From 0cc5f8e517ed621bcd2704dc369821a04e941e57 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jordan=20Fr=C3=A9ry?= Date: Mon, 23 Oct 2023 11:45:53 +0200 Subject: [PATCH] chore: update privacy tree paper experiment --- .../ExperimentPrivacyTreePaper.ipynb | 1830 +++++++++++++---- 1 file changed, 1377 insertions(+), 453 deletions(-) diff --git a/docs/advanced_examples/ExperimentPrivacyTreePaper.ipynb b/docs/advanced_examples/ExperimentPrivacyTreePaper.ipynb index 0fa0288d3..c26ed69a1 100644 --- a/docs/advanced_examples/ExperimentPrivacyTreePaper.ipynb +++ b/docs/advanced_examples/ExperimentPrivacyTreePaper.ipynb @@ -1,7 +1,6 @@ { "cells": [ { - "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ @@ -17,10 +16,14 @@ "metadata": {}, "outputs": [], "source": [ + "# Importing necessary libraries and modules\n", + "\n", "import time\n", "\n", "import numpy as np\n", - "import pandas as pd\n", + "from IPython.display import display\n", + "from onnx import numpy_helper\n", + "from sklearn.datasets import fetch_openml\n", "from sklearn.metrics import (\n", " accuracy_score,\n", " average_precision_score,\n", @@ -29,28 +32,102 @@ " recall_score,\n", ")\n", "from sklearn.model_selection import RepeatedKFold\n", - "from sklearn.preprocessing import LabelBinarizer\n", + "from sklearn.preprocessing import LabelBinarizer, OrdinalEncoder\n", + "\n", + "from concrete.ml.sklearn import DecisionTreeClassifier, RandomForestClassifier, XGBClassifier\n", + "\n", + "\n", + "def basic_preprocessing(df, target_column):\n", + " \"\"\"\n", + " Convert categorical columns to their corresponding code values\n", + " and binarize the target column.\n", "\n", - "from concrete.ml.sklearn import DecisionTreeClassifier, RandomForestClassifier, XGBClassifier" + " Parameters:\n", + " df (pandas.DataFrame): Input dataframe to preprocess.\n", + " target_column (str): Name of the target column to be binarized.\n", + "\n", + " Returns:\n", + " pandas.DataFrame: Preprocessed dataframe.\n", + " \"\"\"\n", + "\n", + " for col in df.columns:\n", + " if df[col].dtype == \"object\":\n", + " df[col] = df[col].astype(\"category\")\n", + " df[col] = df[col].cat.codes\n", + " elif df[col].dtype == \"category\":\n", + " df[col] = df[col].cat.codes\n", + " df[target_column] = LabelBinarizer().fit_transform(df[target_column])\n", + "\n", + " return df" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Loading spambase\n", + "Loading wine\n", + "Loading heart-h\n", + "Loading wdbc\n", + "Loading adult\n", + "Loading steel\n" + ] + } + ], "source": [ - "# Utils function\n", + "# Set up dataset names and their respective IDs for fetching from OpenML\n", + "dataset_names = {\n", + " \"spambase\": 44,\n", + " \"wine\": None,\n", + " \"heart-h\": 1565,\n", + " \"wdbc\": 1510,\n", + " \"adult\": None,\n", + " \"steel\": 1504,\n", + "}\n", "\n", + "datasets = {}\n", "\n", - "def basic_preprocessing(df, target_column):\n", - " \"\"\"Transform categorical values to one-hot.\"\"\"\n", - " categorical_columns = df.select_dtypes(include=[\"object\", \"category\"]).columns\n", - " df[categorical_columns] = df[categorical_columns].apply(\n", - " lambda x: x.astype(\"category\").cat.codes\n", - " )\n", - " df[target_column] = LabelBinarizer().fit_transform(df[target_column])\n", - " return df" + "\n", + "def load_dataset(name, data_id=None):\n", + " \"\"\"Load dataset from OpenML by name or by ID.\n", + "\n", + " Args:\n", + " name (str): Name of the dataset.\n", + " data_id (int, optional): The ID of the dataset on OpenML.\n", + " If provided, the dataset is loaded by ID.\n", + "\n", + " Returns:\n", + " X (np.array): Features of the dataset.\n", + " y (np.array): Target labels of the dataset.\n", + " \"\"\"\n", + " if data_id is not None:\n", + " X, y = fetch_openml(data_id=data_id, as_frame=False, cache=True, return_X_y=True)\n", + " else:\n", + " X, y = fetch_openml(name=name, as_frame=False, cache=True, return_X_y=True)\n", + " return X, y\n", + "\n", + "\n", + "for ds_name, ds_id in dataset_names.items():\n", + " print(f\"Loading {ds_name}\")\n", + "\n", + " X, y = load_dataset(ds_name, ds_id)\n", + "\n", + " # Remove rows with NaN values\n", + " not_nan_idx = np.where(~np.isnan(X).any(axis=1))\n", + " X = X[not_nan_idx]\n", + " y = y[not_nan_idx]\n", + "\n", + " # Convert non-integer target labels to integers\n", + " if not y.dtype == np.int64:\n", + " encoder = OrdinalEncoder()\n", + " y = encoder.fit_transform(y.reshape(-1, 1)).astype(np.int32).squeeze()\n", + "\n", + " datasets[ds_name] = {\"X\": X, \"y\": y}" ] }, { @@ -59,25 +136,25 @@ "metadata": {}, "outputs": [], "source": [ - "# Download the data-sets\n", - "datasets = {}\n", - "spambase = pd.read_csv(\n", - " \"https://archive.ics.uci.edu/ml/machine-learning-databases/spambase/spambase.data\", header=None\n", - ")\n", - "spambase = basic_preprocessing(spambase, 57)\n", - "datasets[\"spambase\"] = {\"X\": spambase.drop(57, axis=1).values, \"y\": spambase[57].values}\n", + "# Setting a random seed for reproducibility across all models and operations\n", + "random_seed = 42\n", "\n", - "adults = pd.read_csv(\n", - " \"https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.data\", header=None\n", - ")\n", - "adults = basic_preprocessing(adults, 14)\n", - "datasets[\"adults\"] = {\"X\": adults.drop(14, axis=1).values, \"y\": adults[14].values}\n", + "# Models with their hyper-parameters\n", + "model_hyperparameters = {\n", + " DecisionTreeClassifier: {\"max_depth\": 5, \"random_state\": random_seed},\n", + " XGBClassifier: {\"max_depth\": 3, \"n_estimators\": 50, \"random_state\": random_seed},\n", + " RandomForestClassifier: {\"n_estimators\": 50, \"random_state\": random_seed},\n", + "}\n", "\n", - "wine = pd.read_csv(\n", - " \"https://archive.ics.uci.edu/ml/machine-learning-databases/wine/wine.data\", header=None\n", - ")\n", - "wine = basic_preprocessing(wine, 0)\n", - "datasets[\"wine\"] = {\"X\": wine.drop(0, axis=1).values, \"y\": wine[0].values}" + "decision_tree_comparison_params = {\n", + " \"spam\": {\"max_leaf_nodes\": 58, \"max_depth\": 17},\n", + " \"heart-h\": {\"max_leaf_nodes\": 5, \"max_depth\": 3},\n", + " \"steel\": {\"max_leaf_nodes\": None, \"max_depth\": 5},\n", + " \"wdbc\": {\"max_leaf_nodes\": None, \"max_depth\": 10},\n", + "}\n", + "\n", + "# List of bit-width used for quantization\n", + "n_bits_list = list(range(1, 10))" ] }, { @@ -86,14 +163,38 @@ "metadata": {}, "outputs": [], "source": [ - "# Set the hyper-parameters and a seed for reproducibility\n", - "RANDOM_SEED = 42\n", - "MODELS = {\n", - " DecisionTreeClassifier: {\"max_depth\": 5, \"random_state\": RANDOM_SEED},\n", - " XGBClassifier: {\"max_depth\": 3, \"n_estimators\": 50, \"random_state\": RANDOM_SEED},\n", - " RandomForestClassifier: {\"n_estimators\": 50, \"random_state\": RANDOM_SEED},\n", - "}\n", - "N_BITS_LIST = range(1, 9)" + "def analyze_gemm_computation(concrete_classifier):\n", + " \"\"\"Analyze the GEMM (General Matrix Multiply) operations in the given ONNX model.\n", + "\n", + " Args:\n", + " concrete_classifier (object): Classifier that contains an ONNX model representation.\n", + " x_train (np.array): Training dataset.\n", + "\n", + " Returns:\n", + " tuple: Shapes of the matrices involved in GEMM operations.\n", + " \"\"\"\n", + "\n", + " # Extract weights and biases from the ONNX model graph\n", + " quant_params = {\n", + " onnx_init.name: numpy_helper.to_array(onnx_init)\n", + " for onnx_init in concrete_classifier.onnx_model.graph.initializer\n", + " if \"weight\" in onnx_init.name or \"bias\" in onnx_init.name\n", + " }\n", + "\n", + " # Extract the shapes of matrices used in GEMM operations\n", + " matrix_shapes = []\n", + " for i in range(1, 4):\n", + " key = [key for key in quant_params.keys() if f\"_{i}\" in key and \"weight\" in key][0]\n", + " matrix_shapes.append(quant_params[key].shape)\n", + "\n", + " return tuple(matrix_shapes)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Experiments for table 1" ] }, { @@ -105,89 +206,114 @@ "name": "stdout", "output_type": "stream", "text": [ - "Average precision: 0.8590584502269576\n", - "Average precision (fp32): 0.856742055225908\n", - "Average precision: 0.8859057196962661\n", - "Average precision (fp32): 0.8978489602543138\n", - "Average precision: 0.8533618889112272\n", - "Average precision (fp32): 0.8702812578992057\n", - "Average precision: 0.5257920855944928\n", - "Average precision (fp32): 0.5313028959377616\n", - "Average precision: 0.5311883185588149\n", - "Average precision (fp32): 0.5435889125308891\n", - "Average precision: 0.4736968702820955\n", - "Average precision (fp32): 0.4878108468555631\n", - "Average precision: 1.0\n", - "Average precision (fp32): 1.0\n", - "Average precision: 1.0\n", - "Average precision (fp32): 1.0\n", - "Average precision: 1.0\n", - "Average precision (fp32): 1.0\n" + "DecisionTreeClassifier on spambase (#features: 57) -> Acc: 0.9103, Acc (fp32): 0.9035, FHE inference time: 1.59s\n", + "XGBClassifier on spambase (#features: 57) -> Acc: 0.9313, Acc (fp32): 0.9362, FHE inference time: 2.01s\n", + "RandomForestClassifier on spambase (#features: 57) -> Acc: 0.9089, Acc (fp32): 0.9184, FHE inference time: 6.88s\n", + "DecisionTreeClassifier on wine (#features: 13) -> Acc: 0.9083, Acc (fp32): 0.9046, FHE inference time: 0.42s\n", + "XGBClassifier on wine (#features: 13) -> Acc: 0.9643, Acc (fp32): 0.9586, FHE inference time: 2.91s\n", + "RandomForestClassifier on wine (#features: 13) -> Acc: 0.9850, Acc (fp32): 0.9813, FHE inference time: 1.83s\n", + "DecisionTreeClassifier on heart-h (#features: 13) -> Acc: 0.6100, Acc (fp32): 0.5998, FHE inference time: 0.45s\n", + "XGBClassifier on heart-h (#features: 13) -> Acc: 0.6679, Acc (fp32): 0.6553, FHE inference time: 5.05s\n", + "RandomForestClassifier on heart-h (#features: 13) -> Acc: 0.6679, Acc (fp32): 0.6644, FHE inference time: 2.62s\n", + "DecisionTreeClassifier on wdbc (#features: 30) -> Acc: 0.9420, Acc (fp32): 0.9391, FHE inference time: 0.52s\n", + "XGBClassifier on wdbc (#features: 30) -> Acc: 0.9649, Acc (fp32): 0.9637, FHE inference time: 1.35s\n", + "RandomForestClassifier on wdbc (#features: 30) -> Acc: 0.9561, Acc (fp32): 0.9526, FHE inference time: 2.37s\n", + "DecisionTreeClassifier on adult (#features: 14) -> Acc: 0.8364, Acc (fp32): 0.8364, FHE inference time: 0.60s\n", + "XGBClassifier on adult (#features: 14) -> Acc: 0.8480, Acc (fp32): 0.8484, FHE inference time: 1.24s\n", + "RandomForestClassifier on adult (#features: 14) -> Acc: 0.8341, Acc (fp32): 0.8341, FHE inference time: 2.42s\n", + "DecisionTreeClassifier on steel (#features: 33) -> Acc: 0.9717, Acc (fp32): 0.9717, FHE inference time: 0.45s\n", + "XGBClassifier on steel (#features: 33) -> Acc: 1.0000, Acc (fp32): 1.0000, FHE inference time: 0.93s\n", + "RandomForestClassifier on steel (#features: 33) -> Acc: 0.9687, Acc (fp32): 0.9586, FHE inference time: 2.18s\n" ] } ], "source": [ + "def benchmark_model(X, y, model, model_params, n_bits, rkf):\n", + " \"\"\"Benchmark a given model and return its evaluation scores.\"\"\"\n", + " scores = {\n", + " \"precision\": [],\n", + " \"recall\": [],\n", + " \"accuracy\": [],\n", + " \"f1\": [],\n", + " \"average_precision\": [],\n", + " \"nodes\": None,\n", + " }\n", + " scores_fp32 = {\"precision\": [], \"recall\": [], \"accuracy\": [], \"f1\": [], \"average_precision\": []}\n", + "\n", + " metric_func_to_key = {\n", + " \"precision_score\": \"precision\",\n", + " \"recall_score\": \"recall\",\n", + " \"f1_score\": \"f1\",\n", + " \"average_precision_score\": \"average_precision\",\n", + " }\n", + "\n", + " for train_index, test_index in rkf.split(X):\n", + " X_train, X_test = X[train_index], X[test_index]\n", + " y_train, y_test = y[train_index], y[test_index]\n", + "\n", + " concrete_model, sklearn_model = model(n_bits=n_bits, **model_params).fit_benchmark(\n", + " X_train, y_train\n", + " )\n", + "\n", + " y_pred = concrete_model.predict(X_test)\n", + " if len(set(y_test)) == 2:\n", + " for metric_func in [precision_score, recall_score, average_precision_score, f1_score]:\n", + " scores_key = metric_func_to_key[metric_func.__name__]\n", + " scores[scores_key].append(metric_func(y_test, y_pred))\n", + " scores[\"accuracy\"].append(accuracy_score(y_test, y_pred))\n", + "\n", + " y_pred_fp32 = sklearn_model.predict(X_test)\n", + " if len(set(y_test)) == 2:\n", + " for metric_func in [precision_score, recall_score, average_precision_score, f1_score]:\n", + " scores_key = metric_func_to_key[metric_func.__name__]\n", + " scores_fp32[scores_key].append(metric_func(y_test, y_pred_fp32))\n", + " scores_fp32[\"accuracy\"].append(accuracy_score(y_test, y_pred_fp32))\n", + "\n", + " shapes = analyze_gemm_computation(concrete_model)\n", + " scores[\"nodes\"] = shapes[0][0]\n", + "\n", + " # Calculate inference time\n", + " concrete_model.compile(X_train)\n", + " concrete_model.fhe_circuit.keygen(force=False)\n", + "\n", + " start = time.time()\n", + " concrete_model.predict(X_test[:1], fhe=\"execute\")\n", + " end = time.time()\n", + " scores[\"inference_time\"] = end - start\n", + "\n", + " start = time.time()\n", + " concrete_model.predict(X_test[:1])\n", + " end = time.time()\n", + " scores_fp32[\"inference_time\"] = end - start\n", + "\n", + " return scores, scores_fp32\n", + "\n", + "\n", "n_bits = 6\n", "scores_global = {}\n", "\n", - "for dataset, data in datasets.items():\n", - " rkf = RepeatedKFold(n_splits=5, n_repeats=3, random_state=0)\n", - " X, y = data[\"X\"].astype(np.float32), data[\"y\"]\n", - " assert len(set(y)) == 2\n", - " assert y.dtype in [int, bool]\n", + "rkf = RepeatedKFold(n_splits=5, n_repeats=3, random_state=0)\n", + "\n", + "for dataset_name, dataset_data in datasets.items():\n", + " X, y = dataset_data[\"X\"].astype(np.float32), dataset_data[\"y\"]\n", + " assert len(set(y)) >= 2\n", + " if y.dtype not in [np.int32, np.bool]:\n", + " print(f\"Unexpected datatype for y in dataset {dataset_name}: {y.dtype}\")\n", "\n", - " key_dataset = f\"{dataset} (#features: {X.shape[1]})\"\n", + " key_dataset = f\"{dataset_name} (#features: {X.shape[1]})\"\n", " scores_global[key_dataset] = {}\n", "\n", - " for model, model_params in MODELS.items():\n", - " clf = model(n_bits=n_bits, **model_params)\n", - " concrete_model, sklearn_model = clf.fit_benchmark(X, y)\n", - " scores_global[key_dataset][model.__name__ + \"_concrete\"] = {}\n", - " scores_global[key_dataset][model.__name__ + \"_fp32\"] = {}\n", - " concrete_scores, fp32_scores = {\n", - " \"precision\": [],\n", - " \"recall\": [],\n", - " \"accuracy\": [],\n", - " \"f1\": [],\n", - " \"average_precision\": [],\n", - " }, {\"precision\": [], \"recall\": [], \"accuracy\": [], \"f1\": [], \"average_precision\": []}\n", - " for train_index, test_index in rkf.split(X):\n", - " X_train_, X_test_ = X[train_index], X[test_index]\n", - " y_train_, y_test_ = y[train_index], y[test_index]\n", - "\n", - " y_pred = concrete_model.predict(X_test_)\n", - " concrete_scores[\"precision\"].append(precision_score(y_test_, y_pred))\n", - " concrete_scores[\"recall\"].append(recall_score(y_test_, y_pred))\n", - " concrete_scores[\"accuracy\"].append(accuracy_score(y_test_, y_pred))\n", - " concrete_scores[\"f1\"].append(f1_score(y_test_, y_pred))\n", - " concrete_scores[\"average_precision\"].append(average_precision_score(y_test_, y_pred))\n", - "\n", - " y_pred = sklearn_model.predict(X_test_)\n", - " fp32_scores[\"precision\"].append(precision_score(y_test_, y_pred))\n", - " fp32_scores[\"recall\"].append(recall_score(y_test_, y_pred))\n", - " fp32_scores[\"accuracy\"].append(accuracy_score(y_test_, y_pred))\n", - " fp32_scores[\"f1\"].append(f1_score(y_test_, y_pred))\n", - " fp32_scores[\"average_precision\"].append(average_precision_score(y_test_, y_pred))\n", - "\n", - " if \"fhe_inference_time\" not in scores_global[key_dataset][model.__name__ + \"_concrete\"]:\n", - " concrete_model.compile(X_train_)\n", - " concrete_model.fhe_circuit.keygen(force=False)\n", - "\n", - " start = time.time()\n", - " concrete_model.predict(X_test_[:1], fhe=\"execute\")\n", - " end = time.time()\n", - " scores_global[key_dataset][model.__name__ + \"_concrete\"][\"inference_time\"] = end - start\n", - "\n", - " start = time.time()\n", - " concrete_model.predict(X_test_[:1], fhe=\"disable\")\n", - " end = time.time()\n", - " scores_global[key_dataset][model.__name__ + \"_fp32\"][\"inference_time\"] = end - start\n", - "\n", - " scores_global[key_dataset][model.__name__ + \"_concrete\"].update(concrete_scores)\n", - " scores_global[key_dataset][model.__name__ + \"_fp32\"].update(fp32_scores)\n", - "\n", - " print(\"Average precision:\", np.mean(concrete_scores[\"average_precision\"]))\n", - " print(\"Average precision (fp32):\", np.mean(fp32_scores[\"average_precision\"]))" + " for cls, model_params in model_hyperparameters.items():\n", + " scores, scores_fp32 = benchmark_model(X, y, cls, model_params, n_bits, rkf)\n", + "\n", + " scores_global[key_dataset][cls.__name__ + \"_concrete\"] = scores\n", + " scores_global[key_dataset][cls.__name__ + \"_fp32\"] = scores_fp32\n", + "\n", + " print(\n", + " f\"{cls.__name__} on {key_dataset} -> Acc: {np.mean(scores['accuracy']):.4f}, \"\n", + " f\"Acc (fp32): {np.mean(scores_fp32['accuracy']):.4f}, \"\n", + " f\"FHE inference time: {scores['inference_time']:.2f}s\"\n", + " )" ] }, { @@ -196,131 +322,566 @@ "metadata": {}, "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "spambase (#features: 57)\n", - "adults (#features: 14)\n", - "wine (#features: 13)\n", - " precision recall \\\n", - "dataset model \n", - "spambase (#features: 57) DecisionTreeClassifier_concrete 0.943842 0.845676 \n", - " DecisionTreeClassifier_fp32 0.938935 0.849004 \n", - " XGBClassifier_concrete 0.953242 0.879400 \n", - " XGBClassifier_fp32 0.947141 0.910795 \n", - " RandomForestClassifier_concrete 0.955468 0.817996 \n", - " RandomForestClassifier_fp32 0.959677 0.841862 \n", - "adults (#features: 14) DecisionTreeClassifier_concrete 0.781405 0.527056 \n", - " DecisionTreeClassifier_fp32 0.784114 0.534578 \n", - " XGBClassifier_concrete 0.795885 0.523122 \n", - " XGBClassifier_fp32 0.801393 0.540081 \n", - " RandomForestClassifier_concrete 0.876720 0.366193 \n", - " RandomForestClassifier_fp32 0.880441 0.386077 \n", - "wine (#features: 13) DecisionTreeClassifier_concrete 1.000000 1.000000 \n", - " DecisionTreeClassifier_fp32 1.000000 1.000000 \n", - " XGBClassifier_concrete 1.000000 1.000000 \n", - " XGBClassifier_fp32 1.000000 1.000000 \n", - " RandomForestClassifier_concrete 1.000000 1.000000 \n", - " RandomForestClassifier_fp32 1.000000 1.000000 \n", - "\n", - " accuracy f1 \\\n", - "dataset model \n", - "spambase (#features: 57) DecisionTreeClassifier_concrete 0.919366 0.891950 \n", - " DecisionTreeClassifier_fp32 0.918713 0.891625 \n", - " XGBClassifier_concrete 0.935449 0.914739 \n", - " XGBClassifier_fp32 0.944796 0.928520 \n", - " RandomForestClassifier_concrete 0.913279 0.881278 \n", - " RandomForestClassifier_fp32 0.923712 0.896770 \n", - "adults (#features: 14) DecisionTreeClassifier_concrete 0.850619 0.629445 \n", - " DecisionTreeClassifier_fp32 0.852492 0.635672 \n", - " XGBClassifier_concrete 0.852830 0.631201 \n", - " XGBClassifier_fp32 0.857007 0.645213 \n", - " RandomForestClassifier_concrete 0.834956 0.516423 \n", - " RandomForestClassifier_fp32 0.839532 0.536637 \n", - "wine (#features: 13) DecisionTreeClassifier_concrete 1.000000 1.000000 \n", - " DecisionTreeClassifier_fp32 1.000000 1.000000 \n", - " XGBClassifier_concrete 1.000000 1.000000 \n", - " XGBClassifier_fp32 1.000000 1.000000 \n", - " RandomForestClassifier_concrete 1.000000 1.000000 \n", - " RandomForestClassifier_fp32 1.000000 1.000000 \n", - "\n", - " average_precision \\\n", - "dataset model \n", - "spambase (#features: 57) DecisionTreeClassifier_concrete 0.859058 \n", - " DecisionTreeClassifier_fp32 0.856742 \n", - " XGBClassifier_concrete 0.885906 \n", - " XGBClassifier_fp32 0.897849 \n", - " RandomForestClassifier_concrete 0.853362 \n", - " RandomForestClassifier_fp32 0.870281 \n", - "adults (#features: 14) DecisionTreeClassifier_concrete 0.525792 \n", - " DecisionTreeClassifier_fp32 0.531303 \n", - " XGBClassifier_concrete 0.531188 \n", - " XGBClassifier_fp32 0.543589 \n", - " RandomForestClassifier_concrete 0.473697 \n", - " RandomForestClassifier_fp32 0.487811 \n", - "wine (#features: 13) DecisionTreeClassifier_concrete 1.000000 \n", - " DecisionTreeClassifier_fp32 1.000000 \n", - " XGBClassifier_concrete 1.000000 \n", - " XGBClassifier_fp32 1.000000 \n", - " RandomForestClassifier_concrete 1.000000 \n", - " RandomForestClassifier_fp32 1.000000 \n", - "\n", - " inference_time \n", - "dataset model \n", - "spambase (#features: 57) DecisionTreeClassifier_concrete 1.714170 \n", - " DecisionTreeClassifier_fp32 0.002973 \n", - " XGBClassifier_concrete 1.550010 \n", - " XGBClassifier_fp32 0.002233 \n", - " RandomForestClassifier_concrete 4.339223 \n", - " RandomForestClassifier_fp32 0.004796 \n", - "adults (#features: 14) DecisionTreeClassifier_concrete 0.575130 \n", - " DecisionTreeClassifier_fp32 0.001656 \n", - " XGBClassifier_concrete 1.411474 \n", - " XGBClassifier_fp32 0.001264 \n", - " RandomForestClassifier_concrete 2.684036 \n", - " RandomForestClassifier_fp32 0.001671 \n", - "wine (#features: 13) DecisionTreeClassifier_concrete 0.433391 \n", - " DecisionTreeClassifier_fp32 0.000979 \n", - " XGBClassifier_concrete 0.940862 \n", - " XGBClassifier_fp32 0.001336 \n", - " RandomForestClassifier_concrete 1.681118 \n", - " RandomForestClassifier_fp32 0.001184 \n" - ] + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
accuracyf1APnodesTime (s)FHE/Clear ratio
spambase (#features: 57)FHE-DT91.0\\% ± 1.0\\%88.0\\% ± 1.3\\%84.3\\% ± 1.7\\%23.0001.589854x
FP32-DT90.3\\% ± 1.0\\%87.4\\% ± 1.2\\%82.4\\% ± 1.8\\%-0.002-
FHE-XGB93.1\\% ± 1.0\\%90.9\\% ± 1.4\\%87.7\\% ± 1.7\\%350.0002.013456x
FP32-XGB93.6\\% ± 0.7\\%91.7\\% ± 1.0\\%88.3\\% ± 1.4\\%-0.004-
FHE-RF90.9\\% ± 1.1\\%87.5\\% ± 1.5\\%84.6\\% ± 1.7\\%750.0006.8751596x
FP32-RF91.8\\% ± 1.1\\%89.0\\% ± 1.4\\%86.0\\% ± 1.6\\%-0.004-
wine (#features: 13)FHE-DT90.8\\% ± 5.2\\%--7.0000.422396x
FP32-DT90.5\\% ± 5.0\\%---0.001-
FHE-XGB96.4\\% ± 3.0\\%--900.0002.915877x
FP32-XGB95.9\\% ± 3.6\\%---0.003-
FHE-RF98.5\\% ± 1.4\\%--500.0001.8261460x
FP32-RF98.1\\% ± 2.0\\%---0.001-
heart-h (#features: 13)FHE-DT61.0\\% ± 5.4\\%--21.0000.448431x
FP32-DT60.0\\% ± 5.5\\%---0.001-
FHE-XGB66.8\\% ± 7.2\\%--1750.0005.0543701x
FP32-XGB65.5\\% ± 6.3\\%---0.001-
FHE-RF66.8\\% ± 6.4\\%--750.0002.6212424x
FP32-RF66.4\\% ± 5.3\\%---0.001-
wdbc (#features: 30)FHE-DT94.2\\% ± 1.9\\%92.0\\% ± 2.8\\%88.4\\% ± 4.1\\%15.0000.520380x
FP32-DT93.9\\% ± 1.9\\%91.7\\% ± 3.0\\%87.3\\% ± 4.8\\%-0.001-
FHE-XGB96.5\\% ± 1.2\\%95.1\\% ± 2.0\\%92.8\\% ± 2.7\\%350.0001.354659x
FP32-XGB96.4\\% ± 1.8\\%94.9\\% ± 2.7\\%92.6\\% ± 3.8\\%-0.002-
FHE-RF95.6\\% ± 1.7\\%93.9\\% ± 2.6\\%91.2\\% ± 3.6\\%700.0002.3711413x
FP32-RF95.3\\% ± 1.8\\%93.4\\% ± 2.9\\%90.4\\% ± 4.1\\%-0.002-
adult (#features: 14)FHE-DT83.6\\% ± 0.4\\%60.4\\% ± 0.7\\%50.3\\% ± 0.8\\%30.0000.598275x
FP32-DT83.6\\% ± 0.4\\%60.4\\% ± 0.7\\%50.3\\% ± 0.7\\%-0.002-
FHE-XGB84.8\\% ± 0.2\\%64.9\\% ± 0.6\\%53.8\\% ± 0.6\\%350.0001.243727x
FP32-XGB84.8\\% ± 0.2\\%65.2\\% ± 0.6\\%53.9\\% ± 0.8\\%-0.002-
FHE-RF83.4\\% ± 0.4\\%57.6\\% ± 1.1\\%49.2\\% ± 0.8\\%750.0002.4181253x
FP32-RF83.4\\% ± 0.4\\%57.6\\% ± 1.2\\%49.2\\% ± 0.9\\%-0.002-
steel (#features: 33)FHE-DT97.2\\% ± 0.7\\%96.1\\% ± 0.9\\%92.5\\% ± 1.7\\%5.0000.455282x
FP32-DT97.2\\% ± 0.7\\%96.1\\% ± 0.9\\%92.5\\% ± 1.7\\%-0.002-
FHE-XGB100.0\\% ± 0.0\\%100.0\\% ± 0.0\\%100.0\\% ± 0.0\\%150.0000.928569x
FP32-XGB100.0\\% ± 0.0\\%100.0\\% ± 0.0\\%100.0\\% ± 0.0\\%-0.002-
FHE-RF96.9\\% ± 1.2\\%95.4\\% ± 1.8\\%93.6\\% ± 2.2\\%700.0002.1841178x
FP32-RF95.9\\% ± 1.1\\%93.9\\% ± 1.5\\%91.4\\% ± 2.3\\%-0.002-
\n", + "
" + ], + "text/plain": [ + " accuracy f1 \\\n", + "spambase (#features: 57) FHE-DT 91.0\\% ± 1.0\\% 88.0\\% ± 1.3\\% \n", + " FP32-DT 90.3\\% ± 1.0\\% 87.4\\% ± 1.2\\% \n", + " FHE-XGB 93.1\\% ± 1.0\\% 90.9\\% ± 1.4\\% \n", + " FP32-XGB 93.6\\% ± 0.7\\% 91.7\\% ± 1.0\\% \n", + " FHE-RF 90.9\\% ± 1.1\\% 87.5\\% ± 1.5\\% \n", + " FP32-RF 91.8\\% ± 1.1\\% 89.0\\% ± 1.4\\% \n", + "wine (#features: 13) FHE-DT 90.8\\% ± 5.2\\% - \n", + " FP32-DT 90.5\\% ± 5.0\\% - \n", + " FHE-XGB 96.4\\% ± 3.0\\% - \n", + " FP32-XGB 95.9\\% ± 3.6\\% - \n", + " FHE-RF 98.5\\% ± 1.4\\% - \n", + " FP32-RF 98.1\\% ± 2.0\\% - \n", + "heart-h (#features: 13) FHE-DT 61.0\\% ± 5.4\\% - \n", + " FP32-DT 60.0\\% ± 5.5\\% - \n", + " FHE-XGB 66.8\\% ± 7.2\\% - \n", + " FP32-XGB 65.5\\% ± 6.3\\% - \n", + " FHE-RF 66.8\\% ± 6.4\\% - \n", + " FP32-RF 66.4\\% ± 5.3\\% - \n", + "wdbc (#features: 30) FHE-DT 94.2\\% ± 1.9\\% 92.0\\% ± 2.8\\% \n", + " FP32-DT 93.9\\% ± 1.9\\% 91.7\\% ± 3.0\\% \n", + " FHE-XGB 96.5\\% ± 1.2\\% 95.1\\% ± 2.0\\% \n", + " FP32-XGB 96.4\\% ± 1.8\\% 94.9\\% ± 2.7\\% \n", + " FHE-RF 95.6\\% ± 1.7\\% 93.9\\% ± 2.6\\% \n", + " FP32-RF 95.3\\% ± 1.8\\% 93.4\\% ± 2.9\\% \n", + "adult (#features: 14) FHE-DT 83.6\\% ± 0.4\\% 60.4\\% ± 0.7\\% \n", + " FP32-DT 83.6\\% ± 0.4\\% 60.4\\% ± 0.7\\% \n", + " FHE-XGB 84.8\\% ± 0.2\\% 64.9\\% ± 0.6\\% \n", + " FP32-XGB 84.8\\% ± 0.2\\% 65.2\\% ± 0.6\\% \n", + " FHE-RF 83.4\\% ± 0.4\\% 57.6\\% ± 1.1\\% \n", + " FP32-RF 83.4\\% ± 0.4\\% 57.6\\% ± 1.2\\% \n", + "steel (#features: 33) FHE-DT 97.2\\% ± 0.7\\% 96.1\\% ± 0.9\\% \n", + " FP32-DT 97.2\\% ± 0.7\\% 96.1\\% ± 0.9\\% \n", + " FHE-XGB 100.0\\% ± 0.0\\% 100.0\\% ± 0.0\\% \n", + " FP32-XGB 100.0\\% ± 0.0\\% 100.0\\% ± 0.0\\% \n", + " FHE-RF 96.9\\% ± 1.2\\% 95.4\\% ± 1.8\\% \n", + " FP32-RF 95.9\\% ± 1.1\\% 93.9\\% ± 1.5\\% \n", + "\n", + " AP nodes Time (s) \\\n", + "spambase (#features: 57) FHE-DT 84.3\\% ± 1.7\\% 23.000 1.589 \n", + " FP32-DT 82.4\\% ± 1.8\\% - 0.002 \n", + " FHE-XGB 87.7\\% ± 1.7\\% 350.000 2.013 \n", + " FP32-XGB 88.3\\% ± 1.4\\% - 0.004 \n", + " FHE-RF 84.6\\% ± 1.7\\% 750.000 6.875 \n", + " FP32-RF 86.0\\% ± 1.6\\% - 0.004 \n", + "wine (#features: 13) FHE-DT - 7.000 0.422 \n", + " FP32-DT - - 0.001 \n", + " FHE-XGB - 900.000 2.915 \n", + " FP32-XGB - - 0.003 \n", + " FHE-RF - 500.000 1.826 \n", + " FP32-RF - - 0.001 \n", + "heart-h (#features: 13) FHE-DT - 21.000 0.448 \n", + " FP32-DT - - 0.001 \n", + " FHE-XGB - 1750.000 5.054 \n", + " FP32-XGB - - 0.001 \n", + " FHE-RF - 750.000 2.621 \n", + " FP32-RF - - 0.001 \n", + "wdbc (#features: 30) FHE-DT 88.4\\% ± 4.1\\% 15.000 0.520 \n", + " FP32-DT 87.3\\% ± 4.8\\% - 0.001 \n", + " FHE-XGB 92.8\\% ± 2.7\\% 350.000 1.354 \n", + " FP32-XGB 92.6\\% ± 3.8\\% - 0.002 \n", + " FHE-RF 91.2\\% ± 3.6\\% 700.000 2.371 \n", + " FP32-RF 90.4\\% ± 4.1\\% - 0.002 \n", + "adult (#features: 14) FHE-DT 50.3\\% ± 0.8\\% 30.000 0.598 \n", + " FP32-DT 50.3\\% ± 0.7\\% - 0.002 \n", + " FHE-XGB 53.8\\% ± 0.6\\% 350.000 1.243 \n", + " FP32-XGB 53.9\\% ± 0.8\\% - 0.002 \n", + " FHE-RF 49.2\\% ± 0.8\\% 750.000 2.418 \n", + " FP32-RF 49.2\\% ± 0.9\\% - 0.002 \n", + "steel (#features: 33) FHE-DT 92.5\\% ± 1.7\\% 5.000 0.455 \n", + " FP32-DT 92.5\\% ± 1.7\\% - 0.002 \n", + " FHE-XGB 100.0\\% ± 0.0\\% 150.000 0.928 \n", + " FP32-XGB 100.0\\% ± 0.0\\% - 0.002 \n", + " FHE-RF 93.6\\% ± 2.2\\% 700.000 2.184 \n", + " FP32-RF 91.4\\% ± 2.3\\% - 0.002 \n", + "\n", + " FHE/Clear ratio \n", + "spambase (#features: 57) FHE-DT 854x \n", + " FP32-DT - \n", + " FHE-XGB 456x \n", + " FP32-XGB - \n", + " FHE-RF 1596x \n", + " FP32-RF - \n", + "wine (#features: 13) FHE-DT 396x \n", + " FP32-DT - \n", + " FHE-XGB 877x \n", + " FP32-XGB - \n", + " FHE-RF 1460x \n", + " FP32-RF - \n", + "heart-h (#features: 13) FHE-DT 431x \n", + " FP32-DT - \n", + " FHE-XGB 3701x \n", + " FP32-XGB - \n", + " FHE-RF 2424x \n", + " FP32-RF - \n", + "wdbc (#features: 30) FHE-DT 380x \n", + " FP32-DT - \n", + " FHE-XGB 659x \n", + " FP32-XGB - \n", + " FHE-RF 1413x \n", + " FP32-RF - \n", + "adult (#features: 14) FHE-DT 275x \n", + " FP32-DT - \n", + " FHE-XGB 727x \n", + " FP32-XGB - \n", + " FHE-RF 1253x \n", + " FP32-RF - \n", + "steel (#features: 33) FHE-DT 282x \n", + " FP32-DT - \n", + " FHE-XGB 569x \n", + " FP32-XGB - \n", + " FHE-RF 1178x \n", + " FP32-RF - " + ] + }, + "execution_count": 6, + "metadata": {}, + "output_type": "execute_result" } ], "source": [ - "# Create an empty dictionary to store the results\n", - "results = {}\n", - "\n", - "# Loop through the data-sets and models to extract the scores\n", - "for key_dataset, dataset_scores in scores_global.items():\n", - " print(key_dataset)\n", - " if key_dataset not in results:\n", - " results[key_dataset] = {}\n", - " for model in MODELS:\n", - " results[key_dataset][model.__name__ + \"_concrete\"] = {}\n", - " results[key_dataset][model.__name__ + \"_fp32\"] = {}\n", - " for metric in [\"precision\", \"recall\", \"accuracy\", \"f1\", \"average_precision\"]:\n", - " results[key_dataset][model.__name__ + \"_concrete\"][metric] = np.mean(\n", - " dataset_scores[model.__name__ + \"_concrete\"][metric]\n", - " )\n", - " results[key_dataset][model.__name__ + \"_fp32\"][metric] = np.mean(\n", - " dataset_scores[model.__name__ + \"_fp32\"][metric]\n", - " )\n", - " results[key_dataset][model.__name__ + \"_concrete\"][\"inference_time\"] = dataset_scores[\n", - " model.__name__ + \"_concrete\"\n", - " ][\"inference_time\"]\n", - " results[key_dataset][model.__name__ + \"_fp32\"][\"inference_time\"] = dataset_scores[\n", - " model.__name__ + \"_fp32\"\n", - " ][\"inference_time\"]\n", - "\n", - "# Create a pandas dataframe from the results dictionary\n", - "df = pd.DataFrame.from_dict(results, orient=\"index\")\n", - "df = df.stack().apply(pd.Series).stack().unstack(2)\n", - "df.index.names = [\"dataset\", \"model\"]\n", - "\n", - "# Display the dataframe\n", - "print(df)" + "import math\n", + "\n", + "import pandas as pd\n", + "\n", + "df = pd.DataFrame.from_dict(\n", + " {(i, j): value for i, scores in scores_global.items() for j, value in scores.items()},\n", + " orient=\"index\",\n", + ")\n", + "\n", + "\n", + "df[\"FHE/Clear ratio\"] = (df[\"inference_time\"] / df[\"inference_time\"].shift(-1)).apply(\n", + " lambda x: \"\" if (x < 1) or (math.isnan(x)) else str(int(round(x, 0))) + \"x\"\n", + ")\n", + "\n", + "\n", + "def format_scores(val):\n", + " if isinstance(val, list):\n", + " if not val:\n", + " return \"-\"\n", + " return f\"{np.mean(val) * 100:.1f}\\\\% ± {np.std(val) * 100:.1f}\\\\%\"\n", + "\n", + " if pd.isna(val):\n", + " return \"-\"\n", + "\n", + " if isinstance(val, (float, int)):\n", + " # To ensure all floating point values are treated as percentages\n", + " return f\"{val:.3f}\"\n", + "\n", + " if \"x\" in str(val): # Ensure that val is treated as a string\n", + " return val\n", + "\n", + " return \"-\"\n", + "\n", + "\n", + "df = df.applymap(format_scores)\n", + "\n", + "# Renaming for display\n", + "model_names = {\n", + " \"DecisionTreeClassifier_concrete\": \"FHE-DT\",\n", + " \"DecisionTreeClassifier_fp32\": \"FP32-DT\",\n", + " \"XGBClassifier_concrete\": \"FHE-XGB\",\n", + " \"XGBClassifier_fp32\": \"FP32-XGB\",\n", + " \"RandomForestClassifier_concrete\": \"FHE-RF\",\n", + " \"RandomForestClassifier_fp32\": \"FP32-RF\",\n", + "}\n", + "\n", + "for original, renamed in model_names.items():\n", + " df.index = df.index.set_levels(df.index.levels[1].str.replace(original, renamed), level=1)\n", + "\n", + "df.columns = df.columns.str.replace(\"average_precision\", \"AP\")\n", + "\n", + "# Reordering Columns\n", + "columns_order = [col for col in df if col not in [\"FHE/Clear ratio\", \"inference_time\"]] + [\n", + " \"inference_time\",\n", + " \"FHE/Clear ratio\",\n", + "]\n", + "df = df[columns_order]\n", + "\n", + "# Drop and rename columns\n", + "df.columns = df.columns.str.replace(\"inference_time\", \"Time (s)\")\n", + "df.drop(columns=[\"precision\", \"recall\"], inplace=True)\n", + "\n", + "# Adjust LaTeX output\n", + "latex_code = df.to_latex(multirow=True, escape=False, column_format=\"l|l|l|l|l|l|l|l\")\n", + "\n", + "latex_code = latex_code.replace(\"#\", \"\\\\#\")\n", + "display(df)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Experiments Figure 2. - Impact of the precision" ] }, { @@ -332,90 +893,148 @@ "name": "stdout", "output_type": "stream", "text": [ - "Calculating scores for DecisionTreeClassifier with n_bits=1...\n", - "Calculating scores for XGBClassifier with n_bits=1...\n", - "Calculating scores for RandomForestClassifier with n_bits=1...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=2...\n", - "Calculating scores for XGBClassifier with n_bits=2...\n", - "Calculating scores for RandomForestClassifier with n_bits=2...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=3...\n", - "Calculating scores for XGBClassifier with n_bits=3...\n", - "Calculating scores for RandomForestClassifier with n_bits=3...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=4...\n", - "Calculating scores for XGBClassifier with n_bits=4...\n", - "Calculating scores for RandomForestClassifier with n_bits=4...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=5...\n", - "Calculating scores for XGBClassifier with n_bits=5...\n", - "Calculating scores for RandomForestClassifier with n_bits=5...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=6...\n", - "Calculating scores for XGBClassifier with n_bits=6...\n", - "Calculating scores for RandomForestClassifier with n_bits=6...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=7...\n", - "Calculating scores for XGBClassifier with n_bits=7...\n", - "Calculating scores for RandomForestClassifier with n_bits=7...\n", - "Calculating scores for DecisionTreeClassifier with n_bits=8...\n", - "Calculating scores for XGBClassifier with n_bits=8...\n", - "Calculating scores for RandomForestClassifier with n_bits=8...\n" + "DecisionTreeClassifier with 1-bits:\n", + "Average precision: 0.408932530057913\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 1-bits:\n", + "Average precision: 0.39404467418842154\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 1-bits:\n", + "Average precision: 0.39404467418842154\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 2-bits:\n", + "Average precision: 0.5783561241072402\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 2-bits:\n", + "Average precision: 0.6114845771874085\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 2-bits:\n", + "Average precision: 0.5692185974635017\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 3-bits:\n", + "Average precision: 0.6809716709773868\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 3-bits:\n", + "Average precision: 0.7447793864050538\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 3-bits:\n", + "Average precision: 0.7371651377079289\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 4-bits:\n", + "Average precision: 0.733924323832143\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 4-bits:\n", + "Average precision: 0.8389496176111542\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 4-bits:\n", + "Average precision: 0.8056925848877744\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 5-bits:\n", + "Average precision: 0.8101306169806723\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 5-bits:\n", + "Average precision: 0.8679259707334989\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 5-bits:\n", + "Average precision: 0.8304046578932958\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 6-bits:\n", + "Average precision: 0.8426735780163795\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 6-bits:\n", + "Average precision: 0.8772631033645845\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 6-bits:\n", + "Average precision: 0.8459941433803378\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 7-bits:\n", + "Average precision: 0.8440934158260279\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 7-bits:\n", + "Average precision: 0.8844115062267996\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 7-bits:\n", + "Average precision: 0.8523380007137836\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 8-bits:\n", + "Average precision: 0.8329951951056004\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 8-bits:\n", + "Average precision: 0.8840570477014416\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 8-bits:\n", + "Average precision: 0.8527017103592347\n", + "Average precision (fp32): 0.8601540794231147\n", + "DecisionTreeClassifier with 9-bits:\n", + "Average precision: 0.8344647983632885\n", + "Average precision (fp32): 0.8240131348716936\n", + "XGBClassifier with 9-bits:\n", + "Average precision: 0.8853757667924429\n", + "Average precision (fp32): 0.8828445864365427\n", + "RandomForestClassifier with 9-bits:\n", + "Average precision: 0.8579231686644472\n", + "Average precision (fp32): 0.8601540794231147\n" ] } ], "source": [ - "def calculate_scores(model, n_bits, X, y, rkf):\n", - " \"\"\"Calculates scores.\"\"\"\n", - " scores = {\n", - " \"precision\": [],\n", - " \"recall\": [],\n", - " \"accuracy\": [],\n", - " \"f1\": [],\n", - " \"average_precision\": [],\n", - " \"fhe_inference_time\": [],\n", - " }\n", + "def evaluate_model(X, y, model, rkf):\n", + " \"\"\"Evaluate a given model and return its scores.\"\"\"\n", + " scores = {\"precision\": [], \"recall\": [], \"accuracy\": [], \"f1\": [], \"average_precision\": []}\n", " scores_fp32 = {\"precision\": [], \"recall\": [], \"accuracy\": [], \"f1\": [], \"average_precision\": []}\n", "\n", - " for train_index, test_index in rkf.split(X):\n", - " X_train_, X_test_ = X[train_index], X[test_index]\n", - " y_train_, y_test_ = y[train_index], y[test_index]\n", - "\n", - " # Train the model and benchmark it\n", - " clf = model(n_bits=n_bits, **MODELS[model])\n", - " concrete_model, sklearn_model = clf.fit_benchmark(X_train_, y_train_)\n", - "\n", - " # Make predictions and calculate scores\n", - " y_pred = concrete_model.predict(X_test_)\n", - " scores[\"precision\"].append(precision_score(y_test_, y_pred))\n", - " scores[\"recall\"].append(recall_score(y_test_, y_pred))\n", - " scores[\"accuracy\"].append(accuracy_score(y_test_, y_pred))\n", - " scores[\"f1\"].append(f1_score(y_test_, y_pred))\n", - " scores[\"average_precision\"].append(average_precision_score(y_test_, y_pred))\n", - "\n", - " y_pred = sklearn_model.predict(X_test_)\n", - " scores_fp32[\"precision\"].append(precision_score(y_test_, y_pred))\n", - " scores_fp32[\"recall\"].append(recall_score(y_test_, y_pred))\n", - " scores_fp32[\"accuracy\"].append(accuracy_score(y_test_, y_pred))\n", - " scores_fp32[\"f1\"].append(f1_score(y_test_, y_pred))\n", - " scores_fp32[\"average_precision\"].append(average_precision_score(y_test_, y_pred))\n", - "\n", - " return {\n", - " f\"{model.__name__}_concrete\": scores,\n", - " f\"{model.__name__}_fp32\": scores_fp32,\n", + " metric_func_to_key = {\n", + " \"precision_score\": \"precision\",\n", + " \"recall_score\": \"recall\",\n", + " \"f1_score\": \"f1\",\n", + " \"average_precision_score\": \"average_precision\",\n", " }\n", "\n", + " for train_index, test_index in rkf.split(X):\n", + " X_train, X_test = X[train_index], X[test_index]\n", + " y_train, y_test = y[train_index], y[test_index]\n", + "\n", + " concrete_model, sklearn_model = model.fit_benchmark(X_train, y_train)\n", + "\n", + " for model_instance, score_dict in [(concrete_model, scores), (sklearn_model, scores_fp32)]:\n", + " y_pred = model_instance.predict(X_test)\n", + " for metric_func in [precision_score, recall_score, average_precision_score, f1_score]:\n", + " score_key = metric_func_to_key[metric_func.__name__]\n", + " score_dict[score_key].append(metric_func(y_test, y_pred))\n", + " score_dict[\"accuracy\"].append(accuracy_score(y_test, y_pred))\n", + "\n", + " return scores, scores_fp32\n", + "\n", "\n", - "# Setup repeated k-fold cross-validation\n", "rkf = RepeatedKFold(n_splits=5, n_repeats=3, random_state=0)\n", - "X, y = datasets[\"spambase\"][\"X\"], datasets[\"spambase\"][\"y\"]\n", - "X = X.astype(np.float32)\n", + "X, y = datasets[\"spambase\"][\"X\"].astype(np.float32), datasets[\"spambase\"][\"y\"]\n", "assert len(set(y)) == 2\n", - "assert y.dtype in [int, bool]\n", + "if y.dtype not in [np.int32, np.bool]:\n", + " print(f\"Unexpected datatype for y in dataset spambase: {y.dtype}\")\n", "\n", - "# Calculate scores for different models and n_bits\n", "scores_global = {}\n", - "for n_bits in N_BITS_LIST:\n", + "\n", + "for n_bits in n_bits_list:\n", " scores_global[n_bits] = {}\n", - " for model in MODELS:\n", - " print(f\"Calculating scores for {model.__name__} with n_bits={n_bits}...\")\n", - " scores = calculate_scores(model, n_bits, X, y, rkf)\n", - " scores_global[n_bits].update(scores)" + "\n", + " for model_cls, params in model_hyperparameters.items():\n", + " model_instance = model_cls(n_bits=n_bits, **params)\n", + " scores, scores_fp32 = evaluate_model(X, y, model_instance, rkf)\n", + "\n", + " model_name = model_cls.__name__\n", + " scores_global[n_bits][model_name + \"_concrete\"] = scores\n", + " scores_global[n_bits][model_name + \"_fp32\"] = scores_fp32\n", + "\n", + " print(f\"{model_name} with {n_bits}-bits:\")\n", + " print(\"Average precision:\", np.mean(scores[\"average_precision\"]))\n", + " print(\"Average precision (fp32):\", np.mean(scores_fp32[\"average_precision\"]))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# P-Error search" ] }, { @@ -423,11 +1042,98 @@ "execution_count": 8, "metadata": {}, "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "100%|██████████| 14/14 [16:43<00:00, 71.67s/it]\n" + ] + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "from tqdm import tqdm\n", + "\n", + "\n", + "def evaluate_model_on_error_rates(X_train, X_test, y_test, concrete_model, p_error_list):\n", + " \"\"\"Evaluate the concrete model on different error rates and return accuracy and time taken.\"\"\"\n", + " acc_scores = []\n", + " time_scores = []\n", + " real_p_error_list = []\n", + "\n", + " for p_error in tqdm(p_error_list):\n", + " concrete_model.compile(X_train, p_error=p_error)\n", + " real_p_error_list.append(concrete_model.fhe_circuit.p_error)\n", + " concrete_model.fhe_circuit.keygen(force=False)\n", + "\n", + " start_time = time.time()\n", + " y_pred = concrete_model.predict(X_test, fhe=\"execute\")\n", + " end_time = time.time()\n", + "\n", + " acc_scores.append(accuracy_score(y_pred, y_test))\n", + " time_scores.append(end_time - start_time)\n", + "\n", + " return acc_scores, time_scores, real_p_error_list\n", + "\n", + "\n", + "plt.rcParams.update({\"font.size\": 16})\n", + "n_bits = 6\n", + "p_error_list = [2e-40, 1e-6, 1e-5, 1e-4, 0.001, 0.005, 0.01, 0.05, 0.1, 0.3, 0.5, 0.7, 0.9, 0.95]\n", + "X, y = datasets[\"spambase\"][\"X\"].astype(np.float32), datasets[\"spambase\"][\"y\"]\n", + "\n", + "clf = DecisionTreeClassifier(n_bits=n_bits, **model_hyperparameters[DecisionTreeClassifier])\n", + "rkf = RepeatedKFold(n_splits=20, n_repeats=3, random_state=0)\n", + "\n", + "for train_index, test_index in rkf.split(X):\n", + " X_train, X_test = X[train_index], X[test_index]\n", + " y_train, y_test = y[train_index], y[test_index]\n", + "\n", + " concrete_model, _ = clf.fit_benchmark(X_train, y_train)\n", + "\n", + " # Calculating num_nodes using analyze_gemm_computation function\n", + " shapes = analyze_gemm_computation(concrete_model)\n", + " num_nodes = shapes[0][0]\n", + "\n", + " acc_scores, time_p_error, real_p_error_list = evaluate_model_on_error_rates(\n", + " X_train, X_test, y_test, concrete_model, p_error_list\n", + " )\n", + " break" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n" + ] + }, { "data": { - "image/png": "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", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkoAAAHeCAYAAABkPCRRAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABrTklEQVR4nO3dd3hT1f8H8HcautOWUSijhbassoRC2aBsoSoiMgQREJyIbAHBsqos8YtFQESWqOyfIKPKRgUUKSBDNhRaNm2hk7Zpe35/xFwSkpsmbdKk5f16njxN7v3ck3Oz7qfnnHuuQgghQEREREQGnOxdASIiIiJHxUSJiIiISAYTJSIiIiIZTJSIiIiIZDBRIiIiIpLBRImIiIhIBhMlIiIiIhlMlIiIiIhkMFEiIiIiksFEqYRbtWoVFAqFwc3NzQ1+fn4ICQlB7969MWfOHFy6dMne1c2Xtv6rVq2yWpnt2rWDQqHA4MGDrVamNQQGBhp978y9TZs2zd67UCCnTp3C5MmT0aJFC1SuXBmurq7w9vZGzZo10bdvX3z33XdIS0sz2O7atWvSvh84cKDoK25l5nwu9+3bh5deegkVK1aEs7OztP9atvi+2EpqairWrVuHt956C6GhofDx8YGzszPKly+P9u3b48svv0RqaqrJMk6dOoW5c+eid+/eqFevHvz8/ODi4gJvb280aNAAH3zwAU6fPl1Ee2R/06ZNg0KhQGBgoL2rUrwJKtFWrlwpAJh1UygUonPnzuLixYv2rrYsbV1XrlxptTKfe+45AUAMGjTIamVaQ7Vq1cx+74zdpk6dau9dsEhiYqLo27evUCgU+e5bmTJlxIIFC/S2j42Nldbv37/fPjthRfl9Lrdt2yb7WmnZ4vtiCzdv3hRubm75vu9Vq1YVMTExsuUMHTo03zKUSqWYPn16Ee6d/UydOlUAENWqVbN3VYo1tig9RaKjo5GamorU1FQkJyfj+vXrOHz4MP73v/8hNDQUQgjs3r0boaGh2Lp1q72r+9Q7e/as9H49edN6/fXXZWMmTZpkx9pb5tq1a2jRogXWr18PIQSaNWuGb775BmfOnMH9+/cRHx+PQ4cOISIiAlWrVsWDBw/wxRdf2LvadjVz5kwIIVCnTh38+eefSEpKMvh8FBfZ2dnIzMyEu7s7Bg4ciA0bNuDKlStISkrC6dOnMX78eCiVSsTFxaFLly64deuW0XLKlCmDPn36YOHChfjtt99w6dIlJCQk4MyZM1iyZAmCg4ORm5uLqVOnYunSpUW8l1Rs2TlRIxvTbVHK77/s77//Xnh4eAgAwtPTU/zzzz9FU0mymPY9dbRWsILIzMwUoaGhUqvml19+mW/87NmzRY0aNfSWl7QWpfxov6vz58+3d1UK7ebNm+Ljjz8WiYmJsjHff/+99P5+8MEHBXqepKQkUalSJQFABAYGFrS6xQZblKyDLUokGTBggDSWIT09HePGjbNvheip8Pnnn+PEiRMAgOnTp2PkyJEm411dXTFhwgRs2bKlCGrnuDIyMgAApUuXtm9FrKBy5cqYOXMmypYtKxszYMAANGjQAICmdbwgypQpg169egHQtGImJiYWqBx6ujBRIj29e/dG586dAQB79uzByZMnZWOvXLmCUaNGoX79+vD29oa7uztq1KiBd955BxcuXMj3ubKysrBkyRJ07doVlSpVgqurK/z8/BAWFobx48fj+PHjBtvkNzj18OHDGDBgAIKDg+Hu7g4PDw9Uq1YNLVu2xMSJE3H06FGDbcwZNJueno65c+eiZcuWKFu2LFxdXVGlShX07t0bO3fulN3uyQHGOTk5iIqKQpMmTeDl5QUvLy80b94cS5cuhRAi39fMXLqvU15eHpYsWYK2bduifPnycHJyMhjonZubi9WrV+OFF15ApUqV4OLignLlyqFdu3ZYunQpcnJy8n3Of/75B++88w5q1aoFlUoFT09P1K1bF2PGjMHNmzeNbpOZmYkFCxYA0Axe//jjj83ex3r16pkdq32u6OhovP/++3jmmWfg7e0NZ2dnVKhQAZ06dcK3336L7Oxsk2WcOXMG77zzDkJCQuDp6Qk3Nzf4+/sjLCwMI0eOxN69e41uZ63P5YEDBwwGbL/55pt6g/h1vxvmDObOzMzEwoUL0bFjR1SoUAEuLi6oUKECunXrhg0bNsh+LnVPFAGAGzduYPTo0ahduzY8PT2hUChw7do1+RezALTvudznyRwuLi7SfVdXV4u3f3K/ExIS8NFHH6FmzZpwc3NDuXLl0K1bN+zfvz/fshITExEREYHGjRujdOnScHNzQ2BgIAYOHIi///473+0zMzMxc+ZMNGjQAB4eHihXrhyee+45/PDDDxbt04EDB/DGG28gKCgI7u7u8Pb2RqNGjTBlyhQ8ePBAdjshBNauXYvw8HDpd8PHxwc1atRA586dMXfuXMTFxVlUF4dk5xYtsjFLut601q1bJ23z+eefG42JiooSzs7OsgMmS5UqJVasWCH7HKdOnRLBwcEmB10aay7WrjM2OPXzzz/PdyDnCy+8YLBdfoNm//33X1G1alWT5b7xxhsiOzvbYFvd7qDt27eLNm3ayJYxdOhQ2ddL7nWQq7N2/ZIlS0SHDh1MDvSOj48XjRs3Nrl/zZs3F/fv3zf6XHl5eWL8+PEmB2GrVCoRHR1tsO2vv/4qxRR2gG1+XW+jRo3K9/PRokUL8eDBA6Plr127VpQqVcrk9vXq1TPYzpqfy/379+dblu53w9T3RQghzpw5I4KCgkyW99JLL4n09HSDbXV/W44cOSLKli1rsG1sbKzR5y0o7WfZ19e3QNs/evRI1KhRQwAQderUKVAZuvv977//iipVqhh93RQKhfjuu+9ky/ntt99EmTJlTL72EydOlN3+3r17okGDBrLbDhw4MN+ut8zMTPHGG2+YrIOfn584evSowbY5OTni5ZdfzvfzKHcMKU6YKJVwBUmU4uPjpW169uxpsP7rr7/W+xH99ddfxa1bt0RCQoI4cOCA6NatmwAgnJycxJ49ewy2j42NlX5U3dzcxPjx40VMTIxISEgQt27dEnv37hXjxo0TTZs2NdhW7of/woULQqlUCgAiLCxMbNmyRcTGxooHDx6Ia9euiejoaDFy5EijiYWpRCkxMVH6IXR1dRXTp08XFy5cEAkJCeL3338Xzz//vFSn4cOHG91X7frg4GDh6ekpZs6cKS5cuCCSkpLEn3/+Kdq2bSvF7Ny5U/6NMfI65Jco+fv7CycnJzF69Gjxzz//iISEBHHy5Enx999/CyGESE5OFjVr1pQOPl988YX4999/RVJSkrhy5YqIiooSPj4+AoBo3769yM3NNXiuCRMmSAeGgQMHit9++03cu3dP3Lt3T0RHR4vmzZsLAMLDw0OcOXNGb9vJkydb/PmUk1+iNHXqVDF06FCxbt06cfToUREfHy/u3r0rjh07JqZMmSIdtF577TWDbR88eCBUKpUAIGrUqCF+/PFHcfnyZfHgwQNx48YNsXfvXjFp0iTRtWtXve2s/bnMyckRqampIjU1VdrXJUuWSMtSU1OFWq2W4k0lSvHx8cLX11c6kH7zzTfi4sWLIikpSZw7d05Mnz5duLq6SgfdJ+n+tvj7+wt/f3+xYsUKERcXJ+7evSt++eUX8fDhQ/k3zEJ37tyR/jl76aWXzN4uJydH3LhxQ/z000+iSZMm0mf1559/LlA9dPc7ODhYBAYGitWrV4v4+Hhx//59sXnzZhEQECAAzT8ICQkJBmVcvnxZ+jz5+PiIqKgoERsbK+7duyd+/fVX0bRpU+k55s2bZ7B9Xl6eaNeunRTz/vvvi5MnT4qEhARx5MgR0bt3bwFASoLlEqW+ffsKQPOP7ciRI8WRI0ek3+ENGzaIkJAQKVm6ffu23rbLli2Tnv+1114Tv//+u7hx44ZISkoS58+fF2vXrhX9+vUzODu1OGKiVMIVJFESQkin6rZs2VJv+e3bt6V1kyZNMrptXl6e9AVs0KCBwXptIuXi4iL++OMP2Tro/uBryf3wL1iwQACaU39NDQg1xlSipNsKsWXLFoP1ubm5okePHlLM6dOn9dbrHryVSqX47bffDMpIS0uTBpj27dvXrDqbmygBEIsXL5YtZ+TIkQKAqFSpkrh27ZrRmBMnTkgHzE2bNumtO3bsmNSStHTpUqPbZ2dnSy1pL774ot66/v37S/W8c+eOiT3OX2EHc586dUoolUqhUCjElStX9NZt3bpVKvvkyZNml2mrz6UQ5p36bypG2xpQr149kZSUZHT76OhoqYwnT8vX/W0pV66ciIuLM3fXCmTgwIHS8/3yyy/5xteuXdtoC0dQUFCBkyQh9Pe7SpUq4u7duwYxx44dk2K+/vprg/Xa3wwXFxejrTUZGRmiWbNm0j+TT7bm/t///Z9U/ieffGK0nrotRcYSpc2bN0v/0Mq9ng8fPpT+kXryH8GePXsKAKJx48ZGty1JOEaJjNIOEH2yf3rJkiXIzMxEtWrVMH36dKPbKhQKzJkzBwBw+vRpnDp1Slp38eJF/PLLLwCAcePGoU2bNrJ1KFWqlNn11Y6h8fT0tNrg1tzcXGlsxwsvvICXX37ZIMbJyQkLFy6U6vrtt9/KltenTx88++yzBss9PT3Ru3dvADA6VqUw6tSpg/fff9/ouvT0dCxbtgwAMGPGDFSrVs1oXKNGjdCvXz8AwI8//qi3bsGCBRBCoHXr1nj77beNbu/s7IxPP/0UgGYQ7sOHD6V1SUlJ0n17D0pu0KABGjduDCEE9uzZo7dOd4xW5cqVzS7TFp9La4iNjZWmAPnyyy9RpkwZo3HdunVDu3btABi+97o++ugjBAQEWL2eWqtXr8bq1asBAC+//DK6du1aoHLKlCmDESNGoG3btlap15QpU1ChQgWD5Y0bN8YzzzwDwPA7fe/ePem1f/vttxEWFmawvbu7uzR2LzMzE99//73e+hUrVgAA/Pz8EBERYbRu//vf//TGYz3pyy+/BAD069dP9vX08fHB5MmTAQBr1qzRG6+m/WxXqlRJ9jlKCiZKZJT2C6E7aBSAdADp0KEDMjMzkZaWZvRWrlw5+Pr6AtD/odA9AFlzJuxGjRoBAFJSUvD222/LzrNiidOnT0sH9T59+sjGValSBa1btwYA/PHHH7Jx3bp1k11Xu3ZtAMCdO3cKUFN54eHhsuv+/PNPpKenA9AMHJZ7L9PS0mR/9LXvZ+fOnU1uX7duXQBAXl4ejh07ZtV9tERSUhLmzJmDdu3aSbM26w6E1u7fkycjPPPMM9J34c0338Tly5fNej5bfC6tYe/evRBCwM3NDc2aNTP53mn3wVQS/8ILL9isrocPH8a7774LAKhWrRqWL19u1nbHjx+X5oy7fPkyli9fjvLly2P06NEIDQ01eaKKuQrynT58+DDy8vIAmP5dad68ufTPi+7vihAChw4dAgC8+OKLssmQr68vnnvuOaPrMjIycPjwYQBAx44dTb7/2gH0SUlJuHr1qlSG9nPxyy+/YMGCBdJvSUnERImMSk5OBgCD03XPnz8PAFi5cqV01pbcLSEhAQBw//59afsrV64AADw8PFCzZk2r1bd9+/Z46aWXAGj+2/L390eTJk0wcuRI/PTTT0hJSbG4zOvXr0v3tQd6OdofE1Nn+ZhqifDw8ADw+JRvawkODpZdp30vAaBmzZom38sxY8YA0H8v09LSpLOPpk2bZnJ73f+6dcvQ/XzptjTZwl9//YWQkBBMnDgRv/32G+7duwe1Wm00Vvv516pevTo++OADAMD27dtRs2ZN1KtXD++99x7WrFmjt0+6bPG5tAbte5+ZmQkfHx+T75225UFuHwHTn7PCOH36NF588UVkZmaifPny2LlzJ8qVK2fWth4eHlCpVPD29kb16tUxZMgQHDt2DE2aNMH169cRHh5u9FI4lijId7qwvyvJycnSd6VOnTomt5cr/+rVq9Jnf8iQISbf/6ZNm0rb6X4GRo8ejYCAAOTl5WHkyJHw9fVFhw4dMHXqVOzfv9+sM2WLCyZKZCAuLg6ZmZkADH8InjyAmENbFgDpwODl5VWIGhq3adMmzJkzB0FBQRBC4Pjx41iwYAFeffVV+Pn54b333rOo/rozHOdXX+16U7MiK5VKs5/bWrQ/1sYU5L3Mysoq1PaA/uchKChIuq+buFlbSkoKevTogfv376N8+fKYNWsW/vzzT9y8eRMPHz6UZrTWtgwa+5FfsGABli5dKh28zp49i2+++Qavv/46KleujNdee83oaevW/lxaQ2G/x08y9TkrqEuXLqFz58548OABSpcujV27dkmtNAWlUqkwd+5cAMCtW7ewfv36QpVnzndat7sKKPzvim5yp1KpTG4vt94a393SpUvj6NGj+OCDD1C6dGlkZmZi//79mDFjBjp06IAqVargiy++kFrPijMmSmRA2yQLAK1atdJbp/3ijRkzBkJzMkC+N905e7y9vQGYTigKysXFBePHj8fVq1dx8eJFfPfddxg6dCj8/PyQmZmJb775Bu3btzf7Px3dH7H8/vPUrrdFAmgruj+iKSkpZr+fxrbXjlUy56bb5ao7VuS3336z2b5u2rQJd+/ehZOTE/bv34+JEydKF9318fGBSqWCSqUy+blUKBR4++23cebMGcTFxWH9+vUYPnw4qlWrhpycHKxfvx6tWrUyaBmz9ufSGrTvXdmyZc1+36w9J5Ip169fR8eOHXH37l2oVCr88ssvUldPYbVo0UK6r53otCgV9ndF93tn7vZP0i1j69atZn8GtOPVtPz8/LBw4UIkJCTg6NGj+Oqrr/DKK6/A3d0d9+7dw7hx4zBq1CiTdSwOmCiRAd0xANrJJ7W0TezaLjRL1ahRA4CmOfrSpUsFrGH+atasiYEDB2LZsmWIj4+Xuk1OnDiBbdu2mVWG7hW3//33X5OxZ86cMdjG0el2lxTk/fTx8ZG6QQr6eXj22WelsWwrV65Ebm5ugcrJzz///ANAM9ZIbqLK7OxsXLx40azyAgIC0KdPH3z11Ve4evUqZs2aBUDTGrty5UrZ7azxubQG7Xv/4MEDvQH1juDWrVvo2LEj4uPj4ebmhq1bt+olN4Vl7y6hwv6u+Pj4SCcGnDt3zuT2Z8+ela2Dk5Pm8F/Q764upVKJsLAwDB8+HD/99BPi4+PRsmVLAMCiRYukYRjFFRMl0rNx40ZpgG7Xrl0NDirPP/88AM0g3oJM/9+pUyfp/nfffVeImprP2dkZU6dOlR6b28VTv3596Qdp06ZNsnG3bt2SBlda62yaovDcc89JMxOvW7euQGV06dIFAPDTTz/lO6u1Me7u7vjwww8BaMZhzJw50+xt8zvI6NJ2GZpKxDZu3Giye0mOk5MTJkyYIP2Xbu7nq6CfS2vQvm9CCGzYsKHInjc/9+7dQ8eOHXHlyhW4uLjg//7v/9C+fXurPsfvv/8u3a9evbpVyzZHq1atpC47U78rR48elcYz6f6uKBQKqYt4x44dst+7hIQE2VZaHx8fNG/eHEDBv/umlCtXDqNHjwagOYHDlv8UFwUmSiT58ccfpW4RlUqFzz//3CDmgw8+gJubG9LT0/Hmm2/qjVkx5skf/5o1a0pnyMybNw9//vmn7LaW/Od36dIlk33huv81mTsYVKlU4s033wQAbNu2DTt27DCIycvLw4cffijVVe4UeUfk7e0t1ffLL7/M95ILmZmZegNRAUiDvOPj4zFy5Mh8xyMYSwbGjx+Phg0bAtAMCteeFi0nOzsb8+bNwyuvvGIyTpe2BeXcuXNGW41u3ryJCRMmyG4fGxtrMhG8ffu2dNaP7ufLFp9La6hduzZefPFFAMDkyZPzTTpTUlJw+/Ztm9bpwYMH6NKlC86fPw+lUok1a9aYPGvT2PZ37941GZOUlCS9z6VKlUKPHj0KU+UCKV++PLp37w4AWLp0qdHuv8zMTIwYMQIA4ObmhjfeeENv/ZAhQwBozqjTTr3xpDFjxpj8zI4dOxYAcOTIEalFVI4QwuBM0PwSe3t9tm3CKrMxkcPSnRwtOjpamr03OTlZxMXFiT///FPMnz9f7xIWnp6eYseOHbJlLl26VIqtV6+eWL58ubh06ZJ48OCBuH37tvjrr79EVFSUePbZZ4VKpTLY/tq1a9LM3O7u7mLixIni+PHjIjExUdy5c0f89ttvYuLEiaJZs2YG22qf98kJ9AYNGiQCAwPFxIkTxa5du0RcXJxISkoSly5dEsuXL5dm1/b09BS3bt3S29bcmbnd3NxEZGSkuHTpkkhMTBQHDx6UJs8EID788EOD7c2dBFH3fTKHNrYwkxEKoZmZu06dOtLsvMOGDRO///67uHv3rvT6bd68WQwbNkz4+voavRzBpEmTpOdr2bKlWLdunTT79I0bN8Tvv/8uZs+eLRo3bmz0Eh9CCHHlyhVRvXp1qZzmzZuLpUuXirNnz4qEhARx48YNcfjwYTF9+nQRGBhodBI9U6/1pUuXpBmya9SoIX766Sdx69YtER8fL1auXCn8/f2Fh4eHqFatmtHXderUqcLPz0+MGDFCbN++Xdq/q1evinXr1kmvoZOTkzhx4oS0na0+l0IUfsLJuLg44efnJz3/xx9/LI4cOSLu378vEhISxLlz58S6devEoEGDhJeXl9i4caPe9pZ+Zk1JS0sTLVq00JskVXe2cWO3vLw8vTJOnDghPDw8xMCBA8WGDRvE+fPnRWJiokhISBD//POP+OKLL0TlypWl55g8eXKB6mrufg8aNEgAEM8995zBOt2ZucuUKSO++uorce3aNXH//n2xc+dOaTZ7AOKLL74w2D4vL0/6fACamblPnTolEhMTxdGjR0WfPn2kyTWNfVe0Xn/9damM8PBw8fPPP4v4+Hjx8OFDERcXJ/bs2SOmTJkiateubXCZncDAQNG8eXMxb948cfjwYXH79m2RkJAgTp8+LWbMmCFcXFxKzISUTJRKON0vdX43hUIhnn/+eXHp0qV8y12+fLlwd3fPt8wyZcoY3f7kyZPSAU/uZsm13rQ/SqZu7u7uYvPmzQZlFtW13hwxURJCc2kI3R9dU7eoqCiD7fPy8sSnn34qJSKmbqGhobL1SEhIEL169TJ5zTjtzdfXVyxZskRv+/xe67lz58qW5+bmJjZt2iT7WdBeM8vUTalUikWLFultZ8vPZWETJSGEuHjxosnrhenenpzN2pqJkjnXsHvy9uR15E6cOGHWds7OznrXOrSUNRIlIcy/1tuTCaHWvXv3RP369WW3HTBgQL7XesvOzhbDhw8363V75ZVX9LbV/lNh6hYYGCguXLiQ72vq6Myf+phKFO1VnsuWLYv69eujadOmePXVV6XB1vkZMmQIXnjhBSxZsgQ7d+7ExYsX8fDhQ+lq6o0aNUKnTp3Qs2dPo9s/88wzOHfuHL799lts3rwZp0+fRnJyMsqVKwd/f3906tQJr732mtn7M2fOHHTq1Al79+7FP//8g9u3byMxMRHu7u6oUaMGOnXqhOHDh6Nq1apml6lVt25dnD17FosWLcLmzZtx/vx5pKeno3z58mjZsiXefvttaexWceTn54cDBw5gx44d+PHHH/Hnn3/i7t27yMnJQZkyZVCrVi20bt0aL7/8sjRAU5dCocDkyZPRv39/fP3119i7dy+uXr2K1NRUeHp6omrVqmjcuDGef/55o7Oba5UrVw4bN27EyZMnsW7dOuzbtw/Xr19HUlISXF1dUbFiRTRp0gQvvvgievbsafEp6R999BHq1KmD+fPnIyYmBllZWahUqRI6dOiA0aNHo379+vjqq6+Mbjtq1CjUr18fe/fuRUxMDG7duoX79+/DxcUFgYGBaNeuHYYNG2Ywb40tP5fWULNmTZw4cQLr16/Hpk2bcPToUdy/fx9CCPj6+iIkJARt27ZFz549pUlHHVXdunXxyy+/4MCBAzh06BBu3LiBe/fuIScnB6VLl0bt2rXRrl07DB482GbzPlni2WefxcWLFxEVFYXt27fj6tWryMrKQsWKFdG2bVsMHz5cGkdkTPny5XH06FH873//w9q1a3HlyhW4ubmhbt26eOuttzB48GC9M46NcXZ2xldffYW3334bS5cuxW+//Ya4uDikp6fDy8sLQUFBaNq0KcLDww0m19y5cyd27dqFffv24eLFi7h9+zZSU1NRunRp1K1bFy+//DLeffddeHp6WuPlsiuFEE9M8kBEREREADiYm4iIiEgWEyUiIiIiGUyUiIiIiGQwUSIiIiKSwUSJiIiISAYTJSIiIiIZnEepkPLy8nDr1i14eXlBoVDYuzpERERkBiEEUlNTUblyZekiwcYwUSqkW7duISAgwN7VICIiogKIj4+Hv7+/7HomSoXk5eUFQPNCe3t727k2RERU3KnVauzatQtdunSBs7Oz1WJtWbYjxFoqJSUFAQEB0nFcDhOlQtJ2t3l7ezNRIiKiQlOr1fDw8IC3t7dZiYS5sbYs2xFiCyq/YTMczE1EREQkg4kSERERkQwmSkREREQymCgRERERyWCiRERERCSDiRIRERGRDCZKRERERDKYKBERERHJYKJEREREJIOJEhEREZEMJkoOZNo0IDLS+LrISM16IiIqeSz5/bdVbEHinwZMlByIUglMmWL4IY2M1CxXKu1TLyIisi1Lfv+tEfvZZ05Gjys8DhniRXEdSESE5u+UKY8faz+cM2Y8Xk9ERCWL7u9/bq4TQkM1ycz06Ya//wWNBYCJE4H162th7Vql0eOKJWU/LRRCCGHvShRnKSkp8PHxQXJyMry9va1S5uTJwMyZACAAKKBSpEGlSLdK2URE5LjShCfShArm/P7bKlY33gm5yIMSM7w+R4R3lGy8AJCZmQk3Nzco8tnHgsS6VqsGp2PH8om2jLnHbyZKhWSLROnBA6BsWasURUREVCgK5CEF3lDBfv+wiypVoLhxw6plmnv8dviut40bN2LRokU4efIksrOzUaNGDbz++usYPXo0nJ2dLSorIyMDUVFRWL9+PS5dugQnJyeEhIRg4MCBGDZsGJQO0vk6b572nibz91Eko7RTih1rREREReFhnjeShQ/M+f23Vax+PCDghDpOF3DKrwvKOCUbjbd5i5KfX76xNiMc2MiRIwUAUapUKdGlSxfRs2dPUbp0aQFAtGnTRmRkZJhdVmJiomjYsKEAILy8vETnzp1F165dpfI6d+4ssrKyLK5jcnKyACCSk5Mt3taYGTOEADR/jT0mIqKSSft7P3VqjtiyZYuYOjVH9ve/ILEzZgiRnZ0t+vU7a/K4oo0fOjRHeHpmCUDzuGJFIe7eNb5Ndna22LJli8jOzs53P20Vaylzj98Omyht3rxZABAqlUocO3ZMWn7//n3RoEEDAUCMHTvW7PJ69+4tAIj69euLuLg4afmdO3dEs2bNBAAxadIki+tpzURJLiliskREVLI9mcxokwNjv/8FjRXicbxcYvVk2fPn7xMVKuRJyVK5ckLoHEIlJTlRctjpAWZqRjNj4sSJaNy4sbTc19cXixcvBgAsXLgQycnGmwF13bp1C5s2bQIAfPXVVwgICJDW+fn54dtvvwUAzJ8/H6mpqVbbB0vl5ho/qyAiQrM8N9c+9SIiItuy5PffGrGTJ+cZPa48GR8UlIK9e3Pg7695nJgItG0LXL6sv91nnzlh7dralu10cWH1FM0Kbty4IaDpmhRXr141GhMQECAAiDVr1uRb3s8//ywACBcXF5Gbm2s0pnz58gKAWLdunUV1tXbXGxERPd1s2eJS0LKvXROibFmh1w13+rQmTtsK1a/fWbYoFZUTJ04AAMqWLYugoCCjMWFhYXqxpqSlpQEASpcuDScn47vs6+sLADhm5dMPiYiIirtq1YB//wUqVNA8vnMHeO45YNgwzZxLU6fmom/fi/atpI045FlvsbGxAICqVavKxmi7z7SxplT47529d+8e0tLSoFKp9Nbn5eXh+vXrZpWXlZWFrKws6XFKiuasAbVaDbVanW9diIiITNEeS8w5plgSW9iyy5UDTp0CGjdW4tYtJyQlAV9/rUmSxo/Pwu7d1q+zpftnCXPLdMhESTtOyNPTUzZGm+xoExVTmjdvDg8PD2RkZGDZsmUYNWqU3vrVq1cjIyPDrPJmzZqF6dOnGyzftWsXPDw88q0LERGROXbv3m2T2MKWPW9eKfTvHw5AAUCgbt1o7N6dV+hyrRVrLu1xPz8OmShZm5eXF8aOHYvIyEh8/PHHcHJyQu/evaFUKrF161ZpTia1Wi3bNaf18ccfY8yYMdLjlJQUBAQEoEuXLlabcJKIiJ5earUau3fvRufOnfOdL9CSWGuV/dlnToA0q5ECu3eHIyoqyyZ1tnT/LGFOQwvgoImSl5cXACA93cT06v+NOzI3OZk6dSru37+PJUuWYOTIkRg5cqS07tlnn0WdOnXwzTffoGw+U2K7urrC1dXVYLmzs7PV30QiInp6WXJcsfQYVNCyIyOB6dOBF14AduzQrF+2TIkqVVwRGmq7OtviGGtueQ6ZKAUGBgIA4uPjZWO067Sx+VEqlfj6668xbNgwbN26FXFxcVCpVGjXrh1eeOEFDBgwAADQoEGDQtWdiIioJNK9SHunTo8TpcaNgenTlejXrxbCw+1bR1twyEQpNDQUAJCYmIjY2FijZ77FxMQAgN4cS+Zo0KCBQTIkhMChQ4cAAJ07dy5IlYmIiEo03TmWMjIAJycgL09zmzo1F+fP2+0iIzblkNMD+Pv7o2nTpgCANWvWGKw/ePAg4uPj4erqinArpK8bNmxAXFwcWrZsiSZNmhS6PCIiopJm2rTHE1F6eAB162runzkDjB2bh379LtitbrbkkIkSAEyaNAkAMHv2bBw/flxanpiYiGHDhgEAhg8fDh8fH2nd5s2bERISgo4dOxqUd+vWLaNdedu3b8c777wDV1dXLFmyxNq7QUREVCJp2xVycoAzZ0pmaxLgoF1vANCjRw+MGDECCxYsQIsWLdCxY0d4enpi7969ePjwIVq3bo3IyEi9bZKTk3HhwgVkZmYalPf333+jZ8+eaNiwIYKCguDs7IxTp07h/PnzUKlU2LJlC5555pmi2j0iIqJirUkT4LvvNPePH1dIlzkpaRy2RQkAoqKisH79erRs2RKHDx9GdHQ0/P39MXv2bOzbtw/u7u5ml1W/fn0MHDgQGRkZ2L17N7Zt2wYhBEaPHo3z58+ja9euNtwTIiKikkV3iPDx42xRsps+ffqgT58+ZsUOHjwYgwcPNrquRo0aWLVqlfUqRkRE9BRr1OjxgO7jxxXo3t3eNbINh25RIiIiIsfk6QmEhGju//svoFaXzJSiZO4VERER2Zx2QLdarcD16yXz6hRMlIiIiKhAdGfUuXLFRz6wGGOiRERERAWimyhdvlzabvWwJYcfzE1ERESOZ9o0zWzdCgUgBHDlSmlpXWSkZt20afaqnfWwRYmIiIgsplQCn34KlCuneRwX542srMfXhFMq7Vs/a2GLEhEREVlMezmTKVM0f3NynDB2rMDSpY+vCVcSsEWJiIiICiQiAujW7fHjpUuVJSpJApgoERERUSGMGvX4vlIpSlSSBDBRIiIiokL4+efH93NzFXjiMqzFHhMlIiIiKpDISGDx4sePa9XKw5QpKFHJEhMlIiIispj27LZp0wCFQgAAfHw0A7lLUrLEs96IiIjIYrm5j89uW7gQSEgA7t9XSGOUcnPtWz9rYaJEREREFtOdTLJCBU2idPeuZvLJkjSgm11vREREVCh+fpqut0ePFEhLs3NlrIyJEhERERVKhQqP79+9a7962AITJSIiIiqUChWEdJ+JEhEREZEO3Rale/fsVw9bYKJEREREhaIdowSwRYmIiIhID8coEREREcnw83t8n4kSERERkY7y5dn1RkRERGQUB3MTERERyXBzAzw81ADYokRERERkoHTpLABMlIiIiIgMlC6dCQBISQEyM+1cGStiokRERESF5uOTLd0vSa1KTJSIiIio0LRdb0DJGtDNRImIiIgKTdv1BrBFiYiIiEiPbosSEyUiIiIiHT4+TJSIiIiIjGKLEhEREZEM3RYlDuYmIiIi0sEWJSIiIiIZ7u658PDQXByXiRIRERHRE/z8NH+ZKBERERE9oXx5TYtSYiKQk2PnylgJEyUiIiKyigoVHt+/f99+9bAmJkpERERkFdquN6DkdL8xUSIiIiKrqFBBSPeZKBERERHpYIsSERERkQztYG6g5Ew6yUSJiIiIrIItSkREREQyOEaJiIiISAZblIiIiIhklC4NODtr7jNRIiIiItKhUDyedJKDuYmIiIieoO1+u3cPyMuzb12sgYkSERERWY02UcrNBZKS7FsXa2CiRERERFZT0gZ0M1EiIiIiq9G9MC4TJSIiIiIdui1KJWFANxMlIiIishp2vRERERHJYKJEREREJIOJEhEREZEMDuYmIiIikuHrCzj9l11wMDcRERGRDqVSkywBbFEiIiIiMqAdp3T3LiCEfetSWEyUiIiIyKq0iVJWFpCSYt+6FBYTJSIiIrKqkjSgm4kSERERWVVJmp2biRIRERFZVUmaS4mJEhEREVkVEyUiIiIiGUyUiIiIiGRwMDcRERGRDA7mLkIbN25Eu3btUKZMGXh6eqJhw4aYO3cu1Gq1xWWlp6dj1qxZCAsLg7e3N5ydnVGxYkW8+OKL2Lp1qw1qT0RE9PRhi1IRGTVqFPr06YNDhw6hWbNm6Nq1K+Li4jBhwgR06NABjx49MrusxMRENGvWDJMmTcKFCxfQsmVL9OzZE1WqVMGOHTvw8ssvY+TIkTbcGyIioqeDiwtQpozmPhMlG9myZQuioqKgUqlw5MgR7Ny5E//3f/+HS5cuoUGDBjh48CAiIiLMLm/GjBk4e/YsmjRpguvXr2Pnzp1Yv349jh07hh07dqBUqVJYsGAB/vrrLxvuFRER0dNB9zImxZnDJkozZ84EAEycOBGNGzeWlvv6+mLx4sUAgIULFyI5Odms8vbt2wcAmDBhAsqWLau3Ljw8HO3btwcA/Pnnn4WuOxER0dNO2/2WlgZkZNi3LoXhkInSzZs3cfToUQBA//79Dda3adMGAQEByMrKQnR0tFllurm5mRXnq73kMRERERVYSRnQ7ZCJ0okTJwAAZcuWRVBQkNGYsLAwvdj8dOvWDQAwZ84cJCUl6a2Ljo7G/v37UbFiRXTv3r2g1SYiIqL/lJS5lErZuwLGxMbGAgCqVq0qGxMQEKAXm58JEybg77//xs6dO1GtWjW0bt0apUuXxuXLl3Hs2DG0bt0ay5cvh4+Pj8lysrKykJWVJT1O+e+yyGq1ukBn4hEREenSHkvMOaZYEmvLso3F+vo6AVACAG7ezIFaLWxaB0uZW6ZDJkqpqakAAE9PT9kYlUoF4HGikh9PT09s27YNkyZNwhdffIGdO3dK68qVK4dOnTqhSpUq+ZYza9YsTJ8+3WD5rl274OHhYVZdiIiI8rN7926bxNqybN3Yu3erAWgEANi79wyUyutFUgdzZZg5cMohEyVbuH37Nl5++WWcOnUKn376Kfr164cKFSrg7Nmz+OSTTzB9+nRs2bIFf/zxB7y8vGTL+fjjjzFmzBjpcUpKCgICAtClSxd4e3sXxa4QEVEJplarsXv3bnTu3BnOzs5Wi7Vl2cZic3IU+PprzXo/vwYID69n8/2zhLkNLQ6ZKGkTlfT0dNmYtLQ0ADA7ORk0aBCOHj2KuXPn4qOPPpKWN23aFNu3b0eTJk1w8uRJzJs3z2iLkZarqytcXV0Nljs7O1v9TSQioqeXJccVS49BtipbN1a3kyYhQQlnZ2WR1MFc5pbnkIO5AwMDAQDx8fGyMdp12lhTbt68KTXb9evXz2C9s7MzevXqBQDYs2ePhbUlIiKiJ5WUwdwOmSiFhoYC0MymLTdYOyYmBgD05liSExcXJ92Xa4HSDuJ+8ow4IiIislxJuYyJQyZK/v7+aNq0KQBgzZo1BusPHjyI+Ph4uLq6Ijw8PN/ydAdpHzlyxGiMdkZuuekIiIiIyHyenpobwETJJiZNmgQAmD17No4fPy4tT0xMxLBhwwAAw4cP1zudf/PmzQgJCUHHjh31yqpataqUeI0cORLXrl3TW//DDz9g/fr1AIxPcElERESW03a/FecJJx1yMDcA9OjRAyNGjMCCBQvQokULdOzYEZ6enti7dy8ePnyI1q1bIzIyUm+b5ORkXLhwAZmZmQblrVixAu3bt8e5c+dQp04dtGjRAr6+vjh37hz+/fdfAMCAAQPw+uuvF8n+ERERlXR+fsDVq0BSEqBWA8XxnCeHTZQAICoqCq1bt8aiRYtw+PBhqNVqVK9eHRMnTsTo0aPh4uJidln169fHmTNnMH/+fPzyyy84evQosrKyUKZMGTz//PMYMmQI+vTpY8O9ISIiero8eRkTM6YrdDgOnSgBQJ8+fcxOYAYPHozBgwfLrvfz88Ps2bMxe/ZsK9WOiIiI5Dx55ltxTJQcdowSERERFW8l4cw3JkpERERkE092vRVHTJSIiIjIJkrCpJNMlIiIiMgmmCgRERERyWCiRERERCSDg7mJiIiIZPj4ANopD4vrYG6Hn0eJiIiIip9p0wClUtP9Fh+v36K0fn0txMQ44YkLbDgktigRERGR1SmVwJQpQG6u5vH9+5r7n33mhLVr60CptG/9zMUWJSIiIrK6iAjN3ylTNH/z8oDJk4E5c5To1+8cJk+uAcDxsyW2KBEREZFNREQAoaGPH8+ZA0ydmou+fS/ar1IWYqJERERENtOly+P7zs7A5Ml59qtMATBRIiIiIps5duzxfbVaM0apOCletSUiIqJiIzIS2LPn8eOePYHp05VYv76W/SplISZKREREZHWRkZqB3D17Pl7WsaNmjNLatXWKTctS8aglERERFSu5ucCMGcCAAY+XpaZqxij163dOmjbA0XF6ACIiIrK6adM0f3W73lJTNX/79r2I8HBOD0BERERPOS+vx/dTUuxXj4JiokREREQ24+39+L62Rak4YaJERERENqPbosREiYiIiEgHu96IiIiIZKhUj++zRYmIiIhIh1IJeHpq7jNRIiIiInqCtvuNXW9ERERET9Ce+cYWJSIiIqInaFuUUlMBIexbF0sxUSIiIiKb0iZKubnAo0f2rYulmCgRERGRTRXnSSeZKBEREZFNFedJJ5koERERkU0xUSIiIiKSod/1prBfRQqAiRIRERHZFFuUiIiIiGQU5+u9MVEiIiIim9LtektLY9cbERERkYRdb0REREQy2PVGREREJEM3UUpLs189CqJQidLatWsRHByMX3/9VTbm119/RXBwMDZt2lSYpyIiIqJi6qmdHmDt2rV4+PAhOnToIBvTvn17PHjwAD/++GNhnoqIiIiKqad2jNKpU6fwzDPPwMXFRTbG1dUVDRs2xMmTJwvzVERERFRMPbWJ0p07d1ClSpV846pUqYI7d+4U5qmIiIiomHpqEyUPDw8kJibmG5eYmGiy1YmIiIhKLmdnwM1Nc/+pGqNUr149HDp0CElJSbIxSUlJOHjwIEJCQgrzVERERFSMaVuVnqoWpVdffRXp6ekYMGAAMjIyDNY/evQIb7zxBh49eoRevXoV5qmIiIioGNOe+VbcEqVShdn43XffxbfffoudO3eiVq1a6N+/v9RydP78eaxduxa3bt1C7dq1MWzYMKtUmIiIiIqf4tqiVKhEyd3dHTt37sQrr7yCY8eO4YsvvtBbL4RAaGgoNm/eDA8Pj0JVlIiIiIovbaKUna2AWl185rsuVKIEAP7+/vj777+xbds2/Prrr7h+/ToAoGrVqujatSu6d+8OhaJ4DdwiIiIi69KddPLRo0KnH0XGKjVVKBTo3r07unfvbo3iiIiIqITRnSKgOCVKxafti4iIiIot3UQpI4OJEhEREZHkqeh6Cw4OhkKhwJ49exAUFITg4GCzt1UoFLhy5YrFFSQiIqLir7h2vVlU02vXrkGhUECtVkuPzcUB3URERE+vpyJRio2NBQDp+m7ax0RERESmPBVdb9WqVTP5mIiIiMiY4tqiVKjB3EOGDMH48eOtVRciIiIqofTPenO2X0UsVKhE6YcffmD3GxEREeXrqWxRqlixIgdpExERUb6K6xilQiVKnTt3xqFDh6Sz4IiIiIiMeSpblKZNm4asrCy8/fbbSC1ulwMmIiKiIlNcE6VC1XTlypXo2rUrVq9ejR07dqBTp04IDAyEu7u7QaxCoUBERERhno6IiIiKqeLa9Vaomk6bNk0ao5SYmIj169cbxCgUCgghmCgRERE9xVxdAWdnQK1+ihKlKVOmcDA3ERERmcXLC0hKKl4XxS10ixIRERGROby9NYlScWpRKtRgbiIiIiJzaQd0PzWJklKpxNChQ/ONe/vtt1GqVPF5UYiIiMj6tIlSdnYp5OTYty7mKlSiJISAEMLsWCIiInp66Z75VlxmFSqSrreMjAw4Oxef67oQERGR9enOpcRE6T8PHz7EwYMHUalSpQJtv3HjRrRr1w5lypSBp6cnGjZsiLlz51o8G3hgYCAUCkW+txkzZhSonkRERGSabqKUkmK/eljC4oFDwcHBeo83bdqEAwcOGI3NycnBnTt3kJubi3fffdfiyo0aNQpRUVEoVaoUOnToAJVKhX379mHChAnYtm0bdu3aZXRyS2N69eqFhIQEo+uSkpKwbds2AED79u0tricRERHlT7frLS2teEwvZHGidO3aNem+QqFAWloa0tLSZONdXFzQo0cPzJw506Ln2bJlC6KioqBSqfDbb7+hcePGAICEhAR06NABBw8eREREBObNm2dWeabi5s6di23btqFWrVpo27atRfUkIiIi8xTHrjeLE6XY2FgAmsHZwcHB6NWrFz7//HOjsS4uLihfvnyBznjTJlYTJ06UkiQA8PX1xeLFi9G2bVssXLgQERER8PHxsbh8XStWrAAADBkypFDlEBERkbynouutWrVq0v1Bgwahbdu2esus4ebNmzh69CgAoH///gbr27Rpg4CAAMTHxyM6Ohr9+vUr8HMdOnQIFy5cQKlSpTBo0KACl0NERESm6Xe92a8elijUYO6VK1fapBXmxIkTAICyZcsiKCjIaExYWJhebEFpW5PCw8NRsWLFQpVFRERE8vS73kroGCVjhBD45ZdfcPjwYdy/fx/NmzeXEqj79+/jwYMHqF69OpRKpVnlabv3qlatKhsTEBCgF1sQ6enp2LBhAwCYNXEmAGRlZSErK0t6nPJf26Farbb4TDwiIqInaY8l5hxTLIm1Zdn5xc6Y4QSlEmjYUECbejx8mAu1Og+ffeaE3FxgypS8QtXBUuaWWehE6eTJk+jbty8uXboEIQQUCgXUarWUKO3evRtvvPEGtmzZgpdeesmsMlP/G+Hl6ekpG6NSqQA8TlQKYsOGDUhLS0PFihURHh5u1jazZs3C9OnTDZbv2rULHh4eBa4LERGRrt27d9sk1pZly8VeuVILa9fWQYcO1wFohuucPn0dgwblYO3aOujX7xyioy9apQ7mysjIMCuuUInSjRs30KlTJyQmJiI8PBzt2rXD+PHj9WJ69OgBZ2dn/Pzzz2YnSkVl+fLlAICBAweaPeD8448/xpgxY6THKSkpCAgIQJcuXeCt2/lKRERUAGq1Grt370bnzp3znazZklhblp1fbHg4UKtWLqZPfzymOS4uGMeOKTF1ai4mT64BoEah988S5ja0FCpRmjlzJhITE/Hll19ixIgRAGCQKHl4eKBhw4bS4GxzeP3XiZmeni4bo52SoKDJycWLF3Ho0CEAlp3t5urqCldXV4Plzs7OnH2ciIisxpLjiqXHIFuVbSp22jTg4UMgKkrz+NgxJWbMACIilABMD82xxTHW3PIKNZj7119/RUhIiJQkyQkMDMTt27fNLjcwMBAAEB8fLxujXaeNtZR2EHebNm1Qu3btApVBRERE5ps48fF9JyeBiAj71cVchUqUbt26hQYNGuQbp1AoLBpLFBoaCgBITEyUHawdExMDAHpzLJkrNzcXq1evBmD+IG4iIiIqnEWLHt/Py1MgMtJ+dTFXoRIlT09P3L9/P9+42NhYlC1b1uxy/f390bRpUwDAmjVrDNYfPHgQ8fHxcHV1NXsQtq7o6Gjcvn0bXl5e6N27t8XbExERkWUiI4FPP338uEqVPEyZAodPlgqVKDVo0ADHjh2TvYYaAFy/fh0nT55EkyZNLCp70qRJAIDZs2fj+PHj0vLExEQMGzYMADB8+HC9Wbk3b96MkJAQdOzY0WTZ2m631157zeSZdURERFR4kZHAlCnAjBmAp6cAAPj4KDBjBhw+WSpUojRgwACkpqbirbfeMnqaXXZ2NoYNGwa1Wo0BAwZYVHaPHj0wYsQIpKWloUWLFujWrRt69eqFGjVq4PTp02jdujUin3hlk5OTceHCBVy5ckW23Hv37mHHjh0A2O1GRERUFHJz8d/AbeC/2X2Qnq55PGOGZr2jKtRZb2+++SZ+/PFHbN26FSEhIejatSsAzdxKI0aMwNatWxEXF4dOnTqhb9++FpcfFRWF1q1bY9GiRTh8+DDUajWqV6+OiRMnYvTo0XBxcbG4zO+//x5qtRr16tVD8+bNLd6eiIiILDNt2uP7KhVw9+7jS5g4+oDuQiVKSqUS27Ztw7vvvot169Zh2bJlADSXFdFeWuTVV1/FypUrC/wcffr0QZ8+fcyKHTx4MAYPHmwyZuzYsRg7dmyB60NEREQFp21RKi7Xeiv0zNwqlQo//vgjIiIiEB0djatXryIvLw8BAQHo1q0bGjVqZIVqEhERUUmgUgkACmRlKaBWA44+BaFVrvUGACEhIQgJCbFWcURERFQCaVuUAM04pdKl7VYVsxRqMDcRERGRJXRPNi8O3W8WtSgVZPC0lkKhQFZWVoG3JyIiouJPt0WpxCVKOTk5tqoHERERPQU0Y5Q0SlyiBGhahpo2bYohQ4agS5cuUCgUtqgXERERlUC6XW+pqfarh7ksSpTmzJmDlStX4u+//8bRo0cREBCAQYMG4c033yzwxWmJiIjo6VHcut4sGsz90Ucf4ezZszh48CAGDx6MpKQkREZGokaNGujUqRPWrFnDcUhEREQkq0QnSlqtWrXC8uXLcfv2bSxbtgwtWrTAvn378MYbb6BixYoYNmwYjh49au26EhERUTFX3MYoFWp6AE9PTwwZMgQHDx7E+fPnMW7cOLi5uWHJkiVo0aIF2rRpY616EhERUQlQ3KYHsNo8SrVq1cKcOXNw7tw5vPTSSxBC4OLFi9YqnoiIiEqA4tb1ZrWZuf/44w+sWLECmzZtQkZGBpycnPDss89aq3giIiIqAZ6qROn27dtYtWoVVq1ahcuXL0MIgaCgIOnitAEBAdaqJxEREZUAxW2MksWJUk5ODn7++WesWLECu3btQm5uLtzd3dG/f38MGTIE7du3t0U9iYiIqAQobmOULEqURo8ejR9//BGJiYkQQiAsLAxDhgxB//794e3tbas6EhERUQlRorveoqKioFAopASpQYMGAIAzZ86YtX2rVq0sryERERGVGCU6UdKKiYlBTEyMRdsoFApeK46IiOgpV6K73qpWrcpruxEREVGBlSoFuLjkIjtbWfISpWvXrtmoGkRERPS0cHPLKTaJktUmnCQiIiIyh5ubZigOEyUiIiKiJ7i7M1EiIiIiMsrNLRcAkJEB5ObauTL5YKJERERERUrb9QZokiVHxkSJiIiIipS26w1w/O43JkpERERUpLRdbwATJSIiIiI9ul1vqal2rIgZmCgRERFRkdJNlNiiRERERKSDY5SIiIiIZHCMEhEREZEMdr0RERERyWCiRERERCTD3Z1db0RERERGsUWJiIiISAYTJSIiIiIZ7HojIiIiksEWJSIiIiIZTJSIiIiIZDBRIiIiIpLh7Czg7CwAMFEiIiIiMqBSaf4yUSIiIiJ6AhMlIiIiIhlMlIiIiIhkqFSPxygJYefKmMBEiYiIiIqctkVJCODRI/vWxRQmSkRERFTkPD0f33fk7jcmSkRERFTktC1KABMlIiIiIj3aMUoAkJpqx4rkg4kSERERFTm2KBERERHJ4BglIiIiIhlsUSIiIiKSwUSJiIiISIan5+PB3EyUiIiIiHSwRYmIiIhIBhMlIiIiIhlMlIiIiIhkcIwSERERkQy2KBERERHJYKJEREREJIOJEhEREZEMNzfA6b8shIkSERERkQ6F4nGrEhMlIiIioicwUSIiIiKS4eWl+ctEiYiIiOgJui1KQpiOtRcmSkRERGQX2kQpJwfIzrZvXeQwUSIiIiK7KA5TBDBRIiIiIrvQTZRSU+1XD1McPlHauHEj2rVrhzJlysDT0xMNGzbE3LlzoVarC1zmzz//jO7du6NixYpwcXFBhQoV0KpVK8yYMcOKNSciIiJT2KJUSKNGjUKfPn1w6NAhNGvWDF27dkVcXBwmTJiADh064NGjRxaVl52djT59+qBHjx7Ys2cP6tWrh169eqF+/fq4cuUKFixYYKM9ISIioicVh0SplL0rIGfLli2IioqCSqXCb7/9hsaNGwMAEhIS0KFDBxw8eBARERGYN2+e2WW+/fbb2LhxI3r06IFvv/0Wvr6+0rq8vDz8/fffVt8PIiIiMq44JEoO26I0c+ZMAMDEiROlJAkAfH19sXjxYgDAwoULkZycbFZ5e/fuxerVq1G/fn1s2LBBL0kCACcnJ7Ro0cJKtSciIqL8MFEqoJs3b+Lo0aMAgP79+xusb9OmDQICApCVlYXo6Gizyvzqq68AaLrznJ2drVdZIiIiKpDikCg5ZNfbiRMnAABly5ZFUFCQ0ZiwsDDEx8fjxIkT6Nevn8nycnNzsXfvXgDAs88+izt37mDdunW4cOECXF1dERoaildffRUq3XeMiIiIbIqJUgHFxsYCAKpWrSobExAQoBdrytWrV5H23zvw119/YdiwYdJjrY8++gjr1q1Dhw4dClptIiIisgATpQJK/W8yBU9PT9kYbetPSkpKvuUlJiZK94cOHYpWrVph3rx5CAkJwZUrVzBp0iRER0fj5ZdfxvHjx1GzZk3ZsrKyspCVlSU91j6/Wq0u1JQFREREAKRjiTnHFEtibVl2QWPd3BTQpiLJyblQq/MKXK6lzC3TIRMlaxM6F5CpUqUKdu7cCVdXVwBAw4YNsXXrVjRq1AhnzpzB7NmzsXz5ctmyZs2ahenTpxss37VrFzw8PKxfeSIieirt3r3bJrG2LNvS2LNnywJoCwA4fToW0dH/Frpcc2VkZJgV55CJktd/lxNOT0+XjdF2nXl7e5tdHgAMHjxYSpK0lEol3n33XXz44YfYs2ePybI+/vhjjBkzRnqckpKCgIAAdOnSxay6EBERmaJWq7F792507tw535OPLIm1ZdkFja1c2RmTJmmWV6gQhPDwaoXaP0uY0yMFOGiiFBgYCACIj4+XjdGu08bmV55CoYAQAsHBwUZjtMtv375tsixXV1eDRAsAnJ2deTYdERFZjSXHFUuPQbYq29LYMmUex2ZkKOHsrCx0ueYytzyHnB4gNDQUgGZskdxg7ZiYGADQm2NJjkqlQu3atQFoJqw0RrucZ74REREVjeIwmNshEyV/f380bdoUALBmzRqD9QcPHkR8fDxcXV0RHh5uVpm9e/cGANmuNW3/Z7NmzQpSZSIiIrIQE6VCmPRfp+Xs2bNx/PhxaXliYiKGDRsGABg+fDh8fHykdZs3b0ZISAg6duxoUN6IESNQpkwZREdH45tvvtFbt27dOvz4449SHBEREdme7jlQTJQs1KNHD4wYMQJpaWlo0aIFunXrhl69eqFGjRo4ffo0WrdujcjISL1tkpOTceHCBVy5csWgPF9fX6xfvx5ubm547733UL9+ffTu3RuNGzdGv379IIRARESE2S1UREREVDhOToB2JiAmSgUQFRWF9evXo2XLljh8+DCio6Ph7++P2bNnY9++fXB3d7eovM6dO+PkyZMYNGgQHj58iJ9//hlxcXEIDw/Hzp07MWPGDBvtCRERERmjPTHdURMlhzzrTVefPn3Qp08fs2IHDx6MwYMHm4ypVasWVq1aVfiKERERUaFpxyk5aqLk0C1KREREVLIxUSIiIiKSoU2UsrIAR7wSGBMlIiIishtHnyKAiRIRERHZDRMlIiIiIhlMlIiIiIhkMFEiIiIiksFEiYiIiEgGEyUiIiIiGUyUiIiIiGQwUSIiIiKSwUSJiIiISAYTJSIiIiIZTJSIiIiIZDBRIiIiIpLBRImIiIhIBhMlIiIiIhlMlIiIiIhkeHo+vu+IiVIpe1eAiIiInk7TpgFKJeDmBmRm6idKkZFAdrYTwsLsVj0AbFEiIiIiO1EqgSlTAKf/shFtohQZqVmuVNqvblpMlIiIiMguIiKAGTOAjAzN47S0x0nSjBnA5Ml59q0gmCgRERGRHUVEABUqaO7fu/c4SYqIsG+9tJgoERERkV3VqPH4vouL4yRJABMlIiIisrN79x7fz87WdL85Cp71RkRERHYTGQlcvvz48ejRmu43AJg40T510sVEiYiIiOxCO3A7LAyIidEsGzgQKFNGszw31wmhofatIxMlIiIisovcXM3A7ZSUx4lSSsrjMUrZ2farmxYTJSIiIrKLadM0fz/99PGy5GTN34gIQK3OQ3R0kVdLDwdzExERkV35+Dy+r02UHAUTJSIiIrIr3UQpJcV+9TCGiRIRERHZlbf34/tsUSIiIiLS4chdbxzMbUdqtRq5ubn2rgYRmUGpVMLZ2dne1SAqkZgokZ6UlBQkJCQgKyvL3lUhIgu4urrC19cX3rr9BERUaLpfKUcbo8REqYilpKTg5s2bUKlU8PX1hbOzMxQKhb2rRUQmCCGgVquRnJyMmzdvAgCTJSIrYosSSRISEqBSqeDv788EiagYcXd3h5eXF27cuIGEhAQmSkRW5MiJEgdzFyG1Wo2srCz4+PgwSSIqhhQKBXx8fJCVlQW1Wm3v6hCVGC4ugJub5j4TpaeYduA2B4QSFV/a7y9PxCCyLm0jraONUWKiZAdsTSIqvvj9JbINbfcbW5SIiIiInqBNlFJSACHsWxddTJSIiIjI7rSJUl4ekJZm37roYqJEREREdueocykxUSIiAxkZGdi+fTuGDx+Ohg0bwsvLCy4uLggICMBrr72GQ4cOyW4bGBgIhUKR723VqlWFrufixYul8t566y2TsceOHUPv3r3h5+cHNzc3BAUF4cMPP8S9e/cKXQ8iKjxHnSKA8ygRkYE1a9bg7bffBgBUq1YNHTt2RKlSpXDy5EmsX78eGzZsQGRkJCZPnixbRuvWrVGjRg3Z9abWmePq1asYP348FAoFRD4DGjZt2oR+/fohJycHTZs2RVBQEGJiYrBw4UJs3LgRBw8eLHR9iKhwmCgRUbHh7OyMIUOGYPjw4QgNDZWWCyEwf/58jB07Fp988gnatGmD5557zmgZb731FgYPHmyT+uXl5WHw4MFQKBQYOHAgvvvuO9nYW7duYdCgQcjJycE333yDd955B4Dm9P7Bgwfjhx9+QP/+/XHkyBGe0UZkR7pdb19/DbRsaRgTGQnk5gLTphVZtdj1RkSGBg0ahOXLl+slSYDm1PgxY8agY8eOAIDvv//eHtVDVFQU/vjjD8yZMweBgYEmY7/88ktkZGSgU6dOUpIEaC5y+/XXX8PHxwdHjx7Frl27bFxrIjJFt0Xp++81SZGuyEhgyhRAqSzaejFRIoejHXMCAN9++y2aNGkCT09PlC5dGuHh4fjrr7+s8jw5OTlYtmwZ2rVrh7Jly8LV1RVBQUF4//33ER8fbxB/4MABKBQKtGvXDhkZGZgyZQrq1KkDDw8P6WC9atUqKBQKDB48GElJSRg1ahSqV68OV1dXtGvXTu+5lyxZglatWsHHxwdubm6oWbMmRowYIV1LzNTrsnLlSrRs2VKa5f3atWtWeU3MpU2gjL1OtnbhwgVMnjwZzz33HN5///184zdv3gwA6N+/v8E6lUqF7t27AwB++ukn61aUiCyimyh1765JilSqUnjnnU5o21aJKVOAGTOAiIiirRcTJXJYY8aMwbvvvgsPDw+8/PLLCAgIwC+//IK2bdtKB7+CSk1NRefOnfH222/j2LFjeOaZZ9C9e3e4urpiyZIlCA0NxYkTJ4xum5mZiXbt2uF///sfgoKC0L17d9SsWVMvJiEhAWFhYVi9ejXq16+Pl19+Gf7+/gCArKwsdOvWDe+//z5OnDiB1q1bo0ePHsjKysJXX32FRo0a4fjx47J1//DDD/HWW2+hVKlSeOGFF9C8efMi7zK6dOkSAKBSpUpF+ry5ubkYNGgQFAoFli9fnu9+p6am4vLlywCAsLAwozHa5XLvNxEVDd1EKTMTGD4cyM5W4N49Txw54oQOHYB9+4q22w3gGCVyYEuWLMGePXvQoUMHadnnn3+O8ePH480330Tr1q1RoUKFApX93nvv4cCBA3jxxRexfPlyvXK+/PJLjB49Gn379sW5c+egfKKd98iRI3jmmWdw+fJlVKxY0Wj5O3bsQMeOHfHTTz8ZXDx16tSp2LNnD6pXr449e/ZIrVFqtRrvv/8+li9fjl69euH8+fNwcXExKHv16tU4dOgQWrRoYbBu1apVePPNNy19OTB16lRMM/PX5/Tp09ixYwcA4NVXX7X4uQrj888/x5EjRzB//nxUr14933jdlraqVasajQkICAAAxMbGWqWORFQwuj+Vu3YB/301AQAKhcC+fZp/jHQOCUWCiZKjCQsD7tyxdy0sV7EiEBNj1SLfffddvSQJAD766CNs2LABMTExWLZsGSZNmmRxuefOncPatWtRuXJlrFmzBl5eXnrrR40ahd27dyM6Ohq//PILXnzxRYMyFi5cKJskAZrB0EuXLjVIkjIzM7Fo0SIAwPz58/XG1zg7O2PBggXYvn07YmNjsWnTJqPdRePGjTOaJAGaM8kGDRokWy85jRo1MisuLS0N/fv3R05ODp5//nm89NJLsrFvvvmmyaTtwYMHKF26tNl1PHPmDKZOnYpWrVphxIgRZm2Tmpoq3ff09DQao1KpAAApjjRxC9FTSLdFCQCWL398XwhNkmSPrjcmSo7mzh1AZozK00bugD9w4EDExMTgwIEDBUqUoqOjIYRAt27dDJIkrXbt2iE6OhqHDx82SJQqVKiAtm3bmnyO0NBQBAcHGyyPiYlBWloaypYtazTJ8PDwwGuvvYaoqCjs37/faKLUq1cv2edt06YN2rRpY7JuBaVWq9G7d2+cOXMGwcHB+Q7kzm96AGOtZXJycnIwaNAgODk5YcWKFXBy4qgBopLmyUTJUTBRcjQmWikcmg3qHRQUZHL5jRs3ClTu1atXAQDLly/Hct1/WYy4f/++wbL8zrIyFaMdqC23bwCkLiW5Qd3mPL+15eTk4LXXXsOvv/6KatWqYd++fShfvrzJbSyZHmDcuHFISEgwWK6dlPKzzz7D8ePHMWfOHNSuXdvseusmwunp6fAx8kuc9t+1Ep5s/SOiomUqUXJ2FoiIUGDKFM3jomxVYqLkaKzcfVWS5TfJoJy8vDwAmu6mhg0bmoxt3ry5wTJ3d/d8n8OcmIIyVfbBgwexbNkyi8vs0aMHevToYXRdbm4uXn/9dfz0008ICAjA/v37Ua1aNYufw5RNmzbh+vXrBsu1iZJ28P62bdsQHR2tF6Mdh7Rjxw7pzMIDBw4AgF494+Li0KBBA4Pn0J65Z48ElIgeM/W/ilr9uOstN7eIKvQfJkrksGJjY42OndEeGLVnkVlKO3i3devWWLhwYUGrVyBVqlQBYHrgsLbFSxtricuXL5ucfFFOYGCg0UQpNzcXAwYMwIYNG6QkyVRrWEGZO73BwYMHZdfduXMHd54Y3+ft7Y0aNWrg8uXLiImJMZooxfz3z0njxo3NrzARWZ2np2aOJN1EqHRpgW7dzmPt2jrS9ABFfdYbO/rJYcmNgdEu152XyBLdunUDAGzduhWZmZkFKqOgwsLCoFKpkJSUhK1btxqsf/ToEdatWwcAaN++vcXlDx48GEIIi2/GznjLy8vDwIEDsW7dOilJMudMM1v4559/ZOs+depUAMDQoUOlZbpeeeUVAJrLsjwpLS0N27ZtAwD07NnTxntBRKYoFICzs/6y+vUF+va9iKlTNdnTlCmGE1HaGhMlclhff/211IWiNX/+fPz999/w8vLC0KFDC1RuaGgoXn31VcTHx6Nnz55GWzPS09Px448/4u7duwV6Djlubm744IMPAABjx47V625Sq9UYOXIk7ty5g6CgIJODtm0tLy8Pb775JtasWWP3JKmwRo0aBQ8PD+zZswfffvuttDw3NxfDhg3Dw4cP0bRpU3Tp0sWOtSQiAHB11X+sbVifPDkPM2YA7dqx641Iop0eoG3btqhSpQrOnDmD06dPQ6lUYsWKFSZPz8/PypUr8fDhQ/zyyy+oXbs2GjZsiKCgIAghcO3aNZw8eRLZ2dk4d+4c/Pz8rLhXwPTp0xETE4O9e/eiTp06aN++Pby8vPDnn38iLi4O5cqVw8aNGy06K8zaFi5ciNWrVwPQDC6PlPkXLiQkBBMnTjS6btmyZQaJrq4uXboYPavP2ipXroxVq1ahX79+eOedd7B8+XIEBgbi6NGjuHr1Kvz8/LBmzRpe543IAVSrBpw69fixv//jFuKIiKKfGgBgokQObP78+ahduza++eYbHD16FM7OzujatSsiIiLQqlWrQpXt5eWFXbt2Yf369fjhhx9w7Ngx/PPPP/D29kalSpXw+uuvo3v37jZpRXF1dcWvv/6Kb7/9FqtXr8Yff/yBrKwsBAQE4MMPP8SECRMKND7JmpKSkqT7ppKd5557TjZROnToEA4dOiS7benSpYskUQKA3r17Izg4GDNnzsQff/yBEydOoFKlSvjggw8QERFh9WSYiArmyTPfdCedtBeFKOipQwRAM0mdj48PkpOT8z29ODMzE7GxsQgKCoKbm1sR1bD40f5nz48mOSJ+j8nW1Go1oqOjER4eDucnB+0UItaWZVsr9qWXgO3bHz/euDEHzs47zN4/S5h7/OYYJSIiInIIT+YrAQH2/4eZiRIRERE5hCe73uw8CgEAxyhRMZeQkIBx48aZHf/WW2/Z7BIfRERUOE8mSvlcAKBIMFEih2PJ2KS0tDSLJlhs164dEyUiIgczbZpmssknEyUnJ2D9+lqIiXEq8vmTtJgoUbEWGBjIQd9ERMWcUqmZTPLJ6cw++8wJa9fWkSactAcmSkRERGRX2vmRtBe91Zo+XYl+/c5h8uQaAJRFXi+Ag7mJiIjIAUREAOHh+sumTs1F374X7VOh/zBRIiIiIoewcaPmmm+A5rpvkyfn2bdCKAaJ0saNG9GuXTuUKVMGnp6eaNiwIebOnQu1Wm1ROatWrYJCoTB5+/XXX220F/o4poao+OL3l8h2vvgCEAJwcQHUas0YJXtz6DFKo0aNQlRUFEqVKoUOHTpApVJh3759mDBhArZt24Zdu3bB3d3dojKrV68ue9aTrS8boVRq+lfVarXF9SYix6D9J037fSYi64iM1IxRmjFD0w2neaxEv361DLrkipLDJkpbtmxBVFQUVCoVfvvtNzRu3BiAZt6cDh064ODBg4iIiMC8efMsKrdNmzZYtWqVDWqcP2dnZ7i6uiI5ORleXl68CCdRMSOEQHJyMlxdXa1+OQWip9mTSRKg+Zubm4vp0+ugVq1cTJtmn7o5bKI0c+ZMAMDEiROlJAkAfH19sXjxYrRt2xYLFy5EREQEfJ6ceMGB+fr64ubNm7hx4wZ8fHzg7OzMhInIwQkhoFarkZycjLS0NLtftJiopMnN1U+StCZPzsPFixeRm1vLPhWDgyZKN2/exNGjRwHA6NXF27Rpg4CAAMTHxyM6Ohr9+vUr6ioWmPbCewkJCbh586ada0NElnB1dUWVKlXyvQA2EVnGVGtR374XER5uv+kBHDJROnHiBACgbNmyCAoKMhoTFhaG+Ph4nDhxwqJE6fLly/jkk09w7949qFQq1K9fH927d4evr69V6m4Ob29veHt7Q61WIzfXfpNoEZH5lEolu9uInkIOmSjFxsYCAKpWrSobExAQoBdrrkOHDuHQoUN6y9zc3DBt2jRMmDDBwpoWjrOzM394iYiIHJhDJkqpqakAAE9PT9kYlUoFAEhJSTGrzIoVK2Ly5Mno3r07goOD4erqigsXLuCrr77C999/j4kTJyI3NxeTJk0yWU5WVhaysrKkx9rnV6vVFk9ZQERE9CTtscScY4olsbYs2xFiLWVumQ6ZKNlC165d0bVrV71lYWFh+O6779CwYUOMHTsWM2bMwNChQ+Hn5ydbzqxZszB9+nSD5bt27YKHh4fV601ERE+n3bt32yTWlmU7Qqy5MjIyzIpzyETJy8sLAJCeni4bk5aWBgBWGVQ5cuRIzJo1CwkJCdi1axfeeOMN2diPP/4YY8aMkR6npKQgICAAXbp04QBPIiIqNLVajd27d6Nz5875Ds+wJNaWZTtCrKXM7ZFyyEQpMDAQABAfHy8bo12njS0MpVKJmjVrIiEhATdu3DAZ6+rqCldXV4PlHG9ERETWZMlxxdJjkK3KdoRYc5lbnv3nBjciNDQUAJCYmCg7WDsmJgYA9OZYKozExEQAj1uziIiIiBwyUfL390fTpk0BAGvWrDFYf/DgQcTHx8PV1RXhVpjX/Pjx47h4UXN14mbNmhW6PCIiIioZHDJRAiCdfTZ79mwcP35cWp6YmIhhw4YBAIYPH643K/fmzZsREhKCjh076pWVkZGBRYsWSWfT6fr999/x6quvAtBMZMlEiYiIiLQccowSAPTo0QMjRozAggUL0KJFC3Ts2BGenp7Yu3cvHj58iNatWyMyMlJvm+TkZFy4cAGZmZl6y7OzszF8+HCMHTsWoaGhqFq1KnJycnDx4kWcOXMGANCgQQNs2LChyPaPiIiIHJ/DJkoAEBUVhdatW2PRokU4fPgw1Go1qlevjokTJ2L06NFwcXExqxwPDw9EREQgJiYG58+fx7///otHjx6hTJky6NSpE3r37o3BgwebXZ4uIQQA80fPExERmaJWq5GRkYGUlBSzzgozN9aWZTtCrKW0x23tcVyOQuQXQSbduHFDmiWciIiIipf4+Hj4+/vLrmeiVEh5eXm4desWvLy8oFAo7F0dIiIq5rTz88XHx+c7P58lsbYs2xFiLSWEQGpqKipXrgwnJ/kh2w7d9VYcODk5mcxEiYiICkJ7AXVrx9qybEeItYTuCWFyHPasNyIiIiJ7Y6JEREREJIOJEhERkQNxdXXF1KlTjV4uqzCxtizbEWJthYO5iYiIiGSwRYmIiIhIBhMlIiIiIhlMlIiIiIhkMFEiIiIqwdavX49mzZrBw8MDZcuWRceOHXH79m29mAMHDkChUBjc6tevb7LsnJwcPPPMM1AoFFi3bp3RmJ9++glt2rSBr68vXF1dERwcjDFjxuDBgwcGsZs2bcIrr7yCqlWrwsPDA/Xq1cMXX3wBtVptEHv58mW89957aNy4MZydnREYGCgtDw8Ph0qlgq+vL4YNG4b09HQzXy1DnHCSiIiohPriiy/w8ccfY9y4cZgzZw7S09Px+++/G1w8XmvZsmWoV6+e9NjDw8Nk+VFRUbh//77JmKSkJLRr1w4fffQRfHx8cPr0aUyfPh0nT57E3r179WLnzZuHwMBAzJ07F35+fjh8+DA++eQTnDp1Ct99951e7L///ovt27ejWbNmEELgwYMHSE5ORocOHVC5cmVs3LgRSUlJGDNmDO7evYv/+7//M1lPWYKIiIhKnEuXLglnZ2exePHifGP3798vAIg///zT7PLj4+OFl5eX+O677wQAsXbtWrO3Xbp0qQAgrl+/rrf83r17BrGRkZECgLhz547e8tzcXOn+u+++K6pVqybmzJkj3NzcxN27d6V1//d//ycAiJiYGLPrp4tdb0RERCXQihUr4OLigqFDh9qk/FGjRqF79+549tlnLd62bNmyAGDQpVa+fHmD2CZNmgAAbt26pbfc2PXZoqOj0aFDB1SoUEFa1r17d6hUKmzfvt3iegLseiMiIiqQCxcuYNeuXTh27BiOHTuGc+fOITc3F5GRkfjkk0/Mig0MDMTDhw+RnZ2NGjVq4PXXX0d4eDj2799vVuzo0aNx9epVvbLPnj2LvLw8KZGoXr060tLSkJKSgho1aqBDhw549OiRXtkVK1YEALRs2RIA4Onpia5duyIsLAznz583qEdCQgLS0tJQt25dLFmyxOjrs3HjRixatAjHjx9HZmYmVCoVXFxckJCQgNzcXNSqVQvVq1fP93VeunQpFAoF2rdvD7Varbfvzs7OerFnz57FwIED9ZaVKlUKtWrVwrlz58x7Y59UoHYoIiKip9zIkSMFAINbZGSk2bFOTk6iS5cuomfPnqJ06dICgKhcubLZsW3atBEffPCB0XhnZ2ehVCqFQqEQAES9evVEtWrVZMuuVq2aaNmypVCpVAKAtJ2xWHd3d9GgQQOpHgDEd999Z7C/pUqVElWrVjVaTkRERL6v8YABA0zue0ZGhhDicddbqVKljL7+HTt2FJ07dy7Q+8xEiYiIqAC+/fZbMW7cOPHjjz+Kc+fOiTfeeEM2UdKN/eqrr6Rk4f3335di7t+/Lxo0aCAAiLCwMINYc29OTk5mxQ0ZMsTsMi2J1d7c3NzMitu4caNYuXKl2bG6r9PYsWOFELZNlNj1RkREVABvvfWW3mNjY2aMxTZr1ky6X7lyZem+r68vFi9ejLZt2+L06dN44YUX0LlzZ2n9iBEj8P7770uPjx07hgEDBsDFxQUHDx7E0KFDcfr0adSvXx+nTp2Cn58f7t69ixMnTuDff/+VYp999lncunUL//zzT75lA8Crr76qF+vk5IT58+ejRYsWAIC9e/di0qRJKFWqFHbt2oWRI0fi9OnTGDduHF5//XW910G3XAD48ssv8fzzzyMvL08qDwDS09PRtm1bPHr0CG+99RbGjh0LAKhSpQq8vLyk12nhwoWIiIiQtitTpgwePnxo8Po/ePAANWvWNFhuDiZKREREReTmzZs4evSo7Po2bdogICAA8fHx+OGHH/Riy5cvj5CQEOlxSEgIPv74Y8THx+Pvv//G6dOnAQA1a9bEqVOnULlyZdy9exe1a9dGo0aNpNi0tDSo1WqcPXvWrLLv3buH48ePS+vy8vIwcuRIg7rn5OSga9euyM7OBgAMGTIEQUFBejEhISF45513kJGRAQB49OgRvLy8AAA+Pj4AgKysLGkcFQBMmjTJoBzd1yk6OlpaXqdOHYOxSLm5ubh48SK6d+9u/EXPB896IyIiKiInTpwAALi4uMjGhIWFAQD27dtncWzZsmWhUqkAaJIGANJcRdrY06dPIyAgIN+ytTHapEYbO2TIEOzfv1+6rV27VtpG2ypUtmxZg+RGy9fXV7ofHBysty43NxevvfaalJiZKke7P9rXFIA0EF53bqdt27YhLS0NL7zwguy+msIWJSIioiISGxsLQHNWmbbl5UnaBOXq1asWx1atWlVaFxISgpYtW+Ktt97CrFmzkJubC0DTitO8eXPs27dPKnvjxo3IyspCkyZN4OXlhSNHjiAmJgaApqVItx5BQUFo166d9DzXrl2T7qekpBjU4/nnn0fHjh1Rr149uLq6SpNdurq6okePHnr788EHH2DLli144YUXsGPHDpQrVw5//fUXAKBu3brw9vaWYrVn6v3xxx/w8vJCRkYGKlasCHd3d3Tt2hWffvopHjx4gDFjxuDll1+WEitLMVEiIiIqIqmpqQBgcFq7Lm2LUFpamsWxnp6e0jqFQoHt27dj3LhxGDt2rJTENG3aVNpOW3aFChWwZcsWREVF4dGjR/D390ejRo3w999/S5f/MFUPLWP1aNasGX744QcpSdTOnVS5cmWDFq1ff/0VALBjxw4AwKVLl6QpC/bv36+XoCkUCgCQEikAGDx4sFR2r1694Obmht69e2PevHn51l0Ou96IiIhKqLJly2LFihVISkrChAkTADweC6Trueeew+nTp5GSkgK1Wo3Y2Fh06tTJrOcIDAzEpEmTZNdHRkbi1KlTSE1NRWpqKmrVqgUAUCqVBrHXrl2DEAKfffYZAKB169YQmjP09ZIkAChdujQAoEuXLlKM9nb69Gmkp6cjMTERS5YskRLDgmCiREREVES0A5eNXeRVS9sqoz24WxJr6uKv2lhvb2+71kPblefq6iobo62fuftjS0yUiIiIioj2CvfaAdLGxMfHA4A0iNmSWO1jU7GBgYF2rYc2+dG2CBmjrZ+5+2NLTJSIiIiKSGhoKADNKfBytIOoO3ToYHFsYmKi1NIiF9u4ceMiqYd2TNKTEhISAOjPIfUkbf1MlaO7P7bERImIiKiI+Pv7o2nTprLrDx48iPj4eLi6uuKNN94oUKz2DDi52PDw8CKpx5o1a4zGalumtGOVjNGtn1w5uvtjS0yUiIiIipDuwOdbt25J9xMTEzFs2DAAwPDhw+Hj41OgWO3Ek6Zii6Ies2fP1puoUjcWANzc3GRfI936mSpHd39sRSGEEDZ9BiIiohLo+PHjegf+K1euICEhAf7+/qhSpYq0fPPmzbh9+7Ze7MmTJ6X5hHx8fKBUKpGcnIzc3Fw0bNgQLi4u0iVRTMW2bt0ac+fOxZgxYwBozhq7e/cuAE0i4uHhIcU2bdoUX375pRRr7Xrs3r0b7u7uAICRI0diwYIFcHZ2RlhYGC5cuCDFlipVCjk5OUZfp0qVKum9xrrldOzYEZ6enti7dy8ePnxo8Jw2U6ArxBERET3l9u/fb9aFXGNjY82OHT9+vNi1a5fZsVlZWWaXfeHCBZvW40nr168Xzz77rPDw8DD7dTJGW463t7dwd3cX9evXF7Nnzzb6nLbAFiUiIiIiGRyjRERERCSDiRIRERGRDCZKRERERDKYKBERERHJYKJEREREJIOJEhEREZEMJkpEREREMpgoEREREclgokREREQkg4kSERERkQwmSkREREQymCgRUbFWuXJlKBQKXLp0CbNmzUKTJk3g7e0NNzc3hIWFYcuWLfauIhEVY7woLhEVWzdv3oS/vz9UKhVq166NEydOoFWrVqhYsSKOHTuG2NhYKBQK/PTTT+jRo4e9q0tExRBblIio2IqJiQEApKWlITk5GcePH8cff/yBjRs34sKFC+jVqxeEEJg2bZp9K0pExRYTJSIqtrSJkpeXF3bv3o2GDRtK65ydnTFr1iwAwKlTp5CRkWGXOhJR8cZEiYiKLW2iNGbMGAQGBhqsDw4ORqlSpSCEQHp6ehHXjohKAo5RIqJiq3z58khISMDly5dRvXp1g/WPHj2Ch4cHlEolsrOz4eRU8P8Nc3NzAQBKpdKidYUtm4jsiy1KRFQsXbt2DQkJCShbtqzRJAkA/vzzTwDAM888o5ck5ebm4n//+x/q168PNzc31KpVC998843etv7+/pg8eTI+/fRTVKtWDc7OzkhJSTG5Ljs7G9OnT0fNmjXh4eGB+vXrY8OGDQb1MlU2ETmWUvauABFRQWi73by9vWVjvvvuOwDAq6++Ki0TQqB3797Ys2cPJk+ejCZNmmD//v1477334Ofnhx49eiAxMRE3b97EypUr0apVKyxatAhCCJQpU0Z2naenJ7p27YoLFy5g2rRpqF69On766Sf07dsX5cuXR/v27QHAZNlE5IAEEVExNGHCBAFAuLi4iMzMTIP1MTExQqlUinLlyomHDx9KyxcuXChKlSol/vjjD734l156Sbz00ktCCCH27t0rAIg33njDoFy5dVOnThVeXl7iypUresvr1KmjF2uqbCJyPOx6I6Ji6ejRowAgdXfpiomJQffu3ZGXl4dvvvkGPj4+0rq5c+eiR48eaNGiBXJycqRb3bp1cf36dQDAyZMnoVAoMHv2bIPnNbYuLy8PixcvxtChQxEcHKwXHxwcjJs3b5rcnogcF7veiKjYEULg+PHjAIAJEyZg1qxZ2L59O+rWrYu4uDj89ddfUCgUiIqK0ut2u3DhAuLi4hAXF4dNmzYZlNumTRsAmmSmUaNGqFy5skGMsXVnz57F/fv30bVrV4P4GzduoEGDBia3JyLHxUSJiIqdy5cv4+HDhwgODsbs2bNRpUoVLF68GFu2bIG3tzd69eqF8ePHIywsTG+727dvAwA2b94Mf39/g3K144ROnjyJpk2bGn1uY+vu378PAKhYsaLB850/fx7Dhw83uT0ROS4mSkRU7GgHcmsTjg8//BAffvhhvttVqlQJAKBSqQySKC21Wo2zZ8/inXfeMXtdtWrVAGgSON1JL6dOnQovLy/06dMn37KJyDExUSKiYkc7PsnSlpmaNWuiRYsWGDJkCD755BPUrl0bDx8+xLlz55CYmIjPP/8c58+fR3Z2NkJDQw22l1sXHByMtm3bYvz48RBCwMfHB99//z3WrVuHHTt2SGfmmSqbiBwTEyUiKnaebFEyl5OTEzZv3oxPPvkEn376Ke7evYvy5cujUaNGUovUyZMn4eTkhGeeecZge1PrNmzYgJEjR+K9996DWq1Gq1atcOjQIb06mtqeiBwTZ+YmomIlLy8PPj4+yMjIQHJyMlQqlb2rREQlGBMlIiIiIhmcR4mIiIhIBhMlIiIiIhlMlIiIiIhkMFEiIiIiksFEiYiIiEgGEyUiIiIiGUyUiIiIiGQwUSIiIiKSwUSJiIiISAYTJSIiIiIZTJSIiIiIZPw/uD/kYjHW2yAAAAAASUVORK5CYII=", "text/plain": [ - "
" + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n" + ] + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlkAAAHeCAYAAACopR66AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACP2ElEQVR4nOzdd1QUVxsG8GfpHQuKCihYMYqKYouo2HuPNVGxxdhrYlfQxBaNwa6xm9hb7L3XTwIWogIKClYElCJSd74/Jruw7i4sCyvF53fOHN25d+7cLbAvc++8VyIIggAiIiIiylV6ed0BIiIiosKIQRYRERGRDjDIIiIiItIBBllEREREOsAgi4iIiEgHGGQRERER6QCDLCIiIiIdYJBFREREpAMMsoiIiIh0gEEWqbRlyxZIJBKlzcTEBLa2tnB2dkbPnj2xaNEiBAcH53V3syTr/5YtW3KtTQ8PD0gkEnh6euZam7nB0dFR5Xun6ebl5ZXXT0Er9+7dw4wZM9CgQQOUKVMGxsbGsLKyQqVKldC7d29s3boV8fHxSsc9ffpU/twvXrz4+TueyzT5XJ4/fx6dOnVCqVKlYGhoKH/+Mrr4edGVuLg47Nq1C0OHDoWrqyusra1haGiIEiVKoFmzZvj9998RFxeXaRv37t3D4sWL0bNnT1SrVg22trYwMjKClZUVXFxcMGrUKNy/f/8zPaO85+XlBYlEAkdHx7zuSoFnkNcdoIIlKSkJERERiIiIQGBgIPbt24dp06ahZcuWWLVqFSpVqpTXXaQvTHR0NEaOHIk9e/bg01XCkpOTERcXh8ePH2PPnj2YMGECvL29MWbMmDzqbd47evQoOnfurPRaFUQvX75EhQoVkJiYqFQWGRmJixcv4uLFi1i2bBkOHDiAOnXqqGxn+fLl2Lhxo9L+lJQUBAQEICAgAOvWrcPs2bMxe/bsXH8eVHjxShZl6fjx44iLi0NcXBxiYmLw7NkzXL9+Hb/99htcXV0hCALOnDkDV1dXHD58OK+7+8V78OCB/P36dJP59ttv1daZPn16HvY+e54+fYoGDRpg9+7dEAQB9erVw7p16xAQEIC3b98iPDwc165dw6xZs1C2bFm8e/cOS5cuzetu56n58+dDEARUrVoVN27cQHR0tNLno6BITk5GYmIiTE1NMWDAAOzZswdPnjxBdHQ07t+/j59++gn6+voICwtD69at8fLlS5XtFC1aFL169cLKlStx6dIlBAcHIzIyEgEBAVi7di3Kly+PtLQ0zJkzB+vXr//Mz5IKNIFIhc2bNwsABADChQsXMq27fft2wczMTAAgmJubC3fu3Pk8naRsk72nAwcOzOuu5FhiYqLg6uoqABAkEonw+++/Z1l/4cKFQsWKFRX2h4aGavxZLwxkP6vLli3L667k2IsXL4Rp06YJUVFRauts375d/v6OGjVKq/NER0cLpUuXFgAIjo6O2na3wJgzZ44AQChXrlxed6XA45UsyrHvvvtOPnfjw4cPmDx5ct52iL4Iv/76K/z9/QEA3t7eGDduXKb1jY2NMWXKFBw6dOgz9C7/SkhIAAAUKVIkbzuSC8qUKYP58+ejWLFiaut89913cHFxASBelddG0aJF8c033wAQr55GRUVp1Q59eRhkUa7o2bMnWrVqBQA4e/Ys7t69q7bukydPMH78eFSvXh1WVlYwNTVFxYoV8f333yMwMDDLcyUlJWHt2rVo27YtSpcuDWNjY9ja2sLNzQ0//fQT/Pz8lI7JaiLv9evX8d1336F8+fIwNTWFmZkZypUrh4YNG2Lq1Km4ffu20jGaTDD+8OEDFi9ejIYNG6JYsWIwNjaGnZ0devbsiVOnTqk97tPJ2KmpqfDx8UGdOnVgaWkJS0tL1K9fH+vXr8/VuTUZXyepVIq1a9eicePGKFGiBPT09JQmxaelpWHbtm3o0KEDSpcuDSMjIxQvXhweHh5Yv349UlNTszznnTt38P3336Ny5cqwsLCAubk5vvrqK0ycOBEvXrxQeUxiYiKWL18OQJzoP23aNI2fY7Vq1TSuKzvX8ePHMWLECNSoUQNWVlYwNDREyZIl0bJlS/zxxx9ITk7OtI2AgAB8//33cHZ2hrm5OUxMTGBvbw83NzeMGzcO586dU3lcbn0uL168qDS5fdCgQQo3PGT82dBk4ntiYiJWrlyJFi1aoGTJkjAyMkLJkiXRrl07lfPjZDLeVAMAz58/x4QJE1ClShWYm5tDIpHg6dOn6l9MLcjec3WfJ00YGRnJ/29sbJzt4z993pGRkfjxxx9RqVIlmJiYoHjx4mjXrh0uXLiQZVtRUVGYNWsWateujSJFisDExASOjo4YMGAA/ve//2V5fGJiIubPnw8XFxeYmZmhePHiaNq0Kf78889sPaeLFy+if//+cHJygqmpKaysrFCrVi3Mnj0b7969U3ucIAjYuXMn2rdvL/+9YW1tjYoVK6JVq1ZYvHgxwsLCstWXfCuPr6RRPpWd4UKZXbt2yY/59ddfVdbx8fERDA0N5fU+3QwMDIRNmzapPce9e/eE8uXLqz0eai5xy8o2b96sVPbrr79m2h4AoUOHDkrHNW3aNNOht3///VcoW7Zspu32799fSE5OVjo24xDW0aNHBXd3d7VtDBkyRO3rpe51UNdnWfnatWuF5s2bK51rzpw58rrh4eFC7dq1M31+9evXF96+favyXFKpVPjpp58EiUSi9ngLCwvh+PHjSseePHlSXsfb21vj569KVsOF48ePz/Lz0aBBA+Hdu3cq29+5c6dgYGCQ6fHVqlVTOi43P5cXLlzIsq2MPxuZ/bwIgiAEBAQITk5OmbbXqVMn4cOHD0rHZvzdcuvWLaFYsWJKx4aGhqo8r7Zkn2UbGxutjv/48aNQsWJFAYBQtWpVrdrI+Lz//fdfwc7OTuXrJpFIhK1bt6pt59KlS0LRokUzfe2nTp2q9viIiAjBxcVF7bEDBgzIcrgwMTFR6N+/f6Z9sLW1FW7fvq10bGpqqtClS5csP4/qvkMKGgZZpJI2QVZ4eLj8mO7duyuVr1mzRuEX8MmTJ4WXL18KkZGRwsWLF4V27doJAAQ9PT3h7NmzSseHhobKfyGbmJgIP/30k+Dr6ytERkYKL1++FM6dOydMnjxZqFu3rtKx6r40AgMDBX19fQGA4ObmJhw6dEgIDQ0V3r17Jzx9+lQ4fvy4MG7cOJVBSWZBVlRUlPyXqLGxseDt7S0EBgYKkZGRwuXLl4U2bdrI+zR69GiVz1VWXr58ecHc3FyYP3++EBgYKERHRws3btwQGjduLK9z6tQp9W+MitchqyDL3t5e0NPTEyZMmCDcuXNHiIyMFO7evSv873//EwRBEGJiYoRKlSrJv7iWLl0q/Pvvv0J0dLTw5MkTwcfHR7C2thYACM2aNRPS0tKUzjVlyhT5l8qAAQOES5cuCREREUJERIRw/PhxoX79+gIAwczMTAgICFA4dsaMGdn+fKqTVZA1Z84cYciQIcKuXbuE27dvC+Hh4cKbN2+Ef/75R5g9e7b8C69Pnz5Kx757906wsLAQAAgVK1YU/vrrL+Hx48fCu3fvhOfPnwvnzp0Tpk+fLrRt21bhuNz+XKampgpxcXFCXFyc/LmuXbtWvi8uLk5ISUmR188syAoPDxdsbGzkX8Lr1q0TgoKChOjoaOHhw4eCt7e3YGxsLP/C/lTG3y329vaCvb29sGnTJiEsLEx48+aNcOLECeH9+/fq37Bsev36tfwPu06dOml8XGpqqvD8+XPhwIEDQp06deSf1b///lurfmR83uXLlxccHR2Fbdu2CeHh4cLbt2+FgwcPCg4ODgIg/nERGRmp1Mbjx4/lnydra2vBx8dHCA0NFSIiIoSTJ08KdevWlZ9jyZIlSsdLpVLBw8NDXmfEiBHC3bt3hcjISOHWrVtCz549BQDyAFpdkNW7d28BEP8oHjdunHDr1i357+E9e/YIzs7O8kDr1atXCsdu2LBBfv4+ffoIly9fFp4/fy5ER0cLjx49Enbu3Cn07dtXWL58uVavc37DIItU0ibIEgRBMDExEQAIDRs2VNj/6tUredn06dNVHiuVSuU/vC4uLkrlsiDMyMhIuHLlito+ZPyykFH3pbF8+XIBgKCvr5/p5FlVMguyMl79OHTokFJ5Wlqa0LVrV3md+/fvK5Rn/OLX19cXLl26pNRGfHy8fDJu7969NeqzpkEWAGH16tVq2xk3bpwAQChdurTw9OlTlXX8/f3lX7b79u1TKPvnn3/kV7DWr1+v8vjk5GT5FbyOHTsqlPXr10/ez9evX2fyjLOW04nv9+7dE/T19QWJRCI8efJEoezw4cPytu/evatxm7r6XApC1lepsqojuwpRrVo1ITo6WuXxx48fl7fh6+urUJbxd0vx4sWFsLAwTZ+aVgYMGCA/34kTJ7KsX6VKFZVXVpycnLQOsARB8Xnb2dkJb968Uarzzz//yOusWbNGqVz2O8PIyEjlVaKEhAShXr168j9EP72KvH//fnn7M2fOVNnPjFeoVAVZBw8elP8xrO71fP/+vfyPsE//iOzevbsAQKhdu7bKYwsbzsmiXCWbTPvpePzatWuRmJiIcuXKwdvbW+WxEokEixYtAgDcv38f9+7dk5cFBQXhxIkTAIDJkyfD3d1dbR8MDDRP/yabM2Rubp5rE4HT0tLkc1k6dOiALl26KNXR09PDypUr5X39448/1LbXq1cvNGnSRGm/ubk5evbsCQAq5+bkRNWqVTFixAiVZR8+fMCGDRsAAHPnzkW5cuVU1qtVqxb69u0LAPjrr78UypYvXw5BENCoUSMMGzZM5fGGhob4+eefAYgTlt+/fy8vi46Olv8/rydwu7i4oHbt2hAEAWfPnlUoyzgnrUyZMhq3qYvPZW4IDQ2Vp2n5/fffUbRoUZX12rVrBw8PDwDK731GP/74IxwcHHK9nzLbtm3Dtm3bAABdunRB27ZttWqnaNGiGDt2LBo3bpwr/Zo9ezZKliyptL927dqoUaMGAOWf6YiICPlrP2zYMLi5uSkdb2pqKp+rmJiYiO3btyuUb9q0CQBga2uLWbNmqezbb7/9pjD/7FO///47AKBv375qX09ra2vMmDEDALBjxw6F+Xmyz3bp0qXVnqMwYZBFuUr2w5Rxgi0A+ZdP8+bNkZiYiPj4eJVb8eLFYWNjA0Dxl0zGL6/czLBeq1YtAEBsbCyGDRumNo9Odty/f18eEPTq1UttPTs7OzRq1AgAcOXKFbX12rVrp7asSpUqAIDXr19r0VP12rdvr7bsxo0b+PDhAwBxkrW69zI+Pl7tF4bs/WzVqlWmx3/11VcAAKlUin/++SdXn2N2REdHY9GiRfDw8JBnA884aVz2/D69caNGjRryn4VBgwbh8ePHGp1PF5/L3HDu3DkIggATExPUq1cv0/dO9hwy+wOgQ4cOOuvr9evXMXz4cABAuXLlVCYbVcXPz0+eE/Dx48fYuHEjSpQogQkTJsDV1TXTm3o0pc3P9PXr1yGVSgFk/nulfv368j98Mv5eEQQB165dAwB07NhRbSBlY2ODpk2bqixLSEjA9evXAQAtWrTI9P2X3WwQHR2NkJAQeRuyz8WJEyewfPly+e+SwopBFuWqmJgYAFC6pfrRo0cAgM2bN8vvjlO3RUZGAgDevn0rP/7JkycAADMzs1zNKt+sWTN06tQJgPhXnr29PerUqYNx48bhwIEDiI2NzXabz549k/9fFiSoI/tFlNndVJldATEzMwOQflt+bilfvrzaMtl7CQCVKlXK9L2cOHEiAMX3Mj4+Xn6Xl5eXV6bHZ/xrP2MbGT9fGa9w6cLNmzfh7OyMqVOn4tKlS4iIiEBKSorKurLPv0yFChUwatQoAGKm9UqVKqFatWr44YcfsGPHDoXnlJEuPpe5QfbeJyYmwtraOtP3TnbFQ91zBDL/nOXE/fv30bFjRyQmJqJEiRI4deoUihcvrtGxZmZmsLCwgJWVFSpUqIDBgwfjn3/+QZ06dfDs2TO0b99e5fJM2aHNz3ROf6/ExMTIf1aqVq2a6fHq2g8JCZF/9gcPHpzp+1+3bl35cRk/AxMmTICDgwOkUinGjRsHGxsbNG/eHHPmzMGFCxc0uiO5IGGQRbkmLCxMvrzFp79EPv3y0UTGpTJkXyqWlpY56KFq+/btw6JFi+Dk5ARBEODn54fly5ejR48esLW1xQ8//JCt/mfMnJ1Vf2XlmWXb1tfX1/jcuUX2i14Vbd7LpKSkHB0PKH4enJyc5P/PGPTlttjYWHTt2hVv375FiRIlsGDBAty4cQMvXrzA+/fv5ZnSZVckVX1BLF++HOvXr5d/8T148ADr1q3Dt99+izJlyqBPnz4qUwvk9ucyN+T05/hTmX3OtBUcHIxWrVrh3bt3KFKkCE6fPi2/OqQtCwsLLF68GIC4lM/u3btz1J4mP9MZh9iAnP9eyRgYWlhYZHq8uvLc+NktUqQIbt++jVGjRqFIkSJITEzEhQsXMHfuXDRv3hx2dnZYunSp/KpdQccgi3KN7DIyAHz99dcKZbIf2okTJ0IQb7jIcsuYk8nKygpA5sGItoyMjPDTTz8hJCQEQUFB2Lp1K4YMGQJbW1skJiZi3bp1aNasmcZ/YWX8BZjVX7yycl0Ej7qS8RdwbGysxu+nquNlc7M02TIOE2ecG3Pp0iWdPdd9+/bhzZs30NPTw4ULFzB16lT5AtTW1tawsLCAhYVFpp9LiUSCYcOGISAgAGFhYdi9ezdGjx6NcuXKITU1Fbt378bXX3+tdEUutz+XuUH23hUrVkzj9y23c15l5tmzZ2jRogXevHkDCwsLnDhxQj48lVMNGjSQ/1+WBPdzyunvlYw/d5oe/6mMbRw+fFjjz4Bsfp6Mra0tVq5cicjISNy+fRsrVqxAt27dYGpqioiICEyePBnjx4/PtI8FBYMsyjUZ5zzIEpPKyIYFZMN+2VWxYkUA4iX04OBgLXuYtUqVKmHAgAHYsGEDwsPD5UM9/v7+OHLkiEZtZFy5/t9//820bkBAgNIx+V3GIR5t3k9ra2v50I22n4cmTZrI5+5t3rwZaWlpWrWTlTt37gAQ51apS2KanJyMoKAgjdpzcHBAr169sGLFCoSEhGDBggUAxKvAmzdvVntcbnwuc4PsvX/37p3CzQf5wcuXL9GiRQuEh4fDxMQEhw8fVgiMciqvh7Fy+nvF2tpafhPFw4cPMz3+wYMHavugpyeGDdr+7Gakr68PNzc3jB49GgcOHEB4eDgaNmwIAFi1apV86khBxiCLcsXevXvlk5nbtm2r9IXUpk0bAOKEZ22WpGjZsqX8/1u3bs1BTzVnaGiIOXPmyB9rOixVvXp1+S+zffv2qa338uVL+UTU3Lpr6XNo2rSpPOP1rl27tGqjdevWAIADBw5kmS1dFVNTU4wZMwaAOO9k/vz5Gh+b1RdURrJhzsyCuL1792Y6JKaOnp4epkyZIr86oOnnS9vPZW6QvW+CIGDPnj2f7bxZiYiIQIsWLfDkyRMYGRlh//79aNasWa6e4/Lly/L/V6hQIVfb1sTXX38tH2bM7PfK7du35fO3Mv5ekUgk8mHtY8eOqf25i4yMVHt12NraGvXr1weg/c9+ZooXL44JEyYAEG920eUf1J8LgyzKsb/++ks+lGNhYYFff/1Vqc6oUaNgYmKCDx8+YNCgQQpzdFT59IujUqVK8juRlixZghs3bqg9Njt/cQYHB2c69p/xrzVNJ87q6+tj0KBBAIAjR47g2LFjSnWkUinGjBkj76u6NAb5kZWVlby/v//+e5bLgCQmJipM2gUgnxAfHh6OcePGZTn/QlUg8dNPP6FmzZoAxAn0slvX1UlOTsaSJUvQrVu3TOtlJLty8/DhQ5VXq168eIEpU6aoPT40NDTTIPLVq1fyu6syfr508bnMDVWqVEHHjh0BADNmzMgyYI2NjcWrV6902qd3796hdevWePToEfT19bFjx45M745VdfybN28yrRMdHS1/nw0MDNC1a9ecdFkrJUqUQOfOnQEA69evVzlkmZiYiLFjxwIATExM0L9/f4XywYMHAxDvXJSlR/nUxIkTM/3MTpo0CQBw69Yt+ZVYdQRBULrjNqs/CvLqs60zuZJtiwqdjInzjh8/Ls8KHRMTI4SFhQk3btwQli1bprCsirm5uXDs2DG1ba5fv15et1q1asLGjRuF4OBg4d27d8KrV6+EmzdvCj4+PkKTJk0ECwsLpeOfPn0qz/huamoqTJ06VfDz8xOioqKE169fC5cuXRKmTp0q1KtXT+lY2Xk/Ta44cOBAwdHRUZg6dapw+vRpISwsTIiOjhaCg4OFjRs3yrO2m5ubCy9fvlQ4VtOM7yYmJsK8efOE4OBgISoqSrh69ao8sSoAYcyYMUrHa5ogM+P7pAlZ3ZwkqhQEMeN71apV5VmfR44cKVy+fFl48+aN/PU7ePCgMHLkSMHGxkblEhnTp0+Xn69hw4bCrl275FnNnz9/Lly+fFlYuHChULt2bZXLzgiCIDx58kSoUKGCvJ369esL69evFx48eCBERkYKz58/F65fvy54e3sLjo6OKhMsZvZaBwcHyzOvV6xYUThw4IDw8uVLITw8XNi8ebNgb28vmJmZCeXKlVP5us6ZM0ewtbUVxo4dKxw9elT+/EJCQoRdu3bJX0M9PT3B399ffpyuPpeCkPNkpGFhYYKtra38/NOmTRNu3bolvH37VoiMjBQePnwo7Nq1Sxg4cKBgaWkp7N27V+H47H5mMxMfHy80aNBAIYFuxiz2qjapVKrQhr+/v2BmZiYMGDBA2LNnj/Do0SMhKipKiIyMFO7cuSMsXbpUKFOmjPwcM2bM0Kqvmj7vgQMHCgCEpk2bKpVlzPhetGhRYcWKFcLTp0+Ft2/fCqdOnZKvkgBAWLp0qdLxUqlU/vkAxIzv9+7dE6KiooTbt28LvXr1kideVfWzIvPtt9/K22jfvr3w999/C+Hh4cL79++FsLAw4ezZs8Ls2bOFKlWqKC395OjoKNSvX19YsmSJcP36deHVq1dCZGSkcP/+fWHu3LmCkZFRoUpWyiCLVMr4CyGrTSKRCG3atBGCg4OzbHfjxo2Cqalplm0WLVpU5fF3796Vf1mq27KzdqHsF1pmm6mpqXDw4EGlNj/X2oX5McgSBHG5koy/sDPbfHx8lI6XSqXCzz//LA9iMttcXV3V9iMyMlL45ptvMl0DUbbZ2NgIa9euVTg+q9d68eLFatszMTER9u3bp/azIFsDLrNNX19fWLVqlcJxuvxc5jTIEgRBCAoKynT9u4zbp1nSczPI0mRNxk+3T9dF9Pf31+g4Q0NDhbU7sys3gixB0Hztwk+DSZmIiAihevXqao/97rvvsly7MDk5WRg9erRGr1u3bt0UjpX9QZLZ5ujoKAQGBmb5mhYEmqfGJgLkq6UXK1YM1atXR926ddGjRw/5xPSsDB48GB06dMDatWtx6tQpBAUF4f379zAxMYG9vT1q1aqFli1bonv37iqPr1GjBh4+fIg//vgDBw8exP379xETE4PixYvD3t4eLVu2RJ8+fTR+PosWLULLli1x7tw53LlzB69evUJUVBRMTU1RsWJFtGzZEqNHj0bZsmU1blPmq6++woMHD7Bq1SocPHgQjx49wocPH1CiRAk0bNgQw4YNk89VK4hsbW1x8eJFHDt2DH/99Rdu3LiBN2/eIDU1FUWLFkXlypXRqFEjdOnSRT6ZNSOJRIIZM2agX79+WLNmDc6dO4eQkBDExcXB3NwcZcuWRe3atdGmTRuVWfNlihcvjr179+Lu3bvYtWsXzp8/j2fPniE6OhrGxsYoVaoU6tSpg44dO6J79+7ZThvw448/omrVqli2bBl8fX2RlJSE0qVLo3nz5pgwYQKqV6+OFStWqDx2/PjxqF69Os6dOwdfX1+8fPkSb9++hZGRERwdHeHh4YGRI0cq5SXS5ecyN1SqVAn+/v7YvXs39u3bh9u3b+Pt27cQBAE2NjZwdnZG48aN0b17d3lC2vzqq6++wokTJ3Dx4kVcu3YNz58/R0REBFJTU1GkSBFUqVIFHh4e8PT01Fler+xo0qQJgoKC4OPjg6NHjyIkJARJSUkoVaoUGjdujNGjR8vnTalSokQJ3L59G7/99ht27tyJJ0+ewMTEBF999RWGDh0KT09PhTu7VTE0NMSKFSswbNgwrF+/HpcuXUJYWBg+fPgAS0tLODk5oW7dumjfvr1S4tVTp07h9OnTOH/+PIKCgvDq1SvExcWhSJEi+Oqrr9ClSxcMHz4c5ubmufFy5TmJIHySjIOIiIiIcowT34mIiIh0gEEWERERkQ4wyCIiIiLSAQZZRERERDrAIIuIiIhIBxhkEREREekA82TlEalUipcvX8LS0hISiSSvu0NEREQaEAQBcXFxKFOmjHzBbHUYZOWRly9fwsHBIa+7QURERFoIDw+Hvb19pnUYZOURS0tLAOKbZGVllce9ISKiwiAlJQWnT59G69atYWhoWKjq6rptTcXGxsLBwUH+PZ4ZBll5RDZEaGVlxSCLiIhyRUpKCszMzGBlZaVREFKQ6uq67ezSZKoPJ74TERER6QCDLCIiIiIdYJBFREREpAMMsoiIiIh0gEEWERERkQ4wyCIiIiLSAQZZRERERDrAIIuIiIhIBxhkEREREekAgywiIiIiHWCQRURERKQDDLKIiIiIdIBBFhEREZEOSARBEPK6E1+i2NhYWFtbIyYmBlZWVrnbuJsb8Pp17rZJRET5ngAgMTERJiYmkBSyutq0HWNiAvOHD2FoaKhB65rJzve3Qa6dlfKP16+BFy/yuhdERPSZSQCYFtK62rRtXLx4NlrPfQyyCqNSpfK6B0RElAfywxWn/HQlK8nEJE8DHQZZhZGvb173gIiI8kBqSgpOHz+O9u3bZzlEVtDqatP2pePH0T7LVnUn3058DwwMxIoVK+Dp6QkXFxcYGBhAIpHg559/1qo9iUSi0bZt2zaF47Zs2ZLlMSdPnsyNp0xERESFSL69krVmzRr4+PjkWnsDBw5UWxYWFoYLFy5AIpGgadOmKutUqFAB7u7uKsvs7OxypY9ERERUeOTbIKt69eqYPHkyXF1dUbt2bcyfPx/bt2/Xur0tW7aoLRs5ciQuXLiAli1boly5cirruLu7Z9oGERERUUb5NsgaOnSowmM9Pd2MbCYmJmLnzp0AgCFDhujkHERERPTlybdzsj6X/fv34/379yhWrBi6du2a190hIiKiQiLfXsn6XDZt2gQA+O6772BsbKy23uPHjzFz5kxERETAwsIC1atXR+fOnWFjY/O5ukpEREQFyBcdZD19+hQXLlwAkPVQ4bVr13Dt2jWFfSYmJvDy8sKUKVN01kciIiIqmL7oIGvz5s0QBAFubm6oUaOGyjqlSpXCjBkz0LlzZ5QvXx7Gxsby9BLbt2/H1KlTkZaWhunTp2d6rqSkJCQlJckfx8bGAgBSUlKQkpKSe0+KiIi+WLLvE02+VwpaXV23ranstFdg1i709PTE1q1bMW/ePMycOTPH7UmlUjg5OSEsLAyrV6/GiBEjst3Gb7/9hkmTJsHY2BjPnj2Dra2t2rpeXl7w9vZW2r9jxw6YmZll+9xERET0+SUkJKBfv35cuzAzZ8+eRVhYGExNTdGvXz+t2hg3bhwWLFiAyMhInD59Gv3791dbd9q0aZg4caL8cWxsLBwcHNC6devcXyCaiIi+SCkpKThz5gxatWqVZUb0glZX121rSjYSpYkvNsiSTXjv0aMHrK2ttWpDX18flSpVQmRkJJ4/f55pXWNjY5UT6w0NDXP1zSciIsrOd0tBq6vrtjVpT1NfZAqH6OhoHDp0CEDOc2NFRUUBACwtLXPaLSIiIipEvsgg66+//kJSUhIqVKigdhkdTfj5+SEoKAgAUK9evdzqHhERERUChSrIWrlyJZydnTFgwIBM68mGCgcPHgyJRKK2XkJCAlatWoW4uDilssuXL6NHjx4AxCV3GGQRERFRRvl2Tpafnx9Gjhwpf/zkyRMAwLp163D06FH5/oMHD6J06dIAgMjISAQGBqJUqVJq2/X398edO3egr68PT0/PTPuQnJyM0aNHY9KkSXB1dUXZsmWRmpqKoKAgBAQEAABcXFywZ88ebZ8mERERFVL5NsiKjY3FrVu3lPY/f/5cYZJ5xtxTmpBdxWrTpg3KlCmTaV0zMzPMmjULvr6+ePToEf799198/PgRRYsWRcuWLdGzZ094enrCyMgoW30gIiKiwi/fBlkeHh7IbgovLy8veHl5ZVpnxYoVWLFihUbtGRkZYe7cudnqAxERERFQyOZkEREREeUXDLKIiIiIdIBBFhEREZEOMMgiIiIi0gEGWUREREQ6wCCLiIiISAcYZBERERHpAIMsIiIiIh1gkEVERESkAwyyiIiIiHSAQRYRERGRDjDIIiIiItIBBllEREREOsAgi4iIiEgHGGQRERER6QCDLCIiIiIdYJBFREREpAMMsoiIiIh0gEEWERERkQ4wyCIiIiLSAQZZRERERDrAIIuIiIhIBxhkEREREekAgywiIiIiHWCQRURERKQDDLKIiIiIdIBBFhEREZEOMMgiIiIi0gEGWUREREQ6wCCLiIiISAcYZBERERHpAIMsIiIiIh3It0FWYGAgVqxYAU9PT7i4uMDAwAASiQQ///yzVu15eXlBIpFkuj169Ejt8Y8fP4anpyfs7e1hbGwMe3t7eHp6IiQkRNunSERERIWYQV53QJ01a9bAx8cn19utWbMmatWqpbLM2tpa5f5r166hdevWSEhIQLVq1eDu7o6AgABs3boV+/btw9mzZ9GgQYNc7ysREREVXPk2yKpevTomT54MV1dX1K5dG/Pnz8f27dtz3G7Xrl3h5eWlcf2EhAT06tULCQkJmDZtGubPny8vmz59OhYsWIBevXohMDAQpqamOe4fERERFQ75NsgaOnSowmM9vbwZ2dyyZQtevnyJypUrKw1V/vzzz9i/fz+CgoKwbds2DB8+PE/6SERERPlPvp2TlV8cPHgQANCnTx+lQE9PTw+9e/cGABw4cOCz942IiIjyr3x7JUtX/Pz8MHXqVERHR8Pa2hqurq7o1KkTLC0tVdb39/cHALi5uaksl+2X1SMiIiICvsAg68iRIzhy5IjCPmtrayxfvhwDBgxQ2B8XF4eoqCgAQNmyZVW25+DgAAB4+/YtPnz4AHNzcx30moiIiAqaLybIqlChAubPn4927dqhXLlyAIAHDx5g4cKFOHr0KAYOHAh9fX18++238mPi4uLk/1cXPFlYWMj/Hxsbq7ZeUlISkpKSFOoCQEpKClJSUrR/YkRERP+RfZ9o8r1S0Orqum1NZac9iSAIQq6eXUc8PT2xdetWzJs3DzNnzszVtseOHYsVK1agRIkSeP78OYyMjAAAL1++hJ2dHQAgODgYFStWVDo2ODgYlStXltcvXbq0ynN4eXnB29tbaf+OHTtgZmaWW0+FiIiIdCghIQH9+vVDTEwMrKysMq37xVzJyoyXlxdWr16Nt2/f4tatW2jcuDEAKMzT+vDhg8pj4+Pj5f/P7MWeNm0aJk6cKH8cGxsLBwcHtG7dOss3iYiISBMpKSk4c+YMWrVqBUNDw0JVV9dta0o2EqUJBlkAihUrhpIlS+LVq1d4/vy5fL+lpSWKFSuG6OhohIWFoWbNmkrHhoeHAwBsbGwynY9lbGwMY2Njpf2Ghoa5+uYTERFl57uloNXVdduatKcppnAAkJaWhpiYGABQusuwdu3aAABfX1+Vx8r2y+oRERERAQyyAACHDx9GQkICJBKJUqqGbt26AQB27doFqVSqUCaVSrF7924AQPfu3T9PZ4mIiKhAKFRB1sqVK+Hs7KyUiiEsLAx//vknEhMTlY45dOiQPLv8t99+i1KlSimUe3p6okyZMggKCsKsWbMUymbNmoWgoCDY29srnZOIiIi+bPl2Tpafnx9Gjhwpf/zkyRMAwLp163D06FH5/oMHD8rv6IuMjERgYKBSoBQdHY3+/ftjxIgRcHV1hZ2dHT5+/IgHDx4gODgYANCsWTOsWbNGqR9mZmbYs2cPWrdujfnz5+Pw4cOoXr06AgICEBAQAHNzc+zdu5frFhIREZGCfBtkxcbG4tatW0r7nz9/rjA5PWPuKXUcHBwwZcoU3L59G48fP4afnx+Sk5NhY2ODjh07ol+/fujdu7fa9REbNWqEu3fvYt68eTh79iz279+PEiVKYMCAAZg9ezYqVKig/RMlIiKiQinfBlkeHh7IbgovLy8veHl5Ke0vXrw4Fi5cmKP+VKxYEVu3bs1RG0RERPTlKFRzsoiIiIjyCwZZRERERDrAIIuIiIhIBxhkEREREekAgywiIiIiHWCQRURERKQDDLKIiIiIdIBBFhEREZEOMMgiIiIi0gEGWUREREQ6wCCLiIiISAcYZBERERHpAIMsIiIiIh1gkEVERESkAwyyiIiIiHSAQRYRERGRDjDIIiIiItIBg5w28ODBA1y/fh1v375FtWrV0LlzZwCAVCpFamoqjIyMctxJIiIiooJG6ytZ4eHhaNmyJVxcXDB8+HDMnDkThw4dkpf/8ccfMDU1xblz53Kjn0REREQFilZBVnR0NJo2bYrz58+jWrVqGDFiBARBUKjTq1cv6Onp4fDhw7nSUSIiIqKCRKsga9GiRXj69CkmT56Mu3fvYuXKlUp1ihYtChcXF1y9ejXHnSQiIiIqaLQKsv7++284Ojpi4cKFkEgkauuVL18eL1++1LpzRERERAWVVkHWs2fPULt2bejpZX64kZERoqOjteoYERERUUGmVZBlYmKCuLi4LOuFhYXB2tpam1MQERERFWhaBVnOzs7w8/PDhw8f1NaJjIzE3bt3UaNGDa07R0RERFRQaRVkffPNN4iKisLEiRMhlUpV1vnxxx+RkJCA3r1756iDRERERAWRVslIR40aha1bt2LDhg34559/0L17dwDAkydP8Ntvv2Hv3r343//+h1q1asHT0zM3+0tERERUIGgVZJmYmODUqVPo2bMnrl+/Dn9/fwDA1atXcfXqVQiCgLp16+LQoUMwNDTM1Q4TERERFQRaL6tTunRpXL16FadOncKxY8cQEhICqVQKBwcHtGvXDl26dMk0vQMRERFRYZbjtQvbtGmDNm3a5EZfiIiIiAoNrdcuJCIiIiL1cnwlKy0tDVFRUUhMTFRbp2zZsjk9DREREVGBonWQdf36dXh7e+Py5ctITk5WW08ikSA1NVXb0xAREREVSFoNF54/fx7NmjXDmTNnkJSUhKJFi6Js2bIqNwcHB606FhgYiBUrVsDT0xMuLi4wMDCARCLBzz//nO22pFIprl+/jtmzZ8Pd3R3FixeHoaEhbGxs0KpVK/z1118QBEHlsVu2bIFEIsl0O3nypFbPkYiIiAovra5kzZw5EykpKRg/fjxmzpyJYsWK5Xa/sGbNGvj4+ORKWyEhIWjUqBEAoFixYnBzc0PRokUREhKCs2fP4uzZs9i1axf2798PIyMjlW1UqFAB7u7uKsvs7OxypZ9ERERUeGgVZN25cwe1atXCb7/9ltv9katevTomT54MV1dX1K5dG/Pnz8f27du1aksikaB58+b48ccf0apVK+jr68vLLl26hA4dOuDo0aNYuHAhZs+erbINd3d3bNmyRavzExER0ZdHqyDLwsICzs7Oud0XBUOHDlV4rKen/Y2QFSpUwLlz51SWNW3aFFOnTsWsWbOwbds2tUEWERERUXZoFbk0aNAAQUFBud2XPOPq6goACA8Pz+OeEBERUWGh1ZWsGTNmoEmTJtixYwf69euX23367IKDgwGIWezVefz4MWbOnImIiAhYWFigevXq6Ny5M2xsbD5XN4mIiKgA0SrIql+/Pnbv3o2hQ4fiyJEjaNeuHcqWLat2SK9JkyY56qQuJSQkYPny5QCAHj16qK137do1XLt2TWGfiYkJvLy8MGXKlCzPk5SUhKSkJPnj2NhYAEBKSgpSUlK06ToREZEC2feJJt8rBa2urtvWVHba0zpPVlpaGszMzLBnzx7s2bNHbb38nidr5MiRCA0NRZkyZTB9+nSl8lKlSmHGjBno3LkzypcvD2NjY3l6ie3bt2Pq1KlIS0tTeWxGCxYsgLe3t9L+06dPw8zMLNeeDxER0ZkzZwptXV23nZWEhASN60oEdQmiMnH48GF0794dUqkUxYoVg5OTEywsLNTWv3DhQnZPocTT0xNbt27FvHnzMHPmzBy3BwDz5s3D7NmzYWJigrNnz8rTPGjqt99+w6RJk2BsbIxnz57B1tZWbV1VV7IcHBwQGRkJKysrrZ8DERGRTEpKCs6cOYNWrVrB0NCwUNXVdduaio2NhY2NDWJiYrL8/tbqStbPP/8MQRCwfPlyjBgxQiElQkHx22+/Yfbs2TA2NsbBgwezHWABwLhx47BgwQJERkbi9OnT6N+/v9q6xsbGMDY2VtpvaGiYq28+ERFRdr5bClpdXbetSXua0uruwgcPHqBhw4YYPXp0gQywVqxYgUmTJsHIyAj79+9H27ZttWpHX18flSpVAgA8f/48N7tIREREBZxWQZa5uTnKlSuX2335LFatWoWxY8fKA6wOHTrkqL2oqCgAgKWlZW50j4iIiAoJrYIsDw8P+Pv753ZfdG7t2rUYPXq0PMDq2LFjjtrz8/OT5wurV69ebnSRiIiICgmtgqx58+YhPDwcCxcuzO3+5MjKlSvh7OyMAQMGKJX98ccfGDlyZLYCrISEBKxatQpxcXFKZZcvX5anfHB3d2eQRURERAq0mvh+8+ZNDB48GDNmzMDhw4fRtm3bTPNkqQp6suLn54eRI0fKHz958gQAsG7dOhw9elS+/+DBg/IkopGRkQgMDESpUqUU2rpz5w6GDx8OQRBQvnx57Nu3D/v27VN53ozrEyYnJ2P06NGYNGkSXF1dUbZsWaSmpiIoKAgBAQEAABcXl0xTWBAREdGXSasgy9PTExKJBIIg4ObNm7h161am9bUJsmJjY1W2+/z5c4VJ5hnTIqjz/v17yDJVPHr0CI8ePVJbN2OQZWZmhlmzZsHX1xePHj3Cv//+i48fP6Jo0aJo2bIlevbsCU9PTxgZGWXjmREREdGXQKsga8CAAZBIJLndFwUeHh7IbgovLy8veHl55UpbAGBkZIS5c+dm+zgiIiIirYKsjFd7iIiIiEiZVhPfiYiIiChzDLKIiIiIdECj4cLLly8DEHNBmZiYyB9rqkmTJtnvGREREVEBplGQ5eHhAYlEgocPH6Jy5cryx5qQSCRITU3NUSeJiIiIChqNgqwmTZpAIpHAzMxM4TERERERqaZRkHXx4sVMHxMRERGRIk58JyIiItIBrYKswYMHY9OmTVnW27JlCwYPHqzNKYiIiIgKNK2CrC1btuDq1atZ1rt27Rq2bt2qzSmIiIiICjSdDhempaWpXTSaiIiIqDDTaQQUHBwMa2trXZ6C/uPlBcybp7ps3jyxnIiIiD4fjdcu/HSh5Dt37qhdPDk1NRX//vsvrl+/jpYtW+ash6QRfX1g9mzx/7Nmpe+fN0/cz3WuiYiIPi+NgywvLy9IJBIIggBADLLu3LmT6THm5uaYLfvmJ52SBVazZwO7dgErVwJXrgBz5ogBVsbAi4iIiHRP4yBr9uzZ8iBr7ty5qFWrFrp06aKyrpGREezt7dGmTRuULFky1zpLmZs1C1i/HnjwAGjeXNzXvj0wdmze9ouIiOhLlK0rWTKyIGvOnDm66BNpKSICSElR3Hf8OFCmDPDdd8CIEUCtWnnSNSIioi+OVhPfpVKpRnmy6PMqWRL44Qfx/xlXPUpIEK9wuboCX38N/PknkJiYN30kIiL6UjC/QiEybx7g7S3OwZJKgVGjxP1GRul1btwA+vcHHByAqVOB0NC86SsREVFhxyCrkMh4F6FskvvKleLj5GSgUyfAxSW9fmQksGgRUKEC0LEjcOwYkJaWN30nIiIqjBhkFRJpaarvIpw1S9xfuzZw9654x2HfvoChoVguCGKA1bEjULGiGHi9ffv5+09ERFTYMMgqJLy81KdpmDVLLJdIAHd3YMcOIDwc+OUXoGzZ9HpPn4pDiPb24pDi9etiEEZERETZxyDrC2VrC0yfDoSEAIcPA23bppclJ4uT4xs1EifLr18PxMfnXV+JiIgKIgZZXzh9fXG+1okTwOPHwOTJQLFi6eV37wLDhwN2dmK+rYcP866vREREBQmDLJKrUAH49Vfg+XNg61agfv30sthYYMUK4KuvxESne/cq5+QiIiKidAyySImpKTBgAHDzJuDrCwwZIu6TuXAB6NULKFdOXLbn+fO86ysREVF+pXHGd1WSkpLg6+uLFy9eIDGT7JYDBgzIyWkoD9WpA2zYIF7h2roVWLMGCAoSy169Eu9c/OUXoEsXMaN8ixaKiVCJiIi+VFoHWcuXL4eXlxdiYmKyrMsgq+ArWhQYPx4YNw44fx5YvRr4+28xdURaGnDggLhVriwGWwMHiscQERF9qbQKsrZv347x48cDAJydnVG1alVYWVnlZr8on5JIxKtVLVqIw4R//CFur16J5UFBwIQJ4p2L/foBI0eKObqIiIi+NFoFWb///jskEgk2b97Mq1RfMHt7cRmfmTPFq1qrV4vztQDg40dg40Zxq19fvLrVq5fi3C4iIqLCTKuJ7w8fPkSDBg0YYBEAMXv8N9+Iw4j//guMGQNkvLB56xbg6SkGZT/+CDx5kmddJSIi+my0CrJMTEzg6OiYy12hwuCrr4Dly4EXL8QkpjVrppdFRwNLlojL97RrBxw5wvUSiYio8NIqyHJzc0NwcHBu94UKEQsLYNgwwN9fXJ7nu+8AI6P08pMngc6dgfLlgfnzgTdv8q6vREREuqBVkDVt2jT8888/OHHiRG73Ry4wMBArVqyAp6cnXFxcYGBgAIlEgp9//jlH7Z49exbt27eHjY0NTE1N4ezsjBkzZiA+i3VjHj9+DE9PT9jb28PY2Bj29vbw9PRESEhIjvpT2EkkQMOGwPbt4kT5hQuBjBdBw8KAGTMABwdxovyVK1wvkYiICgetgqwKFSpg5syZ6NatG3766SdcvnwZT58+RVhYmMpNG2vWrMHYsWOxdetWBAQEIC0XxpWWLVuGVq1a4eTJk6hWrRo6deqEmJgYzJ8/H25uboiMjFR53LVr11CzZk1s3boVRYoUQbdu3VCkSBFs3boVNWrUwM2bN3Pcty9BiRLAlCni8j3HjgEdOqTn1EpJAXbuBJo0AWrUEPNxxcXlbX+JiIhyQqsgy9HRET///DOSk5OxdOlSNGvWDBUqVICTk5PSVr58ea06Vr16dUyePBl//fUXHj58iP79+2vVjoy/vz8mTZoEfX19HDt2DJcuXcKePXvw5MkTtGjRAoGBgfjhhx+UjktISECvXr2QkJCAadOmISAgALt27UJAQACmTZuGDx8+oFevXvj48WOO+vcl0dcH2rcHjh4VJ8FPmQLY2KSXBwSIqR/KlAFGjRIfExERFTRapXAoW7YsJDpO6z106FCFx3p6OVsBaMGCBRAEAYMGDUK7du3k+83MzLBx40aUL18e+/fvx6NHj+Ds7Cwv37JlC16+fInKlSsrDVX+/PPP2L9/P4KCgrBt2zYMHz48R338Ejk5iUOI3t7Avn1iGojr18Wy+Hjx8erV4hWukSOBbt0U53YRERHlV1pFLk+fPkVoaKjGW15LTk7GsWPHAAD9+vVTKi9XrhwaNWoEADh48KBCmexxnz59lAI9PT099O7dGwBw4MCBXO/3l8TYGPj2W+DaNXGy/PffA2Zm6eWXLwN9+gBly4p5ubQchSYiIvpsvogFooOCgpCQkABAvDNSFdl+f39/hf2yx9k9jrRXqxawbh3w8qWYDiLDhUW8eSOulejkBHTtCpw+DUiledVTIiIi9XK0QHRBIbuaVqRIEVhaWqqs4+DgoFAXAOLi4hAVFQVAHCLN7Li3b9/iw4cPMDc3V1kvKSkJSUlJ8sexsbEAgJSUFKSkpGTn6XwxzMyAH34Ahg8HLl+WYO1aPfz9twSpqRJIpWKW+b//BipUEPD991IMHChFsWJ53Wsiorwj+z7R5HuloNXVdduayk57OQqyoqKi8Mcff+DChQt48eIFAMDOzg7NmzfH0KFDUbx48Zw0n2vi/rtNTV0ABAAWFhYA0oOfjMdldqzsONmx6uotWLAA3t7eSvtPnz4Ns4zjYqRW//5Ahw4mOHOmLE6dckR0tLhGz5MnEkyZoo9ZswB39xdo1y4UlSq9z9vOEhHloTNnzhTaurpuOyuykTFNaB1knT59Gn379sX79+8hZEhs9ODBA5w9exa//vorduzYgdatW2t7ikJl2rRpmDhxovxxbGwsHBwc0Lp1ay6unU3ffQekpgJHj6Zi3To9nDsnjnonJ+vj/PmyOH++LOrUkeKHH6To2VMAY1gi+lKkpKTgzJkzaNWqFQwNDQtVXV23ramMF2OyolWQFRwcjO7duyMhIQE1atTAoEGDUKFCBQBASEgItmzZgjt37qB79+7w9/dHpUqVtDlNrpENEX748EFtHVky0owBT8ahRXXHZkximlmwZGxsDGNjY6X9hoaGufrmfykMDYGePcUtMBBYuxbYvBmIiRHL//lHD8OG6eHHH4FBg8Rhx8qV87bPRESfS3a+WwpaXV23rUl7mtJq4vvChQuRkJAALy8v3LlzB+PGjUPHjh3RsWNHjB07Fn5+fvD29kZCQgIWLVqkzSlylWydxffv3ysMAWYUHh6uUBcQg6xi/03yUZdUVXacjY1NpsORpDtVqgDLlonrJW7YANSunV72/r1YVqUK0Lo1cOiQeBWMiIhI17QKss6dO4cqVapg9uzZauvMmjULVapUwdmzZ7XuXG6pUqWKfN6Tr6+vyjqy/bUzfkNneJzd4+jzMzcHhgwBfH2BW7eAgQPF1BAyZ86IebacnIB584BXr/Kur0REVPhpFWS9fv1ao6Cidu3aeP36tTanyFVGRkbo0KEDAGDHjh1K5c+ePcP1/zJgduvWTaFM9njXrl2QfpIrQCqVYvfu3QCA7t2753q/STsSCVCvHrBli3h169dfxYWoZZ4/B2bPFnNu9e4NXLrE9RKJiCj3aRVkmZubIyIiIst6ERERn3UIbeXKlXB2dsaAAQOUyqZOnQqJRILNmzfj5MmT8v0JCQkYMmQI0tLS0KNHD4Vs7wDg6emJMmXKICgoCLNmzVIomzVrFoKCgmBvb6/ynJT3ihcHJk8GgoOBEyeAzp0BWU7Z1FRgzx7AwwOoXh1YuTJ9ThcREVFOaRVk1apVC5cvX8b9+/fV1rl37x4uXbqEWrVqadUxPz8/NGjQQL7JMravW7dOYf+rDGM+kZGRCAwMVDl/qnbt2li6dCnS0tLQvn17NGvWDL1790bFihXlw59r165VOs7MzAx79uyBmZkZ5s+fDxcXF/Tt2xcuLi6YP38+zM3NsXfvXpiammr1POnz0NMD2rYV82qFhADTpwMlS6aXP3gAjBkD2NmJk+Tv3s27vhIRUeGgVZA1bNgwpKSkoGXLlli9erXCHXbx8fFYuXIlWrVqhbS0NHz//fdadSw2Nha3bt2Sb5GRkQCA58+fK+zPmOAzKxMmTMCZM2fQpk0b3Lt3D3///TcsLCwwbdo03L59GzYZVynOoFGjRrh79y4GDBiA6Oho7N+/H9HR0RgwYADu3r2LBg0aaPUcKW+UKydmjQ8PB3buBBo3Ti/78EHMNl+rFuDuDuzYAWTjI0ZERCSnVQqHPn364MSJE9i+fTvGjBmDMWPGyBOPyjKkC4KAAQMGyNf2yy4PDw+F/Fua8PLygpeXV6Z1WrZsiZYtW2a7PxUrVsTWrVuzfRzlX0ZG4nqIffoA9+8Da9YA27eLC1MD4jqK164B48eLE+qHDwcy3HxKRESUKa3XLty6dStWr14NJycnCIKAyMhIREZGQhAElC9fHmvWrMGWLVtysatEuuPiAqxeLU6UX7UKqFYtveztW2DhQnHyfKdO4twurpdIRERZydEC0T/88AMeP36M8PBw3Lx5Ezdv3kR4eDiCg4MxfPjw3Ooj0WdjZQWMHCle2bp0SbzKJcs7JwjA0aNA+/ZAxYrA4sXAf6PYRERESnIUZMnY2dmhXr16qFevHuzs7HKjSaI8JZEATZqIc7bCwoCffwb+WwscABAaCkyZAtjbAwMGADdvMg0EEREpypUgi6gwK1UKmDFDvCvx0CGgTZv0sqQkcR5Xw4ZAnTpixvlMVm8iIqIviEYT37dt2wZATMxpaWkpf6wp5pCiwsDAAOjSRdyCg8W7EDdtAt69E8v9/YFhw8S8XAMHAiNGAJ+kXSMioi+IRkGWp6cnJBIJGjRoAEtLS/ljTTHIosKmUiVgyRJxeZ7du8VJ87dvi2UxMcDy5eLWvLk4x6tz5/S5XURE9GXQKMgaMGAAJBIJrK2tFR4TfelMTQFPT3Hz9RXTQOzYASQmiuXnz4tbmTLiVa5hw8SEp0REVPhpFGR9moqBqRmIlLm5ARs3imslbt0qBlzBwWLZy5eAt7c4gb5rV/HqVrNm4gR7IiIqnDjxnSiXFSsGTJgAPHoEnD4tBlWy9RLT0oD9+4EWLYCvvhKHFN+/z8veEhGRrmgVZDVv3hyLFy/Ost6SJUvQvHlzbU5BVODp6QGtWgEHDwJPnwKzZol3Kso8egSMGycOHw4bJk6cJyKiwkOrIOvixYt49OhRlvUCAwNx6dIlbU5BVKg4OABz5wLPnokT5T080ssSEsTUD7Vri6kgtm9Pn9NFREQFl06HC1NSUqCnxxFJIhkjI6BXL+DCBSAgABg9GrC0TC+/eVNMbmpvLyY7DQnJu74SEVHO6DQCun//vnzhaCJSVK0asGKFOCl+7VqgRo30sqgocdmeihXFZXyOHhXncxERUcGh0d2FADB48GCFx1evXlXaJ5OamooHDx7gzp076Ny5c856SFTIWVgAw4cD338PXL8u3pW4dy+QnCwu1XPihLiVKwf88AMweDBQsmRe95qIiLKicZCVMW2DRCLB48eP8fjx40yPKVOmDH755RetO0f0JZFIgEaNxO2338Rs8mvXivO4APHfadOA2bOBnj3FNBBff800EERE+ZXGQdbmzZsBAIIgYPDgwXB3d8eQIUNU1jUyMoK9vT0aNGgAQ6a5Jsq2kiWBqVOBH38Ur2KtXg2cPCle2UpJEROe7tghDjGOHAl8+614RYyIiPIPjYOsgQMHyv/v5eWFBg0aKOwjotynrw907ChuISHieokbN4pztgDg3j1xCPHHH8UJ8yNGiHO9iIgo72k18f3p06ca5ckiotxTvjywaBHw/DmwbRvQoEF6WVwcsGoVUL26mB5izx5xThcREeUd5lcgKmBMTID+/YEbNwA/PzGRqZlZevmlS0Dv3uJE+dmzxaCMiIg+P42HCzOaO3euxnUlEglmzZqlzWmIKAuursD69WK6h23bxLlbgYFi2evXwLx5wC+/AJ07i3O3WrRIX+KHiIh0S6sgy8vLCxKJBIIgKJVJMtzqJAgCgyyiz6BIEWDsWGDMGDHR6erVwKFDYm4tqVT8/6FDQKVK4rwtT0+gaNE87TIRUaGnVZA1Z84clfulUimePXuGCxcuIDw8HEOGDIG9vX2OOkhEmpNIgObNxe3FC3G5nvXrxYSnABAcDEycCEyfDvTtK17dcnPL2z4TERVWuRpkyXz8+BHDhg3DqVOn4Ofnp1XHiChn7OyAOXPEgOrwYfHq1vnzYlliIrB5s7jVrSsGW717A6amedtnIqLCRCezM0xNTbF+/XokJSVh9uzZujgFEWnI0BDo0QM4dw54+BAYNw6wtk4vv30bGDRIDMomTwayyDFMREQa0tkUWDMzM7i5ueHo0aO6OgURZZOzM/D77+JQ4h9/ALVqpZe9ewcsXSrO22rTBvj7byA1Na96SkRU8On0PiM9PT1ERETo8hREpAVzc2DoUDEFxI0bYkoII6P08tOnga5dxdxcv/wi3qlIRETZo7Mg6+XLl7h69SpsbW11dQoiyiGJRExqum2beHVr8WLAySm9PDwcmDkTcHAA+vQBLl8Wl/YhIqKsaTXx/fLly2rL4uLi8PDhQ6xatQqxsbEYMGCA1p0jos/HxkZcnmfSJODUKXGi/LFjYlCVmgrs3i1u1aqJE+W/+w6wssrrXhMR5V9aBVkeHh4K+bBUEQQBbm5umDdvnlYdI6K8oacHtGsnbk+fiikgNmwA3r4Vy//9Fxg1CpgyRRxmHDECcHHJ0y4TEeVLWgVZTZo0URtkGRkZwc7ODi1btkSvXr1gYKDVKYgoH3B0BObPF1NB7N8vXt26dk0si48H1qwRN3d38epW9+6AsXGedpmIKN/QKgK6ePFiLneDiPIzY2OgXz9xu3tXDKz+/BP48EEsv3pV3EqWFCfUf/+9uHYiEdGXjKuYEVG21KwJrF0rZpFfuRL46qv0sogI8cpX+fLieoknT4rL+hARfYnyfZC1d+9eeHh4oGjRojA3N0fNmjWxePFipKSkZKsdR0dHSCSSLLdPF7++ePFilsesXbs2N58yUYFgZSXOzQoIAC5eBHr1AmSzA6RS4MgRcV5X5crAkiVAVFSedpeI6LPTarhw5cqVGDduHA4dOoROnTqprHPkyBF07doVq1evxvDhw7Xq3Pjx4+Hj4wMDAwM0b94cFhYWOH/+PKZMmYIjR47g9OnTMNVwHZBvvvkGkZGRKsuio6Nx5MgRAECzZs1U1rG1tUXbtm1VllWpUkWjPhAVRhIJ0LSpuL16BWzcCKxbBzx/LpY/eSLetThzppgGYsQIoF498TgiosJMqyDr77//RokSJdChQwe1ddq3bw8bGxscPHhQqyDr0KFD8PHxgYWFBS5duoTatWsDACIjI9G8eXNcvXoVs2bNwpIlSzRqL7N6ixcvxpEjR1C5cmU0btxYZR1nZ2ds2bIl28+D6EtSurQYTE2dChw9Kk6UP3NGLEtKArZuFbfatcWJ8n37AmZmedtnIiJd0Wq48NGjR6hevTr09NQfrq+vDxcXFzx8+FCrjs2fPx8AMHXqVHmABQA2NjZYvXo1APGKWkxMjFbtZ7Rp0yYAwODBg3PcFhGJw4Zdu4qZ44OCgIkTgSJF0sv9/MQJ8nZ2wIQJQGBgXvWUiEh3tAqy3r59i1KlSmVZr1SpUlotq/PixQvcvn0bANCvXz+lcnd3dzg4OCApKQnHjx/PdvsZXbt2DYGBgTAwMMDAgQNz1BYRKatUSVwT8cULYNMmoE6d9LL378W1FJ2dgVatgAMHuF4iERUeWgVZlpaWePnyZZb1Xr58CTMtxgL8/f0BAMWKFYNTxjU+MnBzc1Ooqy3ZVaz27dtnGji+efMGc+fOxfDhwzFu3DisWbMGYWFhOTo30ZfEzAwYNAjw9QX+9z/A0xMwMUkvP3sW6NFDzM01d6549yIRUUGm1ZysmjVr4urVqwgPD4eDg4PKOuHh4bh+/ToaNGiQ7fZDQ0MBAGXLllVbR3ZeWV1tfPjwAXv27AEADBkyJNO6jx49wpw5cxT2GRgYYMyYMVi8eHGWSVeTkpKQlJQkfxwbGwsASElJyfadkkQFXa1aYib5hQuBbdv0sH69Hh4/FmfCv3ghJj+dN09A584CfvhBiqZNBU6UJ9KA7PtEk++VglZX121rKjvtaRVk9evXD+fPn0f37t1x5MgRpStAr1+/Ro8ePZCSkqJyuC8rcXFxAABzc3O1dSwsLACkByva2LNnD+Lj41GqVCm0b99eZR1ra2uMHz8e3bp1Q+XKlWFlZYUnT55g8+bNWLlyJZYtW4b4+HisX78+03MtWLAA3t7eSvtPnz6t1dU+osKicmVxYeq7d0vgxAkn+PqWglQqQWqqBAcOSHDggB7s7ePQtm0omjULh7k5xxOJsnJGdsdJIayr67azkpCQoHFdiSAIQnZPkJaWBg8PD1y7dg2mpqbo0KEDnJ2dAYhXfI4fP46EhAQ0bNgQly5dyvbSOvPnz8eMGTPQqFEjXL16VWWdGTNmYP78+WjdujVOnTqV3acAQJzbde3aNfz0009YtGhRto8/cOAAevToAUActqxVq5bauqquZDk4OCAyMhJWXGWXSC4sDNi4UQ+bNunhzRvFy1dmZgL69ZPi+++lyOTHjeiLlZKSgjNnzqBVq1YwNDQsVHV13bamYmNjYWNjg5iYmCy/v7W6kqWvr49jx45h0KBBOHjwIPbt2ydfy1AWs3Xp0gWbN2/Wau1CS0tLAOJwnjrx8fEAoHWAEhQUhGv/LcKm7V2F3bt3R61atXDnzh0cOXIk0yDL2NgYxioWdTM0NMzVN5+ooKtQQcwa7+UFHDwopoG4fFksS0iQYMMGfWzYoI+GDcU0EN98ozi3i4iy991S0Orqum1N2tOU1qs3W1lZYf/+/bh37x5OnjyJZ8+eARDnUbVt2xY1a9bUtmk4OjoCEOd1qSMrk9XNLtmEd3d39xwlE61atSru3LmD57LMi0SUK4yMgN69xS0gQFwvcft24L/ZBLhxQ9wmTACGDAGGDwfU3CdDRJQntA6yZGrUqIEaNWrkRl/kXF1dAQBRUVEIDQ1VeYehr68vACjk0NJUWloatm3bBiDrCe9ZifpvrRDZ1Tciyn3VqwOrVokT5f/6S7y6df++WBYZCSxaJM7rat9ezCjfti2gr5+3fSYiypdrF9rb26Nu3boAgB07diiVy+5sNDY2VjthPTPHjx/Hq1evYGlpiZ49e2rdzxcvXuDKlSsAgHr16mndDhFpxtIS+OEH4O5d4MoVMWO87Mq9IADHjgEdOwIVK4qB19u3edtfIvqy5SjIunLlCnr16gV7e3sYGxsrXBU6c+YMpk+fjtevX2vV9vTp0wEACxcuhJ+fn3x/VFQURo4cCQAYPXo0rK2t5WUHDx6Es7MzWrRokWnbsqHCPn36ZHoHIwD4+PioXPPw3r176NSpEz5+/IgKFSqgS5cumj0xIsoxiQRwdwd27ADCw4FffgEyZnx5+lRc2sfeHujfH7h+XQzCiIg+J62DrJ9//hkeHh7Yt28fXr58iZSUFGS8UdHa2hqLFi3CgQMHtGq/a9euGDt2LOLj49GgQQO0a9cO33zzDSpWrIj79++jUaNGmDdvnsIxMTExCAwMxJMnT9S2GxERgWPHjgHQbKhwzpw5KFWqFNzc3NCzZ0/07t0bbm5ucHV1hb+/P8qWLYsjR46onNRORLpnawtMnw6EhACHD4tDhTLJycCffwKNGgGurmJurv/umSEi0jmtgqwTJ05g9uzZsLOzw549e/DmzRulOvXq1UOJEiVw9OhRrTvn4+OD3bt3o2HDhrh+/TqOHz8Oe3t7LFy4EOfPn4epqWm229y+fTtSUlJQrVo11K9fP8v6M2bMQMeOHRETE4MzZ87gwIEDCA0Nxddff41ff/0VAQEBqFq1qjZPj4hykb4+0KkTcOIE8PgxMHkyUKxYevndu+LkeDs7YOxYQMtlVYmINKbVxHcfHx8YGxvjxIkTqFatmtp6NWvWRHBwsNadA4BevXqhV69eGtX19PSEp6dnpnUmTZqESZMmaXz+H3/8ET/++KPG9Yko71WoAPz6q7g8z9694kT5W7fEsthYYMUKcWvWTJwo37Vr+twuIqLcotWVrNu3b6NevXqZBlgAUKJECa3nZBER5ZSpKTBgAHDzprhm4pAh4j6ZCxeAXr2AcuXEpXyYiYWIcpNWQdaHDx8yXUxZJiYmBlKpVJtTEBHlqjp1gA0bxLURly0Tl/ORefVKvOrl6CguUn32LCfKE1HOaRVk2dra4vHjx1nWCwwMVLuANBFRXihaFBg/Hnj0SAymundPz6mVlgYcOAC0agU4OwO//w68e5eXvSWigkyrIMvd3R137tyRL0ujytGjR/H48WM0a9ZM684REemKRAK0aAHs3y+mfJg9GyhdOr08KEjMJm9nBwwdCmTIJENEpBGtgqxJkyZBIpGge/fuOHToEFJTUxXKT548iaFDh8LQ0BBjxozJlY4SEemKvT3g7Q08eyZOlM/4t+HHj8DGjeJwY4MGwNat4j4ioqxoFWTVrl0bS5cuRWRkJHr06IEiRYpAIpFg//79KFKkCDp06ICIiAgsXboUX331VW73mYhIJwwNxQWnz58H/v0XGDMGyLgG/a1bgKenGJT9+COQSUo+IiLtk5GOGzcOx48fR926dfHx40cIgoC4uDjExsbCxcUFhw8fxujRo3Ozr0REn81XXwHLl4sT5detAzKueR8dDSxZIi7f064dcOSIOJ+LiCijHC0Q3aZNG7Rp00a+kLNUKoWDgwNKZ5zYQERUgFlYAN9/DwwbJqaCWL0a2LNHzCYPACdPilvZsmKy0yFDxCz0RERaXclKlv12+U/x4sXh5uaGevXqKQVYYWFh2veOiCifkEiAhg2B7dvFfFoLF4opH2TCwoAZMwAHB6BfP3EBa6aBIPqyaRVkff311wgNDc2y3uHDh1G7dm1tTkFElG+VKAFMmSIu33PsGNChgxiEAUBKCrBzJ9CkCVCjBrBmDRAXl7f9JaK8oVWQ5efnhzp16qhd/DktLQ2TJk1Ct27d8P79+5z0j4go39LXB9q3B44eFSfBT5kC2NiklwcEACNHAmXKAKNGiY+J6MuhVZA1bdo0xMTEoGfPnhg/frxCCofw8HA0btwYv//+O4oWLYpDhw7lVl+JiPItJydxCPH5c+DPP4Gvv04vi48X53K5uABNmwK7d6fP6SKiwkurIOuXX37BsWPHUKxYMaxYsQLu7u549uwZjh49CldXV9y8eRMNGjSAv78/OnbsmNt9JiLKt4yNgW+/Ba5dA/z9xUnzZmbp5ZcvA336iBPlZ84U53IRUeGkdQqHtm3bwt/fHw0bNsT//vc/uLi4oEuXLnj37h0mT56My5cvc0kdIvqi1aolpn94+VJMB+HsnF725g3wyy/iFbCuXYHTpwEu9UpUuGgdZAGAvb09/vzzT1hbWyM+Ph4AMGjQICxevBj6ssXAiIi+cNbWYmLTBw+ACxeAnj0Bg/8S6EilwN9/A23aiItWL10q5uEiooIvR0HWqVOnUK9ePcTGxuKrr76Cvr4+Nm/ejMGDB+Mj150gIlIgkQAeHmKerWfPxKV8ypRJL3/yBJg8WVwvcdAg4PbtPOsqEeUCrYIsqVSK6dOno0OHDoiMjMSECRNw9+5dXLhwAWXKlMHWrVtRv359BAYG5nZ/iYgKhTJlxEWpnz0DDhwAWrZML0tMBLZsAerVA+rWBTZtAhIS8qyrRKQlrYKsZs2aYdGiRbCyssKhQ4ewZMkS6Ovro1GjRrhz5w5at26NgIAAuLm5Yfv27bndZyKiQsPAAOjWDThzBnj0CBg/XhxelPH1FbPI29kBEycCQUF51lUiyiatgqwrV66gTp068PPzQ+fOnRXKihcvjhMnTuDnn39GYmIiBg0alCsdJSIq7KpUAZYtE9dL3LAByJjL+f17saxKFaB1a+DQISBD9hwiyoe0CrJGjx6Na9euwTHjmhKfmD59Os6dO4dSpUpp2zcioi+Subl49crXF7h1Cxg4UEwNIXPmjHj1y8kJmDcPePUq7/pKROppFWQtX74choaGWdZr0qQJ7ty5o80piIi+eBKJOC9ryxbx6tavvwLly6eXP38uzusqWxbo3Ru4dInrJRLlJzm6u1ATNhnXmCAiIq0ULy7eeRgcDJw4AXTuDOj99xs8NVW8Y9HDA6heHVi5EoiJydPuEhE0DLK2bduG69evqyyLjY1FYmKiyrKdO3di4sSJ2veOiIgU6OkBbduKubVCQoDp04GSJdPLHzwQc3LZ2QE//ADcvZt3fSX60mkUZHl6emLDhg0qy4oWLYpRo0apLDt9+jR8fHy07x0REalVrpyYNT48HNi5E2jcOL3swwcx23ytWoC7O7BjB5CUlGddJfoi5Xi4UBAECJwEQESUZ4yMxPUQL18G7t0DRowALCzSy69dE9dTdHAApk0Dnj7Ns64SfVF0PieLiIg+HxcXYPVqcaL8qlVAtWrpZW/fAgsXipPnO3US53ZxvUQi3WGQRURUCFlZASNHAvfvi3cd9ukDyG4KFwTg6FGgfXugYkVg8WIgMjJv+0tUGDHIIiIqxCQSoEkTcc5WWBjw88/isKFMaCgwZQpgbw8MGADcvMk0EES5hUEWEdEXolQpYMYM8a7EQ4eANm3Sy5KSgO3bgYYNgTp1xIzzHz7kWVeJCgUGWUREXxgDA6BLF+DkSXEtxEmTgKJF08v9/YFhw8Q0EOPGiWsqElH2GWha8fHjx9i2bVu2yh4/fqx9z4iISOcqVQKWLBGX59m9W5w0f/u2WBYTAyxfLm7Nm4tzvDp3Tp/bRUSZ0zjIunbtGq5du6a0XyKRqC0TBAESiSRnPSQiIp0zNQU8PcXN1xdYs0bMrSXLNX3+vLiVKSNe5ZJd6SIi9TQaLixbtqxWW7ly5VC2bNkcdXDv3r3w8PBA0aJFYW5ujpo1a2Lx4sVISUnJVjtbtmyBRCLJdDt58qTa49+8eYPRo0fDyckJxsbGsLW1Rc+ePeHn55ej50dElN+4uQEbN4ppIH77TbzaJfPyJeDtLSZC/eYbMfDiRHki1TS6kvU0jzLXjR8/Hj4+PjAwMEDz5s1hYWGB8+fPY8qUKThy5AhOnz4NU1PTbLVZoUIFuLu7qyyzU/NnWVBQEBo3boyIiAiUL18eXbt2RWhoKPbt24dDhw5hz5496NatW7afHxFRflasGDBhgjgv69w5cSjx8GExt1ZaGrB/v7g5O4sJUAcMAIoUyeteE+UfGg8Xfm6HDh2Cj48PLCwscOnSJdSuXRsAEBkZiebNm+Pq1auYNWsWlixZkq123d3dsWXLFo3rC4KAPn36ICIiAv3798fmzZuhr68PAFi/fj2GDx+OAQMGIDg4GKVKlcpWX4iICgI9PaBVK3ELDwf++ANYvx5480Ysf/RIDMSmTQP69RPnbrm65m2fifKDfHt34fz58wEAU6dOlQdYAGBjY4PVq1cDAFauXIkYHS81f+LECfj7+6NIkSJYvXq1PMACgO+//x4tWrRAfHw812gkoi+CgwMwd66Yc2v3bqBp0/SyhAQx9UPt2mIqiO3b0+d0EX2J8mWQ9eLFC9z+7/aWfv36KZW7u7vDwcEBSUlJOH78uE77cvDgQQBA586dYZFxMbD/yPp34MABnfaDiCg/MTICevUCLl4EAgKA0aMBS8v08ps3xeFDe3sx2WlISJ51lSjP5Msgy9/fHwBQrFgxODk5qazj5uamUFdTjx8/xsyZM/H9999j4sSJ2LRpEyIzWU9C1r7sfOr6ERwcjA/M3EdEX6Bq1YAVK8RJ8WvXAjVqpJdFRYnL9lSsKC7jc/SoOJ+L6EuQL+dkhYaGAkCmdyY6/LcuhKyuplSlmzAxMYGXlxemTJmS7b7I+iEIAp4+fYpqGVdjzSApKQlJSUnyx7GxsQCAlJSUbN8pSUSUHxkbA4MHA4MGATduSLB2rR4OHJAgOVkCQRAXpD5xAihXTsCwYVJ4ekpRsmRe97pwkX2faPK9UtDq6rptTWWnvXwZZMXFxQEAzM3N1daRDd3JgpWslCpVCjNmzEDnzp1Rvnx5GBsbIzAwECtWrMD27dsxdepUpKWlYfr06dnqS8YhxMz6smDBAnh7eyvtP336NMzMzDR6DkREBUmfPkDbtkY4d64cTp50xNu34u+6Z88kmDlTH15eEnz99Qu0a/cUzs7RYFrF3HPmzJlCW1fXbWclISFB47r5MsjShbZt26Jt27YK+9zc3LB161bUrFkTkyZNwty5czFkyBDY2trm+vmnTZuGiRMnyh/HxsbCwcEBrVu3hpWVVa6fj4gov+jXTxwiPHkyFevW6eHUKQkEQYLUVD1cvuyAy5cd4OIi4IcfpOjbVwoV019JQykpKThz5gxatWoFwyxS8xe0urpuW1OaXtwB8mmQZfnf7MnM5jjFx8cDQK4EKOPGjcOCBQsQGRmJ06dPo3///gp9iY6OVtsXWT+y6ouxsTGMjY2V9hsaGubqm09ElB8ZGgJdu4pbSAiwbp2Y8DQqSiy/f1+CUaP0MXWqPgYMEPNuqZl9QRrIzndLQaur67Y1aU9T+XLiu6OjIwAgPDxcbR1ZmaxuTujr66PSfymNnz9/rrIvYWFhmfZDIpGgXLlyOe4LEVFhV748sGgR8Pw5sG0b0KBBellcHLBqFVC9OuDhAezZAyQn51lXiXIkXwZZrv9lsYuKilI7sd3X1xcAFHJo5UTUf39OWWa8BzlD+7LzqetHpUqVVKZ4ICIi1UxMgP79gRs3AD8/cT3EjFNUL10CevcWl/CZPVsMyogKknwZZNnb26Nu3boAgB07diiVX716FeHh4TA2Nkb79u1zfD4/Pz8EBQUBAOrVq6dQJlsu5/DhwyqHDGX96969e477QUT0pXJ1FbPIv3gB+PgAVaqkl71+DcybJwZb3boBZ86IS/sQ5Xf5MsgCIL/Lb+HChQqLMEdFRWHkyJEAgNGjR8Pa2lpedvDgQTg7O6NFixYKbSUkJGDVqlXyOwUzunz5Mnr06AFATHL6aZDVrl07uLq64v379xg5ciTSMiR4Wb9+Pc6dOwcLCwuMGzcuh8+YiIiKFAHGjgUePhTXS+zRA5AttCGVAocOAa1bi+slLlsGvHuXl70lyly+DbK6du2KsWPHIj4+Hg0aNEC7du3wzTffoGLFirh//z4aNWqEefPmKRwTExODwMBAPHnyRGF/cnIyRo8ejRIlSqBhw4bo3bs3evToARcXFzRt2hRPnz6Fi4sL9uzZo9QPiUSCnTt3okSJEti2bRsqV66MPn36oH79+hg+fDgMDAywbds2rltIRJSLJBKgeXNg3z7g2TPAywsoUya9PDgYmDhR3Dd4MKBmRgdRnsq3QRYA+Pj4YPfu3WjYsCGuX7+O48ePw97eHgsXLsT58+dhamqqUTtmZmaYNWsWmjdvjjdv3uDEiRM4fPgw3rx5g5YtW2LdunXw9fVF6dKlVR5fpUoV3Lt3D6NGjUJaWhoOHjyI0NBQdO/eHbdu3ZIPKRIRUe6zswPmzAGePhWDrubN08sSE4HNm4G6dYF69YAtW4CPH/Oqp0SK8mUKh4x69eqFXr16aVTX09MTnp6eSvuNjIwwd+7cHPWjVKlSWLlyJVauXJmjdoiISDuGhuLwYY8ewKNH4hI+W7YAMTFi+e3bYrb5iRPFq1s//CAu50OUV/L1lSwiIiJVnJ2B338XJ8r/8QdQq1Z62bt3wNKlQKVKQJs2wN9/A6mpedVT+pIxyCIiogLL3BwYOlRMAXHjhpgSwsgovfz0aTEBavnywC+/iHcqEn0uDLKIiKjAk0jEpKbbtolXtxYvBpyc0svDw4GZMwEHB3FNxcuXAUHIu/7Sl4FBFhERFSo2NsCPPwKPHwPHjwMdO0K++HRqKrB7N9C0KeDiAqxeDWRjKTqibGGQRUREhZKeHtCuHXDkiLhe4rRpQIkS6eX//guMGiXevThiBHD/ft71lQonBllERFToOToC8+eLw4Z//QU0apReFh8v3qlYowbQuDGwcyeQlJRnXaVChEEWERF9MYyNgX79gKtXgTt3gOHDxcnzMleviuVlywIzZoiJUIm0xSCLiIi+SDVrilewXr4EVq4EvvoqvSwiQrzyVb480LkzcPIk10uk7GOQRUREXzQrK3FuVkAAcPEi0KsXYPBfqm6pVJzT1a4dULkysGQJEBWVp92lAoRBFhEREcQ7EJs2Fe8+DAsD5s0D7O3Ty588Ee9atLMDBg4Ebt1iGgjKHIMsIiKiT5QuLebVCg0FDh4EWrVKL0tKEvNxNWgAuLkBGzcCCQl511fKvxhkERERqWFgIGaMP30aCAoS10UsUiS93M9PzDhvZwdMmAAEBuZVTyk/YpBFRESkgUqVxDURX7wANm0C6tRJL3v/XlxL0dlZvOp14ADXSyQGWURERNliZgYMGgT4+gL/+x/g6QmYmKSXnz0L9Ogh5uaaO1e8e5G+TAyyiIiItFS3LrB5M/D8uXjnYcWK6WUvXgBz5gDlygE9ewIXLnCi/JeGQRYREVEOFS8OTJokzsk6dQro0kVc1gcQhw337QOaNweqVQNWrABiYvK2v/R5MMgiIiLKJXp6QOvWwKFD4p2JM2cCtrbp5Q8fAmPHAmXKiNnm79zJq57S58Agi4iISAfKlhVzbYWFAbt2AU2apJclJADr1wOursDXXwN//gkkJuZdX0k3GGQRERHpkJER0Ls3cOkScP8+MHIkYGGRXn7jBtC/P+DgAEydKl4Bo8KBQRYREdFnUr06sGqVeMfh6tWAi0t6WWQksGgRUKEC0LEjcOwYkJaWd32lnGOQRURE9JlZWgIjRgB37wJXrgB9+wKGhmKZIIgBVseO4t2KixYBb9/mbX9JOwyyiIiI8ohEAri7Azt2AOHhwC+/iHO5ZJ4+FYcQ7e3FIcXr15kGoiBhkEVERJQP2NoC06cDISHA4cNA27bpZcnJ4uT4Ro3EyfLr1wPx8XnXV9IMgywiIqJ8RF8f6NQJOHECCA4GJk8GihVLL797V0z/YGcnpoMYNUq8i1GVefMAL6/P0m1SgUEWERFRPlWxIvDrr2JG+a1bgfr108tiY8XEpqtXA7Nni3cwpqSkl8+bJ+7X19fu3F5eDN5yikEWERFRPmdqCgwYANy8Ka6ZOGSIuC+jPXsAW1sD/PWXMyZP1sPs2eLaibNmaXdOfX0xSPs00FIVvDEgU41BFhERUQFSpw6wYYO4NuKyZUDlyull8fES7N1bBcuX66N8eaBSJe2TnM6aBYwfLwZUrq76WLasNpo21cfs2UC3bmI/fH3FZKuCkHlAdumS+iBs9+7KmDu3cIYjBnndASIiIsq+okXFIGjcOOD8eXHY8MABAYAEgDiBvm9fcT5X//5AUpK4nI+qK1vz5ok5uSZPBi5eBE6eFNdgfPxYLP/3Xz0ADvL6Bw+KW0ZGRmJA9ccf+ihXrg58fPRx4YLYv2LFxDJA8fy//KKHnTurYs6cwpkQjEEWERFRASaRAC1aiOkdDhyQQE9PCqk0/cpQdDTg45Ne39dXvFPRxES8AjV2rB7WrgWcnIAFC8Q7GbUhOy48XA/h4fby/T4+4tBmmTJioHXtGrBmjdgHb2999O37EDNmVASg5eSxfIxBFhERUQEnG5abMycNrq5H4e/fEd7e+qhRAwgKUhwyPHwYKF4caNVKHxcutMfHj2Jw8+lyPgYG4rqKRkbA2bOAgYEUqal66N0baNoUiIhQ3t68Ad69U+7fx4/iBohXyMqXF/8v9jcIQMXcf1HygcI5CEpERPSFkAVYc+cCM2ZIAYj/zp0L3LsHTJgArFwJ1KyZfkxKCnD8uB4+fjRUaMvREfjhB3EoMCoKaNlSDLDmzEnDvn1HMGdOGnbvFpcAmjNHXCJo715xztXDh+K5ADEgA8RgrFs3cW6YRKLYbwOD9P4WVgyyiIiICrC0NNV3Ec6aJe43MhJzafn7i0OFP/yQsZaAtm2l8PEBHj0S53GtWQN07SoO86kL3jKb5J4xILt0SUyeGhQkJk/t1Su9fmqqOCerMMv3w4V79+7FqlWrcPfuXSQnJ6NixYr49ttvMWHCBBgaGmbdwH/8/f1x8uRJnD17FgEBAYiOjoaFhQWqV6+OPn364Pvvv1fZ3sWLF9GsWbNM216zZg1+UPzUEhERfRaZpUfIGHhJJOIdgcePi48NDQWkpEhQv74UY8cqH5sxeMuYf0vWZsbFqzNeTZs6VYrjx8WATF9fXz7hHRDTTBQtmj6kKM7Jqoz27bP1lAuMfB1kjR8/Hj4+PjAwMEDz5s1hYWGB8+fPY8qUKThy5AhOnz4N008ThaiQmpqK2rVrAwAsLCxQt25d2Nra4vnz57hx4wauXr2Kbdu24dSpUyhSpIjKNmxtbdE24xoHGVSpUkXr50hERPS5KAZDqRg48DG8vatCX1/5SpimwRuQdUB2/rx41+LcueJk+zlzxP0tWkixc2dVVK6cVihzaeXbIOvQoUPw8fGBhYUFLl26JA+SIiMj0bx5c1y9ehWzZs3CkiVLNGqvTp06mDJlCjp37gxjY2P5/vv376NNmzb43//+h4kTJ2LTpk0qj3d2dsaWLVty/LyIiIjyQsYASxYM9e4dhMqVK2P2bHHyu7aJS7MKyNLSgObNxf+HhKQHWS9eSNC370OkpVVW30ABlm8HQ+fPnw8AmDp1qjzAAgAbGxusXr0aALBy5UrExMRk2ZaBgQF8fX3Rs2dPhQALAFxcXLB48WIAwK5du5CSMQQnIiIqJNTN3ZLNs0rTYaoqL6/085YvDzRuLP7/0SMJ6tSJwOzZhXMCfL68kvXixQvcvn0bANCvXz+lcnd3dzg4OCA8PBzHjx9H3759c3Q+V1dXAMDHjx8RGRmJ0qVL56g9XRMEASkpKZBKC+eHkqgw0dPTg6GhISSf3lpF9JllZ/hPl7y8FBe8vnDBQT4nTJYUtbAMHebLIMvf3x8AUKxYMTg5Oams4+bmhvDwcPj7++c4yAoODgYAGBkZoVjGdz6DN2/eYO7cuXjx4gVMTEzg7OyMDh06oGzZsjk6d3YkJycjIiICCQkJSNPlnxxElKv09fVhZmaGkiVLwsjIKK+7Q5Sn9PWBv/8WUzikpgJXrtghORlYuDB9OLOwyJdBVuh/GdEyC2AcHBwU6mpLEAT5cGHHjh2VhhNlHj16hDmyQeT/GBgYYMyYMVi8eDEMDHT7UiYkJCA8PBz6+vooWrQoTE1Noa+vz7+OifIxQRCQlpaGjx8/IiYmBk+fPoW9vT3MzMzyumtEeUZ21Ux212FsrDGGDBHzb+VkQev8KF8GWXFxcQAAc3NztXUsLCwAALGxsTk6l7e3N27cuAELCwssXLhQqdza2hrjx49Ht27dULlyZVhZWeHJkyfYvHkzVq5ciWXLliE+Ph7r16/P9DxJSUlISkqSP5b1OyUlRaN5YBERETA0NISDgwP09Qvf0gNEhZmZmRmKFCmCsLAwREREwM7OLq+7RIWU7PtEk++VvKw7dSrw6JEeduwQv89279ZHkyZSjB6dhsxOoas+Z0d22suXQdbnsm3bNsydOxd6enrYtGkTKlWqpFTH1dVVPmdLxsXFBb/99hvc3d3Ro0cP/PHHHxg5ciRq1aql9lwLFiyAt7e30v7Tp09n+Vetnp4eSpUqhTJlyuDDhw+aPTkiyndMTEzw8uVL3L9/n3MqSafOnDmT7+v26CHBjh2dIFvQ+vJlPVSokIyKFd+hQoX36NMnSOmY3bsrQyqtAkA3fdZEQkKCxnXzZZBlaWkJAJkGFPHx8QAAKysrrc6xd+9eDB48GADwxx9/oGfPntluo3v37qhVqxbu3LmDI0eOZBpkTZs2DRMnTpQ/jo2NhYODA1q3bp3lc0hKSkJYWBiKFCmiUV4wIsqfDA0NERMTg2bNmqmdmkCUEykpKThz5gxatWqVZcLuvK4rZnuXQE9PgFQqBlrv3png9u3SuH27NAwMKmPTJinmztWDbABn505xQWlZ27/8ooe0NKi8OzE7fc6O7Iyg5csgy9HREQAQHh6uto6sTFY3Ow4cOIB+/fpBKpVi3bp18mBLG1WrVsWdO3fw/PnzTOsZGxur/KVqaGiY5ZuflpYGiUQCfX196Onl26wbRJQF2TxKAwODXP2lT/QpTb5b8rLuvHmAt3f6gtbnznXEihWKU2H+/FMft27po2VLcakfIH1BaUPDili40BDe3uI8LkND9dNostNnTZ+XpvJlkCUbnouKikJoaKjKOwx9fX0BQCGHliYOHTqEPn36IC0tDWvWrMGwYcNy1NeoqCgA6VffiIiISD1VS/AsXSpFiRLiEjxlygAvX4p1g4PFTebPPyWIjbWHn58e5s7N/xPl82WQZW9vj7p16+L27dvYsWMHZsyYoVB+9epVhIeHw9jYGO2zseDRkSNH0KtXL6SmpmLNmjUYPnx4jvr54sULXLlyBQBQr169HLVFRET0JchqCZ7UVKByZWDaNODTAa0nT/SwbFkdAICbm+oAS5Zr65PQIU/k27Gn6dOnAwAWLlwIPz8/+f6oqCiMHDkSADB69GhYW1vLyw4ePAhnZ2e0aNFCqb3jx4/jm2++QWpqKtauXatxgOXj44PIyEil/ffu3UOnTp3w8eNHVKhQAV26dMnW8yMiIvoSZcz+/qlZs8RhxG+/BTw9xX2yWTKfppjz9QVcXYFXr9L3ya6S5Zeb8PNtkNW1a1eMHTsW8fHxaNCgAdq1a4dvvvkGFStWxP3799GoUSPMmzdP4ZiYmBgEBgbiyZMnCvsjIiLQvXt3JCcnw87ODtevX4enp6fK7dOAas6cOShVqhTc3NzQs2dP9O7dG25ubnB1dYW/vz/Kli2LI0eOcBIrEcS7bo4ePYrRo0ejZs2asLS0hJGRERwcHNCnTx9cu3ZN7bGOjo6QSCRZbrmxhujq1avl7Q0dOjTTuv/88w969uwJW1tbmJiYwMnJCWPGjEFERESO+0FEqs2bJ26y5X7mzgWSk2WlgrzenTuAoyPg46O8NmN+kC+HC2V8fHzQqFEjrFq1CtevX0dKSgoqVKiAqVOnYsKECRpnTk5ISJDnqHr+/Dm2bt2qtq6XlxdsbGzkj2fMmIFr167h33//xZkzZ/DhwwdYWVnh66+/RpcuXTB8+HDOxyL6z44dO+TzHMuVK4cWLVrAwMAAd+/exe7du7Fnzx7MmzdPaQpARo0aNULFihXVlmdWpomQkBD89NNPkEgkEAQh07r79u1D3759kZqairp168LJyQm+vr5YuXIl9u7di6tXr+a4P0SkKKtgqXfvQMTGVsaJE+J1ouRkYPx4sSw/BVgAAIHyRExMjABAiImJybLux48fhQcPHggfP378DD0j0t6WLVuEwYMHC35+fgr7pVKpsHTpUgHin6DCxYsXlY4tV66cAEDYvHmzzvqXlpYmNG7cWLCwsBAGDhwoABCGDBmisu6LFy8EMzMzAYCwbt06+f7U1FThu+++EwAIdevWFaRSqcbn588y6VpycrJw6NAhITk5ucDWnTNHEObOTa8zd64gAOK/c+akCr17PxSSk5OFqVPF/bLNwED7fmRHdr6/8+1wIREVPAMHDsTGjRuVEvhKJBJMnDhRPl9y+/btedE9+Pj44MqVK1i0aFGW6V9+//13JCQkoGXLlvj+++/l+/X19bFmzRpYW1vj9u3bOH36tI57TfRl+XTOVsaJ8jNmSNG3byAAYMECoGbN9HqpqeJVsPyEQRYVGrI5NoCYYLZOnTowNzdHkSJF0L59e9y8eTNXzpOamooNGzbAw8MDxYoVg7GxMZycnDBixAiVud0uXrwIiUQCDw8PJCQkYPbs2ahatSrMzMzkX/RbtmyBRCKBp6cnoqOjMX78eFSoUAHGxsbw8PBQOPfatWvx9ddfw9raGiYmJqhUqRLGjh2LFy9eZPm6bN68GQ0bNoS1tTUkEgmePn2aK6+JpmTBV2Y58HQlMDAQM2bMQNOmTTFixIgs6x88eBAA0K9fP6UyCwsLdO7cGYCYd4+IdEfdRPl584C7d9MfV6kiDjPmp0CLQRYVOhMnTsTw4cNhZmaGLl26wMHBASdOnEDjxo3lX5zaiouLQ6tWrTBs2DD8888/qFGjBjp37gxjY2OsXbtWfkOEKomJifDw8MBvv/0GJycndO7cWWkpp8jISLi5uWHbtm2oXr06unTpAnt7ewBi5v927dphxIgR8Pf3R6NGjdC1a1ckJSVhxYoVqFWrlsKduJ8aM2YMhg4dCgMDA3To0AH169f/7AuMB/+X8KZ06dKf9bxpaWkYOHAgJBIJNm7cmOXzjouLw+PHjwEAbm5uKuvI9qt7v4lId375RQ+zZ4t3IsqWOU5NFa945adAK19PfCfSxtq1a3H27Fk0b95cvu/XX3/FTz/9hEGDBqFRo0YoWbKkVm3/8MMPuHjxIjp27IiNGzcqtPP7779jwoQJ6N27Nx4+fKi0kPetW7dQo0YNPH78GKVKlVLZ/rFjx9CiRQscOHBAabmlOXPm4OzZs6hQoQLOnj0rvwqWkpKCESNGYOPGjfjmm2/w6NEjlTeFbNu2DdeuXUODBg2UyrZs2YJBgwZl9+XAnDlz4OXlpVHd+/fv49ixYwCAHj16ZPtcOfHrr7/i1q1bWLZsGSpUqJBl/YxX+MqWLauyjoODAwAgNDQ0V/pIRJrLOIR45IiYziEkBJg8Ob08P2CQVVi4uQGvX+d1L7KvVCnxpyMXDR8+XCHAAoAff/wRe/bsga+vLzZs2CDPw5YdDx8+xM6dO1GmTBns2LFD6a7S8ePH48yZMzh+/DhOnDiBjh07KrWxcuVKtQEWIC7XsH79eqUAKzExEatWrQIALFu2TGE+kaGhIZYvX46jR48iNDQU+/btUznENXnyZJUBFiDesTdw4EC1/VIns/U6M4qPj0e/fv2QmpqKNm3aoFOnTmrrDho0KNOA7927dyhSpIjGfQwICMCcOXPw9ddfY+zYsRodExcXJ/+/uezP5E9YWFgAyN46ZkSUO2bPlsqX0qlaVfwaEQQgKCh9aDFjotO8wiCrsHj9GlAzJ+dLoy5YGDBgAHx9fXHx4kWtgqzjx49DEAS0a9dObdoODw8PHD9+HNevX1cKskqWLInGjRtneg5XV1eUL19eab+vry/i4+NRrFgxlQGKmZkZ+vTpAx8fH1y4cEFlkPXNN9+oPa+7uzvc3d0z7Zu2UlJS0LNnTwQEBKB8+fJZTnrPKoWDpqlbAHEO28CBA6Gnp4dNmzZx7U+iQuirr9L//+CB4mT4vMYgq7DI5OpIvqaDfqta6zLj/qwW81YnJCQEALBx40Zs3Lgx07pv375V2qfJYubq6sgmtat7bgDkw2DqJsBrs5h6TqWmpqJPnz44efIkypUrh/Pnz6NEiRKZHjN06FB4ylI9Z2Hy5MkqV2SQJSz95Zdf4Ofnh0WLFqFKlSoa9ztjEP3hwweFlSVk4uPjAUDpqiMRfV6fBln5CYOswiKXh9wKMyGLBJTqSKVSAOIQWc0s/lSqX7++0j5TU9Msz6FJHW1l1vbVq1exYcOGbLfZtWtXdO3aVWVZWloavv32Wxw4cAAODg64cOECypUrl+1zZGbfvn149uyZ0n5ZkCW70eHIkSM4fvy4Qh3ZvKtjx47J7+C8ePEiACj0MywsDC4uLkrnkN0hmRfBKxGlq1o1/f8PH+ZdP1RhkEWFTmhoqMq5QrIvVdndetklm+jcqFEjrFy5UtvuacXOzg5A5pOsZVfaZHWz4/Hjx5muhKCOo6OjyiArLS0N3333Hfbs2SMPsDK7CqctTVNQXL16VW3Z69ev8fqT+YxWVlaoWLEiHj9+DF9fX5VBlu9/f9jUrl1b8w4TUa5zcgKMjYGkpPx3JYsTFKjQUTfnR7Y/Y96p7GjXrh0A4PDhw0hMTNSqDW25ubnBwsIC0dHROHz4sFL5x48fsWvXLgBAs2bNst2+p6cnBEHI9qbqzkKpVIoBAwZg165d8gBLkzv6dOHOnTtq+z5nzhwAwJAhQ+T7MurWrRsAcamgT8XHx+PIkSMAgO7du+v4WRBRZgwMgMqVxf8HB+ePCe8yDLKo0FmzZo182Edm2bJl+N///gdLS0sMGTJEq3ZdXV3Ro0cPhIeHo3v37iqvonz48AF//fUX3rx5o9U51DExMcGoUaMAAJMmTVIYIktJScG4cePw+vVrODk5ZTrBXdekUikGDRqEHTt25HmAlVPjx4+HmZkZzp49iz/++EO+Py0tDSNHjsT79+9Rt25dtG7dOg97SfRl8/ISc2LJ5mWlpgL/pbjDL7/oYedOzedi6gKHC6nQkaVwaNy4Mezs7BAQEID79+9DX18fmzZtyjSFQlY2b96M9+/f48SJE6hSpQpq1qwJJycnCIKAp0+f4u7du0hOTsbDhw9ha2ubi88K8Pb2hq+vL86dO4eqVauiWbNmsLS0xI0bNxAWFobixYtj79692br7LretXLkS27ZtAyBOxJ+nJiOgs7Mzpk6dqrJsw4YNSkFyRq1bt1Z592RuK1OmDLZs2YK+ffvi+++/x8aNG+Ho6Ijbt28jJCQEtra22LFjx2dP6EpE6fT1xeSjGS/gP3gA7NsHeHvro29f7ebg5hYGWVToLFu2DFWqVMG6detw+/ZtGBoaom3btpg1axa+/vrrHLVtaWmJ06dPY/fu3fjzzz/xzz//4M6dO7CyskLp0qXx7bffonPnzjq5emNsbIyTJ0/ijz/+wLZt23DlyhUkJSXBwcEBY8aMwZQpU7Saj5WboqOj5f/PLFBq2rSp2iDr2rVruHbtmtpjixQp8lmCLADo2bMnypcvj/nz5+PKlSvw9/dH6dKlMWrUKMyaNSvXA2kiyh5ZTqzZs9P3rV4NnD8PzJmTBlfXIADqU8LomkTQ9lYrypHY2FhYW1sjJiYmy1vAExMTERoaCicnJ5iYmHymHhY8sisK/EhTfsWfZdK1lJQUHD9+HO3bt4ehoWGhqptZ/Xr1gNu30+vNnQtMnSrW9ffvCEAfGi5OkaXsfH9zThYREREVaO3bKz5u1Ur8d/fuyvD21seVK8i1ICs7GGQRERFRgfbJUrHo1Anw9tbDzp1V4eEhxfnzynU+B87Joi9SZGQkJstWEtXA0KFDdbbsDBERaW/ePHFOlpcXsHQpEBcHREYCv/yiDxeXt7h4sYR8MenPjUEWFRrZmYsVHx+freSbHh4eDLKIiPIZWYA1dy6QlgYMGACsWpVefv9+CcyZkwbZnKzPPWTI4UL6Ijk6OmYr6aama+kREdHnk5YG+VUqfX3FAEsk/vE9ezaHC4mIiIg0lvWVKQm8vfXzbLiQV7KIiIioQJMNGzZvrro8k7R9OsUgi4iIiAq0tDQxwDp/Xhw+TF+IQZDfXahmAQqd4nBhAcIkm0QFG3+GiXRDNqndw0N8LAiAoaGAlBQJmjaVonlzMRD73HglqwAwMBBj4aSkpDzuCRHlhOxnWPYzTUS5RzY/S3a34YcPqejb9yG8vfUVyj8n/qQXAAYGBjA3N0d0dDQsLS2hnxe3SBBRjqSlpSE6Ohrm5uYMsoh0IGM6h1mzgJQUoHfvIFSuXBmzZ4vfm5978jt/0gsIGxsbhIeHIzQ0FNbW1jA1NYW+vr58vT4iyn8EQUBaWho+fvyImJgYSKVSlC5dOq+7RVQoZUznkNGMGVLo6+vnyXAhg6wCwszMDE5OToiIiMC7d+8QGRmZ110iIg3p6+vDzMwMJUuWhJGRUV53h6hQymw4MC/SNwAMsgoUIyMj2NvbQxAEpKSkQCqV5nWXiCgLenp6MDQ05FVnoi8Qg6wCSCKR8K9hIiKifI53FxIRERHpAIMsIiIiIh1gkEVERESkAwyyiIiIiHSAQRYRERGRDjDIIiIiItIBBllEREREOsAgi4iIiEgHmIw0jwiCAACIjY3N454QEVFhkZKSgoSEBMTGxsLQ0LBQ1dV125qSfW/LvsczwyArj8TFxQEAHBwc8rgnRERElF1xcXGwtrbOtI5E0CQUo1wnlUrx8uVLWFpack0zIiLKFbGxsXBwcEB4eDisrKwKVV1dt60pQRAQFxeHMmXKQE8v81lXvJKVR/T09GBvb5/X3SAiokLIyspK48CioNXVdduayOoKlgwnvhMRERHpAIMsIiIiIh1gkEVERFRIGBsbY86cOTA2Ni50dXXdti5w4jsRERGRDvBKFhEREZEOMMgiIiIi0gEGWUREREQ6wCCLiIiIlOzevRv16tWDmZkZihUrhhYtWuDVq1dK9S5evAiJRKK0Va9ePdP2U1NTUaNGDUgkEuzatUup/MCBA3B3d4eNjQ2MjY1Rvnx5TJw4Ee/evVOqu2/fPnTr1g1ly5aFmZkZqlWrhqVLlyIlJUXluR8/fowffvgBtWvXhqGhIRwdHfH48WO0b98eFhYWsLGxwciRI/HhwwcNXy3VmIyUiIiIFCxduhTTpk3D5MmTsWjRInz48AGXL19GYmKi2mM2bNiAatWqyR+bmZlleg4fHx+8fftWbXl0dDQ8PDzw448/wtraGvfv34e3tzfu3r2Lc+fOKdRdsmQJHB0dsXjxYtja2uL69euYOXMm7t27h61btyq1/e+//+Lo0aOoV68eBEFAVFQUmjdvjjJlymDv3r2Ijo7GxIkT8ebNG+zfvz/T55EpgYiIiOg/wcHBgqGhobB69WqN6l+4cEEAINy4cUPjc4SHhwuWlpbC1q1bBQDCzp07NTpu/fr1AgDh2bNnCvsjIiKU6s6bN08AILx+/VqpLC0tTf7/4cOHC0WKFBFMTEyEN2/eyPfv379fACD4+vpq+rSUcLiQiIiI5DZt2gQjIyMMGTJEZ+cYP348OnfujCZNmmTruGLFigGA0jBgiRIllOrWqVMHAPDy5Uulsk/XHPz48SOaN2+OkiVLyvd17twZFhYWOHr0aLb6mBGHC4mIiD6jwMBAnD59Gv/88w/++ecfPHz4EGlpaZg3bx5mzpyZZf0HDx5AKpXCxMQEenp6qFixIr799ltMmDABISEhKtt2dHTE+/fvkZycLK/fvn17XLhwQamusbExUlNTYWlpCUEQkJqaikqVKqF58+b4+PGjUruRkZEAgMaNGyMtLQ02Njbw8PBA7dq18ejRI6U+C4KAlJQUTJs2Te2cqb179+LXX39FQEAAkpOTYWBggOTkZAiCgMqVK6NChQpZvs7r16+HRCJBs2bNkJKSovA6GRoaKtRNSUlB1apVFfYZGBigcuXKePjwoaZvrTKtr4ERERFRto0bN04AoLTNmzcvW/UrVqwodO/eXShSpIgAQHB3dxdGjRqlsq6enp7QunVrhfplypRRWReAoK+vL5iYmAhmZmYCAMHY2Fhtuw0aNBAqVaokmJubCwAER0dHwcjISGX98uXLC+bm5oKpqakAQHBzc1MaLpQ9X4lEorKNqVOnZvkaf/fdd2qft7u7u5CQkCCvO3z4cLWvf4sWLYRWrVpp8S6LOFxIRET0GVWvXh2TJ0/GX3/9hYcPH6J///4a1Z8wYQIA8QoLAAwcOBD79+9HcHAwXFxccPXqVQQFBcnbXrFihbwNqVSK06dP48CBA3j//j0AxWG0wYMHK5wzLS0NiYmJSEhIAAAkJSWp7JtUKsXNmzcRHBwsvxPv6dOnSE5OVqgnG54LCQnBhw8f8PHjRwCAr68vAKBv376QSCT46aef4OPjAwAQ1CxIs3DhQkgkEuzbtw8AsGXLFqU7G//880+l57106VL56zRr1izVL3YuY5BFRET0GQ0dOhS//vor+vXrB2dnZ6X5QerqX716FQDg4uKiUG5jY4PVq1cDAC5fvoyZM2eiX79+2LZtm7zO2LFj8fDhQ/kmC0KMjIzwv//9D3fu3JHXtbW1BQD4+/sr1NXT00OVKlUUzp9Zu1ZWVvI5VLJ0Dnp6evDx8cGtW7ewbt06eTszZ87E//73P/ldg+PGjVNo9+HDh2jcuLG8/oIFC9CmTRsAQLdu3eR1fH19YWpqKn/dMh7fs2dP+eu0cuVKxMTEyNvT09OTB58ZvXv3Tv4ctKL1NTAiIiLKsYEDB2Y6XCgIgvD8+XP5cFmPHj1U1ndwcBAACDt27FCor65tWf2VK1cq1HV1dRUAKAypyeqWKlVK43ZlQ4IZ+5zZlrF+SEiI2tcJgNC7d2+l8sTEROHrr7/OtI1PXydBEIcLjY2Nhfbt2yvUS01NFSwsLAQvLy+V7WiCV7KIiIjyOX9/fwDi3XUWFhYq67i5ucnryuobGRmpbVNW//z58wp1ZRPAM+aiql27tkIdTdqVDQlm7PPgwYNx4cIF+Va3bl0AQMOGDTFnzhx5fScnJ7XtA5APY8qkpaWhT58+8PPzy7KNjK+TjKmpKS5cuKCQt+vIkSOIj49Hhw4dMu1LZnh3IRERUT4XGhoKAChbtqzaOg4ODvK6sv+bm5srzY/6tH5ISIhCXWdnZzRs2BBDhw7FggULUKpUKflwoizBaMZ2v/vuOzg5OaFOnTqwtLTE69evAQAmJiZITExU6LOTkxM8PDzkj6tVq4bbt29DKpXK25bVb9OmDVq0aIFq1arB2NgYAQEB8uM+vTtw1KhROHToEDp06IBjx46hePHiuHnzJgDgq6++gpWVldLzvnLlCvbt24eQkBDo6+vD1NQU7u7umDZtGoyMjDBx4kR06dJFHpRpg0EWERFRPhcXFwdADG7UkV0tio2Nldf/NBhRVT8+Pl6hrkQiwdGjRzF58mRMmjQJHz9+lOehSk1NVWq3WrVq2LFjB3x8fPDx40d5H8uUKYOQkJBM+ywLrD5+/Kj0HOvVq4c///xTHmDKJvxn7LPMyZMnAQDHjh0DAAQHB6Nhw4YAgAsXLigEdrLnffPmTfTs2VOhnejoaAwfPhwWFhbo2bMnlixZorbvmuBwIRERESkoVqwYNm3ahOjoaHz8+DHTOyCnTZuG+/fvIzY2FikpKRg1ahQA5YSfqhQpUgQAUKpUKaWyefPm4d69e4iLi0NcXBy6dOmitp2nT59CEAT88ssvAIBGjRpBEAQIgqAQYGXUunVreZ2MW1JSEqKiorB27Vq1Q7OaYpBFRESUz1laWgJApgsWy67uWFlZyeurS/aZsb4skNBV3ez2WZNFmTMO/2WU3ddJ1xhkERER5XOOjo4AgPDwcLV1ZGWOjo7y+p9OEFdVXzZBXFd1s9vnzOrLyOqq26/pOXWNQRYREVE+5+rqCgCIiopSmo8kI0vsWbt2bXl9dUlEM9Zv3ry5Tutmt89RUVHyeVjqyO52/JQmbWQ8p64xyCIiIsrn7O3t5ekOZHcDZnT16lWEh4fD2NgY7du3V6ivSsb6/fv313nd7PZ5x44dSvXfvHkDANDX10f79u1Vnj+rNj49p64xyCIiIioApk+fDgC4f/++wv6oqCiMHDkSADB69GhYW1sr1AcUl9BRVV/XdbPb54ULF8pzXsnqy1IyNGjQQF4/s9dJVRuqzqlLEkFQszgQERER5To/Pz/5lz0APHnyBJGRkbC3t4ednZ18/8GDB1G6dGmF+k+fPpVf0TExMYGZmRliYmKQlpaGunXr4vfff8fEiRPlbdy9exeJiYkAAGtra+jr68vr16xZE0ZGRvK7AHOrbqNGjbB48WJ5P7Lq86VLl+RL4QDikjrLly+HgYEBzM3NFdoGADs7O9jb2yu9ThnJ2jA0NESLFi1gbm6Oc+fO4f3792jUqBHOnDmjcE6d0TpXPBEREWXbhQsXslxiBoAQGhqarfqBgYEa1/3pp5+E06dP66RuUlJStvqsyu7du4UaNWpk63VS1UaTJk0EKysrwdTUVKhevbqwcOFCISkpSUfvrDJeySIiIiLSAc7JIiIiItIBBllEREREOsAgi4iIiEgHGGQRERER6QCDLCIiIiIdYJBFREREpAMMsoiIiIh0gEEWERERkQ4wyCIiIiLSAQZZRERERDrAIIuIiIhIBxhkEdEXqUyZMpBIJAgODsaCBQtQp04dWFlZwcTEBG5ubjh06FBed5GICjguEE1EX5wXL17A3t4eFhYWqFKlCvz9/fH111+jVKlS+OeffxAaGgqJRIIDBw6ga9eued1dIiqgeCWLiL44vr6+AID4+HjExMTAz88PV65cwd69exEYGIhv/t/O3YRE1YZhHP87M0ao41AxmCYoR1QSM6MCEVu0c2MI+bFyUxAulKCFLhQ0EBQFIVDBVUSbEFEXtqpoEVGQBOPCD1DQIYlQYRT8YMx53kXM0LxHp3pfJqfm+oGbcz/3fZ7lNec5nro6jDF0d3ef7EZF5I+mkCUiSSccstxuNy9evODy5cuRWmpqKr29vQDMzs6yu7t7InsUkT+fQpaIJJ1wyHrw4AH5+fm2umVZuFwujDHs7Oz85t2JyN9C72SJSNLxer1sbGywtLREQUGBrb63t0daWhpOp5NgMIjD8d9/jx4eHgLgdDp/qfZ/Z4vIydOTLBFJKisrK2xsbHD27NkjAxbAu3fvACgrK4sKWIeHhwwODlJaWsrp06cpKipidHQ0qjc3N5eOjg56enrIy8sjNTWV7e3tmLVgMMjDhw8pLCwkLS2N0tJSxsbGbPuKNVtEEo/rpDcgIvI7hY8KMzMzj13z5MkTAG7fvh25Zoyhvr6ely9f0tHRwdWrV3n9+jXNzc1kZWVRW1vL5uYma2trPH78mMrKSoaHhzHGcObMmWNr6enpVFdXs7i4SHd3NwUFBUxMTNDY2IjX6+XmzZsAMWeLSIIyIiJJpL293QDm1KlTZn9/31afmZkxTqfTnDt3zgQCgcj1oaEh43K5zJs3b6LW19TUmJqaGmOMMa9evTKAaWpqss09rtbV1WXcbrdZXl6Oun7x4sWotbFmi0hi0nGhiCSVDx8+AESO6L43MzPDrVu3CIVCjI6O4vF4IrX+/n5qa2upqKjg69evkb+SkhJWV1cB8Pl8pKSk0NfXZ7vvUbVQKMTIyAh3797Fsqyo9ZZlsba2FrNfRBKbjgtFJGkYY/j48SMA7e3t9Pb2Mj09TUlJCX6/n/fv35OSksKjR4+ijgoXFxfx+/34/X7Gx8dtc6uqqoBvQai8vJycnBzbmqNqc3NzrK+vU11dbVv/6dMnLl26FLNfRBKbQpaIJI2lpSUCgQCWZdHX18eFCxcYGRlhamqKzMxM6urqaGtr49q1a1F9nz9/BmBycpLc3Fzb3PB7UT6fj+vXrx9576Nq6+vrAJw/f952v4WFBVpaWmL2i0hiU8gSkaQRfuk9HFZaW1tpbW39YV92djYAGRkZtgAWdnBwwNzcHPfu3fvpWl5eHvAt/H3/QdSuri7cbjcNDQ0/nC0iiUshS0SSRvh9rF99IlRYWEhFRQV37tyhs7OT4uJiAoEA8/PzbG5uMjAwwMLCAsFgkCtXrtj6j6tZlsWNGzdoa2vDGIPH4+Hp06c8e/aM58+fR/4DMtZsEUlcClkikjT+/STrZzkcDiYnJ+ns7KSnp4cvX77g9XopLy+PPAnz+Xw4HA7Kysps/bFqY2Nj3L9/n+bmZg4ODqisrOTt27dRe4zVLyKJS198F5GkEAqF8Hg87O7usrW1RUZGxklvSUT+cgpZIiIiInGg72SJiIiIxIFCloiIiEgcKGSJiIiIxIFCloiIiEgcKGSJiIiIxIFCloiIiEgcKGSJiIiIxIFCloiIiEgcKGSJiIiIxIFCloiIiEgcKGSJiIiIxME/eX0v7gSpPdIAAAAASUVORK5CYII=", + "text/plain": [ + "
" ] }, "metadata": {}, @@ -435,198 +1141,396 @@ } ], "source": [ - "import matplotlib.pyplot as plt\n", + "def plot_metrics_vs_error_rates(\n", + " metric_values, model_name, num_nodes, xlabel, ylabel, filename, red_line_value\n", + "):\n", + " \"\"\"Plot the metrics against error rates.\"\"\"\n", + " plt.figure()\n", + " plt.plot(\n", + " [real_p_error_list[0], real_p_error_list[-1]],\n", + " [red_line_value, red_line_value],\n", + " color=\"red\",\n", + " linewidth=2,\n", + " label=\"p_error=2E-40\",\n", + " )\n", + " plt.plot(real_p_error_list, metric_values, color=\"blue\", linewidth=2, marker=\"x\")\n", + " plt.grid(True)\n", + " plt.legend()\n", + " plt.title(f\"{model_name} {num_nodes} nodes\")\n", + " plt.xlabel(xlabel)\n", + " plt.ylabel(ylabel)\n", + " plt.semilogx()\n", + " plt.xticks(10.0 ** np.arange(-6, 1))\n", + " plt.savefig(filename, bbox_inches=\"tight\", dpi=300)\n", + " plt.show()\n", + "\n", + "\n", + "# Plotting accuracy vs error rates\n", + "plot_metrics_vs_error_rates(\n", + " acc_scores,\n", + " \"DecisionTreeClassifier\",\n", + " num_nodes,\n", + " \"$p_{error}$\",\n", + " \"Metric\",\n", + " \"DecisionTreeClassifier\" + \"acc_p_error.eps\",\n", + " 0.91,\n", + ")\n", + "\n", + "# Plotting execution time per data point vs error rates\n", + "plot_metrics_vs_error_rates(\n", + " np.asarray(time_p_error) / X_test.shape[0],\n", + " \"DecisionTreeClassifier\",\n", + " num_nodes,\n", + " \"$p_{error}$\",\n", + " \"Execution time\",\n", + " \"DecisionTreeClassifier\" + \"speed_p_error.eps\",\n", + " 1.807,\n", + ")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Speed vs bitwidth" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "ap relative: [0.49626943 0.70187731 0.82640876 0.89067066 0.98315255 1.02264581\n", + " 1.02436888 1.01090038 1.01268386], f1_relative: [0.06488922 0.65490682 0.87590196 0.90861806 0.97920588 1.00604989\n", + " 1.00914511 1.00274636 1.00389957]\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "ap relative: [0.44633527 0.69262992 0.84361325 0.95028007 0.98310165 0.99367784\n", + " 1.00177485 1.00137336 1.00286707], f1_relative: [0. 0.65425061 0.88978108 0.9629844 0.98484002 0.99187697\n", + " 0.99896308 1.00004651 1.00134248]\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "ap relative: [0.45810941 0.66176353 0.85701522 0.93668402 0.96541385 0.98353791\n", + " 0.99091316 0.99133601 0.99740638], f1_relative: [0. 0.57332946 0.87035559 0.9402579 0.96505021 0.983713\n", + " 0.99082334 0.99224022 0.99758998]\n" + ] + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlEAAAHZCAYAAACmd272AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACz20lEQVR4nOzdd3hT1RvA8W+6Fy20ZbTQwS4CsgXZlCEge4NiAUUQFAFFhj+GooIIKgqIoDIUFRAB2XuKykYQGULLXi2lZbSl4/z+uCQ0JG3TNt3v53nykLvOPSdJk5dzz32PTimlEEIIIYQQ6WKT0xUQQgghhMiLJIgSQgghhMgACaKEEEIIITJAgighhBBCiAyQIEoIIYQQIgMkiBJCCCGEyAAJooQQQgghMkCCKCGEEEKIDJAgSgghhBAiAySIEnnWwoUL0el0Jg8bGxs8PDyoWrUqgwcP5vDhwzldVavQt2/hwoU5XZUM27lzp9n3zNwjMDAwp6ubZ5w/f57JkyfTpEkT/Pz8cHJyws3NjdKlS9O5c2fmzJlDRESE2WPzw+dKr1+/fuh0Opo2bZriPocPH6Znz56UKlUKBwcHQ/vDwsIACAwMRKfTMWnSpGyps8jbJIgS+Y5SiujoaE6cOMHXX39NnTp1mDx5ck5XS4hUhYWFGX7Qd+7cadExsbGxDB06lKCgICZMmMDu3bu5fPkycXFx3L9/n7CwMFatWsXQoUPx8/Nj3LhxWduIXO7w4cM0aNCAZcuWceXKFeLj43O6SiKPkyBK5Avr16/n7t273L17l8jISA4fPsx7772Hk5MTSUlJTJgwgV9//TWnqymSmTt3ruE9M/c4efJkTlcxV4uMjKRp06bMmTOH+Ph4goKCmDlzJkePHuXGjRtcvXqVAwcOMGXKFCpVqkRMTAxTpkzJ6WrnqOnTpxMbG0vx4sXZtm0bt27dMnzeAgICcrp6Ig+yy+kKCGENzs7OuLm5GZZr1KhBjRo1qFy5Mt26dQPgk08+oUuXLjlVRfEER0dHo/dMWE4pRd++ffnrr78AGDVqFFOmTMHW1tZoPx8fH2rXrs0777zDggULePPNN3Oiutlm4cKFqV6WPHbsGAC9evUiODjY7D76y3pCWEJ6okS+1rVrV8qXLw/AgQMHePjwYQ7XSIjMW7JkCevWrQNgwIABTJs2zSSASs7GxoaXX37ZEHQVVA8ePACgcOHCOVsRkW9IECXyvUqVKgGQmJjI7du3jbYlJCSwc+dO3nrrLWrXrk2RIkWwt7fHy8uLhg0bMmPGDO7fv59i2ZMmTTIaBB0WFsZrr71GYGAgjo6OFC9enG7dunHkyJE06/nXX3/RtWtXihUrhpOTE2XKlGHo0KFcvHjR4raePHmSQYMGUb58eVxcXChUqBCVK1fmrbfe4vLlyxa347///uPVV18lMDAQZ2dnSpcuzdChQ7l27ZrhGKUUixYtomHDhnh5eeHq6sozzzzDDz/8YHF900spxU8//cTzzz9PiRIlcHBwoGjRojRv3px58+aRkJCQ4rFPDhj++eefadWqFSVKlMDW1pZ+/fqZHLN69Wq6detmGKxdpEgR6tWrxyeffGL4QTYnPj6euXPn0qxZM4oWLYq9vT2enp5UrFiR9u3b88UXXxAeHm5Ut9KlSxuWmzVrZjLQPvk4qY8//hgAV1dXZsyYYeGrB5UrV7Z4X8j83wfAhQsXGDFiBFWrVsXNzQ0HBwd8fX2pXr06r776KitXrjR73IkTJ3j11VcJCgrC1dUVJycnSpUqRe3atXnzzTfZtm2byTHmBpYnH2um72V67733jF7b5IPILRlYnpiYyOLFi3n++efx8fHBwcEBLy8vmjZtmurnMPmNFWFhYdy5c4fx48dTtWpV3N3d0zUeTuQSSog8asGCBQpQgNqxY0eK+7Vr186w34MHD4y2ff7554ZtKT0qVqyowsLCzJY9ceJEBaiAgAC1c+dO5eHhYbYMR0dHtWXLlhTrOGvWLGVjY2P2WE9PT3XgwAHD8oIFC8yW8fnnnytbW9sU2+Hs7Kx+/fXXNNuxdetWVahQIbNllC5dWl2+fFnFxsaqTp06pXiu999/3+x5duzYkWY7UhIVFaWCg4NTfa9q1Kihrl69avb4gIAABagJEyaovn37mhwbEhJi2PfOnTuqVatWqZ6rfPny6ty5cybnuXv3rqpXr16an6vly5eb1C21h/4zfurUKcO6/v37p+s1NCe19yOzfx87d+5Urq6uqR7v6upqctxPP/2k7OzsUj2ucuXKJseFhIQoQDVp0sSwLjQ0NM02TJw40bC//r1Ivi65S5cuqZo1a6ZaXt26ddWtW7dMjk3++d+2bZvy9/dP8X0WeYP0RIl879SpUwCULVsWZ2dno23Ozs707t2bRYsW8eeffxIWFsatW7c4duwY06dPp2TJkpw+fZpevXqleo6oqCi6dOlCQEAAv/76K9euXeP69essWrSIwoULExcXx4ABA8z+D3X79u288cYbJCUlERgYyNKlS7l+/TqXLl1i/vz52Nra0qNHj1TPv2zZMoYPH05iYiIVK1ZkxYoVXL9+ncuXL/Pdd9/h4+NDTEwMPXr0SPWSzp07d+jRowdly5ZlzZo13Lhxg0uXLjFjxgzs7OwIDQ1l7NixjBkzhnXr1vG///2PkydPEhERwR9//EHdunUB7X/6+tfdWnr16sX27dsB6Nu3LwcOHCAiIoK///6bN998E51Ox5EjR2jXrl2qd1199913fP/994SEhPDXX38RHh7OqVOn6N+/P6D1vjz//PNs3rwZV1dXJk6cyJEjR4iIiODixYt89913+Pr6cvbsWdq1a2fSI/Xxxx/z559/AjB06FD279/PtWvXiIiI4MSJEyxYsID27dsbXX47efIk//zzj2E5+Y0S+kejRo0A2LNnj2G/xo0bZ/JVTV1m/j6SkpJ46aWXuH//PsWKFePrr7/m9OnT3L59m6tXr7Jnzx4++OADKlasaHTcnTt3GDhwIAkJCZQrV44lS5bw33//ERkZyeXLl9m2bRvjxo3Dz8/PojYEBAQYXkN/f38Axo4da/TaWnrXYnR0NMHBwRw+fBhvb29mzJjBP//8w+3btzl37hwzZ87Ew8ODv/76ix49epCUlJRiWSEhIdy9e5fPP/+cc+fOcevWLXbt2kWZMmUsqovIJXI6ihMioyzpiVq2bJlhn88++yzd57hy5YoqXLiwAtT27dtNtut7cABVvXp1de/ePZN9VqxYYdhnw4YNJturVq2qAFW8eHF15coVk+0nTpxQTk5OKfYYxMXFqeLFiytAlS1bVkVERJiUce7cOUM7atWqlaF2vPvuuwpQtra2SqfTqZ9//tlkn/DwcEMv1tixY022J/+f+Ny5c9Xdu3dTfCS3atUqw3EjRowwKVcppWbMmGHY58svvzTZnry355133jFbhlJKffbZZ4YekiNHjpjd5+LFi8rb21sBavr06Ubb9L0UnTt3TvEc5iTvMUmtN2LcuHGG/f788890ncOclD5Xlkjt7+Pvv/82lL169WqLy/ztt98Mxx07dixd9THXE5VcWr1Mae3z5ptvKkD5+Pik2Pt25MgR5ejoqAD1yy+/GG1L/vl3cHBI8fMl8g7piRL5QkxMDPfu3ePevXtERUVx9OhRJk+ezEsvvQRA7969eeONN9Jdrq+vLy1btgRgy5Ytqe778ccf4+rqarK+U6dOhoGsBw4cMNp28OBBjh8/DsC4cePw9fU1Ob5y5coMGTIkxfPqe4wApk6diqenp8k+ZcqUYcyYMQAcOnQo1TFaKbVD39uQmJhI/fr16dmzp8k+Xl5ehtcrrUHMgwcPplChQik+7ty5Y9j3m2++AcDb25uPPvrIbHkjRozgqaeeAmD+/PkpnrdIkSK89957KW6fOXMmAMOHD6d69epm9/Hz8+P1118HtEHeyel7G829l9aQfFxfTg+QTu3vI3mva3pei4wel9Xu379v+By+//77KaZEqF69Or179wZMPxvJDRgwIMXPl8g7JIgS+ULbtm0NP76FCxemRo0aTJgwgcTERH788Ud+/PHHFO9eevDgAbNnz+a5557D19cXJycno0Gny5cvB+D06dMpnt/R0ZFmzZqZ3WZjY2O4Q/D69etG2/bu3Wt4nlr6ha5du6a4TX95x9HRkQ4dOqS4X/KgJ/kloeRSa0fZsmUNz5977rkUz1OuXDkAo0HomaGU4vfffwegffv2ODk5md1Pp9PRvXt3AI4fP24UhCUXHBycYhn//fefYfBxcHCwITA396hatSqg3Taf/K5P/Q/jggULWLJkSZ6/IzSjfx8VK1Y0vM5vvPGGIb1AWp5++ml0Oh0A/fv357///rNiazLujz/+MAyib9q0aaqfjaeffhow/U9Tcs8//3y21FtkLckTJfK1+Ph4Ro0aRe3atQ2BTHJnz57lueeeIzQ0NM2yoqKiUtymvwMrJS4uLgAm42f0P9hubm6UKlUqxeP1PSzmXLhwAdCCFwcHhxT3CwwMxNXV1ZDJ2pzU2pF8PFlqPQT6/WJiYlLcB7Qgw9wdcU+Kjo4mMjISSP11gMd3nymluHjxotmemtTGnCQfx9W8efM06wba2J/bt29TokQJQLvTcdWqVURHR/Piiy/y2muv0bBhQxo2bEhwcDB169Y1BAkZkbynMaVA0Voy8/fh4uLCBx98wNtvv82ff/5J9erVKV26NE2aNKFRo0a0bNnS7LimsmXLMnToUGbNmsXatWtZu3YtTz31FI0aNaJx48a0bNmSokWLWq2Nlkr+2TD3XWLOrVu3UtwmY5/yB+mJEvnCjh07UEqhlOLevXscPHiQPn36AHDlyhU6d+5MbGys0TGJiYl06dKF0NBQXF1deffdd9m1axcXL14kMjLSMOhU3zWf2u3zqeXoSU4pZbR87949gDSTTqa2/e7duwAUKlQozfPry9Ef8yRL22HJfk+2NaOS1zWtNibfnlIb9QGtOakFyqlJ/tkqXbo0hw8fpm/fvri4uHD37l02bNjAu+++y7PPPkvp0qVZtGhRhs6jL1/P2oP3k7PG38dbb73FL7/8YrjhIDQ0lIULF/Lyyy8TEBBAmzZt+Pfff02O++KLL5g3b54hKD558iRff/01L7zwAr6+vvTq1YsrV65kWdvNychnIy4uLsVtqX0ORd4hQZTId1xdXalVqxZLlizhlVdeAeCff/5h+vTpRvvt2rWLEydOAPDLL7/wwQcf0LhxY/z8/ChcuDBubm64ubmlmQcnM/RBjT6YSklq2/WBQ1plJN/HkoArt0he1/S8ThlpY/Jg9e+//zYE5mk9npwsuWzZsixevJjIyEj27t3LJ598QuvWrbGzs+PChQv069ePTz/9NN31Awx36YH2Gc4q1vr76Nq1K3/++Sc3btxg1apVvP3221SqVAmlFBs3buTZZ5/l/PnzRsfodDoGDhzIiRMnuHjxIkuXLuX1118nICCAhIQEli5dSv369bO8Jy655J+N6Ohoiz8bIn+TIErka5999hk+Pj6ANu1L8kG5R48eBbSBxq1bt06xDP3A76yg//G9d+9eqskwU5tHTl/G2bNnUx1/ExYWZvjBe/JHPzdzd3enSJEiAEZpAMzR/+jrdDrD7ezpkfwSy7lz59J9/JMcHBxo0KABb7/9Nhs2bODcuXOGMWMffPABiYmJ6S6zYsWKhh6a5cuXZ7j3LC3W/vsoVqwYHTt25JNPPuHkyZP88MMP6HQ6oqKiDIP5zfHz86NHjx58+eWXnD9/3jD/38WLF1mwYIHlDcoka382RP4gQZTI19zc3Jg4cSKg/e/xs88+M2zTd7Wn9kP2+++/WzQeJKMaNmxoeJ7aBMkrVqxIcZu+ZyIuLo61a9emuJ9+AHDyY/ICnU5neJ3Wrl2b4iUSpRS//PILAFWrVs3QnWtVqlQxjPf6+eefM1bhVPj7+/Pqq68C2gTCN2/eNGxLPhYtreDqnXfeAbTge+TIkRafP60gNLms/vt44YUXDLMJWHpZ0sbGhtGjRxt6hbLycuaTmjRpgqOjI5A1nw2RN0kQJfK9AQMGGHpevvzyS8P/3PX/s4yOjmbHjh0mx0VHR6eaWsAaateubbjL66OPPjJ7R9s///zDnDlzUiyjXbt2FC9eHIAxY8aYvcQRGhpq+B98rVq1qFGjhhVqn330l2Vv3brF//73P7P7fPHFF4YgQR+opJdOpzMEJcuWLUv1FnXQAown7x5L64dd34tha2uLh4eHYb2np6dhwPnVq1dTLaNv376G3qHvvvuO0aNHpxrsJCUlsXDhQurVq5dqucll9u/jypUrqV5+ffDggeHz7uXlZVgfGhqaao/qtWvXDD2qyY/Lau7u7gwcOBCAzz//3OxrklxsbKzhpg+Rf0kQJfI9e3t7xo8fD2B06aB169aGH7HevXuzePFiLl68yPXr1/nll1+oV68eJ06cMMmobG2ff/45Op2OGzdu0KBBA5YtW8aNGze4cuUK3377Lc2aNTNckjTHwcHB0KazZ89Sv359Vq1axY0bN7h69aphfrvIyEjs7OyYPXt2lrYnK3To0IE2bdoAMH36dPr378+hQ4e4ffs2J06cYMSIEYbgp2bNmhkOogCGDRtGkyZNUErx4osv8uKLL7JlyxauXbvGnTt3CAsLY8OGDYwaNYrSpUvz+eefGx3/1FNP0aJFC+bMmcOhQ4e4efMmt27d4vDhw4wcOZJ58+YB0LFjR6PBxc7OzobLdLNnz+bkyZPExcWRkJBAQkKC0fganU7HDz/8QO3atQGYNm0aVatW5YsvvuDvv//m1q1bXL9+nUOHDvHJJ59QrVo1+vfvb9G4Ob3M/n1s2bKFUqVK8corr/Drr78aso5fvHiRtWvX0qJFC8Ndl/rB6QCLFi3C39+fN998k3Xr1hnmmAsNDWXp0qW0aNECpRQ2NjZpZvK3tg8//JBKlSoRFxdHq1atGDp0KHv27OHmzZtERkby33//sWrVKoYOHYqfn59R76/Ip7IhoacQWcLSufOUUio+Pl6VK1dOgTYXXXR0tFJKm6MrpfnmbGxs1BdffJFqFuTkc86lpkmTJgqM52dLLrW584oUKaL279+fZmZpa82dl5q06pBWWblh7rzUslXrRUdHq27duqU55xpmMqhbckz16tXV9evXTc6b/DP95MPcZ/z+/ftq0KBBac4zByg3Nzc1adIkkzJSez8y8/eRWluSP8aMGWN0XPLs+Sk9bG1t1ezZs03qm9UZy5VS6vr164a/57QeM2fONDo2+ec/NDQ0xTqIvEN6okSBYGdnx4QJEwAt4/OsWbMALQv3rl27aNeuHUWKFMHBwYFSpUrRo0cPdu/enaEs5xkxdOhQ9u3bR+fOnSlatCiOjo4EBAQwaNAgDh8+TJ06ddIs48033+TYsWMMHDjQME+gq6srTz31FCNGjODMmTN07tw5G1qTNdzd3dm6dStLliyhTZs2FCtWDHt7e7y8vGjWrBlz585l//79qfbaWapQoUIsX76cPXv2MGDAACpUqICbmxt2dnZ4eXlRt25dhg0bxubNm03u+jx06BDTpk2jTZs2VKhQAXd3d+zt7SlevDitWrVi/vz57N+/33AJNrl+/fqxfPlymjdvjpeXV5qpJFxcXJg7dy6nTp1i0qRJNGrUCF9fXxwcHHBxcSEwMJBOnToxd+5cLl26ZBgfaKnM/H306NGDtWvXMmLECOrVq4efnx+Ojo44OztToUIF+vXrxx9//GG4zKw3fPhwli9fzuDBg6lduza+vr7Y29vj6upK5cqVGTp0KH///XeWX2pPSfHixdm5cydr166ld+/eBAYG4uzsjL29PcWKFaNhw4aMHj2affv2MWzYsBypo8g+OqVy5z2Yp0+fZvPmzRw6dIhDhw7x77//kpiYyOTJk1McE2GJrVu38umnn7J//37u379PQEAAXbt2ZezYsWnm6hFCCCGE0Mu1Gcu/+uqrVG97zYjPPvuMkSNHotPpaNSoEcWLF2fPnj189NFHrFixgr179+Lt7W3VcwohhBAif8q1l/OqVKnC22+/zZIlS/j333/p27dvpso7cuQIb731Fra2tqxbt45du3axbNkyzp07R/PmzTl9+jSDBw+2Uu2FEEIIkd/l2p4o/S3NejY2mYv3pkyZglKK/v37G+7yAW1MwbfffkuZMmVYsWIFp06dIigoKFPnEkIIIUT+l2t7oqzp4cOHrFu3DsAwn1pyAQEBNGjQAICVK1dma92EEEIIkTcViCDqzJkzPHjwAMCQV+VJ+vVHjhzJtnoJIYQQIu8qEEGUflqCwoULpzgpqZ+fn9G+QgghhBCpybVjoqzp7t27ALi6uqa4jz69QXR0dKplxcXFGc3dlZSUxO3bt/Hy8jJM2SCEEEKI3E0pxd27d/H19c3wuOsCEURZ05QpU3jvvfdyuhpCCCGEsIJLly5RqlSpDB1bIIIo/SU8/aSV5ujnlHJ3d0+1rLFjxxrNmh4VFYW/vz9nzpzB09PTCrXNG+Lj49mxYwfNmjUzmn0+v5N2S7sLAmm3tLsguH37NhUqVEhxmI8lCkQQFRgYCMCdO3e4e/eu2Rfs0qVLRvumxNHREUdHR5P1np6e2TqjeE6Lj4/HxcUFLy+vAvVHJ+2WdhcE0m5pd0GSmaE4BWJgecWKFQ2zpR88eNDsPvr1NWvWzLZ6CSGEECLvKhBBlIODA88//zwAP/74o8n2CxcusG/fPoA8PUGrEEIIIbJPvgqiZs2aRVBQEC+99JLJtjFjxqDT6ViwYAEbN240rH/w4AEvv/wyiYmJdO3aVbKVCyGEEMIiuXZM1OHDhxkyZIhh+dy5cwB8/fXXrF271rB+5cqV+Pj4ABAeHs7p06cpUaKESXk1a9ZkxowZjBw5krZt29KkSROKFSvGnj17uHbtGhUrVmTu3LlZ3CohhBBC5Be5NoiKjo7mr7/+Mll/+fJlLl++bFhOnrMpLSNGjKBq1arMmDGD/fv3c//+ffz9/Rk7dixjx47N1Ah9IYQQQhQsuTaIatq0KUqpdB0zadIkJk2alOo+LVq0oEWLFpmomRBCCCFEPhsTJYQQQgiRXSSIEkIIIYTIAAmihBBCCCEyQIIoIYQQQogMkCBKCCGEECIDJIgSQgghhMgACaKEEEIIITJAgighhBBCiAyQIEoIIYQQIgMkiBJCCCGEyAAJooQQQgghMkCCKCGEEEKIDJAgSgghhBAiAySIEkIIIYTIAAmihBBCCCEyQIIoIYQQQogMkCBKCCGEECIDJIgSQgghhMgACaKEEEIIITJAgighhBBCiAyQIEoIIYQQIgMkiBJCCCGEyAAJooQQQgghMkCCKCGEEEKIDJAgSgghhBAiAySIEkIIIYTIAAmihBBCCCEyQIIoIYQQQogMkCBKCCGEECIDJIgSQgghhMgACaKEEEIIITLALqcrkF/YVa0KNmnEpDVrwm+/Ga/r0AEOH077BCNHag+9u3ehUiXLKrd6NdSq9Xh57VoYPDjt49zc4NQp43WjRsFPP2EHtIqNxc7Jyfyxzz8PX39tvK52bbh+Pe3zTpsGffo8Xj59Gpo3T/s4gAMHwMfn8fK8efD++2kfV6ECbN9uvO6FF2DXLqNVZts9cCBMnGh8bKlSltX3hx+gadPHyzt3wosvWnbs5cvGy++9B/Pnp31ckyawZInxuuBgOHMmxUP07dZNngyvvfZ4w7VrUKeOZfXdtg0qVny8/OOP8M47aR9XogQcPGi8btAgWLcu7WN794ZPPjFeFxQE9+6lfezcufDcc4+XDx2Cjh3TPg7g33+hUKHHy59+qj3Skku+I3Tr1tHqlVdS/vvWS+U7Ik258Dsi1e81C78jzMrl3xEm7c7Ad4TBhAnw6quPl63xHaGUtqyU8UO/rkgR+OwziIvTHrGx8N138PffqR5nl5RkWb1SIUGUleiuXUt7Jz8/03W3bsGVK2kfGx1tvKyUZccBPHxovBwTY9mxyX8E9CIj4coVdIBzasfevm267vp1y8774IHxckKC5W1NTDRevnfPsmM9PEzXhYebHGu23VFRpsdaWt+4ONNlS481Vw9Ljg0PN11340aqx+rbnfhk8JGYaHl9ExKMlx88yHhbb9+27NjISNN1V69qAUZaYmKMlx8+tLy++i9qvehoy47NRd8RzhERaR+XyndEmnLhd0Sq32sWfkeYlcu/I0zafeuW9hrHxj4OTC5dsuy869drnwH9sZa+RgAtWmifW30gFBNj+n6Zc+uWFpSnky7dR5iSIMpKlI9P2j1RRYuaX1eyZNoncHc3XtbpLDsOwMHBeNnZ2bJj3dxM1xUpAiVLooDY2FicnJzMfxA9PU3XlShhQWUBFxfjZTs7y9tqa2u87OZm2bHFi5uu8/Y2OdZsu819uVpaX0dH02VLj32Sh4dlx3p7m64rXtz8F/0j+nbbP/mZsLW1vL52T3zduLhYdqy5z42np2XHFilius7X17KeKOcnfk4dHCxvq+6Jvwp3d8uOzUXfETFeXin/feul8h2Rpuz8jkjeC/Hff1qw9uCB9jh7VquzUiilSIiPx87OzrTdMTFaT05S0uPH5cumr8GTATRovVWvvGJ87JOfr5SOHT9e68XUHxcRYfp+mTtOKahSxfic4eGm35FKoZKVoVMKtmwBV1fTMi2xerX2yIgne87yAJ1S5l59Yano6Gg8PDwIDw/Hy8srp6uTbeLj41m/fj1t27bF3t4+p6uTbaTd0u6CIMvbrZTW0/DggRac6AMaaz+PiYH4eOvXX1iXvT04OWn/iXR0NH6e3uV07Bvx4AHeNWsSFRWF+5P/CbGQ9EQJIYR47FGPhcv16/DPP1oQ8mRgYo3gRliXjU3qD50uxW1Kp+N+fDyuRYqgc3bO3mDGwSHtqzhZxZJL1mmQIEoIIQqCxERt7Mi1a9rj6lXjf/XPr1/HPiGBljld3/TS6bTLfC4u2qWyJ5+bW/fE8wR7e46cOEGNWrWw0/+4pxJ8ZCRgyZL9n7x8nE4J8fFsK4A9rtYgQZQQQuRlCQlw86ZpMPTk8xs3LBuka20ODhkKaNL93MEh08GEio/n6vr1VG/bVrvEJEQaJIgSQojcKD5eGwCdUlCkf37zpvmBxeml02mD2H18SCpRgqsPHuBbtiw2bm4ZD26cnU0HMguRj0gQJYQQ2Sku7nEQlFqAdOuWdc5nY6Pdfenjoz18fc0/L17c0PuSGB/PofXrKd62LTbSIyNEiiSIEkIIa4iJSTswunrVfH6kjLC11VICpBQU6Z8XKya9QUJkEQmihBAiFbaxsVouofDw1AOkO3esc0J7+7R7jXx9tXxfOXVXkxACkDxRmabPE+XjcwcbGzNJF5PJJTM6WGPWF0AZkk6ay/uaC2d0SJXlMzqYtjuXz+hgImMzOmjtnjzZntdee/x/r/ww64ujiqVkwgUCEs/jlxCKf+J5/BNCaeB7HveIUHSpJCFNj1gcuWHry01bH27a+HDD1pcbyZ7ftPWheHUflmz0MhognRPfEatWJfDKK/Ep/n3r5bNZX0jtey0fz/rCk+3OLbO+LFny+DviyZlbkj8KF4YPPtBSj+kfP/2kzb6U0owxSkFSUhTx8YUlT1RucO1a2neF5KIZHTI76wtpTJCQG2d0SJXlMzqYtjuXz+hgIgOzvqBv9717xi9wnpj1JSlJ+/CFhsL587wUeh7fh6GU4TxlOI8vV7HBzP8lz1tWl/u4cA0fruLLNXy4hg+D3/PFqfTjXqOpi3wYO60wJOoglRvk6kVjErPk1HdERESqEzsB+W7WF1L7XsvHs77wZLv13xFJSY9nYLlyxbLzbt4M9+8/DmYs/fwCNGumnVN/7JNtT0lEhPYfpvTL/MQvEkRZiY+PKkizvpBWT1Q+nfUFc+0uALO+oG+3m5vxIOPcMuuLa9Jd/BND8U84j19iKAEJ5/FPOE+VlaGwJFT7Rn5knGXVJcnWDl2APxHOznhWqcItu5LMW/O41+jGox6ke7pCJrfWvzwCnJIFGA7F89ysL3h5xVjUE/WknJj1xcZGu8KZWm+F/rF4sXbFVJ//88ABrTxtuyIxMQlbWxuebPeNG9C2rXHPxvHjpn/D5qxcCUePGh/75GuekpEjtb9tff2joy3PvlCzpvE5r19/fKzxNShFUpJCp9OhlI7t27V2PRlcW2LFCu2REZZMQZvbyOW8TJJpXwpWcjZpdw61OyFBmwD1/HlDj5LRc3Pda5YoWhTKlNEepUsbPy9Vinil5P3ORLv1yc6tOaOLuW3JYmSRBzk55czj3r0ISpXylst5Qog8TimtT95cgHT+PFy8mLFEkU5O5gMk/b/mulKSKyDzrimlXda6fl3rcbl8Wce+ff5cvGhDXFzGA58nL+EK69F3fqb2r+X7KHS6eAoVssfJSWf1YEU/y4u5hxVypGaYNT6fEkQJIbJHbCyEhaXcm3T3bvrL1Om0EbrJA6PkwVKJEjn3DZ0LxMZqQZE+OLp+3fiRfJ3xOCM7oEYO1Tp1Tk7pz/1p6X52dvHs3LmVFi1a4OCg9cBlNDhJz77pKS8rxMcnsH79hgLX42oNEkQJIawjKUkb1JBSkHT1asbK9fBI+ZJbQIBlg1LykYQEbbBuSsFQ8nXWyrqQFhsb6wQxaR3v7Jy1WR3i48HD4yFFi8qsL8IyEkQJISwXHa3dN23ukltYmOW30yRnZweBgeYvuZUpo41UzueSkrS72lLrKdI/wsOtM8uLnqen1mFXvLj2b4kS4O2dyKVLJ6hTpzLu7nZpBjs5eUlGiJwkQZQQImW3bsFPP2G7bBmt//4b+4xccgMta3YqA7jzY0ZtpbQrlGldRtMvW3P8kJvb44AoeXD05Lpixcx35MXHJ7F+fRht2z4lPTJCpEKCKCGEsYcPtYysixbB+vWQkIANkOpFM2fnlAdwBwamPYA7D4mJSXmc0ZPLMTHWO6+DQ8rB0JPrXF2td14hRMokiBJCaN0mBw9qgdNPP5nNhBjj5YVj5crYlC1rGiwVL54vruckJWlXK8+c0T9sOHiwNjNm2HLzphYYWSmBOfB4buDUeov0Dw+PfPESC5GvSBAlREF25Yo2r8SiRdocCU/y9YW+fYnv3ZvNYWG0bdsWmzx+fUcp7Srl40BJmxrvzBn4778ncw7ZAunPfurllfaltBIltP3y4ZVMIQoMCaKEKGgePNBSKC9aBFu3mo5SdnaGLl3gpZe0CclsbbXblsLCcqS6GRUV9Tg4Sh4onTljOkWKJQoVsuxSWrFilmejFkLkbRJECVEQJCXB3r1a4LR8ufmcTI0bQ0gIdOtmOodILhUTA+fOmQ+Ubt5MX1n29lCuHJQvr002W6EClC6dwPnzO+jZsykeHnm7B04IYX0SRAmRn507B99/r00YFhpqur1MGa3HqW9f7XkulJCgdYKZC5QuXUrf7f46nTbOPXmgVKGCtuzvbzq/X3y84v79ByZztQkhBOSBIGr58uXMnj2bY8eO8fDhQ8qVK8cLL7zAiBEj0p1Z9f79+3zxxResWLGCM2fOEBMTg5eXF7Vr1+bVV1+lQ4cOWdQKIbJRVJTW27Rokdb79KRChaBHD63XqWHDXDFaOSlJy8VpbpzS+fPpv/3fx8d8oFSmjJbxWgghrCFXB1HDhw9n5syZ2NnZERwcjJubG9u3b2f06NGsWbOGzZs34+zsbFFZERERNG7cmJMnT+Lm5kb9+vUpXLgw//33H+vWrWPdunUMGzaMmTNnZnGrhMgCiYna+KZFi7TxTk/OyGpjAy1bar1OnTqRE10rSmmJIs2NUzp7Nv3pAAoXNg6S9IFS+fJanCiEEFkt1wZRq1atYubMmbi5ubFr1y5q1qwJQHh4OMHBwezdu5fx48czffp0i8p7//33OXnyJLVq1WLz5s14enoatq1fv56OHTvyxRdf0Lt3b+rVq5clbRLC6k6e1AKnH34wP63KU09pPU4vvAAl03+XWUbcvZvygO70TkPi7Gzco5T8uZdXruhEE0IUYLk2iProo48AGDNmjCGAAvD29mbOnDk0atSIWbNmMX78eDw8PNIsb/v27QCMHj3aKIACaNu2Lc2aNWPLli388ccfEkSJ3C08XMvltGgRHDpkut3LC3r31oKnWrWyJNKIjdUus5m7/Hb9evrKsrPTLrOZC5R8fbN2rjQhhMiMXBlEXblyhQMHDgDQp08fk+0NGzbEz8+PS5cusX79enr37p1mmU4WDoTw9vZOX2WFyA4PH2rZwxctgnXrtJQDydnZwfPPa4HT889b9R77iAhYutSGdeuqMnu2LWfPwoUL6Z+/zd/ffKAUGGg6oFsIIfKCXPnVdeTIEQA8PT0pXbq02X1q167NpUuXOHLkiEVBVJs2bTh48CAff/wxzZs3N7mct2PHDkqUKCGDy0XuoRQcPvw4i3h4uOk+tWppgVPv3mDF/wAoBX/8AXPnwrJlEBdnC6R9916xYuYDpbJltUtzQgiRn+TKICr00a3Y/v7+Ke7j5+dntG9aRo8ezf79+9m0aRMBAQE0aNDAMLD80KFDNGjQgG+//daiS4NCZKmrV2HJEi14+ucf0+0+PvDii9og8SpVrHrqu3e14VVz58Lff5vfx93dfKBUvrw2NYkQQhQUuTKIuvsoEaBrKrNouj2a0DTawtTDrq6urFmzhnHjxjFjxgw2bdpk2Obl5UWLFi0oacHA27i4OOLi4gzL+vPHx8cT/+QllnxM39aC1GbIwnbHxKBbvRqbH35At3UruqQko83KyQnVoQNJffuimjd/fP3LSvU4dgzmzbPhp59suHfPeAxVkSKKF19MoESJv+jduyYlS9qnOMwqv30c5HMu7S4ICnq7MyNXBlFZ4dq1a3Ts2JG///6bDz74gN69e1OsWDFOnjzJ//73P9577z1WrVrFnj17KJTK/dFTpkzhvffeM1m/Y8cOXApgRr4tW7bkdBVyhFXarRSe//6L344dlPz9d+wePDDZJaJSJS41a8aVBg1IcHXVUhls3pz5cwNxcTbs21eSjRsDOX3a02R7xYq3ad06jPr1r+DoqAV1f/+9JcUeqvxMPucFi7S7YHhg5js3vXJlEKUPYu7fv5/iPvfu3QPA3cLpKUJCQjhw4ADTpk1j1KhRhvV16tRh7dq11KpVi2PHjjF9+nSzQZLe2LFjGTlypGE5OjoaPz8/mjVrhpeXl0V1yQ/i4+PZsmULLVu2THfS07zMKu0OC8Pmhx+wWbIE3blzJptVYCBJL7xA0osv4l62LJWBypmrtpGzZ2H+fBsWL7bh9m3jLiVXV0WfPkkMHJhE9eqFgKpAVXm/pd0FgrS7YLU7IiIi02XkyiAqMDAQgEuXLqW4j36bft/UXLlyxRBhmxuEbm9vT7du3Th+/Dhbt25NNYhydHTE0dHRbBkF6cOnJ+220N27WhbxxYth1y7T7W5u0L07hISga9QIWxsbbK1XXeLj4bff4KuvYNs20+1Vq8Jrr8ELL+hwd7eFFM4u73fBIu0uWApau63R1lwZRNWoUQPQosTQ0FCzd+gdPHgQwCiHVEouXrxoeJ5Sz5V+QPnt27fTXV8hzEpMhO3btQHiv/5qmpJbp4PmzbW76zp3hlTGAGbUpUswfz588w1cu2a8zcFBm/1l8GCoX18SVwohRHrlyiCqVKlS1KlThwMHDvDjjz/y7rvvGm3fu3cvly5dwtHRkbZt26ZZXvIB43/99RctW7Y02efPP/8ESDGlghAWO3VKC5y+/x6uXDHdHhSkBU4vvgilSln99ElJ2rCpr76CtWu15eTKltUCp379rJoVQQghCpxcmwt43LhxAEydOpXDhw8b1kdERDBkyBAAXn/9daOUBCtXriQoKIjmzZsbleXv70+dOnUAePPNNwkLCzPa/sMPP7B06VLAfHJPIdIUEQGzZ0PdulCpEkydahxAFSkCQ4bAX39pU7WMGWP1AOrWLfj4YyhXDtq00S7f6QMoW1uts2vzZi2r+NtvSwAlhBCZlSt7ogA6derEsGHD+OKLL6hXrx7NmzfH1dWVbdu2cefOHRo0aMDkyZONjomKiuL06dPEPjn5KvDdd9/RrFkz/v33XypVqkS9evXw9vbm33//5Z9HuXhefPFFXnjhhWxpn8gH4uNhwwat12nNGvNZxNu00Xqd2rUDM2PpMksp2LtX63VasUJLbJ5cyZIwcCC88kq2TZ0nhBAFRq4NogBmzpxJgwYNmD17Nvv27SM+Pp6yZcsyZswYRowYgUM6praoUqUKJ06c4LPPPmPDhg0cOHCAuLg4ihQpwnPPPceAAQPo0aNHFrZG5AtK4XH+PDZvvQU//6x1/zypenUtcOrTR0vhnQWiorSrhXPnms/H2aqVNlC8XTuZUkUIIbJKrv967dGjh8XBTb9+/ejXr1+K24sXL87UqVOZOnWqlWonCpQzZ7Dr1Yumj6YlMlK8+OMs4k8/nWVVOHxY63X68Ud4MsWJlxcMGACDBmnjnoQQQmStXB9ECZErxMdD9+7okmeadHSEjh21XqdWrbKsy+fBA1i6VAueHs3LbaRBA63XqWtXsHCebSGEEFYgQZQQlvj0U8NkcveLF8dpwgRse/fWBoxnkX//ha+/1oZc3bljvK1QIejbV7vLrmrVLKuCEEKIVEgQJURa/vsPJk0CQNnYcHDUKOoPHIhtFiSle/gQVq3Sep127jTdXr261uvUp4+Wn1MIIUTOkSBKiNQopXX3PLrjM+mNN7hTrpzVTxMWpiXF/PZbuHHDeJuTE/TsqQVPzzwjSTGFECK3kCBKiNQsXvx4npSAAJImToTdu61SdGIibNyo9TqtX6/Fa8lVqKDFbyEh4Gk6P7AQQogcJkGUECm5eROSTTbNnDlWuYZ2/Tp89x3MmwcXLhhvs7PTkmIOHgzNmkmvkxBC5GYSRAmRkpEjQT+XYq9e0LataUJNCymlzTv81VfaNHoJCcbb/f3h1Ve1FAU+PpmstxBCiGwhQZQQ5mzaBEuWaM+LFIHPP89QMZGR2hXBuXO1KfWS0+m0hOaDB2vxma1t5qoshBAie0kQJcST7t/XIhu9GTO0ZJoWUgoOHtR6nX7+GWJijLcXKwYvv6xNxyLzXQshRN4lQZQQT5o4UbtdDrSBSalkwU/u/n346ScteEo2Z7ZBkyZabNalC6RjxiIhhBC5lARRQiR3+DB89pn23NFRy3aZxujuf/7RLtctXgzR0cbbPDy0u+sGDYKnnsqiOgshhMgREkQJoZeQAK+8AklJ2vKECVC+vNld4+Lgl1+0Xqc9e0y3166t5XXq2RNcXbOwzkIIIXKMBFFC6M2cCfrJhatUgVGjTHY5fx4WL36KgQPtuHXLeJuzs5ZJfPBgLYgSQgiRv0kQJQRAaKjW8wTa5btvvoEnpnWZPRvefNOOxETj3qlKlbRep759oXDhbKqvEEKIHCdBlBBKaVHQgwfa8uuvQ926RrvMmgVvvAGgjY+yt1d07arjtdegUSNJiimEEAWRBFFC/PijlhcKoFQp+PBDo82PAyhNu3bn+Oorf0qVsv4ExEIIIfIOCaJEwRYRAcOHP16eMwcKFTIsPhlAjRuXSJ06Jyhe3D/76iiEECJXssnpCgiRo956C8LDtefdukH79oZNs2cbB1Djx8PEiUly6U4IIQQgQZQoyLZtg0WLtOceHvDFF4ZNs2drQ6P0/vc/eO89GfskhBDiMQmiRMEUE6NlwNSbNs0w86+5AOr99yWAEkIIYUyCKFEwvf8+nDunPW/USEuyiQRQQgghLCdBlCh4jh2DTz7Rnjs4wLx5YGMjAZQQQoh0kSBKFCyJiTBwoPYvwLvvQlCQSQD17rsSQAkhhEidBFGiYJk1Cw4c0J5XqgSjRzNnjmkANXmyBFBCCCFSJ0GUKDguXNAiJL3585nzrSNDhz5eJQGUEEIIS0kQJQoGpWDoULh/X1sePJg5xxoYBVDjxkkAJYQQwnISRImCYdkyWLdOe+7ryzflppoEUB98IAGUEEIIy0kQJfK/yEgYNsywuLHdLAa+7WFYlgBKCCFERkgQJfK/UaPg5k0AzlfrRJt5nQ2bJIASQgiRURJEifxt50749lsAHjoVotGxWYZNY8dKACWEECLjJIgS+VdsrNHULsNjp3KVkoAWQH34oQRQQgghMk6CKJF/ffghnDkDwO/UZy6DAQmghBBCWIcEUSJ/OnECpk4F4CH2vMo8FDaMGSMBlBBCCOuQIErkP0lJ2tQuCQkATGUMJ6nMmDHw0UcSQAkhhLAOCaJE/vPVV/DnnwCcoiIfMU4CKCGEEFYnQZTIXy5f5uHbYw2Lg/iaEWOcJIASQghhdRJEifxDKULbvY5D7F0A5vMKz45uIgGUEEKILCFBlMg3Nr+2ktLHVgNwneJcfXMaU6ZIACWEECJrSBAl8oWFn9+hytevG5a3dfiCCZ8VkQBKCCFElrHL6QoIkVnz5kHSiDH4cg2Ak2Xb0WdldwmghBBCZCkJokSeNm8eLB60l718DUCcvRuVts1GZyMRlBBCiKwlQZTIs+bPhzcGxXGUgYZ1Dp98iC7APwdrJYQQoqCQMVEiT5o/H159FcYwlUqcAkA98wy614fmcM2EEEIUFBJEiTxHH0AF8S/j+AgAZWeHbv58sLXN4doJIYQoKCSIEnmKPoDSkcQ8XsWRhwDo3n4bnn46h2snhBCiIJEgSuQZ33yjBVAAA5lPI/ZqC2XLwoQJOVcxIYQQBZIEUSJP+OYbbU5hAB+u8rnDO483fv01ODvnTMWEEEIUWBJEiVwveQAFsLH8MJwfRmsL/fpB8+Y5Ui8hhBAFmwRRIlf79lvjAGpBp9U8fXaFtlC0KEyfnjMVE0IIUeBJECVyrW+/hVdeebw87vVoQg4kS2Hw+efg5ZXt9RJCCCFAgiiRSz0ZQL31Fnyg3kV35Yq2onVr6N07ZyonhBBCIBnLRS5kLoD6pMsf6BrO1la4uMBXXyGT4wkhhMhJEkSJXOW774zHQL31Fnzy4UN0tV8FpbSVkydDYGCO1E8IIYTQk8t5Itf47jutB0ofK40cCZ98Arrpn8CJE9rKWrVg2LCcq6QQQgjxiARRIlcwF0BNnw66s2e0nifQpnSZPx/spANVCCFEzpMgSuS4FAMoFAwaBHFx2oYRI6BGjZyrqBBCCJGMBFEiRz0ZQI0Y8SiA0j3auHOntqF0aZg0KYdqKYQQQpjK9UHU8uXLadq0KUWKFMHV1ZVq1aoxbdo04uPjM1zm6tWr6dChAyVKlMDBwYFixYpRv3593n//fSvWXKTFXAA1Y8ajAOrGDXj77cc7z50Lrq45Uk8hhBDCnFwdRA0fPpwePXrw+++/88wzz9C6dWsuXrzI6NGjCQ4OJiYmJl3lPXz4kB49etCpUye2bt1K5cqV6datG1WqVOHcuXN88cUXWdQS8aQFC1IJoADefBPu3NGev/gitGqVE9UUQgghUpRrR+iuWrWKmTNn4ubmxq5du6hZsyYA4eHhBAcHs3fvXsaPH8/0dEz7MXDgQJYvX06nTp2YP38+3t7ehm1JSUns37/f6u0QphYsgJdfTiWAWrcOli7Vnnt5waef5kg9hRBCiNTk2p6ojz76CIAxY8YYAigAb29v5syZA8CsWbOIioqyqLxt27axePFiqlSpwrJly4wCKAAbGxvq1atnpdqLlDwZQA0f/kQAde8eDBny+IBPP9XmyBNCCCFymVwZRF25coUDBw4A0KdPH5PtDRs2xM/Pj7i4ONavX29RmV9++SWgXSK0t7e3XmWFxRYuNA2gPv30icTj48fDxYva8xYtoG/fbK6lEEIIYZlceTnvyJEjAHh6elK6dGmz+9SuXZtLly5x5MgReqcxh1piYiLbtm0DoHHjxly/fp2ff/6Z06dP4+joSI0aNejatStubm7WbYgwWLgQBgxII4A6cAD049KcnLTB5DK1ixBCiFwqVwZRoaGhAPj7+6e4j5+fn9G+qTl//jz37t0D4M8//2TIkCGGZb1Ro0bx888/ExwcnNFqixQ8GUC9+aaZACo+XpvvJSlJW540CcqWzeaaCiGEEJbLlUHU3bt3AXBN5ZZ2fa9RdHR0muVFREQYnr/88svUr1+f6dOnExQUxLlz5xg3bhzr16+nY8eOHD58mPLly6dYVlxcHHH65I/Jzh8fH5+ptAt5jb6tabV58WIdAwfaopQWMb3xRiLTpiWRkGC8n80nn2B77BgA6umnSXjjDS2wymUsbXd+I+2WdhcE0u6C2e7MyJVBlLUpfRcIULJkSTZt2oSjoyMA1apV47fffqN69eqcOHGCqVOn8u2336ZY1pQpU3jvvfdM1u/YsQMXFxfrVz6X27JlS4rbtm/348svaxgCqHbtzhEcfIING4z3c7l2jeBHr6mysWF3377cSaXc3CC1dudn0u6CRdpdsBS0dj948CDTZeTKIKpQoUIA3L9/P8V99Jfj3N3dLS4PoF+/foYASs/W1pZBgwbxxhtvsHXr1lTLGjt2LCNHjjQsR0dH4+fnR7NmzfDy8kqzLvlFfHw8W7ZsoWXLlmYH6i9erOPLL417oKZP90ene+ISrVLYtmmDzcOHACS9/jr133wzy+ufUWm1O7+Sdku7CwJpd8Fqd/KrVBmVK4OowMBAAC5dupTiPvpt+n3TKk+n06GUokyZMmb30a+/du1aqmU5OjqaBGEA9vb2BerDp2eu3YsWacOb9B2Aw4bB55/botPZmhaweDFs36499/fH9sMPsc0Dr6O83wWLtLtgkXYXDNZoa65McVDj0SSzERERKQ4cP3jwIIBRDqmUuLm5UbFiRUBL1mmOfr3coZc5ixZB//5PBlAp3GR365Y227DeV1+BvP5CCCHyiFwZRJUqVYo6deoA8OOPP5ps37t3L5cuXcLR0ZG2bdtaVGb37t0BUrxcp78W/Mwzz2SkygKtU8niAAq0AErfndqrF1j4XgohhBC5Qa4MogDGjRsHwNSpUzl8+LBhfUREBEMeZbR+/fXX8fDwMGxbuXIlQUFBNG/e3KS8YcOGUaRIEdavX8/XX39ttO3nn39myZIlhv1E+i1eDP36pSOA2rQJfvhBe16kiLazEEIIkYfk2iCqU6dODBs2jHv37lGvXj3atGlDt27dKFeuHMePH6dBgwZMnjzZ6JioqChOnz7NuXPnTMrz9vZm6dKlODk5MXjwYKpUqUL37t2pWbMmvXv3RinF+PHjLe7ZEo89GUC98UYaAdT9+zB48OPl6dOhePEsrqUQQghhXbk2iAKYOXMmS5cu5dlnn2Xfvn2sX7+eUqVKMXXqVLZv346zs3O6ymvZsiXHjh0jJCSEO3fusHr1ai5evEjbtm3ZtGkT77//fha1JP/64QedSQA1c2YaicYnTYKwMO1506baNUAhhBAij8lUEPXTTz9RpkwZNm7cmOI+GzdupEyZMvzyyy8ZOkePHj3YtWsXUVFRPHjwgOPHjzN69GgcHBxM9u3Xrx9KKcL0P9BmVKhQgYULF3L58mUePnxIeHg469ato1WrVhmqX0G2Y0cpXn7ZNn0B1OHDWrpyAEdH+PprmdpFCCFEnpTpIOrOnTupTpXSrFkzIiMjDWOORP7www86vviipiEP1OuvWxBAJSQYT+0yfjxUqJD1lRVCCCGyQKaCqL///punn37abK+QnqOjI9WqVePYoyk9RN63ZAmPeqAeB1BffGFBh9LMmVpPFECVKjBqVNZWVAghhMhCmQqirl+/TsmSJdPcr2TJkly/fj0zpxK5xK+/QkgIhgBqyJBEywKo0FCYMEF7rtPB/PmQSvAthBBC5HaZCqJcXFwsSpseERGRam+VyBvWr9fSOSUmasutW4fy2WdJaQdQSsFrr4F+nqKhQ6FevSytqxBCCJHVMhVEVa5cmd9//53bt2+nuM/t27fZu3cvQUFBmTmVyGE7dkDXrqCf9Pqll5J49dW/LRsT/tNPWl4ogJIl4cMPs6yeQgghRHbJVBDVtWtX7t+/z4svvmh2NuSYmBj69u1LTEwM3bp1y8ypRA7atw/at4fYWG25Rw/4+utEbCz59EREwPDhj5fnzAELJo0WQgghcrtMTUA8aNAg5s+fz6ZNm6hQoQJ9+vQx9DidOnWKn376iatXr1KxYkVDlnGRtxw+DG3aaPkxQQum9InGLfL229oceaB1ZXXoYPU6CiGEEDkhU0GUs7MzmzZtonPnzhw6dIgZM2YYbVdKUaNGDVauXImLi0umKiqy34kT0KoVREdryy1bwrJlYG//+LJeqrZtg4ULteceHvDll1lVVSGEECLbZSqIAm2y4P3797NmzRo2btzIhQsXAPD396d169Z06NABnSRTzHPOnIEWLR7PD9yoEaxaBU5OFhYQEwODBj1e/vhj8PGxdjWFEEKIHJPpIApAp9PRoUMHOsilmnwhLAyaN4cbN7TlZ56BtWshXZ2J778P+jkMGzbUkmwKIYQQ+UiunjtPZL8rV7QA6vJlbfnpp2HDhnSOBf/7b/jkE+25gwPMm4dlo9CFEEKIvEN+2YTBzZvaJbzz57XloCDYsgU8PdNRSGIivPLK42RS48ZBpUpWr6sQQgiR09J1Oa9MmTLodDq2bt1K6dKlKVOmjMXH6nQ6zukv74hc5/ZtbRD5qVPacpkysHUrFCuWzoJmz4YDB7TnlSrBmDFWracQQgiRW6QriAoLC0On0xH/6NassLAwi4+VweW5V3S0lsZAP71hqVLajXUWzOhj7OJFredJb948cHS0Wj2FEEKI3CRdQVRoaCiAYb48/bLIux48gHbtYP9+bbl4cS2ACgxMZ0FKwZAhjxNKDR6sDSgXQggh8ql0BVEBAQGpLou8JTYWOnWCPXu0ZS8v7RJehQoZKGz5cli3Tnvu4wNTp1qrmkIIIUSulKmB5QMGDOCdd96xVl1ENoqP16Zv2bJFW3Z316a3q1IlA4VFRsKwYY+XZ83SkmsKIYQQ+VimgqgffvhBLunlQYmJ0LcvrFmjLbu6amkMatXKYIHvvPM4qVTHjtC5s1XqKYQQQuRmmQqiSpQoIQPG85ikJC0DwdKl2rKjI/z2G9Svn7HydLt3wzffaAuFCmm9UPKZEEIIUQBkKohq2bIlv//+u+FuPZG7KQVvvPF4Ojt7e/j1VwgOzlh5Ng8fYvvaa49XTJmi3donhBBCFACZCqImTZpEXFwcAwcO5O7du9aqk8gCSsHo0TBnjrZsYwM//QRt22a8zArLl6M7e1ZbePZZSB5QCSGEEPlcpubOW7BgAa1bt2bx4sWsW7eOFi1aEBgYiLOzs8m+Op2O8ePHZ+Z0IhPef//xTCw6HSxaBF27ZqLAEyco/+uv2nN7e5naRQghRIGTqSBq0qRJhjFRERERLNUPtElGp9OhlJIgKgd98glMmvR4ee5cePHFTBSYlITtkCHY6Kd2GT06g7f1CSGEEHlXpoKoCRMmyMDyXG72bO3mOb3PP4dXX81kod99h82ffwKgypdH9+67mSxQCCGEyHsy3RMlcq8FC+D11x8vf/ghvPmmFQrW340HJM6Zg52TkxUKFUIIIfIWGcSSTy1dqqUy0Bs3znhauwy7eBH++guAqMBAVJMmVihUCCGEyHsyFUTZ2try8ssvp7nfwIEDsbPLVKeXSIffftPGPCUlactvvgkffGClwvWDyYGrzz5rpUKFEEKIvCdTQZRSCqWUxfuKrLd5M3TvDgkJ2vLAgfDZZ1bMf/nLL4anVzOaoVMIIYTIB7Llct6DBw+wt7fPjlMVaLt3axMKP3yoLb/wAnz1lRUDqCtX4PffAVBPPcU9Pz8rFSyEEELkPVkeRN25c4e9e/fi4+OT1acq0P76C55/HmJitOWuXbXM5La2VjxJskt5SV26WLFgIYQQIu9J90ClMmXKGC3/8ssv7Ny50+y+CQkJXL9+ncTERAYNGpShCoq0HT0KrVvDvXvactu28OOPYPVhaMku5SV16aINMhdCCCEKqHT/zIaFhRme63Q67t27xz39r7cZDg4OdOrUiY8++ihDFRSp+/dfaNkS7tzRloODtVjHwcHKJ7p+Hfbs0Z5XrAiVK0sQJYQQokBLdxAVGhoKaAPFy5QpQ7du3fhEP5/IExwcHChatKjcmZdFzp2D5s0hPFxbrl8fVq8GM7PuZN7KldoEfADdullxoJUQQgiRN6U7ugkICDA8DwkJoVGjRkbrRPa4eFELoK5d05Zr1oT168HNLYtOuHz54+fdu2fRSYQQQoi8I9MTEIvsd/06tGgBFy5oy5Urw6ZN4OGRRSe8eRN27dKelysHTz/9OIeCEEIIUUBZ5TqbUooNGzawb98+bt26Rd26dRkwYAAAt27dIjIykrJly2Jr1VvFCqbwcC2AOntWWy5fHrZuBW/vLDzpqlWPM3fKpTwhhBACsEIQdezYMXr27MnZs2dRSqHT6YiPjzcEUVu2bKFv376sWrWK9u3bZ7rCBdmdO9CqFfzzj7YcEADbtkGJEll84mR35dGtWxafTAghhMgbMpUn6vLly7Ro0YIzZ87Qpk0bpk2bZpKZvFOnTtjb27N69epMVbSgu3dPS11w5Ii27OsL27dDlue7jIjQTgQQGKgNvhJCCCFE5oKojz76iIiICD7//HPWrl3L22+/bbKPi4sL1apV48CBA5k5VYEWEwMdOsAff2jLRYtqPVBPpOzKGqtWQWKi9rx7d7mUJ4QQQjySqSBq48aNBAUFMWzYsFT3CwwM5Jr+NjKRLnFxWvbxHTu05SJFYMsWCArKpgrIpTwhhBDCrEwFUVevXqVq1app7qfT6YiOjs7MqQqkhATo0wc2bNCWCxWCjRuhWrVsqkBkpDZqHbTrhnXqZNOJhRBCiNwvU0GUq6srt27dSnO/0NBQPD09M3OqAicxEfr1ezxdnbMzrFsHzzyTjZX47bfHqQzkrjwhhBDCSKaCqKpVq3Lo0CHC9Smzzbhw4QLHjh2jVq1amTlVgaIUDB4MS5Zoyw4OWibyRo2yuSJyKU8IIYRIUaaCqBdffJG7d+/yyiuv8ODBA5PtDx8+ZMiQIcTHx/Piiy9m5lQFhlIwfDh88422bGenxTItW2ZzRaKiYPNm7XnJklCvXjZXQAghhMjdMpUnqn///ixZsoTffvuNoKAgWrduDWi5o4YNG8Zvv/3GxYsXadGiBT179rRKhfO7d9+FL77QntvYaL1ROZJea80aePhQe961q1YZIYQQQhhk6pfR1taWNWvW0Lt3b65cucI3j7pPjhw5wqxZs7h48SJdu3blV/3AHpGqDz+EKVMeL3/3HfTokUOVkUt5QgghRKoynbHczc2NJUuWMH78eNavX8/58+dJSkrCz8+PNm3aUL16dStUM//7/HP43/8eL8+eDSEhOVSZu3e12wBBS4dev34OVUQIIYTIvawydx5AUFAQQdmWvCh/mTcPRox4vPzJJzBkSM7Vh3XrtARVAF26gMx5KIQQQpiQgS457IcftDvx9CZNAjOJ37PX8uWPn3fvnnP1EEIIIXKxdPVEOTg4ZPhEOp2OOH3vhgC0YUchIdodeQCjRsGECTlbJ+7dg/XrtedFi+ZAXgUhhBAib0hXEJWgT7woMm3dOujdG5KStOWhQ+Hjj3NBPssNGyA2Vnsul/KEEEKIFKV7TJROp6NOnToMGDCAVq1aocvxX/28Z9s2LWuAPibt319La5ArXkq5K08IIYSwSLqCqI8//pgFCxawf/9+Dhw4gJ+fHyEhIfTv35/AwMAsqmL+8vvv0KHD43HbvXrB/Pm5JA3TgwdaFxmAlxc0bZqj1RFCCCFys3T9dI8aNYqTJ0+yd+9e+vXrx+3bt5k8eTLlypWjRYsW/PjjjzLuKRUHD0LbtlqsAtCxIyxenIuumG3cCPfva887d9bSpQshhBDCrAz1f9SvX59vv/2Wa9eu8c0331CvXj22b99O3759KVGiBEOGDOHAgQPWrmuedvw4PPccREdry61awdKlYG+fs/UyIpfyhBBCCItl6iKSq6srAwYMYO/evZw6dYq3334bJycn5s6dS7169WjYsKG16pmnnTmjzX13+7a23LgxrFwJjo45Wy8jsbHaVC8ARYpAcHDO1kcIIYTI5aw2EqdChQp8/PHH/Pvvv7Rv3x6lFGfOnLFW8XlWaCg0bw43bmjLzzwDa9eCi0vO1svE5s1aegPQrjPmqi4yIYQQIvexWhC1Z88e+vfvj5+fH2vXrsXGxobGjRtnutzly5fTtGlTihQpgqurK9WqVWPatGnEx8dnuuz169ej0+nQ6XS0aNEi0+U96fJlLYC6fFlbrlZNG3ZUqJDVT5V5cilPCCGESJdMjRy+du0aCxcuZOHChfz3338opShdujT9+vWjX79++Pn5Zapyw4cPZ+bMmdjZ2REcHIybmxvbt29n9OjRrFmzhs2bN+Ps7JyhsiMjIxk4cCA6nQ6lz3ZpRTduQIsWWk8UQKVKWmdPkSJWP1XmxcXB6tXacw8PreJCCCGESFW6g6iEhARWr17Nd999x+bNm0lMTMTZ2Zk+ffowYMAAmjVrZpWKrVq1ipkzZ+Lm5sauXbuoWbMmAOHh4QQHB7N3717Gjx/P9OnTM1T+G2+8wY0bNxg8eDBfffWVVeqsd/u2Ngbq9GltuWxZ2LoVihWz6mmsZ+vWxyPeO3TIZYO1hBBCiNwpXZfzRowYga+vLz169GDDhg3UqFGDOXPmcO3aNb7//nurBVAAH330EQBjxowxBFAA3t7ezJkzB4BZs2YRFRWV7rJXrlzJkiVLGDlyJM8884x1KvxIdDS0bq3djQfg56cl1/T1tepprEsu5QkhhBDplq6eqJkzZ6LT6ahduzYDBgygatWqAJw4ccKi4+vXr2/RfleuXDGkSOjTp4/J9oYNG+Ln58elS5dYv349vXv3trAFWk/W4MGDqVixIu+//z4///yzxcem5f59eP550Gd3KFFCC6ACAqx2Cut7+BBWrdKeu7lpuReEEEIIkaYMjYk6ePAgBw8eTNcxOp3O4rn3jhw5AoCnpyelS5c2u0/t2rW5dOkSR44cSVcQ9dprrxEeHs6vv/6Kk5OTxcelJTZWyz6+d6+27OWlXSUrX95qp8gaO3bAnTva8/btwYqviRBCCJGfpSuI8vf3z5a58kIfjcb29/dPcR/9oHX9vpb4+eef+eWXX3jzzTdp0KBB5ir5hP79bdm6VXvu4aENIq9c2aqnyBrLlz9+3r17ztVDCCGEyGPSFUSFhYVlUTWM3b17F9CSeabEzc0NgGj9gOg0XL9+naFDh1K2bFnDeKuMiIuLM5raRn/+LVu04WWuroo1axKpWlVhhSwMWSs+HruVK9EBytWVhObNsbTS+hQT1kg1kZdIu6XdBYG0W9pdEFijvQVmcrRXX32VyMhIVqxYgUsmMl1OmTKF9957z+w2B4dExoz5k9u3w1m/PsOnyDZFjx2j/qM06lerV+fgjh3pLmPLli3WrlaeIO0uWKTdBYu0u2B4oJ/INhNyZRBV6FE2yvv6yXDNuPcou7a7u3ua5S1atIg1a9bw2muv0bRp00zVbezYsYwcOdKwHB0djZ+fH3Z2ihUrFM89Z927/bKSzbp1hufFhw6lbdu2Fh8bHx/Pli1baNmyJfa5ILt5QkKCxWPuMnueffv2Ub9+fewK0ATN0m5pd0Eg7c6b7baxscHe3j7dw40iIiIyfe5c+WoFBgYCcOnSpRT30W/T75ualStXAnDgwAGTIOr69esAHDp0yLDt559/pkSJEmbLcnR0xNFMHqVvvkmkXbtc+XKal5j4+K48Z2fsOnTI0FQv9vb2ORpERUdHEx4ebnSJNSsppShRogTXrl3LlvGBuYW0W9pdEEi78267bW1tcXFxoVixYjg4OFh0jDV+u3Llr36NGjUALUoMDQ01e4ee/u7A5Dmk0pLaHYV37txh165dAMTGxqanugC0a2f9rOdZavduuHVLe962LaQy/iy3io6O5sqVK7i5ueHt7Z2h/4mkV1JSEvfu3cPNzQ0bG6vNmpTrSbul3QWBtDvvtVspRWJiIjExMURFRREWFkapUqUyNWwnPXJlEFWqVCnq1KnDgQMH+PHHH3n33XeNtu/du5dLly7h6Oho0SWoVfoeFzMWLlxI//79ad68OVv1t9cVBPkgwWZ4eDhubm6UKlUq2/73lJSUxMOHD3FycspzXzaZIe2WdhcE0u682243Nzc8PT25cOEC4eHhqd7db0259tUaN24cAFOnTuXw4cOG9REREQwZMgSA119/HQ8PD8O2lStXEhQURPPmzbO3snlNYiL8+qv23NFRyxCax8THxxMXF4eHh0ee7X4WQghhPba2tnh6enL//v1sGSMLubQnCqBTp04MGzaML774gnr16tG8eXNcXV3Ztm0bd+7coUGDBkyePNnomKioKE6fPp2hy3EFyr598GgsGK1bw6OB/HlJYmIiYJ1r2kIIIfIH/ZjlhISEbBkkn2t7okCbZmbp0qU8++yz7Nu3j/Xr11OqVCmmTp3K9u3bcXZ2zukq5k354FKenvRCCSGE0Mvu34Rc2xOl16NHD3r06GHRvv369aNfv37pKj8jx+RpSUmwYoX23MFBm+pFCCGEEOmWq3uiRBb480+4ckV73qqVNkeNEEIIIdJNgqiCJh9dyhNCCCFykgRRBYlSj4MoOzvo0CFn6yOEECJLhYWFodPpLEpMnZqdO3ei0+kyPetHfiNBVEFy4ADos8C3aAFFiuRsfYTIBeTHQQiRUbl+YLmwouSX8rp3z7l6CCGEyBYlS5bk33//zXQ6mGeeeYZ///032zKB5xUSRBUUSsHy5dpzW1vo2DFn6yOEECLL2dvbExQUlOlyXFxcrFJOfiOX8wqKw4chLEx7HhwMXl45Wh2RfR48eMDnn39Ow4YNKVKkCI6OjgQEBNC+fXt+/PFHk32nTp1KzZo1KVSoEC4uLlSuXJnx48dz584dk7KTj7dQSjFv3jxq1aqFq6srHh4etGrVij/++MMqdWvatCk6nY6dO3eyZ88e2rdvT9GiRbGxsWHhwoWG/WJiYpgxYwb16tWjcOHCODk5UbFiRd555x2TWdubNm1Ks2bNANi1axc6nc7wMDeGZNu2bXTp0gUfHx8cHBwoVqwYnTt3TrWNltq/fz/vvPMOzzzzDCVKlMDBwYHixYvTvn17s1NSjR07Fp1Ox+DBg1Ms88SJE+h0OooXL058fLzRtqtXrzJy5EgqVaqEi4sLhQoVok6dOsyaNctstud+/fqh0+lYuHAhJ06coGfPnvj4+GBra8ukSZMAbSaBH374gRdeeIGgoCDc3d1xdnamYsWKDBs2jKtXr6ZY14iICIYNG4a/v7/hczB8+HDu3LljdG5zsvJ90X8eAObPn2/4fBcuXJi2bdvy559/mj0uMDAQnU5HWFgYq1evJjg4GE9PT8NnWC8yMpKJEydSvXp1w99c1apV+eCDD3jw4EGK9Tp06BAhISGULl0aJycnPD09qVatGqNGjeLChQuG/VIbE3X27FkGDBhA2bJlKV68OO7u7gQEBPD888+zYMECo33Tuux96tQp+vfvT0BAAI6Ojnh6etK8eXOWLVtmdv9Jkyah0+mYNGkSt27dYujQofj5+eHg4ICfnx9vvPGG2e+cXEeJTImKilKACg8Pz+mqpG7MGKW0/iilvv4608U9fPhQrVq1Sj18+NAKlUu/mJgYdfLkSRUTE5Ot501MTFSRkZEqMTExW8+bURcvXlRPPfWUApSLi4tq2bKl6tWrl2rUqJHy8PBQAQEBhn0jIiJU9erVFaDc3d1Vhw4dVNeuXZW3t7cCVEBAgDp37pxR+aGhoYZtISEhyt7eXgUHB6sePXqoChUqKEA5OjqqP//8M1N1U0qpJk2aKEANGTJE2djYqKeeekr16tVLtWrVSv34449KKaWuXLmiqlatqgDl6empWrRooTp37qwCAgIUoAIDA1VYWJihzClTpqjnnntOAap48eIqJCTE8HjrrbeM3u+33npLAcrGxkY988wzqnv37qpu3bpKp9MpW1tb9d1332XqvWrevLmysbFRVatWVW3btlXdu3dXNWvWVIAC1Oeff260/+nTpxWgChcunOLfwciRIxWgRo4cabR+165dqkiRIobXpEOHDuq5554zrGvZsqW6efOm0ec8JCREAWrgwIHK0dFRBQYGqh49eqj27dur6dOnK6WUunTpkgKUh4eHqlevnurevbtq27at8vX1VYAqWrSoOnv2rEk9r169qsqWLWt437p06aI6deqkihQpoipWrKg6deqkALVgwQKTY635vpj7+9a//iNGjFA6nU41bNhQ9e7dW1WpUkUBys7OTv36668mZek/c6+//roCVO3atVXv3r1VkyZN1O7du5VSSv3zzz/Kz89PAcrHx0e1bt1atW/fXhUvXlwBqnr16urOnTsmZU+bNk3Z2NgoQFWoUMHwPlSqVMnkdUr+N5rc8ePHlbu7uwJUxYoVVfv27VW3bt3Us88+q9zc3FS1atWM9t+xY4cCVJMmTUzqs3btWuXk5GQoq1evXio4OFjZ2toqQA0YMMDkmIkTJxq2lSpVShUvXlx16dJFtW3bVnl4eChA1alTJ92/Men5bQgPD1eAioqKStc5kpMgKpPyRBCVlKRUuXJaAGVjo9SNG5kuUoKo3B9EJSYmqtq1aytAtWrVSt28edNoe0xMjFq3bp1huWfPngpQdevWNfo83717V7Vu3VoBqn79+kZl6L+g9V/Sp0+fNmxLSEhQAwYMMJw/M3VT6nEQBajZs2ebtDcpKUk1aNBAAerll19W0dHRhm3x8fGGH9tmzZoZHZfaj4P+/Z47d64CVLly5dSxY8eM9tm1a5cqVKiQcnBwUGfOnDEpw1Lr169XV69eNVm/b98+5e7uruzt7dXly5eNtunb+9NPP5kcFx8fr4oVK6YAdfz4ccP6a9euKS8vL6XT6dScOXOMPsvh4eEqODhYAWrs2LFmgyhAjRkzxuzfQHR0tFq9erWKi4szWv/w4UM1duxYBai2bduaHNe5c2cFqKZNmxr9oEVGRqqGDRsazvtkEDVv3jyrvi+pBVHOzs5q27ZtRvtPmzbNEDTeeOJ7VR9E2draqtWrV5uc68GDB4bA8X//+5/Ra3b//n3Vu3dvBaj+/fsbHbd69WoFKCcnJ7V06VKTcv/55x918uRJw3JKQVT//v0VoD744AOTdj948EDt2rXLaP+U/k6uX79uCHo++OADlZSUZNh24MABQ2A+b948o+P0QRSg+vXrp2JjYw3bLl68qEqWLKkAw3+QLCVBVB6TJ4Koo0cf90IFB1ulyLwQRNWqpVTJktZ+JClf30RVsmRSFpStPWrVss5rtGrVKsP/cO/evZvqvhcuXFA2NjZKp9OZ/BgppX2p6f+n+fvvvxvWJw+ifvvtN5Pjrl27ZuiNSv5ZSU/d9PRBVHAKn+ENGzYY/vceHx9vsj0xMdHQe5A8qEgriIqIiDD0pBw8eNDsufU/pm+99ZZFbUkvfQDyZPD47bffmg1SlXr8GteuXdto/ejRow09JOZcvnxZ2dvbK29vb5WQkGBYrw+iKlSoYLQ+PXx9fZWNjY1RgBsWFqZ0Op2ysbFR//77r8kxx48fVzqdziSISkxMtPr7kloQNXz4cLPH6P8z8OGHHxqt1wdR5nphlFLqq6++UoBq166d2e13795VxYoVU3Z2dur27duG9fre4hkzZljUppSCqLZt2ypAHT582KL/HKb0dzJ58mQFqFopfHFNnz5dAap8+fJG6/VBVKlSpdT9+/dNjps6dWqqr19KsjuIkoHlBYF+QDkUqASb168/Ts5uPbpHj9xv48aNAPTp0wc3N7dU9929ezdJSUnUrFmTp59+2mR7yZIlCQ4OZv369ezYsYP69esbbbezs6N169Ymx5UoUYIiRYoQGRlJREQEJUqUSHfdntQthc/wunXrAOjatavZiUdtbGxo3LgxJ06cYN++fVSpUsWi8/39999cvXqVsmXLUqtWLbP76MeJ7Nu3z6IyUxIREcG6des4ceIEkZGRhnFMZ8+eBeD06dNG+/fo0YNhw4axdetWLl++TKlSpQzb9GNaBgwYYHSM/nXq2bOn2TqULFmS8uXLc/LkSc6ePWsymLhTp07Y2tqm2o5jx46xbds2QkNDuX//PklJSYA2KWxSUhL//fcfNWrUAGDPnj0opahVq5bZgctVqlTh6aef5tixY0brjxw5km3vC0BISIjZ9S+99BIHDx5k586djBs3zmR7Wp/XlN4HNzc3ateuzfr16zlw4ACtWrXi+vXrHD16FBsbG15++eUMtkTzzDPPsH79el577TUmTpxIjRo1cHd3T3c5+vFdKb0+L7/8Mm+//TZnz57l6tWr+Pr6Gm1v3ry52Tv+KlWqBMAV63+JW5UEUfld8rvydDro3Dln65ONHv1eW5lCKfVooGnWBFPWqrd+cKkld9Tov6hKly6d4j76bea+1Hx8fFK8hdrd3Z3IyEhiY2MzVLcnpZQ08Pz58wCMHz+e8ePHp1rGrVu3LD5f2KMbMs6dO5fm5KbpKfdJ8+fPZ8SIEdy/fz/FfaKjo42W3dzc6N69OwsXLmTx4sWGH/GbN2+ybt06nJyc6N27t9Ex+tepUaNGadbp1q1bJu9Rakkb79+/T9++fVm5cmWq5SZvx+XLl9MsNzAw0CSI0rcjq98XvZT+NvTr9e14Ulqf1759+9K3b99Uz62v/8WLFwHt780jk1N2jRo1ir1797J161batm2Lvb091apVo3HjxvTq1Ys6depYVE5a3x2FCxfG09OT27dvc/nyZZMgyt/f3+xx+oAu+fdGbiRBVH73zz9w5oz2vFGjrIoscqWDB61fZlKSIjo6Gnd3d2xs8kaPVHawscm+G32dnZ3Nrtf3djRs2JCyZcumWkblypUtPp++3BIlSvDcc8+luq+3t7fF5SZ36NAhBg0ahK2tLR9//DHt27fH398fFxcXdDod8+bNY9CgQSilTI4dMGAACxcuZNGiRYYg6ocffiAhIYFu3bpRuHBhs+3p1q0brq6uZuujlCI+Ph4vM3fxpvT6g3bH4MqVKwkKCmLq1KnUqVMHb29vHBwcAKhfvz5//PGH2XakFgiZ25Yd70t6mGsTpP15bd26NcWLF0+17ICAgMxVzgwXFxe2bNnCgQMH2LBhA7t37+bAgQMcPHiQTz/9lCFDhjB79myrn/dJ2fndkRUkiMrvZK68Akv/P7xTp06luW/JkiWBx/87NkffI6PfN7vqZik/Pz8AOnbsyNtvv221cvXt9fLySvEW+8xavnw5SineeOMN3nnnHZPt+st55jRq1Ihy5cpx5swZfv/9dxo0aGCo55OX8kB7nc6ePcvo0aOpXbu22TKTkpIM/1lID/3t7EuXLjV7WdhcO/Svr/7zZY65bfr3Oyvfl+RCQ0OpXr26yXp93ZJfSrWEn58fp06d4uWXX07xkt+T9H83165dIyoqKtO9UQB16tShVq1aREdH4+Liwm+//cZLL73EnDlz6NatmyEFSEpKlizJqVOnUvzuiIqK4vbt24Z985u8HQKKtCUPorp2zbl6iGynH6P0008/pXqJCKBx48bY2Nhw9OhRk8smoH1pb9u2DSDNL1Vr181Sbdq0AR4HJJbS95KYy40EULNmTby9vTl58iT//PNP5itqhv5HxlyPQ2xsLCtWrEj1+P79+wOwcOFCDh06xPHjx/Hz86N58+Ym++pfp5Ty92RGau3YtGkT4eHhJusbNWqETqfj0KFDnNH3midz8uRJs59JfS9XVr4vyX3//feprk/vtEEZeR9KlChBtWrVSEpK4rvvvkvX+SxhZ2dHt27dDD17R48eTfMYfbsXLVpkdru+nuXLl5cgSuQx//6rXc4DaNAAnrgWLfK3Dh06UKNGDa5evUr37t1NEk3GxsayYcMGQPsfbvfu3VFKMWjQIKN979+/z6BBg4iNjaV+/fomg8qzum6W6tixI3Xq1GH//v3079/f7DiYyMhI5s6daxQw6XsQzp49a5KQErSMzxMmTEApRefOndm7d6/JPomJiWzfvj3FxItp0Q+iXbRoEXfv3jWsj42NZciQIYSGhqZ6fEhICDY2NixbtsxwCUa/7kmjRo2icOHCfPrpp8yYMYOHDx+a7BMaGsrSpUsz3I4vv/zSaP3p06dTTAoaGBhI+/btSUpK4rXXXjNqf1RUFK+99prZoNje3p6JEydm6fuS3FdffWWUJBPgs88+Y//+/RQqVCjdA71fffVVAgICWL58OaNHjzZqt97169eZP3++0bqJEycC8O6775oNrk+ePMm///6b5vnnzJljcqOC/pwHH42FsOQy4sCBA3F3d+fw4cN89NFHRu/VkSNH+OCDDwDtc5cvZfi+PqGUyuUpDt5//3Fqg88+s2rReSHFQVbIS3milNJuH69YsaLiUULLVq1aqd69e6vGjRubJLQMDw9X1apVM+S96dSpk+rWrZsqWrSo4Rbp1JJtpkR/q3doaGiG66bU4xQHO3bsSPFcV65cMdwC7urqqurXr6969eqlunTpoqpXr25I/vfk50Z/m3rFihXVCy+8oF5++WU1evRoo/d71KhRhtvdK1eurDp27Kh69eqlmjZtqgoXLqwA9dVXX6X2dqQoMjLS8Dp5eXmpTp06qa5du6pixYqpQoUKqTfffFMBKiQkJMUy9Lm8AKXT6Uzeq+R27dplSKJarFgxFRwcrF544QXVrl07Q+6i2rVrm80TZS7hpd6KFSsM6QiqVq1qSLqoT8Jav359s+/hlStXVGBgoKH9Xbp0UZ07d1aenp6qfPnyqkOHDgpQS5YsMTmnNd+XtFIc6HQ61bhxY9W7d29DUldbW1u1fPlyk7JS+twnd+LECUO7CxcurBo3bqz69OmjOnXqpJ566iml0+lU8eLFTY778MMPDa9zUFCQ6tmzp+rQoYMhea0lyTb1f+ulS5dW7dq1U927d1ctW7ZUzs7OhlQiyVOFpJYKZM2aNYYUKEFBQap3796qefPmys7OzmyuK6UepziYOHGi2dcmtfOlRvJE5TG5Ooh6+unHQdSFC1YtWoKovBFEKaXlm/n4449VnTp1VKFChZSjo6MKCAhQHTp0UD///LPRvvfv31dTpkxR1atXVy4uLsrJyUlVqlRJjR07VoWGhpq0OzNBVHrrZkkQpZRSsbGxau7cuapZs2bKy8tL2dnZqWLFiqnq1auroUOHqk2bNpkcc+HCBdWnTx/l4+Nj+OIPCAgweb9///139cILL6iAgADl6OioChUqpCpUqKA6deqkvvnmG6N8Pul169YtNWTIEFW2bFnl6OiofH191YsvvqjOnj2rFixYkGYQtWzZMsMPviU/PDdu3FDjx49XNWvWNCSlLFWqlKpfv76aMGGC2rt3b7qDKKWU2r17t2revLny9vZWLi4uqkqVKurDDz9UcXFxqb6HN2/eVEOHDlWlSpVSDg4Oys/PTw0dOlRFREQYEoCae++Ust77kloQpZSW26l69erK2dlZubu7q9atWxvlTUvOkiBKKS1B6bRp09Szzz6rChcurOzt7ZWPj4+qU6eOGjVqlNq3b5/Z4/744w/Vu3dvVbJkSWVvb688PT1VtWrV1DvvvKMuJPu+T+lvdO3ateq1115TNWrUUEWLFjW8/02bNlWLFi0y+W5PK6g5efKkCgkJUaVKlVL29vaqcOHCqlmzZiZ/x3r5JYjSKZWOwQPCRHR0NB4eHoSHh5u9kyXHnDkDFStqz+vWBSt0ZycXHx/P+vXrDbfGZrfY2FhCQ0MN80Zll+QDbvP6XSXpIe2WdueUO3fuUKZMGaKiorhx40aW3mlnrt36OwPz809lbnq/Mys9vw0RERF4e3sTFRWVoRxZIGOi8q/k18q7d8+5egghhAX2799vsu7WrVuEhIQQGRlJu3btsiVVgRDpISkO8qvkWcrlrjwhRC5Xt25dSpUqRaVKlfDy8uLKlSscOXKEe/fu4e/vz6xZs3K6ikKYkCAqPzp3Do4c0Z7Xrg2pZAIWQljXqVOnmDp1qsX7jxkzJkOZ2/Ob//3vf2zbto1jx44RGRmJg4MDZcuWpV27dowcOTLTwyUseV/UoySj9vb2jB07Vt4XkSYJovKj5JfyJMGmENnq+vXrKebMMadfv37yYw1MnjyZyZMnZ1n56X1f+vfvT1BQUL4eCyUyT4Ko/EiylAuRY5o2bSo/vLmQJe9LfhpgLbKHfErymwsX4MAB7XmNGpDGHGJCCCGEyBgJovIb6YUSQgghsoUEUfmNBFFCCCFEtpAgKj+5dOlxUs2qVaFChZytjxBCCJGPSRCVn/z66+Pn0gslhBBCZCkJovKT5JfyJEu5EEIIkaUkiMovrl6F33/Xnj/1FFSqlLP1EUIIIfI5CaLyi19/BX0OFLmUJ4QQQmQ5CaLyC7krTwghhMhWEkTlBzduwO7d2vMKFaBKlZytjxBCCFEASBCVH6xc+fhSXvfuoNPlbH1ErrNmzRoaNWqEu7s7Op0OnU7Hzp07M1zekiVLeOmll6hWrRrFihXD3t4eDw8PnnnmGaZMmcK9e/dMjklKSmLfvn1MmDCBhg0b4uXlhb29Pd7e3rRs2ZIlS5bIdCkiVwsMDESn0xEWFpapcvR/gyLvk7nz8oPlyx8/l0t54glHjx6la9euJCUlERwcjI+PDzqdjhIlSmS4zK+++op9+/ZRqVIlatasiaenJzdu3OCPP/7gwIEDfPfdd+zatQtfX1/DMefPn6dBgwYAeHp6Urt2bYoUKcL58+fZunUrW7du5eeff2bFihU4ODhkut1CCJHVJIjK627dAn2PQtmyUK1ajlZH5D6rVq0iPj6ecePG8eGHH1qlzBkzZlC+fHk8PT2N1kdERNCpUyf27t3LW2+9xU8//WTYptPpCA4OZtSoUbRs2RJbW1vDtl27dvH888+zdu1apk6dyoQJE6xSTyGsadu2bcTHx1OyZMlMlfPvv/9aqUYip8nlvLxu1SpIStKed+sml/KEiYsXLwJQvnx5q5VZt25dkwAKwMvLi48++giAzZs3G20rW7Ys27Zto3Xr1kYBFECTJk0YM2YMAIsXL7ZaPYWwprJlyxIUFIS9vX2mygkKCiIoKMhKtRI5SYKovE7uyhMpmDRpEjqdjgULFgDQv39/w1iMpk2bEhYWhk6nIzAwkISEBKZNm0blypVxdnbG29ubHj16cOrUqXSf185O6+B2dHRM13E1atQA4NKlS+k+Z3IXLlzg448/Jjg4GH9/fxwdHSlcuDANGzbk66+/Jkn/n45HNm3ahE6no1IqudUSEhIoUaIEOp2OY8eOGW2LiYlhxowZ1KtXj8KFC+Pk5ETFihV55513iIiIMClr4cKF6HQ6+vXrx+3btxk+fDhly5bF0dGRpk2bGvbbunUrb7zxBtWrV8fb2xtHR0dKlSpFz549OXDgQKp1nTFjBlWqVMHJyYlixYrRvXt3Tp48aXRuc86cOcOgQYMoX748JUqUoEiRIjRu3JgffvghxfNZqmnTpoaxeLt27aJVq1Z4enri4uLCM888w/fff2/2uH79+qHT6Vi4cCEnTpygZ8+e+Pj4YGtry6RJk4za/c0339C0aVM8PT1xdHSkdOnSvPbaa6l+pq5cucKoUaOoWrUqHh4elCxZkqCgIPr168e+ffuM9k1pTFRUVBT/+9//qFq1Kq6urjg6OuLr60uDBg2YMGEC8fHxRvunNibq9u3bjBs3jsqVK+Pi4kKhQoWoVasW06ZNIyYmxmT/nTt3Gv6m4+Pj+fjjjw1/x15eXnTp0kV6vrKSEpkSFRWlABUeHp79Jw8PV8rWVilQKjBQqaSkbDv1w4cP1apVq9TDhw+z7ZzJxcTEqJMnT6qYmJhsPW9iYqKKjIxUiYmJ2XrejFi5cqUKCQlRZcuWVYBq0KCBCgkJUSEhIWrKlCkqNDRUASogIEB16dJF2dvbqxYtWqhevXqpMmXKKEC5ubmpffv2Wdzu6Oho1apVKwWoQYMGpau+n332maE+mTF58mQFqNKlS6vmzZurXr16qSZNmigHBwcFqC5duqikZH8riYmJqlSpUgpQf/zxh1FZ+navWrVKAapmzZpG269cuaKqVq2qAOXp6alatGihOnfurAICAhSgAgMDVVhYmNExCxYsUIB6/vnnVenSpVWRIkVUhw4dVPfu3dULL7xg2K9s2bLKwcFB1ahRQ3Xo0EF16dJFPfXUUwpQdnZ26pdffjFpe2JiomrXrp0ClIODg2rVqpXq2bOnKlOmjHJxcVGvv/66AlRISIjJscuWLVNOTk4KUEFBQapdu3YqODhYubq6KkD1798/I2+HQZMmTRSghg0bpmxsbNRTTz2levXqpRo3bqxsbGwUoEaOHGlyXEhIiALUwIEDlaOjowoMDFQ9evRQ7du3V9OnT1dKaZ+7pk2bGj6zTZo0Ud26dVMVK1ZUgPLy8lKHDx82KXvr1q2qcOHCClDFihVTHTp0UJ06dVJ16tRR9vb2Jq+T/n0NDQ01rLt//76qUqWKAlTRokVV+/btVa9evVTTpk1ViRIlFKAiIyONygGUuZ/fc+fOGc5RtGhR1bVrV9WhQwdVqFAhw+fv9u3bRsfs2LFDAap+/fqqRYsWysXFRbVu3Vp17dpV+fn5KUAVLlzYqM5Pykvfa2lJz29DeHi4AlRUVFSGzydBVCblaBD17bdaAAVKvf12tp5agqi882Wj/xFasGCB0Xp9EAUob29vdezYMcO2hIQE9cYbbxiCmgcPHpht96ZNm1RISIjq27evatWqleHLvnXr1urOnTsW1/H+/fuqdOnSKf6Qpsf+/fvV8ePHTdZfuXJFVatWTQFq2bJlRtveffdds4Gf/v3u1KmTAtSXX35p2JaUlKQaNGigAPXyyy+r6Ohow7b4+Hj11ltvKUA1a9bMqEx9EAWo5s2bp/gFvnLlSpMfTP16Ozs75eXlpR48eGC0bebMmQpQPj4+6tSpU4b1CQkJ6s033zSc98ng4O+//1aOjo7KyclJrVixwuhzHhYWZggUFy1aZLaultAHUYD66KOPjLbt3LlTOTs7K0Bt3LjRaJv+8wuoMWPGmP3b69OnjwJUu3bt1I0bN4y26YPz8uXLq4SEBMP6ixcvKg8PD0O5cXFxRu2+ceOG2rNnj1FZ5oKoRYsWKUC1adPG5PswMTFR7dy5U8XFxRmtTymIqlu3rgJUhw4d1L179wzrb968qWrWrKkA1adPH6Nj9EEUoGrUqKGuXbtm2BYTE6Oee+45BahXX33V5HzJ65nXvtdSIkFUHpOjQVSbNo+DqD//zNZT55UgasYMpUqWTPvRvr3pse3bm9s3Sfn6JqqSJZMM62bMMD4uOtqyc5YsqdTBg1Z8UVJgSRD1+eefmxwXGxurSpYsqQD1/fffm/2S1f9AJX/06dNHXb9+PUN19PX1zdK/pU2bNilAde/e3Wj9f//9pwDl4eFh9JlKTExUZ8+eVfb29srR0VFFREQYtm3YsEEBqnr16io+Pt7kXImJiYYeiuRBnT6Isre3V+fOnctQO3r37q0AtW7dOqP1+h7Er7/+2uSYuLg4w/v5ZBDVs2dPBRh6dp78Ud2/f78CVK1atTJUX6UeB1E1atQwu10fdLZs2dJovf6zUaFCBaMgSO/kyZNKp9MpX19fo0A2ubZt2ypArVmzxrBu+PDhClDtk/3xpxVMmAuipk2bpgD16aefptj2J5kLovbs2aMA5eLiYvbv5+DBgwpQNjY26tKlS4b1+iBKp9Opo0ePmhz3559/KkCVKVMmxfpIEJXxIEruzsurIiNh61btuZ8fPPNMztYnl4qOhitX0t7Pz8903a1b5o7VPXoYnyM5pSw7J8DDh5btl9VCQkJM1jk6OtKzZ08+/fRTdu3aRbt27Uz2GT58OMOHDyc+Pp6LFy+yevVqPvjgAzZu3MjKlStp3LhxmueePHkyixYtwsnJiWXLluHl5ZXp9sTFxbF582YOHDjAzZs3iYuLQynF3bt3ATh9+rTR/mXLlqVx48bs3r2blStX0rt3b8O25cuXEx8fT48ePYwG069btw6Arl27GsaBJWdjY0Pjxo05ceIE+/bto8oTSXBr1KhBmTJlUm3H1atXWbduHadOnSIqKoqEhAQA/vnnH0M72rZtC8Dly5c5f/48AH369DEpy8HBgW7dujFz5kyj9UlJSWzYsAGAnj17mq1H7dq1cXNz48iRI8TGxuLk5JRqvVPz0ksvmV0fEhLCjBkz2Lt3L4mJiSY3H3Tq1MlkHcD69etRStGmTRsKFSpktuymTZuyfv169u3bZ/gcb9y4EYBXX301w20BqFOnDgDTpk3Dy8uLdu3amb3pIi36vG2tW7emePHiJttr1apFtWrVOHbsGLt27eKFF14w2u7v7081M3dn68f6XbH0S0mkiwRRedWaNaAfrNi1q9yVlwJ3d7DkbuSiRc2vMz1WoZR6NChUZzhHcjqdZecEyA3pkAoXLkzhwoXNbitdujSg/UCnxt7enrJlyzJy5EgaNGjAs88+y4svvsjp06dxdnZO8bhPP/2UCRMm4OjoyMqVKw15pDLjzz//pGfPnoa7Es2JfjLyBQYMGMDu3btZsGCBURC1ZMkSQBuYn5w+YBk/fjzjx49PtU63bt0yWRcYGJjqMe+99x4ffvihyaDk5JK3Q/8eeXt74+bmZnZ/c+eMiIgwlONn7n8TZvbPzC3++s9USutjYmKIiIigWLFiRttTer3078O3337Lt99+m+q5k78PFy5cAMj0XXJNmzZl9OjRfPLJJ4SEhKDT6ShfvjwNGjSgY8eOtG/fHhubtO/h0gc5Kb0+oAX7x44dMxsQ+fv7mz3G/dEXVFxcnCXNEekkQVRelfyuvO7dc64eudzIkdojI377zXRdUpIiOjoad3d3bGzMB66FCkEaMUeeo9KRSbxu3bo89dRT/PPPPxw8eJBGjRqZ3e/LL7/krbfewsHBgRUrVtC6detM1/PBgwd06tSJGzdu0L9/f1577TXKlSuHu7s7tra2nDlzhooVK5ptT/fu3XnjjTfYtm0bly9fplSpUhw+fJh//vmHkiVL0qpVK6P99Xf5NWzYkLJly6Zar8qVK5usSy24/PXXX5k0aRJubm7MmjWL4OBgfH19cXZ2RqfTMW7cOKZMmWK2Hallwja3LfndivoeSaUU8fHx2NvbmxyT3rsuM8Jcu1J6vfT1r169utmemOTq1q2b+cqZMXXqVAYPHsyaNWvYu3cvv//+OwsWLGDBggXUqVOHHTt24OrqmiXn1rMkUBPWJ0FUXhQVBZs2ac99faFevZytj8jT7ty5w507d8z2Rulv5S5VqlS6ytT/YNy8edPs9tmzZzNs2DBDAPX888+nq/yU7N69mxs3blCzZk2+++47k+1nz55N8VgXFxd69OjBt99+y6JFi3j33XdZtGgRoF2CevJHSt9r07FjR95++22r1F9v2bJlAHz44YdmLzeZa4e+d+jWrVvcv3/f7I+2uelKvL29cXZ2JiYmhunTp+Pt7U1SUlKy/yxY98c5NDTU7Hp93ZycnNJ1SVf/PjRo0IBZs2ZZfJy/vz+nT5/m1KlTlCtXzuLjUhIYGMgbb7zBG2+8AcCBAwd48cUXOXDgANOmTeO9995L9Xj9+6fvWTNHvy2zyT6F9UjomhetXft4ME3XriD/AxGZZC5Hz8OHD1m6dCmgJcO0VHh4uCGXUoUKFUy2z507l9dff90QQJkba5VRt2/fBlK+tJFWvqMBAwYAsGjRIuLi4gwZ182NGWvTpg2gjZlKT0+dJfTtCAgIMNl28+ZNtmzZYrLez8/PcMkreaZ4vYcPH7JixQqT9ba2trRs2RJ4HLxlpZTeA32S1YYNG5odY5YS/fvw22+/ERsba/Fx+p7P+fPnW3xMetSpU4chQ4YA2tRLadHnCNu4cSM3btww2X7kyBGOHj1qGGsncgf59c2LJMGmsLLJkydz4sQJw3JSUhKjR4/m8uXL+Pn50bVrV8O2kydPsmTJErM/WGfOnKF79+7ExcVRr149qlatarR9/vz5DBkyJEsCKHg8iHbbtm2cPHnSaNu8efMMQWFK6tevT8WKFTl79iyjR48mIiKCevXqmc323rFjR+rUqcP+/fvp37+/2XFPkZGRzJ071zAgPL3tmDdvHg+T3X0QFRVFSEgIUVFRZo8bNmwYABMnTuTMmTOG9UlJSYwdOzbFpJMTJ07EwcGBUaNGsWjRIpOEpAAnTpzg119/TVc7zDl06BDTpk0zWrd3715mz54NwIgRI9JVXo0aNejatSuXLl2iS5cuZnvb7t+/z5IlS4yCk5EjR1KoUCF+++03/ve//5mMPbt58yZ79+5N8/wrV65k9+7dJq9ZfHy8YfC6uWD4SQ0bNqRu3brExMQwaNAgHjx4YNgWHh7OoEGDAOjVq5dFY9dENsnwfX1CKZUDKQ6io5VydNTSGhQvrpSZW36zQ15JcWBtefFW4LRSHPj7+6vOnTsre3t71bJlS9WrVy9Dgk5XV1e1Z88eo3brb6l2dXVVDRs2VL169VJdunRRtWvXNiRNrFSpkrpw4YLR+Y4cOaJ0Op0hmaM+8ae5R2Z07NjRKNlkr169VFBQkNLpdIZ8UKkl9Jw6dapRyoZZs2al+H5fuXJFVa9e3fB61K9f3/B6VK9eXdna2irA6HOqT3GQWjvPnz9vSAJZsmRJQ9JFDw8P5ePjowYMGKAANXHiRKPjEhISVJs2bRSgHB0dVevWrQ3vp7OzsxoyZIghceWTli1bplxcXBSgSpUqpYKDg1WfPn1UmzZtDMlIe/bsmeprn5onk21WrlxZ9e7dWzVp0sTwuXnzzTdNjkvp85tcdHS0at68ueF9r1OnjurRo4fq3r27qlOnjiHR6r///mt03KZNmwy5zYoXL646duyoOnXqpJ555hmLk23q8295e3urli1bqhdeeEF16NBBFStWzPD+JU9JoJRlyTaLFSumunXrpjp27Kjc3d3TTLbZpEmTFF+flM6nlxe/11IieaLymGwPon7++XFuqCFDsuecZkgQlXe+bNIKogICAlR8fLz68MMPVVBQkHJ0dFSenp6qa9eu6p9//lFKGbf75s2b6sMPP1StW7dWgYGBytXVVTk4OKgSJUqoli1bqq+++krFxsaa1CN5UsC0Hpnx8OFD9cknn6iqVasqFxcX5enpqVq1aqU2b95s1OaUXL161RD8uLq6qkuXLqX6fsfGxqq5c+eqZs2aKS8vL2VnZ6eKFSumqlevroYOHao2bdpktL8lQZRS2vvzwgsvKH9/f+Xo6KgCAgLU4MGD1fXr19XEiRPNBlH69k+bNk099dRTytHRUXl7e6vOnTur48ePq/fff18BauzYsSmec8SIEapKlSrK1dVVOTk5qYCAANW0aVM1depU9d9//6Va59Tog6gdO3aobdu2qebNmysPDw/l7OysateurRYuXGj2OEuCKKW0z+iPP/6o2rZtq4oXL67s7e2Vl5eXqlKliurfv79auXKl2e+rCxcuqDfffFNVrFhROTk5KTc3N1WhQgU1YMAAkwz25oKoI0eOqDFjxqiGDRuqkiVLKgcHB1W0aFFVq1Yt9dFHH5n9bUjtcx4REaHGjh2rKlWqpJycnJSLi4uqUaOGmjp1qklyVaUkiHpSdgdROqWsfDG/gImOjsbDw4Pw8HCr5LdJU7duoB/XsH07NGuW9ec0Iz4+nvXr19O2bdtMT8aZEbGxsYSGhlK6dOlM5axJr6wccJvdwsLCKF26NAEBAWYvgSSXn9qdHvmt3cHBwezYsYMVK1bQpUuXFPfLinY3bdqUXbt2sWPHDqM5AnOT/PZ+Wyo/tTs9vw0RERF4e3sTFRVlSAWRXnn71Spo7t+H9eu150WLQgq3jgshCq6jR48ajaMCbVD5pEmT2LFjB8WKFTMk6BRCZI6kOMhLNmwA/SzenTtDOu5gEUIUDMOHD+fo0aNUq1YNHx8fIiMjOX78ONeuXcPJycmQHV4IkXnyK5yXyF15ooDp16+fxft26tSJTp06ZVld8oqBAweyZMkS/v77b/bv349SCl9fXwYMGMBbb73FU089lelzyPsihEaCqLwiJkbLDwXg5QW5dEyByDsCAwOtnt/I2vTJLi0RGBgoP9bACy+8YDKvmrVl5H3Rzw0nRH4iQVResXGjNiYKoFMnyIHB3EJkt9we5BVU8r4IoZGB5XmFXMoTQgghcpVcH0QtX76cpk2bUqRIEVxdXalWrRrTpk1LdWZzc44cOcKUKVNo3rw5xYsXx97eniJFitCoUSNmz56d7vKyVWwsrFmjPS9cGIKDc7Q6QgghhMjll/OGDx/OzJkzsbOzIzg4GDc3N7Zv387o0aNZs2YNmzdvTnUmdL2EhARq1qwJgJubG3Xq1KF48eJcvnyZP/74g71797J48WI2bdpkdhLWHLdlC9y9qz3v2BEcHHK2PkIIIYTIvT1Rq1atYubMmbi5ufHXX3+xadMmVqxYwdmzZ6latSp79+5l/PjxFpdXq1Ytli1bRnh4ONu3b+enn35iz549HDlyBB8fH/bv38/IkSOzsEWZkPxSXvfuOVcPIYQQQhjk2iDqo48+AmDMmDGGXiQAb29v5syZA8CsWbNSnIgzOTs7Ow4ePEj37t1xdHQ02la1alXDZJg///xz7rusFxcHq1drz93doUWLnK2PEEIIIYBcGkRduXKFAwcOANCnTx+T7Q0bNsTPz4+4uDjW6zN4Z0KNGjUAiImJITw8PNPlWdW2baAPFDt0gCeCQCGEEELkjFwZRB05cgQAT09PSpcubXaf2rVrG+2bGWfPngXAwcEBT0/PTJdnVXJXnhBCCJEr5cogKjQ0FAB/f/8U9/Hz8zPaN6OUUobLee3atTO53Jej4uNh1SrtuZsbtGqVo9URQgghxGO58u68u4/uRHN1dU1xHzc3NwCio6Mzda733nuPP/74Azc3N6ZOnZrm/nFxccTFxRmW9eePj4+3+ngq3ZYt2EVGApD0/PMk2tlpgVUuoG9rTo0hi4+PRylFUlISSUlJ2XZefZJB/bkLCmm3tLsgkHbn/XYnJSWhlCI+Ph5bW9tU97XG71euDKKyy+LFi3n//fexsbHhu+++o3z58mkeM2XKFN577z2T9Tt27MDFxcWq9as2ezaBj54fDAzkmhXGf1nbli1bcuS8dnZ2lChRgnv37pnMWJ8d9IF+QSPtLlik3QVLfmj3w4cPiYmJYffu3SQkJKS674MHDzJ9vlwZRBUqVAiA+/ppTsy4d+8eAO7u7hk6x/LlyxkwYAAA8+fPp7uFqQPGjh1rlAohOjoaPz8/mjVrhpeXV4bqYlZCAnYvvwyAcnGhxtix1LBykJYZ8fHxbNmyhZYtW2KfA1PQxMbGcunSJdzc3LJ1RnqlFHfv3qVQoULodLpsO29Ok3ZnT7vXrFnD9OnTOXbsmOEHbdu2bTRt2pTDhw+zY8cODh06xOHDh/nvv/9QSrFo0SJefPFFq9ZD3m9pd14VGxuLs7MzjRs3TvO3ISIiItPny5VBVGBgIACXLl1KcR/9Nv2+6fHrr7/Sp08fkpKS+Prrrw3BlCUcHR3Njpuyt7e3bjCxezc8eoN1zz+PvYeH9cq2Iqu320KJiYnodDpsbGywscm+oX36rm79uQuK/NzunTt30qxZM5o0aWIySW52tvvo0aN0796dpKQkgoOD8fHxQafT4evri42NDR988AGr9elOksmKv4H8/H6nRtqd99ttY2ODTqez6LfJGr9duTKI0qcciIiIIDQ01OwdegcPHgQwyiFliVWrVtGrVy8SExP56quvGDhwYOYrnBXkrjwhCpRVq1YRHx/PuHHj+PDDD02216tXj8qVK1OzZk1q1KjBgAED2LVrVw7UVAihlyuDqFKlSlGnTh0OHDjAjz/+yLvvvmu0fe/evVy6dAlHR0fatm1rcblr1qyhR48eJCQk8NVXXzFo0CBrV906EhPh11+1587OkI42CiHyposXLwKkODZzzJgx2VkdIYQFcm2/3bhx4wCYOnUqhw8fNqyPiIhgyJAhALz++ut4JLvMtXLlSoKCgmjevLlJeevXr6dbt24kJCQwd+7c3BtAAezZAzdvas/btNHSGwiRQQ8ePODzzz+nYcOGFClSBEdHRwICAmjfvj0//vijyb5Tp06lZs2aFCpUCBcXFypXrsz48eO5c+eOSdlhYWHodDoCAwNRSjFv3jxq1aqFq6srHh4etGrVij/++MMqdWvatCk6nY6dO3eyZ88e2rdvT9GiRbGxsWHhwoWG/WJiYpgxYwb16tWjcOHCODk5UbFiRd555x2TMRBNmzalWbNmAOzatQudTmd4mBsqsG3bNrp06YKPjw8ODg4UK1aMzp07p9rGtEyaNAmdTseCBQsA6N+/v6EOTZs2zXC5Qoislyt7ogA6derEsGHD+OKLL6hXrx7NmzfH1dWVbdu2cefOHRo0aMDkyZONjomKiuL06dPExsYarb958yZdunTh4cOHlCpVin379rFv3z6z550+fTre3t5Z1i6LyKU8YSWXLl2idevWnDx5EhcXFxo0aICXlxdXrlxhz549HD9+3DArwO3bt2nevDlHjx7F3d2d4OBg7O3t2bVrFx999BFLlixh+/btlClTxuy5+vfvz48//kijRo1o164dR48eZcuWLezevZtdu3ZRt27dDNctueXLlzN37lyCgoJo0aIFt2/fNoxTvHr1Kq1bt+b48eN4enpSp04dChUqxOHDh/nkk09Yvnw5O3fuJCAgAIDWrVvj5OTEpk2bKF68OK1btzac58nvgbfffpsZM2ZgY2ND7dq1adSoERcvXmT16tWsWbOG+fPn079//3S/R9WrVyckJIS9e/dy7tw5GjRoQLly5QAICgpKd3lCiGykcrmlS5eqxo0bK3d3d+Xs7KyqVKmipk6dquLi4kz2XbBggQJUQECA0frQ0FAFWPQIDQ1NV/2ioqIUoMLDwzPRymQSE5UqUUIpUMrRUamoKOuUa2UPHz5Uq1atUg8fPsyR88fExKiTJ0+qmJiYbD1vYmKiioyMVImJidl63oxITExUtWvXVoBq1aqVunnzptH2mJgYtW7dOsNyz549FaDq1q1r9Hm+e/euat26tQJU/fr1jcpI/rcVEBCgTp8+bdiWkJCgBgwYYDh/ZuqmlFJNmjQxnGv27Nkm7U1KSlINGjRQgHr55ZdVdHS0YVt8fLx66623FKCaNWtmdNyOHTsUoJo0aWL2NYyMjFRz585VgCpXrpw6duyY0T67du1ShQoVUg4ODurMmTMmZVgqJCREAWrBggUW7a9/Pb7//vsMnzMleelzbk3S7rzf7vT8NoSHhytARWXidzbXB1G5ndWDqD17tAAKlOrQwTplZoE8EUTVqqVUyZJWfSSVLKkSfX1VkpXLNXrUqmWV12jVqlUKUD4+Puru3bup7nvhwgVlY2OjdDqdSZCglFIXL15UTk5OClC///67YX3yIOq3334zOe7atWsKUI6OjkaflfTUTU8fNAQHB5vdvmHDBgWo6tWrq/j4eJPtiYmJqkqVKgpQx48fN6xPK4iKiIhQvr6+ClAHDx40e+5p06YpQL311lsWtcUcCaJynrQ777c7u4OoXHs5r8BKfinPwtxVIgXXr8OVK1YtUvfokRds3LgR0CbxdktjXN3u3btJSkqiZs2aPP300ybbS5YsSXBwMOvXr2fHjh3Ur1/faLudnZ3RpTC9EiVKUKRIESIjI4mIiKBEiRLprtuTuqVwiXvdunUAdO3aFTs70682GxsbGjduzIkTJ9i3bx9VqlSx6Hx///03V69epWzZstSqVcvsPvqxSykNExBC5E8SROUmSUmPgyh7e2jfPmfrk9c9+sG2JoWWmE6n02VdMGWlel+4cAGwbFzNlUfBZkoTfiffdsVMYOrj45NizhV3d3ciIyONxiqmp25PSik33Pnz5wEYP34848ePT7WMW7duWXy+sLAwAM6dO5dmIsL0lCuEyPskiMpN/vrrcc9Jq1aQSxNs5hmPcolZk0pKIjo6Gnd3d3R5PCmdNWVngj5nZ2ez6/UJAxs2bEjZsmVTLaNy5coWn09fbokSJXjuuedS3TfHb0oRQmQrCaJyE7krT1iRv78/AKdOnUpz35IlSwKPe3PM0ffI6PfNrrpZys/PD4COHTvy9ttvW61cfXu9vLyMUikIIYT8Vzq3UOpxEGVnBx065Gx9RJ6nH6P0008/pToPJUDjxo2xsbHh6NGjHDt2zGT7tWvX2LZtG4Ahr1J21c1Sbdq0AbQUCOrRrPSWcHBwAEhxstKaNWvi7e3NyZMn+eeffzJfUSFEviFBVG5x8CA8ylhMixbg6Zmz9RF5XocOHahRowZXr16le/fuJokmY2Nj2bBhA6D1DHXv3h2lFIMGDTLa9/79+wwaNIjY2Fjq169vMqg8q+tmqY4dO1KnTh32799P//79zY5PioyMZO7cuUYBU6lSpQA4e/Ys8fHxJsfY29szYcIElFJ07tyZvXv3muyTmJjI9u3b+fPPP9NVZyFE3iaX83KL5csfP5dLecIKbGxsWLlyJc899xwbNmzA39+fhg0bGhJaHjt2jMKFCxsu082ePZtTp07x119/UbZsWZo1a4adnR27du3i1q1bBAQE8P333+dI3Swtc9WqVTz//PMsWrSIX375hWrVquHv78/Dhw85f/48x48fJzExkX79+hnu4PP396d27docPHiQqlWrUrt2bZycnPD29uajjz4CYOjQoVy6dIlPPvmERo0aUblyZcqVK4ezszPXr1/n6NGj3Llzh6+++op69epZ5TV60rp164wSDJ88eRLQMp7PmjXLsF4COSGyjwRRuUHyS3m2ttCxY87WR+QbAQEBHDx4kDlz5vDLL7/wxx9/8PDhQ0qUKEGTJk2MMoJ7eXmxb98+vvjiC5YuXcrmzZtJSkqidOnSvPLKK7z66quGsUzZXTdL+fr68ueff7Jw4UKWLl3K33//zf79+/H09MTX15fBgwfToUMHnJycjI5bsWIFY8eOZceOHSxdupSEhAQCAgIMQRTAtGnT6NSpE3PmzGHv3r1s3LgRBwcHfHx8aNq0Ke3ataNLly6Zfl1ScuvWLf766y+T9efOnePcuXNZdl4hRMp0Kj2DB4SJ6OhoPDw8CA8Px8vLK2OFHD4M+vwzLVrA/9u797CqqryB498NHsBQQCBFBQHvVqIiigmGqSRqj3dzaiywaWweu7yWY2Nl5byWtyZnvExmvSna6GReUFFnREwx8IaBzBRqKiLqeAOHixfktt4/6JxEUOF4ztlcfp/nOc+zz157r/1beOT82GvttXbutFyAVlJcXMz27dsZOnToXR9tt6bCwkJOnz6Nv79/pS9Eayq77ek8Wz6Rpjdpt7S7IZB21/121+S7IScnB09PT/Ly8nBxcTHrenX7p1VfyFN5QgghRJ0jSZTelPplPJSdHYwapW88QgghhKgWGROlt3/9C06eLN8OC4PmzfWNRwjxQI4dO8bcuXOrffz06dPNmrldCKE/SaL0Jl15QtQrFy9eZOXKldU+PioqSpIoIeooSaL0dHtXnqZJV54Q9UD//v1rNNmnEKLukjFRekpPh+PHy7dDQ6FlS33jEUIIIUS1SRKlp9u78saN0y8OIYQQQtSYJFF6un2WcitO0ieEEEIIy5MkSi9Hj4JxMdO+feHnleKFEEIIUTdIEqWXDRt+2Zan8oQQQog6R5Iovdw+HmrMGP3iEEIIIYRZJInSw4kTkJZWvh0cDBZc1FUIIYQQtiFJlB5kgk0hhBCizpMkSg/SlSeEEELUeZJE2VpGBqSklG/37An+/vrGI4QQQgizSBJla/JUntBBbGws/fr1w8XFBU3T0DSNPXv2mF3f6tWreeGFF+jWrRvNmzfHYDDg6upK7969mTNnDteuXat0TllZGfv27eP9998nNDQUDw8PDAYDnp6ehIeHs3r1aostl5Kens7IkSNp3rw59vb2aJrGzJkzATh79izLli1j0qRJ9OzZE0dHRzRN46WXXrLItYUQDYesnWdrMh5K2NiRI0cYM2YMZWVlDBgwgJYtW6JpGl5eXmbXuXTpUvbt20eXLl0IDAzE3d2dS5cusX//fpKTk1m+fDkJCQm0atXKdE5GRgYhISEAuLu7ExQURLNmzcjIyCA+Pp74+Hi+/vprNmzYgIODg9mxXb9+nWHDhpGZmUlQUBCDBw/G3t6e7t27A7BhwwbeeOMNs+sXQggjSaJs6cwZOHSofLt7d2jfXtdwRMOwadMmiouLeeedd/joo48sUucnn3xChw4dcHd3r7A/JyeHkSNHkpiYyNSpU/n73/9uKtM0jQEDBjBt2jTCw8Oxt7c3lSUkJDBs2DC2bt3K3Llzef/9982OLTk5mczMTPr27UtSUlKlcn9/f1577TUCAwMJDAzkm2++sdjPRQjRsEh3ni1JV57QQVZWFgAdOnSwWJ3BwcGVEigADw8PZs+eDUBcXFyFsnbt2rFr1y4iIiIqJFAAYWFhTJ8+HYBVq1Y9UGz3a++IESNYtGgRUVFRBAQE0KiR/C0phDCPJFG2JF15woZmzpyJpmmsWLECgIkTJ5rGQ/Xv35/MzEw0TcPPz4+SkhLmz5/Po48+SuPGjfH09OSZZ57h2LFjNb6uMSlxdHSs0Xk9evQAyscsmWPPnj1omkZkZCQAK1euNLVX0zSz6hRCiHuRP8Fs5dw52L+/fPuxx6BTJ33jEfVe9+7diYyMJDExkVOnThESEkL7n7uQO3fuXOHY8ePHExsbS1hYGAEBARw6dIh169bxj3/8g7i4OIKDg6t1zYKCAtMA7uHDh9co3hMnTgDQsmXLGp1n5OXlRWRkJCdPniQpKYl27doRGhpqVl1CCFEdkkTZysaNv2yPG6dfHKLBGDlyJCNHjiQqKopTp07x0ksvERUVZSrPzMwE4MyZM1y/fp3Dhw8TEBAAQGlpKW+88QaLFy/m2Wef5ejRo1VeIy4ujjVr1lBWVmYaWF5QUEBERATz5s2rdqw3btxg0aJFAIwxc+60zp07Ex0dTXR0NElJSYSGhhIdHW1WXUIIUR2SRNmKdOXpY8GC8tf9BAbCli0V9w0f/sucXj/TABelKnYPvflm+cuooAC6dKlefJs3l88XprMZM2aYEigAe3t7Pv74YzZu3MiZM2fYsGEDTz/9dKXz0tPTWblyZYV9zz33HAsWLMDV1bXa1588eTKnT5+mVatWvPPOO+Y3RAghbEiSKFu4cAESE8u3u3SBRx7RN56GJD8fzp+//3E+PpX3XblS6Vzt51ela9xOqepdE6CoqHrHWZlxHNHtHB0dGT9+PAsWLCAhIaHKJGrKlClMmTKF4uJisrKy2Lx5Mx9++CH//Oc/iYmJ4YknnrjvtWfNmsXKlStxcnLim2++wcPDwyJtEkIIa5MkyhY2biz/YgW5C2VrLi7QuvX9j3v44ar33XGuAtTPd6JMyZSLS8XzNK161wR4gPmQLMXNzQ03N7cqy/x/nlH/3Llz96zDYDDQrl073nzzTUJCQnj88ceZMGECx48fp3Hjxnc9b8GCBbz//vs4OjoSExNjmkdKCCHqAkmibEG68vRzZ1dbTdzZvQeosjLy8/PLZ/62u8vDrU2blj9IUI/UZCbx4OBgHnnkEX788UcOHz5Mv379qjxu8eLFTJ06FQcHBzZs2EBERISlwhVCCJuQKQ6s7dIl2Lu3fLtjR+jaVd94hLhDbm4uubm5VZYZB597e3vXqE5nZ2cALl++XGX5X//6V15//XVTAjVs2LAa1S+EELWBJFHWtmkTlJWVb48dW97VI0Qt89VXX1XaV1RUxNq1a4HyyTCrKzs7m7S0NAA6duxYqfyzzz7j1VdfNSVQVY21EkKIukCSKGtbt+6XbenKE7XUrFmz+OGHH0zvy8rK+MMf/sC5c+fw8fGpMO1Aeno6q1evprCwsFI9P/30E+PGjePWrVv06dOHrnfcef3iiy+YPHmyJFBCiHpBxkRZ05UrsGdP+XbbtuXr5QlRy7Rp04aePXsSGBhI//798fDwIDk5mVOnTuHs7MyaNWtwcnKi6OcnCS9fvsyECRN4+eWX6dGjB97e3hQVFZGVlUVKSgplZWV06dLFdBfL6MiRI7z88ssopWjbti3r169n/e3jBW9jzfmdLly4wKhRo0zvjYPmt2zZQp8+fUz7P/30U9OixUIIURVJoqxp82YoLS3flq48UUtpmsY333zD/Pnz+eqrr9i7dy/Ozs6MGTOG//3f/+WRRx6hzNglDTz66KN89NFHfPfddxw7dozU1FSKi4txd3dn4MCBjB49mokTJ1Za9iU3N9c0QP3YsWP3XFLGmknUrVu3OHjwYKX9V65c4cqVK6b3+XdOXSGEEHfQVE0euxGV5Ofn4+rqSnZ2duX5bSIiYMeO8u3kZAgKsn2AVlJcXMz27dsZOnQoBoPB5tcvLCzk9OnT+Pv74+TkZLPrlt32dJ7d3Z7OqyMyMzPx9/fH19fXNID8bupTu2tC2i3tbgjqU7tr8t2Qk5ODp6cneXl5uNw5VU011e2fVm129Srs2lW+7etbK2alFkIIIYTlSBJlLZs3Q0lJ+bZ05QkhhBD1joyJshaZYFOIB3b7gsn3Y1xwWQghbEWSKGvIzYWdO8u3vb2hd29dwxGiKn5+fjWaiVwPdy5ufC9+fn6SRAkhbEqSKGuIjYXi4vLtsWOhjg/UE0IvtT3JE0I0bPLtbg3SlSeEEELUe5JEWVp+/i/TGrRsCY8/rm88QgghhLAKSaIsbetWuHWrfHvMGOnKE0IIIeop+Ya3NOnKsykZMyOEEMLI1t8JkkRZ0rVr8I9/lG+3aAGhofrGU48ZZ9UtNS6rI4QQosEzfifYauZ1SaIsaft2MK5sP3o02NvrG089ZjAYsLe35+bNm3qHIoQQopYoKCjAYDDYbDkySaIsad26X7alK8+qNE3joYceIi8vT+5GCSGE4ObNm+Tn59O0aVM0G60SIvNEWcr16+V3ogA8PeGJJ/SNpwFo3rw5mZmZnDlzBnd3dxwdHa3+H6esrIyioiIKCwvr/EKdNSHtlnY3BNLuutdupRSlpaUUFBSQn5+Po6Mjnp6eNru+JFEWou3aBTdulL8ZNQoayY/W2hwcHPD29iY7O5sLFy7Y5JpKKW7evEnjxo1t9pdObSDtlnY3BNLuuttug8GAm5sbnp6e2NtwKI1801uIXWzsL2/GjdMvkAbmoYceok2bNpSUlFBiXPDZioqLi9m7dy9PPPGEzfrcawNpt7S7IZB2181229nZYTAYdEkAJYmyEC0urnzD3R3699c1loaoUaNGNLLB3T97e3tKSkpwcnKqk79szCXtlnY3BNLuhtVuS6j1nZ/r1q2jf//+NGvWDGdnZ7p168b8+fMpNq5NV0Pff/8948aNo0WLFjg5OeHv789rr73G5cuXHyhO7fr18o2RI0E+hEIIIUS9V6uTqClTpvDMM8+QlJRE7969iYiIICsriz/84Q8MGDCgxo+3r1+/nj59+rB+/Xp8fX0ZMWIEdnZ2LFmyhICAAE6ePPngQctTeUIIIUSDUGuTqE2bNrFw4UKaNGnCwYMH2bFjBxs2bODEiRN07dqVxMRE3nvvvWrX95///IfIyEhKSkpYtmwZhw4dYu3atfz0009MmDCBS5cu8dxzzz3YbKdubjBwoPnnCyGEEKLOqLVJ1OzZswGYPn06gYGBpv2enp58+umnACxZsoS8vLxq1feXv/yFGzduMGjQICZNmmTab29vz9KlS3F1dSU5OZk449gmc4wYAQ4O5p8vhBBCiDqjViZR58+fJzk5GYDnnnuuUnloaCg+Pj7cunWL7ca5me4jJibmrvU1adKE4cOHA7Bx40Zzw5auPCGEEKIBqZVJVGpqKgDu7u74+/tXeUxQUFCFY++loKDANN7JeN6D1FcV1aQJhIebda4QQggh6p5amUSdPn0agDZt2tz1GB8fnwrH3ktmZqZp+2511qS+qqiICHB0NOtcIYQQQtQ9tXKeqIKCAgCcnZ3vekyTJk0AyM/Pr3Z996qzuvXdunWLW7dumd4bx2RdffJJSnNy7htLfVFcXMyNGzfIyclpUPOKSLul3Q2BtFva3RBcvXoV4IEeKKuVSVRtNmfOHP74xz9W2t/ulVfglVd0iEgIIYQQ5srJycHV1dWsc2tlEtW0aVMArhsnsKzCtWvXAHBxcal2fcY6q/phVbe+t99+mzfffNP0Pjc3F19fX7Kyssz+R6iL8vPz8fHx4ezZs9X6N6gvpN3S7oZA2i3tbgjy8vJo06YN7u7uZtdRK5MoPz8/AM6ePXvXY4xlxmPvxdfX17SdlZVF165dza7P0dERxyrGPrm6ujaoD5+Ri4uLtLsBkXY3LNLuhqWhttvOzvzh4bVyYHmPHj2A8ltsdxvoffjwYYAKc0jdjYuLC+3bt69w3oPUJ4QQQghRK5Mob29vevXqBcCaNWsqlScmJnL27FkcHR0ZOnRoteocNWrUXeu7du0asbGxAIwePdrcsIUQQgjRgNTKJArgnXfeAWDu3LmkpKSY9ufk5DB58mQAXn311QrjkGJiYujcuTMDq1h6ZcqUKTz00EPEx8fzxRdfmPaXlpYyefJkcnNz6dWrF0899VSN4nR0dOSDDz6osouvPpN2S7sbAmm3tLshkHab325NPdBicdb1P//zPyxatAiDwcDAgQNxdnZm165d5ObmEhISws6dO2ncuLHp+OjoaCZOnIivr2+FuaGM1q1bx7PPPktpaSnBwcH4+fmRnJxMRkYGLVq0IDEx0dTtJ4QQQghxL7X2ThTAwoULWbt2LY8//jj79u1j+/bteHt7M3fuXL799tsKCVR1jBs3joMHDzJ69GgyMjKIiYmhtLSUV155hbS0NEmghBBCCFFttfpOlBBCCCFEbVWr70QJIYQQQtRWkkTV0PHjx1m8eDFRUVF07dqVRo0aoWkaH374od6hWU1xcTG7du1i2rRp9OrVCzc3NwwGA15eXgwfPpxt27bpHaLVrF69mhdeeIFu3brRvHlzDAYDrq6u9O7dmzlz5pgmaW0I3nrrLTRNq9ef96ioKFMb7/YqLCzUO0yrKSoqYtGiRYSGhuLu7o6TkxPe3t4MGTKEtWvX6h2exWVmZt7339v42rt3r97hWlRWVhavvvoqnTp1onHjxjg5OeHv709kZCRpaWl6h2c1Z8+e5dVXX6Vdu3Y4Ojri6enJ4MGDzf4eq5WTbdZmS5cuZeHChXqHYVMJCQmEh4cD4OXlRWhoKM7OzqSnpxMbG0tsbCyTJk3is88+Q9M0naO1rKVLl7Jv3z66dOlCYGAg7u7uXLp0if3795OcnMzy5ctJSEigVatWeodqVfv27eOTTz5B07QHWmeqrggJCbnrGEl7e3sbR2Mb586dY/DgwaSnp+Pp6UlISAjOzs6cPXuWvXv34uzszPjx4/UO06KaNGlCZGTkXcvT09NJTk6madOm9OzZ04aRWdfBgwcJDw+noKCA1q1b89RTT2Fvb8+RI0dYtWoVa9asYc2aNYwbN07vUC0qOTmZiIgIrl69SsuWLRkyZAg5OTns3r2buLg43n///SqXdbsnJWrkiy++UL///e/V6tWr1dGjR9Xzzz+vADVr1iy9Q7OaXbt2qTFjxqi9e/dWKvv666+Vvb29AtTKlSt1iM66Dhw4oHJycirtz87OVqGhoQpQv/rVr3SIzHauX7+uOnTooFq3bq1GjhxZrz/vkZGRClArVqzQOxSbunHjhurcubMC1MyZM1VRUVGF8uvXr6vU1FR9gtPRkCFDFKB++9vf6h2KRQUEBChATZo0qcK/dWlpqZoxY4YClJubm7p586aOUVrWzZs3lY+PjwLU+PHj1Y0bN0xlhw4dUh4eHgpQcXFxNapXkqgHZPylW1+/VKrjN7/5jQLUwIED9Q7Fpvbu3asA5e7urncoVvX6668rQG3btq3ef94bahL13nvvmb5URblz584pOzs7BagDBw7oHY7FZGdnK0AB6vLly5XKS0pKVOPGjRWgUlJSdIjQOtasWWNKDv/73/9WKl+4cKECVGhoaI3qlTFR4oEZl+m511qH9VGjRuW94fV5gro9e/awePFiXnjhhWqvDiDqluLiYpYuXQrAtGnTdI6m9oiOjqasrIxHH32U4OBgvcOxmJr8vvL09LRiJLaVnJwMQM+ePXFzc6tUPmjQIACSkpK4ePFiteuVMVHigZ04cQKAli1b6hyJ7RQUFDBz5kwAhg8frm8wVnLt2jVefPFFWrRowV/+8he9w7Gp3bt38+9//5uCggI8PDzo3bs3Q4cOrZcJc0pKCtnZ2bRq1Yr27dvz73//m40bN/Kf//yHZs2a0a9fP4YMGfJAi7TWRdHR0QD85je/0TcQC2vSpAn9+vXju+++Y8aMGSxZsgSDwQBAWVkZM2fO5ObNmwwZMgQfHx+do7Uc40NAHh4eVZYbE0alFCkpKdX+o1GSKPFALl68aPplM2bMGH2DsaK4uDjWrFlDWVmZaWB5QUEBERERzJs3T+/wrOL3v/89p0+fJiYmhmbNmukdjk2tWrWq0r6WLVuyfPlyIiIidIjIev71r38B5WuWTp8+nfnz51d4eGDevHn06NGDTZs20aZNG73CtKmEhAROnjyJg4MDzz//vN7hWNwXX3zB0KFD+fzzz9m2bRtBQUHY29uTmprK+fPnef7551myZIneYVpU8+bNAcjIyKiy/Pb9p0+frna9DetPC2FRJSUlTJgwgby8PLp27crLL7+sd0hWk56ezsqVK/nqq6+Ii4ujoKCA5557jujo6ArrN9YXcXFxLFu2jF/96leMHDlS73Bsplu3bixcuJAffviB/Px8Ll26RFxcHH379uXChQsMHz6cPXv26B2mReXk5ACQmprKvHnzmDx5MsePHycvL4+dO3fSsWNHUlNTGTZsGMXFxTpHaxvLly8Hyu8y16cuLaNOnTqxf/9+nnrqKc6fP8/mzZvZuHEjp0+fpn379vTv3x8XFxe9w7SoAQMGAPD999+Tmppaqfyzzz4zbefn51e/YouM2GrA6vtA23sxDij38PBQx48f1zscmygqKlInT55Un3zyiWrWrJlyd3dXCQkJeodlUbm5ucrb21s9/PDD6sqVKxXKGurnvaysTI0YMUIBqlu3bnqHY1GzZ882DTR+9tlnK5WfOXNGOTk5KUCtWrVKhwhtKy8vTz300EMKUNu3b9c7HKtITExUzZs3V61atVJr1qxRFy9eVFevXlWxsbGqQ4cOClAvvvii3mFa3BNPPKEA5e3trbZs2aJyc3PVqVOn1NSpU5WmacpgMChAzZ07t9p1ShL1gBrql4rxia1mzZrVqyc4auLAgQNK0zTl4+NT4XHZui4qKkoBau3atZXKGurnXSmljhw5Yko2srKy9A7HYhYvXmxq1549e6o8ZsyYMQpQL7zwgo2js71ly5aZvmhLS0v1Dsfi/vvf/6qHH35YaZpW5VOHp06dMiWR3377rQ4RWs+lS5dUSEiI6fN++2vKlCkqKChIAerzzz+vdp0yJkrU2NSpU1m0aBFubm7ExcWZns5raIKDg3nkkUf48ccfOXz4MP369dM7JIuIiYmhUaNGfPrpp3z66acVyo4dOwbAl19+SXx8PF5eXnz99dd6hGlzXbp0MW2fO3eu3gy6bdu2bZXbVR1z4cIFm8SkJ2NXXlRUVL0cTL9t2zauXLlCu3btqnzqsG3btgQHB7N7927i4+N58skndYjSOpo3b853331HfHw83377LTk5ObRo0YIRI0YQFBRkmjS5a9eu1a5TkihRI2+99RYLFizA1dWVuLg4goKC9A5JV87OzgBcvnxZ50gsq6SkhISEhLuWZ2ZmkpmZia+vrw2j0pdx7BBA06ZNdYzEsgIDA00z0WdnZ1eZHGZnZwPlT3bVZ+np6Rw8eBBN05g4caLe4VhFVlYWwD3HPBnHeV69etUmMdmSpmmEh4ebVuEwOnXqFBcuXMDDw4PAwMBq11f/0mxhNdOnT+fjjz/G1dWVnTt30qtXL71D0lV2drZpjamOHTvqHI3l5Obmosq7+iu9jEtkzJo1C6UUmZmZ+gZrQ8Y7bi4uLnTq1EnnaCzHuJQTQHx8fKXy4uJiU0Ldu3dvm8Zma19++SUATz755F3vytV1rVu3BsrvKufl5VUqLy4uJiUlBQB/f3+bxqanP/3pTwBMmjQJBweHap8nSZSolhkzZjBv3jzc3NwaTAKVnp7O6tWrq1xw9qeffmLcuHHcunWLPn361Oj2r6idjhw5wpYtWygpKamwv6ysjC+//JJ33nkHgNdff900r0598cEHHwAwZ84cDhw4YNpfUlLC1KlTycjIoGnTpvX27gyUJw9/+9vfgPo3N9TthgwZgrOzMzdv3uS3v/1thUXUi4qKeOONN8jKysJgMDB27FgdI7W89PT0Sk/elZSUMHv2bJYtW0b79u159913a1SndOfVUEpKCpMnTza9P3XqFADLli1j69atpv0xMTH1ZvLJLVu28NFHHwHQvn17/vrXv1Z5nKenpymbrw8uX77MhAkTePnll+nRowfe3t4UFRWRlZVFSkoKZWVldOnSpV6ubt8QZWZmMmrUKJo1a0ZgYCAtWrQgNzeXH374wdQF8uyzz5oSjvpk4MCBzJo1i/fee49+/frRu3dvvLy8SElJITMzk8aNG/P3v/+dFi1a6B2q1WzdupXLly/j5ubG6NGj9Q7Hah5++GE+++wzJk6cyLp169izZw+9evXCYDBw+PBhzp8/j52dHYsWLap3d+M+//xzli1bRs+ePWndujW3bt3iwIEDXLp0ifbt27Nz507TEI1qs8IA+Hpt9+7dVY7sv/N1+vRpvUO1mBUrVlSrzb6+vnqHalGXL19WH330kYqIiFB+fn7K2dlZOTg4KC8vLxUeHq6WLl2qCgsL9Q7Tpurz03kZGRlqypQpKjQ0VLVu3Vo5OTkpR0dH1aZNGzV27Fi1bds2vUO0uh07dqghQ4Yod3d3ZTAYlI+Pj4qKilJHjx7VOzSre/rppxWgJk+erHcoNnHkyBEVFRWl2rZtqxwdHZWDg4Py9fVVv/71r9XBgwf1Ds8qduzYoUaMGKF8fHyUo6OjcnFxUb169VLz5883+wlrTanbpqYVQgghhBDVImOihBBCCCHMIEmUEEIIIYQZJIkSQgghhDCDJFFCCCGEEGaQJEoIIYQQwgySRAkhhBBCmEGSKCGEEEIIM0gSJYQQQghhBkmihBA25efnh6ZplV5NmjShW7duvP322+Tk5FQ6LyoqCk3TiI6Otn3QP4uOjkbTNKKiomp0XmZmJpqm4efnZ7NrCiGsT5IoIYQuQkJCiIyMJDIykueff54+ffpw4sQJ5s6dS0BAABkZGdWqp64nGcYkUghR98gCxEIIXbz00kuVEp+LFy8SFhbGTz/9xFtvvcX69etNZXPmzGH69Om6Luw9atQo+vTpg6urq24xCCFqD7kTJYSoNby8vJg2bRoAu3btqlDWsmVLOnfurGsC4+rqSufOnXVN5IQQtYckUUKIWsXLywuAkpKSCvurGhPl5+fHxIkTAVi5cmWFMVb9+/cHYPTo0WiaxsaNGyvUV1JSgqurK5qm8cwzz1SK48UXX0TTNJYvX27ad7+uw61btxIWFkbTpk1xdXWlX79+bN68ucpjZ86cWaEb784xYpmZmZXOuX79Om+//Tbt27fH0dERLy8vIiMjOX/+fJXXEEJYl3TnCSFqlUOHDgHw6KOP3vfYsWPHcuDAAZKSkmjXrh2hoaGmss6dOwMwaNAgYmJiiI+PZ/To0RWuk5+fD8C3336LUqpCUmO8EzZo0KBqxf3nP/+ZN998E4DevXvTrl07Tpw4wciRI037b9e9e3ciIyNZuXIlAJGRkRXKmzRpUuF9Xl4effv2JSsri379+vHYY4+xf/9+Vq1aRUJCAmlpadLNKIStKSGEsCFfX18FqBUrVpj2lZaWqnPnzqnFixcrR0dHZW9vr2JjYyucFxkZWek8pZRasWKFAlRkZGSV1zt+/LgCVIcOHSrs/+Mf/6gAFRAQoAD1/fff3/ecu10rLS1N2dvbKzs7O7Vu3boKZX/729+UpmkKUL6+vpXiA9S9fhUbrwmowYMHq7y8PFPZ1atXVffu3RWgZs+efdc6hBDWId15QghdTJw40dR1ZW9vj7e3N6+99hoBAQEkJCTw9NNPW+Q6HTt2xMfHhxMnTpCVlWXaHx8fj5OTEx988AEAO3furFAG1b8LtXjxYkpLSxk3bhxjx46tUPbrX/+a4cOHP2gzcHZ2ZsWKFbi4uJj2NWvWjOnTp1eIWQhhO5JECSF0cfsUB5GRkQwbNgwfHx+Sk5N54403OHHihMWuZUyGjInS9evXOXDgAKGhoQwePBiDwVAhCalpErVnzx4AJkyYUGX5nV115ggKCqpyQHuXLl0AZFyUEDqQJEoIoYuXXnqJ6Oho02vr1q1kZGTw9ttvk5ycTFhYGAUFBRa5ljEZMiZHCQkJFBcXEx4ejrOzM3369CExMZHCwkLKysrYvXs3dnZ2DBgwoFr1nzt3DgB/f/8qy++2vybatGlT5X7jnanCwsIHvoYQomYkiRJC1BqNGjXiww8/xNPTkwsXLrBq1SqL1Dtw4EA0TWPXrl0opUzJVHh4OFCeZBUWFpKYmMjhw4fJzc2lZ8+euLm5WeT6lmBnJ7+uhaht5H+lEKJWsbOzMy2PcvToUYvU2aJFCx577DGuXLlCWloa8fHxeHp60r17d6DinaqaduUBtG7dGqDKaQnutV8IUbdJEiWEqFXKyspMScedj/lXxcHBAag8r9SdjEnR6tWr+eGHH0x3p6B8SgIXFxd27txpVhIVFhZmqrsq97qjZjAYqhW/EKL2kSRKCFFrlJSUMGPGDLKzswGq9VSbt7c3AOnp6fc8zpgULVmyBKWUqSsPyrsRw8LCOHLkCElJSTRu3JiQkJBqx/3aa69hb2/PN998Q0xMTIWyr7/+mk2bNt03/h9//LHa1xNC1A4y2aYQQhf/93//Z3qqDSAnJ4e0tDTOnj0LwLvvvkvfvn3vW0+fPn1o1aoVqampBAYG0rVrVwwGA506dTItIQPld4sMBoNpAPbtSRSUJ1mxsbEUFRURHh6Oo6NjtdvSvXt35syZw1tvvcXo0aMJDg42TbZpfNrwz3/+c5Xnjhkzhj/96U8MGjSIAQMG0LRpUwDmzZuHh4dHtWMQQtieJFFCCF0kJSWRlJRkeu/g4EDLli0ZP348v/vd70zLttyPg4MDO3bs4N1332X//v2kpaVRVlZGWFhYhSTK+BTed999R4cOHSo97XZ7911NuvKMpk2bRqdOnfj4449JTU3lxx9/JCAggPXr19OzZ8+7JlGzZs3Czs6OjRs3smnTJoqKigCYMWOGJFFC1HKaUkrpHYQQQgghRF0jY6KEEEIIIcwgSZQQQgghhBkkiRJCCCGEMIMkUUIIIYQQZpAkSgghhBDCDJJECSGEEEKYQZIoIYQQQggzSBIlhBBCCGEGSaKEEEIIIcwgSZQQQgghhBkkiRJCCCGEMIMkUUIIIYQQZpAkSgghhBDCDP8PcRruv5PNSvMAAAAASUVORK5CYII=", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# Plot the metrics vs n_bits for each model\n", + "plt.rcParams.update({\"font.size\": 16})\n", + "for cls in model_hyperparameters:\n", + " plt.figure()\n", "\n", - "fig, axs = plt.subplots(nrows=len(MODELS), ncols=1, figsize=(8, 5 * len(MODELS)))\n", - "\n", - "for i, model in enumerate(MODELS):\n", - " axs[i].set_title(model.__name__)\n", - " axs[i].set_xlabel(\"Bit-width\")\n", - " axs[i].set_ylabel(\"Metric\")\n", - "\n", - " ap_scores = [\n", - " np.mean(scores_global[n_bits][f\"{model.__name__}_{q}\"][\"average_precision\"])\n", - " for n_bits in N_BITS_LIST\n", - " for q in [\"concrete\", \"fp32\"]\n", - " if n_bits in scores_global and f\"{model.__name__}_{q}\" in scores_global[n_bits]\n", - " ]\n", - " f1_scores = [\n", - " np.mean(scores_global[n_bits][f\"{model.__name__}_{q}\"][\"f1\"])\n", - " for n_bits in N_BITS_LIST\n", - " for q in [\"concrete\", \"fp32\"]\n", - " if n_bits in scores_global and f\"{model.__name__}_{q}\" in scores_global[n_bits]\n", - " ]\n", - "\n", - " axs[i].plot(N_BITS_LIST, ap_scores[::2], label=\"average_precision\", color=\"blue\")\n", - " axs[i].plot(\n", - " N_BITS_LIST,\n", - " ap_scores[1::2],\n", + " f1_scores = []\n", + " f1_scores_fp32 = []\n", + "\n", + " average_precision_scores = []\n", + " average_precision_scores_fp32 = []\n", + "\n", + " for n_bits in n_bits_list:\n", + " average_precision_scores.append(\n", + " np.mean(scores_global[n_bits][cls.__name__ + \"_concrete\"][\"average_precision\"])\n", + " )\n", + " average_precision_scores_fp32.append(\n", + " np.mean(scores_global[n_bits][cls.__name__ + \"_fp32\"][\"average_precision\"])\n", + " )\n", + "\n", + " f1_scores.append(np.mean(scores_global[n_bits][cls.__name__ + \"_concrete\"][\"f1\"]))\n", + " f1_scores_fp32.append(np.mean(scores_global[n_bits][cls.__name__ + \"_fp32\"][\"f1\"]))\n", + "\n", + " # plt.legend()\n", + " ap_relative = np.array(average_precision_scores) / average_precision_scores_fp32\n", + " f1_relative = np.array(f1_scores) / f1_scores_fp32\n", + " print(f\"ap relative: {ap_relative}, f1_relative: {f1_relative}\")\n", + " plt.plot(\n", + " n_bits_list,\n", + " average_precision_scores,\n", + " label=\"concrete_average_precision\",\n", + " color=\"blue\",\n", + " linewidth=2,\n", + " )\n", + " plt.plot(\n", + " n_bits_list,\n", + " average_precision_scores_fp32,\n", " label=\"fp32_average_precision\",\n", " color=\"blue\",\n", + " linewidth=2,\n", " linestyle=\"dashed\",\n", " )\n", - " axs[i].plot(N_BITS_LIST, f1_scores[::2], label=\"f1\", color=\"red\")\n", - " axs[i].plot(N_BITS_LIST, f1_scores[1::2], label=\"fp32_f1\", color=\"red\", linestyle=\"dashed\")\n", "\n", - " axs[i].legend()\n", + " plt.plot(n_bits_list, f1_scores, label=\"concrete_f1\", linewidth=2, color=\"red\")\n", + " plt.plot(\n", + " n_bits_list, f1_scores_fp32, label=\"fp32_f1\", color=\"red\", linewidth=2, linestyle=\"dashed\"\n", + " )\n", + "\n", + " plt.grid(True)\n", + " plt.xlim([1, 9])\n", + " plt.ylim([0, 1])\n", + " plt.xticks(np.arange(1, 10))\n", + " plt.legend()\n", + "\n", + " plt.title(cls.__name__)\n", + " plt.xlabel(\"Bitwidth\")\n", + " plt.ylabel(\"Metric\")\n", + " # Save the figure\n", + " plt.savefig(cls.__name__ + \".eps\", bbox_inches=\"tight\", dpi=300)\n", "\n", - "fig.tight_layout()\n", - "fig.savefig(\"accuracy.png\", bbox_inches=\"tight\", dpi=300)\n", - "plt.show()" + " plt.show()" ] }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Compiling and keygen for DecisionTreeClassifier with 1 bits...\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.33 seconds\n", - "Compiling and keygen for XGBClassifier with 1 bits...\n", + "1\n", + "0.17977666854858398\n", + "8\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.01 seconds\n", - "Compiling and keygen for RandomForestClassifier with 1 bits...\n", + "2\n", + "0.16558170318603516\n", + "20\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.37 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 2 bits...\n", + "3\n", + "0.15911293029785156\n", + "20\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.26 seconds\n", - "Compiling and keygen for XGBClassifier with 2 bits...\n", + "4\n", + "0.13340115547180176\n", + "21\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.23 seconds\n", - "Compiling and keygen for RandomForestClassifier with 2 bits...\n", + "5\n", + "0.32805609703063965\n", + "25\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.46 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 3 bits...\n", + "6\n", + "0.5209550857543945\n", + "23\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.15 seconds\n", - "Compiling and keygen for XGBClassifier with 3 bits...\n", + "7\n", + "1.0027987957000732\n", + "24\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.27 seconds\n", - "Compiling and keygen for RandomForestClassifier with 3 bits...\n", + "8\n", + "2.717672109603882\n", + "22\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.56 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 4 bits...\n", + "1\n", + "0.006901264190673828\n", + "200\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.13 seconds\n", - "Compiling and keygen for XGBClassifier with 4 bits...\n", + "2\n", + "0.2832973003387451\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.33 seconds\n", - "Compiling and keygen for RandomForestClassifier with 4 bits...\n", + "3\n", + "0.2766108512878418\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.60 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 5 bits...\n", + "4\n", + "0.38878536224365234\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.71 seconds\n", - "Compiling and keygen for XGBClassifier with 5 bits...\n", + "5\n", + "0.6568257808685303\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.63 seconds\n", - "Compiling and keygen for RandomForestClassifier with 5 bits...\n", + "6\n", + "1.3181705474853516\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 1.24 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 6 bits...\n", + "7\n", + "3.5940380096435547\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 0.64 seconds\n", - "Compiling and keygen for XGBClassifier with 6 bits...\n", + "8\n", + "14.139111280441284\n", + "350\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 1.33 seconds\n", - "Compiling and keygen for RandomForestClassifier with 6 bits...\n", + "1\n", + "0.3134579658508301\n", + "400\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 2.58 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 7 bits...\n", + "2\n", + "0.41345763206481934\n", + "650\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 1.59 seconds\n", - "Compiling and keygen for XGBClassifier with 7 bits...\n", + "3\n", + "0.5010931491851807\n", + "700\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 3.66 seconds\n", - "Compiling and keygen for RandomForestClassifier with 7 bits...\n", + "4\n", + "0.6012833118438721\n", + "750\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 7.72 seconds\n", - "\n", - "\n", - "Compiling and keygen for DecisionTreeClassifier with 8 bits...\n", + "5\n", + "1.1607680320739746\n", + "750\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 2.50 seconds\n", - "Compiling and keygen for XGBClassifier with 8 bits...\n", + "6\n", + "2.4746828079223633\n", + "750\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 15.13 seconds\n", - "Compiling and keygen for RandomForestClassifier with 8 bits...\n", + "7\n", + "7.209601640701294\n", + "750\n", + "Compiling and keygen...\n", "Predict in FHE\n", - "FHE execution time: 24.65 seconds\n", - "\n", - "\n" + "8\n", + "23.54421091079712\n", + "750\n" ] } ], "source": [ + "def predict_with_fhe(clf, X_sample):\n", + " \"\"\"Predict using FHE and return elapsed time.\"\"\"\n", + " print(\"Compiling and keygen...\")\n", + " clf.compile(X_sample[:100])\n", + " clf.fhe_circuit.keygen(force=False)\n", + "\n", + " print(\"Predict in FHE\")\n", + " start_time = time.time()\n", + " _ = clf.predict(X_sample[:1], fhe=\"execute\")\n", + " end_time = time.time()\n", + "\n", + " return end_time - start_time\n", + "\n", + "\n", + "def analyze_and_store(clf, X_sample, nodes_dict, scores_dict):\n", + " \"\"\"Analyze the model and store results.\"\"\"\n", + " elapsed_time = predict_with_fhe(clf, X_sample)\n", + "\n", + " model_name = clf.__class__.__name__\n", + " if model_name not in nodes_dict:\n", + " nodes_dict[model_name] = []\n", + " scores_dict[model_name] = []\n", + "\n", + " scores_dict[model_name].append(elapsed_time)\n", + "\n", + " shapes = analyze_gemm_computation(clf)\n", + " nodes_dict[model_name].append(shapes[0][0])\n", + "\n", + " print(clf.n_bits)\n", + " print(scores_dict[model_name][-1])\n", + " print(nodes_dict[model_name][-1])\n", + "\n", + "\n", "X, y = datasets[\"spambase\"][\"X\"], datasets[\"spambase\"][\"y\"]\n", - "scores = {model.__name__: [] for model in MODELS}\n", + "nodes_dict = {}\n", + "scores_dict = {}\n", "\n", - "for n_bits in N_BITS_LIST:\n", - " for model, model_params in MODELS.items():\n", - " clf = model(n_bits=n_bits, **model_params)\n", + "for model_name, hyperparameters in model_hyperparameters.items():\n", + " for n_bits in n_bits_list:\n", + " clf = model_name(n_bits=n_bits, **hyperparameters)\n", " clf.fit(X, y)\n", "\n", - " print(f\"Compiling and keygen for {model.__name__} with {n_bits} bits...\")\n", - " clf.compile(X)\n", - " clf.fhe_circuit.keygen(force=False)\n", - "\n", - " print(\"Predict in FHE\")\n", - " t0 = time.time()\n", - " y_pred_ = clf.predict(X[:1], fhe=\"execute\")\n", - " t1 = time.time()\n", - " print(f\"FHE execution time: {t1-t0:.2f} seconds\")\n", - " scores[model.__name__].append(t1 - t0)\n", - " print(\"\\n\")" + " if n_bits < 9:\n", + " analyze_and_store(clf, X, nodes_dict, scores_dict)" ] }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 12, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -634,26 +1538,46 @@ } ], "source": [ - "for model, model_params in MODELS.items():\n", - " plt.figure()\n", + "def plot_fhe_inference_time(n_bits_list, scores, model_hyperparameters):\n", + " \"\"\"Plot the FHE inference time against bitwidth for each model.\"\"\"\n", "\n", - " fhe_inference_times = scores[model.__name__]\n", + " # Calculate average inference time per node for each bitwidth\n", + " n_bits_timings = np.zeros((8,))\n", + " for model in model_hyperparameters:\n", + " for idx, n_bits in enumerate(n_bits_list):\n", + " if n_bits < 9:\n", + " n_bits_timings[idx] += (\n", + " scores[model.__name__][idx] / nodes_dict[model.__name__][idx] * 1000\n", + " )\n", + " n_bits_timings /= len(model_hyperparameters)\n", + "\n", + " # Plot setup\n", + " plt.figure(figsize=(10, 6))\n", + " plt.rcParams.update({\"font.size\": 16})\n", "\n", " plt.plot(\n", - " range(1, len(fhe_inference_times) + 1),\n", - " fhe_inference_times,\n", - " label=\"fhe_inference_time\",\n", + " range(1, 9),\n", + " n_bits_timings,\n", + " label=\"FHE Inference Time\",\n", " color=\"blue\",\n", + " linewidth=2,\n", + " marker=\"o\",\n", " )\n", - " plt.legend()\n", "\n", - " plt.suptitle(model.__name__, fontsize=10, fontweight=\"bold\")\n", - " plt.annotate(model_params, xy=(0.48, 1.04), xycoords=\"axes fraction\", fontsize=9, ha=\"center\")\n", - " plt.xlabel(\"Bit-width\")\n", - " plt.ylabel(\"Time (s)\")\n", + " plt.xlabel(\"Bitwidth\")\n", + " plt.ylabel(\"Time (ms)\")\n", + " plt.grid(True, which=\"both\")\n", + " plt.semilogy()\n", + " plt.ylim([0, 1000])\n", + " plt.xlim([0.5, 8.5])\n", + " plt.xticks(np.arange(1, 9))\n", + " plt.title(\"FHE Execution vs Precision\", pad=10)\n", "\n", - " plt.savefig(f\"{model.__name__}_fhe_inference_time.png\", bbox_inches=\"tight\", dpi=300)\n", - " plt.show()" + " plt.savefig(\"fhe_inference_time.eps\", bbox_inches=\"tight\", dpi=300)\n", + " plt.show()\n", + "\n", + "\n", + "plot_fhe_inference_time(n_bits_list, scores_dict, model_hyperparameters)" ] } ],