From 273dc04fbd3a1d652cb9c822da975481896f4435 Mon Sep 17 00:00:00 2001 From: Aadyot Bhatnagar Date: Fri, 17 Dec 2021 13:03:26 -0800 Subject: [PATCH] Re-Architecture of Layered Models & Ensembles (#47) * Fix some AutoSarima bugs. * Harden models to granularities like MS * Correct import path. * Fix save/load behavior for layered models. * Implement generic LayeredModel class. * More helpful model config __init__ docstrings. Create a new metaclass which allows model configs to 1. Inherit params from superclass __init__, and have those params included in the signature for the subclass __init__. 2. Inherit param docstrings from superclass __init__, which means that users don't have to rewrite lots of docstrings for superclass params. * Various minor fixes. Also suppresses output from Prophet while the model is training. * Update gson from 2.8.6 to 2.8.9. Addresses security risk. * Order docstring params according to fn signature. * Fix failing tests. * Allow None models in LayeredModel. * Make Merlion default models subclass LayeredModel. * Allow bubbling of callable attributes in Layers. * Simplify AutoSARIMA implementation. * Minor fixes to default model. * Move LayeredModel to a new file. * Add dynamic inheritance to LayeredModel's. * Allow AutoSarima to use a SarimaDetector. * Add more docstrings & reduce code duplication. * Add auto-seasonality to AutoSarima. * Update AutoSARIMA example. * Update tests to avoid segfault. * Actually use approx_iter in AutoSARIMA * Implement __reduce__ for Config & LayeredModel. This ensures that everything is usable by multiprocessing code. The use of _original_cls for LayeredModel.__reduce__ ensures that the right class object is used when attempting to initialize the object, not a dynamically defined subclass. * change p value to 0.1 * change default regression method to c in KPSS test * Better defined periodicity strategies. * Add distinct AutoProphet model. * Add AutoETS model. * Various updates to make serialization work. - Override setstate/getstate for ensembles - Use a more refined method to check for unused kwargs in config init * Remove depth field & fix _save_state() bug. * Add comment on _original_cls * Add example for ModelConfigMeta. * Various AutoProphet bug fixes. * Fix from_dict() implementations * Simplify LayeredModel._save_state() * Move ensemble.models from model to config. This mirrors the changes to LayeredModel, and it greatly simplifies a number of implementation details. * Fix Sphinx errors & handle docstring suffixes. * Add p-value to SeasonalityConfig. * Move AutoETS & AutoProphet to models.automl * Update benchmark_forecast.json * Implement bubbling for LayeredModel.__setattr__ * Add robustness to None models in serialization. Co-authored-by: Chenghao Liu --- benchmark_forecast.py | 3 +- conf/benchmark_anomaly.json | 24 ++ conf/benchmark_forecast.json | 15 +- docs/source/merlion.models.automl.rst | 28 +- docs/source/merlion.models.rst | 19 +- .../1_AutoSARIMA_forecasting_tutorial.ipynb | 222 ++++++++------ merlion/models/anomaly/base.py | 20 +- merlion/models/anomaly/change_point/bocpd.py | 3 +- merlion/models/anomaly/dbl.py | 3 +- merlion/models/anomaly/forecast_based/base.py | 4 +- merlion/models/anomaly/forecast_based/mses.py | 8 +- merlion/models/anomaly/random_cut_forest.py | 2 +- merlion/models/anomaly/zms.py | 11 +- merlion/models/automl/autoets.py | 32 ++ merlion/models/automl/autoprophet.py | 44 +++ merlion/models/automl/autosarima.py | 199 +++++------- .../models/automl/{layer_mixin.py => base.py} | 41 ++- .../models/automl/forecasting_layer_base.py | 121 -------- merlion/models/automl/seasonality.py | 182 +++++++++++ merlion/models/automl/seasonality_mixin.py | 62 ---- merlion/models/base.py | 134 +++----- merlion/models/defaults.py | 111 ++----- merlion/models/ensemble/MoE_forecast.py | 115 +++---- merlion/models/ensemble/anomaly.py | 2 +- merlion/models/ensemble/base.py | 268 +++++++--------- merlion/models/ensemble/forecast.py | 18 +- merlion/models/factory.py | 31 +- merlion/models/forecast/arima.py | 15 +- merlion/models/forecast/baggingtrees.py | 9 +- merlion/models/forecast/base.py | 36 +-- merlion/models/forecast/boostingtrees.py | 12 +- merlion/models/forecast/ets.py | 34 +-- merlion/models/forecast/lstm.py | 9 +- merlion/models/forecast/prophet.py | 71 +++-- merlion/models/forecast/sarima.py | 85 +----- merlion/models/forecast/smoother.py | 7 +- merlion/models/layers.py | 285 ++++++++++++++++++ merlion/resources/gson-2.8.6.jar | Bin 240255 -> 0 bytes merlion/resources/gson-2.8.9.jar | Bin 0 -> 258075 bytes merlion/utils/autosarima_utils.py | 27 +- merlion/utils/misc.py | 106 ++++++- setup.py | 4 +- tests/anomaly/forecast_based/test_prophet.py | 13 +- tests/forecast/test_autosarima.py | 68 +++-- tests/forecast/test_ets.py | 29 +- tests/forecast/test_forecast_ensemble.py | 22 +- 46 files changed, 1424 insertions(+), 1130 deletions(-) create mode 100644 merlion/models/automl/autoets.py create mode 100644 merlion/models/automl/autoprophet.py rename merlion/models/automl/{layer_mixin.py => base.py} (61%) delete mode 100644 merlion/models/automl/forecasting_layer_base.py create mode 100644 merlion/models/automl/seasonality.py delete mode 100644 merlion/models/automl/seasonality_mixin.py create mode 100644 merlion/models/layers.py delete mode 100644 merlion/resources/gson-2.8.6.jar create mode 100644 merlion/resources/gson-2.8.9.jar diff --git a/benchmark_forecast.py b/benchmark_forecast.py index 274f92a8b..d70500f5c 100644 --- a/benchmark_forecast.py +++ b/benchmark_forecast.py @@ -25,8 +25,7 @@ from merlion.models.ensemble.forecast import ForecasterEnsembleConfig, ForecasterEnsemble from merlion.models.factory import ModelFactory from merlion.models.forecast.base import ForecasterBase -from merlion.transform.resample import TemporalResample -from merlion.utils.time_series import granularity_str_to_seconds +from merlion.transform.resample import TemporalResample, granularity_str_to_seconds from merlion.utils import TimeSeries, UnivariateTimeSeries from merlion.utils.resample import get_gcd_timedelta diff --git a/conf/benchmark_anomaly.json b/conf/benchmark_anomaly.json index add238c64..e35b0ff77 100644 --- a/conf/benchmark_anomaly.json +++ b/conf/benchmark_anomaly.json @@ -89,6 +89,23 @@ } }, + "AutoETSDetector": {"alias": "AutoETS"}, + "AutoETS": { + "config": { + "default": { + "model": {"name": "ETSDetector"}, + "damped_trend": true, + "transform": {"name": "TemporalResample", "granularity": "1h"} + }, + "IOpsCompetition": { + "transform": {"name": "TemporalResample", "granularity": "5min"} + }, + "CASP": { + "transform": {"name": "TemporalResample", "granularity": "5min"} + } + } + }, + "Prophet": {"alias": "ProphetDetector"}, "ProphetDetector": { "config": { @@ -96,6 +113,13 @@ } }, + "AutoProphetDetector": {"alias": "AutoProphet"}, + "AutoProphet": { + "config": { + "default": {"model": {"name": "ProphetDetector"}} + } + }, + "StatThreshold": { "config": { "default": {} diff --git a/conf/benchmark_forecast.json b/conf/benchmark_forecast.json index 53c7d373f..b27c6fbb8 100644 --- a/conf/benchmark_forecast.json +++ b/conf/benchmark_forecast.json @@ -37,6 +37,14 @@ } }, + "AutoETS": { + "config": { + "default": { + "damped_trend": true + } + } + }, + "MSES": { "config": { "default": { @@ -48,18 +56,15 @@ "Prophet": { "config": { "default": { - "uncertainty_samples": 0, - "add_seasonality": null + "uncertainty_samples": 0 } } }, "AutoProphet": { - "model_type": "Prophet", "config": { "default": { - "uncertainty_samples": 0, - "add_seasonality": "auto" + "uncertainty_samples": 0 } } }, diff --git a/docs/source/merlion.models.automl.rst b/docs/source/merlion.models.automl.rst index 794349d72..35ab127e4 100644 --- a/docs/source/merlion.models.automl.rst +++ b/docs/source/merlion.models.automl.rst @@ -1,3 +1,4 @@ + merlion.models.automl package ============================== @@ -7,26 +8,19 @@ merlion.models.automl package :show-inheritance: .. autosummary:: - layer_mixin - forecasting_layer_base + base + seasonality + autoets + autoprophet autosarima - seasonality_mixin Submodules ---------- -merlion.models.automl.layer_mixin module ---------------------------------------------------- - -.. automodule:: merlion.models.automl.layer_mixin - :members: - :undoc-members: - :show-inheritance: - -merlion.models.automl.forecasting_layer_base module ---------------------------------------------------- +merlion.models.automl.base module +--------------------------------- -.. automodule:: merlion.models.automl.forecasting_layer_base +.. automodule:: merlion.models.automl.base :members: :undoc-members: :show-inheritance: @@ -40,10 +34,10 @@ merlion.models.automl.autosarima module :show-inheritance: -merlion.models.automl.seasonality_mixin module ----------------------------------------------- +merlion.models.automl.seasonality module +---------------------------------------- -.. automodule:: merlion.models.automl.seasonality_mixin +.. automodule:: merlion.models.automl.seasonality :members: :undoc-members: :show-inheritance: diff --git a/docs/source/merlion.models.rst b/docs/source/merlion.models.rst index 64570c229..f6663d372 100644 --- a/docs/source/merlion.models.rst +++ b/docs/source/merlion.models.rst @@ -58,8 +58,9 @@ Finally, we support ensembles of models in :py:mod:`merlion.models.ensemble`. :show-inheritance: .. autosummary:: - base factory + base + layers defaults anomaly anomaly.change_point @@ -86,6 +87,15 @@ Subpackages Submodules ---------- +merlion.models.factory module +----------------------------- + +.. automodule:: merlion.models.factory + :members: + :undoc-members: + :show-inheritance: + + merlion.models.base module -------------------------- @@ -94,15 +104,14 @@ merlion.models.base module :undoc-members: :show-inheritance: -merlion.models.factory module ------------------------------ +merlion.models.layers module +---------------------------- -.. automodule:: merlion.models.factory +.. automodule:: merlion.models.layers :members: :undoc-members: :show-inheritance: - merlion.models.defaults module ------------------------------ diff --git a/examples/advanced/1_AutoSARIMA_forecasting_tutorial.ipynb b/examples/advanced/1_AutoSARIMA_forecasting_tutorial.ipynb index 8a387be73..e535f04b7 100644 --- a/examples/advanced/1_AutoSARIMA_forecasting_tutorial.ipynb +++ b/examples/advanced/1_AutoSARIMA_forecasting_tutorial.ipynb @@ -33,19 +33,12 @@ "name": "stderr", "output_type": "stream", "text": [ - "plotly not installed, so plotly visualizations will not work.\n", - "INFO:ts_datasets.forecast.m4:M4 Hourly dataset cannot be found from /Users/chenghao.liu/Documents/research-project/Merlion_backup/public_merlion/Merlion/data/M4.\n", - "M4 Hourly dataset will be downloaded from https://github.com/Mcompetitions/M4-methods/raw/master/Dataset/{}.csv.\n", - "\n", - "INFO:ts_datasets.forecast.m4:Downloading https://github.com/Mcompetitions/M4-methods/raw/master/Dataset/M4-info.csv\n", - "INFO:ts_datasets.forecast.m4:Downloading https://github.com/Mcompetitions/M4-methods/raw/master/Dataset/Train/Hourly-train.csv\n", - "INFO:ts_datasets.forecast.m4:Downloading https://github.com/Mcompetitions/M4-methods/raw/master/Dataset/Test/Hourly-test.csv\n", - "100%|██████████| 414/414 [00:05<00:00, 72.11it/s] \n" + "100%|██████████| 414/414 [00:00<00:00, 650.30it/s]\n" ] }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -73,12 +66,11 @@ "from merlion.utils.time_series import TimeSeries\n", "from merlion.evaluate.forecast import ForecastMetric\n", "from merlion.models.automl.autosarima import AutoSarima, AutoSarimaConfig\n", - "from merlion.models.automl.seasonality_mixin import SeasonalityLayer\n", "from merlion.models.forecast.sarima import Sarima\n", "\n", "from ts_datasets.forecast import M4\n", "\n", - "logging.basicConfig(level=logging.DEBUG)\n", + "logging.basicConfig(level=logging.INFO)\n", "\n", "time_series, metadata = M4(\"Hourly\")[0]\n", "train_data = TimeSeries.from_pd(time_series[metadata.trainval])\n", @@ -114,35 +106,47 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:merlion.models.forecast.base:Automatically detect the periodicity is 24\n", - "INFO:merlion.models.forecast.sarima:Seasonal difference order is 1\n", - "INFO:merlion.models.forecast.sarima:Difference order is 0\n", - "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", - "INFO:merlion.models.automl.autosarima:Difference order is 0\n", + "INFO:merlion.models.automl.seasonality:Automatically detect the periodicity is 24\n", "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", "INFO:merlion.models.automl.autosarima:Difference order is 0\n", "INFO:merlion.models.automl.autosarima:Fitting models using approximations(approx_iter is 1) to speed things up\n", - "INFO:merlion.models.automl.autosarima:Best model: SARIMA(2,0,2)(0,1,1)[24] without constant\n" + "INFO:merlion.models.automl.autosarima:Best model: SARIMA(2,0,3)(1,1,1)[24] without constant\n" ] - }, + } + ], + "source": [ + "# Specify the configuration of AutoSarima with approximation\n", + "#\n", + "# p, q, P, Q refer to the AR, MA, seasonal AR, and seasonal MA params, so\n", + "# auto_pqPQ=True (default) means select them automatically\n", + "#\n", + "# d is the difference order, and D is the seasonal difference order, so\n", + "# auto_d=True (default) and auto_D=True (default) means select them automatically\n", + "#\n", + "# auto_seasonality=True (default) means to automatically select the seasonality\n", + "config1 = AutoSarimaConfig(auto_pqPQ=True, auto_d=True, auto_D=True, auto_seasonality=True,\n", + " approximation=True, maxiter=5)\n", + "model1 = AutoSarima(config1)\n", + "\n", + "# Model training\n", + "train_pred, train_err = model1.train(\n", + " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Full AutoSarima with approximation sMAPE is 3.4491\n" + "Full AutoSarima with approximation sMAPE is 3.4736\n" ] } ], "source": [ - "# Specify the configuration of AutoSarima with approximation\n", - "config1 = AutoSarimaConfig(max_forecast_steps=len(train_data), order=(\"auto\", \"auto\", \"auto\"),\n", - " seasonal_order=(\"auto\", \"auto\", \"auto\", \"auto\"), approximation=True, maxiter=5)\n", - "model1 = SeasonalityLayer(model = AutoSarima(model = Sarima(config1)))\n", - "\n", - "# Model training\n", - "train_pred, train_err = model1.train(\n", - " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})\n", - "\n", "# Model forecasting\n", "forecast1, stderr1 = model1.forecast(len(test_data))\n", "\n", @@ -153,12 +157,12 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, "metadata": {}, "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAGoCAYAAABbtxOxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAADlHElEQVR4nOzdd3gc1fXw8e9s35VWvRf3buNuXLCxKS70hBAgoZg0Ut4kkAbphfCjJQRCugMBQkISSAglAWNqbIN7xb0XSVav23dm7vuHbGFZq96l83kePw/aKXs1rGbO3nvuuZpSSiGEEEIIIYQAwNLbDRBCCCGEEKIvkQBZCCGEEEKIs0iALIQQQgghxFkkQBZCCCGEEOIsEiALIYQQQghxFgmQhRBCCCGEOIsEyEII0YJFixbx+OOPA/DXv/6VJUuW9HKLRHfYv38/U6dOxev18thjj/V2c4QQvUwCZCFEl1u7di3z5s0jMTGRlJQULrjgAjZt2tTbzeq0m266iVWrVjX8rGkahw4datOxx44dQ9M0dF3vruZ1iWHDhuF2u4mPj2/4V1RU1NvNatXZX2Q64qGHHuKiiy6irq6Or371q206/7vvvkteXl7Dz8899xzz5s3D4/GwaNGiDrdFCNH7JEAWQnSp2tparrzySr7yla9QWVlJYWEhP/rRj3A6nb3dtEGvrcH5K6+8gs/na/iXk5PTLe/Tlxw/fpyJEyd26hwpKSnceeedfPvb3+6iVgkheosEyEKILnXgwAEAPvGJT2C1WnG73SxZsoTJkyc37POnP/2J8ePHk5yczNKlSzl+/HjDtjvuuIP8/HwSEhKYMWMGa9asadi2ceNGZs6cSUJCApmZmXz9619v2Pbyyy8zceJEkpKSWLRoEXv37m3YNmzYMH7+858zefJkEhMTueGGGwiFQgBUVVVx5ZVXkp6eTnJyMldeeSUFBQUxf7ennnqK+fPnA3DhhRcCMGXKFOLj4/nHP/7BpEmTeOWVVxr2j0ajpKWlsW3bthav2caNG5k7dy5JSUlkZ2fz5S9/mUgkAsD/+3//j2984xuN9r/66qt55JFHACgqKuJjH/sY6enpDB8+vFF6wI9//GOuu+46br75ZhISEnjqqadabEdLwuEwd955Jzk5OeTk5HDnnXcSDoeBD3tSH3zwQbKysvjUpz6FaZo88MADjBw5ktTUVK6//noqKysbzndmlCEpKYn8/PyGtv33v/9l2rRpJCQkkJ+fz49//OOGY0KhEDfffDOpqakkJSUxa9YsSkpK+N73vseaNWv48pe/THx8PF/+8pdj/g7NfUYuvvhi3nnnnYbjz3yG2+vSSy/l+uuvb/cXCiFEH6SEEKIL1dTUqJSUFHXrrbeqV199VVVWVjba/uKLL6qRI0eqPXv2qGg0qn7605+quXPnNmx/5plnVHl5uYpGo+rnP/+5yszMVMFgUCml1Jw5c9Sf//xnpZRSdXV1at26dUoppfbv3688Ho9atWqVikQi6sEHH1QjR45U4XBYKaXU0KFD1axZs1RhYaGqqKhQ48aNU7/73e+UUkqVl5erf/7zn8rv96va2lp13XXXqWuuuaahPQsXLlR//OMflVJKPfnkk+qCCy5o2AaogwcPNvz84IMPquuvv77R7zpp0iSllFJHjx5VgIpGo02u2ebNm9W6detUNBpVR48eVePGjVOPPPKIUkqpDRs2qOzsbGUYhlJKqbKyMuV2u1VxcbEyDENNnz5d/eQnP1HhcFgdPnxYDR8+XK1cuVIppdSPfvQjZbPZ1L///W9lGIYKBALq/vvvV1dccUWz//+GDh2q3njjjSav/+AHP1CzZ89WJSUlqrS0VM2dO1d9//vfV0op9c477yir1aruuusuFQqFVCAQUI8++qiaPXu2OnnypAqFQur2229XN954o1JKqWPHjqn4+Hj17LPPqkgkosrLy9W2bdsazrVz505lGIbasWOHysjIUP/+97+VUkr9/ve/V1deeaXy+/1K13W1efNmVVNT0+T/UyytfUZaOz7W9nfeeUfl5uY22fePf/yjWrhwYbPnEkL0fRIgCyG63J49e9Ty5ctVbm6uslqt6qqrrlLFxcVKKaWWLVumHn/88YZ9DcNQbrdbHTt2LOa5kpKS1Pbt25VSSi1YsED98Ic/VGVlZY32ueeee9THP/7xRufMyclR77zzjlKqPuh75plnGrZ/61vfUp///Odjvt+2bdtUUlJSw8/tCZALCwtVfHx8Q9D2sY99TD344INKqZYD5HM98sgj6iMf+UjDz+PGjVOrVq1SSin1q1/9Sl122WVKKaXWr1+v8vPzGx173333qdtuu00pVR8gL1iwoNX3O9vQoUNVXFycSkxMVImJiQ1fFkaMGKH++9//Nuy3cuVKNXToUKVUfaBot9sbvsicafObb77Z8HNRUZGy2WwqGo2q++67r9Hv15I77rhD3XnnnUoppZ544gk1d+5ctWPHjib7tRbgtvYZaUuA7Ha7G65LYmKiiouLkwBZiAFKUiyEEF1u/PjxPPXUUxQUFLBr1y6Kioq48847gfpczzvuuIOkpCSSkpJISUlBKUVhYSEAP//5zxk/fjyJiYkkJSVRU1NDeXk5AE888QQHDhxg3LhxzJo1i//85z9AfZrB0KFDG97fYrGQn5/fcE6ArKyshv/2eDz4fD4AAoEAn//85xk6dCgJCQlceOGFVFdXYxhGu3/vnJwcLrjgAv71r39RXV3Na6+9xk033dTqcQcOHODKK68kKyuLhIQEvvvd7zb8zgDLly/nL3/5CwB/+ctfuOWWWxquZVFRUcO1TEpK4r777qOkpKTh2Pz8/Hb/Hi+++CLV1dVUV1fz4osvAk2v8dChQxtN3ktPT8flcjX8fPz4cT760Y82tGv8+PFYrVZKSko4efIkI0eOjPneGzZs4KKLLiI9PZ3ExER+//vfN1yLW265haVLl3LjjTeSk5PDXXfdRTQabdPv1JbPSGsee+yxhutSXV3d8PkTQgw8EiALIbrVuHHjuO2229i1axdQH7D94Q9/aBRoBINB5s2bx5o1a3jooYd47rnnqKqqorq6msTERJRSAIwePZq//e1vlJaWcvfdd3Pdddfh9/vJyclplMeslOLkyZPk5ua22r6HH36Y/fv3s2HDBmpra1m9enXDOTriTDD7/PPPM3fu3Da14Ytf/CLjxo3j4MGD1NbWct999zV6/5tvvpmXXnqJHTt2sHfvXj7ykY8A9ddy+PDhja5lXV0dr776asOxmqZ16Pc417nX+MSJE41ybc99n/z8fF577bVGbQuFQuTm5pKfn8/hw4djvs8nP/lJrr76ak6ePElNTQ1f+MIXGq6F3W7nRz/6EXv27OH999/nP//5D3/+85/b9Ht25jMihBh8JEAWQnSpffv28fDDDzdMdDt58iR/+9vfmDNnDgBf+MIXuP/++9m9ezcANTU1PP/88wDU1dVhs9lIT09H13XuueceamtrG879l7/8hbKyMiwWC0lJSUB9T+D111/Pf//7X9566y2i0SgPP/wwTqeTefPmtdreuro63G43SUlJVFZW8pOf/KTNv2tmZiZHjhxp9NpHPvIRtm7dyi9/+UtuvfXWJseEw2FCoVDDP9M0qaurIyEhgfj4ePbt28fvfve7Rsfk5eUxa9YsbrnlFj72sY/hdrsBOP/88/F6vTz44IMEg0EMw2DXrl3dUlLvE5/4BPfeey9lZWWUl5dzzz33cPPNNze7/xe+8AW+973vNQSlZWVlvPTSS0B9ubw333yT5557Dl3XqaioYPv27UD9/4+UlBRcLhcbN27k2WefbTjnO++8wwcffIBhGCQkJGC327FY6h9jsf5fnK0zn5G2MgyDUCiEruuYpkkoFGpzD7cQom+RAFkI0aW8Xi8bNmxg9uzZxMXFMWfOHCZNmsTDDz8MwEc/+lHuvvtubrzxRhISEpg0aRKvvfYaAEuXLmXZsmWMGTOGoUOH4nK5GqUIrFy5kokTJxIfH88dd9zB3//+d9xuN2PHjuUvf/kLX/nKV0hLS+OVV17hlVdeweFwtNreO++8k2AwSFpaGnPmzGHZsmVt/l1//OMfs3z5cpKSknjuuecAcLvdfOxjH+Po0aNce+21TY6Jj4/H7XY3/Hv77bf5+c9/zrPPPovX6+Vzn/scN9xwQ5Pjli9fzgcffNCQXgFgtVr5z3/+w/bt2xk+fDhpaWl89rOfpaamptk233fffVx22WVt/h3P+P73v8/MmTOZPHky5513HtOnT+f73/9+s/vfcccdXH311SxZsgSv18ucOXPYsGEDAEOGDOHVV1/l4YcfJiUlhalTp7Jjxw4Afvvb3/LDH/4Qr9fLPffcw/XXX99wzuLiYq677joSEhIYP348CxcubLged9xxB//85z9JTk6OWce4M5+RtnrmmWdwu9188YtfZM2aNbjdbj73uc912fmFED1HUx0dRxRCCBHTPffcw4EDBxryhrvC6tWrufnmmzl+/HiXpU0IIYSIzdbbDRBCiIGksrKSJ554gmeeeabLzhmNRvnlL3/JZz/7WQmOhRCiB0iKhRBCdJE//vGP5Ofnc9lllzUsJNJZe/fuJSkpiVOnTjVUAhFCCNG9JMVCCCGEEEKIs0gPshBCCCGEEGfp1znIKSkpHSqCfzZd17HZ+vVl6HFyzdpHrlf7yTVrH7le7SPXq/3kmrWPXK/2661rVlhY2GhhpjP69f+9/Px8Vq5c2alzlJWVkZ6e3kUtGhzkmrWPXK/2k2vWPnK92keuV/vJNWsfuV7t11vX7Kqrror5uqRYCCGEEEIIcRYJkIUQQgghhDiLBMhCCCGEEEKcpV/nIAshhBBCDCaGYVBXV4dhGL3dlC6llKKioqLbzm+1WvF6vVit1jbtLwGyEEIIIUQ/UVdXR3JyMsnJyQNqZc3urGKhlKKqqoqqqiqSkpLadIykWAghhBBC9BOGYQy44Li7aZpGcnJyu3rdJUAWQgghhOhHJDhuv/ZeMwmQhRBCCCGEOIsEyEIIIYQQos1SUlIa/fznP/+ZO+64A4A1a9Ywe/ZsPB4PL7zwQm80r0tIgCyEEEIIIbpEfn4+jz/+ODfeeGNvN6VTpIqFEEIIIYToEsOGDQPAYunffbASIAshhBBCiDYLBoPMmjWr4eeqqiquuOKKXmxR15MAWQghhBCin3I6nV1+znA43OJ2t9vNpk2bGn7+85//zJYtW7q8Hb1JAmQhhBBCiH6qtWBWdEz/ThARQgghhOhGSimUUr3dDNHDJEAWQgghhDhL1DCpDOocrAjy3ok61p2so7A2TFg3e7tpfd7mzZsZMWIE//rXv/h//+//MXXq1N5uUodIioUQQgghBr1AxKAmrFPi06kM6gA4rOCxWzGV4kBFiP0VQTLj7OR4nSS5rIN2RbvKyspGP996663ceuutAMycOZMjR470RrO6lATIQohuEzFMIoYizm4ZtA8SIUTfZJiKuohBRSBKiS9K2DBRSsNj10h1nxv8aqS6LSilqA4alPh8uGwWhiQ6SYuz47LJgPxAIwGyEKJLBaMmNSGdYl+UylAUAJfVwpAkJ2keeZAIIXqPqRRl/iil/igVAR1TKWwWjTi7lXiHtdXjNU3D67TixUrEMDlYGeJgRZD0ODs5CfW9yhbpDBgQJEAWQnSKUopA1KQqqHPKF6EubIIGHptGqsuGpmn1D5KKEAfKg6R67OQlOEhy2bBa5EEihOg5x6pCHKkKE2e3dDqYdVgtDb3KtWGD0lM+nKc7A5Bc5X6vW7tyfvnLXzJp0iQmTpzIo48+CtTnrSxevJjRo0ezePFiqqqqgPqH7Fe/+lVGjRrF5MmT2bp1a3c2TQjRCaZS1IZ1jlYFWXeyjg0FdRysDKIUpHlspLlteOwfDlHWP0hspLptBCIGO4oDvHeilsOVQerChswQF0J0u8qgzpHqMGkeG3GOruvp1TSNeIe1YYTscGWITcUhthT5KKoLE4gaXfI+omd1W4C8a9cu/vjHP7Jx40Z27NjBf/7zHw4dOsQDDzzAJZdcwsGDB7nkkkt44IEHAHjttdc4ePAgBw8eZMWKFXzxi1/srqYJITqoJqSzvzzI2hN1bC70cbImgsOqkeaxk+puPX1C0zTiHFbSPDa8DiuFtRE2FtaxsbD+QSIzxIUQ3SGsm+wuDZDYhYFxLHarRorbRqrbimEq9peHWHeyjvUFdZysCeOLSIdAf9FtKRZ79+5l9uzZeDweABYuXMgLL7zASy+9xLvvvgvA8uXLWbRoEQ8++CAvvfQSt956K5qmMWfOHKqrqzl16hTZ2dnd1UQhRDvUhHQ2F/lxWsFrt2J1tp6v1xKrRSPJVX8LCusm+8tDKBUkI86OPWKQZJjYrZKvLIToHKUU+8uDKBROW+fuW+3hslkaOg0ihsnhqhBmBTitGlleO2keO16n5Cz3Vd0WIE+aNInvfe97VFRU4Ha7efXVV5k5cyYlJSUNQW9WVhYlJSUAFBYWkp+f33B8Xl4ehYWFTQLkFStWsGLFCgDKysooKyvrVDvPpHiItpNr1j4D4XpFTcX20hBWDWw2C7XBrn8PC/UPshN+RUV1HXvLIyQ6LaR7LHgdVjx2CZabMxA+Yz1Jrlf79edrVlQX5XBNlFS3lapuuHfFUltb1+Q1DbACIVOxp1qhm2C1KNLdNtLcVuIdFmxtmJehlELX9a5vdDt4vV4mTpzY8PM//vEPhg4d2qlzGkbnU1F+/etf8+lPf7qhczbWe7Q1buy2AHn8+PHcfffdLFmyhLi4OKZOnYrV2vibm6Zp7S79dPvtt3P77bcDMGXKFNLT0zvd1q44x2Aj16x9+vv12lMWwBPvJMXd/fN6U6B+Ak1iEiFdURo1KQ0rXLqFbK+dFLf0usTS3z9jPU2uV/v1x2tWG9Ypr/MzPNPa45OCk5OSW93HMOsnOZ+MKmymxsQ0N6lx9haPqaiowGbr3RoLbrebzZs3t/s4XddbbHtnf6/f/va33HzzzSQkJMTcbrVaSU1NbdO5urVL5jOf+Qxbtmxh9erVJCcnM2bMGDIzMzl16hQAp06dIiMjA4Dc3FxOnjzZcGxBQQG5ubnd2TwhRBuU+iKcqouQ7Oq+oUlTKYpLSti8ZQuvrXyNEydOYiqF2245PenPjsOqcbImwpYiH2uO17Kn1E9FIErEkLxlIURTUcNkV0mAeLulz1bMsVrqy8aleWx4bBrbiv0cqw5h9sM85R07drBgwQJmzJjBxz/+8YZRh8WLF/ONb3yDuXPn8qtf/YqtW7dy6aWXMmfOHK644oqGmPDw4cMsW7aMmTNnMnv2bA4fPozP52Pp0qXMnj2b6dOn8/LLLwPg9/u55pprmDlzJtOmTeP555/n17/+NUVFRSxZsoQlS5Z0+vfp1q8gpaWlZGRkcOLECV544QXWr1/P0aNHefrpp/n2t7/N008/zTXXXAPA1Vdfza9//WtuvPFGNmzYQGJiouQfC9HLAlGD3WVBkpy2LlnoIxAMUlBQQEHBSU6eLDj93wUUFhaQ4E0gLy+P1NRU3nrzLY4dO8akSZOYMnUKU6ZMZdiwYQ05y4apqAoalPj9KCDZZWNkiosEp1SuFELUpyEcrAwRMRQp7p7LO+6M+mo/GocrQ9SGDMalu3G0Mg/jG6/sZeeppukcnTE528vDV41vcZ9gMMisWbMAGDZsGM8//zyf/vSneeSRR7jwwgv5yU9+wr333svDDz8MQCQSYd26dUSjUS699FL++c9/kp6ezvPPP8+PfvQjVqxYwac//WnuuusurrnmGkKhEKZp4nA4eP7550lISKC8vJwFCxZw1VVXsWrVKnJycnjppZcAqKmpITExkccee4xVq1aRlpbW6evQrU+Tj33sY1RUVGC32/nNb35DUlIS3/72t7n++ut54oknGDp0KM899xwAl19+Oa+++iqjRo3C4/Hw5JNPdmfThBCtMJViX1kQh0XDbm1/cKxQrFmzhh3bd1BQUMDJgpP4fX5ycnPJz88jLy+PuXPnkJ+fT25uHh63u+HYquoqlIKdO3awY+dOXnnlFXw+H5MnT2HKlMlMmTKV/Pw8NKwNdZg3FfoZl+Yix+uQVfuEGORK/FGKaiOke/rXl2arRSPdY6cqpLO50Md5mXF4Ozkhuju43W42bdrU8HNNTQ01NTVceOGFANx888188pOfbNj+8Y9/HIADBw6we/duLr/8cqA+JzgrK4u6ujqKiooaOk1dLhcA0WiUH/zgB6xduxaLxUJRURElJSVMnDiRu+++m+9+97tcfvnlzJ8/v8t/x2795KxZs6bJa6mpqbz11ltNXtc0jd/85jfd2RwhRDucqAlTHdZJc7ecDxdLeXk5v3zsMcrLy1m6dClz580jPz+f9PT0NucOpyQns2jRIhYtWgRAaVkZO3fsYPuOHTz33PPous6UyZOZMnUKs2adT0pyCvvKg9SEDEanuqQChhCDlD9isKcsQLKra0a+ekOyy0YwarKxoI4JGR6yvY6Y+7XW09tXxMXFAfU9+xMmTGD16tWNttfVxe4F/9vf/kZ5eTnr16/HbrczZswYQqEQY8aMYf369axcuZIf//jHXHTRRXzve9/r0jb3r69WQogeUR3SOVwRavekPIVi1ao3+NMTT3DV1Vfzox/+sMsmk2Skp3PppZdy6aWXolAUFxezY/sOtm/fzl+e+QtPPPEEaW4npf4otWGD8zI9xLVh6VghxMBhmIo9pQFcVkuHRr76Ere9/nfYXRqgNqwzMsXdpioXvSExMZGkpCTWrl3L/PnzefbZZ1mwYEGT/caMGUNZWRnr169nzpw5RKNRDh48yIQJE8jNzeWll17immuuIRwOYxgGNTU1pKenY7fbeffddzl+/DgARUVFpKSk8MlPfpLExMSGrIP4+Hjq6ur6foqFEKL/iRgmu0sCeJ3tm/VdWlbGY7/8JVVVVdx3//2MHDGi29qooZGdlU32smyWLVvGvf/3f7z08svccP31pLht+CMGGwvrmJjuISM+ds+LEGLgOVoVoi5iktbPUiuaY7NopHtsFNXVf/GfmBG7fFlf8MQTT/DlL3+ZQCDA8OHD+eMf/9hkH4fDwd///ne+/vWvU1NTg67rfOUrX2HChAk8/vjj3HHHHdxzzz3Y7XaeffZZPvGJT3Dttdcyffp0ZsyYwdixY4H6xei+853vYLFYsNvt/OpXvwLqi0NcddVV5OTksGrVqk79Pprqx0u6TJkyhZUrV3bqHGVlZf2ydE1vkmvWPv3peiml2FsWpCygt7lqhULx+srX+dOTf+IjH/kI13/8+k73GldVV7WpRNIZJ06e5Jvf/AZPPPEE3ngvAFFDURXSGZrkZESyq8/OYu8K/ekz1hfI9Wq//nDNKvxRthX7SfPYer0MZHvvYW1RFzbQlWJCfJRxpwPFgaS1EnBd4cCBA03KvF111VUxS9YNjK9YQoguUeKPcqou0ubel5LSUn75y0epranlwQcfZPiw4d3cwtiG5OczZ84c/vn8P/nUpz4F1C/5mu6xcbI2Qk1YZ0K6B49dUi6EGIiCUZPdZQGSXAO3RrrXaSVimIQNRcRQOPp5CklfJ7NYhBAABE5PbElqw8QWheK1la/xla98mcnnTebRRx/tteD4jJtuuplXX3uNisrKhtc0TSPNbSMcVWwq9FHhj/ZiC4UQ3cFUiv3lATRotSxaf+ewWrBqENZNgrpJv00B6AcG9idJCNEmhqnYUx7AaWl9YktJaSnf/e73ePXV13jowYe48cYbe31VJ4DMjAwuvfRS/vbss022eZ1W4uwWthX7OVIZ7JdF+IUQsZ2siVAe1El09f59qKdYLfX37WBUguS2am9GsQTIQgiOV9cXpm+p3qZC8eprr/KVr3yZKVOm8MgjjzBs2LCea2Qb3HDDDfxv9WpOFZ9qss1hrV+V72h1mJ3FfkK6rMAnRH9XHdI5VBkkdRAFx7qyUFNdhUUDQym5l7WBUoqqqiqs1ran2Q2eT5QQIqaqoM6R6jBprZR0+9e/XuDtt9/iZz/7OUOHDOmh1rVPUmIi11x9Nc888wx3feuuJtstWn0R/uqwztYiH9Oy43HbpZ9AiP7oTEk3r6N9FXf6u2rTAWWVVJSXA2AosGkdW9CpLzEMo10BbHtZrVa8Xm+b95cAWYhBLGKY7C4NkOBoeWKLqRT//e9/+fa3v91ng+Mzrv3Yx/j0pz7F0WNHm82LTnLaqAsbfFDiZ1p2nCwqIkQ/VBXUCemKNM/gmnyrNAtVysWZ3ApTKcoDOhMy3OR4nb3buE4oKytrUmGiN8lTQYhBSinFgYoghlK4bC3fCnbu3InL5WLMmNE91LqO87jd3HDjDTz15FMt7ud1WglETfaWSU6yEP3R0eoQcTIChEXTSHHb2FsWpCqo93ZzBgz5ZAkxSBX7opT4oiS3IXdv5cqVLF26FI3+MYR3+eVXcOToUXbv2dPifiluG+XBKIcrQz3UMiFEV6gN69SGDUmROs1m0UhwWNlZ4icQMXq7OQOCfLKEGIRMpThSGSKxhUl5Z9TW1bF50yYuvvjiHmhZ13A6HNx00008+eSTqFbmeKe6bByvDlNQE+6h1gkhOqugJoKzn+fcdjWnzYLDorGjJEDEkIl7nSUBshCDUFVQJ2yYbaoZ+vZbbzHr/PNJaMfkhr7g0ksvpbq6mi2bt7S4n6ZppLpt7C8PSp1kIfqBkG5S4ovgdQyu3OO2iHNYiRome0oDGKakjnWGBMhCDEInasK4W8k7hvrSbitXrmTZsmU90KquZbNauW35cv705J9azTG2WjQSXVZ2lgbwyfCkEH1aiS+KpmmtLmjUFapCOpuLfPij/adHNslloypocKgy2O7av+JDUsVCiEHGFzGoDOqke+yt7rt//wEi0QiTJ5/XofdSqF7NW75g/gX847nnWL16NYsWLmxxX4fVgtum2HHKz4zc+FYnLgohep5hKo7XhEloQ3pYR5goDpSH2FToY1ORjwMV9fMTXDaNhcMSuWxUEmPTXH1+PkaK20pBbQSP3Up+Yv+tbNGbJEAWYpApqo1gb2PN0JUrV7JkSfsm51WHdDYV+lhfUMeWU34y4uwsGZnEJSMS2zQhsCtpaHzqU7fxm1//hvkXXNDqin8eu5XaiM6uUj9Ts+KxDaLaqkL0BxWBKLphYuvCALkuYrClyM/m00FxTbh+FGlcmotbp6QxMtnF+wU+3j1aw+uHqhme5GTZqGQuGZFAfB9N89A0jRRXfeqY26aRFufo7Sb1OxIgCzGIRAyTwroIya7Wb+rBUJC1a9fyhz/8odV9C+sirDtZx/qCOnaXBlFAqtvGwmGJHK8O8fjWUp7cVsqs3HiWjkpiVk7PBZ/Tp00nPT2dVW+s4vLLLm91/wSHjcqQzv7yIBPS3T0yjCuEaJ1SimPV4U4HpYr682wo8LG5yMee0iAm4HVYmJETz/m58UzPjiPprC/0s/O83D4jk3eO1rDyUDW/21zME9tKuHBIAsvGJDEx3d3nepWtFo0kl5UPSgPMzLG2uFKqaEoCZCEGkTJ/FAUtLgpyxv/+t5pJkyaRmpLSZJuJYn95iPUFdaw7WceJmggAw5Kc3Dgpjbn58YxO/XAY8nh1mFWHq3nraA3rC3wkuaxcMjyRpaOSGNIDw3+3fepT3PvTn3LxxZfgcrb+fikuG8W+CG6bxogUd7e3TwjRutqwQV3EJN3TsdDFRPH09jLeOlJDeaC+XvDIZBfXT0rj/Lx4xqa6sLZwb4yzW7hyTDJXjknmYEWQ1w5V887RGt48WsOQRAeXjU7m4uGJbaoO1FMcVgseG+wsltSx9pIAWYhBwlSK49VhEhxtu0GuXLmSG2+8seHnqKnYWuRjXYGPDQU+qkI6FuC8TA+Xj05mTp6XrPjYec1Dk5x8bkYmn5qWwaYiH68fqubFfZX8a28l49JcLB2VxIVDE7ut6P+4sWMZM3YM/3nlFa677ro2HZPqtnG4KozbbiXbK8OTQvS2gtowrk6UdnthTyX/2FXB+bnx3Dw5npm58aS5W5+LEcvoVDejU918bkYm/ztWy8pDVfxhcwl/2lrCBUMSWJRlYWaiwtYHRqDcdgu1YUkday8JkIUYJOqXZTWJd7T+QDh27BjlZWXMnDkTAEMpvvfWCXaWBHDZNGblxDM338us3Ph2lVqyWTTm5nmZm+elKqTz1pEaVh2u5pfri/ndphIWDElg8chE8ruhU3n58tu461vfYtlllxEfF9fq/pbT5d/2lAVx2Swku+V2KURvCUQNSnxRUjv4d7i/PMiT20qZlx/PDxbmdVk6hNtmYdmoJJaNSuJIdYjXDlbz9pEadh33E95Wx+zceOYN8TI9Kx6XrfcC0wSnpI61l9zxhRgk2lraDWDl66+zeMkSbNb64Pfvu8rZWRLgCzMzuXx0Mo4uKNCf7LJx3YRUPjYhhf3lIVYdrubdo7W8dbSGoc4Io3ODzMnzMj0nDk8XDAsOHTKEWeefz7/+9S+W33prm46pX53Kws4SP7Ny4vH00Qk5Qgx0xXURNI0OBXb+qMn9awtJdtu4c05Ot+UKj0hy8f9mZfGZaRmsO3SKTRWw/mQdbx6pwWnVmJETx7x8L+fnerutCkdLUlw2iusiJDqt5Elli1ZJgCzEIOCPGFSF9DYNJ0aiUd5++y0e++VjAOwuC/LXHeUsGpbANeOSu/zhoqExLs3NuDQ3t8/IZH2Bj+3HillbUP9gsVlgcmYcs3O9zMmLJ7OZNI62uOWWW/jyl7/M1VdfRXJScpuOcdos6KZie7GfGTnxOCWHT4gepZuKk7URkpztD1kUil9tOEWJL8rPlgztkcDUZbMwNcvDReOS0U3FztIA75+oZf1JH++f9GHhFJOzPMzJ8zI330tmXMfvae2V7LZxuCpEepxd7mWtkABZiEGgqC7S5ly4999/nxEjRpKVlYUvavDQ2gIy4m18eXZWt8/SdtksLBqWwJQkgy8nJrGnNMj6gjo2FNTxu83F/G4zDE9yMjuvPlgek+bC0o42ZWZkcMnFF/O3v/2NL33xS20+Ls5hpfr08OR5mR4ZnhSiB5X7o5hKYe1A7uwbh2t491gtt0xJY1KGpxta1zKbRWN6VhzTs+L4f+fX11heV1DH+yfr+P3mEn6/uYRRKS7m5cdzyfCkTnUAtLU9plKcrAkzKlUmILdEAmQhBriIYVJQGyGpjT0nZ1bOq+95KabMr/PwsmHE23t2SNCmaUzO9DA508PtMzIpqI00BMvP7Srn77vKSXJZOT/Xy4VDvczMiW/TeW+88UY+d/vtXHvtx8jKzGxze5JcNsoCUSqDOqltWGRFCNF5SimOVoc7dP85WRvhNxuLmZzp4cZJad3QuvbR0Bib5mZsmpvbpmZQUBvh/ZO1vH+yjj/vKOf53RV8Y14u84d4u7UdyS4bx2siZHsdxEnaWLOkf12IAe5Mabe29L6cKj7F0SNHmDdvHm8eruF/x2q5ZWoa49N6v6chL8HBdRNS+dmSYfz942O464IcJmd6eO9ELd9/+ySPrj9F2Gh9WdWkpCSuuOIK/vn88+1ug9dh5UBFCMOU5VuF6AnVIYNg1Gh3OkDEUNy/pgCHzcJdF+S0WL6tt+QlOLh+YhqPLhvOkx8ZxZBEJ/euLuDxrSXo3bhEtEXTcFk1DlWGuu09BgIJkIUYwNpb2u3111dx8SWXUBZS/GZTMZMy3Fw/sfd7Xs6V4LRy8fBEvrsgj79/fAw3TEpl5aFqvr7yGKd80VaPv+yyy/jf6tWEwuF2va/LZiGoGxT7Ih1tuhCiHU7WhDtUu/eJbSUcqQrzjbnZpPWDEZ/seDs/WzKMK0Yn8c89lXz3zRNUhfRuez+v00qZv35ETMQmAbIQA1h1yCCkmzisrf+p64bBG6tWccnixTywphCbReOuC3L7ZM/L2ewWjU9NzeDHi/Io8UX4yn+PsL6grsVjMjMyGDN6NO+//1673y/JaeNwZYiIYXa0yUKINghEDMoCervro68vqOOlfVV8ZFx9ffb+wmHV+MrsbL4xN5t95QG+/N+j7CkPdtv7JTitHCgPYnZjb3V/JgGyEANYe0q7bd68mfSMDP5XHcfByhB3zskhowdnV3fWnDwvv75iBFleBz9+t4A/bSttcZhyydKlrHp9Vbvfp36SC5yobl/vsxCifU75Itgs7SvtVh6I8vC6U4xMdvHpaW2fY9CXLB6ZxC+WDcdu1bjr9WO8tK8KRdcHsS6bhUBURsSaIwGyEAOUP2JQEYi2eRLGypUrmTj3Ip7fXcGyUUndPlGkO2TF23l46TAuG5XEc7sr+O6bJ6hsZphy7ty5HD5yhOKSkna/T5LLyomaCP6I0dkmCyFiiBomJ2siJLRjEpmhFA+uLSRqmHxnQW6X1GvvLaOSXfzq8uFMz4nnd5uLeWhtESG960etEp02DlXIiFgsEiALMUAV+yLY21gWqaKykg8++IC3zOHkJTj4wsz+2fMC4LRq3DEnm2/Oy2ZfeZD/95+jfFASaLKfw27noosW8cYb7e9FtmgaTqvGYZnkIkS3KAtEUaptk4vP+Puucj4oDfL/ZmWRl9D/l4f3Oqz8+KI8bp2SxjvHarlz5TEK6rq2t9durR8RK6iREbFzSYAsxAB0pvelrctAv7FqFZ6R0/GZNr69ILdDk2L6mktHJPHLy4bhsWvc/cZxnt9d0WSYcsmSpaxa9QaG2f7eE6/TSqlMchGiy5lKcawqjNfZ9vvQrtIAf9lRzkXDErh0ZGI3tq5nWdD45Hnp3HtxPhWBKF/971Heb2WORXslOq0cqw4TkBGxRvr/U1AI0UR7el9MpXjhP69Rlj2Dz0zPZFSyqwda2DOGJ7l47PIRzBvi5YltpdzzbgG+sx4Co0aOJMHrZfv27R06v0xyEaLrtWdyMUBt2OCBtYVkxdv5ypzsbl/QqDfMzInnV5ePICfBwT1tmGPRHlaLht2icaRaRsTOJgGyEAOMamfvy6r3NlNrWJkxaRzXjGvb8sv9SZzdwvcuzOXzMzPZWOjjK68e5VDVhw+CpcuWsur11zt0bpnkIkTXa8/kYoXi0fVFVAd1vrMgF88AGP1qzpk5FstOz7H4/lsnqAl3Ta9vgtNKsS9KdTeWlutvBu4nSYhBqj29LyFd8cfnXsY1bi7fnJ/brmWb+xMNjY+OS+HBJUOJGIqvrzzK28dqAFi06CI2b95MbV3Hhi1lkosQXae9k4v/e6Ca90/6uG1aBmMGwdLJTqvGnXOyuXNONrtLA/z4nRNtWiCpNZqmEW+3cFBGxBrIUtNCDDDtKaz/+/cO4z+2i+9+5Uskuwb+7WBSuoffXDGc/1tdwM/XFpHgsDIzx8vMmTN59913uPqqq9t9TrtVw4zUX/eRKQP/AS1Edyqqa/vk4qPVIf6wuZiZ2XFcOyGlU++rUAQCQSoqKqisqKCispLy8nIqKyoIBINccvHFTJk6pc+kbywblUSc3cL/rSnk0XVF3DU/p9Nt89itlPl1yvxRMuP7/yTHzhr4T0QhBpEzhfXT3K33vqwvqGPlG2+TN34KF47O7oHW9Q1JLhv3XDyEr688xn2rC/nFsmEsWbqUPz3xRIcC5PpzWjleHSY73oGnHWWphBAfihgmBbURkpxt+xv61YZi4hxWvnlBTptHv/bt38+ePbupKK8PghsC4ooKAFJTU0lNSyUlJZXUlBQyszKxaBZ+/Zvf4HK5uO6665g/fz42a+//nS8YmsBtdRGe2l5GfqKDT56X3ulzJrosHKwIkeK2YW9jDvhAJQGyEANIWwvrVwSj/Pz9IuxH1/OFb36lh1rXd7htFn5yUT53vnaMH75zgkeWnkdtXR2HDh1i1KhR7T6fRdNwnC77dl5WXDe0WIiBr8wfRdG2ycUHKoLsKQvyhZmZJLVx9Gvf/v388Ic/4KJFF5GalsqIkSNJTUkhJTWVtLQ0PO7mR4CuvuYaNm7YwPPPP89TTz7JtR+7liWLl+By9e6k5hsmpXKyJsKfd5ST63WycFhCp87nsFqoi+gU1kUYljRwJmx3hATIQgwQZ0q7Jbah9+UfuyoInDpGql0xY9rU7m9cH5QRZ+cnF+XxjVXH+enqAi659FJWvbGqQwEyQILTRkkgSl5QJ9ktt1Yh2qugNoK3jctKv7ivErdNY/HIpDbtX15Rwb0//Sl33nkn8+bOa3fbLJrGnDlzmDNnDrv37OGf//wnf/3LX7nyqiu56qqrSUrsndJyGvV130/5Ijy8rpDMeDvj0jqX6pXktHKkKkRmnAN3O5f5HkgG728uxABTEdAxlWq198UfNVl1uJrMki1ccdmyPpNT1xtGp7q5e34u+8pDHEmezDvvvEs40vGKFF67lYMVMslFiPYKRk38ERNnG+ZPVASjrD5Wy5LTebitCUci3PvTn3L5FZd3KDg+18QJE/jRD3/Iz37+cyrKK/jsZz/Lr3/zG04Vn+r0uTvCYdX44cI8Utw2fvzuSUr80U6dz2rRsGkaxwZ52TcJkIUYAJRSHK0Ot2lhkDcOVxMKBqnev5XFixf3QOv6tgvyvXx6Wjoba5x4MvJYt25dh8/ltluoixiU+Dr3gBJisKkN623+qv7qgWp0BVePbX1inkLxq1/9irT0dD7xiU90rpHnGJKfz5133skf/vAH4jwevvrVO7jv/vs4eOhgl75PWyS5bPxk0RAiuuLH75wk2MllqROdVgrrItSGB2/ZNwmQhRgAasIGwajRau+LieKV/ZXk1Bxk0sQJpKZ0bub3QPHxiaksHZVESeZ0/v7ifzt1rkSnlUOVIaJS9k2INiv1R3HZWg+RI4bivwerOD83nlxv65UWXnzxRQ4fPsw3v/GNbhstS01J4VOf+hRPP/00Y8eO4yc//glPPPEndL1ng8uhSU6+e2Eux6vDPLi2EKMTI1maphFns3CoIoQapCNiEiALMQBUBKJtKo20pdBPYV0Uz6mdLLzwwh5oWf+gofHl87OYNGM2x44cYs3e4x0+l8NqQTcVJ2vCXdhCIQYuw1RUBPQ25buuPl5DdcjgI+Nb/3K/ZetWnnvueX70ox/1yGQ6j9vNx669liefepJoJMKjv/xlk+Xtu9vMnHi+OCuL9QU+/rS1tFPninNYqQrplAcG54iYBMhC9HNKKUp8UTz21tMrXtpfSZI1StGhPcyZO7fb22aYinAnh/p6it2i8aNLRxA3eiY/e+YlCus6nouc5Kwv+xaIds0qV0IMZHURA1MpLK1U31EoXtxXxZBEB9OyPC3uW1hUxM8eeojvfOfbZGVmdqp9Sql29aLabXY+9elPc/LECZ5++s+deu+OuGpsMleNTeZfeyt57VBVp86V4LBysGJwjohJgCxEPxeImoR0E7u15YdLQW2EzUV+JoQOM/m8ycTHdU85sqihqA7pVASi1EYMDAXlgShVIR3D7NtDdV6Hla/f/BGiBzfwg7eOUxfpWIBrtWjYLBpHqgb3JBch2qIiEMXWhhGw3aVBDlWGuGZcSovpEv5AgJ/85MfcfPPNTD5vcofaFDUUNSGd8oBOZcigIqhTEdTbvGKm0+ngJ/fcw+rV/+M//+1c2lZHfH5mJjOy4/j1hmK2F/s7fB6nzULEUOwuDaD38ft3V5MAWYh+riast1r3GODl/ZXYNAge3saFC7s2vSKkm1QG64fiwoZJXoKD6TnxzB+SwJx8L+fneclPcFAXMSgP6PgjRp/Na5s3dTxZSfEUH97Lvf8rINrBh0KC00pxXZTK4OCd5CJEa86MgMW1ZQRsXyXxDguXDG++pJqpFD976CEmTpzEFVde0a52BKMmFWfdx3ISHEzLjmPB0AQuGJLAmFQXEUNRFohSFzZarVaTlJjIvff+H88++1feX/d+m9vSFWyaxncuzCU3wcG9qwsoqO34iFiK20Z1yGBXiX9QBckSIAvRz5X4onhamZznj5q8cbiauVk29u/dxZw5nUuvMJXCHzEoD0SpCESxaDA61cXsPC9z872MSHGT6LI1lJyLd1gZkeLmgiEJTMny4HFYqQjqVLajR6anaGhcc8VljK3ZyY6SAL/eWNyhPEJN00hwWtlR7KcmJEGyELEE2zgCVuqP8t6JOpaNSsbVwv3umWeewefz8aUvfbHVSXmGqagN61QEdCpCOg6bxti0D+9jI1PcJLtt2CwaTpuF3AQnc/O9zMiJJ9ltozKkUx7UW0wjy8nO5sc//gmPPvooe/bubflidLF4u5V7LhqCRdP40TsnqA13POUrxW2japAFyRIgC9GPRQ2TqqDR6uzvNw5XE9QV+TUHmDp1aosrRrWkLlwfFFeHDBKcVs7LjGPekARm5nrJTXAS57C22JtttWikeuxMyYpjbn4CI1Pqe2TK29gj01MWXXQRJ3Zv42Oj3Lx+qJp/7q7s0HlcNgtxdgtbT0mQLEQs1W0cAXvlQBUKuGpMcrP7rF6zmjfffJPvf//72G32mPsYpqIiGKU8oOOLmqTHOZic5WHBkASmZceT4235PqZpGkkuGxMyPMwfksC4NBe6gvKgTm1Yj3kPGzN6NN/61re45557OFlwstXftStlxdv54aI8SvxR/m91x0fEAFIHWZAsAbIQ/diZHoGWHjBnSruNS3NxYNt6LlzQsfSKurCBw6YxLTueBUMTmJgZR3qcvU2F/WNx2y3kJ9b3yEzPiSfFU3/zrQjohHSzV1MwkhITmT5jBtnlu7hwqJcntpXy3sm6Dp1LgmQhmlfq01sdAQvpJq8drOKCfC+Z8bED30OHD/PrX/+aH/3whyQlJTV7rsqQTn6Ck/Pz4pk/xMvYNDepHjt2a/vvYw6rhRyvkzl58czIjiPVY2+4h4WNxvevWTNncdttt/GD7/+AyqrOTZxrr0npHr42N6dTI2JnpLptVIcN9gyCnGQJkIXox8oDOq2tDbK1qL602+I8B3v27Gb2nNntfp+ooYiYiokZHpLdtlZX62uPMz0y49M9XDDEy/h0NzaLpT4XMBjFF+mdnuWlS5aw6o1VfH1uLuPSXDy4tpCDlcEOnUuCZCGaihr1cxdaGwF751gNvojZbGm36poafnrPPXzpS/+vxaXia0I6qW4bI1JcxLcy2tUemqaReNY9bFy6G6WgPBilIlj/hR9g2dKlLF6yhB98//sEgh27l3TUJcMTuXFSGq8fqmbt8Y592T8jxWWjIqgP+CBZAmQh+imlFKX+1ie3vLivkmSXDQp2Mn36DNyu9qVXKKWoDutMTHe3qZRcZzisFrK8DianO5k/NIHJGXEkuazUNKR26ESNnrkhT5s+ncqKSk4VHONHi/LxOqz8cn0xZgd7XyRIFqKxtoyAKRQv7q1kRLKTiRlN7126rnPvvfdy0UUXsWjhwmbPEzHq/3LHpnlaLSfXGQ6rhWyvg+mZLubkehmT6sKicXq+hs41113PqDFjuPfen/b4QiK3TEljWJKTJ7aVNOnhbq9U94dBcl+vTtRREiAL0U/5oyZRQ7XYm1tQV1/a7YoxSby3Zg0XXrig3e9TFTLI8TrIiG991aqu5LBaSIuzMyEjjvlDEpieE09ugoOQbp7umYk29Mx0B6vFwqWLF/P666tIdtn49PQMDlWGePtITYfPKUGyEB9qywjY9uIAx2sifKSZ0m6///3v8Xjc3Lp8ebPnMJWiOmQwMd3TpsVIuorHYSU3wcnMXC/zhiQwIcNNnMPGTZ/5AqbTy89++WsM1XOTlK2axudmZFDs03lpX0Wnz5fqtlEe1Nk9QINkCZCF6KeqQzqa1vJN6ZXTpd3mZ1rZv28/5886v13vEYyaOKwao1K6fxWqllgt9WkYI1PczBvi5fxcL6NT3Gha/UO2ItA9PctLlizhnbffJqpHuWh4fZmnJ7eVdSowlyBZiPaNgCU6rSwc1rS02zvvvsOOnTu5++5vt9grXBk0GJbkJDUudv5yT3DZLGTGO5iSFcfC4Un84ad3UXr8MCuefpbyQJSaHqoTPyM7nvNz4/nbB+VUd8H9J+10kLynbOAFyRIgC9FPFddFWny4BHSTVYequXBYAnu2rmfGzBntWm7VMBW+qMmkTE+HJrB0F03TiHdYyUt0MivXy7whXsalu6iJ1NdX7ko52dkMHTaUDes3YEHj8zOzqAjqPLe7c70vEiSLwc4XaX0ErMgXYUOBj8tHJ+OMUQbu3//+N7ff/jniPM2vqueLGHgdFoYn9+6X/LPZrRbyUxN47nc/Y+d//sL+//2HLK+D2ohBTbj77wefnZ5BxFD8eUdZl5wvzW2jLDDwguS+89QTQrRZWDepDRst1gN941ANQV1x9dgUVv9vNRde2Hx+XiyVQZ0xqS4SnLbONrdbuWwWsr1OZud6sVg0KoN6l1bAWLJkKStffx2AieluLhzq5Z97Kij1Rzt1XgmSxWDWlhGwl/dVYtXgirFJTbYdPXaUyopKpk+f0ezxUUMRMkwmZni6dGJxV0lNTeWvf3mGFY89zKGN7zI7z0u8w0p5INqtE5OHJDq5Ykwyrx2s5lh1uEvOmea2UeYfWEGyBMhC9EO+iEFLNfBNFC/vr2BcmotMW4iDBw8yc+bMNp+/JqyTFmcnL6Fn8447I85hZUZOPBnxdsqDXTdcOf+CC9i3bx9l5fW9LZ+ZlolSiie3lXb63BIki8Gq2NeWEbAaFgxNIM3dNDXi9ddXcenixVgtscMYpRRVIZ0J6fULE/VVw4YN46mnnuJb3/oWH2zbwuTMOIYkOinrprSxM26anE6cw8KKLSWdKvt2tjSPjfIBFCRLgCxEP1Tqj+Js5sEAH5Z2u3pcCu+tfY9Z55+Py+ls07kjholSMDbN3WVlkHqKzaIxPs3NuDQ3lSGjSybxuVwuFixYwFtvvgVAZrydj01I5Z1jtewr73ypJgmSxWAT0k3qWhkBW3WomoBu8pFxTUu7RfUo77z9NosXL272+KqQQU6Cg6wenlzcEVOmTOGxxx7js5/9LEePHGZUqpspmZ5uSRs7I9Fp5abJ6Ww95WdToa/Lzpt6Okje3wX3xt7WrQHyI488wsSJE5k0aRKf+MQnCIVCHD16lNmzZzNq1ChuuOEGIpH69cHD4TA33HADo0aNYvbs2Rw7dqw7myZEv2UqRVlAx9PCbOyX9lWR7LKxYEgCq9esbnP1irNne7f08OrLNE0jN8HJrNw4oobqkokoS5cs4fVVqxp6Wq6flEayy8bvN3eu6P4ZEiSLwaSulSWPz17caFxa09JuG9ZvYMjQoeTm5MQ8PhA1cFg1Rvfy5OL2uPjii/nOd77DTTfdRElJCRnxDmbnetG0+rSx7nDlmGRyvXZWbCnt0nrGqR4bxb4ogWj3BPc9pduegIWFhTz22GNs3ryZXbt2YRgGf//737n77rv52te+xqFDh0hOTuaJJ54A4IknniA5OZlDhw7xta99jbvvvru7miZEv+aLGOhm85NbCuoibCryccWYJGqrqzhy+AgzZrQtvaIy1PuzvbtKgtPGzNx4klw2SgPRTg35jR03FofdzpYtWwHw2CzcNjWdfeUh3j1W2yXtPRMkby/2E4z2XOknIXpaqS/S4hfwzYU+CuuiXBOj9xjg9VWrWLp0acxthqnwR1Wfm1zcFjfeeCM33ngjt9xyCz6f73TaWP2KpWX+zt3DYrFbND4zI5OC2gj/PVDdpefWNLotsO8p3frp0XWdYDCIrusEAgGys7N5++23ue666wBYvnw5L774IgAvvfQSy0/XMbzuuut46623enWpWSH6qpqQQUvzTf5zurTb5WOSWbt2LbNnz8bpaH2Y0R8xiLf3rdneneW0WZiU6WFUiouKkE64gykXGhq33Horj//xj+hGfa/I4lGJjEx28cTWUkJ619yrXDYLSkFFoHMTAIXoqwxTURZseQTsxX1VpLrrR8DOVV5ezt69e5l/wQUxj60I6oztB5OLm3PnnXcydepUPve5zxGJRLBbLUxIdzM2zU3lWavydZW5efFMyfTwl52l1HVhOke83UJhbaTLztcbui1Azs3N5Zvf/CZDhgwhOzubxMREZsyYQVJSEjZb/Qc3Ly+PwsJCoL7HOT8/HwCbzUZiYiIVFZ0vZC3EQHOqLkJcM70vAd3k9UPVLBiaQIrLxprVq1lw4YWtnlM3FSG978727gyLpjEsycX07HiCp3MfO+KCC+aRmJjIa6+9Wn9eND4/M5PygM4Le8u7rL3xDgsn+/mDRYjm1EUMlKmarVt8oibM1lN+rhybhC3GvejNN99kwYIFMUtWVod0MuLs5PajycXn0jSN++67D4fDwTe/+U2UUmiaRl6ikxm58YR0k9ouLAWnofG5mZnURUye3dl19zGnzYI/YnRbDnVP6LavWFVVVbz00kscPXqUpKQkPv7xj7Ny5cpOn3fFihWsWLECgLKyMsrKOlfHr6qqqtNtGmzkmrVPV16vsKEoLA+R4rIQiLH9f8fqiDMCLM1L4uixo9TW1TJy5AiqqltuQ3nAYGyKg0BNNOZ5e1p3fcaGO032V0Yoq1QkubR2T0K85dZb+fWvf820adOJi/OQ74RLsjXe3F3AvHRIdHXNLbUyaHDcGmixl+1s8jfZPnK92q+rrtnx2ih1viiWSOzKEq/uKifLFmR+utbkvqUUrN+wgeXLb22yLWwoIoZipMtFeXnv38U6e73uvfdevv3tb/Ozn/2MT33qUw2vj3Ap9ldGOFJpdugeFkuqBlcOsfH+wUIWZVvIiOua+1hdyOBgQYBcb9tS9vra32W3Bchvvvkmw4cPJz09HYBrr72W9957j+rqanRdx2azUVBQQG5uLlDf43zy5Eny8vLQdZ2amhpSU1ObnPf222/n9ttvB+pnfp45f2d0xTkGG7lm7dNV16vMHyUx4CfZ0/SGY6L4z4lK0lKTmTosixdffJExY8aSntbye1eGdMal2JmQ3reqVnTXZywnU3GkKsTx6jCpblu7esyTk5KZMGE8/3nlFb7whS8AcNPsOG5/+TD/PBLlm/O6ps3KpYPHRXpS2yqPgPxNtpdcr/brimt2MFhLTpqGI0Z+cF3E4LWTJVw4LIv8zLQm2z/Y9QGBgJ+pU6c2WnbaVIryoM4F2fEku/tOakVnr9djjz3GNddcQ0ZGBrfddlvD69kZH97Dkl027DEWUWmvT57v5c0XD/H3gyF+tCi/0+cDiDNMAoYiLc3b5mdLX/q77LYUiyFDhrB+/XoCgQBKKd566y0mTJjARRddxD//+U8Ann76aa655hoArr76ap5++mkA/vnPf3LxxRf3qYe1EH1BS5Nbthb5KaiNcPXY+okt/1u9moULW06vCOkmdk1jdKpr0Py9WS0ao1PdjE51Uxlq/6Iit966nLffeZsTJ08CkBPv4CPjUnjzSA0HK7qmtFF9/l5Y5mGIASUQNQgZZszgGOD1Q9WEDRWztBvU1z5eunRZo+AY6vOORyQ5+1Rw3BVSU1P561//yi9/+Utee+21htfP3MMmZ3qoDutdUoEixWXjxklprCvwsb3Y3+nzATisFkKGib+fTjrutgB59uzZXHfddUyfPp3zzjsP0zS5/fbbefDBB/nFL37BqFGjqKio4DOf+QwAn/nMZ6ioqGDUqFH84he/4IEHHuiupgnRL7U2ueXl/adLuw1NoLSsjIKCAqZOndri+erCBhMzPc0+sAayIYkOcr0OKoPty5FLSkzkxhtvZMUf/tDw2o3npZHotPKHLiq677BaCOkKX6R/PliEiKU2ZICK/UVcV4qX91dyXoabkTEmCvsDAda9/z6XXHxxo9frwgbJLhvDBtDk4rMNHTq0YSGRTZs2NdqWEe9gQrqHyqDeJSvvfWRcKukeGyu2lGB00Zdzq6ZR5u+fcyq69an4k5/8hH379rFr1y6eeeYZnE4nI0aMYOPGjRw6dIjnn38e5+nFC1wuF88//zyHDh1i48aNjBgxojubJkS/42thcktBXYSNhfWl3ewWjTVr1jBv7lzstuZzv6pDBiNTXSR1Ud5sf6Np9b0wyW5ru2slX3XVVZw6dYqNmzYCEG+3cuvUdHaVBll7vK5L2mez0G8fLELEUuyL4rHHDpA3FPgo9etcM65paiXA6tX/Y8rUKSQlJTW8ZpiKsGkyLt3d7KS/gWDKlCn86le/4rOf/SyHDh1qtC3b62B4spPyLiip5rJpfHp6BkeqwrxxuKbT54P6e2NRbbRfjoYNvm4jIfqpymC02XzZs0u7Aa1WrzBMhdUCeQltz3EdiKwWjQkZ9fVS2zPb2m6zc/vnP88ffv8Honp9Sbalo5IYluTkiW0lhLtgidh4h5WiumiX9AwJ0duihkllUMfdTIrYmuO1JDqtzMmPj7n99ddXsWRJ49rHtRGDoYlOPC0sWT1QXHTRRXz3u99tWEjkbMOTXWTG2buk7vCiYQmMS3Px9PYyAl1QUs5u1QibqktLyPUUCZCF6CeKfbHTK4K6yapDNQ2l3YpLSig6VcTUKVOaPVd1WGdIkjNmGaXBxmG1MCXTQ8RU7aqTfP75s8jOzuaVV14BwKZp3D4zk2Kfzot7O1+i0mbRiBgdL0snRF9SdzpdKNZcB0MpthT5mJUTjy3G9uMnTlBaWsLMmR8ueKSUwjAhux8sJd1VbrjhBj7xiU9w8803U1f34UiVRdMYl+Ymzm7p9P1CQ+PzM7OoCuk8t6trSu3atPoJ5v2NBMhC9APBqEkwGntyy5YiPwHdZNmoJADWrFnNBRfMb6g3fi6lFEpBZtzgebC0xuOwMiUrjtqI2eYJLxoat3/+8/z973+nuqZ+OHJ6Vhyz8+L5+65yqrpgyWi7RaO0Hz5YhDhXRSCKo5mO3r3lQeoiJrNyY/cer3r9dS699FJs1g9P4I+apHlseJo76QB1xx13MG3atIaFRM6wWy1MyozDhE4vJjI+zc2iYQm8sLecki64/3j76WiYBMhC9AO1YR2a6ezdWOgjzm5hYqYHgNWr13DhhQuaPZcvYpIZb8fdxhq7g0WSy8aEdDcV7ZjwMiQ/n4sWXcQzf/5zw2ufm55JxFA8vb1zNdqhPs3ilK//PViEOJtSihJftNlUiE2FPizAjJy4Jtt0Xeett99myeIljV4P6Sb5iYMvRezMQiJOp7NhIZEz3HYLU7I8+CIG0U6meX1qWgYAT24r7dR5oD6VTTdMavvZaJg8IYXoB0r9Udwxal0qFJsKfUzPicOmaRSdOkVZaSnnnTe52XOFDDXoc4+bk+11MCLZSUU7cvluvuVm3nvvPY4cPQpAXoKDq8Ym8/qhag5XhTrVnjMPlppQ/3qwCHE2f9QkYprNpnRtLPAxMcNNfIze4A0bN5Kbk0NeXl7Da2HdxG23kuQaXL3HZ9hsNn73u99x8OBB/vSnPzXaluC0cV6mh6qQjtGJ8m+ZcXauHZ/Gu8dq2Vve+fKVdqtGia9/jYZJgCxEH6ebioqAHrPH91BliKqQzuzTQ5OrV69m/vz5jYYizxbSTbwOCwnOwflgaYsRyS6y4h1UtjFFwhvv5aabb+IPv/99Q4m3myanE++w8McuKPvmtFoo9kk1C9F/VYf05gbAKPVHOVod5vxcb8ztq15/nSVLG0/Oq4sYDE9yDpra7bF4PB5+/etf84tf/IKjp7+cn5Ee56iv8x5sf533s90wKZVkl43fbiwmpHfuPhZnt1Lii3QqaO9pEiAL0cfVhQ1MFbu826ZCHwAzc+oD5NaqV/iiBsMG+YOlNZqmMTbNTbzDQm2kbUHyZZddTk1NDe+/vw6oz7m7eXIG24sD/KOTE108dgul/miXLAYgRG8orosQ11x6RVH9Pez8GPnHFZWV7N69u1HKWH0FHo20uLYtXzyQjRw5kjvvvJM777wTw2g8yjQk0UFOgoOKTow+uW0WvjQrk4OVIX7y7slOBclWi4ZuQk0/SrOQAFmIPq4yGG1+aLLQx9jTtYwLCguprKpi0qRJMffVTYVN00iNsUy1aMxm0ZiUEYemNALR1m/oNquVz3/h8/xxxQrCpyfOXDk2iUXDEnhqexn/2FXe4bZYLRqGCTVdMOlPiJ4W1usrsTS3AujGQh+ZcTaGJDWdNPzWW29ywQUX4Ha5G16riRjkJzikAs9pn/nMZ7Barfzxj39s9LqmaYxJdZPktFId7vi9Y8HQBL4+N5ttxX7u+d/JTpWwdFo1iuv6z2iYBMhC9HElvmjM3pfqkM6+8lDDzO81a+rTK6yW2H/WtWGDoUmuZmspi8ZcNgtTsuMI6CYRo/VZ4dOmTmPY8OH8+9//BupXkPrmBTksGpbAk50Mkl22/vVgEeKMurDR3OJ5hA3F9lN+ZuXGN1k+WqFY9fqqRukVSikMQ5HllQo8Z1gsFh555BF+9atfcfDgwUbbrBaNiRkebJrWrjrv51oyMok752Sz9ZSfn77b8SA5ztG/RsMkQBaiDwtEDEK6iT3GBL0tp/zAh0OTq1evYWEz6RVKKUylyJBhyXaJd1iZkhFHdchoU+7c5z73Of71r39RUVkJ1NdG7oog2WO3UBrQibYhUBeiLyn1R3A286X9gxI/YUMxO0b+8Z49ewGYMGF8w2tnKvAMhoVB2mPo0KF861vf4s4770TXG/cWO20WpmTFETJUm77oN2fZqCTumJPF5lN+fvq/k0Q6ECRbNA1DqX4zGiYBshB9WE1Yp7n6bpsKfSS5rIxKdXHi5Elqa2qYMHFizH2ltFvHpcbZGZfWtgkvuTk5LF26lCeffLLhtXOD5Od2tz9ItmgaCiXVLES/YipFeSD2AkcAGwp9OK0ak7Oalnd7/fWVLFm6pFHPclA3yRuEpd3a4tZbbyUuLo7f/e53TbbFOaxMzvRQ08Yv+s25bFQyX52dxeYiP/euLuhQkOy2WfrNaJg8LYXow0p8Oh570wBZV4rNRT5m5sRjQWP16tUsWLAg5kQ+kNJunZWb4GBIkpPKYOsB6ic+8Qm2bt3CgbOGO88EyQuHJfCnbR0Lkt1WC0X95MEiBNSnV+inJ9WdS6HYWFDH1Ow4nOeMkAVDQd5/730uvXRxw2sh3cTrtJIoFXhislgs/OIXv+APf/gDe/fubbI91WNnbFp9nffOVLa4fHQyX5mdxcZCX4eC5Di7hRJ//xgNkwBZiD4qaphUBnXcMSa37CsP4ouYDekVa1avZv6C2IuDhHSTBKeUdusMTdMYkezCZbe0ukpVnMfDrbfcyu9+97tGJd5smsa3OhEke+wWyoPRTg2TCtGTqkI61ma+tJ+ojlDi12NWr1i9eg2TJk0iJTm54TUp7da6vLw8vvOd73DHHXcQjTatOZyX6GRokpOqTlaSuGJ0Ml8+vz5Ivm9NAdF29EprmgaaorofjIZJgCxEH1UXqQ+EYj0Qzqw8NT0njhMnT+L3+5kwYULM8/giBkMT5cHSWVaLxphUF3VtmOyyeMkSIuEw7777bqPXOxMka5qGpjSq2rGIiRC96VRdlDhH89UrAGblNA2Q62sff7hynm4q7FYLKVKBp1Wf/OQnycjI4LHHHou5fViSE6umdfqL9pVjkvnSrCzWF/j4v9XtC5L7y2iYBMhC9FEVgSgxFpYCzlp5ym5l7do1zLtgXsz0Ct1U2KwWKe3WRVLcNtI8dupa6YGxWix88Ytf5InHn6C0rPGS0+cGyc/vbnudZLdd6xcPFiECUYNg1MRhbS5ArmN4krPJxOGCggIKi4o4f9b5Da/Vhuu/5Etpt9ZpmsbPfvYznnrqKXbu3Nlku91qYVSKq0vqEV89NpkvzMxkfYGP+9cUtLk6hcduoSIQJdzKaFxvkwBZiD5IKUWJLxpztnZZoH7lqTPl3dauXcuCZtIrzjxYpLRb19A0jZEpLkJGfVWQlkyaNImPXnst3/jGNzhx8mSjbWcHyU9sK21zkOy2WagK6q2meQjR2+rCRnPzi/FFDHaXBjk/r2nv8eurVnHJxRdjs9mA+nuhUorMePmS31bZ2dn88Ic/5Gtf+xrhcLjJ9sx4O16HlWC08/eRj4xL4QszM3n/pI8H1hS2KUjWNA0Njeo+Xs1CAmQh+qBA1CRimjF7TDY3rDzlpejUKSorq5gwoWn1CnmwdI94h5X8BAfVbeiB+di113Lrrbdw9913s//AgUbbzgTJFw71tjlI1jQNTdOoCjbNLxSiL6kIRHHFKE8JsKXIjwkNX/LP0A2Dt956k6XLljW8VhcxyIx3NLvQiIjtuuuuIz8/n0ceeaTJNoumMTrVTV0bFkFqi4+MS+HzMzNZe7KOB9a2LUj22C0U9vHRMPnECdEHVYf05jpf2FDgI91jY2iSg7Vr13DBvHkxFweRB0v3GZrkRIM2PQgWX7qYr371K/zwBz9g2/ZtjbbZNI275ue2K0iOs1sorO3bDxYxuCmlqAgaMScYQ/0cCq/Dwvg0d6PXN2/eTEZGJkPy8xteCxuKvERZGKS9NE3joYce4tlnn2Xbtm1Ntie7bWR67NR2YpW9s310XAq3z8hg7Yk6HnyvEL2VETa33UJ1UO+SXuzuIk9OIfqg4mZWz4sYiu3FfmafXnlq7dq1XDD/gpjnkAdL93HaLIxMdp2uU926uXPm8r3vf48H7n+ANWvXNNp2bpD8zz0tB8kum4XasNmmJbCF6A3BqIlumjFTu0wUm4p8zMiJb1LhYtWq11l61uS8YNQkwWklwWnr9jYPRBkZGdxzzz3ccccdhEKhJttHprgI662ni7XVteNT+ez0DNYcr+NPW0pb3V/TNKpCfXc0TAJkIfqYiGFSE9ab1AYF2FUWIKQrZubGU1JayqlTxUw+b3KT/epLu8mDpTtlex04ra2XfTtj8nmT+b//+z9+97vf8eprrzbadiZIviA/nie2llLZSm6epkGlVLMQfVRdxEA1MwZ2oDxETdhokn9cXV3Nju07uPDChQ2v+aMmw5Nc3drWge6aa65h3LhxPPTQQ022eRxWhiU7u7Tk2nUTUlk4LIFVh6tbrWzR10fDJEAWoo/xRUw0tJhl2TYW1GG3wNSsON57by1z585pmMzS+BwGw5JkYZDuZLVojE1zt1rR4myjRo3ioYd+xnP/eI6//+MfTeok3zolAwWsPlbb4nni+/iDRQxuLeUfbyjwoQEzsxsHyOvWr2PGzBnEeTwARA2F3aqR7JYv+Z2haRr3338/L7zwAhs3bmyyPT/RiUWrv95d5ZLhCfiiJltOz5dpjstmoa4Pj4ZJgCxEH1MTitLMs4VNhT6mZMbhsllYu3Yt8+fPb7LPmdJuKW6ZnNfdUtw2Uj02fG2ojXxGXm4uP3/4Yd595x1WrPhjo+HNoUlOhic5efdYTYvncNos+CMG/na8rxA9odX84yIf49PdTRYu2rB+A3PmzGn4ufb0l3ypwNN5qamp3HfffXzta18jEAg02uawWhiZ4qIm0nUjUtOz4/E6LLzbyhd9AIsGlYG+ORomAbIQfUyZP/bqeYV1EQrroszK9VJeUcGJEyeZOnVak/1qwwbDpLRbj6gv++YmqJvtWr41LTWVn/38Z+zbt49fPPwwuv7hA2LhsAT2lYco9rWcm6dpUN5HHyxi8Gop/7giGOVQZajJ6nmhcJgdO3cyc+YsAEylUKgmNZJFx11++eVMnTqV++67r8m2rHgHbpu1y8pH2iwa84cksL6grvWVR/twNQsJkIXoQ6KGiS9q4IwRIG86vfLU+XnxvP/+e5x//iwc9sYPEFPVT7jIkNJuPcbrtJLXxrJvjY6L93L/ffdRXVPDPT+9h9DpeqWLhiUC8L9WepG9DiuFteF2BeZCdLeW8o/P3MNmn5N/vGPHDkaOHEmC11t/jrBBjtcR8z4oOu7ee+/l5ZdfZu/evY1eb88qoW21aFgCIV2xoaDlNIu+PBomnz4h+hBfxAQV++GysdBHXoKD7Hh7s+kVvohBlpR263FDk1wowGjHcqsALpeLH//oR3g8cXz/e9/D5/ORFW9nXJqL/x1veXjSYbUQ0lX9Z0aIPqKl/OONhT7SPLYm8yM2bFjPnNmzG36OmIpcr8yh6GrJycl8+ctf5oEHHmiyLcVtI9Vta9ecipZMyvSQ4rLxv+Mtf9GHvjsaJk9RIfqQmrBOrNg2pJvsLPFzfk481dXVHD50mOnTZzTZT0q79Q7X6bJvHZkNbrPZuOuuuxg+YgTfuusuKquqWDQskSNVYU7UNF0F62xWDcoDfXN4Ugw+SinKm8k/jhiKrac+LFHZcAyqUf5xIGqQ7LbhdTYtcyk679Zbb2X37t1s2rSp0euapjEq1U3IMLuk7JtV07hwmJdNhT58rUzC66ujYRIgC9GHlPmjMR8u24r96Obp9Ip17zNj5gxczsY9LFIztHdlex04bRrhDuTxWTSNL33pi8yZM4f/u/de5g+tH2r+XyuTXLxOK0W10T73YBGDUzBqYjSTf/xBqZ+Qrpqsnnfo0CFcLhd5eXlA/SqiQxOl97i7uFwuvv71r3P//fc3uW/UrxLqbHe6WHMWDUskasK6E3Ut7ndmNMwf7Vv3MQmQhegjooZJXbj5/GO3TWNihqf59IqoITVDe5HNojE61UVtB3PpNDRuuukmThWfwldaxJRMD+8eq2lUCi7We4YNk7o+9mARg1N9Dmvz+ccOa32JyrOtX7+e2XPq0ysihonLZpHSbt3s+uuvp6ysjHfffbfJtiGn01/amy4Wy9g0F5lxtla/6EP9aFhVF9Zj7goSIAvRR5ypf3wuhWJjoY/p2XEE/T727dvHrFmzGu0TNRQOqzxYeluax06yy9bhCSc2q5Uli5fw+uuvs3BYAoV1UQ5XtpxmYbNoVAb71oNFDE4VgWjMBY4ANhb4mHy6ROXZzk6vqI2YDE1yYolRA150HZvNxt13383999+PaTYe8epMuti5NDQWDktk6yk/1a0sfmSzaES6sBZzV5AAWYg+ojZsEKsy27HqMOUBnVm58axfv46pU6fidrkb7VMXNRgqpd16XX0enwt/O8u+nW3psqW8/dZbnJ/twqrRak1kr8NKsV/vsuVihegIpRTlgdglKgtqIxT5opyf6230enl5OSWlpUyYMPHMSUj1SAWennDFFVdgsVj4z3/+02RbTifSxc61aHgCJrC2lTSLvkgCZCH6iDJ/BLe96Z/kmTI5M3Pjm02vUArpPe4jEpw2cr0OajqYx5edlc3w4cPZvXUT07Pj+d+xWswW0iysFg3dVPilmoXoRYGoiaFUzC/pGwvrg6Nz6x9v2LiRGTNmYLPW1+D1Oq1SgaeHaJrGd77zHR566CGi0cY1162dTBc72/AkJ/kJjlbLVvZF8kkUog+IGia1YTPm8OTmIh8jk124VZQPPtjF7NlzGm03TIXNohHnkD/nvmJYkgtdqQ7n8S1btoyVK1eyaHgCZQGdPWXBFve3aPUVUIToLb4W8o83FvoYkugg65z67GeXd/NH60tUip5z4YUXkp2dzXPPPddkW5rHTpLL2un6xPVpFgl8UBqkPNDy4kd9jTxRhegD/FETUGjn5N7VRQz2lAaZlRvPhg3rOe+8ScR5PI32CURN0j12ydvrQ9x2CyOT3VR1sBd57rx5HD58iBF2Pw5r69UsXDaNUn//eviIgaW5/OOAbrKrJNB09bxQiA8+2MXMmTMBUEojySWjYD3pTC/yL37xC4LBYJNto1PdnUoXO+PM4kerW6nt3tdIgCxEH1ATMrDGCHC3FPkxqS/v1lx6RdgwSY+TB0tfk5vgwGHRiBjtT31wOhxcfNHFrH77Tc7P9bLmeC16Cw8pp1WjJmQQ7cB7CdFZLdU/3lrkR1c0yT/etn0bY0aPJj4+HsNU2K0a8TIK1uOmT5/OlClTeOqpp5psS3DayInveLrYGXkJDkaluNpUzaIvkU+jEH1AeSAaM/94U6EPr9PKEA9s376dOXPmNtqulAINKarfB9ksGuPS6meD6x1ItVi2bBlvrFrFgvw4qkMGO0sCze6raRpKQZ3kIYteEIia6M3UP95Y6CPObmFCeuOJxRvWb2go7xaImmTE2ZuMoImecffdd/Pb3/6W2tqmAezw5M6li52xcGgC+ytCnPL1n5EuCZCF6GW6qagJGU2GJ00Um4p8zMyOY+uWTYwbN44Eb+NemJCuSHLZcFjlT7kvSotzMDHDQ2VQb3eQPGzYMNLS07EWH8Bt03j3aMu9L3YLVPazHD8xMPgiRswSlSaKTYU+ZuTEYTsreDaVYsPGjQ3zKSKGIs0jo2C9ZezYsVxyySX8/ve/b7LNbbcwPs1DRVDvVJB84bAEgH41WU+eqkL0Ml/EQNOa5h8fKA9RGzaYdTq9YsGCBU2ODegmmXFSFqkvy/Z2PEhetnQpb73xOvPyE3j/RG2LdUI9dqvkIYteURGI4oqRf3y4IkRVSG+SXnHw4EHi4uLIzclBqfqlcBJkFKxXfeMb3+Dpp5+mrKysybZsr4Px6e5OBcmZcXYmpLt5tx+lWUiALEQvqwvH7n3ZWOhDA85LdbBl8xbmzp3XZB+llExs6Qc6GiQvXLSQD3buZEaygS9qsvWUr9l97VaNkGESiMqiIaLnnMk/jlWebUPh6RKVOU1XzztTvSKkK1LcNuwyCtar8vPzufbaa3nsscdibs9NcHY6SF44NJFj1WGOVbe8+FFfIZ9IIXpZWSCKJ0b+8cZCH+PSXBzctY1Ro0aRlJjYaHvUULhslpjHir6nI0Gy2+Vm/vz5lOx8D6/T2obeF43aPrZcqxjYWso/3nT6Hnbul/gNG9Y3rJ4X0E0y4uVLfl9wxx138MILL3Dy5MmY23MTnIxLc1PewSD5wmFeNFqvytNXyJNViF7UXP5xZUjnUGWI2Xle1q5dywXzL2hybCBqkBkvE1v6k44EycuWLWPVqlXMz/ey7mQdoRZWt3JbNcokD1n0oOZGwKpCOvsrQk3Ku5WWlVFWVs648eOB06NgTgmQ+4K0tDSWL1/Oww8/3Ow+eYlOxncwSE522ZiS5eF/x2pQLSx+1FdIgCxEL/JHDFBN8483nV49b2qGk40bNzFvXtMAOWpCilvyj/ub9gbJY8eNxelwkB86QdhQrC9oPs3Cbbd0ejKNEO1RGYydf7zpdHrFufnHGzdsYNasmdis1oZRsFgVfETv+MIXvsBbb73FgQMHmt2nM0HyomEJFPmiHKwIdbap3U4+lUL0otqwgRZraLLIR6rbRs2xvQwZkk9aamqj7UopNCnv1m+1J0jW0Fi2bBkHNvyPVLetxVngFk1Dmer0qmZCdC+lFOUBPWb+8cbC+nvYyBRno9fXb9jQUL3CL6NgfU5CQgJf+tKXePDBB1vcr6NB8gVDErBp9IvJehIgC9GLygNRPOc8XHRTsbXIz6zceN57r5nqFVGTFLetUekk0b+0J0i++JJL2LRxA3MzrWwq8rUYAFs0jWrJQxY9IBA10ZVqkn989j3s7PSLUCjE7t27mTFjRsN+qR4ZBetrbrvtNrZv3862bdta3C8vsT4nuT2jVl6HlRm58aw+XovZx9MsJEAWopcYpqI6Rv7x7tIAAd1kepaLdevWx0yvCBpKyrsNAG0NkhO8XmbNmoXz5FZ0E94/Wdfsvh67hRJfpDuaK0QjdWEDVNMv6fvKgwR0k1nn5B9v2bqVsWPGEB8Xh6kUFk3D65BRsL7G7Xbz9a9/nfvvv7/VffMTnYxtZ5C8cFgC5QGdPaXB1nfuRRIgC9FLfM3kH28s9GHTwF5+hJzsbDIzMpoerCDBJQ+WgaCtQfKyZcvYsvotsuJsLQ5POm0WfBGDcAuT+YToCpXBKG5b0wB5R4kfgMmZnkavr1//YfWKYNQk1WOLWf1C9L4bbriBoqIiVq9e3eq++YlOxqS2PUiem+fFYdV4t48vGiIBshC9pC5sEGPyN5uLfEzK9LB53fsxq1dEDBO33YLHLgHyQJHtdTAhw91ikDx5yhRCoRDn2SrYfspPdUhv/oQakocsulVL+cc7iwOMTHY16h02lWLTxo0Ny0uHdEVmvKPH2ivax2azcdddd3H//fejVOtB75AkJ6PbGCS7bRZm58Wz5ngdehvO3VskQBail5QHdDy2xkFuVUjneE2EqRlu3nv/febPb5p/7I+YZHslvWKgyfE6G4LkWA8Yi6axbOlSgvvexwTWHG8+zcJpsciqeqJbNZd/HDYUe8sDTXqP9+/fR0JiAtlZ2QAokPSKPu7KK6/EMAxeffXVNu0/tB1B8qJhidSEDbYX+7uiqd1CAmQheoFhKqpCOq5zhid3FgcAiK89QUpKMjnZ2U2PVYpkWT1vQMrxOhmW5KSumd7fSxcvZvvGdeS7Ff873vzwpMduoTygt6nnR4iOaC7/+EB5kIgBk7MaB8gb1n9YvSKsm8Q5pLxbX2exWPjhD3/Id77zHd599902HXMmSK4MtjDCBczMicdjs/TpRUPk0ylEL/BHjdOl2ho/YD4o9eOyaRzbuZH58+c3Oc4w63tspLzbwJURb0dvJq5NS01l0sSJ5NfsZVdpsNleYqtFQzcV/qjkIYvuURFoOf/4vIxz8o83bGjIP/brJtmSXtEvzJ8/nxUrVnDnnXeyYsWKNn3pHprkJMFlbXFRI6dVY94QL++fqCNi9M0v8hIgC9ELasMGsUp/7iwJMDHNxfpm0iuCukm6x4ZF6oYOWF6HFZdVI9rMQ2PZsmWU7lgLwOrjLfW+KGpaylMWooOUUlQEY+cff1ASYFSKi/iz0idKSkupqqxk3LhxAJgKktzyJb+/mDNnDq+88grPPfccX/va1wiHw60ek+N1tPoFfdGwBPxRk82nml/8qDdJgCxELyj3N80/rg7pnKiJkBk6hSfOw5D8/CbHhQ2TjDjpeRnINE0jJ8FBXTR2msXMWbOoKi9lKJUtVrPw2KyU+SVAFl2v1fzjc3uP169n1vnnY7VYMEyFzaI1CqBF35efn8/LL7+M3+/nuuuuo6SkpMX9U9x2WutsnpoVR4LT2mfTLCRAFqKHnck/dp4zPPlBSX3+ceDItpjpFQBKyep5g0Gax47ZzNPFZrWyePFivCc3c6gyREFd7JrHLptGZVAnakiaheha7c4/3rCeObPrq1cEojIK1l95PB7+8Ic/cNFFF3H55Zezffv2Zvd12y0kOC0EW+hFtlk0Fgzxsv5kHeE+eJ+SAFmIHuY/3TN47gNiR0l9/vGR3TsaJrOcLaSbeJ1WnDGGNcXAEme34LZZiTTz0FiydCnHtr0HerTZpafP5Lf7In3vwSP6t9byjyed1YMcCAbZu3cf00+vnlc/CiZVePori8XC17/+de69915uvvlmXnjhhWb3zUtw4m+lHvvCYYmEDcXmor5XzUKetEL0sLqwATGW2NxVGmSUO0JVVRWjRo1qst0fNciSiS2DgqZp5CU4mq1mkZOdzagRI8ivO8i7R2ubHcq0WeoXcxCiq7SUf7wzRv7x1i1bGD9+HB63u36ClyajYAPBZZddxvPPP89DDz3Evffei2E0vVclu21oSrU4sW9SpptUt421J/pemoUEyEL0sPKAjvuch0t1SOdYdZikqsNMmTwZq6Xpn6ZSGklS3m3QSHXbYnyN+tCyZcvg0HpO1kYoaibNon7ZaQmQRddpMf+4LEb+8YYPy7uFdEWSy4bDKqHHQDB+/Hj++9//smPHDm677TZqaxsHuU6bhWS3nWALvcgWNC4cmsC2U34CfazqjnxKhehBpjpT/7jxn96u02vSB0/uY9r06U2OM0yF3aoR75A/2cHC47DidTRfKmnuvHlUnzqO5itvdnjSYbUQ0hWBZib8CdFezeUf7y8PEjUb5x8bpsmmjRsbyrsFdJNMSa8YUFJTU3n22WfJz8/nyiuv5PDhw42253jtrQa+C4clYCjYfqpvpVl029N2//79TJ06teFfQkICjz76KJWVlSxevJjRo0ezePFiqqqqgPphm69+9auMGjWKyZMns3Xr1u5qmhC9xh8xMVXT/OOdJX4cFjiy9wOmT5/W5LhAtD5v79y6yWJga6lUktPh4JKLLyazeBu7Slt4sGhn0nqE6Lzm8o93lvjRaJx/vG/fPpJTUsjMyADqn/MyCjbw2O127rvvPm6//XY+8pGPNFpUJMltA40W0yzGprm4dnwy+cnOHmht23VbgDx27Fi2b9/O9u3b2bJlCx6Ph49+9KM88MADXHLJJRw8eJBLLrmEBx54AIDXXnuNgwcPcvDgQVasWMEXv/jF7mqaEL2mNqyjxRg431kSYISlGqfD0bAU69kihiLNIw+Wwaa1UklLly2ldu96imvDzS4a4rZqlMuy06ILtJZ/PPKc/OOzq1dEDYXLZsEjq+cNWDfffDOPP/44d955J6+99hpQP4qV7mm5F1lD4xOT0sn39q05Nj3ySX3rrbcYOXIkQ4cO5aWXXmL58uUALF++nBdffBGAl156iVtvvRVN05gzZw7V1dWcOnWqJ5onRI+pCDaffxxfeShmeoVSCgUkyMSWQcdtt5Doar5U0vBhw8lIT4OyY2xtpti+226hLKg3WzZOiLZqNf84M8by0nPOlHczyIyXUbCBbvbs2TzyyCM8/PDDDb3GOV5Hi3nIfVWPBMh///vf+cQnPgFASUkJ2dn1PWRZWVkNxaYLCwvJP2thhLy8PAoLC3uieUL0CFMpKmP0vuwuq88/9h3fy7RpU5scF9IVyW4rdpnYMijlelsulXT50kuxlx9kSzN5yBZNwzQVvmYqYgjRVq3mH58VIBcXF1NTU8PYsfWr5+lm/YiIGPgWLVqEruu89957ACS6bGia1u++pHf7mG0kEuHll1/m/vvvb7JN07R2f5tcsWIFK1asAKCsrIyysrJOte9MDrRoO7lm7XPmevkiJlVVYTR340B3z4kKsix+fBWljBg+gqrqxte3KmgyMtlG2elAejCQz9iHDENRWxPCEop9v5wwYSJvv7uaY6dKqaiKwxLjluoLGRwhQK5XAhSQz1dHVFVVUWKGCUdNqiKN72EfnKgm3RIk3xlpuH+tW7+O+fPnU1NTjakUtWGTUFyYMt/g6EEe7J+xT3/60/z1r39l7NixANgjYQrrzGYnmvsiJvaon7KyvpNm0e0B8muvvcb06dPJzMwEIDMzk1OnTpGdnc2pU6fIOJ28n5uby8mTJxuOKygoIDc3t8n5br/9dm6//XYApkyZQnp6eqfb2BXnGGzkmrVPeno6el2YpEiIZHfjP7utVVXEh6vA7SYvL6/JsYYjyogcL3GDbGlW+Yx9qEz5CUaNmJ+B5KRkdBNOFhVTZoxnXJq7yT5u3SSqQXq6tyea2y/I56t9lFIUB5xkJVibpFjsqK4lMSmZ3PTUhtc2bNjIVVddRXJSMv6IQVqKlezMuJ5udq8azJ+xj3/84/z85z+nurqa0aNHY4mLsv1UgOS42GGnLWxgDVv71DXr9jHbv/3tbw3pFQBXX301Tz/9NABPP/0011xzTcPrf/7zn1FKsX79ehITExtSMYQYCMoDOi5r4wdLbdjgaHUYd/nhmNUrZGKLgPpSSaEWlmKdMnkSFO5lc1HsPGSXzUJd2CDcD/MARd8Q1BWGok35x8FQkH379jXc04KGktXzBhm3280tt9zC448/DkCSy4bNWl+ytL/o1qeu3+/njTfe4Nprr2147dvf/jZvvPEGo0eP5s033+Tb3/42AJdffjkjRoxg1KhRfO5zn+O3v/1tdzZNiB51Jv/YfU6g+0FpAICaY3uYPq3pBD2/TGwRnC6VhNZsqaRpk8/DWbKXLc0EyGdIHrLoKF/EjFlRZV9ZoEn+8c6dHzBmzGjcrtOjGQoSnFKFZ7C57bbbePnll6moqMBq0ciKd+DvRzXZu/UTGxcXR0VFRaPXUlNTeeutt5rsq2kav/nNb7qzOUL0mkDUxDCb1j/+oMSPXQ9SWnSS8RMmNDlON5VMbBE4rBbSPDZ84dhpFsOHD0cFqtl34hS+yJBGpbbOcFotlAd0Uj3yeRLtVxkycHti1T8O1Nc/PitA3rplC9OnzwAgYph4HJYmnQNi4EtPT+eyyy7jL3/5C3fccQcZcXYKa8O93aw2k0+sED2gLqwTqxN4Z0mA3OBJJkwYj9PReHKCqRQWTcMr5d0ELZdKslg0Jk6Zjircy/biQMx9PHYLZf5oiwX7hYhFKUVV2Gym/rGfUSku4u0f3qe2btvGjBn1I2L+iEl2fN+ZeCV61u23385TTz1FOBwm0WXFZrX0mzQLCZCF6AEVMfKP6yIGR6rCOEoPMm1a0/zjYNQk1WPDFqssgRh0El22+pJtzQS4ly6Yi6Vwd7NpFlaLRsQ0W132VYhz+aMmZjP1j/eVBxulV5SVl1FdXc3IkaMAMFR9mUoxOI0bN45x48bx0ksvYdE0suPt1PWTVC8JkIXoZmYzq0/tOp1/XHl0D9Ni5B/LxBZxNptFIyPOhj8SO8A9f9ZMKDvCphOVqBirNUL9ilU1Yb07mykGoJqQDjT9oh4r/3jr1m1MmzYVi6ZhmPVBdayUHzF43H777axYsQKl6p9puvQgCyGgfva3GWP2987iAPZgFdGgnxEjRjQ9UCa2iHNkeZ2EjdgPF2+8l6z84ZQf2UtBbTPLTts0Sn0SIIv2KfPruGLcis7kH088N//49Bf+QNQk3WNrMvdCDC5nFg5Zu3YtXqcVh9XSL4JkCZCF6Gb+ZmZ/f1AaILPuKFOnTm3yAJGJLSKWRKcVq6X5UkkL5s6Bwj3Nplm4bRaqQnq/eDiJvkE3FVWhpili0DT/2FSKbdu3MW16fYAcNk0y4iT/eLDTNI3Pfe5zrFixAoumkeO194uKOvL0FaKbVYYM3LbGDxdfxOBQZQhrycGGh8nZ/BGTrHhJrxCNtVYq6ZIL52Epaj5A1rT6UnF14b7/cBJ9gy9ioJRqUmoypCv2npN/fPjwIRK8CWSeXgAMhUwyFgBce+217Nixg4MHD5Ie56A/lGSXAFmIbmQ2M/t7V2kAlEn50T0xJ+gZSpEca0xTDHqZ8XaizaRZ5Ofn4XI62L73EJFm9rFZNKqCsVMwhDhXVTAac6LwvvIA+jn5x1u2bGX66eoVId0kwWnFGaPyhRh83G43t956K48//jjxDgsum0akhcWP+gL55ArRjfwRs2Giytl2lgSwVZ8i0ev9sLfltDP7S8+LiCXBacXaTKkkDY3zps8kemIXe8qaL/dW4pc8ZNE2pX495kqesfKPt2/b1pB/7I8aZHklvUJ8aPny5bz88stUVlaSm+DE18cr6kiALEQ3qm2h/nFa7RGmx0ivCOkmaW6Z2CJis2gaud7mSyUtWzgPivaw5ZQ/5naH1UIoahLs4w8n0fvCuok/YuCwtl7/OBQKsW//fiZPmXJ6D40E+ZIvznJm4ZBnnnmGNI8Ns4/PhZAAWYhuVObXm+YfRw0OV4ag+EDs+se6ktXORIvS4+zozTxbZkydgqW2hI2HT7V4jv4wSUb0Ll/EiFXdjZBeX/94StaHvce7du1i1MiReNzuhlrdcXYJkEVjZxYOsSkdt93ap9MsJEAWopucmf3tPGf29+6SAEqPUnnyEFMaelvOoilJrxAt8jqsOC1azFxkh91O/phJHN+zg6pQ7FQKp02jzB/p7maKfq7Mr+O0NA0TYuYfb/0w/zis18+hODe1TIhx48Yxfvx4XnrpJfITHPiaqeveF0iALEQ3aW72986SANaKYwwfOoz4uLhG2wxTYdO0mDl/QpyhaRo5CXZ8zVSzmD9vNhTuZktR7DQLt81CeUBvdlU+IZRSlAWiLecfZ3x4/9q2dSvTp88AIKAbpMfJJGMR2+c//3lWrFhBstuK0YfvQfIUFqKbVIf0mLO/PygNkFx9uKG35Wwh3STVbWsSVAtxrvpSSbEfLpcvugCKD7K5oDrmdqtFQzeRZadFswJRk6hpxuwF3lniZ3Sqi7jTwXN5RQXlFRWMHj369B4yyVg0b+HChei6ztYN60hwWgn10ZpvEiAL0U1KfVHc55Q48kdNDlaEMIoOMCPGBD3JPxZtFWe34LZZYubwpSYn403LZNO2nZjNLTut1U8iFSKW2rCOFiMB+Uz+8eRzqldMnToVq8Ui+ceiVZqmNSw/nduH0ywkQBaiG4R1E1/EaFIDdHepHxXyEagsYey4sU2O0zQprC/aRtM08hOdzVazmDxtFv5juzhSGY653W3TKJNyb6IZpTEmGEML+cfT6ycch3RT8o9Fq84sHFJZeLy3m9IsCZCF6Aa+iNFMebcgltJDnDdpEnZb455iw1RYNWR5adFmKW4bzVVKuvyieVC4l83NrKrnslmoDOrNLlstBi/DVFQF9SYLHAHsKG6cf2wqVZ9/fLr+cVA3Jf9YtMrlcnHrrbfy5z89TpK7b6ZZyJNYiG5QEdSxx8o/LvGTUHWYGTOalncL6SYpUv9YtEOcw0q8wxpz1bzpE8diNcKs330k5rEWTQOlpNybaOLMBONY96IPShvnHx89ehRPnIesrKzTe0j+sWib5cuX88orr+CO+vBLgCzEwKeUotwfxXNODl5Ar88/jhbuZ9q02PnHaZJ/LNopL8GBPxJ7Vb38CVPZv3MLwWYePpqmURuWAFk0Vh0yYgbHId08nX/8YfWKrVu3NPQem0qhaZJ/LNomPT2dyy+/nP/+6x+4YyxG09s63KLzzjuvK9shxIAR1E1ChtmkgsWe0gBmXRkWU2fo0CFNjpP8Y9ERyW4bqpmJeAvmzUUV7GFHcexybx67hRKf1EMWjZX4IjHLu+0rD6KbMOWs/OOtW7cx7fSE45BukuSU/GPRdrfffjt/efpJkhx9L9WrxUShF154IebrSimKi4u7pUFC9Hd1YYNYy0/tKAmgFR9kxvRpTWaHS/6x6CiP3YrHXl/N4twlga+6cDbP/PYR1h8tZ06et8mxTqtGRUgnapjY+2APjuh5Z5aXjlVNZ0dxAAswIaM+QA6Fw+zdu5fvfe97QH3nQH6isyebK/q5sWPHMmHCBHasfYup58/r7eY00mKAfMMNN3DTTTfFrMkaCoW6rVFC9Gfl/ihua+z84/iKQ8y88NIm2yT/WHRGutuKL9o0QE6Ic5OYP5oNmzbBguFNjtM0DQ2NuohJilsCZNH88tLQNP949+7dDB8+/KwFjzQSnDJBT7TP7bffzj0//SmXXbygt5vSSIuf5MmTJ/PNb36TSZMmNdn25ptvdlujhOivTKUoD+okOBqnSgR1k/1lfmyFB5g69ZtNjgsZiuGSfyw6KNFpoSYae4jyvGmzWLttJ8W+j5IV3/QzZtWgKhglxS2BjYDyQOwJxuHT9Y8/Mi614bWtW7Y01HM/k38sq4CK9lq4cCFxDz9MWfEpcrMye7s5DVr8JD/66KMkJCTE3Pbvf/+7WxokRH/mj5gYiiY5eLtLA6iKQpJTUkhLTY15bLzkH4sOirNb0NBiLh19xUXzoHAfmwtqYx7rsVsolXrIgtPLS8eYYAxwrDrcNP942zamz6hfXlrqH4uO0jSNl19+mdzc3N5uSiMtBsgLFixgyJCmk4kAZs6c2S0NEqI/qwnraDEmTH1QGkAr3s/5MZaXPpN/LD0voqOsFo1kty1mLdGpo/KwxiXw7tZdMY91WC0Eo2afrEMqelYwahKJMcEY4FBFqFH+cWVVFWWlpYwZM6b+WN0kzSOjEKJjYqXy9rYWP81f+cpXWmz0Y4891uUNEqI/K/NH8dia9r7sKPbjKj/ErGtvbrJN8o9FV8iIs7G3TOfcTB0NjSETprJvxxb0Gy+MGfygKfwRI+bCEGLwqG1mgjHAgcogo1PdDfnH27ZtY/KUydisZ+53kn8sBpYW74YzZ85kxowZzJgxg5dffrnhv8/8E0J8SDcV1SED1znLs4Z1xYHiWqKlx5k8uWl5xJAh9Y9F5yU6bdBcube5c9BP7GJfRTDmdqfFQnlA0iwGu7JANOby0iHd5Hh1iPPOqn989up5DfWPHfIFSwwcLX7dW758ecN/P/roo41+FkI0VheuX33q3FGXo9UhzJLDDBk2ArfLHfNYyT8WneW2W3BYLeimatJLfPkF0/jzY9Ws2X2cSYvGxzy2zB9lTKqrTw51iu5nmIqKoE5SjHvR3vIghglTs+rTKxSKrdu28smbbgI+zD+WUTAxkLT5657cNIVoWXUoGnP4+mB5CIoPMOf8pqMuhqmwSP6x6AKappERbycQbboyXpLLTsLwSazfsCnmsTaLRtQ0CUYlD3mw8kUMTDP28tI7iwNo2of5x8ePHcdhd5CTnQ1I/rEYmOSpLEQXKfXrMQPdg1UhnGUHOT9GWlJIN0mV/GPRRVLdNiJG7DSLydNnUnJgOzXNLC2t0KiLyLLTg1VtOPby0gA7S/wMTXTiOZ2jvmXrVqafNeFYSf6xGIBaDJC9Xi8JCQkkJCSwc+fOhv8+87oQot6Z1afOXaghpJscL67E9Fc1zPZutN1QpErPi+giLS1VvmzhXCg9zKbjlTG3u60a5f5odzVN9HHFzSwvHdRN9pcHGZX6YXrYufnHFsk/FgNQi5/ouro6amtrqa2tRdf1hv8+87oQol5zq0/tLQ9ilp9g5LhJZ832bswrPS+iizisFrxOa+xyb0PSsKTm89a6zTGPddkslAf1mLWUxcAWMUzqwrGrmOws9qMrGJdWHyCHIxF27d7NlKlTAMk/FgOXfOUToguUB3ScllgPlwCUH2f+bMk/Fj0jK95BIEaAbNM0hkyYxt7tm1Exql1YLRqGWb/YjRhc6ppJuwHYWuzHYdUYkewCYO+ePQwbOhRvvBeQ/GMxcMmTWYhOOrP6lDtGoLut2Ie16iRzZkr+segZiS4rzXUCL5g3h9Dx3RyvDsfcrmmK2rCUextsKoM6Dmvs+9DWIj/nZXg4s7jelq1bmTZd8o/FwCcBshCd1NzqU76owf7Dx7FbIDev6RKakn8sukO8w4pFI2aqxCVTR4PFyqrNe2Ie67FZKZNlpweVlpaXLgtEOVkbYXrOh/WPt27Zwozpkn8sBj75VAvRSbVhI2YZxB3FAVTRXkaOGYcWI0FZKck/Fl3PommkuW0xS7ZlxTuIHzGZ9es3xDzWZdOoCukYpuQhDxZB3STUzPLSW4v8AMzIjgeguqaGU8WnGDduHCD5x2JgkwBZiE4qD0RxxRie3HbKh1a4mzkzpjbZZpgKq0Xyj0X3SIuzE2yh3FvR/u2E9KbbNU0DpeonnYpBoa6F5aW3FvtJcdkYmuQAYNu2rUw+bzI2W/0Xe8k/FgOZPJ2F6ART1a8+FSv/ePPRUrSqQiaMH9dkm+Qfi+7kdVqbW3WaxXOnQ00p6w8Vxdxu0TSqQxIgDxZl/ijuGF/wTRTbTvmZnh3XMAK2des2yT8Wg4YEyEJ0gj9iYiiaBLol/ijF+3cwZMxEHA57k+Mk/1h0J4/disumETGapllMy02E7DG8sfr9mMe67RZK/ZHubqLoA1r6gn+4MkRt2GDa6fxjperrH0v+sRgs5JMtRCfUhHS0GF1120754OQuFs6fF/M4yT8W3S0z3k4gRh6yy2Yhb/x09m6Lvey0y2ahLmLEDK7FwOKLGJgxvuDDh/nH07LrA+TSkhI0i6VhwnFIN0mRUTAxgEmALEQnlAaixMWY/b35RDWUHOCyRRc02WYqyT8W3S/FbSdGOWQA5s+dTaBgPyU1/tg7KA2f1EMe8GpCBs3Ft9uK/QxPcpLiqv8iv3ffXmZMn96QbhE8nSYmxEAlT2ghOihqmFQHDZzn5O+ZKLZs20ZS9lCSEhObHBeMSs+L6H7xp4e+VYxyb+ePzISUIbzy7rqYx9otUBWUZacHuhJfBE+M1fNCusmu0gDTsz8s77Zv775G+cdI/rEY4CRAFqKDzvSwnVvi7XBliOCRHUydOTvmcSFDycxv0e3sVgvJbmvMahVjUl3Yhp3H+nXrYx7rtlso8UmAPJBFDZPaSNMv+AC7SgPoJg35x1E9yuHDh5k6dSpQPwqmSf6xGODk0y1EB1UFo8TofGFLYR0U7uaaSxfEPE7yj0VPyYizE9CbVqSwaRqTps2icN92dKPpdofVQkhXhJrL0RD9Xl3EBKXFrOG+tciP3QKT0usD5L1795KRmUFiQgIg+cdicJAAWYgOKg3oMfOI1279AHtcAuOH5zfZJvnHoiclumwoFTuImTNuKMqTyHtbdjRztMIXlnJvA1VVMEpzt6GtxX4mZnhw2U6Xd9uytWFxEJD6x2JwkKd0BxTUhmPm9YnBI6SbBKMmDqvlnNcVR3ZuYcTkmTGPk/xj0ZPi7BZsVi3mynhTs+IgbxIr310b81in1UJ5QJadHqhKfLGXly4PRjlWHW5YPQ9g3bp1TJgw8ay9NLwOCZDFwCYBcgcU1UbwxyifJAaP+p61pkHHrlIf5okPuGh+0+oVIPnHomdpmka6xxaz3NvQJAfeEVPYu3UTKsZn2WO3UBaISmfAABSIGoQNhT1G/vH2U6fLu53OPz567Cj+QIARI0YAkn8sBg/5hLeTruuU1/ioDknPymBWHtBxWJr++azeeRDMKEvPnxTzOMk/Fj0t3WMnYjYNkDU0ZkwaQ1g3OHr0WJPtVotG1KgfKREDiy9iNrvS4pYiP0kuKyOSnQC88847LFq0qKEcnIyCicFCAuR2+u1vf8vzzz9PcZ2sNDVYKaUoD0Rj5hFv2rCelFFTcMcYupT8Y9Eb4p1WIHYwMzUrHpU7idfeWdPM0Rq1koc84JT7IzhtzSwvXexnalYcFjQUinfffZdFixY17BM2FOkyCiYGAXlSt9OVV17JunXrqAxECMsM70EpGDWJGCZWS+MHTHVIp+rAdmbMntPscdLzInqay2bBY7fEXnY6uz4Ped262PWQ3TaN8oCUextIDFNRFtBxxyjBc7QqTHXIYMbp+sd79uzF6XQxcuSIhn0UEC/5x2IQkAC5nf5XE48jIYVdH3xAnfSsDEr1PWpNg9w1+wqgrpxl82JP0JP8Y9Fbmlt2OjPOTvbIcVSXlVJaVtZku9tuoSKoY0oe8oBREzYwTJp8wQfYdqrx8tJneo/PrJ5nKoVF8o/FICGf8nZac7yWYPYk3l/9P0r9kmYxGJUForhjDE++veZ9bHnjGZcZH+MoyT8WvSfJZcNoJsidlutF5Y7n/feb9iJbNA1DgV+WnR4wiusiMRcHgfr846GJDtI8dnTDYM3q1Y3SK4JRk2QZBRODhATI7XTh0AT8GRPYsnEdx8vrpGdlkDGVoiKo4z4nj1ihOLRzMyMnz8Aa4+Eh+ceiN3kdVjS0mPeraVlxGDmTeHvNezGP1VDUhGVS8kCgm4pSfzRmD3BIV+wu8zf0Hm/fvp2MzAxyc3Ia9pH8YzGYyNO6nRYMTQBXHNkjx7Fh40ZJsxhkfBED01RNelAOFlcTLT7Mwnmxl5eW/GPRm6wWjWS3LWZFislZcZAzliMHD+Dz+Zps99islPklD3kgqAnpmIqY96HdZX4iBg31j8+dnAf1+ccyCiYGCwmQ22lEspNUlw3HqFm8t2YNlUF5cAwmNSEDS4zcvVfeXQfpw5kzIj3mcZJ/LHpbRpyNoN60BznRaWVUZhKu3FFs2rypyXaXTaM6ZKDHWGxE9C/FdZGG1fHOte2UH5sFzsv0EI5EWL9uHRdeuLBh+5n8YxkFE4OFfNLbSdM0zsvyUOAdxbGD+9h9vLi3myR6UFFdhLgYs7+3bVqPd+QUcuIdzR4rPS+iNyU6bTRX/HZqlodAxnjWrn2/yTZN01BK4YvIaFl/FjVMSvx6swHuliI/E9M9uGwWNm3axMiRI0lLTW3YHtKVjIKJQUUC5A6YlOEhYFo5b/pM3np3LYGoPDgGg0DEwB81cJ4TIIeiUcoP7WLG+bHTK6TnRfQFbrsFh9USsyd4alYcZu5EtmzdQiTadFTMqmnUhOQ+15/VhAzQmqaHAVSFdI5Whz+sXvHOOyy6aFGjfcI6MgomBhV5YnfAeZluABLGzWbte2upCcoElsGgMqjHXG7htbWbIT6NeWPyYh4X0pXM/Ba9TtM0MuLtMb/QT8rwYItLIC49l507djTZ7rFbpGpPP1dYF8Ftjf3IP1PebUZ2HP5AgK3btjJ//vxG+0j+sRhsujVArq6u5rrrrmPcuHGMHz++foGNykoWL17M6NGjWbx4MVVVVUD96mRf/epXGTVqFJMnT2br1q3d2bROyU9wMSLZSaErj7C/lvc/2NfbTRI9oLAuQnyMFfLeWfMe5E9iSlZczOPCOjLzW/QJqW4bEaNpD7LLZmF8mhvVzKIhTpuFurARc7ER0fdFDJOKYOzVPwG2nvLjdVoZmeri/fff47xJ5+GN9zZsrx8FUzIKJgaVbv2033HHHSxbtox9+/axY8cOxo8fzwMPPMAll1zCwYMHueSSS3jggQcAeO211zh48CAHDx5kxYoVfPGLX+zOpnVKWpyNieke9lWEWLjgAl59ey1ReXAMaP6IgT/SNL1CoTiycwtDJ80k0dk0eK7fR3peRN/gbeYzCjA1O47K1HG8t25ds+UrpWpP/1QV1EHVjyKcS6HYWuRnWpYHCxrvvvMuF110UaN9glGTZJdVRsHEoNJtAXJNTQ2rV6/mM5/5DAAOh4OkpCReeuklli9fDsDy5ct58cUXAXjppZe49dZb0TSNOXPmUF1dzalTp7qreZ2S4LQyMcPz/9u77zipyuvx4587d/rM9s4u1aUvsBQVFbEFxYaKaOwmakhi7IklMflFE42aRKOxfBNjYtQkaihSLDERCxYQBRcV6bKwFbaX6TP3+f0xsALb2dl+3q9XXtHdOzPPPO7c+8y55zmHiIKsScfx0UcfUeuV248DWbUvTEvXhi82byNsMnPsxKNafFzEkMiL6Dusuol4m44/3FLbaTfEp6Nb7Wzbtq3Z750WE4W1fpTUfu93yhpajx7vrg1S7Q8zLctFbW0tW7ZsYebMmYcc448oUuytf7kSYiDqtrDWrl27SEtL47vf/S4bN25k+vTpPPbYY+zdu5esrCwAMjMz2bt3LwAlJSUMHTq06fE5OTmUlJQ0HXvA008/zdNPPw1ARUUFFS20R+2MAykenRExFEOsQTJ0H+WBeIYPHcLKVas594SpXRpLf3Ekc9bfbSr3oZs0agKHrpL/t/ojUsdNZ2KcQU1t83nxhAzMQS9VlZU9NdQBYTD+jXVFZ+bLHAixpz5EwmELnnQdcix+UvOO5uOPPyYjI73ZY0u9BgmGp9W7Jf3FYPr7CkQUu/b6SbJpBFr4lr9hVx2pmo/RrjDvrV7DiSeeiM/vw+f3NR1T54uQbPNTUSF3wjpqMP2NxUpfm7Nu+2sPh8Ns2LCBxx9/nGOPPZabb765KZ3iAE3TWrzl05aFCxeycOFCAKZMmUJaWst1ZzvjSJ4j1/CQkeJhXXWYuVPz+c8Hn/Dd8+Z0+v30V7GY9/7CE4xgbWgg1Wlp9ru1H32Ib+p8po7MxNpC+9aIL0xmvGVQzVesyJx1TkfnyxYfpkbzkORofvofkt7I17W5VH+4iKuvuqrZ763OCHUmjaNS3f3+XDdY/r7KG4IkJPhIbuG/N8CGmgbscQkclZXGk++9y8UXf5ukxKSm34cNhcVpkGX3D5o5ixWZr87rS3PWbfd9c3JyyMnJ4dhjo6WvFixYwIYNG8jIyGhKnSgrKyM9PRqlyM7OpqioqOnxxcXFZGdnd9fwuizVaWZiuoM9dUGmzDyB9Z8VsLe2obeHJbpBa+kVZeVleOrrmJw3vsXFcZTC3UJbVyF6i9uqY9JoMc94WpaLSucQ6uobKC4pafZ7l1WnxhehTnKR+43i+gDuVtIrghHF53u9TMtysXffPoqKipk+fdohx3hDEVJdln7/hUiIzuq2K3dmZiZDhw5l69atAKxatYoJEyYwb948nnvuOQCee+45zjvvPADmzZvH888/j1KKtWvXkpCQ0Cy9oi+Js+nkZTgB+NprYcyY0Sz/z6peHpXoDsX1QeKszW8p/++9DzCGTGDakPgWHxeKKOy6CUcLjUWE6C0mTSO1lbbT+Zku0EwMnTCNtS1UswBwWUzsqvF39zBFDPhCBvUBA3sr56CvKrwEI4ppWS7effddZs2ahcV86J2yYESR0kr0WYiBrFuv3I8//jiXX345kydPpqCggJ/97Gfcdddd/O9//2P06NG89dZb3HXXXQCcddZZjBo1itzcXL73ve/x1FNPdefQusxt1RmWYCPeamJDmYdvnXgCy1a939vDEjHmCUbwhSJYW6gfuvqDNTB0EtOGtFzezRuKkO5qnpYhRG9LdVnwtVDubXiilUS7jjY0j4/WNO+qB9EocrUvQq1f6r/3ddW+UIt3vw7YUOZB12ByZnSBfPLJJzc7RiN6vRNisOnWr4X5+fl8+umnzX6+alXzSKumaTz55JPdOZyY0k0ayQ4zUzJdfFbu4eZzZvJ/f3uOPaXlDBuS2dvDEzHSWnpFbV0dZXt2kXjCNYxMtLX42JABSQ4zhqebBylEJ8XZ9Bab3mho5Ge6KCgZTqhwN7W1tSQmJjY7zmnW2FXjj1a+EH1WaUMQVxsVdDaUeRif5qCitJiG+nry8vIO+X0gbOC26c3KWwoxGMhffRekOi1MSHNQ649Q5ofp06fz72Wv9vawRIwopVpNr1i37mO0zNFMzUlCa3GpAWiqzbqzQvQWp0XHpmuttp2uDWmMnZTP2rVrW3y8y6pT5QtTJ1HkPssbjFAfiLSaXlHrD7Oj2t+UXjH7pNnopkOP9YYM0l3WnhiuEH2OLJC7IN6mMyEtmoe8oczDqbNnseyt93p5VCJWvCGj1fSKVe99SDg7j6lZLadX+MMG8Va9xccK0RekulpuOz11f0fIxNwpLXbVO8BpNrGrNtBt4xNdU+0Lo7eRX1FQ7gVgapaLd955p1lzEIhu5EyU+sdikJKrdxe4rDqpLgvZcRY+K/Vw9NTJ1AQUmzZv7u2hiRio8oYxmZpfYPx+P199+QUMGc+0VhbI3rBBhlsiL6LvSnGYCbUQQc5wWxjitlCXOobPv/gCv7/lDXluq06VNyRR5D6qqD7YZgWdDWUe3BYTqnIPuslEbm7uIb83lAJNk/xjMWjJArkLzCaNeJvO5EwXn+/zYiiN4084gReXLO/toYkuiqZXtFwe6bOCz7Cl5TA0PbnF2sgAhqGaNWIQoi9xW3Vaa4qXn+Xiq1oYM2YM6zdsaPU5HGYThRJF7nPa2lwM+9tLlzWSn+li9bvvcvIpJzdLFfOFDFKcZvQWggRCDAayQO6idFc0DzkYUXxV4eXkE09kxf/eIRKROqH9mSdk4I8YLV5gPvxoDb7Mia1WrzCUwmTScFlkgSz6LpvZhMuqE4y0XO7NF1aMmjyDNR+1XM0CoovsCk+I+oBEkfuSylbufh1QVBek0hsmP8PB6vdXt1i9whdWpLcSABBiMJAFchfF2XTGpzoxAZ+Vexg3agSujOF89FHruXui76vyhjG1kL8XMQzWrF1LZMjEplzNw/lCBikOibyIvi/NacbbQj3kyftrvJty8vh43TrCbXzhd5hNFNZIFLmvUEpR0kZzEIimVwA4a3aRlJjE0JyhLT0T8bLJWAxiskDuIrdVx242MTbVzoZSDxZdY9ZJp/Dy0ld6e2jiCLV1gdn81VdYXIno8SlMbm2BHFakSeRF9ANJDgvh5utjEu1mjkqysz3gICM9nU2bNrX6HHE2nQpviAbprtcnNAYN/GHV5gbhDWUehrgtbFz7foub88JG9PGONhbZQgx08tffRdE8ZBNTMl1sr/ZTF4hw/PHH8d/Va/B6vb09PHEE2kqv+GjNR2hD8xiX6sDZam1QibyI/uHAJi7VQjJyfqaTryq8zJg5s81qFgB23UShdNfrEyq9QfQ2bl6FDMXnez1MSbPx0Zo1zD7ppGbH+EIGadJeWgxyskCOgTSXlXGp0VuSn5d7GJKawugpx/Dmm2/28sjEkWgtvUKh+PDDNdSkjmu1vJtEXkR/YtFNxNtMBFroqjc1y0XYgNQxU1nz0UcoWtnRR3ShvU+iyL1OKUVpfajNyhNbKnz4w4q46u0MHz6c9LS0ZscEDYNUp7SXFoObXMVjIMGuMyLRisOssaHMg92sMfOUOSx+ZVlvD010UlvpFVu3bsMfikBSDtOGtNxBTCIvor/JcFvxtpBnkZfuxKxBuZ4KwK5dha0+h6Zp2HUTu2slitybGoIRAobC0kYIeUOZBxNQ8vnaFjfnASgl7aWFkAVyDLgsJswmE5MznHxW3oimacyYMYNPNm6ioqKit4cnOqGt9IrFixaRNeM0XFadMSn2Fh8vkRfR38TbdFQL9ZDtZhPj0hwU7PVy3PHH81Eb1SwgGkXe6wnRGJQocm+p8IQwt/PdfE1xA2MTTXy24VNmzZrV7PeBsIHbKu2lhZBPQAxYdBNxNhOT0t2UN4YpbQwS73Qw87S5LF8uNZH7k9bSK4qKi/j8i8+pzJ7B5Awn5lYixBJ5Ef2N26qDpkUbQxxmapaLHdV+pkw/ps1ybxCNItt0jUKJIvcKQylKG9pOryisDVBYGyC7bisTJ+aRmJDQ7BhvSJocCQGyQI6ZdJeV8enRqOKGUg9Oi4mjTz6DRYuX9PLIREe1lV6xZPESTppzJpVBE1OzWk6vkMiL6I90k0ayw4y/hTSL/Mzo33owZQQej4cv26hmARBn1dnbKFHk3lAfiBCOGJjbKC/5XmE9GlD51SetplcYSpocCQGyQI6ZeJtOutNCqtPMZ2WN6CaN8RMmUl5dx/bt23t7eKIDPKGWyyNVVVfzwQcfkJwfLYfUWoMQb8gg3S3l3UT/k+o04w83jyCPTbFjN2t8vs/Pxd++mBf/9a82n+dAFHmPdNfrcRWeUJuLY4Vi9e46JiYabNuyieOOO675MUqhaRpxchdMCFkgx4rbakIzwbQsNxvLvUSUwmzW+dY557NkiUSR+4NKb4iWSoe+8sornHraqayrVIxItJET1/LtR0MpEu2Sfyz6nwS7mRbSkDGbNCZnuCgo9/Ctb81h9549bNm6tc3nirPqlDUG8UgUucdEDEVZQ7DN9Iod1X5KGkKkVWxi+vQZOB2OZsf4wgZJ0uRICEAWyDFj0U3EWXUmpTtpDBlsr/LjMpuYcfIZLFmyBMNooRq/6DOi5ZGCuA9rD93o8fDmm29y0tx5bK70c9KIuFYfr2ma5B+LfslpMWE2RRdah5ua5aKkIURNEC6++KIORZGtJo3dEkXuMfWBCGFDtbmwfa+wHh3Y9cm7nHHGGS0e4wsr0l3yJV8IkAVyTKU6zYxJsQHwWXkjNrOJ9JzhxCWl8M477/Ty6ERbDnSfOrw80quvvsqMGTPY4o9GW04c3nxTC0QjL4l2c5u3OIXoq0yaRkqrecjRGu8FZR7OOGMuO3bsYMeOHW0+X7xNp6wxJFHkHlLeGMTWRuc8A8V7hfWMM1Xg83iYNm1ai8cpFPE2WSALAbJAjqlEuwW3zcyoJBsbSqO97k1o/ODmn3DPPfcQDAZ7eYSiNZXeYLP0ikAwyPLly7noootYXVjPUUn2VtMrfGFFulsuLKL/SnVZ8LXQMGREoo1Eu05BuQeb1cqFCy7kxRdfbPO5olFk2FMnUeTuFjYU+zwhnG00J9pc4afCG0bfuZYzzzyzxUo9EUNhMZnafB4hBhP5JMSQy2pCoZiW5WJzhQ9f2MBu1hg/4zhGjhzJM88809tDFC1QSlHWEGqWXrFq1Vvk5ubiTMtha5Wf2cNbTq8AibyI/i/OqqO10C1PQyM/00VBuReF4qwzz2LTV19RWFjY5vPF23RKG6SiRXer84eJGLSTXlGHJeJnZ8E65px+eovHeEMGaU5pciTEAbJAjiGrbsJt0clLdxBW8MVeDw6LiSpvmF/88h6efPJJysrKenuY4jAtpVdEDIPFixdz8cUXsXp3HQCzR7ScXnEg8uKSyIvoxxwWExaTiXALecj5mS5q/GG2Vfqx2+1ccMEFHYoi23WN7VU+VAs1lkVslDUEsbfRHSSiFKt31zOs5ivy86eQkpzc4nEBaXIkxCHkih5jqU4zIxMdWEz7W3ruL8CfOmQYV155Jffdd19vD1EcptIb5PDOrB999CHx8Qnk5eWxencDY1LsZLVSws0bil5YJPIi+jNN00hzWfCFmuchzxoeh8tiYtGmKgDOPfdcCgoKKCouavM542w61b4wFZ5Qt4x5sPMGI+zzhNv8cv7FXi+1/gierz7kzLPOavU4DQ23TTYZC3GALJBjLNFhQTdpTEx38llZNA/ZYtKo8oa46aab+Pjjj1m7dm0vj1Ic0JRecVD1CYXi5Zf/zUUXXURpQ4gd1X5mD49v9TmCRvTWpBD9XYrDTLCFijtui868scl8UNRAYW0Ap8PBeeefx8svvdzucybYdLZU+ghGpJJPrBXXBzGbaPPL+buF9Vhri4kEGlvdnBeMGDgtJuzS5EiIJvJpiDG31YS2Pw95d12QSl908VVUFwCzjV/+8pf8/Oc/JxwO9/ZQBdH0Ct9h6RUFBQX4/X6OO+443t9dD8CJbSyQlZLIixgY3DYdpVpebF0wPhm7WePFLysBmDfvPD5et47SdtLGrLoJQ0FhjWzYi6VA2KC4IUB8G+eekKH4cE89KaWfcFYrm/Ngf5MjKe8mxCFkgRxjVt2Ec389ZICCMi+6ScOqm9hW5ePss88mKSmJ559/vpdHKiCaXnH43cl//3sRF110ESZNi5ZGSrWT4Wo5QhyMGDgk8iIGCLvZhMNiajHaG2/TOW9cMu8V1rOnLoDb5eKcc87h5Zdeavd5E+3RIEF9QAIDsVLeGMKE1uqiF+CzskYaPB5qtqxnzpyWN+dBtBJGkkPugglxMLmqd4NUp5nMOAvxNr0pzSLOplPlC1PpDXPffffxyCOPUFVV1csjHdyizUFCuA6qXrF9x3aK9uzh1FNPYU9dgF21AU5uZXMeRCMvGRJ5EQNIusvcYh4ywAXjkrHpGi/tjyJfcMEFfLRmDXv37WvzOU2ahtNiYmulD0M27HVZ2FDsrvW3GT0GeK+wAVtRAfn5U0hNSWnxGGlyJETLZIHcDRLtZgxDIz8zmoes9pdOOpCLNzJ3NAsWLOCBBx7o5ZEObg3BCAHj0PSKRYsWccH8+VjMFlbvT6+YNaz18m5hA4m8iAElyWGhlfUxiXYz54xJ4p1d9ZQ0BImPi2Pu3Lks+ve/231el1WnPhChvFHqwXfVPk+QsKHabEzkDys+KqrHVriOc9rYnOcLGyRJkyMhmpEFcjdwW/WmesjV/jC7a6MXhAO5eLtq/Nx2222sWrWKzz77rJdHO3iV1AexHvQJKC0ro6CggDPPPBOA1bvryUt3kNrKBrwDpaviJP9YDCBuqwm01qO8F05MwaLTFEW+cP583n3vPSorK9t97iS7me1V/hY79omOMZSisKbt3GOAT0sb8e3dgyngYWorm/Mg2uQoTe6CCdGMLJC7gc0czUPOS3cBsKGssel3SXadovogyurkpz/9KXfffTdGC7vGRffyhw3KG4KH3FZcsngxZ511Nk6Hg8LaAHvqgsxupbV09DkUSQ5dIi9iQLHqJuKsequL2GS7mbNGJ/H213WUN4ZITEzk9NPnsHjx4naf22zS0ICvq/0xHvXgUeML4wsprG20loZocxDrrrWcc/aZ6Ka2j02wywJZiMPJArmbpDnNxFlNZMdZ2LA/Dxmi5XjiLNFUiwvmX4iu67zUgU0uIrb2NobQtG82uNTU1vDe6tWcf955ALxXWI8GnNhG9zxvOEJaK5v3hOjP0l3WVvOQARZMTMGkwcsHosgXLuCtVauorqlp97kTbDqlDUFqfLJhr7OUUnxd449G+dvgDRus3VWJUVjA3DPOaPW4yP40DWlyJERz8qnoJol2MyFDMTXLxRd7vQQj39yydFhMeEIRyj1h7r//fh566CFqa2t7b7CDTEsbXJYvW87JJ51EYmIiCsXq3XVMyXSS1GZkRSNB2kuLASjBrmO00Hb6gFSHhTNyE/nfzlr2ekKkJCdzyimnsKQDUWRN04jfvx8j0kLXPtG6ukCEOn8ERzsL2rXFDYS+3sC4iXmtbs6D6CbjFGlyJESLZIHcTdxWHU3TmDbETSCi2FzhPeT3STYzO6p85I6fyJlnnsnvf//7Xhrp4FPlDRFW32xw8Xi9vPb661y44EIAdtYEKGkItVn7OGIodA1c7URyhOiPoqlHWpstoi+amArAok3RKPJFF13Em//9L7V1de0+v91swheKUNogG/Y6Y09tAGcHor3vFdZj/notF51/TpvHBSIG6dLkSIgWydW9m9jMJmy6xvhUO2aNpooIB+gmDYuusb3Kz+23387y5cv56quvemm0g4dSisLaAO6DSru98frrTJs6lazMLADe312PCThhaOvpFb5wtL10WzVIheivzCaNJLsZf7j1BXKGy8K3jkrkPztqqfSGSE9LY/bsE1n2yisdeo0ku5nt1T68oUishj2geYIRKryhdtMhGoIRPv1iM5aQhxnTZ7TzrNLkSIjWyAK5G6W7LOiaxum5iby5o5a9jaFDfh9vM1PpCWHY47j99tu5++6724zYiK6r9UfwBCNNjT2CoRDLli3joosuAqJtpt8rrCc/y0ViG+kVgbAizWXtkTEL0RtSXWa84bYXr5fkpWIYsPiraE33iy/+Nq+9/joNjQ3tPr9u0rBoGjurZMNeRxTXB7GYtHbTIT7c00Bk21pO+dbpbW7OkyZHQrRNPhndKMkRzUO+dFIqmgb/+rJ5GaQEu86WSj8XXXIpHo+HZcuW9fxAB5GiusAhF4S3336b4cOHk5ubC8D2Kj/ljSFmt7E5D0Ah6RViYEuwmVG0vRjLdFs47agEXt9eQ7U/TGZGBjNnzuzweSzBbmafN0SlR1It2uIPG5S001b6gHe270PbU8Cl57de+xikyZEQ7ZErfDdyWXWU0khzWjh7TBL/21FL8WE5d1bdRMRQFNVHN+zdd999NDY2tvKMoiu8h92iNJRi8aJFXHTxxU3HrN5dj67B8UNbzz8ORRR2s4bTIrcmxcDlsprQNdrtfHdJXiqhCCzZFI0iX3LJJaxcuRKP19vm4w6It+psrfQTlg17rSpvDLbbVhqg1h9m49oPGHLUONJS09o8NiJNjoRokyyQu5HdbMJu1ghFFN/OS8Wia/xjY0Wz45LsOnvqA4yZNJVZs2bx6KOP9vxgB4HShkNvUa5Zswany8mUKZMB9levqGdalrvNSI03FCHDLRcWMbCZNI0UR+ttpw/IjrNyysh4Xt1WQ60/TPaQIUyfPoOVK1Z06HVsZhNBQ7GnVlItWhKKGOyuDZDQgejx+7sbYMdazj/37DaPUyra31WaHAnROlkgd7M0lwVfOEKS3cx545J5t7CewtrAIcdomobbYmJrpY+7fvozXnrpJXbs2NFLIx6YghGD4oYgcfsbgxhK8fLLL3HxRRej7b+NvKXSzz5PuN30ipBEXsQgkeq04Gtjo94Bl+SlEogoXtlSDcCll17KsmXL8Pl9HXqdRJtOYW0QT1A27B2uwhvCMKI52+35zydfoAcaOOuk49o8TpocCdE+WSB3s2SHmeD+W4cLJqTgNJt4YeO+Zsc5LTqNoQgRRyI33XQTv/jFL2TDXgxVeEIo9c1F5u1VqwA4/oQTmo5ZXViP2QTHDWt9gayUAk0RJ/nHYhCIs+lobbSdPmBYgo3Zw+NYvqWa+kCEYUOHMmnyZF599dUOvY5u0rDpGturfHLeO8iBttJxtvbPNxXeEDvXvUve8ae02znPG46QLk2OhGiTXOW7mdOio6nooizepjN/QjIfFjWyvap5ZCXJZmZbtY9LrriaoqIiVq9e3dPDHZCaLjL7F7WNHg9/e/ZvXH/9j5py+gwU7++uZ8YQ9yEl4A4XiCjirTqWdtq8CjEQOC0mzCZThxp6XDYpDX9YsWx/FPmyyy5lyZKl+P0dS52Is+lU+cJUeELtHzxIVHlDHWorDbBq2z4oLODK+ee2e6xSmrSXFqIdcpXvZg6LCdv+PGSA88cnE2c18fzG5hUtdJOG1aRRWB/mjjvu5IEHHpBoSgzU+ML4w99cZP71r39y9IyjGTd2bNMxX+3zUekLM3tE65vzINrCNcMt5d3E4KBpGqlOM9528pABRiTaOGGom+VbqmkMRhg5YiR5eXkdjiJDtA315kofXkm1QCnFrg5GjwHe+N87uHJyyRuR1eZxYUNh0bUONRwRYjCTT0gPSHVZ8IejFxi3RWfBhBQ+KW1kU0XzKHK8zcw+T4hjT5mDUorXXnutp4c74Oyq8ePeHz0u3L2bVW+t4rvXXHPIMe/trseqw8yctvOPDYM26yMLMdCkOS0EIu0vkAEunZyGJ2SwfGs0inzF5ZezeMmSDkeRrboJq0mjoNzTdM4crOoCERoOqtnelrLGEHsLVnPCqae3e2xjMEJ2nEWaHAnRDlkg94Bku5mg8c3J/rxxySTadZ4vaJ6LDJBo19leHeT2O3/KQw89RDgc7qmhDjj1gTB1gQgOiwmF4k//939cdvnlJCYkNB0TUYoPdjdw9BA3zjYuRsGIgd2sNS22hRgMXPvbTndEbpKdmTluXtlcjTdsMGLECCZNmsTKV1d26vUiSvHlPs+gLv22uzaAo4NNPF756HPw1XPpGbPaPTZkIE2OhOgAudL3gDibjuKbeqJ2s4lL8lLZuNdLQbmn2fFW3UTYUAzJO5r0jAz+/e9/9/CIB47iuiA2PXpxf//996mtq+Pssw8tgfTlXi81/jCzhye09BRNGkMG2fG2djtZCTGQRLutfZMm1p7LJqXSGDRYsT8X+fLLLutULjJEm5Q0Bgw27/O2W4d5IGoMRqj0hnBbO1aG7Z3/vknq5FlkxdnbPC4YMXDIl3whOkQ+JT3AZjYxJM5KQ+CbvLozRyeR6jTz94J9RCtSHirJrlPaEOJ7t97FI4880qmLi4jyhQzKG6Ol3fx+P08//Rd+9KPrMeuHXnRW767Hpmscm+Nu8/kMpUhxSnqFGHzSXRa8oY7lBY9JcTBjiIulm6vxH2EUGaIVgPZ5Q4OyskVxfQBrB0uwbSuvo2Hbp5w5d267x8qXfCE6ThbIPSQn3tZU7g3ApmtcNimVLZV+Pilp3jlP21+k35I1mrGTp/H888/35HAHhL2eIJoWncuXXn6ZvLw8JuVNOuSYsFJ8sKeBY3Lcbeb6RSMvelMXPiEGkySHmc6kBF82OY36QIRXt9UAcPnll7Nk8ZIO10U+INVhpqguyO66QPsHDxC+kEFZfbDDTTz++er/IG0kZ+aPaPdY+ZIvRMfJ1b6HuK06yQ7zIYXwTz8qkUy3hecKKluMIusmjTiribOvuYnHn/oTDQ0NPTnkfi1sKPbUBki0mSkpLeW1117juuuua3bc53u91AUinNROekVDMEJOvFUiL2JQit7q73gUd0Kqg/xMJ4s2VdEYjDBi+HAmT5nMypWdiyIfCBRsr/JT1hDs5Kj7p7KGAGh0aBOdQrHh/bcYeczJJLezeVi+5AvROfJJ6UHDE+34DgrDmE0aV0xOZWeNnw/3NI8iQzRfOWfoUKacNo8//fnPPTXUfq/SEyKiFLpJ489//hMXX3wRqSkpzY5bXViP3awxY0h76RWQ4pDIixicbGYTLqtOoBNh5GunpVMfiPD8xgoALr/8CpYuWdrpKLJuii6Sv6rwUu0b2BuWQxGDPXVBEm0dO9e8tvoTQo11nH1y253zQL7kC9FZskDuQUl2HbvZRPCgkkmnjEwgJ97K8xv3EWklzy7RZubsiy7nmcWvUVVV1VPD7beUUuyqDeC26Kz9+GNKS0o5//zzmx0XNhQf7qlnZk4cdnPrF41A2MBt1XF2cMOMEANRutuCpxML5NHJDs4ek8jKrTXsqPYzfNgwpkyZwooVnYsiQzSYkGDTKShvPGQvx0DRGIzwdbWPtcWNaFrH2krX1tXxzBN/wDzzYk4ckdTu8Qr5ki9EZ8gCuQdpmsbIJDv1B53gdU3jyslp7KkL8l5hfauPHT88i2PPuIAHH/9TTwy1X6v1R/CFImCE+fOf/8wPfvhDLObmbVULyj00BA1Oaqc5SHRji5RFEoNbusuCYahOVZX4Tn46CXadJ9aVYaC47PLLeWVp56PIEK3u4zLrFJR7OrxhsC/zhw1KGwJ8XNzAx8UNFNUHcVlMJHWgzrpC8dvfP0xw6FTmnDiThHbylQNhA5dFvuQL0RmyQO5haS4Lukk7pHXriSPiGJlo4x8bK1qt+2nSNK66+AKWf7iR7YV7emq4/dLuumj90KVLlzBy5AhmTJ/e4nHvFtbjNJuYntV2eoVS0R31QgxmTotOdrz1kC/47XFbda6bls6WSj//3VHbFEVevnzFEY3BYTFhAj4v9x5yJ66/CBuKKm+IjeUe1uypZ2ulHxPRZixJdjPmDlauWLFiJYVlFRiT5zJ/QvPUscM1hgxy5Eu+EJ0iC+QeZjZpDEuwUhv4JpfOhMZV+WmUNoZ46+vaVh+bnpLM6aedyq/+74VBXUC/LZ5ghCpviMbaKpYufYWFC7/f4nHBiGJNUQPHD43Dqrd+UfKHDeJsJpwWibwIMTTeRtigU2XXThuVQF66g79+VkFdIMLlV1zBK0uX4vV1PooM0brywYjBpn3efnEeVEpR5w+zvcrHh3vq2VjuxRuMkOwwk+IwY+tgM5ADvt61i3/88x8Ejr2MmcOTGNqBha+hopVIhBAdJwvkXpDptqHUoReZmTluxqbY+efnlQTbKMh/2YILWFfwBW9t2DLoaoN2RGlDEItJ4y9/+QvnnXcemRkZLR73XmEdnpDBqUe1Xb3CI+kVQjRxWnWy4iydiiJraPzomEw8gQh//2wfw4YOZerUqaxYcWRRZIi2e6/1R9ha6euzjUQihqKoLsCaogbWl3kobwgSb9VJdZpxWfUj2izn9/t54IEHOPqcy/E4Ulkwsf3osT9skCBf8oXoNFkg9wKHxUSm20pj8JtbhNr+KHKFN8x/dtS2+li3282Cs8/gqRcWUVI/OMoedVQgbFBcH2Tn5i/YsX07CxYsaPE4hWLJ5mpGJNqYmuls8zmVgiR78/xlIQarYQk2Qp2MIo9MtHP+uGTe2FHLlkpfUy7ykUaRIbrhrLwxyNfVfa+JUjBi8PleD9uqfFh1jVSHmQS7uUOb79ry9F+e5qijctnknsC4VDt5aW2fv0D2UAhxpGSB3EtyEqz4D8uhm5blIi/dwYtfVOJvY7f4efPm8fWmz3hj3Sbq/AO77FFnVHhCRCJh/vR//8fC7y/EbrO1eNxn5V4KawPMH5+MRusXLF/IINGh45C6oUI0cVl1MtxmGoKd2yh3RX4aKQ4zT6wrJzsnh6nTpnUpigzRRXJhbYBdNf5D9nX0Jm8wwvpSDw2BCGlOC1Y9NuePDz78kA0bPmPaeVex1xPmoomp7T5GRW9VkuSQL/lCdJZc+XtJvM1Mgk3HFzo0inz1lHRq/OGmDlQtsdvtXH7ZZSx+8R98sdfbqdqkA1VDIEJhbYB333ydjPR0jjuu9bqgS7+qItGuc/KIdtIrwgZD3BJ5EeJwwxPtBCKqU1Fkp9nE96ZnsKPaz+vbarnssst4ZelSPF7vEY/DtL+RyK6aaEfS3g4Y1PnDfFraiFKKxA5Uo+iovfv28cTjj3PnnXew8msf2XEWZua0vbkYwBc2SHSY2+wSKoRomXxqetHIJPshaRYAkzKcTMty8fKmKjyh1he+Z5x+OlV7S/j8yy/YXOHts3l43a3OH2ZjuYd1JQ3U1taw6OWX+MEPftBqZHh3bYBPSz3MG5vU5uY8pRSaUrKxRYgWuK066U5Ls/NXe04aEUd+ppO/F+wjLi2LadOns7KLUWTdpJHqtKABn5Q2sq3S1ysVLio8QdaXNmI3m/Z3HoyNiGHwu9/+lvMvuIBg4jC2V/uZPyEFvQM5zL6QIjtOvuQLcSS6dYE8YsQIJk2aRH5+PjNmzACgurqaOXPmMHr0aObMmUNNTTRSqpTipptuIjc3l8mTJ7Nhw4buHFqfkOQwYzdrzU7m38lPoyEQ4e8F+1p9rNls5sorr2TR889S5Quxq6bv5eF1F6UU1b4w60sb+KS0EU8weivz5eefZe7cueTk5LT62Fe2VGPVNc4a3XZhfV/YIMlh6fQOcyEGi+FJtmZpYu2JbtjLIhA2+Ov6fVx22aW88sorXYoiH+CwmEhzmCltCPJxcSMVnmCPbGRWKroZb2O5lwRb7KO1L774ImazmYsvvpglm6tJsOmcNjKxQ+NSxDaSLcRg0u1X/3feeYeCggI+/fRTAB588EFOO+00tm/fzmmnncaDDz4IwBtvvMH27dvZvn07Tz/9ND/84Q+7e2i9zqRpjEiyUX9YFGZMioNzxyaxcmsNmypa38Ry0kknEwyF2PrZp+yqCVDpGdib9gwVrSH6aWkjBWWNBMOKdKcFt1Xny02b2LhxI5dddmmrj6/1h1n1dS1zRiW0e9HwhgyGxEnenhCtibeZSXVaaOxkLvLQeCsXTkzhrV111FlTmD5jOitWLI/JmDRNI9lhxmHW+Lzcyxd7vd3aVMRQih3VfrZW+khxmLG0cVfqSHz55Ze8+uqr/OT22ymqC7KupJFzxya12fnzAG/IIMUpX/KFOFI9/slZvnw5V199NQBXX301y5Yta/r5VVddhaZpzJw5k9raWsrKynp6eD0uzRW9NXj4BpPvTk0nzWnm0TWlrZZ9M2ka3/3ud3j++b8Tb9X4fJ+313PwuoOhFPsao1GhjeUeDANSnRZcVh2F4tXXXuNXv7qX6390PQ67o9XnWbm1hpAB549vuzSSUgo0SJT0CiHaNCLRhu8I9kBcmpdKmjO6Ye/iSy5l2SvLYhJFPsCqm0hzWagPhFlb3EBxXSDmaWihiMGXe70U1QdIc3a9QsXhGhob+O3vfsstN99MakoKizdXYdU1zhnTfltpAF/EkPQKIbqgW1cAmqZx+umno2ka3//+91m4cCF79+4lKysLgMzMTPbu3QtASUkJQ4cObXpsTk4OJSUlTcce8PTTT/P0008DUFFRQUVFRZfGeCDFoze5IyGKKkLN2oX+YJKL/1tXzr8/LeTssYktPjY3dzSZmZm8u+otpsw4mre31DMuyUKqs/v+0/bUnIUNRZUvwu76EMGIwmU1YdM1/EHwA1WVVfzrxX8R8Ae49957yczMpKa25bEFw/DhtlJOyrThNjzU1HpafV1vyMBhNlFXHYrJ++gLf2P9jcxZ5/TmfJkCAUo8Bs5OVnu5doKTZ9bvZX1lMsefcAIrVixn7ty5sR+gofjk6zpcFhO5SVbirKYuz1cgoviqKoAvpEi0m6gNxGis+ykFf/vb35g1axZjx42lsLyCzwvLOXtoHMrfQHsZdUop6gMGQWeQCm9sFu7ymewcma/O62tz1q0L5A8++IDs7Gz27dvHnDlzGDdu3CG/1zSt08XSFy5cyMKFCwGYMmUKaWlpXR5nLJ6jK1wJEeqLG0l0HFo8/oTEJD7cp3hxRz3Hj3UwKtHe4uMvvuhifvu733LqaaeRajJT4g/jtDgYlmA9omL0HdHdc1baEGBndYCQMshIcR9SKslQitdee5UXXniBBQsWMH/+hZj1tjfFvLGjhsKAlR/mDSUp0dXmsRFfiLx0F6mu2KVY9PbfWH8kc9Y5vTVfU+PCfFrSSFInPy8nJSayqtTgha0efn3WPH599x2ce+483K62P59HIpVol81dfoPhdhvuBHXE89UYjLCtzIPTbSPT1j2X0Df+8wY7d+7k0UcfxWa1svSzfeyLOJiXP4Ikd/vz3BiMMDpZZ0hGbOdSPpOdI/PVeX1pzro1xSI7OxuA9PR0LrjgAtatW0dGRkZT6kRZWRnp6elNxxYVFTU9tri4uOnxA53TqpPhMre4I3zh9AxcVp1H15QRaeUWYV5eHiNGjGDlypVY9GjZox1VPrZU+vpMbdDO2F0b4Kt9PpxmE6mOQ+uIlpeX89O77mLVW6v4/e8f5uKLLm53cWygWPpVNUcl2ZncTmMQQylQGgl26TolREck2HSSHDqeTuYia2j88JgMIoZiZZmZo2cczfLlsclFbonLqpPiNFNUH+SzvX52VPkorg9Q4QlR4wvTGIzgDxttnjOrfWE+KWlEN0VzsLvDnqIinn32We666y5sVivesMHr22qYNSyOrA4sjiHaPS9T0iuE6JJuWyB7PB4aGhqa/vm///0veXl5zJs3j+eeew6A5557jvPOOw+AefPm8fzzz6OUYu3atSQkJDRLrxjIchJa3hGeaDfzw6Mz2FblZ9mW6lYff91132PpkiX89W9/wzDCpDrNlDWG+Hyvp1/VSd5dG2BHdfMNL4ZSrFi5ghtvuomjjz6ahx95hGEHpeS0ZX2Jh6L6IPMntN0YBKLpFRkuM5YYFfcXYqDTNI2RSQ68R3CeGeK28u28VFbvbiB/zvksX76MRk/r6U9ddaBuslXXKG8MsqPKzxf7vBSUe/ikpJE1exp4b3c9qwvrWFfc0NQNr6guwJ46P5+VNRJn7b62zYFgkAceeIDvfOc7DB82DIA3ttfgCRksmNB+W2mInitNmkaCVK8Qoku6bRWwd+9eZs2axZQpUzjmmGM4++yzmTt3LnfddRf/+9//GD16NG+99RZ33XUXAGeddRajRo0iNzeX733vezz11FPdNbQ+KcGmE2fVW+ygd/KIeI7JdvNcQQVljS3nxQ4bOpQnn3qK3bsLue22H1NSWkqqw0xDIML6/aXQ+rrdtQG2V/lIPqwla2lZGXfeeSdvv/0ODz/8MAsWLEA3dfxPd8nmKlIcZmYPj2/3WIm8CNF5SXad+MMaH3XURRNTyXJbeHmPiWNmzuS3v30ophv2WmLVNeJtZpIdZlIdZlIO/M8Z/fc4q45Ji3bFq2gM8XW1n53VfpLs5ph1xjucQvGXv/yF7CFDOPPMM4HoPoxXNleTl+5gbGrrG5AP5gkapLvMmGO8aVCIwabbFsijRo1i48aNbNy4kU2bNnH33XcDkJKSwqpVq9i+fTtvvfUWycnJQDQK8eSTT7Jz506++OKLprrJg4WmaYxItLVYMklD48ZjM9E1eGxtKYqWbwEmJiRw7733MmfOt7jttlv5z5tvkmDX0YBPSxup8fXdChcHFscpjm8Wx4ZSLFu+nJtvvpmZM2fy8MMPdzhqfMDXtX4Kyr3MG5uEpZ0LhkRehDgymqYxKsmO5whKqtl0jeuPzqS4PsiQUy4jNTWVW2+9hdJerGKkmzSsejRSHGfTSXKYSXFYum3R6fF6+c1vHmDzV19x8y03N93pWr27nkpvuMPRY4huIMyMs3XLOIUYTOQ+ch+S4rRg1k2EW8iBS3NauHZaBgXlXv67o67V59DQmHfuPH7729+xbNky7r//NxhBL06ziQ1ljZQ19L1ayXtaWByXlJZyxx138N577/GHRx7hwvnzOxU1PmDpV9XYdI2zOlAaSSIvQhy5ZIcZVyt3wdpzdLab44e6eWlzLd/+7g+Zd+48brv1Vj4r+KwbRtq37CrcxU033ojb7eaRP/yBOHccEI0oL9pUxdB4K8d0oK00RMuF6iaaVUQSQnSeLJD7EN2kMTLRRl2g5SjMWWMSyUt38Jf1e6lup97xiOHDeeyxx0hNSeGHP7yerZu/JMluZtM+LzurfX2mNfWe2gDbDlocKxRv/OcNbrnlFo4//jh+//vft9kZry1VvhDv7qrj9NxE4jrQ+lUiL0IcOU3TyE22d7pxyAHfn5GJBjyyppQzzjqbn/7sZzz00G9ZtmxZq3fN+ru3Vq3izjvv5NJLL+Xmm27CZv0mvWtDmYddtQEunJCCqZ29Ewd4QhEy3daY12QWYjCSBXIfk+6yoFAtLmBNaNw8cwiBiMFT68rbfS6b1coPfvADbr75Jh588EH+8fxzJFqhsDbAVxXeFiPVPamo7tDFsd/v5+GHH2HZK8t4+OGHmX/BkUWND1i5tZawggvGJ7d7rERehOi6JIcZh+XIosgZLgs3HpvJxr1eHltbxuTJk3j0D3/gP2++yaN/eJRgKDZ1yfuCQDDIY3/8Iy/+61889NBDfOtb32p2zOJN1STbzZw6MqHDzxuMKDI6WOlCCNE2WSD3MTaziZw4G/WtRJGHxlu5fHIaH+xp4MOihg4959Ezjo7md3/9Nbf/5CcEayuo8IQpKG88ogtZLBTVBZras+omjeLiYm659VYMI8Jjjz3W6Vzjw/nDBq9tq+b4oW6GuNvfdOcJRchwWyTyIkQXmDSN3GQbDYEjO698a1QiV0xO5a2v6/jHxkoyMzP5wx8eoaGxgTvvvLPVRkD9SVl5GbfdeisNDQ388fHHGTliZLNjdtT4+azcw7xxSVg72L46YijMuol4+ZIvREzIArkPGhJvJdRGdHfBhBRGJtp48uPyDt/OTEpM4le/updTTj2VW2+9hfUfvI0vZLCuuIHdtYEj2n1+pA5fHL//wfv8+Mc/Zt68c7n99tux21tuiNIZb31dR0PQYH47baUPCEYUGS6pXiFEV6U4LTgs2hGXl7x8cirfGpXAP7+o5L87a3HYHfz8579g2rSp3HTTzWzfsT3GI+45a9au4Zabb2HOnDncfffPcDlbrsu+ZFMVdnPH20oDNAQjZMdZMHVTcyghBhtZIPdBbqtOitPSamk2s0njtuOyqPGHeWbDvg4/r4bG+eedx0MPPcTSJUt54uHfooJevq7xsaaogc/LPVT7wt2an3zw4tgwwvzpz3/mr8/8lfvuv4+zzjyr3TrFHWGgeGVzFWNS7ExMb780UsRQmE2aRF6EiAGTpnFUsp2GI8xF1tC4ZWYWUzNdPLa2jA3lHkyaxpVXXMn3Fy7k7rvv5t333ovxqLtXOBLhr3/7G08++RS/vOcezj///FbPdXs9Id4rrGdubiLuDuydaHoNQ5HqlPQKIWJFFsh91IhEG542IjCjUxxcOCGZ/+yopaC8c4X1R44YyaOPPUZiYiI33XADpTu3kLK/E1ZBWbRYflFdIObpF0V1AbbsXxxXV1dy++13UFpawh8ff5zRuaNj9jofFzdS0hBi/viUDi24G4MRsuJkY4sQsZLqtGDTTQRbaH7UEWaTxs9PymFogo1fv1vE17V+AGbNmsUDDzzIs3/7G3//+9/7zGbjtlTX1PDTn/6UHTt28MQTTzBh/Pg2j1++OdoQ6oJxHS/tFjYUVkmvECKmZIHcRyXazWQ4LTS0kosMcMXkNLLcFh5bW9bpxazdZuNH11/PDTfcwG9+8wDPv/ACNlM0AmE3m9hZ7eejPfVs2uuh1h9GdfFCVLx/cZzqMFNQ8Bk33ngTM2fO5J577iU+Lq5Lz324pV9VkeY0M2t4x543bCjSXBJ5ESJWdJPGqGQ79V1oUOSymPj1qUNxWXT+36oiKrzRTXpHjRrFY3/8I19u2sS999zT7U1FuuLzLz7nxhtuYPLkSdx3330kJrS94a4xGOH1HTXMHhHfqc12jcEIQ+IsaJJeIUTMyAK5DxuVbMcfabmiBYDdbOLmmVmUNYb4x+cVR/Qaxx5zDE8++STbt23jJz/5CaVlZVh0jeT9naVq/RE2lDaypqiBkvpAs7zCiKEIhA28wQj1gTDVvjCVnhBlDQEKa/xsq/LxebmHzZU+ku06L774Lx5++Pf89Kc/5ZJvfzvm+XLbq3x8sc/HeeOSMXfgucOysUWIbpHusmA1HXkUGaL13+89dSieUIT/93YRnv17JRITEnjggd+QmprKLbfcwqfr1/epUnDhSIQXX3qR3/zmAW697TauvOLKDlXkeW1bDf6wYsHEjkePAcIGpMkeCiFiSlqG9WEuq87QeCtljUGSWunulp/pYm5uIku+quak4fGMTulYO9KDJScl8atf/5rly5dxyy03s/B7CzntW6ehaRpxNp04dIIRg21VfrZW+TC8fpy+eoIRA0NpmLRoUfvo9UlDaaAphW7S0DUN3QSWkId7fvM7AoEAjz/+BCnJ7ZdeOxJLN1fjMGvMHZ3YoeMbgxGy462ysUWIGNNNGmNS7Hy5z0uq88hjMUcl2bn7pBz+36oi7l9dzK9OGYrZpGExW7jxxht59913+cvTT/NXXeeiixYw+8TZmM29d2krLCzk4UceweV08thjj5GRnt7uYwwUm/b6WL6lhvxMJ7lJHd+oHIwY2M0abqvEu4SIJVkg93HDEm2UNAT31+lteRF33bR01hU38vCaMh4+YwQuS+dPlCZN44LzL2Dy5Ck89NCDfPLJJ9y4v7sTgFU3keIwYShFhTcavXZZTB26pbdlyxbu/81vOGn2bL7z3e9i1rsnWlvhDbF6dz3zxibjtnTsNcIKSa8QopukuSwk2s00BiOd2nB2uBlZbm6amcWja8v448fl3HpcZtP+gpNPPpmTTj6JTz75hEWLFvPss3/nwvnzOWPuGTjsnQ8YHKlwJMLiRYtY+spSvvOd73DmmWe2uweiqD7Iqq/reHtXLfs8YexmjSuntL+gPlhjyGBkol3SK4SIMVkg93F2s4lRSXa+rvGT4mj5P5fbqnPb8Vn8v7eL+OU7e7jv1GHYzUcWTThq1Cj++MfHeeYvf+H6H13PHbffQV5eXtPvTZqGVdfabcdcVl7G2rVrWbNmDYWFhdx88y2ccPzxRzSmjlq+pQZDwXnjOhadDkUUNpPWoS57QojO0zSNMakOPi5uwGkxdelOzdzcRPZ5Qvzri0oy3WYum5T2zeugcczRx3DM0cewZetWFi9ezD//9S/OPvtszjtvHkmJHS+XdiR2Fe7i4YcfIS4ujscff6LNqHGtP8x7hQ28vauWrVV+NGBqlour89M5YWhcp87dSikihiLNKZdyIWJNPlX9wJA4C3vqAgQjBla95ZPnjCFu7piVzYMflPDr94r55clDO1xg/nB2m40bbriBNWvXcN/993PWWWdy2WWXtxn5NZRi29atTYviuro6jj32WOZfMJ/8/PyY1DZuiy9s8Mb2GmYNiyOzg5tb6oNhRiTaJPIiRDdyW3WGJ9gorg+S3MqX/I66ckoq+xpDPL+xknSXhW+NSmx2zLixY/n53XdTUlrK0iVL+N5132P2SbOZP/9CcrKzu/T6hwuHw/x70b9ZtmwZ3/3ud5k7d26LUeNARLGuuIFVu+r4pKSRiIKRiTaum5bOySPjSXUc2V2sxqBBhsuCU77kCxFzskDuByy6idxkO19VeElrI5fv5BHx+MMGj64t44H3i7l7dk67kd62HDfzOMaMGcvvf/97fvKTn3DXnXeSmZnZ9Ht/IMDGjRtZs2YNH3/8MW63m+OOm8nNt9zCuHHjejSv980dtXhCBhdO6NjmFkNFc6Yz3bZuHpkQYliijbLGUJtf8jtCQ+Pm47Ko9IX4w5oyUp0W8jNdLR6bPWQIN954I1deeSXLV6zgtttuZVLeJC666CIyMjOOeAwHfL1rFw///vckJibyxBNPkp6WdsjvFYpN+3ys+rqO93fX0xgySLabOX9cMqcdlcCoxK4HDfwRRV6CnMOE6A6yQO4nMtwWCmt1/GGjzVtwc3MT8YcN/vTpXh7+qJTbZw3B1IXmGynJydx///288sor3HTzTVxzzbWgFOs+WUdBQQGjRh3FzJkzWbBgQcyjMx2xpdLHyq01vLe7jvGpdsaldiznsCEQISveiuMI8rWFEJ1j1U2MjcGGPQCLSePns3P48ZuF/Pq9Yh4+YwQjEltfJCYmJnL1VVdx8cUX8eZ/3uQ3D/yGYUOHMWLECEaPGcPYMWPIyMzocJOicDjMSy+/zIoVy7nmmms544zTmz12ryfEr94tZmeNH5uuccKwOE4dlcDUTBd6jAIH3lCEBJuJhFY2cAshukY+Wf2ESYvuCC8o97Sbo3b+uGT8YYO/F1RgN5u4aWZmlzrUmTSNC+fPZ8qUKTzxxBNkDxnC8ccdz0033dxuXc/uEIgo3t9dx4qtNWyr8mM3a8zNTeTbeakdfo6gociJl8iLED0lVhv2IJq28atTh3HrG4X8/O09/O70EWS1k1rlsDs4//zzOefcc1m37mMKCwt55513ePrPfyYYCjFm9GhGjxnD6NGjGTtmDCmpzRsN7fz6ax5++PckJ6fw5JNPkpaa1ux1tlf5+H/vFBMIG9wyM4uTRsTjOMI9IW3xhgymtBI9F0J0nSyQ+5EDtYk7coG5JC8VX9jg5S+rsJtNLJyR3uU2zrlHHcWjf/gDNbU13b7ppSV7PSFe317DG9trqQ9EyIm3cv3RmZw2KqFTlTs8wQgpTkuXL9JCiI6L5YY9gAyXhV+dNpQ7/7ubH6zcyVX5aR2qf27WdcaPH8/xx32zabiyqood27ezdds23njjDR577FF03czo0bmMGT2GMWPHsG3rNlauXMm1113LnDlzWjyfflTcwEPvl5Bg13ngW21HtrsiWtrNRFIXc7qFEK2TT1c/omnR7lSflDR2qMTad/LT8IUMXtlSjcNi4qopzaMdfZ1CsbHcy4qt1awpagRgZo6beeOSyc90HtGi3xs2GJfmjPVQhRDtiOWGPYDcJDv/d84onvyknL+s38c7u+q55bisTtURBkhNSSE1JYWZM2cC0fPOvn0VbN+2ja3btrF0yVLccXE8+eSTpKY2v1OlUCzbUsOfP93L2BQ795wytNXa9bFQH4wwIdUp9duF6EayQO5n4m1mstxWKn0hEm1t/+fT0PjB0Rn4w4p/fVGJ3axx8cSOpyH0Jm/Y4K2ddazcWk1RfZA4m85FE1M4e3RSp1qwHs4fNoiz6iTaJXosRG+I1Ya9A9JdFu45OYf3dzfw1Cfl3PTaLi6cmMLlk9Kwm49sAamhkZGeTkZ6OrNmzWrz2LBS/OmTcl7dVssJQ93cfkL2EZfZ7IiwoTCbTKRK/XYhupUskPuhEUk29ja23TzkABMaN8/MxB+O8LfPKnCYdc4d2/PpEZ2xbEs1zxXswxdWjE628+Pjspg9IgHbEZatO1hj0GBShkNKuwnRS2K5Ye8ADY3Zw+OZmuXimQ37WLSpig9213PzzKxWq1zEgjds8MDqEj4pbeTC8clcOz29S5uiO6IuECE32d6lCkVCiPbJArkfclp0hifa2FPXsduUuqZxxwnZBCLFPPlJOXazxpyjErt/oJ2kUDxXUMFLX1YxY4iLKyandbgqRUcEIwY2XSP5CGuOCiFiI5Yb9g4WZ9W5dWYWp4yI549ry7jrrT3MOSqB703LIN4W27tGFd4Q/+/tInbXBrjhmEzOGdP9gYeIoYBo1FwI0b2kxlU/lZNgw6RFb7d1hNmk8bMTc8jPdPLImjLe313fzSPsnIhSPP5xOS99WcXc3ETuPWVoTBfHAA1BgxFJtnaj7kKI7nVgw54vrKI1yWMsP9PFU+ccxcUTU1i1s46FK3fybmE9iti81o5qP7e8UUh5Y5BfnTq0RxbHEM09HpZgxdaNKRxCiCj5lPVTVt3EUcl26gLhDj/Gpmvcc/JQxqc5ePCDEj4ubujGEXZcyFA89EEJr2+v5eKJKdw8MzNmtUIPiBgKDYm8CNFXRDfsWan1R7rl+e1mjWumpvPHs0eS5rTw4Acl/PKdYvZ6Ql163o+LG/jJfwvRNHh47ghmDHHHaMRti7aVhqw4a4+8nhCDnSyQ+7FMtxWbbiIQNjr8GLvZxK9OHcrIJDv3rS7u9UWyP2xwzztFrN7dwHXT0rlmatfL0bWkNhBmWIIVSww2BQkhYmNYYvSOTjDS8XNYZ+Um2fnDmSNYOD2djeUevr9yJ2/trOOLvV7KGkMEIh2PKq/YWsM97xaTE2/j0TNHxKQbXkc1BCNkuM04LbLBWIieIDnI/Zhu0hidYufzvV7SOnHLzW3Ruf+0Ydz5v9388t1ijs1xc+3UdIb1cMvS+kCE//fOHrZW+rllZhZzcxO75XUOtJXOipPGIEL0Jd2xYa8lZk1j/vgUjh8az+Mfl7F8SyV/3exr+n2cTSfVYSbFaSbVaSHVuf+fHRZSXGaS7WZe+rKSZVtqODbHzU9ndW+lipYEIophCT23IBdisJMFcj+X6rQQb9PxBCO4OrHZJcGm8+jckSzbUsXLX1byg5Vfc+boRK6Yktat9TsPqPKFuHtVEcX1Ae6encOsYXHd9lrSVlqIvqu7Nuy1JNNt4b7ThrK12IFHd1DlDVPpDUX/3xemyhtmR7W/1bSP88cl8b3pGTFPAWuPJxgh2WEmLsYbDYUQrZMFcj+naRqjUxx8WtrYqQUyRHP0LslLZW5uIv/4vILXt9Wyalcd385L4YJxKd0WISltDPKzt3ZT64/w61OHMbWb26VKW2kh+q5vOuw14rSobm9+oaGR4TaTlNh67nDIUFTvXzRXekJU+UKku6zd+kW+Ld6QwVhpbiREj5IF8gCQaDeT4bRQFwgT307zkNYef8MxWZw3LoW/bdjLcwWVvLa1lqvy0zhtVEJMoyVf1/q5+60iwkrx0LeGMzbGlSoO1yhtpYXo8w5s2ItVh72uspg0MtyWaFOitO49R7XHHzZw23SSpLmRED1K7jkPEKOS7QQjdKlk0tB4K788eSi/P304yU4zj6wp44bXdrG+rDEmY/yywssdb+5G1+Dh07t/cQzgCxsM7+HcaiFE5/XEhr3+qDEYYWSiTZobCdHDZIE8QLisOsMTrVT5Ol72rTV56U4ePXMEd83KxhuKcPeqIn6+ag+7av1H/Jyfljbys7f2kGDXefiMET2yIVDaSgvRfxzYsFcX6J6yb/1RKKKw6iZSnFKeUoie1vv3skTMjEyy0xCIUBsIk3gEqRYHM6Fx8oh4jh8ax4qt1bz0RSXXv7qLOUclMiEujL1Wx2rWsJo0rLoJq65h0ff/s0nDat7/7yYTHxU18LsPSxiRaOf+04aS2AObAEHaSgvR36S5LGS6rVR4QyT30HmiL6sLRhiTbJfmRkL0AjkDDSAmTWN8mpNPSxrxhYyYVG2w6hoLJqRw+lGJvPhFJSu3VrMBH5Wqc2kXeekO7jllKO4equF5oK20RF6E6D80TWNsqgNfeYS6QJiELn7R788ihsIE0TxoIUSPG7xnnwHKZjYxOdPFJyUNmE3RKG4sxNt0vj8jg8unpFK+rxqrO55gRBE0FMGwQejAP0cMgmFF6MA/RxQ2s4kzc5Owm3suCtIQMBiTau/2HfFCiNgymzQmpbtYX9rY6fKVA0nd/rbS0txIiN4hC+QBKM6mMzHdyRd7vaQ6zTFdJLotOikuM0l9eONbxFCYTNJWWoj+ymY2MSXTxSeljZjDBrYebsrR2wylUIaS5kZC9KLBddYZRDLcVkYk2mKyaa+/kbbSQvR/LqvOlAwX9UGDsHHk1Xn6I2luJETvk0/fADYq2U6Kw0xtYPAskqWttBADR5LDzIQ0B1W+cJdKWPY30txIiN4nC+QB7MCmPR0NX2hw1BY9EHnpri6AQoielRVnZXSKnSpvGDUIFsmNwQip0txIiF4nq4gB7sCmPU8oQigy8C8uEnkRYuAZnmBjSLyVKv/Ar5HsCxsMT5RzmBC9TRbIg8CBTXs1/oF9m1IiL0IMTJqmMSbFQbJdp9Y/MFPGlFLU+cPE23QSbHIOE6K3yQJ5kBgMm/Yk8iLEwKWbNCakO7HoJhqDAyeS7AsZVHpDVPsjJDvMjE9zSnMjIfoAKfM2iIxKttMYjE2nvb4iGDFoCEZQSiPdZZHIixADmFU3MSUz2gzJHzb67V6D6HnLQClIdOhMTHKS5DBjlco7QvQZA2OVJDqkOzrt9YaIoWgIRggrsOsauckOUpxmnD3UpU8I0XucFp0pmdFGIroWu2ZI3S1sKBoCESIoHGad0Sl2UhyWfnseFmKgkwXyINNdnfa6m1IKT8jAF1JY9OjO9gy3hTirLrcjhRhkEuxmJmU42VjuJcVhRjf1zXOAoRSNwegGabNuYmiClXS3FZfFJOctIfo4WSAPQt3ZaS/W/GEDT9BAaYp0p4WxqVYS7X33giiE6BlpLitjUxVbK32kOc19bsHZEIgQiCiy4ixkxVmJt+l9+lwrhDiULJAHqQy3lYZAhN11AVIdfe/iUuMPEzYUCTad8WkOkp2SnyeEOFROvBV/2GBPbYDkPhRJrvaFcVlMTBvilhQKIfopWSAPYqOS7YQMRUl9kES73icWoIZSVHnDZMZZGZlowykl24QQrdA0jaOS7TjMJrZX+bDqJuJ6caOuUopKX5g0p4XxaQ5pdy9EPyYL5EHswKa9FIeZzZU+fCGDBHvv/UmEDUWVL8xRSTZGJtn7XFRbCNH3mDSNnAQbyQ4z26p8VHhCJNrNPb6/IrL//DU80cZRyXZJpxCin5MFsiDdbSXOZmZrpZcKT6hXblX6w9GyR5MynGS6rT362kKI/s9pjVa3KG8MsbXKhylMj5WzDEYMav0RxqU6yEmQWuxCDASyQBYAOCwmpmS6KKkPsq3Kh8NswtVD6Q2eYISQoZg+xEViL0awhRD9m6ZpZMVFN/LuqPKxzxMioZvTx7yhCN6wwdRMFykuS7e9jhCiZ0mClGii7b9VeUxOHLpJo8rb/a2pa/xh0DRmDHHL4lgIERMOi4m8DCeTM514QwY1/jCqG85l9YEwEQOOHhIni2MhBhhZIItm3Fad6UPcDEu0UukL4w8bMX8NpRSV3hAJNp3pQ1yyGU8IEVOappHmsnJsThxpLgv7vGECMTyXVfvC2C0mpme7e3VjoBCie0jITrRIN2kclewgxWlh014v3lCYJHtsmnJEDEW1P0xOvJXcZEefKc0khBh4bGYTE9KcZLgsbKnw4QmFoQvRZGN/pYoMl4XxaU7Mcv4SYkCSCLJoU6LdzNE5btLdFip9YYKRrkVgghGDKl+YMSkOxqTI4lgI0TNSnBaOyXEzJM5Ktd+gxhemIRDBHzYIGx1bMEcMRaU3zMgEGxPTZXEsxEAmEWTRLqsejcCk7i8HV+OPYAlGsOxvVd3Rcka+kIE3FCE/00mqSypVCCF6lkU3MSbVgR6wY3Xb8QQNfOEIvpCBP2x8c4dMgdIUZk3DYtIwmzQU0BCMMCHdwZA4qVQhxEAnC2TRYU3l4AwPVocZTzBCXSCCYSjQNBQKE1rTwtli0poixA2BCAqYkR0n+XpCiF4VbzWRdlg5NkMpQhFFyFAEI4pQJPqF3hdS+MIGwYhiapabZIdcNoUYDOSTLjrFYTGRHWchLc0JRDfbHbigBMLRKIwnZNAQiNAQjBDen+sXZ9XJS3dJ21UhRJ9k0jRsZg2JDQshoAcWyJFIhBkzZpCdnc2rr77Krl27uOSSS6iqqmL69Om88MILWK1WAoEAV111FevXryclJYWXX36ZESNGdPfwRBdpmoZV17Dq0eoXhwsbimDYwGY2Sb6xEEIIIfqFbg/nPfbYY4wfP77p3++8805uvfVWduzYQVJSEn/9618B+Otf/0pSUhI7duzg1ltv5c477+zuoYkeYDZpOK26LI6FEEII0W906wK5uLiY1157jeuuuw6I3o5/++23WbBgAQBXX301y5YtA2D58uVcffXVACxYsIBVq1Z1S2F3IYQQQggh2tKtKRa33HILv/3tb2loaACgqqqKxMREzOboy+bk5FBSUgJASUkJQ4cOjQ7KbCYhIYGqqipSU1MPec6nn36ap59+GoCKigoqKiq6NMaampouPX4wkjnrHJmvzpM56xyZr86R+eo8mbPOkfnqvL42Z922QH711VdJT09n+vTpvPvuuzF73oULF7Jw4UIApkyZQlpaWpefMxbPMdjInHWOzFfnyZx1jsxX58h8dZ7MWefIfHVeX5qzblsgf/jhh6xYsYLXX38dv99PfX09N998M7W1tYTDYcxmM8XFxWRnZwOQnZ1NUVEROTk5hMNh6urqSElJ6a7hCSGEEEII0aJuy0F+4IEHKC4uprCwkJdeeolTTz2Vf/7zn5xyyiksXrwYgOeee47zzjsPgHnz5vHcc88BsHjxYk499dSYtDUWQgghhBCiM3q8KO1DDz3EI488Qm5uLlVVVVx77bUAXHvttVRVVZGbm8sjjzzCgw8+2NNDE0IIIYQQomcahZx88smcfPLJAIwaNYp169Y1O8Zut7No0aKeGI4QQgghhBCtkrZmQgghhBBCHEQWyEIIIYQQQhxEFshCCCGEEEIcRBbIQgghhBBCHEQWyEIIIYQQQhxEFshCCCGEEEIcRBbIQgghhBBCHERTSqneHsSRSk1NZcSIEV16joqKij7V+7s/kDnrHJmvzpM56xyZr86R+eo8mbPOkfnqvN6as8LCQiorK5v9vF8vkGNhxowZfPrpp709jH5F5qxzZL46T+asc2S+Okfmq/NkzjpH5qvz+tqcSYqFEEIIIYQQB5EFshBCCCGEEAcZ9AvkhQsX9vYQ+h2Zs86R+eo8mbPOkfnqHJmvzpM56xyZr87ra3M26HOQhRBCCCGEONigjyALIYQQQghxMFkgCyGEEEIIcZB+tUC+5pprSE9PJy8v75CfV1dXM2fOHEaPHs2cOXOoqalp8fGXX345Y8eOJS8vj2uuuYZQKATA7373O/Lz88nPzycvLw9d16murm72+PXr1zNp0iRyc3O56aabOJCd8otf/ILJkyeTn5/P6aefTmlpaYzf+ZHpq/O1ceNGjjvuOCZNmsS5555LfX19jN/5kevtObv77rsZOnQobrf7kJ///e9/Jy0trek5nnnmmRi9467pzfnyer2cffbZjBs3jokTJ3LXXXc1/W716tVMmzYNs9nM4sWLY/yuu6a75qyuro5zzz2XKVOmMHHiRJ599tkWH/+f//yHsWPHkpuby4MPPtj082uvvZYpU6YwefJkFixYQGNjY4zecdf01fl6++23mTZtGnl5eVx99dWEw+EYveOu6e35au3177nnHrKzs5s+16+//noM3m1s9OacFRUVccoppzBhwgQmTpzIY4891vS7RYsWMXHiREwmU58qf9Zd81VTU8MFF1zA5MmTOeaYY/jyyy9bfHyPrcVUP/Lee++p9evXq4kTJx7y89tvv1098MADSimlHnjgAXXHHXe0+PjXXntNGYahDMNQl1xyiXrqqaeaHbNixQp1yimntPj4o48+Wq1Zs0YZhqHmzp2rXn/9daWUUnV1dU3HPPbYY+r73//+Eb2/WOur8zVjxgz17rvvKqWU+utf/6p+/vOfH/F7jLXenrM1a9ao0tJS5XK5Dvn5s88+q370ox8dyVvqVr05Xx6PR7399ttKKaUCgYCaNWtW09/Yrl271MaNG9WVV16pFi1a1KX3GGvdNWf3339/02P27dunkpKSVCAQOOSx4XBYjRo1Su3cuVMFAgE1efJktWnTJqXUoeexW2+9tWksva0vzlckElE5OTlq69atSimlfvGLX6hnnnkmpu/7SPXmfLX1+r/85S/V7373uy6/v+7Qm3NWWlqq1q9fr5RSqr6+Xo0ePbrpM/nVV1+pLVu2qJNOOkl98sknsXvDXdRd8/WTn/xE3XPPPUoppTZv3qxOPfXUFh/fU2uxfhVBnj17NsnJyc1+vnz5cq6++moArr76apYtW9bi48866yw0TUPTNI455hiKi4ubHfPiiy9y6aWXNvt5WVkZ9fX1zJw5E03TuOqqq5peJz4+vuk4j8eDpmlH8O5ir6/O17Zt25g9ezYAc+bMYcmSJUf4DmOvN+cMYObMmWRlZR35G+hhvTlfTqeTU045BQCr1cq0adOaHj9ixAgmT56MydT3TnHdNWeaptHQ0IBSisbGRpKTkzGbzYc8dt26deTm5jJq1CisViuXXHIJy5cvB745jyml8Pl8A/481pX5qqqqwmq1MmbMGKBvncd6c77aev2+rDfnLCsri2nTpgEQFxfH+PHjKSkpAWD8+PGMHTs2Vm8zZrprvr766itOPfVUAMaNG0dhYSF79+495LE9uRbre1ePI7B3796mRUVmZmazCT1cKBTihRdeYO7cuYf83Ov18p///IcLL7yw2WNKSkrIyclp+vecnJymP2L45tb4P//5T371q1915e10u96er4kTJzZdlBctWkRRUVGX3k9P6Ik5a8+SJUuabn/39Tnr6fmqra1l5cqVnHbaaV0beC/q6pzdcMMNbN68mSFDhjBp0iQee+yxZl8QSkpKGDp0aNO/H34e++53v0tmZiZbtmzhxhtvjNVb6xa9OV+pqamEw+Gm296LFy8e8J/JjsxXe5544gkmT57MNddc0+rt976kp+essLCQzz77jGOPPTZ2b6IHdXW+pkyZwtKlS4Hol9Pdu3c3C5r05FpsQCyQD3bgW0lbrr/+embPns2JJ554yM9XrlzJCSeccETffu+//36Kioq4/PLLeeKJJzr9+N7SG/P1t7/9jaeeeorp06fT0NCA1Wrt9Lh7U2/M2bnnnkthYSGff/45c+bMafqW3h9093yFw2EuvfRSbrrpJkaNGhWTMfe2I5mzN998k/z8fEpLSykoKOCGG27odH7/s88+S2lpKePHj+fll18+4vH3tJ6eL03TeOmll7j11ls55phjiIuLQ9f1Lr+PntIbf18//OEP2blzJwUFBWRlZfHjH/+4S++hp3X3nDU2NnLhhRfy6KOPHhIJ7a+OZL7uuusuamtryc/P5/HHH2fq1Kmd/lzFci02IBbIGRkZlJWVAdHwe3p6OgBnnHEG+fn5XHfddU3H3nvvvVRUVPDII480e56XXnqp1Vvf2dnZh3yTKS4uJjs7u9lxl19+eZ+51daa3p6vcePG8d///pf169dz6aWXctRRR8XsvXWXnpiztqSkpGCz2QC47rrrWL9+/ZG8jR7Tk/O1cOFCRo8ezS233BK7N9ALujpnzz77LPPnz0fTNHJzcxk5ciRbtmw55DWys7MPiXS2dB7TdZ1LLrlkwJ/Hujpfxx13HO+//z7r1q1j9uzZTekWfVVPzFd7r6/rOiaTie9973usW7cuRu+s+/TUnIVCIS688EIuv/xy5s+f383vqvt0db7i4+N59tlnKSgo4Pnnn6eioqJZ0KMn12IDYoE8b948nnvuOQCee+45zjvvPCD67a2goKBpx/8zzzzDm2++yYsvvtjsNkddXR3vvfde02MPl5WVRXx8PGvXrkUpxfPPP9907Pbt25uOW758OePGjYv5e4yl3p6vffv2AWAYBvfddx8/+MEPuuV9xlJPzFlbDpx0AFasWMH48eOP9K30iJ6ar5///OfU1dXx6KOPds8b6UFdnbNhw4axatUqIHqrc+vWrc0uLkcffTTbt29n165dBINBXnrpJebNm4dSih07dgDRHOQVK1YM+PNYV+YLvjmPBQIBHnrooT5/HuuJ+WrLweewV155pVkFhL6oJ+ZMKcW1117L+PHjue2223ribXWbrs5XbW0twWCw6ZjZs2c3i6b36FqsS1v8etgll1yiMjMzldlsVtnZ2U27hisrK9Wpp56qcnNz1WmnnaaqqqpafLyu62rUqFFqypQpasqUKeree+9t+t2zzz6rvv3tb7f5+p988omaOHGiGjVqlPrRj36kDMNQSik1f/58NXHiRDVp0iR1zjnnqOLi4hi9467pq/P16KOPqtGjR6vRo0erO++8s+nnfUFvz9ntt9+usrOzlaZpKjs7W/3yl79USil11113qQkTJqjJkyerk08+WW3evDk2b7iLenO+ioqKFKDGjRvX9Pi//OUvSiml1q1bp7Kzs5XT6VTJyclqwoQJMXzXXdNdc1ZSUqLmzJmj8vLy1MSJE9ULL7zQ4uNfe+01NXr0aDVq1Ch13333KaWUikQi6vjjj2967GWXXXbIjvDe1BfnS6nojvtx48apMWPGqD/84Q+xfdNd0Nvz1drrX3HFFSovL09NmjRJnXvuuaq0tLQb3v2R6c05e//99xWgJk2a1PT41157TSml1NKlS1V2drayWq0qPT1dnX766d00A53TXfP10UcfqdGjR6sxY8aoCy64QFVXV7f4+J5ai0mraSGEEEIIIQ4yIFIshBBCCCGEiBVZIAshhBBCCHEQWSALIYQQQghxEFkgCyGEEEIIcRBZIAshhBBCCHEQWSALIUQ/UVVVRX5+Pvn5+WRmZpKdnU1+fj5ut5vrr7++t4cnhBADhpR5E0KIfuiee+7B7Xbzk5/8pLeHIoQQA45EkIUQop979913Oeecc4Dowvnqq6/mxBNPZPjw4SxdupQ77riDSZMmMXfuXEKhEADr16/npJNOYvr06ZxxxhmHdDoTQojBThbIQggxwOzcuZO3336bFStWcMUVV3DKKafwxRdf4HA4eO211wiFQtx4440sXryY9evXc80113D33Xf39rCFEKLPMPf2AIQQQsTWmWeeicViYdKkSUQiEebOnQvApEmTKCwsZOvWrXz55ZfMmTMHgEgkQlZWVm8OWQgh+hRZIAshxABjs9kAMJlMWCwWNE1r+vdwOIxSiokTJ7JmzZreHKYQQvRZkmIhhBCDzNixY6moqGhaIIdCITZt2tTLoxJCiL5DFshCCDHIWK1WFi9ezJ133smUKVPIz8/no48+6u1hCSFEnyFl3oQQQgghhDiIRJCFEEIIIYQ4iCyQhRBCCCGEOIgskIUQQgghhDiILJCFEEIIIYQ4iCyQhRBCCCGEOIgskIUQQgghhDiILJCFEEIIIYQ4yP8HdW7QMtNE0WAAAAAASUVORK5CYII=\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -168,9 +172,8 @@ } ], "source": [ - "# Visualize the groud truth, actual forecast and confident interval \n", - "fig, ax = model1.plot_forecast(time_series=test_data,\n", - " plot_forecast_uncertainty=True)\n", + "# Visualize the groud truth, actual forecast and confidence interval \n", + "fig, ax = model1.plot_forecast(time_series=test_data, plot_forecast_uncertainty=True)\n", "plt.show()" ] }, @@ -190,16 +193,29 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:merlion.models.forecast.base:Automatically detect the periodicity is 24\n", - "INFO:merlion.models.forecast.sarima:Seasonal difference order is 1\n", - "INFO:merlion.models.forecast.sarima:Difference order is 0\n", - "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", - "INFO:merlion.models.automl.autosarima:Difference order is 0\n", + "INFO:merlion.models.automl.seasonality:Automatically detect the periodicity is 24\n", "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", "INFO:merlion.models.automl.autosarima:Difference order is 0\n", "INFO:merlion.models.automl.autosarima:Best model: SARIMA(2,0,3)(1,1,1)[24] without constant\n" ] - }, + } + ], + "source": [ + "# Specify the configuration of full AutoSarima without approximation\n", + "# Note that the default values of all the auto_* parameters are True\n", + "config2 = AutoSarimaConfig(approximation=False, maxiter=5)\n", + "model2 = AutoSarima(config2)\n", + "\n", + "# Model training\n", + "train_pred, train_err = model2.train(\n", + " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [ { "name": "stdout", "output_type": "stream", @@ -209,15 +225,6 @@ } ], "source": [ - "# Specify the configuration of full AutoSarima without approximation\n", - "config2 = AutoSarimaConfig(max_forecast_steps=len(train_data), order=(\"auto\", \"auto\", \"auto\"),\n", - " seasonal_order=(\"auto\", \"auto\", \"auto\", \"auto\"), approximation=False, maxiter=5)\n", - "model2 = SeasonalityLayer(model = AutoSarima(model = Sarima(config2)))\n", - "\n", - "# Model training\n", - "train_pred, train_err = model2.train(\n", - " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})\n", - "\n", "# Model forecasting\n", "forecast2, stderr2 = model2.forecast(len(test_data))\n", "\n", @@ -228,12 +235,12 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 7, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -243,9 +250,8 @@ } ], "source": [ - "# Visualize the groud truth, actual forecast and confident interval \n", - "fig, ax = model2.plot_forecast(time_series=test_data,\n", - " plot_forecast_uncertainty=True)\n", + "# Visualize the groud truth, actual forecast and confidence interval \n", + "fig, ax = model2.plot_forecast(time_series=test_data, plot_forecast_uncertainty=True)\n", "plt.show()" ] }, @@ -260,22 +266,39 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ - "INFO:merlion.models.forecast.base:Automatically detect the periodicity is 24\n", - "INFO:merlion.models.forecast.sarima:Seasonal difference order is 1\n", - "INFO:merlion.models.forecast.sarima:Difference order is 0\n", - "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", - "INFO:merlion.models.automl.autosarima:Difference order is 0\n", + "INFO:merlion.models.automl.seasonality:Automatically detect the periodicity is 24\n", "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", "INFO:merlion.models.automl.autosarima:Difference order is 0\n" ] - }, + } + ], + "source": [ + "# Specify the configuration of partial AutoSarima \n", + "# We explicitly specify values of p, q, P, Q in the order and seasonal order,\n", + "# and we set auto_pqPQ=False.\n", + "# Because auto_d=True, auto_D=True, and auto_seasonality=True by default, we\n", + "# can specify arbitrary values for them in the order and seasonal order (e.g. \"auto\")\n", + "config3 = AutoSarimaConfig(auto_pqPQ=False, order=(15, \"auto\", 5),\n", + " seasonal_order=(2, \"auto\", 1, \"auto\"), maxiter=5)\n", + "model3 = AutoSarima(config3)\n", + "\n", + "# Model training\n", + "train_pred, train_err = model3.train(\n", + " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [ { "name": "stdout", "output_type": "stream", @@ -285,15 +308,6 @@ } ], "source": [ - "# Specify the configuration of partial AutoSarima \n", - "config3 = AutoSarimaConfig(max_forecast_steps=len(train_data), order=(15, \"auto\", 5),\n", - " seasonal_order=(2, \"auto\", 1, \"auto\"), maxiter=5)\n", - "model3 = SeasonalityLayer(model = AutoSarima(model = Sarima(config3)))\n", - "\n", - "# Model training\n", - "train_pred, train_err = model3.train(\n", - " train_data, train_config={\"enforce_stationarity\": True,\"enforce_invertibility\": True})\n", - "\n", "# Model forecasting\n", "forecast3, stderr3 = model3.forecast(len(test_data))\n", "\n", @@ -304,12 +318,12 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 10, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -319,9 +333,8 @@ } ], "source": [ - "# Visualize the groud truth, actual forecast and confident interval \n", - "fig, ax = model3.plot_forecast(time_series=test_data,\n", - " plot_forecast_uncertainty=True)\n", + "# Visualize the groud truth, actual forecast and confidence interval \n", + "fig, ax = model3.plot_forecast(time_series=test_data, plot_forecast_uncertainty=True)\n", "plt.show()" ] }, @@ -334,42 +347,45 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ - "INFO:merlion.models.forecast.base:Automatically detect the periodicity is 24\n", - "INFO:merlion.models.forecast.sarima:Seasonal difference order is 1\n", - "INFO:merlion.models.forecast.sarima:Difference order is 0\n", - "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", - "INFO:merlion.models.automl.autosarima:Difference order is 0\n", + "INFO:merlion.models.automl.seasonality:Automatically detect the periodicity is 24\n", "INFO:merlion.models.automl.autosarima:Seasonal difference order is 1\n", "INFO:merlion.models.automl.autosarima:Difference order is 0\n", "INFO:merlion.models.automl.autosarima:Fitting models using approximations(approx_iter is 1) to speed things up\n", - "INFO:merlion.models.automl.autosarima:Best model: SARIMA(1,0,5)(0,1,2)[24] without constant\n" + "INFO:merlion.models.automl.autosarima:Best model: SARIMA(5,0,1)(2,1,0)[24] without constant\n" ] - }, + } + ], + "source": [ + "# Specify the configuration of AutoSarima without enforcing stationarity and invertibility\n", + "config4 = AutoSarimaConfig(approximation=True, maxiter=5)\n", + "model4 = AutoSarima(config4)\n", + "\n", + "# Model training\n", + "train_pred, train_err = model4.train(\n", + " train_data, train_config={\"enforce_stationarity\": False,\"enforce_invertibility\": False})" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "AutoSarima without enforcing stationarity and invertibility sMAPE is 3.4972\n" + "AutoSarima without enforcing stationarity and invertibility sMAPE is 4.2063\n" ] } ], "source": [ - "# Specify the configuration of AutoSarima without enforcing stationarity and invertibility\n", - "config4 = AutoSarimaConfig(max_forecast_steps=len(train_data), order=(\"auto\", \"auto\", \"auto\"),\n", - " seasonal_order=(\"auto\", \"auto\", \"auto\", \"auto\"), maxiter=5)\n", - "model4 = SeasonalityLayer(model = AutoSarima(model = Sarima(config4)))\n", - "\n", - "# Model training\n", - "train_pred, train_err = model4.train(\n", - " train_data, train_config={\"enforce_stationarity\": False,\"enforce_invertibility\": False})\n", - "\n", "# Model forecasting\n", "forecast4, stderr4 = model4.forecast(len(test_data))\n", "\n", @@ -377,6 +393,28 @@ "smape4 = ForecastMetric.sMAPE.value(ground_truth=test_data, predict=forecast4)\n", "print(f\"AutoSarima without enforcing stationarity and invertibility sMAPE is {smape4:.4f}\")" ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# Visualize the groud truth, actual forecast and confidence interval \n", + "fig, ax = model4.plot_forecast(time_series=test_data, plot_forecast_uncertainty=True)\n", + "plt.show()" + ] } ], "metadata": { diff --git a/merlion/models/anomaly/base.py b/merlion/models/anomaly/base.py index e7ec83e90..a039abccb 100644 --- a/merlion/models/anomaly/base.py +++ b/merlion/models/anomaly/base.py @@ -20,7 +20,7 @@ from merlion.post_process.calibrate import AnomScoreCalibrator from merlion.post_process.factory import PostRuleFactory from merlion.post_process.sequence import PostRuleSequence -from merlion.post_process.threshold import AggregateAlarms +from merlion.post_process.threshold import AggregateAlarms, Threshold from merlion.utils import TimeSeries logger = logging.getLogger(__name__) @@ -32,6 +32,8 @@ class DetectorConfig(Config): """ _default_threshold = AggregateAlarms(alm_threshold=3.0) + calibrator: AnomScoreCalibrator = None + threshold: Threshold = None def __init__( self, max_score: float = 1000, threshold=None, enable_calibrator=True, enable_threshold=True, **kwargs @@ -73,15 +75,14 @@ def post_rule(self): return PostRuleSequence(rules) @classmethod - def from_dict(cls, config_dict: Dict[str, Any], return_unused_kwargs=False, **kwargs): - # Get the calibrator, but we will set it manually after the constructor - config_dict = copy(config_dict) - calibrator_config = config_dict.pop("calibrator", None) + def from_dict(cls, config_dict: Dict[str, Any], return_unused_kwargs=False, calibrator=None, **kwargs): + # Get the calibrator, but we will set it manually after the constructor by putting it in kwargs + calibrator = config_dict.pop("calibrator", calibrator) config, kwargs = super().from_dict(config_dict, return_unused_kwargs=True, **kwargs) - if calibrator_config is not None: - config.calibrator = PostRuleFactory.create(**calibrator_config) + if calibrator is not None: + calibrator = PostRuleFactory.create(**calibrator) + config.calibrator = calibrator - # Return unused kwargs if desired if len(kwargs) > 0 and not return_unused_kwargs: logger.warning(f"Unused kwargs: {kwargs}", stack_info=True) elif return_unused_kwargs: @@ -96,6 +97,9 @@ class NoCalibrationDetectorConfig(DetectorConfig): """ def __init__(self, enable_calibrator=False, **kwargs): + """ + :param enable_calibrator: ``False`` because this config assumes calibrated outputs from the model. + """ super().__init__(enable_calibrator=enable_calibrator, **kwargs) @property diff --git a/merlion/models/anomaly/change_point/bocpd.py b/merlion/models/anomaly/change_point/bocpd.py index ba92f084b..6edf3abe8 100644 --- a/merlion/models/anomaly/change_point/bocpd.py +++ b/merlion/models/anomaly/change_point/bocpd.py @@ -117,7 +117,8 @@ def __init__( def to_dict(self, _skipped_keys=None): _skipped_keys = _skipped_keys if _skipped_keys is not None else set() config_dict = super().to_dict(_skipped_keys.union({"change_kind"})) - config_dict["change_kind"] = self.change_kind.name + if "change_kind" not in _skipped_keys: + config_dict["change_kind"] = self.change_kind.name return config_dict @property diff --git a/merlion/models/anomaly/dbl.py b/merlion/models/anomaly/dbl.py index 5cd757a00..0481b3923 100644 --- a/merlion/models/anomaly/dbl.py +++ b/merlion/models/anomaly/dbl.py @@ -96,7 +96,8 @@ def determine_train_window(self): def to_dict(self, _skipped_keys=None): _skipped_keys = _skipped_keys if _skipped_keys is not None else set() config_dict = super().to_dict(_skipped_keys.union({"trends"})) - config_dict["trends"] = [t.name for t in self.trends] + if "trends" not in _skipped_keys: + config_dict["trends"] = [t.name for t in self.trends] return config_dict diff --git a/merlion/models/anomaly/forecast_based/base.py b/merlion/models/anomaly/forecast_based/base.py index 375692491..3226e653e 100644 --- a/merlion/models/anomaly/forecast_based/base.py +++ b/merlion/models/anomaly/forecast_based/base.py @@ -7,7 +7,6 @@ """ Base class for anomaly detectors based on forecasting models. """ -from abc import ABC import logging from typing import List, Optional @@ -17,11 +16,12 @@ from merlion.models.forecast.base import ForecasterBase from merlion.plot import Figure from merlion.utils import UnivariateTimeSeries, TimeSeries +from merlion.utils.misc import AutodocABCMeta logger = logging.getLogger(__name__) -class ForecastingDetectorBase(ForecasterBase, DetectorBase, ABC): +class ForecastingDetectorBase(ForecasterBase, DetectorBase, metaclass=AutodocABCMeta): """ Base class for a forecast-based anomaly detector. """ diff --git a/merlion/models/anomaly/forecast_based/mses.py b/merlion/models/anomaly/forecast_based/mses.py index 5a82da9cf..042bd0633 100644 --- a/merlion/models/anomaly/forecast_based/mses.py +++ b/merlion/models/anomaly/forecast_based/mses.py @@ -17,10 +17,14 @@ class MSESDetectorConfig(MSESConfig, DetectorConfig): + """ + Configuration class for an MSES forecasting model adapted for anomaly detection. + """ + _default_threshold = AggregateAlarms(alm_threshold=2) - def __init__(self, online_updates: bool = True, **kwargs): - super().__init__(**kwargs) + def __init__(self, max_forecast_steps: int, online_updates: bool = True, **kwargs): + super().__init__(max_forecast_steps=max_forecast_steps, **kwargs) self.online_updates = online_updates diff --git a/merlion/models/anomaly/random_cut_forest.py b/merlion/models/anomaly/random_cut_forest.py index f8b33ed02..68bba4ba8 100644 --- a/merlion/models/anomaly/random_cut_forest.py +++ b/merlion/models/anomaly/random_cut_forest.py @@ -33,7 +33,7 @@ def __init__(self): import jpype.imports resource_dir = join(dirname(dirname(dirname(abspath(__file__)))), "resources") - jars = ["gson-2.8.6.jar", "randomcutforest-core-1.0.jar", "randomcutforest-serialization-json-1.0.jar"] + jars = ["gson-2.8.9.jar", "randomcutforest-core-1.0.jar", "randomcutforest-serialization-json-1.0.jar"] if not JVMSingleton._initialized: jpype.startJVM(classpath=[join(resource_dir, jar) for jar in jars]) JVMSingleton._initialized = True diff --git a/merlion/models/anomaly/zms.py b/merlion/models/anomaly/zms.py index 69efaa030..e9b105f4f 100644 --- a/merlion/models/anomaly/zms.py +++ b/merlion/models/anomaly/zms.py @@ -24,19 +24,16 @@ class ZMSConfig(DetectorConfig, NormalizingConfig): """ - Configuration class for `ZMS` anomaly detection model. + Configuration class for `ZMS` anomaly detection model. The transform of this config is actually a + pre-processing step, followed by the desired number of lag transforms, and a final mean/variance + normalization step. This full transform may be accessed as `ZMSConfig.full_transform`. Note that + the normalization is inherited from `NormalizingConfig`. """ _default_transform = TemporalResample(trainable_granularity=True) def __init__(self, base: int = 2, n_lags: int = None, lag_inflation: float = 1.0, **kwargs): r""" - Configuration class for ZMS. The transform of this config is actually a - pre-processing step, followed by the desired number of lag transforms - and a final mean/variance normalization step. This full transform may be - accessed as `ZMSConfig.full_transform`. Note that the normalization is - inherited from `NormalizingConfig`. - :param base: The base to use for computing exponentially distant lags. :param n_lags: The number of lags to be used. If None, n_lags will be chosen later as the maximum number of lags possible for the initial diff --git a/merlion/models/automl/autoets.py b/merlion/models/automl/autoets.py new file mode 100644 index 000000000..cb4522f2a --- /dev/null +++ b/merlion/models/automl/autoets.py @@ -0,0 +1,32 @@ +# +# Copyright (c) 2021 salesforce.com, inc. +# All rights reserved. +# SPDX-License-Identifier: BSD-3-Clause +# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause +# +""" +Automatic seasonality detection for ETS. +""" + +from typing import Union + +from merlion.models.forecast.ets import ETS +from merlion.models.automl.seasonality import SeasonalityConfig, SeasonalityLayer + + +class AutoETSConfig(SeasonalityConfig): + """ + Config class for ETS with automatic seasonality detection. + """ + + def __init__(self, model: Union[ETS, dict] = None, **kwargs): + model = dict(name="ETS") if model is None else model + super().__init__(model=model, **kwargs) + + +class AutoETS(SeasonalityLayer): + """ + ETS with automatic seasonality detection. + """ + + config_class = AutoETSConfig diff --git a/merlion/models/automl/autoprophet.py b/merlion/models/automl/autoprophet.py new file mode 100644 index 000000000..2864a97a9 --- /dev/null +++ b/merlion/models/automl/autoprophet.py @@ -0,0 +1,44 @@ +# +# Copyright (c) 2021 salesforce.com, inc. +# All rights reserved. +# SPDX-License-Identifier: BSD-3-Clause +# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause +# +""" +Automatic (multi)-seasonality detection for Facebook's Prophet. +""" +from typing import Union + +from merlion.models.automl.seasonality import PeriodicityStrategy, SeasonalityConfig, SeasonalityLayer +from merlion.models.forecast.prophet import Prophet + + +class AutoProphetConfig(SeasonalityConfig): + """ + Config class for Prophet with automatic seasonality detection. + """ + + def __init__( + self, + model: Union[Prophet, dict] = None, + periodicity_strategy: PeriodicityStrategy = PeriodicityStrategy.All, + **kwargs, + ): + model = dict(name="Prophet") if model is None else model + super().__init__(model=model, periodicity_strategy=periodicity_strategy, **kwargs) + + @property + def multi_seasonality(self): + """ + :return: ``True`` because Prophet supports multiple seasonality. + """ + return True + + +class AutoProphet(SeasonalityLayer): + """ + Prophet with automatic seasonality detection. Automatically detects and adds + additional seasonalities that the existing Prophet may not detect (e.g. hourly). + """ + + config_class = AutoProphetConfig diff --git a/merlion/models/automl/autosarima.py b/merlion/models/automl/autosarima.py index a73be8d0f..5b1518911 100644 --- a/merlion/models/automl/autosarima.py +++ b/merlion/models/automl/autosarima.py @@ -4,37 +4,47 @@ # SPDX-License-Identifier: BSD-3-Clause # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause # -import logging -import warnings +""" +Automatic hyperparameter selection for SARIMA. +""" from collections import Iterator -from typing import Tuple, Any, Optional +from copy import copy, deepcopy +import logging +from typing import Any, Optional, Tuple, Union import numpy as np -from merlion.models.automl.forecasting_layer_base import ForecasterAutoMLBase -from merlion.models.forecast.base import ForecasterBase -from merlion.models.forecast.sarima import SarimaConfig, Sarima +from merlion.models.automl.seasonality import PeriodicityStrategy, SeasonalityConfig, SeasonalityLayer +from merlion.models.forecast.sarima import Sarima from merlion.transform.resample import TemporalResample -from merlion.utils import TimeSeries, autosarima_utils, UnivariateTimeSeries -from copy import deepcopy +from merlion.utils import autosarima_utils, TimeSeries, UnivariateTimeSeries logger = logging.getLogger(__name__) -class AutoSarimaConfig(SarimaConfig): +class AutoSarimaConfig(SeasonalityConfig): """ - Configuration class for `AutoSarima`. + Configuration class for `AutoSarima`. Acts as a wrapper around a `Sarima` model, which automatically detects + the seasonality, (seasonal) differencing order, and (seasonal) AR/MA orders. If a non-numeric value is specified + for any of the relevant parameters in the order or seasonal order, we assume that the user wishes to detect that + parameter automatically. + + .. note:: + + The automatic selection of AR, MA, seasonal AR, and seasonal MA parameters is implemented in a coupled way. + The user must specify all of these parameters explicitly to avoid automatic selection. """ _default_transform = TemporalResample() def __init__( self, - max_forecast_steps: int = None, - target_seq_index: int = None, - order=("auto", "auto", "auto"), - seasonal_order=("auto", "auto", "auto", "auto"), - periodicity_strategy: str = "max", + model: Union[Sarima, dict] = None, + auto_seasonality: bool = True, + periodicity_strategy: PeriodicityStrategy = PeriodicityStrategy.ACF, + auto_pqPQ: bool = True, + auto_d: bool = True, + auto_D: bool = True, maxiter: int = None, max_k: int = 100, max_dur: float = 3600, @@ -43,28 +53,10 @@ def __init__( **kwargs, ): """ - For order and seasonal_order, 'auto' indicates automatically select the parameter. - Now autosarima support automatically select differencing order, length of the - seasonality cycle, seasonal differencing order, and the rest of AR, MA, seasonal AR - and seasonal MA parameters. Note that automatic selection of AR, MA, seasonal AR - and seasonal MA parameters are implemented in a coupled way. Only when all these - parameters are specified it will not trigger the automatic selection. - - - :param max_forecast_steps: Max number of steps we aim to forecast - :param target_seq_index: The index of the univariate (amongst all - univariates in a general multivariate time series) whose value we - would like to forecast. - :param order: Order is (p, d, q) for an ARIMA(p, d, q) process. d must - be an integer indicating the integration order of the process, while - p and q must be integers indicating the AR and MA orders (so that - all lags up to those orders are included). - :param seasonal_order: Seasonal order is (P, D, Q, S) for seasonal ARIMA - process, where s is the length of the seasonality cycle (e.g. s=24 - for 24 hours on hourly granularity). P, D, Q are as for ARIMA. - :param periodicity_strategy: selection strategy when detecting multiple - periods. 'min' signifies to select the smallest period, while 'max' signifies to select - the largest period + :param auto_seasonality: Whether to automatically detect the seasonality. + :param auto_pqPQ: Whether to automatically choose AR/MA orders ``p, q`` and seasonal AR/MA orders ``P, Q``. + :param auto_d: Whether to automatically choose the difference order ``d``. + :param auto_D: Whether to automatically choose the seasonal difference order ``D``. :param maxiter: The maximum number of iterations to perform :param max_k: Maximum number of models considered in the stepwise search :param max_dur: Maximum training time considered in the stepwise search @@ -74,27 +66,36 @@ def __init__( the length off the period is too high (``periodicity > 12``). :param approx_iter: The number of iterations to perform in approximation mode """ - super().__init__(max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index, **kwargs) - self.order = order - self.seasonal_order = seasonal_order - self.periodicity_strategy = periodicity_strategy + if model is None: + model = dict(name="Sarima", transform=dict(name="Identity")) + super().__init__(model=model, periodicity_strategy=periodicity_strategy, **kwargs) + + p, d, q = self.order + P, D, Q, m = self.seasonal_order + self.auto_seasonality = auto_seasonality or not isinstance(m, (int, float)) + self.auto_pqPQ = auto_pqPQ or any(not isinstance(x, (int, float)) for x in (p, q, P, Q)) + self.auto_d = auto_d or not isinstance(d, (int, float)) + self.auto_D = auto_D or not isinstance(D, (int, float)) self.maxiter = maxiter self.max_k = max_k self.max_dur = max_dur self.approximation = approximation self.approx_iter = approx_iter + @property + def order(self): + return self.model.order -class AutoSarima(ForecasterAutoMLBase): + @property + def seasonal_order(self): + return self.model.seasonal_order - config_class = AutoSarimaConfig - def __init__(self, model: ForecasterBase = None, **kwargs): - if model is None: - model = {} - if isinstance(model, dict): - model = Sarima(AutoSarimaConfig.from_dict({**model, **kwargs})) - super().__init__(model) +class AutoSarima(SeasonalityLayer): + + config_class = AutoSarimaConfig + require_even_sampling = True + require_univariate = True def _generate_sarima_parameters(self, train_data: TimeSeries) -> dict: y = train_data.univariates[self.target_name].np_values @@ -109,7 +110,6 @@ def _generate_sarima_parameters(self, train_data: TimeSeries) -> dict: max_dur = self.config.max_dur # These should be set in config - periodicity_strategy = "min" stationary = False seasonal_test = "seas" method = "lbfgs" @@ -129,36 +129,26 @@ def _generate_sarima_parameters(self, train_data: TimeSeries) -> dict: trend = None information_criterion = "aic" - n_samples = y.shape[0] - if n_samples <= 3: - information_criterion = "aic" - - # check y - if y.ndim > 1: - raise ValueError("auto_sarima can only handle univariate time series") - if any(np.isnan(y)): - raise ValueError("there exists missing values in observed time series") - - # detect seasonality - m = seasonal_order[-1] - if not isinstance(m, (int, float)): - m = 1 - warnings.warn( - "Set periodicity to 1, use the SeasonalityLayer()" "wrapper to automatically detect seasonality." - ) + # auto-detect seasonality if desired, otherwise just get it from seasonal order + if self.config.auto_seasonality: + candidate_m = super().generate_theta(train_data=train_data) + m, _, _ = super().evaluate_theta(thetas=candidate_m, train_data=train_data) + else: + m = max(1, seasonal_order[-1]) # adjust max p,q,P,Q start p,q,P,Q + n_samples = len(y) max_p = int(min(max_p, np.floor(n_samples / 3))) max_q = int(min(max_q, np.floor(n_samples / 3))) - max_P = int(min(max_P, np.floor(n_samples / 3 / m))) if m != 1 else 0 - max_Q = int(min(max_Q, np.floor(n_samples / 3 / m))) if m != 1 else 0 + max_P = int(min(max_P, np.floor(n_samples / 3 / m))) + max_Q = int(min(max_Q, np.floor(n_samples / 3 / m))) start_p = min(start_p, max_p) start_q = min(start_q, max_q) start_P = min(start_P, max_Q) start_Q = min(start_Q, max_Q) # set the seasonal differencing order with statistical test - D = seasonal_order[1] if seasonal_order[1] != "auto" else None + D = None if self.config.auto_D else seasonal_order[1] D = 0 if m == 1 else D xx = y.copy() if stationary: @@ -174,7 +164,7 @@ def _generate_sarima_parameters(self, train_data: TimeSeries) -> dict: # set the differencing order by estimating the number of orders # it would take in order to make the time series stationary - d = order[1] if order[1] != "auto" else autosarima_utils.ndiffs(dx, alpha=0.05, max_d=max_d, test=test) + d = autosarima_utils.ndiffs(dx, alpha=0.05, max_d=max_d, test=test) if self.config.auto_d else order[1] if stationary: d = 0 if d > 0: @@ -183,9 +173,7 @@ def _generate_sarima_parameters(self, train_data: TimeSeries) -> dict: # pqPQ is an indicator about whether need to automatically select # AR, MA, seasonal AR and seasonal MA parameters - pqPQ = None - if order[0] != "auto" and order[2] != "auto" and seasonal_order[0] != "auto" and seasonal_order[2] != "auto": - pqPQ = True + pqPQ = not self.config.auto_pqPQ # automatically detect whether to use approximation method and the periodicity if approximation is None: @@ -276,7 +264,7 @@ def generate_theta(self, train_data: TimeSeries) -> Iterator: if np.max(y) == np.min(y): order = [0, 0, 0] seasonal_order = [0, 0, 0, 0] - elif pqPQ is not None: + elif pqPQ: action = "pqPQ" order[1] = d seasonal_order[1] = D @@ -289,50 +277,32 @@ def generate_theta(self, train_data: TimeSeries) -> Iterator: elif stepwise: action = "stepwise" - return iter([{"action": action, "theta": [order, seasonal_order, trend]}]) + return iter([{"action": action, "theta": [order, seasonal_order, trend], "val_dict": val_dict}]) def evaluate_theta( self, thetas: Iterator, train_data: TimeSeries, train_config=None - ) -> Tuple[Any, Optional[ForecasterBase], Optional[Tuple[TimeSeries, Optional[TimeSeries]]]]: + ) -> Tuple[Any, Optional[Sarima], Optional[Tuple[TimeSeries, Optional[TimeSeries]]]]: - theta_value = thetas.__next__() + theta_value = next(thetas) # preprocess - train_config = train_config if train_config is not None else {} + train_config = copy(train_config) if train_config is not None else {} if "enforce_stationarity" not in train_config: train_config["enforce_stationarity"] = False if "enforce_invertibility" not in train_config: train_config["enforce_invertibility"] = False - val_dict = self._generate_sarima_parameters(train_data) + val_dict = theta_value["val_dict"] y = val_dict["y"] X = val_dict["X"] - p = val_dict["p"] - d = val_dict["d"] - q = val_dict["q"] - P = val_dict["P"] - D = val_dict["D"] - Q = val_dict["Q"] - m = val_dict["m"] - max_p = val_dict["max_p"] - max_q = val_dict["max_q"] - max_P = val_dict["max_P"] - max_Q = val_dict["max_Q"] - trend = val_dict["trend"] method = val_dict["method"] maxiter = val_dict["maxiter"] information_criterion = val_dict["information_criterion"] approximation = val_dict["approximation"] - refititer = val_dict["refititer"] - relative_improve = val_dict["relative_improve"] - max_k = val_dict["max_k"] - max_dur = val_dict["max_dur"] approx_iter = val_dict["approx_iter"] # use zero model to automatically detect the optimal maxiter if maxiter is None: - maxiter = autosarima_utils.detect_maxiter_sarima_model( - y=y, X=X, d=d, D=D, m=m, method=method, information_criterion=information_criterion - ) + maxiter = autosarima_utils.detect_maxiter_sarima_model(**val_dict) if theta_value["action"] == "stepwise": refititer = maxiter @@ -342,31 +312,10 @@ def evaluate_theta( else: maxiter = approx_iter logger.info(f"Fitting models using approximations(approx_iter is {str(maxiter)}) to speed things up") + train_config["maxiter"] = maxiter # stepwise search - stepwise_search = autosarima_utils._StepwiseFitWrapper( - y=y, - X=X, - p=p, - d=d, - q=q, - P=P, - D=D, - Q=Q, - m=m, - max_p=max_p, - max_q=max_q, - max_P=max_P, - max_Q=max_Q, - trend=trend, - method=method, - maxiter=maxiter, - information_criterion=information_criterion, - relative_improve=relative_improve, - max_k=max_k, - max_dur=max_dur, - **train_config, - ) + stepwise_search = autosarima_utils._StepwiseFitWrapper(**{**val_dict, **train_config}) filtered_models_ics = stepwise_search.stepwisesearch() if approximation: @@ -391,11 +340,10 @@ def evaluate_theta( logger.info(f"Best model: {autosarima_utils._model_name(best_model_fit.model)}") else: raise ValueError("Could not successfully fit a viable SARIMA model") + elif theta_value["action"] == "pqPQ": best_model_theta = theta_value["theta"] - order = theta_value["theta"][0] - seasonal_order = theta_value["theta"][1] - trend = theta_value["theta"][2] + order, seasonal_order, trend = theta_value["theta"] if seasonal_order[3] == 1: seasonal_order = [0, 0, 0, 0] best_model_fit, fit_time, ic = autosarima_utils._fit_sarima_model( @@ -409,6 +357,7 @@ def evaluate_theta( information_criterion=information_criterion, **train_config, ) + else: return theta_value, None, None diff --git a/merlion/models/automl/layer_mixin.py b/merlion/models/automl/base.py similarity index 61% rename from merlion/models/automl/layer_mixin.py rename to merlion/models/automl/base.py index bbc56b7c9..f968c977c 100644 --- a/merlion/models/automl/layer_mixin.py +++ b/merlion/models/automl/base.py @@ -4,35 +4,61 @@ # SPDX-License-Identifier: BSD-3-Clause # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause # -from abc import ABC +""" +Base class/mixin for AutoML hyperparameter search. +""" +from abc import abstractmethod from collections import Iterator +from copy import deepcopy from typing import Tuple, Optional, Any -from merlion.models.base import ModelBase +from merlion.models.layers import LayeredModel, LayeredModelConfig from merlion.models.forecast.base import ForecasterBase from merlion.utils import TimeSeries +from merlion.utils.misc import AutodocABCMeta -class LayerMixIn(ModelBase, ABC): +class AutoMLMixIn(LayeredModel, metaclass=AutodocABCMeta): """ - Base Interface for Implemented Layers + Base Interface for Implemented AutoML Layers This abstract class contains all of the methods that Layers should implement. Ideally, these would be generated by an existing mix-in. """ + config_class = LayeredModelConfig + + def train(self, train_data: TimeSeries, **kwargs): + train_data = self.train_pre_process( + train_data, require_even_sampling=self.require_even_sampling, require_univariate=self.require_univariate + ) + + candidate_thetas = self.generate_theta(train_data) + theta, model, train_result = self.evaluate_theta(candidate_thetas, train_data, kwargs) + if model is not None: + self.model = model + return train_result + else: + model = deepcopy(self.model) + model.reset() + self.set_theta(model, theta, train_data) + self.model = model + return self.model.train(train_data, **kwargs) + + @abstractmethod def generate_theta(self, train_data: TimeSeries) -> Iterator: - """ + r""" :param train_data: Training data to use for generation of hyperparameters :math:`\theta` Returns an iterator of hyperparameter candidates for consideration with th underlying model. """ raise NotImplementedError + @abstractmethod def evaluate_theta( self, thetas: Iterator, train_data: TimeSeries, train_config=None ) -> Tuple[Any, Optional[ForecasterBase], Optional[Tuple[TimeSeries, Optional[TimeSeries]]]]: - """ + r""" :param thetas: Iterator of the hyperparameter candidates :param train_data: Training data :param train_config: Training configuration @@ -41,8 +67,9 @@ def evaluate_theta( """ raise NotImplementedError + @abstractmethod def set_theta(self, model, theta, train_data: TimeSeries = None): - """ + r""" :param model: Underlying base model to which the new theta is applied :param theta: Hyperparameter to apply :param train_data: Training data (Optional) diff --git a/merlion/models/automl/forecasting_layer_base.py b/merlion/models/automl/forecasting_layer_base.py deleted file mode 100644 index 99f42564f..000000000 --- a/merlion/models/automl/forecasting_layer_base.py +++ /dev/null @@ -1,121 +0,0 @@ -# -# Copyright (c) 2021 salesforce.com, inc. -# All rights reserved. -# SPDX-License-Identifier: BSD-3-Clause -# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause -# -import json -import os -from abc import ABC -from copy import deepcopy -from os.path import join -from typing import Tuple, Optional, Union, List - -import dill - -from merlion.models.automl.layer_mixin import LayerMixIn -from merlion.models.factory import ModelFactory -from merlion.models.forecast.base import ForecasterBase -from merlion.utils import TimeSeries - - -class ForecasterAutoMLBase(ForecasterBase, LayerMixIn, ABC): - """ - Base Implementation of AutoML Layer Logic. - - Custom `train` and `forecast` methods that call rely on implementations of `LayerMixIn` to perform the training and - forecasting procedures. - - Note: Layer models don't have a config but any calls to their config will bubble down to the underlying model. This - may be a blessing or a curse. - """ - - def __init__(self, model: ForecasterBase, **kwargs): - """ - Assume config also inherits ForecastConfig - """ - if isinstance(model, dict): - model = ModelFactory.create(**{**model, **kwargs}) - self.model = model - - def reset(self): - self.model.reset() - self.__init__(self.model) - - def train(self, train_data: TimeSeries, train_config=None) -> Tuple[TimeSeries, Optional[TimeSeries]]: - original_train_data = train_data - train_data = self.train_pre_process(train_data, require_even_sampling=False, require_univariate=False) - - candidate_thetas = self.generate_theta(train_data) - # need to call evaluate_theta on original training data since evaluate_theta often trains another model - # and therefore we might be applying transform twice - theta, model, train_result = self.evaluate_theta(candidate_thetas, original_train_data, train_config) - if model: - self.model = model - return train_result - else: - model = deepcopy(self.model) - model.reset() - self.set_theta(model, theta, train_data) - self.model = model - return self.model.train(original_train_data, train_config) - - def forecast( - self, - time_stamps: Union[int, List[int]], - time_series_prev: TimeSeries = None, - return_iqr: bool = False, - return_prev: bool = False, - ) -> Union[Tuple[TimeSeries, Optional[TimeSeries]], Tuple[TimeSeries, TimeSeries, TimeSeries]]: - return self.model.forecast(time_stamps, time_series_prev, return_iqr, return_prev) - - def save(self, dirname: str, **save_config): - state_dict = self.__getstate__() - state_dict.pop("model") - model_path = os.path.abspath(join(dirname, self.filename)) - config_dict = dict() - - # create the directory if needed - os.makedirs(dirname, exist_ok=True) - - underlying_model_path = os.path.abspath(os.path.join(dirname, "model")) - self.model.save(underlying_model_path) - config_dict["model_name"] = type(self.model).__name__ - - with open(os.path.join(dirname, self.config_class.filename), "w") as f: - json.dump(config_dict, f, indent=2, sort_keys=True) - - # Save the model state - self._save_state(state_dict, model_path, **save_config) - - @classmethod - def load(cls, dirname: str, **kwargs): - # Read the config dict from json - config_path = os.path.join(dirname, cls.config_class.filename) - with open(config_path, "r") as f: - config_dict = json.load(f) - - model_name = config_dict.pop("model_name") - model = ModelFactory.load(model_name, os.path.abspath(os.path.join(dirname, "model"))) - - # Load the state dict - with open(os.path.join(dirname, cls.filename), "rb") as f: - state_dict = dill.load(f) - - return cls._from_config_state_dicts(state_dict, model, **kwargs) - - @classmethod - def _from_config_state_dicts(cls, state_dict, model, **kwargs): - model = cls(model) - model._load_state(state_dict, **kwargs) - - return model - - def __getattr__(self, attr): - try: - return getattr(self.model, attr) - except AttributeError: - try: - return getattr(self.model.config, attr) - except AttributeError: - raise AttributeError(f"Attribute {attr} not found in underlying class {type(self.model)}") diff --git a/merlion/models/automl/seasonality.py b/merlion/models/automl/seasonality.py new file mode 100644 index 000000000..4c74fb3fa --- /dev/null +++ b/merlion/models/automl/seasonality.py @@ -0,0 +1,182 @@ +# +# Copyright (c) 2021 salesforce.com, inc. +# All rights reserved. +# SPDX-License-Identifier: BSD-3-Clause +# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause +# +""" +Automatic seasonality detection. +""" +from abc import abstractmethod +from enum import Enum, auto +import logging +from typing import Any, Iterator, Optional, Tuple, Union + +from merlion.models.automl.base import AutoMLMixIn +from merlion.models.base import ModelBase +from merlion.models.layers import LayeredModelConfig +from merlion.transform.resample import TemporalResample +from merlion.utils import TimeSeries, UnivariateTimeSeries, autosarima_utils +from merlion.utils.misc import AutodocABCMeta + +logger = logging.getLogger(__name__) + + +class PeriodicityStrategy(Enum): + """ + Strategy to choose the seasonality if multiple candidates are detected. + """ + + ACF = auto() + """ + Select the seasonality value with the highest autocorrelation. + """ + Min = auto() + """ + Select the minimum seasonality. + """ + Max = auto() + """ + Select the maximum seasonality. + """ + All = auto() + """ + Use all seasonalities. Only valid for models which support multiple seasonalities. + """ + + +class SeasonalityModel(metaclass=AutodocABCMeta): + """ + Class provides simple implementation to set the seasonality in a model. Extend this class to implement custom + behavior for seasonality processing. + """ + + @abstractmethod + def set_seasonality(self, theta, train_data: UnivariateTimeSeries): + """ + Implement this method to do any model-specific adjustments on the seasonality that was provided by + `SeasonalityLayer`. + + :param theta: Seasonality processed by `SeasonalityLayer`. + :param train_data: Training data (or numpy array representing the target univariate) + for any model-specific adjustments you might want to make. + """ + raise NotImplementedError + + +class SeasonalityConfig(LayeredModelConfig): + """ + Config object for an automatic seasonality detection layer. + """ + + _default_transform = TemporalResample() + + def __init__(self, model, periodicity_strategy=PeriodicityStrategy.ACF, pval: float = 0.05, **kwargs): + """ + :param periodicity_strategy: Strategy to choose the seasonality if multiple candidates are detected. + :param pval: p-value for deciding whether a detected seasonality is statistically significant. + """ + self.periodicity_strategy = periodicity_strategy + assert 0 < pval < 1 + self.pval = pval + super().__init__(model=model, **kwargs) + + @property + def multi_seasonality(self): + """ + :return: Whether the model supports multiple seasonalities. ``False`` unless explicitly overridden. + """ + return False + + @property + def periodicity_strategy(self) -> PeriodicityStrategy: + """ + :return: Strategy to choose the seasonality if multiple candidates are detected. + """ + return self._periodicity_strategy + + @periodicity_strategy.setter + def periodicity_strategy(self, p: Union[PeriodicityStrategy, str]): + if not isinstance(p, PeriodicityStrategy): + valid = {k.lower(): k for k in PeriodicityStrategy.__members__} + assert p.lower() in valid, f"Unsupported PeriodicityStrategy {p}. Supported strategies are: {valid.keys()}" + p = PeriodicityStrategy[valid[p.lower()]] + + if p is PeriodicityStrategy.All and not self.multi_seasonality: + raise ValueError( + "Periodicity strategy All is not supported for a model which does not support multiple seasonalities." + ) + + self._periodicity_strategy = p + + def to_dict(self, _skipped_keys=None): + _skipped_keys = _skipped_keys if _skipped_keys is not None else set() + config_dict = super().to_dict(_skipped_keys.union({"periodicity_strategy"})) + if "periodicity_strategy" not in _skipped_keys: + config_dict["periodicity_strategy"] = self.periodicity_strategy.name + return config_dict + + +class SeasonalityLayer(AutoMLMixIn, metaclass=AutodocABCMeta): + """ + Seasonality Layer that uses AutoSARIMA-like methods to determine seasonality of your data. Can be used directly on + any model that implements `SeasonalityModel` class. + """ + + config_class = SeasonalityConfig + require_even_sampling = False + + @property + def require_univariate(self): + return getattr(self.config, "target_seq_index", None) is not None + + @property + def multi_seasonality(self): + """ + :return: Whether the model supports multiple seasonalities. + """ + return self.config.multi_seasonality + + @property + def periodicity_strategy(self): + """ + :return: Strategy to choose the seasonality if multiple candidates are detected. + """ + return self.config.periodicity_strategy + + @property + def pval(self): + """ + :return: p-value for deciding whether a detected seasonality is statistically significant. + """ + return self.config.pval + + def set_theta(self, model, theta, train_data: TimeSeries = None): + model.set_seasonality(theta, train_data.univariates[self.target_name]) + + def evaluate_theta( + self, thetas: Iterator, train_data: TimeSeries, train_config=None + ) -> Tuple[Any, Optional[ModelBase], Optional[Tuple[TimeSeries, Optional[TimeSeries]]]]: + # If multiple seasonalities are supported, return a list of all detected seasonalities + thetas = list(thetas) + if self.periodicity_strategy is PeriodicityStrategy.ACF: + thetas = [thetas[0]] + elif self.periodicity_strategy is PeriodicityStrategy.Min: + thetas = [min(thetas)] + elif self.periodicity_strategy is PeriodicityStrategy.Max: + thetas = [max(thetas)] + elif self.periodicity_strategy is PeriodicityStrategy.All: + thetas = thetas + else: + raise ValueError(f"Periodicity strategy {self.periodicity_strategy} not supported.") + theta = thetas if self.config.multi_seasonality else thetas[0] + if thetas != [1]: + logger.info(f"Automatically detect the periodicity is {str(thetas)}") + return theta, None, None + + def generate_theta(self, train_data: TimeSeries) -> Iterator: + y = train_data.univariates[self.target_name] + periods = autosarima_utils.multiperiodicity_detection(y, pval=self.pval) + if len(periods) == 0: + periods = [1] + return iter(periods) diff --git a/merlion/models/automl/seasonality_mixin.py b/merlion/models/automl/seasonality_mixin.py deleted file mode 100644 index a12c3f821..000000000 --- a/merlion/models/automl/seasonality_mixin.py +++ /dev/null @@ -1,62 +0,0 @@ -# -# Copyright (c) 2021 salesforce.com, inc. -# All rights reserved. -# SPDX-License-Identifier: BSD-3-Clause -# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause -# -from abc import ABC -from typing import Iterator, Tuple, Optional, Any - -from merlion.models.automl.forecasting_layer_base import ForecasterAutoMLBase -from merlion.models.forecast.base import ForecasterBase, logger -from merlion.utils import TimeSeries, autosarima_utils - - -class SeasonalityModel(ABC): - """ - Class provides simple implementation to set the seasonality in a model. Extend this class to implement custom - behavior for seasonality processing. - """ - - def set_seasonality(self, theta, train_data): - """ - Implement this method to do any model-specific adjustments on the seasonality that was provided by - `SeasonalityLayer`. - - :param theta: Seasonality processed by `SeasonalityLayer`. - :param train_data: Training data (or numpy array representing the target univariate) - for any model-specific adjustments you might want to make. - """ - self.seasonality = theta - - -class SeasonalityLayer(ForecasterAutoMLBase, ABC): - """ - Seasonality Layer that uses AutoSARIMA-like methods to determine seasonality of your data. Can be used directly on - any model that implements `SeasonalityModel` class. - """ - - def set_theta(self, model, theta, train_data: TimeSeries = None): - model.set_seasonality(theta, train_data.univariates[self.target_name]) - - def evaluate_theta( - self, thetas: Iterator, train_data: TimeSeries, train_config=None - ) -> Tuple[Any, Optional[ForecasterBase], Optional[Tuple[TimeSeries, Optional[TimeSeries]]]]: - # assume only one seasonality is returned in this case - return next(thetas), None, None - - def generate_theta(self, train_data: TimeSeries) -> Iterator: - y = train_data.univariates[self.target_name] - - periodicity_strategy = self.periodicity_strategy - - periods = autosarima_utils.multiperiodicity_detection(y) - if len(periods) > 0: - if periodicity_strategy == "min": - m = periods[0] - else: - m = periods[-1] - else: - m = 1 - logger.info(f"Automatically detect the periodicity is {str(m)}") - return iter([m]) diff --git a/merlion/models/base.py b/merlion/models/base.py index a365dbd03..efcfb67f4 100644 --- a/merlion/models/base.py +++ b/merlion/models/base.py @@ -24,46 +24,25 @@ from merlion.transform.normalize import Rescale, MeanVarNormalize from merlion.transform.sequence import TransformSequence from merlion.utils.time_series import assert_equal_timedeltas, to_pd_datetime, TimeSeries -from merlion.utils.misc import AutodocABCMeta +from merlion.utils.misc import AutodocABCMeta, ModelConfigMeta logger = logging.getLogger(__name__) -def override_config(config, config_dict, return_unused_kwargs=False, **kwargs): - """ - :meta private: - """ - to_remove = [] - for key, value in kwargs.items(): - if hasattr(config, key): - setattr(config, key, value) - to_remove.append(key) - - for key in to_remove: - kwargs.pop(key) - - for key, value in config_dict.items(): - if key not in kwargs and not hasattr(config, key): - kwargs[key] = value - - if len(kwargs) > 0 and not return_unused_kwargs: - logger.warning(f"Unused kwargs: {kwargs}", stack_info=True) - elif return_unused_kwargs: - return config, kwargs - return config - - -class Config(object): +class Config(object, metaclass=ModelConfigMeta): """ Abstract class which defines a model config. """ filename = "config.json" _default_transform = Identity() + transform: TransformBase = None + dim: Optional[int] = None def __init__(self, transform: TransformBase = None, **kwargs): """ :param transform: Transformation to pre-process input time series. + :param dim: The dimension of the time series """ super().__init__() if transform is None: @@ -72,6 +51,14 @@ def __init__(self, transform: TransformBase = None, **kwargs): self.transform = TransformFactory.create(**transform) else: self.transform = transform + self.dim = None + + @property + def base_model(self): + """ + The base model of a base model is itself. + """ + return self def to_dict(self, _skipped_keys=None): """ @@ -89,20 +76,32 @@ def to_dict(self, _skipped_keys=None): return config_dict @classmethod - def from_dict(cls, config_dict: Dict[str, Any], return_unused_kwargs=False, **kwargs): + def from_dict(cls, config_dict: Dict[str, Any], return_unused_kwargs=False, dim=None, **kwargs): """ Constructs a `Config` from a Python dictionary of parameters. :param config_dict: dict that will be used to instantiate this object. :param return_unused_kwargs: whether to return any unused keyword args. + :param dim: the dimension of the time series. handled as a special case. :param kwargs: any additional parameters to set (overriding config_dict). :return: `Config` object initialized from the dict. """ + dim = config_dict.pop("dim", dim) + config_dict = dict(**config_dict, **kwargs) config = cls(**config_dict) - return override_config( - config=config, config_dict=config_dict, return_unused_kwargs=return_unused_kwargs, **kwargs - ) + if dim is not None: + config.dim = dim + + kwargs = config.get_unused_kwargs(**config_dict) + if len(kwargs) > 0 and not return_unused_kwargs: + logger.warning(f"Unused kwargs: {kwargs}", stack_info=True) + elif return_unused_kwargs: + return config, kwargs + return config + + def __reduce__(self): + return self.__class__.from_dict, (self.to_dict(),) def __copy__(self): return self.from_dict(self.to_dict()) @@ -110,6 +109,9 @@ def __copy__(self): def __deepcopy__(self, memodict={}): return self.__copy__() + def get_unused_kwargs(self, **kwargs): + return {k: v for k, v in kwargs.items() if k not in self.to_dict()} + class NormalizingConfig(Config): """ @@ -161,6 +163,11 @@ class ModelBase(metaclass=AutodocABCMeta): config_class = Config _default_train_config = None + train_data: Optional[TimeSeries] = None + """ + The data used to train the model. + """ + def __init__(self, config: Config): assert isinstance(config, self.config_class) self.config = deepcopy(config) @@ -187,6 +194,15 @@ def __setstate__(self, state): f"'{name}' is an invalid kwarg for the load() method." ) + def __reduce__(self): + state_dict = self.__getstate__() + config = state_dict.pop("config") + return self.__class__, (config,), state_dict + + @property + def dim(self): + return self.config.dim + @property def transform(self): """ @@ -239,6 +255,7 @@ def train_pre_process( :return: the training data, after any necessary pre-processing has been applied """ self.train_data = train_data + self.config.dim = train_data.dim self.transform.train(train_data) train_data = self.transform(train_data) @@ -383,7 +400,7 @@ def _from_config_state_dicts(cls, config_dict, state_dict, **kwargs): :return: `ModelBase` object loaded from file """ config, model_kwargs = cls.config_class.from_dict(config_dict, return_unused_kwargs=True, **kwargs) - model = cls(config) + model = cls(config=config) model._load_state(state_dict, **model_kwargs) return model @@ -415,7 +432,7 @@ def from_bytes(cls, obj, **kwargs): return cls._from_config_state_dicts(config_dict, state_dict, **kwargs) def __copy__(self): - new_model = self.__class__(deepcopy(self.config)) + new_model = self.__class__(config=deepcopy(self.config)) state_dict = self.__getstate__() state_dict.pop("config", None) new_model.__setstate__(state_dict) @@ -423,56 +440,3 @@ def __copy__(self): def __deepcopy__(self, memodict={}): return self.__copy__() - - -class ModelWrapper(ModelBase, metaclass=AutodocABCMeta): - """ - Abstract class implementing a model that wraps around another internal model. - """ - - filename = "model" - - def __init__(self, config: Config, model: ModelBase = None): - super().__init__(config) - self.model = model - - def save(self, dirname: str, **save_config): - config_dict = self.config.to_dict() - config_dict["model_type"] = type(self.model).__name__ - os.makedirs(dirname, exist_ok=True) - with open(os.path.join(dirname, self.config_class.filename), "w") as f: - json.dump(config_dict, f, indent=2, sort_keys=True) - self.model.save(os.path.join(dirname, self.filename), **save_config) - - @classmethod - def load(cls, dirname: str, **kwargs): - from merlion.models.factory import ModelFactory - - config_path = os.path.join(dirname, cls.config_class.filename) - with open(config_path, "r") as f: - config_dict = json.load(f) - - model_type = config_dict.pop("model_type") - model = ModelFactory.load(model_type, os.path.join(dirname, cls.filename)) - return cls._from_config_state_dicts(config_dict, model, **kwargs) - - @classmethod - def _from_config_state_dicts(cls, config_dict, model, **kwargs): - config = cls.config_class.from_dict(config_dict) - ret = cls(config=config) - ret.model = model - return ret - - def to_bytes(self, **save_config): - config_dict = self.config.to_dict() - model_tuple = self.model._to_serializable_comps(**save_config) - class_name = type(self).__name__ - return dill.dumps((class_name, config_dict, model_tuple)) - - @classmethod - def from_bytes(cls, obj, **kwargs): - from merlion.models.factory import ModelFactory - - class_name, config_dict, model_tuple = dill.loads(obj) - model = [ModelFactory.get_model_class(model_tuple[0])._from_config_state_dicts(*model_tuple[1:])] - return cls._from_config_state_dicts(config_dict, model, **kwargs) diff --git a/merlion/models/defaults.py b/merlion/models/defaults.py index 4b3e76999..d781229d0 100644 --- a/merlion/models/defaults.py +++ b/merlion/models/defaults.py @@ -6,45 +6,35 @@ # """Default models for anomaly detection & forecasting that balance speed and performance.""" import logging -from typing import List, Optional, Tuple, Union +from typing import Optional, Tuple from merlion.models.factory import ModelFactory -from merlion.models.base import Config, ModelWrapper -from merlion.models.anomaly.base import DetectorConfig, DetectorBase -from merlion.models.forecast.base import ForecasterConfig, ForecasterBase +from merlion.models.layers import LayeredDetector, LayeredForecaster, LayeredModelConfig +from merlion.models.anomaly.base import DetectorBase +from merlion.models.forecast.base import ForecasterBase from merlion.utils import TimeSeries logger = logging.getLogger(__name__) -class DefaultModelConfig(Config): - def __init__(self, granularity=None, **kwargs): - super().__init__() - self.granularity = granularity - - def to_dict(self, _skipped_keys=None): - _skipped_keys = set() if _skipped_keys is None else _skipped_keys - return super().to_dict(_skipped_keys.union("transform")) - - -class DefaultDetectorConfig(DetectorConfig, DefaultModelConfig): +class DefaultDetectorConfig(LayeredModelConfig): """ Config object for default anomaly detection model. """ - def __init__(self, granularity=None, threshold=None, n_threads: int = 1, **kwargs): + def __init__(self, model=None, granularity=None, n_threads: int = 1, **kwargs): """ :param granularity: the granularity at which the input time series should be sampled, e.g. "5min", "1h", "1d", etc. - :param threshold: `Threshold` object setting a default anomaly detection - threshold in units of z-score. :param n_threads: the number of parallel threads to use for relevant models """ - super().__init__(granularity=granularity, threshold=threshold, enable_threshold=True, enable_calibrator=False) + self.granularity = granularity self.n_threads = n_threads + super().__init__(model=model, **kwargs) + assert self.base_model is None or isinstance(self.base_model, DetectorBase) -class DefaultDetector(ModelWrapper, DetectorBase): +class DefaultDetector(LayeredDetector): """ Default anomaly detection model that balances efficiency with performance. """ @@ -71,7 +61,6 @@ def train( if train_data.dim > 1: self.model = ModelFactory.create( "DetectorEnsemble", - enable_threshold=False, models=[ ModelFactory.create("VAE", transform=transform_dict), ModelFactory.create( @@ -91,10 +80,9 @@ def train( ets_transform = dict(name="TemporalResample", granularity=dt) self.model = ModelFactory.create( "DetectorEnsemble", - enable_threshold=False, models=[ ModelFactory.create( - "ETSDetector", damped_trend=True, max_forecast_steps=None, transform=ets_transform + "AutoETS", model=dict(name="ETSDetector"), damped_trend=True, transform=ets_transform ), ModelFactory.create( "RandomCutForest", @@ -108,47 +96,40 @@ def train( ], ) - train_data = self.train_pre_process(train_data, False, False) - train_scores = self.model.train( + return super().train( train_data=train_data, anomaly_labels=anomaly_labels, train_config=train_config, post_rule_train_config=post_rule_train_config, ) - self.train_post_rule( - anomaly_scores=train_scores, anomaly_labels=anomaly_labels, post_rule_train_config=post_rule_train_config - ) - return train_scores - - def get_anomaly_score(self, time_series: TimeSeries, time_series_prev: TimeSeries = None) -> TimeSeries: - # we use get_anomaly_label() because the underlying model's calibration is - # enabled, but its threshold is enabled - time_series, time_series_prev = self.transform_time_series(time_series, time_series_prev) - return self.model.get_anomaly_label(time_series, time_series_prev) - def get_anomaly_label(self, time_series: TimeSeries, time_series_prev: TimeSeries = None) -> TimeSeries: - return super().get_anomaly_label(time_series, time_series_prev) - -class DefaultForecasterConfig(ForecasterConfig, DefaultModelConfig): +class DefaultForecasterConfig(LayeredModelConfig): """ Config object for default forecasting model. """ - def __init__(self, granularity=None, max_forecast_steps=100, target_seq_index=None, **kwargs): + def __init__(self, model=None, max_forecast_steps=100, target_seq_index=None, granularity=None, **kwargs): """ + :param max_forecast_steps: Max # of steps we would like to forecast for. + Required for some models like `MSES` and `LGBMForecaster`. + :param target_seq_index: The index of the univariate (amongst all + univariates in a general multivariate time series) whose value we + would like to forecast. :param granularity: the granularity at which the input time series should be sampled, e.g. "5min", "1h", "1d", etc. - :param max_forecast_steps: Max # of steps we would like to forecast for. - :param target_seq_index: If doing multivariate forecasting, the index of - univariate whose value you wish to forecast. """ super().__init__( - granularity=granularity, max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index + model=model, + max_forecast_steps=max_forecast_steps, + target_seq_index=target_seq_index, + granularity=granularity, + model_kwargs=kwargs, ) + assert self.base_model is None or isinstance(self.base_model, ForecasterBase) -class DefaultForecaster(ModelWrapper, ForecasterBase): +class DefaultForecaster(LayeredForecaster): """ Default forecasting model that balances efficiency with performance. """ @@ -181,42 +162,4 @@ def train(self, train_data: TimeSeries, train_config=None) -> Tuple[TimeSeries, # ETS for univariate data else: self.model = ModelFactory.create("ETS", damped_trend=True, **kwargs) - train_data = self.train_pre_process(train_data, False, False) - return self.model.train(train_data=train_data, train_config=train_config) - - def forecast( - self, - time_stamps: Union[int, List[int]], - time_series_prev: TimeSeries = None, - return_iqr: bool = False, - return_prev: bool = False, - ) -> Union[Tuple[TimeSeries, Optional[TimeSeries]], Tuple[TimeSeries, TimeSeries, TimeSeries]]: - """ - Returns the model's forecast on the timestamps given. - - :param time_stamps: Either a ``list`` of timestamps we wish to forecast for, - or the number of steps (``int``) we wish to forecast for. - :param time_series_prev: a list of (timestamp, value) pairs immediately - preceding ``time_series``. If given, we use it to initialize the time - series model. Otherwise, we assume that ``time_series`` immediately - follows the training data. - :param return_iqr: whether to return the inter-quartile range for the - forecast. Note that not all models support this option. - :param return_prev: whether to return the forecast for - ``time_series_prev`` (and its stderr or IQR if relevant), in addition - to the forecast for ``time_stamps``. Only used if ``time_series_prev`` - is provided. - :return: ``(forecast, forecast_stderr)`` if ``return_iqr`` is false, - ``(forecast, forecast_lb, forecast_ub)`` otherwise. - - - ``forecast``: the forecast for the timestamps given - - ``forecast_stderr``: the standard error of each forecast value. - May be ``None``. - - ``forecast_lb``: 25th percentile of forecast values for each timestamp - - ``forecast_ub``: 75th percentile of forecast values for each timestamp - """ - if time_series_prev is not None: - time_series_prev = self.transform(time_series_prev) - return self.model.forecast( - time_stamps=time_stamps, time_series_prev=time_series_prev, return_iqr=return_iqr, return_prev=return_prev - ) + return super().train(train_data=train_data, train_config=train_config) diff --git a/merlion/models/ensemble/MoE_forecast.py b/merlion/models/ensemble/MoE_forecast.py index 36ce914b3..56f73ac44 100644 --- a/merlion/models/ensemble/MoE_forecast.py +++ b/merlion/models/ensemble/MoE_forecast.py @@ -6,12 +6,9 @@ # """Mixture of Expert forecasters.""" __author__ = "Devansh Arpit" -import json import logging -import os -from typing import List, Optional, Tuple +from typing import Any, Dict, List, Optional, Tuple -import dill import numpy as np import torch from torch import nn @@ -23,7 +20,6 @@ from merlion.models.forecast.base import ForecasterConfig, ForecasterBase from merlion.utils import TimeSeries, UnivariateTimeSeries from merlion.models.ensemble.MoE_networks import TransformerModel, LRScheduler -from merlion.models.factory import ModelFactory logger = logging.getLogger(__name__) @@ -100,7 +96,6 @@ def sorted_preds(preds): """ if preds == []: return [] - out = [] s = sorted(range(len(preds)), key=lambda x: preds[x]) out = [preds[i][1] for i in s] return out @@ -181,7 +176,7 @@ def smape_f1_loss(output, std, target, thres=0.1): ########################## End helper functions ########################## -class MoE_ForecasterEnsembleConfig(ForecasterConfig, EnsembleConfig, NormalizingConfig): +class MoE_ForecasterEnsembleConfig(EnsembleConfig, ForecasterConfig, NormalizingConfig): """ Config class for MoE (mixture of experts) forecaster. """ @@ -260,29 +255,43 @@ def __init__( (B x nexperts x max_forecast_steps). The second variable is None if nfree_experts=0, else has size (nfree_experts x max_forecast_steps) which is the forecasted values by nfree_experts number of experts. """ - super().__init__(config, models) + super().__init__(config=config, models=models) + for model in self.models: + assert isinstance(model, ForecasterBase), ( + f"Expected all models in {type(self).__name__} to be anomaly " + f"detectors, but got a {type(model).__name__}." + ) self.loss_list = [] - condition1 = config.nfree_experts > 0 - condition2 = len(models) > 0 + condition1 = self.config.nfree_experts > 0 + condition2 = len(self.models) > 0 assert (not (condition1 and condition2)) and (condition1 or condition2), ( - f"Number of free experts (nfree_experts={config.nfree_experts}) " - f"and number of external experts (#models={len(models)}) cannot be " + f"Number of free experts (nfree_experts={self.config.nfree_experts}) " + f"and number of external experts (#models={len(self.models)}) cannot be " f"greater than 0 at the same time, but one of them must be non-zero." ) self.moe_model = moe_model - if self.moe_model is not None: - self.optimiser = torch.optim.Adam(self.moe_model.parameters(), lr=self.lr, weight_decay=0.00000) - self.lr_sch = LRScheduler(lr_i=0.0000, lr_f=self.lr, nsteps=self.warmup_steps, optimizer=self.optimiser) - self.nexperts = len(models) + @property + def moe_model(self): + return self._moe_model - for model in self.models: - assert isinstance(model, ForecasterBase), ( - f"Expected all models in {type(self).__name__} to be anomaly " - f"detectors, but got a {type(model).__name__}." - ) + @moe_model.setter + def moe_model(self, moe_model): + self._moe_model = moe_model + if self.moe_model is not None: + if self.optimiser is None: + self.optimiser = torch.optim.Adam(self.moe_model.parameters(), lr=self.lr, weight_decay=0.00000) + if self.lr_sch is None: + self.lr_sch = LRScheduler(lr_i=0.0000, lr_f=self.lr, nsteps=self.warmup_steps, optimizer=self.optimiser) + else: + self.optimiser = None + self.lr_sch = None + + @property + def nexperts(self): + return len(self.models) @property def batch_size(self) -> int: @@ -927,26 +936,8 @@ def save(self, dirname: str, **save_config): :param dirname: directory to save the model :param save_config: additional configurations (if needed) """ - state_dict = self.__getstate__() - # remove items that should not be saved - for key in ["config", "train_data", "models", "moe_model", "mn", "std", "optimiser", "lr_sch"]: - state_dict.pop(key) - - config_dict = self.config.to_dict() - # create the directory if needed - os.makedirs(dirname, exist_ok=True) - - paths = [] - for i, model in enumerate(self.models): - path = os.path.abspath(os.path.join(dirname, str(i))) - paths.append(path) - model.save(path) - - # Add model paths to the config dict, and save it - config_dict["model_paths"] = [(type(m).__name__, p) for m, p in zip(self.models, paths)] - with open(os.path.join(dirname, self.config.filename), "w") as f: - json.dump(config_dict, f, indent=2, sort_keys=True) - + # Save MoE transformer state separately from the rest of the model state + super().save(dirname, **save_config) state = { "model_params": self.moe_model.state_dict(), "optimiser": self.optimiser.state_dict(), @@ -956,13 +947,12 @@ def save(self, dirname: str, **save_config): with open(dirname + "/torch_params.pth.tar", "wb") as f: torch.save(state, f) - # Save the remaining ensemble state - self._save_state( - state_dict=state_dict, - filename=os.path.join(dirname, self.filename), - save_only_used_models=False, - **save_config, - ) + def _save_state( + self, state_dict: Dict[str, Any], filename: str = None, save_only_used_models=False, **save_config + ) -> Dict[str, Any]: + for key in ["train_data", "_moe_model", "mn", "std", "optimiser", "lr_sch"]: + state_dict.pop(key) + return super()._save_state(state_dict, filename, save_only_used_models=save_only_used_models, **save_config) @classmethod def load(cls, dirname: str, **kwargs): @@ -970,27 +960,21 @@ def load(cls, dirname: str, **kwargs): Note: if a user specified model was used while saving the MoE ensemble, specify argument ``moe_model`` when calling the load function with the pytorch model that was used in the original MoE ensemble. If ``moe_model`` is not specified, it will be assumed that the default Pytorch network was used. Any - discrepency between the saved model state and model used here will raise an error. + discrepancy between the saved model state and model used here will raise an error. - :param dirname: directory to save the model + :param dirname: directory to load the model from """ - config_path = os.path.join(dirname, cls.config_class.filename) - with open(config_path, "r") as f: - config_dict = json.load(f) - config_dict.pop("max_forecast_steps") - # Load all the models from the config dict - model_paths = config_dict.pop("model_paths") - models = [ModelFactory.load(name=name, model_path=path) for name, path in model_paths] - - config = MoE_ForecasterEnsembleConfig.from_dict(config_dict) + loaded_ensemble = super().load(dirname, **kwargs) + # Load the MoE model state + config = loaded_ensemble.config if "moe_model" in kwargs: - moe_model = kwargs["moe_model"] + loaded_ensemble.moe_model = kwargs["moe_model"] else: - moe_model = TransformerModel( + loaded_ensemble.moe_model = TransformerModel( input_dim=config.dim, lookback_len=config.lookback_len, - nexperts=len(models), + nexperts=loaded_ensemble.nexperts, output_dim=config.max_forecast_steps, nfree_experts=config.nfree_experts, hid_dim=256, @@ -1001,8 +985,6 @@ def load(cls, dirname: str, **kwargs): time_step_dropout=0, ) - loaded_ensemble = cls(config=config, models=models, moe_model=moe_model) - state = torch.load(dirname + "/torch_params.pth.tar", map_location="cuda:0" if config.use_gpu else "cpu") try: loaded_ensemble.moe_model.load_state_dict(state["model_params"]) @@ -1011,9 +993,4 @@ def load(cls, dirname: str, **kwargs): raise RuntimeError(f"Found error while loading parameter states/optimizer states of the moe_model: {e}") loaded_ensemble.mn = state["mean"] loaded_ensemble.std = state["std"] - - # Load the state dict - with open(os.path.join(dirname, loaded_ensemble.filename), "rb") as f: - state_dict = dill.load(f) - loaded_ensemble._load_state(state_dict) return loaded_ensemble diff --git a/merlion/models/ensemble/anomaly.py b/merlion/models/ensemble/anomaly.py index 1743c39ec..df1fef7c5 100644 --- a/merlion/models/ensemble/anomaly.py +++ b/merlion/models/ensemble/anomaly.py @@ -64,7 +64,7 @@ class DetectorEnsemble(EnsembleBase, DetectorBase): _default_train_config = EnsembleTrainConfig(valid_frac=0.0) def __init__(self, config: DetectorEnsembleConfig = None, models: List[DetectorBase] = None): - super().__init__(config, models) + super().__init__(config=config, models=models) for model in self.models: assert isinstance(model, DetectorBase), ( f"Expected all models in {type(self).__name__} to be anomaly " diff --git a/merlion/models/ensemble/base.py b/merlion/models/ensemble/base.py index 61327983d..1af0f0f5a 100644 --- a/merlion/models/ensemble/base.py +++ b/merlion/models/ensemble/base.py @@ -7,14 +7,11 @@ """ Base class for ensembles of models. """ -from abc import ABC import copy -import json import logging -import dill -import os -from typing import Dict, List, Tuple, Union +from typing import Any, Dict, List, Tuple, Union +import numpy as np import pandas as pd from pandas.tseries.frequencies import to_offset @@ -22,32 +19,25 @@ from merlion.models.ensemble.combine import CombinerBase, CombinerFactory, Mean from merlion.models.factory import ModelFactory from merlion.utils import TimeSeries +from merlion.utils.misc import AutodocABCMeta logger = logging.getLogger(__name__) class EnsembleConfig(Config): """ - An ensemble config contains the configs of each individual model in the ensemble, - as well as the combiner object to combine those models' outputs. + An ensemble config contains the each individual model in the ensemble, as well as the Combiner object + to combine those models' outputs. The rationale behind placing the model objects in the EnsembleConfig + (rather than in the Ensemble itself) is discussed in more detail in the documentation for `LayeredModel`. """ _default_combiner = Mean(abs_score=False) + models: List[ModelBase] - def __init__( - self, model_configs: List[Tuple[str, Union[Config, Dict]]] = None, combiner: CombinerBase = None, **kwargs - ): + def __init__(self, models: List[Union[ModelBase, Dict]] = None, combiner: CombinerBase = None, **kwargs): """ - :param model_configs: A list of ``(class_name, config)`` tuples, where - ``class_name`` is the name of the model's class (as you would - provide to the `ModelFactory`), and ``config`` is its config or a - dict. Note that ``model_configs`` is not serialized by - `EnsembleConfig.to_dict`! The individual models are handled by - `EnsembleBase.save`. If ``model_configs`` is not provided, you are - expected to provide the ``models`` directly when initializing the - `EnsembleBase`. - :param combiner: The combiner object to combine the outputs of the - models in the ensemble. + :param models: A list of models or dicts representing them. + :param combiner: The `CombinerBase` object to combine the outputs of the models in the ensemble. :param kwargs: Any additional kwargs for `Config` """ super().__init__(**kwargs) @@ -58,20 +48,19 @@ def __init__( else: self.combiner = combiner - if model_configs is not None: - model_configs = [ - (name, copy.deepcopy(config)) - if isinstance(config, Config) - else (name, ModelFactory.get_model_class(name).config_class.from_dict(config)) - for name, config in model_configs - ] - self.model_configs = model_configs + if models is not None: + models = [ModelFactory.create(**m) if isinstance(m, dict) else copy.deepcopy(m) for m in models] + self.models = models def to_dict(self, _skipped_keys=None): - config_dict = super().to_dict(_skipped_keys) - model_configs = config_dict["model_configs"] - if model_configs is not None: - config_dict["model_configs"] = [(name, config.to_dict()) for name, config in model_configs] + _skipped_keys = _skipped_keys if _skipped_keys is not None else set() + config_dict = super().to_dict(_skipped_keys.union({"models"})) + if self.models is None: + models = None + else: + models = [None if m is None else dict(name=type(m).__name__, **m.config.to_dict()) for m in self.models] + if "models" not in _skipped_keys: + config_dict["models"] = models return config_dict @@ -93,46 +82,52 @@ def __init__(self, valid_frac, per_model_train_configs=None): self.per_model_train_configs = per_model_train_configs -class EnsembleBase(ModelBase, ABC): +class EnsembleBase(ModelBase, metaclass=AutodocABCMeta): """ An abstract class representing an ensemble of multiple models. """ - models: List[ModelBase] config_class = EnsembleConfig - _default_train_config = EnsembleTrainConfig(valid_frac=0.0) def __init__(self, config: EnsembleConfig = None, models: List[ModelBase] = None): """ - Initializes the ensemble according to the specified config. - :param config: The ensemble's config - :param models: The models in the ensemble. Only provide this argument if - you did not specify ``config.model_configs``, and you want to - initialize an ensemble from models that have already been - constructed. + :param models: The models in the ensemble. Only provide this argument if you did not specify ``config.models``. """ - msg = ( - "When initializing an ensemble, you must either provide the dict " - "`model_configs` (mapping each model's name to its config) when " - "creating the `DetectorEnsembleConfig`, or provide a list of " - "`models` to the constructor of `EnsembleBase`." - ) - config = self.config_class() if config is None else config - if config.model_configs is None and models is None: + msg = f"Expected exactly one of `config.models` or `models` when creating a {type(self).__name__}." + if config is None and models is None: raise RuntimeError(f"{msg} Received neither.") - elif config.model_configs is not None and models is not None: - logger.warning(f"{msg} Received both. Overriding `model_configs` with the configs belonging to `models`.") + elif config is not None and models is not None: + if config.models is None: + config.models = models + else: + raise RuntimeError(f"{msg} Received both.") + elif config is None: + config = self.config_class(models=models) + super().__init__(config=config) - if models is not None: - models = [copy.deepcopy(model) for model in models] - config.model_configs = [(type(model).__name__, model.config) for model in models] - else: - models = [ModelFactory.create(name, **config.to_dict()) for name, config in config.model_configs] + @property + def models(self): + return self.config.models - super().__init__(config) - self.models = models + @property + def combiner(self) -> CombinerBase: + """ + :return: the object used to combine model outputs. + """ + return self.config.combiner + + def reset(self): + for model in self.models: + model.reset() + + @property + def models_used(self): + if self.combiner.n_models is not None: + return self.combiner.models_used + else: + return [True] * len(self.models) def train_valid_split( self, transformed_train_data: TimeSeries, train_config: EnsembleTrainConfig @@ -159,7 +154,8 @@ def get_max_common_horizon(self): horizons.append(h) if all(h is None for h in horizons): return None - return min([h for h in horizons if h is not None]) + i = np.argmin([pd.to_datetime(0) + h for h in horizons if h is not None]) + return horizons[i] def truncate_valid_data(self, transformed_valid_data: TimeSeries): tf = transformed_valid_data.tf @@ -177,121 +173,71 @@ def truncate_valid_data(self, transformed_valid_data: TimeSeries): def train_combiner(self, all_model_outs: List[TimeSeries], target: TimeSeries) -> TimeSeries: return self.combiner.train(all_model_outs, target) - @property - def combiner(self) -> CombinerBase: - """ - :return: the object used to combine model outputs. - """ - return self.config.combiner - - def reset(self): - for model in self.models: - model.reset() - - @property - def models_used(self): - if self.combiner.n_models is not None: - return self.combiner.models_used + def __getstate__(self): + state = super().__getstate__() + if self.models is None: + state["models"] = None else: - return [True] * len(self.models) + state["models"] = [None if model is None else model.__getstate__() for model in self.models] + return state + + def __setstate__(self, state): + if "models" in state: + model_states = state.pop("models") + if self.models is None and model_states is not None: + raise ValueError(f"`{type(self).__name__}.models` is None, but received a non-None `models` state.") + elif self.models is None or model_states is None: + self.config.models = None + else: + for i, (model, model_state) in enumerate(zip(self.models, model_states)): + if model is None and model_state is not None: + raise ValueError(f"One of the Ensemble models is None, but received a non-None model state.") + elif model is None or model_state is None: + self.models[i] = None + else: + model.__setstate__(model_state) + super().__setstate__(state) def save(self, dirname: str, save_only_used_models=False, **save_config): """ Saves the ensemble of models. :param dirname: directory to save the ensemble to - :param save_only_used_models: whether to save only the models that are - actually used by the ensemble. + :param save_only_used_models: whether to save only the models that are actually used by the ensemble. :param save_config: additional save config arguments """ - state_dict = self.__getstate__() - state_dict.pop("models") # to remove from the state dict - config_dict = self.config.to_dict() - config_dict.pop("model_configs", None) # should save/load models directly - - # create the directory if needed & save each individual model - os.makedirs(dirname, exist_ok=True) - paths = [] - for i, (model, used) in enumerate(zip(self.models, self.models_used)): - if used or not save_only_used_models: - path = os.path.abspath(os.path.join(dirname, str(i))) - paths.append(path) - model.save(path) - else: - paths.append(None) - - # Add model paths to the config dict, and save it - config_dict["model_paths"] = [(type(m).__name__, p) for m, p in zip(self.models, paths)] - with open(os.path.join(dirname, self.config_class.filename), "w") as f: - json.dump(config_dict, f, indent=2, sort_keys=True) - - # Save the remaining ensemble state - filename = os.path.join(dirname, self.filename) - self._save_state( - state_dict=state_dict, filename=filename, save_only_used_models=save_only_used_models, **save_config - ) - - @classmethod - def load(cls, dirname: str, **kwargs): - # Read the config dict from json - config_path = os.path.join(dirname, cls.config_class.filename) - with open(config_path, "r") as f: - config_dict = json.load(f) - - # Load all the models from the config dict - model_paths = config_dict.pop("model_paths") - models = [ModelFactory.load(name=name, model_path=path) for name, path in model_paths] - - # Load the state dict - with open(os.path.join(dirname, cls.filename), "rb") as f: - state_dict = dill.load(f) + super().save(dirname=dirname, save_only_used_models=save_only_used_models, **save_config) - return cls._from_config_state_dicts(config_dict, state_dict, models, **kwargs) - - @classmethod - def _from_config_state_dicts(cls, config_dict, state_dict, models, **kwargs): - # Use the config to initialize the model & then load it - config, model_kwargs = cls.config_class.from_dict(config_dict, return_unused_kwargs=True, **kwargs) - ensemble = cls(config=config, models=models) - ensemble._load_state(state_dict, **model_kwargs) - - return ensemble - - def to_bytes(self, save_only_used_models=False, **save_config): + def _save_state( + self, state_dict: Dict[str, Any], filename: str = None, save_only_used_models=False, **save_config + ) -> Dict[str, Any]: """ - Converts the entire ensemble to a single byte object. + Saves the model's state to the the specified file, or just modifies the state_dict as needed. - :param save_only_used_models: whether to save only the models that are - actually used by the ensemble. - :param save_config: additional save config arguments - :return: bytes object representing the model. + :param state_dict: The state dict to save. + :param filename: The name of the file to save the model to. + :param save_only_used_models: whether to save only the models that are actually used by the ensemble. + :param save_config: additional configurations (if needed) + :return: The state dict to save. """ - state_dict = self.__getstate__() - state_dict.pop("models") - config_dict = self.config.to_dict() - config_dict.pop("model_configs") - state_dict = self._save_state(state_dict, **save_config) - class_name = self.__class__.__name__ - - model_tuples = [ - model._to_serializable_comps() - for model, used in zip(self.models, self.models_used) - if used or not save_only_used_models - ] - - return dill.dumps((class_name, config_dict, state_dict, model_tuples)) - - @classmethod - def from_bytes(cls, obj, **kwargs): + state_dict.pop("config", None) # don't save the model's config in binary + if self.models is not None: + model_states = [] + for model, model_state, model_used in zip(self.models, state_dict["models"], self.models_used): + if save_only_used_models and not model_used: + model_states.append(None) + else: + model_states.append( + model._save_state(model_state, None, save_only_used_models=save_only_used_models, **save_config) + ) + state_dict["models"] = model_states + return super()._save_state(state_dict, filename, **save_config) + + def to_bytes(self, save_only_used_models=False, **save_config): """ - Creates a fully specified model from a byte object + Converts the entire model state and configuration to a single byte object. - :param obj: byte object to convert into a model - :return: `EnsembleBase` object loaded from ``obj`` + :param save_only_used_models: whether to save only the models that are actually used by the ensemble. + :param save_config: additional configurations (if needed) """ - name, config_dict, state_dict, model_tuples = dill.loads(obj) - models = [ - ModelFactory.get_model_class(model_tuple[0])._from_config_state_dicts(*model_tuple[1:]) - for model_tuple in model_tuples - ] - return cls._from_config_state_dicts(config_dict, state_dict, models, **kwargs) + return super().to_bytes(save_only_used_models=save_only_used_models, **save_config) diff --git a/merlion/models/ensemble/forecast.py b/merlion/models/ensemble/forecast.py index 8a98bd1f0..3633bcb7e 100644 --- a/merlion/models/ensemble/forecast.py +++ b/merlion/models/ensemble/forecast.py @@ -27,7 +27,8 @@ class ForecasterEnsembleConfig(ForecasterConfig, EnsembleConfig): _default_combiner = Mean(abs_score=False) - def __init__(self, max_forecast_steps=None, **kwargs): + def __init__(self, max_forecast_steps=None, verbose=False, **kwargs): + self.verbose = verbose super().__init__(max_forecast_steps=max_forecast_steps, **kwargs) @@ -42,7 +43,7 @@ class ForecasterEnsemble(EnsembleBase, ForecasterBase): _default_train_config = EnsembleTrainConfig(valid_frac=0.2) def __init__(self, config: ForecasterEnsembleConfig = None, models: List[ForecasterBase] = None): - super().__init__(config, models) + super().__init__(config=config, models=models) for model in self.models: assert isinstance(model, ForecasterBase), ( f"Expected all models in {type(self).__name__} to be anomaly " @@ -84,9 +85,9 @@ def train( # Train individual models on the training data preds, errs = [], [] for i, (model, cfg) in enumerate(zip(self.models, per_model_train_configs)): - logger.info(f"Training model {i+1}/{len(self.models)}...") + logger.info(f"Training model {i+1}/{len(self.models)} ({type(model).__name__})...") try: - pred, err = model.train(train, cfg) + pred, err = model.train(train, train_config=cfg) preds.append(pred) errs.append(err) except TypeError as e: @@ -117,7 +118,7 @@ def train( t0, tf = valid.t0, valid.tf valid_windows = [] preds = [[] for _ in self.models] - pbar = tqdm(total=int(tf - t0), desc="Validation") + pbar = tqdm(total=int(tf - t0), desc="Validation", disable=not self.config.verbose) while t0 < tf: next_tf = to_pd_datetime(prev.tf) + h dt = int((next_tf - to_pd_datetime(prev.tf)).total_seconds()) @@ -154,12 +155,15 @@ def train( # Re-train on the full data if we used a validation split full_preds, full_errs = [], [] for i, (model, cfg) in enumerate(zip(self.models, per_model_train_configs)): - logger.info(f"Re-training model {i+1}/{len(self.models)} on full data...") + logger.info(f"Re-training model {i+1}/{len(self.models)} ({type(model).__name__}) on full data...") model.reset() - pred, err = model.train(full_train, cfg) + pred, err = model.train(full_train, train_config=cfg) full_preds.append(pred) full_errs.append(err) err = None if any(e is None for e in full_errs) else self.combiner(full_errs, None) + if not all(self.models_used): + used = [f"{i+1} ({type(m).__name__})" for i, (m, u) in enumerate(zip(self.models, self.models_used)) if u] + logger.info(f"Models used (of {len(self.models)}): {', '.join(used)}") return self.combiner(full_preds, None), err def forecast( diff --git a/merlion/models/factory.py b/merlion/models/factory.py index 7218dfc30..4aff3c50e 100644 --- a/merlion/models/factory.py +++ b/merlion/models/factory.py @@ -8,7 +8,8 @@ Contains the `ModelFactory`. """ import inspect -from typing import Type +from typing import Dict, Tuple, Type, Union + import dill from merlion.models.base import ModelBase from merlion.utils import dynamic_import @@ -57,7 +58,9 @@ ForecasterEnsemble="merlion.models.ensemble.forecast:ForecasterEnsemble", MoE_ForecasterEnsemble="merlion.models.ensemble.MoE_forecast:MoE_ForecasterEnsemble", # Layers - SeasonalityLayer="merlion.models.automl.seasonality_mixin:SeasonalityLayer", + SeasonalityLayer="merlion.models.automl.seasonality:SeasonalityLayer", + AutoETS="merlion.models.automl.autoets:AutoETS", + AutoProphet="merlion.models.automl.autoprophet:AutoProphet", AutoSarima="merlion.models.automl.autosarima:AutoSarima", ) @@ -68,16 +71,22 @@ def get_model_class(cls, name: str) -> Type[ModelBase]: return dynamic_import(name, import_alias) @classmethod - def create(cls, name, **kwargs) -> ModelBase: + def create(cls, name, return_unused_kwargs=False, **kwargs) -> Union[ModelBase, Tuple[ModelBase, Dict]]: model_class = cls.get_model_class(name) - signature = inspect.signature(model_class.__init__) - if "config" in signature.parameters: - config, kwargs = model_class.config_class.from_dict(kwargs, return_unused_kwargs=True) - init_kwargs = {k: v for k, v in kwargs.items() if k in signature.parameters} - model = model_class(config, **init_kwargs) - model._load_state({k: v for k, v in kwargs.items() if k not in init_kwargs}) - else: - model = model_class(**kwargs) + config, kwargs = model_class.config_class.from_dict(kwargs, return_unused_kwargs=True) + + # initialize the model + signature = inspect.signature(model_class) + init_kwargs = {k: v for k, v in kwargs.items() if k in signature.parameters} + kwargs = {k: v for k, v in kwargs.items() if k not in init_kwargs} + model = model_class(config=config, **init_kwargs) + + # set model state with remaining kwargs, and return any unused kwargs if desired + if return_unused_kwargs: + state = {k: v for k, v in kwargs.items() if hasattr(model, k)} + model._load_state(state) + return model, {k: v for k, v in kwargs.items() if k not in state} + model._load_state(kwargs) return model @classmethod diff --git a/merlion/models/forecast/arima.py b/merlion/models/forecast/arima.py index 0dc5574d8..b41e8c73e 100644 --- a/merlion/models/forecast/arima.py +++ b/merlion/models/forecast/arima.py @@ -24,16 +24,11 @@ class ArimaConfig(SarimaConfig): _default_transform = TemporalResample(granularity=None, trainable_granularity=True) - def __init__(self, max_forecast_steps=None, target_seq_index=None, order=(4, 1, 2), **kwargs): - if "seasonal_order" in kwargs: - raise ValueError("cannot specify seasonal_order for ARIMA") - super().__init__( - max_forecast_steps=max_forecast_steps, - target_seq_index=target_seq_index, - order=order, - seasonal_order=(0, 0, 0, 0), - **kwargs - ) + def __init__(self, order=(4, 1, 2), seasonal_order=(0, 0, 0, 0), **kwargs): + """ + :param seasonal_order: (0, 0, 0, 0) because ARIMA has no seasonal order. + """ + super().__init__(order=order, seasonal_order=seasonal_order, **kwargs) @property def seasonal_order(self) -> Tuple[int, int, int, int]: diff --git a/merlion/models/forecast/baggingtrees.py b/merlion/models/forecast/baggingtrees.py index 229b1b87c..93373aa63 100644 --- a/merlion/models/forecast/baggingtrees.py +++ b/merlion/models/forecast/baggingtrees.py @@ -143,10 +143,11 @@ def train(self, train_data: TimeSeries, train_config=None): f"'training_mode = autogression'." ) self.config.max_forecast_steps = max_forecast_steps - logger.warning( - f"For multivariate dataset, reset prediction_stride = max_forecast_steps = {self.max_forecast_steps} " - ) - self.config.prediction_stride = self.max_forecast_steps + if self.prediction_stride != self.max_forecast_steps: + logger.warning( + f"For multivariate dataset, reset prediction_stride = max_forecast_steps = {self.max_forecast_steps} " + ) + self.config.prediction_stride = self.max_forecast_steps # process train data (inputs_train, labels_train, labels_train_ts) = seq_ar_common.process_rolling_train_data( train_data, self.target_seq_index, self.maxlags, self.prediction_stride, self.sampling_mode diff --git a/merlion/models/forecast/base.py b/merlion/models/forecast/base.py index bad028852..8f6f5e4cc 100644 --- a/merlion/models/forecast/base.py +++ b/merlion/models/forecast/base.py @@ -8,9 +8,8 @@ Base class for forecasting models. """ from abc import abstractmethod -import copy import logging -from typing import Any, Dict, List, Optional, Tuple, Union +from typing import List, Optional, Tuple, Union import pandas as pd @@ -26,11 +25,13 @@ class ForecasterConfig(Config): Config object used to define a forecaster model. """ - def __init__(self, max_forecast_steps: Union[int, None], target_seq_index: int = None, **kwargs): + max_forecast_steps: Optional[int] = None + target_seq_index: Optional[int] = None + + def __init__(self, max_forecast_steps: int = None, target_seq_index: int = None, **kwargs): """ :param max_forecast_steps: Max # of steps we would like to forecast for. - Required for some models which pre-compute a forecast, like ARIMA, - SARIMA, and LSTM. + Required for some models like `MSES` and `LGBMForecaster`. :param target_seq_index: The index of the univariate (amongst all univariates in a general multivariate time series) whose value we would like to forecast. @@ -38,15 +39,6 @@ def __init__(self, max_forecast_steps: Union[int, None], target_seq_index: int = super().__init__(**kwargs) self.max_forecast_steps = max_forecast_steps self.target_seq_index = target_seq_index - self.dim = None - - @classmethod - def from_dict(cls, config_dict: Dict[str, Any], return_unused_kwargs=False, **kwargs): - config_dict = copy.copy(config_dict) - dim = config_dict.pop("dim", None) - if "dim" not in kwargs: - kwargs["dim"] = dim - return super().from_dict(config_dict, return_unused_kwargs, **kwargs) class ForecasterBase(ModelBase): @@ -65,16 +57,9 @@ class ForecasterBase(ModelBase): """ config_class = ForecasterConfig - timedelta: Optional[float] - """ - The expected number of seconds between observations in an input time series. - should be set in `ForecasterBase.train` if the model assumes a fixed - timedelta. + target_name = None """ - last_train_time: Optional[float] - """ - The last unix timestamp of the training data. Should be set in - `ForecasterBase.train`. + The name of the target univariate to forecast. """ def __init__(self, config: ForecasterConfig): @@ -93,10 +78,6 @@ def target_seq_index(self) -> int: """ return self.config.target_seq_index - @property - def dim(self): - return self.config.dim - def resample_time_stamps(self, time_stamps: Union[int, List[int]], time_series_prev: TimeSeries = None): assert self.timedelta is not None and self.last_train_time is not None, ( "train() must be called before you can call forecast(). " @@ -144,7 +125,6 @@ def resample_time_stamps(self, time_stamps: Union[int, List[int]], time_series_p def train_pre_process( self, train_data: TimeSeries, require_even_sampling: bool, require_univariate: bool ) -> TimeSeries: - self.config.dim = train_data.dim train_data = super().train_pre_process(train_data, require_even_sampling, require_univariate) if self.dim == 1: self.config.target_seq_index = 0 diff --git a/merlion/models/forecast/boostingtrees.py b/merlion/models/forecast/boostingtrees.py index cfc896cd9..2eaaea50c 100644 --- a/merlion/models/forecast/boostingtrees.py +++ b/merlion/models/forecast/boostingtrees.py @@ -65,8 +65,7 @@ def __init__( :param n_estimators: number of base estimators for the tree ensemble :param random_state: random seed for boosting :param max_depth: max depth of base estimators - :param n_jobs: num of threading, -1 or 0 indicates device default, - positive int indicates num of threads + :param n_jobs: num of threading, -1 or 0 indicates device default, positive int indicates num of threads """ super().__init__(max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index, **kwargs) self.maxlags = maxlags @@ -146,10 +145,11 @@ def train(self, train_data: TimeSeries, train_config=None): f"'training_mode = autogression'." ) self.config.max_forecast_steps = max_forecast_steps - logger.warning( - f"For multivariate dataset, reset prediction_stride = max_forecast_steps = {self.max_forecast_steps} " - ) - self.config.prediction_stride = self.max_forecast_steps + if self.prediction_stride != self.max_forecast_steps: + logger.warning( + f"For multivariate dataset, reset prediction_stride = max_forecast_steps = {self.max_forecast_steps} " + ) + self.config.prediction_stride = self.max_forecast_steps # process train data (inputs_train, labels_train, labels_train_ts) = seq_ar_common.process_rolling_train_data( train_data, self.target_seq_index, self.maxlags, self.prediction_stride, self.sampling_mode diff --git a/merlion/models/forecast/ets.py b/merlion/models/forecast/ets.py index d58f4bcce..9e7d101f6 100644 --- a/merlion/models/forecast/ets.py +++ b/merlion/models/forecast/ets.py @@ -17,9 +17,10 @@ from scipy.stats import norm from statsmodels.tsa.exponential_smoothing.ets import ETSModel +from merlion.models.automl.seasonality import SeasonalityModel from merlion.models.forecast.base import ForecasterBase, ForecasterConfig from merlion.transform.resample import TemporalResample -from merlion.utils import autosarima_utils, TimeSeries, UnivariateTimeSeries +from merlion.utils import TimeSeries, UnivariateTimeSeries logger = logging.getLogger(__name__) @@ -44,7 +45,7 @@ def __init__( trend="add", damped_trend=True, seasonal="add", - seasonal_periods="auto", + seasonal_periods=None, **kwargs, ): """ @@ -56,9 +57,7 @@ def __init__( :param trend: The trend component. "add", "mul" or None. :param damped_trend: Whether or not an included trend component is damped. :param seasonal: The seasonal component. "add", "mul" or None. - :param seasonal_periods: The length of the seasonality cycle. 'auto' - indicates automatically select the seasonality cycle. If no - seasonality exists, change ``seasonal`` to ``None``. + :param seasonal_periods: The length of the seasonality cycle. ``None`` by default. """ super().__init__(max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index, **kwargs) self.error = error @@ -68,7 +67,7 @@ def __init__( self.seasonal_periods = seasonal_periods -class ETS(ForecasterBase): +class ETS(SeasonalityModel, ForecasterBase): """ Implementation of the classic local statistical model ETS (Error, Trend, Seasonal) for forecasting. """ @@ -102,6 +101,13 @@ def seasonal(self): def seasonal_periods(self): return self.config.seasonal_periods + def set_seasonality(self, theta, train_data: UnivariateTimeSeries): + if theta > 1: + self.config.seasonal_periods = int(theta) + else: + self.config.seasonal = None + self.config.seasonal_periods = None + def train(self, train_data: TimeSeries, train_config=None): # Train the transform & transform the training data train_data = self.train_pre_process(train_data, require_even_sampling=True, require_univariate=False) @@ -111,18 +117,6 @@ def train(self, train_data: TimeSeries, train_config=None): train_data = train_data.univariates[name].to_pd() times = train_data.index - if self.seasonal_periods == "auto": - periods = autosarima_utils.multiperiodicity_detection(train_data.to_numpy()) - if len(periods) > 0: - min_periodicty = periods[0] - else: - min_periodicty = 0 - if min_periodicty > 1: - logger.info(f"Detect seasonality {str(min_periodicty)}") - self.config.seasonal_periods = min_periodicty.item() - else: - self.config.seasonal = None - self.config.seasonal_periods = None with warnings.catch_warnings(): warnings.simplefilter("ignore") self.model = ETSModel( @@ -176,10 +170,10 @@ def forecast( if time_series_prev is None: forecast_result = self.model.get_prediction( - start=self._n_train, end=self._n_train + len(time_stamps) - 1, method="simulated" + start=self._n_train, end=self._n_train + len(time_stamps) - 1, method="exact" ) forecast = forecast_result.predicted_mean - err = forecast_result._results.simulation_results.std(axis=1) + err = np.sqrt(forecast_result.var_pred_mean) if any(np.isnan(forecast)): logger.warning( "Trained ETS model is producing NaN forecast. Use the last " diff --git a/merlion/models/forecast/lstm.py b/merlion/models/forecast/lstm.py index acace3f5c..f6deb59df 100644 --- a/merlion/models/forecast/lstm.py +++ b/merlion/models/forecast/lstm.py @@ -44,19 +44,15 @@ class LSTMConfig(ForecasterConfig): ] ) - def __init__(self, max_forecast_steps: int, target_seq_index: int = None, nhid=1024, model_strides=(1,), **kwargs): + def __init__(self, max_forecast_steps: int, nhid=1024, model_strides=(1,), **kwargs): """ - :param max_forecast_steps: Max # of steps we would like to forecast for. - :param target_seq_index: The index of the univariate (amongst all - univariates in a general multivariate time series) whose value we - would like to forecast. :param nhid: hidden dimension of LSTM :param model_strides: tuple indicating the stride(s) at which we would like to subsample the input data before giving it to the model. """ self.model_strides = list(model_strides) self.nhid = nhid - super().__init__(max_forecast_steps, target_seq_index, **kwargs) + super().__init__(max_forecast_steps=max_forecast_steps, **kwargs) class LSTMTrainConfig(object): @@ -252,7 +248,6 @@ def __init__(self, config: LSTMConfig): if torch.cuda.is_available(): self.model.cuda() self.optimizer = None - self.timedelta = None self.seq_len = None self._forecast = [0.0 for _ in range(self.max_forecast_steps)] diff --git a/merlion/models/forecast/prophet.py b/merlion/models/forecast/prophet.py index 5ed84ef9e..a9df3c55a 100644 --- a/merlion/models/forecast/prophet.py +++ b/merlion/models/forecast/prophet.py @@ -8,18 +8,53 @@ Wrapper around Facebook's popular Prophet model for time series forecasting. """ import logging -from typing import List, Tuple, Union +import os +from typing import Iterable, List, Tuple, Union import prophet import numpy as np import pandas as pd +from merlion.models.automl.seasonality import SeasonalityModel from merlion.models.forecast.base import ForecasterBase, ForecasterConfig -from merlion.utils import TimeSeries, UnivariateTimeSeries, to_pd_datetime, autosarima_utils +from merlion.utils import TimeSeries, UnivariateTimeSeries, to_pd_datetime logger = logging.getLogger(__name__) +class _suppress_stdout_stderr(object): + """ + A context manager for doing a "deep suppression" of stdout and stderr in + Python, i.e. will suppress all print, even if the print originates in a + compiled C/Fortran sub-function. + + This will not suppress raised exceptions, since exceptions are printed + to stderr just before a script exits, and after the context manager has + exited (at least, I think that is why it lets exceptions through). + + Source: https://github.com/facebook/prophet/issues/223#issuecomment-326455744 + """ + + def __init__(self): + # Open a pair of null files + self.null_fds = [os.open(os.devnull, os.O_RDWR) for x in range(2)] + # Save the actual stdout (1) and stderr (2) file descriptors. + self.save_fds = [os.dup(1), os.dup(2)] + + def __enter__(self): + # Assign the null pointers to stdout and stderr. + os.dup2(self.null_fds[0], 1) + os.dup2(self.null_fds[1], 2) + + def __exit__(self, *_): + # Re-assign the real stdout/stderr back to (1) and (2) + os.dup2(self.save_fds[0], 1) + os.dup2(self.save_fds[1], 2) + # Close the null files + for fd in self.null_fds + self.save_fds: + os.close(fd) + + class ProphetConfig(ForecasterConfig): """ Configuration class for Facebook's `Prophet` model, as described by @@ -33,7 +68,6 @@ def __init__( yearly_seasonality: Union[bool, int] = "auto", weekly_seasonality: Union[bool, int] = "auto", daily_seasonality: Union[bool, int] = "auto", - add_seasonality="auto", seasonality_mode="additive", holidays=None, uncertainty_samples: int = 100, @@ -56,8 +90,6 @@ def __init__( By default, it is activated if there are >= 2 days of history, but deactivated otherwise. If int, this is the number of Fourier series components used to model the seasonality (default = 4). - :param add_seasonality: 'auto' indicates automatically adding extra - seasonality by detection methods (default = None). :param seasonality_mode: 'additive' (default) or 'multiplicative'. :param holidays: pd.DataFrame with columns holiday (string) and ds (date type) and optionally columns lower_window and upper_window which specify a @@ -72,13 +104,12 @@ def __init__( self.yearly_seasonality = yearly_seasonality self.weekly_seasonality = weekly_seasonality self.daily_seasonality = daily_seasonality - self.add_seasonality = add_seasonality self.seasonality_mode = seasonality_mode self.uncertainty_samples = uncertainty_samples self.holidays = holidays -class Prophet(ForecasterBase): +class Prophet(SeasonalityModel, ForecasterBase): """ Facebook's model for time series forecasting. See docs for `ProphetConfig` and `Taylor & Letham, 2017 `__ for more details. @@ -138,26 +169,22 @@ def holidays(self): def uncertainty_samples(self): return self.config.uncertainty_samples + def set_seasonality(self, theta, train_data: UnivariateTimeSeries): + theta = [theta] if not isinstance(theta, Iterable) else theta + dt = train_data.index[1] - train_data.index[0] + for p in theta: + if p > 1: + period = p * dt.total_seconds() / 86400 + logger.info(f"Add seasonality {str(p)} ({p * dt})") + self.model.add_seasonality(name=f"extra_season_{p}", period=period, fourier_order=p) + def train(self, train_data: TimeSeries, train_config=None): train_data = self.train_pre_process(train_data, require_even_sampling=False, require_univariate=False) series = train_data.univariates[self.target_name] df = pd.DataFrame({"ds": series.index, "y": series.np_values}) - if self.add_seasonality == "auto": - periods = autosarima_utils.multiperiodicity_detection(series.np_values) - if len(periods) > 0: - max_periodicity = periods[-1] - else: - max_periodicity = 0 - if max_periodicity > 1: - logger.info(f"Add seasonality {str(max_periodicity)}") - if hasattr(self.timedelta, "total_seconds"): - period = max_periodicity * self.timedelta.total_seconds() / 86400 - else: - period = max_periodicity * (series.ds[1] - series.ds[0]).total_seconds() / 86400 - self.model.add_seasonality(name="extra_season", period=period, fourier_order=max_periodicity) - - self.model.fit(df) + with _suppress_stdout_stderr(): + self.model.fit(df) # Get & return prediction & errors for train data self.model.uncertainty_samples = 0 diff --git a/merlion/models/forecast/sarima.py b/merlion/models/forecast/sarima.py index b36785cb5..beb62c22a 100644 --- a/merlion/models/forecast/sarima.py +++ b/merlion/models/forecast/sarima.py @@ -17,7 +17,7 @@ from scipy.stats import norm from statsmodels.tsa.arima.model import ARIMA as sm_Sarima -from merlion.models.automl.seasonality_mixin import SeasonalityModel +from merlion.models.automl.seasonality import SeasonalityModel from merlion.models.forecast.base import ForecasterBase, ForecasterConfig from merlion.transform.resample import TemporalResample from merlion.utils.time_series import TimeSeries, UnivariateTimeSeries @@ -32,14 +32,8 @@ class SarimaConfig(ForecasterConfig): _default_transform = TemporalResample(granularity=None) - def __init__( - self, max_forecast_steps=None, target_seq_index=None, order=(4, 1, 2), seasonal_order=(2, 0, 1, 24), **kwargs - ): + def __init__(self, order=(4, 1, 2), seasonal_order=(2, 0, 1, 24), **kwargs): """ - :param max_forecast_steps: Number of steps we would like to forecast for. - :param target_seq_index: The index of the univariate (amongst all - univariates in a general multivariate time series) whose value we - would like to forecast. :param order: Order is (p, d, q) for an ARIMA(p, d, q) process. d must be an integer indicating the integration order of the process, while p and q must be integers indicating the AR and MA orders (so that @@ -48,7 +42,7 @@ def __init__( process, where s is the length of the seasonality cycle (e.g. s=24 for 24 hours on hourly granularity). P, D, Q are as for ARIMA. """ - super().__init__(max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index, **kwargs) + super().__init__(**kwargs) self.order = order self.seasonal_order = seasonal_order @@ -220,72 +214,7 @@ def forecast( ) return forecast, err - def set_seasonality(self, theta, train_data: np.array): - theta = self._correct_theta(theta, train_data) - self.config.seasonal_order = tuple(list(self.seasonal_order)[:-1] + [theta]) - - def _correct_theta(self, theta, train_data: np.array): - y = train_data - - order = list(self.config.order) - seasonal_order = list(self.config.seasonal_order) - max_d = 2 - max_D = 1 - stationary = False - seasonal_test = "seas" - test = "kpss" - - # pqPQ is an indicator about whether need to automatically select - # AR, MA, seasonal AR and seasonal MA parameters - d = D = pqPQ = None - if order[1] != "auto": - d = order[1] - if seasonal_order[1] != "auto": - D = seasonal_order[1] - if order[0] != "auto" and order[2] != "auto" and seasonal_order[0] != "auto" and seasonal_order[2] != "auto": - pqPQ = True - - if any(np.isnan(y)): - raise ValueError("there exists missing values in observed time series") - - # check m - if theta < 1: - theta = 1 - else: - theta = int(theta) - - # input time-series is completely constant - if np.max(y) == np.min(y): - return iter([0]) - - xx = y.copy() - if stationary: - d = D = 0 - if theta == 1: - D = 0 - - # set the seasonal differencing order with statistical test - elif D is None: - D = autosarima_utils.nsdiffs(xx, m=theta, max_D=max_D, test=seasonal_test) - if D > 0: - dx = autosarima_utils.diff(xx, differences=D, lag=theta) - if dx.shape[0] == 0: - D = D - 1 - if D > 0: - dx = autosarima_utils.diff(xx, differences=D, lag=theta) - else: - dx = xx - logger.info(f"Seasonal difference order is {str(D)}") - - # set the differencing order by estimating the number of orders - # it would take in order to make the time series stationary - if d is None: - d = autosarima_utils.ndiffs(dx, alpha=0.05, max_d=max_d, test=test) - if d > 0: - dx = autosarima_utils.diff(dx, differences=d, lag=1) - logger.info(f"Difference order is {str(d)}") - - if pqPQ is not None or np.max(dx) == np.min(dx): - return theta if theta != 1 else 0 - - return theta + def set_seasonality(self, theta, train_data: UnivariateTimeSeries): + # Make sure seasonality is a positive int, and set it to 1 if the train data is constant + theta = 1 if np.max(train_data) == np.min(train_data) else max(1, int(theta)) + self.config.seasonal_order = self.seasonal_order[:-1] + (theta,) diff --git a/merlion/models/forecast/smoother.py b/merlion/models/forecast/smoother.py index cabeab356..b11b1ffbf 100644 --- a/merlion/models/forecast/smoother.py +++ b/merlion/models/forecast/smoother.py @@ -36,7 +36,6 @@ class MSESConfig(ForecasterConfig): def __init__( self, max_forecast_steps: int, - target_seq_index: int = None, max_backstep: int = None, recency_weight: float = 0.5, accel_weight: float = 1.0, @@ -61,10 +60,6 @@ def __init__( \text{if} & \space\space z_b = (b+h)^\phi \cdot \text{EMA}_w(l_{b+h,t}) \cdot \text{RWSE}_w(l_{b+h,t})\\ \end{align*} - :param max_forecast_steps: Max number of steps to forecast ahead. - :param target_seq_index: The index of the univariate (amongst all - univariates in a general multivariate time series) whose value we - would like to forecast. :param max_backstep: Max backstep to use in forecasting. If we train with x(0),...,x(t), Then, the b-th model MSES uses will forecast x(t+h) by anchoring at x(t-b) and predicting xhat(t+h) = x(t-b) + delta_hat(b+h). @@ -85,7 +80,7 @@ def __init__( errors of the estimated velocities over the models; inflation=1 is equivalent to using the softmax function. """ - super().__init__(max_forecast_steps=max_forecast_steps, target_seq_index=target_seq_index, **kwargs) + super().__init__(max_forecast_steps=max_forecast_steps, **kwargs) assert 0.0 <= rho <= 1.0 assert 1.0 <= phi self.max_backstep = max_forecast_steps if max_backstep is None else max_backstep diff --git a/merlion/models/layers.py b/merlion/models/layers.py new file mode 100644 index 000000000..c22dab757 --- /dev/null +++ b/merlion/models/layers.py @@ -0,0 +1,285 @@ +# +# Copyright (c) 2021 salesforce.com, inc. +# All rights reserved. +# SPDX-License-Identifier: BSD-3-Clause +# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause +# +""" +Base class for layered models. These are models which act as a wrapper around another model, often with additional +functionality. This is the basis for `default models `_ and +`AutoML models `_. +""" +from copy import deepcopy +from typing import Any, Dict, Union + +from merlion.models.base import Config, ModelBase +from merlion.models.factory import ModelFactory +from merlion.models.anomaly.base import DetectorBase, DetectorConfig +from merlion.models.forecast.base import ForecasterBase, ForecasterConfig +from merlion.models.anomaly.forecast_based.base import ForecastingDetectorBase +from merlion.utils import TimeSeries +from merlion.utils.misc import AutodocABCMeta + + +class LayeredModelConfig(Config): + """ + Config object for a `LayeredModel`. See `LayeredModel` documentation for more details. + """ + + def __init__(self, model: Union[ModelBase, Dict], model_kwargs=None, **kwargs): + """ + :param model: The model being wrapped, or a dict representing it. + :param model_kwargs: Keyword arguments used specifically to initialize the underlying model. Only used if + ``model`` is a dict. Will override keys in the ``model`` dict if specified. + :param kwargs: Any other keyword arguments (e.g. for initializing a base class). If ``model`` is a dict, + we will also try to pass these arguments when creating the actual underlying model. However, they will + not override arguments in either the ``model`` dict or ``model_kwargs`` dict. + """ + # Model-specific kwargs override kwargs when creating the model. + model_kwargs = {} if model_kwargs is None else model_kwargs + if isinstance(model, dict): + model.update({k: v for k, v in kwargs.items() if k not in model and k not in model_kwargs}) + model, extra_kwargs = ModelFactory.create(**{**model, **model_kwargs, "return_unused_kwargs": True}) + kwargs.update(extra_kwargs) + self.model = model + super().__init__(**kwargs) + + # If no model was created, reserve unused kwargs to try initializing the model with + if model is None: + extra_kwargs = {k: v for k, v in kwargs.items() if k not in self.to_dict()} + self.model_kwargs = {**extra_kwargs, **model_kwargs} + + @property + def model(self): + """ + The underlying model which this layer wraps. + """ + return self._model + + @model.setter + def model(self, model): + self._model = model + self.model_kwargs = {} + + @property + def base_model(self): + """ + The base model at the heart of the full layered model. + """ + model = self.model + while isinstance(model, LayeredModel): + model = model.model + return model + + def to_dict(self, _skipped_keys=None): + _skipped_keys = _skipped_keys if _skipped_keys is not None else set() + config_dict = super().to_dict(_skipped_keys.union({"model"})) + if not self.model_kwargs and "model_kwargs" in config_dict: + config_dict["model_kwargs"] = None + if "model" not in _skipped_keys: + if self.model is None: + config_dict["model"] = None + else: + config_dict["model"] = dict(name=type(self.model).__name__, **self.model.config.to_dict()) + return config_dict + + def __copy__(self): + config_dict = super().to_dict(_skipped_keys={"model"}) + return self.__class__(model=deepcopy(self.model), **config_dict) + + def __getattr__(self, item): + if item in ["model", "_model", "base_model"]: + return super().__getattribute__(item) + base_model = self.base_model + is_detector_attr = isinstance(base_model, DetectorBase) and hasattr(DetectorConfig, item) + is_forecaster_attr = isinstance(base_model, ForecasterBase) and hasattr(ForecasterConfig, item) + if is_detector_attr or is_forecaster_attr: + return getattr(base_model.config, item) + return self.__getattribute__(item) + + def __setattr__(self, key, value): + if hasattr(self, "_model"): + base_model = self.base_model + is_detector_attr = isinstance(base_model, DetectorBase) and hasattr(DetectorConfig, key) + is_forecaster_attr = isinstance(base_model, ForecasterBase) and hasattr(ForecasterConfig, key) + is_layered_attr = hasattr(LayeredModelConfig, key) + if not is_layered_attr and (is_detector_attr or is_forecaster_attr): + return setattr(self.model.config, key, value) + return super().__setattr__(key, value) + + def get_unused_kwargs(self, **kwargs): + config = self + valid_keys = {"model"}.union(config.to_dict(_skipped_keys={"model"}).keys()) + while isinstance(config, LayeredModelConfig) and config.model is not None: + config = config.model.config + valid_keys = valid_keys.union(config.to_dict(_skipped_keys={"model"}).keys()) + return {k: v for k, v in kwargs.items() if k not in valid_keys} + + +class LayeredModel(ModelBase, metaclass=AutodocABCMeta): + """ + Abstract class implementing a model which wraps around another internal model. + + The actual underlying model is stored in ``model.config.model``, and ``model.model`` is a property which references + this. This is to allow the model to retain the initializer ``LayeredModel(config)``, and to ensure that various + attributes do not become de-synchronized (e.g. if we were to store ``config.model_config`` and ``model.model`` + separately). + + We define the *base model* as the non-layered model at the base of the overall model hierarchy. + + The layered model is allowed to access any callable attribute of the base model, + e.g. ``model.set_seasonality(...)`` resolves to``model.base_model.set_seasonality(...)`` for a `SeasonalityModel`. + If the base model is a forecaster, the layered model will automatically inherit from `ForecasterBase`; similarly + for `DetectorBase` or `ForecastingDetectorBase`. The abstract methods (``forecast`` and ``get_anomaly_score``) + are overridden to call the underlying model. + + If the base model is a forecaster, the top-level config ``model.config`` does not duplicate attributes of the + underlying forecaster config (e.g. ``max_forecast_steps`` or ``target_seq_index``). Instead, + ``model.config.max_forecast_steps`` will resolve to ``model.config.base_model.max_forecast_steps``. + As a result, you will only need to specify this parameter once. The same holds true for `DetectorConfig` attributes + (e.g. ``threshold`` or ``calibrator``) when the base model is an anomaly detector. + + .. note:: + + For the time being, every layer of the model is allowed to have its own ``transform``. + """ + + config_class = LayeredModelConfig + require_even_sampling = False + require_univariate = False + + def __new__(cls, config: LayeredModelConfig = None, model: ModelBase = None, **kwargs): + # Dynamically inherit from the appropriate kind of base model. + # However, this creates a new class that isn't registered anywhere with pickle/dill. This causes + # serialization problems, especially when using models with multiprocessing. So we maintain this + # class (cls) as a class attribute _original_cls of the new, dynamically created class. This is + # used by the __reduce__ method when pickling a LayeredModel. + original_cls = cls + config = cls._resolve_args(config=config, model=model, **kwargs) + if isinstance(config.model, ForecastingDetectorBase): + cls = cls.__class__(cls.__name__, (cls, LayeredForecastingDetector), {}) + setattr(cls, "_original_cls", original_cls) + elif isinstance(config.model, ForecasterBase): + cls = cls.__class__(cls.__name__, (cls, LayeredForecaster), {}) + setattr(cls, "_original_cls", original_cls) + elif isinstance(config.model, DetectorBase): + cls = cls.__class__(cls.__name__, (cls, LayeredDetector), {}) + setattr(cls, "_original_cls", original_cls) + return super().__new__(cls) + + def __init__(self, config: LayeredModelConfig = None, model: ModelBase = None, **kwargs): + super().__init__(config=self._resolve_args(config=config, model=model, **kwargs)) + + @classmethod + def _resolve_args(cls, config: LayeredModelConfig, model: ModelBase, **kwargs): + if config is None and model is None: + raise RuntimeError( + f"Expected at least one of `config` or `model` when creating {cls.__name__}. Received neither." + ) + elif config is not None and model is not None: + if config.model is None: + config.model = model + else: + raise RuntimeError( + f"Expected at most one of `config.model` or `model` when creating {cls.__name__}. Received both." + ) + elif config is None: + config = cls.config_class(model=model, **kwargs) + return config + + @property + def model(self): + return self.config.model + + @model.setter + def model(self, model): + self.config.model = model + + @property + def base_model(self): + return self.config.base_model + + @property + def train_data(self): + return None if self.model is None else self.model.train_data + + @train_data.setter + def train_data(self, train_data): + if self.model is not None: + self.model.train_data = train_data + + def reset(self): + self.model.reset() + self.__init__(config=self.config) + + def __getstate__(self): + state = super().__getstate__() + state["model"] = None if self.model is None else self.model.__getstate__() + return state + + def __setstate__(self, state): + if "model" in state: + model_state = state.pop("model") + if self.model is None and model_state is not None: + raise ValueError(f"{type(self).__name__}.model is None, but received a non-None model state.") + elif self.model is None or model_state is None: + self.model = None + else: + self.model.__setstate__(model_state) + super().__setstate__(state) + + def __reduce__(self): + state_dict = self.__getstate__() + config = state_dict.pop("config") + return getattr(self.__class__, "_original_cls", self.__class__), (config,), state_dict + + def _save_state(self, state_dict: Dict[str, Any], filename: str = None, **save_config) -> Dict[str, Any]: + state_dict.pop("config", None) # don't save the model's config in binary + if self.model is not None: + state_dict["model"] = self.model._save_state(state_dict["model"], filename=None, **save_config) + return super()._save_state(state_dict, filename, **save_config) + + def __getattr__(self, item): + """ + We can get callable attributes from the base model. + """ + base_model = self.base_model + attr = getattr(base_model, item, None) + if callable(attr): + return attr + return self.__getattribute__(item) + + def train(self, train_data: TimeSeries, *args, **kwargs): + train_data = self.train_pre_process( + train_data, require_even_sampling=self.require_even_sampling, require_univariate=self.require_univariate + ) + return self.model.train(train_data, *args, **kwargs) + + +class LayeredDetector(LayeredModel, DetectorBase): + """ + Base class for a layered anomaly detector. Only to be used as a subclass. + """ + + def get_anomaly_score(self, time_series: TimeSeries, time_series_prev: TimeSeries = None) -> TimeSeries: + time_series, time_series_prev = self.transform_time_series(time_series, time_series_prev) + return self.model.get_anomaly_score(time_series, time_series_prev) + + +class LayeredForecaster(LayeredModel, ForecasterBase): + """ + Base class for a layered forecaster. Only to be used as a subclass. + """ + + def forecast(self, time_stamps, time_series_prev: TimeSeries = None, *args, **kwargs): + if time_series_prev is not None: + time_series_prev = self.transform(time_series_prev) + return self.model.forecast(time_stamps, time_series_prev, *args, **kwargs) + + +class LayeredForecastingDetector(LayeredForecaster, LayeredDetector, ForecastingDetectorBase): + """ + Base class for a layered forecasting detector. Only to be used as a subclass. + """ + + pass diff --git a/merlion/resources/gson-2.8.6.jar b/merlion/resources/gson-2.8.6.jar deleted file mode 100644 index 4765c4afeb2cdfa224422b174a4e5d94ffc9f714..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 240255 zcma&OW0Y*$vMpS;ZQHhO+qPY`3af0}wr$&9aYPpShY zy&iB)3(Ek`BfHEy$oe*|_}2&gRc%NKP*IhH;A|NOzh?mgN_UMXp4K zi7K`13*^zv#$lpljr&4soZqbEYKoeRj5`6+I5?9ZAIz}~L@O%&DY zXEaf2oD`{X4<4;hL|E|S8QGeyIlr^6Z!UfmbEVK{>ec{vuQZ#4lMT%6xlEg-Cva*=i2n| zV&`AUM(eInwtxN|Kv}HY4EG3&p8Y9i&-J;|Vy8bpcN%i~L;udttUG(`DI*ppdG+1u&SN^s^$Kb^mr2oI^`9V&&n;0iS{#1&bx zhkA~G^aA|%O$Pwj_>Z{(@_o~d?QQ>OTKs=4p#NoIZf|dHWBNCS2>)(q?rd-OH`eHX zVeRbVWNP?#*MjjEPEMv~Hm1fde=`*3UwB#CxtKcH8QT1f8{yx%{qLjV{hhm!rJc#& z3{CvM`ntMU{@u0X{KXF#+S%E=7`j;6+d2Qw&=3IY-xC5RRn&gw+k^KT2@t=}w)Q5j zHm0I?=!9g3h7b(X5;GVEjVAhcmC2T2j=*e!!gSdpU-blzS!onxtsM)0u8 zzX{=nb4>k0E{)F#fehS=Cc<=2l0cH2l$~7>CZFR{Sb#*rPcdt17pvdF&d|4mR;Us>XKa!?% zXY)W$pwBcS+s9-O0G~n5ryC!;BdltJ3hxrxEj>gwqnfsPT4l0t-*X7HIfY+)U*ZAJ z1EbqO;bdb7`oGVc0QO0p&2RX`_$J-IDmU?O z{QK{zlB{a0jID~|Yu8{%NP_~MUu>%drGkv4VMAFBD@wV51e5nu@a#OnCNbC)-JKof zY_jwdKpZ#wl}@fy|2zt;o*6Fp19|3ocssAK`q-|e3x%xDZB9R_!0otjR z&IpRWZ***lsLrvR-f<$kNY{9F@ymp^V8j#K^>4)durx!Crm_>;4$Fhr#)*JJEtp-E zk*rkblJrGYPV>wVp7yRhuMNnM2)*gKEWVH<3C&513_%QrMk-KA%MdQDv|Pn%NmH@X zza%^8< z@Tln!Oz(_(aM4*#!FkrqKV=(8q@vVWT?4#KEOx!7QgE@VlgQ>?($2b!PE6uSC(>aA z8R&2Vi@ZE+%#~&wL6L9r;nm4Z>c z_vNkvs7&U@PD-jV+Yy(?Gc^$_g0ROVA@+-cl*c0HNT!VBgkU;?JTZ`DXe2zqW~DvQ zonam_9X8{1G!fDGQ&J5E;~n5s+FUc6q(v$7e~Ln~*AMZ=UoyFbtOBk&hP853JYiV1 zZWjkV-2%pLPS=Z3P`lNdhwPg+`A*3(A0DiV$@0Lp#7n@yc6yfPO}O=8+x>X*7=HAC z{XooNDt($_?RdjvEyesOp5Q=UJN0Na587OWKA0^Zk_jUzRLL)RYlzw+5G5L%Cx;Hr zS7L3@OLD4{8Wz5~i{P!Guc@pT`P9xP#*ow*Q2OUegO0xKLXn#n6Dtco^?!b z`2QUhlcgxhuD@CI^P5r0{(?=VzEM%c)Y;U@($L1z%hc(gRN5UkDFw`k67r=y8e`jS zyBS&8!ML-4$gYL1q;S$DfFmE8#6p=KrGh^l48J3cI!l2z$t9t5VshOPYekK+Zf5_RH7H$5jNL{I04kS;Du6QU9 zJ)yyp#t$eAmjQV<1|(W-*5te(o}z8DG&RM-fz2c+2zGPN%5q9hY`Pu{ZE{k0h1rc|hO zie>RNpZNp-16{6kcB=}3Kdqf>wbm5_Y-UfNDb8Pe2yW1pM%Ub$U&T- zx)3;#Fh_=xia1V{|B5P|9;C!I$WdGzjv43)R_3V`bJ{w2hISpH|34W z4mED50mKv}iVF;dliq0Ff)kFCfnZ)4p$21Cg$qlT4ElvrgT=;hf(%+^OLA3549mp0 z_VJtnm%lkvKU@r%s_nzjSb+O(8Yy~R0nd!u8W(WZ#h(5w5q zG+ugEmr_}2Gpj?+(&_b7O2bqTrewemMGu6Rx-8PiGtkBDvP~Cp5y)`$D z9$naOKCDH#S33$C!O=oi5jUsu>4eIf5aZkYlon-D&6Jco4dSVOTej!{CpS?pXwo;$ z2}he^+dNJ+R`yN?(#e+lcUbOU3n9mvR?P&|HI~{a=XWmhn5gtzK#^E~Wcdb3(4wtI z0=Mw+GnV7LwYf^876LAl4!e$}5`%lD1!PQGXFz4>HB5hqdUarSe!$2?zK1%Q%`#&X1cae3Hb#mJBWXouVhU>9$ByFK`K~{Rw%X0hifo;>T zm132gCsn#vD&ufhkg^uxa@d|AiFX)84Po+GAmiSEkYwWIT~+91B2 z<5o6{R#9+zH&qQ8Qe;A;)P zu1~(v|1IMcdO|~T;;la!OBMB(66gq-eCaXMTGa+`DpMYZ;2^SsTrhgVktf| zksOa_M&oHUu za5q9h_~zC4cl?l4eLg|I&tCrEPn@#iJVS!_8jqbSyxmmdm#{fd_?h8`o)>|kxHjcr z`ntBvR0gnthbdl`It0BT768xEdHonX7R&(_ts&!-=7=zrN=Vu-FEG81}^MTM)!~wEpY}8oSLkEHLC{czE_>q zYa?8o!)+JrGXDG!(*?q<3Ok9!O{Hhmy}I#lu76+3Fk)KrJ^zo#xj>2=;yWM!fL1U7 z0GhubpnojO#avxnolI3c9ZUsH3?06A0{-MHM_3Zf=Y&7kDNR%i;)VY zC!iq#5zsFqY104PU?Xo<-z2@Vdx-q($f79067QbzZ0MaoFJc^Zd}@UWcZ{@f|0nA> z>+3oT_w{xq(;dJpKAJct9m6P1)qxZak{POI1aUYjOb0pk9O?*l)4+n>DKZr%mN{z{ zHA}bqfP*yhyg0)6{H4^iM~7ts&mA4pDAR15t*b)X9XxYu`shAVI%zJ2)pk+aC~jn= z72^Q4RHA%5%ksW+dCT_dDB3oXXN$AcKtnskI^wC1@(27hl<nf3~pG_%c4%cQJ?JmSZ-8B|`yUXr#^{1G@;_gvS z1pEt40U5#sr}A8gsz^SJpL8;p-ITTB-TJ@ws#(0GpGGMR>D!8Re! zXNf@cKT3?%QE2AUw2??vLQafVP@L24Go18woY1iuMN1QuA3lY#4iYzW3{^jPx$=$2 z$+7H%gY2MAUu=By>lc)$vGKQED(w7g)Az)q!%;KPf>lTyr9^?K0^L+R(;rt*f%RN< zLtBh?D4OiTw;C*DdGJ2rsKDaX7$c;iBy5_RlTAGtxdJw4?|KzCPSiY7G zoe~R2l5`Mm2-+jM2$4z4CX`0XO`2e?SsVdekm8cKEH=t=6w9CFH)?hM)XO5Q#Jw-_ z4sSduT`1UOZa{1(Wv)f=j2+UJfQKW%HzU&{CYqS^eX$cH^a&yI5mb7O&^yhfUsjif zD;he@qA$UXEFRDxAD(%eqrqhH@YiASc7nts-8Xhlz8w)evHy?U21vK(=8(^z%BBHmsV)m z`UUpyE0MTV^ByW9000Gb006B2NA3KxQii=D4U|>U9^0QtJtbc|iknDH>kEt|wV@)) zQ0Es)5KaPYq0n%xEf$d^5;IDGsEDXy;I6p2n65a^C<1K#xdOQMdZN!cBKJ8WxiTl) zVkZ1%eCX-vrP1HBzv%z;yj{7?@J_nVc0ctuZeRMM0K~*iv9%;`85{GPKdrC%T==Gc z42<9zz)#p2_-l34nE8oG8&Q4r#a^nOft=?Y*!#ooG7$4Cy zt5vvKQF){ytyR3^BDGhzBXj&nfir?eqPvam`i~f;8Xd-7KXX_wS3%+STsZGBu z6Z^(4i=E(uLwZCtQ|p=`XRESn$P_B~^Kz<8y7`(RXzRwVM2@Uv%#^d~Rh=;VhCk$N z@Rz8+xOL(&xKWX|efHhaaZKT{P0oz9qak5JY4 z0zJ|tabxQ4h%apjbAzRJ#Mho}d|5Jx_(Sw!lRb-Z?TECm;xN}D~r$8P#9jJr92+hCz|1IEYmN_v!Bb z)B{7d+H6*(b^YBEJRJ>d3;TS|x2C(bSDm4LY}S{;I93H0>Q&gI;1W*1dYp=f3HiUzO%i zou!pkmj2;zIF8Z^-c`*BRA^_wXGG0!hccqQYna!KLR1nQdyiy(7aU-1=RE2I;&{e1 zF0WY5n)cOlis%Cgh~}Kc`1FqC9%dg@LH)AI0!}Q*ppLrG_APXnf(vNay|smO zQL3wZ7k^ltm-c!r2s9M0&(oRN(3G{NM-S#u7V}E}gbJ#${OVp`=>am0&z3;Xr1>!9 zkR+VV%gdgkU0A_`6|dOW1Y=hA0vB4PKY&(KDZLB`J1kR6% z2)^Z0%~#+!iv$}W1TI|{Frm*T!X^-(V3a7Lo5LPHTay$mSCV3;8t*&PU&OmOCvcZ# z)ydXu{RP2&;0pHM9|NY(j)jqtPJebi5RR9!(8|_~`8Wd!0dD`KURX?s%#rzO@-cEb zU$&+a?S&K?VaZy#f)Du|;NSQT>of2eS!XOi*9!)&f1tlnwG5@m?BMo-SQytG;xN`d zvi%$VffoK%lAaDtvg3G|RmfU77qPp+@n>w% zm)~jW7w~d4;{B)oBav}^=zHaHc}^&!sAL6l;+6%h!b@neWjtIcy!;{^?@@uYvduE$ zl`!!_bX}W?<{3b^TiP1VLvyW=uJsYSbrRhvEkcqUkVv)zVa@$zB?3AyE~G>_hmw*o z5Oc0%T6mwvpn+h;1DrWw;q=iGzD1d#S$bs47Oq7tGuLaDN`hF9*#2`BVrS{0F-;aU z@yJ|hH$756YL$>_SAo>AK5PukeYl9X^@C*f*Vj<;j!>Q1_1);Sr5+V``R$`cLQC`a z4m+S$lB4N$l&F1Z%<%ZU4dsYg+=Te+PO6u}Qe>nZ_i$GekBI1wZuQ8SDB>5Uk3R$* z!L36V*U3}xt5+SR6i|0@rPf>)Q-MLZ9%0R0z1X4F8zThEuO^QshRpuEuMaVyqvJ8} z?5SB%E9b6cFlZZ{9;8q{yQ{QPmWE=hMz_3=<>$apx+e53IBs#5$h78nfEKGuUDOlJ zi*wPwsVGzHK_9}V!pIS>x9j0GKnV-6aDP}w0EV+BCuJ`Mf+Q6*Y`fQh70n}H zUIo&yveO_6-{T>0%}_AO=OZ&!iB3znim2a=o=xGwxzCJ#tiyO~=Gpr?-&S)PD%0LS z3`mMGEjMM%XkgmM+Yu*OoHvS>Fm$;wF&^nzD9kC6c5E6WVI@vDC3Fsl+6@$QT3R$) z9OO5%otPoo{>(b~%y1lrh_cY}Nl1Dl2COf64iA&1#7maS9~dcKmVd6K?zaavQY*8&admuhAnd+J z0E!xz<38>K_QsgM9{U6&JE`;t);?~ZNRZZoO;HHo24Yl`Vbx|dOVhj&ukjUEhh0cC zRxdhee-0~_T6Ru%ubapr8UG4H^ES#y~qbc%PO z9>pVc%J+cdR6>ucdUaocX)%x zlQbyn6^%+s0+E6fbdD4>$h-^cXi}={aHQ^bxPzkuuQ3jWInl|1U8u?zEg7qhBm7gDyI{g!{Dc5AiWz*^cR}rCl(|3*HP|?w{IVs*t-_djm z_KR8fP-G1k(;@}kx^!?%DXX&DQF?;GJ~2sK8}>sJ9VcEw2F&UU!**;jp78qO!v{q2@`dNWbUS3DJ*bw9%fS{salw}pw{vl4z# zw#O#CQNDnZ!pHr@S9xah;v>1ggxVHd3q}1cuR<;U+d(BX@Icg(zGF-X3I80w3ZvXtQR zc2p*qpUayz;|8xseK+>D8>Xa)hEg@Yh7X9vN7Yav4Mv?&?s@+m5=^=)VV~|#qrt4w*xjot6D`M9{tlu|#_Z?UCncj?{PyW@qjsXQ~QJ03eoB)K++UKGFBU2;GF-Y^M<7G>Jt-XVRpXX^JA zD8J-nig(rmCeuH{IKjO$+X~N_EWNGI&S`T%3QSdi=NSTVS1>zRY=(U0jR#?;qcXX? z)ookibV^+V=w|}WiGTbGZ(3heDcI-bDpy=JWbU4;2VzWoQCpI- z2l6fi;D}J%xS$t2u9fJd#_upJNVAa@bkRvGO+U$ldq=;j>R4@TZNeost(iF}wyf+e z1RrnF3*Oa(zoLu2_`U>o@s+Nwlit7&KPB7!ZvB`s_&2YxT3f?Ghe?I3DjcX4HSo(r z-U|l!k8M?c@<} z(&d$SG|Pj=$-z9VptVLCY;QAW3%&pbpSOX_l{OwV=a<>b>ewSCw)wjnCq--Fla7>7 zZgH7mFAq~lV&G0zH2 z96>Igv}kDS3euqO(&}!dmcSn<>1Pw($qb21vX5tu%KD*Wzczgu z8Y>gVnZ;8Zy>l7c;Ukg4?`&DXs{g5nHF|r(l(lJTmf1;P;n23~Nu{+|hFUG2iQ67a z%@1IPjEqrfc}Zm%t29HqJFfy#A4y{=iMe)IS>3p5Y=zp&I{Favp67nUYI8Z~wSwzX zmN?X&R@OBng;o|XP)J*ghxK+P>nx>&%w4$Izq_hq}Yo6Mu+X}`F%`pnj)?+`r93^>(tMgQavPY)n zgs5E?SP#ZOv*OxGN8hvk-Y)*#9?j0S46JQvIhqRdLx8*G`UbfBwYn77wX!!&3M7dL zXoy}&@+HJt5k=2=p>?FayWG@_(PtN~aC@NmiTri1{P*_8`FTO2@j#)KzCQiYWJtWs z`hvz>d1OjcjuT_Vfei<=e3|eVA^n4|vt$NrddUHKb4yVkELdz{aJYd=}9FYSf5w~#0Jn~%Yw z*Ix4OFXG+qu$jS|kL|q=V(%{uzWHL+EdK4p1!aIydLs+7?iKz9Ed)wTHjp3`vr|Hi1vm7`_peqEcI>H9_ zxIt0A3JjUyYgN8uZpn|rF_l&LPP&I5JAD~lzwQJvR^VT1yp{P;+U}-56iy~u{|tJl z_iVz;srNMFN9eq?_@?blv_1`rq#}!FA2?BcFsvo#uEB%1-#OkGr80No!?|%fzZy;; z-~^DsSG#cFZtT|SG9(WwP2k~8ri?JnZ0af$EXzf*y9_rL-`Z&MXxErByAAxA;Grsa zm~uF}3`)Oc6X%E#NdY%h6FOClSu6N+60~5vw`RP*Hlgv$5ht40oE)b%b*eZ48HRLY zcgUlKDkmF#R_(mox+QRfL+jElEtla}fX6T7S-L25^x;O=TN@vq4TM>?k;cgVwZy%( z5|0+noQIJ{+WHFw8cSDRn3uph7Cn~+iV1djp-Qc=l4coeH!1$CGZ%x;rDqHN2zOpt z7HaBZu)w(zS9dPAfzyv2;?cM<##NKF*cWrer5S%4jo6qw92}6@!TcK_nLzYi3J=Jb zz|?#wz5bNm5HAnZj9pfX;FWyX-vInwkd-^M9KgQ7mAkeY;Jv}nJ5LUj?}&Ck-FMxB zO5YG5-ViSyiTZd$^B%!~);b|fzXZ^~Q22x_UN8W6kwF4u=Rf#NL;L+SAo(J_qe2|c<;NgnCVjPx0q z*%_f{&6wzo)O;}-Z!xE4VvRAoQYha^QpV*?IeDg>UV^9kYYjc$&24acQxEUSS3N#? zoS_^%)Ifh?a2Uk=ObEvjBt_>0l9oJxAd@8$7qCQYmrQF6!|^YsaRNzIizn`D`Zg)i zJ%GSr7KzzWM&tyVlZhvWm`36Rnze~1UY$a0A5WVMmiA2k>7j$Mx^Y3E2IF~QkkzLX zjf^a%w=l9=8(XiBZ`{i_;_DMd8COaR5t@epC#qXs=S?C3BxEaG*M*wj;FB5CjaFfx z48vi)C?ctpRROC2C{$!17C9|>#j`xOR^@(lal?}BoXFj|;MS5CUhGEK#$SH_rzzXk zlAAk`8$Z5;xm4QP)z+VsAPU@x+y=~(+r-WHM|;Ut zP%^teE`BB3-~k~-3!ET3ond}CFoN43m_AHk1k9k{yzr!q^N^ZV&?WV%V2~ZFFhL*) zMY4U35q)yoFc$aL0uf78@cTgbTP(vAa`o`x+37rk2xBxbm=`dEJDyeyNSW#(Y_-Jk zBQ$PFV#f7m2NMWYWNX!!Kbyw`^}`a{P{4cV7}|)k;DhdmrAu~{iIA0LpLz*Ev&ED^#ADDTIS@u7&vAqK4RCei6*Dq zDV3U{ziSK}9vj|V21H`2Ovrhn>kJq$XNk1uGF)_)FPt@XI21{fmNjwOzU@Gi4qIaX zoCFD#%j61MRI?=aO5OKLv0sOAOCF@odbHJk+BRok996Z@D+YFGHZw}bwCps!5FFHm z;%49=FZxp8kri4pH;j{KJo0D(CTj>aWWp9n+=Mi>L7y_{+{Iu6A#V`R%eV{{d&KZq z+pkNy3|w4Ou1nwr&RlJD;Qaw-XEr-{c>(GX*A2nFTFEe^7utOR>jAZ$Yf3)|eSzwM zKgr*JM_>2Oturi9UAmXyz&m9$ny+PB2PtRZyn5Saloi92H7Q4cGKxx%*1+Ueu!*MW^~D`>-V)IYz( zXypiVmvT=~C>R3!Ha=p)Fm6N0%f54uJ(ghj(dt0a(_l{^9}>MiKq6z>a>C6a86v9@ z?bi$P$$ImFQd9{Xwg=nhWd}mm*zpkuFC_*0VNh)M`(exlm?)LO-h~Xgx?xUSEYzzv zWKh1DjnWaj^(U{hH06ORt`G771Z|9 zvanh=vPBK8m31p91_9a~&g4$V;l`06=6F!C3n1n&*Z~M1p4A(Pmw{&g zj7u_t5P@u2n*kz|NX5Pc@>F4S6wMeDnNx{(lwo*OPKZ@ns2ybM48yMe?G%~nP{}^t$2m`(my{$uqV98PFl1&0^`b1-GOTJv` zj#QZO+{qsh>2(Dx^2W9zN!c^rSi{OqmZ&>TOb(_Uh5#)m%5R!?OC2YE;DEukEHc#s zrJ8XZtfB~7Ib6-!j#jOf%AL4=*AhAFTv_Yw9L*2lEX{7VR=63OL3Q}J$SBEIgIl#} zNM3!H+RL6sk2B8&R|vN^svO;wgsJ+_a=zVpx>YfGi*yT?=!BxZW>syg#}gPx3Nih> z0g-gk&^ir4+A@yG*Fl|NJ6BT3GndZHy&2AsA-C@)V5qt+OFAOf*YbNbp^nS3D)~)T2dMZsHytl zU@40^YbidTZa{_;nAuysw+6K_d{1=Hg-3@bk9sW!96{?=AmJ_uw;fG{8vyx^$Px5~ zRB|hBNVW|L$n{v2ddOBAALIIR z-URx19d8)4ieSQ7Va8|+5wOXJXU6ES!mWbEq+-Wa_Bq4tAezsVe#^T*m`C+|n`q2g z?ja7!bYoS&a1O+5WXZV0X`*mI;sI$b9jHy%;*yg_eTcEUiRalrf!IFPMIzwm_&>os z-+j`?_)%o0gcR5Mh`fadc=Kh;0x>d+nAs$aqnsS=#bPalZdZ|p8bYd%k1Um z1Ns6Mze@)KOuF(-#&UTS@-x9NFh1}4klU*7-pn`>`nXBIQkE$u( z103d4KN;u|VxOP}fpTCu$c5NWMF9N;z0IDFnN8@e59i13$`?keB#|3u~C@~AejN#;1XoWIy1jEZHPTg$bU#T;1RJuWOFeJ zVVOYKP|$(al!}2)Z_W5?I3QD+#m-K^Zpvub7m7TDyNP_<=83g!BEG$p0_}GRuqR-1 z5?X+y76@87UPCb$MG}_lw!BWWF%MdOE=1V07-^MpW}F`3LJvwi>*;0%X{seb1Ya#Y`*T^(ddarS_jjv3i3~pH;tA27h== zT!5Lp_?W!>{pW?U|8dCnpp7kKHj^?|q{Mj}NBu5@ezbB)v|eq-`Uz_Rz&6+n4;8<1 z+%kCYE=blNv>zCQH39nsyzv|o6P&5qE}Lb))W>qT6{-(%Oj)DS~9mstsB+Y z4x;-Gl}2D25WZ0CkmiEf@p`C0s9Uc6B2sUr6KX8GtOk$e3f75d4ouVvXI>W436(Z7 zo488~Pm3e8%3d~b6wQ%;{rDRDrtdeY&x+vq1dbM8ntQycF zWlib7rdR5Uef7X5%vRZCL7HOAVpy7Prelf6mm8FG1oB5jNhRRKa!uKCQLN=saAi8^ z)PbdoENA+r!=okt`J^2RF)4?A&m|@Vd3Z6f^B1;jDNO#D;47uWDO#PBHi2>nD$atf z1xKrJybHKY{1r&53%X3ITR)9A9_<$FsMJ&c?ZSrxQOfe(A~7$(^g`eqk!>K`DdCMI zw?XnBh5{RY9WvNH*9C&jFkAa%H)vJeFdo!hAYbAuhUu~$@0RrWoF&lofg2MJv@KuK zPDpEQ5UA}a&fJhuvo@JQV<9ut*zvaOx#%}I)oG$qEcVciRe4Z6?OikfsS3x%21<^8 zU7IN6)|qirjMDgWgSj63VRJOawL*;Ka)pm|_P9t&x*We!#N!NGgmv4P7AI9O4y=KYEE1B94un1Kx|k~{6u>p9|+GE;yWw|BkiW93iy ziqfG#62zu-n>fpcz`-QIL|LddnRErai2zKp%-tyGl9Afvt@3RMd>ShT1Lv+r&o!H<6`W50m^y~CtQcE2_I zyR&5W`~AO&ZN6Jx4P8vlJ^$HoVjEo#HNb!obY0u5DCpqgJ}(Gdv&9__=M$h|Y$1s( zQMgmd`3%69^t>Vp{_M{+DP|hATRxXFUM8qB z&=xmKDkI_1kASrzyOnMLG1Rew5up5w%)CtcgiIdIjlHeAz!t-#TwVinT4T{hx9e>r zud(rpde9D$s!UON{>iWNJ~&^4b{0AR?ybDA%ylZ z5kcrHW)?j@Czxr|Gb0O!T^H$WdDPNCLN81V;bz11HKP+9%YEdg+#?rZm=6hL1?iN0 zbGXx;iMTwm7G2Lr7iXueYX+AQss0(yS`F?^lA&Lf-VhA!84?Ul`rtJa2jFrI1{gh5 zP_Dfg&@j#BiyBCnr%Zz*|jdXmUL-(6Pr! z5FwTz1H|}ZJ@SHlg$U_I>)bw2z`mFHcoX#asi!() z@=kcApt^iB5TuAI2RXlc^vn2ym99_@A~l-$96s=gjwbB>wEgHl8I)xY_|)1&x`qGu z0xWc?z@7gFnC5Q{0?A*rvHh0@LHR!v2;2|$D|}#JV8mc}ZeVn7V0L0)GkZ$~Q$tx( z6)IvOMzFb6Lk%JySwmeeK3_#uO9x2SB=H4HO*naRz7nQ_70vQazBRsHzP?^^6~11M zzBxVUaS$OH$r=gixyM<#>IqtkM%CXByVi16c2;`ftaG6K>8GCIq{X5F=tUMmL3905 zu$c(dGPUc5P>f6tpvS3bq@^9EM*!L6Q8va_@puxv#DO?`?!{<~Ob~gQlOtn&V|_ya zKv6VNFjFuz0({hQqeFdQLkMWxz~E^|Apaxz{inzgfe8m>>N_13|G!0!{{pCJC0To9 zMg$)_AvRd-^^iCvin#@Nq>2U8*mW=%(S z@QcbBbcY*d<{YUb9tb3wO z@Qi@*HXIovYxA|-1{oTMOkw=`ioBaMgNV#v5WajAS}8xYMl!!aG8{#6(O=&sj3O#W!Ud&ep*KXkQVbYEQF=q-)iwYLm zraJ`Vl8#i6d8hCN*pGylOQfA28dB#^y4`d#1=;f8;C*_0_dE83`*?WUfCqR|Y%(lh z0ET6k-E622*3CITO`w#TC62XuA|d>6oQh3K6UuG`M)p z$LBYii}>D!)(h7PFzh)CEnvG~6a9)~2ksbGBaQ`!{rz;Pse9%)iqEhrFH%d-gBOW5 zxX{lK^-*K-44fM=VTduz^H>-M<-AP%nbHXezj2iwqN71e=#^ad6k}Md(J@xLyc;+( zM#A~K6mTe$5FVmOAuF%f0uPe@H)Xexgt(*lpF@zZF0-`C(aj$pSx-wMmB%jI&?4$V zN`{wS{#xm}u^~S4JP*ZASGD7`Ya{;An&4<=7-+NDqeC5mpa(cQ{i-2{&{)hSC0oXs zpEikn;Op%+BI$yh55+%1<={_-PSSsKfO40R1an5P@%p=-Q0yX^&XhsiLlObMbXFOx zfyiG`NFAgIo$1HD@{nPTZ#{y@sLvNmZ(KFKXvf+t@Z(eQzdl=2Z+UT}9a zz6ZXT&Yk-`xWq5CLL!+-kc_Nkr&2A!q%D;5o1WY?qmb7uok_N(;Q7i!lUzbRHsx!! zfJUcOHg?EblG)&qyD7Dlp@+yd>zQ8|h2pnFU8UcDz)ZTqWNOH!*13$0Dr_y1gKcc zl0beH(d4zU5(Aq%n0e_LNm?Q*RMOa06A}tEW3dscrJ1c7GN({;^fpm&F4HoDThu^I zxy}%@9-*K5voJ7?kslwQl^!3L9v7XJq8}fh7E&6UAPBIhN5}ie`-cO9U}$1sreSLQ z^j0H?{cfQOBcyQygJYP6{EyJ}(vELxzC)({4xR2VLjS)>=N}?`Vf*jDAnaLsQb6PR74R$4blI@*jWZu5^Y(!fWp0SFWlkc4u!kBySux)UfkW?-QC@t z!YSMxYIyx#$MoyznV5eWu`?qhGV`3h&&s{NQ|boXaKaHsS>1y9r0$|{yM}x&7+kYP zG@y0)nKm}rIhi5O+q2F04cA~QBsf0WJ+VP;)F4iOPt5`=2arg;4d#1Ly~k;XV^85M z*HgL8pk`W3i3yXvxYxeXhOtZAqk2H;F?@U)6UxZ(l%=wnj2!4#)&wu1BG9B#4o{8P4$%WZ$e27z_(vrTnLRP6l{$Z^B znq=gl0=EELvltkRc^de?Yg2xfEjeCLqr|;Q3|cQZ0b+`1lJP3_lJpIObiNYEltJlV zmb*BSB<{)nSYE7p|1eK58s*Q_g+IKE_sBo63Rq&Gk;py>(S`jsWvr9B{JRRVIMx&3 z^k5SjxyrHvrO`*H4cmcyL|NNvoqaT zl{5kQ@}7!&@Su%)-z}NZJ?kXd;bAkjoEqfIx^|i#Z%-u&-WDObvixPz${w4TJ|JHa z@99r0OLjOiXb7$Hp_z`)eNuL=Ix??F-d}cXd%)OZu9R_moxyD>XVE4pM$1=$>fl;^ z8&v(R_#!grjs^^aj5Pv%0RX0f(a0^B0d!BoC|6W=(E8GTVR2y$bkt3`I>p#crMXB= z19hbvIvd!V&VVP>&KjZ!L{zWR-^3x!iXq0WrkEP9h$ff<;-1M&jU{lAvdOnhn{Aad zsU&~I*3`EqPhtqSC;DT#Oi{5O3+r8Yvl}il$M)Out%u0Xuvpx)HEjY2xhLES%`woi z9v=w2qHf3vD838E6wr^CRx_AoTUnT7n`%c)f0e12?gK0~azD$-(`eGj${aWGH1=y- z5GqQ@=cmk7XXvzMVGjqJpOFsP9~b#!EkIAM@yhcfmhEq)zc*Z+fcJF70B7Ud$fiQA zHxAl`hvI8=3NV`!t~z}Lff?;dk)HEHLlhUBAxRZ1w}stvVcPL8iff{!+RIS zhHyAV6vnV5@ltY*MMj&qQp40_1CiCCOj>0o4l=IymDHf5_VK?(-THRbw+OE_7E^1$I4jnpa%) zm73@XY`rY*Rgy7;@V;xU({ZTtnuMWSS7B!m_v;KSxP`!|KczCBm-Bsh8pOd@E?4O? zEZe2nV_vV2#%7bwUXaG7N_m3@+%IN#?8fU)N31)^tlJ0yr3y>Ph;_IFk>LsUQR=+P z2?HI8x@MShk(Kb2j~H3>y`{|BE0}%G1w6IPT9nO&4+4=xx|edR)rJ)Ee$gH`D1u*( z?jMKN>^geo{g#B+-NlWTmzyv@AlNXB8^MEjy=e;Y#P05EN%Z;EQ}(+e4jwIXGj5kX zhL*6~@Wk3KZvI8>q1vfcnjW=Pc)+FA_XDl=Q-#8nT;5Q(n;B=#6@1sUvxCU|R5<<5 zG|NgXY6Mmlvip$H)Pc=h0+WD1wCl>dXFaRL^b_6QBFyRJ>=X?SG>!(9V(i;SWf$ zFXBio@CRW1)%yDjV8Criy(+HQN!=8%{YP+O)m#5(V-QcBzi!>r?2kvDpTbHM*&Y;i zkAwog+kRI?a}ZqN0GvVA!D!8$J3$Sx4_E-E5O1slXsc&!o`)F1u%YQ^@^M9~Js z+I2OOlWI%OzvdSBdXeLM2$?+YeE)s27A&eLo-n(EGR<5`yOcdF9RZ`%RN)$cmc0$4*MJf=0i|K!=H) zUj|jS9iw;w4t5QM6d(tWQrV~_bmN<7=6KxJ81U6AN)Ds+Leg@cZV4fsLuA4wH_o?s zD6mw?P+xVLrBvc0{Zf~ctTl0i-K?pZ(pRmzbluxawG7u8s1HW$ zR+?v|hlhjB3eVk-0fE1t=TDo+!p$Jp*Bn9*&ppHhhLs5K63u2$-R0%T z74(>Cs2XLH`dG!W-dRgQWWcV-b?@S(UGGA|63Fl3ymm9`*Y!BspgnDxDM7U>>VQnWN<`Te+;#(b&($oFeT`Y&kQ8=7~h*XaPo%6IRdHv6C#bIBO< zn2-XJdG?#ikdCtYX~-%buTXV4$bX^5H9q$IEcehF!7hgc4)h1ZwfyLg>{K@SB{q46 z<82xWZ}92j_CEj$g{cUbj-bYx&Z$op8X=kyHx^TJ z<6e3943Nd;3xkZ|!qK9_ATd!Y@}e^1MWJntq6}eFl^D#lheC~T`oEELNKT5Se*zm9 zK`rAW3$=k3@uTu7_OngRB2ZajCo%g{2AutE{PAjMnI<}h*HES%If}57zBrL*QYU|r ztAn4olghf(KU09g$nUekV|)KiBc6*3#((&#ng9QxQ>_20nH4u|7Jeb|Y^1I-(81>w zqK+5c3(U{9?t&0v;TE-p+{ zW$OAknbyG6(epR+1N8bkEe^WPn%b~6tgBV38s#qC! z`5D*o<@nu@G{g$lDe>vmYjoBH&^$B#*xLOqH+c1&NAyKJ0a1nIPG0jkXP!4q1F&aN z{IJ)&`TDc4c@4&6Cs-v&x(S$z>!1%lb$#)h?_83A$XUtpfV(rFED*X;nnka(&cQ&7 zBqj}TLOop~KV+;W4mp$*#5~W&ANF5ONn1P(3NiV@PHjbk{sTLO{HN}ci$Z^EFeFf# zJD)@vl4Viaio@)8UAQ%*bSs=C6K=fzLk|TC7u@z2{zh&%IfcdgPdqjtHBU@*NV`(H zYWG3N>o-GFt~ABc5j1~}G&0klGzryd?*Sz$wjTOHlS^mam#poxa~SQ-4-EU5;9ZMy z?~iagrD~G~RxDPQCpdWXi_6)8+E{N#W{Fzlo6Bc-z5c1qYj6mkEntbG5PEczYp1BxOE?KFCn{3{u4X}`vsn=_!oFe zY1GuDcZ+dZYoi-R{Qb4o^8a$U6aNoY`+u>ull=X4PLWF^V*I+UBxgg3lWZsXN z)|}JP^)k(I#{2y9&z>8aC`=z5#)bmZpR7_I+z=CSkIV-8teL%IeayKqSSof!I^dW7&8hjNLiN@cq7KYzvBx2OS+&UqY)@HFD zk>6?+Fzt%(5r+R=-e_hQhJ5gcJOiSse7HnDCE4Lxkuyi;^GLe*K=1do5=Aca_-qDtzp^87x*zgJ2i zS-DX+UjtS;$!yI3oKKUKu4ET}VfxtMP*2me`{CtQ1jepCK z8A??Eo509vuUg#p=9RiT1qO{sDm4}hm`WiSJNQUtWBO75BOBm!W&T)u!D&)mc<~xC z0#KXXs<&vbHpeMfx5sn7`=~~ynE?njBpbdoz>O(M&dd=B8cow;2`SX#8a8QcGj>W_ zPZXzZ2UJ+P$luZMB9IY(Hh&p_Qa&xSgr&ox7Vde5 zsHof-^&H4-OX-VlEG(AWu9aVMdMw?+ji|Eh&}z1)qiFj(A-RWcmCpbcA*e|CVN@u? zl{&t|-S%^Btv=AMl!^T$B@WUsP&0qQIXf6Iw$33l;xK&|@*8^+%w116EPsg3A1nJi zDk@ZMeSz678@F&_E9Uq!g|}HhE#^p-C*G+{+242r%^;b&dsl^LORdo@Ysd{m`r3?Y z3S6{hE8;MfI;7_!nS@*U8jb^eD~$T2&k@`)HKjIJ4_Ok1ZofGI2U9n#PZe%0p*_A*Y zBnWk^pcC+e`&HN4dk_I#b5sF>Ciz>CH@`?D>dlZS9};$KjPy?Seoi7NMFLr0FdUp3 z!gsJZY|a=vqv>AW{0#3OYGdubenV9tT4>MqaC_M9y5w8}OyaL*x?sgCAc_KjB_y7p z##`8|nlNUUZ|;9F_f09zB~c9X`%-(6LZCJCT#;JRvGzH15<&Kja%GsMosJ(hDNx{R zGwe(I!7|})LeHpx1o;gRMR-X-rLdo*@1k+&ADF{K>v&SXg5|)buSCPL__|>LF z+b02z5lj+gK z(-Xlrsr<@)s~+4>Oj!nD3M{erVC%`43sbCBH!WGq`*jbMLu0dz&FxL2-Zr___7N<}>ijI7O7 zF8_-7y9-#sP*GKkL6V4Qi-!)k9z&;G-}K+=Bp1zj%Vn@D(_RY8i@+2>vW%5)qS~0t z-oGuum#_KHDI#n^C%sIEgjnhKaIBFEye#Bj$$nps9no)}KRwIkzYs`i0+J-CL@!S} zYM!{ESQF%>QN~K|Qj879fV|pgszkq%iK>jhQ19LIe$%;OGN89BeTJ?>-?NRdo#ROG z3>T^cXHA%J>Ismabm|g3w%i;yjuMj=I!)3dLn|8-qqP$BZHyi)pmlm6(L}8WF9^2L zP$+DQ-4-}17-vF1GELq*{GZ|eWNIq85P#kX(t(%-(3{|`0U|1pmKnSHuE-Sw1~9{wJ0K5b7L zF_I+o1{srKB9W5NfFiG<3H0iri>Vo*gc$khi2>3mG*VRNsOrtlDV4w%u~SNO>nYGr zyH|>unyNaSSK4!{h~jrRp1P#AgUmiY?;^h-Dn3^nPmax8z!D&@-M3zVv4?p+!C4sq zoRV2Qo=BWbl2JH*ft2!bUdcU&P&}NZNQdf9q0Oj7oN-GMjko-sIM77Qy*MX9J$~|W zK20ZmUlft*eoo7MCgJrZUh`*L14+v)@RE5zG|~(8VQvq#Jl{gxAf6`9sElNSq%6*; zjl@(^fqD#ve(8lF6~=hjLtN$c1lOJIxpYp{evDPzp&B@y@I*zBbyWaM^2Q`icRwf0 z+rBEA@_cg4`x*?Sdzdr!DV(zqr)$ys`%)!1@SyAe;8-F-(Jc_KJ1(Ru;gAGe7Moh~ zL8?lEUVbs_0@a^yxIHX4&Z|f>F`|?ym{aIP4?`FBEFs=&Oi785nfM#ydGi{ zUaO)&C~u(09|pc^6N#LZSBu2?eDYP`44W)KRQTf3O zR1sttIpT76Fwp8nL&mU?6EG8!B}u6EhW2Yw`IN$F-_s>d6x4KAFoDZMroxsP20@GF z!osFsy~v{tX!bw@K^gzqrLt@_=3sK_9QLMCyAWn%!U1zCs*D2*xz2(@1{5kPvdK8~ zKqM!4@x_db$eGeqbrNNC#Xm~Y*>VbHXWZ31(=?QERP=cnipuI*tVM-pyZ32_Oc7+_ z88uzj3NMr4;?|TwP9@eR>YhzKT^@UrV+oN|=ch^nGt(;SYOHxmS2VNRgyv;Jp>!9X z=9=_#$;&b-;_hpxWfI2=;&K|Cb)^DSf!}>`p2o=~r6KMXY;@J!UV~tc58}(IB1;l) z_}ePT&}1pI`;OCw8&u((D_x4^ zSe!ct;t%H>3j{4rLzM>em+C?ozS~+?R+Fh$i&aR0Ouo z?aK1v(0IBEO>!|RrP6Q#{N`_UtjH|9iX(iN;=*nl%7R(CRWJzyXUY8uOl22iW-gB^KrR z5vJziYBJtzu45Mx_gbx^5dW!%jkUZ3ZK%W0d!E1)LU zg@x2oMh8g1*+-+W#1 z#hRi5951+4>7$1&rw=QA#icvzTBk!&#$MSjhniS1Ecfa{%&Gv*a;u|(DnsCuG=x%4 zV$erpWjPpo5q^lbyF)zjYcqpiEy5q zy}4B`aRwwCYW$s z3n$m%Yb%a9ei%C2nC7cW2G=R9sOqn5BCjZ~vDWAdJgMDkHoIX zr#1}kTr*SFI-A%8fwLthTkte&PjW1aK3*9rsEEMEK5d)vn|@4^lN~yQ7RP9sEU#v0 zPh2_E8Oq;jBpQaJ)m*q)>YqGltatVDzIOzj)j;wmok29 z57_$l8R81&1_y+^{$36LTigoaH>G%X@&T3=+5Jn-I-Q?s(Nm-wr61o!Y$gGgg1T}~ z{1f{5sQHRwKj}?&76)-3$2X6sCaQn~HxDjZMVA zE64YUJEW1+h~2u1fAUX`;a@1p=;=1`o2^j5$&Hn0~7`7>{O&L;Dl9N%2lJFaz zRU39wknA#4ZYN>Y7wb}gnXj*+e= zmwsnAXk@^yK=&Z$vi+(0)+K%Q{Y3X3L4OtzO2NF*McjKJLapg~8T^!2k$Pgnq1D{} ziz6kY-bh2Pp_-+r*dnwjC|zma7;@DC;sM;2{qbW1RQ8CYz<$Nc#$;)^1M}yC*Y6@k zuzO^0n)g$bTt%}&CEs~u*60`!XK9rY&j@nq$FdZ)TwY?{cB^G^)Z{|)hue7!kk@zw z@!O=7ZiHFilOB-QkZ$8EfUTGx!~)W=+I+xdBifhqkXdZMM!K$k=8}lE7oE7o=S!tE zMltj_m>gD%k<@MOuxJDA0^zuM%`{7jnpU41e#hN)@k3Z})#3b+dnF86cp$XTVZh0( zEUM0Nyow%#xRXHKP05x?!$U0Y*(@zAV|boftd@l-z|s?*q-2i=`)i6}3Pg+82FRk< zC;agNVxf#nhX0rE!Ko)tM;N3iF|N>eBXcN}Y@N)+)7d!wbpz*p?vu~NzEN=k@=~QEsGDyH@v+6yfOU_~u@^FH_9owrJ>E!@ zr>?WJ&$*7SUg+OAER64Rb1 zcF#~g#>k036iWz+bwuMjkx3H+5ypsF&%+l5#98*khj;C8JQE8Fcp@j(mVC_%+88bN z4U{^HTtblY(d!GE?mjO+rX07^gtEaLj`3gxZ8)15*YY@e81N1G*n~3_(vo9WpmQU+ z%H~b2BP`R{y|%-ES+d?6B>t5dc28ZqWq8AcC#wUGE^>Q6g7_V_U^sHNEy*^7`1q`-zMAoT#?jfm zztVbzKyUf>;O@Dm3~(73odDeAHsKT}(ui}%yOG|A@6Es-ai8o&D^Y&Gz1*M7d_zjL zq}wx?Rh*vk{k3jW$n_EMBAYwmV>CH7xfxjvd3WntutC>mI>g7G6$R(_+CTUPcRG^_ zpHslq{p))u=8^8M&^zQyV#sb{OkARW87e<%2jAU^8%?24X&#Ea=*Nw0HuVG!whZg1xm~kmN{Uq*cQ5$K-og}Tazi>$h`ZWpq z^+|&vcx1~>4ATmdPr@cVlWA1)pb~yrAtR4;@8xeP!Do!^LmkFv-lISo`M$sp!vW2- zeNQB8oJT8Bl9|m&nrg&b6hK9^gE!{X=oawO@v&eev+E>n%@VPno1`9c?frDBh=<*V z-T|T6@m=VM$qXHZ1&7j!Y~GBneV+S1_vJ(s77&YWz$vW#U6q}yNiO1vIV`w<7gN;D z&V`Scv#4xFw#Xc7L1}=Q_Skxf zQJOkd+F44L2j&EKGEW3U+7X3$iaWU&CEom%f`4HOK{|nJ*2u<`iraxV>%8ac@3yb5 z4we(h)nC(FPgL7ogYCbhUSxct2-*=u8W-zt2ZP(>xE)VX2VK6It8WX_+Z9~rTkA1> z;>U4x$NWiryGAj8cM=clko>Dn#+Mn5&)foWQtP9=7*}%nv?G!4siR7C3nExMC@N*- zTCde$LP->#p(C@UcVR@MJY5Uza6t91Sly;fQZ>jfx{PmebI;su*{-*nZx*?%8oDg3xIw8rg9f|lpUd9ehzwx=hAKWp;=J0bp83G8FV*{b)oC zFK&hlJF%6UVmgRD+3w6xlbdVaf7<@*5p-Aj#HiaxA;DhtJzkhe)>Y2()BT4kK|Npxz*RUJ6KRVl9iZ0qY_?~NO$H_&Z|vee(%A0#WA zfthWP*N$r;2k7nqp%Q8yyqLkiX14W*_Eb= zCdrbL2lL~RS{)d~&DYt1=9fD8GOEoBd(^V)qs-R6esk|ZFIeQQHp6N>Pu}Yy?k?AA z_>SY~nZ3e_<*)(!Ed?E2?3&J!2^8BOM%X*b$J1x7LGO>vD+JGcAJZf<1*1CDsS$y}yHSMKkX_S0Rl zIXO3u(R!m-em||o)0q4`T=dLmLBx+zzxM?4c2K1n^zb}RH0oXmJ{g=ppVQXX2bbDg zuYLdfT-ppXGu$G3zM|L&O&BeLRWj3%?+t==Ph|O$Vh8#W?46au5d9glcJKOV|M`W$ zWf?qyXP(VyP2CAh^}7 zKf#GVOR2u+aV{xX4Ar2N^$(#GE=UyKPYNxjB9K<3RH{%Y({audNfOCW;=GX4{-c7H zuT2&CSax}zdn!Itu5%v`vp`KHP*Y~*p^7I;P>#!4_&Sg5+<9AyqWtSYbz$)vi{fF~ z%PC2d5WMol5&%tgU@7?2$Thq=XR++`$s(X#Q(EDvjHpRgS_QGx zz*$aO8Cc4+Xjm@QRXM&?Q7+I`M%4tGK9kO#WB&nOffrl>m>=Wuhu{)juq#gR46d_n zsV}xyQu);tK`(4v_HE@v#hCzR^=xyJ{JwoyCd z!~OD$eqp_bI8g5uxp^&=@XA0!8rof<3;%L9`dyt4t@HqUMaAEOc?EisPsY#GiXC$& zB+<@p*rm`kTO-|FBhB`EMzvypA8ES}qbE)piUbpRGC0rtEEjI&0=)6OoLV|Z8O=!& z#Dp1EE2J10Y-6_#z4ZrMD z8lgpjd*w)gnlOVBt%!elKc;dtB~?h71+>JH!lWEaYLSXwclp2%WuWAu#Jd2CUYEsS zG*wJS+owwt41MX{jjA(8%ob(TGlQTua zaYxuW%5Nx3Qiq{{9Q~G2*I2{~M0ttAYv+-n3b%+>E0Aas_PxK&i{~!(Z;MR(pnOMd zsgHEcB5iRt%wzmy2ZLy{FuW4-A8^Ocxz?NehOwy@Vpp z%Zh@$tauf?6kJ=X_v0_?0wq6xi*xmILC3UXVYw6bs5EAqXwmWr&rEov%~x^yyigXC z?pxgBCNxYYN@E|Zo3Kfh&3V%Sv*NLDC4_H^&*rA4t%N;PERx&)aE|DAx&^UOKcG$@ zeck0%p_9>^ZJR6OJXW?cFJ)+|g)=xExLvlA{^QoBPblpIE_({&){5W1%PRu;Hq1Ve zN*8_R&lqEdOTi*~JSr_opA#)fIu8F)NEaMaFtp5CbbY6uoU5ICy{J9a#GHlNsY7xR zb!>IDP;om9VG{(;;u2Ja^GQ{8U-+p8oI=kqQWI9ek}ZRb^^nj5S8^4OTy?X4)4G4f zVzgR?O+SP)Q?+u>i0u+>s|<@_Mms#05^VZXR;%DvaL1y|xlz_j?s{~f1tPx6+jZ$W zuw*IFP3FVj`0A&Q0^+{(0~D_ka1Z5y{j(~2u;r9e;ne)L%#=c!N0RUX*I9LDA^a4) zS%_Xf?UiXt#WxT9)L5hXQ^MJTcxb7jQMpURwRFWC=n;re%JxV2f%QYWOVGB|wMnZ+ zh$<^g3%8G~Fh4V9fsuS}MAO=as;it{9m=P<_R2gmcG~61v+9m%4V&Pwj!fX5MuVgF zsg}++;^SdNeR|-Y9r@`lkF8^XXqQq$qDH7k1p!UrK*t|7PW&D6dq-^@rldBgm6sx| zL=j!tfvlwEp&JED8Fk&e0%6mw(g?HNBS1j`EL2~O2UM}GEy;CLpR#QZAeg4TIsVJX z+R1>ZZ4U|2PH`1|+wvJ&U=Opl8F|a72W+|{@^0FGd*sfsY2Vq}j{Bkl59koMAFO@h zmaNLPy$WHSb8!n=^3HTji+>Gq6BdQuAV^ZoYq04Z6MYJ`e_2XBxtiE&w{~zcx6)*5 z<`vo`qJTSm8AFa7nzjxObPXE_FyD*26m~afl0k$7JNG*R|Kf@=(*fZiJ5C&QJ2Fwv zW&Y;rt#qS0?wK@r=pFWELHCP53BvPLWGN+oe4CGQa$*Vy!^AqVG&0DYLmP;)!L&3Y zw7Bp6Q9zCzrIY@A1J|3)Jws>>fFN-w_`y+3hwQ(^g{+piPv5cB{aEkRnm&yn&eu#J zK6y))I#eBCvTTT#c4%w5Y-?r9(^pB|--pW9Fbxl>*m6c47%|U{KuDb~kA_feYl^tb zn|LWm`cwu58Cd!GNNg>WQPC^ArZZW~l8buoM;!jZ^cUl73j>Yraz-Gmft>N_jp@lT z?d@gq!+-LFfAXVm@`G^lBO&!okYNrpgmTy}5eGcgJTppo7>YMq(axXYhh@~9Rpjs< zuOaS6*$p)0{fMk73^goQ%TIY%5Ln%fHJ)Vl4n9s(S2nQjk8{Rd$7wny_S}U1J5oAc z9s6pRWLI|HoI&oa+EJ@HeytK~k;sg@2A{AyXnaJJC4E#TU5jEU_|Hs9)^4_BffRKeIw7+L$tL%?tBSL9`Bm@ zjeL6)OPT^#Z$9?)!~)Sd?4zmsia|xuy%Oz_xjWHu zy8%G*w%I5q>m~39rc~3DdL5!;#G#5Fb;$AMfL4s}3yUqR%qs9D+#UUTbd!gDf@iEh zP(#~E#O`t{987=N@sXQ1R@v+4@twccD5L{2!RXK^3tgYyzQXXTG5dg8wEOU(nuV6X zWR&in!@h(Sb9KLM6w9h@<$&O@2e+|R@7b^jE@L$zW)IQCs*unA%HqOvH18s=dpWtOb z{R;^Z;Ww$`@_-2*StP)SxCw4wyTI-L%Hkie&bDaq2(q0;fr1%@ z2;n9qD6l;KnY9skx)%8C@iwDkDwCn7~Ge$=36DHjilNPTY+pb`yiC2)qKFTC3 z?~vC%lM+@=!SHjy?j7kg3&O6!K6NG2XW-PZZ{5~2L<{Cl{t){l;1vO}mcwoF8DAEs z%59pY(~RnUfcZMle3!2nQdEjQtr(~v-4ikg9_3dznkR+lsx?_Bp^E3qtkwsDX+(8#syj(9Y2R%jaag9_Z2>()B|H+f^Bl!Q-t`!Q z2IKQ;_fTyh5?|EaC;!U6evrgm)b2-*40{v}+9!l}eC-PHuCh-rb)+{}umoJ2L@m+l z^oa;_i+ROpKIrlAyJ~f-Q2|PdmD65)w2W5$T(~4Fr~Q})Io;aHAgN{6e*`O?_u`xW z8C?Dfiu^N(b3y%~UV+Baq9SCJ0BQ53wb@^`470SvO!r|HoeuMoeaNE1u=H?Gj|XVocn7|ZJRON z*|DX#W(eB#azo3e3*JO={FuooxK8JQAjmMhY2rZm%Q8~1#GRC=F7|Rn!nsB>|81fH zI06!spYX7YB86Qos3rX3^)2+CCFxs!$G3hX=2_R)F*y~A0^O%uqZXwx$nfbG;E}!q zqD%EMkHDcD?T(4dXeaU^Sc57~r9)-x5EVHoeTMlqTD?$0kTXB?fYnN=$^C&z&pZ2K zHFABEz^XA~?Mv4X@*To?2FwJF{|=4tERAR~2YlDmpA%CRJwNquZbyiabS8%W%wUEi zt4x#Za05q#;xtEwmCVrKBTyRFnSKRF!qX|?@NZRYWMm@DvijJws}848*HP-u z+N|e89C6J(wcmi*#ILg7T|z zCA4(r8dqWq9-`^XT=R|Fam5cY%vJV<=3T_nmbm{B!u|@e|3|2KN9R9J3fI-_2R+uV z2rnux?CR&Jio^4vW;KMnBNbyy*)wqjA9(lLG}Q6)Y`gTiBjPsDnMCR#+4ZF6F3h9i ztKdDmesz{6JX3V3>mEQDuqAcZjvL&1)3`KZ(5&FRX&+Qb)}gehV0L&ZMz9%o*pQ~& z3T)tNrOa}ph)ezp^wGFzoo^VSEp)=$X0+_%HGp5Gxz>?x31c1cEG~4W5FL3tb+%>5 zustc1wE?Ln-zY73V`C3=DCT=UDTd{f6AaBrv|1XhR-$*AhHQ$mQ_Hj1H94}Wo*v9; z`U}>WLq35Bn`(^lfb^^|XVdF@&`Tk$42$oFsyc6)(Kep7`5&G|H2e)ONDcS;P7DP6 z#V-^d>A*H#&eOx#eVqzdfi5fi%SoNt(;+^(J+Z;oFwt)m1V6)34Ydao>r1M<uxM z($J7OV921OjwIYHEr;rE;v5N5)YaCi4W@-zQ9f}iQjgaHI{lRHgjped`A5Cd{GU`0 zQhf)YkEpE>K4TIleSaI><*+Y#p$Q*Zrdo8zU>)IDA#4x+xTCEE>{w*jG5-OlsLQbUdTrrL;0bK}Vqul)@I3 zxSl1Sia!*+GA5k!BRtsivUUu#wIt}e-IC=gqAHohM@%a%M*a+Qh8K!By; zoF4C9hUC~#PqOAZW;sLpNtPC6sJ1>;9i4DNVYC_UK%Hz@4L{SQD0ft1NL}UIM5U9K z4R~GOcrQ9N%8EMSrj_w7dWC7n)sUH);YyWLzDdduUDpNvq%hWl>^p9it;0xdfnK^O zAd<(0jmW{=&u|{U*;l4`DC=44c1@%shot8=SQ~GBivf9hmVtL!a5J`{>H);Zd2QK) zzDJ??4!5es-e%Ngv#?~jBJ<=`u79sc#BCi5_x8GX1{xvEay}0w(x2IW z=BR)^v?X`OM=Q40jqgnFl#aV}o$!YI~cndCQ2RkU2<-7W_IFzQKeJ2YXxd!-X zi_EZ`C5ec70VY`i!xDC&gCz}BrM;R(^oWBc=yV;~p4QjS4H^k{!OB6-%SKks#Q-On z5!Q}_c^gdjB`xt|)b6kZVMpty^OY7rQm&}m zF>v|m9;6TnLQWkqb&_zU5u`Bt%jDLBiZ)WBq5>kngI}CdNh{?!6Lu~Oi#5>Mt9Lo2 z7@CQFKiAQhGYICH9+@R|AMwIw5qYx)1yqqeh+@-_@@Od#1q#Uxx}Fim6ZY5D1iD&& z{LrqVo||cpNE*gQRr|t4H=&nkoT=GYj{I2`$<)_QxVCw_@qMJ%xk!I-mQD_yBpx-D z9Z12YhASqFiC{UmJ)2GKffTO3elsDX$`4P^NO|6#xH99%pw)n_UonfCiP;R>Uw<{B zry8AE-U$Q*msK*RYwoaWYVAZc*|?ZXe!Qg&+0`}YrZl0oEnfj9+hfAJv%|MT#?FZ+ z>s<7Bk1;ga3ew-Q47cKEV^LPc-vtK5x@pZ8lA@h{Xv)qdIll%%uUwI$JD-vJ_|^<$ z!jrsk2#)QiPK_ZX62-VR80mIGrjU$1%aC}tTW?+VE0Tb`>xXmDA8yx6-ejgm8sfpd zlurrouQ!g)xLahuVWg!vkP)R#oU^F9YAp=kYDXP*U;(%`Y>OC{)i)v&I3zG4=g`1P ze>wUe=uWM@{zFqV&zTuMrc;mf^u^W8>YTc4m0Uz4sh>t6{x2SE|<;~=L z>_F4G=|R11A3OieO_l;#AG=z|&=h>EM|Q^B3tYvas*rBkwb(19ljIQZ6N4>a_A(en z-KSrvvmeMc)O0hx@TLb`io-_I@*OmwaNp}qBVhl9J=#vMZdzXq;~h8j{vp_p5bBJg zeM}iVYo6y#GtTN|Bqb*49V>y#RX?Rgyj_cmgbteCwNet&za<*glIbTJXL0e3MuGnU zQ|jlmXV_i#RVN0l)K5M=Ka!a0{im^ao8M1jqq40(>zt%(d#0K@V^U3y+!I`kcJiWU z9_bux8`g|OJ1eU3MjkH;@kaMnN;6Vl6e@L`E?P}!4y~L*ySddJP#B9yHSq4~ z=c5(dnEPxp&zWV!m=PQxy;ZNx+lke-H=`a$WvK%+rT{h9)bl^2%%NX30zQ-VQj+y@ z57nfxY{n+ujpv1gF0^nvk<5180zm^pu@ZPVAPb0%)N-0)x#$*4(X z2On^ZTm*Pb7TBxsw_u}X)!N82xe6QZSDG8NyRl6#`;s7WpNo#g`H5#YjicZt|A-iM zRHwI+Y1Kw+J}gaZdXVw>Ydw)jwwDlaXQLq9--~vWV}S9zVD4gt8O;2YI(%oI0nl8- z$;vUM7L{M99Mhz1Nd|MsHF9ymUUK0PWQ`xJdAPNbA)KVUFX30VVW?h?EKN={JsE7# zRUQ2}MMLa6#`2evg}ysx-lYj8ecWRDW2^dn^%vydTi?kAVVK~)`gz;F`g#ASy=4DO zAEJ`2rLoPwOhx}`4@tRn=@1)&4EQOhLraC$XNqh-5uvw+-5krDo?2Ahw0cSAjrp}D zgQz8uNWo2m?z`x)XXw%|l71V9>FhROo1>|T@B8cf7fX#h=2u^SjHjAHoRYon2vRZ@ zBX)>0qsc09g?nhq?wctkMg~@r1Xq(W^ei>D;BMiY>kqu{J-N5^PtEJ8ZRh%=b;$&XC~DY zxeb4Sa+Q-Ilt|5y8J5db-fEgS>W+%PrqE&BnGEYDRI(!g>{dmw8|wgC7$x#%rWe`; zbWC4|_<;VJER79o7%8%Nau_*>Pe;{UHr)rU3hfzU zePsog3i$`BKL_FtPlq4%iW@H2$}(67%G@QF^%psrN*@MG70t=trwJGiH5@YXHR&rF z+rzZ0++-@Ya;5UV<1`LA&_e03BUyS}yzbpJ(ajySU$o#DKk+A3a6hcDPK?2kJY$l< zTlqa&+_wfnn8*VFKya{uRwKP2{Y3pJ!fC)Xpbszx7y?8~GQaG;g11kw>=cwwlf^?W ztE{d$PtcaX5BJ#|TF5k{@mh=^m`c;n%#9J`pP_5x|0V+2i}Mo&!-r+Zzjo_I0g;8^ zM!RNqw2IvjPD2v{1Wq8e3Q);R0=NC2!>X^Kx`;{bk^nT;UrZ+qP}9Vs!T1yKkSp zpVNJxKL6JGyVe|Y&hd^HGG9_$3%S|~N?c)Ug;xlk7fLOGdhG?S_G@$&c$u}*jM!Y_ zBE}ygEF1!8r#pEczW-dn|9d_ChO%OA`PS))eb*H8|2wZ&*3jD2&d~;7@(-fCsJk)1-r2(LU&WQ2 zAS=HvfE2PFE<8tagrha44(yMOt5vR3+EVaKz#{%v$UVxB`sB+>>(EELze*$s@4r4v zB37fNiQ{o^4Ba08PB34aKTU79djYL$lpBHFtD@B!VjA@X`6fi0Nv>qo><@)q| z4i^v)nk|Z%ml@F14$|ce*b)z#6Gtz%PG!)x^Fd|rLlnh*F4oN3Gcc={rC|G8ND|eo zLeuS>5BobaJTfl1H08>kxcBloGji}ks}QVvaWzOB1HoEB}de~Fi}De zzqaM6p19l0+9`BL#KIs=o(HJ|chXGn7=GgC#A0ZQy1Wbf>4*>D z3Eiu=z1d2p2-`-v0}JOCiI5#*Y*+5z?z;56i3spS%mHI8|F0-%RQO>3Yp3Zj%@ z?yoU*T!KF4=x9$lwD}O*edl+W_|yjMaV=YGoJrAt#h)FipiZu3hkQ{T6CCJSZPDUZ zGeUi(3_q@T=TAFdBJ0;H7<*1Fb^87X!5bwRMi3I|+qjVM|48swceMC_5@j_!3{aPU ze|63>5_?!nGcgT8j*$I|i@yBT3SWu{1px}Z7uM@Pa<>9yG8%6ZgWFCbEx{(qP9j_H zZvz{=$<-j49=BqefwFe>qwN%**VW5xy~xXM-ZrB%Az|ja6DwN0Df1Pv`|0Sk^ZL!_ z>FIq4Qr{h~$7qRWKhoBE2!Qdp=fbDsFn>b|u75&Wv?#a`tHTc z`hZ?>#BeTe^4p`CT67Q?s|bCFyFr(XkKlPcO?(7NuJwp1%3jCVl+0OUEkE8PIu>WiH*$@1>%^?!u zo|o@EgrK`>#|NC?MwtB@zVE~IVURRs3tj~r*)x+To>C~I9Vw-0e=8A{(bY<7eg>nr z{|@y4JWxy&*PsMXDx~Qf7BXc})-(N{G`g|XcGEVKUT>Q`Z?fKAZpsB{r{$W+AN!Bw zN(l2LoY+B`Db+FXm%u0Bh1(nP;*IGUUu5-elC&X0-+b0yCG1ck@3u#wWk zJnFNJfr}VtqQ*c0aKXiY^o}w)iP&<@|3+WJ-gR8UlhVB= zNrf0PRIE!VaSo28t!bCdUYAQ5^U!8;$d{%fJhWFbd6(jbUU!nRaQ%(n2Rln$HCHQ- zLG~SoFB6=!VO#y)i`n}_14(~45JdNc9}c+$j@_`&!8E5erpd324w$!6 z0O%M|WR;_)HBIn>N4+a~FFM>$hroMaX1f17}sg01X}z;BVFFz!A|M-M4) zC~D@4ERWKQJjsX;PhxRPA+ArnyO#1Q4{qYBcQUPkU^18P*cHcXp6eWfsTNj zm?kSWl7Bh^99$u@KR)fuaAox(xTnN*D6s{RY?!n%bv+j!eV%~cLOjxOOj$)XYuvI) zH|2Y?NYf2>G?m^#K;{5e#DMk7*@~sV!dyHd0d|xX7VS(%6ig6R-iab|uK(u5kHj}u zL|R*&j-Txsl2QiJDw#)JXd9SYZKC2`{EiZ^B&fIfEat#cj*)w7M6^}X>eIli@iydZ z3JO`vE6yax?7Sc2)}>=9wTD#Ny=n;=F3UH$8HiYX4>v@;p=boJCI}wr+O+0uY;@tK z80obd9W?Fp8Jn6@3Lo5wpoOayuWK#iCx16Kw5OQ^yVR4`Nlg$nQ3+q7M^=LfnIuFb z{?LqNd=}fg>7_D_Ry>-5krU`-EgmY6{jft~O?hHULXNS$%%?%yn4hhUb5kLr-fsGc zSJ7zql8_c@q;a#NFM2$Aj%ESWs(Cp|KwXCwj{bG0wLaE`_=_H*JZG28vnx*<&%Zrg zUjK^bl_g16owFRk1DO@X(}$CoIVJIwzlb}T=lw=YT5!-`A^%xX-e4aaWaWnKKHfqzI$W zRAMExUE!A_{^>&l)v=*`eXN;VI;kGpfK0ORMAG>nF1zX2vymM3{hwOD=I1Kle53dx zMq&M9!+iZ;fLuliBKe)jE94cSCgc@i{9{a9q8ukM#}Ph|!bfZv{lN0^V|?M>WQ9wp zh1Am`cwv^1ouXT+3wH!6x}L~sL$bqjDjlMjWd;I$VdAYGC2zdY6O+;@^iT&dBx*Xm zenl;ez|BAX2~nL)ya@sWSA#wfJ1n{f>g87AitgB3NH~J=F^*|J8u!Z6x`t|#`z80? z2>08AF1>E{Ase2A!ETDcr5nQwwait9jY1FHP=CGMdO`eZga9%tF#^j@4L0S(XkPi? z#C2a6TC`AdfoWGr9=_gB6D$|>4qSMr;+f7U$TYgg6BJuS#5Cz}C*()E79k{<$-+?v zl7u1D<4MB^a|1y-nt(k(yaQDv12Pn~ui2ygwP6FqnT-*~ge+l!PY8#&3zoH)svh6_ zVn@YUPjB*cfh9aDjk*J7V7B;!t4#3N(9l|J+9gC>CPI1ue+On2FPnRqVZRVMPUweH z6;HvZJ&e8Fbf-rTD{@O=H|R#bM-M0Z^oK`J<%fbj$~8E=lG~hherr(Ai!i#e(3j;Q zg_Q8Vh>tU}HLqa@I$aa6-VzZe*h6`duJn!Kn2b!tCzSk*4Yl$aFMJVyJ}>0tTiEFY zz_q*c4o>R*F@l+AI?3h5G&=VkX1Amz!m{fYo2s26|JgLw2w}U(;#($Ha;~~^*_qTB zHljJEZd`H%q5=S~A)#!7m-`VCYa>p&9^Pt=p|xUbZRBOoeLjTHgyB}_BUD3%mrc3C zr(EOX9-)??;S(j&5I@v@!*aPQx~5>zE_Yx@HHw_o=G{LTiv;j(_g7duoql}PoPfFo z-?qQ#4oXL^ZGxV4UBCXhO4Sr|%(D4Iv7`b+wF2>P#WspDJlO%nd+jf8l?@Zs2WFQy zhzM@D)QUYBc^s`AD|21K+NULcyE~daS5we9%>ZkDbB8Lbz?|MeX&t$Yx+-QVnOzP= zU6k@;0y4=M4H60#Qi_2bC}>PRe|hX|r#Zsbg$hab@N^RI!e5b$As*^v(xQvGeS3(k zCge4cBe?`3$lM70USUZu47NtGaY6)xD3592b4sXlN&yM9NC+zt5j>oxMQ$nn^%>vbIj*^~jJgpa-Za%&N5nRMH8l_~8aPnc26j%)8W< zy|L)Yi+7GBLM)8>kn7UCgX&xp^mWLzJLG&Fg5USXgN~`uX7ojW=?)6HP8y|ZMe4e@ z?_?UJ-9r@QYj$cZDmuVKRNG=AZ0G5~dH(VXe7KXPxX>Q>T30?JHasyVE5KV|-25ZI z#S?yukCj0N^s3Ap;VjTw3a84C&ar9E|I^+sVb)hh>b_zguLKzW$mQspdxAeK3z!4fSp`N)_yd8uDq`IM=Xd$?`p3y*=%LCr%x{ zpwSGA?@4%uRjr#W))*m9T^#g>dtH*Gr2Sfp@Te!ptw<&f@s*QD1 z%|>X{^8Nb9+N3`$v+P)D?@pqtSmkH5H^Bxd=V8)bMdbl5z?R7=Y88R_uhbF5$ElcieD-Z1o0Iusyq1(I7Il>J~*O;PjOvT%$ zgSAdKOD|cG>ZA1sRwejDv<)~=*}?10w&4R>2YH7fdm(;cwYlEMb zY=)hDjF)aq2+;e9kyx9^ab8Yl-9Nri!+F}MX>&1||MVa_$8GmWmlosMOTnTy#$P=H ztCOelrr05MY^_3wKNPllK^xFsqdN5M76~dBKg%|YVSHe-T$)PJ76Y<&*BROOM3H>7 z;FVV^MC?JfDrQ7>KPl zc6ACfMVcF>CJMZe>b~FW{r9dq`28Uc4S54niW~)cEu`>;N(58|z7vT?8Webh1CaL* z^$1e@^2cHll)Ma;nDoCf{ftebP;vLof@Q2}!|sd*bRw)?CP>YS_-nJzXbPUo-uCEd7%(W2z`NcYh-zbJ4f)Qp@yo8un5`MRy!_Kn&m(XY4bMfygTE`raM|@H6ehd*!-^=EI2bGkI zg>fvuK_!!KkCOibNULmYXAk%%NE@6mF4rf37VNh^Ts=5>SW7XGCx##)E3~90 zpJ3JNtH0K=p;DZ55VSu@D{|&uAp09sluJ0)?RH?p7T&}qQ$PW z4Epz{Jaljy_5+9q*4faU`Ed!co+z@k2HC~C8=RrqaHYh6&A6S5N!$*mv4mMKOx=jT z_*^x*#}RZO`EOVVc}5KQY*RWco_-!H``;c@J03^iff%g;*@4UGRWy2yqj+6`TXQ32 zpG>7XgV_9${Xk2#xnGFZMWvPh;U**29GUj;|I4cDU)D0w%DQ&*3~0Q1(upBa$xK`j zfwl5$$+Xmaz?4w3R5&Uy=t{HrWF!ZLdOr?^C>#Xesdo$s@t=X;DF$S8><2hI~M&A=I#>!xlI4J=Fq^lI5{*oW z-E+l~^|Yk%9y!b(^zeo?o|c$zdxSEDOJpn;1@X=cA0w$$L{Hzg6{_1}SmzLI*4BH# zq#kDZF(Ll6II=t}H8#eO2>oklp(YwH^kY4Aff>uOlw&K7^F_2eN}v$LNS-PiFP^sW z)#Q7iv8CWPLZfW*P_HE769wd{RS7R4TuE}S`eBnR*IhSUovEkjZ7px~d$_Rus+Bx{ zzn!{@OzZiuQZl?ELK+@jPYu%~y0NeU8h^Z3L86mJPt0vk z9g$V#3Z;~Vo^gC~<|+?K-NP|2_93Em?~Y+f>C@eGOuAbVU!;*|ZT9e7V%RhHFV_e( zrBP<(E0X5KgFI1b;H-B4d!b`m+b?;4FP-)8LMQ!i(Bprg#gg*>ffonIPyVCMgQpLx zBmR`^Ai}^li{B6|v-Esr$O|hRP6eSp@o7w8*bnPa` zoTGnD02EW3D25_RtF6@pp9Gf#-74@JW8Xf9h`a0Y1?Cfi5Vp%QJ-o?lCl#u##1hNNk6@!Ak}D`}ZUj6V@pqRF)< z)=>Wf|L?!g<(+3k4;~08iRiz_d>sL%)&OH?`hOU%{g-K4qXx9Q$|CxwJ`=N@=^cuM z!7sz`a4=a@2qnQlh$s*t9YO`r-j>FGhM-v!vI!ZW#!9VOE$xPkr4`Xg*v7(2E2zBy z8(quI<|fN>rE=X$t4NjHuhwe-lXhRhlK0D%*Y%duH0M+17mXYKs@rlWkf|lW7B@O) z)IJiY-l06Q-^O+$hWp@BPV*$O`O@~mI5ei+#HJXhutj^Am+Q)9kB@zjAMeN4jnnbeX^Wr9w56Cb62xJl%g!Ofqn?JcFvn{KMwbCAmP!`n=JpfOBE+L) z@PmymHXC+TQChg+A=;j8YQg*Cf#qDB%C#InU|gHP_mN+k$ZSR%H%Nri>)3^xMBoNev+Id>WANYx3LX_zNDW%Y+@=z$_Y>@)`64prdOY|%=8mih~5ia%Wun=}vr@@0(2K-p$ z(a4a8S&P+hdg4Up{E+>GOEiK~PS6QW8>eO_*iiP6zRfygNsf3ZSEQP)Os5G9-u}L+ z#C~D>(tQ|saw7KTg)(nF{kwNnK*w(^K<{e>hPUAx_6Z=; z?W6qnXQ=yy@a(Xk$+z+X?{mg`*4>AD+BpLMvWyp6<3pNNq}c8Xlo@?Blk{ydt|^iZ zRZ9JxW)|su(r|(u*@D;_?uN^R?$Qms9$7bYm&S&B+V?0n92xK9f`=q%;YS&(hxbO$ zfX|F!hI`iz;eGDu(W_ndyU7h1+a2%lkF~Fi@qwu7*M@ksVz?pIu{2MoXqTWykxo2M zZi!IC{HdtC-dmO@!;){g+e6N%h^{@a0Wcp-kKWY(6G6LMmTjUcPP01l6d|yfuCc*V zVQC^6alNmV+gxuBpWGzM4ar>wDwKS*HDLJ>eN!j(Zjq)L7o%x&u94o?r(?DDmodLc10ZNywNHW-lt9BX&dfYhu-0an zzJD92ROjiP#m>guT2q!~X*x@j;COzO8o|~&ibzXWIi8|z4Ex031W?g|Ilm%8h?y`x z1+mF7ZAKkvwF)5Wl@zscEf>h=4A|V@K(bsVouj-H)uKR2+p}Pjcy@A9>op6Z+gFeE;&Qun@b)ZFv83 z^{bNLk$YF6+GplE=w6L<7c6Gq;MkqTgVhl2P`04MBFXiq8GItKT%QqNs>wuyz-QVQ z&C9W}FLI8HR981am9G6$et`jcdTHNn$<72COm$%x>SDT)WhtSZ5g4uPZ!T%9;sqzA zNGA0JC&rx6IfNmjuMdh#(NnpkM)jz*nAxWnw_I;Y14t^&Oy<&kT9U@XQb(p?wL1N& zYc`oasY0pb(S-8h8;jxI0%TP($}w~_-jXDR#@kjU$|9dI2d<42a(O{ctDVx6vU>@7 z?u&+~{gfnGT+6!&7WNgV4TEMm+dAH8Yt2Slw}SdcKs`LhfLdaV5GOcKa%V_?NWl4)xMU3&u>VL;J87DMp9lG%A~63T)8^G`Y)oD$mY@b zlA#s76`M2R7q=wLxR&~-UgOD>C!O~%;he;{Wr#5t%pY=t+fcbz49WQL>&cW^Fe%#;gns7MD-F7|qH#9O(8C(n zaYH(Xl>;8Wb6M1Dsc<(#+U+@?BdHgerJtI7+|G_W%$>`h0BvuDjx#x3`mcqaBTF*!lZ`Ec@QldY&O4|mY0}#<4%Rwo04A4<`jAm z{}{uL+4J7-!(+zcC13$(p^kW`>g)%rNrCf{7>Y1w4j&dUbONH!29X5$^ zyFbU@Gr{0>=C}C(s(d1>d_+R%>#tys?F${oq==nYmapmJm0aK*#C>JCKVTNwsF&E` ztQfwBDD zb)uU92vnJoE>f~b!LLYM;|@l-Axsq(xiJJyD|4Szqv5-vB2ufoUzpbTW&8eBIKv40 zW0ZY_4Dkh%{eboQ0diGC-2gQuBTH96R`d=_xt;V^bd0H0H)(H4%r>A*lft2euTP^- zQv8LF@@Iy#PK6IxY=G9^1>PI)UkN^SU6S);N0dxF>aeNI4jx6XF` zk!~xU9ph=5)aP9E~Y4UiQ0B}Ym2q^Sth+x zVdw@@4gl&kO?j;+|fx+b8Y!Tp`jI}dsRWyaz zI1#vQ*fWOYabvYQaiCFb?TlhBkFh|7IRlIxlA}a!LX@+p)zN2eMnacmtpu-@N%A_D z7|WS@%NTa1XR5JNE4$RAY!5y$jkyA}R!yInew$IJ<`B`4#?((8uz2m~*J-g-a$Ran zWg2tR9?|Ix_nEah{&*Tsw<0%fu7|ek5dG~K2e84GPIXCTnCui~Y}+zwU?9kM-aM+#h?f*9+=R3hh3r7`GHv=q01{I~RHI7X|3lw~AN!UuTL9MJ3 zw;`NR3sE>NB-apvJaNvFm-mRm|bWZ0J_&X2puVQtDuTC>&&S>D)k&BCu zYro3BUsl&}+g~3y+&~yNr4d9MN;fGHKRPOnkgh4*)kUdUz^$mzHue%Z!Ovlw(TD9b z4{Q1eqK-uFO=^Sh5cEOHC9s>;Hor z{*u|4Jf*GH^{4n=bX>;l+U{I^snJHfxfsjS>S->0gMZF?UxtmFSGrx>&AR_P?W8p1 zB+vMKP3;vTTHQU4m$$ogb8!xQa-ukOiS2JD^Q3C2)05+Cld%R_Ej%Eqqm+|0!!ZVr zl>Bx_at$|9o+(qVl_DhBHUhgw3PHGnFAZ&T5r$N8KEwx%_+SmLI7T2^APi5n#Xo^$ z(>4VrILb9+My_ZG;dI{oJ}fXKQ}X4cGJTAr$?%9W4%sqR{NgoX8C(k%Lv$emZT&nL zPf9nM9BhBt{xF}m)1E&n%8zKo? zq$W{~b>#U9G)By|hhl|*TmVtUtYdB2o`EJzIsCp6&ow8v8~(W7)Og6{+F%rGI;ss< zzgEqZL}L*-OO*1^=Y>5BmW#bT()4l%Wec<9&`tOBLW+Fa59!GZ$#Nb{Ut`}ROWEWz zKfdVRyV&G%fJD}0sw_X!FAa!TrZ20}b( zx}!EnXPN#oy#Lt}ec(?J%ou6~|4Ic+5PFMJMey4Kuul#bC`%>1S#C$KTDMzw#O_iz zE4gdNRr-QAj?}~8(-c067d2zH0^dU!HFxhkIe|qu z(sFoOXc=`B(2!@8DMrg7Z6|dA7#s1&pJA4m*1)1IH+w?Xz|F~A5v~bX6_np$4!YsY zdloW?ZTBK!BakJcoiXMJy$?dkOKpI4&wO+=?PS>>%89={z6nBXCJ$d%Zjm6KvWaoM z&&*}gQZGfaX?F>B*H8^p5GE2rnUeL99P!mE`cxZ}OUbRc%J(t!-n+mT(kHId7^+2^r~0!4wT!cd?f`5T0l1f5q}dOBWNM^qmfdssSYRxl#3 zDvvNc?2;~Nww0?k_L*2QjiZy8VC0mv8>kz(bX#(Ypy7_Z?KYLXFl(zNXIK^FG$oH? zGl-muNL00egoiSmigKYQYVYq0xFh+dC?A4*yvQsm$>#6GygY!jgPUt&8SjINch8p< z5G>(9K!D>M{L_i311P=nD-v?ra6m=td29)Zmb{^&f<^)wQ^OEIbZGqK5clenyA{rL!1#u(;?6;uW_Z+3v^G_6TjJr-*m6b6hC5O3jX8Uh=MLd+S zI@GY*L_OHn$K{|$4OLi3o^()+Au0GtU`b^*+tRs~g^fR<4H`@DjGpr8tl7B&{n7OwxAbtJ26JL3#ueGVmbJbL=X0*?2%T&;R*U*$jL!@J3Z7t|p4`;u*wH zz0&~0kFRp$k-TK4vamgSW5f3r6LEf{%jYE|igx^RA&Jg?lLP}V@APiQ zu~V-z>hyl!5dbcyc@TL``P>zNf!0y71DW{}3ZrKPmAN#34aLOG^`4S>c!oK<#?)AHT&gHqgUJkFWc=`ZBE8;pi^nF;$S>2Gp~1@>aAJ>ROpV^O zU^>vTfzhf@#F{*v{=0xS8THxenqUDXu7Mo{WuB1Zd0@NLnWYL4p z7@e-zO5+gdy(gEtBFP=*jWa`*FB<3%axiM&gsJFcc?X`w325;;9zvg7tAcdn#cA|Z zXOtz1n2h9M%AA`yQ&>xg(s>~$tIdC*mS#e;L_N7!}KigDD&7`JsH?3-`Tn7 z6_(HFSMxVN=&YVuua<9cIx2UYTk?paHQ}#!+^|tH4|U(ObK5P~Sl)rJEg0+rV9L7|m_WSl24?K*=~w8`tm`o<>Wa@9hpG^7in z21=o+`5AMHPHdj2%&XP(Q=20+7x!ENxXpGDHZf+i7NXfc&TSPu)(U4one+T5a502y zYgwR9_Qp_uBV>67xIu*P!BkpfmI1}X%EN8R&n3}7E7F5qTzK5w+a~UNB!VnBU+F-XP>)& z!*L5Obj~w|;u0{ai$pmO+{XO?o#B^`Ef%K72E>Z7#bdrCdat`8yUp8xEeihu9KDNP zV{d}aYS;I~g=}e0?31yRZ9JQB-w=O*z2 zzYCy%HQ!@US7?uF15rFv&vg{Z8RrItk&AOHg7pX#!AI?=hz@J61Q-byhHJ{Bp8Ce! z&EqBv(KdB7{c=g2b#JyQZ3-xgRJauFPcSZr_n89k90cSCq(>9O@MdF?E!8QPH{2`A zbDwr4-PJ2~<$x@A{DQ#NN||j_!CGx1weTEWBLsQy3Ao^jHRX{!FP@1s(}8)@U3ycT z-;i3ER&ZG1E&!B;7EQ_jfK&Q_VL8}~{M zg^IyfD>mEkE6r7T#OA(%UDd$ncw-HJ=v%qb6W5AQ@`_*QkEU;jA>-?W+p1#)6%uW# z>8ouWjlWZ?vK%q`KUHSiJmNcJ*HW^D+|ji$N;cO^xfkZ=>s~g9V=+unwFMOL>r}0& z*L8*-6Uf&kacQS|?SBUGAQWhL)e)ZJ9f`b~^h`nA&zz&TON*+XCBsl%MDeIH?On(fulc zl5Igt?{s$WeTS+6LMtL1Ihz?&aju8320YloS@j7nMR99V%grVy7~^q_2e@augxg;h zDIC6T;C0rF_xOU)LYQywO*k4n_zk1LWt1!* zs;52)nVFu0s)P9cALIC{$MAN7?|JG27!VNqf79_jEc~SLTOsNCs!RzO(2R+%ik~%mK-+^So^0Aok@`%BuBqa*{kp-3ci}zMi1|C%_ zZWKHmd^>CW!P^$}VF?kWB83N=1jR$Z=ABQS$9&J~uK2I;5$=InX_{nERtb<>vN_et zPwsJAys(&~MmVEnF{TZS)mkp5U5yoi^*&V<+9bu=8mMt0_zbcpnwA+^hNpOj1$woi zop(PAxT{P?tcct{gak!iwsrE}e-*0i;7NN}cxkR~h;746wVDY&)`y~k&z zSyG|kI|f|RQ=LK}q5*`MrOBd0v(u;oW+vk8 z55=hIO_K>SF)O|cUb82jjMBbqH)AONe~IeK`p5PmZ|6OVY|-b_A3zKB|l=`adKeim{3% zE#n=6;eae|46GH4q->*xrFIsXXrYUbif09eJXopqAEwF-*3-I8-{F01ePJ&DhDzg> zlI*_Fx@TBS75XtoQZ6U0 zDL72dpEbU0Igs>)_1LXm1{sECk;TSl{_U*K)kWKkcWG+qEof&UezBKQr(1nlvQ!24 z*jeOnC&ME$IKbUvw1+~Mq0v+2FHVNPukC+fafsT&^o!?I!IzM@Pa4#a_2%@+ci15# zLEFK0-Kb~MdIx~F^cp;o{){01pa}R|H!N3^qpGLowju(*eLlhwI|%Iw*n{N3lQXhz z2cttI%*fQ_pO}5()%U7&VkE~m@NbAHng8HZD{TBT%)9jRZL3!AWwenm_yhM3>P{ZG$hRHQ4S=4Z$ekv(#4 zD`#gQ5G6)JcNu05c!%_P%Q6%T#Lm_>Rw zLQp2#s1-7SK|B*S2?u^2cAnBRe6112tv*$Cd`WkRB7d(8*{Z0NbX(eF+Ce*qGEeU+ zdvQQp+W9_{i)E%CDnxc4O_6Zhggc8*gj3SV3cpaoG4Q^khzG+hvm&3`%x;#w$N}S- zJJykFsG8A$p9JlvP1-byI#-mPzr*WooJa)i z?E$tXhDO$ae+vFcgk2mRzmu~6*V{h}zA#}s@;j_?xO<&tRF~ef(eh)qOtrj7H7df6 zJXBl&D(OCVIDBKCO?NKIS(!cDFBuW0&%f^j_)~Gu8hg4P>L=gDl!+0aZ7amskU`FlDP5ae-66#ToIWF;B< z7*f!`Q!PSL^BHvA71PDSZ}Mo8^=P0Id2WYVQM}{-E|b2Qp}U$th>TF=B;(*G)pqsH1n@Clos3Y4Q`Z6qZSv;|LbGvxynZ;$WD; zat-=(6(`=Mnm~vqTx)|w*RMh`(%L9z$B~*hKb%jpcTXk^qRGY2`vW$7CZ3ODU$`c( zgL&9o`t3KIX#5ZPA7Qzmom`tPTn}UvdvV9{i=0mGaF5}c*i*AM$-UoZ#g`Q@HPO;0 zSakWiz4BBm*uG+y2mRQc9Kzbg59m6TCwOI5@r|sT9$hi?(jA>trS_M){eSChAUCDV zbD#pP@2R7j3pJXVdBr~V(~Z=q9@QF)BV;!y&>P(nd3YBp6E1Q^n?&_c>F5-uNUndf z{ikL$VbKTs;Co2{{|~P!?EgA8ly$BC5#E?sAc?AuV^vA;H-C*{6{jM9d|~+TSAKjNvGX3D|g&mqCjSKVo^Y}2!S(&f!i9FO;Qf69mbXk9QNeh3wLeW942-J8Hpaa9z^Oe-;d)b+ z`X15cl?CU;MYFW!OYS`1VGT4aLjyrSQ^X_AN!IWQ!d4EG@P3n@!ow%Bf=XvteHj z`vd1()vM@M?ShgiCPX|mk2~;{0jJq%h8f7ULrmBBd`6>|kWjsIFYQj@nq)Czi3R66rUqGt|q)HrYEh{oQBFI77A{%(*?%E_F&- zxaC7q2MKX#*iEg?!~@kkLay1cSRto96m8*Cm&6zzr}CAqOqNM8vYBzF@Y5T}AR6e?Jq>XJ)%|wy z9|_O~{OkuGXS8E!W|yrtv0huirEfv+*JIbfz5)e^nEw1z%@K1-yCRlv=*YcDV0!cu zHPXFE5xr$2ORfL7`10MQ%5E;iViN5N`zfl!hNJO)?MN5o7?^1MRCU1C&fjE~ol3i0 zj8`C=5T*V~yJ~*srm5#afN+G!|719uwAABJAr;=@VA8ceMHQi4wj1%TCudipy2w%> z7N+2H6@7Kg)z=W*#S|&5At_Bxe){2?DKa^RYuo5Ijw0H8syYFXHj=z@`{XS$={aDR+!}&oK(!Ho#qI|gSKZCk@Z$F^M#l^+c1xF*x{o`mS0uX$`Tmb8n zMT0;DJuG&eO1SY@UjQi{0ki#DUjivNV|t%bH~ymMD4PUn01CJ~HYPjcKz!SDx35br zIVL;0>8p*cm#+GS;FSA-l52!~DAvZZTn#MxJqz-=x2;x=UUQZdE-MSQ6hbnh@P6<+(t z9;)bbM9itKWa2T8E`5w9z9s5{FA`LxuzQ)lx}6DHATBOG8B8fMGF~M#<&=V@8hYxh zt&S-xBuvoKF^TzV$(jNGSPiWV<%noU8NTZ1+Wb184aiqaGKTR20`OSQlRtG{X$Z!O z2BPg*@v$eXdRpzYOLD!|hD$>h*-Z7o+vF=o-4XcFOc&a3@ERU{m4SV2&th6(7C&*h z?p`I>q#FYo$;s_wjo$Y)>5<13gWuLU|pjODQA_pxM)n>Ps_cV zS15^UobJ`OvG3N|bnoy}Y{g0Y=c$M`OtY0HAKA(~}sQ`+nAl20S$ zg^xU4?^DWNiy4LwVF&T8u&iZZAKG@CGviG{KVj&wj~KJHh%MD2vB;BuBhuMUl+X=^ z9`VZO$BA2DnA;Mj%AXfNm>0w6Ip7rbLg1V^qd0JFw9ap2;JL%4Za-t!U+`||i$`Wf zv?LMV0CkIYk{rle_p=l3H_3?qcI{wB=lT#1(y^L0H6VJzx;4XG-J%%!{S%FEw11$_ ztVuw`au;Recc1$$e#H6pyn_$^cyl>dy0#a5Wkl9sriZidM=8km3cAs@ta(6~UznaekVH~q#416pHee?ao670t5^o_ z7wbu)sAx(W6!^F0mTLu57+P$#RgUnI27w=d@)}lh2K+!j}Hd$jth1s z_iosq`i0-F@ljtAq2DgMFFr9}yG1|W1Va%_UJAwi9F?hj4)fy9Py${BAaV*6LaxlP zb6QyT*c(Tu1^a;bNSS90%bRsdkj}ud+^}mIzz+$4T*m7qMH&Cz4VZeXd;M80}SO2;A zW4NFpMx4Kv7%Rw&AIXAN3u%M}Paio9qEVW^bq-Q5Xhs=O$PmJ8+nHLe)tc{e9=hT( z*+Q?~>#IbQWG1sX(^n&?CqcNNj{#$I4GD(g;%H2-a@A!|nyuLfD6d%`1Lz^!z%Ze) zhLGeRU(ML$uZ8xtz+a%Uivth<395!BHqINVG5_HXr|fZiWg{#X zKg&gPm>~}~G;>u`qiF5W#Ipb%YqJReaioh7yqRDNpkK_Kap%ZO{E>qd9JtK%(B>> zarC|m=Gam<40jn{x=NuV#7-4kDw4Ows?NI3RS`a`o^C$1XE8B6gyeF+f^P3gl5idm zwC`a)3B}jwVNPuuxG2mb17p>8Gs4R$>}tt6^a}*^l`+n~4$eX(Kgf8B_dFhLi%Sm} z%b&7T&-9Y{JV7gN$`!03Oa6*Nj#Xn9i1!uNKJ?|42jmtnadTyaJ?y7a(JZZJnHJcd zeMyvzbvSmU*P7Ep=mGJV|0_kZS=qlj&AK=HPzN%{+9W&>}LDtsNx@~k91Ll_J<)2lgoq{In}hZFgm zkDNmK%qu64ey!3;$m~`K{PHpGxmL^+T8vnG;mCIcsIba8?JV1IzvtY}TY>R~7(a zMN!CxgR_MMt-p>dR}~e)GZ;~KIB~b<1>d!BaUY0x+&Ddtd>IQSGC8;psEIyVac-|w zp&=S5Eq2hu6#~1A96>CyLTZ%`Ht7t?rx*D<0?%Vy&BVtu5PjC4Ts(wmGfO@F6H8@y zjIML$kr#!EtkW~yG%oUMSEE%v>l8V`FmE)4`K6#3<4JJS+llvkQ@w0lBa5?XQpF1=%nXdo*0h5n&*IJL3+Q&q* zBE_oi!jKg%q6L;HOTTt#uz)NX(Q(ve2c+j!+33sVtw(;dTrbGr*2KGAUITHTNtnk4 zdh!OnyZTL4hPxC~AHP4K! zXj}&M2n9UIm@J!}E@Ob9xWk)0>E-d$efOPRCc!qa2CTws{>G(P^$E!(F4GY#}T6asN(zRd7Z`ujktC~HA z(&hlCD|}t#@gic)V#U@<^UoP+Dr%NDVDqD$G)%CjUwT?V?19^7Ri3VV6bZcssn z@`G~VI4@Oln`^eZ&nciDx9I6_2Gq4>lg)u|ZUo}pZ1nTT-=)lf=NdwivQO(2Qt2W60!G(Et7`+Uq0_f)lGot^ zq@k2l*@3DNa=e3BGpB*2dn>G|(AEK_+YUd6YJX&T_c*@>s!E$V+En)|GVhxN@YpRF zs}F=ukS7I1Ym1!p9-EW}WgLM$$`ku4|N2-HUZTX<^1fQ51b~`p($1zHS z>}@uvNr6K!v{=S{f9uPONGS7Vfvk$bk+eZ*>dK45mV$}zJ!a%xctLVxruV0+?OExq z533M74ZNbQQ9ya~Kr)KXAK;a9pOlM`6I`M&j5!}ws1SK;KA|MU$t0M{H^UkJOl?j$ zlr;xw0yTod%o7#(=5$@m+91!`usjY$B<(?E=in1TIF&PX(XUHY1wWztQ+W4jJ;g`g z!(^`nk*gpjeE$i2+445j)>@TFK8QBbIwyMt4To|6s)Abo$g-Ee<)<9!;Y@ud{#hG? zn*KEdH6&lq^)gzR&mxV3%9+H1!3DhaszBPwYLxcBPkJtu*6T8ZI!D@3T_ zY2TyQX3GnT-@tdTDLz)qWo-{HYNAUO^}jx-df*xrz*ZJWSCsp;sWtijgi6>h;j_{H zhLrR|UZ9V64VY4i_aL{;Ppx&m3sVdtn1+arB#njcS0dspsl>rVp!wlqsYL;joh6)y zYV+erua0c~pzf7pk4u3tI#(2kPweQyCR0@50979R&*B*;U4dD7h&&-<9EtZYk+NLT zGCl+ZcO;>EWb04nPqJ?aP4Y)<4p1M(89B{rc=SKF2}*MHvhEX6=^lwT6KRjayhW#Eb8}K zKPU|XQ!}@siVEbcdIhD#W&7fFvRs#P8yNKyRrK$F9BXg28oL71S7sRllq#>5*zveN zTLkpDRtCwS$Q$Ob`(NMf7X%1}d<0R>L^#I@=5arv;O~(ByhHeZ!ctwb+JubROreHa ziLD8Du1T*t(Cf`&`^4l(GtdN7|KK1fu=7{O%2DtO$$)FaSQvbOT$&e0vp;I90DV8L zoz^!acsh_SecwHFD6@-O*T>>OD7L*~vui1@Fz&ks>srgpzH6E-Q!E^uA~PXKfIEWZ zKgNECLd06XhrGDgT;Pvn6XKYbBnHDeGHxc=uo6VOdz)(#3j1WGsP1KEqDhMnMWoqh zrVe}8xO2~pw(CgCN=>xG7M_T=!~T?$<|jdzT;EyJQbu=n6Qjz#kC&N>{ef4qS92sD zNh_%)PUSS=0t4|Vstd+n>OWhq8tSp(H+g1W@{OK9!6|qSCEA_;OCmu0N5lnPhtYx7c6b!ogY+>8N`hXn+NC(H>!iE!3DGrhS*^Uu=|P8d0d z#4ZjTwzAb!YkFXIBr{u@+u~2KRf36lF*@Qw6@-yp&?MVpx&W z3ia{79Q(c3TnJ%a$zi~!eHE`o#FSa9(}XGUG!fp^5~{))sE;+R_*M>@NFgkij<40C zin*G`&+j$gE!MGnM?E#+W3zn+YZEPl3_Qp?Onaw}HTl$E&Tw()*DQ$!vUTI8lcGM3 zg!r1h9N#TGY7JhWo%_hnsl27K}&IWc!Scj8n z1kr#drhz8pfUB?2ge}ZI$&eXt$spg=u?MrtlNrQL4Pqah!CpB9S=@1^4-u<(FPlO) zEkLjHCb?vZcBFV-r6Ygt4CgEjB6WloI+M7~<9VZPPcnJNA3Q)FL$qh%pXfc1J}U}b zmteeFdL}*3@pPr_99W%-d@<)7Tdhy5I`I32jNWOqCqFxaT_22lqs;DSwFBSZzo6^) zwcaWB#H`+2b|qZzPQJ*)-|OKXbb4@HpPawqXKzJ)Umb+K0Q;u*=6iqmJS-c}MF~XN z4n+#{V|A?Cl1BD(z=oRSg{3vxhCA&T26 zey#B4iVwFP*tXrel;~}e+x9^^KUv=9xTW+1{!ACY_TPzeSKi3)1$nR&(%f~O(pPUR zuA7E>P*yLBLVu&*weCuw7xi!{?oudRSW++FxT}DdYr%eHWj#`b zIoRJ?!m9X|b_VM+2wSE|erHaMWM-bRl60V4tv6ak{KblY`Q~E++_$2a{ptWNMT7_CeJ?VWt4t79+OWHci z)1%*=YJCv56CVV)-$ZFbZ}-!CE7vB@zGtZuCqTPrD#|l6=M~7Rkmf2ITu6eFI7n83 z=~XNH7En)A-p3$3(8`4$%lTO0xns>|$GXDjP3V?gYIi#bBl`WwW!t zC)WKU19YY(g!aP3>SNzLCC8`K42D2V@frfE%&86!U`dOe#iWmq4==n(t zusk;RTZc`6J6~gL+m(2*+A9<{75O$DT<6DkrOtz;m^j~#uh!;#u|0|y76%UC3=pj-h=YF9hEFd9NJUj|o`MTN~kA8z^UrPh5j11Z5)) z7Q%-yvKYaV5?m;f>L?=BOLUeYv`M1>yRKdcgyWPFSrFb>6cKom`40Yq%5Hj|juuZO z(~=zG_nq`{>wVsK8~uK{*a!piyeSXQk&NUFYHWtw#U;o^yd~`!2rJ!Mi!R0LnYf7# z*XEtP$%f>}MT5Gu@_yka=;^36boGybH3SD!1Sdqh+^;?xe0|7 zFz6Y-apLH>g{iq`!15(4h#UNkJU@HVDt6Jwv$bHr@)5e%3R2`#QE^J5#D%Z15Auh^ zl9#BG$V61*Ft`Mz%=V$lTl|RvC85kuHWv?75hprWjVoaZ4cNH76xazTWx38kqoh!p za_^V>PjxI@I145}&BXzwVS-qz3O|FIJbycL`*8QpLq*AmQ8qn61AIAAIm!-d37Zv& zX|96uL3FJ8GAPj{$)A$`qn(nEv4bGlBpb;C)sbZr?#!;r^Av|kN`da;&GEQ=wC6}Y zt^PAZ<&dv=oo&=4*jNP=A^?$<#3jv~$+&F8+pX(QU7wMd3TPTAV2Cac)b$VYCM+?$ z3rgFGK?(&F+Zwdy2-af%79PV{J#if1nBgkqKIhkq;DDw05Wpc}P6t=FNxuJMake1p zB%LpNf|+ziYSO&5$`9uhvlr^9(VjJH)d(s};bRN(U3hQY^mIl?kLjQ1AUah^a%3V_ zMDz*n9@3>2sbDCM=Ap=wV;rHsnYw6pZ~KM>Q~Ajscjxf3ROEMx*tCkN&D<~`hd+0^ z+R`#ouN2vPUBDdaFbovksZHuenhlwqEoG&$7fPP8VW=8;+k#hnSSe`yVS6#`3(`ek zAi9yin@v8_vO>9bc=zeF1+Nw+Po%Tjin=-(w?#?_ln?Hjm~ms2iP&ge$eIj=cT+ zbqZCYM(4`#`KPAeaQG03(#qxjGJasJ&qwA8V)}w+rV46I$xANODA_yOv6E)RMqj9y zYbG*VM1~?zP~Zo-4GjI=U=2mW>f5Tc_fZSJT4~163F5h`B~!75z7F&8HyR%?zaMZ? z#xBc)%UY3%4~VytZ#c_#UnWUs>o=>~m-5hS_CdVe^lQMX^zQnLR+c7lmUf`EJv@)f ztK~3sfIoN0@Ds^;Ay?8a9LlK3v`Sv5xfCV~cGO6lg3g{?#NbUk0=5sQ{aASJgg}12 zPHU9UMka3CGfjISswZ7_rYNeeq?Bi4K6=KXnWITRsw$J(zH$EaUQx`8Q)a+VGE26; zITqd)1}C~CP-f5!;7MIcaH^)DlS1{WOKj@yKtJdf>1i%{Yq>je0xtX&wZSEX4NsQn zm>d-PXrJij^8mN8f61xN%ixA?GjYctmtWf=aXkUO7%@C8%mK0n)d!4V*GOo%wsu*- z_o&}gL$ro}LAD0k1T%k;*WS;8(C-EoJ6U$;q91zC2co_w!|J(PR3A5@-&T+Jcsj}- zSj7%?ONb8=okC@wyn;MU4oIsWz^op*t)tR($6vNb@@{d0Up8i4ihif>Uw1*95{ikj z>RUhT3g`e^l7Gw0*1Jbk`S6bN(cf0J$?iPs%A6@qwwvTlg-U5Q7GQhY@L#QJk zd7j!@EtV`i>CttY74}2TL7tpE?V#OzpAs8aU?K67g3zWhQi8f9j zAqn3N-U|yV!D}Hft^aicw$CGCCe)xo`T@1CEXEv*>76~)O<+)>#!oBZ< zI(4&$C?snehscpMTF4%E@Xu$|`mZxe9E#{R^oiLG(-`a{Y|W;oS@a~9{Ej$YjJmZu zwqmap*)YXWXXNqIOa+3XKv-yf4VOFpfE-d;!BHewRrW)|tVZ!ja5B=L$2CDHBkq=U zMcuz}(~V=q9#_M;LEX4%+G8ZxTTaAUbUr3511*INqEP9)9C95rGG)FMWw}+*(cD4` zG%_T?ZJi%uycZC5VEi$r2pMfIJ6P}_!Pk}Id_p&9>#-7#6@y_vwg{jv0%KKJz%!Tf z=rI&dr8B%Ns&&GDF5*46E>e$e;y+q96xrYe7+V5gde>7Oty|+yyLYk6LkQdz^X9M@ zMTxW&@#@%w{}Q>9bJ3T025a*23JoQD~7mHmRMQTmz^X+M4AOk4B zajwj{=jz^q=xt)m2d32_F&NYATDzny}dXBi` z*6$x}3(stKVC7`)uat%9hDA z-pZEkw9Q|;vuUg9_Vr8+?Z(YatLd}ve?Ndw>;8}-`Kzjb{(qFJG5@!2KuG~b2=S|h zj!$Q%Whs)9npmw?V(*86EMXTbwiLCwx%+SYa|icrHZP6nZ)p_lLWD0Mzhs9e>7>vN z!^ZZU^>(|SjP;Dpe;VG^#Z-gbX|in^H&&w=<@3i;xLfXJfQBF{u0={mY4ApZTWg1rbW%7M39Y zZrlQM^%sk^oMI?5lJ?!06&3NS&oB2;)q^ga3C4BfQ!W-S1D*b+Y@AWp_Ehftj)_af z$ROWZ8s{)Kk_OeyiIaJz)qE}kY^M@&a4!Std>y1XW1&%u#uDA%C5bR!V||KF*v3-$ zuCtU9pj^YkSX`D~t#qLFK8yK}BD$94hja#e%r)-e#4o3Ve@J6^jTRfn`!?pzUy? zv3|%8z2fP>HoC~o#Aq+(D;C(Ad-afaSnY9#wpLS|7;wAz?QcW21sK?jhO?`4m(}V` z-C7P^iB{ZSUa>~!CX?$2JZf?G93w&jj_S0dcTupfJxL8ZunU=;gII}&dpOgd*v}Q> z)8v!p>uhjdlQ*>WanXsC(L&crM0mMz6( z2iAk-=UL5qwr?R9Uf{>j5%TS22>8F<5X6kUx`r8bZ{eKc;1P4&`kLDJz)X_f7WKrc zR-YbWWWo=sxo8;K7Z zSDClPXoa>r<;TmK^M7W3I9sEw#(rp6e~ zXv|ug^;Urdb|eZ&aH5}vc-9)%&EK}x0BS73<*3N_QKQ+euxJ1 zISe2Z>-VGH!VIM)Gk8xZNN%cL^`+sDUD8V)p%x(v$uFV}sx{~h%Bo(Nl7B(}5_6?cUjqEl@}XkSb$9T@y>WNGi82izG%$E;z#x zrh1bnm;9C=dhXD9H75OXQ$w3$f^oENWq3SD&~%H3%GJ6H;{L0dioz<8N<{R{o$K)* z_xpEZu2m1s)cy+)oWB6!`A?bQzW^d>C=SN#{3^+O8cRT+R0@F6_I6`cwfKJHha34n6R?xo=m%Y!ZGQdH+q}}&z3yn4liPEUz)aO=8|K;ZuxaZK9ek~KNX1v0 z5UougczD8eWpP4kkPDp)*xrL(m5Nz?E2X<)jZPJ#iBjvd(UcoxI`VGAOnf(5%NW(` zkR+m@C_gbXZrCN@1 zh-*rdz)mIVc-vfovYR}th9$-m_YmE~$-QJuXSuLBEM`nmd$luUZqnSo;<5RfNXC-X zAHXQ$%bCtM*J-heykdIxbd?xHL1h)qx0tIhCW}s2i+fCDU1AJ5JBe=KPZ|dRI2tg2`^i$r1cCoMq{4HgeTA2Nd z&SMn&r3VF3vscOGs>`dJ@p(O}`2h7{#FY57)WR11l2=*IF*+@s@wOif3V{#;eR0Sf zN{GB39`~j@jMtaG-aXx*>AQ77$ss;CIOY6TN4SQ=0afMGw;S^jZ4 zBi`Am(4&z<={j#K)})O{VW(kRyiV$3?e^)ohK_uctcJ0nm@1G^zhfmA=Wupc$inYX zLdHZir#Ms?_w>8JAfIgQP%ecWMbmhGvu1Wpz4y z5@)s#_PnTW?mCUmxBn&ZWZDYk`~8;>BL4CL?|=Ha|9?IZa0w;}RhzdA=% z@9$|1m5*iHcV4-UI1gPE?2Lkjz+#DA^yg2~rWe7*=i+D+qjj4W9gqlYfx_=04OvEV z{`8lEh~5#1U(sc58x80;{R^y4S6kgq*IZ^TpKs@b52C?WCC4bw zBsSt;Me?r9Q>GKu?u{dbv|3hvh`7(QhWL!m{c%w;OQ%<@qHDI)QWGg#?l7t7ID_0= zf^E*r%KCYkt8T!sE{xi|(RLAU3|Vukrs9On-rTu<`RZ*!r9pjiTpq_zo7KTGyBk3< zA$lX9WJr*MG8t2?W|Y>X7Q{Rss5T$nV0!NOTb_ERL;gbqW^;rXZV7kNv~3g~g z^yV~vdt)zev-Dz}Xvr$7!yfZ^U$2uc#)y2pnJ$X^p=OMxxYWk+wT7poDKj^C-{M_(_^A8hewC92;w6#g4%?A{vFpmxS#XATr-<*m4ccK_kr- zHaDG}EG~FH?2Az~c(9E=&0n_!>B%^h$gAVsQ zJb3*3I+*G0?EOq3AV9_1j()gk`^IdP$85_GAVGKpQ{yAFZFdD!+_xDEjk zZa(cEFseY{!>3S{Hh+xrHko5R*}0dFI(s`par_$y!=?<{;`?q0D@g#VhmNC)vTc8g zj1F1%$)6me)Ln$Vqd3IFIHY|A$$J&iSC)kjT;mTg@)sFd<*GZ4^n+SGgr9i$fK7b& zvlzav?H}|6w}dNuLM@j6weKFDD<+ zT|8Iar5k4`6JDpAEitq#_)K_xPIsxYOPcBhP~rAvej-&*Q(CA=(Bh+$@mI+a5vE8? zExdF=Q3_)9b$$q%vvO%jeFH13XW|6X#MT@k!}fHB@Th2MOkQH8%;E9tc@$knV zmZI^jJv93Y!fd*TtI(vp%nd{4I?mmpVqWEKQ%mN&gEY(ltHl*GkikRxOZ+HKqOu6x) z?=`;9@))=9MX%oP;WOyNmM9*PSkawj`_000u*%cftGILE5;D{W<%dfA+mA?c$)<_Z6-WdBp`_PaeV= z$*n)21;=pi4pwW)SxORc@7PY5w(6`qM244k+Yw9CR*_HodzCdAZ2wuV)EG)ut!3)n zw4Z6)(`edm-S)Ffa6j;J<4Y{yGfKh0Y!&RMMz>rb3)<%2SgTcTdz9R@rrhVI zld11FjK>@oEq&wC(^lwzay8tlgu5@>!8RYOQ?0l|B40Buy*)Rro2#e&IsidVfGpNy8b;MIz7XRDuIYhX^Fac6NM z`x`FSAGjZ`ex;NRJ_jy=J<*wtE>FI+nY^!5G+(B(SuLHfjHIgFpA|o&wy<>x?S!t| z{s@uPByf=}v%;=tU_3$0ER7z;AXWMGx|IA#qSgvZIOX9n67E_Rl4xpmNS&Nc_ zqXDlu2<`juTT&4j7DVI7NWVMOE3_Z3Nvq1-X2p}@vf0I&!uo5O6siMmdmsl_3BF=4 z@z<>OE*)j;JMIaGGaw1XOQIMGS5=|~gM0>G)3mWv?J*jZdmW;NQCg2UX_FbK^_zJ$ zyDs9(VmJv-flAge4$L8%`E#)Wm$F4Te_&cu(}S`#5u027*TCh~2#AbG-!5wDE7RD` zLEm_Mb#aRF6YHwaRtnM{!kkkaq2uc>#MgjM$S0>9!Q^wIp|vhlbjHDL+1(h>bg*PH zt=HB>J$SBOQhDNEv!CH(v*E5Z0Z89nI-h6}Zbw&0O?af8k#Zu5Ov;BD`#cqYrw-3N zQrgAdnkH~)S8%I`{iEmr0l#29pRhxn@lN6+dP(o1ALv)c^hLHqUdbIOG zQz*N4id=^D7N!rturCUWgWZ4j<_PR%MfXcifxQV~^wr((B>+wX@4H3B?RRJ)Ea)@S z0`f3}RkchmN9qX*7oV$CEM(#F^C(G-P{tjMm&OvPqCYNK(x@&X%4xiE#t=mzeQJ?1v+%#};(ntJAwQx60qL*<0WtnZ zQut>{;(ugZ|Jo|HwYvSkP>gefr*r$3jD!B0pQ{T zve~lnVH+zIKE1yheAd)sY^v30gk;c~=+!TunmcW_doG^pH%DzdFKsL7GJR)1KabO( zqH7{L-$(C^uDW}@uDPDKoMvBho$PXb&t&32(rH#h=OQr%Xuohrj$WdCb657o==2*H z5kpNnz_i2B@nh(O?~1GNu;K)^x|GYq~PWA@judj|ajs z=)v`>$LBXu8FkqD+0IqtK(8HGhsTFs@88%cP#`t7|qeN6Vzx zr;t^G1w57qU6uzDHqd9u9cj3dd3h@H1@8W_CaNV#)uj!$9Roa#U?%?g}2-gLg~KtKoGyD0wuk z>CFQKwr{e>TdrK`+PZ1{{p9Mps*}syHjdY5J-M!udp&XLpx(lKRp8A9Yz~sPdXS@+nHChjv~UKz3$3$`PysbT}!hM&~-VpLn74& z#Ly9fYA83Skl*_BN0Cq9B!D)(rFM7|kM#a*oLN4<3X`W@YNebh!Q`t0XofAv0W2Y8 zAaZ*+rvBAR#? z^bJ`8fqaPuRXtscXTH>%Fypb^*>nRHHb=LQU^XA9a?oK78E^-tT`3}XwlPO&Md8U+)RTcWhW?{h) zhbHeVxOCpgrqNxYQY*+qdy|mwqsv0oahukm)zjF@_Gl~}&%{})s%4*0)d@}|T`JR@ zgwMZ+gvs8^>Yj}VrxRt$EEZ9#=v*i<*d&^${3cAX=q3})y5ibIm7>v}t9n{E*2TUV zINHhZHS#%9Ql?>h(Dhzao)$2=xD|>{q!R@;R7?~ib9XYP2z0HcmNk0i`^DF)8hrka zq5u<7`>a!IQ@f=}zvv=gPgo$typk@4u~Ny{_IgoDSnYU0JDqk-fD)VI+EUrMs-37$ z*uzen_e&n#H;US`oCsIpaWG$324IwB%9sOl0{dCpJ@K(e@N1m&u031Zz|FLG zg-C@5o6$z%CN+T|&gWEIHEjL#4U4SJ@IUw*LF5(-Zm?Ad%0?(;CLY#SO9e zSu;gMVFlBA;(3o(9>rqmY&^R$FH06>h*J}r2eE3U6mH1zBpN^e6QTxv3cDnf`1D5< zsjHZMjN>coRroGmTgLd(?#PB`X9xWxmU>N~`TW#XDZM@~+oX~-2@dfKBEPkimFULd z0_2XujAKTFU$*R#V;50i;c4mJWN{W!7ZisCps+~-kfe*$UnuPXvLv=mW8%^!2n*ke zahGvnJ1R)hpjFd!SP^cWP>r-szN9d{Bs;rgUBlQ6F!I0)qxyb=?1>B`d4YPQ=DB$f zzW@Zh-wKK*x?BAybr$I?BRh+qfmHdLfy5seOf4d5&?)Px-Hm{(0o!)`$djFnYMoks z^`x9uU()d}=_?wC7$P^*O)lBPTKaWs{$IRn#`M*C zO+%g_50z5C*<{XUN}o_j9%C&85`r&uOe|HdOlr>0A5v{*J^l**efl?Ys4-_t@k4=H z&q$qPxeKEv z6p4-#^Cz~fd)=-c=!)Ko=MRQNKyD;B}FT}|7%`Hnv$Ic6q^P}76d zb>aKPgPDX#Xf|s>cl%vad8B*X8UXRmy*!H535%cdjR4BfuSw2Fwvs2xfzJDwQ926+(_4 zAP|3)^-ij)ug(;fDJFP}7jcFtC-|g?As!b(pDRYmFv6cRLE$JWu+0kLsP8G0+-d4S zN&sxuEBILf7j-->NyZ#?KKQ(#fIszfmV4rm=fLs|z8F?IjkOf>wq5=6Cypx>XaXF& zoug`|ea^PDN=4taVsCn4JxCl}k=`g)4Cfo}1q2Bzp(>CZyx|l4FNgdZbWc(?$HW?$ zU?m%*S1nd-2Sl4d)DAE<<0=LI+Ga{#w9`E}(So^)m^S^|rhDpUzwF4vT6e2w_}$9* zH8rsM-K+Q~4!cKF#&>R75{=6okIEenv9>%@NQU9W;c~~A)%G`UTqR%(4njYwX61Mz zCJxFrlv(~@RwxXi^!P^GS``UrddI*?C}G{hS(<>DD02Dmu26fbCik!=^suI2E0!N` z((0AZxD+j*-z(GOk!|yEZOU{I?hNz)O7MuktCM-PO}pcxH`yrmMxe4|93fPT<^C-b z_}p&$0z&Z;JveuDM6|Rf-*_bS;=>b%Z3S&3$s4}05;~NFDS=aFV`Uy+Ex~k5X^Err z4(^*GNt!(?6qXSI9fTtEQ7vFvYYCfBTKNKl#>mwM!VxZz0^d zBYf6o@;sy(!5r%eVir-GP$3Y{m4q!F7KR%W477oLl?fbsEv0tY z%=x^~)~e0!M!CZy>}#Bk0>wrkAk4${k}7#?{jnD|-7lINK3E8(h!*OOV)x{W5!Qfq zdQ`7*3kE*n^6g%Hm`98}9!U+tvHpbn93ND%U=A3Gg3hojXh$$6&BtKAa%+$+yznv) zJQ@FSMJTpdk>;>9ivM;NM%DYK9-c_=q^@!=*Vtkr-O0RtHIeLU!hJDO6b_JSSWGDj zN?fZfX|L#3-_Vn3FXO2iJO!ra2MlagNBeSUzZBKAw8?agy)0rQkFM6FDFV>E6wX!8 z!G3&{)`wS*mpaf_3KIT#St!-K>$&D`Uv;;JYYdR&-6IRdfv_u%+w3<)mFs6uqGq|b zc;X!tJ+-Ro)ks=}urIwm!(P3&nt`TeBMT@7*_kkDNm6kA(HUhzU^9e%p~S3?n2V>t z8{O6y(ROeUDRdzX8JoYyPO(ClF{D1IXx=dVNkRYU^Ya2hmzPXIV{H8e_fm!Pgx7ng zZ}D~My}Jpw7HKQAaoyH={?)+#1cRs&yk{}JrZZs?GV*G2ebMJ*I}RedZ{dLOPn+mq z4yIU4oHXl0;cZJA)Y@zc;U7+oLkDL<9@f}Gob3xj%b_!@C3~QqdHs>AI{scxkScON z7Q2DqKTN9%WwE+`i*nP}I;+1~1A4upr-JIftQtP$q%v7!saqSj!fxt(M|)|XQ4|^yD@Dg#lv&A8`o3w-Sl@Bv4s8DHQE3bD7_3T<8^(j=crtDQ zp+kw}k_&LB6|BrY*E7VjRF<6^hpE!cOXyq0guxx2`k}JMP24@L!zItW75*6wD)Bg( z!$E}}u06jil3psvNs}867HjnN3E<_Fz_pwV3iYX7vfc<4u z#U+8we4sUbg#=v7Z1?|g_Kwk&Z(H1+jc5eMHMG~ zZol_Nzx)29@95Fre&^#EdtvUi=bCfmMO*AHE7efWJ+YowaMO_= ziZc_FIcPU0V4gmC9425T@3X9z4k+JDlZf&+6&_(-6X2X_+mK8R%@4TmE-JgJCDK)(A- zfUGUDri5VT385^ZueDN}!LTw#77>`d{$FDmn~Pjn z4yrg~ASTJ^x$_HI$Fb6BS>0HtyjV9w3%f)92uNi~jS+pt4#03ol_^UwT^vV=GP2Gf znJ61ZSEr?^*Q6Kitm`fy5BqYIZmn%F(j~@;lG&v ztBhZH^Lp@{8nsmvw^*KL#0=;Bx~>$-x><5GMR#-H6o8I9R~G0pE9-n+lB0?0q={g@ zxjS#|d9o1!8rNh@sq{GWAe?;1Ph{;$96sH&cd!kn2u9B^T-%8^C)1p3&tuj+eFi6) z8z;pYlCC9@O;cYCb(*Y?icx*xV{)e)3^6iiW`PcuM$SDQ7&D)(m3qA>_8HJxcRQMw zA(!$Bp_M7qG&Wt8(pwH)${1W`6$OL5Smo_OjD)*sGmjfAEbLP}#Z}P|G5_UhtTJ3aR1GgmAlT$ruQi z9)g56f>hxqCFc0KNqv_<$^@=S@+Qf~q3RlqT}oD$$nE@e`ZZ&ScY<3qTbCoRNkTh2 zg{EAlu`LffQ|j7)ea>O>xs|nuxpm}?o3#pxN)gkHkFR>;0v z(Njx)n7v#q4pG%!j2spbK#ZQH7bnj(1zW$ltWX4;D_A`ikFn;L%7Sa(AI6V9kd1*S zF|d|UyxtbKcvrW4ivl{c^<<@2Yl-<9GT72Tnig)=16P_u=es}CoHg~7!|GhvH!Ve# z$R~{bzv&t{%1dd}VmPc%3TJyTpePSkz$4oCx z3~K{`(E0DgzPuIh1Vr>bfvAJEeF1sJll%nRpv4zn6JVptkpGnoIichk4FgBSuBbR7 z>f5`6s5d5x6BW(M$-ZC8b^^j8y~;lS9oItNNwGk{CWwi9Q@ts^A!JtHx~S?o%YyWA zUbOo=ZhI`dBfm{h0wZx6#d}kHs0FEBNi;wZmrDZIBZ}QIf&G@#<`+#<{D=jq+*>RA zZIu=2H&JWSnGSZx{m<%X==ggv*_S8L<27%RyYeIWX(JDpH%`r#!B_*J2Gx`q;K$ek zVAM=fSK}LZ=-%;T=j09;0z5X)Ma-qsmsNnTil+HmZ!YO_ELRF0&2oZbeM*J;1m$u9 z%g_tt(96+8M6wq|ea8%v71faDS=a-fAHTznG*4FMA=(|skI=fe!%y#B{H|;-(}I4M zuZg^kQnu+z)1Fbe`_Sl%{CYg}Z{K{x40c-gV_T%}uRXw3&mF4t5gHqSwA*^!tzR_k zZ_==Z0pYp**$#JPWr)tw*4rCcjV}&Nwb9!U#Jb!o))t167z;_(dxN;KM9xm~FOa?- zPGN^~M&0IhgjPl$=3&V)r}NTI7StYdaF0ehJH0bTzNHWEkQ6F?u{0?62Pi&8=4La4 z*>iD!o+EuYpxUNe5+)>4N-Fao8y)qOIEc%3J45c4wg$m8MjyzQL^UcUJc&+L- z8!;*Y=XJwQF!YaABkVJfWV|-3mlk5njW9oRXzW)Tv0L4120MAx_Sk1sz1JJT#CUmb z(4Kt;8Yl38D}i0nInI$|uD1QGrEG|V%`F`rS@NL%l=Z=1h30Z4Cw zsS>*C#iUM$?+n>xGs zlGbRPvCdVy@#%N`+@3#{4_~#QlDi|QMaZQHXPz4C0(tiQGE#kR+>z1yy~#`o`rn&N znZ}v~!MZv_X9^k&1`6L{=@>ZrtR7MCEOmJjkJ*l?bY1*LcKmi95h$yc`AzE(@JBp@ z1K=MGMhQE+Hl*HYoKyUQ@jhREF!A>J-6DNVxLl$G=g{zM)$is!`v6rR1;S#6VlSOP zBBSqEFZRkaCrAwMUw<+)}!)mcA+5>>aED#Kl?+X5d*^DUe=h{{F7ke2y~A0(@)zOZ@O3RjB^5eo!E8l4$#DWwpaN(kpM{l`rWTq4p$Izt+^ouCq)1GR#=&UU2b|WVoh;0qqI#7| zv-ya#p%L4*C<4wWjMqOuucr>)&9?b)=^&EGa*m)aKd#&1Q^8 zG~feQzpP&J!SmI$yVOcMAE?3fJ^aT&bAhIw*NYh|1d_16?CKq zfR@&7mDS5WCyV7?>0!Gi`Xaf;+(>;;RMga7zv(J9rY>;Ic|KDaVELZF-V_Yy5srT_ z_kBLn&5?u8lx?TfTt(PzZ*k1F^g=_Ul<1=JLm=Sh;q!lZ*~!n zuWiyEw^x6B!z~qX7bDR>thWmOs);}v zn+vN`@bMb$+N1@d#7IYrSr@1xcut6q^52pt26~GV$$B>s&$FOq4>7y)%EF3_MCX!%1g9rnkmgQ z`6+c$DEwr3*{?Ij8PETt5=D<@&vOR!u@Z&*@`eBZ%Sox&IvJRn{J))3wyK5}sv4GW zXpl5>6cLC58reRma6pTtmO%Aue+BqlnNG+USri#dnza1jyw6T+=10J`dwxrxoKJ_g z#~lA+XUlau9Fie;F3=qHh-c?9$IfHM?Woqz_Z`3Y>ytpAZ4_k~3C2DnwJPSIrF0cN zwfO)!wgYymiMB$XK3Q0$p|1*DEyPVn+KxO@AjRbHk69BkV=UOS_@sD6Mi8Z8Hi+mn zAlq&~3#Ly{RNOkak_vm}$0%&1Hj<8^E1GpLInyr_X($JwVHs=|W?PA}aV;)73p_ci z_N<`w0IT(;jp7xT z#Y@VUE^QS!>}Q*#kL$qoyhhT%KOJ}kRx>~rm_u&U9Y}d9qD!}aSnEi|T4*SFdkOZi z7|f4tgKSMZms;~Avv(+QZJe>wjf`q)dY-lg_K*fMEVRU&b;M7sX?V(K&rwA8x1JWX zG^gIUY4m+na;H(~Y!~M;J$bG=oy_Vg$5qQ^&w3BsQVq!gGw)FF;%V2NfDTlfam zOvSJ+Lxca*oEtUL#9cHM24@s&I%ulEn%;=%$N7AZ%kz!W$12~@USPmV{Z^>s=s&(83BM+bc zEvjKucxzd8KThx$DMf+TU+I_5;%CsRH?k34T@}?xm z`q+WBJ(B2Xkw3D_O^PG8rrZX`*pn7#qRu?Ivg$UnWJcd6e@T;3fsTtLxUv`Q%S?vv z(ZRH$O-TwndQ{RF9o^RmY9mj$mZi6%=!{FC433K0&I*tgh}?Tm*Rvp-nMUZr*hFv` z%LSzcszjQZOC>q9N=wNdql%sq>rihG^N^@rSFHY!z3p|KxG3?BXJ#YMHWJyT886%C z?V4FnZ4e!XqUIYE`*m=;>~B-7>>VYizG@;^@r0Jm%%^n2)r~ZeigGqCA7z^iUbs)= zUA>R%UAfQW&5-YMnMp?6<}N^Ht&{DKg79{IMP_iF(Q;U|T>oVVe`?29qp8xLoy|nO zOEG)svY@wbJLa1(0iBoBJ!a+@^^3p=eVAltfLltCIW}ShSs|*%YsiVGZ6*y;6^<897}4-W?|t^8u3C#%eVA;FcXsQ22Yd}ZXHUQh!~60uDx!{i>!~ir}o4y z@`-T*o3-20;C=f9XPax95LCc7bucN8-$B{#G^>aqO>i@-?V@a$J2rD#a#)_46F4xn ztu-P+^S0+w;h8_J;JMvN{!CGR`h6kvGhC%;gN>{?o4zQ|qS@Lh`WaUjTbWq43tJXR z|2=|FcnIu?^O-R3mS6~db*4VU{`=QAa}o5G0}ahAz;d;n?~Jx_!uy&Pns0%5PFg zKr&$zJ6|(%;0ERK4<%^9+ndENHPV-*IrL0wD;fzDH2a;K9?XJ7+&7Nv3;D%yz1_nkTHNNp)iXKrh2leM~V_T=6$ z-#e%Q3WIV1$DScL%dsGIdqCsI9 zaYm7Fb5z38fjZf>>9n;W!e5cKx&({~fA#xemW`6=0%8NgX`7knX}9UD+@9{wFZP(N zLf~akWxJMPCRI;u82v(9BM(R>rFKu|jLJ#Vjy1u??z70INZ*FZ%A*n|Pt;GfHK!I< z$A2hC0DQR?$4^ECM`>AFrsTJkH4kI4N;Iu>Ag{~UzBRSS%-=rG<$b%P*_NVem}Lqq z)v-;!6BA(Yt7HyY1L02+l2#d1)5_-an>k22H{SG3_+7A4`hDXeQ^L5Lmtfz(>N3kn znO%e%r>G;<$)RmvCt5tj>u40Vy*R7Ux#K0K5QkWhO)w0bkb`tnFfreUB3XwfSK(&Z zLOvv#rBd%aig46#X`5#R1Yw0$A;>D-Y+F!TtET$%>Qx6{O<(F|7;Qbi`)XVumCRG04*%kRJkMaI6;oT zo@L<*!BSEpAY{%d@6H#o3%zZL@N}kVSg0h0JL#D|m zTwEhNCi1G(Q?7o=RI4sn!6`hpD@esK6@5kb(yCr}U0PhSoN7@XaC8}&mu}Wq*LpD> zn|l~pmyPi@g`2Q9!)~DCF&T$5+}>Wq_B;{RtG2T{k(JnBv^ZAot1&m#aDwYbU>A{% zGUytD9{T|ozLRbWRN4TWJjLiUm5g;PcC>Y61*cw)QUgbpFjEzm`?1No;!3H`t>m8n zb8)f5D(y_NP7c9wu1qsS%WbrB0L(BAuWRW^Br;4RETlv1 zdij*CWQ>zP7t9@;&j^xYoqt!L2rbp9VUU0g%NO+f)%#P&W@37w3O1`VRLh;r>g|mI zts$DFEwV2{D=Y)++NnQkOVe*cMshr?j_WQ7RLrgg|c=X)-9UlH0>+oOgN7yzONNV;Kzh!&0Zu^kS_z8 z>QF3^DU8G}AqN%I$3;J}B>Jm0Ng!w#E%pr(Ki(~WHCD`*;36vQ^#l!3Lfc=M41);9 zW3-=nzuyNA7BLQqfP@NgI|~0i3h!9%ubrhy6bz?uZC4ZrYyM(3(3E=5iu!AZlWGsA zeQpXtoF96{mGIVJXOJ_kPPA!RtHO?qa>OOdvoOLIhe0XfEm!Mi{()aXdQVW9f&x|C zXqVydS(=0>BC-pt;2*$D{l}~Gf5|`m*C!e4zw=aS!xohZiC4B|je?eyu2_4#xQR+K za8U6?tD*}$jvB&U9Br_$(WSKpuk+SZ;0<)&Xct5UG#qia5UJ5NP7oyAy!p@5(_~h9 z{rme!s@)fYJzey#Z5l!lt%931`P!3ig4RK86Y>sX%vJ#7)zfdLD5b+il7kke&a};_ zs!r&SQPWvdsAd;YYsAOA<;CaB7wpG>Y@@LPbGVS;qyjF;!7b%CAwne>>#!#vL!tAPqadBp4c=qFSA8VhB4ZEeu{Z$z6R=fRu zEF1+jAKz{MWJkR=gGHxxUq0R*;O$1j&C=F`m<=>C5`a>$Z4q6~J5~S@f>y1?v?*3o zD(DIM8YF7$VsRP&=Y2;1QYnegV_F2~tK2Rx{ahRSgH}KDZZ!!`p&dl31;g!i{V!gO zo`Sa*w7wCic)lm`Iczf=izTNTqBKpbwzFfaHd!ZsTy?Rn^EsJ^@h6zS2MtznHhmSa z-9rKc=N~DHa)y>BM$Z2gFaL@bn`i;penupbPjed^mpws%(0RYV6go#BO<(ZO?{LIT zi{SH%1MR=R5J*O0a4ZKH9Ql15@aMGh_>&JKCqR=5(ni%}+XRtHFI>pq(wL#IMBvX9 zquE}PmI_*2;n2YRV4ZNfF3$8`nzH@e8yjO)9raQ0@LgW-v@5TVSQpEr6N&c zS*Vep-ISqLC4@xL!$ur1JK#l0bdT)L`@14^*6vjV%h&b|^zZ-Nk2xhQ3z#a0KnAzq zKLMV<2!hgg7V@5{e(W6eSrNE+!)g9EJE`#NxdL z(B$pw`Ih-l(LJbpxBK$L;80b`h(*c6dX%1W!A%W@@VlGcFsPZy?e@nUr|b3ux!lhW z%wBjvG;z3=LzF&Mh{jSFBuHWL_O!qglI$!zZx)p;_f;MICkqdW%tp#f>eytK> z9Z&C4M9Z@E7IPf-^4&-uM0mOHcKP>oq#_I#=A~s2BU&Q0M`(^rlYfq79Bp9=i77wz ztv-WuPbIbCg|IMQed@Q6H%4n{EM7?QK$bv*KL8SR+_4Mc2cXrcV%-vaMCnP~3#n0Q zq8R}Cg|TmhvR)LaJ75z&UuF(n`&^mozM3G$<827BvX*;cy(M}>*#|Ex2m`fs3;_X5 zIfH0~eRy(v)LRxw*^Y*gjZHK74aVk#0vPRaeR4qz*NWgMF!g>vD!P6!|K{7JnJbtN zjeRhd!=K23Z9G231mX9Jy%z(y11Acgw39#6fl;0amG3b+AsqC5^M4wtlU#@NOI{{! zuq1SPu0wr>w;UpTjiMJI7flTNO-1%B5_M-WdCQ9Ngid=?39>LpuRqEfxSy00<%IGLz}2#?vrfEpum%Y@_R;7PaLO28~qK8 zCB+k-a*jaYG2_^s_#No2cX#Q}V53XA`=-?2s7=meTfJ_5zKNEgQ)J)BQ;5sIS@e|r z{-0Jk^;Sop1I+zYVDA6pW#>OrO<@CP6Hz-y8v|!~M>|&w;MDnFOp zOiXE8qU=ysZJ<;{iY{pT`V<7PsfS?oFQ^sFte#^%l_;&b5L-=sW%{Gys^@RUW%l@! zaQq;fl@;QQt5a5v+3jjGbSJz0bVSbY6HFhamKkd>DX0@$dl1p$E$kEyB`uyYQvg;; z&__=E(FA}L8ESYVKAaUb!hB>xE3TeYPqsLLoz*AD49l!)42V}qQb?$$HjwG(7COth zTz>-*lJm8WB<8wSp$)N~bFE%7wHdkPn8XidwVb2LoK892FvO=yd~4p)PoqxxnW6j8 zRbh0x6moplGC{Y)hVQf0sygX*bRRfYO+Hhb?!EQ6y@o!KTTH7GX@_%zG4Q>cQPo=r zfxC6viFcdid>e8HT|y&Z&w9gYpr-k!({q37)&jaFb~TcPtZqE*k_^4RNO*pwQHRRb zZuh0?*d?CHnLmHirQHt#HhvTP#+SAAh$oW@>hlsPYNw!am2{dIdcTwW&%lc3*1Q}mm`FG z=IYiRv`NF~x&vXfznw=tl@v_&h(XXEe?V8)=CcV$0eDI8;UZxdl>%O5@=U=VobCNt zHjv9JP_%3j&BE;nk3ZmbAXH-{p~x30*Gx8x3ci(Z@)WgiIABf88Eh7b3D&0M33#`5 z$JFx6&p@38fA(=m-TLf%iQ_p9a>#i{iaiZ?-pI5(`$j%tkaT-P?TjLK+kv?d$vb_* zTmjf{>jWL+cU57osQR$+^}p^_xAZEC!!pkAAaw3eoBh4v*SV+3J^^8lH!$&;|Eb~s z%T589`bKuP#z1Rvpd0AFp|PUOUu*C_nQ1|75}L3{ig~k0^JIZooGmKkK~D*SnpF8H zhfNNuHA-eHj@ICQ!FnL2lFB4vWWLHv8YATNmk@-2oSTiu?aArf#N3`Q0I?}y5X4(` zmiW1P!~WTu{P3)pB>(}_uNLjokPWysPovmA_j1TUgYHYoxo}Y*!f`{bwatx# zfXcff^B}0JwEU%(rBTNuM}PsB5jmY&2ua*#V}%`M^%Z2ls^e}*O~QdpNP!R0#Xt-! zPy8{{;@Ktf=OkUM)X4fNVlcifJAKM%;Td9cvXn1f_T%Fc|6SVU!--otuH)-(-p54C z{B!PvP=)V1m0CZeBa=2_nVJpU3sSORtfKapa^30$jJWf}Y ze2@J+j=;kB$cD<E6EFrR|5jW5IOh}S1O;B!!J1JpDM z+-Y1VBGhBxwc8Q#aN*c@Y zM+5=_wecJ#snZ60uO~G;J8(KPLa%*uQ54WX}QTTA!a3*#~E-5Cp&x9E}ERN8s63Y$Up>X z)ucPD+<5Ny##FtXj0H!TGLD2}N%> zfkzR6MZ;#n+gOpwYysq0jA(S|th88qZT`MnLxh_|R?gNSknpf=3HEs$AlrvDBIA5g|%EK0#(hWL7a8@AvT6deVPJcd&g-wTbEVEj`E z=*a}&c!j5Z#E-9KjWldR9~<=!VQNsiq|_efoNWNd$;N6>L0-}9B%z*rZo|pzDd|*? zJdEj!L4My1eeVk1nvR87Emz86-)ECn^a*H_eKwSNhB@VJ=J{QW=vWP<$OTpo9;d`g z(k9!Dn}(3F1zRRB+}QXH>*Bg)aUf{kM6ge0lY1y={_@oovkSxj8<_nwhw&+D5eB)U zTzqv7$(U_C+0{L|1f`g4+7Nxohz4TE{+}nt1kHePkAWzf0te*9iZtl8n11qDSSXO z6T7}QoT|JQS^t-t`Emw`!6Hz^>IipdC58w4ODc!&Z@7aD{!pwiWWHbcVI2<8%i-d} z3RYJ}j<=nzoBuVHxY|>NmukBxjG!T?d7(VMV?)#;s9{mQR4$3`gyVT1o$OytQx9j9 zjYcSLA(_(-<}FmpG>&H*3j|Ijbjph_nE~b)e@!K*d0CL*Xn<3RJ8O1~vA9@t;z}P> z2e$q2$rY!%gwk=O3k$l0oPD5uvPVyu;{maWZlo^=oYj=-r+?30{ysN?UKFCRvD5hX z;fxJh;?rVNE1;9%Bhbn4b|eVmP$y!pnDV37Sb{8Pw5wvO$IRm!CeX=nIUeX_c*9~) zYP!EC>TkmeN?VB0J0cSTpQ>b93|k&wJM@c6t`tll`4Y_*Ug+I881C|uA)c7FY^<(& zXhX~}*KGW13L!ZS#-=@4^mEvAP5Z=$jL8E`C)|>@8Cn|T0VhiNZ|H|=VrAX{Id|caZ|43Ffdzj2J9^Q^x+5f}K zuxC6SQ61Q<`GI5Se?;m3XQBQ3=sEk}qvxyjYUd~phc0+zhkbU)LmhqpE+~>b7~^p7 zb5Z=iRG#8?`RDk1G5P+gw1)SQ6dx2ufL0nKFcfU7buZ5=-WiW8r|C=I&-YvSUZNCZ z@Hs+ZHb`Sn`~*h4pfF^ZnZSF#VZwtqCSqc8gM@s`Ip>Q5_$dSE>a-Hm8C*-g+Cyy6 z>6o0@KP+TwjmzB$ja6CsRP9|R)E%ehE#)}l6}DFv8xxkkW(&(rY!5^t4v8J*JPUA= zRSCWB83I|YEo>p&8B5EEk1QEK%T3{KH?^7@D&kdbvU|K_<57*V%88`z4O<5r%kl1> zc0t`LKzqlgFw3W$Ejwz#)hEyGU_`@))Wr3&1yx2Gs0?VkJ1Pmv$MXwu$0LqQzjz4E zq*pkZD<$nf6zFWVxi<`|ZB9c#bx`UpGvljpNTk_U)ve1N5oqkYR^h#RWxp7os^oQKMxNJG?@l5nEo?OHMx?jvb&^QK1IqzeL;G z&ft6f?7{0SPbh!F&!}mc?Wfrm^sS*5U5hB4L;O&HQ1{`zMr;6eUMU1|P8s!tAk_KJ zIClp_dxPB_dry%!r!zKcJ>b)j30sLo@wH}X4l82};l|}k`|FHw^~zB`!|dpB1&W!z z8Io+RerxAy@Dnv+XZxhm_o%$=Tc<&hh-0T)wXoXon$H3B(Ppe_aDaTrJKOOgkeXfc@DHQy;FkzI0*{>QWV(M7mD;dMW7^BJ< z_Y=-(i7>}pyngjBf^LPOJ{;|1bns>bw}U?y)R^);BjS(&iW#|Y2u=^@*K>rrFEWXs zP=~{p2dmF$+z(Gga(SrB^6Mkn$3AbKSHy+)FncMvL7ADbh2iM_x6(&%`$LigtGqnW zdX(ay@b$lKkpErSRc+LOLo;8BAUF+7;qOR`)L}>@9hQ1c;B!cV!ACFmn5;nGUt9zZ z-XpV_O*RIq_lhc2+#xi0kS=4a!HE@T4;^JJ!lv*G_QzPG=#s$q#1{S6Gf)NFiItkZV{IN!1xGwi@Ru`PD0QN-Wa1CTA+uyWE1cOQ>+! zEmw;YW_-&*HC{wE=|{1_y2!GI6)>b63n~hzB~rs!*f^72vwjO~1M~CGL-!L6bvgsg z7D_zR0;;timRxKKH&GC*(*$Xn<4}i%-CY%#=NU7@B&X-LNT?27ao9E9g1T^@&jD6aF-@svR$m6Y9 z;-vxkf@!8Fk9&hNd+|nLL!fX3w(>rYF#L>pCm=ujmMw#*)iY%^XmL=A*=1MYVN{u- zkVD3VhQ68lEVR&#peJ|H;psSfey@oj$+530Qy+T7v!c$~hYGp-7T3=>dp79sR<{4` zOkVt}=u747@2^5LBZvBJOErZfvD>uF zOxb&mJR%)Q2u9tBN=va{$mBr#KzadvG6U% zY3S8$Hj&06trDUJl@r0U1Sw|N}tu{3uGFLw6)c1m}zAQL^uNG?iF83@8J}hxeO)r@gJ7m@-Y=oA|BnjOx zKrt)8uTKoX{Y2ecidkDZDbn~w_ayu0jpp}C*wuxAM&|ZVImg;=mDsc_l(BBqJJ$D9 z4%s*{u5IB~j_JHqZ=K%-*ZIe+5_bMdJR&eGyN566hwh<Z{5mU8B``?ndHK0AyBDx}!h% z;sm=clUX8WZXmEXIW}by7^r-Dy11BWrP$Fvp?qrd26qas-Llq&K5@{!^+Q{{DM$%u zQOjQq$H}>tmgIgWw8VN@qWEI{p!ztrB)Y|sgmR{F>py8vDS5H zurYui(vUq`^buyzN|<4RgpEptEF3-;u&6qiQJH7vYO1q zQC$8ai&0>dmb}M>EM=4$wEJ)pVqW!{1_t< zK|lIB(@$u3uXLOSb&cA%8%rR+=moUs28K4M7O{G;}n%il6t1)h$!O z65<_`M5llIjhq7%e-4f3w-QCS2Le+^U<&;n8;P|c8QpOb&Buq6#u|BBe0lq zs3~A=U9R6d`s=X^hNCV#W1$f0Jn07Y?-QW6t@VN;AXO3*NR|9YA*X*)Bn7Ok?Tr2( z&9h|HH8mhJ@-rIzTad(1n7=%Fl@zp8gp^jWB1pcB2DPyDX(Tnj7mr zdNB+>)~`i;{-s&;Gw-1<^9u_**<=l5>Z4rkJCklb*N@eYnXfB7Js$|WMBRu)K}`Iz zkn#aazEM)#$h{WQm)KJjZ9v!`5t17&sq!#J9^#^TtD=sLypANSVqA&M2HtK0(N-uzeNKBTRs(Zcm*t1R`=_tpj$^6g(-}VLY&+sa~he~ zmFxJm0rrvz>#z~WkWqmLw8=zHi*j|`k9 z&W18dv2t=`zgpq@H0%>GktL@%O=;lMrzA3Ux?XAY{ZU+X6wXZVN4{mt4+f^gok0AL zbgLvg1;yYdK*N_q0z&pZiJDB_BU@+tB}dRMI{bo`zyVId^3q#*0l=jN?*Pw)0^UR) z^_}qVct8H8Q;=~;QsN!Jc&KKZt?gb3Hy+HhPftFHE=C1bu9&_ta`H{fGK9Gslu&GN z3++ZFPL#Zjl`t`bux?nFjbDK$!A2*grjjmF^y^ijDm7fIWU$EMkQOdR-H)&bn!%I- z`U2wVS;8Kkwy^9Qh>MO$=9jHL^B+-<6#P&EoBeW6>5IpQrm$?<15tJ9UC;~b{ZhI# zTsDi`g4MQoKzsH+wh)3WF$Vg4L=DH9!rZU{;jCYAJU@6iBMFheb;Rfqb$RCo=B`$s z2xIO9u7;Y0k17Ru77&=#3_v5brshcE`gns@UmR=kA2zJ^<2)!{ArRc<_y>67JGIT+ zT|#_R${&?Me?%v0duYMWJHjz2DzyxQ9&8D7W2BE<)r3Hqzk$d^nny4gBC9N{{qeIj35BnkXzwIF2JZr2maSw6HdiFt$id%dq5CCfhr z0tNK{zd(WUKf|YgQ)1?eG=M;XDisJ6lqKR;Th9KX#2_Urx|{#slo*a0Jf6QOG3bCi zN&goRZ?u0Q1?1{w7Ct@}v*FZh{_ZTW)ge_N9|kio>J=RTEkAkb34x0aFsAH`mp~y3eNZ;KZiHxra*+stzsYs6y+>$ zsBsa4FYoe6jN8o&Yalnrz4;)dJj&tr0+#z-(dCo`7JdSO0?&rBC)-CbV-;)0Po<9- zv50drPhJ3e{REX5^W3IMN)sl|ly7d`)n7nC56Kh=6vRRsz{MyU>)na(T9!X_j~2AU z{v5T(Buf;_A_U{6AHWg>uL|=Hl2j0b0!&wL*-T_B!a5>Ytb$!V9+e)+<-|h0`|Ef< zRhbs)-6m&WK2xv@xJ0noo83R9vwlmr|F{tjXL{((W_{sU&K#c9PR=11|n+rImj`kq? z4ZmWpojpvq$G!U#fLmyX&0$V=Khv~J3gpH-{edfvO=Zr0mEMCe%di-7vMpeJP<5Et z*k}`K3F=fAnmV6*|3A4gxxSR<6u?104={ZGksI?raN$2ps%RxGI}`<^4+MFjgJU7DdNpnn)25e!6&?Mo;Drj_cjKVOb zY(j)}Tp$K42a@-7J{w2_F|OTzjj<(7{bWbfb|T`A@^>KIFg^8wdHLj~;}k^NcsL z{Ux=>?HWcm6Bc`IybM0l+Ijc6i%4&s>ailV&Te5xLEI1@LKNp@f4`H`Sy+D8lOa|_ zWZ)@j1DJq{#lz5i|3qn}l;aL&!-PGf$3Wp)$ub|CtqCOm_Wb8>i)vB&b$JH@@vvw= zmswGeBR^)jDBX2TEZYQ=c+;ZPaiO|e1@~lQW=Gg6kzHtimrW9x;Ccbd#L4)ula?AJ zo*uzB?=WFo3;$o+U&=w|g=jzX?7agUFcpr(EQ}E@?XLK5_MCfdgN8U$0 z{X>gTStm$;jgBKxpgQbw{JmuPhVo!@03Vk0fDrZjKizHq%k!NKs4i(>V*(WDweSQA z-2-jlC2Z`0iy+npPEOgX682~+SRdQcGxx^x-_xb!(hakrUstf5S2PLU0*cXf^rf4O5X$~KF2$L=;^qly4VJ;_;7FNWY{RA-14)tABpwAqaC zq1FyyKK8e1dNXL z{TXYfb_}sPQ?1=Qk{;Qfub2E@RT}O(GVRfE`EaJX9Y*wo8rG85a@q{*%r0%}%=}WM z%lZeF@=}&xCwJLmeR%jypAc;Ml~xfYTQ=n5^giL$pOY=DnQq1Qlq*D?YrI9y(41vb zyS3EIv}EGaA%r{nJp^4m(&o+M&4rO@qONiK3-518>DvnS(;7Z=_n^bLIR?a2EW8Hr zCWK6l-|$#yR;QDkKTfL-jOj?Tr?_0#o63#ORXcp?S|1cy)cEt6UQ~U2l#RjTw^-TZ zQjBR1OBoX;x8Yyn7)k;Ol+EC4=?)HEDvw{BSJ+v%iS`^SZ0J?Tt3rmy$k`-a-*r3{ zj0k5Eprfa~)piIfXsyT!kVkMCwbSbztOHj4b8sntac<6vZyMz+T(*~Fn@p~FDC8-M zq($)@U#lL;RO+qgaLF?2kBxR@cV#jyr=g5QC}^mYxv-M;CCTA})rJ6;%U5gnz5xAg zkV2IKC3_Bl{Cx>{xjhSbCnmW8y0j65mEUj(R%r}sGy7p7h3+xN0nY~pc&T3Sm+rbl zUvYr?f?QP#zfp9Tf5YmUw9BHTJHN5P*B-zN&XXWXP8}Ag2tlouNqzcZqZHML8`66x zoECMyd-Va7DJRatD+lFF?edk`Ip2?L6Ca@`EW4cJCk8P$66jWDiNBU6tXEz}9sa^{ znJrkTtqeDmD*Yf{m#K+&xbR&#&=Dn7Vg?g6wX0|Q<*`xJ`tS+K0+~E#Gs4b-0wKB% zI1P6DK@h=6y@92*UtBZm4aLE|J=`_qO`BHMNHJeYduHP}B_P_l6KCUB`1&yY>8&#{ zi3Uq{ovzw3o zGz6|wPv2m`z5FI(NDH*}E<9(-tqasCkBHbS^IB7|rIo+2k!z;ETD^DWioHh2+#eWo z1>s3RCYM_oHuE9e8O;aC!XDmf+ssV~A|X?}1j<=A6>whl_UVg<>!_!OWVxd*sXOG=8; z*UEX_OTL0Iw;G=}0(|AVI3t>pxl_5kFgXQ!KB3|#yv89+a}>P-5$|Bx8#d2Dj#aAe zp|~f6*zjuyPPKgmUs93T(Bd}KNC_6MouTA6Rn7qN zwv|tP&OI>Nx8xwsU$tbT29}^W9uS=q=7qsY50pKkuEHYO5wCBbjw)xtYGD&$UuFG> z25CC4aBL_Ghmfc*YV>3B@e!QO@lE+!xojGLF-lY{t+P%$2WfOw=L(rIxe&Z zS1IKaeh@-$v`W?iGh;Loq9#Htq+oJt+m&i-|Grw$w7irvI8{Da<*%gn3lFC84?_t` z^CsNEkrKH!MqX5owy3hU0kVz=2^k{?TE9;|qiQgk^e$3*R`Ux#xNX}fqgamgl?wjxt!0_ynii{fWkZHrW1ET;`704{_$Nvwr z9MMe0mak;;e2l^W|KjW%gDhdzEnV!gZQJg$ZQHi3F59+k8(p?-qsw;H)ORp*&$$zK z=H7@M89QSC+7X#=X1=-Bv-I0g9^hUy|E4ToJdyiWmYhrF4U2 z`C_kYqO`a)VXbeLSrwi4@pZ^Bz+}N_WpQ&^wsPF&z`0LEG9LeiFZR)aKI>9Xrj&<_eVAg$?b;o)9?s7Machw1| zhgU6x(BB5`am^fIsy0!dYIs8(!5nc1N!yI?5B8$Q88tZ zZELJfP2O6FPX@0sXWdT;HNrFqD#({rOWn^SKO;#}KwRreso5Dnh7!@{A2*g1t}@m* z7HoYQw2){{g9%mgpx`sb8forbmO~T?DP%I4S(HM>8pOGA{!!O9I@@-C@^df_YKQ~~ zwQ*7_8&G42Mdv&1w46_Gwv~2<-C4b-yacj6Nir*|$I58jiH?t8gibaNbO1!SwH%G| z2VUVRej?k(uRMl%#_@e4S=ge}K8O;)i|_2hGOaYWb9V1;J-c`V6}m#_^@A?c#BulU z$exDJ^VhllJGOXf*Xr)%up>iePd_T(3iHa(L^Es5Lf;o&U89-2O!;ghSSu-576X2I z_JeDDDIas!Pb-5Tfs)CziS=qzSgciK zRrS((9oj7ttJH<9)WcGKLh`xJ&6TETqkrVdC!k%##i2=XL&gVJ>#a4~*7CKf(7Z<* zC50|NoHpJAs5E8j^;m6DQ9(p9;OZ|=7coW$>qdzls>Io9qn0x;St=-noto9^8iptm zIu;c}%-FV7^7PdPiB|Fq00JnR4I{#NP&3q^>#WIK$a7PKRRg>zANPNSGhi#*$2K9k zDd4NDzwe{`Nz<>Ow@({#PfQXG<@UKI@HMdiG>~<@CMSF8C&jOLcmB?QWfj5XmSg$cEE#b>Sq5>HTppJAqH0CETY&2~!vOPkme z9%O(lcCaKgbb5xAXPEQ`)MY;H&kz3aT)(lZW@2o5%tu6M8TXha_$tq=#C|GpbD~VP z-|iGe`Gf3rsBnn>ZgNE2vm@l1YC{6%Ut_H-O84!IUedt8+R&lqaW?)+0 z-bP}wacTf%XQHn_o7Q7(UW9{I$bHKHnATS3@Y8i+J<8YPSfR6OVLhgh_Tdt)8~!{G z>Jp)H6+Wz;=c>6L(YR;jLFyqS_YqXuaIL>@`}Ap1jtS!@|GDSnIn7ZMc$IYD6@PG; zk3oXf<-pbLbDlQ}?fYv4W&)-sW`z0ahIUPKixj*yqG0^QAfZE7cpSzIddpdU)wtmd z(d8*Rf{;!Sb_05&@T6XG-NYn{I9Ci&iQ<=FeB@oDPON}W7G-g$_HgE@>3cX6Gjz3E zRX;|})c*Y@z#&gFMQEdIGg!3zU1`8O?SsVkrhk*f-$G4N8%bVKVdh-nW(JXPxX98v zMP5kGQG^OT)6RPRyWI~?<@c%ohJ!>V0$eI3B==Fg9^vfU8gSwbz(aRLbA;>6rT^*4 zMs2tZn#eF^y9GK~lPcgz74Xa%AiNu733qXq3}?2rR0d{uRMb2XUcU2yathu%msf=e zJ&02HtQ~8faCm|8TBMD+<#*hyjsnDHsvWLtkWRFzSOuxx1n`~*iB}DXdNik9(RQAk z=1I?D!2z+m^atqqCCkbtj6Tv=k&iz&Uo9BKql`cIJ@0Q=w>4&y%m-?HAP(`fJYD~v zU9*I(zlH_lc>gVV4ba3=Hu_tav+=E-{~x`-{nKftY^{nVjO_Dc&MKZ32FANVO*?5# z+HF-~rX0U1sxjYB1M&Jk058}WeBH*x!G&$+6Z$JGn`o!*_{7yWz^C+;lWQxjFa#j4 z>*beGU9httOB?a^sl2Wb=G*s|fp zywMcSY2y%6y8RXegW`4!QuJ=p=+eb=iARIWA)az*?@YjanlTqhWHk_GO*agxL^uu; zWMpMas%DZ#xn9)p>DtOIgu7euPUyci71?usC#7gbc;!E1AB^LyDx{%wSJnEKSw(J| zqsChe_Gz@5>|wUmN>|i|!lNQujrM=wGDhByB<=noaBU(<$R%hGOo0npB+egs(%^5U zC}j$_RKSZ$ASa9E{8MIEn0_G7?eZ?TGldF+wY?XYt%9LaoT)4iXTX8taAbGd*C z?1aVM!C!yDmD|A7+-`xrNiP$?>)PyZYAwU0fZV}gTo9f<MiI@AVE@cDVfx zi_yswpg^IZUyb8y((NdF)`&!-eqJzC+WA9!v*I$N+u+{k`IWE~m26`lz!EEhPIjrRrHVbWvb8c=Gi?{m_r~iN;jr4@ZLn-rN^6-#+FR!G&vB`5ce%nVGp4TqSc|h~L}+?hi>Y zu_T}0Y21arj3Tg;?1t}TL9b9(2+pJDqBlH40D`CR_b$5}Jj7|R*T8Md8canwam5U^ za21T7hKHc_F}nCq&j6-Sk37L*%&jqwL0q$M=(xPSFWaLC^ns2edNTmz=#_N(`?hPL|>2_=R+ih*r1MKG-f33ZW3^Tcl#i;H*C9!QJLrDx%I0^p?D z7vM=5-r@eW4^`s({?Li|{l-oW0Py_*FtoL%GqtrfwKAqN{U$Hb{om6S|9)iKRCD*z zT6E?SZRk|bc8RIF@jukgo^{hlNT$Z9`Bh8Jm26l`kXm<73=?TGU}j|1oLnCWpq>tb zs72l(><7S_j0_JNf`}xLYo0($+$e4~pWG}iCT1pPRJPrAH90SaS$gn!D{D0EeYNR1 z-FBS)mUWf+)Xj4H9wFq!QFE^V?7bdx1cPUt%XWBEQ}hxo;=51OvU=At=cQ`mJsr{A zGkiP=j?zJrVMX~&KJ8Lz2QT&B!%=<8Zd}2EHu7$gel&2e7tSfQXJ@J;r z+><;D=XK9}kJa`jcm7%>(sMU!pb}Vr^w}AG~3K% zjlOQM8HGWFGhR3+HFczv@mp!7Q~#LE#ma2Wb2oZY(}jN0s-3Wzx4E7TNw#t$@z0;= zvQtZa4arU-jPd(of=0rtLXQSE1W6Yo&nJ?|rZfC=%O+EX$w@k#)i&ix=VC<)E6O>L zv?QD|CDA^$)!77z>PRwuFUDjwb{o-e&MG{oCCA-a-O`lykpuUQWi5$K;C{P;b%sKh zc;)3Kvw`)G$OrUfrZg2<=b*~3ye^k;M)d%dgOT={#wpiBB{Wt*z}be9j!GK!ZZ)~Y zsR6~&?8zHk_TQ{-rf?HwF=YE`&Qj6khaEdD(vZA^Gp5W>E*DI5oW_ib>Wbwo$tx(5 z!-uPXoJ=_Of0-;IUuaPAVV4}~xvDIdn_t4T`zyc5)4Y*ejSgtk40sjk81l})zt7VI zhQn(CbpGTxa7tD=-1+SDRzl* zU!}AFdO6xn#NOLVFRp!J=fJRmQFg=nP*kC z>-2v}kVZMLOp213KEggY-3y5VllGRvx-~HNxYV;cZ`{BEzp@Q~Flxp&V`(R` zV<@}%>X4hleLuoQW-`|b1eR}+%5dBtAni7{9WhS*1LWQWD5;%`o|^Y-wc)UQOAUx1 z%hti0ocM?Bv}Fn=moi=}wBxF=A`~?U;qdkPu#!=*P(c&)+=!C+%>6X(E|zTNMYvkU zcML-c{TkCT)=jhw*Ymu?Bu|@X0tET5VI$Z`nT!u%ij9FhU(S4frv;gDkad``#M_&< zRX4do1N*fTnclu3BbO*y8QYBcbZVF9DwK`|i%T2J)oR7{H42C@jIG8xy5I&|mIIyl z%|fp$#V5`QU~&x4k!FS1 zSC53`IDum^0s~TuKoC`j`sbxB89v1iY9AkTlJ2kg{eL+ZR3oh(nbq+ z8KI1zvGOqQ2loJ;3MGHHBVvS#gM}KI;kr}^@?s2Sccy}A(iPmO2z!x-Du@c=HO-zT zoBEMzF7Go3>Xq6WvV`a+6Nd;C$Di5$?xkLcHhE^-kvoMk~2Tv38Er$~+3CD!?8^G+Qys)~@OCRSLJU10-{iF0}#z{5dv z#r1i4jA8PX( zcqxD6=%Zj-jOK2PhKF>NphPr|)wS~M<&IamM339J;Cw0~pOi=y)5bD`UYMc#4_-L9 z#h*^K#J6YM0Wn`8Hw*3r#!l{A&-DA)Sr_?+m%KQt>(Z z$@NQ$WsWVPb3=3X_A27$?)cGWMuUUDsCVoICCL&}S=Iqq53k$QQ7Rx>=8<+KOY*lsfH7(SN<{ z|CDg{CkZEW(#Hs89&RVCOE>16%drC*TMjJ2>$EoO$ zik*G6j>;)AQ+6%dpyk!(cOJM!K8;K0<% zHMOJ`IHIu>lu{bObNw*BC~rwRXaR-``QT^nhQtgM~%QBILyF$ z&_Ll4zj7Id>6)9-iF<^>hFa@fJgLijJq3O4lYFUP73hoJp2B(J*A|#7cZ|h;0lEFq z+P4-_R5*58X!V~$LXP-F1DinSm1z>1=X{y_29gB2$$!iM$tB7JCHIIT+PAb%4Yt?{ zR4o{IV+vI52`V^TnjgdE0JCJyf1&K~6h~LO$o?h*c7#ia#!Ey7G&rSNwyH&o{mC); zQz#>cn7MowDy09p#Op~U zm!W~Y{tN{ZCu6u%xQxILtRLoMRMkj#;a#;-@r-hB=1E9c$=nCs`LUz0g{ z2c+1aRk>ZVvzWbCTJ6Bu-AKp}?wNZzKvKE@sh54Ln^8nm&*`{C#9}i{MVoImAWzGDL4s?WOYau~5(ZN4LO1%+%bm8aDzJ zy+A>H_#Sqd8&ZE~p+vNHU|^I1$k3@AT;TJ1o}!BVEDPOS;fH#3<@Q$RUx(qGYP^RS zqYZL#TrBPSQ*y3c@K}EFSdKV+q%rxZAzx`+4d0{+um1ZFCM8B5=*W*>RKPV%!6e!f zJFV(*>sB?w_EXj}p0Ou$hGes%bbu#tA=Ls6!k-4#>4^e{5Kd*G}htQ2)}y zd?ZBXwAV&~0?2(o1KRdV)TQ<+? zfWr4AjD9To<1;=#yeK!N0{|d|aicI90@jI21}KF1`W5LGk0GveQqeNQ zuKPWh1%tYenM2{}kqu9BbdiyL)gS&T5nPWI+2KdE@Nq~Nw;ZYmAe*$zu2A(lV&s(f z;4g&0$SRR6e2dPPax*YHLf9}hx!L;kfKq;3pAL=6xJ-b z`Vw}~rKFs6=uTjbzm6>%1;|N}fdpaZ#)#y!Fd;?lDiR;Ue$dkdXQ}Ac#&qV??MMU@Lhtj9lyr@5(Ez^a*YMjBBy13xL`a zxbe1Au1jG2feH2sNO&h$yJ?GX(*w)~4Ln2!7=@5INzUC1dXLOmnv=!cOY4FAZjU|t zD*ZTv6?%VF7qexOwZ;G5Gfr~PvvNk9Qs#(8YCa_X*=CRyAuqcDcQjkOq~QVJst_B! zQi?907QMI`5Z-fSNC#J@VF1xZqAONAPVbFI-H!84lBy2Ybp$D&v?OYk8dJJ4k{n%H zMPzXnDj-@SalExkqy#gN@fGAJqoCGmo|=4POD?BzS?|9q-xy+j)W$N-<&t*_qYNw? z_ooh?Ho@)?AaRlyZ#o|l_Gnm8vP)rtl zIi6vr3cwS78u&CPSb9=ap{&qBN2tM#Ve**ddX3RyhjOh?zx@{xduP+)*0su8YYQRD zSg^KOX-vgF>(Hca%dzOii@awyc9l8+fo5Vq&LQBWdD{WRd-jAvNCP%ggx*t={s5aU z4K5!9bzNmRzVbj|;Go3@e+pMe{j>(a?+ulG2sa54++1KrIr-xx$=WHj`%x`jmulXV zBvac*b+{8iUrlNK+}a(VIpCm>QuK#dUJ%%y1u>|UJt;pRvz}tyar}72Ov^F!*pU@# zo@Q@;4U@2E%x+kh`q?y3XA z1xf;EYn>78RZ?H+bCchQf&_XKxkmDz*kjY*tcFIg6gtv;yOL=Rnw&V;|PLbJCg1cxBruI~9c38H00c?)gM5@exk~xH<8)XNhUl$?|91LvuuAU zcROU!6KpJEBd*D#2cQ=d>qT@07jf2Ax~5*%(Z|U)BZOuL?h{;F+#N-xJW)GM1XLdw zO*(H%$YDiER$4-R3gMwTDHk(pk*OjDH&M~j(rz=P-j=7gGpHZN|P9W1^EHibaj zACxwE4KJ)~6FJ2=t(Uv&=5*Wq+O3bbS&%(v6F95)9ow&vM|~M%g5f^iZh$`gyQLt< zuPKlLOuI2?!^WD5q8)!*wrd>u&3~=n4C|P9R*CTgEzEP^F~S|h#)A0cTWH8a0D_PG zA9s=*$RHv&?LOi8s3JOjSI@)xg&O*WYki?mX0gB`>GTW=Fp*08tmxIKk@ygAgzAw8 zpaM;OVmi-K10!g^;^=^gbb&-Jfke82A~%2|vA@3qi{Ju?;DU(Y0*c^*{x!k^6M6TJ z=<6NSeIXIit)ROal6yl^z8~&#Fr=g{;QUz%QP|!pyCD zzMfV>@7fsC<;j;qDjXxR&L=&YA zb&r#Qa+KS2RdJG6Yj&fws%wb6k+NXVEaJM^^gUpzHI?L8XWXZNBAI7il=zQhU!K@$ z3#U7}zM%fK3-*j4apwB109^X-DggiCFo{yMmR;b3_etg~RizvTBqRi)G`oT=+o8fH zU^o!7ka@8qgSeSjNUm3FF1*P5q>>g23;P7{O|)+g>BJ#kF6?NSH_bM6F?l_{n%)LD zk{{j0EWT&7CJX^1RfmIFAq^)_BR3rCB?4}MegVdnqPUq=w(ml!-sM?Jb zcERTsb|*s2xvLVCm%FPYwo`gaqJLn+K%bIlDo&Y3%@nz=8qzW79DFX(Dopv3uC!jMFP(yQ@B!wdK-ATK(T=4FlyU*5F&UbR> zwL*csdNFvz>nPG&SYbVe9lC?_DOENz5oEfi1$A=;UCZIh9_5^b*F!CayMeOBVjdJ* z_M)HCcKNM8ee#$s%66aOggX7KGh(O_Otj;jH3zBP(t_U=OkU!o_Q$W)CJO@p%Z%pU&+AXHa9K zxd^u5P1Y8r7ui{nh9D48S}c2s9ZF!kU+7%Go6e&r`thGj*dN2%A0O@d02d2P z7kMpk?sD@Z8o7?ak6vY7;3}A$LaL`=jJk??eoMGOK+4cF7DX<&0j>O;+OTXCYV%@! z=6hkc2=zs_dX;Zz?QD#PufaSnIC&+XI74XuMhY5CK^-G|&Jx@kki5l9OW-zf11DjX zj~k?xk>}JGuJqZY?EH}lWzJ|%bGGao!*b=nOCeA1{;icuXmY1Y^!;2z|1B$|_^(>I zat`L!-?$SO z)@Fwdjp+je@4SY00|)~HsS)!c8HhwG733G_OCIXMnVy>Pb-k$B=Jf%xgGQT2_oJ~V zj-tg*`blc9wHG)b9;S#`LNX`{d$=&m5sEXZW&_b+9H0b zN3?boX!pfy9MbDp!?A$tV$_DIOcj2ty=Bi3P{Do?*`Ks}?nBj!8TWJcq!HR4HAhIK zh|F{`z4>he5RYRbWp@w3s-}!nXUi7Bre+bEVa29nX2g{YYNm@ajVk)O-$b!r0k*A* zOvctJn4jFOYP8>nykLea6>pS9LCAE9;YF2%wHs5dOFYL_L(ekFMxoNq~3IfL&C?(>F=95%cXsOs&yCG@m z@E8(9QEv?-G%;m$shl#QOg8yX6p-wK_h|n+D*hF9I|4gAV|zeDAkWWTdguEkjn!w6 zyWbSsX&r$D=*2a`H8!snX4d$%K_aZ&pq^E5@U{@`^b#sDS(>Lmys_;8n)NxkVKjbc zU5pkz9w3}Swd)YtdW$T37U~FC)Z`Uy%~U86Ze9wtBb?_m)RrvH&Iof*t-Mh=ih0|q z%`!Iqh}JZ51rDKBwAGtnGzRXVZYc4SA7{oyub*%NW*?tm)>AWiT+|)IlS-r`9Xk`t zk{En=(&t?~eBEAqNU3x^@dn*oNEh;!+LU4^%__?gXFN>vNR=WDF#km{$vl#%k7+Q* zT19%ey!T1{iUO!;RKgI7p>ZsUK1}^_Mypt&8d_}tyIdQ31_2{ks2`cVkCPqNTn-)C zINMzcV0#W2?{*D`C>@m?>BMTDX84C_&TtH2H1##i!(a>vgCR}mH1$U)BLfVprY0W; zn?$FeOC0eQNtooqGk3gO4lxgY7@T;1e#`h5nf#8NREY1tF_bkjxHw6^w+sAl(aV2> zlPhWLu4wE;%g}Tuey16@;8KX>g6={Ib)*6Yy;jYJ3av`Phb<)&+}H9vLo#hK zjb-oJ>n*yEJz_ z%$50!+@@5fp?$mADrXc!c{g*1(GIb0ffUYS=8T!LSGowRJo?lX#*0{!<%a1S%&Jtt z+^Fhs;NeGf;f;XnVF;|e0Bxj?Ub31H?72)|$Ue0mdw_!)7 zKVCR%QXNs8ah(yK#4H~Bm-~?B8F4xy*_1}`UrzP5<`f>!WAt^O|8`9I0Q3|4i~RbR zKJS0@Apg&)rzmavzYVeJ$qGDUqIu>`Sj5RnT}5Lgzm)!R#42HDRdqNYuFO`W4fOZ} zmxCaJ4TRwZr8);7Ly8GiG!0MN9z_jjU-9yI1J(vyV4~W!L{nb6x9WoSx zQlqDSC_<&XyAdk{=1~Z&yqBI>j4i2Gll?6bM;Xvl(Rc-c8N>5`#PGuQ(k zyoi6EbHm1^phY%YF`E`@&Ci|~cElIji# z?sJ@wj`9B-_`xEE0*$LyAit3GS}f#?qYC~o`e$y9plTTJx5hEKw_RohGir9Ps#bx$ z^`DkN<6d=PW|3{k>iO&|j`{;GRNjL3SL~#x6hd#uxOb>2=DG6y4L)Ows@>uY+g;Xg zDBf=lOqrPMa>XIpe?8nHvlXfT`L0#~-{*q=GC1TvS`9>vZN6;=hWri=`tJYcG*J3G z_uWOay~rj{iMA*dQsn|v4a8PWD3`pf>{X6f-f%*0sJ?Y#X|oFA6UoPqj>{~ckD1~1 zPYh%Bri}#!E5YS=x8Lz^==W9Ko*o}C`hYitdp$;f3bV~n1C8G-PxSlcrYKr5eCDaP zt#gy~Hf?X#ppKB2kkE|yvX`HhoX43m2K-Sh=yxlb3Z^=Ce82f&?K4@R5JaakxoOT2zx~_9!#pM1cVf;s zu{Jaev3{8Eku(ULqLUOg)If7+w$tSKCP5h}s@GFDp9!G2S_FeMk&9{5r&>GSj-`}x zR9U1e2UsevBd611+d!C@l__%+Z-{X|&{3DWmX6_UEB)LmG~$?VlP36Vo`N3ZnbCMm zh_b#NHO*F>muhK7BK>2us-MIbqc&N?(m8luCE^NnDiwzut?ojttyBg$OwQa0f#%+1W5blQOxs zG9Mrv@?}sG1;dDB;=}bT%-b`h)J(Q0Lt`~V=)=1}b^$xRV8PsBHdHBh-#SG_Fro{pK4GBbE1iHcS{hN zCENnd3oH1aDMkM*y9$zjowGkl=~}7<1@U%a0AD3A z7yY3^lu*d?2^A#ZqCB>LQkh0{Xs*KcghQdhW5b*JARBZ#7g4exV!N0;uG@C6rfq(& zm@j~9L`_-?x7I*LTz0qf{BSZ@4cKfBH<8i{s;W6)0rwJbiPHN?A;jbmiSeg~yShkG z18jzRMrOZI8R73Bd03UXG70c4K1+-zqB!FC7SI0P%(pxSec}&#hBZ}=q&n2N>3Y0_ zgDkP)0<|$B`?{NaQM*)M0t3{0*I?7Vf@`8Y!5-V$N`)G9YdCOX@X0+x9&DuM8cE!k zX`|Rj8^W}Oy)eioG1FCHoi7^mE^Pe)XBY*o(1=Gwf871c= zFYK5JE5H6pKC=#=Eg+A|8?~a&9_=iT`X-;Tq2(92mm3-@lqG2I!cORXYyPeBQ1mb! z0AMx@PM$Lv$&`hZ2h*AaIy#@**u7{>fz#Txc?ww3cxZx;M)Sq*zA8*YA-@&$<~NUF zE5~ULobzdjij)DI=J&RhDHlO5YIQZ=J9Jk@*d@>@vCCUU`(}! zv>Z0>62uRfOhRV-3Q@}#r%Nx8Gwi(k-;Of#sh&VNcELjrZQGDj-x0(G-~KK8=trP0 zIms=<7pk$EXgPOZ!Z}ksil0XlL{;un?n*~X9JJ3bgtW6-zm2VOPajSoJ&oT%{~zyt?r)n3KKD;A5dgSDY5*0pD7`Z0Ws-_rc4Y>`(m7iYIO)qOe6^bB0SvslI#`-&E$bLe&zN|Ner3h z8M`hj7I-=6m)WW3m&Td7BvA#7ym2@=jE+DCV!7?DMXU~tqx{E{$14$RHYQRfs=f1b z9HSi{Hk2&q{L#{Q?8Fxk`>}(Z&P1z$`qG%qWlKfUlj0ef=6cie6}D@Q@pGsrmBi;H zg0StoE@;CJ5jHM0&<^U{PMSwW{eUgdWz@jY`Xz--mn!%twZ=VhOo+vfk!O0uZ5*5$ zl$Y5+OV&(DCcImLN-e8-BsuoT>OYy3Ya=zc{@4%VG*gqH*Q{u=v$dGJu#U^DX=HUNJMFza2z?~`Drln?Oq4<%=)91xU<)T`RT^X8Qx9|aCi7iJ1`Qt7OezW7y?qO&S+tDx-P%*`pp}yYVu%jN68SryF z9PCWpW05p#psNxr0<9J>!>#+OIcrj`ho-d$tJq+oSUoxp+$+0;?qQ%0d=PCk zo>%#WS=szJAJnr`v{`|kaMpH0ml<1HVcZbBl9`fhNYZih0R&^|er%3~g5Mymqk-hv zwDj8HC3O4HYg#}I?5L-MYbsAv3Wb<~HGx=^`@v(ad5kutkrgr#W*OsV1R@bP`MBMn z_cVGtsDWy|idnyVQd0+OoUz6)-r0tVsC9PI3IqZycEu7kp8-|0d&>ik;X9GDsGbUC z+zYT~PF+Dx;V#E>Y`8JQmx?l-D)~LxONtsM4;u!-Kj2&BuJ(CF2TNV-bcCybV_F@_ z-jsa8uApnJiF$Z|-&a}w{T6KbYc0FzyI4~Fzl!C5yw9k9mqhyU!FR>V?;|J<6- zMKI_U0IYM5Llg)ENrw4h6|l-Mgpi7l3PtCJPHJ$j9~>z%9G+PSi^z0KLC85`sZy ziO`?oCM}skQIgZucp^N(v(G|i=mQ^<9=mtgKq?}^ok0_PhaVtvat1su{;BQM>)fcqwclB2TgZLoFKgfvs_8}x%> zbCRp_xVG}7zfQm$T!8ODEwpQ*KK^HEdPF(cW(D+IVpkE9mKauYx7^C=LYA@fd|{T| z+LBnPgoBy(TH=N^ku2phw4r($vX^jRRfgyo|QrRh4mqU_;q za}p!9sw?%Pl=IYFy`tI?r|g{^#~+zF5iV=xL8Mz^`@qWS*78RKoy~{BF)@1P11Hy_ z_$s2U4^$pNSyhGChiS-i@q{C{pv8d6VXgj`emT2j`TAK^(|5~$WtzVDdQ&ooNm)EVcgHWJaL%dn*s9SavZp@v&#`sLRyOTuoKeNFQOmS3N8nlO?%QJ>?IR zS;XSi{B0W1u7p*{XCG$J_73E+MG=ljS1r$04JeAEu_LIa#N5kiY8TT2}E=KGh_pAyM@et4r*aCkFMtRLP0JwL*9MN^*$#%yyV z;Br&o@^j$y!?y>xdV-XB{K?*7nR&vkd47b*<|dqf+jy@D_H>S?3$P#{e8iV@Hrgv(_hn2^pqIOr8J!Lh^q*yp9Nbq6A-;2oz|B-r}NIV6od2K{`6~C1t*K|DY%3mtC+KXmm{@>Z`(B|_rj%Nh9ry2w`SL^IvIvBK)o!DCEQ;~ z5aOg__mn^uP_9ImAa@d$f|dJ<)DqvK#8=VHd2A1a&CXlbDhH=gLBM>mxb>39qJDV| zE`(UInq4l^osIPlfZCk_#2j2A=ws&W^?d8AAm{aMHUC0i)Di-fkuw(F3;R{We*}lH zuTK2TvADbsY21J^MtrS?Bi?qn`D*PYWEd-l@akF&b@BCYRC4Gr3niNGZjRD-@AbJll?&3FHq8(0(7PP0y>a=U3RIc5yCu6qU^sj`m38V|n{j$XZabBeViBF?u+ zd~pEG_ZRP&n<<(_Ba6Lc{Y0DVCHK?RMB2;E-8Z&`t~QGwQkHQ+uI)SwqAnQ+$QxzIh4 z*;T2XpGIBDD4A*MMHk(7JfKAxG}Z}#QWOYf@v#wzi0M6jX$i_KpZDE=z`h&)2Kycwenzj95|0DN@;qBe0h%j>#YIKAIVsjjj$2dS`}M4P{fpFZ z0B#5LCIp^MCvHp+Q672wZ2vntsk^!74qLw9efjBfNToGZQEuNpSX@podpgX>8DqO6>YU)Da+J+Em({Sa6QuIymM@7w$ zzrJ~_*iC+Fq!8+@yu`>YpD-8BEj7h4(+s7gvE5T4vO4dfJFQu}qHDciJBVmVjbP4|qbh1Gs{1Wcngb^dAJZ{XlbN2=VRZiIb$B!e7TjOL1# z@H}UiVd$){ZAL*k`9yuea97zu0y>2*wzOgc!onjC`CGUpD%(qHM*GxwJNWLL@Jno) zYVqY7I}s-?=V?5HEFEcJXjZifW!4N_#V8*zG;rt9v)K*&tdN{u4zf~j=FAA@+*3NxZeR z!ih$v!#bw-g67`)YHCEXDd1qRi|WfQ-Nxr!FwchDV1Tr;sf3m`+aB|Dg9mO?wh zmA=E`T9m=E!aWrr%w*+CVtVRFIwbVp4*py!${Q;7qQYtX_AfTMdsG}%p3YObs#+;; z#=}-E?iu=J7Vb0!yY4U{SW(T#n9W#r-6~{!c_X>jD=^E@ z96Y}_(V4rI-rCjHsjK0mWeEL;GkC-a?fS9nADIL}-6*!GAb5>bjE>Bp8Q12^GGiFK8w+wn3DFyNchtaA8 zDiQ!OU3&#{1}L2fK}(Q&Cq8#%xgNy);PX{|9Yj)>Xwq0gX)V`#4X)V);Vh!sMasWrG=e-WoloIcE+X zMfFu$Tav66XWwlp?3)3j`Cd(4Pf* zycRcS*T-xq9q;e3fe%50d|DT{Ib>(W`!F<0V!9ML{frduv$li=sE=8AvmDZHXF&H1XuZtZbx zo(9UKrVfLyp+&1PHpUP{Xcg{g?xDF-5QYhF@CqpV1QUKC{z0*_-z93my;s8XpcI&% zM+#u>mT33|M2qN$Pq;z9sOevt9#iVmcbzlS?d~@)PA<6?>;u1~U;l^B@ud{;g@$sl z1}uDvpG25p;NWL`eK>`BKPbDOu=;R47b2y!D+EyygVwOYN=6WoYg#}DlM~s3Wvq^& zgs=snMFBraM_^H@(Y7Kf;VzJP)E7-*bnJqnK2Trbd-QAq<}wP8^k5*@4`tlp(#$?} zcOXXxOn3MClFp#+TU3uR$@hXrr$OzZB)czv2cXG9p{&*)1l;0p$Km(9|5^%a1XfJ| zzKzBG$N&IL|5Yh4Ft;(H`)l3*^$f*tq;L1lzxw<3QKb5&h^>tJdC6s#%ACOmj({AM zDuj;TAe2ZBff+IY?pP<>m?Y#{wYKe|czE^6 zmVD@H%a-M5lS-h#(sH%G;n;nZ={Vi>et+is38lx|Et88<9NYjQM6VcvIx#jBKgqs2 zF-l68@dA?J|1EE*{Z)d9?w`d6F@;LK*QwPS+bgWK=3jDyMsXXt@wg1~8Qf#}EDALF zT$1>e5^T!CRkG_2=-nzuf8n%6o-`5PyA&cZDA2BEHxV&){Vt%BOJIm%a%=K9=kLBS z!Y=$4&zEnsmK@iT&1&tZP1LsMWl^kFsALhmt~k7E|Wi=6wl)X5@>d3wLQvIauE*Y zy#sfq z(V8Y)sZ>(I8{4*R+qP}n75j~?if!Arom6b6Voc7Q?w;=ReQUbsCp>HIXJ7ZebW4ToQgMsES(386+L>K$GoOlj3@1p(p`EVwXNW>5#s!$k zG=){#p^z$(mC7krPK;A42ULArw5ocr*I{fK8Z*5gD5(pORhpq@0 zqdlAe8>U2+#MCwBDS@Kn5hzEqmdeft#j;IH6qLUBitZMeJIt2y&#lubF%L*EKp_e$ zgPc}?Ly#06taEZwVf*zSUu>yclJFsc-39oKbE}F-z&MGykTgzpjC5%uuIy^3z^`bD zfRmA|F)c25OK#B*EuCGaHe{|U^p8#gvCy<EfKLHH_z=I7|SrbQc*ecb6GH za~JB*;N0q4$K0B|p`|@z4>5qbI&sItGk=mF7JE{c^(92Pw|r6_HhapX3z`Ih<-Q`` zr*rlWum0i5S}67B5QcmD)y`?eQ|LY^$-w`D@?MpJ#k-HLi`oFEj`|?L%xQ1jHQ}mE zSLMyMvYjITl!Wg6DZn5l)d)`2&CkaBint;^6J2ZK2t^_Ld7Ua5pBK+ex_YkBHBm3_ zAP7U#IsV$zrU-2Fk1T#-xWE<2L*6M=1|h1G-{Vc}p2p zClWA}qoHnX-v3ODQ93rlkIhyu>2DRh=X19M)#uOU8FDG6R8?$|nJI?y2EMI8n>zja zZE~w7zMTEk1FgyWuGc$?5B1ph8(&TMic+8R?a8zp1TmOyp3y6Ucs@c$N{TKqXN+3o zWyamnOOpj+ujE`3EyM8=nWljhh2{Ymg^m6uoJQ*_(Wnqi_W3B^jVMrJdpd(HF@DfY z5O5tL?N&5-o|J-f2~UK@gn_&=veH3ix%dbrByonr6ZC7BhF8u9cTUd4OsNNaHID>^ z;(`~Nqr!)-eYcWU@T@2#9ktp6$z`}1vG(e`52ArVl-9M$q6f;49=Y@F70iB=%Sei$ z%7{_?O?N&a2XwA$d|Ls`=G`#Ew(-Utri=Uy(1F@Ob2uf|yq04qx13>;Nj)6Gbzgzb z5CY2hp&vj`W?aLZ=I`~OYQAWA>aG{h!Mc~F|ZMGDl_AhM~%0?M%-PMC1!DyXm^Bh z8lSe9oXF4_k|`F2Iy?)+xQLJ%od#aB6}A#T90O8dTXaebEFB; z5Z9*V{-EjiFZ%Roq~|o?Yqlo&0@eTP+Pi-O>i+>#r~l~fQu(K|qbRY8Ix>uW3rx6{ zO+#53Ss0#xf?`XOnHzRpLSti!b-lC??)vjsN`hG-=bef4G_VVO+FH#TpTZ_1i^=OD z^UC{(v0d-e`{~CY@MM|$opgUlBvw+H0sGA!7*IADV*0`>yR?>QE>ddgvx6#sFK|JUrFRv3d`th)F=_eFITausMoBxP-^A(QRYotAG*r4W?wwbm4f-ljvbECwF5ihBd~+1;%GpL;`3qPV#eb) z#VuCh#S7J~n zR9t;sE?Rm39;3}d_dCYvcNEux-E}Wj%=6q}kqIn!V*`K`YKZ+LAY_-JG}D%Y!(K%P=t*L3G_BI2j~50=H>f|@w{|}4fqF^PX*?@E@Byr zt5F&HPy$scYY`mL{oz^Qhs(T(4|qSr+A49{QjuHl);wtlh$d3*Ef#Rzk_A8?mOz|l z3$9^t1l4v72k;TL=1~$`rro_#p?X+@NW?b9kA1F*I55!LeW6Np(931!mKuQU@W5e(S&07l$>X%QpoC(;kl zhA#I_lT0Ia zwj$D=nv|qdkl|XdgvUK6Lc`_qvcU09;&TF5VwAGYRZsn=>dUnLc-t}ke#?9J@h}a? z2cjMb4W;^A-t6omkW5-=Z7i4Pn`CXP+q2}?*!g@tYhjM zrOZ6xu04X`JH_{ z3oetL=(>V1dA^7k@@qcS*KFi>X?guV2C7UwGZVMD`mFu(9|E3TX$DR8jG4xgAeYr9 zVw^nDKCYH+8B9erH|#lBD+8r^LkkVTUlznTkR-UY2k53Pfn)>w(dG_2eZ+>vs7I~S z`tHfwA-s^!K%R;O&sFIKo`50+VQ)?IV^dD?u?Hih=VjSuXG)dk1o{efBik;G zO&}@;X=u6ZPy%Ciyn}Q=-G#Yu3|ELv?L&4En;2qvX0Qwblg^0;fq~cW^)QX}1j!lF!gW~^ zLeC6>TcvE3RH|TM6ql94Jn%-wY=s-@Y$XS5GG*=l2V&0;Qo8qKP+#^NOD$$x{Hj1Mf>_;$A5G`iyi&Q`@o3a%| z#zvFT_kr!@P(EmT13Ar>FNSc*K4zY=EZ%66pwnDkA$ul+uUl7dgM~WMET+7M5Ozaw zY+7q7wE}ErN<>U(c$n)`n!R*5^t~rqOl3M&kI$k+7If<7YcV)&F-6zz55>w`NxN;x z@s8VR&TIU9QGT5;YsrHs<{)jQlY#9%X9K!|HjtD zx3elfwNSFC;V&-u{@iXEM>*7v;(jbWFdK*d4N?w~wsoOM`U-je7&qA`Y~3m{qn`Vxkj}{5YUAzd|=qd z9?5TWuOj_E@qaJrwS{ODP;LDoV_zrA)Fb#6&-^!&Ti~6`e-7%Cgw|ltDr3PkYr$#7v73UvJmVBHVN z-BT^Z<#Vj>b=9XK$q~bI-rw2s%a_d+CUGT*L_P8qT_X!_$z=im)`0^~g{MMZ-C^{a%r|>7`z&o8nJHfn+ z_@ym{^i5nr@aq|`_5(d4*k_MryKmTd(+49d>u2MYsW{^sfQ zc*CjtYQqV58TavYRGb5HK6pgP$)NNTgG{6&Na?$5Bzu@f9P{iqZuhkE=`eRNGtkt2 z4v=ancbadI)zI!jJsIDpBFzH`^>z8B2TUiP5RTN6jOgrJg7DDTfrrB)y!qkgLxiEr zB0pl0^&#jn)(nf{nkuCySSxQ>;tZypC!|FYi_&Hm#R+XN5hiO?#^vuv+y_a709^)? zGv^2)K0GlUZI6IJvyG^7K=gci?v!!-T^WkXq01bzTn2*(U%onK_cR610~pOD{oifn zyL5yYY_dglZHYMB33_R*T9)}RU5QeTgMz;2YWqi{tH|;Qp`mBw=Sh(0!dkV6Ws_(O=M@350yT+JRqsV7kz?_QvnWIe7|Th{ z>7bO)GE9kH&8!);Q48>Hwd9AHGtpii_G2BGK+#Z9i2Tll+#P${BHtxd!$`8kL{Lin zL3y@v62Rb;`hbH6JEwZh!~UB{A0DR1c}ELuWk#k-4wmE2bSXxc&%L4V`E4mDAJ(Eq z5A8-@6HLWmgQUe$WBG8ypPJFK!>VeCJv zaWlcq#OFP-Ajv_@)*+XaJPEPA@j(V1_&bQ!kSPu#Ju1FFT0LBdTZ3Kzt>v3+Nvo{Y z%pAlQ1jCgsUj=y+8)Z1a&R;Ew1;xehzPK4y`@5Pbc=IkSxyMGTgbB(puOkPk>5g(+ z4=ni_>vK^9J5fc$9A=LUM&VC;b%yJfj3(nv6i%!f4Eu{_BCKe}kXL)4ka)NJzCkgN zimG#UgfXbX+Pd^KjBi15!{8CcVZ`}kqta{1M=F|XHF2@pxQnsc*nDZRQlP9oMD{v; z#ZOUiZ0JDlINgaEWlzQ#EddECgOZCKI+!dhUK!<2=b{|li}(yLPpbP-Peu$X^Q3zy zB}?TN>&ks1t8$2;v!FC@BKDB?B>u9KlQ=PM3v^kP*<`BNBZp0TwC?G@Dt0m#T%V+) zGm6N&Z>t|JtuZ8(CVQf{`^?U>HK`-eK+56D~vxT@0-TLX2=NrF^s35$d|QKU`<*3nUyCvZE5PY*yj<>bC_g`H%Np zdaYhk*oIHo^rK7tyA3GKbgg=GXwQ1{X+!AnNde404er5%N@nAmd#I*&#QS$q+_=wm zjAR0E9}Q=~O8{70PHwe;ys{<*#5(zpP2>OwFauu@Nc7@lK7SJI zUYD|Ltkga=_K0?Ne?G_)x=cq#SCkH<1;jprCkQK*2!+Nd#AJR5nIJY;zMYCG#gaRY zuAFje+8gU-_PHM>?MrbR*v-4R9*h!coss&m!AI}p3@vX;9v`LLpR?SbwcMY|K)>(9 zCtD(WZJ@I6VQ@1vzn!Fv_DX(*gsuofz!0;3L?{1Tp~9a@r^-beJ+ctvy!jVp_6#6s z^pC8^+E>=&f558$XA*$n-$hwvEv0#WZG1mo9&ma9tjgx%CE|wM z0g%#?GAZMU5f&3Ase_pju-uB*>SG>Y&`{a*L1;1tZBl?CT@_a@K8 zqp7C2=8?ICTIi5QEB2zdhiC$~?FxEx3xbwOxYE=V}5} zz^pKrZX|oHV1&f2j)nMyz90Z5u%(~gf#ib$z-$qpyF6~V=E>2thu!ZlpUotXis+@2!&%( z?kT63Jcx2rvWGCab=i$RNvzl7*G6_0P=|8+#mRA5S+|KiUCzO?E8 z{*C{irZFdCCYJvQb*lI&VTqvovi{VL#|0OJ))dZ<#uPkGGPFSvwkV*n`MxHssp)4E z&(is@66uzad>X|!WVcgX!(@{hhw(nm_a-e{e5Lz~v_Qxo(0;}9dilyb-F3C~{&-*A z4U#$N#z1@%hkLk--s=r8dQ#Vd_Ix4+%GH|h=7oD>Gfarxq-Cct`%*-VJ520Gh$ioa z?UsdGkwO9bA6qjX2H|87$Q7vgtRecK+@5`G+qIW2I>rO_fd>4W4gvQVNFv}?8j?Gq zRD}99&?Pg)n$>(M8)E5!Q8l4(bju_0}pP zLL9jK@k}})&%T5xc9_Zg(Qm>sYa)<0cdU%#cLm#BO!Yl^j=oC8YGcfzn1m+7t zF{vqZJJt51NMU*m0o;{#~-F?fbUA;Dw`&w<` zV%8QRH=tAD1~1|_=%RuQKr^~(S8?pMZ6xI*N)*}cf{ad)PU(JaxYNihoi)ruYOlwj z0$r|OTq?^W$Con4j~S=PEI08!;Xc*qr$}V#$^N@_l2LNTy7skpUy){W2njTkw)tG0 z`6{nj71J@ZrC`k+EaVsWqU9Luy#1Pfh|%K&(OY+Apkij` z2r%A{8RgM&V&^MBVNLM@c|^CBtAano40Lg5tAx<``WStogo@Utya;|eKv`tPSGZl% zdUT65HHI2Pq!G8r#8MXtrgPMt5MIKs6tUpcw@CdC_BXuLqn>dJ8cFrjcg33hYcsIa z3@?Xk)jKCe1Bg%Z4hFxKY{1h(U+cT(Yvt=tMB2f)PxzH1(L;yM<~x{qgH?sm?G{FH z8BWnRgh6O#@Wr$ku7#mGCC8Gle7>AA9b@4&bx$h~~ za?}43{U)j@3KsmjB~rdv_WxIa{r4D{^133D06b4WiO>>QfIEmnYZ(C~wE!}67~%H- zkVO2%7%9e(U=!BHRr}{3egkIrAT*VVIq&4-*(Slrwcm*wnRs|MecHU_uCDmDeZQfL ziV86SzVIRB#9dE<$kOBY4uL4g5Ie4gd5;kcMU<6!9}*CHb{P?gF}%szdD2-y?w(9ES~V08=se;#FrRQYb+ziH z{?TYT>=1!HjdALxGfwobzF4rF7?)8tp~2F=MB!mpfb@j40klYYd%VuMrQ2vJ{3{dp zR}pQR*Is!wlu1=L{mZ>q5s#uUrhDsdGKTb{(j+rJ1BEI1)K-dv0z8Q)&R)~hrr3P+ zE#R6p6MHxj2%*WYdz?B17_I=Oq|P(MpyE`=9;(mOhLMC!F}3j9@MYCfve(n`I9;ts z_I4h+=kLKLyENox5jb+z0d4Q;0oq~r02~$0p_$Wsbs$5taR-~88sGE%M)vs=`n{~3pm9?$6&n8Ps6o1k54Y<=^5v_10t5k=kEJ# zO{6Tk;<~mXeVbXv&9RWP^Rn8%;hL2QVt_}o5=Y@_2F&BU7z?I_>TFy?GILQ0MsqVs ztk>@k;X#-UbC`;fsDmi+e~Cb~2VK_E^3XIWj^IiP`DT##geT1we1%gBI9>Uz-qnI& zQM>YYbatWp_(tgjKWzT&=fc~O8pLp!S4&18HyL)G{9?&JYPc~ZHE;kiu$G} z!R8RTyGfdsPs%uMaaJINL(zUrko^;KYPJe-|7L)op)uiw86_c_a9{{HtK8s+j@@~D zcXr6EXc3KhcOpkqvzCB82v!BZ*m7MUZRfmb&-cZMe-c9{Ix?r)Zo(fd>mlbJa}Lh&LcP zSPHS5`(<8qij&9&?-YIrH0#tsIN0gs=RUs!xu~#ZE{P>a;wGCxYZUx2!o&o#Bh@YZ zfoxtNe*w9A4b3pD9@am{8!~nUdGA4fp7!aODyar||<2VbC1Ch4skAdH$4?#%5n z;7CiEvWSzp@l1@nZ9?GI+TU2<6lwRuEu)wG%9#$i74oi3IB;0T2y2s$(>|Cz{_14*Mn$urH7gjge_CAtdMU7kh+Y6##g}9u2S~uy!F+9#p8zR8 zl<0;qk&n!7NEx$<>nO61|&ch(7gV>V=jJ$HzX%f!YL5uo2X#ax=(Zc3GTbBOMF7+|xFQVzpV0pZXb(G9}TJD05$P%LzVdksHjF zfK^WSj_ioaAMUGjq{#KRD+-u-S?om#lV?}vFrQiy7qPoy*^BI|aN04WgHQ!Rl^UA4 z&JwL2=37=vu!=oc7BfmY-O$4sjLYm3Q+jr@4hcYgvDsh^W9~A|WwCyd<$IAgrhnV5b~R4c#L)Grof z3mO29JxDz<#-ThR9S(NFM7(y1pJ-8{Pt2~GdNq*#F9_9hgEA2=e?09R*91OtrQ*;m+ z{3JubvI$a673EnR^EO47%1LsC}U9jQ984 z&83;ZA7I=V?kh#*Qw(iLWQh7~1}*chff=+vM|xw3a^IB@d-6AzEDiejO*+R|CdJtB zLwl0mwA|^3=Wo9R%LAK5aiTwmdn8-11}fV5zp+%ucLaus=;$nm&V#cT*;H}kB2PEW zQka1j`UY;hQ{c8l*VM&>HO=zpkVNtEB2_!C+)C7%b;L%U#Ox%M;U;R>^R7O1x`@4H z>&gx~{nJk({k&0DD>0y$uz6L668(&}<}Kk`15G+{Hjbi(8>TDUcsmC`6E+i}XgWtG z`_Aasc=%=N9fxV{&6bi}5TSk?<&yJx?f^jH0@=BLh?1hxmI^cA`QzP?o%8pGgSm1! zZ!&!%?V5{Z33 zI+=_i4j^IYwG*78cCEaHh=|z4MkM<>1k-W#4cj8`9Zz7?TO)JT&atgP#f^h=x+?^Vv~fGJA-D3DpFWyFUu^{}4+IOCX*=akUYi zND@y5yo2c!o;h(&6iTx%e3jX|i4!^$ZxMLeLHB!BMe;D7(=w-@5Y3vEMbh0Xq`Sp! zpHq#FXU9tq5>T<@Xim=P&?WE$Ue&Nkvx{&czBA5V|4UEXdzXp)&=-W1Lx1}w_}^+l ziYBJkCPx2gYx_@>!@mhY|Bfkz$**$QD#*j#DOsBap}E{^V9hQF)J-+~&4WKtYLW?+ zDe3$JbaML7O)yfkRwn@B(bMWi3*9{*wr{zB0ftHxJ% znM^rOm#t=pSAM>)pgr{GAqg@b-u;KBvFWHjjJTgQ>&1!dFo~r zNPH475v^$FAzec5R02L7BHmHlGN5U~jh-X3lj^@)rQ8$t79Bk5nxa=T>LUamp}Sgc zIx?0P=5Z*N{mEjT==pmYe=ySWb&s%w@V&|!mJm*j`)AJp3jGd)Vb~h5sI8Y zDFYc#Ov$-v>XZAA+#Js~v>29bG%a#mWcqbkxTPkfA|t{G19|t@H!>PvwvU)QT?54L zb#s4H0^<^5Dk7i)51j;SDr49WinDi;=C4DuQDS!&PwTHiW3XDk8jh2g+ClZh(diRy;PiD(Y46SFp&-FoY{ z0;_j%P_7ie9ZRI;zo`?Zz4*F2K0@$Gl7nu*i!cxE=d*m=M~b(n*@l89ib&7eem z*&rZx2zI&zDxeUDtj00kVv_0Bfb8ccYBHD&Yu#e6Cm`3er9@=->EMZ%y3LYci)6%QawuSVdT_zyO#DGd>mYyiLBJi|J5Q3F;!e}+oyyD*) z8&x)Ze|b`Vt1HID_o4)?E0)B66}HHc1h=5PzY>K*&n}$hNoNa>$=&A%p~iqin~9AQ zfr22}vjmZ3+*0Rf$B03Nml$gJzOb)PSi~*7M^~`;>uy1G1hpnq*HVvP-B*(*H1ZdA zcrhc#sSuw<7(x;tXy5#oXh4`H)CFW-&PYHFL;M>GOyFzS)I9z@5A$gv9>OQi8V+X2 zVB$xd4>L~?Z`nYSf36QRbVcmIX{;%)C~}UxIj;y*&g|pb5lSkN$pnISo)4*5ibJq` zUIU|aK71wHt_M`5pyn@Nh}g06&8|&SMYQ-MmOTpHyg9NF30W6D&yH%!Z|s%f(&yan zf}f{8$8!vZHXQmkDRX*rN>{VFz9M1|jr-E5m5Chf^guOYGqp4()m>y~1p&8^V@y1u zd^)HOO+n}qv|2fM(KDYQA-ErK45CNiKYf2mGN=l|jTsjgT(cQJYvsMkD!QP-B$gt^ z2XE~pO7_P~)-V$w`vkEDv`*pXL+-RqWI=AkKC_1}+Y5l_`q#BTSrspGN$z$3S2{LS zt@us@_A6tC`#-TW{zEbUM@B%=#Na=gFe=p{-M`ioeN;2gVMS-7jz=16G=Xb16=F#B z8P?u>7hg&p!Ho$!C~N{XrRI?`5@kA_fUutXp-5gxLToX|nhKAp7?tr|fI9h{yq8o) z>AMWHgiuqFO|q%QnyE#yAk04BW2XlbQMAY$-NVuAW&4+0KIQjtb64%Ri}6jTEdRUR zpb_u+z%UFqEge!Fx4|C?4_=b-d0yILN_#*;T}J#tgFh3m@(z+CHw{@%;$|^`drt^w zN_98${X*mU!D972+J9O(55$j(E+=Vo8^e7nl+;^RHale#y6)9$)koL$Q~wX1kGek^ zOb79m9=uDJ`JbGQr+BzZ_tl;hABbHMx-P=$>aklEx-MG5PYvBJgFkvy4v-%*$#zhC zcB8j88eK%npO^h^H#{0$c7J~AzJ+eQv`UEJMN%>R=>daMmk19d`$OvfYsB*)hKH1&N{D2QcbMmtYVyA zUR*+k!j6$PDCM$x6MHH(FeRUFScrsHEQkiD`^y`v=@BFIwvcS_!Jn#>)Q9iR4%nN0 z<`OOGhZvijNIPGP20#Cd@Tr{9`)Sy$4pkO&Wi&e>PRvQ^iM)aSRgDpDuvLN_(KR4% zEpILfd#(6fq)wRJs@S`Yi7O3sg(XYqcPMva&ysew-9wDf%2y?ztr$Lw6kkDZQG_N40pUtTgBOo8e*lsUV@KR3CgH|J zwpQjFVUCTh_HsbCw`^PK)r7Rj%pAtUFp;#a{x)<5C!!E(P6<9g>f?q01fcqK)Y9|l zfJa6m*g=^O@FLhLF^26IlBGpy%?hj5-?(q&p2`5N?ije$A(X3y5M;+xaBcSpJ~13u zQfl_X(zIC7<%(!1K-q?d9)$PyHoTZdD<;}%l^vPWY6pI_$4|LUuef^#k!zWG@lUA5 z)!4#nh3^0?##*tbDboKB^<`AF}jX^upAmQZQpNGpigQzyDxIK9#& zlFp+(T>{r|0Ud60NWMdn){vtV!JTl`s_o72oG*@^5OVHm>N=V|Pax_3e&&~HW%0WP z>g|G`Irm;SNBTp?6Z$2?4gMv=0^v)y_*80{F82)nxFKS>sDIaw$MiBtHVn(s(&SQ8 zrDqi<5ZHD|8<`@}{4{QCG#jkrm?tz9>Lr`{x=tORP1lEfPEwN5HBDhF6ksYp1XZ0xVxMod)_dWV`!or4 z+a-@QRNp^ZL))wtzEXB_>DQW0QngGGZB4TqH3Jph-QhV)WX{aAye3LeUJ{itBbR`-nGr-X=kebyvfgOk_C- zk907?W8p+$%eIA$R!yL8i=}*#K{)4eWJ+UY1gb9It^_yE(V45Tx^q z38EjUjPX@Z zZsCpgx39!0g@CoBIaD?X&t;2s`WvaC?HOPPFso7DVaLn>jt38gf0Ii#e++v;;;zCE_*!)N3zD&H-ucd0 ztFW1OHQt)$ndL&ADZKE-ef3B4q~}@6g@|Xnw;?9ggG}xLHC^R`SUIVA#v)w4C1&O8 z=~@vydq&|b`JikwJz8%GTUbZ@Qi(~)A`_}A17<+fFDZ+Uzs^0RnpRXm1~1>3xO-X8 zTh+C{tfmv~=s4YS5$hI}TrGPAW8IJTfOn*gU@wlYt(f+kFLA8b>zaKlQlbOy=-T%p zMd~C{r6ovvTR`jQtiV?D+?t&jR%^GE*9FJ0RU^C1hu2Bkde%to$rXO*lLJW^mEQ0m zxFp8{tFqetOkwp*XV;y&@$b-w&HfQw9vg>qH54?Uf+ zj94jRS~Cvr>&Wty%ucNw)9N(hD#{e23TQ9(yk*1>-hs>(JgMs)LJy7@!^^oaseQ9w z3E<5fhD#FIU~>AEPi~k3o~|-Fuf&|{P@HG9hoTTIW+hbz>%?uu;EtmfXrVuM2E}SK zyV+@n!?{V8S`XZDa8hMZHBW8>sCQ~kj#&oK|D8SMBXXSoB^{_JT^|Du_qF|d|n; zWMDmjHaH%GJAB*?nJ&f3u_k36qbFe^qL>~}j)_UOR!x@SO19oA(*QDaY39Aj(w3)v ztJG{Ar-x)sV^MLn$`ms*bXy<(2=tw)xngT~Kf zpnJ0DJ9GK7*=-;BpJVl67OzkwQR$L9bDD}~zsrY=e%WpffoRp%_d^-Y=*lrC!`qVxoxZkES@?8?JA4c|bkh;yoFgk-t91E`vVm z^PZSkmw_=c(ZjXYm>Jrp86K4`h-#;#ae?(4S5qEt3XKzJB$5YV3 z=;{Ui(p2utD^T@}!e2NmU!yxlT9>0U;acMA13x$}MS)~k9Fv97(O}7KF)ZT`^J?y- z$(YZT8mFh19I>x>PI1+OV9R!jpLPh(&Eaoc!XG+$Pq@VuSM{bJzn*+kL)`~NxQ_{S zB?lh4!X~w=D(;d0ywl^3^lQCn4X$^3z;?*QO!d4HS$-PrzC+%8B)a>F0{F-J$QU={C4G#@6H6p}*tGzP(Cv$O1+-t_+mMCsSSVyrQRO}}2vj|bQ)Vq z4)4?kAaSM_25?mS8ZNdrYU=#g6EvDjp>V@@-MG>;*BQfYcIGB!^4`Sa=Gxp&K(zN- z@`XpK%-)*8t>9u+K@Cyusyl&pUUIt;=P1Cdi(s{6Fe|PganP$5YQM4-u4Mnt0?y~5 zb`A74$IkA^8$v$cGG(e$6|7El{uC&~49j8MMGz?$!YbOK7R|_YaVIje+6o%W)V(VQ zYlmi|NXn@pH-{j_)rb#FOX4Rs?FY(tHNww?lc)L&JH*nztKU(ZLWgl&2h%Y&ux(F;&rvfUOBPa+(1Kt(L8!qVp~pK>BnW zHO%Z4m!JbI(p8y9{t-f<4!rAvUahWcbkvaDi5%02RO;2ahjBpCIXZx-*O3bl7qR}^ zdQU|w;wY=X?OIybO+2BZ4b{!H#jM;(4&&zWoHg)6T5?APd&?G@6H|z~iP(D-C*O8_ z2;YpWzP2i&hP&72yEljS04hzt2V6=rSFr*@p;7896Nu2ll@LX-UW7`BZLiUNi*foO z1&W8^{3Oj&RU$zu^DQ4 zE6y}xMpK!&sXe9@B|-X!k_CtDdh(>r&Q&W6OgZP0{&A|HvKRo(`M9OrK&M!$rnlxr z4b)8hO_#lYE^L6U^`LzVvi}W&EadG;LmsKX-6y2|_0!EPv2j?a63%RqM2kBmsw#;6 z)IJn3YZCU8v{g-fKA#B^941#-a1!iymZ%V%QW!7Jym6=aoDgTIMa+?!Z0~KbQqXM< z!UsJ2kcA381Rp;kgwYrZQ;RT{5LEqsp+3-)zEtpoXoXC8i%?YY7>Dqj=Y$AUeefIA z=HDv=3Jcg2&^P2N`91<|9wX&+tAaGLG)YO`C3_rbfe1E;LsR_1T6B0C0}P3HLu5__ z{=I)`s0~3n2eHMFlW8!B__M zt%OF4V;c#ye~{p6q2gl}SMxN_rh1>zwMOIInUsm89}Fah*MDnYXs<_n zABHQ%ezLP8?XNO3k!BIi`mN~?#r|kW)i1wq1ZbTt!~bW}7qW>(QDhZ>`a4d*djZRFm07#$5dcW^ll%{IrDiR{5Xl zPPprJ*P`uQv4ch#tb_ESsvdV9_u0M5_W`I2nW36KbC{X6DWToipC6dM+cu<8{(aWBz_*)j!r&zTwucpfC#$#-S@}P>9s%L(d@7EveIUxXJ`|dM+KqtvZIsfk;%xf`bsR@i5Skpu}ic5vN$@& zOE(0bF51JfKnH2_k>e+DEzBkt@5n1=?`fiiw;wjiR<*E}&2FK`m&>H*!lSYfOhmP(MoIJ`;jDKN< zzPBIwCtgg~PqqmQ`m zS+px_uMpsq)v+AHq4UoF;_RHFBa63v-5uLz$F}XHW81dvjw)8gwr!_l+qTtF$GF+! z?sLc4`{C|;?n8}H^;Tn4t@WR4&fok#j>h_v$kC@rt5r-@51@+@t)*^~YZaU#U9o<+ zC}oAt#LdTF^{`kfQGNSg*5dHXcm4H{y?7M+`GF>(xFcQBTuO5KJ*-a?DqEmcs#fN zW;ZN`F62UWqP8odj%_Cr8~8dt6WK*AbuGyCVEB9N@;etV?R}cavnqVnAZQxdxbQ81`Jgua)*u&`HZEiwS;+^(kctv_3z5;=8J(^obiccfNL)_ z5J+DGA>7e=^#i`m`YtK~NHt}y1+uu@OOw0h%WMYMUTjN-&P&F~V@C>)lf)>Gozd3R4zev@hAP#7o>QuHW|}HjR?6i|SQ1hTx!>#{y^H5S-xMUwetc zVmeKWUbdIxFJQG0SEh|0T@N2kNwJa*5g$oZFK`OOaUM#| zz?UEd%1erX;*6@%uEAt4UgKpiUxVnWTmu0qHxU5!n{>eP;Td>m-i|E1{jM$e9}5rl ztL^z~emNS>f?ZIb(p{3u*=r{L*;@E@HlK)TR-c#_HlF|nHlL775LgOniU4qM^-H(8 zM2@+_2AC;HXnht;m@M+IzE{1@yNO(?$+LKG3(nL&5B&B%B)ou;h6orYtIYs;K*Y_} zq60%8+$elf5Kg*ws8_%ct`VvRv)E2UFUzGuE8R4os`0OW(s@nYw6hQf2Jb}r-1=)K zWgp$y4VuXg1L4V24jsfSE`#8z!U}DsIS#P?-SCB`gvIcP8?ua8wb%yIDOe28@7`$x zN+pgz<;i|>b7*M8sh2s@7Ex&AVxo=}Mp%=(zS1+h;L6S2>4c;uzRKFli9X^@0EF<| zWyUtIun787>)nR=?S{H`kbKoy&l_yub@^?H44d#+qYCOpm*rG>xi1R5$DJp{rQht%Hd$|3@NKNqt61ry%IJiidNnnk;jS&EED^^>XY)J25yUQ>YWDb zO%%G#l)(GzzscPV%xUG@;x5^PKWttWcP+Jf7XD z#LHm1`PFK;n7FsER%UW%wjCXqZ;fyu)1MX0iKK~^L(8264(>JNS>)spKrUU6(~ta7 zB*Yk<8&plQfko0d%mU;(OcM0nxMgQb{G!P%utUqgg7N7pxc&?j+?<9=OwJQD9rrdV}JHu8fvS-NS^)AXH;K@wm?-VTa+vs%!H*>JbFP=;dsnVEwm<_w+wH(TN zTMz1}O-9qgfnGA?JBEMDf!<~VLWigs(${FNSq6h`4X+M+D!xNe(;F>mCOWE80nRxm z3B(p1-Pe;UI2=; z?^$OlicV}8j3(C5#zGL+vT!IfL+gBlAVPx~XG2em&gHet9`O_S4?lOyx%W*%h zGdn7e16f!7u{oX#qv1(f+nTx^!CDT$y_uxAu1NO%5_i8j;twc&OlbConK~+)AQ-c` zG2!)Ti`8aOr_8;m?LC--=EZ>L*QZg;N|X6bS?T)rfAL|3{)SMQl!eq|1@pB%s$3F*jMd}<i9Be6tnyWw8+`Gs|MQ9V-%4|5QfAJ7o@i>e%DAd%epj;f z(#b`^->hs@VOGe$x1(9*2`UyY1QIO>3)QNo9LTNIUodCDH~anN;55eHmZ;hv|A`-h z&9na$jqRWKcSANe4Z?GEG~4-V)p^QyYP>7K|Kt6R;hQfKjuvA+QOLMr)SfC;7x`#v z#PT^~wNNj*BspnY`5rv{-&XR`uMcEtb2G6VY^)?Cqm2=j5!eu>15UL5iA0GqlPtl* zo1F$06B;oi$xF#09Om@vh2iEKnU>QT*<{P;rYrh&5pXYs`%v~9|Hzw5lO#AVM+Kf^ z*chTke4(FX7>X0~Jo+YB;7F}`Bl&%`Sc-pk3$gd% z%9&oJ;ub63(yc==Lc{dXh34%;Z6(&?bB;y@3O1jjMQLs~WP-z#xG=QJNf(}ki@MQjG3NcJkMxbv~=d9Oo}+C-ZGjYzd9Odw{MeE z9E1mqKxQ8*2ls-q=*Nj~IoR&i8xyr05CXjibDGx~FXs{#Dn!as`Sd^o+b3zak-7Yz zG@w;@#=!C$)OwlTff7Aa%c5mC$?;kX1&vMY=gFO!tyZJO?5c>%T$b|vyt@ZBEkEBg zFg_wO)R{^O(5FvY!#&YqxwnS$W5~)pAzQFeKG)qL9rI`QW9sQh~JR zBfH5lcO&N>-xJyvhwR!2Rs* z{Hz|PQBodD8n=aDAc!)KwXNcG++=cV~Oa;s>5L^Ms}~ zt?10vA@l~$OE0574C}s|$lkwgq7Ag}W#4Q!Dbnw|Tncp`9;@y)H}N1IVEuzd!r0qC zw14D4b-5t)T+>QIIFdJ=+SZ1WI`B%U#7liJp^%=Ng7U^H52WE)))SjIi)tmg5W3n1 zd;(@3!+>aJ8KNZfvAMKv!;7{HrmG66@M0@*a}6OUzM|olBD-f|g?yi5nZpXQ9_IZx5bs?eN;4>o-Z|+xsO;&)Wk?;IR$u*A{2lTnlu*itA9*s^S!PIKh)VXz?eic2$Mq2>&`MuDW>gaH<(Yq zPq|O=9be`6e?IX1S}&#IjI{*+V@MuIYxV^^(>eNLyhGvq|j}0ohF}r2!9l)|hGRVmYn$qFYo$4={e;K2~mA zQ-pY+W|M2pDM|}Zs6u;;XEVhOS44GUX0dhHYq3p9-qlnqRf__})H>jh^~41uSlyW# z4^cJBAhDp(2s@TjO{H=V7Tp?4^@gll>4vR4Z;s%AWS+`)HG6AKKmiaR#OF|Bsii^# zfW52YCX*@-u9M(ODCORg5gu%Lk{f(vUV?H#H^ri+Mso73G(u5Dwrz*rI zmuDq~!+c#xA@X93?OG`=Ku>N*>2^9WWGRf5uM*g?+n)7V4$(0G$`QQ$U8`@Pw>b+R zP~-<|eJa&3Y4yTu71D0(94t0fY>{-W;XekJ>2V0IptvnoJYyOK5d8JhU2)-(B}eoo zgg9x8g*)f+j^FEQo_h;6H9kDSfg*z~;{|;@_Bmzhw_yJCf86+M5^H#oG@>HkCo#>9 za_`HD_>7w?Fo}b;^+T3qw2-X}^dU3x8~qGUvbfr?*xaHnq)s1}Bzfqs&VwU!D59Ut z9JAs(>R>rUU#qaT^W3#@mLzyD*v1UK%X&r*wvh3ub3;`1IaVT@y{GDF2Ds*WI~VK$ zVDCM~_j#b!f`$ZBwdDC1RC|>n<2JDK7mzs$MjRHQaSPQu)X0$}LDpClPWJI^fr9r8 z8$0)9okFohPCw^zjvD*@5(>d0WKBN4ng9M}uJ>c@YcuK2mmB8!*LxVzgR=U`ir)zq zIVI&?ljO!laCfj^fXHiTW)PmE5|?iiX!nH?>ZBsw_El{>^ADPw;=RG<*G$n?)Gl0# z5N-TS9_CV$?Cg=wAqU*!m#lIpq{yZ9?oXaz(nH)&Uko z`|A!;{Qt2z$MMfH=f41TGu677RC-8lBdT?JJ)dvi1GE1-;i_SOdSQ1`V~5XT^+A3Mk+6u?~3c=RCQFwyqoo0f*T(KD;WtXuXf zWHHh*sgKaZng=js6msR%DsSi<#1Jia*U#xP-NLOUnnIgz96WdC7G?ij)#Ni?$5uW` z>$E)u9KsCm6?fZcSo#{$Pz=rSO0@e(d@5u*;PuYK-JB3jcS>t(`lcV4n~-FBm>VV4 zIP;BfqQB?Gdy;-3tTVv+DD#%~ibt!W^kn?!+(ulCSje2J4Yf@$8EF9zsG>%tQtiA= z?BmWKo{PSjoqwz*vJ;AM4GF(zbt-Uyrv}-Q( zCzoN743E7=n&xixiX}KAo2Wv8flfb(3}x!VHI_W9Odeo&PcfcV##r2o3@bf$7e;rq zH(^AWw1qas{z_Z2RukMAr`k|Y1i#wiAF__=vi4tjc`~b5^d6%)&HQZoK!b||rc$n$ zLvU9kxiB_BjDj0wKfzq=qt7vA)|qDwf;YFjWg~|IJeMq(gR?=?2>bq zW=gp0;IG2M-!LuUB+0x(B1x^qKy~^|)6-0lH6@QzN^=Pvx~w4sYD%BN7$uWzyb*gb zNkiQ7Y^JY>-txUM&k%2;bH7J(is&&*@eO|WhEMA1<$AzCen5!6;7ecmlJ&x_vdw;q z@Ogrp;0ZudvCSd$s{+x%h*@9)*yrY1`rW;pp~F25}Nna6yuzq zeT45Il#gwOdLcy)$xZVE=V_TM`jjN7(!@#!g{%u;|S5OPT{e>#B#QT5QbaHjE zv|*4`mgi(=WK#Wl|4*%WmIl0!%3{)I4hP|%$pcenY%E4nBINH0f>2N~Nc_RUKZH;c z!QvFC`zOXtOqfwwQ294>t?4$GE}Px7>5S+Cl^W>mX#9dHshc{M&YCNhRIL}?Hf>#7 zI=)_;RS_nikCQSeBek2G&+(66CWD+FX;mq2KL=nNPLf(@nI2e zf`=VsIM3YyXofc|_Ulqi_Z=osimsLDVfcG258>eiy(b5$VRy7;Wr*&IgHJ8w&!L`q zh;H*8Y|w3v>&k#*wCQc?8$!-s`d8O`p-()6FP)f>H-SJOMLxg6W?#wK4?W4wN_Ian zy3TS^zsh|!P=T9#jGBrhjzqxz*-ddk4$e~=@kdZV3*1}5+Ih~x{xO7~;f`(7eN2o$ z`!xfE-{~Mqr`?V|{9DG_jk;BXe#Nl8OXvCE_s){1&&?ix%Ez~*$n9ML__vgpZCsD- z9!R~$J(0hhmtqObMY}W=cmPgUN9c_V?u;5+T{@`B-}Ie48*xw+#_;%Q zA0EujDheB!QpGI0CT7l@Oa{;sQ%lXs)$o^e%StkgTr@0k;S@OB`(0}4)Ru~4U9?}W zF*PVWwx$g5sUwxddb0C<$qL_G_mahJ?5(ugh2*lIoON6qgz9&AvuUT{CE&125=ve0 zbt+%}Q5>UFmHIF2zQ+5gL%yLoFZSfs>RO$lqs2#{EMCcC_h#A0mV-EBgIT6wML)K+ zrET6kSAh=7)3hnZM0##+MJ|N9J#W0ikyMJosA!(>2UY3FVgk;Ix|bHXmpw06Tpi~; zGY1sh%ZQYC|GW8X2bYAdmGgZF$;||R0L6J=6gqmM>`JNSkL)llse!DsD>-c!RYwVO zywydFtA@>*NNvolqx+E8d6&uC^zb1wDr{@};$7-NS5<2K=>g@@x89-GSrJr@1Gn2% zZtWo`ad&^td{WN~sLxadFH*yhgfS0$z(;$)ht=VL5I^-u!YxeZEA1h=x;=H;DUmv> z3?gR|MH!W?n9?X^mW7#dVh*==->IBkwmLZ_{w6hcv_xMT+F;1N=Kj!n5}@imgRH-K z@Z_Na_wS?CcOocWAx>~4?74^D>Ksr3|(619nX=mcTr`)_v>t4+Xm@(R7mL+2LdF!>*mQ$zdeKPGE~C8G#M;nfXzv}6n8qwS=32&B zq!d^zlR~XrUbHZ<$!W7!(JF20l}E*50;Iwml~FbuwLu+r_)Uc=9c?f~DCY+#;8U%# zGiqv&SiE`DRcW`DElc1U+iGjH{qQlH$$WlGQ)0MFR6d%vS5X;8x6LYMg~G)6iGEkA z_)VEUJ|r;m12ZMfy^SvWXfgVJrXZx!vXUG|eR!{<>R`jl%!0aZT}s_tiB!Fr+!KFq zR@i6}b3Qe4Q4u4}sVrEgO$GzD{NtNLYQ5-lyP`)!(NLmky-o^xv5p;KwVFfkb2a;B z8$i%dEXnpEsZwpQ+6yLGAbAPLT#rB8f;bOinUEM4u zsEcaLNu%<3f}+#9w=_^TaH}q!In*otBp%HMY9UAQ9@T`9H-qoBsxN$}0a>1{@Jh9D zx_?7VK1+^<+Wmf1qFh(XHjK8YELP=jNY;C*Is5%Rc59|TB-%pn+CGyGCPCBV(`!v5 zIN3Q;V+P7+hTjKDWNI-!CI)rF^?1=>w*Z^1J1*er`)xba9A`?|X|fQ8En*4eTkapq z)Z(6JlcFu1i7T&t<}?*s!l~<~h)nl42-F!`ps+tc5u9xM!CM$ zwe{vSQl2-aG(op3AsrU8e&Mxf3QU!(lPB+K@lVo%y8W1F2`%CUqUu3~b{&;D^FrBT z#XwFeDhS90Kq_i*fp2nwkCzEHpeJIg)vl;9l8q-S=ZZ7kJLr~0n5$Yc+!f4#jyxnd z*k!%`(4p9|tEs^aBwz&WKyw96-3C#@cyM6{&)1149wFqN$f3w)Sb#-5AN5H-HA7@k zD9oKuJ1>4cKH@+*m|`W2>3qBp$!rPNN17bR2s-oy<~@}}0ePHF{NwZG@ryW)PK)#P z8RPXTPfRtrWf_jwI^WXet7ONkCl7}XMM|OCg06yAuv{RiQ9n$>4P_8d+az<+u}rIT zU+J+vUW8iToa@U2UqllE#kVqr(Be0e$Pn|SLWT`PkHuAP+S2Q64bv+J!Qbku`?*>7 zr5X#Oi7d@)*{`iXsBiV6r;SPM%o(YQh2Zpg9 zt&6aDOmO;`F?mdxi0e@lH4&xRbG4fpPpIkmArPPGhh0~|vo=zG^v34_f8mJFm3maf zD=#QtC1SC7>9`ugd`@@fu-mQpu-o7zKZo{ulaB2hUOs#jd}M(4EVa$9vd!p>uswv`>{j#Pvvb{S7V3+)*+ zNaHljz#qO=_1$PMQ+dw9?DCNRqq_eZxeQ14N+8T#GVe6;ZkcdMIl|Vipe8D1-WV@H z@d$T7C4`JZp(gAO1m5QwhwX*hwiurVuA6WFo0K z%80?1PK}hp{xbe%O4j^V&g}mKtBrgOXvh4e!Pp%QCf5nItLyXWvMEQY9(_6}J%w6$rv=yrerTd3vz0TcZ1MGqX# zs*vMM0{&AV*skiT&uy8HRgFDvu%$rws;%qvH8)~DuX(gT1iYcn(JCm;yxfYNol7#; zZwZh5?-hu9`JIP!2m80_L^^pf?d-?w!d~Ymr%Uk~HEJ~~F%D*$3(?#Ge_?K@w#7Sn z#E@pLp)bK{du7+|9<(R*g%bGG?b-Z6n;vw!Y#kCwVQH_2b-Iwb`A4O_S*Wk6guglv zz#KBXjlglJaM{+R>u0gPV|^s{-(QjStT>NfJFbur4it7f?gz4*!Sef1-BI4xY#WjN zZ&G{!A>P2>1B?>ez3(1?xEF}-5cg}_&e;9Eo}j%U5wNDvBDmNQxXd5F_^H)*m^4F& zp?}H$q>vkB^N?!ZeMRYQn2RL*N$5xEecibchIydup=j3|d+<#UDB>a39Z5Lu{1pC% zRx`ft5jpeVJQ>i!#;`R@zOSx@?pL+A$DswgwaC8LvPS+Uw;sK1oip&dW_}&uM#H~= zu-E7YMrgS`fPG06Wf6;dnvEgrEr+)TySJo=rt2*!bBWH>s6JTGHFSf!x44z~72Q^v zFzY{dkPbHnKHHCYf_# z!0s}8#8;2*FJ2LG+5rE+<+Bz!kvF|Ny*s1VB$0D`BI|t*zSgJIqqCat5L2L?Ch1bz zZ^Z5!ZeM%vn^u7knvu0`i^qk(-u8I;1Y@gqc*(AkEtC`kTCOB(3a$;SJdDqcu-3SN zSeovILvskX-!ytt|o_OfJ5ol|&fEW*21CW$tW$?p#%dvI8YI~A`L&R|AG7^CYUx=1TGV)123 zV6}{z*X5`5S{Y2L^5U_@2p8wA#Y`IZognHPuP|SN7eVnTh{_0ADNrnxD#a)-xK@L{ zpRFEj-IS`M!n9Q{nZ<(U8a;4e*#sS;X!T)BtN2D3WC@vPv_^uGBVSvb&nv1n@$U&! z;5>bMO<=gkLN~Rl%tp<+x|pXY^G4ASHm!NF&$nYst()(c6ObL@SHELxli?qT&J|dZEkh%HjRdQOvA0$0ksAv8 z`Dcl(=*aO1o2v5t0xf>~#;o!xKHxRnz`x?r`$_vM>k_Vep{_|#n)%J0DXo2dxs0a2 z-#nA5*KQBdR~ITPjT_BsO7Euy`po)%A5iL`KmJ)ZaX6qGWW*??I*aT4t9MGgX4CHy zV_vVQTD+uKmNQb|AmvLsSsSf(DNQ%afBv?n!%YMFK30PB#}S+EKmarTSi2y{6LU>s zKSlXh)mDCJE`E!5ejQ@fsn1uQB$nXG{$Gcfo9m|Z%wMq_*{@iR;Qu6F=u9eR=W6@k zrkek|)mNmht&aLNEI>eY8Zr6x3k(VqHiA9aLaQPStnR0fBK?vmRZ%m3AC`C@oGf|7 zgI-nF>bddjk;r9&v<@A;b;{o|-HOlr3jcihqE{IsC`nNpeX|EHbN%l3OWy}l|F^q4 z-*4X6(NSumocYXSNRD~nB?J47Q9-E(9+Kjq#RGe$F_zdhJ5jjIdgx#N1`mAdVK@mH z0AvF*Yo;qv3>VSqo}=8!#Fsk*!z8LNZnTYxAQDe922CU>kKRJjd4^Rf*-8U$vEX>& z&Wh6ZG<+tt9m~u??7lTm>$GsF*7@C1-z-fb+Jn-F%$0;7=Ia^OLjWGmrz;z^)nsO&$wq7W4qrh#XYR!UEe$aSdMnU| zGh!lh1#vhL>9uF;!Ws2oS&j=h)I99XRQSWyVs4nsRgI8>@CHRdIT5nit$u zB^x+T4f))_COBYfa)2b5c;JJ=)WPT|6QhCBU>emdi49Gt-*U^sP}jv=)^Gn3gCE_+ARJ){Fr~=GCNWuLK$ERhE6Yp2hf*ek39QbAwWs!gwPyii z7l7v6QoN?$Qg+Zub>H)iK}6FTLI+ggZu`)Fb!*zYWv$L?;E0CregQp}L zj=PlIbPR6IPtJ^p65w9OUQ)GV_|1xDx0*p)APN|ujhz3rvTtlF5K~;i40>U|7H~SgrAjK+bjqxGiKZi%QwX&J%_1mxxPXH`@;BhUC#{ z?w^VVk_RHOpdpi}e0!Nyq^?qW{==a8JSnr0|NZ8K=Sz~3qE7REO(Rm3D*S`}LY1eg z<*=%yAoBp>;)f?i&qp9wh@2nv#&_5ak<|tKp8~ygJAvI^S+h=|C4*P zY&up7%~2O+#d*zY>Ks)MI0zJR%GcVn0B)?%P(ev2anh^q2CGC`P0PNw6p3#nr0-DpOZ3(SQ{!}gZ$%ygiN5)k$@)0;BTUz&9TUTiMD$_{ z5Ltdhdm~*rhj`=ri_X3oMRPQ*K7AJy}!_Y5P zxSp(*{@I?{5hCaN$h+UO-BN4}LpZJr6} zfoCgpLj00j9imlZEAFW+8H|WWj5Z?5#HV2#kWi7U^+_NT`XO)YoP1_F?#5)8OQG#9n=Y?Mf$1MmzAlp~e4hY{@~m^{4#W%E4PksoTwgE+H4+hgO9kC|?xr3{4w z1&qZ;RT)g)vO(E`R1dsQmUi$U*jsq5DRH`tHnCp;tAAk|l`)E(LIv*bGhSZtogLE@ zdxCsZl)7`5(=CBB;xf3y8@IDWL{_CNuKAB-_C1M_*SZ5iv)i+^9k`CyQsfkU-O8l1 zyk(r*NvRz5cuT+EL4>QmD{EGFYezY)7rlHTZg3q#7kvCh@Fs|OJclApZVcI(JSHI_ z*iw4;j<>llAsmglzY-j(KSH#j-Ou^+4cWevI772A)cy+rNEzol(OckeMJSOrEgo0j zmyDIt9})ryyCA;vaCA6WS@b}&coR~IvoB;zCoE{dZ0URnRHhrF`9`zIc zDDCLz;q*vi{U60Q7~@wc5<-L8SQAFp)l|1~aK|RJwMMoq+<8K;oOFu@a?1M@|5^>H_9b(BepN$lUo(`S z|MB1Ve?|QKi}A8c-BJNh4DG{96U)sV2_!gJ6!mA=?nHVj7DyOKKrj?;2z=^*wVS&J z1OPiTi;R@uChXBydclG>a5!gROwl6gCM9{6I<|hSf!5x?R5e+`(ulP?n@wK9n^uzBI{hRc4AnM)hRS&PuNXB0|n;vmU`o6);_JI&-N9+>PttXFt zccO2af-~QdvwXxIQ5MN^wSS!ru--$>fSt&h6r|MXUj+vs*~`%5x_x0KbSwg5Io@@= zShj6o&D#PIYYrAvL8>uolX?%Ao8l~k-;3=S6WO2t61XcY2bEtS09+=kEQ)O`fM@clTYqtWdq$C- zmiyAsGWLMDY(g}*DolvU-WATME;@EO5W}S<9gq)B^SSvWCm(!+&M8#NjLNs(-MK~w zHO|YjNoeoEmL5yZ20LA;$Ra{>QbE^)7b@DRv^2U`uLRQMsv>_h+aXq-y%Q*3l=gLf zFqGK>^R69T=Iq?44=RwFVhj(%InnWf0$e0X$kr-``;sdH7@PRRL8_uBpOIzgup-Hn zb;+?!36;Pif=8Tu?XXZr8mU?`BVJF7Zx%0->5jJ;>r~gNhbxy<~d;`(YOJu_5Itj ziVe(?J!~M=MW}eo*eDM8BD;yT&FnH;CY0o-v^ffeSuBHcTKT%XJ#zZO??31Ik4Why zI7l{qfjHz8)#jAx5njxx=QK69YEeyvUK0?|@iF)q<`o3&Z%#pd@`fLxl67G;NQwYk zVwhS%QKAU2jURjB8>BmC$X}am$SsB$zt+Vz0(itW20g#f0>oE?z93haH{S@*u8~CE zB_-vM5sd{0+^UQ5Vi_m0NTsO|3ecTdwK=Sd7YL8{&7tiTIqowkf)lw6!{=Yot}E4sOfDRa)32dhH$tB zijnVts98xCBWfjOtolkSF?Wt?*=s)MRx|cFGoWic1^TVeM9zJsBK&#Mp&cNwFF*Wt zqr!c>)8;^SNCekwiH+@snlK3Fv=&*RJ;oR4>}1_N->=V;w9P63Hd2PmGQ z3gImr7d8`-sZ;He7#lu!Ukj@UI|yW!%+ceNE5erd9rTyS@j3%DkunlfN}^*{r5A_b z5#E%t$E}e{P#v&9#{W$`CDXXQZh4a7(bSw=9D9s2$zwu?6y%s69FZuzw@jI>D2*wT zse0p;VYIl8)KWva3$IX8fg!aNoy51r`#Wn-GKWQ&(+#WAdkv;*zFjeYxwrMou2v7! zxZaRj&VY&U=6Jqi8z7^XA~b3gjWL%h<>I3gL3%b`8)m_ws9ko60LjX?!^OKfysP={7q8m1-Z7{X4A+v}>9X>FKUk=ROX zMpx42RQbb>IU;R=V!l(ep6~%Pz7rN9UuorPud5bb$iSl5k=i%rJ8i+#Xie|_ivu!>em5r1U(;iwvTHNf5?@ohr zq`^R;)*SX<7bpbrT(3+md?=Bfy~cmwiETx}Zr2~0OYZ-9 zcjW$!!~uPyk9JrG3=tiSkx-0+fXX8(gpz;ITuCdj9~d|tf1siE52Zr1N8b(jaC)Fw z;enY}FK0Hji=Qq6E}lNM#c-wQgoDrE4O>vVg=SwGhwCyrJ*a+ahI~!6Db@>6vm^J= zmd753?$26s@)k`*Tc+N9a6_&+-}-8X5tKkfR^Szr3u((=%=V*jU3u?9Z=+uQ6r6d0 zsRzqH>=s!>toFxXBF=NQ15#zRRGE$4mUB|pb5mzW?N?a09rakZ&2afeStQ09qWyES zw)3@4T5FVH)LtfR2T**&j5s{*bAK%=a_Qnb#kmXd8I;{~GWg?ovq%uJw8n7BrWt*6 zEJ9~*FR9gYPGV>X{+Nl)ve-nTfdB?)yumBrIFz!pH#L_c(an=k@ly720f3YkBFm^n zS0fQ>E@d|U5e3LgpgVNRm|^{2VGe0Oy{SD1pYa>~{wFpN%QH1S?ad7g&9u+(iiR3~ zO8hVx^GT^2QDyP1xh|jx9^`q_kxMivi>Q#~Dj5Y01|0TbbGi|B;tTbb7>DTtYaLaA z0_&h98S2>5l1U!pY5QjP%OmwxuWk&X%E`*XNco6RsRzIg-P6Vqi@Lc^#FM{2tY1a( z7a5ZcVCu7iY)q#rgV~^rZ+zZezL8W)k=e7_2ree#@g^YRi)%8kQ(^uBP~8y1?0>h$ zH*Qs?z=q@SeT382)~UL!Jf4flJ>sDYPdvs_%g*CVuGGChFC@@uV=XI+6jq>rF}m?O zOl@`#zSz4OcY$&^v}7K2_(Sc09C_QJ#4KA*Ohv%6fDxJV7?#Y?E>$KGSua?Kh#9>5 zrn>0j7DE5nd2RCYR^buO8%8&J;;PnP)Pna~)EV>EL!iwa(iQjTDI}Qwk5~D;oY$4$ zpA9-qAJV3=;(RVRa&6!c!ArVW80nfUULaWv(nuo&BJvA7e9im(qeWs384}4P(LJy* zSw0u>&+G;XZV)T!8qFN^*mCpdx80mxf8oSJsCD_b2nb+4q(XA+Oe!|#FRw7%*ZbK_ zY);W}!r)b5P9|35z(z03%pt_l7s{&W*(ugEZiw<_>2&ixNz#(>xnNF$ zUe#~zaA?;Ny!We`P7nP90tAErWixYqgb?VWF+qq-@=DL8@sB5iWZEBgue^W zr|IP&&v*nsOb~JV^{%5tsRj&=t~A>37*9OiQM6uy&a3Vd4*CCU4elZaajyH7H*EQz z=M7o@slq8zJy-gN3MV}{!#=^Z1`Xm#xzskCSQ&lQZZ~my$?fkI*g&`2&_Xl?%c+{GSk#=gO4W%)3F<_3O zB2^*nhcaRgu23rO!y0Ttst#qOTV^4ekn$AyvgCY4{o$$@TA6g1oR?@64NiUv>@#?5 z?3gj}TVFeI_0ACU@xBUixwA_dwf9KJ)MCdJ-7R@_T4;)i3`ClaN13}%pyA7$6<-x% znGy5b{LV*S8o?1USfYlaCaKCWYOBvibxS^mn{z3y;^kOstr~ACf#a0_%N*2fk>2OV z0;M(d_H^~JT2>s<{r;%tlcfh(WXA8P{J6p6VnLq6D!L8MHZBv%SvQuo&wO*OY1=j8qL z2~W{cK<7}URqDd-vT!OfQchPNy;!KePTF9RivRh}lGc=t!`!awHd4nNQMpCw`%5MvDYxY)kB$5uE^a>2CRQMbA(#}1YGy|^8G}# zWg1Igna%-q-}1BwJG6sn%Zy{s;d%~&QPi&TU3xV(6Nlu0XcT+n&g88o6Uk={5gjdK z|HWpw%fC@9aBKz>6%h~G?lW-j@O@bMuQib??1ST%GJC&sg2mMs;PLBF;_Zib2wx+{ zXkg4;&(~ZoEmNzG8Aqha4Cl15BA?brw+*-I#$SJ~jQi*=h{!&5}F zS25k%NiJ-dQ4pv&C3FkasNt*JH&$h37|ti`-O5!RD{g^GP}m~s*-DV;5ra<9fKQa;#}am^{-fSHPGRv+qU9%;4>wo#mG7J7~$KriS$ryfFb zZef-)a1rXP!Yz#7QPxNDd4jewFO9&^v_UsacWquup3t^xDpC^DKK`Dr?s)yqE(A$h zHHHI`_?k-ECbE#&f)9SI^XD$r0lM{9q)Iz!z*SC%?!U9H!H7(K9=}Qozb`()|8NP4 ziSfTm3T@?eHKdOiIufKkf*|N!pdgAdTwo#)c7d*FCKbm5Gytd^lrZzvE1obYy3g0_>$SW$>C*~3|Rh`FMs_P`jt zqIIxB-s35;CFkb4EfB${w?Bp11=|`#6}8y*g(|?I(_d3C-N^|cC^FO-{oFT{zDs{< zjUk!NBB$1p{Drdi)uenrFQ&*Xwq189q20WLTeC{5Y{)zOQj}wBZ8_UaHv0BAxk_uZ zTxK>mc38?B-OB{}D6EB*jIS7|T7Mbr5~{F+S$<}N$c^11d~aoYvD64=;% z=j%?B8Nr2%>1Mp)R0evP`Sf$^K=c5lE=s1@Zd>}8VJC1u4W16rwDjQx#A@Hbj`i<2 z6niUejCsW%&M?@yqGE9;%%4jM-OiErfd78uAU^R9GpV30{ulf z$CscdlI_r;CJs>#Lxv6XjQ@>p|aTOJ?7`EQPnGOD=bOXnUIb@UMYZjD-}zCBT@Lr@;X z-%vF)6bXTs$bAhHEqV0FA`CTmTz>eQg{gJnr#z^BF*v9jke6_C^qLJr2yWELAf#%; zbYgokvSLdAr~=n8FuY8?)yHv>W^(K(G^nbBip8rKihRO(01tkqPnY-EZ^NzSl}D@) zzRy#H;B2y_L|lpOIiF3-2JWf*codBwCbNqGW?nC8Oy`-yc02r%dag%sA0H>_Ulk74Zx=s zH89`X)IX8}6N{tX`Kb~8D<0_`_GWeeq#*R8E$jQI*hxYD_r-5jtm)v=sNP5dbHt+M zg6Obf@KbYIbW}TdFBTY?c7A47Yq<7Ufc8JpO@iSc^oJ zygnbpZ>W?NLEfeBNeGJv#tYDe*|~mA&WP~Kh1VLSIHx|ps@8 zM91I5Jf7Yj-A<8-EQA;gONstOv4ueNDMec438TAOAj5}+i{D3J=pf0qQl>M}=3}8( zVPV>)VlYiax>@~YO3f$P@c>;6%frA#mPj4KxhD~-ICtR|WJX9=jrmps#3IF895DVvoJOS)P1$u9?GW7FhqA%(v z1y=NfSORgf581h?Sz6E~Krzwb!l}~xZzT}GEU`vEb)cgB!o`ML0_YZ^e%NaoQp zvTj(IU4#^(>zLhrxbq2S*Y|V9dRk>>IIxlP84QQGm37!^=sk916Tv{OsUl|B7tIF- zzMLo6Dh5XmIVHQovDQ#a1V?hF>j#NY6E)RU^`fi_j}9C2#JG@EW1}C>Xn(~r`Dvw- zmThOBoliu0Q@nB~rJa{$LzR!E8Wv<&UH_r{TTldUl_!!CiK8V)re z8aROvq27T0{<3(<&qU3>@I>GL2cBU0=Sw>Ihxf||o6H`I^NBVz*bjIDX>8yFD`Y`< zGtOBD5+wrASu%A$&7~-u;J*CpZ#}v(W-;G)f!w~np!2#vM&F;WqYpLDaw!8k3+t+2B>kyTl zVd7ewmgn!BJEdv{vn6T8WgBpTu>ksT8>Ft$NMRlJaL+Fjj>(KfPl{zfZcsP=({Sy) zUfpKeER4ME7Bx;w?Ud1*t!tD7uAq->ay>KWrY|qRPXheRv9O{yiRA4OWMnW^?o!L|CJS>r#=`Kq}d|v#afhBeJU!ynerf_V@LBkDCC}(%(C9A4Xd7E zpocuLMp9kuO3SIhHyabP_@vuND`ZN%2OyTlxGT$hXTxFX*NbNeTFzb?w3|jr*C`Si z)2YIW(y3}D!5sX_k2q1Gqrv4sO0^ZYv-N<xS;-s7tMyDeVAq|8ddw^2S>1?pLN!Cc#K5F=*<6SI&6 z;&tq7wP0xJ>w*MmTYgBdzqYw&AsECBLppMKh6w))TLd!-e_;##FKj{2TR|Mp+E6iI zc-5I|kA0rO&6#jFd)C-V%vULJnW0LG^ZvSGvy$Wv~_5nvoGKeph} z{qpK?-BI?+{O8-7%Qrpn$K5YX#OSo65Mcrmkt_mW0@RAr^wSX2 zG~B}m?!#<0fIj}M9#y*d_`67Tc8sm^eqT$dncZp1fKTI`2E46S@(9)Y@q%3ai`Tb^g?4^qpdKD(Mt-MWFgk-9NRIR zOE)Vo=Ev?Qk=2~4qq!dfJ=^wqB1`3JO4|8j^Dc~)=A5L33RwoFG0Li?H>VzS#^L5cggY@%s|Jk0A4>Hz<%n<=x`0oh}iKK%=jzDH~fLMz)maSU^V1UV~%2!_=G z4*fdPHvd*qp5Aq&tx+7yTc|@fe}1AUj6@&si{pL-(Gjr|h%9p5@6coW(0DoGI(^6( z`>6BV@epoN===0H(AA?`+q9dp!FFjn{Ie|q!wHyX`7>gMaGmfXVI!X+^FN^uxYtHz zq8<13GsrXZ5Q~x9Mq(pAfd3xm(q{3Or9j`lU3`T)%fIdX{%)A?Pqss%yrv}zBL?qE z>P5w#h2>e{=9)r{=8t@DWl(h$40U=lO7ZNv`W5rvER?D0H@448a?`Fq6Wo~AJUkJw zMl#~#@4UC5?%s!ws=v9^B;`eNLtPe`fpsKrclE3z8%n<`5v(3xKn-}aF9P2Q8U86h znGt{0W~C~nfzz(~&3hotksW65tw-f+V^aC>uArQFc(cRcDpqG#MbSt@Wm;I<_@I*L zV05n5mD6z6=u^I+_WDOGdSxUYh6MZdo#xCqf;CNsY!#;A;gzR#v|zUJ*S0dZtDOVUo2Y&%~aCqGWpqGRE=Ln9h$xZM`!d?^?fNjQbtz4H#&xtwRU} zK|ly7vN$R#aIEdpSrJX^Wn4S$Zg(F&>CC{>3cg52Cz~28rS)8TYI?f+VQOaUtMkVN z#4mN4wVIn{!68vJPHd?CNQz}QKXdF(yU3K<>-jz!ArE!&FQcX}uW&jc3+Y5op`Y^N zA5|)}H+I9+VEdRf($&2u8U${S${iND$t#U)h5 zCxlMGY$QhN+V2iRs*^g2Z$;w6B|(uLs>iSS?sN?bPUIIhNM4Btuk9!`dG4cln~OxY zUw*taGVCj~XQ?Rfw3pX$olmle_PEP#^X4eQ7e<2_elWnb;mfXJ$>SI!bXvKrGlZQ5GqnU}k2Disu4 z63%SA!512zk7?i68JN^pTI5b`LDoC@SH3b^|1#5ED#R}lANj3W1+TQym*LOH0a%Hm zR(HGH20N7A_45}Q*K|AC#!JEh?6vtLuIp1c)vp2atz~@mImpNp+H0L_y#cM0k_-{L zTqZ$$qq0u*V^1Pm#jbSq__hXtXAcCCq0Cb9f)MGxj4y4=4BdEbX5e})KawQVFl{dZ z=hT=`TEwYQnMcRuCoI4WzX)f?Gq^-j_f^H?L=d-PpAxEx-Gk;&wlMD5G1#%~r2&Q; z1o-{5b&3VHb0Yp$yJ!;R5^K+bxV+8utPnP7gW%d%~;%BXR)bI^F zBjT9^Pgu5eEKUUTF>Dlf|G&_TALf!5)xYYv_ScwDO)|9}orNCevRK1oP!bXokQp zH8xdE0|h2VM(fI`=(J%VVH9S(H0rjS7+3Qak#MvDwqs--TOS^ELhsqY(bPVu=Mf@0 zly{lpJW9PWlSF zL&<@0qwgnFl^44~S5rxa%{Li(>J4ve=~VP1m`%)88TzpbzX5j*Ux6XE0{Ue3+}R+p z6JptvXuAL!dY;iWYJ^pKHL9hhgM+%GgHw~J*m$;aT%orBS<99|`R67b@=?acHA(9u zNA>0(6^s}6VMLvp+MM5J1H@}Lfk!EY;}sX<#5$%s6G}oZF9h;syPZ;9UP&v7rX(A8Hx>J^}=u#H@8+u%f<_+jW>iPT|`&r|2@ zj^{2)S+y!=)7Pf6>5|w`6WR!m!2^+z$=M@_H%`WM}!sWWgyk$44w)j%2 zL)O`|-Or$@($j_L}}Rudo9#o#AjO!ojs{G15K zDq>Tvcisf2&+3p$!K+WO+BLaE)&WvUSapk$GRdF8Pqx#X3md*8eelKWzPO!ByRo&71v<>%v%9hM@S{G<*Y~e0?Zi9I({Dy-Pa9tv83cjH(a& zFhy2_P-yNp+y7bzAJq`slbfRA`q}4eiR2XO_W6yzOk#6@9rc%|JKQzYuC3Mng57)|t|H2`1KzcyrqR>u25dnX1+KHWRXlrhXc zW|xw>XSOaK{PLZpp5@(C4$KK{Sv?vMeN~(ug*Osyi=OyCtf@eObfc*e?j&?WgY5KB zfSan5f)P)FNR~R2JKrkKs;aMg%CU|}pAh2UPVx}-YQTHw&9$*BZ!^2Mz^t>jF{X65 zt7NR-;D2#RD9?gDU)%glh4ik4mD$>|mp92Ezt6Z{9i13=P`cfDP5C@!Ew}mxR-OqR zIQY|WIX7jC$*tWaMpUU2ckOp}T9@UW_mkO^9pi~1bXK&v8y#@a#3z$<(CL&i(>Qtz zi@xRSD*~p4sV{AzaNmLlN+=CB6y=h2F#Ja+Q?V#}hoEbBMA=1ZN11iIT56?VCX|n;5MyV4Vy5f8crS#4eU+_VMhRF9eOCLn`|t8$OW-+mL6Vq2r^{Xfnkyl zO4fpLplPCHplRb6O3Qbxdke_&=}}`X!gH#sNd0iN7er#u&Jxqq1OY22+}bcs(_8Al zN7_#tk2#vJDEt03s``I0Dt5Os{)=V$r6|buzp;Rlrs6-A-)x-&L*Hl#bP$9~p$idg z9MA=9K+sYV!T&H?FCmG?rxlcRkEmuKrp#alljFZhf33b76w5D^R%Kk+O{bo_uFg5r z@w>acKp3JTA(bM+^#uiB$Of;)hY}*p!3f7GAPrS8j!(fj=(hMCYXxuTqzvle)} z9&>w=>pgn+;DltU438Q!tdFFs4z$*0%bJcb`!;=dhQy3{^yS!T!?~%8F5A|Pr!53w z!~JMB=4XQvE5RN1_7s7wdTRjD4d(zB%r&`#Z9XBp!AXaprUd<)qd~85=qbB zlgEc$ml(U9R)DYyjG;6|MEralKe(lOKz$=-3!e2sBdta^0X`v(a|!-g+Uw%{Lj;d2 zFRv~E?Aecx>HU@Lftca2W+Z_R&@wT=s43b zWV$s-P#2oqv@1)-z6_}S_+88b(-*R|J4|?4o&V&73~qeBGNUP;Zyz(7$doQyz&U3c ze}&Z{=@J!pmUX5i80|Dvhs42niH-}i%O-?5N|}8iBqS$Fv!1Ew0!zgv-4nt^WsoVW zEF@1zgnCuU$v{Me9SKnR6Ft_WK1peiYEgK)EsP^bc;C9VP4=;J)WrDMCWf)!PSNci z=w<<{si{@*gtcuBlPXaBP}w?_&8(!e4eqC*g0Gh!?`w#r(!&sbZqF}@qKv&P4)Ra$ z@XMi|#cjTqW9%hCd@)Ka)FaqwsbJY81m4UR@E|&k%I|`|4x0w1Db17FFi5VT#hlbxY*GyqOzf$4 zKuOCp=6Tr{hM|&c%k4{&$2j6qD}_d30;cR5ePHM7ihkCZNxgjZW89cVCWLlk3{A_F zyVMQJy_6Q7u>bw}8Ph!5aDOHGj;|-k{%;e#tbxUszw7@1Wh6|0VY0Z3f4D*fZ2|oT z-q5IMTzm<99 zff7wG8J+Z;tmo8?XWsJh;smM*^rX}p5=4q`i6glx>$AkFqHf70$=@%E6Cstv0>DxO zEJfcVvAHs#uP~#|v8*ZAvKUsqPq5z^5IQW+;EjvrElD+W0OqE`;P5lWl`P$Kv zZlMdz?7ep#GZzd;itPXnf2tUdyk#+J_&<+f?GCSFmjsQitP4Eb-b7=N!7R#*`Xp8_ z+|pyBeUggA3yc+V21^!Ip5-Gg^i?TaO+)?>tu(rdF}zZSR*6J{Ua(mCy>^dx-s!-Z znLAuEw`X+X-bipj=xbgunnnH+MdAMI3E*!DwN)2M?-6&i?cLGUlrbc~%WxI{SU*uwzzwct}>Vq(KD;3rojLG7h(m(v`KPGdEa9<)NhkST2aBd{Hq ztFAoxb`9nU?^yGVbHMbTMTV*<9VTYa$7y&MQ>@Y^x)`mcLt4Vf{xS>gWVl<|eVG_4 z$`6nH#v5&fYtKSD*9y&3o31D;VlLxx)^5tlb!isgOR+b-41b$LS6*@`H}^0X9ov|` z*N_MB##&C$smTmBNT#oI+l4_ZV6D#N>gn`j7ogIZsSU&h89SM&^_Kau!eUlXyOrxB zsKdI>j3IPl>GpraV~W@c`w4iL4l_DVl9v*jveA_bB(%T700=F~#9n?+y(12qQR`UQ zcgZBSQH+PEt|#oNd~C((zn0$Cu=Y8%R;CRfBA-_-rlN;XwG9Mbo%Mt#|c z(kZTL0Wcki7bkTcQW4;oW%SEP_ZG+37Scx`paFq9zwzJ>K49vYfdZ%{!Q=iL?rx!8~WBRsLQ@%4|ZisV{i}NqEut#HoLgXofuuHf8(D<`Vz;?vL@W6+)tt zrtG{N!iR}%FfFQTM>rA1Xi#fz#ZD$1@0-4pJoi3%r*o z{`8Zc*8@A|UWP(!i;*wc#dxZZ>0vsj?(5xS^VK(*?IvG1HSj1yib$J`=}n1?%-KzI z5K}u9&m%y_Fc?UB3_8Z*Ltalv52>KVvb+p~V2J7K6(Z7}jtCMPIu2=g}x8;VK!z$esejV8Ft zwI*2$`(}ex_7$7#PGnQ=UgoNhruEX?XF%O)%{DbZgBfS19?|huqOHdol^4?Byu_3_ z1;*tA{Lhvr8Z8ti5#btur`#MB7+vO^^pcN?&NY^+dyM@UZyIPUefLV+dMF5=2V8;i z23wQXuPkG-ZHuZ*zs_ZPvCi|~wJvKf*!uR=jG?i!bj=@PX&{~>>!aB$9!|`Q%g%Fl zn9FmEXs9PWT2gNzli7+ogLIQracN6vsT=L3_SD0X%g#NVhh#^$b5#*Bm}@aHeeIF8 zTMP_~Vv8rpJp%OOxU@m^_DNH2=O&g#`hBL{K(i3h}WyWjyNnHQG}79CFK77CoYzi zz*TG>JoRblfd??e47v;56E)ZP@d0qEd{%F4NsoOKBX2FvqZv&@iXPSPP9|CUI{7R` zEnQ{?2M~Y$=`P6n1#^tgqg~UVAyfJU4gMThUED)c7dvntwZWM?@P994$xYkvxL<`V z@@oS4?{Url@uKsuQ8%y0*F`7VHJI-2@KquY-)NmVGr0(fnE{;;pLLxj!<}WzD$PXx z0^L9Vyy#q^P5B{>>e!mPVrr6=nz`}*Ul*O&YGeH(q|qWYt!gf{al!s!;pucXtt}HV z<&vb%{)C#97YrQf5@HNe?1e~EFA&zT;2pTg3KnF9(XMV(6E7G}_f_YNC|OjbvlA(g z9)3w@6AYZ!c_g6*J0hLJ?h59yRCu1&rCgiDRU&rFsPNn!l258TzL={taLFc#rNw5c@dCdZZGgpb!gx@`(ExxM0x@R+=yG`!@fmTGaao$L5mJ<| zc!Sk;I=K3~AUF@PIFkq~L8RpXq^Qvd<)Ml{Q#0(c0q8#ySw=_&>EbVl6n(aLq0y|O zRYMQc8msn#X!m0>ingq1OiOQ&{~m9Rd-F~-N^1&Vixu8@P_}`< z?=7WnJ4HbVA)vM~zqCMwK~y9P8ffr?i~ulWHtSa3eZ+NZe+VY3JOlY87*Cn>#+ZG> zcPT1vKe|5|o1D_!^7$5G*a$`2LY-Xu%`XIkA<7f_5@PyS3qJh*2g5J1lxsr;JR~2I z1HM8%jb1}AX26Gb75wU}d+&78>TWwvks~pB`jUHYj2V8(&WV6x( z>(Ixxk!3iQ)paycgX4(0r-G>5vnwEsv~xp>AYaFUc)8*qL|>x5^JMalys0weMNs+f z6Njl6s9CO6R4JrHmKu}WE;B?0(WCc9BFMT{R){|gUb@Sb;53_+hR5HUT1_<5VhxWW z;gw@3EJulJOwcML)aH#QtrN9z6E|5zU_?dPVfH(T&ZU{l^<)NlvP#T^#6`HdzVNa@ zZ(bNFkMVI;6qD-weFe_Tp-5J(wB={z`D{X%F^}T4mIyRrnOZ_J2zL-F_=C(~Ohbg0 z84{*Ox#h&PZm|j-Wl$EE3?TNdKBo-J;0l!8dq%e`Z3 zrI;iWv~j3zg3Qo1dr?A+QgAG*Hf+`6#PoolV*TmFEI|oKZ*ekE(jSTl7xx- zNNMn8Ox_x?CpBY}l3&^-1_PtUD2(?pCB3Ud1 zipr7PIDZ;|4xwIpkh<$|ynr zR##cDKns>x(8hOaBW?~j+!vI%2kgJ6m4otPw8pP8Hu3-A82K01JyYpI<{ty!{wA6P zBDorhuv^jxe`CbA0#qSLjCj9L>yQ`QHr%m3m*$m?uiF&9U)_vkF=o48rzt~jrqCJ` z0yU$P$uzwoJk*9btOM)XP=NX7j)M=x^$T#Phb**H`wtXfK2qkZe|fgx*rmpnneT z2oQ|0z$B?MmuIOu2dKmHEH!M%J<++2*IlTL&)R6ysJdm==vnB~iAkt+Z4)APNi|-0 z{}8Hz{WFLyrk1-)<<8wVQMhKH;UrE>#2B=H0u0#LK}7A}=e);o%I=P5>;JBw-hEKi zzfo5DJ0@Om7@F*nHGI)$fbG_+G(nRnYA3$KeQ$bB6VTdNKY13WG_1|_eWt37$Lguz zhEG?+YUQ<~L;9_?@UzN>IhR?|E5WhinAJe@k}%w_dMmD6{w5+4JROh%K{;|w2D<$2g8Vb_E5mT*P0#Vn@XxHJ&H=jf+dtXt)b zfWpp8%lukJKMm-2Dm5^@Z~iRtpmd4}>u+FKL@cJJ4>Pr~(-Gdnd^v04X4weFo<^el zdErHqk+##im=MRS`PU(F7=(ys>cWkf0CVuTv`~3#IONrixCxe0c;f@AQ2oRoXsULJ z3Lyq*YE|+`OHk5f@e*6d_*<~I8zd~?7vw=s5w&;`pcThx@DarJq140j1l(MX2k8suJTlIoo6=2e+0{oRr=22*bQl?oqGMC{pA#1P9Fy zY6<-yU=r(!*SKA(3}HmWLXu~otaXt|=U(X8?pkQ(MCi`pyMwVL%|I%gbkHypfnID;ye zXo{;;UD6G@Y*-|gX32`6_lJ>1GxfPPf~}tK%{70D-whiPcL)|Wn0GKZt=VCDjKtsw z(**AfF-`+Ur8Sh-X$3;g{1JR%fn<`s``0y?(gbF4#sI%}qbSq54!CHeP=l5wr*2E} zTp`G6$Raxg`G2ZEm8?BPi|2+H+tCBsanDB~e%S94Rn4;no}kN2cEKc^#~YXLVEC*N zs~~&pc(8djW#8fis{FQDV;%&N%qpwzgfP-to%7Xll3KG?mob!h!nVoWiwz5J8Mw(0 z8>VnnQhEzeMzwrW7eP(NaQO5DNDkR1b-s>VYvke8$=$##77mj)Kn2?~z28>tVRJ*g zPTJjg6B%z28$7NhNjDbKhjU9)0ZJ8Qwt2EGTD}Zt2NmC_f=#Kud~W*X7=zNn%EcK0 z_;e9rfI5I7xdmAeHAvZ14A*H7a|TR^T9uC0i3Gc+7uJ{A29NLjoit$+MuXEKhqPN9 z)q0nZXLX{kMpvtVbvdFEOm13$MXqNh`N4=stZ%!{22pgcviMm5vCG>*lQgu5Oux1X zNA~4KmYgI&Z0=oS1*pCxpgDbl{wNLeel6$@({*-JET)L&KwsUwiNt~5qGxgbNk-v^xLM>W#PE8{E9$%$&^6%<>p!6T5I~#`BM_L5ZTLf@ zF9bc;JPT4+;4LCdMAR9D_hLG*C2k#O?Mt#`J(B#2D7s8?b**pw980~=0`0{(4n==I{w9D8QNK(tm7+B93cJ2xq|7R3@0@hH$7}W4NFB! z-KK&<`kafPV6gcxzr-*V8d`6HJeBes5m3-wPUiY_t2!t~{I6Pi^zpdvIJLLge%@Z= zP5z5AZYdaB5B57wxwo2sBVWU|3ud;-Av2%R9vo)&pWblplxLjqrRQsRFW(jJ)K1H9 zT_NYBz%g@!FxcZgARgUxTU-OR*xb8TYM#PEk3v-@#}OHJp^w;Kw7wKlQTdfnRe zY*e1xi04cBEy?D1(F#hP6gla~wNe_Pw4hC-jajdoVkeq*-lCXzoUS2TNo=8olIbR@ zM};UanLsHUfK~%j-jFeJ`Mb3L3JPSp2}d4SALAV4aZsTMEGtsUQly-ogk-%6qC&a^ z8g&*^e#X5pD=UeEFENw{iR`NDT$ z);?b@>N0G7tO=ET`9_gGQp`rBYl1U|GgS-NqyS0gK8p+aKB$cthAMnfNeng9QrL2| zl8LG_FJ(*Y1`|p4$dm~FIDN~!Hj&%keBLz?0yDRK+CBO>(>M$1PLjIPSi?rb^iseh zY@u?2==PQ_iTWa=_Wd~3L$tg|IU6Wz23S)}dy67oJ6N)9D&?^pm7F}h#T2w!N8@7+ zBNbUU=W~ZE2TU$MQZ~B_PUkbjE*syWMWRESKF{uJay4!!tW8wlX|wum2b&!|&@xc}pkPTEt8| z6&Hpnj}0^9nBg}j@l8f=+tzNr~uZyi`+q-_JBvtF+OjeA#aCvHO4`{*AYO7 zIjh?XrjoVP35ntq!-`}V_UNP37jhZWYR`+bi_F+O8zCk*gsKwjy^KN`M*^k(TY)4x zVL~d3Xmpuee>Rw_hxxn(hNu;L1E922)FC_Uc!XCDRir^CAo<#T9bA8VYMwAY=xpg! zo*S$ELoqzFbdaZUneyBqY97fkQS>1+y___n^LR!eS+%~XGEPow_Ri$guc=Xi6I~lX zEWnmenAxPiHEI+vHNB`ILqQ6lO^rFx!B&}7{0$+WwqB~Jd1;#ID9xN#8q_%n&85W^ zTZ*{CS7vFO-`{2=vXrPT%n5IVaEGO%LyQqUEh=jAC9^W3iCdwmfLKwH$28@J!4xso z>p@$=O%9g2P%`9KkSV1qbnifM>aOJ4jntWj=wSM4t=xEtdzDhkHgm&8>W@zEvlO>m zK$neowyv%_`lUwBvO;jIaKzQoaRvCHo!vS~1jAkZ^j2e{?1UaQkyG_LakXG5@YuTF z>`c;n&*F$Y+$H5%q}UySmouy&QMfEwxO|hL{ywvll2Q?HlB&@4(*p;hZH0m2fJaqs zIeag!TBtgUt2&5$Bf%x=CVa7`o_Qm(G&yh=BW65pJpoK&^jJqq+&sssh&xw?_X!3e7Zt3$bw*^>AaM}m0GQ2*+@Q!41x2GL5z$$6@Qv4Gs}HRm@g^+8X~O z6eF#trYQ4qb15V>XmuL4z8KXW;}eSmMBJZlJK7n6tLVUO^JhAjkSnIfy- z5dzeVJpxd~2kB=zi={BU?gT7CJKy;@NEw*D7JF}2`SiIKD= zYlHlFWE|#qit)np85tSl_dkb0ALJ<;dj;=nBN*&0g2bBvEG@Hr~a1A|qoc zLiJg7o~dKQA$PH&Hkm!D+b#edip3v0*39TRVBrZGxv4r11N|Xhz`#rBtB}V&*HEV- zMQn_mj3`YU?;qbpNRW$49NMDLXQRbBr>LQ_Eur^AbC}UjN1AyF!NY;my=P!fYE${@iF({O+>0=-%<%ng-!z_J9XnwVm9Kzf&U zohT|_BqPVIJCb{l8|5lqjefH*CA&}M1}_Vx*j_H9m(U(fY{1Z|SyOkm;5=Tqyso4*ig`EQ<2kig(I+y96R(xN}sIT$CM-lB8U#!(bpyO9g5}Mk+fv*a(AVaT`L_p7-3%7a@ zO}&w8*md=NAr^kb)Ca|0>P8rBC0N>d2J=6dJ&tQ1FYoUVzeGCN^mhY8aZIEzM$Yo} z4A30OUKnh979DzfBtgOfD|In~o%-TOCJ&fQ=>n0=0HuI^{L^7&yh~2yY?}VW9_Kf2 za+H@%7L)nh5R3|IO81r%Zux-|3)S`QyVeU4--bXXo`fh{)xe4s&m(fL{KJeG z+B6+4N<^8FOssy5Dq1Fm`_p9~u@|x19ewj6hy4sL7j3Am&5B^c?Of;KdJ9c0F7|&I%3lb z))GtbM1-A>TDah>AY*7R`j~39*U%y``fnpSz7`+r1I^6kzj68ZN?K4G; z5w4UG`Xq;lzMY+pwM2mDGe-s~9ZpV*gnd&B)xdekRbQ<$5@c%D`mH?+P0`=6bk5Al zmXFUA!ZPi6O;4>zTo`sNSvfR<9C`#+7i!NFQO{~ljUWFAHoOgf@Ln)l z`(5EEVM1|tWlX(UCDYoVc{c*l)~{ab%^v)@?_3rG_TJ?z#~m|#a&-BVvdDll()6bN z$O&!`@xF67@n07|JhYafc(C8TMHBtMs7wFfsFVLqt?5*Qa#Q*O1GY@@?01K;i9mSi zK)@vN0SSX8@j0MIP(dO&6uFgrCOUmRMr4_(1SO5(%WX>K%9?|?=<6DYq=W|Y<$;S$ zF4Y_729}TBwPuJfTaRSS@#7?*`|;5ou3Mc)SqJ<_S)P*}UiYaW-%LU^Q1h>2!nEC% zBU{~s`S22>hxhmFpP&4G52Gtv^&wNa>`6Q?=WkPT-ZemY&Q#_>yz)#>Q{zujELZ2D zOnKB`>MGdd!%?`&3>v%X48lCs`Js~zCo3I4g@yTZ6l|AMbp<6WuZLGtKLv$3cq#T$ zQ;|t-(DgTyB-@V7xb83VDr{Z-c-~UH@qYV0`7ZavQ<=}FV0$?3gWTbz7;il?o6=3X zcMA1Qa1dwI9?s{auX_L0ncZY($>AOwwf-h3;YJgzTk+ZxsF2dkiEKRO*CI}6N-dZ^ zF&_IK@adbGDoG24QIJ1yKJbh<58@d#NnS13Np8=0()Wh^DqbvW_;lxs8?ypB6vE~Z z#461gM@j$h66r>5>&K8HtE!^?xjk&YrRrX2g`q0s2BLjdTSJ|6j+O-Ls0lAJm|FLJ zB#7(@HoHG&heU$rf6b!7cR34DAXR9lZO);D3L}dWp+r`!ESf)8T~aVgU=&5JYrz zR~-gtZW>xfIFvG%WYG&)!m5pVlab*w*oEtaVO&5-55{~+3qN>^gSX-j~Ho}I$>+hXWRBoeET!SJ0Ji5WE3JbrNooBud*9vYk6pa3{8+j3Mx~DCi)Si!ZeBVAW}E zn=+ClzNL(KOphIZ!BE_)`_pPvZjDKXYI??7_7ieCQG$aCe45R~k-uF)K&}@bZNOI-EtZWj7I19BEGQ3R&P1f#CN(Jm>Hr-B z;k(}&r_n^=`<7x4b!QgDsoed5v6eQ>RV4kuxnajYpHz1qORZ(U?whh7raTVl@k-cMIBirWTrPN@Nrk* zikOs9ELz?R@H}j6XA<2(v^bO~Fb6Gy?GNQU$-X7bJdOvtH}&Pc3LGnFP&{cA)B!nx zBAfU_p6y##Ct&ihH$i-N7ZicPuY<_qMXkkH&uj&1q!ohjnU*@at5Og->f5zX#_l{c zPMd3n9!X~=slo5^j?nA?)L%lK5mPw$y}cm^6B3(|8S{d;J}w?rEdAwb%pNm?NZ3gD zL!?(1^-a2W3w53B`h1j}=`?FBtL$+S3w-j_OqHhR<%iB^(ir2!}TQNDVf zRI1$rTq&~L$r|-8r$1o_2QJW5$$`!_T1cwqPXv4gP}zV&H2KOklgvN==oMrno+;wS9PO_<>i<+r}!>6Ri#7hY?#uKAO zWAg;weP^@nB{K>7EVn2_@+5eN$s2JCc+G!KwzA#lK6DXa$?fy)iFdW&(?iV$qRPX5 zpDXZ)VV|xW1hgA9C>PN9OT z#F7GW%G}}}(sR2MuF@y(mo&>hEtTmfH}zl0MJ)x6mHbd=_}ur}%-)j}>alyvuKIfM zmS<3JiwUK_CH&W@gd%D6>y_FkVp5YiGjf@mwrcU@x{?t_9pjZGezl!UP1WV_a0lsVoYv^P+={liPi!72^`c zz$FsHWq`EJakn$9gH+ktGYU;)&3g5VP-_n@5z87|6yGH?Rvtc*Q4IBX449* zf*l*W3w5?x4>No0*%|k?1wCc|%=h|2Xj5nZwd}EPr+fvbxY~st+N;fwK7m{!AsE$5 zh6HAthQp4F=Q?X$yO)X%+Ny;{Y^EOyJ^(^Z%s*zO2;5O{_AhZ^L`y^ED z74PlhB|K-g)iIb+SL#*yNoO=2p;#R%PQwQ?0_jr=a*a^?l|=IE=r#OStEf5& za&$l(~oHP1<2DjP>zjv0HHVdWlOZvJUG{46gL1(o@oV(mP`0{i}Sf(AGY_CIS zVIz3EGzlOR592fClTdxP;E}oMhgDrRMp`}VFJaD8q+SI#*NBMsOffWl2-VOQg~l>H zOWIHys3PG-*Z4BnGC>+6r zEVyeig*a7r;^&Kjl%G&kl_w*57c%aQa%i4)gu2+?8w;9Ct4I@tTLDq|CCJcZ6fKji z(J|;@B-`1CTYp-5!FfKlBzXqz-%~oOAQa!=L=^@vD5i1SF+~;MCs$|X;i1t@i$~*6 zRGkjKcT#t6UO0PdbKjIzx&1HB&Vfl2Zp+e@thCKa+qR8L+qP}KY1@^yZQHhORvME% zb9MB#k-)}%#*=!P|;&nV5V>N{Lv7u zl3Ux=CBN$-(k8PMohPdP`3lhe6L#9#HDN;~9*s>wOj6w>t5#bK9yvj2Ld$>txg%`E zH~IM3gz%H2D0bTMk(Cs%ZV%@|Z*!%WxiL!Yh`zsz_j;|99rV`XO7`K(XVVLU!xJiY zi}0w2?(xFw7jg@Q*ITaT3mdcR%uNer*FICD89AI_fb)lp^CkmOJt zJ-s7paR=4d7M33uS?auSZDQfZ=DL;j^RZO%pv#E0=V6#8ob z>sVJ1!~m;hH~^T&oOeBNWxkFCRKu~gs_Y`#aRHyv@@QciokxVp>it+W^SJUfi#1zpmnFwh=4J9knkp)XLLKFTvL_v!N!Y$r zIdQCv?Og@6~kyDxlP;ZdOSw{3>@}$S>e)jMA z+QPgF(5B-#T9bC*ViyDnJy(L7aOeaT63*(bXvIZ(LImH_b)dgOGRVGVk?$8v?&W#S zZjE{wiG(l!FMSRc0QXHbqt;gpA;393+mki|i07xPPq?Uz$n&>zQgad12XLH9#2S*w0jozbC$5Pw>~55=?NJ&;?z}l7}nvY4m~oJ0xg6SnJ1S0zR2cZ%2dv_rb0cXWbR9^ z2eLbBTCFJ3i#Q=has~u9?yDW6PDHxr8tK=+ZI>t+@hldmC*HQC7CU!U>X8;8@mOa> z4U95Wh|XY~?%UVk{tmpvr|y_0#E+_f%P6ROAi)@q_R{2qZV|F$qMJY7>|^WYsPmE> ze4@}k{g^(eLcN8@v56@f!w*>oa^xi|1U0j@gzi>*C?BpLWOm8hW<=>w4&?1FR{P|3ywB(ygW*=5y*s9ACmA3)J7m0a#g#`B3C zGnL~tzIvcscs@Kuwe`cR#X}qann7LuhJh`f!Oqq;e9O%J>*2QmQEtZyW#?Z%MjN*q z7j^i3l)@`f>^$OP?5cSdXKA1Q_@UK$uJvy-YzFoh(3ean@W()z&`(p2+z1{J@xQ$O z>8V0w@nll?u2HLRqvF2_fcO{cHc3hQ+e(4LgUyA8Mu^XU@)JxjKH#zi5w?LJS0O%D zt~gp~KXKwfrnXsBIF37B`V{^|y1NL3=?47{$4#7xo1XhVBUN^qGJU6%M|LLiaZ7vC zb9%Mi^zY-gYBi8_US2-Fs*_jPC^PL&*D`7*Z$fR@@9j-LHDvxdr7`^m< z0*lT%7xEpe3H;*7B3JIDslRe}7EAxz<0!KZTX?Mz-KyWt;`lcOkXfn@TLosQ3Qhf8 zu%vcG^T`-N?+wPZbGrJG574V~Lg#VDoXMJOraaSe$>UHwO;q$AK~4cI1sfdz-~u;Ppzn(Hno&d3w`tXAjmD#xu-4IGWv5k2Va}v#h%a zHP$m835XOqw%QO6EFVTo5Gn}%K5||I0|R}4y#ITMnd**J$|V#Gei1{qq42=t#NM5M z9jJqtDS_SX3=ulUxKRw@v(d}9^0Sjlg&%TQw14v@r>;$re0>+Yph5EB4gp-uR7tp| zYuq+l^Ugm6i;6MU@Y8r1tUA@Zd6L{@~9x?!bx;a%$P|81K%0rB(4bRk3r(aKd$QsA(;f0j+*E!tbFbQX<`I@coV zGWc^mbZ@J7u_xTjH2b9+VMrT}Z(5xA0gOD$mav~jE+hvrd;bqEBlo2*&d=`(?fdr; zE&pkGRI>cvJKi9Ln#+`Ohy%xz2enJc;$rf=`gZ$ua`}FKK11#xZmpd3r3QC^EsJsuDPXWY=0-`mCH5KC zr`F5@$S9=OP6iT#i-&mYZ#Mdvg&ah|6{Z;*OhdjiL>$YgF*qX_5yrnU<`B!zMZdYy z@+j?nvqI)7GBO}LNr(Us1_+^cQ^)eeW@&Wao2P#h@4R+B;-IuJ%@DyiT-l7}T$a0( zL~N% zk(=Lnz2=x(a;A4^t0n35IMWn(MK;&qP%cQ1@0!T8sVLJ>hk====f2r#--F9~GvgK)Sis8BG0?xq64*>gi=DA|$ZIw4%?d9H3m zUVUAny~3QsHD1Dn)NoGE_J#NI+k-q~6SZDNvoYr|`Cs?v^c(DjMlh-du%#P=NkT6Z zS-JsLKXod@{e1jut!~|Lqn-BS=Hx8ZvS-FS1ifbytOpT)6)GD}eQ$7FIkP+ccJAfI zeNb5Tixy@|p=N4t)vmc_HWz}A8g+BTQ#F@n`*AYJQa5psEH|Y7OTd+oz9yt8il@Q$ zf-vs1gor~UM1(XUFE<*xy~g)aF#{XLO^Shf=s`9xPr}xO&5kJP^Z_@W%xpL} zW(okO(X4CdJ@$3)~`Sn{-RywQG7!`Pb5$%#wW_c+N!BLfof|q+qK^)Xlpp4+h%#JHv@Q!w`cC>e>_fG z6@`nGV1E1%!~O9?;D21TmC?8RFMGECV!#mpw}hq2|5OG{-0QgsK!IT*sDT00LAcTe z%maV?5n_cEDpbfuiK3-V_KX7bEF&r#7n?L18l#%f2GiSq`tuJoHeIzVpLgXhwJn#0 z>b>T$C8VSoK>)pZA9Oih^E_ESeaiy7yS6sn&^^-12SbGKqk)4fCA${zRxP3`tfO|7 zFuspFoDHqDXugPN9&|Q6QCK$dNg2k6XqYyrke50qD;2w}u$tT$;Zd|!N!P5HoJK2G z0^lwq-{Flmu7+e+H+C(GMXv646jv_$cpFp;x*{9iI`6Q)PX-`1NxMF<@NNuX@E$h^ zcxY8(Mz19$zS2c8mTrPmJf{M9+3|laZuIb9vAgQK9iM#HwUk^KA9pvM5 z?k6|neSgDYsy$}HmyO)C0c=KY^MTDZQYFpCZesoIcsLz~Kj)U818GF#qd_t?^|SAnf`uk$#ig{FCx}|Tqd^6Y;9de>q%5bI&+-a+Hn=-( zeyv0}P)?We3CM3Gze7KiB3Z>|4y%>H~m7g1j+T_NZlC_>=xG{0}cny zg^2@)=M=Ef6w+cyl1r*~-P`3~l+XovTL&yNrw#NCixg zUsqVW6sq)Q8K^Z{0!eKC``HWw(DK6t-G`*I)rE zzo3S#jFHV5YZPYc*EjM37FyEGPaPkRv#YU@R%dHi5zHBE`bu~hOUIKO9O9x-@$z*L z2N@Z1uT0Hop7IR3=m)cNXEEd4WDh-4xMDL5jODEhhRx)nC2 zG}HK2X&$^SG4&ZUV$U)*p~Qja(F)*qD$|M#kmDIv%@z_5zw0L0~FoMt9d@!vB+tizfjSRDx5J3bYOfPTHI5U*9)u=PGWUVKyv)+BTu>hnuHl&3hqV&OpOjRAeQUD3hwtKo5Ui)^HAFDP}`MYXA-SdzT4U zWor)gtsu)Aa?dK$>!`XM#`cy{$+Vg^JDiUZrFSyRowjjG!_F=0B~_IZ)Me%ZEG$y1 z80`_=7v!{IRf(8c6jXUxBvtbVvvMgMWK<1nek&R}JGWpJm)}|hRFN9wL_6A(h?M6_ z3`-IsesPXc;-Z%j4OB?9rB@W4oQWqDmX}6`k0sAIg@JKUs_HxOo#vKuQ!31#V7=UG zYE_kc#W844&MZpjnv|}#r4ub^4a!X^!j#|G-&`^(dQj_2XyX%b^tVO#CeJqXX1a#>d|v z5ECX#dx!7LrpxAmni*4xF4e2*ERIASWza`c%L)?;>9KUqSoBRDcLfnJ%@7370m-dm z!N(qKJT$Fyw(!21ip1c#_CFWTcVqLC$#it;qkT%2%%Fm#Ypd>e*5%9@is0eX>6H>& zA}UL~1_~&1N^O}6k%AY#cXhA4@ZP$SZb{@9gE@jiwmzy09>wvSP1$)9N@zD%S`Bnp zUCWUoY$~P~Y*#$X8hE;trR7ZNFC*<^ROMw-lv3j;CYRmvo7?UhX|pJhY=C^x!kk0y z4#FidSSk?| zHx=WVJOcGFZNe0&ge}~JVbmI^o*UxFtG-_pNQ?w$R`huhg+G5^dsl)0nf#fhRM<>#jHT%=1O?! zWwyIPl|<=~6Lp`#4;LfxW3Vpf%tC#N<8h=TGKVRjiFDQCVyShpwb4c_uMkTr4$ovq z^zN5cj`crv%|tYwWL8v;-j5xZHMW67yw0yUM;mFE^GyQl6GW_;6ZY&lIKJ>~8VpFP zgwVbN%v!YD}6gH(6% zPLeh_E;yfCnL>R7HEWf`>)4-V=xP!i(i4fMO`kY883q8r#nbM5LzlusV_6t&4qO0Z zBNx4v=`DXmT#Q!Su6B1LB~8`#m)_F%$Y_b`4p>9+sE;M6wKmrKgVS|tD0@?lPNrn^azJ>n1h1ES@_c}Jka~=a zLH!SVVNdpcz3ZO%;EJiUo<&p`jQW_06G3t1xxM6Q<*F6cR#-<83aiX**&X!7sG?W^ z7LbCLe3G2KOnT3BaoQLseP@rzK`SQ%5iAR@Bai0PxKqXSqJNrd57oAM<;;b3&H560 zv$Hl|6Y3(fnM;&)m10F^KE3L=lZnN$UQ_E;y{W?C26I_vNu&Qj+bR#YpkANhC#h52E8U+;_Eg7%hPj=5(*9Qc|H*&x( z!}AS$bRTks_U}4rUXiWhumMxG)@;K`k8B7pp!2XtnSp(@jtLmIU^3l7+75v0E%?@O z#xtlb1-1(X`j!gOZFG9L>mqa2#C186*PpW6+_FKjGDQ4xdCMI9n zp^?#FX4_~6ufxxf_Fa%9N9m$nNG6BB!?rt9W1z*&6In@zilh5$YqcALR@hv@ zf;Txt&P*pvjTPc0%uAHjTM*C@n@uV3ef)J5u01}5KT#<{(o6>zQ0q?(t_G=FSK`|# zX;tsYXlVD5q=u>~gdEPeVID6A$zib{^_udbTW6T5=jSN86KcA5@qH;AjSJPG<|f1>_*GZJfqlZ0 z4_tOmRNP)DOt6~nJ*BNYF5nVJoCK(}-V>Xy2ka{*j1|t%oBGo1I~omoZAhURZoojE zNJ=*`!mH`AJn0d2>v66Rm9P7?!2hK3HH|A_$;!DPBkh}~OwD)iD^kIqFqZSvGSO+L zr4)=8rh7UNAl(VrT@0GRRt9!>{%M_;I`~rGv;6}05lLps0MD7R;W?F690OOhmCj?T zniSVS#8u&@c+ukhVi<*{kk~sJofS^{?#H*FWrg_(G*^-QSyk4$%MpE(6=Y=IJWyipZuYUzaK4r02j4NVab1 z%v#JeW^(+Lf(vrul%NQ7gCy#GeJ1E^*g;Jem1dTj>r+lexW%~mobz{OS#VPgOj%vd zMRXsYuQO}+)`Uhu0xatZ)BPtw=^nXLeGLCsU32>;wtVw#b>89CB5*i2!C*uqp&7DP7%0~#t!cW6W?VDJiXWIo) zsb5;+#y$-hqg|-tlpF}~|6NnI@Q#Aw+E2#GQJ7~{&t7pmIs|are*H;iL21{M2hvb+ zr^o+sKLkS?F5*hjM&cSFW^7%p%_LBuSc6(LxnNb~!swkq-K8Gj1#qk$_(i*s{uR{>+T z>VRutW-8;P`NOK!y=7J~VZQhW@5(a?jzb#e?_Wsv8bnuKp{#mjFzsS08?8wTQTv_e zGZK%XcvBC<%~EuYsYyB9D&2D|)K<}rnk?v8TG(oj&OawkZE|i==3T<`(w82H%GnBduuZNB7D$V0Wv>x>{55@C4XK}^CLPSvl z=8&3VSQ#5qD?YZ6To7>)p{8J^$z$7IGm>PXkamC}-=Y_g1 zdThHU%{!*erQWV`gS>f7n}sSgzb3A7gExy=y(2N+K)o-3XzPA64DsfpdBa8U4A4A+ zUDeTLxm-|X+qXVIvj^IkKA@K{@F+#NkTw0POr)3*Sx2Bs4tpIriE%0RXRLIuo1%4> z5*Yi)99oUfF)VbDf{l{#)muk$sccS}8}ACuc6^v7y5jb{ZMv_DF-tHcU{@5)`T7UE z8-=DQnCg2ntn=?fCK&%yyixu}HGZ>&WHqp`L97uG>B!j{X~Yo*u-zyiG-%ij<#U8Q zi@^K7vCW8|v0VZ1i?_YElbF7VZDdh1aT=1Qb?vL)L1$O1GOzc`%eLQmuDo1qHfHI* zP(vuTLKd3ANO0KZg49L&@O3Pw4S9RjGf&E^@>IjnNzJ}Hd`989G&Y!7k>UwVBrvI) z#kD0)*X0T44DL>_Vxu)%>E-yRT&A<`dTwwYI#5@1>-l^3O|0PfdSfM=VgtwgJ&wMb zplNOWwV^hWv7t4aq~kG@B(JTN2N+aD=T>!-R4M$?2QQBXV{WoH)ImE}TE!LmW*H&7 z?U96$q;c9OZeV%pUivj4U?JC8HmM9M6P61h%{7YKGFy&5MTY5_zgvx7chk)@CzUV` z#~RB-@JxFP?}n04451F|4VJ>I+gP%LAgIiRTx9Ho>nN!Ehp*stFT!@s_d4mw#`w2+ z%7{g}j-_3zMC3C2nYP;RLV3&KvXeBo#!}g?W$s|4#&DGz@#L$~reR}HPum>Es4-3t zzjF6#qjAubsL$L2NN~r+Pu#*yIq%%6_JbP$@-yEce;Dsd$r*;waTqq)hEy2z^@lM# zQoTAHqOUtWUkpGDk1^n2mjT^7MZ;voMMuLkdI)ZYOZm_%L7zIyob8FlGyxiGdd8js zwB@x;P7mW0^eS(gSg$^2&Secf2M&7BQf2P#IjrMsrx!e-XSXkLuy^KI%$ZjWjB z?HqC|25ohiju&u}zXJPmo)_^5$4puWGbim)pVRv|Di=$LnIJ17Mz`ug^uyFHnK{8y zT%K{2xH1+Gqgv=CXyoU>Xm}k@ytXdFdcJAgk7{)JJ}G|1_!D^raY}C>7hfX2f#jth zQ^>=?zY<97<08p9XlF&Z@Q`d1qQzHk!|>%+#Y&K!OPn)3+I{&;CCrR}P#k*kvn-6T z|2Z2(y{qlP8iEj!nv;^dqZIX3v=a`NEuxGA=5bMzVH7LWg)L8miXliY?ArIVhGRgh zW7~(>(+G8xiXR)R-$&lgU9@l8K(rcdCeOSoSX-M5z!KC|HZ@^|0|$VEm?g~J?#FBO zCm-6F)SK^`J}=_$u$h{f21Q3#=6uLUPj6n#z}N#5J6>~q3pCS?SF4VXIlfdF3bIU<%I;7klrGr-^lWT&JOM%N2| z0C_8^eaD`<6Ua|XtQ$#YR9Xt2-85d~T7nN#vUJUQxXQDnd_YYHy zg~}`!wej{`v?gI|j<#Dv(uIcMgY67`^aQ!S8DS4D{G>{qg9pnx337^eo8vj%dhGjB z-k)&R+V-kAlRv#!CG#hJvs$#YU^@$%I8(uwoTkdT{OumSpu%zzU;zl8sux+wq7onHW?R-{Bn4@LD;ffhRt z*zoQ^^bST-;u>1&#BbQM0F9c;iJQor1EHSAkfiXQsQI2KvqtPnXD?lTtj3D`=Zpb+ z-mkVHXit+7qm7SqzFNkY;4M+s;eocq>)OSpzMWFcwHBG5i@EFDx9<<|Clufy`|mZ; zeeZzQm5iMr1xNDWS~NIPFCzZ11Hg7IbMxTS1*i8Y$`jB& z!Ipm)EbITM_Xs#R=)3-*g=;bgF}6a|9DL=bAs==gtB&nVac=Yg3P z^p|`IdQkhv)0@k=qqsx8gG5fTzb-G|u;2{yP>>hOg(r-*UaWpRuJg2y#%$+w17q~Y z2Ozwn!sy8XNDamPgY<)CuON55!QgM?hKLBzUb}581}e@PaT8RNFRh1n9E%{@4xJ!jp%j#w z>XYSbB8rl953bU)6leWJG0jntQ;kk4I||)plPs}(Zz1+`;#FrUi|A<#MF){5%2N9~ zB8w@}ZX)$>9D=L%I2Z%w3HxV^J%e&@osKB-DAO|LVV7N8N?fLBWdYgHbfE%M_xWSi zF+)>yj;qujiJjLY*npANT@hhx;UG+JN)3q3`V5Ne__!1IsxrG9M2vnn)Oih!wwDTx zqE}?;n0DtjKvnyVB{72T*|ii9nFZ7el{jNob>qIQM@3IqvekMWYCN2Idb4Wvc5?J4e248$v`kB<)5Il7_aCh1H>g0{S3A!WfwrdgScoOC=XQritqdbzL%ZW ziSi#Fqx65jSN`YNLdjYYSrx^Hmaf&PJy~vBZXHV85+ZJKDV!xS9D`yI9) zom|rl{BWPUmbl|vp}ID6zbYE7zQSJJz%tsa-Jl?lA{y@i8`8YVr^;U0KqSA#*t`xi z%}f?EigWz;xkb_0lDTE1EE)tdWiusxnl-eQFt>)Q zje~axFC}$k5?*VJ_OPq-!0@mEmxwB*r>`e9sKjv(S_JwNd&qnR- zDyM2BrBBB_R{|UFO=p~BYDj?xLcdOy1SWd(KcBZC(y)5%@?##)i{6Woi(9*lJRArN4O~ULzq7#LW~jtQw=A}@Qgz=DxwLMJdF(;xF8-kZP4`tH zSM$5ofgRd#k|*RjEP@v2C^bBa&Mqjw^U^P$eM&tQDL)1B6KL}th-hE)!~<}JA9qwH zX6114fmu2ktULSk>kH>}8#6`!{TDG2OuvM~4?6u_I!VvL`EK`D_}Z^1^^g-+VDr~3 zLTnBspZZsf9E5AdDv~xvHsa#>jfWZb0X|C}M70Ae(6QThw2HUuYtzBhFZup+T-LzH zXrVnA(S1*;$Hn7^ErO8=x#QKK9d6UW#If5kd@Gj`UCLgYK}=EB-quv3xv{9@FTc)U zp{%XoM(l`vwTqm7H|}G5THcHuH|l7F35WN8Qc{R+7iyQjE7a&a9EbmZY>q;Cx7qwQExFTo(VPO;rj zkYxJamyi_BH*xW_u}Y0TCrKUYs>DU!yMVKQuD=Onn(oTHrgO# z8=M=<@pyWHCe(0o0ReC=E23ddZ(r}LJY7e-BrcE6)zoaW9CNm0I{Pq+2P^zkS}*Dp z@=7LwnRdEZtLK@$6B0=&m@b$CP&viy-~~DB6z5{BISk zCu`VF!PVs^CJ=($aTmB{+-&f#B+rz}{ASDET6*#dWbZmCQ7B=XY^(H?Ns)eURGM#7 zf>bJt^RN>XIaDzXd>r_knMD-ouZ1XGUOZwe=#va>2_32djl+U4zUsrH#? zs;uHvtPR5&R;6sZX$=4AHgYRI$mTY2TdCzcT;p<=C)1p)P7urGC%XQ*dn#2TON^6uVo9JDs!=OZO^TG z43c~4)Y8&+Rz3RSSU$3nlIrg^>;lPfaC+tx7O7VB0c>Iq#-UZX%F?D`-jx>vg31|Q zWeAQRo9@6GQD0yYd1c_pE+$GU*XLdkW}Gbj5&LKm#3tl*ZzSmtGZFieSwqyo7o$&& zAnRWM+#n(dO`i-w#O_|}$hghn9wtjWp9Z?poqEtRaq6OW^WhzwUH*AtnECi2iEkUe zg;DfRcmIfZNtV_3oe}>3F@5^K#>9WY1cGhUQxu@wL|Q}FGOEx!0>}x;*8&t0PuAyT z&&C)vn>7c%cJOcDPpDq~!bpeMkbOhKu)Jqn^RriNYWye)C#N!bJZZN%jx%1K-zK?! zki{v^0+1;siZ$p%=M^}KQtI_H6=wtQp=e7W48>{tDfe1R2m=sXwG`~sd&z~&R#1w# z`Vb7ubk{Q5q%BZItO_x%e#gWHX-22-s*^9iMl4Pzx>L>olK=o!ZjFYEu zbe^Skbe5o7x0~5_sx<}MK&5c?)jYT>qCBjUk??BJOBPVUXL2(L+vC`vV8{-#xhHGC zHk54tatBzoFeQ#~Gbh7n+o^UlB@i6@QDDHa?u`C{>bf}68b%ROk3%0}6qc_(8_8oi z#(D{^kiVl;vJ1`}O1bYiNN*yi!Z0-Rxk#l&2b6 zKoy2pz~}E!_ODtqkC)qfd=gBzqx{v{@3F)sc}~t$M*(QsLoCe7$XkGOb>Q8GTKP;> zvAgr?OCB{W$M2*zkjzCzbQq_YBdA~T)<8oIu?Wvr=?V6;XsObd?t{w*Ll$Y};EUh)w#i`Y^b#7Q6hQd%LhqTpT;^c228p&X(^@Pq##1j-ghbHh4tg96mQWRIY+*t>M)n z7D|#;kI$d1j|vHVO)32TA8_v4p|ccRd&SqHdI|#XF;)Qq5Mf9nJ?oyIxna=?!$`d8 z`wc&hd=nN3tqWzlg?t{@*gwQy?ZH2$%14!6dHTxDD%rT44XHI0v%RszIzKQF1$?-! zU`0I;UOLup(f>P0l=$x?7S6Yn*!Yb`;P@XK{{Pec_xqO@0a=njSD+%NJfj>JGq@jxZ#wt__$m$ZiQ%mK2U0?{W@Te8v+Q{Ng`ek zd%cQ=Y|v@Byo^-ZN3s{HuvFzPN3kTO*?_HkN&=^1_%mx484tC89o@13Jz-RR<>4NDmFgq)gB4>*~V}Zb+ z$M^||7|^wbw&4L)UY)P-;VSJsWikR3@m?|BA$_8#6AJ-~8i!DhwFSM*K)|c~Sr}Ct9Zls${&Hj6XYn1Cbdqpvw zf?Zb1HkBX$Cc)Pq8zNKfh~0SY3bT2mlEC(gmGsgh7|VO$8L&GvtH)!T+s(7Zf@d4m zk4iNyj$=Jfkm;1I|MGJF9d^3awe$;tW?j_Ow74pAf;#d5&EX|u*@TpypKjHKz5@t9 zuft$Yvn2^#=JoHq7moHqmqfmM{LQ?r)E$Xh-8Gok=ri=QZy^Q#Rvv5kk4ErD(mu8P z<{KfWdFkeM*kXHdI}aQH2M^pq{-^rG3|(&x<7L!LdiffU z-7f>92N2#)U#+WuVp}9)inH_ok_9QwHsv7tM4!-wJ``@4*>Wnm5Biy!4Lu2$> z<_}mawXgKP9$dS=Ur$fo#~asN$6F60-Eeroe&%i|piet?03zZVI~)oD)@TE_GEfdT zto~V~gS}IjVewz~I8^=+9Lc3J~=wV?Oz(_Wp!qjr$P^%!l$ zr&$ug@B~CT~jqO0e&`Mm~#$r%&Pc5@?_W;tPU^?S~LO)vQ_#R zM=JyLgaXoJ`7jwAIwSO?y&73o`WVM6V>g75tS1K=fb-{lv;83t0J>C-{Pk+j4N`Op zV~2DAjaNw136~Q`KNY);QHw-?_3O>1sOq z)48>c13!e|k<@yjL)QnP`fE*T0U?StC{P5JW>!2-e}y}U+qQ&YFQ$1Y07B6dOLI@- zvU@}qK{%KXA!caJ42#P7)uIy)KY#`#r}9>= zrGy#f34A|&zRSR6jE2g&T^r65wXZ>XSq>`-5cnBNs`jgBd9AZiCudl_F_a5YYfi$& zJZd^leIPxgY^jKI&G3ra>~DWUp!E0ML`lYvi4qJZ8eQE`f__kWjZ3E_$HUbl6}2@isAbmzd(+Cq1m4l; zaHy;Nm5LbU@}LNytV9L8x-y=I<^-mhDA0WoX0>fa3#U1|^jSL>Gj>I+y83WgQBU{Q z;oj*lX*^_f$2ojYZ*OXHxW`XXOd%d6RjzTwaK=g@-tS;zh+9-EKn2;$~w&zVVZ)zdny3n#-@4oXSuJSOkY{k`Iye5CIXKtdF48=1`ZuB~}M za|`pjIaY2uQXN-i8X@EIrzLsFMFpml4^t!$!-#OKvjW9@toiA^VS&*U{ZNvEadKWj zBwD-A#t(Aid4HWgo}7aT1wQ8PUe5al)EawIUJMCrpc%QgFf0U87f6y3l-;TK;L#U@ z3I7h&Xp+-YAZbxu`35NmYm*KKl=t=WYh@foR6&X!{e9kJJ4KSf)pP`2 zuX)dRv0|jC$;zy}iaM)iY{`yiu>=i}4ZGm_!`s zZjR&LGI5w=<~45~m6(4iSeKDa*LQfAZbOraY#wd-c^$i}(-LE|-C+6+@Jc1>pWZn> z5;k|m+wu6AAU(sJa(!2?dSfgN#1py*H?t=?i~d8 z9;O3g>@Zew#4^5E*K*W*(*0!YFw}@=yw|3sz~*~3cAIWMFg|}f6_mw{PDjgw;T;No zb$wc-GxYV@_R`ZxQh4uQ7|nb?OW7rwS4iB!ccidfj`6|dWuOsz`6xY=NZcNtH$>or z_yr>s7h%u~Aos=*<@ERtWVdaami8>zd*;3zTJoij-G!bvP{Pr5sMY6rSNPShYT{h1 z=W8@J#y!O2KuzjhRfaGU0p0f)!`^$n|HFmvb)Dz=IVZAYF#waOk6)^cD%YU2!&@^0 ziu(b?10v&)yl=Vj3u8lqyZIsBxAbq(5>C0U*6@V1l?nQ+#`^-Nzd*FLvt$tSQ`vP@ zu`1$&16+x>MGohSN>vRtBnd4%Yrwxc2V^WR8^#^Sm%h@<{~YtVP2Y&^=lh;d_9kA4 zd0XLAogn)`arLdoE!^SzUgZnZB|YAwM5azfYlm(0F%LTpCpS&IDcU1(`R9u{Va&0< zPFTtZxp(nfM}aExdy+8P1F~4Z88U>`6-9(9Rm8{ERZq>8Z~g-o1KTwp;`4QGd0)^( zE@8^%@4xKE z6t4CWoNmcWM;iO?=xSG~V5p1g3p!4}76j#;EE<=N0o-i0?z#k>fQVJP~rVS!LBEN$hHW`RN|Nb5N|w2mfufd;CT zQeonF0%&!uEucY%6iCB{f~`zsW3M2jM#`v0EMXQ6>TGKZx{u?X2x}XuS1y$8?gzTg zc5i~tf$SX#ny$krJ5c(%=K@iuWK5cvjkapn%km|=HBqy&A#}WYV-E+SO{ZRdKqIza zA=G}@uzmo6#;$xN7WT4X+pL%S3-2RP0m4Dq7<0az#wpv~6VzeRVykY5faFGLSK&e* zrbG4TD?j=p;I|)y|1D7|%01smVnf6_A`e`;2;uF3(d{5r2k7V!MFygoIw6{pHJ6MjYWOTa$ZxgEo|JgxQ@7)B&E&ZdpAsrn~ZZa5wwdPVjJhTr=MsonSy$3PMn-LM`w@vz_o9Mkl<`F%BGB z4&J>m2!_m`vFe5~nIjnYaC{+j@2Kj=Oue}M19k6cHzRj;^)X;>TF0Lg%((~}Vg`U| z%Fi|CuzePpQ8!x0yUJ5~d=Auuj&zo~15LmoI!^5uEe{)1{-=&=?(4X?s-%6RFYMI($D1*`aIQH>(| zoBe-&UwMKQQ&o3FYC}YV5LDk9F+>>fDc_;}6!wjp4QU}t6GzV{tt1I0hnS*a`z&(` z4Ms=Dmx_+ZMo0CkOM0);T0!%zEW&LDWvcku4d2JjJB!Ve6gHAhGva=2_TZkZ(3Rw&z^UmzhTMM*;;>z3Rw(}jJ=c9>YncKbNcmZqNzS4nc2L+miM^tiz9 zdG=8MWG_6F%Ozn@xooAqhR{17J-jG&&^YwV8CJfn4Zgvms7v5iZHK)(`X!%-`Xu1( z{TLfY5;)*CQ-dkbWRFI-Rm(N(T%j$kkPmAVMpa?my?^-JL)_$hV3!CFgNpEkTr|bs zx6z_Zm-*HafvuxjsYXQzceRi|_s0O(-BG;@g}}3de#mw>$lfC)CHe&y79JscRmI4L zDwwAD3Y<5{nDybqkbnLv7g3#zzSp+%WhOCAh#H1s17~#4$}2D&M}kQP!FAam8*APP z*ApwO-e$t7d!m&|FIgDJ!`HZzxUUb+_FLi44z7hdQ!R+FYp45%!kP>Vr&f|uZ>4?T zXh>bIg&HQ-M-0q&ye_G5kn}vtWcVjWFwBP2v~-#>X?5yhU|?N1sU>OA!Ot*icU$R{ z$cVihwxzh?Rh}H^(>c*j9NE!M0rSc4oA}39ueD8rI z%oC0L4O&_mUZ$n3 zhjG9A3}wq;z)Qu9Rffmj6k+M5{swOspkFLt&I$jc>q#O9MMfRLvW6fN2S4T=rZ(!r zG0Uj3yb-w|gfLC6-weEkpn$hKJ}=){G_Vn}A^gxzhvylK)#VR(I~h|fj&;o64Py~A zyOFV(OPot5^T;L3Lki05d|In+#~*b)H_ zi|pVFB@~sD-=3jA#sry)Sp@=-j=Y*6%CO1;C(IZvpK@+CZgMVf1O@%SXbmjFFLWdi z>TjwhYATNxL|CJGSHDNFBmu!%swMV#0%| zZ=l=OO2|J$)^nT9SLIFQ%3p8Hzs+U9x7$S(YRcCa>gHK;i?oBzF57dW^-baTT6DW_ zt=dPWyKj@}GPO2nk*FH$=C$(Z++llm(P|G$PygffBwM)5R3)O;aH(lEJ0MiS^-4KWzWEGuG(rG&qc14kg@jn!wXbWt6hA2jzBClCVB+|1*2-UxQ zf3Yr6ovpZ`CJ6LPp8{JdBPmuHE!d>vG=j$YMFtUOt$Fx{F6|SPnCVBQvybbuy4{kt zDLK2YSh9_2iBxbs@1ruhRMb1^T9S%+s!lVPVvgOv8t059h@&s8U=p`WQ-_i!F9@e;c)5Ukg;ta&jPs^vGHhB5g*WR#qS@3cxIn+R&rRRgS$u>00V?o#5i?G)D3%QrfLMMEg#39fjUxHBld%wA^wmq<&$kE_9C;Zd>KeEVZcXMJwpZf!W)nRGU=@G3&T=-Qid*q*ucC0u}f1<8%qQdSnZ;hAh2X+WnxX6Hbn6bM^AmTj>LTIFP@bY5wzwe zvtx1()|ys&Re8y-Qk^^-Clod~+`nfPeScx04E5D|f4$!#6>?t*2y-s!37~#&bbVmajkyBN)rHCdjf1MyTo>{d!oyN70nU%7N zB)~N~KCMi!KP#l?--=8~-z)6%WpGoQ3RBY?5NFm?VGc3=a?%t<>!mBsDUku>tfWWR zkErs3I#GX9hu>(F7x_6%UbatDC!N7-tPLGxfX;8y5p_YcJrWT`mJtl+ux@GoCO`7k z%W(HkcNCJ7!Du**z|a_tz~ESk30^FIrkjD6^6;TacoCRK*{qbgC=5NyLXFyZ17d^~ zu8&8q=#TQD12)Tou1yJ@>;ANqXPYJLLOHdjH+=H~*^|ooMzqG%!vOwX{78A_(O$cDR1>;F z?9y~x-P*+1N2J3zUQ9%wZY@H@h1}4UdG`i5b)SDs&yNX53c~TV>!gsi+5-scc_qZ4 z4%FsRn5{uvRFOvX1B7N`tTj;H)}9?7H-l6P#VB%*7a4kSM$Pm_<}>ycr5W`>eTkd) zcx0Wc4h`9&-0Ygyzz~V~yD9dA(Wb3j$ZrxHb-H1$q%$-TZfo8rICU&dH7Ed}Mc)r+ z{^K;*r{icyL1#1^nRfkH(@w&xZC=TrW7@GBBp$!^6{z&}3_%z8r!Ux!P9GikA?x9) zpA|ad>W9>QQO!M%YqXb0^4*J2L`5i1&{u~AJ-94q@}@2MpH}O+9ro(&ecjWB&u+b? zfECd31unBpW%?rMnpKR#0w+1|X~o_+W9|voy@qvJY ztI;8hMdz@f00q-K=-{ss1#>d;xjZ3fCXUdoVN9-A&2@L_zG;#)_{pr#;9JUP3K?-| z)8c%tJ051GOw0~XR~|kF>@Zt(S309y?hx+i9dR$seTvI*n||U^*?c4DMWJ6`K4kqs zr3l@>e}L?@5wpRIz#tVmBR2zq8Ls&9x*H=5?Z;uIvKfAmw!ZaH%BT*c>6UXOA?q_!3B5F~2v1W5O*FyQ< z>i4?iFpXXM0tu;kJ2ZtOEv|Q_u;;(#MXov2Nh5l*DsM5#*7XNTVE-~)$PqLD>Mg{@ zkhgF36FMu>QJ8O&BPMjHzKvE(El7V4G}Xc9QWWHsIoB!6N&>{jjooV$EaerkQ0AE+ zAo4-$d;iTVhGF5Wh`+1yBo)JzfOIg>D;sG1vjaLDPp!B}AQ$co+bQrYf0PdknwdNO zfGK*+Y)?yEiu0?KXvx=fMbv)giz8esVrla6r__V0$Tm#H8TBwbvkCP$y(7w9nUKbm z>UYD)>N4Hqb1r^<;ks)|RHTGzsEK9K!}G-Y>D18}r^Np8a;ggUWFpMe1u~l0W!A4E zCGiV0reZH#&V0fpibeS4T2zYMQlfR_g1>(tOQu+fJPgvul1#gsGT%Hh-<)LKdMj*% zLpMC=&NOK=tO1D_(VZFC=s7!PU3}x*pZThvLP;i%pM{QGO-2 zA;d)qS`W(-_pR{AV9KL;wC0+X&Lt*QF@J7WxX;aY$83Z(*5s_Gui_XmfErwY)tB=4 z$!T0$*vpSDHo5TiPTsMb_~-KFW+s$Bh<$!{3eR3s8-&f+3ujL$7W@4+Z(*r^)?$tb zC53)%NBnyIa3pHwS(md>&wWeP=7N>z0=+}j#OVwrXm>rRD-O-;8-gEV>;--1EBX^F zBHZ^{KZ4;Yn&BXhBza`hkmw1M^6=x9>so$x0Qm>_p;yq?@ECPBZ}pmG^O9CRQv2fc zhB2uPE*_oL#K5T*r_e87p_-p3mImT;$U8>+ZV6C&ZZ!44Gz#)dpdXwG{c^NwMbe?S z2R0`em1WZ9v-BL-K`0ALhj=5t8V0DPT&ZN*5#Fkk^>PN4ipN^KKx^DocDrI3qB`QL zpE##)&x<4-wp#ZGUXP|n~`%4VwH?0&c9 z3YjC%KUOd-m%Q;aEN#m9rD{}lB`&M6<`Y6KH^ZX7<b-@`3oo!`c1ry% zq>yIMs`AR5{-TwH@v_uwWuBgu(m0Cvd1VFvel}&Pg*fHVk~G&R9{0)*YZW5O znx5SK=q|43^`8e!R0~#b7}F`nM^yHiyxAH$LovbG6=}C>gLKDa)3jb;5D?K77FK;F z8NnxB!~j#+H1P8wSqOP6jk%x=c(5N?s$p#aeAcCI8aLV4+Pv=o(I2Un-<4dI1w~7I ztxIjuD1rjP-YHH}3cL!CxJ4+OLn-lFf9lS0hgN#g9KXbM zXu7E|rep|E4FVouo?_@!C{$-p*vv@~3B`H%SOxa(cbmNjbF2q|M*0OV1w=%NsyX20 zq^3ZyzeegjPSG!Gk8{zghw~nqGMtNWinn)x`ZB}t2T-jt3KLT)zY6?#{7_Y8m0h^; zu?ZWG;t)YO*WG4xP9Uy-@y*0?Xog&y|r;w4c8R1EZ=z-FYw=sc=yO?7`_X;@*216N4(Vn4Ut!FCi@GR!f zJpWO5 z!C5(sxKNh!7x$J+`N`Cak}_>~7wKND++dz;ie0zad8Eon8^C>x8U|lgVq#&ls#tb) zU5T1aY-7l_0j!q>Y&3vFBEj0Y|No zF0Q5Dd}T0>$L)+3>J~|Q?^(3Qq6c$K@$Sncg5U#jx_ep9f*K8w4n0@h1|Acax%bQQu@j&=ZBX?fgt`vO@AwkzQZo1C)^w~O^s5-)N-Od>2*6FO_6&cBilJ`;_ z*;pk3su&m39k`++($MH-uvF31?N)`j$8KsmBa}sX>uLTTCs4kF=u^8(?RaX9!tJ02 zm+BjAr|AQ#qle!%B6m<9?x0T*WADgfef&X%IcR`T$QOVbsLqqXD7R6W>yKV95r$lAIBpqi-Nj5 zq7_x!Y?>NO7zIjy*8+Tf8Wf5UvgIc($bmR6c=mZ<2SpkaKJ`ST{IXT(+`u1_xNNwJ z8}8MohISZ(I=}_&Xb=mw=1GC|dA=C`_-H7i@%J#R_i0}p)D#%qys-(*dbP_VO=HG> zN!>Z}b;LK=2aW7b3n<77ftIA5DEq^9*AtHQ4zPz|0->n)sI*0}?gJ#`u5hOsN?)W} zI6in|C~M5264H|JiTFZX$%bU!n`Ja3Oq0Md1R@2};Eh;*HZra~wef=xUSqOx$2(7t zdB10sY_LiJndO{qN&8FdJm_>nXGt#>J|{wZLE#S+{KNIq?#BH27i4^Zf4EDIF3QJ} z?&|nE)0+C)^L4n^R`6ynWqgMLYQHSKG13Nsm{j~d@D)pg^qYd*+#}436F^yR;TyGY zgmsB?)Hm}{ApXIYpB3fD6MugAH)!|+u*)ZrJ}@Y+dU7G%WJSn(P)~Up@k%yLSdrZL zd1}$K5#d=um~Z}g0wlAI$76epAJOBK2LWCvr_=NaeTmV6VwjTfjdiI`a9d_8jU7V< zSUy8fjUO|~{nVYr?Z~@3k%$Z0bu$UPvWzpv2s4YKu>vP&u0dQ;U#|62uqYf7!_5MD zC(gMp%>{nrU1#~SB?GliZm_85|`4RlqI!W+?cq_s`?ysn7!!ow(!Bw-$ zwtOu(DLucl?{kc_^mz19>4UI2p4zh|D5glnCFel4MBkEp5<$J11YAx2T=#AdGs;;e z|A;T#Nlw-}AYakA3HbMuT(OV6>#wni{XZX@SpHS^uKfwQK>Kje!HM`XNm3MOhu$rW zyoRQiARY~2VfGdAsuf&hL0WZKu%vz$j7Na0kNY5AWmX<+X((kObJNwuW#%cD{_I9@ zCjbNs7!@o5Pm2|_5~$1vSkK@hpjBdGh=YOnmZ4T)A~V{OplHUI$_zx#Vy6T8%^*jN zwqXVT^hjf6D@TZ~$3XJvXfC|TFte85z~0(iz!%+o9WCg#MZ;Q3F?Y7yUV?*AXxHH} zCA8w+n90q}uBx;}7;gHsg_jR6}2XoS!RAA3=kw`Ms1(;s4TUpeeVAJ_)VVu5`w zz!fI@I8jw;2S;gH#)3LO5UB4~FFhJh(ImP>2hi>SNR6^$vn5yJvJEfQ=CrjnkFhcc zEB?p}@Ght*Fy0#1?$UL8RxhT*-L^AdpN-VgZjV+7W5h1%ravksK-q55iPjixfeTC3 z-Mg;KO}!HlN6j+tQDY9*WI0x+Gg2M_WU2Rz;-XX2ReRC4_?M`=?IMaU{yC)aamqJ- zmZ%<^Q!f<=tkAuCyo3Zss+rYxcr&^}4d z?sH@`$}JIVVs(Z+g9FW<6Vge*U(|j=C^YrKxV_auA?x`K^#|-PK8LKB*k0{IidOm<afP2l0JIKrDHF64<$+Oiu{I6&Fp>hDnGj}e0W z<}V;1T%Q|`?SCDc{{2_@pAq>~WqsQvAvAu@*F<*x+BTaD&svBJ6`pm?Ql&I>k_#v; zof5^hOETW1nQ^kp5Bhvq*q7^gz|FKn^smCzpow1B;~dtLB*XsxcaVlqNfez9%zgim zfn_LS4d^ROnQEKtIjw*uNvp^2T7ctlnJLa+ds6FLW@=1LOl;OFm*g6qKjG~(P}Kol zjVlX}(Is@`g!zc7u2e3j_aad}Y&ul>($&c#F)0Mmn($XT)K9f0URP_bA{WgUFnce; zf^qr@j}6Jd@UF4w0cI(g>3I&egranr1>R+vlzC(Fjr< z`@JbBsu92(kh_h=%aGXUXA=ndk75UQZ&wRso=up#;nrndNWUYB0dhsBT^f=Pk8<$o zJr>QG@ZXuw_|3WcC%nN(V^(i(lKscOdZ{hTJh#;Kem>#T86~3Eb@|`SCa4e1v8y+5PA2%>QXt z{tqi5SpK2WYy2aw4YP)f2@#?bwL)bDCbmyYLj?*;kI?`cT%n$BhCpk~xn=h$owwQS z?7U9Y=bkZLJqzb5*tzTE`k%VJ(7sgu8B|yn@I)m7gSTRNy4-S|y>Ok$+v?BT{(~@} z;fp)K^FYIeCiKMz%nL4@{ESwF^(C4qko3znB)OwhQgISIloR|IdXCa%?T4qKpmF*)fWZsoowcpuHFLdOLEUofPCimvO z4vC)pQs%FY*RPw`x#^QUV#G?eoL=0+A^E7cP>{zuu$H@Oq(nXy53`N7{#wiX z?cF}-Ua%vxq1?Ne@Lex!Cmlg0Z&o-QTV9KrLoEI}Mu!<9F6vhW234l5lCaJQw^}t< zJPdmBYVB@tRCGtlKFhYcvv^Xqm*U84N7-&y6z;Q-(G}qPB+InC-2Q@~HP5;p$ z{3Sz3zT(llOh{SsCP>xZ&(9tRnr#5mU>^(mi~aosfs$tuTSas^tHvf)t#+cTYT{e@ zE?0Sq$r|%Fj4jUK9?oAdtNq#F&MV3v1oZC9U=Xkr8&d~^P=|DQ1(LoxXpss(0_`st5BUS$xjEiW3O3j zH7HL%lOa+zM!6+uBcg;Czxm@xaQ8Q99DwV1uSctL!N-=99*$P{g}qo>6C*+u=a?UK zh|l1YB@#bjmfR*#_LIg6a{VOhlcp=iM@n23O9{8|Sec$F?1xN~DSZF_=4kBaU~GP! z@_b@L>Z#iptNnZjz0vs1T_^6QC&S8vMGf(;xdrE}ZMb1H{l4+o%ol2}oe>zgJ{uPsf_4l&hT zZ5F+jpZH!Zyap1=U8QA!s{i*n{B|g^(fS2yr=UzdGllfkoOJ9h@ni2$Y&~~oo5J#6 z(^avVXexai>8i}Qow>aAF{Y5A5|?9N+NW^l_lk&E8N!}mqC%`P z^)Gq1ZcF|W;yRZ5;PDSgGU~fOe~Y?M^dO)81^h_=oZ0`w8{{9(iN8vPHpI~QvymL> z)b{5<-$AlkzTl0C(Lcc9iHJ(+U{VJ0v;d1u7~^ovj}nhK-r@YN0(QA?6IhtQ_?W=V ztqf0Z&YV1o%knau0bg6XyT0CVTi$F}_vin4qYXq-c4va09puiLz5hkp++`FNjmWUq zG>je@j}O0x3Pmo*>xG^#GoA69HL@De44Qc+1T_&O6yU#~HZY*4rU?p66{;?M9IcAd zd|OUP`$41p9m(;xv(dKdfl$*YvVF5@g(eT~T|@THy-FpTqelXbUF3*eC-D(9w?gZ+ zupQ^LtKFr)(&EBDxozdS6ku#}^aFz7q#INhrk$e)S8t~zK1vofmG?t`2hMNn)&m?M zq%@`VV^jp~v*PT{LjF@zSa`UAYZ+JrYklxGuf$Rc7B7^Jc#+`rtDwwGn-3 z0&fW(75*3Tga4EGv6KX!sA+JWpvT1pv0=OuWwgVds)LC&tN2hAoNA??d5pa^8y+aT zqecU2i3rMBvTV_`gedyOI?X(HHC`!`^AF3|aw}Dt{`$S)hr+H_-@5jMN50kAPyEr; zR<16~^%V@8dbvxrj;n%e41OV7qm-Qmwz19k@NPw`02^z`Gf#lFK)IA;l3Jj4Z)khu6XFkmU*|>AqIDnq z9M@!cXdVrnD+lb6nKSOpRXUU#-EsfYtzG)}g5KQ(GwgP#bi1mj(_Yzx>?7`NMVJ|9 z4yjDg*|$lJcyJ4qONAWfL)9EOYnt34vpdg@4X%X?mYiNKJj$AYFUcMZ`(WLiBXuw@g>i>;M*9r%U&(z!`K9LqLGjN<=maO~(uus^JeJ4Vjx(9|{cqlX zPzFGfSgPh|i;--|Uz~LdzQA~f;HRJr>o;yD&xaCY!-kEx7!HBKkir)#e<~8rkO5>} z!^jy;EIEZdpVwYrL+8+yxcvx`BA0b3yQ%a)6bW|z*#fdpMM5$`<;87qFD>x@B1yVcQO(qS9%~526tknrcTqm9j)e=WN+m9BJ7dwsfb#yqBkKfji{&Rk0z0_gZj1Pj zm$dX0$v6SGY$BIZzq->5%2A`zJpFenrTMaEX!028f^M9yQ(^j=nx8GAIn(f4&u3$6DF;rvI4S)ZNtS)+j(^e9kn(CoZHKM*$h=OTTKt>jcv4rO%&#gB@?eISD6#OMg z_zK;U7tRTk;Qu8_7=u&2`L`HRLk{m?nNo5*09-l?u&&NfIKTl7z;8NfNrB zl0+Nj2R7g<9CcrK6u%n74MEypl0>M^;|Be*!88{g`l$OoGp2h%a#6Pam%fmYx_-3F z^aA9R3cq~Y?myL)nM7xteC+RL;<$xazvypLxW*GS->DRF9ent{6;&t^Tto(lI3cC~ zB}wE&ujo4Xf-Cj^diOuyRiAB`|3F4g zG|eUSY1Ov0$g?||rxvLR(}~-tqACFf`h2yMc#Wr;a5-^h0-^$h`ilrpV{+&HUCRV3 zSZfvTa^y>bScv5*p7Qvga!&Y8jvVs(KMoW@4!jD8gQO8Oo%q8TZ^>mlBeQWJnZ{;!tgwJRVaA*LN4k+40PmpjmLsi?7$HNNUJ~Ry*J_zvu|bz5 z5m#4vZ*$l06!Jy88ggGp9_aiXXn`MH%gA#-Z2b=7NC+LZdF+yVmPTzM?jpg&X`y*Ye&*4eHTdc-&;vt=*|vUJP# zsCKjrCX<_wEA0CZzox00W(`)_^RO3}lEMkANg{Tn67cqQ^7rd2yYIN*H^FGf4IB~d z!@;NdHa97=T?nI-T}$~&vOUh@03d=sB=e3`cO&bhYnu`|-I*vI#PT;D;)$0YOOmJQE%ML?i-WTUFG`EP%5h(3*8<9UWg8`9V{p8 zebC*f_P^j>=cF>Xm9FJ@(zI<9tYD<}Uu%$KxaAB}<3~Rm>_g$pQk_vXTYyio=_@R@ zxz`w_O0=>PbiY$M)lHc}>v75DH_tL&>oSe&<3BajkcPkV%0}`><>Ywln$X#x**vLN^)LaW{o<#SV=1JxJ~T05i9xIuDq}3x-p$Bn57+Sx{E_IQot0M9SwZI67?292d zx+J8w75I?wtSI0^)@mO;Y*;WN`Ue;NV*MyFNH#}7=I@DRGIq&<0?m!B>R=OfRXys zLJME(5~^dyj))UwlyRG}YBgn=FGeCI53Yi!IgxdybaW3e{i@dyIEK||BGW>-(-1kU zk{ZjAt!Q-uPKr%)5t+-tAJtOqxnBYrFq&!QDLZB6I*Ys|-lN;!3e;-x$61X(&}mRq zD*S!vsrhkpIsI<$wKcEB{rH`%ySZG_<)!zaXwUB5y4l#; zu=Rg?yGI}3YNAB1ucvh- zDu>V9*>aoPAA;>M>FO|L+UDf!f>tpOi+u@bx2v^%i5QL$>s4~IlkmMA)FSEO3(@M`?UE`>fK z$2?t5S!?fFmbczWL6o-$mcS}wOcVkR2jfl(Xe!1|XRHlGEwMHdTWeygkrEYT!I>$J zrm9+)n~R=;pZJq08FtOny1?yCS60ht9#0v*Cj(?)75j7Ce$J!DF1r7axqMPF4dPxl z{+j}53@?4i)SX|hgex6Umdq8Q~Y(nlilKplG z7gp%?%l|5x!1+qBBANXgjw0{N>7V*fK#myFcU&tEsYPiPYk6N#UNZ7PL>*w>9fC=V z4LX$DKAtqiNnDRvgD{IwlkFjNP(1W?VP*?5Z$axB>kw26=EE%ya?vk(wVq`NL-;3GuAN(&opJiHDpmc*515BV4mWA zI}mp)lxgn9983?!VSQvZoyF_Pn{wBxgPL;u4mw60;?Ft_bCZ zik)q&8dtuAfmT6M^@S_JtkjOU3l4yB|EwC?8;(=Dal5aIXBa7X&Zez{v}I_$H` z)N#QpHr8`?rtH6SZn78_S+*P?!0e-!pgP<%5Sl8A9+@Hf48A5E(M4ycF{_O$V9itr zB<{^bb>uFvns%AO_O+aS5?6}4U6t_`O~7Sm#VF)#Sr!hsUM}BWmQv(Xr|N|2OwL7^ z7)id%Dtii=cW`$oyg_|(hEkC<_m8k%N`iZ%^-uJpni7jk!YhGQ8-%XO85tZZ49QjO zwLR6k)G*EvohJ8@)VCRF!Sa7dU|lCOyg@gV+%%oO%aP6C6HjDoVvX0sg`jz9hXwC~ z_HuWm#dP@CSSZa(uc{xa&)F8-xJ-_Bmf0sV&ix=Cb31c2R0mHIu{M}Wg+;@9VLZ5^Hq3(z%}v8`e5@u6F3r}$oRbotKb;kCZ@(z4cD2lyl%ukvWRJx{U2pyt)?oQI zE280k9(zJ;SlHSN9|pm3S016eQ4*-@r3>UWxgn~n^3nu;nuCE*&q_eaXRQ%R9Dxyb z+f>#VY=52g_3GBk3Vzdw&bAL}9ycyG7J~W*l#1QD zm3=IH8Z#EVRTxrx9sZH2T7|RU#N5W-j0&#Hl-w1;_UP9{rLAG?+AJkJC-Mg-ax(Wm zQypluNVT>0(S4+iQQu~=IrLEe5SDlo+gB&(*WAxlcrc9Y4Is6MYmG($&1HW9sb`aU z%CH%~7e*POWZS1M)F@$i+@+i02gLVSVLXK_!8(r(8W3L%V%+yUoA)6IEL=e+ZqMXi zRO}(np>@HLf2OAHuJ+=(Z@0r3mz;4wbQe6eUeJABJ0+w&3sIov26L6>33!pekhLaj zQzrou`I7+L-oH_yEY9f91T`0SCTosc+LUyrAG}|7r>T!C@hnqtt}WYlguint^^4$F z>Kq+|EyeDL8zt!R6JE)(h4p=lKTIlg8uO$3MfCfZVSkb#MqfN6#1`7C?v1hc(#h5V z(u=NRLFF3eL-@CkSFOE4{+gb1O{;~=?ukGF?KfgJC%-RW9*imvUosp7S4;v!@e37V zUe%dI6y!RsZyn`Y_A_3TM#=q8*D#S4ZKkrL6?>1#dAe|OIu_*%=;b+_rq24GUJ<@zd#i^?Q$&Ot= zU7R7S0N!k=|8r;@_{{umJYY|KLRasbw4>)UfyT{UE2JOi9Du?5(lJ#S4uye8`Wh`@d~70jkR``_Ebjn2-?GUPCQAXDf{be`rYBQJQoDd~3g*&Z z+l`byE)n2|K@UVD$NJVZTLb{ahnTVNh!gSyR;*wEU&Q!(f&>exCms3PC69jG*8oby zerKDJs{Hz*v#Md+H=?5)PjXS2INz4a`@p3z{=}5G$17wV%{IF8?-v2mO=LCuXUa?e zGno26QhC%JT#YTvh3)JdO^w|gUC91Xpb@LO;qW+iZMf4{1+&imKgqBz@=Gu!CwG{^Waqs2c=)w^;9hI)XNW7xo*5O7yN|1_|m<3&z-BtU~#A3AE|YUQ)yq%yFskkp%0n zb%!>{Dsw)?KfPGKx{4e`RLjxcwSE&mjVL$BX=U;_Xyr;c@218VjNm(qvvDkniHdX` z&tOtwR|e{bZF*VY+fvwVnpVgenn2p>Fad}Dq)3_<-Z_g5qU)Diq?S7{q>`AAu%^bJ zkf;?C&^j)~4N?11T<&WxN|W$$!*8pclG-+r*Hh@1#~^5&(w{vHXfa@7X4hc9z)uI2 zWz>YXDDtJAV>V23PW#F*omk7GLZ0q})>@e0j++Qrh!Vg_k;D59jljz;-$8?I%ebk)xFIiL@vLYM*F?UO0Odm&rq-|Oq7(HyjH;MI?A+c4eP2$hgVdO7; z;?e5J))?D~%(#nja_$Nj7~H>m73bckX#ZM=ZQo8m z*P-}+5Ylcja`Q}R8TjF9lc$H@Oy=4e_Mf+3?@)m!UZo_F;WTP|GNQr)@{42lt(Cix z!6c$AT#fI(>cWx4?o^;RMc=_7ucV-_@wv)wT=NUpFFP%}j=Iu-?o*zj4$*;Mhuj{@ z=(q3s_MH9g@=nHJ-Blb`hY%mJJYZd$qw&L@gEmg!f4~gserslUf^WWu868CkVU2nwNy4ImCOu2r(j%es;GLtc#&P$ofXwbvbj zha|Y;uqehE&F-47yg~6Xk8zSSK6C6Y@1aa?<3T?{33Y(18yg_Kw!JK5Q7ymQDKOJI zYvaB3d28jDdoWLoI$ZF^i^@BAX={k{cG`br=Mxndh0TJ**TrF}uu=12WUDdBm9e}A zQ&uvy5-WKpHGIi2K<_cr%1gvQhy7ZVmCdu#*#6QH-G)_^#^XGh*w}=t(ge_5LV@+D zWY9HsJVjF+j+FuB-im>=9G5!^F3JTWT`v+!6cQp&S(qCqvwc-8oHEr63rsuz0l=GKO=_C2L zi@J57;v=J!I85!=a9BVE)?NxsY!g8EYjZfOw_!+g2mtz`zwUJiuO(H}5bWY;lZ#YQ z5#gUB*5gfT$yTK-f?f}su9i7NUG)F0YG~!4BctkbRwl&*0ipb#)~L9bshzvozcPYa z2)^p(*nje08=JSjj*jj+z+(~ZVtL}C3af<;l?s z@4#Kz)@=<}Dz^)=8( z?~ES0nLCmXTo%4`*}C!bMvR^cnQ4LO0ckwrb&+Gro{YJXsw~uZP%kW3lbVb5s>qb)g|!8?GljjJO_G%yZIuxu z&&JJ(QMbx}tq5s|R!DqA{xDaxQ{TxDSFn@_WlbclO?75WT;VDd)LC=(Ft#Z!rNW6> zW)@Mhh@7&Da2+B#`#}t!>oD*PleQ50qk?)BwwYmQbIJYpROe!wr+4u(5~eoVudO|HK!5ST)ufE3hlaW*|;bh zdA6)w#K?v-n_xC^)P6N*=@CW2^@V7=Mq-B)9GtL?BbziMz>YyAIWKy>aB=st!Nw_c zl~5@F$lc@$Yl8B)LuIH=CwXXC~!y0JBtG$ggu_D9XJwlBxqN5{fF=abrS? z^&WyQ+68;w<9cILJuH`TW8)xhj^I-nSiccH<#&}qt(UCO+FedqwmmQ~4}PgpIaD}5 zwL5rNyMsI*_E8jl*qdEFU`L zs72sotcGKL%--Q~%-+<}pHvF=6=#*%n6bO%p+;@7Xe?OKq3)}gu%i^ZKTbMl|Tr9&?FU(V-zyAqzCePwAc z27mv+9DQ3aC_^)?e2>?4<54;xx>nxv<)G1e9?xTDE-ub{9qJMR#i3K^nf&dOFe`Gh zXm5$9GG3o3;wAG8pn+}>wGJ;+I}^>ox8kH(yY_A<2N=JK#*|_b2SSNH>unDwZGIY0M0Lalp-Mw*U@KAlmL`ln zAx5^EDv|DPDp!Y#?#S@y5>a|^`Q1~0{Mmd=M&QGUmGs-e1I~}>j^fae{oI}wu=J3% z0Zb=(^o(I-YExup%j#KbR>Ze+o>AcO<0m@Z?oFWXLjiUpf1S;qBp1iSaS`_>FbOwc zJLfhG+-Om@++aucFceQ>q9$N^Fhi_AXR#M}x|$BAe?wziPoL4d%FA%mR8AW}?8F+Z68e-_ zbgWFh_Q#xP&jeY0NvRp#DMw=If^%b481k;1mqtO&Py8?rl~jgW3}$eCgsLP@R zJml)P!hw-?dx1&yb|8rj>*}C0(}}=l)khh10rc88+oRAX`zvNMdcLHGhg=hL zJAfZP6rr?)7mmu!s2p5sBKu9ih>!eDz^ZpC&E=0ep>%wPkQ$gfS?Hq35df$3_N*=Y zB^<+2mIb9QXD6K84aKM_vvt-e+7R1gylR_%zzpQ;b3{aLhA*1r4ZBZVA|pvlgL12_ z-V9xEZ0`aeS@v77$F}3g`f5<(w7id&0spx6BPW`Z;HNE!Ajj4%)y6OVwFUcdsT36+ z@-5p<&U%|AT`I zZp=9mXrDu8%->T7+sx7UtsYGt-8j-5?fty%31q|YB=T`zo7+?D#v!hhwwS4j(4b5UMEA%Ym^EbOjA%QG zaXwBVUU#NS%BT@^5Tnd0czxn|CHZcwGV;@r-I}>CI=WF3Z!1=a0VDcq&>bnflVc^r z-cNshTPi2wyQQhdBS+T}Tr*b*pf0Orj5)%P^CT$wc%s1U|68z5B(} z!ffcVN6;K%fIO-o z-|V4RD*Hrfj>#6YRGV4q$y`CxIO^~z@~H-DreBQqm{5N5w@!d|FkjY~kMt#tD7zMc zrWs0Z6OppJ0oiNO@U63{gv-)LwD8r8dw3(_H)`w&>)}ald-98`&&2At+LIept&EgZ z#==46QmlYsj4fIcN0?GO&fqZXJ%Y4bc{*y((8oldDjF`saoXZbi;#A;b{n?go+AV; z^C%kU2%1Y8S9!Yn_s?d@N0Sk$6qw-^reV{S!p7`@Zn!-=u6R9bXu1(L`>cAs=#2~r z9*hDZ3T649Hjq+tNaMMiaLMs8yeuL7OTp*ws(A^c3NZV9v*^jvTxSx{y2n6Rx|p=q4e^6tjBYmD-L5&zu5&&R(Sgi!z=VKYmcBn*Fv z(KIx5Z$)9KFHB*Uqodr+1RaN=m${ahW2 z$v!yRGZohBs#GL$UEQ+dZ<}DV4#74y5aTe^Q&@S?p4`JSTOWsyV%DIAjd$TF$SWj| z6`oplqEE#ull6ZK5O;QCGYh5hJrwIUl9-8Z?+D{2!9@L8RYwkDH!H*gCWRu$Y_rk|rXXF6bdtorR9;DV0M38Spngw<#N*~XyQm1e}>y6)2(xim!;k29q{V2 zFo;F-lk9jzt`pkyOm2MMu2{jTg6#0fB*!(vN0h z>Um1JCyDccRp3YStzwQ8XMMxQ3soPR0bXq~dsh(e?@2YNvWsNHb#VHkpe=y(3}P+) zFEd!W!%=``TZMJRz9TLstUFwH<)nt<+QZ*ee^4M$@c(KxRsFwqJO9(LB}UoCZbAOb zhiwU^Ll#{q`@x1%iM(xM2_9L{Fg1f(;Nu%DgKvkkTXkcn&U)Z$E%VQ%^i%j-F^pLn zr>s6CanGCf)>)64NvDZ_$yp69GCOZ7NzL;^QGsdva^a?z7t{zVDH@2!v#%ZFnRRm~ zZZSr8ID-f+S>eUA-#BP6124kcy#kV_+t#0{zM=~tklVr1G zTq}u+7{Qs`fLRai6>7>zr5djK6i@@f*?!vRPcCYm23E_Fqir^yg~gbnp%AItyQ2p< zSAVrr`mTTb&LX?gJ(gTeh=U}0J!i3BoeTzk&lI#G`jsAZ_` zTfKesA<}81u<$9~RXP@b+DeW-2&p}XTBf^)s~}sTlbRZ2rU%*VBU-}zoWJE(yj z2D5tQWlAmuWundw;l6^ENkN9lR9t9P&eb$H7i%a6qbds}3J(<&yOvLS2dRhbbf2G7 zMEZgKax?DI&HQ7s6OxI?H^k%>N$D|$v_WIp!Qa)gqocqKE=}NjLbL1%@fAjuBaEp^ zTtZzB!EX{UQ;Qq|eDQV<&FQrn{4nk-7&Uo(Gm%A-Y(m{??=Vvw#6?7;bvbR%tm>v- z!(*mysS$6{=qiDtUie`R^w%7F$_q*!m{r+>iY$_bA4xUyzQAk;m$rxpS8JLhIWIa; zZuJPG`pUlVZ_xkVif8N<3>5#mm?K|g2+IFmk4e(N)j-w6(do;8>|c}E7?lmh4OM_n zgUKWW&JrPvM;fS+o0nv;~ZT zCZ^y1J&$;=7&)cRI6wW7Hz$K;O!B&aty|@>i1mIxJ!pMHGZ4P5h;)!HHy5q;Cn5s+}==KK&UQjAiLmj{}Y%13WYYn(+!TbahrKvhapuK7<(iiF1ft(NQxoT^D zG<(fFOQjx7#ix-7x68~uL#5u(=5@f8hMO14bz_x*q|tcY{7GS*e(mE!Bqv3Ef$r%m z<3>|oi;2~0#Wd7lm^FUkO51qeo|=P|HkZoOourR5p49VQ*TeO6zUlmIgl*FE4xzHV zlfU$=IBFYp|c5yh@970u}>fYRKhPGqG0bQK{xq zO(Qny01^S@Mf>IiDBi`o$Tv*S`Lexi^o3I+po>AdJo6W9t1tS5JGpiDpLsIKvOfW> zp;G9}w-64narHlTp<(52OG3+DB!gmMVax0_2Qh|7vj6a9OI)E_nHX&_cm0w^dHE$n zCz(6GVNrZ|-61>g+>K`Kz*DiTBvYa z98zD>+1NBTUP1s>am8+%O9sdeSFTaL64W8S*RNc(Fg~EnJ%PFfF`vXM#^qW^zDL zHn}mNA+I~_CuVy7P9BxZ6Ft{0{^t8e6xy4)GJSDQ1RvFOe<9#WIlPoVSVkr@YQ6!lr?8lfUuHg*@ayp6 z#YkS}ndjSL4{qiXZZ`!-{jES}%l$WJJi?$v{bKijDAQuyVH`oF?>tKp2VrhJQ@u~r zSD6*;<{n5zf0H44i~ihUG9M6oNuQ^DW4L^Xm}45$6@O$mbu9_i!0vaG_ds00JLS?> ziXCS`xR%$2{K}a0ikdGc<&)8$Ykr+_3@lLP5`7MTn|_RD^@bPm1#K6NN=Xmnb%y}5 z5#0yU%e8bbtvrxFhc)cP6RyFlP#e-EMJ(8^66^0?rcCi^`Uqaxf`_^~bf1>vcxL-| zJgU1xE7kc*00CbfNeurxJo@*!jI5op$-iK!TV=xu*#z*%CrgG^h7I5uf(a#vwOC9N zjxQ;7L`eP^xQ=N)G|8%y7*bb}6&#@UL6v?t3|>aZy$eHAG{e5F&^5z;9`w3z;ynkh zhm9h_ZRWP6jY+rHFX7npwFozPqun#T%%|&((Z`w%Q-GyJb2@-X3QzA5KQ1I)nlVK# z-b@%?WbGN(vX0PRqGWb?lbNnAyzPFb0< zdKh09=Tao(#REEX$QCpv))R(F|yB+Df-w7IPj-zHCQB z?7$k`%rivS(XgzT4ScNTaU9~V6lM$6`x@X;xs;e%XB+g|b@6#_K7kC8`@&GCa;eSS zGM!Zl*~}%>$&MW&CqRwHh-^ z`>xhX_0U9=EljOwZy=uMUfs3QHVZT;Wc-T=of+d5YjIvG)hEI}rz-jo#F?jsedn*L zbE-@J#VL!ZsLn~E%)-e~`BFuZrQ@CWm$evx6b#pIjXE za=Hp+x7CSij|n{QO$$F%?!PHulFuW2m>)k9%WgW14FTocO(HndgHYhcm8I=!1-^5i z`G_OfkWq!N;zzB{6K?Z|xg!e|tKsOuAS3BZ#e^a{mJ*wsNarn_l`+|xNmBRY;q5!$ zBYnR!kVMlHtG3I*Ci6`XuTA8OM4kA&uQyD2M{D2?Z#DdOij0@9=N&F^mk?vo znKV2+l)Y$n+$NMb?Lg@ZRpj%&O~A(V1n z?T+b=<9#1Gcqq*^H72%u)o6K(g%7Z4^2=Q+MrEiG^R^mnY(l_^?d+6f-}aD*k7OJXlzB#vsWI~-1x`u)u(MB|Gdkp^n?p%TAIaeLYEiq-d|glV!9AZdaVo%^s{{`Coo-~Bq?r0Ub^p@x4N%ez1b1hM?kC6{` z(0n^ke_k~IJ|JvowkwT9_ZFVJh#?YZX}Io?+xS7+?H3A%dRuxvu9{k}G08LY7xJ2z?={LM{}I;2g?{D!>f9ZS%UhYJl)q%edck?1CwVj(ET zNU+WfJdr++f(Tx05c!NZ9CIOv1pX zz(g^5h4@YKyuSebp5>-S^>-QB28F>4mw_Zr*C7q>)$g6cZ0=>oU$87&ZPaP}Wx*v> z6v3xsvD&o1R7jfrBnG1*$4h!x+FMARrfsc=fpZmnCjO|SGmXE0Yed2w>)yw+l{cY9i+v+WGgs*ocQOE@+?_dnE?S5(^ui7LUrzu0(fz;H+>7W{? z>1{J7?J4t{=wRV`8Z))k7&LJpPA9Lgihi5EhgW;$i3(0e)d~-rTh=5s%dzz;+%f1v z#tdar*7rM+T0+oQ=oYF~Os|avx{M9D9T))jq%*6`j_S#&i8cocy{fj_YgDX8trYI8 zTQFX)D4c^{CQ1O-70fwM?X#mBIbVM4qZt|W?9_?r48xWs?Tx!B{~PWGN}Ds3Q9zHO zzGi2YDb0Q*U2mN+O&w3`kG&TD8@V2w+mD(b+7puFc|3w887cG{47G^a{A3MfohPM> z5D#l$BDQiW-dQrR{g&I_8z41yd`YPja~q+^7?t*kBAw@%0)u1`8LY_*vn=iM3|38f zR&fTpA_|X~JtPk;l5UF2?~w~eY&4yz2g^3C9h5;hwS{|Hz*KEUyE-uF=tDZNpukn? zi*>fUxYtQKJp8@6uQ1$#Kq)qN~ zO25o#r)P|J+MLm4284grX*G%nW@k99#5TEFSgHwzS+{!SUL~q?Hx?t!6j8~+iW1(M@IKhDgD8}iBi8bvLAInjR; zlqhi6S3!RG!YncMvkIXhUBXYU#AMH5-qj(y+We~rMpQv~eA!b%XN3Wc`ha9|JclIP zc8LRE?cyYOA*lpI?tF@i{xBmXkIxV;AO^+{x6=pd-}yeEWg)LbyLkwWCHMMi>ef1kMi90+dkp}B&U+l7k;yTHK+cGhG$@FYv*k6m*uzssnGgqNtq32C{31*1Z&9a zJogLGyZru8kOGAJT86&}W;B2;xT&+-19!1aNfOI8SQU=^6EUZs6Bsai~|M(^%(isAA%HwlngYc)>n9mY7+I(1hQ1%z_gFgJpNm!@HAb|i}ve4puY};{2xH}e;$a4yS<&$e;q{pq~xGH zLg=5f;k6a13lua(7X@ddz8b0m3cz`Yrq0r`1!t&|tdnFC;G(Ggiw8;W5U3CMP7BBM zXX@81_VK~P6;2=MGtMRZnoFHnxiI(Y#Hku7KyEfPctiz#U43{x)W%meA|h&^P|#+o zclW$o=9t&52~pue*OYbtAwrXUK!6X@i04QQ=PY5*T349;x+L048NKbiY_FvzPnsjz z!&`!V!t50rnji!JD1t6a5~@X3K8ys2ABn&9rd2L_Xb0jk6nF+ zeQW2;v%96QC#NHzIlzD|9ApwZ;mrpi$b;q1m05>}oY~IRAk_>J9N6{v9D-C3+)$39 zFie0=QT<5)p(P=qtdO(@*vTAn5nx5nDlbYXmBSxFZR!2??@0{#Z{%T#ub2~m#hm8< z5c9vZcLav+)i* zn#n@0=JC{^guIRxnFC9HAD9zr585}h>^LH+fQ5q8xEDtNSi~}yM62}^X7Hsr#W*>J zNOf#Kid*W1%t8uvB+EeV$@S^r^b!!r4f(d=dj3Xy^FxlgK>9k9v9JF>5#Rpjd!SC(C09W-L9O7F~tV z%p}W1i*e^9^@%5OkPexjWk0i#sO(qPD>9Vy`1|&j_xKQKTZ2X?`bXAZ*Npwf=V_jS zoM;WOxYlZpu&@VkpRE=@K>z)!pKo#QBm7^6RmsBE=)chAe}whvzr*UUCMT#xu1v!K zwQqqHR%g&M&sZF8S}NFaO=LF{|K;&~#7 zrrN4cz;h!F8U<`h>3N~FBpXwE@^>-^i|mOr6%@F*4$?-AAL$X;ZDb`535iAzMV!3zD=?`wrXvdU z*)gMk-}&CN?qzN*9V?jHlu}GUm0k=HYSs$pE2a5l&XBu&H;c(R=XnAF;pV0J1m1Jv z0;%S9U$K(6^W47|g zUQVW2O+*10@kx53x_qLh@$^_9Ms`MSn!2JU zNK@$m>CgljRXw!=9~lFk&JG_1w0UYX&d=11;9&KQJvSg>NFrc0dtVC$$4V_zxH+7C zai~qCB&qu-$|wFALu2^?BY-jqQP8&zv^8`Lbc}R#e*c7!gn*0y&@nU6`*id*e`;6y zDB!=rY2pjtL(vzM9(?ui{s~T{?QG5dE0_3>)Z$;2;lE;%qHH7gr5WgZy)s_|NOvjh zD)(1v2}qW?wE-repOOY1`98j1rf5ZTwr)%({Gswa0zD#h_LKaeT}MbX+jJw*>AL%b z<7?E_ZsYgy`H9RAL`OUnFBeZC?jB5nSVWjgLVTB=)fGDH8YHkuZC7(+(Fvn5{b*Hr_10ol;X3$BdHR?2RwY;Hg*R(v>`nKx z&gi=^G4Shk*2J@vq})xdm6u5l!LOFQccMX?yZeU>Yd(*G&_6SV==0dK>$b~q)vrGs zb{8Vh5w5$;)W^-kMiL4-Ld~i?8?W76>+@Cs(`>MAJNf7gVUK}kOVqe;QJJ0;9J!0y zCg@j+?TK4um&f2?o-(JgR~XZ(Iq8KMxZ=!$kj{m!rcNF@%Dse2t~HrQEETt&t)AVk znKNkZfn%d0w)ZgA18AV3?Y!jZ`y`97Z*;C>6kau&6(<$7ND0MnH60c<9NybWfB0!{ zp%_c)X%4{rUd2q3>jvmIj3_XW@hg7jY9$x!dg4KXVo>DLd%K{*Ff2atSkk17yG~i0 z?9MU%HlcOXHNAC3?c^C%(}>DieAFd7J3!FEIuHc9r?^f!@a6|#@PB7>fQB(@$|9Kc zTYx0gL+tjy4RM7z07@Jsh2RSqN9YX}x2L42-(-W)#vi#$pnBObK4OJ(vu6umgiSbt z@en?P4{vWDX9^$vBY`G?D>2&rEeRF<`T)q?PbnQ1^a>%vUShOk@n}6hKDL&zxJrjm zYF@l9bC5e>_zvML-kWO-78Apg4OF=`i1UrUUCY#PUHTcCsap-N)p$>T!>&ybr+WMS5oo+nxoMw4Ce;@SM z_5z#twIXH9Z-I!C3^9=I)_s#1@Qr55hY}Y@U?30}B?dMlCT18mPl7>*DMn8>jc$Y~ z&d?48QsP$~5`gUKigWs|i#m^Az}?G7w;LEP3i+CT6EGLfhcD$H(-oEdhmxiH;YP|cnOk34Q+x1 zhKo)|D)xm!nKRLRn&kN%bPYClkdM#aXM((#8@r@Dq*J-J1SUo-Og6g0E2p4Ah>%E= z+nOOAJ(w5g;D=;+Yo5cTsv{RyGIDgi2TGBf|Cr#DyqI%5q;ei5rT!`hONL3sHRKYY ztUpZA2r~>;VQbE_L8(a7lOIP0Nu*BtDb|S{5H4GJveZ#*Nqi1qk~yzo5yAK+s^q}J zQC#X`y2v_(b)G!pUKOqQZqreab){VKotKD3sMtpKG&xUe9y!#~bkTSmQ9n8c!$wo4 zsuY#*VLU#QU}o>TD%6o8K9ZYvVD+#J&2#|tyj+=HFrl=i=s-3;(1uy!fEspjYdh;jCzk#z%%n_Z-^ z%Fz-01BmUTx+A~XG>k9Fklbk7F1iC?Z}MFt47>wmjJ$m&Fy6E;wd3SFlp>#)NyeUj z6tEu);*^EajBtfL!NY;$V^f*oKq1)HITXcYS#Me&poa2ntdSJE!2#KF zN0g08QeC+snc`W1kvQ3J0{dmTVy{DVP;a{?@&&qT^+1xKs@!u1x$Sj;lO@r42!E0b zaL|Pqoz}fDgF-MuKiSWv?L%BwO$oF$hIaVNqH+$TN~3ija>pRM;`<>Di?li68d8J1 zwHEQPgI9$7dV9ethkZSfU*CV^xDo6Dle@xR(Fy8w-N9agUwzGXTBG+K@89x%L37}y z1`R_Ql)VSaL?fgviY^WLa#@P1Qr@>xZkhyaxT$(y9d5))3g5!uL}%|&Vkc|R7$5&FfJW9_+ri~E*va9tGn7M6%Q4?iq>OMu~kG{e5GZuq{I zpbvWFjD}Tfzlv$J7F8tPEp3UkhDwW>XwORnJ#(JrixbPcRX6$ z2Q^qZ=F(&O5PprMcqh1!G7I=P?l66gdle?`?Vo{Zhn_ceyX8S(%gmE2!^cGK-Zf_J zy;8_yk;_tps&p0C7ETKevnnO`fanPK1WZc1_tX+_?d&kRV;ZN!Iulf_69wD{+XJWa zbde+OcZM6}8SIu>!%^-^!|j|O*sjxsp17Sx+dP7BUS^2tVlq=$0H-y!D9yMFHg=;} z3~Lp`cNof)bSE&bxrg9Zi1xK^H1s(x{z{;jrPfIM9H(jXii1Rgp zxOQC0C4-&NGp$-pi6QkakNt?^bgTE8s_A^|9^6(|K34U+>iujK;W@h)m^E4TJWr*BQ&RJJGW8+D9EcKQU=C3V6BK}_FcE+ z>y^lXg7Wo!)jb=$;b?RGrimF&vZNVZ-F&^@*F>j>At*hRb&Tn=MzwJtVA^piOd3Yz zbslcqeTR3kGh4pYJ4#Sbwq|0&X@AEQjwRcD|MA0}_{taEIZ+FP zDOt7&b-rNVf!ciI7I^Va&`6M^i=K7;AqGw>uw0V&Rv6RAAQ|RbttAs?*gYMe=f+7R zb*b8Kk(0!!SXk^dc#VkXN5>Gg9|>a{&0@@V2Wp#=W1i0jgGmIHwN2_&` z!@9n@$HhIggv&AE=U`A8#-%^rTA|B)h8S?Z=W~3xE3A$88b)(6kvDb&1+xZXp{8&i z^neAu1AM8Ty6y6M-ZiC2@=|nBR}9*piXF820HZ^U5eA0C8>9_NRK56i+s6c-cfz_v zC}r^c0)L$p28rL)rHb#&repAu)bq|1C+fStzk&a~958N(|sOWn0KD&-%7AUpYPn<94pD&d^^Z-O>Jl!DW zx^rA-sa5qHty;6Lez8pYOj{2ntcfh6rB(CXlGePwsOqKR<<9ak?PUkQ7P5=S^)Nlf zw&UG2aDBI~ln>L-J zTjYkHWGC11cM_Ww`$}z6{ach!_HowNDOSJ6j~5-Un4q4eGQa_wwY!Z_yk)x&l&D6v zSbFgvXeJLNY(1AFYMbaOMM0FP5*!NL__H6CfaiQ{-eNEh&178H_~ zjz~uC!a_hCIeYa!^JU6L+~J-kH)Ymnaefh0Pr+^$CFU&+O5DK)R8Ns1O|JZHG1g8B zLT;iYcg?`iFJ%a<0Fg+GZG0$tvu34!7uUZnSRt5y9^b3$k zABzcb`{DNmdVk8vR-niiOy67B3Bnl-z{&QDuK z+^72NvuVfevzrSnZOzoX9^5b!`c`mKyUraxML68DWa_<0iKi zYsX-&F&xRrRy9M{P^cJ4DlzN^CJd&KXB`D20E<=7l#m((RTkd=sh&R`4uw3akV+b} z){q^Qh3YjX(muSm1_e`hxYzi>&Q#C_l&P+T5ZmCU;xsM;rSp+o zD|cW!%e8=3z`NWa1CR)WjB$juOxtZWsUz_(l{QV}c0mI1Jydd;$`ydgy?BTgVS>Fd zGH!t-Av~Q0fKd+l!NDb)N9b}?fHbXRVBTLl0sM%n&>u;d0NK-XY@Gb*1~TLII+(tF zcv8yJHjOxLrfYXmI3f7Gl6+<8yFy_^h+`X;7c&UM)VExccMVWePFg?BX;YQ#YyE(v z*sNrwmY*6Ja_Cl-URuOmlRXz#Kl$5wT$Cb*yT|KnO1QURluTRXOkD*g-b8!w$cd~P z`R+)iH1rkGU@=(5LtUz#II-~kXzW&f*0UG7eq`I8@{9ATKI;2mu|1+El$%ul4k*X2-3C~!(i|+xFC9c< zxVeeISP9+N{SbFI>yy;7sGt69o8D)6UxHDS0fUzHgDBySzsHdy?v_9B;*uqQn6@r^ z51w)Gae7KXBoZcDp7s_iMXPCEd^wQK3E4r81P$VO1~#Pd;YkA*V_i-OwVZZ8JF<15 zSAJc{<;wW6p@M-pBNF%Xeyu0<_d7a6^p#?ohKOreE7pCCyaP8b)aB;p_RORS?fxb~ zM@JHvsiFj?<4Ixsw z*=o?@@3QDBYe9E!;<+_T6!g(|x)cUWeL`jr?QiXyn-}CnV2>-NVIHE!>%eZ?I)a1! zaF>McCSc^pON)O+_5`C{>`$|NfUul2c8(?XEM%{;Dvf_aaHz@$p(Z;6Y%ka61YwEl zNftl4hu*}s+pI=LPuIijr{~DHlgllBM(f$zetuEF* zD~gv{m!VD4qIxPor?0?T>6Wmvw!wr`{3?!ye}OJyOu;U-_dOD)2~AU-&asX+0(%{L zjU9SL0`Rzuw8IXyxJxTeA3ey0vl$`3W^3Qaf~*Z~d6o&rSr%iPkWYDefSk|DnNRG( z@Vh~A*G4+vB!qp$paoX0pCr2xNY@gmYYONE`)Cr`(+o)6_ICq0tad>&d%!+GZR7;i z`1w#2+?uQ3wQrPNt6yA8E@Z~Y`zf9YvV#riuOAxcYnlJG43G}FPX{ktW|ausF%V;R zoIs1O7^H`P*}7Ikn>RKzyr#6G)N*8jZkvaoBC?zyGCWK!9kpduNd$oV5dFaf$`yx< zX+kcY=Ib#QkKyPTI<>aH_8_?aKxjHLJQXd4E}1q@$!Wu=Ur^rCEIA1yfq@DoA;JBV z$vA3t1~tt{BVj7drU|1r5gU4QJn3*7F4L87rs_HMdi2%O?x!!atf9gVG3p}m(Khmk zePOFH%+0N;c%`%=nM9?sqLjEw68hlyK{>ZJPv$qC9mi4CN+#iNK+xGW)?uKK6z6X0 z=YxuT;QO$&8<=({jCw1vBXW(iDJ-SW4(^sCbxptWn>_SRj2b3+gOwWWK_suxWpxfJ zDV$>ZSGC+8eao#J&tp@MuzANmLduO?K>etyc{dbYA&Vf;3(mPts~`^f5<|iZkhm0r zbkv8%d^fpLs6h!z8z(DW9r#%Ar#=P>UF8K+K})fJo-}(3RZ#7fAtS(7L8VIG_t9@z+cz48wc|3kI^*#33Mn|SRL1w7E`pH zOWjN=l{0=+l*AtM$^Bvf{ekhYzB^I9wy^PsX^Ics#4YdoKBl2J-<2~>)r*Vt!bQd# zp5%CaF335qc7U7G8{@W`8matAzsfCXWOz#(JE(E$yK2*eX0n5<>XPkb08T0}{2ole zOoxyvS2R$9QOE$vbMOEW z25gn?0d$NG7I4tU3(9qJua^aL#@#0c(6;UaIn=X(1h>d41Ef`KzCMgu5@k8&AXi3Q zky7gD9CPSX#U$jTE>-f4A;M-1Xj@o>gWd$mm7DzZNz5-G!xn{hg6$!wO1-(ExsQD0 zj!7IBCzEnpbeZIHZ7K#!B}@JvL@bCn$;ig>KOI%s*>W^>l)ONwd8fh65=vy#L}WX? zz0vXNU*kUi#!eY4KH~VR0e|vULH!dG>94_^l83Fcf%}(Kf{FcKBklj}#3g8ajf?;x zvsa1cTWRYZDhePqQgdp_#DjIn8 zvTcQ$xPebv-JWa9)wAMDj>ESy!C_{>T_wz&!|id;yJ!Cv?9KHIdzQvjkehHKR_1|q zCU%e!O6^&Q+~~vM6ezFr$EkT~E@v#MEFeN}I=tP8HSEg;muOLxGI-hV9p8Y&>ReH{ z{vCIK7K$ETFg*w~YZj>Ih(4Tq)q0Y1AV=u>=%wjAt2RH)*)+L1+0|-FeE?mJX^SG- zh|tFvIw{>S)IjE!y4lpsgT*Rat8T(YSk3K%C6v~6tk2*j?v(P;(u6*h2I9lVl1*&e zgP8CjBVyIr%;n$a-y#pCBUfJpBT&?D-|+sSc>0g|tD~dAe~wU6)PB2RsiAy|Ci1j! zw;+!~YMfJ-NLW)0pGQU;~U%?%>_Q{v>P!RUm2T+?|3+-+4xP z;1u2PyW~`My4f%{*|7(M8t7?eI-2l&txH#VLH~q?lIk{%r8=AEOg4*`sW#UtlO@`g zE^LeiR~Fk=Mqo*S+3;<`d^FQ?l1sMr%=tKKj`yrWrUCqq*0`#RxHXTBt3`yXOU?=c z)>CPOGMA+pt%{5ahbsF4?AiJ`FZCDnC`fvs|8Ecsi)59)vH;PBi+%!(CHE@aI)#Z42POP&8>6|&*i%M zeQKQEJ$Z`26W!)6-rDv3Bb5VAR5fbpPEfAACsb6GPVWM~9|tC|TwGNh(gt%39~{>{ zUPtZQ3}aKAN}0|k^M!GCH`1!3x(9XVm+!$Y0yyi*Ojb2DPrMn4+R!Zo2ZuU~GhdQY zB5wv#A}t`<(?;|BviHa) zBYW?ekyW;&M7CsxWJR`&WUo?2_FfUCGWws}damp8@AqEMt=Ij0&Uc>gIlnVXCUH9l zdYUwRIcCn8Z{zDP__Hd=ZcAwjwO*;tlo9KSXuG&-7-IS4JKFMcR7_EJM^aNp-HkRE zdaqb1PdPgJje;1$OL=tL)JbHT(Na#43rxj$uSyq5T*axqBKrb!=^5kav9NVr1Z|N? zuBmm8J5dyRIiT|iE)WQem!dW5ESWxBoNgV^nKSRub(F}tdndJobF|l5(c{53m@czC zLtj(;oI2w^i&V>c?iB6X;@ZOZ=QvkV+{`eDt(#(G`hQ_ElPdu^wmb8Fb z|G?Fstf<=afw>EAP9ih8u$82&y%=YVP;V*;*+AiQqx#m1qBaXGPn2$9>m}oD<)RF> zdsg#1GocSMsJp0N!Cq_V#nh+zY8^d4y>b4T7G8DYSrP@&_%DLX#KI3Q80c*wm98#@ zBxwt8(l!-58T^=OJArAyW;8F(T)8GxOFT+aeJ=PLLACz5k3@44UvMb;^Qx;^O^(RRp zS9HXqgWBJB1f_8}C#Z(*Zdk-iEya}oh=0K(ceZeqi83ARg7?JQj1Z~B{z8^jiYQBB zw#15opLY4u7CVUp{rAB|VP11GjZvbFW*nw(y+4eeVe7T@=*-`cTQMP`-c!^alOd<# zS(mz4>-Sj`T-IW1j(_TXVaS)0ih8G9n%_wqPx77<>=9M=(l1}NVW_CFpLuNXfd1U$^Xgy`bP$0}ZvHp{VG_2o1KHMX#uw%C zBHx1us?&=DPUO^MDmx&Z!&G%BmVe23{pId$+$gu=B@@k;iCQuvS&v7v1Y7Fc4Uxk? z7W;h>7OujORYHia^zZEFn^OHSqxzf)BZfNhW~_AU7qhtJ659Evh@Q-=s{RG*+v^kO z2h*7z63+8v4aVkV48BDu)i-@*EHe85dGwiz?lmfbhA-OQO~JK|?}n$VOsr9ynGI4I zdIt&WdA{l67M=*8n)83okR1O0%oCaoCv<%|zju@EZ*D9bPTF0&u)4tAv`#X9i(KWS zfZwBtr%egWPb$_Q&3?jusZEHhwXI=iTZhb3iy+j+`!dj%Brabvpc?+&n;Z}M`x|l zs%RtZ1Q&*+-{~huZoRrT?{?*^XnD8`Z}e*HqDyt!fN1&LAZ{gc#%1YBv9}8p825LA z=D&0W_I!@E^K$`za$~6=uz)LqRi-jMgY>Cl%l6F+E_#M8TT!d}eaShBEx!2Adt8Hi z<16zRb{Bwa(~0R7d=Vx{}&6`@up>qman9=FNaoDGguYfzRz% zRL`ZgNaZZ0v^-_5J6k1t=H0i)RC`-XoL_zqij0jU(O z8ag?f9;o*M@5#s^i{J)cHz&R0b@A2BMlty&#NKFp99CR(D%J}3S7tWxE=w$nW1lik zdJ^s^8-1oB>J54L?k%_am>1W^-ue!P0r`e@28XH1(Cdgm?NiQ}js9!2TZ;`i85(nM z1<~AIvxCjoxK4xv6=y>?b+|u&a`7!`iW-cS4I&KkE^=$OuUK)M8Oly$i>Z39VlKV1 zcX*z5q&4c{T&j(DmS2lecxKy?IIk_l;D~p^0bvorE|4Vk|9cON@Tw&Rg=V#)c@r!d0Jo%7|%<5|T$I8EU9yY^xFTUWWgvipOWX(9}Y# zYeNx+r=AwE#k~!RXF8`tY?G$bG_7P6NVU3y*ITHygWl0KeP9!TS>DN=Qea6Nu(TzQ-FsilSHO$A20qDT4XaPhwLX}7qVa3WVF3UB8P z(T4{>mMGPf%!(Lc5XV()ICg>E{x#+9n!E@+MT=<>mEmV8^e1D2k3+-%7cX}yk`xazIf1)?R-5!0rp<|Wz&(J##V zZw3S#YO7RzleWGaOHTV;)VzrmtAYf#gkRBQt9{Yp^-yISPKhWxWuF^rkDHS_|1J@$-Z}zvJa!Vo>%PQhv1X7|F`N+Mb-CRYT}G z$uj=3QskP&h0{o1BeF}sf2$RBi)DF>^%k#A%m!R0o4%I*`E0WsUtP<^0Nr$-yQG%y zKag2%>Nj~u`bMG_D^kp0c9MLEVJTE??yK270UgTYAAH+X!D=BPu!iD1oTC3{C_DXL zB&0E@K`Mn?gB9Q}f`&;{ZJ;xXj~i}7A9;2yKpUS-hR?4E@p9LBZuaw?4C(hCE^kXs z3w~6jsHaalJ-VS3>Rd%@=ua~>pA^4XP*C}NbA@7e_hYIr;){^$@<|eJggeBsCh58i z+UoUAyM7f{$Wn+fkW?@yb;NN-Tjd@152b6u>R_%gFzBrpJALBq+C9!)wHk`Vub8p^ zw8-elGN+|*ZT-}ov}-g-3Uf1TC)o!u^Ya`Sku8xUg?9^A-tS@g52cJ=@~_WHe1LR8 zdHCMCuBVJ=$$&~&LRn5?p^9^?d7g&|@tWpJzOehni7aa7f~x!X@5SUViLb7?zWY?% zM~cthV)7~)z5St#KXw15mhyh5uCAdbB1f`!<=(@tA^pU9B`I3>9?ay6>u=+xF{T%q zd3U>Mps;Q}cs~;(Rzs*tq|{pK;!TAzz^AVMz=N}H-5fohaXJcz`@++VO%$rYt7W9E zd0((ha)l+fxJ+MZk40--+a^Fj#f;xN!&i&*!l0Y_ExWOX)!^qo&L=m+XnU7>6tSXJ zGrz0ScE0}j8D;lE-J7gY<_CS1<#z*m)vV~o!;^p zZiN$sy0~*|k-}0%eF9>i81uMmM_Xul?8cB{Ts5@sqD@erZ8|N$y~YwU_cfq2M{TFW zxura1WmA8AIW8G<-Q_kujT!3&vAtZ(BD&$|w2Z{pPc*PFFn8GabDC-iES}Y$^pB4D zG-O+ZV;t~RR(+ehmSfEpH=1ZhqsEdMRZ+9%vh1YWieOo}vm3bQ+N)twpH9rm_X*C2 z`1r8kV$bMjfdOB!$MW9cAYoCCVKFYB_i0kf#kq2d>}l8I^Hh9F;ftE2I1-`TeF^eW z_AR5rHvY38f^%Nrkr=wYbxxiA_*N{&j&HPUW#%2;^Q>fbk;w`*+3QoMs#iHx3Zg{? zXHGo5omaA2sTM3yntDgrK|tus4I2<^HUdKJ3SOk|8#C3H)v{IXN=tet(YZ_d#^O;t z&%E)}mMoi+>n>8sC5>nduN-YN$!KJ26K?JnA#A{0DeAT#CS4)*F8?~SFwe-FA@M>K zO)4|+ADzxj$X^`@k2h1ikw&+Q3f0T-zN_SyK=*jPpgld5?(thEQ>Rm| z7*Bgn6 zO?X~?TjPTL<78&0W|1reOpZH-L%J<8BLph4gdR-8cCLygxn;v^RX0q@0ykRHt1$92vMT!i9GY?N2Y%ECWQ#|mF7l< zzlP9CrNn1(K`o0ui!_<1HoAV+auZlCn8C=Td;~tBU0Pp>q;}bibod170;x9=`GzFb zc{?M^lgAB^>9=|DhDOJz3X0Imh+D)@@%LuV5GM6G`S)Hf$}zP(n=O+-d~GDXlZv0= zGKbXI_NI3|c4xowJzuJGi>yJRqY_Ql%XV0rZ!MCjHC_qn4G?%YMf&Dco!onpe^a8K z|?UQkt#lq0lq;6JHZs>N}%fo%h zben;C&+08M?Ym^~HydkRGuV>0nI!UG9+&M6rHsZp@B7qGOG93Z}x*4`f-x>#fVaX?KxFV?Nr>l3M9Xe{omZ*kO|6yQno&J_lQDHQo6Y&7zxQ{Hn=xM3 zRTq|o+UC^tHnC!A7pOZ)yu(||F9RZJ+j%ukD`3qr1!L!IFxyfw##Q zZBNj1dJ}o|W^k)}yg=J@usnq!mAm)dR%H1UlLzz_Te#7|@jE$wKJr&4MwM>%%$Upj zbWNX?_i3HhWZJ!!IhM1j6XP?NKGuIOZsXj|O?uBZDc`I*_NMDMH!lh{Hr6?Q)ZY6- zwzTbjZuiz+Yt7tEgC7An`T8BFr%#JA$P47Ie|Hnv58u6hAEnuuHqOFan@JRpjS~7@N(1$t@NKc!Rl;H)+1%`K?9we)ln{3 zR8syfYS1g4Y+{0DFxjD(KEW;&QWJoIW)PuQ*}D3&mL|}EUMoCsn)bu&4I(t$6{5tn zi6`pAIhgQE7bRL>6W8 z&8rvjx#XR_&wkyo5{VI^HY$8q`S@Ne2Q|CZ;;x~wR8oA7AvHcHRo0`C3u)4?N0?qu zT`$s!Dre+3C=?UJV-OgzZ}PlxmYEo9e&PeaN#pyydG26Q1HF~PHQMvTQeuY8X3V<< zr+DOoL>PO>wg~wY92-Y$ncI=rhp-nJ!-Og#viM&q1 z8O)!KOi6}PHfrR1-HUF1XJX;A>^bkso>WV@JGabT-q;Y6hQ7MCOGqF^XUp`^o(6O=Gff6B_)(;#a)s<+;e7p(-Gs}QeRcKI8>hUv5x z{=K}!$M+gfYA;2ZAcjAZCA4A+Mz<2m)(X{q*h_HBZoEUC&IVg}yF{K=VNd9U!S{v^ z`He|*MMY-5iLWTMtyiP(oD9IIQV?{&^=H7Tl;xecJ&%mcxIX!QZgJK3`b%1aH#SXN zg$bz4A7$HbBr)@6;p@cTG~kmcb0~fwhf$P4jQh0rePN{f*{{j1m!tg_U#rbtpYD#k z=rpO(b$!{s`$ZWG8e21BvWp11)mHZnLq*@Z(jN#y<^+OcO|>cWB8{IhX>YYlBwlJZ zKA&SmN9;hH(ApQvA2_+3;^e_z?ds3c&VoGl{k0uyR?$a0A+15(b>a{Fe1V$LSV@m^ zP2z{Ltb%XTmeFr3m@WUPH}u%`)Fr{h`{*Y&r0~N`JFod=Lfwz;5ocswkD_h!7Ypyp z`3huaL#Ir*+XdV8A~n3cJZoNY3v;}aW#stiVo}18b2n&aP~loD-FO3r>*Sg=MnDOj z%m<#}nRy`sJ4N%FSzXeVA#NWKGhEFY1Ce>fBhAaxL0vZa(WVP+Je#gkQpSh`B;NV5 z_eWlYQ5w?mZdtlSHRLk95y#WZ>Se8zCu|m&ADE|o^E3=!q2sZp^&7F8ll7JpLz`c5 z?M~cr>z{e8(|`I-QJhfMW9!m~-!BezPh(W#BfXZ$;xOZpr#HCso$*mG!s%NDZ!Dr@ zqFQmIaifvt^`&ZZ8QzM=$2|=NCteaRB*cl`%#F(n#ZZ0{>R~`P8_uMl7LH1yM}5Z1 zg*-%y=1h8kY&pWi231}XFQ0qE0Sk#C4_P|a#PKfvAfr%i_DIm8vbtEL#-iHnr0}&F zL;a1_W5qYtMwwcfVQ30TY^*&^Rz*Rr*A((kFHO)Fi4c-=a7)u96YI_Fr)3ttfO65v4<#Kd zAEv&%^KYPj(w#6<&`naFET_j=JzVKDJFaqf<#pEPa24-my3bs!xfAobA`@prB+*1| zWG%Fbh&Yi>kQ%@2HPrs@vFkNxPH{u6N3w2X)VT^&9%rP)FNK|EM<+3G>W$z|P&F)P zvYv^+*b1gX#uDV9{yg0kP9(Lw!*;H?l95icB8R7*0w-;6`-iJv z%ChESxUa@A-jy&Q+GC)hlDeeQt>RxXo}f}hD{k?ln~6u0RghKd2l7OXqptX;yWA^e zbKu_@J7Tn1b>mlt?}^*nckE2xMO2_WkGXg@dR#5>6wjvZuNkPz|3P+70W03J10yD5 zYOSB{42h)SH+303VS3*)AALwLvRD*Y%x8WYrif4$pnbK?KAOs@@gn2?O#+c(31Zq|#47IPZz5G#Ln~1qf|kGd z*~}r6#+;vLm&Cb$Lhy0gkc^6zJ#{yVNAI-!RaO$$fKOKT**k*K(}|CeT)16M(dSrH z>ga|G(msE$+?Ej$ToV7_ohQ8o_SP)t69W|w6E`o=kk1QexSXOLPx z07r8y*b;pZHQuBo$yVx9=9kdyC*S0(ysNKxJ==QTurfB~(b7Dxi=dz!uXJ!{QPO>f z_s;bfCpZy18r>|)L{)Hx9$&MV*7HfwDAv1@*waGKF^Vu($AMnL;IEBR@ygFtaGF1% zzO>u=2yV7Q=t@IaKXQ7h%}Gkt}RQ3yCod!rDs|(KyKv5f14qbOnDZ=@k7L7(#OY< zIax?W_3Ft>cr<5vnzf!tYcKNc^ix_<%IV!teRDfi%@ft=;jRmHO&=CZH)0v00jW8H z7h%n{fSMrvL~<>z`fFwgjY}F#-#nWNP`{A8pgn7=r)R#arz5yjGc&q~(-ZLB>Eoip zrM{q;OZxBEk>0OMKI2=TRH6~PYoZlqEc!UnES%z6fLq}E3&d;-L3W?c?HgxwP-+P02Jh!$4QqYDU&_BB7>j_dcjAdOO0>%b8Fr?7 z9fJ<`spxN~k`X_jyxPL~sJ2niVvfJw&hN7R1WD;fm*5;1;c8-B-U`v`cGQN(K>_l~ z4}LG23Bg4>?%{be*dft=FKmNec=EO(Q@*s5c|U~IZ#B1JOR0hv&Q}$-?e~@TQ|NZ| z)#oTob1(hwc;?+13g%9FXs(^3?v0o!qQmix;n6wd%_%jO6zL`!9=dV!)@zclDKxok zYPb5fm#Y%{uQ<$G>A?VMzmrtOvFZ81Pj{njg>I)$q=rH&V`bO0LqPaJKfC&*t+DC8aX=m=RPdpIwzVUIPDPFd>@6BU7@S;T*o8 z1oxz6?&twEBb^Su?N*Ez?mh#mM&-0s86qBmA`cGukN}c{rk%=>zKv2A8 z`2FOjprO>Q=X>c|z9lYY*JcLkm@wL3T-zpE#ckibP$qUN^M#BkU2}+vFlt!rbY}y~ zwy3Un>Vwf8f}14HnBNezilb^`pO|S&x)l}dx;{6Zp5%{nkSj)S&yLf+k{QGL@8X|RH6D_5rSMS54lMA#|A z(oTIH0hv3=>~G5Wt4O9Y%E_+q7;1l@v8_u+VcVRfx0cwET}!-ku4S5RXLeFE^c*Q4 zug8^=2L*8%Sqt0a0&Ta& z7;ko%4d)yB<=gocH$y9Ka$?Y5jd+dQK-_JZ;!?fpbM5+2L~icP#pe|CZ+adsGuqy| zv6ARy;v2j?*S4oPsHxQRDmQ7XZ_z3$W*{hcQD!l2$PjD!!T^F-H9d1prftk6^PGMR zZcKNzSvnkdjoi?QR&la|7RvLq!nAD&>WpYC7o|xboHud|seFL@neoK@BmAsF{LH)< zJkMmOkVr-j%OYM%Q;mt;i!$bqiq6r0VYx}+SaT+}np}{O*7>>MaIke-0=aBQgJ|ei zA=c~8qF=t7&Q4Lx3Ey2VVynK9`LuZbh53YIcf(~$)oLp_Cn?eUm)AQdeWe0^B=h)w zNq;9F=n6RomC0Ck0DL3NO3_Gs2+wX2sCUtmNWR+Ii_{G{DM;vw2S~WB6`0X}GzNZG-9u*68CiXxS7eOd&-N_msYFmwSZQWmlS2jj#oQOIQTlE}0hBrq z%Pd^B$^wTchtGI8xwcFX(79)Rk4&=k)%3*Va_t{vb|=d=4s~jAUdAn4+%6bd{oX*0 zWD{(dOTDw*Pdv~TBc z5WRYXQ1fkX6+bnn@tbj<&Ql56+G~~!f2RTvGY(d+ikp>RrD!W7?v! zf=%~?&a>e z7=_p~Tt&W=Z~dNq4YwRq-bl8w9F2AIzQ4O;v+!vyv}=nS$Q>jM6PK|P(ZG@Z$amM9 z9wjPyq)s>oG1HHbB)rbR4%&*;<>tSIU7nl8DDW(swMUfKLnxHh_WBE+^0tz^%}Z4i z82v@Vocx}RBZRXhK~ubyvCR&4uc`5{J97&InLF1MpI8$Wn4M~V`oocw?(S^PV&#rs z?p?=ctb+I5pBHj(R^P%Tlq5nU6eMF{i_&dW@RTgsEbBwaxs?_}k>4Gaob}8#q6g=7 z>>gsGSLi*d%!;~u-=lS@`Q&(lFAjPXzq|JmD}b3=DaAW3$e9lZCGQl9l4@R8@|Ubz zC-iP2Pi59cdHQdj6MR!t?@KSt*${Vf!3lQ*`Uu= zvvpVGgJV1Fe`I4Pd^R`>Te8z{5( z$laclA=h>etOn{qTcMr;v7Zdy4jpzp3FIB0xT*8+gGqjFCKj`|b1}W)@+%!AD~Tu~ ziQwkHiXpr@g(J=P^ zC7Dkmo$lsX7ua4!rPDW=q@~lJUB~dMZELc|T%tN5L~Z(tMzDpFKk5zVdTCZBA!ACH zrIoG*mW-wZtKB#%5I5SNGs+&zW0t6JDsr76*~!!zWD`J?!woow05gX zdtr1-vil@dtg)9W(cGcOHp<_Nra2l<+cnxwd% zS*mU37|}z6Ih@hp~hsA|KKJacRy>&&OHmuVGAvubX!R;h_|zKmKcQIvG_CmT>^*ZE@<Zgs0!c$rj$&5CPHzUF;<1X4zDHM#Y-3Tve zu&JR8t+c>&;gw&j`3dnWFYy+m(jB5YpgJCGma=CrW0_nYFJC zxRB2jB9dOR$+6Ksbp!VTS*%w!?Km#ljp_^IX_XdG{`gcZ3p(cUT@$CA&hq`6#P_rSGezIeG7wO%IXqiE!CN z(vfN#RFNV%l1x)<&yUlT@Ha2>8gl8WbJY3drdU{bYMTwamXVJXH8#BT#9V;z>=J%+(>2PWLHqHkh{L%S^JGjvM`(=r`|!X%a%ulS}3cQaCb; zP(5QXd7DZ^vZ}bbsiOL#jFKyzz1Aj$Z`o%}vP=@eC$ z@jvUnp2)}|d7M9+qw-$%O5EbYJF*W~60toC(v{S;?l!w*El?5DDxCFpK_W5Yec8PF z$ZC_$`%1%)WJ_=<)9xP9K>gk4A-yt%%e_Hvr&AL;UR5e+`FMW$AmV3`d%_4*Jji=z z0?4q-$3#OKFJ4=?+Bm`55L%{Ce!kH70{Pqs_CtK|~EwXfli*6Bc z$tqVd&iaIraCv-_F2{8I{$bI$$W~vZ&E|44pMZ&TSw00_9(l$SvA5SPCh^9LHk7!K z6SOYhl{0B9Od$GE_T@U>cZxgj#_dyIPEO^s*WPbfeueKskP{WH;Fs*y&zt12*hEom z=3oAJQRqWNH>O^De^HpZ@AEBU2j!)b(&g96tapb-zTjh?VvZXzE4*#xmhPYlPQ%lW zRxB~qd_?`;g|A`0theB*De;1Iew^7?>;%bS`^WLCFXNZb$4*+v6yHPg%!snNeSd{P zL4zSnz1;~TuJu#q(EV!Snwx|@I*XzoWgfd1BY6r^Y<{Wg8~&K_P6x3jwK}t?^MJUUGa@UqxHX z(1+CW$=%{y(Akn^6qCYJ_@tzzKB`ygVkZ&4Bj%b=vNjdC$S1_K^i&K*yu^=|W zH0SCtZqQ1R!0e^zJl&uj(z}$Y5~MF3?MvzChd!hBq}1o|4k0U+#-sKq_GEZIMoqGC zKP#L*cA01B_ML&gGuvmFhkPUrNbd0j&^Rn>Sx;0W4q{^Jh;4-^UXJncc$jP z!G%DB3pYr*uSSv6F^(~f>7V)fK=Wh1qyFwkXRmDknr89gD|Xu_jJULxI(!vEo$IsI z0y$LaEHMUm`i9RD+dRSEJ+KSX#(rD819aWgpzA*DT!fz&)B0DN2U2;*nyWA`!(xNnXF!R(mlCe}~qQ1o!gim_mav@W}_)g#BX|?hWvPYa8 z@4}{&`X)Pes>km?Zo2l4ps) z?7(Ef#1u}5@yBB5!p1ipM=0id$-?^gpa=wn5BlOtKo@@>9Pmj&LQUj?jG`o)t)ZK# z9owJ55A0v2f&I(i=tb6FM=!D(LykTEdHy1N1ju^$19XQXK*5|u=mxD`>$u_60ajRz zq5rDD{%;ru(8L?IHc!>{l!n;}Zhm~0)1z;EHmFC<^j?%CJxyDjaQ|qO^;i~bDEXkG z4$u0S`)PVwnJX5C_Ur34b3VySx8r-A#|!FYPn*`IbR@M0J;<^_3@eOSC&slk^4v0( z?{4+ORIHwHFr69ltkAtbGuO9j!_Jpk*TTCsA6G2Qv=le#r>}u+i#}@6mGoxO&tO;Y z3@t88*SQmSk&BBbX%;_k8|I?Y)YWU5+?5T8Dl61CjNw3TBgDQDa;K}r8qYELzSa^S zVngq(CXKwPf%4H^``?%{k<9`2v;M#DbG?qy({5xP3og9zFWL#vr0(q zL*5zl1)B57w8D<)Wntw(DS|yY)+x*j{DGy^nrA*x^}KdU7MZS}eE#FhgvDk3sby8B zLK)WRN^AFlQOL1k?Ztl8ZVR$^GHn0m2~FFAm#DOgs5Rh7_~aQLyecF6|KCJN6dV0@kyYX zmf8yDmxjp#c70j?ObK6>p72ir#qxbov0oHaz*NG*KxA5k4%diaU+9k5`|c6Mcmj!) zWa?`uKCxcrFD%$@7*`FbBBY#Ma0^y_nVi*}K$|S7VG)lf6h&r}_7cgmst;?-$d1&I zLEY0x{IuEft&BqG)bYZb3-+J4&(GAJPOf5cAIlJtPMkZ-U+eh- zv%p@;#oNP6u(Hzk{1jeot3ss429J@CmfxLK1vBW26m~rXs_y?L;6^DW9U4Ruy7_0KF@xQX>os&{fQ)OQHKakWKRy6?&~T4VFki)S(V z^N)$e?C_jt=NI=UI7mFB8(X?qfxxl_zaZDg{!?T)IN3XxI=NVyI;*-Vb;^Njjbs#* zrE}HR6j7NqU~~G}8s?2oZ?@83QyG zHphVI{{7%0tcXB9{uSrHubo5~0dk9Je)fu9-vkTn1IYYHNYGLjOi zYOD&9{~;hC3?ImZ;NFSf4}_QuWEzl}Kr;NVsRzFScGmCvhS2YC{62{Gs4wFF`5pp< z${hNOzf&K4+OtsKjydGp&Oh%YIHp+|1fmhx zKy?rnvSbGyQj9Dk-AB->JFVO@WV-JJzQc7H z_0YVQi3$PXswM&g$sbtoGsB-7Y(zWQS_tytIHd|x<;mcC-QX9rsUPye;_NdubQOG1 z@UlBj0P&%n$XXm08&;#gFAX~g>vW0E&JIN6 zc`*8*glaOaBrGhX$r3iEwx)K!rwR|kN#$~TMg!wa!}w&M94y>VpM=w}-g|)*pMkuv z`7u=q7712fp2JO4!TJhk9H`JAWD9C2d54u@VfUqoKc^}klo&DjQN%g`>;r(%QMjW9 z3;0vwi-$Lj?pt3pBG8>SFbA}S>9t^yeis%Hi>R0yf;E@$8rWw^+b9P1!mug)8gu^< zI>X;y=l_5`ysh+N7+jd3treiHbbk=R&kX;qTStPb*;|{!Nt5v>t+`$Ri%SArE+|Z? zV7QnHhBjvQPPV2dKV1`g6Zb(=;m2TzoB#!1_xXdGxEczV0D`OOWcl}K=YycC*yRi_yciIXjtq9SgiK9suP4^Bo>SSqa>0;>y$2A`QExW->C=j%+kh6aGeS{abTZ%eNcR?gbf|f&*DAfEr37r1)hvO4D zND*2Fh(Is$_sQ_*czFcLuc8e0qE#9(S0^XXH~fs3!;NwO?M;$tz=#N=yW5RNFz!>> zzK=A#hm$N=+Mj}y^GxUmGigop5hOonACJhS?iN1Mf&jJ?qUOJpw}K>~d(oSdnL zswtcj=9xxN!ZkoZ3J5MkDe5&C4js&rf1~>sfwZBsg_@J8se+*cjf9W8n5C(ohSBqYsjv#^<`sX$aU=K4(^TSQSF%lpMIW{sB2mwm?zfXq$`;8;8VTBNJ zvylOPAN-ovqhqKIavpye2n=W`@V1U+`lBj1HDE+KMH8gdPXpB?$6pyBu|WInC~QMd zOB;`$V@4B2S6d@^gOo9XCj8G}B&WmkYa06xLxw;9?osfTcGjjQ`z(F937#+Ezv2Rl zfn+=AY;ltS5xyTm`k!9~oy_m~8WCe-Q)f6)Ev_l_P!1480b7gwX+QXx;g5Re2$DY? z@l_{N$o;LCRDMk`ffp~o$lwidSP2|1FM^&~5)%R83KZS5v;PO3x}CG3nd$#amcM$r zsTs)T06YK;X#YML{$I%d53>K4Mq~2Pv?`!D0g@whV#Ggx1PydT`4j7hyTSt;35jta z$sVAAc7+$T8^ zAfY@*kibfEbg0}J^+Tfta$E(a8y%EE`1p^Y`CXA7ZiL)7iO5qxKWG3KOt}4hGW^qo zj=eZCG5k(8?=?-@0+4_Z zMuIlUBe3^1d*4SI4)(+Klp(xk ze{BHCAU8n|1{Zx_VM_zLba2TSC`TaojX`Q%8CwT9vwJU+8-^x<#t?uk&?y&RaJG#qdWXae3$?L{C60?-e1y*BT41k3)IlQOjf@n$Rn zIbQGRkW0{}!H3NGC&2WjY~U1oxW<5u1~|e{TibnI0{CZPLR!xM2>g9|+V_#>(&7I4 z#U{r54Il^w{SWl$FZ|XKB(Oq&ffLkb_=ACM?KhN?35b|sh+c#oLA2jAnqaTJv7ys3 znGiG7O6>yh0X5*v&>eT#og;`~B|6%n z^XBPs#QQj$;d)$>GS`ECU>p&c;UP}ZKM`sgLkcUA) zQ3Svi(27c{IhJC7pm4!iJ+WJBdi51(e;!y*B>LBhAvNlam&XzOnQw60FUfD;g(pz3 zX7+|QHb|P_Wj1u=r@T6j4-z(z^zv)6=18x4ps-}HphE;oVJq+cfk4F-+}8Qa1vH%u9l$EM!{xbbc3u_o=2ji(3|gMDk>ltN zy;|j9I>Bo;!BYeB7=T_Gnji#pZbwH7l)^aS3>gyKv60q+3%`YN;S2NsM8E;Zg+Do* zLec>QSun}Oe)-=B;JjS%NiW1218Lq8yzPdLbJDfr2!1Zfl2KLS<6-Ae2g|eI)ZZV5 zO<6gCnH7MUg`vDk0Rib>tRP|MYP-)YhdWg&_9tHk5G0R)-JwhI4`|1+sXE#m@GE$s zuHF#K?*U}00IHLQl1T^sIF=(*3)IEK2mRTH!5Sap&v7vR%#V8<>w%i(3bRo zyOqPZ+ZgC^{vlwv;8c7ll&0R07fL=L!$3#m9k3GU9|+*hdiZUeocadLPY)te49YNX zL3efx0d#ryvp5H3EM%`C{HX~+`XxyVpt1%Sm9bGCO9<7A!5Pgz-%Q7i1rm7xwLzXRmP!g1~9E*Rz=-^6ZbDN)?{*jfc5`KNp1Dj}zS&&{vQ+~b=-D=H z-s31>3rskZJr#9x36PamgTN=C-Fo5DaU@W3z?;FvnhiGR0TRf-1SYfKzY)NRn}=oN zIFJ(g4)8o^OB4$qNAQ1kGtZ~eLs5VXs4y~6i5*7)r6evmR*>+h(t8FX2D~9XAcMQ) zu>=Pq2ToBrlb3zwIZ#t4Ox5Zkbu0x`{(>{nEs(iqI1H3D2~&jm%Kj4py!liNSCQOm z06qieRH0opPw}4!IN?lFZRAfbKL#p-OsGQ>ywLnN0yt%)zy<423NWpGm+D{%mIoF` z9NmW;m8}j>Q4S6zry=iPVT)FJi{sd!yp{t_*^S5T#iarWY(Q8*_b2Vv|3m<%N)cG` zB|8D|8({h?Tl;^)hZ7PNs^VE{VERUi{NOmN$>}(Juoh*Xj{b~Dk1QL{i9M;T1JUOM z#=+3;rKIPl*H;20#g>Z+F^ASi_~ICAJe5x|L$ zdA6ID^Z-8(;DhBOf1eEh(um{m|IVfVwgJ3eH6^}=0OGU>Ffy=#^&UsZAe0s07=e2I zP0;|rHvvVE@LvfavmL$B$Ki|F+d#G?fj8g#n2xM}UQ6dBL$-&7!I&I1?l`u69KfgG z#0bsjP7*TUw9i4>f}ZGZNIZ_<&=@&9S7C~e;gx_;?4djOmXtf~IGX)(9FD&Ilu<rC-+9Q!4p;e=D?h*!*bPyqqJ;ot|d zZ6oN{k9OQYy2HTG#fVK;~log^G zL-PqJP7nkLbU!=Vc^m_@2#32aT{Z3kBqc(o%AxVkzxgM8IG!6xa#DZ_;7h=mpyAy= z;d8<%JvC}c^BVzyJWT1y4VL8}9U9OIf>WQXL|(UK1Oy5&8j=AELjQpPjt?!}Al)km z7KWriXdh~rJP!Y0$9K4maXyCTaf9@v2ht$)GL-iDBNz@Y8Gut5^y~yvDgpF4AOrMJ zRdMMEbg1`*$1d$1n0gVH3lu~R=1A$Gde%>1&IGOz{wzR0TrZxwh77kr%lTkhKJons zP*w5Rl>P1+qp}!_*0K zFv#JpE?hsGr3chP0J?tYg~GF7Fbr4U&_;#72-k7ybeCt>Yoez;UG<+^Q(%12J!&QL5GYAE-cdiUCn@DS*xZCH@rN(dZI49PFLppfY?r-3*qMA%H1KgbPrpnx~IOg>*$n zcBHM zEUmJG|KkZfvoXcp=(S9S?Bu z){_6wEWw(KgXaIcxdz@F=z|B){s*^m6&CKN{~tV#_Mp8Fo@ePfw>PmA`3p^f4w~fkwqV-Ifo*e0~_r-dF1dz zv;MR-!{7N|V1FN>^`8z;{ya$R;K4)xq2B@1U LEN=v=6@>o*Edn-* diff --git a/merlion/resources/gson-2.8.9.jar b/merlion/resources/gson-2.8.9.jar new file mode 100644 index 0000000000000000000000000000000000000000..3351867c1c8f342d1a6b9d78b14c1e924ae64b47 GIT binary patch literal 258075 zcma&OWmIJ8vMov#?(XjH?(XhR0g1aq;ZEW1?(PnSL*ef3?(T4_ckg%3zJ2@jc{wuw zMZ{Vo){GgkzA+;eq=7+DfFK|sfILifm4W`_g8~EsBrB>SNGmBPMlUNUCn+YXtU@O% z_Avqk6r`ePyTX9r9i8xruy61+lp_#r#Gk0r*o?}*ib{Jk(y>4fm%APp>h->zBq5>7 zaqNU09 z0r86BIp_%itRK9ebvb8W{J$K(BKPbAcKC03czF@Ev_oe!Dbf&->8gtBu@|q^>Z!t( zGVh~*js}d0Cu$?wn8C?niyy&xC+i4vt$v{{XESveZH(iT!1i!#D@-9K%GZ1Cbi2mG z4^^8rK1(BXg6gVIJQ1Hyd^fz<74pt)Y9 zNiqrj#v9;oVTHMm6F*s*Qts8MM&Xf>QrM;TQ+d6-Ou6Eyh_#HgEB`WEDpJ>R;bWS=R4Gm6 z9TR!y&|T6i;ciHS?8{0emcZ#T*W$Bzbx{JzVX*8KT4Ko}ot%o7vC2|J%^i(bnc4tkM6*+R@3u*x;Y;1>ru=)o#!hdr6->-`IPws~1Hb(z(>AyY|;{Vmx*~$E$?j84^ z0yx_LlPk{OxEk2l*g6?FncLbp{>NSamG|I(d;S;XtPNa@|MU6()tKb}HU5o5x<64y z2e36Vruz$m{u+Vu|7V2%3iBsh{}r}>g!!Wa{*CW{S&Nw)TN%k2Sex6JD%e_?1HO~6 z(gCat933N-b!}%EPx@N(grJ1tN5B$=f>at@>F?_Z*n>$x zud97S{M@#{uksPt&4Y?qkt6qXUSHN6qc0nMKA)eVy8#WfDJ{kZn9Z_lwcWK!J-Tc4 zHDRwfmN@VjBa^)pGTF>?*c;r}iX2iRezWqs3{Yz}w?BH0j{J-?u6iXN#QFf3U8sX$ zJgoAsLb#zEV++W|v01^80qc=On63#DNRp$nlZ(RSQ(Ou&kVv?kA-O66Sr|%b>{>w| zhmquTeV$n16pBQN4Wd@uhoS>FB9i>LA#LO4tuSYS2G27HVCB`~we;|PG8SG^eKu&4 z;=DgR3y~eVl8#I9mP+&k0JFc)@RZ_kv0+&_o2(nIegrJE+hHu2a**_`_$cCdz}P7A zn@|8&$J3geBw;b!2xjf=(XoSAe`XCm5*492K@@Tk5gyHcWt`0tYO}b&F3fefc@?sH z%XyqRefVaqezbGZ^}~?{MqT;*Sl%hufmJ*5xTbfFnWdnkOB}vW~ZtZJPK?;P1RJwzs>nzm_5d7^jsQ!uqDg->gD z{2tFeqf2l6xQ%}7gLF&k?x_Zm<(?F}+1$_|^OiU2Dc}|T55h%9=YhEX0tEC7{3qf5 zJ&yjaZp43c;Qt4mL{)2LWOWp8n>qtRnlvau#e9e*GLjk<#7aS}0*c%(S~+I zLB`1!HV9{<w@er-B*%3J8=i;EK_%ZpmkP1>f;2BF+2{`o2sNMsnieU+Ejd|VWJIY|i} zWx*&iqv@zOk#3@=zW>u2>t^r9|HPUcv&@}U*zpA;EgzK@fC*;MJywcBQi*wDX74T5 zNS271sTNf*UXf>tBE9Qqt7r$6Hcn6oBisn)G|;4)hPR5i&*f~54k5LbRLJRMXGo5; zTB~BLhgWos!d@45tZA$m8y;0OsyRD^Mk~#VqiH|(otFBs-N(Etz?gEB&aFsM8qq+P z^+wLG#}*AYc)_cm$gIFJHpdxYk=dH}XaWvic~bJkUD9Ej;hu36 z#d0!=APYU7Z;^+am5JK814QyeHj*~{jYoCr)J^aX3hmdls)U^m@R{al1yJ&_yhPP% zbDcn}tijARP}QNF*m+rP<{RQNWd=tg6-f5jWP~0{$YMo=T*;Jyj6e=2$Xf=I6pffG ztc`>#&I61S_U#tJ);b~@?+SvRVB&3_5fVoz4s)qTE*tlXeX36vyXFgR*lhn-A1D+aMQXJ_n2c?yK+dMyT-$d<_+;Qj=YjBSOzw6qEQ!JnlkD$MqpY`sm_7Ytia%4ox?!?}TKFk#C!lfwg zw8MhrCIJtJPdc?NlrSwh9zus+b;&=+_`n`|#WfCE$02_HQ#!r7UQR3jXlhY`xx;@8 z!l1m6%qzeI-4oLJltZO=yF8S+-vE^mzvzPZGk{G`vSJ59coXZSn3&qEh33iQ>$boi zog6FVwg3^1hUij=GfR?PLm-1mAoC?5n|`#yxEf*YV81sk?HRdKP4@Ia zR!;i6KZ*7rfk&9)1GQItmJpz;wG5^;Z4PM_x?lvMcKhWje2QIVnzC1PM>`USZX0$d zg9LY{Dk{J!C|Q!yZ((FnyXoYZn=23F?o#twhB*^Zb| zDPTsF;7{ejDC-XEwea#b#;qAdb}e)zg`=MWIPxJ0ER?AcD){3;@LR&DlN2cQ0n)&d z6Of0WPeZFG`>5TN4@y?WON20ZEj#)RtdlY&A-lbks>fdDnu&?7Ct^{t`<(45k)|(- z)aAORU*&1h75C+#C$xmsEo-8FwM$S}$@6{plJjHQ$qTHdz+(G05u%{znQEr6|cRe<$eUccc056uSLo#{QpD zsQjlBI3fN(i7 z9d<|NQ{uH_mfpK26DpHwR^~Kh;?C#p8%HMJqe&2u;tIUSmf3W_t6Ls&r9aE;O z)Uf|9N*Ix535QGFPHQg>DY?WZ;=xZldUFVr%(uGWs=K&WL$dH9T?v!JB#{kbdbi~| z+RL!uEH_~8!`39-78bkYWbw@Jxw*{N%S6hwvKK#*+?KVa^om z@g|z0;)8i$fcQy<<8TwweEsHM;3ljjn@JOr^(|X1q>R-%<~39Gm_M%17Fyw?LCg0l zAPUoKS0}q!hJ7eZQLSdn+fZcA(Mqyac*|389s@1iqdcB7uRDY?#U@JIzB8H(haD}a zK07LrUjO`}H~F16+8{$rhwgBnDR28J>mK))WPh*5S8AR*WFHLtEftL9>!9GHBx+1@ z^>O-_81}xRd<-#2n&D1YRq%3;%RObYv4YQA#pSX)dal?}lRQ+^l#)Y{ebnRp{Ozk% zY1#;;ES7$%MBG&t#NhBSxEGCqu;BX7awiV1G?ydhRc3W;nT%h`Y=C$yKryRwjqJ`{ z5@6P)P>YVciu|D7SW{dLIE8_3hnUQ)I<-f$wjHN=;8Je>p+X6O)z+M%9A0(Xz*iIz zb~V%6l;SNH^;BUMqkBM`PG(h~QUTEKYAS6lHz4`IKe*6yq?k6^DoW2$jj!$?_^DV= z6^XyLFr;nhR6qmLPOxAE!q%PckQ9W2+y1srnNOH8qT}Pl%gx1`IwHfGej+>f@N07i zgl`WSZ3o%?EA3I7?r2D(G#kYu!Q}Fb4)_yw5RkENR6kflYL9b6MgF}wUIRhTyuui@ zr%ttD%$#2v!^9R%PKl$0QliTD4fWQe@(9ue#GDnjQnn`~_Z`zCKh}m@8W!OVX$a%<7V+5@Q#PXuz17}Ac$Yks$(gH&{JlNQPfN5gohPwTUn|9+~X0mOBljsV`_T% zlXv}%sJb@$v?glOKJ*NTUgXJs;f3%sKIiLN$WbL@{D&}CM=&t|_SPc|C$0sq=e?%` zaEQr7#jahXN_Iii0#iWs#Dbe@@i?Dh!Yd`tws+uBcYU};_3PTCqkgJoC=RyxwnE$3 z8h@^e=rdN~nc^>#AF8uLQ3X|HV=%04Lsng4)lpM5Y}bT+vsyoH#i;Y_ zL}8%yxuGw^gjTYn>OI1&Iv7{Xq_iWEsst}t4k9>e9;e*s=r}B_Fp3r{wr{_Oundv6 z0@kgbJe(v46|>EXu#ws*3)I8*=O3g_*R4DaCv7-TJFS1<%}3NZ)mDsoTCs0?5Ki; zo*=y3XdXH%gWevtPem)iL!3>>rREAU_EN%I^lpH%puYFaR7B+aFnu;6iW0vOWk5rM zs2bK?VQ|{8Ujm2jtrtq#U%{?ZTl@fpDY_94+e^5m^rU_8~UHv}HWonubq>U#1R z{y+(@gfEh$1vrGT&08geQDNyA&~|Ce3cwi%hHLag42WAYGpP#%-QVStjAGhhOfNt> ze4@<~+-~}r!BnvxQ*Xw<5-P*Zrd(?}Q0OO-Ook*vhmaaN?!A&9^0cNGgPL`A$chSZ zi}=9P%r*M%Xdg@QB zc;{Y2k$whA2!e1h5Md@i2_{I*?i@C_VGU=e zrH1*o%U2s+E$d2c)ctJ7&2)C#B$}u5&y+RSP2CCaNzTU&bo}R!XrHgXsWaPbYP~M2 zOY3LRE=?ip#PuH5R#>m`V-GX0UTiM>9=>7V=lX^|O?~ZF*<*Nh0lg2)HTwps?*nkh zC!^m=-2%hJOP|um@al!(JTmmMa&{!CvhsG!sj_l+G^w)kcfM2UE-PhQ|aaI zs8Z?W?>MJ+7=#cbkbA^;B=#27H=m#3KG4l?G}uj-?{hKzlf z4<^Z5F(m8p6=wM?=Va)8$kJyoFZ(RDvz$6ob;c|w&%L3Jzi63!7AD)49BNZNC|~2E zZjPP-@o^4iOvft|DjMYV$^->Lr9x4ovSByhqLfjxWsS_1NE8zVgvnsLmUL(`O*ZvG z@`xX_x^RRZ4)binS@VE{_GYr=2^{R6O-&5uZ#;omWKrA-pG^%wm%b{b%GdQ_-S0D` z$~Ota(>L_#cypjC8+iwgg`Xi9zSe9hSFVXp8d6D})kG!F^qVu$*Y}m00eiOL${jM7 zq)^@xpVx%yv0bMIepiFC`rYLDz|R9G6Q32C_*dW#5zK;;M?JG}n`*pR%=msc>G)o` zJ?Ih*id%s47LlGG@DmZ1f~0bp(M8jZMV>LEtcKw}tBcE5emK^im)90nys|yh_C@jb>G>N;;hZNIs;R=YG8Nl&LH6>~az}$#_4*eNyV6sR1s~6G>x_%5^s@mzpTnq0wOdT;TY~;nkftu%>f}>i& z^16!V^p=hGjE=fA14h743HA8)+M>KBlfw4#Stae&eggg(WzBDk3a5OXl~sAU)}`&J z?QqlNtS!`yn)8kKsN>e7)E3%n9KC`9W=`75(g48wPMC?&Ud_sbzPzS7!_xdY_2~&T z*oaW)gv~dPmWaSv9Gz3>F@V11jVtI+5+*QZv24c8Mj&!jl2{zb+_nQZk|LK#*N*0j zRqP`wOQ?`HKe`M9n@1Pd{Igq_%<*zzK3j2Xgm$}r^I?yJ^6_^AA6L!HX4j7;2^ow= zjPPN|B0=F}elYXxp*tztSBnHucN#sMx}CvXk}tdgYc5R93O*hj7-bP^m}3$t=kQxU zeVf}Wn2u3E4m1LGSwQCXEY^nfudjU7o`G?V2SpE6K)&|Wu%E$$!6aDt4(V};&dWnn z_7TKt`Fz7@KkR2Em7;`HiSwIp&h2RL3P`fd+dk=UYZT-lET6qRy|{AonChMy`)TkE zoypy;ckQI#fP=e_9?h-28i^J*RKVtB=N`LJJl_PFRSOxKDsR4l@7vJj`9t=}?H%+g zeRpWbdM7M|XQ?e9et}wKynA_H7(Zz5yZ@I*ux}o}d7*4o++A?F0`Ad?{X6$G&8)j{ zr=1%Bn68|`p$BZ@AkqMkO!!mSb% zq8c|WEMg1lIJZ>X`5HV%pY&X9>>WDhLwqZ3?2QS4e~VUm^Eo#900(0d3wi%^T8K7^ zd+@zX5G_0HDJY|u4v%%V0G^_PlV>8*B&bt=o}z3f#%nM%1^D=S9y$1A<6Bc=$pit1 zhVgTO6XL{zDBsisU4+lfM(Oi^iKa>F%}WdgGF%+3Pz;8tXG)L@Bjn2Ck%?ScG^Dt8 z?FT`>ixpKRF}!t$V#bgz!=RDWVeVVrH=E0TBKE@y9+lBi%^dKbqOEDaFkpcn5=qqSN>e!m#r|=u4rm8`$=8~g6hxWkxXwP7mhS6 zxt13}X0F}NKDAxU0~Vhi_4wNj~# zEtF8F^ZN$^Skn<)QQhhx?A_yzLIOxtQk9e+v!oCJ!OF27LH$^RooVHX5rT!O!LA__ zGC=k7wIg|qKKYp~^_$<)0ednwO^qIj|1DZXYGEv`>h6l)gZlpIgL7`jk;@Y2PTdcU z(P;hhh0ey04KQ1LuCY0ZMy@4%0qzwZwF@5vV{pRUV^J`s{?y?TD`T>;AFwWYv|qan zIK<+k@R5ff>|)rY#tHFBeMLj3kaMTpT$`)G*HMG!I~6?6dV<(u1!a-!`be$wb8BNS zOD?U5jGp_x2yQw^4bh)DtYEmD_uoM=W|@@v>obs+BE)5k(r{2b!~sp+oxxkka?uVD z3s)VT--B{dY@)Sl+T761(VA#=>d^+2@jzIF3>p_B$DyW0Aj4z7^d7Z&1ll>9U|^|5 zO_yyqkaVCZp|rWE3vDomXb>0Tg0MoJoEUeD&` zU^B7|`l+RhfmKHK7^BSq8TC&tMo|0fi|o!6t3(l%)h$S`-uBXO0h5xswG1MqRXp2o`cW((LNU?K~OiM#<$?v`rs~93syvEMfOb zL}%&dIDI1Y0w59XkYG)$m?|+4Kd+`D4Ac2%j;4dy>Mni(sUXDeqUx1k;X-C|JwD!* z6(@_6-umUQ@>M?8ZU378-p(D6i}x)=0|qRFI+vHJL|%tA6ar34gzX*9cgB4eRB3Kv z?S5WN368b9^th;@t{w96m2VegdGk<=bGyjsS;O&XfWAlx>9CjuhAFpytotvib%~rv zG;QXC()?4CrHX5C1B5-zK0)Vc<{>&Mtm}%~)%He5fQNDo|J+6S!z^ZgqzlWenSXD! z30kVH`V;G$R_eFntJ7y#2>NBTAHNO^vIS5ctRa$Ba|>1SLp}PSK54rY57;R^!9O#5 zeWF3ryUCOp5!H2P(iuTJgFg++lAWOSO4XCyUe$IFWc_Rnx_ZaP@)%MzQB4BgTb|%< zydfcXd=$1kIC5%Go&*e_P)k(`6l3K?!BA6G_Ud?oIJc;SjBBle_7!3$NrE^_DfD+y z*s>x4F{t~3IH*(GIb*hJPqbtCSK80kOkKL$+|RgBn**0BPr7qF0s{o7!YXIK{Be7z zP`%lWEzK2Ery&D-KZhRmyV)|jLz!8E#sRmO1>&m%JxRT2HP1|4;@gi%XZ(zlcK&i4 zrwS?Qs>p2}DZu#s*FSobrhdLcY`&}fjvElD$%@GmKW`lVrmNcN;saln{h5k7@Hn6I z(&kARbO{iJ^7%o)BQx-=D42K@jGx?Ayqp)J@KdydU#ULTB+ADO#*(=TZ?I1X zJ*O-^^&9F(SuaOCMg;aYMJiZVFm!rDV)~txQE=}!{7%=R-pwsrsyEsXvH(fQ>ms7L zdHD@Fcuts`9WTv5qz^WqkRH*Xk4~Wk#oT5)Ufg}Ovcch?owKH?RrRG)Xw(^49|L?{ z5|@|NAeLfOHgM>GQ+vyu%PZ8GOpR@-O^Zu+9|jerrOiWXmMb1>7}r=)WoQr2?fIT4 zbCJb%sP_1E(K>he?Mrz2NaB#w6j8^|sw?=XkUrSd%5Kvu{2-rxez`oRfHd>D4-x6L=^pwEy*>Nm$;Ci6RFyda;RE|n+s{>{ZV z01X8V3an*!8`+ez>kDIdq_)fnO&aHUiMhBut<~cU5~a>LtW(a4$#>84n%yO()Jydz*Ef$K{1^d)a!vMtxB$*Vg(WT*u-9iL-rjQILRa^46z(936owcMH0+HdQH-FqUF*gA=V&^QWfd= z8O%NS=KEhAwpPLhu`y{>Yd1ooBMPw3$f6?&IQ`uNErr?cGnnE_#+1-Kw~q{tIL_+g zcP7kRK@g0l4U`sX63PyUU)nVcl541qDCJ{sNQ-GuDP}mBUN|_#O>}#!)R*X&K~?(3 zcAL+{M6u{)V+upVoPMYbZP{(ZarvM3&ZA`QsNzbHUjVbAHHByTRs*mCW&%d`?E^Ao zOl7fw4K9OAHVT9bCBbGQDLE{>5{q&;FTVlq{V)z&%4EGc%`DIz42HeWGu7=OLiU@#FK+i0D!(ggHQU&zPDWZC_a z$;Lajv+Q+xmJcVTCi@PXm8Um*NG3lX-(oB*6zxHIuG~U$htZE>u&$-N|G=`y(oDu3 zDv7~QA7=&)mbWZ}@0B5f&A|#L(gYPu-)EJljao=t^zD_(V#n?#Qo*9M%ogz*K3QvG zC}vo9oh7x}!dz1ms|pKSYxxQft#kNVWFSWr-HJ}T7@LQ10U-T004w)NSfy!4giM-;SXZY3>$DQtVHLEtmGZ3lA z6!zW|0M6WG+wR@XZpeSetz$@)&p(BH)_smCs>1ptLlqXyE9oE0{JaQvG7$e%g_Hfk z7rmJ8y%3-`dvRIb6DiUykFTlmbEg*!oxz0+Ctb8Fn@ZN3h(E;<7Sg}q2DKoZLRLKv9D6Q# z9F|N}wzxugye9z0F-Kj&hd=xY`89t@Kys7%;@#X#-DrBPEv>S-%wG4ZmmdF;nTu+AI^S0GLu(q&jO`EtdvmMz4y$D5P; zkBoFV$msg(~dp_*XX0S-RV=MrFRAae7qWRe#ZM(a=<3OkQ<=BLPH^4ec;_&!k)WVh@Cz- zj~~%3k3Lu%{5_BIyDy)pj~|yQk6*d?`X46;UeTXENIeT5H=*I$Vy9obdLDld3LhW1 zoj&m3H5WO`W}YTf+x7+7saNY`GmH#RSjDOCmJzJFyvjnGSUTkV2* z^;~c3pe814mtdOj%b*Lo`q*DzalhISCV>jPZ2&V#(ZgSaKwbd}b%7W42AnCJDXI)v zA#uyRA#Q-MVVV-D-I`>?hwE?H!wQC3!c$ZV$8Z6eYsd~-`~os+3j^=oro!~4Q9lMU z0?5l*9wY3*31m?ZvH(3-;oJgo=yE1~p|IN@0%n*<1mOB!SK+Jz$=K~Lf$QgurZNGA zv}B9`p|m+5^yOA%G<-?fb0~uEHlQOXN;`Al0^71nX0`+>I`bn&@KXHV#O#4C(}Asn zqsCF68MUr}^L?6nVyrz^WX?sGSN^>G(oq?On zBnm*s!~h~b(13SNfnIqFQ#nzw&cM&Lmty8DSa6~tF%`WIi|Rf~WKqh}aOPiTHji7W zCSD$Xzbx${7>T%I->z>CQ?X*!NT+o#Qhq;eA>7bLqh%eTmYz3n(%1&A)zo8Tp@Da< zjaZ3?U)dy9PE9<|YbR8GN;ibjT%U<2RQiB21y;U~ z#|?~XOH%`8QV^F1UiRv1LvQ#L7`-37UU1_TR7HULHi{eUgzw_Ej2rMo0Q7Wj(T0OLF{x}NSq4IIk>?z(Yc}F zkvuV*1uBa@GwVFDss%cWJr!k<$h9gWi#<1G(MTQJw366aS#(y1X3RV`P|MWcffflZ z_MGas96D73!>tmwSI;2SKsYH}eL=CDDcB>!J)M>jTgY&$Ah*jHWVm3^dnv>)#DTfH zz3q=Eo1(Cd0o50(& zrunNkO1V7=z0|DiF4$zbVVs$=K=&f@e@`-~)69$r7DqTJ11u?i%@~VX zIAZi44Fh+TA>znL>K%!fndl<5x!Ea*pN*5lL+Agl88@8KDXt)lLnF?Z7bJAhVvWM{ z&i&Ypg>diJe^e`0G#r?xNUtVqJh*k~zisDCT!q(* z`eSfDpylh2^2aSZ&1E5%3bF(_9Zi;TR%3RMY9MYt_vW5{-kK_B_&^5t2U*O(ntUIG z1^8Y{;TnTWTzioWSo0(0qB>eeNg+^ektWd%BsZ-uOUI4~k^Jz61&W#}1u3>fXzWDc z#YA>SnzLP3H{>` zZ(zW7q7^0x<(w-!Gw6$k{dFXmNwR_arU6&CZZyrrNg)WM=iq*YV2zGA+GU2@uL{9X z%S>1{Cf9xXtDs#=U&v&1C-&^{GgQ@Y5K^RF3~g5kGQ9Au9wZ!P3gLw>@HoJO+^a&| zrNFm&klDmdtSQ;K$Km}3%*B>%dkhYfwWY!{liIhFk$$ze=iJ(~_zJe3C0qtq!Zh_T zmFW)IU5M_h70(C7c#{gYj+=Wxm~hd`trFjcV=t*f(c7=M zUU1me_w?f==Y29>NCuXP{o}HEK}M5tz0y5Wy@1z z=p+SPcxKp+uG|zB1!>Avyw2oTQmB_SpXU&kJ=mbpw&(&$uO1}N?_ygj57DgvJxsf= z$urc5!cb|UW@#B+y5?$FfM97t$}MG3BY@GO;a)McG-EXkqHj3!S?9{`+&UyZje{_^5b9}M*J(+Q#_mi90err9Xz-5 z><1*$`Y$;(S?qv_&|Z$a_@3Me`a?TV4a)nnL^P4esXZ2RF^+JHG{iJ=oiFBcj@Po_ z8;rC|_FK-y%c^-R1q0@UvrRdf?*nG0t5XwFc`ililQOZG34&DFaB$uvMz@*_>65|v z&o~-GheD{?@b_;RW(Uyvp!{fRH&D6mN;14vwn`nK%$<@#_a=rCu6^%g;np&968AXG zW_n}4_yp(~RUKEUBK8e|7wA<+f@U0g&Ys|oh~ZCVTwsv@|ESjV+2eju@B@?huP66Tap=OxK`4K4?Hm8X7#MyGC=+7n<8^nNYN ztTNPTR)_C>;AjrPj0nvJQQI)-KAEcJTEc2tz(|LXYz{5}j{M5Z-);1UrLx+Hr*03~ zR(*Q)_+s`eC+}Do#X#cl{cEIjW^w+o+Vl@C_+Y81Q6&Sx1!;+8F0>hjteo1!aH~Rn zl{O0mn<>p#YaYw=*hWWz?L$nB(RurWUv)(gB3_{t>pv>i3u>EyeSFX-S5${q=7&}u z0t{&@Y8R?gMS%{dKd*{%o(P8&G<#8N03jpfYCia|M{u!rmm&`rzB1nA9E`E#Ye7^X zxUPOt?LdrGixH^6UPWVah^%1EP*iMYD*7!*qlR(iPV8?62sU)=gokp-7S7-W$>|T> zm6dX{2y4N|pvVb(-LzP$J39WKsAq48IvLS|a6TwEPn%l`A1$2_=PkZ$b;^j+cwk4f zOUFW5@lIjn$0+hM1){8Zx=0P9-P%Siq@+=A%q;BpU`^X}Uz^&cT>8#innq+xYBL?j z3Ws`dHP%@Nm88t~3D!bwUB_dOX?+mDGBE7H4Qf~hWq=)NSO~;_Md41V*q~Zzhgz4x zAA4J{zfcxCBV0rk==u~eNYi>SlwAzeqybHui%+NOgW&{mk8AhWqXzdoGArqUABGj5 zT&jh2EJ0Jupc4jk1rm8-{<>D*V)&qv*{|+_`2fQi_=JOg0zZ1?3*3DYR~XzB!GA() zxM|)7`GCv#wK*WO!@Cr*+qBanZMQc62FQ)mOHk`5f}qX_UY?X+CWFx>7hr6%&U?qQ zT1>3hRiALt0*T^aGFbMfEZl6ioxMa~G@xRF_bJh_m@-3;@L z(~1U~{bIB(r3kr^h$&XsfsxK&W!UepoixHQ7uiF}E_ZHfS>A{WNiaG&qh^V_LPM7l zM-qMytyA$$22pxRf>X%bH2 zJ|-wZY*Cb}Dk8<1FtwX3l~A|NU${;niKNeMnCwSNP+py%XazJRdnH>g9Fdj3+!9D6 zLz$>LQ&q4SN;It)&0#GqRtLJJAt5&X5?GVrU`N1JF>g(CqZ`CUdEdhIlGaX(4+0*= z^n#$|T&P^Z5BkCY8e@U7JzpJ4XflaqyOadz0i`TK-$Jg;LW%Wf2?@GANN*J`>u%*o zVK^$);ybj0yW|#Y35Bfyk+tm8A6bwAHPs|?elsEoyM!^6sBrn+m1jaAw!-q{ya9aFFk348=mGG;9#z zdWdlRNqY!23~|1a9nAT|35s3D0!^5se&|D?^a&+E@h0c-=Ny8OI1^@2WtB-K<+p>D zLX6F*)JW3}#e@-lqw)(?qLhgpQJ{4+Xmdgti;W6?R^p^dC=w200Zx))&Xwaraxc)F zq8|^2<*-4H;>I4dPCPh1whuahrkTor{)(Agcmhpv*s6ej7t|)GN5KOmPP{?1z?(T@ zg>wwr{9ZJ$#SMtRlCU&4#g2t`s?x{L|yg#&_-D-?{tvg1!LmMV5nBa zoqUW4s@^Yc!u%aw#C1|s2!xM>49j4F_SwIXq zR-jeS^kiEbu*uvrgTOE{~Kt3n9i&AlAZd7Oz0;==8CctIx5*1vpr_SkA& zh$D3Kw#98sE{XO3YQ`5lY&vjj$7?8)*G+vcVB;M!`pSku{{AI@VBD@H5KgEKb);a` znHvB1!JRCGWf&SK#EDZAaO)Sv)i9hg^Qx~!Z=y^*+QNdu5U;LfHP1?cMuUj#DQE=X z^ka;u`HW}8CTbGU6v? zp}P*ZSN9x~=26!fKf<)89JL;jR#}FcI$E5R4Y0<7lKbl+_0T5SQ1E#d^BxNh_lG8Q zIMjG7?V8K9K`B*Yua|PUi$Ry$YrC*q3;`IVGpE9V= z1M&;Sw=j+sInY+it~H5y!DMH4r-*pJSkJ_7C1wRT&M+StWCuYP9B#4C+}gF?1z?ld>9EyPh?d*~L`B{Ry)ET-?{{Nc-}J;M%FI;20t% z+un3NqPE^@3I9{-H1pim1USOG(Dtur2fv&{wX^n0?;Ldz7 zp4p6%Ll57deGX<%c2K(o>&VR*fzI(~<-%>rM^u|s-f*nMf{bX4MH&=bJTzHaSdTH4 zpOP}A3_B|31dFNJoz}9iH?CL7QIE*@^@UB#j@%RIIcq8q%>)TSpcxow9ibck5y7vo z+A;Z$Q{8NSFPBc|-^(7(rn4;P2LF_2M`R4mIsM)#bJqGN?au$Q2=!ZHY~W;U`u$HS zk9A}%R4)TY;AM5AqM)6V>$D(n)jD@5oR_}_z)TWZB7du#^9hJA;b~D6{B!$hyR+S+ z^8y}7I|4MY`v*g$VCzI#$z$1k;Jw3QNqL+aCvgCuwyvn=+)CEXe6;c}1@3rcvcj%F zDfGtut2bZ=L=PFe(pKvi1OvrqHLT0X`5C8;VSWmO2VD*v?5qihK#+GuDGY*hei%(; zBildpN)58R&6&T|NRz*R|E;I~t6u4UT-ss&w~AU-TX|Li;REjXapD&$6%oa~vVEfM zq8w3_*z$d%{TrN-S`eWbNo5e9p!_K2lcW5}EnqG_j!w?qXot0UBPik2ubIQk<5}yr zUF-Ga@2{7K^gv!3iHf1zOl`Tr(*b(9_o0Q~egR#Kw1fy^3b)WVfHP4cBs0+L2}2Si z;V}jVJ1e`2VSKH9Kn>nXX`H|;Luwm7bXc`q)01w;5bT zPG5;C!b(e3Ux)QJ<4QnO)N`kAqbnPt7%wS?rrDnU7&|AV{W$SV9&~MUfZ{> zA$dFhR4aHV1e!8h!GhC#{|qI2O4PM)C!#$FUv{!3m8qh_rpUCl-k?$%L-h&-09A$% zd!^FDK}I4&mD1bbk12w}BP`Tg;m$Qh*5o^(;B7FOOYec5=bZj4{PhB+k0v!;X zdvhehQ@;oI@&^IJFDeJe8S7k8066tmWpT*} zGrgzN^SkFco|3?ymiqCHl9y#szeOZXDh|?e(6t6wNlHQ1QUYVyaxSLI4em!?5(#pE ztR9gDv8GU6)hz<)Cc;w~gSlXnyyY=`;Db5+){ZK%M6NACeRGM?C4#}F?wJ<-#vs~? zs!l%6y$wnD78j);(^^i5O;om9xRI|>!)bK7k{0jL8JEa1VQ)G3b3tRC+Jhg2mq?0N z-d?t1UuR1wcE^+F4uxjgnoz!zmo~&^F#*2fj|ipnO1>bxt_KD1{p81RC%Qdy&Zt>m zHhfdxI0z~SannWH1biV%7pNOa8ZFz_k7h(iV{Wds4mqdoI&487TAR=(_&@4$H(n4` z?)M_y5Ac7g%Rjb5|GJd<$Et|(Ulv8U?`;?Pz`($W!SGzb=v=^T#K0zY=JLk+GsenP z#DWcBvn%@RMBX#{e>!=67F5jbAz6~d=FK(Ws)$~$;hdAoaiyUUe%yW4wb zb)v^W1g9lx#HD5*W@M|!X(<|3hN_lY%30W0=!LRQf%c>xe-9-s6cs=(FbfQv>XCv? zN1&FeUNL}TWU>Q2OhzLuZZkdr$|R4l0$9Z2N$?T};PAN?qA@Z-atihXr3)CG6_$+lle2u$K4 zQ{n@xa5=8m0XmNYkE}ny(kLwuJQF$CLH!Z5Pl!a>o&_rd>NYf+>V|~~*U`B*(&;d? z-fGhL*A{+3DTD4nz08z7b=WPGLB1MO=b9%b(_H1Yc~J|h zlc-g@PdL+#Oz2XSaNk6O4w%cY8V73#;d6~>85;K)No!eiw*w-|c~|oaC$S4*gGuL2 zaa2pZ1uvdoV7zq)0Ay{xrt3fhK=2sG!e{u+lnF$58iVlpgV18>z9o|B6_UXqs&x}I zbaO%|A-l>WU9L*9DTtRUfys)p#3BxC$eefHv1I(Rga$QCpUW43!5fU|`l^rI=TZQ+ zBTIP-I4j5JopD0O;qe63E8GLC_!@~eA;=XTpC3FyVlV z{f0-`?@IcgyW8Kbp#M=ta?Vy(fA%-8I7Qo;KP2qQNlm9*AxCzA*<59C?K&K8Fh#-+ox_vM%>2L*^Hwtf{-Eerhc-rayes%tY>oo^skaaf>-p7YGpFV) zV8NO`{j--`e15~JutKl^y-+QGgU*wXJhn47(a#um;I?5k;%IQ#jmLdW-4pvke1;`? zk!pG#yl}L>nI49S_bRg|;Owvw1B?Nlhx`~Q$9dw9r1mfHs~4%kIvTWu?ujLjQ3iz? zZ9~=b+W`|pB%B+?U;5Jt;UPK|GIF}i@F3|oC_4-##O=lJ_CY>>nxvEuu6ev?JkE)f zA3AM9i>L=G8JxTOX{G8$2Ybo$+!s1rR1ecG|MH8}1V=N$K%2xK>~9MI-NVu8Q4QXQ z#$q-qS_h!ue4f;qzZE07kY%q!5{S>rEYM5au@wpgM_j1`Z*s_Y$F*@ltA1< z5&=KASLiQ;$X`%M?Ij7Fe2aPEA;TJ8e*lqHXCWI@Uh$B!G-^aX$}XCNC!hPJ;RPEm zakxVB@L{_pq`NG@jRBy!Gp;ul7t*pR3x_xTqM(W1(;V*FNo|N2La zk^T)+OWHU(8Q1`fg@6B@gOlyQw%z1*K`2hXckA;D0i2U&Ir1Jk!iKYwQeq+joy5{2xq4~)P6>9{@iGsLJ6_N+4Y=`RnLz{GxkFG1fuaovauEhVJHZlHd z5x%ua669N(YRW3k7R@Uj-Uz4~(X=1PP!WF%s0kOl#;iGFiz6>@Kz&koP`F-0Jm>YV zSi$SlIuoRfjJJ=ciSu=B^8dimp9l^DM7jN~QybQg)7w=u$IN2*OT7u=vtPN(Ws7Y` z;Uw2pzR9R&Qb>sbl{vrLy4HfeL))dgPw75*bQ&GP#QBt=vYv<-;84==a|$JfNMML| ziXE4oflRI!mMCunno{X5)unhmN`BU7m8Y0_e=qh->_$y4)OzhhoC<}O^t~CPLKWhK zsrox8$^8u2VAz;OM`y~>z}v1)A}CpKxS&Rkdz0w5nsa1`E}%)oE&r3CXAr3KEkG0c z#b2y!9S^bpaVN2BPfi`a@8Zz6i~sKuLioQt)54ZURz}wUevA`T*S^sP7+<)MmRhRv z2y!<7B~6B0c@M>1Sm63UpcX7>9yO92u+V86j&<@S9a{~Lx2KZ$Z}Sk`8GbS;C65iv z9}utb_w>gWMO&O{G=!G9kjzKt-bq_mZRuAe?=M?6T_7CMSIRiuPGB~a)2QPV!==kH z>R?(uYgE0>cp@_A4*HD!OjUwC{tV20!x0-$eP|wpkuE44z_rD_!s5c{XejG)HHxw8 zN;46f`szwIbk@)}?fy?F?NvnK@Fn-I|$s{LYE9x8L$I*nFW4*E5CMcMX`L)h`nROTGBYQ2mRs-awn5=G@n%4e=JY#Nz zX6R^`j}Q2skvC*{6re)UdGw>jm5ipDmgdHpCfebX9v#ORZhfrQas;L1DKsf$B@XMj z8hh1EaAig0vlC{jQ*>I>&I~gvadf0g4N*;Dj>P+x*U%P;J1A;)-Z-sREnv zYAdy{0StBlg%LCfKuXS`z;OLmYLJ?&FQPJpS*yg@UdH9VoEn(a4zMli+Owm+aZm@t zO`kxsG<+8+&q5JMI998@!D<;Xsp_KKUB(wCQ4|uj{&Nb;En~DP51_5%a<-8pAXh>N!;T+|V$?}1{yS*QiY(|;dh3=B7;Euaa zQUe{nji>p&N+SAiTu`l5I!<*y<4`oK3am8ZUY))<*I+31r(~w{Qhrd!er!DD zQk4#ak{ya&men$8EOzP4IcY4aq&G;0`}xea9f00s_^PAKs`c*}sr(``VjZ3UL|FVi z{Q$(^&V}qswE=~kU$jSc zieQ(+dq*KvI}V;X+mf)lJ2+ADa${!u_-h7nL%6UmHx2$CSe-phf4zTom28*A!JcfbE#Hw-K40 z3a6e-vM$Ax)l^nHP--YDyFaxX6+OrSP3K`>_c99vzt=?j=Y(y8N z5fPn2`ZaB$Ai#`#2J1@5GBo>FQx{95$o=iCfdA#`-rXNLyDDm9w~BdDQDan{b)pCI zYGxJRO93}q>INRVgMgrq(XHohqW^Qz@7)iUf5EqZDPa0;=GRf8#u2jL1 zhR`+omS$qF&~9jvrY;6O1H?eDo!mVuES?+L_uGU#UBZrZ3_2CP?_{A6FD|H$Q_0pZ zN$+V4hWb(sY`dVUHu%Ad>(^ceNZCx3dZ$1I``BfZflQ_=}x8NXHoq(HUIcSw#dPZ;dz-w^;(6`lB~jt)*I{XD%#ZfmBVYVZ|Csk)h63r8icawqj}iaqe@&%t7q@ zK7GP&@-82JWYvh6CY7Br-r3Q1wdpX~akV+I_4WP&+)eKKw;zFTA+Fa6H*4pP#Cp51 zwo`5_i3=)u?gsVH31-&RM5_gn_>+xAre@X>x~uuu)Zi$r-oH2P2pO!@6OiUS!5UB~ zi&%$CagcRppJTp=$-k7+&r|XsU3H>qzoPOWQ>xfi`o1hHS$W_JyH-&#@@J{c{CTn) zN(Ht(Ko5ktxgy8#59}Eh8_aso6ZqwBqo4F&)^%pNp2XlcaPRI$KbTQ)yzLx6>gSnm zuy*oTckNg|ajjhr$17n!Xx>?549Fcz`mWj;YivJ9tzGQ$R&MmDmggjpOTA--o%5~O}3bKMzs2X%isPB?fpjT#Rb{_4K# zPUu>h>LuCmpt-Uuo^$nd*juxbYYvebvA`sGkZ)#I-7QwuKVlfFCgD!OEL*tntQ~j` zmgFPJC>#AxNHT90Fkh#K@&M9cm`iAq$ zQsGz8$Mzs>n$L!a&MgKNfkYDAZx9)ZHP~-WL0q&i-%E|Tm)vWl45OiXjkprb&@DcR z#0)6>BByQWPTeOmTgpu(v7a}>ajd976eK5V-Q$Q)j<&6gCt2hwN=d@t>0P-zn(v*} z9opqWe)Wkg+u@ZpGPg-Rz}KL(*l5-(YJ*nLE|$q^lsj~nE6p(wWGd1; zh7^2BaZYa2MXr7snHg>k<`mdgIh-3$rfK(xP z5?9>MS!NAV7&tO0knA*XND%ASuR(Zig(`$d*BNHw+USE$U0!@=+ZQCjvzD^lVeZVv z^M2nbO{3LVWuc=+5R>{lqMR;}A23xD2OmfZVVq~;4f-u7rOclO1{;5a%Qhl_{{b#T zJgK?lrqJ8y4-Sy#$t977V4at?9vZZkGyOH(WyLJDL_BQl?)NT^PF^(j%YchUD7Upnc&WNe->VH4x-4W>nr$A5wkX5xbI9=?-_WO?FHM{jt7z`R1XZwngYX*tXv+<00VxP%o_5nk)v&I zC=fp$hR>ysVn_tjo!QMplX}&8W?!SmQd8qnIcy5$9x#?c30w#}W?O`rdlNj|PZi*@ zNM{>-huOC)nQ-BBxqt5zFT=&;ti!JAG=wm=q_yhI^k8Ho}k z3y5oySbm*l)LJE%rKW9zTWV@3GOU7(WlmU`cX6DUugGm|?;N?S9Pc(z)e7!mUN5rc zZdH#P$;c(j)rwwxEH681LVHk9&S7dwJ0e}GuBgR#!Sq+pGd-~$Ak^KZDGgIbOPz#< zL58g=M-X8t%;d=;&X889U}ET>LJ=$EnJ)G*^Or&v8(BpqpE`n-|2xgt9Xc(6%wlpN z-DasaAbe@P42Jr|P>}omhSNF!qq#~S3tlPEZX_mz%J_OXMjbVVu}B}!3>5?;f5)0b z5{4J{9VVmJHn@V8X6bbhp_-~NWkEOgw$n9tWNj}j!uf&SiYCfGTbnn%IESnKfvg(w zN}{1>6X{43re^Rjr$>ZlKZaE>VQ;}vic|;RQVz75=aP1AKSIG(ykHg;8dMk?;2_6` zd(y8}dXc$)8eR@i67y$7*3^xhkx;`c5QuLNxMD{Lh07K(Ht~j+ILG?stwUU zSNu8rlqVJGDa3LVRgwpl zXZP^`kyy3R4Z?4KSHyn*5p(drCRQx}inj?$SF&@zFubj?smoBoiDVEn{5ilC<|!$V zGu{D4zkm%zx)LfA0hOfMA+JvuQQm#OLuUG~SNKPtH!CoqV;R@(FFV?vyfT}nI$V4{ zKi^@u@opK@L#XngP);(Dge z2UeLmp1A=<-TwSGo;G8MXTAGimdJDh~O5oQ98vi z4^BlY0Hs11rql)sbKA?kvHU>0R3b(|N*t)6uV(gweYW3cWR*o`$Z7Ka+jrziD0?;4 zp!6XscckQNSoC+L)dfbYY~0+1jhF*L5?`ZUO7x*BFTk-x+0SSVRX>rsb4P`DL#^I5 zW55+i`r4Fg0!*}NBm5wlI=Jg1k%UM28io^WBb55M#{tYCIjK7P527R#-CmTT8Hyl&gZwSfn{R|6^?GomHwgz8dTKjIFBj2I zMSNKw5Nzx!!gr83EUsuAsE!1Z_m|ZM4U2<+gX7N{3U68^h zAVooj1q9x}`djFXs!$f^A8ub5dnOd;lE?cd zNigg80E)W#cAeS(=sN#X-tuqm??3iR6^!%@jqLxg;%s~?4#F2Mtw#Z#m0xMGa+xn$ zX=X041|KL*4e0bwLc-VD4(rpAhevxCJ16BJzuNjA1$`jxgkiePZY=DDf=Xvv%r32_#4qB!5aDN_(p)&6LEK z5*&#kaoL?BPcW9c$*h*(idg%h`p&R+?`p|d@{=r8%UQ6McdBkmg~YA^_04sYUYaLX z40wDsAh03>PC@)v*8~XpaJ1^RdW!|++`Z)xv#bHr=U1d9Blv0#`1 zD4S5$mdOmVpmsTq<+gk#BX1H5435A6@QFP6nduHJdkcpY4=bpV+eFg#=|t;n1OI#F zm3!KnKV=j5I-%6EFU&GxD-Af9%hunck+O!;eJvng*4gmtT$vxijo46I2hr*c8ybdP zhtM~3M<})6)`T|OY!qL0r(cDA8?EYgIPynuh>!zS=d3teb!xC)qqdc1-10%MW15-O zg7RZ`?UDnNs%d(kobVKnB!w_i1|s4vFys1@-R-Hbwn-8mp#k}yujDnnZ4%h24K#>y zG55kFtFz_nUs1q!fy*dLN-FX2Vv(%?xG1|RWV)RsXvsJA#NaRA)&3H5JNj@YHH=T) z`0^3yoy14sE?qAP);F8`zd|m8MxAXrgqqKD>E}o}?`C%QHlT}fqVU>g4^GqB&N*Z5 ze@Wt1q82CPa2+j0*M>}up$yq=(+v%letd92ccn?&A zzPZ@Nc9c2LA(FTBlRjYDsQQD#tZg&%f(6|YiC7XQhC5}X5SJP`-|KOVeHFL7ETC#r zD@>@#pmAJCDy#v0f@{cE%3~m(IVYx=`Q35Bp<;6RYy139wqgZ?8u*XzazOpN-2dPAY03YG6z>0VtAB}8hlkrY zh3DbxsNKkzDTO4yJJ5&>1A&AD2!)8-^d}IAIsxckX$$}mxS+o}!ji@SECNCWT4}9V z)IxD~ty1~CS5rkrh11evg>{7{EzwJc!`9eX8bZ<)f7fx_)#X#B1HVbzo$KK;2s{vm zh-imnD*oBYJqd|?4hb?TCBP6sG60ZzoJDcVA`}lJQS4xukSqYu#^R`6N$|#H=g#55=uJ@p(;yW6++DE%9lka)nV63~NXh z0P|u<@rGd}lBDWn;#z=s~NIY;=YD>qw5J5zBaw*a}fTWSujj%k0kr+ zagz7%g7x^uS2OGn348VtV0Ot%G<>V48Mu+l`#HMSg&0+{CYCP^z}ow~V3Xu6HvQ** z5%;mo$#f~m(5k0n(tXRC=+m%AOw!})u@3)al0mcVFT1X+#~E*Y+}IcR*W)HW_g}4e zw=)5JuNg6Tx3hZumHPM7S0{S(yt@UWkbZ#HRy84ctNI$oMX_NTYD!%-1sx@MwvPJd z(&Rej0mX1?NS1_H1ts03Wt0kwRNM-yXhc8}I=U=%W_@*gZhNxF-{W2#LH0Ix{Tr0b za{b-;-CQa&BEi-TB`v;@g=|D7q(KxF6|~7XvFt1Yt;!F$}9zqcH6=7@X}rJ*~2r_G$lq0eL+nm!U^j%9R6qv_$m8k_<2|E)tppN3oM-$v1ZyO>+qK0@MS=e5xqfp`3b4eY*lB zaL7;ROJ%9VRL8@jh0f~Z$Xcw zbu}er1^G=r61~*jxg;tl!Iu#hF%EqGUU8nZG*@KydnVg%~oP-)v>2f`q9!>1B zh@sO|3|9xp%z+8X#D1*iw{{}`&eF^p%f&&}!IlkO?5c@3aNE30z|-7`yf zW$MGrB9@efR9FYpZCp?6{OMIx4O}Eth^f%%W0db`#Mh@YpXegk5{DLt?R%XBztRQx z>)%RWQERw1+lvalTTiFX?jYqFY%{5iHHLQ0LRs0C*dXg#>^08Y9nP?y&pW7B9e}vm zEf%N*!c|*K<_bCZVW?#*qQHZaoPWQZ>L0B_#hGJ*l~wSl(nqlks5&fmk!;SZDVX$( z{ZU@qAIL*P-;=HhHmzW26k8q+SY&}zP?JhNritrd9^#$NL%5MPTtd@WRql9mpAnqNS3GzNK!Vg3F)krt1( z$e*9W<>L;9XvW*Shptvp)x}G8b5ixGu8ofBE-D+GmkD!Ns{N{@=fcl@$r4GyG8r9A z0%5;q$}1~uP~75moXga;x5{JNN?Ytw#uamajGHWWTEQQcovM}?lvZSzo*|wLL+omg zVVzD@6jN8~ z_AZy|pq1(D{EQ;1F6~Sf==$BC?4q~ub1CMi<-C7RiJwzv=S%#gcnO7_W!@o6K>K1$UX~O!IzPhNSa}IUd-Uv~*m%Z92h$)JwJJ=RoqeE?y+5&k z>Ds)!CMDT!p`o$%ZmQhIx;R=onwrJAxPY)0d~UU-m4L&SfA8XlcT5O~D`*V(HAW-o zbu)>q*eyvT!s}=8B=k1r$p??l&uw@nk~fIp@$yRD0S{>BL!lo&{aL`Si1Y}rjx6C6 zGF354j1jnp3@JCQ=Qn$GXU%bD+2p{kKSMyCf8s+>2&xAna~V^VR2Y#ej6s!C9;W6F zJ&7R`mbp$)E#}muS|$-EbTqe6zO9=YIQwSi7Xd#bd;!+yfUpELe8G|5boETK_e?nZ zl8a^LG%C}|gh=#Ui;U@ck+V%sT7K}-k3a+71pxdKx8&kT`)5H@uqX{6^ygs8%QaZ%@xvg267<1(ieY}OJ{O!Az{jX+EdX7Vjh z$LNcUwv+vUu!!qiu6{^3w=-&%prb`@(<}lxn-pAtJhS8De@4V$77aj3>-%EQOX4W2 z5KHDZ2vD!fuE+nH}}=aTLQTU z?~uWG=k-@YtI$^1s%BClZg=u)y_!!!Jy=<<=g^r;*2bfqie6kW96C!3bq^iF8lrga z#LR9wz5pexIlnf{WM17*oth|Lbw>v=Om%6$kNNQtzlrK?1ZK*CD)!)!p*LrGro3e6 z^c^|^M>d~kA%I?E)k@uKWODHUffRF+_+ZBdufGAZ9rxdv)8~jh;4yHGD5boIj8@hf9qI?~B!=YAlmF&uy3OLEQst(@yL1 z7MQp^jJWUWg4Tdu0iCzAOF*wx$G-XJrizUi2*L~O<>_^_+`S=JBWUmm*fMdMae=&= z;KXf=&N2?=GUGXHT`37l;<(Vi+9M8;eUzx5_y;7vpr;XE*#y=axkFDbEGd!v>etuj ze~YHT*Fw+T^DZXf0?Yo*;A8X&=-L{0I_V_2Z(@56`0yF-q!3MvS*kFE{O}IHI=h-~ zvuxHL&5Bu=InFia2sC`-smaZ)Z;s8Qn;I{bfRp1i2+{?vv=0-HFit>CQVpZ-u@=;e zKR`f*1SxVpL2-9DL`|UB8g=?}(McM@4jWbTsE}mRy;*{3V_1`vKvohG2(4pr^aZ10AQg~1q!@*O>@jC9nyF^Ai4qzr_oiUx! z@JcElXE#=OF44JY&rVJp7P?-@p>nWNU0E-!i0cMnuyMLAA(OV&Clgc^GXDHCa~R5gHB|Z)YLNCwRpI~!vYpwdSq}4g0Y`t@PQF&|A>={z#NuH#sebCnI*;M8UpY+r`>e1p^ zRpyDGbmfxpWas8X{)uzkpB0ete77d^?v{=01Lw3m^^dP@pr3}JJ{7k()YsfXy+L3PZ6cw1Xnv#&(KBmb|=`$ z>%j9zcz@Qat^zXCthWUm@L?!Z0InoA61(yJS=i&=(_Qc*3P1NZr&2hdh#8jD`UkR$ z{L_3w=Doz~-oWlMe~%L8PmK(7_*#cJ+&$*J>9hz9=yGO6I{Mb?ZEJ(yk&g$<&VXbQ z`hPLB^ACwXBe|h;+e-9MNc1V4Mx~$(@|;El&BNQB#&BWzcHL;Dpi#PF`lv>hs?w0P z6QRdtv-Rh=#qAH2*&l>1UAImILGs82yW97T`*nnC)^c{uD=o@;dP(QM0`3~`~@ zY)gAo=Dd@&JQ+K?=uiMEd;7{eJd?Jx;+KYHAOuZM@U5<}zturuqOZu?^;gqy<~ z=?+tA;P-?e(u4?AMvyg&sJEDaPily#hA(FQw6JYKha$3(tEitE$7z7~UkHfm1sWv!T_|^LHi{BA05U(23LN1Q@kXSKkaxu7ZZHczNxvuDCsX1~)yVi#pP{8Ax_`X7_na z4`n%0kbTl?wa|#B7o=9W3N3kUuCWR&iL=JZVDcw01eWxHwL{EbgCcsdLVAOS6ggPp zdg0aXX`obB_1BA-gVqp#dNlP})OX7+z;nc?k*@ix9;4#@(0?>{H6{3-64TX}@@X4Z-R z@w6TR^q$^7^~>A-8M9xOj^FKl|Bl&piV$X}56gX-&;yuSXxd8389LsJThW0WZAQB| z$D=y(F*otDkBZBe%>9SOv4EBs{2r1!jKvA0mauG#TK$UNj5~TnDk)u*FD(!tvwqvbncl?m(rBW+;ABw=QPKDNPHSb$X z4pR%%Ne zEd+x;^%ZCG=_To@qcQ|-XLw+gl8h*7%IU_(fk^b!Q?~lw?cx5#^QMwP)#$%IJ;@dk;=m z*kp^@5mk1ATR+Qhhu;wbau0-;?X#oz3MxBW^ql+|MStSpN!-MFkD7AN3U-WrDzc>j z&n4k?!aZ7Gz#;__wdk0djme^XH|;nf*`R^6U@;l8VdJh1J=dmz=2iTVLtk&nRr;{Nqv1oF4LeW4<0vs_v}Z?^Od zoP~*Lj5+D*o<_+dK&7x_u&%RULJrE5&tjKn@PWMCnMEr~-uY6TMU68xy6${^EYhni zfiquyb zv_@h|obpw_`pCIaCK>vts6TV8OgLJ(r{r}_aq?=3j5R#Ralw2$e=^J|l7e$+s{7h-rD2 zC+(6N@XXD)i@8!BaY|YU6(?#?d2xZZ7l8{Z4Ocl;fG~{3)$d1 z>R&=lp{HfJv4!e`ID|C=foa2QR`sQ3N*R^Q1X9q15*ktJ8c>s~7WBV|O8MtbuN|yV z+02jrBrW=GX!(RRV)@KZ4(K+px~jJ!d`;6S46g(cXx@`V(`W7xhgO?>avnLRcTb^v zzJkl>Weu3qZs*l#h;mjaW-KxlIR($P#yWi^rQcC-o9InU@hy`-hsSucK_o7!ji?)t znnYzx0c|X$GoyaK5*<)5QF3y|;&6c95+!^UU2PMeu3h~c&)h&pc-V6OeWH5yVuxI)8N2CT_dvQ~1 zM2tinZ$o};%!DXF_ViNTNlSs(*?t6{$Bo>Tyl18@E))BfTKLSY$NhxTU1xDD)|yp1 zDTYM~6J6CoYLW#|smd|M(rO@~_0Yn@#>Baw5DqxA5pkiegEdPTNf~@#UxC{jAZ||? zA?8fOy{$QsydBNoaJ#as72>N2)54 zN8F=O?v%=$i#KDr5=tQ@5tk72HJ?u#hFPrale<92$lLI}S*2%`UmLA;LZ+%u zt5I2tE?Y6QnReEXZ-myY*SwLmVz-|4j=I3Y_j88OsOPpTS`2S%XsZ!zLx(qi+F>}G z6s`Yom3pJioWbk9TO~~F*xC`kcyg8Wz`!?;**3Uf^epA|`!bUS6!Hr5U6@HMWah*^ z(mYIO7RA@2FC%@#jW+Plq^w5N|9WGlUHTMsS`{&$)yc71Epse-#JX70DGI4)KbBq9 zyl20#a#wr{0h}p|6~@N~*r0P7T>c7+8Kk}O7u+!~ zL*eh#fD7K^bRKkX_O-s7;T)u+>p6RO?RoRv=ZGc*l2cN4Kz*!nD2XXNK@Ek%?D7*N z_YUgg6b#cFj*yWmN-(`BDz0(hNyr>a&~z#Nw&Gc^gT<<$6qkU*@w0w}kvceJ+q5ex z*ca*hX_QYw5y0k;i`+56x(lR(-x&C$;q4#G0e<>A{M}yr$UWd~vUY|1qHg1%on|TP z(QzSfUE5Lp&}h=K`8dc%`DU4bzI*_E-sIYy)=xaE86^{kv33Z%TC_a==oH~Rk7}%S zrPZuub2g*;*QVGf0@hs+XV51VIRtYQ%;pp&1Wa*&nAbPF7A+kS4tyDC0v3+P_e2X1 z6|S(@p1&*|;e7JB^UqQ%O5L7u9+{8vCtgT3Vz1oBkU4pn(N>HSFWAO|((ikg`?-MO zr8!2ascw!A?p$}W#ADv+>?fsu**`xXcis3D!3|>Lu`l+MxyXUbnux0@hICcRoGy$? zZCRsm0iTxqNXZ4<99Lf8}i>I~>ZXvBq#WZ6DUd5II zP8`XN0la~4^3Dj%EwV-&7={lnKT8}|wZXhc;j>3)eHE&Kqd^y^UKtnL=&7{Uv!W18 z-k6^t7@lPHFX8nskC>kBlHXp&KYGSL2**Eil8dt6FuZE(-W6(djx zCp*tzhnt&PuZFtzxtkY~!xMjDkus>=R?9ifjr6P!uXGnz9x0^|z92VG@Vn+fHy+rH zdxiV?F$OvMA%YHIp@$QSX_IWRs^2v_2S^tOaPay=y7%{Rhs_IY6_mjsks%(Ic zHlkAG%*ppKcKcUBG7H3bRHNEbOfJ^rf}z+V`T7-7& z)xKk}O3loeG%V`cafJT$mw2Q?EiKwV5uvro8TBFWOQwB=S<}=uIOfinc@s>!TGwi*i?bx2z;63fp#^V z^iIFXG~CfEplc=KIKn&nl~`RbU~>qfJ^^7*hd=z{Nq~Osrxl9xK?E~HuNV1&Gj}h^ zDBA@i`Oxs3;>}67RMl<#A+esV8zl7V&QosFJ@?A$Dzp_!e=iOCp~3V(yT4SGr}O%7 zDpx0rsjzH^&2%B;g_ZCe>DlOsMxY-36!^Z+mIzeSXZY3G?VNfJwNXl-%kV^Tmfi^| zCq~qT3x>vo!eN;0PnV7G!bh?G;3Lpm00n6l2^s{!cVm_@)nStKqh``sUQe}cxU;M- zLF;|a7f@%GA7@rR)?!x%6GH}f#J5N@d3|~nB35XJ?%)7@7RYP)MpK?Xg@<{s9hsns zN!^A?eO`Um#H`}or`7@z@X*Gf)rGD2;K87}6`* z`Q!08q+0b0_w2Lpx$X1k<~;tu+pEE|m{0o69q7Y+RxZtl*=Z5KP|@)e)jLWx(Q={V z3)us-f?q&>4jW)o6q7s1KR-q;_Z1ULF|VR+j_>89C+i25gI`id@aLAEzjgC>vn0(J zxC*R|u;(AfUJ=X$Xc3X03OvN)v0Bg=_O!M(-ot7Zay>t7hOsyO^l*xqO`9ha4dx@S zVZD(AfYiN*k#8AWA7NBlK8iUV7_z1^W7#=;AdK<&OW7feBF2q~0sEdp;yv2}{cwN8 zTP%3;Y)-ssxa;l)HC(Gzf2>OTo+o&SRcUlD&8iG4hAeRLG^tKpNT+MA8oFCZo_Z^4 z4-YJm9@gYVYqS7?aJ)eY9Voj#EXCz^^&Fnj-cN6d;D6w<>3tbCw~yJn#cTvV*|mNG zjo$n0NPQ5GFYyaBd{XYdn!sfF7&Lt1YnA&%Tinyz5r^z>-&fhCed^YqZZSlDhCtjC zyGCX#n%rMqw{HgX*WCIH5O|s%Kd&;_&NGZK@cgxWMv0s>0gV)5D=2N=(~nl3Ra&3% zg>>HdwO)Ez*$#tOwlD^F#0bU|fZJoEWNDBK(V3s^MBPJdWs|BvsqHpLMq1J=7Le5b zS775DyBg=YDI%$fovD%<)DDUTG;GIEDw_-U&ns(7_ScjWr*|-OoaMr5riI!wrvl{{D{-+W9 zyxB0|Uq;w(U>^pNT_*z;KBmEKY6F^2i%6j~#!QyH*xSf3$5Q$H zABh&YF|a?QWRyc>8BCj@>{v(lK%S+v6(I@Se-glmfOth{I!(Y7JYC7h9VeRK2IY8t z8n`B}2UMuEiWaA5@2?F*VYbP|Wi_anBp3ussc?$7s7u{##!{w|ltblWPC7=J3(#9X zj>QP!Vi8JG26im|h}hD!V!Z0_Wd4JtxWisZX!@6MjzSba5BMYM1jBIrSEYQIVq^-2 zo|0ZUBXkjVoxFZ}*vX)Jvc-6hpaC8l_Fc`qny>qwCHb$j06|i}TqAUfmO0EJ6RH;G zKxcAi?F~9Bj-kXx)xTfIoKW#b=iJX>(Yfdkoc^(i-7|O>(e(HUmG2Wr_jbL>)wHrM z4S5unV{D%uKsN~%2pR@^?Y))I_%kXU zi*+yfp7FZ=dhhY_p~_jTsXXyQGd1pA{Ly#c*& zq{qqcJwC3!Wl^I+?@|^xxp|IbF?$=7GF=~*LHWWA15=L5wUxv(5k$CiW$K4srGtL; z(HoYQDp?~$yuL!ldmqQ`!MIG{86BB&&4D@S=8R@Gb@|cB~ATPb&b$;7TX<4^+xd9$Zy+{Hl z^otv0kjSa8zubuD9qBaG74hc{l}r4nS;=_p$ao*k;~VfW4SCh$|3imO&p${i3a6jo zSxUZMb1N`XWcQiAm8X}b4|oIVv=xS^dmV<4A2b>7_sL7U!B79=8O^f?9m&|)%d%$? zQf!SO=4F{Hf$O0vLC*2}WEaNt7GA?LA6280(7H(rL}!a!qsBPP-GihH*Q;RDV9Dc_ zG`>pmPJpFR4WHP(yb47fA%7+QD>naJjCCR<+iHk~RbVmK%N*^BEq$DRiwy>#)0C~- zsS`Uxqbf3)1ZPeTxMQp?P<1s~v+ny)9Q(~!KFD~v6%2#cL=zutVUA&e#gPH)R+AD~ zC(WOjLfO17H9*D;t2sen&D5sXw1V4|!pr4JFCxUuhFmsss_sQ(Cd5OUd8VQtVjAXy4QBBN*Ro$J|-4MB%XIn_jP$_2Zft49U> zt>;w%fIhQ8y%7x8Hi@r!{V}Pmw~qCo4aS}vv6|7>*jbqgVVk*T_bpKxZfWFPI0#Gm zisWb@X1kfRLpo44!dpTdTQ#oej!aJIc#B*FH~#PMW^lV3)er`Ugr;yJg`A*z4iqvh zvV-D>4Erg4cHx}Bc@uFk@$$*$(IP>d;0cFnG^|5MffnQ9 z%`-H{$5MmN4pi2xMLRNeD6%Q2t~2S;suP&5L+N4g6I7nYlf6fWI5@0FOu>h7W|KvR z4nPH|5fmF*QoIx9qfC4AOIk7@t&OAVMXc<$-?O=A4ydxkb>*z2s!O$V0xaO5hiv{mIT_>Pp z&2{!`-OcEREXRk6WutS9!FJqk43Yi;!DCc0?hH)S+Oi9DJYUKYi`Pf~^QP-6> zOc~|~=XqrD466~O3$aWLU#PG-n@(=RAE>(?! z-PC>m+!W=&7Q1!;k2g-67-9QM)aY7gsN0UuFByBoEB5TT*1i!`6bkZAIGLVgZ?jV1 zDl1oOJ{7dq!|uOF@^kIghWNT#x0SnwJx$XwPO&}qx$+}8Xj!c_eJvN%k8ccw{~(Rl&_WsG$O~y zDlLZvOOn+CQ@l zluxRGd{3_srP!im$|}={4V^LAbD4bw#V73oI+j*;8v6d1s8|{14yuc&q9=bbyNdPL z2DY#s9He5%V${~cdN2VJopqF(U01KaYjkxNmMzsrzwU+2+p04_58KO6D@!n|M%4}Q zvg5y%d)X8dn5l3;C<#**F zCpyOTjXffPxrg~3O^zPG>!!9nGPIHzUo-9yF1a^g)hmbtt=W|cbLi?;?huqn#9w#F z&;-3)VH?uOhgH<+<0dX%ROlM-`BRnAqN)@A?2)SC=$#v17U&(%nMsVUg$3=xLJ4|@ ziWTAOpMV~-uemL3iXT6&r2bRC;s10H@qg<#RI;%!vi|q%(SJ>iBwaeUi48#b6Ugb% zQla*kAexPZ|5?FmjAcnpE~so+z9jR)@bT*swICA7yQ$NC7ajBnS@=cLYi&Q7*%H&@ zU}Eg^{`wBv?ZgxPt0y$qZ-Uuq=pOTVENSQam^V^Sq5`o znWyNo_981?=|g{^tT7RE5})xv!#*unlfJCJHB`I8RiyMDl)_NNX9f`$9V z>+Veh-OPULMH9BslR!cl&%+Yi*a!^CGX@!~rSGHpeRCkVu{;Aq3>ZjXv*DjWy}x>q zgp&-D3_af^>HtHOB+JXrD_H9o>sDUrBpCqW|FHIsQI>{Vwq`^IGi=+oonhP74%^NQ z+jfR++qP}nR>wJYx^H!L)wyHb(f{@y`{(=Z^{#KOHJ`bty1e2%MqBzm*kf~GAzPQq zXE6k4DnmatGlHLchNgqJO$59f`$r4}51Iq-+N~QISPq;A^_s=eDrQY26;+tQe+;2n zkV7=dAsWm9cSqiC3E)l2QJi`EADNVsM)!vL6i@U zRd#Z0h5;O*2e9+u?KEER)faN*E_iZO`s@)J%Q{&?V2G|JBtL(b0OdnT4zC$LLad-# zeu$)jlhh3HKuI-9hho1nrL5FUk%YZySY3^Gm!o1PJkDD|E5h0nPK*TuTI{Kz^p_;p ze2$j9B3H<2!6lr>xni@wZd<;K-71|qZbpq1BNmsKu+fJw3x@#8$#(7s@b1Uf*`p}#_148GsGr#?bHWk-e{n$2obVg3L{^2ko7lLAj`QGPk z`G4*6y8V}+Z|47bpI64f%EZ>e8esg7eR&Z#BY>Tgx$VFHY>9C)a%%zzL0h3hvm}Su znxksKzF0V#W!fdp`F{k=WB&x*A)(bJUQ}2GKiWPhlEA(H`791wiIgIa#kn?ceS8>a zzA}57+G_IzT2n7G1ie#1snO3JU`iTQCzH-+QXsM#BT?b{^mq;x5DSk z;q==S3z!u{Ewf5y(6aSLX77a;!Fevy$lcX9t&^c(dnh1@Xi}!>a>|2F%?J&TNi0da zv?K1hc+LnPIM*xy?OK?L#W%wZn1xrweKSuIM|{Z^aVv-y*Tt)8eyStxGPQCH-WE33 zPm$w6Xa}BrO{hB){$kju*pPg@8YYGvCmLS}PEwO|WCA7py7{9b zS0HWjS{nA&H~g*t?Y3T|I&RG(!hZ0mC1eGcI$C1ec;ky7(4y1JBS~@h=_d!Kylpq0 zaa8hhmOl932Dbp^;PS=?eoWrJ4JxO&k_~9&xm{B^_9Q zi0-`e+mC;00d~2Ttkh2>>Azx650#N8RsE+XWbuBk(ajF<0zmkR?m%Z|)oGuXc z>g0_)CKo$={sS@tY02FD?bQ?gemMS*7roUS%>P$DvAVlH@>1$o#|$H}yOk6Z(;&n! z*`Jumi$5)}CGe2oAU}6QdVGg(mm!TuVvVD4+DN3t*(BIWWa@map#wL#>Lt=*mQB); zRxi<7Pw;qMJWbaMJ#FV~(mUefrms3MBgGmsUg5i*4o^C+-n^fl-UlJ{+;F=M7ispw zZL9_X=#RV3eA@PN*Q8*2hxmFseR{*sQeb+{eSu}~Z)w=+*Q}kd)L^#PFK1rQmOeMO z&(}>}&)j@oeUYBmGM%qU5w_INp6sJeL|+7Z+&0R)wSHH(FP>h{eEcsU>^>p*lx|8r z{I`3_FNe`D*T{VD3GrT1JyXx`a!{WdJzdQ=nD1rSUqQc*`hE{T%YpG9fo{3%nP?xR zciHVFKhx8lKB2ycNnKllT~R#S`kepfsQ5bQ`x<%v4R+OP9}aiN%l96H-&MKo4TgU$ z#10AJ-;3eRAYsB5xB@t|Vy4P33w`^i#N64itUcGxURTux*Mp5lOk0QBGt?yu$<+C;E1w(RSx}k-jrc z%t^-!NKP}!(RtJru%e2w(o6#cLk+^xIDj?@_%~*d8@v4I;o=U>>?BF^4lN>(dyRxe z#TxMp4lB&Cp}GEC&ZH2)Vw`eM=r-D=adi`^=ZU7hDa|i-aF6==M{zmry^y#WaWB(= z;+duz^QrS_#EF*7nSBAF-%Km`OIWXSBf>&l@v}rAu-N9-lDZg&y*5!WVPj0x=tuxA znAne=5hh1r8?L!j)J44Yi6`xv`W65*9SJPTbO~%B(t6GPb8xZ`l-#sm)9>4?HnFp4 zQccB@VTB;~p1?Vz(T+=TW$f|(xSzRy%|dnee0&KS8;<^KIBAr7|3s>^4fsN0A|8UOKLYU`uuL%_c z^H6}O6gl#8;XkMzi3@l4+=Bj8Dk2X3LUr{Qv}{y4c6)C$j=jt@&+SkEX(jHzQ~abd-KE> zH5cG`$GT}Mt?=N)uXrWW>iQqp2ZroWZXsh-Xz?3%^gXlyYH9L zj~Ui)^>Vsw;VVBIOMs6RVTnmQogM)dK$Ux}fSBXEF^(4h<^oS^quox_rY<3=FQuGu z*om@^vDqpj*2(W64o!l5lgDBPEa?!wyGlq~DWx_A%o=M=zA7)DxwPy=a>UO2F=kaV znpAT@rPZSvm+riDos$la$#;KE*b|I`|7r~9j;cj##>Pe$YJ#3tqux%_HkZDkF)9DS z9S>5lQvSNyJa+uBzOFUJ?BA)DutsVOuYpYP5;?pQK)@s}684v7H2t&4&Q&*=X{7wo z1e6?KJ9FVco(#xgDFHFc<|2;=r$g?9y8_T~nRaWCG|>RoSqKPr3CH_pR$ion`MvYaHQgVyINFM4q_Nf567 zZ9F5*&5W0|YgGOTPSo!L!1M7eB<{87SlCe^TjGqsE-pmxH4$G9Zj=AxfOm3VPjzG< zR~K#SnntS2)-RnXG@fuafWvMw`fMnRb@#XCchhqvaGqgoA)}DqkwKo`A3zSHIHBBj z_$A`9U?bwP5Z)064q>(DGb(Mus3it`Jt5-FZbdKL;A7*GNz`C_Py}i^+&%?ObpMUNeQ^;TOuTUd z{g(sY;M*)Z`)XyDVhV0pn+VvyVxt^V&>D8jQaT4~68j|fTnYBt0xmqS_aN$@gg~zg z!K4~O3pCA?hYW-FU6KF1-FSlkX#fW@Ej9$rN)9yPL~mOD;KXrT6I?J?bcSk^PaL}1 zOZipy%gcZMor-5FJwLSn0z8IsyBh&N!j&)q{&Xg`5|9KGfi6!9dWh=}gu`*@ zefV1t1u`H55xeSL%0KJYK%80VAxwzk=6D1!@H?QHyUA*?JukLYoOSfZPv@9IBT~rQ zp!%i@Xk4YgjtmT}M5mmC#H7Qd`ti1*R&cYphZyz>e#Qu*DOU32f7(IW$xd~+ce5fk z7j%KF=ec)tqE3CdcUOGK+aX5B z0_!dkVu0S473xS`D~w7@mw!UaOcz1_l8DCi)S9pRHnM5&Wdu#i&p zXG8wP;PaKk%5t0~XqhjUU=K|r@hbQe&KTsbMkXb)kkh*h&uUCw4KbX9FO0|y$L|@E z@WNnY7#$;sKY;X@0ye9NJgevzM~eWr92UmIX;SE#nId>K%<4Cp`DZl-T>0P6HFN1#iFfc~?en!>b7Q$RV5fQlccpOcw=)6%R{b;%2pp1f%LP&~-ouotm5 z#VerJB~DM9Osiei$3F0VcP!wD8f98fGz`L-JCslIB~zFV>`NVGj40HhWHb zJ{1qj=d=qJW-L)WCnIZ*OL*(6UC?P(IBQjrmRO})lBdp`!zNirDCM}FSgJs_Ae|}N z!TBzm_KwFWIgXV1t~uG*((8lw3H7FtaHpyIYhi{s+~y#OZcA)%+Cii}gRGq=AAzvM zpNh6uy96&hUz8@vrZL4M)K`4qhu3ELB2)g^d}gzQtvwD@p|zkGhA(?Zr`#P zZ;-Yg8!{_!t;q(pqaf_%sJ9OCb+Ib0>^;iyI;37;taMXheU#s1WPjEt_@lC62-oU) zki3*5`JA!MUHUKP(x|Saa!lGVo0I|_E%A~@6qY&IX@26qVA%`}Yo3axzg9uRGKDwj zT0WYSL`8_W^>n_Lz5MDQ;#pB10_)zNtxwJMm~s+YcA~AYQxZ+k6OXY{4RLGmWLAA+^VOUu4I0+x6M0YfBD37Kk8~+f9z7H+x}*G6)6m+v%5Mtol1Em`1bBlX zE9bO+ZPhA+-;7ES+1MGj8FZt4>!p%piq@zftG6BDy-yU0hs%!h_f4h|a2c)vB1^-e z0lVw@9G=Wi9jvH{N5#H>Wqkh3`((rpZs9Fay>Od5bHA*@QOx4zZ_p`Av?1oV=$^9q z*MOo5YKlIzD?g_<813JpbQiWq8xC$3tQ(1No8)@0heUyIwz8{Hm@3d*D>hQ#237T@ zuJzryXyf$-+1KaxODb^W>o$|Z7AWFV<@=1s8>&;_67)md-PgfM^2;5Gic|73RAA7j zV*EBTjzGrQGYymph3U_tndKbDyzeMxtjg));El=Y=RM9cWbj= zV5^MX7%S4N+z%EGHz;Bo7%V*XSAi?wVAh~s{3*6YYUID>c%^W58*a+K8JMzO@AZs^HTxM)Qi8FcX?2q7Q?sHLUC9OmDk4HQaus(EQ4WEmqe~3;zY6@V^mt zm5gld0RJKC2F8ua_6nc`eGwCch3tO#Xq7-)u3Jb~E9X&GAd=Vvg1CC(W*OzZ=+(&R zynygW-3l=>GSeqtI$x%wWxw4j(Q-iK%hUw^7d$Sv%cRnr4QqpaMhBObw*r4Hd+RH9 zeTa7RDRX_t?cfn5v#u~DtcM^6a&GBuaN2Tog4W7VuSPo_EAcucCuW$WQZO670FGX# z(<h*Qwhn#j9{x5`k65{= zHK~p&1i6LBfoZ=1Z0>Mx0le(;I%O5g+;OUHRfc8R_xZ<~?Al8p4@7c516#4~f4F0w z4$hh#6%*@n zM2F2Xp~K|q`%X|vb)VdJKLqnfZ}H3WUqY><(QO#P?eyQA9WMQ3D$yRm;t%fwTCB&laO{9{J?HmIN zudY;lP(&gV7r1|o+-f2%^)4_aqzo0dG8C%f3?3QDet|C9{vd_@uXpNg0|LBf;CG6C z8O1XsrKkDHPi>aL8!r1F(u{t zDCVayES#@+zlL!~^14q>I3j*D&2bo0G^*$^jJq3ah2pi8gt2Z}3^aOJgK7^8jJI6^>4HTvmh=2rr}>ZJWGccZ>Kr+h zEm6!fa5gKeU0_mo)4Zr4Us`M#9+qlrBM7*@Rg_?3^%wflZaV+;rD)30Wrw*!T5UxT zaAE`x<@Fa28`vuHU6AM!Fl)gPHaW;w60z}o^5n|6mmsbLSr@&KiRG)#Yp#yu6V%q` zH~L*1=suMSo`-KBSfNQBA7)azXIN1Eqsxgwig*_$7C`;4*9r(#uE65D1zpXRz8-Uc z%oh^jAL8-Zt;s{O${fL@(%@5$PtF{rLCHH9=7nB(l&+moOi4Yu>-GsZ3*z$>^308H zo(pt4=DwwBfyNZd%sd6s?ARYqRO;9(UH>QqqD|o`_y1cU*nbE0{)gHuA@^VO=D^sA zf7F4%se`Jpzs1||P*BuhMtNGW*z=%*^TGiBd5IEA+&Q#|2JO`5YhW*my-dM~Um-df zEsl;?uCmP8dRO>BQ8n?RNFuaanvJjtFbO|f1YV=;+GgQ#4mKJlA?L}A(?_8NcLH^G z-~dQ@(U2udLyo=RIn_(<2y(qnfu`x~Iah9vBRZcr8?Kp5Avt6jX$=uYq zfSOo+?#}4+(CRDD0LdXmkU7;+mjZK$QRX$8in2R&DZU__^8f^V(8j`dd(!zrsaHxc0YBn)sXofNKL%ggC4X`Ko}TXE=7QCZm`ZpBSZ>H?xtx?@}hN7?wq%0)8T z(EKs{LL7oWUJ)Th^7XGdhng@B5s%VnHR{&Q?dS(hOI3e^bciyCm!A}C8H!>`X;(xR z#`HPdS5tmdpfS@D@#{3lqwdV-Pa$kWWO}LEl_T}r@@nu?7=unq5j4!tU+Y6|BrGzC z@y0`nHMn*~>+4>=>6{Rd7*=+RUf;*L_HQf0e+LI~0GL<-jGXBI!Fl^n)4B%rpKi(v zsGoXF%(f=CNaFf`3_?ReWlX>ofBAz){1DV8kO%2$Zs=nOm@y_Bmj-I6(45iKs!v~B z772%LD5$W6-1W28vDj#8v?x<7)2XlwSI+rrxdJe0_2w^nyJb3&&$pEAB^T=7<1 zmpXt^{l9_k<>kgsT~! zs^GRAj70a{h~ZX^5Pp*Mo%J5Vy1&)I{GPc{gZ1EPHhAIOFn~P>pMt7{Uvve}!+n6} z@(=CDfvsE=7h_2|4|VkRdVF0!8B3lrCrYF(!H^aw4hddz3KAOeFpz;dTGcna@LRS} zG$%K+8;=ww9wCDrXmGY(x2=rO#0d@3@@Q2J+#B;RpD9uPLCXs_JD#sq}bqi!QIKOxR9L4U+28N!wHqn7>863;R z6_=7M4*6Mpkn2_+4l!ypsT7(kUzBp*Z0E{@14T9MqFbiIL*{-9pOaHWo(P^_iWf#D0WXXgV>ax$=#N|APgNyAf z-*fkwU5OP7zN-ot2t4!fLz*i(QUXxK5(md9{a6a{C&pyfqxz$166S6DWgMuYTsdRAp7@5($f+7>L>~mUrkNx9JfphG~iP|_5Jz~?( zX97V2tFZi|om}U0R>d@SRoaE|B*-D@xXXAvfq0>YPK}GSdfkG<;}}X+p4j<(XvOxh zFvx2{-atm1jjvob*ZT5yP&<$1V9`fJFC-UjZ>D~wcD%5xg5dcRK8Y6fn(^u0sYV^? z$3=Ds!5o-umQJ^CgW$Pj{nv0nyH%<7R=M0dy&ELST7eW z1b#C=(r6v57iv0eMiev5yH*hIb9c9H&C>7doS@O2v3B2RyNW1p@X9_d@JCCAYf^0s zvowk}acU%~__M_3FjWj91*Nr~(p+hlJd2%faz+JItvPjo*+5$4hTh*W+MUuYV-+!) zmEor_frT{nb(V4qV~Mb>;l%

yfw!S+$Phln!w6PCuVFtB@xxIJM5C zjX584g0y>bC&0g4RQ{bt&9S>5XRbh8^T?6IIiuz+RPQyzEwd!(7anNX`~&W;$>?MS z#Y5epb!_OM<7+s8-zm0kNggJ7)#;SG6RM-X^WF*C6Mw4)21?s!TQ6j@N5;e5`|$TQ zVQC~*xJc+?2{k`d(F!&w(TvS2dgmk5>rrDjeVPydWuobZVr~!6 zs==DVOTl99E?oMfJw#i<1Gszf{1?_;5DL-@qF)jOnnsDHzd8QMLKWqD37m@h9-C~N zf+c>b8$C5|_qOdAq5(DS?}tPDg>*)gv2a+1Akv8rbadL&k?o>w8 zQe+0`>xN~eqhLbUL*jD&0H_3(P8-0RzJDNHfOZcjJW=k&?Epzpcn^JiUwl zpakM;#OIdQlctlL=ark>UI0r2X8x*ckyC?u{@h)q5|aS-O{1nQpmxteQ0OSO=S~f zlgw4Ef<$#Ka}4F;7YJX)PP%hYZx}Btl%9)7R>^cH*mcDy+x95A&Yw4P(>dnp9-$f6 zn<$caiu?iiM@lF$*H)?A9;X75rZTJg$aH4kH@CD>Wj^Ot^>drM1KK8ELMz`abE#X?% zTXJV_HCuJr0T5@~p|Gb@Th*E;Z;`3}w> zgdW8RS)<@QjS^~)pjclMnPIUJkCyy{?^NMU&G>S|j-fucYsg~-pfY-Dm%)T!D7Y7G#BQ{Yf7 zo&)H#NgHKgG}73NFaoY=ejwk)02qGGrT}&m|4prY(yC6U(b&k*ar_fF^4xMMr)=VS zG~bR~hw`}t>X5TlDUKu48j{;B0y8$JZWMr74Q&oz>m2z)rb8MFO)4r0d)3{kyY^iy z=LpzI;FW3-rO5$8-zT$v$jK@3+-z)K!((U2Y!Ae1yvx$(qLPn!=T#l6*@C>l<^?Q4 zNXelFo|lu&N6JFl*}Xw|(bd8SzfqXHBO)+Rhw*C>nXghKwz#n2k!x=Vv?9wfW8==M zEGJ<|33B{>ks_;dLs?yxX@stlU^Vg_zc8|9!bM%_Gg85zmD(K^b#b$op{md-?{ZfV za-z*(PuNlIR8%X781h%bO=rG`)5Bn!>q1SXx5Wn(nW@TjOCK_8D<=SxtNsd&tI-Bc zTLrae6d+~SpCWtAl58a&>#H_gZRLKNp@g z;bKv$oaV5Ad_S_wO*O2Mf#M6yvo!lcwqEZMXuckEo zF?Ch(V4-~16lKRtaLNeMs-$&MFh<63$MrEyrhoN>F0*l%KMVa~y5;fN)EuRGkVT`r z&uN?rU1K)Y_%`HMF@1D2t;&*9H z)VIO?swWxmZ|PBAzxHn& zaS?)J^&m2)BmJP$y9`dVn3Y7X+``XjE2R_+y#7^K4vG^gdiE@Q0n|6Ox4CkdGVV9b zpC6{r&FRO=dmS+DS9lplbM#6UyJMc;&%Ad<>DKRz3b(oI`^0-Y#JrWt5l9nCn*bx0 zbj`MY&lhpaLaT3e@VQ23uP{2e_Ip58bYFCA#X3z>oWDrVS=d&H=Etr^QJg@YT%32# zA)N85a;c9Y_TUa%$^B**Qaj3!#Jc45B}W-03o4t++n@DsPzlShjISqt_#fkT!oPXu z3f5ANmU@PEf8qOI%r!#sQ~GN|=x3Xyrg#YrsIRXF8m?6~6dV#3z8`ctvJfd;88?OD zf~x$AfrSvlC+RZ0F(JD5!xy}(L;3`VQ#HFkUf5@~I$y55f82g`bf}h(!d*2*2#5^; zWCjO^sKnc1NjJMt=!x=mLJb2J2vUfnM(4OZga(KE^P>HjoCsA(pdrj#tz8u0K#0+o zv5VACRJPdN>f+D=GrnOTHA~c0p=@~_jy+AZpjTzH$=GXXm0jPYA5HshiNsxi+A@Dv zmR^d&G^?)0B)_zzsy$&v>rnh0QP1a9W+O4*k*@yJpyp@ueccFd-Lyu%%WlhroWylF1hK7fHx15RvhSfwH;s#}8Srj)47!yg5Ab`_GCFga&K6yh zI|K-|Uz$8kpDmU#h8ypeZ6~UK?m63B$%V_#oaf%-Mp{3F|;^MtLaD&FLp%s-24WXddKJg2F-zziy^Hq-cSB?sbEZG7n(=9PZnfFoigeglr1=y;k_g=sCncugcz@t!ct?X$26W<0rE*2KcMlob z#BT0flk)KI33{z3BvTz&YDH(;d2hjBfh5G7u1k|P-8hdH(a z5{DtZI84lHOrIHGfj~jrh2C`s!27|+e;-or$~{9i47;rZnzvT>v%E?2zl`1_%J;7V zeU*{!WkrNTw~0` z3|!#nasSeS_$d;KoTIFRepkiM95Td{05#3Ic~iDPc`5s^R`Aoj<}XDqQd?ggL;xxn z`p>b)Nz8-Y+jFcg;0LTTmLqXc;QE~V!dS;K+ZI{^EEfgg<_f4yCq>&06lWxcdQgU#kt?{Wetd-V0FEKYGWr-Oqo(+)@HE!m=RW5O!XoNpZ8N zOt7qP9=o+SbaiC3_|*I8G6w^VV#d6BBJy!!xUi>I;UJ_ov)72%gZcZm-8}@r=lItY zBgyvTVUm@9kb$X-iYmy)Eg^2E_Brx0p{5n*#S}{84j?yl{+TpBO4fF#`mzD!zu4g4 z;|t3F#Rh+Dl>0}M#$R42AbMO1fEF=8`q}o|Cl8{bE+G}121|%~1o#gTnqnX!1Y5-c zIy5;ox>>yKd5x!UT=CbDka#hDf~)Rso$gld+?^ZT-!yV9atyYodVcV`O&AiQhSs|| z_Z)sb5Z8x_JzbanXsoBq9b$-J(s!cYlh$A(#Udl$aS-Cl=YyaE0_9e0cFpGz^|3|c z@;k!un5Sg7LK~<<#y11h7&o5Ag17LQj7T2|4Btvz9{9Zp`)7y^ipv=eJHcD2!)0s* zU`syy?&|K=fqj%dkGqf88c`z`lUO*e{PAuaJUfW`<2f?=CE()iwA}U*ten->iF;GZ ziW-XQNdnHNag-tW-S=l1d<&j;P4G2(ZE6}e?cfzb0)Z-A@;z@19CORihyoNb*N#qW8g^jxHn)r{;bh`DWUd@e%XldKicT;G?@PT zJ5jW9FtzxHbz&s{z^}kfNrb%XGA|8|V5R({e&8?sDwHW!(<2Wqvc%*YA^tgDIMsRZ zN9+er!kH}I)jQGFdOJ3DarQWQi1JGv^Ofbw_Gnk6OPyRZhUf6a4Hu3h%`9e$X4V5y z0y6D5tyzuSC$4x$Q;J$OWt?FhC7^o5sE`!cZ8tSW4XH%I`#UR*Xo9@%@w?PkTSmwb zU3X~A2lr6l5ryj4Vz=2rmpeF!^oF87vzS=Jn_Nz8vr=m0O0Hl$YZi#XHy38F44ZK+ zJ-Zb2II~#OVTU-Ck$VxK7RlwJ<#hO;HVZ2yDe~GpUEA~9+;CKleuE?311m2}2HwNd zln>wSsdO>emn!x#Fnh2c%xC{9&U+(7oWg${^U&AN|AnJ}e+vKGG0U4;>Hiar{@XF% z|JO19QkLdZCsm-Nh1fC02&yuyRZo@7@gDd4YV9VHa-&K56_nwcyT4IhH1YJdE@Jl%@Tg2LhU9E8r;#qu~*+6IcEHRaUp1zYBm904m#f znabLgsvE&fVeN=Otii{L--MCh^GxU&$n@wVDi9I)eQQBkLPbMGOGf4S6%!ESli~wB zWDItjn81ACPmfZg}^qz;ZkP(PB z56?FUGm1qHs#|rIX`KX0^ZE|7n%Y5piPdw|QGz{5P4bKe2-$E!$3!x%p<0&e3~i!e zsTM4x&`?N)y@z(b1q5WRO;r!9a;0ILLM>N;&V)63V#qf$j?Ng(?Kh#Nr@E58rP%10 z2X-nxS*m`p1(KmiKeYu}vjFLlIfJ*LM6W1OX)I*V_$j23)$mXF~O#-8O&6ztN6Hy*;heH|rqgis0FC@5r>R3biQI|gjKyvj zC)v~w>A6~jDGc0jRF`@mLFA$os}uk5fY$uOD$`{j`I z370zndPaZW8jGj$O177mTdoPH0^UQ)MxuoW=Q=`zG%l)`PW@jRru zw`>DR6Cp^!PGzQ(0(bt?M*j}}pBwJ*vHp$tk=ScDK+=nS#j4aVZ5(mIIa&VDa4X;s zY4A_1_?Ui}{5t8RjmMSWVlD*^O2S&}sRm5%N;2aJEAOx1VSVdsid;n`)^x+@+)$8v_^s7~Uz8Gp9UWF`yQBR|E(_4GEn5Al+@& zX}7VK&xgt zO7TCWB{U>Ge}@DI1}QjmRr2g*M29Zx#9Fg-QuMUSwE(v|OAy3%oQ%+N^af zyFmSYyKKVN@wnoQb&C)Kb<}D-vPwLI({PP%Ux$%7u2hA!hDNEe=Wf+0^%^A5@ zr2o47TvMOG)^3ZwyHKqn2{GO{;I$B`k5t`+1yyC(iL!y% zZ^ha+VkE6grOHJq`DOrRxaX!W$w;4sXZ%T;RZw95a+4O_&j1)!qDsY|4;=SM__2r zW-T4U6-2Wsk*w#Km}c{?nEv&&REW4)dt9^=paL=9 z`)m;!yJ&=3-S?iQxp?<}`a=e5n+(!mLY0`|)`5j7&&n^0IwMw+P&buw8&A#}8e7gYAAz>yaQe2^#7XXIsdi zS`GAf@JJxmJcmaM74n^@I0&W)6c!2CTjj;3@s3ncjDR$bmZo)qDJX>rF-GO;7r6t? zRon*hb5rE#4)zKiy50v0@{^KB&25f|L&mc(#8%rvv59CC`x%g=DlT*1_1C7hZ7PkV zysz88`!53(|9<=bj>!CfY6kzvDJh@Yf2rVqTGkF*H9-3FKomsCkVXgc9V65g$jm6@ z9`{<|fXvlMHXbKB$j3taJ|S;f3&OnxMzZ)#@UYL)I5|9>DZewnU4Q~ju3O;LRTHM9 z+h4k7*-vt|KJRu{b^@AoHNvOMtOE%V_0bY$C7zBT(BoZzO79A`H>IgwUuB96`o*rq2*hc<*l$ z4o1Z!kpg(p3aS`UG)K*rM9fp!A_s!$B(alg$O=pjKTpr?j~E$YXBIITaQk8nQFOFO zsB~1>7dAfK0Di$b=OsOA8W4B3-gmL$#w^=WB|CQZc!cn3SEM}WUqgKNGQtkg;0jsf zvg=cBHS&3CZai3$@Pbtm_mTE2h`~~&hjT6YW`rk* z^pYnPjDl#7Lh`nZtoenGMzc&~7$@-qE@k0zZEOQus_Jf+E zpX$0Gw-j`ccRy9*iwDB;^kvM*&#yUVnU9-qv`EMS9IFrC z1si7R8(7gryZp*|mkti#?t*RXSM7U8CZV0fUuA|{wNdW^dXR1rpyBKypk?jQgK#BX z19!w(K04L`saR5z%j znZA#5>R)MrB$0x~kuHRCN&*wC?kVIz!STZ@v z#uxy_XJkNUqNML+U=X_>+v-_2%FRiQ%P)l=&cS! zpYwJLC`lZ(tH1{-FDrF`{!!Hv127mJuL$Jr%lWS*u|H0 z`z0EW;qG;hClp%_3g955Ua6b!sVMkVdEtctILb0%WePjy3U#B1tIkRu=X|W%Bz6Ql2Ps8?0 zUE!hG!cVZQsOY^d#pnRU#?ALpM>|#&>NZ*RNaeA46>v1~dj#Hbp|T(r{(Q zLc=c!G}jtfx=rzhriCP;6mLmWhfpjKh%|p^Cf&I#f%eLteU8}xww9Ch$|Y_SJinr6 zoTrYTwJf8Oe2zwmdZz#>MxVP&?!&DR6KMtIS6~Dk#_p%7aQV_wzWVGLx9xaUv7YY* zG)p}^*Sm}5=v-pZ*e(P=5Iif?ElXB(YG5;EH7=8P@d*!ci9si~eoHI>)kqJbKA^PU zt1?7ZKa|D3wc52WP8Qtfc%!{Un80k9)End|)!WGUjp;O=>xSK`zro1|<8J&TSy*xm z1JQnoIZQ3;l$k|863s$U?+uDPA=w^;JwDkSWL``nnU^OdguYnONY>8ydWMr1&$qNL z(#;IR@l^|=J>+DGH>w$1Y)*GGFo{NQdv?zDEXqG@WBAy9jK9TI!%)97V$||%boVb# zbtJ!^`2CAd`TdKY{=a(_>c0t1{s&I=&vBjvIZbhluek|*E~j-NjC!xR0W)=3{p{5T zMtm_5JVWRnv93W|tOZjN`kGeoC2hl3{JN<{K5vEVi5A$ILLHQYiA7f9i4FHJPW1rT z#ol%!(|^ODbT!vq8U&xoYQmzTUwk1-S?d{Ckg-L8W3O$C^{SgFg4O6g{++*Be*(C7 zm&a@x2YOP#u4fW4V+@pd$E*v%^s!%;wKbZIzoMvKge|nFB-`?5@YqyTL`cHb?wkoX z4I2!wY{J_Zz3E-GEM3cr$Es1XR2byU?d57szSAvr1uSBV7*xPSuqK*4BfinQG=8r{ z^7YZ|k4}m)w;VL07#HgE!fROZ$_l7=Z)M?;{P;wI7|rBcbW9HCDy?1+Et+ykI>+>( zoUuI>l`HYU1Ham?s&^iW-tGL6ci|q(7-iz+%cyug8~69Cy{gS)x^gLr_5fKsl*`p7 zbHrFK7e5HbtoK3_78_pRiPjnkPq3!e3Xs#_%8m)pQ!vx$@>2xe*MQxO@y4MI3h3V=Xmpm)@q{D z$#PhV^}HsiWvGnJ4^Tau>Xc%zch(E&)4nx{vEMnKr7t8oU9ek0p_$8^xe6^HMw!N$ zu3nvomJ{Ay+<;YqgWpS=&UdALwKkz-`?BIZdXCX6Xp2$aDnmsWz4>_W#O$y|INASn zTsf|PKx>1ZMS()7p3^%uvB5pV=9V!i86@5|%ThnS;Ll6K841TbGo?HW+)*VyYC$A1 zf-4xv7ZWfwa;P&F`)!$LU{_PKpbJ*M^*%Fl)#eHwyT3;hyb;~OpV{N2=AEwr5K@TG z*n8uhO}2Aji_dj1b2K&x*m9XlkQTf+qeyq}94f6zYSvpQ6+c|eZY}6#@`C3zSg6$AyDr`AjJX^k>cf8@s?d%1oY>%pPtVmq%O z46R?kd(gpB<2Oe`8F?FQTS=^%BI!Pn-xljRNiNi>v@sq1Ah&HyK z_urcHtA&9}) z<3f?g<4=j!@1vw_JP5=_5{_w#M8Zcz`NPTJ$9fdl0&lkOUbKvs^JSl(4PTB&QH+4k z%_Fm#e@{$Dj6W^2%uViWN208}M$lS{L;awT@A=b)IdbeOE^&EtrWJd~f)^kZeo&8xo$`vdfG5`9i)%%O0RE5@<#Zb~49TP{MaC#$`$jOM3({oGW> zP+XBui1v7Jvp!Q0t-mh8tPr{Xy!RWfE|K|(oIv#(b`KQ<4J==>A8^{__l|v?8-H`1*J!^D&hkRAu^gvT%KTPXoxfWW7LL$ww9Av58yLMdn7W znA4z&s~Iy0^@~W)-ZShmh5fk^ou8^dZr|rDBb#o81$yazlpRl;{aMZ#%HL=GBX4#1 zAJf!0f8!wjEFaj}>HZ_5m!SNrjA@MKW#p`@gNBPoNSwn{O8lLPtf4^*S9G-AiVxRm zcFjyL$WSAYg5ANnJ6g?sS}o$*Q%J_(-7vNe z!bxi3(eGoLBPzI<;|D{=Y9|AvB!c-P`yl~ zTYg$@_}lbk9LSYhHZOR=pH#!6NVMEYxAe5!D7TQb_Ig@yx4FqleHuvilE1W&Z^dQr zLL*PnOUvsRO$~2Yq0<71$Ixo+*{qA$gU&QM)cGBIqzuR292RD#n)RN#WM)#%Bv_@$ zEgcK(qzZ3o#(zOBdIp#Mgk3^o_ZH!As(AMg6v?6&N`Jv)*N2vtj#OLTJupe-gyxr; zk;tN{ZD@1Phb*oXePlSqTXu?${8YiPSJv3%CvF(UdT~NrP?_WG^G;XjlQaV3BT%ZiCEU$sI3Q3D>eQqvc@IWIaxgJCBrnq zb&a2UR^k4BrNrx}sdTDoh^+N>ehSW|2KF&kGUnu7Mf{K|<1)tfUU4c%veQ@_K88Vf z*>5~yhMGef&0@Em*pv&MON)*9g|Zx}`#Z`*;MUi~UkJNW3&2t)G;8%%4Yb(hL51TE zR)SMe;ZiE1%wP;fEF0~3Cuq#}yH?-Lp^M2zZSBg>T)wxlIk9=3>qil%A6v|z*X{eC zysb4$Rza=F=`pVQxc4I z;du;neUmzAGTnl18>Uk;`=~DGG(9{=6OV@)2_?yQ6c`Im_V!Z57 zDac{5h2YZHMRHy2GiXp01vwc<=LY6X|4j*JO3qle<_0&EI|QAnlKfiyP)?)vxn47+ zTu@USyo~-VumVyb(qJx+#qa%UO>5-5cUI1=Y^OJ5m$TOs@GHd$MV}%RjAaRYUp$%P z7q_F3fb6asrFG%Bcx&zD5nKuo*#WnJhRZ(b|x{er_QA^{>AA}-rD?_7^uVbJI(U72I6AK{)Ov(}n3;MNI zQ-oJf6JkSFt=msQ2ZWR*^3@M9om9?dM&2e74q!&aDdUXXMgM0AhQi!x9RgGvLqpXC z6{N`>Fhgm!129cV*%C*Ji_%HPNNCHl*BioRvm@B)ByuUei`t{8Y>KlIG1YO0BDk zdrXd)T1OXnbe?(lbNHl?9>kaxI@)-a`?lUG@@PHtUc`ZQX@rsEuH1D|XEtXhTW2EV z%C-T(%=dVK+mexQj<0F|^*1v!U*l{4w-%V0ev!sb4lL^ijGWb5zHgc+$1S4HRUGT( z+QbWjjm}iRE@+M;qc2RpzO%s0w}>p@jB^a1-}?2HPZSIP-<-i0P{-J~>&rI|Z=f|l zS|V{X0NpqgpFMpU7X8zT^R;9{JqfF^{shG za>{hHgSFW??E(M}JeU+y>q#KQZml{QE(eOPOfRiuLu|~Su*qhnK|K2&Dqnz1r7J#u zS)9ap*J1M&1<6d)My)D&QCm@}Xl#c-xNc?@U8&x`t^Jk01X4p2dHw2z6Gr||b=dz@ z1N)NM)^#v6{s*1VjNpU-;Drj%epCO1;60m4YQ$X|(;F zBR*MCh6e6N>nZlpW#i&opb1DZa52J3B4hDIGOW?m_M% zUk%113aIA-`9g_oR&RA%iUuFo%TS@}ufV4W&gSSM*)ji#1+T6M=at%Km>XFC3-2Xv zVUzb7rB=Y8%qg!*{4%35Ym)#TQe0~CAEBfn)M2B>f^2Y_JokUji%&Zj5I24irf>gJ z4*Ykf{CCyTKPZ)quARN%zY*nt7*TORUbp~lV~hxQh5^rBeu=*C({b>yRB+(W4{=O>WeGzxKIn1bVhSCcNic~ZX%vFLUTgdd&gk+epB={6r66`q@&yY)B$ zR-Ez2>R*8BeQx|%#4}vg6JY{e{N;s4AK%jVyujLecb`!mPG5+H4ZN3m$<|FH`)1GR zUP`d}19T>z`c10Y2>Eq7$csQ<2m(s{vlF{lVepnvF2}+NTki+ii}^Q0eh}mgP6zR9 z6w%V*2!~&6NGC)`WAP(VIE~5E2^oJ=L1v7an#W)a@70wq|E~GFW!DkSE;;#k4@~6{1GTqgV$~O ze_3_yU&!+RL!kW6orH>i4J=WqE>%$=`uI+-3^XkGH^wilh>;h_ME4Qu0D_s#i`#*# zCuvGKmE2=?-We}x%XacarR#UTX4P>y^OH?2CEH(ojVl=+OpIl;KA&cBe*oD)M+P|C z_V%E1)!QI6GCYs<RO*+VDO*$(#A1yHR;=nLJdN?!ES7ML{>Mjrah1NfEn@IQ+fX&lef?N?p6)LT%slk1GVC=%HeZn>0xwYlY- zjN0`U6jp92=eLrdMsDDcN7#&l1D&NGmb{ez^y9I0ua>!Pq-Wf$Y$lh7^q{UiW7ea5 z{w@E!e{)~{OHsLIWeQ?VqJ10QAIQS^CjJtkcX7tlHD~Us=4RQC)dLrRqz_>_2Uo&%O~;$NUMUENB!1QB5>n?} zp%l^-6^81*nkXq~36Fly<;@D^xwgF3W$zm4T{fCay@|3WudXgZB&%md4DfhM3jO}R z(um4xBYmt*FbxygqCijcU3QMD>vm%9EuXF>Z$$O@3*vW0MP*M!ew`zSX8P91(Zr)V zFN4P0{eHRFObjVTkwgPBvUanXYo_}>R`O(tmFT6Z{GOXyQ8Z#_gC;>oq@PUj#NZZo z;Zh`p2&%F7JXoWJ?n-W%mHaYUbD4}j24&9(ET2wbrzWg`0PRKUwtVxbf*WS}c5f0^ z*QB-x>Hv+YoIOg`#`tAfCHp0V+hz2ia>qvWOSfvuSB$^u&E)vGR`<>vGAD^sPXrq} zxf8gtm0D?!5{$gYH?sDDng^y?mM%MX(ZBcRI))TNs-ipN9bJ@m_zJvk4G|E33 z7{FTgPag(z23?m6m+a>BiR9GUgq4HuIA_i0FK^GKHktK5eLeuCmL~aOV;L7df&H9ETKU3jmw**325WHovtqeu4 zg&|bZ{(YASHZ%p5j@C#Ivq#`{6P1uL2F+VE-PRzVu4dP4?v_b{#s~FMaI@FZgU?EU z+1aNoss+#U#)0~R`RUKr87;bJtj({K$OVDESr<2cazA2l@2 zMi*=L2V?Ed-XAY7SY6oYz$_}`yG#q1Yh>A>_W*QFr#ijeoPlfNC5AVlIdWe*1Oyj9 zfAe={3CIPFOTGK!tQ?`(KY|7|#%DV)4&^BL4c^4I9aHxjLZ7BFl-zx4LYbSm zV@qMOdIVy$6*S;4_M~f3x5>zM2I))w978g5M3852rE6a~934dh+IRR#HtL&tZ?0iI zRF{AXmI3GH)i)mi7br&Cl%e(cl)Y$1UoKhU3!TS-@;Slr9f4k{SLXM8rMZDhO72=|3aeu4t8{dp zR^uinG>WTCz-$(-IF@u~hERg0kEa-cRrI(l&=n%KDz#>-N*Dh49G6PwmNd8}+Gqrh z4VV7K4F+9N1oe=X?nmrfN_<)J#&!6Gn-m4L+{6ef^LAYsycN7Ts~~OCm>7AXuR4RV z{IeiB(ak=Eo;Ud8;H~u^4rmQ{r^2C@m{t z2*b5Q`}*iYG3i|S>br{bP}F3wXmvp%M{kMr@S_nNjvmW8Ipf(|}q@k4`lc4XMr zuTP$*jr3QWsNcapU|h5A2E?)qHdm3clkJTSly< zUjZoD3q@B_uPV$h*acf)v?#VQg+_yJ;JHmURNsL18TjyV_2KCg74?ZMHXL=~s-;+G zwP%%-G@cVpJ$9X~pKB?Gaxqtu%BD_rpxag)RFxdPGMlLGb!J&(Tu*lp2EGGNaWtH{ zGZ5Q4ve37CwMm-`>8J(P9@e9*P} z+BsyKLByG+#54%|cd;dvhC)amFh4Vp)%IgT|N zVU?k^Uk>KN+P}-k-Y0K#o15vXw%=(;_L@33TVm+bm)?mgF3#qko zOo#Y#yu8m~z&0zPp}ww;va`U~dEefL_*2D|1DxcNct5`2X!Sa@kN{$$i?xk%O0%@C zE1GY_EUMSNsc&GOQbzSK-m&jjP@B)7OYU}wkMWalU9pnKPRe=8*CZ9SjZ=>puq(T- zC_H^u46n5$h-%#ah|Y-Ufnu2uP~ayDmV%x(goC24f7h^?>v?@~FDjsRiCO0WGkOOj zO$1ofiK8%?P0PWmEOMjsXMKA{;|v$|YoSD@A$jpc(#uvxp9DwpE$p+}RG0IU7Bn|l z9S|o;aTWAiaFzm&7cz@dnjlCIi*(qEGzPU`%Ii&wM6C6oHot|>Lwi!O0~}Y+TIwkW zSzB39S-Xu#Mk{$1ftlmaBHzB-YL_Kv1VR~du6RUy5AG57;>67I-D5a_35~- zP4tNEsF~2Rc5~Xn&th)Koyu#yfiQ{<0(*H(^1YkFowV~3Y;2Q{ zC1hJbSn|G2hyuQ_GQE!2qi&fXLc#Um?nlGYY1E5iizf&g-rrKKuBKSSYu`b^+ZJ>6nwZwlF${ho4%YvBrJ_vnZR z*3HuuxX7b<=!jG0#xN%g@5DznmbiPFbA=)&tp@b=u3dT(^lS=%Gw}^3FnCS8&iu>* z@avz9rB~|5)_gdAj9&lfy2O2gHQW{VWvtTnDsy}GJ|g?zTAL;6hRGtS!9P&^ji`ly z=HSLnkn~RI1_2_NS(dU>NV;o8^r#ptuE?hl?NwnEoYoIlP+kTaq5z(I4y(5aW%ms2 zyLLUz^#c0`K0|EwlE{5E@UZ>Iax327;PW5DpL~v{7XP(!w_gR^MPuHP>v3E}n%+%~ zLN$%jr`ph+0^9(MMJU$NfYKzK5Hf;LAEQb*kHqwBdd%fuxIqKnOb?1d+!ny}jEayj zfLKz^CI-a>TofaAhRO6emWd%vG_An0g(M&Hj{A`_!swv=TH@)(lX0y1vBUk+dxCS6 z^YA2rImWK;BFB61>vP8Na*+{_KW-r{YCpp4tG)K)`oWJowC-=RjKSt zshIaU(KzNCbX+B^t7Jjv{==z6NB3TuL`VPLs>E*^*NkCjjyvbO$F*0g$Bx1FOLq|( zYhG@hn|Zkx&==v0_JOa@Y8l^1e|`2DCeE7HHDMSQZ9i!3l`YPi=J`<0E#)J?{c~im zRifjvU-UC#xRd@FmiTjcPni5&rGQ!CB}D9Vzu>d*Qy1-HQQUe)U5l6Rn*^!WHoO%! zvTbE@XUS9`=`}>um0jhWcpl%a9)F5(d>90^C$9pX>h=%C+!e^0SQ4}3xtl5}r2A46 zIE7psSiys$JO6A)znmqFR|yIxgy~h|*lH7Z7PF%e6PlR~W6Bdm1(Xm$V*odN%y55c zpa>fZHux}qen?-Y9WG>^b zW^-gtLF?-Ry}8zB2cmJ7*3Jj&Pj~FPnQVYb3EgRm8s!SznW#je`;rh(GmseSgm8v>!c9c(+L~_k3<( zSalgVi0Ih8R*}1m6w(je7c;*j{aJTNzTq^kz%p$9Ci^5NO_(fUM5zKSxZflU1ieHV zeL68*wv~_Im$0>A-C63i^1Pz7zpsLpQ0A^Tofiy5{{ZnTK;96d{_L(@8R;hyldB@_ zInc(p->vq2W{nxulq$EjhHu-L(_IkqQcjRd+bnLg z1}0PjK)iaif#M#SY;zx=Tu+wCm699qO|_ew}s|D@z~f)&r0au?0WpU!{BgjNh6V)WcYuI5hjccaV4B+MgQts05@8yjoll<1U<(~#&LFf~+LQ?D&XAhTS*ns3a(V=a zzAosV<8QbP;ybLWF@%1e_ufRFNYa83vIP+Kd%E3$Cl5DH6oU6#^Uzcj#`&morqC=y z-ZVuVO#lO1-yTgISR5;{EJTT31@S_NQG_1iVxQXu7H`zVl$PR`{H@UTY#*@}r(g@Q zT9;B?D3hedWIKjHD=||T2BVwITK+UY?he+D5u)s}+BUvBS+&htxvg4NI-JXJy(>1i zuab6uXlaumf_fi2W3}R1X()`OLnxJ^#cZEbH9taQh?U~moIupkyg=5Gv0W=Mh(AsP zLaTyp5HUzU9Bd%t3q^Zh`+6zrbYI3)AS%*#$#C!`$Ed)^NiWGY%8Ti~_BD1L7+5y2 zjI{~0c-J|Q@M z_B>v^&sTk$WRy`$*wcg>R95q)k!wJsBzQMiY~PuQsn+-{@k~nmHiD81H$# zI@-3oej5Zv%`nSq*gm>$Eh@dq1A=!JSkx2g2P7vCR+AB;F4>Pa>q;>DJ-UNm_w;ZZ z83JEZ=&#qcS+C2J=IK_b#mbUL!a2qil@F|na;%qy;CQ8H3ss1SzS7?3kPuL?drHmmalrOH__6T*A+jN`hgNdqDZqneh_*0SmPYwhd$L&@xR z>Md|E=`!OZ+^tSaYE$;h)8l1>Z~UG<)fYK2fE4)(*kY*5(yv?g9whAe8*ff9qRca- zz`@0#h`&3g<(sY&Db1`{{ovDp)+syJ{b)~4=ALi|y2Gn`<{QrP;1VgxEtV(LESFNb z4eEm`nDpxHl`!hn6xM1u@mGnjdh)){bsf3y@kJe(j0(!BO~jeU{GZOQ1FpvX{~twU zR8pcP+G%GbMSJfd+UvCU)Ic<7NTt2Cw4|k!k|LE-sgQ(*_L4$G{@*8D=Q+y^wo9 z**Q<~pEOYmiW#db`+qJ@8 zI@@k?c=NL@Utcdh%PxVr-!Ac`n|A12Qk#3KnUr10#?D5ocHNv9pAdZ`*J=KFammpqXzr%&fX@R%rN8R;$yQR8o@8pukw}-tYWg`9VwJ%cn%iQJnycMH7!lk|=n|wA;|48ip zS{r`xXdX$3W`cH_Yocq5+sP7D$+YX7%JJ2TW|c-6$*KG&N!r=nOiuXkDot^7f6`dH z$*%8=&gV2Ax;`hZonqI=dI~nNWwf(iQQelQ9O`k=(ItAy`}9K4p^P%BuIzH|*dtfs z_qWk7ppVUNFBPOCisPdE6hT2gD%!F8^2n7t+FUP=F^-%1JR<797#kXs6Z?>vE|a0m zUbXysQrUIIdK9_U@w)i8;~P(z5l<8KRnnWx-z1kWG0v9TM!ZODuS-@UB+s+`v&6T^ zU46E6eZmjeT>V|no0K%CG$h%ZNcgLap^hJDQ0*SMdtABX@(_QMwZ8u4exrf8{^4Sm zqhbm#E1utVn&OCl;lSbguyH;xNUgf5$H_2iyJ55`HBo|e%BfK13bI+YsF?7VCg)jo zy7->Sp~yePY`NvWTQ@AjIBqqgR)Ll4M<|7p3HWP{ad|fFgClWu-Qi|cZB!3DUwl!o z7LN6N@q{W%u{DMIi!`rx{PE3G=Hll>$*1H}ALe(EeBRi;pwaC7EkQ^+V8?Ke|8h9< zqw8c8yd6@nDB@8ruXYZPERnL;+uo7z+e=~GI;2}4VJ0tMa#OSBNJB;3^OSLEHnU^z zR)$}T*Gd{K_|O)_Nj{&jv0&fz=_q&akJAtLng@mSx8w_L>tAZ!x8Yl}|M>u}M{1fU zLauTyQ$&S2%xN-pg`FV+g6Rd*-ZL#V7jbS=FEDZ;AJ~MATY~bVg5|PZ}sz z>TNRal%y}F>Nm=aXsdWC5y%tGFmCY;f%GL7m7(%Yq#^{`W;}b>VOlCCkSnX8xvd(Z7 z<;l;jdf%r$*gd#LZIE(5{fwfPf6?RHlLx-WFE;9NHIQ1}Sm6H}Bh{NLZCxlRXQ&wC z6Dxf6V924TC()_3q&gSBe(Mj7j(KO8JN7EJy*ePjI-r~Gq!M3Zap1kDO&XF`9X-W7 zO_@^E46L{KdMv5@y$muM+$x4dcbM`IXlQHs#eE;QsiMyxw~_W#Ov_EC3rowOOASErrsiw!Ej<)dWe!io=r{~XCZfJexe$?pHrSlT)5 zoXYPUkx}2yd_uDRniJREl-!fzw^?|oI^+ZHqB-0nS9qolTvN>AIw!G(I*6l%X5pm9 ziatHJjn;Bfl7U@QNBoBssaL)-GqDl}q+=5(ysE{1u3h>vea>Gmk{pG~1lt_(Pv*1Y z7~Zd)(UEX>ceOF|a7VlS-v6krq6Ow)&joA@i@UQ)Fz)28%NvhJKR8AI`YS5&bVD%5 zP9nLj5l?MS?=YJ|$M+lNG_t<@9-uuMW6pbmMJHtBW5dY@C!fToy^$%}&6%zJJZX0F z{0}3w&RXx4sl)!)N^_XQLL=^-elmFIewPcm4*I^f<(b5lC!*?>GR<7~H4BpjjSntf z4Xz139Q=%dv*^gfpl@2M#C-Y#b1{ac=A=D*xAShv1ebdFo1rL~S}7iw z`_{-zWPjPTfcg7MW5t$7bKLp^Spz~rX5s3C*%QiIp7wKIhFsAn88tY0G$Li4ciflT z-*axiit&(^c%`mdh5qIhsZE$(^Ti>mv&I!BqF_wcKExBe{|*86%kJlrX#eJ1z3N_ z8$~y3&XTfnlf5cPX}?&;?pa+mR>7)Lpk%I^EyvLD9myn& zL%zh<79M3Jt9hpnxdk6&AD~TY$tDZ_5@B#yB!H}{AeU40_D$X{$)nyULUKXB0diXsK?5KM~>FSH90%BFz+OLT2K;n@aYHDe7ilxmfMSkxJr zntR(Lyo zRMRT#Nd24}u-Wh&@kma6q|e2rePX?(4=%8a3pOTfnRKOCB>QUZeju7OHPQd}pxbu*>94;~O~_ zlgw$zx6I|9CIS>^x1z!@B7*Y+H%?wo(8ZS69Rj23ifUfTo=|rOES`Ex8GOQtesnx za&mM>pLPG09#YjLREZ=giXO*OtB;V&3+HcUUznif+ZqzMk@7R~NWW_Eb|DK-{)ryZ zWz;@&%7XZnxsVg?1%+CM&yC;hOGg#EBxS{qBy{sCRHalziS1 z>S8bTP^t1p0{8nuDp@^N(fU2bmpI;?GBB9o&Q%L*p$VLG)>`1Qwwk7A59}~Fc<^Dr zT>2{q??q#`s++!N*zQ&9uY}6=KU(>b?b~;rO)J#birA@9r$+Lwy{qjhijGxT*6m44 zLX&oH@4sQA(36+hJA^uuLN*d3bN{Mn=I#ivw<2?HZO*idWR`gZq|8);+hv$=BvnX( zYG$@*^WO@)rE3ShTlIXULXARsow_yY(G6{^D%*xK{VTK;ZA#T8OLI<3qm#KOyPrDt z*>i205B1tvt9e_fL)dSKSVWH`!Q}(7NS5)B!yO!HUS7#f;Pgi_0$R6D#o+&X+a zUM*GCJN?{fLU+GdLQGRMn^t>2wIu4McedR`qFK@czeNhEIXhc|`q@0wz#9uKW#eZU zMTlDjHZuo`vL(JOS56rb&H`;7#%p=AnDC)?IM>lAhWPV;$Byxt(suLJk~0vv+LAnGmhUY2|})LA;|MLmO^Xr;KTP#@VJP zk6++CJsY{7jeSew#<2yFSXJd5Vf(OEr9-FYO-{Wmi*{5l-P-ALQ$#sbF>T^@5=|?& zxNI`Z!&|)E19nF4WV?Pg5M8_K=-u8#f2ON@YF|&8OL0;&-64ua=9!*DGL$;I%u2)3 znv?0x7R{H!!)g)^*&k3Te)r*QSaG9$-Tsgo>z#LON)6>@WN1C6M1H162wrxUEF-%8 zWs7ELGyV5tUu!Ko>`%73L|qMj-IPA4ZtjpCm|#j~vpXls!Ykt0p5D=L`<5b?O5FkB zP3#9Wb^WJactq8(8wrF@ic@^rw8d_lbDySwtf`Z>lGS47Fn7R@!v~3P#(J5$9M|~t z_|~nee)Hqh>^YY5ZPN<*YL*9c?zfl-5$`V9KzGbO&;HD|vlKk^alSWM2PsI-*76K8 z)E*oZGobu%f_L$*WY1u?FVTqyXtEt0{iMs0le+EKll)bfR_}e-a_!nizi_pUexZd# zHT)`VR&RMtY%V;0G_LH|@F<>$at{Unl`NwBjaozz0x9S;2c|)ma*?K)qozWJIs*0Q z3(~Buea=|EpuC}wT25tYr*&euqN;5ybk2}>@JPJM8+*|j$21@2{i#Cj3R!F)3&$?A zmVX{i)zh_t4I%E-h(l-L2d&l*NDFK23PW zb}6G<@pC7lH9OkSG?!t!#O8Oh@n@>d8^!1!t3)pzT)T7SnL^3T zvtZ9%=}GO6YgA63_xb!r!r!D|vl*y}h*x)Cq9?0*Cwb-JA-zfUheN!LA(bjsOeN+# z3{!nK&XepAI_*BYy3gV0mR2W0>s$lz(f}$UMU8u$W9LI@g}lGYS8aFs_GZew)WJyN ziT&|ZVNnaW$|6R#LWb;osprPl!#f8{7uAG_lTIJMreyK3Bx%pj%Fo6-zcF5YIp~z$ zFg#Mke?Pi$<}sx^RbJ$E75`Mvo}($=Q;!(SEH74NPMvrY-nredt*129>cX8bd!03= z%PVGDGI(qM5;k`}KVGOZ^RXSGLvlNb?zPs5?RKv)N#h!MWSFh{&hSU?P>($7j>h zH?au@_Z+X16vn#Ctzx#T>8**8623~p_m(AFD@pscix@l;+)4RTMpaY7(uC#Y<>qzn z8IYw=zhn@OH*gjfjoAPC!ZoqP{l0>_A{&b1!qIu^uPK6OOGU?yy)QHf{!Vvof4U4^ zgNsuI8+-q!OQe2>_Q~9oV(mpEp7rkdk^aD|FS>WH>^|6kUe<&* z>c}M)=jl_rY5`w*xV{+j9lDVEYUWGc*9$j}j@Ct+@C2FgoTcqlk7Quue8=_9X!mlA z?tGDp(T{mI-o#~E9BSMq zpRcZ=m7O8J^VYXH5v7`zS2C4avGtc|#6`hvXH;e{DhnvNE|pZIy?P~NWz{irE&bhn z;g3DXOyazl5f;MgTqUWZZQOUPhjp_R6JD)kip?u)*mMZr!vxX~0PWMbFeDK<$s}aNL(f3Bo;rfU7wE}8s5#C#rcHJ>$ zh!JjW)1a6--?jfxvz4@SM3x1M=R3-=?;Xda2KootZ6Er17?II38E`O`>qyhuO)oBO zS_=6=Yq0t{9Ch$))K89g(PUbhaN6gujgb&s=O2#>r9oe5otx$J90y=-^a=#g?~{Ml+h5p?le+f^dS zWk!0hX}kl4OfzhQ4t6z&>lO#F<=5WB4V;aG?}Pq(QIe4-bXO9iEXnF{`1kIVn8BHL zqN?`Z6e?K~YUGVcCE?}@3L2K0HVU?-cPyzwWyez<4V%f5?A(!F^r%ewN6MpG_j4tC zPVcvJw)8PFQ?SwCVh#`vrljLJUcyy8_`N&xy>?YQeTJaG%h2~J-NWtQYX_q-7LQyLH`x`2& zF6CM$=1bdVzh|NQ_|8i{v`o6SEVaha_|k^~zJ1ro#CT$Attv$Q+J$ahIHnhab_sS- zjZk z*lTTbK9^^qE<5|Zrv1>2b7~-~blcH%wsf`$4R*POhM}QD@uY35CW--{`e+k|$z~tO zQ%ehIfUS4gil>@%D76x^w4WR@zjBF0@eY;NOtorUUUVlZ&qb5ID4|15JToQwT)>T-e5PYz@6JU0co}rtqCq@jvY(EkJ2dYR`&VWW3HCt_rsMNG zXsH%qwB*s;sjTB)Q5_Divb3~RXk?$>(%zj+GD%BxYxV5CaGQC-`r1;n&SEw z)tqmOh7+Bjb7dR{KJ!3r4n$svFvUZk%TuL|sseV~INCDkAmOe|M}|M~~tw zrOHg=%iEH4NLS=1tfI^4CUP8!->Z*<%+0?zg{?2ygtr0uQ#BVrIZ??R7w#33&Uqr>=UC-X1o1@+p zM&C$mwXKL^+F5NGapYNNc$H)59a;Gv64`^h9Q#s}=(TxbS+;SV(Q=UHr{U-!2@LMm zX%aK+xto@}TVvj+k*{|5RF3W?j+#NG-PtL-8pio4dA%PG>;8Q4PTqn~G+O73_NbG= z28Du~R(|JPyciAN7z|G@CKtAay*}=Kj&xsI*y= zR@?F*Uf!!=xHYY$<@(tUp*DALdD=c3Ju-GUi58Cuy?nk$y~>uG@LWn!Dw!Sl^cm4p04#KG0&i z;Z;-pA^%*c3H7DK`hz5Dp92~sOXWMpVuL+mWDYU=Y&qXVl(g^VqJW_9rCo=wI@jJB znksukqrBn!OkAv(@xfE+JsG~vC3nR8I(8+K7p`0}U)QW9AEDDctMhU%F|;=VA6Dd8y&!0nTrxeCCeVdfIN5&NH~Ui*4Iiy8Dj9 zlf^-)s%K7Jbnyk3p4}7idMAULIZO3<%+MlG^a%bu6xJd7Nl5Zt7kD8m8`;5bb zg|+<;e&N~{zjK7{;MmbkZrP6>4SkBtpDmDmtNbPE@N_-c=S`e>F*1s-`w-nWy_OWQ zkre|r#~=wdSBnlY8Ku7O2k5?=GFb$!sE@x*qMfK8;-kA;amsIESkvbHW!@s4GZFsf zUIDj)S8hn?WWKlDZ~bF#aB-jNc%h|D|009M-qr5m{NypWd7VCkuvytTJNKH~FID!N z9`NgVUiLQk%XiBDq330zT;)RxL?@D2NOlu<(iu5>?kGQ5C`YXJB50;&j$~TA*TGZt ze#FOB_qYABCFrXOwh2kAt7TGpIgTjRW--%E8IP z!ye6J<7nyht9m*Y`3BrhN)gn#r6rxHW>if)%J*JKNQlp9tQQ;G35i!c$w4_yZ;D@v z#?zeeCr8ilYLJqu_Vl#3AAM@Gb~EXKiDVw%opjlcT-rj0Bs=x=H=Uqyy41C_6b)|5 zje0MU85?lmhWk^euM9U9Vir3a^{i zpLBMXLDDOJFy;6)WwDDVgD z`Oi00Bvch-Wwf+;Rb*G6Kt|g@v=s&3YXPRauziXYh7bg>|3kt2nkSaE;nYRoYtZIp zLf_a1BkK75jo3-==|mn-!FJ*t@LV++ES8gxe0H664Xs z0pg0bvtcC;i*jLr*eYXy0R|8PgkFwG$Pu8ocl>?sNs=nGhm$eaGL&>VHK1FL3^(7fAIqG@@Ie zqY7VvQ^8@20W7v(AnveZIs7Z@L#@SxC7s%-(g)PE2c#gl89sv(HrTLoV1^hUQmi5; z+dL3<6*LkW!~cEu>_8&HC+pF_Pf=jD4w@Gbp%C*U1>#)?t%GCgli|h$qdkx@)((B* zz=RZem0Y=M5@f0>FcZQ3w~^q}-Can`HHl?C+&x?|-!zO{m*FWbz3w2+JoHc~S{U*Q zIK_)gm$-2%R7R>&!S_4C58TvMB1jwzQ^SY4n976%AGt0;tqQ#73NS3()Pgf6oZv zX($i1+TlIG){g-!Jeh__B4MG1s(RS}z8@T0=tPsY_`5)8=%RR5@bACRtTYl!!O_j# z6kLKW1>6x_(~R5DMhW8|iUAG}faLv)2R<#!AR#gJh+N&mZZ2KfRPXSl450j*cKAnLwu94{LI}JQ5aavWz|20qyvENeLS+u31s|x_I8vw$ia99@!xF+!-+)ca1f0_gZbaw`r18!jkeI(NFQVOz&CfXDni@?`F zo1Z=J=)h_Ed1Vxe9VW56(fa}^%2tIXEaCKhuQVhk()aN=*dBZ5 z7k&WtHE1d;Ou>yANLWa41vRXj7O++7@m+Ml2$Tac`>|CTm5GG>QzJuRg~t&fJE5pK zkE4NOy#~Ml7vqsGL4sn$C~j?z4Hp>i0K>5tl-MD^dv=R`3P7mYhctn8YS5fB1BH`?ecC;`xvq%4m5Edy9S69$Cto;qQF`}QJ zqkRt;Hz0H;&`f|4Ltz*nS&rf+nY3}*wg^PxBE`<6KCJ{u)|MCuWKu8d^YUT<_9;m9 zaFaZGN&pt4!5ANLmEDj@t`XC}O`<}fv_S#Eg<%jB$`X`l*j3$s&c(9Uvp})%d-EK2 zq;lku9^(bRq)3B8L9a{y``LEnx~F;c2e``ab+MFWH-!oMuQ05EyVNn7*&e^hZ@V zYCuSerw=~-I0~Gp82GIKObV%0LfEE0Hum0YV@8W#?-InVqz0FY*Tx_md_XWmD{6nA zS411|hZ;gjxY6%SQ$7{~ih;5nAi#LyXbGT0{m-w0PT}_qm4vxD+6_lkLmHCSJrS=2 z2qJ;4B`_#}E$9i5{OO3*UD1lBo~Fk%e@&@@)zx2QxCS^7oh#i6M;GXse}WlCFfzn= zCK&z)IvqzhQ%m&!LI#xqMRUg=H38Y2foC1VpaR7U%l|<3ztTvWj6K!_6q|vQ!4snf z8vz=4LirQxNLPR)i6NJf83dB70vfn0kZ=*8z(}|lAOE}s2RkwD$w}oc0%{@wL~w6b z;wC`yYl9#LK5h|Yqvjq$@ADi513O8WF?7Ia7jL%&*Anb+_|wW-=zlNaO#^)KfDdkb zGXVm8e;isGT$LsZ#s#MC>aI?nHvgR($5a|bQ*t&zbr~qk4^Roh#Iia}faX_k3&_l! z94$bl2r4_I+8}(=_l_H0f!M48c^Gc_`lAH!5h(_5j|I6`esXWi7!YD3qTl@?LVyIr z<`^GYa15LwNnRXbPK7SOJ_*7EZV-7f0whRD@P`USbM>u12`i989h7ctFoSHCAVBlG zB1IYj(oaEeGVut|4>aKpEw252{ACDWVkiaUBa46_k_Bq{_c#5S*&sL4Y{%{hTjdC_ z{K|Ncj z!%R?+?{2WPQ;Ps$HvsY=+;`EY1Sl{XkMWU30ME=o!^y60AuwCc%7ZhXXA6W$O6wp=t zp~=wN2ny&2Uat`c5@5lMIXSc=h&OWy=q7XgAqScImcC37H0Pg$=u6LBCP0Ra1{}gb zDr2^DT%HE_G>DK63L$`xp(l)wEXQ#BYt^O)9?t)WhB#Dn~KqDe>v>w&W{I6iDWH)>r;&Vwa;bo9(9)Z;) zc;X~3To3=(KxADJB1{6wv%?~~qm90ksjjex+Ao1VZ&@mWcuVBRtSV(uO z>>{R9AddQhd~oBu0IL>wM?h>k%NVgUg2@k{|cc>W@KQD+E&nk}&&(H?BwWXUOh}>B+y1f{$-+ z{N1-iz!%c?VY_rAn3f`J7RVo|>L2EKD6ZffvFFYQe^Q!ut9( znT8=1KBFABWYU6GSary-UD*W`<%CJFawTr7hzL(iR{|-X0x96Jq6NlDgb6U(@5UmC zqqK*DQSM~3JwGtJ0>bRJGyg;&fM=slX4?-xT|khFAXu6IHv$}sAdsP_&j}380>8Qs z0ui2z%NN%pKo}Zn1&HE&R>_WcAogtNu_=hZu1B%9A5B3^P521EfDYK3hNl#N??BEz z0X(T2(1^pLLxz;-Ukokd=;452dE5f&G{%W|1GZ|wCJ$pvrdW@SVDkFi>mXSnUu#n- zW)1|dKn02#xHE^+tjCHWk{gQQm9DMiN6%jyO3y9&BjjPal#K_79&t!wxvSPtUK=KM04DNF~kE}-mGZ?NJ zM95;#WlAHV4bR9f!Q$Az5#X7_jjtReX8;mh1#SWF684F$M}XBWHnc)7Q$f#$kZ!lCz5ADU>-o2+Dd9Y0z|ilZ;@HX`<~%#5Hml(4_r~zGV76G z4aD#%YV5T9_0>5`tZ5%%3JFZ;LI>c%R_R= zmFiC`fSz$Tppt}X?O6GGG%z`^2CewKfIk1cnG{HzS*+L#w;%8P3m$sjTo>e+hu zFmnjtC{&;*iL$0qwg9$P0MlLaFngctU5|iJl(69KzaTF&VT#nr1YUL&F&@$FUyl_g zmLQ%n@M;rddO7GIg@NwiV_>~E>k+{1g3obR!dW60f#Sq~^TI0@r+4cyz(qj1E@YUW z+B;fvKxw!&|3ZLc{02$TduuA43lOLvG&H*O zPXu_JcKR&cY87y)a^O;Mr*&Fh4OzxHkzw%@(_U zCIieAqb`7^!+pM;i2yuwj9bga9?CmzS{}~MPOfDP4WuM^owks-tpQV}gFATpT$ z#_9?~}K9OMT3!Tp12GBWR?d4J=j4pH5*~twDDx^=2ue+KQ z04i7~LV@j$7*DYhVN|H5C0LPP*mZRp+9lKie&F4zCYYDQRxqYIbim?0o}H{yZ$ySE zaDrBY)Bl7(!K#fp9wmIwDW+=$qWc42xM~^Ik)RlrDc<_gqCVk$d1TK#dYwKMhjycdF=&>D@~0Gb*A zDj~SJE`gPKY|SD97U{GQsH42(9AtlJuRI@22YrS}P(1C$*EuFB_**{%;LyQPxPpHe zBf;SjjBl-@C42N3mOD;b$Etb{_{^gx3Drv_KtO2vEw#C{l>oG%jX zzoU$}HY-8Fgn4;9D-4b;3<TZgp}R>j z@6>}U)duVa|6guhvI2_c`+ZX>COUK>6y~LP@KU`k{U6c)-qMFj0^P5Nc@^AFn5$W5 c5F>BfQ&-wV0{vevz(2gIDAdavASI*z50xZ}vH$=8 literal 0 HcmV?d00001 diff --git a/merlion/utils/autosarima_utils.py b/merlion/utils/autosarima_utils.py index e438b5115..c6bc23650 100644 --- a/merlion/utils/autosarima_utils.py +++ b/merlion/utils/autosarima_utils.py @@ -4,14 +4,16 @@ # SPDX-License-Identifier: BSD-3-Clause # For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause # +import functools +import logging import time -import numpy as np -import statsmodels.api as sm import warnings -import functools + +import numpy as np from numpy.linalg import LinAlgError from scipy.signal import argrelmax -import logging +from scipy.stats import norm +import statsmodels.api as sm logger = logging.getLogger(__name__) @@ -160,7 +162,7 @@ def _refit_sarima_model(model_fitted, approx_ic, method, inititer, maxiter, info return best_fit -def detect_maxiter_sarima_model(y, X, d, D, m, method, information_criterion): +def detect_maxiter_sarima_model(y, X, d, D, m, method, information_criterion, **kwargs): """ run a zero model with SARIMA(2; d; 2)(1; D; 1) / ARIMA(2; d; 2) determine the optimal maxiter """ @@ -225,7 +227,7 @@ def detect_maxiter_sarima_model(y, X, d, D, m, method, information_criterion): return maxiter -def multiperiodicity_detection(x, max_lag=None): +def multiperiodicity_detection(x, pval=0.05, max_lag=None): """ Detect multiple periodicity of a time series The idea can be found in theta method @@ -233,7 +235,7 @@ def multiperiodicity_detection(x, max_lag=None): Returns a list of periods, which indicates the seasonal periods of the time series """ - tcrit = 1.96 + tcrit = norm.ppf(1 - pval / 2) if max_lag is None: max_lag = max(min(int(10 * np.log10(x.shape[0])), x.shape[0] - 1), 40) xacf = sm.tsa.acf(x, nlags=max_lag, fft=False) @@ -253,10 +255,9 @@ def multiperiodicity_detection(x, max_lag=None): clim = tcrit / np.sqrt(x.shape[0]) * np.sqrt(np.cumsum(np.insert(np.square(xacf) * 2, 0, 1))) # statistical test if acf is significant w.r.t a normal distribution - candidate_filter = [] - for candidate in candidates: - if np.abs(xacf[candidate - 1]) > clim[candidate - 1] and candidate * 3 < x.shape[0]: - candidate_filter.append(candidate) + candidate_filter = candidates[xacf[candidates - 1] > clim[candidates - 1]] + # return candidate seasonalities, sorted by ACF value + candidate_filter = sorted(candidate_filter.tolist(), key=lambda c: xacf[c - 1], reverse=True) return candidate_filter @@ -285,7 +286,7 @@ def nsdiffs(x, m, max_D=1, test="seas"): D = 0 if max_D <= 0: raise ValueError("max_D must be a positive integer") - if np.max(x) == np.min(x): + if np.max(x) == np.min(x) or m < 2: return D if test == "seas": dodiff = seas_seasonalstationaritytest(x, m) @@ -313,7 +314,7 @@ def KPSS_stationaritytest(xx, alpha=0.05): """ with warnings.catch_warnings(): warnings.simplefilter("ignore") - results = sm.tsa.stattools.kpss(xx, regression="ct", nlags=round(3 * np.sqrt(len(xx)) / 13)) + results = sm.tsa.stattools.kpss(xx, regression="c", nlags=round(3 * np.sqrt(len(xx)) / 13)) yout = results[1] return yout, yout < alpha diff --git a/merlion/utils/misc.py b/merlion/utils/misc.py index 78a837b4e..58e8a7899 100644 --- a/merlion/utils/misc.py +++ b/merlion/utils/misc.py @@ -6,10 +6,12 @@ # from abc import ABCMeta from collections import OrderedDict +from copy import deepcopy +from functools import wraps import importlib - import inspect -from functools import wraps +import re +from typing import Union class AutodocABCMeta(ABCMeta): @@ -18,8 +20,8 @@ class AutodocABCMeta(ABCMeta): also inherit docstrings for inherited methods. """ - def __new__(mcls, classname, bases, cls_dict): - cls = super().__new__(mcls, classname, bases, cls_dict) + def __new__(mcs, classname, bases, cls_dict): + cls = super().__new__(mcs, classname, bases, cls_dict) for name, member in cls_dict.items(): if member.__doc__ is None: for base in bases[::-1]: @@ -30,6 +32,102 @@ def __new__(mcls, classname, bases, cls_dict): return cls +class ModelConfigMeta(type): + """ + Metaclass used to ensure that the function signatures for model `Config` initializers contain all + relevant parameters, including those specified in the superclass. Also update docstrings accordingly. + + For example, the only parameter of the base class `Config` is ``transform``. `ForecasterConfig` adds the + parameters ``max_forecast_steps`` and ``target_seq_index``. Because `Config` inherits from this metaclass, + we can declare + + .. code:: + + class ForecasterConfig(Config): + + def __init__(self, max_forecast_steps: int = None, target_seq_index: int = None, **kwargs): + ... + + and have the function signature for `ForecasterConfig`'s initializer include the parameter ``transform``, + even though we never declared it explicitly. Additionally, the docstring for ``transform`` is inherited + from the base class. + """ + + def __new__(mcs, classname, bases, cls_dict): + sig = None + cls = super().__new__(mcs, classname, bases, cls_dict) + prefix, suffix, params = None, None, OrderedDict() + for cls_ in cls.__mro__: + if isinstance(cls_, ModelConfigMeta): + # Combine the __init__ signatures + sig = combine_signatures(sig, inspect.signature(cls_.__init__)) + + # Parse the __init__ docstring. Use the earliest prefix/param docstring in the MRO. + prefix_, suffix_, params_ = parse_init_docstring(cls_.__init__.__doc__) + if prefix is None and any([line != "" for line in prefix_]): + prefix = "\n".join(prefix_) + if suffix is None and any([line != "" for line in suffix_]): + suffix = "\n".join(suffix_) + for param, docstring_lines in params_.items(): + if param not in params: + params[param] = "\n".join(docstring_lines).rstrip("\n") + + # Update the signature and docstring of __init__ + cls.__init__.__signature__ = sig + params = OrderedDict((p, params[p]) for p in sig.parameters if p in params) + cls.__init__.__doc__ = (prefix or "") + "\n" + "\n".join(params.values()) + "\n\n" + (suffix or "") + return cls + + +def combine_signatures(sig1: Union[inspect.Signature, None], sig2: Union[inspect.Signature, None]): + """ + Utility function which combines the signatures of two functions. + """ + if sig1 is None: + return sig2 + if sig2 is None: + return sig1 + + # Get all params from sig1 + sig1 = deepcopy(sig1) + params = list(sig1.parameters.values()) + for n, param in enumerate(params): + if param.kind in {inspect.Parameter.VAR_POSITIONAL, inspect.Parameter.VAR_KEYWORD}: + break + else: + n = len(params) + + # Add non-overlapping params from sig2 + for param in sig2.parameters.values(): + if param.kind in {inspect.Parameter.VAR_POSITIONAL, inspect.Parameter.VAR_KEYWORD}: + break + if param.name not in sig1.parameters: + params.insert(n, param) + n += 1 + + return sig1.replace(parameters=params) + + +def parse_init_docstring(docstring): + docstring_lines = [""] if docstring is None else docstring.split("\n") + prefix, suffix, param_dict = [], [], OrderedDict() + non_empty_lines = [line for line in docstring_lines if len(line) > 0] + indent = 0 if len(non_empty_lines) == 0 else len(re.search(r"^\s*", non_empty_lines[0]).group(0)) + for line in docstring_lines: + line = line[indent:] + match = re.search(r":param\s*(\w+):", line) + if match is not None: + param = match.group(1) + param_dict[param] = [line] + elif len(param_dict) == 0: + prefix.append(line) + elif len(suffix) > 0 or re.match(r"^[^\s]", line): # not starting a param doc, but un-indented --> suffix + suffix.append(line) + else: + param_dict[list(param_dict.keys())[-1]].append(line) + return prefix, suffix, param_dict + + class ValIterOrderedDict(OrderedDict): """ OrderedDict whose iterator goes over self.values() instead of self.keys(). diff --git a/setup.py b/setup.py index b8736811b..35499cfb8 100644 --- a/setup.py +++ b/setup.py @@ -7,7 +7,7 @@ from setuptools import setup, find_namespace_packages MERLION_JARS = [ - "resources/gson-2.8.6.jar", + "resources/gson-2.8.9.jar", "resources/randomcutforest-core-1.0.jar", "resources/randomcutforest-serialization-json-1.0.jar", ] @@ -20,7 +20,7 @@ def read_file(fname): setup( name="salesforce-merlion", - version="1.0.2", + version="1.1.0", author=", ".join(read_file("AUTHORS.md").split("\n")), author_email="abhatnagar@salesforce.com", description="Merlion: A Machine Learning Framework for Time Series Intelligence", diff --git a/tests/anomaly/forecast_based/test_prophet.py b/tests/anomaly/forecast_based/test_prophet.py index 9d5eb2546..77c7467d9 100644 --- a/tests/anomaly/forecast_based/test_prophet.py +++ b/tests/anomaly/forecast_based/test_prophet.py @@ -12,10 +12,11 @@ import numpy as np -from merlion.transform.normalize import PowerTransform -from merlion.transform.resample import TemporalResample +from merlion.models.automl.autoprophet import AutoProphet from merlion.models.anomaly.forecast_based.prophet import ProphetDetector, ProphetDetectorConfig from merlion.utils.time_series import ts_csv_load, TimeSeries +from merlion.transform.normalize import PowerTransform +from merlion.transform.resample import TemporalResample logger = logging.getLogger(__name__) rootdir = dirname(dirname(dirname(dirname(abspath(__file__))))) @@ -33,10 +34,8 @@ def __init__(self, *args, **kwargs): self.test_len = math.ceil(len(self.data) / 5) self.vals_train = self.data[: -self.test_len] self.vals_test = self.data[-self.test_len :] - self.model = ProphetDetector( - ProphetDetectorConfig( - transform=PowerTransform(lmbda=0.0), max_forecast_steps=self.test_len, uncertainty_samples=1000 - ) + self.model = AutoProphet( + model=ProphetDetector(ProphetDetectorConfig(transform=PowerTransform(lmbda=0.0), uncertainty_samples=1000)) ) def test_full(self): @@ -83,7 +82,7 @@ def test_full(self): # posterior samples for reproducibility logger.info("Verifying that scores don't change much after save/load...\n") self.model.save(dirname=join(rootdir, "tmp", "prophet")) - loaded_model = ProphetDetector.load(dirname=join(rootdir, "tmp", "prophet")) + loaded_model = AutoProphet.load(dirname=join(rootdir, "tmp", "prophet")) scoresv3 = loaded_model.get_anomaly_score(self.vals_test) scoresv3 = scoresv3.to_pd().values.flatten() self.assertAlmostEqual(np.max(np.abs(scores - scoresv3)), 0, delta=1e-4) diff --git a/tests/forecast/test_autosarima.py b/tests/forecast/test_autosarima.py index 57a08f53e..b8c07c278 100644 --- a/tests/forecast/test_autosarima.py +++ b/tests/forecast/test_autosarima.py @@ -6,7 +6,6 @@ # import logging from os.path import abspath, dirname, join -import pytest import sys import unittest @@ -15,8 +14,7 @@ from merlion.evaluate.forecast import ForecastMetric from merlion.models.automl.autosarima import AutoSarima, AutoSarimaConfig -from merlion.models.automl.seasonality_mixin import SeasonalityLayer -from merlion.models.forecast.sarima import Sarima +from merlion.models.automl.seasonality import SeasonalityLayer from merlion.utils import TimeSeries, autosarima_utils logger = logging.getLogger(__name__) @@ -785,23 +783,28 @@ def __init__(self, *args, **kwargs): data = np.concatenate([train_data, test_data]) data = TimeSeries.from_pd(pd.Series(data)) self.train_data = data[: len(train_data)] + self.train_data = self.train_data[:-50] + self.train_data[-49:] # test robustness to missing data self.test_data = data[len(train_data) :] self.max_forecast_steps = len(self.test_data) - self.model = SeasonalityLayer( - model=AutoSarima( - model=Sarima( - AutoSarimaConfig( - order=(15, "auto", 5), - seasonal_order=(2, "auto", 1, "auto"), - max_forecast_steps=self.max_forecast_steps, - maxiter=5, - ) - ) + + def run_test(self, auto_pqPQ: bool, seasonality_layer: bool, expected_sMAPE: float): + model = AutoSarima( + config=AutoSarimaConfig( + auto_seasonality=not seasonality_layer, + auto_pqPQ=auto_pqPQ, + order=(15, "auto", 5), + seasonal_order=(2, 1, 1, 0), + max_forecast_steps=self.max_forecast_steps, + maxiter=5, + transform=dict(name="Identity") if seasonality_layer else None, + model=dict(name="SarimaDetector", enable_threshold=False, transform=dict(name="Identity")), ) ) + if seasonality_layer: + self.model = SeasonalityLayer(model=model) + else: + self.model = model - def test_forecast(self): - # sMAPE = 3.1810 with pqPQ train_pred, train_err = self.model.train( self.train_data, train_config={"enforce_stationarity": False, "enforce_invertibility": False} ) @@ -819,21 +822,44 @@ def test_forecast(self): self.assertEqual(len(pred), self.max_forecast_steps) self.assertEqual(len(err), self.max_forecast_steps) + # test save/load + logger.info("Test save/load...") + suffix = ("_auto_pqPQ" if auto_pqPQ else "_fixed_pqPQ") + ("_seas" if seasonality_layer else "") + savedir = join(rootdir, "tmp", "autosarima" + suffix) + self.model.save(dirname=savedir) + loaded = SeasonalityLayer.load(dirname=savedir) + + # make sure save/load model gets same predictions + loaded_pred, loaded_err = loaded.forecast(self.max_forecast_steps) + self.assertSequenceEqual(list(loaded_pred), list(pred)) + self.assertSequenceEqual(list(loaded_err), list(err)) + # check the forecasting results w.r.t sMAPE y_true = self.test_data.univariates[k].np_values y_hat = pred.univariates[pred.names[0]].np_values - smape = np.mean(200.0 * np.abs((y_true - y_hat) / (np.abs(y_true) + np.abs(y_hat)))) + smape = np.mean(200.0 * np.abs((y_true - y_hat) / (np.abs(y_true) + np.abs(y_hat)))).item() logger.info(f"sMAPE = {smape:.4f}") - self.assertLessEqual(smape, 4.5) + self.assertAlmostEqual(smape, expected_sMAPE, delta=0.0001) # check smape in evalution smape_compare = ForecastMetric.sMAPE.value(self.test_data, pred) self.assertAlmostEqual(smape, smape_compare) - # test save/load - savedir = join(rootdir, "tmp", "autosarima") - self.model.save(dirname=savedir) - SeasonalityLayer.load(dirname=savedir) + # check that we can also get the anomaly score (since model is SarimaDetector) + logger.info("Check that we can also calculate the anomaly score...") + score = self.model.get_anomaly_label(self.test_data) + loaded_score = loaded.get_anomaly_label(self.test_data) + self.assertSequenceEqual(list(score), list(loaded_score)) + + def test_autosarima(self): + print("-" * 80) + logger.info("TestAutoSarima.test_autosarima\n" + "-" * 80 + "\n") + self.run_test(auto_pqPQ=False, seasonality_layer=True, expected_sMAPE=3.4130) + + def test_seasonality_layer(self): + print("-" * 80) + logger.info("TestAutoSarima.test_seasonality_layer\n" + "-" * 80 + "\n") + self.run_test(auto_pqPQ=False, seasonality_layer=False, expected_sMAPE=3.4130) if __name__ == "__main__": diff --git a/tests/forecast/test_ets.py b/tests/forecast/test_ets.py index b4baa6f72..5d2371e68 100644 --- a/tests/forecast/test_ets.py +++ b/tests/forecast/test_ets.py @@ -13,7 +13,7 @@ import numpy as np from merlion.evaluate.forecast import ForecastMetric -from merlion.models.forecast.ets import ETSConfig, ETS +from merlion.models.automl.autoets import AutoETSConfig, AutoETS from merlion.utils.time_series import TimeSeries, to_pd_datetime logger = logging.getLogger(__name__) @@ -99,16 +99,7 @@ def __init__(self, *args, **kwargs): self.test_data = data[idx:] self.data = data self.max_forecast_steps = len(self.test_data) - self.model = ETS( - ETSConfig( - max_forecast_steps=self.max_forecast_steps, - error="add", - trend="add", - seasonal="add", - damped_trend=True, - seasonal_periods="auto", - ) - ) + self.model = AutoETS(AutoETSConfig(pval=0.1, error="add", trend="add", seasonal="add", damped_trend=True)) def test_forecast(self): # batch forecasting RMSE = 6.5612 @@ -123,6 +114,17 @@ def test_forecast(self): logger.info(f"MSIS = {msis:.4f}") self.assertLessEqual(np.abs(msis - 101.6), 10) + # make sure save/load model gets same predictions + logger.info("Test save/load...") + savedir = join(rootdir, "tmp", "ets") + self.model.save(dirname=savedir) + loaded = AutoETS.load(dirname=savedir) + + loaded_pred, loaded_lb, loaded_ub = loaded.forecast(self.max_forecast_steps, return_iqr=True) + self.assertSequenceEqual(list(loaded_pred), list(forecast)) + self.assertSequenceEqual(list(loaded_lb), list(lb)) + self.assertSequenceEqual(list(loaded_ub), list(ub)) + # streaming forecasting RMSE = 2.4689 test_t = self.test_data.np_time_stamps t, tf = to_pd_datetime([test_t[0], test_t[-1]]) @@ -142,11 +144,6 @@ def test_forecast(self): # streaming forecasting performs better than batch forecasting self.assertLessEqual(rmse_onestep, rmse) - logger.info("Test save/load...") - savedir = join(rootdir, "tmp", "ets") - self.model.save(dirname=savedir) - ETS.load(dirname=savedir) - if __name__ == "__main__": logging.basicConfig( diff --git a/tests/forecast/test_forecast_ensemble.py b/tests/forecast/test_forecast_ensemble.py index f68141992..59491f4ec 100644 --- a/tests/forecast/test_forecast_ensemble.py +++ b/tests/forecast/test_forecast_ensemble.py @@ -14,8 +14,8 @@ from merlion.models.ensemble.forecast import ForecasterEnsemble, ForecasterEnsembleConfig from merlion.models.ensemble.combine import ModelSelector, Mean from merlion.evaluate.forecast import ForecastMetric +from merlion.models.automl.autoprophet import AutoProphet, AutoProphetConfig, PeriodicityStrategy from merlion.models.forecast.arima import Arima, ArimaConfig -from merlion.models.forecast.prophet import Prophet, ProphetConfig from merlion.models.factory import ModelFactory from merlion.transform.base import Identity from merlion.transform.resample import TemporalResample @@ -36,8 +36,9 @@ def __init__(self, *args, **kwargs): model0 = Arima(ArimaConfig(order=(6, 1, 2), max_forecast_steps=50, transform=TemporalResample("1h"))) model1 = Arima(ArimaConfig(order=(24, 1, 0), transform=TemporalResample("10min"), max_forecast_steps=50)) - model2 = Prophet(ProphetConfig(transform=Identity())) - model2.model.logger = None + model2 = AutoProphet( + config=AutoProphetConfig(transform=Identity(), periodicity_strategy=PeriodicityStrategy.Max) + ) self.ensemble = ForecasterEnsemble( models=[model0, model1, model2], config=ForecasterEnsembleConfig(combiner=Mean(abs_score=False)) ) @@ -68,17 +69,12 @@ def run_test(self): # generate alarms for the test sequence using the ensemble # this will return an aggregated alarms from all the models inside the ensemble yhat, _ = self.ensemble.forecast(self.vals_test.time_stamps) + yhat = yhat.univariates[yhat.names[0]].np_values logger.info("forecast looks like " + str(yhat[:3])) self.assertEqual(len(yhat), len(self.vals_test)) - y = self.vals_test.np_values - yhat = yhat.univariates[yhat.names[0]].np_values - smape = np.mean(200.0 * np.abs((y - yhat) / (np.abs(y) + np.abs(yhat)))) - logger.info(f"sMAPE = {smape:.4f}") - self.assertAlmostEqual(smape, self.expected_smape, delta=1) - logger.info("Testing save/load...") - self.ensemble.save(join(rootdir, "tmp", "forecast_ensemble")) + self.ensemble.save(join(rootdir, "tmp", "forecast_ensemble"), save_only_used_models=True) ensemble = ForecasterEnsemble.load(join(rootdir, "tmp", "forecast_ensemble")) loaded_yhat = ensemble.forecast(self.vals_test.time_stamps)[0] loaded_yhat = loaded_yhat.univariates[loaded_yhat.names[0]].np_values @@ -91,6 +87,12 @@ def run_test(self): loaded_yhat = loaded_yhat.univariates[loaded_yhat.names[0]].np_values self.assertSequenceEqual(list(yhat), list(loaded_yhat)) + # test sMAPE + y = self.vals_test.np_values + smape = np.mean(200.0 * np.abs((y - yhat) / (np.abs(y) + np.abs(yhat)))) + logger.info(f"sMAPE = {smape:.4f}") + self.assertAlmostEqual(smape, self.expected_smape, delta=1) + if __name__ == "__main__": logging.basicConfig(

L>?j{cRxW@qowtz`tBmabwfNy`ZOiNO(| ztPOQ`NeCA;Zgld)I@_cPxxdBIkFZBV#M-4yAdk~;V`GQntmocK6}ON%O}kCM`Jg>8 z`_?3~nGkUdTLM)YB=gr_0+s@t~e!_ji+u-8VBzS+A%-nJ`Sss3Wxu7Od>x=SHEs zmsk1u=v{7uyO+y975Iq+Et zzc?+oOOi_O#Ob-u>m&A(5@c{JZpT^JmmSyjn`CWjc_Xbf8fabf>ly%cu;~4&@lk@D zU_9X`I>A$0!HjFQ&(1|Hqp=Q&AlWLz;ayCX${EGxT&%t}-!}I&t^j7<)AD9}P8E zlXTr^6m9v*9tE4M11Ay=njs1<22cwsE*isb7hOC9n;P)rFQt47feUb^_Lh!{KVRE*ffFte``_x59!X+b>ym{?4wmu2Thv=`d^IjDGV#Wo z2(2|HyqM0)cfC(^PR!^N=I~=)3xzRD+|=Uaa9uwloqMfW35OZ@FN@PaY zM7d^!i+M3zFhO1Lbb~L@sEZ!V(6Cy>vi>C@R=ZxPWqIdU`Q55=nnk_bm5@qaUh}V~ zW8Jq7ExWj-%NbKPt`0c&=Jv@qg^h6k_DP*$8z+eNOI!~1J~(OGAISE3j%T0AAiBwE z_gt=j3o)qKPgWN{nmnyP1^5doc`5c|M7dwsQ{s4{Xj8ZdhELGnAR71NxddXV!RhVe z3DakjAv%zUjH6ud&hU1P(RrQttv`S&o(L))5#W0J%GslPgNHCEqUV(4sylfl=6MIc zf!TNa%);w+;@h0%Lw7I*cIum>A5Frx%x5*geZ z`)m*ruTv6P!Fj=sbmIVlN>kDWO7;lYW$`PVfe2T)$$~;xhJYz0?&B&HJQrkmYUTHH zlj`4DJ`eJz=pla%vkno#zo4@2F<(D^Tvk)pLrzM|(B+dAzC%-PB|M0XGPUR=>@JGh z_?2o<*f;a_s`pBWz3@>Ir8{YtdxJ*%X+F&JUVHtC^RDfbm@9phPryt`mPfhLQH^zv zhY^vkvxTs5?~~8_q318RQ#N4RJFdE}|Fm3$tQTMCdCuIy1}7}G!Rogpk6;e5y6#$r-; zkUgD3T&hYw$wpSAb*1)RU(%|voRkdDUe!2*+app|=SgH?5QuG2^g z_l{-NwXK+bXcyqM1NJbMw~@9ufta1^hpretu4?Bj36_LD{oWewNv!k^eE9dY2PL=KFtLArXVkb%6%HP{K6z1bqEi;HG}p2!h9z)$-V$Mxmu6g1Icp92x2qFnU0a6utpo#-dtFpRp1C zvov!#aHUj&*P+-**2GKNpd&3qm7QA2xejS-;E8F}1)#ZN^2C&CN}ZfdNJAP`H@VN^ zxtCX~$x^{}p+1>m#7TQdr#;kb+UkJzG?r#bZqiiu)3#kC)gcC8jU|=roXjxMA;Q?Y zX;{yIpXao3T8DufpOZem9%JZD?Eg-NW)T12RTfK)ym*?#Kpvf2Hz%zq2K>SZg?tfv1)z&nzRWNkYL4%}`k$PKV*yJ@qz-AYPEPvIjmR8`@hZ1PQ z3#A0*7=V+<&sy;E{_)shes&LMzxIyI?sx)w=i&WRq$dB>Vam)I3EVP#e*SUwyW-)? z@(O0_>*Jam2>rSwjBs7?Iw=gTy}}UTio#7zgqj7+k_u&gH=Yyh49W?0$Sz&(5Y?*q zcsm@I0mZ8HxE~h`3{_`&ms=VvDNNVuj{b_WS_S5BakqlJA*3F=%toi?beKrT@HCrI zN)YoHf_Nk?Gia6O8JCdLCd0$P>|&aYX+y)0iTYqExHpKlQzAwJyTn3uoi17!t}6wnoV2hpS9YPX5;3SYSVJH>if=UDG55xH9A{W zeFcwHbBp2S?JC(=m<5{{FG^lyd&pp(P$_YIa(HbtQYWi{1w^!$aFV7wMB$Q>-)u{) z;)KgFWyrQr1SQ&pVO2}Q36=AuplmEakt)mud4m$~ufh~X2}BBn;HotH#*u8;BtZp6 zxTH_Z77oIl%$ePV_y=W3yc}1gjdC;^98$(0T11PTzs4idSJC zdb|RS5_9b$S;8UagI6+ZTUoTFqX(7O;Jf$O_I2%ZgQ1L@-q(-L zI*R3n+3*Qwi1;bZ1%X(uAmlZ#gcR8E%4tvt|~kfrhy1P zIWUHNmjm02Dx4zmZWELfom#x{uC_Hp=w|S43?0FZm^NL8?WT;Fy>prv$0QhTJ~%0` zh&c4C&o#^trR9*amD~r64*TLwGmB5DW73wHJ|U{(WM?c3Rr{^{lG|nuxaQ1#7SxY! z^CV%zmm#E`HsT1r3qZJ;xK&Bd%2+tXUZG4!~Ey zBx*}2$&{RyK=J?a!eIcyKYbp}@&#S-9>C#K$oF}gNG7|S%t#BBd>@v8Qn2&#%II$z zB2mC1QN<#mfEA?SdV&7~k^)1SAzxl1^#>~nDzB8(RIHM=h#n%=kW|9Vudv+8T!PS$ z3%Z1v7OtA;XJUmEjt*k{;Spf1FcEs06|`djkWn`H8W%+2QPA(bB|Nx2*w0pwJK zA}aYLJd~kSl=IaQyALm54&)mmd~j~D!ZV~K8>#WRxd117SC{xw-g{-Q?k`IqXo7w} zKZjY^r(+R&5PGFo1jLjfzw(x|=wcF0IRgcG^*9u!`ayum;Mnm2&gMzyrxlD=2-w5C+H|^3&OCtzM4l|>Swp3UJT%?d%aG!XScN!STs2_-h0lG)~^&=hXX0qoruvWAt84owlZ_kY%GAXG|j z!1uoH+IQXlp9Og0|E6aDLp7CmFt;{$GI#mc>?Bc5%L#h`^K&q+{n5ia8gR7B2+GmRxdbZrbpjNuoa4JG z$9A3eh~xWRyC0aW#(ww}<#VSWI!b%-HblltFqEz#WX9s$9k|fiv&2U|p43^Y@wwT1 ze(W$j!~7*Pv&kor5X1f|UVV9ejw$$SH1l(CIcA&<9y{}3|GU-tFy7prxvwed{k=V} zm4QY&@^*awTY3$m}Ft3I+H2D(C9vOJgv@TlgB#7@OOq2L%pXN;Mfk^h#IwV z-lV^M9lb@5kTr2CEj6Du5&6!6vSn)eAPGzIM1war`BdFcfY%lC)C9)JW;l(zZyNiw ztFZ0CM7DKk8~QzQG!chtRdJ{`sm)pd!kvjyT`C>x@Ec8@I*>>P^Fpb zef9;+o57kanuqzt7)y4@pDBf;0EmgPVGGD5^&9!}#>^gEY?b`Rg;Ua}D;bPNAa+9V z7R`}om_vwQ{Bzfi>y>~1?x?hj796WcCK_w7VA1W+fk|QM4$-r`Bh6uLcBf+{eJ2W| zmRmfdUd~;k(OEvTUM^i@w^!^mHRlpWYQSD?yJ8__9O%4f<+NF>vb+OdnKRh=L6xF_ zwQ=bnI$LzCpi;Hz>ln84u&%?D-bSi8&`Tz+N;^$p*meXaX_427_KuFnqvx(nxkN?pKo+=^t zb2Ea1KCVmlI=pj#WP)2IqxA1MUYn)+trC8j%=9ll=|7*-6t?=bE@sgh>yf{hv-i0B z)*Uucf@eLVC@ugKItY|=z^&Z(KhuAwVTp$5vH`K8Z}ONeirnce%WU!1V~N0i07vei zR@)h)vfB1OaUojR5qqa^=h>!Y2_kv~2;Wd@f>{H5VX*@nHll-o4jDXo<1Fidwx#Po zl>@ioLy(;JM1!pvl%`?Bi8?^FliCRB@&TD5_Af_(x&dAh|Qy%G#L4}aLYcCRr8 z7y!#2>G_Ebq{;V6!KqgKU8iGU@l-Qp*sLtrm!;G|_W|j2mDehc=^7%{I~tDWx^)$Q zf!zU6K%4C{sL8iQwEj>yRm*V@&K~0ifs&1JErfRW7sf+wFOLjqssI=Y6@+TYA)olf z+|J>|4bnEYH~w)>o^flkE@|{D441zU>5DTegY}*S>lgs!`K3h?L-A%|k}cLMmDS%V z$Z?-^CfwF3c4q%rYX1X{rI|F-s*JhPNNVmevP$sd-rMh-E82ud;;d*o+*BLtO=s~< zVQyV=eoEeXxs-fZezIXxb;GR8=ExUOLff^3h=IQ4?^4@ekYaP(ot>;+#;a>rJ2!*h zLn8IMqQT;zFD-uhfM;mS;NQ3#pSbn*5K&URHx9g+QCYG%MRoZ_W%)yNBdEpdBKVOi z56-iLnayLNn0Fy79^@w2x@Fgg8y1XvH(lw5Dzy06xaTres`_6gSdH1kPurWNF0S0m z*%Zq9UoBW{L$5TKWnmk8`nHw*pJNTx{K0Q!hEH6}-igb@z(2FZLDT&+`_SbKkFo^d44Z+i! zZ;S8|*t+a}3q%9`$b5h4$%h=g?LdVYF9)w71P_ZS{5Y4eHAEGwQY%2R4{pd)Z{&z= z6lK}ud*jc#93I3f{Z7;@56`yCL>!k%wl{sdD~#7kBi8*3ToZ1tttakqU_TW~p35*% zE?8G>0wNn5-lf?A@O{=i1^1QV@+z0k9Tya)o)rvhd(n;YZCm?e>M zPDo!+`6`hAG3w{kDX=KON<`gydEan#edO%=dVdGo#^eGdR96;3czCG$h8_~(?bL*l zlUX6t2CY&?r>`VW^-6(GDfd=y!YXgUCPEqQ1eL^HW=S*C`7;`)?~E>o8x(l<%C%rS zVz!1pV}uhpdan{HQ*uYm9x4S*q5w#kW5w~>vWRYt5fH8dN;6%FFad2a-kvsBQq(Dl zj%P7F*Yk)ADrcgK?b!F8L^tmAWMhhIU_5;L68TFt6@{)7&5_t!|#B!aMt|F|KW zRjK&o9;3w#i8^e6F-#O?T1Q{0;bPiRUlv&FRZ*r*P^hVf92101Cu^i>o|a*Fie;Fm zR~_7b|7{L)nZbw|mh+bYztGdBR?h1|V)$Yar$`z?yWZKgLT5G&W$tM#WxF2&;6D;-dr6K+dBbnb5<-%hO4fL75}Elq3^?8!ldadUVDJGNj7t}O$v zu^B0rWJuWde&@7g#~|=X0KpOqeV#osb$yYBUd%k6%2CAgr2M;Y;Ezt-ZO_Cz4dQ*y ztqL5+bO$W+vgY`)wq~)}Zmb12|9U9*#M|fRe|ET)kD+b&-$Uek(Eq=mLFRvH6)M(_ z$SNp4Hfbj5PK2cL0#ryftL;GMl)F-vH36#BloX+f$3+u|adcK!ixY(hrY8~k%uK{v z9|{pu8^+^gqLzH;yrxe)M;{q&zdJuZ;DMTMjF8an#PIQp7^2%Kz5L{0y-ct3^Tn*AC#fmuk?3P(L$pc`083M#}?u!9{Z(`U5Zh~gkk^`V*;Zb=+!S_D#K$KN9 zVF~vD6#K`*djD#XaMBiXNODJ^u_mh6h*+k7(7mN<-$AnUKpm~?)GhAE<`>3NDr5?` zq(s-b=0)TF*=fEnZ6#C9{+0?#bIh2Sq4&Dmr2NlNEuZkmsV^2;3XG|T5dGjq8dMI| zZ8wXv0$O-hq3sDqz_Nmvv>b@VJ)u8^&v5bCXwrerYNJ(C-e2a0?pnQb&EStwk}_E- zjbB6L{F!4*7X1ldn2%j*r4S(~=9z44W~3?)T%EK{xECe{UcYS3#m;w=YIUkEiWe(k z9y1U1nC(X{NpoEyn3D$j*R4Z`o1kQhfpY=zg@v3kzY|#)NY@By-b|f4WynNVt+w5 zIneU={Iwq}bang>_25Xl(QE>ot%EcB!52zsoVpkzhXXu|UAw1(vf%^XOx8$J@8Z}Sr zf(yuC8?i*h*NqukI=z!=C%n&i z>V|pf60B<2|682)(>i4e$zfkrJUND3t;p$TU^QN(4{Vs%q~U3fulc6eNz$(^_TJSE zV_c7Q+62K?X{}bSrkb#)K5;N{#28LFE|1cnu2st}AeCLDz^V@$EfP%QK@b3)Dp{_NS5gmGu28)lG zQkAkom54A~@?bFm$b`F`q0sd?Hl5i7CnfgK--+;0y}rF4z@G}cR#;PYkVJgvlg1`| zTd5Ma7jMbPKvUPnz|c+X2D?#_ZLk=}Yt5#6oskeQCLDWIM#Z=eBp-=eZedLGp@ySR#EhMRhT|7apd_ zPRxV2FJ4+5GpAz5m^S%7(VL3$l!$>g!Qq%!mkk1ph{Jc@CnP%JDe_~}B$f%9qc9ej z)A4DHqClvDGIjbhWk=q{YJc!X94q~Jm#+d*(wYb-hvDkC-`JmI?;cF(gcA#&cl&I3 zOgtY)J}`}*`*YAa^jmM(k$4|+XdyWu9b6mET=!%YyD>+x3!IK_Fpr@bSd%l>i9MW? za~I`M)sa%hm~?qMJ#tjbSU#c``+Zm)970+}_o&(x$GD}Hu??&n?wwKeQtcg7C3Y7& zeGj$P5F3(a*^vHLchnJ01?o-AyrLg_X@;s)kE#ttVKVC!s12_1JiPN2apyTAjUu|p zbae8QBv(Xi|KVxVHU zAS+dRf)oDaLj{`QV-fMmHqKonj7Vo@s=NmGi+RWR`SEw-lWt~;hN$!aRx-Iwr+AMh zGgGhlJH0@z5nRtw`WnDXn|1pl!AD=*>Vg)qY#5a{I;FMXF%}_eQiVeqLy;8Xuy7?V z0v5DIyz`Vy>y|OOkzFX=g)B}yV4~Lo>-rFd3T$1#`g#Q{ZkW;|0>6=}M8@{|hLxi0 zqh5Sq)b1rPJxPnbkEn7=zvje5GPUH2Z#~|j_0=tc{XssH#KO)^MluzsPU49?1hRk6>qyVnp0G*|YU25+9?753Pb2UsQ^CQqZj}tbq zO(5ktsQP#hK(OmMuii4L@l+lbwEU|mwoFL6^y}zBhbIS&PM^QMVHAGJ1+mg@T8Hsw zQ8ux1DjO2j@5y3)V4taY7T&0yQ!+&bi3R6!`@a#-z(QL=UWl~dMpa9)=R*Zysn?~t zCeg`;Zh}uDV9C$twTN*&@%cMPDZK7;nx3Sa{KvPE2g`u)N+@IWK(qJ#1pAas z@)~?sOVGp5yS+)6g{VTwuq%|5%5p?ZG_Q}4@@bEtW>K|6r8e{6=PKXCFcV1OPg35XSuR#*BFk*k&K#CpW zpUW72{C@s8LL}z|e~TOH71(sIN>$4Q90*Y$2;@b!2kEyh)CiRZe2F1^kVS4x>e0Y| zN~z!!nie*BzB)gp$NGHTCF=q6M8ofKgt=4d#_u{)R!ZK8{^(N||Bf5{X_I{LsQJUz zP^7X_D!vJcLSLm17S{5IkjfmJWsohDGz!(FOXp!L>)QAdR%JZ>Mpe36ndk<4hla1) zOets3*1Z|Gd&-4&2@AJe@Qi0rrn!~IgGil$N0NaS{4v$&kD z3GK3Io`D1R0=~&15puX&p#p02dZudMQPIV_bEWNUkog44CDv0!yER9{`|6?2k0W5h zu@jYk8(UxF8FnhIGErWEECQsuORdVe>FdVsdjW!BLf_+|EYcG91NmfF^Zf~zz9bd6 zHkmH?+wSZg@v1@#foQ1w&lS{_Q5PQrFlQ5lkotraS-GkE@6D8nQ5>6w-(yH3O(!bj z04YO>OV^L~@jv^T4pfb6)fua$CtLPqSerBU$RfDp_L4(Cd2)zJ2Qaec8Ka7vr2ppc z=-{4FoM7~j;wrg`_60?0xycPVa?mdfu_QRRN5iYH0aH-e-NqwyehOhgjkcCuj1 zC50xnG84u`*2Fl-ZfRLXEmSk?gPfTY4Ft7Cnvei;O-R!rVp`$ay&1IJm2AW{6@^)< z+P_|Zm=h>ZkHaO5fg0i%-O5N35mgcSawF>{0lx*0A^IIXr~BciaCWKwFV5aExY91% z7VV@v>9Awlw%tiOwr$(VioMdYZQHhO+qSKnZ|{9h?S0O@-;Z0hR@M7^)_lf%FvfT~ zdg=6>F6;u&6ifksiJoA@JiPFQOOS{HO&G=@m_`(cqm{%U_x$A3c2g3`AU>lUl5Wv^ zgI2eZuv#hKnt8C7l_t!btmAH=2R;!wJM`Zwi(iHHL>nWIbkPNI(G+n8Y@@xA1OUDk z?YXYeDW83hG3M(e;=z{)ox?xKf@SFTy!;bhqjj4@Z6P9#aAWpS@6x^b-xF>acAg?+ zJITfde~z$V{0^^~PehS=0BK;!F z4NobsOXr`g=h(~!`u6KNqJ;eKS&z&A;lJ_!=SgDuuO~_I_@8ClAFQxWDd;Kob-y{F zkZt9QMm_70S7hnnh`zp(8VQD8g?H4p$A<; zdQxiLc}0;~if^H_QlUf;y;$bZSx|!K9DzfvS^@OMS6~{EtI}vx{-9v$ho5{l@>^H4 z&2K_fE)H(N%-SOWnCXvowZkkaf}pz4l<- zForYrmtQKbU8R0qO^-rqL1y~c95>G*EaJ7kwWOrB5qhsXs7kZF%BOqn=2NWvGECvn=j4lYOq5^wEPtuy=?krlXg`QQB!vz_ZW7 z^eP&iJQ(LtUGho!H4Z9cz)`VB{Q@~bg~M&)0*gg{2RXI)hiyS1XwHgRt)q#PN&u#V z;tUJ(ylx3k*vMn7gK}`jjXm;5#%8 z=01Iz8lkx|1O{o`COnPxNC8bRaGzT`J5tyT&D0t{LHac3r)mDLEIX{+&YxJvj!1SK zD~;1D$v6Pmg!MND;YZQ+pA;D%Vjt8eXpT8Vb0ExTFqcj_dBH=SD;Q90fO{534b zjdchfFs@C|7uU%8HuX_>2D^K@OzQXq&9{(NHoE}VxWT8F({`S?!}Y}+N$T#uN`q5- zQ(PUr-U>n27f|)qB~5*UM0XJMTqTn-x_T7M3l2N4ZF=CHWd$4+YsW>MXEFuOxTwFk zc0{XrBY#*!<#(TVZyA3F0XW z1$6~QMk$iB6vL89k}tnYhWt=uWfQ?P`bFXgM2v{cv5MkeK)Z!BwVV(%=k9%5TS}{1 zK~a}JpD)ryGGIp{4>m5nHaQNtE(fokO5rj<>2)d~b09ivu1@i9?%q@J+n&k;5B-J) zFR9-|K;E8YpgS9{g1)U>>F^_ObvoTfqJ9`|KjVj7`|*B|4efr`UwMO(_h#N-=kRLR zT-k$AUM1pBY@r0Yvix2A(WcTD!`%t^OQS8}k+AoZ0ju*Sw7>D`9^pztb`y+0jjDlA ziU5`ZMm!q%)Z84zSS*;gjh}jcmXKmz#5aa!R-~D?;Mkx=U-5`CFH-!yti}-7VWQ4B z8sk7++_1NXpRlXa!qxh1bg#{4ZU|c|z0%#`UJfgcv$%1p0uS~D;xtzd4~_hqR7=Zr zxDB6`vRpseW*C(JdO_92M1**`+11;|QA@_Au}d2*QY5ytg~-!$nwNihg`Cd-hbm@x z=!07Ec|eqL21)ArsKxY`^YzI+7F(mu=30l{{R2uWYq=>}K!y1-&N7nONkm_3O3ZZn z`gDY1tUfjB9$j$1L2*8)jMX}-uZvO+6D7HMaDQ(3I}caR@OU)N`Wl6qnB$+8#y>F( z@6K%P#aayZ0Ix45cwsSB^wCKKtJ6Knj#M@KbBIVws`w5ul$*t%fn(dd9Z4zHuI}8tAw|?w>Aj#VLq|%Pp<(5BW|IEJ5JMGj8Rm?fE9F1u zHF|}*$4zUI{bQn~H!&271~!D!EfY$<9UnMusfadST73@bApy`Y?(Vt6@xdPeGLDDy zDX>Zl2_eo3HN0b|aNnWDQuu=U_}HSAU*M43JVN01X~`cEgjU3&r?zhaKTfnKLh z6?&7HuO1oy_+A}_JJ`I*VmOGpRIjqNMV*n|iu|x@n}k+9%>qhf-b?I{Cf%DwicV&?yOj@!+1@qHK3Fbi zB%AA_^cJf*9Yc1!3j{c4dMk8$$1t!;F2i_dZJR%M#Bq!HrL+-L{f2_^>&Ma z24~ifRI*&MK^|c%Nn{0vkLdHa3)42L`cP>z`O68>poEB+(t#_U)MDx&hEgK_Fi(mX z70POgmBgcs-t`-~q=9q0KUG#UG2|nv?npxh3a%VuC@o_~J=9Z)Jk(@mkWWgzvjvle zh6N58mEw+6&#HtSEnF+HAZjeG7Lr}m%@PVOLlv%CCCNTF%rz5z9zu?vA%MdKQJ2nP z(Yhm}O8;xf5-&BGJQqc`d^+1zwQ^6-7)jgsc*)l}C2u7N$%dA;-Abp?VeP6{8zbI8&Ov7Wf59?zbT*;YT07Wa&p0H_cQU9KYZKrDy@cnhb*;|zpFKnEucwoVCNYf z&mPz}VWQ9fsw7=)fZ@|(48G&dz+}1h{(!X$Ws50xXr+Q@ioPn;{^jpHpUepYf>Pj(Eoh_(qU}c*P+w71axfLcuv=`t#LTw1F3Z@Mo-82}?bmV({)n)z2ZSN^(MPI;?2_zGMS!zRw? zGF%FwCN~l9k`yS5==&E_BYSmqoYPY2v$6#NHLFqZW&x|J=>e-7RIJN>ehJoN5iC4F zCr#?UxJJW}vc^V9iA|zagC+^u>J6)FYk1@~uEqwx!;tg*H6%>6hf%CoLwMaa#~Ra_ zr2%6$jhUo`#m_ zSsX1E!dX>x7YnQN=`O|FT!I#2?KRO`5|KiJqeV`6r$OTCt#O#asI^)Th#R9aFX%xJ1;=J^n6KxsJ<5k&W-QLHQUjK_S`))v? zGTEX&lOoJ%ow&td2{h;M9!gpiOLG-!{5^#&tlOw|HgD3F7X~GCxdyF&JvNyVh-AL9i|cU zLF~9&qPoI`oq?6ina9wO3f^h!uu*y|7=tV};&d+MA()0{n$WTAtn6sy02Xi6I?UZt9s0)dVAAw3eie z^~FLyDu%j6Vk91M-8hC%{Vsj3ii1~FsR!D0 zr>pK>mu@k@>M2$a*%sDV)G(WAC~UC#PZ*q|A7hjw@k#U2j7dR0sK|+DpaQ3ATf{f3 z!#ON5M<>`FpnoxPKxlYqeX)jHM`Tq#ou1&bn?9J|;PAWl_`a>a!uy?Pyfh@llAGEz z;XOb~PD67a_#x(iV(D$ukmq8W?2Lr{PP6AEId)q(k_bVKP)RR$*|xQwR&Ah(^bs~?tS6OH(MVgIM``5d zF8OTbh~vGI(cQkOnP<|VU=d8~prWEEijm^n41ECQB)l>9pd?S46T~x#P{o2cqx4y_ zOjY-b$Tl5^$+vDd0Xz{F6gTTvu`y%nOci;Dv;*#lTZsdEey$0Qh!vzFclKRbD!H-4 zam7o1C^u+vJ1U2M9QL1hea}zum{*TUkzCb3K7B^**pxicQgF zvq$v)3_*3wCD*=MDNQH>3-m~D6Ge~6(w!!`YNN>%1oFqqYSXh+$$(9mPUj)(Im`Ws zm&3akpu3#nC&VYY_&b5BHfP?UsxUek1q?c{fP);SSwpfQS-RWW&MFCRQgPBgdwt23 zD{-&*9bH3SY10;&_mW}I1H@sLVbCiT^NABC^Dqh1o`)#AjKz|>bDBG7$`%4v&tOyd zmRS?nLrRD_?f|h3vl+2GNj4wCJP(zxo~ ze#nrB>s<(}QJfBzY~DFh-W^?R{dLf_8SmQ}dS$236hhN_gi=-`JK#OKPm)F67w(!+ zpwn4<4Mkf;ANo1=CG@lQ=yQ6G_b8>4$!@gJox~cQ_D)xLFW8E*b8JsvZtDU(>QT;T zH*Vy!joY14q{Jstf$K&;y>%l5cqa-1)TN8mhf-U_bLT2=i&C$wNz8fS;D9W)CqH`! ztGD167_Di|@2PKCGt~E}9ur<8wsHpSFW;`@kgrS$68-ho*&~F4WKXOy{t~pAU6EG8 zv^{8D(6WaW0Aq68oyCCMo`VLyEjP{C3p&}|Pvg#J!AOQ{nOx&N0}nfu-d(Z1%;#4K|%|6GtY zeZ#UM_D_iY$RLWIxeY?8tmpSs))QNZakY1)61C%-gt0%xY5tVO!u4sPw@(uf_6|%p z+!~bUt6jzLIY$`31LUUc@Z89 zv6JA6_UZSVlpYL=-V$m?er>0`Z@HzJlTl$%jFlbPhS%^3rbfy zslOmLUJNE`O~~B3S^Z3$Pt63 zn}gr8-y7vzs6`!%Q=S_FpBom|Bu@>aKB=5zLHfWYny(2{9Mvo~uKD5N6reZ?lu+^l zk2n7*QnI_o*_yaX*&y&tFv^Vh5oO~8=T)Y~hp!SKMXdR!462`TnjW%Ee()`5{_U6cGX%Fj z;$@Y_KH4)1PYeqnV5 z#})}JF7T`&)o*NPliGwi{u(pWELgyopiRx`P=n9Mi)>*B)8dy);7B?xI)HMH;*v}* zcg-y=^_`%z;__>%1h)+}qia?Ms|MvlyIqFOPW63Jpvk9s8tI1Dk0PAFW~o}OL7V=x za_jc|YRLJ=zB#f)078|2^jffM4RoS743!bvf+(CiNyK}izXYZyDd#a>+|(yJcqe@9{D z9+PlJISJ@Y=reWbh(1Z4MqG84E%rv)K%oIjsS58DKA$({OCrcE$=8qx`Ip;bY*?QQ z=NG6oJgS>3PvjnY%BaAl)RcCDN~VPxB#ZrPpioB?n-;;S=hksHXhS;gZn;H(u`As$)l*;kEhrr~yj;LPWh9h3hq3PePW%4th z1l!o|971|}1-r6Ot&CSc*HGJKEi6Es?7QV6wCu?Gg-6sNTy$9og89|(uL?$b_^?J~o+8vEKH=qA8-HiN>y^(_a?VBR) zzs)3A{;lKr-)MjT?eo8ypQ^_$NQ)?+;nc6gxge|hxAK17=!VN#O9SjnJ;e-BF-yPj zzq1er@Z&=3nGIlw{hrAaYsn+ligA>{w~D3x`YxaGgJl;JoDo==736!6_zd`fOsjhw z4-<|dQ4{Us^&0hb>3m#w8T@=YTM7E+c2)d0T{MKC#Z03DXb8q zW8^CAuR8bWRT>0a1`6c4g~t;oUPt$p6v90~yAP(LKRMk5z~3j{HgshLyu9kYWb}yK zCP5enByXbtFGE#B@#x>8wnYflIqn8wDnm@alA%9!x6h!xLcm2K$E$T2bT7)BP&)5{ zD8^yFcq&{-GP+Kdj<8*sZ3G|#i7|9GH=A{qn$0N}{)S9oIhv)hLUV zA`!G8pp6E&iWizCfFaqN0ue`t*npc1ZB$#=U46e3cuDWKrf^afq&M zNrKQn+YvwOjouPdfgBs$J2dLQFJ?v$#Z#N|+FI$?1q<-xcW>$#vBTvFSg4&y>U0Gn zi}XOK+HuroB*~+L`Du!S;@T0kO*Nnv56aV|dvM=;w<8hd0Wv)h&Js;{I)+cN0erhL z0i#K=^bxKj6^JhJ)@I3-y#v;?AL3vEJctEZqJR10>7h$RoI*Wlw<#;cRJ&dgdAov8 zhK}|}-UH8Q?X<`tl3gtSgsRh32SnOxb%l(n$<88;C`R$g8Bx`SQ{6JOBV#zb&l#36 z8hfTKlPMC^I+p&OeyH1o#r+v0E??{;;SI+8xNj;Wr2X5(SVoB*i}&%iJi|+y<1o ztJ{8YxfGfP$omEfZVX8$#6sekT@eL|TEX);hssG&E0Hp-7a7eqeVAP+Q3^LP~~k2X1A@p>c|1@rfQt~ zqd{K3GFHfId|c45cnaIN1;lYu!^H9)^zwo0DhgEq?z}C6M}xz!c?0ITuvgmdWhay| z{_t>%uI0Tp-xjbr>6er=ty_47H;+(H?R7<~wASOcl!^Q}n^Eor*g59;wQsN#I%wyK za8;xOk7H|#`Jx50RTgM!pfFLm{_tYAq%*g;elJ6~d}`JH%(#T%czp!eU3xD(eqz{x zVFpP9M1QvewuAf&a2oK8t23{_cDMvh_-j;%-ynD7g%~5yJ<|Hx@pKB5Shg#eo)+K_ zId?pf$F8>F`6aC*;n`vbb6F#I|Km4k`DHs3iS%b3>d@qhVF>0PreamqBy1EzdP5i| zT-g$UDcfm5(ofdc8glS3Q35B+7Xf>>U8`Rn5ArdZIZ-;Kmt)&7o&((e zkHI&b+TDXqe^UWJGZ6=*GR}I5;y!D_iH~>E3`R?w-8~yU9_)bQ2-U(!mjU0_R_`QaXUgSV7{FUHYnhq?KY9y za*)ZT!yBA{$}N|oSASlbfkuwyPO!I|k=vOc7Zc-Flul#FRrUTs$<5k@g zJS5*W3|{uluo?PdCX=OC!Q`%^(Os8@H3-w3 zz?t1LaxNRlHKZsUapt)HP1N|}7 zCF_7`AcpHaNiOo81M2&5Y)8SAgLLjJfrbSlJI@lCHKg(8#ZkM_;F6vHEV zVRl2`X@8Pt))q#F ziuOiEQhK(;5=O2HMvnh=2bHvBW@XX6CX|oVfd=@0b9M&%JR!i|e&f&*nrkNM))>j{ zN~uT-XIqFZMvw9Bg`=h5lcoTU1NlatG8$41Y`;nnp2K%Q#v}J->&4_IUgtOJ+^paH z%7Gzieq2e3)mZ|51vAESYYBZ;Y6DsrkLW-)Kv>yJdqk)kX+rZDA!OQu6ulP?%#i_@5n4n`=SPZZSA(+%9D)eS9fXP5cK2r zsGTpeYs^`st_e%T7HlkOu?X=rjr*eqv*H6&yyc;|K4(|=;Puna=y%clh;WDeI2vk8 zl>F?Q=(lOdf+2V&3b9A*#G;wxnF_Bg3!2HgI(^f^ANiIGR_hkcaV!qzQTPYW3?UA7#PnplUf8pM-|ZX<7V;8= zo+J@LN#VN)%DYX09KdI$>?p~?41YaqdhD6llqT+@J?8^)QTCbRd0><~p=>950gpwj z(g0FQf|U2fqO`?fRvD;YOAMDFJL+*@)=aB!czCe!w0Tm+kAbeJym_oYN6W? zb)YFlS{u4MmZ0ektL6?ehAwzcux3^;<}8NT zt~QbwBPQ<%Q-JtdkzA@XJ?c`a@+ekb!{rxsuT%4Am6p|Nq1$3ZE9%z-hQ?6=@O9AnDpJ=GF4|bu zmD{#nmTCDSag5Vc1@!Ga!1XcB0LxecM_m{8UMKn8`VX^EF4$b7=@I>TYb@niTS*Q) zTHu}+y7PfuBCY(;u-M7-auW5^Pnm8~!6i0b)p5g8vH^F&OZS5ONUNnqdRA?H&D*#silTdox(Xt$;cwy_Nuv0YG{Dt7!wg^cWNsGmkC8>Gnh_c zep)5VDVtm*Ar?R>2_pN9))q)X4?;HN=dcgk%nxy>_c~QV0_dPTd>%V$pOer$a8SqZ z(9LgXRhl16pqub-d#5BvI0QUD6q@+$?X!Y~s6jf2gXc_%N3SQ@NTFlu@`F56hHbXj zR04Rlzu7!(`PuPK|zxbBUzR&NX?1ji(d4NK&kwXtTP!) zcf5r}aj!Q60mGMFEpiW}sHi0M<;FVKy8FZ6;VPF8sL8b;l^e;B#*`KcqMlGnxBu}xXnCaWdu#bG-wjppmf@lUYAv70@9@x*= z138F;KAuqSqEZ&|KBPg?teS% z{~Is(oh`+_q^Wveder|s`gBFDuY3=ghiTnwTE2=f3pB^?E;c@r*&K_IfdO&l<@fAk zewdNoa``Ha0JGe#$z9M>2;@NDPb3}Y>&2(7ScH3N4Ra7YYAL3}iFK~W3?kjl#T?$WnoEk=nW@8dYQ6iFr3FVWHB1VGy&ww-{Y&ibCy|XOQ%y z2~_B@00WPRCpX6WTsjB13b8t|AKS?f_oK9-Mb?Bst*Hdqk$_rdF{Y4|@=;HwV)jzb zYq&8ROlW3V2kO*qaWcylX~g6Yf58T-SN{sUT9Z`SMw*fyy`*r*vj%2hGVx*q^~qE_ zab;Wv>#z~$#Zrz^;~KKYoYU93uK|lK0woszh(a#D4P$B2S^!OJft7hlcCyj^!%Ye~ z$MFuIy@QAXjFyD*IBjrhOI_~gP@R}7@n76K@N7c#KBMNRJ;uN+v{{(q-E}&_OFd2) z^B~@+P?T|0QXC~1yTRzo5N&Oeuy=2=uo;<632!JEO>v^L0vr(tT7enqWF+KqBt`+3 zGoB?4C{2jjlV~w7G%|``tWZ-yUOm3wllg}x&C;?LYSzgF0&SD_cb@l50%C(*wuUwB z(JJ0jH#ha`l=jf`d5!>4PR$~q3ZpO0TmA7snNo)|;FU)sIQT5{sNagd(M-@oGCb>Aa_FzUj! z2c8sc&^CQO}3&+zJ!!Q;@eY2kKwiUjZQ99tJ^Ae+n|2m zcg7+G!q_U1!HDjwm})n@ZpSCiK0a=D$$S=u}_*14drEceJGW^PfQVyxB8Co6kgMDJ4fOVwP)uyJJDx zi;N^Q>>(l+taXn?x?19lD|ehfq;j0x3CGFF?0^eB$Wh#eYQZJ)n1`t@D}mF3T@q$g zu4ZLz)4(csCbC6j+;Y(!ZRN)K`MX=Hc|t0PX2owFe_?FAqM5_0A~Mfps-@ES3JG(b*or_b zdvUhZ=nU&u3o&+wIos?zq(tdbNO_j1a02ul+7%k-70T02&DigrDU5{5H)0^jL@ct6 zlOB3x#-{yb^WnOC-sb1avB@`+>{9fue&iQhv3{d53V>8UN?V1lwh*YwepGxf3krv( zO0|y=7+VZLFpxCzR#o0s;e(D!72)sZVtEPyebnZ{HWrxJ=FXfYN!h$%B+TSH0bNGH2DjYaE&`#jwCTnWx$_^g3 zzMtW>mzr!cQlxIzbrhztvBm<6lrtC9Q|eCDYz1n=NP|TL#j~cvSDC7q<1(SYMCjc7 zHG_uvl!)@vm{VUU-Ou8H-0e-VR^m!QFG*g@0>iGfY--D|ZCOCvmlqP(+#nP>C?5 zxAOd8zr1j?z%I=IcrI!c@0`>t%rzJZ)%=q_WpXS>0;3*Tqv;H_oXWMO_2T_@Jh??% z89)bg1kq181@okSSKp3_QOK~mM-o%9(6nyg#(sj(1$6gACa`xx8s_#HTxgs=!1*FQUsyCz-wWvXI8c##ewCEglwwoYlb01y$nAvAxV zwxFqfxM$TV=ZC9jda?ttXFUutjL}_zh@cJ4RUjt@rl;S3K2F_A{9H^gT)F^L*I-sP z5}Ra!SyJB8x?A94XKO+Ii=?=|B1`|E*~J&!=l$&DCRW z(8=4Ta#Nc#O100g$+AmdC6qldI-rR-ER5JNpc({AHBabmtcJ>2Wepx{mpEb;TgqT? z&mL(tB$Ajns;iqd0!J{dR&1rfdV*CXm04;g&n!Z}bd#-f^TRoCfi`CKLyq@s?9KR~ z-SzgM{nDN3W#%-~J>xIxl^mQCy}R5}TV3?|&2dx@MWBSixJ1hIc|s3Spvu83H~N)@ z=H7um9g0J+V}hG2cH*#TGxBae-By09t6v=@a|Zc3Kc3W{7;lR|*{DfAvI28P$u%>b zltC-J|60cj>p(sMkWR zjzqL80W+EyE$W4)1&R6^j_^d$$9j3{_-fAZ)AqS8*qds}U12(!LvRf3ifbska{bl; zs`feJo{AkM`LcvbQi_y()!b_*ZS_G57HhlPdu{Vw#PG%826Yze+Zpif>56MN(#4jg z>nDCzoJ?aXN_HY%IO#c#;jj?c~;4> z7^(YCbxMXKqg*Z-z84{8B|}3iYs!X|wQIo{c%M@)sd-lo7F%(=R;z(re(|$`s_U_a z>QeLqoWbClpTm0UnRfb_#8w|q@_MnVn+xMc_(hWInq;U*t?Yt$)liMxl#x+ts>tPv zf%;C*G=y#v-O%Y^O`DJwufiE`{B&%vV{8U0w68J%y3c$H$*mxISewNJNwHZy*$EOb zu=4tF5th%2d7!wEauzbcCZy2fk(wScW{oqU$OE-}(k}w6=A274YBYs$|Iw+dq*oaG zj_hQdqvcMoLz8_CD>(}#!-gaJHw!8$rd>qZ!dWRpqhixlql`*{&%_X|VAn%P3Ji!1 zNv>qjB`8E-y#zj$PnA6M=l+jddKDycO3-Z0;{$(z1$3BpcHG4KGyqSr+$mpwIt<+x zT6zZUrwZ2tU4Bf#h-@W6YAqFRDAWOo{!fNzgAj8Y`_7~;S{@Ku3#h+!Mic0^UCZ?v zXpuB0HTbe-h5~wy0mK8+_`}w=-@gwbas9<1#dn}iubxr}`X-s}=Yc6KzNuP?&_s(S z73M|edT@*eTh3`%Wn@lkDx0e^H}@7Us;Yq_Ws(Bt-0#*u%8-5Tt)JcI`4r|!E;(76xvU)*est@+O^{p26%%zkekA^W=wp>=0veE*$xMLPQ$nqV~?jAr^W( zbtvP=|I&&HxgQ`fk|1oAw!B-#VeuW>Wm4HFRd%{h8GQ`*Of?`yw&N~dnjjy`ry{lD z@K<)i)t=wb#TGZaAaeI2Sy;1-a ztuU9yrA6nMX&ZgTgc1-*DDLVCKp-|o)Lu(J6I4~hIS+R}#!!8&D48zYbr^{0E;YDR ze6G@y6i#)5AwXeeqQOhy*j@`>0&(an*I-Mt5UF+BSQixtgwWGUO9+dJ79-DAIPcQ0 z(+@-32YI}y9*!B@_60gQ5fOMP_;w6TXSz-pZjK0zdE*Bftm$T9&pws-6=yAqqt0M% zl23%4YOqQqn}f6Rm?72?N4Ys0)0mCjiet- zQm(`gPd=3kaTzoT^xgLwlOCN*KIFIPTR1Z8K@!z5s|>x>g*UvK@wH2ZeD`Vze|Vyx z?U!DeP-GG>e=QDjel1R&TA|X!E%4%NPuu`OIbdDbK#O!CMxl~8zCv6>FQC(4ZtCoS0!^pAw)b%iNH@yt*xK8=qWt4a(%D zi=#tC3>{$`MYf|DhTT28`x9w)=e0-8#d2}uz&7~S7V2)>nOsp=NEVZP>cIQ9`#VImr-ZnW&A^my`s&%IT zd`TibFjP-&Z)Po&ACAl9r8W18|Jab>izDP0TPq&4Ekui!J0o(fM!%2OY>#BoO}##20!VMj|uqX2f?K5-y%1W@p9aAStDiE-A*~peGkJdjGFL8D#=#fH~BKS7E zupgCojgVIjRs&K%(Tgzkt&_VxLBmZabCK)J%Z5E&V19-hB~N}@(9EejVwYAnP}iPB z1$MZ|O`-a|OS>5kYmx(Z7aV&v$-}=e5Y-id{y_6OUs!w9O&=qWeCeyvvW-;}4jOE3*7F=cR`9X#%?xqy zlIbNrA(oKb#n|hGK=77@=Vgw3@X$hI1R_Dkp3Yj?W626>bZ^RHC?28jeL8JAxi7@I z*pXm?&kcUZf_|7>tDp=*Cc$c~>FP1OIGylcQ;rDrrVd!< zp8LT)*}3s-CPTR8pF;$JOvoL-)tfRkRV{opvmNrRcm*#&ZUV$5GKs8w+Ont2Zg?-l zO2tDj(eX3??#U7c;v-LUdd$ZJCCD<=%X8Grfz{6n7rkMI8y800WP}J1H zDB)qTI~P^P)`BpSHE;!;B~4LDz{RX~?%r`f~nP%zi6E^f89Y~k8h*AAVhf1c=* zphwJuB%eiUFk)bBB)GuzCvf){a5tBurOK~W%<0}%!jpVpv=7eps8kVJ8$DTSWdQaEcJU|AqpF!%;5j2eYH`3)_N$oAR4Z3XD$5FWA3|p)d&v ziYlq{V$jdD3BP<`rl)R1@&?6lh`6rhyg7v4d`Hc7i&m^I%_|LGQuof`@041IDe>Y;(x)1LnAo)%ItmDWMym;?ml88}kbhySBj( zmGShw2Ym?RN-!9K_%`gH+vbUaB@Vy)AWe#M7BFnVaVEK1A{SfyB77|P(~T>ge%3=H zhkrZ7CAE11OCi@Ny8Xd}Rz!USXJc3LZ z#fMmp6AGt|tVhD;7f7e{1wLj~*O0Wg099~w3$$Z!gJ#>OTjp!zac&U>w#;0!Z%fBf zjsy5wETSrj<9`@LYbU18Zv8s=X<&{C7R@ldffe1DOZ-!-%QB(Rzd}`o_-T%C;Je#| zEyT%-I3{G{lS&Ojzk1H{&$z^}SA1k9ef%~T@Rwjih!MlV23P#YyV+O8U5RSHlPKA} zerVT1Xe+^On)RK{SCzn4#o^rPJ>7`dmceZ`eIk;-L3Zs6Zr#uq;RSy3fR9Y-Kg1^8 zGrLLeBC`E0oyU0L;1LwO<8YqhXxetb&brkUqr1^`3r3q7UHgp|&)@!&)eTwrTCZ%U z|HNwRPVPq<-p;9K2^=trg{@9|M>-iI!k%Ul_X&70pdCL*nRSLUG|3SfjF5yFF^+06 zXMV&3C7jAE3p3^Q0ijKT(BW8qJ3N$Q>)}T(LBep1=CG{+p^|OH1a|t*oDWZ_Q|@8~ z=HA4Kw}>=C7AGVvnDL`1^q$8Aad5p0wbIrqxEesqEPYqf*U7AM;`4S-+UcOAK zFE`u(aL+COwh39Eq3~KTe)o4re+*$o&rXa}%)%@sZtN6jyUxJ;4!E#)AOFHFGYIxb zp0BTlW644?WUeVN?Yt19)s)>ZY^f$fXvv>(%)C%)Mte;!E02n43BqAzY5K| zpsCD|AS;H$ji0*g=*hEQbv5ilvUd|Z-(PFUUTF`c75({QR+ zDJMA!zKm~G9!RAqa(o=+$$w5Gk)BL1L}VoSZSIO6V3&O&T6)DNk|<3bj!`{gb6cdv zo)fA5)ZM=rd5uQ=l0rmBa3MyV8RWZd%fgbATf=;W5fQ5##-*o9=) zUBCK#fSzj(q##;=)w?_sP$A(L{CfBnG1Tvf&?BMz2~8=C-}!%k@zD!hX3fWIwnGno z&XJumo76AMbyu~I3Vkfpl@MmQCT$w?A^`J^?rJ^!5zjnUmt-|JB;?3=Wy6^`BaJJ< z4LCNN(jKl|Ilf484ImwTlgVpk6Mg(bxiJ`f%(JyO=nBc*r*JHnA#LKLclb>4)2Fh4 z_Db>vaWco|#`1+lM_}`i;)Oi$2n{=X6IerM7jwZigJV|1)|$2VAhUN8#Bc>PbM@dlh{D(>nUnn4%Z-zNutTI#o`rFxhoyI3uZ{D53L-FCpctul?$f!b0{*MhOF~* zHPhM|Cz&ElmijD?Km;RSu(l}8ABMEF1@hW4hmM5ZV<0k3T;b%kGw5F0n7brHDk%cK zB3G^6Nd<A6!Wh6w7E3r&_P7x*+JQuzOePS~elYC(6wCrWv&TSjAdJ5iFWFLJ3Re*b zAFm5rlx1v`CDoD9Xr3P#;jBgqj)FC!SMywpsRym=D7xFuOWlK5=o^n>{%m5_Y5Y^h z+1&__1)3-V8%DNM3I;m-HWO>vcF6|E^v`xg(c!zm1o@gwf?7HIWjsgLL4=~8*!NeH z-xq+d9<7HD#C5#9VMl%t$1oKp4-9MCI9JHT@Mq?!4m9#^!9BHDgaQjSM8@*UI);cx z9Y6DX2j{;##&aat`49{aIs^h+vI#hugMLRK-${Hu_V->Tm*5P`$H~=Lv3s6za+6pM zpoS?@FEK`!aJf^cC8{=x_LlI$X+gW6l^4=Pu>I|-*_BS~DHs>;FI)yj(h;;|`g!`e z2Yn)kO?OSwG%mGwm|7ghI;+&?k9tL27iW`?(iUE^dpA=_D+=h) z$o`owwN905iKdti6>HdwXW|E2ixJqGm*s0%;506bXoK+zD9{XtLc%x9cxw!shPZ*i zZTERnfXy%U=o{CDOU|V|V*c>2x4(pcI3=gc-0NZYWV@uyB8P+B2FGdUfvjoZ)e^T%tYZPbKGxABpxE}NycBMMq}vG6>YvG%ABImGp|n=G`%6ZAPkAgUtbgLd*IL<&IN(x z2FY^6WVr|UKwIunS?*yh4b@gf-&RGV=@Qa*3wM0Ma)V&GVX)i-w%jvX8tVUC5bm&8 z9tt&Rh`v;3x!L*Dz|!mU@uHg@kRq* zroKUWYIsYR_Vlas#l5Rq&tR~;S{uoGepe-07^eIwS^u4I-~k1>lwPN7pI`wpXx_D1 zb<=4&rjqxX^sw`QdUi#uHM(p zG(@CWmOG}Ghg(GRvC4j2X_$F$&^V55NN3VL z%~9#PJZ(tg@a>g#<&EEHE^$Ph=+h=2g zHx#CK8q>+m%$3ppPBYV)H5pJk>l5bb>B37( zeFs+@2IZOj0=0*Wn#bq(&MQ{vrH!QG3jqogh1rd3K&EhM71CRv8wOhB;?Urbi0=vh zn%b}6f}#Wya2JCh~RtFk_vio#c`kM7v<3zJvrBnOeFg^$DTf7tW3{2seNOj@8WuMHfgNCm7}9T@k5NWKv=CF4kkqrLU?*QLx1GUT{cU?Rz`+aVVzQiscg8Z zz>p1VL_WtOy36*@%kkZJlMP<&%`cieSpfvY!- zV5*VlB-?V#Vjbnfab$0dEH4#Mz*ncNBnK_baUQ?j#O6r{KLw^*H`K7W#+b*OTOi`S z=8-y|=e5i)2jq5^WzzQN7Ue=`ZtrZunV1^iav_rJ?LGrKH`(cgR?oI?R>PdaJ|9@R zMLhz+CBXA_3HtuXg=1ui6WU!Fx6bkU?16zra%EmxUCJyM^fRl1&OvT*SdJ;rYZs({ zhAQ=^hy#p(QSTq$%KyDCe1A7o`BzG_ti6%Ap0ggi{9mcfW;WJ;tHyAkuh5HqCsoMA%rYYxVA}u7()7LjS{TbN9-wfS9 zV%%jW*jZWKK3zS(Rt2WOd_}>6Qmv~p%oeAE6C#UY!QV7>1|JA++x4rV4eR$25AQvi&Sp}A{PjJ6(+2q;p42=4sgKE$t)L7}A&|woH)-#~9^zBt0mM}6C zQY+`Y zux~zr$a!9LIog!MPp)!hMi7f6ISYKku7R=@vHkw&rt+ngi8YptXspt6ji{s_-OA09Gvuz~%VgE64l)vT{n+4tmB$|7+dSl{GYwRFJg< z_(?E?qrAx=iFZK2SuN_zbCl)UY?b&d`?LcL5rt!tlO$#PW?tVn20wf{p0XNzb>4rv z-DGRpIfHf$z5Y(rAL)HjydE>N687 zCn7uQst*`pRE`Fb$=Ys2QtD?Q^eXXgz~81q0=r6JyRX7Y4U~ly`3m=?;JNoiTHpE9k}6XPGbvOY&MB4JY{>Am)eR z%E*^)F6UHS2I*lw&_iTQRG2|*#43jwp9;e$;7#Tm-;c$yq&i}&$x*YH8$n;6L$*Jp z@kM>jm7|=`tSS?fzmzeguHB@w;x;fJ&S~Y^2z{pQWsnif{mEis6qLo1KYsdb(_6hV ztp;?N;)_-AgHo=wUeHZG@_-?roz0Q4q|hHLrE-P}qBV0c=No&fgI}G8zgb#c{n1-e zf^+VA1MZPgZw6)DHaq@o5pDje18QhZdrvUEiOgP}Yt}eu0~JTfgTddR+HTn4^iCYPfWj-U@I{_rb?E|_$n{V zvXU4{X{gS^>|k6bgQ0(aFL?wCbb?QfJZ@w~iBf|$`xatmHq%KenCi;_f~Ou0H<6$x z9NV0vec|Wk1~y8da#?aD3>Rw2&g{sby(OTfrlE@V)Ph;;3U^sswI3IC?GplbjWFEM zz-*ojfr7V-gz6k_MZqH!88)ft5kcE`mG2h-`G#LsqpRrQU#aBr+eXPF@RGRB#zTte z3F<7-_Q=Zx6@u2vwPB`MYRtxr^~PBnp?sh)f3{`#bV}pr{4Fxp(NF%p8tW{{&MEMW z?90aTwxTzZ)(#>+%I{>+lH*2YSUG^iM}{sats~SyO6)YZ9iL*y)zEOW)!Bz6{9etf z1EFyWL9W{~BIH)Z?_!|R7Q<;5ZK7vCH;RXz@wI)KjS}N67oqs=2&7q@VG~o3mhec(2XpwZD^%N)QN*@6|qp6T)h*rc!!)Y$}*b-V59B}T0i1g(LK zUa%w>WEj^y9M?-|ngz|E5KUhv!9%W`%XJ?`lls?@5|pbRFM=EwFLSc@)P? z#WBwK08EUAF=ZHC?PGsy;_n8r3%f}A@edN9J=QDP6d)|_6YAd&kN+kCvj4vv9{=;$ zkk|Obru`_OTED3A2c}TxD_@XxwMC+(&+8ME5~ULgHbufK?5&bso=95m$J-07)W)TY z`E%+IFs>Ga=iuw%O;}AnPq<8^Wp;Laez8Ss6!=yIRL zN~ah%VPD~I=sJa93{N^hR2&vNcBFc&sXjitH1b<9gzFdk+{n=&{}2^p!?^5*qWXR$ zdV#uy7U*U1GHG3N#LUg}bQbA3<%R@#%@n8m|3{xyo5A{IPnUDrVThjk}{~ah_(N$Rfmnnb(r&VAs^U87>;&elP^toS#DR(_!CL0jWSgLaxhBxFjzri_(h;AO64#zq+ z^+p@DyB^O%_$@%_r$5dH>q8`&R$s#+-~xU`hH+xeY3OD+TO&FVgBEL?SGXA2%@Zrv zgFS4IV3{S%fJvge5%2COu1dt)1)_=VLxqNMcMBz+8DfkS_A0nF)3lD%in_VojRhV6f_0ss45 z`CH-te-cLjCDizfd=sx^Av3Fs{A;74%rP#N4Zbjl?Qp=`7d%8@50?}zoRH|IJd6V> zpID2&+bhS*o7Q9M8~t?O5t`JS)PUq` zunc)gTWLegP+MI+$DyU@WpuV!uS=>}uf6x{el&A~qKMXD?th=N4v!?78l&1MznvQ>`8b_(vf#jO>3CC|E<*n7TmHB2sx_>~-R zL3(@qyVq0N#PJ5OcloYDtM<&j1bjPYn=5l=F)x;$3bZZFW0+rSo1+dR&@rJZl-CH> ze)tvT+N|0%^`OoR^bn>{yT>0_DqC1f4-wWcsodZ*hZIz>XC%y_=4PK2E{s;>qG-ju zedhE*K2#{o9l~@5_8@>YRr-=JQJV5)`ayuCyGoub8n6-vvUTP8kq0zDfs-cJGP(yP9splVR3?FzUe-{u^OwY3V*2N{Ramt~qKD^}ZdfLvVxrF19LcXE7x(RxKhh{^`xZq3;wp;Vms{)9 zSb^Po^ONrlc*kH1NC`L?ZYvkQ);fwGDA=_A;pu5CExG#reK67H3+}cy3RIJt07xVM zT1B?zmVK%RJ?nXw~PJ(98m%466>+BlNQS=ciEA$M{9 zDZ?4d;e&NJx^D(MJhX&oaet@>1FG}TU`SMahgcJ)9cAb}GIWR{MviB>|2cVvk#9oe zp_28P{K5PlT36MI<$dHc7A|GQ?9={}Dp5riv|&l}D76?ZlSVJTUOR8_X96$b7}=#< z7>!UdjHETvwIr%i`6u@%m_K~|L9uk)aGrD~l#3xOzoqaJKVc6+34-~OvFZc+zI2JJ zsXsb+iJ8eo0o~l>;=t6$=MMS~J2OT@o$IS0_Kjxq`*1KKQZ|mu%n^X(l)|XhxFZ{F z%jMxhz`@wm36~B?=HdgBvu+Sx$~u$-5&&1OM6=3QQOM~GfbtVIbTT`Set4hMJy(e1 zahnjrgp%3frkQSHx!34n*eWN$%(VeYG^4$_tlr~B?aX<5LGB)Oh~{|`oyIW1G@Ey* z_@1P0(R6ZX(IoBQjioBGemX66Kk@|er}D?a2?=chD*wmdR{pPy+}-~Iy{zi#f^%v^8t)fZgxFjUqBAH4&w zk6qr3Mow?y0fZQELVl{SigYVKLdn@P*&9j|l*JI7$$Vt%bHYM?vkOd0h~G@34ww0< z9`obYpWDO3Ov*!^a&EsCR$6U}t3z4&lb{^*M$nZ2L(E*Y?DV=6l`;W%l1^rPpQ&DV zvhQ~YuH088fm1fGLg>G&-+=%86X^NrG-UuZ1Pb^!{?BFmzqb?r-{GbIJOlsZ+keWC z@)b7~0Qc>W3!_+Ky=ZR)05v5f3NH#ogHKAbntZ{uFF8i9;LdD$Ent3}rgoX_3GB(` zuJ~mpkmDeQX3ge~G~v}#%}7^;_T-Xhh$-2$U=6rON;l&}jmNpIP1&ORyFegJX3w3Hq_4HB>v zFmDUpc`jFydKgT-fQw?D>W{cT1-|wvB#S6Ahk>YXqmtB7#Or)&)Oh{`UDgAC?^3of zMj+54fxk%l276ow1tS^Wy~Bo~?7S;yf;!e6Znr6}tXw!i@cCZNMn+fUcD zjdj3i4?X>|zvpNyEMqdkP7;=*Jb^aP{9R6lY5kBuSNLX)a^XTT6LS(~oh2>c!|;_n zTq4_XWv;r;b96yQ8EeyaHG}*@14pfO`j$bwmo;gIE}R&}0`tN^vR9~Ca%kT~(!9x6 zEMN|yucdXG#bv>A8abk~!6Za$(;g>aELvjOX{_OJntQzAzSRhGCV$lBr6?f8Dl|YZ z7P5owZtCjG9cqXUqx@Zuv~!5MZHv;K`2;>%ucgff(P0GLiSNxLE1b@ge?(tAou6Lu zBHAiOb@^C9Ob&04o-DQO(PZpg&MjLy8_g_--A?M>S44rVGC7zduZ0zqbMqoh!B==w5ZH5}*RLpRM{Jsby>=L5yqR|h>ChZzD3A4USVyIwi5XW@}Kc$Q&eW(y`x;ko5~&clD5>4${5K6EIZtV zdPkPUd?XD51Breg!HWB-0<7eKKSv4u_Is)F68;liSxn;-Z-(QC^0@&EnlnM8Q2Enq z)HJf9K9fIqrm$}z-9)8+f{s9j&LBf3Oj3mN7l8tVHGZBcXi!WIT{Ju*jwp2%G%pFn*n;;f7QqII4hM|Q9w6aK|D60X`ZY_ zQMrap9zHy$3F^s@%c>fL$vdZ#Kec?C=~%3=;!I>I@s<8T$yvwSgx%z^9dBeeorwwL zguPW-hQZ}xEpRiv`FK#q>+`EFVkHB5Uz}enhGrj}*;~*tG-6UTT`C`>0>7t>=%W!A zd}yHlwdg=v*dW7!5tXQFTs6_$C`MYh3#`(${kbulD z%TRpwOC_oR%W3EGd1I@=8`d$LKqm8P%G8O3<5hhe^4Pcfb=@S2g!UBe`?eATfH?l} zq+yhLlNrZzy-|70<>1bDxSV*hGTCGOabp=}G&7${Dbxn@8nyRF2c5Eq04zu2h6DEo z!RZF*CW@Gv&$i{NLvKZWyTfx&;`%I#Iz~CXnY4B^)w~pqu267xsX>d<`c}ue^6)vf z(TO*2-MP(gTxMP)+uFAV!-ZTV%&@gZsm#iE5AZcu!Um0yidBqWix(q{Vbtt^4vz%Z zZ^6Eow!yIY6`NHSPJnbbimEz5csGT-GbF{))CxExuB+3UpH04SDkp zZCMfI$0(jIiF`Ei;_62Hg!Eb=`)35<5aj}l*oS0n5lYPI?42uRYBV|?It?dl4@C=W z1+NSDzwGLhRG@4;gptr!eIJPhHuaXYe6r#qsOAld1&mJaCCdwvzmfYT3kwW+TB*Ip zZ$GHq(e@rUmDQ)SDktNWq^vpLUAbplNLXT?FPto8X=99Mo@K>4@03a=-?!{5qgS}t zu`dTsqsT)hyHleu+jm&>!a7^*vHkFB^wW6C_;3Qn7lxF)!)XsR&7U1Jy};Q7dyVz` zRyPEW{+;!Q8HO%;qM7tAq%gsu8BuTh84%)XGu+WL)Uf9r=wJ`Dli%WyOTbPg@ zHsq|p?Gupv>T48POj}@GO7>zu+)=JsB&G;aGFt}U+oUgnc8)7S=t78K=XBMIohD)M zbcZj5Ch2JyU_?IFDFQ-u_yJyCnawO1#>Gu|2OSDIr{MDL5wT@>4)k5$GaROD%(^;)Xo`B&%@DnW+_*ya#Ihs{@~@3Ib`5Eop`SXKP? z(Ob$87v$X-IJ!{Vrs?VE8*k@K*o`pcEL` zSR0x-0tP1knq|pL{gG07rY89{iK#=z$!AT$&k*^dvo$CY`#r_*tCMFV?$_BVS16b) z+FO3}^49?>6jvk=A^N2_uQo_ba}I*Xm2tiLxG^@78JpSpg-c{y5D4x@l`(p{TEA!N zIy*QmVx9|^ey>6EIA9fe`KLi-w`(z|uU^Nw_;j$aC&HcUX6}f-#`4o{tYZjV>q1+R=`gn<7)aj3p^Srl7=l4f0#aQ;QSKN=W=GmtlF@bVF zHcK_y!$ad%oj8_~a3jt;z|%~L3`MOLAMeW`z1;P6U!jdvgnX#ATh(=+V=fRG@cYr<}u_WEmZn@rGO|HyErG2U>_&S3VuX>9)_T! zex!qCY*U=a{;H{ynfrQ#OTL;FCP&2**%cA}?CN*scLObcx6le!HhrKZyHG@qdaIhH z+Qm$T$bai0?m9BQnZy~u$E89teyK>WU%+q-%sjFHu_;;|Z->K5w!x*Mmg7q4JQ|`J z0kclhjj|>d)+zN(f^xG?H{#oEEZ%i+@&?k31;U$|c%@%!Q=)MpXUOwq=0A{Tm1~mp z698p;{0HX#@5AM9FK&Nf%YRcj{a=nUe^lSP!;4mm* zR_1qB>LdW2JMk*D*t-B=lmnrk6CBC&lpE%@tzO{(k#7Nc#n9 z2;hqm@!i)9h0ztg$MPHG9k|rNo{_dTD!(f9HiE!sf+@1!Xx2WJ1e*m(<-B0{+eL<) z#2Ee99_2%}cKvw?`JMTAgJZ9~zk!{?B-Vn@()wO~dk$fr#jm*RHltpKHLBkkcdmqr&i|HXtbcV)n*rc@#YQ5hpOoQ|({01=x zssuTn6Ugal@iR&f0W;mu6Y$U|# zX5@(u^D9_}uzH|4_*TbFbl(Yr-xU++@)9l>K^*RGf~MU4U|_@$hEEC+`r*;_4cqy= zKYkE#zr#?yoELFqSP3=e5`gD+nzBUk7no2ny*OG7WGgjrN*#l|0bXg*HqOlwv6 z&ywnrg8#bc*0$1byy?YtehVw~cm2Ans6$Yp(4nc|#qTC_{@{7+vv%aGILTLnyZ0^p zi*?K;DoMMs3m1jr^4T5vi+k((g+Bi|UG@&kMVcP}`JHFuSM7%`16Y#GWM`NkH2EYC+kmgmi&AIQE|$aFs0N0*@(QN~bP$#|!U8!1}O z0@p$`Aj|WbQLoT=XIt3YiV2u17qx3pDgY)?!8jkXIJ&Zbk6fnkD_{IMvNepryP-ey z`6q2OK2_0hReArah<>KY$i+Bpd=i9JbG-29z|UpPBTs?{6o|lHm%yA_?^cazGZZ!? zjtKi&<{nh;7z%W*s-N68H0|k#?>y27m{Ifj^#Uu6yua%sWncE#3eCp2eID5ac3;+P zKumy4_NUQ3OW8!+eca=T53Ao$_EkiRJh%{!e}^JtI~+tWU>|Z(I9T9dS$BAcGO5@? zq!w{8?%hppFt7vjG~`Tutz{?FET`k&VKA9F8tWe!kiJIypr zzu;D}W}USk7e3&o4EzL}BSpEg>N6;aZ=^3^Pd#8#@JImt+jQTiVc6yI2zsL6K*~d# z=cHTaBm3lQ#>L|%?dBI~T|VJ2oua5Kk-)vqtqwV+{HI)ZUPGSpc0CY3UZeV%*jSunoF&y zOZ?B7HD;CP-kJK192Hyut|h9(KG#m5b_**r=qF~%8r(~(b7#y>v)?D%jVKE-PfN;g zT_(xxeW80%41+T+6I@xe2Dnyu2h)qb(A7Q=v^QjAoKb&8g?iEOMbN({o>3{3ov*eO zM-QdFED9{K-#QAY*?T7cF>&Ii-~?+KQey98s=`V$5%(;<%6|9*ZGTj)Aoqj>0x1GT z1*;@kMcewql${i3Q&$%i?QR8Gz-mFxq1EpcV!N@P`!tHuN?wadGOGSWxu_Fy+zmWC zQUz0C1+6%Sk!jwUIiCfCTtV+=?bZ_Dc76OwEB)O9XNld4@37UPC80iX5KJ!c^o?saVvGtiTTjYH;xV##-bNrL#tCMk|Q>=k559V+xDpU4omdo!k;)^x_C>z zK>53RMeP@I`hiZV5uBoUI0E?y-XN*&FnZt}OZ~`gjfq|kcr|22QNfo))eOz0XNtwy zJ6&r=%?MV%^agFFcj~9G{FCR`;z<}KL2NM_&*E3vXu&w2=XryKWH^H!5{E* z6hZI=E#^9P-=^XD{SW$6!23Ivq!{|+4P7RU$!`?iK;FgQWhclF6<_N+hjZI&v2=-j zRPC3Nn{U?|0ZGp5UYnoytlq0+Gl z0lWBvh=Vo0xXmxhz>L^x5ox}^_SkXlJO-yy>#X#Y@8p%rI07iKL7j$~{9{W_?pul& z1&v{5Z4c2o^c5(fx8sx$pbHrT$n8_>V{Aj@le1|J;_pvFF3|1Q;RDtWL6^}-bDK=aHb(8!qO=On4UjD%`(YB_BlkYa>YOmp~8#7}#=1 z1rh`37?@*iQ(m*oy)!e<{dc1^Ra(6aW(u4We9Dy{=IqRJ*I^(`6Szt0BVhZv9c?A4 zrzw*I1jnb=@JM!TQ5Y2-{MyiKMTHg(ty1h?m9&PUpEj# z4Mv$CPy-y#Cf;XA4LzH$uA$LZWwDpdv6GN)&6A9e9=H1@x1$Y$`hlTwtz|tQA$aMs zj)2;`7tH-AR8ytZs4x-pS)>_)F}ZFwuzGG6 zTkKm$+BLtkbWoLddTL70>84K0e@uIWnstv@K@mhMH#Y4zRordWbq6m$&)J137ijt+ z^DO;sKPi>F-N&4Z9@@g^Y z;*3u%b)&zSb$P2yWI`IuP&@1${YN6Jbd(7DhF~M>L{_4Q)>Y1B_92s)jkf})5Cr4a z{tL>!YoLAx{`of&dI`=)QMMJ*GZ-DO%Ks3YVjtLokfD4#|AD0PB~_ey2*{u$^%EHdfDv zs63CnK(c7(a8=IR!DvJ-A`1!aq2D}F{B7s)j3JZPAQ)?`Yf>?^m**X)6Msm?$2)fejucgt$~lMK|!*E{Vd)gd#JyA726;2KXzc>kn|T|rl=l&&*NB&529JS6jvyF3?K7v` zmr|N#;;6sNR7avs$GO)=TVJOY&@OC!c@UL7$7s-9HIt~$B~wU<0QV${P_x2{uOfgu zYRL^Ip;iL@;j=(+L7qY6o8ayif7)s#-j4XekG_C8{sH0D`?cQ*xeYx%Xv#Q_5W~!R zHmNQVbG+NTY9q7H>ogMqHSdvJI59^tQW7(8r+6%){`rv!&zN5^m6St~B$|T**;#)8 z_u^%VZ+(z~f~TZuuxi_0yj&4T>82i?cXxO%Y7%|r*`imL(@QX5;WINDe~5gfDz|@i za3`s71Xt2TT++W`O+44jpxu%RnHP01#P4@^Yw{7*)`ga13OkMeMedR*Y_*h?Jm_yQ zh{0s6aoqiC0&zhDxXVKGck*Wo9NIdLAg79S0E0jc2&CJ*GErzW!D!}v&Ma6W| z#Z%1|cQjMZ!zpCu)gij0RrCYH1lMfbt|Eu2aiMzq;Lr38g0(6*Wi}V=H0~Pd&x=$t zm-=-V06;JWSL z=9o+8g!@>=LhW8_YS!4e)PQ3pt3#+zrrLq5W37HNA|<;xG5&YjWzofzr?c@ z?hvEUzD|=iAAWTOiKWnvS5%?L`N0g$3#$@Eq8Rk$gf5q$Sf9lkG>L-dTNRRl{ImWS z^0DG7#U-G#=wErn?L@3N&v6i05EaOyFLrAEC?M)%HGZ>+GmIF~a#hWvb&`!5d}>W@ z@;;T5(%>K(0g@77hKB|IjM*}ULl7{_0U0TCI0_wy$viF#M8bwcRfjoZtI`F6Ov%;d zMP|)C)pv?_b@9_Cf-)3BR*TLYD2nPk;$tICEk^KqI<7~)ztkcN94RsG(@re42Qbn* zz5|W{pO4Z2QGn&$w03_}a-AdVDSTG0#g1wRQsI@M%aI-QB+Hl0{~zke1CIR<)TZ3I53B%GB}&1!y3083ph8lzgN=foKLIXga# zie0$!nwp%h=AZ?I_Tm-j#@rR^M!^gPAcwvi#`)xDX?vr9dS|ijD&kZxs$n>Hh1sD6 z>qSlW!A-zMKMYEvYWh6N!WHg@$tKmcL0txNL;#$vMsMKJN>CtJejpyoGh*$`a+?4a zA1Bc?&qmOI9!%y|Lx>EsN5CWswrWUd&a&f*vt~596qHD!m_|bT3b7c6x%%)%S+Y2I zV4<3EU-C4}wCT|TP@|nXQ;L5Fp}N9ujCcNIut7LxM^_?dN6#W=N3;Z1$NW{g3qTQ+Pnxn0k=*@{K8E&uO@h4~GFKp`r$Jzde3Ax`F@ocepbMJ{#=lD`8*XhEF*y zZa(wHrJd?=@Ww$pK@x+500&p`j9P{j#VTG+E?aTf708f7&@kU!Vh_&8AvZ0ioSf}? z7YSEEuSS|tx$fwedO9{asnZ(I=xtnTgzaW0&~uY$(tAZ^*G#?$$xT6zGM(-U5m&s} zi3<-u@3lokmc#-*Pl)X|nob(y13ib)lm3)K^o$H?C<`3Vnw{_TMDYm@<7zlG39MB+|^bxbp>ts z@Yjo6WeVs8$ZI5bV(o`Zap${rMfe}uMuM@2(4S zpe7GqR0M#Tz5z*vnugHo!!h0qnqG1}*xB$7bHP%5e*^PRQGpN&@w;$KCuCHU@$!at zAI*1BJGg;lv_;7siyae`n&1I#myC;rbwhyVIumbJ2@>uv@Pe0nIKZ+``b4-ii{ty< zXog3-(F_=FF6^5bMZEpP@9J%RB_|KS(L{s(_c7yta}NH00SUT)0f~QdcxLj{06>B~ z5db6<#iEuPPX2Is;N#_8P5;5+VXeUC{KMfv;mQ*CegSyS{)HqE$`=@Uco z(!MVBD+35TXjx&eC|uyOW9L6Xuu!-R$y%c&PgkKi+n2&}`;r8M$;84V@|=6jD{?BP zi)+=+!qV*kmNT^VjyM7B18dXcKmx@U5g@(tGUnG5SaAO5x7h@S%_jO4plc%@JRlNo z#n9V63q8&#G75aN?EoObxvKb+`NN;Cj47pE;UhvMV^xPpcaq8$7OtZWvtRcJ!Vz3+;8R$_g)48LgHTpq*3Ew@Rn`_)aYhfg2L%S9x^gxZuneSnm+@x&PG=2+VD`}66l%a=j630)9>WZc3l zqTg(gW?alaSA%c}GmPMSts(o7#i55wD7CZQAsa5<4$4g;W<7088TA6?NQIWAZ&oy^$9*)yurWW`3~)xX8KFSuje!0p5z zA|xw!KZHW|h9GB38x=Lz1=|-cHErQF@7o=#A{gZ}ZL^(C)?zFiWku{Xzp=M|YwD*FUmH5$92SnX*(gVTvoWkv98MT8Roux8b3I~4qwByuC zsH;l(P@Y0-E4L*ls-FiojBTu^$3fvFD7)=RAE2G^JtnOA%BN&@-#^ncT39M!zll~e zYD?$V3+US|@Gx5&f%0z5d|sKA3)8H~+Tn@@g?l+o2?HH?G022zub`n@#~4K$=OvB^ zR8`8k#v3x&LzW3`fqOfx5lH!0^O41lMF$-VVn$e5|CjDY*_Ny=FPKQpJZwwR~ z4R)0{zj!Gbo7(=cSI_T#<^78MoyVU)2g;iA7;K)cE+5|$9$#j8d0U{fzFrPYg=taX zhVgY@pa;-VskrN-cbh`P>DjQSRLQ`;8n~iI7MVt;isW|zkzc7ow-7}ROOXWFq0b!f zHII+z@%v*400-{bvO!XY43k+lJ19yPyK4GV%3fgC!?-*x#o@~?x`^`T6jHBG6lm>c z85y?K?O7tqYQ3BCnI;bwuZ|y(!mN%$%w6iJmvWdi6#xbS))S6r-0d15ebOmMc4ujY zs~bVq`{CfeygoJ307Fw6pN@rm0Etm&CbR*;$2O8tda?9Haut0@CIX>~$2c1C(t+Vl zrZ900tZ(Ftg-gZYjDw#l<7k5nL0!DD_0TLgBjXV)r-J7ILN8=qKhFo($4`hMhO;tO zKDaiB*tQj|mQoo^++O=N^u)kBX=ycJ^~!OoCs61Rrw607VaEl#nuO?y18hz!0ENnl za=j*f6}33oLTwJ+Q({a=NZ8x5-{X7wFwF=pVa(X}FV7J)1wJ@LwcgVSSGRS{jMR&P zg@#iiF5&w1esMEXD-Ct|+`X;Md=+D^iCRNHW0PODXYfia_4GLigQ;|y$<-j1K1Y5T zsbqU>b1MT)BHSiJ4pN*FV-9U(vc&|FI}^{XuQvtEG**kW@X)G=%Tci|i~}qR1Jys3 zQLUpnFk^H_N)P-2w^&X^>}gwM8B?}^U5bpkE94C2>#zWm>$CvoCcfVACLf%NZMU?k zD&-38E-@I|9pV^}1sdpG-!+S)?okTu=_;Jc5xu~G#jBP8)+>@xCkVrt>qe%wa2NC# zX&~bIfKVEkcHLa-w@;&?@8dGa4%@Mkq}+R(PzTriY9y$Yh`h;!{_@8z1LMC*E)f#u z`Hu+_gJ~OxH7e>=P^Dq&FB=&XVhK;Ogc2u9jXpMD6X849SNa(Yayn>0-z zK7IpXe&&KnY~j7H+)-~qF&u}NN#Jn0WZb+qFB3+$TVC~8 z+^K1QPfj0i24dC2(P>R>R-S^4x5^acBq!(bJuFfao=Dy!4w_RUE3_3kfs6~^ez{`e zX9SSK{56Gnd@u_fO$GAakpwLW3w__AM`H4dnDQGYf3ydgGVD+858uxQY4P<7nP$Um z?*&fL*aXc5SP+^-C$z5E4nik{CnUuCE16ZorbCB1Eb_d1!v<7m7@(CL-JyYp(#qj? z@P|KPj_+XAh1yJ$cL-!Y;r8x;VZ-z4i@*7i_D{OXz4<;idFi|bXqXzX<`n6DnSs_# za)7bkWtIPCuoSdv2WCcc$k`mT329OV^!S7aZSwX7%8OW_?1}RKW9^-yGws%`-PpFB zRBSsH+o;&KZQHhO+qPM;ZC7@_iM`jHtL@qTwc3A@qa2R+e(upn?^nk>Kwtrk>h#}| zZk!Vzmr>Fr=E~0(AL{z!6GZv=Q)BQt6jG+==Wf!LV-#zD!%!}YvyX05t{k|Pi=G3a z*@N_px6-JXbU`>__l}8baH&-NAU>~B6#*JQ!m9sOuMj6aVqVyt+jw|t5e@~#c1x+f zW4FDb{2`ttY$CjzdT@31CuO-TVYX$D{~6{Y%K0g`8{xJ#sR{YvY$Aq=eX=t$=$3-Q znJr)QoPCK;di*L8&%Ol;4lBRisaRIzA`Cq7x*0VeA#P4zIP$UN^hg>Lc~Yrdd97yI zkzZ1S?_!egO%ar(&oppZK&8%VR#eHSXW~rmQ~!wAMU}NR3?Hi`IOx+Wo>=1Ta+Hg5^QER0jhnAKfV-Xqysb8x?42uNt zl`>YH@~*xLzkF7=|zvh;O)d$+8l{ z{;&%|Z@&120-IADB2i@C*V{qCJ%1DKChs=*>Av5Mb^hbN{l9rTX8-59t!43-^CF>8 zyrG)4xygGm##?Ai#J1{cvVYVguM5m+|h#E4fFSPPa)CZ#RE$@FK;=C?4+ zkurudB)fD>T*=G8CCwq80w$CDK};?q=TSEe^V4>i7+Nv^;60XPQ|Ro%i{5$S}xc&S+_VdsVi2+-A?U zPfR-elk2K*uBT4BrTEQysGDfl)z*`QQ?a!NG0QagKDU6pC7*El#80VD4%mV@e0aWt zVh(=?Cy(%#mbyow*E5saJ4lgiwnO%CKeB5KG;_7o2bwM7jS6DaTuUhFI`n4*wOh<< z4S5fTz~O$H7`|uWO(IiR;{`{GT$&nH>%}outAqpSN^N-XNZz3)96 z;6Ls-{|9lY|7zfou`x9I`yuL5-f%!RLizH_*s`3rrYmeDYP$&un={rnjT~1%>Bz#y$Yd*%-;We-bhIa>ch}C34F<5D@Kqd_ilLJ}Bu~ zYc}oDeV=vh;(c77F@xuQgXrtkI`6NY30(j_)8OhiL><9_MprLZ$6m9iPt`Bh@FiK< zHMvejscj{|HC^XQ6S(5WqBeP71rq^^nxbR~>Mt2=uucm2Dy7$lGD~K);TU}!r9Vdv zb&)lVQccnj^pn+wtYOGMtMMZKbXA0x!)h(!sGp)BHN!#5VOR~zFi_s_rZTB1N@0z# zbztf@entjXmSj{ka&#se9U{X}T|EbxX~v4{FZ;;~AV1;zc7DURdb=gDBUr8sZZ4m- z$;1HN2cO1_TnhnJ5m*M+2t8KZi)NMkzNqidF_rg>8J46CZym6^KZ8E}eSV<)K!K+) z?Uw&0%Ybi1+0LWX%Z;bORsFb#7c}hPX|W$0|G`1CBE_FiFw3gJ$QWVxC`vwvd;v=; zO+IB;s+3B@k(^Z}z20CypF&+CR?w9bo#A6^v;kXeXs_Zepxfvc>m*y#Cq)}Ajax=1 zUG8Vq`FpWcO_@;L7~1HEL3U%SI*(JN^Su=;rwB8*-zAo_uGj#pO=jBIfCqfH_j}5Z~+!vy)(`X)Wb5GPreij6D}+ots+-UeSPUJh>ZEkKL6)xJ`r84 zWqlz~R+Cen{4deOqggd|Mys>6YM<0&Cqq$-k57g-E}zfsh7r8CtW(fWyjgHftNE6A zGD4C18slx92K>@B8;zFhgJ_kffX0#||1W2iwSAHhEe~ifGlUZofpo|;W`gov=tZ4zVYGz(Mrbp%jRSnpxWJVX|9T=Y_N*KLYVZ{!dI8=~FkdY`mO{lUndALeZRDA?x$N2(Aakr@O>0hBXVdE}Bhj{X@ zcv^el`S-+cgR&SXe1hmbx-lQvT0!P&jg ziB6qCnKo z;%1RraKv@!$f-8WEH<_swm_A6caxcQ0bb;rts&Y(JcYnP?nw~8=~$G#r2*Qf64X{V z*Ttl%tj?dB*sCKIN@iO*X|IZ1M~GXm3s`OnhF@awEDe`IKUn_mcVLMT5iT0hMB)~E z=zfAM^7`q8pZ~ds>*#98NsO(6cx&odonK+RS15++*Ubs!#S7?nY&)LQ*kU%)b-7d` zYHhGDwWruf%I)+z_Zg?b5*79{o>kGIIahf-oRuQjx~#_>z~eI)sh#1J_lmL$M>au) z$G3|YafV9eHayO#zle&*SHdXoi`9puWYpq~&eWx5?#f!sBh>VoWZf>GN2<{&KkFsO z6x`Da%w+SC+75`Q@Nt$-An~`Uo@Vbt-%ZY67g;hG04V=&ZO#w&|GGB!&wQP-l`6I{ ziZ|H2Wjrk`tXG4YR?@n(>zcxBIYCoYW4@m{(#=Bvey|b5hPAQ1Gu!MRn6I#G;+?wV z6BnNV@6tCe?%!#JApm(Luk=hVSLY5US6Q$7Jvu(Xw$SIlQ0Oa4Vjdps;D}qZ27bIK z+3+naW(K)I=V+*n5{7JGZqA*fySHcb@pL6SjP6N7N1$D4xc0`{jHTLyb?B9PW|xNM zDcvd$R8T3y2dTy+JpiSJirvj>aFuh7*gA)V7$9y}@z_8G-zWX>4DldLajL<&V<(G> zTo#Lbkw{@j(#X3(mc@=&;ar2;tIL2{C=sQj!u%mWQ8KDk$eImr7)wp>dHrPN1$1kM z5pfPvR=-}zeV4=^o}^+RD8$l@h*Tn%hYY#mhH~N+k~Yu5uqvq#KFW^Xn01N1xB5VR znEI$=he6vqM4JfDmW?3hg|2Wx8;6wA?YAfx6t|WsVkcQS-Q|n)06}fGWnrPMEr_pS(gWXmuT~!+nkBVqD+y}#BjC>eP+C3w5X(CO? zC2S8&fe%_D$sc`I7igs@WeT@Yz>i8G{}Ib|R%Tn6ejv}|{2{nAjRuRoy%(3Qf~iuR zsVon#&xvY(WP94#UT9C^h|SRFB8D$((G?yCBvkE(!pR<5S~8m z61A)HRWwNN`2kgSxcvc#*~uHAK%t;ljq79F?I3&Jh)nbAqF}hR6HIHX;wqzC|H1p^ zji?mu$L2~P=j)4>iwFPZoRisaOQ^On`ZcRr+G=7n(KD;7RPT|k^9QCd^M>f}HA~HL zfd|3S&U_arAt`(Jg@M% zshKByC39Vf-~0g{m?XGZl6PMkPoWQ^2;3Bh!N-rFH|T3br?CssTi#&+!Bd0>=Uq-- zk~Fv*;IE9D!-pk0ECgNesIe16yZ`U|9|k=^&HkiQg?1 z7w@D!k`Z=h&CD9)zH9F^#Ka*An-{M)pO4H)V5<-Q+{qfv zcs*|&ZaU0-eiIeEwlUpyg^4&bus0Wl`0jFFu;E!2u~?5>$_R~EuLi;b0Hra66J8HINiYMlCzEJT#PjuBJ1Fvpx=p-7Yvj~8V^ z|GKUp_BmH=ONq6Gb80mvq!=f~IAmxe>nxlGwXqUHx7FTlwKAlU!DeNYzSc}`GF{w$aXor_}(Lz>~F1UflL-64GB~ zUdfXrrLUOBI$5@rA$1icYVlm05Vo!cH2#C=cHNFko#?%{(V>K?8_^=D1#_avU`1DQ z6KLR(q=S|09Z79#4L8bBIIoh*94o=>P)T)k9e@73g#Aaw>|JKt^yks*vW$}+FhZ*P zZ^x7-<77HSlp@ADL!vDl$q7j3K2EM&83Ss>iAsZdWo4UyEmq186hZCAt{M{;X4K?9 zDI=?~X)#X?64V5jhPmw1Xux!3#YKxK0K3do!LDEORK;ybCwZzD$n=!W% zox{Avpsyhr5J}?nfH(RZR3#0Kl;F?k*T2y(Co9XcHv@OsWf^3%X+4037hTB~Dm~|G zVaaSv7~{eQkxo?*_b1rkN(|3017s@hA7tbuNBqP(;fIv101TwQY5$@93Cr&a$X)4DJgM)yK}4?2o?Dpe=(ydJ9%Fhxx{w?s1VBq*kmEFJL?@LG6ZnJX8ZcU# zlMI!NnMirK|J<+}dpESF)W`-0tAVt}Ql0smwS-0O;U=BxsdOo`>!_Z`q`8SHoy5FV zac&DS)5}EZO#hnI$RnoQh6OIy@PgDsWoUWTGJ>Ii18W9=gO)=;FkYj)I9I$Ua{zO8l!MJhWQN&t_3(1=n7QS|LNl$2 z>bXk^Q<<0v85>|d+GQwFM)*iI!+v22@mk7RWHvt2eDfyqFIq)M^MWZ(VbdIx4yX3S z=2J&ZJ+?kGoJtw|61c49(8`TdxU6czLa6(Z=DF{d1{-lo$mB3Uz(UC+ixm~J*@#EC zXc-@&oOv^%p4zz%q;!j?@|Zq>oVK7GS}#IKtjkdY zVWdyiCxT!|LAAo2YWt*|IYW|8+QZ7V8b;JwX=28d4roW_j)I52GO`PToZzR*Sr&4$ zpAWv%gzf4jHDhqe1G$6?m-!W`l*5a^a-^r(ioWrF zGTXW6PH)L58;}Ae&gpC@@^_k9)wxu5yXniM#1)oe65 z=#V5JlX5hOI!DyHx5ZfKysgwj5Q(OkpwZCDT;RM4*WFQLi0fiQn?JjP zDW=3>3s6x&W z^ycv1;X3y+5f7@RK5Li_rVpt{wn5L2xpf4MMuL~`Ybf0)g)Rc;5yB^ig!sU{6B5!(vkAbKPnja^&vBdiCt1U5|RN5mjIej=O)tJS;#G3)cT*Ixs)pQv|k zx*&oeX@{^m?iH&j4V@AeGXn8c}q^+8zkTaixXp{uwk-XhTPZlb^rum+KW9z9C327Qmy# z#IN1sHA&f!2RtX!dFlTF@#=fbfuI_uJ!hUoI6rS*SC+IGo2+Msf20NNu|_{5yplIn z?TD9w*Zi;+K~m=E2{8@6*#@7O!oDn886<(n_(~#2&%V``1gL1T@-) zqT!-D-){uSjkrZhgqYd|pxp#qNW7h9a(m^==WWvxQwl*U--CZRza!KF1+Ib;a;V)} z82oU6rV(=E*|7tTeiqTaBUAe@?mb#KX#)-U8p-X_Fk{&gIbO&jn4Rfc*8w2!32k^s z0&t@1b;O^(o~ZrG>3{A4GK<;4iJsM}-D_yI^TZx|J%1qS3Znj#@__Ogo}Kr6N}iQ( zkUIrPTSOQXu5DV2{lkL^Ay@BkYIYYbbFW2UXQiAq(-@7ZG_5;y-2@>HO7}ru_&4ia zJ@(bS9}ij`GS_PGrn+mK(#c%T}tY(#{Ulrl_*cMv)Rj9-2 z!%%t_plMWUTGFJ%jRjo**tdvS?yybW5bgf79(s_*I{lDQ_^6n? z>#rBj)JPs^)2>Hl#e`$qP`aQ^^Z_Ek2=9XEbLLhB@t;}vC1p%s4grCmYeuo3nX+XP z?B{eZSAi*u9Nb=GPT+WSmit_J1ah7M3mF$8P=OAoOo zvS$V>rznQ8KBthh9-CL3mLsvU4t_txqGYViEMdzS)B>+aU#*i}jY&-T*-MD}T@S8f zs$MA@beleZnA%lr+>mX{60@N;!Lq2C%xq|C*{}{c_E2PEbn|@_KX77S70f>+q!X?* zIs7RXAr9BbDS+6~-x@XjojG5dpPcdr!tG}Oic%zt7z^yQ?>6L4pruoKdK6}Zk!{wI zOE5n%VUL4vc|a%=YBOgFp$5sc;{2Irf8q{?oV0W1_Vt4q@5+i0^SMmS+b_#eZ_@=jjUxbE-6nqWIACdgCd#4gPv4_H^r`4d_+k zGlvUGbAl?a|EU)AhzNZ|OvzeLL2Nk&p~2W6Tc;D`@O*(-4!9Tf)z4o-PQK|FpL`2` zKtc!pUTa1R7k{|9t^#Rvo?G(`@&qFjww0o!=olE9ctxKF&DuF-OYJmYCEzVjyMxtVV?GG%!)Us$tFN^si(2VOv z+vgawN#CJ(%bJu8Xh+8h&%Un{d}P$1K@)(Y>8OauO6UXi>pRKx2>!-hbD07+xhi*f z)g&c`pA$GOHMN}>*xtGIFs%H+wS4X}(ZTgz{qC64Qc*pts&+$g+TA~(6zMsN4FI0G zARCdmCFvFP1U^y;H-ew9Y!0FZ7q~sY9;t!F?HG-v zN8MGF!}_WjQbv8?KuIRZM7B~f*5FAiMf=!h-vl34SKk@(NHBInErUS;iE^~bchxMI9+oy7ayFLi&j7YH$1n$fb#5r_u zgReR4>dfwFOWVho8+hmF6aSF31J)T33I2mr-wO@UXaXRo*F%_9JHR-B7HR~3$$IV4;pR~sf zeo-&XP@?$jSZ^rrQB#(V9FAruU_8kZ$chTPGdURc`RC;fYKE}`mjSw-1-jme;R$B* zLn5nSLTyplO;Mf7sydZ%6)GzqFt_l%%(xLg3g^m>;uv{&YnZ6rp7tYZ?-<$|V}?`y zMXO>i;_^68fvMv$Cj1DmaMHe!te#Ap4~nQM{yE7T&qeCtBhV|F^$|wAvp6#cL0hkJ z*wba3AtJK@&j}tI{7o6+tYE1MNRUpzQIV|MuB-Xjn=mHUq;G@Ftt*Q*^aQ}k_d9@nSaV-)Wd zOs@KqXWv)UNx#33y0;|a5`XqD+-~VveZ$c!}4HCHt8@UM_xd|P)2_MM~ z5b24D;Rb=>hJoP*f^iBi(hU{)08AVMNF4LGzp@`S7#!Am2cwt<2dzPI#9uZH{f)S6 zM#SJ9V;}r1KWrnXt(f>GlR<*ZtUYu?r#Be*k7Zs+250sNVvamR(hcx}fv;D}W@fkc z-}F_s56J*4;01!PmYQ}=${*2BI*4l2D;$=}(VMVxm(Rk?VSiTT#VDPRFO)Vp_^Ut4 zRs2>u2l1|%tteK{IFQ88zS2`C4rdNW;?VJjuAY8Hy2ySYUA3ZpPglq~w95hDMqYje{CT6JzD>eb`bz)v*8)Km zpQUz0P66sbBjDQH@bN*Y(Y1jNV8t$Y-)2oK>jJlruuoMspU@~7C${^tA~H@l$nm-m zdVUq?OvDA~iP=vTO~0k6o|${d@c;9*zE!pYOf4TgJ( zfg506f$^j$Zs(Nk`jdWOh*)%d_;LZNb|Z&f^1FuJi;!^bss!cb?rMwel%A669auBa zr{tN4Q>IZfMQ*5ubPPEKUr4kHQ@*Av?O8e4CBXnZXDayc$Y&P!_`a=FaoA%`g}^)1M1y* zj-syLG9D0+GR&-bkuzREs{of492^)VUf3OCeUXiBUbu@3yZ!c6g>6lOVlGplcujI+@{P%=e{zy3nWjC^gh z{^k_V81C}5*Rh(?yQ0d=ZBC@(FA@<>w6_cphppD)fD4W32Z!jqL2v~K0|%)Q^CTUN zL@pH+_}QO4+=DwaJ?Z0eS+mXO4P*<0zJTFJV@DE2`#m@@cR2gSn_kxEFX#WIuF zP51@-_}Q?tq6`-XhIJE+6Z8^<1@g)5wE5}UyRoIA3WtB7fi#}(#1Y3*wRJl*bhlpw z8Z}+SjZlN49@3P~9@=e`2~$~1>*|ytRjw})l)6W2Re^S2zQrNGjW-ozym4fw$r&Wy>oy;tl;abHDbx9B^ zog#ULbkCZGFYgFC0ejtSYHN03CeeM%22y+m{IPa#vVBv)+Di}qevq0~kEzW^GJXhd z_V>}sma3Uaxy&je=Q$`P(uu~iR99%J*jKv&S?S0)GDA^s4HOIsWp$~XGLcL+IXfyy zcELxq|2>t!>MvVDTYMusKm#C;KfCl!56kLnFCh1Q6x(SXfdv@FHNiF3Z|0^}1hqjT ztURC|RqzNlknQvmDlu6ar(j+lC=qR43AH0$h9j@&Cp?*UJR5UDM2*p%Cmc$sL z{ye8uEKv=u)`wfE4Lyg1l`PbY%-+Y%4r?xliENzfE(Nf;0E~CNfk%>#N{)17HA^!% zBc3-HM;uFi3v<^WN5*7G6FN-=3uR=0W!2E&=VX)U6m*Uw`Ar%ox%k2p@0vrxOArPx zo}b?`@%2M~M@}ll=WpIGBIA2i;_tT|`v1rc`9Co2DA^ksN$c7EgSbP&@DJyg(ziqY zn}rS5Ouf^rU!`{4L}XuQS@s69TVy`Ze-4bD9XSS&aM`@EX6(XnSHoPE{;~sqBZ@h5 z@v})iA;^#;h2xmZWZIL7&)fSQP8Xcceso6wj>GaGZ(lEaata48uW-LE`UlLn)f`en z8W(lGLw_BeFre;7(T$519te>xy$A92U?b0rbC4N$HZv|Jl#cz%1;?FQB#lm}m$!QO zy++`ob0M-bhBGDfkqS7>dNmswj4A~`j+9Jrf6L1(>5Tadww+6_m*^T(NwIV9tY)QL znYa}i-JZ1U3`5Zizk^zZl85LA51SUP)&10sSmffBz1K)GcjgO9n^Kv&*4wvJLHBrayW~r6IRMz=@Ojs*mGAHA5u-0>vu;Zt5N}TqpJOZyC3nT7b2d!0kHBS zjG-Pz>3*4?ktL?7u7>CWK`?j0zLu>mXEswKGTEmIwFdN(;V9>E+0bjA)Xx!?KFauPEhHx}+M2;r4)J3WOs!|YTJ3rgthq~1AgzmtgLqq=6T&|i0i(V*D zVc>a%2O0y=L9}`5`XrMgl`j~TiX6g6)Kl35@dExY%y7Xon(8?1NaDKibN)gKp?2~@ zs#lLFYax^>%1#o=DkBS>+$_D*eVx^7q6>CDt_hv@5eKF-Puz8>jwr6U?*?BI7We(D zeJHbxIBk(^N<)M%$9fwx3ip?B`no^Rf2}oO)4*+{-(LWNZ*Rf>KWFrRmHGMSQB;(+ z`G3R1>B$PbKORpKU-a2(g-G*&5kY^fVuIE+f}M{4xck40#7_qSq& zz`P29l@HRBOR**OYO;M2ag+f)HBDk3;U;OhhgY5BQ_dL*HA6iJ!Ak@eIk#-w3YtIW zD(2Ectpqp{!;V;G#XSfiz=8YSeUQGYd<0VCeivY9g)8teyJ1D3*RwCbZ7+B#K*(EV zpR1)wogMJ6Dmq;3#>Gf8dr{{f6-3hn8>J{(7lnhBA%K5peVBZdF|bnLN4~Q@3k*UI z;eUA$pVHWO+v?xfft~f0F z*D3l0^b`Aw+xwp_i2tz~=D)@)M2)P!BNhe%_V#*i|CX{)`a9#_S);wk8W7QxM!Qae zuGbzBRRA_zqkwEHGE4CtHQUcoo+YTQv))C6gVZ#g+LtF@6&^Qjo)Zbd50cp;;YfhlM{$1i>9ja1&_^K|ftqsGh)bkJ3tE3Vm zU z(YncT+QkOyBJHO>dh%UijL$%dyRHceX_oJ-!Y>ULKw;+w`g%w<%%8yg?rC`}b@`_1ASIf1WBtHBo_C2t|aT2CE zky00IMMtzY7v=&Mpn}z*&8|YC2`A4nM1s@jGac*9)iXBd>U@Q9L1&dp>9o&AiAr>j zdfsI(GPs0<-zgVZRX)X;w^!bTrt&x>a!Aml#KIsoQb{po>JlzZ8a`2Fn_$hMMV_lY zNUD}>Y!8Ie&e^c_lAzEtphTFej+R4r)QuTrzJh2=Aqp~%+Yr zV>9{Vmm3Nx;$@0>vbh4|C-4P6{bzq{vj|(D|0IDUt~AcA_yrShor~t{(0ao6)tVC; zjGJQ1;mh*DG#mFGIXl9sDMY!1h%ZvZ0YcB5)rtht1I12y&xiP1(1!L4c$e7mLIB!_ zpS@AGSq`|%E}zv8lq=5A)`IX|HR}eLGc|@AoKgVZ?AODue zY@rX&IQ^c@0ROTh{QH&jKN27RI}qxhQ~8_t_}5gPUTNNt#+JbDCgk@cgKgUBCsYOv zFAYs>4iV|%!nWA2Z>rdW?FolQN5DZa@kY_Vur8w1B5}W&e0iP1o>}er`gnodVaa|k z-Ti|!?z*#^7ek5XX6oj8u?s0^C2E#*$(YEtF#?m+ohBBXlsA8;>&%%vCKhS%Qze89 z<+e^NaqC!@G^0F`5W|!v!JB$sY2MibBV7>>6L~0Jv}{iEREkV}Fw>B4d<;BJMl>w( zea~`VIi`>JPnwumtPPy(;Ez6oJmTLjrB#bmA{)S9WEkl~BaZCGu6DofY(0qvDqT?P z6e2_^=ZP}bB04`+=cyHMh2thkxQ2>c%36I4jm62I6SI`=C5|^(L+ZqnqMwXd zS&-pnfvI&Fdcm?8DpAM4O?SWqpE(VTO($I(^LST~1dqRh%*ZBsp3$kPpf6R=ka&7l^%gOa4t)&C&5D{Q!BFl|HxfyHBBs_;+9yGUa%Tw;& zYx26hR&^e@bw{b#Eo9N@M&&E)Dy*^Ga(T?2h2x>Fm$2NYfo1PQ-al+bvKAP5MlVrhMU zpUJ_P`{@~9ryJ1Kcf*N#R{~gBvualz_|UUkd%zf$4Ws;uzqrZ+<{7*$O(cvV3`ss7 z3s?NFB!RG3fuc#nDrSeS^VnZW0(Y30^`Nc+2EihmCy;@0e)C)I44FXO+b|+SCtdyO z;qUA&4>*-uIZR*la{tp`+1-f{;UtY1ihK7DXk9gnkO09y!GDYHGH{@LM|TC&Os^Ov z=Or&~nTaaD{t2P83ZE+=kIEafq|P4eERXtz(6OQC7r2!h7%7w`XzjvHYJbNdEb~xx zF&_b7H}y|mG8um;3n>q!HT~@9d}-tGq%i?bYuDl}U`6Mp2|gOj7r+0iFb;+K4i%W) zK839wr`dBYq#-F%25^}@*i@!m205$M)qH=@yE4MAfUf8y*r;d8&g9x~qwB=4i=l$C z)Ed%q*mz2iK4CM7mA9Pyf(ne7k<14G=xmoLUwb!YV zelxbLz=^6=T6-C`1<=|ua*!tCKHzp9Z)1Yi%E3c)P-qVvd7m#AGR!a-9(o7*ge)|? zW^QoM21CA}6mj3;XMRK&@tK2lhqW(0we++@_Wj*QcFQ);>4;Ab7yOPj40tV$2GDA_ z6>6za36dRYd*VqoKT9)}T@4bC#KYSxdM~5ptzHRy2gpOG0STFBt1LhEG+7b9EkQz@ zbB@u$JtJMM8F?HJf}~Lch2*fpt0kGSF^-uD3>B-PF_~fg_GpQy6~!|~8(GjSh1_4H z)?g+xrp;yh%l(3YtOA3qtw`j?LOZ!5(FS5H4@jT1^+F_UtrSrYwqeAIv=2?f?vn=MS!bVWm z6y&FTLU1jQstCb0P__*lID2s(M~$PR0YEnRQflC6y^=zvJr#nJTBDpeCZuAA&~sgq zHqOf_BNth~^X4>(27H^J zrm(I3?Y&b*3Vs!fo4f)@n&ZU4TVt==;4tea$B7R;U2V`I%N(!zTsUpF4)I`)$*e=R zO(=nub@gdfFhc6HEWga+4+1!Y%SR{+1SXX#=0mlBAj<2@g>@ui^i%o;3%dX(lPU%e zJ-QrBiy|`~r>;f9dw_6ShZ_JbtsD~}P&I@CQ6&qFx+p@mGEQm6yP4RCJf#a?-ML&$ zVX`JfBsEgU(%gM*=@IkUOuRxa@|SBPl4mGlAzsMO4$yp8q%ZWRr$5x9Et@En z{t-(xcr0fb!Ch8ohfHRIAn)Z@Qyuw(rYcjQ18A;js^Q-XV4G#sH)u7^T%MH0tE{#??-sexL4V3~aO%Y^ zgTF)3?rU006IOF;u_N3FZ#g3nSw!MSo6*2LETRr=nS}>7X zf*+WR9~9$_M0IK+FWM8@^iW}1ye$N+g>dOOR5bc|vyb7vq}XT35zU@oX~rG|_h?EF zc$A6|wBS}Ox%*(2DHQST8X5uWQnPUCArdb4ehF{QL$WRn#HQR0mS{a|^lk1AMAL7? zpflge5WVJh)(k?pcQtjmp(8Wiwb&Z9NB)Ykqloz7dQ>V`q0WwSh~^z?WTU(Gf(f!6 zolW^@Sac`SBUr+RP{?sLkz>tc4!t~-=~&6HK~XZ)@L1Fk1o4TWD!;$aCmK|`&QasP z?B}uR#`Lb_9d>=8yTR%0<$PnJ`b_uN^lSm`bz1tKoMHchAFe1M-Y*@I31Q4lD~->PZR-a@hufC5t2JwVSM26zA_OX}=!PSr($Go}7kR`?x2JEgPNAbZ}X695#Ey7Rr zAs*BSnBDp5XxkrI@mVG(QdVcOkA+1WL4{PjErGkF4LX;h^8LT2(UmnN<-bGZG|(-k zoKN$UG+?X~v&@I9!m=%0Mq89&U`pN?{%lek57vyDVLYYB`B#0ti~)>?2cQFMiDnV6 z^H1%z4R7tjTycp8u?SfQv+ea|uGq$LO4X?%k&|Q{hX^yD87VTS*|kNW~O;BG^)F1hc7C)CAS(}P*&w8j6o6|EsiX~)^K=> zrc)Nq%xfd)EK+aIUhTDbA>me*#6*#mXctasR*PEde;t_Mrvf*Txc@e*q88tIQ65l+V=kQSxT_NV49ckp#u`bXOBcG#;^8!bhj z(nP?QK)Q3#?yB2DK?y0<@G8Raa-ase74$~IUi4N9b$Soawd*};1ou;p^K;3T%XUWK z;HL&cir+tbcX#iVrt&~#ChwKHf8&jNDFPS>1vJ79LKIX?iR`uws_VMDit z)hN8o5@T1tBOG+ZLN}DI2d@nGLJ2fLPKse2KUh5SnLNr4IiUBQFyOXHd zXGFDIa8g8!Xl73zkg)fSK6JVLxL|aM%RPns$Y*fU=>z$EAgE6D zYoDBpIYKZsL;L{G0aa?M${^I!%f2xD>zdu-obbVjWJ0`8nY_EH_#>F9jsL?WkkP~Y ztx!FU|0bl{q$;w))z}d=1m_e=6#Z9)Pkxg`k+C!42jp&-`HNA16IOpUR9<40e~Yw4 zQ{?)jZID+|qb(K}vQO?UCeiKSwdrQaQ>8sNpVC-{k+mxx7Wzr9Jf|7mhRXd?7^0`x zHp$1sCGw1F`C{K^*7(GjPmP}FXHyvD@~l5ur=rh)yO^91j(X;Q zk3y>dI12xVRAWgaH$@}Ie~R5%N?LyvKYjz|x0-3y5UITV)pGGyi&Q)*%%tRi@_Gd` zx@9-Uob}R3J2VWwlE_NK;Qx9*y`8p>kf*mZtIu+{-t@T2N@Kd;KW@1O(AK94KuKmC z)#ZT3ovAmhHoUoamKlN?>8xJdAmw8YpXtl zSh&+NTW3VbK8RkcKP{_!G#FkKp}}1 zG2z*Da-AWZ@tnAHD4F2VSVJx;R_U)r?NQ4Uxjgzhz{%0MleqPJGR#vm&$@ZubZl*$ z-2WYRQlV_T-c0I7e@AX#tagSId#$Zxs$iH?e&73u5pB@eEG?B~UGS?bS#pZCOriNQ zy0d$=Nx9&aibQ#d-fD+y%|&xNc{vXv_^Ov5(g@fL1B`N&o<0H{tbBD*o=J?wiSrKI zGCx`s#zvpyk3oicodr|j<;26sL#~M5Wvb%(u+muu3^v2G-`ynp8*EU9N<0VDP=&By zR(g0Hby6Yah?JgKc6ePo{t9{1$Csg6vCGdA^uR-V39>+>hbeuyLb9K@->P@D<|~X! zwI;zW>e#pHO(gn75iV$t?gDXKAU3&Ai$F9ah8|vv{>!A(koH?yQTo5HdNcj=y8I7pd&A$b?ZYE47?o1uaS+%Z=Zh&o^M!DDXsEX* z#oEbn>&kn+FRC}bWPXG2JD|5A2yEJM>otQl&(G~lqC z&2}RKU5zeQl^d-iv$LY&zr81GSwa^{LX*N35eM7i1&EqPNn``M*(C*|VM(_BAxpD8PDUO+DDJ6|;p9@ib@g2I+ znzkz{w{<-RO4x|>AO@q(;N#r4aErA?2BZ`iAfCqC#}3Ay6A@p6&cF-yzxg1~tniUe zieD>4;~Ck(Da8rzxhYs%Ty7ijQU;Z>%@`(@bI09w&x)E`?srGPHfmAlpZ50w2TwUr zIsIs$PN+GDRi-=73&if9&yUw!7o$U zTv0RHrN-MLbmxR$;aFFTuhiIzIC4A9d|QjPrGa5s)hd)(Gw>9nyur~SoW{=QHVLvq za(X#`lzK5|MldJG^}2jVc%0B$0w;WL$1NkfjWqLER=5|^!_+?f-_DWM@TU#m^cdSp zGJh4L4zxj8998Y5*R9wER0d_J2~M2MeX@+dfc~}UrLnXv{e{f@&x|XP|JgP9UqV~| zR@J7g;jkfu@<)YAqq$&F=3Emy9U$*Es?^NPy$*XmJ+Zy@RBC?=W>-88U89QH$gmCB z8=xoP$4(Gw; zh2#|!822fsARVVkYgI=~mGk5k2KDR}`@p#~_iDWkzZju9TDb~_Rk3suvAGVN`sc)jZln-msnTIE?P}>N`I&QEr^;6xfWcG@ zLcy&%Ivk_&n1N99MyQ%Cz|UbstP^`b;F8L*k}0KI^ea}@ki7>s9Y*czu6`9a7`!!m zMn{&TjMA@-gxi`6>4Ig?+~_7kWkm>#NoA4!G5(@?T|k`OPWC?3OqM7>bQZSA*rLJ0 z0?VGbW{k7}FMjmLAs!CQEd4{hHG23p@&w-LlWj&rd7OMiME2 z+d&E3=F!CJ#e%qygLNU%qPfq8`&uKBzk3@{3&QYeF}vi6o1^n?wuwu-tn^qxZ%H^u zzTH)y$t(Hv!5IaeOP_NaQ`e;t54V{BhoDKkt&2RIGqd7-)Kv~9*=1S$WW?{^HD3CF z5eK)tSwMI4rZcGFqPjtC-dQJ#)>U{eQ1(z3=m^o4ax1oC3S3w+GLDi{$54G!4x)RU#pwIvF-LRYR<$$tKM^t?_`B_F| zXt=za2B|;-$+ZlL7niaPC<=)fZtq@oP%(9fqx#mq@|tPUjYNrKfiyC4&<>MQ-!@rx z;~@IV{joIwD=0J`xBE&T`=7-} zx;-3-AFh%Ix56_ZBLfYMlgz;9*$)y1_}$;W8C!J_X9XR^ccKUOx?o>~hPlnAJUrm^ zbm-0nly1;8fQUaMwtAy=)LmVJW@&J^>I8?-z zD*OkzetLTW!zB9Hs2gc{q8t8dxixvPhb|i$JjWJ+)CyaS`WC6aDq<<4d|Yywr7~skfx#n1r3#_p)e9w(fn$X9!}$xDph(TPOicU3^g%r` zxwl--J-s4t+4@fF3ML7;R;_NiC>~sWuqGe4+OlT(+oa+vFgIWAtvhyIWjaoEzTKbs zen9H+cFE)-7X#|Sh3FJRP{v0G<0sfx#z#nKGhTjV1bk^#e)*Lkpao>{fls1P?RIFj z#P$eltp=3bppxH4u0JmQ_yBZUK8pfPJeMSXrT|Qtxk`52zk9dH(Oo!gk|mAD_bi4; z3<$KT*^Ng`UcU+Guu#`EPHttQ7cXR}!QYZJEadRY{!6)IT( z))a?UjJ(3W`?p0hE7D~OuOTff^1tb^aMBHP|7lRSDdKiolS_ko`pVoj6ZVv73^Bs%@oh$1QKX+XR$rZRB{mx<>a7{tU{}yo=}G*jfyAy#raE80ymol1#9l{ zZ30_k5<0Bs-5(l7^*kU>R$4vyL}`&_f&saGsXw6&z*55Rzlo}($Z35zf!?Ptw#Gt0 z4~Np;_=jA{ahgJHGPh@mU$$DjbhHc(XOUu8O`45^G?n*cmnqE_yB6YbdG{^R+%&3; z%M?R2oSl$Y9}Z$@(`Y=c(XB-xJG+{Z<@)3OymS0mT_BJ->^?px&6K@Koz z{Syc)tV)u3+HQUpV~j>!I!hIK%4?wvw(AbDzhL)SSCWLz;!L17hfN<1wyR!qbZ3jA zC@6=sQ=aA!uD!w#K9q%<@KA-D^w61`U|$BuX73ut=FANZ%^4d&ANuOV4Hwt!S#C)5 zSxv@=0QuhHS!u}hIg>VM;yVo2712JelUG>H4-b|i$?ZdEuBlgB$6*h_`=lg&{|kzH z6?$f`UfM1yee61_g8);$NqaD@| zo~<{=JzVIw*$*FO+?Wk$3Xv5p<&>R>(IM;&b*pp!XQB*JvEja~Ho8d<)o>m!-S(89 zpi9$al8mV;Si;kj^c4-fn}ODKdi7gmmQB1l`>6+-6ZKuMcjO;xvF|rN8gP|m-sf8r zX*uws(B0f4SNQR~1P&DBU7}9twMI(}yCavz^F*G>dBmCqW2Mqf{VDRz{m}C3eNEVn zR#zgC01URdNT2mcU?MwO{Y_Cm;LPtowF$LaQRTQ(3eUwo;1?45^UF!g29#vu!xa(5 z=nqfOuALiRIUe0OI1)1@AMsS(6XZ(@|4<(lJ$CK8mbU)PibT{_tv!%ff}Iv^ugU)) z>>ogGU7aX?r1Oc!=tI7YARnv>AHmyj;}x_=Dn?~nBs<|gn2h&7KrV7 z+G3i0@HNxkGBAxNjOyk{S&g0>Gi$?1GdTpO`|T1xjw8%1BmY#`d$nu+bUidv{2azm zd4Z$iYR$VeOCbY~oVW<(i!*C)F@<-Ak4aNZbH9tzUJwiW?Pg_V%r0)#R=<-J~ln{k&1=y<7pK}oAMP?bw z9gWV`O|WhrqTA-q%(+=9Yc#pj&5mNnRl$Zb2*vD$$sH1;L+M{u8uTKv; z&xL7b4bJ={a~C;Mk7{gkYpY3pqyg3t*QV+AsNwr}A5D$GipjUHY(Nt$N_^ z7380Wt0U2W2g@n{LxxkFSWOiXO123iRLiQaq=X~{hfhwvDZ#`AvnH;-KFP9H)(d<6 zc`PZ;B%kxnOmZ5?gEnQQYK2E`osq@p`Ivd-b;QuF`|0)kWBX^a^!-k{KR6-_iFCi+ zMh`SFt27Z^(UomlOB5#wmDJfmwZG?26EuYSYNrJ(K~;u_0Q|0FR1iVm&7@uXst)`B z=B}N%L$^)WmZt#F>aLhX`^FO$^vcXc0=0X$Mv7w>f~#la_Wf64+2x{gT5DAbc(KbB z%xbDtOHSlkxqakWWm;Lp5H~!xxE=FbKSSl2g1SzD*CQjsfHWnai0(rd>DKIv=h-q4 z|Jc#PQQG$CPPlNaV9GdZV8ZwUXxNys_zf}h71&WNYU{65MJ?TU?eQciL+c>ZV@k5i z4vmAP(w~PoG7CLaf=?Me%7wm*3>hh_WT_{3=o?uy-Ni6F=-&}zmewoaW28hu+UNLT zd&7BFmJy-}rcNI3tX<~KACT7L_sdb>Hj;qJr1u!s_-P#bN2s0~d*H{aZRYvz)}_tx z((i~*+JtmnJm(NBSpT?%*CJXJ8x z^MZxPG2M*xp(IfNcB8#?B4~A4SabkKYKc)gGz1rK&ZojKqqcWAr+%0$^q+#^*22V) zTTJX6_am8??t>CIi4^&W4mpN&QAGMq!(P)CGx!jnQ!GUVz_3m;BDk$QwQaK3M%Xm7=hSggqNMktdjjswpzHs^-- zm^LLJhw}aY*ShbqQnw5BweA!DucVfL<$M0En)mOs_K))Ve>!phSZA}o@(GX*uzgHi z6GkRs@Ck(Y<02q~q;4; z6soG!_bIX;?-i#`bZ2QlzCX%6yi8DK1ZnpgPo=qTKDl;x*MGj;&iQ^@4v@r%r(TgGVLxV8b2mm8Hzo zFc7ffj2ZkpQ{y)psHTXoxpZPTo|v!(=vO*RcV977z~^AE;{ho*;Q-Z}^zbE+TL2X= zl4DH8b6hws<+LZ((0!X#ff`c;1P1Pjub1gnbxFIc6ZGR1Z!X?ShV#&z`_jak35Tvz z*8R5A@AWuFh^`Z7O!cjYVnzsq0EKBS;m!=m8|kh!R(c)-58y7-XA>Wd->ViQsxz@Q zU2_gpzU={0lGZazsCD+Cht93+iobVt#^?p_bLkBb>k5}9Fdk!f)E#Jm4L?c;)SQ$? zny@%7&|EzF9UM+R;`&S$a}Z%xE7%W>*eM`zs+w9?+b|<)l3Cxw*M)(*;;`3Dk*b%^ z(nMr_2QHk3^T_X=aIasHLQAhNdoHsX6k7L@zk#D`m>vAR^l`Ot%-;nfz|nDkL8@5cb+lar&vTUAes1DPyaL}(!qBLm45Qve)^df?PiA&J zzIs3pZw+5|^PTbv?IUPN%mVl_D22J5x-9>Y;(G=W3&Lrffno2F>pT~!X5-6am6Kv$%79?cR?uBp`I^u-&^+HF@iRZ-He zyuGLAz-Zpx*no$nhG0VL@o{L6;t`Lf8`8LI(~goo@66d7RCXV~Cnc}ay_#pSp{?uc z@b0r`G)4{y0?!32^84kcG*+ug)U8UUfbcmEp5N%mrp}pt?#QT4TKy#uv#jX&^jDE+ z14Xd!5SJRcLbmwdD{~i?Fc~`b4rLK0=iC|hBy6ZyH|BJ~0|J~`879=dp0pxP>6e=7 zu(lp{uq8EX$mxiT&1+gOgAVR#QV-EcjU*qm6%2cuNI*%bc{wFy6NqF1`{MXtKO^k5 z;an>Pe9oPj1HV8o2I^T14yj4WGzmnb406~HK!*dhqOh*=jj5SU@N1!q6DS!;fXMNq z!9p1RMB1~Q%qB-I7-j@OGKe%Uz-b98SKXkOTvDDg$)ko>%QLDTCEYs=0D3fA%mhW! zw)0&|&Q^*)B?{ zIDx$s!{P+IPpcA#hodC&UH~p1I?Ox~uc|+p$}rb~O<3gd z_5}Gvb9iW^?e*jK$v0W~F4U4q!gver2($7FDg)vGf@O?f11&a(tNv^&7$Yvc_!WXI zTz;VbdZZSKdZgh5(?-cD3+tj5zXxdZF25-+OerDFgNA|v<#~uHH8DkpPy#*1YuHJb z&B<+rcwj&?mO+M2{pbqM9PG)>y03u*KTjvDDWjrAoTdN=a@Ur_1EI8?splY&s5;8M z;EWPlV?WN|f%5VQi9tvOv)Z1UIwH$SdCuIvdnBgfu)~S+-q#*4IU!1JC~0ZEAV?OX z^E$1Va&PJMK14{4w^`|<$H@@v86|2X>Z`itoF8hN{89?4uNRSw z?u>R%yhG;Lo-2x#zE^T_-^xVW6S0FM+^uQk5sR?ryo;7Wy|fmkn=-KjWC`5 zmJr(V`CN2GF!f7>)M|QTluidd+VIl;J(2nZ7S{Am%HiBgc zB}kG8*KQPbO9tsa zZkYh*lnUw~DEN4A3buUhL5U;-N*-qaEG8|QvyVDii}J<`uEdol$aaH1EYA6GYzjxT zru0!MiHea+JVp#*U`LNqDlcn3U^6T9rcqKeG7u1M%I+krEpR88TA2NJZHt8BnCqD4 z1{uL=tT%7yzeekVfGgYcmjJdO_TR0slK&AWl5n)O`7ij4-^Rw)N#E(K{`{XNt%$y% zldXgMKL&26s=DKx3ety-fo60iRS;DRQa&GE1~#e$Igz=Tn3x|xE>seHL(Xa-bU2=J zO#ehcRpSq2`ePMs$Ls`ou=$B3=UkUF0Cx`k_+D$ozJr+_YVPLw^myH|`)b`W`ZDh0 z`KTo4`}x2T0SCPzC_1TdN01_rOaxn~dK}ZtH!iodim6aH5L4jPKKAc55N_1pz-u7g z1bZ@oCL_!O2=sLLr20+9pW%;GlMHF?T7qy<*`g1Jgn0_W%mxWUmV|%ABI$wAp|2Vg z$2C<+jk8qUFvsanIgLw+AQY!fFNhIXW57@RQXZ4LA9foc7KG~3pO`*}2Y7SGbhJG| z1)8o$RzO9~rRPl=#ov`9D<3+~GRbDp3-cDJVRTQCb3cMmPtZMVDcz;R$6%2zsA-AE z*^JXkY1T5&h3bfxu^$xnzSP(~877x62YEHEzmsDCMjMjbax3ULdw22=^DUqrl2?%A z;6p;r$j+0%(S$T>5y~e}>CY;xYpG1ttBg`NZ!lN%GIi&+ryht-9dN#9U z&_vF|xz>^$W==*cSEFG(?5}r+tV`Lo*a^o*R+U0pz8-DP!g#PmA{hu}mvvSLP}ZUzmd2QtzSWYs zOgFAJ<|$ThTvC(|(ZAh%RshR98%g|j*3M*RRFTS>f4;*kzZJsTUWuCyZYDbKk@=n+ z#AF4yq~K17{T&~q-+{M-U7*4;UQxmd__B23$d9wvHeMMMz7B zj+)^uNOtIFcu6Qx!RUz8YVwh?hU%}lSS_4|SS>8xv{*@CmL5Vo?cS2-uP9=)Xs$S& z@o6OwhG|WHaZ3Htiyd0%EKDA0rBA2g9G#2!3{MZr`wI~D3>{!jPgt}W$cmt1|4em)Pu6E^abZ9$;h+<((c>JhjVKT zah36&$nCy{Ww0xsjw}yxl0p<4_RruEN8HtX1Tv3kb5|b;_i6?Z1k=Pq4U8}>h?Ui1 zO2!72e0<(F5s@X+B#;I2dq*g2?lV^J^E}#689CQtUq;@#6VHO=eVX%pd)5u84(kHM*S6(}Y)7&y&6P_fj`eIyh$mOb*<$XS^9Z~e?v8X7`g#Gx{b`kD0@@T#b z$@=+7uP1CazwGoKXUDmU?$%T32`q^CRlBeYPDs^+I$}P(R1O2MnedN`2H?rIse!S2 zm$-><_73C$rlCs_Fm)S}IIMXfQ?n4T*<-I`RgXti*`yV~h0uQ7LhXKmKYdHCF zZ3*ph>2yC<+3dP`k6WSrF?in8=QjYubjQ#hzQSLUWd2nxy=hlOUAw4;NR;Y0ft|I| z8jzfpK~jt^AwONr6#DaZ+PhazvlehzOe6=z=<*r)m#bWJ8nPend32h;Nnjd2VbYH- z_3qXoG}1Ne%^*GM&87?>!zTDKdeylG4yu@pZtfwP-VyHKMR4Li*U*#ke|oDsLH!8; ziOb2W;g?g=AO~9``>}x(00yFuM0nrn24*VG5kuc^hZccWlFaK*jP={OJR38$SCuWi zoz0&Y{Dd~sfx!j2192Xq7ylW|QaN0{F%lt}4@^3U6^3`Ga#Er6j=d|Vf{NzGYKd+3 zhjII2+&WhCE{;2ccv@$K9!&7jI~je;o1*(i8Q13w*Jmx)r&6>p(9p@I@Ln6R%zG&8 zGzrgl;h@a37^2F4DIKaJYe_V&Yk}k z{KWrB+c*-@|F=3oNmFr-59uR0836DD`9YiOpy>I%SDHV9Okvu<9GlNgX07o@Z*f9# zsPC}p{da6!tekiGXw$jS1&$&k5>J`Q?7OTnpA1~yZXaN}2xw%CU5c;G#}GkLVeWvT z%HU@`#UWiyJ|Zjg5mpGMFd*1W&}g8jiqJdop89r2h9Q}qE*g4+K}fttS?rs2u^yTw zetMqe8Z$&lhRDhlzcwfNb)DsEe*QMS;49~l(@58HLn`;lJ`tv6oFLS)@>77ro128I zm720`*Qfc}EbzzytlMvCoh}bvIThTmjknjX0!Ep6&QHPBO)`p@R?)(Ft(pU*%VzZC zaUS=xPvh9>_inIP1PUL6)<#9~-VpY)+iPJj(aytC$2PK?9&Lu2TT^ zL}TK6XNyFTt_yW;2P-<9y}?H*sR&r50E^?RY*6tPA~;C6pFFDj82gt zYxiZ~MmJndi=|`ZmL1_bFq?P;qCUJ%A!T0sbR)mQ_*AyQx~CEw<=tBCpylikUM)w? zAyqp}*A5?~RZRS=yB~4Lr&jZ_bQi8D;fj>lCDMlox3y>X0J_7yaYY|>e+*eEFVSZx zi+c2mPlFkb%a^SIZrM%I`L&3d)gxg#4z;V7zeN%j^-a*d1`5w0syb(&wM0<1?n>=W@p^b)0H z2L>YdY%(_QbySeJ#^Q`rO~nCSjJb47rd*8ejd~~n8jX8w!qG=iQSH?SaXt+evzDX7 z(#nUG>DpI`9T+D+K7KTRbtimszm>c>vf7w?D;jmqztqEGDL?K+CZ!asI*?Zw2`AR2 zaOtGo{@}MnFWEO4h%9PbQ;`YBHwqR5qAwy0V2MWMD5meDuh=6# zbQ&*^9-{mz7I&gD2+OPZiCTJ5!l(jb`dDOoiLFWAvV>*pv_4oD4dcWhBqRDoN9h25 zJH}=2cqB#BH@uQ18;(;a9boZt{+tk$mYCi~4q7%iju|YU(HmGnu<&TJT#@S;Xg*;Q zyi)IF7Oy>`PPXENn~NTQ-O6(ZBqz||mbFfO7Cp_UAA7MUEii+eg(jb9E)z6z$&m*y;#X6M*&=>p6E`l8ns(6OJd@|2xN#3R z75L-GOcngrzaPJ$G2ssrazYeA|6uftQiCrVmiO_~*&&mn>+l~mJO**)b?9#ZbJdmL zU8URLK#F%rklL}mdpwN(R$M2A2v{N?eqJHRM%TQoXkp9o<`~1CEkv~YD}Y(&?pJ2l z`{?!GdIf0Q@LYi}05j#^0+|2SKK6ePVE&I(LCN|{qxltbgOkY(Lx%kO&uGOX2s^(a zfwdSEJQI2PGUQCm+L;FymgmI7w;RwGbi^bfL|;GNQ8v%2(oFJTl8ubCEbqg#w5<1= zyEE5sl4=ryLk-`d8GoA4m+VDHxBvVTwRe97!hU^3|jG`uwkG$QeMhFBjReW)h~ij-Q*R4jWINTBm1 zWj-7#I?9xHy_sAaQO95w8JKxG6giAcqG>0TWHX0LY-xO1M|yb~V}>A<7JTO}sEZdc z?GdDJBTqIJP(tS;&+kVdMD^u`F%Pk}I<$dV(_g#v ze|QFY`Wj$jVXwgm8`5j7@B*p4$$2B=QUu&&Wwpb*T3=|zj2mIm-T@!1MZ2VQE|LC( zEP+Eiyd&Dtdm+=|iFT_$6W<7KPq;gzq?D{v3p zIoT>5ak}};I?}{hej5A)^-M=z9(Fy(*Lr8b-$$@HpQ|t4C#i-q5x(3`Eu^YkS>_gj z-)6boiHXw82z9eUZMY=10!WZ{_v$+|iA_h*G=#zyuo-jB3X6<`ZD~HX&*A(IEXmjq zI-uRsa9W>hV1mRgN?&0RMmPwbt5fnZ592ks2GQx5pF z4}V8?&48p1Gkl52PO1m8Ot~lcU3~DQV}e%0pa&m#gyv$g;lNNlB$sSJk*JaiQJ zRTaZ_P?EiqGJHRs6eVYzU5!GFX6$_DpqIMlD53+971kJ7BVuVZz4g*-jjq|nM!r(`b}XJ&@TNwP z_Q%J~;R%daf+Uo`ja$3pp|%Ty@wOkK9TKDcAv~m{B#Lbx`Wi(!^!rc+piN7 zrN01euiOI2sbjV%Mh=7H>6x$+r+#~tA$G1RLj1)x{Zh3h{eaj(n5j~TfFf+tUk>r+ z6O6a+mu1@=EV0NWtjee>Y+wk@18#Yk>d4zqvu(djZd z_cs|czrwvU($&dKkp!Ro)6mxr)%&vPq%U8Q2?{=ObLgZ;i~UsNKee)&1~8Ora+SC4 z_7W?cKE+SRzta`@n(9^oQC-TLVw(WulIP;IN}W|LkOQ*`xeeq-jBXlTnrga58v1e4 z(rZw+1a%>u$H7TOqZLgpJdkgN;b<1Y1gh$UP=7`7h<$5pR9^S-^Pu=vSAv1(Nda6} zB7yfRWS%4O)12b|N(2@yyJ&_xoi!{bZ=dfw6*?^HbZn$B1Q_w2#disYO*K9?^cWO4 z@xcb5`F%ZtVlJUQ+QJ3DyLpjelwTn_7P@?DJ{sI15q?-17AZY=kV^inNAaN;Xkogu`vJxi63#^Ox!^{<^4(idEQKr zm9hP&u_insNI7z5Ji-t;Gf%5W$f<0L1 ztC>6>VbRCNeJPZxMD}*NXjLLp)ig$xT_h)YepkRT2CiTMEkuWg0Awj@tt_0#nRk#N z>&dTeen{%nDCLjgIEGu zCvghEciP6Yz}I76*us|V_?f=6{w7 z{DraprN>h+*8d-mZmHdp8yzI>^E}}F}xl<;3i_UO#hi-ek$bt z!NB|2ueym+MnX))*+Cbr_Og05aYRGT49lLVEFqm(Jffeb6^5)42_!~kYbNH&#n#Ct zW!UHUyjPv>jD*k*v$V4%ub1s#Mvh6}$BHAJ_x`zWZSx1l%TXm<%Ylj*BnQS4fH{oG z;X6cz%m=$?BlOK6*^zT|^htBJD}IpdxC7a-QB$Kwb?BVCT$CKzAzrnu>?B*=*cC^N zY+6y^4|$y~vU*zc=Q>d7=kUr7y!x8ab4!L-(ll@FK2|QXK@jMz#3$P@wyni$Y}SvJ zK$VWt9bK;Ly#UNliYecOopu{9*;F509iQBtkOQ?q+>vWT_GeOtw-6{^ zhBv9csg390^FKwP8Q1yKZ!u8a4tq;BZz6rR#zUW@*YTLNo^p{}?so7sVDKZl16S;1 zW}(+eMMh)}6$UnPG}1|KzYtN4mO!MFEF=`C18-X5))68K6+|TROkCYeb3H9xWy$RL zNc^)P7V}!O!qP`sHQLv7pt#Uv11ZVj3$XBPN zi!$GD#*jk#=(XX;+ia`xUPVOiXY_?_z;+aHZ+}sZ~aYIy7hvDCrLkRDQxOv^@ z#;qf>qcI9S_v;Zgk;X@kws#&�ZTP84lLYxUGr}64p)z&OFDC7^%+EVQrWJ1{Cyn zDiO%tU@-b-+*}#ZTb1)#l!d6m)NCOkY?#H34Y#jz>64>LcFCn%k=I4MhIxwfbrV-H z(d*?CGq|X-6g9y;XO*tsD$U*WOk!9N#Dk;y%oY2+k?6BpMnWm6vh0N@vX}@8l0||{ z3W`Tdhv@rqVH+BnIv;-ugcUPHF7UV)RTG{FDdE{E@Crp41G+i^-%h%5G9i*VwpB z1k_pP*iH3N@7k-N2jNQWPqp&(IYdb@loAJXNr}xLki1&|vjnR1T1sBG2|7Q|DCtG=LZa>);nmZ51p~SFW>S%FljuDXgA4)P zo^yKRlakpV{caA^%P7dm3kRkVFzgCLIr@A>dsRp;@$5oF7ZOX#!PB46o8=V}SjuJ{PBPQVN#%=5QPmd4 zQ@Y!<&dQ{o7N&s6Xia+|RY$SX!*DD}(>bfAseyy=tO>^2Jo^{-$#ZAwouIuu3#A^o7ms@*MFpj4B($&HAj&RjR3p9(S! zjYi*tL%Ih>t)?2KY=FV7^^}_}`*9ON+}&AfTL(L4TSuBY?;Y|-W_5BnNV>5Zeys+{ zbjlvTc2y6&bmM8c^o!GhY8Mxl3dmZrShUD2J6p2V)fJ>CEO9(&fgrgp8^WW=uHdHG zTkNypyqLDW#A>y9 znH%Y)al?Q#qEoqYx<*w6%}fu5c9rs(QHxZN8g1WCc~zo|q*)nfZ>dL@H+te`<3@p3 z+&Q@@i>&B>V?~{E3!|F(EIa6Bd32ga2eq<-ERE&)+c2T!`t~AfE>|Ad8r4Cb_ zQMIOZD6e)~n$meeAA}ZXdeuDKKLi79=qJqHawq^?pB}|rk0GvLf{Trn+k3$-XNQ;3 zGQfkAg27Bi$|xa=PSlSnZv`SE(^_58sKRVy&LQL`&6|N$ShuiJ7l)+i3xL$1@=dfy z#XHibi06fZFzg#`lCp;pr9rB|_L@nmCkV{aMuy6xZ5(Y-gwxWkTioO>eH`UIF3lx_ z#;M ztiF;XUL$NcW-f@rW`O&8AA9fm~M!3?epS#&Lxx6TUk&P5@>^GQT8GV}E1-U>!o zlC#%Me&tHE^TKWgl@Gos9fH{;5oFiWw5nE4r1Gb~eRJ$;eU`eVJ-C)lwdA9GQB{Ad zc+voF_4(GQ{(W~Ck;-jV;t>VOw@cU#Tx4E2DvZR5b@&7g{?^~f#>$H_sCd=m5odO*y7j{QVu z!V>fwqj^zcn16t>V11n`!0)()I;6xBG@PSfnP%JB8wHx&qf;mJeU$V)z?x|Q#=_b! zz16=By*C5BqN%Y3W3+t+f2s81{ zEyX;gv0=cWa!0&be#yt_I*7CMf;B&?WW!wJLVe9*fo&q;X^z2lCEIXjOWLD$k3lK9 zJ6stfJP808t;0#bCr)2-gui2Aa_6EIyi+<#Z4p~_;vCT?i6$K~u{yyxO}FbUoqv1m z!b(}N&I>Q}ZJt8gZdI`wMF&2_e@!K57f2IS?&kG0BUT53r6a-V2kPkN(^`z3;3nB6 zXjNI(zc$I3@VFuTw68>Y%?{2%f34qg z1DCveRT{X*E>P(@d*gvYjYu^=YwRxHYciK?IT{@wHI5J^V_^aIBHhg!!`yjc`b)*6 zNS{KmzM-RTYv{n)M6WW_&ZvHLkPz!L`{aZWYC|-w1y_)iCwy$BC_IxVA}gVP^+d^4 zuJA?Z&gPAJz`~@i74+&L&2^=Ps=OV${K=jgb8^n z^J!?#9z=fX>l+$0RC7Yf$n0Zr)I1}qW@^@Pwn6Ey!ga-Sr85QA&Fi3ryHq`))-7oE z%F?-pHaUj853{A$^van0#QxJC!Jh4yv&Vkr!c~&;=-$1@ULAQvMPp+%J~t0~&A%qI zwV!Tgu`gR(R?%+CfB9Y-=*2$D=96;k8#CN3yU&Yvi;FN2vuynS6?B6r@jQpsPYAQi z4|*d&@DUrOHZ?fPpo+5z%gAkVgYqeU<6@f)VmoIzcl{AL`4y5n4kgm9GL^jWu?o!(WpEE=bC^UP&`}o^GJk}9o>l@ixzQh^2m4#A8p6(MSPhi(pM&}Wn zoq(V-t=BTj?1=>%m+o_f6r+nO>K|gZbSUf7233`>OD&YjFsM|V&Ey@oRUK>jnYSsJ zarGPVM7O&#rBnlU2cRAAjC_NHWmeM38CaQC1`D=pGKAzjK!@~88u=~cVG2kx;tbLz zkN1Kum@#QqK#eH7uBg&~W;W?TtX-^0Mtjo9T9&8vXfDgq3f5~xKXecnZ5Rg5eQRKc z?lEZpKAil{QgTk4E&mv?zt!9%f`c4`ysQaV}SZt>oN+&uubD zL2Qpv3zw1ez9MoCA;X(YPkf4})xLgRq0!dr^F@eRz<6MlC$5!@@Au zF$DEo@#Gi-9tpUr=!&-Wl+EW9?4ulQ`y$XM7ll7gVenayXCB;j#wyXd&1hXS-`?fB zFT|5^+KzeUEameS^Rlsx>osJtt$Jhv+H|b@TULV#EEq49RCluq-D_5NkBh@qyi=+& zK&>{EesOk%G($?VF4%PjP(-D_V|%c4OGTiJeEM5Mx~}o)=7_5;X;j^wLAJj#vtR1i z);sRmx3xR3TPVWmU8|=;*x|>-I%GF%ENxTQVNMKd7dKlgE*;+l=gAoU&c*O!yx{PnW3! zPSbAg_Lo2JbEw~ru75#mw~&QMI<#Pku$uIW;gt~}i@-7BYsE2O5B{hI*ii?aIdEh4 zW+t*Ktt-*j?4k(1Vz3UDn{0XzheK_oA$JY#{bZHAVGmTL5C^cF?5uDetgnzooC)i= z1WnjcDgIK~BOaqbC(tQZGvR8sEI!tgV09*)pu@b;e?^}7HHb_9czS+6;G4cCRu#WyTij?BA%}@TZ7TO3Kgszb&HyzdpPe+x(;*xzF zh^B3r?>y1Z2DUTXHESqS(DHU#m6;C;sUtVM*GRD(fDc{LTt&O4JT#G~xETJ%mo_cQ z>73y>9_N6!wVUrn4y{J%(=g09Dinw69Ca|X9a{Uws7eFY$dF35Dzs_qWvhvY$`mQf zh+Ubwc}OGoyb0Yx^s)9LKf9%EjzM&{nnX3(Q?^e(+SShoH-wryB-vt=1 zV9FkR6;|2d!ZI1zC;Lmo*(P*kZ#M3Wy+{daX&Cx0G(l-cMKVS*>5o^0Kd#wLe{t8n zX!)--5z8zCe`H8*wqykuD^dW_YGK38hR-qH!Ust~cY3d@SZ`(eO*wsdv?;;jw)#}N zwxqrx?n5+Jg!;5M$}?~$8-Oj*-5=by8@a475~?Xy9DYVQq-77sHM!HQetrASK_=!6 zczkuN9JL@QS?dClOiduff8hO}3T=PmMe*MP7@`W8mi9B@V)Va+M45g2#I-t}YD1)! zX4>Jxyndbh?TOU3@FaO1f_~QVqCLXU3A~9BX38|J!!sTYAylS>XwLQ6LufMbu9OSj z;SQ8+(yKn`SE~e^P4;!d;%nazF#jk?{irER6&l|dMF%Q?+_m1oi~1q!C2~C?TFRY= zv1(My9Av&Yr=!bYTegrQ0gcTKYNjK2bO^%_N>n8QBq)KvNaPwO;kG7Eci@rJ*C5&B zu9Z7W4DR%~ezE&^5~tUQhlTXFRkhc;qJ|F4#vB88_c&gLoZr^l1Dm!pXs-|x>@yumZkCvY@7 z)mSF`cCvrEpytw{4Pox|*M|ZC#3wp4*NuQezyjTB?Ch^%eQmxmjwVx{#~&W?>ZZJL zj+t7UQv2c2S!%m<#M5n-hZC7QX+Wx{1`M(ghv(Ju*HS5Yc@0VR&^7xEF2B^zok$p& z`#`?O@c3`;3-0xZ?L%{>*iCeHq&+Axec9Kk*KSRIDD*{vvlKa@*f>SB9%X4Cb|p4g zF)FWa(Qic7CX|#go)|^FCGTk5w(UZ>!B*z?Y_vI{ydJHJ@YEWvv^S2hN;a8ZWX#rI zU<3!8%1wz|ZB}egb;4e+xfE~Zi5@gcV;-arR`@A&6Omb-(s;X)b z?fK%&Kl60I6Ndm3>M9^uOaqe7flJKlkgb``Lu;0G#XLnGXUgT}Jf?qvYf+^= z9Z1!PppJP~L|Ya^nk=r?*8x3+Os#JJjADDGlaVqdIYS-KJSq&ClNp&vk4Q#()l+2N zN=kjv8xb@8IDU)dWM61-d_%KBML_$ z5$tyb@%TbSH%>Xv#>msmemiBr75A(>=AaME5)Q;cW^aktRMVaIxoAlud6i6=5||g? zS_qDgVnYM)o9H#XcZ&J@sJcg+VGW+ajo`BZVaWB62rRr?+4Uc`N4#Nszqmj!`6pC* zq5By7;7v7m?&^G8x#V%$L}K*kL}JHRduja_iW}ghXPN`bJFI<)Y|23i7t8zMW4UwJ z=RHxdu2KaHdGh>l0dxkXA~=FZSvTF{Ic2TV_)CDrx19YXC;3$_$pa^X{?_0R=rkJS zy|1R6;N%fLBD@Pzw7va^A0K`$L5m}LM9&y`C1AnON@?3|IC`tmWD%H@efT#fiJVftnbr%*-xtlAk9c8;3bD z93J~!Z@rYfM;~$BGpJWqo`C3Q%VSyiL#G|qhPva3k>?1D6;x%ncqauabL}MODkypC zV%;zyvI?#7n~w+8&=_(dU0XlPvHnXvg$YHqLo5@gQ0rHSxX&-wJt!Ce@%*R77xJG4 z@y~n7zasztnfI4*wl*+!_^0dbze&=p{|f9bp0CzMDHFa4K((>ZjBT7IO@V`kr~l&p zp>Lequv3DK0amE7yoELCZsK0opw9)7rY@ATm1T=0S?qTc>|3puUY|D)n105UNcJ#} zBoL{r_MooN>x^>ra?EmuXxzIyTvykDYC=#C#U+o{y5zh_O})#WXrO~_Y#TErtwEt}=1cJBN|55;@{!!9jPBs2Xl< zLD3hZRb<#X9(rFQ21;s21Yuh`zcop$IZS~w{Wq0_cx>dXj1TYM%d~kBdxYE<$?y7a zN&f$=C;OLFjepGFB-JNJBx5w6u6D^}t7JF%kpWFSg9T7zpbf!E}-4#~Dv=xpE{5vrl)lDUJBf$A&@O%6ch?9q6=4Zw1Tq#j# z8#GAR_-x#OqsjPgxD=*1sa63FEyk}P(p77)An4d~bx3hhq?*8_NK%fPJy1ERP(n#p zDO>ymAw1Qa@GN}5Ogu04Y6awoRgiq& z^(%Rd{Ldnr5Uw!rn4>nn6D+wXJSQ77^C$riZMPV-b!Ye`#+oIavc33KmuQd|P5USe)P|a3R+f1) z_FRuWpz;+&t}^n@=dqN}=>z^=D&JG8urXUx@YrA#vvmf4stk1KPVZ~Cgg zj7}T^T>79A66k8dg*w}=2;kbS?xPbED5uRdzj3gA)OlMzOs23MB{rm}JfsZVwj^*^ z$qe$C*)2&Uzl`5gXRALI7F}d2f2XOs(2xM%M|G+TDD3QYO}K7KVsRpnK;jVco$f$! zq4HGkU0$5J{}=g#x;CXB8sg)56K*_Ccs;3VyqH0*m4nMBLiNB04Yz}d+cFu>S0 z1SWa**WO|=7*EqlxgC})+5GJQ!Qmf#72|U3SGw|FlkFByV}9V+NGGx z)nizJorcpgamDcywFUXd{HCxv?P=sGdgrOnGPOqgZ789T$=H#SYIpF^gcJ-_C5q0zoEJ?C_0PvAXz}< z?bVV!eLw6NTyhXrhE|woz%aHUvO1IKPGcYQr9vC+3@^aQe}H5`Lnr+#gr43ji7v0< zhEdsDXKsUfs#9NR>Xcm@KKnbeuDYmFi*cSEbYM4Pu{m)mBJ!3rGfp+Gkz^VM%>&3Q zeNd^?K|-EXf{R^U3s$Y%fu@*TJr5muyeQI&%;hf~lQXv5{JnNadeUE68#$3@oXJo@ z95aCTTN!#3~JQpkE#>~%W z_bTx+=&rtX>dwY)9jjGYTv_c$2WDHN?1*${h4aGcA{9__CxJtIjrr!exp)wNZYJnP zuNCpp#^#3r$<{Ck>WA5(dG=Fy{kN_;S>pcGd4;wp1y|7C-Gw(_fdZQ|kV#3}BB!Jj zsf+_*InosSuApa_D#dp6`8;05*?8Pp>3m%R<-Qwz4q&GC)&(U~X(83>(@%5Zmpm53 z+3)K?opmWFnpjYO40w;>-gBY0Srec_R1N5AHP$SGLAOR$M?4fiAgSmK|7avR04PJx zIi~PLm+a;IP+38q1x(<}1Td1@W)|4_X$cDDYuJonlC5tzL`hi(oMu!rdSg=^R`qzO zce{kKyfK|XUD>sVwF4H^M~8RbXxu_~ijy zj-_U^u~~xLCo7%I;MF2KdS%d>HMeHM{yzcs@NUqu`CjPnK?1?ZMgx554!8GN#m5!c zpEp^ZRmXuWtA3d5uSGF%#BJ@(Jr1C)2Vh=|Qk+-B`@TuLKOArel|Cmm`oc{dl#PEH zvAQzi^lORMWm2Wizp3s$nStcTf)zBRQ_iOJwQ?l_7c(79gmu5CAV~2#RNgsOuoTki zO?R8mjA}@e`c7Nw_zVE5Fa#e%D2>ZQ8Zd%++n-e~34zCJ^Frk<4&~bb(8wPBLMkb~ z(YVo-S=@@}P^KO?X}A3U7|Xa#36ulAx1|9;|MOV(zm$uZ{snKyR=Je>2QG72KVh90vxJcg+yWn%(cBG+h)5f)xDy{&2$1WO*B;(}=Un1jHdLWt< zf^`G)HwDRQ`ulQuzrVbE{p09XBq&r1YkstB)DmML8pCwxmoj6i;I~%drZQRCHt`+h zti!Ess4$-{8#)@Pb_g3Jx_*c~yzcr^AsO!~OhnwJfH5qXIf#t6uJfsG%8?X^KCLKoC z%2;=}Wi#6x8_=X`tLwYAd<%#RCBj&U0gOd*==vvcab*}}Gx;NVZ%5n)%!5(uM_LQ^ zGOebgOa~#e^??yS)deni^+8<_{s}rxQ+20_ z!+(W5>TRy?$JuU=Q$IU9s)MI#ZWX6B$E@Da^csgW_AlLd$&m`SG+k&%V60nPrUz(i zEB1PR;(k!!ow$dCg$DdiukD!Xh!s~~j29OiG1}GZ67YixWD^+27Y=4?_PY||oj~B$ z&w>Gzyz+u{ErKvUucXey++_l*0CkY5m=lZF zk0*@|2#X7}yaHmfAd??^7zODh_7MexNAwB^`5vaO_&V57q^)tI5D!O4jjG`H?|va4q$gEUZX|KaE~Nd^LkCno zDnqd{++qKtEVwvM&Rf0*zo~y4{Qj3^!N~fr!S5fQ90i2WmR4cm!ALhnP*Ga2824o& z=8%trLJiBw~McvZzsVB#E6++;(i;sh%t>a0rS|^s<+wmI=$aGOr zjzdi12qG6jDN$Dz$u$LsbYg1A<8)^Ao=&_&T6f7cccrIr*?|F-&kZ<`v-;cIqQebI zLSL=4q}6nmqvlo&fafvcq)O^((6QQ5WVO0-5sqEH03Qhqh>sGT-k0taYs<1*#SDJ8 zoa-3*Bp#Od&7tmTK);$|RoTIc{T5r{6c}6xInr8}X~kNkO@Lzsk#mn$WOxy~I$J|G zxe`;R7mx{RbOr#h35_TR^l}h>%)1HU4JEg16+fwtO{!!6vL2vc znl1kfd^%tsY&y3DM8*Hn;xC~M$MAW7J`e1f_il{s?XEMcC&i`p2J$f^53uz$b^e95 zGS$-#-HUf&B&{hIJ)g;6C)>cw*%UZ9d1FSxPc-2hnc@-u87`-5+#^KOc&Tn`%rw3Q z(-#3;L>Vk8dteT=`G5s+3iozuqta~0>ZA?uf^Z10EY)b?<(N#4vuBAKq<4mSaNVV# zSB5>=&rHTu&Wfds%-=;d&iy=})rav6gYDK+ldoDr=7 z-gr2!rQiOMHKc=k$>5SToKBK@oaC%K0LH4fh&bXODvd>VF0mIGO9Sp=Jpf>ZwP-8Z z?GP}Wb-mP=$RpkSE0E_hNG%!g)O9&Nw)P(VP+E-{mT~D(Yt9Kzv#okAv5Gah1^^;H z?wH4SsImk&S{y}fb!j|xH4slZE;-4r!xdd?#&Q~Uk~+OZ|e?pgV@tDbW;Q$iI3kF={ty%=t_^UT}8K zr=SkFY;THMt?8YU6+v3jgsj@z2nIw;!_dxclvHsoY3mh zt`HPM1`+0}%f%47eO1DOd?^mr$tg3(nK0%Z7fO=_6y2!Ql)f1$3{NCEWjYwFWhPc6cd#`O?i_sZSos(; z+uPJ=Ly}p=3NumEBfNB3Zy9&Q*SgS7xfL&4qGCR?ot2vTT$i71*;i^noq@t*o3k=D z$W|h{fOov;*o1lRc}vRNG)j&g&?K=n!rn#*!E$quB`J0F%8iR{{^+a=YgC{qNOts% zMf>#M_mK|w5t|Svp=^#EstxBaNKy}N;rK>Yi+@ba{$5}Y0XY%D?p8_qhz7Yu>TnOS z*<)4OgX0>~!a=$2AJu8%9{mdYc!(_LQ2Z%%b`aZ3IF4ml?hum~)^1+h5Jj(nN1DE0 zm9Y5smRfhXyUlzCt=uIpVf|mM^Hs{%T~*{u*UuGe3*~s{fe% z9|vi{#D}-qcioo!TiyOw+3$bm75)EYzyGM*Y*nj&BvWiA{z!yfK(mlWwLFYp`z4@E zB}LnWW?q-StC$dI0l?```w7d;n0}++7l;22T9((4^0vBt;Mjg`3~3?ckybUC?RA{x z^TheQk^TC2d;9B$J^D8Fi2x5IKbf(2#do?Faef|Rq9Bi`5>GnP!~(P_1NNW`k1|6z zC!;JYam+E(kqHfjxgd}K%wQT~D*E1?QXi>z%?^w=r2v|VH>sHHa6)tyD64^nr@_Wz zq-ZzAuBYRJC{emb7H+om5I(&ee=b(k46R&rY^m816OzzaRX5tis_XQkQG0>S(7cw1 zkF&-Ex2J*71edFUmAX)rk$mc(ln0@4O7qy9^@LfSYYije-zA6xOj%(_wxROu*V(DF z47T*=NOYud()9dtR?=i@nu)He%*V}vghyM3LHFbOu^=uhp`vdoro$<)R-8vbxTIpH zk;QiIuafSLXs%|JNma>e7=)$V$&lzuu_-#)^LJsBFdZ#424#HqbGwDjDXtutsaeVtUZ1e7HR+^o%W;w)8={#SnsbfQH?ON+v4KANBvzi? z7w}Ob`s1S-8Ol^ABdq!WyXeK^8wD2qaXQq7w_w7x{OTG}I~a1*P4}V#WE~{K;|qt~ zbz#e6XBarhky?`jm9D`2J@75%xv21HRMbiaKvkcT0degfHcvvIgy>*fV{^QStg=>0 zp^a*WWPfv^;f*<_s%@2V+Qc_eHOlTt!-md9ldiI7I_q-dT4gtngzlV*t&Z}K`Yp{yt*!ua3gWnp}*xlZhj+q)?@89-hP$TXr+Eo2CxbQ&6LPSI?@v8tBTjxZ%`a{r3~ zK6y$Q;Bwb3qd<4r-_VEERvLNlUu?&(=Q=YAkx2hZW^5wQfnj6J`1<^X#nPv@cGqwP|eI>S7Q2L6IJ+sBT8zc!SA zs8gx%GZv^u5wNrWFb@-6+qn5+pYXY}MTr@m5)Blk*Goo`qomMFp~Q7= zqZQ)uDb6;~=UuV}c_R?|7u}l9uNGqd0rU?zbe|Qz-r`9hO{Tx@mCY{jsZjPN@M_O< zPi7(7+!^k;`<2@q%2CK3fjCGnX%SEb*(N`e=SwtlfQD@kGIw;m1!_KQci+DVXpHI< z`@*u=XZMIdzvI2xpkjd~V-tp&Lkd4=mc3J0asXgS)IcUm?bl?~XY4}fQvNB5NIb*Ibm$fgv^Z0x1nQ2yZPs`#Z`;ekTy`Mk=B;lH<*H|x%1Ci zOXVMc)spL`jZ16i_iwX0()jCnN(O1PZgcZB{`rdc`5SS}d%W?*8FxI*^F-n=eE?0* zw%Z1Q>pnX%BFu&Nw1Wuiu{#Jw|Bk_SQ-- z-cxz-rHS}8+`9nYW440{vdw-|6>y9)vrTo2&*86kb+Z@t!aelXg${8WnBc9*>s!>~ zBRTh}E7?``ZKkE|svz;L+Ghpfzb!zktxRT53f(`uEeXiQdPyhx3<_w4eJ@-)&t2R< z2KP1Cv1xvYjrC)@p$GRp9YX4|-O+=4&s@7zvuxC>9IL>QnZ!lHu{%HiSwwUNn{S!<(13t9D> zu8Vu)7bLk69B%rjJ5!5_!bX-<3G=S8sS^jIKGfv&pO%ywxJ%k)B^d@zYUW=N8BB;U@(8gmAc{TRj5+xunbRC8VGYH)j!7U3k=M7FsH6o*Xj)%%s&HVaY~nZHp@S^ z?8O)wO|uLt2QaNHZ1NX43$>A6W=zl~GxG8(^T6Hgc;Xd~q*C?AMDm3`Da%He60uIy zJT<{Q?RYqU)pIN`u|vYXjY^3Pe3LFKXc zzTv-f!pQ6gu6L_kTEmcHZhjmEBpw%#Uug=SBnBagogV#QW1xhC&`R z_NS*wt*%1@CPS}E7Joc8UBsySJe7l?$L>!YJ!5f0?t7TVE+0?q#})WF5!pJA$8ym z-n=Y7IJ>w~hF&e5bSk}s8cG#K;-t8DWvC|llhM{+Hh?Sez!b$6PG(9w_+Bxkw&^NdxFo@$UB}8-v9pQpO~qsreSsV6nWeGU zz|JcmDV<4n+@+kMSSheXCY4ILqIhv|lf!zivQ65?GoO;#IDryDM zbd3HmzMOApAupiX*08xFa_RP6N2S9`wmgw@e5<|5=F{7BHtY30U5WlaN%?5TPDNz| z)h4@y1ri-i0`9d`Mu_CwgkSTRUye(NfIAY+*>1MHLye+Q?pK^}&XvsX0~s zx|Euk5{X(1nFsFPoRHxX`a)XNk|J8VV|lPly9^pk#pe(Ev<8va4n_CI;^8Ed2JKYT z5^Y=j8ddwg*BZ9X_D}%>(PW#anI#|ZKE{_f2TrLa?LB(7`5V$y=kT8_|6jL8IaWP z@YmW+D}~CIa-*hI%vJ&T172e1fEYOC zSasimIIRbd3%}*cA7U&lVAewzyO^t~ zbg8cE##q5J5rR4#AunnyrVK0M@rqCD-_t=@!7RJE=aH}UlDRb+sRSLwdI5 zfH&ULfNYO8IHkH@I*%d7U!_OG9liurNH5zYBEf@WlPGKQ5_Xd-Cw5F87uM8?G@j%*O#6nCfD)7q%Kq#tnf^BkwO_U2Xq9&oM)~%>AkWM5i=ZP^s*z1%> zngJ~6?+a%^MxPSxZL|LpXp`^Q*4APN;xUBoKye05-vv=XyK`a&FVqVw9{tQekwcQr zGzX1*J?fWyX#vkBSC~Jca$4dJQ^10>H^E38*ZzDXl-Uw$h%!Eo6|nCQ%zr73PT+Pj z_KPo&$1P?*I{h}VF~sXuotS8F$CWX(J3K=m7I~G&9ZO^E; zG03PI0()<$8Q@~smuf1EA+#{7W4p2Xq`K3MnK2@^HM8~2PJdCGoFz}omS^VIGxhBm zp7f4tf5j{28yvxWwkpQpHpc2_LgzMNBx*oX)PR>}%hPIMIH98D1BZX58*y0y%ic&O z=!?%!AjqY{F;Nzc@oWzp{&Q?e62Heu_Y3Vou|L9n9`IKOZGr$uI{^TY`ACYHYRRSrW$V(EM*5d4g&-7g8ly^nC zGJWq#Z?@+L2tp#8Sgoh1&Y^Euqay}-XqHRz=bOPK2jsJF7)6Kag3fV%CobV8A`=c_ z9Pt6EJh_LEsNhyel!p;d@^gWdPGKJNiKAH16>?kIT$VCtaJq@i9}RfWsGobyT$VV1 zGG$4lfSBmUQ>kY~oQlGVRYGR-x6Z3!^w*4UsLF2Tr|kw0**TQwyL4Rt$nxQ{z%xCZ zN107dwM}Msl+y?U-A`1ADC$j|rC@KDWWih{TBtXw!VwkV$wnYBwzSE~m{Kt5_pO{? zZV1IVqL=Y_G*BKvL)4BV^n4L()j)=OS<3U~rk96&pEUzFh~-$aSN!2_lKH1e_sjUh z%8@p!!1qh~ec2R0wGHDS6JfVXzs21PfGXkl3c3zy5BBde2(|NLJJ^odggnnrPXENK*Q(a4#M+x` zEXHsJJVM`6Zi{tsiz3Y4KwW~-^vSN>KWRic2$a9P1i9Q8tGiUo?q8(7#`!3G34C7)ubZxoaP$LZcSYO2xPu?M z3BvB8JyG}*PA?JfD76#o?oqQ(PE!G`tn^!RWczBGsJ_)pd+eGpTT5(vt!rfOa_ceM zR=I>j?w-doyABDyqFnlv3awwI1DN_fGAA>CP_OTn{0+=gXxH6J)LrltSqSybe!=l+V!>CCku3qWGrWRYX$QQ$ZI2z=7X=>U_ z*P^f0L*qidD;l~#Ob{jHJLxNJF`2h+bo`Yd?M@h0Jhez z)T_N(U>{qkl`iRAHekr+6Jb~P;FDhYGb}TE-3EsfcfI}j@&(#P_3)BSB}Xth7NkN+ z)&xuoMtKC6>*re227FnDCl>V~>>vB1TMjW5RNBGs9Nuw_(V`a?GNOI~=OgdsfKf=bqPI6CvbG4RfI+9um`H07mZ^HENYJM-Ce%zFXkcOD_$L{GeuF<_NZ z(y|26G|E)N{NOrux&hV(&~+2a&PtOu-4teX>KoL+!DVAq@Z!~{EzOc!A>coV++(%k z6zm0BV!WQwbxDsWjDZVu9kqcG?u$KC0GW;2bv02B52lUcVN4pc67L_!7Miy}my-~k z5m&$C>QdmI2+tJ%PSW8I4?jT%Hj`JxdW9Q2&!}CV_#Th$jC#3154Ymr^S-j6vRH&g z`4|e+2w`pm8W0% zJ>W-=Kh`SD^*~<}-%nNcuih#M%foH;DyWCAKK1@yzKX+pvHORD;9FP+xBYDpvB3H7 zUi=0AZ#K-GNJMR2Z2r?u=6{H@#cCGz$N-$L?y1{^%a$Y|7z>0bqK5?aU|^ea)t|H} zfh-DgE}3+v{jROHY+zm2f6;h)t@Oo)?RLzmnS#6(&78%4pP|<(j@WH6a}Yg#u@OpU zDnK6zKXre-{>8uka^>f}?}Pere_IkwWqq_lB;0>QKsTg%+bTQk(xx%&xY6%W8T~^A zlQ$eniav}?xX(4VZRHs?AGG6TT`@SDf#Z!+22-Sotys0( z-LG0n$JcS8h3$gOj)ifCW-jqwunq0TfWFSjvn5cu%|))n58e!1fER2*l8ZXy(n{W1 z$4Y60;8I>4RL?iK^cNw*M6hng0J;Zp&9ZafM1_TB?Ab}mn8G$<;~TNVL?pH*Nlx9Q zL`r73v&F&otlhGw=!S8BtIl$?Zm2@6i`KfWIU{+4Qk_I-eB6ZN%VlW(5TKA2Z2>6) zwbN+LpMma%!-oq{O$r}q+hKqekGFXdbD+Y*=p_Tw#u#N?vx;YumX&H08aWq1vgFLf ztHKi6^0nJVx`~Hsw3%aq;>KxdP0kvAp0H0B+sS#g+m_0dwv0eY+e&RMTTDV%1AJh zf>(8(6MnjhjT2iXhQ#E?H#8&Sg*do>r~qbaCnE=Uas0sk7ECNz6-6#nt3cLhPd2%Q zb;dZrG@-Mq2zNpCtU#RnxMiEiKv@tDk3h`zb0d%Jt!xNIJ9ZWNDDpSN?XSSFTgdF{ zsOzg$xc=oroH$B%9C-?NB>6cC9NaXa=fMToW@LOK=$yVr=s&v_VD~pTBlpm{y}F?l z_p-2|ryB8W880n)HFY)yk zmT(WY>uJx>Z`o$8s+%7Lxi}qSl$r^64VbYsOtmJv1EF)g;SU3CU(SHTI)A+^o@EH3 zQj$^Ia6VrI0zdoxB=yoI(x7Eh=IemexG&t|hkWo)D~^Kk&(X&?4B6Xc@P#ivBG*v^ z$!JM5U^6kqIW22eMkWMg%ZI|c)?A+;&?6!Eih%uKAey)e1S zZZUAXf$URx@>dtMnn0Nteg^JByRnBoIXOKo-OP1>H#gM~3*mNAkWH&K59l|K5ZYjc z9`m|4O6zSRyjPyfH+f=pA8twi+*9yrck@l8KfngwhW|3)@xTd?Z6@4i2)q`)`uVzo z5Gf~QaToS3Okul6sHZW8K$a~csZ4&Rbmfp7Z*_2LMnGe%`t)c8@sW36f$G6SDri*TDS1%#YE$O!786W=m)z^y84?PzUc& zPGSkVM?X=%m~)b*+2$u7MAT3V3Bv_*M-Va06tI=MLo93w{Zd_6MN}l$42e|?6x_5A z4psuAC?vNKewuxFlAw5W-y8%K!I}i-hPVo4LF?w5MUo4l1$g#KITTtYwJ*!Trfz9b}p78hJ-N`{xt_rerXFP_m`5N6;)1Fn#zMRqP2BY3@#sp zBk5)d;{0a!?KC6F5@Z3DXBS!p_j+vvyDD}Nw$Iq(J(ycy7rXeJX@A>QsH#6T$9GA^ zH{8LbehU!zToTC~4(5urIirZ?jkyy{WUu^c>0-)8X>X~?8D|Pk7aysoJJv`Y(Inqg z0o`0V;bY%rY8^+KDq1f$Oj*)WC3cLjVpA7D1K>-<`}z!vGEk4hV+C4uMM>v~qP*2M@)}$|fbj zyA6N#kzO?C2^`5C99J|?zD-S;ql#-7Z=|vFD+8p6TNtwRxq7v;$@-$<~Yez8m2W4S`u8 z#B6Ud2c#v^JT3pTL6!%|S1Pv&iF7_;_3 z_}YU-0B{XjU2@;ya`P{X;Exhph9tJvM?5!$<)Dg-pP|lE)#fGE<_TxI=jRC@rkf0I zq(OxMD9Z#)HsjEt5kk$wM^)9tu8GpHW`Q80$^ZmmPvz7_^7es}f{x5$T}`(?!^@b1 zVzP-bTma}0$M5@@!oR^2OHNr<7$SWxAu! zZ@nv4R2||T4y{Op?8< zz%|9lILnww_PD~kNp&r3GF!$JWT!N_3Pst>Lvor0Iy}8{dO{x(^8-gDbmHv9n?4Cx zWaKqw6d948OljxTwRft~%|)J*;85|gxM*gTcx>;ELH+UupCXd=;nWCXy1?vT|=gifNdln1qzLx zlVmoeQc}ifsG<~gW3Q3D;dN>=Wt%rmaET{Jz4M;TeW+4|J5N5e4Grush`8IRberh1 zK9C(2#`gTf%6d!15^|};+P?h|3`ov*KXU`)<=Lx63%%%fuxgwjhd`wbEt#bZ;VGIB zG8LAo2Xu>%kDR-$g;$0j1TsnH>T<{xW6Jvs`N?B>o`D)m8Hy?;(=q|*#K5_QHl^&a zYo!tagXW01zlo-0nzq+1PBPt_TT)8mj3Z<23Qh#ESdAE3eXYWboG7E9IVpMspL3b~7C>AXDwO!fP>4F$F7*NUS zGxFXZFLZ8)%IKyFjv2H*k(lg_-5vH&Ku!T6O4WuaQGp8wt*i~;T*|Bfr4$h zZ2lKWc=4Qn8C!XgqPqHwuHlGmL_qJ>pMPyRR~6%rV%(tyv)}N>_!6-A7D2v?;P5Br z{zkCSqA`>?B{y*FJ_kliLJH*J5Lcz0a0IDUVealvly&9P9oHS1gU^3P z{sH<^xIDg(yY_!O?*5nSc+CG|)@K9092dk8KBpaNMkodq2?emt1Aeq1?~_?b_mPnn zfl|ojsRIE@{?0E;@3Vv;?`x<@*fa4<70KU_S8`}=gY zxo!6e$`JK;0Arx_?osC_7u~@PFWnJCh%{9Nr?G;v17{cwoiQW*2!7F>6dKIkz9m|e zLl|d3X(fWD%v$;5&C8cL9NRKB6aqZ z%Pyw(CRv28o2>2jeh%qxBLl9LI*KqU;6IL#v%9V)n~oXXa_?j9oS22DPW|8b!2hysikx%A6Wh;{10K_r=nKLOK+`#~cpg>!7+As**xUP3qW>fmds-Eu8 zByE*r{`MjV84#FGbT>A8>~DxFZK!aFp*%)>5{)BV>wdc7SOO{LIccAVg)B|yQCiD+ z;$s6O+eS%bCYEnMEnFC5%`RlDB}ZsSqzN2#TlMZwU$U0VDCr zY)kKm%z$$n-h|*b1E@H!qv1a$aBvt{s*3py^7oKTNlLkOdMsI~8FxJe2BXi{=mX_I zhq1zinveCt`46`_lwRUp44wid|q)< zoniW0cj(@FOu29RV@~X($4IV2VX#q>ckp~s2^1)#d<(Gj6aiAgKh_DL<&##3Di zyk?6~%FJ>LhbpK#WF2K>8KIEhkj6xBWRXxq-9uY=`bV77%~Kci_m1npcK0K!`$kP_ zmUn@;WImiCD5tiJgTPW+jZ(wsQ^yIvD+MF&PpndA{q&L3M=~2-Mj&-anOC%rE!roE za1Pk2G5e6b$+$Qu z;1_$Kzh3qvtniYiM+&RnxR|pFyLHRB{@mVCNVAW0C3!O`*v0>cQZyl3Lmd3QyNB{` z(c=HhB4=dy*Uqh~)^~9BcadMwfD8x_Y(OC(B+yLs3y3%ZsWAqww)q1CMJyY*HNe6$ zb1+Yko@w0?dpF)CEWb2_t61cGVD@Mn;?K?xUahjO{um0W;7sN7>G)kM@7DO6_v;n? zhuUp+Ae+qii+~_xBD5{>*Z?jbpNc}13aJHVnXwcP+gZ31_b3d`EL$C)stB(3l=0FA%H6dZ^dLJnu|ib&3@Cx#k~-hLn60-jP2t z>+ITEtxYBi)8_Zw*u(B(+XU&$%q$#IuG+#gkQ*_09jC9(O*NbOk2I$kSLSw>fbGTR zEIUojwfY6lWmapw1s}6b8LR!iBoaT30bDm7pRmaZQv-T}ga^c%nm|P9)<(^MqoVzq z^Up!owUry$krJ(b6v*@zfoDnuM2Lw8?|=*>cgloW$#jzy4L?zW0%n4MzD8wpZ{By5 zzCC1Zt{uCoV2!z9^FUaYsaH==8>wx`t1$@}3i?$~CW`lX+U19_gZ}!Nx>2dC6Aemx z$B^tSKOv5&KTwk;wug|D&CCZP0-mNGOA<~iwYfQBr=sARjAo`bn=|d8`bWQ#6W(#T ztuS6FlQ_wbOJXDww`*IPu)tTa0NcUYEjTkxjmn|F_l%7(sMh8P6`sd+^yBojhkZ~r zoXgY$$_-JisA#Gx>vkcvc?joDd4JyR-%MVv^z?ivMQz%Zy<~K^sUu@>VB7Uvm*2z_ z4~HSC43PzG^|wUh9^DBH5lDM`7w*}3Pvl6tLZt#WdSwGXnO!gfU`D#~1s=~yAN93*M#8nu32IZ!;|?MR@w(MVNxI zE#itSltV;=5PW<<&hogafa31G7j5k{R#b>}+rQAyc?qTC%vRI!EYU zp?sMEJZiCmv7I%C1DFu;80s~z3b9^^@Xy1O?1huU#LxETfG?s4#W{aVe^sohz~XwH zaC{3yB0yerh>#$9^->`4mH_-z;6hB(NrQwSk!HN#6_WJxexlWt<-M3gmBRZBbR1Ts6-cGLs$kv zPCi@Qy~iSNbT~P^uU&i&QXrQzawzoOv^p|}xVlPG6jfOPmP-NnQ~}y%0bc5mhc==M z21*PDnk)wDR2l+p8d~+@nU;QJi~l~&_*}G7AVZ6Jv&3k z7&+4U-*Wf>C&J;E0PBWljG-!D#;ow;IR>I;c8NH}ZFIYt!dpc=*(mq{lpKD^Qjd5& zMjVlnWIb_h5fNu4m5wrH*8*|pLbV6bs<5#JnXy@Pf?6TvUO_`U`u5@Qb@g9BjFy~g zQ8XuJtG^$-_vl%A`gOh|q2np`@QOGG(+7LCa_|y0OhomHU2S)=3$hJ_&`jg%dtR7T zj^{kJQ>Y+p=3Z>`_+y^~U(vSUr^(DOe%dNv!=~;1Q!0@6kHmQTUg7ioHxj~sZL7ln zqg3z@{_%g(+)2d0naIL+w$3K*&XP9v*8kGr0h@BjiU^-_S{*j4QleGi1^#8lU@b#_ zngRzSi!4Is1KH^7PYS8lZkd;cv+Rsyx-Xy^1SGoO9=s$6xz3c&&?;T$>8#Vq=c%l8 zr>Es>e10Gr1Mc7-34%!F2!ZO*dq8-hH;{25?nHyiP%|Vn(h}063@?|hnm%k}c?VLa zfbuj`(POD%TdYXal?^5p9ZE|v4NCPa^sVI=PmSt$UfF!i?@TyAXmQ%aO1lsEcBsY? zEHUTQPHkL;s4ePZfn3U5mWI-!&(i$h7UP>p#=%EIFS`sde@o+1#x#K$T#y!ahi0vl zqq{Y&_H;Yj4U`UsXoLaH2JGT8#ew5jbZT3h2T0miK^q2fMM`&yKT!+?Se;1?tP8r zl82Sd$&`~vj`Lr`0#cZoJ@XXm7>(Gc2l*@nR*S0KY)`XHA}gQ~1dh7NiWMTs87~C9 zUcbW$ZLv@gPBVQ;`{XxSy{Q~Sv&(dg@^Fjh3p0YKYj7JyUo*FdZ10i~UvPvT7I0rf zx%?r&Gk2VV^eawRh#KxVTxdC)l5z2U0DqhTxc0W`%WVP=p0 zqMJB@5TV}xiEr?~hunYtcJYz_zZ_tgS^lH4ViJAi7Zgy2a~tYyjYb7j2HAt}=u|xJ z!Ms<+2nVWJ*LyJx>&`P2DK6(-F}|4R;Pngedj~PTRFRFMcR^nAqmN-H(F)^>o;|Lv zk2aj-uPyxE-(7#`gFo-J`A0E2?Juqw{6%k)G*dCY@c!rJES`Tm6@{{4QnO5w;4?lX65Xl zO{Fa*YW+hwrQ}U?&QX$D@3PJA12kqzsOd4n0t|#aGDD%iGqW>bI!ZegesS)aJ%fdZ zy*>Fs?;FoncK+I=guvjQWj^;vg$7tYVQSWqvC@>2EKe#`r!rmwSax;fOhBDTP*-ZW zPy;B6@a9Bu5>j-PnMj31b+$DBFr9md(%z=1=;oK)pI!gL5H@mi^zS(%@i`P~h|`*2 z7I8r8J}8A`2{s4s`hwow@p@-owohBE(W=tNr5DmfkBFri9sCA#weF& z9dV^^3nkCcI_SnUlHnc5xwR7)PApcOE6DxE5R7=I_!e+HhW@LcsXY*aA|dq#6x;)h z_2mF4XCLfUx~t#nj@>=V?SL?=WI2w7ZkN#ne9in}E=~AaIG*Uv5Ajh?=yjUS@nIjw z1LIh-)O_$Xn9h-;@b^DaJO78ZcZ$-i%hq)>oMGFx?F`$tZ96hUU)aj9ZQHhOE5r7V zzpB<*yXv%7)plF!V&2T_K1P2>$NQky(^4jVg}ujDl987J0Y&-t{rmTC|5}S?`HuAbq})hu$csXAu=eEb&hB+&Bbj$) zg4N@5$bKL8Mc`W@qx-Ux8Sz&gR;m&jIGxH--UAtqtS|>3eJVd&(~5_8MU}+E>m3F+ zv0D2|iUt}gvx1t22h~JJ<3H-1+4W})zGVyQulHinE5q^7B-pR-G-oCetf{hOtI+ii zuRJXy`LhkbwpF;@l%z7@_zqv1*(OyXRsD5a!*XWbY)_ru*8>Bke!{QuDdVGLu~+w4 zTVE?KLD;72H-vMN0FU($NSQ!CM^s(symxyIk*pngC3YVaCObHiF^;FibbKW0=%4ZU z)cF7IVk{3yOCg*lj29NLw#)qCO=PhnX!Ly-B>&CT3|zaPl<3wVvr_&IfPu`dH|_b; z|9Fl6zN66B>5#Jm-Wo_P^%V}Sra!E^GESsf6z}$e>k$@^KCwa=_Bp=tLVUE6zLh~l z!~T)hykOphO|MRzYOMWt0~;Ch_=^&tZ{N?WmOZ~wN}Uv1snF9f22%(*W^B0rsLzKduoeb=*SxY%t3;sldo zLMZa&Pna@s&XBLFlE@5`5VUWX_qEl(CObepWv|f^)wK5{$lsJ+Vy_1Zv)G!ju6LhK zuzkm&UmoUBs^!MdX@6OPf@e)-Hqtw5Y}!>&1--HZsZ34MV?8!79U;!37Fr$L>3sPh zI7gm|;VO{ueFUBG{;89=Ob2&A*^{6^jKf7fB(9r2ptZvIIS-t2_4*3?cF zt{GeL(D3oSk^HprrBPUh%U~aQBhrb&#z^lXCmG$7O^fPA=R@se{9+ai6vVexIK5$Qi~kw9^(%-;eq$R_SeloDipLrFAHu3elWR^zy3cEwefpKl%FI1TuyFpi7L50mX8G7miZ)fFPC>hKq=B5Jm zSczYcyN0jG5L*s)GW*BHFtGz-*^FpA9|~%o(JgA2Rc1A+xw)N#y1kuKi>SzCwqabc zCm&hcjzQ(;CLQuo`uP<}%OgkC=DjM$i^mY6ZgoxesChr}+I7&;?}G92^Kp}#%buEUWvGwcHckvp=NW$=b^Y7Jt+fVd2Pm@3mfcvw4oNU=q!FKGBl7MlYAiFkthMH z+j;cfdF(D28y7j@TpU;`B4$vTV-!$%FfH4_$1EY`ccCeV2=kvxvJ`vsQO}(Y>VG(+ zB;@cyAYZf@38#sjak{P4^*7VfmsgMx%SvX!;0)ILZ)R9cb%PumpkAtqf#dZHz$U}) z$FNU_v4K41%>6)bh1C}5>Ar2Y+{QA6Y5#?en*JmwLDS(Q(R1}V?hO{Y`l?U80yGu2 zk)vZ5oTwW=1hp%XG6eE@>XOy|+({{?Ude3s+ITiy92;s%8{s){AhI*f&&r9c9N_M< zLZh(RDN|6eT-%Vl>`v7hUqY2=DQ99SUom^C)nsnV_7JFpN1J>g)C-LNxGr)i&$PPd zeyBEMTzPWY_ISd(^dK8ly>hAu##aH!H}=P|cyDN)WMlG?3jkM{Qjo7@oU8iu=rqw@ zRqoYd>Z`UG{N#t}5$J@U9pO|-Y{vD@o8bIe6;dI1`3Y9DCZEXKPbvkgVL4nX^)vX% zZklsp!;h@x8ecvOPdy;%LZh$KIk^=8uxlv^%j=m6>baJw+$6eZ>vIpT?2>BQ>|j&lQMd2gEcV}^i87lGnbvpD7BsL+C2u&U-(EW!5hH^T=Mh&4=g}e@^ z=zzqmTJ}akwycn#P#qv57ts_hv@Kz=+Ng(#ts`DpreBP0#^?7%MlUg8;}b}NBOLNht5GtA|XqcT7TKGu922gZZ%9!JV02WB7* z6cu?OF+pl3o5znHB2VC-DTpwCTZ0zz4cj{THA*y~QDophOVauY{#POPVJ>-5<4erx z{8Po?Uj^NNCe#1h;Q~QB7h7Xd3lnSOzo+*KC7myhA`0&uosPOHC4sL$iik~ezl=B2 z;%@>Vr8#k6(D+Unw;_{Z=fahZ+@~sUbLqG4FQs=qDVeXyk7UH=bu^jcHP!NHdU%`c zM*i)>U_uD2fl^w3AQgLY-PvJmzy@6GHdFt%75)t|9#v($J-b|iY$}W{nc_&IH?1M~ zqQR9BBa{lBB<-yjO!Dd8NrtRZ)-k)Z^c}NJ$v~Yz740YNs6Qi1Y~|jvgcr zQLl!)hdx{zy9&0mdkf6EYa3(Ahr7xq2K52w7laBd*z+|_&s0e7+E^Ja&3n0%3<~>< z>s8T-aR(*a9aogkQ#SIeZ(wB^P(cGfjh1tMZ!x*InZ}4Jci^s#W~Fvo-TFM4KiM;$ zd|{BIE!^pV0TZ80GJw-56{d0Y7#0Jo*H;8gOEW*(0^!~TPn1v^Y)Hx_n_&2l4yGbe z_I5$Hu87j}l=f1aHua+Lg`BA&PKNZN48hQc?x1a2^=dX@N_CuE92(f2cEa{R%v$tN zRCl>p{E%~ElT3ZQtPteyod-rqz9^XsCP8M2QbA@7V<^o(wC^mxmrakDXcPXSs*E%U zSARhy_Ub4$OGyy0cE+s<<21XW{wwlC_3CJVzOv}WR~z8J_2a)r-hX!Olryyaf*k%| zDw2fhA0jgD>@SO`p-&;HreqN+BBD(ynrWz7MS49JGFrO)Xpu8$dmlP&cRT zm3nANF-e;eX{a=2_081{)R|B1)~8RD-KXr0m&~PB9s*}Bq_&QO{kC{ny%CuvtvPwJjik<0NZHz)9^HwAZvvSaweu-g^qARInEWk3DMEdwOZ`I%lD; zS$g@Hz;=j@T90;v2T6MMbt5?x?IyHHwL`xg2Wpe|P`uBQotp4+EG=vqRxOufbDj|t zx^uaK&XA5Z15Gi?+@9}Ry@}7ZbED2qLpec=ADmLqhEi<_N!32zA|~6(%i-}$VQc5f z%lzh1=8E%s(({nMXr=)DbLaO)p3Tc}J|WL_ca%x}4>J|O_$6i@>+h~|Lf767Kg;Iy z3EZTq@So!`RGQqRuPQmJ)a}OVZ^4${Oj4@5I=!owi_X4P|5BhkN~vO9;# zh4U8NLM~t#ML9S{BN^BRIda$rK@GekBzo^CUU&_~_0iNALOo~fsjYD60C;EBk}i~5n^2|VT;r%Wq? zcJd_mS9k)NUqZOP!n6IK!t-CBH_ZQ8r_8D7pTAJ>8?}Hljm!1&#$-W zzcj3eCJR+qLcOO8IJuAGRV*)k0<^{+v zQaJ&*!8Z1_8uo~q+!}HPxOu_K@-e+qE5hx5GVkLZ!v(~#%h;O1PB?J*lm^AQl&jK| z8QM4Osuz}~(KX3T$cMLTsJij1codeaSWIDNrN|&gO*Am|d6$pOIh-7;ZxFoWuGn*< z(&Jc?r&^kuid2}1e*@Yu#%S|n7dnb6ON?S`DAfyS_te}~QTsAE?6m?kMmJu}eh7{k zKLo{QSWb$tFq$u{Tg_g-Q?7@a&bNPj6qReS8@IU+OWB@2th`1My$(m|GO{+ADycHd zP{8&{u@|+A=N#acnI^l-qHmnr;yff&f6>!ZN&6g~>uN_fK+0$c3$nlb8)l(4^%Z*K z{7sBe+C`H6X;ewdsfGX)A1Q2hIT8+nyhdg;!}CNGhl&iJLpj`MD2i=Y;Tm@%1kj>_;M=M zSuY)@Ftb`oCn(rRC%vmw0*g?0nxp z9&o!|cob*8>kP2M1X*|pNwTBDjClx)Aml*}qO2<-4%(nXMwFR%<8n$mLIMW!4MLq;s%g?mJFVjmoo6Q8W$;--JRmdbrXS_jk8-p`pYlkm5?Hz)b_44m6 z!)*IwJdjREy`d}Xc9wW%V}6cAMGAfd6=Z~Fu2oSC&bRnZyh7bviW z>XOS1&Thk1AP1xI0#f(nb>+#?g49ipBkz}Xd*eE&jHyN7>j=B=;z<@-Un%U#X=*MC z#|kBkV+;pfj%!cIj265r%tzj=t-Dm?Y6>*F>UWo|YP~utG)apQ)~&KT`o-tV0b7_M zvu+KDH;}VZ{h&T<60;Vjir8D=uY!W#CRhOZaUH#waSdTWle;_<(zPMw`L_`BGw{dQ z%Yxg~r+mcHio9#<0#}?8IW`n&CAfiE@U}wl0-cEUIk@D*Qlb_G9Qsu`we05lc8CU- z)03Nd0X|nmVk4CSANZOh)C7A^oZ=-Jt{8N2%AjppOeLs(hrD)4;JjH!mrpu0`Q(4b=32GKE@r`f1%Zz|3Ea0){WDl6cJmcv07r> z)RZ~2BHTLIzAPy3fdud)uJ@K`KIU8u;DE63G&gJLamB(297idF$QV&0;H)I2%A{&c3fmdOESV}H+QOu z7YygS%0G-KnN*~+6Tcrl{gcin7&xzTNkR>GL^^~$6fI(@@VsnFxHgF^MeLVR;knzT zo<>{cYywLNe(#vLki0@d{sfg7I8zT?WtJAGrBBeq4#g}LWW@k911ydc+LI!V1_qcR2y+$& zT|Sp0K0R(SE>r6@LYndwZ=lLv7uR4H1m__ZXA)rrKw1Wa6g3i|GFW~;HN!3!i2gH? zWtdcuF8-WI$#;7f3e7rNE%Y$8p>hvEyC0KYxMfXaR&tH}SNtV6Zo}h#rH;sd>O1@o z)fn?X)L4PaisE0-(G18G1n?CHy&v-kAqv`rgizuru-PPvnn~~(`H{qZ(w0mP&nWyq zd={u6RWuJEo@9~^&PS?ESZv|vnqKLSHu-$oj?NzC^u9@N`*OEBXuzrM+q3%PNsBTB z<%cPQ5=sLYX-&BN8MUGJJTaKRz(a$7frrM+hKrj=5PT;iV}Yxn0GWT7KhnuzY?F|k zz^hc_VMwaut&~bRnfRdqB2H zZ?XtGwaeQZ$j0}N?cAXxar?HX zkQwT^wd7&QKMC+NQYZ9=8$dUNqV6mZ_P3(oDH5m1=K>qMo~sodre;%|+#PIU)wJbN zo0MtfJ2*S69g9qrQj<&G<}0rbdYL6-r9YWdB2zV)T2)+^YaF90(o6s!$H$dKSeiE> z`{k)RD41WWt?u<|O481<(WA~>UFAj>Wm2guf4339J#2X_)CBhZ>rb5-o%sj}?bNBKn}+U>Q+esPB+p4Ei}HTVAF4sA?C zFvB$I|BXB3JPi}970B5C(M|^tDs%p)D4_;?L9H@Tmkjipw2PJ(%F zoGRuV?bTRm74`ALGCy;Gxt0|9P0E#3+|}`hd(`09o)y^9za*lR@5b$m8;X3l`S@#- zhhk%+77{29Br{(%A3soV3kZj#N{B^@_!q&*$MQ0`y0tfmpSm}Qq&0F6d_Md(LnPrB zf+(E$?+Bt@4hd1t;9)7lj)&@~$fcP^lxQDNuK+ko5IaP|F9cCc2YGaMI2Pegid%Cf zCOXG*#B;*26$aVXl_&$yqT#m4Q@$^ba5~|6c7&VFjLjOwiedhF;0-JY)u7;#N(AwR z?E#J7Owe_sZH%bDoR8@74%z&sJ??q!F&7cb*||9l$flwdO(GYU*W*lkRORCu=y`lS z6*NcX`XI>&Zf;@wIR(e1r`)!`RJVwB?q3Wd{EdGwh)|5I=gq$INbkS(?EBA;{{Nq6 z-~adMSDdh&`+@MmCUH*E%tAX2xS5-7wGbqwp(2DMAPG@K5mzjc7|SV&?2v49aV1`( z+y!$-o%|JuAn5ZA4~p?j6r~5Kx}Etp^6=;3+9SQ$S0hoUpGg&JI8U44q}8vC%@)Ic zE=G&ZcCDDfx~*f0=~@hz&%uY+HCPi~Xa~-yVx1!t6u*sL{NfBb2ZKqB7Vu`opCCyl zhf&+`LMT+31JYAHoh0d8wC5M{;*AhzT2kSIhj$~6#iKO&h;)%9--${0aK~~pP|2e2 zarkLUlnL$~7jnGG$bW#+Uh*bCqzKDRgn80Eh>Lv&$F|Kyu^Sg4ql{~vGgxh@-5yhx zv$W7r!oQkG1e2}TV z-Vaj{F6|Xyv_lTF`BT)S`0c`h&+Rg~MXQJ>JvA_`;YH!q?jhQobR)nu&EQ?n+6EUa z4oZ4S;qUM9@t+yUnnm7PBE%d`~~~$wbgDmK4(c3G>e%t7~E= z{PC>!KyBMgTt|8(bHwzvUu*)@L6;*Pzsl5ma7OE=sjg|cb|L!f)U&%)u{!;bW?(u7 z?lh{bYmeVZjN22iv~%=-z%7MT&QKTL&#I079scra$d`8h)rKVcx9{}->{|A(&6s~+ z=@KJk;D7`Wg7luTzp%+T-MvY0>RzZyg3TQGFH8D>)`BXNiJsl?-+vd@{-?R}AG5U?${K%vWOL3xZ8(Hd=Kn{296c_!J?liEID>#AU;@K{bz6j_$K%cwTwjBB9Xrv-(;mugMkFA zJ2>OyG-N{1qpi)co?TlJvPzM`X-bI*F%q_ znTx{REA_wN?D)(Z#aXj36S7PZfU+w~-N+jSP?9Q`+il%_G zv-R2i6_uAiG8Ns~bl8CE^zZ&r*QFGx0#D#ed)TS5w9y*J`1r92%t_m9K%H=E8#=iK zsZFGtXGBKa>K0B~7@@=ue3BsC^CMB}Dr!UG}8RAHy zd#Mx5Yt%fK+A0iE!i)9kJ?A>%F?_NGij>w9iWv;IYlo>e7NSu( zv^bQ+3bS!y`V;F7Ooj^aoJ|RCF6y3hD6|-u8#FPGVb$yt<^G(gx9+*QV1%d@=1MbB zU|``?NmmIubxfn{m4#*p)9iFN5jL6-b|%S%Dl~g3{zp`bA&LWr?I_EPvtwUcQjSe6S@IOJxGw z8HHUxHy5t8Pinx|%wSVPPm-eHCbm0<&Ey#3&8NvTDm#B~DBa<14n2W;7q1K=c+TiwH{+1YD3;6hcjy zxKD%yPt6pWXu@?|dmBg6VTj@Ah$)JrLjpEjHewLH^_Lz!U?oQZ`MPny{wa_B56>9m zKgPcjRkoDAoM%65k!p?GY=q3A9v}*78zsK$Xeq=Hutq8?ftU*kl-hKxtOJ5f*y~mZ z-mSdLs^){IYnV^dm_KBqJTGeX!3}?n?`PJTO}$ULP4XUHWc~cSq5jpTuH=BIM8!Ag z4C65_evaHSUfLY$1;wW#FFzAq;=7()OOZUKOz}G&4_+B&*QvxOM^H>c1PLD1o6kDp{){qsd)^xmoOs6f_a?-=v#rLXv)mbAE;BT z=Ug3JTG!1rj8oXY04Oxf`4$@?7M{79G^kF)K`13Gl$6K9|QNiGhBsZ9? zZ~jEK4rL;U@j5do-V+QStFyGKOF=t`PxF*>-`PfS3{mTBovIFynnH;waEEy25}#;9 z)Tm|0So=Yg8s~nPnXM;j6(b*6_c^9IbKUl8IaxyCSzz>5`-cHYc~q;*4Zya{)d#8p z8dLVH*h3@l3oM~@r>92eM2q5+0wQae~oW6VK0RuY(GulZy`UykVa<-)vVG;X9$ z*FOf##&WV1?Ay9qGv4fh%la5HW}f$}adm{&r=jx>e4x!&&Dy5TzDUi!TGn3Bv}3up z@u$8%botC1w!ex=|Fkw8>V=Sc?neTuPaLu2@bCGc5bC30&NyFV<2D*+--Jz!P>-=f z@}*LU0(PgKccB40&exw8Eomooeap15Bb9;CSuXjLtX$Fm zuu>9g=dw~o{(+>6uRQ16h7uKL;so6%W$zLMHHZtUVH4yheuHA1C9)1)?0nY7o*9nD z@PIbCr+ra#LzF7ENhf?wVXN0kf2T#a-CM^_-y)2u#&-|obJmy3sWsym$|RoTh${~( z8}CLz=F6Gao&FUwU%A7DC`AvT{d$}K4dF&B{O%Y&xP^aA9FB=^n0zSj-z|@SqEj^qa=iy58$PK>MJbk9T(5GI% zdH>E&KtjpiFZR(Up;`69ZJhJI)%DkH5@cZob^djmEc|Z{lQR8dma0H`!xl*Zk>Ee)C#_{)*UxcTRISF(0 z15`py+l-#tOcRDFSA|x5O>l1R)ZZ863Bpo064o!>1Mdor`O5LGkZn1%*&@;JK?Et2 zK3hKM&WFkFn8O#=qs&5Wv{YD_IU->b5oA6?>v5WF#tx&d9yb607hG%xXKBJxVt^Ev z^ZJ6lT#=N0mce@!_+-pehdm$>G1Sg!%&kKS@Qf%6^L$;GO(uT*g5!4ic{xd86&J1L?5714qi0hWjDoSQS$!W`+cVB^Vl*3jKpS5`b5T;KNlo;+c|9%N1-9 zeY7;$eG|kSc$08g7I11HZzz=ty@cBQX3DA9`6=Wnl9GJ0)xQP^2GT(5P{b{0 zAcU#aMdI=?f@DFbpZ+!i9SDN{sq$5miT~dOnK1nq$V62`4qFZ7C4fV#D9>2O86IK#IQUuzPwSn~WA~#eA@2B2`0}=0T3)n3bZ6uw_B1 zUE<5^)?6?fuiKwrs0f6j>6}MtbsKfsuT`JE@3=p(`C{!+jF4VpZ-^q@I3MxI_q5Zu zwE^)Gh)Q_Qk{I3f5Qy-tJjV~&Ljk%UFR3AzkU-;-%~9Ct>-+mjA)Js9_G!>iAw=9E`}<=3J=6g?e+WR`S1y=UGG$^F*W47syz3WD=8( zRG9O|(Umg;VpX7&!b%)kfVnC}D#QtqTPss%{1s_RNdWO0Mx5#F5#9o@N8tdxMQ15i zmjZ;$r8{koy4GB{g9OyxFj$)mXSAXcDo!BY9NHw2a9!G9u~2+++=V|@ROore7D4vP zl*OL1j4jYQ8iy*)YL$Q(gl-m{?h>75k~0+*#u3kz=?W1bMOh6Q`iMlZ%UlJfIcP?c z#L|ZH6VmXvws6bEX-c!Lzyjms!35Mc5+XF={SkpT0acwE7R%W0H#@C%mh0tiL&&mD zq;+;}AUB&F5pnkb5EY*9#cQ3CI(7;MUhA{BZu8|LNgHib?W-|I=NZ#Ct%ixxR6kzP z=vC&XyL4s)P_$FC(To=iBH=eSPoFu5Xw7x^JR9{%XN=PaF)`h$NpcB{ zFk4{)h^f=p4{}3uVW3HY%(g5IP1gH-FyZY*f4rNHs!6ljP5T~KIUZ4#hq>zwLbwlu zAah2;Tck8*Dh(2Qlk5TW9Y#r9R>g2$0 zuPixiqJTbnFfT<-^I<|B-E#E;3$;a5u;~wU3>|EfXH!1Pv2)Kf+Z>P7obKL}< zGCtn{_kB(~UVfEQZ{lmK!%?k`wWFH3)oNRv^S;XH45s9jQwCh*#URws$=DHv&vT0d z#^p9fu>%RCI7wDIh$)oXQ9*3A^%00|~q8h;Q?;1hu4iNHYU3m}hIM-Vb-% zGL{t!gkjmd6!(_$UufwCMcwRy^~35<*m#l%PX#8pf6O?1{gXM0t|E%Z7`HK;h0~)# z#tgG0U=O@`?YIvbGB`PPnZf;Z z*?@+gDK7aTu9_?n`xuv;OqV^}#Q@}Tbl7tEBXxeR>nwql>OkB$v*PA)JqYG#6-N&KI@Ks) zJzA9z>-OalA@4BZ2)RTgob>bL5kPm4#rkbGEt4KDdSG>ja8yY66z1J1AdLTP5xDe; z^tz3>gE~XMjK8hN5{7|b`R@D4f5iwBR~p}|lmD|7gxaREKp)QYw)tMJt;c0EXk-#QtgkN|Kfj&9(W%xMSwJ3$D;%8$YT?@59 z>AQBX{eYZU!MdW8)EPcUbrdo8dBI&PeXg4IJ0i)SBZ|t^%Q#^z8|a=)rkA(Bb&S+t zIuAO&GK1s4b(8<`ALl3j-C)_@JHH+3knYM0s2_Gr@$9#Uv57!J!;K2zP^Zey~H6oTT0@a0zJG8L@>T=aEKL{dUSg|eWXI$D3+^oQKmerG4&K3@Zl)j zWd}^$bqAoIYW>kkhmw_#pTfcdIP$m4s5${(3#8#y)K7pgM{lJbYAQ0R4Z6N2l4QHF z8MpmKUd63T$>%MlYoE6tlkf79UMhUP`P)NrALNcNMR@CxS(NTFJyWP}f&(}s4sgCF zy;b|KF6^c|OOAKgsCCzXgljFZE~P6kpaM#BXR`6%zZP*qf7gH+5aY4$0iV8^tC2KQ z7y|-;^MGf>c@WQ_NOEhyPI9`(lYZ3aRq|rlz^A#KUz_LCp%6BOAXaL{I7tQkkVrFb zT|fRVvZ^N9m($JWSEAvKRuHO6ZYbJoy*1cD=VV2&j+*cyi>ZCrOM=LrV7n_hJ17E} z|22yS-{~SmfmE)Qx;cjuDvT^jgc4c4vS{&OMsm*x48)9=ET7KDG531~jSVEURplXV zJT-1@I5(RS1(I|U&AsL(um=atvakCDWk;bwwY!D@iv=u9QxMU~Lv0A0xp8nA;ZWK_ zibX$g39BaNO;(l*30nxJad3}f8e$!dxd^@9SeYcxMbKo7%qUy|osH<6@=rwrV6jvc zXof5sUXo}#DovRQr!@fzYIX|Se~Y0zkw~mu7Q=61EMhW$gLrj9IIC*kq6|-4II&|k z@q5sjrry)t+@eO6fV_NL^ z3x?8G?R|@JnGGfxs@WND>Cf-ei4q)C;L~iTPW){G0`fihC~vW%m2eG8@s${lnatkD zjE+lG9#(PCK`V}Zw}%z_Fa7EDAct8Ff0DVUwuXY_V3K{pmJ7NmqlcRy4z+@Ch60Y1){(qQh|(PD)qg zeie~cZkslgv|8D`Mw^CToIG9N`=dNfnJ9V zZA_v&h?a*E`4*ssuzjI?Cs{Xy8OQNJcV>RPmqBCs^-3oVg1R6lkYp40rTGIp22aoQYn^hi2$DNTOY zcZ4QKpuS@2^q7KyA8qy7nBTD(nK93a>*C^3#nN6b#~d)ziG+=XKSX-;P~T*Fwouo} zuKtXWGo5CRWtKipVu4Sdnyb-tzewsX#sgM4g-UaBz&_=#vm%tvv}qKt>L#44Gq5pd+(TkA^pI7LiU`mOcs2z;$!%6zmhT4?jj;0i^R9OP21jihGrM8KC1nFUjTrce=zl>_}Gw~3+2&@YligvBckwQn=im1;tU z6N3&q2utXvBi5DBW69#?Y6?D)SnhrkomQcqraDSz{~&SKOM~E!%T~w5LIgzRAbNm+ zejY>HAjsE9Xp*SHS-MZdU7OnIESCbhsFgA>bV_PMyp%9wGBHv(Hc#Nwdp6rvJdEc35&qX8z#ZNAe=afxE!ys`m#0^O-5bXJw4 zlcvB<5d1lPuPsG&;gqN2I4O?PEE{@$z$*=6{vdP_INUFZ(y$=M@rYa~9n^>oMc^b4 zofXap!}Zi^ZZ4@czxr{Q8PW5)(;>+vO{%_KO}O@W))`fAT4R;7V?%YK&Nk^|W{o|& z;NCQ&|K30IyE+%z)ZKqAeeB&STY)aBa;1mzZZ)D$AeTr8M)j5@f!?O!u;=2r%EN%Y z%0)CPk>*s<7+K(CQ%a9k`NbMH3TZ{MFjgn3*yC+QqT+0y5h<-$L#iEXWqI$uHAk@w zv9w1C3q>OaJL zvHTEW)I7ef6+MWgbR#%-gSA87sxG|>Np%cv!$BWHdvVk^p>nTiZx=7&Iit0%5uNzf z-tXcH{TK}5H1W+ddFepK)v#Q5+hyXW&86s~(W0@fztYI=cyc4*vRe8kFJ`J>{+K69 zGn*NKY__|kzbfhbB$=w&Y`+ML9T+SOVhhLCnt5!X0qV@W1(}sMFcMK#IESOut0G=A z&j}DJqc%(chcsyu>|&IrBkVLZoMdruDBF$Y5nklo!?@_AIE7h{11K`8`Q;Ba1|flC z46?oo{YrnC^u{9;>qDh!_+UmL18PC8VQT;4NPb=Y`cd`r$`gQ_$4T+3ZNw%}(XtFY z2M7D$!4B`@Ke@sBgO z)du*zv&7U{NY!65S8bqq-7W~atHt8nWiG)Nr-Q<>ZSZ7!?Yava!P_NCFtYK`zEi#l zRksVC8JqrCRi$I3RkHyS7CeO-m2h(nh$@bom-&9gbnkd`?6O~tt3`IuKJjogzgC0h*oprc%-`oSv z^QkSx(|`B=yS)-Z=?zX)ap0U{8n+EoROx+kbyfi$3f-(|B>qIr`QS$fb=T&(iW1&@xKYhLGFgl&2WI&Wm% z(*>Xl0(Q#AEq+}!4vk$=Oj5%%qee#z9ywll{0qAM)E+wgGwJBal<0%AFlNf>fsG8X z<^bo)V0)pTzCJ?Ygub_f_j0M58Ti`lM*i-`Z`%Wc!y6)Yjqsq4?)l93EBG1;ucu7= zCv5bN3lANXecN=kR>V-eA<{z#6pzV(A-i2yMth}keWGJg)YP`9v@BrLY^wSVxAVGf2)oSX|GSkxEj z8>0gzrIWc^;ReRu_~<)lty7?uGzk`WNGcsv=upMT{PA7{tyP}%=HjH4BKM{Ac{_Cq z`^B%L@1rQ|KA^H-k2rni#j1<#YCS&Eg+#U9X;M#VUqqXryXg=7yStyw?e)$<@2qc3Y%1kLX4;j-$UY#p23y(@lx zLi|-<=ApF;#e@C!Et=?mu_Mg*4-$8xij~}#4G3>jRgjBE*M33!ydoZSh$5K45BcgJ zhKRdEHeO0-NgJkZB5y*ezl-$^C0H1JnXko4oSLAKETt@J-DHm@(>a|ktzXX`zcAcn z2=NL5h?sfuWsCp^KQfYaR98h&gvWkIJyHHbV~UWl1HZsvq(V$V=GGGKI+%~EZ8RbM z8A%5~>u=ke`*mOcwKUCM0y(2`Nz1_`)3e>SBNz^D&m;j2Mj2arEKRntlQJwR%~#9S z+F5Y5IgK%8m${wB5==`$cBRw9G?*+o3(Ol6h`~G(B7LVX%XDnI(Q&T&hN?;=UjfGT zOuUG5ONhS)+BGA&8*K~G9a{-qF}zzxruBdzg=E%e;=HUh2dzYOxukC*^0Z7v>3S~0+@s$!3sWVlc8VGx9ixufcMdTi;L*6frY+*Jp8!1R z!U=otlWQ=9^hljm$oQQXod5!E^xGE66Fa+2DZN^~1GJ@AMhdr69vDeP75Z53glA6j_;6zI>R39yuy5H zDI-Hw=h=PlBtqfxa~6K}O|C-(lKlLZ;v-D^8t7R{{-gzsgV*(AX-3wCYh7A3S{4|{ z(RoG)`KAbP? z!XE5NmE`U-N52a+C|mgaRh#=tPUf9QqU)dplbCw-yGXI!bMd$Up#Aj7qt#(;F9k&; z=w^8O3Lf;fS@lA3KU}+h(_yaCMxUL&53{Q(d9ET0u(znIwdJ%@^b*dV`7rYicp}lD5=1TTkFQ(3oCI~ z;HrP>lXeEW|H|ria>H*8f#_%Mf+U=5Rt*(TK$FV!GV{LVbh}8uUb@WU1Hljx5yO`~ z;fS1Jf!m9?nWVZ*zxkLq*lx$7D?SmyO_raoK{djnqvo#Mp@&riRWxd}3f=<_m44{_ z;W%3+;nx{=KI-UXA@Jg~=--!kPFiKj`AEB^dh5=J$yW;kly}@Zo8|nqhm;wTInh14FdCj}?ur%h=(o?;Sx z39>$^x$kGf9ji!F}=mK+;?ouqZ=3SL+M> z#dlv$g`IQW;s+=L?vV50Hm0U?rTy@Ox`D>ERFjc&pRmukG5U7HTa!-q?CSid?|I8+ zXS~-)0~Y#>wX5TULm@G%a@l1GDvfUp%SOn8^SlO*H9B6a%3Y?5s%lMA>Am6xLjf{j zFG0(tz1#I{j>&s}fkhf?8Q=@M1cTF^;CIGhz775YyRNK@*aWjKI?~ND3T|pI)~5q9 z#Au~Y{}5YHfdmr3K^60fdk*#IvR22>@1AAQDFUj7w~nK(kj8I)und8t_97M<#-tV- z(Yw~WN*QT+f7HyN<%HvnU-{59`~aJWq(K$Z2hYEIQo1ZeRV~Gqg@!-IK?uy%WHq4n z7FaWXCnXrTjrlf%{?2ru5xoEBocE(oB@|1XLojzEkcH8&p96fwMalF7U>|qKDfNOC zhabUD$+GaS_R=EZpZertMy!94g;gnn7$r%&3S_1agVm;)Gd|%%;l=0V!c1g6`Sq#E zOQNUw4Ck0O>T4t9gmHqi9lT}t(f8d)a&-~!S!cHNLzct5^ilZ_RX6dMr?5O7{Mq*i zp5Yx>BQx!mD|hVqk$;hZL|ql^?g@rF^(-h*6J5=IgLp|8E}F|C%II{>qbIQ^p_#T8mWj z_)o5=35t^FSF|Cj380FEi2Y9vtFnuXYZ<0)VbOX8dXl-JFDv*a?R7SiIe|lD+-{~1 z>rNNzUTz;Buh-DsIJ&DCgY`lFU`t*Aorc!B*A&Be-b-#ZH4GDfcubaaG| z^-gGz2VmbHCSTY1cuLlmkt9i3g#jkfm|zjBq4qnw$4XI#^w*9+*MN}4;`v122LUnQ zP7)!+e&k4j#9V_yMS0>r;l_K%GZso4(;N|e-EEAixHV1hG*|t=Wi@=zj1u4{moqQ? z$gsADc_M$6+9rAH297eB%y(mR@xpVpuI{iIxKV|bYVZf+BENCs9lRv+j)8Hxicd{g zM+#LK4ej{XzOIS8{nrk}ccJlkh^zaN?U3r#Nip0nC(QdBF@>4^=i@*NEAG^GSX*)W z?N&rZK2uNiR;Y$)r)H7(O4x<()P@?VrWhxcRz2M%(Yj!$dLm3$y3p?eDzrQ@MxdC0 zP@T>uf%)Pkm0}T>4mv^G*{u?gzGFH^pS=g_+S%(ECFAE~p_6 zln5b0xckufL88opYa9{4wSQf?pq#}XGL)PsN>v}O%nWxsDxZN~wmFfGq54{$B5E6# zch~Io|6%Q&f^-d+Y~4A_wr#s=mTlXZW!tuG+qP}nwr%Uw>eVZvd&Q3Kvm^H9fBQ$| z_hydF@r)AdJrqMDZSMJ&fN`kT_z!j{6&P~#{#8{03TMMk%IwyCiJ8Zqb4U0YTDnZ? z%j*6ts|^JVGw5+dx5(uZ+9^YKTOd)>#WcnYo)C9Pm`#O#y6In?3x9G5XD5iqC#=rd~N13vyj#B(^m{VNQdW1SsSU?{7J&<|D#Fz0G&o^GMt-ywuz!m#M5tpc{@?c zg;9C4toH>;tr)ybvSs_WrHf62aDK^r4yEQVb5xWj>dirfKgz7nGEKZ;dcTo>GSPam zq4)(W&0>30bFlW_YLXFht0plsZ4ZwcCdmePz zUvoWMJpXJ4c6Y6Zcl@5joK1TfqS4wuxU@cmNm6=CvDxtiecGw$QYf-%6&)jIO zdLl8b;u6yh4^hyqkijqio~%^tGDB-{qJ>4$SR`IEqq7^VT=9W8jrdBVTRH2KUftNV zC=@xn+K^v4?c;1vE@%sFc>eu>_I}m_v`XCdf`)aW2ZnXKfx|(m5;1ryDe;ypiZ*xQ zpW-^@!^w&}EbeT9ns%Xs{f^mH+jWy(%1YREh5mln0q!6hr*%EK8Si@xgQ|9$30pRB z(F$TQaGCdSu8}BdHgpl`Z^yw3j_muUMcqOK|4!QVgvK>_k=30s)7N@zBgrq=+D7_A z0HIHL)))B-y=x;dGh%H56MTo*1c|(%KO&ZZvgY^|0-cV&2*Mm$hE=Qaf7!KZbvcoPZ#w3$gMQ8ktU=En;#+I@(z~3$UmlPljAeXm6Raa#RU+l#G#04gh zKHUFHR={gO6}dpb2<+XC^zT+KHhV~mj8?7=E%Y)XhCd|$etunH?NW&1yScySXbITz zt3TyOtk1-LxD$A_pAnw`Eo@+fsOil1hK0@ki9qOT0ZxM%klcbQrV?5fdyGM-u}|N~ zCvZqfGcQ$K9M-P-Mrxh4enlXApz#~QVGJ!-Qed!?e8ubcK`eMg$ZoYt%H}c~1l>pE zYq%&aDJ-p(Ssl^Pz>o$4)FUgx{4{m2D+@$b$Sl&sGP+_M3K%&S{VyS=@=-eFVhS43*G%o;7kZa(Lsm?O`pP~P^YEw$t(6imFzDu8X#gPo$?`lO0%MP z3psT8)eVB-MI;9N7a@wJMWQ+MaL^UHF~fzNT@~rRLEkF6TYNs$Y^O)C8><8Ud&j3j z;gD}v6{5Au8`hp@$BR#jff@vkLw!C8hjQfSS4?b``$QEpa{klE5bv~G;0Q_00eQZ{ z^>@pHA5v1nf`T5o`2Z3QRPnlOWlSEW`QYu29)a}^7rn!hQ$5IS;tI?pVr3E)>ZoCF%j%B)e~MX9P#S>z9`7<>s;o^RK+Zno^f_mh=>DoW z9me#QQcAU&G~1nz5vFxA$)2{cOG3{r>LgZ`62AP?qR594XyDrFTL8}ll{gGGU z{v)oEKbVf`n0RPQCN`Z}9LO4(% z)|OULbaEz|SXf>f5jK`I;}8nWMxmnXz;l{g%1I$Ve}eIPtD#v{?h#9`H94~=nQK(K z+LlJRpgAZz5$`gwyeL;EpQCvJ;^wE`XGBDhOA{{G2d{-pF-U-91`cvq1+X^{3k#Ox zrTvX`MdUcK5BtDgi}56#ef0~p@AXguaVBmS@yQP2IsNl70MLlh+}7>~j(^!aP%~pJ z)}?Y)oynH)R|)9J*u27sTyiXpJqC62udBQeutqSP`+)dXG5=E!CJu_`IZIeyO+`ZB zTzk&N%iY+#coHqG+GwAmITH{+$=a&xon<*wx&lbpbXuj@mXPuix1K!GoMKysd<6f6 z_g&o^H>{_2gi9jX#b7qSfVG#(f?M&=m10)jgd)n#m1YC&Ro8Na5R0<$1+UrRB7-e~x1ck&nlF?hf2UsQYJp>W-! zS<%16XY^Bl`pnV@o48cNp9DxD<%z`d2ybx68+Ory{Zq<=m%W%l*7cWC0!+{ZC96$*mcvkq+ z!2aY#>=8q`Ce0usbUHW=Jydo*WMH1OV?6eaLLR{x;Hhrbyy07au*dA|zSo(QwD(iz%} zM@?}x5@DETI`FVyFue5SSYTD>@wqOR)Qy%#$!GZ*H&?<+FSA_rs>Di%9H{#AemNP4 z9s_qVWfba?ACDs(k=jjpO{A$57fY;*tc^Becm$hMvbiVOqjtZpvaRRTH4{?1lUh*P zdp>ns)>!)!ay!0Z9c`pu&NuO`PY^O^OxUtwVR^%{sM8~;5I}klFlh!}+dTrA5MUJV zF!dBj2WN)`mI+}KgArS;NR?%o!^9(~WRMU`o6w0G54TG9mbElMv{^>qXPexr#Pb|E z0BRYQ^K&`;v1D@{%v!c*-%*cq$J|){$;8ZGt8eB%4^Y{~IZ51LyI_B5WeoB5*Qiw# ztz&(WqOFOyOG_Y}HhyMfrymFcD4urZ8M+i48p}j$v*QRdG;q>sncnh+$3|<#?rL{6 zP}ERqf9);(h=>xa?ts=8jr?4KSZiaxU)XiMB;h=;B0%fEH5+=sI9UmB(3DVZ(8a16 zM%w5J*$P2#0k^kdjOwK(8?FQrmvl{E|$5C>;|}E`g-lgESpM&bJ_2!|5+T*5sH@`h#7AkuT?j^pGNId@DbbLkrU}d>L&Z`3wY6G}0V^!dz=4}=LTAPk#)k57Vx}cZ zYAtZ6@Xf~LxL&^6^4D%(f?vqw!KucB3&{1SdRK!~tt)Zu6f`P#q|`L~2ogip2H7mpNUI1(ZZ_fFx=_jwd=92 z50$R_H9>Mxc$&r)F{EXk5E1rGlc(mp_Z28%PZ-L1X&7m>Qj-hD3)9@~@Dc9#Y%T^( zpey}5+;dvzB@VvT_N>1Fyo8b%(?POlthi346~;gmtR?dpt0qO&;jxuD$zL_Ozv)LI z$;I|gMyEy(iO$%Pyy{t|oEhAfq8#Tci5*eBZwzqJ? zdnQSgR<1+lh=dg;!>O`L-Hm%IL1Ddlq`{#TBrce$U|_OAKRNTZjXq0q#ax{JZ0*At zPf;V{;4WA#M1VY?bO&ksx2f2@srWa>I=h}~i$-b2i0VC+Oo<2Q`o(ivWmXC&dZ_c< zRI^)mpZRz`CFdNWp?v8+ECC0L95(k68jMj#48mW!j!dxt-Di9ci=O;q3!Z2cDL}Vr zq`(A?f|s&;=nG0A-~9gn zYknr&|7ZR?2A2QCoTx&&ApIP>9UsxBuln-3fIx#I&>!@w=i~&)3qV=YQ6@E|t5+h+N#rY^=NI`A%(U(1a+*PgA;VKIHZAFyYnW?TnrjFJWHBM1PqiLr*|+R5 zDdFw2qiZYV?63=08Q(!ebVXUyy-5jMPxLq2#7CqfBjse0Zy%ka06)IK%v%{c_(!ZG zA+bYYy0&G`a-g*-3^^ks-LFAe$lq`|FgT$8gluHFyS#DJZj8@(a(?>EzN5SObaP~b}7zI&J+>i-G(>Ekux@?47p2XGmhnsz<4#~K@f(!3K zIgYQC=73z{T(`R|!E4Y3VgJ4uy!zo=Yi2F9bE1D3UHL&tO?S0{8Fgh&}`7 z>&_deB;n9E##E2e0d^T;Te?Z{niHxwIJ{k$#Uai%`UJchemV(LtD8Mi6S_=z^I35eXM;fB`oyOqA~rry|@~$Yaq! zn35iETKP-Nt($^C0^4f>-j6|D>3T^}Iq8@w-`u5UT^~_{o@na>rGZLMmA7p4&$j4s9!!4?^MZz+R)J}hQ`i1lxzn(&B7F?r)?Ll$zB zn*U4RpDcq$^IWu3tn@U61e7X!lmZ2YW^Yc$HO^}FAsONc9Ci7}5ozIPKm5SFZQ=Fr zNa2}<3e^A@#^&sVzrZuz^p9W(WRbh71!~$lHTuF&Rhgz0X=aU67BGSN2hl=pjX#Gz zw+&$uGkCcS_SkhrhVWX>9yxo)FRc;&6fnE!=+kAa5yu8U7mW9;{I9X}p^6cWJn6mHqQs`r3!z503*%lDwS9zmI39eN=E(uX?Km>|yg$w6?GU!L-lLF*S#f1~=)Lufm0`tY$ z3?dW46+E7ri6YL&>3NI-05*i~&HHO*dpM^lYB3c<#UZ%r5{kvc4h#jA3Uyg2a?kYt zd@_+<4aTxY!apt2O1ovNyG^5?hf3A=4XATF3_4&bXj?ztQ8y0P^;#uF+t;Bpmi`jQ zTt-*E$AAH}q;_mdx?7Sj_DxtTMEzSnpE_Xph(H3pmUj|oG5^?zAU?mc9=@I=E}RTC zM92pt0sw;D-}OEU7U;Pjpxt*;S9*FiO>g*Fx^W|aB5k?uKs%m+3lr=x+3;AAN3ztP zsu_Dpse5bIhBT)9xO8Esc+Zc)aDGz*<4q+9ik?mL0|zPI+HR>B($)~E#JB88jeNI^ z$6?C2;y0?C&GU+rE;9e^mYeXQfG zDYi#`enJZCeeNn-T7ZFlI)o%%yitjGmOMjEAa*w;r6geWRNl+P4hytxm5%e2Gg+3& zXCI+jHahmgRv*zy{M7Tqj`{LJ(oz2oIG-JasE&lkl+TaB_7`Ft0$KVgj_E?Evq2cN z;GT*l1U_)_l?bfqoJ;6duhUxSHU7qApJrucBtzPox%xMZ4Zc$GhSBI8LG9BtY#XaQ zr!Hdo`-y1BJj2P2s%~%q!W(`Q+()v2f0j~AO$Z*YCvj1dv%s3Pk5BfICP|n^ZHmsK zZqE-Isd@kwX)kYbPC2ix1rZq(kA1gt1r}apsR?_eDuj?UYQr>L?U36Y_XL!hE_(<9 zqu!*VNO!Jf^H3~`D@G*_NQULE>DCz@@rhSWZ57&@WxYBq*(*@zF*VI8Jc0ui%yu4!4#M_>Kp}&YT^k~_SFF;)^=qJ36lA=E8=2RqWEnc@X)JD>VWMo%|vX>#+8IcbrCCy(RH7uFgNmW2l65Fm#Ib*^&?nl zYe<3gp2-y-w2*@rpCuM*{6QNl%D`=^m50-|(>e#%0ZBwL-gS`5MK9Sc5lKj~EY*`< z*h=uJ9uEx?s4Qt#kT%vp;T~KS+;olrr$jGL#QC6oiyCKiV{nY;t!ozdmZ!D}Z0 z(ulzV<^oga%?d-~|Dax)AmmL85}AKn5>Q)OZfY()tm)HkS*pKPzo@LFGEa;|d>z65 zn63BMA$9oKkw$9XuhyHdJ-54>oxQ)`-*JD@>=m=hzsSQvf=B4nM~Qh9RuXe*itU(F zrMH;nrSM{f(|hC}BS&h7@45j1dz8m60_nBr?iGb1^sgvf83No&NCb#4luxJo)y(OC zC9LBlL${yY(rN0TDMxS6+4l;j_Nph!1I>nkp_uTFsU>&#XoUK)%h>Qy!>A@U_IbMng*Z|`n_ zz-oSWfWnv&L*0{(XQYyD2}}d1$c2o5HV7btbs28sy*D{Qw3SvZ<3NfCX+n1wqD^7z z^AzTeY90<4@{T+7bNwk*jXimogXGIc8t_^2BF~M)J*569z#xznF9$ix8whxh}=dD<^Z`5XiC8 zV1xQx5_769QJ8%jGxPp3+XuKYFNrfJ#i=X7D2Z(=4P{RnHcW>TYU}Cl*eS7hvYVB| z&Y{Ta8GrSF`p6Cq>1l%J^<-k6P=Ce)ALn)ncBQCASl-T3x@}#ivG9QTL=4-hS0b}v zqZAY-5sa2S+I6J2O)S>2iWEmPrr9h|T*UR9S3Hq0Z2{Ie)J*npYSE3$HZYr$?#Of&oA}r##55S~=il{g9ttRQat5p|<77}A%^@OZS^9=#m zij7A5pe#AhDj7w0wa-;Y3Pz~XA+rmw#!i%vs|5m)4``R>*b}=UQIm|8G1*N2KrmL` zPwW7;|C?%G7l~ohmK%)C&wk`2!N4;hdh{K!t`d+)Vgzc{Kj454b9N$L3e`5cho7|0( znQHd#y-8QVcOug<AL}3-RuZHKeRa!XZ@#)jWJ1W_5P+Yx93#B{j!UEhQMKk~xyB%+UMCOfhjGJ!pziN0cKXqvR0yg3sB4 zLSF($@+6|48V%>RV_}_*K<3)te5(>Q`xls56d;jk&`|`w|LCD(ENOAIH|L>#KqL(I zQ_;2rSt1)wFf4f5KqX(q|4~>)k)E1(l@tU7bQK<7ytfpMr8G>ETHMVygInaCW}Pi1=|(%o6PvhOu?T(O2vJ{^6X=6zu)kDiPIH8`Q{sB&f{8 zZbrxWCRLe@bcYgPQRfPp!g!=_aKQ?yEL@U)4#<5To0=1yR&T#zSELfydn;W$Dw)XK zN&`4{hpK(ytyHsKao@=6eb77<@F`C}WfItS+&QE6I;l#Mj2EU^h^NrlUwl`s4Qq6{ zUU?_wu6+?omiT+_w#*%x(#EuR1}y^Xd5(eS<>)Y*=4&L7CL9NDzDK!2PmYjw3I(2z zB%I*TxaWh?5Nn`!J|Y2hPDC7sJ<3o(3+BO}(tyX2SAT`cFY-XOvnvEo#EBfX;3HfL z4^O_Q+pMzZX_4!=17o}KNT3FkSpYOTBZNxbB;&!$fTt7h-;Gy49RoTZGhVXI&0VMn z)UXg{R63b|Rw|ewFJzl-F){BpE3)O+04Xmn{H<*VGa@UN2pmVMZ@emaMD_^k#Z2gh z$q!z};h~rxil5zyQH%-enPJOE!5~8f>djBSuUzR%c3gYnaqv1!zcUIHw^h%_uF(nh z7Jg?o{2zo@3ZxeZte@ih^Z)XG=3kC*MazFsQM|0cS`9XnX5Qp2Elv5L@#`v^nlQou z{6GLr;^%Jn<23t|4y}ypO?Qo77IAk(f|t-s*yn#PX%1gDCSfMwNO>8av+9qt?m3UM zxQxGFpQn?5arX*9=C>Hiy)%Wcx9B%aPdF|2ae2_(1`3Y$ff-S63}*!1zsT=V5*`zy z5TO7w)Mq{AvH9kHEClho(VLL@&W05)sv8?C1paZJFFxcaE@}S?!gM=gqc6&*uBO&W zZdBgQyKY(!i)Ku8tJ!Dr&pD+mUWIAcKNSsJd(4!QVaZtgD>2Ekfns?%%0b*0FlGt1 zG5HMKM2K_AO1XeLBNA9xI;Q^|dj3*n#l8syPRH*Dy?jkaEO(UGv31;x6~JND(4e7G zIReSlFQ0{y!U|m4(`t6G)U2|pY-fKCR!n`{*mE;K6K>MYasv|`p`d!!?MMD8I}<7` zN*YSK8GmI-`=R}3N^|&ea<~j9aVpCb!K7|lxwFVAUwWj%>M{4B<8JjU;aJpC$zNJk zWjw3&O4T}i0?oVWAg=12Z+(7J3E2`DI+PfbEkcZH{RL!OJi)SJhXWHiVpVl8ZJ;gk z`qsav5z)=pkSD9Q)y7dq`%P(s%S@A%~^}czq0N`bCTzNYh=l6l{YPdLu-+eO7(= z8Mr`hoIR{t4D0y5+PtI2I7o76v-=B?GG6<{(!gG_LgE!i!;%szD=jy2-9Z++?Gt-i z=G!MrkB)8%_MB|OIwialT$=1pusoXiF5Dh^A`*?=^uW;e4T*F6xli*AOd;K*5G~>d zWY7yJcM3P}8C%eg#;mS>VUOG8eRzyIa#8&i7>gSoJkq6S0&~5~K;#cKs$|x~sOmCx z;dOlaTMXv(7C_}jMDj0hT9YJJa^9cuNA5pyB;e?m6jkt^8Et|T%>?9C_xMu#&@@e+ zaki;*22{o2pfxz1T{X!1T}1QcDb5||P6PO(ceG;bg@{j`&(z%f zdaZwJZEbC7t6915+uF#fE`2i~Mbzu{h-2${>%H52E1>wV`vd(i3Bl4Zp%!JB>*Rh% z3gullV3rcHzDK^RjgmbQ58mEMJKJzSSdd$M4^<(0E9xQ3myof;zib;c&sVbasLo@% z`4rqi?TQ%Uh;!59^zY_D!tsTp4l6%+JnWnyl~mQ#)lX7b}t=@ro2LM z@5G<}gD1lRb}p)a-35Dd{P2+>=_>Yq;V9Y23PE?}?1WRGh@(XV+lWfTN^IgJ*OMOD z2nw-u73___geBY*_*IV`FC%dl?wkXt6XCGLo$I?%7#08lJhmA&T%7&^3d@fjItqeb zPh0w91xIGyBD-A~tnb#rfz=pi)bW!P*6o@+JXaV&9bkx>!u&Pg<3y{uF&F<0)MV&*6U!9v=X)eih ziCT<`jza^8T#LS(WRS=5%T!L{dSkKVd2Iv9Lm|g9Y{*{qxOwNTl%0g26{)a-3!yCp zEaN@M#Js*TyUva|J?BAVVV2G$mFv{?AaZu0wr5amo?!a0l~X}^ZpOVYoz*%~a8wB$ zoXB-GiVRb%-I08l51>mho7mXw@|-qvln!) zs$JBV!XV>A3&gDX8_ExLm}$zF$e`^#VBqfg8+Xp?U04@681ID}Qnn9t@7WubkI2%D zY6_DVj<4KY$Zy3PTIlZ~Z=?^EehC+~{_mTHAQhX?pzxT58;Za4HVK!xa4vj^zr|sVR5rY@T`X^j(^V66jbNQb*vjNBuQwT)H#ipB?Ac488*i0Qr_OzUsYJ?WqoiQ@o(-4#uv?GMMs1u40j@>;I3>=Q?I10;vHS)Ha z#fE-|%1qxgOcutk8M7)S{qmb@UtyjYD^}(xmu}W=%^GzvG=hA_M14H1T_7Vtk35AM z4e*Lw*_O!^l*bM`kh896c6TLXI8r-I+4`YQX+#=~JsmqR5-%HDNi&}~dCVjSG&@YJ zKjeAp3aZ*O)UG$&%k8ZqR8E^n8nuCGti(p;RDcWpg`Hp{!o9aCB2^d@9fux5SJcQf zvgPh%j6IqYk)BhD?a)2m-mR;GI&5fqeJezEU2h3Jq5q4*;fbs<`L}W;N|p;2`El$Sx|IKES={9#>|8X|s!F5F6b@}f1?W{(2}i^PrV$;Iz|=C%6X(_z46@=yb0E-p{>Qv6eAe{Y)+{e zSfCQ9+yXPuH=%4%7#K_FCA>_5d_{biFmh;ce;ir zgsED>hkI6&F1IR^#^u66(`YGsNOrWE(_haAYPx1|f&F2cUnyKMyQ`K@NQq+-X)k-C z?k^c-KY0ryUqk6Ic}tb;#eHs;C0OGIcUbvOoYYiAdeHCu%O1i8&E6b@-r7*t(RCM3 z+M<%H4)$y;Z;kL!DER5#i`*d7*JmGDQ%ej}6BzlL_;9gg;2jNknGT}r zQC5XP#9^p)K>K*NT0hBQ?dAMMN~eWBrU-L@?3U58e(x1WHo$b`@>zETY15@J*@|BD4~G%_VB zUDhvME=q%8ueoo&)9)6$c)(6iCStCWK)@Myr8BzZq@^Ue`(tbuMhQ9Gl|(|c{4zy& zogP${d`8jz*FEr)9u8?S_-S!s;9($;joW0bZ=%q96t{|g`Ymn`)8=D0WJtqqB@_0wmL z8YJhqI~w^dsWx}c>nRa>!tFZi@`%q(r;R4JCbPjor%#+(S@WR4q@HgI@l!eAPm$Ae zg7tBY3}|5c!hC^=Gq>SwXw6vA?r5z0U3~HQyF9huo3fv*Vb6Jccqnzh0>)8b^~4@^ zkzn=6jfo*aYAxW5p$xEL^B>aW`Bjdf)P%j%Y5}bCB&LF%+csqs4Mh8BUq_E~X&_L%fGAbdd8VmrE^B$a_m+LqdhRold zc+{H%V6y7TCjlSgJMla0SS%ar3wz4`MXIbjSLms_{3;*tF>m7EBi{&`3TPkyF~c?? zH+N+Gq@hV+{a3m9-+N%@|Fs9^v$NB6{V(U~|I(W?mCohof3i(sq%biQ_yWd+;A%{0 zd4E;U$XfyA0UH*Jp@hkcaY3 zFio{e>qU*Mq!VRyieh9kkPN4_w63xvEw4rqolB^N?ySk4W? zs(ZfC&+2e|L6lICPNQj%n&rHdIr%cWh2G&WFPCPP7hkReaN=s&PU@IZylLM_)e#{w zDn+iL3JWvJBQ>nn5G%3a>V!sH`(;c$Z0W^HsH}EHReZj|Teo$^K*d=jcDzc`rRDIB zeGypOp#wNHguG%?eUe;Fcu}JE!BtwO!mN)lx+yYZior=`N1>~9qB(~5E!ci`oXRXk z5goO@@F3zuSxSFLL@@=*O@!`^U0~H78-w3GLH~@QdqD23!x4ENMQZvy^sPT)>c)L(w~C(7${b9J zPY$%Ix#)tZzhjQOj$Jip!#uvuo&ZUUTbGld%qe2fB2jaqz_=MaW?#4vJP z>%4ItB>=Z~*M(9@P6_z|xA89fyCHejF$&4V5w5ky*m zV6lr!VSf_Bz?hkRFc$tpO&YWNA?%WU1$@PN6?yeYbId>4fBP{mr8a4wdE0ng)#(K} z09cSbA8y!Z9k&`!ukvhv++q6A$~Miw4)?igirT*ys%bIxtDw;6%J0<;ETg>H4D$b0 zK;a%>L6|rCQr;^Yh~PCFoB#U*tIx!4C!Jx$R7~Xc9-TLlXr`b`wS=?~0)J`Z60)JtGfg4LSX2;7{dxwzJADyoTi^22+Fzf9%`>$1-1<8L>LKflUB zzr_EClBBeJ*}rU{A`%LOsA~PE-UgsCw{uPQNkA+-qT>eq zB88~HGq&{xun|aX zSV?XHidz;4*qXf|n)s5?1QMfk=ambSX3Sk_=otfqTJt^p>fhzVxpQCTUe;L zZU|jkwnabnz;zdY(#g<%SIE}LwA!&k>Q8b7zl4U&7|Mbg>?6!ckzi{IvJ=v`;7OEb-In7towmSL_p^$oW6J5nw9#2LW!Et3F~4Z*Aa4J{k)TA_-#je&)zcz)wyhIN3)oC{v{zyfIO z_5-Ej{rcK?Fy&jW{~Vjy|0zmf4@!97UE*o+_+blwWJ30MHDHI+*gs+Hb`00TX+)c% z*J==5n7Ow##b9nM^7z}QGf*INE3gqWd|&k9RWY+ouN|w;Zx41Ra*`UL2iA zDKraqZP{FOm?HXAYP{M_6AA^nj5bJF2j<4GJ)K^l2sB(=fCV|16;U&%wXgS8p01-@ z5|u~gYG|~Xk2#t%o_!j`0T+HLt`~I*cq9=+O*>qy)pO0>2?!+@OczWAQ94BL-~>4S zEzU(-a~zCDprhS2n^Mq4t;bKLq1X!<h{>zX;G!RNVpV8*AarL>@p5p}9$4`r|qkOR->&_a0^*`0bqoI95-YRQ{z z;5>~7j7=l~nS~Z1JP4MrRMy0=8SMv2N>aC;f~m_*h$jHN<1BDVzuDkjNt!AB(YUj= zmYzNQS-TEOoVU;G>4P^aby3wAA%9)M_Ikaerjaqhp9G+rPDyk8jNnW6~z z_c`-Mbb!<(XUzy=gdygAc;ZVUysmdW-9%C>;BDhTmr$SKS99>ogyL%z9dK{A;q|AJ z07kaJY?_6uf7;YdyYiv|k=etl^g(cA((IVS>kG^xuJr6#MTALY`&E>5{+*SzT-G8MfJ6L#1ivQ-M~x zQw~}tPMuV5K0O1o%D*o3GoC)hajnC)(2D*^D)K5`RT>()RA;(%f=|Fm=?kB<`XfAsXcc%IT zlwIGrVVixw`b5%kD4F(cvzE&kkzSUhVDwF=)_LLd5#dAQ+#AjlSs zLC}|&Lb1O^1TUfj0P_gTOoBJH0}%tH7;p6B)7fTdgBd~in#YGh1I7q!zBdVns$UJf z)(nkQaA}T|;ya>Sb_*W*`bw_E_KPG^oYzTG?yC<}URLA+UujCHGxkyW9ct?p%g9_9 zuXXLZidxNPl=!dnMh z87=UgR{cByvg678mT6A@{Loi{v1DM99l}!nWz-d0Du&sz)lM=Uo3W!J;mV8jpFtu8 zG+lmkOW&oP5yh>ULe0C0i;m%-7A2m_!mCL}_EY5RR+;J=vR`Q$u_?>8NW%_PJCLEY zj#POi`2%UXRI?6j(CbUJ5y%*ePc9cu5<7~)qj0SC+1$q28Y_)DG3)GCI$|Oy&W;&< z>+Er&a%a_!upy$s;@|^MGEF_J0vePyOg8nJ`H}X<(?;gNbDQ(U1j`BPsG>v)nMFkh zRg_(pgu$zDJ0J2{RkN*DI|LsPDaAC9SQ14rZomPHD$w&JS;O(55f_=BZ%LTZccV_8 zrjz!#H-)fKARf*|tm)&Q%aNh}cx^2LXt}6%SRbePxRDGYuwo#WevPLhaNDsn{)i1h zYUEkF&~cANzoqx6*MMPh)oSk%ALdY9XzTs6u4mt&^N0cmm@7R4U0B-zU;h-mpnOKd zw`O_}3y+I7-z~oL4 z9eyH`L#gsu1zI2JLQ(jVruMLMvq@dia#wiJ;zVb*(0t$MrV+aI`uIHk! z6ez2u*4V0DNZy%pXvyP!v7xGL0@Dosz)*bk#k_FfG0#1$5W~aDBl;!`3e_30E4F3~ zoIuiO_UI?5(i5*Ce%FUi$|gplq)Rl|`|UC%K8`pA;{^6W=6w-pvLu9Aja z;4PJIuX$0a(}F(fw;){XE1Gb-XD_h;!n&n`wC)rCqunrJ-47Z2`BX^$kH;Qt|7*MP ze>9T+=ekNoL*-{q_>#5~Ao1rdFrjNGSqOrYoL!RuXz=k%N>57&YwDwmur)~O9ToIz ze9A4iZsN(gw$v^yCn(n{=gH~L_s-YaG$t9~9l@*Mp7$^Q)r?^SIKi}W8f6?re zhB$d(4B1NF8Ho=t8M!c0?zDwO+X@Uw9k|4!itR9VF_Z4uZy?Y|uLE3#um;igne-q_}A^r{FPgh4(y;?Sir^N`hh%tYtCr8cTKvpEzP#LF}zO+(K)8CVzS5*XA zq&BnYkZXcrI*A7yR8c(qv>9i3Qb29qrk<6wOtKi6VP;xa#g$|^$%M9I88nk)Wb#DU zUpSDvxP6tW4;GO+P9ZR!U{u0bxJbM>HZH+FDOfBiv~+CXS72+%ptHR^GAt2Uo5ZBg zCUtbgva}!`GGqwT%y$)lB#~%2XJ5Xx25wg8SfAu3qk_yaxp+8NaB2v_j)xk3;jEy~p(wepwwknye8>^3T-T#T=mD#$;zqA&NxX zq(EypI41)Q=2^TdYFDoZH;3ey^hM;jj<;c9rxo#NaeF7HqYx*IVkQrVQ^u{L2E5tS zYF%BQ*u*DN4V6y^^l^h9{yK5^+hMOq(U27?4Tpz;a{EZuLKTLJ?BytixFidJWlxF! zR5Wi!?IOdW7T(eApWn}DRo~fkR$l`dYIR7Cn){k(8_h-;$7{p`axpdt^x&DS#5-L; z(6B@myu~i8APE*y$A%((6i~lOvw>6x7&_u7M{hF#Gec5(h${8=IPZa_wa_J&?;3YAZ!K|0tXg*s z>@oTR`Qlwjj=Pn|9QLabq>-dgHNW{rz+qmpxgEOL7R1I4D~OE?<{Nwr4Uq+$WGBUI%eMi(zznyST_`^SQL(iB2FO}L z$qQA6KERsJ*Sa?38moCpld;^2oXB;`R-%_fG=_t3Lx?u+zze8(CKxH#Q5WlT0W$7^ zZnp@Y;1cx|(}x%B_#}JeR~i_(bp5p)9g68UaT?++_i2~7JWU^Hz|PrcP1!Ty+_~Af zVxwW%TAh-me}QGlgJ92BFZl0%fY0vNfzbmnPlxZ;)ts0XvFPHg|BtnIjIOlJx<;!i zNyW~NZQHhO+qO}$ZQHhO+qRvGlRA02-yWym9_Q;GJDqM9CdDM8K?4VuXj{}DYGrw*& zR=gO3Dk8)OEhqP7QZ(FnG`llacDRqfaWsU#C($v#e7+mM{N?`7+fMxdu?7FrJ&>s)gKHMtM+6Z0U(^7WZY-afeOSj#zL@J(CewJRJM8(&bQyk$28xTYLxlmn-}eee?_(b zG;cAPx$TMqWj|v_hbr_%TxbG;9@?3!DT5j|8Wmzu&Ypxi zHf^}Jy|%q?HN7g~I-tv)94B(_i#W__Z0fvGWAmnW0=n_IFT2JspW*nUout4d`f#Wb zoL;G%tUy#9>iX;$k6l43e>_G9mP=Zmo&Qe7ta~`0su%@M{xYBn(J@oBKW8Ot^F;I}`#3m!c zXx9fZ0ua1ywuco9e6FSRk{NtqEm~3RI7>j=>_kfn0Fe3N$^vG9_mRxG3eJTH_!1`n zstZMca`G%O7z~J1LR6{Y&9}{pqEjReq@(=QN`(>kc>klM)IXdSKk3L!*V?9jJdH{o zbeck9wR~tt8N{IAQ;ksJ(+y!gyT8y&3sNaBhXrC<{5a<^HLKDKL{ioFs+`CISZ62X6{Qg(Oba zFk@@cE2b*`%N|KISJsqNrFBe6oJQQNmEGn^O?`Z5k=e}Kbj#Cbi{*Z6s-xfS^&AxY z$1nR1em{(wfLyH~7|Fw;G12~LFtIss!StBKeDF#1nAE-lhhX4_QPByy-;!>! zy5Q&egTh-X3|mzF20Kh@XX^uY!8Nl+X2^WQpIRca$b+b<%hB41Zcx3=QK=JqEJ(*u z-3f2d#_WSg#CBMaZ0IWXb}M$osRXit%OkTAp=6SDzjj*DeiEwMkn#mNX;$gGh5_pV zX-?WgCtILZ#pkSklQD8dPd-Oml9`vij@!jWY*{fjgrgT|!M8wxl{(?Ye{ZS0Y?6 zvrp^0GAFO2W)bva192wrx@l%j+qMHQqwjz~xzKjR>F3U zjW^A;(>WTKP@;!1C7xEHQ04A^II;X@l2(~H*72v!r3p)uWkr*5oS{`d1YZK(-~$$c z;a-oAHC0n0M3aqcoA=ds9HJ_}d+D$?q2hO|awrKK&KD;wLQD$T{XNC&g?sX}Kz%3; zoa7v*(@1QYpFp&19yz zZ@)|?opRdguD84UX~ES?`c*e@v8#=ohKR4$B6VIy_+8)& zCtIcofRlf(Js>g*i)inwx)^60{V^&?xKx1nSbj%YS?%1;Ns*x)(8@;J5D+)1s4BH6 z&N_z>W~@3VW*j(}Ap;Z*x|kbrh(_L8$POa~p2n()w2ntVP+@GwMhad?0HZVXpr!ymTt@(=5-@#DKx z*3@9{jL1lEAhgOK_cn^XFBS&Rv(CaZOP`d84SLY-yj z7aQC!U(~e)s6$G==YLwSzC+j3nV<~3ZW)N7NQv5H&csV z6GUx2Bx)D^`J)QEgl!RGBZq!khijPavOAWlB~ibAHunxo!8|_~7dfKnEX7%y$CVb* zfDR(iF6b>#5@}9cUD^3Pw?Twl zP!Z*ev$KjMi4)JSPD~<3%Oxc|(oYER(AQLNz;3 zQy3hL*ph5e?jPO|9V$elFn(QW={Jm+5?bjuCI!%R^)E!p={GDA&9b0MD?nfw71+mf z)C<#S`n@v77^QA2Vt2jE4IvFr4h$6>ypzXjw#yL7TY%Wu#e8TZIcnuWVajGeTa(q6 zvR3Fh4Oan+UrNWZIoz_?U@F)=eb<3I+_E5%mmC9kx8?Dhl*3bbG)`+au(4;7udg?} z$?Kz2wmRuK4cIKlE(J0ySDHB&j9i!KNeb^AG8{~=VI9!2n#I+m-QIo_CrsMdqiHut z3aY)xGslVq^(TdrD(D;u`g?dJjH^yWhp152%aeDftxC>EfBmpoQV9I>XK~$=XZ-k0 zkrDgYb#vY6fCHwU;ZMyp*|fa}#&DBN4S&?y4(9a08Nq4F`dAjnpfOHIwg_!FoL)| zRS!mVFsmGk+Z8get@0NfbUzE_w%+f{{m|bCgYPA|?@6d>!9&_7B&l<$bGYVV;9fdzPCZfo4U(3TdDd|P~82P6*^vB0+c z+={)~t%-V&KEB;?ltqv)+oOK>6+{@;6lO8q#%m8=vbL~hZtN0xIEgIcCaC7By~Q`M z-#UM%VqfyK|MtqddQ)*H{-@_b z9VxaO?Jb9OT`%zJMzufUTfZKUp!u!Qd2xHwX%&OAPGM!&Ssg!y+9{tdHq!lksR>*h zykPeQZdn_MTntQJO*AgV93K+?IZ+5cL|#jCV)F0A>ufB6Ot(Y1*T+N=@HeO5xf_jl zD>z}cyedIQREzn*>#op0Wu0Lz0{F-iRxZg%S7C=sXjT330$?u#^vDy|f{Sw$heb?j z<8I{0BbIcb8i&ekEg_=nqa}JzK4`TA!Owwa6=(;!nLy&5Jkx&8gJn~@<)1^L6|#wrvAk4OD?yK`>AbSDVwVdgwv>8W?T2u+Z;p@tSD!y ziyVY+^NkF>g4i<>!h%Uvq*6hNRm%VOLYd00qX#L7h12U#=ssZPvYZu`0^ZqPV(2vR zh?1JwE;0}dC)zSTNMAa@-s2L|fYS{WGy-r_X-q?NF>a~9iFr+G4;8uNN^?<;0Z&F@ z-zvH#!3`#0sxdp+&%tT@o@qsT2PCsbAQ18rrErHIt6SASGe9S=kNU0faa_K}f=9Ee z4}g9_{45Trs`DWMQ8{heUety4RVlZP{O}-a5q^5~g03REVF!5I(j0={2?}y3DzpbN zx5wx50Ouf`=~7w-ht>cY_X{;#Zt@~G`zOPQ@?gaj&+X2M29F7|OU5Ks9i$5SY7(hwX0)}^CGJKbJI6I&-TaqbZcEmlusqFf@vUB@GVE0KqQ#S2mxMFvO zw%;gpp!4)&SOvOg0@tVO~)kXqhsQ8&7(U9#sW6lFKsS0#NvtzXJxcJ)wG${`}@v!)% zQS9?#3<5C84&d-WQTP5PM8Yif)))A)n8_#lcQ@<*b`HY)KRzPLf1iU~N!3Xt6$X4W zw`@eJk^13CrfBM$Ct3<3P!i~;TTVTi`4#28XDlEgGQ3Rli1Q5kK=B)BVg+s3X5u-& z;@JAf7g?xF;xYB5d|-QGqj{44{MxAL{szTsYTBO?gk&v57X@4JkV}Y2&TZaP@Drq& zC^haNDlD3baj%N->qC;Ea6c~h>`Yh$VLf>r`4MUY-FWT0U63q&;|N@q`fS#D?<1GBmWKRj5t>sb_|C@HQaT1XX&;71`^x!E7SNGutwNC2g!CeZ5HT zj%po@79667DmeEXX)B=yi*+=@kGJU*DZ*{LKI4a14S^c7 z=y04`j;Q$*6YEd z&@+tnZE?!Mwr-G5xW_w3#m#l>F;CAPPu_qJjJR!chH)XVe48ks(7L~u$Wfk+{ba%= zH=WjT_TvWow67&X7+3ED%yUqR*F-Qb8Ft?TjB>yqYLb$M+g-c5BV=)=AG`GKpw~*mjte;p5fBd4?dZF zM-k|lMqsD!p6Y`;67(=2EkBiBfqH?myaS1u`98X}`Q~LZ#?8bO`Zb)-C8On6JK3di zwyFnOsaX<7=D9k3^6g-6-Q!C-vBOVfb$KE%zITVl?yQr}-ur^a3*cq{-4JW?_LW=; z_}e#K@PF=B|Em1>&sr^UhcC^Cy^)cMy_utt{ok8arV6CT*UFGL35u}`y=GhtT%T1a z*eJ0fNVw*Y25|7Y)4E(jUPN=dxJ*j__*F+5($eLoFyzLjMe~X^agEX@CF(d31L!4X zbt|*RS?1Anrgt(%^uk#>@WCN0q0~SIeaNKS>{Ma*))@epi zm1>mVUEz6QJ5KEA~DmM;z&yXP1yAggmdroLJt~V4Q z>9+E??~M`Lq$}i_?g~R1^i$%p{z-+yE~)DHxg>-MIAAzyI2UGCkWQHJuvl+Iln_|+ zQ5N3`kXeyf;@mrg`ueV-Y-MOS;Mwaa78W;Wsr8ghahZ`qhV{b%i0B(!k#ME8Ii~&L!T=T z)a2<|g&b34+!q-O0d0=5E|cNI4jaTS@XGgvVU)$w+O+GTjb z4NAYE>wl$hQ;}|zL(8mTEIHAub7t(>>3l%TK-wG)zDA$=1s;M3wO?rQC0}9c&qP@m zhb1L=bD67F-nKaN_f>-YB961u!l+v`+#p}Xd@%y4Tpc4VblhOcfhCDd5&AO8E3_z_ zU|v1Mf3!EaYaOT~=xp_NFFF1>VNobKH1%j{<+Z&5Bj$zBbdY8lJ}9YGe`H!;*CVFa z=@qrK9%y@v8OTvwvtMw987m(9PHqp&kmYYoZkO=%ogoR=DeDrbNP|k;g%W@>nDba^AQx-v%onic4 zVqAwB=T7WKNST@jZAI1}abgGVg!Kf?;P4PaV|>~*byZa}J!z}sLw?$>BJr%K*=Jd5zX@`Qb1^reRQBM9D!UOKZBY#Ub zNB>!|N7GWYSKT6fRy4bX9$gf$%;v1Jo&Ynn_aRKU3u&R3Z7k#()xKZCe#>Hde9 z(hHx`i+uljZ`1*n@$~dGaSS z4~IqSx)h5;+|;Bx!+;u(lW>@`hC+xtJX?}9?Z)Oj`qpAEC)o;CAgY8l-zTl11v3Jc zK>sM><1c-Yv|{M+TM|95UTanXu?ERl7mh!X2VA2m^@MtMAC7$DiU59GW79s=S*s2{ zJS58$>lEk#QNvZTW6p~uHg%~iB#8{DLG=J_Yf4>%rCJrG5v5)&O`I?+l7>68%2dJi z_-YGT{>Buzz%&=9UNc^QZ36f?s{X8;@H+g<+61m9iAZ3JfPtEqr`0OZO!?g{dpbDd zD1N2SrhGf{r})j!9mCI+^4ydVt-=(%mvAaK+{}K=Zg`Dm4kEM;XUiRq1DGzM`=WM%iT**W{IgcCE`{SN;o-cbohNdQODx z9F^$)bq<(7&9OAdmLqpdwNaS} zPtSa$1I*TlFnmqE!_^3xd&^cXwlE9LL7}|E_Sa7duOFg3PE?<8pF?zt zk^HyOx$a2eA7+%TercUj`pFs+A9F&K-2NQ&+R=Fv+oIIdInrR)R_-{B`KmLHAdLw+ z&OeR(5*U-d4z!l8Sl1rbQ9;?kM8H5x(n#9qHM9Ggu+KH&5Ut$folO6-CC&SRCrGzV z*BBH0oB9LYmC%pQ^Hf*%38qnE8@wYWsyK@3e0AYN6GjrbSuIXn%b&%Pqn7I>iZsWGkn&p?}ne_RQL7 z4qT24BKBso@WgCU%^TP%GLiMNBhBR*1&+nte@5+!3)VvY^pu=(Y6&(n)F-LH{QBby zxaT&Lb3?O>6YrcBVhcDAJ_1|HaxWL_9RMU4;P#1xDy3*G2znF8<~t5{N)GdynK{J- z2WQ^&x8GXbV>G!0tFT`-G=*~7M`&6=7Pnjh_rr3_8g>%s7R1K40t-@+B^*n0dp}y1 z=`Vdn%&zI2qlRnIWpch3tJ2HNeJW6<7oYYsFnebk@F*Z$3N{C$rk|!N(DudCCs3B; zPyiIP*fQ)6t*gMl3>W~g9DS80s{+f9^pEF)&-I-Xp)d6v+#(s%^(8tZdiG*JM-j-k z_vuHOk0$FmQouirw}iUy!nnE2Mn|a!m!1*wpY5d$?%R;>HbdIZ{HibxnO}{PWwV6^ z@EWV@3^Jy^YiE7emUa>;COp?mTC?P2LfSiJ(w>I9KR-B)L~~o+RTzN;6@7H1G0G+6ZGZ< zGoTVJpfW(O4eZoY=@KbK0PPAZ1m>xndiN1CzI6i;ZH~d-Mrxn?2px4l?HA^ly-j?r z@BA@juX}LpYhMJT^ptUO?2wt`Lam&yjyaKQ^7r%}{HPUjZ@iB00+n!$l9=sS1+>8IqzRkB0=1APA%J8+;{%AwqrHKG6&4u8r9SjRjcW1@;+- ztxKqn;mSz$iCYZS{ZB%eg?koXXgYCoF*H^7@q?07L8TOvg_@dw`D%wzS=*`Y`T`VF zVLF=AsaKSYaRcVdoz@4z-x9+5X|cKEQ(a6g6fz0k)AU=o)-qL>n-~{oCMwb#EjK5m z0m>?7{=LQNP?;(_x+S9Q>h1>G<8GNF4C`iD4vym!5~v1XUG+}E?}$W1l_9tu5BNfR zydTQFt%3=|{6fu)`Mg3HH24X$)V^2dCvBjODeA{&!QxOl^*A(gj`CRx^+iWWmh!BQ zL5Bl{QCfsPe!wGQ<+{Z$71Yw5uLq2tPGNdd3VLiC)MKBsV({haZjmiGDvfOotCS{$}xdjf+=ERp-Gl1ON?K5m>9Ru4XHb3GTd!^cPn#T1*Ym*$X zJapTEJxhw12s17y%)uyvzn2JcsS*DS%@d>#&W4*opz3=9|AZwaLDJ>Wf%B9gZ2&b3 ze*B;|f&P*3iABFYEcFPabWuFMZ9B4i3%L-ou2?)eX{cmM42LTM$;H^CeM?p6&4&?* z#jB5lHyB1rk>_dj)bdwG*!Qwknr6jyy&=0&^bH~$2RROQ(za2+PPoL$yF{-yI0Rql z+EZ)Od3RcwSe*B?!n}xb1ALh5_MLR7Ka|^|%w5A?*DwQV7{p<$sU=P?zI8$y(hDO> zCG5sCZJQPb-@<12topI>j84)z$LIIIM$y`;92|eyrU?C8)x79`97X>-)%?H2q9|)P z%qzovB5}y!LyGoOfm0D_Vh!t_K%1v05E3%#Vv6!>{-7FZqWvLcZ-fTc$!o*o2dX_z zaf*bxmMXo?KA%1FG!Jt^~_(woUuuljkMI15uG{6tC2iL8XPb1 z+XvpGIAa2gtoHP|`(4 zLK>hFA7Osi0cv_@3;Li!lX3^Q&D5nkVwX00XFdiK#$h-MABGEUMskD$T%NNQqtmI7 zQYCVBSQr``ZX4CXHr9m~&LVJ|qh=aoCF3~Fk2DHBf&CfUXs1Q|T#r^&aOmlsouX)? zcc$mphBE8!A*3b8t}gNWx{DKaIezofbj3UihtT{360-iACzb`_n9HA1=E}VkQ9`;* zKCWB(st8S;zd*pn@nk9K8zOq);6+y!fHMP>6snY4JT)oa)Rcs)0S|RCB_l0_yEEki z-sX9~%M4l+Gp)vU=dj8rxO_=gtpc4~by6zi3Ea6?bsf~Z-OXTB)S z+CFlqT(Q_v)Ldwm>hBa#QbuD{>>yqfaIIGsOL-mv_y7&Px~p^-_szr6vO{!-t&|Zq z_0|dw{=tkeDOZKT^%eYndA)qfeFp#xIk{#EeHnSxPFa9cMDlNKPm{P&1Ui0)m3z(db+Zk(zjp>qR786#ZYr(74hBe(RB3jD)Pa ziX;009D0N!Hn%R$vtG~g=|H92;)@wuB32h>SzyUHb|S9yoVGYtJba(RaB+)I}h}{xYda&!*f}+uP{cots1pd;J zGmqe}wg_^PbN*zWA!Y@3Vg5!#{aE{(UZy0UPh8o-{$No1Ko?K^T|l6&)kM4+@iIT} zez-`H3zageK=5Zq&yoq84)RU&fctx6a^p|e9u4o(e1?yZ?EARFpG;82bj61A2%?BU z)mA0iLT(x^)wc9$qq_G!)qYOJcayc^HN)*}hr-OMRZYckNYG5_K~K#KL~EdK}{l@}7rv z)Ll~6gsNxRfuU%cZt|9MTzGbC)nyZE;{50?-vwfRG`?$rolhK#LBtGa0@xIJ;zzt3 zL%1BfUvp~?(#=>T`i#!qj*nN}Ct6az_W7%`$C{!@8h#0k#9;m}!YX>k|AAGC8h<6q zeHKw^sVcX+e;541RUpXU59h@B!H84HL*ZPvtVxVU=p~#@;6D?)6!-mazY^s*>osa~ z8+GF|9gZ&8wq2~(v%9;zKx+b)i6uE%V*4w=f!w^1Bq#K?1Ywb(R@AA{UG;QB@iY+R zOQeDvtsi6X>MvE;1a#CyKe^c}Z+3c0T@BGF->UYp*nMZ7g6Q(2ghq5P#Yg?ErojcI zRoX9Ji=8K*EU=p`a%rsa0I&u7s{p+6$KFGkrRRW|E!6c*X;>be}9?Ws0`ptxwzsm1_93I_@`k3wH^uCGy0RkW|H;-s+1 zE5Ria(t7RrP~nH@W(+ToFW58|4SwJ5V`0Uo2nMIf(naSOxC=n2YO~Alh&R(`x~oJ? zMCkirT3q~%BL5|x9B)eadxky_U4uqP=@M5hFQ;mKuN|J9f zvbNr0ANmrNrS2l{Y#*yQauBU#`|_Bn9&kR?{hvlNC<81A1Jp3L0GsP8tx7O~&qqwC zcoyj>bfTBQHI?s7e$KedoTCHclNXhCjT8chqRDNkV<5zt6Ofqep^TS|XM}WcGCVtBe@#y; zjg^l-5tAAmK@2?x)P>SCrm2rt)8o~sAO8{vg}8| z+4$A-zuWzM>Dz!sb#in$%W=aBpBi29ftat&=$(JwWMBu~1AE=tAG`UR9lkDZj8fj0 z&W`1Og}wbNG0gZM!Oj0BhW{Z6O_bBJSl~tCVtI>Y(yDASKX<7FIhSW&Qz?{90^py6 zs;U>rtX>du#7&J5mVQ!YLqNPAi6UNZo<5`MQzv5jIH`(o6f8c@@_ScwZPQSZwvLw3O#XsMx%D2(QW;SR zp{;gl?lHW8k_aaoM!|vfizws|>tIwT)eKg*KL_7^ z<>QIcig_%VqBjHY<;#{K9v4@GCpDHgH#Zrnt0)-<$3yoNJ1iDQ(+YAOaaB)=yUcqk zYAW?C)m#~@zqa<4j6%qA%)wGSeQ&HFHpbL%3}@!=t0M@(pKV>)?X9TCQMGv9oD7E+ zZQ1KgfL93V&-iw`K7SDq8UCyfBK4VX&E)1_49C6!UNz9T$nn$bkhq6P#&(As|I@h` zcw(1cWh(e3^%<)^Q|p++4|dq}{avK{=yxZXNs-H@lIHyr7Nu@1psIaJ%YPOFlj`P5 zL-Q2#Y1pTb!>;=jl#udP`_ra)t5q6$h@D;km65>u zg?j(bs^R}UE%+yqumd6eKSrD9e{FXE(Gso0qlN%$Tu@ms1rppx1R#AMM21v__A~XO z9A&Nu*lbxPZ8C2(-@f*~voK+of#q$o`qwI;%gZ@CuipdM^{=*efgpf@=$ar9ECfwlr)hl7OPTYPR6tL|N z5nG1wX*_1rZc{ z3vsD-)(ujE=J93GtIiQ`E}9YAdg7@27d+|+{M$Cmp^HXJ&+7Kn%Vg@I-Kc4~reb2@ zWjUl2nxBdrB!!Rd6jV0QLy4&!Bz>Lk!9sCi(4#Q;7bR!G z{mT_{CTfG66pU$VoSEA@<)gw5=|sX|T`bH{tU-UUtxYeAl136*360C_c@qU;aWwOQ5peF3B-v$Dv2VZV>*tn>7)p$1s!|1Q-QeqjI zNGup0spcz?ey?Vox;4)@;TjuVws_CPOlqoaB&Qc}Zu`S9WYy7IhFEW+)geG&0xW8! zbM(vjK-2&zHPVAZ5oIj_n-E)aw_lRf7Qb@xt@s1^i1=+#iXVJmo7CeSeL0z|EF+Kvis*GiOFm>%`1pvHf(Vjzv!cE}7o&?~M{slb0$kQdRmi66yH~n{CnM$!ybi-$=-Bmg~;GL9!Qly92?j0{n{5 zhe`tSObIRka=+pO!r}z|Rd9dgO&UMW^w*MkGkZ+RxOV4p@zb=^E&HvXgf^JtXSSP? z{tfnVA{_4Wd%H3g&cZH}O990cV#C~2nM+dMD9lrYw2?}tF0c0-ORjsM_bl-)ep$OeqsQtzz+5}R^VrM%p0oO`;^-gs7Ut!%U3xmh!Zj+9DasoGSxbq*dR(2 z&U_kw+UR**>6+~(J)@WtxM2*TMM|M{quUJb7O}=vq%otl8x!iCw#6%=z)aQX|(OL?DK>pxr7 zuIf?LD5!3~%Va&^-sVmASe8ORD#NDwMe? zhX<%7tS*Yt9(|~`%mA82={--Nw@o@rBd-w;+YQjHWLegdFEFe{*OBw3EO~<(<%F%c zyOIdUQ?k|iV!B-O^_Wl--rZlA`zQ4m*Xho)R9qv4c zmH5=nCsQOF?TAh}U+PBgw#)*UpAwXH`^)8~PV+9+VbeJP<23hS0$HJXDu+`FSL=OI z8R9qXn{W+H+~VwK)3B_+Nlz8DT7$UocW2WUYrrm6d=PFjW?R)AGGT%cH~mGU^eWCY z46W|WcV60|rX#xMNDNI>C2R_$Srdb?g9<-cp(wmSu4pK(ZMNGHw-FQsv(OOuzcw1LDyLpl>Iu^ zAU92XjU8_yB|7%Z4`3q>j_F6*j^J47lT2h&_Mjp=G!=YP@hAxqSzlUM5lMJ{E?~Pq zU<1*rtO#~Y{G7q$&7_EUMWL~r0kuSECVjPZ1pXFrcxChh81y=|;$Nmh|cww*}d4ebeEjz#~ zQv^X1$`6C|(XjU({d-sVqX<1(_ysceA^*86{OerfzXY29(9kJrIUorkab-qdUa+P> z=irtQ!a3EeG5i8M88QtCrL;yc1VW0A-*jd$oU3QzfZU~dKJ&R@yMd#ul-DWYFIzx) z=SV8%FjyH=Pd98XYCnGLJkIFyy6$*BF4p;`DY)26%L#gHhc~_)J#iL?#_ep2rb1XH z?zyK6md#G8x7K_9EU7;u8A>>3vs|ntqZf;Xi<~pbY zZN{b^YVPei91V1zg$^3mm_xLa)Tn%`dD5&NR;`ez-muG*R{~F7JsZ}5aw0gm2HUyF zyWE*5BE_Wfl-|%dM3W4LNYPDgOSBDG5pQYS-Q0Oa_EOYLE7^F$dDcET4{9ce`R1oS z)W(dQ=h`oG#9~pF3LIOcqmu(Nh4o8lN0jN-;ym1&fvc=Od_Jy)nD&_#24SuG5y@+r zb+FKSl6d(;j$>6s8eOIh)O)3A`G#kh$8j%7<+@y5p1tFhai(Ni*+1wp-dR!!rA?Oo z%C$}74yrWMd?Cuh9FKMt>$D=2Y=0U(W4V#m9HgzbB(AZ9n(i1s7(SFnYr=qUj{;>` z&e;x#fVxXqNPl(^AqiyaLIG0pO?&XDXHkUxmc^wLYEvrIMzbSWm^%(c~b^{t14Y+>v8qYQvp;6o{uq3pLl=SQMYN8 zf@N@lZNw<2kIHjEffT4EIqCE88*ROpW2!9Yr8t|7`@CrQ?!TCu5UqRVWlp%Dsx0lD zoHh7@;TO%0i6g% zhg2;3gEQEFyYv#o<S^0)E-%R(sme+@(awPOEU*(>e8P-cPFYRU3tu0ni*m(VDHYTPieu%;X& zqSvN|__n%no@dZ^pw}>N-MrA-$pKJArq>*w!4i`u49et$Ib)NlEUx71jQdH? z*ZW7VUkW3oru|}x1@ly>J!oW^iQ}!bQ{^XpARs780d%CvcmO+1dHCKMO=^mkJlz01 zMni~sEMu@C#3$4=;maQ;rsFAoGRnwDEer+|YG2MO7ae+=444XCZ)3zyxoXBU!~r`~ zB}NJ38QQ-kx3p)%o3t8Qv?JReJ!F2v%3}s&fCSCo zLT4Ezb5otLh-DFDTtN}jM@I+=plmE&OPh?a=|Th%K)ceH_|deoEz=3fDIb8;LCGzY zv)7qUwhDW5TO@Qq05pk;YYRDPT8pU0AcWD zPMVxv1&~Y+OO0kcsv~5?GzqrR#Wr8iL*tKc20QchrUa3$0U|6~qA(zc88-%~pwSvE zP6U%u2*{HP^$-*U`_~fY?wf)*M$B1*%%DALWdvZO;BKsD<9B&8+0pXO)dp(j6;hkH z=^Dqdg8DU>Gx1hg8(zPc-xq%-ux6W_M3S6MdJbhB`w1^}CPIT`bfk~cJ}@FsQ%0{7 zLWk(7aMrHcLEa7n`t#J;_tN36df7X#i9=ECE^7)ay_A%Xs3^E+06v8+kf#9a5WWFX z7d6)(Z(uet$bEoFIF?cHz4K827c|&G9v?svtS!Wd(1RjAX>{qkPhX^88 zZWNIx<3|Iip-=n*zE!Jqw}AKK8Y`FRtKHA{sq#_9SD-!$(@GW&Cof8Mg)A?0uPtxC z5ZvzH9{#b|g51^4??8VQv&z}ao8?zAhy7nI=6?`9|Bw}Y`4IPG!Gk2EtB(1YKSpWVE6wsuC^nZvzk+qzDE<5=lMxc&J_1 zX>M_>8H5*e z;iI;D(hP>yMX!XV0)moVO1felNQL$@NVVPxXUi#^x5Vz*6tY~jZXuvXFx6YtVw)_N zkyeGStk%##rWqCUg?k;K6;hOG^LIgvnI>|D`z0nUMD_XlDa@R33vK0cQYAHHsI#wN1%?BQCuBXGCn&N6s2B*Tb=2LW`e0@_68AmRb2+vY ztgp%+nNb+C9>ev)Y)_@g2WQ;STEu&wItn|CW_JqR=8OzFTv;HsA)<|1SvtE7FJ`O> z_l^lZ-=aS=}A&%utpfB;)!3(3yc6`Q$Nkaf(>yI3nsF(*=8ZV;7aHsAQdUm*AKx+4+uRtzp6JkFD`j#=VS8k zr4vM4aJk5tz!L^~)X!kvfV2%FDy`t0U4H_c!B-_>guK!qkG^?(Vsdp!=PdgyiPX(b z=lr4Xc6tAsJ0I9B9VXRRHERFcYWy$3@IQu8O50z>!#6S_lqP!n`BOO?w0U+U1LoB*|Nb;j$P5 zPo?RA_x;bIa@9D-(bVUQcE$%Qp|2{?Fp$;O=A4IWmY|0duDj?Cx)>p$*8Jc$QX3=q zedo_@*$r$?+197jG^YWUx#@j}QM8m^{YYwFmr_|WCXt-05Fhzues(RfMySy$Ajq0) zYjY@KKH}UPV$el8Xw}f0^n%jawN~L6esDjEbWb>hSp~++EDV{GBRX{a*_-2c-Ig1K z_`d(nHCw0Cs*_omJZg2cmuHHL)q+|R`dVz@{&w5RT!DW(Xs%1vQoZxNe?^R=zV0^6 zWJ57OOIc+XbPPXDWio?yBTry7QZZK%bI&kxqBulDt1(bOZDeW!enTK z5sGB2#@>|#N4c)L07V{zG2_Hut{)rpy=%cPT^>4eP023!93yR>sTz7xK`zQT`qcgp zImN;+bN(+I7LrtvpeXjnIW{-SqDqp!RR{;C#>Yn&R5`q&~+W zgM3|odtZiQ+g$=J)TZ%87GrHb{Bc3QTNZ?!#4vRhofo4wlkx;vCn5U9tlpn>`;}O0 zZUow}k10lc9Mxf1!$;41*Cz|^O9ST3&tQhvBw{ho67Xb7bF$lrCwR_L;woe2iTfIj z{n%>X5lz(?-Xe&JrA-)~7oPx!Mm;a!aEWLW4;2>yP2__&S9HNImIOzZ`||RMSA3>- z9rwAqe?}dQX+JnJx4(#-54k9^71xO9D+Kc4Gs;$y6Q~%-P5fO(n@gO8UCOKFsMw+Z zT7Z==u-wC`4fS!A;{W079oREr(`C_4y5ppyH&(~CZQHhOqr*40ZKGq`w%Jj~ws*c+ zGwZCG>zwcGz5hc!_1vl|`&PQ&yAcu!sqO}3Psqg*)5n=1&%CIW_ESF*pk%+}ADw9z z-&LUTXM_IhU!1Qz&ZhW$iPHb}6902p_5Z&V`%jbUe@>@VG+mLE0q-QqGU?ytXmZiW z9E#z{K@v4J!V2q&V_2q>fCOm8M=>(#QA0}-@kFboTg^u`h0x#HMQsVbw-z`$Ml6DJ(&)puOa%%7IejtF&OU8 zk-ua0#TgxOoV_Lk&4qShm(Hn229g4$n+TR)8$pS~mpsKF74k~g^RdWyePJ6|@bv=BzIMKM=7y36 z^OQV!l>efI`LvZG^Hat}8qFNTh7H(P>+pGqW+xT6+LF9`dazC%&$xZ6@bn~lrC}NL z$r7H}?a8p_%vnamRzv8{`lC;)q`WpqdF(lZZ<13&BvO_Pv%751r?1cR3DSvUReV(j z#{zV;1QU|-uDr(0?9DQ-5T6ENiSVjB`{>UF{vF}k2ijq^vH1nTrF~^!@pnx0dDF* z|KFe;?9E9L%^nsea#P~V%6rPwmN^#=qr)vl*0HqHOTBV4|!wa|x}p zaG4fTeLaIQBGk`s?uG*eXZu?kuB}G2Pt#4akJI4dM^%}dYM-Z@ZXd*>T%YKp+5jmg z&j7Rj&_1IVjghVyn}14m+hF>u6RYJ74cti$e@|3#onm#-s6Ck)M;GDB7!17Gcf`4< zex&^*CAwQ)(p*EW;pq;**q`;weykzOM8*gVAC1-3@`lqAZo`n4#&>ZJCpISr++Vkd zCBLf{chGStj2Ntzp$M%txCh26WsaWk+4bFNWgHhs*-QMbQO$woLoqub3;9|Gs0?T31F@XvZsuiGDuJK=a{&mdzr zCo@kxK9CKGEp+!aQA!W952=Dsu2fU9!<&l9kNF!DbQf6#tX>jstC|;b zC45f$7i@xW5Tyr&Q*h%3cy(z3wTG3W{%O>H8ueU4V>%A}k67<~9btXc@G~VPhl)N2 zTUK3EQM!yg*i(f*k3mu2CTA|Ak+zjbG_?Z6ZCxL6RW7a?z`a?gkw|F>9R+qn%B3i;8UpwaLyC1W8t7bh8`xrw+sJWvL%5EjENt^4!a^N~ zQ|RQF<@>aO>ux4EmLyi|h9y$E2H=(&bbY-a38KciSM~xuDB8s)iN!WFi3G+2jET_) z1WNh1l(zFxz2x2`r#tF%;sjjmu$u~pgqF3$zeu!;Bj8jHsgG{E)o9Sr(kn2YVaNT7 zQYwP$Ww{cM(W*yT$Gs%z4$P&I!4G#pD@_crh7EX31aV=+h+(~Z2Vf-^F2t010%}&W zr&~)c+RFDE?p4Qq3g(o0&4`AVrY}kKsiO$s)eXyi!~1o(M$tWKVjIL7MLx#%LhrQk z2g^enA}mKzqfUm2*-IRtu--xcf+YM!XgTxe0gv{ly7T{8to`Rej`i;rYf|Pmmc~YZ z$vQusHb33noE+R0jh+7chDXWVc9jpoTM~9~zJLireCjrbTr^xT9v2co9)?6tk&DOY z#aiu@uO?lV{Qpu|G#cP=L4>s=U0B7;Z>+o7t5wKQEHM=S$Sl%erX^lgD|geoNK9 zUEj5~{xZSuEIOQg{czXzt$Q|JmMKTWAWZTlYki0|Sk9-Gm7wndwHNp7ea`-ue|}et zbu_Lxn;DmG-%Ax6dNgK4qoTCH1JB&G_bC7tE*e)mUU?Lc!F&UXKsPZxwJ|iz;y<2q z=-(O*r4q!1ELW1~0{leVBUeg1UZg8GO3#oyj3ex%^f7GPirY!!8#z$+5rb^NtA@G> z&n-`LxfF}vw{ncsN|QJ*Jzl4ANuF!P)E$nvz6(v+duyr+egpmC_w$P_X@_?5Vg$VH32h;~M>cYJOIs<{X`=TYv3m^qy*3~d|i(X8o(sGWBb z2n>x`iHwO^N(q(Pi(0LK9r|)yh^!P)drpj)8Tj0ZFkvp=d7EW_q9^+BQN_mw@h^gM zn^XB82UoBdIDHKtm_@5LIAS!|s=g%eH$Ke#fo-6?{)3ZK$N*}L&rDp{;z#ZN;{jGY zn5fy2ezeK8X8R+`=n!RZgPfaKRP-?A&&u@zQlMrkcU%WI^lPovYnc^uTm=JCXSnBf zGyh6lIVSFJOm$%ai(x{P^=-ngj0s!J;(~chxeVkUjWh(*JMB zlKvfJh20FToQ?ilfvg7SrEHAxk^NFrx8bR0SRDxmK0Ji}qDK(YN^KB@-$$PS7qcyz zd`4HbzQK9XA=s@zDap`9nV6W6W)Kt#B}L>{UtM2ce+?r=URg2zO2*9d%zKdGd5&Rf z`5L(8J>|Li;GRA8al>`7qBw-R0rQ(_0I$2)E;{VOg+!}Wk)X^w;y3pWWnfAWXHy61C|4IM=m46 zOan@~e|>EReKESnLYW#FyK%T}VtC1f?MmJT!@!M?L^F0H521Hs4>4$``t=Zk@vDTb zWG5y}6~!!arUds_bRZJcr=MBdRmPtINlY5`#p;4Wdwa)lP~$9r{ks}70*o+{ zR#0c1zSq`5=}Kr4=7yoxG7iI<$q=1|)DYqF-kAd2XdrICkj%@9h0T5JXwx3>@%5gg}X zs#fBdMgvy~qkKoy7h6_jM<5=nPQ5*8Mz(@{&@(5yT;)t|eV8WZxFN?F%mvMP%P(v) z56s`*!v@=&H6H^!9vk8$By0epjwiUa%WvH*&6EQvTcm65<&4xrtR8KspW`fh_qJ`L z`}4dq6*$fc)=Od<`|F93!dz4q#;*RT)fN`FaiYo|;}k5k2`eJOTZ#i$muR4!H5meu zgX1le3U}lbI4<^#q6)AK?Ke15N;IXBH1L*?xnqe6A(w=YQe>#LSKJ&T7!}3jeygvP z2f3y&sLlog^lrk^l19_hgoa%1V*zC=U79)mnCDiuW8nZ)?j2L&vK!$&nnBNm7>v z@~e+$&ccL0->Ut06oPTdI|p^7*O*kg`G)bn_NmAd&&zY6E&3u%Z##4Fr$BeB%S zwbzjFa)%x9m@?=+x{(J}Iu-pyLa#KL7OcMq%mR^*6#Bvd zGAV1&z+hx6=k4!pIxMX-+qXWFY^zmmxjRJan8q*k+A#>nehOm~ar4{`c4QXuZTfM! zZsFv4dq(r9nIZ+r*6E7vNomlqJ(Sj4WQ~lf9*E;Eu1LB|jiMFcm}a4O{cEX6aliW% zP+VH=rtNr=5V{2y4dHI=8~q7b-4r{-``FcTauH#au@^5zU&RHHQO((8f&0t7-#Akz zC9sP|j@{Hl$1DW7aKpnJZ6SZbFlB@Xon{_cifVJ=uKsX^$M;a|%f3QG`MGe3`d+k6 z*(}p1-R#GYJ3BkV7}&R^{3_VD_IU7R(jH8$kMnT}AyDL&$qP<*>5^tsk6#6wD9NrEg93D*khXg>&qE({=NXS;Ujb3G#R}?UV zaCS6nr`5kT$`7JW5eAE^--*$I*%`#iJW&3~G%on{B9gS%=}Q}qvX@T@H%7Rv7#K?= zppO017CuX1b>%@uW%O2n5Cb?I1xDMUh8GTLF@ZfY+twP7RS{5{;jn4{qAWogDbBwi1BRvj`k1 zCXqk%;P=ggfrRnu$e>prdl`Kp8#POTP!n(N+~47t(Wl>NyJDkIH#4c)NZUGm4D!*O zaUiHu{!~knAYFF|VNXR5wT>4m?beqpe~;))Z5;?g4tIK*uGxG$8MhUDHo`Tzq`Xml zo#-i!4JDZ0wjwJz7;1vh0Z80qII39hj(iGV)w3cGS{CC`(hkxhpA6n7Pab%}99bZv zsI`uMn38Zyj6UZiMEbjuz8Vu&o0HBnY{?iUar;+fLZ(Fj76gx0`Ml7N;RSRDuR!j= z`fO&0f+`teTT~}-rj4MC!Df0ZY*B}nlg;#*(hys`Y_DwAL-eU8Jm7m~+lv&Q^@$NG z0B@lBupDEBe?lI3WlipUpqMcpI#Od$K8^X-^6i_0q3sYzZX#V@hKd-s#ms53CxWZk zG@a2Aj|XRO!cJ!s==;|cse2i_)F49fI=2N{>UiO1^AGQQ{Mb_FF@^pRMsdo~*wkuq zZJgp<%RZR8*@(;=<@&f?!9byO;hPw1PW!g>m{>WqRf53W{@s;DzMr6MMhXQrKVu35 zHnBuW1es)dro1FiZn8cxU3_2-Zodd?)E!n7(Bp_r7@OOhpVx=2zPQ2?CSJbQqsI4{ zFwHK#uf$XLDs(_ad|^Cq(Q@`)@2EhtZOWnoKbXjdxxLdGXG%-76b}5o8p@2i^!3&+ zdrD+m;|!|}V;9ETGo~%i3Ify{T|5Q#nT>iYQXZFz*5w=`@-62SqVi2Zcl8Z9l#bs7 zTpL4A1~4N`4A2(YUT98u3_-V;WI(CT+zO#{MKNqlS~;LML7Z&gQ?^ClW(17`n?g*W zL)aDH3?dUZkB<|V1mikTJr&S#R5!-EHhsDF>TuF7`~?q|p8MR`>l@c{=uBg>_vDE< zJa!nQyV41fubQxLU)r7rAG8)(fxI?|0oq#nWb9-ITKr=qLwhKqHKSzHY`LLZskiaO ztKUt4-Xs}c#`x`GaPjcj}*7Z8@z)?IT|V+7PS%)c0d0XpeNwC4-!G_QA_E%3jpgtuR*~)KsL6S zHN7V&j({b()2|)r^GY>nzqI0`Yf5B}bF=ufg2b-b2BBCOlc@62%p}d<&^6tKWeojG z9@$8xN{^SKj1ig<+Ph8u$gjud|DoLYEj1QdifjO8BWA#CHD;wm@z-IL|XXRzt2S z?TKnGNp{d(*w6ITDYOU&_hs-teJ{i%bUn85wsK!D80JR?%ETqerUVa1;~JDbBv>pS z;u}bk@|2h-`h-}G9G!uNAv$1wSSbaI^z(2DVtapjDj%g6xYe=tOlDTtxs#~V#1A-1 zN?B1d6x?7&=~LdQ5jTGcobDsHxQ=6TmxSUK=`8z%ERVHy53fWvRjax2#K??FCW-ha zBGNH|lRK@`9=Ieh)oG7XE3iC?P9w`FHPObR1_qelj#LLziqDQwS&Dj=@_Od@jB`ik z-q6_3M%|qLrO>ld#y+S17FXzlq&<9@H&qk!K?M8C>7wI$hqj_fhO zKRbVz>9WpzXQ)6%)2j|Hue@TAk@!K02B8;nRTB~wRZ&~fzK3U$cFR$BNT?W3}y0v+e#ycZ9Jb zdtcdj8biz(Tx)t@q;u5xIm}8Dw2Go;86svW?poe{>%Zn)wpd6;XH{VUXig!duf$I} zK|gQ=cAX9TdTF}xbo#ADym_rp`X7oz;RxsYPwGgjMI?z0U~2pL@@z)d?ZL|`2-|Fd zcDbh0QqydG*>5ZGi5_|2@w)?A;;A2ic=|}U3fl>qX(Y&pDqo{YcUMOxy$W|j*j^yt zc3o%Bxg;#$k)R(Udu6+RO60D; zji19LroWA|{M+tb(aFJB-`bIg-_Fk1#z^15>fg+n3Kh>}K2>PkbsH|`>O55b2H|=B zNqJO&E_$K5o&xcFUWAyOL1T>3V+KvvrenuP5)veT5*}Z8<90^PRf8XakA@fPEw1P6 zrstV&ZwEuPUr@M4@q@{~?qf4a8OILPGd0!LcWi`XsLhUJmY}DYqI<2H+12b@J$q5# z8sIXdDC4`Cy~qHrLKX%0Pc=P9+ z1YVa!;HIJ5`Bj2jP+8Sas zhFT&E589nua9Y#tz<%U1jF8dq@8o2q5`X0H8n+@2#mW-&p7@E{I!mxi|KO!-HGBCzd(ZY^QLm8x8I)~Qg8@Mr?m+|`4QOI89k4Cdwq-h={DM3O)))E6|0S3t*X=-h4^XBH_+@G z?t4vy(eX862Wg*Z03t78A?Vum@LPH=GxeK+{pkh>ERH2kp3m&pb`khZglZDpjY1|q5%3fVjH%ry2A1e z+Fw&D4=sJNfX|kn?cZAd|J6 zq8uoH=U4rPvw63}myq8y@l`cbmZrgM3JznOnq0Vg`-%?WRMCk2(S82pGllFzds}!m zD)ueT^FE#J?4Zy8Ek`Zi%Czv&PUU&@DZy~m!(AbATt5XbfC2D6bu+i!r3I7!TclXg zw#>)I$Eja`@0np<-dL<@BWq8d_eZft{*4uvha=SxzG|wbu}oy0^xEhWHu2hjp{W#*D&QHpC28Qk)jhcTfDjFBoMi0cea zIuRK%O<}f4DN95DShThPoU#~57y=X!J(r7r{Z<3j<~B2}fK-M3croJK!BjQY2F1wj z6J-31q<9-iQmejT@9SdG+L~+nJ((Xgwn65A=nSLW0oFu0I<~qK|A`pf#5@!Khggfd z#>DdEk6^AdSXH?nrowZ?83Z~NUcn|fh;xWYtFl@i=@s=|27eekBnQ1jBFg#myWj`Z z(VsKzC{8H2VV7ifOVf#KtKutXe7>^op4uSpo-Jz(WVJZj;_v2D68bLsD#i|upAm`wwe*ctUQ<|80ch76k3oVA@*xGZ5c{FG z4G`m_jlLlvAtA@b#lc7`;wuqaNaCs+#XY`S_;;Q)H{F^(ryV6x4JG}c77wva z%Q`}(TGQgO$Cdg%BbepNA`L~Y{=B|IZk2NG?M)~vNp*tm;Un!zUGobQtI3kFzt$jq z6DxT-iLoPI4`(F46I93D<#49{_-K%I%;O57tfY`w9hAnIWZ(%7YM;)u|!WHq=;SJNeJHK^{*{)DygM zOd0%m#|*8_@jZ#C%o;236%QW8Ia+@6-eFbblJdzY%argWGs*fvYc&8dKk}SiLo77U zTy@wBrpHY2b_V+Fp&`ggpKP|-Beo?pu9+K|RmaB+DOB-?UlULgec=+)9xl43>NgCW z>}4Ua_)#Jt3J$K=PNNT_pE%=#H$(gk-O|`_jj8=X3gz)ZnpPrfWX-%_@4QuJ#;F6% z%AUJ)K~cIyX?y7Y|A zCW{o15u#Kv>z`QX)#R?r;8A0?s^sba@>5=`-|>6fOjd;;9HAE`$K$!!Az-OnInWm3 zqPOXbVFvMJB)vJF-&(2nx3P8b=>hqDSSP{=T%zXLLrjs7BwSuZXJu)_ImO4pRKBc3 z=ysvN@|Z4Y(Y}3}Sf)eLGb3oZNc;@a?k^g4e#|85nxhcx%Hm33P*)Y%rbv?x#Xl}d zHHd=z4x*}EF%b5+V#0zRa*XDB!e2t9NcNILind011*FG%6=V|X{A#m1l2edbn_cd z$K(>w?<*e;*?VUxvk2b@zD(Rkuz0}>`+&8GgeRs1^SD8RSc~j{=w?~C6_xGE9m5&4 z;R#jZm8$mZkRax6mW%duEKnqR*S`fWZoosI?YT|Jvfs1*#ZL@f0{&v3+%))a=PCcW z@}d7be){*7kBqI6@qf`&hw7RmvN7PpdxRvt2n)JHt|V93%rLF~8={bTDIXYhM@}Pn zh}%k$WGw=VbDjQg}PNvEcyt1cY_ zjFyN9)7}t0;7BM!*8UAU$yD7EjFb|FItC1AkMP1gtR!C=IgJ_UXrR8?A2)vL#QM1I zOQl9Pl0-^WM={Djma*B{ z0x@gq4RVWW67hg>OBnz?e{YG+;UYSBID8TLPP4W>mfwuYsOcYON104oGwhBOWf=2? zd3Be>xWuJW{1eOUOx@R$C(5fLjfPg=KkMoAe++ajp|e=wI4(vnsxr_`>5A3kxppQW z_44r{b^`8G1dWniNG2$eSQqZI}Zia-(^P zOHv&0m`UP#ZP<0Sg=h*|NtH!>d1MHw4fChwv#sGsqveI|bsqJ>^-1{UY?i65vE{U@ zGY&JP&UHeP8F@n!8=@oC$Y8bZ9ehpRu8trUG=-y(QPDEz>Kk!RUpLeO;B2Q0$^hFd z1}ozh!A=L?Rq|H`v!Hj-mF#F6$(AEp8>K1Hyq|!eJ3dOOwobc|7aVu%&xQ zR(qn^&qOZ11aJu^Dcv-7ZgB)xP$YXk(;a9b0jj=aF(QVsOIksX%v&F#(8hR_e$SYG z<@ylILxc;%klV1`aD4$#tMVB-fysot(~KjIOb`xam!dre&`^9CQfk5!^dhIw$o^+m z38$FSJ~k6v@tjX{LuO z%vIoK&ebmFJxy+aa#Xwj=M_99=tW9>t$4H{0cWi>fuv;2E7}_xd8^UVqF_hBBW~>@ zNZ2D{+kh;a!~y+PoocZ^d4$1umd{{0TY%_k7$weRv-5fW1;;S!3@PFK;^v<%Qkxiw z%e5KmD`V(E8v0_Ym99X#I?72%3d~d{TIJsy`Hx^{JF2kL>KJZT3*Cmwb{sz%aFK!s zaCA^oH~D#b!i8-0(BuZe=3h_=hJNWqel4o>K7(Flr(0E-*6FhEQU*EH2FqCj`(cL= zye{pTNR(ymYCz$RfiowliiqK`T*%o)O1l8}2OQ)r0O%Mm;s%3gkEP8NGz#j{Chn~? zYj{tbe8l|7X1@TldQl#3O!K)L& zNEb%K`xejtSE8QiJL$5|r=q=b-w_k5=fXCMAsX%$YTNu2uJIM>FZO1n5W>~{Y=Z#* z7RmOX)#cwF7dqNo(JIdA>L!xa8;{14zB2|7QdrM zaBm^PN?Mopr@Kuf^Cv%W)h4X^WP3tz#c0Szia3!8f}bQR5Qy=8D_uM5KfI4UYrZbd1ZG9A-^vlkAHC~M)U=C>?eYuT56Td5qZGf%J0lr zel!b26UXG*0~|zXiMu^h33W-%Xu{ONc@u{c`ibCNHY{ros&I?M!gwVV}KtR z6)_|`phoH)X+#C7m^xXU;kI6J&$Qr(>x7fx;&^B{tNT&F!vu6ItGmf(ThpioTEz-+ zwE08_V4}*_?or`2D{nx-rP!i4Jsagp@HJZG7-QXKC z$U+3(p8}9N1s+c6bJD=9ZPaoqyCdlKY&}9CNqc?48+9|)5UYVKS~(rh_&O@6M=js;x8V$P_H*7fAZG{9ij*-Gzt=DZN2GHJ0w1lnE@(IT8#8vv% z$2W2};L_1bDG+|)u_vAkn(oQT<+RXz6fwrlY9Q8e?~PdH$rbhuE<5ow$AW|Zw_C4M zUN@_@ZrJoe76y)o$Y|G&lKe=CmvwiWp2zfELnqn-(k)|q6 zo^<~tYO`2)l8RLy+Kb8GKw1`AZKlpq|BI@{yL!M0o zQqobvF;0o1OH|MFPkKvm+~R6~Qdr8&48?)w(dyL|3{&)1&JoynRU-5FH3rCNbdoA* zcH>#ZYwC{zE?B`9FG>EpvjQVinsQUa{U>j=@doPla|LY&UCpl?-iX(n=%Vo4#sx8( z4K&ZCn@q=qV+C3ATilM&_tLTtis{pnAkR61|g4Mh7y}iBaCd zCM-C-{15l=M4MA}is(b9DSvh`87mFvh>Sw28PEr+2ea`%eu%dT@1_V?GLWCJoSWTK ziI8=|xQ^f!bs}b44@T=ncs&e9`9GXl2a4#pGiN*e_P3ix`g-(>z%QW zM*g2zKmFTz6UYB0*8i(k&iqT@#0U7)R6zAZq5y0mmCFS9K|_A{L;5RM55v-tMN5v% zo24Z%DL)E#_Lk5QGJcokc}&M1s!Nxm@<-7JZp3eJZ|$EJZvFImg4RQzWn$8o5=7dC z18@Uo0>eZpNSNpKB?A114D2+hG7u+_QPkl)VlI^i7*XnApDT7#HtQ$GjO)g8Op1$= z(NI+}W?R}d|B9|kP{wwR`=rESHE2hO@-af(f5Z(;6%q>ck}~ zV>Ed(tjA}xi69TZF+TUSG!uae%$+l0)T)#x1EQLDJ4CJo9qLRIYS4N6k(5ij63s9# zb@&6g+r&NE4UM7z^7B*{Up?6O4uwk`&KVk1&Gu04=#Lx5aSc+lra~^JRk2k|tqX9Q zf!|hVe81Pvs9IL1iL8NOfF0rhPUay9(Fi*M!bDzrqSp4_5rKtUvTOaovjdp1v zUr*kS=#$*4ztwa<#2OPOf}sLaO}Yqf@tVE5n#tI?3h_Ih_X3(Y(-p*HeuQh0-ZZoD zxXHopWXHj}=lS>po^B^dbNB+4R6=b^UwbGRa=1Erv++)2C^51zP--JyXI)P$nbI2E zkoyoCcvVC?Q1|X58pv|-XhhERHTdQ8xYpvQ71It5t-&yNqC(&aPc#Alnx{1Ba?|<` zJ8%6K`5)A>g9HidO?h9pEN~vNxKQu)Wx0;8aWko^7RV4u0Y)Zn>N5Sjxw%Qzz@0?e zjAaEemuga@%D7AR0Re1aEj{2vrkf5kn~$FA^=r2OvoxwRLkx+!5J9`mS_8O#M1#Tk zsr6OXcV_uqA&o3b_jP!vBkK*H>Gl0zE&Gl~8dIEYfeMzdO%h*Ma}_d;lB~hrrV-P# z%FIYsnw1k#H8A9-))$Vvp-b-Tq;tWL$#mw5zAOVg+f)vS&kQ=@ZeJ$WXhW$fkkX| zD9@;sgO-(--CwjJtf9`tm?l~#s+WChA2k@kLZcrsxZVxoE;T()7YgcL`rB1(4g#Rc zL#Qm)CHhVoVeJwa3s1K!%0bVhq2s9>>VlDMrczSKzeiuDA+4{$h%!jr*ZS?lGzb4On37dE}xiytXEYPv%BQh<(Vp!KGF0Xb6ujjAYDf(R~o z<)fMi>?CESaEdP^Qd8PS&cQSPZ_4dw`5f(1LqvTYvHpYrc{}iXu#)4n&T#!(PlZUXbK-N0V`PfxQtQ{P@5;yG2^<8g< z?@{1eXCxa}ULgW2cF4Ggi8XRnP>bZ-R(;5LuC70qjQRgJ|;BR%|5(b8cMr za{bQ0)%`)s{K3oHu4Cinb}1QP&`p3|wKrmV9{TM5C`e?3m!8i<=Fi+96c2fpaf!L2 zML!PNCKB!@KxS&>F~)LoKaZ_ZZTisBw5$7}U%8qT0>iLj=qXq$XK40}R(`s1E4P4; zNL;96C;nZdN7yY!#S~F7$)zjFW#-o7cXJ!xeXKwi$H2zB-~;k1sa;t2s1{gzs&$F> z^WQcOn|*IWFTR~H^)Llvdd*f9yN!u&|B#KAM5zejeF8B4ZvptfLybzGW&QtKMXzXW zJ4c7$vu*-MxdF3r=nHb9<7<*2{52H5zKjo9Pz*GNmjX7+72AVi!vLK_~CRl+TSc?^STP{ogO%odf8X* zW}nQMZt9~9Tle#Nu$rOh!uouWg3Cw^+c8A{NhvCnL-vyyHLdki%GC&2Q$F<{4?fS( zvL9qBSjKyYii{cgcv-pPMsw#}<}8&Fi=lUf)v}>&d<@Cx)Kalju<&}Ab@dEde6cb| zzaiBz`AVNLQVs68i@JK<2BPS$K)_$ba1FmzrsX6CBMzBqr1yF4&~6h&nHP5 zwMeywj=u_iW{;-;+!OVxh*zDdh(lrH7@3?+TfSlP{^PS&CuDi2{zT;b-y-sFh0OnI zH2=qhqPigl&`9Kv7}!ov|c^>$gPi69b6W_2pEshi)BM zY?;Qv1qD3iXnt8At`;@wtj|;Kw-0vXGWI9edH66ce^c#+7IaE|DS9unVOXcaw2`iC zwOL?Uu;_DE8oQF7+HPz$W?wt5eF|i?{-_ghNnbj0yJpNZ2{Bh$*!Gztw8ZK%rFP7H zjOg04m}|09rkj)8(-YGr-)H95q_`;Ne}ZtsPoG<2p@5Cwa(nOP_rvNu=^)Syt8Vzk z+-&O+jW(hdwNFbQ7&ixx(aY|d&TvAj5!JFJq_bRh--|lKTyN5G?!hyFGJ)?n=pg*^ff==Ua3Va?45r`rqOw`02jg6UiJ%pfMKWt+HkXhrcPJjVPP z4rt-g>Iq$P)HhBaD5UMrggCq{?%EyuxqyCl@gQR)<^z~^sWz^J4*CuA#%EZ z`VrAZVf;I~4Dx0GZz#u{@Eq(nhg}>t#BSW@Ij|oibwb(%!;tq`TVS5Z*xC8oB#?WE zA|T?u35^Iv_E~xWw0qdsXt^IRY%@3T0@1xkOCbQkqgYfA*Xb8llP!D&aj~|LMg&iY zmN>Rrn8Kh=oPN#2tsXv@08X)bx?9X445*po`~_Ldn8oSzZdZUPS(+Te1~-vecAh~7 z@jFvyMm}VGh@unJxSTtULfB2p05jEZ4u)p_7Ncr^;e^ARx+qOIUn+`Wu$9QT4YJDA zoNDNOM?~^#%wn|M2}Sb(#7y&Nth<^2H~Q%RsJs8!LaSIRuBjk!FQLQ0@Iyj<$2T{CmX5GT zBqE9tFeeA1S_p>Wl)Rvd(1WTo)Q;!#mOQ21>YI#8t=VQCMO(B{Vj0ry%TK9299uk` zt`y%cZFRka|3-8%UKi04f!{$g(WE2Ihr)w6xi3^4hlrUz){-|&K^S-1@d?O9!o~*j6)R2KHBFo}6N|Z+!}#S3g1TvdHIl}$KpIS{x0madro8lW_E*jhy3{-2!{IsZ#?>7v{cPPmX-6b$wgDLW!WBDx_F<|IK2R$4OU8{pV*y;ARg5uB; z?*Xj3<*Gw z7)B(6=TYc_t5ryC28!b39Zq6y^t)X}Y)$iXE^#T$@U8igAQr*xg~5&*b>FSVC`~E8 zo!jBi#3*Nmkb}q4W6q1rF18xpG%(kpKj7NTqA{B*_d#67bxuU*p@?@v{-=>&{*pF) z@}}LXHg1rPUNjX>U)G50sgwYkEH>=i+OOcvCLSlpr&Q=mbs;J3oiYgh+r# zz)U>+=lNCT=dE~i7gi?p7vwm%AeBEjl=*cLSmoFW(9A20h@xmF+?d;8DicPK`EEO4 zgXIA_(tlZ+L%AKo zAI~C|X#5&-7s4<99ihSeBvi$nHNX>3@-(wJMe}f{zVSt?XGq%=fkhApKoy8 zf~OBk8DEW*~4nx29VhVJ+nwZlZ-2Vrv^?LRS=L1@=NqOLRm8O zsvS&B8AlLK6H`{@IvYq-r*^ZYnT|-?O=9ZY@f3n>?KY#X!1(>Sq zpHDq1Cqz7G4_agVl9r#Cf#lkPlIoW00{bW1bvQ4kFnqEd>ZPcmC8(b|XF1He6navp z)~M~FjKmt9c_FpFDU+kLBW#u`>eurk{(9iaFgof?HIpLsd5VR@kQ9b{2k^%e99wG= z3VP(?3aa+%Xn8o%uek5#L!)~b;eN{;lJt;)V_3vXbR8vTqSFqh($!N;#U;EOQ-@$n z)zk`?7)g8=DDBYeVyzk`p5l~By@};)!cbdKOlR!ib8ho=A4@fj9 zhSv?{sl*pXlRIM5L{^Xy+Jv!1Cj0Y=k<{FMD0Du~_YFYo?p=>LNtsUDWyaexc&Lb1 zctgKGr4j1nD{`~L>S>w|WV8X1mn+}&n6sK)Ckh?S>=mS