4-us8k-cnn-salamon.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### Workbook 4 - Implementing Salamon and Bello's Convolutional Neural Network "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is a Keras/Tensorflow implementation of the 5-layer CNN described in Salamon and Bello's paper (https://arxiv.org/pdf/1608.04363.pdf).\n",
    "\n",
    "First, import the libraries. As with the earlier notebooks, the audio processing is handled by a library called librosa, if you haven't already installed it on your local system, do that with: pip install librosa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import glob\n",
    "import os\n",
    "import librosa\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you're going to run this code with the full data set, this notebook assumes you've already run the feature extraction code in notebook 3. \n",
    "\n",
    "If you have a powerful computer, and don't mind the training process taking longer, you can set the all_folds flag to true - which will create a single large training set of 43722 examples, using the first 8 folds. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data_dir = \"data/us8k-np-cnn\"\n",
    "\n",
    "# this will aggregate all the training data \n",
    "def load_all_folds(): \n",
    "    subsequent_fold = False\n",
    "    for k in range(1,9):\n",
    "        fold_name = 'fold' + str(k)\n",
    "        print \"\\nAdding \" + fold_name\n",
    "        feature_file = os.path.join(data_dir, fold_name + '_x.npy')\n",
    "        labels_file = os.path.join(data_dir, fold_name + '_y.npy')\n",
    "        loaded_features = np.load(feature_file)\n",
    "        loaded_labels = np.load(labels_file)\n",
    "        print \"New Features: \", loaded_features.shape\n",
    "\n",
    "        if subsequent_fold:\n",
    "            train_x = np.concatenate((features, loaded_features))\n",
    "            train_y = np.concatenate((labels, loaded_labels))\n",
    "        else:\n",
    "            train_x = loaded_features\n",
    "            train_y = loaded_labels\n",
    "            subsequent_fold = True\n",
    "        \n",
    "    # use the penultimate fold for validation\n",
    "    valid_fold_name = 'fold9'\n",
    "    feature_file = os.path.join(data_dir, valid_fold_name + '_x.npy')\n",
    "    labels_file = os.path.join(data_dir, valid_fold_name + '_y.npy')\n",
    "    valid_x = np.load(feature_file)\n",
    "    valid_y = np.load(labels_file) \n",
    "\n",
    "    # and use the last fold for testing\n",
    "    test_fold_name = 'fold10'\n",
    "    feature_file = os.path.join(data_dir, test_fold_name + '_x.npy')\n",
    "    labels_file = os.path.join(data_dir, test_fold_name + '_y.npy')\n",
    "    test_x = np.load(feature_file)\n",
    "    test_y = np.load(labels_file)\n",
    "    \n",
    "\n",
    "# this is used to load the folds incrementally\n",
    "def load_folds(folds):\n",
    "    subsequent_fold = False\n",
    "    for k in range(len(folds)):\n",
    "        fold_name = 'fold' + str(folds[k])\n",
    "        feature_file = os.path.join(data_dir, fold_name + '_x.npy')\n",
    "        labels_file = os.path.join(data_dir, fold_name + '_y.npy')\n",
    "        loaded_features = np.load(feature_file)\n",
    "        loaded_labels = np.load(labels_file)\n",
    "        print fold_name, \"features: \", loaded_features.shape\n",
    "\n",
    "        if subsequent_fold:\n",
    "            features = np.concatenate((features, loaded_features))\n",
    "            labels = np.concatenate((labels, loaded_labels))\n",
    "        else:\n",
    "            features = loaded_features\n",
    "            labels = loaded_labels\n",
    "            subsequent_fold = True\n",
    "        \n",
    "    return features, labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training a Convolutional Neural Network with Keras and TensorFlow"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, the imports we need and a few configuration variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tf.set_random_seed(0)\n",
    "np.random.seed(0)\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, MaxPooling2D\n",
    "from keras.optimizers import SGD, Adam, Adagrad\n",
    "from keras.callbacks import EarlyStopping\n",
    "from sklearn.metrics import precision_recall_fscore_support, roc_auc_score\n",
    "from keras.utils import np_utils\n",
    "from keras.regularizers import l2, activity_l2\n",
    "\n",
    "frames = 41\n",
    "bands = 60\n",
    "feature_size = bands * frames #60x41\n",
    "num_channels = 2\n",
    "num_labels = 10 # test_y.shape[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This method defines some evaluation metrics that will be used to evaluate the performance of a trained model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def evaluate(model):\n",
    "    y_prob = model.predict_proba(test_x, verbose=0)\n",
    "    y_pred = np_utils.probas_to_classes(y_prob)\n",
    "    y_true = np.argmax(test_y, 1)\n",
    "\n",
    "    roc = roc_auc_score(test_y, y_prob)\n",
    "    print \"ROC:\",  round(roc,3)\n",
    "\n",
    "    # evaluate the model\n",
    "    score, accuracy = model.evaluate(test_x, test_y, batch_size=32)\n",
    "    print(\"\\nAccuracy = {:.2f}\".format(accuracy))\n",
    "\n",
    "    # the F-score gives a similiar value to the accuracy score, but useful for cross-checking\n",
    "    p,r,f,s = precision_recall_fscore_support(y_true, y_pred, average='micro')\n",
    "    print \"F-Score:\", round(f,2)\n",
    "    \n",
    "    return roc, accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This method contains the code that defines the successive layers of our convolutional neural network (CNN) described in this paper by Salamon and Bello: https://arxiv.org/pdf/1608.04363.pdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def build_model():\n",
    "    \n",
    "    model = Sequential()\n",
    "    # input: 60x41 data frames with 2 channels => (60,41,2) tensors\n",
    "\n",
    "    # filters of size 3x3 - paper describes using 5x5, but their input data is 128x128\n",
    "    f_size = 3\n",
    "\n",
    "    # Layer 1 - 24 filters with a receptive field of (f,f), i.e. W has the\n",
    "    # shape (24,1,f,f).  This is followed by (4,2) max-pooling over the last\n",
    "    # two dimensions and a ReLU activation function\n",
    "    model.add(Convolution2D(24, f_size, f_size, border_mode='same', init=\"normal\", input_shape=(bands, frames, num_channels)))\n",
    "    model.add(MaxPooling2D(pool_size=(4, 2)))\n",
    "    model.add(Activation('relu'))\n",
    "\n",
    "    # Layer 2 - 48 filters with a receptive field of (f,f), i.e. W has the \n",
    "    # shape (48, 24, f, f). Like L1 this is followed by (4,2) max-pooling \n",
    "    # and a ReLU activation function.\n",
    "    model.add(Convolution2D(48, f_size, f_size, init=\"normal\", border_mode='same'))\n",
    "    model.add(MaxPooling2D(pool_size=(4, 2)))\n",
    "    model.add(Activation('relu'))\n",
    "\n",
    "    # Layer 3 - 48 filters with a receptive field of (f,f), i.e. W has the\n",
    "    # shape (48, 48, f, f). This is followed by a ReLU but no pooling.\n",
    "    model.add(Convolution2D(48, f_size, f_size, border_mode='valid'))\n",
    "    model.add(Activation('relu'))\n",
    "\n",
    "    # flatten output into a single dimension, let Keras do shape inference\n",
    "    model.add(Flatten())\n",
    "\n",
    "    # Layer 4 - a fully connected NN layer of 64 hidden units, L2 penalty of 0.001\n",
    "    model.add(Dense(64, W_regularizer=l2(0.001)))\n",
    "    model.add(Activation('relu'))\n",
    "    model.add(Dropout(0.5))\n",
    "\n",
    "    # Layer 5 - an output layer with one output unit per class, with L2 penalty, \n",
    "    # followed by a softmax activation function\n",
    "    model.add(Dense(num_labels, W_regularizer=l2(0.001)))\n",
    "    model.add(Dropout(0.5))\n",
    "    model.add(Activation('softmax'))\n",
    "\n",
    "    # compile and fit model, reduce epochs if you want a result faster\n",
    "    # the validation set is used to identify parameter settings (epoch) that achieves \n",
    "    # the highest classification accuracy\n",
    "    model.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=\"adamax\")\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training on a minimised data set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data set is a large quantity of data, which is quite computationally intensive to process. But if you've checked out the git repository for this project, I've included a selection from the original data set, which should allow you to run the CNN and see how it works. Obviously as this training set size is much smaller and contains much less diversity than the full set, lower accuracies will be achievable when models are trained on it, but it will allow you to run the code and explore how it works. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold1 features:  (5446, 60, 41, 2)\n",
      "fold2 features:  (5388, 60, 41, 2)\n",
      "fold3 features:  (5852, 60, 41, 2)\n",
      "fold9 features:  (5118, 60, 41, 2)\n",
      "fold10 features:  (5218, 60, 41, 2)\n",
      "Building model...\n",
      "Training model...\n",
      "Train on 16686 samples, validate on 5118 samples\n",
      "Epoch 1/12\n",
      "16686/16686 [==============================] - 64s - loss: 2.1939 - acc: 0.2140 - val_loss: 2.0159 - val_acc: 0.2532\n",
      "Epoch 2/12\n",
      "16686/16686 [==============================] - 64s - loss: 1.9663 - acc: 0.3076 - val_loss: 1.9394 - val_acc: 0.3214\n",
      "Epoch 3/12\n",
      "16686/16686 [==============================] - 64s - loss: 1.7583 - acc: 0.3674 - val_loss: 1.8850 - val_acc: 0.3376\n",
      "Epoch 4/12\n",
      "16686/16686 [==============================] - 66s - loss: 1.6103 - acc: 0.4147 - val_loss: 1.8866 - val_acc: 0.3912\n",
      "Epoch 5/12\n",
      "16686/16686 [==============================] - 64s - loss: 1.4991 - acc: 0.4506 - val_loss: 1.8545 - val_acc: 0.4017\n",
      "Epoch 6/12\n",
      "16686/16686 [==============================] - 64s - loss: 1.4288 - acc: 0.4700 - val_loss: 1.7745 - val_acc: 0.4388\n",
      "Epoch 7/12\n",
      "16686/16686 [==============================] - 64s - loss: 1.3687 - acc: 0.4893 - val_loss: 1.9818 - val_acc: 0.3974\n",
      "Epoch 8/12\n",
      "16686/16686 [==============================] - 63s - loss: 1.3354 - acc: 0.4960 - val_loss: 2.2752 - val_acc: 0.4125\n",
      "Epoch 9/12\n",
      "16686/16686 [==============================] - 66s - loss: 1.2968 - acc: 0.5032 - val_loss: 1.9761 - val_acc: 0.4902\n",
      "Epoch 00008: early stopping\n",
      "Evaluating model...\n",
      "ROC: 0.863\n",
      "5216/5218 [============================>.] - ETA: 0s\n",
      "Accuracy = 0.54\n",
      "F-Score: 0.54\n"
     ]
    }
   ],
   "source": [
    "# change this to data/us8k-np-cnn if you have the full dataset\n",
    "data_dir = \"data/us8k-np-cnn-mini\"\n",
    "\n",
    "# load fold1 for testing (or 1,2,3 if you have full data)\n",
    "train_x, train_y = load_folds([1])\n",
    "\n",
    "# load fold2 for validation (or 9 if full data)\n",
    "valid_x, valid_y = load_folds([2])\n",
    "    \n",
    "# load fold3 for testing (or 10 if full data)\n",
    "test_x, test_y = load_folds([3])\n",
    "\n",
    "print(\"Building model...\")\n",
    "model = build_model()\n",
    "\n",
    "# a stopping function to stop training before we excessively overfit to the training set\n",
    "earlystop = EarlyStopping(monitor='val_loss', patience=2, verbose=1, mode='auto')\n",
    "\n",
    "# now fit the model to the training data, evaluating loss against the validation data\n",
    "print(\"Training model...\")\n",
    "history = model.fit(train_x, train_y, validation_data=(valid_x, valid_y), callbacks=[earlystop], batch_size=20, nb_epoch=12)\n",
    "\n",
    "# now evaluate the trained model against the unseen test data\n",
    "print(\"Evaluating model...\")\n",
    "roc, acc = evaluate(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training on the full data set (intensive) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the training process, it will create a fresh model for each trial, load the next fold of input data, fit it to the model and evaluate the accuracy achieved. When all the input data has been used, we can calculate the average accuracy for all the folds. The parameters used here have been chosen to allow computation to run on a reasonably powered laptop in less than 10 minutes. When running it on an 8-core high-memory cloud instance, more epochs and larger training sets can be used, this will produce greater prediction accuracy, but come at the cost of much greater training time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Train on (1, 2, 3) Validate on 4 Test on 5 ***\n",
      "fold1 features:  (5446, 60, 41, 2)\n",
      "fold2 features:  (5388, 60, 41, 2)\n",
      "fold3 features:  (5852, 60, 41, 2)\n",
      "fold4 features:  (6048, 60, 41, 2)\n",
      "fold5 features:  (5689, 60, 41, 2)\n",
      "Building model...\n",
      "Training model...\n",
      "Train on 16686 samples, validate on 6048 samples\n",
      "Epoch 1/2\n",
      "16686/16686 [==============================] - 74s - loss: 2.1965 - acc: 0.2171 - val_loss: 1.9184 - val_acc: 0.3054\n",
      "Epoch 2/2\n",
      "16686/16686 [==============================] - 67s - loss: 1.9443 - acc: 0.3093 - val_loss: 1.7758 - val_acc: 0.4347\n",
      "Evaluating model...\n",
      "ROC: 0.798\n",
      "5689/5689 [==============================] - 9s     \n",
      "\n",
      "Accuracy = 0.34\n",
      "F-Score: 0.34\n",
      "\n",
      "*** Train on (2, 3, 4) Validate on 5 Test on 6 ***\n",
      "fold2 features:  (5388, 60, 41, 2)\n",
      "fold3 features:  (5852, 60, 41, 2)\n",
      "fold4 features:  (6048, 60, 41, 2)\n",
      "fold5 features:  (5689, 60, 41, 2)\n",
      "fold6 features:  (5080, 60, 41, 2)\n",
      "Building model...\n",
      "Training model...\n",
      "Train on 17288 samples, validate on 5689 samples\n",
      "Epoch 1/2\n",
      "17288/17288 [==============================] - 71s - loss: 2.2377 - acc: 0.1872 - val_loss: 2.1070 - val_acc: 0.1317\n",
      "Epoch 2/2\n",
      "17288/17288 [==============================] - 74s - loss: 2.0207 - acc: 0.2793 - val_loss: 1.9992 - val_acc: 0.1932\n",
      "Evaluating model...\n",
      "ROC: 0.768\n",
      "5080/5080 [==============================] - 6s     \n",
      "\n",
      "Accuracy = 0.27\n",
      "F-Score: 0.27\n",
      "\n",
      "*** Train on (3, 4, 5) Validate on 6 Test on 7 ***\n",
      "fold3 features:  (5852, 60, 41, 2)\n",
      "fold4 features:  (6048, 60, 41, 2)\n",
      "fold5 features:  (5689, 60, 41, 2)\n",
      "fold6 features:  (5080, 60, 41, 2)\n",
      "fold7 features:  (5277, 60, 41, 2)\n",
      "Building model...\n",
      "Training model...\n",
      "Train on 17589 samples, validate on 5080 samples\n",
      "Epoch 1/2\n",
      "17589/17589 [==============================] - 75s - loss: 2.1972 - acc: 0.2083 - val_loss: 2.0948 - val_acc: 0.2248\n",
      "Epoch 2/2\n",
      "17589/17589 [==============================] - 82s - loss: 1.9407 - acc: 0.3164 - val_loss: 1.8980 - val_acc: 0.3421\n",
      "Evaluating model...\n",
      "ROC: 0.769\n",
      "5277/5277 [==============================] - 7s     \n",
      "\n",
      "Accuracy = 0.34\n",
      "F-Score: 0.34\n",
      "\n",
      "Average R.O.C: 0.778\n",
      "Average Accuracy: 0.314\n"
     ]
    }
   ],
   "source": [
    "data_dir = \"data/us8k-np-cnn\"\n",
    "all_folds = False\n",
    "av_acc = 0.\n",
    "av_roc = 0.\n",
    "num_folds = 0\n",
    "\n",
    "# as we use two folds for training, there are 9 possible trails rather than 10\n",
    "max_trials = 3\n",
    "\n",
    "# earlystopping ends training when the validation loss stops improving\n",
    "earlystop = EarlyStopping(monitor='val_loss', patience=0, verbose=0, mode='auto')\n",
    "\n",
    "if all_folds:\n",
    "    load_all_folds()\n",
    "    model.fit(train_x, train_y, validation_data=(valid_x, valid_y), callbacks=[earlystop], batch_size=32, nb_epoch=1)\n",
    "else:\n",
    "    # use folds incrementally\n",
    "    for f in range(1,max_trials+1):\n",
    "        num_folds += 1\n",
    "        v = f + 3\n",
    "        if v > 10: v = 1\n",
    "        t = v + 1\n",
    "        if t > 10: t = 1\n",
    "        \n",
    "        print \"\\n*** Train on\", (f,f+1,f+2), \"Validate on\", v, \"Test on\", t, \"***\"\n",
    "    \n",
    "        # load two folds for training data\n",
    "        train_x, train_y = load_folds([f,f+1,f+2])\n",
    "    \n",
    "        # load one fold for validation\n",
    "        valid_x, valid_y = load_folds([v])\n",
    "    \n",
    "        # load one fold for testing\n",
    "        test_x, test_y = load_folds([t])\n",
    "        \n",
    "        num_labels = test_y.shape[1]\n",
    "        \n",
    "        print(\"Building model...\")\n",
    "        model = build_model()\n",
    "\n",
    "        # now fit the model to the training data, evaluating loss against the validation data\n",
    "        print(\"Training model...\")\n",
    "        model.fit(train_x, train_y, validation_data=(valid_x, valid_y), callbacks=[earlystop], batch_size=24, nb_epoch=2)\n",
    "        \n",
    "        \n",
    "        # now evaluate the trained model against the unseen test data\n",
    "        print(\"Evaluating model...\")\n",
    "        roc, acc = evaluate(model)\n",
    "        av_roc += roc\n",
    "        av_acc += acc\n",
    "    \n",
    "print '\\nAverage R.O.C:', round(av_roc/max_trials, 3)\n",
    "print 'Average Accuracy:', round(av_acc/max_trials, 3)\n",
    "\n",
    "# using all folds: best ROC = 0.91, f-score = 0.592 (50 epochs)\n",
    "# using 2 folds: average ROC = 0.792, average f-score = 0.335\n",
    "\n",
    "# if you want to save the model, uncomment this...\n",
    "#filepath = \"models/salamon-cnn-model.h5\"\n",
    "#model.save(filepath)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluating Predictions using a Pre-Trained Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To save time, once trained, a model can be saved and reloaded to generate predictions. The following cell also shows a summary of the CNN we've created."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A pretrained model has been supplied, enabling us to run predictions without creating and training a model from scratch, if you want to use it, run the following cell:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                     Output Shape          Param #     Connected to                     \n",
      "====================================================================================================\n",
      "convolution2d_14 (Convolution2D) (None, 60, 41, 24)    456         convolution2d_input_2[0][0]      \n",
      "____________________________________________________________________________________________________\n",
      "maxpooling2d_9 (MaxPooling2D)    (None, 15, 20, 24)    0           convolution2d_14[0][0]           \n",
      "____________________________________________________________________________________________________\n",
      "activation_20 (Activation)       (None, 15, 20, 24)    0           maxpooling2d_9[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "convolution2d_15 (Convolution2D) (None, 15, 20, 48)    10416       activation_20[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "maxpooling2d_10 (MaxPooling2D)   (None, 3, 10, 48)     0           convolution2d_15[0][0]           \n",
      "____________________________________________________________________________________________________\n",
      "activation_21 (Activation)       (None, 3, 10, 48)     0           maxpooling2d_10[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "convolution2d_16 (Convolution2D) (None, 1, 8, 48)      20784       activation_21[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "activation_22 (Activation)       (None, 1, 8, 48)      0           convolution2d_16[0][0]           \n",
      "____________________________________________________________________________________________________\n",
      "flatten_5 (Flatten)              (None, 384)           0           activation_22[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "dense_9 (Dense)                  (None, 64)            24640       flatten_5[0][0]                  \n",
      "____________________________________________________________________________________________________\n",
      "activation_23 (Activation)       (None, 64)            0           dense_9[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "dropout_10 (Dropout)             (None, 64)            0           activation_23[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "dense_10 (Dense)                 (None, 10)            650         dropout_10[0][0]                 \n",
      "____________________________________________________________________________________________________\n",
      "dropout_11 (Dropout)             (None, 10)            0           dense_10[0][0]                   \n",
      "____________________________________________________________________________________________________\n",
      "activation_24 (Activation)       (None, 10)            0           dropout_11[0][0]                 \n",
      "====================================================================================================\n",
      "Total params: 56946\n",
      "____________________________________________________________________________________________________\n",
      "None\n",
      "fold10 features:  (5218, 60, 41, 2)\n",
      "ROC: 0.91\n",
      "5216/5218 [============================>.] - ETA: 0s\n",
      "Accuracy = 0.59\n",
      "F-Score: 0.59\n",
      "\n",
      "Model R.O.C: 0.91\n",
      "Model Accuracy: 0.592\n"
     ]
    }
   ],
   "source": [
    "from keras.models import load_model\n",
    "from keras.utils.visualize_util import plot\n",
    "\n",
    "model_filepath = \"models/salamon-cnn-model.h5\"\n",
    "model = load_model(model_filepath)\n",
    "\n",
    "plot(model, show_shapes=True, to_file='models/salamon-cnn-model.svg')\n",
    "print model.summary()\n",
    "\n",
    "data_dir = \"data/us8k-np-cnn\"\n",
    "# model has never seen fold 10, so use that for testing\n",
    "test_x, test_y = load_folds([10])\n",
    "roc, acc = evaluate(model)\n",
    "print '\\nModel R.O.C:', round(roc, 3)\n",
    "print 'Model Accuracy:', round(acc, 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using micro-averaging we can create an ROC measure for this multi-class classification problem, by converting the predictions into boolean values, each representing positive or negative predictions for each possible class. The resulting ROC plot is shown below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEcCAYAAADdtCNzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11cefe690>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve, auc\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.style.use('ggplot')\n",
    "\n",
    "# obtain the prediction probabilities\n",
    "y_prob = model.predict_proba(test_x, verbose=0)\n",
    "y_pred = np_utils.probas_to_classes(y_prob)\n",
    "y_true = np.argmax(test_y, 1)\n",
    "\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(num_labels):\n",
    "    fpr[i], tpr[i], _ = roc_curve(test_y[:, i], y_prob[:, i])\n",
    "    roc_auc[i] = auc(fpr[i], tpr[i])\n",
    "\n",
    "# Compute micro-average ROC curve and ROC area\n",
    "fpr[\"micro\"], tpr[\"micro\"], _ = roc_curve(test_y.ravel(), y_prob.ravel())\n",
    "roc_auc[\"micro\"] = auc(fpr[\"micro\"], tpr[\"micro\"])\n",
    "\n",
    "plt.figure()\n",
    "lw = 2\n",
    "plt.plot(fpr[2], tpr[2], color='darkorange', lw=lw, label='ROC curve (area = %0.2f)' % roc_auc[2])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This saved model achieves an accuracy of 59%, and represents some of the best performance I've achieved with this particular neural network architecture."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Showing Confusion Matrix\n",
      "              air-con     horn children      dog    drill   engine      gun   hammer    siren    music\n",
      "     air-con      514        0        9        0        1       40        0       70        8       58\n",
      "        horn        6       65       12        3        2        5        0        3        1       38\n",
      "    children        9        2      442      102       14        3        1        0       43       76\n",
      "         dog       13        5      101      279       12       16        0        0       27       62\n",
      "       drill       31        3       26        4      290       43        0       67       66       93\n",
      "      engine       43        0      157        0       15      366        0        9        0       50\n",
      "         gun        0        0        5       18        3        1       45        0        0        0\n",
      "      hammer        6        1       16        0      149        6        0      399        0       11\n",
      "       siren       53       11      164       40       14       11        0        0      256        4\n",
      "       music       43       11      102        0       65        2        0       29       14      434\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12febd8d0>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0MAAAHcCAYAAAAUb3FQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XlcVGX7x/HPsAyI7IsCgoIIopn7VpoL5VaZlampqWnm\nVqm5a+67uO+iaaVptme2uJuZT7mLCoILoKwiIIICzgzM7w8Tndym5/cMZ2Cud6951dzncOY705kz\nc819n/uo9Hq9HiGEEEIIIYSwMFZKBxBCCCGEEEIIJUgxJIQQQgghhLBIUgwJIYQQQgghLJIUQ0II\nIYQQQgiLJMWQEEIIIYQQwiLZKB1ACCGEEEIIYf5qV2lpku2evnzAJNs1hsmLIVO9aKXd6csH0ORk\nKh3DLKmdPQAoyExTOIl5svfwln3nEe7uO/L6PJza2YPc+BilY5glp8BQ2W8e4e77KuvUEYWTmCf3\nuo25ff2q0jHMkp1bRQCyo08pnMQ8udasq3QEgQyTE0IIIYQQQlgoGSYnhBBCCCGEeCKVSqV0hP85\nKYaEEEIIIYQQT6RSlb1BZWXvGQkhhBBCCCGEEaQYEkIIIYQQQlgkKYaEEEIIIYQQFknOGRJCCCGE\nEEI8kRUygYIQQgghhBBClKixY8fi4OAAQIUKFXjttddYtWoVKpUKf39/+vfvD8CePXvYu3cv1tbW\nvP7669SvX/+x2zW6GLpx4wZarbb4vqen53/zPIQQQgghhBClkFJTa9+tQaZMmVLcFh4eTvfu3alR\nowbr1q3j6NGjBAcHs2PHDubNm8ft27eZPHkytWvXxsbm0SWPUcXQxx9/zMmTJ3Fzc0Ov16NSqZg5\nc+b/82kJIYQQQgghSgsrhabWvnz5Mrdv32bWrFkUFRXx5ptvEh8fT40aNQCoV68ekZGRqFQqQkND\nsba2xsHBAW9vb65cuULVqlUfuW2jiqGLFy+yfPlyrKxkvgUhhBBCCCFEyVGr1bzyyiuEhYWRmprK\n7Nmz0ev1xcvt7e3Jz8+noKCgeCjd3fa8vLzHbtuoYsjb2xutVoudnd1/+RSEEEIIIYQQpZlSw+R8\nfX3x9vYGwMfHBycnJ+Lj44uX3y2CypUrZ1D8/LM4ehijiqGMjAyGDBlSHEKGyQkhhBBCCCFKwr59\n+7hy5Qr9+/cnKyuL/Px8ateuTXR0NDVr1uTkyZPUqlWLoKAgtm7dik6nQ6PRkJycTOXKlR+7baOK\noWHDhv1PnogQQgghhBBC/BthYWGsWrWKyZMno1KpGDJkCE5OTqxZs4bCwkIqVapE06ZNUalUdOjQ\ngUmTJgHQvXv3x06eAEYWQ1ZWVnz22WckJSXh4+NDnz59/v/PSgghhBBCCCGewMbGhqFDhz7QPnXq\n1AfawsLCCAsLM37bxqwUERFB27ZtqVGjBlFRUaxZs4bJkycb/SCm1r5jGHOXTSqe6U6v17N/9yFG\nDJxUvI5/lUp8s3MDTWu0NzjhymD5jvUM6TOG40dOl2R8RWm1WmaFL2T3vv3Yqe3o1aMbfXv1VDqW\n2ci+cYPZC5bw19FjlC/vQM+ub/BWty5KxzILsu88mUajoVvvfowbOZwmjRoqHUdxM5esICk1jTXz\n7gyzTku/xowlKzgdHYNPBS+GD+jHsw0ffz2Isk7eV/f8fOAgs1avQwXoweDf369cQpG+iDkR6zkT\newFvL08+6NWDZ+rWVjSzknJyc5mzYAl//HkYe3s7Xm7flqGDByh2jofSdh08xOTFy1GpQK8HlQpa\nNGrIvHGjuHT5CgvWfULMpTgqeLjTr2tn2rVornTkUkFlqRdd1Wq1NGx454O8cePG/PzzzyYN9W8F\nhQSwd8fvzPxoUfGb/vZtTfHyij5erNgwB7Xa9pHbmDpvNGo7tcmzmpsFS5dzJiqa9auWk3Y1nfGT\np+Hj7U37Ns8rHc0sDB/3EZrbGiKWLuRWXh4Tp8/GysqKHl06Kx1NcbLvPJ5Go2HMR1OIi09QOopZ\nOHIykm0799Cgdq3ithFTZxFUpTKbli/ktz8PM2bGXL5euwKfihUUTKoseV/d0+bZpgbFjV6vZ+Tc\nhfh5V6SChzu9x3xEVX8/NsyZzsFjJ5iwaClbFs7Dx8syr4M4M3wRmVlZfBaxgqzr2YydPA1XFxf6\n9HxT6WiKiEtMolXTxowd2L/4R3C12hatVseoOfNp0bghkz4YzPGzUUxfvgp/H29qBldTOLX5U2pq\nbVMy6hkVFhZy5coVgOJ/m5Og4ADOn4vjetYNsjKzycrM5tbNOzNJtG7bnK3b11JwX3H0T13f6mSR\n04bnFxTw3Q/bGTtiGKHVQ2jVojl9e/fki6+/UTqaWTgXe57IM1HMnT6ZGtVDaFivLsPfG8SGTVuU\njqY42XceLy4+gZ593yU5JUXpKGahoOA2s5etou5TNYrbjp46zZXkFD4a9h4B/n683bUztWuGsm3n\nbgWTKkveV4bUtra4u7gU3/b9dZT0rCzGDejHsbNRJKalMW5APwIq+dKr08s8HRLM9v0HlI6tmEN/\nHuatbl2oGhhAw/p16dD2BY4cP6F0LMXEJyZRrUpl3FyccXd1wd3VBUcHB+KTkki7do133+yCb8UK\ndHy+NdWqVOH42WilIwuFGFUB9O3bl9WrVzNo0CBWr15N3759TZ3rX6kaHEBC3MOLtOdaN2X5go8J\nn7b8ocsr+ngxaFgfpo2bb3FdybHnL6DV6ahX594vb/Xq1CEq+txDhxJamqTkFJydnKjs51fcVr1a\nEJlZWaSmXVUwmfJk33m8YydO0qRRQz7fsFZeD2Dlp5toWOdp6j99r1fobMx5qgdVxd7+3iUb6j5V\nk9PnYpWIaBbkffVoeQUFfPLt9wzo2hlHBweiLlwiJCAA+/su+VGneghnz19QMKWyXFyc+WXnbgoK\nbpN+LYNDfx2hZo1QpWMpJiExiSqVfB9od3Z0BGD7nv3o9XrOxJznSnIKNYIefVFOcY9KpTLJTUlG\nDZPz8/NjwIABBAYGcuTIEfzu+3KoNBsba/wr+9LyhWd5f1R/VCrY9fNvrFr8CTqtjunjFwDQoEmd\nh/79pNkj2fTx11xJSC7J2GYhIyMTF2cnbG3vDR/08HBDq9WRmXUdTw93BdMpz8PdjZu3bpGXn49D\nuXIApKSlAXfOJfLxrqhkPEXJvvN4XTu/pnQEs3E6OoZ9f/yHLyOWs+mbH4rbM7Ky8PrHfuLu5kJ6\nRmZJRzQb8r56tO9370Vtq6ZjWCsAMrOz8XRzNVjHzdWF9MzrJR/OTHw0egQTps3kmefbU1RURJOG\nDRj8zttKx1KETqcj6epV/jh6nDVbvgS9nrBnmzKwe1e8vTwZ1ONNVm7awvKNm9Hri+jXpTMN7xvC\nKyyLUT1Dy5YtIyEhAYDU1FRWrlxpykz/SuVAP6ysrci7mc+IgZNYOGs1L73ahlET33vi33Z8vR1e\nFTz4JOKLEkhqfvILCrC1NTxPSv33fa320cMKLcXTT9WkYgUvZoYvJC8vj6vXrrFm/afAnfPoLJns\nO8IYWq2WmUtWMHLQuziWL2+wrOD2bYMv/XBnWJQlv7fkffVo2/b+RpcObbD+e0h7wW0N6n/uPza2\naHWWu/9cSUykRvUQPo1YwarF80lJTWXBMvP5vlaSrqSmUVRUhEO5cswbO5Khb/di5+9/sOSTjRQW\nFpGYmkanNs/zSfgsxg8ewJYff+bA4aNKxy4VrFQqk9yUZFTPUFZWFq1btwagU6dOTJs2zaSh/o24\nC5dpUfcVcnNuAnAhNg4rKxVzl05i3tRljxxa4O7hyogJgxjy9hhAuSvqKsnOTv3AB6zm7/v29vZK\nRDIrtra2LJo9g7GTp9O83cs4OToybPAAzp6Lofw/vthZGtl3hDHWbt5KZT9fwpo/88AytVrNrbx8\ngzaNVovdfcOeLI28rx4uJi6elKvptH+uWXGbWm3Lrfx/7D86LfZqy5sICe4M656/ZAU7t32Nl+ed\nCSSmjB/DwKEj6N+nF+7/6EUr66r6+7Hzs4+Lh8RVq1KZoqIiJi1aRnBAFc7ExPLlisUAVK8ayNWM\nTNZ+8RUtmzRSMrZQiFHFkEqlIiUlBV9fX9LS7lTb5uRuIXRX3MUr2Nja4ObhSlbGw7vMm7Vsgoub\nM+u3LjUohFZ9Fk7Eso1sWF32T5Kv4OVFTk4uOp2u+IJUmZlZqG1tcXF2VjideQgNCWbb1k1kXc/G\n2cmRK0lJWKlUFj3bFci+I4yz67eDZFy/TovXugGg1ekoKiyi5etv0rfbG1z8x0x7mVnZeLq7KZDU\nPMj76uH+OnWamtWC8HC994Xey82NS5cTDdbLyr6Bh4V96b8rOiYWJ2en4kIIoGZodQqLikhNS7O4\nYgjunRt0V4BfJXSFhZyJvUBQlcoGy0KDqrJ52/aSjCfMiFHF0Ntvv82SJUu4ceMGbm5uDBgwwNS5\njBbW7jkmzvyQNk27UFhYCECNWsHk5tx8ZCEEsOfXA5w4eu96QtbWVvy4/3OmjAnn0IEjJs9tDkJD\nQrC1teXU6TM0rF8PgOMnI6lZI9QiZ9f7p5zcXIaOGc/iOTOLP0j2/36I0OohODg4KJxOWbLvCGNE\nzJ+FTldYfH/zd9uIuXCJmeNGknr1Kp98+Q0Ft28XnwQfGRXN0xZ8wre8rx7u7IWL1KtpuF/UCq7G\nxm3bKdBoinuDImPOUyvEMqdG9vL0JDcnl8zMLDz+PrfsUnwCKpUKP98HJxEo63776wjzIj5m+8er\nsbG2BiA2Lh6n8uWpVLECuw4eMlg/PjGJShZ8HvC/oTLuDJtSxahnVK1aNcLDwxk7dixz586lalXz\nmXHj2F+nAJg8dxSVAyrxXFhTPhw/iE/WPP48oPz8ApITU4tvSVdSAbh2NeOBnqayyt7ejo4vdWDW\nvAWcjYpm/4GDbNz8BW9176p0NLPg7OREQcFtFi5fRVJyCrv27WfdZ5sY2LeP0tEUJ/uOMIa3lxd+\nPt7FN2dHR+zs1FTyrkj9p2vhXaECUxcsJe7yFT796lvOxp7ntQ5tlY6tGHlfPVxcYhJV/SoZtNWr\nGYq3pyczVkYQn5TMpm0/EX3xEp2eb6VMSIXVrlWTakFVmTBtJucvXiLybBQz5i2gY4d2uLhYXq9i\nvadqAjBn1VqupKTyx7ETrNi4mV6vvUL7ls+RlpHJ0k82kpx2lf1/HmbT99vo0fFlhVMLpfyr8m7T\npk2myvFfy7mRy6Deo/GtVJGtP61j0qyRfLXphycWQw9jiVOXjh4+lKdq1qD/e0OZGb6Awe++Q5uw\n1krHMhvzZ0zlavo1uvTpx6p1G5gybjQtmj14/oMlkn3HOJZ4PqIxrKysWDhlAtdv3KDX0JH8uu8A\nCyZPwLuCl9LRFCXvqwddz8l5YMiTlZUV4aM/JDsnl77jJ7Pj4CHmjhqOt6dlXnDV2tqaVYvDcXF2\n5t0PPmTk+Mk0alCPSWNHKh1NES5OjiydPIG0axn0GTWeeWs+pnP7tvR+vRM+FbxYOW0i5y7G0Wvk\nWFZv3sqQXj14Kayl0rFLhbI4tbZK/y8qgGnTpjFlypR/9QC1q8jO9TCnLx9Ak2O5U8g+jtrZA4CC\nzDSFk5gnew9v2Xce4e6+I6/Pw6mdPciNj1E6hllyCgyV/eYR7r6vsk5ZxhDyf8u9bmNuX7fsa889\nip3bnaFn2dGnFE5inlxr1lU6wr/WIrSTSbb7e8w2k2zXGEadM3R38oR27dqZOo8QQgghhBDCDCk9\nDbYpGDVMbvXq1QA0bdrUpGGEEEIIIYQQoqQY1TNkb2/Pp59+iq+vb/GMNi+88IJJgwkhhBBCCCHM\nh4qy1zNkVDEUEhICwI0bNwA5IVgIIYQQQghR+j22GMrMzMTDw4NmzZo9bjUhhBBCCCGEKHUeWwz9\n9NNP9OnTh3Xr1hm063Q6ZsyYYdJgQgghhBBCCPNhpbKwi6726XPn4pLPPPMMWVlZpKenk56eXjxc\nTgghhBBCCCFKK6PKu127djF16lTq1avHkCFD8Pf3N3UuIYQQQgghhBkpixddNaoYcnNzw83NjYKC\nAp566iny8vJMnUsIIYQQQgghTMqo2eQcHBw4cuTOlad3795NTk6OSUMJIYQQQgghzIvFXnR10KBB\neHl50aNHD1JTU+nXr5+pcwkhhBBCCCHMiMpE/yjJqJ6hcuXKERgYCEDv3r1NGkgIIYQQQgghSkLZ\nmx9PCCGEEEIIIYwgxZAQQgghhBDCIhk1TE4IIYQQQghh2criRVdVer1er3QIIYQQQgghhHl7sXZ3\nk2z3l9NfmGS7xih75Z0QQgghhBBCGMHkw+RuZ6eb+iFKJTvXCmSdPqZ0DLPkXrshAAWZaQonMU/2\nHt5obmQoHcMsqV08AdDkZCqcxDypnT3kmPwIdq4VZL95BLWzB4Acdx5B7eIp+84j3N13chNiFU5i\nnpwCqisd4V+z2OsMCSGEEEIIIURZIxMoCCGEEEIIIZ5I6QukmoL0DAkhhBBCCCEskhRDQgghhBBC\nCIskw+SEEEIIIYQQT6SSCRSEEEIIIYQQomyQniEhhBBCCCHEE8nU2kIIIYQQQghRRkjPkBBCCCGE\nEOKJZGptIYQQQgghhCgjpGdICCGEEEII8URWqrLXj1L2npEQQgghhBBCGEGKISGEEEIIIYRFMmqY\nXH5+PidPnkSr1Ra3tWzZ0mShhBBCCCGEEOalLF501ahiKDw8HDc3Nzw9PU2dxyR0Oh2LV6zmp193\nAtDm+daMHTEMWxvLO2VKV1jIyk1fsOP3PwAIe6YJH/bthY2NDXMj1vPj3v2oAD2gAoa9/RZdX2yv\nZGRFaDQauvcbwJjhQ2nSsD4Ah48dZ+nqtcRfvkxFLy/69HyT115+SeGkyom/fJnZ8xdx+mwUbi6u\nvNnldd5+q4fSscxCYlIy4YuWciIyEodyDrR7IYxh7w3C1tZW6Whm4dddexg3eToqlQq9Xo9KpaJ1\ni+YsnjdL6WiK02q1zApfyO59+7FT29GrRzf69uqpdCyzIcedJ9NoNHTr3Y9xI4fTpFFDpeMoJiEx\nifCVEZyNOY+rsxNdOr5Ery6vAXAp4QrzVq4h5sIlKnh60L9HN9qHyY/8lsqoakCv1zN06FBTZzGZ\nhctW8dvBP1i2YB4AYydNw83lU94b2F/hZCVv+cYt/HHsOPPHjQRg8pKVuDg5MeDNN0hISuaD3j1p\n99yzxeuXL1dOqaiK0Wg0jJ0ynbiEy8VtlxOTGDp6PAP7vU3bsFacjopi6uxwPNzcadHsGeXCKkSn\n0zF42CiaNmrA5PFjiE+4zNiJU6ng5cWL7dooHU9RWp2O90eMJjgoiM0b1pGZlcWk6bNQqVSMHPa+\n0vHMwqW4BJ5v1YKJY0ei199ps1OrlQ1lJhYsXc6ZqGjWr1pO2tV0xk+eho+3N+3bPK90NMXJcefJ\nNBoNYz6aQlx8gtJRFKUrLGTYxGk0qluHCcPeIyExiY/mLMDL053nmz/LiCkzaPFME6aMHMrxyLNM\nXbgU/0q+PFU9WOnoZs9iL7pauXJlLly4gFarRafTodPpTJ3rfyb35k2+/n4bUz8aS52nn6LO008x\nZEA/omNilY5W4m7eyuOH3XsZP+hdaoUEUyskmP7dOhMbFw9AQnIyoVUDcHdxKb5Z2heUuIQE3np3\nMMkpqQbtu/buJzQkmH69euBXyZcX27bh5Q7t+GXXboWSKiv9Wga1a9VkwugR+FeqRItmz9K0cUOO\nnTipdDTFnY2KJik5hZlTJhJQpTIN6tXl/UHv8vOOXUpHMxuX4hMIqRaEu5sbHu53bo6O5ZWOpbj8\nggK++2E7Y0cMI7R6CK1aNKdv75588fU3SkczC3Lceby4+AR69n2X5JQUpaMo7lpGJk+FhjDmvYH4\n+XjTvHFDGterw4nTUcRdSSTlajoDe/Wgkrc3r7R7geDAAI6fPqN0bKEQo4qhc+fOsWTJEj788EOG\nDRvG8OHDTZ3rf+bkqdOUs7enScMGxW2vvNielYvnK5hKGZExsdjb2dHw6aeK215s+RwLJ4wmK/sG\nuTdvUcXXV8GEyjt2MpImDRuwae0q9Hd/sgbavdCacSOGGayrUqnIvXmzpCOaBV8fb8JnTkP9d7F8\nMvI0x09G0rRxI4WTKS+gSmVWLVmAvb2dQXtubq5CicxPXEICAVUqKx3D7MSev4BWp6NendrFbfXq\n1CEq+pzB8chSyXHn8Y6dOEmTRg35fMNai99ffCpWYPb40ajVd4Ymn4qK5uTZKBrXr4OLkxMqlYpt\nO3ej1+s5HR3D5aRkagQHKZy6dFCZ6B8lGTVMrmPHjrRo0cLUWUwiMTkFH5+K/LJzD+s+3Uh+fj5t\nwloxdMhAiztnKPnqVby9PNn1x3/47Ltt5BUUENa0CYN7dCUuMQkrKysitn7FnycjcXVyotvLHXip\nVen8//7f6vpap4e2V/bzM7ifmZXFzj37GNivT0nEMmvPv/wqGRmZtGj+LG3CWikdR3Furq4G4/T1\nej1ffPUtTZvIFza4M4wwMSmZA38cYsWaj9Gjp21YK4YMeMfiz6nKyMjExdnJ4HXw8HBDq9WRmXUd\nTw93BdOZFznuPKhr59eUjmCWXuzZl4ys6zRv3JDnmz+LSqViyNtvsWLDZyz7+FP0ej3v9OhKo7p1\nlI5aKpTFYXJGVQN79+4ttcXQrbw8kpNT2frNt0ydMIabt24xY+5CioqKGD38A6Xjlai8/AJSrqbz\nzY5djBvYn1v5+YSv3UBhUSH+Pt6oVCpCAgPo+mJ7TkRFE752Aw729rRu2ljp6GYlv6CAEeMnUcHL\n65HFkyVZvmAe165lMH3efMIXL2XsiNLTc1wSwhctJfbCRbZuXK90FLNw5UoiRUVFODg4sGjeTJKS\nk5m7cCl5+fmMH/Wh0vEUlV9QgK2t4dBk9d/3tVqNEpHMlhx3hLEWTZ3Itcws5ixfxaI1HzN8QD+u\nJKfwavu2dGrfhpiLl1gcsZ7qQVVp9WxTpeMKBRhVDOl0OsaMGYPv30OoVCoVw4YNe8JfmQcba2tu\n5eUxZ9pkKvn6ADBy6BA+mjbL4ooha2tr8vLzmTb0PXwqeAHwQe8eTF+xhn2b1tO2+bM4lb8zbj+o\nsj+JqWl8t2uPFEP3uXnrFu+PGktyahqfRazAzs7uyX9UxtUMrQ6h1RldkM/E6bMYOfR9bCys1/VR\n5i5YzFff/cDiebOpGhCgdByzEFQ1kN93/oSzsxMAIdWCKCrSM27ydMaOGIaVleVe/s7OTv1A0aP5\n+769vb0SkcyWHHeEsUKDgwgNDiK/oB9TFywhsIo/p6Nj+ObjVXeWVwvi6rUM1mzcLMWQhTLqyNGz\nZ+md1tPL0wNra+viQgjujOnXaDRkXb+Ou5ubgulKlqebK9bW1sWFEEBlXx80Gg3ZObm4uTgbrB9Q\nqRJHTp8t6ZhmK/vGDQYOG8n17Gw2rFpKJR+fJ/9RGZV+7RrRMbG0eq55cVvVwEC0Wh23buXh8o99\nydLo9XomTZ/Fr7v2sGDODFo+10zpSGblbiF0V9WAKuh0Oq5fz8bDgoeCVfDyIicnF51OV/zFPjMz\nC7WtLS7Olv2eAjnuCONdy8zk3IVLtLjvx9yqVfzR6nRciEugWkAVg/VrBFfj829+KOmYwkwY9RNc\nQEAAx48fZ9u2bRw5coTKlUvPia91nq5FYWEhF/+eMQ3gUlw85R0ccHVxUTBZyasVEkxhYSFxiUnF\nbfGJSTiUK8embdsZNXeBwfqx8fEWP6HCXVqdjvdHjSMnN5dPVi9/4BwiSxMXf5kPx0zgenZ2cVvU\nuRjcXF3lCwkQvngZO3bvYUn4HMJals4hxqay97ffad2hk8GspOdiz+Pk6GjRhRBAaEgItra2nLpv\nVqvjJyOpWSPUonvM7pLjjjBW/JUkRk+fQ/aNnOK26PMXcXd1wcvDnbgriQbrx11OxM/Hu6Rjlkoq\nlcokNyUZdXRdvXo1np6edO/enQoVKrBy5UpT5/qfqezvR6vnmjF5xhyiY2I5fjKSpasi6PxqR4v7\ncPH38aZ5wwbMXBlBbFw8p87FsHrLl3R6oTXNG9Tnr1On+frXnSRfTefrX3ex4/dDvNXpZaVjm4VN\nX3zFudjzTJswFns7OzKzssjMyiInxzJnCGtYvy5BVQOZOG0WcQmXOXDwEEtXrmGATChB5JmzbN76\nFUMG9KdGaHUyMrOKbwIa1qsLwPQ587l8JZHf//gPi1espm8vuXCmvb0dHV/qwKx5CzgbFc3+AwfZ\nuPkL3ureVeloZkGOO8JY9WvXIqhKZaYuWELClSQO/nWUlZ9spF/3rnQIa8XVa9dYHLGepNQ09v3x\nHzZ+9S09O8s5wJZKpTdi/sVp06YxZcqU4vuTJ09m+vTpRj3A7ez0/z7d/0hefj7zFi1lz/4D2Fhb\n88pLHRg2ZKCi44vtXCuQdfpYiT9ufkEBiz/ZxG+Hj2BtZc2LrVowuGc3bKyt2ffnYdZ//R3JV9Op\nVLECA7t3oYUCV692r33nMQsy00r8se9Xr3lrIpYupHGD+vToN4Bz5y88sE7dp2vxyerlJZrL3sMb\nzY2MEn3Mh7mafo1Z4Qs5duIkDg4O9Oj6Bv16KzukVu3iCYAmJ1OxDAuXrmDjlq0GbXq9HpVKxck/\nf1f0Rxi1s4dZHJNjz19k/tLlREXH4OjoSJfXOzGgb29FM9m5VlB0v7mroOA2M+fNZ8/+3yhfvjxv\n9+xBrx7dFM2kdvYAkOPOI6hdPM1i37lfnSbNWbtiicHMlkq4u+/kJpT8tR3TMzKZt2INx0+fpXy5\ncnTt9BJ9unYGICr2AkvXbSD2Ujye7m707NyJ119sX+IZnQKql/hj/n+91WSASbb7+eG1JtmuMYwq\nhj766CNGjx6Nq6sr2dnZLFiwgJkzZxr1AObwwWuOlCqGSgNzKYbMlbkUQ+bIHIohc2YuxZA5Mpdi\nyByZUzFkjsyxGDIXShZDpUFpLIZ6NRloku1uOhxhku0aw6iukW7dujFp0iTKlStHfn4+Awea5oUQ\nQgghhBBCiJJi1HiN9PR0bGxsSE1NpaioiDVr1pg6lxBCCCGEEMKMWKlUJrkpyaieod27dzNhwgRc\nXV1NnUcinhhZAAAgAElEQVQIIYQQQgghSoRRxZCzszNeXl5PXlEIIYQQQgghSonHFkNbtmwBQKfT\nMWvWLAIDA4uX9egh06AKIYQQQgghSq/HFkO+f19w01cuvCmEEEIIIYRFU/oCqabw2GKoVatWJRRD\nCCGEEEIIYc6UnuzAFJS7+p8QQgghhBBCKMioCRSEEEIIIYQQlk2F9AwJIYQQQgghRJkgPUNCCCGE\nEEKIJ5JzhoQQQgghhBCijJBiSAghhBBCCGGRZJicEEIIIYQQ4onK4nWGpGdICCGEEEIIYZGkZ0gI\nIYQQQgjxRGVxAgWTF0N2rhVM/RCllnvthkpHMGv2Ht5KRzBbahdPpSOYNbWzh9IRzJYckx9N9pvH\nk+POo8m+83hOAdWVjiDEI5m8GCrITDP1Q5RK9h7eaHIylY5hlu5+qHRt2E/hJObpq2MbSPttn9Ix\nzJJ3qzAACjJSFE5inuw9fdHcyFA6hllSu3jKMfkR7h6T5fV5OLWzBzmXzikdwyw5B9UAIPP4Xwon\nMU8eDZoqHeFfk3OGhBBCCCGEEKKMkHOGhBBCCCGEEE+kQnqGhBBCCCGEEKJMkGJICCGEEEIIYZFk\nmJwQQgghhBDiiazK3ig5KYaEEEIIIYQQ5u/GjRuMGzeOSZMmYWVlxapVq1CpVPj7+9O/f38A9uzZ\nw969e7G2tub111+nfv36j92mUcVQfn4+J0+eRKvVFre1bNny//FUhBBCCCGEEKWJklNrFxYWsm7d\nOuzs7ADYuHEj3bt3p0aNGqxbt46jR48SHBzMjh07mDdvHrdv32by5MnUrl0bG5tHlzxGFUPh4eG4\nubnh6SkXXBNCCCGEEEKUrE2bNtG2bVu+//57AOLj46lR4861rOrVq0dkZCQqlYrQ0FCsra1xcHDA\n29ubK1euULVq1Udu16hiSK/XM3To0P/B0xBCCCGEEEKURlYK9Qz99ttvODs7U7t27eJiqKioqHi5\nvb09+fn5FBQU4ODgYNCel5f32G0bVQxVrlyZCxcuEBAQUNw99rjuJiGEEEIIIUTZotQwuf3792Nl\nZcWZM2dISEhgxYoV5OTkFC+/WwSVK1fOoPj5Z3H0MEZVNOfOneP48ePF91UqFStWrPi3z0MIIYQQ\nQggh/pVp06YZ/Pe7777L559/TnR0NDVr1uTkyZPUqlWLoKAgtm7dik6nQ6PRkJycTOXKlR+7baOK\nofnz5wOQm5uLo6OjoidPCSGEEEIIISxbr169iIiIoLCwkEqVKtG0aVNUKhUdOnRg0qRJAHTv3v2J\no9mMKoaio6NZv349RUVFNG3aFC8vL8LCwv7/z0IIIYQQQgghjDRlypTi/546deoDy8PCwv5VnWJl\nzEpffvkl06ZNw9XVlddee42dO3ca/QBCCCGEEEKI0s8KlUluyj4nI6hUKhwdHQFQq9WUK1fOpKGE\nEEIIIYQQwtSMGibn7e3Nli1buHnzJj/88EOpu95Q9o0bzF6whL+OHqN8eQd6dn2Dt7p1UTqWWUhM\nSiZ80VJOREbiUM6Bdi+EMey9Qdja2iodrUQN/KgPFf0qMH3wfIN2a2tr5m2ewp97jvLtx9uL2zt0\ne4F2XcNw83Tl8sVENi35igtnLpV07BKh0WoZMHsuQ7t1oX5oKABXs7II3/g5UXFxVHR3Z8gbnWlS\n66niv/lm3z6+33+AzBs3CPKrxODOnakV9Og5/ssKjUZD93cGMWb4+zRpYHjFa61Ox5t9B9CmdUsG\n9eujUELzEH/5MrPnL+L02SjcXFx5s8vrvP1WD6VjmRWNRkO33v0YN3I4TRo1VDqO2dBqtcwKX8ju\nffuxU9vRq0c3+vbqqXQsxc1cupLk1DRWz50BwMHDR1n12eckpqYS4OfHkD5v8WzD+k/YStnxy+9/\nMCviY1SAHgz+/d2yRVhZWTF/w2ccOxuFi5MTb3V8kdfbPK9o5tKiLM4bYFTPUP/+/fH09KR69erY\n2dkxaNAgU+f6nxo+7iOSkpOJWLqQGRPH8/nWr9ny9bdKx1KcVqfj/RGjsbOzY/OGdcydMYV9B35n\n2aoIpaOVqFqNatC603MPXda5f0f8An0N2p7r0JSug15l8/JvGN1jCueOx/LR8g9x9XApibglSqPV\nMv3jDVxOTTVon7ByNW5OTkRMGEe7pk2YHLGWtMxMAHb9dZhPfvyJga+/xoZJH1EnOJgxy5aTeeOG\nEk+hxGg0GsZOmUlcwuWHLl/7ycZHLrMkOp2OwcNGUcnHh282f8aEMSOIWP8pv+zcrXQ0s6HRaBjz\n0RTi4hOUjmJ2FixdzpmoaNavWs7k8WNYu/5Tduzeq3QsRR05FcmPu/YU34+7ksj4OfPp/FJ7vlqz\nnPatWzB6xhxSrl5VMGXJeuGZJmxftYwfVy1j+6plbFu5hJAqlWndpBFe7m6Mnr8IrU7LJ7OnM/jN\nLiz7/AuOnY1SOnapYKVSmeSm6HMyZqW5c+fStm1b+vfvT4cOHUrVNYbOxZ4n8kwUc6dPpkb1EBrW\nq8vw9waxYdMWpaMp7mxUNEnJKcycMpGAKpVpUK8u7w96l5937FI6WolR26kZMKE3MacuPLCscjU/\nwjo9R3KCYSHQ8uVm7PhqL0d/O0F68jW+WPUd2Zk5NGhRt6Ril4iE1FQGzwsnNTPDoP1ETAxJ6emM\n6tWTKt7e9GjfjqeCqvLzof8AsOOvv3itVUta1KuLr5cXA157FXcXZw5FnlbiaZSIuITLvDXgPZL/\nUTTedf7iJb7/6VcCKvuXcDLzk34tg9q1ajJh9Aj8K1WiRbNnadq4IcdOnFQ6mlmIi0+gZ993SU5J\nUTqK2ckvKOC7H7YzdsQwQquH0KpFc/r27skXX3+jdDTFFBTcZs7y1dStWaO4LT0jg26vvMQbL3XA\nt2JF3nr9VcrZ23M25ryCSUuW2tYWdxfn4tv+w0dJz7rO2P59+fPUaZLTrzH9gyFU9vGmzbNNeanl\nc5y5cFHp2EIhRhVD5cuX59ixYyQnJ5OSkkJKKTpIJyWn4OzkRGU/v+K26tWCyMzKIjXNcn4leZiA\nKpVZtWQB9vZ2Bu25ubkKJSp53d/rTNSxGM6dMPyQUKlUDJrUl8+Xfc3NnFsGy75Y9R17v//doE2v\n11PeqWydSxd5/gINQkNZNXYM+vvao+MTCPb3x16tLm6rHRREVFwcAO926sTLzzU32JYKFTfz80si\ntiKOnYykScP6bIpYiV6vN1hWVFTE1Dnz+XDIAFycnRVKaD58fbwJnzkN9d/7z8nI0xw/GUnTxo0U\nTmYejp04SZNGDfl8w9oH9iVLF3v+Alqdjnp1ahe31atTh6jocxb7Wq38bBMNaj9NvafvDVNuWr8e\nH/w9FFdXWMi2nbvR6rQ8HVpdqZiKyisoYMP323i3y+s4OjhwIvoc9WuG4njfhThH9+tD39c6KZiy\n9FCpTHNTklFdPDk5Ofz8888GbfdPa2fOPNzduHnrFnn5+Tj8PfFDSloacOdcIh/vikrGU5Sbq6vB\nWHS9Xs8XX31L0yaW8aUk+OkgmoQ1YGS3SbzSq73Bsld6tycnO5c/dvxFm86tDJZdioo3uF/nmVr4\n+Ffk9OFoU0cuUZ1atnhoe+aNG3i4Gg4JdHN25tr1bABqBAYYLDt8Noqk9HQa1qhBWdX1tVceuezT\nLV/i5urKS+3a8PUP2x+5niV6/uVXycjIpEXzZ2kT1krpOGaha+fXlI5gtjIyMnFxdjI4p9XDww2t\nVkdm1nU8PdwVTFfyTp+LYd+hP/ly9TI2ffvDA8sTkpJ5c/BQ9Ho977/dC5+KFRRIqbwf9uzDztaW\njq1bApCcnk5FDw/WfvUtvx78g/IODrzZoR0vt3r4Z54o+4wqhkpL4fMwTz9Vk4oVvJgZvpCJo0eQ\ne+sWa9Z/Ctw5EVPcE75oKbEXLrJ143qlo5ictY01gya+zacLt5B/y7DHwqdyRV7u2Y5xvaY94q8N\n131v6jsc+PkQCbFXTBXXrNzWaLD9x1BZWxsbtLoH30+JV68y59PPaPdMU0IscIhYwpVENn7xFVs3\nWNZ5eMZavmAe165lMH3efMIXL2XsiOFKRxJmLL+gAFtbtUGb+u/7Wq1GiUiK0Wq1zFy6kpED++NY\nvvxD1/F0d2Pj0gVERp9jybpP8PP1ofWzTUs4qfK27fuNLu3aYG11ZzBUXn4BOw4eolWTRswdOZzY\n+HgWfLIRFydHnmtgOZNMiHseWwy99957BrNGWFtbU1hYiK2tLYsXLzZ5uP8FW1tbFs2ewdjJ02ne\n7mWcHB0ZNngAZ8/FUP4RBxBLNHfBYr767gcWz5tN1YAApeOYXJcBnUi9cpUj+088sGzQpL58t+En\nMq9ef+w2/IMq8dHyESTFp7B29kZTRTU7altbbhUUGLRpdTqDYXMAccnJjF62nCo+PozsaZkzhU2d\nM593+/TE20J/kX2SmqHVIbQ6owvymTh9FiOHvl+qzkkVJcvOTv1A0aP5+769vb0SkRSzbsuXVKnk\nS1izZx65jqODAyFVAwmpGsilhCt89ePPFlcMxcTFk5J+jXbNny1us7a2wql8ecb174tKpaJ6QBUu\nXk7k+z37pBgygtKTHZjCYz91lixZgl6vZ/369bRp04Zq1aoRHx9f6i66GhoSzLatm8i6no2zkyNX\nkpKwUqkstsv4fnq9nknTZ/Hrrj0smDODls81UzpSiWjWtgmuns58dmAVADa21lhZWbH1yMegh4CQ\nyrw5+HUA1Ha2BNUMJLhWVeYOXwpA1RoBTFj+IZfPJzLvw2UU6goVey4lzdPVlUtJyQZtWTk5uLvc\nGzoXk3CZMcuWE+Tnx5z3hzzQk2QJUtOucurMWWIvXGTF2g0A3NbcJiomhrPR51ixYK7CCZWRfu0a\n0TGxtLrvvLKqgYFotTpu3crDxUXOqxIPV8HLi5ycXHQ6XXHRnJmZhdrW1uLOx9t54CCZ16/TsvOb\nAGi1OoqKimjVuTsfL5xLXn4+tWuEFq8fWNmPyOhzSsVVzF+nz1CzWlU8XF2L2zzd3Cgq0hv82F/Z\nx5vjUWVrqLsw3mO/odwdl3v16lWqVasGQGBgIKmPmDHJHOXk5jJ0zHgWz5mJu9udN8P+3w8RWj0E\nh/tOnrNU4YuXsWP3HpaEz+G5x/zCVNZMHTgX6/u+oL/csy1VQ6uwdOLaB+bQHz57ELGRF9n22S8A\nePl6MmHZh8RFJzB/1HK0Gl2JZldazaqBbN6xk9saDXZ/9waduXiRp6reuY5QyrVrjFm+gupVqjBz\nyCDsLOyaVXdVrODFT19+btA2ZvJ06j5di75vdVcolfLi4i/z4ZgJ7Pv1R9z+/oISdS4GN1dXKYTE\nY4WGhGBra8up02doWL8eAMdPRlKzRihWVkbNB1VmRMybie6+H+G2fP8jMRcvMX30h2zfvZeDR46y\nZcWS4uXnLl4iwN/vYZsq06IuXKJeaKhBW63gID4+dZrCoqLioXPxScl4e3kpEbHUUVH2eoaMOno4\nODiwdetWjh07xpYtW3C9r8I2d85OThQU3Gbh8lUkJaewa99+1n22iYF9LfuihwCRZ86yeetXDBnQ\nnxqh1cnIzCq+lXWZV6+Tnnyt+HYrJw/NbS3XUjIM2tOTr6HVaLmZc5PszDvXyek/thcFeQWsm7MR\nB0cHXNydcXF3xs5e/YRHLRvqBgdT0d2d2Z9+RkJKKlt27CQ6PoGOf//Sv/iLrZSzs2NEzx7cys8n\nKyeHrJwc8m/fVjh5ybKyssKvkq/BTa1W4+zshJenh9LxFNOwfl2CqgYycdos4hIuc+DgIZauXMMA\nC78QrXgye3s7Or7UgVnzFnA2Kpr9Bw6ycfMXvNW9q9LRSpy3lxd+Pt7FNyfH8tip1fj5ePNK2xfu\nXIB142YSU1LZuu0n9vx+iL7d3lA6domLS0oi0K+SQdsLzzTF2saaOWvXk5iaxo4/DvHLwUN0louu\nWiyjxq4MHTqU3bt3c+LECfz8/OjSpYupc/1PzZ8xlenzFtClTz8qenkxZdxoWlhQL8ij7Nn3GyqV\niqUr17B05RrgzrA5lUrFyT9/t7hf2h7l/ilb7crZUeeZp0APy3+YZ7Dedxt+4quIB2f0KQvu/x3I\nysqK2UMGMW/j5wyYPQdfLy9mDR5EBXd38goKOBp9DhXQY+Ikg2289WIH3nmlY4nmVsLjrs5dFq/c\n/W/Z2NiwcvECZoUv5K1+A3BwcKB3jzfp0dXyvqg9iewvDxo9fCgz582n/3tDKV++PIPffYc2Ya2V\njmVWKnlXZNn0KSxat4Et3/9IJe+KzPtoDCFVA5WOVuKu5+Tg7Gh4friDvT3Lxo9hwScb6T1+Eh6u\nLozq25tn69VRKGXpUhaPSyr9Yybnj4yMfOQf1qlj3E5TkJn271NZAHsPbzQ5mUrHMEtq5zu/mndt\n2E/hJObpq2MbSPttn9IxzJJ3qzAACjJKz7XQSpK9py+aGxlPXtECqV085Zj8CHePyfL6PJza2YOc\nS5Z3Po4xnIPuXFIh8/hfCicxTx4NSt+EFhPajTfJdmfvnGOS7RrjsT1Dhw4deuQyY4shIYQQQggh\nROlncbPJDRw4EGtra3Q6yzpBXAghhBBCCGGoDNZCjy+GVqxYwbBhwxg2bFjxGMG755SsWLGiRAIK\nIYQQQgghhCk8thgaNmwYACtXriyRMEIIIYQQQghRUoyaTW7v3r388ssv3L5valzpGRJCCCGEEEKU\nZkYVQ7t27WLcuHGl6vpCQgghhBBCiP8di5tA4S5nZ2e85Mq8QgghhBBCiDLkscXQli1bANDpdMya\nNYvAwHsX7OrRo4dpkwkhhBBCCCHMhgoL6xny9fUFwMHBAbVajYODA1988QUdO5b9q8gLIYQQQggh\n7imLw+SsHrewVatWtGrVisOHD1O/fn1atWrF9OnTOXr0aEnlE0IIIYQQQgiTMOqcIWtra7y9vQGo\nWLFi8TWHhBBCCCGEEJahLJYARhVDnp6ebNmyhZCQEC5evIi7u7upcwkhhBBCCCGEST12mNxdQ4YM\nwcXFhZMnT+Ls7MzgwYNNnUsIIYQQQgghTMqoniG1Ws1LL71k6ixCCCGEEEIIUWKMKoaEEEIIIYQQ\nlq0szhsgxZAQQgghhBDiicri1NoqvV6vVzqEEEIIIYQQwrzNfmWKSbY74cdpJtmuMUzeM1SQkWLq\nhyiV7D19uZ2drnQMs2TnWgGAtN/2KZzEPHm3CuOHD5YpHcMsvbp8KAAFmWkKJzFP9h7e5F9LVjqG\nWSrnVQlNTqbSMcyS2tkDQF6fR1A7e3ArOU7pGGapfKWqAFyPOqFwEvPk9lR9pSP8a2WwY8i42eSE\nEEIIIYQQoqyRc4aEEEIIIYQQT1QWzxmSniEhhBBCCCGERZJiSAghhBBCCGGRZJicEEIIIYQQ4olU\nyDA5IYQQQgghhCgTpGdICCGEEEII8UQqmUBBCCGEEEIIIcoG6RkSQgghhBBCPJFV2esYkp4hIYQQ\nQgghhGWSniEhhBBCCCHEE8k5Q0IIIYQQQghRRkgxJIQQQgghhLBIMkxOCCGEEEII8URlcZhcmS6G\nNBoN3d8ZxJjh79OkQX0ADhz6k+URH5OYlExAlcoMHdifZk0bK5xUOb/u2sO4ydNRqVTo9XpUKhWt\nWzRn8bxZSkcrURqtlgGz5zK0Wxfqh4YCcDUri/CNnxMVF0dFd3eGvNGZJrWeeuBvdx8+wrbff2fF\n6FElHdukMvNusD32DxKup6K2tqWOdzXaBTflu+jfOJESA6gAffH67uWcGfNcLwBOpV5gb9xRsgtu\nEujmQ6fQFng4uCjzREqQRqOhe78BjBk+lCYN7xxzDh87ztLVa4m/fJmKXl706fkmr738ksJJlaHR\naOjRfzBjhr1H47+PydcyMpm1cAmHjx3H1cWFvj3fpOtrnRROqhytVsus8IXs3rcfO7UdvXp0o2+v\nnkrHMhvy+tyTlJLKgpURnDobRTl7e9q2bsl7/Xoza/EKftq1p/hz/a5KPt78+PkGBROXHF1hISs2\nbmHHgYMAhD3ThBHv9MHGxoajp8+y+vOtJCQl4+XhzluvvkzH51srnFgoyahiKDo62vCPbGzw8PDA\nw8PDJKH+FzQaDWOnzCQu4XJxW1zCZUZPnMqooUNo1rQJe377neHjJ/LDlo1U8vFWMK1yLsUl8Hyr\nFkwcO5K7x0w7tVrZUCVMo9Uy/eMNXE5NNWifsHI1gb6+REwYx6FTkUyOWMtnUyfjfd9+fyI2loWb\ntxDs71/SsU2qsKiQT0/+jLejB+81eYObmjy+PrsXgE6hz/Fi8DPF6+Zq8lhz9HueC6gHwIXMRL48\nu5uXQpoR6lmFP5POEnHsB0Y164Ha2laR51MS7hxzphsccy4nJjF09HgG9nubtmGtOB0VxdTZ4Xi4\nudOi2TOP3lgZpNFoGDfV8Jis1+sZOmYC7m5ubF0fwbnzF5g8ex5V/P2Li0lLs2Dpcs5ERbN+1XLS\nrqYzfvI0fLy9ad/meaWjmQV5fe7Q6nQM+2gK1QID+WzFYjKvZzM1fBEAYz4YxLAB/YrXzci6Tv/h\no+nV5XWl4pa45Z9+zsFjJ1gwfjQAkxYvZ8PX39Oh1XOMmj2fd7p25vlnm3D2/EVmr4rA3cWFZhZ6\nzPm3yuLU2kYVQ19++SXZ2dkEBgaSkJCAjY0NWq2W559/nldeecXUGf+1uITLjJs684H2tKvpdO/y\nevGvjn26d2X9xs2cjoq23GIoPoHqwUG4u7kpHUURCampzFj/4C9lJ2JiSEpPZ+XY0dir1VRp782x\nmBh+PvQf3nmlIwCfbv+JzTt34V+xQknHNrnEG+lk5ecwtGlXbK1t8CrvSttqTfjp/CFeqt4MO5t7\nBfP3537H36Uiz/jXAuDQldPUrliN5lXqANCxenPOZ1zhZOp5mvg92LNWFsQlJDBuyowH2nft3U9o\nSDD9evUAwK+SL8dORvLLrt0WVQzFJVxm/LQHe5v/+OswiSkprF22ECdHR6pU9uf4qdOcPhtlkcVQ\nfkEB3/2wnZVLFhBaPYTQ6iH07d2TL77+xuK+7D+MvD73RMXEkpySxuY1y7G3s6OKvx9D+vZm0Zp1\nfDioP+UdHIrXnbN0BbVqVKdLp5cVTFxybt7K4/tde1k0cQy1qgcD8G63N9hz6E9sbW0ICQyg9+t3\nvrtW8q7Iyahz7Dx4SIohC2bUBApqtZr58+czfPhw5s+fj6enJwsXLuTw4cOmzvdfOXYykiYN67Mp\nYqVBF/GzTRrx4ZCBAOh0hXy3/We0Wi11atVUKqri4hISCKhSWekYiok8f4EGoaGsGjvmvgFfEB2f\nQLC/P/b39ZLVDgoiKi6u+P7xmFgWDh9Ki7p1SzBxyfAq70rfei9ja234e0mBVmNw/3J2GueuxdOx\nevPitqy8HKq4Gv644O3kwZXsNNMFVtidY04DNq1dZXDMafdCa8aNGGawrkqlIvfmzZKOqKjjpyJp\n0qAeG9esMHh9jp44RaN6dXFydCxu+2jUcN59u5cSMRUXe/4CWp2OenVqF7fVq1OHqOhzBq+bpZLX\n554q/n4smzMdezs7g/abN28Z3I+MOseB/xxm1HsDSzKeoiLPxWBvZ0fDp2sVt73YugWLJo7lhWbP\nMOrdtw3WV6kg91ZeCacsvVQqlUluSjKqZygnJwf1318KbW1tyc3NxcbGhqKiIpOG+291fe3xvVUJ\nl6/QuVc/ivR6hg16F19vy+wV0up0JCYlc+CPQ6xY8zF69LQNa8WQAe9ga1t2hzPdr1PLFg9tz7xx\nAw9Xw3Nc3JyduXY9u/j+8tEjAThxLsZ0ARVSXl2Oah5+xff1ej3/STxD8H1tAPvjjvN0xSAqOroX\ntznaleNGgeEHcnZ+LuXV5UwbWkGPOselsp/h65WZlcXOPfsY2K9PScQyG11effgxOSk5Be+KFVm5\nbgPbf92Fo2N53ur6Bq++3KGEE5qHjIxMXJydDI6/Hh5uaLU6MrOu4+nh/pi/Lvvk9bnHzcWFxvXv\n/RCn1+v58oftNGlQz2C9DZu38kKL5lS1oB89k66m4+3lya6Dh/j0mx/IKyjg+WebMrhnN/z/MQoo\nMzub3X/8yTvdOiuUtvQpg/MnGFcMNWrUiEmTJlGtWjUuXbpEgwYN2LVrF/6l9DwJT08Ptqxfw6kz\nZ1mwfBX+fpV4vuVzSscqcVeuJFJUVISDgwOL5s0kKTmZuQuXkpefz/hRHyodT1G3NRpsbQzfHrY2\nNmh1WoUSKWt77B+k5mbwQdOuxW3X83OJzbzMe43fMFi3jncwv57/k+qeVQhw8+ZYcgzJOdcIcvf7\n52YtSn5BASPGT6KCl5dFTxBwv1t5+fy0YxcvtG7BkrkziI49z5xFS3F1daZV82ZKxytx+QUF2Noa\nnrOp/vu+9h+9spZIXp9HW7AygvOX4vh8zbLittSr6fznyDE+W7lYwWQlLy8/n5Sr6Xz9yy7GD3mX\nW3n5zItYT2FREcP73ut1Lrh9m3HzFlPBw53X272gYGKhNKOKoTfeeINGjRqRnJxMWFgY/v7+5OTk\n0KZNG1PnMwnH8uWpHlyN6sHVuHApni+++d4ii6GgqoH8vvMnnJ2dAAipFkRRkZ5xk6czdsQwrKws\n9zJUaltbbhUUGLRpdTqDYXOW4seYg/yVeJZedTtQofy9c8vOXL2Ih4Mrfi6G50w1rlSTa7ey+fj4\nNvRANXc/GviGkq+7XcLJzcfNW7d4f9RYklPT+CxiBXb/GNpiqWysrXF2cmLymJGoVCpCQ4I5f/ES\nX3+/3SKLITs79QNf6jV/37e3t1ciklmR1+fh5q9Ywzfbf2b+1IkEVr73I/WeAwfx96tEzeohCqYr\nedZW1uQV5DPtw/fxreAFwAd9ejJ92eriYuhWXh4jZoWTdu0aEbOnWdzEUcKQUcVQZmYm3377LYmJ\nifj6+tKnTx8qVCh9J41fuBRHXn4+de6bHrlqQBVOnTmrYCpl3S2E7qoaUAWdTsf169l4WNCQg3/y\ndEgeV+MAACAASURBVHXlUlKyQVtWTg7uLmV/eui79Ho9X0ftIzLtAj3rtKOGV4DB8tiMK9SqUPWB\nv1OpVLxcvRntg5tyW6elvNqezZE7cS/nXELJzUv2jRsMHDaS69nZbFi1lEo+PkpHMhtenh4U6YsM\nxosHVPbnyPGTCqZSTgUvL3JyctHpdNj83TOdmZmF2tYWF2fLfP/cT14fQ3q9nqnhi9m57zfmTZ5A\ni2eaGCz/z9HjhD33rELplOPl7oa1lXVxIQRQxdcXjVbD9Rs5WFmp+GDqbLJzclg1Y7LBesIyGfXT\nf0REBC1atGDmzJm0bNmSNWvWmDqXSezc+xuz5ht2F5+LPU+gBY2lvd/e336ndYdO6HS64rZzsedx\ncnS06EIIoGbVQC4mJXFbc+9XyDMXL/JUYKCCqUrWT7GHiEy7QK86HXjqIUVPYk46Vd18H2j/43Ik\n++KOYWNlTXm1PZpCLRezkghyr1QSsc2KVqfj/VHjyMnN5ZPVyx84h8jS1a5Vk4uX4iksLCxuuxSf\ngK+Fzu4ZGhKCra0tp06fKW47fjKSmjVCLbqn/i55fQwtXLWWXfsPsHD6JFo9ZHbKqJhYGtR+WoFk\nyqpVPZjCokLiriQWt8UlJuFgXw4nx/KMmBlO7q1brJk19YFziMSTWalUJrkp+pyMWUmr1dKwYUPK\nly9P48aNDT64SpNXX+5AYnIKK9Zu4EpSMpu//pade/fTv7dlXrCtYb07J19OnzOfy1cS+f2P//wf\ne/cdHlXRuH38u+m9kAQCJKEHgvSuomIUFSsiIlVAQaqABbEiIIh0kC5WEOyKHURFmo9Ib6EngRRC\nSCGUJGQ32fcPIJoHQ/b5vW7Okr0/17WX7OzJyb3jnLOZnTlzmDVvIf0vLQPszJrVq0eVSpV4/f0P\nSEw9wYpVq4lLSOTem9qX/cMVwLHTaWw6vouOddpQPSCMsxdyix8A2XlnKLAUlFg44bJK3gGsT9zJ\nocwkTp3PZsWunwjxCaR+aI3yfhuGW/bRp+w/eIjxL47By9OTzKwsMrOyOHPmrNHRHMJdt8Xi5ubK\n+CnTOZaUzPer1/Dtjz/xSBfnvKbKy8uT++7pxKQp09m7L4616zawdPlH9O7RrewfdgKqn7/sjtvP\nR19+zaB+vWlQry6ZWdnFD4DUtJOcz82jTk3nO+9GVg3nplYtmThvMQeOJrAz7gALl3/MA3fEsuLr\n7zmYkMDLwwbh6eFO5unTZJ4+zRknW+FTSrJpmlxhYSHHjx8nKiqK48eP2zvTv+rv0y8iqlVlwcwp\nTHtzAcs++YyIalWZMWk89evVNTChcQIDA1g0ZwbT5syle98B+Pn50e2hB3nMSTuHf/9ewsXFhdeH\nDmbK0g954vXJVAsLY9KQwVSp5BwjZntPHgVMrDr8B6sO/3Gp1AqYeL3jEM4V5AEmfNyvnKffsHIt\nbq3dks/2/oK50EL90CgebhRbjumNdXGZ0Iv//nntbxQVFTFo5DMltmnWuBHvLZxrQDrj/f2c7OPj\nzaLZ03l9xhwe6TeQ0JBKvPDMSG66vp2BCY01etQIJk6ZxoBhI/D19WXIwMfpGHur0bEchurnol/W\nb8JkMjHv7feZ9/b7wMVpcyaTiT9/+pas7NOYTKYrpsI7i3GjhjHznaU8OW4Srq4u3HPrLQzp1Z2B\nz4+lqKiIJ//rvmdNGtRn0cRXDUp7bTFR8ZaTM1ltWJw/MTGRxYsXk52dTXBwMIMGDaJmzZo2/YL8\njNT/34wVkldoNS6cTjc6hkPyDLp4PVrab78anMQxhXeIZeWTb5a9oRPqPHcEAPmZFfeeRv8/vELC\nyTuVUvaGTsg7rDoFZzKNjuGQPAJCAFQ/pfAICOF8SnzZGzoh3+oXp1hn79tucBLHFHzdtXej14U9\nJttlv0M+esEu+7XFVUeGhg0bVvwtntVqxdPTkzNnzjBnzhxmzXKupRpFRERERKRiuWpnaPbs2Vit\nVt555x06duxI3bp1SUhIYPXq1eWVT0REREREHIDRix3Yw1U7Q5fv8nzy5Enq1r14XU2tWrU4ceKE\n/ZOJiIiIiIjYkU0LKPj4+PDxxx9Tt25dDh06RFBQkL1ziYiIiIiI2JVNS2uPGHFx1Zbt27cTFBTE\n8OHD7Z1LRERERETErmwaGfLy8uK+++6zdxYREREREXFQJme7ZkhERERERASgAvaFbJsmJyIiIiIi\nUtFoZEhERERERMpUEafJaWRIRERERESckkaGRERERESkTC4Vb2BII0MiIiIiIuKc1BkSERERERGn\npGlyIiIiIiJSJi2gICIiIiIiUkFoZEhERERERMpUAQeGMFmtVqvRIURERERExLG932+aXfbb7/3R\ndtmvLTQyJCIiIiIiZXKpgENDdu8MnU8+au9fcU3yjahDQU6G0TEckkdgKAC5qQkGJ3FMPtVqqe2U\n4nLb+XTwbIOTOKZui0Zx5uh+o2M4pIA6MRScyTQ6hkPyCAgBIGvXFoOTOKZKTVuTtXur0TEcUqUm\nrQA4uXGdwUkcU5X2txgdQdDIkIiIiIiI2ECryYmIiIiIiFQQGhkSERERERGHVVRUxOLFi0lNTcXF\nxYWBAwfi5ubGggULMJlMREZGMmDAAAB+/vlnfvnlF1xdXenSpQstWrS46r7VGRIRERERkTIZNUtu\n27ZtmEwmXnvtNeLi4lixYgUAPXr0ICYmhiVLlrBlyxbq1avHqlWrmDJlChcuXGDs2LE0adIEN7fS\nuzzqDImIiIiIiMNq3bo1LVu2BODUqVP4+fmxZ88eYmJiAGjevDm7du3CZDLRoEEDXF1d8fHxITw8\nnOPHj1O7du1S961rhkREREREpEwmk8kuD1u4uLgwf/583nvvPdq3b8/fb5Xq5eVFXl4e+fn5+Pj4\nlCjPzc296n41MiQiIiIiImUyejG5YcOGkZOTwwsvvEBBQUFx+eVOkLe3d4nOz393jv6JRoZERERE\nRMRhrV+/npUrVwLg7u6Oi4sLderUIS4uDoAdO3YQExNDnTp1OHjwIBaLhdzcXFJSUoiKirrqvjUy\nJCIiIiIiZXIxaGiobdu2LFiwgFdffZWioiL69+9P9erVWbRoEYWFhVSvXp127dphMpno1KkTr7zy\nCnBxgYWrLZ4A6gyJiIiIiIgD8/T05KmnnrqifNy4cVeUxcbGEhsba/O+NU1ORERERESckjpDIiIi\nIiLilDRNTkREREREymT0anL2oM6QiIiIiIiUydZ7Al1LKmRnKDEpmSlvLmDP/oMEBQbQ7YF7ebTb\nQyW2OZ6SSveBw9j43Re4uDjvbMGEY8d4fdpMdu/dR3BgEN0f7kK/3j2NjmWI5NQTTJu/iJ179uHj\n7U3HDjfz5IB+uLu7cyozk9dnzWPzth0EBQbQr0c3uj1wr9GRHcK4SW+QlJzCOwvnGh2l3GTlneGH\nwxs5npOGh6s7jSrX4fbabXFzceXE2Qy+P7yRtHMZhHgH0rFOO+pWiiz+2fjsFH44vImsvBwiAirT\nuUEHKnkHGvhuys/EOfNJOZHGwjdeA2DD5i0s+OBDkk6coGZEBEP79uaGVi0MTmkss9nMpKkzWPPr\nWjw9POnT8xH69+lldCzDWAoLmf/hR6xavwkrVm67vh2j+vVmylvv8MO6jZgA69+2r1Y5jM/nzjQq\nbrmyFBYyf9lHrFq/EYDY69vyVP8+uLm5kZGdzbQl77Fl914CA/zp88B9dLnzdoMTl6+cc+eY9eEK\ntsbtx8fbi66330a3jhfr4Pddu3nry69ISU8nKjycJ7o8SNvGjQxOLEa5amfIYrGU/oNlLFNnFEth\nIU++MJY2zZvy0lNPkpiUzIuTplA5NIS7YjsAkJZ+ilEvjaPAbDY0q9EsFgtDRj5Lu9YtGfvCcyQk\nHmPMy+OoHBbG3Xd2NDpeuTJbLIx48VXq1arJ0vmzycw+zbgpMzCZYNSgAYx84VWCg4P46K157D98\nhFffmEGNiOq0bdnc6OiG+uPPrXz5zXe0buE89VBYVMiHu3+gil8lnmjZhXMFuXx1YC0AN0U1572d\n3xATWosHGtxCUs5JPtq7isebd6aafxg5+edYsedHbq3ZiuiQGqxN3MryPT/yZJvuBr8r+/tz5y6+\n+elnWl76gyP+eBIvTJ7GqIH9uaFVC37d9B9GvzaZz96aR7UqVQxOa5zpc+ayZ18c7yyYS9rJdF4Y\nO56q4eHc1fE2o6MZYu6yFWzcup1pY54BYOyc+QT6+/FU/0cZ2uuv4ybrdA5DXp1Iz/vuNipquZu7\ndAUbt25j2vOX6mb2fAL9/Rn4yEM8O3k6wYEBvD91EgcTEpk4fzGR1cJp7UR/8L80bwEFZjMzn3mK\n3Pw8Jr39Hq4uLrSMiWHswsUMf+Rh2jZuxLpt23lp3gKWTZpA1dBQo2M7vAo4MHT1ztDIkSMxmUxY\nrdYS5SaTiXnz5tk12P9VekYGjRrUZ8yTQ/HwcCeiWlXatGjGtl17uCu2A2s3/s6kWfMIC61kdFTD\npZ/KoEmjhrw4+mk8PDyIrF6ddm1asXX7DqfrDO3bf5CU1BOsWDwXL09PakRGMOSxR5m5cAmtmzcj\nKfUEi2dOwd/PlxqREWzftZfdcfudujOUl5/PhDem0rxpE6OjlKvkM+lk559hSKuuuLu6EeoTxG21\n2vDjkd/x9/DF082DBxrcgovJhTCfYJLPnGRT0i4ebng7W1PjCPcL5caoZgA82OBWpmz6gPjsZGoH\nRxj8zuwnP/8Ck+cupFnDmOKy9IwMHrn/Hrre0wmA3l068/4nX7D3wCGn7Qzl5efz5cpvmT97Og3q\nR9OgfjT9H+3FR5997pSdoXO5uaxc8yszXniWRtF1ARjQrQu//P4Hvt7e+Hp7F287/e33ua5eHbrc\n4RyjH+fO57JyzS/MeGE0jaLrATDgkYf4ZdMf/LFzFykn05n36kv4+foQVa0qO+MOsPfQYafpDB08\ndoy9R47y4aTXiKhSGYDBDz/E/E8+Iyo8nIdui6XzrR0A6H7nHXz4/Y/EHY1XZ8hJXbUzNH/+/PLK\n8a+pVqUKk18eU/x85959bN+9lxdHDgdg459bGfbYo0RFVGfQsy8YFdMhVKsaztSJ44uf79i1m207\ndvHymGcNTGWMGlERzH3jNbw8PYvLTCYT586dZ8uOnbRu3hR/P9/i1158argRMR3KmwsW0aZlC0JC\nKrFz1x6j45SbUJ8g+jS5G3fXkqfPC5YCsvPPUM0/DBfTX1Nvw/1C2Xh8B3CxI1UzqGrxa+6ublTz\nD+V4zskK3Rma/8EyWjZpTEhwELvjDgDQrkVz2l0aUbQUFvL9z79itphp3KC+kVENdfDQYcwWS4kv\nGJo3bcpb776P1WqtkHP1r2bXgYN4eXrSqtF1xWV333ITd99yU4nt9hw6zIat21k27fXyjmiY4rpp\nfGXdzF26ghbXNcTP16f4tdED+xsR0zAnTmXg7+tT3BECqBsZQVZODlHhVWh9XUPg4rln1e//wWyx\ncF2dOkbFvaZUxPPQVTtDL730UqlveuLEiXYJ9G+665E+ZGRlc1O7Ntx2840AvPL0CAC2OdEfb7a4\n7d7OZGRkcnP7G+gY28HoOOUuODCQNi2aFT+3Wq188tU3tGnZjOTUE1QJC2X+ux/w3eqf8fP1pVfX\nB+l8950GJjbWzt17WfPrOr76eBnvf7jC6DjlytfDm9oef3VcrFYrm1P2Uju4On4ePqScSS+x/en8\ns+Sa8wE4W3Aef8+a/7U/H85cOGf33EbZvf8Av276D58sfJNlX6y84vXE5BS6DxmB1WpleL8+VP3b\nHy/OJiMjk8AAf9zd3YvLQkKCMZstZGZlExriXDMaUtLSCQ8L4aeNv/PBV9+Qm59PbLs2DOnRrcRU\n/Q++/IZb27WhZkR1A9OWr5STJwkPC71YN19+falu2jKkZ7eLr4WGsPjjz/jxtw34+njT/d5O3Bfb\nwejY5SY4IIDzefnkXbiA96UvOdMyMgE4fe4cVUJCOJ6WRt+x47AWWRnUtQvhoSFGRhYDXbUzNGrU\nqPLKYRezXnv14oXvs+cxfcFbjB42yOhIDmvu9CmcOpXBhCnTmDprDmOevrb/3///mj5/MQePxPPh\nojeZOnch36/5ldtvac+sia8Sd+gwb8yZT1BgAB1uvN7oqOXObDYzbtJknn9mJP5+fkbHMdyPRzaR\ndi6DwS27YgXWJm5lc/JeWlVrSOrZdLafOEChtQgAc5EFN5NriZ93c3GhsKjIgOT2ZzabmThnPs8M\nGoCfr+8/bhNaKZilc6azK24/s5e8R0S1qtx6Q7tyTuoY8vLzcXf3KFHmcem52VxgRCRD5ebnk3ry\nFJ+vXsPzgx7nfG4eU5e8R1FRESP79gYgLSODP3buYsmkccaGLWe5efmknkzn81U/8fygAZzPy2Pa\nkncpLCokNy+fH9dv4NZ2bZgy5mkOxicw/e33CfL356bWLY2OXi4a1q5FWHAwM5Z+yNN9enE+L4/3\nv/kOAIulEIBKAQEseeUl9hw+yvxPPqV65crc7ETXv8pfrtoZ2r17N7fddhsrVlz5zW/Pno6/4lhM\ndF1iqEtefj7jps7iqcEDcHN1LfsHnVDDBvWhQX1G5+fx8oRJPDNiuMMukmFvU+cu5PNvf2D6+Jep\nXSMKV1dXAvz9eOWZi9fQNahXl0NHE/jsm++dsjO0cMm71IiK4vZL862d2feHN7IlZR89Gt1JmG8w\nAF1ibuW7Qxv44cgmKnkHcH1kE/6TtAsANxdXLNbCEvuwFBXh+19/AFcUS1Z8Qo3q1Yi9ynHi5+ND\ndO1aRNeuxdHE43z6zfdO2xny9PS4otNTcOm5l5eXEZEM5eriQm5eHuOfHErVymEAPNmnBxPmLy7u\nDP36nz+JqBpOTJ3aRkYtd66urhfrZsSw4roZ3qcn4+cupFlMA/x9fXl+0ABMJhP1a9XkcOJxvvzp\nZ6fpDLm7uTFx2BDGL36Le54chZ+PN4Me6sL+hAR8vS8eS34+PtT18aFuZCTxKSl88cuv6gzZoALO\nkrt6Zygk5OKQYXh4+DWz/PSpjEziDh3hlhvaFpfVrhGF2WLh/PlcAgP8DUznWNJPnSLuwEE63NS+\nuKx2rVqYzZfqKjDAwHTlz2q1Mm7qTFb98htTxr7IzddfbENhIZWwFhWVmDJaMzKCLdt3GhXVUD/8\n9DOZmZm07XDxQmWz2UJRURHtbu3IH2vXGJyufFitVr46sJY96Ud45Lo7qB9as/i1plWiaVolmnMF\nufh5+PBnyl6CvC4eSwEefpy7kFtiX+cKcqniWzGnP61et4HM7Gxueejiql+X20qHh3rw9ow3yM3L\no0lMg+Lta0VFsCtuv1FxDVc5LIwzZ85isViKv4zKzMzCw92dwADnOh/DxVFDV1fX4j/2AaKqVaWg\noIDsM2cIDgjgj5276dCmlYEpjREaHPTPdWM2E1YpGCh5jVlUtaps27vPgKTGqRcVyYeTXuP02bP4\n+fiQfDIdk4uJoiIre48cpVHdv64RqlmtKnuOHDEwrRjpqj2cZs0uXkPx+++/06FDhxIPRxV/PInR\n4yaSnZNTXBZ36DDBgYHqCP2X+IRjPPXci2SfPl1ctm//AYKDgpyuIwQwY8FbrP51HTNeG8ut7f/6\nJrvJdTEcTkiksPCvb/SPJh6jWrhzrnj1/uJ5fPnxMj5f/gGfL/+Ahzrfx3UxDfj8ww+MjlZufjzy\nO3vSj9Cj0Z3EhNUqLk/ITuXjvasB8PO4ePHygYxEagdfvJYhIrAyx3JOFG9fUGjmxNkMIgMqZlta\nPGUiHy94k+XzZrN83mw633UHDaPr8eG8WaxZv5E35i8qsf3+I0epGVlxF5IoS4PoaNzd3dm5+69r\nWrft2EXDmAbXzBeS/6ZG0XUpLCwkPim5uCwhOQUfb28CL03RjTtylOZ/W6XQWTSKrndl3SQl4+vt\nTeP60Rw5lkRhYVGJ16qGhf3Triqks7m5DH9jKqfPniXI3x83V1c27txJdFQUv27ZyswPl5fY/mDi\nMWpUrVrK3uTvXEwmuzwMfU+2bOTr68uWLVtISUkhNTWV1NRUe+f6P2vZtDG1a9bg1SkzSTiexPr/\nbGbe2x/weO+Kfx+P/1WrFs2oU7sWL4+fRHziMdZt2MSc+Yt44rG+Rkcrd7vj9rPii5UM7teHmHp1\nyczKLn7cFdsBNzc3JkyfzbHkFL5f8wvfrf6Zbp3vMzq2IcKrVCGyevXiR4C/P16enkRUr2Z0tHKR\nlJPGH8m7ia3Zmmr+YZwryC1+hPoEcSQriT+S95Cdd4Zf4v8k+Uw67SIaA9CyagzJZ9JZd2w76eez\nWXngN4K8/KhTqWJ2AMLDwoioGl788PfzxdPDg4iq4dx/x+0Xb8C6dDlJqSf4+Ovv+Hn9Jvo/0tXo\n2Ibx8vLkvns6MWnKdPbui2Ptug0sXf4RvXt0MzqaISLDw2nfqgUTF7zFwfgEdu4/wMIVn/DA7bfi\n4uLCifRT5ObnUyvSeRZOuCyyajjtW7Vk4vzFV9TN7Te0w83VlcmLlpB0Io1V6zfyw7oNPHSX89wy\nw9/Hh/yCAuZ/+jmpp06xdstWln33A/3uu5e7299ISvop3v7qa5JPpvP5z7/w65at9Lm0zL84H5su\nCjlz5gw//PBDibJXX33VLoH+f7m5uvLmpHG88eYC+g5/Gl8fb3p17Ux3J/3D9Wrc3NyYP2s6k6bO\noPdjT+Dj48OjPbvTs5vz/THyy/qNmEwm5r79HnPffg+geCnbLWu+Y9H0yUyePY/uA4YSGlKJF0YN\n46Z2bQxOLUbYdyoeMLEmfjNr4jdfKrUCJsZ1GET3Rnfy45FNrInfTFW/UPo1u48Az4uLBwR5+dOj\n0Z38cGQT649tIzIgnF6NnfMDuHp4Fd6c8Cozl7zLiq++oXp4Faa89BzRtWuV/cMV2OhRI5g4ZRoD\nho3A19eXIQMfp2PsrUbHMsy4J4cw6/1lPPnaZFxdXLm7w00MvtQ5zDpzBhMQ4KQLuYwbMYRZ7y3j\nyQmvX6qbmxnc8xHcXF2Z++qLTFvyHn2efYHQ4CBGD+jHDX9bMdUZjB/8BNM++JD+r04grFIwz/Xt\nw/WXlq2f/tRI5n7yKZ/+tIaqYaG8NnQwdSMjDU58baiI1wyZrP99R9W/GTZsWImbrrq5uWGxWPDw\n8GDWrFk2/YLzyUf/naQVjG9EHQpyMoyO4ZA8Ai/e9Cw3NcHgJI7Jp1ottZ1SXG47nw6ebXASx9Rt\n0SjOHHXea3KuJqBODAVnMo2O4ZA8Ai5eP5y1a4vBSRxTpaatydq91egYDqlSk4vXc53cuM7gJI6p\nSvtbjI7wP1v55Jt22W/nuSPssl9bXHVkaPbs2VitVt555x06duxI3bp1SUhI4KeffiqvfCIiIiIi\n4gCc7qarl2/8dvLkSerWrQtArVq1HPqaIREREREREVvYdM2Qj48PH3/8MXXr1uXQoUMEBQXZO5eI\niIiIiIhd2bSa3IgRFy/k3L59O0FBQQwfPtzeuURERERExIGYTPZ5GMmmkSEvLy/uu0+rsYmIiIiI\nSMVhU2dIREREREScW0VcQMH5bmktIiIiIiKCRoZERERERMQGFXBgSJ0hEREREREpm6bJiYiIiIiI\nVBDqDImIiIiIiFNSZ0hERERERJySrhkSEREREZEyVcBLhjQyJCIiIiIizkkjQyIiIiIiUqaKuJqc\nOkMiIiIiIlKmCtgXwmS1Wq1GhxAREREREce2avQCu+z3rmlD7bJfW9h9ZOjM0f32/hXXpIA6MRSc\nyTQ6hkPyCAgB4PCyLwxO4pjq9XlIbacUl9tO3qkUg5M4Ju+w6rx27ytGx3BIr3z3mo6rUlw+rvIz\n0wxO4pi8QsLVdkpxue3kHNpjcBLHFBjd2OgI/zOXCjg0pAUURERERETEKemaIRERERERKVMFHBjS\nyJCIiIiIiDgndYZERERERMQpaZqciIiIiIiUqSLeZ0gjQyIiIiIi4pQ0MiQiIiIiImWqgANDGhkS\nERERERHnpJEhEREREREpk8ml4g0N2dQZysvL4+uvvyYrK4uWLVtSo0YNwsPD7Z1NRERERETEbmya\nJrdw4UIqV65MWloaQUFBLFy40N65RERERETEgZhM9nkYyabO0NmzZ4mNjcXV1ZX69etjtVrtnUtE\nRERERMSubF5AISUlBYDMzExcXV3tFkhERERERKQ82HTNUP/+/VmwYAEpKSnMnDmTxx9/3N65RERE\nRETEgVTEm67a1BmKiopi0qRJ9s4iIiIiIiJSbmzqDK1bt46VK1diNpuLy+bNm2e3UP+WiXPmk3Ii\njYVvvAbAhs1bWPDBhySdOEHNiAiG9u3NDa1aGJzSWGazmUlTZ7Dm17V4enjSp+cj9O/Ty+hY5c5s\nsTDqnfk8cce9NK1VB4B533/F6h1bMZnAar14gd+Ajvdwf5sbeHzuNNJzTl+xn47NWjLi3i7lHd8Q\najtXKigooOeAITw3chhtWl48t7w2dSZffvs9JpMJq9WKyWTi2SeH0vPhittOQqqHctfge6heP5Lc\nM7ls/X4zf3y1CQC/YD/uHno/tZrVJvdMLr9/voFtP24p/tnqDSK584m7qRxVmYzkDH5a8gPH9x0z\n6q2UOx1XV3c6J4fXp8/mjy1b8fX1oVe3rvR+5GGjYzkEtZ2Sflq/kVemzylx7r25bSumvvgcaemn\nmDR3IbsPHCI8LJRRj/fl+pbNjY58TaiAA0O2dYa+/vprxowZQ0hIiL3z/Gv+3LmLb376mZaNGwEQ\nfzyJFyZPY9TA/tzQqgW/bvoPo1+bzGdvzaNalSoGpzXO9Dlz2bMvjncWzCXtZDovjB1P1fBw7up4\nm9HRyo3ZYmHqVx+TlJFeovx4RjqPd+xEh0bNist8PD0BmPX4MIqsRcXlO+KPMPf7L7m3VbvyCe0A\n1HZKKigo4PlxE4lPLPmHe3ziMZ4ePph77ri9uMzXx7e845Ubk4sLPcb3IWFXPN/P+4bQyFAeMVqb\ndAAAIABJREFUHN2Ns5ln2Ld+D91f7c350+dZMnIhVetW4/5RD5KZmknirnj8QwLoNeFRtny3mS+n\nfErzO1vS7eWezH9iNnln84x+a+VCx9XVjXr+JQouFLB4zgzO5+by8oTXcXFxoefDDxkdzXBqOyXF\nH0/i1uvbMmboQC6v++Xh4Q7AMxOnUKdGJB/MnMK6zX8yZvJ0Plkwm6qVwwxMLEaxaQGFypUrEx4e\njru7e/HDkeXnX2Dy3IU0axhTXJaekcEj999D13s6Ua1KFXp36Yy3lxd7DxwyMKmx8vLz+XLlt4x5\neiQN6kfT4eb29H+0Fx999rnR0cpNUkY6z7y3kJOns694LTnjFHXDqxPk61f88HC72PYDfHyKy7w9\nPFm69ie63xRL7fBq5f0WDKG2U1J84jH6DBpOyom0K15LOHaMhvWjqRQcXPzw9PQwIGX5CAjxJ+Vg\nMqsWfsfpk9kc2XqYhJ1HqdG4JnVa1iM4vBJfTv2UrNRM9q3fw84124loEAlAm/vbkRafxtqlP3P6\nZDZrl/5Mdlo24XV0XAnsP3iIXXv28caEscTUj6ZV82aMGjaYd5etMDqa4dR2rpRwPJm6NWsQHBhI\npaCLDz8fH7bs2kNS6gleHDaYmpHV6dv1QZrE1Oebn34xOvI1wWQy2eVhJJs6Q56enrz++uusWLGi\n+OHI5n+wjJZNGtO88XXFZe1aNOfJx/oCYCks5OvVazBbzDRuUN+omIY7eOgwZouF5k2bFJc1b9qU\nfXH7nWb59D3HEmhaqw7T+w/m7285+9xZzuXnUT0ktMx9fP3nJlxMJh5s196OSR2L2k5J23buom3L\n5ixdNK/E+8/MyuLM2XPUjIo0MF35yjmVw1fTPqPQUghAREwUUdfVIGFnPDWb1CZxTwIXci8Ub//j\nwu/Y+Mk6AGo2rsWB3/eV2N87Ty0iYefR8nsDBtJxdXXJKakE+PsTFRFRXFa/bh0ys7I4kXbSwGTG\nU9u5UkJSMjUirvwiZd+hw9SvXQsvL8/isqYNG7DnoPN+Of6/qIj3GbJpmlzz5tfOPMrd+w/w66b/\n8MnCN1n2xcorXk9MTqH7kBFYrVaG9+tD1SqVDUjpGDIyMgkM8C8x0hcSEozZbCEzK5vQkEoGpisf\nd7ds+4/lSRnpuJhcWPbbGrYdPUSAjy+d29zIbU1LXmNmLrSwcvNGHu1wB+6uNh1OFYLaTkkPd77/\nH8vjE4/h4uLCvCXvsumPPwkKDKT3I125v9Od5ZzQGCPffxa/YH8ObznI/t/jaHRLE3JO5dCh9200\niW3GhdwL/PH17+xasx2A4KqVKMg38+CzD1OzaS2yUjNZ885qUg8lG/xOyoeOq6sLqRTMufPnyc3L\nw8fbG4DUtIujsadzcqga7rxT3tV2SrJYLCSnnWTDn1tZtOwjrMBtN17PEz27kZGVTWil4BLbVwoK\nJD0j05iwYrir/vV29OhR6tSpQ3Bw8NU2cxhms5mJc+bzzKAB+Pn+85z80ErBLJ0znV1x+5m95D0i\nqlXl1huc5zqPv8vLz8fdveR0HY9Lz83mAiMiOYykjFNggjrh1bi/zQ3sORbPvB9W4uXhwY0xjYq3\n2xi3B4DYJtfOFwb/BrUd28QnHsdkMtGgXj16PfwQW7bvYOK0mfh4e3N7h5uNjmd3n7y2HP9K/nQa\nej93DOyEh7cHTWKbsX/TPj6duILwutXoNPhe8nJyOfTnATy8PYnt25GNn/zG719soHFsM3pP7MeC\nwXM4l3XW6Ldjdzqurq7xdQ2pUjmMiVNn8PLopzl7/jyL3nkfoMQCT85Ibaek46knKCoqwsfbmykv\njib5xElmLnmX3Lw8LhSY8fivyz083N0pcPI25Myu2hnas2cPderUYdOmTVe81rRpU7uF+r9asuIT\nalSvRuyN15e6jZ+PD9G1axFduxZHE4/z6TffO21nyNPT44qTZMGl515eXkZEchj3tGrHLY2a4ud1\n8dvHmpXDSc3K5Idtm0t0hjbE7eHGBo2KryVyFmo7tnmkywN06ngbAf5+ANStXYvjySl89tU3TtEZ\nSjt6grSjJ3B/dxUPPNWF43HHyT+Xx/dzv774evwJqtQKp+XdrTn05wGKCos4suUgf37zBwAn31lF\n7WZ1aHJrM37/YoORb6Vc6Li6Ond3d2a+/hpjxk6g/Z334u/nx8ghT7B3/wF8S/kC1Fmo7ZRUOyqS\nn5a/S4DfpXNvzRpYrUW8PG02ne/qyPnc3BLbF5jNeHl6/tOuxAlctTPUuXNnAB5++OESFze5urpi\nsVhwc3OsaUGr120gMzubWx7qDoDZbKGoqIgOD/Xg7RlvkJuXR5OYBsXb14qKYFfcfqPiGq5yWBhn\nzpwt8f8yMzMLD3d3AgMCDE5nvMsdocsiQ8PYEX+k+Lm50MLuxHhefrh3eUcznNqO7S53hC6rVSOK\nP7ZsMyiN/flV8qdq3Woc/vNgcVnG8VO4urmSczIba1FRie0zUzKo1aQ2AGezzpKZUnKqSlZqJgFh\ngfYP7gB0XJWtQXQ9vv54GVnZpwnw9+N4cjIuJpNTT3kHtZ1/crkjdFnNyAgshYWEVQrmcEJiidcy\ns09fMXVOSmH0BT52YNMCClOmTGH06NHMnj2bMWPG8OKLLzJs2DDWr19v73z/k8VTJvLxgjdZPm82\ny+fNpvNdd9Awuh4fzpvFmvUbeWP+ohLb7z9ylJqREaXsreJrEB2Nu7s7O3fvKS7btmMXDWMa4OJi\nU9OosN75+QfGf7K0RNmRE6lEhP617Oax9JOYCy00jKpR3vEMp7Zjm5nzFjLiuRdLlO0/dLhCL6gQ\nGhnGwy/2wDvAp7isar1qnM85T/KBJCrXDMfk8teHaVhUZU6nX1zNMeVAEuF1qpbcX0QYOelXrvZY\nEem4urozZ8/Sb8hwsk+fplJwEG5ubqxdv4kG9aPx8fEpewcVmNpOSWv/s5m7+gzAUlhYXHbwaDz+\nfr40qh/NofgE8i/8tZDLrrgDNKpfz4io4gBsXlp7zpw5TJw4kTlz5lCnTh1mzJjBqlWr7J3vfxIe\nFkZE1fDih7+fL54eHkRUDef+O26/eAPWpctJSj3Bx19/x8/rN9H/ka5GxzaMl5cn993TiUlTprN3\nXxxr121g6fKP6N2jm9HRDNemXgO2Hz3Et1t+Jy07i2+3/Iff9u7koev/mtp0LP0klQODnW6KHKjt\n2OrmG2/g981b+OjzL0lOSeXjL77ih9U/069Xd6Oj2c2xPYmcOp7O/aO6EBIRSr3W0cT27ciGT9ax\nb8NeiiyF3PtkZypVq0TjDk1pGtuMrd9vBmDz178T3aY+re9rR3B4MLf1uwP/EH/2rN1l8LsqHzqu\nri7A35/8/AvMmLuA5JRUfvp1LUs+WMag/n2NjmY4tZ2SWjRqCMDkeYs4nnqCjVu2Mfe9D+nT5QFa\nNGpIeFgY42fNI/54Eh98/hX7Dh2m89/uBSelq4hLa9s0zy0nJ4eAS8Osfn5+5OTk4OfnZ3j4/0X1\n8Cq8OeFVZi55lxVffUP18CpMeek5omvXMjqaoUaPGsHEKdMYMGwEvr6+DBn4OB1jbzU6liH+3pwb\n16jN6Acf4aP1v/L+L6upEhzMcw92JyYiqnib7PPn8Pf2/oc9OQe1nX/29/Niq+ZNmTzuZRa/t5Q5\nC5dQvVpV3hj/Ck0bXXeVPVzbrEVFfDx+GZ0G38dj05/gQl4Bf6z8na3fXezwfPjye3Qach9PvDmM\ns9nn+GHhdxzZehiAE0dS+XTSR9ze/05u69uR9GMnWT52Keeyzxn5lsqVjqurm/baOCZMmc7DfR+j\nSlgYrz4/mpuvcp2wM1Hb+Uugvz9vTniZ2W+/T99Rz+Hr68NDd99J364PAjDjleeZ+OYC+j41hoiq\n4Ux96TnCdcNVm1xDf/rbzGS1YQH6t99+m3PnzhEdHc2hQ4fw9/cnJiaGjRs38txzz131Z88cdd5r\ncq4moE4MBWe0jOM/8QgIAeDwsi8MTuKY6vV5SG2nFJfbTt6pFIOTOCbvsOq8du8rRsdwSK9895qO\nq1JcPq7yM6+8qbCAV0i42k4pLrednEN7ytjSOQVGNzY6wv9s02vv2GW/N77yuF32awubRoYGDBjA\n1q1bSU5O5uabb6ZFixakpqbSsmVLe+cTEREREREH8PdrPisKm64ZysvLo6CggODgYM6ePcu6deuo\nVq0anlqGUERERERErlE2jQxNnTqV4OBgQkND7Z1HRERERESkXNjUGbJarYwYMcLeWURERERExEFV\nxAUUbJomFxUVxeHDhzGbzVgsFiwWi71ziYiIiIiI2JVNI0P79+9n+/btnD17tnhJ7Xnz5tk7m4iI\niIiIOIhr6bY6trJpZKh///54enoSHBzMTTfdRJcuXeydS0RERERExK5s6gx98sknjB8/nuDgYLp0\n6cLq1avtnUtERERERByIyWSfh5Fs6gyZTCb8/PwA8PDwwNvb266hRERERERE7M2ma4bCw8NZsWIF\n586dY+XKlVpiW0RERETEyTjtNUMDBw4kNDSU+vXr4+npyeDBg+2dS0RERERExK5sGhlydXXljjvu\nsHcWERERERGRcmNTZ0hERERERJxbBZwlZ9s0ORERERERkYpGI0MiIiIiIlKmiriAgslqtVqNDiEi\nIiIiIo5t67QP7LLfVqP72mW/ttDIkIiIiIiIlK0CXmBj985QwZlMe/+Ka5JHQIjqphQeASEAXDid\nbnASx+QZVJm89CSjYzgk78qRABTkZBicxDF5BIbqvFMKj4AQzhzZZ3QMhxRQ9zpAn+el0ed56S5/\nnqt+/tnl+rmWGDVNrrCwkIULF3Lq1CksFgsPPvggERERLFiwAJPJRGRkJAMGDADg559/5pdffsHV\n1ZUuXbrQokWLq+5bI0MiIiIiIuKwNmzYgL+/P8OHD+f8+fOMHj2amjVr0qNHD2JiYliyZAlbtmyh\nXr16rFq1iilTpnDhwgXGjh1LkyZNcHMrvcujzpCIiIiIiDis66+/nnbt2gFQVFSEq6srCQkJxMTE\nANC8eXN27dqFyWSiQYMGuLq64uPjQ3h4OMePH6d27dql7rsCzvwTEREREZGKwtPTEy8vL/Ly8pg5\ncybdu3fn72vAXX4tPz8fHx+fEuW5ublX3bc6QyIiIiIiUiaTyT4PW2RkZDBhwgRuueUWbrzxxhLX\nL13uBHl7e5fo/Px35+ifqDMkIiIiIiIO6/Tp00yaNIlevXrRoUMHAGrVqkVcXBwAO3bsICYmhjp1\n6nDw4EEsFgu5ubmkpKQQFRV11X3rmiERERERESmTUavJrVy5ktzcXL744gu++OILAPr378+7775L\nYWEh1atXp127dphMJjp16sQrr7wCQI8ePa66eAKoMyQiIiIiIjYwqC9Ev3796Nev3xXl48aNu6Is\nNjaW2NhYm/etaXIiIiIiIuKUNDIkIiIiIiJlM2poyI40MiQiIiIiIk5JnSEREREREXFK6gyJiIiI\niIhT0jVDIiIiIiJSJpNLxbtmSJ0hEREREREpUwVcP8E5psmZzWbGTXqDG2+7k9hO9/PesuVGR3IY\nqpvS/fjTzzRtdzPNrr+l+L9PjXnJ6FiGKygooGvfAfy5bUdx2ZGERB5/8mluuOM+Huz9GD+s+cXA\nhMZLOHaMgcNH0rbD7dz1QFfe/3CF0ZEcUkFBAQ92783mLVuNjuIQJr65gMHPjy1+/vrchbS59yHa\n3te1+L8ff/O9gQmNp8+s0qluSqe6kauxeWQoJycHs9lc/Dw0NNQugexh+py57NkXxzsL5pJ2Mp0X\nxo6nang4d3W8zehohlPdlO5ofCK3dbiZl8c8g9V6sczTw8PYUAYrKCjg+fGvE594vLjMbDYzcszL\ndGh/AxNeGM2WHbsY+/o0oiKq0yimgYFpjWGxWBgy8lnatW7J2BeeIyHxGGNeHkflsDDuvrOj0fEc\nRkFBAc+99CrxCYlGR3EIf+7czTc//ULLxtcVlyUkJTPq8X7cdetNxWW+Pj4GpHMc+swqneqmdKqb\nf4+pAg4N2dQZevvtt9mxYwfBwcFYrVZMJhMTJ060d7Z/RV5+Pl+u/Jb5s6fToH40DepH0//RXnz0\n2edOfxCobq7uaEIi9evVoVJwsNFRHEJ84jFemPD6FeVHE49x4mQ6Qx7vi5+vL9WrVeXTr75h645d\nTtkZSj+VQZNGDXlx9NN4eHgQWb067dq0Yuv2HeoMXRKfkMiYl181OobDyM+/wOR5i2jWMKZEeUJS\nMkMerU2loCCDkjkWfWaVTnVTOtWNlMWmaXJHjhxh7ty5TJw4kUmTJl0zHSGAg4cOY7ZYaN60SXFZ\n86ZN2Re3H+vlr/udlOrm6uITE6lZI8roGA5j287dtG3ZgqUL55ZoH4EB/gB89d2PWK1Wdu2NI/F4\nEg3rRxsV1VDVqoYzdeJ4PC6NIu7YtZttO3bRrk1rg5M5jq3bd9C2dSs+fPctnWuA+Us/pGWTRjRv\n3LC4LCMrm7PnzlOjenUDkzkWfWaVTnVTOtXNv8tkss/DSDaNDIWHh2M2m/H09LR3nn9dRkYmgQH+\nuLu7F5eFhARjNlvIzMomNKSSgemMpbopndliISk5hXUbNzFv0dtYsXJHbAeGPvF4ifpyJg93vu8f\ny6tWqcLwgf15c/E7zFm4hCKrlYF9e9GmZfNyTuh4bru3MxkZmdzc/gY6xnYwOo7D6PbQg0ZHcBi7\n9x/k101/8MmC2Sz78uvi8oSkZFxcXFi4bAX/2bqdwIAAena+j3tvv9XAtMbSZ1bpVDelU91IWWzq\nDGVkZDB06FDCw8MBrrlpcu7uJa/z8Lj03GwuMCKSw1DdlO748SSKiorw8fFh5pSJJKek8MaMOeTm\n5fHCs08ZHc+hFBYWcjw5hS733c2D99zF/kNHmD53IfXr1iX25huNjmeoudOncOpUBhOmTGPqrDmM\neXqU0ZHEgZjNZia+uYBnnngcP1/fEq8lJiVjAurXrkX3++9h2+69TJ63CB9vL2JvvN6YwAbTZ1bp\nVDelU91IWWzqDI0cOdLeOezG09PjisZecOm5l5eXEZEchuqmdHVq12L96u8IuDQFLLpuHYqKrDw/\ndgJjnh6Ji4tTLMRok29XrWHX3ji++vBdABpE1yMtPZ2F77zv9J2hhg3qQ4P6jM7P4+UJk3hmxHDc\n3HRHA7loyYpPqVG9KrE3trvitYfv7cRdHW7G3+9iJ6luzRokpZ7g8+9XO21nSJ9ZpVPdlE518y8z\nek6bHdj0qbxu3boryrp27fqvh7GHymFhnDlzFovFUvxHSGZmFh7u7gQGBBiczliqm6u73BG6rHbN\nGlgsFrKzTxOiYfVicQcPUbd2zRJlDetHs/Tjz4wJZLD0U6eIO3CQDje1Ly6rXasWZrOF8+dzCQzU\nsSUXrV63kczT2dzStScAZrOFoqIiOnTtxW+fLy/uCF1WMzKCzTt2GRHVIegzq3Sqm9KpbqQsNn29\nHRgYSGBgIAEBAWRmZpKRkWHvXP+aBtHRuLu7s3P3nuKybTt20TCmgdN/u6+6Kd0vv63n1k4PYLFY\nisv2HzyEv5+fOkL/JSw0hIRjx0uUxSceI7JaNYMSGSs+4RhPPfci2adPF5ft23+A4KAgdYSkhMVT\nXuPj+bNZPncmy+fOpPNdHWkYXZcP585g9tvv89T4kqs3HjgaT81I511QQZ9ZpVPdlE518+8yuZjs\n8jCSTa2gY8eOdOzYkTvuuINBgwaRlZVl71z/Gi8vT+67pxOTpkxn77441q7bwNLlH9G7RzejoxlO\ndVO6Vs2bATBh8jSOHU9i/cbfmTVvIf379DQ4meO5547bOZGWzox5i0hOTeWXdRt4b/kn9H7k2hg9\n/re1atGMOrVr8fL4ScQnHmPdhk3Mmb+IJx7ra3Q0cTDhYaFEVA0vfvj7+eLp4UFE1XBuatua/2zb\nwSfffE/yiTQ+/fYHVq1dTx8nXnxCn1mlU92UTnUjZbFpmlxqamrxv7Ozszl16pTdAtnD6FEjmDhl\nGgOGjcDX15chAx+nY6zzrsjzd6qbfxYYGMCiOTOYNmcu3fsOwM/Pj24PPchjj/YyOppD+PtN16qF\nV2Hx7KnMmv8W3R8bTGhICCMGP879ne4wMKFx3NzcmD9rOpOmzqD3Y0/g4+PDoz2707Obc3YOy1IR\nb+D3b2jZ+DomPfcUS1Z8ytz3l1E9vAqTnnuaJjH1jY5mKH1mlU51UzrVzb+nIp6yTVYbFlkfP358\n8b89PDzo1KkTzZo1s+kXFJzJ/L+nq8A8AkJUN6XwCAgB4MLpdIOTOCbPoMrkpScZHcMheVeOBKAg\n59qZyluePAJDdd4phUdACGeO7DM6hkMKqHsdoM/z0ujzvHSXP89VP//scv1cS+KWfGyX/TYc2N0u\n+7WFzSNDOTk5BAQEcPbsWQ4cOEBgYCADBgygSZMmZe9ARERERESubRVwaMimzlBMTAzdunWjWrVq\npKWl8fnnn9O1a1fmzp2rzpCIiIiIiFyTbFpAISsri2qXVoYKDw8nIyOD8PBwrcIhIiIiIiLXLJtG\nhoKCgli+fDn169fn4MGDBAUFsXv3bt08UERERETESVTAWXK2jQwNHz6c4OBgduzYQUhICEOHDsXL\ny4uRI0faO5+IiIiIiIhd2DS04+Hhwd13312iLDo62i6BRERERETE8Rh9g1R70EU/IiIiIiLilHTR\nj4iIiIiIlKki3ihbnSERERERESlbxesLaZqciIiIiIg4J3WGRERERETEKakzJCIiIiIiTknXDImI\niIiISJkq4gIKGhkSERERERGnpJEhEREREREpU0UcGVJnSEREREREylYB55SZrFar1egQIiIiIiLi\n2A4v/8Iu+63X6yG77NcWGhkSEREREZEyaZrc/8GF0+n2/hXXJM+gyhScyTQ6hkPyCAgBIO9UisFJ\nHJN3WHW1nVJcbjtHP/rK4CSOqU6PB3VOLoXOyaW7fFyNin3G4CSOafavM9R2SnG57eRnphmcxDF5\nhYQbHUGokDP/REREREREyqbOkIiIiIiIOCVdMyQiIiIiImXSNUMiIiIiIuKcKl5fyLZpckuXLrV3\nDhERERERkXJl08hQcnIy58+fx9fX1955RERERETEAZlcKt7QkM2doccee4yAgABMJhMmk4nFixfb\nO5uIiIiIiIjd2NQZWrBggb1ziIiIiIiII3PWBRSSkpJYsmQJ58+f56abbiIyMpKWLVvaO5uIiIiI\niIjd2LSAwnvvvcfQoUMJCAggNjaWzz77zN65RERERERE7MrmpbXDw8MBCAgIwNvb226BRERERETE\n8VTAWXK2jQz5+fmxZs0aLly4wKZNm/Dx8bF3LhEREREREbuyqTM0ePBg0tPT8ff35+jRowwZMsTe\nuURERERExIFcXlX6334YyaZpcl5eXrRv3x6z2QxAWloadevWtWswERERERERe7KpMzR58mQsFgu+\nvr5YrVZMJhPPPvusvbP9aywWC7PmLeS7H1cD0PG2Wxnz9Ejc3Wy+ZKrCKygo4JFHH+P5Z0bRtnUr\no+MYrqCggJ4DhvDcyGG0adkCgFMZmUyaMZvNW7cRFBhI/17d6fbgAwYnNY7ZbGbS1Bms+XUtnh6e\n9On5CP379DI6VrkzWyyMeGsug+66j2a1S35JZCks5MnFb3JTw8b07HB7cfm6Pbv4aP0vnMo5zXVR\ntRhy9/1UrRRS3tENo3Ny6Zz9uKocGcZDT3ahRsMozuecZ+PXv7P2098AqF6nGl2e7Ez1OtU5lXKK\n797+gYNbDxX/bPNbm3FHn44EVw4ifk8CX8z9iszUTIPeSflz9rZTmoKCAno89gTPjRpB21YtSryW\neOw43fo+zp+/rTEo3TXIWW+6WlBQwPjx4+2dxW5mvLmA3zZs5M3pUwAY88p4ggPfZ9igAQYncwwF\nBQU899KrxCckGh3FIRQUFPD8uInEJx4rLrNarYx47kUqBQfz8TuL2X/oMGNfn0KNyMgrTq7OYvqc\nuezZF8c7C+aSdjKdF8aOp2p4OHd1vM3oaOXGbLEw5fOPSDqV/o+vf7Tulyte23H0MNO+/IQBd95N\n63oN+H7LHzz//lssHv4MXh4e5RHbcDonl86ZjysXFxeemDyQwzsO8+msz6gcVZlHX+pNTkYOB7ce\nYuj0wez9fR+fzPiMmg1r8tj4fswdNZ/kwylEt6hH7xd78vXCb4nbvJ/299/A8BlDeL3fFMwXzEa/\ntXLhzG2nNAUFBYx5dUKJz/PLUk+k8eRzL2C2WAxIdu0yekqbPdh0zVBMTAw7d+4kIyOj+HGtOHvu\nHJ999TXjXhpD08bX0bTxdQx94jHiDhw0OppDiE9IpFf/gaSkphodxSHEJx6jz6DhpJxIK1G+8Y/N\nJKWm8sb4l6kRFcldt8fywN2d2L13n0FJjZWXn8+XK79lzNMjaVA/mg43t6f/o7346LPPjY5Wbo6f\nSuept+eTdjrrH19PSDvBTzu2EhEaVqL8m82/c3OjJnRu157qIaE8cde9eHl48tueneUR23A6J5fO\n2Y+rwLBAjh84zudvfknmiSz2bz7Aoe2HqdOkNq3vaEl+bj4fz/iU9KRT/Ll6C1vXbKPDwx0AuLnL\nTexYu4v1X24gIyWDlQu/4ULeBVre7hxfVjl72/kn8YmJ9B44hJTUE1e89vPadfR8fBDeXl4GJBNH\nY1NnKCcnhw8++IBZs2Yxa9YsZs+ebe9c/5odO3fj7eVF21Z/3ST2/rvvYv6saQamchxbt++gbetW\nfPjuW1itVqPjGG7bzl20bdmcpYvmlaiPLdt30rp5M/z9/IrLXnp2FAP79TEipuEOHjqM2WKhedMm\nxWXNmzZlX9x+p2lHexLjaVarLjMfH8p/v+WioiJmf/MFj3XshL93ydU307KziImIKlFWq0o4+5Ou\n/OayItI5uXTOflxln8xm6cQPKTQXAlDruprUblyLQzsOE1IthKRDyViL/qqHlKOp1Gy1ac/0AAAg\nAElEQVRYA4CQqpVIjEsssb/U+BPUalizvOIbytnbzj/ZumMXbVu1ZNlbC66ogw3/+YORQ5/gmSeH\nGpROHIlN0+RSU1OZNWuWvbPYRVJKKlWrVuGH1T+z5P2l5OXl0TG2AyOGDtL8dKDbQw8aHcGhPNz5\n/n8sT05JJbxKFeYveZdvf/wJPz9fenfrSud7O5VzQseQkZFJYIA/7u7uxWUhIcGYzRYys7IJDalk\nYLrycU/rdqW+9sXv6wn08eXWJs35YevmEq8F+fmRcSanRFl6zmkCfXztktPR6JxcOh1Xfxn3ySsE\nVApg3x9x7Fq3myqRlYmMjiyxTaXwSvgFXjxuzmafIygsqOTrVYI5l3O+3DIbSW3nSle7pnf8i2MA\n2Lx1W3nFEQdm08hQVFQUhw4dwmw2Y7FYsFxD8yvP5+aSknKCjz//gnEvPscrzz/Lml/XMXveQqOj\nyTXkfG4e3636iYysLGa/8Ro9H+7C5Jlz+G3jJqOjGSIvPx9395LXt3hcem42FxgRyWEkZ5ziy983\n8OR9//xFwy2NmvLjts3sSYynsKiIVdv+5EhqCpbCwnJOagydk0un4+ovb7/8Lm+/8i6R0RE8OOwB\ndq3fTWR0BO0fuBEXFxdqxETR9s7WuLq7ArB97Q5uuLcdtZvUxuRiot3dbYmIjsDVzdXgd1I+1Hak\n3Jjs9DCQTV/D7d+/n+3btxc/N5lMzJs3z26h/k1urq6cz81l8vixVK9WFYBnRgzlpfGTGD3qSYPT\nybXCzdWVAH9/xj73DCaTiQbR9Th05CifffUtHdrfaHS8cufp6XHFB2zBpedeTj4He843X/DIzbcS\nFhj0j6/f2aI1yRmneGnZO1itVprVrsvtzVpyLj+vnJMaQ+fk0um4+kvy4RQ4nMLXC7+lx5jufL3w\nW5a/8RFdR3bhwWEPkJGaybov1nNL15sB+OP7zVSOrMzQqYMwuZg4tO0Qf67egreft8HvpHyo7Yj8\n39nUGZo+fbq9c9hNWGgIrq6uxR+6ADVrRFFQUEBWdjaVgoMNTCfXirDQEIqsRSVWUakZFcmf23YY\nmMo4lcPCOHPmLBaLBbdLU5syM7PwcHcnMCDA4HTGST99mrjjx4hPS2XZrz8BcMFs4VBKMgdTkhjf\nqz8mk4kBd95D39vuJK/gAgE+vkz+bAVVgpzjXKRzcumc/bgKCAkgMjqCff+JKy5LO5aGm5srXr6e\nbP91B9t/3YFfsB/nss9x4/3Xk3Uyu3jbbxZ9y/dv/4Cnjye5Z3Lp+0ofstKy/+lXVTjO3nak/FTE\n1eRs6gytWbOG/9fefYdHUfVtHP/uJtmEBNJ7o0OCiJRIFWkGH+ABRFSK8CCv1KAEQVBBBBSkKyV0\nEOkg0kXpCIiFIi2EhBrSgRQIkLab7PtHZCGEQFQ2s2R/n+vaS3ZmdnLPccqePTPn7Nmzh5yc+786\nPCvPEL3wfE1yc3O5ePkKVSpVBODS5SvY2dri6OCgcDrxrKhVswbzFn9Lbm4uFhb5t11cuhKNt5en\nwsmUEVCtGlZWVpw8fYagunUAOH7iFDUCA1Cri3X3bankam/P4sEFx2Cb9P0aaviV582XmgGw+fdf\nyM7R0uXlFlhZWpKVk8Opyxf58PUuSkQucXJOLpq5H1ce/h7837h3+OyNcdxNz3/Wx6+6H3du3cGz\nohcvd3qJb8ct507aHQBqNnqOCycuAvm9yWlsNOxZvZeM9AysrK2oWrcqK79cpdj2lCRz33dEyVGV\nwnGGinWE/PjjjwwbNowJEyYYXs8Kfz9fmjdtwmdfTCQiMorjJ04xc+4COr/WXk4Qotj+06ollpYW\njJs8jauxcWzfuZttP+2iy+vmOeiqjY017du1YcLkaYSfjWD/gUMsX7WGHt3eUjqaotRqNV7OLgVe\nGktLypUpg3O5/F9nvZyc2fDrAf68dIG45BtM+n41Xs4uBFWtrnD6kiHn5KKZ+3F16dQlkqKT6P5R\nV9z93KnRMJD/9mnLrpV7uBF7g4AXq9O000s4ezjR5p1X8Q8sz8GNhwBISUyhZZfmVKtXDXc/N3p9\n1pPk+GQij5pHl+3mvu8I8W8Uq2WofPnyuLq6PrMXqi/HjWbyVzPp+94QLC0s6NCuDYMH9lM6lskp\njU2f/8aD5WFrW4b5M6bx5fSZdHmnL64uznwyLJSmjYruUay0Gz5kMOMnT6XPoMHY2dkxsO+7BLds\noXQsRfydQ6dB9Rp0adqSrzevJ1ubQ70q1fmg45vGC2eC5JxcNHM+rvLy8lg4cjFvDO7MB3MGk5WR\nzc/rD/LL5vyOapaOXcZrAzvQ7t02xF+MZ+6H80lPSQfg7G8R7F61l+4juqCx0XDuSCRrJq9TcnNK\nnDnvO08i32+eolJYlip9MTqg37NnDxs3bsTDw8MwbcyYMcX6A9k3Hz0yu7mzdnQnJz1F6RgmSWPv\nAkDmjXiFk5imMm4+su8U4d6+c2nNJoWTmKbK3TrJObkIck4u2r3jakjLYQonMU0z9k2XfacI9/ad\nrJSkJyxpnmxcnr1b7WO3/2SU9fq1U26okmI/M/TBBx9gZ2ce42AIIYQQQgghCiqNrWzFqgy5uLhQ\nuXLlZ/Y2OSGEEEIIIYR4WLEqQ1qtluHDh+Pn52eoEYaGhho1mBBCCCGEEEIYU7EqQ506PXokdSGE\nEEIIIYSZKH13yRWva21/f39SU1O5ceMG169fJyrKPLqqFEIIIYQQQpRexWoZmjp1Kj4+PsTGxmJl\nZYVGozF2LiGEEEIIIYQJMdtBVwH69euHl5cXn376KXfv3jVmJiGEEEIIIYQwumK1DKnVanJycsjK\nykKlUpGbm2vsXEIIIYQQQghTUgq71i5Wy9Crr77Kjz/+SO3atQkJCcHd3d3YuYQQQgghhBDCqIrV\nMnTnzh0OHjxIdnY2WVlZXLhwwdi5hBBCCCGEECbEbAdd3b17N5988gmOjo7GziOEEEIIIYQQJaJY\nlSF7e3vc3NyMnUUIIYQQQgghSsxjK0OrV68GQKfTMWHCBCpWrGiY1717d+MmE0IIIYQQQpiOUti1\n9mMrQ97e3gX+K4QQQgghhBClxWMrQ82bNy+hGEIIIYQQQghTZrYdKAghhBBCCCGEki5cuMDq1asZ\nM2YMSUlJzJ07F5VKhZ+fH3369AFgz5497N27FwsLC15//XXq1q372HUavTJk7ShjEhVFY++idAST\nVsbNR+kIJkv2ncer3K2T0hFMlpyTiybH1ePN2Ddd6QgmS/adx7Nx8VQ6gnhaFGwY2rp1KwcPHsTG\nxgaA5cuX061bNwIDA1m0aBFHjx6latWq7Nixg8mTJ5Odnc1nn31GrVq1sLQsuspj9MrQrcjTxv4T\nzySHgFpkpSQpHcMk3TtpRi1br3AS01S915vcvhKpdAyTVK5iAABZyQkKJzFNNq7ect4pgo2LJznp\nKUrHMEn3vuhL+Tyaxt6FXSPmKR3DJLWeMhCAnFvJCicxTRoHV6Uj/G1K3ibn6enJhx9+SFhYGACX\nL18mMDAQgDp16nDq1ClUKhUBAQFYWFhga2uLp6cnMTExVKpUqcj1qkskvRBCCCGEEEL8Q/Xr18fC\nwsLwXq/XG/5tY2NDZmYmWVlZ2NraFpiekZHx2PVKZUgIIYQQQgjxTFGr71dj7lWCypQpU6Dy83Dl\n6JHrMVpCIYQQQgghhDCCihUrEhERAcCJEycIDAykcuXKREVFodPpyMjIID4+Hn9//8euR3qTE0II\nIYQQQjyZCQ262rNnTxYsWEBubi4+Pj40bNgQlUpFmzZtGD16NADdunV7bOcJIJUhIYQQQgghxDPA\nzc2N8ePHA+Dl5cXYsWMLLdOyZUtatmxZ7HVKZUgIIYQQQgjxRDLoqhBCCCGEEMI8lcLKkHSgIIQQ\nQgghhDBL0jIkhBBCCCGEeKLSeJuctAwJIYQQQgghzJJUhoQQQgghhBBmSSpDQgghhBBCCLMkzwwJ\nIYQQQgghnsyEBl19WopdGbp16xZardbw3tXV1SiBhBBCCCGEEKanNHagUKzK0OLFizlx4gROTk7o\n9XpUKpVh9FchhBBCCCGEeBYVqzJ08eJFZs+ejVr9bDxitOvgYUZ/NROVCvT6/PGhXq7/IlM+Gc6h\no8eZt2I1sYlJVPD1IaRnNxrVraN05BKXk5NDt//rx4ghg2kQVLfAvOirMbzV612O/LxboXQlT6vT\nMXTpPPoEt+WFCpUBSL1zm3k7tnLqyiXsbW15vWFT2tZrUOizCakphC4JY0yX/1HTv2JJR1fE+Blh\nxCUmMX9y/o8iSddv8MWMME5HROLl7saQfv9H44f2K3OQk5NDt3cHMGLIezSoV3D7tTodXXv3I7hF\nMwb8Xy+FEipLzjtPptVqmTBlOrv37cdaY03P7l3o3fNtpWOZDHMun+S7t9gQfoDLKfFYW2qo412V\n9jWaYKm2YO3JPfx6NRxQAXpAxes1X6Z55fzvN1dSE/j+zAESb6fgUdaJzjWbUcXVV8nNUczYCZOI\njYtnybzZSkd5Nplry5CnpydarRZra2tj53kqLsfG0qJhfT4a2Be9Pn+aRmPFldg4Rk6ZTuj/9aJx\n3Trs/+0Phn85le/mzMDbw13Z0CUoJyeHj8Z8zuXoq4XmJSQm8f6IT9DqdAokU4ZWp2Pq5u+ITb5u\nmKbX6/niuxU42Nrx9f+FcCkpgZk/bMTH2ZUXKlYu8PmwHzeZVXkdOXGKLTv3UK9WTcO0oWMnULm8\nPytmT+fn3/5gxBeTWL8wDC+zO67GP/K4Ali4dHmR88yBnHeKZ9rM2Zw5G8GSubNJunadTz4bh5en\nJ/8JbqV0NJNgruWTm5fLgt+34GXvwrCXu5KencGqE7tQqVS89lxTEm+n8lrNl3nRN8DwGRtLDQBp\nmbeZ+9tmmlZ8gd5Bbfn1ajiLjmzjs1fewU5TRqlNUsTvR46xcesPvGiGP4KLohWrqSc5OZmQkBBG\njRrFqFGj+PTTT42d61+5EhtHlQrlcXJwwNkx/1XW1pZrySm89d+2vNHmVbw93Hn7tfbY2FgTHnVB\n6cgl5nJ0ND36DiQ+IbHQvD37D9D93f6UsbFRIJkyYpOv8+GyBVy7lVZg+vFL50lKS2X4a13wcXHl\n5edq8coLdYmMjymw3I/H/yDvXo3bDGRlZfPlrLnUfi7QMO3oydPExCcwKnQQFfx8eeetztSqEcCW\nnebzC//l6Kv06DeI+MTCxxXA+YuX2PTDT1Tw9yvhZKZBzjvFk5mVxcbN2/hoaCgB1avR/OWX6P2/\nt1mz/nulo5kEcy6fq2nXSM64RY+6r+Je1okqLj60C2jE0dhIAK7dScXfwZ1y1raGl5VF/u/dBy+f\nxMfBjQ41muBq50CHGk1wsXUg9ub1x/3JUiczK4vPJ02hzgu1lI7yTFOpVUZ5KalYlaHQ0FAmTZrE\nkCFDGDJkCKGhocbO9a9ciY2jvI93oekN67zA+716AKDLzWXL7r3otDqeD6hW0hEVc+zEKRoE1WPF\nwrnoH/oSf+i33wkN6cew90MUSlfywmOieaFCJab26seDxXHm6hWeL18Ruwe+oA38Twe6vNTC8P5G\n+k3W/rKf99q+hrnUh+Z8u4KgF56n7vP3W4XCI89TvXIlbGzutxzXfq4Gp89FKRFREfnHVV1WLJhT\n6LjKy8tj7MSpfBDSDwd7e4USKkvOO8UTdf4CWp2uwJe1Oi+8wNmIc4XKzRyZc/m4l3ViQMPX0FgU\nvKEnU5dNetZdMnKycC/r9MjPnk+Oo7ZXlQLTRjTvToB7eaPlNUWz5s6nfr26BNWtrXQUYWKKdZuc\nhYUFK1euJD09nYYNG1K+fHnc3NyMne0f0el0xCUlcejoceavWoter6dVk0b06/YWVlZWAFyNi6fr\n4GHo9XkM6vk2Xu6muS3G8FanjkXOGzfyIwD+OHa8pOIork3d+o+cnpiWipu9IysP7GHfmRPYWdvQ\noX5jgl+oZ1hm3k9b6Vi/CV5OLiUVV1GnIyLZ98uvrFswmxXfbzZMT05Nxc3FucCyzk4OXE9OKemI\ninmrU4ci5327eh1Ojo60ezWY9Zu3lWAq0yHnneJJTk7Bwb6c4VoF4OLihFarIyU1DdeHjjNzY87l\nU9a6DNXd7rcs6/V6Dl45RYCbP0m3U1Gr1PwQ+RvnrkVjp7GhReW6NPCvAUDy3ZtoLK349thPnE+O\nxc3OkU7PNaWCs5dSm1PiTp4OZ/e+A2xau4JvV65WOo4wMcVqGVqwYAEtW7YkNzeXGjVq8O233xo5\n1j8Xk5BIXl4etmVsmPzxhwzu/T92HDjEjKXLDcu4ODuxbPokPuz7LgtWr2P/b38omFiYosycbPaF\nn+Dm3duMeuNt2r/YiAU7t/HH+XMA7DtzgtQ7t3m94UsKJy0ZWq2W8TPCGDagL2Xt7ArMy8rOLvDl\nBEBjZVWgK35zFR0Ty/I13zF6xFClo4hnQGZWFlZWmgLTNH+912pzlIhkUqR87tsYfoD4W8l0qPES\nSXdSUalU+Dq4MbDRazQqX5O1p/ZyMiH/EYAsnZZtEYep4ORJSKNOlHfyJOzXjdzKvKPwVpQMrVbL\n2AkT+XhYKOXKllU6zrNPpTLOS0HFahnKycmhZs2abNiwAW9v70JffExJJX8/dq34Bvu/dvgqFcqj\n1+sZPX0mw/r0Rq1WU9bWlmoVK1CtYgUuXr3Kd9t/okWjwr2ECfNloVZT1qYMg9q8hkqlorKnN1eu\nJfHTn0eo7uPH0r07GNu1V35/+/o8peMa3cJVa/H39ablS40KzdNoNNzNyCwwLecZ6nDFmMZOnErf\nXm/jaUYdSYh/ztpaU+hLfc5f723kmSopn798f+ZnDkef4d0X2+FZzhnPcs4E+VbH1iq/DLztXblx\n5yaHrpymtndVLFQqanhUMPQs5+vgRtSNGI7EnSO46otKbkqJmLfoG8r7+/NKi+ZKRxEmqliVIY1G\nw8mTJ8nLy+P8+fMmXRkCDBWheyr4+qDL1XHqXBQWFmpqBVQ3zKvo52dWzzaI4nEuV84wptY9Pi6u\nnL56mT8vX+B2ZiYjVy4hvwvTfOPWLadLk+a80biZAomNa9fPh0hOS+PlTl2A/B748nLzaPZ6V3p3\neYOLV6ILLJ+SehNX50ffv24uEpOucfJMOFEXLhK28BsAsnOyORsZSXjEOcKmTVI4oTA17m5upKff\nRqfTYWmZf3lOSUlFY2Vlts+bPcjcy0ev17PqxG6Ox0fRO6gtNT0rGebdqwjd41HOicgb+T03OtiU\nxeOh54nc7RxJy7ht/NAm4Mdde0hJSaFB81cA0Gp15OXl0bBFML/vN5+Ofp4Wsx10tV+/fqxYsYLb\nt2+zbds2+vbta+xc/9j+3/5g8vxF/PDNAiwtLACIunSFsnZ2/HHyFIeOHmPVjGmG5SMvXaKCr49S\ncYWJCvDxZ/WlveTm5WHx1/haMTeu4+HgSOPqz1HD9/6Dp3n6PAbOn8n7bTtRt3JVpSIb1YKpE9Dp\ncg3vV23cQuSFS4z/eBiJ166xdN33ZGVnY/NXa9CpsxE8HxhQ1OrMgoe7Gz+sW1lg2ojPPqf28zXp\n3aObQqmEKQuoVg0rKytOnj5D0F9d/x4/cYoagQHPzDh/xmTu5bMx/CB/xkfRp/5/ec7j/ph2m8IP\ncv1OGv0b3n82L/bWDTzK5j9DVcHZi9hbNwqsK+lOKuWdPEsmuMK+XRBWoNv+5avXEnEuislfjFUu\nlDApxTp7bN++nSFDhvDVV18xbNgw3N1N95aPujXzHxicOGcBMQmJ/HLsOLOXraBnp460f6UF8YnX\nmL9qLbGJiazb9iN7fvmVd958XeHUwtQ0rfE8Fmo1YT9uIiE1mf3hJ9l35gRt6zXARqPB08nZ8PJw\nzL/gOJcrR1mb0jlmg6ebG75enoaXfdmyWFtr8PH0oO7zNfF0d2fstJlcvhrDt99tIDzqPJ3atFY6\ntqLUajW+Pt4FXhqNBnv7cri5mkenG+LvsbGxpn27NkyYPI3wsxHsP3CI5avW0KPbW0pHMwnmXD5X\nUhM5cPkEbQMa4efgTnrWXcOrpmclIq5f5cDlkyTfvcWByyc5FnuOV6rmd/jTolIdwpMu8/OlE9y4\ne5MtZw9xK/MOL/oFPuGvlg6eHh74+fgYXvblymFjbY3vI3odFsVgrs8MxcXFcffuXeweenDaFDmU\nK8essZ8y45tl9Br6EXa2tnRu8yq9Or8GwMyxo/h6ybes3vIDPp7uTProQ6pVrKBoZqWUxqbOf+PB\n4iijseaL7r2Zt2MbgxfPwblsWQb+pz1BVao/8bPmRq1WM33MSL74ejY9Bw/D18uLaZ+NxNOMeml8\n0OOOKznmpAyeZPiQwYyfPJU+gwZjZ2fHwL7vEtyyxZM/aCbMtXxOJV4EVGyNOMzWiMN/TdUDKmZ0\nGMw7QW34KfJ3tkT8gqutA+8EtaWic/6XfX8nD/rU/y9bzv7C1nOH8S7nQkijTjjYmP53OiFKgkpf\njM75Q0JCSElJwd7eHpVKhUqlYsGCBcX6A7ciT//rkKWRQ0AtslKSlI5hkmxc8pvuo5atVziJaare\n601uX4lUOoZJKlcx/9a8rOQEhZOYJhtXbznvFMHGxZOcdPPpDv7v0Njnt2RK+Tyaxt6FXSPmKR3D\nJLWeMhCAnFvJCicxTRoHV6Uj/G2pJ48YZb3OtR891ElJKFbL0Ny5c42dQwghhBBCCCFK1GMrQxs2\nbKBz587MmDGj0K0NoaGhRg0mhBBCCCGEEMb02MpQvXr5D9/VrVuXjIwMLCws2LJlC23atCmRcEII\nIYQQQggTUQqf+3xsb3IVKlQAYO/evfj6+nL69Gm6devGsWPHSiKbEEIIIYQQQhhNsbrWVqvV1KhR\ng4yMDJo0aSK9AQkhhBBCCGFuzLVrbZ1Ox8qVKwkMDCQ8PBzdA4NXCSGEEEIIIUq/0tggUqyWoZCQ\nEDw8POjYsSPp6ekMGjTI2LmEEEIIIYQQwqiK1TLk5eWFl5cXAI0bNzZqICGEEEIIIYQJUptpy5AQ\nQgghhBBClDZSGRJCCCGEEEKYJakMCSGEEEIIIcxSsZ4ZEkIIIYQQQpg3lar0taNIZUgIIYQQQgjx\nZObatbYQQgghhBBClDbSMiSEEEIIIYR4otI46KpKr9frlQ4hhBBCCCGEMG23Ik8bZb0OAbWMst7i\nkJYhIYQQQgghxJOVwkFXjV4ZSr90zth/4plkXzmQrJQkpWOYJBsXTwCSft6ncBLT5Nm8JTnpKUrH\nMEkaexcAUk8fUziJaXKuFST7ThE09i5SNkW4d1zdjbukcBLTZOdbmazkBKVjmCQbV28Autfvq3AS\n07T6yCKlIwikAwUhhBBCCCGEmZLb5IQQQgghhBBPVBo7UJCWISGEEEIIIYRZkpYhIYQQQgghxJNJ\ny5AQQgghhBBClA7SMiSEEEIIIYR4MlXpa0cpfVskhBBCCCGEEMUgLUNCCCGEEEKIJ1KVwkFXpWVI\nCCGEEEIIYZaKXRnKzs4GIC0tzWhhhBBCCCGEEKKkFKsytH79ejZs2ADA0qVL2bx5s1FDCSGEEEII\nIUyMSmWcl4KKVRk6duwY3bt3B2Do0KEcP37cqKGEEEIIIYQQwtiK1YGCWq1Gp9NhaWmJTqcjLy/P\n2LmEEEIIIYQQJkRVCgddLVZlKDg4mGHDhuHn50dCQgIdOnQwdi4hhBBCCCGEMKpiVYZatmxJUFAQ\n165dw8PDA3t7e2PneirGz5xDfGIS8yZ9AcChP44yd9lKYhMTqeDrS0ivHjQOqqtwSmXk5OTQ7f/6\nMWLIYBo8VAbRV2N4q9e7HPl5t0LpSl6OVku/LycxuMub1A0IAOBaaipTlq/k7OXLeDg7E/JGZxrU\nfM7wme/37WPT/gOk3LpFZV8fBnbuTM3KlZTahBKn1WqZMGU6u/ftx1pjTc/uXejd822lYylCl5vL\nnBVr2HHwFwBaNmrAB717YmlpyaQFS9i6dz8qQA+ogNB3evBW2/8oGVlRsXHxTPlqJn+eOoVtGVte\nfaUloYMGYGVlpXQ0xclxdV9cQiLT5i7kZPhZbMuUIbhZU957txdWVlZEXbzElLD5nL90GT8fH97v\n8w6NzPl6/u4ARgx5jwb1CpaBVqeja+9+BLdoxoD/66VQwpLVZ2RPPHzdmRAyHYA6L9Wiy8BOePi5\nkRCdxLq5mzj9+9lCn/PwdWPiqjFMGTKTyBMXSjr2s6MUDrr62MrQhg0b6Ny5MzNmzCjULBYaGmrU\nYP/WkZOn2LprD/WerwnA5ZhYPpk4lSF9e9M4qC77Dv/G8C8msn5hGN4eHgqnLVk5OTl8NOZzLkdf\nLTQvITGJ90d8glanUyCZMnK0Wj5f/A1XExMLTB85Zx4Vvb1ZMPJjDp88xWcLFrJs7Gd4uriw6/c/\nWLr1Bz7q9T+q+Prwwy+HGTFrNis+H4uLg4NCW1Kyps2czZmzESyZO5uka9f55LNxeHl68p/gVkpH\nK3Gzl6/ml2PHmfrxMAA+mzEHh3Ll6Nf1DaLj4nn/f2/zatPGhuXtypRRKqritDod7w0dTtXKlVn1\nzSJSUlMZ/fkEVCoVw0LfUzqe4uS4yqfV6QgdNZYqlSqwbPZXpKTdZOzUr0Gl4p2ubzDgw5E0a9KQ\n0cNCOX32HB+OGc/ir6cQWK2K0tFLVP71fPwjr+cAC5cuL3JeafTciwE07/AS5/48D4BPRS9Cv+zP\nypnfcfLXcOq3qMvQqYP48K3RJCemFPhs31G90GjkB5knMbtxhurVqwdA69atCQ4OLvAyZVlZ2Uyc\nPY/aNQIN064nJ9OlQzveaNcGbw8Perz+GmVsbAiPPK9g0pJ3OTqaHn0HEp+QWGjenv0H6P5uf8rY\n2CiQTBnRiYkMnDyFxJTkAtP/jIwk7vp1Puz5NuU9Pen+n1d5rnIlth/+FYAdv/9Op+bNeLlObbzd\n3OjX6TWcHew5fOq0EptR4jKzsti4eRsfDQ0loHo1mr/8Er3/9zZr1n+vdLQSd2sWfNoAABnaSURB\nVOduBpt37+WTAX2pWa0qNatVpU+XzkRdvgJAdHw8AZUq4OzgYHhZazQKp1ZO+NkI4uITGD/mUyqU\n96dendq8N6Av23fsUjqa4uS4uu9sZBTxiUmMGzGU8n6+1K1Vk5DePflpzz6279qLnZ0to4cOpoKf\nLx3+E0y74Jas/H6j0rFL1OXoq/ToN4j4xMLXc4DzFy+x6YefqODvV8LJlKGx1vDuxz2JOnXRMM3Z\n3Ykd3+1jz4YDJCem8OPq3WRnZVO1ZsG7OF7p3KxUfskXxfPYylCFChUA8PT0xN7eHkdHRw4cOICt\nrW1JZPvH5ixbQb1az1Pn+fu3NDWsW4f3/2oi1uXmsmXnbrQ6Lc8HVFcqpiKOnThFg6B6rFg4F71e\nX2Deod9+JzSkH8PeD1EoXck7df4C9QICmPvRCB4sjYgr0VT188PmgS+ttSpX5uzlywD07diR/zZ9\nqcC6VKi4k5lZErEVF3X+Alqdjjov1DJMq/PCC5yNOFdovyrtTkVGYWNtTdAD55u2zZoyfeRwUm/e\n4vadu5T39lYwoWmpUN6fuTOmYWNjXWD67du3FUpkOuS4uq+8ny+zJo7Dxvr+fqJCxZ27GcQnXSOw\nahUsLCwM86pVrsjpiEgloiom/3pelxUL5hTaP/Ly8hg7cSofhPTD4Rl5tOHf6hLSiYjjkUT+ef9H\n7jN/RLA2LH9oGLWFmuYdXsLS0pILZy4ZlnF2d+L1Pu1ZPGF5/n3MwuwU65mhmTNn8uabb7Jz504a\nNmzIsmXLGDNmjLGz/SOnz0Wy7/BvrJs3ixUbCo+HFB0XT9eBg9Hr9bz3Tk+8PNwVSKmctzp1LHLe\nuJEfAfDHMfPpOr1js5cfOT3l1i1cHAve7uZkb8+NtJsABFasUGDeH+Fnibt+naDAQMxBcnIKDvbl\nCjzj4eLihFarIyU1DVcXZwXTlaz4a9fwdHNl1y+/smzjFjKysmjZsAEDu7/F5dg41Go1C9Z+x28n\nTuFYrhxd/tuGds0fvd+ZAydHRxq8GGR4r9frWfPdBho2eFHBVKZBjqv7nBwcqF+ntuG9Xq9n3ZZt\n1K9bGxcnR85GRhVYPiHpOjdv3SrpmIp6q1PRnVl9u3odTo6OtHs1mPWbt5VgKmVUfb4S9VvWZUTX\nsfy3R+tC8738PZi8dixqlZq1czaQnJRqmPfuxz34cfVukuKul2RkYUKK3bV2jRo12LRpE02aNGHv\n3r3GzvWPaLVaxs+cw7D+fShrZ/fIZVydnVg+cxqnIs4xY9FSfL29aNG4YQknFaYuOycHK8uCh4eV\npSVanbbQsrHXrjHx22W82qgh1czkdoTMrCysrAre6qX5671Wm6NEJMVkZGaRcO063+/Yxcf9+3A3\nM5MpC78hNy8XPy9PVCoV1SpW4K22/+HPsxFMWfgNtjY2tGhYX+noJmHKVzOJunCRtcuXKB1FcXJc\nFW3a3IVEXbzMynkz0ev1LFy+mnWbt9G5fVsioi6wdcdutFrzedb1caJjYlm+5jvWfrNA6SglwsLS\ngr6jerF8+loy7z767oybKbf4tNcEqtWqQo8hb5EUd4NjP5+gadtGOLo68MOKnXKbXHGZa9faOp2O\nlStXEhgYSHh4ODoTfbh+0ep1lPfxpmWTRkUuU9bWlmqVKlKtUkUuRcfw3dbtUhkShWisrLiblVVg\nmlanK3DbHMDl+HiGz5pNeS8vhr3dvSQjKsraWlPoy1nOX+9tzOiZMwALCwsyMjMZN3gQXu5uALz/\nv+58HjaffSuW0PqlxpT768eZyv5+xCYmsXHXHqkMAZOmfc13Gzfz9eQvqfTXbdnmTI6rR5saNp/v\nf/iRqWNGUfGvH5zGfTSUSbPmMX3uQny9vejeuSOrHnE3iDkaO3EqfXu9jaeZ3PnSuW97EmOucfTn\nE0Uuk3k3i5gLccRciMO/ig+vvtWS86cu0v39N5g8ZCZQOsfPEcVTrMpQSEgIp0+fpmXLlhw9epRB\ngwYZO9c/svPAIVLS0mjWuSsAWm3+ALHNO3dj8fRJZGRmUiswwLB8RX9fTkWcUyquMGGujo5ciosv\nMC01PR3nB3qKi4y+yohZs6ns68vE90IKtSSVZu5ubqSn3zYMxgyQkpKKxsrKbO5Pv8fVyRELCwtD\nRQjA39uLnJwcbqbfxsmhYHlU8PHhyOnwko5pUvR6PaM/n8BPu/YwbeIXNGvaROlIJkGOq4L0ej1j\np37Nzn0HmDz6E15u1MAwr02rFrRp1YKU1DRcnJ1Yv3W72fUM+yiJSdc4eSacqAsXCVv4DQDZOdmc\njYwkPOIcYdMmKZzw6Wvcuj4OLvYs2T8bAEsrC9RqNYv3zWJsn0mUsbPhwpnLhuXjriRSrVYVXmhU\nk7IOdnw678MCFaERM0LZtOQHti3fUeLb8iwojZXGYn17S0tLw8/PjwsXLuDo6EhKSgoeJnjSWTB5\nPDpdruH96k1bibx4ic+Hf8C23Xs5dOQoq8NmGOafu3iJCn6+SkQVJq5GpYqs2rGT7JwcQ89fZy5e\n5LlK+T3QJNy4wYjZYVQvX57xIQOwNrPxUQKqVcPKyoqTp88QVLcOAMdPnKJGYABqdekbg+Bxalar\nSm5uLpdj46j01/nkSmwctmXKsGLLNmISEpn28YeG5aOuXDH7DhWmfD2LHbv3MGPKRJo+piXf3Mhx\nVdD0eYvYtf8g08d9SpMHnik7fuoM6zZvY8qYkbg4OwFw4NffCapdq6hVmQ0Pdzd+WLeywLQRn31O\n7edr0rtHN4VSGdfn/adiaXm/M4223YOpGFieOaMX06x9E+o0fYGRPT43zK8YUJ746ESO7DtO1Mn7\n4wmpLdRMW/8Fi8Yv49Rv5v2D1WOVwnGGirVFu3btMryWLFnC1q1bjZ3rH/F0c8PXy9PwKlfWDmuN\nBl8vTzq0fiV/ANblq4hNSGTtlh/Yc/Awvbu8oXRsYYJqV62Kh7MzX367jOiERFbv2EnElWja/9WD\n3Ndr1lLG2pqhb3fnbmYmqenppKank5mdrXDykmFjY037dm2YMHka4Wcj2H/gEMtXraFHt7eUjlbi\n/Lw8eSmoHuPnLCDq8hVOnotk3up1dHylBS/Vq8vvJ0+z/qedxF+7zvqfdrHj4GF6dPyv0rEVc+pM\nOKvWfkdIvz4EBlQnOSXV8DJ3clzddzoikjUbt9C/19sEVK1CSmqa4VXez4ffjh1n7aatJCRdY97S\nFYRHRtH99aI7CDIXarUaXx/vAi+NRoO9fTncXF2UjmcUqdfTuJ6QbHjdvZ1BTraW6wnJ/LztMB4+\nbrzZvyMevm68+lZLGr0SxJZvfyQ7K6fA567F3wAgLfkmGXfMo2dYka9YLUNDhgwx/Fun0/HVV18Z\nLZCx+Hh6MOvzMXy16BtWb9qKj6cHk0eNoFqlikpHU0xpbOr8Nx4sDbVazZchA5i8fCX9vpyIt5sb\nEwYOwN3ZmYysLI5GnEMFdP90dIF19Gjbhnc7tC/R3EoZPmQw4ydPpc+gwdjZ2TGw77sEt2yhdCxF\njB08kK+XruD9z7/EQm1B2+YvM6B7FywtLPg8dBBL1m9kzsq1+Hi488UH7/F89apKR1bMnn0/o1Kp\nmDlnPjPnzAfyb4dSqVSc+O2gWbaAPEiOq3x7Dx1GpVIRtmQZYUuWAff3kyM7tzJlzCi+mreIsCXL\nqF6lEvOnTiy1X/aL43HXc3O+1t9ISGZy6Ax6ftCFNt2DuR5/gxkfzyPmQtyjP2BePdj/M6WwowmV\n/m8OXpCdnc3IkSOZPn16sZZPvyTP5DyKfeVAslKSlI5hkmxcPAFI+nmfwklMk2fzluSkpzx5QTOk\nsc//MpR6+pjCSUyTc60g2XeKoLF3kbIpwr3j6m7cpScsaZ7sfCuTlZygdAyTZOOaf0tw9/p9FU5i\nmlYfWaR0hL8t41qMUdZr6+FvlPUWR7Fahvr162f4ZUGn09GuXTujhhJCCCGEEEIIYyt2ZWj//v1o\ntfljrERFRT3hE0IIIYQQQghh2opVGVqxYgX9+/fH1tbW2HmEEEIIIYQQJqg0PoNWrMqQn58fNWrU\nMHYWIYQQQgghhKkqhV1rF6syFBQUxKhRo/Dx8TFMCwkJMVooIYQQQgghhDC2YlWGfvrpJzp27Ci3\nyQkhhBBCCGGmzPY2OUdHRxo3bmzsLEIIIYQQQghRYopVGdJoNEyYMIGKFe8PUNq9e3ejhRJCCCGE\nEEKYGHN9ZqhevXrGziGEEEIIIYQQJapYlaHmzZsbOYYQQgghhBBClKxiVYaEEEIIIYQQ5k2lLn0d\nKJS+G/+EEEIIIYQQohikZUgIIYQQQgjxZObatbYQQgghhBBCKEGv17N48WKuXr2KlZUVAwYMwMPD\n46msW26TE0IIIYQQQjyRSqU2yutJjh49ilarZfz48XTv3p1ly5Y9tW2SypAQQgghhBDCZEVGRlK7\ndm0AqlatyuXLl5/auo1+m5x95UBj/4lnlo2Lp9IRTJpn85ZKRzBZGnsXpSOYNOdaQUpHMFmy7xRN\nyubx7HwrKx3BZNm4eisdwaStPrJI6QjiaVHomaGMjAxsbW0N7y0sLMjLy0Ot/vftOvLMkBBCCCGE\nEOKJlPrRyNbWlqysLMP7p1URArlNTgghhBBCCGHCqlevzp9//gnA+fPn8ff3f2rrVun1ev1TW5sQ\nQgghhBBCPEX3epOLiYkBYODAgXh7P53bU6UyJIQQQgghhDBLcpucEEIIIYQQwixJZUgIIYQQQghh\nlqQyJIQQQgghhDBLUhkSQgghhBBCmKVSURlatmwZKSkpSscwWT///DOrV69WOobJKKo8Zs6cSW5u\nLnPnzuXUqVMF5mm1WgYNGlRSEU2WlMPjPap8Tp48yd69e7lx4wajRo0CYNCgQeh0OiUimpzNmzdz\n6dIlpWOIEiLXo6cnOjqaDRs2KB2jVJCyNG+lYtDVXr16KR1BlAKhoaFFztPr9agUGnXZlEg5PN6j\nyqd27doA3LhxwzBPyvC+1157TekIQjyTKlSoQIUKFZSOUSpIWZq3Z64ylJmZyfz588nIyCAtLY3W\nrVvz22+/0bdvXw4fPsz58+fJysoq1P/4xYsX+fbbb9Hr9Tg7OzN48GDi4uJYunQpFhYWWFlZ0b9/\nf/Ly8pg5cyaurq4kJSVRpUoV+vTpo+AWPx0XLlxgwoQJpKenExwcjLu7O+vWrUOj0VC2bFkGDhxI\ndHQ0q1atwtLSklatWrFt2zYCAwOJiYlBpVIxYsQIypQpo/Sm/G05OTnMnTuX5ORkcnNzadCgQYHy\naN26Na1atWLQoEHMnDnT8LmsrCxmzZpFRkYGHh4ehunjxo3D3t6eu3fv8vHHH7N48WKSkpLQ6/V0\n6dKFGjVqMHz48FJRdvDocoiOjuabb74pcOy4uLjw/fffc/ToUezt7cnJyTGUR2n2cPno9foC+0jj\nxo1JSkoiODjY8JnSNKJBbm4uixYtKnAMLF269JH7/+LFi7ly5QoODg5cv36djz76iPXr19OkSRPS\n0tI4ceIE2dnZXL9+nY4dO9KsWTNiYmJYunQpAOXKlWPgwIHP7LH0oJycHMLCwrh58ybOzs6cO3cO\nb29v+vbti7e3N7t37+bWrVs0a9as1F2THr4elS1blp07d5KXlwfA8OHDiYmJYdOmTVhZWZGSkkJw\ncDDh4eFcvXqVtm3bEhwczIcffmjYz7y9vXFwcODcuXNoNBo++eQTsrKymD9/Pnfu3AGgd+/e+Pn5\nERISgq+vL76+vvzvf/9Tsij+lsTERObOnYulpSV5eXm0atWKP//8kyFDhhTYpnbt2rFw4UJycnLQ\naDT079+f3NzcUrcfQX5L4/Hjx8nJyeHmzZu0adOGY8eOERsbS8+ePVm4cCELFy4EYMaMGbRu3Ron\nJ6cC5RgaGkpSUhK7du1iyJAh7Nu3j927d5OXl0dQUBBvvvmmwlspjO2ZqwwlJSXRpEkT6tevT1pa\nGmPHjsXZ2dkw38fHh3feeafQ5xYuXMiQIUPw9vZm//79xMXFsXDhQgYOHIi/vz/Hjh1j2bJl9OzZ\nk8TEREaPHo2VlRXvv/8+t27dwsHBoQS38umztLRk1KhR3Lhxgy+//BKdTscXX3yBo6MjP/30Exs2\nbKBevXpotVomTJgAwLp162jatClVq1Zl1qxZnDhxgsaNGyu8JX/f7t27cXd3Z8iQISQlJfHnn38W\nKI9JkybRqlWrQr/W7969G39/f7p27crFixc5e/asYV7Tpk0JCgpi165d2NvbM2DAAO7cucOYMWOY\nPn06GRkZpaLsoHA5hIeHs2DBgkLHTufOnTl16hSTJ08mJyeH4cOHKx29RDyqfABeeuklXnzxRX7+\n+WdlAxrZ3r17Cx0D2dnZhfZ/jUbDnTt3DF+CQ0NDCx1zmZmZjBw5kqSkJCZPnkyzZs1YsGABISEh\n+Pj4sG/fPrZs2ULXrl0V2tqnZ8+ePXh4eDB06FASEhIYOnRokQMIlrZr0oPn34kTJ/Lyyy/zySef\noNFoWLhwISdPnsTZ2ZnU1FSmTp3KpUuX+PrrrwkLCyM5OZnp06cTHBxMZmYmTZs2pVq1anzwwQf0\n6tWLrl27Mm7cOGJjY/nll194/vnnCQ4OJikpiblz5/L555+TkpLC1KlTsbOzU7oo/pbTp09TtWpV\nevToQUREBHFxcYZj6MFtmjFjBm3atKF27dqEh4ezcuVKunXrVur2o3uysrIYNWoUv/76K9u3b2fC\nhAlERESwffv2R7bCP1yOGRkZQH6LfXp6Olu2bGH69OlYWlqyevVqsrOzsba2LunNEiXomasMOTg4\nsH37do4cOYKNjU2h++7vXUx+//13du7cCUDPnj25deuWYV6LFi0ASEtLw9/fH4DAwEDDfcyenp6G\nHd/JyQmtVmv8DTOyihUrAuDo6EhycjJeXl44OjoC+du+Zs0a6tWrV+BirFKpDM3Grq6uz2w5JCQk\nUKdOHSD//62trW2B8sjOzn7k5xITE6lbty4AVapUwcLCwjDPy8sLgJiYGKKiorhw4QIAeXl53L59\nG6BUlB0ULgdLS8tCx86qVauIj4+nSpUqAGg0GkMZl3aPKh/gqY2MbeoePgZyc3O5fft2of3/+vXr\nVKtWDQB7e3t8fHwKreveZ1xcXAzHTHx8PIsXLzas29PT08hbVDLi4+MN5yVvb2/s7e0LzH+w9bC0\nXZMePv/a29sTFhaGjY0NiYmJVK9eHQA/Pz/UajV2dnZ4eHigVqspW7YsOTk5hdZla2uLr68vAHZ2\ndmi1WmJiYjh79iy//vorAHfv3gXy979nrSIE0LJlS7Zs2cKECROwtbWlVq1ahnkPbtO9VrUtW7YA\nGK5dpW0/uufBfeDeecXW1rbI7Xu4HLt162aYd+3aNfz8/Azn8e7duxs5vTAFz1xl6IcffqB69eoE\nBwdz9uxZTpw4UWC+Wp3fJ0TDhg1p2LChYbqTkxNJSUl4enqyZcsWvLy8cHJyIiYmBn9/f86ePWv4\ngvug0nQ7yz329vZkZmZy8+ZNHB0diYiIKHLbS8OzDT4+Ply8eJGgoCCuXbvGmjVraNasWaHlHv5/\n7evry/nz5wkKCuLKlSvk5uYa5t0rFx8fH1xdXXnttdfIyclh06ZNlC1btsAyz7qHy0Gn0xU6dry9\nvfHz82PHjh1AfkcC0dHRygYvIY8qHyg9//+f5OFjYOPGjRw6dKjQ9vv7+3Pw4EHatm3LnTt3SExM\nLPb633vvPVxcXIiKiuLmzZvG2IwS5+fnR1RUFEFBQSQlJXH79m00Gg1paWl4e3tz5coVXFxcCn2u\ntF2TMjIyWL9+PXPnzkWv1zN+/HjDNhbnGHrUMvc+7+PjQ+XKlWnSpAnp6ens27ev2Os1RUePHiUg\nIIA33niDw4cPs2bNGqpWrQoU3CYfHx/at29PtWrVSEhIICIiotC6Stt+VBSdTkd2djYWFhbExsYC\nhctxy5Ythu8EHh4eJCQkoNPpsLS05KuvvqJ37944OTkpuRnCyJ65ylC9evX45ptv+PXXX7G1tcXC\nwqJYXz769evHvHnzUKvVODo60q5dO9zd3VmyZAmQ/8vJgAEDCq3nWT1pPo5KpaJ///5MmzbN8Kvb\noEGDDPf3P7hcaRAcHMy8efMYO3Yser2e9u3bk56eXmi5h7c3ODiYsLAwxowZg5eXF1ZWVo9c9/z5\n8xk7diyZmZm8+uqrqFSqUlN2ULgc7t2D/vCx4+7uTu3atRk1ahTlypXD0tLS8OtaafZg+Xh7e6PR\naJ74mdK2fzzqGHhY3bp1OXHiBKNHj8bBwQGNRlOgtbUo7777LmFhYeTm5qJSqRg4cKAxNqPEtWzZ\nkrlz5zJ27FhcXV3RaDS0adOGxYsX4+bmVuD279J4Xr7H1taWqlWrMmrUKCwsLLCzsyMtLQ13d/cn\nfraocrn379dff5158+axZ88eMjMzDc9+PKtlWLlyZebMmcPGjRvR6/W0adOGixcvAgW3qWfPnixa\ntAitVotWqzU8OlCa96NHUalUtGvXjpEjR+Lh4YGbmxtQuBx79epluFXO3t6eDh06MGbMGNRqNfXq\n1ZOKkBlQ6c3l5wEhhFGlp6fz+++/07p1a3Q6HcOGDeOzzz575K/bwvwkJCQQHR1N48aNuXPnDsOG\nDWPOnDlmUWF+lHud/dSqVYukpCS+/PJLZs2apXQsIYQwO+Z5FRJCPHXlypXj4sWL7N+/H7VaTatW\nraQiJAxcXFxYuXIl27dvR6/X8/bbb5ttRQjA3d2dmTNnsn79evLy8kpFz15CCPEskpYhIYQQQggh\nhFlSKx1ACCGEEEIIIZQglSEhhBBCCCGEWZLKkBBCCCGEEMIsSWVICCGEEEIIYZakMiSEEEIIIYQw\nS/8Pqt2iA8geF5wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12f6e8f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sn\n",
    "import pandas as pd\n",
    "\n",
    "labels = [\"air-con\",\"horn\",\"children\",\"dog\",\"drill\",\"engine\",\"gun\",\"hammer\",\"siren\",\"music\"]\n",
    "print \"Showing Confusion Matrix\"\n",
    "y_prob = model.predict_proba(test_x, verbose=0)\n",
    "y_pred = np_utils.probas_to_classes(y_prob)\n",
    "y_true = np.argmax(test_y, 1)\n",
    "cm = confusion_matrix(y_true, y_pred)\n",
    "\n",
    "def print_cm(cm, labels, hide_zeroes=False, hide_diagonal=False, hide_threshold=None):\n",
    "    \"\"\"pretty print for confusion matrixes\"\"\"\n",
    "    columnwidth = max([len(x) for x in labels]+[5]) # 5 is value length\n",
    "    empty_cell = \" \" * columnwidth\n",
    "    # Print header\n",
    "    print \"    \" + empty_cell,\n",
    "    for label in labels: \n",
    "        print \"%{0}s\".format(columnwidth) % label,\n",
    "    print\n",
    "    # Print rows\n",
    "    for i, label1 in enumerate(labels):\n",
    "        print \"    %{0}s\".format(columnwidth) % label1,\n",
    "        for j in range(len(labels)): \n",
    "            cell = \"%{0}s\".format(columnwidth) % cm[i, j]\n",
    "            if hide_zeroes:\n",
    "                cell = cell if float(cm[i, j]) != 0 else empty_cell\n",
    "            if hide_diagonal:\n",
    "                cell = cell if i != j else empty_cell\n",
    "            if hide_threshold:\n",
    "                cell = cell if cm[i, j] > hide_threshold else empty_cell\n",
    "            print cell,\n",
    "        print\n",
    "    \n",
    "print_cm(cm, labels)\n",
    "\n",
    "df_cm = pd.DataFrame(cm, labels, labels)\n",
    "plt.figure(figsize = (16,8))\n",
    "sn.heatmap(df_cm, annot=True, annot_kws={\"size\": 14}, fmt='g', linewidths=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "-----  air conditioner -----\n",
      "Top guess:  air conditioner  ( 0.914 )\n",
      "2nd guess:  siren  ( 0.02 )\n",
      "\n",
      "-----  car horn -----\n",
      "Top guess:  car horn  ( 0.873 )\n",
      "2nd guess:  street music  ( 0.066 )\n",
      "\n",
      "-----  children playing -----\n",
      "Top guess:  children playing  ( 0.306 )\n",
      "2nd guess:  street music  ( 0.199 )\n",
      "\n",
      "-----  dog bark -----\n",
      "Top guess:  dog bark  ( 0.652 )\n",
      "2nd guess:  children playing  ( 0.244 )\n",
      "\n",
      "-----  drilling -----\n",
      "Top guess:  drilling  ( 0.595 )\n",
      "2nd guess:  jackhammer  ( 0.103 )\n",
      "\n",
      "-----  engine idling -----\n",
      "Top guess:  engine idling  ( 0.876 )\n",
      "2nd guess:  street music  ( 0.033 )\n",
      "\n",
      "-----  gun shot -----\n",
      "Top guess:  gun shot  ( 0.621 )\n",
      "2nd guess:  drilling  ( 0.087 )\n",
      "\n",
      "-----  jackhammer -----\n",
      "Top guess:  jackhammer  ( 0.794 )\n",
      "2nd guess:  street music  ( 0.041 )\n",
      "\n",
      "-----  siren -----\n",
      "Top guess:  siren  ( 0.97 )\n",
      "2nd guess:  street music  ( 0.013 )\n",
      "\n",
      "-----  street music -----\n",
      "Top guess:  street music  ( 0.729 )\n",
      "2nd guess:  car horn  ( 0.056 )\n"
     ]
    }
   ],
   "source": [
    "sound_file_paths = [\"aircon.wav\", \"carhorn.wav\", \"play.wav\", \"dogbark.wav\", \"drill.wav\",\n",
    "                    \"engine.wav\",\"gunshots.wav\",\"jackhammer.wav\",\"siren.wav\",\"music.wav\"]\n",
    "sound_names = [\"air conditioner\",\"car horn\",\"children playing\",\"dog bark\",\"drilling\",\"engine idling\",\n",
    "               \"gun shot\",\"jackhammer\",\"siren\",\"street music\"]\n",
    "parent_dir = 'samples/us8k/'\n",
    "\n",
    "def windows(data, window_size):\n",
    "    start = 0\n",
    "    while start < len(data):\n",
    "        yield start, start + window_size\n",
    "        start += (window_size / 2)\n",
    "\n",
    "def extract_features_array(filename, bands = 60, frames = 41):\n",
    "    window_size = 512 * (frames - 1)\n",
    "    log_specgrams = []\n",
    "    sound_clip,s = librosa.load(filename)        \n",
    "    for (start,end) in windows(sound_clip,window_size):\n",
    "        if(len(sound_clip[start:end]) == window_size):\n",
    "            signal = sound_clip[start:end]\n",
    "            melspec = librosa.feature.melspectrogram(signal, n_mels = bands)\n",
    "            logspec = librosa.logamplitude(melspec)\n",
    "            logspec = logspec.T.flatten()[:, np.newaxis].T\n",
    "            log_specgrams.append(logspec)\n",
    "            \n",
    "    log_specgrams = np.asarray(log_specgrams).reshape(len(log_specgrams),bands,frames,1)\n",
    "    features = np.concatenate((log_specgrams, np.zeros(np.shape(log_specgrams))), axis = 3)\n",
    "    for i in range(len(features)):\n",
    "        features[i, :, :, 1] = librosa.feature.delta(features[i, :, :, 0])\n",
    "    \n",
    "    return np.array(features)\n",
    "\n",
    "\n",
    "# create predictions for each of the sound classes\n",
    "for s in range(len(sound_names)):\n",
    "\n",
    "    print \"\\n----- \", sound_names[s], \"-----\"\n",
    "    # load audio file and extract features\n",
    "    predict_file = parent_dir + sound_file_paths[s]\n",
    "    predict_x = extract_features_array(predict_file)\n",
    "    \n",
    "    # generate prediction, passing in just a single row of features\n",
    "    predictions = model.predict(predict_x)\n",
    "    \n",
    "    if len(predictions) == 0: \n",
    "        print \"No prediction\"\n",
    "        continue\n",
    "        \n",
    "    #for i in range(len(predictions[0])):\n",
    "    #    print sound_names[i], \"=\", round(predictions[0,i] * 100, 1)\n",
    "    \n",
    "    # get the indices of the top 2 predictions, invert into descending order\n",
    "    ind = np.argpartition(predictions[0], -2)[-2:]\n",
    "    ind[np.argsort(predictions[0][ind])]\n",
    "    ind = ind[::-1]\n",
    "    \n",
    "    print \"Top guess: \", sound_names[ind[0]], \" (\",round(predictions[0,ind[0]],3),\")\"\n",
    "    print \"2nd guess: \", sound_names[ind[1]], \" (\",round(predictions[0,ind[1]],3),\")\"\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Although this model can correctly identify all 10 samples, the accuracy across the entire test set is still only 59%. From my investigations, there do seem to be certain recordings the classifier finds difficult to identify correctly, these tend to be the very short recordings, 1 or 2 seconds long, and recordings that consist of only rapid repetitive noise. \n",
    "\n",
    "I managed to get slightly better results with the CNN than my best FFN, an ROC of 0.91 with the CNN, compared to 0.89 with my FFN. My max f-score accuracy for both is about 60%. Getting higher results proved difficult, and there may be diminishing returns for effort given the constraints of the data set, (8000 samples isn't a huge amount, and it does include some repetition of same sound source).\n",
    "\n",
    "In their paper Salamon and Bello report an accuracy of 73% - but it's worth bearing in mind that their CNN is optimised for a different feature extraction process. I was feeding my implementation of their CNN features extracted with Piczak's method."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the plot of accuracy indicates whether the model could be trained further (if the accuracy trend is still rising for the final epochs). It will also indicate if the model has over-learned the training dataset, it should show comparable performance on both train and test datasets. If the model is overfitting the accuracy graph will show excellent performance on the training data, and poor performance on the test data.\n",
    "\n",
    "Ideally, in the loss graph, we should see comparable performance in both the training and validation datasets.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "History keys: ['acc', 'loss', 'val_acc', 'val_loss']\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA74AAAGRCAYAAABCG2/LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nXxEREREREfE0Nb4iIiIiIiLiaf8fcntZNzZEj5oAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x122f80bd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(16,6))\n",
    "\n",
    "print \"History keys:\", (history.history.keys())\n",
    "# summarise history for training and validation set accuracy\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "plt.title('Model Accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'validation'], loc='upper left')\n",
    "\n",
    "# summarise history for training and validation set loss\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('Model Loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'validation'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Accuracy plot on the left provides an insight into whether the model has reached a plateau or could be potentially trained further. It should also show comparable performance on both train and validation datasets, otherwise this would be an indication that the model has over-learned the training dataset.\n",
    "\n",
    "The Loss plot on the right should show comparable performance in both the training and validation datasets. Should these parallel plots start to diverge, it might be a indication we should stop training earlier. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Next Up\n",
    "\n",
    "To summarise then, I've managed to implement a general audio classifier capable of distinguishing between different sources of sound. Next, I intend to investigate whether this classifier will be able to effectively recognise some very different sources of sound: birdsong. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}