjqKPDs2NsB95NDSABtBAdzWAeU9QQlCCBtAAGkADaCApDWDedzJR1odKYd7T1ybV15IE6G4SgLalbdEAGsC870PoSR1rEMC8J14nXkcDaAANoAE0EIUGMO8RYhRCJHlA8gANoAE0gAZCNYB5XyMNxVdiIIB5T9xN3I0G0AAaQANoIAoNYN7HEBpShggJYN7TR0fRR4fmRtiPPBoaQANooLsawLwnKCEoQQNoAA2gATSQlAYw7yNMc1GkEAKY9/S1SfW1JAG6mwSgbWlbNIAGMO9DQjf26SEBzHvideJ1NIAG0AAaQANRaADzHiFGIUSSByQP0AAaQANoIFQDmPc9TKN1o8qY98TdxN1oAA2gATSABqLQAOZ9N4JLatE4Acx7+ugo+ujQ3Aj7kUdDA2gADXRXA5j3BCUEJWgADaABNIAGktIA5n3jSSoOOB0CmPf0tUn1tSQBupsEoG1pWzSABjDvpxP8cZbkCGDeE68Tr6MBNIAG0AAaiEIDmPcIMQohkjwgeYAG0AAaQAOhGsC8Ty4JRoHfIoB5T9xN3I0G0AAaQANoIAoNYN4TnkLASQDznj46ij46NDfCfuTR0AAaQAPd1QDmPUEJQQkaQANoAA2ggaQ0gHnvTDSxMX4CmPf0tUn1tSQBupsEoG1pWzSABjDv4w8cKeFMCGDeE68Tr6MBNIAG0AAaiEIDmPcIMQohkjwgeYAG0AAaQAOhGsC8n0kii5NOTgDznribuBsNoAE0gAbQQBQawLyfPLDjCJ0kgHlPHx1FHx2aG2E/8mhoAA2gge5qAPOeoISgBA2gATSABtBAUhrAvO9koqwPlcK8p69Nqq8lCdDdJABtS9uiATSAed+H0JM61iCAeU+8TryOBtAAGkADaCAKDWDeI8QohEjygOQBGkADaAANhGoA875GGoqvxEAA8564m7gbDaABNIAG0EAUGsC8jyE0pAwREsC8p4+Ooo8OzY2wH3k0NIAG0EB3NYB5T1BCUIIG0AAaQANoICkNYN5HmOaiSCEEMO/pa5Pqa0kCdDcJQNvStmgADWDeh4Ru7NNDApj3xOvE62gADaABNIAGotAA5j1CjEKIC2AedQAAIABJREFUJA9IHqABNIAG0ECoBjDve5hG60aVMe+Ju4m70QAaQANoAA1EoQHM+24El9SicQKY9/TRUfTRobkR9iOPhgbQABrorgYw7wlKCErQABpAA2gADSSlAcz7xpNUHHA6BDDv6WuT6mtJAnQ3CUDb0rZoAA1g3k8n+OMsyRHAvCdeJ15HA2gADaABNBCFBjDvEWIUQiR5QPIADaABNIAGQjWAeZ9cEowCv0UA8564m7gbDaABNIAG0EAUGsC8JzyFgJMA5j19dBR9dGhuhP3Io6EBNIAGuqsBzHuCEoISNIAG0AAaQANJaQDz3ploYmP8BDDv6WuT6mtJAnQ3CUDb0rZoAA1g3scfOFLCmRDAvCdeJ15HA2gADaABNBCFBjDvEWIUQiR5QPIADaABNIAGQjWAeT+TRBYnnZwA5j1xN3E3GkADaAANoIEoNIB5P3lgxxE6SQDznj46ij46NDfCfuTR0AAaQAPd1QDmPUEJQQkaQANoAA2ggaQ0gHnfyURZHyqFeU9fm1RfSxKgu0kA2pa2RQNoAPO+D6EndaxBAPOeeJ14HQ2gATSABtBAFBrAvEeIUQiR5AHJAzSABtAAGgjVAOZ9jTQUX4mBAOY9cTdxNxpAA2gADaCBKDSAeR9DaEgZIiSAeU8fHUUfHZobYT/yaGgADaCB7moA856ghKAEDaABNIAG0EBSGsC8jzDNRZFCCGDe09cm1deSBOhuEoC2pW3RABrAvA8J3dinhwQw74nXidfRABpAA2gADUShAcx7hBiFEEkekDxAA2gADaCBUA1g3vcwjdaNKmPeE3cTd6MBNIAG0AAaiEIDmPfdCC6pReMEMO/po6Poo0NzI+xHHg0NoAE00F0NYN4TlBCUoAE0gAbQABpISgOY940nqTjgdAhg3tPXJtXXkgTobhKAtqVt0QAawLyfTvDHWZIjgHlPvE68jgbQABpAA2ggCg1g3iPEKIRI8oDkARpAA2gADYRqAPM+uSQYBX6LAOY9cTdxNxpAA2gADaCBKDSAeU94CgEnAcx7+ugo+ujQ3Aj7kUdDA2gADXRXA5j3BCUEJWgADaABNIAGktIA5r0z0cTG+Alg3tPXJtXXkgTobhKAtqVt0QAawLyPP3CkhDMhgHlPvE68jgbQABpAA2ggCg1g3iPEKIRI8oDkARpAA2gADYRqAPN+JoksTjo5Acx74m7ibjSABtAAGkADUWgA837ywI4jdJIA5j19dBR9dGhuhP3Io6EBNIAGuqsBzHuCEoISNIAG0AAaQANJaQDzvpOJsj5UCvOevjapvpYkQHeTALQtbYsG0ADmfR9CT+pYgwDmPfE68ToaQANoAA2ggSg0gHmPEKMQIskDkgdoAA2gATQQqgHM+xppKL4SAwHMe+Ju4m40gAbQABpAA1FoAPM+htCQMkRIAPOePjqKPjo0N8J+5NHQABpAA93VAOY9QQlBCRpAA2gADaCBpDSAeR9hmosihRDAvKevTaqvJQnQ3SQAbUvbogE0gHkfErqxTw8JYN4TrxOvowE0gAbQABqIQgOY9wgxCiGSPCB5gAbQABpAA6EawLzvYRqtG1XGvCfuJu5GA2gADaABNBCFBjDvuxFcUovGCWDe00dH0UeH5kbYjzwaGkADaKC7GsC8JyghKEEDaAANoAE0kJQGMO8bT1JxwOkQwLynr02qryUJ0N0kAG1L26IBNIB5P53gj7MkRwDznnideB0NoAE0gAbQQBQawLxHiFEIkeQByQM0gAbQABoI1QDmfXJJMAr8FgHMe+Ju4m40gAbQABpAA1FoAPOe8BQCTgKY9/TRUfTRobkR9iOPhgbQABrorgYw7wlKCErQABpAA2gADSSlAcx7Z6KJjfETwLynr02qryUJ0N0kAG1L26IBNIB5H3/gSAlnQgDznnideB0NoAE0gAbQQBQawLxHiFEIkeQByQM0gAbQABoI1QDm/UwSWZx0cgKY98TdxN1oAA2gATSABqLQAOb95IEdR+gkAcx7+ugo+ujQ3Aj7kUdDA2gADXRXA5j3BCUEJWgADaABNIAGktIA5n0nE2V9qBTmPX1tUn0tSYDuJgFoW9oWDaABzPs+hJ7UsQYBzHvideJ1NIAG0AAaQANRaADzHiFGIUSSByQP0AAaQANoIFQDmPc10lB8JQYCmPfE3cTdaAANoAE0gAai0ADmfQyhIWWIkADmPX10FH10aG6E/cijoQE0gAa6qwHMe4ISghI0gAbQABpAA0lpAPM+wjQXRQohgHlPX5tUX0sSoLtJANqWtkUDaADzPiR0Y58eEsC8J14nXkcDaAANoAE0EIUGMO8RYhRCJHlA8gANoAE0gAZCNYB538M0WjeqjHlP3E3cjQbQABpAA2ggCg1g3ncjuKQWjRPAvKePjqKPDs2NsB95NDSABtBAdzWAeU9QQlCCBtAAGkADaCApDWDeN56k4oDTIYB5T1+bVF9LEqC7SQDalrZFA2gA8346wR9nSY4A5j3xOvE6GkADaAANoIEoNIB5jxCjECLJA5IHaAANoAE0EKoBzPvkkmAU+C0CmPfE3cTdaAANoAE0gAai0ADmPeEpBJwEMO/po6Poo0NzI+xHHg0NoAE00F0NYN4TlBCUoAE0gAbQABpISgOY985EExvjJ4B5T1+bVF9LEqC7SQDalrZFA2gA8z7+wJESzoQA5j3xOvE6GkADaAANoIEoNIB5jxCjECLJA5IHaAANoAE0EKoBzPuZJLI46eQEMO+Ju4m70QAaQANoAA1EoQHM+8kDO47QSQKY9/TRUfTRobkR9iOPhgbQABrorgYw7wlKCErQABpAA2gADSSlAcz7TibK+lApzHv62qT6WpIA3U0C0La0LRpAA5j3fQg9qWMNApj3xOvE62gADaABNIAGotAA5j1CjEKIJA9IHqABNIAG0ECoBjDva6Sh+EoMBDDvibuJu9EAGkADaAANRKEBzPsYQkPKECEBzHv66Cj66NDcCPuRR0MDaAANdFcDmPcEJQQlaAANoAE0gAaS0gDmfYRpLooUQgDznr42qb6WJEB3kwC0LW2LBtAA5n1I6MY+PSSAeU+8TryOBtAAGkADaCAKDWDeI8QohEjygOQBGkADaAANhGoA876HabRuVBnznribuBsNoAE0gAbQQBQawLzvRnBJLRongHlPHx1FHx2aG2E/8mhoAA2gge5qAPOeoISgBA2gATSABtBAUhrAvG88ScUBp0MA856+Nqm+liRAd5MAtC1tiwbQAOb9dII/zpIcAcx74nXidTSABtAAGkADUWgA8x4hRiFEkgckD9AAGkADaCBUA5j3ySXBKPBbBDDvibuJu9EAGkADaAANRKEBzHvCUwg4CWDe00dH0UeH5kbYjzwaGkADaKC7GsC8JyghKEEDaAANoAE0kJQGMO+diSY2xk8A856+Nqm+liRAd5MAtC1tiwbQAOZ9/IEjJZwJAcx74nXidTSABtAAGkADUWgA8x4hRiFEkgckD9AAGkADaCBUAz0077ecSeqKkzZNAPOeuJu4Gw2gATSABtBAFBrAvG86zON4HSGAeU8fHUUfHZobYT/yaGgADaCB7moA856ghKAEDaABNIAG0EBSGuiReb+vmX3QzNZ0JBnW92pg3tPXJtXXkgTobhKAtqVt0QAawLzve1hK/T0EMO+J14nX0QAaQANoAA1EoQHMe4QYhRBJHpA8QANoAA2ggVAN9MC8397MnjazvzezzMyu9iSX2JwWAcx74m7ibjSABtAAGkADUWgA8z6tIJLSTo0A5j19dBR9dGhuhP3Io6EBNIAGuqsBzHuCEoISNIAG0AAaQANJaaDD5r2Wx7/FzL71Q9Nexr3+rphauooTtUkA856+Nqm+liRAd5MAtC1tiwbQAOZ9myEfx06YAOY98TrxOhpAA2gADaCBKDSAeY8QoxAiyQOSB2gADaABNBCqgY6a9+ea2e8Pmfa5eb8q4QQYRf8fBDDvibuJu9EAGkADaAANRKEBzPv/EaDxDgIFApj39NFR9NGhuRH2I4+GBtAAGuiuBjDvCUoIStAAGkADaAANJKWBjpn3R5rZFz2mfW7eX1hIKPE2XQKY9/S1SfW1JAG6mwSgbWlbNIAGMO/TDSgpeasEMO+J14nX0QAaQANoAA1EoQHMe4QYhRBJHpA8QANoAA2ggVANdMS839vMPmRm3ysx7mXga1Y+/9IngHlP3E3cjQbQABpAA2ggCg1g3qcfWFKDVghg3tNHR9FHh+ZG2I88GhpAA2iguxrAvCcoIShBA2gADaABNJCUBhI377czs6fN7O8CTPt85v2ZraSmOOi0CWDe09cm1deSBOhuEoC2pW3RABrAvJ92GMj5EiGAeU+8TryOBtAAGkADaCAKDWDeI8QohEjygOQBGkADaAANhGogUfN+SzNbY2Z/WsG0z837lYkkuyjmeAKY98TdScTd9z3wRnbl6idH/h55/ONJlD/0WlLc7+Zb3zFSXzF47sUvdrbOxfrzvtkY7MGHP+TU00PrPoKeuA5EowHM+/FBG5/2lgDmPf10NP008Vmz8Rk84YkG0EBqGsC8JyghKEEDaAANoAE0kJQGEjTvtez9f6hh2ufm/Yreps+6VXHMe/raJPraU0+7Ju97Nnu97IpHkih/nRvyAw5YvFld8/76gYd/prN1rsOJ74QlvM45b61TT+dfcBd64joQjQYw77sVZFKbxghg3tNPR9NPE3eFxV1wghMaQANd1QDmPUEJQQkaQANoAA2ggaQ0kJB5f4SZfTo3gSZ4XdpYOooDzZJA78379Rs+k91w06ao/+6658eT6g/buEnFvLe3jVfMexJBdX5jmPfopo5upv0dzPtZhoScO2ICmPfkRnp/LzDt6xHnI25CA2gADbg1gHlPUEJQggbQABpAA2ggKQ0kYN7vambvM7PvTmDYv20emdlxESe4KFo4gd6b99fdsLGo6yjfH3Lo8Un1h23c5GLeY963oas+HRPz3p186pMGUqgr5n14AMeevSKAeU9upPf3AilcwygjsRYaQAN90ADmPUEJQQka6KwGXnrlK9mT6z818rfp1a9UqvNT6z89coxYnoH6ymu/OVI21fnFTV+uVMc+XPBcdYRfmsFexOb9PDNTwufvGzLtc3NzSa9SZt2tLOY95n0S12bMe8x7V8zEtvCYqU/m/X0PvJGdceaNI3833fJKEv1dn3WNed/dgJOaTUQA854cKdcvNIAG0AAaQANRaADzHiFGIcQ+3zRT9/BEWFVWN9/6jtz42uz1ltter6T7LbfcarPvy5Q74cSLKh2jatlD91/32MdGyqbyrbrs4SjKF1qPWe336OP/zsnv4lUPwC/i62OE5v0WZrbGzP6kYdM+16eW3+df+gQw7zHvk7i2YN5j3s8qLuvKeftk3l92xaN5rLLZ67LllyXR33VFc3XqgXmffmBJDVohgHkfcR6gTl/Hd9rLucIWtmgADaCBdjXQC/N+46YvZ7fd/u7snHNvy45afFq2z74Ls+233zmbO3ebwQ3mVlvNyebPX5DtvseB2QEHLslOXHpxdtXqpzKZYu9459e56SRwQwOJagDzvt0LSBcu0Jj3aWokMvP+ZDP73ZZM+zwRflgrqSkOOm0CmPeY90nElJj3mPddiPFmWQfMe8sw7+OPsTHvpx0Gcr5ECGDeJ5r/m+V1n3PHf82jjWgjNIAGUtRAZ837V1//rez6G1/Kjjzq1GzOnLl58rvy67bz5mfLV1ye3f/gB5NItqUoQspM59mWBjDv0VaZtjDv09RIJOb9IjP7dMumfR63HJRIsotijieAeY95n8T9BOY95n1Z/MTn4+MnzHvM+xR+I5j344M2Pu0tAcx7zPsk4vUUrjOUcXy8OCs+Jy69KJs3b4eRv1vXvhPt0/+hgcg00DnzXjPlr1z9ZLZgwV55wrux1z32PChbvWY9s/EjE/GsLnacN84gpNgumPfxt1GxvVzvL7/ysZFnaF50yf2NBROY92lqZMbm/S5m9rqZ/euUjHvFMfv1Nn3WrYr33rx/fuMvZw8+/KGJ/tbe8R5vbH/S8ksnOrbK9viTP9/YNcZ1XUth2/oNn3FyfPaFL3SWzWNPfMJZ502vfqWzdU5Bi6mWEfMe8z4F7WLedyvIpDaNEcC8J+dL7IcGOq2Bo485y3k/fdMtr3S63inEZpQxzRx1m+3WKfNey9zvu9/hzg6oyQS7znHP/R+gQ+NijgYi1wDmffoXvQMOWDzSp++6676N/fYw79PUyIzM+63NbN0P/v6myZgi8Fh7NZaO4kCzJNB7876Jm5pnnvufRq4L+e9IhlkT5+AYaV4baDfaLRYNYN5j3seixXHlwLyfZUjIuSMmgHkfeZ5vXL/W5GennLo622WXfaL+u/qaDdz7oNfKGsC8556pyb6SY7Wrp86Y97pg5c+wzxN4bb5uscUW2Zprn63cQSLodgUNX/gWNYB5n74eMO/Tb8Pib7Kp91M277cwszVm9sdtxhUlx94t4gQXRQsngHnfQHIF857rQlPXEo6DltrSAOY95n1b2mryuJj34QEce/aKAOZ9A/F6k33VrI51zLHneAcMl9y7T+17qy57GF/CoVetyvzCS78y8vfipi/D613fyPpk3msVNZcWXn7HV9GC47czq/6W8/rvyzth3l9w0d1BF8YFu+ydLV9xRXb1Nc9k997/k9nTG37p7R/whuc/n2kG5i23vZ6df8Fd2aJFy0oHA1y86gF+6PzQ0UDEGsC893f+qVwYMe/Tb8M2tDZF8365mf1OBDfnO/cqZdbdymLeNxAzYN5zXWjjusIx0VWTGsC8x7xvUk9tHQvzvrsBJzWbiADmfQPxelv91jSPi3mfbmz49DO/6PSJdtxxN3L4PTPvTz5ltVMLl1z6EFqgr09CA8mb9xdefK/zR5gn2rfYYstsydFnZvc98Eb2jnf+dqVG2bjpy9maazd4l+LHvE/3Qj7NgI9zzU4nTZn3MbehHheS93fF166MwG3bvI+5bSmbv++Ygnl/gJl93MzeLP6uZvh+/kQpKL4cCwHM+wZuEDHv/X0j1w3YoIE4NIB5j3mfwm8R8z6W8JByREYA876BeD2FPrCsjJj3ccRUZe3k+hzzfnzb9WnmPeb9eC24fj9si4tZ0ub9DTdtyrR8vS+ZvueeB2f3PfhTlQx7n0DX3vneTMcrngvzPi4x+9qO7f1tJ8z79Nse8z79NmyjD27RvF9gZq+b2XeK1/sI3m8bWVKL4tQjgHnfQDIwVvP+ldd+M9v06m80ct9R7De17ONLr/x648ctniP29xtf/rVe1f/V138r0xKPMbWLyqSB7TGVaZpl0W/wlde+Flz/WZr3026ny654dLMcSR4zLVt+WTCvttpy2izaqkdbx8W8rxfM8a3OE8C8byBeb6vfmuZxMe/TzUVh3o9vO8x7y5h5P14j0+xrOdf4tkjWvH9y/aeyrbee57xR1A3j4iVnZC+90mzSQ0kLLdG/1VZzBufFvB8vLn588Jm1BjDv09cg5n36bdhGP9CCeT/XzO42s7/Mk86Rvc7pfJqsHxXEvG8gGdi0eS/T/cGHPzTyt+7Rj3qNJz1qS6t/HX7EyZkey7Xlllu+fU8yd+622Z57HZI98fQnvd8f7hdfff3rg/Nfevkj2fIVl2cHH3JctsMOu2Rbb73t28fV/cf8+Quy3Xc/YPB4rzPPujm78+73VTIUh8/r+//6DZ8Z4SFGz77wheA66dgvvvSrzuOMY3P/gx/Mzjjzxuygg4/Ntttux0L952Za6nL/A47KzjrnluzewapqX69UHl99tf2xJz7hLGtVE/2ZZz/rPM6G5z7nLKvuL2+7/d2DR7vtu9/hmfSTX3/0fued98wWLjwpu/iS+7NHn/iE8xjj6lXnMw0YWXvHe7KTT7kq22ffRZvpUGWTDg859PhMGvQNlNeKd67f1UPrPjKVOtSpd/4dPRfziqueGOQTdtlln0wr+eVtstVWc7M99jwoO/a4c7Krr9kweARf/r3ia9vmvX5DehTgytOvzRYdvjxbsGCvbNttt3+7nJrgoN/Prrvumx18yLGZZh3dePPLjQzAkA6Lbavfa86n+Lp4yemb7Vf8juv9w4/U04byQlot8bTTr8sOP2LFoE/edt78zco0b94OAxbqV1acfGV2/Y0vDfqnYpv18T3mfT+CT2pZmQDmfQPxehf61FWXPjR4NriMzrp/iqOK18bie8V3dY+bf++2298VfVw1Cy1g3o/PI0o/RS3m72+65ZXO6YmZ9+O1MIvfJ+es1iZJmvdKBihhkHcuw69Ljj4jUyKkLTHc/9BPZzvutHuGeV9NbG21B8elHXwawLxPXxuY9+m3oe/3Ocn2hs37S83svw3HEpH9f4vKaSe+ECMBzPsGkoFNm/dVkjv33P+B7NDDTvDegxT7jYfWfXjsvcjzG385u+rqp7Mjjjw526ZguhWPEfJeRqpMVB1vkn61+N1TT7vGWcfLrnik0jluvvU153E06KF4Pr2/8+4fGxjzIXXO99ljjwOzW257beRYw8cO+b8r3tB5Hnj4ZyodX22Rl6/4eu75d2x2HN3Prr766YHRWNyv7L309/CjP7vZsULqF7KPTHsZ0hqQUlaO4ucadKABCMVzaJZ6cZ/8/dy522y2X/E7s36v39CpK9dsNoAiL7fvVZMJVp52zYgR3LR5L5NVv6fjTzh/kIfwlads+5w5W2cnnHjhYLBKXd777LvQ2bZl5y77XAZ7SJk04OrWte/MTjjxomynnXavXZY5c+Zmxx1/fvbI4x8POm9I2VLbB/M+xlCRMkVAAPO+gXg9tf6wrfJefuVj3utU3UFrbZW1S8etcn/XpXqH1gXznpn3oVphv9nn5JM07zUSyHfzt/c+C1tZsnJYrJpVoaX0h7fz/9mLmjagDXINYN6nrwVXMl2ziPI25jX9Nq7Thg2Z9yeZ2W/74omItn83giQWRWiGAOZ9A8nAWZj3Mjg0I37c47qG+wyf6SvjSYZ9vpLX8Pfq/l8zTddc+2wj18dpmvdie9LySyuxHWak5bmrLGnuuu644g2dx9eOrmNoW4h5r1n4Wl1huB6h/5d2lAz2laHOdt3bjhscH1K2E5delKk9df7UzPvb7/qRbP78XWq3yQ477prddc+Pv90mTZr3Z519czY8kzykPcbts+WWW2Wnn3F9pv6oql5mad6fe97tmUz+cXWr+plWTlGf9/I7vlqZRVV2se2Ped9McMdROkcA876BeD22/m5W5cG8n02+CvN+PHfMe8z7WfWJnHf8b9PFJznzXjMCtFSd66ZMy9hNazlBF8ymtj21/tPZdTdszHRzuuLkK7KlJ10yWN7t/AvuqpxAUpn0/M27731/pufRaWk9JbiUXNHrmWfdlF25+snBcdtcrSBno1kmWuZPy/apPkoaFMuj8inZoVkfd93zvsy3xGR+vDqvepyClni84qrHs7PPuTU7ZeXV2YlLLx6UY/mKKwZLdV6y6sHsxptfGSz118cb+TpcXd9R4u6xJ38uu+mWVwfLy2pZRbW3NK2ZLZqFpFlc99z3/lYSJqmY91ouV5rUcrla0ifXo2byXH7l49l9D7yRaR8X43WPfczZH6667GHn/q5jjNv24qYvZ1rCVktznnPubYPZRUuXrcpOWrYqO/3MG7ILLronu+a65wbtPO44dT9zJdNTNO+17O7d9/7EQO9q12K/p0T/VaufGvQ3vnauy6/q95TEVfJZfaD6Rv1WtWyu+mX1mfo9Vz1mG/tPaN7vb2YfN7M3XbFEhNtk3t/7w2X9tbS//m4yszWFv8vN7NzC39lmduLQ38If1PuQwt9eZrag8De/c6m3+CqEed9AMnDa5r2eda1lp6v2DQ889NPO/lJxaNVjVdn/lFNXZ7pfmqTvnZZ5r+dRaxnrKvXz7atlzCeptyve0LmaNu91HdVS+L56VNne1LMa9RiISWYwF8usQQmKeVIy71evWb/Zoy+K9anyXrkI3T/qt9ekeb/ffkc0ohdXXfT70+MtqvQXszTvDzromNZYqA9ocgWTKkxntS/mfXyBIiWKggDmfQPx+qz6tdjOi3lf3ahqog0x78dzx7zHvG/id8Yxxv/OmuKTnHl/+53v9d6wyfhtCkxTx5EJr6UEh/8eeexjm5VVSQ6Z63vuebC3frrhlukTWjaxWrzkjEzL47lu1oe3aUS/jJo2nkUog1Fl1/M7h89b9n8luI459uzsmuuez2Qkhta/uJ8SSNfd8OLg2aSaaVB2zuLnmt2y//5HDp6jJ4NVAxCKx277vQYyDOtH/1eiqelz65iuc6kMoedSsvq8C+4cPJ+z+KzWIlPXe+lUyXEZhEpsh55v3H5Nmfdq/2EuSsqNO3fIZ0qWnX3ubUEzfbbffqfBwJJnX/j8Zudt2rxXX6TEowYR6Jm9VWYZ6vet5SeLM49COBT30XM7i6xdGtIyl8V9yt7ruZbFcxTfa7CU6/vqb4r71XmvvkLPIDtq8WmZyuzS/fA2zSbSDMSmlsTVuYfrp2VGh+vz1DO/ODDpi89LHS5b/n89/1Xmfp3ZWsPnrfv/mub9zmb2upn9S14XXi1Ilz9k9m0zy/++ZWZ/XPj7QzP7ZuHvd8zsK0N/XzSzTxf+NIDig4W/9/2wfdRG+tv4g/M+PfS3bmgQwy2FAQwazKBHIBQHMawcGsCgAQ2LCwMYNJhhn8IABg1m2LrFtCHmfQPJwGma9xpQpSXe6/QVesyWq49r27xXWfXsZ9e5Q7dNw7zX4NhJZ3oPt4sGPIbWcXi/aZj36zd8pjHjXnXXjGHfIJHh+vn+//SGX8p23HG3YI2HrBihZ537TMHYls3XvXqVWLfsPlL3M/c+8EYy5r10dOBBR7+9YoJPJ8XtXTXvxUL3fJr4UKxvl9/7fqdDfatiF/5BoE8EMO8biNe73HdWqRvm/XReeHGDAAAgAElEQVTMreE2wbwfzx3zHvN++DfD/8f/ZmbJJznzfsnRZzqTC7pRbmOW9qSNc+ll65zl1UzW/Ngy2Rcs2Mu539CN02BJt/x7vlc9N2f/A44KOt7w8fV/JTBk+q9/9rNvl9F3rrLtGr3+1lKYW9YuT7GMMsCeXP+pSuWSkahnYhaPM8l7JaTK6t3k5088/QvOpNLeex/WeDl0zGE20oPKMK5Oz734pUxLOsrUG/5+nf/LBF5zbfiAAV/ZmjLvXYk6lwHqK4dr+w03bcq2226nyrxkrhYHbjRl3us5vsced27mMsvrtOEhh55Qa5Z2lQRyaLnGPaNXs81cx7l41QNjNe9q0+K2Bx/+UDbJTCn97hRQy6QqHrfq+1133W+kfro+5MfRAIOLLrk/UzLdxWHcNq2AUHVGYn7eSV8rmvdzf2j4/sW4+vBZsJFfWSuJsv1eYbBCPmjhTwoDFjR44T8WBixo8MLvDg1Y+LXCYAUNXPjNMhZKpE/6++j696dp3l948b1OvW+//c6D50avuvSh7Nrrnx887/v6GzdmF1x0d6ZBaIpXNeDT1RY+836vvQ4ZxMwazKsVszRDe8Pznx8YSS+89CuZBpspplWfrdnNZYajVj1ynT9k2zTM+xUnXznCVtwWLlqWaUa5Br/pGqM4R6txibVii3G/oW222S6TGR1Sx+F92jbvNVjTdX+mwdMapKwVeLR0+7pHPzoYSL32jvcMDGANpBxXZ+mm7mA69Tf77Lto7PF3233/QTkU1yjeFzd9T5y1ktoxx57jHKCo9nKVOybzXvfM4wa56zcmPeqeRPdCuan77AtfyDQ4RyuKuVYs2HnBXoNB6676a+W5Ye2V/d8VT+q8Wm1CfZRyCXpuu+7dN778a4M/DRTR70cDWqSvskGk+s2XlSP/XIO6dd7876jFK51tLSM83yfkNWTVMNfMe907SIdaDUy/mxEWz342e2jdhwcDxHWvM67N1WZa+TCva9dfMe/75EdT1woEMO8x7xu7DmDez8YQw7wfzx3zHvO+6zFul+qXlHmvG+a5c7d13hzq5jXGhikz7/W5Zk24bu5d27TU+Lh6KjAImRHhOvbwNs0AXXvne8eeb1xZlGRcsMvewXUbPr/v/1UejSDTriy56TuPb/uhh51Ym8k4XuM+8y3XqqTTuO9V+UzHctVZ5y47zp13/5jzu67jVdl2/AkXTPQM0xjNe5mlGuhQhYNrX61koWM1Zd6Ltes8k2zTQICbbnlr+dAyDeWfd8G8V/LRNeCjDksN8NDvK+dT9XWcea9kv5KddcqVf2ebbbcfmExVyzXp/hXM+yvN7L/k5eUVgz52DWDej080qO+Ylnmve47hgU0y5bSaS9ny7Crjcy9+0dl3F817DZqUGV82SNLVZz72xCeyRYuWeftwXU9zs9H1/XHb2jbvNfihGJ/r/QknXhhkvN959/vGzl7XoIBxdfN91rZ5rzoX+x/pS4a+Hh3gK5O2S2urr356RIvFY+nxReOO4ftM5nPxOMX30r4M15DHqmmgyfDjD3z37LGY9+prxw3u1sB6DRrxscu36zemmHyLLTa/n/fVfxLzfocddx08skqTAHQPkJch5FUDL/S7Hi5n3ubaXve+UgOO8uMUX/WIqJCyVdknN+/nz98lO+2M6wePFivrj4eP//zGLw2+68vBqD/yrZwyfKzU/495X8HOZdc+EcC8x7xv7Po1C/Neq2nq0aSKHxVr5o/P1asez6u4UXnEqtfPSa55ipd0bdVjezXYTjGoYnat+Kn3KpceI3rLba8PBrLWvYfJy9h1814rmClOFTPxW77i8kE761GnWlVK94k5C9frNMx7xapa4VMDKzWoV4/NVeymxxXr0aG6B77q6qcHg1A1MNZVzia2SWPF+DR/39Tjx0LKKBbSpDw25YvFQo+RVrvpcdY5C30+vNpuyPHZpzyHlDKjpMx7zcDIf2TDrzLpYmyIcea9niE9XI+y/+s5lr56+p6tV3bMcZ/LgNIMYd85fduVsNxp5z0q129cWfLPQs17XQTy7zT5Ogvz3qeVKrMkfG2Vb/clanXufB/fa1vmvdpNq23UDSpjNO91YW5Kj7rgx2zeq55KjFXpn1M37xWENtW++XE0IKsKw+Lv1Gfe6zelmVj5OSZ51WoQzzSwUkux3GXvA8z7D5jZb01SL76L0T8LDWDel994Tcu8H27/I448pbYZXuzTZN5rYORtt7+7sgFXPI7eKxGg2f7DZc3/HzKTdfiY+r8vJhy3ko3rODff+pq3bHkZ9Zzwqo+q0SpcvnhBg4+V0HKVZ9y2ts37vL56nT9/QeXl7nXT7hugrccPjKub67PHn/pkJvbFcuXvZbBXfQSSHjGhldvyY/heYzHvlSTzlVGPsav67HOtwFEckOI7dh3zXgMtlexu4vqg1R18AwvUfi6tlG2bpnmvwcYyQiZN6qtO0vjWW89z6uCII0+uxaKMVWyfY973yY+mrhUIYN5j3jd2DZiWea/BoFptR6tz+eLF4dhku+12zJYuW1U5Jg29lmnlsCuuemLweJ6qk1tUBz3+UQb/rWvfWRoDaVWgwxYuffsvH+w3XGetRFTcL+S9VvQJrXPb+2kloeNPON8byxXrq5Vxzzv/DmdM24Z5rxzjfQ+88dZKdIuWZVoRrViesvdaTUyDTWRyT8JRj7wrtqvPm9LqYsX9yt6fsvLq4HKJhQbbaiD04UesyDTpqaz+xc91L3LWObcMVt6bhAXfLc8tpcAoKfP+9DNvcIpdF4GyWQuzagyfea9lr13L1+niqeXcNOpMF3kZ0EouHHf8eZmeea0RQ6666IJY/KEPvx88R3nZqkxLaGq02xNPfzK79/6fHCTLlhx9xtjl43TR1PKdrvP6tp249CJveZRs0wVYy4vqwqOOWRd1jS5SIknLjK659tnBKCQlN4cv8iHmvRK74zpHHVedqEY15cuSqgxaAvKRxz6W6YehUVgnLr040yyrIs9ZmPdaAtG1nLmSf6+89rVKbeNqMx1DxyrWU+91Tp3b9Z3iNpd5r+TcoYedMBj9p5Gdmiml9lbCVW2odlayVgNSXEtOFsty0cX3lZahWJ78fWzmvepbrNfw+91222/w29coWXHSb0LPPtfIST27fDjJqiShr4+pmrB3zbxXu2jwhBKNt9z22qDP0O9DZVO58mVs3+pD3Angt3Q0L/hxF+dfePdg5KhGj+rPlZzXErP55yGvCiBzTQy/Nrlsvka9Drdp8f/q3zWycrQffm7Qvq5rQv59tf24egzXK/+/z7zXNSY/dv6qPlOJW2nnxptfGfSDMl7UFy5ceJJ3tpa+f+RRK72M87I0+Rpg3r+Z14tXTPiUNNCEOdPkby3GY83CvJeRU3dp8mkw1LXQpXPd+Nc5/zTNey29XqeMupa66qxtWh2h6jGnZd4rkVU2G8ZXdl87Kx6smugad69Wd8Cg+i9fsjRvqxjMew2gUUyWl6n4qriz7qPjLh4zICA/Rx3z3qeHuts1IcI10ED33HpUR9XjTtO8r1q2sv11z+9ajUDb9OiBsu+n/jnmfQU7l137RADzHvO+sf6/bfNeg9k021oTKvJYo86rDM86K4G5roMa0Kmc0ricfNUyqn7jJhbK9K16zND95Z246jnNbS9u+nKmR7e64reyeih/qsdzFVdtatK8l0mte0dX7rasbK7PFY9q9ab8cV1VOZc9Hsl1zpBtBx64pFQHemSVVrgo8zhCzqd9NBlOA2zaXJmgKl/2n/6AgKTMe42CcQlcz4GLVTw+Y01GfLEuby3DuTHTRc5XFyUNXSO+ZKL5TB917DLKX3rl173H1fk0a1KjgYplKr7XM8i1xJyvbMXtuuC7LigaBKCBCFUT05r5IPM3TwaFmPdavrBY/vy9lkes84xmMdaoJy15OQvzXny1nGhej+KrlhUq8q/zXscoHjN/r3OGHC8379XGmh2iRKwSYyHf1T7SvRKtWg4yP3fxVYm+Os8wjcm8V1LVN/JQwYWesT7u9y9O0r6eH1lk4xrUoc/rmvc77rT7YNke/U6KwV1ZW2q1DdeyoXlZdTNQdgzX565kup617tq3zramzHs9C1b6z+tbfFUCUI880fJl48oojcskL363+F5to6B93DGGP3OZ91qKv7hU6NtL95YM1JEmZAQVy1R873u+83CZmvh/gHnvLWexzLzH2I9NA1VjpCZ+T6kdY9rmvWL22G+YNdNcsytcetYg1aptPC3zXvcoVctW3N/1LHAxqDIzIj+eK97QsareN5StwFN3iXuVU4PVfbOmr73+hWCWitmGB4Tm2tGgzZxJnVfFVb54SOeIwbwfl0jXgNk69dZ3FDfvf8BRzt9hzjcG815l1SMH8zIVXzXDv2r9UzbvVVffQBYNyq3KIrX9Me/75EdT1woEMO8x7xvr/8fFHHUfV5Nfa5R/8cX/xWt76HvFmFXiybwcxdcXX/rVwUz70HNW2W/cpK4um/ca9NtEOyveySf/NWne+3LSVdrWta8GA9w7ZiJWUXfF97M0732Dg131q7JNnlzVSbVFJryfvuHeJPOkzHvXzGCJXbOjm4TS5LF85n3xRyrTvMzU8ZVJSYIDDlzivPmWYVR1GUqNECqWrfj+pGWrgjj7liHU83Z89QjdrgRa2SACMRmeLa96aORT2XfLyqHk6N33/sTE9Sg7j+tzzXIutkf+XrOeXftX2aZj5McrvoY861Hn0QVVidI6BnuxnBqoMWxO5+Wpk+CNybz3LVGuwKLKRVjB1riBNjmvqua9luDXb3TSmYVvLce5jVNPdZ4f6Uqmx2bea0kkn4GgfrjKDED1X75nMqltTz7lqkq/d5d5n2tEr1WX7lXCf/c9DnS2b+g1ovibr/s+wLz/uJn992JdeY9Rn4IGMO/Lb6ymbd5rdau6fdU0v+eb9bvm2g2Vyz8N817LVWvVq0kYKdZx/a414LfqcV3xho7dpHmv5T+rlmt4f5/pqmdIDu/r+7+WVXVx00yXSWN5nXNcHBODee9r6wMPOrrSwFUXX8W6Lrb5tljMe1856+RVUjfv1z36UWebaeVDVxt3aRvmfQU7l137RADzHvO+sf6/LfNe8b1vIGYec9R91cSiOtc6Gff77LvIeU2tW5bi9/po3mtQrCYRFjlM8l5L7ivnmIJ5r3pqQIlW8a2ixy6a92Kheq294z2VWFThxr7leahZMUrGvNdSMK7Z3BKwnvU4K4Bl5y0z7w859ITKM9GL5/TNmBaXOhdcdeK6UdX3h/80U1PL7RfP73p/5FGnjnxXy4Hr2K79m96mGc7DZdf/q5qZTZdr0uPJIFwwtIS/6qWAbZJBCfquK+jTuXTOSctd9ftK5LrqqWWSqj7fMBbzXkvN+/ovPT6iKiOtarD3PgudOs+1P0u933TLK86ynbT80sp1dSVYYzPvtcR8zn34tc6sHf3uNPNt+Fj6v2azVVkad5x5rwS6EpZV9afHOrj0rJUlJh38EVqWAPNeSZd5P8gybTSzf3GxZNvodR4ms2eCeV9+0zRN8179WtnqVaH9Vtv7+VaSWb7iisr9/DTM+6qD0Vz8fKabBgZUve9wxRvqE5s076sOqnbVWTOjXX11lZXBjlq80nmMRYcvr6wVVxm1YpmrjNo2a/Ney8K7YhiVbdLZZjkL30BknSMW816xpmtihFapy+sR+pq6ea96upY31Sy3UAap7od53yc/mrpWIIB5j3nfWP/fhnl/9TXPOB/5Uoy9NIlkr70PHSy1rhUyNVlH8b0GgfpWPC1+v87kO+X7iscYfi9PQZNeNANcz2KXGa9B0nqkrVau0j2LBuBqmffh7+r/48x7PX5TjwvO/3yPERCXfJ/QVy0LP4vrvFZ+c+XGi2wU02pCpx5HKwaXrHpw0M5a+dSV59d35RVNy7zX6qGafLb0pEsGK7VKh1rVWPeBevRmiBbVTlUe6aBZ6sW29a1UrXuS4n5l7xctWlaqg3Ez71VX3WtpoOzgN3ne7T9ksXrwGFWxKrat670eRVFnVb1Z6Jdzlue1QhklY97LNHYJV9uuuOrx0h9QKJCm9xtn3iux9NQPltKe5JyHHHq8k4s676pJq7wcMk99I7tCEm0uU0UX5/z4bb/qwurSStXkW9vlrHN8PcvIVTclTeocT9/xJVx0rrrHnPR7mqnsqufaO99bqUyxmPe+IFbP0K3LSktk+ZKPYjdL8151UlAy3IYKVqrW15VMj82817OPhuuq/x98yLG1+2E938kX+GmFlFCO48z7OgML8vMefMhxzjpPuvRbfvyyV9d1ZqgNlHTJ/x1mZr889Lmz/Owze/O6722AeV9+kzNN836a8WtZv1f2uQZPueKCOo8Xm4Z5r0culdWp7HPV2ZeY2ljyKJjhY7viDfVHVe8ffMvmK2GpZe+Hz1v1/4r/XP2kHmsTeiwltVzHmGRJ/+Fz77XXIc5zzNq8v/nW15zl0pKfTQ3UGXf/H4t5r/Zyrd6npHZVDr57ST2vdFgXsf5fEypcvwmtChdrmZsoF+Z9fqvAKwQ2I4B5j3nfWN/ftHkvE0crJbmuWdq28857DnKB4yZ4ySu474E3Bo+E9R1H8bUeDRl6rVG8rBjCdTw9gkyGcZVrqh4ZqlV9Dzr42Lfvb8aZ98Pl9E3q01Lsw/vG+H+1kfLFLp7apns+5ZmfWv9pb33EWwMjFHsXj6MZ3L7JYJqIVZVHcdl8PaJT98+Kt7VqZ8ixNNlNq3aNmzEv76uux+VbEeySSx8KKl9IHfJ9ijlcDQbQo4jlSzz7wueDzqXB+Hrkqu8xaWpH5Z9nMdEyryOv5fmqphklY94/MGYJuutueDHoR9A0vJDjjbt512ijkGP49tGyhr6LY9VE0/A5NPqt2Lnn79URl82sdJlFdZY8Hy5T6P9l8OblLb6qEww9Rqz7abSZMzG7/5G16+YywXSOKiPbmub16uu/NRgBV2w/vdcovSrnisG81+9FF+3huoixnl1UpT7D+/qWTtW5Zm3e33jzyyN1VrmqLsfqSqbHZN77fpNq3zqz2ottLHN9WDf6//z5u2T6jRT39b139cc6hmYY5c+78n133HbfNWL1mvVB5Rp37JDPXP3WEKuieZ9nYy5lKX3M+SGdOH9js9wH8778Zmia5v1VDTzyKaRPa2ofV7yhJF7V47dt3ivpqFWEqpbLtb/q5/rNrn/2s5WO74o3dNyq91Q+836ffRdWKo+rrtrm07+u677vFLf7kpmqa5WVfYrHdL33Jcpmbd6fdfbNTr1ohpKrHnW2Pbn+U85ziHFM5r1mPLl+O+MSwS4eXTDvfY83m/RezcUrpm2Y9/ktAq8Q2IwA5j3mfWMxQZPmvcxQ30Q75Z8Uv5fl64evQVetfso7GEAz90PzTnqMqium2H33A7L1Gz4zEU/NNNas/CqT1nzxbirm/bXXP+/kKcaaCFpl+XTFMrvtvr/3eMV2q2ve73/AUQOTOlQvwzrU/1VOPWKsWJ7i+1vXvrOWjnz3JG2Z9xo4r9VZJ8m3Pv7UJzMxLda/+F7HdzFkW3kuKUVGyZj3eiZ0UajF97fd/u5oResz73Vh1Y39JKLxPa+wiecpKqGmC0KRc/6+7Jnvrs5WCbFJ6lrlu74ZKbr4VTlOrPtqNm/eFsXXOoMTfEus6hyzrv9Ri08bqaeW26lSrhjM+3vu/8BIPdRuGi1XpS6uffV8+aIGiu9nbd5rBY9iefL3VZ7/rjq7kukxmffinNet+KrZTK42q7JNM/SGR8nm51A/F3Isn3l/3gV3Bn3fdw7NsM/LUnytsiqA79gh22ua98rKsJS+47E4xTbk/WwHOGDel99w+cxLafec89ZW7tt8yR0d76F1H6l8vJA+zLePZhTo/kDxi0y+FSdfMRixv3jJGdlhC5eW/rlmoWt5Pd/5fNvbNu/1OC3fuatu39Mzw/uRxz9e6RyueEMaaMq8P/6ECyqVx8dBs6Jd/bQe8eD7TnG779FGGvhR3G/S91pu1VXOWZv3Cxee5CyXTP1J65x/X79j33KtTZv3egyAEjpaJlVLpmpZUi2FunDRstL+wrVsfp1+LxbzXsvLakWPVZc+lMlEqMLCtxpFaLydt31qr5j3mxm2/AcCOQHMe8z7xmKCJs37444/3xnDyF+4+poNtct80y2vOieJKSYIPe7Ojke8vhVTfLh2uSa5pvru71Iw7zUAwzc4Wfd6t1dckVYcteqz6u6KzYvb6pj3ynlP0lbF7yoH6nv8lGbfF/cNfT9N875JFrrn863y2kS+OZQf+5Xnp9pmlIx5P86kavLH0TRwn3nfhMHue0aJblibqIeW9yh24vn7ssSoEoz5vsXXOheBOvXwJXV33+PARparrFOmJr+j5xsVuebvTz/zhsrtru/k3y++6hxNlrnOsfSIhmKZ9F4BYZVjxWDeK5k2XA/9f5Ily3MGr76u51W6lz6dtXmvMrqW+qn6OAZXMj0m89733FjdpOXtNMmrb3UFPc8q5Lg+837SZKSSfVoCeFjb01piegLzPk/KHGpmXxwuP/+frXENf8sw78tvjnxxnvRTFqO6+k1fckfHa2p2uOu8+TZdy2U4nbRslfdxKZP8NpTQ0zny84W8tm3eKykTUo6QfXyzErRqW8j3831c8Ya4N2Xey0zMzzXJq4xhtemwJnRNDjmub/C3njEa8v3Qfe73rJo3a/NezzEfZqf/X3/jxkbrr2VeXedpwrxf99jHBs+G9dXFdd4q2+66532VWMzSvNcgnbPPuTXzxbtV6u3at8rMttDfRkz7Yd7ntwW8QmAzApj3mPeVroPj+vWmzHtNnvCtvKvnx48rQ8hnvtyw8uhly5W//I6vOmOeJnyPkLK79vHd36Vg3mvAhCsm0TY9SsBV35Bt9z7whvMeoniuafk248qr1VpduWTpv84qDtM078fVq85nG577XKYB2sU20nvdC85yteQ6deE75XkuH6NkzHvN9h4Wa/7/OqOOfECa3u4z75evuLx2h5uXUbNWcgbF16aeOey7YGiJvbwMrtfLrnAv9awRYjKcppEI1dI8RSb5ez1/8a57fnxs+V11immbnuFZfKZMXjcFIVWSs9rXNfJOx676nNBQPpqNrQuxnl+k5yeN+9Nsibxu+WvV2WMxmPc+8/Whdc2MQHWtUCBebZj3SvDo2VmaFah+Zlz76TPXkl5Vk9euZHpM5r1vhHGdlTBcv6MrVz858jtQ+y45+sygfsyVzJwzZ27l5dRcZdt23vyRsi1ecnpQuVzHq7KtAfM+z9CksJT+75nZIYW/xT8YeHDi0N9KMzu38Kd6rSn83WJmdxf+1pnZ00N/G83s9cLf+8zsg4W/j5vZp4f+NADiK4W/3zGzbxb+/tDM/rjw9y0z+/bQ35t5H88r5n1IPzAt815Lu4eUZ5J99Oivtgy44u/pxZd+tVJd2jbvNSt4Em7F77qe2626yzwu7lf23hVv6DhNmfdVBy+OK6/LvFdCa9x38s8uuOjukWu36qnBiPk+TbwqoVTUYP5+1uZ98VmUeZn0qoRmE/XOj3Hc8ec56z+Jea8BKb4V2Ip1mfS9nlOa1yPkdRbmve6jDj3M/Zz6Setf/H7VFcNCeMW0D+Z9fjvAKwQ2I4B5j3lf6To4rl9vyrw/9rhznXHF3nsf1sjzr3U92GHHXZ3nKJsw6TPKVeZxbNr8zFemFMx738Bk+RxVH4swzFgrgRXjnOH3MZj3KvPpZ1zvLOc111VfUTll814szj73NicLrXI23L78v75BHjO7ZMz7+x/8oFOs6mhuuGlTtIL1mfd6XvAkwpDx6hp1J4N80s48L9fDj/6sk7kuGPk+rlc9h8dlLucXBS0juHTZqkG7Pb/xl8cey3X8kG0ajZafz/WqZ/dodKJ0VcXwDjn3NPbxrYpQZXaA9nWx0bGbqIOWeFFSWgNVtNSLjHfX+apse2v2WNizvlWHGMx710CSrbaa09jsSs0ydDFswrxX4lW/JT0Hco89DvQ+C8t1ft+2qsvGupLpsZj3GmHsSqBrlGhT/YoStS6WGogU8jt1mfehz8UtO75rKS8lUsu+18TnDZr3ytTEvpT+1zZLJ3X/P2qPBYW/fQoDFzSIIZbBC8MDFzSIoWzwwl+5fs/FbUqcNPEb6fIxpmXey+Rri6NiJN9KVUU9NPW+arzdtnnf5PPFUzHvz21gVlSuR1fsEWre+2ZWHX/C+Y3qXfeDLv3O0rzXzDHXqkEqp2az53ybeNWzWV31r2Peq9y6b9WAItcxm96m5XOrMJi2eX/hxfdOjcV1NzS7IkMVrtPYF/O++0E1NaxFAPMe877SdXBcf92Eea84fs6crZ0xQJUc8Lhy6jNdX10xxWlnXD+Wx+NP/rzze0uOPmPs98rKM8nnqZr3Wt7eFeerXZp4FPC44+scsZj3vhW8ZMRX1UXq5v26Rz/q/H2dtPzSyiyqsmP/OAYDJGPe+57Nrc7lqtVPRStYn3k/aYeo2a+ui+qeex7cGItXXvua86Kh2bRlP2AlGFzlG96mBIpGlZ1x5o2ZRvk/9+KXSo9ddm59rtn9oTOZ9IxHBRV63IAuEK++Hm4Ou8qiOuiHNcnfple/MpaDljMcZqn/a5a3q0yubb4Z4VWXShw+9mNP/lymEZauZW5cZa66TdofPqfv/zGY99ttt9NIW+22+/7BdfDVLd+uZT5dDOua90oQatCF7zlDrnNV2Va2ckder/w1ZvNeSxi56r7Pvgsba1/1Ba5z7LjT7kHncJn3GryU853k1XXsgw85rpFjl5WrYfM+z+rIGI5xKX2ZwvzrBoHTXL/n4jbM+/IbpGmZ923NzJBxP8k1VvGVYlfXny/Zg3lfritXvKHfZlMz72Mx71ec7DaVZTaXXXurfL7p1d9wxi+zNO/1HM1if1t8ryRvlfqV7esbJFHHvPe1WbH8vvca2O/qK7RNn7m+F7N5fzWMInEAACAASURBVOrKNc4yu+oxvK0OC8z7weOUFLvwDwJ9IoB5j3nfWEzQhHm/5tpnndc+TcooW9K+LF4pfv7E0590nqcsv+XzJ+bPX9DY5MJiOUPep2rer7nWvWS+JkjqHjKk7mX7jFvFaVKvquzcoZ+/451fd8ap++13RGUGqZv3YuZaOj90Qlcoc/YrzxfMilEy5v24m+2LJ3jmR9vgfea9nm05ybl9SxE2/bxC12xpjfovCxDU0coQH75pLvu/Zo1o2R/N+NHyhWXnGcdQMyh8SyOOK4c6Rc0IUttteP7zldvpltter1zv4fIoaBpXN/FdsGCvkfMoIfbipi+P/a6Oq3207/B5dUwde9y5fZ9pxQcNwvAlgobPVff/Mkx9ZRjePmvzXvp1zZKR8Thc1rr/X3vne0faUWzrmPePPfGJ1pfjXHT48kp1dyXTY5l5r8c/uHTc9Oxz1yhrmTchmnEZ7BowFfLdsn1cx07cvM+TUlpyXku9O9t3Btu/nBeM1+QJYN43kAxM3bw/+pizxvYtir2POPKUTMubK6bUSljPvvCFoPhMSTJXH4V5X34z7oo3xLJr5v1pp1/n1MgJJ17USGyQxw7SnEuLszTv9fgIV5m0TY/1ysvexOuZZ93kPFdV8943qz2vh1bzOujgYzOd78abX87ue/CnMvWRISvx6RFM+XGKr7Ga91dd/bSzvHnZdc+lXIgGTmhlxvseeGPAImRQnFaeyI9TfMW8x7xPPvKkAnUIYN43EK83cS3twjGaMO99E6+qPpIyhKcGBBSvg3qvHO+4a6lyyK5JS/ruytOvbTS+CqmD9knVvF9x8pUj/MVR94+hdS/bb5wmYzHvVQeXFrWSc1n9hj/vgnnvyr0qTzxcV/5ffs+fIqNkzHvB1Y90+CKi/6tzixW+z7zXjfUkZVYiycWiyWUoVT5XZ6nzhjwTXRdwPZtDSQVXWUO2afb8JasezDR7ow6vp9Z/eqLZTVoZQEzvvf8ng88/DfNeLPTsTBdDPSO7jJXvOdp1n8epQE6JZld5mt6WknmvkZGu+h962ImlbVTWhvnn6ktc56hq3usZ9q5n1LuOPcm2Lpn3GmDkYnHU4tMaa1+1s8+MGXcDlevDFeRh3gflcLR0u54J/4+uNp7yti8FlZidUiCAed9AMjBl8/7Wte9yXjfUp2jAqYy6SWZVaDatq3/CvC+/ke+Lee9bHU2PcchjhyZetSynS4uzNO91b+p67JzK+cjjH2+0/r7kaxXzXoPIfY+i0z2qlrFd/+xna5f7qMUrnW0Uo3mvle1ckwrUdmpTDfyfZADGMcee42SBeY95n0JwSRkbJ4B530C83kQs0YVjjDNKlYMLqaNvVVmtmBny/Sr7HLZwqfN6qBVWxx1n3OQ9ffb4U+MnqI07dp3PUjXvfStMyhepw8H1nYfWfdjZxoqp2jLvlbvUvYHiba14HLJS8c6OCYuK+apOOIzNvNcAW+lTK4yHsthtt/2cbaZHubramG3l9/4pMUrKvPc903BaM/3qNKzPvNez1uscL//OPfd/wPnDbTrx4jJ+1KFrBlBelrJXzSLXc9RdM711rJA/LV0qU7zsXK7PNftZN95a6ifkXL59jjzq1KB6T8u81+oLruVR9Xx5F4fiNu0zXE8dS8cs7hfyXnwXLzl95HjF4287b/5gEMRZZ9+cXX7l49mta985uFg/+PCHMt/f0mWrnMdMybxXoqnIIX/f5CCbh9Z9xHmOKub9k+s/VbpKhYKFE5delCnpqGW7br/zvdk9973f235qV5fp3CXz/s67J398RchvzPVseWlJM8jKvu/qwzHvKyV49vvB89Y/nv92Z/T6uUolZueYCWDeN5AMTNm8d8Vf6le0oowGnJb16eM+H2dMYt6X38D3xby/9PJHnHFj06sGKSHtumbO0rzX70f3JK5yTTqwfvi36ZvJXcW8P/Osm51l1UwbJXGGz1n1/75lU2M0788973YnC00SuO32d03MYuGiZc7jY95j3sccVFK21ghg3jcQr1e9JnV1/0nNe61C7Mr7KpYJNf+rsNUjUJ1x0gNvjL3WKj/o+l6+TXVQrCkT+pHHPjb2WFXK69s3RfNeuXXfSraTrt5c5CQj3TfJsinzXvwvXvVAdsyxZ2d77HmQc0XaXBtVX1946Vcq6WfW5v36DZ/JLrn0oUwDRfWoax/7qhy0f53Voota4H15jiAGRkmZ9yefcpXzYqAR6SFLw80CeFvmvZaEd/2wFy48qVInVsbEZb7pvCEzPoePraXar77mmcFz2efP38VZfled8m262J93wZ0T1U/czjlv7WBJvTodpp4zrdFRw3Ur/n9a5r3O6Uq6iNO4UY36zBX86VjFeoS+v/b6571tqYE1a+94T+WRcTq3L2GVknmvfinXb/FVy1uG8i3b7+573+88RxXz3je6VoHjytOuyWTul5XD9fkOO4z+zrtk3muAQrFd8/dHHHlyLV4uhtrmSjSHjjjFvG8sqXOumf3nvI2n/PpLjdWCA82aAOZ9A8nAVM173yOv1J9reWffNSB0u2+ZcvVXmPflN+Z9Me+vvf4FZ+yie5xQrYXsp9lgrmvlrM1712PHVM6qhnUZA62y5ap/FfN+p533cB7jooYeGbjnXoc4j1+VhW9p/2XLL2tMU654VnwnzQ3k7ahnqLraC/Me837WgSPnnwkBzPsG4vW8f+3766TmvS/3r2uWcrjK5TX5p3jQdT3U6mFlbenLK7qOp/PokU16JI5mY8u4Ljt+lc9TNO/lmbhYaZvuf6vUv2zf3Xbf33muScz7V1//erbm2g2ZBqu7PAdf3apurzrgfRbmvQbVX3Pd8wPvSff6VesYur8efVvW1nxengeInVFS5r0v2SBR333vT0Qp2LbMey2R5/oxNzWjMheu61nLGiyRf173VRdmPS9aF+rjjj8vq2LmX33NhonPr3K/9MpXsrvueV927vl3DIKeUDNfSR/NqPbVXTM3Jg2eQi/MGgzh0sHZ59zqLZ8+c31Hx/LVadx2jaJzHU/PG6y6nE3xPMtXXOE8bkrmveqjZ5MP89l7n4W1WBf55O+1isHw8fX/UPNey/S4vi/DWMvC5+ep8+rqP7pk3mtQg4tdk4Mz1FdqadTh82hp5JA2cSU7m7pOuI49rZVwfMuJFTgp6dL0v63NbJ2Z/UPhPCNt08Jnv9B0RTjezAhg3jeQDEzVvPc9r7mpftN3PVefhHlfftPeF/NeCVLfdeq5F78YFFuExB+6D3CdZ9bmvWvgs8qpmd0h9QrdZ/vtd3LWP9S812BrFz/dh2969SsTl/XV138r23rrec5zxGbeazl8FwvdZ2hGYmib+PZTrK3HlrjOgXmPeT+ziJETz5IA5n0D8bqvz+3b9knNe02Gcl2fpr3tmuueK73eKo70LfFfVl49wlPPdb/iqieyJuLRFM17X5nFronYr/jb89331DXvtQrEXnsfOhWtVjWsp23ea8DNpCs/l/1e8s8ffvRnS3+XxXbnfXlOIEZGSZn3z2/8ktPEkGg1Kz9GwG2Z975naWt55aY4+Gbw7NTwzAyVVzfN6x796GAJHV9SJe+ctttup0wj0pqqZ34cMdVowlNXrslcM4bz8+tVz9XLvzfL140v/5rzWYjSgcs41zbXEtxKBOlYVeui5x4VueTvNdtEbVr1eMX9tbxOfrzia2rmvR75UCy/3ktfxbpO8v6q1U+NHF/nCDXvzzjzRuf3NUpwknIpuByut/7fJfPe10fK1J6EXfG7WgbJxVE3RcX9fO9dBjvm/cS5oGkvpf+JiUvMAWIhgHnfQDIwVfPet6JQU7Nodd13XS+0DfO+/Ebdl8R64OGfCbre5tdhXztrsHC+z6SvrtksmtURclzdC/ieHa4BoSHHCNnHdz83a/P+tNOvc/5Omnz03DOeQfb6LYaa92s9y88eceQpjbSR77FbKmNs5v1d9/y4s82aWnFQEwp8fSfmPeZ9LAEk5ZgqAcz7BuL1kFihD/tMat7fcNMm7zXKd+1qY/uVq58Mij9kvGt5/EnKoIl1Ws3yttvfHXROl458Rrjys679Y9gmI9bFbcstt2q8zL7HBdUx7xU3Kr53lb2NbTGb97qXkr/SRr1dx8S8L7/Hj+G3PWkZkjLvVVnNTnEJVoauZlJPCqTp77dl3quc22yznZNF1QSdr86+5zlrxqPvO01t10ilQw493lk/tb+4NnUu13Feee1rg1UBfKPwNdI/Fr2dcOKFTk5qv+G6+dpUxxjeN+T/eoaN6/eo0aEh3x+3zz77LnIeOzXz3pcMfvaFL0zMSPx8owhDzXvXkpkKaF2DP8a11/BnvsCzS+a9GLmCVCXQ6wyGGWao//tGWh940NFB+sG8bzXRc86UltL/aKu14ODTJIB530AyMFXzfumyVc64pilzaPGS053HV5xW9d5Ag1Rd8d1lVzwSdO3Jr2c33/qa8ziHH9Hc42UOOHCJ8xxaiSAvR8irL17rmnkvFr4l3bUaWgirsn00U9q3ROSszfvrb3zJqRfd8zX1GD7fymj6TYWa95rh5voNamWyMv4hn6+69CHn8XXOqua9Zsm5yqo+L6QsZfv42uzEpRc3cvwrrnrcWX7Vqan+uayOs/pcj0J0td3QNsUu/INAnwhg3jcQr8+qX4vtvJOa9754YKifDunLJ9on1LwXf+XJ9PikXXfdd6Jzqo5agl2Pq6zarima9/fc/wEnLw26rVr/sv19941VzXsNsNTjVsv0uO2222fKYR573LnZ6Wdcn1148b2Dv8uvfDyTtlx/rolwOk+s5r1WrnWtPjvMRu0pXR973DnZaYEsXJMwdVzMe8z7gr4GI269P0aNACjrGJr6fNyFSzehTZ2nqeO0ad77klV33v1jjXC4+JL7nW2+9KRLGjl+GWMt56fRdgUhvv2+6WdK+8qi1QBcy56rTLfdXv7MH99xm9xexZCvYvSHlHH5isvfbpO8nTQqUIMfQr7v20erIOg4+TGLr6mZ96esvNpZj9vv+pGJGOXsDjroGOfxQ817V3ChZ0/lx6/76kuEdcm8F5t99zvcyV+Bd112xe+df+HdzuOvODkseYt533qeae4UltL/UOu14ATTIoB530AyMFXz/qjFpzn781tue33i64WYjHv8E+Z9+Y19n8x7X2wxWImrgdXN9BzyYuxefD9r8/6pZ37RW7amVh4Y98zXUPP+Eo+5rpUDinFinfdaHW3PPQ/2cqhq3vsebXjc8edPXFbVz2d8aABznfoPf8c3YFy6xbxn5v20AkTOExUBzPsG4vXhvrav//ddw3SN0VLjZVx811gNktSjFKf1p2eZl5V1+HOZ+PInlIeeP3+BN+4oxomu93PmzM2UXxw+/rj/p2jeSw+u+rcx837RomXOc1Ux7zVhSSszu8qsbTvutPtg0KomZ9aZHOZ7RG+M5r0md+oRyz4WO+y46+CRzTLbX33965W0LJ37YlXM+/J7/HH9RCqfJTfzftOrv+Ht9DX7ftyzyGfRKG2a91re3dUxnHJqMzeyGgnkOn6VEXeTMtczpV1LQ6pTnPTYod/3La+owQ2hx2hzP10EXRcJJeBkgufn1nvX8i36bp0LqY7rGq2ni1J+zrqvvue4S4+pmffXXv+883ek329dPvn31N/5kvUh5r0eP+H6jZ919s0Tl23J0Wc4j13VvNfz44fLuHODv/9HH/93I8fX+bSqRM553OuKk690fl8jKMd9L/Sz/fY7wnn80JsnzPupJYLaXEr/p6ZWC07UNgHM+waSgama9zKyhq9n+v/qq5+e+Hqh67br2Pk2zPvyG/s+mfdPPP0Lzvsr6WVSc1grS207b75Xj7M27xV/uVadUt2PWrxy4t/iE09/0vuYP50j1Lz3PRbr+BMumLiMt3uW5M/7i6rmvQYg5d8tvjY12H7Ntc86j69HrP3/7N0J0H1nfR/2B7EIJIEwFovYhMQOAgEGSYAQ+y6wjFnEIiQ2sQowq0AIxI4dO64bYhLcUNtxQ4uX1KnruKWJm6QZTzptOtNJ0k7GbiZNOm4nzVJvjWPDbX8vOdLVfc9y77lneZ5zPprRvO//vfee5XO+73mfc773nrPveLrpeTe866u1067WQ3mvvB97YGj6WQoo7wcYrzftd9f282PL+/ibXP1N2v46xqexx9w2J7fK/dDPnVxFN+5tH+eNt9en6/voBfY9BxbrUWJ53/YG0+jChtw+TR8GPaS8b3uzbpwnPfaKVufe+361GcmxvI/b4DVl+GmXXrX5/Jf+xlHb77zzHlw7feV99zH+kL83c02ruPI+oNou8xZ/BKbAjEuMxaeyu+Y1Znkf92ev2znEO9riU+tdy9b2eJx4qCvNY35RqLe9dujH6i61c7e73WOyZWi6bPWVz3nDZMvQZfr8F1xfm4VXv/b2d0fG93V5idd2Tb/p8bp7GcU775qev+/PH/2Yy2uXNZa/tPI+TpDWucdgtc877rYN2w4E9invb/rkL9cu2/OOyEQs3003/0rjmwoOLe/j0rq7frGP23Y45vtjy/umg6l4l2nfN8VU69N0D87YN8f9VKvntX1V3k9+Qigupf+PdjN75L+/OvlamOFYAsr7AU4GllreN12K/tgrWsXfsbqr6Gzvd5T33Qf2ayrvY9zQNNaON4Xuc4zbNPaIcnk7e7vf51Dev/il72xcxhve9ef2Gl81rX+8AWB3nbf/vW95f+11X6qdTozrmua9z88/98Vvb+5734fWTrtazkPL+3e/92u107vv/S44almr9Yl73lbLtv01TupWz+nzNc7nNL2Ro5qP8l55P9aA0HSzFlDeDzBe77NfXuJr2s7Z7fPJ+6ZLqcffqVtu/bWj/g7O7f3BD/3cSb8Tb/aLc/zV396mr/EJ/Ci491nuEsv7z3z2rzcaHHIefB+fGKPVOe9b3se5zqZyfajbJtV9+DCWObfyPt6YUtdbxbI++SnDXIUqPqxct72U993H+Pv8PuT+nCLL+3jHStMvRoT5xS+9Ya+ded+N86nP/BebCy+6ZK9PZo5Z3scBZ1wip+4XON6t33f94nVxIrFuulPc7353uR/4oEeeWpbYce0+b6x/N52QOLbgHHJ5mz5Bc9HDn3ybU3y/u02jBIzX9l2Wuk9Xx/1uIpt9pxkD2Kb7ZMbyHzJoaTrxdeglausu4X/IpeXPf+DpDMe6xCWw+jrFrQmaBlwx7X3K+7gaw24m4t8/8NTjLpvfVFDEtA8t7594yfNPLeMQt2ao3I8t7+MdsPGu5zrH+KRQNZ8+X2OgVzfduFXCvtNT3s9yYujMlNKnUkp/WLf9evzsp2ZZCzMdQ0B5P8DJwFLL+7inX93vf4zlP/3ZX997v769/4+/QU1XaNmel/K++8B+beV9nAxueqN2vAGxz5u1n//Ct9RmfDuLOZT3cbK76bZoD3jARXe4ctn271vX903HHdvrv2953zQ+jWkd8waDuO3S9vLUfX9oeX/r536jNktxPBdXYuhy63q86Y3QsexxFYGu1zc93na8Urko75X3YwwGTTN7AeX9AOP1pn3v2n5+bHnf9IGb+Dv13vf/TO+/gblthzi/GeOoug+IVX+T4+u+b3ousbyPQrzpyqpxb/mhtll8Ir7pPvX7lvc3fuAv1Y4n400YMS48dlnjMvTb2337+9zK+w9++OdrlzXeXH/oMXidW2yvM844o3YeyvvuY/w609J+VmR5H8hx0NpW8MU76sfYGHGf6uryLvtcVnnM8j7WLy6Rv70Tq74/++x79z4R+J4bv95oO+Ul82P94pPJZ511r1Pr+IDzHz7K9q3LzGuv+dSp+Yfz1BZ1y7b9s3hDSbX9q69xUi7emRj/152gi9dsT+PQ75tOAMXVCg6dVjw/Sv/YttXy130tsbxv2g/EOxXjEzB9rKKcr/OpfrZPeR/zrTt5GVdP6HtVgHe956cbBxaxbIeW91c++/W16xmfSu/jtvuappOj++zfq2lddvkP1i5j/K3oO3BtO6l+yOXClPeznhx6UErp56vfySO+/visa2HmQwoo7wc4GVhqed9WQB3yhsDqb0/cZ/DCi55U+/dnd39z6ImDplLrlVd/8KC/vdde9+Xa5Yur6lTrcezXpss+xvHMIdNeW3kfNpc86fQbJKvs3Pve99/Ece8+hvFm0Lj/ePXa+Np0nJ5DeR/rFFdQ217e7e/jPqCHXuozxr91Y+rt6cb3+5b3sYxN9xGNNxjE7/8+26Z6Tnwq6IpnvbZxnbeX89DyPubR9MGKY67wVi172/Tvd78LNnEbsO3n7vP9c577pr0slPfK+yEHgaZVjIDyfoDx+j774jU859jyPgrduzd8WOTlr7jx4L9/JZjHh5yayuW4NdM+5ypLLO9j29R9eDHGaPue391n+0bhuz3u2/5+3/K+6fZO8eGrfZah6znR+W0v1/b3uZX3Tbd3etzjnzWIxbvf9xcaLZT3yvut342TQXtjWOIdAF2/eGM83nV/xzghMdRlZOJEYXwKcrsA3afcaSrt3vO+vziI2cc/+UuNf9Qe+cinHfwJ6LjcddxPemvj3/Z9nEDo+kR1XJXg6ld9eDPU/Vje8KbP3Tb/7WVqO8kZJwqH8o0THXWfWI9lOfSk4Bi/A9vTfPVrP1Fr9YIXvW0T/2/7Vd/Ha7ancej3TX+k4pPBYXfI9GIAFtu1WramryWW93FSs+kyN5c86QUHX1498t10CaHKbd/BXdM7W/vcgzc+pRUne6tlqPt6aHn/utffUju9uKfQIflqeu4Q5X2sd9M7ZKOgiHcRN82/7udxoNF0wvb7vu8BB01PeZ/FSaLnHXkp/a9ksRYWYggB5f0AJwNLLe9jf3/++Y+o/ZsWfy/j3oD73g8v3uAVBd7239koS5v+Finvuw/s11jef+ozv7a5173Ou0OO7pipO20ufsJzNnHFqrpjwI/d9K1NjMfqjh2b3mCeS3kfb65sW/eHXXjJXlcni+OdOIF55pln3cGx7qpdYXtIed9kGNOJN2DHvrBuHLn7sxhX1l3Ov+nEeJ/yfvfNG1WO4pNCV1z5ur0sd5d7+9/Pangzb8znggsu3sQnE7ef3/R9PK/pynHVMm9/Vd4r74cY/JlGcQLK+wHG60374bX9/NjyPrwe/ogfuMMYo/o79YhHPnWvv30lmrd9WGmfKw5EV1I5bX+NsV/OHk0fDHrSk4e5/Hqse1PxHk77lvdN97sf6krYMZ3t7bb9fW7lfdP97uOKZENk7eVXva/RQnnffYw/xDaYexrFfvI+4OIdaI+/+NmNIY5f7rPPPnfzspe/5+B3p1cbJj7dGZ8urruXZA7lfSznc557baPBYx93xd6fwP/wR//K5n73f1jjtK55w2c6dzzVu9vintQvfNHb9r4vc+W9/TV2QnEFge2ddPV92ye7q0/qxkmNKPIP/eREtQxxMqbpDSLx6YJDy+lqumN9jU9g1BW6sax1n4aI5x76qY3dZf/kLb/a+MmaZ115zd5GsRxxcrDavm1fSyzvw+0VP/jBxvW79LJX1p4U3fWOf8e77preCLDttm95/9KXvbt2uSIfMa+6Zaj72fve/x+1ngStlu3Q8v7DH/tm7fLFFTlinnXLcsjPhijvY36R92odd7/GScJ9PxUUf3POu+9DGqe172C6MlDeZ3OyKC6lf3PPS+l/MZu1sCDHCijvBzgZWHJ5H6XY7t+I7X/HPjs+3V5XRMV9EGMsHu/i334zcfX6KLdi/F39e/ur8r77wH6N5X2MFb73ifEza3OznaG4z2i8gTCK0njjSNMxWrwmbhcV5fj266vvcynvY93jkut1v0vVssZYOAr0GA/vHvfFuC7eBB1vWK6eX32NMWqM7at/b389pLz/xM2/sgmv7ddvfx+33IjzAB/40M+deiNwHP/GlRPiPEbdFQHiihWxL9meXvV9n/K+7VNc1XTj+OWBD3rUyRsPonDY/b/tE0o3f/pXa49zq2nHpxLjagof+JGfrbH4zU2c9Io3GNQdK8etR554yfNqLZT3yvtjB35eX6SA8n6A8Xp1PmLtX4co75vO2cWb8OKc7BKN4wOBdX+z4+9+XF6/a53jDarVGGH765S34O1axrrHf/g1N9Uud9yqc983eddNd/tnMf7aNtn+ft/zjU1vMO3zIbDtZau+f9CDH924jIeW901X23rJy97VmaNqedq+Nr3BNH73216372Nx7LW9jba/V953H+Pv65zz84ou7wM2PgUQBfV2eOu+j51+HJTFjjDuR1H36YGYXvzhiwP55z3/upMTD3XTqn6WS3kff9R2P31TLWN8jUs3x2WW49O/dWGMdX7RS25oPTkQb5Koe+3uz6ryvpp/fAooTmpEcRmlVLzhYvc1u/+OExXh3/SHOi4j88Wv/LeN06nK+2oZ4sRG3MM7PjUSf8B357f77/gEeNxPpundjTHdq1/1kc7p7E53in8/5Qde0vm7ULnEc4dYprZLbj76MZefnERpmk+U9vHHvboVRbVsJ7m58PRtAOLxUsv7yGzbH937P+Cik5NsTZeAirIiBh2797ppGtTsW97H1TJ2Py1UbYc4URsnGduK57gMcJwY3F2u+9//wpM3T1XTqr4eWt5HdpouiRs5iSsHxJts4jYWcXmtuFrH7v8f+fg3G7M+VHkf92SKS3dW67n7NT5JH7fgiOfV/T5EruMKGXVvFKumFb9rda9t+5nyPrsTRn0upX9rdmthgfoKKO8HOBlYcnkfBWBcGavar7d9jXI0/q7E/zGWbXtufEo4jm2U97dfLe7QK2SttbyPMcR1b/nR1uPAtuztPhafwo8rucVVh3Yfi39HrtvGLVM/Fifu6pZz92dxVYu43VWM19veRBtj0zjxGZ+22Z1G/PuQ8j4s4s3wddPZ/VncYzTGfHFO4J73vE/jm6vjdfF43FKt6Y3Tfcr7WNamAnx3WZv+Hfu5gSEugAAAIABJREFUtu3f9Mmm3enFOYQTi/Mf3mkReYzjhLjC4u504t/Ke+V93wGf1xUtoLwfYLzetj9f02NDlPfxd6rpzYbPvOI1rX87S7aON9fV/W2OXqdrveK4qO61Ob2JtG4dYlvXLXf87I3Xfr5zveumuf2zOI7ePXe7Pb99y/vLLr+6djnj6tPb8+vz/bve+7XaaVfLeWh533SVgKFu7RS/g9WybX+NDxL3Wf/t18QH1ranufu98l55v5WJ20+EbP3wtgDNddn8KtBxcuBpl73ituWpW8bdn8Ufvjhwvc99HnjyKcc40VD3rvTd123/Owrpahmavo592fxqvvHp1KYTdtUyxx+pOGEYRXa8M+jJT3nx5oKHPaF1xx2vjXuQ73vf5t3yvpp39TVOLESZH8Vx7OBe8MK3bmJHGqVkFFMxr6ZBSUwj3lkY76iv1rvu6255X827+hrb+rGPe+ZJZmK+L37pO0/eLPCMK159cj/utk+SxDTCcPfTF3XLMcfP3n7DT+39exDPHWIZ44oNbYVjmMWJpPgdfd4Lrj/Z5lH2hmOUw9V22f4aZWy8gWP7Z9X3pZb3YR0nyrp+T+MyTnEp/fgdjTI3LpsUV5GIk4GVQfU1CuGmewHtW97HcsXvYTXNuq9xEizenRn3yozf17hPZNziIE5g1j0/9qUxiLjXufc99Xif8r7p9hl18677Wds9eocq78MxBpFtJ3Rj2cLmkY+69I774Qsu7twPx5uWmt6A1fZ7rLzP9sTRi1NK/7gurzU/uyXbtbBghwoo7wc4GVhyeR/7609/9tc3973vQ0/9faz53d/rOfFJ1vhUfky7aYzhk/fdB/ZrLu8jO3GiLMrpvjmM10UW4/czpheZrJtWjEvi8Zz+b7rSWt3yd/2senP/UOV9HHM+4YnPrbXsWpa6x+OqhO//4DdO/Icu7+PNvm1X8atbnu2fdZX3YdFUsm9PZ9/vY37vvfHrJxZN01XeK+8PHeh5/iIElPeZ/Z3Oacxw6LIMUd7HPJs+LR3nYuMDNYcuVwnPf8hDH1c7/tm3YG463zzU7ZXHMozxdN1YJq5s1fRhr32XJW7TVjft6mf72jadr49P5O+7LHXPi7FeU9arZTy0vG/q5Z76tJcftazV8je9OSCuwlU9p+/XuA1rtd51X5X3eR1T9t3OXa8r/pP32ysYn2q8+z3OaQ12XdgP/dk559zn5NPC+3yKvGknMdQ92bfXPy6Zt/sJ5kPXbff5URjFJ3O359P2fVd5vzv9Q/4dpdf1b/2xzmXpKu8Pmefuc+ONB9UJ0jaHuR6LTMYlLXeXe/ff8Zx98rvvesR9ydvedLE7/7Z/x718YtmaBgMll/fhGSeJuj5B1+ZTPRaflo+Tb/F7X/1s++sh5X1cFeBRj7q0djrb09zn+/hk0nVv+d7v6VDlfbgdclWJ3eWcqryP5bzxA9/YxN+I3WU45t8PfshjN30PMJT3WZ9AumtK6QMppd/vyMcnsl4LC3eIgPJ+gJOBpZf38bcirjIVb6Dt+N3vfDw+6bp9RRfl/e1vOPfJ+8NPZsQbS57+jFdtYix3SDbjUp7xCfF4Q33kO/7/2E3fqp1GXDGqek5OX+MKTm2XqO/yiJPn258EG6q8D6O4TGrTZfi7lmv78XiT/Edv+s9u8x+6vI9ljf1RvPH4jDPuXLv9t5dn9/uu8j6mH7cDiMvf77720H/HFcLiww9VBpX3t+87ayxj7OI/AmsSUN4PMF6v9q9r/zpUeR+3wqnZP5/8LD4s1eeDHjlvm/h733TedN8+penKmPEBqJzXPa7027St4371fZc9zh13jc/2Le+bch0fkD3mDQZN0932OLS8j/PT26+vvn/IQx7b23J7G8TVhKtpbn+ND9u1XTV6exp13zdNd3seyvvDj3frrHP/2aLK+8C++dN/bRPvnunaIW2Hfd/v41Pj8UntfT+FHsszZXkf84sTmhc9/Mm1O45917N6Xnyy9nNfrL/Ec1Owxyrv4x7Q+95/e4zyPorpyFXcoqBp3XP5eVz6pdqGTV+HujzM9jq/8uofOfhk3+7yXXHl6277Q7/U8j7MYtD0/d//oM7ttOtT/TuuGlLd732I8j6WKT4tc2yBH/f4fMc7b7+iw5DlfZwQfu7z39z5CfXKaPvrlOV9WMa9iuvugbq9TPt8f7LfufSqo/Y7yvsizjN1XUr/o0WshYXcR0B5P8DJwCWU9/G3Ig7m45368bdzn78J28+57/0uOLk0d0xn+3/l/e0FlPL+jtnYzknX9/Hmkri8e1yNoKnIj+PiuIpWjLHqbq/UdJnHuC1Z1/znevxDH/1PNo95bPsnXLZ/D6vv4xNCcXu47eUesryvpvv6N3725IqB1Xz3/Rqfto8rzW2/uSKmOUZ5Xy1rvNk6LnMfV5uK45Z9lnWf8r6aflzav+mkfNu84gpZsd+NUqCaVnxV3t++76zxU97vM8LznCUJKO93xpfb+0vfHza+aisjqyvh7GvaNkaJc3lDf9Aszu/vc1n+N1//5c1rr7nl1Dhj3/Wqe17c+rfm79HJ7XX3faNCXNG0bhr7rFPdMk31s+gdmsZOMf4+9Bgnljs+lBm9Sp3H9s/2Le+bzkPHtF716o/fYYy1r9tb3v4Te72R9tDyPo5rttex+j6ubvvxT/5Sr2XdXqe4GnE1zd2v0ZNsP3ff7+O8etOtpLfnobw/bH+8r39uz1tceV8Bxx+ZZ115zSCfRI8TF/GOl7g/dzX9fb9OXd7HcsVlRuIqBOed9+DGHcj2L/vu93F/6b6XVI8TkbHDjUvQ15VGu/Pq+neUf3Hgv3uA3eYfJwvi0xOPe/wVm/jD1jWPtsere2rv+8aBtuWa6rG4ZFLbp+DjsbEuqxROcSmfNtO6x+LkS7yTdNtoyeV9rGe8CSguKdR0UrTOKfIYA9Dtq2E0DZoO+eR95R5XPIh7fx56BZO4Z1JcMeGTt/zqHbbhkOV9tYyxb4/bCcSlpJoug7VrN3V5H8sa++FXv/YTvd+kEZ/IHOIdwXX74bj8WOV5zNe6aV940ZMGmXbXcsXVCHa3886/46RLaf89N6X0D3fWI9bzR0pbEcvbKKC8H+BkYNzDMA5g6v7f/lRp136kejw+1Vo3rXe++6uT7M/iBNQP/tCHN4+/+MpNlGw1+4CT2xPF+CpuXxMH8/E3plr+7a83vOurteuyW9ptv6bu+zgJUWcSb06re37Tz2K8UjedQ09WNk0/fh5XNKqbxyFvth5yOnGvyrrliU+it63HIY/VTf/tN/wHg01/e1kiO3GS7N3v/dom5hGlfIzFmjJYvfbVr60/6Rq3bauek+vXD37o505uXxWfVK/7QECMP2Mc+uznvPG2S9Dvrkvsi+q207E5iLF6FNdxG7z4dFPd/iKWOcZo8ZzXv/HWk0/u7y5f/DtuRVe3jPFhiLrnH/Oz2BfE70bkJ/ZhdfM9dNwbFlEYxBvs68akYRPHKPGG6biC1zVv+EzjG2Jjm9ct082fvuOxzTEGOb626Z7AO7lS3jcO6zywUAHl/QDj9Rz3eXMs05Dl/U03/0rrm37j712fYnfbJcZ3MeaL83vVucrtx+u+f/krbjwZj9z73vffvPyq920+9ZlfO2ocEfOPqzrt/C06+XecD61bhrqfxTnRumnEOC56i7rX5PKzKMDrlj1+FseL+159INYnchNXvmqa3vbP9y3vIydxy9ft11bfx1WT43Zch1i+7vWf3vv87qHlfSxH3bnpWN54U/EQb3r5vvucX2sRV+Y6dHwbt43tujVxZa28V95XWYjxWG0Iq5/Hgc4hv5RTPjcO6t75nj9/UvbEvSLiF7bu3tHVusTBbhyox+Xe4qRDHGQes7xxObb4xdv9/9B7T/ZZhrhUSZxkifuLx72p4+C1Ws/tr3Ef+bhcSFxVoPo0b5/51b0mTvLFuscnqmOnGPebb/OPsj1Kq7gcXix7bL+66e77szjhFIOXV/zgB0/eUf/gBz+ms5iMZYzi/6Uve/cg78Lad1mHfF5cFmY3c9W/q0uaDzm/7WnFH/G3vO3HT06ShOV21qrvI4txv9fLLr/6pLSvOwEYJ1GqZd7+GifZt+fX9n28w277tdX31f042167/dgbr/38qelU90jcfl6f7+N3JD6dEycB635Hq31SnCCMTwTtziM+7VSt1/bX7UtB7r6m699xwjsG4PFpmaY3wMS7AB924SUnn1752Cd+8dRyxTyue8uPnlq2IcuQeLNQfNIpyoq4pcb2+m9/37Yfj3Xdfm71fRQXXU77PB774be+/SdOsh73Aa3bxvF7EQcQ8Uax5zz3TSeX3t9n2vs85/q3/plT6zfUCf66ab/jnf/hIG5d67bQ8j7GW9Wl9H+v2l+mlG5c6ImxNa6W8t7JwM59ZPxdiuIvxuRxciJOuBw7Hu7ap3p8HQf+U27nGONv/R277fv4BPiUy3HsvOIN5DHOvfEDf2kTxybxCZ1jLgd67PLsvj6ukhfHEvGGmFi+WNZD3vS+O72S/x2fVIv9ZZzIPLG46Vurtdh3Oyrv1zgUtc57CCjvjdcHG6sMWd7Hvj3OwdS9sbAac8WHtR716MtOzkHFPn6fvwdx7BFvrItPpJ977/vdNmarptk1jaq8r54fyxfnEuNDkPF3uev11eNxnjZ6iaYP6cS5tEPenBBvwKvegFAt2/bXCy64+OTqnnEb2HhjZN2b+OJnQ5S71Tru+zWO/dquqhzGcY647Y0SMS6Ky+zvfjgrTOLc5LZF9f2+5X2sR4zpq9ftfo1t+KKX3ND5Idi47WjdFWAjh/FmkN3pxr/7lPfPee61tdOK6d397mef9BdXveLGk3On8TtWl4W2DmD3d2B7ucM7PoBWd6Wy7TzEm2rrrq4Rt8duesOu8n4dx/CRx+1MNXxfbnm//YtQfR8HtPGJgfjFeO/7f+bkYDwOdOMPVvWcJX6NIjve7R8nA6v1nuOEYOX/Ix/+yyfLEcsTnwSPd+RPdXIy3jwRJ0bjD0VYxKeXIxP7Xn5nifkYa53i/plRhIZzta3XelKpyzgG1zEQud3qm7OfdIqCPK5mESfBYrlicBBviuhaF4/XDyLq9sPhOdW+bynbZcHlfXU+aftS+u+ufuhr8QLKeycD/f2UgcVnIMY0TSeZ4uTDUsYi1qN+rMulHBflffHjSiswjoDy3lhtsLHK0OV9/I19zetubvxQyHapE5/evfCiSzaXP/2HNs97wfWbKCdfdtV7T76Pq4A+4YnPPblyTduVW2N6XX/X24rLeP0559xn8+jHXH5ydeQXv/SGzdWv+vDmmjfcenLF3Fie+PBkfKCw7YN+MZ0oq7uWZffxJz35hfuUXq3Pedd7fvrg+e4uR59/R1/TdPn8ajtHMRwfVI1b48aV3OLqn1EUx+2Amq5eEB8ei21fTWP76yHl/We/8O3O5YtliKvLxXaPN3PE7RCi9I9MNh0rxJs04kNB97//hbXL2Ke8j/6n6U0h2+vf9n282aNpO8aHDOve+LI9vTPPPGvzuMc/6+TWZHFlhfg9jivfxodXm64gFb8T8WGwpisnKO/LGXM3ZWefn6+yvN8HxnPW8QtgO9vOMiADMlBeBlZQ3lenol6YUnpR9Q9fixdQ3jsZ2HjQ729ReX+LbLP6bRZXXto+WVV9H1d7i0+Kc6t348Jl6gwo74sfV1qBcQSU98brg41Vxijv42/Ftdd9aa97YldjsGO+dv1t6irvj5l39donXvK8Xh9suuXWX9uce+59a8el1bS7vs5V3od7fAAuPhnetYz7Ph5XGY432Q5R3sfyxZVVY3y/7/y7nhdvJIliO6Y9ZHkf04vbqXbNv+3xtvI+ph+3Ed73cvdt89l+LPYfMW3l/bqPEZT3BiWDDUpih+J/BjIgAzIgA2NnYEXl/TinpEx1LgHlvbGisbIMLDoDccWmuA3Q9omn6vvHPPYZi173scc+pm98PXQGlPdzDQfNN3MB5b2x2mDjlbHK+/h7ELfV3OO8SO2YrBqbdX2NYrbrb8+Y5X2Uuc+44tVHXakyrvobt0fuWtemx+cs78M+rhjcdH/5pmWu+3kU99WbaIcq72P5Drlffd1yVT+LrP3wa266LW9Dl/dxu964PXLfNxt0lfdhEbfcHaLAjysqXP2qj9xmobxf9zGA8t6g5LadQdcfZI+ve2dh+9v+MiADuWRgj4PUOOniPwK5CSjvjbuNu2Vg0Rl43vOvazw5+oY3fW7R657LGMlyGK/vmwHlfW7DRMuTiYDy3lhtsPHKmOV97OvjU9RxqfSmS5BXxeihX+N8S3xSue2e6tXfmvh0eHwyfvfe6ofOc/f5UVi++71fG2RbxGXNY30e+KBHNY5Td+df/Xvu8j6c4zbDj7/42Qcve6zDyf3nX/yOzRe/8rdusxyyvI/li1usHpPB889/xOY9N379tuWLaQ5d3ld5jdsmX3rZKw++osE+5X3M4/0f/MbmvPMe3Gtbxfa6/wMuOpV75f26x/bKe4OSO+wcq52Zr+veMdj+tr8MyEDOGVDeZ3Jqy2IcKqC8N+427paBrDLw8U/+0uYtb/+JQZYpPs3SdL/SOKG3fdIw5zGGZTMGXksGlPeHDuM8fyUCyntjtUHGRfG35M3Xf2XziEc+tfb/j3z8Px1sPlHix7ziXudnnXWvg4vDs8++90kB/0M//NFN3B+8z9/BuPpSXEY97rn+mMc+fXP22ecevBxnnXXu5mmXveJkOvFJ6T7L0fWaj3/iFzfXXvflk3uPX3b5D26e9OQXbeJT6U3b6YMf/vlRlqNrOesej3vBP/Zxz9zEfeGrNxc0fY0cxP3U67bnC1/0ttr1fed7/nzvdY0xRVzy/n73u6Bz2WKZzzjjzpuLHv6UzevfeOsmsrO7vk+99KraZbzpk7986rm7r93n3/GGjvfe+PWTZb7iytdtnnbpVSe3E3jUoy6tnW9cAWKf6cZzvvDlv3nyppoo4pu2z/bPY3teeNElJ1cx+MKXf/PUfC5/+tW1y1S3bfddRs8r53hHeW9Qcmqn4Be4nF9g28q2kgEZWGMGlPcrOXW2vNVU3ht3G3fLQFYZiE+5xMmj+CRSXPbys184/J70H/n4NzcXP+E5rSenfOreeHWN49Xc11l5v7yBpjUaREB5b6yW1Vjt0L8lUeR/6CO/sLnmDZ/ZvOglN2wuf/oPbX7gqS89KSajEH3ms167iSslxWW5owy++dO/Otr6RlH+9ht+ahNvCnju89+8efozXnWyLPFJ/fj085XPfv1Jif6mN3+htmQ+dN3X8vxPf/bXN2968xc3UcKHY3yS/gee+rLNM575w5u42kNcLWDON83Gm4Nfe82nNs97wfUn+XviJc8/Wb7Y/i+/6n2b69/6Zzaf+dxvjJa7nHJw082/snntNbfsWLz05HfhZVe9d3PdW35s85nP/vVVWOS0XUpaFuW9QYkdhAzIgAzIgAwUlQHl/SAnpkxkegHlvX1tUfvakg5qLWu/crgq76tPf9z1rmeeFPFxQjc+jRLlXp1tFPZxmdYo7bs+/ROX+aybhp/122bcuA2VAeX99ANBcyxCQHlvvG7cIgMyIAMyIANZZEB5L4hZBHGoA1DTcTJDBmRABpafAeV9ESe+LORpAeW9cbdxtwxklYHd8r4q8be/3uMe9zy5j2VcBvOcc+6zufOd79L6Kfvt18Y9LG/9/H+V1TobJy5/nGgb77eNlfenB2p+QiClpLw3VjNukQEZkAEZkIEsMqC8F8QsgugAe78DbE6cZEAGZODvbJT3TqwVKqC8N+427paBrDKwT3m/XcYf8v1DL7h4c8ut/2VW62sMZRwtA7dnQHlf6GjSYo8toLw3VjN2kQEZkAEZkIEsMqC8F8Qsgugg+vaDaBYsZEAGZKA9A8r7sc9Zmf5IAsp7427jbhnIKgNjlPd3vvNdN8957ps2X/jyb2a1rsZW7WMrPuvzUd6PNNoz2dIFlPfGasYvMiADMiADMpBFBpT3gphFEJ0sWN/JAtvcNpcBGeibAeV96efEVrv8ynvjbuNuGcgqA3FJ+1e88gObix7+lM5713d96v6ud7375tLLXrn5yMe/mdU69h1reJ1x6tIzoLxf7XjUircLKO+N1YxjZEAGZEAGZCCLDCjvBTGLIC79wNj6OfkjAzIgA8NlQHnffsbJo9kKKO+Nu427ZSDbDHzmc7+xeevbf2Lzwhe/fXPxE569edCDH70566xzG+9vf8973mdzwcOesHnGFa/eXHvdlzaf/cK3s103Y7DhxmAsl2OpvM92vGjB5hVQ3hurGc/IgAzIgAzIQBYZUN4LYhZBdBJgOScBbEvbUgZkYOwMKO/nPaNl7r0FlPfG3cbdMlBcBr74lb+1+fRnf33ziU/91c3HbvrWyfdR+o39t970jSdlYNwMKO97j+e8cNkCyntjNWMcGZABGZABGcgiA8p7QcwiiA7Mxz0w58tXBmRgSRlQ3i/7jNmC1055b9xt3C0DMiADMiADWWRAeb/gEadVO0ZAeW8fncU+eknnb6yL85EyIAMy0C8DynuDEoMSGZABGZABGSgqA8r7Y85Hee2MAsp7+9qi9rUOsPsdYHPjJgMyUEIGlPczjgjNOmcB5b3xuvG6DMiADMiADGSRAeW9IGYRxBIObi2jkzAyIAMykEcGlPc5n++ybC0CynvjbuNuGZABGZABGcgiA8r7lhGbh9YsoLy3j85iH+3cUx7nnmwH20EGZGDODCjvDUoMSmRABmRABmSgqAwo79d8Pq3odVfe29cWta+d8yDVvJ0kkQEZkIFxM6C8L3pMaeHHE1DeG68br8uADMiADMhAFhlQ3gtiFkF0YD7ugTlfvjIgA0vKgPJ+vLNVpjyqgPLeuNu4WwZkQAZkQAayyIDyftQxn4mXK6C8t4/OYh+9pPM31sX5SBmQARnolwHlvUGJQYkMyIAMyIAMFJUB5X25Z8NWvuTKe/vaova1DrD7HWBz4yYDMlBCBpT3Kx+VWv0mAeW98brxugzIgAzIgAxkkQHlvSBmEcQSDm4to5MwMiADMpBHBpT3Teea/DxzAeW9cbdxtwzIgAzIgAxkkQHlfeajRos3l4Dy3j46i320c095nHuyHWwHGZCBOTOgvDcoMSiRARmQARmQgaIyoLyf61yW+R4poLy3ry1qXzvnQap5O0kiAzIgA+NmQHl/5KjOy5cqoLw3XjdelwEZkAEZkIEsMqC8F8QsgujAfNwDc758ZUAGlpQB5f1Sz5Utfr2U98bdxt0yIAMyIAMykEUGlPeLH3dawX4Cynv76Cz20Us6f2NdnI+UARmQgX4ZUN4blBiUyIAMyIAMyEBRGVDe9zsT5VWzCyjv7WuL2tc6wO53gM2NmwzIQAkZUN7PPi60AHkKKO+N143XZUAGZEAGZCCLDCjvBTGLIJZwcGsZnYSRARmQgTwyoLzP80yXpeoUUN4bdxt3y4AMyIAMyEAWGVDed47bPGGdAsp7++gs9tHOPeVx7sl2sB1kQAbmzIDy3qDEoEQGZEAGZEAGisqA8n6dZ9IWsNbKe/vaova1cx6kmreTJDIgAzIwbgaU9wsYWVqFMQSU98brxusyIAMyIAMykEUGlPeCmEUQHZiPe2DOl68MyMCSMqC8H+M8lWlOIKC8N+427pYBGZABGZCBLDKgvJ9g5GcWJQoo7+2js9hHL+n8jXVxPlIGZEAG+mVAeW9QYlAiAzIgAzIgA0VlQHlf4nkwy5xSUt7b1xa1r3WA3e8Amxs3GZCBEjKgvDc2JVAroLw3XjdelwEZkAEZkIEsMqC8F8QsgljCwa1ldBJGBmRABvLIgPK+9kSTH+YvoLw37jbulgEZkAEZkIEsMqC8z3/gaAlnEVDe20dnsY927imPc0+2g+0gAzIwZwaU9wYlBiUyIAMyIAMyUFQGlPeznMgy0+MFlPf2tUXta+c8SDVvJ0lkQAZkYNwMKO+PH9iZwiIFlPfG68brMiADMiADMpBFBpT3gphFEB2Yj3tgzpevDMjAkjKgvF/kibI1rJTy3rjbuFsGZEAGZEAGssiA8n4NQ0/r2ENAeW8fncU+eknnb6yL85EyIAMy0C8DynuDEoMSGZABGZABGSgqA8r7HqehvCQHAeW9fW1R+1oH2P0OsLlxkwEZKCEDyvschoaWIUMB5b3xuvG6DMiADMiADGSRAeW9IGYRxBIObi2jkzAyIAMykEcGlPcZnuaySPsIKO+Nu427ZUAGZEAGZCCLDCjv9xm6ec4KBZT39tFZ7KOde8rj3JPtYDvIgAzMmQHlvUGJQYkMyIAMyIAMFJUB5f0KT6MtY5WV9/a1Re1r5zxINW8nSWRABmRg3Awo75cxuLQWgwso743XjddlQAZkQAZkIIsMKO8FMYsgOjAf98CcL18ZkIElZUB5P/hJKhOcRkB5b9xt3C0DMiADMiADWWRAeT/N4M9cihNQ3ttHZ7GPXtL5G+vifKQMyIAM9MuA8t6gxKBEBmRABmRABorKgPK+uJNgFvh7Asp7+9qi9rUOsPsdYHPjJgMyUEIGlPeGpwRqBZT3xuvG6zIgAzIgAzKQRQaU94KYRRBLOLi1jE7CyIAMyEAeGVDe155o8sP8BZT3xt3G3TIgAzIgAzKQRQaU9/kPHC3hLALKe/voLPbRzj3lce7JdrAdZEAG5syA8t6gxKBEBmRABmRABorKgPJ+lhNZZnq8gPLevraofe2cB6nm7SSJDMiADIybAeX98QM7U1ikgPLeeN14XQZkQAZkQAayyIDyXhCzCKID83EPzPnylQEZWFIGlPeLPFG2hpVS3ht3G3fLgAzIgAzIQBYZUN6vYehpHXsIKO/to7PYRy/p/I11cT5SBmRABvplQHlvUGJQIgMyIAMyIANFZUB53+M0lJfkIKC8t68tal/rALvfATY3bjIgAyVkQHmfw9DQMmQooLw3XjdelwEZkAFP7rMxAAAgAElEQVQZkIEsMqC8F8QsgljCwa1ldBJGBmRABvLIgPI+w9NcFmkfAeW9cbdxtwzIgAzIgAxkkQHl/T5DN89ZoYDy3j46i320c095nHuyHWwHGZCBOTOgvDcoMSiRARmQARmQgaIyoLxf4Wm0Zayy8t6+tqh97ZwHqebtJIkMyIAMjJsB5f0yBpfWYnAB5b3xuvG6DMiADMiADGSRAeW9IGYRRAfm4x6Y8+UrAzKwpAwo7wc/SWWC0wgo7427jbtlQAZkQAZkIIsMKO+nGfyZS3ECynv76Cz20Us6f2NdnI+UARmQgX4ZUN4blBiUyIAMyIAMyEBRGVDeF3cSzAJ/T0B5b19b1L7WAXa/A2xu3GRABkrIgPLe8JRArYDy3njdeF0GZEAGZEAGssiA8l4QswhiCQe3ltFJGBmQARnIIwPK+9oTTX6Yv4Dy3rjbuFsGZEAGZEAGssiA8j7/gaMlnEVAeW8fncU+2rmnPM492Q62gwzIwJwZUN4blBiUyIAMyIAMyEBRGVDez3Iiy0yPF1De29cWta+d8yDVvJ0kkQEZkIFxM6C8P35gZwqLFFDeG68br8uADMiADMhAFhlQ3gtiFkF0YD7ugTlfvjIgA0vKgPJ+kSfK1rBSynvjbuNuGZABGZABGcgiA8r7NQw9rWMPAeW9fXQW++glnb+xLs5HyoAMyEC/DCjvDUoMSmRABmRABmSgqAwo73uchvKSHASU9/a1Re1rHWD3O8Dmxk0GZKCEDCjvcxgaWoYMBZT3xuvG6zIgAzIgAzKQRQaU94KYRRBLOLi1jE7CyIAMyEAeGVDeZ3iayyLtI6C8N+427pYBGZABGZCBLDKgvN9n6OY5KxRQ3ttHZ7GPdu4pj3NPtoPtIAMyMGcGlPcGJQYlMiADMiADMlBUBpT3KzyNtoxVVt7b1xa1r53zINW8nSSRARmQgXEzoLxfxuDSWgwuoLw3XjdelwEZkAEZkIEsMqC8F8QsgujAfNwDc758ZUAGlpQB5f3gJ6lMcBoB5b1xt3G3DMiADMiADGSRAeX9NIM/cylOQHlvH53FPnpJ52+si/ORMiADMtAvA8p7gxKDEhmQARmQARkoKgPK++JOglng7wko7+1ri9rXOsDud4DNjZsMyEAJGVDeG54SqBVQ3huvG6/LgAzIgAzIQBYZUN4LYhZBLOHg1jI6CSMDMiADeWRAeV97oskP8xdQ3ht3G3fLgAzIgAzIQBYZUN7nP3C0hLMIKO/to7PYRzv3lMe5J9vBdpABGZgzA8p7gxKDEhmQARmQARkoKgPK+1lOZJnp8QLKe/vaova1cx6kmreTJDIgAzIwbgaU98cP7ExhkQLKe+N143UZkAEZkAEZyCIDyntBzCKIDszHPTDny1cGZGBJGVDeL/JE2RpWSnlv3G3cLQMyIAMyIANZZEB5v4ahp3XsIaC8t4/OYh+9pPM31sX5SBmQARnolwHlvUGJQYkMyIAMyIAMFJUB5X2P01BekoOA8t6+tqh9rQPsfgfY3LjJgAyUkAHlfQ5DQ8uQoYDy3njdeF0GZEAGZEAGssiA8l4QswhiCQe3ltFJGBmQARnIIwPK+wxPc1mkfQSU98bdxt0yIAMyIAMykEUGlPf7DN08Z4UCynv76Cz20c495XHuyXawHWRABubMgPLeoMSgRAZkQAZkQAaKyoDyfoWn0Zaxysp7+9qi9rVzHqSat5MkMiADMjBuBpT3yxhcWovBBZT3xuvG6zIgAzIgAzKQRQaU94KYRRAdmI97YM6XrwzIwJIyoLwf/CSVCU4joLw37jbulgEZkAEZkIEsMqC8n2bwZy7FCSjv7aOz2Ecv6fyNdXE+UgZkQAb6ZUB5b1BiUCIDMiADMiADRWVAeV/cSTAL/D0B5b19bVH7WgfY/Q6wuXGTARkoIQPKe8NTArUCynvjdeN1GZABGZABGcgiA8p7QcwiiCUc3FpGJ2FkQAZkII8MKO9rTzT5Yf4CynvjbuNuGZABGZABGcgiA8r7/AeOlnAWAeW9fXQW+2jnnvI492Q72A4yIANzZkB5b1BiUCIDMiADMiADRWVAeT/LiSwzPV5AeW9fW9S+ds6DVPN2kkQGZEAGxs2A8v74gZ0pLFJAeW+8brwuAzIgAzIgA1lkQHkviFkE0YH5uAfmfPnKgAwsKQPK+0WeKFvDSinvjbuNu2VABmRABmQgiwwo79cw9LSOPQSU9/bRWeyjl3T+xro4HykDMiAD/TKgvDcoMSiRARmQARmQgaIyoLzvcRrKS3IQUN7b1xa1r3WA3e8Amxs3GZCBEjKgvM9haGgZMhRQ3huvG6/LgAzIgAzIQBYZUN4LYhZBLOHg1jI6CSMDMiADeWRAeZ/haS6LtI+A8t6427hbBmRABmRABrLIgPJ+n6Gb56xQQHlvH53FPtq5pzzOPdkOtoMMyMCcGVDeG5QYlMiADMiADMhAURlQ3q/wNNoyVll5b19b1L52zoNU83aSRAZkQAbGzYDyfhmDS2sxuIDy3njdeF0GZEAGZEAGssiA8l4QswiiA/NxD8z58pUBGVhSBpT3g5+kMsFpBJT3xt3G3TIgAzIgAzKQRQaU99MM/sylOAHlvX10FvvoJZ2/sS7OR8qADMhAvwwo7w1KDEpkQAZkQAZkoKgMKO+LOwlmgb8noLy3ry1qX+sAu98BNjduMiADJWRAeW94SqBWQHlvvG68LgMyIAMyIANZZEB5L4hZBLGEg1vL6CSMDMiADOSRAeV97YkmP8xfQHlv3G3cLQMyIAMyIANZZEB5n//A0RLOIqC8t4/OYh/t3FMe555sB9tBBmRgzgwo7w1KDEpkQAZkQAZkoKgMKO9nOZFlpscLKO/ta4va1855kGreTpLIgAzIwLgZUN4fP7AzhUUKKO+N143XZUAGZEAGZCCLDCjvBTGLIDowH/fAnC9fGZCBJWVAeb/IE2VrWCnlvXG3cbcMyIAMyIAMZJEB5f0ahp7WsYeA8t4+Oot99JLO31gX5yNlQAZkoF8GlPcGJQYlMiADMiADMlBUBpT3PU5DeUkOAsp7+9qi9rUOsPsdYHPjJgMyUEIGlPc5DA0tQ4YCynvjdeN1GZABGZABGcgiA8p7QcwiiCUc3FpGJ2FkQAZkII8MKO8zPM1lkfYRUN4bdxt3y4AMyIAMyEAWGVDe7zN085wVCijv7aOz2Ec795THuSfbwXaQARmYMwPKe4MSgxIZkAEZkAEZKCoDyvsVnkZbxior7+1ri9rXznmQat5OksiADMjAuBlQ3i9jcGktBhdQ3huvG6/LgAzIgAzIQBYZUN4LYhZBdGA+7oE5X74yIANLyoDyfvCTVCY4jYDy3rjbuFsGZEAGZEAGssiA8n6awZ+5FCegvLePzmIfvaTzN9bF+UgZkAEZ6JcB5b1BiUGJDMiADMiADBSVAeV9cSfBLPD3BJT39rVF7WsdYPc7wObGTQZkoIQMKO8NTwnUCijvjdeN12VABmRABmQgiwwo7wUxiyCWcHBrGZ2EkQEZkIE8MqC8rz3R5If5CyjvjbuNu2VABmRABmQgiwwo7/MfOFrCWQSU9/bRWeyjnXvK49yT7WA7yIAMzJkB5b1BiUGJDMiADMiADBSVAeX9LCeyzPR4AeW9fW1R+9o5D1LN20kSGZABGRg3A8r74wd2prBIAeW98brxugzIgAzIgAxkkYFByvsHnP/wzYUXXeJ/BjIgAzIgAzIgA6Nn4Mwzz9qklNr+j5Mu/iOQm4Dy3gFgFgeACrFxCzG+fGVABkrIgPI+t2Gi5clEQHlvvG68LgMyIAMyIANZZGCQ8r7jBHrbyXWPtZcPfPjIgAzIgAzIwOEZUN5ncvbLYtxBQHnvADCLA8ASSiXLqPyUARmQgXEzoLy/wxjNPwhUAsp743XjdRmQARmQARnIIgPK+8MLASUKMxmQARmQARnIOwPK++r0k685CSjvHQBmcQCoEBu3EOPLVwZkoIQMKO9zGiJalowElPfG68brMiADMiADMpBFBpT3eZcPyiHbRwZkQAZkQAYOz4DyPqMzYBblNgHlvQPALA4ASyiVLKPyUwZkQAbGzYDy/rbxmW8IbAso743XjddlQAZkQAZkIIsMKO8PLwSUKMxkQAZkQAZkIO8MKO+3T0H5PheBzvL+7vc4Z+N/BjIgAzIgAzIgA1NkYI9bYMbYxX8E1iTQWt6fccYZxuqOV2RABmRABmRABibJwN3udo99+oe0z5M8J+8iw/axfWRABmRABtaSAeX9mk6xlbOuneX9HifR1/I7bD39vZIBGZABGZCB+TOgvC9nnGlJhxFoLe+N1fUDMiADMiADMiADmWXABslsgziInf8g1jawDWRABmQg3wwo74c5eWUqwwoo7/PdZ9if2zYyIAMyIAMycDoDyvthx4Kmlr+A8v70fsC+kYkMyIAMyIAM5JsB5b3yXgZkQAZkQAZkoJgMKO/zPzG2xiVU3ud7sONA1LaRARmQARmQgdMZUN6vccS67nVW3p/eD9g3MpEBGZABGZCBfDOQPp+S/xnIgAz0ysCP9ig8/zvWvaztq/2tkoHvZeDKdZ9zsvaZCijv8z3YcSBq28iADMiADMjA6Qwo7zMdVFqs0QSU96f3A/aNTGRABmRABmQg3wyMNigyYQIEli9wTo/y/kvLZ7GGBAgQILAyAeV9vgc7DkRtGxmQARmQARk4nQHl/coGq1Y3Ke9P7wfsG5nIgAzIgAzIQL4ZMHwjQIBAbwHlfW86LyRAgACBBQko7/M92HEgatvIgAzIgAzIwOkMKO8XNBC1KnsJKO9P7wfsG5nIgAzIgAzIQL4Z2GuA40kECBCoE1De16n4GQECBAisTUB5n+/BjgNR20YGZEAGZEAGTmdAeb+20ar1Vd6f3g/YNzKRARmQARmQgXwzYPRGgACB3gLK+950XkiAAAECCxI4L6X0Gv8zkAEZ6MjAbx14y6l/3TE9+x37XhmQgb4ZiLGL/wisSeAx/qYap8mADByRgV85cBwfhei1R8yv7993rzM2lIHlZGBN4zTrSoDAwALK+4FBTY4AAQIECBAgQGCxAr944Em/312shBUjQIAAAQIECBAgUI7Alw4cx0d5H+fN/UeAAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uSh5WmkAACAASURBVBkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQmFxAeT85uRkSIECAAAECBAgUKqC8L3TDWWwCBAgQIECAAIFVCyjvV735rTwBAgQIEChLQHlf1vaytAQIECBAgAABAvMJKO/nszdnAgQIECBAgAABAn0FlPd95byOAAECBAgQOCXwjpTS74z4/z9JKW0O/P9fj7g8sa6/lVI685SEHxAgQIAAAQIECBCYV0B5P6+/uRMgQIAAAQIECBDoI6C876PmNQQIECBAgECtwD1SSv/0wHL90DI+t+e/vVbCDwkQIECAAAECBAjMK6C8n9ff3AkQIECAAAECBAj0EVDe91HzGgIECBAgQKBR4HUrKu//fkrpzo0SHiBAgAABAgQIECAwn4Dyfj57cyZAgAABAgQIECDQV0B531fO6wgQIECAAIFagTullP72Cgr876aUnlUr4IcECBAgQIAAAQIE5hdQ3s+/DSwBAQIECBAgQIAAgUMFlPeHink+AQIECBAg0CnwpJTSny68wP9mp4InECBAgAABAgQIEJhPQHk/n705EyBAgAABAgQIEOgroLzvK+d1BAgQIECAQKvA1xdc3v9RSumC1rX3IAECBAgQIECAAIFmgQellP7iyP//zoHj8T8ceXlifZ/dTOIRAgQIECBAgAABAgRSSsp7MSBAgAABAgRGEbhfSunfHHjCcFPI828dRcxECRAgQIAAAQIE1iTw64WMfYcao/+fKaV7rWkDW1cCBAgQIECAAAECPQSU9z3QvIQAAQIECBDYT+DDCzwh+c9SSmfvt/qeRYAAAQIECBAgQKBR4DEppX+3wPFyU9l/XaOEBwgQIECAAAECBAgQqASU95WErwQIECBAgMDgAndNKf2vCzshec3gSiZIgAABAgQIECCwVoE/u7CxclNx//dSSmesdSNbbwIECBAgQIAAAQIHCCjvD8DyVAIECBAgQOBwgZcv6ITk300p3elwAq8gQIAAAQIECBAgUCsQl5H/3QWNl+vK+++mlK6oXXs/JECAAAECBAgQIEBgV0B5vyvi3wQIECBAgMDgAku4n+d3UkpPG1zGBAkQIECAAAECBNYu8M6Fl/c/v/YNbP0JECBAgAABAgQIHCCgvD8Ay1MJECBAgACBfgJLuJ/nz/Rbda8iQIAAAQIECBAg0CoQl5P/HxZa4P9hSumhrWvvQQIECBAgQIAAAQIEtgWU99savidAgAABAgRGEyj5fp6/l1I6fzQZEyZAgAABAgQIEFi7wLMXWt7fvPYNa/0JECBAgAABAgQIHCigvD8QzNMJECBAgACBfgIl38/zI/1W2asIECBAgAABAgQI7C3wrYUV+P9bSunue6+9JxIgQIAAAQIECBAgEALKezkgQIAAAQIEJhMo8X6ev51SOnMyITMiQIAAAQIECBBYq8BDUkpxmfnNQv7/4bVuSOtNgAABAgQIECBA4AgB5f0ReF5KgAABAgQIHCZQ4v08rzpsFT2bAAECBAgQIECAQG+Bzy2kuP+bvQW8kAABAgQIECBAgMC6BZT3697+1p4AAQIECEwu8MyU0ncLOSn57cl1zJAAAQIECBAgQGDNAvdIKf3TQsbKTVcI+NOU0iVr3ojWnQABAgQIECBAgMARAsr7I/C8lAABAgQIEOgnUML9PP8kpXRxv9XzKgIECBAgQIAAAQK9Bd5YeHn/073X3AsJECBAgAABAgQIEFDeywABAgQIECAwuUAJ9/P8yclVzJAAAQIECBAgQIBASndKKf3tQgv8f5VSOs9GJECAAAECBAgQIECgt4DyvjedFxIgQIAAAQLHCOR8P89/mVL6/mNWzmsJECBAgAABAgQIHCHwlJTSdwos8D9wxDp7KQECBAgQIECAAAECKSnvpYAAAQIECBCYRSDn+3m+axYRMyVAgAABAgQIECBwu8A3Civv/1FK6a63L77vCBAgQIAAAQIECBDoIaC874HmJQQIECBAgMAwAjnez/MfpJTuMszqmQoBAgQIECBAgACB3gL3Tyn9m4IK/Jf0XlMvJECAAAECBAgQIECgElDeVxK+EiBAgAABApML5Hg/zxdOrmCGBAgQIECAAAECBOoFPl5Ief/X6hffTwkQIECAAAECBAgQOFBAeX8gmKcTIECAAAECwwrkdD/PXxp21UyNAAECBAgQIECAwFECd0sp/ePMC/w/Tik96qi19GICBAgQIECAAAECBCoB5X0l4SsBAgQIECAwm8B/nMEJyX+bUnrEbAJmTIAAAQIECBAgQKBe4OoMxsqblmX4sfrF9lMCBAgQIECAAAECBHoIKO97oHkJAQIECBAgMKxA3M/z/2k5Idh2snCox74w7CqZGgECBAgQIECAAIHBBH5j5rFy05j7/0opnTvYWpoQAQIECBAgQIAAAQLKexkgQIAAAQIEshC4acYTkv88pXROFgoWggABAgQIECBAgMBpgcellP7djOPlpvL+bacX1U8IECBAgAABAgQIEDhCQHl/BJ6XEiBAgAABAsMJzHk/z2uHWw1TIkCAAAECBAgQIDCKwJ/LrLz/+ymlM0ZZUxMlQIAAAQIECBAgsF4B5f16t701J0CAAAEC2QnMcT/P30op3Sk7CQtEgAABAgQIECBA4I4C35dS+hcZFfhX3nHx/IsAAQIECBAgQIAAgQEElPcDIJoEAQIECBAgMJzAlPfz/G5K6bLhFt2UCBAgQIAAAQIECIwq8L5Myvu/MupamjgBAgQIECBAgACB9Qoo79e77a05AQIECBDIUmDK+3n+bJYCFooAAQIECBAgQIBAvcCdU0r/88wF/h+llC6oXzw/JUCAAAECBAgQIEDgSAHl/ZGAXk6AAAECBAgMLzDF/Tx/P6X0wOEX3RQJECBAgAABAgQIjCrwvJnL+1tHXTsTJ0CAAAECBAgQILBuAeX9ure/tSdAgAABAlkKTHE/z09kueYWigABAgQIECBAgEC3wF+dqcD/Zymls7sXzzMIECBAgAABAgQIEOgpoLzvCedlBAgQIECAwLgCY97P83dSSncfd/FNnQABAgQIECBAgMBoAhellP7fGQr8a0ZbIxMmQIAAAQIECBAgQCAElPdyQIAAAQIECGQpMOb9PH8oyzW2UAQIECBAgAABAgT2F/jyxOX9300p3Wn/xfNMAgQIECBAgAABAgR6CCjve6B5CQECBAgQIDCNwBj38/wb0yy6uRAgQIAAAQIECBAYVeCclNL/MVGB/52U0tNGXRsTJ0CAAAECBAgQIEAgBJT3ckCAAAECBAhkLTDk/Tz/NKX0xKzX1sIRIECAAAECBAgQ2F/g+onK+5/Zf5E8kwABAgQIECBAgACBIwSU90fgeSkBAgQIECAwvsCQ9/P86viLaw4ECBAgQIAAAQIEJhOIy9j/vZEL/N9LKZ0/2RqZEQECBAgQIECAAIF1Cyjv1739rT0BAgQIEChCYIj7ef6rlNJ5RaythSRAgAABAgQIECCwv8DlKaXvjljgf2T/RfFMAgQIECBAgAABAgSOFFDeHwno5QQIECBAgMD4AkPcz/PG8RfTHAgQIECAAAECBAjMIvALI5X3v51SOnOWNTJTAgQIECBAgAABAusUUN6vc7tbawIECBAgUJzAMffz/IcppbsUt8YWmAABAgQIECBAgMB+Ag9KKf3BCAX+VfvN3rMIECBAgAABAgQIEBhIQHk/EKTJECBAgAABAuMKHHM/zxePu2imToAAAQIECBAgQGB2gVsGLu+/PfsaWQACBAgQIECAAAEC6xNQ3q9vm1tjAgQIECBQrECf+3n+58WurQUnQIAAAQIECBAgsL/A3VNK/2SgAv9PUkoX7z9rzyRAgAABAgQIECBAYCAB5f1AkCZDgAABAgQITCPwlw84IfnHKaVHTbNY5kKAAAECBAgQIEBgdoHXHDBW3rQ89ydnXxMLQIAAAQIECBAgQGCdAsr7dW53a02AAAECBIoViPt5/n7Licbtk5BfKXYtLTgBAgQIECBAgACBfgK/uedYeXvcvP39v0wpfX+/WXsVAQIECBAgQIAAAQJHCijvjwT0cgIECBAgQGB6gU/tcULyd1NK95p+0cyRAAECBAgQIECAwKwCl6SU/nSP8fJ2Yb/9/btnXXozJ0CAAAECBAgQILBuAeX9ure/tSdAgAABAkUK3GOP+3m+tcg1s9AECBAgQIAAAQIEjhf4Cz3L+3+QUrrL8bM3BQIECBAgQIAAAQIEegoo73vCeRkBAgQIECAwr8CrW05I/o8ppTPmXTxzJ0CAAAECBAgQIDCbwH1SSnH5++1P1O/z/QtnW2IzJkCAAAECBAgQIEAgBJT3ckCAAAECBAgUK/Bf15yQ/G5K6VnFrpEFJ0CAAAECBAgQIDCMwI/UjJXbCvxfHGa2pkKAAAECBAgQIECAwBECyvsj8LyUAAECBAgQmFfg8SmlP9k5KfkL8y6SuRMgQIAAAQIECBDIQiAufx+XwW8r7KvH/m1K6RFZLLWFIECAAAECBAgQILBuAeX9ure/tSdAgAABAsULfG3rhOQfpZQeWvwaWQECBAgQIECAAAECwwi8dGusXBX1dV+/OMzsTIUAAQIECBAgQIAAgSMFlPdHAno5AQIECBAgMK9A3M/z//73JyVvmXdRzJ0AAQIECBAgQIBAdgK/1lHg//OU0jnZLbUFIkCAAAECBAgQILBOAeX9Ore7tSZAgAABAosS+GBK6X9PKZ21qLWyMgQIECBAgAABAgSOF3h0SumPWwr8a4+fhSkQIECAAAECBAgQIDCQgPJ+IEiTIUCAAAECBOYTuGtK6cr5Zm/OBAgQIECAAAECBLIW+PGG8v63Ukp3ynrJLRwBAgQIECBAgACBdQko79e1va0tAQIECBAgQIAAAQIECBAgQIDAygTumVL63Z0C/7sppUtX5mB1CRAgQIAAAQIECOQuoLzPfQtZPgIECBAgQIAAAQIECBAgQIAAAQJHCrxjp7z/2SOn5+UECBAgQIAAAQIECAwvoLwf3tQUCRDoEPjvU0r+ZyADMiADMiADMlBCBq7vGNd4mEApAp8yBncMIgMykFL6g39f4H8npfQ/yYRMyIAMZJqBXy5lgGU5CRAgQIDACALK+xFQTZIAgXaBzc67/f07JQYMZEAGZEAGZCDPDHyyfVjjUQLFCHzDGNyYWwZkQAZkQAZkoJAM/HYxIywLSoAAAQIEhhdQ3g9vaooECHQIKCfyLCdsF9tFBmRABmRABk5nQHnfMbDzcDECyvvTv9/2eUxkQAZkQAZkIM8MKO+LGWJaUAIECBAYQUB5PwKqSRIg0C7gwCjPAyPbxXaRARmQARmQgdMZUN63j+s8Wo6A8v7077d9HhMZkAEZkAEZyDMDyvtyxpiWlAABAgSGF1DeD29qigQIdAg4MMrzwMh2sV1kQAZkQAZk4HQGlPcdAzsPFyOgvD/9+22fx0QGZEAGZEAG8syA8r6YIaYFJUCAAIERBJT3I6CaJAEC7QIOjPI8MLJdbBcZkAEZkAEZOJ0B5X37uM6j5Qgo70//ftvnMZEBGZABGZCBPDOgvC9njGlJCRAgQGB4AeX98KamSIBAh4ADozwPjGwX20UGZEAGZEAGTmdAed8xsPNwMQLK+9O/3/Z5TGRABmRABmQgzwwo74sZYlpQAgQIEBhBQHk/AqpJEiDQLuDAKM8DI9vFdpEBGZABGZCB0xlQ3reP6zxajoDy/vTvt30eExmQARmQARnIMwPK+3LGmJaUAAECBIYXUN4Pb2qKBAh0CLQeGD31aS/fvOglN/ifgQzIgAzIgAzIwOgZOPfc+7aOS1JKyvuOgZ2HixFoLe/PP/8Ro/++GeM7xpEBGZABGZABGYgMPO7xV3SNwZX3xQwxLSgBAgQIjCCgvB8B1SQJEGgXaB2gv+0dP7n58o/9Hf8zkAEZkAEZkAEZGD0DD37IY1vHJcr79kGdR4sSaC3vL3nSC0b/fTPGd4wjAzIgAzIgAzIQGbj6VR/uGoMr74saZlpYAgQIEBhYQHk/MKjJESDQLdA6QFfeO5BzMC8DMiADMiADU2VAed89cPOMxQgo770hyhs0ZEAGZEAGZCCLDCjvFzO+tCIECBAgMI6A8n4cV1MlQKBFQHnvYDGLg8WpiiHzUULKgAzIQL4ZUN63jNg8tDQB5b0xuDG4DMiADMiADGSRAeX90oaZ1ocAAQIEBhZQ3g8ManIECHQLKO8dLGZxsKhMy7dMs21sGxmQgakyoLzvHrh5xmIElPfG4MbgMiADMiADMpBFBpT3ixlfWhECBAgQGEdAeT+Oq6kSINAioLx3sJjFweJUxZD5KCFlQAZkIN8MKO9bRmweWpqA8t4Y3BhcBmRABmRABrLIgPJ+acNM60OAAAECAwso7wcGNTkCBLoFlPcOFrM4WFSm5Vum2Ta2jQzIwFQZUN53D9w8YzECyntjcGNwGZABGZABGcgiA8r7xYwvrQgBAgQIjCOgvB/H1VQJEGgRUN47WMziYHGqYsh8lJAyIAMykG8GlPctIzYPLU1AeW8MbgwuAzIgAzIgA1lkQHm/tGGm9SFAgACBgQWU9wODmhwBAt0CynsHi1kcLCrT8i3TbBvbRgZkYKoMKO+7B26esRgB5b0xuDG4DMiADMiADGSRAeX9YsaXVoQAAQIExhFQ3o/jaqoECLQIKO8dLGZxsDhVMWQ+SkgZkAEZyDcDyvuWEZuHliagvDcGNwaXARmQARmQgSwyoLxf2jDT+hAgQIDAwALK+4FBTY4AgW4B5b2DxSwOFpVp+ZZpto1tIwMyMFUGlPfdAzfPWIyA8t4Y3BhcBmRABmRABrLIgPJ+MeNLK0KAAAEC4wgo78dxNVUCBFoElPcOFrM4WJyqGDIfJaQMyIAM5JsB5X3LiM1DSxNQ3huDG4PLgAzIgAzIQBYZUN4vbZhpfQgQIEBgYAHl/cCgJkeAQLeA8t7BYhYHi8q0fMs028a2kQEZmCoDyvvugZtnLEZAeW8MbgwuAzIgAzIgA1lkQHm/mPGlFSFAgACBcQSU9+O4mioBAi0CynsHi1kcLE5VDJmPElIGZEAG8s2A8r5lxOahpQko743BjcFlQAZkQAZkIIsMKO+XNsy0PgQIECAwsIDyfmBQkyNAoFtAee9gMYuDRWVavmWabWPbyIAMTJUB5X33wM0zFiOgvDcGNwaXARmQARmQgSwyoLxfzPjSihAgQIDAOALK+3FcTZUAgRYB5b2DxSwOFqcqhsxHCSkDMiAD+WZAed8yYvPQ0gSU98bgxuAyIAMyIAMykEUGlPdLG2ZaHwIECBAYWEB5PzCoyREg0C2gvHewmMXBojIt3zLNtrFtZEAGpsqA8r574OYZixFQ3huDG4PLgAzIgAzIQBYZUN4vZnxpRQgQIEBgHAHl/TiupkqAQIuA8t7BYhYHi1MVQ+ajhJQBGZCBfDOgvG8ZsXloaQLKe2NwY3AZkAEZkAEZyCIDyvulDTOtDwECBAgMLKC8HxjU5AgQ6BZQ3jtYzOJgUZmWb5lm29g2MiADU2VAed89cPOMxQgo743BjcFlQAZkQAZkIIsMKO8XM760IgQIECAwjoDyfhxXUyVAoEVAee9gMYuDxamKIfNRQsqADMhAvhlQ3reM2Dy0NAHlvTG4MbgMyIAMyIAMZJEB5f3ShpnWhwABAgQGFlDeDwxqcgQIdAso7x0sZnGwqEzLt0yzbWwbGZCBqTKgvO8euHnGYgSU98bgxuAyIAMyIAMykEUGlPeLGV9aEQIECBAYR0B5P46rqRIg0CKgvHewmMXB4lTFkPkoIWVABmQg3wwo71tGbB5amoDy3hjcGFwGZEAGZEAGssiA8n5pw0zrQ4AAAQIDCyjvBwY1OQIEugWU9w4WszhYVKblW6bZNraNDMjAVBlQ3ncP3DxjMQLKe2NwY3AZkAEZkAEZyCIDyvvFjC+tCAECBAiMI6C8H8fVVAkQaBFQ3jtYzOJgcapiyHyUkDIgAzKQbwaU9y0jNg8tTUB5bwxuDC4DMiADMiADWWRAeb+0Yab1IUCAAIGBBZT3A4OaHAEC3QLKeweLWRwsKtPyLdNsG9tGBmRgqgwo77sHbp6xGAHlvTG4MbgMyIAMyIAMZJEB5f1ixpdWhAABAgTGEVDej+NqqgQItAgo7x0sZnGwOFUxZD5KSBmQARnINwPK+5YRm4eWJqC8NwY3BpcBGZABGZCBLDKgvF/aMNP6ECBAgMDAAsr7gUFNjgCBbgHlvYPFLA4WlWn5lmm2jW0jAzIwVQaU990DN89YjIDy3hjcGFwGZEAGZEAGssiA8n4x40srQoAAAQLjCCjvx3E1VQIEWgSU9w4WszhYnKoYMh8lpAzIgAzkmwHlfcuIzUNLE1DeG4Mbg8uADMiADMhAFhlQ3i9tmGl9CBAgQGBgAeX9wKAmR4BAt4Dy3sFiFgeLyrR8yzTbxraRARmYKgPK++6Bm2csRkB5bwxuDC4DMiADMiADWWRAeb+Y8aUVIUCAAIFxBJT347iaKgECLQLKeweLWRwsTlUMmY8SUgZkQAbyzYDyvmXE5qGlCSjvjcGNwWVABmRABmQgiwwo75c2zLQ+BAgQIDCwgPJ+YFCTI0CgW0B572Axi4NFZVq+ZZptY9vIgAxMlQHlfffAzTMWI6C8NwY3BpcBGZABGZCBLDKgvF/M+NKKECBAgMA4Asr7cVxNlQCBFgHlvYPFLA4WpyqGzEcJKQMyIAP5ZkB53zJi89DSBJT3xuDG4DIgAzIgAzKQRQaU90sbZlofAgQIEBhYQHk/MKjJESDQLaC8d7CYxcGiMi3fMs22sW1kQAamyoDyvnvg5hmLEVDeG4Mbg8uADMiADMhAFhlQ3i9mfGlFCBAgQGAcAeX9OK6mSoBAi4Dy3sFiFgeLUxVD5qOElAEZkIF8M6C8bxmxeWhpAsp7Y3BjcBmQARmQARnIIgPK+6UNM60PAQIECAwsoLwfGNTkCBDoFlDeO1jM4mBRmZZvmWbb2DYyIANTZUB53z1w84zFCCjvjcGNwWVABmRABmQgiwwo7xczvrQiBAgQIDCOgPJ+HFdTJUCgRUB572Axi4PFqYoh81FCyoAMyEC+GVDet4zYPLQ0AeW9MbgxuAzIgAzIgAxkkQHl/dKGmdaHAAECBAYWUN4PDGpyBAh0CyjvHSxmcbCoTMu3TLNtbBsZkIGpMqC87x64ecZiBJT3xuDG4DIgAzIgAzKQRQaU94sZX1oRAgQIEBhHQHk/jqupEiDQIqC8d7CYxcHiVMWQ+SghZUAGZCDfDCjvW0ZsHlqagPLeGNwYXAZkQAZkQAayyIDyfmnDTOtDgAABAgMLKO8HBjU5AgS6BZT3DhazOFhUpuVbptk2to0MyMBUGVDedw/cPGMxAsp7Y3BjcBmQARmQARnIIgPK+8WML60IAQIECIwjoLwfx9VUCRBoEVDeO1jM4mBxqmLIfJSQMiADMpBvBpT3LSM2Dy1NQHlvDG4MLgMyIAMyIANZZEB5v7RhpvUhQIAAgYEFlPcDg5ocAQLdAsp7B4tZHCwq0/It02wb20YGZGCqDCjvuwdunrEYAeW9MbgxuAzIgAzIgAxkkQHl/WLGl1aEAAECBMYRUN6P42qqBAi0CCjvHSxmcbA4VTFkPkpIGZABGcg3A8r7lhGbh5YmoLw3BjcGlwEZkAEZkIEsMqC8X9ow0/oQIECAwMACyvuBQU2OAIFuAeW9g8UsDhaVafmWabaNbSMDMjBVBpT33QM3z1iMgPLeGNwYXAZkQAZkQAayyIDyfjHjSytCgAABAuMIKO/HcTVVAgRaBJT3DhazOFicqhgyHyWkDMiADOSbAeV9y4jNQ0sTUN4bgxuDy4AMyIAMyEAWGVDeL22YaX0IECBAYGAB5f3AoCZHgEC3gPLewWIWB4vKtHzLNNvGtpEBGZgqA8r77oGbZyxGQHlvDG4MLgMyIAMyIANZZEB5v5jxpRUhQIAAgXEElPfjuJoqAQItAsp7B4tZHCxOVQyZjxJSBmRABvLNgPK+ZcTmoaUJKO+NwY3BZUAGZEAGZCCLDCjvlzbMtD4ECBAgMLCA8n5gUJMjQKBbQHnvYDGLg0VlWr5lmm1j28iADEyVAeV998DNMxYjoLw3BjcGlwEZkAEZkIEsMqC8X8z40ooQIECAwDgCyvtxXE2VAIEWAeW9g8UsDhanKobMRwkpAzIgA/lmQHnfMmLz0NIElPfG4MbgMiADMiADMpBFBpT3SxtmWh8CBAgQGFhAeT8wqMkRINAtoLx3sJjFwaIyLd8yzbaxbWRABqbKgPK+e+DmGYsRUN4bgxuDy4AMyIAMyEAWGVDeL2Z8aUUIECBAYBwB5f04rqZKgECLgPLewWIWB4tTFUPmo4SUARmQgXwzoLxvGbF5aGkCyntjcGNwGZABGZABGcgiA8r7pQ0zrQ8BAgQIDCygvB8Y1OQIEOgWUN47WMziYFGZlm+ZGWwYagAAIABJREFUZtvYNjIgA1NlQHnfPXDzjMUIKO+NwY3BZUAGZEAGZCCLDCjvFzO+tCIECBAgMI6A8n4cV1MlQKBFQHnvYDGLg8WpiiHzUULKgAzIQL4ZUN63jNg8tDQB5b0xuDG4DMiADMiADGSRAeX90oaZ1ocAAQIEBhZQ3g8ManIECHQLKO8dLGZxsKhMy7dMs21sGxmQgakyoLzvHrh5xmIElPfG4MbgMiADMiADMpBFBpT3ixlfWhECBAgQGEdAeT+Oq6kSINAioLx3sJjFweJUxZD5KCFlQAZkIN8MKO9bRmweWpqA8t4Y3BhcBmRABmRABrLIgPJ+acNM60OAAAECAwso7wcGNTkCBLoFlPcOFrM4WFSm5Vum2Ta2jQzIwFQZUN53D9w8YzECyntjcGNwGZABGZABGcgiA8r7xYwvrQgBAgQIjCOgvB/H1VQJEGgRUN47WMziYHGqYsh8lJAyIAMykG8GlPctIzYPLU1AeW8MbgwuAzIgAzIgA1lkQHm/tGGm9SFAgACBgQWU9wODmhwBAt0CynsHi1kcLCrT8i3TbBvbRgZkYKoMKO+7B26esRgB5b0xuDG4DMiADMiADGSRAeX9YsaXVoQAAQIExhFQ3o/jaqoECLQIKO8dLGZxsDhVMWQ+SkgZkAEZyDcDyvuWEZuHliagvDcGNwaXARmQARmQgSwyoLxf2jDT+hAgQIDAwALK+4FBTY4AgW4B5b2DxSwOFpVp+ZZpto1tIwMyMFUGlPfdAzfPWIyA8t4Y3BhcBmRABmRABrLIgPJ+MeNLK0KAAAEC4wgo78dxNVUCBFoElPcOFrM4WJyqGDIfJaQMyIAM5JsB5X3LiM1DSxNQ3huDG4PLgAzIgAzIQBYZUN4vbZhpfQgQIEBgYAHl/cCgJkeAQLeA8t7BYhYHi8q0fMs028a2kQEZmCoDyvvugZtnLEZAeW8MbgwuAzIgAzIgA1lkQHm/mPGlFSFAgACBcQSU9+O4mioBAi0CynsHi1kcLE5VDJmPElIGZEAG8s2A8r5lxOahpQko743BjcFlQAZkQAZkIIsMKO+XNsy0PgQIECAwsIDyfmBQkyNAoFtAee9gMYuDRWVavmWabWPbyIAMTJUB5X33wM0zFiOgvDcGNwaXARmQARmQgSwyoLxfzPjSihAgQIDAOALK+3FcTZUAgRYB5b2DxSwOFqcqhsxHCSkDMiAD+WZAed8yYvPQ0gSU98bgxuAyIAMyIAMykEUGlPdLG2ZaHwIECBAYWEB5PzCoyREg0C2gvHewmMXBojIt3zLNtrFtZEAGpsqA8r574LayZ1yaUnrwQtdZeW8MbgwuAzIgAzIgA1lkQHm/0NGm1SJAgACBoQSU90NJmg4BAnsLKO8dLGZxsDhVMWQ+SkgZkAEZyDcDyvu9x29reeL7U0p/lFL6SkrpXgtbaeW9MbgxuAzIgAzIgAxkkQHl/cJGmVaHAAECBIYWUN4PLWp6BAh0CijvHSxmcbCoTMu3TLNtbBsZkIGpMqC87xy3re0J70kpVWPVf5FS+kBK6S4LQVDeG4Mbg8uADMiADMhAFhlQ3i9kdGk1CBAgQGAsAeX9WLKmS4BAo0B1QrT269ve8ZNZHEhMVRqYj4JKBmRABmRABubLgPK+cby21gdu2Crvq7Hq/5JSes0CQJT3ChvHWTIgAzIgAzKQRQaU9wsYWVoFAgQIEBhTQHk/pq5pEyBQK1CdCK39qryfr8BQHrGXARmQARlYWwaU97VjtTX/8K015X01Zv1vUkpPKhhHea+wyaKwWdvfGetrbCUDMiADpzOgvC94RGnRCRAgQGAKAeX9FMrmQYDAHQSqE6C1X5X3pw9qHOgxkQEZkAEZkIFxMqC8v8MYzT9SenNLeR9j1++klL6VUrqgQCzlvfJeeS8DMiADMiADWWRAeV/gSNIiEyBAgMCUAsr7KbXNiwCBE4Ha0r46Uaq8H6ecUPpwlQEZkAEZkIHTGVDeG53uCLy+GpN2fP2D//91t6aUzt55fc7/VN4rbLIobPwtOv23iAkTGZCBtWVAeZ/zkNGyESBAgEAGAsr7DDaCRSCwNgHlvROHThzKgAzIgAzIQBYZ+P/Ye9fgPar7zrNB3MRFEmAuwiAJgUA3EEiALghdECAkJCEkkARCF2QQQiAECBBIAkmAADuxNySOHTuJM/Y4dmwndhw7dhJi57JTW1uzs/tiq2a3ditTWzuZmX21qa2ZF1s1U7O9/j72EUf9nNP37v/pfj4v/tX99NN9+pzP+T79/53+ngvm/aiFoZnl3ZJh2ifj2H8fRdHeKIrGZaY89idg3vPcDeK5O2oGFeXFlEUDaAANDGsA837sA0NyAAEIQAACQRPAvA+6esgcBPpJIPnS84zPVUbeb9l2LL5r4YbSf4sWb4yXr9ger17zbLxx06vxvue/HJ987694ycWLTjSABtAAGkADPdUA5n0/g80KpdpU0Lw3cez/EkXRQxXu28almPc9fY5hCg2bQjCBCRpAA2ggbA1g3rcR+nEPCEAAAhDoMAHM+w5XHlmHQFcJmJeczm0V8/7OheudaZZ8CTtI66yzzo4nT74xXrP2ufiNo9/HvOGlJxpAA2gADaCBHmkA876r4WRj+V5fJW6MougnURTNaSx31RLGvO/RswtTKmxTivqhftAAGkAD6RrAvK8W1HE1BCAAAQj0ngDmfe+rmAJCIDwCqQZ7aOa9/QL37LPPjhctfiQ+fvKnGDcNvfw8+vYP4/HjLxn6u/e+3TBviDkvFdJfKsCn/3x27v5g6Jmj59DuPZ/juTMCzx3M+/ACxTHO0Vo79iu5/1+jKPpGFEVXj3FZkrfHvB+BZxpxS//jFuqYOkYDaKAPGsC8T4ZpfIYABCAAAQicQQDz/gwcfIAABNog0Fnz3rzAvfjiS+N9+7+EqdPAC9Ajx/7UqY97lj8O7wZ496HRTxl4eVVVA9t3vOt87uzY9T7PnRF47mDetxH6deoe95t4r4btf4yi6GgURRcGQgDzfgSeaVX/J3I9cRUaQANoAA20oQHM+0CiQ7IBAQhAAAKhEsC8D7VmyBcEekzAaZKYF6Qhj7w3edT2vPPGx888+xHGTs0vQTHveVHQxosC7oHObA1g3o+2HjDvexxxlivavXa8V9P+v4uiaG8URePKZam2qzDva45b7f8l7I/2/xLqn/pHA2gADRTTAOZ9bfEdCUEAAhCAQD8JYN73s14pFQSCJtC6eT9t2q3xg2v3Zf7dd/+eWCO8Z8+5J5506dWp+dTL3AsvnBgfO/4jDPwaX4Ri3hdr8PKCAF5ooLoGMO+rM+yyDjHvg44ZxyJzy2oy7F1x5P8YRdHKsSjUr+6JeV9jzNrl594o511L4ixfsX3o751Tf02brsO/j21PnBiq03tX7aJOO1ynXXhOnfrw74Z0p+dLX2bveu6F33GW782j36/lt4V5P4YRIbeGAAQgAIEuEMC870ItkUcI9IyA62Xm6WNNjLxfes+WUo2LAwe/Ft8+f3V81llnnc5f8oXugjvWlEq7C43Rscgj5v1om2hjoTnuieYw70dbA5j3PYsyqxdnSTLWa+Dzx1EU3VI9q4VTwLzHyBr5dos6art+02+f/MnIs+lyTHzb7fcP1eu5555PnfLMa1QD733w8yHd6fmyaPHGRu/b1m/14UdecZbvhRd/r5byYd4XjuO4AAIQgAAERosA5v1o1TelhUAQBJwNAPMSJSTz3jSKdu7+ID7nnHOd+T777HHxWyf+vJbGi7nfKG8x70fbRBtl7VP2sdM+5v3YsQ9B95j3QcSGIWXiLhOTNrz9L1EUfSWKoqtaLHxr5v2SuzfHdy3cUPhPHW5X3bc7XvvQ/vjRLW/G+1/4SsyI6NF+Rtf9fwLzvp96wrzvZ73W/fuvOz3M+2q6w7xvMQLkVhCAAAQg0EUCmPddrDXyDIGOE3Ca4OYFaYjmvRp5Gza+7M33Y1uPYN7X1Ksf875aA7juFxKkR32MggYw70db55j3HY8q68/+fBOTtrT9T78owokoisbXX5ShFFsz78877wJv3FyUqzrKXnX19FjLWx1+83vE3DXF3KPw/91VRsz7fv7Px7zvZ726fsMhHcO8r6Y7zPuhOI0DEIAABCAAAZsA5r1Ng30IQKAVAqkv80I177We2cRJVzrzfufC9bxIrOlFIuZ9tQZwSC8zyAt12RUNYN6PtlYx71uJ/bp0k1uLmss1nf+PURTtjaLo7AZhddK8t/meddbZ8cxZS+JXXv1mL2Pv5w98NR4//pKhv82PHu5lecciTsC87+f/fMz7ftbrWDwjitwT876a7jDvG4z4SBoCEIAABPpAAPO+D7VIGSDQMQJOA9y8mAvVvFcjbuGijc68z5p9dysv1NSB4N33f9bKvbIare9/9u/jE+/+Ze15GUXz/p1TH8fvf/bvameZVYeu79/74G9i/bm+C+2YNFj3VLqnPvzbWC9hQiurKz96FjShG5VfzxrXPcfq2Mn3Pm40PyGb902XvY461TOsjnRMGife+Yta0zPp+raY9x2LIpvP7mwTk47R9n+Iomh5Q8XsvHlv6uScc86L+2hoP/f8l51tDc0A5nuGcbyYeYR5X4xXV/SFed/Peg1df5j31XSHed9QtEeyEIAABCDQFwKY932pScoBgQ4RcL6UMi/jQjbvtf6myae9vW7K7NpfqD3/4u/G969+Or555uL48suvjceNO/f0vc87b3x8+eWfjqdNuzXW2qB7nvlCY2bbu+//PN6x64NYL7punHFHPGHiFfH55194Oi/icM4558YXX3xZLANk7i0r4ntX7Yp37v4wPvr2n2VykTH1wou/d/rvqac/f0bahvPt8x84fY59vm+/TQP25Ve/OZS3lw59w1l2Tfe6dt3z8c0zF8WTJl0Vn3XWWafLe8EFF8VTps6NV6x8Mj5w8GvO6+t4gSGD+rkXfmeQjzvuWje458UXXxqfe+75p/MijSl/1147M168ZHP8+PaT8dG3f1hrnl4+9M+HuImlq4zHT/40fmzr0fi22x+Ir558wxl51Qv8K6+aFr9x9PvOa5PpyZjW72vdhhdjzZqh35H0e+65n0zxq+l+NdPGNZ++adBpZ+vjb8dvHvtBrvST9/N9fuW1Pxwq/0uvfN15j7dO/Dh+ZPNr8Zy5y+LLLrsmPvvss0/X1YUXToin33B7fP8Dn4nfOPInzut9eTj53l/F2544MeB65ZVTY7HUb04jGy+55LJYzzalq7rypVHn8SNv/TDeuOnV+Pb5q+PJk2+Mz7/gotPl1DNQz5/J18wYfL/5sTfit0/+pFS+pGX72bFm7XOn72OeOdrqt2qfl7Z/0FN3efm8efT78cOPHDqtcfs5q2esyn7Np2fE8xesGfwWjtdkcKvDlKtcro5i6izy9N7fiJcu2xpPnXZLfNFFk05zkyb1zFC95CmzOgnteupz8fIV2+MZN901+L3Z9S3+qnM9mz597c3xnLnL4xUrdwz+HynPee6R9xzM+w5FkO1k9Sb7OTCG+x9HUTSn5iL3xrw39bJu/YFanwd5nxtNnYd5X80IylMvmPfNM85TD3Wfg3nfz3qtWyd1p4d5X013mPc1R3kkBwEIQAACfSOAed+3GqU8EOgAgdMv+82LN3sbsnnva1xcP31eLS8OZZTLuLriiimpjGxeZn/ixCtivcCsy7R+68Sfx3cvfSyWMWjuUXQrY1rmnwww3yj9/S98pXT6afnJa+TW0YCXwZvMi4xQO20ZvzKetW5r8lzf5+k3zK/NxJf5/eiWNwedBmTM++6Zdlzm3Ow5SwdGn122svtXXXX9UD5kTtrpyaC8Z9m2oQ4jrnzKDLevtfelvy3b3opnzV56hiHsSsd3TIb2TTcvjPc9/2Xvfex7Zu1PnTp3qPwy5u3rZK7rJXPetYOlL/1ufb83k7aeNavXPJv7963f8h13PhQfO/7jM/Jn0qu6ffa5L8YzZy0+o1OCrx7s4zK177xrXXz4yB8XypeeSXY6dexfeunVhfJgmD2996P4ppvuGnSYKJIPdbRYtHhjLNPfpFVmq84ZrvvavyfNcqHfz6euuM55rn291qROy4dG1S9b8cQZxr99fd59PXelYT3b0u6X5zvM+w5Eju1mcXpeHbZw3n+OougrURRdWROCMTfv1SHwwbX7vH+r7tsd333PlsH/a3Uiy2Ks/83qVJTnt96FczDvqxlBeepYs+3ctXDD0F/ds8jkyQvn1FffmPf1sUSX+Vli3udn5dKV7/2a9b//H2qKf0gGAhCAAAQg0EUCmPddrDXyDIGOE0h9EReyea+R8FZD4vS+RsO6GiNFjj3z7G/mMkZc97ePaQTyiy9VG7Wtl1r2iEo7/bL7Bw7+vpPRKJj3u576bKxR9WXYyYzVqOAiWkqeqxHmMjnL3N91jYzcJXdvrryEQ5Z5L5N8woRP5c63b9S+RgLbswq4ylT0mGYsqDqdeZZ5r1H4l142OXf57TJodgJfB5ZDr31rMHrdPj/vvjoJvXr4jyrp0danRsDfcuvKUmW086zODRq1bqedth+CeS/TXZ1h7HKU2ddo9byj3V1Mssx7mePqtJI3b6vuf8pbDxppf8mEy3Onleee+t/pKleRY5j3HY8q68/+lDzaa/mcf4qi6HAURRdULO6Ym/da073I73Pvvt8adDxM460Zqrqy5E9W2THvqxlBWXz5vr98Me/7W7ch/24x76vpDvO+YlTH5RCAAAQg0HcCmPd9r2HKB4EACaS+uA/ZvJ895x5n3jVKqEqjUqMH7Wmw015Q5vlO03/LMC6Tp02bXy88AjRPnkbVvNd083XUrUbYavRrmTrVSOw8dVT0nOun35Y5wjstv2nm/VOf+fXTU7jnzZdvanf9PvOmUeS8666bFWs6+7Qypn2XZt5rGvMLxl9cKd/qyJOcWv3gy/9sMA15kXImz9XU6L6OAWnlTX6npRuKdM5I5sP1WSPpNLV78l7Jz2Nt3u/d98XK9ZAsv37nZZ4Raea99O36nSbvbX/2/T/c9sTxQjOP2Gmm7WPeBxjldT9L16Rpboy/+7dRFO2Mouiskpg7Z96b5/f6h186Y6mhZD1odhBzbpe3mPfVjKAu1z15r1b3mPfV+KG/cvww78txM3rDvC8ZzXEZBCAAAQiMCgHM+1GpacoJgYAIpBpSoZr3Wr/dN3V1lTz71l20X0qOG3dOfMWVUwdrDGvdeU0tnjWdvdYKLpqvffu/lPpiVHkaP/6Swdr2muJaI2bn3rI8NnlKrldsl2EUzXuNLjNriNsstK/pp2+4cf6An9Zc13rWyXOSn+9dtavUi+ks817rvV919fTB2u+qS61lLcNOukvmIfn55pmLc5mlpoFub12moLStEef2et/Je+qzayaBlw59w8kny7zXLBPKi8qt8qs+rr56+mC9bde97WNaMqPsaD+feS9j3DdVsFn3XaMM85j7WqrBMFe6qms7/2ZfHUx0T6WbZ9YNLT9g0i2z1UjKPMs36HehUdEDXU6dG2tqepNn31YdXbLyNJbm/Z6nP59rJoiJk66M1UFEZZ8yde5gPXlfmc1xLTGRVfbk9z7zXr8n/RZM2q6tdJN8TrieUy++/AeZxr1mJ9Ha9nqm/PJ/y4pB2XUsbeYSzPuAorv+ZEVT1KdqP4Dv/2UURfeUQN5Z817PLt/zSvWhDrbJ51sXP2PeVzOCuljn5LmeOse8r4cjeizGEfO+GK+kvjDvS0RyXAIBCEAAAqNEAPN+lGqbskIgEAKpL0SLGs52A+DOheudaS+9Z0ulF3oazTjvtlXOtCddenVp83LtQ/udaeolpEwR3XPPM19wTs+t0aUaubrgjjXekfIy+F9/83u5yq70ZFi6XkjLSF2x8slYBkzayE599/ob34012vz2+avjCy+ceDo9n3mvtbdfe+M7p/9UJlcetK61fV7W/qkP/zZXuW39lN13rXmvtcs18tkuiwznlffuiF89/O2hvImdXti6XjyZNDRd/e49nxu6NivfSfN+8uQbY01tLW2pU4rv+ndO/XWsEcJa6zxt2vn1Gw560/ClreMu814alFlnymy26ujw8COvxJoaX+vAm3S11vnj20/EM2ctifOa96qXlat2xhrdf+TYn55Oy6RpttKmpu5fuGhjqtEsM8FcU2TrMu81Tb7MWlNubWVya41wLTGRHFWuKfBllvrqR5p5dv9vx/o9XDdl9hnpqiPAosWPxFpzPdkBQUa/DO60DgI7d39YqtyaISHNjJU+H93ypnct98Nvfi/Ws/Oiiz55vti8tP/I5tdS86YZCexnyIaNL5/BxqS3cdOrZ5xnX5Pclxaz6l9Lmvg6geme0v6Wbce8utQ9NVOLOlGZPCa3RUef+sww/W9Jpi196rej34WeHeb/gXg+8+xHsf7Xupb5UOeDZFr6LBZLl22N9f8hqe0kS/F9cud7g+fRxRdfejo9zPtAIrt+ZeMyl14DPfajKIpuKIC/0+a9/lcpvnLVhf6vtBn7JZ9RdX3GvK9mBNVVD6TTvXpwtaEUH1OX3avLLtUZ5n01fWHeF4jgOBUCEIAABEaRAOb9KNY6ZYbAGBNwvnQzL+JCM+9lSsxfMGximPwWNUpMY1SGmm86dY0+fuW1P8z9suH5F3/XO3J7xow7c6Wj/Jgy2dsrr5w6MORNvots9RJVxuqUKXMG5kyea2Wm2vc3+5qhIM/1Y3GOy7w3+TZbGcY+czmZ5527P4h9sxhovegT7/5lIRYy79UBY9nyx0uvV651zjUy3ZTH3srgTU7PniyT67PLvJfZbKetkeDqZOC6Pu8xjbzXUhJLlj46MP/zXmef98aRPxky1U0+lbbWbrfPz7PvMu+T5dda428e+0Fm2jLEbUPT5E1brav+0LoXzuAqM9XViSSZb63Lrjqw0zP76lCRPD/rszqEXHHFFGd60ujWx986bQhnpaW12DUDgMmPvVWHB5n8WWmY77fveNeZzo5d7+dOw6Tl26rTicxvO59mX78h5cF3bfL4Wyf+PL7ppru8aaV1Skmm5TPvbS2qo8eDa/fF6tCSvD7rs5ZqMOW0tzLgivyfs+8jo//JnadizXyBeT9mEd34KIoutf401fz0xN+cKIoWWH9Loyi6z/pbH0XRY9afpoPfm/g7+Ku13rXeu/5O/GLd9w+tv4+iKPqK9feNKIq+m/iTwf2x9fcvoij6V9bfv46i6N9Yf/+HrdUO7P/nX5X/ihxq6LR5r+fAfffvcT5TVE8HDn6t8DPKfrbk2TedlvKcW+acrpj3TXBoIs0yddDWNSGWN7Q8FclPH837IuUvotum0i2SB/vcrA6c9rlt7edlNNbmfd58luWmzvOuOEhLvJVN074O8z5H5MYpEIAABCAwygQw70e59ik7BMaIgLMBYBoFIZj3Mlk0slBmp88QU37tKantRkjWvhp5mjbdlNneymh79/2fFW4MaVTiRM/U67v3/Fpmeq7p+8866+zBFOZZ5cnzfd5GeR/Ne60RLhM0Dydzjn4HMsxsbZh9jXw15+XZasT22yd/UugaV7oy7mZ4DEONhHZdk3bMZd6bMmqrqfyPvFXcFE/eUy8Xjh3/UeH8JdNRZ5Q5c5c76+Te+3YXTt9l3tvl19ThyRHxyTzZnzVLgn292VcnIXuKeo3sP/nex7nzq9H9Li3q+ZBntLmdxxUrdzjzqOn8NauCfW6e/VMf/p13tor5Cx7MnV4b5n1yBgxTP1oaQJ1j8pTXPkfamDN3mZOnZouwz03b95n3Jn9a+mPXU8Vn/DD3vO+BzzjzqOeSOafKNu//lrR7aGkGU17P9ogjXuqzef0foij6p8Tff/WwyWLH9+1Nwf9//6qDw/kOvZpDnTfvteyKT4tlZifyPRvUKVEdyhYuengwc429nI2ei4q51XlTbQHN9nLs+I9LPdPUGUtxivl7dMsRZ/mWr9h++hxzrm/rm+3KlFUd/pLX6pj53t6qw+i2J04MZvpSee2ZX9TR9OrJN+SOBfS8VocuzXaj/1PTb5gfa4kYezYe7WuJHM0WtHjJ5sFsXnXEgnaZNLtQsvz6rJjCPi9tX7PPuNLwxVf6n6dZlLQcjb08kXiqU6Nivs2PvZE6I1Zafop+l0/fMyx9V4+js/KomFIdBdXWsJcTO/vscYPPil8fWP2MN14M0bxXbJvUia8zt54Fmx49HM+9ZUX8qU9de3rpNcXbeh8hLprdqEgHTTGXJs2seGpbmaXJFNuL87Tr5w24H3p9eGa4rDor+r10pxmm7rhr3WC2qeQygOrMqkEMGjihWbjEpOg9ip4v81uzRy25e/MgT4aP/s+og7Y64Gs2xMe2HolPvPMXQ/lpy7zX7Iaa7e7WeasGz127o7/aeZdf/unBMlt6bqrzcZkOty52mPcmfGILAQhAAAIQGBMCmPdjgp2bQmC0CThfSpkXcU2Y9zKGNB1x1p9GmOrlnD3i0OTL3ur7xUs2lZ6e8+FHDjkZXHvtzPidU/lNtWQD65ej+ccNpX399NuGGprJa2fNvnvoOo20Tp7X9Oe+mfd64aKZEcpwS46WNhrUi42io+/L3N91jUY72y8dTZ401bnr/LRjaea9XkhotHva9WPxnX6f6oxhym22GlFdND9p5r1mvCjzLPCNRDf5nDDhU/FbJ4obDHremTTs7YaNL+Uut2YQcE3vLxNk/4Gv5k4nyVkvJZNLVCiP6rSQV0NNm/d64ebqAKGXbZpKP1mmvJ/1EtE1M4LulbYkhp1+lnmftQSBnZZr3/UyXWaP69yxOpbDvHfq3/4tsB/8GvGjVIf/exRFmzxNjc6b93qe+n5vj209WvnZInNZhryWOvLdx3V83Lhz4tvnPxCrw1uRZ5mMcVd6VY4pL2l50DM4mb6WLbKvUUdmLfVkG+vJa8xnnxmp9BTLaBauubcsj5NGnbk+a6tYeuasxYPlwuw8lt1wrk1WAAAgAElEQVR3dVhWHop0dF234cUhhkojGc/s2/+lOC3eS5ZdcZI66TYV5x985esDnSr2St477bMMdP0uys6Yk1ZXitWUtm9GumS+1A6/eebiIRPfFW+M9bT5atsn86/OGjYP/dY0YCBtWSU7Df2+1Zknq50g81adIfL+7vQ703JaauvZ+atjX7N9Lbhj7ekOCXZ50vYVz0obZTr45sm3OnypQ0NaHuzv9DxUh23NJGbSb9K8V8eCHbs+GHT6yXo/ZedT+6p3zT7n6nBg8p5ni3nviaY4DAEIQAACEGiHAOZ9O5y5CwQgYBFIbSA1Yd4nGzNlP5uXRxrhmqex4zpHo05c63Uqba0p77qmyDGNNk2WT429rJGdLvNCax4XuXcd5/bNvNfIgrJcNLJWJm6yPvV50+bXS6dbNj/mOo2yT+ZJGivy0lNppZn36uBi7hfaVuyT5dfnItO0q0xpL3P1srdMuX0mtMmvRt6USVcjxkwa9lYv4vKmt2Llk840NBo/bxq+8/Tyzc6X2dcILd819nEft7qmzdeSDSZP9vb+1U/nyp+d1+T+9h3vOdN+aP2BXGmnmffq+FV1Ok4tr2CXWfsycZLlGMvPrv9/yTzzGXO+gxr473+xVMDdVvyt3c6b94o1fHWhUYllnyUyS2VcKR73pZ/nuEz/1Wv25n52hmjeP3/gq86Oir7yv/TK153c1ZkiryHpS9s+rljzzrvWFZo9yKWHNsx7tfdksJbVk0ZeF53dyFVWc0z6VkfMsvkx9SB9P/Bgfn2b+/u2imHydBAx97e3yovdXuiiea/OFL4lleyyuvbVMcA3K4U6WUy+ZkapZ5nyU5f29DtQrFu0M1SyvOqwoE4tmgXNp6UixzXLodroyfvk/awZFF/51YxhTZn3qkPF4Xnz5DtPsyvsefrzpblh3ieiKD5CAAIQgAAE2iWAed8ub+4GAQj4GhbmeKjmvUbtP/WZXy80paKrEfn03o+cjbB5t91XulFl30cvAQxLe7tx06up6bvWbde0fXbabez3ybzXS0aNEKvCTS8+7Xo0+5o6skq6Va5VRxCTD3v7VMEXAz7z/qKLJtY21V+Vcvqu1Yhm18vPosa4z7y/7rpZpetWoyukO7tezL5mDCj70ksGrj1VrklTLwZ9nOzjmorWnnLYXK8XtlVHhOg+yp9LT9NvuD1X/po07zXyyTXqSTNYZI2ashn69vVi1DX6XqPSfNfYx9PM+yrT5Zt7uF48aupX830IW8x7jHnzTOrh9v+Loui7URRN/1ULpPPm/etvfs/5P051t/nRw6WeLTJJrvB0liyrCXVSkqmT9YwLzbx/eu9vDKaKLlJutT1c5bzv/j3euiqSfvJcxUlVptNu2rxXzHPrvHsrl11xWx0mqmaD8HUGTrLN+1nLSOXRt0sX5phmb3LF03nzYM4zHSG7Zt5rZjbNpmbKUWarOFwj9w1TbTUYwDVLWpH0FVcefbva0mWKcTVjRpH7Zp17w40LKrcbxEtLdmTdK+t7xfaaqaUJ816dWuzp+7PykvW9ZrVY/3D+2dJsPWHe8/4WAhCAAAQgMKYEMO/HFD83h8BoEkhtLIVq3ptGkdZh1FqNZdalV0NI66mZtOxtnnXp7YZU2v6kS68eusft81ef0bBPXu9ay1xmndamS57b5Oc+mfcyZ6uy0kgZrbdna0X7GslU1oitmidd7xolIxOwSNous1Vl0yjlIumMxbmuqfOXLttaKN8+877syxXDQevHJvWiz5qS05xTZuvKr3SQJ619z3/Zmaci69Jn3Ucj3JLl1m8nz8vlJs17/U9L5kufi6xLn1V218h+vRCWsZ91rc+81xILRdb/9d1HBlay/KqXY8ebXz/Xl6fkccx7zPukRnv4+T9FUbSgDyPvtUSUr37KzJaiUeNaXsuXpjkuM0xTy6tT2JQpc2JXrG3ONVs9/7KewyGZ9xpJ6orvTHl8W9/MYVnmvcwvjWDVM1idUrXOvaYU1xTtvnuZ49dPnxdrhqrk8zzP56bN+4WLHvbmX9PVT5x4xeAvz9T1VTu7aUmDpvQtAz9L37762PbEcW9nU1PHaodKG1oSSrP4qIONy+xXp1Wl1yXzXp0yfPWiMqrDq54xrjag4WO29ixvWoLg4osv9epPzzHNAJjHGFYHFF/9ZR1Xx1X9pk0efVs9b9Qe1HNV2zzPH7VHtGRWVh5c3yuuzdOhQIymXT8vnj3nnsFMab66UpvrzaPfd5Zz0eKNpfL42NYjuZaQUFtUgy/EWcvXZXXY0O+kzFJYmPej+cKWUkMAAhCAQDAEMO+DqQoyAoHRIeBs4JhGXejmvcmnGpi+l1WuxqI55hrZo3WP8xhMJo2srdbcNPk026w1uWXsmXPtrV6Y2Ou6Zd276vd9Mu/XPLS/VKM9ydC3jnkZ/SXTLvtZ03naOtH+XQs3FCqvz7zfufuDQumULUOV61xrWBZ9yeUyw8XxhRd/r1L59bIpWTf6XMbUsBlpzclkunrBmOfF7X0PfGboWqX15M5Tlcpq50+jxJP502dN+W+f59pv0rx3dSpQvorOVOHKtznmy79vNKS5TlufeV/XbDC+9PWyMTlazM5Xm/uY95j3rmdHj479fRRFt/6qmdH5kfeasttXN1lLRCWfK28e+8HARPWlp9hZa5ub6ZGT179x9PuD0Yzq7ORLQ///ktfZn/UcfO2N75z+8y2FonTs89L2s2Z9cq15f+VV0wadEpLlmDbt1gEDdXIwo901243usfXxtwZrj/vi0aR5r9G8y1Y8Ee/c/WHqiHK1O7REmabIT5t2v2inUcO9SfPe1ZlOMaNmQEsur6T4SUsU6BpNC55kbz6Ls8l7ka3aVOooYNJJbgf6Xn+gmr7v31M4b/o9pZnSGhWtmM7VOUO/Oa3jnjR5ZfTLAE6WMcQ179WGSi4pJJ3fs2xbrM6uyY6TiuXUTtf678ny6bNicelIvJLxjK6Riaz3K8l3DfoN63maZuTveeYLhetXGk3rwCITWW2KZ579aKgNod+EfvvzF6xJ7dyhjhpFfgvmXF97RBzFSpxf/tV0+OYabfXMU8exebetGqoDGfyueilj3muWR42Sd6WnY5rN6vHtJ2LNAmfnz+TxwMGvxUvv2eJdpkDPmaJLs2Hej85LWkoKAQhAAAJBEsC8D7JayBQE+k3A2yBRo6QJ814NPBlief727vutwQuDteuej2VeJF8O2I0pvRDQ9JLJxpPvs0ZRu0YMqKHtu6bM8ZX37hhirIZgmsmmBrRdNntfL3c2bjpUeZq6PGXpk3lfRBtpbJIvP03d1LHu/fGTPx28jFD9atpJGY164SIjPu3PNQ14UfPaZ95rBEMajzq/02/yuee/PBiJoPXRf1n+jallFxfXFPBFR0e5zHu90HK9rCxSZk2XbjRib7Ne5mfdQyPF7fTMfp5p730dUGQ+ZN037/e+JUM0qjErDZ/5XbXDg+7rG3nkW6c0K6+u7/cf+KqzbvJ0hPGZ6/of6LpX0WOaFtZoJbnVOpxao1prWBdNt87zky+7k/nkM+Z+RzXwf0VRtDOKorOsZkWnzXuZWq5OsKof/V+WwZL32aCY2NfZTQbHQ+sP5J7hSKNAZTa5dKK0Xnzpa7nzpZjElY5m/cpbtqzzXOa94g/7vppprOrSKYpfNbpc8UOezmSufMvw1lTZdt7MvtphZf6XNmne2xzVOVvGex5dqgOEb8S0lm5zsUk7Jn1rdgLDyt4O9L3uhUL6XnCHW9+aJUGGYVpe7O/EQjNX2Pkx+0pLOs/DSyPMfb9fk562IZr3tkaUx5tnLhqM3rY5ufbF2dX+Uhq3zlsVq9O4XfYpU+fGrx7+dmbdaCkS30wieZdgsvO7e4+7M63ydsmEy3O/69G7CcWJdpns/TzxvZ0vPYN8s3poBhBfJyQ7De0rtk57P2TyWNS8V/vXt4yCOohp6cVkXnyf1eZzPeeVN3UeKTIwA/PeiqDYhQAEIAABCLRPAPO+febcEQIjT8DbCFODognzXj2QfY2brONq3GiKMV9jSj3l846UPXDw951lnz1n6WA0jRqNeToYZJ2j8pqGo73NMkiyDAy97FGPb/XSlyGT1hkgi6vv+z6Z98eO/7i07mw+O3Z94KxPvXy0z8u7r5cX6uDhGj1v66Xo/owZdxbKj8u814vOvOUoe55G/Ky6/6lYo8ySL9CKltk+v+ha9S7zXi+DypbLXHfLrSudeqm6BIZGBNnlNfv6zZp7+7Z6KWbON1s9Tw69/u3BCJesZ1qe7zWS3aRtb/XSyZcvc7xJ89416k3/N9Rx4eVD/7yWZ77vGfHoliOZZfeZ91VNG8NW26y1RfUydeq0W2KNiHruhd8ZGnVmp9XEftb/PltP7GPkd0AD/yWKov/mF38THS2OTpv3WlbGx79orO8zJPR81gjLos8aGY6uUdfKb5H4aKzMe5urpq6voyOlRgOXMdeT7LVMlKZot/No9leu2lm4rpo0702+LrpoYu72oSmv2la+Ubd5jUWT1sOPHHLyGuj7uS8WZiZ93730MWea6qRo7pu1fXz7SWcaisc3PXo4dzq6jzrgumbCMnWgbYjmvZ0/me5FOu36Yl3F1Kpbk7aeOVpzPqs+zPc+Y1vxWZHfsH6rl1/+6dP5MPnRVh0PNIOHuWeerdps+i3Z6Zh9dQQoYkKrk4S51t6qg4CWMciTH3POM8/+pnd0u0m7qHnv6+is5VpcI+1NXnxbtft88e1D617IXV7f/0q1zXz3LnJcAwgMM8/2HxyxDIcgAAEIQAACo0IA835UappyQiAgAqkBemjmvWl8qIe/1j9zNSpkAr77/s8yGzCaes51fVvHsqYT1cigtGkMk/k8/4KLBi8kZf7oRWdyOj7Drsi2L+a9RhoVKXfaub5OH5pONO265HevvPaH8U033dWYBrVWafKeaZ9d5r2MzrRrqnynUQi+qQ2T2i7zWaMBi+TPZd5r5EuRNFzn6kVgMv+/nN4+/4hEV7q+pTU07bDrfPuY78VbMp9NfNaMCnZeXPtNmvdFnql1l18jR13ltY/5zPsy5pWdrr2vZ0+eUUqm/OrEoyll1clGs+Hk+f9q36/ovu/lpskPWwz7DmnAniLfFfp31rzXMisyqFx1cc4558ZZMa79XNBIedcaxvpfmWfGEjste1/GlToiufKY1+gYa/NeBmDVmXpsJnXty6RzjQ4uEzc1bd7LiN6959cy//+62Mjwc+kna/kFO600fauzn31ukX3pW8souPKnjgd50vJdr9kZ8lyfPEftRrVHXXnSsZDNe70/KGI+m7Jrpi9feXVcbSmzxIW5Js/2jrvWOdMtsk661mv35a3sFPyKA32drTV7U56yaQYCVxo6pmdunjSS52iWCF9ZdbyIea+Y25WWOijoHVTy3nk/61nuGoCidPV7zpMO5r0rlOIYBCAAAQhAoDUCmPetoeZGEICAIeBsnJgGS6jmvRo3h177lreXdR6TRGuUmXKOxVYmcFYjTT36XY28PPnVCz8tNaARm3kbhMn89MW819SXybKV/awXDi7+c29ZnvseGza+5F0r0ZV2mWNFR567zHu9yCrLKe06jUDWC7wy5cp7jUaqpeUh+Z3LvFcHoeR5RT+7zHuNnCmaTvL8sua9ZujwjSTLy7bKeXlmqGjKvJfpXCXvVa9ddd/uzHr3mfdlpzhO6sZ81hqmF100qRQP/Xbn3rIilulQRycxkyezxbzHnK/6WwvgetcU+SbutredM+81Q4s6K7qMF8O96Ohrn+ly+/wHMp+Z5rnh22pqa5Mve7to8SO50h5r835djk5fvrI3fXzzo4edbIt2NmjavF9wx9pcde3ipRH2tm7MvmZpc53vOuYz27TOuOv8IsfU0dv1W8xjVGotcVMee6v16t86UX62srWJ6eLttEM278uaxlqKwS5jcv/Jne+VqmeZ5Mm09LlIxwpN1e9KQ22TIjpLnjt/wYPOdK/M2XFay6K58qV0k/fK+1nLuPimpte98vwmzL1mzb7bmT/VtTmn7Hb1mmedae/c/WGutH3Pk7wd0rLyzch7O0RjHwIQgAAEIDBEAPN+CAkHIACBpgk4GxCmQRWyea/Gx5K7NzvzL9Mtaxp53xSGpuxNb/OOitB0znpJVCU/l19+bbxl27FMJskGXV/Me00ZmCxb2c+aKs9VF1oDNE+ayTUQXWmZY3rJdellk2NN0ac1NjWCTNNhJv80MtZcY7Z1mPfXfHpGrjLlKbc5Z+OmV2ONpjP5TNtqxgSt83qVKf/UuUNlFwtXBxfM+//WWXfHT/40F/u0eqnynab7N1rwbZsy7zUrQZW8V7323grmvUbL+3iVPa7RQ3qB63rxn7esGmWpqXXLdhBz5R3zHvM+r/4CPE9T5H/kmSLfFc+PuXkvI+GJJ9/x/un3vXbd8/HiJZsGcUjW80JTIRd9Hlzz6ZuGns3qZKb11V3PiKLHFCcktaLlY/KkM5bmvWKbMiOB85SrjnO0HJWrM2DRta+bNO+lVy2JU6W8rhkGJk66Mneait+T+hM3rW1eJV/mWk3JnkxfS3GZ731bXyfQop1vkum/+/7PvWvBh2rey+ROliPvZ/0OkvzNZ7XhZCrnTcs+T9fZU++bNPPOrCZ9uZ7XOqaO6Pa9iu4rfvS15dRhKi09vZ9xzbSi8qlzWNq1Wd/5loFQ2nnNe7UV1Mna8DZbtcWz3i1l5U/fa/p8V71qmZc812Peu0IpjkEAAhCAAARaI4B53xpqbgQBCBgCQ40T00jRNnTzXuvx2vm197N6IGt9Mfv8tvfzmvemIae1KhcuethpVObNu17wFFmnrS/mvQxgw7Hq1sdEU+Bnpb3rqc+mak5rUcvc37f/S7FeBmWlZ77XC4WkBkI073851aLfuFfnhNVr9sYaFVxEp65pPzHv3eb9WBvYY2ne+2bNSP52mvocmnlvnh8aWbh4yeb4wgvd65jm4THt+nm1rMmsPGHed968/3+iKPon6+8foyj6N9bf/xpF0b+y/v5lFEUfJ/7+JIqi71p//zGPDsf4nL+LougWE1zn3I65eV8ns9tuv7+w2ay1jV15UDxknlFVt+s2vDh0DxlYeeKMsTTv85pNVflUuV7xdbL+7r5nS6G6a9K8LxoLu1i4OlBLP3k6qcjkTPLR5+k33F6IkStf5pimKnfd4+jbP0y9h9i4rnvp0DdSrzP3TdvetXCDM+1QzXsZomnlyfrOZ0avvHdHpXSvvXbmEEfNmpSVH32/afPrQ9eqvtXWynN91jnqtO7Sz4Nr96Wmr840rusU+2XdM+t7dXbyLduQ93mqZQlc+Vuz9rnK+TP5Vye35D3Uic18n7bFvM8ZXXEaBCAAAQhAoBkCmPfNcCVVCEAghcBQ48FuTIRu3uvFiW8N46x113yNH710kvHf9N87pz7O1UhLNuDe++BvBmvaa71FrUPs6r1t12FyX6Zm3rX3fEZ1nqmvk/lu67NrBJdMqbrurxGwSab6nDUFoepNM0K4rp18zYz44Mv/rHQeNbNCMt2iLyxd0+bXOfJeoxV80xlqffr9B75auvyuUU2Y927z/u2TPxnSirSjJRKafuYp/TeOfj+znhsbeX/0+86y65nRRtnVcSLrOeSbNr+JkffJvOj/qTrEPfDg3lhruJ5//oVOXslnjfksE+fIW+lmQfKers89Mu//X8vAlpn9HywDW2b2v7YMbGNm/4uEif0jy8CWmf2NKIq+Yv1plPeHib8TURQdtv4O/uLee62/nb+492PW3/ooiu5L/C2NomiB9TfnF/eebv1d84sR5pdafxekxJl1fPW/GZ0FuFW9iulZJQraC/Neo0u1FNX7n/37zGdc8jfvm3p9/cMvFU4rmbb57Js6/8mdpzLvMZbm/fYd5abbNuVuY6tRwMnfZJFlpJTHJs37PB0GszgpjWQZ9TlPW2rzY284r81qo2blyf5eU+e78pc2XbtGx59zzrlD1xWNne182Pta9s2Vp1DNe3VitPNfdN/1O1D5tXRd0bTs82+5deUQx3Hjzs2V5vwFa4auVZ7Umcm+R9l9X+cATTmflqaWTnNpQ523067L+51ruTLdL69572Ku66vOCmDnX2VNMtBsHCfe/ctMBr73V2rH2Pcou8+0+SUiOS6BAAQgAIFRIoB5P0q1TVkhEAiBocaD3ZgI3bxXw8Q16kNl0EjCtIbLtieOO8t+99LHUq9LS3MsvhsYLs9/OZaZr9GP48ad4yyXXa8zZy3JVca+mPdiUualsqs+ZTLbLM1+1hqET33m153Xybgv25HD5O/CCycMpR2aee97Aa6pa0+88xe59GjKm9zKODD1YLZFX0COypr3el64prnVqKEk17H63JR5f/K9j4d0Ir1IP2NV1uR9x9K8T+ZFU7bqeSczX6NgXS/6ze/NbK+ffltlljnMe43U/q2EYf2BZVYb43q/ZVjLvN5hGdYyrx9JGNYysG3DWvs3W4a1zOukaT0hkFiu79lQRwfn73cMj5sp8qtooNPmvTr4KJYuOy20njkyVFx1qBl4ks+ksp+1frfrHpqBKytNX+yyYePLmddmpW2+93VsrGtadXOftK3iUM0IJrNZI2dXrNwRaxpnjaBO+7tkwuVDbPMuI2Xy06R5v2Vb9TWqff+XX3/ju5kaWLT4kSE+0qJmojLlr7pVJwKXvrXchS9tX1tm3m33ea/xpeU6fuz4j5x5CtG81xTpVZ5hKr9raQ7VSZ4Oqy5+5tgddz7k5Pju+z/LrCdXZ3blad/zX8681tw/basODy7dTZp0VWr6vme+Onyk3S/vd77l6fKa966O9hooUdd7BJVj6+NvO9nlWeID877v4S7lgwAEIACBwAlg3gdeQWQPAn0k4Gw8mMZYF8z7yZNvdJYh6wWEpiY35bS3s2YvraXxmLeRWfd5J9/7q1hrvmlKRrtcyf1n9/92Zjn7Yt6r7HWtneprcGdNjeh7WVF0+YSkXjSi37XuYGjm/YqVTzr1+PTe38jUYbLM9me9THGNEMa8d4+8Fzu9WEs+DzQFrMxtm+1Y7Tdl3qs8F100PDX8L1/c/m0QZfeZBG2MvM+qb5k7O3a9H2uJkKR+7M/qqJSVVtr3Ocz7I30MxihTKoH/2dZYAPtlpsh3FbDT5r3q4YILLoo1ej7tN532nWaQctXn/he+Er/2xncK/R185evOWVQUZ7nipDxre4+VeX/OOeeVZprG2/7u1cN/FD+w+plY7SgXH1e95Dmm2ZDs+2TtN2ne795TbeSz8r5u/QGnRqW3rLL5phbXLDd16tvVKTOtXaJODa66XL3m2cwyZZXZfD9hwqeG7hGieV/H7GyupRXEV9O4Gx5ltkvu3jzEUOmqQ1JWeppeP1nH+p3rPUHWtXm+V/vT1alTWkxbG97X0aGOmZuU7z2eWR/ymPeakULtoSS3K6+cWvj3qt+3Zn1xzeylzl/Je+hznnczmPeuUIpjEIAABCAAgdYIYN63hpobQQAChoCz8WAaFF0w730j7zXtWVrjUw1fVwPtkksuS210pqUZ2neaclMvSkx92luNBsnKb5/M+527P8gsbxYPfb9sxRNOnhp9lna9poa3+WtfLwPSrsnz3fOemQBCM+9dU0pOnHhF5d/aode+NcRVbDHv/ea9z3yt2pEij17znNOkea/ZSZK/Q32WQZMnb02fE7J5b5d9955f864revv81ZVYYt6b8IytReB/cv1ux+BYlSnyreKc3h1z816xgswU35+Wz3F1kEuyX7psa6nfva8DbjL9Jj7nMXPGyrxXW8R+5ta5r3XYb513r/N/YR2cNaNSkfw2ad7X8b9d04y7uORZ7kqza7mubeNY2oxgPvNQHZSL1F3aua64P0Tz/vLLP125zHPmLhuqZxnbaXzyfKfZAF1aOXY827x3zcR38cWXVs6TnW+xc+Xv+Mmfeu9z7bUzh65RXu10q+y/8uo3h9JXHvM8733vPVxlbOJYnncVmPen4yd2IAABCEAAAmNBAPN+LKhzTwiMOAFnA8c0SEI37zXNnW/N+6xp89UwdK0VrrJXWX+7SoOziWt96y3mMY7VC95owd7WsYZkE2VVmr5pAvOMsMqTJ00LbbMw++phn3b9BeMvHroua3aItPTMd1o30+TB3oZm3k+69OqhfM6ctTiVmSlj2vaxre61EzHv/eb9vfftHqoLaWfpPVsq10daXeX9Tp2ObC2bfY36zpuG7zzf2rV1PR989817vCvmvcrj62ShTjl5y+s6D/N+xKNSd/G1VILzudDS8TqmyHeVbMzNe3UAdP0O7WOa4UYjCO++Z0usEeE+5hrFbV+XZ3/ipCu96fnuU9fx226/PzO/Y2XeFzXA87DWOepoqimg62LoSkdtq7z50XlNmvd1tOeqmPeumY5czJo4ltbG0G/Vdc+duz8sVHdp9TzDMUtPiOZ90faCq8wu817vJ1znFjlW1rzX6HpX/Rb9bWblVZ27XPdJW1LC9e6ljtkPTF7fPPYDZ57ymPcvHfqG81pXGZs49tjWo5mawbx3hVIcgwAEIAABCLRGAPO+NdTcCAIQMARSGymhm/eaVtPXeLp/9dOZDSAZ/K7rQzGxTEO0ylbT2o0ff8lQOc+/4KJMPlqL3MUnT8eIKnmucq3PvK/jRaheCLimptRU3Gl51jR8Lo5aTzTtujzfTZ12izPt0Mx71wv/+QvWVC7/zTMXO8tf9GXcqKx5L01pvUuXHvWSWc+LPLpr8hyN6nbl7/HtJyrnTf/TXGnr+VB1zdM6mHTJvJehN2HiFUM8s6ZMzeKEeW/CM7YWgf/O9btt6djfRlE018pLnbudMO/t3+yh178du2YSUl3ot190rfqmjeQ0jcy7bVXm/5SxMu+vvnp6Zt7sesmz/+iWI84Y1sVII2HVsUIjazUNvu/PNStDUYOwz+b9eeeNH/of6eLdxLFb5/n17WO+d99v1aa7uXnSUlAAACAASURBVLesGCo75r2/Y63rN1zWvH/j6PeH2EtjaiO77lP2mG+JvgMHf997H9fyVZpFsWwektf5Oi7kMe+ffe6LTm5N/D5daeoZnSxP8jPmfZ0hGGlBAAIQgAAEChPAvC+MjAsgAIGqBFIbKaGb9xoF5Gr86Niupz6b2QDSSwrX9eotX9faa8lG11h8njJlzlA5tWRA2pp0yqe+dy0tsOCO6qZrUxx85r3queoInDUP7R/iqHTTXpCpnO++/zPndXfc+VCmRtM46eWIS7861gXzfu4tyyuVXyaCb51WzHv/C0L9rjU62qWdRza/VqlO0vSa9zut+ejKW5V1lc291TnBtQ6o7rdl27ExL3uXzHsxdY2sE8vj7/xFaZaY91XDul5e//euZ0LDx+qeIt9VMZ0z7/W7P3b8R/Gll012Pqdl+L598ie5f/++/+EN1+0g76Nk3qvTnqvzqeGs5+59D3wm1vI56qia1T4w/1Nds1Fh3n8Sf42lvtPaJstXbHf+fp/e+1Hu367RgG/rHo1+fm3p++6bdtw1ZXvR9oIrfXdZx27kvW/6dy3j4Mp/2WO+pahefPkPvPe5+OLLhrSnzsNl85C8TvGnea7Z2zzm/VNPf955rZ1Ok/uY964wiWMQgAAEIACBoAhg3gdVHWQGAqNBILWRErJ5rynZXKN51aiS+f7Oqb/ObAhq5OBVV093Mpi/4MHM65MNxlA/uwztrNHipiwaoZ9sqE6/YX6wbFxlNfnXmq6mXEW3Wr/PZ/rt2PVBZrqu0Ul6WVo0H+Z8afeGG+cP1Y0pa2jmvcssrjobwuw593jLX/Rl3CiNvJeGfFOmaiS1jBmjs7HYvvTK1531qun+68jPynt3ONPXyJ8iplMdeUmm0TXz3jXqSuu86vmULFvez5j3oxF8Fizlz83/tha2/zmKoo+iKJpQMI9lTu+kea/fstb71m/dVR93Llyf+/fvWvpKsdYLL/5e439pUzub51UfRt7reayR8666ktG+b/+XcteX4WK2ijWT6WLef2Leu2aWUPurDX2nLee1es2zQ/WmeszTnjF1n7W94cYFQ/dg5P0n2sjip+/Ljrz3dRq/7LJrSv/WXfmdPPnGoTqWjt448ife+6h9lnxmaGk5V/pljuneyfT1OY95r5ljXNfOmr20ld/sWyd+nMmBkfdlQi2ugQAEIAABCNRGAPO+NpQkBAEI5CXgbKSYhkuo5v2x4z+OtWa7yWdyq6n68jb4Ht9+0pvO+odfyp1O3vtlTc2s0S5VjI9kPmQ662VJkpHWqUue6/p83ZTZQ9deeOGE+NSHf5vreleaTR5LM+/FQOs0l7n/4iWbhjgoPY0+y6pT3c/VSUQjcg699q1S+Vm3/oAzP6aeQzPvfdP76wVmmfp4dMubqeXHvE9/QSiTWmtMGr3YW42ka2L6/Dy/E2lBLx3PPnvcUN5unrmolFaS+jr69p/Frk5JYqCR5HnzmUw37XPeNJs07/V/Je9oyrSymO/UQU4vXG3taF/TLJtzymwx7/OGbyN13l8lddbQ5yanyHdVWGfNe/22tTyVqx40wltrB+f5/U+Y8KmhNHT9ex/8PNf1ee5R5Zw+mPfPv/i7Q4xVbzLyqnZYk1Gf1ADm/Sfxl2tpGcX+Wk6rii6rXrtx06GhelM95llzO++9XaPcMe8/0UYejmXNe6XtGuBQ59ryusekS6926khT1/vK53qvoGd+Xe8+fOvW5zHv1Skt+TzT5yqd/30cyh7HvHeFUhyDAAQgAAEItEYA87411NwIAhAwBJyNFNNwCc28l2G87YkT8SWXDE+5ZvKslyIHDn7N22hMNpbUWHRN+6j01Jh84MG9tTQoNY3bg2v3xVmjrWUsXXXV9bGmTqvj5eXSZVuddbxi5Y5cjBYueth5vTo9JFmG8DnLvNcomOde+J1CeVcnDqOv5HbT5tdzpeVb4qGMWSjjWtpM5sX+HJp5v+r+p5z5lalfVOdPPPlOPG6ce8SfYYB5n/2C0PfyVgxnzLizlhH4Mot37v4gnjbt1lhrSeZ9Rrg6u2iUZ9pIsrxp67yH1r3g1KPKPnPW4vitE3+eO6+++6rs23e8F+tFZd5OKk2a9+oU8alPXRtraQTt+/Kd9/h99+9xMlRHp7xpuM7DvDfhGVuLwE/Ms72h7b+PomhnFEVnWfdsY7fT5r06eemZ4qoTPUddv+/kMdeyTkpPy+Ikzx2Lz30w7zVrjauOnvrMr1dm7O7AdW2hdH3rrxfpWLBuw4vOMlZdLkua86Utoy9Lk1OmznXmq2zH3az75f3+8e0nnPlaVdMMR8rH+PGXDN0D8z47NrfrsIp5f8mEy4f46zlQR3yrPL5z6mNnW3TcuHNSfxc33bzQma88M6HYbHz7mj3C9bzLY95rwIPr2qqdUn15LXMc876N0Ix7QAACEIAABLwEMO+9aPgCAhBoioCzkWIaLk2Y91rnWwZMnr8XX/rawPCRQbpw0cbY1xA1+dV22fLHUxuNrobS4SN/HGs0uZ2Ova8e10UNX91HnQ3E8K6FG06P9NT04a48mGMy7829NWJj5aqdcdracea65FaGqKaHdq1Zr2Oamjp5jevzkzvfO50fky9tNXJ1w8aXKo/acd2zyjGXeZ80uvWyccu2tzLLr5Ex6uTgW7NS05DmHVG7d99vOTmK5bzb7ovTRikYHife+Yv4nmXbhupUL0rsutF+aOa9OtS4tKi8ajpCvTAx5fRt9aLo3lW7hl4WucqPeZ/9glAdl2bNvntIO0ZLev6oo0jRzhWqv1de+8PB1PxXWDOkFDHvfZ2ONKJvx673K49ak7Gu57opa3Kr0UR6RpSZgUCjfrR2sD0iMRTz3pRTa44uW/FErBGZvt+b77j+r2nZBd9zUWaX79o8xzHvmwr3Op3unxnt1rxtc4p8VwV02rzX79lnAqqennk2e/3sBXesdT6Ht2w7Vuk5kudZk+ecPpj36jyX/N2oPVV1Jha145Lp6jMj7z+Jv9TmdTGqc4R7Hh0nz/GNTp49Z2ktvzu1q13lxrz/RBvJOnF9rmLeu373qpOn92Y/l115SR7b/8JXnHWsuD95rv3Z11lHsb19Xtl9X8fSPOa97ul616R49+jbP6wlf2XLZa7DvHeFUhyDAAQgAAEItEYA87411NwIAhAwBJwNL9PgbsK8N2k3sZ05a0np6dxlLLmmmLPzqYawTAudq0acPXW8jFW9rNjzzBcGI+xvnXevc1rqIua9fW+tD754yeaBmSYz/+R7Hw81IjW6Xy8aNcp50qSrvHWbtwGrhqLMKxk9dl7sfRmyWitaRvb0G+YPDDGZYvafljkwjc6mty7zXkss3Dxz8VAZlF+9QLPX+NbLTPX+X7vu+dg2Hu0ya1/rtL786jcLlcu1PrRJd+KkK+M1D+0fpGlPHagRsnpBsmLlk84ZJ267/f746qunD5UtNPNe9a6Xgqa8ya1m09BvS51K7BfK6kAh41Mjx8QoeZ1+864p+THv870g1HNLS2gkudqf9czSC0R15Dn85vcGo23M71j6PPLWDwcmsEZ0q5OVa9S80iti3uu3Zechua8OG/p9alpW+1lj9m+7/YHM36ZGH1151bTU+8jEX3rPlsE6sHq+a6p4U3btHzn2p7FG9m3c9Oqgk5bvmRGaeW/z1NTJ+p+w6dHDg1lrXB2JTrz7l4Ny6jeq8+3r7f083A0/3xbz3oRnbC0C37d1VtP+30RRNMe6x1jsdt68V7zi+x8y7fp5p5+Xvt+7bwaYebetyrzWl2adx9Vx16W3DRtfri1/V0++YegeiuvqKodGjSbLcNNNd1VOf+vjbw2lq/tg3n8Sfyk2SLLX51vnja2+FWeff/6FQ3lTp027DVJWg1qizFVuzPtPtJGHbRXzftHiR5x1oBkF89w765yHPMu3ZT27NXuaSxt5ZwTMyperva/75X334WurakbErHu38T3m/ViEatwTAhCAAAQgcJoA5v1pFOxAAAJtEXA2oEyjqkvmvUY3lBkhaje0VF7fWsiGSXKr830jEJPn6nNZ896VlqYOl7GuNUMvuOCi1Lo01+uFoMucsTkk9zc/9kautM09kts3jn6/tQavz7x/89gPYtfakyav6rih3vautbbNOWar+paRmeSU9VkzScj0N+n4tpoeXHlxTUdqX6MXvjJfu2Lea5rQPDqVrmXmp82GIQ6arlcdQ1yjSzDv878gVOcV1xqUttaS+3oBm9XZKXlNEfNevyXfiLVkuq7P6lCU9XvU9zLffcaTK10d+2XZ05dtSF4bsnmfzKs+q3PExRdfOnhmZj2HzPX6PdYxHSvmfVvhX6fu8x2jsxq2/xhF0ZZASt95817PUY2Y9NVLVjtCnSVd1ypWUIfUPM/xJs/xrYGspVfqum/T5r3LpK2jo5XP5MK8/yT+ev3N7zn1rbbjWOvb16F43/4vVdb2LbeudJYb8/4TbeR5flQx732zoqiDbZ57Z52jTvuuZ7c6ZKVdq07ArusUQ6Zdl+c7xaC+ZdXymve+DmWa7j9PHpo+B/M+kOiNbEAAAhCAwKgSwLwf1Zqn3BAYQwLOBpRpVGW9dEtroNy5cH1q2uYeVbeXXjY5furpz9fWoNJUgr6Ro1XzquvrNO+L5mfq1Lmlp7lfcvfm0vUZgnkvrWrGgosumlS6HOItcz/PlPu+34bWwC7S2cNXx1dddf3AeNR9umLeK697nv58rg4SvnKb43o5bNZHxLyPTmtanVR82ks7riUJ7rhr3el0DOc6t0XNe80u4qrbPHnKa96LiUaV3z7/gUbL3jXzPg9j+xx1mtKSL2kay/sd5v0YRoTh3vpbtt5K7psp8i8JqJi9MO81UnfyNe4ZXDQzTtZv32cCafrjrGub/t7XuUBLo9R176bN+/POG+40OnvOPZXyLy7JJanM7xLz/kyDVjMEGTb2VrOk1aWhMumsXrPXmS/NoFQmPXPN2yd/MujkaJfV7GPen6kNw8y3rWLev3Xix4OOmIa9vc0bk/ry5evUpNn41GHFd505rg7Wdn7MvmaaM+eU2fqMd6Wf17xXp15fR/6q+StTpuQ1mPcBRXBkBQIQgAAERpEA5v0o1jplhsAYE3A2nkwjKlTzXubnDTcuiDVlo6ZtTjZsqn5WmppWrugofMPNtdVLrptnLopl3qblTyMx9OLLlUbZY3pxd//qpyuzUt7LdGwIxbwX90OvfzuePPnGUnzV8UJr16fVX57vZGBX6USgJRk04t7cq0vmvfL87P7fTp0FIUvnmipfL6VM+V0GLyPvi70gNCyf+syvx+oYklUHRb7XshV6SVxmlNmpD/8uXr3m2cHsIkXuWcS8N2XfufvD2PdCsci97XP1vNRyGHlnO7n/gc842b/y2h+e1rvJb9GtlkDxTetv57nIvmZfWLlq5xlLKRTNV/J8zPsxjgrDvP03iujSce7PA5gi30W2F+a9fsNP7jzlfHapLrScVPJ3bn/2ze6k0fd5jCA7rbr3NZLToadBZ7e67tW0ea/lX5Jl0FT6VfLvG1mt+2Denxl/PbrlzSH+4jTQ9xvfrVQPVerQZ1Jq9q9XD/9R6XwtXbbVWV6VGfP+TG1k1V8V815pz5p9t7Mubrhxfun6Vbpaoir5TNHnPEul6Pq1D+13Xq90s5j4vleHX7XTXfnSsbzmvdL3Pd/UGU2xtC8PbRzHvHeFUhyDAAQgAAEItEYA87411NwIAhAwBLyNHDV0qpj3anCOH39JpT9NBy/jZ8qUOYP1slfdt3uw9nDZ0aVFG1XHT/401tSYRadVNg1HlX/O3OWxGlp6SVLk/jKZ160/MGh4a/pik2aRrV7OiVnd5rl67Kvhffv81YPptj91xXUDc81X323Vl/j6ps232avhrbX6Jk26KhfXiy6aOFhzPa8BZ9/Lt691wu9Zts255qSrjjWaQS81ZPwn09TLhCR7dW5Jnpf2ecrUuUNpXD/9tkJppKWf/E7m+8p7d2QuDWCzuH76vHjn7g+G8qROMcnyywBM3jPts9Z/Taah333aNXm+09TvyXTVcSPPtWnn6HedTFefpau06/J8p7VQdz312cGzp+jU+KovjViRJpXH5w98tXJ+lGeZ+JphRSMx9UzVSE2tva5no4uDOujkKWvyHN1HU0BrzUy9wLb1l2df03Xq5aXyeeDg1wrnQc9VV3lePfztwmkly2Y+v/bGd2Kt16wpj7U8R55yJc+RGaTfr5n9wqRdxxbz3oRnbC0CqSZ3Up/W538bRdGjVjqh7aaWa95t99X2u3eNvhanup7RqaPvp85NLce77/988Dy36u30c0nGdp2xl3lG6f+c2c/aanmqZN4000DWdXm/b9q8V+yUzH+VuvetdW/ugXl/pkErfas9ZvjY2zJLmeXRVV59z5m7zJkvmb76Tee5l33OwVe+7h3trXJj3p+pDZuda7+qea+OU7be7P2ys8jpOjsde3/7jndzaUbLdfmmt88a5ODipGMrVj7pzZfyWMS810x9vplFFi1+JFcZffl0HddvLe/vDfM+tDCO/EAAAhCAwIgRwLwfsQqnuBAIgUBqQ6eKee9qnHT5mAzwx7efjJev2B7Pu21VrLUCZSDpTy/GZs5aHGsNSU2lqfM0pZyMoLrKLMNFDdo1a5+L71q4YXA/kwdNyShjd+4ty+Ol92yJH9t6NNYLlLwNwbryGEI6ecx7k0/Vz77nvzyYleDWeasG9aj6vO66WQPjctnyx+Pde36t8owF5n6urUbQyyi9Z/njsaYxlfGnPGiJA2nK1GfRzh+ue4V4TCMlxFgvXfQSUb8llV+dCWSgLln6aKxRS28c+ZPafkshcgg1T6qfp/f+xmD0uzoiaM1HmauDOpoyJ54x485YM0GoI8qmza/Hmhpf09CHWp4i+ZJhpM4ymoVlwR1rY3XwOF32qXPjGTfdNSi7nhObHz0ca51Y8SpyjxDOPXzkjwcdFtRxYOGih2PNbDH9hvmDOlY963+LOkzoJfKjW44Mlh9p8n8L5n0IoWFwefiqbRDk2A9xinwX1N6Y93qWpY6+d3Q8tJ9/O3Z94G2PXDdldqznlH1+2f29+744eMYpTs+bhp79Sc2pQ2XVqafN/Zs279UuSeZfnxVnvvfBz3NzUH5VT1kd2zDvhw3atvStGEz/w7XeudFX2lbnu7ShY2qXpF2b/E6/0YmTrvSmpzQx74e1keRof65q3ist1+xkqgv9jou+51F7wPf7V6fZvJ1GlC/fUl3qaFZ0enq1E/VM9mlZx4uY98qflo/wpae01CnHrqsy+3r+Prb1yGBWQw0YyZMG5r0rlOIYBCAAAQhAoDUCmPetoeZGEICAIeBtmKjBUrRRl6fRwTnFGu7wKsariHkP22Js4QUvNIAGmtYA5r0Jz9haBL7ke4nuOK4p8mdb14a82yvzXp16fMsSqUNe1rPjzrvWedskF144MX74kUPxO6f+OjOd5H3U+Xb9hoODjplGL9ueyGduKi3fFM8TJl4RK52qy3c1bd5rFKnP2FKn0TxL2qiMWtYluRb0uHHnDNUZ5r07Trhz4fohVkaPv9T3KzXq+3ju34k6YJp8JLfKc54OmeoUo+WK7Ou1xJ39WfuY925tJJ9Z5nMd5r0GE7h+p7+sjwvidRtezDTdZcrrGXruuRcM1anSUV0/9/yXc2tO5VOndC0dkdSIPmsGqm1PZGv41Id/O/RcculOaRY179UZN22pKc3OpqXGynRmPXDw9+MVK3ecMc0/5n3IoRp5gwAEIAABCJwmgHl/GgU7EIBAWwScjSbTkMK8L9bINo1ttmPHDfN+7Nije9ijATRQVQOY922Ff526z2+auDRl+39GUbS5U6WKol6Z9/rtP7nzPW+7QkufpD0fNJJRs36k1HGsKeyX3L05fuLJdwaj8W3jXAaT1qiXWaWpnTVzkZbdchnXRcz7o2//mXeKZ+VVhrZMHt1LM5Uk/zRjTVq5mzbvdW+Z9D6uWqLswbX74lde/eYZRpRmp9IsXvevfjq+9LLJQ9dr9hnXlPyY9+44QPrWWuO+etBxLQW0eElOfS/bmqLvbOPTaPLY8R8Plj7z5UtLjEkDh1771hkm7zunPh7MHKbZ6JKGqZYJ0NI8yTQx793aMHWR3NZh3itNLcOXrAv7szpdaTk51bEZUa5R4a8e/qOBuZ+1fKBm90jmPc/nrY+/nZovPUt1ztsnf3JG+oMOWQ+/FGvZCbsc2l+6bOvQMR0vat4r/1qyKmvpQrHTMmGalUD/K9ShwJRdLI++/cNBxwaNmFcHNS35lcyzPmPedyx6I7sQgAAEIDCqBDDvR7XmKTcExpCAswFhGhWY98Ua2aaxxnbsuGHejx17dA97NIAGqmoA834MI8Jwb/0FE5c6tl2ZIt9Ft3fmfero+ylzTpsavueERvlqORZHPXuPaTTo+Z4RnL50ipj3yuvKe3d47++7hzmuUa++8up4G+a9TDnfKFeTT201ZbWmPr/kksuGDFn7PJn5R976YXz99NuGuGDe++OAgb4dyzDYbJP75fSd37yXBvcf+Kp3VLWdH2l54sQrBiOj7eP2/jnnnBc/98LvxLfdfv+QNjDv/dpwPSPqMu+VtjqF2PWUtn/B+Itzn6tlpYpMl58sp5ZnSMuL+e6888bHky69Opa+zLHkVh2o9BtLHtfnMua98vrSK18/Y4S8K+3kMT1rpfXk8bTPmPeuEIljEIAABCAAgeAIYN4HVyVkCAL9J5DasMC8L9bITjZI+dw+P8z79pmjc5ijATRQlwYw7/sfeJYo4ec8L71/1qEp8l3F7p15r+dA2uj73Xt+LdXI1vXvffA3g9GTrhHzHh2ktmVc1xQ17zWaUmuJu9LKOhaCeS+umvkgOe19Vt5d32t69EOvf3tQj5j3xf/3S9/3LNvmnBHCxbvMsTxTjif/Zz+996P4/PMvLKVxk8dx486Nt+94b6ANzPsLMp91yTpIfq7TvFfHqntX7apVd9JxFeNe5dX1mk3FaKjs9sqrpg1GvmvGAFcaZc175fHNYz/InBXGdc8ixzDvXSESxyAAAQhAAALBEcC8D65KyBAE+k/A2cAxjQ3M++IvZZINbz63yxDzvl3e6BveaAAN1KkBzPv+B54lSphsJP+7KIp2lkgntEt6ad7LJHKNJlfb4rops3MbWs8+98X42mtnprZTTHsl71ajylesfHKw3nLR55bKtWHjy7GmEc97P50Xinmv8mptck2TXyT/9rmaKl9TQxt2mPfl//8/u/+34xz/7wrVlUbFl9W36vSlQ9+Ir7rq+kL3NPqQrp559jdPawPzPizz3vxmtU67OuCYeiuznTDxinjHrvdP17VJu8r24UcOFR6tbvI+c9bi01PrN2Heq1zqZPDI5tcGs5KY+9ax1TNg/cMvnTHdfhpHTb/vuu8LL/5eLfWxcdMhZ/rWPf8htECK/EAAAhCAAARaJJB8L5H1f1PfX9xi/rgVBCDQQwKpDxrM+/IvZdIaXnzXHFfM++bYolvYogE00LQGcpgZR3oYi1GkdALv/OrFsZkivy8N4NbMe5kt48dfMvR34ODXannhn3wuPLnz1NC9zP2fefaj3PeUYb7nmS/Ec29Znmtab8tgGLRvNHpfaxJrhKhGFlcdJapyKg2tb6x1npUvpa91jC+8cKKzzFozOcnH/jxt2q1D102ddkvqNfb1Rfe1xrkM3jzT6Buemo56+453h/Kk2QhMvZqt4vAiebr3vt1DaSitvCNRda+Nm151pnHg4O8Xyosr3760ZXS7zi9yTPpWW7uqvrXOd136lvmpNdLV0cXUf9pWOlq2/PGh+rpz4fqhOrlkwuWVmRXhmzx3+g3zh/KkmCN5XtHPt89/YChddWYomk7yfP1Oze/K3r514seV0n73/Z/FMoGvcqwZn1bXV1wxZaCNd059XOn+yXKaz4ff/F48f8GD8TnnnJtLe+oklpxlQvq1WZl9/Q8w96myfefUX8ebHj0cT506N3VpER9HTft/w43z49Vr9sYvv/rNwnna/OhhZ/lefPkPCqfl4oB5nx6M8i0EIAABCIw8Acz7kZcAACDQPoHUxhHmPUaNq2EX8jHMezQbsj7JG/pEA+kawLxvPxDswB2PR1GkKfJndSCvRbLYmnnfh+eOTBMZlA+u3RdrneUZM+6MP33tzYNR/trKEJk9Z2m8aPEjg5GMMv3tUeJ9YFBnGU6+91fxrqc+OzBeZ81eGqvDgDoiiOVNNy8crBGtkaavvfGdWkyhOvPex7Skb3Vsya3vpz/fqL5PfahOKh/FK+/dMfhdaRYM6eO662bF0svyFdvjXU99Lla++1gfo1ImGcgybBcuejjWzBr2M/XGGXfE6oSxYeNLg7Xf22Kizglbth0b5EkdLq759Ix48jUzYnV0UkeJNQ/tj+syq6uUSf9f1FFt5aqdsWaamHb9vNP89DsRv1vn3RsvW/FELNN93/NfDv73gnlfJITjXAhAAAIQGEECmPcjWOkUGQJjTQDz/nPpRkKVRh3Xts8W87595ugc5mgADdSlAcz7sQ4Lg7z/p4LMVfVMYd4Tg2N8ogE0gAbQABoIQgOY99UDO1KAAAQgAIFeE8C873X1UjgIhEkA857GYhCNxbqMH8x7TMS6tEQ6aAkNtK8BzPswg0Vy1QgBzHti8F7F4PzPbP9/JsxhjgbQQF0awLxvJNYjUQhAAAIQ6A8BzPv+1CUlgUBnCGDe8+KwVy8OMe95gVHXCwzSQUtooH0NYN53Jn4ko9UJYN4Tg/cqBud/Zvv/M2EOczSABurSAOZ99cCOFCAAAQhAoNcEMO97Xb0UDgJhEsC858Vhr14cYt7zAqOuFxikg5bQQPsawLwPM1gkV40QwLwnBu9VDM7/zPb/Z8Ic5mgADdSlAcz7RmI9EoUABCAAgf4QwLzvT11SEgh0hgDmPS8Oe/Xi8Prpt8Xjx19yxt91U2b36X/BfwAAIABJREFUqox1NdBJh5c9aAANhKYBzPvOxI9ktDoBzHticOJTNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCmPc0FoNoLGKStW+SwRzmaAANhKYBzPvOxI9ktDoBzHticGJwNIAG0AAaQANBaADzvnpgRwoQgAAEINBrApj3va5eCgeBMAlg3tNYDKKxGJqBRH4wNdEAGkAD7WsA8z7MYJFcNUIA854YnBgcDaABNIAG0EAQGsC8byTWI1EIQAACEOgPAcz7/tQlJYFAZwhg3tNYDKKxiEnWvkkGc5ijATQQmgYw7zsTP5LR6gQw74nBicHRABpAA2gADQShAcz76oEdKUAAAhCAQK8JYN73unopHATCJIB5T2MxiMZiaAYS+cHURANoAA20rwHM+zCDRXLVCAHMe2JwYnA0gAbQABpAA0FoAPO+kViPRCEAAQhAoD8EMO/7U5eUBAKdIYB5T2MxiMYiJln7JhnMYY4G0EBoGsC870z8SEarE8C8JwYnBkcDaAANoAE0EIQGMO+rB3akAAEIQAACvSaAed/r6qVwEAiTAOY9jcUgGouhGUjkB1MTDaABNNC+BjDvwwwWyVUjBDDvicGJwdEAGkADaAANBKEBzPtGYj0ShQAEIACB/hDAvO9PXVISCHSGAOY9jcUgGouYZO2bZDCHORpAA6FpAPO+M/EjGa1OAPOeGJwYHA2gATSABtBAEBrAvK8e2JECBCAAAQj0mgDmfa+rl8JBIEwCmPc0FoNoLIZmIJEfTE00gAbQQPsawLwPM1gkV40QwLwnBicGRwNoAA2gATQQhAYw7xuJ9UgUAhCAAAT6QwDzvj91SUkg0BkCqeb9Y1uPxIde+xZ/MEADaAANoAE0gAYa18DVV09PjUuiKDrSmQiLjEIgnUCqeT9z1pLGf2/E+LRx0AAaQANoAA2gAWngvgc+kxWD/0N6WMO3EIAABCAAgV4TwLzvdfVSOAiESSArQOf7KIIBDNAAGkADaAANhKEBzPsw40lyVZxAqnkfhfF747lHPaABNIAG0AAaQAPSAOZ98ViPKyAAAQhAoD8EMO/7U5eUBAKdIUBDjIYYGkADaAANoAE00BUNYN53JsQkoxkEMO957nbluUs+0SoaQANoAA1g3mcEdnwNAQhAAAK9JoB53+vqpXAQCJMAjTAaYWgADaABNIAG0EBXNIB5H2Y8Sa6KE8C857nblecu+USraAANoAE0gHlfPNbjCghAAAIQ6A8BzPv+1CUlgUBnCNAIoxGGBtAAGkADaAANdEUDmPedCTHJaAYBzHueu1157pJPtIoG0AAaQAOY9xmBHV9DAAIQgECvCWDe97p6KRwEwiRAI4xGGBpAA2gADaABNNAVDWDehxlPkqviBDDvee525blLPtEqGkADaAANYN4Xj/W4AgIQgAAE+kMA874/dUlJINAZAjTCaIShATSABtAAGkADXdEA5n1nQkwymkEA857nbleeu+QTraIBNIAG0ADmfUZgx9cQgAAEINBrApj3va5eCgeBMAnQCKMRhgbQABpAA2gADXRFA5j3YcaT5Ko4Acx7nrtdee6ST7SKBtAAGkADmPfFYz2ugAAEIACB/hDAvO9PXVISCHSGAI0wGmFoAA2gATSABtBAVzSAed+ZEJOMZhDAvOe525XnLvlEq2gADaABNIB5nxHY8TUEIAABCPSaAOZ9r6uXwkEgTAJToijiDwZoAA2gATSABtBAFzQwIcxwilxBoDCBy4jBaYOggZHVwImouBF4D3oZWb10IT4jj/1vR0wuHOlwAQQgAAEIQKA/BDDv+1OXlAQCEIAABCAAAQhAAAIQgAAEIAABCEAAAmcQOFjCvL/hjBT4AAEIQAACEIAABCAAAQi0RQDzvi3S3AcCEIAABCAAAQhAAAIQgAAEIAABCEAAAi0TwLxvGTi3gwAEIAABCEAAAhCAQAUCmPcV4HEpBCAAAQhAAAIQgAAEIAABCEAAAhCAAARCJoB5H3LtkDcIQAACEIAABCAAAQicSQDz/kwefIIABCAAAQhAAAIQgAAEIAABCEAAAhCAQG8IYN73piopCAQgAAEIQAACEIDACBDAvB+BSqaIEIAABCAAAQhAAAIQgAAEIAABCEAAAqNJAPN+NOudUkMAAhCAAAQgAAEIdJMA5n03641cQwACEIAABCAAAQhAAAIQgAAEIAABCEAgkwDmfSYiToAABCAAAQhAAAIQgEAwBDDvg6kKMgIBCEAAAhCAAAQgAAEIQAACEIAABCAAgXoJYN7Xy5PUIAABCEAAAhCAAAQg0CQBzPsm6ZI2BCAAAQhAAAIQgAAEIAABCEAAAhCAAATGkADm/RjC59YQgAAEIAABCEAAAhAoSADzviAwTocABCAAAQhAAAIQgAAEIAABCEAAAhCAQFcIYN53pabIJwQgAAEIQAACEIAABKII8x4VQAACEIAABCAAAQhAAAIQgAAEIAABCECgpwQw73tasRQLAhCAAAQgAAEIQKCXBDDve1mtFAoCEIAABCAAAQhAAAIQgAAEIAABCEAAAlGEeY8KIAABCEAAAhCAAAQg0B0CmPfdqStyCgEIQAACEIAABCAAAQhAAAIQgAAEIACBQgQw7wvh4mQIQAACEIAABCAAAQiMKQHM+zHFz80hAAEIQAACEIAABCAAAQhAAAIQgAAEINAcAcz75tiSMgQgAAEIQAACEIAABOomgHlfN1HSgwAEIAABCEAAAhCAAAQgAAEIQAACEIBAIAQw7wOpCLIBAQhAAAIQgAAEIACBHAQw73NA4hQIQAACEIAABCAAAQhAAAIQgAAEIAABCHSRAOZ9F2uNPEMAAhCAAAQgAAEIjCoBzPtRrXnKDQEIQAACEIAABCAAAQhAAAIQgAAEINB7Apj3va9iCggBCEAAAhCAAAQg0CMCmPc9qkyKAgEIQAACEIAABCAAAQhAAAIQgAAEIAABmwDmvU2DfQhAAAIQgAAEIAABCIRNAPM+7PohdxCAAAQgAAEIQAACEIAABCAAAQhAAAIQKE0A8740Oi6EAAQgAAEIQAACEIBA6wQw71tHzg0hAAEIQAACEIAABCAAAQhAAAIQgAAEINAOAcz7djhzFwhAAAIQgAAEIAABCNRBAPO+DoqkAQEIQAACEIAABCAAAQhAAAIQgAAEIACBAAlg3gdYKWQJAhCAAAQgAAEIQAACHgKY9x4wHIYABCAAAQhAAAIQgAAEIAABCEAAAhCAQNcJYN53vQbJPwQgAAEIQAACEIDAKBHAvB+l2qasEIAABCAAAQhAAAIQgAAEIAABCEAAAiNFAPN+pKqbwkIAAhCAAAQgAAEIdJwA5n3HK5DsQwACEIAABCAAAQhAAAIQgAAEIAABCEDARwDz3keG4xCAAAQgAAEIQAACEAiPAOZ9eHVCjiAAAQhAAAIQgAAEIAABCEAAAhCAAAQgUAsBzPtaMJIIBCAAAQhAAAIQgAAEWiGAed8KZm4CAQhAAAIQgAAEIAABCEAAAhCAAAQgAIH2CWDet8+cO0IAAhCAAAQgAAEIQKAsAcz7suS4DgIQgAAEIAABCEAAAhCAAAQgAAEIQAACgRPAvA+8gsgeBCAAAQhAAAIQgAAELAKY9xYMdiEAAQhAAAIQgAAEIAABCEAAAhCAAAQg0CcCmPd9qk3KAgEIQAACEIAABCDQdwKY932vYcoHAQhAAAIQgAAEIAABCEAAAhCAAAQgMLIEMO9HtuopOAQgAAEIQAACEIBABwlg3new0sgyBCAAAQhAAAIQgAAEIAABCEAAAhCAAATyEMC8z0OJcyAAAQhAAAIQgAAEIBAGAcz7MOqBXEAAAhCAAAQgAAEIQAACEIAABCAAAQhAoHYCmPe1IyVBCEAAAhCAAAQgAAEINEYA874xtCQMAQhAAAIQgAAEIAABCEAAAhCAAAQgAIGxJYB5P7b8uTsEIAABCEAAAhCAAASKEMC8L0KLcyEAAQhAAAIQgAAEIAABCEAAAhCAAAQg0CECmPcdqiyyCgEIQAACEIAABCAw8gQw70deAgCAAAQgAAEIQAACEIAABCAAAQhAAAIQ6CsBzPu+1izlggAEIAABCEAAAhDoIwHM+z7WKmWCAAQgAAEIQAACEIAABCAAAQhAAAIQgEAURZj3yAACEIAABCAAAQhAAALdIYB53526IqcQgAAEIAABCEAAAhCAAAQgAAEIQAACEChEAPO+EC5OhgAEIAABCEAAAhCAwJgSwLwfU/zcHAIQgAAEIAABCEAAAhCAAAQgAAEIQAACzRHAvG+OLSlDAAIQgAAEIAABCECgbgKY93UTJT0IQAACEIAABCAAAQhAAAIQgAAEIAABCARCAPM+kIogGxCAAAQgAAEIQAACEMhBAPM+ByROgQAEIAABCEAAAhCAAAQgAAEIQAACEIBAFwlg3nex1sgzBCAAAQhAAAIQgMCoEsC8H9Wap9wQgAAEIAABCEAAAhCAAAQgAAEIQAACvSeAed/7KqaAEIAABCAAAQhAAAI9IoB536PKpCgQgAAEIAABCEAAAhCAAAQgAAEIQAACELAJYN7bNNiHAAQgAAEIQAACEIBA2AQw78OuH3IHAQhAAAIQgAAEIAABCEAAAhCAAAQgAIHSBDDvS6PjQghAAAIQgAAEIAABCLROAPO+deTcEAIQgAAEIAABCEAAAhCAAAQgAAEIQAAC7RDAvG+HM3eBAAQgAAEIQAACEIBAHQQw7+ugSBoQgAAEIAABCEAAAhCAAAQgAAEIQAACEAiQAOZ9gJVCliAAAQhAAAIQgAAEIOAhgHnvAcNhCEAAAhCAAAQgAAEIQAACEIAABCAAAQh0nQDmfddrkPxDAAIQgAAEIAABCIwSAcz7UaptygoBCEAAAhCAAAQgAAEIQAACEIAABCAwUgSu++qfAAAgAElEQVQw70equiksBCAAAQhAAAIQgEDHCWDed7wCyT4EIAABCEAAAhCAAAQgAAEIQAACEIAABHwEMO99ZDgOAQhAAAIQgAAEIACB8Ahg3odXJ+QIAhCAAAQgAAEIQAACEIAABCAAAQhAAAK1EMC8rwUjiUAAAhCAAAQgAAEIQKAVApj3rWDmJhCAAAQgAAEIQAACEIAABCAAAQhAAAIQaJ8A5n37zLkjBCAAAQhAAAIQgAAEyhLAvC9LjusgAAEIQAACEIAABCAAAQhAAAIQgAAEIBA4Acz7wCuI7EEAAhCAAAQgAAEIQMAigHlvwWAXAhCAAAQgAAEIQAACEIAABCAAAQhAAAJ9IoB536fapCwQgAAEIAABCEAAAn0ngHnf9xqmfBCAAAQgAAEIQAACEIAABCAAAQhAAAIjSwDzfmSrnoJDAAIQgAAEIAABCHSQAOZ9ByuNLEMAAhCAAAQgAAEIQAACEIAABCAAAQhAIA8BzPs8lDgHAhCAAAQgAAEIQAACYRDAvA+jHsgFBCAAAQhAAAIQgAAEIAABCEAAAhCAAARqJ4B5XztSEoQABCAAAQhA4P9n706A56juPME/SUgIEAiBBeKWBBgE4r5vMJcxNwLMDeIQN0IIEIhLCMThGbeHbs/Y7u7x1R6ffdhtt9s2btpHTEzExO5sTOxEx0zsTkTPdk90xGzv7M5sR29PX7l+f5Puqr8yqzLrn5mVlfUh4h91Zb338vN+9Vfy/9bLJECAQG0CwvvaaDVMgAABAgQIECBAgAABAgQIECBAYLwCwvvx+uudAAECBAgQIECAQBkB4X0ZLdsSIECAAAECBAgQIECAAAECBAgQmCAB4f0ETZahEiBAgAABAgQITL2A8H7qSwAAAQIECBAgQIAAAQIECBAgQIBAVwWE912dWftFgAABAgQIECDQRQHhfRdn1T4RIECAAAECBAgQIECAAAECBAgQCCEI75UBAQIECBAgQIAAgckREN5PzlwZKQECBAgQIECAAAECBAgQIECAAIFSAsL7Ulw2JkCAAAECBAgQIDBWAeH9WPl1ToAAAQIECBAgQIAAAQIECBAgQKA+AeF9fbZaJkCAAAECBAgQIFC1gPC+alHtESBAgAABAgQIECBAgAABAgQIEGiJgPC+JRNhGAQIECBAgAABAgQKCAjvCyDZhAABAgQIECBAgAABAgQIECBAgMAkCgjvJ3HWjJkAAQIECBAgQGBaBYT30zrz9psAAQIECBAgQIAAAQIECBAgQKDzAsL7zk+xHSRAgAABAgQIEOiQgPC+Q5NpVwgQIECAAAECBAgQIECAAAECBAj0CgjvezXcJ0CAAAECBAgQINBuAeF9u+fH6AgQIECAAAECBAgQIECAAAECBAiMLCC8H5nOGwkQIECAAAECBAg0LiC8b5xchwQIECBAgAABAgQIECBAgAABAgSaERDeN+OsFwIECBAgQIAAAQJVCAjvq1DUBgECBAgQIECAAAECBAgQIECAAIEWCgjvWzgphkSAAAECBAgQIEAgR0B4nwPjaQIECBAgQIAAAQIECBAgQIAAAQKTLiC8n/QZNH4CBAgQIECAAIFpEhDeT9Ns21cCBAgQIECAAAECBAgQIECAAIGpEhDeT9V021kCBAgQIECAAIEJFxDeT/gEGj4BAgQIECBAgAABAgQIECBAgACBPAHhfZ6M5wkQIECAAAECBAi0T0B43745MSICBAgQIECAAAECBAgQIECAAAEClQgI7yth1AgBAgQIECBAgACBRgSE940w64QAAQIECBAgQIAAAQIECBAgQIBA8wLC++bN9UiAAAECBAgQIEBgVAHh/ahy3keAAAECBAgQIECAAAECBAgQIECg5QLC+5ZPkOERIECAAAECBAgQ6BEQ3vdguEuAAAECBAgQIECAAAECBAgQIECgSwLC+y7Npn0hQIAAAQIECBDouoDwvuszbP8IECBAgAABAgQIECBAgAABAgSmVkB4P7VTb8cJECBAgAABAgQmUEB4P4GTZsgECBAgQIAAAQIECBAgQIAAAQIEiggI74so2YYAAQIECBAgQIBAOwSE9+2YB6MgQIAAAQIECBAgQIAAAQIECBAgULmA8L5yUg0SIECAAAECBAgQqE1AeF8brYYJECBAgAABAgQIECBAgAABAgQIjFdAeD9ef70TIECAAAECBAgQKCMgvC+jZVsCBAgQIECAAAECBAgQIECAAAECEyQgvJ+gyTJUAgQIECBAgACBqRcQ3k99CQAgQIAAAQIECBAgQIAAAQIECBDoqoDwvqsza78IECBAgAABAgS6KCC87+Ks2icCBAgQIECAAAECBAgQIECAAAECIQThvTIgQIAAAQIECBAgMDkCwvvJmSsjJUCAAAECBAgQIECAAAECBAgQIFBKQHhfisvGBAgQIECAAAECBMYqILwfK7/OCRAgQIAAAQIECBAgQIAAAQIECNQnILyvz1bLBAgQIECAAAECBKoWEN5XLao9AgQIECBAgAABAgQIECBAgAABAi0REN63ZCIMgwABAgQIECBAgEABAeF9ASSbECBAgAABAgQIECBAgAABAgQIEJhEAeH9JM6aMRMgQIAAAQIECEyrgPB+WmfefhMgQIAAAQIECBAgQIAAAQIECHReQHjf+Sm2gwQIECBAgAABAh0SEN53aDLtCgECBAgQIECAAAECBAgQIECAAIFeAeF9r4b7BAgQIECAAAECBNotILxv9/wYHQECBAgQIECAAAECBAgQIECAAIGRBYT3I9N5IwECBAgQIECAAIHGBYT3jZPrkAABAgQIECBAgAABAgQIECBAgEAzAsL7Zpz1QoAAAQIECBAgQKAKAeF9FYraIECAAAECBAgQIECAAAECBAgQINBCAeF9CyfFkAgQIECAAAECBAjkCAjvc2A8TYAAAQIECBAgQIAAAQIECBAgQGDSBYT3kz6Dxk+AAAECBAgQIDBNAsL7aZpt+0qAAAECBAgQIECAAAECBAgQIDBVAsL7qZpuO0uAAAECBAgQIDDhAsL7CZ9AwydAgAABAgQIECBAgAABAgQIECCQJyC8z5PxPAECBAgQIECAAIH2CQjv2zcnRkSAAAECBAgQIECAAAECBAgQIECgEgHhfSWMGiFAgAABAgQIECDQiIDwvhFmnRAgQIAAAQIECBAgQIAAAQIECBBoXkB437y5HgkQIECAAAECBAiMKiC8H1XO+wgQIECAAAECBAgQIECAAAECBAi0XEB43/IJMjwCBAgQIECAAAECPQLC+x4MdwkQIECAAAECBAgQIECAAAECBAh0SUB436XZtC8ECBAgQIAAAQJdFxDed32G7R8BAgQIECBAgAABAgQIECBAgMDUCgjvp3bq7TgBAgQIECBAgMAECgjvJ3DSDJkAAQIECBAgQIAAAQIECBAgQIBAEQHhfREl2xAgQIAAAQIECBBoh4Dwvh3zYBQECBAgQIAAAQIECBAgQIAAAQIEKhcQ3ldOqkECBAgQIECAAAECtQkI72uj1TABAgQIECBAgAABAgQIECBAgACB8QoI78frr3cCBAgQIECAAAECZQSE92W0bEuAAAECBAgQIECAAAECBAgQIEBgggSE9xM0WYZKgAABAgQIECAw9QLC+6kvAQAECBAgQIAAAQIECBAgQIAAAQJdFRDed3Vm7RcBAgQIECBAgEAXBYT3XZxV+0SAAAECBAgQIECAAAECBAgQIEAghCC8VwYECBAgQIAAAQIEJkdAeD85c2WkBAgQIECAAAECBAgQIECAAAECBEoJCO9LcdmYAAECBAgQIECAwFgFhPdj5dc5AQIECBAgQIAAAQIECBAgQIAAgfoEhPf12WqZAAECBAgQIECAQNUCwvuqRbVHgAABAgQIECBAgAABAgQIECBAoCUCwvuWTIRhECBAgAABAgQIECggILwvgGQTAgQIECBAgAABAgQIECBAgAABApMoILyfxFkzZgIECBAgQIAAgWkVEN5P68zbbwIECBAgQIAAAQIECBAgQIAAgc4LCO87P8V2kAABAgQIECBAoEMCwvsOTaZdIUCAAAECBAgQIECAAAECBAgQINArILzv1XCfAAECBAgQIECAQLsFhPftnh+jI0CAAAECBAgQIECAAAECBAgQIDCygPB+ZDpvJECAAAECBAgQINC4gPC+cXIdEiBAgAABAgQIECBAgAABAgQIEGhGQHjfjLNeCBAgQIAAAQIECFQhILyvQlEbBAgQIECAAAECBAgQIECAAAECBFooILxv4aQYEgECBAgQIECAAIEcAeF9DoynCRAgQIAAAQIECBAgQIAAAQIECEy6gPB+0mfQ+AkQIECAAAECBKZJQHg/TbNtXwkQIECAAAECBAgQIECAAAECBKZKQHg/VdNtZwkQIECAAAECBCZcQHg/4RNo+AQIECBAgAABAgQIECBAgAABAgTyBIT3eTKeJ0CAAAECBAgQINA+AeF9++bEiAgQIECAAAECBAgQIECAAAECBAhUIiC8r4RRIwQIECBAgAABAgQaERDeN8KsEwIECBAgQIAAAQIECBAgQIAAAQLNCwjvmzfXIwECBAgQIECAAIFRBYT3o8p5HwECBAgQIECAAAECBAgQIECAAIGWCwjvWz5BhkeAAAECBAgQIECgR0B434PhLgECBAgQIECAAAECBAgQIECAAIEuCQjvuzSb9oUAAQIECBAgQKDrAsL7rs+w/SNAgAABAgQIECBAgAABAgQIEJhaAeH91E69HSdAgAABAgQIEJhAAeH9BE6aIRMgQIAAAQIECBAgQIAAAQIECBAoIiC8L6JkGwIECBAgQIAAAQLtEBDet2MejIIAAQIECBAgQIAAAQIECBAgQIBA5QLC+8pJNUiAAAECBAgQIECgNgHhfW20GiZAgAABAgQIECBAgAABAgQIECAwXgHh/Xj99U6AAAECBAgQIECgjIDwvoyWbQkQIECAAAECBAgQIECAAAECBAhMkIDwfoImy1AJECBAgAABAgSmXkB4P/UlAIAAAQIECBAgQIAAAQIECBAgQKCrAsL7rs6s/SJAgAABAgQIEOiigPC+i7NqnwgQIECAAAECBAgQIECAAAECBAiEEIT3yoAAAQIECBAgQIDA5AgI7ydnroyUAAECBAgQIECAAAECBAgQIECAQCkB4X0pLhsTIECAAAECBAgQGKuA8H6s/DonQIAAAQIECBAgQIAAAQIECBAgUJ+A8L4+Wy0TIECAAAECBAgQqFpAeF+1qPYIECBAgAABAgQIECBAgAABAgQItERAeN+SiTAMAgQIECBAgAABAgUEhPcFkGxCgAABAgQIECBAgAABAgQIECBAYBIFhPeTOGvGTIAAAQIECBAgMK0CwvtpnXn7TYAAAQIECBAgQIAAAQIECBAg0HkB4X3np9gOEiBAgAABAgQIdEhAeN+hybQrBAgQIECAAAECBAgQIECAAAECBHoFhPe9Gu4TIECAAAECBAgQaLeA8L7d82N0BAgQIECAAAECBAgQIECAAAECBEYWEN6PTOeNBAgQIECAAAECBBoXEN43Tq5DAgQIECBAgAABAgQIECBAgAABAs0ICO+bcdYLAQIECBAgQIAAgSoEhPdVKGqDAAECBAgQIECAAAECBAgQIECAQAsFhPctnBRDIkCAAAECBAgQIJAjILzPgfE0AQIECBAgQIAAAQIECBAgQIAAgUkXEN5P+gwaPwECBAgQIECAwDQJCO+nabbtKwECBAgQIECAAAECBAgQIECAwFQJCO+narrtLAECBAgQIECAwIQLCO8nfAINnwABAgQIECBAgAABAgQIECBAgECegPA+T8bzBAgQIECAAAECBNonILxv35wYEQECBAgQIECAAAECBAgQIECAAIFKBIT3lTBqhAABAgQIECBAgEAjAsL7Rph1QoAAAQIECBAgQIAAAQIECBAgQKB5AeF98+Z6JECAAAECBAgQIDCqgPB+VDnvI0CAAAECBAgQIECAAAECBAgQINByAeF9yyfI8AgQIECAAAECBAj0CAjvezDcJUCAAAECBAgQIECAAAECBAgQINAlAeF9l2bTvhAgQIAAAQIECHRdQHjf9Rm2fwQIECBAgAABAgQIECBAgAABAlMrILyf2qm34wQIECBAgAABAhMoILyfwEkzZAIECBAgQIAAAQIECBAgQIAAAQJFBIT3RZRsQ4AAAQIECBAgQKAdAsL7dsyDURAgQIAAAQIECBAgQIAAAQIECBCoXEB4XzmpBgkQIECAAAECBAjUJiC8r41WwwQIECBAgAABAgQIECBAgAABAgTGKyC8H6+/3gkQIECAAAECBAiUERDel9GyLQECBAgQIECAAAECBAgQIECAAIEJEhDeT9BkGSoBAgQIECBAgMDUCwjvp74EABAgQIAAAQIECBAgQIAAAQIECHRVQHjf1Zm1XwQIECBAgAABAl0UEN53cVbtEwECBAgQIECAAAECBAgQIECAAIEQgvBeGRAgQIAAAQIECBCYHAHh/eTMlZESIECAAAECBAgQIECAAAECBAgQKCUgvC/FZWMCBAgQIECAAAECYxUQ3o+VX+cECBAgQIAAAQIECBAgQIAAAQIE6hMQ3tdnq2UCBAgQIECAAAECVQsI76sW1R4BAgQIECBAgAABAgQIECBAgACBlggI71syEYZBgAABAgQIECBAoICA8L4Akk0IECBAgAABAgQIECBAgAABAgQITKKA8H4SZ82YCRAgQIAAAQIEplVAeD+tM2+/CRAgQIAAAQIECBAgQIAAAQIEOi8gvO/8FNtBAgQIECBAgACBDgkI7zs0mXaFAAECBAgQIECAAAECBAgQIECAQK+A8L5Xw30CBAgQIECAAAEC7RYQ3rd7foyOAAECBAgQIECAAAECBAgQIECAwMgCwvuR6byRAAECBAgQIECAQOMCwvvGyXVIgAABAgQIECBAgAABAgQIECBAoBkB4X0zznohQIAAAQIECBAgUIWA8L4KRW0QIECAAAECBAgQIECAAAECBAgQaKGA8L6Fk2JIBAgQIECAAAECBHIEhPc5MJ4mQIAAAQIECBAgQIAAAQIECBAgMOkCwvtJn0HjJ0CAAAECBAgQmCYB4f00zbZ9JUCAAAECBAgQIECAAAECBAgQmCoB4f1UTbedJUCAAAECBAgQmHAB4f2ET6DhEyBAgAABAgQIECBAgAABAgQIEMgTEN7nyXieAAECBAgQIECAQPsEhPftmxMjIkCAAAECBAgQIECAAAECBAgQIFCJgPC+EkaNECBAgAABAgQIEGhEQHjfCLNOCBAgQIAAAQIECBAgQIAAAQIECDQvILxv3lyPBAgQIECAAAECBEYVEN6PKud9BAgQIECAAAECBAgQIECAAAECBFouILxv+QQZHgECBAgQIECAAIEeAeF9D4a7BAgQIECAAAECBAgQIECAAAECBLokILzv0mzaFwIECBAgQIAAga4LCO+7PsP2jwABAgQIECBAgAABAgQIECBAYGoFhPdTO/V2nAABAgQIECBAYAIFhPcTOGmGTIAAAQIECBAgQIAAAQIECBAgQKCIgPC+iFIz27wdQvgzPwzUgBpQA2pADaiBCaiB/6mZwyO9ZAgI7zNQPEWAAAECBAgQIECAAAECBAgQIECgCwLC+/bM4qdDCIkfBmpADagBNaAG1MAE1MCftOcQaupGIryfuim3wwQIECBAgAABAgQIECBAgAABAtMiILxvz0wL7315wZc31IAaUANqQA1MSg0I78d3DCm8H5+9ngkQIECAAAECBAgQIECAAAECBAjUKiC8r5W3VOPCe4HNpAQ2xqlW1YAaUANqQHhf6jCv0o2F95VyaowAAQIECBAgQIAAAQIECBAgQIBAewSE9+2ZC+G9IEQYpgbUgBpQA2pgUmpAeD++Y0jh/fjs9UyAAAECBAgQIECAAAECBAgQIECgVgHhfa28pRoX3gtsJiWwMU61qgbUgBpQA8L7Uod5lW4svK+UU2MECBAgQIAAAQIECBAgQIAAAQIE2iMgvG/PXAjvBSHCMDWgBtSAGlADk1IDwvvxHUMK78dnr2cCBAgQIECAAAECBAgQIECAAAECtQoI72vlLdW48F5gMymBjXGqVTWgBtSAGhDelzrMq3Rj4X2lnBojQIAAAQIECBAgQIAAAQIECBAg0B4B4X175mJgeH/wIWuSm2950Q8DNaAG1IAaUANqoPYaOO2Mq4d9QUN4P75jSOH9+Oz1TIAAAQIECBAgQIAAAQIECBAgQKBWAeF9rbylGh8Y3q897oLkzY/91A8DNaAG1IAaUANqoPYauPHm54X3pQ7jGt1YeN8ot84IECBAgAABAgQIECBAgAABAgQINCcgvG/OelhPwnthTO1hjC+A+AKMGlADakANFKkB4f2ww7axvi68Hyu/zgkQIECAAAECBAgQIECAAAECBAjUJyC8r8+2bMvCe+G98F4NqAE1oAbUQCtqQHhf9jCu0e2F941y64wAAQIECBAgQIAAAQIECBAgQIBAcwLC++ash/UkvBfYtCKwKbIi0zZW7qoBNaAGul0Dwvthh21jfV14P1Z+nRMgQIAAAQIECBAgQIAAAQIECBCoT0B4X59t2ZaF98J74b0aUANqQA2ogVbUgPC+7GFco9sL7xvl1hkBAgQIECBAgAABAgQIECBAgACB5gSE981ZD+tJeC+waUVgYzVtt1fTml/zqwbUQJEaEN4PO2wb6+vC+7Hy65wAAQIECBAgQIAAAQIECBAgQIBAfQLC+/psy7YsvBfeC+/VgBpQA2pADbSiBoT3ZQ/jGt1eeN8ot84IECBAgAABAgQIECBAgAABAgQINCcgvG/OelhPwnuBTSsCmyIrMm1j5a4aUANqoNs1ILwfdtg21teF92Pl1zkBAgQIECBAgAABAgQIECBAgACB+gSE9/XZlm1ZeC+8F96rATWgBtSAGmhFDQjvyx7GNbq98L5Rbp0RIECAAAECBAgQIECAAAECBAgQaE5AeN+c9bCehPcCm1YENlbTdns1rfk1v2pADRSpAeH9sMO2sb4uvB8rv84JECBAgAABAgQIECBAgAABAgQI1CcgvK/PtmzLwnvhvfBeDagBNaAG1EArakB4X/YwrtHthfeNcuuMAAECBAgQIECAAAECBAgQIECAQHMCwvvmrIf1JLwX2LQisCmyItM2Vu6qATWgBrpdA8L7YYdtY31deD9Wfp0TIECAAAECBAgQIECAAAECBAgQqE9AeF+fbdmWhffCe+G9GlADakANqIFW1IDwvuxhXKPbC+8b5dYZAQIECBAgQIAAAQIECBAgQIAAgeYEhPfNWQ/rSXgvsGlFYGM1bbdX05pf86sG1ECRGhDeDztsG+vrwvux8uucAAECBAgQIECAAAECBAgQIECAQH0Cwvv6bMu2LLwX3gvv1YAaUANqQA20ogaE92UP4xrdXnjfKLfOCBAgQIAAAQIECBAgQIAAAQIECDQnILxvznpYT8J7gU0rApsiKzJtY+WuGlADaqDbNSC8H3bYNtbXhfdj5dc5AQIECBAgQIAAAQIECBAgQIAAgfoEhPf12ZZtWXgvvBfeqwE1oAbUgBpoRQ0I78sexjW6vfC+UW6dESBAgAABAgQIECBAgAABAgQIEGhOQHjfnPWwnoT3AptWBDZW03Z7Na35Nb9qQA0UqQHh/bDDtrG+LrwfK7/OCRAgQIAAAQIECBAgQIAAAQIECNQnILyvz7Zsy8J74b3wXg2oATWgBtRAK2pAeF/2MK7R7YX3jXLrjAABAgQIECBAgAABAgQIECBAgEBzAsL75qyH9SS8F9i0IrApsiLTNlbuqgE1oAa6XQPC+2GHbWN9XXg/Vn6dEyBAgAABAgQIECBAgAABAgQIEKhPQHhfn23ZloX3wnvhvRpQA2pADaiBVtSA8L7sYVyj2wvvG+XWGQECBAgQIECAAAECBAgQIECAAIHmBIT3zVkP60l4L7BpRWBjNW23V9OaX/OrBtRAkRoQ3g87bBvr68L7sfLrnAABAgQIECBAgAABAgQIECBAgEB9AsL7+mzLtiy8F94L79WAGlADakANtKIGhPdlD+Ma3V543yi3zggQIECAAAECBAgQIECAAAECBAg0JyC8b856WE/Ce4FNKwKbIisybWPlrhpQA2qg2zUgvB922DbW14X3Y+XXOQECBAgQIECAAAECBAgQIECAAIH6BIT39dmWbVl4L7wX3qsBNaAG1IAaaEUNCO/LHsY1ur3wvlFunREgQIAAAQIECBAgQIAAAQIECBBoTkB435z1sJ6E9wKbVgQ2VtN2ezWt+TW/akANFKkB4f2ww7axvi68Hyu/zgkQIECAAAECBAgQIECAAAECBAjUJyC8r8+2bMvCe+G98F4NqAE1oAbUQCtqQHhf9jCu0e2F941y64wAAQIECBAgQIAAAQIECBAgQIBAcwLC++ash/UkvBfYtCKwKbIi0zZW7qoBNaAGul0Dwvthh21jfV14P1Z+nRMgQIAAAQIECBAgQIAAAQIECBCoT0B4X59t2ZaF98J74b0aUANqQA2ogVbUgPC+7GFco9sL7xvl1hkBAgQIECBAgAABAgQIECBAgACB5gSE981ZD+tJeC+waUVgYzVtt1fTml/zqwbUQJEaEN4PO2wb6+vC+7Hy65wAAQIECBAgQIAAAQIECBAgQIBAfQLC+/psy7YsvBfeC+/VgBpQA2pADbSiBoT3ZQ/jGt1eeN8ot84IECBAgAABAgQIECBAgAABAgQINCcgvG/OelhPwnuBTSsCmyIrMm1j5a4aUANqoNs1ILwfdtg21teF92Pl1zkBAgQIECBAgAABAgQIECBAgACB+gSE9/XZlm1ZeC+8F96rATWgBtSAGmhFDQjvyx7GNbq98L5Rbp0RIECAAAECBAgQIECAAAECBAgQaE5AeLfc/WcAACAASURBVN+c9bCehPcCm1YENlbTdns1rfk1v2pADRSpAeH9sMO2sb4uvB8rv84JECBAgAABAgQIECBAgAABAgQI1CcgvK/PtmzLwnvhvfBeDagBNaAG1EArakB4X/YwrtHthfeNcuuMAAECBAgQIECAAAECBAgQIECAQHMCwvvmrIf1JLwX2LQisCmyItM2Vu6qATWgBrpdA8L7YYdtY31deD9Wfp0TIECAAAECBAgQIECAAAECBAgQqE9AeF+fbdmWhffCe+G9GlADakANqIFW1IDwvuxhXKPbC+8b5dYZAQIECBAgQIAAAQIECBAgQIAAgeYEhPfNWQ/rSXgvsGlFYGM1bbdX05pf86sG1ECRGhDeDztsG+vrwvux8uucAAECBAgQIECAAAECBAgQIECAQH0Cwvv6bMu2LLwX3gvv1YAaUANqQA20ogaE92UP4xrdXnjfKLfOCBAgQIAAAQIECBAgQIAAAQIECDQnILxvznpYT8J7gU0rApsiKzJtY+WuGlADaqDbNSC8H3bYNtbXhfdj5dc5AQIECBAgQIAAAQIECBAgQIAAgfoEhPf12ZZtWXgvvBfeqwE1oAbUgBpoRQ0I78sexjW6vfC+UW6dESBAgAABAgQIECBAgAABAgQIEGhOQHjfnPWwnoT3AptWBDZW03Z7Na35Nb9qQA0UqQHh/bDDtrG+LrwfK7/OCRAgQIAAAQIECBAgQIAAAQIECNQnILyvz7Zsy8J74b3wXg2oATWgBtRAK2pAeF/2MK7R7YX3jXLrjAABAgQIECBAgAABAgQIECBAgEBzAsL75qyH9SS8F9i0IrApsiLTNlbuqgE1oAa6XQPC+2GHbWN9XXg/Vn6dEyBAgAABAgQIECBAgAABAgQIEKhPQHhfn23ZloX3wnvhvRpQA2pADaiBVtSA8L7sYVyj2wvvG+XWGQECBAgQIECAAAECBAgQIECAAIHmBIT3zVkP60l4L7BpRWBjNW23V9OaX/OrBtRAkRoQ3g87bBvr68L7sfLrnAABAgQIECBAgAABAgQIECBAgEB9AsL7+mzLtiy8F94L79WAGlADakANtKIGhPdlD+Ma3V543yi3zggQIECAAAECBAgQIECAAAECBAg0JyC8b856WE/Ce4FNKwKbIisybWPlrhpQA2qg2zUgvB922DbW14X3Y+XXOQECBAgQIECAAAECBAgQIECAAIH6BIT39dmWbVl4L7wX3qsBNaAG1IAaaEUNCO/LHsY1ur3wvlFunREgQIAAAQIECBAgQIAAAQIECBBoTkB435z1sJ6E9wKbVgQ2VtN2ezWt+TW/akANFKkB4f2ww7axvi68Hyu/zgkQIECAAAECBAgQIECAAAECBAjUJyC8r8+2bMvCe+G98F4NqAE1oAbUQCtqQHhf9jCu0e2F941y64wAAQIECBAgQIAAAQIECBAgQIBAcwLC++ash/UkvBfYtCKwKbIi0zZW7qoBNaAGul0Dwvthh21jfV14P1Z+nRMgQIAAAQIECBAgQIAAAQIECBCoT0B4X59t2ZaF98J74b0aUANqQA2ogVbUgPC+7GFco9sL7xvl1hkBAgQIECBAgAABAgQIECBAgACB5gSE981ZD+tJeC+waUVgYzVtt1fTml/zqwbUQJEaEN4PO2wb6+vC+7Hy65wAAQIECBAgQIAAAQIECBAgQIBAfQLC+/psy7YsvBfeC+/VgBpQA2pADbSiBoT3ZQ/jGt1eeN8ot84IECBAgAABAgQIECBAgAABAgQINCcgvG/OelhPwnuBTSsCmyIrMm1j5a4aUANqoNs1ILwfdtg21teF92Pl1zkBAgQIECBAgAABAgQIECBAgACB+gSE9/XZlm1ZeC+8F96rATWgBtSAGmhFDQjvyx7GNbq98L5Rbp0RIECAAAECBAgQIECAAAECBAgQaE5AeN+c9bCehPcCm1YENlbTdns1rfk1v2pADRSpAeH9sMO2sb4uvB8rv84JECBAgAABAgQIECBAgAABAgQI1CcgvK/PtmzLwnvhvfBeDagBNaAG1EArakB4X/YwrtHthfeNcuuMAAECBAgQIECAAAECBAgQIECAQHMCwvvmrIf1JLwX2LQisCmyItM2Vu6qATWgBrpdA8L7YYdtY31deD9Wfp0TIECAAAECBAgQIECAAAECBAgQqE9AeF+fbdmWhffCe+G9GlADakANqIFW1IDwvuxhXKPbC+8b5dYZAQIECBAgQIAAAQIECBAgQIAAgeYEhPfNWQ/rSXgvsGlFYGM1bbdX05pf86sG1ECRGhDeDztsG+vrwvux8uucAAECBAgQIECAAAECBAgQIECAQH0Cwvv6bMu2LLwX3gvv1YAaUANqQA20ogaE92UP4xrdXnjfKLfOCBAgQIAAAQIECBAgQIAAAQIECDQn0KXw/q4QwuXN0VXek/BeYNOKwKbIikzbWLmrBtSAGuh2DQjvKz/Oq7JB4X2VmtoiQIAAAQIECBAgQIAAAQIECBAg0CKBLoX320MISQjh3RDCMS0yLjoU4b3wXnivBtSAGlADaqAVNSC8L3r4NpbthPdjYdcpAQIECBAgQIAAAQIECBAgQIAAgfoFuhjexwD/r0MI74QQltZPWFkPwnuBTSsCG6tpu72a1vyaXzWgBorUgPC+suO7OhoS3tehqk0CBAgQIECAAAECBAgQIECAAAECLRDoangfA/z48+chhLiPC1pgPWwIwnvhvfBeDagBNaAG1EArakB4P+ywbayvC+/Hyq9zAgQIECBAgAABAgQIECBAgAABAvUJdD28T0P8fxNCOK8+xkpaFt4LbFoR2BRZkWkbK3fVgBpQA92uAeF9Jcd2dTUivK9LVrsECBAgQIAAAQIECBAgQIAAAQIExiwwLeF9DPH/PoTwlRDCIWM2z+teeC+8F96rATWgBtSAGmhFDQjv8w7XWvG88L4V02AQBAgQIECAAAECBAgQIECAAAECBKoXmKbwPl2F/5chhLdCCEuq55xTi8J7gU0rAhurabu9mtb8ml81oAaK1EDHw/uVIYRFczpqG++bhffj9dc7AQIECBAgQIAAAQIECBAgQIAAgdoEpjG8T0P8Pw0h3BVCmFebbrmGhffCe+G9GlADakANqIFW1EBHw/vdf3Zoti2EEL/IubTcYVqrthbet2o6DIYAAQIECBAgQIAAAQIECBAgQIBAdQLTHN6nIf5PQggnVUc6ckvCe4FNKwKbIisybWPlrhpQA2qg2zXQsfA+flHzphDCH4cQ0uM/4f3Ih6zeSIAAAQIECBAgQIAAAQIECBAgQIBAXQLC+5//EffvQghfDCHsXxd0gXaF98J74b0aUANqQA2ogVbUQIfC+1NCCD/tCe2F9wUOSm1CgAABAgQIECBAgAABAgQIECBAgMB4BIT3/7ACK/4x9y9+Ng3xdKq7jmE6hPcCm1YENlbTdns1rfk1v2pADRSpgQ6E9weEED4TQvjbjOA+HvNZeT+Gg11dEiBAgAABAgQIECBAgAABAgQIECAwWEB43x/ep6ux/kMI4crBdJW/2unwfuOmzyf3PfCJnX5efOXbAusxfWnhjbd/vNN8xDl64MF3zMmY5qRIoNalbdTgdIfIr2z/XubvoEef+DW/g1rwO2iCw/uFIYR4fPvfckL79FhPeF/5oawGCRAgQIAAAQIECBAgQIAAAQIECBCYq4DwPju8T/+w+8MQwtq5Ihd8f6fD+2PXXpCa9t3eefebQpoxhTSvvfEHfXORhhyLFi02J2Oaky4F80X2RQ1Od3j/8GOfzvwdtHLl8X4HteB30ISG99eGEP739N+zIbfC+4IHqDYjQIAAAQIECBAgQIAAAQIECBAgQKA5AeH94PA+Bgt/8/5pVz9Q87TUFt4/9sSvJ039PLHpc5mhi/C+fSGd4LR9c1Ik8O7SNmpwumtQeN/u+Z+w8P6oEMLvDQnrZ39ZRHhf84Gt5gkQIECAAAECBAgQIECAAAECBAgQKC8gvB8e3qd/7P2/3j8N64LyzIXeUVt4X/KP2en+jnS7xx57C+9bsGKySMArOG13cFZkDid9GzU43TUovG/3/E9IeL8shPBWCOF/jHCsI7wvdHhqIwIECBAgQIAAAQIECBAgQIAAAQIEmhQQ3hcP79Mw+49CCJfXMEnC+wkJvSc9ME3HLzhtd3CWzlOXb9VgO2vw8Y2fTa5f98xOP4889pnML2eNWqPC+3bOfzqfLQ/vdwkhPBpC+PMRQvv0eE54X8PBrCYJECBAgAABAgQIECBAgAABAgQIEJibgPC+fHif/tH32yGEVXPj73u38F54X2kwlgYwebeC03YHZ3nz1qXn1WA7a/Cyyx9I/53ruz3/wtsq/R0lvG/n/Ke/Y1oc3l8UQvi3cwjt07oW3vcdhnpAgAABAgQIECBAgAABAgQIECBAgEAbBIT3o4f38Y+/8TSt74QQ9qxgMoX3wvtKg7E0gMm7FZy2OzjLm7cuPa8G21mDwvt2zkvTn/0WhveHhBC+WEFoL7yv4KBVEwQIECBAgAABAgQIECBAgAABAgQI1CMgvJ9beJ/+Afg/hxA2hBDmz2GahPfCe+G9Gmi0BpoOA2f3J7xvZ0gsvG/nvMz+/NT9uEXh/ZIQwpshhL+qMLiPx29W3s/hoNVbCRAgQIAAAQIECBAgQIAAAQIECBCoR0B4X014n4b4/zqEcPaIU9V4eP/Bo85Ibrtje6U/d93zZmb4+NCjn8rs5/kXfydz+7pDCe3/NBGcCujG/TlQg+2swabC+5e2fSfz34UHHnzHvwst+CJTC8L7eSGEm0II/6ni0D49ZhPej3jA6m0ECBAgQIAAAQIECBAgQIAAAQIECNQnILyvNryPfxD++xDC10MI8fSuZf5rPLw/+5x1ApIWBCTjClAFp+0MTsdVD+PoVw22swabCu/HUXP6LF5zYw7vTw0h/MuaQnvhfZmjU9sSIECAAAECBAgQIECAAAECBAgQINCogPC++vA+/aPwX/xsJreFEBYXnFHh/RQH6eMIlASnxUOscczPNPSpBttZg8L7ds5L078TxhTeHxhC+EwI4e9qDu7jsZqV9wUPUG1GgAABAgQIECBAgAABAgQIECBAgEBzAsL7+sL7NMSPp3u9q8CUCu+F942eCUFwKqBrOgyc3Z8abGcNCu/bOS+zPz91P244vF8UQojHpP+9gdA+PT4T3hc4OLUJAQIECBAgQIAAAQIECBAgQIAAAQLNCgjv6w/v0z8S/0EI4fgB0yu8F94L79VAozVQd/g3rH3hfTtDYuF9O+dl2Oep6tcbDO+vDiH8xwZD+/S4THg/4KDUSwQIECBAgAABAgQIECBAgAABAgQIjEdglPA+/pH1pBb+/OoY/vCb/gG46O3fvH862OUZ0y28F9w2GtwKTgV0VYd9ZdtTg+2sQeF9O+el7Odrrts3EN6vCSF8b4zHbsL7jINRTxEgQIAAAQIECBAgQIAAAQIECBAgMF6BUcL7okG17fJX9f/X908Pu0vP9Hc6vH/19R8kL7/63Z1+Xn/zD0sF1q+98d5ObcR2t+94N7edZ5/7enLVNU8kJ5x4cbLigMOTvZetSPbYY+9kv/0OS1auOiE5+ZQPJx+99eXkxVe+k9vGXEOQ9P1vvP3j5IlNn0vW3fRccv4FtyZHrzkrOejgo5J99z0oWbJkn2Tp0uUzY1y1+sRk7XEXJBdfuj5Zf98/TmLImbZR1W0XgtMdb/0o2fDQryQxbDz+hA8lBx+yJtlnnwOTPfZYmixZsizZd9+Dk1WrT0hOO/2qJAZRz7/4O3NyfPX1dzPrL9Z3VfMyu53tO36Y2ee27d8bqc8u1uCOt35cqdHsOYiP8+b+9TffG2keZvcRa+i+Bz6RXHX148npZ1wzU7fx91Ws591223Pmd8TBBx+dHHnkackpp35k5nda/F0yu51RHr+y/Xt9fh+6+O7Mf7/PPvfGvu2yfqf3PhfbHTSe+Pnt3T69v+217w9836A28157adt3kltu25bEfTjyg6cny5cfmuy1dPmM7bJlK5IVK1Ynxxx7bhL3ff39H0/ivzV5bY3y/Cj7Gv9Nir+3Tj3tyiTOfayF3XffK/nA8kOSww5bm6w97sLkuhs2J88+/41Kx5ruX43h/T4hhHdCCH87xuA+1rjwvucg1F0CBAgQIECAAAECBAgQIECAAAECBNohILzPD9gzw4uK/9D870MIV7xfCp0O749de0Gm5513v1kqdLj62icz2znjzOt2aueRx381Oeros5J58+Zlvmf2XM6fPz85/oSLk6e3fHWnttIwY5TbrS99K7nmuieTNcecMxMUze63yONFixbPBHYbN32+srFNcnj/zHNfS84+Z91MSF/EL91m3rz5yRFHnprc+8AvjeQY5zFtq/d2zz33SWI4N0p9DHvPUUefmdln/NLJsPemr3e9Bp/c/MVMoxh+pwZzvY3evXOe3v/orS+N3MejT/xacuFFdySHHnpsMn/+gsz2037ybpftc0By2Yc3JMOC8kH7H9vIa38uz8d2B/X78GOfzux35crjB75vUJuzX9vw0CeTNcecmyxYsEtmX3n7t3i3JcmZZ12fxC9/zW5zlMcPPfLPMvs/5JA1O7Uf+4xf0FiwYGHme7LGHL+kFPsYZWx576khvI9fWNwQQvg/s/ZhDM8J79vx/yJGQYAAAQIECBAgQIAAAQIECBAgQIBAj4DwfrzhffqH+W+HEL486A/XcRV23h/Yhz2f124MP4e9t6rXmwzv42r+c867OYlBbd6+D3p+4cJdkxtv3jpnmxgaHX7EyUn8UsCg/sq8FvfptDOuTqpYmTqJ4X0MKM86+4aRg85e6/jFjue2/lapeY6rd3fZJTtQu3v9x0q1VeSztfXFb+bu6wMPvjO0v2mpwUkL7z/8kYdmzgrRW49zvR/PNnHTR18YWhNZddfF8H7L1t+cObPJXF1j6H/Bhbcn8QwYWXZFnysa3l97/VNJ/LLWKOOO/z7EscazaxQd16DtKg7vLw4h/K+j7FeN7xHe9/wPgbsECBAgQIAAAQIECBAgQIAAAQIECLRDQHjfjvA+/qF+4Oljhfc/vwbxsJX38VTiMTCv4o/98VT7g4KNYa9dfMk9lYwja1/iKZ/neqrkSQvvH338V5O9996/UtN4+YQHH/5kqXmOn8WsOYmn7R9WE2Vfv+LKRzL7iqf5LhLQTUsNTlp4Hy/nkFVDVTwXLxERLyNQpta6Ft7HM2vE08tX4Zm2ceBBRybxCwFlXHu3HRbex8/zmWddV8mYTzn1iuSNt38y8ljTcVcU3h8RQvh66tiyW+F9O/5fxCgIECBAgAABAgQIECBAgAABAgQIEOgREN63J7wf+Ed74f3w8D6G0asPrya4jwFDXMX40KOfGjkAqTM4jeOLQXY8HXoatJS9naTw/u71b+eueJ9rGLRw4eLk/g3DV7GnvnEsWX3ussuimet3p9tVcbv//qsy+4q1VaT9aalB4X3oq5MY3hapj3SbLoX38VIsZU+Rn/V5znpu6d77JfGSHalbmdth4X08o0hWn6M+d+31m0caZ+8+zTG83+Nnx5rbQgh/Neo+NPA+4X3P/xC4S4AAAQIECBAgQIAAAQIECBAgQIBAOwSE98L7Of+Bv/eP/Xn3mzhtflxxmvXH/gMOOCI5/4Jbk+tueDq5464dyZ13v5HcsO7Z5PQzrkn22usDme9J29lvv8OS1954bySjrOB03rx5yX77r0xOPOnS5IqPPDwzlnj688c3fjZ56ukvJfEa0Pfc+7Hk8is2JGuOOWfoNY/jdY6LrMDOmpdJCe/ve+ATQx323feg5Iwzr01uvuXF5IEHfznZuOnzyRObPpesv//jyZVXPZYcfsQpAy9fsHjxHjP+WU6zn4vXto/XuE9rpPf2uhvmHpil/cXrofe2nd6PNVQ0QJxLDcbrqE9KDXYhvI9fIjls5XEz11mPv5/i74EY+MZ9i/UczxBx6+3bkvPOvyWJq8DTesi7LRPeXvShO2d+H8bfifHn4EPWZLYff5em2xS5je2m9Zx1W/U17x9+9FNDv+QTz7YRr2V/6+2vJg8+8k+TJ5/6wsyXd2JQfeza84e+P5414eVXvztwv7L2dVB4H+cqax7jGTbiavxrrts0M/fxi0NxnOddcGuyfL/DMt+TtrPrrrvP6UwBcR9GDO/nhRDuCiH8WTqWFt8K79vx/yJGQYAAAQIECBAgQIAAAQIECBAgQIBAj4DwXnhfOoTICiaGPVd3eL/PPgfuFGSsOebcmeBr0Nhef/O95CNXPjIwsInByaA28l7rDU5XrjohiafhL3va5Rde/t3kQxffPXAl6fXrnhlpfJMQ3m9+9ivJbrvtudPcpmHQgQd9MIkrbYt8gSGGdEceeVpuWysOODx5/c0/LGQZA9R0DL23hx56bKH359VM7/Nnnb0us49Vq08s3EdVNXjRxXe1ugYnNbyP1zY/8aTLktvvfD3ZvuPdwvMa6yReRmJQPcfwduuL3yzVZlp/l13+QGbtnX/hbSO1l7Y7+7bK8D4G6nstXZ457vgZjWfGiF+YGvYZf/GVbycnnHhxbjuxrfhvy+x9GfY4L7zfc699k4ULd+3r7+CDj07u3/BPBvYRT4sfv8yxZEn2F4niOE897cqBbQwb8wjh/ekhhH/V+zux5fd/EkJ4b0J//ngE2yU9x//uEiBAgAABAgQIECBAgAABAgQIECDQUgHhvfB+Tn/cH/bH//T1usP73j9iz5+/IFl303Ol9uuee/9R7uruGBCn+1Hm9pJL702OO/6i5PGN/3yk9/f29cSTn809S0A8lXP8EkLv9kXutz28j4H8YYet7Qu10nmOq8/POfem0vsd2zz3vJsz24xtx3CviF38IkA6ltm38QwKRdoYtE2czz32WJrZRwzUBr2397WqazAGjbP3Nz4edw1OWnh/yKHHJBdfuj55adt3Cs9l77ym92OAe9XVjyfx85A1L7HW023L3E5ieB/PBJBlEJ+Lq+03PvWFUhY33Lgl1zW2edsd20u1lxfezx5znLMiX0ZK5zP+vsk7E8iiRbslr77+g1LjTNuNtyXC+4NCCF8MIfz97P3xuP+SFmP2EN639H/GDIsAAQIECBAgQIAAAQIECBAgQIBAr4DwforD+w8edcZMABFDiLn+xNOT9/7Rf/b9psL7GGLFVayz+y/yOK5wz/vDejwFe5E2ercZtsKzd9si9zc9/aVk8W5LMsd4x13l97nt4X08dXjefMQV5UXM8raJp9jPanv33fdKtm3/XqG24+rYrDYuvGjwqcLzxtT7fJzPrLbjSu1tr32/0Phie5XX4ObfSOIlBrLGNs4anLTwvup5+chVj2bOSfx9ET/nvbVV5P6khfebn/ly7mUxYoAdz1JQZL9nb5PnGus/nj5/x1s/LtxukfA+nuFi9hiKPI6Xgsn6TMbnynzZZ3ZfBcL7Pw0hbAkh/L95/XteeN/7Px3uEyBAgAABAgQIECBAgAABAgQIECAwXEB4347w/n8LIfxg0B+51x53wUh/1I9/jB/UblWvxeu4z/7Df+/jpsL7uBq7t98y97fv+GESw9ssk6uv2Thyu2XGMGzbuMo2a3zxNM/D3jv79TaH96+98d7Mau6sfV173IVJXHE8e3/KPI77vv/+qzItr73+qUJtX3Pdk5nvX7p0eamVs1njjqflztr3k0/5cKGxZbVZ1XNXtrAGJy28r2ou0nbi5yGu5s+qmbvueat0zUxaeH/a6Vdl7nv0GPWyJ6ntUUefldv2LbdtK2w7LLxfffhJpb4MkI4vvT1s5XGZ4zz5lCsKjzFtK70tEN7/bVbNea5VgX1vXVh5P/z/i2xBgAABAgQIECBAgAABAgQIECBAYOwCwvvxhvd/8bMK2BZC2DWE8OlBf/AW3v90JoC4+trswDTa7b770tLXjU5DivQ279TLcwlA0raruI2nU89a+bz33vuXDmjaHN7HUCzr8xCvDf3c1t8qva9Z9vF60Vl9xBA0a/vZz8VTnu+yy8LMNu594JcKtTG7zfj4hZd/N/f68g88+M7I7Wb1NcpzbazBaQ/v4zzecdeOzFq88KI7StfMJIX323e8m+y66+6Z+x4veVJmdXzW52Hzs1/J/ZwfeeRphW0HhffxjDGbNv9G4bayxnn9umcyDVasWD1yuwXC+8w+s36veq4Vgb7wfuz/22UABAgQIECAAAECBAgQIECAAAECBIYLCO/HE97H68LG68Ou6Jki4f3Hfh7QZwUT6XODwvtRTzmcth1v805XvuKAw0cOQHrbr+L+ccdflBmYbH3xm6XG2Obw/sgPnp65j2efs67UPg7yjqFevFb77FBp3rz5yYuvfLtQP/FLNbPfHx+feNJlhd6fNb6rrnkis81ly1bMeUV/Vn+jPNe2GhTe/3Tm2uZZXyaJK7rLzvEkhfd5X1qIn8P4BZ2y+561/UknX575mfz574rvFOpjUHh/1NFnFmoja2zpc1ue/0bmGOfPXzDSpRNiu8L7VgTumfOa9e9OgeeE9z0H3e4SIECAAAECBAgQIECAAAECBAgQaKvAMSGE+zry8+0Cf7is8o+go7b1r0MIZ2YUhPB+juH9xqe+MOcA5OktX82c17iqPw1Jxn178aXrM8d4591vlhpjW8P7eE33BQuyV7Q//OinSu3jsLk65dSPZFrGQHDYe+Prd69/O/P9CxcuTrZt/16hNmb3c+BBR2a2efEl94zU3uz2q3jcthoU3v/8i08fWH7ITrUTr/le9jITkxTen3nWdTvtczwWiGcoiZdCqaLe45k08o4vbr391UJ9DArvb/ro1kJtDNqXOMeLd1uSOc5nn//GSO0L74X3GceqniJAgAABAgQIECBAgAABAgQIECBAgEBhge15f1xvyfN/GkK4K4QwL2ePhPdzCO93223PSlYlv/H2j5N4CuPZNRNX7dVt9AAAIABJREFULw4KTpp87drrN+80vjje+HyZcbQ1vM8LxJcs2ad0CDnMIy+cKhqU73jrR8mee+6TOR/rbtxSaj7iWDdu+nxmW7Emn3nua6XbG7b/o77ethoU3v88vF+58vjM+nml5BdJJim8P+DA7C+7nHjSpZV9XuJZOuIXuGb/uxAfn3PuTYX6GRTexy+NjfpZ7H1f3hd/ntj0uZHaz/v9mOXguYkI+q28zzkA9zQBAgQIECBAgAABAgQIECBAgAABAvUItDW8/8sQwlshhD2H7Hbj4f3xJ1ycPPbEr1fyE8Oz3hBh9v1j12afXrzsavG80+avWn3iwP5nj2fQ47zrJ7/6+ruV9RH7j9c2v++BTyTXXLcpufyKB5MLLrw9OevsG5LTz7hm4E/c16yg5PIrNpQaX1vD+0suuy9z/4448tRS+zdojtPX8lbUxhX56TbDbs87/5bM8a5cdULhNtI+zj3/o5ltrVpdvq20zUG3XanBLoX3MSje/MyXk9vvfC258qrHkvh5OP/C2wb+Tkh/Z+y1dHlm/cRTqg+qg9mvTUp4H788k3WpgPj78cqrHy+1z7MNZj+O17fP+r27+vCTC/WTF97vuniPQu+fPZ6sx3n/Nmx46JMj9SG8n4hAPrMus2o1hCC8H3Ig7mUCBAgQIECAAAECBAgQIECAAAECBKoVaGN4H0/lv6rgbjYe3ld5/fCsIKH3ubrD+9h+b39zub/nXvtm/jH8hZe/Nac+Xn/zveTu9R9LTjr5smTJkmWZfeT8wb3QtjHgK7PfbQ3vjz/hQ5n7e+rpVyUvv/rdSn8efPiTmX0dc+x5hS2ffOoLmW3E1fJlVtTGIDKeXSCrBmKIVmZu87btag1Oengfv0QRv8QTv6SxcOGumTWQVRdFnyu78npSwvvnX/ydXKsHHvzlSj4z6Wcp/n7N8t5334MK9ZMX3i9btqLQ+9NxDLo9es1ZmWO8596PjdRHgfD+77JMPNfa0F94X/CA3GYECBAgQIAAAQIECBAgQIAAAQIECFQj0Kbw/t+EEM4vuVvC+zmcNv/U064cKZzICkKW5qxe3frSaOF9vO7yZR/ekOyxR/Zpl6sKOs4487pSBm0N7w89bG1mAFWVU5F2yq50P+jgozLHXPT0+7EO8y4XsGjR4mTba98vNbez67rrNTip4f3mZ7+SxDOgzJ8/P7N+itRqkW3KrryelPB+0+bfyHV7butvzekzM/szdMO6ZzP72n33vQr1kxfex9P+z+5r1Mdrjjk3c4zxS2OjtFkgvP/PIYR3Qgh/U6QObTP2UF94X/LA3OYECBAgQIAAAQIECBAgQIAAAQIECMxNoA3h/Z+HEDaGEBaMsCvC+zmE92WD60FBRpXhfbyG+fL9DssMVKoOMk474+pSAU1bw/t99jmwEa9B/itWrC5lec11T2aOedk+ByRvvP2TQm2tPe7CzDZOPuXDhd6fV9PTUIOTGN5fd8PmWlbZZ9X1Aw++U6qGJiW8zwvEo8Err/5+qX3O+/ykz996+6uZn8/4xYs33v7x0L7yxnrIIWuGvjcdw7DbMYT3f/L+sc7RIYTfz6o9z409sO+tW+H9CAfn3kKAAAECBAgQIECAAAECBAgQIECAwOgC4wzv//r91WdLRx9+EN53LLx/6NFPJbvuunvvH85rvd+V8H7PPbNPHd9kCLTffoeVCtRe2vad3Gtvb3joV4a29dK238t9f9ngtTfgi4HhNNTgpIX3F150Z62/C2Z/VsrW0KSE93FF+ex9jY/jJSt2vDU8UO/9rAy7v/7+j2f2Ffsr8kWBjof36aHP1SGE/5g1Jy197t+HEP5oin52TyfKLQECBAgQIECAAAECBAgQIECAAAECBJoQGFd4/24I4dgKdlB436Hw/rkXfjvZfffBp8nfdfEeyQePOiOJp1a/6aNbk/s3/JMkngb6mee+lmzZ+pu513bPW+XdlfA+noZ63EFP2fA+Bn9rj7sgc9ynnPqRoeH9tdc/lfneeD3sIqt6s4LHIjW4uCM1OEnh/bobt2TOdW/NxzM2nHTyZcmVVz2WxBXf8YtAT2/5yszvhnj5jpdf/W7mz6rVJ2a23dXw/o67dmTu7/z5C4Z+5rI+M4Oei4a9c9R7/4WXf3dof1MS3sdDoUXvn4Hov/catfT+XL5wWcFhnyYIECBAgAABAgQIECBAgAABAgQIECDQbYGmw/v/EEK4qkJS4X2Hwvtjjs2+9nAMMA444Ijk1tu3JTve+tHQwCcrTMq79nJXwvs999o3MyQ74shTk9PPuKaRn4svXV96bvKuWR9Xvr/6+rsD24unzs4Kt+IXO7JqoMhz01SDkxLev/Dyt5JFi3bLnOu4Wvykky9PnnjysyPP+erDT85su6vh/aBAffuOwZ+5Ip+h3m3uvPuNTNv4uX39zfeGztkUhffpYdGBIYTPhBD+Lut3W0ueE96ns+WWAAECBAgQIECAAAECBAgQIECAAAECNQg0Fd7/3yGELSGEXSveB+F9R8L7GL7lBRPx+uWvvTE86OkNjWbfv+LKRzLb70p4v+++B2Xu39XXPjk0IJtt1eTj+GWMvFP+33zLi7lj3/T0lzL3N4a5T2/5au77Bu3btNVgE+F9XAmf9bn+6K0vFZ6j8y+8LbONBQsWJrfdsb1wO3lzf8CBR2a239Xw/vGN+b9r4xkK8pxGef6mj76Qabto0eJC/UxheJ8eIp0aQviXWZ+dFjwnvE9nyS0BAgQIECBAgAABAgQIECBAgAABAgRqEKg7vI+rx74YQti/hrHHJoX3HQnvzznv5syQ58CDjiy0QnNYsHThRXdktt+V8H7V6hMy9+/c824uFJIN86vz9fPOvyVz7IcfcUru2C+48PbM90SHUcc6bTWYG96vWD2y4Wz7NceckzlPRcP7ePmDvC93xGvMz+5vlMd7771/5hi7Gt4/+/w3Mvc3hsIbN32+EtN0Hq74yMOZfS1durxQP1Mc3sfjm3khhJtCCP+pBYF97zwK72s6oNUsAQIECBAgQIAAAQIECBAgQIAAAQIEokCd4f2PQggn1swsvO9IeL//itW94cAv7t9+52uFQp40LMq7PeHES37RZm8Q0pXwPp6doHe/0vtHrzmrEr881yqef/KpL2SOfd68+cmWrb+50/h3vPXjZK+lyzPfc+PNz++0fdExTlsNxuvBp3XSe/uBDxw8suFs68NWHpfZR9HwPobJvWNL7++6eI/ktTf+YM7jjGf02GWXhZl9dDW8f/X1HyTxDBWpZe/toLNdzJ7bIo9PPOnSzH4OPOiDheZuysP79PBpj5/d2fazsxf9f71zNcb7wvt0ZtwSIECAAAECBAgQIECAAAECBAgQIECgBoE6wvs/CSHc9f6qsRqG3Nek8L4D4f0bb/8kiafAnh1GLFiwS1LVNZiXLVuxU/uxv66E91dd/Xjm/sVrx8/1kgNFQrq5bnPQwUdljj9rdfX6+z+euW08Ffe2175fKBScPd5prMEXX/l2pmNc6T7bZ9THe+61b2YfRcP7+OWd2b8X4uPjjr+okjE+9OinMtuPfXQ1vI9zuXfO78Oqz9Sx336HZfqedPLlheZPeN93vHPI+2cyyjTN+pzU9Jzwvm9aPCBAgAABAgQIECBAgAABAgQIECBAgEC1AlWG938RQngphLBbtUMc2JrwvgPh/Suv/n5mGBFXV48aGva+77kXfjuz/RhsdCW8f/ixT+fu470P/FIljr2mVd+/5ronM8f/geWHJDFY7+0v7ywK8ewDvduVuT+NNRi/1JEV7sUv0rz+5h+ObJm6P73lq5ntxz6LhvfXXv9UZhvnX3DrnMcXx3nFlY9kth/HWDa8v/yKBzPbOq+isaaueZ/1lSuPL2xyzLHnZY51RYWXTNjy/DeSePaMrBq76ponCo1VeJ95/HNRCOHfZrk29JzwPnNaPEmAAAECBAgQIECAAAECBAgQIECAAIFqBKoK778dQjismiGVakV434HwfvMzX84MePbbf2WhgCcNtfJuL7nsvsz2Y9DRlfA+Xht8yZJ9Mvfz+BMursQxz7eK51/a9p3Msy/EOYqro9M+Xtn+vWThwl0z97Ns2Jq2GW+ntQbj2QqyAr+NT33hF+a9TmXuX3v95sy2Y39Fw/tLLr03s42sMzKUGVvcNn4pJG9leBxj2XrK+wLKWWevm7Nl775VEd7njTXud1XXvb/swxsy5y72ES+V0btPefeF97nHQ/PfP8PRf4meDf8I73OnxQsECBAgQIAAAQIECBAgQIAAAQIECBCYu8Bcw/v/OYRw7tyHMXILwvsOhPfxuuZZ4UMVp+9+/c33krxTd8c+uxLex/ArhoRZjnH1axVhbF7AVtXza4+7IHP8p59xzS+CvhvWPZu5TbwsQvwCw6hjmdYaPPSwtZme1697ZmTLOAdxLpbnnDI91mjR8D5vNXsVp3e/N+fyC+lnqGx4f9NHX8i0nMsZIbLquYrw/tnnvp573fuzz5n7lw3imRv23fegTI999jlwp7NpZO1nfE54P/TYaFkI4a0Qwv9I67aBW+H90GmxAQECBAgQIECAAAECBAgQIECAAAECBEYXGDW8/7MQwvoQQlz9Nc7/hPcdCO+37/hhZsgTQ4jntv7WnELEiy+5J7ft2H6Xwvunnv5SbiC3avWJlZwKPS9kq+L5u9e/nTlXi3dbksQaiX0ctvK4zG3iPM9lDNNag2edfUOm5+FHnDwnz5tveSmz3TRYLBrer7txS2Y7K1edMKfxvfbGHyT7778qs+10jGXD+7vueSuzvbmOdXZdVxHexzbjuNJ97b1dsGCXZNPTX5qT70cGXI4gnk1h9j7lPRbeFz68OiqE8Hu981jjfeF94WmxIQECBAgQIECAAAECBAgQIECAAAECBMoLlA3v/zqE8E4IYa/yXdXyDuF9B8L7GNzEVfZZYcPlV2woHPTMDoAeeewzyfz5CzLbTfvqUngf9z/vevBxf2NQO9uoqsc73vrRnNuObeSd+v+W27Ylm5/9SuaXE+bNm5fE66vPdV+msQZvvuXFzM9HPFvDo0/82kimcVX34sV7ZLabfu6Khvf3b3gns534uZ7LnJ93/i2Z7abji7dlw/tnnvtaZpu77rp78tob741kmVXTVYX3t9/5euZ4476vPvykkb/s8/SWryS75sx/vOTF1pe+VdhCeF/6uOmSEMIf9dZxDfeF96WnxRsIECBAgAABAgQIECBAgAABAgQIECBQXKBMeP9uCGFN8aYb2VJ435Hw/qSTL8sMkuKq6xgGZoVYg5579PFfTXbffWlmm71hRtfC+3j69xgW9u5j7/0zzryu0iDxwUf+abLmmHOTq6/ZWHqOsuYvL1T94FFnJB+6+O7M/Vq1em6rsNNxTGMNbtv+vWThwuzr3h908FHJ9h3vlprXGKh/YPkhmfPUW4dFw/sYeueN7+g1Z410qYRLL79/6PjiWMuG92+8/ZPcLy3EU+qndTbX26rC+3hpgwMOOCLX4sSTLi3t+/wLv50s2+eA3DbLXu5AeD/ScdTCEMLGEMJ/6/3MVXhfeD/StHgTAQIECBAgQIAAAQIECBAgQIAAAQIEigkUCe//XQjh0mLNNb6V8L4j4f1td2zPDXxiGLj5mS8XCr9igHbjzc/vFGDHlcRZ4UXXwvsYDOatpk73P4ayZYPJ3sDxhZe/lVx3w+ak93rpV139eKH56W0n6/7Gp76QOU/z58/PPTtDnO+stso+V10N/niiavDkU67INI/1svrwk5MXX/nOUN/4ubv19leTJUuW9bWVd+aLouF9nMNjjj2vr820juPtqaddmcRT4BeZ65e2/d7M9r3vj/fzfjeM8hnJG2tciR4vAZBe/qHIePO2qSq8j+0/9Oincvc/2qw55pxk64vfLOS74aFfGRjc77V0eRK/LJK3X1nPC+/ndEx1QAjhcyGEv5td83N8LLyf07R4MwECBAgQIECAAAECBAgQIECAAAECBAYLDArv/2sIYUsIYdHgJsb6qvC+I+H9jrd+nOy332G5Id2iRYuTCy+6M3n2+W9khj8vbftOEle3HnLoMTu1sWDBwuTsc9bt9HwMMLoY3scQ7LwLbs3c397Q5pBD1sysmH/iyc8OXGEbV9PG69Ffetl9yarVJyYxSO9tJ96vKryPY49fLpjdft7jWBfbXvt+Zk1khYGDnqusBg9Zs9P421yDT2z63MDLS+y2254zc79x0+f7nKNXvDZ6nPuDDz56p31eccDhyXHHX7TT83Euy4T38fT98dIIeTUQf2/c9NGtyauv/6BvfOlcP7n5i0lcbZ91Jo447ljTWW2PEt7Hz0lWW+lzsV7j5y66xC9NnH7GNTv9XPShOzP3I92fKsP72OaHLrln4Jjj/F908V3Jk099YadxxUtdrL//48mJJ1028EsA8Usco3gK7ys5vjo5hPDTtAYruBXeVzItGiFAgAABAgQIECBAgAABAgQIECBAgEC2QFZ4H1dpfTGEsDz7La16VnjfkfA+hkh3r//YwJAuDR2W7r3fTOAWT9e+ctUJyfLlh2YGyun2V1/7ZLLupucyA6quhvdxJXS8xn1qMOw2npp82bIVM8F5vN71AQceOfN40Cn4e9usMry/5ronC4/75FM+vFOgmIaco9zWVYNxn9pcgzEw7p3PvPsxhN1zr32TPfbYe+D2MSh/6pl/kcT5yWqrTHgf5zHvkga9bcex7b9idXLkkafNrBiPIXkca+82vfd3332vZPOzX0mOOPLUzG1GCZvjFxqyvsjQ2++w+/G084Nqt+rwPp4+P+9LFrPHGs32339VsnLl8ckHPnBwEq9hP3ub2Y/jFy/imToG7VPea8L7yo635oUQbgoh/PHs+RnhsfC+smnREAECBAgQIECAAAECBAgQIECAAAECBHYWmB3evxdCOH7nzVr7jPC+Q+F9DHDi6u4RwoTc91x40R0zoVGbg9O84KqK56+86rEkrvqu0jSrrSrD+3gWhaJjvn/DOyOFgoNsq6/Bn6+kbnMNxlPPzzV0TusiniI+BszRuKrwfvuOd0udkSEdS97tokW7zZwyPo6xyvA+tvf0lq8mcbV6Xt/Dnm86vI9jfv3NP5xZPT9sbGVfj1+ouO6Gp0f+jArvKz/22iOE8FoI4S/LzmXP9sL7yqdFgwQIECBAgAABAgQIECBAgAABAgQIEPgHgTS8/z9CCHf9w9MTc09437HwPq4Yv/iSewaegrknRMgNyBYs2CWJq51jMBV/2hycpmOs6/axJ369smA2y37vZSuSex/4pV9YV7Efa4+7IHdu0zHEMwXEVcNV9NfbRrU1uOkX42t7DcYvTRx22Nqh7ql/1m2shY09p1evKryP8/PCy9+aOdNGVr9lnot103sJgKrD+zjWLc9/Y+SxjiO8T+s/fgknnoWjjGfetrEWRjl7QTqWeCu8r+1Y7OD3z7D093nzN+B54X1t06JhAgQIECBAgAABAgQIECBAgAABAgQIhPDczxC2hRAWTyiG8L5j4X0a3Nx7/8dnToc/IEDIDZjiNax7A8TYZtuD03S/67qNgfStt29LDjn0mFy3MtZLlixLTjn1iuSBB3+5lgD9rnveGjrO+CWPurxiu9NYg6+98d7M9c132aXc2RriCuszzrw22bb9e31zUmV4H+ckXmM9znvRyzn01vQuuyxKzjnv5mTba9/vG2Md4X1alzF8Pu30q5J99jmw0CVB4njHGd7HcT/z3NdmLlMQ57TXr+j9xbstmZmjV1//QZ9zalLmVnhf+5HZBSGE/6Xo3L6/nfC+9mnRAQECBAgQIECAAAECBAgQIECAAAEC0ywQr4M6yf/VFt5fv+6ZJOsnhgllwoe5bBtXM2eNIa7qLNPuk5u/mNnOgw9/slQ7g/qM17DOGms83fWg9w16La6qvuOu15Nj114w8Brb8ZrK8TrM5553c/L4xn+e2V+8/nbW+DY8VM4gXtM6q5345YBB+9Km12I9XPbhDTOnC49B27DgJga58Vria4+7MImn4X9i0+eS+GWAOvcphrRZzr3PbX3xm7WOIe7ftNbg8y/+TnL5FQ8mq1afkAwKcffbf2Vy4UV3Js/m/E6KK6975yy9//SWr8xp7l5+9bsz44vXXh90iYV4TfbDjzhlpm5ffOXbmX3evf5jmWOMBlXWePzSwCOPfWbmSyG33bE9WXfjlsx+b7ntlYH9vvDy72a+76573hz4vrL78vwLvz3jFv0WLRq8Gn/Jkn1mfj/EfweqCO3TsW596VuZ+3rn3W9Utq/r7/94Zh9btv7mSH3cePPzw36n/knLDroWhBAeDCH8l2H/FgjvWzZzhkOAAAECBAgQIECAAAECBAgQIECAAIEWCtQW3qfhgdufn3Z+3A4xLI4rQh9+9FPJ3evfTu64a0dy3wOfmAmSt+/44Ughy7j3qS39x4AshorR8/Y7X0vuvPvNJF5LPp5u/9nnvp7ELyy0ZazjHMe01mAMnTc9/aUkftknfpnmnnv/0czn8JVZq+zHNTfx8//kU19INjz0KzPji2F8/NJArN06Lqswrv0cZ7/xd0D8gkY0jr9/45cP1t/3j2fqIF7OYJxja1vfExjep4d2e//sLExvhRD+akiIb+V9KuaWAAECBAgQIECAAAECBAgQIECAAAECBHYSEN4XOG1+28IN42nHFyLMg3lQA2pADVRbAxMc3qcHWB8MIXx7QIAvvE+l3BIgQIAAAQIECBAgQIAAAQIECBAgQIDATgLCe+G9VZ9qQA2oATWgBlpRAx0I79MDrUtCCP8uI8QX3qdCbgkQIECAAAECBAgQIECAAAECBAgQIEBgJwHhvcCmFYGN1avVrl7lyVMNqIFJrIEOhffxgGthCGFjCOH/6Qnxhfc7HYp6ggABAgQIECBAgAABAgQIECBAgAABAgRSAeG98F54rwbUgBpQA2qgFTXQsfA+PdbaN4TwTgjhb0MIwvtUxS0BAgQIECBAgAABAgQIECBAgAABAgQI7CQgvBfYtCKwmcQVosZsZbMaUANqoNoa6Gh4nx58nRxC2DV94JYAAQIECBAgQIAAAQIECBAgQIAAAQIECMwWEN4L74X3akANqAE1oAZaUQMdD+9nH4N5TIAAAQIECBAgQIAAAQIECBAgQIAAAQIE+gSE9wKbVgQ2Vq9Wu3qVJ081oAYmsQaE933HaB4QIECAAAECBAgQIECAAAECBAgQIECAwJQJCO+F98J7NaAG1IAaUAOtqAHh/ZQdhdpdAgQIECBAgAABAgQIECBAgAABAgQIEOgTEN4LbFoR2EziClFjtrJZDagBNVBtDQjv+47RPCBAgAABAgQIECBAgAABAgQIECBAgACBKRMQ3gvvhfdqQA2oATWgBlpRA8L7KTsKtbsECBAgQIAAAQIECBAgQIAAAQIECBAg0CcgvBfYtCKwsXq12tWrPHmqATUwiTUgvO87RvOAAAECBAgQIECAAAECBAgQIECAAAECBKZMQHgvvBfeqwE1oAbUgBpoRQ0I76fsKNTuEiBAgAABAgQIECBAgAABAgQIECBAgECfgPBeYNOKwGYSV4gas5XNakANqIFqa0B433eM5gEBAgQIECBAgAABAgQIECBAgAABAgQITJmA8F54L7xXA2pADagBNdCKGhDeT9lRqN0lQIAAAQIECBAgQIAAAQIECBAgQIAAgT4B4b3AphWBjdWr1a5e5clTDaiBSawB4X3fMZoHBAgQIECAAAECBAgQIECAAAECBAgQIDBlAsJ74b3wXg2oATWgBtRAK2pAeD9lR6F2lwABAgQIECBAgAABAgQIECBAgAABAgT6BIT3AptWBDaTuELUmK1sVgNqQA1UWwPC+75jNA8IECBAgAABAgQIECBAgAABAgQIECBAYMoEhPfCe+G9GlADakANqIFW1IDwfsqOQu0uAQIECBAgQIAAAQIECBAgQIAAAQIECPQJCO8FNq0IbKxerXb1Kk+eakANTGINCO/7jtE8IECAAAECBAgQIECAAAECBAgQIECAAIEpExDeC++F92pADagBNaAGWlEDwvspOwq1uwQIECBAgAABAgQIECBAgAABAgQIECDQJyC8F9i0IrCZxBWixmxlsxpQA2qg2hoQ3vcdo3lAgAABAgQIECBAgAABAgQIECBAgAABAlMmILwX3gvv1YAaUANqQA20ogaE91N2FGp3CRAgQIAAAQIECBAgQIAAAQIECBAgQKBPQHgvsGlFYGP1arWrV3nyVANqYBJrQHjfd4zmAQECBAgQIECAAAECBAgQIECAAAECBAhMmYDwXngvvFcDakANqAE10IoaEN5P2VGo3SVAgAABAgQIECBAgAABAgQIECBAgACBPgHhvcCmFYHNJK4QNWYrm9WAGlAD1daA8L7vGM0DAgQIECBAgAABAgQIECBAgAABAgQIEJgyAeG98F54rwbUgBpQA2qgFTUgvJ+yo1C7S4AAAQIECBAgQIAAAQIECBAgQIAAAQJ9AsJ7gU0rAhurV6tdvcqTpxpQA5NYA8L7vmM0DwgQIECAAAECBAgQIECAAAECBAgQIEBgygSE98J74b0aUANqQA2ogVbUgPB+yo5C7S4BAgQIECBAgAABAgQIECBAgAABAgQI9AkI7wU2rQhsJnGFqDFb2awG1IAaqLYGhPd9x2geECBAgAABAgQIECBAgAABAgQIECBAgMCUCQjvhffCezWgBtSAGlADragB4f2UHYXaXQIECBAgQIAAAQIECBAgQIAAAQIECBDoExDeC2xaEdhYvVrt6lWePNWAGpjEGhDe9x2jeUCAAAECBAgQIECAAAECBAgQIECAAAECUyYgvBfeC+/VgBpQA2pADbSiBoT3U3YUancJECBAgAABAgQIECBAgAABAgQIECBAoE9AeC+waUVgM4krRI3ZymY1oAbUQLU1ILzvO0bzgAABAgQIECBAgAABAgR/2tBpAAAJ50lEQVQIECBAgAABAgSmTEB4L7wX3qsBNaAG1IAaaEUNCO+n7CjU7hIgQIAAAQIECBAgQIAAAQIECBAgQIBAn4DwXmDTisDG6tVqV6/y5KkG1MAk1oDwvu8YzQMCBAgQIECAAAECBAgQIECAAAECBAgQmDIB4b3wXnivBtSAGlADaqAVNSC8n7KjULtLgAABAgQIECBAgAABAgQIECBAgAABAn0CwnuBTSsCm0lcIWrMVjarATWgBqqtAeF93zGaBwQIECBAgAABAgQIECBAgAABAgQIECAwZQLCe+G98F4NqAE1oAbUQCtqQHg/ZUehdpcAAQIECBAgQIAAAQIECBAgQIAAAQIE+gSE9wKbVgQ2Vq9Wu3qVJ081oAYmsQaE933HaB4QIECAAAECBAgQIECAAAECBAgQIECAwJQJCO+F98J7NaAG1IAaUAOtqAHh/ZQdhdpdAgQIECBAgAABAgQIECBAgAABAgQIEOgTEN4LbFoR2EziClFjtrJZDagBNVBtDQjv+47RPCBAgAABAgQIECBAgAABAgQIECBAgACBKRMQ3gvvhfdqQA2oATWgBlpRA8L7KTsKtbsECBAgQIAAAQIECBAgQIAAAQIECBAg0CcgvBfYtCKwsXq12tWrPHmqATUwiTUgvO87RvOAAAECBAgQIECAAAECBAgQIECAAAECBKZMQHgvvBfeqwE1oAbUgBpoRQ0I76fsKNTuEiBAgAABAgQIECBAgAABAgQIECBAgECfgPBeYNOKwGYSV4gas5XNakANqIFqa0B433eM5gEBAgQIECBAgAABAgQIECBAgAABAgQITJmA8F54L7xXA2pADagBNdCKGhDeT9lRqN0lQIAAAQIECBAgQIAAAQIECBAgQIAAgT4B4b3AphWBjdWr1a5e5clTDaiBSawB4X3fMZoHBAgQIECAAAECBAgQIECAAAECBAgQIDBlAsJ74b3wXg2oATWgBtRAK2pAeD9lR6F2lwABAgQIECBAgAABAgQIECBAgAABAgT6BIT3AptWBDaTuELUmK1sVgNqQA1UWwPC+75jNA8IECBAgAABAgQIECBAgAABAgQIECBAYMoEhPfCe+G9GlADakANqIFW1IDwfsqOQu0uAQIECBAgQIAAAQIECBAgQIAAAQIECPQJCO8FNq0IbKxerXb1Kk+eakANTGINCO/7jtE8IECAAAECBAgQIECAAAECBAgQIECAAIEpExDeC++F92pADagBNaAGWlEDwvspOwq1uwQIECBAgAABAgQIECBAgAABAgQIECDQJyC8F9i0IrCZxBWixmxlsxpQA2qg2hoQ3vcdo3lAgAABAgQIECBAgAABAgQIECBAgAABAlMmILwX3gvv1YAaUANqQA20ogaE91N2FGp3CRAgQIAAAQIECBAgQIAAAQIECBAgQKBPQHgvsGlFYGP1arWrV3nyVANqYBJrQHjfd4zmAQECBAgQIECAAAECBAgQIECAAAECBAhMmYDwXngvvFcDakANqAE10IoaEN5P2VGo3SVAgAABAgQIECBAgAABAgQIECBAgACBPgHhvcCmFYHNJK4QNWYrm9WAGlAD1daA8L7vGM0DAgQIECBAgAABAgQIECBAgAABAgQIEJgyAeG98F54rwbUgBpQA2qgFTUgvJ+yo1C7S4AAAQIECBAgQIAAAQIECBAgQIAAAQJ9AsJ7gU0rAhurV6tdvcqTpxpQA5NYA8L7vmM0DwgQIECAAAECBAgQIECAAAECBAgQIEBgygSE98J74b0aUANqQA2ogVbUgPB+yo5C7S4BAgQIECBAgAABAgQIECBAgAABAgQI9AkI7wU2rQhsJnGFqDFb2awG1IAaqLYGhPd9x2geECBAgAABAgQIECBAgAABAgQIECBAgMCUCQjvhffCezWgBtSAGlADragB4f2UHYXaXQIECBAgQIAAAQIECBAgQIAAAQIECBDoExDeC2xaEdhYvVrt6lWePNWAGpjEGhDe9x2jeUCAAAECBAgQIECAAAECBAgQIECAAAECUyYgvBfeC+/VgBpQA2pADbSiBoT3U3YUancJECBAgAABAgQIECBAgAABAgQIECBAoE9AeC+waUVgM4krRI3ZymY1oAbUQLU1ILzvO0bzgAABAgQIECBAgAABAgQIECBAgAABAgSmTEB4L7wX3qsBNaAG1IAaaEUNCO+n7CjU7hIgQIAAAQIECBAgQIAAAQIECBAgQIBAn4DwXmDTisDG6tVqV6/y5KkG1MAk1oDwvu8YzQMCBAgQIECAAAECBAgQIECAAAECBAgQmDIB4b3wXnivBtSAGlADaqAVNSC8n7KjULtLgAABAgQIECBAgAABAgQIECBAgAABAn0CwnuBTSsCm0lcIWrMVjarATWgBqqtAeF93zGaBwQIECBAgAABAgQIECBAgAABAgQIECAwZQLCe+G98F4NqAE1oAbUwP/fvh3bKhADURTdUqACKkDUQYyohyIpBOJJjEZaNN7xCQisn3wf3cDiiSkaMN4v9gp1XQIECBAgQIAAAQIECBAgQIAAAQIECBAIAsZ7g80Ug41fr+7761WePDWggSM2YLwPbzQHAgQIECBAgAABAgQIECBAgAABAgQIEFhMwHhvvDfea0ADGtCABqZowHi/2CvUdQkQIECAAAECBAgQIECAAAECBAgQIEAgCBjvDTZTDDZH/IWo/9kvmzWgAQ3s24DxPrzRHAgQIECAAAECBAgQIECAAAECBAgQIEBgMQHjvfHeeK8BDWhAAxqYogHj/WKvUNclQIAAAQIECBAgQIAAAQIECBAgQIAAgSAwHO9P58vn8Xz5MNCABjSgAQ1o4O8NXG/3z7Zto887vGIcCBAgQIAAAQIECBAgQIAAAQIECBAgQIBAI4HheP/jC/TRl+v+Nh4f+PDRgAY0oAEN5Bsw3jd6hLoKAQIECBAgQIAAAQIECBAgQIAAAQIECEQB431+ODC2MNOABjSgAQ3UNGC8j+84JwIECBAgQIAAAQIECBAgQIAAAQIECBBoJGC8rxkfjD7cNaABDWhAA/kGjPeNHqGuQoAAAQIECBAgQIAAAQIECBAgQIAAAQJRwHifHw6MLcw0oAENaEADNQ0Y7+M7zokAAQIECBAgQIAAAQIECBAgQIAAAQIEGgkY72vGB6MPdw1oQAMa0EC+AeN9o0eoqxAgQIAAAQIECBAgQIAAAQIECBAgQIBAFDDe54cDYwszDWhAAxrQQE0Dxvv4jnMiQIAAAQIECBAgQIAAAQIECBAgQIAAgUYCxvua8cHow10DGtCABjSQb8B43+gR6ioECBAgQIAAAQIECBAgQIAAAQIECBAgEAWM9/nhwNjCTAMa0IAGNFDTgPE+vuOcCBAg0FbgC9gKpCAkxx0WAAAAAElFTkSuQmCC"}}},{"id":"94f82fab","cell_type":"code","source":"data_7_9.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"c1ed8613-e8b8-4b50-b674-147a2f58a03e","cell_type":"code","source":"data_4_9.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"3b3f8560-5211-4215-ab48-c57d4bd71af9","cell_type":"code","source":"data_10_clean.info()","metadata":{},"outputs":[],"execution_count":null},{"id":"99bd614b","cell_type":"code","source":"# def objective_CB(trial, X_train, y_train):\n#     params = {\n#         'iterations': trial.suggest_int('iterations', 100, 500),\n#         'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.2),\n#         'depth': trial.suggest_int('depth', 3, 7),\n#         'l2_leaf_reg': trial.suggest_int('l2_leaf_reg', 1, 10),\n#         'loss_function': 'Logloss',\n#         'random_seed': 42,\n#         'verbose': 0,\n#     }\n#     # skf = StratifiedKFold(n_splits=10, shuffle=True, random_state=42)\n#     # cv = RepeatedStratifiedKFold(n_splits=5, n_repeats=2, random_state=42)\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n\n#     base_model = CatBoostClassifier(**params, task_type=\"GPU\", devices='0', cat_features=cat_features)\n\n#     #Tuning threshold với F1-macro\n#     scorer = make_scorer(f1_score, average='macro')\n#     tuned_model = TunedThresholdClassifierCV(estimator=base_model, scoring=scorer, cv=5) #Sử dụng TunedThresholdClassifierCV để tìm ngưỡng tối ưu cho xác suất dự đoán\n\n#     #Fit mô hình và return best_score\n#     tuned_model.fit(X_train, y_train)\n#     trial.set_user_attr(\"best_threshold\", tuned_model.best_threshold_)\n#     # Lưu threshold tốt nhất vào trial\n#     return tuned_model.best_score_\n","metadata":{},"outputs":[],"execution_count":null},{"id":"804d969a","cell_type":"code","source":"# def objective_CB(trial, X_train, y_train):\n#     params = {\n#     'iterations': trial.suggest_int('iterations', 100, 500),\n#     'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.2),\n#     'depth': trial.suggest_int('depth', 3, 7),\n#     'l2_leaf_reg': trial.suggest_int('l2_leaf_reg', 1, 10),\n        \n#     # Số lượng mẫu tối thiểu để chia nhánh\n#     'min_data_in_leaf': trial.suggest_int('min_data_in_leaf', 1, 20),\n\n#     # Cân bằng dữ liệu mất cân bằng\n#     'scale_pos_weight': trial.suggest_float('scale_pos_weight', 0.5, 1),\n\n#     # Lượng overfitting detection\n#     'od_type': 'Iter',\n#     'od_wait': trial.suggest_int('od_wait', 10, 500),\n\n#     # Early stopping điều kiện tự dừng sớm\n#     # 'use_best_model': True,  # Giữ lại mô hình tốt nhất trên validation\n\n#     # #Số lượng bucket hóa (bining) cho biến liên tục\n#     # 'border_count': trial.suggest_int('border_count', 32, 255),\n\n#     # Tăng tốc\n#     'task_type': 'GPU',\n#     'devices': '0',\n#     # Các tham số khác\n#     'loss_function': 'Logloss',\n#     'random_seed': 42,\n#     'verbose': 0,\n# }\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n\n#     base_model = CatBoostClassifier(**params, cat_features=cat_features)\n\n#     #Tuning threshold với F1-macro\n#     scorer = make_scorer(f1_score, average='macro')\n#     tuned_model = TunedThresholdClassifierCV(estimator=base_model, scoring=scorer, cv=5) #Sử dụng TunedThresholdClassifierCV để tìm ngưỡng tối ưu cho xác suất dự đoán\n\n#     #Fit mô hình và return best_score\n#     tuned_model.fit(X_train, y_train)\n#     trial.set_user_attr(\"best_threshold\", tuned_model.best_threshold_)\n#     # Lưu threshold tốt nhất vào trial\n#     return tuned_model.best_score_\n","metadata":{},"outputs":[],"execution_count":null},{"id":"15ea16e5","cell_type":"code","source":"# #Sử dụng mô hình CatBoost cho các thực nghiệm\n# def CatBoostRun(data_train, train_name, data_test,index, delay_not_delay_ratio):\n#     # 1. Tách train/val từ data_train\n#     X_train = data_train.drop(columns=['label'])\n#     y_train = data_train['label']\n#      # 8. Đánh giá trên tập test (data_test)\n#     X_test = data_test\n#     # X_train, X_val, y_train, y_val = train_test_split(\n#     #     X, y, test_size=0.2, random_state=42, stratify=y\n#     # )\n\n#     # # 2. \n#     n = delay_not_delay_ratio #Undersampling tập train với tỉ lệ 1:n (delay:not delay)\n#     delay_idx = y_train[y_train == 1].index\n#     not_delay_idx = y_train[y_train == 0].sample(n=len(delay_idx)*n, random_state=42).index\n#     undersample_idx = delay_idx.union(not_delay_idx)\n#     X_train = X_train.loc[undersample_idx]\n#     y_train = y_train.loc[undersample_idx]\n\n#     # 3. Xác định các cột categorical và chuyển đổi thành kiểu category\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n#     X_test[cat_features] = X_test[cat_features].astype('category')\n\n#     study = optuna.create_study(direction='maximize')\n#     study.optimize(lambda trial: objective_CB(trial, X_train, y_train), n_trials=30, timeout = 1500)\n#     #optuna được tối ưu dựa trên chỉ số F1-macro được trả về sau mỗi trial trong hàm objective_CB, optuna sẽ tự động học lại từ kinh nghiệm của các trial trước đó để tìm ra tham số tốt nhất cho mô hình CatBoostClassifier\n#     best_params = study.best_params\n#     best_params.update({'loss_function': 'Logloss', 'random_seed': 42, 'verbose': 0})\n\n#     # 6. Train lại với tham số tốt nhất trên toàn bộ train_4_6_full đã undersample\n#     best_model = CatBoostClassifier(**best_params)\n#     best_model.fit(X_train, y_train, cat_features=cat_features)\n#     best_threshold = study.best_trial.user_attrs[\"best_threshold\"]\n#     # 7 Đánh giá\n#     print(\"Best parameters from Optuna:\", best_params)\n#     print(\"Best threshold from Optuna:\", best_threshold)\n#     # Đánh giá trên tập train\n#     # train_metrics = evaluate(best_model, X_train, y_train, name=f\"Train {train_name}\")\n#     # Đánh giá trên tập test\n#     # acc_test, best_precision, best_recall, best_f1, roc_auc = evaluate(best_model, X_test, y_test, name=f\"Test {test_name}\")\n#     #Chọn ngưỡng tối ưu cho xác suất dự đoán\n#     preds_proba = best_model.predict_proba(X_test)[:, 1]\n#     best_preds = (preds_proba >= best_threshold).astype(int)\n#     #Lưu lại kết quả dự đoán ra file csv\n#     submission = pd.DataFrame({\n#         \"ID\": index,\n#         \"label\": best_preds\n#     })\n#     submission.to_csv(f'catboost_submission_{train_name}_ratio_{delay_not_delay_ratio}.csv', index=False)\n#     #Lưu mô hình tốt nhất\n#     best_model.save_model(f'catboost_model_{train_name}_ratio_{delay_not_delay_ratio}.cbm')\n#     #Trả về các chỉ số đánh giá trên tập test và tham số tốt nhất","metadata":{},"outputs":[],"execution_count":null},{"id":"c7ee2cd5","cell_type":"markdown","source":"<a id=\"num_statistics\"></a>\n## <b><span style='color:darkorange'>EXP 1|</span><span style='color:#8502d1'> TRAIN 7-9 TEST 10 </span></b>","metadata":{}},{"id":"46af1eff","cell_type":"code","source":"# # Tiến hành huấn luyện mô hình CatBoost trên bộ dữ liệu 7-9 với tỉ lệ delay:not delay là 1:10 và test trên 10\n# CatBoostRun(data_7_9,\"7-9\", data_10_clean, index_10, delay_not_delay_ratio=10)","metadata":{},"outputs":[],"execution_count":null},{"id":"6e919799","cell_type":"code","source":"# # Tiến hành huấn luyện mô hình CatBoost trên bộ dữ liệu 7-9 với tỉ lệ delay:not delay là 1:20 và test trên 10\n# CatBoostRun(data_7_9,\"7-9\", data_10_clean, index_10, delay_not_delay_ratio=20)","metadata":{},"outputs":[],"execution_count":null},{"id":"571b1e2f-3e95-4df0-a965-ec6d86128669","cell_type":"code","source":"# #Sử dụng mô hình CatBoost cho các thực nghiệm\n# def CatBoostRun_param_specify_custome(data_train, train_name, data_test,index, delay_not_delay_ratio, best_params):\n#     # 1. Tách train/val từ data_train\n#     X_train = data_train.drop(columns=['label'])\n#     y_train = data_train['label']\n#      # 8. Đánh giá trên tập test (data_test)\n#     X_test = data_test\n#     # X_train, X_val, y_train, y_val = train_test_split(\n#     #     X, y, test_size=0.2, random_state=42, stratify=y\n#     # )\n\n#     # # 2. \n#     n = delay_not_delay_ratio #Undersampling tập train với tỉ lệ 1:n (delay:not delay)\n#     delay_idx = y_train[y_train == 1].index\n#     not_delay_idx = y_train[y_train == 0].sample(n=len(delay_idx)*n, random_state=42).index\n#     undersample_idx = delay_idx.union(not_delay_idx)\n#     X_train = X_train.loc[undersample_idx]\n#     y_train = y_train.loc[undersample_idx]\n\n#     # 3. Xác định các cột categorical và chuyển đổi thành kiểu category\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n#     X_test[cat_features] = X_test[cat_features].astype('category')\n\n#     # study = optuna.create_study(direction='maximize')\n#     # study.optimize(lambda trial: objective_CB(trial, X_train, y_train), n_trials=30, timeout=600) #optuna được tối ưu dựa trên chỉ số F1-macro được trả về sau mỗi trial trong hàm objective_CB, optuna sẽ tự động học lại từ kinh nghiệm của các trial trước đó để tìm ra tham số tốt nhất cho mô hình CatBoostClassifier\n#     best_params.update({'loss_function': 'Logloss', 'random_seed': 42, 'verbose': 0})\n\n#     # 6. Train lại với tham số tốt nhất trên toàn bộ train_4_6_full đã undersample\n#     base_model = CatBoostClassifier(**best_params, task_type=\"GPU\", devices='0', cat_features=cat_features)\n\n#     #Tuning threshold với F1-macro\n#     scorer = make_scorer(f1_score, average='macro')\n#     tuned_model = TunedThresholdClassifierCV(estimator=base_model, scoring=scorer, cv=5) #Sử dụng TunedThresholdClassifierCV để tìm ngưỡng tối ưu cho xác suất dự đoán\n\n#     #Fit mô hình và return best_score\n#     tuned_model.fit(X_train, y_train)\n#     best_threshold = tuned_model.best_threshold_\n#     best_model = CatBoostClassifier(**best_params)\n#     best_model.fit(X_train, y_train, cat_features=cat_features)\n#     # 7 Đánh giá\n#     print(\"Best parameters from Optuna:\", best_params)\n#     print(\"Best threshold from Optuna:\", best_threshold)\n#     # Đánh giá trên tập train\n#     # train_metrics = evaluate(best_model, X_train, y_train, name=f\"Train {train_name}\")\n#     # Đánh giá trên tập test\n#     # acc_test, best_precision, best_recall, best_f1, roc_auc = evaluate(best_model, X_test, y_test, name=f\"Test {test_name}\")\n#     #Chọn ngưỡng tối ưu cho xác suất dự đoán\n#     preds_proba = best_model.predict_proba(X_test)[:, 1]\n#     best_preds = (preds_proba >= best_threshold).astype(int)\n#     #Lưu lại kết quả dự đoán ra file csv\n#     submission = pd.DataFrame({\n#         \"ID\": index,\n#         \"label\": best_preds\n#     })    \n#     submission.to_csv(f'find_threshold_{train_name}_ratio_{delay_not_delay_ratio}.csv', index=False)\n#     #Lưu mô hình tốt nhất\n#     best_model.save_model(f'best_catboost_model_{train_name}_ratio_{delay_not_delay_ratio}.cbm')\n#     #Trả về các chỉ số đánh giá trên tập test và tham số tốt nhất","metadata":{},"outputs":[],"execution_count":null},{"id":"059dbed5","cell_type":"code","source":"# #Load model tốt nhất\n# model_path = r'D:\\Python_code\\DS_108_BTVN\\Kaggle\\best_model\\best_catboost_model_7-9_ratio_10.cbm'\n# best_model =  CatBoostClassifier()\n# best_model.load_model(model_path)\n# # Lấy ra các tham số của mô hình\n# best_model_params = best_model.get_params()\n# # In ra các tham số của mô hình\n# print(\"Các tham số của mô hình CatBoost tốt nhất:\")\n# best_model_params","metadata":{},"outputs":[],"execution_count":null},{"id":"b5f3f9eb","cell_type":"code","source":"# # Chạy lại thực nghiệm với các tham số tối ưu đã tìm được\n# CatBoostRun_param_specify_custome(data_7_9, \"7-9\", data_10_clean, index_10,10, best_model_params)","metadata":{},"outputs":[],"execution_count":null},{"id":"48b375a1","cell_type":"markdown","source":"<a id=\"num_statistics\"></a>\n## <b><span style='color:darkorange'>EXP 2|</span><span style='color:#8502d1'> TRAIN 4-9 TEST 10 </span></b>","metadata":{}},{"id":"af91278a","cell_type":"code","source":"# # Tiến hành huấn luyện mô hình CatBoost trên bộ dữ liệu 4-9 với tỉ lệ delay:not delay là 1:5 và test trên 10\n# CatBoostRun(data_4_9,\"4-9\", data_10_clean, index_10, delay_not_delay_ratio=3)","metadata":{},"outputs":[],"execution_count":null},{"id":"fc61d9e0-151f-4f6c-8d74-11f47b9f975f","cell_type":"code","source":"# # Tiến hành huấn luyện mô hình CatBoost trên bộ dữ liệu 4-9 với tỉ lệ delay:not delay là 1:10 và test trên 10\n# CatBoostRun(data_4_9,\"4-9\", data_10_clean, index_10, delay_not_delay_ratio=10)","metadata":{},"outputs":[],"execution_count":null},{"id":"7477b707","cell_type":"code","source":"# # Tiến hành huấn luyện mô hình CatBoost trên bộ dữ liệu 4-9 với tỉ lệ delay:not delay là 1:20 và test trên 10\n# CatBoostRun(data_4_9,\"4-9\", data_10_clean, index_10, delay_not_delay_ratio=20)","metadata":{},"outputs":[],"execution_count":null},{"id":"2e3fa123","cell_type":"markdown","source":"## Chạy lại mô hình với các tham số và ngưỡng chỉ định","metadata":{}},{"id":"60d76e0e","cell_type":"code","source":"# #Sử dụng mô hình CatBoost cho các thực nghiệm\n# def CatBoostRun_param_specify_custome(data_train, train_name, data_test,index, delay_not_delay_ratio, best_params):\n#     # 1. Tách train/val từ data_train\n#     X_train = data_train.drop(columns=['label'])\n#     y_train = data_train['label']\n#      # 8. Đánh giá trên tập test (data_test)\n#     X_test = data_test\n#     # X_train, X_val, y_train, y_val = train_test_split(\n#     #     X, y, test_size=0.2, random_state=42, stratify=y\n#     # )\n\n#     # # 2. \n#     n = delay_not_delay_ratio #Undersampling tập train với tỉ lệ 1:n (delay:not delay)\n#     delay_idx = y_train[y_train == 1].index\n#     not_delay_idx = y_train[y_train == 0].sample(n=len(delay_idx)*n, random_state=42).index\n#     undersample_idx = delay_idx.union(not_delay_idx)\n#     X_train = X_train.loc[undersample_idx]\n#     y_train = y_train.loc[undersample_idx]\n\n#     # 3. Xác định các cột categorical và chuyển đổi thành kiểu category\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n#     X_test[cat_features] = X_test[cat_features].astype('category')\n\n#     # study = optuna.create_study(direction='maximize')\n#     # study.optimize(lambda trial: objective_CB(trial, X_train, y_train), n_trials=30, timeout=600) #optuna được tối ưu dựa trên chỉ số F1-macro được trả về sau mỗi trial trong hàm objective_CB, optuna sẽ tự động học lại từ kinh nghiệm của các trial trước đó để tìm ra tham số tốt nhất cho mô hình CatBoostClassifier\n#     best_params.update({'loss_function': 'Logloss', 'random_seed': 42, 'verbose': 0})\n\n#     # 6. Train lại với tham số tốt nhất trên toàn bộ train_4_6_full đã undersample\n#     base_model = CatBoostClassifier(**best_params, task_type=\"GPU\", devices='0', cat_features=cat_features)\n\n#     #Tuning threshold với F1-macro\n#     scorer = make_scorer(f1_score, average='macro')\n#     tuned_model = TunedThresholdClassifierCV(estimator=base_model, scoring=scorer, cv=5) #Sử dụng TunedThresholdClassifierCV để tìm ngưỡng tối ưu cho xác suất dự đoán\n\n#     #Fit mô hình và return best_score\n#     tuned_model.fit(X_train, y_train)\n#     best_threshold = tuned_model.best_threshold_\n#     best_model = CatBoostClassifier(**best_params)\n#     best_model.fit(X_train, y_train, cat_features=cat_features)\n#     # 7 Đánh giá\n#     print(\"Best parameters from Optuna:\", best_params)\n#     print(\"Best threshold from Optuna:\", best_threshold)\n#     # Đánh giá trên tập train\n#     # train_metrics = evaluate(best_model, X_train, y_train, name=f\"Train {train_name}\")\n#     # Đánh giá trên tập test\n#     # acc_test, best_precision, best_recall, best_f1, roc_auc = evaluate(best_model, X_test, y_test, name=f\"Test {test_name}\")\n#     #Chọn ngưỡng tối ưu cho xác suất dự đoán\n#     preds_proba = best_model.predict_proba(X_test)[:, 1]\n#     best_preds = (preds_proba >= best_threshold).astype(int)\n#     #Lưu lại kết quả dự đoán ra file csv\n#     submission = pd.DataFrame({\n#         \"ID\": index,\n#         \"label\": best_preds\n#     })    \n#     submission.to_csv(f'find_threshold_{train_name}_ratio_{delay_not_delay_ratio}.csv', index=False)\n#     #Lưu mô hình tốt nhất\n#     best_model.save_model(f'best_catboost_model_{train_name}_ratio_{delay_not_delay_ratio}.cbm')\n#     #Trả về các chỉ số đánh giá trên tập test và tham số tốt nhất","metadata":{},"outputs":[],"execution_count":null},{"id":"13d11046","cell_type":"code","source":"# #Load model tốt nhất\n# model_path = r'D:\\Python_code\\DS_108_BTVN\\Kaggle\\best_model\\best_catboost_model_7-9_ratio_10.cbm'\n# best_model =  CatBoostClassifier()\n# best_model.load_model(model_path)\n# # Lấy ra các tham số của mô hình\n# best_model_params = best_model.get_params()\n# # In ra các tham số của mô hình\n# print(\"Các tham số của mô hình CatBoost tốt nhất:\")\n# best_model_params","metadata":{},"outputs":[],"execution_count":null},{"id":"4f7b4e9c","cell_type":"code","source":"# # Chạy lại thực nghiệm với các tham số tối ưu đã tìm được\n# CatBoostRun_param_specify_custome(data_7_9, \"7-9\", data_10_clean, index_10,10, best_model_params)","metadata":{},"outputs":[],"execution_count":null},{"id":"965eff9a","cell_type":"code","source":"# #Sử dụng mô hình CatBoost cho các thực nghiệm\n# def CatBoostRun_param_specify(data_train, train_name, data_test,index, delay_not_delay_ratio):\n#     # 1. Tách train/val từ data_train\n#     X_train = data_train.drop(columns=['label'])\n#     y_train = data_train['label']\n#      # 8. Đánh giá trên tập test (data_test)\n#     X_test = data_test\n#     # X_train, X_val, y_train, y_val = train_test_split(\n#     #     X, y, test_size=0.2, random_state=42, stratify=y\n#     # )\n\n#     # # 2. \n#     n = delay_not_delay_ratio #Undersampling tập train với tỉ lệ 1:n (delay:not delay)\n#     delay_idx = y_train[y_train == 1].index\n#     not_delay_idx = y_train[y_train == 0].sample(n=len(delay_idx)*n, random_state=42).index\n#     undersample_idx = delay_idx.union(not_delay_idx)\n#     X_train = X_train.loc[undersample_idx]\n#     y_train = y_train.loc[undersample_idx]\n\n\n\n#     # 3. Xác định các cột categorical và chuyển đổi thành kiểu category\n#     cat_features = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n#     X_train[cat_features] = X_train[cat_features].astype('category')\n#     X_test[cat_features] = X_test[cat_features].astype('category')\n\n#     # study = optuna.create_study(direction='maximize')\n#     # study.optimize(lambda trial: objective_CB(trial, X_train, y_train), n_trials=30, timeout=600) #optuna được tối ưu dựa trên chỉ số F1-macro được trả về sau mỗi trial trong hàm objective_CB, optuna sẽ tự động học lại từ kinh nghiệm của các trial trước đó để tìm ra tham số tốt nhất cho mô hình CatBoostClassifier\n#     best_params = {'iterations': 416, 'learning_rate': 0.0192814727903692, 'depth': 3, 'l2_leaf_reg': 10, 'loss_function': 'Logloss', 'random_seed': 42, 'verbose': 0}\n#     best_params.update({'loss_function': 'Logloss', 'random_seed': 42, 'verbose': 0})\n\n#     # 6. Train lại với tham số tốt nhất trên toàn bộ train_4_6_full đã undersample\n#     best_model = CatBoostClassifier(**best_params)\n#     best_model.fit(X_train, y_train, cat_features=cat_features)\n#     best_threshold = 0.5659902159380762\n#     # 7 Đánh giá\n#     print(\"Best parameters from Optuna:\", best_params)\n#     print(\"Best threshold from Optuna:\", best_threshold)\n#     # Đánh giá trên tập train\n#     # train_metrics = evaluate(best_model, X_train, y_train, name=f\"Train {train_name}\")\n#     # Đánh giá trên tập test\n#     # acc_test, best_precision, best_recall, best_f1, roc_auc = evaluate(best_model, X_test, y_test, name=f\"Test {test_name}\")\n#     #Chọn ngưỡng tối ưu cho xác suất dự đoán\n#     preds_proba = best_model.predict_proba(X_test)[:, 1]\n#     best_preds = (preds_proba >= best_threshold).astype(int)\n#     #Lưu lại kết quả dự đoán ra file csv\n#     submission = pd.DataFrame({\n#         \"ID\": index,\n#         \"label\": best_preds\n#     })\n#     submission.to_csv(f'catboost_param_specified_submission_{train_name}_ratio_{delay_not_delay_ratio}.csv', index=False)\n#     #Lưu mô hình tốt nhất\n#     best_model.save_model(f'best_catboost_param_specified_model_{train_name}_ratio_{delay_not_delay_ratio}.cbm')\n#     #Trả về các chỉ số đánh giá trên tập test và tham số tốt nhất","metadata":{},"outputs":[],"execution_count":null}]}