diff --git "a/Adversarial_sample_using_ART.ipynb" "b/Adversarial_sample_using_ART.ipynb"
new file mode 100644--- /dev/null
+++ "b/Adversarial_sample_using_ART.ipynb"
@@ -0,0 +1,592 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Adversarial Robustness toolbox (ART) example \n",
+ "\n",
+ "The script demonstrates a simple example of using ART with TensorFlow v2.x. The example train a small model on the MNIST\n",
+ "dataset and creates adversarial examples using the Fast Gradient Sign Method. Here we use the ART classifier to train\n",
+ "the model, it would also be possible to provide a pretrained model to the ART classifier.\n",
+ "The parameters are chosen for reduced computational requirements of the script and not optimised for accuracy.\n",
+ "\n",
+ "* reference: https://github.com/Trusted-AI/adversarial-robustness-toolbox/blob/main/examples/get_started_tensorflow_v2.py"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Image Classification\n",
+ "\n",
+ "* Date: 04/06/2024\n",
+ "* Author: Pankaj Maurya\n",
+ "* Type of attack: ART tool attack (FGSM)\n",
+ "\n",
+ "### Metadata\n",
+ "* Dataset: MNIST\n",
+ "* Size of training set:60,000\n",
+ "* Size of testing set : 10,000\n",
+ "* Number of class : 10\n",
+ "* Original Model: CNN model trained with ART classifier "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "\n",
+ " ![\"Open](\"https://colab.research.google.com/assets/colab-badge.svg\")
\n",
+ ""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "'\\nDescription: Uncomment and run to install libraries. Needed for running first time only. \\n'"
+ ]
+ },
+ "execution_count": 1,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: Uncomment and run to install libraries. Needed for running first time only. \n",
+ "\"\"\"\n",
+ "# !pip install adversarial-robustness-toolbox\n",
+ "# !pip instal tensorflow==2.10\n",
+ "# !pip matplotlib==3.8.2"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "C:\\Users\\PUM8KOR\\.conda\\envs\\AIShield\\lib\\site-packages\\art\\estimators\\certification\\__init__.py:29: UserWarning: PyTorch not found. Not importing DeepZ or Interval Bound Propagation functionality\n",
+ " warnings.warn(\"PyTorch not found. Not importing DeepZ or Interval Bound Propagation functionality\")\n"
+ ]
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: import library \n",
+ "\"\"\"\n",
+ "import numpy as np\n",
+ "from art.attacks.evasion import FastGradientMethod\n",
+ "from art.estimators.classification import TensorFlowV2Classifier\n",
+ "from art.utils import load_mnist\n",
+ "\n",
+ "import random\n",
+ "import numpy as np\n",
+ "import matplotlib.pyplot as plt"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 1: Load the MNIST dataset"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: Load mnist data with art functionality. \n",
+ "\"\"\"\n",
+ "(x_train, y_train), (x_test, y_test), min_pixel_value, max_pixel_value = load_mnist()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "'''\n",
+ "New Block: Additional block added for creating function for visualization of images\n",
+ "Description: visualize sample images. \n",
+ "'''\n",
+ "def plot_images(x_train, y_train=None, rows=3):\n",
+ " #visualize x_train and y_train\n",
+ " ROWS=rows\n",
+ " random_indices=random.sample(range(x_train.shape[0]),ROWS*ROWS)\n",
+ " # print(\"the random indices are: \", random_indices)\n",
+ "\n",
+ " #taking sample image from x_train dataset\n",
+ " sample_images=x_train[random_indices,: ]\n",
+ " if y_train is not None:\n",
+ " sample_label=y_train[random_indices]\n",
+ " else:\n",
+ " sample_label = [999999]*x_train.shape[0]\n",
+ " \n",
+ "\n",
+ " fig,ax=plt.subplots(nrows=ROWS,ncols=ROWS,figsize=(12,9),sharex=True,sharey=True)\n",
+ " for i in range(ROWS*ROWS):\n",
+ " subplot_row=i // ROWS\n",
+ " subplot_col=i % ROWS\n",
+ " ax[subplot_row,subplot_col].imshow(sample_images[i,:]) #,cmap='gray_r'\n",
+ " ax[subplot_row,subplot_col].set_title(\"No. %d\" % sample_label[i])\n",
+ " plt.tight_layout()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "((60000, 28, 28, 1), (60000, 10), (10000, 28, 28, 1), (10000, 10))"
+ ]
+ },
+ "execution_count": 5,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "New Block: check shape of image and label after loading. label are one hot encoded\n",
+ "Description: check shape of traing and label data\n",
+ "\"\"\"\n",
+ "x_train.shape, y_train.shape, x_test.shape, y_test.shape"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "'''\n",
+ "New Block: convert label from one hot encoder. it will randomly plot rows * rows labels with labels.\n",
+ "Description: visualize few training dataset\n",
+ "'''\n",
+ "plot_images(x_train, np.argmax(y_train, axis=-1), rows=4)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 2: Create the model"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.autotrackable has been moved to tensorflow.python.trackable.autotrackable. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.base has been moved to tensorflow.python.trackable.base. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.base_delegate has been moved to tensorflow.python.trackable.base_delegate. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.data_structures has been moved to tensorflow.python.trackable.data_structures. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.graph_view has been moved to tensorflow.python.checkpoint.graph_view. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.util has been moved to tensorflow.python.checkpoint.checkpoint. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.checkpoint_management has been moved to tensorflow.python.checkpoint.checkpoint_management. The old module will be deleted in version 2.9.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.resource has been moved to tensorflow.python.trackable.resource. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.asset has been moved to tensorflow.python.trackable.asset. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.python_state has been moved to tensorflow.python.trackable.python_state. The old module will be deleted in version 2.11.\n",
+ "WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.saving.checkpoint_options has been moved to tensorflow.python.checkpoint.checkpoint_options. The old module will be deleted in version 2.11.\n"
+ ]
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: Import librabry for tensorflow model\n",
+ "\"\"\"\n",
+ "import tensorflow as tf\n",
+ "from tensorflow.keras import Model\n",
+ "from tensorflow.keras.layers import Dense, Flatten, Conv2D, MaxPool2D"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: create class to design architecture of model\n",
+ "\"\"\"\n",
+ "class TensorFlowModel(Model):\n",
+ " \"\"\"\n",
+ " Standard TensorFlow model for unit testing.\n",
+ " \"\"\"\n",
+ "\n",
+ " def __init__(self):\n",
+ " super(TensorFlowModel, self).__init__()\n",
+ " self.conv1 = Conv2D(filters=4, kernel_size=5, activation=\"relu\")\n",
+ " self.conv2 = Conv2D(filters=10, kernel_size=5, activation=\"relu\")\n",
+ " self.maxpool = MaxPool2D(pool_size=(2, 2), strides=(2, 2), padding=\"valid\", data_format=None)\n",
+ " self.flatten = Flatten()\n",
+ " self.dense1 = Dense(100, activation=\"relu\")\n",
+ " self.logits = Dense(10, activation=\"linear\")\n",
+ "\n",
+ " def call(self, x):\n",
+ " \"\"\"\n",
+ " Call function to evaluate the model.\n",
+ "\n",
+ " :param x: Input to the model\n",
+ " :return: Prediction of the model\n",
+ " \"\"\"\n",
+ " x = self.conv1(x)\n",
+ " x = self.maxpool(x)\n",
+ " x = self.conv2(x)\n",
+ " x = self.maxpool(x)\n",
+ " x = self.flatten(x)\n",
+ " x = self.dense1(x)\n",
+ " x = self.logits(x)\n",
+ " return x\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: Create architecture of model. \n",
+ "\"\"\"\n",
+ "model = TensorFlowModel()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 10,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: Create loss object for model. \n",
+ "\"\"\"\n",
+ "loss_object = tf.keras.losses.CategoricalCrossentropy(from_logits=True)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 11,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: create optimizer\n",
+ "\"\"\"\n",
+ "optimizer = tf.keras.optimizers.Adam(learning_rate=0.01)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 3: Create the ART classifier"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 12,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: create classifier with art library\n",
+ "\"\"\"\n",
+ "classifier = TensorFlowV2Classifier(\n",
+ " model=model,\n",
+ " loss_object=loss_object,\n",
+ " optimizer=optimizer,\n",
+ " nb_classes=10,\n",
+ " input_shape=(28, 28, 1),\n",
+ " clip_values=(0, 1),\n",
+ ")\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 4: Train the ART classifier"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: train classifier. \n",
+ "\"\"\"\n",
+ "classifier.fit(x_train, y_train, batch_size=64, nb_epochs=5)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 5: Evaluate the ART classifier on benign test examples"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "'''\n",
+ "New Block: convert label from one hot encoder. it will randomly plot rows * rows labels with labels.\n",
+ "Description: visualize few testing dataset\n",
+ "'''\n",
+ "plot_images(x_test, np.argmax(y_test, axis=-1), rows=3)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 15,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Accuracy on benign test examples: 98.05%\n"
+ ]
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: check accuracy on test data\n",
+ "\"\"\"\n",
+ "predictions = classifier.predict(x_test)\n",
+ "accuracy = np.sum(np.argmax(predictions, axis=1) == np.argmax(y_test, axis=1)) / len(y_test)\n",
+ "print(\"Accuracy on benign test examples: {}%\".format(accuracy * 100))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 6: Generate adversarial test examples"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 16,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(10000, 28, 28, 1)"
+ ]
+ },
+ "execution_count": 16,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: Create adversarial sample with FGSM methods\n",
+ "\"\"\"\n",
+ "attack = FastGradientMethod(estimator=classifier, eps=0.2)\n",
+ "x_test_adv = attack.generate(x=x_test)\n",
+ "x_test_adv.shape"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 17,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAABBAAAAN6CAYAAAA3rbpCAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy89olMNAAAACXBIWXMAAA9hAAAPYQGoP6dpAABt9UlEQVR4nO3df5zVZZk//muYgQF0GEWEgWVEEEUTBUNF1BTUFbE00i21TKzUUrAUzYRM88fuaG5KFqH9EMjNnyW6KmGJgh9X0KQ1V1MSQkVhKClmAHWAmfP9w28TJJ57hjkzZ348n4/HeezOuV9z39e8jQu4eJ9zCjKZTCYAAAAAsuiU7wIAAACA1s8AAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQKBRZs2aFQUFBdG1a9d46623PrA+evToGDp0aIvWdO6550ZBQUF84hOfaNFzAdq61tDT/17D9h6VlZXNejZAe9Ma+vrq1avj8ssvjzFjxkRJSUkUFBTEggULmvVMWo4BAjukpqYmrr/++nyXEc8991zMmjUrunbtmu9SANqs1tDTr7nmmrjjjju2eeyyyy55rQmgrcpnX1+6dGnccMMN8dZbb8UBBxyQlxpoPgYI7JDhw4fHj3/841i1alXeashkMvHVr341zjrrrOjTp0/e6gBo61pDTx83blyceeaZ2zwMhwF2TD77+ogRI2Lt2rXxxz/+MSZPntzi59O8DBDYIVOnTo3a2toGTTa3bNkS1157bey1115RXFwce+65Z0ydOjVqamqaVMMdd9wRL774Yvz7v/97k/YB6OhaQ0+PiFi/fn3U1tY2eR+Aji6ffb2kpCR69uy5Q99L62eAwA4ZOHBgnHXWWQ2abJ5zzjlx5ZVXxkc/+tG4+eab4+ijj46Kioo4/fTTd/j89evXxze+8Y2YOnVqlJWV7fA+AOS/p0dEjBkzJnr06BHdu3ePk08+OV599dUm7QfQkbWGvk77ZIDADvvmN78ZW7ZsiRtuuOFDM7///e9j9uzZcc4558R9990XF1xwQcyePTsuvfTSeOCBB+KJJ57YobOvueaa6NatW1x88cU7Wj4AW8lXT+/evXucffbZMX369JgzZ05cdtllMX/+/Dj88MNj5cqVTfmRADq0fP5ZnfbLAIEdNmjQoPj85z8fP/rRj2L16tXbzcydOzci4gOvf7rkkksiIuKRRx5p9Ll//OMf43vf+17ceOONUVxc3OjvB+CD8tXTP/OZz8TMmTPjrLPOivHjx8e1114bjz76aKxdu9ZL1ACaIF99nfbNAIEmueKKK2LLli0f+vqq119/PTp16hSDBw/e5vmysrLYZZdd4vXXX2/0mV/72tfi8MMPj1NPPXWHagZg+/LR07fnyCOPjJEjR8Zjjz2Wk/0AOqrW0tdpPwwQaJJBgwbFmWeemXWyGRFRUFCQk/Mef/zxmDdvXnzta1+L1157rf6xZcuWePfdd+O1116L6urqnJwF0NG0dE/Ppry8PP761782+zkA7Vlr6uu0DwYINNnfJ5vbe33VgAEDoq6u7gNvhrVmzZpYt25dDBgwoFFnvfHGGxERccopp8TAgQPrH2+99VY8/vjjMXDgwLj99tt3/IcB6OBasqdn86c//Sl23333nO0H0FG1lr5O+2CAQJPttddeceaZZ8Ztt90WlZWV26ydeOKJERExbdq0bZ6/6aabIiLi4x//eP1zy5cvj+XLl2c965hjjok5c+Z84LH77rvHwQcfHHPmzImTTjopBz8VQMfUkj09IuIvf/nLB56bO3duLFmyJE444YTGlg/AP2npvk77VpTvAmgfvvnNb8Ydd9wRS5cujf3337/++WHDhsWECRPiRz/6Uaxbty6OPvroePbZZ2P27Nkxfvz4GDNmTH322GOPjYiI11577UPP2WOPPWKPPfb4wPMXXXRR9OnTJ8aPH5+znwmgo2qpnh4Rcfjhh8dBBx0UBx98cJSWlsbvfve7uP3226O8vDymTp3aLD8fQEfTkn09IuK6666LiIiXXnopIiLuuOOOeOqppyLi/TsiaLsMEMiJwYMHx5lnnhmzZ8/+wNpPfvKTGDRoUMyaNSvmzJkTZWVlMWXKlLjqqqvyUCkAKS3Z00877bR45JFH4te//nW888470bdv3zj33HPjqquuij59+jT1RwEgWv7P6t/61re2+XrrlxgbILRtBZlMJpPvIgAAAIDWzXsgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAEBSUb4L+Gd1dXWxatWqKCkpiYKCgnyXA+RIJpOJ9evXR79+/aJTJ7PLjkRfh/ZHT++49HRonxra15ttgDB9+vS48cYbo7KyMoYNGxbf//7349BDD01+36pVq6K8vLy5ygLybOXKldG/f/98l0EL0teh/dLTOx49Hdq3VF9vlgHCPffcE5MnT45bb701Ro4cGdOmTYuxY8fG0qVLo3fv3lm/t6SkJCIijowToyg6N0d5QB5sic3xVMyt/zVOx6GvQ/ujp3dcejq0Tw3t680yQLjpppvi3HPPjS984QsREXHrrbfGI488ErfffntcfvnlWb/377dCFUXnKCrQlKDdyLz/f9zu2PHo69AO6ekdlp4O7VQD+3rOX7S2adOmWLJkSRx33HH/OKRTpzjuuONi0aJFuT4OAAAAaAE5vwPh7bffjtra2ujTp882z/fp0ydeeeWVD+Rramqipqam/uvq6upclwRAC9LXAdoPPR3YWt7fNreioiJKS0vrH96UBaBt09cB2g89HdhazgcIvXr1isLCwlizZs02z69ZsybKyso+kJ8yZUpUVVXVP1auXJnrkgBoQfo6QPuhpwNby/lLGLp06RIjRoyI+fPnx/jx4yPi/c+LnT9/fkyaNOkD+eLi4iguLs51GQDkib4O0H7o6cDWmuVTGCZPnhwTJkyIgw8+OA499NCYNm1abNy4sf5TGQAAAIC2pVkGCKeddlr85S9/iSuvvDIqKytj+PDhMW/evA+8sSIAAADQNjTLACEiYtKkSdt9yQIAAADQ9uT9UxgAAACA1s8AAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACApKJ8FwAAAEDbtO7zo5KZZ26Ykcwc9O8XJDO9pz/doJpoPu5AAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASCrK9Ybf/va34+qrr97muSFDhsQrr7yS66OgXauaOziZWTz8F8nM2H7Dc1AN0FwKiouzrteMOTC5x+un1yUzPZ/qksxs/JeCZKbuIxuyrv/hyFnJPQoLWte/X/y5dmPW9Umvn5zc44byB5OZsXd9PZkZOPXZZCbqatMZgBby9ohMMlObSf8+VXVITTLTu0EV0ZxyPkCIiNh///3jscce+8chRc1yDAAAANBCmuVv9kVFRVFWVtYcWwMAAAB50Cz3EL766qvRr1+/GDRoUHzuc5+LN954ozmOAQAAAFpIzu9AGDlyZMyaNSuGDBkSq1evjquvvjo+9rGPxYsvvhglJSUfyNfU1ERNzT9e71JdXZ3rkgBoQfo6QPuhpwNby/kdCOPGjYtPf/rTceCBB8bYsWNj7ty5sW7durj33nu3m6+oqIjS0tL6R3l5ea5LAqAF6esA7YeeDmyt2d8GeZdddol99tknli1btt31KVOmRFVVVf1j5cqVzV0SAM1IXwdoP/R0YGvN/vEIGzZsiOXLl8fnP//57a4XFxdHceIjrABoO/R1gPZDTwe2lvM7EC699NJYuHBhvPbaa/H000/Hpz71qSgsLIwzzjgj10cBAAAALSTndyC8+eabccYZZ8TatWtj9913jyOPPDIWL14cu+++e66Poh2oGXdIMrN+j/T/THvdtigX5bSYqrmDk5nFw3+RzHx1Vfr6RWxuQAZoDp2G7pvMrLymMOv670felpti/jU329RktmRdfzdTl9yjUyb97xfFBc1+k2S93Tp1y7p+18DfNGCX7snE82d+L5k59d+PTWbq1q9vQD0AbUumriDfJdAAOf/d+e677871lgAAAECeNfubKAIAAABtnwECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5bsA2rfXrxmVdf2Vc2Yk93jkna7JzC237dvgmppb4ZDByczi4b/IyVkvfGt4MlMcv83JWcC2OnXvnsycc//cZGb8Tuuyrj+/aUtyj7OnX5TMxBHZz4mI2LiqJJkZ8FBd1vWCLdnXIyKiU0Ey8rfBXdL7tDGlr21OZorX69lA61FU1ieZOWvMk7k5bHP69wbyzx0IAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASUX5LoC26+0vj0pmXjlnRpPPufHCzyczxfHbJp+TK3+9OTf7PPJO12Sm+Fet5+eGjqagS+dkZvxO65KZlzZvyrp+5TGnJffot+LpZCb+Mx1pTXr/Ot8VALBlzz7JzBW9fpWTs/a8P5OTfWhe7kAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACCpKN8F0DoVDhmczCy5akaTzxn9pXOTmeJf/bbJ5+TS218elXV9yfCmX5eIiOkf/0QDUstychaQP+9lCrOub1nxegtVAgDb+tuQnXKyz0+r+yczXZ96OZmpy0UxNIk7EAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJKKGvsNTz75ZNx4442xZMmSWL16dcyZMyfGjx9fv57JZOKqq66KH//4x7Fu3bo44ogjYsaMGbH33nvnsm6aoHDI4GRm7hO/yMlZI64+P+t6r18tysk5LWnJVTOavMdhz/9bMlO6dFmTzwGaz6bheyUzD72zJJkpK6zKul7Yo0dyj9rq6mQGABrrrwdmcrLPa+/1SmbqNm7MyVk0r0bfgbBx48YYNmxYTJ8+fbvr3/nOd+KWW26JW2+9NZ555pnYaaedYuzYsfHee+81uVgAAAAgPxp9B8K4ceNi3Lhx213LZDIxbdq0uOKKK+KTn/xkRET87Gc/iz59+sQDDzwQp59+etOqBQAAAPIip++BsGLFiqisrIzjjjuu/rnS0tIYOXJkLFrU9m5VBwAAAN7X6DsQsqmsrIyIiD59+mzzfJ8+ferX/llNTU3U1NTUf13tdZwAbZq+DtB+6OnA1vL+KQwVFRVRWlpa/ygvL893SQA0gb4O0H7o6cDWcjpAKCsri4iINWvWbPP8mjVr6tf+2ZQpU6Kqqqr+sXLlylyWBEAL09cB2g89HdhaTl/CMHDgwCgrK4v58+fH8OHDI+L925yeeeaZOP/87X+cX3FxcRQXF+eyDADySF8HaD/0dGBrjR4gbNiwIZYt+8fn069YsSKef/756NmzZ+yxxx5x0UUXxXXXXRd77713DBw4ML71rW9Fv379Yvz48bmsGwAAAGhBjR4gPPfcczFmzJj6rydPnhwRERMmTIhZs2bFZZddFhs3bozzzjsv1q1bF0ceeWTMmzcvunbtmruq+VCFQwYnMxMfeTgnZ424evt3lWyt121t69M3Xr9mVANSz2ddfeSd9P/WS09clswArVvhgt8lMz8+bkwyM/WJB7Ou/+mSock9Blz1dDIDAI11+OF/yMk+nQoyOdmH/Gv0AGH06NGRyXz4/wAKCgrimmuuiWuuuaZJhQEAAACtR94/hQEAAABo/QwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACCpKN8F0DiFQwZnXZ/4yMPJPT7e/b1kZsTV5yczvW5blMy0JqlrFxHxyjkzmnzOlTd+IZnpFW3r2gE7pnbVmmTmrP++IOv67790U3KPj3a+OJkZOFXfAWBbhR/ZJ+v6Df9yewN26Z5MPPTa0GSmLF5uwFnkmzsQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkoryXQD/UDhkcDIz94lfNPmcEVefn8z0um1Rk89pbf56c272eeSdrlnX2+O1A3ZMZvOmZKbwvYKs690KuiT3qOvc4JIAoF5dt+y/gfQu7J6Tc955ZZec7EP+uQMBAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAICkonwX0FEUDhmczMx94hdNPufEMf+WzPRauqjJ57Q2b395VDKzZPiMnJx15Y1fyLreK9rf9QWaT78nt2Rd/+Pp7yX3ePQzNyYzX3xycjLT9aFnkxkA2o9Nu3ZtkXO6VBe0yDk0P3cgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBU19huefPLJuPHGG2PJkiWxevXqmDNnTowfP75+/eyzz47Zs2dv8z1jx46NefPmNbnYtmziIw/nZJ/Dnv+3rOulS5cl96gZd0gyU/yr3za4pmze/vKoJu9R8saWZGbJVTOafE5D9bptUdb1wiGDk3vUNuC/E9AxFM/N3m/P/+Nnk3vM3//+ZObNY9P/ZjD4oWQEgHZkxb8VNnmPP9e+k8wMmPOXZKa2yZXQEhp9B8LGjRtj2LBhMX369A/NnHDCCbF69er6x1133dWkIgEAAID8avQdCOPGjYtx48ZlzRQXF0dZWdkOFwUAAAC0Ls3yHggLFiyI3r17x5AhQ+L888+PtWvXNscxAAAAQAtp9B0IKSeccEKccsopMXDgwFi+fHlMnTo1xo0bF4sWLYrCwg++xqampiZqamrqv66urs51SQC0IH0doP3Q04Gt5fwOhNNPPz1OPvnkOOCAA2L8+PHx8MMPx29/+9tYsGDBdvMVFRVRWlpa/ygvL891SQC0IH0doP3Q04GtNfvHOA4aNCh69eoVy5Zt/13np0yZElVVVfWPlStXNndJADQjfR2g/dDTga3l/CUM/+zNN9+MtWvXRt++fbe7XlxcHMXFxc1dBgAtRF8HaD/0dGBrjR4gbNiwYZu7CVasWBHPP/989OzZM3r27BlXX311nHrqqVFWVhbLly+Pyy67LAYPHhxjx47NaeEAAABAy2n0AOG5556LMWPG1H89efLkiIiYMGFCzJgxI1544YWYPXt2rFu3Lvr16xfHH398XHvttSaXObJ4+C+yB1Y1ZJfnc1BJQ7XkWdk98k7XZGb6xz/RgJ22/3Kcv6tdmn0dIB/Kns7NPnUfOyjreuc/vJ7c4+Xr90pmitam/4jSZe/0m7kVPVmazPR8ZVP2cx59LrkHQFv0ucMWNXmPr7+Z/vNz7cuvNvkcWodGDxBGjx4dmUzmQ9cfffTRJhUEAAAAtD7N/iaKAAAAQNtngAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASUX5LqCjuGXwvsnMtXMHt0AlEYuH/yIn+xz2/L8lM+ue2z3r+ivnzMhJLY+80zWZmf7xTyQztUuX5aIcgJx5b0tufqu+8Np7kpl13+6ezHxy5+9nXf/s0s8l91g45OZkZubfDk1mLtptSTJTOLIgmamNTNb1kT+9JLnHgKufSWairjadAciRwv32TmYu2W12IpH+M/bip9N/z9krFicztA3uQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgqyncB/EPpicta5JyxMTwn+5RGut7NX949J2elTP/4J5KZ2qUtc30Bcmm3SVuSmcFf/0oyU7S+MJmp7b0pmVl+YO+s612/vlNyjy/XnZPMdPrr+mTmlH1GJTNbdkr/3Cv/tSDr+qvn/CC5x/4FE5OZAVcuSmYAciXTOd3/enTq2gKV0J64AwEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACCpKN8F0HbVjDskmVly1YwmnzP6S+cmM8VLf9vkcwBaoy1/ei2Z2ef8dKYhCoqLk5kXC7tlXa975w85qaWuAZmiN99KZxqwz75P98y6fsfxZck97j3r5mTm8nsmJDO1Ly1NZgAa4o2PZ+9tDbHgvc7JzJDpq5OZLU2uhNbCHQgAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQ15OOR61VUVMT9998fr7zySnTr1i0OP/zwuOGGG2LIkCH1mffeey8uueSSuPvuu6OmpibGjh0bP/zhD6NPnz45L57mUzhkcDKz4Kc/bvI5I64+P5np9atFTT4HgLRMTU060wJ1tLSCLl2yru/bJf0Z5/t3zr5HRMT6fXZJZrq/lIwANMgJ/7a4yXv833vlycyWFa83+RzajkbdgbBw4cKYOHFiLF68OH7zm9/E5s2b4/jjj4+NGzfWZy6++OJ46KGH4r777ouFCxfGqlWr4pRTTsl54QAAAEDLadQdCPPmzdvm61mzZkXv3r1jyZIlcdRRR0VVVVX89Kc/jTvvvDOOOeaYiIiYOXNm7LfffrF48eI47LDDclc5AAAA0GKa9B4IVVVVERHRs2fPiIhYsmRJbN68OY477rj6zL777ht77LFHLFrkNnQAAABoqxp1B8LW6urq4qKLLoojjjgihg4dGhERlZWV0aVLl9hll122yfbp0ycqKyu3u09NTU3UbPWay+rq6h0tCYBWQF8HaD/0dGBrO3wHwsSJE+PFF1+Mu+++u0kFVFRURGlpaf2jvDz9Rh0AtF76OkD7oacDW9uhAcKkSZPi4YcfjieeeCL69+9f/3xZWVls2rQp1q1bt01+zZo1UVZWtt29pkyZElVVVfWPlStX7khJALQS+jpA+6GnA1tr1EsYMplMXHjhhTFnzpxYsGBBDBw4cJv1ESNGROfOnWP+/Plx6qmnRkTE0qVL44033ohRo0Ztd8/i4uIoLi7ewfIBaG30dYD2Q08HttaoAcLEiRPjzjvvjAcffDBKSkrq39egtLQ0unXrFqWlpfGlL30pJk+eHD179owePXrEhRdeGKNGjfIJDAAAANCGNWqAMGPGjIiIGD169DbPz5w5M84+++yIiLj55pujU6dOceqpp0ZNTU2MHTs2fvjDH+akWHKjcMjgZGbuE7/IyVkDHzk36/o+t/l0DgB2TGGf3snM6k+nf8+bfelNWdf379wluccvN+6azJS88tdkpjaZAIjoVFKSzAzpvrzJ59zy+NhkZu94psnn0HY0+iUMKV27do3p06fH9OnTd7goAAAAoHXZ4U9hAAAAADoOAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASCrKdwHkVuGQwcnM3Cd+kZOzRlx9fjKzz22LcnIWAO1H9WcPS2b+tm9BMjP1M/clM58rmZfMrK2rzbq+95z073dl/y9db8nLi5MZgIYo6F+WzHypx8Kmn7Ml3dvoWNyBAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkFSU7wJonMIhg7Ouz33iFzk5Z/SXzk1mev1qUU7OAuDDFe4/JJnZvGu3ZOb1E9OZLf02NaimbEp33ZjMPHvw9GSmUxQkM1uiNpn5z79+JJmZdd+/Zl3f+5qnk3sAtKTaV5YlM0f/378lM2fs8dus60N+8rd0LckE7Yk7EAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJKK8l0AjbNm9O5N3mPE1ecnM71+tajJ5wDQdHVd079VL/tSYTLTues7yUzv0o3JzP8Muzfr+gVvHZHc44VNtcnMD9Ycm8w8+98HJDP9/+PpZGaPSGcAWpVMJhnZ6YQ/JTP/HbslEksbWBAdhTsQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIatQAoaKiIg455JAoKSmJ3r17x/jx42Pp0m0/G3T06NFRUFCwzeMrX/lKTosGAAAAWlZRY8ILFy6MiRMnxiGHHBJbtmyJqVOnxvHHHx9/+MMfYqeddqrPnXvuuXHNNdfUf929e/fcVdzB9bptUdb1sbcNT+8R2fcAoPXILHkpmdnnCy1QyP/vxPhoIvFuco+pcWgDTlqfTPSPpxuwDwCQK40aIMybN2+br2fNmhW9e/eOJUuWxFFHHVX/fPfu3aOsrCw3FQIAAAB516T3QKiqqoqIiJ49e27z/M9//vPo1atXDB06NKZMmRLvvPNOU44BAAAA8qxRdyBsra6uLi666KI44ogjYujQofXPf/azn40BAwZEv3794oUXXohvfOMbsXTp0rj//vu3u09NTU3U1NTUf11dXb2jJQHQCujrAO2Hng5sbYcHCBMnTowXX3wxnnrqqW2eP++88+r//wMOOCD69u0bxx57bCxfvjz22muvD+xTUVERV1999Y6WAUAro68DtB96OrC1HXoJw6RJk+Lhhx+OJ554Ivr37581O3LkyIiIWLZs2XbXp0yZElVVVfWPlStX7khJALQS+jpA+6GnA1tr1B0ImUwmLrzwwpgzZ04sWLAgBg4cmPye559/PiIi+vbtu9314uLiKC4ubkwZALRi+jpA+6GnA1tr1ABh4sSJceedd8aDDz4YJSUlUVlZGRERpaWl0a1bt1i+fHnceeedceKJJ8Zuu+0WL7zwQlx88cVx1FFHxYEHHtgsPwAAAADQ/Bo1QJgxY0ZERIwePXqb52fOnBlnn312dOnSJR577LGYNm1abNy4McrLy+PUU0+NK664ImcFAwAAAC2v0S9hyKa8vDwWLlzYpIIAAACA1meH3kQRAAAA6FgMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgqSjfBfyzTCYTERFbYnNEJs/FADmzJTZHxD9+jdNx6OvQ/ujpHZeeDu1TQ/t6qxsgrF+/PiIinoq5ea4EaA7r16+P0tLSfJdBC9LXof3S0zsePR3at1RfL8i0stFxXV1drFq1KkpKSqKgoCAiIqqrq6O8vDxWrlwZPXr0yHOF7Y/r27xc3/dlMplYv3599OvXLzp18uqpjuSf+7pfE83L9W1eru/79PSOy5/VW57r27xc3/c1tK+3ujsQOnXqFP3799/uWo8ePTr0f9Tm5vo2L9c3/CtVB/Vhfd2viebl+jYv11dP76j8WT1/XN/m5fo2rK8bGQMAAABJBggAAABAUpsYIBQXF8dVV10VxcXF+S6lXXJ9m5frC9vya6J5ub7Ny/WFD/Lronm5vs3L9W2cVvcmigAAAEDr0ybuQAAAAADyywABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAoFFmzZoVBQUF0bVr13jrrbc+sD569OgYOnRos9bw5JNPxsknnxzl5eXRtWvXKCsrixNOOCH+53/+p1nPBWhvWkNP/3sN23tUVlY269kA7U1r6OurV6+Oyy+/PMaMGRMlJSVRUFAQCxYsaNYzaTkGCOyQmpqauP766/Ny9h//+Mfo1KlTfOUrX4np06fHpZdeGpWVlXHUUUfFvHnz8lITQFuWz57+d9dcc03ccccd2zx22WWXvNYE0Fbls68vXbo0brjhhnjrrbfigAMOyEsNNJ+ifBdA2zR8+PD48Y9/HFOmTIl+/fq16NnnnHNOnHPOOds8d8EFF8SgQYNi2rRpccIJJ7RoPQBtXT57+t+NGzcuDj744LycDdDe5LOvjxgxItauXRs9e/aMX/ziF/HpT3+6Rc+nebkDgR0yderUqK2tbdBkc8uWLXHttdfGXnvtFcXFxbHnnnvG1KlTo6amJmf1dO/ePXbfffdYt25dzvYE6ChaS09fv3591NbWNnkfgI4un329pKQkevbsuUPfS+tngMAOGThwYJx11lnx4x//OFatWpU1e84558SVV14ZH/3oR+Pmm2+Oo48+OioqKuL0009vUg3V1dXx9ttvxyuvvBJTp06NF198MY499tgm7QnQEbWGnj5mzJjo0aNHdO/ePU4++eR49dVXm7QfQEfWGvo67ZMBAjvsm9/8ZmzZsiVuuOGGD838/ve/j9mzZ8c555wT9913X1xwwQUxe/bsuPTSS+OBBx6IJ554YofP/8xnPhO777577LfffvHd7343vvzlL8e3vvWtHd4PoCPLV0/v3r17nH322TF9+vSYM2dOXHbZZTF//vw4/PDDY+XKlU35kQA6tHz/WZ32yQCBHTZo0KD4/Oc/Hz/60Y9i9erV283MnTs3IiImT568zfOXXHJJREQ88sgjO3z+9ddfH7/+9a/jpz/9aRx22GGxadOm2LJlyw7vB9CR5aunf+Yzn4mZM2fGWWedFePHj49rr702Hn300Vi7dm38+7//e6P3A+B9+f6zOu2TAQJNcsUVV8SWLVs+9PVVr7/+enTq1CkGDx68zfNlZWWxyy67xOuvv77DZw8fPjz+9V//Nb74xS/Gb37zm3j22Wfj7LPP3uH9ADq6fPb0rR155JExcuTIeOyxx3KyH0BH1Vr6Ou2HAQJNMmjQoDjzzDOzTjYjIgoKCpq1ji5dusTJJ58c999/f7z77rvNehZAe9VaenpERHl5efz1r39t9nMA2rPW1NdpHwwQaLK/Tza39/qqAQMGRF1d3QfeDGvNmjWxbt26GDBgQM7qePfddyOTycT69etztidAR9Naevqf/vSn2H333XO2H0BH1Vr6Ou2DAQJNttdee8WZZ54Zt912W1RWVm6zduKJJ0ZExLRp07Z5/qabboqIiI9//OP1zy1fvjyWL1+ePO/Pf/7zB55bt25d/PKXv4zy8vLo3bt3Y38EAP5/Ld3T//KXv3zgublz58aSJUvihBNOaGz5APyTlu7rtG9F+S6A9uGb3/xm3HHHHbF06dLYf//9658fNmxYTJgwIX70ox/FunXr4uijj45nn302Zs+eHePHj48xY8bUZ//+EYyvvfZa1rPGjRsX/fv3j5EjR0bv3r3jjTfeiJkzZ8aqVavinnvuaZafD6Ajacmefvjhh8dBBx0UBx98cJSWlsbvfve7uP3226O8vDymTp3aLD8fQEfTkn09IuK6666LiIiXXnopIiLuuOOOeOqppyLi/TsiaLsMEMiJwYMHx5lnnhmzZ8/+wNpPfvKTGDRoUMyaNSvmzJkTZWVlMWXKlLjqqqt26KwvfvGLcffdd8fNN98c69ati1133TUOO+ywuPPOO+NjH/tYU38UgA6vJXv6aaedFo888kj8+te/jnfeeSf69u0b5557blx11VXRp0+fpv4oAETL9vWI+MBHq99+++31/78BQttWkMlkMvkuAgAAAGjdvAcCAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5buAf1ZXVxerVq2KkpKSKCgoyHc5QI5kMplYv3599OvXLzp1MrvsSPR1aH/09I5LT4f2qaF9vdkGCNOnT48bb7wxKisrY9iwYfH9738/Dj300OT3rVq1KsrLy5urLCDPVq5cGf379893GbQgfR3aLz2949HToX1L9fVmGSDcc889MXny5Lj11ltj5MiRMW3atBg7dmwsXbo0evfunfV7S0pKIiLiyDgxiqJzc5QH5MGW2BxPxdz6X+N0HPo6tD96eselp0P71NC+3iwDhJtuuinOPffc+MIXvhAREbfeems88sgjcfvtt8fll1+e9Xv/fitUUXSOogJNCdqNzPv/x+2OHY++Du2Qnt5h6enQTjWwr+d8gLBp06ZYsmRJTJkypf65Tp06xXHHHReLFi36QL6mpiZqamrqv66urs51SQC0IH0doP3Q04Gt5fxdb95+++2ora2NPn36bPN8nz59orKy8gP5ioqKKC0trX94TRVA26avA7Qfejqwtby/be6UKVOiqqqq/rFy5cp8lwRAE+jrAO2Hng5sLecvYejVq1cUFhbGmjVrtnl+zZo1UVZW9oF8cXFxFBcX57oMAPJEXwdoP/R0YGs5vwOhS5cuMWLEiJg/f379c3V1dTF//vwYNWpUro8DAAAAWkCzfArD5MmTY8KECXHwwQfHoYceGtOmTYuNGzfWfyoDAAAA0LY0ywDhtNNOi7/85S9x5ZVXRmVlZQwfPjzmzZv3gTdWBAAAANqGZhkgRERMmjQpJk2a1FzbAwAAAC0o75/CAAAAALR+BggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5bsA2q6if+mXzLxy6R5Z1/9t9OLkHtf1XtLgmprq45/+YjJT9IfXsq7XrqvKUTUAAACthzsQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkopyveG3v/3tuPrqq7d5bsiQIfHKK6/k+qgOqWhAedb1P03Ivh4RUTi8Kie1/NdBtycz+3XJPqPq1IAZVl3UNbimpnrovp8kM1947fis62uPyFU1AADA3xX26JHMHP0/q7OuL3und3KPVaf3Sma2rHg9mWmPcj5AiIjYf//947HHHvvHIUXNcgwAAADQQprlb/ZFRUVRVlbWHFsDAAAAedAsA4RXX301+vXrF127do1Ro0ZFRUVF7LHHHtvN1tTURE1NTf3X1dXVzVESAC1EXwdoP/R0YGs5fxPFkSNHxqxZs2LevHkxY8aMWLFiRXzsYx+L9evXbzdfUVERpaWl9Y/y8vRr+AFovfR1gPZDTwe2lvMBwrhx4+LTn/50HHjggTF27NiYO3durFu3Lu69997t5qdMmRJVVVX1j5UrV+a6JABakL4O0H7o6cDWmv3dDXfZZZfYZ599YtmyZdtdLy4ujuLi4uYuA4AWoq8DtB96OrC1nN+B8M82bNgQy5cvj759+zb3UQAAAEAzyfkA4dJLL42FCxfGa6+9Fk8//XR86lOfisLCwjjjjDNyfRQAAADQQnL+EoY333wzzjjjjFi7dm3svvvuceSRR8bixYtj9913z/VRHdLLV/fOur70X7+f3KMu6nJUTXr+NGdD9nrvrjw0uUddFDS4omz+c89fJjMDirokM+eVLcy6XhEHNrgmgJRO3bsnM0uvPyCZWTD+u8nMv97x9WRm0PUvZl2v+5A3TQaAplo1YWgyM7nnE1nX39n1heQeJ4y6OJnpseL1ZKY9yvkA4e677871lgAAAECeNft7IAAAAABtnwECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5bsAGufQvVc0eY8XN2WSmc888NVkZq9fvJfMdF69Luv6lj+9ltyjIQoHD0xm/vCrPsnMgKK/5aIcgAbrVFKSdX39fbsn93j1gBkNOKl7MvGHs6cnM3f/W/Z6frdhQHKP+397cDJT0G1LMtPpL12SmX5P1SUz6/tn/+PQxn9J/745YG7698SGyBQWJDNrDu2azNQmLs3Ama8l99jy1qpkBiBXNo1N/95w76U3NmCn7D3yYzdfktyh751PN+CcjskdCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAElF+S6Axqn+eG3W9fG7fCq9Sc2mZGTw6sUNLSmrLTnZJW3diD7JzLjuf8vJWec8cF7W9b0iN9cO6BhWXHpA1vWXDvhBC1XSMKfv/JcmrUdEfOek53JSS2FB+t9Bas+oy8lZSWe1zDG58vyX0r9DTx14aAtUAnQEBZ27JDNvfD7733MiIgYWdU1mLq88JOt63+8+ndyDD+cOBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkoryXQCNU7uuKnsgtd7KFBSl/yf4p2uyf5ZrRMRLExryOem5mZft8avNOdkHaP9WX3J4MvP7c76Xdf2hd3ZN7nHT1z+bzKz8RF0ys1PPd5OZXOjTY30y8y87rUtmpvSdl8zURUG6nsLs12bXTt2Se+RKXWSSmT9u3tTkc77wg0uSmb7hs9KB3PjjTQclM0vH/DAnZz306Mis6wNjUU7O6ajcgQAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBU1NhvePLJJ+PGG2+MJUuWxOrVq2POnDkxfvz4+vVMJhNXXXVV/PjHP45169bFEUccETNmzIi99947l3XTRmRGDcu6vuKk7sk9/m/CLclMXYMrym7Y/3wxmRnw2JIcnQa0d5nCdKYosocu++Xnk3sMfGBRMrPPA+laWpM1DchcFIfn5KxOwz+SdX3jgJ1zck5DFNSmM10ffrbJ5/SNp5u8B0BERMFB+yczN53w82SmUxQkMxVrs/friIi9f/RW1vUtyR3IptF3IGzcuDGGDRsW06dP3+76d77znbjlllvi1ltvjWeeeSZ22mmnGDt2bLz33ntNLhYAAADIj0bfgTBu3LgYN27cdtcymUxMmzYtrrjiivjkJz8ZERE/+9nPok+fPvHAAw/E6aef3rRqAQAAgLxo9AAhmxUrVkRlZWUcd9xx9c+VlpbGyJEjY9GiRdsdINTU1ERNTU3919XV1bksCYAWpq8DtB96OrC1nL6JYmVlZURE9OnTZ5vn+/TpU7/2zyoqKqK0tLT+UV5ensuSAGhh+jpA+6GnA1vL+6cwTJkyJaqqquofK1euzHdJADSBvg7QfujpwNZy+hKGsrKyiIhYs2ZN9O3bt/75NWvWxPDhw7f7PcXFxVFcXJzLMgDII30doP3Q04Gt5fQOhIEDB0ZZWVnMnz+//rnq6up45plnYtSoUbk8CgAAAGhBjb4DYcOGDbFs2bL6r1esWBHPP/989OzZM/bYY4+46KKL4rrrrou99947Bg4cGN/61reiX79+MX78+FzWDQAAALSgRg8QnnvuuRgzZkz915MnT46IiAkTJsSsWbPisssui40bN8Z5550X69atiyOPPDLmzZsXXbt2zV3VNMnGU0cmM72++lpOzvrRoB9mXd+1U/p/F3U5qaRhDu6ffl3fX/cfknW99qWluSoHaOP6P/q3dOhr2Zcf++yNyS0u+9jJycz/PbxvMjPgl2uSmdo/Lk9m2pq65/+Qdb3b8y1TB0BbtOLywmTm492rkpmfVqffoPPXVx6VzHR77dlkhh3X6AHC6NGjI5PJfOh6QUFBXHPNNXHNNdc0qTAAAACg9cj7pzAAAAAArZ8BAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkFeW7AHKr6F/6JTM/+M9bkpn9uuRqttQlR/u0jJ8O+E0y89R/d826fv495yX3GDh1UYNrAtqwP72ZjBz2v6dnXV980N3JPX6+52PpWialM2sveDeZGfnrr2Vd7/N4+o8WpXf9NpmJutp0BoBmterSw5OZGz76s5ycNe2/xicz5Q88nZOz2HHuQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgqyncB5FiXzsnIfl1az9xon4fOb7mz9l6VzPz3vnOSmaO6bsq6/rmPL0zu8fTULskM0PbVrV+fzOw+IXvfPvj0Sck9qke9m8wsOvoHycxunbolM8tO+FH2wAnJLeJrF41KZl65bFgyU/jE79KHAfChCg7aP+v6k1/7z+QePTp1TWYO/d0ZycyA76Z7el0yQXNrPX+TBAAAAFotAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASCrKdwHk1pYVryczpxx+SjKz7Nz+ycyuL2eSmdKfL866vk88m9yjJQ2/74vJzMtH3JF1fWqv/0vuMeSWC5KZvb/6TDIDtH21a/+adb339KeTe/Senj5nwrBzkpmlXyxNZq4ce3/W9c+XVCb3+F6/RcnMR74yJJnZ84lkBIAsln29c9b1nTsV5+Sc3a/ukszUvfdeTs6iebkDAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIKmosd/w5JNPxo033hhLliyJ1atXx5w5c2L8+PH162effXbMnj17m+8ZO3ZszJs3r8nFkhtbXl+ZzOx5RTrTHu3x6f9LZv7nT3VZ1w8uzr4eEXHCqN8nMytKSpKZuvXrkxmAiIi637+czOz9tfQ+9+y0d9b1m77yb8k9/nfyD5KZPxw5K5nZ7z8mJjMDpy5KZgDaoz/eemgys+zoWxOJguQe+/803YsHPNdyvbhgxP5Z1zNLXmqhStqnRt+BsHHjxhg2bFhMnz79QzMnnHBCrF69uv5x1113NalIAAAAIL8afQfCuHHjYty4cVkzxcXFUVZWtsNFAQAAAK1Ls7wHwoIFC6J3794xZMiQOP/882Pt2rXNcQwAAADQQhp9B0LKCSecEKecckoMHDgwli9fHlOnTo1x48bFokWLorCw8AP5mpqaqKmpqf+6uro61yUB0IL0dYD2Q08HtpbzOxBOP/30OPnkk+OAAw6I8ePHx8MPPxy//e1vY8GCBdvNV1RURGlpaf2jvLw81yUB0IL0dYD2Q08HttbsH+M4aNCg6NWrVyxbtmy761OmTImqqqr6x8qVHfPd/wHaC30doP3Q04Gt5fwlDP/szTffjLVr10bfvn23u15cXBzFxcXNXQYALURfB2g/9HRga40eIGzYsGGbuwlWrFgRzz//fPTs2TN69uwZV199dZx66qlRVlYWy5cvj8suuywGDx4cY8eOzWnhAAAAQMtp9ADhueeeizFjxtR/PXny5IiImDBhQsyYMSNeeOGFmD17dqxbty769esXxx9/fFx77bUml7QbZ/76K1nXXzlpenKPm/v9v2TmpAPPTWYK/uf5ZAYgl+o2bsy63ve7Tyf3OPYPX05mZsz4XjLzzOe/m8yc9sgFyYxeCrQ1dR87KJm58MjH0vtEJuv6DWv3T+4x4MpFyUxLyix5Kd8ltGuNHiCMHj06MpkP/x/ao48+2qSCAAAAgNan2d9EEQAAAGj7DBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIKko3wVAWzNkn7eavMcT7+6czBS9vSGZqW1yJQAtr/hXv01mznrx7GRm8UF3JzPLPtM1mdn7f5IRgFbl9YnpPwVeuOuryczLmzdnXZ939dHJPXaKZ5IZ2g93IAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5bsAaE3enHp4MvP8kO9nXa9rwDkXPPKFZGbvpYsbsBNA29Np+EeSmZ8Ovb0BO3VJJorX+rcSoG1ZeUX6z6O/O/KmBuzUOZk4bdbkrOsDHl6S3CPTgEpoP/yuCgAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkFSU7wLag4Li4mTmja+PSGZu/+L3k5mJN0zKut7rtkXJPTqqTiUlycxPz0n/N+hcUJh1/dfvdE3use+P/pbM1CYTQGvXkN8fMjU1LVBJyykcPDCZuf6B25OZ/Tt3SWa+97fBycygH/8pmdmSTADkRsGI/ZOZW7/4w2SmuKBzMnPeytHJzICKJVnX29vvUTSdOxAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgqaky4oqIi7r///njllVeiW7ducfjhh8cNN9wQQ4YMqc+89957cckll8Tdd98dNTU1MXbs2PjhD38Yffr0yXnxrUXtoR9JZv73/O/l5KzPXfho1vWZPU5I7rHHHcuTmbreuyYzW0q7JTMpXd54O33O6yubfE5ExJqf90tmDiquS2aq6jZlXb/wnvOTe+z50qJkBmj7lt5yYDKz76WvZF2vW78+V+XkREFR9j86vPWfXZN77N+5S05q+a8/HZLM7L56aU7OAsiFDf/+bjJzVLqNxgVvHZHMrDqsdf3+QfvQqDsQFi5cGBMnTozFixfHb37zm9i8eXMcf/zxsXHjxvrMxRdfHA899FDcd999sXDhwli1alWccsopOS8cAAAAaDmNugNh3rx523w9a9as6N27dyxZsiSOOuqoqKqqip/+9Kdx5513xjHHHBMRETNnzoz99tsvFi9eHIcddljuKgcAAABaTJPeA6GqqioiInr27BkREUuWLInNmzfHcccdV5/Zd999Y4899ohFi9yyDQAAAG1Vo+5A2FpdXV1cdNFFccQRR8TQoUMjIqKysjK6dOkSu+yyyzbZPn36RGVl5Xb3qampiZqamvqvq6urd7QkAFoBfR2g/dDTga3t8B0IEydOjBdffDHuvvvuJhVQUVERpaWl9Y/y8vIm7QdAfunrAO2Hng5sbYcGCJMmTYqHH344nnjiiejfv3/982VlZbFp06ZYt27dNvk1a9ZEWVnZdveaMmVKVFVV1T9WrszNO+4DkB/6OkD7oacDW2vUSxgymUxceOGFMWfOnFiwYEEMHDhwm/URI0ZE586dY/78+XHqqadGRMTSpUvjjTfeiFGjRm13z+Li4iguLt7B8gFobfR1gPZDTwe21qgBwsSJE+POO++MBx98MEpKSurf16C0tDS6desWpaWl8aUvfSkmT54cPXv2jB49esSFF14Yo0aN8gkMAAAA0IY1aoAwY8aMiIgYPXr0Ns/PnDkzzj777IiIuPnmm6NTp05x6qmnRk1NTYwdOzZ++MMf5qTY1qqo6t1k5veb0vsM65LOTNx1adb1Cy96NbnHTWftm8yM2fkPyUxD6k2ZXT0gmbn57vHJTFH6P0H8ctiNDago/UNdteZjWdf3/KZPHAHe1/fxwmSm+r7ds66XXph9PSKi9tU/NbimbAr79E5mXv72nlnXF370pgac1L1hBSXULNotJ/sA5ErRv/TLun5Cv/SfsWszdcnMr34/NJnZJ55LZqCxGv0ShpSuXbvG9OnTY/r06TtcFAAAANC67PCnMAAAAAAdhwECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5buA9qDuhVeSmdPmn5/MvDJuRi7KSbqo5x9a5JyG+EKPlcnMhPO+l6PTuiQTR/7v55KZ3ue/k0i82cB6gPZulwV/Sma+ff0jWdfL59ck9zjhd+cmM3WZgmRm+rA7k5kjiusSie7JPRriuD98KpnpX/F0Ts4CyJXafrtlXf/Gbg8l93h+U20ys98Nf03XkkxA47kDAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIKko3wV0FPt+9Q/JzIHf/Goys2m32lyU02JGHfBq1vXZez7WQpVEHDgzfX33vPLZZGZLXdv6bwDkT+2aPyczn/1F9t609LPTk3v87pCfN7im5vZuZlMyc+jic5KZPb+8KpnRjYHWpvKwkibv8Zfa9B7v7LNbMlP86p+aXAv8M3cgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBXlu4COou6dd5KZPb+5qAUqaVlrE+ufiBEtUkdExJ7R/q4v0Pb9y8LarOu//OSuyT1O3elvuSon6d3MpqzrHz//q8k9yh96NpnJflUAWqceb2xp8h73vX1IMlNXVNDkc2BHuAMBAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAICkRg0QKioq4pBDDomSkpLo3bt3jB8/PpYuXbpNZvTo0VFQULDN4ytf+UpOiwYAAABaVlFjwgsXLoyJEyfGIYccElu2bImpU6fG8ccfH3/4wx9ip512qs+de+65cc0119R/3b1799xVDADtSNeHn826PmvRiOQe15y7bzKz+7FvJTOv/7Esmdl32l+yrnd9NfvPA9CedXswew/8xIPpnh6xIX1O6LXkR6MGCPPmzdvm61mzZkXv3r1jyZIlcdRRR9U/37179ygrS/8hBAAAAGgbmvQeCFVVVRER0bNnz22e//nPfx69evWKoUOHxpQpU+Kdd95pyjEAAABAnjXqDoSt1dXVxUUXXRRHHHFEDB06tP75z372szFgwIDo169fvPDCC/GNb3wjli5dGvfff/9296mpqYmampr6r6urq3e0JABaAX0doP3Q04Gt7fAAYeLEifHiiy/GU089tc3z5513Xv3/f8ABB0Tfvn3j2GOPjeXLl8dee+31gX0qKiri6quv3tEyAGhl9HWA9kNPB7a2Qy9hmDRpUjz88MPxxBNPRP/+/bNmR44cGRERy5Yt2+76lClToqqqqv6xcuXKHSkJgFZCXwdoP/R0YGuNugMhk8nEhRdeGHPmzIkFCxbEwIEDk9/z/PPPR0RE3759t7teXFwcxcXFjSkDgFZMXwdoP/R0YGuNGiBMnDgx7rzzznjwwQejpKQkKisrIyKitLQ0unXrFsuXL48777wzTjzxxNhtt93ihRdeiIsvvjiOOuqoOPDAA5vlBwAAAACaX6MGCDNmzIiIiNGjR2/z/MyZM+Pss8+OLl26xGOPPRbTpk2LjRs3Rnl5eZx66qlxxRVX5KxgAOhIatf+NZn5l+ufTm90fTqyd7yerie9DQDQTjX6JQzZlJeXx8KFC5tUEAAAAND67NCbKAIAAAAdiwECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQV5buAf5bJZCIiYktsjsjkuRggZ7bE5oj4x69xOg59HdofPb3j0tOhfWpoX291A4T169dHRMRTMTfPlQDNYf369VFaWprvMmhB+jq0X3p6x6OnQ/uW6usFmVY2Oq6rq4tVq1ZFSUlJFBQUREREdXV1lJeXx8qVK6NHjx55rrD9cX2bl+v7vkwmE+vXr49+/fpFp05ePdWR/HNf92uiebm+zcv1fZ+e3nH5s3rLc32bl+v7vob29VZ3B0KnTp2if//+213r0aNHh/6P2txc3+bl+oZ/peqgPqyv+zXRvFzf5uX66ukdlT+r54/r27xc34b1dSNjAAAAIMkAAQAAAEhqEwOE4uLiuOqqq6K4uDjfpbRLrm/zcn1hW35NNC/Xt3m5vvBBfl00L9e3ebm+jdPq3kQRAAAAaH3axB0IAAAAQH4ZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZINAos2bNioKCgujatWu89dZbH1gfPXp0DB06tFlrWL16dVx++eUxZsyYKCkpiYKCgliwYEGzngnQHrWGnj5//vz44he/GPvss0907949Bg0aFOecc06sXr26Wc8FaI9aQ1+PiFiyZEl84hOfiLKysth5553jwAMPjFtuuSVqa2ub/WyalwECO6Smpiauv/76vJy9dOnSuOGGG+Ktt96KAw44IC81ALQn+ezp3/jGN2LBggXxqU99Km655ZY4/fTT4957742DDjooKisr81ITQFuXz76+ZMmSOPzww+O1116Lb3zjG/Hd7343Bg0aFF/72tdi8uTJeamJ3DFAYIcMHz48fvzjH8eqVata/OwRI0bE2rVr449//KMmBJAD+ezpN910UyxbtixuuOGGOOecc+I//uM/4uGHH441a9bED37wgxavB6A9yGdfv+222yIi4sknn4yLL744vvzlL8cDDzwQRx11VMyaNavF6yG3DBDYIVOnTo3a2toGTTa3bNkS1157bey1115RXFwce+65Z0ydOjVqamp26OySkpLo2bPnDn0vAB+Uz55+1FFHRadOnT7wXM+ePePll1/eoT0BOrp89vXq6uro2rVr7LLLLts837dv3+jWrdsO7UnrYYDADhk4cGCcddZZDZpsnnPOOXHllVfGRz/60bj55pvj6KOPjoqKijj99NNbqFoAsmltPX3Dhg2xYcOG6NWrV872BOhI8tnXR48eHdXV1fHlL385Xn755Xj99dfj1ltvjfvvvz+mTJmyQ3vSehggsMO++c1vxpYtW+KGG2740Mzvf//7mD17dpxzzjlx3333xQUXXBCzZ8+OSy+9NB544IF44oknWrBiAD5Ma+rp06ZNi02bNsVpp52Wk/0AOqJ89fVzzz03Jk2aFLNnz46PfOQjseeee8akSZPilltuia997WtN+ZFoBQwQ2GGDBg2Kz3/+8/GjH/3oQ98te+7cuRERH3ivgksuuSQiIh555JHmLRKABmktPf3JJ5+Mq6++Oj7zmc/EMccc0+T9ADqqfPX1wsLC2GuvvWLs2LExe/bsuOeee+Kkk06KCy+8MB544IFG70frYoBAk1xxxRWxZcuWD3191euvvx6dOnWKwYMHb/N8WVlZ7LLLLvH666+3RJkANEC+e/orr7wSn/rUp2Lo0KHxk5/8pEl7AZCfvn799dfHDTfcEHfddVecddZZ8ZnPfCbmzJkTRx55ZEycODG2bNmyQz8LrYMBAk0yaNCgOPPMM7NONiMiCgoKWrAqAHZEPnv6ypUr4/jjj4/S0tKYO3dulJSU5PwMgI4mH339hz/8YRxzzDGx8847b/P8ySefHKtWrYrXXnstZ2fR8gwQaLK/Tza39/qqAQMGRF1dXbz66qvbPL9mzZpYt25dDBgwoKXKBKAB8tHT165dG8cff3zU1NTEo48+Gn379t2hfQD4oJbu62vWrIna2toPPL958+aICHcgtHEGCDTZXnvtFWeeeWbcdtttUVlZuc3aiSeeGBHvvyHW1m666aaIiPj4xz9e/9zy5ctj+fLlzVssAFm1dE/fuHFjnHjiifHWW2/F3LlzY++9927iTwDA1lq6r++zzz7xm9/8JtauXVv/XG1tbdx7771RUlISe+21147+KLQCRfkugPbhm9/8Ztxxxx2xdOnS2H///eufHzZsWEyYMCF+9KMfxbp16+Loo4+OZ599NmbPnh3jx4+PMWPG1GePPfbYiIgG3dZ03XXXRUTESy+9FBERd9xxRzz11FMR8f6UFYAd15I9/XOf+1w8++yz8cUvfjFefvnlePnll+vXdt555xg/fnxOfzaAjqgl+/rll18eZ555ZowcOTLOO++86NatW9x1112xZMmSuO6666Jz587N8jPSQjLQCDNnzsxEROa3v/3tB9YmTJiQiYjM/vvvv83zmzdvzlx99dWZgQMHZjp37pwpLy/PTJkyJfPee+9tkxswYEBmwIABDaojIj70AUDDtIaePmDAgA/t5w39PQGA97WGvp7JZDLz5s3LHH300ZlevXplunTpkjnggAMyt9566w7/XLQeBZlMJtPCMwsAAACgjfEeCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBXlu4B/VldXF6tWrYqSkpIoKCjIdzlAjmQymVi/fn3069cvOnUyu+xI9HVof/T0jktPh/apoX292QYI06dPjxtvvDEqKytj2LBh8f3vfz8OPfTQ5PetWrUqysvLm6ssIM9WrlwZ/fv3z3cZtCB9HdovPb3j0dOhfUv19WYZINxzzz0xefLkuPXWW2PkyJExbdq0GDt2bCxdujR69+6d9XtLSkoiIuLIODGKonNzlAfkwZbYHE/F3Ppf43Qc+jq0P3p6x6WnQ/vU0L7eLAOEm266Kc4999z4whe+EBERt956azzyyCNx++23x+WXX571e/9+K1RRdI6iAk0J2o3M+//H7Y4dj74O7ZCe3mHp6dBONbCv53yAsGnTpliyZElMmTKl/rlOnTrFcccdF4sWLfpAvqamJmpqauq/rq6uznVJALQgfR2g/dDTga3l/F1v3n777aitrY0+ffps83yfPn2isrLyA/mKioooLS2tf3hNFUDbpq8DtB96OrC1vL9t7pQpU6Kqqqr+sXLlynyXBEAT6OsA7YeeDmwt5y9h6NWrVxQWFsaaNWu2eX7NmjVRVlb2gXxxcXEUFxfnugwA8kRfB2g/9HRgazm/A6FLly4xYsSImD9/fv1zdXV1MX/+/Bg1alSujwMAAABaQLN8CsPkyZNjwoQJcfDBB8ehhx4a06ZNi40bN9Z/KgMAAADQtjTLAOG0006Lv/zlL3HllVdGZWVlDB8+PObNm/eBN1YEAAAA2oZmGSBEREyaNCkmTZrUXNsDAAAALSjvn8IAAAAAtH4GCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQVJTvAmi7Cnfrmcx85n9ezLr+uZLVyT1e3rw5mfnalyclM10efS6ZAQAAYPvcgQAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBUlO8CaLtWfmnfZOaMkkezrtc14Jz9OndOZjZcWJ3M9MxeCgAt5LXrRiUz+x+1LJn5/Rv9k5nBZ/5vg2oCSCns0SMd6ts763Lt0nRva6laIiKWfXunrOsvHXV7co+Z1eXJzIzp45OZzhsyyUyv+19KZmqr038vYMfl/A6Eb3/721FQULDNY99903/RBAAAAFqvZrkDYf/994/HHnvsH4cUudEBAAAA2rJm+Zt9UVFRlJWVNcfWAAAAQB40ywDh1VdfjX79+kXXrl1j1KhRUVFREXvsscd2szU1NVFTU1P/dbXXrAC0afo6QPuhpwNby/l7IIwcOTJmzZoV8+bNixkzZsSKFSviYx/7WKxfv367+YqKiigtLa1/lJen34QDgNZLXwdoP/R0YGs5HyCMGzcuPv3pT8eBBx4YY8eOjblz58a6devi3nvv3W5+ypQpUVVVVf9YuXJlrksCoAXp6wDth54ObK3Z391wl112iX322SeWLdv+R5YUFxdHcXFxc5cBQAvR1wHaDz0d2FrO70D4Zxs2bIjly5dH3759m/soAAAAoJnk/A6ESy+9NE466aQYMGBArFq1Kq666qooLCyMM844I9dHkWc1u2byXQIALaxw8MCs669c3Du5x5KTv5vM7Nwp/S+ea/Z8N5n5/AkXJzNd5v02mQHat8LdeiYzVT/fJZl5/IC7sq4f/fv034m63bJrMvP6Sel/B355/A+SmU6Jf0+ui7rkHhN6vJ7OTPlek2uJiBgy+rxkJqqz/xW365rC5BZ73rM6maldtiJdSzuU8wHCm2++GWeccUasXbs2dt999zjyyCNj8eLFsfvuu+f6KAAAAKCF5HyAcPfdd+d6SwAAACDPmv09EAAAAIC2zwABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASMr5xzjScQyetjyZeea0zlnXRxZvzlU5ALSAV88ty7r+8vhbGrBL9t8bGmr3wuJkZlOPwmSmSy6KAdq0dw7dK5l5/IAfNvmc/zfsnmRmVPmkZGa//3g9mTnhoQuSmQ2TqpKZlEP6vJHM3Nzv/zX5nIiIl//11pzsk/LwhN2SmZunnJHM7PTLZ3JRTqviDgQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAICkonwXQNtVu+bPycylL3866/r/G35nbor55W652QegnSrcrWcys+ySIcnMH878QSKRm3+b6FxQmMxc9/b+yczO9y7ORTlAO9f9jepk5kuv/2uTz5k5YH6T94iI2PLWqmSmSwMyPec1vZbXevRIZj5VdnrTD4qIpefvnsz84TPfb/I5J+/0t2Tmjq++lsy8+8sml9LquAMBAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAICkonwXQMfWqQEzrE5RkMx0+1ttLsoBaHUKRuyfzLw1pjSZ+dIX5iYzc3Z5NJmpS6w/U9M5ucfI4s3JzOZMMhK/W1eeDsVfGpABOrral5YmM385PL3Pgb/L/ufW81aOTu6x6Ns/SGZmTk73v19OODaZiWf/L51JqK2uTocakmmAwRcvT2YOefNrWdfnfPU7yT326rxzg2vqaNyBAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkFSU7wJouzrttFMyc2jvN7Ku10VdQ05qYEUA7U/N9RuSmec+MiuZ6dSAXtqQjnzA7K9mXf/o0UuTe4zc89FkZnXtu8lM9VXlyUxh/CWZAciVFz6aybr+2nVDk3tMGJP+K9pRu/4xmRl+2/8lM89dMiLretHjS5J7tDYl/1qZdf24eRcn97hhzL3JzNrv7ZnMdI81yUxb0+i/mT355JNx0kknRb9+/aKgoCAeeOCBbdYzmUxceeWV0bdv3+jWrVscd9xx8eqrr+aqXgAAACAPGj1A2LhxYwwbNiymT5++3fXvfOc7ccstt8Stt94azzzzTOy0004xduzYeO+995pcLAAAAJAfjX4Jw7hx42LcuHHbXctkMjFt2rS44oor4pOf/GRERPzsZz+LPn36xAMPPBCnn35606oFAAAA8iKn74GwYsWKqKysjOOOO67+udLS0hg5cmQsWrRouwOEmpqaqKmpqf+6uro6lyUB0ML0dYD2Q08HtpbTd6errHz/DSv69OmzzfN9+vSpX/tnFRUVUVpaWv8oL0+/IRIArZe+DtB+6OnA1vL+9vZTpkyJqqqq+sfKlSvzXRIATaCvA7QfejqwtZy+hKGsrCwiItasWRN9+/atf37NmjUxfPjw7X5PcXFxFBcX57IMAPJIXwdoP/R0YGs5vQNh4MCBUVZWFvPnz69/rrq6Op555pkYNWpULo8CAAAAWlCj70DYsGFDLFu2rP7rFStWxPPPPx89e/aMPfbYIy666KK47rrrYu+9946BAwfGt771rejXr1+MHz8+l3XTCnTq1TOZ+W6/+1ugEoD2a/duG3Kyz9u17yYzH/+Pryczvf6Wybr+n2f+dwOqSf9r5lWrtv+JT9vs8vJbycyWBlQD0FIG/cfvk5mqGbskMy8+0D+Z2W+nVcnM12ffknX9ky99PrnHe/f3SWYa4q/Da5OZ3Z8tTGZqE3/92KlnQXKP2TeMSWa6/+mZZKY9avQA4bnnnosxY/5xQSdPnhwRERMmTIhZs2bFZZddFhs3bozzzjsv1q1bF0ceeWTMmzcvunbtmruqAQAAgBbV6AHC6NGjI5P58H99KCgoiGuuuSauueaaJhUGAAAAtB55/xQGAAAAoPUzQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSGv0xjgBAy1lzw17JzL4fn5jMfOTfVyUzu69clMy8++jA7HsUFif3aIjl1++XzHSrfDYnZwG0lLp33slJZsW/lSczzx88PJk575bXsq4/fsA9yT3igHSkId7cUpPMfOWuC5OZzr9fnnW9Idd3y5YtyUxH5Q4EAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACApKJ8FwAAfLiuDz2bzOzzUHqfLQ04640rD09mXhj6/UQi/W8T+947MZkZ/MDiZAago9ry+spkZqcGZPb/ty9mXX/pqNsbXFNTnfPHzyUzXf7f/yYztbkohg/lDgQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJKK8l0A7VunHMyoOhcU5qASgI6t0/CPJDNTPntvMlMXdVnXX9yUSe6x73deS2a2JBMAZLPqssOTmZeO+n7W9VTPz6Wr93owmbnhX05MZra8tSoX5fAh3IEAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQVJTvAmi7Nn6kLJmpi7omn7M50+QtANq1wv2HJDNn3/tIMvPJnd5OZv63Jvu/PXz9kguSe3Rf/Uwyw4erGXdIMrOpR2HW9ZJ7FueqHCDHGvJr/K0x6b/G/c8ZNzbgtK5ZV4/+/RnJHd59tHc6M2pDMvPSx2YmM2tOHJDM7PbjVckMO67RdyA8+eSTcdJJJ0W/fv2ioKAgHnjggW3Wzz777CgoKNjmccIJJ+SqXgAAACAPGj1A2LhxYwwbNiymT5/+oZkTTjghVq9eXf+46667mlQkAAAAkF+NfgnDuHHjYty4cVkzxcXFUVaWvr0dAAAAaBua5T0QFixYEL17945dd901jjnmmLjuuutit9122262pqYmampq6r+urq5ujpIAaCH6OkD7oacDW8v5pzCccMIJ8bOf/Szmz58fN9xwQyxcuDDGjRsXtbW1281XVFREaWlp/aO8vDzXJQHQgvR1gPZDTwe2lvMBwumnnx4nn3xyHHDAATF+/Ph4+OGH47e//W0sWLBgu/kpU6ZEVVVV/WPlypW5LgmAFqSvA7QfejqwtWb/GMdBgwZFr169YtmyZXHsscd+YL24uDiKi4ubuwwAWoi+DtB+6OnA1nJ+B8I/e/PNN2Pt2rXRt2/f5j4KAAAAaCaNvgNhw4YNsWzZsvqvV6xYEc8//3z07NkzevbsGVdffXWceuqpUVZWFsuXL4/LLrssBg8eHGPHjs1p4eTfyn8tzHcJAG1e4eCBWddfubh3co9bx96ezIzp9l4yU5dMRDy+4SNZ19eXp39v2LmkJF3L+vUNqKb16NSAn2n1Fw5IZnY96a1k5r+G3JzM7F6Y/V+MT77nkOQeQO69O/7QZObYq59KZnp1TvfIdQ1o6qN+MTHr+j5XvpTco3T9smSm8L96JjP7VHwlmfn4eb9LZn519EFZ1wef+b/JPfhwjR4gPPfcczFmzJj6rydPnhwRERMmTIgZM2bECy+8ELNnz45169ZFv3794vjjj49rr73WrU8AAADQhjV6gDB69OjIZDIfuv7oo482qSAAAACg9Wn290AAAAAA2j4DBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACCp0R/jCABExPz+yciUPecmM7t0eibr+n5dcjXrz80+l+z2Yvb1b2Rfj4h47qLCZGbl5t2SmavuOz2ZyYUrP31vMrNX5z8nMwcVP56LciKiOJnY77EvZ13fO36Xo1qArRXu1jPr+hPTb03uUReZnNSy772XJDODL16cqCU3atf+NZkZMnFDMvPqvN7JzCtjfpJ1/aDLL0zu8S/XP53MdFTuQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEgqyncBANAWPbrfw8nM5kxtA3Zqmd+KF9UUJjN1mc7JTHlRddb1lVt6NHmPiIgjuqb/jePUL/wgmWkpnQvS13dzJr1PVd17ycykN05KZvrP8Uc8yIc/fXXfrOt18ZvkHnVRl8xcvOpjycw+V77UgLNaj06D9khm5ux7VzKT+pmWXPi95B4nX39IMtNRuQMBAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgyQABAAAASDJAAAAAAJIMEAAAAIAkAwQAAAAgqSjfBdB2dVmXnj91ysGMqnNBYTLzbs90pluTKwH4hxNe+Xgy8+CQB5p8zjH/d1oys/GRsmSmz/efbnItERFx2IHZ1xe/0PQ9ImLNITsnM+/1Sh91xvgF6VAOFBbUJTO3Lzw6mem3MH3WTr94JpnpFs+mNwLarN+u2SOZ6bn+jy1QSe4UbHgnmfnVO7smM+O6/63JtWw5ZkQyU/T4kiaf0xa5AwEAAABIMkAAAAAAkgwQAAAAgCQDBAAAACDJAAEAAABIMkAAAAAAkgwQAAAAgKSifBdA2zVo9spkpu7L6c/FTtmcSWe+e8WMZOb6R45PZmrX/LkhJQFEfH2XZOSYvS5s8jGl89Of473z2j81+ZwGW/xCi+zRZ3HTj4mIePqqLrnZKAf2jmfyXQLQzAbd8kr2wDktU0dblOlWnMx8tLiyATtl3+fKPx+S3KHo8SUNOKdjatQdCBUVFXHIIYdESUlJ9O7dO8aPHx9Lly7dJvPee+/FxIkTY7fddoudd945Tj311FizZk1OiwYAAABaVqMGCAsXLoyJEyfG4sWL4ze/+U1s3rw5jj/++Ni4cWN95uKLL46HHnoo7rvvvli4cGGsWrUqTjnllJwXDgAAALScRr2EYd68edt8PWvWrOjdu3csWbIkjjrqqKiqqoqf/vSnceedd8YxxxwTEREzZ86M/fbbLxYvXhyHHXZY7ioHAAAAWkyT3gOhqqoqIiJ69uwZERFLliyJzZs3x3HHHVef2XfffWOPPfaIRYsWbXeAUFNTEzU1NfVfV1dXN6UkAPJMXwdoP/R0YGs7/CkMdXV1cdFFF8URRxwRQ4cOjYiIysrK6NKlS+yyyy7bZPv06ROVldt/w4uKioooLS2tf5SXl+9oSQC0Avo6QPuhpwNb2+EBwsSJE+PFF1+Mu+++u0kFTJkyJaqqquofK1em39kfgNZLXwdoP/R0YGs79BKGSZMmxcMPPxxPPvlk9O/fv/75srKy2LRpU6xbt26buxDWrFkTZWVl292ruLg4iovTH9kBQNugrwO0H3o6sLVG3YGQyWRi0qRJMWfOnHj88cdj4MCB26yPGDEiOnfuHPPnz69/bunSpfHGG2/EqFGjclMxAAAA0OIadQfCxIkT484774wHH3wwSkpK6t/XoLS0NLp16xalpaXxpS99KSZPnhw9e/aMHj16xIUXXhijRo3yCQw0q5HFm5OZgi5dWqASoKPILHkpmdl5SdPPqW36FgC0kNq1f826/olXPpnc4+F9H0xmnj7ormTmqv89KJlZctAOv6K9UQp365nMfPQXy5KZvoXdkpk1te9mXf/l/PTfS/eKxclMR9WoAcKMGTMiImL06NHbPD9z5sw4++yzIyLi5ptvjk6dOsWpp54aNTU1MXbs2PjhD3+Yk2IBAACA/GjUACGTySQzXbt2jenTp8f06dN3uCgAAACgdWmZe1YAAACANs0AAQAAAEgyQAAAAACSDBAAAACAJAMEAAAAIMkAAQAAAEhq1Mc4wtYy79UkM8/UdM66PrJ4c05quXjVx5KZurV/zclZAACwIwo/VZ3MHHXPZ5KZBQfencxc1XtJMvOxRz6bdb3m17sn94iCdOSULyxIZqb2+r9kZsWW95KZEx64NOv63pcuTu7Bh3MHAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQFJRvgug7apd8+dkZuIPL8i6PmvitOQe36sck8y8edU+yUznd55LZgAAoLnUVlcnM7uemf4r2uj/Oj2ZWXDg3cnM/wzPnqkbXpfcI3fS/7Y9/idfT2b2vvbpXBTDh3AHAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQJIBAgAAAJBkgAAAAAAkGSAAAAAASQYIAAAAQFJRvgugfev3n09nXZ/6n4c2YJfqZKJzPNfAigAAoPWqXfvXZGbX07YkMx+5+sJk5rTR2f+sflXvJck9cmXfeycmM/tM+79kpi4XxfCh3IEAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQZIAAAAAAJBkgAAAAAEkGCAAAAECSAQIAAACQVNSYcEVFRdx///3xyiuvRLdu3eLwww+PG264IYYMGVKfGT16dCxcuHCb7/vyl78ct956a24qBgAA6MBqq6uTmcEXL05mliT+PfnkOKTBNTXV4EjXW9cCdZBdo+5AWLhwYUycODEWL14cv/nNb2Lz5s1x/PHHx8aNG7fJnXvuubF69er6x3e+852cFg0AAAC0rEbdgTBv3rxtvp41a1b07t07lixZEkcddVT98927d4+ysrLcVAgAAADkXZPeA6GqqioiInr27LnN8z//+c+jV69eMXTo0JgyZUq88847H7pHTU1NVFdXb/MAoO3S1wHaDz0d2NoODxDq6urioosuiiOOOCKGDh1a//xnP/vZ+K//+q944oknYsqUKXHHHXfEmWee+aH7VFRURGlpaf2jvLx8R0sCoBXQ1wHaDz0d2FpBJpPJ7Mg3nn/++fGrX/0qnnrqqejfv/+H5h5//PE49thjY9myZbHXXnt9YL2mpiZqamrqv66uro7y8vIYHZ+MooLOO1Ia0AptyWyOBfFgVFVVRY8ePfJdDs1IX4f2T0/vOPR06Bga2tcb9R4Ifzdp0qR4+OGH48knn8w6PIiIGDlyZETEhw4QiouLo7i4eEfKAKAV0tcB2g89HdhaowYImUwmLrzwwpgzZ04sWLAgBg4cmPye559/PiIi+vbtu0MFAgAAAPnXqAHCxIkT484774wHH3wwSkpKorKyMiIiSktLo1u3brF8+fK4884748QTT4zddtstXnjhhbj44ovjqKOOigMPPLBZfgAAAACg+TVqgDBjxoyIiBg9evQ2z8+cOTPOPvvs6NKlSzz22GMxbdq02LhxY5SXl8epp54aV1xxRc4KBgAAAFpeo1/CkE15eXksXLiwSQUBAAAArc8Of4wjAAAA0HEYIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJRfku4J9lMpmIiNgSmyMyeS4GyJktsTki/vFrnI5DX4f2R0/vuPR0aJ8a2tdb3QBh/fr1ERHxVMzNcyVAc1i/fn2UlpbmuwxakL4O7Zee3vHo6dC+pfp6QaaVjY7r6upi1apVUVJSEgUFBRERUV1dHeXl5bFy5cro0aNHnitsf1zf5uX6vi+TycT69eujX79+0amTV091JP/c1/2aaF6ub/Nyfd+np3dc/qze8lzf5uX6vq+hfb3V3YHQqVOn6N+//3bXevTo0aH/ozY317d5ub7hX6k6qA/r635NNC/Xt3m5vnp6R+XP6vnj+jYv17dhfd3IGAAAAEgyQAAAAACS2sQAobi4OK666qooLi7OdyntkuvbvFxf2JZfE83L9W1eri98kF8Xzcv1bV6ub+O0ujdRBAAAAFqfNnEHAgAAAJBfBggAAABAkgECAAAAkGSAAAAAACS1+gHC9OnTY88994yuXbvGyJEj49lnn813SW3Wk08+GSeddFL069cvCgoK4oEHHthmPZPJxJVXXhl9+/aNbt26xXHHHRevvvpqfoptYyoqKuKQQw6JkpKS6N27d4wfPz6WLl26Tea9996LiRMnxm677RY777xznHrqqbFmzZo8VQz5o6/nhp7efPR0aDg9PTf09Oalr+dOqx4g3HPPPTF58uS46qqr4ne/+10MGzYsxo4dG3/+85/zXVqbtHHjxhg2bFhMnz59u+vf+c534pZbbolbb701nnnmmdhpp51i7Nix8d5777VwpW3PwoULY+LEibF48eL4zW9+E5s3b47jjz8+Nm7cWJ+5+OKL46GHHor77rsvFi5cGKtWrYpTTjklj1VDy9PXc0dPbz56OjSMnp47enrz0tdzKNOKHXrooZmJEyfWf11bW5vp169fpqKiIo9VtQ8RkZkzZ07913V1dZmysrLMjTfeWP/cunXrMsXFxZm77rorDxW2bX/+858zEZFZuHBhJpN5/1p27tw5c99999VnXn755UxEZBYtWpSvMqHF6evNQ09vXno6bJ+e3jz09Oanr++4VnsHwqZNm2LJkiVx3HHH1T/XqVOnOO6442LRokV5rKx9WrFiRVRWVm5zvUtLS2PkyJGu9w6oqqqKiIiePXtGRMSSJUti8+bN21zffffdN/bYYw/Xlw5DX285enpu6enwQXp6y9HTc09f33GtdoDw9ttvR21tbfTp02eb5/v06ROVlZV5qqr9+vs1db2brq6uLi666KI44ogjYujQoRHx/vXt0qVL7LLLLttkXV86En295ejpuaOnw/bp6S1HT88tfb1pivJdALQ3EydOjBdffDGeeuqpfJcCQBPp6QDti77eNK32DoRevXpFYWHhB975cs2aNVFWVpanqtqvv19T17tpJk2aFA8//HA88cQT0b9///rny8rKYtOmTbFu3bpt8q4vHYm+3nL09NzQ0+HD6ektR0/PHX296VrtAKFLly4xYsSImD9/fv1zdXV1MX/+/Bg1alQeK2ufBg4cGGVlZdtc7+rq6njmmWdc7wbIZDIxadKkmDNnTjz++OMxcODAbdZHjBgRnTt33ub6Ll26NN544w3Xlw5DX285enrT6OmQpqe3HD296fT13GnVL2GYPHlyTJgwIQ4++OA49NBDY9q0abFx48b4whe+kO/S2qQNGzbEsmXL6r9esWJFPP/889GzZ8/YY4894qKLLorrrrsu9t577xg4cGB861vfin79+sX48ePzV3QbMXHixLjzzjvjwQcfjJKSkvrXSpWWlka3bt2itLQ0vvSlL8XkyZOjZ8+e0aNHj7jwwgtj1KhRcdhhh+W5emg5+nru6OnNR0+HhtHTc0dPb176eg7l+VMgkr7//e9n9thjj0yXLl0yhx56aGbx4sX5LqnNeuKJJzIR8YHHhAkTMpnM+x8R861vfSvTp0+fTHFxcebYY4/NLF26NL9FtxHbu64RkZk5c2Z95t13381ccMEFmV133TXTvXv3zKc+9anM6tWr81c05Im+nht6evPR06Hh9PTc0NObl76eOwWZTCbT7FMKAAAAoE1rte+BAAAAALQeBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACQZIAAAAABJBggAAABAkgECAAAAkGSAAAAAACT9f8JXgVIgYhvVAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "'''\n",
+ "New Block: convert label from one hot encoder. it will randomly plot rows * rows labels with labels.\n",
+ "Description: visualize few testing dataset\n",
+ "'''\n",
+ "plot_images(x_test, np.argmax(y_test, axis=-1), rows=3)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Step 7: Evaluate the ART classifier on adversarial test examples"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 18,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Accuracy on adversarial test examples: 36.05%\n"
+ ]
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: check accuracy of model on adversarial sample.\n",
+ "\"\"\"\n",
+ "adv_predictions = classifier.predict(x_test_adv)\n",
+ "accuracy = np.sum(np.argmax(adv_predictions, axis=1) == np.argmax(y_test, axis=1)) / len(y_test)\n",
+ "print(\"Accuracy on adversarial test examples: {}%\".format(accuracy * 100))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Additional Steps"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 19,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\"\"\"\n",
+ "Description: Select successful adversarial sample. Prediction that is different from true label\n",
+ "\"\"\"\n",
+ "adv_x = []\n",
+ "adx_label = []\n",
+ "for i, img in enumerate(x_test_adv):\n",
+ " if np.argmax(predictions[i],axis=-1)!=np.argmax(adv_predictions[i], axis=-1):\n",
+ " adv_x.append(img)\n",
+ " adx_label.append(np.argmax(adv_predictions[i], axis=-1))\n",
+ " "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 20,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Successful adversarial sample : 6514/10000\n"
+ ]
+ }
+ ],
+ "source": [
+ "\"\"\"\n",
+ "Description: Print to check successfull adversarial sample\n",
+ "\"\"\"\n",
+ "print(\"Successful adversarial sample : {}/{}\".format(len(adv_x), len(x_test)))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 21,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "'''\n",
+ "Description: visualize few advesarial dataset. label are from model on adversarial sample\n",
+ "'''\n",
+ "plot_images(np.array(adv_x), np.array(adx_label), rows=3)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3 (ipykernel)",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.9.0"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}