{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"### Intro\n\nHere is an attempt to implement an idea of using the parallel RNN branches for each asset and use learned information to compute the vector of targets at each timestep.\n\nSteps:\n* Fill gaps in the training set like it describer in the competition tutorial notebook;\n* Use `TimeseriesGenerator` like generator class to build the dataset;\n* Build the model. Simplified  structure:\n    - Lambda layer needed for assets separation;\n    - Masking layer. Generated records (filled gaps) has zeros as features values, so they are not used in the computations;\n    - LSTM or GRU layer;\n    - Concatanate layer.\n\n![изображение.png](attachment:99932047-786c-45c1-afb0-14c19c3b6183.png)\n","metadata":{},"attachments":{"99932047-786c-45c1-afb0-14c19c3b6183.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"### Load and prepare the data","metadata":{}},{"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nfrom sklearn.model_selection import KFold\nimport tensorflow as tf\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\nimport tensorflow_probability as tfp\nfrom sklearn.preprocessing import RobustScaler\nfrom tensorflow.keras.preprocessing.sequence import TimeseriesGenerator\nimport gresearch_crypto\nimport gc\n\nfrom plotly.subplots import make_subplots\nimport plotly.graph_objects as go\n\npd.set_option('display.max_columns', None)\n\nDEBUG = True","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2021-11-13T11:31:46.446001Z","iopub.execute_input":"2021-11-13T11:31:46.446694Z","iopub.status.idle":"2021-11-13T11:31:54.444167Z","shell.execute_reply.started":"2021-11-13T11:31:46.446508Z","shell.execute_reply":"2021-11-13T11:31:54.443174Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train = pd.read_csv('../input/g-research-crypto-forecasting/train.csv').set_index(\"timestamp\")\nassets = pd.read_csv('../input/g-research-crypto-forecasting/asset_details.csv')\n#for assets sorting \nassets_order = pd.read_csv('../input/g-research-crypto-forecasting/supplemental_train.csv').Asset_ID[:14]\nassets_order = dict((t,i) for i,t in enumerate(assets_order))\n\nif DEBUG:\n    #train = train[1000000:12000000]\n    train = train[10000000:]\n\ntrain[['Count', 'Open', 'High', 'Low', 'Close', 'Volume', 'VWAP','Target']] = \\\ntrain[['Count', 'Open', 'High', 'Low', 'Close', 'Volume', 'VWAP','Target']].astype(np.float32)\n\ntrain['Target'] = train['Target'].fillna(0)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:31:54.446476Z","iopub.execute_input":"2021-11-13T11:31:54.447176Z","iopub.status.idle":"2021-11-13T11:32:57.831713Z","shell.execute_reply.started":"2021-11-13T11:31:54.447117Z","shell.execute_reply":"2021-11-13T11:32:57.830755Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# VWAP column has -inf and inf values. VWAP_max and VWAP_min will be used for replacement\n    \nVWAP_max = np.max(train[np.isfinite(train.VWAP)].VWAP)\nVWAP_min = np.min(train[np.isfinite(train.VWAP)].VWAP)\nprint(VWAP_max, \"\\n\", VWAP_min)\n\ntrain['VWAP'] = np.nan_to_num(train.VWAP, posinf=VWAP_max, neginf=VWAP_min)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:32:57.833687Z","iopub.execute_input":"2021-11-13T11:32:57.834143Z","iopub.status.idle":"2021-11-13T11:32:59.409141Z","shell.execute_reply.started":"2021-11-13T11:32:57.834085Z","shell.execute_reply":"2021-11-13T11:32:59.407956Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Get the series of the 'real' record ids for futher matching\n    \ndf = train[['Asset_ID', 'Target']].copy()\n\ntimes = dict((t,i) for i,t in enumerate(df.index.unique()))\ndf['id'] = df.index.map(times)\ndf['id'] = df['id'].astype(str) + '_' + df['Asset_ID'].astype(str)\nids = df.id.copy()\n\ndel df","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:32:59.411940Z","iopub.execute_input":"2021-11-13T11:32:59.412636Z","iopub.status.idle":"2021-11-13T11:33:42.257652Z","shell.execute_reply.started":"2021-11-13T11:32:59.412592Z","shell.execute_reply":"2021-11-13T11:33:42.256512Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def add_features(df):\n    df['Upper_Shadow'] = df['High'] - np.maximum(df['Close'], df['Open'])\n    df['Lower_Shadow'] = np.minimum(df['Close'], df['Open']) - df['Low']\n    \n    df['spread'] = df['High'] - df['Low']\n    df['mean_trade'] = df['Volume']/df['Count']\n    df['log_price_change'] = np.log(df['Close']/df['Open'])\n    return df\n\ntrain=add_features(train)\ntrain.shape","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:33:42.259138Z","iopub.execute_input":"2021-11-13T11:33:42.259456Z","iopub.status.idle":"2021-11-13T11:33:42.818173Z","shell.execute_reply.started":"2021-11-13T11:33:42.259412Z","shell.execute_reply":"2021-11-13T11:33:42.817087Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"scale_features = train.columns.drop(['Asset_ID','Target'])\nRS = RobustScaler()\ntrain[scale_features] = RS.fit_transform(train[scale_features])","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:33:42.819772Z","iopub.execute_input":"2021-11-13T11:33:42.820580Z","iopub.status.idle":"2021-11-13T11:33:49.603342Z","shell.execute_reply.started":"2021-11-13T11:33:42.820534Z","shell.execute_reply":"2021-11-13T11:33:49.602209Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Filling gaps\n\nind = train.index.unique()\n\ndef reindex(df):\n    df = df.reindex(range(ind[0],ind[-1]+60,60),method='nearest')\n    df = df.fillna(method=\"ffill\").fillna(method=\"bfill\")\n    return df\n\ntrain=train.groupby('Asset_ID').apply(reindex).reset_index(0, drop=True).sort_index()\ntrain.shape","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:33:49.605104Z","iopub.execute_input":"2021-11-13T11:33:49.605413Z","iopub.status.idle":"2021-11-13T11:33:59.547209Z","shell.execute_reply.started":"2021-11-13T11:33:49.605375Z","shell.execute_reply":"2021-11-13T11:33:59.546098Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Matching records and marking generated rows as 'non-real'\n\ntrain['group_num'] = train.index.map(times)\ntrain = train.dropna(subset=['group_num'])\ntrain['group_num'] = train['group_num'].astype('int')\n\ntrain['id'] = train['group_num'].astype(str) + '_' + train['Asset_ID'].astype(str)\n\ntrain['is_real'] = train.id.isin(ids)*1\ntrain = train.drop('id', axis=1)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:33:59.548927Z","iopub.execute_input":"2021-11-13T11:33:59.549983Z","iopub.status.idle":"2021-11-13T11:35:00.199196Z","shell.execute_reply.started":"2021-11-13T11:33:59.549935Z","shell.execute_reply":"2021-11-13T11:35:00.197925Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Features values for 'non-real' rows are set to zeros\n\nfeatures = train.columns.drop(['Asset_ID','group_num','is_real'])\ntrain.loc[train.is_real==0, features]=0.","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:00.200633Z","iopub.execute_input":"2021-11-13T11:35:00.203065Z","iopub.status.idle":"2021-11-13T11:35:00.899220Z","shell.execute_reply.started":"2021-11-13T11:35:00.203021Z","shell.execute_reply":"2021-11-13T11:35:00.898203Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Sorting assets according to their order in the 'supplemental_train.csv'\n\ntrain['asset_order'] = train.Asset_ID.map(assets_order) \ntrain=train.sort_values(by=['group_num', 'asset_order'])\ntrain.head(20)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:00.905258Z","iopub.execute_input":"2021-11-13T11:35:00.906840Z","iopub.status.idle":"2021-11-13T11:35:07.475147Z","shell.execute_reply.started":"2021-11-13T11:35:00.906764Z","shell.execute_reply":"2021-11-13T11:35:07.474008Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"targets = train['Target'].to_numpy().reshape(-1, 14)\n#targets = np.expand_dims(targets, axis=1)\n\nfeatures = train.columns.drop(['Asset_ID', 'Target', 'group_num','is_real'])\ntrain = train[features]\n\ntrain=np.array(train)\ntrain = train.reshape(-1,14,train.shape[-1])\n#train = np.expand_dims(train, axis=1)\ntrain.shape","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:07.514434Z","iopub.execute_input":"2021-11-13T11:35:07.515173Z","iopub.status.idle":"2021-11-13T11:35:09.928521Z","shell.execute_reply.started":"2021-11-13T11:35:07.514997Z","shell.execute_reply":"2021-11-13T11:35:09.927456Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Dataset creation\n\nSamples with a duration of 15 records (minutes) will be formed from the `train` array. Each sample has a target vector corresponding to the 15th record. \n\n![data.JPG](attachment:364826a2-fe7e-4ac8-89cb-5d1217ae2d02.JPG)","metadata":{},"attachments":{"364826a2-fe7e-4ac8-89cb-5d1217ae2d02.JPG":{"image/jpeg":"/9j/4AAQSkZJRgABAQEAYABgAAD/4RDiRXhpZgAATU0AKgAAAAgABAE7AAIAAAALAAAISodpAAQAAAABAAAIVpydAAEAAAAMAAAQzuocAAcAAAgMAAAAPgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAANCh0LvQsNCy0LAAAAAFkAMAAgAAABQAABCkkAQAAgAAABQAABC4kpEAAgAAAAM4MAAAkpIAAgAAAAM4MAAA6hwABwAACAwAAAiYAAAAABzqAAAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMjAyMToxMTowNyAyMToyMDoxNgAyMDIxOjExOjA3IDIxOjIwOjE2AAAAIQQ7BDAEMgQwBAAA/+ELHWh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczpkYz0iaHR0cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iLz48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOnhtcD0iaHR0cDovL25zLmFkb2JlLmNvbS94YXAvMS4wLyI+PHhtcDpDcmVhdGVEYXRlPjIwMjEtMTEtMDdUMjE6MjA6MTYuNzk4PC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOmRjPSJodHRwOi8vcHVybC5vcmcvZGMvZWxlbWVudHMvMS4xLyI+PGRjOmNyZWF0b3I+PHJkZjpTZXEgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj48cmRmOmxpPtCh0LvQsNCy0LA8L3JkZjpsaT48L3JkZjpTZXE+DQoJCQk8L2RjOmNyZWF0b3I+PC9yZGY6RGVzY3JpcHRpb24+PC9yZGY6UkRGPjwveDp4bXBtZXRhPg0KICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICA8P3hwYWNrZXQgZW5kPSd3Jz8+/9sAQwAHBQUGBQQHBgUGCAcHCAoRCwoJCQoVDxAMERgVGhkYFRgXGx4nIRsdJR0XGCIuIiUoKSssKxogLzMvKjInKisq/9sAQwEHCAgKCQoUCwsUKhwYHCoqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioq/8AAEQgBvgRSAwEiAAIRAQMRAf/EAB8AAAEFAQEBAQEBAAAAAAAAAAABAgMEBQYHCAkKC//EALUQAAIBAwMCBAMFBQQEAAABfQECAwAEEQUSITFBBhNRYQcicRQygZGhCCNCscEVUtHwJDNicoIJChYXGBkaJSYnKCkqNDU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6g4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2drh4uPk5ebn6Onq8fLz9PX29/j5+v/EAB8BAAMBAQEBAQEBAQEAAAAAAAABAgMEBQYHCAkKC//EALURAAIBAgQEAwQHBQQEAAECdwABAgMRBAUhMQYSQVEHYXETIjKBCBRCkaGxwQkjM1LwFWJy0QoWJDThJfEXGBkaJicoKSo1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoKDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uLj5OXm5+jp6vLz9PX29/j5+v/aAAwDAQACEQMRAD8A+kaKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAIrq4FpZzXLqzrDG0hVMZIAzgZwM037RJ5zJ9km2gZD5TB46fez7dKL7zBp9x5LBZPKbYzEAA4ODzxVbXb+80zRLm903Tm1O4hTetokojaXHUAnjOO3ek3ZXY0r6E/2qbyt/2C43bsbN0ecev38Y/GnfaJPOVPsk20jJfKYHHT72fbpXHaX408T6xpdtqOn+CvNtrmMSRv/a8IyD7Y4PtXaWzyy2sUlxD5ErIC8W4NsbHK5HXHrVWaJuRC6mMRb7BcAggBN0eT7/fx+tO+0S+ZGv2OfDAEtuTCex+bP5ZqeikMrC6mKOfsNwCuMKWjy30+b+eKX7TLmP8A0Kf5+vzR/Jz3+b8eM1YooAr/AGmXMn+hT/J0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/wB/H61ZooAg+0SecyfZJtoGQ+UweOn3s+3Sm/apvK3/AGC43bsbN0ecev38Y/GrNFAEH2iTzlT7JNtIyXymBx0+9n26U0XUxiLfYLgEEAJujyff7+P1qzRQBB9ol8yNfsc+GAJbcmE9j82fyzTRdTFHP2G4BXGFLR5b6fN/PFWaKAK/2mXMf+hT/P1+aP5Oe/zfjxmj7TLmT/Qp/k6fNH8/Pb5vx5xViigCsbqYIh+w3BLZyoaPK/X5v5Zp32iXzJF+xz4UEhtyYf2HzZ/PFT0UAVjdTCIN9guCSSCm6PI9/v4/WnfaJPOZPsk20DIfKYPHT72fbpU9FAFb7VN5W/7Bcbt2Nm6POPX7+MfjTvtEnnKn2SbaRkvlMDjp97Pt0qeigCsLqYxFvsFwCCAE3R5Pv9/H6077RL5ka/Y58MAS25MJ7H5s/lmp6KAKwupijn7DcArjClo8t9Pm/nil+0y5j/0Kf5+vzR/Jz3+b8eM1YooAr/aZcyf6FP8AJ0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/38frVmigCD7RJ5zJ9km2gZD5TB46fez7dKb9qm8rf9guN27GzdHnHr9/GPxqzRQBB9ok85U+yTbSMl8pgcdPvZ9ulNF1MYi32C4BBACbo8n3+/j9as0UAQfaJfMjX7HPhgCW3JhPY/Nn8s00XUxRz9huAVxhS0eW+nzfzxVmigCv9plzH/oU/z9fmj+Tnv8348Zo+0y5k/wBCn+Tp80fz89vm/HnFWKKAKxupgiH7DcEtnKho8r9fm/lmnfaJfMkX7HPhQSG3Jh/YfNn88VPRQBWN1MIg32C4JJIKbo8j3+/j9ad9ok85k+yTbQMh8pg8dPvZ9ulT0UAVvtU3lb/sFxu3Y2bo849fv4x+NO+0SecqfZJtpGS+UwOOn3s+3Sp6KAKwupjEW+wXAIIATdHk+/38frTvtEvmRr9jnwwBLbkwnsfmz+WanooArC6mKOfsNwCuMKWjy30+b+eKX7TLmP8A0Kf5+vzR/Jz3+b8eM1YooAr/AGmXMn+hT/J0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/wB/H61ZooAg+0SecyfZJtoGQ+UweOn3s+3Sm/apvK3/AGC43bsbN0ecev38Y/GrNFAEH2iTzlT7JNtIyXymBx0+9n26U0XUxiLfYLgEEAJujyff7+P1qzRQBB9ol8yNfsc+GAJbcmE9j82fyzTRdTFHP2G4BXGFLR5b6fN/PFWaKAK/2mXMf+hT/P1+aP5Oe/zfjxmj7TLmT/Qp/k6fNH8/Pb5vx5xViigCsbqYIh+w3BLZyoaPK/X5v5Zp32iXzJF+xz4UEhtyYf2HzZ/PFT0UAVjdTCIN9guCSSCm6PI9/v4/WnfaJPOZPsk20DIfKYPHT72fbpU9FAFb7VN5W/7Bcbt2Nm6POPX7+MfjTvtEnnKn2SbaRkvlMDjp97Pt0qeigCsLqYxFvsFwCCAE3R5Pv9/H6077RL5ka/Y58MAS25MJ7H5s/lmp6KAKwupijn7DcArjClo8t9Pm/nil+0y5j/0Kf5+vzR/Jz3+b8eM1YooAr/aZcyf6FP8AJ0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/38frVmigCD7RJ5zJ9km2gZD5TB46fez7dKb9qm8rf9guN27GzdHnHr9/GPxqzRQBB9ok85U+yTbSMl8pgcdPvZ9ulNF1MYi32C4BBACbo8n3+/j9as0UAQfaJfMjX7HPhgCW3JhPY/Nn8s00XUxRz9huAVxhS0eW+nzfzxVmigCv9plzH/oU/z9fmj+Tnv8348Zo+0y5k/wBCn+Tp80fz89vm/HnFWKKAKxupgiH7DcEtnKho8r9fm/lmnfaJfMkX7HPhQSG3Jh/YfNn88VPRQBWN1MIg32C4JJIKbo8j3+/j9ad9ok85k+yTbQMh8pg8dPvZ9ulT0UAVvtU3lb/sFxu3Y2bo849fv4x+NO+0SecqfZJtpGS+UwOOn3s+3Sp6KAKwupjEW+wXAIIATdHk+/38frTvtEvmRr9jnwwBLbkwnsfmz+WanooArC6mKOfsNwCuMKWjy30+b+eKX7TLmP8A0Kf5+vzR/Jz3+b8eM1YooAr/AGmXMn+hT/J0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/wB/H61ZooAg+0SecyfZJtoGQ+UweOn3s+3Sm/apvK3/AGC43bsbN0ecev38Y/GrNFAEH2iTzlT7JNtIyXymBx0+9n26U0XUxiLfYLgEEAJujyff7+P1qzRQBB9ol8yNfsc+GAJbcmE9j82fyzTRdTFHP2G4BXGFLR5b6fN/PFWaKAK/2mXMf+hT/P1+aP5Oe/zfjxmj7TLmT/Qp/k6fNH8/Pb5vx5xViigCsbqYIh+w3BLZyoaPK/X5v5Zp32iXzJF+xz4UEhtyYf2HzZ/PFT0UAVjdTCIN9guCSSCm6PI9/v4/WnfaJPOZPsk20DIfKYPHT72fbpU9FAFb7VN5W/7Bcbt2Nm6POPX7+MfjTvtEnnKn2SbaRkvlMDjp97Pt0qeigCsLqYxFvsFwCCAE3R5Pv9/H6077RL5ka/Y58MAS25MJ7H5s/lmp6KAKwupijn7DcArjClo8t9Pm/nil+0y5j/0Kf5+vzR/Jz3+b8eM1YooAr/aZcyf6FP8AJ0+aP5+e3zfjzikN1MEQ/Ybgls5UNHlfr838s1ZooAg+0S+ZIv2OfCgkNuTD+w+bP54ppuphEG+wXBJJBTdHke/38frVmigCD7RJ5zJ9km2gZD5TB46fez7dKb9qm8rf9guN27GzdHnHr9/GPxqzRQBB9ok85U+yTbSMl8pgcdPvZ9ulNF1MYi32C4BBACbo8n3+/j9as0UAQfaJfMjX7HPhgCW3JhPY/Nn8s00XUxRz9huAVxhS0eW+nzfzxVmigCv9plzH/oU/z9fmj+Tnv8348Zo+0y5k/wBCn+Tp80fz89vm/HnFWKKAKxupgiH7DcEtnKho8r9fm/lmnfaJfMkX7HPhQSG3Jh/YfNn88VPRQBVe8lSJX/s+5JZwpRTGSoyBuPzdOe2TweKtVR1hY30xxNIY08yP5gu7neuOPrxV6gAooooAKKKKACiiigCtqWz+ybvzt3l+Q+/b1xtOce9ZvjC/udM8MXN3ZXb2c0ZXEyWD3u35gDmJPmYeuOnXtWpfeYdPuPJUNJ5TbFYAgnBwOeKnpMaPnfw18VPEkmp6PaxX+kasG1K8WW0iufsc8qDzCoeORcRrwCuTnOAecmvoO1me4s4ZpoWgkkjVmiZgxQkZKkjg46ZHFcbqHhn4dPeRaZq8Wky3FvNLfC0vJ1dg0zHe7Ix5BaTjIwCRjHFdnb28NpaxW9rGsUMKBI40GAqgYAHtiqumiWveJKKKKQwooooAKKKKACqWq6omlWqTSQTTl5BGkcO3cxOf7xA7HvV2sPxT/wAetj/1+J/6C1AEf/CVH/oB6l/31b//AB2j/hKj/wBAPUv++rf/AOO1UooAt/8ACVH/AKAepf8AfVv/APHaP+EqP/QD1L/vq3/+O1UooAt/8JUf+gHqX/fVv/8AHaP+EqP/AEA9S/76t/8A47XnfjK1t7nxDB9pgjm22vHmIGx859axP7L0/wD58bb/AL8r/hXfRwTqwU7nJUxKhLlsev8A/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47XkH9l6f/wA+Nt/35X/Cj+y9P/58bb/vyv8AhWv9nS/mI+uLsev/APCVH/oB6l/31b//AB2j/hKj/wBAPUv++rf/AOO14rqkOmaXpVzey2FuVgjL4EK5JA4HTueKxdCkitrK/i8S29qbmwQTyyi2QZiZd2QFHYhl/wCA1DwNnZyKWKurpH0J/wAJUf8AoB6l/wB9W/8A8do/4So/9APUv++rf/47XhCX+kiznurzRHs7eGHz/NmtUKunqChYZ9jg+1OS70pbW4uL7RHsI4IxKWntkIZT02lCwJ/2c59qf1H+9+AfWv7p7r/wlR/6Aepf99W//wAdo/4So/8AQD1L/vq3/wDjteCy32npFcIdBMF2tu88MM8EWZlXqQQxHGRkEg+1T6OtvfeGbXUJ9Gi8+S3WQwrFEC5xn5ecDPbJHviksC3e0vw/rsH1q3Q9z/4So/8AQD1L/vq3/wDjtH/CVH/oB6l/31b/APx2vnKwvpL3UbKI6C0cdxdTMS0VtjyUyuOGJ4YqSevpmtKKIt4vW0uNPsobRrWSRIhErMxV1AYnHGQeAPx9lHBOVrS38vK43iUt0e9/8JUf+gHqX/fVv/8AHaP+EqP/AEA9S/76t/8A47XkH9l6f/z423/flf8ACj+y9P8A+fG2/wC/K/4Vp/Z0v5iPri7Hr/8AwlR/6Aepf99W/wD8do/4So/9APUv++rf/wCO15B/Zen/APPjbf8Aflf8KP7L0/8A58bb/vyv+FH9nS/mD64ux6//AMJUf+gHqX/fVv8A/HaP+EqP/QD1L/vq3/8AjtY+jf8AICsP+vaP/wBBFeXaZplg2kWZNjbEmBCSYV5+Ue1ctDDus2k7WN6tZU0m0e0/8JUf+gHqX/fVv/8AHaP+EqP/AEA9S/76t/8A47XkH9l6f/z423/flf8ACj+y9P8A+fG2/wC/K/4V1f2dL+Yw+uLsev8A/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47XkH9l6f/wA+Nt/35X/CuZ8UMbW5jh0ayt99tEb26UW6NviUgbORwW+bkc/LUSwLgruRUcUpOyR9C/8ACVH/AKAepf8AfVv/APHaP+EqP/QD1L/vq3/+O14Xc3ekQtax22krfTXURniit4Y8mMYy2WKj+Id8802a90qN4IoNCkubiaEz/Z0tY0kRAcEsHK9+MDJ9qbwNvtfgH1q/2T3b/hKj/wBAPUv++rf/AOO0f8JUf+gHqX/fVv8A/Ha8LmutLR4orbRHuriSETm3jtUR0Tplt5UDnjGc9eKpNqNlPrWkQ6do8U9pfJKXk8iMFSuARhiCNpzkY+meaTwNvtfgH1n+6fQP/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47Xg3iIJp8dtFp+iLNLcTogkiigwBuyy4dhyVDY4x7iqmn21zrkDvFZR6bAbuQGSSCBpFjX5QqABlzuByW/DPBpfUneyf4en+Y/rKtex9Cf8ACVH/AKAepf8AfVv/APHaP+EqP/QD1L/vq3/+O14R4eht7yO9juLa2uVtbpoI7n7OimYADJIAxkEkHGBx0FbH9l6f/wA+Nt/35X/CrWXtpPmJeLSdrHr/APwlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O15B/Zen/8APjbf9+V/wo/svT/+fG2/78r/AIU/7Ol/ML64ux6//wAJUf8AoB6l/wB9W/8A8do/4So/9APUv++rf/47XmnhaztrbxZA1vbxRMbaUExoFzynpXoFcNak6U+Rs6qdT2keYt/8JUf+gHqX/fVv/wDHaP8AhKj/ANAPUv8Avq3/APjtVKKxNC3/AMJUf+gHqX/fVv8A/HaP+EqP/QD1L/vq3/8AjtVKKALf/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47XjdrpenmzhzY23+rX/livp9Kl/svT/+fG2/78r/AIV6ay+TV+Y4fri7Hr//AAlR/wCgHqX/AH1b/wDx2j/hKj/0A9S/76t//jteQf2Xp/8Az423/flf8KqapDpml6Vc3sthblYIy+BCuSQOB07nih5e0ruQ1i03ZI9q/wCEqP8A0A9S/wC+rf8A+O0f8JUf+gHqX/fVv/8AHa+fvD2yO3vbbxFbWrXdkqzPILZFBjZdwIAHYhl/4DVqC/0iW2kuZ9Ee1tVgNwtxLaoUeMc5BUtg45w2D7VP1HTWQ/rWtlE93/4So/8AQD1L/vq3/wDjtH/CVH/oB6l/31b/APx2vCobzSWt5bi70VrKCOHz/NntoyjJ6goWGfY4PtUE2o6ekMo/sBoZzA81sk9vEPPCjJxhuOoOGKnHalLBcqu5fgCxV/snvn/CVH/oB6l/31b/APx2j/hKj/0A9S/76t//AI7Xhuix29/4btb+40eETywLIYhFGC5xn5ecDPbJHXnFYdrfy3WoQRr4fdY7i/dQTFbYWJF2svDE8MMk/XBPApywLi7OX4AsUmr2/E+jv+EqP/QD1L/vq3/+O0f8JUf+gHqX/fVv/wDHa8EiiLeL1tLjT7KG0a1kkSIRKzMVdQGJxxkHgD8fapf3MbeKNNtNPsbMWBumguZTApMj+WzbF46DHJ9cD1o+ouy97fy87D+s76H0N/wlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O15B/Zen/8APjbf9+V/wo/svT/+fG2/78r/AIVf9nS/mI+uLsev/wDCVH/oB6l/31b/APx2kbxZsUs2iakAoyTut/8A47XkP9l6f/z423/flf8ACtHw3Z21t4kc29vFETptwCY0C5+eL0rKrgnTg532Lp4lTko2ParedLm2inizslQOuR2IyKkqno//ACArD/r2j/8AQRVyuA6wooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigCnqm/+z28uATtvT92VLZ+cc4Hp1/CrlUtWAOmtum8kb4/nwTj51449en41doAKKKKACiiigAooooArakEOk3YlYqhgfcVGSBtOeKmmmit4zJPIkSDqzsAB+JqO/3HTrny4xK/lNtjIyGODxjvmo9T0uw1mwex1eyt760kILwXEYkRsHIyp4PIzQB4VqGnXth4muoG0Bda+1R6lG96lxb+Vcm5kRomkZnBUIq7TkcbRtzXt/h+zl0/w1plnc3H2qa3tIonnBz5jKgBbPvjNeTf2BZ6hZ6nrei+AfBKaRp09xEba8swLicQsVdt4XbHkq2AQ3GMkV63oVzbXvh7T7qwt/s1rNbRyQwbNvloVBC47YHGKpfBb0/W36il8evn+ly/RRRUjCiiigAooooAKw/FP/HrY/8AX4n/AKC1blYfin/j1sf+vxP/AEFqAOQt/FMdx47vPDn2WdWtrWOfzzBJtYsWyN23aAAowc8kkDkGugrnI9Pvbb4j3Ooi0eWyvdPig89HTELxvI2GBYNyHGCoPviujp9ED3f9dAooopAcX4r/AORii/69R/6Gayq1fFf/ACMUX/XqP/QzWVX0OD/gRPIxH8VhRRRXUc5k69pMmtQ2toZAlp56yXQDsjuq8hVI6fNt5yOBVNvDT2msR32kzt88DwXC31xLcBgeUPzMc4btkcMa6CR1ijaRzhVBJPsKqaTqa6vp8d7FbzQRSgNF523LqRkMArHAPvg+1ZuEW/P+kWpSS8jDsfDt9bXEskMOn6fEbd4vskDvLBMxOQzRkKFA54HJz14qD/hELi6tby3dbXTLee38pbW0leaEvuDByrKoGCMbQOcnJrsKrXN9BaTW0U7FXupPKiABO5tpb8OAaTpQtZ/1f/hxqpK+hiaf4eaFpd+k6Lp+6BovMsYsyOSMZ3bV2j/Z+b61f03T7qDwvFp135azRW4gDQTOAcLgNuAVlJ68dPU1q0VSgkmu4nJs5jSPCCafe2s8s9w/2a1SNcX05zJnLnaWxtOF+Xoe4rWfT5W8Sw6gGTyo7V4SuTu3MykdunymtGimoqOwnJvcKKKKokKKKKAO80b/AJAVh/17R/8AoIrzjSv+QPZ/9cE/9BFej6N/yArD/r2j/wDQRXnGlf8AIHs/+uCf+givJy/45HoYv4YluiiivWPPCufTwwl1ql/fapNOZLhwkYtryWILCowqnYVycljznrXQVQ1DVUsLi1t/s81xPdOVjjh254GSSWIAAHvmokov4iotrYwk8N3sGlWti9rpupRWjuIPtMjo8aZ+QrIFYggcYA7DmnyeHb7+zbO3nhsdVkgVgJLqWSOSIknGyUBmwBgdicZzXU0VPsole0Zx7+EJ1mtbmeHT9bmS1FvMupDjgkhlYq57kYI5GOa0F0O4t7nSJ7SGwjNmZVlghUwRhZMZKABuRjocZ56Vr2t9BevcLbsWNtKYZMgjDAA4/UVZpqnHdev6g5y6mNrOhDVr+zmeSVUtg7YjupYiWIwpGwj1PPXt3rN/sDV9P0G0sNGnQs2Ptzz3suW45EZKvsyc8gD8+R1dFDpxd/MSm0UNJhuLaxW3ns7WzSIBYo7adpV2/Uov9av0UVoQFFFFAF/w5/yNVv8A9e8v80ruq4Xw5/yNVv8A9e8v80rtbmH7TayweZJF5iFPMiba6ZGMg9iPWvAx38ZnrYX+GiWiseHw/wCV4VOh/wBr6o+Ymi/tB7nN2M5+bzMfeGeDinzaF53hcaL/AGpqUf7hYft6XGLrjHz+Zj7xxycVxvrY6V5mrRUNpb/ZLKG382WfyY1TzZm3O+BjLHuT3NTUMFseYWv/AB5w/wDXNf5VNUNr/wAecP8A1zX+VTV9VHZHgvcKyde0mTWobW0MgS089ZLoB2R3VeQqkdPm285HArWoocVJWYJtO6OfHht7LXIr/S53w0DwXC3lzLPuHVCNzHo3bI4Y1WsfD19bX3nRQ6fp0axOjQ2skjw3DN0ZoiFCgHnjJ5xmtrUNVSwuLW3+zzXE905WOOHbngZJJYgAAe+av1mqcG9OhfPLqcevhG4uI7qB0tdLtZ7V4Gt7GV5I5GY5D7GVVXHPABznrVnTvDbQlhJpOiWR8hovPs4cySEjG7O1dg9R83XrXT1Ws76C/WY2zFhDK0LkgjDKcEUeyht/X9ahzyt/X9dCrpNjcW3hyDT70RrJDAIC0ErYYAYBDYUqT146etZOleDlsrqymuLi4Y28BDbb+cgzMwLEAtjacdO/cV1NFW4JyUnuSpNKyM6TT5X8SQ6gGTyY7SSErk7tzMpH4fKaybnwZai802TT2uIorW5M0qNfzkYKt90biAdxGenGfXFdPRQ4Rdrrb/O4KTWwUUUVRIVc0H/kYm/7B1x/6HFVOrmg/wDIxN/2Drj/ANDirlxf8CRvh/4qPXNH/wCQFYf9e0f/AKCKuVT0f/kBWH/XtH/6CKuV88ewFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAUtWIGmtuh84b4/kyRn51549Ov4Vdqnqm/wDs9vLnEDb0/eFiuPnHGR69Pxq5QAUUUUAFFFFABRRRQBW1IKdLuw7+WphfL4ztG084qzVfUP8AkGXWI/NPkv8Au+fn4PHHrVigDwpfATeLbzXLy10zTbSC41G4juLa81i/EkjK5UtIiOFG7GQMY2kV6v4U8K2XhfT/AC7PzvMmRDP5l7NcJuA/g81mIXk15nH4T8vUtVbxB8KbjxFcS6hPKmpzXNqzzxs5KZ3yggBcKB6AdOlew6bGsWlWscdp9iVIUUW2QfJAA+TjI46cccU46Q08glrN/Ms0UUUgCiiigAooooAKw/FP/HrY/wDX4n/oLVuVi+J4ppLG2aCCSYxXSuyxLubGGGcfiKAM6ioPOn/6B1//AOAzUedP/wBA6/8A/AZqAJ6Kg86f/oHX/wD4DNR50/8A0Dr/AP8AAZqAOS8V/wDIxRf9eo/9DNZVdLrmh3uq38d1bx3luVi8sq+nvJnnOeGXHWs7/hE9V/56z/8Agpk/+OV6+HxVKnSUZbnn1qE5zckZdFan/CJ6r/z1n/8ABTJ/8co/4RPVf+es/wD4KZP/AI5W/wBeo9zH6rUMDU4YrjSbqKeNJY2iYMjqCDx3Brz23toYtO8Ppcvp1npkmmByb618y3ac7SSw3ou8joTnvXsP/CJ6r/z1n/8ABTJ/8co/4RPVf+es/wD4KZP/AI5WU8TQk73/AA9f8zSNGrFWseXF7G3s9NtNTltb63k85o7rUVeK3A3cRrG5IY/3ST0HBNUNPis5dL0SfUUt3gtdYniMk0e1Ikw+1cNnYuduATx8vtXsH/CJ6r/z1n/8FMn/AMco/wCET1X/AJ6z/wDgpk/+OUvrFG97/h5p/oV7Kra1vx9f8zye6vraDw/e6dLKq3p1rf8AZz98KbkMGx12kd+lSXFgTpHiG+sYf9PGouhnRC0ixbk3hcEHG3JwCPzr1T/hE9V/56z/APgpk/8AjlH/AAieq/8APWf/AMFMn/xyl7ejbV9Lben+Q3Sq32633PL7Czs3j1A6frdotvJZMsq6LYuioeMSHa7jeB2GGIPtV3wnLaR6pJbWEWnzI1uGe80tisRIOMPHkhXIOc5JODntXof/AAieq/8APWf/AMFMn/xyj/hE9V/56z/+CmT/AOOVUcTQi009vL1/zJdGq01Yy6K1P+ET1X/nrP8A+CmT/wCOUf8ACJ6r/wA9Z/8AwUyf/HK1+vUe5l9VqGXRWp/wieq/89Z//BTJ/wDHKP8AhE9V/wCes/8A4KZP/jlH16j3D6rUOo0b/kBWH/XtH/6CK840r/kD2f8A1wT/ANBFekWizWljBbiw1BhDGsYY2rc4GM1yVv4N1O2tooI5rgrGgRS2kyZwBj/npXBhK0KUpOXU7MRSlUSUTPorU/4RPVf+es//AIKZP/jlH/CJ6r/z1n/8FMn/AMcrv+vUe5x/Vahl1zOv2NrceLtCkltYJZf32GkjDH5UyvJ9DyPQ13X/AAieq/8APWf/AMFMn/xyj/hE9V/56z/+CmT/AOOVM8XQkrX/AAKjh6qPJNDgEs1i9zqtlBrKXANxGli5vWO75kdvMJ2H1K7cYIxxUV7eWT3Fld2sVnaX39qRrKAWku0UyYIkc8qCOxyMEAV7B/wieq/89Z//AAUyf/HKP+ET1X/nrP8A+CmT/wCOVksRRVtdvL0/yNXSqNt2/E8muILKyTWPLhghlh1uCeYIgDLAGjbecc7fvHPTr703XLq01W/1trWVZ4GGnJ5kbZVv3x6MOo56ivW/+ET1X/nrP/4KZP8A45R/wieq/wDPWf8A8FMn/wAcprEUVFRvp6eVg9lUve34+dzy3V7C0s/EL2ty+ladpi2ym0S+tN8G8sxfZh0CvyPU46Ut/a2kdlpbanqtndCO1ZVTWLd44pueGUsfkkxxk7mxzivUf+ET1X/nrP8A+CmT/wCOUf8ACJ6r/wA9Z/8AwUyf/HKX1ihrr+DF7Kr2Ob0Gc3Gg2cptpLXdEP3Ujs7KOn3m5P1PNaFan/CJ6r/z1n/8FMn/AMco/wCET1X/AJ6z/wDgpk/+OVv9eo9zL6rU7GXRWp/wieq/89Z//BTJ/wDHKP8AhE9V/wCes/8A4KZP/jlH16j3F9VqEXhz/karf/r3l/mld1XL6P4fvtN1Vb24W7n2RsgRNOeP7xHOSx9K6Hzp/wDoHX//AIDNXlYmpGpUco7HoUIOELMnoqDzp/8AoHX/AP4DNR50/wD0Dr//AMBmrmNieioPOn/6B1//AOAzUedP/wBA6/8A/AZqAPN7X/jzh/65r/Kpq0o/CGqRxqiyz4UADOkyf/HKd/wieq/89Z//AAUyf/HK91Y2ilueU8NUuZdFan/CJ6r/AM9Z/wDwUyf/AByj/hE9V/56z/8Agpk/+OU/r1HuL6rUOC8R2MFx4o0WQ2sEtxtn2NJGGOVTK8n0PI9DWJ4dgSS40yR9Uso9UVwbqGKyf7W5wdyzN5h492XHTGOK9Y/4RPVf+es//gpk/wDjlH/CJ6r/AM9Z/wDwUyf/ABysfrFDm5r/AIGvsavLax4/e3dk9xZXVrFZ2l9/akaygFpLtFMmCJHPKgjscjBAFTxxWGmwXHlx29u9rr4luQiqpii3HYzY6Lg8E8V6z/wieq/89Z//AAUyf/HKP+ET1X/nrP8A+CmT/wCOVMa9FO9/w9P8vxG6VRq1vx9f8zybVLi11W+1UwOJ7eTUNOTep+VxuGcHuOeoqxrNlbWviFrW7k0vT9MS1T7Il9ab4A25i+zDoFfkepx0r1H/AIRPVf8AnrP/AOCmT/45R/wieq/89Z//AAUyf/HKf1ihazfW+wexq9jyq+tbSG3sX1DVbO9ZLPakerwSRLLydrRsxyknQH7zYANdppE5udGs5mt5LYvCp8mVizJx0JPJPuea6H/hE9V/56z/APgpk/8AjlH/AAieq/8APWf/AMFMn/xytIYujFvXfyIlh6kktDLorU/4RPVf+es//gpk/wDjlH/CJ6r/AM9Z/wDwUyf/AByr+vUe5H1WoZdXNB/5GJv+wdcf+hxVY/4RPVf+es//AIKZP/jlWtM8PX+n30l3Ml3cN9mkgVE054/vFTnJY/3P1rDEYqlUpOMXqa0cPOE1JnpWj/8AICsP+vaP/wBBFXKq6ZE8GkWcUq7XjgRWU9iFAIq1XkHohRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAFHWGjTTHM0ZkTzI/lDbed645+vNXqq6kZVsWMEqxPvTDswUAbxkZPqOPxq1QAUUUUAFFFFABRRRQBX1D/kGXWJPKPkv+85+Tg88elWKrakVGl3ZdPMUQvlM43DaeM1ZoA+fvEPhq6Nzd6j4ig8RyTG51C3uJLSW5YbmO6zeJIzgoFAXgYBOG56e3eG0uI/CulJfW4trlbOISwA5EbbBlfwPFeQ6ld+JdSvr+21SbxFBFp019d3Rs0mjy+7y7KKEoBvUghyBkZ+9XsOgG/bw3pp1kY1A2sZugO0u0bv1zTj8H3fr/AF6WCfx/f+n9L5mhRUN1dw2cQkuCwUttG1Gbn6AH0qMalatPHCGffKoZR5T4IIyOcYH40gLVFUV1iyeF5VeTZGQG/cPnnpxjJ6U7+1LTzIU3SbpgCn7l+cnAzxx+NAFyiqQ1azKytvkxD9/9y/HOOOOefSg6tZhYm3yYm+5+5fnnHPHHPrQBdoqn/alp5kybpN0IJf8Acvxg4OOOfwpraxZJCkrPJskJC/uHzx14xkdaAL1FVTqVqs8kJZ98SlmHlPgADJ5xg/hUf9sWX2cTb5NhbZnyHznGemM96AL1FVf7StftP2fc/mY3Y8p8Yxu64x0qP+2LL7OZt8mwNsz5D5zjPTGe1AF6iqo1K1aeOEM++VQyjynwQRkc4wPxqNdYsnheVXk2RkBv3D556cYyelAF6iqf9qWnmQpuk3TAFP3L85OBnjj8aQatZlZW3yYh+/8AuX45xxxzz6UAXaKpHVrMLE2+TE33P3L88454459aX+1LTzJk3SboQS/7l+MHBxxz+FAFyiqLaxZJCkrPJskJC/uHzx14xkdakOpWqzyQln3xKWYeU+AAMnnGD+FAFqiqP9sWX2cTb5NhbZnyHznGemM96k/tK1+0/Z9z+ZjdjynxjG7rjHSgC1RVH+2LL7OZt8mwNsz5D5zjPTGe1SDUrVp44Qz75VDKPKfBBGRzjA/GgC1RVFdYsnheVXk2RkBv3D556cYyelO/tS08yFN0m6YAp+5fnJwM8cfjQBcoqkNWsysrb5MQ/f8A3L8c444559KDq1mFibfJib7n7l+ecc8cc+tAF2iqf9qWnmTJuk3Qgl/3L8YODjjn8Ka2sWSQpKzybJCQv7h88deMZHWgC9RVU6larPJCWffEpZh5T4AAyecYP4VH/bFl9nE2+TYW2Z8h85xnpjPegC9RVX+0rX7T9n3P5mN2PKfGMbuuMdKj/tiy+zmbfJsDbM+Q+c4z0xntQBeoqqNStWnjhDPvlUMo8p8EEZHOMD8ajXWLJ4XlV5NkZAb9w+eenGMnpQBeoqn/AGpaeZCm6TdMAU/cvzk4GeOPxpBq1mVlbfJiH7/7l+Occcc8+lAF2iqR1azCxNvkxN9z9y/POOeOOfWl/tS08yZN0m6EEv8AuX4wcHHHP4UAXKKotrFkkKSs8myQkL+4fPHXjGR1qQ6larPJCWffEpZh5T4AAyecYP4UAWqKo/2xZfZxNvk2FtmfIfOcZ6Yz3qT+0rX7T9n3P5mN2PKfGMbuuMdKALVFUf7Ysvs5m3ybA2zPkPnOM9MZ7VINStWnjhDPvlUMo8p8EEZHOMD8aALVFUV1iyeF5VeTZGQG/cPnnpxjJ6U7+1LTzIU3SbpgCn7l+cnAzxx+NAFyiqQ1azKytvkxD9/9y/HOOOOefSg6tZhYm3yYm+5+5fnnHPHHPrQBdoqn/alp5kybpN0IJf8Acvxg4OOOfwpraxZJCkrPJskJC/uHzx14xkdaAL1FVTqVqs8kJZ98SlmHlPgADJ5xg/hUf9sWX2cTb5NhbZnyHznGemM96AL1FVf7StftP2fc/mY3Y8p8Yxu64x0qP+2LL7OZt8mwNsz5D5zjPTGe1AF6iqo1K1aeOEM++VQyjynwQRkc4wPxqNdYsnheVXk2RkBv3D556cYyelAF6iqf9qWnmQpuk3TAFP3L85OBnjj8aQatZlZW3yYh+/8AuX45xxxzz6UAXaKpHVrMLE2+TE33P3L88454459aX+1LTzJk3SboQS/7l+MHBxxz+FAFyiqLaxZJCkrPJskJC/uHzx14xkdakOpWqzyQln3xKWYeU+AAMnnGD+FAFqiqP9sWX2cTb5NhbZnyHznGemM96k/tK1+0/Z9z+ZjdjynxjG7rjHSgC1RVH+2LL7OZt8mwNsz5D5zjPTGe1SDUrVp44Qz75VDKPKfBBGRzjA/GgC1RVFdYsnheVXk2RkBv3D556cYyelO/tS08yFN0m6YAp+5fnJwM8cfjQBcoqkNWsysrb5MQ/f8A3L8c444559KDq1mFibfJib7n7l+ecc8cc+tAF2iqf9qWnmTJuk3Qgl/3L8YODjjn8Ka2sWSQpKzybJCQv7h88deMZHWgC9RVU6larPJCWffEpZh5T4AAyecYP4VH/bFl9nE2+TYW2Z8h85xnpjPegC9RVX+0rX7T9n3P5mN2PKfGMbuuMdKj/tiy+zmbfJsDbM+Q+c4z0xntQBeoqqNStWnjhDPvlUMo8p8EEZHOMD8ajXWLJ4XlV5NkZAb9w+eenGMnpQBeoqn/AGpaeZCm6TdMAU/cvzk4GeOPxpBq1mVlbfJiH7/7l+Occcc8+lAF2iqR1azCxNvkxN9z9y/POOeOOfWl/tS08yZN0m6EEv8AuX4wcHHHP4UAXKKotrFkkKSs8myQkL+4fPHXjGR1qQ6larPJCWffEpZh5T4AAyecYP4UAWqKo/2xZfZxNvk2FtmfIfOcZ6Yz3qT+0rX7T9n3P5mN2PKfGMbuuMdKALVFUf7Ysvs5m3ybA2zPkPnOM9MZ7VINStWnjhDPvlUMo8p8EEZHOMD8aALVFUV1iyeF5VeTZGQG/cPnnpxjJ6U7+1LTzIU3SbpgCn7l+cnAzxx+NAFyiqQ1azKytvkxD9/9y/HOOOOefSg6tZhYm3yYm+5+5fnnHPHHPrQBdoqn/alp5kybpN0IJf8Acvxg4OOOfwpraxZJCkrPJskJC/uHzx14xkdaAL1FVTqVqs8kJZ98SlmHlPgADJ5xg/hUf9sWX2cTb5NhbZnyHznGemM96AL1FVf7StftP2fc/mY3Y8p8Yxu64x0qP+2LL7OZt8mwNsz5D5zjPTGe1AF6iqo1K1aeOEM++VQyjynwQRkc4wPxqNdYsnheVXk2RkBv3D556cYyelAF6iqf9qWnmQpuk3TAFP3L85OBnjj8aQatZlZW3yYh+/8AuX45xxxzz6UAXaKpHVrMLE2+TE33P3L88454459aX+1LTzJk3SboQS/7l+MHBxxz+FAFyiqLaxZJCkrPJskJC/uHzx14xkdakOpWqzyQln3xKWYeU+AAMnnGD+FAFqiqP9sWX2cTb5NhbZnyHznGemM96k/tK1+0/Z9z+ZjdjynxjG7rjHSgC1RVH+2LL7OZt8mwNsz5D5zjPTGe1SDUrVp44Qz75VDKPKfBBGRzjA/GgC1RVFdYsnheVXk2RkBv3D556cYyelO/tS08yFN0m6YAp+5fnJwM8cfjQBcoqkNWsysrb5MQ/f8A3L8c444559KDq1mFibfJib7n7l+ecc8cc+tAF2iqf9qWnmTJuk3Qgl/3L8YODjjn8Ka2sWSQpKzybJCQv7h88deMZHWgC9RVU6jbCaWIs++FSzjymwAOuDjB/CmHV7MW6zl5PLZioPkPnP0xmgC7RVb+0Lb7Ubfc/mKNxHlNjGM9cY6VH/a9n9m8/fJ5e/ZnyHznGemM/jQBdoqsNQtjcxwBn8yRQyjymxgjPXGBUa6vZtA8weTYhAY+Q+cn2xk0AXaKq/2la+bDHufdMoZP3T8g9MnHH400araFJnDyYhID/uX4ycccc/hQBcoqodUtB5PzSfvvufuX55xzxx+NH9p2u6Zd0mYM7/3T8YOOOOfwoAZrHlf2Y/nhynmR52EA53rjr74q9WRqmrRLpaSW7NuldQmYW5Add2QRxx6/hV06jbCaWIs++FSzjymwAOuDjB/CgC1RVFtYskt1neVkiZtgZonGD75HA9zxV6gAooooAKKKKAIL/cNOufLkET+U22QnAU4POe2KnqtqRQaTdmVSyCB9wU4JG055o1G9Gm6dPeNb3FysK7zFaxGSRh/sqOWPsOaA3PPdR+MEtvfTxaZ4R1LUbaP7SY7lJ4UEwt22zFVLbsKT6ZPYGvQtNvP7Q0q0vQqqLiFJdquHA3AHAYdevWvC01PStc0m8m0zR/G+nXC6heNbix0x5lj3lkljDGIhRIRuZSCUYnnivavDVv8AZPCulW5s2sPKs4k+ytJvMGEA2Fu5HTPtTS93Xy/X/gf1YJfFZef/AACa7A/tKwzNsO98Jg/vPkPH4deaxfFHiPVbDWNN0Tw3YWl3qeoJLMDeztFDHHHt3ElVYkkuoAA9a2rsj+0rDMO873w+T+7+Q8/j05rjvilfrpFhbaz/AMI7catLpSyXkVxbXiwNbFQCd3zKzRsPvAbuByDxU9VcaV9iRviFMqWlpcaetlrI1S3sb+xmct5aykgSxsMb0OMq31BAIIruq8kg8VeE/G3j+0h8QaZcaZqWmNbS6bLdQywSySuWJjPGCuVGMkhs5HSvW6u1lr3/AMiL3lptZfqFFFFSUFV72/t9OhWW7corMEXajMST2wAT2qxWH4p/49bH/r8T/wBBagCx/wAJFp39+4/8BJf/AImj/hItO/v3H/gJL/8AE1kUUAa//CRad/fuP/ASX/4mj/hItO/v3H/gJL/8TWRRQBr/APCRad/fuP8AwEl/+Jo/4SLTv79x/wCAkv8A8TXmnive3iAATTootUIWOZkGdz84BFY+xv8An4uv/AqT/wCKrup4KdSCmnucs8TGEnFo9j/4SLTv79x/4CS//E0f8JFp39+4/wDASX/4mvHNjf8APxdf+BUn/wAVRsb/AJ+Lr/wKk/8Aiq0/s+fdEfXI9j2P/hItO/v3H/gJL/8AE0f8JFp39+4/8BJf/ia8P1i9OlaRcXhnumaJPkT7TJ87HhV+93JAqlousS3Gj3E+s3MltcWUjx3e26lVEK85GWPBUg9e9T9Rle3MivrStex77/wkWnf37j/wEl/+Jo/4SLTv79x/4CS//E14TDr+nTRzv/aF5EIEDyC4knhIU9Gw+CQfUUsOvadNDPL/AGjeRC3AMq3Ek8LKD0O18HB7HHNH1F/zIPrS/lZ7r/wkWnf37j/wEl/+Jo/4SLTv79x/4CS//E14BfeJ7W20W/vbafUJZLOLzGt5HuI356Eq3zBT/exjr6VoLfodJS/eXURGyBiqm4LjP+wPm/Sj6i/5kH1pfys9v/4SLTv79x/4CS//ABNH/CRad/fuP/ASX/4mvnXSPFUmqX1lbg6gPtCSysxS6VQob93hjxyDyemfTIq/p2q3d7r11Zyrd28MUEcsfmXMnmNuZhkjdx93gdfX0pRwUpWtJajeJSvdbHvf/CRad/fuP/ASX/4mj/hItO/v3H/gJL/8TXjmxv8An4uv/AqT/wCKo2N/z8XX/gVJ/wDFVf8AZ8+6I+uR7Hsf/CRad/fuP/ASX/4mj/hItO/v3H/gJL/8TXjmxv8An4uv/AqT/wCKqG8EiWM7JdXYZY2IIupODj/eo/s+fdB9bj2Paf8AhItO/v3H/gJL/wDE0f8ACRad/fuP/ASX/wCJrmryR18PzyKzBxaswYHkHZ1zXn0SuYUJubokqM/6VJ/8VXNQw8q97PY3q1lStc9m/wCEi07+/cf+Akv/AMTR/wAJFp39+4/8BJf/AImvHNjf8/F1/wCBUn/xVGxv+fi6/wDAqT/4qun+z590YfXI9j2P/hItO/v3H/gJL/8AE0f8JFp39+4/8BJf/ia8c2N/z8XX/gVJ/wDFVzuueIbnS9Ygggkmktogsl+7XEpaJHcIpGGx1yTnPAqZYGUbXkiliovZH0L/AMJFp39+4/8AASX/AOJo/wCEi07+/cf+Akv/AMTXh1/q1pprql1eX3mMpYJC88rBR1YhMkD3PFQT+ItMtwhbU7t1eMTboZZ5VVD0diuQq+5wODR9Ra+0g+tJ/ZZ7x/wkWnf37j/wEl/+Jo/4SLTv79x/4CS//E14Xd63Y2cgSa/vGYoJCIZJ5dqHozbM7V46nAqD+3438SW+lxS3siT2puFnSadlPIxyOMYJ5zjoO9H1GV7cyD60rXsz3v8A4SLTv79x/wCAkv8A8TR/wkWnf37j/wABJf8A4mvA9a10aVeWlsH1J5J5QGKLcygJgkkFMgnjpyRnOMVXsdV1fWrCCXTGmiEssm64uWnCxoG+T5C6s5YY5Bx1+lL6lK9lJXH9ZVrtH0J/wkWnf37j/wABJf8A4mj/AISLTv79x/4CS/8AxNeE6Hfz6jbXH2maYTWtw9vI0V1Lscr/ABDLZHXpzg5GTWnsb/n4uv8AwKk/+KqlgJtXUkS8XFOzR7H/AMJFp39+4/8AASX/AOJo/wCEi07+/cf+Akv/AMTXjmxv+fi6/wDAqT/4qjY3/Pxdf+BUn/xVP+z590L65Hsex/8ACRad/fuP/ASX/wCJo/4SLTv79x/4CS//ABNeY+FWkXxMI/PndGs5WKyTM4yHjwcEn1P5129cVWm6U3BnVTmpx5ka/wDwkWnf37j/AMBJf/iaP+Ei07+/cf8AgJL/APE1kUVkWa//AAkWnf37j/wEl/8AiaP+Ei07+/cf+Akv/wATWRWd4gkeHwzqkkTskiWcrKynBUhDgg0AdR/wkWnf37j/AMBJf/iaP+Ei07+/cf8AgJL/APE145sb/n4uv/AqT/4qjY3/AD8XX/gVJ/8AFV6X9nz7o4vrkex7H/wkWnf37j/wEl/+Jo/4SLTv79x/4CS//E145sb/AJ+Lr/wKk/8Aiqq6neDTNLub2a4utkEZfH2mTnHQfe7nik8BNK7aGsXFuyR7b/wkWnf37j/wEl/+Jo/4SLTv79x/4CS//E14BoetT3Gl3Umu3ElrdWLlbrbcyqijG4MMseNpHf1qzDr+nTRzv/aF5EIEDyC4knhIU9Gw+CQfUUvqMv5kP60uzPdv+Ei07+/cf+Akv/xNH/CRad/fuP8AwEl/+JrwqDXdPnhuJRqF5CtsoaYXEs8LIp6Eq+Dg9jjmql/4ptLbR768t59QlltIvMNvI1xG7DsdrDO3P8WMe9J4JpXckCxKbtys+gP+Ei07+/cf+Akv/wATR/wkWnf37j/wEl/+JrxKK9V9LS+MuoCNow5UNOX/AO+PvfhisHTfFb6jfWduDqCm4lmJzHdKPKUkIdx45+XJ6DOODTeBknyuSuCxSaukz6L/AOEi07+/cf8AgJL/APE0f8JFp39+4/8AASX/AOJrwTTtVu73XrqzlW7t4YoI5Y/MuZPMbczDJG7j7vA6+vpUDeIJ5PGFvpNsblrVkk825N1JzIoB2L83bIyffHY0vqUtPe3H9ZSbVj6C/wCEi07+/cf+Akv/AMTR/wAJFp39+4/8BJf/AImvHNjf8/F1/wCBUn/xVGxv+fi6/wDAqT/4qr/s+fdEfXI9j2P/AISLTv79x/4CS/8AxNNk8TaXFG0kssyIgLMzWsoCgdSTtrx7Y3/Pxdf+BUn/AMVVyxLrZ6spmmdTps5IklZxkAepPqazq4OdODm3sXTxMZy5Uj2uiiiuE6gooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigCjrHlf2Y/nlwnmR52AE53rjr74q9VXUhK1iwgiWV96YRlDAjeMnB9Bz+FWqAKupmUaXP8AZmVJdnysxAAP48VaqjrXlf2LdfaN/leWd2zG7Htmr1ABRRRQAUUUUAQX3mDT7jyWCyeU2xmIABwcHnip6rals/sm787d5fkPv29cbTnHvWR431y68PeFJ7zTkja9klitrfzQSiySyLGrMB1ALZx7UAeQ67rWu/a7u91Xxvqehr9o1C2liQxxw2ksZ3WqcpnEkYz1y2Rg+vtfhu4lu/CulXFwk0cstnE7pO26RWKAkMe59a4zV/DHirR9HutZj8a3WpXVpEbmSzvbKD7LPsG4ptVAydMAhiR713ek6guq6LZahGpRLu3SdVPYMoOP1pr4WvT9fz/QJO8k/X9Py/US63/2hY7ZxGu990ZYjzPkPGO+OvPpWJ458CaR470Gey1O1t2uvJkjtLyWHe1qzD7y8g9QDjIzitm8aManpweMs7SPsbdjadhzx344qt4sj1Sbwdq0fh9imptaSC1YHBEm07cHsc9Kl7FR3Ry2gfC4eHfFEmqaX4m1WOBkgR7JpfOWTy85DtLvbB3cBSu3nrmvQK8C8JRRt4pNn4a03V7S7TXYLkST208Qjtvs8a3Hmu4AYsysMEklsEete+1o9r/1sv8AhvkzNWTt/XUKKKKgoKw/FP8Ax62P/X4n/oLVuVh+Kf8Aj1sf+vxP/QWoA5aHxXpc/jC58Nx3Ef2+3gSZk81cndn5QM5yAATxwGHrV+z1Oz1C4vIbSbzJLKbyLhdpGx9obHI54YHIyOa51rN/+Fk6h9qhuxa6jpUMMc8CyKoZHlLAyp/qzhhgkgnPHNJ4Z8EJoWt6pfyXN3J9ovPOtl/tO5kGzylX94rNtdshuW3cY54GGrde343B76f1oaNh4oh1HxRNpNtbS+XFbeeLpxtWQh9hVQRkgEHnp6ZrdrA+zz/8LGNz5Mnkf2V5fm7Tt3ebnbnpnHOK36S+FfP82L7TX9bI4zxT/wAjD/26x/8AoclZNa3in/kYf+3WP/0OSsmvosJ/AieRiP4rCiiiukwMTxBpt3q9xYWsMklvaJL5888ewsCnKKAwPVuehHy1ROhXttqWoBt+r2eo2mJhdPHHmVeADsUcFTjIU/d57V00kiQxNJKwREBZmJ4AHU1DYX8GpWi3NoZDE/3WkiePcPUBgCR79DWbhFvz/pbGim0v69Tm7fTdSlhuYJLG6fTWttn2HUbtHdpMg/JIpYgYzyxznGMdaq3Oh6zqNlPAi3kVpGYZIba/ulaVnRwxAkjYkAgYyWJz6V29RPcwx3EUDyqssuTGhPLY64HtSdKPVgqj6HKyaDJeaZqoh0y7tLi4smt43v8AUWndicnbje4C575z7Vs3SX1z4WkS2jntL025CJujLq4HAz8y8/yPatain7NWa7/8H/MXO7p9v+B/kc1ovh6807VTLLqV29vBbxQQo4h2uoByDtQEAEj0JxyTWhDZzp4su7xkxbyWkUavkcsrOSMdejCtWiqUUhcz18woooqiQqvff8g+4/65N/I1Yqvff8g+4/65N/I0PYa3PQb3/kW7j/r0b/0CuCh/494/90fyrvb3/kW7j/r0b/0CuCh/494/90fyrysu+18jvxn2SSiiivVPPCuUXwzeamuqzajf3Vm2ouyNbxCFl8ofKgJZGPI54I6+tdXVO51S0tLyC0mdzcXGfLjjiZyQCASdoOAMjk4FROMZfEXGTWxzrWGpNY6fLe6bePqMEBhe5sLxI5AQcchmCspwG5zgn7tNm03VxHDNNaXUmqG2EbX1hdImW5wJEYhSBnqA3fAFdhRUukn1H7RrocbJoeoR6pLd39tdag91bxLI2n37WwWVVw24eYgKnqDyRzxV+z0qfT9a0uW3sdlrHYNaukUwcW53KwyXILDgjIBPtW9Bcw3SF7eVZVVihZDkBgcEfgeKlpxpxTuvX8/8xObasznde0rU73U1ubC8uYVt7SUxJF5ODMcAD51PUZyfpjHNQTWutaPotlpmlR3d8pUJNchoA9ugUDCAlAT6E5x1OeldTRR7Na67j53poZ2iRrBpqwR6dPp6RnAjnZGZu5YlGYHJJySc5zWjRRWhAUUUUCNHwx/yNaf9eU3/AKHFXcVw/hj/AJGtP+vKb/0OKuwv4bi40+4hsrr7JcyRssVx5Yk8piOG2ng4POD1r5/G/wAeX9dD18L/AA0WKKyH07WT4V+wLru3VvJCf2p9jQ/P/f8AKzt/CjU9O1m68OJZ6drv2HUgqBtR+xpJuIxuPlk7Ru5+ma5Op0GvWZ4k/wCRU1b/AK8pv/QDWkoIQBjuIHJxjNZviT/kVNW/68pv/QDT6i6HE0UUV9UeCFYviLTrvV1s7GBnhtmnElzOhXcip8ygBs5JYL2I4Oa2qKUoqSsxptO6OVm0G+t9VuX3Saxa39m0Vyl28cfzL9wfIg4ILDOCelMt9N1KWG5gksbp9Na22fYdRu0d2kyD8kiliBjPLHOcYx1robnVLS0vILSZ3NxcZ8uOOJnJAIBJ2g4AyOTgVcrP2cXs/wCv6ZfOzirjRNZ1LTb21T7XBalImgt9QuVeQyI4Y4kjZiFIGMlic88VJLoEl9p+oeVpV5a3MtjJbxyX+pNOxLY+UDe4C5A5yDx0rsait7mG6h822lWWPJUMhyMg4P6g0vZR/r+vMftJL+v67FNhdzeHWEUc1pdmAhUzGXVgOBn5l/mKx9E8OX+n6nC8+pXbQWtpHFGGEJDkkl14TOOF5zk+prqaK0cU5c39f1qQpNR5TJSxuD4nvbohkgmsooklVhncGcnA9gw7Vkw+Fbqw1zR2tNRvZrO0WXeZPI+XOOOEBO45yeT7iusopOmnb+utx87/AK9LBRRRVkBVmz/49tV/7Bk/8hVarNn/AMe2q/8AYMn/AJCuXGfwJf11OjD/AMVHtdFFFfPHrhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAFHWFjfTHE0hjTzI/mC7ud644+vFXqp6pv8A7Pby4BO29P3ZUtn5xzgenX8KuUAVdTEp0ucWyK8uz5VYAgn8eKtVR1oRtot0J2ZIzGdzKMkD6VeoAKKKKACiiigCC+8w6fceSoaTym2KwBBODgc8VneKfDdt4s0CTSry6urSN5I5RPZuqyoyOHUqWBA5Udqv6kEOk3YlYqhgfcVGSBtOeKs0AcBP8KpLm3kguPH/AIykilUo6Nfw4ZSMEH9zXa6ZYRaVpNpp9sztDaQpBGXILFVUAZx3wKtUUCtrcq3JlF9ZBJVRC7b0LAFxsOAB355rB8dWxTRZdXl8SatolrpsEk0/9miImRQM8h0bJGOMY61tXnlf2np3mB9/mP5e0jAOw5z+FVfF39n/APCGav8A2zcPa6ebOUXMyKGZI9p3EAggnHbB+lTLa5cfiSPONGl8T6ZrWk3Xii+8T2umXlzHHA099ZzKzv8AcSdEhVk3Hj5Wbk9RXr9eJWPh46tJpFqdW+JQhtpUa0lu7WARxHGFdi0fYHq2cV7FpdlJp2mQ2k99cahJGCGubrb5knOcttAHtwBWj2+Zmt/kW6KKKkoKw/FP/HrY/wDX4n/oLVuVheLGEdhaSPkJHdoWbGdo2sMn86AKNFU/7Vsf+fhfyNH9q2P/AD8L+RoAuUVT/tWx/wCfhfyNH9q2P/Pwv5GgDmPFP/Iw/wDbrH/6HJWTWxr8Ml/qwubGezaPyFjImldCCGY9kPHzCs3+zb//AJ66Z/4Eyf8Axqvaw+JpQpKMnqebWo1JVG0iGipv7Nv/APnrpn/gTJ/8ao/s2/8A+eumf+BMn/xqt/rlD+b8zL6vV7GR4giWfw5qCOXA+zuco5Q8KT1BBrj7Fb+Wa20y2Zjbw6ZDPGsmqTW7FmzubcqsWA4G0naPSvRX0q8ljZJG0t0YEMrXMhBHoR5VVbnwul5DFFeWmhXEcP8Aq0ldmCfQGHisZ4ii5XUu36+Xn+BpGjVStb+tDlhPNdwafZ3l3NeXX2RppHsb7yYXAOA5kXazH6cdSR0rL0JpNWvvDd5eXE00g+2IJFuXwQj/AC8gjdwOpHzAc139z4a+2iIXltolwITmISyM2w+2YePwpP8AhGFMcaG10PZFJ5sa72wj5zuA8ng579aXt6Ld3L8x+yqWtynHW+qzf8I5oBe+kM8+piJiZTukAd8qeckYA4qtJcXtvoN/qcN7ctcHU5LbdLdOI4ojPt6chcD+LaSM+1dwPCsa3TXIs9CE7OHaUO24sOhz5Ocj1qddDmSF4kTSFjkJZ0EzgMTySR5XOe9JV6VtZ/n5f5P7w9lU6R/Lz/z/AAOLaDV7aO+t11SzsQ0MZjjl1SS4dGL4zvkQMoYfLnnB6VqeHZzHfXmnTpeRXESRytFPd/aUUNnlJD83JB4b04FbVv4XS0tZLa1tNCggk+/FG7Kr/UCHBqSz8Otp0Ri0+HRbWMnJSCV0BPrgRVccRRi7835/5f12JlRqNW5R1FTf2bf/APPXTP8AwJk/+NUf2bf/APPXTP8AwJk/+NVr9cofzfmR9Xq9iGq99/yD7j/rk38jV7+zb/8A566Z/wCBMn/xqo59Jv5reSLztMXehXP2mTjI/wCuVDxdC3xfmH1er2O3vf8AkW7j/r0b/wBArgof+PeP/dH8q7i4vLObSpbVbpAzwGMMQcZK4zXHppd+kar5umHaAM/aZP8A41Xn4KtClzc7sdeJpynblRHRU39m3/8Az10z/wACZP8A41R/Zt//AM9dM/8AAmT/AONV6H1yh/N+ZyfV6vYhrkdbh8vxtDdwec08el3EkaLO4UsrJgbQcY55GMHjPQV2n9m3/wDz10z/AMCZP/jVRnRrk3Czn+yTMqlFkM77gp5IB8rOOBx7VE8TQl9r8+xUaNWL2OH0KTVHOk38l7biO6XdMZNTklNyNhJCRNGFVgecLjGCDmoH1CT7Rot9Zz3ypqF6Iy9zfEmaM56Qj5F6dgpHHfNdxB4ZFrdSXNta6HDcSffljkZXb6kQ5NMXwlCkzTJY6AsruJGcMwZmByGJ8nqD3rP6xS097r5+XkX7Kpr7pxGkg6da6aILi4jWbXp45Q9w7BgDKADuJ64H1PJyalvb24utTvoItQuFi/tm2gzDOy7UMY3KCDwCc5x3967R/C6yLOslpobrckNOGdj5pHQt+55/Gli8MCBQsNrocah1cBJGADKMKf8AU9QOAe1JV6NkubTTv5f5fiN06jbfL37ef+Zx90uoS6/e6bbSt5FlbxtD52qzW7AEHMhIRjJzx8xI46c024/tOWOO5udQW/SKxR5Rpmo+S0bckygcLIGwMb8Dg8GuzvPDX9oMhv7bRLoxnKGaRn2n2zDxSXXhhb6SJ7210O4eH/VtK7OU+mYeKPb0rP3/AOtfIXsqn8oywuUvNNtrmEu0c0SupkGGIIzyPWrFSjTb4DAk0zH/AF8yf/GqX+zb/wD566Z/4Eyf/Gq6frlD+b8zH6vV7ENFTf2bf/8APXTP/AmT/wCNUf2bf/8APXTP/AmT/wCNUvrlD+b8x/V6vYt+GP8Aka0/68pv/Q4q7iuM0C2ew1n7Zf3FmqLbvEqwyO5JZkOeUXH3P1rp/wC1bH/n4X8jXj4qcZ1XKOx6NCLjTSZcoqn/AGrY/wDPwv5Gj+1bH/n4X8jXMbFyszxJ/wAipq3/AF5Tf+gGp/7Vsf8An4X8jVTVbq01DRr2zjuo1e4t5IlZgcAspGT+dAHG0VN/Zt//AM9dM/8AAmT/AONUf2bf/wDPXTP/AAJk/wDjVfQ/XKH835nkfV6vYhoqb+zb/wD566Z/4Eyf/GqP7Nv/APnrpn/gTJ/8ao+uUP5vzD6vV7HF63D5fjaG7g85p49LuJI0WdwpZWTA2g4xzyMYPGegqro0uqbNL1GW+t1iuU3Ts+pyTNcjYSQkLRhVYHnC4xgg5rujo1ybhZz/AGSZlUoshnfcFPJAPlZxwOPaoIPDItbqS5trXQ4biT78scjK7/UiHJrD29FO6l+fd+Rr7KpazicOl7J/aOgz2s98IdSlZXa5vy7TxlGIPlD5U6A5XaRwO5wmhbrC18PpDPOiT6lcrKrzuwbAlwDk+oBx689a7aLwnDBL5sNjoMcnmeZvRmB3/wB7Pk9eTzSv4Vjlhmils9CeOd/MlRnYiRv7zDyeT7miOIpLXm/Py8vIHSqPS39a/wCZxc99cXkt5HFqFwIm1+O3DQzsCsflruQEHgZz0706ddSn1jVLK3nKRaeiCBp9WlgaNSmfMbCN5gznlyR8uMV2sfhrylCxW2iIokEoCyMBvAwG/wBT1AAGaS68Mi/ljlvrbQ7mSLmN5pGcp9CYeKXtqNvj/Psl/wAEfs6l/h/q5yN9/aUpS4nvReJFYpJKmm6gYHifBJlUcLIrdt/HHQ11dhcpeabbXMJdo5oldTIMMQRnketPuvDC30kcl9a6HcvF/q2mkZyn0Jh4q0NNvgMCTTMf9fMn/wAarWOKopt82/r/AJGcqFRpaEVFTf2bf/8APXTP/AmT/wCNUf2bf/8APXTP/AmT/wCNVf1yh/N+ZP1er2Ias2f/AB7ar/2DJ/5Cmf2bf/8APXTP/AmT/wCNVYt7Sa3stSNxLas8tlLDFHbyPIzsw6covp79a58TiaU6TjF6mtGjUjUTaPZqKKK8Y9IKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooApasAdNbdN5I3x/PgnHzrxx69Pxq7VLViBprbofOG+P5MkZ+deePTr+FXaAKmqCQ6XceTCs77PljZdwb2x3q3VLWAjaPciSXyVKHMgBO33wKu0AFFFFABRRRQBBf7jp1z5cYlfym2xkZDHB4x3zU9VtSCnS7sO/lqYXy+M7RtPOKs0AFFFFAFW5877dZeVt8ve3mZxnGw4xnnrjpWT49kMPw71+RYraYrp8x8u6/1TfIeG9q0rzyv7T07zC+/zH8vaBgnYc5/Cqni/V5tA8Gavq1rAtxNZWkkyRMOGKqSM+1TL4WVD4kedeD/EOlXF1pEMHxgm1J3Marp8kVsDMeP3Z/d7x6dc+9evV4zYeJ/Ftr4mgj1TxBpdzZx6nZ20iWumKjTR3EYZHRi543ZU+wJz2r2atZbX8/8AL/Myjpp5f5hRWfLfXbzXEOn2sMslu6q3nzmMEFQ2QQjeuMYqQS6l9qRTaWvkFRuf7U24HHIC+XgjPGcjjnHaoLLlFZ6zax9nYtY2ImDDagvXKkdyW8rIPTjB+tSCXU/OhBtLTymUea32ptyHuFHl/MPckZ9qALlFUBNq/lylrGyDgjylF45Dc85PlfLx6A/hS+bquYMWdnhv9d/pbfJz/D+7+bj12/1oAvUVR83Vcz5s7PC/6n/S2+fn+L938vHpu/rSGbV/LiK2NkXJPmqbxwF54wfK+bj1A/GgC/RVMy6n50wFpaeUqnym+1Nuc9gw8v5R7gnHvUbTax9nUrY2JmLHchvXCgdiG8rJPXjA+tAGhRVMy6l9qdRaWvkBTtf7U24nHAK+XgDPGcnjnHao/O1j7Nu+w2Pn78bPtr7duOu7ys5z2x+NAGhRVMS6l9qRTaWvkFRuf7U24HHIC+XgjPGcjjnHao1m1j7OxaxsRMGG1BeuVI7kt5WQenGD9aANCiqYl1PzoQbS08plHmt9qbch7hR5fzD3JGfamCbV/LlLWNkHBHlKLxyG55yfK+Xj0B/CgC/RVHzdVzBizs8N/rv9Lb5Of4f3fzceu3+tHm6rmfNnZ4X/AFP+lt8/P8X7v5ePTd/WgC9RVAzav5cRWxsi5J81TeOAvPGD5XzceoH408y6n50wFpaeUqnym+1Nuc9gw8v5R7gnHvQBcorPabWPs6lbGxMxY7kN64UDsQ3lZJ68YH1qQy6l9qdRaWvkBTtf7U24nHAK+XgDPGcnjnHagC5RWf52sfZt32Gx8/fjZ9tfbtx13eVnOe2PxqQS6l9qRTaWvkFRuf7U24HHIC+XgjPGcjjnHagC5RWes2sfZ2LWNiJgw2oL1ypHclvKyD04wfrUgl1PzoQbS08plHmt9qbch7hR5fzD3JGfagC5RVATav5cpaxsg4I8pReOQ3POT5Xy8egP4Uvm6rmDFnZ4b/Xf6W3yc/w/u/m49dv9aAL1FUfN1XM+bOzwv+p/0tvn5/i/d/Lx6bv60hm1fy4itjZFyT5qm8cBeeMHyvm49QPxoAv0VTMup+dMBaWnlKp8pvtTbnPYMPL+Ue4Jx71G02sfZ1K2NiZix3Ib1woHYhvKyT14wPrQBoUVTMupfanUWlr5AU7X+1NuJxwCvl4AzxnJ45x2qPztY+zbvsNj5+/Gz7a+3bjru8rOc9sfjQBoUVTEupfakU2lr5BUbn+1NuBxyAvl4IzxnI45x2qNZtY+zsWsbETBhtQXrlSO5LeVkHpxg/WgDQoqmJdT86EG0tPKZR5rfam3Ie4UeX8w9yRn2pgm1fy5S1jZBwR5Si8chuecnyvl49AfwoAv0VR83VcwYs7PDf67/S2+Tn+H9383Hrt/rR5uq5nzZ2eF/wBT/pbfPz/F+7+Xj03f1oAvUVQM2r+XEVsbIuSfNU3jgLzxg+V83HqB+NPMup+dMBaWnlKp8pvtTbnPYMPL+Ue4Jx70AXKKz2m1j7OpWxsTMWO5DeuFA7EN5WSevGB9akMupfanUWlr5AU7X+1NuJxwCvl4AzxnJ45x2oAuUVn+drH2bd9hsfP342fbX27cdd3lZzntj8akEupfakU2lr5BUbn+1NuBxyAvl4IzxnI45x2oAuUVnrNrH2di1jYiYMNqC9cqR3Jbysg9OMH61IJdT86EG0tPKZR5rfam3Ie4UeX8w9yRn2oAuUVQE2r+XKWsbIOCPKUXjkNzzk+V8vHoD+FL5uq5gxZ2eG/13+lt8nP8P7v5uPXb/WgC9RVHzdVzPmzs8L/qf9Lb5+f4v3fy8em7+tIZtX8uIrY2Rck+apvHAXnjB8r5uPUD8aAL9FUzLqfnTAWlp5SqfKb7U25z2DDy/lHuCce9RtNrH2dStjYmYsdyG9cKB2Ibysk9eMD60AaFFUzLqX2p1Fpa+QFO1/tTbiccAr5eAM8ZyeOcdqj87WPs277DY+fvxs+2vt2467vKznPbH40AaFFUxLqX2pFNpa+QVG5/tTbgccgL5eCM8ZyOOcdqjWbWPs7FrGxEwYbUF65UjuS3lZB6cYP1oA0KKpiXU/OhBtLTymUea32ptyHuFHl/MPckZ9qYJtX8uUtY2QcEeUovHIbnnJ8r5ePQH8KAL9FUfN1XMGLOzw3+u/0tvk5/h/d/Nx67f60ebquZ82dnhf8AU/6W3z8/xfu/l49N39aAL1FUDNq/lxFbGyLknzVN44C88YPlfNx6gfjTzLqfnTAWlp5SqfKb7U25z2DDy/lHuCce9AFyis9ptY+zqVsbEzFjuQ3rhQOxDeVknrxgfWpDLqX2p1Fpa+QFO1/tTbiccAr5eAM8ZyeOcdqALlFZ/nax9m3fYbHz9+Nn219u3HXd5Wc57Y/GpBLqX2pFNpa+QVG5/tTbgccgL5eCM8ZyOOcdqALlFZ6zax9nYtY2ImDDagvXKkdyW8rIPTjB+tSCXU/OhBtLTymUea32ptyHuFHl/MPckZ9qALlFUBNq/lylrGyDgjylF45Dc85PlfLx6A/hS+bquYMWdnhv9d/pbfJz/D+7+bj12/1oAvUVR83Vcz5s7PC/6n/S2+fn+L938vHpu/rSGbV/LiK2NkXJPmqbxwF54wfK+bj1A/GgC/RVMy6n50wFpaeUqnym+1Nuc9gw8v5R7gnHvUbTax9nUrY2JmLHchvXCgdiG8rJPXjA+tAGhRVMy6l9qdRaWvkBTtf7U24nHAK+XgDPGcnjnHao/O1j7Nu+w2Pn78bPtr7duOu7ys5z2x+NAGhRVMS6l9qRTaWvkFRuf7U24HHIC+XgjPGcjjnHao1m1j7OxaxsRMGG1BeuVI7kt5WQenGD9aANCiqYl1PzoQbS08plHmt9qbch7hR5fzD3JGfamCbV/LlLWNkHBHlKLxyG55yfK+Xj0B/CgC/RVHzdVzBizs8N/rv9Lb5Of4f3fzceu3+tHm6rmfNnZ4X/AFP+lt8/P8X7v5ePTd/WgC9RVAzav5cRWxsi5J81TeOAvPGD5XzceoH408y6n50wFpaeUqnym+1Nuc9gw8v5R7gnHvQBcorPabWPs6lbGxMxY7kN64UDsQ3lZJ68YH1qQy6l9qdRaWvkBTtf7U24nHAK+XgDPGcnjnHagC5RWf52sfZt32Gx8/fjZ9tfbtx13eVnOe2PxqUSaj9qVWtbUW5XLSC5YuGx0C+XgjPGcjjnHagC3RVATav9lZmsbIT7gFQXj7SvclvKyD7Y/GpBJqP2iEG1tfJKgyv9pbcjdwq+X8w9yRn0FAFuiqKzat5MxayshICPKUXjkOM87j5Xy8egOfanGXU98GLS02tjzz9qbKc87R5fzceu38OtAFyiqQl1T9/mzs/l/wBR/pbfPz/F+7+Xj03f1pDLqvlwEWVnvYnzh9rfCc8bT5fzceu3096AL1FU/N1Lzpx9ktfLVT5Lfam3OewYeX8o9wWx6GmNNq4tkZLGyM5J3obxwqjthvKyfyH40AX6KqmTUPtUii1tvs4U+XJ9pbexxwCuzAGe+T9Ki87V/su77DZfaN+PL+2vs2467vKznPbH49qAL9FVfM1D7Yq/Zbb7Nt+aT7S28HHTbswRnjO7pz7VEJtX+yszWNkJ9wCoLx9pXuS3lZB9sfjQBfoqoJNR+0Qg2tr5JUGV/tLbkbuFXy/mHuSM+gpizat5MxayshICPKUXjkOM87j5Xy8egOfagC9RVMy6nvgxaWm1seeftTZTnnaPL+bj12/h1pBLqn7/ADZ2fy/6j/S2+fn+L938vHpu/rQBdoqiZdV8uAiys97E+cPtb4TnjafL+bj12+nvTvN1Lzpx9ktfLVT5Lfam3OewYeX8o9wWx6GgC5RVBptXFsjJY2RnJO9DeOFUdsN5WT+Q/GpTJqH2qRRa232cKfLk+0tvY44BXZgDPfJ+lAFqiqHnav8AZd32Gy+0b8eX9tfZtx13eVnOe2Px7VL5mofbFX7LbfZtvzSfaW3g46bdmCM8Z3dOfagC1RVATav9lZmsbIT7gFQXj7SvclvKyD7Y/GpBJqP2iEG1tfJKgyv9pbcjdwq+X8w9yRn0FAFuiqKzat5MxayshICPKUXjkOM87j5Xy8egOfanGXU98GLS02tjzz9qbKc87R5fzceu38OtAFyiqQl1T9/mzs/l/wBR/pbfPz/F+7+Xj03f1pDLqvlwEWVnvYnzh9rfCc8bT5fzceu3096AL1FU/N1Lzpx9ktfLVT5Lfam3OewYeX8o9wWx6GmNNq4tkZLGyM5J3obxwqjthvKyfyH40AX6KqmTUPtUii1tvs4U+XJ9pbexxwCuzAGe+T9Ki87V/su77DZfaN+PL+2vs2467vKznPbH49qAL9FVfM1D7Yq/Zbb7Nt+aT7S28HHTbswRnjO7pz7VEJtX+yszWNkJ9wCoLx9pXuS3lZB9sfjQA/VN/wDZ7eXOIG3p+8LFcfOOMj16fjVysbV5r1bRPtFjbPbkxecVu2DK+8cKPL+YA45JGfQVbWbVvJmLWVkJAR5Si8chxnncfK+Xj0Bz7UAO1bP9k3OIfPOz/VYJ3e3HNXKx9Xk1P+yyBFaQK0f76Y3bgwnPO3Efzcf7tWxLqn7/ADZ2fy/6j/S2+fn+L938vHpu/rQBdoqg1zqUaW7SWdr8zhZwt2f3eWABXKDdwc4O30Gav0AFFFFAFfUP+QZdYj80+S/7vn5+Dxx61YqvqH/IMusSeUfJf95z8nB549KsUAFFFFAFW5EpvrIpEroHbe5UEoNhwQe3PFUPF+o3OkeDNY1CwtxdXNrZyyxQsu4OwUkAjuPard4sZ1PTi8hV1kfYu3O47DnntxzVqeeK2t5J7mRIoY1LySOcKqgZJJPQUpbDjo0eDaVexXnjPRNZsJdNu4rG8ttG01ILGFVu1Me65mjIG5Qm/jadq4I7177XH6b8SfAWoalbafpOvadcXcj+XBDbjLFm7DA/Ouwq29P6/rfX5k9f6/r/AIYq2xl+3XvmMpj3r5YBGVGwZz+PrXnWoy+Ldc8Va7Z6Tr8+m3VncxQw6cIowjWThN9yrMuS+S+CDgFMYr0Gz8r+09R8vfv8xPM3YxnYMY/CvHfH+qa7o/xI06W3vvEmn273U3lySWcd9aEeSf8AVxxHzWHqrdOSOlRe0tSvss7XwfPNZ+ONZ0Gz1e+1nS7K3id5r2UzPbXJZt0XmEZPyhW2nO3PocV1etanJpdpHJBbrcSyyiJEeTy1yQTkkA+npXnnwN8T3OteGUs7rU9KvmtotzC3kcXSMXb/AFyNn8wfbFdv4p/49bH/AK/E/wDQWq3pYnqyt/wkOr/9Amx/8GD/APxmj/hIdX/6BNj/AODB/wD4zUNFSMm/4SHV/wDoE2P/AIMH/wDjNH/CQ6v/ANAmx/8ABg//AMZqGigCC78dSafMIb+LRrWUqHCTavsYrkjODF04P5VB/wALIh/566B/4Ox/8armvFP/ACMP/brH/wChyVk16VHBKpBT5tzjqYnkk42O7/4WRD/z10D/AMHY/wDjVH/CyIf+eugf+Dsf/Gq4Sitf7OX834Gf1x/ynd/8LIh/566B/wCDsf8Axqj/AIWRD/z10D/wdj/41Xl/iO9lsdElNqcXM5WCDH/PRztB/DOfwrN0rUptI0fVLfUWnvJtIZjuyXkmiI3IcnqcEj/gNZvAxTacvwLWKbV0j2L/AIWRD/z10D/wdj/41R/wsiH/AJ66B/4Ox/8AGq8jh8Tyf2PdalPaQyW8MQkSSxu1nV/YnCkEd+Me9NbxX9m0eXUL+C2WFSixS2t6s0Upc4A3kLtwepIxjuaPqUF9v8A+sy/l/E9e/wCFkQ/89dA/8HY/+NUf8LIh/wCeugf+Dsf/ABqvEb3xlJc6Lqyae9l9vtbQzpJaXonjC9Cd237w/uleeOa6Rbi9XQ0nMMBuhEGaNpyI89/n2Z6f7NCwUGm1L8PX/IHiZLeP4/13PSv+FkQ/89dA/wDB2P8A41R/wsiH/nroH/g7H/xqvE/D97rd3qGnrdxwiE2z3Up+2OzESNlMqUA45AGcY78YrS0yW/Hiy+hv7kSD7JFIsUYIjjJeQcZ68AZJ6+3SiOBTt72/kDxLV9NvM9a/4WRD/wA9dA/8HY/+NUf8LIh/566B/wCDsf8AxquEorT+zl/N+BH1x/ynd/8ACyIf+eugf+Dsf/GqD8SYACTLoAA6k62P/jVcJVe+/wCQfcf9cm/kaX9nL+b8A+uPseqnxLqqxl20uwCAZLHUWwB658ms3/hZEB6S6B/4Ox/8apt7/wAi3cf9ejf+gVwUP/HvH/uj+VcmGw/t762sdFat7K2h3/8AwsiH/nroH/g7H/xqj/hZEP8Az10D/wAHY/8AjVcJRXZ/Zy/m/A5/rj/lO7/4WRD/AM9dA/8AB2P/AI1R/wALIh/566B/4Ox/8arhK4/xDf3/APa0tzp80q2+iRpNcRo5CzFjllIHXEYJ+rColgYx3l+BUcU5bI9r/wCFkQ/89dA/8HY/+NUf8LIh/wCeugf+Dsf/ABqvJ9U8R/Yr61tLWK3kkuYjLG1zc+Qjj+6rbW3Nz09O9RXXieaHUI7GKwiFy0CzmK6vFiZs5+WPhg7DBzyB05pfUofzfgNYmT+z+J67/wALIh/566B/4Ox/8ao/4WRD/wA9dA/8HY/+NV4/qnjCDT9Qksh9jSaGNZJVvL5Lf7wJ2pwdzcew5HNNs9du9T8T6f8AYlQ6ZdacbjDyYb7y842nkZxjOOSfqLBQcuVT/D1/yD6zJK7iexf8LIh/566B/wCDsf8Axqj/AIWRD/z10D/wdj/41Xkmv3eqx6lp9vp0Nu0byGRme7eIsEUkqQEOF+7zk56Y71X0m2v9X0O1udUvJLaCRpLh0t7lw0isdyZkG1kCg4wOwHPal9RTbSlt5B9adrtHsf8AwsiH/nroH/g7H/xqj/hZEP8Az10D/wAHY/8AjVeVeF557jS5Xlkllt/tDi0kmOXeEH5ST1PfBPJGDW1Wiy9NJ834EvFtO1ju/wDhZEP/AD10D/wdj/41R/wsiH/nroH/AIOx/wDGq4Sin/Zy/m/AX1x/ynodn45l1CVo7CHR7qRRuKwauXIHrgRVc/4SHV/+gTY/+DB//jNcD4Y/5GtP+vKb/wBDiruK86vT9lUcLnZSn7SHMTf8JDq//QJsf/Bg/wD8Zo/4SHV/+gTY/wDgwf8A+M1DRWJoTf8ACQ6v/wBAmx/8GD//ABmmyeJdUhiaSXTLBI0BZmbUWAUDqSfJqOszxJ/yKmrf9eU3/oBoAl/4WRD/AM9dA/8AB2P/AI1R/wALIh/566B/4Ox/8arhKK9f+zl/N+B5/wBcf8p3f/CyIf8AnroH/g7H/wAao/4WRD/z10D/AMHY/wDjVcJWbr+oNpuh3FxCN05AjhXP3pGO1R+ZFTLL1FX5vwGsW27WPTf+FkQ/89dA/wDB2P8A41R/wsiH/nroH/g7H/xqvG9J1C40PTtVs9Te4vpdLHnKxYvJNEw3A5PU5DD8KsQ+J5P7HutSntIZLeGISJJY3azq/sThSCO/GPep+pQtrL8CvrMukfxPXP8AhZEP/PXQP/B2P/jVH/CyIf8AnroH/g7H/wAaryP/AISn7PotxqWoQW4t4gpjltLxZo5SxwAGIXBzjORjnrWXf+NZJtJ1OKwex+3w2bXET2l8LhAo4JJCcMM52kYPr6KWDgvt/gOOIm/s/ie4f8LIh/566B/4Ox/8ao/4WRD/AM9dA/8AB2P/AI1Xm8M14dHjlaGE3flBjGZjsJx/f25/HbXM6Hf65eahponjhMMomvGZb12JjZsKCvlgYAbgZwcZ4xVSwMVLl5vwJWKbjzW/E9u/4WRD/wA9dA/8HY/+NUf8LIh/566B/wCDsf8AxqvJNNmv18V38Wo3QdRZxSiNMiOLLyDjPXgDJPX26VSi1S9v/GWmzRTsmlTxTiCIZHnbQv7w+oJPA9BnvU/Uo6e9v5f8Ep4lpvT+tz2j/hZEP/PXQP8Awdj/AONUf8LIh/566B/4Ox/8arhKK1/s5fzfgZ/XH/Kd3/wsiH/nroH/AIOx/wDGqkT4gNLb3E1vBpV0ltGZZVtdV8xgoGTwIv5kVwFWbP8A49tV/wCwZP8AyFY1sEqVNz5tjSnieeajY9rooorzjsCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAKupGVbFjBKsT70w7MFAG8ZGT6jj8atVR1jyv7Mfzw5TzI87CAc71x198VeoApawUXR7kyRecoQ5jBI3e2RV2qmqGQaXceTMsD7PlkZtoX3z2q3QBQ1nyPsMf2rzNn2u2x5eM7vPTb17bsZ9s1fqrqInNqn2WJZX8+HKsAQF81dx57hcke4q1QAUUUUAVtSKjS7sunmKIXymcbhtPGas1Bf7hp1z5cgifym2yE4CnB5z2xU9ABRRRQBTut/9oWO2ASLvfdIVJ8v5Dzntnpz61jfEJtMk8BazZ6zqcWm29zZSoZ35KjbyQo5bHoOa2LsD+0rDM2w73wmD+8+Q8fh15rF+ItrocvgTVLrxNo8Gr2llbST+RKozkKfut1Ru24YIpPYqHxI8+ktpYPEGk3bapoUcGoajYXMslrcs8kNwiCIxxIEyVkUKCTtwCc17RXh+gXUukeLLZZfAfgzSVttRt7Ga8soiZoTNGHRlPlqTncEzn73bHNe4Vo/h+f6L9Lf1vmt/kv1KtsJft175iKI96+WQBlhsGc/j61R8R6noOh20Gq+IngiW2c/Z5ZIvMkVypBEYALFiueFGcZq1ZiManqJRmLmRN4IwAdgxj14rzj4neJtAa8tltvEeix6xpEkyvYX935QYSxNGRuwdrAMCDg9x3zWUm0tDSKu9TuPDGr+G9RtRD4Ye2EcEMb+TFF5RjjkXeh2kAgEHP1z3BpfFP8Ax62P/X4n/oLVwPwcnh1K8N1JqmkT3VhpFtpi2um3f2jEcZP72Rto5YngDOAOprvvFP8Ax62P/X4n/oLVrNJPQzi29/60OQt/FMdx47vPDn2WdWtrWOfzzBJtYsWyN23aAAowc8kkDkGr2n63BqdxqcFrFN5mm3H2eUMAN77Ff5eemHA5xzmsqTR5n8fXtxdacLzS9S02O1kZvLaNCjSEq6MckMHAGAe+cUzRPAWi+HNS1HU7PS7CS4muDPaiGyjje2Hlqvlo3YEgnsPmP1qdLa9vxv8A5FPfTy/L/MjXXfEOn3ujtr0FgkWrXH2cWUKt51oxRmGZN5WXG3Bwq47Zrrq4fQ31y716PU/E/hjUI7vJig2zWr29jGTyRibezEY3Ntz2Ax17ijoLqcZ4p/5GH/t1j/8AQ5Kya1vFP/Iw/wDbrH/6HJWTX0OE/gRPIxH8VhRRRXSYGXquhwazeWbX4jmtbYs5tZIwyyORgE544GeMd/aq8fhi3stSe70Ux6WJbYwyR21ugBbOVfHTI56g5zWnqN/BpmnT3t26pFChZizBfwye56Vnad4ls7jRrbUNQutPs1uf9WBeKyn23EL83YgdKytT5vP+kaXnby/pkVv4duY76a/lv4UvXg8lZbS0ES4yDuZSzbjx+AJxiq7eD/tMlxcXl3Et3KI9sllbeSqsjblYqWbc2e5PTit271Kx0+FZb+9t7WNzhXmlVFb6Emqdx4k0231SwsWuoTJfKWiYSrtIHTvzknAx1waHGmrJ/wBXBSmxsukXt/pt7aatqSTLdQmECC28pUyMbsFmJPP97HtUt1pT3/hyTTL+WGV5ITE0gg+TPZthY9OD16irYvrRoY5VuoTHK+yNxIMO2cYB7nI6VGmr6bJcPBHqFq80YLPGs6llAOCSM5GDxVWhaz6iTluuhQ07wppOmanJe21jaI+1Fi2WyqYsAgkEdznk8Vdj07y9dn1Hzc+dbxw+Xt6bWY5znvu9O1Pt9U0+7tXubS+tp4I/vyxTKyr9SDgU+zv7PUYfO0+6guos43wSB1z6ZFNKK0XqJuXUsUUUVRIVXvv+Qfcf9cm/kasVXvv+Qfcf9cm/kaHsNbnoN7/yLdx/16N/6BXBQ/8AHvH/ALo/lXe3v/It3H/Xo3/oFcFD/wAe8f8Auj+VeVl32vkd+M+ySUUUV6p54VgweENKIuJNUs7TUbq4meWS4ntlLfMeFGckADAHPat6sW+8RxWviSz0aI2z3Fwpd/NuRHsUEDAGCWY5yBxnB5qJqDa5vQuPN0IU8PXsGj22nQalBLDDGYSl3ZCVHXnbxuByBgZzg+lMbwvOmlRaZb38L2Sw+U0V3Zib6spDLtPPQ7gMDArZg1KxurqS2tr23mni/wBZFHKrMn1AORTV1fTXvWs01C1a6T70AnUuv1XOalxpta9fMfNNGZF4duNPm8zRtR8jfBHDMLmHzt+wYVx8y4bHc5HTirX9kTDV7PUFvd8kFu1vN5sQJmUkHPykBTlfQj2pNJ8SaZrEBltbqH/XtAqmVSWYE4xg9wCR6jmrk2p2FuJDcXtvEImCSF5VXYxGQDk8HHP0prk+L5/194nzXszN1nwrY61eNdXUNs8v2V7dGlt1kKknIfnrjsPc+tQX3hUvY21lpE9rp9nGd09sLPKXJwB8+11OOOR378cVrSazpcUMMsupWiR3H+pdp1Ak/wB055/CnXWqWFjJEl7fW1u8xxGssyoX+mTzS5Kev9eY1KegtjFdw2+y/mt5nB+U28BiUL6YLt/OrVJS1qZhRRRQBo+GP+RrT/rym/8AQ4q7iuH8Mf8AI1p/15Tf+hxV2F/Y2+p6fcWN7H5ttcxtFKm4jcrDBGRyOD2r5/G/x5f10PXwv8NXLFFZD+FtGfwr/wAI21nnSfJEP2fzX+4O27O78c0an4W0bWPDiaDqNn52moqKsHmuuAmNvzAhuMDvXJ1Og16zPEn/ACKmrf8AXlN/6Aa0lUIgVRgKMAVm+JP+RU1b/rym/wDQDT6i6HE0UUV9UeCFZmraLDrM1mL0rJa28hke2ePcsx2kLnPGBnOMHtWnRSaUlZjTa2MIeFrW11IXmieVpTGB4ZFtrdAHyQVbGMZBHcHrTbfw7cx301/LfwpevB5Ky2loIlxkHcylm3Hj8ATjFS33iOK18SWejRG2e4uFLv5tyI9iggYAwSzHOQOM4PNaEGp2N1dSW1te2808X+sijlVmT6gHIrNRpvbz/wCCW3NbmGfCAujeS391ELi5jRN9lb+QFKNuVyCzbmBA5J6cYq7Lo9/f6dd2mramkyXEDQgW9sIlXIxuOWYk/iB7VcTWdMkupLaPUbR54wS8SzqXQDqSM5FV9I8R6brNpHPa3MQ8yV4kQyqWLKTxgHqQN2PTmhRp7L+v6uPmnv8A1/WhM+nPc6C+n30kMzvCYmfyMIeMA7CT7cZqhpvg/SdM1EXcNlab0ijji22yqUZc5fP9456+1akup2ECSPPe20axP5cjPKoCNjO05PBxzio31rS41gaTUrNVuf8AUFp1Hm/7vPP4VTUHLme5K5rWQi6Yv9tXN+7h1uLdIDCU4wpY5z3zu6Y7VQPg/SE1qx1C0sLK2+y7yY4rRBvY42tkYwVxx9e1al1qdhYyRx3t7bWzynEazSqhc+wJ5qzRyxfy/wCHDml/X3C0UUVZAVZs/wDj21X/ALBk/wDIVWqzZ/8AHtqv/YMn/kK5cZ/Al/XU6MP/ABUe10UUV88euFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAVdS877C32bb5m9MbsYxvGevHTNWqo6x5X9mP55cJ5kedgBOd646++KvUAUdaMa6LdGdWeMRncqnBI+tXqq6mZRpc/2ZlSXZ8rMQAD+PFWqAKGsrC9jGLiRo0+12xBVc5YToVH4tgZ7Zq/VPVPM+xp5Nutw32iDKOu4AeamW+qjLA9iM1coAKKKKAK2pFBpN2ZVLIIH3BTgkbTnmrNQX3mDT7jyWCyeU2xmIABwcHnip6ACiiigCldkf2lYZh3ne+Hyf3fyHn8enNYvjLXTZpb6HZ6Out6hq6yRx2UkgjiMar+8aRiDhAGA6EkkDFbd1v/tCx2ziNd77oyxHmfIeMd8defSuY8XaJ4pn8WaRrvhEaQ8lla3FvLHqksqKRIYyCvlq3/PPvjrS0bs9hrTU5fT9GXS/Fuhf8Jl4WjscvHb2F3Y6vLcwGZAxiWZGCksBu2uQ3PHHFet15zdaP8Rte1DSE16LwvBY2Wow3krWVzcNKRGc4AaMD8yK9Gqr3jr3/wAtf67E/a+S/UqWok+33u+FUXemxwuDJ8g5J74PFcz481e706fRrGxu7TShql20Eup3MSuIAELAANgFmIwN3H14rpLMINS1ArLvYum5MEbPkHHv61yvxHWO7bRNMfQtH1R766ZI5daj3W9uRGWz0J3MBgDipZRH4V1+4tfFmoeHtU1yy1lLe2injv0SOFwzsy+U4Q7S3y7hgDg9K3/FP/HrY/8AX4n/AKC1c34b+HmnRXrTa14V8GIYtr276XY4dHBzklhx7YrqfElrc3Njbmzt2uHhuFkMaMqsRgjjcQO/rVPZE9WZdFRbNV/6AF9/39t//jtGzVf+gBff9/bf/wCO0hktFRbNV/6AF9/39t//AI7Rs1X/AKAF9/39t/8A47QByPin/kYf+3WP/wBDkrJrq9W8Mahq1+Lp9O1m3YRCPbDNZ7SASc/MxOfmPeqX/CDX3/Pvr/8A39sf8a9ahjKdOmou+hwVcPOc3JGDRW9/wg19/wA++v8A/f2x/wAaP+EGvv8An31//v7Y/wCNbfX6XmZfVKnkcrq6PLol8kas7tbyBVUZJO08AVx6WN3aX1vcXcmqW8MmmQQKbO0Wcqyg7kZTG5XORzgD17V61/wg19/z76//AN/bH/Gj/hBr7/n31/8A7+2P+NZSxdGTvr/V/PzLjh6iVtP6t/keYpFNpK6bbwWt3FbJaMi3D2QuLhSW4jOwEIMc9McDpiqXh60vLNdBmu7S7QRTXkcgaBiyGR8qWCjABx977v4V63/wg19/z76//wB/bH/Gj/hBr7/n31//AL+2P+NL61Rv1/pj9hUtbQ8mt1uk0vRtMOnXvn2upq07G3YIi72O7fjDDkdCcd8U6XRryXwnfxw20yTNq0k7IqASSRibJIDDByoBGQQcCvV/+EGvv+ffX/8Av7Y/40f8INff8++v/wDf2x/xoWKopdf6t/kP2FTy/q/+Z5Y2mW1/Hf3Mx1+73wxo0klrHA3yvuXamxCxU88qRjgZ6VqeHZLuS/vGmjaW3KR7byayNtNKwyCrKQC2OOcAc4Fd/wD8INff8++v/wDf2x/xo/4Qa+/599f/AO/tj/jVRxlGLvr/AF8yXh6jVtDBore/4Qa+/wCffX/+/tj/AI0f8INff8++v/8Af2x/xrT6/S8yPqlTyMGq99/yD7j/AK5N/I103/CDX3/Pvr//AH9sf8abJ4CvJYnje31/a6lTiax6H8aHj6Vuo/qlTyN29/5Fu4/69G/9Argof+PeP/dH8q9DltNSlsntW0G/8t4zGcTW+cEY/wCetc6vgS9VQot9fwBgfvbH/GuHCYiFG/N1OnEUpVLcphUVvf8ACDX3/Pvr/wD39sf8aP8AhBr7/n31/wD7+2P+Nd31+l5nL9UqeRg1zmr2V3c+KI2tUdQdLuIlmCnajsybQT2PGfwr0H/hBr7/AJ99f/7+2P8AjR/wg19/z76//wB/bH/GpljaMlZ3/pWKjhqkXfQ8q0LTju0iOc619osF/wBRJbRRwwtsII8wRrvU/wCyzZ4Jqu/2+6Oiq1jcwvb6gjz20GnmOG3GT0Yj5h7qccknFeu/8INff8++v/8Af2x/xo/4Qa+/599f/wC/tj/jWf1ulpvvf8vPyL9hU12/q/8AmeUR29zbW9rvs7pjYa1LPMEt3Y+W5kwy4Hzj5h93OKa9rc3+pXM50+6EMutW0qiWBlyixgb8EcDI79O9es/8INff8++v/wDf2x/xo/4Qa+/599f/AO/tj/jQsVRVt9LdO1v8gdGo77f1f/M8v1Kxmh8TahNcSalFbXcEccTWNok4YAENGwMbleTnnAOahuNOXTo4xZx6rHM1gkGy4shdx3CjO1JNgO0jPPKjnqcV6t/wg19/z76//wB/bH/Gj/hBr7/n31//AL+2P+NL61Rs9/6v5+YewqeRzOnmY6bbG6hSCYxL5kSHKo2OVHsKsVvf8INff8++v/8Af2x/xo/4Qa+/599f/wC/tj/jW/8AaFLszL6pU8jBore/4Qa+/wCffX/+/tj/AI0f8INff8++v/8Af2x/xo+v0vMPqlTyKXhj/ka0/wCvKb/0OKu4rA0rwtfaTqBvE03WLiTymiAmms8AEqT91hz8ora2ar/0AL7/AL+2/wD8drysRUVSo5RO+jBwgosloqLZqv8A0AL7/v7b/wDx2jZqv/QAvv8Av7b/APx2sDUlrM8Sf8ipq3/XlN/6AavbNV/6AF9/39t//jtQXtlqV/p9xZzaDfiO4iaJys1uCAwwcfvfegDgqK3v+EGvv+ffX/8Av7Y/40f8INff8++v/wDf2x/xr3Pr9LzPL+qVPIwaK3v+EGvv+ffX/wDv7Y/40f8ACDX3/Pvr/wD39sf8aPr9LzD6pU8jz7V7K7ufFEbWqOoOl3ESzBTtR2ZNoJ7HjP4Vk6NpzbdKjmGstc2CHFvLbRRwQtsIIMgjXep6fKzZ4Jr1f/hBr7/n31//AL+2P+NH/CDX3/Pvr/8A39sf8ayeLo3vr/Tb7+Zp7Cpa2h5HEL66vPDpazuovst1meCPTzDDbZRhgEjJAPGVO3ue1SWsNzZ2ukPNZXZGnalcG4CWzsQr+ZtZQBlx8y8rnrXrH/CDX3/Pvr//AH9sf8aP+EGvv+ffX/8Av7Y/40li6S7/ANW8/IboVH2/q/8AmeTLaXN7NPK+n3KxyeIIplWWEgmMIo34x0469qlv9Pmi1/V/tcmqJb3yoIvsNpHOsqBdpRiY22c56kD5s+teqf8ACDX3/Pvr/wD39sf8aP8AhBr7/n31/wD7+2P+NH1qja2v9JL9A9jUvfT+nc8tvrAWBT7KuprO9gkBWayF3FchQcJJtB2sM8nKg56mus08zHTbY3UKQTGJfMiQ5VGxyo9hXTf8INff8++v/wDf2x/xo/4Qa+/599f/AO/tj/jWkcbSTb11/ruQ8NUaWxg0Vvf8INff8++v/wDf2x/xo/4Qa+/599f/AO/tj/jVfX6XmT9UqeRg1Zs/+PbVf+wZP/IVq/8ACDX3/Pvr/wD39sf8amh8IalbWt7HBp+qyy3Vs9urXU1ptTcOvyMD1x6/SsMRjKdSm4K5rSw84TUmeoUUUV5R3hRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAFXUhK1iwgiWV96YRlDAjeMnB9Bz+FWqo6wsb6Y4mkMaeZH8wXdzvXHH14q9QBR1ryv7FuvtG/yvLO7Zjdj2zV6qupiU6XOLZFeXZ8qsAQT+PFWqAKOrhTZR77g24+1W/wA4BOT5yYXj+8fl/Hnir1UtWOLOP/R/tH+kwfJzx++T5uP7v3vwq7QAUUUUAVtS2f2Td+du8vyH37euNpzj3qzUF95h0+48lQ0nlNsVgCCcHA54qegAooooAo3jRjU9ODxlnaR9jbsbTsOeO/HFXqq3JlF9ZBJVRC7b0LAFxsOAB355q1QAUUUUAU7TP9o3/wC52fOmHwf3nyDn8OnHpXF/FbT3vINFnm0Vtd062u2a809rmKGKRTGQrMZGUEqxBA5/rXaWmf7Rv/32/wCdMJk/u/kHH49ePWuX+IPhaTxLcaH5mnRapYW104u7OUrjbJGUEuG4bYTux1645pO/Qa8yl8OF0JNQvl0XwbB4bkMSmR47i2kMwz0xDIxGPfFeg15Z8JvC1/4e1a6EuhQ6TDDp0FlcSIIsXtzGzZmXYScFSMlsEk9OK9Tq5W6Eq93cKKKKkYUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUViXvjTw3p2vxaJfa1Z2+qTOiR2kkoEjs5woA9yRitW8u7fT7Ge8vZkgtreNpZpXOFRFGWYn0ABNAE1FZWg+KND8TwzS+HtUttRjgYLI1vIGCE8gGo9e8X+HvC7QL4i1i0003AYxC5kCb9uM4z6ZH50AbNFNjkSWJJI2DI6hlYdwehp1ABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAU9U3/2e3lwCdt6fuypbPzjnA9Ov4VcqlqwB01t03kjfH8+CcfOvHHr0/GrtAFHWhG2i3QnZkjMZ3MoyQPpV6qmqCQ6XceTCs77PljZdwb2x3q3QBT1UMbNNlx9mP2mD5yxGR5qZXj+8Pl991XKo6w0a2MZlhM6/arcBQxXB85MNx6HDY74xV6gAooooArakEOk3YlYqhgfcVGSBtOeKs1Bf7jp1z5cYlfym2xkZDHB4x3zU9ABRRRQBRvPK/tPTvMD7/Mfy9pGAdhzn8KvVVufO+3WXlbfL3t5mcZxsOMZ5646VaoAKKKKAKVmUOpagFi2MHTc+Sd/yDn29K5X4kWetzJpF5odhcaktlcSSyWdvMsbNJ5bCFzuZQVVyCRn3wcV1lqZPt97vmV13psQNkx/IOCO2TzWR4u8W/8ACLxWiQaXc6re3rusFrbMikhELuxZyAAAPqeKmVrajW5zHw18H6h4W1u5WW0a2txptvFdTFwRf3uWeSYAEnjdt3HBOPavSa43wT4+k8YXUkUuhz6UBZw3kJnuI5DNFLnayhCeOD34PGK7KtJXvqSrf16BRRRUjCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAPPtf+DuieIfiTZeNLu/v47+zmglSGJkERMLBlBBUnkjnmuz1zSYde8PajpF07xwahay2sjxkblWRCpIzkZwavUUAcZ8OfhlpXwzsL200a7vLpLyVZHN2ykqQMcbVFV/iR8JdH+JsunPrN7fWp09ZBH9kZBu37c53Kf7oru6KAIraBbW0ht0JKxIqAnqQBipaKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooApasQNNbdD5w3x/JkjPzrzx6dfwq7VPVN/9nt5c4gben7wsVx844yPXp+NXKAKWsBG0e5EkvkqUOZACdvvgVdqnq2f7JucQ+edn+qwTu9uOauUAVNTMotE8idIG+0QZZ22gr5q7l+rDKgdyQKt1Q1kwixj+0q7J9rtgAhwd3nptP0DYJ9s1foAKKKKAK2pBTpd2Hfy1ML5fGdo2nnFWar6h/yDLrEfmnyX/d8/PweOPWrFABRRRQBRvPK/tPTvML7/ADH8vaBgnYc5/Cr1VbkSm+sikSugdt7lQSg2HBB7c8VaoAKKKKAKNmYzqeohFYOJE3knIJ2DGPTiuC+IPifT7TxboOn3eieIJbiG6aWK80+0dxjymLIuFYShlyrKMEDnjFeg2xl+3XvmMpj3r5YBGVGwZz+PrXI/EnSrrWH0S1K6g+lyXTx3q2EjowLRlYmYoQ2wOQT2HBPApMaMn4Z6fDb+Jr6Sy0/xCllDYx29tPrdsbbyIxIxFvGhRWYDOd5JPQdq6/xdFHcafaQzxrLE92gZHUFWG1jyD1rgvg9p0+n61dI1jqcJj0y3iv5NQMxAvVZhIsbSE7lIw2VyvTFegeKf+PWx/wCvxP8A0FquWlv66/167k63d/60Rgf2DpH/AECrH/wGT/Cj+wdI/wCgVY/+Ayf4VZW8tnvns1njNzHGJHhDfMqkkAkdgSD+RqepGZ/9g6R/0CrH/wABk/wo/sHSP+gVY/8AgMn+FSLqti+sPpSXKNfRwid4ByVQnAJ9Mmkm1jT7fWLbSprpFvrpGkhg5LMq/ePHQfWgDlfEEUVjrAt7HT9LSLyFc+ZYqxJLMOuR/dFZvmyf8+ej/wDguX/GtbxT/wAjD/26x/8AoclZNe3h8PSnSUpLU8ytWqRqNJh5sn/Pno//AILl/wAaPNk/589H/wDBcv8AjRRW/wBVo/ymXt6vchu78WVnNdXFpo6xQoZHP9nLwAMn+Kq+ka0mtaal7bWGlxqxZTHLpiB0YHBBAY4ORWX4see6Wy0iyjSaa8mDukkmxfKjIZskAkAnavQ9azFFza3ut6dqQGmpqNu13FJZSPNsIASQr8inP3TgDvWLo0VJ+7/w++5oqlRx3O182T/nz0f/AMFy/wCNHmyf8+ej/wDguX/GvP7H7E1rf2VnJZxxSWm6TUtGDmNSGxh4lJAYg84JJAOcVWkm26Xe2+hx2PlDyHnvNI3mApvw4MSnAbbyQrZK9SOKXs6H8v4/194+er/MehXuorp9jNd3VrpCQwIXdhpgOABk8A1Klw8kautno+GGR/xLV/xrz2XTrS40DXRp95p19A1if9G02yKQLIMkODvdd/sMHoT2rpLp7FfAjyWNrY3lnHa71hOPJYAZPQEdR6daao0rN8my/wA/8he0qXSUt3/l/mX7XX47uW3jisNMDXHmlN2mp92Ntpb73QkjH17U601+3vtUuLG1ttIkkt0V5GXTV2ckjAO7kgqc1yfhfQ3tNeiD6JpUJsrSON5opMvvYEl/9UMsQADyMA9Tmt6zjSLxtexxIqIunwBVUYAG+TtTjh6TtePf9QdWdnZ9vzRu+bJ/z56P/wCC5f8AGjzZP+fPR/8AwXL/AI0UVr9Vo/ymft6vcPNk/wCfPR//AAXL/jUdxcyw2ssostHJRCwB05ecD/eqSq99/wAg+4/65N/I0PC0bfCHt6nc7e50rS4tIlul0mw8xIDIAbZcZC59K46OeV41Y2Wj/MAf+Qcv+Nd3e/8AIt3H/Xo3/oFcFD/x7x/7o/lXnYGlCpzc6udmKqShblZJ5sn/AD56P/4Ll/xo82T/AJ89H/8ABcv+NFFej9Vo/wApx+3q9w82T/nz0f8A8Fy/41maj4kg0zUrKyubHTDJeNhWXTEKpyACx3cAkgDrzWnXC3NnqXiJ9aurKztZYZ/9Ftpprpo2TyifmVRGwP7zJ6jOBWdTD0Y2tEuFWo95HdebJ/z56P8A+C5f8aPNk/589H/8Fy/41weo3dlq9rpupajc2ETNbMr2eqofJLg4ba/AWQEEdCcdqrXrWU0VteXv2SKRrJRHpuqtJuXGcGCY87jxyAW+70qHToK/ur7ylOq/tHovmyf8+ej/APguX/Gq51NF1JLA2ukfaHiMwX+zB90EAnOfUiuEvmF5q7vrM1np0MlpC1pHqtu02wlTuCN5i4kB64y3Sr9jaWtv4s0WW9eO4nfTGSO7uIfLkmkDLjhvmDbSeOuM+9NUaTlbk62/P/ITqVEr8x1l3qX2NrdZLHSWa4lESBdNXqQT/e6YBNUH8Tx+XD9n0qwuZp5JEigi02Pe4jbazfM4UL7kjqKw/FtskuuwzPoumXqWtpLcSSXTYLKoxtP7tuBuyBnk56Y5S01S38G+GrGzu7fT7O9uRiKOOcLHI20ZkkYqoX369gM1PsaN3eNl/S/zK9pUsrP+v6sdTp2rLqdu0sNjpaMjmOSKTTFDRuOqkbuv0yPSrfmyf8+ej/8AguX/ABrH8NLZLprtZahBqLyTNJc3EEgdWlbk9CcDoAPTFbFbxwtGyvEydepd2YebJ/z56P8A+C5f8aPNk/589H/8Fy/40UU/qtH+UXt6vc0fD6w3uufY77TdLkja3eUGOxVCCrIPU/3zXUf2DpH/AECrH/wGT/CuX8Mf8jWn/XlN/wChxV3FeNioRhVcY7HpUJOVNNmf/YOkf9Aqx/8AAZP8KP7B0j/oFWP/AIDJ/hWhRXMbGf8A2DpH/QKsf/AZP8Kp6vpml2Gh315DpNgZLe3klQNbKQSqkjPHtW5WZ4k/5FTVv+vKb/0A0IDifNk/589H/wDBcv8AjR5sn/Pno/8A4Ll/xoor6L6rR/lPH9vV7h5sn/Pno/8A4Ll/xqOe7NvbyTTWmjrHGpdj/Zy8ADJ/iqSue8XvPcWNvpNnGsk+oyiMq77V8tfmfJAOBgYzg/eqZ4ajGN1EqNao3a5o6Lrqa5pwvLXT9MiXcUaObTEDoQehAY/X8a0PNk/589H/APBcv+NcS63VnqWq2eooumQ6raGWOSxlebY6AKxHyKc7SpwB/Caq2P2JrW/srOSzjiktN0mpaMHMakNjDxKSAxB5wSSAc4rL2VH+X+lf/LqX7Sp/MegebJ/z56P/AOC5f8agvdRXT7Ga7urXSEhgQu7DTAcADJ4Brz8zFdH1S30CKyOIonku9HLGFhvAceWDhX25PytkjqRxST6dZ3Oi6x/Z17pt7E2nuWtdNsSkRcYKuTvdfMH4N+VJ06NtIdP8yoyqX1l/Wn+Z6Mlw8kautno+GAI/4lq/41Qg1+O5lhSKw0wmaWWJSdMT/lmSGb73TIx+Iqtusf8AhDS1jb2V1ZrallhyBC4AyRwCMcHsa5nwzojw65Z+ZoWkxPaWaPJLG/zbpDnfjyh8w2EYzwD1q5Yekqiio/n/AF0M41ajg5OR2Npr9vfapcWNrbaRJJboryMumrs5JGAd3JBU5psviK2i8QQaMbbSWvJomlCrpqkIo7t83Ge1UbePyfGV+lqiJt06AImNqj55cDjoKxLa01ex8WaML+2snuZvtDzTx3TsZCQm44MYxgYAXngYyOtJ0KSt7vX9bFOpO797p+lzuvNk/wCfPR//AAXL/jR5sn/Pno//AILl/wAaKK2+q0f5TL29XuHmyf8APno//guX/GrFsEns9SE9lpwaOxllikgs1jZGUcEHJ9f0qvVmz/49tV/7Bk/8hXPicPShScorU1oVpyqJNntdFFFeKemcFq32KLVfEN9qYzDazB3c5OxFtomOAPxOBUYTRf7Nhv5fJgtplRkkmfYPnxtHJ6kkDHqareMrJdR0vxpatbrcmQNtiZA25vskRXg98gEe9cpdeH38QfDHSY/COoMkCpa/6JYfZhCzLKrSOSyEhlIYkBhyvIPIItW/l+N/8h9vn+h2epjRtHsJLzUFjihjHuSx9FHUk9gOajtJNFu/D0OtGJYLKW2W63THaUQruy3OBgVNLp5t9DlS9u5tSlijkdLi6SIOCVI42IqjgkcDODXKWem6zqfgTwjFpkNjc2MdlBLd291dPB5xEa7Fysb5XPJGBnAHTNHR/L8b/wCQuq+f4WOg8OXei+KdBg1fS7ZvstwXEZlXaxCsVzjPQ4yPatT+yrH/AJ91/M1zXwue7bwUv2yCGHF5c7PKmMmf375zlVxg5HfIAPGcDsact9BI88fVLxppvKg05I1mdFDW8jHCsQMnzBnp6Uf2lf8A/PLTP/AaT/47VZP9ZP8A9fE3/oxqfXuwwlFwTcfzPKliKqk1cm/tK/8A+eWmf+A0n/x2j+0r/wD55aZ/4DSf/Hahoq/qdD+X8xfWKvcp654uk0CxW6vYLF1Z9gWK0ct0JJwZhwACT7CtBdUvXQMiaYVYZBFtJyP+/tchqIvNW8WOlnZ213b6dbmKRbm4MSmSUDOMI2SE+n3qywVl8MW1hrF3b2c2lXZt5Vu0MttLhfkEh+X5SpUgnHIH0rD2FG7938/Q19rUstT0X+0r/wD55aZ/4DSf/HaP7Sv/APnlpn/gNJ/8drzdjYXOnWpmawsbW3uJUjiu901hdKcHcrNgL1+XsOcA0yeY3UOkfa1tNP0kRzIRfo9zalw4CklmT5SMlC3HoOlL2ND+X8Q9pV/m/A9CufEE9pNbRTrpqvdSeVEBaSnc20t/z144Bpb/AF670+wmu5odPZIl3FUtZCx9h+961wMljYx22gz3lxbX1pHqT7biS3MccSMjYVd5b5NwGDkjpjoK2vGVn9vXTbZdOsr15bpQpumxtA+cgHY3BCEH+tV9XpcrfL26vrb/ADD2tTmS5vy6X/yOgXxLMXvA66ZEtmQJpJLdwoJUN187sCKNL8Sz6vZm6tItP8sSPH89pIpJVip4831Fc54O01EsrqeTSNOtI7qdnX7Md24ByFBGxcAYBHXrnAq94V/5Blz/ANf91/6OaqhhaMrXjuvPyFKtUSdns/LzOh/tK/8A+eWmf+A0n/x2j+0r/wD55aZ/4DSf/HahorT6nQ/l/Mz+sVe5N/aV/wD88tM/8BpP/jtT6df3E2r2dvdQWDRTyFG8uF1YfKxyCZD6elUqm07/AJGDTP8Ar4/9kesq2FoxpyaXQunXqOaTZ0PiNItL02OWztoPNkmWMGVWZQDk5wGHp61z39pX/wDzy0z/AMBpP/jtdF4z/wCQTbf9fSfyauXrDB0KdSm3NdTXEVZwnaLJv7Sv/wDnlpn/AIDSf/HaP7Sv/wDnlpn/AIDSf/Hahors+p0P5fzOf6xV7k39pX//ADy0z/wGk/8AjtZlr4we71260mOGxE9suSzWjhH6ZCnzuSNwzx3p+pX0em6Xc3s33LeNpCB3wOlcSttq+i2Omare2Nqn2W4aa7ljumaRlmP7zK+WAACwP3jjZWUsNRUkuX13LjWqtb/l/X/Dno39pX//ADy0z/wGk/8AjtH9pX//ADy0z/wGk/8AjteeagbKz8RTXAez1K7a6Rkt33xXsfIGI2HLIOuMBSM5NFubOx8RIYGtNTu3vSDjdFfQhic7+7ouR1AXAHXipVGg7e7+LKdSra/N+B6H/aV//wA8tM/8BpP/AI7Va18QT3r3C266axtpTDJm0lGGABx/rfcV51p0HnXERvdVs7XWFvD5i/YXa9P7z7u4SZMZHGdu0A10Xha3srTV9eihit4bn7aWZEVVfyyqlTgc7clsds5p06FGbXu/j6f5ilVqRT978DcvvFc1g9ws0Fixt7b7Q+y0c/LkgD/XdTg4+lQXXjS5trieOPTre5Fqoa5kgs2KwAjPOZwScc4UMa4jUdJ+16veFPD+jN9o1CO3R3kwSV/eMf8AVH7wyGP4c4rZ1bxBYbj4cF7Y6Y4iVLtzOqJboR9yPONzY4HAAHJ7Cs/YUnG9rdt/X/Iv2tRPf8v67nZQ6xd3ECTQrpbxyKGRhbSYIPIP+tp/9pX/APzy0z/wGk/+O1SsRbrp9utiyPbLGoiZG3KVxxg9xip66/qdD+X8zn+sVe5N/aV//wA8tM/8BpP/AI7R/aV//wA8tM/8BpP/AI7UNFL6nQ/l/Mf1ir3Ok8NeXqtncte21v5sFx5WYVZVYbFbOCx/vY69q2f7Ksf+fdfzNY3gr/jz1L/r9/8AaMVdLXh1oqNSSXc9Sm24Jsp/2VY/8+6/maP7Ksf+fdfzNXKKyLKf9lWP/Puv5msbxL5elWds1lbW/mz3HlZmVmVRsZs4DD+7jr3rpa5rxr/x56b/ANfv/tGWtaMVKpFPuRUbUG0YP9pX/wDzy0z/AMBpP/jtH9pX/wDzy0z/AMBpP/jtQ0V7n1Oh/L+Z5f1ir3Jv7Sv/APnlpn/gNJ/8drP1rxXPoWnG8ureylUMFEcNo5difQGYdBk/QVarltVN5qXi2KGxtLe6i02EvKtxOYl8yQELyEbJChuMfxVnUwtFLSOvzKhXqN6v8jrIdXvJ4UliTTGSRQysLaTkHp/y1p/9pX//ADy0z/wGk/8AjtebhVGgR6XrFzbWL6VeeWUuUaa1lQqTGJD8vy7WABOOV/CnsbG60W385rHT7S3u3VFuA09hcjHUE4AXn5ewIOAaj2NDpH8fT/Mv2lX+b8D0X+0r/wD55aZ/4DSf/HarXPiCe0mtop101XupPKiAtJTubaW/568cA155PN9ostJEiWlho6yzo7XIe5s3bjY3LJ8h+bbn5QcYHSntYWESaNcXVzZ31lHqjbZjbeVDCjRnCIXZvk3gEEHGcY6Cj2NFvSPbr6f5j9pUS1l36dr/AOR6Dfa9dafYT3k8WneVBGZG22shJAGeP3vWol8TTGS5WSPTYVtUV5ZJLZwq7hnGfO6gdfqKwvGdob7T7S2Sws7x5rpEX7U2NnOTj5G6gEE9h69Kr+ENMRRf3L6NptpHPcMF+zNu+423bjy1+UFNwOepzgU/q1JzcVH8/wCupPtqnJzN/l/Xc6bS/Es+r2ZurSLT/KEjx/PaSKSVYqePN9RUWmeLjq91fQWK6dIbGURSv9lk2lsdAfN5x0rn7CHULjwvfQ6Q8SXEmoXC7pXKYXzm3YIVsHGcHBpPCkVzbeINat5LO1toozCu2C4ZwpEYAAyi5GOc8c9j1oWHotxXLuvPsOVWoubXZ+Xc7L+0r/8A55aZ/wCA0n/x2j+0r/8A55aZ/wCA0n/x2oaK2+p0P5fzMvrFXuTf2lf/APPLTP8AwGk/+O10XhG4a51DS53jiim+3ywOYAyq6i3kboSe+O/auXro/BX/AB9ab/2FZv8A0mkrjxlCnTppwXU6MPVnOdpM9D1ho00xzNGZE8yP5Q23neuOfrzV6qupGVbFjBKsT70w7MFAG8ZGT6jj8atV5R3lPVs/2Tc4m8g7P9bkjb78c1cqlrBRdHuTJF5yhDmMEjd7ZFXaAKuo/aPsqfZGVZPPhyWxjZ5q7xz327vf05q1VDWfI+wx/avM2fa7bHl4zu89NvXtuxn2zV+gAooooAr6h/yDLrEnlHyX/ec/JweePSrFVtSKjS7sunmKIXymcbhtPGas0AFFFFAFG8WM6npxeQq6yPsXbncdhzz245q9VO63/wBoWO2ASLvfdIVJ8v5Dzntnpz61coAKKKKAKNn5X9p6j5e/f5ieZuxjOwYx+FcRBon/AAnviXxBJrmp6kllpd99htLKyvJLaMBY0ZnYxkFmJfucADpXd2wl+3XvmIoj3r5ZAGWGwZz+PrXL6n8J/Bur6tdanf6ZO13dyeZO8eo3MQdsAZ2pIAOAOgo66/1t/wAEOhH4US50HxtqnhY6jd6hp8VlDfWrXspllh3u6NHvPLL8gIzkjJ5ra8U/8etj/wBfif8AoLVB4a8BeHPCF3cXWgWMlvPcoscryXc05ZQSQP3jtjknpU/in/j1sf8Ar8T/ANBam3ewdXb+v6Z5xDpx/wCFua3cWk0kV0dIt2iaaWWSIMXlXJiDgMBgccYOSMEk0eDdN8SWviPxBLqNzaC2l1Le4/s2SM3H7iMB42MxCrkY6Nyp59Ovk06yl1CK+ls7d7yFSsVw0SmSMHqFbGQD7VZoTt91vxuD1f8AXaxyNvZW+mfEaYafbAFtIaQqp5kczE8k9ST3JrlbH+04viT4fv8AWvD19Dq14LoXErSWzKqbVCohWUnYg5OQCSSQCTivU/s8H2r7T5Mfn7PL83aN23OdueuM84oe1t5bmK4lgjeeEMIpWQFo93XB6jOBnFKOlvK/5v8AzB6t+f8Akkcj4p/5GH/t1j/9DkrJrW8U/wDIw/8AbrH/AOhyVk19FhP4ETx8R/FYUUUV0mAhIAJJwB1JqG0vbW/h86xuYbmLJXfDIHXI6jIqr4gjaTw5qCpK8J+zud6AE/dPHII56Vymn3+rbLXS7J9QlS30+CbzLZbXeS4OAfM2jYAMcDPqfXKVTlla39a/5GiheN7/ANaf5neUhYAgEgE9BnrXMvf6lNBZW95cXFjqDQGSa2sbeOV8g43Fm3Iq+3qevBrI0y7vdZ1fw5eXd3NHKBdoyqkYDbGA54PJAGcHHHGKHVSdrB7N2ud9RXJwa3fP4e0a5e4/fXV+sMrbF+dd7DGMegHSqj6zq9vpN5qX2yS4f+0XsobcRRBUHnbA2SASwGerAdM+tHtV2/rT/NB7N/18/wDI7eiuUSfxSkF4kFtdPiJWgkvjbCXdu+ZR5TbT8vIyBz1Jq/4fv5Jp7m1uru6lnjCP5N7bLFNGD6lAEcZHVRxg8mqjNN2sJxsrm5RRRVkBVe+/5B9x/wBcm/kasVXvv+Qfcf8AXJv5Gh7DW56De/8AIt3H/Xo3/oFcFD/x7x/7o/lXe3v/ACLdx/16N/6BXBQ/8e8f+6P5V5WXfa+R34z7JJRRRXqnnhVe4vrS0kijurqGB5m2RLJIFMjeig9T9KsVyOtebbeOIL1LmbEGl3EohVUw21k+Xlc4PfnPAwRzmJz5df67lxjzOx11Fcjouo+ILmSwupYLqS2uVD3LTfZlhjUrkNHscv1xw2ePSoX1u+S60y4ttQurq3vbsQMWtY4oNpzymRv7cHLA9emKXtFpox8j7nZhg2dpBwcHFLXCaHPeadZ2KR3ssi3WtzwyiRE5UGUnoo6lQT+mBxVq/wBY1M6jeWtte+SF1W3tUYRoxRHjBYDI5OSTzmkqqaTtvb8bf5jdNptX/q7X6HY0VyFzqGtNrF1ptm9/P9hijPm26Wu6V2BOZBIVG3oPkA6Hmlu73xEiwzXy3NhELVXdrKGO4CS87xIpyxXpjZ780e1Vr2YvZvuddRUNpcJdWUNxDKs0cqB1kQYDgjOQKmrYzCiiikBo+GP+RrT/AK8pv/Q4q7C/muLfT7iaytftdzHGzRW/mCPzWA4XceBk8ZPSuP8ADH/I1p/15Tf+hxV3FfP43+PL+uh6+F/hoyH1HWR4V+3roW7VvJD/ANl/bEHz/wBzzcbfxo1PUdZtfDiXmnaF9u1IqhbTvtiR7Scbh5hG07efriteiuTqdAiklAWG0kcjOcVm+JP+RU1b/rym/wDQDWnWZ4k/5FTVv+vKb/0A0+ouhxNFFFfVHghRRRQBXuL60tJIo7q6hgeZtkSySBTI3ooPU/SrFcjrXm23jiC9S5mxBpdxKIVVMNtZPl5XOD35zwMEc5TSNS1+c2N5NDdNa3Cb7l5/swhiUrkNHscvjOOGzx6Vkqutmv62/Q0cNLnX0gYMMqQRnHFcYusX41DSWi1C7ubfUZTEzyWscUWChIaMEb+2RksMde1Q+HLi7sLHQ4EvJJI7vULlJVkRPujzTgYUY5XP19uKFUTdrf1p/mDptK9/61/yO6orjrrWNTklu4be98kjWo7NHEaMUjaNSQMjBOSeTmmzahrsup39jYtfznT9iLJClr+9cqGzKHKnHIHyBeh59D2ytez/AKV/1H7N3tc7OiuU1C919GimuVurC3+yq7tZQx3Ajl5LiRTlio4xs9+a6W0uEurKG4hlWaOVA6yIMBwRnIFWpXbXYhxskyaiiiqJCrNn/wAe2q/9gyf+QqtVmz/49tV/7Bk/8hXLjP4Ev66nRh/4qPa6KKK+ePXOOb/kYdb/AOvtP/SeGpqwPGUl1DpfjSXT5xb3EYZ0lKk7cWkR4wQQcdDng4PPSsWPWtWsdB0e3OsaPFdSWCzM01rNNLLgDhYFkDYwRl9xyf4eaV9/L/g/5D7ed/wt/mdzRXn9j4z13XY/DaaVHYWcurW9zLM91DJIIzCyr8qhkPOTwTkZ9sFbnxd4ibQ31+xTTf7OF8LRbeWKTzSonELS7w+OTkhNvAx8x6VVne39b2F0ud/RXndz8QdUe6v5tM0+a4t7G7e2FnHpF3NJdbG2sy3CDy0Oc4UhunJGeNTwzNqk3j7xSLu9jltI3txFB5LBo8xhgAS5A4JyMcnnjpSWuoPR2MNP9ZP/ANfE3/oxqfTE/wBZP/18Tf8AoxqfX1FP4F6Hhz+JhRRRVkBRRWX4keePwvqT2kohmS2kZXwTtwpPGCOfQ54NTOXJFy7FRjzSSNSiuXXUtQtdL0+E6jp0c7WglJkgkkeTjtEr5xyPm3HJ7VDY+JdU1eTS47JLS2N3aS3ErTRO+0o4XAG5Tznv0pOolLlGotq511Fck3iHWH0+11WFbL7DdXccKwtG/mLG0gTfu3YJPXGBjPU4pkvi6+ea6lsrSSaG2uGhFumn3EjzbThiJVGxTnOBg9OSM8L2sf6+X+aH7OR2FFc7okl9J4r10XFyklujxBIvLYFMoCMEsQOOoxyeeOldFVRlzK5DVnYKKKKoQVNp3/IwaZ/18f8Asj1DU2nf8jBpn/Xx/wCyPWOI/hS9Ga0v4kfU6Lxn/wAgm2/6+k/k1cvXUeM/+QTbf9fSfyauXrly/wDhP1N8X8a9Aooor0DjCiiigAornNQkvh4802OG6jS2NnM7RMjHdhkzzuxnkYOOOeueK8fia8TVtOiknsLuG+maEraRSEQkKTnzSdr/AHSMbVJ/A1n7RdTTkdtDq6K5G08Q6xPp1nO4sfN1K4NvbIInxEQWy7Hd8w2oTgY54zTrjxNqOl/2lHqENvPLbSQRQm3jcCQy8ZK5Y8egyeO+aXtY2uHs3ex1lFcLrGuatP4f1WEb4zFbiRL0afPbKfmwybZCCG6chjx2rtbcSi2jFw6PLtG9kQqpPqAScfmaqM1LYmUeUloooqyQooooA6TwV/x56l/1+/8AtGKtLWL7VLJrIaTpH9pia4WO5P2lYfs8Z6yfN97H90cms3wV/wAeepf9fv8A7Rirpa+ar/xpep7dL+GvQzdUvdTtbywj03SPt8M82y6m+0rF9lT+/tPL/Qc0l5fapBrtha2mkfarCcP9qvvtKp9lIHy/uzy+Txx0rTorFGgVzXjX/jz03/r9/wDaMtdLXNeNf+PPTf8Ar9/9oy1tQ/ix9UZ1f4cvQ5uiiivpTxAooooAKKK5DTr/AFG1TXZri/s8JqPlRtcq6rH8qf7RzwRhRjJ788S5Wdv66f5lKN0dfRXFy+LdRhh1KFBa3NxaPbCObyJIo5BK+37rMTxzyCR+VaQ1DW5dQn0+CSwM9nEss8rQOFkLFtqKu/K8LyxLfSp9rG1yvZtb/wBf1c6KiuOj8XXtxa6aixLFc3Vp9pmkSymuVT5sBQkfPPPJIxjvSRanquo+INCkbNiri4EtvLbyLv2EDdhipwRyuRxk9aXtYt2Qeza3OyooorUzCiiigAro/BX/AB9ab/2FZv8A0mkrnK6PwV/x9ab/ANhWb/0mkrz8w/hL1OzCfG/Q9D1jyv7Mfzw5TzI87CAc71x198VeqrqXnfYW+zbfM3pjdjGN4z146Zq1XiHplTVDINLuPJmWB9nyyM20L757VbqjrRjXRbozqzxiM7lU4JH1q9QBV1ETm1T7LEsr+fDlWAIC+au489wuSPcVaqhrKwvYxi4kaNPtdsQVXOWE6FR+LYGe2av0AFFFFAEF/uGnXPlyCJ/KbbITgKcHnPbFT1W1IoNJuzKpZBA+4KcEjac81ZoAKKKKAKV2B/aVhmbYd74TB/efIePw681dqldkf2lYZh3ne+Hyf3fyHn8enNXaACiiigCjZiManqJRmLmRN4IwAdgxj14q9VS1En2+93wqi702OFwZPkHJPfB4q3QAVh+Kf+PWx/6/E/8AQWrcqve2FtqNuIb2ISxhgwGSMEd8igDnKK1f+EX0j/n1b/v/ACf/ABVH/CL6R/z6t/3/AJP/AIqgDKorV/4RfSP+fVv+/wDJ/wDFUf8ACL6R/wA+rf8Af+T/AOKoA848VRzf28HS1upUNsg3Q2zyDIZ+MqDzyKx9tx/0D9Q/8AZv/ia9f/4RfSP+fVv+/wDJ/wDFUf8ACL6R/wA+rf8Af+T/AOKrtp4ypTgopLQ5p4aE5czZ5BtuP+gfqH/gDN/8TRtuP+gfqH/gDN/8TXr/APwi+kf8+rf9/wCT/wCKo/4RfSP+fVv+/wDJ/wDFVp/aFXsiPqkO7PHZ7eS4t5IZtO1Bo5FKOv2GYZBGD/DVCfw5azrAG0nVIzBGIkeGC6ibYP4SygEj2JNe4/8ACL6R/wA+rf8Af+T/AOKo/wCEX0j/AJ9W/wC/8n/xVS8dN7xQ1hYrZs8Ol8O2kogH9j6lF9nTy4/ItrmLCf3TsAyPY5FNg8NWdtDbRW2j6lClrK0sIjtrldhY5PIHQ5+709q9z/4RfSP+fVv+/wDJ/wDFUf8ACL6R/wA+rf8Af+T/AOKo+vT35UP6rHuzwpfCtgtxHMujaluim8+Nfs9zsjf+8qY2r15wOe9WRotuLGezOjXzW88jSSRtZTsGZjuJ5Xjnn27V7Z/wi+kf8+rf9/5P/iqP+EX0j/n1b/v/ACf/ABVCxskrKKD6tF9WeGJ4as0hljbSNTlEwCu89vcyuQDkAOwLDB5GDwasWOjx6c0j22maoZJMB5Zra5ldgOgLOCcDJ4z3r2v/AIRfSP8An1b/AL/yf/FUf8IvpH/Pq3/f+T/4qhY6a1UUL6rF9WeQbbj/AKB+of8AgDN/8TRtuP8AoH6h/wCAM3/xNev/APCL6R/z6t/3/k/+Ko/4RfSP+fVv+/8AJ/8AFVX9oVeyF9Uh3Z5BtuP+gfqH/gDN/wDE1FdRXUlnMiafqBZo2AH2Gbkkf7teyf8ACL6R/wA+rf8Af+T/AOKo/wCEX0j/AJ9W/wC/8n/xVH9oVOyD6pDuznbuKR9BnhVSZGtmUKOpO3GK4CJLlYUU6fqAIUA/6DN/8TXsP/CL6R/z6t/3/k/+Ko/4RfSP+fVv+/8AJ/8AFVzUMRKjflW5tVoxq2ueQbbj/oH6h/4Azf8AxNG24/6B+of+AM3/AMTXr/8Awi+kf8+rf9/5P/iqP+EX0j/n1b/v/J/8VXT/AGhV7Ix+qQ7s8g23H/QP1D/wBm/+Jqu+nCS/S9fS9Qa4SJolc2U/CMQSMbcdQK9n/wCEX0j/AJ9W/wC/8n/xVH/CL6R/z6t/3/k/+Kpf2hU7IPqkO7PDbXw3Z2dwktvo+pAxkmJGtrlkiz12IQVTqfugVEvhPT0kiddG1T9xKJYVMN0ViYHPyKeFHsAAa93/AOEX0j/n1b/v/J/8VR/wi+kf8+rf9/5P/iqX12X8qK+rR7s8Nk8N2kkcqNpGp7ZZvPIW3uRsk5+ZMD5DyeVxnNJD4ZsrfPlaNqWTOtwS1tcsTIowGJI5OPz717n/AMIvpH/Pq3/f+T/4qj/hF9I/59W/7/yf/FUfXZL7KD6tHuzxC90GC/uRcXGl6oJtuwvDb3MRdf7rFANw9jkVHN4as5in/En1KIJGIttvb3MKsg6KwQAMB6HNe5/8IvpH/Pq3/f8Ak/8AiqP+EX0j/n1b/v8Ayf8AxVH12X8qD6tHuzx6KCSGJIodNv0jRQqqthMAoHQD5adtuP8AoH6h/wCAM3/xNev/APCL6R/z6t/3/k/+Ko/4RfSP+fVv+/8AJ/8AFVX9oVeyJ+pw7s8g23H/AED9Q/8AAGb/AOJo23H/AED9Q/8AAGb/AOJr1/8A4RfSP+fVv+/8n/xVH/CL6R/z6t/3/k/+Ko/tCr2QfVId2ea+FoLg+JBM9pdRRJaSIXmt3jGS8ZA+YDP3T+VdrWr/AMIvpH/Pq3/f+T/4qj/hF9I/59W/7/yf/FVxVajqzc2dNOChHlRlUVq/8IvpH/Pq3/f+T/4qj/hF9I/59W/7/wAn/wAVWZZlVn69DJceHNSggQvLJaSoiL1YlCAK6X/hF9I/59W/7/yf/FUf8IvpH/Pq3/f+T/4qgDyDbcf9A/UP/AGb/wCJo23H/QP1D/wBm/8Aia9f/wCEX0j/AJ9W/wC/8n/xVH/CL6R/z6t/3/k/+Kr0f7Qq9kcf1SHdnkG24/6B+of+AM3/AMTRtuP+gfqH/gDN/wDE16//AMIvpH/Pq3/f+T/4qj/hF9I/59W/7/yf/FUf2hV7IPqkO7PGH04SX6Xr6XqDXCRNErmyn4RiCRjbjqBVS28NWdpMkkOjal+7z5aPbXLxxZ4OxCCqcE/dAr3L/hF9I/59W/7/AMn/AMVR/wAIvpH/AD6t/wB/5P8A4qp+vT/lQ/qse7PCYfClhBNbyR6Pqm62ffBuhumEJ6YUHhRz90ce1PfwzZvCYjpGqBTMZ12wXSmNznJQgZTOT93HU17n/wAIvpH/AD6t/wB/5P8A4qj/AIRfSP8An1b/AL/yf/FUvrsv5UP6tHuzw2Dw1ZWybYdG1FR9pF0c2twcygYDZI68fjTrzw/bX90bm40rVBKyhHaK3uY/MUdA4UAOPZs17h/wi+kf8+rf9/5P/iqP+EX0j/n1b/v/ACf/ABVP69O1uVB9Vje92eG3Hhy0uXVn0jU4ysYixBb3MQZB0VggAYD0ORWjFBJDEkUOm36RooVVWwmAUDoB8tew/wDCL6R/z6t/3/k/+Ko/4RfSP+fVv+/8n/xVNY+otkhfVYPqzyDbcf8AQP1D/wAAZv8A4mjbcf8AQP1D/wAAZv8A4mvX/wDhF9I/59W/7/yf/FUf8IvpH/Pq3/f+T/4qn/aFXshfVId2eQbbj/oH6h/4Azf/ABNXLKOb7Dq0j2tzCi6bMu6a3eMEkDgbgM9DXqf/AAi+kf8APq3/AH/k/wDiqRvCujOpV7MspGCrTOQR6EZ5rOpjJ1IODS1LhhowlzJmvRRRXEdJxF9aRX+p+IrS5XdDcTiKRfVWtogf0NYkXgiK3a3kt9b1WKeK2FpJMjQhp4VJKo/7vAxkgMoVueuea7y68NaZd3k11Klws0xDSGK8mjDEKFBwrAdFA6dqj/4RLSvS+/8ABjcf/F0AcXo3gnTtDOlfZJ7txpUU8VuJXVsrKwZt3ygnGBg/nmuY1Hwpquo3MunW1lqGn2b6kt2w+2QPZcTCRnUACbc2D8hAQMT9a9b/AOES0r0vv/Bjcf8AxdH/AAiWlel9/wCDG4/+Lp3d0/67h0scafB8CX1zNZapqdjb3k3n3FnazKkckh6tu2+YhOBnYyg/iavW+hxWniG71aC5uEa8jRZ7f5DG7KMK/wB3duA4+9j2rpP+ES0r0vv/AAY3H/xdH/CJaV6X3/gxuP8A4ukB5GY7mKe4VtP1DP2iU5WxlYEeY2CCFwR70fv/APoH6j/4ATf/ABFeuf8ACJaV6X3/AIMbj/4uj/hEtK9L7/wY3H/xdehHH1IpKyOR4WDd7nkf7/8A6B+o/wDgBN/8RR+//wCgfqP/AIATf/EV65/wiWlel9/4Mbj/AOLo/wCES0r0vv8AwY3H/wAXT/tCp2QvqkO7PI/3/wD0D9R/8AJv/iKiubd7u1ltp9O1Fo5kKOPsE3IIwf4K9h/4RLSvS+/8GNx/8XR/wiWlel9/4Mbj/wCLpPH1GrNIFhIJ3TZ4QnhiSJonifxAkqQi3eRbFg0sYJKq37nAxkjK4PvT9L8MRaRJaNbWmst9kgkgjEllKfldwxz+75II4/rXun/CJaV6X3/gxuP/AIuj/hEtK9L7/wAGNx/8XSWOmnflX4lfVotWuz5vfQNVubqC1TR9es7WG+W62tAXtl2vuOzCeYSfRgAMn0Fbv/COsl1NJbLr1rDcSebNbQWkqo7nqc+XvXPfawr3P/hEtK9L7/wY3H/xdH/CJaV6X3/gxuP/AIupjjJxWy/r/hhvDxfU8Wh0t7fV7i/itdWU3KqJYfsEmxiowG/1e7OOOuParv7/AP6B+o/+AE3/AMRXrn/CJaV6X3/gxuP/AIuj/hEtK9L7/wAGNx/8XVrH1FokiPqkH1Z5H+//AOgfqP8A4ATf/EUfv/8AoH6j/wCAE3/xFeuf8IlpXpff+DG4/wDi6P8AhEtK9L7/AMGNx/8AF0/7QqdkH1SHdnkf7/8A6B+o/wDgBN/8RVnSo7h/EGm/6DfIqzlmeSzlRVGxuSzKAK9T/wCES0r0vv8AwY3H/wAXR/wiWlel9/4Mbj/4upnjqk4uLS1KjhYRkmmcd4xjlk0iEwwzTFLlGZYYmkYDB5woJ71yn7//AKB+o/8AgBN/8RXrn/CJaV6X3/gxuP8A4uj/AIRLSvS+/wDBjcf/ABdZ0cVOjHliiqlCNR3Z5H+//wCgfqP/AIATf/EUfv8A/oH6j/4ATf8AxFeuf8IlpXpff+DG4/8Ai6P+ES0r0vv/AAY3H/xdbf2hU7Iz+qQ7s8j/AH//AED9R/8AACb/AOIo/f8A/QP1H/wAm/8AiK9c/wCES0r0vv8AwY3H/wAXR/wiWlel9/4Mbj/4uj+0KnZB9Uh3Z4pe6Qb6+t7t7TVo5IFeP93YS4kRsbkbMZ4OB0wfeqEPhIRPY7v7flTT5A9rE9k+2MAEbeIgSMHGTk+9e9f8IlpXpff+DG4/+Lo/4RLSvS+/8GNx/wDF1P12V78q/Er6tG1rs8EvdDltdAt7Kx0jWbr7NN5kbR2zpNGck7kLJtJGcYOAQearab4aurpdTfWdO1rF9JEyia0kMyGMcMSibVOeQF4GK+hP+ES0r0vv/Bjcf/F0f8IlpXpff+DG4/8Ai6n65K97If1eNtzw1vD8k9jdW1//AG/ei5j8tmms5BtX/ZVYwoPvjNaVvFcw26Rva6pOyrgySafLub3O1APyAr2D/hEtK9L7/wAGNx/8XR/wiWlel9/4Mbj/AOLq1j5rZIl4WL6s8j/f/wDQP1H/AMAJv/iKP3//AED9R/8AACb/AOIr1z/hEtK9L7/wY3H/AMXR/wAIlpXpff8AgxuP/i6f9oVOyF9Uh3Z5H+//AOgfqP8A4ATf/EUfv/8AoH6j/wCAE3/xFeuf8IlpXpff+DG4/wDi6P8AhEtK9L7/AMGNx/8AF0f2hU7IPqkO7OL8GRSx2F800E0HmXe5BNE0ZI8qMZwwBxkH8q6OtH/hEtK9L7/wY3H/AMXR/wAIlpXpff8AgxuP/i64Jyc5OT6nVGPKkjOorR/4RLSvS+/8GNx/8XR/wiWlel9/4Mbj/wCLqSjOrnPGcUslhYtDBNP5d3ucQxNIQPKkGcKCcZI/Ou0/4RLSvS+/8GNx/wDF0f8ACJaV6X3/AIMbj/4uqhJwkpLoTKPNFpnkf7//AKB+o/8AgBN/8RR+/wD+gfqP/gBN/wDEV65/wiWlel9/4Mbj/wCLo/4RLSvS+/8ABjcf/F13/wBoVOyOX6pDuzyP9/8A9A/Uf/ACb/4ij9//ANA/Uf8AwAm/+Ir1z/hEtK9L7/wY3H/xdH/CJaV6X3/gxuP/AIuj+0KnZB9Uh3Z5H+//AOgfqP8A4ATf/EVjT+GVmluXWHW4jPOtyBHYyfupQAN65iPJA5ByPavdf+ES0r0vv/Bjcf8AxdH/AAiWlel9/wCDG4/+LpPHTe6X4jWFiurPBD4QSSW6luE16aW6MJld7OTJMTblIAiAHuAMewqPW7HU31Zp7HQ9e3SwCJ57KArvGT8siyJwBnhky3J6V7//AMIlpXpff+DG4/8Ai6P+ES0r0vv/AAY3H/xdQ8ZJq1kUsPG97ngVh4VmttK0xDHrFtf2Vt5H2m1sZMkHBZcPGQRkdxV06ExkspRHrgms3ZlmazldnDfeVi0Z4PoMY7Yr3D/hEtK9L7/wY3H/AMXR/wAIlpXpff8AgxuP/i6r69P+VfiL6tHuzyP9/wD9A/Uf/ACb/wCIo/f/APQP1H/wAm/+Ir1z/hEtK9L7/wAGNx/8XR/wiWlel9/4Mbj/AOLqv7QqdkT9Uh3Z5H+//wCgfqP/AIATf/EUfv8A/oH6j/4ATf8AxFeuf8IlpXpff+DG4/8Ai6P+ES0r0vv/AAY3H/xdH9oVOyD6pDuzyP8Af/8AQP1H/wAAJv8A4iun8GRyx3OlmaGaEvqkzKs0TRsR9mk5wwB7V2v/AAiWlel9/wCDG4/+LqW08N6bZ3kd1Clw0sRJjM13LKFJBBIDMRnBIz71jWxU60eWSNKdCNN3RPrHlf2Y/nlwnmR52AE53rjr74q9VXUhK1iwgiWV96YRlDAjeMnB9Bz+FWq5DoKupmUaXP8AZmVJdnysxAAP48VaqjrXlf2LdfaN/leWd2zG7Htmr1AFPVPM+xp5Nutw32iDKOu4AeamW+qjLA9iM1cqjq4U2Ue+4NuPtVv84BOT5yYXj+8fl/Hnir1ABRRRQBFdxyy2c0dvII5XjZY3ZchWI4OO/NRAah58e57byto8wBG3E45wc9M1aooAoquq+S++Wz8zI2ERttA75+b6U7bqXmQ/vLXZgeaPLbJOedvPHHrmrlFAGPdLq3262xd2cfztsTa43jaeCM8469ulWyup7YsSWm7/AJa/u2557c8cetJeNGNT04PGWdpH2NuxtOw5478cVeoAp7dS8yb95a7MHyh5bZBzxu5549MU1l1XyU2S2fmZO8mNtpHbHzfWr1FAGRapeDWNTMdxbshZcIY2yjeWNuTnBHc4qzt1X7OMS2fnbuT5bbduPTd1zmnWmf7Rv/3Oz50w+D+8+Qc/h049KuUAVcah9pzvtvIx02Nuzj1zjr+lR7dV+znMtn527g+W23bj03dc4q9RQBVA1Dz49z23lbR5gCNuJxzg56ZqNV1XyX3y2fmZGwiNtoHfPzfSr1FAFPbqXmQ/vLXZgeaPLbJOedvPHHrmkC6ntlzJabv+WX7tuOe/PPHpV2igCkV1PbFiS03f8tf3bc89ueOPWl26l5k37y12YPlDy2yDnjdzzx6Yq5RQBRZdV8lNktn5mTvJjbaR2x831qQjUPPk2vbeVtPlgo24HHGTnpmrVFAFHbqv2cYls/O3cny227cem7rnNSY1D7TnfbeRjpsbdnHrnHX9KtUUAUduq/ZzmWz87dwfLbbtx6buucVIBqHnx7ntvK2jzAEbcTjnBz0zVqigCiq6r5L75bPzMjYRG20Dvn5vpTtupeZD+8tdmB5o8tsk55288ceuauUUAUgup7ZcyWm7/ll+7bjnvzzx6UFdT2xYktN3/LX923PPbnjj1q7RQBT26l5k37y12YPlDy2yDnjdzzx6YprLqvkpsls/Myd5MbbSO2Pm+tXqKAKpGoefJte28rafLBRtwOOMnPTNR7dV+zjEtn527k+W23bj03dc5q9RQBVxqH2nO+28jHTY27OPXOOv6VHt1X7Ocy2fnbuD5bbduPTd1zir1FAFUDUPPj3PbeVtHmAI24nHODnpmo1XVfJffLZ+ZkbCI22gd8/N9KvUUAU9upeZD+8tdmB5o8tsk55288ceuaQLqe2XMlpu/wCWX7tuOe/PPHpV2igCkV1PbFiS03f8tf3bc89ueOPWl26l5k37y12YPlDy2yDnjdzzx6Yq5RQBRZdV8lNktn5mTvJjbaR2x831qQjUPPk2vbeVtPlgo24HHGTnpmrVFAFHbqv2cYls/O3cny227cem7rnNSY1D7TnfbeRjpsbdnHrnHX9KtUUAUduq/ZzmWz87dwfLbbtx6buucVIBqHnx7ntvK2jzAEbcTjnBz0zVqigCiq6r5L75bPzMjYRG20Dvn5vpTtupeZD+8tdmB5o8tsk55288ceuauUUAUgup7ZcyWm7/AJZfu2457888elBXU9sWJLTd/wAtf3bc89ueOPWrtFAFPbqXmTfvLXZg+UPLbIOeN3PPHpimsuq+SmyWz8zJ3kxttI7Y+b61eooAqkah58m17bytp8sFG3A44yc9M1Ht1X7OMS2fnbuT5bbduPTd1zmr1FAFXGofac77byMdNjbs49c46/pUe3Vfs5zLZ+du4Pltt249N3XOKvUUAVQNQ8+Pc9t5W0eYAjbicc4OemajVdV8l98tn5mRsIjbaB3z830q9RQBT26l5kP7y12YHmjy2yTnnbzxx65pAup7ZcyWm7/ll+7bjnvzzx6VdooApFdT2xYktN3/AC1/dtzz25449aXbqXmTfvLXZg+UPLbIOeN3PPHpirlFAFFl1XyU2S2fmZO8mNtpHbHzfWpCNQ8+Ta9t5W0+WCjbgccZOematUUAUduq/ZxiWz87dyfLbbtx6buuc1JjUPtOd9t5GOmxt2ceucdf0q1RQBR26r9nOZbPzt3B8ttu3Hpu65xUgGoefHue28raPMARtxOOcHPTNWqKAKKrqvkvvls/MyNhEbbQO+fm+lO26l5kP7y12YHmjy2yTnnbzxx65q5RQBSC6ntlzJabv+WX7tuOe/PPHpQV1PbFiS03f8tf3bc89ueOPWrtFAFPbqXmTfvLXZg+UPLbIOeN3PPHpimsuq+SmyWz8zJ3kxttI7Y+b61eooAqkah58m17bytp8sFG3A44yc9M1Ht1X7OMS2fnbuT5bbduPTd1zmr1FAFXGofac77byMdNjbs49c46/pUe3Vfs5zLZ+du4Pltt249N3XOKvUUAVQNQ8+Pc9t5W0eYAjbicc4OemajVdV8l98tn5mRsIjbaB3z830q9RQBT26l5kP7y12YHmjy2yTnnbzxx65pAup7ZcyWm7/ll+7bjnvzzx6VdooApFdT2xYktN3/LX923PPbnjj1pdupeZN+8tdmD5Q8tsg543c88emKuUUAUWXVfJTZLZ+Zk7yY22kdsfN9akI1Dz5Nr23lbT5YKNuBxxk56Zq1RQBR26r9nGJbPzt3J8ttu3Hpu65zUmNQ+05323kY6bG3Zx65x1/SrVFAFHbqv2c5ls/O3cHy227cem7rnFSAah58e57byto8wBG3E45wc9M1aooAoquq+S++Wz8zI2ERttA75+b6U7bqXmQ/vLXZgeaPLbJOedvPHHrmrlFAFILqe2XMlpu/5Zfu2457888elBXU9sWJLTd/y1/dtzz25449au0UAU9upeZN+8tdmD5Q8tsg543c88emKay6r5KbJbPzMneTG20jtj5vrV6igCqRqHnybXtvK2nywUbcDjjJz0zUe3Vfs4xLZ+du5Pltt249N3XOavUUAVcah9pzvtvIx02Nuzj1zjr+lR7dV+znMtn527g+W23bj03dc4q9RQBVA1Dz49z23lbR5gCNuJxzg56ZqNV1XyX3y2fmZGwiNtoHfPzfSr1FAFPbqXmQ/vLXZgeaPLbJOedvPHHrmkC6ntlzJabv+WX7tuOe/PPHpV2igCkV1PbFiS03f8tf3bc89ueOPWl26l5k37y12YPlDy2yDnjdzzx6Yq5RQBRZdV8lNktn5mTvJjbaR2x831qQjUPPk2vbeVtPlgo24HHGTnpmrVFAFHbqv2cYls/O3cny227cem7rnNSY1D7TnfbeRjpsbdnHrnHX9KtUUAUduq/ZzmWz87dwfLbbtx6buucVIBqHnx7ntvK2jzAEbcTjnBz0zVqigCiq6r5L75bPzMjYRG20Dvn5vpTtupeZD+8tdmB5o8tsk55288ceuauUUAUgup7ZcyWm7/ll+7bjnvzzx6UFdT2xYktN3/LX923PPbnjj1q7RQBT26l5k37y12YPlDy2yDnjdzzx6YprLqvkpsls/Myd5MbbSO2Pm+tXqKAKpGoefJte28rafLBRtwOOMnPTNR7dV+zjEtn527k+W23bj03dc5q9RQBVxqH2nO+28jHTY27OPXOOv6VHt1X7Ocy2fnbuD5bbduPTd1zir1FAFUDUPPj3PbeVtHmAI24nHODnpmo1XVfJffLZ+ZkbCI22gd8/N9KvUUAU9upeZD+8tdmB5o8tsk55288ceuaQLqe2XMlpu/wCWX7tuOe/PPHpV2igCkV1PbFiS03f8tf3bc89ueOPWl26l5k37y12YPlDy2yDnjdzzx6Yq5RQBRZdV8lNktn5mTvJjbaR2x831qQjUPPk2vbeVtPlgo24HHGTnpmrVFAFHbqv2cYls/O3cny227cem7rnNSY1D7TnfbeRjpsbdnHrnHX9KtUUAZd1Z6je2PkTzWqsZkYskbY2KwYjluvHWtSiigCrqYlOlzi2RXl2fKrAEE/jxVqqOtCNtFuhOzJGYzuZRkgfSr1AFLVjizj/0f7R/pMHyc8fvk+bj+7978Ku1T1UMbNNlx9mP2mD5yxGR5qZXj+8Pl991XKACiiigAooooAKKKKAKtyZRfWQSVUQu29CwBcbDgAd+eatVRvPK/tPTvMD7/Mfy9pGAdhzn8KvUAFFFFAFO0z/aN/8Avt/zphMn938g4/Hrx61cqlZlDqWoBYtjB03Pknf8g59vSrtABVPVNSTSrH7TJDLN86oscO3cxY4H3iB37mrlYvir/kExf9fcP/oYoAh/4So/9APUv++rf/47R/wlR/6Aepf99W//AMdqpRQBb/4So/8AQD1L/vq3/wDjtH/CVH/oB6l/31b/APx2qlFAFv8A4So/9APUv++rf/47R/wlR/6Aepf99W//AMdrzvxfaW9z4iU3NvFMVtUx5iBsfO/rWJ/Zen/8+Nt/35X/AAr0KWCdSCnfc5KmJUJONj1//hKj/wBAPUv++rf/AOO0f8JUf+gHqX/fVv8A/Ha8g/svT/8Anxtv+/K/4Uf2Xp//AD423/flf8K0/s6X8xH1xdj1/wD4So/9APUv++rf/wCO0f8ACVH/AKAepf8AfVv/APHa8S1mPTtL0e5vP7PtmaNPkUQr8znhR07kgVmaI0Ftpt/H4igtnudMJaeb7Mg3xldwbCj0yOP7tQ8DZ2cilirq6R9Af8JUf+gHqX/fVv8A/HaP+EqP/QD1L/vq3/8AjteDjUNISzlurzQ5LSCOISrJLaxlZFJwMFCwycjg4PtUgu9KjtLi41DRHsFgCkie1Q793A2lCwYk8YBz04o+o/3vwD61/dPdP+EqP/QD1L/vq3/+O0f8JUf+gHqX/fVv/wDHa8AvtSsYNPvyug/Z763tWuI7e4gi+dR/ECrEEA9RkH25FaFjDbT6DBeyaIhmeEOYBFFvY47c7eevLULA3v723l6/5A8Vboe3/wDCVH/oB6l/31b/APx2j/hKj/0A9S/76t//AI7XznpF5Jf6hYRS6EyJcNPOzmG22mENhOjE4G5c9yfUVp2kJk8VS2t3p9lFb/ZfNjhWJWYHeRlmx3HYdKUcFzWtLfy8rjlieW91t/wx7z/wlR/6Aepf99W//wAdo/4So/8AQD1L/vq3/wDjteQf2Xp//Pjbf9+V/wAKP7L0/wD58bb/AL8r/hWn9nS/mI+uLsev/wDCVH/oB6l/31b/APx2j/hKj/0A9S/76t//AI7XkH9l6f8A8+Nt/wB+V/wpk2l6eIJP9Btvun/livp9KP7Ol/MH1xdj2L/hKj/0A9S/76t//jtH/CVH/oB6l/31b/8Ax2sSyJ/4Ru3Pf7Iv/oFeaWWl2BsLcmxtiTEuT5K+n0rlw+HdZuztY3q1lStdbns3/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47XkH9l6f/wA+Nt/35X/Cj+y9P/58bb/vyv8AhXV/Z0v5jD64ux6//wAJUf8AoB6l/wB9W/8A8do/4So/9APUv++rf/47XkH9l6f/AM+Nt/35X/CuX8SzCz1BV0uytvLsIhd3yi3Q74ywG3kcHAc8f3RUywLjvIqOK5nZI+h/+EqP/QD1L/vq3/8AjtH/AAlR/wCgHqX/AH1b/wDx2vDLy50q2uIre30cX08kXniO2gjyI/73zFR+AJPtTJr3So5kht9Dkup2gFw8MdqivHGTgFlcqc5B4GT7UfUbfa/APrV/snu3/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47Xhdxc6XHOsFtojXk/lLM8UNtGrRIc4LbyuDweOvHSqa6hZXXiHTLaw0iGaxvbaSYzCCMYIKjuQRjJyMZyRjvhfUdbc34B9a/un0B/wlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O14L4gMdhNZW9jofnPcTqPMhigwQMsyfOw5KqecY981W060udc09JUs4dNjkuZC0jwQtIIgTsCABl9MlvfGeDS+pO9k/wAPT/Mf1lWu0fQf/CVH/oB6l/31b/8Ax2j/AISo/wDQD1L/AL6t/wD47XhHh6G3vI72O4tra5W1umgjufs6KZgAMkgDGQSQcYHHQVsf2Xp//Pjbf9+V/wAKtZe2k+Yl4tJ2sev/APCVH/oB6l/31b//AB2j/hKj/wBAPUv++rf/AOO15B/Zen/8+Nt/35X/AAo/svT/APnxtv8Avyv+FP8As6X8wvri7Hr/APwlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O15r4Ss7a28Xobe3iiJsZgTGgXP7yH0rv64K1N0puDOqnP2keYt/wDCVH/oB6l/31b/APx2j/hKj/0A9S/76t//AI7VSisjQt/8JUf+gHqX/fVv/wDHaP8AhKj/ANAPUv8Avq3/APjtVKqatxot7j/n3k/9BNAGt/wlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O147DpenmCP/Qbb7o/5Yr6fSn/2Xp//AD423/flf8K9T+zpfzHD9cXY9f8A+EqP/QD1L/vq3/8AjtH/AAlR/wCgHqX/AH1b/wDx2vIP7L0//nxtv+/K/wCFVNUh0zS9Kub2WwtysEZfAhXJIHA6dzxSeXtK7kNYtN2SPav+EqP/AEA9S/76t/8A47R/wlR/6Aepf99W/wD8dr570KSK2sr+LxLb2pubBBPLKLZBmJl3ZAUdiGX/AIDV1L/SRZz3V5oj2dvDD5/mzWqFXT1BQsM+xwfap+o/3h/Wv7p7v/wlR/6Aepf99W//AMdo/wCEqP8A0A9S/wC+rf8A+O14Ul3pS2txcX2iPYRwRiUtPbIQynptKFgT/s5z7VU1DU7G20+8K6B5F5HavcQQ3EEX71V6sMNjAyCQSGx2pSwXKruX4DWJu7KJ9Af8JUf+gHqX/fVv/wDHaP8AhKj/ANAPUv8Avq3/APjteI2EFtcaFBeyaLH5zwhzAIotzHHbnbz2ye/OKwLC+kvdRsojoLRx3F1M5LRW2PJTK44YnhipJ6+macsC4y5XL8BLFJq9j6N/4So/9APUv++rf/47R/wlR/6Aepf99W//AMdrwa0hMnimW1u9Psorf7L5scKxKzA7yMs2O4HQdKpT3Mc/i/TrWxsrMaY0ssM0nkqTLIqEkDjopHJHfjsaSwV0nzb+XnYPrOr02/4c+hv+EqP/AEA9S/76t/8A47R/wlR/6Aepf99W/wD8dryD+y9P/wCfG2/78r/hR/Zen/8APjbf9+V/wrT+zpfzE/XF2PX/APhKj/0A9S/76t//AI7Udx4yS1tpbifRdSSKJC7tmA4UDJPEteSf2Xp//Pjbf9+V/wAKnsLW3tptS+zQRw7tHu93loFz/q/SsquCdKDnfYuniVOSjY94ooorgOsyLrxJbW19Pai2up3t2CyGKMEAlQ2Mkjsw/Oov+Epg/wCgff8A/ftf/iqwdWv10u68SX0kbyLbzCQpGjMzYtojgBQT+QNZVp4ytpPAMHii6triOJ4ElkhWJlYFsAhQ+3IyeD0I5GaN/wCu4bHZ/wDCUwf9A+//AO/a/wDxVH/CUwf9A+//AO/a/wDxVYep30tjZeba2ct5Mx2pFHgcnuxPCqO5/LJwK5W78a3cXw80XWmksLK81R7dC9yCYId5y7EblOFUMfvDpR/wPxH0uejf8JTB/wBA+/8A+/a//FUf8JTB/wBA+/8A+/a//FVzfh+7nvtN+0zatpurJI37q402Exxken+tkyc9wfwrVoEX/wDhKYP+gff/APftf/iqP+Epg/6B9/8A9+1/+Krx6WytZ7+/kmtoZHN9cZZ4wScTOOppP7Msf+fK3/79L/hXpRwEpRUubc45YtRbVj2L/hKYP+gff/8Aftf/AIqj/hKYP+gff/8Aftf/AIqvHf7Msf8Anyt/+/S/4Uf2ZY/8+Vv/AN+l/wAKr+zpfzE/XF2PYv8AhKYP+gff/wDftf8A4qj/AISmD/oH3/8A37X/AOKrwTxI8em2cSaVptlNf3DlYY5YMqQoLMTjHYfmRUyXuhHR7TUJLaFY7tVMSJbb3YkZ2hFBYkc8AdjUfUdWubYr61ony7nuv/CUwf8AQPv/APv2v/xVH/CUwf8AQPv/APv2v/xVeFG/8PCzjuRDC6yuUREs2aQsOq+WF3ZHcY47043vh4WUV0Ird45mKxrHal5HYdVCBS2Rg5GMjHNP6j/fQfWv7rPc/wDhKYP+gff/APftf/iqP+Epg/6B9/8A9+1/+Kr5/v8AWNFtYbCW2sI7mO7uhbkpaMTHwc5UISGGMbTg9fQ1b1+bT9F0ia6Gmxyy7D5SpZtIC+OA21TgE45OKn6lo3zLQf1nVLl3Pdf+Epg/6B9//wB+1/8AiqP+Epg/6B9//wB+1/8Aiq8A07ULC/uL8JpW9LRUAX+znjZnIyw+cD1GB179OataGIdSt7l7zS7SCSG5khCJGpwFPGT3NNYFt2UhPFJK7R7r/wAJTB/0D7//AL9r/wDFUf8ACUwf9A+//wC/a/8AxVeO/wBmWP8Az5W//fpf8KP7Msf+fK3/AO/S/wCFX/Z0v5ifri7HsX/CUwf9A+//AO/a/wDxVH/CUwf9A+//AO/a/wDxVeO/2ZY/8+Vv/wB+l/wq5o9lawa9YPBbQxv5wG5IwD0PcVE8BKMXLm2KjilKSVj1b/hKYP8AoH3/AP37X/4qj/hKYP8AoH3/AP37X/4quQ8aIsnhwo6hla4hBUjII3iuM/syx/58rf8A79L/AIVlQwrrR5k7GlWuqbs0exf8JTB/0D7/AP79r/8AFUf8JTB/0D7/AP79r/8AFV47/Zlj/wA+Vv8A9+l/wo/syx/58rf/AL9L/hXR/Z0v5jH64ux7F/wlMH/QPv8A/v2v/wAVR/wlMH/QPv8A/v2v/wAVXjv9mWP/AD5W/wD36X/Cuct9Rt5fFbWL6bZjT3d7eCcQjc0yAMwJ6YwWA46qal4Bppc25SxSavY+hf8AhKYP+gff/wDftf8A4qj/AISmD/oH3/8A37X/AOKrwuS98PRX32V4oBJ5giLC0JjVz0Uybdob2JzyKPtvh77cLTyoPMMnlBvsh8vf/c8zbt3e2c0fUf76D61/dPdP+Epg/wCgff8A/ftf/iqP+Epg/wCgff8A/ftf/iq8Le98PJefZmigDeZ5Rf7KfLD5xsMm3YGz2zmo9JutO1TVNSs10uOM2MwjDNakBxtBzkqB1PT0wehpLA3dlJB9a0vynvH/AAlMH/QPv/8Av2v/AMVR/wAJTB/0D7//AL9r/wDFV876vrum6dfXUEWl7xbWzu3/ABLJTmTOEGQuNpw3PQ9jVqaK7lja4tNP022tooFlL3doczNjLKFyDGB6sCeenFT9T0upf1r/AJD+sd0e/f8ACUwf9A+//wC/a/8AxVH/AAlMH/QPv/8Av2v/AMVXiWlx2Gp6TaXy6dBGLiFZdhiXK5GcdKt/2ZY/8+Vv/wB+l/wrX+zpfzEfXF2PYv8AhKYP+gff/wDftf8A4qj/AISmD/oH3/8A37X/AOKrx3+zLH/nyt/+/S/4Uf2ZY/8APlb/APfpf8KP7Ol/MH1xdj2L/hKYP+gff/8Aftf/AIqj/hKYP+gff/8Aftf/AIquI8CxpFYajHEioi33yqowB+5iPSuorzqkeSTj2OyMuaKZf/4SmD/oH3//AH7X/wCKo/4SmD/oH3//AH7X/wCKqhRUFF//AISmD/oH3/8A37X/AOKo/wCEpg/6B9//AN+1/wDiqoVznjZFk0WCORQ6NdIGVhkHg9qqEeeSj3JlLli2dl/wlMH/AED7/wD79r/8VR/wlMH/AED7/wD79r/8VXjv9mWP/Plb/wDfpf8ACj+zLH/nyt/+/S/4V6X9nS/mOP64ux7F/wAJTB/0D7//AL9r/wDFUf8ACUwf9A+//wC/a/8AxVeO/wBmWP8Az5W//fpf8KyfEJh02xjXTtPspb65lEVvHLCCpPUk4wcBQTUywDiruQ44pN2SPeP+Epg/6B9//wB+1/8AiqP+Epg/6B9//wB+1/8Aiq8Itb7RZNBtNTuLWCJbhQBGLfc5fugUAkkEHgDPFO+3+Hfsa3PkwlWkMQjFoxlLjkr5e3fnHOMZxR9Rt9tD+tf3We6/8JTB/wBA+/8A+/a//FUf8JTB/wBA+/8A+/a//FV4Z9s8PfYUuxFbmORzGii1JkZx1UR7d24YPGM1Rv8AWtFto7J7bT0uFuboW7hbNi0Rxk5UJkN0+UgHn2pPA2+2v+HD61f7J9A/8JTB/wBA+/8A+/a//FUf8JTB/wBA+/8A+/a//FV4brcmmaNpM15Jp6OVQlBHZmTLYJGdqnA9zge9Z+m6lZ3t1dR/2TuFtbxsyjTZI2aQ53KN4A7rgHnknpR9Rs+Xm1D61pflPoL/AISmD/oH3/8A37X/AOKo/wCEpg/6B9//AN+1/wDiq8J0U29/aXUt9ptnbNBcyRFVRSFVfU+tVdBvrfWdX1KFtJtobW3ETWzmIbpVbd859AccD0+tCwLdve3D60tdNj6A/wCEpg/6B9//AN+1/wDiqP8AhKYP+gff/wDftf8A4qvHf7Msf+fK3/79L/hR/Zlj/wA+Vv8A9+l/wq/7Ol/MT9cXY9i/4SmD/oH3/wD37X/4qn2/iW2uL+3tDa3cL3LlI2kjAUkIz4yCeymvGv7Msf8Anyt/+/S/4V0/gRFi1CzjiUIiaxIFVRgD/QWPA+pNYV8I6MeZu5rSxCqSsker0UUVxHSVNUEh0u48mFZ32fLGy7g3tjvVuqWsBG0e5EkvkqUOZACdvvgVdoAo6w0a2MZlhM6/arcBQxXB85MNx6HDY74xV6qmpmUWieROkDfaIMs7bQV81dy/VhlQO5IFW6ACiiigAooooAKKKKAKtz5326y8rb5e9vMzjONhxjPPXHSrVUbzyv7T07zC+/zH8vaBgnYc5/Cr1ABRRRQBUtTJ9vvd8yuu9NiBsmP5BwR2yeat1RszGdT1EIrBxIm8k5BOwYx6cVeoAKxfFX/IJi/6+4f/AEMVtVi+Kv8AkExf9fcP/oYoA5T/AISvS/8AhMj4aNxGL/7MJ9hlXJyfuAZzuwN2MdMGr9nqdnqFxeQ2k3mSWU3kXC7SNj7Q2ORzwwORkc1z15ZsfiSXuIbv7JfaR9lWe3WUBXEhYgyJ/qzg8EkexzTfDPghNC1vVL+S5u5PtF551sv9p3Mg2eUq/vFZtrtkNy27jHPAw46rXz/O35BLRu3l+X+Zo2HiiHUfFE2k21tL5cVt54unG1ZCH2FVBGSAQeenpmoNN1/WNXv7n7BpVidNt717VrmXUHWRthw7CMQkdcgDeM46ipfs0/8AwsY3PkyeR/ZXl+btO3d5uduemcc4rlm0e1BSDRPDUuk+If7QEr3Yt3cCP7Rukb7UVwysuTs3Z+bG3ilHW1/61/yB9fl+X+Zq+Kf+Rh/7dY//AEOSsmtbxT/yMP8A26x/+hyVk19FhP4ETx8R/FYUUUV0mBka5o7601lBJJsso5vNuAkjI77R8gBXkfNg5yOlVV8OSWOsteaXMSk9q0M6Xs8s+4jlD8zEkDLAjI4NbtxOltbSTzHEcSF2IGcADJqDTb7+0rCO7FtNbpKAyLNt3FSMg/KxHP51nyRcvP8ApF80kvI5+08OX0M0rxW+nWMJgaI2UTvLbzsSDuZCFC9+AD97knGKhbwfcXdjeW8gtdPgmjjEVnBK88AdGDBiGVQAcYKgdPWuxqtPfQW11bW0rES3TMsSgE5IUsfpwKTpQ6jVSXQwrbw2Ta3sbaVo2mG4tXgDWEeXJYYyX2rgf7OD9a0orK7k8MfYbkRxXQt/KBgncKCBgEOArDseBx71qUVXs42a76fn/mLnd7nN6R4Si0zUkn864aOC2jihX7bMQGGd2VLY2n5cL046CtQWEv8AwkZ1DcnlG1EO3J3bt+fyrQopqMVawnJvcKKKKokKjm/495P90/yqSo5v+PeT/dP8qAO9sv8AkW7f/r0X/wBArz6x/wCQfb/9cl/kK9Bsv+Rbt/8Ar0X/ANArz6x/5B9v/wBcl/kK8nL95fI9DGbRLFFFFeseeFc9F4WjubrULrVpZ2mvJT8tteTRoIgMIpClQeOTkHkmuhqhe6tHZ39pZC3muLi63FUi2/Kq43MSxAwMjpk+gqJRi/iLi5LYwh4cv10qztLi103UjaK0cUk8skUiLkhWWQKSDtwCAByOtPuPDt89jaQTRWOpywQmP7TcyyQzRsc5KyKGYjBAxweOtdTRU+yix+0ZyEnhGZbqK6nt9O1yY2scEx1JcHcufnVtjnnPI74HNaMei3NtqWlXVtHZKtrDJBNFEphRQ5U5RQG6FehxnPWtWzvoL9ZjbMWEMrQuSCMMpwRVmnGEfiXqJzlszD1nQG1XUoLnzpUWCCQKEu5Yv3hwFPyEY43ZPXnHNU5NB1TT9DtNO0GVAowLtri8l3EAAERsVfZn2Ax2GeR1FFDpx18wU3oUNJhuLaxW3ns7WzSIBYo7adpV2/Uov9av0UVoQFFFFAGh4Y/5G2P/AK8Zv/RkNdzXDeGP+Rtj/wCvGb/0ZDXY39r9u0+4tPPntvPjaPzrd9kkeRjcrdiOxr5/G/xn/XQ9fC/w0WKKyhoePC/9i/2pqR/cGH7ebj/Sv9/zMfe98Uy78PfavCq6J/a+qw7Y0j+3w3O26O0j5jJj7xxycc5Nch0LzNiqerf8gW9/695P/QTVmGPyYI497ybFC73OWbA6k9zVbVv+QLe/9e8n/oJp9RdDz6H/AI94/wDdH8qkqOH/AI94/wDdH8qkr6o8EKyde0mTWobW0MgS089ZLoB2R3VeQqkdPm285HArWopOKkrMabTujnW8NPaaxHfaTO3zwPBcLfXEtwGB5Q/Mxzhu2RwxqCx8O31tcSyQw6fp8Rt3i+yQO8sEzE5DNGQoUDngcnPXitq91aOzv7SyFvNcXF1uKpFt+VVxuYliBgZHTJ9BV+s1Tg9i+eVtTj/+EQuLq1vLd1tdMt57fyltbSV5oS+4MHKsqgYIxtA5ycmrFv4ZY291E2laLpxmtXg8yxizIxYY3btq7R/s4b611FVrK+g1CFpbVi6LI8RJBHzKxU9fcGl7GG39f1qP2kt/6/rQr2tpdf8ACPJZ3gSOcQeUfInfHAwCHAVh68DisnSPCCafe2s8s9w/2a1SNcX05zJnLnaWxtOF+Xoe4rp6K0cU5cz3IUmlZbGcdPkPiJr4svktaeRtyd2d2fyx71lP4NtodT0qbT3uIrezkd3je/nPBUgbQWIHJ56ZGevSumopezj2/q9w5n/X3BRRRVkhS2/+u1D/ALA93/7TpKW3/wBdqH/YHu//AGnXLjP4Ev66m+H/AIqPc6KKK+ePYOH1KBrq+8SW6EBpZQgJ7E20QrkV8DQeIvh1YaTr2lwWmpWlpHarPcwRTvFs2glCGPytt9QcHkDpXZ3MyW+sa9NMdscdyrscZwBbRE0mn30Gp6bbX9mxa3uYlliYjBKsMg47cGhdfl+tv1B9P67f8AjstJsNJsHtNIsLayhOW8q2hWJSxHXCgDNc3pulanpngnw7C+k215daasZntZQjSqQhGYXLBA4JHJOMZGRnNdjRTu/y/ANlYwPDthdpqmr6reW72Q1GSNktJHVmjCIFLNsJXcT6E8Ac1v0VTh1S1uNWutNictc2kccky7SAofdt57/dNIDgP+Xu+/6/rn/0c9Opv/L3ff8AX9c/+jnp1fT0f4cfRHh1PjfqFFFFaEHP3ekX2o+JzeNdz2EFrAI7Z4BExkZzlyQ6tgcKOgPWs6PQru0sZLK4srjUYbe8M1pPDdJDOquMkqQVGQxYEZUYPGeldjTJZUhheWQ4RFLMcdAKydOK1v8A1uaKb2OWOn6lLYRvqdld3jxXLPbmK6SK6t0IGAWUqjHrkbsYx1qtLomsT/2fe6gt1eNbPMpgiuxDcCJ8bcujIpYbeRkAg9TiuvtLqK9s4bq3JaKZBIhIxkEZFTUOiuoe0aOSfRZ4rG1nstLmjkj1JbuS3e886aQbdpJZ2wGwRxuI461q+INOutVtbWC2uJrdftCNMYvL4VTu53A9wOn45FbFFV7NJW/rS3+Qud3uYnh3S7zT47mW/vbiaW4ndzHKIsD5sK2UUclQvfHsKsaLaT2i332hNnm3ssqcg5UkYPFWoL6C4vLm1hYtLalRKMEbSwyPrxVmiMUrNdFb8v8AIUm3e/qFFFFWSFWNM/5Ddh/13H8jVerGmf8AIbsP+u4/kayrfwpejNKXxr1Oj8Zf8i//ANvEP/oYrlK6vxl/yL//AG8Q/wDoYrlK48v/AIb9Toxnxr0CiiivROMgvHnjsZ3tIvOnWNjHHkDc2OBk8da5N/CepWvh+3+y6ncz31rILtLdlhWN5s7mGdgbnLDlu/NdnRUSgpO7LjJrQ5O6sNTTUpZNIsrqxnmmWRpkuka2ccbi8bHIYjg7V64+alk07UodVZtIs7mxaS6Eksguke1kTPzExk7gxGfuqOcc966KW+ggvrezkYie5DGNcHkKATz+IqzUqnHo9h87scTb+G5YDNZXemXl4j3TyiZdUeO3Ks5fLIHyGGegQgkda3dKtLmz1zVzLA3kXUyzxThl2n92qlcZ3Zyp7Y962aKcaahawpTcr3ORv/DmqXuo3so1C8jjuLqFAq+RjyFw2eUJ4YsAOp75qbVV1ae7XT5dNvb3SkRfNmikgD3bf3WBdNq+uBz04HXob28h0+xmu7ptkMCF3YAnAAyeBUqOJI1dejAEUvZrZMrne9hlsxe2jJga3+Ufun25T2O0kfkTUtFFamQUUUUAdH4J/wCPTU/+v7/2jFXTVzPgn/j01P8A6/v/AGjFWnrWjf21DbR/2lqGn/Z7hZ91hP5TSbc/I5wcoc8ivm8R/Gl6nt0v4a9DTorN1TRv7TvLC4/tHULP7FN5vl2k/lpcf7Mgx8y+3FJqGjfb9W0++/tLULb7CzN9ntp9kVxkAYlXHzAY46VgaGnXPeM/+QTbf9fSfyauhrnvGf8AyCbb/r6T+TVrR/ix9UZ1fgfocvRRRX0x4gVgX+k32p+Jop/tU9ja2cB8iWDy2aSRzhuHVsAKAOn8RrfopSipblJ2OPXQbu0tryyntJ9TtluxdW0yXKQTqWBLlSu0Bg2f7oIbr2p507Up9PJ1Kyu7wJdeZaqLtI7u3TbjO9SFJySMbvunknpXW1Wtr6C7uLmGBiz2sgjlGCMMVDY9+CKz9nFaXL9o3qcpPous3UdndXy3VyLW5kZbdboRXAhZcAeZGyqXB/2sYOMmpX0WeO2gurPSriOSPUUupIZL3z5pVClclncqG56biMDrXX0Ueyje/p+Fv8he0exkeItPutU02O2tJ5YN80fmmPZ/q9w3Z3A9vTv6jIqLQNJvLCe9uL6+upnuJ2PlzCLBAwqv8iA52qOM49s1uVWnvoLa6traViJbpmWJQCckKWP04FVypS5ieZ8vKZEehSXmlanY3rzWyXV5JIGhZSzISCOoIwcYII6U3RtFvrDxNqN1cXlzPbywwpG0ohHmEbs5CKCNueOnXv26KihU4q3l/lYfO2mu4UUUVZAVveCP+Qra/wDYYf8A9ITWDW94I/5Ctr/2GH/9ITXBmH8Jep14T+J8j1aiiivDPUKerZ/sm5xD552f6rBO7245q5VPVs/2Tc4m8g7P9bkjb78c1coAoayYRYx/aVdk+12wAQ4O7z02n6BsE+2av1V1H7R9lT7IyrJ58OS2MbPNXeOe+3d7+nNWqACiiigAooooAKKKKAKtyJTfWRSJXQO29yoJQbDgg9ueKtVRvFjOp6cXkKusj7F253HYc89uOavUAFFFFAFW2Mv2698xlMe9fLAIyo2DOfx9atVRs/K/tPUfL37/ADE8zdjGdgxj8KvUAFYvir/kExf9fcP/AKGK2qx/E8UsukJ5EMkzJcROUjXc2A4JOKAMyioPOn/6B1//AOAzUedP/wBA6/8A/AZqAJ6Kg86f/oHX/wD4DNR50/8A0Dr/AP8AAZqAOT8U/wDIw/8AbrH/AOhyVk10et6Hf6pqK3Vst1bjyljKSaa8mcFjnIZf71Z//CJ6v/z2m/8ABRJ/8dr2MPiqUKSjJ6nnVqFSU3JGZRWn/wAInq//AD2m/wDBRJ/8do/4RPV/+e03/gok/wDjtb/XaPcy+q1exzfiCGK48OahHPGkqG3clXUEZCkjg+4zXEPawQtpcN4+mWWlnTI2hF/ab7dpjy5xvRQ+MHJyTzjvXrf/AAier/8APab/AMFEn/x2j/hE9X/57Tf+CiT/AOO1jLEUJO9+3T1/z/A0jRqxVrf1p/keZSvZ29jpNnq0ltfQskpS71NXigOG4Ty3J3NjGNxzgEg81m6RHaOugz6glu0dvqd1Cskse1Yx83lqA2SozjapPHFev/8ACJ6v/wA9pv8AwUSf/HaP+ET1f/ntN/4KJP8A47S+sUea9/w80P2NW1rfj6/5nk/262i0C2015VF6muKXt/40BuiwLDqAQRgng5p40/b4W1m/06AC/a/nR7hIy0gi8/5wMENjaCcAjv3r1X/hE9X/AOe03/gok/8AjtH/AAier/8APab/AMFEn/x2j6xRtZv8PT/Ir2dTout9/X/M8ttbOykgvzYa7Zx2z2hWUaPYOsatuGHba7Dd2I4JBOelafhGW1XULq3sINPeLykdrrS3It2bkbSmSqv34JJGM9q7/wD4RPV/+e03/gok/wDjtH/CJ6v/AM9pv/BRJ/8AHaccTQi00/w9SJUarTVjMorT/wCET1f/AJ7Tf+CiT/47R/wier/89pv/AAUSf/Ha2+u0e5n9Vq9jMqOb/j3k/wB0/wAq1/8AhE9X/wCe03/gok/+O0j+EdWdGXzphuGP+QRJ/wDHaX12j3D6rU7HT2X/ACLdv/16L/6BXn1j/wAg+3/65L/IV6HDHNDpsdp9hvyEiEW77K3OBjOK5SHwdq0MEcQnmIRQoP8AZEnOB/11rgwdeFJy5+p14ilKolymfRWn/wAInq//AD2m/wDBRJ/8do/4RPV/+e03/gok/wDjteh9do9zk+q1exmVyut2ME3jfS5fssElyLS4ZHeME7l2bDk+hJx6ZNd7/wAInq//AD2m/wDBRJ/8do/4RPV/+e03/gok/wDjtRPF0JK1/wAC44erF3seT+HoEefTZH1Szj1RGBuoYbJ/tkhwdyTN5h492XHTGOKhF5YyX+j3VlFZWt0+oKlwkbNJdIpJG2aQ8jPTDZ54HSvXv+ET1f8A57Tf+CiT/wCO0f8ACJ6v/wA9pv8AwUSf/HayWIoaa7O+z8v8i3Squ+n4+p5PHFYabBceXHb272uviW5CKqmKLcdjNjouDwTxRqlxa6rfaqYHE9vJqGnJvU/K43DOD3HPUV6x/wAInq//AD2m/wDBRJ/8do/4RPV/+e03/gok/wDjtEa9BJK+it07JL9AdKo25W/HzueX6zZW1r4ha1u5NL0/TEtU+yJfWm+ANuYvsw6BX5HqcdKjvrW0ht7F9Q1WzvWSz2pHq8EkSy8na0bMcpJ0B+82ADXqn/CJ6v8A89pv/BRJ/wDHaP8AhE9X/wCe03/gok/+O0OvQ11/Bh7Kr2Of0ic3OjWczW8lsXhU+TKxZk46Enkn3PNXa0/+ET1f/ntN/wCCiT/47R/wier/APPab/wUSf8Ax2uj69R7mP1Wp2MyitP/AIRPV/8AntN/4KJP/jtH/CJ6v/z2m/8ABRJ/8do+u0e4fVavYZ4Y/wCRtj/68Zv/AEZDXc1y2i+H9Q0zWBfXAurgCB4RHHprx/eZDnJdv7nTHeui86f/AKB1/wD+AzV5GJqRqVXKOx6FGLhBRZPRUHnT/wDQOv8A/wABmo86f/oHX/8A4DNXObE9U9W/5At7/wBe8n/oJqXzp/8AoHX/AP4DNUV2JrqzmtzYX6+bGybvsrHGRjNAHn0P/HvH/uj+VSVpJ4R1ZEVfOmO0Y/5BEn/x2l/4RPV/+e03/gok/wDjte99do9zyfqtTsZlFaf/AAier/8APab/AMFEn/x2j/hE9X/57Tf+CiT/AOO0/rtHuH1Wr2PPvEdjHN4u0+WG1hkvBY3LRO0YLb12FDk9wScemTWZ4bhhN3pksWqWYvVXNzb29i4uZDsOVnbzGxz3YDkDHWvVP+ET1f8A57Tf+CiT/wCO0f8ACJ6v/wA9pv8AwUSf/Haw+sUObmv+Br7Kry8tjyH7XYy32j3VpFZWt1JqCpcJGWkuVViRtmc8jPTDZ9B0qWySw0y2tGVLe1FprsrXZRVXyUPmhC+PujBGCeMYr1n/AIRPV/8AntN/4KJP/jtH/CJ6v/z2m/8ABRJ/8dpRxFGOt/w9P8vxKdKo+n4+v+Z5PfS22pnV2iYS282s2S7h0dcRg4PcH1qXVLG1tvEF1a3cul6fYpDH9hS9sy8Y67vKxIgV9x7Dd0xXqf8Awier/wDPab/wUSf/AB2j/hE9X/57Tf8Agok/+O0OvQas3+HlYXsqt9vxPKtRt7SAW0moalaX8y2KKkOqwvC0hGcPCx5RzxnALcDpXc6fMbjTbaZoZIDJErGKQ5ZMjoT6itz/AIRPV/8AntN/4KJP/jtH/CJ6v/z2m/8ABRJ/8drWOLoxbd9/IiWHqyS0MyitP/hE9X/57Tf+CiT/AOO0f8Inq/8Az2m/8FEn/wAdq/rtHuR9Vq9jMpbf/Xah/wBge7/9p1pf8Inq/wDz2m/8FEn/AMdp6eGdTtob6Vo7q6llsJraKNNOeLl8cklz/dHp1rDEYqlUpOMXqa0aE4TUmevUUUV456JwGv2kd9/wlVrLAlwsr7TE6hg3+ixYGD1rzTT7jw3/AMI14Y0+5stFNudMyZtQf/RhcfKHjEQ+Vp92Sc4Yc46mvVryR7TxBq3m2l8yzXCPG8NlLKrL5Ea5DKpHVSPwpv8AaC/8+Wp/+Cy4/wDiKO/n/wAH/MP6/L/I8l8M2tvrun+DLTV40vbcXupIYZoyEKqX2qUYkgAAYU5xgelR+IYtCh8O6+mtrCniWO9ZYFPE6weaohEffydmM4+XO7PNevf2gv8Az5an/wCCy4/+IrAu9CgvriT7Zc+JprGSYTPp0llK0JYEMBkw+ZtyAdofb2xjiqv71xNaW/rqcPqtsLrXteTXNa03TL/7UfsDXenST3iRbF2NaMsyk/RFPzZzmuo0Ky0+y+K+stPHarqFxp9q8UrRKks/3xKyjr1C7v8AgOe1df8A2gv/AD5an/4LLj/4ij+0F/58tT/8Flx/8RSTsrA9Tz//AJe77/r+uf8A0c9Oq/J4fvftNw8T3QSa4lmAbRbrI3uWx07ZxSf2BqH9+5/8Et1/hXt08XRjBJvoeZPD1HJtIo0Ve/sDUP79z/4Jbr/Cj+wNQ/v3P/gluv8ACtPrtDuR9Wq9ijVPVrdbrR7uCSJZleFhsZchuPStr+wNQ/v3P/gluv8ACj+wNQ/v3P8A4Jbr/Cpli6EouN9/IpYeqnex5VaTaI+l6VZS22liMaepEl2d0bS9GVIxwZAc7ujcgd6XRLaDWR4Xj1ONbtPsl3uSZcgkOoAIOenoemPavVP7A1D+/c/+CW6/wo/sDUP79z/4Jbr/AArL6xQ5ruX4GnsqttF+J5PfDTF0txf+WNfGpKGz/rtn2gbffy9uMfw/jUt/CJ9V1ZdU1Oysrz7QfsrT2by3KpgbTARICfoqnnOc13j/AA/vJZ83OpatcWwmEy2kukXDRhgdw52byAeQCxHtWv8A2BqH9+5/8Et1/hURr0er/B+X4lOnUWy/I4jQbaztPGWtKyW6XkoikU7FSSRCg3MB1wW6+9dTV7+wNQ/v3P8A4Jbr/Cj+wNQ/v3P/AIJbr/CuiOLoRVr/AIGMsPVbvYo0Ve/sDUP79z/4Jbr/AAo/sDUP79z/AOCW6/wqvrtDuT9Wq9ijVjTP+Q3Yf9dx/I1N/YGof37n/wAEt1/hU9ho13balb3E5u3SF95VNGugTwe5HvWdXF0ZQaT6Fww9RTTaNbxl/wAi/wD9vEP/AKGK5Sus15G1fSzawwajC/mI4d9KuGHysDjGwelYP9gah/fuf/BLdf4VzYPEU6UGpvqbYijOck4oo0Ve/sDUP79z/wCCW6/wo/sDUP79z/4Jbr/Cu367Q7nN9Wq9ijRV7+wNQ/v3P/gluv8ACj+wNQ/v3P8A4Jbr/Cj67Q7h9Wq9jidft7JfGGhXd7Db4AmjWWVV+/gFFBPfOce/SucF5ZSX+j3dlFZWl2+oKk6xs0l0ikkbZpDyPTDZ54HSvWf7A1D+/c/+CW6/wo/sDUP79z/4Jbr/AArB16HNfm6329P8jVUqvLa34+p5bbWFtH4ZW9a3UJJqcq38yr85t/PbIJHOzIXI6Yz71FcfZxN4i/4RMqsP2e0ybRSyBd779gUjI25+6RznvXp194U1C9tvKF5qFswYMstvpF0rKR9QQR7EEH0qLT/Bd5YSTTG6vrq5nx5txPotxuYDoMKoAAyeABU+3obJ6enlb+vzL9nV1dvx87nm8+nWdzousf2de6bexNp7lrXTbEpEXGCrk73XzB+DflXd6ObQ6NanTTCbbyx5fkY2Y9scda2/7A1D+/c/+CW6/wAKP7A1D+/c/wDgluv8K1hiqEOv4Pz/AMzOVGrLp/Wn+RRoq9/YGof37n/wS3X+FH9gah/fuf8AwS3X+FafXaHcz+rVexRoq9/YGof37n/wS3X+FH9gah/fuf8AwS3X+FH12h3D6tV7Gr4J/wCPTU/+v7/2jFXTVzvh63k0a2uUuItQne4uPOymk3KhfkRccqf7mfxrX/tBf+fLU/8AwWXH/wARXiVpKVSUl3PUppxgky3RVT+0F/58tT/8Flx/8RR/aC/8+Wp/+Cy4/wDiKyLLdc94z/5BNt/19J/Jq1v7QX/ny1P/AMFlx/8AEVma/E2r6ekEEOoQukqyBn0q5YHGeMBR61pSko1E33Imm4NI5Sir39gah/fuf/BLdf4Uf2BqH9+5/wDBLdf4V7n12h3PL+rVexRoq9/YGof37n/wS3X+FH9gah/fuf8AwS3X+FH12h3D6tV7FGuBl/s7S7vxLvs4DNJdxMyB/KPlMqfO5X5vL3ZLdjznvXpn9gah/fuf/BLdf4Uf2BqH9+5/8Et1/hWc8VQk73/AuNCrHp+J4/dG2x4gt9PazWzks7d1TT12Qk+bgsuDhj0ywx2Hat+8sNFsPEC2+rW9rb6Z9lBs43ULF5xY7yB08zG3HfrjvXoP9gah/fuf/BLdf4VQvvBmp3dwJodV1azbZsZYNHuCrD/ddGAPuMH3rL29FLR3+Xr/AJmnsqre34+n+R5nbDfo+gi5uLaDSTFPltSgMsBk8z5fMBdRnGcFj1z3q7HY2MNx4fuLu4tr22W9mEdw9uY40BQlEXeSdu77pyR0x2r0ey8J3On2UVpaG5SGJdqj+xrs/rjk1P8A2BqH9+5/8Et1/hTjXoLXm/BidKq9LfkUaKvf2BqH9+5/8Et1/hR/YGof37n/AMEt1/hXR9dodzH6tV7FGir39gah/fuf/BLdf4Uf2BqH9+5/8Et1/hR9dodw+rVexRre8Ef8hW1/7DD/APpCaz/7A1D+/c/+CW6/wrb8K6dPYaxp8bxXkrPqElxJI2nTwxxr9ldOS64HIHU9TiuPF4inVp8sX1OnD0ZwneSPS6KKK8s7ilrBRdHuTJF5yhDmMEjd7ZFXaqaoZBpdx5MywPs+WRm2hffPardAFDWfI+wx/avM2fa7bHl4zu89NvXtuxn2zV+quoic2qfZYllfz4cqwBAXzV3HnuFyR7irVABRRRQAUUUUAFFFFAFO63/2hY7YBIu990hUny/kPOe2enPrVyqOorOkltc28Mtx5DktDCVDOCpH8TKOM560v2+482Bf7JvMSgF33w4iycYb95njr8oPtnpQBdoqiNQudsx/si9BjxtG+HMvOPl/efj82OPfig6hc7YT/ZF6TJncN8OYucfN+8/H5c8e/FAElsJft175iKI96+WQBlhsGc/j61arFs71xq2rCPTb1mRlJJMIVyFUAL+8zyOfmA98GrT6jdLBHINFvmZyQ0YeDcmPXMmOfYnpzigDQoqmb2cXEsY0y6KxqSsoaLbIQOg+fOT05AHrio/7Ruvswl/sW+3lyvlb4NwGPvf6zGO3XPHSgDQoqn9tn+1+T/Zl1s258/dFs+7nGN+7OeOnX25qP+0br7MZf7Fvt4cL5W+DcRj73+sxjt1zz0oA0KKpi9nNxFGdMugsigtKWi2xkjofnzkdOAR6ZqNNRumgkkOi3yshAWMvBufPpiTHHuR14zQBoUVS+33HmwL/AGTeYlALvvhxFk4w37zPHX5QfbPSkGoXO2Y/2Regx42jfDmXnHy/vPx+bHHvxQBeoqidQudsJ/si9JkzuG+HMXOPm/efj8uePfil+33Hmzr/AGTeYiBKPvhxLg4wv7zPPX5gPfHSgC7RWe+o3SwRyDRb5mckNGHg3Jj1zJjn2J6c4qQ3s4uJYxpl0VjUlZQ0W2QgdB8+cnpyAPXFAFyis/8AtG6+zCX+xb7eXK+Vvg3AY+9/rMY7dc8dKk+2z/a/J/sy62bc+fui2fdzjG/dnPHTr7c0AXKKz/7Ruvsxl/sW+3hwvlb4NxGPvf6zGO3XPPSpBezm4ijOmXQWRQWlLRbYyR0Pz5yOnAI9M0AXKKz01G6aCSQ6LfKyEBYy8G58+mJMce5HXjNP+33HmwL/AGTeYlALvvhxFk4w37zPHX5QfbPSgC7RVEahc7Zj/ZF6DHjaN8OZecfL+8/H5sce/FB1C52wn+yL0mTO4b4cxc4+b95+Py549+KAL1FUvt9x5s6/2TeYiBKPvhxLg4wv7zPPX5gPfHSmPqN0sEcg0W+ZnJDRh4NyY9cyY59ienOKANCiqZvZxcSxjTLorGpKyhotshA6D585PTkAeuKj/tG6+zCX+xb7eXK+Vvg3AY+9/rMY7dc8dKANCiqf22f7X5P9mXWzbnz90Wz7ucY37s546dfbmo/7Ruvsxl/sW+3hwvlb4NxGPvf6zGO3XPPSgDQoqmL2c3EUZ0y6CyKC0paLbGSOh+fOR04BHpmo01G6aCSQ6LfKyEBYy8G58+mJMce5HXjNAGhRVL7fcebAv9k3mJQC774cRZOMN+8zx1+UH2z0pBqFztmP9kXoMeNo3w5l5x8v7z8fmxx78UAXqKonULnbCf7IvSZM7hvhzFzj5v3n4/Lnj34pft9x5s6/2TeYiBKPvhxLg4wv7zPPX5gPfHSgC7RWe+o3SwRyDRb5mckNGHg3Jj1zJjn2J6c4qQ3s4uJYxpl0VjUlZQ0W2QgdB8+cnpyAPXFAFyis/wDtG6+zCX+xb7eXK+Vvg3AY+9/rMY7dc8dKk+2z/a/J/sy62bc+fui2fdzjG/dnPHTr7c0AXKKz/wC0br7MZf7Fvt4cL5W+DcRj73+sxjt1zz0qQXs5uIozpl0FkUFpS0W2MkdD8+cjpwCPTNAFyis9NRumgkkOi3yshAWMvBufPpiTHHuR14zT/t9x5sC/2TeYlALvvhxFk4w37zPHX5QfbPSgC7RVEahc7Zj/AGRegx42jfDmXnHy/vPx+bHHvxQdQudsJ/si9JkzuG+HMXOPm/efj8uePfigC9RVL7fcebOv9k3mIgSj74cS4OML+8zz1+YD3x0pj6jdLBHINFvmZyQ0YeDcmPXMmOfYnpzigDQoqmb2cXEsY0y6KxqSsoaLbIQOg+fOT05AHrio/wC0br7MJf7Fvt5cr5W+DcBj73+sxjt1zx0oA0KKp/bZ/tfk/wBmXWzbnz90Wz7ucY37s546dfbmo/7Ruvsxl/sW+3hwvlb4NxGPvf6zGO3XPPSgDQoqmL2c3EUZ0y6CyKC0paLbGSOh+fOR04BHpmo01G6aCSQ6LfKyEBYy8G58+mJMce5HXjNAGhRVL7fcebAv9k3mJQC774cRZOMN+8zx1+UH2z0pBqFztmP9kXoMeNo3w5l5x8v7z8fmxx78UAXqKonULnbCf7IvSZM7hvhzFzj5v3n4/Lnj34pft9x5s6/2TeYiBKPvhxLg4wv7zPPX5gPfHSgC7RWe+o3SwRyDRb5mckNGHg3Jj1zJjn2J6c4qQ3s4uJYxpl0VjUlZQ0W2QgdB8+cnpyAPXFAFyis/+0br7MJf7Fvt5cr5W+DcBj73+sxjt1zx0qT7bP8Aa/J/sy62bc+fui2fdzjG/dnPHTr7c0AXKKz/AO0br7MZf7Fvt4cL5W+DcRj73+sxjt1zz0qQXs5uIozpl0FkUFpS0W2MkdD8+cjpwCPTNAFyis9NRumgkkOi3yshAWMvBufPpiTHHuR14zT/ALfcebAv9k3mJQC774cRZOMN+8zx1+UH2z0oAu0VRGoXO2Y/2Regx42jfDmXnHy/vPx+bHHvxQdQudsJ/si9JkzuG+HMXOPm/efj8uePfigC9RVL7fcebOv9k3mIgSj74cS4OML+8zz1+YD3x0pj6jdLBHINFvmZyQ0YeDcmPXMmOfYnpzigDQoqmb2cXEsY0y6KxqSsoaLbIQOg+fOT05AHrio/7Ruvswl/sW+3lyvlb4NwGPvf6zGO3XPHSgDQoqn9tn+1+T/Zl1s258/dFs+7nGN+7OeOnX25qP8AtG6+zGX+xb7eHC+Vvg3EY+9/rMY7dc89KANCiqYvZzcRRnTLoLIoLSlotsZI6H585HTgEemajTUbpoJJDot8rIQFjLwbnz6Ykxx7kdeM0AaFFUvt9x5sC/2TeYlALvvhxFk4w37zPHX5QfbPSkGoXO2Y/wBkXoMeNo3w5l5x8v7z8fmxx78UAXqKonULnbCf7IvSZM7hvhzFzj5v3n4/Lnj34pft9x5s6/2TeYiBKPvhxLg4wv7zPPX5gPfHSgC7RWe+o3SwRyDRb5mckNGHg3Jj1zJjn2J6c4qQ3s4uJYxpl0VjUlZQ0W2QgdB8+cnpyAPXFAFyis/+0br7MJf7Fvt5cr5W+DcBj73+sxjt1zx0qT7bP9r8n+zLrZtz5+6LZ93OMb92c8dOvtzQBcorP/tG6+zGX+xb7eHC+Vvg3EY+9/rMY7dc89KkF7ObiKM6ZdBZFBaUtFtjJHQ/PnI6cAj0zQBcorPTUbpoJJDot8rIQFjLwbnz6Ykxx7kdeM0/7fcebAv9k3mJQC774cRZOMN+8zx1+UH2z0oAu0VRGoXO2Y/2Regx42jfDmXnHy/vPx+bHHvxQdQudsJ/si9JkzuG+HMXOPm/efj8uePfigC9RVL7fcebOv8AZN5iIEo++HEuDjC/vM89fmA98dKY+o3SwRyDRb5mckNGHg3Jj1zJjn2J6c4oA0KKpm9nFxLGNMuisakrKGi2yEDoPnzk9OQB64qP+0br7MJf7Fvt5cr5W+DcBj73+sxjt1zx0oA0KKp/bZ/tfk/2ZdbNufP3RbPu5xjfuznjp19uaj/tG6+zGX+xb7eHC+Vvg3EY+9/rMY7dc89KANCiqYvZzcRRnTLoLIoLSlotsZI6H585HTgEemajTUbpoJJDot8rIQFjLwbnz6Ykxx7kdeM0AaFFUvt9x5sC/wBk3mJQC774cRZOMN+8zx1+UH2z0pBqFztmP9kXoMeNo3w5l5x8v7z8fmxx78UAXqKonULnbCf7IvSZM7hvhzFzj5v3n4/Lnj34pft9x5s6/wBk3mIgSj74cS4OML+8zz1+YD3x0oAu0VnvqN0sEcg0W+ZnJDRh4NyY9cyY59ienOKkN7OLiWMaZdFY1JWUNFtkIHQfPnJ6cgD1xQBcorP/ALRuvswl/sW+3lyvlb4NwGPvf6zGO3XPHSpPts/2vyf7Mutm3Pn7otn3c4xv3Zzx06+3NAFyis/+0br7MZf7Fvt4cL5W+DcRj73+sxjt1zz0qQXs5uIozpl0FkUFpS0W2MkdD8+cjpwCPTNAFyis9NRumgkkOi3yshAWMvBufPpiTHHuR14zT/t9x5sC/wBk3mJQC774cRZOMN+8zx1+UH2z0oAu0VRGoXO2Y/2Regx42jfDmXnHy/vPx+bHHvxQdQudsJ/si9JkzuG+HMXOPm/efj8uePfigC9RVL7fcebOv9k3mIgSj74cS4OML+8zz1+YD3x0pj6jdLBHINFvmZyQ0YeDcmPXMmOfYnpzigDQoqmb2cXEsY0y6KxqSsoaLbIQOg+fOT05AHrio/7Ruvswl/sW+3lyvlb4NwGPvf6zGO3XPHSgDQoqn9tn+1+T/Zl1s258/dFs+7nGN+7OeOnX25qP+0br7MZf7Fvt4cL5W+DcRj73+sxjt1zz0oA0KKpi9nNxFGdMugsigtKWi2xkjofnzkdOAR6ZqNNRumgkkOi3yshAWMvBufPpiTHHuR14zQBoUVS+33HmwL/ZN5iUAu++HEWTjDfvM8dflB9s9KQahc7Zj/ZF6DHjaN8OZecfL+8/H5sce/FAF6iqJ1C52wn+yL0mTO4b4cxc4+b95+Py549+KX7fcebOv9k3mIgSj74cS4OML+8zz1+YD3x0oAu0VnvqN0sEcg0W+ZnJDRh4NyY9cyY59ienOKkN7OLiWMaZdFY1JWUNFtkIHQfPnJ6cgD1xQBcorP8A7Ruvswl/sW+3lyvlb4NwGPvf6zGO3XPHSpPts/2vyf7Mutm3Pn7otn3c4xv3Zzx06+3NADdaMa6LdGdWeMRncqnBI+tXqybu81CfS5fsulXcNwzBEV3gyuf4ziQjA/P2Na1AFDWVhexjFxI0afa7YgqucsJ0Kj8WwM9s1fqnqnmfY08m3W4b7RBlHXcAPNTLfVRlgexGauUAFFFFABRRRQAUUUUAFFFFABRRRQBUtRJ9vvd8Kou9NjhcGT5ByT3weKt1Sswg1LUCsu9i6bkwRs+Qce/rV2gAqjq+pHSrD7QsBnYyJGqBtuSzAdfxq9WL4q/5BMX/AF9w/wDoYoAg/wCEjvv+gSv/AIFD/wCJo/4SO+/6BK/+BQ/+JqrRQBa/4SO+/wCgSv8A4FD/AOJo/wCEjvv+gSv/AIFD/wCJqrRQAl345TT5FS/gs7Z2GVWbUEQkeuCKr/8ACyLD/nppn/g0j/wrlfFP/Iw/9usf/oclZNejRwXtIKfNv5HHUxXJJxsegf8ACyLD/nppn/g0j/wo/wCFkWH/AD00z/waR/4V5/RWv9nL+b8P+CZ/XP7p6B/wsiw/56aZ/wCDSP8Awo/4WRYf89NM/wDBpH/hXluv6g2m6HcXEI3TkCOFc/ekY7VH5kVlaJeXGkWWp2OpST3s2mDzlcsZJJomXcOvJOQy/hUPAxTs5fh/wS1im1dRPZ/+FkWH/PTTP/BpH/hR/wALIsP+emmf+DSP/CvG4PFbvpFxqUlnDJaxQ+aslndrODz91uAVI78EDB5p/wDwlP2fRbjUtQgtxbxBTHLaXizRyljgAMQuDnGcjHPWl9Sh/P8AgH1mX8v4nsP/AAsiw/56aZ/4NI/8KP8AhZFh/wA9NM/8Gkf+FeHzeMmudP1KCyawN9DZPcxNaXwnjAHBywXhhnOCMH1643tPuL6Tw7b3EkULXjQK2wzHYxx3fZkZ6/d496I4KMr2lt5ev+QPEyja8fx9P8z1L/hZFh/z00z/AMGkf+FH/CyLD/nppn/g0j/wrwzQ7/W7zUNNFxHCYZhNeMy3rsTGzYUbfLAwA3Azg4zxitW0e/HjCWK8uQ8bWQdIY8hE/eEZ9zjqaI4JO3vb+XkEsU43028/M9e/4WRYf89NM/8ABpH/AIUf8LIsP+emmf8Ag0j/AMK8/orT+zl/N+H/AASPrn909A/4WRYf89NM/wDBpH/hR/wsiw/56aZ/4NI/8K8/qOb/AI95P90/yo/s5fzfh/wQ+uf3T1UeJrwpvGloVIzu+1jGP++azv8AhZFh/wA9NM/8Gkf+FQWX/It2/wD16L/6BXn1j/yD7f8A65L/ACFcmHw/t29bWOitW9lbTc9I/wCFkWH/AD00z/waR/4Uf8LIsP8Anppn/g0j/wAK8/orr/s5fzfh/wAE5/rn909A/wCFkWH/AD00z/waR/4Uf8LIsP8Anppn/g0j/wAK8/rkfE15qEmpP/ZUsqDSIFu5kjYgTktxGcdfkV+PcVE8Coq7l+BUcU5O1j27/hZFh/z00z/waR/4Uf8ACyLD/nppn/g0j/wryDUPEwt7mzgs4reR7uHzomubnyEccfKrbW3Nz09O9NvfE0ttfW9kllEl1NAJ/Lu7tYc+qIQGDsMdMgdOaHgoL7f4DWJk/s/iew/8LIsP+emmf+DSP/Cj/hZFh/z00z/waR/4V43qfi+HT75bIizhuRCs0iX18tuFDZ+UHDbm47ce9Q23iG71TxBo7aesf9nXlnLK6vLhshlB4CnJGcDBwcnpgZX1KDdlP8A+sytfl/E9q/4WRYf89NM/8Gkf+FH/AAsiw/56aZ/4NI/8K8g8RXWqR3Wn2umwwsJ7hcs100THaC5U4RvlIXBOe+MVX0qDUta0mKfUbtrWGS4kmZLW4ctLGSdi+ZhSgHH3eoA6ZIo+oq7Sl+HoH1rS7X4ns/8Awsiw/wCemmf+DSP/AAo/4WRYf89NM/8ABpH/AIV5H4YmmntbsmWWazW5ZbOWZizPEAOdx5Ybt2CeSPWtyrWXppPm/D/gkvFtO3Kegf8ACyLD/nppn/g0j/wo/wCFkWH/AD00z/waR/4V5/RT/s5fzfh/wRfXP7p6PaeOo7+UxWMNncyKu4pDqCOQPXAHTkVc/wCEjvv+gSv/AIFD/wCJrzzwx/yNsf8A14zf+jIa7mvPrUvZTcL3OylP2keYtf8ACR33/QJX/wACh/8AE0f8JHff9Alf/Aof/E1VorE0LX/CR33/AECV/wDAof8AxNI3iW8RSz6UiqBkk3YAA/75qtVPVv8AkC3v/XvJ/wCgmgB//CyLD/nppn/g0j/wo/4WRYf89NM/8Gkf+Fedw/8AHvH/ALo/lUlet/Zy/m/D/gnn/XP7p6B/wsiw/wCemmf+DSP/AAo/4WRYf89NM/8ABpH/AIV5/VDWtQ/svRbq8C73jT92n95zwo/EkClLL1FX5vw/4I1i23bl/E9P/wCFkWH/AD00z/waR/4Uf8LIsP8Anppn/g0j/wAK8V0S8u9Fg1Kw1aWe+msolu0ckvJKjLyBnrh1YAfSrFp4okm0q51F7OKW2igMyvY3az7sfwEYUhvbkcHmo+pR6z/Ar6zLpH8T2L/hZFh/z00z/wAGkf8AhR/wsiw/56aZ/wCDSP8Awrx6LxTt0e61O9gt/ssMQkWWzvFnR88bckLhs+2Pes648amexvoLJrA3q2clxA1rfLOqhRzuIX5WGcgYIOOtKWEhFXc/PYccRKTso/ie5f8ACyLD/nppn/g0j/wo/wCFkWH/AD00z/waR/4V5dps97N4ft55YoWvGgDbPOOxjjjLbcjPf5ePeuc0zUNdvNS09ZYoBFPcT3Tbb6Q5iU7AuPLAwNwIGcHHarlgFGXLzfgTHFNq9vxPdP8AhZFh/wA9NM/8Gkf+FH/CyLD/AJ6aZ/4NI/8ACvIbSS/XxhNHe3IaNrLesMfCJ+8Iz7nHU1nyape33i7SJra4aPSnlmhSMZH2grGxLn1XIwB3xn0qVgk0ve38vO3cf1p3em3+Vz27/hZFh/z00z/waR/4Uf8ACyLD/nppn/g0j/wrz+itP7OX834f8Ej65/dPQP8AhZFh/wA9NM/8Gkf+FO/4WFC9rcTW1va3QtommkW31BHYKBk8AV57S2/+u1D/ALA93/7TrKtgvZU3Pm2NKeJ55qNj3OiiivOOww7rxJ5Oo3NpDp1xObZwjurooLFFfjJz0YVH/wAJPN/0B7n/AL+x/wDxVYes3kunz+Jry3ga4lglEiRLjLEW0RxyQP1FcufiNbaZ8PbDW9cNvb6jc2kcy2U91HE0+cAsgDN8vOR3A6gGjf8Arv8A8MB6J/wk83/QHuf+/sf/AMVR/wAJPN/0B7n/AL+x/wDxVc4+trqWjm68Kz2Wpln8vzo7lXiiOMkkrnOP7o5OR06jmD4vvU8DeGJ7q7EN7rZjilu0iDGPMbOzImCCx24UYPJHB6Uf8D8Q3Vz0r/hJ5v8AoD3P/f2P/wCKo/4Seb/oD3P/AH9j/wDiq5TwnfXF5b3i3Gpf2jHDPsikmgMFyq7Qds0ZRNrAnjCjIIOK6Cm9AHP45to5GSS02Opwytdwgg/99U3/AIT20/59x/4GQ/8AxdeYmKNry+JjUn7dc8lf+mz0vkxf880/75FelDAc0VLm38jhli+WTVj03/hPbT/n3H/gZD/8XR/wntp/z7j/AMDIf/i68y8mL/nmn/fIo8mL/nmn/fIq/wCzv734C+uf3T03/hPbT/n3H/gZD/8AF0f8J7af8+4/8DIf/i68W8TTXUcVvY6Lsi1C7ZvLfy1bYqLuY4Ix6L/wKli1+zbRtOuzaPLPfYSO2iVd5kAJZcsQoxtbqR0rP6irtc23kV9adk+X8T2j/hPbT/n3H/gZD/8AF0f8J7af8+4/8DIf/i68ck1q2jht86Zdfarl2SOzMSLISvLcswXAHOd2D2zSyaxapBbEadctc3JYR2fkqsp2/ezuIUAeucHIwTkU/qMf5/wF9af8v4nsX/Ce2n/PuP8AwMh/+Lo/4T20/wCfcf8AgZD/APF14bqHiMKlgdO02SR5b4WtzCyRh4jgkpywG48YIJGO/Sr/AIkvJdM0WaaxsWmnZdkbRiPEbnhSQzDIyR0zS+pRs3zbeXz/AFH9Zd0rb+Z7H/wntp/z7j/wMh/+Lo/4T20/59x/4GQ//F14npeoXl7c6lu0qZRbbI445zCMyBQWGUY8ncPUYHrxVnw9LPd212dRELyxXkkQ8uMBQoPAH+JprApuyl57CeKaV7fiex/8J7af8+4/8DIf/i6P+E9tP+fcf+BkP/xdeZeTF/zzT/vkUeTF/wA80/75FX/Z3978BfXP7p6b/wAJ7af8+4/8DIf/AIunxeOIJpBHDZmR26Kl1CSfw3V5f5MX/PNP++RVrS4o11ywKooPnjkD2NRPAckXLm28io4vmklY9Ll8Wm3jMk+lzRIOrPPEAPxLVX/4T20/59x/4GQ//F1geMwG8PYYAj7RDwf98VyPkxf880/75FY4fC+2i5XsaVq/spWsem/8J7af8+4/8DIf/i6P+E9tP+fcf+BkP/xdeZeTF/zzT/vkUeTF/wA80/75FdP9nf3vwMfrn909N/4T20/59x/4GQ//ABdH/Ce2n/PuP/AyH/4uvMvJi/55p/3yK5W11e4k8UhpAh0i4mksoB5a8SoAd2cZ5Icf8BFQ8Ak0ubfy/wCCNYptN8p7v/wntp/z7j/wMh/+Lo/4T20/59x/4GQ//F144NZtZL6SCDTbmeOKbyJbmKJWSN/QjO/jIyQuB60Q61a3N40Vvp1zLCsxt2ukiUxrIDggjO4AHvtx70LAxe0/wD60/wCU9j/4T20/59x/4GQ//F0f8J7af8+4/wDAyH/4uvGX8QWCOz/Ypms0l8l70Rp5Svu2467sZ4yFx70mi6jLqOr6vbXFh5MdncCOJyqYI2KecMTnnPTGCO+RQsDFuyl+A/rTSvy/iez/APCe2n/PuP8AwMh/+Lo/4T20/wCfcf8AgZD/APF14NrWu3tpqF7FaaRcFLa0JBAg2mR2xG/L5x8p4xnnkVdvINR+yyXs96umW9vbBwnlRMxcDLebkEY9kI781P1NWb5tvL18/If1h3tb8f67ntn/AAntp/z7j/wMh/8Ai6P+E9tP+fcf+BkP/wAXXk+kub3RrO6urVIZpoUeSPb90kZIq35MX/PNP++RWv8AZ3978CPrn909N/4T20/59x/4GQ//ABdH/Ce2n/PuP/AyH/4uvMvJi/55p/3yKPJi/wCeaf8AfIo/s7+9+AfXP7p6pD4u+0x77fTJZUzjdHPEwz+DVJ/wk83/AEB7n/v7H/8AFVxvgdQtnqYUAD7d0A/6YxV09eZUjyTcex2wlzRUi3/wk83/AEB7n/v7H/8AFUf8JPN/0B7n/v7H/wDFVUoqCi3/AMJPN/0B7n/v7H/8VTJvFpt4/MuNMmiTpueeJR+Zaq9c740AbSLcMAR9qTg/RquEeeSj3JlLli2dF/wntp/z7j/wMh/+Lo/4T20/59x/4GQ//F15l5MX/PNP++RR5MX/ADzT/vkV6f8AZ3978Di+uf3T03/hPbT/AJ9x/wCBkP8A8XR/wntp/wA+4/8AAyH/AOLrzLyYv+eaf98isbxJNPDbW9npOyK/vZRHFJ5atsAG5mwRjoCPqRUywHKr834f8EccXd2sezf8J7af8+4/8DIf/i6P+E9tP+fcf+BkP/xdeK2mv2x8P2N5c2pe5uW8j7PEi7mmGQyckAcq3UgcVNLrdtDbwtJpl0lzPMYI7RokWRnAycEtsIwCc7se9L6jH+f8B/Wpfynsn/Ce2n/PuP8AwMh/+Lo/4T20/wCfcf8AgZD/APF146+sWsdtAz6dci4uJGjjs2hUSsygk9Ttxgfe3bfes7UPEmw2S2GmSNK98trcwOkYeI7d23lwMkYIIJGM89KTwUV9v8O//DgsS39n8T3P/hPbT/n3H/gZD/8AF0f8J7af8+4/8DIf/i68f8QXh0rRZ7q2sGuJFQ7fLEfyHHBO4jjOOmT7VQ0rUL+6vr2OTSph9lgjQJMYFDTYy2SjHGQyn0GD0PFP6ilLl5vwD602r2/E9v8A+E9tP+fcf+BkP/xdH/Ce2n/PuP8AwMh/+LrxjRr1jp1/cay1uv2W6lVnVAqIi/4epqv4e1C81LXNUS/gjhgRIZLWHYNyIwbluM7jgHHaksCnb3t9dv8Agh9aavpse3/8J7af8+4/8DIf/i6P+E9tP+fcf+BkP/xdeZeTF/zzT/vkUeTF/wA80/75Faf2d/e/An65/dPTf+E9tP8An3H/AIGQ/wDxdWrHxbFfX1pbrZSBbqVolmWWN1VgjPztY9kP5ivKPJi/55p/3yK6LwMoXVLUKAB/bD8Af9OJrnxGE9jDmvc1o4j2krWPVqKKK4TqCiiigCjq4U2Ue+4NuPtVv84BOT5yYXj+8fl/Hnir1UtWOLOP/R/tH+kwfJzx++T5uP7v3vwq7QAUUUUAFFFFABRRRQAUUUUAFFFFAFO0z/aN/wDudnzph8H958g5/Dpx6VcqnaZ/tG//AH2/50wmT+7+Qcfj149auUAFYvir/kExf9fcP/oYrarF8Vf8gmL/AK+4f/QxQBx3/CUx/wDCenw19ln3CzFx5/kSFSSxGN23aBgfezjPHXir+m6zb6pealb26Sq+nXP2aUuAAzbFfK4J4w46471k3ukTS+Pvtc+nfbdMvdMNlPnyykZDlvnViCysDjgN78UmgeAND8P6vfajbadp/nXFz51s0dkkbWi+WqFEYcgcE8Y+8eKcdtfP8/8AIJbu3l+X+ZPYeILy88YzaXJYG2s0s/PjeXiWQiTYTt/hXjjPJ68VlWXjK91PVIjp82lSQy3LRDS3lC3hhWQo1wCZMY4J27M4HXPFbf8AZ9z/AMJ4dR8r/RP7N8jzNw+/5m7GM56d+lc1F4Yv7fSIfDdtpZiji1MXq6oskflhRP5vTd5nmbfl+7j/AGqUfs3/AK1/yE93/XT/ADLvin/kYf8At1j/APQ5Kya1vFP/ACMP/brH/wChyVk19FhP4ETyMR/FYUUUV0mBmatosOszWYvSslrbyGR7Z49yzHaQuc8YGc4we1Vl8L2lpqi3ujCLSyYHhkS2t0UPnBVsYxlSO4PWtPUb+DTNOnvbt1SKFCzFmC/hk9z0rO0/xJZ3GhwalqN1YWUdx9wi8Vl+m4hfm9R2rJqnzee/6Gi57eX9Mgi8OXX9otqE2oQpeCEwpLa2YiyCQcuCzbzx7Dk8VEfCAujeS391ELi5jRN9lb+QFKNuVyCzbmBA5J6cYrdutRsrK3W4vby3t4WxiSWVUU/iTiqV14l0u1u9Oga7hb+0GIhdZV2kAZBznkE4Ax3NDjTWj/q4KU+n9WAaXf3Npc2+raklwk8DQ7ILbylAIwWOWYlvxA9qeulPJ4cOl30sNwxgMJk8jCEYwCULH27/AJVb+32ZhEwuoPKaTyw/mDaXzt259c8Y9eKjj1bTprl7eG/tZJ4wS8SzKWUA4JIzkYPFVyws0+v9fqK8t+xm6b4P0nTNRF3DZWm9Io44ttsqlGXOXz/eOevtWiNPxrh1HzetuIPL2/7W7Oc/pTrXVLC9hklsr62uI4jiR4plcJ9SDxTrLUbLUo2k0+8t7tFOC0EquAfTINNKOlvUT5upZoooqiQqOb/j3k/3T/KpKjm/495P90/yoA72y/5Fu3/69F/9Arz6x/5B9v8A9cl/kK9Bsv8AkW7f/r0X/wBArz6x/wCQfb/9cl/kK8nL95fI9DGbRLFFFFeseeFYcXhPS3nurjVbO01G5uZmkMs9srFV6KgznAAAH61uVi6p4jhsNcsdJjNs1zd5Yia5EWxQQPQksc8DAzg81ElDTm9C4820SvF4cvLXSYtOtdRt3tow6CK7shMmwklRjcDkA4znBA6Up8NTxaTBptpfQtaRw+U0V7aCdT1+YYZcHnGDkYxxWvFqdhPePaQXtvJcx/fhSVS6/VQcimpq2myXzWUeoWr3S/egWZTIP+A5zU8lNr18x80zKg8NT6Y8Umjaj5Mi2yW0puYfOEipna2Ay4YZPOce1W30ec6lp98L7fPaxvFK0sIPnK20n7pUKcqOcEe1GmeJdM1UTm3u4R5VwbfBlXLNnAxg/wAWDj1xV2bUbK38z7ReW8XlFRJvlVdhb7oOTxnt601yNcy9RNyvZ7mfq/hqy1m/juryC2laOCSJRNbrJy2MNk+mDx71TufCZXSbXTtGuLbT4IyDcotpkXWAB8+11ODjnnnp066763pUdrFcyanZpBMcRStcIFkP+yc4P4VJdapYWIjN7fW1sJTiPzplTf8ATJ5pOFPUalNWE0+C7t7fy72a2l24CfZrcwqq46YLt/SrdICCAQcg9DS1qZhRRRQBoeGP+Rtj/wCvGb/0ZDXc1w3hj/kbY/8Arxm/9GQ12N/Y2+p6fcWN7H5ttcxtFKm4jcrDBGRyOD2r5/G/xn/XQ9fC/wANXLFFZQ8NaSPC/wDwjotP+JV5Bt/s/mP/AKvpt3Z3fjnNMu/Cmi33hVfDl1Z79JWNIlt/NcYVCCo3A7uMDvXJ6HQvM2Kp6t/yBb3/AK95P/QTVmGJLeCOGJdscahVGc4AGBVbVv8AkC3v/XvJ/wCgmn1F0PPof+PeP/dH8qkqOH/j3j/3R/KpK+qPBCs3WNHj1oWsV0ytaxTCWWBk3LPgHCn2yQe+cVpUUmk1ZjTa2MI+FbO21OG90VYdKkSOSKT7LbovmKwGOMYyCARkHuO9Mh8OXI1J9QuL+JbvyGhSW0tBEeSDufLMHIxx0HJ45qbVPEcNhrljpMZtmubvLETXIi2KCB6EljngYGcHmtCDUrG6upLa2vbeaeL/AFkUcqsyfUA5FZqNNvQtuaWpiN4R+2SXUup3UTTXFv5Beyt/I4yGDHLNuYEDB6D0q8NK1C5tZ7fVdTWeGWBoNkNsIs7hjcxLMS30IHPSra6vpr3rWaahatdJ96ATqXX6rnNVdJ8SaZrEBltbqH/XtAqmVSWYE4xg9wCR6jmjlp/D/X9ahee/9f1oTW+nONBGnX0kNyfJ8lm8jCMMYGULHt1559qztM8F6Tpt5BcJZWbNBBHGh+yqGDqSTJn+8ePcY61sS6jZQLKZry3jELBJC8qjYx6A88E5HHvUTa3pSQwzPqdmsVwcQubhAsh6fKc8/hVNQcuZ7/1/mJOSVkKdOB1ttQMmQ1t5BiK/7Wc5/pis2bwbo7apYXlrp9jbfZJGdkjtEHmZUgZIxjB578itS61SwsZIkvb62t3mOI1lmVC/0yeatUcsX8v+HFzSXz/4YWiiirJClt/9dqH/AGB7v/2nSUtv/rtQ/wCwPd/+065cZ/Al/XU3w/8AFR7nRRRXzx7BxN/b/a9S8RW+7b50wjz6ZtohWJY+GJbn4fW/hvXikflW8dsz2UpbcqYwwLIME7RkYOPU9a2NTv7fS77xFfXsixW9vOskjswUAC2i7niqFl4z0G68N2Wtz6rZWdneICj3F1GgViOUJJxuHQjsRQuvy/W36g+htuu6JlHdcDNcnbeE7y18K+HbdXtn1PQikke5mEUjCNkZd2MgEMedpwcHB6VvXOv6NZWcV3eatY29tMu6KaW5RUkHqrE4I57UyfxLoVrci3uda0+GdoxKIpLpFYoejYJzj3p6r+uwX08ivoWlXdrfajqeqCBLvUHQvDbSM8cSou1QGKqWPUk7R1xjjNbdUbjWtKs7+Gxu9Ts4Luf/AFVvLOqySf7qk5P4VTs/FWl33iu/8PwXEZvbGNJHTzVJbdnIABz8uBnjjcPWlvsGxx3/AC933/X9c/8Ao56dTf8Al7vv+v65/wDRz06vp6P8OPojw6nxv1CiiitCDDufD39o+IZL/UJpRFHCsVqttcyQsuTlyxQqeTt4yelUl8M3FrZXVjDBY6hYtdefDb37uxAYfMN5DEHdkg4bOT0611NQXl3DYWU11cuEihQu7EgYA+tZShBJt/11NFOTdjnl8PXy6X5FxBYagDcGYWt3K7LCCBhUlKlhg5OSvfAxVZ/B9y8NlNcrZ6jNaySn7JfM0kQjfHyB2Vm+XAwxX1GBW3b+JdJm0i11GXULW2guVBQzTovOOVznGR0I9qtXOr6bZxxyXeoWsCSLuRpZlUMOOQSeRyPzFJ04DUpmK/h64TTrX7HZ6Xaz296t0Le2UxRHAIwWCkk4P3to6DitHXNH/tqG1ikd0SO4WSQRzvHkDnA2kZOQOvTqOasSaxpkV19ml1G0ScAHymnUNg8DjOecj86fNqdjb3cdrcXtvFcS/wCrheVVd/opOTVcsLW8/wAv+GJvK9yjoGhDR4pmeWaSaaV3YvdSyrgsSvDk87cAnGTjqasaTYS2C3gmZG8+6kmXaTwrHjPvTLHXrHUNYvdNt5ka4siA6h1JORzgA5wCcH34rTpwUbJx9P6+4Ur3dwoooqyQqxpn/IbsP+u4/kar1Y0z/kN2H/XcfyNZVv4UvRmlL416nR+Mv+Rf/wC3iH/0MVyldX4y/wCRf/7eIf8A0MVylceX/wAN+p0Yz416BRRRXonGQXguGsZ1siguDGwiMhIUNjjOAeM1zcvgmOPQYrexurkX9vtlillvJjH5wO7cU3bQCc9F711dFRKEZO7KUmtjl7/w/f3momeKCwtZWkRjqFvLJHMFB+6VAxJ6ctjnpxS3Xh+9uNW+0QxWNo5uFka/tpHjmdAfuNGBhsjIyWI5zitS916xsNastMuJkS4vAxQM6jGOnBOck8D1wanTVtOlvmso9QtXu1+9AsymQf8AAc5qFGF9O5fNKxzlv4Q+yXUiLpeiXUb3DTC8uod0yhm3FSu35iM4Dbx244ra07Trmx1nU5m8pra8kWZWDnerBFUqVxjHy5zn8KsLrOlu0qpqVozQqXlAnUmNQcEnngA8c0611TT75Gayvra4VV3MYplcAc8nB6cH8jThGEfhFKUne5g3ng/7bf3VxLPOBcXcUhC306/ulwSuA2M7s49M8EVJf6XrM2oxJFFaXWlW6r5VvcXsiu7jndITG5bB6DPuc9r934l0m20a61OO+t7mC1U7zBMjfN2XOcZPYVoWl1De2kVzbSJLFKoZXRgwI+opKENo+o3KS1Y6AytAhuUSOUj50jcuoPsSBn8hUlFFbGQUUUUAdH4J/wCPTU/+v7/2jFXTVzPgn/j01P8A6/v/AGjFWnrWjf21DbR/2lqGn/Z7hZ91hP5TSbc/I5wcoc8ivm8R/Gl6nt0v4a9DTorN1TRv7TvLC4/tHULP7FN5vl2k/lpcf7Mgx8y+3FJqGjfb9W0++/tLULb7CzN9ntp9kVxkAYlXHzAY46VgaGnXPeM/+QTbf9fSfyauhrnvGf8AyCbb/r6T+TVrR/ix9UZ1fgfocvRRRX0x4gViXugHU/EAvL6WQW8EHl2yW9zJE4Zjl2JQg9AoAz61t0UnFS3Gm1scsPDE9rBf2lvFZ31hcTrPHbag7yckfOCx3EHcAwb5up470J4evl0uWCeCwvEluPNFhdSySRQrjG1JCCQc85245wAOtdTWVZeI9OvrzULeK5iDae+yUmVfQEt16AnGT3BrLkgtDTmk9TEk8IXMtvbyXItLyS3uXlSyu2aaFUdQDGHZS3BGQ238MVZbw7cpY27WdlpVnPBepdLb2oMUbAAjDOFyTyedo7DFbUGs6Zc2slzbalaSwRnDyxzqyofcg4FNOu6StvHcNqlkIZHKJIbhNrt6A5wT7UclNfh+Frfkg5psi13SDrVjDbNI8aieN5Nk7x5UHJGVIzx6/XqKi0PQU0mS5maSd5p5nY77uWVduflyHJG7aAM4z2zir0uq6fBZi8mvraO2JwJ3mUITnH3s461Vu/Eem2d7p1tJcxE6gxELiVduMZB68gnAGO5qrQUubr/mTeTjboRw6BHNY31pqY8yG6u3nAildDgkFeRgg5Hao9K8ODTPEN9fpLM8U8UaRiS7llYYzu3byc9Rjrjnpmt2imqcVbTb/hhcz1XcKKKKskK3vBH/ACFbX/sMP/6QmsGt7wR/yFbX/sMP/wCkJrgzD+EvU68J/E+R6tRRRXhnqBRRRQBT1UMbNNlx9mP2mD5yxGR5qZXj+8Pl991XKo6w0a2MZlhM6/arcBQxXB85MNx6HDY74xV6gAooooAKKKKACiiigAooooAKKKKAKVmUOpagFi2MHTc+Sd/yDn29Ku1UtTJ9vvd8yuu9NiBsmP5BwR2yeat0AFYvir/kExf9fcP/AKGK2qyfEltc3WkqtnA1xIk8UnloygsFcE4LEDp70AZVFRbNV/6AF9/39t//AI7Rs1X/AKAF9/39t/8A47QBLRUWzVf+gBff9/bf/wCO0bNV/wCgBff9/bf/AOO0Acj4p/5GH/t1j/8AQ5Kya6vVvDF/q16t0+m6xbuIxGRDNZ4IBJH3mPPzGqX/AAg19/z76/8A9/bH/GvWoYynTpqLvocFXDznNyRg0Vvf8INff8++v/8Af2x/xo/4Qa+/599f/wC/tj/jW31+l5mX1Sp5HK6ujy6JfJGrO7W8gVVGSTtPAFcibG8tb7Tru4fUraD+y47cNZWqzNG45ZWQxuRnjkAdMHtXrH/CDX3/AD76/wD9/bH/ABo/4Qa+/wCffX/+/tj/AI1nLF0ZO+v9X8/MuOHqJW0/q3+R5hHFLo8OlxW9reJbLHL/AKTLZC4uIyWyExGMIpznpjAAOKpaLZ3lomkzXNpeKsGp3Rk3W5LqHDBWKoMYJI5AwM+leuf8INff8++v/wDf2x/xo/4Qa+/599f/AO/tj/jS+t0b31/pp9/IfsKlraf1f/M8jMd3Ho8Glf2detcR6yJXIt2KLGbguH342kYI6E474qxJo15N4N1a3gtpEuJdRml2bAryp52TjcMHKjjPB4r1T/hBr7/n31//AL+2P+NH/CDX3/Pvr/8A39sf8aX1qjazv/Vv8ivY1PLe/wCf+Z5Z/ZlvqIu55jr94GtPJfzbWO3ONwICrsQsw6jgjqO+K0fDz3b6pcM6PNb+SgF5c2JtZ2YE/IQQNwAOchQBnHNehf8ACDX3/Pvr/wD39sf8aP8AhBr7/n31/wD7+2P+NVHGUU09f6v5+ZLw9Rq2n9fLyMGit7/hBr7/AJ99f/7+2P8AjR/wg19/z76//wB/bH/GtPr9LzM/qlTyMGo5v+PeT/dP8q6L/hBr7/n31/8A7+2P+NI3gS9ZSpt9fwRg/vbH/Gj6/S8w+qVPI3LL/kW7f/r0X/0CvPrH/kH2/wD1yX+Qr0aKz1KKyS1XQb/y0jEYJmt84Ax/z1rnY/AV5FEkaW+v7UUKMzWPQfjXDhMRCi5c3U6sRSlUSsYdFb3/AAg19/z76/8A9/bH/Gj/AIQa+/599f8A+/tj/jXd9fpeZy/VKnkYNYGq2txN4q06SBH2raXKGUKdqM2zbk9un6V3v/CDX3/Pvr//AH9sf8aP+EGvv+ffX/8Av7Y/41MsbRkrO5UcNUi76Hk2h6ZIsOk2t42tG6sHDGE20SQxuAQT5vljcpyejEnPPeo3+33R0VWsbmF7fUEee2g08xw24yejEfMPdTjkk4r13/hBr7/n31//AL+2P+NH/CDX3/Pvr/8A39sf8az+tUdN97/l5+RboVHfb+r/AOZ5U9vcwwTqbO6drXXPtjhIHO6IsTuTj5+D0XJ9qhvoLnU77UZk0678iW/sCnm27KZEVhubaRkAd89O+K9b/wCEGvv+ffX/APv7Y/40f8INff8APvr/AP39sf8AGiOKopJa6W/Cy/QPYVLt6f1qeZatZzweKp7uWXUYbWe0SKN7C1WfkFiyMpjcjOQc8A9+1V30yPTILU2i6vBKLLyB5tkt0kq5JEcioDtIJ7bRjjPp6r/wg19/z76//wB/bH/Gj/hBr7/n31//AL+2P+NH1qj5/wBfPzD2FTyOV0kznR7T7ZbR2s/krvgj+7GcfdHsKuVvf8INff8APvr/AP39sf8AGj/hBr7/AJ99f/7+2P8AjW31+l5mX1Sp5GDRW9/wg19/z76//wB/bH/Gj/hBr7/n31//AL+2P+NH1+l5h9UqeRR8Mf8AI2x/9eM3/oyGu5rn9K8K32lal9tTTtZuJBC0IWaaz2gMVJPysDn5B39a29mq/wDQAvv+/tv/APHa8vEVFVqOUTvowcIKLJaKi2ar/wBAC+/7+2//AMdo2ar/ANAC+/7+2/8A8drnNSWqerf8gW9/695P/QTU+zVf+gBff9/bf/47Udxbanc2ssEmg34WVCjETW+cEY/560Aedw/8e8f+6P5VJW6vgS9VQot9fwBgfvbH/Gl/4Qa+/wCffX/+/tj/AI17n1+l5nl/VKnkYNFb3/CDX3/Pvr//AH9sf8aP+EGvv+ffX/8Av7Y/40fX6XmH1Sp5Hn2u2d3deIbM2iOP9BuoxMFO1HYJtye3/wBasvQtOO7SI5zrX2iwX/USW0UcMLbCCPMEa71P+yzZ4Jr1X/hBr7/n31//AL+2P+NH/CDX3/Pvr/8A39sf8ay+t0ea+v8AXzNPYVOW2h5E/wBvujoqtY3ML2+oI89tBp5jhtxk9GI+Ye6nHJJxU0dvc21va77O6Y2GtSzzBLd2PluZMMuB84+YfdzivV/+EGvv+ffX/wDv7Y/40f8ACDX3/Pvr/wD39sf8aSxVJd/6t5+RToVH2/q/+Z5Lc21zqDanKNPuhFPq9nIiyQMC8a+XlsEZxwc+nfFWtSsZofE2oT3EmpRW13BGkTWNok4cDO6NgY3K8nOTgHNeof8ACDX3/Pvr/wD39sf8aP8AhBr7/n31/wD7+2P+NH1qja2v9K3cXsal+h5TcacunRxizj1WOZrBINlxZC7juFGdqSbAdpGeeVHPU4rsNPMx022N1CkExiXzIkOVRscqPYV03/CDX3/Pvr//AH9sf8aP+EGvv+ffX/8Av7Y/41ccbRi29df67kSw1SXYwaK3v+EGvv8An31//v7Y/wCNH/CDX3/Pvr//AH9sf8av6/S8yfqlTyMGlt/9dqH/AGB7v/2nW7/wg19/z76//wB/bH/GnL4N1G2gvDb6dq0889nLbJ9pmtAq7wOfkYHqo9awxGMp1KbirmtLDzhNSZ6pRRRXlHecNqsL3F14lhiGXkk2KPUm1iArg9Mur6z0zw5G2m3dosGlC1lul0aSe4SZQqmLG3KIcZ3EFWwOeOfR7u11ODXdTkh0i5uobidJI5IpYQCBDGp4ZweqntTNurf9C/ff9/bf/wCO0Lr/AF3/AMw/r8v8jzDwho1+kfhEahpl4htL7UnlFzb4MW4vsLYAUZzwRwc8VD4jQaX4N8RaZqek3U15PqLXMdyLRmikRpVMcnm42gquF2k7vl4GK9V26t/0L99/39t//jtYreDom1BrxvCupF2m89oft0fkNJnO8w+f5ZbPOduc81V/ev8A1umJrR/13/zOKvdDuPt3iSx1eTxEE1S6MkcWmWMM0d1GUUKPNaFvLZcY+d1xgEV1OkW01h8QtU86C68q7sLUQTtEzKxj8wMGcDaG+YcEjOeM1023Vv8AoX77/v7b/wDx2jbq3/Qv33/f23/+O0k7KwPU86/5e77/AK/rn/0c9Orbk8F6w1zPJFZagizTSTbSLU7S7FiM+d6mm/8ACFa3/wA+l/8A98W3/wAer2qeNpRgk+x5s8NUcm0Y1FbP/CFa3/z6X/8A3xbf/HqP+EK1v/n0v/8Avi2/+PVf16iT9VqGNVe/jabTrmKMZd4mVR6kiuh/4QrW/wDn0v8A/vi2/wDj1H/CFa3/AM+l/wD98W3/AMeqZYyjKLWo1hqqdzy+xuLuG106JrK5tlXTkgMyaa8kzSDgxklfkUYyCw2nOc8Uzw3pt0snhv7XZTp9ntrtZPOiI8pi64B4wMjOPXtXqf8AwhWt/wDPpf8A/fFt/wDHqP8AhCtb/wCfS/8A++Lb/wCPVH1qjzczf4F+wqWskeS6gPsXh99PvbGdrz+1FlM5gYo264BD+ZjbnaQMZz2xipbnS5ftWs2uoPrG2+uC6R2VrHIk6EDH7xozsIxj5mXGMivQ0+FNxHd/aE03Ut3meaIzJCYg+c7hH9o2A55yBnNaX/CFa3/z6X//AHxbf/HqiOIo9X/w2n+RTpVOiOH0eGS08TatHLHOBcCGSKRo2KsFjCnLgbd2R0zmt+tn/hCtb/59L/8A74tv/j1H/CFa3/z6X/8A3xbf/Hq2jjaKVrmTw1RmNRWz/wAIVrf/AD6X/wD3xbf/AB6j/hCtb/59L/8A74tv/j1V9eoi+q1DGqxpn/IbsP8AruP5GtH/AIQrW/8An0v/APvi2/8Aj1T2XhHWbXUILmSw1CUQvv2Ytlzwe/ne9Z1MZSlBxXUqGGqRkmy34y/5F/8A7eIf/QxXKV2us6Vqur6abU6LfwnejhxJbNgqwPTzfasP/hCtb/59L/8A74tv/j1c+ExFOlBqXc2xFGdSScTGorZ/4QrW/wDn0v8A/vi2/wDj1H/CFa3/AM+l/wD98W3/AMers+vUTn+q1DGorZ/4QrW/+fS//wC+Lb/49R/whWt/8+l//wB8W3/x6j69RD6rUOJ1hJY/EujXYhlkijE0TGONn2s4XbnaDgcdTwO9c6/2+6Oiq1jcwvb6gjz20GnmOG3GT0Yj5h7qccknFesf8IVrf/Ppf/8AfFt/8eo/4QrW/wDn0v8A/vi2/wDj1YvFUW73e9/y/wAjRUKiVrf1r/meXw6Zcp4XgLWMreTq0lzcW/lfPLH5zHIU8n+FgO+BiqmoQya1f+I10q1njaS3tGZWi8p5gHYt8rjqVGPmHOOmK9Tvvh5qeo23kXlhqDpuDDH2dWUjoQwnBBHqDTbH4calpyOtrY6iDIdzvI0EjufdmnJP4mp+s0dr6f8AAsV7GpvbX/g3PN7nTW1Gx1Sa1fWbq5fTZLdTeWqW6nPITaI0Zj6HBHJ5rsNPmE+nQSKsiAoPlljaNh9VYAiui/4QrW/+fS//AO+Lb/49R/whWt/8+l//AN8W3/x6tY4yjHr/AFr/AJmcsPUl/Xp/kY1FbP8AwhWt/wDPpf8A/fFt/wDHqP8AhCtb/wCfS/8A++Lb/wCPVf16iT9VqGNRWz/whWt/8+l//wB8W3/x6j/hCtb/AOfS/wD++Lb/AOPUfXqIfVaha8E/8emp/wDX9/7RirpqxtC0TVdFtp4/7H1C4aebzmYvbLg7VXAHm+iitPbq3/Qv33/f23/+O14tWSlUcl1PSppxgkyaiodurf8AQv33/f23/wDjtG3Vv+hfvv8Av7b/APx2syyaue8Z/wDIJtv+vpP5NW3t1b/oX77/AL+2/wD8drP1rSNV1ixWD+xtQgKSCRXEls3Iz2833rSnJRmpPoyJpyi0jjKK2f8AhCtb/wCfS/8A++Lb/wCPUf8ACFa3/wA+l/8A98W3/wAer2vr1E836rUMaitn/hCtb/59L/8A74tv/j1H/CFa3/z6X/8A3xbf/HqPr1EPqtQxq426W7tJ9eVNOa4aa8hnTfatKnl7Y1LqMYZlIJ2g54r0v/hCtb/59L//AL4tv/j1H/CFa3/z6X//AHxbf/HqiWMovqVHD1F0PIdQtL2/udckFtfXEdzZQLG0tmY/MIkOQF2g8Ds3PXtiuhu4/sHihry5spri1mshBEYLcy+WwYllIUHAYFeTx8vJrvf+EK1v/n0v/wDvi2/+PVQvfhhfahcCe4sdUEoXYXhlhiLL12kpcDcPY5FZfWaSWj/rX/Mv2NR7/wBbf5Hl+l2lwNP0C/jN9bWcUNwhNlAszws0nHyFHyMAjIHH0rTt9P8AsM2jXMEOovbrfTSSG4gHmL5iMA2yNRtUsfQYzzivSYPAerWtukFvYXsUUahURI7UBQOw/fU//hCtb/59L/8A74tv/j1VHE0I9WJ0ar6f1r/mY1FbP/CFa3/z6X//AHxbf/HqP+EK1v8A59L/AP74tv8A49W316iZ/VahjUVs/wDCFa3/AM+l/wD98W3/AMeo/wCEK1v/AJ9L/wD74tv/AI9R9eoh9VqGNW94I/5Ctr/2GH/9ITUX/CFa3/z6X/8A3xbf/Hq2fDXh3U9M1Wx83TrsRi9e5nnmaAKgNs8YGFkY9dvY9a5MXiadWnyx7nRh6M6c7yPQqKKK8w7QooooAqamZRaJ5E6QN9ogyzttBXzV3L9WGVA7kgVbqhrJhFjH9pV2T7XbABDg7vPTafoGwT7Zq/QAUUUUAFFFFABRRRQAUUUUAFFFFAFGzMZ1PUQisHEibyTkE7BjHpxV6qpt7oXU0kd2BHImFjaIHY2MZznke360w2+ofZ1UX8YlDEmT7PwR6Y3frQBdoqt5N59qL/a08nGBH5PIOOu7Prz0qP7PqH2bb9vj83fnzPs/G3HTG7170AXaKrCG8+0xsbtDEFAePyeWOOuc8c1GtvqAgdWv4zISNr/Z8BR3GN3NAF2iqvkXvmwn7YmxVAkXyPvnuQc8frTRBf7Jgb6MsxHln7P9znv83PHHagC5RVQwX37nF7H8v+t/cff57fNxx9aPIvt0x+2x7Wz5Q8j/AFfPGeeePpQBboqk1vqBiiC30YdSfMb7P9/043cU8w3nnSsLtPLZSI08nlD2JOefpxQBaoqkbfUPs6qL+MShiTJ9n4I9Mbv1qTybz7UX+1p5OMCPyeQcdd2fXnpQBZoql9n1D7Nt+3x+bvz5n2fjbjpjd696kEN59pjY3aGIKA8fk8scdc545oAs0VSW31AQOrX8ZkJG1/s+Ao7jG7mn+Re+bCftibFUCRfI++e5Bzx+tAFqiqYgv9kwN9GWYjyz9n+5z3+bnjjtSmC+/c4vY/l/1v7j7/Pb5uOPrQBboqp5F9umP22Pa2fKHkf6vnjPPPH0prW+oGKILfRh1J8xvs/3/TjdxQBdoqqYbzzpWF2nlspEaeTyh7EnPP04pht9Q+zqov4xKGJMn2fgj0xu/WgC7RVbybz7UX+1p5OMCPyeQcdd2fXnpUf2fUPs237fH5u/PmfZ+NuOmN3r3oAu0VVEN79oVjeJ5QUBo/I5Jx1znjnmoxb6iIGU6hGZCwKv9m4A7jG6gC9RVQQX3nRMb2PYoHmL5H3z3IOeM/jTRb6gI5Ab+Ms2NjfZ/u888buaALtFU/Iv8w/6dHhf9Z/o/wB/nt83HHHejyL/ADN/p0eG/wBX/o/3Oe/zc8cdqALlFUjb6gY4wL+MMud7fZ/vc8cbuKcYL7zpWF7HsYHy18j7h7EnPOPwoAt0VRNvqJgVRqEYkDEs/wBm4I7DG6pDDe/aGYXieUVIWPyOQcdc5555oAtUVR+z6j9n2/2hH5u7O/7NxjHTG79akEN79oVjeJ5QUBo/I5Jx1znjnmgC1RVEW+oiBlOoRmQsCr/ZuAO4xup4gvvOiY3sexQPMXyPvnuQc8Z/GgC3RVIW+oCOQG/jLNjY32f7vPPG7ml8i/zD/p0eF/1n+j/f57fNxxx3oAuUVT8i/wAzf6dHhv8AV/6P9znv83PHHakNvqBjjAv4wy53t9n+9zxxu4oAu0VUMF950rC9j2MD5a+R9w9iTnnH4Uw2+omBVGoRiQMSz/ZuCOwxuoAvUVVMN79oZheJ5RUhY/I5Bx1znnnmo/s+o/Z9v9oR+buzv+zcYx0xu/WgC9RVUQ3v2hWN4nlBQGj8jknHXOeOeajFvqIgZTqEZkLAq/2bgDuMbqAL1FVBBfedExvY9igeYvkffPcg54z+NNFvqAjkBv4yzY2N9n+7zzxu5oAu0VT8i/zD/p0eF/1n+j/f57fNxxx3o8i/zN/p0eG/1f8Ao/3Oe/zc8cdqALlFUjb6gY4wL+MMud7fZ/vc8cbuKcYL7zpWF7HsYHy18j7h7EnPOPwoAt0VRNvqJgVRqEYkDEs/2bgjsMbqkMN79oZheJ5RUhY/I5Bx1znnnmgC1RVH7PqP2fb/AGhH5u7O/wCzcYx0xu/WpBDe/aFY3ieUFAaPyOScdc5455oAtUVRFvqIgZTqEZkLAq/2bgDuMbqeIL7zomN7HsUDzF8j757kHPGfxoAt0VSFvqAjkBv4yzY2N9n+7zzxu5pfIv8AMP8Ap0eF/wBZ/o/3+e3zcccd6ALlFU/Iv8zf6dHhv9X/AKP9znv83PHHakNvqBjjAv4wy53t9n+9zxxu4oAu0VUMF950rC9j2MD5a+R9w9iTnnH4Uw2+omBVGoRiQMSz/ZuCOwxuoAvUVVMN79oZheJ5RUhY/I5Bx1znnnmo/s+o/Z9v9oR+buzv+zcYx0xu/WgC9RVUQ3v2hWN4nlBQGj8jknHXOeOeajFvqIgZTqEZkLAq/wBm4A7jG6gC9RVQQX3nRMb2PYoHmL5H3z3IOeM/jTRb6gI5Ab+Ms2NjfZ/u888buaALtFU/Iv8AMP8Ap0eF/wBZ/o/3+e3zcccd6PIv8zf6dHhv9X/o/wBznv8ANzxx2oAuUVSNvqBjjAv4wy53t9n+9zxxu4pxgvvOlYXsexgfLXyPuHsSc84/CgC3RVE2+omBVGoRiQMSz/ZuCOwxuqQw3v2hmF4nlFSFj8jkHHXOeeeaALVFUfs+o/Z9v9oR+buzv+zcYx0xu/WpBDe/aFY3ieUFAaPyOScdc5455oAtUVRFvqIgZTqEZkLAq/2bgDuMbqeIL7zomN7HsUDzF8j757kHPGfxoAt0VSFvqAjkBv4yzY2N9n+7zzxu5pfIv8w/6dHhf9Z/o/3+e3zcccd6ALlFU/Iv8zf6dHhv9X/o/wBznv8ANzxx2pDb6gY4wL+MMud7fZ/vc8cbuKALtFVDBfedKwvY9jA+WvkfcPYk55x+FMNvqJgVRqEYkDEs/wBm4I7DG6gC9RVUw3v2hmF4nlFSFj8jkHHXOeeeaj+z6j9n2/2hH5u7O/7NxjHTG79aAL1FVRDe/aFY3ieUFAaPyOScdc5455qMW+oiBlOoRmQsCr/ZuAO4xuoAvUVUEF950TG9j2KB5i+R989yDnjP400W+oCOQG/jLNjY32f7vPPG7mgC7RVPyL/MP+nR4X/Wf6P9/nt83HHHejyL/M3+nR4b/V/6P9znv83PHHagC5RVI2+oGOMC/jDLne32f73PHG7inGC+86Vhex7GB8tfI+4exJzzj8KALdFUTb6iYFUahGJAxLP9m4I7DG6pDDe/aGYXieUVIWPyOQcdc5555oAtUVR+z6j9n2/2hH5u7O/7NxjHTG79akEN79oVjeJ5QUBo/I5Jx1znjnmgC1RVEW+oiBlOoRmQsCr/AGbgDuMbqeIL7zomN7HsUDzF8j757kHPGfxoAt0VSFvqAjkBv4yzY2N9n+7zzxu5pfIv8w/6dHhf9Z/o/wB/nt83HHHegC5RVPyL/M3+nR4b/V/6P9znv83PHHakNvqBjjAv4wy53t9n+9zxxu4oAu0VUMF950rC9j2MD5a+R9w9iTnnH4Uw2+omBVGoRiQMSz/ZuCOwxuoAvUVVMN79oZheJ5RUhY/I5Bx1znnnmo/s+o/Z9v8AaEfm7s7/ALNxjHTG79aAL1FVRDe/aFY3ieUFAaPyOScdc5455qMW+oiBlOoRmQsCr/ZuAO4xuoAvUVUEF950TG9j2KB5i+R989yDnjP400W+oCOQG/jLNjY32f7vPPG7mgC7RVPyL/MP+nR4X/Wf6P8Af57fNxxx3o8i/wAzf6dHhv8AV/6P9znv83PHHagC5RVI2+oGOMC/jDLne32f73PHG7inGC+86Vhex7GB8tfI+4exJzzj8KALdFUTb6iYFUahGJAxLP8AZuCOwxuqQw3v2hmF4nlFSFj8jkHHXOeeeaALVFUfs+o/Z9v9oR+buzv+zcYx0xu/WpBDe/aFY3ieUFAaPyOScdc5455oAtUVRFvqIgZTqEZkLAq/2bgDuMbqeIL7zomN7HsUDzF8j757kHPGfxoAXUftH2VPsjKsnnw5LYxs81d4577d3v6c1arOl065uovLvbxZEE0Uy7IdpHlyBwOpzkqBWjQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAH/2Q=="}}},{"cell_type":"code","source":"# timeseriesgenerator-like class, except it using target from the last timestep insteed of last+1\nclass sample_generator(keras.utils.Sequence):\n    def __init__(self, x_set, y_set, batch_size, length):\n        self.x, self.y = x_set, y_set\n        self.batch_size = batch_size\n        self.length = length\n        self.size = len(x_set)\n\n    def __len__(self):\n        return int(np.ceil(len(self.x) / float(self.batch_size)))\n\n    def __getitem__(self, idx):\n        batch_x=[]\n        batch_y=[]\n        for i in range(self.batch_size):\n            start_ind = self.batch_size*idx + i\n            end_ind = start_ind + self.length \n            if end_ind <= self.size:\n                batch_x.append(self.x[start_ind : end_ind])\n                batch_y.append(self.y[end_ind -1])\n\n        return np.array(batch_x), np.array(batch_y)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:09.937562Z","iopub.execute_input":"2021-11-13T11:35:09.938814Z","iopub.status.idle":"2021-11-13T11:35:09.956140Z","shell.execute_reply.started":"2021-11-13T11:35:09.938762Z","shell.execute_reply":"2021-11-13T11:35:09.954796Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#last 10% of the data are used as validation set\nX_train, X_test = train[:-len(train)//10], train[-len(train)//10:]\ny_train, y_test = targets[:-len(train)//10], targets[-len(train)//10:]","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:09.958687Z","iopub.execute_input":"2021-11-13T11:35:09.960152Z","iopub.status.idle":"2021-11-13T11:35:09.973955Z","shell.execute_reply.started":"2021-11-13T11:35:09.960085Z","shell.execute_reply":"2021-11-13T11:35:09.972199Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"BATCH_SIZE=2**10\ntrain_generator = sample_generator(X_train, y_train, length=15, batch_size=BATCH_SIZE)\nval_generator = sample_generator(X_test, y_test, length=15, batch_size=BATCH_SIZE)\n\nprint(f'Sample shape: {train_generator[0][0].shape}')\nprint(f'Target shape: {train_generator[0][1].shape}')","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:09.975843Z","iopub.execute_input":"2021-11-13T11:35:09.977438Z","iopub.status.idle":"2021-11-13T11:35:10.025167Z","shell.execute_reply.started":"2021-11-13T11:35:09.977373Z","shell.execute_reply":"2021-11-13T11:35:10.023687Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Training","metadata":{}},{"cell_type":"code","source":"#https://github.com/tensorflow/tensorflow/issues/37495\ndef MaxCorrelation(y_true,y_pred):\n    \"\"\"Goal is to maximize correlation between y_pred, y_true. Same as minimizing the negative.\"\"\"\n    mask = tf.math.not_equal(y_true, 0.)\n    y_true_masked = tf.boolean_mask(y_true, mask)\n    y_pred_masked = tf.boolean_mask(y_pred, mask)\n    return -tf.math.abs(tfp.stats.correlation(y_true_masked,y_pred_masked, sample_axis=None, event_axis=None))\n\ndef Correlation(y_true,y_pred):\n    return tf.math.abs(tfp.stats.correlation(y_pred,y_true, sample_axis=None, event_axis=None))\n\ndef masked_mse(y_true, y_pred):\n    mask = tf.math.not_equal(y_true, 0.)\n    y_true_masked = tf.boolean_mask(y_true, mask)\n    y_pred_masked = tf.boolean_mask(y_pred, mask)\n    return tf.keras.losses.mean_squared_error(y_true = y_true_masked, y_pred = y_pred_masked)\n\ndef masked_mae(y_true, y_pred):\n    mask = tf.math.not_equal(y_true, 0.)\n    y_true_masked = tf.boolean_mask(y_true, mask)\n    y_pred_masked = tf.boolean_mask(y_pred, mask)\n    return tf.keras.losses.mean_absolute_error(y_true = y_true_masked, y_pred = y_pred_masked)\n\ndef masked_cosine(y_true, y_pred):\n    mask = tf.math.not_equal(y_true, 0.)\n    y_true_masked = tf.boolean_mask(y_true, mask)\n    y_pred_masked = tf.boolean_mask(y_pred, mask)\n    return tf.keras.losses.cosine_similarity(y_true_masked, y_pred_masked)\n\ndef get_model(n_assets=14):  \n    x_input = keras.Input(shape=(train_generator[0][0].shape[1], n_assets, train_generator[0][0].shape[-1]))\n\n    branch_outputs = []\n        \n    for i in range(n_assets):\n            # Slicing the ith asset:\n        a = layers.Lambda(lambda x: x[:,:, i])(x_input)\n        a = layers.Masking(mask_value=0.,)(a)\n        #a = layers.BatchNormalization()(a)\n        #a = layers.Bidirectional(layers.LSTM(units=24, return_sequences=True))(a)\n        a = layers.GRU(units=64, return_sequences=True)(a)\n        a = layers.GRU(units=32)(a)\n        branch_outputs.append(a)\n    \n    x = layers.Concatenate()(branch_outputs)\n    x = layers.Dense(units=128)(x)\n    #x = layers.Dropout(0.2)(x)\n    out = layers.Dense(units=n_assets)(x)\n    \n    model = keras.Model(inputs=x_input, outputs=out)\n    model.compile(optimizer = tf.keras.optimizers.Adam(learning_rate=1e-3), \n                  #loss = 'mse',\n                  #loss = 'cosine_similarity',\n                  loss = masked_cosine,\n                  metrics=[Correlation]\n                 )\n    \n    return model \n    \nmodel=get_model()\n#model.summary()","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:10.027872Z","iopub.execute_input":"2021-11-13T11:35:10.028269Z","iopub.status.idle":"2021-11-13T11:35:41.188106Z","shell.execute_reply.started":"2021-11-13T11:35:10.028163Z","shell.execute_reply":"2021-11-13T11:35:41.187122Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gc.collect()","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:41.190213Z","iopub.execute_input":"2021-11-13T11:35:41.191506Z","iopub.status.idle":"2021-11-13T11:35:41.459958Z","shell.execute_reply.started":"2021-11-13T11:35:41.191461Z","shell.execute_reply":"2021-11-13T11:35:41.458723Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#example with 3 assets for visibility\ntf.keras.utils.plot_model(get_model(n_assets=3), show_shapes=True)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:41.468073Z","iopub.execute_input":"2021-11-13T11:35:41.469055Z","iopub.status.idle":"2021-11-13T11:35:48.136998Z","shell.execute_reply.started":"2021-11-13T11:35:41.468864Z","shell.execute_reply":"2021-11-13T11:35:48.135705Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"tf.random.set_seed(0)\n\nestop = keras.callbacks.EarlyStopping(monitor='val_loss', patience=10, verbose=0, mode='min',restore_best_weights=True)\nplateau = keras.callbacks.ReduceLROnPlateau(monitor=\"val_loss\", factor=0.9, patience=3, verbose=1, min_lr=1e-8)\nscheduler = keras.optimizers.schedules.ExponentialDecay(1e-3, (0.5*len(X_train)/BATCH_SIZE), 1e-3)\n\nlr = keras.callbacks.LearningRateScheduler(scheduler, verbose = 1)\n    \nhistory = model.fit(train_generator, validation_data = (val_generator),\n              epochs = 30, callbacks = [lr])","metadata":{"execution":{"iopub.status.busy":"2021-11-13T11:35:48.138743Z","iopub.execute_input":"2021-11-13T11:35:48.139047Z","iopub.status.idle":"2021-11-13T13:15:34.963979Z","shell.execute_reply.started":"2021-11-13T11:35:48.139002Z","shell.execute_reply":"2021-11-13T13:15:34.962855Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gc.collect()","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:15:34.966225Z","iopub.execute_input":"2021-11-13T13:15:34.966849Z","iopub.status.idle":"2021-11-13T13:15:36.121454Z","shell.execute_reply.started":"2021-11-13T13:15:34.966787Z","shell.execute_reply":"2021-11-13T13:15:36.120412Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(1,2, figsize=(16,8))\n\nhistories = pd.DataFrame(history.history)\n\nepochs = list(range(1,len(histories)+1))\nloss = histories['loss']\nval_loss = histories['val_loss']\nCorrelation = histories['Correlation']\nval_Correlation = histories['val_Correlation']\n\nax[0].plot(epochs, loss, label = 'Train Loss')\nax[0].plot(epochs, val_loss, label = 'Val Loss')\nax[0].set_title('Losses')\nax[0].set_xlabel('Epoch')\nax[0].legend(loc='upper right')\n\nax[1].plot(epochs, Correlation, label = 'Train Correlation')\nax[1].plot(epochs, val_Correlation, label = 'Val Correlation')\nax[1].set_title('Correlations')\nax[1].set_xlabel('Epoch')\nax[1].legend(loc='upper right')\n\nfig.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2021-11-13T13:15:36.122977Z","iopub.execute_input":"2021-11-13T13:15:36.123903Z","iopub.status.idle":"2021-11-13T13:15:36.689780Z","shell.execute_reply.started":"2021-11-13T13:15:36.123853Z","shell.execute_reply":"2021-11-13T13:15:36.688690Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gc.collect()","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:15:36.691633Z","iopub.execute_input":"2021-11-13T13:15:36.692111Z","iopub.status.idle":"2021-11-13T13:15:37.892743Z","shell.execute_reply.started":"2021-11-13T13:15:36.692065Z","shell.execute_reply":"2021-11-13T13:15:37.891704Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### The correlation coefficients by asset for the validation data","metadata":{}},{"cell_type":"code","source":"predictions = model.predict(val_generator)\npredictions.shape","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:15:37.894790Z","iopub.execute_input":"2021-11-13T13:15:37.895450Z","iopub.status.idle":"2021-11-13T13:16:18.560487Z","shell.execute_reply.started":"2021-11-13T13:15:37.895402Z","shell.execute_reply":"2021-11-13T13:16:18.559570Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('Asset:    Corr. coef.')\nprint('---------------------')\nfor i in range(14):\n    #drop first 14 values in the y_test, since they are absent in val_generator labels\n    y_true = np.squeeze(y_test[14:,i])\n    y_pred = np.squeeze(predictions[:,i])\n    \n    #get non-zero targets (assuming that zeros indicates 'non-real' rows)\n    real_target_ind = np.argwhere(y_true!=0)\n    asset_id = list(assets_order.keys())[i]\n    asset_name = assets[assets.Asset_ID == asset_id]['Asset_Name'].item()\n    print(f\"{asset_name}: {np.corrcoef(y_pred[real_target_ind].flatten(), y_true[real_target_ind].flatten())[0,1]:.4f}\") \n\ndel predictions","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:17:45.863292Z","iopub.execute_input":"2021-11-13T13:17:45.863732Z","iopub.status.idle":"2021-11-13T13:17:45.932402Z","shell.execute_reply.started":"2021-11-13T13:17:45.863657Z","shell.execute_reply":"2021-11-13T13:17:45.931332Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Submission","metadata":{}},{"cell_type":"code","source":"#test_df = pd.read_csv('../input/g-research-crypto-forecasting/example_test.csv')\n#sample_prediction_df = pd.read_csv('../input/g-research-crypto-forecasting/example_sample_submission.csv')\n\n#test_df = test_df.iloc[2:12]\n#sample_prediction_df=sample_prediction_df[sample_prediction_df.row_id.isin(test_df.row_id)]\n\n#scale_features = test_df.columns.drop(['Asset_ID','group_num','row_id'])#test_df.head(3)\n#test_df.head()","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:16:18.640860Z","iopub.execute_input":"2021-11-13T13:16:18.641285Z","iopub.status.idle":"2021-11-13T13:16:18.646770Z","shell.execute_reply.started":"2021-11-13T13:16:18.641223Z","shell.execute_reply":"2021-11-13T13:16:18.645449Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#placeholder for first 15 samples\nsup=pd.read_csv('../input/g-research-crypto-forecasting/supplemental_train.csv')[:15*14]\nplaceholder=add_features(sup)\nplaceholder[scale_features] = RS.transform(placeholder[scale_features])\nplaceholder['asset_order'] = placeholder.Asset_ID.map(assets_order) \ntest_sample = np.array(placeholder[features])\ntest_sample = test_sample.reshape(-1,14,test_sample.shape[-1])\ntest_sample = np.expand_dims(test_sample, axis=0)\ntest_sample.shape","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:16:18.648539Z","iopub.execute_input":"2021-11-13T13:16:18.649116Z","iopub.status.idle":"2021-11-13T13:16:21.506619Z","shell.execute_reply.started":"2021-11-13T13:16:18.649074Z","shell.execute_reply":"2021-11-13T13:16:21.505418Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# for test gap filling\nexample = pd.read_csv('../input/g-research-crypto-forecasting/example_test.csv')[:14]\nexample['asset_order'] = example.Asset_ID.map(assets_order) \nexample = example[['Asset_ID','asset_order']]\nexample","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:16:21.508858Z","iopub.execute_input":"2021-11-13T13:16:21.510028Z","iopub.status.idle":"2021-11-13T13:16:21.540291Z","shell.execute_reply.started":"2021-11-13T13:16:21.509973Z","shell.execute_reply":"2021-11-13T13:16:21.539158Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"env = gresearch_crypto.make_env()\niter_test = env.iter_test()\n\nfor (test_df, sample_prediction_df) in iter_test:\n        \n    test_df['VWAP'] =np.nan_to_num(test_df.VWAP, posinf=VWAP_max, neginf=VWAP_min)\n    test_df=add_features(test_df)\n    test_df[scale_features] = RS.transform(test_df[scale_features])\n    \n    test_data = test_df.merge(example, how='outer', on='Asset_ID').sort_values('asset_order')\n        \n    test = np.array(test_data[features].fillna(0))\n    test = test.reshape(-1,1,14,test.shape[-1])\n        #get test sample, shape (1, 15, 14, test.shape[-1])\n    test_sample = np.hstack([test_sample, test])[:,-15:]\n    y_pred = model.predict(test_sample).squeeze().reshape(-1, 1).squeeze()\n    \n    test_data['Target'] = y_pred\n        \n    #test_df = test_df.merge(test_data[['row_id', 'Target']], how='left', on='row_id')\n    for _, row in test_df.iterrows(): \n        try:\n            sample_prediction_df.loc[sample_prediction_df['row_id'] == row['row_id'], 'Target'] = test_data.loc[test_data['row_id'] == row['row_id'], 'Target'].item()\n        except:\n            sample_prediction_df.loc[sample_prediction_df['row_id'] == row['row_id'], 'Target'] = 0\n       \n    #sample_prediction_df['Target'] = test_df['Target']\n    #display(test_df)\n    #display(test_data)\n    #display(sample_prediction_df)    \n    env.predict(sample_prediction_df)","metadata":{"execution":{"iopub.status.busy":"2021-11-13T13:16:21.541840Z","iopub.execute_input":"2021-11-13T13:16:21.542961Z","iopub.status.idle":"2021-11-13T13:16:58.072145Z","shell.execute_reply.started":"2021-11-13T13:16:21.542917Z","shell.execute_reply":"2021-11-13T13:16:58.071059Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}