{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The Very Basics of Musical Instruments Classification using Machine Learning\n",
    "## Short-Time Fourier Transform (STFT) and Convolutional Neural Networks (CNN) \n",
    "\n",
    "<br>\n",
    "\n",
    "<p align=\"left\">\n",
    "<img src=\"./img/businesscard.jpg\" width=\"300px\" alt=\"Business Card\" align=\"left\" >\n",
    "</p>\n",
    "<br>\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[name: \"/device:CPU:0\"\n",
      "device_type: \"CPU\"\n",
      "memory_limit: 268435456\n",
      "locality {\n",
      "}\n",
      "incarnation: 10604460152652463385\n",
      "]\n"
     ]
    }
   ],
   "source": [
    "# Imports\n",
    "\n",
    "#General\n",
    "import numpy as np\n",
    "import itertools\n",
    "\n",
    "# System\n",
    "import os, fnmatch\n",
    "\n",
    "# Data\n",
    "import pandas as pd\n",
    "\n",
    "# Visualization\n",
    "import seaborn \n",
    "import matplotlib.pyplot as plt\n",
    "from IPython.core.display import HTML, display, Image\n",
    "\n",
    "# Machine Learning\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder, StandardScaler\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "from sklearn.metrics import recall_score, precision_score, accuracy_score\n",
    "from sklearn.metrics import confusion_matrix, f1_score, classification_report\n",
    "\n",
    "\n",
    "# Deep Learning\n",
    "import tensorflow as tf\n",
    "from tensorflow.python.client import device_lib \n",
    "from keras.backend.tensorflow_backend import set_session\n",
    "from tensorflow.python.client import device_lib\n",
    "from keras import backend as K\n",
    "from keras.models import Sequential, Model\n",
    "from keras.layers import Input, Convolution2D, MaxPooling2D, Dense, Dropout, Activation, Flatten, merge\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.callbacks import History, EarlyStopping, ModelCheckpoint\n",
    "from keras.models import load_model\n",
    "\n",
    "\n",
    "# Random Seed\n",
    "from tensorflow import set_random_seed\n",
    "from numpy.random import seed\n",
    "seed(0)\n",
    "set_random_seed(0)\n",
    "\n",
    "# Audio\n",
    "import librosa.display, librosa\n",
    "from librosa.util import normalize as normalize\n",
    "import IPython.display as ipd\n",
    "\n",
    "# Configurations\n",
    "path='./audio/london_phill_dataset_multi/'\n",
    "\n",
    "# Display CPUs and GPUs\n",
    "print(device_lib.list_local_devices())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Configurations for Google Colab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Not inside Google Colab: No module named 'google.colab'. Using standard configurations.\n",
      "processor\t: 0\r\n",
      "vendor_id\t: GenuineIntel\r\n",
      "cpu family\t: 6\r\n",
      "model\t\t: 142\r\n",
      "model name\t: Intel(R) Core(TM) i7-7500U CPU @ 2.70GHz\r\n",
      "stepping\t: 9\r\n",
      "cpu MHz\t\t: 2904.002\r\n",
      "cache size\t: 4096 KB\r\n",
      "physical id\t: 0\r\n",
      "siblings\t: 2\r\n",
      "core id\t\t: 0\r\n",
      "cpu cores\t: 2\r\n",
      "apicid\t\t: 0\r\n",
      "initial apicid\t: 0\r\n",
      "fpu\t\t: yes\r\n",
      "fpu_exception\t: yes\r\n",
      "cpuid level\t: 22\r\n",
      "wp\t\t: yes\r\n",
      "flags\t\t: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase avx2 invpcid rdseed clflushopt\r\n",
      "bugs\t\t: cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf\r\n",
      "bogomips\t: 5808.00\r\n",
      "clflush size\t: 64\r\n",
      "cache_alignment\t: 64\r\n",
      "address sizes\t: 39 bits physical, 48 bits virtual\r\n",
      "power management:\r\n",
      "\r\n",
      "processor\t: 1\r\n",
      "vendor_id\t: GenuineIntel\r\n",
      "cpu family\t: 6\r\n",
      "model\t\t: 142\r\n",
      "model name\t: Intel(R) Core(TM) i7-7500U CPU @ 2.70GHz\r\n",
      "stepping\t: 9\r\n",
      "cpu MHz\t\t: 2904.002\r\n",
      "cache size\t: 4096 KB\r\n",
      "physical id\t: 0\r\n",
      "siblings\t: 2\r\n",
      "core id\t\t: 1\r\n",
      "cpu cores\t: 2\r\n",
      "apicid\t\t: 1\r\n",
      "initial apicid\t: 1\r\n",
      "fpu\t\t: yes\r\n",
      "fpu_exception\t: yes\r\n",
      "cpuid level\t: 22\r\n",
      "wp\t\t: yes\r\n",
      "flags\t\t: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase avx2 invpcid rdseed clflushopt\r\n",
      "bugs\t\t: cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf\r\n",
      "bogomips\t: 5808.00\r\n",
      "clflush size\t: 64\r\n",
      "cache_alignment\t: 64\r\n",
      "address sizes\t: 39 bits physical, 48 bits virtual\r\n",
      "power management:\r\n",
      "\r\n"
     ]
    }
   ],
   "source": [
    "# Only for Google Colab\n",
    "try:\n",
    "    import google.colab\n",
    "    if \"GPU:0\" in tf.test.gpu_device_name():\n",
    "        !nvidia-smi\n",
    "        config = tf.ConfigProto(log_device_placement=True, allow_soft_placement=True, device_count = {'GPU': 0})\n",
    "        config.gpu_options.allow_growth = True\n",
    "        session = tf.Session(config=config)\n",
    "        set_session(session)\n",
    "    else:\n",
    "        print(\"No GPU Detected. Configure the Runtime.\")\n",
    "    !git clone https://github.com/GuitarsAI/BasicsMusicalInstrumClassifi\n",
    "    !unzip ./BasicsMusicalInstrumClassifi/audio/*.zip -d ./BasicsMusicalInstrumClassifi/audio\n",
    "    path=\"./BasicsMusicalInstrumClassifi/audio/london_phill_dataset_multi/\"\n",
    "    os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'\n",
    "    tf.logging.set_verbosity(tf.logging.ERROR)\n",
    "    from tensorboardcolab import TensorBoardColab, TensorBoardColabCallback\n",
    "    \n",
    "except Exception as e:\n",
    "    print(\"Not inside Google Colab: %s. Using standard configurations.\" % (e))\n",
    "    !cat /proc/cpuinfo\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Signal Processing Parameters\n",
    "fs = 44100         # Sampling Frequency\n",
    "n_fft = 2048       # length of the FFT window\n",
    "hop_length = 512   # Number of samples between successive frames\n",
    "\n",
    "\n",
    "# Machine Learning Parameters\n",
    "testset_size = 0.25 #Percentage of data for Testing\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aux Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>.container { width:80% !important; }</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Function to Display a Website\n",
    "display(HTML(\"<style>.container { width:80% !important; }</style>\"))\n",
    "def show_web(url):\n",
    "    html_code='<center><iframe src=\"%s\" width=\"800\" height=\"600\" frameborder=\"0\" marginheight=\"0\" marginwidth=\"0\">Loading...</iframe></center>' \\\n",
    "\t\t% (url)\n",
    "    display(HTML(html_code))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find Audio Files, Generate Labels and Get Duration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 1 = viola_Gs4_025_pianissimo_arco-normal.mp3\n",
      "Get 2 = viola_A4_1_fortissimo_snap-pizz.mp3\n",
      "Get 3 = viola_A3_1_piano_pizz-normal.mp3\n",
      "Get 4 = viola_Gs4_05_pianissimo_arco-normal.mp3\n",
      "Get 5 = viola_Gs3_025_fortissimo_arco-normal.mp3\n",
      "Get 6 = viola_A3_1_piano_arco-normal.mp3\n",
      "Get 7 = viola_A4_025_piano_arco-normal.mp3\n",
      "Get 8 = viola_A3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 9 = viola_Gs4_15_mezzo-piano_arco-normal.mp3\n",
      "Get 10 = viola_A5_1_fortissimo_arco-normal.mp3\n",
      "Get 11 = viola_Gs4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 12 = viola_Gs5_05_fortissimo_arco-normal.mp3\n",
      "Get 13 = viola_Gs5_1_fortissimo_snap-pizz.mp3\n",
      "Get 14 = viola_Gs5_025_mezzo-piano_arco-normal.mp3\n",
      "Get 15 = viola_A3_1_fortissimo_arco-normal.mp3\n",
      "Get 16 = viola_A4_15_fortissimo_arco-normal.mp3\n",
      "Get 17 = viola_A4_05_forte_arco-normal.mp3\n",
      "Get 18 = viola_A5_1_pianissimo_arco-normal.mp3\n",
      "Get 19 = viola_A4_1_piano_arco-normal.mp3\n",
      "Get 20 = viola_A3_025_fortissimo_arco-normal.mp3\n",
      "Get 21 = viola_Gs6_025_pianissimo_arco-normal.mp3\n",
      "Get 22 = viola_Gs6_05_fortissimo_arco-normal.mp3\n",
      "Get 23 = viola_Gs6_05_pianissimo_arco-normal.mp3\n",
      "Get 24 = viola_Gs4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 25 = viola_Gs5_1_fortissimo_arco-normal.mp3\n",
      "Get 26 = viola_A3_1_forte_arco-normal.mp3\n",
      "Get 27 = viola_Gs3_15_fortissimo_arco-normal.mp3\n",
      "Get 28 = viola_Gs4_025_fortissimo_arco-normal.mp3\n",
      "Get 29 = viola_Gs6_05_piano_arco-normal.mp3\n",
      "Get 30 = viola_Gs4_15_pianissimo_arco-normal.mp3\n",
      "Get 31 = viola_A4_05_fortissimo_arco-normal.mp3\n",
      "Get 32 = viola_A5_1_mezzo-forte_artificial-harmonic.mp3\n",
      "Get 33 = viola_Gs5_05_piano_arco-normal.mp3\n",
      "Get 34 = viola_A4_1_mezzo-forte_arco-major-trill.mp3\n",
      "Get 35 = viola_Gs3_025_piano_arco-normal.mp3\n",
      "Get 36 = viola_Gs4_1_piano_arco-normal.mp3\n",
      "Get 37 = viola_Gs5_15_fortissimo_arco-normal.mp3\n",
      "Get 38 = viola_A4_05_pianissimo_arco-normal.mp3\n",
      "Get 39 = viola_A4_05_piano_arco-normal.mp3\n",
      "Get 40 = viola_Ds3_1_piano_pizz-normal.mp3\n",
      "Get 41 = viola_A3_1_pianissimo_arco-normal.mp3\n",
      "Get 42 = viola_A3_15_fortissimo_arco-normal.mp3\n",
      "Get 43 = viola_A5_1_forte_arco-normal.mp3\n",
      "Get 44 = viola_Gs3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 45 = viola_Gs5_15_pianissimo_arco-normal.mp3\n",
      "Get 46 = viola_A3_05_fortissimo_arco-normal.mp3\n",
      "Get 47 = viola_Ds3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 48 = viola_Gs6_1_pianissimo_arco-normal.mp3\n",
      "Get 49 = viola_Gs3_15_pianissimo_arco-normal.mp3\n",
      "Get 50 = viola_A3_1_mezzo-forte_molto-vibrato.mp3\n",
      "Get 51 = viola_Gs6_025_fortissimo_arco-normal.mp3\n",
      "Get 52 = viola_A3_1_fortissimo_pizz-normal.mp3\n",
      "Get 53 = viola_Gs6_1_fortissimo_arco-normal.mp3\n",
      "Get 54 = viola_A3_15_piano_arco-normal.mp3\n",
      "Get 55 = viola_A4_1_mezzo-forte_arco-minor-trill.mp3\n",
      "Get 56 = viola_A4_1_fortissimo_arco-normal.mp3\n",
      "Get 57 = viola_A3_025_piano_arco-normal.mp3\n",
      "Get 58 = viola_A3_05_pianissimo_arco-normal.mp3\n",
      "Get 59 = viola_A4_1_piano_pizz-normal.mp3\n",
      "Get 60 = viola_A3_1_piano_arco-glissando.mp3\n",
      "Get 61 = viola_A4_1_pianissimo_arco-normal.mp3\n",
      "Get 62 = viola_Gs6_1_forte_arco-normal.mp3\n",
      "Get 63 = viola_Gs5_05_forte_arco-normal.mp3\n",
      "Get 64 = viola_A4_1_forte_arco-normal.mp3\n",
      "Get 65 = viola_A3_05_piano_arco-normal.mp3\n",
      "Get 66 = viola_Gs5_05_pianissimo_arco-normal.mp3\n",
      "Get 67 = viola_Gs4_1_fortissimo_arco-normal.mp3\n",
      "Get 68 = viola_Gs5_025_fortissimo_arco-normal.mp3\n",
      "Get 69 = viola_Gs5_025_pianissimo_arco-normal.mp3\n",
      "Get 70 = viola_Gs5_1_piano_arco-normal.mp3\n",
      "Get 71 = viola_Gs3_025_pianissimo_arco-normal.mp3\n",
      "Get 72 = viola_Gs4_1_pianissimo_arco-normal.mp3\n",
      "Get 73 = viola_A4_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 74 = viola_Gs4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 75 = viola_A4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 76 = viola_A4_025_pianissimo_arco-normal.mp3\n",
      "Get 77 = viola_A4_15_piano_arco-normal.mp3\n",
      "Get 78 = viola_Gs6_05_forte_arco-normal.mp3\n",
      "Get 79 = viola_Gs6_15_fortissimo_arco-normal.mp3\n",
      "Get 80 = viola_Gs6_025_forte_arco-normal.mp3\n",
      "Get 81 = viola_A4_15_pianissimo_arco-normal.mp3\n",
      "Get 82 = viola_Gs3_15_piano_arco-normal.mp3\n",
      "Get 83 = viola_A3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 84 = viola_Gs6_025_mezzo-piano_arco-normal.mp3\n",
      "Get 85 = viola_Gs6_15_forte_arco-normal.mp3\n",
      "Get 86 = viola_Gs4_15_fortissimo_arco-normal.mp3\n",
      "Get 87 = viola_A4_025_fortissimo_arco-normal.mp3\n",
      "Get 88 = viola_Ds3_05_fortissimo_arco-normal.mp3\n",
      "Get 89 = viola_A4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 90 = viola_Gs5_1_pianissimo_arco-normal.mp3\n",
      "Get 91 = viola_Gs4_05_fortissimo_arco-normal.mp3\n",
      "Get 92 = viola_Gs6_025_piano_arco-normal.mp3\n",
      "Get 93 = viola_A4_1_mezzo-forte_molto-vibrato.mp3\n",
      "Get 94 = viola_A3_15_forte_arco-normal.mp3\n",
      "Get 95 = viola_Gs5_025_forte_arco-normal.mp3\n",
      "Get 96 = viola_A3_025_pianissimo_arco-normal.mp3\n",
      "Get 97 = viola_Gs4_05_piano_arco-normal.mp3\n",
      "Get 98 = viola_A3_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 99 = viola_A4_1_piano_arco-glissando.mp3\n",
      "Get 100 = viola_Gs5_1_forte_arco-normal.mp3\n",
      "Get 1 = saxophone_A4_1_piano_normal.mp3\n",
      "Get 2 = saxophone_A4_1_mezzo-forte_normal.mp3\n",
      "Get 3 = saxophone_A3_1_piano_normal.mp3\n",
      "Get 4 = saxophone_A3_1_fortissimo_normal.mp3\n",
      "Get 5 = saxophone_Gs4_025_fortissimo_normal.mp3\n",
      "Get 6 = saxophone_A5_05_fortissimo_normal.mp3\n",
      "Get 7 = saxophone_Gs3_05_fortissimo_normal.mp3\n",
      "Get 8 = saxophone_A3_05_pianissimo_normal.mp3\n",
      "Get 9 = saxophone_Gs4_15_forte_normal.mp3\n",
      "Get 10 = saxophone_Ds4_05_pianissimo_normal.mp3\n",
      "Get 11 = saxophone_A4_05_fortissimo_normal.mp3\n",
      "Get 12 = saxophone_Gs3_025_fortissimo_normal.mp3\n",
      "Get 13 = saxophone_Gs4_05_piano_normal.mp3\n",
      "Get 14 = saxophone_A5_1_forte_normal.mp3\n",
      "Get 15 = saxophone_A3_025_fortissimo_normal.mp3\n",
      "Get 16 = saxophone_A3_15_fortissimo_normal.mp3\n",
      "Get 17 = saxophone_Gs3_05_mezzo-piano_normal.mp3\n",
      "Get 18 = saxophone_Ds4_025_fortissimo_normal.mp3\n",
      "Get 19 = saxophone_Ds4_05_piano_normal.mp3\n",
      "Get 20 = saxophone_A3_025_mezzo-piano_normal.mp3\n",
      "Get 21 = saxophone_Gs3_05_piano_normal.mp3\n",
      "Get 22 = saxophone_Ds4_025_forte_normal.mp3\n",
      "Get 23 = saxophone_A4_15_forte_normal.mp3\n",
      "Get 24 = saxophone_Gs5_15_pianissimo_normal.mp3\n",
      "Get 25 = saxophone_Gs4_1_pianissimo_normal.mp3\n",
      "Get 26 = saxophone_Ds5_1_forte_major-trill.mp3\n",
      "Get 27 = saxophone_A3_025_pianissimo_normal.mp3\n",
      "Get 28 = saxophone_Gs4_025_forte_normal.mp3\n",
      "Get 29 = saxophone_Ds4_15_fortissimo_normal.mp3\n",
      "Get 30 = saxophone_Ds4_025_mezzo-forte_normal.mp3\n",
      "Get 31 = saxophone_A5_05_forte_normal.mp3\n",
      "Get 32 = saxophone_Gs5_15_fortissimo_normal.mp3\n",
      "Get 33 = saxophone_Gs3_025_forte_normal.mp3\n",
      "Get 34 = saxophone_Gs5_05_pianissimo_normal.mp3\n",
      "Get 35 = saxophone_A5_1_pianissimo_normal.mp3\n",
      "Get 36 = saxophone_Gs5_05_forte_normal.mp3\n",
      "Get 37 = saxophone_Gs4_025_pianissimo_normal.mp3\n",
      "Get 38 = saxophone_Gs5_025_forte_normal.mp3\n",
      "Get 39 = saxophone_Gs3_1_piano_normal.mp3\n",
      "Get 40 = saxophone_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 41 = saxophone_A5_15_pianissimo_normal.mp3\n",
      "Get 42 = saxophone_Gs5_1_forte_normal.mp3\n",
      "Get 43 = saxophone_Gs4_1_forte_normal.mp3\n",
      "Get 44 = saxophone_Gs5_1_fortissimo_normal.mp3\n",
      "Get 45 = saxophone_Gs4_025_piano_normal.mp3\n",
      "Get 46 = saxophone_Gs4_1_piano_normal.mp3\n",
      "Get 47 = saxophone_Gs5_025_fortissimo_normal.mp3\n",
      "Get 48 = saxophone_Gs3_1_pianissimo_normal.mp3\n",
      "Get 49 = saxophone_Gs4_05_fortissimo_normal.mp3\n",
      "Get 50 = saxophone_A4_05_forte_normal.mp3\n",
      "Get 51 = saxophone_Gs3_05_pianissimo_normal.mp3\n",
      "Get 52 = saxophone_Gs5_15_piano_normal.mp3\n",
      "Get 53 = saxophone_A4_025_fortissimo_normal.mp3\n",
      "Get 54 = saxophone_A5_1_piano_normal.mp3\n",
      "Get 55 = saxophone_Gs5_025_piano_normal.mp3\n",
      "Get 56 = saxophone_Gs5_05_fortissimo_normal.mp3\n",
      "Get 57 = saxophone_A5_15_forte_normal.mp3\n",
      "Get 58 = saxophone_Gs4_05_forte_normal.mp3\n",
      "Get 59 = saxophone_Gs5_05_piano_normal.mp3\n",
      "Get 60 = saxophone_Gs5_1_piano_normal.mp3\n",
      "Get 61 = saxophone_A5_05_piano_normal.mp3\n",
      "Get 62 = saxophone_A3_15_pianissimo_normal.mp3\n",
      "Get 63 = saxophone_A5_15_fortissimo_normal.mp3\n",
      "Get 64 = saxophone_Gs5_1_pianissimo_normal.mp3\n",
      "Get 65 = saxophone_Gs4_15_pianissimo_normal.mp3\n",
      "Get 66 = saxophone_Ds4_05_pianissimo_subtone.mp3\n",
      "Get 67 = saxophone_A5_05_pianissimo_normal.mp3\n",
      "Get 68 = saxophone_Ds4_025_pianissimo_normal.mp3\n",
      "Get 69 = saxophone_A3_05_fortissimo_normal.mp3\n",
      "Get 70 = saxophone_Gs4_15_fortissimo_normal.mp3\n",
      "Get 71 = saxophone_A4_05_piano_normal.mp3\n",
      "Get 72 = saxophone_A4_15_piano_normal.mp3\n",
      "Get 73 = saxophone_Ds4_05_mezzo-forte_normal.mp3\n",
      "Get 74 = saxophone_A3_025_forte_normal.mp3\n",
      "Get 75 = saxophone_Gs3_05_forte_normal.mp3\n",
      "Get 76 = saxophone_A3_1_pianissimo_normal.mp3\n",
      "Get 77 = saxophone_A4_025_piano_normal.mp3\n",
      "Get 78 = saxophone_A4_1_pianissimo_normal.mp3\n",
      "Get 79 = saxophone_A4_1_forte_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 80 = saxophone_Gs3_15_fortissimo_normal.mp3\n",
      "Get 81 = saxophone_Ds4_05_mezzo-forte_slap-tongue.mp3\n",
      "Get 82 = saxophone_Gs4_1_fortissimo_normal.mp3\n",
      "Get 83 = saxophone_Gs5_15_forte_normal.mp3\n",
      "Get 84 = saxophone_A4_1_fortissimo_normal.mp3\n",
      "Get 85 = saxophone_A4_05_pianissimo_normal.mp3\n",
      "Get 86 = saxophone_Ds4_05_fortissimo_normal.mp3\n",
      "Get 87 = saxophone_Gs4_15_piano_normal.mp3\n",
      "Get 88 = saxophone_Gs3_15_pianissimo_normal.mp3\n",
      "Get 89 = saxophone_Ds5_1_forte_minor-trill.mp3\n",
      "Get 90 = saxophone_A3_1_forte_normal.mp3\n",
      "Get 91 = saxophone_A3_05_piano_normal.mp3\n",
      "Get 92 = saxophone_A3_05_forte_normal.mp3\n",
      "Get 93 = saxophone_A5_1_fortissimo_normal.mp3\n",
      "Get 94 = saxophone_A4_025_forte_normal.mp3\n",
      "Get 95 = saxophone_A3_05_mezzo-piano_normal.mp3\n",
      "Get 96 = saxophone_A4_15_fortissimo_normal.mp3\n",
      "Get 97 = saxophone_Gs3_025_mezzo-piano_normal.mp3\n",
      "Get 98 = saxophone_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 99 = saxophone_A3_15_forte_normal.mp3\n",
      "Get 100 = saxophone_Gs4_05_pianissimo_normal.mp3\n",
      "Get 1 = oboe_Gs6_025_forte_normal.mp3\n",
      "Get 2 = oboe_Gs5_05_fortissimo_normal.mp3\n",
      "Get 3 = oboe_Gs5_05_forte_normal.mp3\n",
      "Get 4 = oboe_A4_025_fortissimo_normal.mp3\n",
      "Get 5 = oboe_A5_025_fortissimo_normal.mp3\n",
      "Get 6 = oboe_A5_15_forte_normal.mp3\n",
      "Get 7 = oboe_Gs5_15_forte_normal.mp3\n",
      "Get 8 = oboe_A5_15_fortissimo_normal.mp3\n",
      "Get 9 = oboe_Gs6_025_piano_normal.mp3\n",
      "Get 10 = oboe_A5_1_forte_minor-trill.mp3\n",
      "Get 11 = oboe_Gs6_1_piano_normal.mp3\n",
      "Get 12 = oboe_Gs4_025_piano_normal.mp3\n",
      "Get 13 = oboe_A4_1_fortissimo_normal.mp3\n",
      "Get 14 = oboe_Gs5_1_piano_normal.mp3\n",
      "Get 15 = oboe_Gs4_05_fortissimo_normal.mp3\n",
      "Get 16 = oboe_A4_15_piano_normal.mp3\n",
      "Get 17 = oboe_Gs4_025_forte_normal.mp3\n",
      "Get 18 = oboe_Gs6_15_piano_normal.mp3\n",
      "Get 19 = oboe_A6_05_piano_normal.mp3\n",
      "Get 20 = oboe_Gs6_1_mezzo-forte_normal.mp3\n",
      "Get 21 = oboe_Gs6_05_piano_normal.mp3\n",
      "Get 22 = oboe_Gs5_15_mezzo-forte_normal.mp3\n",
      "Get 23 = oboe_A4_025_forte_normal.mp3\n",
      "Get 24 = oboe_A5_05_fortissimo_normal.mp3\n",
      "Get 25 = oboe_As3_1_piano_normal.mp3\n",
      "Get 26 = oboe_A6_15_piano_normal.mp3\n",
      "Get 27 = oboe_A6_025_piano_normal.mp3\n",
      "Get 28 = oboe_As3_05_fortissimo_normal.mp3\n",
      "Get 29 = oboe_G6_025_forte_normal.mp3\n",
      "Get 30 = oboe_A5_05_piano_normal.mp3\n",
      "Get 31 = oboe_A4_05_piano_normal.mp3\n",
      "Get 32 = oboe_Gs5_025_forte_normal.mp3\n",
      "Get 33 = oboe_Ds6_15_forte_normal.mp3\n",
      "Get 34 = oboe_A5_05_forte_normal.mp3\n",
      "Get 35 = oboe_A5_025_piano_normal.mp3\n",
      "Get 36 = oboe_Gs4_15_fortissimo_normal.mp3\n",
      "Get 37 = oboe_Gs5_15_fortissimo_normal.mp3\n",
      "Get 38 = oboe_A5_025_forte_normal.mp3\n",
      "Get 39 = oboe_Gs6_05_forte_normal.mp3\n",
      "Get 40 = oboe_G6_025_piano_normal.mp3\n",
      "Get 41 = oboe_Gs5_15_piano_normal.mp3\n",
      "Get 42 = oboe_A6_1_piano_normal.mp3\n",
      "Get 43 = oboe_A4_15_forte_normal.mp3\n",
      "Get 44 = oboe_Gs5_1_mezzo-forte_normal.mp3\n",
      "Get 45 = oboe_As3_1_mezzo-forte_normal.mp3\n",
      "Get 46 = oboe_Gs4_15_mezzo-forte_normal.mp3\n",
      "Get 47 = oboe_Gs4_025_fortissimo_normal.mp3\n",
      "Get 48 = oboe_A5_15_mezzo-forte_normal.mp3\n",
      "Get 49 = oboe_Gs6_025_mezzo-forte_normal.mp3\n",
      "Get 50 = oboe_Ds6_05_mezzo-forte_normal.mp3\n",
      "Get 51 = oboe_Gs4_1_fortissimo_normal.mp3\n",
      "Get 52 = oboe_As3_05_mezzo-forte_normal.mp3\n",
      "Get 53 = oboe_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 54 = oboe_Ds6_05_fortissimo_normal.mp3\n",
      "Get 55 = oboe_Gs6_15_forte_normal.mp3\n",
      "Get 56 = oboe_Gs4_1_forte_normal.mp3\n",
      "Get 57 = oboe_A4_1_forte_normal.mp3\n",
      "Get 58 = oboe_Gs6_05_mezzo-forte_normal.mp3\n",
      "Get 59 = oboe_As3_15_mezzo-forte_normal.mp3\n",
      "Get 60 = oboe_Gs4_05_forte_normal.mp3\n",
      "Get 61 = oboe_A5_1_forte_major-trill.mp3\n",
      "Get 62 = oboe_A4_05_fortissimo_normal.mp3\n",
      "Get 63 = oboe_Gs5_05_piano_normal.mp3\n",
      "Get 64 = oboe_A5_15_piano_normal.mp3\n",
      "Get 65 = oboe_A6_1_forte_minor-trill.mp3\n",
      "Get 66 = oboe_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 67 = oboe_Gs4_1_forte_minor-trill.mp3\n",
      "Get 68 = oboe_Gs5_1_forte_normal.mp3\n",
      "Get 69 = oboe_Gs5_025_fortissimo_normal.mp3\n",
      "Get 70 = oboe_Gs5_025_piano_normal.mp3\n",
      "Get 71 = oboe_A5_025_mezzo-forte_normal.mp3\n",
      "Get 72 = oboe_Gs5_1_fortissimo_normal.mp3\n",
      "Get 73 = oboe_A4_05_forte_normal.mp3\n",
      "Get 74 = oboe_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 75 = oboe_A5_1_fortissimo_normal.mp3\n",
      "Get 76 = oboe_Gs6_15_mezzo-forte_normal.mp3\n",
      "Get 77 = oboe_A6_1_forte_major-trill.mp3\n",
      "Get 78 = oboe_As3_05_forte_normal.mp3\n",
      "Get 79 = oboe_As3_1_forte_normal.mp3\n",
      "Get 80 = oboe_Gs6_1_forte_normal.mp3\n",
      "Get 81 = oboe_Gs4_1_piano_normal.mp3\n",
      "Get 82 = oboe_As3_15_forte_normal.mp3\n",
      "Get 83 = oboe_As3_05_piano_normal.mp3\n",
      "Get 84 = oboe_A5_1_mezzo-forte_normal.mp3\n",
      "Get 85 = oboe_A4_1_piano_normal.mp3\n",
      "Get 86 = oboe_Gs4_15_forte_normal.mp3\n",
      "Get 87 = oboe_Gs4_05_piano_normal.mp3\n",
      "Get 88 = oboe_Gs4_1_forte_major-trill.mp3\n",
      "Get 89 = oboe_A4_1_mezzo-forte_normal.mp3\n",
      "Get 90 = oboe_A4_025_piano_normal.mp3\n",
      "Get 91 = oboe_A5_05_mezzo-forte_normal.mp3\n",
      "Get 92 = oboe_A5_1_piano_normal.mp3\n",
      "Get 93 = oboe_Gs4_15_piano_normal.mp3\n",
      "Get 94 = oboe_As3_15_fortissimo_normal.mp3\n",
      "Get 95 = oboe_Gs4_1_mezzo-forte_normal.mp3\n",
      "Get 96 = oboe_As3_1_fortissimo_normal.mp3\n",
      "Get 97 = oboe_A5_1_forte_normal.mp3\n",
      "Get 98 = oboe_G6_025_mezzo-forte_normal.mp3\n",
      "Get 99 = oboe_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 100 = oboe_Ds6_05_piano_normal.mp3\n",
      "Get 1 = trumpet_B3_025_forte_normal.mp3\n",
      "Get 2 = trumpet_Cs4_05_pianissimo_normal.mp3\n",
      "Get 3 = trumpet_G3_025_pianissimo_normal.mp3\n",
      "Get 4 = trumpet_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 5 = trumpet_E6_1_forte_normal.mp3\n",
      "Get 6 = trumpet_Ds4_15_pianissimo_normal.mp3\n",
      "Get 7 = trumpet_A3_05_pianissimo_normal.mp3\n",
      "Get 8 = trumpet_Ds4_025_pianissimo_normal.mp3\n",
      "Get 9 = trumpet_B3_05_forte_normal.mp3\n",
      "Get 10 = trumpet_C4_15_pianissimo_normal.mp3\n",
      "Get 11 = trumpet_Ds4_05_pianissimo_normal.mp3\n",
      "Get 12 = trumpet_F3_05_pianissimo_normal.mp3\n",
      "Get 13 = trumpet_As3_1_pianissimo_normal.mp3\n",
      "Get 14 = trumpet_D6_05_forte_normal.mp3\n",
      "Get 15 = trumpet_Cs6_025_forte_normal.mp3\n",
      "Get 16 = trumpet_A3_05_forte_normal.mp3\n",
      "Get 17 = trumpet_E4_025_pianissimo_normal.mp3\n",
      "Get 18 = trumpet_As3_05_pianissimo_normal.mp3\n",
      "Get 19 = trumpet_E3_05_forte_normal.mp3\n",
      "Get 20 = trumpet_G5_1_mezzo-forte_normal.mp3\n",
      "Get 21 = trumpet_G5_025_mezzo-forte_normal.mp3\n",
      "Get 22 = trumpet_B5_025_mezzo-forte_normal.mp3\n",
      "Get 23 = trumpet_F3_025_forte_normal.mp3\n",
      "Get 24 = trumpet_D4_15_pianissimo_normal.mp3\n",
      "Get 25 = trumpet_E4_15_pianissimo_normal.mp3\n",
      "Get 26 = trumpet_Cs4_15_pianissimo_normal.mp3\n",
      "Get 27 = trumpet_D4_05_pianissimo_normal.mp3\n",
      "Get 28 = trumpet_E6_025_forte_normal.mp3\n",
      "Get 29 = trumpet_Ds6_025_forte_normal.mp3\n",
      "Get 30 = trumpet_E3_025_pianissimo_normal.mp3\n",
      "Get 31 = trumpet_B5_025_forte_normal.mp3\n",
      "Get 32 = trumpet_C6_025_forte_normal.mp3\n",
      "Get 33 = trumpet_D4_05_forte_normal.mp3\n",
      "Get 34 = trumpet_As4_025_forte_normal.mp3\n",
      "Get 35 = trumpet_F4_025_pianissimo_normal.mp3\n",
      "Get 36 = trumpet_A3_15_pianissimo_normal.mp3\n",
      "Get 37 = trumpet_Gs3_1_pianissimo_normal.mp3\n",
      "Get 38 = trumpet_Gs3_05_pianissimo_normal.mp3\n",
      "Get 39 = trumpet_F4_05_pianissimo_normal.mp3\n",
      "Get 40 = trumpet_D4_025_forte_normal.mp3\n",
      "Get 41 = trumpet_Cs4_1_pianissimo_normal.mp3\n",
      "Get 42 = trumpet_C4_025_forte_normal.mp3\n",
      "Get 43 = trumpet_C4_1_pianissimo_normal.mp3\n",
      "Get 44 = trumpet_A4_05_forte_normal.mp3\n",
      "Get 45 = trumpet_G3_15_pianissimo_normal.mp3\n",
      "Get 46 = trumpet_Cs4_05_forte_normal.mp3\n",
      "Get 47 = trumpet_Gs3_15_pianissimo_normal.mp3\n",
      "Get 48 = trumpet_B3_1_pianissimo_normal.mp3\n",
      "Get 49 = trumpet_As5_05_forte_normal.mp3\n",
      "Get 50 = trumpet_Ds4_1_pianissimo_normal.mp3\n",
      "Get 51 = trumpet_Gs3_025_pianissimo_normal.mp3\n",
      "Get 52 = trumpet_C4_05_pianissimo_normal.mp3\n",
      "Get 53 = trumpet_Ds4_05_forte_normal.mp3\n",
      "Get 54 = trumpet_E3_15_pianissimo_normal.mp3\n",
      "Get 55 = trumpet_Gs3_05_forte_normal.mp3\n",
      "Get 56 = trumpet_As3_15_pianissimo_normal.mp3\n",
      "Get 57 = trumpet_B4_025_forte_normal.mp3\n",
      "Get 58 = trumpet_As5_05_mezzo-forte_normal.mp3\n",
      "Get 59 = trumpet_E4_1_pianissimo_normal.mp3\n",
      "Get 60 = trumpet_F3_05_forte_normal.mp3\n",
      "Get 61 = trumpet_Gs3_1_forte_normal.mp3\n",
      "Get 62 = trumpet_Ds6_05_forte_normal.mp3\n",
      "Get 63 = trumpet_D6_025_forte_normal.mp3\n",
      "Get 64 = trumpet_Gs5_05_forte_normal.mp3\n",
      "Get 65 = trumpet_Gs5_025_forte_normal.mp3\n",
      "Get 66 = trumpet_B3_05_pianissimo_normal.mp3\n",
      "Get 67 = trumpet_A5_025_mezzo-forte_normal.mp3\n",
      "Get 68 = trumpet_Gs4_025_forte_normal.mp3\n",
      "Get 69 = trumpet_A4_025_forte_normal.mp3\n",
      "Get 70 = trumpet_D4_025_pianissimo_normal.mp3\n",
      "Get 71 = trumpet_E4_05_pianissimo_normal.mp3\n",
      "Get 72 = trumpet_F3_15_pianissimo_normal.mp3\n",
      "Get 73 = trumpet_D5_05_forte_normal.mp3\n",
      "Get 74 = trumpet_Ds5_05_forte_normal.mp3\n",
      "Get 75 = trumpet_As5_025_forte_normal.mp3\n",
      "Get 76 = trumpet_A3_1_pianissimo_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 77 = trumpet_As3_025_forte_normal.mp3\n",
      "Get 78 = trumpet_G3_05_pianissimo_normal.mp3\n",
      "Get 79 = trumpet_C6_05_forte_normal.mp3\n",
      "Get 80 = trumpet_A3_025_pianissimo_normal.mp3\n",
      "Get 81 = trumpet_C4_025_pianissimo_normal.mp3\n",
      "Get 82 = trumpet_F3_1_forte_normal.mp3\n",
      "Get 83 = trumpet_B3_15_pianissimo_normal.mp3\n",
      "Get 84 = trumpet_G5_15_mezzo-forte_normal.mp3\n",
      "Get 85 = trumpet_A5_025_forte_normal.mp3\n",
      "Get 86 = trumpet_Cs4_025_pianissimo_normal.mp3\n",
      "Get 87 = trumpet_As3_025_pianissimo_normal.mp3\n",
      "Get 88 = trumpet_F3_1_pianissimo_normal.mp3\n",
      "Get 89 = trumpet_D4_1_pianissimo_normal.mp3\n",
      "Get 90 = trumpet_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 91 = trumpet_G3_1_pianissimo_normal.mp3\n",
      "Get 92 = trumpet_E3_05_pianissimo_normal.mp3\n",
      "Get 93 = trumpet_A5_05_mezzo-forte_normal.mp3\n",
      "Get 94 = trumpet_G5_05_mezzo-forte_normal.mp3\n",
      "Get 95 = trumpet_C4_05_forte_normal.mp3\n",
      "Get 96 = trumpet_E3_1_pianissimo_normal.mp3\n",
      "Get 97 = trumpet_B3_025_pianissimo_normal.mp3\n",
      "Get 98 = trumpet_F3_025_pianissimo_normal.mp3\n",
      "Get 99 = trumpet_Gs4_05_forte_normal.mp3\n",
      "Get 100 = trumpet_E3_1_forte_normal.mp3\n",
      "Get 1 = flute_Gs5_1_piano_normal.mp3\n",
      "Get 2 = flute_A5_1_piano_normal.mp3\n",
      "Get 3 = flute_Gs4_025_mezzo-piano_normal.mp3\n",
      "Get 4 = flute_A6_05_pianissimo_normal.mp3\n",
      "Get 5 = flute_A5_025_pianissimo_normal.mp3\n",
      "Get 6 = flute_A4_15_mezzo-piano_normal.mp3\n",
      "Get 7 = flute_A5_05_forte_normal.mp3\n",
      "Get 8 = flute_Gs5_15_mezzo-forte_normal.mp3\n",
      "Get 9 = flute_Gs5_1_mezzo-forte_normal.mp3\n",
      "Get 10 = flute_A5_05_mezzo-piano_normal.mp3\n",
      "Get 11 = flute_Gs5_025_mezzo-piano_normal.mp3\n",
      "Get 12 = flute_Gs4_05_pianissimo_normal.mp3\n",
      "Get 13 = flute_Gs5_15_piano_normal.mp3\n",
      "Get 14 = flute_Gs6_1_mezzo-forte_normal.mp3\n",
      "Get 15 = flute_Gs6_15_piano_normal.mp3\n",
      "Get 16 = flute_Gs5_025_piano_normal.mp3\n",
      "Get 17 = flute_A4_025_mezzo-piano_normal.mp3\n",
      "Get 18 = flute_A4_05_mezzo-forte_normal.mp3\n",
      "Get 19 = flute_Gs6_025_piano_normal.mp3\n",
      "Get 20 = flute_Gs4_15_mezzo-forte_normal.mp3\n",
      "Get 21 = flute_Gs6_1_forte_normal.mp3\n",
      "Get 22 = flute_A6_05_piano_normal.mp3\n",
      "Get 23 = flute_Gs4_15_mezzo-piano_normal.mp3\n",
      "Get 24 = flute_A5_15_forte_normal.mp3\n",
      "Get 25 = flute_A4_025_piano_normal.mp3\n",
      "Get 26 = flute_A4_1_mezzo-forte_normal.mp3\n",
      "Get 27 = flute_Gs6_15_mezzo-forte_normal.mp3\n",
      "Get 28 = flute_A6_15_forte_normal.mp3\n",
      "Get 29 = flute_Gs4_025_pianissimo_normal.mp3\n",
      "Get 30 = flute_A5_025_forte_normal.mp3\n",
      "Get 31 = flute_A5_05_pianissimo_normal.mp3\n",
      "Get 32 = flute_Gs5_05_piano_normal.mp3\n",
      "Get 33 = flute_A6_1_fortissimo_minor-trill.mp3\n",
      "Get 34 = flute_Gs4_05_forte_normal.mp3\n",
      "Get 35 = flute_Gs6_1_pianissimo_normal.mp3\n",
      "Get 36 = flute_Gs6_025_pianissimo_normal.mp3\n",
      "Get 37 = flute_A4_05_mezzo-piano_normal.mp3\n",
      "Get 38 = flute_A5_025_piano_normal.mp3\n",
      "Get 39 = flute_A6_1_forte_normal.mp3\n",
      "Get 40 = flute_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 41 = flute_Gs4_025_forte_normal.mp3\n",
      "Get 42 = flute_A5_1_pianissimo_normal.mp3\n",
      "Get 43 = flute_A6_1_mezzo-forte_normal.mp3\n",
      "Get 44 = flute_Gs4_15_piano_normal.mp3\n",
      "Get 45 = flute_Gs5_05_forte_normal.mp3\n",
      "Get 46 = flute_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 47 = flute_Gs6_15_forte_normal.mp3\n",
      "Get 48 = flute_A4_15_forte_normal.mp3\n",
      "Get 49 = flute_A5_15_pianissimo_normal.mp3\n",
      "Get 50 = flute_Gs5_15_mezzo-piano_normal.mp3\n",
      "Get 51 = flute_A4_15_pianissimo_normal.mp3\n",
      "Get 52 = flute_A6_05_forte_normal.mp3\n",
      "Get 53 = flute_A4_15_piano_normal.mp3\n",
      "Get 54 = flute_A5_15_mezzo-forte_normal.mp3\n",
      "Get 55 = flute_A4_1_mezzo-piano_normal.mp3\n",
      "Get 56 = flute_A5_15_mezzo-piano_normal.mp3\n",
      "Get 57 = flute_A5_1_forte_normal.mp3\n",
      "Get 58 = flute_Gs5_1_mezzo-piano_normal.mp3\n",
      "Get 59 = flute_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 60 = flute_Gs5_025_pianissimo_normal.mp3\n",
      "Get 61 = flute_Gs6_05_mezzo-forte_normal.mp3\n",
      "Get 62 = flute_A4_025_pianissimo_normal.mp3\n",
      "Get 63 = flute_Gs5_15_forte_normal.mp3\n",
      "Get 64 = flute_A5_1_mezzo-forte_normal.mp3\n",
      "Get 65 = flute_Gs6_05_piano_normal.mp3\n",
      "Get 66 = flute_A4_05_forte_normal.mp3\n",
      "Get 67 = flute_A5_15_piano_normal.mp3\n",
      "Get 68 = flute_A4_15_mezzo-forte_normal.mp3\n",
      "Get 69 = flute_Gs6_025_forte_normal.mp3\n",
      "Get 70 = flute_A4_05_pianissimo_normal.mp3\n",
      "Get 71 = flute_Gs6_15_pianissimo_normal.mp3\n",
      "Get 72 = flute_Gs4_025_piano_normal.mp3\n",
      "Get 73 = flute_Gs4_05_mezzo-piano_normal.mp3\n",
      "Get 74 = flute_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 75 = flute_Gs6_05_pianissimo_normal.mp3\n",
      "Get 76 = flute_A6_1_piano_normal.mp3\n",
      "Get 77 = flute_Gs4_15_forte_normal.mp3\n",
      "Get 78 = flute_A5_025_mezzo-piano_normal.mp3\n",
      "Get 79 = flute_A5_05_piano_normal.mp3\n",
      "Get 80 = flute_Gs6_025_mezzo-forte_normal.mp3\n",
      "Get 81 = flute_Gs4_05_piano_normal.mp3\n",
      "Get 82 = flute_Gs5_1_pianissimo_normal.mp3\n",
      "Get 83 = flute_Gs6_1_piano_normal.mp3\n",
      "Get 84 = flute_Gs5_05_pianissimo_normal.mp3\n",
      "Get 85 = flute_A4_025_forte_normal.mp3\n",
      "Get 86 = flute_A4_1_pianissimo_normal.mp3\n",
      "Get 87 = flute_A6_1_pianissimo_normal.mp3\n",
      "Get 88 = flute_A4_1_forte_normal.mp3\n",
      "Get 89 = flute_A5_1_mezzo-piano_normal.mp3\n",
      "Get 90 = flute_A4_025_mezzo-forte_normal.mp3\n",
      "Get 91 = flute_Gs6_05_forte_normal.mp3\n",
      "Get 92 = flute_A5_025_mezzo-forte_normal.mp3\n",
      "Get 93 = flute_A4_1_piano_normal.mp3\n",
      "Get 94 = flute_A5_05_mezzo-forte_normal.mp3\n",
      "Get 95 = flute_A6_05_mezzo-forte_normal.mp3\n",
      "Get 96 = flute_Gs5_05_mezzo-piano_normal.mp3\n",
      "Get 97 = flute_Gs5_1_forte_normal.mp3\n",
      "Get 98 = flute_A4_05_piano_normal.mp3\n",
      "Get 99 = flute_Gs5_025_forte_normal.mp3\n",
      "Get 100 = flute_Gs4_15_pianissimo_normal.mp3\n",
      "Get 1 = cello_Gs5_025_mezzo-piano_arco-normal.mp3\n",
      "Get 2 = cello_Gs5_15_forte_arco-normal.mp3\n",
      "Get 3 = cello_A2_05_pianissimo_arco-normal.mp3\n",
      "Get 4 = cello_Gs4_05_pianissimo_arco-normal.mp3\n",
      "Get 5 = cello_A2_1_fortissimo_arco-normal.mp3\n",
      "Get 6 = cello_A3_025_pianissimo_arco-normal.mp3\n",
      "Get 7 = cello_Gs5_05_forte_arco-normal.mp3\n",
      "Get 8 = cello_A3_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 9 = cello_A2_1_forte_arco-normal.mp3\n",
      "Get 10 = cello_Gs4_15_pianissimo_arco-normal.mp3\n",
      "Get 11 = cello_Gs4_025_forte_arco-normal.mp3\n",
      "Get 12 = cello_A3_05_fortissimo_arco-normal.mp3\n",
      "Get 13 = cello_Gs5_05_fortissimo_arco-normal.mp3\n",
      "Get 14 = cello_A3_1_pianissimo_arco-normal.mp3\n",
      "Get 15 = cello_A3_05_forte_arco-normal.mp3\n",
      "Get 16 = cello_Gs3_1_forte_arco-normal.mp3\n",
      "Get 17 = cello_A4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 18 = cello_Gs3_1_mezzo-piano_arco-normal.mp3\n",
      "Get 19 = cello_Gs3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 20 = cello_Gs5_1_forte_arco-normal.mp3\n",
      "Get 21 = cello_Gs3_025_fortissimo_arco-normal.mp3\n",
      "Get 22 = cello_Gs4_05_forte_arco-normal.mp3\n",
      "Get 23 = cello_Gs4_025_fortissimo_arco-normal.mp3\n",
      "Get 24 = cello_A3_15_mezzo-piano_arco-normal.mp3\n",
      "Get 25 = cello_A3_1_mezzo-piano_arco-normal.mp3\n",
      "Get 26 = cello_Gs3_15_fortissimo_arco-normal.mp3\n",
      "Get 27 = cello_Gs5_05_pianissimo_arco-normal.mp3\n",
      "Get 28 = cello_A4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 29 = cello_Gs4_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 30 = cello_A2_05_fortissimo_arco-normal.mp3\n",
      "Get 31 = cello_Gs3_025_forte_arco-normal.mp3\n",
      "Get 32 = cello_Gs4_1_fortissimo_arco-normal.mp3\n",
      "Get 33 = cello_Gs5_1_mezzo-piano_arco-normal.mp3\n",
      "Get 34 = cello_Gs3_05_forte_arco-normal.mp3\n",
      "Get 35 = cello_Gs5_15_pianissimo_arco-normal.mp3\n",
      "Get 36 = cello_Gs5_15_fortissimo_arco-normal.mp3\n",
      "Get 37 = cello_Gs3_15_piano_arco-normal.mp3\n",
      "Get 38 = cello_A3_1_fortissimo_arco-normal.mp3\n",
      "Get 39 = cello_Gs3_15_pianissimo_arco-normal.mp3\n",
      "Get 40 = cello_Gs5_025_forte_arco-normal.mp3\n",
      "Get 41 = cello_Gs3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 42 = cello_A3_1_mezzo-piano_arco-minor-trill.mp3\n",
      "Get 43 = cello_A4_1_mezzo-forte_arco-harmonic.mp3\n",
      "Get 44 = cello_A4_025_pianissimo_arco-normal.mp3\n",
      "Get 45 = cello_A2_025_mezzo-piano_arco-normal.mp3\n",
      "Get 46 = cello_Gs4_1_forte_arco-normal.mp3\n",
      "Get 47 = cello_Gs3_025_pianissimo_arco-normal.mp3\n",
      "Get 48 = cello_Gs5_15_mezzo-piano_arco-normal.mp3\n",
      "Get 49 = cello_A3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 50 = cello_A3_025_fortissimo_arco-normal.mp3\n",
      "Get 51 = cello_Gs5_1_mezzo-forte_arco-harmonic.mp3\n",
      "Get 52 = cello_Gs5_05_mezzo-piano_arco-normal.mp3\n",
      "Get 53 = cello_A4_025_fortissimo_arco-normal.mp3\n",
      "Get 54 = cello_Gs5_025_fortissimo_arco-normal.mp3\n",
      "Get 55 = cello_A4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 56 = cello_A4_1_forte_arco-normal.mp3\n",
      "Get 57 = cello_Gs4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 58 = cello_A4_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 59 = cello_A4_1_mezzo-piano_arco-minor-trill.mp3\n",
      "Get 60 = cello_Gs4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 61 = cello_A4_025_forte_arco-normal.mp3\n",
      "Get 62 = cello_Gs5_025_pianissimo_arco-normal.mp3\n",
      "Get 63 = cello_A2_05_mezzo-piano_arco-normal.mp3\n",
      "Get 64 = cello_A4_15_mezzo-piano_arco-normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 65 = cello_Gs4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 66 = cello_A4_05_forte_arco-normal.mp3\n",
      "Get 67 = cello_A2_15_forte_arco-normal.mp3\n",
      "Get 68 = cello_Gs3_15_forte_arco-normal.mp3\n",
      "Get 69 = cello_Gs4_1_pianissimo_arco-normal.mp3\n",
      "Get 70 = cello_Gs4_15_mezzo-piano_arco-normal.mp3\n",
      "Get 71 = cello_A3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 72 = cello_A3_15_forte_arco-normal.mp3\n",
      "Get 73 = cello_A2_15_pianissimo_arco-normal.mp3\n",
      "Get 74 = cello_Gs5_1_fortissimo_arco-normal.mp3\n",
      "Get 75 = cello_A2_025_pianissimo_arco-normal.mp3\n",
      "Get 76 = cello_A2_1_pianissimo_arco-normal.mp3\n",
      "Get 77 = cello_Gs4_05_fortissimo_arco-normal.mp3\n",
      "Get 78 = cello_A4_1_fortissimo_arco-normal.mp3\n",
      "Get 79 = cello_Gs3_05_pianissimo_arco-normal.mp3\n",
      "Get 80 = cello_A3_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 81 = cello_Gs4_025_pianissimo_arco-normal.mp3\n",
      "Get 82 = cello_A2_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 83 = cello_A4_1_mezzo-piano_molto-vibrato.mp3\n",
      "Get 84 = cello_Gs4_15_forte_arco-normal.mp3\n",
      "Get 85 = cello_Gs3_1_pianissimo_arco-normal.mp3\n",
      "Get 86 = cello_Gs5_1_pianissimo_arco-normal.mp3\n",
      "Get 87 = cello_A4_15_forte_arco-normal.mp3\n",
      "Get 88 = cello_A2_025_forte_arco-normal.mp3\n",
      "Get 89 = cello_A2_025_fortissimo_arco-normal.mp3\n",
      "Get 90 = cello_A2_15_piano_arco-normal.mp3\n",
      "Get 91 = cello_A2_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 92 = cello_A2_05_forte_arco-normal.mp3\n",
      "Get 93 = cello_A4_1_pianissimo_arco-normal.mp3\n",
      "Get 94 = cello_Gs4_15_fortissimo_arco-normal.mp3\n",
      "Get 95 = cello_A3_1_forte_arco-normal.mp3\n",
      "Get 96 = cello_Gs3_05_fortissimo_arco-normal.mp3\n",
      "Get 97 = cello_Gs2_025_pianissimo_arco-normal.mp3\n",
      "Get 98 = cello_A3_025_forte_arco-normal.mp3\n",
      "Get 99 = cello_A2_1_mezzo-piano_arco-normal.mp3\n",
      "Get 100 = cello_A4_05_pianissimo_arco-normal.mp3\n",
      "found 600 audio files in ./audio/london_phill_dataset_multi/\n"
     ]
    }
   ],
   "source": [
    "#Find Audio Files\n",
    "files = []\n",
    "labels =[]\n",
    "duration = []\n",
    "classes=['flute','sax','oboe', 'cello','trumpet','viola']\n",
    "for root, dirnames, filenames in os.walk(path):\n",
    "    for i, filename in enumerate(fnmatch.filter(filenames, '*.mp3')):\n",
    "        files.append(os.path.join(root, filename))\n",
    "        for name in classes:\n",
    "            if fnmatch.fnmatchcase(filename, '*'+name+'*'):\n",
    "                labels.append(name)\n",
    "                break\n",
    "        else:\n",
    "            labels.append('other')\n",
    "        print (\"Get %d = %s\"%(i+1, filename))\n",
    "        try:\n",
    "            y, sr = librosa.load(files[i], sr=fs)\n",
    "            if len(y) < 2:\n",
    "                print(\"Error loading %s\" % filename)\n",
    "                continue\n",
    "            #y/=y.max() #Normalize\n",
    "            yt, index = librosa.effects.trim(y,top_db=60) #Trim\n",
    "            duration.append(librosa.get_duration(yt, sr=fs))\n",
    "        except Exception as e:\n",
    "            print(\"Error loading %s. Error: %s\" % (filename,e))\n",
    "\n",
    "\n",
    "print(\"found %d audio files in %s\"%(len(files),path))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max. Duration: 3.2536961451247164\n",
      "Min. Duration: 0.3657142857142857\n",
      "Average Duration: 1.082833560090703\n"
     ]
    }
   ],
   "source": [
    "print(\"Max. Duration:\", max(duration))\n",
    "print(\"Min. Duration:\", min(duration))\n",
    "print(\"Average Duration:\", np.mean(duration))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Trim Silence and Recalculate Duration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 1  ./audio/london_phill_dataset_multi/viola/viola_Gs4_025_pianissimo_arco-normal.mp3\n",
      "Get 2  ./audio/london_phill_dataset_multi/viola/viola_A4_1_fortissimo_snap-pizz.mp3\n",
      "Get 3  ./audio/london_phill_dataset_multi/viola/viola_A3_1_piano_pizz-normal.mp3\n",
      "Get 4  ./audio/london_phill_dataset_multi/viola/viola_Gs4_05_pianissimo_arco-normal.mp3\n",
      "Get 5  ./audio/london_phill_dataset_multi/viola/viola_Gs3_025_fortissimo_arco-normal.mp3\n",
      "Get 6  ./audio/london_phill_dataset_multi/viola/viola_A3_1_piano_arco-normal.mp3\n",
      "Get 7  ./audio/london_phill_dataset_multi/viola/viola_A4_025_piano_arco-normal.mp3\n",
      "Get 8  ./audio/london_phill_dataset_multi/viola/viola_A3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 9  ./audio/london_phill_dataset_multi/viola/viola_Gs4_15_mezzo-piano_arco-normal.mp3\n",
      "Get 10  ./audio/london_phill_dataset_multi/viola/viola_A5_1_fortissimo_arco-normal.mp3\n",
      "Get 11  ./audio/london_phill_dataset_multi/viola/viola_Gs4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 12  ./audio/london_phill_dataset_multi/viola/viola_Gs5_05_fortissimo_arco-normal.mp3\n",
      "Get 13  ./audio/london_phill_dataset_multi/viola/viola_Gs5_1_fortissimo_snap-pizz.mp3\n",
      "Get 14  ./audio/london_phill_dataset_multi/viola/viola_Gs5_025_mezzo-piano_arco-normal.mp3\n",
      "Get 15  ./audio/london_phill_dataset_multi/viola/viola_A3_1_fortissimo_arco-normal.mp3\n",
      "Get 16  ./audio/london_phill_dataset_multi/viola/viola_A4_15_fortissimo_arco-normal.mp3\n",
      "Get 17  ./audio/london_phill_dataset_multi/viola/viola_A4_05_forte_arco-normal.mp3\n",
      "Get 18  ./audio/london_phill_dataset_multi/viola/viola_A5_1_pianissimo_arco-normal.mp3\n",
      "Get 19  ./audio/london_phill_dataset_multi/viola/viola_A4_1_piano_arco-normal.mp3\n",
      "Get 20  ./audio/london_phill_dataset_multi/viola/viola_A3_025_fortissimo_arco-normal.mp3\n",
      "Get 21  ./audio/london_phill_dataset_multi/viola/viola_Gs6_025_pianissimo_arco-normal.mp3\n",
      "Get 22  ./audio/london_phill_dataset_multi/viola/viola_Gs6_05_fortissimo_arco-normal.mp3\n",
      "Get 23  ./audio/london_phill_dataset_multi/viola/viola_Gs6_05_pianissimo_arco-normal.mp3\n",
      "Get 24  ./audio/london_phill_dataset_multi/viola/viola_Gs4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 25  ./audio/london_phill_dataset_multi/viola/viola_Gs5_1_fortissimo_arco-normal.mp3\n",
      "Get 26  ./audio/london_phill_dataset_multi/viola/viola_A3_1_forte_arco-normal.mp3\n",
      "Get 27  ./audio/london_phill_dataset_multi/viola/viola_Gs3_15_fortissimo_arco-normal.mp3\n",
      "Get 28  ./audio/london_phill_dataset_multi/viola/viola_Gs4_025_fortissimo_arco-normal.mp3\n",
      "Get 29  ./audio/london_phill_dataset_multi/viola/viola_Gs6_05_piano_arco-normal.mp3\n",
      "Get 30  ./audio/london_phill_dataset_multi/viola/viola_Gs4_15_pianissimo_arco-normal.mp3\n",
      "Get 31  ./audio/london_phill_dataset_multi/viola/viola_A4_05_fortissimo_arco-normal.mp3\n",
      "Get 32  ./audio/london_phill_dataset_multi/viola/viola_A5_1_mezzo-forte_artificial-harmonic.mp3\n",
      "Get 33  ./audio/london_phill_dataset_multi/viola/viola_Gs5_05_piano_arco-normal.mp3\n",
      "Get 34  ./audio/london_phill_dataset_multi/viola/viola_A4_1_mezzo-forte_arco-major-trill.mp3\n",
      "Get 35  ./audio/london_phill_dataset_multi/viola/viola_Gs3_025_piano_arco-normal.mp3\n",
      "Get 36  ./audio/london_phill_dataset_multi/viola/viola_Gs4_1_piano_arco-normal.mp3\n",
      "Get 37  ./audio/london_phill_dataset_multi/viola/viola_Gs5_15_fortissimo_arco-normal.mp3\n",
      "Get 38  ./audio/london_phill_dataset_multi/viola/viola_A4_05_pianissimo_arco-normal.mp3\n",
      "Get 39  ./audio/london_phill_dataset_multi/viola/viola_A4_05_piano_arco-normal.mp3\n",
      "Get 40  ./audio/london_phill_dataset_multi/viola/viola_Ds3_1_piano_pizz-normal.mp3\n",
      "Get 41  ./audio/london_phill_dataset_multi/viola/viola_A3_1_pianissimo_arco-normal.mp3\n",
      "Get 42  ./audio/london_phill_dataset_multi/viola/viola_A3_15_fortissimo_arco-normal.mp3\n",
      "Get 43  ./audio/london_phill_dataset_multi/viola/viola_A5_1_forte_arco-normal.mp3\n",
      "Get 44  ./audio/london_phill_dataset_multi/viola/viola_Gs3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 45  ./audio/london_phill_dataset_multi/viola/viola_Gs5_15_pianissimo_arco-normal.mp3\n",
      "Get 46  ./audio/london_phill_dataset_multi/viola/viola_A3_05_fortissimo_arco-normal.mp3\n",
      "Get 47  ./audio/london_phill_dataset_multi/viola/viola_Ds3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 48  ./audio/london_phill_dataset_multi/viola/viola_Gs6_1_pianissimo_arco-normal.mp3\n",
      "Get 49  ./audio/london_phill_dataset_multi/viola/viola_Gs3_15_pianissimo_arco-normal.mp3\n",
      "Get 50  ./audio/london_phill_dataset_multi/viola/viola_A3_1_mezzo-forte_molto-vibrato.mp3\n",
      "Get 51  ./audio/london_phill_dataset_multi/viola/viola_Gs6_025_fortissimo_arco-normal.mp3\n",
      "Get 52  ./audio/london_phill_dataset_multi/viola/viola_A3_1_fortissimo_pizz-normal.mp3\n",
      "Get 53  ./audio/london_phill_dataset_multi/viola/viola_Gs6_1_fortissimo_arco-normal.mp3\n",
      "Get 54  ./audio/london_phill_dataset_multi/viola/viola_A3_15_piano_arco-normal.mp3\n",
      "Get 55  ./audio/london_phill_dataset_multi/viola/viola_A4_1_mezzo-forte_arco-minor-trill.mp3\n",
      "Get 56  ./audio/london_phill_dataset_multi/viola/viola_A4_1_fortissimo_arco-normal.mp3\n",
      "Get 57  ./audio/london_phill_dataset_multi/viola/viola_A3_025_piano_arco-normal.mp3\n",
      "Get 58  ./audio/london_phill_dataset_multi/viola/viola_A3_05_pianissimo_arco-normal.mp3\n",
      "Get 59  ./audio/london_phill_dataset_multi/viola/viola_A4_1_piano_pizz-normal.mp3\n",
      "Get 60  ./audio/london_phill_dataset_multi/viola/viola_A3_1_piano_arco-glissando.mp3\n",
      "Get 61  ./audio/london_phill_dataset_multi/viola/viola_A4_1_pianissimo_arco-normal.mp3\n",
      "Get 62  ./audio/london_phill_dataset_multi/viola/viola_Gs6_1_forte_arco-normal.mp3\n",
      "Get 63  ./audio/london_phill_dataset_multi/viola/viola_Gs5_05_forte_arco-normal.mp3\n",
      "Get 64  ./audio/london_phill_dataset_multi/viola/viola_A4_1_forte_arco-normal.mp3\n",
      "Get 65  ./audio/london_phill_dataset_multi/viola/viola_A3_05_piano_arco-normal.mp3\n",
      "Get 66  ./audio/london_phill_dataset_multi/viola/viola_Gs5_05_pianissimo_arco-normal.mp3\n",
      "Get 67  ./audio/london_phill_dataset_multi/viola/viola_Gs4_1_fortissimo_arco-normal.mp3\n",
      "Get 68  ./audio/london_phill_dataset_multi/viola/viola_Gs5_025_fortissimo_arco-normal.mp3\n",
      "Get 69  ./audio/london_phill_dataset_multi/viola/viola_Gs5_025_pianissimo_arco-normal.mp3\n",
      "Get 70  ./audio/london_phill_dataset_multi/viola/viola_Gs5_1_piano_arco-normal.mp3\n",
      "Get 71  ./audio/london_phill_dataset_multi/viola/viola_Gs3_025_pianissimo_arco-normal.mp3\n",
      "Get 72  ./audio/london_phill_dataset_multi/viola/viola_Gs4_1_pianissimo_arco-normal.mp3\n",
      "Get 73  ./audio/london_phill_dataset_multi/viola/viola_A4_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 74  ./audio/london_phill_dataset_multi/viola/viola_Gs4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 75  ./audio/london_phill_dataset_multi/viola/viola_A4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 76  ./audio/london_phill_dataset_multi/viola/viola_A4_025_pianissimo_arco-normal.mp3\n",
      "Get 77  ./audio/london_phill_dataset_multi/viola/viola_A4_15_piano_arco-normal.mp3\n",
      "Get 78  ./audio/london_phill_dataset_multi/viola/viola_Gs6_05_forte_arco-normal.mp3\n",
      "Get 79  ./audio/london_phill_dataset_multi/viola/viola_Gs6_15_fortissimo_arco-normal.mp3\n",
      "Get 80  ./audio/london_phill_dataset_multi/viola/viola_Gs6_025_forte_arco-normal.mp3\n",
      "Get 81  ./audio/london_phill_dataset_multi/viola/viola_A4_15_pianissimo_arco-normal.mp3\n",
      "Get 82  ./audio/london_phill_dataset_multi/viola/viola_Gs3_15_piano_arco-normal.mp3\n",
      "Get 83  ./audio/london_phill_dataset_multi/viola/viola_A3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 84  ./audio/london_phill_dataset_multi/viola/viola_Gs6_025_mezzo-piano_arco-normal.mp3\n",
      "Get 85  ./audio/london_phill_dataset_multi/viola/viola_Gs6_15_forte_arco-normal.mp3\n",
      "Get 86  ./audio/london_phill_dataset_multi/viola/viola_Gs4_15_fortissimo_arco-normal.mp3\n",
      "Get 87  ./audio/london_phill_dataset_multi/viola/viola_A4_025_fortissimo_arco-normal.mp3\n",
      "Get 88  ./audio/london_phill_dataset_multi/viola/viola_Ds3_05_fortissimo_arco-normal.mp3\n",
      "Get 89  ./audio/london_phill_dataset_multi/viola/viola_A4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 90  ./audio/london_phill_dataset_multi/viola/viola_Gs5_1_pianissimo_arco-normal.mp3\n",
      "Get 91  ./audio/london_phill_dataset_multi/viola/viola_Gs4_05_fortissimo_arco-normal.mp3\n",
      "Get 92  ./audio/london_phill_dataset_multi/viola/viola_Gs6_025_piano_arco-normal.mp3\n",
      "Get 93  ./audio/london_phill_dataset_multi/viola/viola_A4_1_mezzo-forte_molto-vibrato.mp3\n",
      "Get 94  ./audio/london_phill_dataset_multi/viola/viola_A3_15_forte_arco-normal.mp3\n",
      "Get 95  ./audio/london_phill_dataset_multi/viola/viola_Gs5_025_forte_arco-normal.mp3\n",
      "Get 96  ./audio/london_phill_dataset_multi/viola/viola_A3_025_pianissimo_arco-normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 97  ./audio/london_phill_dataset_multi/viola/viola_Gs4_05_piano_arco-normal.mp3\n",
      "Get 98  ./audio/london_phill_dataset_multi/viola/viola_A3_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 99  ./audio/london_phill_dataset_multi/viola/viola_A4_1_piano_arco-glissando.mp3\n",
      "Get 100  ./audio/london_phill_dataset_multi/viola/viola_Gs5_1_forte_arco-normal.mp3\n",
      "Get 101  ./audio/london_phill_dataset_multi/sax/saxophone_A4_1_piano_normal.mp3\n",
      "Get 102  ./audio/london_phill_dataset_multi/sax/saxophone_A4_1_mezzo-forte_normal.mp3\n",
      "Get 103  ./audio/london_phill_dataset_multi/sax/saxophone_A3_1_piano_normal.mp3\n",
      "Get 104  ./audio/london_phill_dataset_multi/sax/saxophone_A3_1_fortissimo_normal.mp3\n",
      "Get 105  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_025_fortissimo_normal.mp3\n",
      "Get 106  ./audio/london_phill_dataset_multi/sax/saxophone_A5_05_fortissimo_normal.mp3\n",
      "Get 107  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_05_fortissimo_normal.mp3\n",
      "Get 108  ./audio/london_phill_dataset_multi/sax/saxophone_A3_05_pianissimo_normal.mp3\n",
      "Get 109  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_15_forte_normal.mp3\n",
      "Get 110  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_pianissimo_normal.mp3\n",
      "Get 111  ./audio/london_phill_dataset_multi/sax/saxophone_A4_05_fortissimo_normal.mp3\n",
      "Get 112  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_025_fortissimo_normal.mp3\n",
      "Get 113  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_05_piano_normal.mp3\n",
      "Get 114  ./audio/london_phill_dataset_multi/sax/saxophone_A5_1_forte_normal.mp3\n",
      "Get 115  ./audio/london_phill_dataset_multi/sax/saxophone_A3_025_fortissimo_normal.mp3\n",
      "Get 116  ./audio/london_phill_dataset_multi/sax/saxophone_A3_15_fortissimo_normal.mp3\n",
      "Get 117  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_05_mezzo-piano_normal.mp3\n",
      "Get 118  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_025_fortissimo_normal.mp3\n",
      "Get 119  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_piano_normal.mp3\n",
      "Get 120  ./audio/london_phill_dataset_multi/sax/saxophone_A3_025_mezzo-piano_normal.mp3\n",
      "Get 121  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_05_piano_normal.mp3\n",
      "Get 122  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_025_forte_normal.mp3\n",
      "Get 123  ./audio/london_phill_dataset_multi/sax/saxophone_A4_15_forte_normal.mp3\n",
      "Get 124  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_15_pianissimo_normal.mp3\n",
      "Get 125  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_1_pianissimo_normal.mp3\n",
      "Get 126  ./audio/london_phill_dataset_multi/sax/saxophone_Ds5_1_forte_major-trill.mp3\n",
      "Get 127  ./audio/london_phill_dataset_multi/sax/saxophone_A3_025_pianissimo_normal.mp3\n",
      "Get 128  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_025_forte_normal.mp3\n",
      "Get 129  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_15_fortissimo_normal.mp3\n",
      "Get 130  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_025_mezzo-forte_normal.mp3\n",
      "Get 131  ./audio/london_phill_dataset_multi/sax/saxophone_A5_05_forte_normal.mp3\n",
      "Get 132  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_15_fortissimo_normal.mp3\n",
      "Get 133  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_025_forte_normal.mp3\n",
      "Get 134  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_05_pianissimo_normal.mp3\n",
      "Get 135  ./audio/london_phill_dataset_multi/sax/saxophone_A5_1_pianissimo_normal.mp3\n",
      "Get 136  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_05_forte_normal.mp3\n",
      "Get 137  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_025_pianissimo_normal.mp3\n",
      "Get 138  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_025_forte_normal.mp3\n",
      "Get 139  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_1_piano_normal.mp3\n",
      "Get 140  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 141  ./audio/london_phill_dataset_multi/sax/saxophone_A5_15_pianissimo_normal.mp3\n",
      "Get 142  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_1_forte_normal.mp3\n",
      "Get 143  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_1_forte_normal.mp3\n",
      "Get 144  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_1_fortissimo_normal.mp3\n",
      "Get 145  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_025_piano_normal.mp3\n",
      "Get 146  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_1_piano_normal.mp3\n",
      "Get 147  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_025_fortissimo_normal.mp3\n",
      "Get 148  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_1_pianissimo_normal.mp3\n",
      "Get 149  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_05_fortissimo_normal.mp3\n",
      "Get 150  ./audio/london_phill_dataset_multi/sax/saxophone_A4_05_forte_normal.mp3\n",
      "Get 151  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_05_pianissimo_normal.mp3\n",
      "Get 152  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_15_piano_normal.mp3\n",
      "Get 153  ./audio/london_phill_dataset_multi/sax/saxophone_A4_025_fortissimo_normal.mp3\n",
      "Get 154  ./audio/london_phill_dataset_multi/sax/saxophone_A5_1_piano_normal.mp3\n",
      "Get 155  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_025_piano_normal.mp3\n",
      "Get 156  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_05_fortissimo_normal.mp3\n",
      "Get 157  ./audio/london_phill_dataset_multi/sax/saxophone_A5_15_forte_normal.mp3\n",
      "Get 158  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_05_forte_normal.mp3\n",
      "Get 159  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_05_piano_normal.mp3\n",
      "Get 160  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_1_piano_normal.mp3\n",
      "Get 161  ./audio/london_phill_dataset_multi/sax/saxophone_A5_05_piano_normal.mp3\n",
      "Get 162  ./audio/london_phill_dataset_multi/sax/saxophone_A3_15_pianissimo_normal.mp3\n",
      "Get 163  ./audio/london_phill_dataset_multi/sax/saxophone_A5_15_fortissimo_normal.mp3\n",
      "Get 164  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_1_pianissimo_normal.mp3\n",
      "Get 165  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_15_pianissimo_normal.mp3\n",
      "Get 166  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_pianissimo_subtone.mp3\n",
      "Get 167  ./audio/london_phill_dataset_multi/sax/saxophone_A5_05_pianissimo_normal.mp3\n",
      "Get 168  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_025_pianissimo_normal.mp3\n",
      "Get 169  ./audio/london_phill_dataset_multi/sax/saxophone_A3_05_fortissimo_normal.mp3\n",
      "Get 170  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_15_fortissimo_normal.mp3\n",
      "Get 171  ./audio/london_phill_dataset_multi/sax/saxophone_A4_05_piano_normal.mp3\n",
      "Get 172  ./audio/london_phill_dataset_multi/sax/saxophone_A4_15_piano_normal.mp3\n",
      "Get 173  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_mezzo-forte_normal.mp3\n",
      "Get 174  ./audio/london_phill_dataset_multi/sax/saxophone_A3_025_forte_normal.mp3\n",
      "Get 175  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_05_forte_normal.mp3\n",
      "Get 176  ./audio/london_phill_dataset_multi/sax/saxophone_A3_1_pianissimo_normal.mp3\n",
      "Get 177  ./audio/london_phill_dataset_multi/sax/saxophone_A4_025_piano_normal.mp3\n",
      "Get 178  ./audio/london_phill_dataset_multi/sax/saxophone_A4_1_pianissimo_normal.mp3\n",
      "Get 179  ./audio/london_phill_dataset_multi/sax/saxophone_A4_1_forte_normal.mp3\n",
      "Get 180  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_15_fortissimo_normal.mp3\n",
      "Get 181  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_mezzo-forte_slap-tongue.mp3\n",
      "Get 182  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_1_fortissimo_normal.mp3\n",
      "Get 183  ./audio/london_phill_dataset_multi/sax/saxophone_Gs5_15_forte_normal.mp3\n",
      "Get 184  ./audio/london_phill_dataset_multi/sax/saxophone_A4_1_fortissimo_normal.mp3\n",
      "Get 185  ./audio/london_phill_dataset_multi/sax/saxophone_A4_05_pianissimo_normal.mp3\n",
      "Get 186  ./audio/london_phill_dataset_multi/sax/saxophone_Ds4_05_fortissimo_normal.mp3\n",
      "Get 187  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_15_piano_normal.mp3\n",
      "Get 188  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_15_pianissimo_normal.mp3\n",
      "Get 189  ./audio/london_phill_dataset_multi/sax/saxophone_Ds5_1_forte_minor-trill.mp3\n",
      "Get 190  ./audio/london_phill_dataset_multi/sax/saxophone_A3_1_forte_normal.mp3\n",
      "Get 191  ./audio/london_phill_dataset_multi/sax/saxophone_A3_05_piano_normal.mp3\n",
      "Get 192  ./audio/london_phill_dataset_multi/sax/saxophone_A3_05_forte_normal.mp3\n",
      "Get 193  ./audio/london_phill_dataset_multi/sax/saxophone_A5_1_fortissimo_normal.mp3\n",
      "Get 194  ./audio/london_phill_dataset_multi/sax/saxophone_A4_025_forte_normal.mp3\n",
      "Get 195  ./audio/london_phill_dataset_multi/sax/saxophone_A3_05_mezzo-piano_normal.mp3\n",
      "Get 196  ./audio/london_phill_dataset_multi/sax/saxophone_A4_15_fortissimo_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 197  ./audio/london_phill_dataset_multi/sax/saxophone_Gs3_025_mezzo-piano_normal.mp3\n",
      "Get 198  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 199  ./audio/london_phill_dataset_multi/sax/saxophone_A3_15_forte_normal.mp3\n",
      "Get 200  ./audio/london_phill_dataset_multi/sax/saxophone_Gs4_05_pianissimo_normal.mp3\n",
      "Get 201  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_025_forte_normal.mp3\n",
      "Get 202  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_05_fortissimo_normal.mp3\n",
      "Get 203  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_05_forte_normal.mp3\n",
      "Get 204  ./audio/london_phill_dataset_multi/oboe/oboe_A4_025_fortissimo_normal.mp3\n",
      "Get 205  ./audio/london_phill_dataset_multi/oboe/oboe_A5_025_fortissimo_normal.mp3\n",
      "Get 206  ./audio/london_phill_dataset_multi/oboe/oboe_A5_15_forte_normal.mp3\n",
      "Get 207  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_15_forte_normal.mp3\n",
      "Get 208  ./audio/london_phill_dataset_multi/oboe/oboe_A5_15_fortissimo_normal.mp3\n",
      "Get 209  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_025_piano_normal.mp3\n",
      "Get 210  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_forte_minor-trill.mp3\n",
      "Get 211  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_1_piano_normal.mp3\n",
      "Get 212  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_025_piano_normal.mp3\n",
      "Get 213  ./audio/london_phill_dataset_multi/oboe/oboe_A4_1_fortissimo_normal.mp3\n",
      "Get 214  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_1_piano_normal.mp3\n",
      "Get 215  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_05_fortissimo_normal.mp3\n",
      "Get 216  ./audio/london_phill_dataset_multi/oboe/oboe_A4_15_piano_normal.mp3\n",
      "Get 217  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_025_forte_normal.mp3\n",
      "Get 218  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_15_piano_normal.mp3\n",
      "Get 219  ./audio/london_phill_dataset_multi/oboe/oboe_A6_05_piano_normal.mp3\n",
      "Get 220  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_1_mezzo-forte_normal.mp3\n",
      "Get 221  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_05_piano_normal.mp3\n",
      "Get 222  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_15_mezzo-forte_normal.mp3\n",
      "Get 223  ./audio/london_phill_dataset_multi/oboe/oboe_A4_025_forte_normal.mp3\n",
      "Get 224  ./audio/london_phill_dataset_multi/oboe/oboe_A5_05_fortissimo_normal.mp3\n",
      "Get 225  ./audio/london_phill_dataset_multi/oboe/oboe_As3_1_piano_normal.mp3\n",
      "Get 226  ./audio/london_phill_dataset_multi/oboe/oboe_A6_15_piano_normal.mp3\n",
      "Get 227  ./audio/london_phill_dataset_multi/oboe/oboe_A6_025_piano_normal.mp3\n",
      "Get 228  ./audio/london_phill_dataset_multi/oboe/oboe_As3_05_fortissimo_normal.mp3\n",
      "Get 229  ./audio/london_phill_dataset_multi/oboe/oboe_G6_025_forte_normal.mp3\n",
      "Get 230  ./audio/london_phill_dataset_multi/oboe/oboe_A5_05_piano_normal.mp3\n",
      "Get 231  ./audio/london_phill_dataset_multi/oboe/oboe_A4_05_piano_normal.mp3\n",
      "Get 232  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_025_forte_normal.mp3\n",
      "Get 233  ./audio/london_phill_dataset_multi/oboe/oboe_Ds6_15_forte_normal.mp3\n",
      "Get 234  ./audio/london_phill_dataset_multi/oboe/oboe_A5_05_forte_normal.mp3\n",
      "Get 235  ./audio/london_phill_dataset_multi/oboe/oboe_A5_025_piano_normal.mp3\n",
      "Get 236  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_15_fortissimo_normal.mp3\n",
      "Get 237  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_15_fortissimo_normal.mp3\n",
      "Get 238  ./audio/london_phill_dataset_multi/oboe/oboe_A5_025_forte_normal.mp3\n",
      "Get 239  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_05_forte_normal.mp3\n",
      "Get 240  ./audio/london_phill_dataset_multi/oboe/oboe_G6_025_piano_normal.mp3\n",
      "Get 241  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_15_piano_normal.mp3\n",
      "Get 242  ./audio/london_phill_dataset_multi/oboe/oboe_A6_1_piano_normal.mp3\n",
      "Get 243  ./audio/london_phill_dataset_multi/oboe/oboe_A4_15_forte_normal.mp3\n",
      "Get 244  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_1_mezzo-forte_normal.mp3\n",
      "Get 245  ./audio/london_phill_dataset_multi/oboe/oboe_As3_1_mezzo-forte_normal.mp3\n",
      "Get 246  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_15_mezzo-forte_normal.mp3\n",
      "Get 247  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_025_fortissimo_normal.mp3\n",
      "Get 248  ./audio/london_phill_dataset_multi/oboe/oboe_A5_15_mezzo-forte_normal.mp3\n",
      "Get 249  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_025_mezzo-forte_normal.mp3\n",
      "Get 250  ./audio/london_phill_dataset_multi/oboe/oboe_Ds6_05_mezzo-forte_normal.mp3\n",
      "Get 251  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_fortissimo_normal.mp3\n",
      "Get 252  ./audio/london_phill_dataset_multi/oboe/oboe_As3_05_mezzo-forte_normal.mp3\n",
      "Get 253  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 254  ./audio/london_phill_dataset_multi/oboe/oboe_Ds6_05_fortissimo_normal.mp3\n",
      "Get 255  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_15_forte_normal.mp3\n",
      "Get 256  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_forte_normal.mp3\n",
      "Get 257  ./audio/london_phill_dataset_multi/oboe/oboe_A4_1_forte_normal.mp3\n",
      "Get 258  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_05_mezzo-forte_normal.mp3\n",
      "Get 259  ./audio/london_phill_dataset_multi/oboe/oboe_As3_15_mezzo-forte_normal.mp3\n",
      "Get 260  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_05_forte_normal.mp3\n",
      "Get 261  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_forte_major-trill.mp3\n",
      "Get 262  ./audio/london_phill_dataset_multi/oboe/oboe_A4_05_fortissimo_normal.mp3\n",
      "Get 263  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_05_piano_normal.mp3\n",
      "Get 264  ./audio/london_phill_dataset_multi/oboe/oboe_A5_15_piano_normal.mp3\n",
      "Get 265  ./audio/london_phill_dataset_multi/oboe/oboe_A6_1_forte_minor-trill.mp3\n",
      "Get 266  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 267  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_forte_minor-trill.mp3\n",
      "Get 268  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_1_forte_normal.mp3\n",
      "Get 269  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_025_fortissimo_normal.mp3\n",
      "Get 270  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_025_piano_normal.mp3\n",
      "Get 271  ./audio/london_phill_dataset_multi/oboe/oboe_A5_025_mezzo-forte_normal.mp3\n",
      "Get 272  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_1_fortissimo_normal.mp3\n",
      "Get 273  ./audio/london_phill_dataset_multi/oboe/oboe_A4_05_forte_normal.mp3\n",
      "Get 274  ./audio/london_phill_dataset_multi/oboe/oboe_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 275  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_fortissimo_normal.mp3\n",
      "Get 276  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_15_mezzo-forte_normal.mp3\n",
      "Get 277  ./audio/london_phill_dataset_multi/oboe/oboe_A6_1_forte_major-trill.mp3\n",
      "Get 278  ./audio/london_phill_dataset_multi/oboe/oboe_As3_05_forte_normal.mp3\n",
      "Get 279  ./audio/london_phill_dataset_multi/oboe/oboe_As3_1_forte_normal.mp3\n",
      "Get 280  ./audio/london_phill_dataset_multi/oboe/oboe_Gs6_1_forte_normal.mp3\n",
      "Get 281  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_piano_normal.mp3\n",
      "Get 282  ./audio/london_phill_dataset_multi/oboe/oboe_As3_15_forte_normal.mp3\n",
      "Get 283  ./audio/london_phill_dataset_multi/oboe/oboe_As3_05_piano_normal.mp3\n",
      "Get 284  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_mezzo-forte_normal.mp3\n",
      "Get 285  ./audio/london_phill_dataset_multi/oboe/oboe_A4_1_piano_normal.mp3\n",
      "Get 286  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_15_forte_normal.mp3\n",
      "Get 287  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_05_piano_normal.mp3\n",
      "Get 288  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_forte_major-trill.mp3\n",
      "Get 289  ./audio/london_phill_dataset_multi/oboe/oboe_A4_1_mezzo-forte_normal.mp3\n",
      "Get 290  ./audio/london_phill_dataset_multi/oboe/oboe_A4_025_piano_normal.mp3\n",
      "Get 291  ./audio/london_phill_dataset_multi/oboe/oboe_A5_05_mezzo-forte_normal.mp3\n",
      "Get 292  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_piano_normal.mp3\n",
      "Get 293  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_15_piano_normal.mp3\n",
      "Get 294  ./audio/london_phill_dataset_multi/oboe/oboe_As3_15_fortissimo_normal.mp3\n",
      "Get 295  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_1_mezzo-forte_normal.mp3\n",
      "Get 296  ./audio/london_phill_dataset_multi/oboe/oboe_As3_1_fortissimo_normal.mp3\n",
      "Get 297  ./audio/london_phill_dataset_multi/oboe/oboe_A5_1_forte_normal.mp3\n",
      "Get 298  ./audio/london_phill_dataset_multi/oboe/oboe_G6_025_mezzo-forte_normal.mp3\n",
      "Get 299  ./audio/london_phill_dataset_multi/oboe/oboe_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 300  ./audio/london_phill_dataset_multi/oboe/oboe_Ds6_05_piano_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 301  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_025_forte_normal.mp3\n",
      "Get 302  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs4_05_pianissimo_normal.mp3\n",
      "Get 303  ./audio/london_phill_dataset_multi/trumpet/trumpet_G3_025_pianissimo_normal.mp3\n",
      "Get 304  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 305  ./audio/london_phill_dataset_multi/trumpet/trumpet_E6_1_forte_normal.mp3\n",
      "Get 306  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds4_15_pianissimo_normal.mp3\n",
      "Get 307  ./audio/london_phill_dataset_multi/trumpet/trumpet_A3_05_pianissimo_normal.mp3\n",
      "Get 308  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds4_025_pianissimo_normal.mp3\n",
      "Get 309  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_05_forte_normal.mp3\n",
      "Get 310  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_15_pianissimo_normal.mp3\n",
      "Get 311  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds4_05_pianissimo_normal.mp3\n",
      "Get 312  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_05_pianissimo_normal.mp3\n",
      "Get 313  ./audio/london_phill_dataset_multi/trumpet/trumpet_As3_1_pianissimo_normal.mp3\n",
      "Get 314  ./audio/london_phill_dataset_multi/trumpet/trumpet_D6_05_forte_normal.mp3\n",
      "Get 315  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs6_025_forte_normal.mp3\n",
      "Get 316  ./audio/london_phill_dataset_multi/trumpet/trumpet_A3_05_forte_normal.mp3\n",
      "Get 317  ./audio/london_phill_dataset_multi/trumpet/trumpet_E4_025_pianissimo_normal.mp3\n",
      "Get 318  ./audio/london_phill_dataset_multi/trumpet/trumpet_As3_05_pianissimo_normal.mp3\n",
      "Get 319  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_05_forte_normal.mp3\n",
      "Get 320  ./audio/london_phill_dataset_multi/trumpet/trumpet_G5_1_mezzo-forte_normal.mp3\n",
      "Get 321  ./audio/london_phill_dataset_multi/trumpet/trumpet_G5_025_mezzo-forte_normal.mp3\n",
      "Get 322  ./audio/london_phill_dataset_multi/trumpet/trumpet_B5_025_mezzo-forte_normal.mp3\n",
      "Get 323  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_025_forte_normal.mp3\n",
      "Get 324  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_15_pianissimo_normal.mp3\n",
      "Get 325  ./audio/london_phill_dataset_multi/trumpet/trumpet_E4_15_pianissimo_normal.mp3\n",
      "Get 326  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs4_15_pianissimo_normal.mp3\n",
      "Get 327  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_05_pianissimo_normal.mp3\n",
      "Get 328  ./audio/london_phill_dataset_multi/trumpet/trumpet_E6_025_forte_normal.mp3\n",
      "Get 329  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds6_025_forte_normal.mp3\n",
      "Get 330  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_025_pianissimo_normal.mp3\n",
      "Get 331  ./audio/london_phill_dataset_multi/trumpet/trumpet_B5_025_forte_normal.mp3\n",
      "Get 332  ./audio/london_phill_dataset_multi/trumpet/trumpet_C6_025_forte_normal.mp3\n",
      "Get 333  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_05_forte_normal.mp3\n",
      "Get 334  ./audio/london_phill_dataset_multi/trumpet/trumpet_As4_025_forte_normal.mp3\n",
      "Get 335  ./audio/london_phill_dataset_multi/trumpet/trumpet_F4_025_pianissimo_normal.mp3\n",
      "Get 336  ./audio/london_phill_dataset_multi/trumpet/trumpet_A3_15_pianissimo_normal.mp3\n",
      "Get 337  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_1_pianissimo_normal.mp3\n",
      "Get 338  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_05_pianissimo_normal.mp3\n",
      "Get 339  ./audio/london_phill_dataset_multi/trumpet/trumpet_F4_05_pianissimo_normal.mp3\n",
      "Get 340  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_025_forte_normal.mp3\n",
      "Get 341  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs4_1_pianissimo_normal.mp3\n",
      "Get 342  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_025_forte_normal.mp3\n",
      "Get 343  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_1_pianissimo_normal.mp3\n",
      "Get 344  ./audio/london_phill_dataset_multi/trumpet/trumpet_A4_05_forte_normal.mp3\n",
      "Get 345  ./audio/london_phill_dataset_multi/trumpet/trumpet_G3_15_pianissimo_normal.mp3\n",
      "Get 346  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs4_05_forte_normal.mp3\n",
      "Get 347  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_15_pianissimo_normal.mp3\n",
      "Get 348  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_1_pianissimo_normal.mp3\n",
      "Get 349  ./audio/london_phill_dataset_multi/trumpet/trumpet_As5_05_forte_normal.mp3\n",
      "Get 350  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds4_1_pianissimo_normal.mp3\n",
      "Get 351  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_025_pianissimo_normal.mp3\n",
      "Get 352  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_05_pianissimo_normal.mp3\n",
      "Get 353  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds4_05_forte_normal.mp3\n",
      "Get 354  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_15_pianissimo_normal.mp3\n",
      "Get 355  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_05_forte_normal.mp3\n",
      "Get 356  ./audio/london_phill_dataset_multi/trumpet/trumpet_As3_15_pianissimo_normal.mp3\n",
      "Get 357  ./audio/london_phill_dataset_multi/trumpet/trumpet_B4_025_forte_normal.mp3\n",
      "Get 358  ./audio/london_phill_dataset_multi/trumpet/trumpet_As5_05_mezzo-forte_normal.mp3\n",
      "Get 359  ./audio/london_phill_dataset_multi/trumpet/trumpet_E4_1_pianissimo_normal.mp3\n",
      "Get 360  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_05_forte_normal.mp3\n",
      "Get 361  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs3_1_forte_normal.mp3\n",
      "Get 362  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds6_05_forte_normal.mp3\n",
      "Get 363  ./audio/london_phill_dataset_multi/trumpet/trumpet_D6_025_forte_normal.mp3\n",
      "Get 364  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs5_05_forte_normal.mp3\n",
      "Get 365  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs5_025_forte_normal.mp3\n",
      "Get 366  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_05_pianissimo_normal.mp3\n",
      "Get 367  ./audio/london_phill_dataset_multi/trumpet/trumpet_A5_025_mezzo-forte_normal.mp3\n",
      "Get 368  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs4_025_forte_normal.mp3\n",
      "Get 369  ./audio/london_phill_dataset_multi/trumpet/trumpet_A4_025_forte_normal.mp3\n",
      "Get 370  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_025_pianissimo_normal.mp3\n",
      "Get 371  ./audio/london_phill_dataset_multi/trumpet/trumpet_E4_05_pianissimo_normal.mp3\n",
      "Get 372  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_15_pianissimo_normal.mp3\n",
      "Get 373  ./audio/london_phill_dataset_multi/trumpet/trumpet_D5_05_forte_normal.mp3\n",
      "Get 374  ./audio/london_phill_dataset_multi/trumpet/trumpet_Ds5_05_forte_normal.mp3\n",
      "Get 375  ./audio/london_phill_dataset_multi/trumpet/trumpet_As5_025_forte_normal.mp3\n",
      "Get 376  ./audio/london_phill_dataset_multi/trumpet/trumpet_A3_1_pianissimo_normal.mp3\n",
      "Get 377  ./audio/london_phill_dataset_multi/trumpet/trumpet_As3_025_forte_normal.mp3\n",
      "Get 378  ./audio/london_phill_dataset_multi/trumpet/trumpet_G3_05_pianissimo_normal.mp3\n",
      "Get 379  ./audio/london_phill_dataset_multi/trumpet/trumpet_C6_05_forte_normal.mp3\n",
      "Get 380  ./audio/london_phill_dataset_multi/trumpet/trumpet_A3_025_pianissimo_normal.mp3\n",
      "Get 381  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_025_pianissimo_normal.mp3\n",
      "Get 382  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_1_forte_normal.mp3\n",
      "Get 383  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_15_pianissimo_normal.mp3\n",
      "Get 384  ./audio/london_phill_dataset_multi/trumpet/trumpet_G5_15_mezzo-forte_normal.mp3\n",
      "Get 385  ./audio/london_phill_dataset_multi/trumpet/trumpet_A5_025_forte_normal.mp3\n",
      "Get 386  ./audio/london_phill_dataset_multi/trumpet/trumpet_Cs4_025_pianissimo_normal.mp3\n",
      "Get 387  ./audio/london_phill_dataset_multi/trumpet/trumpet_As3_025_pianissimo_normal.mp3\n",
      "Get 388  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_1_pianissimo_normal.mp3\n",
      "Get 389  ./audio/london_phill_dataset_multi/trumpet/trumpet_D4_1_pianissimo_normal.mp3\n",
      "Get 390  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 391  ./audio/london_phill_dataset_multi/trumpet/trumpet_G3_1_pianissimo_normal.mp3\n",
      "Get 392  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_05_pianissimo_normal.mp3\n",
      "Get 393  ./audio/london_phill_dataset_multi/trumpet/trumpet_A5_05_mezzo-forte_normal.mp3\n",
      "Get 394  ./audio/london_phill_dataset_multi/trumpet/trumpet_G5_05_mezzo-forte_normal.mp3\n",
      "Get 395  ./audio/london_phill_dataset_multi/trumpet/trumpet_C4_05_forte_normal.mp3\n",
      "Get 396  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_1_pianissimo_normal.mp3\n",
      "Get 397  ./audio/london_phill_dataset_multi/trumpet/trumpet_B3_025_pianissimo_normal.mp3\n",
      "Get 398  ./audio/london_phill_dataset_multi/trumpet/trumpet_F3_025_pianissimo_normal.mp3\n",
      "Get 399  ./audio/london_phill_dataset_multi/trumpet/trumpet_Gs4_05_forte_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 400  ./audio/london_phill_dataset_multi/trumpet/trumpet_E3_1_forte_normal.mp3\n",
      "Get 401  ./audio/london_phill_dataset_multi/flute/flute_Gs5_1_piano_normal.mp3\n",
      "Get 402  ./audio/london_phill_dataset_multi/flute/flute_A5_1_piano_normal.mp3\n",
      "Get 403  ./audio/london_phill_dataset_multi/flute/flute_Gs4_025_mezzo-piano_normal.mp3\n",
      "Get 404  ./audio/london_phill_dataset_multi/flute/flute_A6_05_pianissimo_normal.mp3\n",
      "Get 405  ./audio/london_phill_dataset_multi/flute/flute_A5_025_pianissimo_normal.mp3\n",
      "Get 406  ./audio/london_phill_dataset_multi/flute/flute_A4_15_mezzo-piano_normal.mp3\n",
      "Get 407  ./audio/london_phill_dataset_multi/flute/flute_A5_05_forte_normal.mp3\n",
      "Get 408  ./audio/london_phill_dataset_multi/flute/flute_Gs5_15_mezzo-forte_normal.mp3\n",
      "Get 409  ./audio/london_phill_dataset_multi/flute/flute_Gs5_1_mezzo-forte_normal.mp3\n",
      "Get 410  ./audio/london_phill_dataset_multi/flute/flute_A5_05_mezzo-piano_normal.mp3\n",
      "Get 411  ./audio/london_phill_dataset_multi/flute/flute_Gs5_025_mezzo-piano_normal.mp3\n",
      "Get 412  ./audio/london_phill_dataset_multi/flute/flute_Gs4_05_pianissimo_normal.mp3\n",
      "Get 413  ./audio/london_phill_dataset_multi/flute/flute_Gs5_15_piano_normal.mp3\n",
      "Get 414  ./audio/london_phill_dataset_multi/flute/flute_Gs6_1_mezzo-forte_normal.mp3\n",
      "Get 415  ./audio/london_phill_dataset_multi/flute/flute_Gs6_15_piano_normal.mp3\n",
      "Get 416  ./audio/london_phill_dataset_multi/flute/flute_Gs5_025_piano_normal.mp3\n",
      "Get 417  ./audio/london_phill_dataset_multi/flute/flute_A4_025_mezzo-piano_normal.mp3\n",
      "Get 418  ./audio/london_phill_dataset_multi/flute/flute_A4_05_mezzo-forte_normal.mp3\n",
      "Get 419  ./audio/london_phill_dataset_multi/flute/flute_Gs6_025_piano_normal.mp3\n",
      "Get 420  ./audio/london_phill_dataset_multi/flute/flute_Gs4_15_mezzo-forte_normal.mp3\n",
      "Get 421  ./audio/london_phill_dataset_multi/flute/flute_Gs6_1_forte_normal.mp3\n",
      "Get 422  ./audio/london_phill_dataset_multi/flute/flute_A6_05_piano_normal.mp3\n",
      "Get 423  ./audio/london_phill_dataset_multi/flute/flute_Gs4_15_mezzo-piano_normal.mp3\n",
      "Get 424  ./audio/london_phill_dataset_multi/flute/flute_A5_15_forte_normal.mp3\n",
      "Get 425  ./audio/london_phill_dataset_multi/flute/flute_A4_025_piano_normal.mp3\n",
      "Get 426  ./audio/london_phill_dataset_multi/flute/flute_A4_1_mezzo-forte_normal.mp3\n",
      "Get 427  ./audio/london_phill_dataset_multi/flute/flute_Gs6_15_mezzo-forte_normal.mp3\n",
      "Get 428  ./audio/london_phill_dataset_multi/flute/flute_A6_15_forte_normal.mp3\n",
      "Get 429  ./audio/london_phill_dataset_multi/flute/flute_Gs4_025_pianissimo_normal.mp3\n",
      "Get 430  ./audio/london_phill_dataset_multi/flute/flute_A5_025_forte_normal.mp3\n",
      "Get 431  ./audio/london_phill_dataset_multi/flute/flute_A5_05_pianissimo_normal.mp3\n",
      "Get 432  ./audio/london_phill_dataset_multi/flute/flute_Gs5_05_piano_normal.mp3\n",
      "Get 433  ./audio/london_phill_dataset_multi/flute/flute_A6_1_fortissimo_minor-trill.mp3\n",
      "Get 434  ./audio/london_phill_dataset_multi/flute/flute_Gs4_05_forte_normal.mp3\n",
      "Get 435  ./audio/london_phill_dataset_multi/flute/flute_Gs6_1_pianissimo_normal.mp3\n",
      "Get 436  ./audio/london_phill_dataset_multi/flute/flute_Gs6_025_pianissimo_normal.mp3\n",
      "Get 437  ./audio/london_phill_dataset_multi/flute/flute_A4_05_mezzo-piano_normal.mp3\n",
      "Get 438  ./audio/london_phill_dataset_multi/flute/flute_A5_025_piano_normal.mp3\n",
      "Get 439  ./audio/london_phill_dataset_multi/flute/flute_A6_1_forte_normal.mp3\n",
      "Get 440  ./audio/london_phill_dataset_multi/flute/flute_Gs4_05_mezzo-forte_normal.mp3\n",
      "Get 441  ./audio/london_phill_dataset_multi/flute/flute_Gs4_025_forte_normal.mp3\n",
      "Get 442  ./audio/london_phill_dataset_multi/flute/flute_A5_1_pianissimo_normal.mp3\n",
      "Get 443  ./audio/london_phill_dataset_multi/flute/flute_A6_1_mezzo-forte_normal.mp3\n",
      "Get 444  ./audio/london_phill_dataset_multi/flute/flute_Gs4_15_piano_normal.mp3\n",
      "Get 445  ./audio/london_phill_dataset_multi/flute/flute_Gs5_05_forte_normal.mp3\n",
      "Get 446  ./audio/london_phill_dataset_multi/flute/flute_Gs5_05_mezzo-forte_normal.mp3\n",
      "Get 447  ./audio/london_phill_dataset_multi/flute/flute_Gs6_15_forte_normal.mp3\n",
      "Get 448  ./audio/london_phill_dataset_multi/flute/flute_A4_15_forte_normal.mp3\n",
      "Get 449  ./audio/london_phill_dataset_multi/flute/flute_A5_15_pianissimo_normal.mp3\n",
      "Get 450  ./audio/london_phill_dataset_multi/flute/flute_Gs5_15_mezzo-piano_normal.mp3\n",
      "Get 451  ./audio/london_phill_dataset_multi/flute/flute_A4_15_pianissimo_normal.mp3\n",
      "Get 452  ./audio/london_phill_dataset_multi/flute/flute_A6_05_forte_normal.mp3\n",
      "Get 453  ./audio/london_phill_dataset_multi/flute/flute_A4_15_piano_normal.mp3\n",
      "Get 454  ./audio/london_phill_dataset_multi/flute/flute_A5_15_mezzo-forte_normal.mp3\n",
      "Get 455  ./audio/london_phill_dataset_multi/flute/flute_A4_1_mezzo-piano_normal.mp3\n",
      "Get 456  ./audio/london_phill_dataset_multi/flute/flute_A5_15_mezzo-piano_normal.mp3\n",
      "Get 457  ./audio/london_phill_dataset_multi/flute/flute_A5_1_forte_normal.mp3\n",
      "Get 458  ./audio/london_phill_dataset_multi/flute/flute_Gs5_1_mezzo-piano_normal.mp3\n",
      "Get 459  ./audio/london_phill_dataset_multi/flute/flute_Gs5_025_mezzo-forte_normal.mp3\n",
      "Get 460  ./audio/london_phill_dataset_multi/flute/flute_Gs5_025_pianissimo_normal.mp3\n",
      "Get 461  ./audio/london_phill_dataset_multi/flute/flute_Gs6_05_mezzo-forte_normal.mp3\n",
      "Get 462  ./audio/london_phill_dataset_multi/flute/flute_A4_025_pianissimo_normal.mp3\n",
      "Get 463  ./audio/london_phill_dataset_multi/flute/flute_Gs5_15_forte_normal.mp3\n",
      "Get 464  ./audio/london_phill_dataset_multi/flute/flute_A5_1_mezzo-forte_normal.mp3\n",
      "Get 465  ./audio/london_phill_dataset_multi/flute/flute_Gs6_05_piano_normal.mp3\n",
      "Get 466  ./audio/london_phill_dataset_multi/flute/flute_A4_05_forte_normal.mp3\n",
      "Get 467  ./audio/london_phill_dataset_multi/flute/flute_A5_15_piano_normal.mp3\n",
      "Get 468  ./audio/london_phill_dataset_multi/flute/flute_A4_15_mezzo-forte_normal.mp3\n",
      "Get 469  ./audio/london_phill_dataset_multi/flute/flute_Gs6_025_forte_normal.mp3\n",
      "Get 470  ./audio/london_phill_dataset_multi/flute/flute_A4_05_pianissimo_normal.mp3\n",
      "Get 471  ./audio/london_phill_dataset_multi/flute/flute_Gs6_15_pianissimo_normal.mp3\n",
      "Get 472  ./audio/london_phill_dataset_multi/flute/flute_Gs4_025_piano_normal.mp3\n",
      "Get 473  ./audio/london_phill_dataset_multi/flute/flute_Gs4_05_mezzo-piano_normal.mp3\n",
      "Get 474  ./audio/london_phill_dataset_multi/flute/flute_Gs4_025_mezzo-forte_normal.mp3\n",
      "Get 475  ./audio/london_phill_dataset_multi/flute/flute_Gs6_05_pianissimo_normal.mp3\n",
      "Get 476  ./audio/london_phill_dataset_multi/flute/flute_A6_1_piano_normal.mp3\n",
      "Get 477  ./audio/london_phill_dataset_multi/flute/flute_Gs4_15_forte_normal.mp3\n",
      "Get 478  ./audio/london_phill_dataset_multi/flute/flute_A5_025_mezzo-piano_normal.mp3\n",
      "Get 479  ./audio/london_phill_dataset_multi/flute/flute_A5_05_piano_normal.mp3\n",
      "Get 480  ./audio/london_phill_dataset_multi/flute/flute_Gs6_025_mezzo-forte_normal.mp3\n",
      "Get 481  ./audio/london_phill_dataset_multi/flute/flute_Gs4_05_piano_normal.mp3\n",
      "Get 482  ./audio/london_phill_dataset_multi/flute/flute_Gs5_1_pianissimo_normal.mp3\n",
      "Get 483  ./audio/london_phill_dataset_multi/flute/flute_Gs6_1_piano_normal.mp3\n",
      "Get 484  ./audio/london_phill_dataset_multi/flute/flute_Gs5_05_pianissimo_normal.mp3\n",
      "Get 485  ./audio/london_phill_dataset_multi/flute/flute_A4_025_forte_normal.mp3\n",
      "Get 486  ./audio/london_phill_dataset_multi/flute/flute_A4_1_pianissimo_normal.mp3\n",
      "Get 487  ./audio/london_phill_dataset_multi/flute/flute_A6_1_pianissimo_normal.mp3\n",
      "Get 488  ./audio/london_phill_dataset_multi/flute/flute_A4_1_forte_normal.mp3\n",
      "Get 489  ./audio/london_phill_dataset_multi/flute/flute_A5_1_mezzo-piano_normal.mp3\n",
      "Get 490  ./audio/london_phill_dataset_multi/flute/flute_A4_025_mezzo-forte_normal.mp3\n",
      "Get 491  ./audio/london_phill_dataset_multi/flute/flute_Gs6_05_forte_normal.mp3\n",
      "Get 492  ./audio/london_phill_dataset_multi/flute/flute_A5_025_mezzo-forte_normal.mp3\n",
      "Get 493  ./audio/london_phill_dataset_multi/flute/flute_A4_1_piano_normal.mp3\n",
      "Get 494  ./audio/london_phill_dataset_multi/flute/flute_A5_05_mezzo-forte_normal.mp3\n",
      "Get 495  ./audio/london_phill_dataset_multi/flute/flute_A6_05_mezzo-forte_normal.mp3\n",
      "Get 496  ./audio/london_phill_dataset_multi/flute/flute_Gs5_05_mezzo-piano_normal.mp3\n",
      "Get 497  ./audio/london_phill_dataset_multi/flute/flute_Gs5_1_forte_normal.mp3\n",
      "Get 498  ./audio/london_phill_dataset_multi/flute/flute_A4_05_piano_normal.mp3\n",
      "Get 499  ./audio/london_phill_dataset_multi/flute/flute_Gs5_025_forte_normal.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 500  ./audio/london_phill_dataset_multi/flute/flute_Gs4_15_pianissimo_normal.mp3\n",
      "Get 501  ./audio/london_phill_dataset_multi/cello/cello_Gs5_025_mezzo-piano_arco-normal.mp3\n",
      "Get 502  ./audio/london_phill_dataset_multi/cello/cello_Gs5_15_forte_arco-normal.mp3\n",
      "Get 503  ./audio/london_phill_dataset_multi/cello/cello_A2_05_pianissimo_arco-normal.mp3\n",
      "Get 504  ./audio/london_phill_dataset_multi/cello/cello_Gs4_05_pianissimo_arco-normal.mp3\n",
      "Get 505  ./audio/london_phill_dataset_multi/cello/cello_A2_1_fortissimo_arco-normal.mp3\n",
      "Get 506  ./audio/london_phill_dataset_multi/cello/cello_A3_025_pianissimo_arco-normal.mp3\n",
      "Get 507  ./audio/london_phill_dataset_multi/cello/cello_Gs5_05_forte_arco-normal.mp3\n",
      "Get 508  ./audio/london_phill_dataset_multi/cello/cello_A3_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 509  ./audio/london_phill_dataset_multi/cello/cello_A2_1_forte_arco-normal.mp3\n",
      "Get 510  ./audio/london_phill_dataset_multi/cello/cello_Gs4_15_pianissimo_arco-normal.mp3\n",
      "Get 511  ./audio/london_phill_dataset_multi/cello/cello_Gs4_025_forte_arco-normal.mp3\n",
      "Get 512  ./audio/london_phill_dataset_multi/cello/cello_A3_05_fortissimo_arco-normal.mp3\n",
      "Get 513  ./audio/london_phill_dataset_multi/cello/cello_Gs5_05_fortissimo_arco-normal.mp3\n",
      "Get 514  ./audio/london_phill_dataset_multi/cello/cello_A3_1_pianissimo_arco-normal.mp3\n",
      "Get 515  ./audio/london_phill_dataset_multi/cello/cello_A3_05_forte_arco-normal.mp3\n",
      "Get 516  ./audio/london_phill_dataset_multi/cello/cello_Gs3_1_forte_arco-normal.mp3\n",
      "Get 517  ./audio/london_phill_dataset_multi/cello/cello_A4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 518  ./audio/london_phill_dataset_multi/cello/cello_Gs3_1_mezzo-piano_arco-normal.mp3\n",
      "Get 519  ./audio/london_phill_dataset_multi/cello/cello_Gs3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 520  ./audio/london_phill_dataset_multi/cello/cello_Gs5_1_forte_arco-normal.mp3\n",
      "Get 521  ./audio/london_phill_dataset_multi/cello/cello_Gs3_025_fortissimo_arco-normal.mp3\n",
      "Get 522  ./audio/london_phill_dataset_multi/cello/cello_Gs4_05_forte_arco-normal.mp3\n",
      "Get 523  ./audio/london_phill_dataset_multi/cello/cello_Gs4_025_fortissimo_arco-normal.mp3\n",
      "Get 524  ./audio/london_phill_dataset_multi/cello/cello_A3_15_mezzo-piano_arco-normal.mp3\n",
      "Get 525  ./audio/london_phill_dataset_multi/cello/cello_A3_1_mezzo-piano_arco-normal.mp3\n",
      "Get 526  ./audio/london_phill_dataset_multi/cello/cello_Gs3_15_fortissimo_arco-normal.mp3\n",
      "Get 527  ./audio/london_phill_dataset_multi/cello/cello_Gs5_05_pianissimo_arco-normal.mp3\n",
      "Get 528  ./audio/london_phill_dataset_multi/cello/cello_A4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 529  ./audio/london_phill_dataset_multi/cello/cello_Gs4_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 530  ./audio/london_phill_dataset_multi/cello/cello_A2_05_fortissimo_arco-normal.mp3\n",
      "Get 531  ./audio/london_phill_dataset_multi/cello/cello_Gs3_025_forte_arco-normal.mp3\n",
      "Get 532  ./audio/london_phill_dataset_multi/cello/cello_Gs4_1_fortissimo_arco-normal.mp3\n",
      "Get 533  ./audio/london_phill_dataset_multi/cello/cello_Gs5_1_mezzo-piano_arco-normal.mp3\n",
      "Get 534  ./audio/london_phill_dataset_multi/cello/cello_Gs3_05_forte_arco-normal.mp3\n",
      "Get 535  ./audio/london_phill_dataset_multi/cello/cello_Gs5_15_pianissimo_arco-normal.mp3\n",
      "Get 536  ./audio/london_phill_dataset_multi/cello/cello_Gs5_15_fortissimo_arco-normal.mp3\n",
      "Get 537  ./audio/london_phill_dataset_multi/cello/cello_Gs3_15_piano_arco-normal.mp3\n",
      "Get 538  ./audio/london_phill_dataset_multi/cello/cello_A3_1_fortissimo_arco-normal.mp3\n",
      "Get 539  ./audio/london_phill_dataset_multi/cello/cello_Gs3_15_pianissimo_arco-normal.mp3\n",
      "Get 540  ./audio/london_phill_dataset_multi/cello/cello_Gs5_025_forte_arco-normal.mp3\n",
      "Get 541  ./audio/london_phill_dataset_multi/cello/cello_Gs3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 542  ./audio/london_phill_dataset_multi/cello/cello_A3_1_mezzo-piano_arco-minor-trill.mp3\n",
      "Get 543  ./audio/london_phill_dataset_multi/cello/cello_A4_1_mezzo-forte_arco-harmonic.mp3\n",
      "Get 544  ./audio/london_phill_dataset_multi/cello/cello_A4_025_pianissimo_arco-normal.mp3\n",
      "Get 545  ./audio/london_phill_dataset_multi/cello/cello_A2_025_mezzo-piano_arco-normal.mp3\n",
      "Get 546  ./audio/london_phill_dataset_multi/cello/cello_Gs4_1_forte_arco-normal.mp3\n",
      "Get 547  ./audio/london_phill_dataset_multi/cello/cello_Gs3_025_pianissimo_arco-normal.mp3\n",
      "Get 548  ./audio/london_phill_dataset_multi/cello/cello_Gs5_15_mezzo-piano_arco-normal.mp3\n",
      "Get 549  ./audio/london_phill_dataset_multi/cello/cello_A3_05_mezzo-piano_arco-normal.mp3\n",
      "Get 550  ./audio/london_phill_dataset_multi/cello/cello_A3_025_fortissimo_arco-normal.mp3\n",
      "Get 551  ./audio/london_phill_dataset_multi/cello/cello_Gs5_1_mezzo-forte_arco-harmonic.mp3\n",
      "Get 552  ./audio/london_phill_dataset_multi/cello/cello_Gs5_05_mezzo-piano_arco-normal.mp3\n",
      "Get 553  ./audio/london_phill_dataset_multi/cello/cello_A4_025_fortissimo_arco-normal.mp3\n",
      "Get 554  ./audio/london_phill_dataset_multi/cello/cello_Gs5_025_fortissimo_arco-normal.mp3\n",
      "Get 555  ./audio/london_phill_dataset_multi/cello/cello_A4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 556  ./audio/london_phill_dataset_multi/cello/cello_A4_1_forte_arco-normal.mp3\n",
      "Get 557  ./audio/london_phill_dataset_multi/cello/cello_Gs4_05_mezzo-piano_arco-normal.mp3\n",
      "Get 558  ./audio/london_phill_dataset_multi/cello/cello_A4_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 559  ./audio/london_phill_dataset_multi/cello/cello_A4_1_mezzo-piano_arco-minor-trill.mp3\n",
      "Get 560  ./audio/london_phill_dataset_multi/cello/cello_Gs4_1_mezzo-piano_arco-normal.mp3\n",
      "Get 561  ./audio/london_phill_dataset_multi/cello/cello_A4_025_forte_arco-normal.mp3\n",
      "Get 562  ./audio/london_phill_dataset_multi/cello/cello_Gs5_025_pianissimo_arco-normal.mp3\n",
      "Get 563  ./audio/london_phill_dataset_multi/cello/cello_A2_05_mezzo-piano_arco-normal.mp3\n",
      "Get 564  ./audio/london_phill_dataset_multi/cello/cello_A4_15_mezzo-piano_arco-normal.mp3\n",
      "Get 565  ./audio/london_phill_dataset_multi/cello/cello_Gs4_025_mezzo-piano_arco-normal.mp3\n",
      "Get 566  ./audio/london_phill_dataset_multi/cello/cello_A4_05_forte_arco-normal.mp3\n",
      "Get 567  ./audio/london_phill_dataset_multi/cello/cello_A2_15_forte_arco-normal.mp3\n",
      "Get 568  ./audio/london_phill_dataset_multi/cello/cello_Gs3_15_forte_arco-normal.mp3\n",
      "Get 569  ./audio/london_phill_dataset_multi/cello/cello_Gs4_1_pianissimo_arco-normal.mp3\n",
      "Get 570  ./audio/london_phill_dataset_multi/cello/cello_Gs4_15_mezzo-piano_arco-normal.mp3\n",
      "Get 571  ./audio/london_phill_dataset_multi/cello/cello_A3_025_mezzo-piano_arco-normal.mp3\n",
      "Get 572  ./audio/london_phill_dataset_multi/cello/cello_A3_15_forte_arco-normal.mp3\n",
      "Get 573  ./audio/london_phill_dataset_multi/cello/cello_A2_15_pianissimo_arco-normal.mp3\n",
      "Get 574  ./audio/london_phill_dataset_multi/cello/cello_Gs5_1_fortissimo_arco-normal.mp3\n",
      "Get 575  ./audio/london_phill_dataset_multi/cello/cello_A2_025_pianissimo_arco-normal.mp3\n",
      "Get 576  ./audio/london_phill_dataset_multi/cello/cello_A2_1_pianissimo_arco-normal.mp3\n",
      "Get 577  ./audio/london_phill_dataset_multi/cello/cello_Gs4_05_fortissimo_arco-normal.mp3\n",
      "Get 578  ./audio/london_phill_dataset_multi/cello/cello_A4_1_fortissimo_arco-normal.mp3\n",
      "Get 579  ./audio/london_phill_dataset_multi/cello/cello_Gs3_05_pianissimo_arco-normal.mp3\n",
      "Get 580  ./audio/london_phill_dataset_multi/cello/cello_A3_025_mezzo-forte_arco-col-legno-battuto.mp3\n",
      "Get 581  ./audio/london_phill_dataset_multi/cello/cello_Gs4_025_pianissimo_arco-normal.mp3\n",
      "Get 582  ./audio/london_phill_dataset_multi/cello/cello_A2_1_mezzo-piano_non-vibrato.mp3\n",
      "Get 583  ./audio/london_phill_dataset_multi/cello/cello_A4_1_mezzo-piano_molto-vibrato.mp3\n",
      "Get 584  ./audio/london_phill_dataset_multi/cello/cello_Gs4_15_forte_arco-normal.mp3\n",
      "Get 585  ./audio/london_phill_dataset_multi/cello/cello_Gs3_1_pianissimo_arco-normal.mp3\n",
      "Get 586  ./audio/london_phill_dataset_multi/cello/cello_Gs5_1_pianissimo_arco-normal.mp3\n",
      "Get 587  ./audio/london_phill_dataset_multi/cello/cello_A4_15_forte_arco-normal.mp3\n",
      "Get 588  ./audio/london_phill_dataset_multi/cello/cello_A2_025_forte_arco-normal.mp3\n",
      "Get 589  ./audio/london_phill_dataset_multi/cello/cello_A2_025_fortissimo_arco-normal.mp3\n",
      "Get 590  ./audio/london_phill_dataset_multi/cello/cello_A2_15_piano_arco-normal.mp3\n",
      "Get 591  ./audio/london_phill_dataset_multi/cello/cello_A2_025_mezzo-forte_arco-col-legno-battuto.mp3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get 592  ./audio/london_phill_dataset_multi/cello/cello_A2_05_forte_arco-normal.mp3\n",
      "Get 593  ./audio/london_phill_dataset_multi/cello/cello_A4_1_pianissimo_arco-normal.mp3\n",
      "Get 594  ./audio/london_phill_dataset_multi/cello/cello_Gs4_15_fortissimo_arco-normal.mp3\n",
      "Get 595  ./audio/london_phill_dataset_multi/cello/cello_A3_1_forte_arco-normal.mp3\n",
      "Get 596  ./audio/london_phill_dataset_multi/cello/cello_Gs3_05_fortissimo_arco-normal.mp3\n",
      "Get 597  ./audio/london_phill_dataset_multi/cello/cello_Gs2_025_pianissimo_arco-normal.mp3\n",
      "Get 598  ./audio/london_phill_dataset_multi/cello/cello_A3_025_forte_arco-normal.mp3\n",
      "Get 599  ./audio/london_phill_dataset_multi/cello/cello_A2_1_mezzo-piano_arco-normal.mp3\n",
      "Get 600  ./audio/london_phill_dataset_multi/cello/cello_A4_05_pianissimo_arco-normal.mp3\n",
      "Calculated 600 Durations\n"
     ]
    }
   ],
   "source": [
    "# Load audio files, trim silence and calculate duration\n",
    "duration = []\n",
    "for i,f in enumerate(files):\n",
    "    print (\"Get %d  %s\"%(i+1, f))\n",
    "    try:\n",
    "        y, sr = librosa.load(f, sr=fs)\n",
    "        if len(y) < 2:\n",
    "            print(\"Error loading %s\" % f)\n",
    "            continue\n",
    "        #y/=y.max() #Normalize\n",
    "        yt, index = librosa.effects.trim(y,top_db=60) #Trim\n",
    "        duration.append(librosa.get_duration(yt, sr=fs))\n",
    "    except Exception as e:\n",
    "        print(\"Error loading %s. Error: %s\" % (f,e))\n",
    "        \n",
    "print(\"Calculated %d Durations\"%len(duration))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Duration Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Duration average: 1.0031891156462587\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAGRxJREFUeJzt3X+YXVV97/H3B/KLEDDQQDoCJVKphUIz6qCotQTRPlRFuDyIUrSRhzRPe7XXWH8hVsXa9uJ9LBSrrY0/ICpS0Yj8KGi5MZFqBUnoxECjxUuJUHKTJjhAcCyGfvvHXgcPk3Pm7Jk5+/xan9fzzDNn77X32t919uR8s9beZ21FBGZmlq/9uh2AmZl1lxOBmVnmnAjMzDLnRGBmljknAjOzzDkRmJllzonABoqkWyQt79Kxz5f0D22s7x5Jy9LrSyR9vo11XyzpU+2qz/qb/D0CawdJ9wOLgb3Ak8C/AJ8FVkfEf1V0zEuAZ0fEG6qof8KxrgJ+B/jPtGobcCNwaUQ8Mo26HoyIP57CPpcwzbamZPL5iDhyqvtaHtwjsHY6IyIOAo4GLgXeDXx6OhVJmtXOwNrk/6T2HQZcAJwMfFvSge08SI+23QaYE4G1XUQ8EhE3AK8Dlks6AUDSBkkrattJepOkb9Uth6Q3S7oXuDetu0LSA5IelbRJ0kvT+tOBi4HXSdojafPEY0jaT9IfS9omaaekz0p6Ripbko63XNKPJO2S9N6S7ftpRNwJvAb4BYqk8LT2qHB5Ou4jkr4n6QRJK4HzgXeluG9M298v6d2Svgc8LmlWWvfyukPPk/RFSY9JukvS0gnv3bPrlq+S9KcpSd0CPDMdb4+kZ04capL0mjQUNZbew+Pqyu6X9I7UhkdSDPPKvFfWH5wIrDIR8V3gQeClU9jtLOCFwPFp+U5gGDgU+ALwJUnzIuJrwJ8DX4yIBRGxtEFdb0o/pwLHAAuAj03Y5jeA5wCnAe+v/wBsJSIeA26lcft+C/hN4FeAhRRJcXdErAaupuhdLIiIM+r2OQ94FbAwIvY2qPNM4Ev8/L34qqTZLWJ8HPht4KF0vAUR8VD9NpJ+BbgGWEXR27kZuFHSnLrNzgVOB54F/DrF+2oDwonAqvYQxQdXWf87Ih6OiHGAiPh8ROyOiL0R8RfAXIoP7jLOBy6LiPsiYg/wHuD1E4ZePhgR4xGxGdgMNEook2nWvp8BBwG/SnEtbmtEbG9R10cj4oFa2xvYFBFfjoifAZcB8yiGp2bqdcDfR8Stqe6PAAcAL54Q20MR8TDFtZHhNhzXeoQTgVXtCODhKWz/QP2CpLdL2pqGJMaAZwCLStb1TIqLujXbgFkUF7Vr/n/d659Q9BqmomH7IuIbFL2PjwM7JK2WdHCLuh4oW54uwD9I0caZetr7lOp+gKJtNTN9n6yHORFYZSSdRPFhUrsO8Dgwv26TX2yw21O3saXrAe+mGJY4JCIWAo8AmrhtEw9RXLiu+SWKu5p2lGzCpCQtAF4O/GOj8oj4aEQ8H/g1iiGid9aKmlTZqj1H1R17P+BIijZC8eHc7L2d0vskSelY/95iPxsQTgTWdpIOlvRq4O8oblvckopGgbMlzU8XNi9sUdVBFB/c/wHMkvR+oP5/1TuAJelDsZFrgLdJelb60K5dU2g0/l6apLmSng98FfgxcGWDbU6S9MI0hv848FOK22prcR8zjUM/X9LZaWhrFcWtrLenslHgdyTtny6kn1K33w7gF2oXyhu4FniVpNNSvG9Pdf/TNGK0PuREYO10o6THKIYV3ksxjn1BXfnlwBMUH0xrKC6aTubrFHe8/CvF0MVPefrwyZfS792S7mqw/2eAzwG3Af+W9v/DKbRnonel9j1M8R2JTcCL0wXZiQ4GPkmRKLYBuynG3qG4pfb4dIfOV6dw/OspxvN/DLwRODuN6QO8FTgDGKO4NvJUvRHxfYqkeF865tOGkyLiB8AbgL8CdqV6zoiIJ6YQm/Uxf6HMzCxz7hGYmWXOicDMLHNOBGZmmXMiMDPLXF9MbrVo0aJYsmRJt8Mo5fHHH+fAA9s6B1lXuB29Z1DaMijtgN5vy6ZNm3ZFxGGttuuLRLBkyRI2btzY7TBK2bBhA8uWLet2GDPmdvSeQWnLoLQDer8tkra13spDQ2Zm2XMiMDPLnBOBmVnmnAjMzDLnRGBmljknAjOzzDkRmJllzonAzCxzTgRmZpnri28W5+iC913ecP2VH3pbhyMxs0HnHoGZWeacCMzMMudEYGaWOScCM7PMORGYmWXOicDMLHNOBGZmmXMiMDPLnBOBmVnmnAjMzDLnRGBmljknAjOzzDkRmJllzonAzCxznoba9nHB+y7nlOOO2GcqbE+BbTaY3CMwM8ucE4GZWeacCMzMMlfpNQJJ9wOPAU8CeyNiRNKhwBeBJcD9wLkR8eMq4zAzs+Y60SM4NSKGI2IkLV8ErIuIY4F1adnMzLqkG0NDZwJr0us1wFldiMHMzJKqbx8N4B8kBfC3EbEaWBwR2wEiYrukwxvtKGklsBJgaGiI0dHRikNtj/Hx8bbEuvToRQ3Xd+J9WHr0IubPmbVPDP1yDuq163z0gkFpy6C0AwanLVUngpdExEPpw/5WSd8vu2NKGqsBRkZGYnh4uKoY22psbIx2xHrF2vUN169aUf37cMXa9SycP5fN23Z1/Njt1q7z0QsGpS2D0g4YnLZUOjQUEQ+l3zuB64AXADskDQGk3zurjMHMzCZXWSKQdKCkg2qvgd8C7gZuAJanzZYD11cVg5mZtVbl0NBi4DpJteN8ISK+JulO4FpJFwI/Al5bYQxmZtZCZYkgIu4DljZYvxs4rarjmpnZ1PibxWZmmXMiMDPLnBOBmVnmnAjMzDLnRGBmljknAjOzzDkRmJllzonAzCxzTgRmZplzIjAzy5wTgZlZ5pwIzMwy50RgZpY5JwIzs8w5EZiZZc6JwMwsc04EZmaZcyIwM8ucE4GZWeacCMzMMudEYGaWOScCM7PMORGYmWVuVrcDyNUF77u8admVH3pbByMxs9y5R2BmljknAjOzzDkRmJllzonAzCxzlScCSftL+mdJN6XlQyXdKune9PuQqmMwM7PmOtEjeCuwtW75ImBdRBwLrEvLZmbWJZUmAklHAq8CPlW3+kxgTXq9BjiryhjMzGxyVX+P4C+BdwEH1a1bHBHbASJiu6TDG+0oaSWwEmBoaIjR0dGKQ22P8fHxUrEuPXpR07LR0dGm5bW6N2zc0rB82ciJJaJsHdv8ObP2iaFfzkG9suejHwxKWwalHTA4baksEUh6NbAzIjZJWjbV/SNiNbAaYGRkJIaHh9scYTXGxsYoE+sVa9c3LVu1Yrhp+aoVw5PuXyufiSvWrmfh/Lls3rar7XV3Wtnz0Q8GpS2D0g4YnLZU2SN4CfAaSa8E5gEHS/o8sEPSUOoNDAE7K4zBzMxaqOwaQUS8JyKOjIglwOuBb0TEG4AbgOVps+XA9VXFYGZmrXXjewSXAq+QdC/wirRsZmZd0pFJ5yJiA7Ahvd4NnNaJ45qZWWv+ZrGZWeacCMzMMudEYGaWOScCM7PMORGYmWXOj6q0KWn2iE0/XtOsf7lHYGaWOScCM7PMORGYmWXO1wgGkMfxzWwq3CMwM8ucE4GZWeacCMzMMudEYGaWOScCM7PMORGYmWXOicDMLHNOBGZmmZtyIpB0iKRfryIYMzPrvFKJQNIGSQdLOhTYDFwp6bJqQzMzs04o2yN4RkQ8CpwNXBkRzwdeXl1YZmbWKWUTwSxJQ8C5wE0VxmNmZh1WNhF8EPg68MOIuFPSMcC91YVlZmadUnb20e0R8dQF4oi4z9cIzMwGQ9kewV+VXGdmZn1m0h6BpBcBLwYOk/RHdUUHA/tXGZiZmXVGq6GhOcCCtN1BdesfBc6pKigzM+ucSRNBRHwT+KakqyJiW4disor5CWZmVq/sxeK5klYDS+r3iYiXNdtB0jzgNmBu2ufLEfGB9KW0L6a67gfOjYgfTyd4MzObubKJ4EvAJ4BPAU+W3Oc/gZdFxB5Js4FvSbqF4ktp6yLiUkkXARcB755i3GZm1iZlE8HeiPibqVQcEQHsSYuz008AZwLL0vo1wAacCMzMuqZsIrhR0v8ErqP4nz4AEfHwZDtJ2h/YBDwb+HhE3CFpcURsT/tvl3R4k31XAisBhoaGGB0dLRlqd42Pjz8V64aNWxpus2zkRJYevahpHaOjo03La3VPVj6TfWvl8+fM2me7MnX3mvrz0e8GpS2D0g4YnLaUTQTL0+931q0L4JjJdoqIJ4FhSQuB6ySdUDawiFgNrAYYGRmJ4eHhsrt21djYGLVYr1i7vuE2q1YMNy1rVb5qxfTrLrNvrXzh/Lls3rZrynX3mvrz0e8GpS2D0g4YnLaUSgQR8ayZHCQixiRtAE4HdkgaSr2BIWDnTOo2M7OZKZUIJP1uo/UR8dlJ9jkM+FlKAgdQzFb6YeAGih7Gpen39VMN2szM2qfs0NBJda/nAacBdwFNEwEwBKxJ1wn2A66NiJskfQe4VtKFwI+A1049bDMza5eyQ0N/WL8s6RnA51rs8z3guQ3W76ZIJGZm1gOm+8zinwDHtjMQMzPrjrLXCG6kuEsIisnmjgOurSooMzPrnLLXCD5S93ovsC0iHqwgHjMz67BSQ0Np8rnvU8xAegjwRJVBmZlZ55RKBJLOBb5LcYfPucAdkjwNtZnZACg7NPRe4KSI2AlPfUfg/wJfriowMzPrjLJ3De1XSwLJ7insa2ZmPaxsj+Brkr4OXJOWXwfcXE1IZmbWSa2eWfxsYHFEvFPS2cBvAAK+A1zdgfjMzKxirXoEfwlcDBARXwG+AiBpJJWdUWl0Pa7RIx9POe6ILkTSO/wYTLP+02qcf0maKuJpImIjxaMmzcysz7VKBPMmKTugnYGYmVl3tBoaulPS70XEJ+tXpplDN1UXVud4KMPMctcqEayieLLY+fz8g38EmAP8jyoDMzOzzpg0EUTEDuDFkk4Fao+Z/PuI+EblkZmZWUeUfR7BeqD5Q3bNzKxv+dvBZmaZcyIwM8ucE4GZWeacCMzMMudEYGaWOScCM7PMORGYmWWu7PMIzNrCU3qY9R73CMzMMudEYGaWOScCM7PM+RqB9QxfPzDrjsp6BJKOkrRe0lZJ90h6a1p/qKRbJd2bfh9SVQxmZtZalUNDe4G3R8RxwMnAmyUdD1wErIuIY4F1adnMzLqkskQQEdsj4q70+jFgK3AEcCawJm22BjirqhjMzKy1jlwjkLQEeC5wB7A4IrZDkSwkHd5kn5XASoChoSFGR0criW3p0Ysarh8dHWXDxi1N91s2cmLDfefPmfVUrJPV3aysVflM6i6zb618/pxZ+2zXrrqnG/d0jI+PV/a302mD0pZBaQcMTlsqTwSSFgBrgVUR8aikUvtFxGpgNcDIyEgMDw9XEt8Vaxs/b2fViuGmZZOVL5w/l1qs7a67VjbdusvsWytfOH8um7ftqqTu6cY9HWNjY1T1t9Npg9KWQWkHDE5bKr19VNJsiiRwdUR8Ja3eIWkolQ8BO6uMwczMJlflXUMCPg1sjYjL6opuAJan18uB66uKwczMWqtyaOglwBuBLZJqg2gXA5cC10q6EPgR8NoKY/C96WZmLVSWCCLiW0CzCwKnVXVcMzObGk8xYWaWOScCM7PMORGYmWXOicDMLHNOBGZmmXMiMDPLnBOBmVnmnAjMzDLnRGBmljk/qtIGRqPpRE457oimZeCpRszAPQIzs+w5EZiZZc6JwMwsc04EZmaZcyIwM8ucE4GZWeZ8+6j1Dd8CalYN9wjMzDLnRGBmljknAjOzzDkRmJllzonAzCxzTgRmZplzIjAzy5wTgZlZ5pwIzMwy50RgZpY5JwIzs8xVlggkfUbSTkl31607VNKtku5Nvw+p6vhmZlZOlT2Cq4DTJ6y7CFgXEccC69KymZl1UWWJICJuAx6esPpMYE16vQY4q6rjm5lZOZ2ehnpxRGwHiIjtkg5vtqGklcBKgKGhIUZHR6d1wKVHL2q4vlbfZOXNyiYrnz9nVmV1zzTuMvvWyufPmbXPdu2qu8q4J6qdj1b7bti4pWH5spETG67vhvHx8Wn/O2ikW21udzu6aVDa0rPPI4iI1cBqgJGRkRgeHp5WPVesXd9w/aoVwy3Lm5VNVr5w/lxqsba77pnGXWbfWvnC+XPZvG1XJXVXGfdEtfMx07p7wdjYGNP9d9BIt9rc7nZ006C0pdN3De2QNASQfu/s8PHNzGyCTieCG4Dl6fVy4PoOH9/MzCao8vbRa4DvAM+R9KCkC4FLgVdIuhd4RVo2M7MuquwaQUSc16TotKqOaWZmU+dvFpuZZc6JwMwsc04EZmaZcyIwM8ucE4GZWeZ69pvFZv3igvdd3nD9lR96W4cjaa8q2zWo71m/co/AzCxzTgRmZplzIjAzy5yvEZi10Gw8GzymbYPBPQIzs8w5EZiZZc5DQ2YVm8mtkvX7nnLcEU8te0jK2sk9AjOzzDkRmJllzonAzCxzvkZg1sc8VYO1g3sEZmaZcyIwM8ucE4GZWeZ8jcCsi/p1+gpfmxgs7hGYmWXOicDMLHMeGjKztvPQUX9xj8DMLHNOBGZmmXMiMDPLnK8RmFlPyfH6QrdvI+5Kj0DS6ZJ+IOmHki7qRgxmZlboeCKQtD/wceC3geOB8yQd3+k4zMys0I0ewQuAH0bEfRHxBPB3wJldiMPMzABFRGcPKJ0DnB4RK9LyG4EXRsRbJmy3EliZFp8D/KCjgU7fImBXt4NoA7ej9wxKWwalHdD7bTk6Ig5rtVE3Lharwbp9slFErAZWVx9Oe0naGBEj3Y5jptyO3jMobRmUdsDgtKUbQ0MPAkfVLR8JPNSFOMzMjO4kgjuBYyU9S9Ic4PXADV2Iw8zM6MLQUETslfQW4OvA/sBnIuKeTsdRob4bzmrC7eg9g9KWQWkHDEhbOn6x2MzMeounmDAzy5wTgZlZ5pwIpqHVFBmSlkl6RNJo+nl/N+JsRdJnJO2UdHeTckn6aGrn9yQ9r9MxllWiLf1yTo6StF7SVkn3SHprg216/ryUbEfPnxNJ8yR9V9Lm1I4PNtim589HSxHhnyn8UFzg/n/AMcAcYDNw/IRtlgE3dTvWEm35TeB5wN1Nyl8J3ELx3Y+TgTu6HfMM2tIv52QIeF56fRDwrw3+vnr+vJRsR8+fk/QeL0ivZwN3ACf32/lo9eMewdQNzBQZEXEb8PAkm5wJfDYKtwMLJQ11JrqpKdGWvhAR2yPirvT6MWArcMSEzXr+vJRsR89L7/GetDg7/Uy8w6bnz0crTgRTdwTwQN3ygzT+A39R6k7eIunXOhNa25Vta7/oq3MiaQnwXIr/hdbrq/MySTugD86JpP0ljQI7gVsjoq/PRyN+HsHUlZki4y6KOT72SHol8FXg2Moja79S04H0ib46J5IWAGuBVRHx6MTiBrv05Hlp0Y6+OCcR8SQwLGkhcJ2kEyKi/lpU35yPZtwjmLqWU2RExKO17mRE3AzMlrSocyG2zcBMB9JP50TSbIoPz6sj4isNNumL89KqHf10TgAiYgzYAJw+oagvzsdknAimruUUGZJ+UZLS6xdQvM+7Ox7pzN0A/G66K+Jk4JGI2N7toKajX85JivHTwNaIuKzJZj1/Xsq0ox/OiaTDUk8ASQcALwe+P2Gznj8frXhoaIqiyRQZkn4/lX8COAf4A0l7gXHg9ZFuL+glkq6huHNjkaQHgQ9QXAyrteNmijsifgj8BLigO5G2VqItfXFOgJcAbwS2pHFpgIuBX4K+Oi9l2tEP52QIWKPigVr7AddGxE0T/r33w/mYlKeYMDPLnIeGzMwy50RgZpY5JwIzs8w5EZiZZc6JwMwsc04EZmaZcyKwgSHpyTSd8T1p/po/ktS2v3FJb5L0zLrlT0k6vg31XiLp3yX9ySTb/HJq255m25hNl79HYAND0p6IWJBeHw58Afh2RHxgCnXsn+aWaVS2AXhHRGxsR7x19V4C7ImIj5TY9qk2mrWLewQ2kCJiJ7ASeEv66v+bJH2sVi7pJknL0us9kv5E0h0Us2G+X9Kdku6WtDrtfw4wAlyd/md+gKQNkkZSHedJ2pL2+XDdcfZI+rPUQ7ld0uJWsUs6RT9/WMs/Szqove+O2dM5EdjAioj7KP7GD2+x6YEUD7R5YUR8C/hYRJwUEScABwCvjogvAxuB8yNiOCLGazun4aIPAy8DhoGTJJ1VV/ftEbEUuA34vRKhvwN4c0QMAy+lmH7BrDJOBDboGk0RPNGTFLNk1pwq6Q5JWyg+3FvNk38SsCEi/iMi9gJXUzwxDeAJ4Kb0ehOwpEQ83wYuk/S/gIWpTrPKOBHYwJJ0DMWH/E5gL0//e59X9/qntesCkuYBfw2cExEnAp+csG3DQ01S9rO6idSepMREjxFxKbCCojdyu6RfbbWP2Uw4EdhAknQY8AmKYZ4A7qd4uMh+ko6ieORoI7UP/V0qHqpyTl3ZYxTP353oDuAUSYvSLJXnAd+cQey/HBFbIuLDFMNRTgRWKU9DbYPkgDTl8WyKHsDngNpc+N8G/g3YAtxN8XSsfUTEmKRPpu3up3j+RM1VwCckjQMvqttnu6T3AOspegc3R8T1M2jHKkmnUvQg/oXiwehmlfHto2Zd5ttHrds8NGTWfXuAlWW+UAbs6FxYlgv3CMzMMucegZlZ5pwIzMwy50RgZpY5JwIzs8z9N8cLx2grc0LpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "durationDist = pd.Series(np.array(duration))\n",
    "plt.figure()\n",
    "durationDist.plot.hist(grid=True, bins=40, rwidth=0.8,\n",
    "                   color='#607c8e')\n",
    "plt.title('Duration Distribution')\n",
    "plt.xlabel('Duration [s]')\n",
    "plt.ylabel('Counts')\n",
    "plt.grid(axis='y', alpha=0.75)\n",
    "print(\"Duration average:\",np.mean(duration))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Short-Time Fourier Transform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<center><iframe src=\"https://en.wikipedia.org/wiki/Short-time_Fourier_transform\" width=\"800\" height=\"600\" frameborder=\"0\" marginheight=\"0\" marginwidth=\"0\">Loading...</iframe></center>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_web(\"https://en.wikipedia.org/wiki/Short-time_Fourier_transform\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Spectogram Array Shape: (1025, 87)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxgAAAH/CAYAAADUo3t3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzsvXecX3WV//88n8/0Puk9gdASOtJUVBSxoOK6FhRWxbZ2V9e1oa4oq4v+1J+urruiImJBsSOiYFlApIQOIZCENNImZTK9fsr5/vG+A598mHM+k8lMZpJ5vx6PPDIz55b3vfd93/eU1zlHVJWIiIiIiIiIiIiIiIixQGqiBxARERERERERERERceggGhgREREREREREREREWOGaGBERERERERERERERIwZooERERERERERERERETFmiAZGRERERERERERERMSYIRoYERERERERERERERFjhmhgRERMEYjIxSJy20SPYwgikhaRbhFZNNFjGSscitcEICIpEfmxiLSLyK0TPZ6IiIiIiMmNaGBERBxCEJGzROR2EekQkT0i8ncROW2cznWpiPzIkX86Uba7RaRfRHIFvz+oqjlVrVPVJ8ZjfEVj+ZGIDBacv1tEXj3W5xnLayp1/8ZivPuAc4BnAnNV9bkH+NyISJWI/JeIbE2uf72IfHGcz/kuEfnzeJ4jIiIi4lBFNDAiIg4RiEgDcD3wDWAaMB/4LDAwDucqK7WNql6WKNt1wPuAvw39rqonjvWYRoAvFJy/TlV/OZYHH8k92Zf99/X+7e/5S2AxsF5V+/Z1xzEa12eAZcApQD3wQuChMTjuhCCJCMXvb0RExCGLuMBFRBw6OApAVa9JPOl9qnqTqu6liInIl0WkTUQ2iMhLC/4+T0SuSyIfj4vIOwpkl4rIL5JIQCfwLuAS4ILRetRFpExEVESWJL//SES+KSI3Jse8VURmi8g3EmrOoyJyYsH+C0Tk1yKyK7mW9+7rGJLjHCsityTneFhEXlYgu01ELi74/e0icnPR+N8jIo8Djw1zTVUi8lUR2SwiO0TkWyJSlcheKCIbReQSEWkBvrOP465KzvVuEVkHrEz+/j8iskVEOkVkhYicWbDP5QnV6RoR6RKRh0TkpAL5p0Vke7LvoyLyHBF5D/BN4OzkuVySbPteEVknIq0i8isRmW2Nq+Bv70r26RSRT4nI0ckYO5JxWcbIacAvVXWHBqxX1R8XjLtFRD4qIo8l8/cKEakskL8qudZ2EfmbiCwvkC0Rkd+KyO7k31dE5GTgawXX3JJsO01EflIw5z4qIlIwH/4ruR/rROQDIpItOM+dIvI5EbkL6AXmicg7kzF3SXjn3lqw/UuSv30qGddWETlPRF5ZcN8/vC9zJiIiIuJAIRoYERGHDtYAORH5gYi8VESah9nmDGA1MAP4EvC9IQUJuAbYAswDXgN8QUTOKdj3lcAvgCbge8AXgJ+NcUTidcDHk/EpcCdwBzAd+C3wZQi5DoRozd2ESM25wEeKxlsSIlKRHOf3wEzgQ8DPROSIfTjM+QQF+PhhZF8GDgNOAI4ElgCfLJAvAOqARcB79mXsBXg58Azg5OT3O5KxDN2zn4tIecH2rwKuJDzHvxAUaSQYb28BTgIagZcBW1T1W8AHgZuTZ/0FETkP+HRyrPnAbqCYLlc8LghUqxOB5xGiEt8AXku4R6cDFm3tTuBjiYFyrLHNG4AXAEcn5/xIcl1nAt9Krm068EPgN4lBUA78AXiU8AwWEgyZ+4uueU5yjv8FypPxngu8G7gwkb0vua7jkmt5zTBj/CfgTYQoTAuwHXgp0EAw2v+76PoWAxlgDnA54bm9hjCfXgh8XkTmG/cjIiIiYsIQDYyIiEMEqtoJnEVQzL8D7JIQkZhdsNkmVf2OquaAHwBzgdkisjDZ92Oq2q+qDwDfBd5YsO8dqvobVc2PhiozQvxSVe9X1X7gN0C3qv4kGe/PeEpZPRNoUNUvqOqgqj5OMHpe7xz744kHu33IIw08G6gA/j9VzajqnwkKp3ecYnxBVduK74kECszbgQ8m8k7gP4uOnQUuTa5htPf086raPrS/ql6dnC9DMAKnA4cXbP9XVf1Tck9/SDAohsZSDSwH0kmUYINxzouAK1T1oeRZfRQ4R0TmFGyz17gSXK6q3YkCvwb4vapuUtU9wE3sbYwU4rMEQ+jNwH1JhOYNRdt8XVW3qeouwn0ekr8T+Kaq3ptE9q4AKgnGz1kE5f4SVe1Non63DzeAJCLyasI70p3Mua/x1DvyOuCrqrpdVVsJBnwxvquqq5O5llXV61R1QxKV+TNwSzKmIfQS5mYW+CkwG/iyqvYk93Adwxu2EREREROKaGBERBxCUNVHVfViVV1A8KTOI/FQJ2gp2LY3+bEu2W6PqnYVbLuJ4J0ewmbv3AmdZigJ+ZFRXsKOgp/7hvm9Lvl5MbCowGBoJyi5hQpuMS5X1abk39B284AnVFULtiu+7lKw7sscgiL7YMEYrwdmFWyzQ1UH9+FcJc8vIp8QkdUi0gG0AVWEiNAQWgp+7iW5p6r6CCF69HlgZ0JZKjROCzGPcJ9I9m0HOik9X0b6fPdCopB/XVWfCTQDXwWuFpGlxvk2JWOEMFcuKZorM5OxLgQ2qGreuM5CzCF8MwsT+AvnyryiMQx3/cXP6vyEIrYnGdcL2PtZ7SoY25ChNqJ7FhERETGRiAZGRMQhClV9DLiKYGiUwjZgmojUF/xtEbC18JDFpyg6X2ESskVjGStsBtYWGAxNqlqvqq/Yx+NsAxYW0MRg7+vuAWoKZMMZMMX3ZQg7gEHg6IIxNqpq4wj23Rc8eQwRORd4P4G61ERI9u8DZPhdiw6k+gNVfRYh4lEF/Iex6TaC4j503kZCJMCbL2OCJNLwVULxgmMKRAsLfl6UjBHCXPn3orlSo6q/SmRLZPiE6+LxtwD55NiF5xm65u0Eyttw43naMUWkFvg5cBkwS1WbgL8ywmcVERERMZkRDYyIiEMEInKMiHxYRBYkvy8k0ETuLLWvqm4Gbgf+M0nIPQF4G/BjZ7cd2MrZeOMOYDC53ioJ/SeOF5Fn7ONxbidQgz4sIuUi8gLgPODaRP4A8GoRqRaRo4C3Gsd5GhIK0neBr4nITAlYICIv2scx7gvqCZz9XQTq1+cIhkJJiMhyEXleQgXqS/7ljM2vAd4hIsdJSFr/IoF61WJsv19InvNzkmddLiL/DKSBwuICHxCRuSIygxCJ+Vny9yuA94vIqckzqEsiBzXAbUAXcJmI1CTP+VnJfjsIxmc5gKoOAL8m5CbVJtGTf+Gp3JNrgQ+JyBwRmQ78W4nLqibkc+wE8iJyPnD2KG9RRERExKRCNDAiIg4ddBGSuO8SkR6CYbESGGmlmTcQkpC3ERSpz6jqn5ztf5783yoi941qxKNEwkk/j5BMu5GQZPxtghd9X44zALyCkMC+G/gv4EJVXZNs8mWC13knIcHW7Pth4MMEGs0KoIOQZ3DkPh5jX/A74FYCN3894Zp2jXDfauAryT7bCdSbfx9uQ1W9npDncB1hvsxh73ydscYA4dnsIDyLtwD/oKpbCrb5KfB/wFrgYZIcCFX9O/ABwvxoJ+R+XBhEmiHMoxMJBQ6eAP4xOd4fCXNrp4gMneedyf+bCNGG7/KUEf5NgsG6ilB84HqcEtGquptghPwOaAX+AbhhxHckIiIiYhJD9qYeR0REREREHFxIkvZfo6qTqVP9qwh5P0dP9FgiIiIiDjRiBCMiIiIiImI/ISL1IvKihK63CPgUIRIYERERMeUQDYyIiIiIiIj9R4rQq6KDQJG6DztJPiIiIuKQRqRIRURERERERERERESMGWIEIyIiIiIiIiIiIiJizBANjIgxhYhsFJEXJj9fIiLf3d/jjCVE5GIRmbBEUBG5VETMSkQi8oiInF28rYgsEREVkbKxPN9Uwgju/UUictMoj/3kcyux3dkFFYnGDEkJ19Uj3PYqEYnUnTFA0lTy8NJbljzOqNfKZP9Rz92IscN4vd8REQcjooExRZF80L4wnudQ1S+o6tv3YUzzpvrirKrHqurNE3HuA6V4TgYFdziDTVV/rKovKtruWSJye6njjedzE5Erkr4P3vn/dqCrFYnIuSLyfyLSJSKtIvKAiHws6Ysx0mO8OXkOI1onRORDItIiIh0icmXSswMRqRSR74nIpmQ894vISwv2G3re3QX/Pr3vV703kqaS68fgOPu0Vg6z/9PmbsShDRH5kohsFpHOZN5/cqLHFBFRiGhgTF2cx+SruX4eofZ8RMRkwWR4T14yCcawF0TktcAvgJ8Ai1V1OnABoZP1cB2shztGM/AJ4JERbv9iQgO9cwj9Wg4HPpuIywhduZ8HNAKfBq4VkSVFh2kq6DZ/2UjOG7H/2NfI64HCZB3XCPE94BhVbQCeBVwoIv9YYp+IiAOGaGBMQSQf9qOAO0SkWUSuF5FdItKW/LygYNu9qErFNBMReWPiPWkt9qAMs+35CZWkXURuFpFlRUMbVplLvJNfE5Ftyb+vFXguzxaRLRI6/e4Uke0i8paCfaeLyHWJl2cFsLTo2M8SkbsTj+jd8lQXX5IxXiYif0+8ojdJ6BLs3dshT+k/J2PdLiLFje4qROTq5JiPiMip1v3eV4jIYSJyS3LsPwEziuQ/L/AA3yoixyZ//2fgIuCjiXf3d8nfPy4i65LjrZJQ23/oWEck5+oQkd0i8rMC2TEi8icR2SMiq0Xkdd55nOvZKCIfEZGHRKQn8VLPFpE/JGP6czKfh6UnOPfz1uT/9mQcz5Th6XPnATeIyP+KyJeLjv1bEfnX4vN483WY6zPvbyI/gdAcblfy3hxXIJspIn0iMqv42kVkWTJ/25M5dr5xfvf9N/YR4KvA51T1O6q6B0BVV6vq+1V1bbLd6SJyT/Lu7RCRrxYd6j8JzfN2e+crwJuB76nqI6raBlwGXJycu0dVL1XVjaqaTxoBbgD2tbN78bVelTz7PyXP6BYRWVwgVxE5Ivn5ZRIiJ50SPMuXFmw3tC68WUSeSN6XTxbIC+mQVSLyIwlraruEdWl2IrtYRNYnY9kgIhcV/P22onG9R0TWJtteJiJLReSOZHzXikhFwfbvEJHHk/f1OhGZN4J783V5yoN+r4g8p+h6fpFcRydwsYTyvZcUzPd7RWRhsr25Dg9z3otF5DYR+XIyZzfI3tGqeck17Emu6R0lxnWphHXxR8m4HhaRo0TkExK+KZtF5EUFx3iLiDyabLteRN7JCFAwB96SHLNNRN4lIqdJWN/aReSbRdf5dxH5RnJfHhORc4bkyfvWU3CKPHDESMYSEXFAoKrx3xT7B7weuCb5eTrwaqAGqCd0Z/5NwbYbgRcW/H4p8KPk5+VAN/BcoJKgdGSHti/a9iigBzgXKAc+CjwOVCTycoKiUV98XuBzhK7Us4CZhG65lyWys5Nzfi45xnlAL9CcyH8KXAvUAscBW4HbEtk0oI3QgbiM0Mm6DZieyG8mdEQ+itDl+GZC4yzv3i4hdH6+Jjnn8YROyoX3pD8ZZ5qgZN053P0uun9Dxy0rcf47kudQmTyXrqFjJPK3Js+5Evga8ECB7CrgP4qO91pgHsEZcUHyDOcmsmuATyayKuCs5O+1BG/yW5L7ekrybI+1zuNcz8bk2c8G5hO6ON8HnJxcw18JHceH5sKWYfYf0f0kKKu3Ffw+N5kvktzLzTxVea8Z6APmjWK+bhnJ/U3kHwf+M/n5SuDzBbL3An8sPi7hPXgcuASoAF6QzIOji+8/Jd5/45kck9y7JSOYi29Mfq4DziyQnQ7ck1z3zcDbRzAXHgQuKPh9RjKO6cNsO5vwnh1T9Ly3Ejp2fx+YMYJzXpXcu6E17utFc0SBIwqewfHJNZ1A6Dr+D0Xn/w5hLTmR0OV72TBz852E7t41hDXiGYQO9bVAZ8FznMtT79TFw4zrumS/Y5Nz/YUQ9WkkdBt/c7LtCwjv5ynJNX4DuHUE9+afkvlTRuhY3wJUFVxPhtCdPJVc80cIHdaPJrxTJyb7u+vwMOe9ODn2O5L7825CN/mhd/MW4FuENekkwvp7jjOuSwlz5cXJ+a8mGKefJLxL7wA2FJz/ZQRHlRAiZr3AKdYaVLDf0Bz432RsL0rO+xvCWjG0vj2v4DqzwIeScVxAKIE8rWh96E6Oux5YMJJ1Nf6L/w7EvxjBmJp4GUmkQFVbVfWXqtqrql3A5wmL5kjwGuB6Vb1VVQcItIS8se0FwO9V9U+qmgG+TFjchzxVzwUeTMZQjIsI3tKdqrqLQIt4Y4E8k8gzqnoDYcE9WkTSBOXp3zV4OFcCPyi6D2tV9YeqmlXVa4DHgFcUbPN9VV2jqn0EQ+Wkkd0aPpuc82GCMvOGAtltqnqDquaAHxI+tPsNCc29TgM+raoDqnorQVF5Eqp6pap2Jc/rUuBEEWm0jqmqP1fVbRq8wj8D1hKUQwj3fTFBye7Xp7oovxzYqKrfT+7rfcAvCfNlNPiGqu5Q1a3A34C7VPX+5Bp+TTA2xgPnERR4Tc6rwJCX9jXAHaq6bZj9Ss3XJ1Hi/kLBu0qgIxXOowuTvxXjTIJCf7mqDqrqX4Hri/YdOv9o3v+hqFjL0B9E5KeJB7ZXRIauNQMcISIzVLVbVe9Mtk0TFMD3q6q1XgyHOoKCNYShn+sLNxKRcuDHwA9U9bHkz7sJ78ZigsJen2wzEvy+YI37JPDMIc97IVT1ZlV9OHmWDxEM8OJ7+VlV7VPVBwkG03DvfoageB+hqjlVvVdVOxNZHjhORKpVdbuqevSyL6pqZ7LNSuAmVV2vqh3AH3jqvbkIuFJV70uu8RPJNS7xboqq/iiZP1lV/QrBOCnMA7pDVX+T3I8+4O3ApzR43lVVH1TVVka2Dhdjk4boWY6wps8FZifP5SzgY8ma9ADwXfZ+/4rHBfA3Vb1RVbMEI3sm4f3JEJxUS0SkKbnu36vquuQabgFu4ql1YSS4LBnbTQSHwjXJWjG0vhWuZzuBryXftp8Bq5P7RTKWywlz+RTCt6Tw/YiImFBEA2OKQURShCjCH5Pfa0Tk2xJoTp0E6khTogSUwjyCVxcINAWg1dl2U8G2+WTf+cmfPK77XvsmPxeG8FuTD8MQegnKyEye4mYX7msdd0g+v+D3loKfh447EhSfs3C8xceskrHhAs8D2nTvsPmT15dQFC5PKAqdBK87FNGoCiEib5KQvNsuIu2EKNDQ9h8lePFWSKDhvDX5+2LgjKF9kv0uAuaM8rp2FPzcN8zvI30m+4on52RiZPyUp5T0C7EV1FLz9Ul49zdRaI4hREAgRGuqReQMCTSdkxi+U/Q8YHOR8l48r4fOP5r3f+gdnzv0B1V9vao2EaJLQ/u+jRD9eyyhvbw8+ft7gIdU9Q7nHMOhm+CRH8LQz086JZL17YfAIPC+gvF1q+o9iQK7I5G9SEQKj2ehcI3rBvYwzPNMnsv/SaCbdQDv4unv1kjWkx8CNwI/lUCx+5KIlCfv9QXJcbeLyO9F5Bhn3CN9b4rX5m7CM37afCmEBFrqowl9p50QGSm83s1FuywkRISLYa7DEqqjDSXlFxpTT95HVe1NfqxLjrVH93ZUFc/94nHB0+/N7sR4Gfp96PiIyEtF5M6EgtVOWCdc6myJc3nr2dZk7RnC09aSxNC5P9n3s0RETBJEA2Pq4TSCd3lX8vuHCV6nMzQkiz03+bsk//cQQvVDKFQSt1OQ0CkiNQTP23DYRlA8h7aVZN+tyZ/OA34/kn2BRcnfSmEXIcRc6G1c5Bx3SL6V/UfxOUcy3v3FdqBZRGqLzj2EC4FXAi8kKANLkr8PPeu9um4mSux3CMrY9ESBXDm0vaq2qOo7VHUegdbxLQl89M3ALaraVPCvTlXfPdx5xhB7zdVESZ5pbOuOIfGCPw/4U8GfrwFek9yXMwhRmeEwovla6v4SKBt/GVJ0EoPhWoKRcyEhejhcxG8bsDBRtgvHMNy8LvX+D4fHkmO5CaWqulZV30Cgf3wR+EUyN88BXiUhF6iFEMX8SiH/3MAj7O3xPxHYkXjBh9aU7xHoUa9OvM/m8JL/vescQuEaV0eg9Az3Pv+EQEtaqKqNBCrMSI6/98CCt/qzqrqccG9eDrwpkd2oqucSjLvHCPNnf1G8NtcS1nFzHZSQb/Ex4HUEOmoTwXteeL3F79hminLghjt/gkUE5fpv+lRS/rEjvJZpIlIY1Sqe+6NefyTkUv2SEIGfnVz3DYziOY8Q85N5PQTvW1LG8Pc3ImJCEA2MqYdCygWE8GofIdl1GvCZou0fAF4vIuUSkpELaS6/AF4uImdJSBj8HPacuhZ4mYickyhvHybwgm8XkcOAygI6QzGuAT4lIal1BvDvQMneDoli9ivg0sRTu5yQKDqEG4CjRORCESkTkQsIeSXXlzr2CPDp5JzHEnIRflZqh/2Fqm4i8No/KyIVInIWe9MM6gn3vJWgiBeXKd5B4GgPoZbwMd4FIbmR4GEn+f218lRCcFuybY5w/46SUACgPPl3mjyV1F98nrHCGkI06GXJHPsUgbYxHHYR6CbWOJ5D8LIPUVNIvIS7CJSLG1W13dh3pPPVvb88/V2FoMReQIgIDUePAriLYGx9NLn3ZxPmwU+H2bbU+/80JB7VDwOfkZAc3CwBRxKUe5Lr+ScRmZkYRkP3Kkfgli8jRGBOIpmzBPqRh6uBt4nIcgmJ/Z8i5EgM4X+S476igPoyNJYzRORoEUmJyHRCcvnNCV2oFM4rWOMuI1D0hvOC1xO85/0icjrBCNxniMjzReT4xEDuJFCmchKKG5yfGAADhIhOzjvWCPET4C0iclKiQH+BcI0bnX3qCc6bXUCZiPw7e0eXhsN3gctE5MhkvpyQPIsxW4eT53I78J8SkuVPIETSRkqHK4UKwpqyC8hKSC4fz/LAs4APJO/xawnz+4ZkHr+z4N07nZCT9ZdxHEtExD4hGhhTABLoKxclvxZTkb5GyIXYTUhMLS4T+2mCV6SNoAQ8qdQk3N73Jn/bnmwzbB8LVV1NSAr8RnKuVxAUgUGGV6QK8R8EJeQhQpLgfcnfRoL3EULOLQRl5PsFY2oleAc/TFC6Pwq8XFVHWtXGwy2ERNu/AF9O+LYHAhcSvOt7CMri1QWyqwkh9q2EJM87i/b9HrBcAl3nN6q6CvgKIVl3ByGB9e8F258G3CUi3QTP7b+o6obEq/4iQjGBbYR7/0WeUvb3Os/YXDYkyuJ7CIrMVoKSbc3HXkK+wd+TcZxZtIlF2buGEAGylHsY4Xz17m/itXySyliwz5DxMI/Aox/u2gaB84GXEt61bwFvMgz4Uu//sEj44K8jvNObk/2vBa4gcNghlNd9JJkfXwden3DP25PoV4uqthDoTJ2llH1V/SPwJeD/CPN4E4lBlESD3kkwWFrkKVrN0Lp3eHJtXYQo0QDD5KQY+Elynj2E/I2LjO3eA3xORLoIRuW1Izx+MeYQnDedwKOEteRHhO/1hwnv1B5ChO09ozzHk1DVvxDW+V8S1vGlhHfXw42E+beG8Bz6GZ56VIivEu7JTYRr+x5QPQ7r8BsI0dltBArhZ1T1T+4eI0Sytn2AcB1thPX2Omt7CdXuLtmPU94FHEl4vz4PvGYoYge8ikA56yLMj28k/yIiJgWGqi5ETAFIKHX4ACEpd9I8eBG5AfimhgTtgxoSEiM3AOVFeSERBxFEZBXhY75qgs5/OuGdOL3kxhHjBhG5ilAV6FMTPZaIqQURuZhQXe2siR5LRMRoECMYUwuNwL9OJuMiwc0Er2RExIQjocJcPVHGRQFK0pUiIiIiIiImI6KBMYWgodzqNRM9jmKo6peKOdOTGSJyUQEFo/DfiDoSj8H5hzt3txQ0ujpYICKLnOtZVPoIYw8NpV0vn4hzF4xhhaoOS4E6UJiIeZZQSoY75/7QTEqd8xHjnBYVakpA9q7gtNe/iR5bRETEyCEi0yQ0C12b/N88gn2uEpHXJD/fLKFh7gMSKsf984jOO/mc2RERERERERERERERI4WEgh4Xq+rFRX//EqEAxeUi8nFC5bePlTjWVYRKhb8QkZuBf1PVeyQUA1lHqKI26B0jRjAiIiIiIiIiIiIiDk28kqeaDP+A0Ml+LyTVyL4pIqtE5PeECmbDoY5QaKRkBbuxaO4VERERERERERERETH5MFtVtwOo6nYRGc54eBWhJ9LxhHLjq4ArC+Q/FpEBQlWzDxY0ojQx5QwMEYmcsIiIiIiIiIgJgkce8VSU0aovXh/AiVCJ7PGE9i82VLO7VdVqoDqhePGLT9fW1pG01tk33HvvmkcIpaCHcIWqXjH0i4jcRSgDX0doNPlAIvqYqt44wtM8F7gmMRy2ichfi+QXJRSpmYT+ZX9Mem+ZmHIGRsAUveyIiIiIiIiIJ7F3w/u9kUpVm7Jcrmc8hnPIIJ22+y7mcp2mbARpwa5SO5Fobe3grhXfHvPjlqWf36+qp1pyVT0D7BwMYIeIzE2iF3OBndahSo1FVXeJyH2EflvRwIiIiIiIiIiIKEZoND88ohHhG1mqGVPmGRGHLBTI2/NpAnEd8Gbg8uT/3w6zza3AO0XkakL+xfMZpqGsiNQAJxOanrqIBkZERERERERExD4gJeW20ImKAOTzA6M6p4itspWXNZmyrGMo5fN+hfhS8oiDApcD14rI24AngNcOs82vgRcADwNrgFuK5D8WkT4CFesqVb231EmjgRERERERERERsQ/IO9778UprUM2assHM7vE5qYOKcqvQEOTytlHjRYbEzReZmIyRkUMnNIKhqjcTGhcX/70VOKfEvgq8z5CdPZrxRAMjIiIiIiIiImIfUF42zZQpvpKZzbaP6pye8q0ToHoPZiwqvw8v76W8bEaJc24b1TkjDjyigRERERERERExZji8+eWmrFlnm7LNutKU7ey4a7/GNBqk07WmLJPdcwBHEjARRsR4wMt7Ga3RMmkwOXMwJgTRwIiIiIiIOGgxo+EUU3ZE6nRTNiA+D35Vzx/sfQdbSg/sEMey5teZshl524jYklpnyna2H3gjwkNM8o7YJygjKoM1VRANjIiIiIgJxswGswKhS7dwE02ZGK/vgcbuzvuKyq++AAAgAElEQVRsGbasofZo97gza44xZWW1J5qyOXqYKavTGlO2PrXGHc/6tutd+VjDqx4E8Fj7r0yZlyuwoOlsU/aMpreZss6UHTHIqm8s5rDzJVr715qynr6N7nEjIiJsRAMjIiIiYoKxq/OeiR7CIYk5jc80ZaemnuPu26JtpuyRXju60V25y5R19G40ZftDuamuXGDK6irnmLK+jH3O7r717jm9ikY1VYtM2c4emwa1JXOze87xQEW5zfkfbS+HgwnecwTfWIwoxsQmeU82RAMjIuIAwPMGemUAp8IHLsKH52nPZH0KR1nannd9g7YiXF8135Q9s+w8U1aT9j8pG3I2v/re9u+ZMk+BHsi0mrJpqcWmLFWiWs08sZN4y6rtHINp1JuyzfU2tWpl+zXueDxFr29gy6hk+wNvPEuqnzWqY+7IPmbKWrseMGX7g4movjSZEA2IiPGC6BTji4mIRrsqIiIi4sDjyGmvMmXNebvk5W7ZaspGSx1qrjvWlZ9Teb4pq0jbxsltmbtN2RNtfyo9sIgxh0dB7Bm0jd7e/ifGYzijhkeJdMvmescsQYVLe93MHefY+PXPyN7rdbWeSJx6ypF61y1fG/PjljW8fNJes4eoaUdERERMYhzWbEcMpulcd98B6Tdla3v+YsqW1p5tysqpNGVbsg+641m759eu/ECirfsRV/6LEvJDAV6VpLn1p7n7lok9D1p6HzZl/QMHvszooUJB9IwIr/Qr2LLJ1mgvlbLnFUA+P4kjLpO3k/eEIBoYEREREZMYG9pusGXjdM5Vgz8dpyMfWHhRiiPLzhr1cddkbjVl7T2Pjvq44wGvGdqs2uWjPu7OPvs6J8KIOFTg5YR45V1TqQpTtj+lXw90743RdjmPmHyIBkZERMQhiSOmvdKUnZa2Fc9S/qe7sveaso1tN5YaVsQBhBelWMH4RChObXqHKWvGzqlaJ6vd446WCuYpl1vaJ1fPgarKeaZsRvVRpqyl28/P8BrblZU1mbKUkwA9Xrkb43HcUt2xPXhGxGhzCw9dxCTvQkQDIyJiEmNa/fG2rPxwUzag3e5xt7T91ZQdKs2cHt/zW1uGLYs4uOApOXMafJqPhx1OUrFXYOGe9u+M+pwHGh6n36vABZB29t3RbVOkPIPHi3xsz3aYsnkNZ5oygL6cXRHMK3N8qGC81nQvP2NqGhgRhRg3A0NErgReDuxU1eMK/v5+4H1AFvi9qn5URE4HrhjaBLhUVX+dbP8G4BICu20b8E+qurvgeK8Bfg6cpqqHBtEyIiLBni77Q70HWxYRMVXgKTLb2m0q08GGuU3PNmXlYvfX8NCVsRX6Xd12OVnw+7N4EQOvD8YcjjRl67K3m7KBvO9QyeaisjsemIhu5pMaMQdjL4xnBOMq4JvA1UN/EJHnA68ETlDVAREZIoeuBE5V1ayIzAUeFJHfJbKvA8tVdbeIfIlgnFyaHK8e+ABw6HeTioiIiIiYstje/veJHsKYYEv7zbYMW+ZFqmbVn7QfIxp7lKIkVVTYnc69srFToaSun6w+2RtlK+LkyUw1jJuBoaq3isiSoj+/G7hcNbTdVNWdyf+9BdtUwZPxPEn+1YpIK9AAPF6w7WXAl4B/G+vxR0QcKEx3Po5pp6JGzkmGG6+a8VMBHg8c4Ljql5mySrWf1+aUzbEfbfnSeU3PdeXTZKEpy5MzZZsH7DyTrl6783HEoYO6apuCCSCSNmU9/XbvjdFSZ7z9WjruGNUxAWqrl5iyvoEdoxpPKUrSwKDdD2Wqw0tkjzi4cKBzMI4CniMinwf6gX9T1bsBROQM4EpgMfBGTcx4EXk38DDQA6wF3pv8/WRgoapeLyLRwIiYUJzc/BZXPkunm7InxK61/mjbtaMe01THjIZTTNkRqdNHfdzH8ytM2YHmc5eiAE22Wj7ee7JI7GpHO/I2//7O9v/drzFFDI9SnbwnE0qXNrWdMT19G8d4NBFTGpEi9SQOtIFRBjQDZwKnAdeKyOEacBdwrIgsA34gIn8AcoSox8nAeuAbwCdE5AvA/w9cPJKTisg/Ax8B7HIRhxC8Mndz6k42ZR6nFmBrh82BneoJXfe3fX+ihxBRBE/Z382hn9g5GeG9J/cfwHFEHFqYbKVN02m7WhhA2ikpm8sP2jKnuMBkw4Eubxsx+XCgDYwtwK80tA9fISJ5YAawa2gDVX1URHqA4wj0KFR1HYCIXAt8HKhP5DeLCMAc4DoROX+4RG9VvYIkiTx08j604fE0YyfZ8UFp3mj0akREVFcuMGXlZXbTt0y2x5T1Ddh0HA9efwjYv94BETYWNZ9rymbpElO2DZtiOF7J/F4jwnyu15SVMgRyNjvxkMGUNCIUyE/B6zZwoA2M3wAvIBgGRwEVwG4ROQzYnCR5LwaOBjYm8uUiMlNVdwHnAo+qagfBMAFARG4m0K2mTBUpr3b3nDqb09+ftxe+qpTvdamQOlPmlSz0SqZu67rbPl+ZP55lVfaHqkbtqioPZ/9syhoqbAVoS4f3EfMNjLK0naA42mosXu5Gv7MfgDjjnVNtl8YdVPujWkuzKavAvv6mfKMpK4VHnfoOs1N23fwObA50S7ffjbq8zJ5b06vsKjhZtb2snoJ05LRXmbLj5WhTBjDohOs3sd2U7VabHlObsiOk7dnN7njaemwl8Yia55uyrgpb2W9x3p/qcntOLin3aXI1eVu5fCRrd0HPOUm6Xv7KzIZT3fEsSp1oyiqc3J+W1EZTVoW9pvdgl3YFaBuwj9vdt84eT7cdq+qvsb9Pu7secsfjYbTe9FzONmwjIiJ8jGeZ2muAs4EZIrIF+Awhx+JKEVkJDAJvVlUVkbOAj4tIhtDn6j1DpWhF5LPArYlsEyOkRR3q8BRPr0rHwQTvGgHWpm8zZeVOxRGv0+6CCpu3/+KG95uyPkepAGhL2R/rSq1yZHYofWfaVhA7Ulvd8XjGZEfW3rdn0Fb06qqfZcpSjkGzPuU3PPNKUM5P2Z2I02ovbxUp20h4Zs3r3fHkHYWkRWxvehrbCD962mtM2bSc7WkvK/er1ZSl7UTcx7v+z5SdUvUP9kEd72sunXHH07ZXjY69sbL9GlN2QtNFpmyZYxB3O0bdYzk/Mfj5FbYDY0GFbfTdnbUV4RdX28/5YfWT5+9v/5Epm+0YJ+XYc33LgK3sl0rm97z7XsS2qsLOR2vrsc9ZV73IlFWm600ZQEefnefmIZe3DYxofEQ8HbHRXiFEJ3fNrzFHoEjF/oLDoRRvdJbT9K1zwFZK+wd3mTKwFaCmGr+KiVcpqbxsminLO8bAzDq7w/POLpu3X+dUIgHo6lljykYbSvaeV75EXXhPAfCrqthGjThdb72P8ZlN7zJlAOVqK+YPZm8yZZ2Ot3x/UFkxx5SNtjpMQ60diVhUYSuP9SWiP3f3/MSUlaVtxbOparEp8yr27E9Vqyon6ug1r5vVeIYp68vYdfq9Kmzgd3HO5vtN2eJa29DuytsG+v7QfJY0v9iU1WFHnNLOtzBfIifPi3KNR0ldj9IW6WxTCdl7VdUP900QTj3hcF1xw3+M+XHTCy+atNfsIWraUwzNjgJdKo+g1VHYyssc+lTK9nR5H4b9KbXqNQBqql1mysrEphosbTrPlKUcQwn8so2e8n160zvd41q4t8v2BgPUV803ZXMrbEMyW217qD0KkDpu7634ntLWfls+o9pWzOdX2JSSzf02m7LXMZYBahwP7PQam5a1vcOOuE2vWGrKZuRnmrJUiXr78+rtTtZ7+m0aS07t5/zMxveYsibHWw6Qcbx7q2R0DNf23g2O1D5fqZ4CntLuYW2b3SW+qe4YU+Z11Qaodt7ZjW03mrJpjmNodpm9FpaiSPU492+0TooljXbUKIu9vvj01cmXBB4RMRUQDYwphrZum47ieWbB91B7pf5K9RU40OjqtxVIjz7lJahWV9hcbxh9OH1F+7dHtV8plKdtRbglY8+RDicSU+k0j6qrtOdWfdqfd41VttfbM2oGxM4XKXdyYsrLSjxLJwLW5pT29HJmWnptWs2GgRtMWSmnQCplG/5ejlNt2jai1uXtHhmZrB856+yzczS8xFiv5LAXjdsfJ4WntHuoqbKpPBXO86iqnOsed161XQEwW2VTBauxn7NH2ytVEMRbD6vK7cja9EY7yrWh7femzPv+5B2D+GCD907HYiEHAeIzehLRwJhAeJ1JARodipCn6HkeKy83oVRYe27Ts02ZOJ7bjgGb/+qV1C3lYfSUDu9D7iWkZ9WmPngeu47e8akZ73Hz6/O2IrxFV7rH7Rq0vYgVDrfaU4I8Skl/1u5jUIofPc8prTw3b3v+16mtCPdlbO+slxgMsLTcpsA4+hqrB+ychzk1J5iyw6pea8qeSPvzbk/Glu/petiU7XSUnDnV9lgbnZwYgI7yw0yZl3DchU2z9CoPLW98jinzij0ATHOiSrPUVq53Onk4XTmbQudTSSHnRA+7c3YkeHOvXV68VCEND17Vr/ry0TmV5jbZz2t/mul5RVFyOdso9rpqjxeiERFxqCDmYESMCUZbpcPj1ZbyztZX2R+xQSdi4HHzvfFks7ZS6hl14NMfPO+b18vByzOZ71BjwOdlt+pGU+bd19oym8rjeUqnqe+5zYr9kd+pdtKwF2lYKMeZsjq1FSeA1lSrK7fQ6BiEHU4RgFXtdrPF6fU2DQygOm2fc57aCnSdkw/xgN5sykrRfLzKXtXOOT3sEjsq4nnv56tNOQLoV7sfwcbUKlOWyduRM8+wb6y0Ix8A7f0bTdn8mmeYsoa8vU60id2penOXbZgAHN7wAlPWp7ZDYbSl0r08paYK/97t6LGN6dhV+2DDZM7BOExX/O6zY37c9JI3T9pr9hA17QmEF2IGmF1j50t4CXiDToKvl4xdW+HXhd/VaXOk02V2SNwriVpbaSul5Wnbowl+kqZnRHj3fXr1EaZsUO1mgoNZv+55vUOhqhHHY+44GJvKbOpQW2aTO54esZXknEM76s/YioMX+Ug5JY7va7/SlAEsa77AlM3H9pjf02EnBndW2kpp/4Df/9qj/B1eY9M/bu/8linzaDVHN9kVi6rwjaHd2NGhrWLntuSwjd75KduY7hX/PdiVsw3CAYci5VE7vXu3sMYuRXv3oE09A5hRaRtDXpK3R/lrz9vvZX/Oz3noH7SNgZaUHbHc4ERGvDU2W6KXw9Y+2/nhdQE/r+kjpiwltqNqrdjH7Mzbhhv4je0iIsYMsQ/GXogRjDE5ZuRMjgdKGWCVZXZpQo8C431splfbfQxU7UTljkE/MXgw22XKPGqRd42ekelVHgLIiG1EbO5dYcpmVdtJod54mrGV8mbHsw+wPW1TTlbv+YUp8xLke8U2wtvVf5bdGVvRqyu381DqxTbgu9SmuPTl7HLNi8ps+hj4vRzKnXWwX2wFenX2b6asVNRxXplNr5qet/M+usWOnG3jMVNW7RjvlSUS0j00qh0V2J2ylV2PgrlM/Shop7PvGr3TlJ2RtiMNm7Hncqv6Topl2FGTFrHn8+puO7dltB2uZ9f59661zzame/tHV8I2YqIwiSMYxx+mK667dMyPmz784kl7zR5iBGOE8OgoXq1sdbzBpeAp2N5C7DVuG8zYXimvEhRANmd/4LzxeJ125zQ+0z2nhz1Onfaj6uwKMMvTtsezP2ffuxVqVypZWGl/bAFqKmxDYYfjmevP2RGDSifPZI/6Dc9aOuwGdV5uUKV4Bo9tgK3q/J0pS6ftPiAAZ1a9zpSdOOMTpmxnxo44PTr4V1NWqv6/14l4ft42ULeLXbXJ611zatM7TFl13r939w/aSbOel9lr+ja9wo7yeT1UwDci0k4lNi8Kmsnbz1nFfp83dNj3BmBx84tMWY/YDown9oyOAtTb6Pf9OQE7B+4YsdfRWwZ+bcq6e+2IUqny2T11dmSk2cm1mV9nz60NbX5UycLmNrvxYcTBBS8vE2AwM5kpbbEPRiFiBKMAHsfzyPKzTNkesT1Wo10wxwtecyQvqRwg63iQvDKsFU5VnqZqu95+l1POEGDQqVizuN7+GJc7XW+rHf59d8pW9ndl/J4LHsXDKx3c3W/fg/oq2wCdV+Z79GY7pU9XYidTeqWKj6y3FTKvjK+ndAJsTdue1Fq1kze9DuGewvq3bp+yVVlun9Pzhp7d+EFT1pi256S3RM+q9nMeNvXayreXY+BFN1YM2mVYZzidzAF6cjY1b17afg+Wpuw8nUfztqHU6lCSjsbv5H1Xr13q+aTaV5uyzdh0pSqx5+SivG24Afyty6b8Hd5kv3tLck4hhJQd/Sn17fLKFXudzmsdBdLtW+L1L6r0O9p3ZO3vU4dTAdGrbCZOvpE45c4B8jk7on0wVcvykuc99oZ3X0tjMkcwluiK33xmzI+bPuKtk/aaPcQIRgE83v6qSjtK4ZXg3B8saDrblHVmbMVzdqXNS690Eim3OR1oAbqdhbim0k6Y9PjK3gelt4SB4XV2nZOzle8FTjTB8z08nLXnQO+gn/g7t9E2UL1Sq3OcCkoVYj/LR52IAcAupxfIAicBekmdnVS8JmtXITslZXv9W1I2TQNgbs5+zkuq7CSVrf12su2dA780ZfMa/AR5r2pRTaX9TB4Xu2TqTLU9voelbNrV2h7/Q31f5g+mrLvP7h9xWNNLTFmdU464FM1nk0N3O6bMXkP2ZOx3pBZbaZ/hGBHr5EFTBjCtxjaWatSO8s0Tu9dFL7ZieUvHf7njuWDmJaasJ2sr39d3fMmUeQ6nUriz439M2RHT7E7wzXmbKthXYb+zHmXLi9oDVDpNSbNZP3JkntNlJ4yeueA9kzLH+ChVdXE8MNp7d0gjRjCeRDQwRgiP5uPJ9gcebcLDeHUw9uDRLTwMZmzFvLpEXXjvOlc32p6VdRnbu+SV1PUq5MyothNCwU+arUrZH7+M2h9czzA5qc4ubQqwDfvebcjdbcpG21dg4zQ7OlalPjXvbx3fMGUPZm3PpZf34iVye53eAXoqbH55U5kdkct7/Rqw15BdOdsQWIxtDAIcUWkbtmudnKKUOHQlJ89im/ilVnfm7PLatzq0PnEiYF4HbI9uMdsxlsHvafJ4tb2GNIltKO3M2tdfql/QdZ1Xm7KaCp9WYuGEBpt+WAoPddpG+rq2602ZOHTaBqdBn5eQ7jXyBL/6nUcJzTv0u/FCVNoPYihINDCeRDQwRgiPd+3BK8lXKiFyYdM5psxTPHc4pU0rnBr/pRLSR1vOz/PIeA3ovOZ9UKK2uVNhqqHcjjgtdbo/r8WuolWd8hOVPYpQT972PO1yqFXplM2/76vwIypeSdkBpyJW2vEEHtdgVzsqy9tLzWMDo+dPL66wPdQNTmWzNZV2Uuy0siXuORflbSPiCbG9rOsdpevw5pebsoVqR5s2pXxnwtYu21icW2/nDc3M2UryUrE9++UO9Qwgl7aNtyonurqyz86X8HLVPOfP/vD2t2VsQ6rdieZ6UdlZTiNGgOPF7r+yW+xIMHbAlha1DR6PkgRQVWHnJS6ottkcNWoP6P6275uyWEwlIuLgQjQwCuAtYHXYPPF+7FyAVMr2dOUdJRigtd9ONq11aApVFbbymM3ZhkmpRdrz4HsepNF2sS4Vvq8qtz9w1Wnb+PA85p0pW7nu7nfKRPb5TaBmNZ5hyrzeAHvKbC+qpzyVMgarnKaJHqXNKwTwaLetBB5R90JTVsqQ9Iz7bVm7vn2LQyfY7URidqltSAJO7Md/32sd76xH8fByUJ5wvPfgrwUN2FSV7pRT9Sxvr5Prud8dz06nud+z6t5syo6ssZ0tG/ptap6XIN+Cn8y/tf0WU7a80W6A6SFXZUcyZ+f9CMZ2sd/pHTl7VnrfNS8yVJuyDXSAmip7jfUKPjzQdpUp84p+eD1ESq0hlRW2U8mjFk1Eo72IgxnqJ81NMUQDowBePfUGJ5k0k7INBc+IKBUSX1x1pinzPOK9Dk2h0olg1JXog7G7+1FTVu30s+jqtavnzHb47jUpP/nX68i9rc9WdHKV9kfeywlpcDylc6t9usXmHrtq05FV9j1YVmUr1/f2f8+UeQ0DATr6bTpKQ5XdXyPtKO17nOe8qu2n7ng8eFFAjwKTdbqHe4pDqaaJXpd4r9uwpwRtKKEgjRZe0vnK/h+P+fkOaz7PlZ9Wf6Ep2yo2FaxZbYfKrCo75+zetu+aslJVkjysbBv7e2evrgFeEZKTy+yqeV5jSC9Xa279Ke542gY2mjJvrnuUpDInOXp2lU0H7CpRecjr4RQRETE+iAZGATJOVaLuKtujV6M2baTeSRTs7rMVMoBNKZvGUcpjMxp4tCKAI+odLzR22UYv6dzD2j12ecX9gXfvPNpVKmW/Lh4PvBT+PvDNUe3nectL1YX3qih5Cr1nhE+Ety/rVGNJORGwvMNzri3zjbOerJ2D4UWqFondA8KrROdRqyYCTU6BAK9zNkDW8WwvzNt5OnnHGNjmrD2eESHYTd0Aqh1De0m1TVfyKEBbsTuAt/f5fSe8nLNb3Lja6DDaHMBS8PIavJ4vW7puHofRRESMMWIOxpOIBkYBvFDpY102/aOm0vauDTgN3/aHkuSVsqt2IiPza2zetVfdA2DNoK1Edzvc4jonWTvtKMml4JV39Wr1r2uzn+U0x0uYydkG6PnNHzNl4CtId+b+bMqq03bEyWvatT/88nlNdjfqmWJXO1qI/R5UpGyDZlPOTwz2cH+HnfjqJbr3iucwcEjrwAZsj3Bdyo7k3dNmlxn1vNP/OP3jpmxJnV+mtjNjz7u7HIXWo2x5BQua874neZaTqLs1b6+V5Wpf52JsY7q/zlZYvbLbAAsq7Qpuldhe+Ae6f27KUmInOJ9Q+0p3PNOrbcrS9Ar7uO2D9vPa4OQMec0fwZ8HXv5GW49NTatwnAKe8e5FngGqnHysGN2IGDMo0cAoQDQwRgiv9r3XNbm1fKMp29nlV6joGWWHUS/PwivfWleidvcZ5S+zhY6e49XbzzqKTH+T37xuFrbyXa42hWx2g+2ZTDlezXza8aKq/4Grc/prHO0kb85P28puWcoea1uT38DwD+1fNmXTxL4/DXnbQz0gtnd6a942Iu5vtxM7wefRH9lsK2VNOefeORTDreJz8738nuOwjd7jp9my+5y8j0fUbob2q012J3OAc5s+bMpmqm0MNJXZa8Ecp/fGxm7b6AW4R+2kc48e06O2UdfZZ9P9zq+38zq6ndKuAGVOlC/rOId21dnJ2kvVjmKVYmwtqrWLOnRn7PFsdnI36tVxYKT8ngteP6GZajurNtTba8gytamm69L2e+lV4gPY1WsT0LzcqHzeniPjVT0yIuJQQTQwRojpTgOpbuzIh9cxeW6jrViC/8HtythlNj3PXItTh96jegHscUppeh7z0UZb+kv0wVjeZJdY9CrStDgKpBc1asYe60PtPifbow8d13yRKVuXudc9roXmlG0kAJzX9BFTNrPS9oZ6yDsKUnPOTrJcNsuu7w/QOmjP58qc/Y6cMs026hrKbdnuATsKAbCy3S5a8JDT66JZ7fnTiH1/etXOJXnXgk+bMoDmCtsIvaPVLmhwe9ZOcD6+21636tN+ROU07HyjIxrsffudKqQ/UTsi6XVzr3PKpQKcMsMp0OHM9UX9zzFltc4Xd1uPX2q1bcCWlzvOhlMqbVrjyn7b+Cj1PXiky2642OkYWSnHG9XpVEfck7ELXtSV+b2oZtbYtD7vW9rWZ1fxi4h4OtRfHKYYooFRAE+59HoODOZt6oynWJbqytnq9GQYbQ7GbCchskf8+tuitkdvtKUivZB4f4kyvl71mL7+rabsmCanC++ArdDXVtrevgVNzzNl4Fch84whr8KU15isZdCekwBt2EpyV5/thW5P2XOkHftDPQNbyanL+H0wpqXt+1ODve9f22zlySuNW+1UJwM4Ruzo0Ckpm1azJm972o9yDMINeVsR/vEeP3n+6IqzTdl0p6v08dhGhDpVUrxrBNjYbyfB/7bddtScWfcmU1buJA3nnF4fK/K2EQXQ0GFXrurP2RGD+/N2w9LWQTvvbmmF73Ca60ScMs54VmbsXL5mJ1o5J+/3ITqt1k7Y3yH2t2ua2sdNOVGjhjLHGZX3G07mnG+tFwGLiIgYPaKBUQCv1GqV2N6caWl7kS5vsj9+pdDoeJ6mVS01ZV4S6vq2G0xZKYNnRoNdVaS6wla+Z9fY1BDvw+B/MiDnRGq8a2nJOr0lnGjLlgG7MtX0SjtBFeDRtmvtczpG1hGNLzVlf+i1G12VlfDOHpW2y9TuTNnzpzFvG0pHp+2eFDlHKe0Vn6oymLffyxaxqTMt2DX+l1fa1bkez93ujmdj2k7U/VurbbgcOc3uE7Ld6T1S6UQyX990gSkDn3XT5dBqHs7a3uJ+sY3X5fgFHeZU2B2eOytto7fMcW4crrZRV+XkPDy/3DYgANJiRwU8KqVX3c6r9tTo5HUAbBb7vfSMjxOwDeJGJ+K02nEYAHTgUK+cEsgeXXRdyi5jvDRv59pUOus2wBMp2+DZlbPpiV70PZawHR94hVYAstkD37F8xIg5GHshGhgF8JKu27N2Mlx3yvaweh4ZL2ICft1vr0txWdrm6pY5yW6pEkrpbqeBn1cta2PXje5xLRw9za81Py3n1PGvse/P6u4/mDKvtOkZDXYuQIXTSA7g6Cb74zinylYgM0649Xa1r7Era/fsAFiTs6M/e9rtj7yXaPlQt+0JPKzG7ihd43UCA2ZjRxRmqW3wZMU2+ury9jt7UsqulgbQ7dA4tqXtHAOvtLS3TjSLHaWpLvMrIQ3k7PnTl3U6eTtjXZK33/XDGvw8rp399nuSGXRKBztRrKq0fe+8SMMjzpoOfi+HakdJnqu2Yyjt3FevEARAudrrc5dT8ME757W77Vwsr9gDwNkV9js9p9o+Z+uA/Uxq+m0n1pcIdSUAACAASURBVKqUHRkqhS1dNlXZq4znObFG23g2wkfsZH7oIBoYBfDKFnoczz39dti7b9C2tuc32MoawOJK2yOcqbIpW25n6Nxjpqx/wPdYeZ6FeVV2cl7OoWV5EZU5Ob9PSKkPsoWTam2K1E7ZaMo6sBO5y/EVqzJ1mr45ebFP5B80ZXPSNq/4LKeTLkBnxlbm7muylfbpYnexPrHaVji61PZOexWCANak7UjEUkfZne5EW2aU20b47oyfqLwuZysrJzTYc+vIlL2G5B074bG87X1t7fCbodViX2d92lZYT62wq4V1Z2zF+zfdfvSnycljqhVbae/O23Pkpj47knd4lZ0Psd2JZAL0DPpVlCykHUfNEVU2lXK2Q3kEmJ62oxQ7++3vwQqnSp0HLxIDcH2Pfd+nZZaYsiV5e93qSNkOno6MTbXtGfAr0TVU2ywDr9lgq9OQMyJiWMQIxpMQj097KEJEdKztKo/60OeUXhSxFzaATN72lHrdWbudD6NHAWqosPMowK/EUVVuKzpt3faHvLzM9k5nsn5lJi8CNLvRNs7STpKhR9nyIjheDgrA9Gq7bK5X+92j7VWmbdpeqZr6Xl7MsmY7eb7N4dg/M20rT82V9rwbdLzsAD2Op/3G3p+YsunVtvFxXo1NGymVo7e2x6YI3TN4nSnzcjuWpmyD8IgK21AqFcFwcn9p6bNphD05R4ZtgA2K30unxsk3Wl5nryEeXcmrXDXo5GDUpHwF2jklv2v7qil7bsN7TdmWtN1MsF793J8lYpeB9qI4M51owo4++/60Z/wyvt1O1KRf7OT6R/vtPjuNTu+RupQdsa505hX4+Xrj0VMqYjyRvVdVfQ/aBOHU5Qv17h/alftGi9SpH5q01+whRjBGCE+B7MrZodL5YnvvO8VXoDuythJ4TPnZpmyV2gt4c6WtdFWIv0h7RkR/xvY8efCMCK9LM0B9lf1MdjlGTaNTltArQeo1tvMoQOD3Vnio324ouKzOLg1crU5ya5XPD/YMDC9fxMOvsRN4PZzYbJcSBT+Z/YXVbzBlNw/+xpT9Tu2eLi2ddoQC4Iz6t5my4ypfZMoez68wZZuwaWkru2yqZK/TkR3gvKYPmbLaMo9fbh9zTrk97zYP2LQ9gPsyNj1xY69DVXFyqjx62e5+O/pVKmK7vPn1puzUxreastu6r7L3q7OPeViJ4gKza2xDodvpd7KywzaId6XsaPfqHp/aOpixHVmek+LsKjtvaLeTefdAlx0x8XpYgZ/n5q3r+RLNZyNseIwQrwFmxKGDaGCMEHknoasmZXsYN2RtpaK90/74AZQ5PRBW9Hzb3ddCd5+dvDnZUOqjMVBme/C9brHjEfZ+rMPvOl5eZifBN9XYCfs1efvD2JuyaUcNDh0HoNnJb+lVp+GZk4ha6/QlmePUxa8usQx5UVavH4HXmd7rvO7lEwGsyds0IC/iNLPcbqY36OR1dGJXRPPyggAayu17+1DGjnKlnWfSkXFKm6bsOQCwrOIFpszLi8lgr79e6e26Svs9OLnKTjgHmJ+yHSq3OzlML6qzn0lV2jYSGir8qnmbumwja7VuNGWDaXst9Ay3GidyBr6B4RXE6Kqw99veZUeJZ9XbeWzN4udN7crZvWRio73xwZQ0ImKS916IFKkJhFehAkZfpeKw5vNMmRelaMv53tB+J/mqysnP8Lp8946ymeB4YU6jTZ15QeXzTdn8Wp/u5ugGVDm77u6zF6vreuwma509q93xlKJ0WRjI2Fzn5lqbW11fZtM7SiWktznX8mKHjlKdtt+vx/IbTVlfifplm7tsA2OOU/9/R7edpDq33m4qOUftZPUusY1B8KNRnue2odo+5+xKuyrcwpwdhQCflrRa7Fy2o9U2wr0u8evytjE0gN/otDFvRxR2pW2jz8P2PjunKpvzc3+OrLWrXh1fZudGDTgKj1dWeGfeVsoB5qaOMWVeyfPH99j9Mzx4lNhSFRAjDiVMYorUsoV699V21Hi0SJ3+4Ul7zR4mh6Z9EMD7GItTRnJard3HoDbte4hyjtdua8dtpuyJTltWVWF/NL2yuAAnltn0jxlOCDpXbRuxnZW2l7nZqYYFML1qdNO3td++r2vV/uCuH7TpXA9n7GgCwMNtPzRlzXW2wra0zK6Nf2qZbUg+1mQr9OB78EcLzwCdV2Z7H5ekbMME4Oi5LzdlvVmHGjJoU2C8pPwufFrEgnq7OMN0tQ23xbXHmbJt2JHFLThdiMVfQ5Y022VRe/P28/IUtsPzS0zZLvGpkmvYaMrm5G2j5rr2L5qyk5vfYspKGREeZjkRZHV6RLSInWfhrbFnpPw+GHXltieiZcA2Tm7ttZuAek6so+rsuQNQqfb6/Hi/bUh51am8CNjuATviX1/hFwTx8hl3OlGTSJGK2GfERntPIkYwRogFTWebMs+zUumUmGwrUSbRSyqe2WAbs171j4qUPZ5SXqCerO297ui1vV1TYZFuqLXpL+CXQB5tkuHhXqTKyc8AyDm9J2Y4Xbc9WlYGOyl0a8ampXkFCwDae+y5tajRVla85phVDv3FoyuBXwhgVspO5l/Z5nd7t+CtLy9p/KC7b2O5ve+6jN1DZJfYhrZXuMJbXwAW522vd53jqPFIB1vFjpD2YI/Vy50DOKfyJaasudJW9jsHbZqcV9Bgu1N6HODuXrup4im1NuXxnOk2PXFWlT0er3s6wE4n4LK2w14LrmuzjUUP0xyKVD7vR/s7PWqwszZ7lEcP3nofsb+YzBGMBXr3lf8y5sdNPeujI75mETkG+D5wCvBJVf1ygewlwNeBNPBdVb18BMfrVtU6EVkCPAqsBgToAd6iqibFIEYwRogKx1BQ5/O3vuMmU5bL+V5vj0LllVBMO9GWnX0237S83PeGZpweEV6OwXjVC5/bZDeL8yqO7MnYHsalads7vSdlX8eMEl1vVw7Y82BRk81LPwGbctPgJOnenfUpUmvafmXLHKdDyvHqljm9CrzeLA2VfnnOWfX2PVictw27cmd5y+RshWRTyr93Z6bONmVe6eTOZru5n1fq2qs+5fU/ABh0lN2UUyYph+1sOAy7sd2iEk2yNmXsKOAdGbuc6lHltiHZ6uQfeFG1+nLf6/2rtv81ZV7PIC83wVM8mxyKIcBpNXaCeN457o2ttgHmleV+os0uFgL+eBdW2LqQR+Ft6bVphF452aYSEQzvHdrebufTREQchNgDfADYK8lMQtnS/wbOBbYAd4vIdapqd459Otap6knJ8d4JXAKYVVqigVEAr+HO/LxdF74xZXuL59XZYf+tqbXueLy63x29tkem3Gmmtz/dR5tqbe/sni67Cs5oS9GWylHZ4UR4pOE0U1ZbNtOU3dNxlSnzIjFbKv0P3Jwa2/vW0mvfu4FquwLM9p32h9H7iIOv6Hhllzf32BWWvKo8z2n8V1M2q8Knwj0+aHvaVzkJ115hBi/R3/OUAqxyemg80WU/E49ykqqyozh12NXUbmv/b1MGsLDJ5u0f5XRGPkxt5fHewetN2ZaUvU4CVDoRjlTKft/v6bzalHkFHTz6YVfGryK1oP5MU9aZtfftyNtGnzfWmZV2dAfg7i47grGgzh7rnkHbeO3otb9BFeW2kwb83DovkdszFOqrbGfDDGdubei1acGlEHM7IsYMkyDJW1V3AjtFpLgE5enA46q6HkBEfgq8EtjLwBCRw4CfEOyDPzqnagAnREw0MPZC/6CdbHqf/M6UnVpxvilbVGFXOqoctEvYAjwg9kfM+1ANDNoyT/Es1Vip16FGTHe8zHu6bD6uV4rWq64EUOtU79o9YH84p1faRt/JDf9kyrz67ZVlfjdqj1Yzo9rO05mNcw8cZ/EGp4EhwDOa7FKrp1TZ0ZgZTkdyr2v0BifLvVRp0xqnclVTmV03vzFvz60TG+2E/Q0lIhidDrVmXp2drO1RpJ7d+D5TdmS1/aBfsPCTpgxgl9Pn4GGHoilO6dcmJ4/AqzIGfoLvuU12/fhp019hylZlbTrXYuy5XONUdALoztoGasZpctnYaEc3NuRsxXtb3m6CCnB87StN2Qy1I8jHzrAjTs5QS6LVYb7+oXulKfOcFIfX2JEqr39RqTLhq9ps48z9djmFGcaLBuU51vbHQRhxUGOGiBTST65Q1Sv28RjzgcLFcgswHGXj68D/qOrVIlJcRWWpiDwA1AM1xv5PIhoYe8H+qM5yulGv0XtN2a277d4AnicH/IpGRztlRncM2D0gPOrDhj2+UuoZJwNO5Z1nNdqVfranNpqyUhzpDkexUKfB1jPLbCXQqzV/cs2bTFm/k2wMcNOATf9YgO25bHPq1O9os0sge1EIgCeyttE32Gfzp6f12IrMHU7Tuxm19vtzDPbzAOgQe249u8L2UHsdp7c6dD+v3C7ATLGNml3Yc9LzCP+945umbFXOvsYT0naEAmBA7Ge5LWcrT7PLbAV6qZ5gytLqr2nzG99vyp5I2QbPAwMbTVmZQwmtxqbt9eX8XJsyR9Gbo7Zz42G1nRsZsaMbu7r8zuIn19lJ4HUOXfJPnRtN2epuu9dFXv1GezWVdiGJM8rtwgzHOflqXuPMpdV2I882pzgH+KWn+xza3kTkUkw2I8IzeLxmwVMh93JYjE+S9+4xyDsZjhM73GCfDbw6+fmHQGHSVCFF6gLgCsBMVosGRgG8qEDGSfycJTZ1KFvvJJqmbSoTwOa2v7jy0aDTScYuVbrU84p7YeZNjbYHtjdr019KLe5eh3AvMvLLdtvwH9zl994YLTzlckXOS4C2X1EvfF/KOMs5CdBexasyh2O/pMFWAM6qsCMf5X6ONz/rsOddo9pG7xNpm0aYS9n3zjPCAXqc0rBeFPBYp2ni/W3fN2UvrbE913OcLs0AA84rVNdp54R0O+vdGrEdKqXe2eOwnSZLXRqqrdB7eQQtTnUuL+EcoC1lK55ppzDBEqfKVovYhTJKwetK3uU4OGqxK/xVOs1Te0qUEJ/lULq2q73+PLLHNiK8PEAvYf8IPcWUAbRV2s9yW9Y2tHv9FKcpAc/gmWzG0MRD3aIB44EkyjDUfOc8VbWoL1uAQu/YAsDadiRW0nWEZHIT0cAYIbZ2OJ1/0zYNqrnGNj5a+/wcDC+C0evkLng9ELzmN16Pg1LwlN0t7TeP+rgevMpNCypsWoDbdKlEcz8LHtcb/JwZr+dAe49dotRDS4ffVbvKyRk5rvkiU+ZRZ44W20DtGLQ/RB0539N1StlLTZnXiXha3q6GVa2213uV2nkdAI1OZ/H+vB0Z2dpnK+YLm+1IRNpJxu4pETlrdUoBbU3ZOV6rO2wq0+FNdnWl2Tk7ugPQLjYdziuW4aGmRMTJwuaU3+jUKzm8Qm8xZV5/CK/0+LJ6mwYG8JjaOWf1YjswTq2wvffLq+2u2oN1o/fE9mTteTej0a6y05qyHU7bsnau2vouOy8I/PzKUv1HIiImM1T1vwnJ26VwN3BkkmOxFXg9cOEw2/09kf0IsJUBOAuwE7yIBsZe8LzMJzj815nYXqBeJ8x8S/9/ueM5t+YNpixdZisdW8ptT0+f2J7JUmHmcqdJXwN24nSvQ3Fp6be5uh7NCaB/0LnOCvucx4htuOXrR5cY3J/x6/9PdwwQT2GdUWnnZ3iVq2aIbfQCdDmJqH/vshNqs05VnpXDRmADUk6FqWUNfkflo1K20rqwzFYcVg3YXtRbOr5tyo5vfqM7niPFfl6D4lSUq7ENHq9s8O+67YpfpQxQr5y1l3B9bOPrTNmylK14p9P2HAB4MGsb2jVq5zF1OIbk4Xk7TynjrCF92MUnAAbFNnynp+zGdhsG7zRlMxxlv8b5jgB0Zuyk6qpKe1+vH8yWnK20l+pw7VXxO0qdxpEV9rxrdSLa5U4xlVLIOnQ4r/pdJmPP54OpU7XHThhwylWDz+yIKIIy4X0wRGQOcA8hCTsvIh8Elqtqp4i8D7iRUKb2SlUdjgbyL8BPRORfgF8WyYZyMAQYBN7ujSUaGAXwSmk2OR+/jON5uz9rc1y9JlgAv91j1wuvrLCVFQ9pp7xiqkTVJi+5c1W3nQTvVYryFKDylL3wA+wYtBV+r1b/7d0/MGW5nG2YLHLKjFaLrxzknXrqHv1uz+BG+5xOAYEjSpQcbnaoPGnsxmXrU3ZFu41t9lxf2GAnYW7P+tzzRqf51nanI/f6dt+racGjiAGsdhwR9VV2ZOjZZfb7Xu0kHK+psNeeXqerNsDujB2te2GV/Uzqy23F6tFOu7z2+hJRgUbHEVGnNpWnJW9He9c51fi8hPxzKvz8lScG7bVgj9N7w4sg1zgV9bocWhHA0go7B+PRHvvd86q77Q9a2u1IX1+tvebfPWDPH69k9Sliz9dBJ9oPsMuJjOzI23O2twRN7GBB34AdrZxsEMdRBSPj7kxlqGoLgf40nOwGwE22VdUNsBeX9fLk7xuhRBWPIkQDowDeYvLn/q+astrqJabMWzBbB5zmP/h5BJ4xMF6LSbfXrMiBR8fp95JtK30l+Vl1ZvllOlO2clDVYCvmy9TOFcg6/SHu6P+1KQMYzNrlZmsrbaWrzDEIV7b9yJRtrbe7RgMcmzrblHmJwYNqe7O892C5kxhcVuYnYWzGrryTd/IlPAqd1wW9MW3TpwB+2/YVU9baZY/1OmyD2CvacHzK9npvzfrrfabMiVQN2obdrD573dqRtiOd7Rm/eejOrG2gLqyynQ0L1KYdtajd18arNOc1GgS/T4iXj+YlFOcdZ5QXyfx/7L13nF1neS283n16P2e6pqgXW5YtWbaFKxYugG1aCNgGTAyh5Qa+EEgg9ctNfrk3PxKSfJBLkgsJJCGF0MGmGRv3JslFsmR1aUbSjKZoZk6Z09v7/TGyMyReaztjyxpZ5/knMY/OmX32fvf7PmWt9QDAMSHMsDLMVdE2tXFuiyq2ilmCAICI6qIXeEfungpP/DNlvn52hnnBaaqs4cbqbPd4eGKrzuDqPOG0LdN2JnWGXtBOs0ztQrLWJO8Xab1JLp9XEm1dNXNhwGqZ2hUBXrk9UeXt+0lBhlOVtyMZPVhJEboiIX6I+b18A496eSdmJMflDAGtmZ70827LlT4eyHSHebCrpFafaHJ8PaCnSk8WeNB1eZjL5irbUv6W9KuA5ECRiwsoEm88yNvwbnALZUq+tC/IA+x9JV5FzQoCb5fgbgDAFR08Qa2KiG3bNOcf3J/9HPVdk/gE9aUdDllzs1Ehi1oTsIj1Dl87nX6d8KhUMiQSzbvLnFO0prme+vpC/HoUtwXQU7cVx+AJu436EuBrS0EBAeBEnVf+jzj8WR6f4dczEONdkaM5zjMBgHVxnhSnmvzsckSFWnVIK5YXaYIuHeSRPN9/aoJ7eMYHu69KW8CTvNf02W1/+ysv+/c61/7Bgv3Nyk5LgmGM+QRmsVsWwE4A7wfwZgB/COBcAJustU+c/Ld+AF8EcDGAJoCPW2vvN8aEAXwTwAoADQB3Wmt/+0X8bZpgeARZe1GMq1ScKPLNvS/C10QMukKfA9/43OYczMfcBtupBCMqiMqqeq+m3roNPIt7eWdEHUbKqmK6ekYMpXKbyj5fUxW0xTGOge5rcIw4APgFRGrY8Ar1ZIN3sZRkYcrDg6dz1KwPAIcs78iNNzgcpSgqjKqi+VKkKdU8iw1xHnRNCzK2CoQjAsoEALkq/y0nqry7kRdcrcczfMK12gcA4JwAhyUdqnPIjVKMU6b4NOvEDBUAiAh5M3XXDxT43lMTJG81MwkAhqo8mZz08MKRmgfjFVLpy8T0awCYrPGiydPN+6gv6PDfudFwWf26eC8frWk4ZEOcXTNiX2/ZQrQFnGCs7rPb/uYjL/v3Otf/zwX7m5W94hApY0wfZseYr7XWlowx38AsY30LgLdjNpmYax8CAGvt+caYLgA/NsY8N6b5z621951MQn5mjLnBWvvj+V6b4icolY5glFdPDqV/SH1ekdAAgBFSiKfCXorknIJPqaCjS8xHmCxpPLeaHv5qMdWGV0PLMnEt2ziZ5Yo0agjfBodL0R4yekAds0frWo5ZdT8UvOFUJX3KthQ4bC3vcKEIn+V7T8bhSfiGhu6COiISDjs8ydwHHnRdnfg16lOJKwB0+Dj8bNrwLuilSa4klhEcJpUszohBegBwQiifnRvn667Dy7smakZGe51/JwBMe/g6GC6ILkXwnfJ7mQ2K+RAAUBe8Mr8QEGg3vPhxEBwitSfLu7JBF07imcRBaNmZbLYFkZpjp4uD4QUQMsbUMDsN8Li1dg8AmP9arVsL4GfA7Ah0Y0wGwMXW2q0A7jv5v1eNMU+BEFterIX9HM7U0eRDhZQ1kq+nvh6hlw4A4w6vslZVu9jhCc9UlZM+51slBPRgNyU/OV7mLfGa4GcAmttRqfCK3pWiyrw8xJO+81M84XPrAz4wxqt9cR9/Df0O/5vpKodsuUm/jqe4fPLOPE9comLKt4K7jRY4/6DW0N2mrgSval7qbObX4+PV2cdrXPt+sqormov9vJCkyPwHSryqqzoqm5K8IjYEjQM/0nya+lTH6VwxoNUnjo2QSxdUHb0bHd6xVIkS6hyb/3BVrGU/LxoBwCYPJxWPl3jycbTBA/OpJi/E7PFq1bz1hvNQ1sc47yNb5/vEPsOLNG6DYPuFetcG0YmoNpWyF+d/XZDkipmHK1paumUta9krb694gmGtHTHG/DmAowBKAH5qrf2p+MgOAG81xvw7ZoeEXHTy/z4/xtgYk8QsxOrzL+XalOTjtgAPdoNeHtArLkAImkzaJwZPDQmsqgraq3WOA3ez9SlOqg43ePXtySKvPPVHN1HfjLivABDxcIhZPcTlXZ8s86DjkRkOS1tbeTv1DVgeeAPAYTGxfN8Evz8K4tEmcM6Lg1x5CADa6ryrVInqCcfM8k1+71aHOTTm2jh/RwAgLMmkPGQdLHGiv08oxm3y8oF4gB5CNwwOl3xTlAdIHR18K85UeUCmeAIAsNhzHfV9Z+oz1Le54xrqawvyax0v6iGFkoMhEsIdlRH+QZF8bPDxmR2XiAGgAKC0B9IVft9rDR4kbzCbqe++7N/L61kV40WBguB4JT28o5KyvEizI8O7cQAwaOcH01XQ1/MVv6fJ9/vjLmeFOvVUAWMiq3mAzNTgWUBzOxba8DoFnV5o17og7DTL1C4kOx0QqRSAtwJYBiAD4JvGmNustWw3+wpmeRlPADgC4FHgP4CsZnb1fw3AX1lrX7A8ZIz5MIBPAdAnirCuMJeDHGhytRpVsht1xKEJIN3kWPixtB6kdipsR5rLuypTG9RIgcNfUoI4DmhZ1FNhu9L/yn0v4XvVQEW/gM4MOjwhfmBcE/bXpm6lvlVCsWe/4cnrZh8PShWB99GsHvCoMP9KuapQGpLfy2x+QC9325riBYWBGZ4QlwznSqgZPADQ7ud/87LEr1LfSI2HZLsbXERCvSOATphLQiLaJ4oxXuErCi7JaFEn4dkaT5YSPh5Appr8mEkbXqi6JvYh6gOAIcGN2p35d+pzHH5/+hNcvETtSwCQE7AjBZlV0NaHPfxzXg9/XpWqlvhVNt8kQpkaPHumWSuJaNl87XRApK4DMGitPQEAxpjvALgcs1MD/4vZ2dX9vIyKMeZR4OcAwl8CcMBaS2VYrLVfOvnvTpK8//uWrgxRXyLAoSF5cIWpsSKHaQCaE6Imk1ZqnAyo5jy4cT4UnllNslafU7KN2bIe/PdKm7o/18Z/XX62P8RhHMMlHgSNGx58q/kZbrY7zQMSnkIA/cnN1LdDzCoYnHqQ+hTR381WB3nF8zXtHDHpF0v98IyGl+0VXJORAlcTy1Q5DGqVj8sKL/JyPPu+hp5xcNckf85rE1wFKGj5em0IaWC3oPSSIMffdwR5KyJb5du2nFYuYJbjVT1A7BkxrXtzk880UUTukuG8oIPgsE4AGMr8hPoUjG6ln1f+91XHqW/7DF87s8Y7TmqvVOeB4k05Rnf8lam5UfUGT6ZPB4/rbDcltgMADQFBPO1mAbwEkZBXm52OBOMogEtPqkCVAFyL2e7EC9rJf2estQVjzPUA6tba3Sd9/wtAAi7TBF8OU9CiKT8npvWCdzfOD2uVpO4Qr5JNVfghVm7wA7c9IIjsQjUFAIp1fsgrGEu6wq9nV50nEauEDC0ABBP8etVviQrlHQGtxoRwbmlupT4A2CuCoOHs/fKzzBTvZXnqTfKzh9NcdaVDEMRHsg9TXyzMMdkXhXkw63fZhtRcjkCTr+ecgBZ9L8fhH2uDegBmHbyj0B7iMJYVYhbIoMO5UfdMcaGIgSRPsADgjXFOyC42eaLQEFwSNVU76tG8hkcqHEKWLvK9IOHhUJ52wY/r9vAOT8SrOQY94PAqJVOr1uvB/L3U1x7hPAoAeGPik9SX9IlOZ4UXnI6J3qvqGAA6iag3uG999Beprwsc9nlcFFtmRCEPAIpN7p8U0Gg19K0lYXuqTHORFrbZFkRqjp0ODsYWY8y3ADyFWajT0wC+ZIz5BQD/B0AngB8aY7Zba98AoAvAXcaYJoARAO8FAGNMP4DfA7AXwFMnyeFfsNZqIOs8TU3oPZ7h1dnuFA+60k1dgd5X5hWtiSrfFOdL1lZdCAAoVvkmrWYgdHp40DXZ5EOy3CYqK/N6OUyhIUjFC60d7BOTf6fElOYlXq1od2PyU9T3NDhhUsEt1nu5oEGHUHv6Qe5L1AcA7WJgnsI6PzDB30tlzzQ4RwcAAj4NS2I2GODXep6Y5bChjT+rybquws+IZGhpkFcKj5Z5QaWrwRP/lMM7HwAw0+TXUzDzG1w2Zvge8kyR75MXBPmwRQCoiU5NyvI1UHL4vn5dhPPY1GA/QHfOyjXeqZlvAeNU2ZOZL1OfUhxcE+DJtM/hnCoAqIjOfULALKdmuDhFy06NtbpGrx5rDdp7kaaIYBeaq6jPIyogaatfpD11Jr8vnQAAIABJREFUrjqz0scVToqGBwd7M9+mvvOT75bXUwUPZhQUbEWTd2qOisqt25wQRfpTWOdnS1zJuFzhkBO1BgaMnpytBhyWmhzTvtjwqnfR8EQpYzW/px+cZ5EWFeqqgGV1Wj5XoCFgIxXBMQA0XOeo5TBD9UzC4jsfzv2tvJ6wSKYV72NRks8t8Rl+PWN5rgTlNk1YJYRtoqBQrM4v2PeIvwcAA4GLqG+6yTvBKTGEbqzGCyqqKOK4KF51C96dmrNTrvOOQaHCq/AXRLiMMQAMiKnkKjfpEp1wJRKQr+lKcl3EDkcb/L4PW9418YkE1QdOVs/U9QT5fJlDwV4KRPOVtvlCz15dtoDnYKxaZLf95S+/7N/rvOVPFuxvVna6ZGrPOOtzuN582vKAftLwQK/bavUcn6j6KvnJkIe3mRX5bL4kbkBj81USMZzn0KJoUEsDDwsMsKo8qWt9YzvfHJbHxMAzlzdppsY7YKUG/97BGYHnrvPnfNTqit5The9Q33wPXF5HBtpjG6hPDUwEgHbLA6t3xnnXpCKCpy0lHpAsSvDpxgCw2nIIWTbIE9spcFLssICenZ+4mfqWiMnQADAInjA/K4jBa5P8bypbYbRS+LlJDuUZLfJu754yT3gui3Go4DMlHliqQgwAhC2HCHUJvZCiR8jCRvi8ioDVydmyGE8UQgKiuifDIVsHmvx8Khle9QeA1U0uTHBhmO/dq+tcDv7eCi+qHcvqeTlng509SUTLXg3WSjDmmKowVkT1frjJVTF6PbwKdl5ET0pdVOESk3eXeCdCDSaLhTnOtz2gp/Be6eMJdFHwPgJilsPlKQ4N6QhpTkhD7LWlGOcgKIjk9iIPSL564rvU515J5tU3v48nCqqjohKlfI3/DkDDfOabYCiC71ujXKZ2Pf/5AICghz+wvSIGenyGd42s4YsnaDTJ8KjhCXN3kxcNrgmJKcUdXK75/uoj/GI0qkbCDK8TmP4weDCbbfK98PvZP5XXE/H+LvWV6vMTgzhc5N2EODhB/ryg5ovsK/MqfFEUaipilsNyeyH1rY1okcOC4MD9NMfVl/bnuaJcMsyV+vyG71kAcNfM31DfRbiN+jrA36/LPHyQ5+EUT2hy0Ep0R3M8gVfCJy175U2pTgILD8b8c2bRGrQ3x1oQqZ/z8YW9IcnlFZUs4V4jVGXKQ9QHAI0mP6h6I3pSM7OIINGVBeQGAA6nue55V5wnHwURfM8U9VCzlnFT7fI+IT8JAJc4l1Jfws/fAzVb4qoO/h7kRXD0UFYrIdUNP1AKhsNRDk3fQX0vhaCpiJ+dCZ4orAH3Xd3B38uASLDcLCbO6iGB0Px+ju9bSkVqLL/jxVzWC9p8E1slVVwSAzc3x7hMLwD0BTlcZ7UQmFA5X5pv6Xgko4NkNWPleI5LrVoxI0NVxP0+nYCdSdAixS+sNQT0V8jttuzUmLuaZXXBwoUuXrnIbvss51nN15y3/+mC/c3KWh2MOaYy430l3p5tC/LKvyKA+30d8nqCQl4wKvgJR6ocdrTa92bqaxdKJACwIsG7H2oibLquwDMtm6+p4GA4c7/87DC4XxHk6wJf/oAevH5KTMlPqiRCkUmLaqgbgPYo5xQVqzxI3OXhCkKPDPJAxopquU88Kzdz67rNx9z2NPU3lfBAb5jDh/YV+MJ7uPzX1LejyeE4ALA1x5PpeIVDwbxCTtUjOkNuE+RVMUaJQdze9RvU1xnkwZyLqKDsqBzL87M0XedZVtTh96cp3ueMSBIAYFvle9RXLHP5aGVKYOLVNAdDmeJ4NZta7ptZCwb26rFWgjHHFERqs4+3boerHMt7aTuHjYy6TNVudzgHI+7jj+6iANeaVzZU0h2MSYcHT2ronQpYdZVMbzTzDZCuSXyC+paG+T0PenhtsuHSCSzUuN/jcpAzSwup4sMC7w8Az2a+Rn0qiTgVppIEQA/RUkRd9T5fAP5enteloSEK7z5a5GtWSbTmSlyitSkgSW6Hca3ONeNV5X9JkHe4yuD7xFpwrhoArIzzgCQk3q+nprm4QMbh6/VqIdPb5tU8pUEvhxk+nf4H+dn5mBp8CADr+/lwTEHPkB2V6QpfP0cLWnwhLgpSqQDnxyX8/P0aL/HA/CnLuzSj+aeoDzg1MKizJYlQNt8k4lVrtiVTO9daCcYcy1d5y3dUqOcsCXBM6UGhYtLv1eDzbINXekpVHlxOzFPucV9ZT3++IMS7H4rbsdrP4TpKXand6jkYiz28i7O3OcR9lsM/KkWu2hQQ1cf9hpPuAd1RWNnG1WNUxXOt4fK/V4Y1n6anyQcD3pP9S/nZl9vWh7V6Tl+UvyfFJn8PjhpemTwOHjzuz2lNfSXO0OvhyfT/6DmXX0+C+3IijnHLTWdqPIBUnIc1Sb7uVDL9xJRWxvt+jisIeURFOG8FHEec58scLiCwKCyicgDtQo734tDvU9+OIi/EdBnOfVqb1CTvvFgH/zjFlfEaAiKlpmq7mepWOWJIbG+Yw3uXC+L4m2JcrTHv58UEAHiwxrv67eDdqIzYJwYFZLhlmnfoET5VFGnZmWUtDsaLtFs6OTmxXVRr1HC6yYoA5AK4v/wt6qsLPfnlYR7Qpy2vlI4Lcjgw/2rF+SnOX2lr8uDxySofBgcAXtGezYjhSafDVDU97iyivqLlErbH0hy2pyBAAHBTlFdDF0f5ei6IIOdAnifh5yU4nl0UrgEAh0WE/dMSV0Kq1njVMhrk9zzl48RXQENZcgU+q+Di5Ieob7WfK+tc3c3TCI/R+/dQgT9LR9z3fVmeuH39xJ9Qn9sk76WGizoELA9Kq0LmWBH2j4q59G5CCDcn3kl965L85kW9/HqmqvxZ7krrbtRkhe+/FSsGIwo1qEWGF2lGrU60Ryy/t0GHF92OZPi+dXX8Y9QXEt3KI0Z3bBuCxzVS4t2Pco0Hu3oieYs4fupsAcvUruix2z7D4535mnPzny/Y36ys1cGYYwqqsb2xn39QzMsbaPBqp8eFzBT1c6m/sexj1Le7woMuhRtd4TL9OWi5IstgmSvdvJSBefO1uBjOdqFXT2pmdsLhVVRHSOYCwPE6n9ewxvCZJmra8LY2/jzc7Dvp/0t93hxPBkoVfZAzuyfLt5pUlFfvASDi5TC6xUFOnB7z8PkIvyCkTd3sQJMnILtjvDI32OASpb01via/d4wXIh6qcLlhQJNUVTKQr+rgm9kycJUkAIiKWQY1McE3ZTh0UXFt6uD7QI+fw8AA4FiRF3G+kb2T+vIl3jlLiX3JbaibSlCnHd4Jfn8nfyYDIZXUaOhiRkhvHxZo291NTrhW84vKopurZqgA+rw8k+xMmiyuyNpe0eltNDR0vLmQVaRa9nPWSjDmWCLEuQvrHQ4BivtVxZdv4LvqvJsAAG8Oc2nPKS/nhGxvchhCHbwK1t7UcycmxSFWEgFJMMAPonqDZ2du+PKQn7fovaJFf0GcwxTWJPgmnfDygD5b18ni/aM80VwW56/hiii/nqssJxs/7dJl/rtpPq3aunBfmHnE3JbeOA/mrvbpQG9dit/bpRF+2NSaHEL22CQ/qONa6wCv7eLV2XyDJ+mPTwi5XQHP3CBmCrzN/y7qA4ATAf6+H3J4J8YRU8f/9/L3UN+ysO7KTlb5Wv/yEA8uN3Xye35OjHe4MjV+79zekWfy/B/cIrobkwFRLRdDNW/svklez2s6+ZpdHOIk73HBs/jiEL+eYYd34wDgci+XXb68i5+Jq+O8gLE7w/eQ+wt8vXYbDq0CAK+An6UrQ9SnBmeeKltow/SUuqYSxlHX+qqGQZ1lqCBlLYjUHFOEYzVl1RFI6K2ZL1Lf2hSHqQDAhOiaqE6Emv6sOA+vDfBhaLN/k/vURNg2oVSiTHzl7PUI31SFV0P31Pg9GLeHqC/m8PWxb5rD2QDg6gQnm/pEnp8zHNP+li5+PR0BffN2ZvjD3JXjgV6bhwcHx5ocUvHaBK+GLhNJFAA0LL/W+8Z4QNsZ5O+Iqk4fdvgaAICNHt5xWS50YfvC/MBNV/lv/JMjnFD8rrb3UR+g+RId4r1cHeOfU4Mhn0nrZ/lYme9pezO8G9OT4MHseZb7Lm5XQhnUBQDYMc33EAUTU3DISx3OI7i6R19QRECvHhQNp7AQJVBE7r2GQ6AAzZGbMEPU97oA7zoOROYHN/5ZTitB+QT8Tp2J42VerCsLxbhGQ3ORToUttMTk1NkChkgt77Hb/oQXYOZrzrv+csH+ZmWtBGOOKVWV8wIcwjDtcJWb6RqHKCS8A9QHADVBLA86vArvExNhB4t84JAa6gZoLfF0nsNRepOcEzJV5AGHUg9qmbalKQ0Du9x3EfUtifIgOeHj+8W0KF7/c5pLtKoAEQAK4EHQ4zkO9XIc3nE6VRhpBbMMC9npq3y883F5l0tLRZgKkr2iYvDjEicNV4T6Xd1FLlTBG6RamOBbxQXfKOXncLaCSIgB4FJnM/UtF22u8SK/535BOPIYTUb6RubfqO9i/1uo77DD9+ajaS3s8UpbMsKT97ifk7HrgsgOaLn402GnQt711WLq3gBAs1lYsMF2K8H4eWtBpOaYaofuMTxA2ui7gfo6HV7l2VXRm/tCG0LX4edt6PY2Dkc5KOA4Z5ItFxyVFZZjqwGg288lMQt1HpDckeXTcpclXsevp8khfQAwVuGB4HeyX6c+lYSGgxwGprTmL0lxDgoArAjzrski76epT/FXflT8IfVdG+DvMwA80eDKO2MlzrVRSfgPPRwumWv+MvUt8vN7AwCjVV6kuD/7Oepb0/YO6vN7OI9izwznJgDAFZFfor5wmAftO8EFKFRweWzmUeo734WHkxeDTg+L/FTNa9hd512Bw2ktaqFmXUwFebJUbvKLvTD1fupzk+LdnOBKdGHRYX+sztWXskVekGuI5HSJ6IoAAESR63QkH2d7EqHszL43LZnaudZKMF6kqWD/cT+v6KmqXFuMY+gBoD+5mfomCvygalp+MKoZBwprCQCHs1z9Y6GpZixL3Uh9uToPkm9NcinePhHLrYxq4lnMy59Jpclb21cVPk59dx3nScKzIiADgAExr+Dq0LuprxIUszdUpVQkGFNNDSdI1Hm37g29fM0qNZ/+LK/4ClG42espcCjhIZHYNQI8kQwaHlz3hXhyGvXpqndnKEZ9Hd7fob5yk1+rCrxvin9EXo/qxkyKuGJxlSehbQH+/uz1cK7aPjvE/yCAhIfzN4JieM2OOg+Sjxe5YpGq3gPALclfpL7+CF8HmepS6ns0zRMT1bEGgPPifG0l/fx6nGnONbnXzwVK+gOcrJ6HlmafLh2U/vnY2QNJalnL5metBONFmiKwdkX4RhyO8oNxjV0h/6aSHjw3wiEuHT7eYlSwCDVIDgA84rMNkbULnrvkdahDCgA6xJyssAgSp6u807h9SlRDhTLK7+3/LHcCWJ16O/U1hXrOKkFUviDFK8mJGS0XqgZhHcjzgL8CHlwqPfmVYrjhzYv5uwUAvUFO4t2f5wHrQ+N8/eyp8mndK4RqFQB0h/i2uSHJYVknyvxF+EmJD6pM1K6nvpzmVEtT4gIeWWzg+8uejL4gJcV6oMKD3TVBDi+7M805M6+Pc/iUyS+lPkBzNStNQWA1fA9pirktb4/zBAIAlsf4eh4RSoZfndL8MGYdAU2c3pfj+8SYw0khV4Q5bO2jydup79FpTkh/JPsN6gO0qqBjeOe1XOH8DEVwbtlZbGJvONuslWDMMTUYpjPKidNBwxVOehpcQalmeGAJAMcN36QL4JvtQTFA63QMB1LQogB4W+BEQVedfGpYj5h+vM7wdnnZCLx/huP93UwlESubPNE85AxSX3mac3hWRXXQvl7MeLysQzyTKg+ge4MCe+7wNfnQCZ1I/mCYR3qPNzjUqSQ07BVkK5fSg/9OTPOJ3Cooe12Qdyz/dBnnzATFvXOzmI8/k2NFnmQqhaUDBZ5pH3L04La2Kk9Co5ZXxB+o8I5cQsipDonpdCtdWN49YqC7T8wfuay5jn+wm/uOigIGAHx7jCuNTRqeMM+30j5c4LLKADAM7led+0NFfiZ2hnkXx+O4sPKFqd+peEGhIIc4nw6FqZYtcLNoQaTmWCvBmGNqumTK4cFcyXLVnbCAPgyJScMAsCfNqzKKOF1rinLWSzBFvtoY59hqr+WBzLi4BzGP1mE/UeaBntL/P5Ti0XXYcN9FyQ9QX0moPQHAOodX7bxC5aVW4wnPvdn/j/qebmh4wxdHOJzp0uSvUF/OcIidmgUyLQY8rofGTz9pH6K+88QMEY+Qjw4G+NY3aLWmfqnGk/uCj0M1vjz2F9SXr/EOz2iVJ8tTjiYqhyxPFg/UueBDRgwMDAnZ6SUhLTm8RCQDGzr5/pKv8TVyd26I+r4z9Rnqc9Iig3CxtQnekUyI4aHP1jnMdIlfvwdpw6GdY/mnqa8jwuGQi8AhfRU/33sAIGB5C3mP4fzCsJAXPz7DkxYF73WzhcZnbFnLzgZrJRhzTA14qYhpemHwA6UuCH+Lmrw6AgBjAgO70IhpSuovZTmWeabBK12TOY5XBjTP4pog16lX0LN7cl+ivkujPIla7PBJzACwq8E7EesNh3FscHjFszvBW/sNl+7YE4ard6l5JxHLByQNgD/nohigNQw91O0icGlPlaTf2qFgEfzvra+fI6+nEOPfq2A1/1TjyYeCLi4N8a6RU9Ldn74gTzDWBN5GfU/7hqgvC/7OKngQAMS9Yhpznq/ZuI9X4dc4vCsSSnJOSJfh3B4AqImqd9LLC0dJP7/W9iLn/mRF1xWA1OX2x/hz/lgP774PhHmHx7gMbpup8wR+pMShTltOcBjd9+t/Sn0eD0cKeIV8NgA0BW/oVT2ToWWvvLU6GM9bK8GYY2uSXFXkPIfDWJTWfFm0XwMuLd9VojrbkeTB9SQ44TotplEfmP6uvp42fn+UksuTmS9TnyISqqmlAJAUAe1B8Cr0TSletfuFAT6vIujw53wgr5nB51gOnUlX+fc+XOC/Y5WHd9VCXj17JJXg5PGf5f6e+hzRkRsJ8UCvKnThD4tqOQAcEupUlwY4bv3xEzxg6wryarlPZR/Q58dXRv8X9anZNYe9PAGrGR4EXpXSfJGkULgVdCP0VRZR38HSfdR3g1CJAgDBjcZ3crxj+9bozdR3VY8aBMbvz34XXYrhAg9K/WKNjBT555rirJgQcucA8NYUT2yXRvheoOSjv3eM71t1l0ApLhKptgC/Pz0h/h7cmPwU9R1wOGTW7exSU+vVlG/FvTwdsy5a1rIzyVpzMObY+hSvulwY4Bt4XJCR8zUxsEpwJQDgRJW3qIc8nNjot0JGMs0lSDek3ievx2P5Qb7Yw0mYaokpKdFyQ98fFQem6/xUfazCD6PFQQ5T6BODpdIiqQOAJzJ/J/3zMaUytsLyqiUArIrwamBGDCncYTksLdXkwZyqFvdHtO65EkraNsOHXe1vPEJ9FzjXUp9XjnAEttv7qW+D2Ux9G9t4sKJ24f1ZvpYDjr7WS7v4O9vm4+9Xm58XRrI1/s7uzuma1c4039PuLnyV+hSm3+vlXbUbY7yDUWzyxA0Akl6+LndaDrmZrPFAWKnUbUjp/U6lvYdEgePRKQ4terLKpXGLFc7rAICgnxd4ukO881oS08zfFOHvpUrqDuQ1LHin5XLFJ/JcWjrg4+darc5JM62uyKm0BTxob2mX3fYHt7zs3+t84AsL9jcra3Uw5phq/debvDo7KdRh1PAk1fkAACM+m2hyHOuRJsfjJiIc/jEJzQlZajmW97EGx8l3CAhQpsZxxcfSHK/8UkzJ8aYDvHIbBw9kvIJ/AACxMO+adAZ4ZTJd49Cq4cz93AfuA4Aphw8DujLCn9c7/JyoHBND+Ip1vpYP5bQaS6+Y7ntRjEPTuvK8y/dQjQdWal6Fm93n4cFlT+D/ob6ruvn9uXERvz9+R0PhcjVeLd43wzH0/zjE2xtLQjxRCnpcigIiTFaqPH4f3++6o+upT8GOYo6QoQOweRFfd7eFlAIg942V1aBKnSy2+dUkePG9Dg/oVRJxfuJWeT2KgzHYeJL6aqLy/7MSh8VGwYP9qqPhZePpLdSnCODFBdalWGjSuOos9Xg4tPOl8GkWutkWROp5ayUYc6wJoTQhaoxK8lO12R9r3C+vp1jl1dk1AT5kzS8mGCteQ6awR17PsHmA+gJ+Xr2eavJqXzTIq2CqrQ3o3xkRvBjFfalYXtV8JPsF6nObPqqmDcfEwZn38DUQFd/pZodLPCFUCmWFGU4qVqpN63xcanXMGaI+AFiW5gnYQIhjr/2iun9TmPMPGiGtIvVQjVdDiw1+fx6oPU59/7KTJ9PXCIlfj9GJ7eoYDwIrDb6nKTWovSVeLV7kaP5KtMnf2ZtiH6S+6Qb/m2OC/HzM4WIPboPtekK/T31HAxx7prgkJ6rzD1jzht+DqCDzZwX06prYR6mv6RIo7TFbqa/hMlmb2ZEMJ4cHRMekVBmW36tnPC0sWdGFlkQoU+feqzmJaNmLs1aCMceOpvnmdm+Ct0M7ajzQ67O8In6llycJAFAV8pQTQrkqX+OkWaU+5UYcVxXGRoNDH9SGWazygGy6oiUvldpRUFTXDjt8SOFQms8juCrBK9ABaD5Nm5hNoqBgO7L/RH2bBIE1ZHXCo9SH9uf5PagKorLCK0f9HLZ3hY8P0AKAmTpfd/dWeNB+a/Jy6lMKpYMzOrCqVPhe4DH8vmerOghiNuXwxK3HanGBgoBoTpR5ELjYcrnQqOHvVqdY5wCwJMYTIr/gOFWbvHt4MMeTluma+I0JPmMGAE6UebGhaecnmTrk8GLL8qaeO7FCwHUeqHFFsD8a4CIJ3YH5QeEAYHeOw5memubJUE103e61XBlPJREDKX4tAFBu8PPyRE4PJX2lbaElEcpaPJQXsLOMdqCslWDMMTUH40JzBfXVxGCVgMM36Z1Wk1tvbuNwlMkyr6BtNJwQuXOGV6dtUsMtUobzUHKWV8l8hlfXeps8OQuHdLASElWp89v5Z2+LcChjdNlG6qs0Obxjpq7hDfuy/LNLeHyEj6z6NPUVGnxtPZXWr/ZYkXMwzvd+mPruqfyU+tRBXRbT5Q9VNLzhjV08sLohyLtcx4XC0o4pHjwuiel79+nkDdQ3XeF/89m0GOLo5+IL6/xcFjYsJI4BYLTIf+cew6FgV4qpySkxOXt1XB+udx/n96AzyIP2J0s8uNwU5h3JRWG+9xzM6Sp70MPfL1XcL4iE2COI/sc8GqJ6fTtP+i5x+MTyB8fnB60as7x7CgCO4Cr1ObyDvNfy37k2xWFZ8oyBi+SwzpX493rbqM+Kzkeret+ylrVI3i+LKSWkTV4ejOiQFJgCl80dM7z132N50O4RMykCEJIz0OTXfc4O6lsPPnVcwVjiPv2cRsu8a9IX4lVWFRzcU+aVQEWcXhriQ8IAYKrCA72nLK/Cq67SJ5f8v9S3xiXQKzV4YPr0tJhSLG7ekQonun94GU9oOgShGAD25Xmy+PgE/2xczMGYKPHn8WD1Dnk9q328C9jv4YHVW/r59fhE9T5b4+/IHpG4AkCmyp/lmgT/3s4Av550lf/NB8f17IRRwztgq7282/tUg8M3R0t87/nVHs412pDU667Y4PfnqWnuU7BY1Xm+tVd3o1L++Q1N/IODf0Z9Nwo1uammnvz3WPZvqC8SWkp9bUHOUdloePFHnQduginTdb4uRzx8Rs9Ela+7XIl/zgqoLaC7+mdDV0ApRFoXeeQFTfJe0mW3/u47Xvbv9fzK3y7Y36ys1cGYYwrTrjgPUcuhIT6xkajDBgCG7Hbq6zK8vb+nei/1naqBQ2rDnPTzTs3aEJ9gfKTGyeoAkPTyjspEmVeegk1e7Uo6vBpaF63rssCzA/pZ94NXJs9LXkJ9D2d51ft7Oa0A80sdvFNzSQff/LsCghBZ50pRJyp8fRzK621oyySHWzTEM9nV4Nj8vVmuJKbmvQDAkx5+bw9H+Hv5nR2c/3Rb1+9R33WL+Nq6vkcHMgrm8s0jPBBWCksK7tdwCQ66LO9GbalzYrBXwLIcIff9YJpX4b88yRN7APhAB1d8Wsy3fAQ9vFDT2+BJxIPj+lnONLh/e5Pz49QQ1EnLiwJ1l1k6ykoVDtOt+Xki+aMCV9tTykx+n5ZrrtV5YqsgSQE/HwzZaLjoHLeMmnsS0bJXg7USjDkW9vPNXw1SWxTmB5yqqwRLPAgGgHidw7IKTV6RWe3nFdYTgaXUpzgobrYkyUm8i5q8YjUjBrBFPbqil6nzClK3gHMVHV4hOlTkB/UN0duoT0koAsAih2PIn2xyhZOQ6Jo8mf8X6ntT4mPyeu6f4IfjsIcrV6lqcWeQk7HVMMpEU78HqlL69vbfpr6YXUp9pQTHbKsJ8QDQEEO7Zso8qVFwi0KdB3OPTfJtujesu47Hi/wg9wqOl9pfRhqCS6JfA+xNf5P6LkpykvdFYU7wzXh4B/nrJ/6E+q5weUfKIr4+IXLQ0SL/4M46n2ujVAwBoN/y92uxw5W0QoKPNuaZHy8I0EH9a8JcqjMseErHYzxBLxm+Z1Ws7rbka/zeZvKck6fkkc8We2ndhrPQWipSz1srwZhjuTIPWA+GedUyVOXBrAocVGAJAJ3OMurbNfN96lOVHgnnEqRhAKgYHnQsAa/0hARU5aHaLuqbLOhA76bYL1OfUPjFoSaHnr02xDHAYYHJfriquy2Lm/zgvNTLeQR3zvwD9b1FkM4VPAgA4mKmSbzKlYAOlvi6KwvVs5gYlgc/D44AYHGKJ6/KukXi35bnVdSkUDYDgAo4Z6S3yfkSxx2efKh5MAkxZ+dYQR9m/zL1Feq7KvQu6lsR4AnhTJ3DAdUQNQB4W4rD+kaLPOFZGef3YNsJfg/e18OVoNxM1QzUAMMJwf05XuczF5QqHgCsDPJn0hXiHcJz4/y+9oWdTPgmAAAgAElEQVT4vuSSKyJT4/vEzixfB8MFfiaWS/w3PpD9Z+rriPOOLKCHXIZDS6ivUBqS33s2WCuJ+G+Yta0EY46dhRwMxxqy2SxNvpF+bkWTb6ZpUYVfZDgkwI1joNSFfIK7EBLjckviO4eruuWrZFqXpfjMgcH0j6gvGOAB2ZqwDiyVrHCv5UFi2OH3/eoefhCtiPKyZcKn4Q0+VS2u8785WODVPgU7mtBQeGzP8CRrl+W8j4aAD13m5WtgxvILmnI0mfSGBJ8hcoEQJugO6GfCrGF1aHWsxKPL/TkhBCAUnYIeMfNGJBjjJY0994vvDYvtZ0wE++1Bvu6u6tT3fF0bF5nwiHdkssjJ2rkqf392ZDkc8tm026Rqfu96Bad4b1bwaapCic/lLL6xn//ONTGe9Kq9Z3uGY71GipoluD7Jn3XEy9/LLVO8o7JlineX9xmenCneIQA0wO/7/irf704VpLhlL8UWMAdjcafd+lu/+LJ/r+djX1ywv1lZK8GYYwqLqSRKu7288vTdzF9R3x+t+CT1AUBCDC57WgwKTQX4wbhIHIxuK0EMeMa0gAzszHIM/XkJHjisjOkrinq4PyOIsUcFh26qImahiMtx62AsERKUM4YH+08JmdpPLP4d6guIwBIA6uK9/+zgH1Pf7T2cK6CqqEoWdmNScx5qIuD/i/18ba0IcU6IkmidgFaAWeFTcEneiVDvXq7Gf2O6wp/VDwqaRzCS413Sj/dzgu+yKP+bPhF3PpPW606lQ0fzPAndaTh/5WLDeUrX9vIsKubVydlYmT/LfxvnSkiXhHhH+ydFLiKx2c8hsYCW+FUcsD0Zfl9/VuRdgYSo7APAJqFcFfPy+76+jf+ODsHxmhAFlRE9yBv/nrmT+hZ5eVf/YPE+6lOTvN2sxd94KbbAE4xPvf1l/17Pr31pwf5mZWchRMrSROL81Hvpp9TQIVUl/GgfTyIemNC74rQIdMZxiPqGR++nPjURt1bjsxEAoC3GoSyFKicc1+q8w5MF5zX83QSfxwAAG/2chJkWswNiTR54JoWkbleQdxPcZjkMV0RlrsIPsb7EldS3X8hsLotqid9N7Txb/PbFnNegApnOAK+iKriJ6tIAwNNCseeh7J9S3/IQT4Z6hATydEnDy56p82ndU5l+6vM7vFq8Osafx8YU990sJqsDgAXn8Bwp8GDusUn+wLpFIhkRRRFAFylUl2uN5XvPiOFdke3TPBm8slMnQxcm+XpeGeUQuwMz/Ht/Ocz5cXnR4QKAraK6r4j3nUHecfMIuXPXOU1Cnv14lQffdx7lAw7fEuMQVUekp27dnw4fh4Ltmv5X+dn5mJoJ1LKWnS12FiYY3AYs5xFc18fbukmf2GiFxP90mX8nAAwJLkW3IE57U/yw2ejwQHi0rqsq4w6v2l3t5/AypX4yYji3JejlxGgASIi5Jd0Ox/w/bh+jvir4fT1W4QHQZSHOlwGAlaLLdX+WSw5v8vOp0o/Uf0h9+ZyGlz06w7smCtb3lOVTrDeaS6mvJrqDy6IainFHnhPvFW9oa4XDGyrgyb1HqBIBgE8MMVwR5QmqEho7KmRGlQ8u0tJB0eW7b5zfg24/35tKftE5FN0WQCeoayI88T9Y4AHrcXCVuuEilxmdOc6TdwBYEuX7S0ScnDum+X5XF4WqoIC2AsC5cb62VHyt8pb3d95OfWLcCQCgLr53TKzZa5x3Ut+E2GP3ii5WAxqaN57lU8fnayqJcJOaVaqLjuDieBz+vtdFIa/pIpvbspfPbIuD8by1Eow5piQW1QCtiTLfLLyCbawgJQAwkFlKfUvDfHMLeTmv4a4Cx7EalwOur8GD6NEmT04i4MFKGJwweoWXS9gCQEgQYxdHuc+Z4aRqI57XjwpcAWeRSE4BwCsokx/s5URUVWMNFjnnoWq1xKRSRZtu8MBzvdlEfbvB1VjekeKJbafOs3G7fzP1/Wyay2FeKKqWYS9f64Wahs58ffKz1NfT5MpEHT4xATvEt+LO4PxkgwENSbplMb+ep9P8/mxPzw/yCGj4phoa2Bnkc1T68xxatEeoAJ2o6gGPqRoP9HwuqnHMEn7+nPsiev+dETGimr3RJ5TG7svzJPz1cc59AoCYj9+DvI/vv3fMcNjnyjDvmsTAVasmG7xIAwBeLyeP12qCA2b477BCTc7NGEwb0PCpxvyVg1v2SphFi+Q9x1oJxhwbcoaoz5laSn1rk7zS1RPmi22/hp5jUzs/4NQa/kGGHxrdTY4Pdlx0Q5YE+CHvFQeuUqFob/AgpzukoSqKpKokJpfH+QefmeYVtHenuPTilPqDALpEANkV4vdOVXxzNX5I+YyubCtYn6nyIHFPkx/k14Q28OsR62N/VgfJoyX+oryxg5P5L04JnoUqGFT0ulse59yXwzlOJh2t8sD8pyUOB/x0Px/clPTpdTdW4evuCT4aAF4x+O+CpIIRystBf4hf728c5lDB/7OS4/3XiYGBGyu82PLAmA4QV8jZmfz+HGjyrmymypXE3uDhhQ9AJ2eXd/H3/XVdvLL9qSg/Y7xeLWFbr/P3ZCTHb97y2Aeo7/AMv6+7S2LBukzqDkV5gqHk2f2iix7w8e90U59ym7VzNpuS8waAWp3DsVu2sKyVYMwxj5DuvKidH6pJARkYEUWyuI4BsTHFgxU1uHSxGPY1VhID8VzgDbkqDw7youp7b5Vjbn8xxiFAJdWDh654XiRw62uTHB70wdU8wag2+CmWr+iH+egkh39c2ckx5FE/L1umyxyq82xWS15+7hifdXFDkiuyLBNa/JNlvgbunuKdhnODnBcEAJ84hz/LZUkesI3O8Hvw1wf4d35whd4W1yT4+slV+TpQzyQ1wt+DGUEAVzMyACAtRAvifr4X9IpGxKY2vqktjmni6zPTPCj7cCeH9W2b5nvBe5dz7lipxu9P2MMLJgDwuJC/DYiANi44XhHDO8/9Lh2MW5fywCqV4MnrsNh7/mYP55K4NPKwqY2fT+e38aTmHUv4+tmf4de6aJp3MHJVPTPpB3k+xFFZtcaTGuVr2fxNyeyfEdbqYDxvZ6GKlLEsr1qb4gSzc5yl1Kc07C9s54dGd0BXH3fn+OGolGUO5Plho6qPfRrdgLDAcwdFxVNJFiqkQVDIK7qZY4RUpAg6dud4V+DOKT4npQyNuQ2CBxZp8CA5BH7g/no/V6bqCPDDHwDuHOb34Nwkfyg7pvmzVF2Kq8Sg3WURramrnpeSlJ2q8s8NFcTnynpPLIjs/vIu/r4vDvFkcX+eJyZHRMy+J6fX3Zo4X3dXdPA10ubnvuOCGDwiChgAcCTP721IdNWq4tBW3cN3CCGk3pBed0+keUK4X9DVflLgeP/f7ueKV0pqFgBqTX5/MqKbuW2avweKdO9W4FHd1W9lOAzqL1a9j/qWR3hlP1fnv+NjB++lPgDY7LuG+g6LKd8Bwbd6psa7ji1521NpC1hFaqDTbvl1Xiyar3l/88sv+jcbY94D4LdO/mcewP+w1u446XsjgM9jtuf399baz7yI78tba6PGmKUA9gDYh1n0dgHA+621lAR3FiYYHusQcvA1sY/Sz50rDuplopUeEUF5qaEhSbsy/LMPuUwbZqYIql1NEQUC8Ahi2qjhWOeo5TeoZPih2mV1ZTvh4YHOUIMfGmUjpE0NVwHqDPJDvOxyGD9Q43KiMfDqW8PwQE89yyvCmnR+Pi8kY1GQ/03/PJO+n43za23XIlLYk+FRULsoJSty67Y8T+rUQEkACFqeiXeKhHBxhOOH1CC5tTEedCkJXwD4nMBh3tTL38tdaf6cF4WFXKhQpgKAJ6tc/c4D/n5lRBKeEEM+NwX5exAVHAIAeI1QWvOIAsa4gKVlBCrLukBUvzz5CPXdluI8lO3TfL87rGDBVieLfnBocEBMD99R5MM6Xxt+D/XNiCLOMPR5OJ7n3MOgn4taqLk/pQqHkCnJ+5a9VFvACUZ/h93y8VOQYHz6K/+dBONyAHustWljzA0A/tBa+xpjjAfAfgDXAxgGsA3Au6y1nDyJ/5Jg/MBau+7k//4RAJdba6lSxFkIkWpS/KNfELr6heqcTxw2QwW+SauqHAAcLvANNSgkL1NNHj3mDf/OnNHwhnHD8fermlwO86iHfy7f4G3/HWIGBAAE/Dyw6I5w+c5ey5WiCk1eZb4n/XX+ORfM7fXJ36C+mIcHVscbXKp4yuGwo315/Sy3FjmEocNyDKySTl7j53wIj+EH7v6sxsIfNTy43DXBJSZZIQEA2sVE+4iHBxwAsCvDK5dGBIl94DyC6xpc0chneAaW8usu6LXdPIl4eop/9kSVB1YPVHZR31rL9wEAWOXwlkJ7gL8HQc9S6hvM8yLFN7P8ne0PaGnpe2e0mhgzJfjgAz8PBs0R+b1WCDd8Pc0VlsbKPLi+NMCHgg05ugpfAId2Hs1wXsP6FFeuGgHnrwREQtN0CegV7KbR5GeiR5yzrSRi/uaV3BatHlkocZnwBWGnuWhv7c9JPT4O4Lmq6SYAB621hwHAGPPvAN4K/Lw6izFmGYB/w2x+8BPxp+KA2ARwViYY3ByhIJSpct9hAVeaKPHgqdLUwYHa4Bc3Oc9iUkxGHmtw2caYRyshBQ3HLE84PFFoa/LvDTn8O6NtuqOivnephwfJasjciTo/bKo1jotQgxgBHViMi4FNaQ/vxPQ1eLAWMvrVXuTl2Gtly/38cwpuUawLHkVMM4OzOZ4wq/u+yMchLlsbPCBb3lwrr2dtkgfR4+CBzJOZL1Nfuu011Fdu8E6dgs0AQF+IB0EeUVB5YFxMJK/zxHbQq4PkGnji0l/k3QYlQJF2eNIb9fM9ZBV0l687xBOMrOCj1UThKO7n93yFj/ObAMAY7lcSyIjyRGp7jscHmbp+lkEPfy9VAHmo/BD1xQJ8f0maPurLVriEupsFxODMal3GTy2bp9Xr/J1Vvpb9t+0DAH588v/vAzAX5z0M4IUOns8D+Ftr7VeNMf8Z1rPCGLMdQAxAmHz+eTttCcbJds0TAEastW8yxrQB+DqApQCGANx8ssXjB/BFABdjVnXx49ba+09+hx/AFwBsPun7PWvtt+d7Tdkmr4QdnuG3KiVgGisT/JA6IgYyAcCq+jnUlweHcdQMP8QvMLyKGnXR1FejviuiuuYRwUG1yStExugEI+zwZxISMqRBgfVWEJeE80Hqexa8qgsAy5o8OIgaHmCHGnyex1NNXiU0LrIqPsurgSrRDBZ50K7IrY9kv0B9i5J6gnHcw4OO9Q6X0lSSw46A46iAFQBeE+EwumiZr9l06k3UVxNDy46JeZwDQqUOAIoChrkrLfg0SqdfJK9hAYcEgG7Lk+LFItEsCd6LR8iEn+Pj+53bcDZlKlHYVuJcrT6xPjJVDUlSEM3zU/w5BwQ/bqlQV/pF8zZ5PWr6/GDwKuo7XOKFmr1NPqOo7uXd5cVBLp8NADOBpdQ3OiPma7jMs2DmcREQaE3yfnWaBXCKGlsdxpgn5vz3l6y1X1IfMMa8DrMJxnPt8Rd6YV9oc7gCwHOtzX8GMHea7SFr7YaT338LgC8BoEPQTmcH4+OYJYw89yb+NoCfWWs/Y4z57ZP//VsAPgQA1trzzWzE+WNjzCV2tj/5ewAmrLWrzezkGq1v5mIlgb1eGecbcUjwLJTsXlRohQP6EKs0BKa9KvBc4gx7ym6R19MPnvDkDA/KrFDj7weXza25zHLYYXlQf0GFS6buNhyv29ngVbKbujl05vaYhoZ0B3nS5xj+OyeFOtXeHA8AFBkb0FXW0TpXSVoV0u1rZg3zK/P6HAC8sZ0HZWvjPOi4oI13E4IBDueq1fR7OSUCpMMznJ/RPckryT8t8M7iaudc6ts2ra/1rhwfQrdMBPs18c7+xqIbqG9jShOVV6Z4RVjkgyhWeXA9I5S7MhW+4e2ecRl0KlCG00Ix7Rwv30P8Ss7bJd+5roe/0+ck+JqMi73nyQnOc1MCCgBwWTvPfGfaBeyzxIPv6RpfW88IJbEHKjxJAICYKFblgkJJq86vVXEwWglEy15mm2QcjJNdhg+d/M8brbXHjTEXAPh7ADdYa5+T2RsGfi7g6gcoue3FVF/uAPAP6h+clgTDGNMP4CYA/xvAJ0/+z2/FbCcCAP4JwP2YTTDWAvgZAFhrJ4wxGcx2M7YC+GVgNuo9mXC8JN24YRGw7snw4PLhOtdvvyX+euqLuEB8Vdu7XFcHFX+sM3VO4L3U0Trs7UEezCT8vM0cE79TfCVcOJhI+S+iPlULPFbiycBuQW7dKSq+D0zoSak9AR54qqGAyjamOPwuKQZ6AcAXhznWOSwUryYr/G/uMxzSd1NiHfW5zU5YGeX3tiIq9A+P83d2T46vEDfSeZ+Y5dAd4Nd6cTu/1oLoVi7izSbsy+pz4Loo75z5xEvy+h5ebOmN8MRtqqwf5iOjPNDL1l1GRxNL+fg7uyTCE57+kFZaU3NmlHLg4Rn+vUvEu74uofeQpVEe0O9I8+7hVJX/zV1pvn7KDV2KjXh5t0oNkV0dUyIS/Hp6wvz9uaS+nvoAYKbO7+2k4AjOVw1KcbEAPRuqZWewnYZBe9bavwbw18/9tzFmMYDvAHivtXb/nH+6DcCqkxyLEQC3Anj3C3zlIyd9/wKAqy7Mdka4agdOXwfjcwA+DfzcGOdua+0oAFhrR81/4GN2AHjrSULKAICLAAwYY567cX9sjNmM2R/6MWvtfwEIG2M+DOBTAGT51SvIlG0Bfqs+2saTiGMFId/qEkFnBLdDEcS31XnVcrSyk/oqNU78BYCN9bdTX0DAq+IOv69qGvdUVav51MAPqh4/h/I4EB2D2vwGIOWtrtzmRbBy9xTPi1f7eYXxoAiSr9F0Gnz2HB58p4W8q1K66Qzw+RlNy4OjnVmtj/zlQzw4OC/Bo281J2Vflq+tR2o/lNdzlY9DndYkeYB9QYKvu3cv5ddjRDCyNq6T06iQiFbn4O4cv6+7cvw3fmtUTEUGsCnOn/VwgT9nn8PX+sEa54S8rYsP2tuQ1O/69d38na41+fX4enlgHhNzbU6UdHL2Z8/yZ7Iuxd/Zr09zkZipOg+u2738fQaAHjHEcEeR39vBGb73REREsprnUIi06VBmf5Y/r7Xly6lvR5hDpMp1fl6WK1yYomUtO8X2BwDaAfzNSZhw3Vp7sbW2boz5GIC7MCtT+xVr7bMv8PmPA/g3Y8zHAfxnysFzHAwDoAqA48ZxGhIMY8ybMAtrevJkYuBmXwFwLmb5GkcAPAqgjtlr7wfwiLX2k8aYTwL4cwDv/c9fcBKr9qWTf39e6WVGMFjDXn4bVUt8X0ZXrBQJPOjhgUUD/HvXe3kylPdr5aEtWU5SXZvg04aj4JCBcTHd+InqHfJ6FPSqXWDzC3UeBF3tp3BCBESQMyG01AHgoOWwoyWGt+gPVTikJASeuN19XAcrD3n4ZxU2vV9UYKfK3JepirkKDR2UdhkeWUyLJHxwRgSsgmNwmfdGeT3HLa/g56Z592dXmu8TK2M8eAyKXXqsqKvMBdEViAmIZk48r72WB6VtVvOmns5yKOWqMH/OKhlKVfnn/mrsx9R3c4nDcQBg2wwPIFPg1fvFYZ5EDRb4frcLj1IfAFzp3Ux9h3MaEsnMJ5TWBgT/CwAiQv3u6QpXWkvhzdQ3Ijqkj5S4kIg6CwDgSInLhJfKnDPT6jS07L9tp3nQnrX2gyCBv7X2RwB+5PL5QQBz4SyfOfm/DwFCyu0F7HR0MK4A8BZjzI0AggDixph/ATBujFl0snuxCMAEAFhr6wA+8dyHjTGPAjgAYApAEcB3T7q+iVlCy7yt3XBM8ipB1m4PzG9BJfyaVP21NOdEhIUU7TkQsAgPDzgyVpNb35z8depTwfeQkAjcXvwu9b0u8kvyehR5fHVcEKcFyXtIwBvGq7yieY6fJ1GArrKmRXV/RYA/55+UefB0rcMTJQCYEjKkz5od1BfK8mAuaHlwvcbDq51v7HSbv8LfL5FDYKjEuwKrIrzDdaigK9vtIuGJiGKDgq3NlxeoujQA8Ex1lPqu9vMAsiPI3+dKhu+TPWGNL1sWE2pQ4rb/NMeTmtVezuMKCuGBuksgcJNYl6qTpyZgN8QMlWR1s7weN84es0v9/DzwOxyaN1HWBbBnLO+Gq0LWs3iC+lIOf5a70/9OfQOpa6kPANpDvOBUEcpVU3kOm1YytSHB8QKAYlnN0NDQPWYKlnU6EqWFdj2vlLXUi//DXvEEw1r7OwB+BwBOdjB+01p7mzHmswBux2y2dDuA75/8N2HMDgQsGGOux2y7Z/dJ352Y5W3cC+Ba/Cc935fTjuRFRS/DV1R3mCcmeSF1CAA+ww8jj+Xfu7X5APUtNpx/oCrpAPCU5QfD8hqfK5AyvNp3rUgijhi+CQPAWofLTObFlLWJEr/vd5XvpL5r/BwaExGqVQCwuME5Kj8ufJX6KuYm6jvXcPWlguDaAEAavFv11uil1JcT0dPdQn6ys8ZhEQ9OaCickkBWwyFVErG7wJPpZQGBxQBwR+Fb1Pfe2M3ys8zU1ORclfs6RSIAANcleBKhtp8HMnxwZp/DtTQUrwMAioI7pniFt3RyuM6dE3x9XB/ivLJvz8hCHn4zyt+9qOB9TFWEepn4+V0hfRxPiS56QyRLEfFQnihxLtZoUw+vCzm8+DHiDFLfJofvL6N1To7uSnBFzHMsF/UAgKrlydIe8EKeUpGKhXnS4nV08dAE+XuZL/FkWtlCC9odDz/3WyT4s8NO6yTvOQnGm4wx7QC+AWAxgKMA3mmtnT45PfAuzBb5RgB8wFp75OTnl2BWRisJ4ARmx5ZLQexZiNQLb+Qr2/gExv4GD2YDQrYxJ6rTlyT0QC/1aI7keSs57uPXoyrXw4ZX2QFgjcPlOZXiVbrCg11Floz7dbTyrIAwXJjkQaIKrMqCWa8m/x4v6IBezddo8/BEMiW4P5Oiwrg8rg+4I6L0r/6mgjqpirBfdLhU0AVoyeHlMe7rDvLrueMYfw/aA/re9UX43wyJIrNaIj2i8awGeUa8ev9u8/M/OlTgv3O/OP8XC5G6aT0zEZ0CuffTUaFKJKY4/0IPT96V7ZzWpcYlMf4w1X3fPsXvea8gGaiiCACsFN2fxWH+N9cm+cNU717DZcbKjOCV7c7xRfLEFHVhX0HA0hy+T45YDVHNObyLfmCad9H9Ps6B83u5wtR8k4SzxRzD147Pp+OiSnV4wU7yvqi3wz7+YV6YmK/5/+irC/Y3Kzutg/ZOzrO4/+T/P4XZLsR//jdDwAtjfk4mGq99ua4nZPmGsbN5P/XdEOIL6h09HDKgglkAeJrviWgTU2+V2sYaL49k/nZCD1Y6IIiNfSVeSV4p4EpKRarsAisuGA5ZaghoxNqkUk7h9643yA/xmYSGL1zVwyMvv4f/jiEBSdovZDaFiiYA4Nw4f/XTYqhkUQx9my9U8EcjmiC/KsEDi6s7eSdm/Qpehb/9Wn6DXGYUIjPI78EfbuHwoQvEuvv+MO/ivLmfP+ffGuRYdwC4OcF5BiqpUcX0DUl+rWu7RPQIoK2XJxEfdVHvYpYd5uvnvkEOzcspeTvoIYVlEXxnhWJRR4PfWLdu1BsW8QNhxXJ+331xvu6mh8TskbK+P70dPHEZSHDOWVIMthuY4Tosan/52jENDU80llJfR+JXqe+JAodlqdkbu10SjLMVPvScNUVHqd7ga6dlZ5ad1g7G6TDVwbg68Wv0c2/o4YHegJCt9AnZPTedcaUeo743JaqWXsMPTSXLCGiceNgzP5JhRSQtpYYO2hU2X/lURTjm4xtfw/JrzdZ0VFoVv/P+cf7Z9yzlm20qxAO9vJifAQCH87zC+JQgI+/L8kQp0+CB3ro4f382teu1sy7J70FEPK/JEg86ns3y398b0mX4VWLmgFe8B1kh4Xq8yK/1RIW/l2qQHqCHoalw9jVixsEFS3ni5vHp8yQ3ze/BsTQv8PgcvvsEhVKWXzyPvJitAQDTFZ7xnBDv1+4cf3+CYmbSRhdVq8tWacgos8ERXhGeEMpVo2W9h0yLwYCdAX7fVVdN7c3TQt3uiWm9/z6Q5apOadG577CcWxcXnLMtNQ61BYCagF5Va1zYoymGAZ89Vl+w1fyLejvs4x88BR2MP251MM4Q88DjeeGNod3LK6XqMA56xOEnfCpwAIAfDAu5Q0H4S4nuxmiRb/x35DgXAACuD3NJ5LpIlpTiVZ/gqGSqugxfETrtS0R1MqCGXYEfuBMl/vceLu2nPgCIWB5gL/bwAOCrg5xHMFrkAdBbBvSrvTrGA8glEf47iz38e4sNfu8SPt5pKNb1tT58glc1Fd5dNQi3TvED3nHRsG/zcQ5CRUzkfp24dyujPLi8MMVhI34RXAM6wI5FeYJaF+pTe47ybuX2jMBPAThSUCRv/sCGhPrSOgGHHJzhyWJ/RAfQ7xgQSVYb5/Bc0cefZTzBA0S35Gx6gt/bx0Z5V2CkzNfdoCgWu9DKcEJw2dpEa3q8yH9nUxTVbuzjF/S6bp2cXZjipOtnsrzLtSfDv3evw+XgI+DQKgDINQVEM8r5jKUa72LNF5blNrPD4xVCCXUtDDMfM0LhD3AfSHlazVrY06witZDsLEwwGpS4NdzglYOVVX6oLhbt8rJoiR/O6wp9e4B/7yOlg9S3ssJhGmqK9aYgn3MB6Em7vSEe7Kqp0Xtn+CF+TkzPR5gs8+99dJo/yzVhHrCmxD1Xvtea1dQHaP6GELXC50f+jvpuTnHRtDuHNSfk3CRPXMQ4AqwTMB8FrSo3VeKmN2QFJewSyAgFsRsICVlY9UAADBd40Kqmz39/mH9uRZSv9dd2CTikmNIMAGNl/l4OH+H7T0GQsbdP8b8ZdOlkDkT5fnisyBOeNh//HQXBXVDBiBsh/VCer9mnM/x5tQkCuGeYX1DFhfMwKaaSf2GcCyxcF59ZG0MAACAASURBVLyS+r6e/kfq6wlzQRAAsGKe0MocD5IPO7wYUwdfW4HRS6ivU5w/ADAukqGDNd7d2FO+m/rKFa7QpiBAblYsSxrpy25ukKxTkUQoCwU41xMAiuUWv+VMsbMwweAWtjzoGBfV60dEBfa4iNbWchEOAFqxp73JK1b7PHy4YmeDt3zVcDpAKxONlXjwVBFB1z3Zz1PfUedt8nraGjzpaxj+N48WeVIzJbpKw01OJCwZXhEHgFSTV7QC4H9zY4gnffsq/Hq2Zr4or6dhP0195SZ/zo8KYv0SD1+TubpYH2JuCwCkBUFzbY2rsfSE50dWV9cKAB5RYdsjJH47wa91tMjXwEf2fpP63hy9lfoAoFAX8CEBs2wXFeiuEL/WgwU9Swd5vscuCvKAfrjE39khIaBQNPx62ipcuAMA7h4Vg04bfM2mGzxRGnL2UN866GnU5yZ5EL3ZzxXlRsq8a/KutvdRn1sCdrjAvzcP7ktZ3k1Q++iMOH8yOf3O7jJPU990hZ+XKtj3enmhyutC5KrWNCn9bLZXOsF62a0lU/u8tRKMF2kJoWikyJIKAjQqWsUAsFNMWc0ZjoOeKvDuxtrIbdS3u84HDgHAMvBWclZgQ0c8nDzeI6QH3eyI4RrlYSGheLzJg+SJGf6d7RGuJx+Czha94MmQguSoYXpThgfe/cnN8nomLL8HCcMTzXP9LiPCiT1Ue5D6Oj0r5WcvCaygvvaAmL+S50GgCpL7hMIJ4ALdq/AgUQXQarjh+U0+02S6pgOrvhD/m/kaTz5U16ggCh816M7ZRI3vExsjXNZydYLDgwp1vl6P5nli76ZeVhLJWUnMKkh5+D2PWr4+go7uaO9K83unFADbLYdgjohZMQ+VvyGvZ23wDdS3ws+LDVNV/jezDg+8R5ucyD7lImmuzE8g0wCEdhlgRZei+gpX/VvWsoVorQRjjim52V6RRIjZdRDIGPRpBBDOL3Ptd5/hQVfDdzn1lQVvIWD19OeSCB4WB3lwkKzxYU6HDW9PXxLgUC8AiHjPpT6Fv68KyJY3ehX1HcrzKurysFYxUTCOXE0EK0Iytigqk5s8XGse0FLGKvAsNkTQ1eC/4w0iGKkK3gIAlOrc/9PKEPWd4/COgZobMOoyYKza5PdAYcifrPBE+7CYIXJl4B3Ut1pMAAeAZfy1xAmhEnRYDJxU3Y2U5VVdQL+XFbEMdqe5c1udY+H9oivdLyrpALBc3NuYj/M3Pj/yD9T3nvb3UZ+CUQIabhspcVjJpCj+qIn2b4vzYhQAZKpCLUusEZ/h9/WI4P6kFI/N0VLF+aaYg+F9kvo40FbPyHAzxTNQA/xadgZYi4LxvLUSjDmmKkjHeGyJ1Vz8BANCkcZNJanc5I9nf1YFpfx7M6LgOVDX+tNqHoEiXIcFyft4mU+D9YjJ6gAw46Ibz+wbGU5mf3/n7dTXKyrQJ8rzn4Mx4eFY3lSZV2C7xETpxVGXKrwg1GYbfJF0+nlHZXWCBw5q9khTSijoQXNrBezo/Db+/qiZFB1BjedW3QbFC7ogsJT64h3cp6ST1yd1MuQX6kvDYg7GmoSAfRb5d8ZdgmRFHG4Xt31JJ/+bF5d5oeGAIDFXXWTClYqt4iklQ3zfyohheSdctKVVp3N9O3+Warp8SKhadfhdBATEbJuyKDYcLoiwI7OWuvY2+UDFe6Y5Vw0AOuIbqc+47D+nwhZSEuFG8lbky4X0OxaEWbRI3nOslWDMsXahvnTjIk4+G4jy4NEI2b2HxrkaDQCsjvJT7BxRmcyLiu9BQSwfLeqN5nCJy3MmHB5c3l/mk49v7+DKVNMugzA6xVSzsSI/4L5+AcetD0Q47CgacJkiNk9zDK/6loTM6FCOr609LlLi2zK8hd/v5xmzIs2q4HpJVJDn/XpDrgry66Y2QQwOCYy4kBl1xDsLADGxDopC+vTpNE8IleDDbqFk4zE6GVK/ZKzE35G4KH5c18Pfyw3tquYLtKX4Xlmrivc5wze8Td38OU8UhBzztNhEobtRYyUelN6Wuob6BnP8nveFtarVdWKWzprkCeprCInsnWJNTovnAQBr43xdLm3jEMz1Qv52Z5Tvhf0ZDs9cF/x96gOA7818m/pW+jgJvjfFyep78j+kvkRoqbyeqRnO1XopBPH5mOvcjVMg2+Q4fN9qCoWtlp1Z1kow5piqTKqZFSphLdd5wNHmUiE6R+jt+4T8rZoBcblABbjVcVSy5Bi+KTTt9dRXF8Rp9fcALcHp93Gf+t50nuPWnj7BOzxbp3XHQDU4FDSiJ8SvdbWQGb1CDMECgMVhHlypJFQRPw+KhEcNVFwR0QfKkjjPllTwdCTHf+OTad6NanNJeJaEeaDXF+EB9OViCF13QLRBhXRyd1BXEGeEGtSmDv5QLm7j5OjeGH8eWTFXAQDu2j9Afc/m+PWsjPLf2RXg7+wKsYde1qFx8j84znlVf3b0/1LfZ1d/hPpGC3y9Xtyun+XyGH8mM1UetI+JZzJe4fd8gm8vAIBykydvGTFTSf2OS3t5l2JplHNtHphQ7w9wXorDDJVtmVAbN593UDR6/4WCLuaecLmqM99e1UlEq6nzvLUSjDmmlGW2TvOM+7jAMufFAb88oiviz2Z4delIcX4wqK3TPDhYEtIa9s+W+ebf5/BujOoMHSryjTjiUp1N+vn3qqKLUtZZJEj5ioPRJ7opACAQZJgQlf9DImgfFYHDiqiuhiro3poY/5vfPcp/59sX8/cnLBLin4zpoLQrw/1PTPJqX9IvJKKL/D1YGtJqag8L6GJXiL8Hl3eI6rXotlwmkIsFF5jlGjGIUA25rAnfkxMctve1IXk5WBzl1xv18r3yd4/cQ323JV9PfUdL/OZdkNAY+luXceL02wZ4wLonw5/z+1fw964toIOujBj8tyvHk6xPH+SQ0FtTv0R9Sm0PAPqE1POPRvi6u6SdP5OBMN8nVsd4xnNlh27ZZkQCNljkvrLgW3VZ/h6UmzrhOZT/kfS3rGWvBmslGHNsQoyoT/j5JhQSwdM/HuOQm9emNDFtSBAtD9e5itRxcNJjt8PJ4T8uPSavZ8Cso76RJv+dowJOcBTPUF++wg94AFhhOCH7SHUr9b3GxytPjxS5Gsk5XkUapi4AwGCFQwbKhgeXSod9SYMTuZdNa/5KVOCy1dySmmjf/9Eh/rwu8HMS6sEyr+wDwLoID9r7IzwhVBB7pXb0YGW7vJ7jxaeo7/W191LfcJ6/B9UmDx5XxvneU6jrbkvAw5/zPTMHqE/JKp8b4YHlPWUuqQsAtwTeSX1pwfLuczg2f3CGr8lhMdjvX4/pgD7r8OJHT3MR9aUdvjf3CEWnhyo66Lw19W7pZxbxcwW7reVB6ku4KOPdkb+L+tL5Z6lvzOGwWJPm70hETM4eMXwtA4APfM0eTv9AfvZst2CAq0dWa3zvVl0KR0Cqz/Rp5S1ayn9YK8GYYwM+3jFIiuFJAaEn/9GlPDg65vIeRcW07g1+jnU6V8yHUApKiSr//QAwZfhmYkVfMNbk31tt8na522TSpP9N1Lc08BbqU8opfQ0eOCii9oBH47lrhgdBQaF084YwDyoqoro2YzW+IWb49SoI2QHwZ5JtjFDf4zUerJ3r6CGFhwr8vic8PDAvinkeh8CVY8bSOtG+OPkh6vMIQuSPi1+jvnMC11JfLcsDvW3Ne6kPAN7fzpPptQ6fAxEVQhEjYiDeZr+eXaOSvpSQHF4PPr9nV5kntlfGOG4/UdMcuKzYD70C1tgL3k1Q3dMVQV4wAYA9BQ7pihve5bs1cQP1qfNgrKS5AF0eTq73xPl7WTG8MxKw/N7tbzxCfTUxe+SlmCJAu3IXhKmgvVbnZ2Kj4QK9mocpPgQAlCtc6XG+dqYnES17cdZKMOZYXgzyOVHhG3i2piqT/O+5YVzvKW+hvteH+PyIgJhErA643dCB1et911FfQgQHOXETqpVN1BdP8U0YACKWb4xLBERIBTkzohpaEjyTTE3DfNaHebdK6fGreQQlAb8rCHI4oGVaFUysnucdsOVJHnC48PWlhT38WSrlnZDla/INDifiBqM82Af0HIyYIKl8NMG7GzkRy6m5E78a40k2AMR9/DlPlXkSoYaZq8GZeau7AhcJEq/i6fQE+QUtE0nEU1MckqTU7QDdycvW+QNTs0fGKvx6rojxDikAiGYdxBLBUTEPZqrOg30lPQ4AlxkuPw6/8AlTyWKvV0wWd4lkxhp8wGFCzAzK1/j1NMQslKJL9/1UBO3ztVc1H+KVNosWB2OOtRKMORYQMrVqA1ca7evbecBxXkKvxKj3MupTQamCTfgdfkp9OMBnFQBanrIk/mZnkN+D8w3vGPgc7gOAqQrf4I+JQ1XJ7W5r8Lb/e5K8OqsOfwA4KGSF436+7nrD3Jeu8t9hxLA8AFgV569+W4A/y5SQqQ17RTArzjAFcQG05O4SgelXiaS6VjeVQfVM1HupvtcnPnhZh3gePh0cVIQC17CYHq4SjNeIwNOtqnvPBK/C3zLAOwYxL997lADHlFALy7vIXC+J8ecc8fJ7pySQc1X+/qSFhC0AzIiiQU9YkMc7+LWWGzzhcxtEqNazUpRTBaeGEfNXmhxeVneJ6qYdvh+mxSydkI93uZpiDoZ1SbTDQZ5Mliocpns6ZGGljG1LwvbnzKIFkZprrQRjju21HP6xss6nON+8RASPgmS5ZVqTqpWpw1HEqxgUbe8Rl8niT1uOPb8xwvkAT2U4t+XJGpf6uzrwC/J6lL67T2x8/VF+GF9d55ASlbipIAcAIqKyrVSkFA/nodqj1LfB8M4QAPz1+Pep792pt1PfsTw/OFXnY+oltDAezXJxgWU+HnR0C+K96iw+k9Hk1jHDr+f6+FLqCwkS8+eOcyLubW184Nn6lNZ+28pF2tAumm5b0zwRWOTjwZqCiAHAtV0c7jUibntOBNequKGEet0SyYmSkBivcI5BVAwbPGE55yHW5NBWAFgFDmlTPJ19Wb7YFSztnpyGqG4K8utRioxb6rupLwi+tvKGB/T7KvdRHwDUGnxx1eqcH1cs8eGYHq+AFLtEmU3R/Qj4eUfudHQ+ZNHglEjY6uGhzaaLBnvLFoy1Eow5lrC8WhGd553qivCN7bVCZhUAns3yzXZEEKc7RQXa7/CDaLsqMwO4JcmTiIgIniJeXvG80vMu6usQvwMAiuL2qY6Tqs4ebXI9+XyJBw439eqx7CeEHKRSTqnF+HPuyHHMthv5991RnkR0icBTafX//+y9eZhdd3Ulun7nznPdmmdVlYbSPNuWJ9nGZooBYyDMCRA64fGSbvLySCch3XnpTvo9Op2v06ST7oTXkACBR6CBMGPA2NjGsiyPmmfVJNVwq+rWnedz+g+JuOBprUPKsSWhu7+PD0tb59zfOec37GHttZXBFpJF5fw6ALg5yg2vZ+b5JNjVynULguN/tqjhbq9qG6K6zgD/TeWE/uc13IlIutBZK7m5gz/ns6JlxcYoN54mChzbGfHoVN6uJL82IJoCKtrTrICoHs1w3eYWPfEWRIawNM/hgP0iq3a+wLOyQY92FhVkS9FH//Iwf+f9gjL2PVVd5L1U5Tj6M3m+hnrzW6guKwJnJ3P8LL3ZJRilggJHMrxPk1/UOno9/BmrLqTvlSofjyOcj+tBvC71jNWr2cFoQqR+Qq47B8MyAYSCl+divzHOcfKqoZdPFHl/9zy/5w2teqHc3s3Djx5xGOfKPA2/JFL0N7VqTHJBdGetCShGxMsNpEdT3GBdHdUOT1Tcd1WCR6W6V/ND9Xe6xDvw8kPBzutDobbI54hHBGyMWKE1YSDOTupN+slZzhI0U+E/qiLJ0wJCNxDh193arjMG24Z5p/MPreHX2SU+niNP8nX5phE9nlic7wVlATtayvMPfWKJ01p+7hw3VnojegsfFEnS3a38/ag97YPruGE50MO/FaAJBDJp7qSHBK1ytJPrCmJ/OTEt+H8BrBJUvR1+blwe4luPhO1tTersz5tW8+h1Sxefs/PneaDq6xPc4bFc4G6rwjwb/toRTvhQq4tmryKotihoer8zrQuVT+c5QcmWxFuprtfh+8QJc5zqzpV4Zh4AfCLzqqQusi0rlZe7sZ+b1Ou6WWdTrh257hwMwKHYQNEAG4ez/EA5kOLG5S+u4ou3N84NXQCoio14Ns8th7BX4FiD3DjqccmoeIXeIwwSn4+PZ88gf+lewdwFAB6hVxnqhQlu6E0f5Ib5TIlfN1/VS6lFvIMjWW4ErY3y627s4tmWREKzdKwpcrjBA9P8HdwuUBw3JPlc3zXAqTtbRG8AADACDD73HDcspkSX4gsiS7FvXrOpdQX4N9mY5AZAZxsPKATEmh2J8fmh1h0ABMV9VZNCWdcg+q88LBrpAUClwe/71CJfQ4q4YkuCz7vt7dxYGW7TjfbOzPMIfm+IBz9qgnzi9g6+h25qFXg2AKUKd16/8QT3tPtD/P08s6AatmqHp+bweXlWzINb2vka6UnwMzEq4L17GzpjcHc3z5pcKPH3+sAF/pt1UWexs+VX5Hiey32e6hqituNaElVnYgSTY8mlQP5FkHe9LPISoMauWbnuHAzbqaJYnrisTlH2RURg+1W96qDmG/izczqK8dwSP6i+uMCjJ5tErwtFfVu1tUHfIoo7FKWhWnCqj4FK+wPAtMBI9YoPNi7qGmYFRaBK8LTovnboEdAihfU+scRfwqMpHn1cqxMY2CwckN9cv7IomS2KAb93hvfB2LKof69U59vUl6d45F91T28XKKi09nfw9Sl+4+Eozwzd3cUdl+3tvI9Mh0tDOCWPT3I890BkZVSR4wXuaD+W0ou2VWD+/R5Bj5zh+0tVND6cq/AI9G0d2sEYFF3A8yKa3uIXGZU63+/+1X7dnO3tQ3wPWazytXcmz687UuaO/4iXz2UA+MG0aEIX5L/5+BwPjg0IaN76OD+fWl1ghKfFO1BF+ZKNMLeD6p4Gp9R9MeL1cphuvc7ns2X4vHupMhjMzvp5l2aR9wty3TkYSrwiUnpOnPE3CKx3XEQQx4o6rXt8iS982/BZ3BPmn1Wl6EsubHVzJf4sPosbDknBqa8K0qsu0PMRwYSkDJL5OjeskhaHaVRF34m4T3sYZ7MrY5FSDfr6bO5FeBS2CsAR0R17TZwbkMeWBFtYiK+fGQGfemROY72TAX7fpHjtj6b5d/aJQsK4KLoHgKGIrtFgcjzHD/kjWe4I3CYgZOcKeg9RHdufWuTz51yev4OgyCZEBDsXoMkp2hQVrVjr/aL8SXEL/Kej2gtvC3Bjblhc6hc+liIX2KbbcuCM+CZDEf6gVZvvL1tDvMbgaElnVF7TwdOZih45Vebr4LEFnnH6kTgP2412zibBjd2o2EfHzWGqm1p6WP6mErdCZiaNFUKkrjYYVFOuD2k6GMukWxhI93RxeINX1EPMCTiBOvwB4BU93HpaledFhqrYVhkHIa+uwVB0ooPCqUkKG0iZcnUXZqZnRVG6KkbuCvEDRRVHjxX59wq5rKRNSf4PVMQ8AH4YFxvKA9MD8ot58O1pPtf7Azz62BB1OGvi3OpS/Q8AbZSN5/n3Gjec7rG/Joo3XKRd0C6HRBT+27PceHrHADdmn1zkxsgaAaEDAJ/YmxTU6/E5URyd5B9sU1yPJ+bjc1blPjI1/psHM2JtVfj3UBTHADAS5dcqDoUlkU1Q0uJCdW0LbIgqdFcwy7jIaBfq2uN5cI7P5wPlL1HdQIT3cLo7tpnqFkVKMif6kgDARh9v1PhYnfebChqeUelt2Ut1qn8GAHQFeGf6VIWjE7LF01QXDvJn9IhmetnCCaoDgFh4LdUVynyPDQW48+r38HNEdYG/6qVZ5P0T0nQwlolXFCD6rRXWH4gD3nYxoMcKgsM+z8fTI3j6k37Vq0CPRzWmUmvqaJqPVcF8OwXNKAAMC2x6u3BqVIPD+TI/qHxGHcb8noCGjWarXLu7lRc9Kq55NxxoXXihI2FFQ8rvuSiMucczPBr62g4OYwGAqQK/72cXOb3r62OcmUlRxlZcDojHF3ixvyKKuL+XZ2pU/VerWLMPz+ot/E5+xstAxKAwvguiwaNi53LTq71AfC4UxdpTFNCtLrBG2eehwg36w4t8D9nZzr+X27z74jynjV1vcbz7LV0i+xPiLy/W7Ub6wZ3i19nv59eJ+aNgyg2RiVHQX0B/y2idQ5XDIrvR4XCDvujnFL4AYIkmoF4/rxeJ+znU1FmhZRv36X5Ttrhv1M+zWKoPhqpfUbUbAFAsa/rkplw90nQwlomKmGdr/DQKiuJnxcby7WkNtVDp/Q5hfKtMwzHR8G0kptO2SzV+3wN1vuh7GnxTXLA4u8d2ozcaFRVXtMI+ySjLv/MTWW5YBlxS3gvCGVAUv5PCkVRyqqgZyrq83InoEtmoNgFX6gjwA6WvzA+ib6f4ewWA9UGOBX+z6I6teo+o+p7DaU2+kBRNDE9nuff6ZJZ/ky1h/ozfLD5Cdb/VcwfVAcDhDA+LKwf1yTzH5m8I8G/pEU44oKE8KoMxXlxZBkyRbBVFwTkACESO7POgsoMHUvymnS5p0MPpz1Ldlo6PUN1js/w31yX4/DgpYKYAMBzj1yroomKbU1JW+F4XeaLGe2/kbD7X5wW0yCPqGhbzZ+R4qjW95zXl2pVmDcYLct05GMb44PddPqynNrDpEt8xN7Zww6E3xnXvDOiih48Ihpz3r+ab24QonD4wzw/qgktn25joZbCxwSEnCXFwrhEdVhdFp24A+Pt5vonvCfLxtAust5K39nDDquTiB6ju4ZLeVUaS+e/Z0PjyhiI0ELtCUEGAzvP53B3ic7nVZaxxP/9elgh7q6jls8KJWBfTXdC/nPsu1f1i/FVUd2OQO6F+EYi438OdiN6gXiPbW7iBNCbqNxJ+HtVsE/1p5oTTAgAzZb4OYqJ+wxFgyhMZPu/Kom5qVNQaAcBqMS0jHr5mO4XH89isoNaO6n3pk9t+n+oWVc+OBh9PUWQTDtgH5HiQu4GqPHF+BimI4TOLvN6oM8Dna1iluADc4dlEdeU6hyudrvIAWMWI81u3RcKMaB6qGJZUVsASMCjbdimwFPJSFIirGhTbXhn5xNUiTRapF+T6czBgwee5/Oo/X+CLNyLqEzbyTLGEQRVq+vX/xU2cxWQqx0+/J+b5fXMiC5GqamMlKpr0bW3lm1tAGKV/cPKPqe73V/9bOZ41NseGfmL+61T33sDrqW5A9A04JwLbboxXKpOlDOGMgE+povu8rTf+u7v4CaigKiUR9b1vgM+BE1l+3WhCZ/IULCst8O6q4HiPaGM953IW/3qIOxGqJkQh/qbFmSoSSkiImgYAOCbotTtEU8AR4WM9u8gne4+LYaWKfxNevv+qsUa9K2MIcmN+axfMRKpx5oTYJ9Yl+I/OC+gmAJzN8bm+WZxBwtaXTS5fH7lFjufGNn5x0MNfvCI7UE6fqsVy2X4xLObzEUFc0e/hsEaVIe0pawauySQPgJ0p/ZDqFDNTa5Q7UUrcnASP4ft6KvvUin7TCKclGtKQrXzp5Ip+sykvv1x3DoYStWEMhcUBJ2gJHeFg/PeTOoL2zmG+CM+K6OOOJB9rv2DAmXEJHKjItoJb7M/xaM2HVv0bqhOkXgCArNgX7/C/huoUk01QOBidIvMxGtPOWXtAMIIJB2PfAp8jXSFurAwpXmUAD83w0/r1/Vx3ocQnQUrU8CjYXm9YV7eqmhkFVfmlEd4nRDV8q4qIL6CL2Z9a4JaeqsdyHIUv52OdFJlVANiS4BHhOUG1+oioUd0sSL8UrBMA7urmEeFUkc/1tZ2cxjdT4Hva1yZ4TYzqDwEAKdFw8oxAIB7I8Xm3J8HHo6CSACBiODIo8NZRbpS2DHFv2hKkJwDgiJqr4jQ/S0cmufG9VOXzOSMCck8u6j1kURBp5Gt80k7afL6O1Z6mOq8wygGgLjIKrcERqlMOhso0NMTvLeYOUd2LEZWJaTQEBXSJ6656cQCI8+F6k+vOwbCdOkrVy5+e6crKaiku5AXMp8oX/fY2bciovhRbheFQFA36Hprli/6WDp3b+8j496lur/cV/L4JDi1SWF0VXQMAxSZ6Uwd/73d08g1spJsbMtFObpB4kzqGZkThvarw3Zvjc7Iyx1/AsbOiIx6AHsFTfyLHDbaVZjfO26LhWUTT1I5E+Ly8ezM/cMNrRUGxoE5GQ088W9TFjExzg+TMaW5YrRFFQydy3OtV7EEA8LHj/Du/po+/g5tE3f3uVr5+Nm/Q7DmBQf6co6GVQSM6Szxl8KEF3hsgdUqnWzY9+CjV/dHQ26lujSj0DwtoVdGlWdzb1o9TXdtOAZ1pE3AU0SS2ltLeoifC13uU28jYnOQBp9Iin5N/+gTv77RLNPkEgKfTfN4l/Fy3KcKj6QHPvVSXcslG/UPuCaqbSH+P6hSj01yGs2EpUY7ARVGQLT5/FJxL/ppwlICLNlxTrg257hwMoEE7ZapldjSrsgmiKFYwwAxH9ELpFl2328P8MPZ7+cEwnOBjLQpnCAD+Lny70PIN3jJcNyc40d0O3EBMUFCKJmKre7gRGO3j30Ttw42c3kwb8yvbbL0io+Jv5c8/sKT50mN5bnRERO+WSRFlLovI/28Oc8rLdQk91u52btB6hfPhCEzF4rP8Y1oC0gcA0R5+X7+AqvR18ucsFrkjoDIfs2UNLxtdy5/liNjTdiV5AGOogzuLbjZFI83n1sw+UROS5Os52MF/VLWDCQiaXgDYf+vdVFep8XcwU+JrZLbMv7Nb479Yt6I1F72PZvm3PPown7BHlnRH++4gj4rfuPk81flEPMGb59/yg5s4JWpRvFcAiHh5nwxVozJeFBBMsU/MlXQUOyTob7sTN1OdYnQqrbAJn7sj8M9ftayciHCQd4EHrm6IlINmkfdyuQ4dDC6zVX6IdSbhJgAAIABJREFUBfLc0lNQHk9C0SRqQ2aswCNsM8KwKNZFIbeL0a5EbZmWgJwoDH1NpBNzogARAL4wzSNhH17NDVp7kkf36xP8/aixns7rlPhSTVyb5TvSZpEZ2Sa6cW8f1JHkka3cQFob4b+5VTRnVRtruI/PDxPSc7IhaI7LC/za1Dg39M4t8sNYQTEA4Kmn+Le+u4t35Nw6xNlqem/gwYSBpOjHcIHDcQDAiDTfK1X6UGTVajPc0M2f1+9u4hyvHZvM8Uzw2Dh/5388+RjVfXU7p/xct1a/u7Y9/N0Z0eF503meBVVRLEuwMgFAY56PJ3+cf69Mmq+DvGBHVIXjADBb4efTp7/Osw339vGxrovzYEJvB9e19nMnCgB6a9zAXpzhZ3uH2CfOCWhe1KUob8georqnHc7ImC3xDTjk55mzAgmqAlemCZ/KfDQzFD8/YpzrrOTdGOOsxK8abX0L1e3xj1Ldzja+0agiQkAXEh4XMMUHi7z76KnFr1DdjuT75Hgy4AfyXJnTAN4cuJ/qfCr9KrtHAI9Wvkx1Q0EeBboxyDnKC8qwEpC1XENv0jWHf2vFX+4BnwObE9wgG9JESNgY4wbtmiQ/jJOt/CC3BUTqiHDqFgTuGgDGiyLLJc4iVbvxeJYb+6q7MaBpSFXh8GCYz+dbRFZg9RqecfNqAi4sjXHDfCwlnCzxTY4KCN1pF/j0jlY+18cKfK6r4vCV9lW4tV1X82/v5vtdopM797aAreXS/HvMZ0W6EsBZAcVV9SL7UgLeW+aLJOHR6/Lh2gNUd2/4tVSnmr12uQQbmKgGqW6i2G9PZ/kcyQi2oxN4Uv6mI86DxcIxqmPoi+tL6k87jrP7So/icrKjvd15+HUcOrdSafnUp6/aZ1bSdDCWiTKw13k5FlNRfqoGQB0uRXSqWdyXZnj0vmZ4hLEPPMphyRwF8N3iZ6hulWAcma/x7qNK2n0r77asZK3DAcJPOz+iuk0O70A7Y+mMQdlwnPhEjv/mhijfrKZt3vF1wPCOuABwQ7SX6k7n+CEWE0ZHrsHn3eYWbjyVXIyDotBnBPNZ0MONlXydX/dA7q/leG6IvZPqej3caFc1VS0+7kTd1rUyKl4AWBQsW5+a54GIm7yculN1lD5f0T1E1kU4VEUZl4/Oc6c3ZnGHZxJ8Xd4c0X12CqII7GyZw902RPgcGBB1CwEX21pldJ9a4I7/7jaeaTglel1M1/S3VMEPFTTxiQDfWYvDX2LgDfHKRhvep5Y4q2AowM/2Wp2/V7/oJRQQkCwAKFT5+V2uXJDXNuUqdjDa2p2H7v3ndzCSn7k2HYzrDiJljB8Bf/dldWGbG0FtoihUQe5U1rtHk0ihKhIcOyLcUVgS2PMHijyDcZuf07cCwCtC76K6MWeS6pSj0Nngm3u9oTM8ix6+SSdsfhh5BKbtRnMbv05UOBfqGq8cA9fXo7uo7lyVFwPu8L6a/6CL4flI4RTVHc18nure0vZ7VNcf4oZMWrQpninrSHLCyxfRjMMNz7Y6D+/nRSdZu6GbFM6Yc1zZ4NmxJLhB0i6M66OiRMWFeAhjeW5ADtnc0f5+5QdUl6tOU51xIQzt8r2N6nxiXXb5+N48V+OR5FXW5fd6QHezB/R6XxviTsTBIs98PCEK0lU3bgBYI+rntib5YXJsiTv+qTo3zKMuTEgxsS7LgiihKCi0U5UTVNfqv5PqpurPUx0AtMe2U9187iDV+bx8367WFROS7jbdEtlAdde7g9Es8v75kevOwXCcBmr1y5/YKoKvmvC1iEZginrw+UVtQJ8v8YPz0dLnqO7VkfdQnUc04/lG+k/keHa1vJ/qqkYUnTv88KuDv4OipaNSfQ1eDHbQ4VziJcP5wqM2N0rjhhvQAdEBHACmPbxAMenwbEKfl7OGLFocVtPtaBapPqeP6toS/5LqVOOyFos74QGRTbBcGFeUAdlpcQPgGYc7Z5sMbxJ2b/K35XhOGW481AX0YcrwaHpWFNQqY7bNpWmkMqIPOY9T3RbDM5JlQcudcOlob4TRfi7Pv3Ma3DBvNXzNKmiiYtsDgD5B9ZwV/WlShgdbIuAVzo83eCYTAI6n+X43XuGQnE1+Xqze7+XjUWsdAJ6xOSw2KJzpsOG6nT4OrVo0fL/r9K6jOgDIOBwS6RdV57bN53os2E91BdEzCgCWBAzqepegyCgBQLGsnbcrLc0i7xfkunMwFItUzPBUuzrIveLQXC2YotpdcuK+RW7Qxjzvpbq4XzQFtDkT1GRSY88PF75JdWsj/BBL2HwDn/NwthGVZgeANvBC7r0+3gwtYPH3XhRZE3XgqiaEALAevOgxLJo4qijqTIX/5pw1L8ezySccjACfd8rRVrTCihW2xaffXVw4byHhuNzR4NmoC1VusLpBBbdY3Ok7ZPHMUNLmNLUTFjc40hWerXxdiI8FADaKQu5o/i6qe67GD3GPapLl6Ki3avT5vfz/oLq7ojy4MSYM+obh++9oja9JAPBb/HhsDfA5sirPjd3zHk41u8nhTi8ATIE/5x7/G6iu1ce/iVo/p1wi6UMN/v7Ubx5t8EJlBe9tsQXjlf0w1QEr7/WgOk5nSmNUp1ibroS8FN24X4woatwyaSPQlGtPrkMHg0tVRR8F9/2bBlUHWr5hhl2KW0sJvimMebhuPM/hH+cM32gt6NTkYJQXTica3In4UeYvqG5D8q1Ul2ro2o2YhzsYXRaPgijjctLihtWQw405nwuXuCoQV/wnYY+Iogrc8VY/j64BwGSZX5sVzDJJP58j8xU+74olvg7qLo7koJ/jmXUEn7+7rjDPGAR1YBsCcYKenH7vTF4d5murN8znlipiBoCTAmOvyu88Yi+og3/n9oB2MBTj3ruDv0F1McGGFSsOUZ0K/tRc3t3ZLHdOvOJBkh4eqOLhFGAR2igNOzxTM2M4XHR1YIjqVJ3Jei8PQrjJoTrfR7ttvjcr5z5lcbKDXEG9WaA1xtnElPPRKaBVDWGYl2qCSQw6eFatcehVw1ZZfb5P2KIg/UqIosZ1BHz1ahcHurny9SZNB2OZBEXEalOSWx0+izsfiiHnbEEb9E/N8/sq6ldl7K6zd1BdQsCnAKBLdFwui0LctMWbUqkIY8ijG7ANWzzjojo8b45z4/KeEH8/E/mVEyJ0Coy9gltMFLg163MEl7gLOL9XZClU1kTdN1Tn62fK5lGptZY2ZNS3VEZyp2gRMSHO6e9meRYCALb7eO3CSJjDP54s89qNjRY3Hj2SAlp/5w2C1mqqwPeXVRVuBKrfdHN4FINQpyC9UHfN1vkaKYDj71oE5BEAesI8ei2SxCgJJroREfUfDOueJueL3PDqDfG9eyLP38FglP9mv4D+AkBdLL54hdfdqVqb+TI/D7J1gTCIcCZHAAhaPEgRbOHnzPQSb7YYDq2iulJZOzxXImvQlJdBHMBpdvL+R2k6GMukJrzqcWFcDoguzQkfP8TdGGBUTPzr+S9S3V2BN4l78smvDDlAH5zqWWbrnO1oi3Un1SU9mkWqRXRgHStyC7Jm88M4YHGjXUXLcy7nRa62Mudk1uYVvk9k/orqjtk8YgcAt3g51tkW8yDg4e/upM0hHDMNDgFS9TsAcFeIF0QGlKctzNKsIEIIOZouVK0hVY/1uhB3TPzC6FLQs6Np1XwNGBYsE+o3czY3SlXBddyF/1+tgjYR34h4hROe54bnsI87Eeq9AkBExH9E43WURAqsXXT+O5LTGQxV36IYuEIio6ICGG0u2aipHD/bVM3VIVEE3wqVpeFZkT5w1jMAqIM7LmnDoWBtcR5w8onvUatrBq46qQMFgFCQB1yKJQ6xc6N1b0pTXm552R0MY8wAgE8D6MZFAqaPO47zMWNMK4C/BzAEYAzAWx3HSS+7bhDAUQB/6DjOn176u3cA+AgunlsXALzbcRwNPodFcZWKfnFvJz+NBsL8MM7W+CklbBwAwJ5OflBtb/CsgLpvxMsjmiqqC+hI+7hgR7nLz43ZfnFSqz5gAPBYSlAztvJIcqu4b19IFzYycWtguCXB38950fl3byfHHddszujk1gU9L5oxfuO8cs74HFD9RYZj3LhOu2TE14iO7R0B7tltTPJD/D2b+NzJ5HSh8kSODzgl6mKUEz5e5Gu9LPqL3NGts6AqwSGzoKJ4/uZ2vkZ6Q9pZXJUQzdLa+LwTyWXsGeOMcaqfR05k3ABgXvSWOLjE18+auChkFwRl93aLNvAA7ujk7y4Z5LruXq6LrBGTQOHZANRT3Gg/fYjDV48v8edsCHjJ80scRtgd0sa16s+yX/R1sSyecdoY4jVVwegr5HhUv6VjZW7CZEIcJjZbOUJ1ubLokCqgpAAQDnCkQK6os73Xo1xnnR+kXIkMRh3A/+k4zjPGmBiAp40x3wPwXgAPOo7zUWPM7wL4XQC/s+y6PwPw7R//wRjjBfAxABsdx5k3xvwJgN8A8If6522KR9xf+wa96v0WzwpEfdzIGS/yzUsVgANAuzCeyg2+KWRqXLckmkApQwYAFip8PNtbeORJmbojwnhUhw0A3NfPDcF2P3+3W1p5pLCjlTsCsX7+/J6YNuhll16Lj8epcGOusSDmR0qP58IFDhkYinDjcl44A1nB06+ixbd26PTPQJg7A8M9HOuc2CigRR187nQKGk0AWFPgcK+6+Calaf5N0vM80n5kQRhrOR1l7g3w+aPaj9zZxZU72rmR07dWd9oLrOPPaZICEimIGXZkRNO7BT4/Khd0MCF9ns+R4RT/JseEgzoc42ukW3wrABgd5JH/+DZ+lFt93AGDVzQ6Fe8VAPwJPt4N/dxZXDfP2aBK0/z97Fng73VsUdOEd/j5OmkIxr0uEWtYExUMZS5n6cE0f++1Cs8q9TR4jVfOx+twkn4e/GlA73dlm69pTejdlOtdXnYHw3GcaQDTl/47Z4w5BqAPwH0A7rz0zz4F4GFccjCMMW8EcBbA8l3LXPpfxBizACAOYGUd3S7JOt/eFV23VOWb10yZG/sfvE17/6FV/FqjaHlU5MkNl6VE3VccVEqMuqdLBE3+pkjRq990atwhdEQ/goYoCAWA6hi3zC0BBfd28rnlG+WHqrdfQ2f6PfxhxrPcWbxJdJyeK/LTeHOfgEXs0Kk8q1NAlhz+8pwlnlmc4LEEhEL6wG3dysfrH+XRWf8GUb9yjL/XqJg7PbM66p0TBfu7N3GceGgT/5ZWh2gTX9V1U3aKG57T/5PrCiW+DoZv5Yawdx03rsND2qC39gtHqsTH2hnm4ymIjPaO27mBCAC+rYJ6WjSctKd4AOPoAzzQEPa71NMIh6hzK19DPrFGfNv4+4mM8TXSPqEZr1LjfM6+cp3YJ2b5fD6V5fcsigAgoDOLm3yctvybxX+gun7fTqo7lOZNct36TgT8PIMRCQ1RnYKJVWs8S2MEjBAAnKu8D0azyPsFuaI1GMaYIQA7AOwH0HXJ+YDjONPGmM5L/yaCi47GKwF8+MfXOo5TM8Z8EMAhXHQ8TgH49RczngGLH0Y1UbhzXKRY93Zw79+lPhPlCb6BVwWcyRfkBpDx8EPDZV3DI6I56lojGGCMKBpWOkDzTTsCJ1Zb5O8gNyvoFaf5Af/DlMbt39zGjY6SgDO1CxayDf3caE/u1O8usp1/zFeOcs54R4S9R6vckFEZAxPTUXhbdCk+8S2BS2/nc6Au3vl3TvN+AwBwY5Ybnl0dPKaX2C3qITbxiHjbCDfWWssa620i3MEwcUFLLdjLnAvCYP2qLlTu7eJ72rjI1Cjx7+f37FrgRntwl/694O38/azbxdelU+MGkPGITSuh+f8hMsgL3+ABA0vs+bEgd16/dE6PZ1WYj2dHha+RtjE+1phwMDwi0BCK6aBAfy/ff6szfP6EFoWjZPH3GlbfGcB8WfRRqfJvstnDoVcL4OeBEreC81KF93AKBrgztNKGeMaFJvxqRyA1i7xfkCvmYBhjogC+BOA3HcfJigZM/w7AnzmOk1/+b4wxPgAfxEUH5SyA/wrg9wD88WV+69cA/DYAGe5LiKJh1a87LDZwv9hoShkdOZgQ0clvXuC6C0U+nieLPGp5V1xTbG5I8Gc5meMG22aRSo96+T19gj0H0HUEtRVGEZQjea7AjS439pwfprjhpQpGPzrJi/l/4Hkd1dkHdMFocgs/xDxdPCtg+Vw4XIk4FX7Y1M/qsZYn+buNiILao5O8f0Sqwp0at3r8Dx/gGZ6PbOYOz2iDGwAt27lj4unnUWbTqh1bKTVBQDHNAyOVU9xYi/h1JvPUFMetz5T5GkmJeogfpvj3eGONv/PVYR4RBwDfWgG7ifGxyp1HAbQzirAaaFzgcyQQ4fvo1CR/jtkCnz/pqt5DVeT7L/fx+/77zTzyvzrHv0nrCN8nvILlEQAaoiC9tMDnVklknMoiSDEuzgoASFf5tz5tOCFG1PAg6JLDsziWYIi0bV0Ep3pWKLrZlfYCaTJsvTgxxtwH4I9w0Wit46KN/dgl3WtwsbTAA+B/OI7z0Z/hfnnHcaKXEgLHAJzAxW2uAOB9juOcYNdeEQfjknPwJQCfdRzny5f+etYY03Mpe9ED4Mehp5sAvOVSjUULANsYU8bFrAccxzlz6Z5fwMW6jf+fOI7zcQAfv/Tv6A6vEDcVYXiGhIORFUWGZ6Z1BO1snm/SR5f4hhkTTC5bRcSh4lJ0/pjI4D9R5ZviF5f4wThkc4aggaA2nhTdoSpgzYqCANV5vUVABtyyUceW+CauOkN/sPtdVPctgVcuCqMCANae5O8gI+p0ooLNJyScaRXRO5PXjvbpHP/NbsGe87TIfDyQ/TOqe1Pyt+R4AqJj+X85xrfU3SmOg44e4c+4NcHTlfsXBVwJwI4W7gwcExCPBQF16hEZ0kdntXc2VeHPsjbC393JAnd4/MLQnSlxJ3PdNHd2AKAkcPQZYXyrNbIuxjMfZwvckQSAQ2luXN4m0FOHREG6qDXGmOisDgA/rPDgx0iQN3T927Oc3rX9Ap+TU4/zM09F/QFgIMQdf0V8ptjm/r/U/011a1vvl+MZL/Cu7bEgP6OnSk9QXa3O641Uw0A3UU5Epcqz3QrqdLXDnFYqjnNVFHk/COBrjuM4xpitAL4AYL0xxgPgL3ERDTQF4IAx5muO4xz9J9z7jOM42wHAGPMBXCRZeg/7x1eCRcoA+ASAY47j/Odlqq/h4kA/eun/vwoAjuPcvuzaPwSQdxznL4wxvQA2GmM6HMdJ4eJL41buzyBFsdsuVLlRMSv24VxccKmLFCsAjIkoSE9Y0TbyCMARh5ep3GBrLvEFtYmLTTpX50WxUz4eCVzr0VSrfuFFHMlyp0YVpOdF+PrpJZ7a7/ZrQ0/RfiZEVGqmyOdkXkSgR+LaaP/TKRp0wEBjkOp2ioi5Ih4YExHETUntnalyoyfmOURINea6PfYBqpt1oZgMiCZ0bX4+n784yyOMN0S5UXEozd95v0sC44EZvv88l+YOWFVEESVNrWoQASBd4xHqMcGk1eblWTWfKABXjHonMnreqaCBgsfkhe30N6kzVPeWpKZaHY7x9/OJCR79WR/kjpTaQ1pEU00AuMF5I9X1CnpgRYf+6CI3krs93AE7Z+nyy1yJw73CDl8jp6yDVLcj+T6qW3A4rAgAWsOcnSpT5nTf9YaCXCuD/qVpXuf1Ckawht5Hm/LSiOM4y198BC+gym4EcNpxnLMAYIz5PC7WPv+Eg2GMGQbwOVz0D74jfioOQKaBr0QG41YAvwTgkDHmuUt/9xFcdCy+YIx5P4AJAL+obuI4zgVjzL8D8IgxpgZgHBeZqFYsdQFz+dEsjzypg+h4hp9wr+7TeOWOgHB4KvxHj9k8Ir7F4t2oj4loBAAcLn6T6kI+HvH0e/nBMJb9IdUdbWgsfL/o8OwVHs/+JW541gw3rFoER/tUVbPnqPsqyYjiTdXw7FRGF3mPOENUVwE3Or6c4QfuRmc91QVF0b1q+OYmYdG35IYQnz+qE3PDBe42LbqSP1jh+/Fe36uoTgGLBDu0K8VvUpS3bEpwI/DgEjcO5mo8opJwcbTva+PRa2W0K4a72RLfJ8+JKHxDFXEBGAzz96P6TlQa/DnuS/CM7ZL60NCR9p1RnqlZUEx0wti/UNEZjFYvP78Kdf6bzzrP8JuKZ3xi8ftU1x7fxS8EMN84SXV1ARHyCmjRiHcT1dkWDxgAwNna41QX9PHss6qHuBKiYFDK4QkHeRDLq4o9AWQLnI73you5Koq8jTH3A/h/AHQCuPfSX/cBWO69TuEiQuin5WMA/rvjOJ82xvx0XfPqS3Z7DECYXP+PciVYpB4Dh6ne7XLtH/7Un/8KAO829k8Uv4iEbWnlryorbMdBkWlo9WvD6kcpHrGKiuBSj80b9ZwBdyI8LtNBRV28hh82MfDDz4qIhkz178nxVAxn/Vrr579ZE0b78Tp3eNr966guXecNkABgyPAD0BH1PccMPxhzDf4t53LPUR0AdMU540i7xXtW1B2eiTkI/pt3em+gukVRvOom42I+h2p8HTwjGv/t8elIctnm80c13wqLwulDosnaUJA70oGgrnk4l+Nza67MHaVpi9dq9dm8VktBmQDA7+HGU1UY5ioxomA+B7GP6iIeDZFqq3MD8rEFns3s8nIn6+u5v6G6gDAsAcAW9Mn3x99BdeUGP2embM7mM2Y/LcdTLvI5u9v/BqrrNnx/OVV/jOoGkvdQnSVddMDydFPd2TSnlGuLbac6j+gfUTAuTRP9fG9StLGqAPpqa7SnYFDKicgWeHb9WhD7pSnybjfGPLXszx+/BP2/rDiO8xUAXzHG7MXFeox7cHm7+3KT5lYAb770358B8B+X6ZZDpN6Gi6UHr2HjaHbyXiYhgb8fCvNNOijw5YU634TiPo1DvKtrZbj1hohAZwTTxNklwd0JvdnOFp6iulUJ7jf6BEdrUBSAu8lUlUOkOrwc4tFjcViWx+HvfK3ZI8ezZPhBXgI3VhIOPxjTDndquuPcoAeAfvBIag3c8AyDZ6qSNjfYHq4coLpVLtC8oOF1TLdGeSSsLAxWb51/y7NlwUcMYCTIDcFO5w6qUxBDn9iKZyrcqesMCY5jAHHRWdxn8ffqL3MjMCfgU8bRht7JHIdlPZj9b1S3tuVeqpsT9V97/ZwIIe7Tx5/KHp4zh6gu2thNdbeH3il/c6VSaQhHssH3Qp+oX1ntkcFJpLwcyvPD9J9T3SsS/wfVtfh4hms88xDVsd5WP5aAn++j0RCf6x6RwVDZ0y6X7PtJhzvwAYs7qFfCiVipU6PqPoICWhVpuVWOZ3qJBwF/jmXecZzLbiyXsgy/eumPv+A4L1T7O47ziDFmtTGmHRczFssnZj8uNqi+nPwsE+1rAHjEBE0H4ydEORiq6VtApPb74xxqMC+YbADgaFZwgosvpzqS2yI6cmviN+R4QuAGST12J9XFIRgsxFhjktULKIrI3FHwyH+lwSsi52x+nYK4tIX0WM8XuVNz2PDI/7kqL+qr1Lgh/MrIL8vxKHjVUoMf1i0e/u6WwK8bdbjjFvNqrLeCcaQF/CMiMCV7wzwbpd4NoLOH43k+nm8v/SnVbUn+EtWttXi006XUBgmhnynz50wLCGa7j6/niq2DAhEvD7h8eNWHqU5lKR7M8rWVFHUECnIEAKvb+XPe6eNFzDPC1n1qkZ8HHX4NDVHkHUpWWzwoUBaOSaahYZYJhwcpViX4+goYPgf2BnZQ3WSMByIeKUg7B10Rvv/Ewfe0c2VejF3087PUcqFaTXj5mp4uPU91HlGH0hD1GS9GlBOhajDsBid0mMvsX9E9r3q5AkXejuP8JS4WbwMAjDFrcDHT4BhjdgLwA1gAsARg7aUai/MA3g7gchGPH13S/R0AzjID3AaAF5Wh6WD8zKI6c1bKLqc8EdVxG9AHYF50TR4VnZgTpW1U1xbSz5ESkIpZwzMjx21egGcLeNCe+p1yPCFhrCRqPJqeAI8Q9Xg4LEJBDRqOXkoNsUmHDP9ebf5bqG7eP0F1bgfcmMOhRQMW5//PCaOjCh7xzRrR87UhOg0DqDn8vVds/t4Lgsb48QaH372vjUfLL96X607XOKHBrpb3U13S5nNgXZKvy4CLzfn5C3xdvqqdG1YTdV671+vhBkBbQK+DmOiJMymcs9Vxft+NwlibKvGMyWhcV8gviF4FioxW1fLNWNNUV6pqVsFRD9eXhKMQEvVPPeJ7Hc2fk+O5IcAhszlRPF4VTujhMt+XVvv4fL07+i+oDgDOGN7U1hb7S2+IOzwLDi9IH68+KcfjsYWTXlPwqpXXq61UFMXtSqlolbwU97zO5M0AfvlSbXIJwNscx3EA1I0xvwHgAVykqf2k4ziXK2j5EIDPGWM+hItsr8vlxzUYBkAVgFx4TQdjmSjPs8XHF3arn1scR0UTvqILU5vqM3dYNABaK8KWZwv8unaRoQCA9Qn+LIEcP+S9Fp9m0/Zxqnug8Ck5nvvinMWjz+KH8azNI/+jFseXfy37Car73xOclQgAyg3+bjNV7gx5hOOy2fDDL+UIgx7A9hB/TlnjLIKarRafHzZ45E05bgAwKYyyPcE1VBcVxuwQuBMRcOkgP1Pl4z3X4IbFsOdGqiuAw6BshxvCilobAN7QxWuRFMPS5jCfk2cLPDKpivkBHRRYm+BzXc3JMzVO2nCw/HWqi3h41gjQ2Ra/qKf5TpYbsxstTlWsoGeAjiQPRvl4QoJtLytY80bNkByPX6yTjiD/lpMFDhU8WX2E6oY9b6G6oKCOBoCJHM9ERIOcYWoxx6FwA0kO/XUrVC4LJyIa4ONZKrwooszLis+rHVvdMG9l7FRuzfSUXF2VJj8pDq58J2/Hcf4jfrJuYrnuWwC+5XL9OQA3L/urj176+zEA/yS+46aDsUxUcl81bqva/FCNiB4ZY3k9EVdz0iKMiIieYlXZFOfR+2kXj6dmbyitAAAgAElEQVQ/wh2XjUluQPsyHP+aMzwKdGOEF3EDQEJUfo4VuSXsE7UUCrL1jrZfpbqoS9fxbtEIqr/KjcB9i/wgqouxRqAZylQPkYTA7VdFxqAmIpOdIjt2JCfa0gMYMbzOoifM157qT/Pvz/011X2oXzuLcdFM7o3WfVSn/JbJIncwRCkJyi4BTUWh3RXiNy6IXjHdAX7GKEMX0ONdqAgDWiQbdkT4+jlnczafnrAOqLQH+AdLlfn7UZF9te5ConEbAMyUuTHXHuTfpDfEx2qLmpnhmFtGW9R9VPhY5wx3CFXxc02wfikYJaAZjTp8HHpVDfG9aTL9INW5wXw8IhhTrPD6jJdCVP+Ml0qUs6wa+126uCnXiDQdjGWSF4eq6oNRFawBx7JcpxwIADiU5itJRWdHE1w3rWvhpJzM8Aibwq0r3d4Ax8a2B7XR/oM0T6fvjnKnZn+Bb6ibQgr+IXDpAb3rBWW3d37f9w3xyP/Ti3wj7naJMyjmsznhoGZr3HGbEYZDpMYNvb3tHCMOABHRuGyqIIzSJNcpJ6JN2524vZ2/g9MFfvEc9yEQ9/EPdiLDjbW2gDYCowJnqfYCReO7tZXf083h2d7Cgxh+i8875ZxtS/Dr1mZfS3UT2q+VDfM48Ax4vMr7VsVtHi1e4+PQRAC4p4fPkWNLfKw3DfEgxc6kcNBdSEg84nulS6orO4c67V/gxdGq0enprN5/h8xtVNcqavKsIC/YLwS5I9nt8MJxAOgTdTGqBCzX4Bt33uELOmPxM2+mrmlfCxU+2+uCQMAyHFrVEEX5bgX7V7tc6QzG1SRNB2OZlEQloUgYoDPAN2Ij0v5ujni7oKAUvhB+5/gfUd3fbvt9qluq6OjjOuG4KONJ1ZIcSfMLb+nQlp7f4qnkLtFt+O4udZCLBmNBbujF/bogsivJizu9An6Xy/KD+vZuYejV9dKeyPFM1rEcPxiemOMG7W1hDrtSNfDKcAC08fD2Vfww6gxz/P1btvCCSJ9gjAOAxRkeTg8vcMMh2s7nVqrM3/nuVr4uT+V1tE+RU5REauTGNq7zGb7fbUhoaN5At4CG9PP7Kjrr8nm+1tdM8ShOQ2SeAWA6z79zSDDRtYqgicouuwVUdif5fH7nKHfuW4f4Husb4k6LcWma6OT5nlc+w7/zY89xJ+I9q3nDwCVBiuIxugt6W5AXpE+I2h+fcARU9lTwA1y6L9ed558ZJcF+V3M4o9xUge/3VZ826Os2/86qk3cD3INXGaVrXeymg/GP8vP7lVcgipVoqcZfVU9QFPCKyHW3MFgB4M4efljHovzQ+Ne7OYPQUpbTpW4XvT4AoNzgep+IZjVEhue+Pn6IKepFANjdxt+Binj2dvIajOig4O5u5w6PiWpnyAhYiRpsmwCfO6KQsjGtC+V6T/F30DjFHYWwhx9wJ4VtqTIqayLaObujgx+Aa1fx+Rxdx9+rlXCxAISE1/M50lcRB26GX1eYEk7EFK+HWBPTztnRDDewOwLc4fniBJ/Pf7CDP2PfNp0W8I7w8ZgWQbkr6iGi6/lzRHKitmVWdxruVBP6DM+Qbo7zebdvgT/jSERnDNZ2cCei8ya+F3iGeIQeiqq35IKvV3CvHXx93bOBr1lH1AiWzmj6aCUTBf7e18e4tV8Xy2tdjM8fNyPzQom/nx/O8B/tCfP91xEwsYzICkwUea8YAKjWuNO3UtHUt035eZGmg7FMbLFAl6or80rVBjXcojfM7tUi6t3GN0Uj6gGSdX5oOmqwLmKURb8ydkXtJUA/p9SFhDMQ4pFJGdFzGavi2XSETv2mifJDyuvCCBac5ynzjiA3ypSuxccP8eEoD8ttGFGAEyA8xHWeNlFrIr6JU+BOTSOrDT0rzL+JleTjsdoE1XWOr8vurOiqLShaAWBzC8/UKCjGn/8C73EQXMUXtKdXR5LlGhLOgHKmjU+skQT3bD0ujFfWGc6ktbaTr59TcxwGtTHO592o+FYA0LmVX+vpFZh/0c/DPs/PoNIJ7WDkF/j+034D39M83aLbe1L0TijzoEl8Uo/11lb+nOcX+ZztjHOHuVjhe+x4VuOfVWaxW6R7FYz7YIW1NACKHr6/1Gp6/3UrAqf3FbUdtguhwTUrjoHz0jTauyal6WAsE5/o5L0qoihKBTZfMExFYytjYACA+oIwSsVX1TqXhaEy5sKJkPcVNDhutV7SQhLOop3jB7UtqmLtEr+noPwGAOTnuVOTy3OjtEUYHbFVoqN0v27AFtzED/kbhkX2Q2SVdpV5ZNISjVuMKFAFtANmq4jnKf5+Tp3lWYFDaW0k72rnhufgANeFR/k7CG3h32Ngk+izI7opA5AtsDUERjgu6nvM6YWwdJDr8oJxb980L+S+sYvPu/7NnDbYNyIMXQDRW/k8iAqHpyvPxyPFBZJk/HxNNya4AZ0/yefPqXHdzVxJTUDM7P18XiYH+RrxDwoYVAefH+vu0fPOEZRprQs8Qp+fFg0ws3z+PDCjSTbWiczjeUFQoorZywKSNFvhdRZer66B0yxSXJRjciUKy5vy8kvTwVgmulCZX1cWG21fiBusxaKG1aSO8w3sVIYffkXB///VKf4gmwXTEQDMCaimKlg/v8KarcGwzqg8K/YohY9Vogqcny1yY+WmGIdMAICoz5TyydSzVPcHg7yniVuX+E2t3ADoGeQR80CXyAwJdqX6HHemy7M6mjU7zSfX6SVueCrHf/8iN2QUTh4A9s3zg3PbLNf1H+HPuTrBHcnBYW6QBYdctnDRiLA6y3XZab43Hb7AITcHMy70nKK8Je7j6z1b49/yL/fxb/nmCb4ub27TGYO+Fh71bV/FM3Iqu+xU+TNWUhoqmE1xo/WpCxzW+PAc/5ZDwo9027cVnfPkaX7jW87w6xYEUmB3kg+oL6odjLBoiles8kzEGXHOqveq7AUA2Jfi80DRqD+x9Ff6xkQUM5Mj2LlejKhKtnCQMwPWG6KgE0C1xjM1V1ou0tRe6VFcPdJ0MJaJapB0vqScCH7P86UVWpYu1z6W4p9OUToerJ+guh/NaazlOof3XTiY5dvJYJAbiE/UOOPKtux6OZ5kgDtEBQH3mpcsSdwwPw8+1q/lz1IdAKzO8sLPvOGH424vv+4L49wguUlE+wDgQpkb5skUb3xXEenfxSpfI71B/l6Vg+4mJ3P8WkXvagTSV3WNBgCVAf/+NH/O3aKG53SBR5J3FHnk2hzRp9lkkc+DoEd0cRZNQMviBfxwRhvJO9q4M3BCIEZVQfoNLTwCe1r4ELmazlT55rj+5gXuYDyV5t9L+HsYcCEXUGyFx7N8HZzN8m8yJUgC3Dp5qwaYPSLb8sUJ7igMhvl1S1V+HvpTOmP7zUUO+bsjzslClKNwWrzXr6Uv24bgH6Wn5Vaqa6wwY6DkpXIilKiu434PN1KKZd5A9lqQZpH3C9J0MJZJRWyYfz33KNX9hyHebXlAMNmMi8IzAEgJVqe1IsWqjMCbAuuobqLEOcgBwCOiIAnR7fN0mUdgXxHlnbMjLpCtojiP0xW+oR4QTlYFPGpZaXBr5a7QHXwwAMICCvZoeYrqAh4eLa47fH6ktW2AuTJ/t0HRREwdqsdxhuo6G/wQXy86zwNASTiLDbFmT4qu2puDPLL9qZn/IMfzCy2/TXUJH4+GPirgdyMxbjwV6nyfCOqko2SbUz1NDi7y73zG5o0PRz2cjhgAsiKCnxNN3yKCllsZgUlBLe0x2jkr1vm1nznHv1eLSEwrVOeJjEtfpDj/2G2CJjvUyvfmI2KjmDa8mB8AouCBo3113hBus8Uzr2UBwXxykZ+lj+b/X6oDgJ7YDVT3tSyvQWh3+Jl41j5Adb0tuodTm1lFdXlF9x0aorpCaUz+5kpFF2QLEhJRWF6o8L25I86pgQEglX1C6pty9UjTwVgmZYdvtm9K3E51qsFYyMujEWvjmsVkaYEbXidFfw0F5Z0u8/Rj1ujxKJl2+OY/ZPFNeqbEU9eLDcHXB6DHxyFkIWHQrwfnLy+K6NHT4KnZqgqXA1io8vd+c5BnahTmVukezI7J8Wzz8RT1sQp/72qOJMAzHxujfC6rxn6AzkQo3WYPdyKyohu3ciAA4IHcf6O6dXHeIbzT4eM5LBy3x/I8tL83yr8joN+PsDuxu507StY8dyLmahpXE6rwyGVccHeKHp+YE7VRPovrAi44FsXKExdz9sEsj8CWRbZy0OZG50U9f3cKelYUDnqb6Li9v8ij/gCgUPRt4M8y7/BATbHCHbfj2E9198Q+KEYDZCCIEsADPNPmNNVlS/w7b4nwhpsAYESDQxi+j15wgQ+93CLpZkVAst7gEN28YOK7FqTZB+MFaToYy6Rg+OHYHeIGUqkhjH0v3/nd0pZxcW3Czz/dP8zzDXPQ4hFxr63DoQr/eUfiX1HdU+DZH8uortoam2/VeFTqXINH0PZ4eITEAh/PHRY3HpVDAwDzdVFwXOKb7UiQz7ulBp+vdUsTCCjnZCjMDZnnS9w46LP4waiyUfNl7Zx9r8xpFG+w+Bz4TuFzVPf2JKdyPlbU2HzlRJzMfofqBmO/RnU9QdHZt86/h8rUAcCA6KytpqyC43QKlpt4w6VwWmQilDMkavllo0rV2yhV1lCUiQaPJK/2cYjhRi8PqKh1t7pFM79VBYT3k/Nfp7r1Fm8yNxzg+0uHGZbj8TqCuEEZ0EIsES0fBV/rWRHgAoCsxffYs+lvUF1SZNjXRO+hujR4hN7tN7sTN8trX25RWQqIgFzQzwMqtQZ3+EoVbr805dqSpoOxTDIWZ//wW4L3PLGyKuZcVRd5//k5Di26t5MfcIpN4lnnGarrc9bK8dya+A2qe7rKN8yQjxe+tlo80qUivgAwa/GMQqczRHUHGs9TXb/Nsxtp0Q11r093bh31CsaRIk+1lyu8eLNq8ah33WjnbL7GHRBLFNSerD5CdR7/K6jOV+BzIFPXY73dfxPVdYe4Af0vIu+hOmWYtxgNXSwLhqVkZA3VqShqtsQNpFMOhwTc6Xs11QHA2Rw3DkZFE9DpIn8/88IwbxcRcQD4YY5HfQfAG2DWRDCmO8hTMWMlAXk0GkfYbXhtR1A4NadLPOO0LcbvKQ05AF6Br7rVy+eBIi9RNWcp55wcz3bwdekV0et9jQeprlzn765S4+fh6ijfewBgwR6numiI793ZEs/itPhElqZySo4noIxvAS1Sje2uhKgMRrXOv9e13q2biYNmDcZyaToYy2SpzjeTUmMz1ZUbfDM9lub85EHRnA4AfnOEGzpjgjRj0OGwiSPghdxPFT4vx3NL5N1UtzrIIWTHCw9QnREGouPyftI2/179hneLLYAfYj/KfZzqdsT585ddIFLKINng8ELu58Cj9zNLXBcJDsnxBIJ8bsVsHtXcKgxan4hofr/CHdBbvK+lOgCIC8yfggOKpBHmKtzBmgc/GAEgafM1HTc8WhwGDyjMWDxavsHmBaFBj44Uq+j+vnm+iSQ8iiFHNDB0OVtvDPF1qWptnq9xx8RT5saaykrHHZ1tafHzjILKjLRZKuPEs9LP5jTl8G2tPEPYH+Hf62/nP0V1o6G7qW4QG+V4AhZffJkGX1+9Hp4VmLM4JClX5Eb7XP0k1QHaaM+XOEGHKsbucPhc9ov9FQCmK4eozhJGezDA4Yll0QdDiWKYAjTSwhEZjEZjZcXqqjj84n2vYopbpwmRWi5NB2OZzOd4dL/Y/gtUd6bAI2hzZb54W/3aKFVFiDlRgKjSxRWHR1HXxzj0AwAKDt+kfQ4/4NZHdJSVyZHcV6X+9fH/jeoiovOvVeUO2LSIZk2DH3CrqjyjBGjqOkUukKvygtq+hC4kVHI4/Vmq25B8K9WtxhDVhUVx+K76q6jue4VPUx0AvNn3K1RXFpTMP6py7vcRUYcTdDTz25jF58EtPl7AaosIdarKt+JRUQSvIEcAsFQR9Qmiec0Zm0dK4zYv7p3N6z1tS4w7Z1VRLxEVv1kGN2QqFsesW45uhjZT4ftdZ4BD2kLGpfKeyJTFyScAoFznZCIK7rYrwOsBwg4/uyIebR6UG3zfmvBwo90nfnOpxDMNqsDZTeoCTqokU+ZBrKEAX+tREaQBgHKVBxRqFnf8V+pEWBbf0yJBniUHgHyRO31uWbeVSCKsa5EWc1exg9GUn5Cmg7FMoiFudCjs7HyF7+6nMvzw2+TSd+L7s9wZWBPh0bd9GV6Euqvl/VR3pvq4HE9VdJNT1HI7ku+jOo+Iem+IvV6OR8EmvBa/b0AYABu8d1FdQzB7jzd0N9RFwx2FLRbPYOwNvJnqnnEeozrlmABAKMAPFQWverz2barbab+S6pRBdqvzDqoDgPMVvg6qwriMWtyYPdDgWbWbPNohDtgaSshEOZlJ8KjdhSKPBu9o087QhQJ/P50CWvRsmUeE0xbfexLQsMZ0hRv1imFp0nCK6H5wkoSAw+edmpMA0C4oulSm5iFR5L0jwKPebwhyBwLQzqTq3VI0PODULrItrS6dzr+S4/VGvYbvad0OrwMcDL2d6hTE8ECO11sBQFxkzlT1Rq3OfzMU4OvHKxj+ACDg41C54gqdCCUKkhTwakfbDnGDf6XMVer8iXs1E90ieG+oq0FefkLgq1eaDsYyWRPg8IbxPDcud7bxzWRbK9cdz2jvfyAoGloI+ZWef0N1ok4QfeYN8r5zjiCqF2e1rdhYHP6Medk6HDho+Eazp84LuRWsZMDDI0/pOsdsnzfnqQ4A4uAZDq+wVi4Ito1CjdcMrQ9w6AMAZEU9ScPhDkafdzvVHbY5ZOvGGs+2dAdcOnmr4t86X5dxcYjfHHkj1dXUIgFwssCLwFNV7gy0+niWb0OcQyoGxDbQHdAwhOEIn+un8wLTL76XR8zX/fWn5XgSAZ49/PiFP6G6vYIlaEHAy0YtHZ1VopwIVcOzSjhZedEBHC77XUXAMEOCRGGDwPsPxfhvqt4jAPCLCR4AmhPsgKvjHHqWFhk3uyh6ocTeSXUAUBOEIecT3LlXGQwVqAoL8hIA6AiMUl3Vz0kCppYelvdlomBQ2ZIuqg77OTRvpeL18D2/KpASTbm2pOlgLJNBwyMr/RG+mQyEVtZEbEE0JgM0D3tGHHDqsFGHZtynp8NQkBvJk3m+ge+3eWakx/Bo8LDoAQEAa703Up06yAeifPPvFrbuXJm/H1+GR8gAYMzwg0plxzrAHZ5OPzeSDxkO9wOAhCigX29xI3CxziEnm/38HfQLNiPVG+Gins/1rYJr9UyWr0tBhIROQZd6UbhB8lSJO5qeGm+md6bGjeRb2gVJguhQDADjRf5+lP2osoPtAe4o/Xo7X5MAEPfyH/2zxL+muqx4zFxNdHoXF6pO7wAwKxrtLNV4sOGUgNDtNLyu4ZHKc3I82y1eBziaEBDVOH/OJ+f5M3aH9TqICRt6WOyxS6Jbt2p2+1D+E1TnBu81Ag7YKnpSzFR40CQb5HvhKr+uIxit8qzbuOEGf0tkAx+PgDJ5PHyN1BucCAEASrWVN0JlouppCqIv1NUvplmDsUyuQwfD0MYxqsHYloSIWEVW5nG3+gURPYDZMt+kz5f4WHXfAK6sqRbF0A2kfBYf62KaR71jFk99pOqikh3A7hg3viPCSIwLqEHcx7+zolA8uKQjyXtCnF1ossTnT4eYIyp6/wr/Hjkev/A0VXO2iI9H2lcJJ0I1AiuLZmgAMC8aTqo+BwqCqBztR2c15eX2Vv4OdoFHH7tCfE52CLhOwrdy7vuCqFFRMmvzbGW3wx2lDpe6sphPd6tmMiXoox+Z499rfYx/qyNZ3fdndyuHgvXYPBIxYm+lunnRsGKPj18H6JqQjiB/Px0B/pu3dvI1ctalLZKCZUVFUcgF4bidFTTYNwmSkYqjabkLhs9nL/geO5zktZcBW2cplEwZTmO7KBivbAXAEcXhiu0pHOBNUAGgUNJsYv/s8hJ0Mm/KlZHr0MFwaGFSQ0RPIl6+sD3CyCnX+Qa+KAo7AcASRd5qc+8RkaeQMKyeTGnaxi0iWpwUhsXmOHcECqJ774aQdsBUn6x9BY6Dvi/II1Zl0dNEYSs3JzQjjZItfm4EnReHsW7wqA3LCwX+rb8jIoWvifIank5hQCunpSAydYCGo/gt/pub4jx6rTKLW5IaspURGRflvCrklZpbabFPFASDHaDrGvrEY75O9HlI+PiDLLlEOw9llJPFr2sR+8vdXXz9yOcP6zU7I/w69ZR5MdmrolO1YksDgK0tfLxqLwx6+G/2Cl1EBNwAIOwVdW7iTFwT529vqcq/Za7B96yM0cxvZwucXnswyvtOnM89SXXrIjwbla7qszQgMMVh0U+o4HBYrKR+9QgosmDRAgCfl2dQGzYPAhrDF7RHBBbrDQHFBvAS1JX/s0mTpvYn5Tp0MLgoo8xjuFf91CJPPyaEY/L0oj6MFVynPcgn8bmcwqXz31RsRgCwIPCxShQe96Sgmt3r1b0lRA0mkjbH3z+U4ofR2rDIigjomTJkACBV4u92o4i0twa4Ll3lusm8jgINxXg66i7nvfJaJoKBE6dFFv7BIqdsBIB3t3FoSFeQry8FgcmJyL4b1eozeV7Q/+YebpjvmxMGUpVHQwthvk1PFfSaPJrnNTxv7uNF8Aoldr60MicK0E7EsQz/lvd08xvXxWBPZvlYH8jzZpwAkLR5pmZnlH9nlalqD6ozRg5H1gFuEEb7hRKfW88KQp6xos7M/+pq/k2OZvlvKnbEbSKIdTrL9ztPVZMLnBERfEe491VR5zZheBCrz9a1Pz6RkUuKfjATVU7sMZTk5BRjaU5qoQquAaAi4Ju2zTNH0RDP5up3zqn0rwVpQqRekKaDsUyUYREWkZ5Ncb4Rq+jjI3luXAPAW0QvAwV1UmnmV8e5UTFf1o3/FOb2Y1N/TXW3hd9FdVHDnTPVuRYA4n7+wbYm+H0HRdHsrIhajsb4HFBQHQDwiUPVZ3EHzBGwrGfSfG7d2683OdVHpUUUIyvsea9wiCfE772jlTPOALrYNCLWZYufO1nb27jhXW3oyO27hkXfGwGVC3bz75Wq8O81U+LPPxzV3zng4ev98BK/70iM37dbEDqMxjScqzPE9a8fELVsddE9PMCNnNWC4ndXZR3VAUBKQPO+MM2NoD1xXjumsnHJgA443T/Av9dAmOOZeuNc967dfGF6QnpPq+X5ePcWhVM8z7/JIyk+X7eKQMxEQfed6PLzpptfyvA+IW9t/z2qUxmngkrZAnhFKz+EIsIqy9b+JdXNCX9wyuyguglbMyA+U/kk1amGgSoz4ubUNOXnQ5oOxjLxCQ8j6uOHX1eEb9JdAorx6W2aK7sgKPKU7X1rO7f0HuLQT9zSqQ+4/Sm+af7WwAeobklki9sD3BHoCOoDbn+KR/TeMsh1oy3cAWtv4d8y1MIfxM/PRQCAJynwsSIzUl/gzsdr5gUJwLym4PzUcb7BK+jV6/v5fceFE7G7beXkfaMxfuPRfh5hjA3zOeDtEwaJCz2ns8gx/6tO8ZBwTnyTsXk+gWbK/LpzBY0Dv6Wd71uq5irg4WtvayvPAA6s083iAsM8Qm11CciSWCOOyI4NnuMFo6UL2jk7N86hIRtE/dffnePz7v5B/l7b/No5G27h0JG+bXyNeFfzgmPTJcgpXEg/ILqAY4F/lORpvmZHxni0fGGez48jizxjDQDjRR40CXrfS3UdIuO0Mc6fvyekIVIRL9/XKyLAcaHE109Q9C1ZEtnuaFmzVe5NcKdmyeJzctzH4WU5QWvvEXAuAGi4QaiusLhlca8naToYyyQgctRmhcC/L4zxaNav79TYx0iH2MCFL6B6Ct1S4BuNLeoPAOBNFZFmFq+nKrI4KjIZDepN+i1rBX44wa8N9fLBerq4MWclxCEW1tkfBFZWEOiv8IPIlxIO6AENb3j3Os61/h7RQX02ww/5quhwfXs3j5K1d+pq0lCXqH8SjpunX3yvpDJmXeoakiL6mOCGle8k14VjPCI+sMTn5I0iMwYAj0zyAs5dHdwZGhjijkKgX0BVOnRdg+nnRjtEQbZKL5syX+u+CDfIjJcbswAwGuOGcNs4H+u/FYxO37vAoVV3j3KjCwDinEAIVrdwIvoFzWhYYNbEewWg10kXd8B8oiA9FuXOq3+MB4YSCb3fbc7yNXT/CHcISxU+1lyFf+ewC7ubkoksD7qdF0yGqpYtU+W2xKKl61f6HQ7ZMgKKXPXzzLTqqdUW0n2GJgXcqylXlzQdjKY0pSlNaUpTmtKUpjTlRYjjNGswlkvTwVgmKm5ZavBXdXqJR2veMcwxSUGRYgUAI4LiisnNKzKM3giP1rixw4Ua4h8IBIwtin8dEXUxgjMfACwBlbckAZXYAKri/QjmJYhCdgA6b6qKf1aYbw316+tCvRzm4xWd3Xptft32Mo8wmrDI4Hh11LsxzaNd2SP8ORMBwXCS47h9p/4ierGu8Hv54/w3+27mO5MJ6czY20WRqmImMKILel1Ekuuq2AaAT8HPUvy+jlpfKpIu1lY9rb+zr4tvMP07xOYsfnO0zjNVxq8hs3aaR+krz3PYSFCMx85wWFZ1TEO2RH0vgqM8Y2AUW5ZYe984MEx19796jN8TQEuQ71u2IEWZP8Wf49tTHJ2wXUBtAaDuiMZ3Iqs/IxI14zm+RlI2z55mrBl+UwBVi/9o1p7mF4qtsDPE03FRcHIFANCVq1debGVfXGfSdDCWiWjwjEyNvyrFVuMRcJNaUUMxzpzm6ceZIk/RD8X5ZqLG82LEuEA1mDREjYqi6QWAUo0bV2o8mQr3PlpDfDNVX2vGpcjw4RQ3oncl+UGucPJ3jvKtNjIkhyOhYCYmvDMB9ZLbqqDndDIa3qAMANX9+NA3ONRgUcyBB2b0t9yUEH0FBBSsbz3/zv5+8V7V94joWhv5TQSGXn0TWxQql2f1njb5HH93M3m+Rop1rlCjvGcAACAASURBVJsQzQRfv46vkeQGPVYrKd5t9CVYIwIOCQAQfYEE6yfmvsWN6+kFDq0q1HVdQ584Z/Y9yaFg29s5JKd/gI/1dTepfgwuRl1ddAgXfpQtzqe1Ue5hfX9WB00UzfyGFv4sIQHjVnGqec/KHAEAyNncKV4snqG6Wp1/52hoiOoyotFgU64tuQ4dDAuWpXnuLyeZGjdkzuS5biffZ5FKcQMIAJ5Z5FFEhbecEWxQM2W+YR52ieidrnLM8t1tPJqTE70uBAMnvC7crwnBjT8hYP3bkvw5p+f4wZASB5GisAWALlGw/nunePfnD/TyqN3z5zi+Pjal8dNrhvi3DLRx49ITW1njNrvA33lxWt9zaZF/k2yZG3pTwgl/Os0tsl4X9hxFcXs6zdds/nm+Lgd7+WEcTvNvZQlKVABwBLV0Iy+oX4t8Phcz/DnOzIoaCwDfmuYG7Q/mebblPYP8vmrlnZ7h1/UUdO1P1wjXe0SmxrjxzRKxXWjABWwd8+d51vHRKY6hr4ngmFvj1f90lM+Ddw/zZzme5pma5+a5U7NNML9FXOr1KqIOMF/lz5EqCYIFUTieFWceAEwU+WFyXqw9r+Hr/Rx4Xd187STVtfl5E1hAU8rW6jxzpvpg1ISznK1c2w6Gqke93uQ6dDBs2pAmLzaFCyW+sLe18MiTasIngroAgDnhDPSFeCRQNd/K1PjmtaVVGysbHe4ttYgO2BD+nOo67pYVOZvjz9IvYGJzFf6cyolQvQHCLnCuSdEceod/iOrGCqJ7eJq/2Js7dGT78HPc0JsUc11RKKpvqaQvqLF5C4IBRTnaITHWg4vcIGnx621RRRgfTvH3PiwKjp8QhlXHCf5+PuvSZPf+Af4t1T4REvS/P5jhz785qfeQIQHRfJOfFyN/aoI7Wfd18+v+foIHcTa6NMeMzfB38IVx/hw72vh3Dgl2LjfJir17XYzPkZL4zqfEHioeA4DuHv5NHjORiLaigEhNFDl05pszmr1spf2N5kTnddWL6mBWj6dLwEKnGjzYcLj8Xarzefh5sJQ/SnVFvyY7UJkIOPx7BQPcsY0HeI8Mn8J4A8gWjkh9U64eue4cDMsEECadnGsCP50QUOcWH3cwwmFuyHiF4QRoJ0JFUVXX8aCIrj2/qBvtDccEY48IdinD/Jl5Ptb1LdpYSYoD8HRWdUHn32tjkkelhAriHAKgjW+PyNSo6/zipe/T1ObyWvW9/q/Tf0J1v9T5YapL+PlNn19w+86KPlpl8lYI23MJQf3NzGmqW2t4l/jvL3LWprd0cT75Zxf5unPxI3E0y9/tgQUeob+9kx/yOwUpUdUlaJIRnb6Xqvxb7ozzH/32DI+idghjJSBoPQFNFzoY4bqs6PR+tMAdgcMN7S3u9K2murTo2aESwRWxwSzoEgy5T5wo8bne7eHBDUUV/9Q8f3e2C87nkQpv5hkr8yyXJWolDokofNxo3vKM6EpeAz+f3Lpur0QqVV2DoRAftsNhYoXSGNWFfPydt4qAG3B1OxgOTLOT9zK57hwM26nQReqR2Ee+gfXH+EGtbJWyoMADdKOwYoNbu2UVsRLdcp+oHZTjWVhYT3WPVf4n1b0z+W6qiwvDM+3Ckjhf5s/ynEizdjo8EjYnut6eF12T5yui4hFASVTQn7X4hrm1whskRbx8+baI9woAdeFMH8ny+Xxz7FepbklQIRrRuXZ/nUfXAGCkyg2rk4Zfu8vaRnWWANYcLusDd97hhqBPsAvEHW5YKSO5YDhkrc/SkKT+CN8nlBMxKeb6o3M8HXdTu84KpAUMKF0RVJqiwaNyIlLCCCykNbV00eG/GRNFDw+X+V54e+BNVHd3TNNzqmBDTjR26xQwOmXQq6aAAPB8kUPaekSDx8UGnz8e4dTlDb+u3fC1BQATjoDDKY4N8MhRDNzpPes8K8fTZfEmj15hlvm8fL3X6qItuxDzIoqSvV7+nVWX7/nsM1RXEvUZ14I0i7xfkJUBqpvSlKY0pSlNaUpTmtKUpjTlMnLdZTAsK4hI8PKFs1URIjqwwL3SXa2CVm6RY4CfF8VuABAQ9RtpASfIC1x6UaRmcw0duV2weFHxPaG3Ud3hAo+slAUF3mqLw0YAoCbwn0pChn+vgyVOKzzi5ZmPzqAGLJfqPIIfb+ykunHwb5Ksctz+k3Ve1AcALeD42IzF8VWrnVGqe955jupusXdR3a3+TVQHAG0C69xW3E51z9Y4lGmDNUJ1qzyCmQHAavMKqsvUObxhQDA+LVV4pDRhc4hCp8DBA4BAkGFSkHepiPgNbarZooaqnMnzH+0Wa8gW9XEPVb9Ndbd6X011eVunSKvgGZUWH3/vtQKPtMd9PGtSEExHAJCvCTrVGscz1cX8ea7O18iQMyjH43F4tuE4OLtQ2OJnYloUKg/ZnNq0KjINAJAw/CyZqR+jOtsRjU59fD0nPL1yPN02J0U5YTiSoC92A9Vl6/zd5cucRarhsg4aDU4fHQn2U11NsBJUBWGaglZdC9Is8n5BrjsHw7bLyBVPXVZXFDUPm5P8QKnYfKP1CVrYVgGBAoBFwXyRE07E/nmeDvYJFoo15kY5HoUNnbJ5unzW8MOmLtKo56pPyPGs9t3CdR7uDCmMfQ0CGyuMp6gL41VccL8rPvmywzfpWeGcjWcekuPpir+H6nZ7eQfWyRqH8vQ4HMqkIEktosYCAE5l+DcZbRHMTIY7QwpuMpXXxspUldNzKinXuQE9HBcdesVh7EZYtCSuVQdhTADsFaGBoOIHAEwY3q26GxwilAY3cnrFfFXMOksW1wHADuymusdrz1PdqyN8bT1e30912+rcCQeA9gCf62sDHCZ2NM/X7BZR16GcZQCIG87Slne4U7MADl8NGR50S4uO015n5aZMt4c7LgoitcrmxnWbXwec1BkUE8GYbJ3vhYct0elcwI5S2aeoDgBCAf6cRtgT5QpfX8EAd8DKFU1b3pRrR66Ig2GM+SSA1wGYcxxn86W/awXw9wCGAIwBeKvjOGljzCsBfBSAH0AVwG87jvODn7rf1wCM/PheSizjR5AsmLCHG4GdQe4opCt841fnf1b01gCAx1N88T5a5BHqEYtHnp5znqS6ndZNcjw/qj9AdWqjKdV4BsMnCsjW+G6T47EEwk/VGFxw+HimneNU1wJ++E2UdGOlHiMqY4V02txROm04jnVv7APyvkmBIc8Jw2LKOkF1DRHtG3T4t5wuaoNe+W6qoNYr8OWKDUsV3QNAUHTAnDY841Sq8ihzscHnh2K1eqjM5wAA7A3wGp6y8LJqNt/vFoVDvFE4fADwuhbu9M2J3gBlgb+fqvJ3cJPvXqrb7BugOkAHFHrqQ1SXBh9r2eHG/vfKn5bj2eu8i+p2JLmx/6oujpMvikDVt9LaARsSWYHj2W9S3ZrYPVTndbhhPlHjhrDXhXpeGdEDybv5fcHHE/PyoGPIhbb8gRI/h/d4uGMb9/L15bd5ZtEjauCcOM+gA0BDRMBqDe4MhIPcDkkE+dqrC+cDAFJZHXi8kuIAzSLvZXKlajD+FsBrfurvfhfAg47jrAXw4KU/A8A8gNc7jrMFwHsAfGb5RcaYNwHQhOZNaUpTmtKUpjSlKU1pSlNeFrkiGQzHcR4xxgz91F/fB+DOS//9KQAPA/gdx/kJOoYjAILGmIDjOBVjTBTAbwH4NQBf+Fl+23aqKJYvn6YX2VlJC3tsjkeDVYR1OKo93bJolLE3IuAEAs+9ocbTry6BW+z08EjPvspXqM4jmHV6/ZzpJ2DrNHObxWEBCRH1rVd4RG/G8EjYjIBUBB3N3X0UnO1oqT5OdSOePVzn8Oh0UTAPAcCYw3HHaxxeE6EyVTPZA1S3GOfY4XMWx4EDwJ1BDoERpQKYLgqKaJHdcGvw2CHgD7kqx5ePORxWYzW2Ut0z5TGqu8mrYTUzZR59/EH+k1T3v9h78zBNz7rO93u/e1W9tW/dXb2mu7N0ts5KFlbZQoABZBH0iI6IOEfOMDNeOuqMsxydc+FxdA5zBtGIKKLoOArKgQgoSBCSAEnorJ2k972ruvZ633r39z5/VCWpxPp+31BVIdX095Orr3TXr+7nuZ97e577/m3v6P1ZKlPmHa18MFTun/859YdU9raun6SyS4XGdr7BzVDV+goAhSbXyG1M8X5W94zCb2xr/mZZn+9U7qSyK5vvoDIRXVyaCr6iU/tgjJZELpD8O6nscIOvE0NJHl1pd4prQY9HHQFxc88rqezUzDeoTEVC2tXDTauySR0/+rLI33sVoT1Ucy8buAajIqJozZZ0YrtajfvkxVZpwAkqTO1wikerBIBzWL8aDGAhVK1ZYD35YAzHGM8AQIzxTAhhOS+otwP4boxPB1/+NQC/BQid9PdAsc4XzEem+KDpyojY3UXuXNXbQhW4RcRaPzynXpz8Ocoi9OJ8Q4daTQmF10COb3jOFPmH1XyG29U+NvPnsj43d//vVDZT5kNbmbj0Bq66VR8HD059ksoA4MYebrLUkeLO2ger/OU3nOUbARW/HQAmq0eprJbiZix7wdXpV3Rxh+tsgo/lvWEPlbWiKBJzqdj4yryuxTeyNL+bCdz8rjvBk0tNCmf+7Q3+ITNX03byM0K5e1X+bVSmPr2VCdmZef3RfrTM7cTf3ftTVKbWWBVO9f4mDwEdozbNKzT5h9Ulkft/DSX5h16nMCtqF2GMAb25V/7hM2KIPDLJhR0q0QWAhpgH+4p/RWUX57njfXvkbdcQTvcTRW66CQBJ8cHfn+cHGFPz/PBDvUsLNR184XCC+yVuaywfhAYApgI3sTs4/f9RWW9eOMjXxqjshaJS5+vAdFKb5q1rYuv3x4XEetpgSEIIlwP4DQCvW/z3XgC7Yoz/ehltyHPL/gyAXwAgs9+UxMnTGWEf3JHmJ5rvHOabCJVVGwBOFleWs+LMPI/ms72dvxibQX8cqI/Wo1PcP2Owi9uU9kfhKNfzPlmfe6Z/h8qU/WelxjOXJoUt73CenzqpEzJAOzlPxRNUlk5yzUhDON33Ru67AQAhw7Uf+5vcEfWq5q1UVhefpacif4mlm9puv3+eqxb7M7ys+hBWia7UhwMAFBL8o70eeJ/kxMfTw1M8d8Jw9wepbB+0D0YqiKzSQmU7V+MnjIcSPA/IVJWPZQC4NfkqKusTzv5KU/VQlZ/AXhz55vVQYvlgH0+hTtNHklyDoSI65RPc12Yick0mANycup3KJkVOIBEjAZuEM5K6JgDc3+QHR1d2vIXKDlXvprKZeb5R2NzDx47afAFAEAcKbUl+wFPO8IhyNTHXv9siD8ZQ3E5lFbGuTwUe5TCKXEuFMi+n3pUAUKvz9U7l3sikedTFotDKvhDJBM2Lw3raYIyGEDYuai82Anj6iySEsBnAZwG8N8b41Nb/ZgDXhRCOYuE5hkIIX4sxvvK5F44x3gHgjsVreX9pjDHGGGPWDDt5P5v1tMH4HBacuD+8+P+/AYAQQg+ALwD45RjjN5/65RjjxwB8bPF3tgP4/HKbi++FdOBmHB0pLlNJkztESMfJFpmqFVcHHjBrqI1rBQ42ePSKrhaxuwciN/Ho6+Rq5p0JrsHYJmwxlakXAFR6eVSVgSY/eTrVzk8KE+D9rDQGrcwuO5tcE3FTktteHwc/Idpf/iqVXZnjKnEAuDTDT5dGy3z8TICHaL1v9o+obKiTm1ZtDdz/ANBaio6UygTPJ9j9jS9T2XBKt113VNmzuZJ067JWnwv0df9LKru7+CdUdnsXz6wO6AhUh0vcTGFLnms+2svCHBI6d82c0BwlE9yspE+4Y70U/AT2yTluPbu1wXOhAMBghmszlXlQCdzUdLTJo//1JnVUqw1Z3gi9QvujwhE/NM2j323v4JGpAOCGJtfoHqpzjWUqydu1s52PrZHIZWdbaAzSItu70uQls/yeocnXnqzI9QEA4zhFZYXITfPK1WleH/ESUmZQKicFoLN1KxpN7geozI3NDw4vVpjaP8OCQ/dACOEkgP+IhY3FX4QQ3gfgOICnvMQ+CGAXgF8NIfzq4s9eF6OwuVghXWneHFvy/MOzM70ypYgwHwcAtItQd+kEr2s/+Mdj7zxPEvYtEYYWAAaSfIOxN8FDD6bFR/u+Bn/hXiLyKgDA5altVJYTCQJ2J7jvwoPzfHFXtInwigAwIByDp2v8o6tdxJPfkuUOvr1ZPbVPlNTiz8dzt3AkTAtfEmWqo54R0D4PKqDBGLh/T3uSm6qoJHwA8HDkZhylJr/njgRPbtgpwgbf2vFeKhtskWhvUzv/CLqkm384KJMkdc/ypP4oVaHAlc91T4aPgbow3TxX4ptT5egPAJ8e+7+o7MreH6eyjPDxUmamO5vc9h4APnWOm4T+h4t+jsqGsrzthnL8w/tEC8/G3izvyz1JvtFUoYNV6OT7y9z8rqddz1nFfORzdrLMfSU2ZLZTWV+DHyYAwGSSf74MJ7hp3kyG+2rNtvBDWSntWb5uzdbFhifotYmRaFGuKUzBXnyCnbyX8GJFkXoPEf2TMEUxxl8H8OstrncUQMscGMYYY4wxxrwQ2Mn7GdaTidSLTkea7zwvyvOT0vYkH1GHi/yUZ7amR6LSYMyJsqqcyjD6kiqP7gEA58AjWEwkuCnPhsjNlWYaXFVcCdtlfSaqXL1/Sx8/TVens1e08bqqU/9NbTos4WyVj59eYQJUExFytkduUlFV8ScBNISKujPJ6zPa5GPgprZ3UVkRvO0aIlsuoBfsgTa+hG1LcZO/dILLRFcBAMozu6isO81P35QWdEaEs56q8HKtMnlPVPh1+4VZjRo+6ppX9up5cLLIG3eziPScS/B7BhEtrF2YtrbKIP+BkX9PZUJxhpJQxQzVlJZPOyr/UJ5H2RKvLiSF22GviK8w39D1UabBWTEwiyK537fO8XUiIdqnU2hMWnF46vNU1tPBzSV7RCbzSosIZQMNXt8BEdhjRiTk7OrlWhNl3qsc4AGgAB4URZk6Kcf7RpprTxtNbTteriyfZsCsP7zBWEJGqMw7U3zBaAinnq+Mcpv1N23SdppZ8VI9nuST90+muG3+NeBRpDa364+Dnpr4EBZfgTWxCF3b5LHNezJaVbolxetbFlpUFdYyJzZnV3bzF0pRxYkEcKzBI8s0RbZYlbl1qI23j7IRB4AN4GZJ+2s8tPJO8SJXFieDbdzOuUt9HQEYzIm8C+I9Pt9YWVSi+6e52h8Aru7mz3J0jr8cc0m+3D42z+32e8U8OCVyfbQqu7GNryEqEbHafJxrEXlImbuVRV+mRUSwgpjr1/HvMRzl5xMAgKv5XkCO9ZPzvJ+nqnzDM5jT82CizNtO+fodLPB+VpHW1DUB4IFx3vC3DPM26BEmxWr03N7HD1RGS9x8F9AHPFf2cZPZk81xKpsX0ebahZkcABwIPApks8HDhJ9L8gO543PfpLLr2vnhz1yCHxoBwFyNh42dm+eR2JQ/TbF0lMqymZVvFtcDNpF6hhcrk7cxxhhjjDHmB5DnpcEIIbwMwN1xSWaiEMK1MUYdhP08Q+UVyiT42cpklTfj6zZwLUW+xQnRifmVnTC+Jc8duYfEKZlysAOAhnBQvEg4k6rT/WNzXHV7dR8/7QOAqjjuOjTL77m7m9dVxYwXAYvQpjoEwMsHdRIthjppV6eoQ1oZhSdnedturXNV+yU9/EQ8J0wFRTdL8yBA93OXUHKNiRPfG/tFXpsU11AAQF6smheLCmUT/BR1Szsv9+1x3tGXdOvgAiPC57ooTv4nxDwoir6stFhD5ur8piPC178vw8v1Z/Q6wUiL5I8AMFrm7d4uxsCAcKo+J8bkP4zpk+TXDPNxqbQN24XpmcjXiv6M1ka9abM6n1xZlKA3jfCF6+4xXtlWDvsrzY7dHvkEGsnygBfKAR4AMkUeWVHlSL1/7h4q29rJcxR1NHi7dolnBIB+ERDjLvBEljmxjsY27pTfykRqPbMQpvbFrsX64fmaSH0JwHdCCO+KMT6VseXjgEjrex6irFyU2mtrOzdv6BMvv1aqtHMV/tGhPg4uFd+yXz7NbVzftrVFwrOseqHwxmsTa+2uTj4EW3yrICeuexN3pcB3J/mFL+tWGzB+zZ15HetPbUK3ifHTnlpZxIz9s+KrAsCrh3l9DxT4OFBme70Z3kCfOc5lP7xVv4yVeeKpEp8jN/dzs7S7zvGX6muHeWIpANjWw8O71oXd+sk5ftjwyAz/un7pEP84OlvWbdeX5mX7xOZssIePj9NlXrAzpT8su4V5zEMz/Fku6+J9cuNWbtI2NcP7udrQbffYNF9IT5T4fO4SbZ4Vpq0/tYN/sALAxhxfux+ZFeNngG9cjhV5+2zr0GGktojr5vv5mlaZ4203PsHbYFMbt9t/bFab005X+Zqm3jPpBN/APylOozaolx6AK3r5ddWB3K8Nf4DK1B6rQxz+tHrPHpjjY6sSfpbKRgP3lag0uOl4T1Yn/jtZOSrlLzbrJQ9GCOEGAPcC+JEY418u/uw2AB8BkATw8Rjjh5/HdQoxxvxiOoj9AJ7AQmD+IoB/HiMPq/h8TaSeAPCbAL4WQnjKiH99tKIxxhhjjDEGIYQkgN/AgnJg6c8+CuANAPYAeE8IYc/3eOlDMca9McarAXwSwK+oX36+GowYY/x8COEJAP8zhPAJqCPrdUwikUNHbvl440q9P1fjJxJpEaVDOYBf1Km9DF92ET+Zy7av7GT7nwtZvfr9d8lRidVji5OAlFDhJ4WZwnuFVUmynd8zKGfkFiZSiiDCsQShao8ilM31Z7ijNgCMHeYajm8/xiMsDYi2u3+S1/UX9/DT0O0D2qm6Z4RrItJ9ou0yvE9eKU4JE+369DF0C/uzujCDOsodNHc9yU8Jv3Z8I5WpiDwAsGuAt3tNJAq7YssoleVH+NqT7G5hriTsB35ImMole/irKvTythsQJlm1o3rcpb4rVJYT3AP8CXFC/0PD/NR7dzfXjAHAtt08X8Mrh1a2dl8xxddQkZoFAJAa4otBopdrG9rFGOg8yrUiqft5f3S2MGs8V+EajJE2vr70C1lamE2X63oenJ3nmqOaeO/lhTZXmeao75ByC01eUuSlKNR4zq0rA5eV0tdQ2aEKH+cAcBJ/L+UvNuvkw/j/APBXAG5Y8rMbARyMMR4GgBDCnwN4C4DHlhYMIewA8Gks7A++KO7RBYhkU3j+GowAADHGAwBeBuDlAHT6XWOMMcYYY8xqGAgh3Lfkz8+wXwwhjAB4G4DffY5oBMDSbJUnF3/2XD4C4GMxxhsAPDez484Qwr4QwiEA/wbAb6tKPy8NRozxmiV/LwJ4VwhBG8qtU5rNCgqlI8vKQp/Y5QvT4nFhX/+SQZ4fYvNmfYLWtoPv/xKd4ii5hcPbSgkrva7yjlbXbHU/EWJR1TWqox7h8CePiFp5dim58NMJIhQtMnzcJbfr+my5kttI//prTvL6rHQMiOcI+RYO8Ok+LqvxE+o4I7KVz/LnD3ntixQ6xdFulj9nZoj7YIxczuvzo3M8e68cywCCcB4PKtt7Gz99hMgtgZJ20GyeE/4tFd6XoYP3iXyOPt5XmV7t3HrJVq5hvrjMQ3cG5XS+0rkOADmeURlV3nbNM9zePVkR+RFahI+OJTH3cuK6QgOYvphrhnYMc23CtiLXuAFAVA50qpzI+VIZ57LRE3pN667zeaJs+EsNFQKZy8otcpoo0sLv7rIeft15YWQxU1VjS8SHBsDd3F98YnzBfDDGY4w8MsCz+X8A/NsYYyM8+xtpuYot17m3Anj74t8/hQVTq6c4FGPcCwAhhB8BcAeA21hF5AYjhPD/kgo8xb9U5dcj6WQHBjqv+57LTYoJ8YpB/tIs1cRHoDDjASC9zmsn+WKb7OCTvlkSKnFRrhVBmQiJDUZQmcJafcyqTURNqJLnRU6TGd4+TeHH3SrwhXJsTGWFU6iIdpTq59dMDOqPp9DBN6ihW4TzEZsaJFY4fkSEFwDyo7Up8szMP8I3GIcP8cgoWza22PgP8Pme3sg/EhMiCZ1q89ApZKo/gJUfNqiNS1HkdDmp227mQd7XM1P8OTu7+D2zwtQ0t0UkKRxuMUfEh3AYFO2eFjK1OVvNPDjNzaumv8v78luHeW6JXT06qtWocBDf3MXn5cbLeOK21KCYP518k5loEU0NHbxto0ia1BAbsGqB9/NDE+JQBMCxef6cyiH7rIhs9uCkCNzRPEFl+abOx3VNp0gms0KKdd4ff1P4izW/3w8yIYSfA/D+xX/eDuB6AH++uLkYAHB7CKGOBY3F0gm/GQA7KXk+ll6fA/CH6hdaaTDuW/L3/wzgPz6PmxpjjDHGGHNBsbIAzSsnxvhRLDhvP8XTTsYhhD8C8PkY41+HEFIAdi/6WJwC8G4AP7rMJb+5KPsTAD8mbv1SAIdU3eQGI8b4ySUV/VdL/32+Um9WMV06tqysLDIG75vgp96vGuLlMklxIt5Ca1s4zGVPnuDq8qpQh54t81O53Z381AkA/n6Uqy5fNshP0E7P83sO57hpSFKoZgGgLpxUK6INToj65ER//Zcj3HTobf3LBw54iiu7+eljVTxHrclPrPqzIofIMM9ACwA9g7yvcyPC9CwntFEikYxySK+d0xOhPMGvOyvCu56c5s7RysH5M9/eKevzho38ZPfvv8hNI969g5txbLxojMrSPbw/kr0tzGoEDXHiGYX2tDrJ63PmpDYNOVPgwQVSYr6Pj/NT1PakGFsP8X6+asM5KgOAnmE+RzIDvA0S+ZXl5WjO6XlQmxJBSM7xE/w/e5xbM798kGucDs/ovrxviq+jbxDhtT/zt3xevmqY12dIhCru6OTvEQBIivDI5QKfQxMz3Hl8ssTXHuVUDQCfP8WfJRf4Z9lQjvfz8QbXDO1O8kzn1RafxJ8v3ktl5SZ/74/NfIvKNnTfTGV709TiBgDwj9gv5S82mEL9kQAAIABJREFUrYLTvFjEGOshhA9iIbJUEsAnYozLJTL5EIBPhxA+hAVH8aXsDCHsw4K5VRXAT6t7Pt8oUsC6cY43xhhjjDHGMGKMP/mcf98J4M4WZY4AWLoD/PDiz48CEPbT/5TvZYPxA0EykUF3bnm7U5XR83Wb1Ik4t0V93RXLa0sAoFnTNutHTvFTu20itOfRczxE4FaRPOlwQSd6un2En4qfmONl753gtrMXi0R7ZXF6DwDTwi/moUl+gnbDIG/3iQqXva57O5W1MnX/y+P8utf28xPPu85y2/ObRbruJ+b4iRUA7DrLNSqNh/jDnBSJ7VTCpg1Z3h9ZYXMMAJUGr8+9E3z83NTP7/nwDH+O6/u0Q803xvnJrkr+eNcZnv0xPcqdqlVyQ5WEEADahVxpxx4UmiHFljYdPnu0wvtL+AzLTNWjZX6qu6mNn84ePLSZ3xBA7QCXbRbPeWyeP2OfCK3d6tzzoWm+hlzaxdvnii7ui3TPOD+hVwkDAaA/w+/5P4/zd9DlXfy6D0zyuTV6ml+z1ennwVl+z91dvF3VfL5XZBbfJt5rAHBFN39f/v4oN2u/HG+gsscrX6GynsxbqawQdELFJrimsz+xjcp6+7h/TzXye5Yif+etdyK+/yZS65lWTt5zeGbutocQntKHBSzkxmgR/sUYY4wxxhhzIdHKB0OHFzgPaTTKmCgun9m8v+c1tNxmkXBnV69IiCeiALXioia3qVRJ8TZ2c/vOqQI/mRxp5yddAFCu8+GST/PTnGt7uUyFwGtlxzorInRd3M1PqB8VyaU2tvN2HRI+Oq3O0C4SJ1pKc/aSQa6lmBNRrSa0STL+3SHuTvXK9v+Nyl6/aWU2/6fL/PlzLTQYXxAR0+YjjyA0lOOns/sm+QnaGeGjAwD9Qjwu4lkHeUbNZbMiFO+eHl3XqogG1S4iv50qimhz4jGOi8g6ADAsFCNKAzY5r9qHFzxW4LJLuvX60i3s9vdN8+d8coZPzMEcL3dZi/rs6V6ZlfLx+RYRlggTVe1L8sQMr88OoQx/fI6vsVWhkLuok9+vJLScALCjkz+LGj//OMbficr+oCAS8wLAmFDX3dbB19+zDe7zMJi7lMqmI/cnOvXsPGv/tCzxWQWARhsf6/UGfwn1pLh2I9tc2XhdL7SKWH8hccGZSBljjDHGGLO2BMSWxo4XDhfcBiOEJLLp5W051YmEOqHPd/CdekMoBaI2V8bUOPftOCN8HlSEnANCg5FrEbWpKCIz9WX40dMBcapZFidPI8J+GgBUCo2HJnmfzDR5p/Rk+am3cN3AREUvKl+f4AkXr+vm0blGRKj+kjjtUyeBAPAvNr6fyjLiOT97imvr6uA3vbmH+xMN5HTbtSX5+HnjRn6CH4Vm6DKRd+KIUg0ByIr6pMXxflYMWOFOg1KDa41UDksA6BOHgdMtcrcwlJZvc5seeG1CWzVa4ae+Spv5wCTX2I608X4+rU3PMSsSzc1UlWaE+5wpjdJ0Tc+DU9O87LY8LyuUudInpNVJbHsfHwfjQoM6JZLXHSrwTmlL8XfehLgmABwR120K7XMjrswCQWkoAJkjFj0it82M0EZNVZdPIAwAhSSPUhdbeA2kUzzy29jcw1TWENqWiQQvl+t+l6yPOX+44DYYxhhjjDHGrCURNpFaygW3wWjGCgql5RNMKFtMpRW4+xiP6713nvtRTIk42gBw3wQ/TT9a5PVREVfEwT7+eOoLsj7v7X0jlTXFUNKn6SK2e4sTPUVvltdnvqROoPk194sTxGqLLLzHwiNUlpq5StSHj4E2cSI+JzKZA8BYgQ+EgojisUNoeOYb/J6H5vhxebnBT3wB4A0j/GS7XeQtmRJR2ubEXN/cof1MiiJHxJnSyiKg3FU4Q2WvyF9EZV3CTwDQWln1HAMi30kyqHmg52yhzq97SmgUThdFVLh+frJ9dI6Xa+X7o/xilDZK+VQp/5WJsq5Ph9CoCHcaHC2quvK5lU/pNe0+/mqTmrW/nP08lb1v4M1UVhZOOgfnuC9WK6YCz2uTAdcY5AN/fytfCQC4pZdrdGui2ZW2ZWaehz3Lt/E8TaWqzgezpfMWKotCwzM3z9ugLcu/mcwPDhfcBsMYY4wxxpi1xj4Yz+ANxhLmxNHBeIWfam4TuSWKVX46O1bSEWDuEQcLG9r5ScavHf19Knt33/uobHviOlkfpeFRbXeyyk8yjid4Vs6ZKs+cDQC5FI+L/urMq6msBn6qeXaeP0exzk/oh9r0qfc7ul9GZQ1xYiUOQ7Fvkp+Wb8vrSBy5JB97pTov2ydOtucKvF07UvykdDM38QUAHC7wslvaeeMpn6J+8Rxfm9Qnehdl+qjsbBCR38S4623yayr/jPYWSaNFl2CsxNsuKYzEN7Rz2dmSfrkqDeGZeT6/MuLo/5h4yME23kCPzehT764UX7u/Xvsylb2ri2t6DxW4/9dgRs/ZYRnhjvdlm1AnPCKyg2/NawefmtDajpW4H9ObOnj7KNeFjBjrqq8A4IHmd6msP/J8KAmhA9yQ5RNzQ7twnoM2oykJzeLxxEEq29LzKiqbFe/S3Z2v45UB0B352nQuPk5luewmKuvMcA1GW9Rtt96xidQztHARNMYYY4wxxpjnjzUYSyiJE+pHZ/npktJgJIS98oPTWoOxQWzkVUz9f72ZRwg6O8/rs7d9WNYnL2yAjxb5afq1XTxK0u7azVT2Fdwj63NR83Iqe7I2SmXXdPDTk09P/jGVvavnvVRWb3Fskc+K2O+irDA9R3+Wa00KypAXwFCb8OER/awid3Wl+XIyUVE+GHoZahc+RXN1XqHRFqfpjKs7eFZtADhc5KfQlcBl/U2eyXtjmvsRjAm3DjU+AOBkka9pAyJN8UyVj5/eDG/XgRYh7M+IqHrFOn+YiSaXqUzE7Snel5d0adWZGuvvifwU/jtzXAN2YxcfAyprNAAcnuVtsLeHV7YnzcfATpEf4tCcXtO+WrmXX7d5JZWdLvOIenPVKXFN7kcwCZ7nAQBOFr5FZflOPkZ2JIeobFBoQVvxtenTVDYEEVUw8jY4leBRpELgdR1o6ve+WpvSYh7sj9+ksoryUTmPv0pjtAZjKdZgGGOMMcYYY9aM83ivuPbUIj/pqYsDYRWjvSlkO/M63v6pEj+hrgvjfHH4iJrIOdAqd0JZyK/u5aeByn5acX31Binfl3iIyq5J8MhMqj7vG/wJKpuo8AZ4uMw1JgBwa4rbo6rTYhU5Jaci2bQI3z4r4vh/rvgPVPby9CtWVJ+UOEErqRTOAFQ0H6WoOTXPT3zVafmOFv4rA2kur1U3UNmmLFdJTlW5hidf56fM/UIzBgAXdwutkohaNFbmUcY2Vnn0nJQK8A8dgaostBS7Ovn6crrI18m7S9xmvS1qDcaVOa7pzAptw7CI/KZ8qtT8AYBbhkVkPDGHZoWWb1q8giYqWj2WC11Udvf8n1BZXrTrS/B6KquJfA33Tv8ulQHARb1vorJZ8BwRn5n6DJW9O/mvqUz1M6C1FN9u/D2VdYn3yJGpO6nsyt4fp7LOqC0plD9WPvC1UI2PMrgGoxsiDfx5gJ28n8EaDGOMMcYYY8yaYQ3GEtKBH0vt6eJHPW0pcVJa46drhwo68lBBnDwdEva4KXFCf7bCDbq3t+u8HJ+Z4/HLX519A5UNtvFhdmaet2ulqVUqm+NOKkuINlCHrMof4kkRL3xnitvqAjqOf7HB22Agy0+IukXK7VZKgf2zPPtxXtgdT9f4SfvDtW9T2ds7ueajqBV5+F/T3H76ljSPfKa0FLkEn+t/W+SaMQDY3OB20OXA59c9tSepbGvcRWW5JJ+XrTIY92dFhmfhS7GxjZ9qzooM4J8tfEXW5ypwnyvlE6Ei68w2ubbl8bnPUtlFPbdRGQDUm1wbdaTEbf7bA49odHBO+OhUtOZM+U31in7+9LknqOxYiftRvCynMyqfrvDITJ05HplpZ/IlVNaWFO+KOj/1Hurm1wSAY7NcK3tRN4+iNNx1LZWpqILfanxd1mck7OH3TF1KZWXwdXtr72upbAZnqexc5Fo+ALhEWBKcTp6gstFZvo5el+djq09oiM8H7IPxDN5gGGOMMcYYswoiIAz5LjxsImWMMcYYY4xZM6zBWMKwSJbWm+Fq+GqDm1s8McvV/i38ITEjHHEP1Li5zkiCJ8bZ28sdTYVPLADg0sRLqawvx4dSRdjrZBN8j5sWMgDYmdPOaYwHZnkoRBVS9yWd3GTinMoQBUD5uQ/luEpYhSP+78d/jcp+dvOvyvpsyHKzm+F4GZV1i8xuVwkHcOX8+8SMsLkBsK15CZXlknyMlBtcdq7Ok6xdleLPDwDFwPt6HiKkboObKVTDdipbqaM/oE3lOkT437KIEjBV4bI3tvNkX4BOwDYlgih0igx97YGv29d28dDSY/EolQHaBGZrrpPKSiIiSHKFppuADmjw3Qnerj1NHoZ1f/k4lR1o40nUAOAN7e+hsobwcs6nVXhkPreOhgeprCvB12YAiB28Lw/P8KSJI123UJkyg7o63CrrowLKnMEBKovg5c7OcFPSjtxWKhvKcXMtADiMR6nsxCQ3iexo205lOeFYvtKgMOuFKAL7XGhYg2GMMcYYY4xZM6zBWII6DVR70gnhnPfgFD+tOVdeeTK0a0RSPHUA0Caesa1Foqcr8j1Upg5Se0VCIuVo2ipsbkbUVyUYu6yDP4fqEZW8bjWOXTVROIqTwB/u/yUqe3RuWt5zbxdvA9Wuj0xzJ9WrerlWZFKceittAgBMJ/iz7Ig8FKJK/AdwTV5HSk+EiTpP7FZJcCfvN7bzQAh5cUI/KlSLSRUvFTrE5EnR7Mcq3KG2ITQ4jRIP0QroPlEag3tq+6gsH/qpbFCc3o8Gvf4qjYI6aT8deSK5wQZvn4o41QaAdIJrw1VY1C1pfs+rhaZzRmiUAOBclY/1E0muGclV+XM0wNs1HficrUeRjRJAM/LrdomT9g4RTna6zp/xgeTXZH0GkhdRWbPJtVGjs9+hsp4OoelN8fV+vjlBZQBwdoYnvE0k+JpfE+vk8Rx3LB+o8ySN6x37YDwbbzCMMcYYY4xZJY4i9QzeYDxPisLPotLgR12TIoykOgUDgKIIU3uswX0wBsSpy7ky7/JWyYHUye6DtWNUNtwcpLKRNn4C8kRZn6xcm+fX7REhXL9W4rbFr2rn9vfTor9qTX1uMZjjoSuLQjPSluLP8ReTv0Nlt+R5wsBWKG3UkAibqxZWZT+eaJGY6LIkD3n5RImfFt/cw0+2O9Pcbn9cZZQEUAncZ+TGti1UdqTIT/SKJX4C2xv4iW9X1BqM0/O84b9WfpjKbs5dQWXHyzxE6/YuPs4B4MAs14CpMOGvaePhQpVfw74CXydvb+chcwHgoQIfW2cSh6hsJO6msq6Mbh/FY6VxKtsr10J+TeV3t1H4WwHATI3LNzb4nFX9PJLj74MDJf5eO504TGUA0JniGv88uJarAa5NuDLFw8KOivCtALCxMUJl5QQPRdvZzkOzFyt8rE8VuB/FgAjF24pmk8/nqpBlhQZZzWdzfuENhjHGGGOMMavECoxnsJO3McYYY4wxZs1YdxqMEMJRAHMAGgDqMcbrQwjvBPCfAFwG4MYY432Lv/taAB8GkAFQBfALMcavrvTeIuIl5oQ6uNzkKr154bi4Jb9ydXl/YyOVjYpQkMrMSTlZAsDZCld5vrJrO5V9vcCd4Tqr/Dm2prhKHNBOmBWh+r8hxZ3hZoVn+Y5OZeakzy3OlLhZTVPYpj0cj1DZz418iMoKLerz+Bw319mQ5SEEVSCEJ2Z4KOeNbbztVOZjAAiio187yM0bvjw+RmU3dHKTkkt69LLYVuD3/LOpT1PZu3p+lMryaW4G9e1ZbvowKkzEAKAuHLKvS11OZX899xdU9vMjPDzpGWGSBegM4Sq8a1uKjwFRDFd38L76/TO/xQsCeO/Qz1PZjqYKeMHr2i6eQ4UjBoAr0/xZjhX42nxamCTdMsTn3sfP6Yz22yM311EZ7S8Xc29SmCduy/KADvM1bgIFAMnI53Q+8rmnzDc7k7zthoMOdf0N8MziV0ceDr6W5ut2V5q/S2dr/F1abeggG4Nd11PZTOkolWVSvL9GmtyETpk3r3cigKbD1D7NuttgLPKqGONSg9NHAPwwgN97zu+NA3hzjPF0COEKAF8CwI0bjTHGGGOMeQGwidQzrNcNxrOIMe4H/ulJZozxu0v++SiAXAghG2PkR6nGGGOMMcaYF4z1uMGIAL4cQogAfi/GeMfzLPd2AN9dzeZCRc85WRJx6rl2GoNtvIkHc1qVdmSOq4snKtwM6rIeHumnLEyHVPZVAEgFfl2V3ff23m1UprJ8iwBKAICqMI1QpggdaWGmINpnWkQE68nqvrxSZB2P4szjmsCzrM4JM6gO8fzAyjO6i3QN6KrzsX6gyCOj7O3hWZEBPbamRJ/8s2FuijElkod/+5yY0ACuH+CRbn61+8eobLrK++SeCZ534uV9/DlaofLBKIucN3W8k8ruG+fmL5d08zUC0Fl6Z2tcppYmtW4Pt3Hhf9j5C7wggP3TfCKopWlrJzdJUuNVzS0AGBD5hC7u5mOyS2RsP1Xi7fP6jqtkffrFmteT4WZH58q8XKnOn1GZ0F2d1YYLao2tyncQr2uhxufWdF1/hlwBniF8X/MuKjs3ex+VXdfzPipLZHjUvHqLT6aJ+SeprFrjkc2ULJnl/XxeZ/KODlO7lPW4wbh10eRpCMDfhRAejzF+XRUIIVwO4DcAvI7IfwbALwDghrPGGGOMMcaYVbPuNhgxxtOL/x8LIXwWwI0A6AYjhLAZwGcBvDfGuGxw8kUtyB0Lv5+MicTyp2zq1PfkPN+WjvDDYDwxzU9dOlscWamTlZcM8pPCojiBHuYHXbi4S9enIpzZlTYhl1SnR7ycUP60vGebuGc2wWU18YzqYEU5agNAMnD5tDi53ZDjnVlu8P46WtR9qRJAq2dRbXBxN++w5Cw/0WylbdnawZ9lQ5a3z2ydP+Rglj/jZS1yOcw3+MATyjGIQzvc1M+1OEWurMS0mgQArurlbdCV5m1QFLl9RnIi8zx0oIhcktf32Dw/Zd3SxhvhRImX60nz+5XFMwLAk0J8bT9v12MF3j5TIjv2tQM674Sy7t7TyU+hG8Lp9NEZPtZvGdB9OSe0DXMih9P2Dt4nF+X5M05Wefso7RegT5WrQtgrloKyyI1VqPMxCWit28X1N1DZ/sA1HwVwZ+0bE9xR+1R9hlcGwGaRGyrdxp/z0fpXqKwnzRu2QzfdusaZvJ/NunLXDyF0hBA6n/o7FjQSj4jf7wHwBQC/HGP85venlsYYY4wxxhjGutpgABgG8I0QwoMAvg3gCzHGL4YQ3hZCOAngZgBfCCF8afH3PwhgF4BfDSHsW/wz9OJU3RhjjDHGXKjEuPZ/zlfWlYlUjPEwgKuX+flnsWAG9dyf/zqAX//e7tJEs7m8Orkvw3vy2l7uFarMam7o5Wrm3zugVeI/vYsr25TT+bZeri5XY7XeIn5zWpj5ZIXpw0yND7OBLG/XqarWlfZmhO2IIIhWSArzqYQol0lqc4JTRW4idGM/j20+XeWmcOkEb/MruvWqdGCO10eZcz06y8dsTrTdG0d4XTMJYdMHYCjHx0gmwdu9t407Ix+b5THaO1J6XOVEX1eavH3qTWFSIubIQzM8QMCrhnXbKeuzr41xM4V3buVO5xVhGrK1m5cDgINT3A3uzVt4To/7z/VT2Vu3naGy8SK3Xy22MGO5opuP2bvH81SmTGZv4qks0JESkQcAbM/zQAmZFB+TuRQfI1s6+Ng6JtYsALhmgPdXVZgnKsZK3IZ3i3DYn2txv6KQt6d4P6u1UL1HSmKOAMBohcvH+LKFW/r5/CnUuuU9GTtSfG4BOjCByhvVFd5IZf3CRlflkVn/BDRF7pQLjfWmwTDGGGOMMcacx6wrDYYxxhhjjDHnI+ezSdNacwFuMJJIJpdX/XaJiCNXDUys6G5K2/epa7nKGwCC0C+RQFgLiDAGUYU4aBX+QOm7VNkVlksIU4NWBBGmI9HOKxRE6KqQEWrvFkk7rkwLUxYRTz1WeWSQOCfM9ka1ydYl4/w5T45yVftFeW5WcrjAO+yq/ikq27RJm9W0bRImbb28PqGXm1vszAtzlLpuuzjHTRCbM1xWH+WmV+UJPn6umOGT/QtHdPz/a3t52/7i9dy0KJfn47Vts8gbsFFP2u2Dwvws8LI7wXOTxBlhsjXJx131hDaFmzrJzYc25vm8VOZcu7v4mr9rO88bAADtu/lzJoa4OVMYFNHZ2/jYurpVQP853j6Y43Y+zXHedo0pbnZVGeO3K07ryG8NETEsKSIOZkQUv5wwf060iISk3sN1bjGLaoGv2+USlxXmeV+NzYvQkgBmavxhzlW4bL7B71kSS+wVXToPEZaNFWrWIxfgBsMYY4wxxpi1w2Fqn403GMYYY4wxxqwSZ/J+Bjt5G2OMMcYYY9aMC1CD0UCjsbwNaE6E/ewb4oaRmU6RLXaK7+HaLtdhAMPuDVzYzTP/IqvtUVdMQjmFCGcTWU7IWoSriyu97kqfQx1NNFehGBVlVQsoWa5FfXLCE61vhc9ys5DF9FZRGeVQBEBkfV1pf8Xayn0wVH+pExs1K5WMB9QF3t+qr1JiiU8JnyLV5qJcVI5jgDY+rwqfCPWc4ppBlMu2OGrcsMJ5cJESrrTNAcQ2bisfRT9Htb5URExU1R8LF+YydU/R7ilRTsnyjRZzti584NR8V9ctc38rzAlHCgAY59mzm6PcTyc3x/skX+R92Vviso1Fncm7XOB+FpUKH3eNBl8L6kJ2vmMFxjP84PayMcYYY4wx5vvOBajBMMYYY4wxZu2IsA/GUrzBWIKKNJrKcfVseoSrEBM5rtIM3SLMHwBkRKy7DhFaTplFKFpl0BRmNTGpQriK+qzUtApoEcdXyJTpQ0Op0rks1FqYE7Qyu2Eoc5QsNy1S5hQApFlSTCqzGtWXL4wpnAwsXuWmTmFujstmRGjcVqYhqu1UuysTF1UuI0xn1LwDVt4nCmVu0moeTPGwsUGElJUIk9DYLtpVraEAYlasz2mxNr8QpputUPNgeprLzo7yawqzGgDaFLeDhxyOyiSyTZgNq7WQl1pArflqvpd4yFS5arUwkWpOcHn5gAiDrSy2ZvmYTKVFKPSo119lBlWt8T7JZsQ6IXhkvG9F5dYF0XkwlmITKWOMMcYYY8yaYQ2GMcYYY4wxq8R5MJ7BG4znSWVOqALrwnxqK1cV1w+2MAk4zFXbqYt5tth4jkehiCWuDk4M6KhWKgNraBMZlYUpWKyu0HQILTJrC+Q9RfvILM0T+jmq0yKTbFZEVRHBwlKbuYlHYjAv64M+fuGgTEeU2Z6KkKNoZT5WEdFaZkSm89N8fk19nZsojJ4TjQ5g54086kqyiy+piR18zspsy51iXipTndWgTJ3mhelMgfcHAMSjPFv145/lbdfXxfur92Je19QIX3/DVt4fABB6xDhYqYmqMgltNQ+UydK0MPkb5eN19mtc9tOf3yar87uvPUFlnTv5syR7uWlVYouaB6LNW5kKKipirE/xd2lziptPNUa1ednMIT5G7j7C271Y589ZEtnK+4S5khqSAHBinq8xJ+Z54ZcN8nW7LSmihQXbGK2GEMIrAfwNgCOLP/pMjPH/XJTdBuAjAJIAPh5j/PDzuF4hxpgPIWwHsB/AE1iwECwC+OcxxidYWW8wjDHGGGOMWQXryMn7H2OMb1r6gxBCEsBHAbwWwEkA3wkhfC7G+Nj3cN1DMca9i9f7AIBfAfAT7Je9wTDGGGOMMeYHlxsBHIwxHgaAEMKfA3gLgGdtMEIIOwB8Ggv7gy+K63UBkGY4dvI2xhhjjDFmlcQX4M8KuDmE8GAI4W9DCJcv/mwEwFK7xpOLP3suHwHwsRjjDQDOPke2M4SwL4RwCMC/AfDbqhLWYCyh1uT2hEdO8dBpu9ITVJbt57bDD3xbZOoGcMUuHkJw5j5uG6oIgXf5/zo4KMu+YycPhThf5na19471UtnePu5nkhZ2mgBQrPJnaYoggqmg7HX5c9Sa3CZ7vq6nUnuK28COV3jYxr4Mb/PODLcdziR1uNBGFP4JZW5bfKbE66rmj7LzvaSLh5MFgGSCL7GZJLf1DiKM8VyFz+dCTfs1fOcLvOxNQ3wtyLfzsV6Y520+U+ZtPinGTivOijn70o1jVFZtiHChLUJednfweTBZ4vPrTIH7odRO8n6+pJ+P80pN+C0AyKQmqezwNPcVmBd28kHYlw/ntN1+bxuXp1UYUjEvSzWeJ/6mIb2m7TvG31+zB3kbXCn6pFTj46Nc5+/SiRbzoCQyR29sE74CKd52h+aGqSyT0O+usqjPMeHz8OAkv+7Gdn7NQo3P9VPz/B0DALUmb58dndy/8tNH+PipCTuiN29p4RSyznmBTKQGQgj3Lfn3HTHGO8jvPgBgW4yxEEK4HcBfA9iN5SMrL1fbWwG8ffHvnwLwG0tkS02kfgTAHQBuY5X2BsMYY4wxxpj1yXiM8frlBCGEnwPw/sV/3h5jPP2ULMZ4Zwjhd0IIA1jQWGxZUnQzgNNYnuezTfocgD9Uv2ATKWOMMcYYY1bBgklTWPM/8p4xfjTGuHfxz+kQwoYQFrLXhhBuxMJ3/gSA7wDYHULYEULIAHg3FjYJz+WbizIA+DFx65cCOKTqZg2GMcYYY4wx5z/vAPAvQgh1ACUA744xRgD1EMIHAXwJC2FqPxFjfHSZ8h8C8OkQwocA/NVzZDtDCPuwYG5VBfDTqiLeYCxB2ZAru+wmNxuFMvMdzuuY8Q8e4DauyuZ/osrreqTIbWM35LTd6JePbaSyfIqXVVd9dLqbykILLd2ssHXuSq0sv8acuGZC2E8fFe0KaJvTfmF9qv8fAAAgAElEQVQ+3CfyTpRF3PPpmj716FzhzFd9OSfcPnKieQ4XuY8OAFwk7PYrYs6mRX+1OhVS9Ap790cnuW3+2dO8L4dz4hlFP397UnfkrQO8U/qzXPYXR/jaM5Tlo6CVXv2u57oMLuGWIf6cYxWubN8hxsfsKPcraxP+OwDw6AyfmGo8F8T7QFFqcHt2ALipn/tgKFOEjx/k0msH+JjMtLBvePejX6WyXxp5I5XN1rSvH+NkiVdI9cfCPfnYOlzk/gnCBQMp4Rs2VtaN9+Vx7uO0OcnXw5MN4TtX5f40ycAf5GST56YBgHzk/k+PzvI1ZC5w37qL09x/5VDh/P4sfbHD1MYY/weA/0FkdwK4s0X5IwBuXvKjDy/+/CgAkYzmn3J+96QxxhhjjDHrgBd7g7GesA+GMcYYY4wxZs2wBsMYY4wxxphVsIq8FT+QeIOxhLoYGXPC56E4x211E0V+0UKF234CwGOz3PbxzpPc9nG4jSum0kJndfeYjod9KHGEym7NXkplKpb4SDu/X09aT9V5YZv+8DTvrzPzvO0u6eYNtKmN255f2qkNr+dF3POieI5jwrejPcXbZzC78rZ7YoaX7crwclMVXm6izNtHmAcDAObrfJ4MiOdUtvD3j/MxcDiekfW5JruFynZ18Ycp1rlsUvhNqZ6crep+PlDg1717jMe3v36A17Vd5KfJJnV9bh/h11VzZKLMr/uQyA3Qn+XrQI+QAcD+ad4+2/J8zc+KNVbNn/4Wc/Z0ic8D1Sdv3cqvWWrwe6r3IQD8+83cz2K/WEMOivw0A1nhUyV8Hg7qVDoo1bm/TS7J7xnE4hRF+8xW9fvglm7uh1IRjm4zc/yb4Bj4uvXw5Keo7MaeD/AbAmgPfNw9Er5LZSemvkJlW/t+icqqtjH6gcEbDGOMMcYYY1ZDtA/GUrzBMMYYY4wxZpVEG0k9jZ28jTHGGGOMMWuGNRjGGGOMMcasggibSC3FG4wlqKQ6jSicN0vc4e+hSZ40Z7SsswPdN8FHairB63Nf6RSV3dyxmco2tYuMbwAuTV9GZcp5cazEn+Ne4Vh+ea+uj3IG2yKcx7vS3GlNOcFPVvkzHp/XffnQVInKZlCgshs6uTPgiSL3Bpyr6SRiZ+o8A+Qtvf1UlhXj7sQ8f8ZG5HW9to87LrbikHDuVM7jNVGftw1sk/d8aJKP2RNFvqRe1MmvWRJO9xXRlZfxvH4AgDbhdH1xF59f6iX5ySPcQf7ybp2HaV48y1YxDIbFZUsN3uZl4cR8SswfALhpkLePCiBweJYLT03zAdse9HpXi7zxbuzjg+vMPH9OFZTgwIxun6E2XnZLB5c9MMHHT0eKByX4xMRdVPbmjpdTGQCcLXOH/f4Mfx+UGyKgQYK/LNT7GQDumjlNZcebD1LZzclXU1m+ycfARb1vorLeKBYmAHVh8jNaeJjKgnDmrzT5WC43+Bgw5xfeYBhjjDHGGLNKrMB4BvtgGGOMMcYYY9YMazCMMcYYY4xZJfbBeAZvMJaQSXB7y5u28iQ21erKmvGa3qKU12Keyk7Nc+XT1WnuZ1ESNtBT3EwVAKAsck8Ke+ZekTzp0h5u/7pvcl7W52yYoLJXdfE2uLiL17UmHvIvT3K/hbeNdPOCAJqRG5E/Ms3LtYmhNdLOx0BS2AcDwESF+1mMtPEV8lyF9+U7tnIb8lqTl5vlJtkAgEmRTE754VzJ3Z9weTcfd0f1tMR9zfup7JrStVS2Lc/v2SGSJo6IBI9/cbxMZQDw/p38np3C6eyecZHgMcEHZVYkLQOAYTG29k+LxJHC3v+u8qNUtqWxncoeD7wfAWCo8goqGxMLaW+Wt932TjEoV8HWDt52l3er+cyvuatb+5WpJI9/cI637Y/1XU9lR+d4u16JG6msWNNfdco36BszY1TWFvma9pI+vuYfK2gfuCtyG6isr8wdq+5u/B2VvaebJz6crfFx9/UKT5YHAJfFy6lsV/6HqKwtcqeqngz3szijHLXOA1QCxgsNm0gZY4wxxhhj1gxrMIwxxhhjjFkFEdrS40LDGgxjjDHGGGPMmmENhjHGGGOMMavETt7PsK40GCGELSGEfwgh7A8hPBpC+NDiz/9TCOFUCGHf4p/bl5S5KoRwz+LvPxxCyL14T2CMMcYYYy444oKT91r/OV9ZbxqMOoCfjzE+EELoBHB/COGpsAn/Lcb4X5f+cgghBeBPAPx4jPHBEEI/gBYxaTgq+WZHH8/e27+VRz14a/0YlTXmtLXepad599RrfG9YKvHIMaMFHtnhsRketQrQEZZ6M7wNxis8GsmhAm/0HxZRiQCgWOeRoqZr/LodSV7X9ix/yB/fxjOexhaWlzvyvD6lOo9wkhGDciDLn0NlngeADeI5O9P8uttFtBp1WnGusvLsrDf08QzhHWmeNbna4DX6hzE+1jeISEcA8FPJm6lMZa3vSnPZkOjLQp0/x7+8WEf62dTOs8TPVXmfVPv4mGz08nUpl9ARYJKBt0HfkMhmLqbXlbOXUNlIF3/Gd/Xo7M9DWb7miwBcyCV5ufkG76+TJT1HNrfxV1tnis+D0TJ/H5wWgfpUxC8AeESkM39L13VU1pfl1x3O8YYdLbev6JqAPlXeOMejNu3p4e8gdc+ejJ6Xj0zxeXJdH3/O12R4pCgVia4s1sItVR7VCwBKDX7dHY3doj78midElK2DtVFZH3P+sK42GDHGMwDOLP59LoSwH8CIKPI6AA/FGB9cLMPjlhpjjDHGGPMCYCfvZ7OuTKSWEkLYDuAaAN9a/NEHQwgPhRA+EUJ4KqjzxQBiCOFLIYQHQgi/SK71MyGEAyGEcy94xY0xxhhjjLmAWZcbjBBCHsBfAfhXMcZZAB8DsBPAXixoOH5r8VdTAF4K4McW//+2EMKrn3u9GOMdMcbdMcbB70f9jTHGGGPMhYV9MJ5h3W0wQghpLGwu/jTG+BkAiDGOxhgbMcYmgN8Hnk7peRLAXTHG8RjjPIA7AfBUusYYY4wxxrwANF+AP+cr68oHI4QQAPwBgP0xxt9e8vONi/4ZAPA2AI8s/v1LAH4xhNAOoArgFQD+2wtSN7EVC53ciS4hnL0SHdwZEAA6s9yprz7LHeyS43xIVuu8Ppc0tWNwtcnLphP8nv0Z3nib2vgQbOWoXBTOr8rZdiDLy3ULB2cVBOBwQTv1be/g1+0W7TMgHAnbk7zNv3FOT+0uMS4v7eQPGoSTrnJ8zYjxkQy6n9XY6szwOaTq+vJB3ubHijoQ3f4ZXrY3w58lK55Dtc9sjfflrryOadGW5vJk4Pe8RLRdU8zLiSpfCwHg785yR+bbNgrnaDHXr+rT92R0CcdoAOjL8LbLi3ZVZGorD3ZQbfI2mKsLx3uxTuzI834+V9Hnj1vz/J5l4evfnhTO0WL9HRSBENTa3IrXb+J9UhQOzmekg7y+5yuGedvmRBCSKObepAj8UhVfqT0ZfUSeF2XVx690rG/n758z0zyYijm/WFcbDAC3AvhxAA+HEPYt/uxXALwnhLAXCz40RwF8AABijFMhhN8G8J1F2Z0xxi9832ttjDHGGGMuWCIi4vls07TGrKsNRozxGwCW26LfKcr8CRZC1RpjjDHGGGNeZNbVBsMYY4wxxpjzEWfyfgZvMIwxxhhjjFkl3l88w7qLImWMMcYYY4w5f7EGYwkqalFxgkcqSR4tUVkUYRbqBR36olzg3VMpc9mp6S4qOz3PI+SoaDUAUGjw+vaJ6B8qism0uGdORNYBgBkRyCUpmvbkPBfmk7w+k1W+Hx+v6HOLTW38nhvaeFnlLzZR5ZE4dvEhAAB4bJpfeFREtTowy8uNlXiHvG6ER2pJt4gA863JPJXtzvN52SciTOWSIgqbiKC0IOeyuuiwY0XeX9MiAsyVXWUqKzd09LKTc7ztVNnJKp8HwznerjM1XZ+9vbx9aiKKXbHB2+eRST7uTjamqCyf0mmR1D2Hs3w818R7ZLzC23WsoifCVIXL6mLI3tDH1+ZvjPE1dmte12e2JiLciZByqq6PzfJ27UyvLKITAFzWzZ9zrs7rekpct0M84xkRxXABXnaTmEKPi2iF01V+z6qIhtXRYgEuiH5WqDGQFc9YgY6uuZ6JsInUUqzBMMYYY4wxxqwZ1mAYY4wxxhizGqI1GEuxBsMYY4wxxhizZliDYYwxxhhjzCqJjiP1NN5gGGOMMcYYswrs5P1sbCJljDHGGGOMWTOswVhCSYRh/daxjVTWd5aHVXtyrp3KTszr/V1OhHLrFiH7Rsv8OUo8YiFqOiosOkU4OxXiV0TIw6QIzZgIOuTlWHllRwUdYtSfLfM+UeEV+7M61J+K9KdC0T45y2XJwO+ZbnF0sE2EoFTzYDDHZZkEDxl7rMjrcmhWxBsGcHUfD1351ye47Ko+Xp+eNB/sqzmBOjbHw99u71zZcjtR5c94rqLnSIFXB3lRHbUUjFZ4qOvDc7rxujJ8/ExmROhXUaEeMfeStT4qOzzHrwkA9xXGqexlPUNUpuq6f5bHPR3Jtcn6HChNU9kbhnup7HSZj5Fr+/n9Wr0PusX7QEQoxVSVC8viZVHQy4TkVEnPE0Z7itdHrbGjqwhT24i8ruotkxX1GRLr9kyLqLCZBC8rXkGSivgOORYeWdlF1wnqfX6hYQ2GMcYYY4wxZs2wBsMYY4wxxphV0rST99N4g2GMMcYYY8wqsYnUM9hEyhhjjDHGGLNmWINhjDHGGGPMKojQATIuNKzBMMYYY4wxxqwZ1mAsodjg+62pGg8dN1fn4QWPFnm5korfCh2m9pPHJ6lsV46HLKw0+P5ahT0FgLk6j3m5I5+VZRnKIWr/bEGWbU/wsJbpwPuy2uDD/k8nPk5l7+x5H5Vt6tB79bvHeFy+pGj2Up33VybJ71mqiziAAHZ18RCufzl7D5W9JX8TlaVEE5yd588xXa/wggBmRJjWEdHuj03ze36zdIDKbszulvUZFtFE99V47NMtkc/Lo2Koj4kwowMtpt1Ymcv+cbZEZVvaeSjag0Ve2fbA+woA2st87p0VA+ieGg9deVPqCirrENf8duEslQFAW+RtoMIRn2xMUVkv8lT2cPmMrM/9M39AZVfk/z2VqdCvObH45EUY2lZlvzrOQ/xe3clj4x4v8AHbn+VrVr1FbOkHpnl44D2dnVR2VNRne56PD9U2APD3s4ep7MbqDio7U+JzdmMbr8+cCPF7pMjbBgCS4l1aiLx9BpIdVFZr8rX5je0vl/X5/emvSvmLTbQTxtN4g2GMMcYYY8xqiM7kvRSbSBljjDHGGPMDQAjhlSGEfSGER0MIdy35+W0hhCdCCAdDCL/0PK9VWPz/9hBCafG6D4YQ7g4hXKLKWoNhjDHGGGPMKlhw8n5xVRghhB4AvwPgthjj8RDC0OLPkwA+CuC1AE4C+E4I4XMxxse+h8sfijHuXbzeBwD8CoCfYL/sDYYxxhhjjDHnPz8K4DMxxuMAEGMcW/z5jQAOxhgPA0AI4c8BvAXAszYYIYQdAD6Nhf3BF8V9ugBwhzN4g2GMMcYYY8yqWQc+3hcDSIcQvgagE8BHYox/DGAEwIklv3cSwEuWKf8RAB+LMf5xCOHnniPbGULYt3jddlL+abzBMMYYY4wxZn0yEEK4b8m/74gx3kF+NwXgOgCvBtAG4J4Qwr0Algttttx26FYAb1/8+6cA/MYS2VITqR8BcAeA21ilvcEwxhhjjDFmFUTEF8oHYzzGeP1ygkUtw/sX/3k7FjQT4zHGIoBiCOHrAK5e/PmWJUU3AzhN7vd8HuJzAP5Q/YKjSBljjDHGGLNKYlz7P/p+8aMxxr2Lf04D+BsALwshpEIIT5kx7QfwHQC7Qwg7QggZAO/GwibhuXxzUQYAPyZu/VIAh1TdrMFYQlnkJhvo4ImVJqu8Gf9m4hiVvaZ7q6zP8SKX9Sd4wqb9lTEq25LgSY6eiGwzu0B7kyfOac7xWXBvg/sJ7U28hsoew72yPrckXkFlgzneJ0EkFHxP309T2XilSmUnJkVGMwCFwDtzW3KQyr5e4213deNVVLalrV3WR431d3TxZHpjJZ4g6USFJ5nrTPAkUDNhllcGwFiJP8uhCvcx21debu1c4G3d76WywyVdnxMlPrZu6ePJ9FJi3CUDnz8qH+ffn9VJsv5u+reobKDrWirrqLyOyra38bXnT8d/R9Zndye/7rVxJ5W9NHMllX2x9BUqe3n6lVQWwccyAHQGPmaLDZ65rBJ44shv1b9JZTuSN8r63Nr9QSo7W+L1UR8omQQ/Y3y8IF5A0NFyrujoo7KZCl98aqJP7qp+m8q2NGW0TIwlT1HZ3aP/SGW35Xly1Qfn+NqzMcWT9wFAf5O/hx8qjVLZSIK363iZj4HvxG9QWW9iC5UBQBY8s2gz8P6ajPw7ZCv4t0+hpuel0cQY94cQvgjgIQBNAB+PMT4CACGEDwL4EoAkgE/EGB9d5hIfAvDpEMKHAPzVc2RP+WAEAFUA/IMJ3mAYY4wxxhizal7sMLUAEGP8TQC/uczP7wRwZ4uyRwDcvORHH178+VFA7DaXwSZSxhhjjDHGmDXDGgxjjDHGGGNWQQTQXAdxatcL3mAYY4wxxhizSuI6MJFaL9hEyhhjjDHGGLNmWINhjDHGGGPMKnEMrGewBsMYY4wxxhizZliDYYwxxhhjzCqIWB9hatcL3mAsIcnzYMlEWLkkV4q9oZcnlAkQNwRwssiT+6UDVz5lm1kqewwHqeyV7ZfK+hwq8KRe+RQfSjfhNio7Hnhyvz1NnvANALrS/J4dKd62KaG3m63yvtzYlqGyzQne5gAwXuJJCo80eEKid3a/mcrGRHKtdEKPrXuKJ6nsurYRKuvO8sb7anUflf2zDp4UMVPiiQYB4OEar2sq8DHw3oGfpLKjRZ4YsZAoyPqo9nlimidjzKeTVDZb43N9vsn7uVWSwut73k9l5cDnc1YkYEuL+fPOvp+V9TlcnaSyvhy/8EMzPInjHvB1ojvD23xzc4DKAOC4SBTW3+QJFbtjN5VNzj1MZTf0vF7WZx48gZ9iplmisk7wZIKbMjyhIgA80HiSyi5qdlFZUiSc/Gb5f/H6tF9DZb0JnVh0c/IyKntJ7x4qm67xuTeS5sn0HmwekPXZk9xFZZORJ/Cbb/J1QrXda9reQ2W1KLKuArg/8kSEW8IVVDYV+HMMZXdT2UNVvt6vfyKio0g9jU2kjDHGGGOMMWvGutpghBA+EUIYCyE8suRnvxlCeDyE8FAI4bMhhJ7Fn6dDCJ8MITwcQtgfQvjlF6/mxhhjjDHmQqaJuOZ/zlfW1QYDwB8B/8Se5u8AXBFjvArAkwCe2ki8E0A2xnglgOsAfCCEsP37U01jjDHGGGPMcqyrDUaM8esAJp/zsy/HGJ8yPLwXwOanRAA6QggpAG0AqgC0QbIxxhhjjDFrzFNO3tZgLLCuNhjPg58C8LeLf/9LAEUAZwAcB/BfY4zcg9AYY4wxxhjzgnPebDBCCP8OQB3Any7+6EYADQCbAOwA8PMhhItI2Z8JIRwIIZz7vlTWGGOMMcZcUKy9/uL8Td13XoSpDSH8BIA3AXh1fCYG2I8C+GKMsQZgLITwTQDXAzj83PIxxjsA3LF4rfNX32SMMcYYY9Yh57dJ01qz7jUYIYTbAPxbAP8sxrg0cPtxAD8UFugAcBOAx1+MOhpjjDHGGGMWWFcbjBDCnwG4B8AlIYSTIYT3AfgfADoB/F0IYV8I4XcXf/2jAPIAHgHwHQB/GGN86MWotzHGGGOMuXCxk/ezWVcmUjHG5dJN/gH53QIWQtUaY4wxxhhj1gnhQktrHkIyJhIdy8p+5/Kfp+V250tUdqSYo7L7JrmSqNHUbd+fC1RWbvBySlascYehZOD3A4CeLJePl/h11WXvqx6gsqHmBlkfRRJJKtudX77/AeBv579FZZfEq6isPaH36mqafRf3UVmtOU9lb+18DZVNVcQgAFBt8P4abk9TWW9GjxFGUhQ7Oqfr+nh1jMrSkdc1I85PrunpprKqrg6mxS+crvD+6kvxdUKREhNovFaWZXOBt0F3mrddKsHvOVurU1k6aKV4LsnlbSl+z+Pz/DlH2ni7bmrn9yvU9fp79wwfdyOJXirbkuftmuPLEoq8WQHoNWRCzPeOFG+DgRW+YwAgK57leIEXLtb5g+7qyq6oPrMtJu25aoXKupMZKpus83GXDbwB2pL6fXCkOUplIxikso4Uv+7x6jSVqbVwZ3sXlQFATizeNfENc6BYoLLeJJ+ze3p5fwDAfzn0H+6PMV4vf+lFIp8ajlfl373m171n5r+v22dWrCsNhjHGGGOMMecfETGcv1Gf1pp15YNhjDHGGGOMOb+xBsMYY4wxxphV8JSTt1nAGgxjjDHGGGPMmmENhjHGGGOMMaukeR5n3l5rvMEwxhhjjDFmVUREbzCexiZSxhhjjDHGmDXDGgxjjDHGGGNWQQTQdJjap7EGwxhjjDHGGLNmWINhjDHGGGPMKrGT9zNYg2GMMcYYY4xZM6zBMMYYY4wxZlVEazCW4A2GMcYYY4wxqyACDlO7hAtwg9FEs1laVtKZ4gMjm+SydiEbyCaprNSgIgDAP05MUdnerl4q29XJr/n500Uqe9eWvKzP7506TmW3926jssemK1T21t5LqaxYj7I+xwpVKru4O0NlTXHZn+y/mco+M3GIyt7Su5NftAWXNm+hsmKdlxvKcdmXiw/Je7616xoq68oEKntwgrf5azfx5aTY4NcsN/SC/Jq+DVQ2VeGdqfo5LYxDO1K8rgAw3Mbn9J7AJ9/nxsao7LaBYSpTz5Gf53VpxVW9/DlPL79EAgCSpTSVReg5253hDa/afWu+jcrKYmzVI6/PbFXX9bLcAJWlE/yeDXHZXJKXe2KGr5MAsKcnK6R8HOTEEDk4W6OyM/VZWZ/XD/H2qTX5nO5K83VC1fVkgS+GxQZ/DgDY1s7Hz2COj8luUVe1ah0r6LG1J7eFyu4dL1BZp/hkuyrfR2VqbolXJQDgr8dPUtkb+/hzAPx7Ymue16czpdvOnD9cgBsMY4wxxhhj1pKIJlqcHF9A2MnbGGOMMcYYs2ZYg2GMMcYYY8wqsQ/GM3iDYYwxxhhjzCqIiM7kvQSbSBljjDHGGGPWDGswjDHGGGOMWSV28n4GazCMMcYYY4wxa4Y1GMYYY4wxxqyKaCfvJXiDYYwxxhhjzCqIAJrRJlJPYRMpY4wxxhhjzJphDYYxxhhjjDGrwiZSS7EGwxhjjDHGGLNmWINhjDHGGGPMKokOU/s01mAYY4wxxhhznhNC+IUQwr7FP4+EEBohhL5F2W0hhCdCCAdDCL/0PK9XWPz/9hBCafG6D4YQ7g4hXKLKeoNhjDHGGGPMqohovgD/fU81iPE3Y4x7Y4x7AfwygLtijJMhhCSAjwJ4A4A9AN4TQtjzPT7gocVrXw3gkwB+Rf2yTaSMMcYYY4xZBRFYb07e7wHwZ4t/vxHAwRjjYQAIIfw5gLcAeGxpgRDCDgCfxsL+4Ivi2l0AptTNL7gNRggZ5DKblpW1JbntXHe2wmXVNJXd3M+vWWkGKgOAz84cpbLZwgyVXdq1jcp+dU+SysqNqqzPH13dSWWn58tU1p/NUtn1vQUqy6drsj7j5RyVPT7Hy9UilwVw4b/asp3KTpb4NQFgd75OZaMV3idXdfMLf/lsG5XtgT6Y2NPNF8GUGJYbcnzJuPccL/f2rXz+AHx8AMCWNt52c3WuhC03+INMiqE+VhYDBMBl3Vx+3zgv93/v6aCy6Rpvn/EKb/P2lFZCK+lglq9NW9u57OvnMlR2uWgbAEiGlb18swl+3b4MXycyCX6/Yl2//g4X+XPmkrw+XSk+XtPiObZ36PrUmvxZLttYpLKzYp28sZ/PkeEcX18W6jNPZRty/J6pwNtgRwe/5us38LrM1XTb5ZL8/TSY47JKk6/NXzzTTWV5/kkAALixj8/3Vw2p7wJeLiGKNcW0TIr+AICXD/ZR2dfP8bLX8GLyHfytc/ZhWIaBEMJ9S/59R4zxDlUghNAO4DYAH1z80QiAE0t+5SSAlyxT9CMAPhZj/OMQws89R7YzhLAPQCeAdlL+aS64DYYxxhhjjDFrS0R8YRLtjccYr/8ey7wZwDdjjJOL/15uC7rcVu9WAG9f/PunAPzGEtmhRdMrhBB+BMAdWNjELIt9MIwxxhhj/v/27j1GrvK84/jvt7te78W7vuAbN2PC1eRi4hAjigO0ThoaKU1Ig4CiRKWlKCpQ2ipJq1ZtU/UfpKZKkYKaIkQgbUmEoCSkQQQaKXFF0sZADcRAuTrGMWBzCbbXa+9lnv4xZ+3B9fvu2Z3xjGf3+5FW9swz58wzz7zn8s57LkCbsX1tzUndtYfnXK6Dh0dJ1RGLE2senyBpe2K2+WGtqvskXZB7AR0MAAAAoE7NPsk7Im6eOKk7IrZLku35ki6U9J2al26UdJrtk213q9oBue8ws3y4iEnSlZm3XifphVxuHCIFAAAA1CWOlvtgXCLpwYg4cHJWRIzZvk7S9yV1SrotIjYfZtobJN1p+wZJ9xwSmzgHw5JGJF2dS4IOBgAAADADRMTtkm4/zPP3S7p/kmlfknRezVM3Fs9vkZS/8sMh2uYQKdsLbN9t+xnbT9s+ryb2edthe3ErcwQAAMDsE5IiKg3/a1ftNIJxk6QHIuLTxfFjfZJk+0RJH5G0tZXJAQAAAGiTDobtQVXPVv8dSYqIEVWP/5Kkr0j6ot55MgsAAADQJDHlO2/PZG3RwZD0Lkk7JX3d9mpJj6p6Isp6Sb+IiMft/E3rAAAAgCMidKTug9GW2uUcjC5Ja1S9u+D7JQ1J+pKkv5D0V5NNbPsa28/Z3hmRvssqAAAAgPq0Swdjm6RtEfHfxeO7Ve1wnCzpcdtbVL1pyGO2lx86cUTcEhGnRcQSu10GbQAAANAeQqFKw6zULg8AAA4ESURBVP/aVVt0MCLiVUkv2z6jeGq9pMciYmlErIyIlap2QtYUrwUAAADQAu30c/71kv61uILUi5KuanE+AAAAwIHL1KKqbToYEbFJ0jmZ+MrmZQMAAADgcNqmgwEAAAAcnUIhriI1gQ4GAAAAUCcOkTrIEdHqHJrKdqT6Vd9d+8XkdB9atS0Z6+hM17B7UTrWtWROMiZJmju9/p87MvcEycRiX/4SvjGcjleG0gvV+J50DToH0vl0Hdubzcf93dl4SoxnVgAj6V8fYu9oMlYZzv9qMbYzHR/Znb7WQrZtLZ7+suvOdN07+qZ37YfK3nRd9+9M5/rL1/qy831zT7odrDjurWSs55h0Pp39uWUkm8607X8tXYOHNq1MxtYsfT0ZW7hkb/Y9u/qmt7HrnJuJzU8XKEbybXJ8KB2f7gX+xofTsdzn6FqWX//mlpEYT3+Oyq70enL/a+n3e3NHfjno7UmvfwZXjCRjczKfM7cvNLIt/X6S9ObL6eVycNG+9HtGuq47X52XjB1/2q5kbM6xkzSeSvr72vV0erJ5KzLrkKWZxpXbBk/CPenPMu1t3lC6fYy/MpSdds/z6VjnnHRdOzKx3LJeyTc7Dd76L49GRPJw+Vbq6uyLwb7TGz7ft/Y8ftR+5hxGMAAAAIA6MYJxUFtcphYAAABAe2AEAwAAAKhD9TZ7jGBMoIMBAAAA1IlDpA7iECkAAAAADcMIBgAAAFCPCEVwH4wJjGAAAAAAaBhGMAAAAIA6BSd5H0AHAwAAAKhLcJJ3DQ6RAgAAANAwjGAAAAAAdQhxmdpajGAAAAAAaBhGMAAAAIA6cZL3QYxgAAAAAGgYRjBqdDmSsd5j07GuE/qSMQ/MTb/hMQP5hDqcjlXS+WTtG02GvHs4O+n49t3pYGc61H3inPR7Zurj5fOz+ag3U9ue7vR8c/McS98kx6Pp2nW8NZSbqzqXpGv79n+k57vo9JFkrOu4TLtbPpjNRx2N/20hV4POBXvTE1by7W7n7v5kbN6p6ek6j00vX57fm55wIBOTsm1Ew+nva86p6e/5N096ORmLsfSy3rks3Qaqb5peMN2dWWj7e/LzTenOb1LmVDK/7r2dbiPx9r5kbNfGdKxnZfozdiyfZP07P1PbzPq3Y8eudKwv/RkH9qU/hyT1LU23u7nvXpCecCDzXWa+r953pdurJB33Rnp599zMursrve4Z2L0/Pc+FS9LzXJheR0jKbvcWnpFZnnPteW56G6O+zLZJym+/OzPLZW6fYHQsGfK+zHZkUb52C1ak22WMZraXmXVPDKXzGX8jvxwc3biKVC06GAAAAECd6GAcxCFSAAAAABqGEQwAAACgLiFxkvcBjGAAAAAAaBhGMAAAAIB6BOdg1KKDAQAAANQhxH0wanGIFAAAAICGYQQDAAAAqAv3wajFCAYAAACAhmEEAwAAAKhb5s7wswwdDAAAAKAuHCJVi0OkAAAAADQMIxgAAABA3RjBmMAIBgAAAICGYQQDAAAAqEtInINxAB2MGs7FetODPe6dk4xt+KeeZOyCv86XP557Nf2ey+enp9v+y/R0/d3J2Ohzb2fz2b2lMxnbNzw3GevpHUvGBlcNJ2OdHblvJF939afrrp7MdG/sTseWL0yGtn5zb3o6SSsuTddnybn7k7EYS9f8gr/sS8b+845JFu2R9HdS2fpWMtZx+rJk7NF/Tn/GNZ9Mf8aOOaPJmCSNjKeXvY7+zOfsTE+37+EdyVjPpauy+WhnejmpvLIrGcvVrmM0feWRh+/oT8bWXZOuqyTFcLq2lVfTbbZjzTHp6R7bkp5u1bHZfDR/XjI09L1tyVjf+9PTDXwg3e5y67uxF9PrSUnqOiuznpibjm2+K/2eq9an13ebf74km897O19LpzMvXYPK9nR77Vg2kH7DRZmYpOEfpvPpW5f5LAvT7bmy9eVkrHNhen2n/t50TNLohp8nY3PWrUhPmNkGVZ5I59px4buz+YzcvSkZ6/7U+9ITPpdeRrRkMBmqPJPel+g4YUF6nlK2BuO/GErGulYvT8bGnk+3yfGhyOeDtkEHAwAAAKhTiA7SBDoYAAAAQN04RGpC25/kbfti2/9r+3nbf9bqfAAAAIBmsz3f9ndtP257s+2ramJT3l+2vaf4d6XtYdubinn/2PYZuWnbuoNhu1PSzZJ+Q9JZkq6wfVZrswIAAMCsE9H4v6m5VtJTEbFa0kWS/t52d4P2l1+IiLOLed8h6c9zL273Q6TWSno+Il6UJNvfkvQJSU+1NCsAAACguULSgG1LmifpTUljks5Vif1l2ydLulPV/sEDmfcZlJS+Iozav4NxvKTaSzlsU7WIAAAAQJPE0XCS91cl3Sdpu6QBSZdFRMV22f3lmyT9Y0R8w/a1h8ROsb2pmG9fYvoDHFMffjlq2L5U0kcj4uri8WckrY2I6w953TWSviBpgaT5kp5odq6z3GJJr7c6iVmGmjcfNW8u6t181Lz5qPk7nRQR+Ws6t4jtB1T9vhqtR9K+mse3RMQtiRw+Lel8SX8i6RRJD0laLemjKre//Iak5RExantQ0vaImGd7paR/j4j3FK+7TNJVEXFxKul2H8HYJunEmscnqNpre4fii7hFkmw/EhHnNCc9SNS8Fah581Hz5qLezUfNm4+at4/czvaRUowy/H7x8GOSrpJ0Y1RHD563/ZKkM1Vyf7lQZuThPklfz72grU/ylrRR0mm2T7bdLelyVT80AAAAMGNFxM3FiddnR8R2SVslrZck28sknSHpRZXfX364iEnSlZm3XifphVxubT2CERFjtq+T9H1JnZJui4jNLU4LAAAAaLa/lXS77SclWdKfRsTrklRyf/kGSXfavkHSPYfEJs7BsKQRSVfnEmnrDoYkRcT9ku6fwiSHPW4NRxQ1bz5q3nzUvLmod/NR8+aj5iitGMX49URs0v3liHhJ0nk1T91YPL9FUu9Ucmnrk7wBAAAAHF3a/RwMAAAAAEeRWdXBmM5t0pE2WT1tn2n7J7b32/78IbEttp8sbjv/SPOynjlK1P8Ttp+YqLHtda3Is92VXW/Y/qDt8eIygRPP0c7rVKb+ti8qarzZ9o+aneNMUGJ98oWixpts/6xo64uKGO28TiXqv9D2vcU6/ae239OKPIGyZs0hUsVt0p+V9BFVL9e1UdIVEcFdv6ehTD1tL5V0kqRPSnorIr5cE9si6ZyJk48wNSXrP0/SUESE7fdJuisizmxJwm2q7HqjeN1Dql6r/LaIuLt4foto59NWsp0vkPRjSRdHxFbbSyNiR0sSblNT3T7a/rikP46IXysebxHtfNpKtvO/k7QnIv7G9pmSbo6I9S1JGChhNo1grFVxm/SIGJE0cZt0TM+k9YyIHRGxUdJoKxKc4crUf08c/AWhX+WubY13KrveuF7VK26wY9tYZer/25L+LSK2StX1TpNznAmmun28QtI3m5LZ7FCm/mdJ+oEkRcQzklYWlyEFjkqzqYNxuNukH9+iXGaCeusZkh60/Whxp3VMTan6277E9jOSvifpd5uU20wyaZ1tHy/pEklfO8z0tPP6lGnnp0taaPuHRZ0/27TsZo7S63PbfZIu1jsvYUk7r0+Z+j8u6VOSZHutqkcHnNCU7IBpaPvL1E6BD/Mcv+hOX731PD8itheHUT1k+5mI2NCg3GaDUvWPiHsl3Wv7AlWvj/3hI53YDFOmzv+g6rXGx+3/93LaeX3K1L9L0gdUvblUr6Sf2P6viHj2SCc3g0xlff5xSQ9HxJs1z9HO61Om/jdKuqm4D8GTkv5H0tiRTgyYrtnUwZjKbdIxubrqWVyrWRGxw/a9qg4Rs0Eqb0r1j4gNtk+xvZjjpKekTJ3PkfStonOxWNLHbI9FxLdp53UrU/9tkl6PiCFJQ7Y3SFqt6jHtKGcq65PLdcjhUbTzuk1a/4jYJekqSXJ1ZfNS8QcclWbTIVJlb5OOcqZdT9v9tgcm/q/qTWF+dsQynZkmrb/tU4sNkWyvkdQt6Y2mZ9reJq1zRJwcESsjYqWkuyX9QUR8m3beEGXWM9+R9CHbXcXhO+dKerrJeba7Uutz2/MlXahqzSeeo53Xr8z6fEERk6p3UN5QdDqAo9KsGcGIiDGXu006SkjV0/bnivjXbC+X9IikQUkV23+k6olqi1U9bEeqtsE7I+KBVnyOdlWm/pJ+S9JnbY9KGpZ0Wc1J3yihZJ1Tlol2Xpcy9Y+Ip20/IOkJSRVJt0YEO7hTMIV2fomkB4vRogm08zqVrP8qSd+wPS7pKUm/17KEgRJmzWVqAQAAABx5s+kQKQAAAABHGB0MAAAAAA1DBwMAAABAw9DBAAAAANAwdDAAAAAANMysuUwtALQb28dI+kHxcLmkcUk7i8d7I+JXWpIYAAAZXKYWANqA7S9J2hMRX251LgAA5HCIFAC0Idt7in8vsv0j23fZftb2jbavtP1T20/aPqV43RLb99jeWPyd39pPAACYqehgAED7Wy3pBknvlfQZSadHxFpJt0q6vnjNTZK+EhEfVPUu77e2IlEAwMzHORgA0P42RsQrkmT7BUkPFs8/KelXi/9/WNJZtiemGbQ9EBG7m5opAGDGo4MBAO1vf83/KzWPKzq4nu+QdF5EDDczMQDA7MMhUgAwOzwo6bqJB7bPbmEuAIAZjA4GAMwOfyjpHNtP2H5K0udanRAAYGbiMrUAAAAAGoYRDAAAAAANQwcDAAAAQMPQwQAAAADQMHQwAAAAADQMHQwAAAAADUMHAwAAAEDD0MEAAAAA0DB0MAAAAAA0zP8Bwg8hvSM0lKUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1008x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# STFT Example\n",
    "y, sr = librosa.load(files[10], sr=fs, duration=1)\n",
    "y/=y.max() #Normalize\n",
    "duration_in_samples=librosa.time_to_samples(1, sr=fs)\n",
    "y_pad = librosa.util.fix_length(y, duration_in_samples) #Pad to 1s if smaller\n",
    "y_stft=librosa.core.stft(y_pad, n_fft=n_fft, hop_length=hop_length)\n",
    "y_spec=librosa.amplitude_to_db(abs(y_stft), np.max)\n",
    "plt.figure(figsize=(14,8))\n",
    "plt.title(\"Short-Time Fourier Transform Spectogram \\n %s\"%files[0])\n",
    "librosa.display.specshow(y_spec,sr=fs,y_axis='log', x_axis='time')\n",
    "plt.colorbar(format='%+2.0f dB');\n",
    "print(\"Spectogram Array Shape:\",y_spec.shape)\n",
    "ipd.Audio(y, rate=fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Encode Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6 classes: cello, flute, oboe, sax, trumpet, viola\n"
     ]
    }
   ],
   "source": [
    "# Encode Labels\n",
    "labelencoder = LabelEncoder()\n",
    "labelencoder.fit(labels)\n",
    "print(len(labelencoder.classes_), \"classes:\", \", \".join(list(labelencoder.classes_)))\n",
    "classes_num = labelencoder.transform(labels)\n",
    "\n",
    "#OneHotEncoding\n",
    "encoder=OneHotEncoder(sparse=False, categories=\"auto\")\n",
    "onehot_labels=encoder.fit_transform(classes_num.reshape(len(classes_num),1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train and Test Sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create Train and Test Sets\n",
    "splitter = StratifiedShuffleSplit(n_splits=1, test_size=testset_size, random_state=0)\n",
    "splits = splitter.split(files, onehot_labels)\n",
    "files_arr=np.array(files)\n",
    "\n",
    "for train_index, test_index in splits:\n",
    "    train_set_files = files_arr[train_index]\n",
    "    test_set_files = files_arr[test_index]\n",
    "    train_classes = onehot_labels[train_index]\n",
    "    test_classes = onehot_labels[test_index]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convolutional Neural Networks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<center><iframe src=\"https://en.wikipedia.org/wiki/Convolutional_neural_network\" width=\"800\" height=\"600\" frameborder=\"0\" marginheight=\"0\" marginwidth=\"0\">Loading...</iframe></center>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_web(\"https://en.wikipedia.org/wiki/Convolutional_neural_network\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# CNN Model\n",
    "model = Sequential()\n",
    "\n",
    "conv_filters =  16  # number of convolution filters\n",
    "\n",
    "# Layer 1\n",
    "model.add(Convolution2D(conv_filters, 3,input_shape=(1025, 87, 1)))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2))) \n",
    "model.add(Dropout(0.40)) \n",
    "\n",
    "# Layer 2\n",
    "model.add(Convolution2D(conv_filters, 3))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.40))\n",
    "\n",
    "# Flatten\n",
    "model.add(Flatten()) \n",
    "\n",
    "# Full layer\n",
    "model.add(Dense(16, activation='sigmoid')) \n",
    "\n",
    "# Output layer\n",
    "model.add(Dense(6,activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 1023, 85, 16)      160       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 511, 42, 16)       0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 511, 42, 16)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 509, 40, 16)       2320      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 254, 20, 16)       0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 254, 20, 16)       0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 81280)             0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 16)                1300496   \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 6)                 102       \n",
      "=================================================================\n",
      "Total params: 1,303,078\n",
      "Trainable params: 1,303,078\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loss Function \n",
    "loss = 'categorical_crossentropy' \n",
    "\n",
    "# Optimizer = Gradient Descent\n",
    "optimizer = 'sgd' \n",
    "\n",
    "# Compile\n",
    "model.compile(loss=loss, optimizer=optimizer, metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def featureGenerator(files, labels):\n",
    "    while True:\n",
    "        for i,f in enumerate(files):\n",
    "            try:\n",
    "                feature_vectors = []\n",
    "                label = []\n",
    "                y, sr = librosa.load(f, sr=fs)\n",
    "                if len(y) < 2:\n",
    "                    print(\"Error loading %s\" % f)\n",
    "                    continue\n",
    "                y, index = librosa.effects.trim(y,top_db=60) #Trim\n",
    "                y = normalize(y)\n",
    "                duration_in_samples=librosa.time_to_samples(1, sr=fs)\n",
    "                y_pad = librosa.util.fix_length(y, duration_in_samples) #Pad/Trim to same duration\n",
    "                y_stft=librosa.core.stft(y_pad, n_fft=n_fft, hop_length=hop_length)\n",
    "                y_spec=librosa.amplitude_to_db(abs(y_stft), np.min)\n",
    "                scaler = StandardScaler()\n",
    "                dtype = K.floatx()\n",
    "                data = scaler.fit_transform(y_spec).astype(dtype)\n",
    "                data = np.expand_dims(data, axis=0)\n",
    "                data = np.expand_dims(data, axis=3)\n",
    "                feature_vectors.append(data)\n",
    "                label.append([labels[i]])\n",
    "                yield feature_vectors, label\n",
    "            except Exception as e:\n",
    "                print(\"Error loading %s. Error: %s\" % (f,e))\n",
    "                raise\n",
    "                break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Not inside Google Colab: No module named 'google.colab'. Using standard configurations.\n",
      "Epoch 1/30\n",
      "450/450 [==============================] - 82s 183ms/step - loss: 1.1988 - acc: 0.6200 - val_loss: 0.8557 - val_acc: 0.7867\n",
      "\n",
      "Epoch 00001: val_acc improved from -inf to 0.78667, saving model to best_model.h5\n",
      "Epoch 2/30\n",
      "450/450 [==============================] - 82s 183ms/step - loss: 0.6286 - acc: 0.8622 - val_loss: 0.5856 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00002: val_acc improved from 0.78667 to 0.82000, saving model to best_model.h5\n",
      "Epoch 3/30\n",
      "450/450 [==============================] - 83s 184ms/step - loss: 0.3707 - acc: 0.9600 - val_loss: 0.4101 - val_acc: 0.9133\n",
      "\n",
      "Epoch 00003: val_acc improved from 0.82000 to 0.91333, saving model to best_model.h5\n",
      "Epoch 4/30\n",
      "450/450 [==============================] - 82s 183ms/step - loss: 0.1949 - acc: 0.9911 - val_loss: 0.2499 - val_acc: 0.9667\n",
      "\n",
      "Epoch 00004: val_acc improved from 0.91333 to 0.96667, saving model to best_model.h5\n",
      "Epoch 5/30\n",
      "450/450 [==============================] - 83s 184ms/step - loss: 0.1132 - acc: 1.0000 - val_loss: 0.1741 - val_acc: 0.9867\n",
      "\n",
      "Epoch 00005: val_acc improved from 0.96667 to 0.98667, saving model to best_model.h5\n",
      "Epoch 6/30\n",
      "450/450 [==============================] - 82s 183ms/step - loss: 0.0806 - acc: 1.0000 - val_loss: 0.1424 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00006: val_acc improved from 0.98667 to 0.99333, saving model to best_model.h5\n",
      "Epoch 7/30\n",
      "450/450 [==============================] - 83s 184ms/step - loss: 0.0594 - acc: 1.0000 - val_loss: 0.1288 - val_acc: 0.9867\n",
      "\n",
      "Epoch 00007: val_acc did not improve from 0.99333\n",
      "Epoch 8/30\n",
      "450/450 [==============================] - 83s 185ms/step - loss: 0.0503 - acc: 1.0000 - val_loss: 0.1375 - val_acc: 0.9733\n",
      "\n",
      "Epoch 00008: val_acc did not improve from 0.99333\n",
      "Epoch 9/30\n",
      "450/450 [==============================] - 83s 184ms/step - loss: 0.0418 - acc: 1.0000 - val_loss: 0.1053 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00009: val_acc did not improve from 0.99333\n",
      "Epoch 10/30\n",
      "450/450 [==============================] - 83s 184ms/step - loss: 0.0371 - acc: 1.0000 - val_loss: 0.0912 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00010: val_acc did not improve from 0.99333\n",
      "Epoch 11/30\n",
      "450/450 [==============================] - 84s 186ms/step - loss: 0.0327 - acc: 1.0000 - val_loss: 0.0840 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00011: val_acc did not improve from 0.99333\n",
      "Epoch 12/30\n",
      "450/450 [==============================] - 84s 186ms/step - loss: 0.0294 - acc: 1.0000 - val_loss: 0.0999 - val_acc: 0.9867\n",
      "\n",
      "Epoch 00012: val_acc did not improve from 0.99333\n",
      "Epoch 13/30\n",
      "450/450 [==============================] - 96s 213ms/step - loss: 0.0267 - acc: 1.0000 - val_loss: 0.0799 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00013: val_acc did not improve from 0.99333\n",
      "Epoch 14/30\n",
      "450/450 [==============================] - 88s 197ms/step - loss: 0.0235 - acc: 1.0000 - val_loss: 0.0754 - val_acc: 0.9933\n",
      "\n",
      "Epoch 00014: val_acc did not improve from 0.99333\n",
      "Epoch 15/30\n",
      "450/450 [==============================] - 89s 199ms/step - loss: 0.0220 - acc: 1.0000 - val_loss: 0.0789 - val_acc: 0.9867\n",
      "Restoring model weights from the end of the best epoch\n",
      "\n",
      "Epoch 00015: val_acc did not improve from 0.99333\n",
      "Epoch 00015: early stopping\n",
      "CPU times: user 12min 52s, sys: 15min 58s, total: 28min 51s\n",
      "Wall time: 21min 8s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "hist = History();\n",
    "es = EarlyStopping(monitor='val_acc', min_delta=0.01, restore_best_weights=True, patience= 10, verbose=1 )\n",
    "mc = ModelCheckpoint('best_model.h5', monitor='val_acc',save_best_only=True, verbose=1)\n",
    "\n",
    "\n",
    "try:\n",
    "    import google.colab\n",
    "    tbc=TensorBoardColab()\n",
    "    callbacksKeras=[hist,es,mc,TensorBoardColabCallback(tbc)]\n",
    "\n",
    "except Exception as e:\n",
    "    callbacksKeras=[hist,es,mc]\n",
    "    print(\"Not inside Google Colab: %s. Using standard configurations.\" % (e))\n",
    "\n",
    "\n",
    "model.fit_generator(featureGenerator(train_set_files, train_classes), \n",
    "                    validation_data=(featureGenerator(test_set_files, test_classes)), \n",
    "                    validation_steps=150, \n",
    "                    steps_per_epoch=450,epochs=30,callbacks=callbacksKeras, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_history(history):\n",
    "    loss_list = [s for s in history.history.keys() if 'loss' in s and 'val' not in s]\n",
    "    val_loss_list = [s for s in history.history.keys() if 'loss' in s and 'val' in s]\n",
    "    acc_list = [s for s in history.history.keys() if 'acc' in s and 'val' not in s]\n",
    "    val_acc_list = [s for s in history.history.keys() if 'acc' in s and 'val' in s]\n",
    "    \n",
    "    if len(loss_list) == 0:\n",
    "        print('Loss is missing in history')\n",
    "        return \n",
    "    \n",
    "    ## As loss always exists\n",
    "    epochs = range(1,len(history.history[loss_list[0]]) + 1)\n",
    "    \n",
    "    ## Loss\n",
    "    plt.figure(1)\n",
    "    for l in loss_list:\n",
    "        plt.plot(epochs, history.history[l], 'b', label='Training loss')\n",
    "    for l in val_loss_list:\n",
    "        plt.plot(epochs, history.history[l], 'g', label='Validation loss')\n",
    "    \n",
    "    plt.title('Loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    plt.legend()\n",
    "    \n",
    "    ## Accuracy\n",
    "    plt.figure(2)\n",
    "    for l in acc_list:\n",
    "        plt.plot(epochs, history.history[l], 'b', label='Training accuracy')\n",
    "    for l in val_acc_list:    \n",
    "        plt.plot(epochs, history.history[l], 'g', label='Validation accuracy')\n",
    "\n",
    "    plt.title('Accuracy')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_history(hist)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "150/150 [==============================] - 18s 118ms/step\n"
     ]
    }
   ],
   "source": [
    "saved_model = load_model('best_model.h5')\n",
    "test_pred = saved_model.predict_generator(featureGenerator(test_set_files, test_classes), steps=150,verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_round=np.around(test_pred).astype('int');\n",
    "predictions_int=np.argmax(predictions_round,axis=1);\n",
    "predictions_labels=labelencoder.inverse_transform(np.ravel(predictions_int));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall:  [1.   1.   1.   1.   1.   0.92]\n",
      "Precision:  [0.96153846 1.         1.         0.96153846 1.         1.        ]\n",
      "F1-Score:  [0.98039216 1.         1.         0.98039216 1.         0.95833333]\n",
      "Accuracy: 0.99  , 148\n",
      "Number of samples: 150\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      1.00      0.98        25\n",
      "           1       1.00      1.00      1.00        25\n",
      "           2       1.00      1.00      1.00        25\n",
      "           3       0.96      1.00      0.98        25\n",
      "           4       1.00      1.00      1.00        25\n",
      "           5       1.00      0.92      0.96        25\n",
      "\n",
      "   micro avg       0.99      0.99      0.99       150\n",
      "   macro avg       0.99      0.99      0.99       150\n",
      "weighted avg       0.99      0.99      0.99       150\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Recall - the ability of the classifier to find all the positive samples\n",
    "print(\"Recall: \", recall_score(classes_num[test_index], predictions_int,average=None))\n",
    "\n",
    "# Precision - The precision is intuitively the ability of the classifier not to \n",
    "#label as positive a sample that is negative\n",
    "print(\"Precision: \", precision_score(classes_num[test_index], predictions_int,average=None))\n",
    "\n",
    "# F1-Score - The F1 score can be interpreted as a weighted average of the precision \n",
    "#and recall\n",
    "print(\"F1-Score: \", f1_score(classes_num[test_index], predictions_int, average=None))\n",
    "\n",
    "# Accuracy - the number of correctly classified samples\n",
    "print(\"Accuracy: %.2f  ,\" % accuracy_score(classes_num[test_index], predictions_int,normalize=True), accuracy_score(classes_num[test_index], predictions_int,normalize=False) )\n",
    "print(\"Number of samples:\",classes_num[test_index].shape[0])\n",
    "\n",
    "print(classification_report(classes_num[test_index], predictions_int))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compute confusion matrix\n",
    "cnf_matrix = confusion_matrix(classes_num[test_index], predictions_int)\n",
    "np.set_printoptions(precision=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to Plot Confusion Matrix\n",
    "# http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "    \"\"\"\n",
    "    #print(cm)\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot non-normalized confusion matrix\n",
    "plt.figure(figsize=(16,12))\n",
    "plot_confusion_matrix(cnf_matrix, classes=labelencoder.classes_,\n",
    "                      title='Confusion matrix, without normalization')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Find wrong predicted samples indexes\n",
    "wrong_predictions = [i for i, (e1, e2) in enumerate(zip(classes_num[test_index], predictions_int)) if e1 != e2]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['viola' 'viola']\n",
      "['cello' 'sax']\n",
      "['./audio/london_phill_dataset_multi/viola/viola_Ds3_05_mezzo-piano_arco-normal.mp3'\n",
      " './audio/london_phill_dataset_multi/viola/viola_Gs3_15_piano_arco-normal.mp3']\n"
     ]
    }
   ],
   "source": [
    "# Find wrong predicted audio files\n",
    "print(np.array(labels)[test_index[wrong_predictions]])\n",
    "print(predictions_labels[wrong_predictions].T)\n",
    "print(np.array(files)[test_index[wrong_predictions]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}