07_SimpleXml.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "SimpleXml7.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "metadata": {
        "id": "VrIDlK2mvNsS",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "\n",
        "import numpy as np # for multidimensional arrays"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "8tEeyPXoDTBA",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Constants\n",
        "INPUT_WIDTH = 25\n",
        "\n",
        "# Word index\n",
        "word_index = {}\n",
        "word_index[\"<PAD>\"] = 0\n",
        "word_index[\"<VALUE UP>\"] = 1\n",
        "word_index[\"<VALUE DOWN>\"] = 2\n",
        "word_index[\"<STRING CHANGE>\"] = 3\n",
        "word_index[\"<CHILD CHANGE>\"] = 4\n",
        "word_index[\"<NUMBER CHANGE>\"] = 5\n",
        "word_index[\"<UNKNOWN>\"] = 6\n",
        "word_index[\".\"] = 7\n",
        "word_index[\"entry\"] = 8\n",
        "word_index[\"document\"] = 9\n",
        "word_index[\"validationxml\"] = 10\n",
        "word_index[\"browserxml\"] = 11\n",
        "word_index[\"arrangement\"] = 12\n",
        "word_index[\"text()\"] = 13\n",
        "word_index[\"@label\"] = 14\n",
        "word_index[\"@chronoid\"] = 15\n",
        "word_index[\"@type\"] = 16\n",
        "word_index[\"@rtti\"] = 17\n",
        "word_index[\"base\"] = 18\n",
        "word_index[\"dlitem\"] = 19\n",
        "word_index[\"node\"] = 20\n",
        "word_index[\"member\"] = 21\n",
        "word_index[\"displayxml\"] = 22\n",
        "word_index[\"sheets\"] = 23\n",
        "word_index[\"sheet\"] = 24\n",
        "word_index[\"root\"] = 25\n",
        "word_index[\"@Y\"] = 26\n",
        "word_index[\"attr\"] = 27\n",
        "word_index[\"textstylemanager\"] = 28\n",
        "word_index[\"textstyle\"] = 29\n",
        "word_index[\"Entry\"] = 30\n",
        "word_index[\"@X\"] = 31\n",
        "word_index[\"renderstylexml\"] = 32\n",
        "word_index[\"bodychecksum\"] = 33\n",
        "word_index[\"bodychecksumxml\"] = 34\n",
        "word_index[\"executablexml\"] = 35\n",
        "word_index[\"BrowInfo\"] = 36\n",
        "word_index[\"transform\"] = 37\n",
        "word_index[\"primattrstack\"] = 38\n",
        "word_index[\"primattr\"] = 39\n",
        "word_index[\"renderstyletable\"] = 40\n",
        "word_index[\"style\"] = 41\n",
        "word_index[\"attributesxml\"] = 42\n",
        "word_index[\"attributesets\"] = 43\n",
        "word_index[\"attributeset\"] = 44\n",
        "word_index[\"attribute\"] = 45\n",
        "word_index[\"occurrence\"] = 46\n",
        "word_index[\"@style\"] = 47\n",
        "word_index[\"feaxml\"] = 48\n",
        "word_index[\"fearesults\"] = 49\n",
        "word_index[\"simulationresults\"] = 50\n",
        "word_index[\"resultsummary\"] = 51\n",
        "word_index[\"RelativeReactionForces\"] = 52\n",
        "word_index[\"constraintReactionForce\"] = 53\n",
        "word_index[\"@maxComponent\"] = 54\n",
        "word_index[\"propertiesxml\"] = 55\n",
        "word_index[\"propertysets\"] = 56\n",
        "word_index[\"propertyset\"] = 57\n",
        "word_index[\"property\"] = 58\n",
        "word_index[\"@value\"] = 59\n",
        "word_index[\"EndTreatments\"] = 60\n",
        "word_index[\"EndTreatment\"] = 61\n",
        "word_index[\"Lengthen\"] = 62\n",
        "word_index[\"workgeometryxml\"] = 63\n",
        "word_index[\"workplanes\"] = 64\n",
        "word_index[\"@blue\"] = 65\n",
        "word_index[\"@green\"] = 66\n",
        "word_index[\"@red\"] = 67\n",
        "word_index[\"@ambient\"] = 68\n",
        "word_index[\"@diffuse\"] = 69\n",
        "word_index[\"@internalname\"] = 70\n",
        "word_index[\"definitiondata\"] = 71\n",
        "word_index[\"testingaddin\"] = 72\n",
        "word_index[\"apixml\"] = 73\n",
        "word_index[\"assemblychecksumxml\"] = 74\n",
        "word_index[\"component\"] = 75\n",
        "word_index[\"bodyoverrides\"] = 76\n",
        "word_index[\"bodyoverride\"] = 77\n",
        "word_index[\"@count\"] = 78\n",
        "word_index[\"FDesign\"] = 79\n",
        "word_index[\"Object\"] = 80\n",
        "word_index[\"Calc\"] = 81\n",
        "word_index[\"thl\"] = 82\n",
        "word_index[\"@Val\"] = 83\n",
        "word_index[\"dcxml\"] = 84\n",
        "word_index[\"features\"] = 85\n",
        "word_index[\"feature\"] = 86\n",
        "word_index[\"@health\"] = 87\n",
        "word_index[\"@Z\"] = 88\n",
        "\n",
        "# Train data\n",
        "train_data_raw = []; train_labels = []; train_ids = []\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 67]);                         train_labels.append(0); train_ids.append(\"dm_at_appendfilenameinviewblockname_0004#2\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 66]);                         train_labels.append(0); train_ids.append(\"dm_at_appendfilenameinviewblockname_0004#1\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 74, 75, 76, 77, 33, 13]);                                                        train_labels.append(0); train_ids.append(\"feat_807811#0\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 26]);                         train_labels.append(0); train_ids.append(\"907031#1\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 65]);                         train_labels.append(0); train_ids.append(\"dm_at_appendfilenameinviewblockname_0004#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 14]);                                                              train_labels.append(1); train_ids.append(\"break link_catia_part1#1\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 19, 21, 26]);                     train_labels.append(0); train_ids.append(\"tc5356#5\")\n",
        "train_data_raw.append([4, 1, 7, 10, 72, 73]);                                                                           train_labels.append(0); train_ids.append(\"colorschemescoll_001#0\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 18, 27, 27, 28, 29, 21, 21, 13]);                 train_labels.append(0); train_ids.append(\"idw_etsy_d1207#0\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 42, 43, 44, 45, 60, 61, 62, 13]);                                                train_labels.append(0); train_ids.append(\"curvedframe_trimtoframe_spline#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 34, 33, 13]);                                                                    train_labels.append(0); train_ids.append(\"bug116303#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 35, 36, 30, 30, 30, 30, 30, 30, 30]);                                               train_labels.append(0); train_ids.append(\"md_mbs_c1213#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 48, 49, 50, 51, 52, 53, 54]);                                                    train_labels.append(0); train_ids.append(\"cylinder-30x1000x1500n#0\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 26]);                         train_labels.append(0); train_ids.append(\"tc5356#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 55, 56, 57, 58, 59]);                                                            train_labels.append(0); train_ids.append(\"curvedframe_trimtoframe_arc#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 18, 19, 19, 19, 19, 19]);                 train_labels.append(0); train_ids.append(\"idw_replacemodelreference_c1102#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 32, 46, 47]);                                                                    train_labels.append(0); train_ids.append(\"windshield wiper#6\")\n",
        "train_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 14]);                                                              train_labels.append(1); train_ids.append(\"bimcompexptbuildingcompmtd_002#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 9, 63, 64]);                                                                        train_labels.append(0); train_ids.append(\"parker union tee 164ca-264ca 164ca-2#1\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 14]);                                                           train_labels.append(1); train_ids.append(\"break link_catia_part1#0\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 15]);                                                        train_labels.append(1); train_ids.append(\"tp_change_diameter#0\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 31]);                         train_labels.append(0); train_ids.append(\"907031#10\")\n",
        "train_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8]);                                                         train_labels.append(1); train_ids.append(\"rt_core cavity design_1170823#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 35, 36, 30, 30, 30, 30, 30]);                                                       train_labels.append(0); train_ids.append(\"md_mbs_c1213#3\")\n",
        "train_data_raw.append([3, 2, 7, 10, 9, 32, 40, 41, 68]);                                                                train_labels.append(0); train_ids.append(\"windshield wiper#0\")\n",
        "train_data_raw.append([3, 2, 7, 10, 9, 32, 40, 41, 69]);                                                                train_labels.append(0); train_ids.append(\"windshield wiper#1\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 15]);                                                        train_labels.append(1); train_ids.append(\"tp_flip_orientation#0\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 14]);                                                        train_labels.append(1); train_ids.append(\"tp_flip_orientation#2\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 42, 43, 44, 45, 79, 80, 81, 82, 83]);                                            train_labels.append(0); train_ids.append(\"dac_chains_calc_10201020#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8]);                                                               train_labels.append(1); train_ids.append(\"rt_core cavity design_1170823#2\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 32, 40, 41, 70]);                                                                train_labels.append(0); train_ids.append(\"windshield wiper#2\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 63, 64, 78]);                                                                    train_labels.append(0); train_ids.append(\"parker union tee 164ca-264ca 164ca-2#0\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 15]);                                                              train_labels.append(1); train_ids.append(\"break link_catia_part1#3\")\n",
        "train_data_raw.append([3, 1, 7, 10, 9, 84, 85, 86, 87]);                                                                train_labels.append(0); train_ids.append(\"smmsb_createpunch#0\")\n",
        "train_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 31]);                         train_labels.append(0); train_ids.append(\"592453#0\")\n",
        "train_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 18, 19, 19, 19, 19, 19, 19, 21, 88]);     train_labels.append(0); train_ids.append(\"dm_migr_r11_coolingtower#0\")\n",
        "train_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8]);                                                                  train_labels.append(1); train_ids.append(\"5456_moveeop#0\")\n",
        "\n",
        "# Validation data\n",
        "validation_data_raw = []; validation_labels = []; validation_ids = []\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8]);                                                                  validation_labels.append(1); validation_ids.append(\"5456_moveeop#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 14]);                                                              validation_labels.append(1); validation_ids.append(\"bimcompexptbuildingcompmtd_002#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 15]);                                                        validation_labels.append(1); validation_ids.append(\"tp_change_diameter#0\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 15]);                                                        validation_labels.append(1); validation_ids.append(\"tp_flip_orientation#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 14]);                                                        validation_labels.append(1); validation_ids.append(\"tp_flip_orientation#2\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8]);                                                         validation_labels.append(1); validation_ids.append(\"rt_core cavity design_1170823#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8]);                                                               validation_labels.append(1); validation_ids.append(\"rt_core cavity design_1170823#2\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 14]);                                                           validation_labels.append(1); validation_ids.append(\"break link_catia_part1#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 14]);                                                              validation_labels.append(1); validation_ids.append(\"break link_catia_part1#1\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 15]);                                                              validation_labels.append(1); validation_ids.append(\"break link_catia_part1#3\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 17]);                                                              validation_labels.append(1); validation_ids.append(\"break link_catia_part1#5\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8]);                                                                  validation_labels.append(1); validation_ids.append(\"break link_catia_part1#6\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8]);                                                                     validation_labels.append(1); validation_ids.append(\"break link_proe_part1#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8]);                                                            validation_labels.append(1); validation_ids.append(\"copy_paste_part#1\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 14]);                                                     validation_labels.append(1); validation_ids.append(\"copy_paste_part#2\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 14]);                                                  validation_labels.append(1); validation_ids.append(\"copy_paste_part#4\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8]);                                                      validation_labels.append(1); validation_ids.append(\"copy_paste_part#5\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 14]);                                               validation_labels.append(1); validation_ids.append(\"copy_paste_part#6\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8]);                                                   validation_labels.append(1); validation_ids.append(\"copy_paste_part#7\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 14]);                                                           validation_labels.append(1); validation_ids.append(\"bug106347#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 14]);                                                  validation_labels.append(1); validation_ids.append(\"bug109424#6\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 15]);                                                           validation_labels.append(1); validation_ids.append(\"bug117219#0\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8, 8]);                                                               validation_labels.append(1); validation_ids.append(\"bug117664_1#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 14]);                                                        validation_labels.append(1); validation_ids.append(\"bug118455#1\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 14]);                                               validation_labels.append(1); validation_ids.append(\"dwg_crosspartproject#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 14]);                                                     validation_labels.append(1); validation_ids.append(\"dwg_crosspartproject#34\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 15]);                                                              validation_labels.append(1); validation_ids.append(\"futureanycad_update_catia_assembly6#17\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8]);                                                                     validation_labels.append(1); validation_ids.append(\"futureanycad_update_swx_part1#0\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 15]);                                                           validation_labels.append(1); validation_ids.append(\"futurecad_derive_part-dv-update#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8]);                                                validation_labels.append(1); validation_ids.append(\"localsave_catia_2#112\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                            validation_labels.append(1); validation_ids.append(\"localsave_nx_2#59\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                            validation_labels.append(1); validation_ids.append(\"localsave_nx_2#63\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                         validation_labels.append(1); validation_ids.append(\"localsave_nx_2#123\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                         validation_labels.append(1); validation_ids.append(\"localsave_nx_2#127\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8]);                                                            validation_labels.append(1); validation_ids.append(\"localsave_nx_2#362\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#1\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                             validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#2\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8]);                                       validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#3\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                   validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#4\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                          validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#5\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                       validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#6\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8]);                                 validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#7\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 14]);                                      validation_labels.append(1); validation_ids.append(\"selective_catia_assembly8#10\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 16]);                                                                 validation_labels.append(1); validation_ids.append(\"createtextboxconstraintproxy_001#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 13]);                                                                 validation_labels.append(1); validation_ids.append(\"createtextboxconstraintproxy_001#1\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 13]);                                                                 validation_labels.append(1); validation_ids.append(\"createtextboxconstraintproxy_001#2\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 17]);                                                           validation_labels.append(1); validation_ids.append(\"skipallunresolvedwarnings_001#2\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 17]);                                                     validation_labels.append(1); validation_ids.append(\"skipallunresolvedwarnings_001#4\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 11, 12, 8, 13]);                                                                 validation_labels.append(1); validation_ids.append(\"translatoraddinsavecopyas_001#2\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 13]);                                                                 validation_labels.append(1); validation_ids.append(\"testsmeasuretoolsgetlooplength_001#1\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 17]);                                                           validation_labels.append(1); validation_ids.append(\"additionaldisplaysettings_ipn#4\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 17]);                                                                 validation_labels.append(1); validation_ids.append(\"part_ps_delete#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 15]);                                                  validation_labels.append(1); validation_ids.append(\"asmb_conn_1483947#10\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 15]);                                                     validation_labels.append(1); validation_ids.append(\"mirrorcopyoffsetoriginwithrefer#7\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 14]);                                                                 validation_labels.append(1); validation_ids.append(\"tfs_147642#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 14]);                                                                 validation_labels.append(1); validation_ids.append(\"tfs_151789#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 15]);                                                     validation_labels.append(1); validation_ids.append(\"cm_smoke_ipartcolor#5\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 15]);                                               validation_labels.append(1); validation_ids.append(\"cm_smoke_molddesign#1\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 15]);                                            validation_labels.append(1); validation_ids.append(\"cm_smoke_molddesign#21\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 18, 19, 19, 19, 19, 19, 19, 21, 71, 21]); validation_labels.append(1); validation_ids.append(\"bug_168114#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 17]);                                                              validation_labels.append(1); validation_ids.append(\"translator_jtexportoption_006#4\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8]);                                                         validation_labels.append(1); validation_ids.append(\"invalidpath#8\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 15]);                                                  validation_labels.append(1); validation_ids.append(\"patn_multisolids_02#1\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 17]);                                                  validation_labels.append(1); validation_ids.append(\"ipt_psur_c1132#3\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 17]);                                                  validation_labels.append(1); validation_ids.append(\"ipt_psur_c1132#6\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 17]);                                                        validation_labels.append(1); validation_ids.append(\"ipt_psur_c1132#386\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 17]);                                                        validation_labels.append(1); validation_ids.append(\"ipt_psur_c1132#389\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 15]);                                         validation_labels.append(1); validation_ids.append(\"am_migr_r11_topassembly#1\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 17]);                                               validation_labels.append(1); validation_ids.append(\"am_migr_r11_topassembly#6\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 17]);                                               validation_labels.append(1); validation_ids.append(\"am_migr_r11_topassembly#12\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8]);                                                   validation_labels.append(1); validation_ids.append(\"am_migr_r11_topassembly#15\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 17]);                                            validation_labels.append(1); validation_ids.append(\"am_migr_r11_10000142208_iam_000#357\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 15]);                                               validation_labels.append(1); validation_ids.append(\"am_migr_r12_201-01#15\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 8, 15]);                                         validation_labels.append(1); validation_ids.append(\"am_migr_r12_201-01#169\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8, 8, 8, 15]);                                            validation_labels.append(1); validation_ids.append(\"am_migr_r12_201-01#171\")\n",
        "validation_data_raw.append([4, 2, 7, 10, 9, 11, 12, 8, 8, 8, 8, 8, 8]);                                                      validation_labels.append(1); validation_ids.append(\"am_migr_r6_7000ma016#71\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 48, 49, 50, 51, 52, 53, 54]);                                                    validation_labels.append(0); validation_ids.append(\"cylinder-30x1000x1500n#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 35, 36, 30, 30, 30, 30, 30, 30, 30]);                                               validation_labels.append(0); validation_ids.append(\"md_mbs_c1213#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 35, 36, 30, 30, 30, 30, 30]);                                                       validation_labels.append(0); validation_ids.append(\"md_mbs_c1213#3\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 34, 33, 13]);                                                                    validation_labels.append(0); validation_ids.append(\"bug116303#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 72, 73]);                                                                           validation_labels.append(0); validation_ids.append(\"colorschemescoll_001#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 55, 56, 57, 58, 59]);                                                            validation_labels.append(0); validation_ids.append(\"curvedframe_trimtoframe_arc#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 42, 43, 44, 45, 60, 61, 62, 13]);                                                validation_labels.append(0); validation_ids.append(\"curvedframe_trimtoframe_spline#0\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 74, 75, 76, 77, 33, 13]);                                                        validation_labels.append(0); validation_ids.append(\"feat_807811#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 63, 64, 78]);                                                                    validation_labels.append(0); validation_ids.append(\"parker union tee 164ca-264ca 164ca-2#0\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 63, 64]);                                                                        validation_labels.append(0); validation_ids.append(\"parker union tee 164ca-264ca 164ca-2#1\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 26]);                         validation_labels.append(0); validation_ids.append(\"tc5356#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 19, 21, 26]);                     validation_labels.append(0); validation_ids.append(\"tc5356#5\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 42, 43, 44, 45, 79, 80, 81, 82, 83]);                                            validation_labels.append(0); validation_ids.append(\"dac_chains_calc_10201020#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 65]);                         validation_labels.append(0); validation_ids.append(\"dm_at_appendfilenameinviewblockname_0004#0\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 66]);                         validation_labels.append(0); validation_ids.append(\"dm_at_appendfilenameinviewblockname_0004#1\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 19, 18, 37, 38, 39, 21, 67]);                         validation_labels.append(0); validation_ids.append(\"dm_at_appendfilenameinviewblockname_0004#2\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 18, 27, 27, 28, 29, 21, 21, 13]);                 validation_labels.append(0); validation_ids.append(\"idw_etsy_d1207#0\")\n",
        "validation_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 31]);                         validation_labels.append(0); validation_ids.append(\"592453#0\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 26]);                         validation_labels.append(0); validation_ids.append(\"907031#1\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 21, 31]);                         validation_labels.append(0); validation_ids.append(\"907031#10\")\n",
        "validation_data_raw.append([4, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 18, 19, 19, 19, 19, 19]);                 validation_labels.append(0); validation_ids.append(\"idw_replacemodelreference_c1102#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 32, 40, 41, 68]);                                                                validation_labels.append(0); validation_ids.append(\"windshield wiper#0\")\n",
        "validation_data_raw.append([3, 2, 7, 10, 9, 32, 40, 41, 69]);                                                                validation_labels.append(0); validation_ids.append(\"windshield wiper#1\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 32, 40, 41, 70]);                                                                validation_labels.append(0); validation_ids.append(\"windshield wiper#2\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 32, 46, 47]);                                                                    validation_labels.append(0); validation_ids.append(\"windshield wiper#6\")\n",
        "validation_data_raw.append([3, 1, 7, 10, 9, 84, 85, 86, 87]);                                                                validation_labels.append(0); validation_ids.append(\"smmsb_createpunch#0\")\n",
        "validation_data_raw.append([5, 2, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 18, 19, 19, 19, 19, 19, 19, 21, 88]);     validation_labels.append(0); validation_ids.append(\"dm_migr_r11_coolingtower#0\")\n",
        "\n",
        "# Prediction data\n",
        "prediction_data_raw = []; prediction_ids = []\n",
        "prediction_data_raw.append([3, 1, 7, 10, 9, 11, 12, 8, 8, 8, 14]);                                                           prediction_ids.append(\"idw_atxt_c1110#0\")\n",
        "prediction_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 19, 21, 31]);                     prediction_ids.append(\"idw_atxt_c1110#1\")\n",
        "prediction_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 19, 21, 31]);                     prediction_ids.append(\"idw_atxt_c1110#2\")\n",
        "prediction_data_raw.append([5, 1, 7, 10, 9, 22, 23, 24, 9, 25, 20, 18, 18, 20, 18, 19, 19, 19, 21, 31]);                     prediction_ids.append(\"idw_atxt_c1110#3\")\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fI6IrfgnDYws",
        "colab_type": "code",
        "outputId": "a1b6dc49-9f9a-4c6a-e153-1ba9d8704b8a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "#print(word_index)\n",
        "print(\"word_index length: \", len(word_index))\n"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "word_index length:  89\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "pL4FYiOSp_Yx",
        "colab_type": "code",
        "outputId": "5e3851f6-c3f8-4409-8947-33629828ecd0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "cell_type": "code",
      "source": [
        "# Pad train data\n",
        "train_data = keras.preprocessing.sequence.pad_sequences(train_data_raw, value=word_index[\"<PAD>\"], padding='post', maxlen=INPUT_WIDTH)\n",
        "print(train_data.shape)\n",
        "\n",
        "# Pad validation data\n",
        "validation_data = keras.preprocessing.sequence.pad_sequences(validation_data_raw, value=word_index[\"<PAD>\"], padding='post', maxlen=INPUT_WIDTH)\n",
        "print(validation_data.shape)\n",
        "\n",
        "# Pad prediction data\n",
        "prediction_data = keras.preprocessing.sequence.pad_sequences(prediction_data_raw, value=word_index[\"<PAD>\"], padding='post', maxlen=INPUT_WIDTH)\n",
        "print(prediction_data.shape)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(37, 25)\n",
            "(103, 25)\n",
            "(4, 25)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "5ts5dy3kqgVX",
        "colab_type": "code",
        "outputId": "c585baba-14fc-4f01-b915-c3e831db9334",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3723
        }
      },
      "cell_type": "code",
      "source": [
        "model = tf.keras.models.Sequential([\n",
        "    tf.keras.layers.Embedding(len(word_index), 16, input_length=INPUT_WIDTH),\n",
        "    tf.keras.layers.GlobalMaxPooling1D(),\n",
        "    tf.keras.layers.Dense(16, activation=tf.nn.relu),\n",
        "    tf.keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer='adam',\n",
        "              loss='binary_crossentropy',\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "# Train\n",
        "history = model.fit(train_data, train_labels, epochs=100, batch_size=4, validation_data=(validation_data, validation_labels), verbose=1)\n",
        "\n",
        "# See final loss and accuracy\n",
        "results = model.evaluate(validation_data, validation_labels)\n",
        "print(results)\n"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_1 (Embedding)      (None, 25, 16)            1424      \n",
            "_________________________________________________________________\n",
            "global_max_pooling1d_1 (Glob (None, 16)                0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 16)                272       \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 1)                 17        \n",
            "=================================================================\n",
            "Total params: 1,713\n",
            "Trainable params: 1,713\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train on 37 samples, validate on 103 samples\n",
            "Epoch 1/100\n",
            "37/37 [==============================] - 0s 8ms/sample - loss: 0.6955 - acc: 0.4054 - val_loss: 0.6942 - val_acc: 0.2621\n",
            "Epoch 2/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6821 - acc: 0.7297 - val_loss: 0.7033 - val_acc: 0.2621\n",
            "Epoch 3/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6711 - acc: 0.7297 - val_loss: 0.7073 - val_acc: 0.2621\n",
            "Epoch 4/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6636 - acc: 0.7297 - val_loss: 0.7115 - val_acc: 0.2621\n",
            "Epoch 5/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6549 - acc: 0.7297 - val_loss: 0.7135 - val_acc: 0.2621\n",
            "Epoch 6/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6466 - acc: 0.7297 - val_loss: 0.7131 - val_acc: 0.2621\n",
            "Epoch 7/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6390 - acc: 0.7297 - val_loss: 0.7124 - val_acc: 0.2621\n",
            "Epoch 8/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6304 - acc: 0.7297 - val_loss: 0.7103 - val_acc: 0.2621\n",
            "Epoch 9/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6208 - acc: 0.7297 - val_loss: 0.7080 - val_acc: 0.2621\n",
            "Epoch 10/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.6101 - acc: 0.7297 - val_loss: 0.7084 - val_acc: 0.2621\n",
            "Epoch 11/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5978 - acc: 0.7297 - val_loss: 0.7078 - val_acc: 0.2621\n",
            "Epoch 12/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5848 - acc: 0.7297 - val_loss: 0.7042 - val_acc: 0.2621\n",
            "Epoch 13/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5708 - acc: 0.7297 - val_loss: 0.7010 - val_acc: 0.2621\n",
            "Epoch 14/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5556 - acc: 0.7297 - val_loss: 0.6882 - val_acc: 0.2621\n",
            "Epoch 15/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5392 - acc: 0.7297 - val_loss: 0.6796 - val_acc: 0.2621\n",
            "Epoch 16/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5219 - acc: 0.7297 - val_loss: 0.6705 - val_acc: 0.2621\n",
            "Epoch 17/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.5018 - acc: 0.7297 - val_loss: 0.6521 - val_acc: 0.2621\n",
            "Epoch 18/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.4805 - acc: 0.7297 - val_loss: 0.6361 - val_acc: 0.2621\n",
            "Epoch 19/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.4568 - acc: 0.7568 - val_loss: 0.6126 - val_acc: 0.6990\n",
            "Epoch 20/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.4323 - acc: 0.9459 - val_loss: 0.5867 - val_acc: 0.8641\n",
            "Epoch 21/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.4066 - acc: 1.0000 - val_loss: 0.5598 - val_acc: 0.9903\n",
            "Epoch 22/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.3805 - acc: 1.0000 - val_loss: 0.5318 - val_acc: 0.9903\n",
            "Epoch 23/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.3541 - acc: 1.0000 - val_loss: 0.4981 - val_acc: 0.9903\n",
            "Epoch 24/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.3286 - acc: 1.0000 - val_loss: 0.4664 - val_acc: 0.9903\n",
            "Epoch 25/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.3037 - acc: 1.0000 - val_loss: 0.4402 - val_acc: 0.9903\n",
            "Epoch 26/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.2799 - acc: 1.0000 - val_loss: 0.4175 - val_acc: 0.9903\n",
            "Epoch 27/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.2563 - acc: 1.0000 - val_loss: 0.3926 - val_acc: 0.9903\n",
            "Epoch 28/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.2341 - acc: 1.0000 - val_loss: 0.3610 - val_acc: 0.9903\n",
            "Epoch 29/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.2120 - acc: 1.0000 - val_loss: 0.3395 - val_acc: 0.9903\n",
            "Epoch 30/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1920 - acc: 1.0000 - val_loss: 0.3144 - val_acc: 0.9903\n",
            "Epoch 31/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1728 - acc: 1.0000 - val_loss: 0.2889 - val_acc: 0.9903\n",
            "Epoch 32/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1561 - acc: 1.0000 - val_loss: 0.2603 - val_acc: 0.9903\n",
            "Epoch 33/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1400 - acc: 1.0000 - val_loss: 0.2389 - val_acc: 0.9903\n",
            "Epoch 34/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1258 - acc: 1.0000 - val_loss: 0.2223 - val_acc: 0.9903\n",
            "Epoch 35/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1135 - acc: 1.0000 - val_loss: 0.2068 - val_acc: 0.9903\n",
            "Epoch 36/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.1024 - acc: 1.0000 - val_loss: 0.1916 - val_acc: 0.9903\n",
            "Epoch 37/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0926 - acc: 1.0000 - val_loss: 0.1778 - val_acc: 0.9903\n",
            "Epoch 38/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0838 - acc: 1.0000 - val_loss: 0.1651 - val_acc: 0.9903\n",
            "Epoch 39/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0759 - acc: 1.0000 - val_loss: 0.1543 - val_acc: 0.9903\n",
            "Epoch 40/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0690 - acc: 1.0000 - val_loss: 0.1448 - val_acc: 0.9903\n",
            "Epoch 41/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0625 - acc: 1.0000 - val_loss: 0.1373 - val_acc: 0.9903\n",
            "Epoch 42/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0570 - acc: 1.0000 - val_loss: 0.1279 - val_acc: 0.9903\n",
            "Epoch 43/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0519 - acc: 1.0000 - val_loss: 0.1191 - val_acc: 0.9903\n",
            "Epoch 44/100\n",
            "37/37 [==============================] - 0s 971us/sample - loss: 0.0472 - acc: 1.0000 - val_loss: 0.1130 - val_acc: 0.9903\n",
            "Epoch 45/100\n",
            "37/37 [==============================] - 0s 958us/sample - loss: 0.0434 - acc: 1.0000 - val_loss: 0.1079 - val_acc: 0.9903\n",
            "Epoch 46/100\n",
            "37/37 [==============================] - 0s 924us/sample - loss: 0.0401 - acc: 1.0000 - val_loss: 0.1033 - val_acc: 0.9903\n",
            "Epoch 47/100\n",
            "37/37 [==============================] - 0s 912us/sample - loss: 0.0371 - acc: 1.0000 - val_loss: 0.0989 - val_acc: 0.9903\n",
            "Epoch 48/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0342 - acc: 1.0000 - val_loss: 0.0960 - val_acc: 0.9903\n",
            "Epoch 49/100\n",
            "37/37 [==============================] - 0s 983us/sample - loss: 0.0318 - acc: 1.0000 - val_loss: 0.0930 - val_acc: 0.9903\n",
            "Epoch 50/100\n",
            "37/37 [==============================] - 0s 942us/sample - loss: 0.0295 - acc: 1.0000 - val_loss: 0.0899 - val_acc: 0.9903\n",
            "Epoch 51/100\n",
            "37/37 [==============================] - 0s 935us/sample - loss: 0.0275 - acc: 1.0000 - val_loss: 0.0874 - val_acc: 0.9903\n",
            "Epoch 52/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0257 - acc: 1.0000 - val_loss: 0.0854 - val_acc: 0.9903\n",
            "Epoch 53/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0241 - acc: 1.0000 - val_loss: 0.0833 - val_acc: 0.9903\n",
            "Epoch 54/100\n",
            "37/37 [==============================] - 0s 995us/sample - loss: 0.0226 - acc: 1.0000 - val_loss: 0.0816 - val_acc: 0.9903\n",
            "Epoch 55/100\n",
            "37/37 [==============================] - 0s 944us/sample - loss: 0.0212 - acc: 1.0000 - val_loss: 0.0795 - val_acc: 0.9903\n",
            "Epoch 56/100\n",
            "37/37 [==============================] - 0s 916us/sample - loss: 0.0200 - acc: 1.0000 - val_loss: 0.0780 - val_acc: 0.9903\n",
            "Epoch 57/100\n",
            "37/37 [==============================] - 0s 927us/sample - loss: 0.0189 - acc: 1.0000 - val_loss: 0.0766 - val_acc: 0.9903\n",
            "Epoch 58/100\n",
            "37/37 [==============================] - 0s 957us/sample - loss: 0.0178 - acc: 1.0000 - val_loss: 0.0755 - val_acc: 0.9903\n",
            "Epoch 59/100\n",
            "37/37 [==============================] - 0s 957us/sample - loss: 0.0169 - acc: 1.0000 - val_loss: 0.0743 - val_acc: 0.9903\n",
            "Epoch 60/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0160 - acc: 1.0000 - val_loss: 0.0732 - val_acc: 0.9903\n",
            "Epoch 61/100\n",
            "37/37 [==============================] - 0s 980us/sample - loss: 0.0151 - acc: 1.0000 - val_loss: 0.0724 - val_acc: 0.9903\n",
            "Epoch 62/100\n",
            "37/37 [==============================] - 0s 942us/sample - loss: 0.0144 - acc: 1.0000 - val_loss: 0.0712 - val_acc: 0.9903\n",
            "Epoch 63/100\n",
            "37/37 [==============================] - 0s 960us/sample - loss: 0.0136 - acc: 1.0000 - val_loss: 0.0702 - val_acc: 0.9903\n",
            "Epoch 64/100\n",
            "37/37 [==============================] - 0s 938us/sample - loss: 0.0129 - acc: 1.0000 - val_loss: 0.0693 - val_acc: 0.9903\n",
            "Epoch 65/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0123 - acc: 1.0000 - val_loss: 0.0685 - val_acc: 0.9903\n",
            "Epoch 66/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0117 - acc: 1.0000 - val_loss: 0.0679 - val_acc: 0.9903\n",
            "Epoch 67/100\n",
            "37/37 [==============================] - 0s 991us/sample - loss: 0.0112 - acc: 1.0000 - val_loss: 0.0674 - val_acc: 0.9903\n",
            "Epoch 68/100\n",
            "37/37 [==============================] - 0s 999us/sample - loss: 0.0107 - acc: 1.0000 - val_loss: 0.0668 - val_acc: 0.9903\n",
            "Epoch 69/100\n",
            "37/37 [==============================] - 0s 979us/sample - loss: 0.0102 - acc: 1.0000 - val_loss: 0.0666 - val_acc: 0.9903\n",
            "Epoch 70/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0098 - acc: 1.0000 - val_loss: 0.0660 - val_acc: 0.9903\n",
            "Epoch 71/100\n",
            "37/37 [==============================] - 0s 924us/sample - loss: 0.0094 - acc: 1.0000 - val_loss: 0.0658 - val_acc: 0.9903\n",
            "Epoch 72/100\n",
            "37/37 [==============================] - 0s 982us/sample - loss: 0.0090 - acc: 1.0000 - val_loss: 0.0656 - val_acc: 0.9903\n",
            "Epoch 73/100\n",
            "37/37 [==============================] - 0s 908us/sample - loss: 0.0086 - acc: 1.0000 - val_loss: 0.0654 - val_acc: 0.9903\n",
            "Epoch 74/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0083 - acc: 1.0000 - val_loss: 0.0652 - val_acc: 0.9903\n",
            "Epoch 75/100\n",
            "37/37 [==============================] - 0s 948us/sample - loss: 0.0080 - acc: 1.0000 - val_loss: 0.0648 - val_acc: 0.9903\n",
            "Epoch 76/100\n",
            "37/37 [==============================] - 0s 948us/sample - loss: 0.0077 - acc: 1.0000 - val_loss: 0.0646 - val_acc: 0.9903\n",
            "Epoch 77/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0074 - acc: 1.0000 - val_loss: 0.0645 - val_acc: 0.9903\n",
            "Epoch 78/100\n",
            "37/37 [==============================] - 0s 940us/sample - loss: 0.0071 - acc: 1.0000 - val_loss: 0.0643 - val_acc: 0.9903\n",
            "Epoch 79/100\n",
            "37/37 [==============================] - 0s 913us/sample - loss: 0.0069 - acc: 1.0000 - val_loss: 0.0643 - val_acc: 0.9903\n",
            "Epoch 80/100\n",
            "37/37 [==============================] - 0s 951us/sample - loss: 0.0066 - acc: 1.0000 - val_loss: 0.0642 - val_acc: 0.9903\n",
            "Epoch 81/100\n",
            "37/37 [==============================] - 0s 918us/sample - loss: 0.0064 - acc: 1.0000 - val_loss: 0.0641 - val_acc: 0.9903\n",
            "Epoch 82/100\n",
            "37/37 [==============================] - 0s 931us/sample - loss: 0.0062 - acc: 1.0000 - val_loss: 0.0640 - val_acc: 0.9903\n",
            "Epoch 83/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0060 - acc: 1.0000 - val_loss: 0.0640 - val_acc: 0.9903\n",
            "Epoch 84/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0058 - acc: 1.0000 - val_loss: 0.0639 - val_acc: 0.9903\n",
            "Epoch 85/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0056 - acc: 1.0000 - val_loss: 0.0639 - val_acc: 0.9903\n",
            "Epoch 86/100\n",
            "37/37 [==============================] - 0s 987us/sample - loss: 0.0054 - acc: 1.0000 - val_loss: 0.0637 - val_acc: 0.9903\n",
            "Epoch 87/100\n",
            "37/37 [==============================] - 0s 970us/sample - loss: 0.0052 - acc: 1.0000 - val_loss: 0.0637 - val_acc: 0.9903\n",
            "Epoch 88/100\n",
            "37/37 [==============================] - 0s 957us/sample - loss: 0.0051 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 89/100\n",
            "37/37 [==============================] - 0s 957us/sample - loss: 0.0049 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 90/100\n",
            "37/37 [==============================] - 0s 948us/sample - loss: 0.0048 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 91/100\n",
            "37/37 [==============================] - 0s 955us/sample - loss: 0.0046 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 92/100\n",
            "37/37 [==============================] - 0s 936us/sample - loss: 0.0045 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 93/100\n",
            "37/37 [==============================] - 0s 920us/sample - loss: 0.0044 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 94/100\n",
            "37/37 [==============================] - 0s 981us/sample - loss: 0.0042 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 95/100\n",
            "37/37 [==============================] - 0s 992us/sample - loss: 0.0041 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 96/100\n",
            "37/37 [==============================] - 0s 987us/sample - loss: 0.0040 - acc: 1.0000 - val_loss: 0.0636 - val_acc: 0.9903\n",
            "Epoch 97/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0039 - acc: 1.0000 - val_loss: 0.0637 - val_acc: 0.9903\n",
            "Epoch 98/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0038 - acc: 1.0000 - val_loss: 0.0637 - val_acc: 0.9903\n",
            "Epoch 99/100\n",
            "37/37 [==============================] - 0s 964us/sample - loss: 0.0037 - acc: 1.0000 - val_loss: 0.0637 - val_acc: 0.9903\n",
            "Epoch 100/100\n",
            "37/37 [==============================] - 0s 1ms/sample - loss: 0.0036 - acc: 1.0000 - val_loss: 0.0638 - val_acc: 0.9903\n",
            "103/103 [==============================] - 0s 95us/sample - loss: 0.0638 - acc: 0.9903\n",
            "[0.06378819418899281, 0.99029124]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "cPIfRF8INgc_",
        "colab_type": "code",
        "outputId": "1c56dfc0-3555-4ca7-8f33-e536e8972cac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        }
      },
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "history_dict = history.history\n",
        "history_dict.keys()\n",
        "acc = history_dict['acc']\n",
        "val_acc = history_dict['val_acc']\n",
        "loss = history_dict['loss']\n",
        "val_loss = history_dict['val_loss']\n",
        "\n",
        "epochs = range(1, len(acc) + 1)\n",
        "\n",
        "# \"bo\" is for \"blue dot\"\n",
        "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
        "# b is for \"solid blue line\"\n",
        "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAFnCAYAAACPasF4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xt8zvX/x/HHddjBbBg2RiqUwyb6\nqm9fTM1hGFKpMIWk0lf6Ir7fb/jROqAS5dDJN+mgg5Um+eYQab4dlIovQlF9c8phcxg72HZd1+f3\nx9XGbGPs+ly7ru15v912s8/nuj6fz9vL5Xp93sePxTAMAxEREfEb1oougIiIiFwYJW8RERE/o+Qt\nIiLiZ5S8RURE/IySt4iIiJ9R8hYREfEzSt5SpSUlJZGQkEBCQgIxMTF07ty5cDszM/OCzpWQkEB6\nevo53zNz5kzefffd8hTZ44YOHUpKSopHztW8eXMOHjzI6tWrmTBhQrmu99577xX+XpbYltX48eN5\n8cUXPXIukYpir+gCiFSkxx57rPD3Ll26MH36dK699tqLOtfKlSvP+55x48Zd1Ln9Tbdu3ejWrdtF\nH5+Wlsb8+fPp378/ULbYilQlqnmLnMPgwYN57rnn6NmzJxs3biQ9PZ177rmHhIQEunTpwmuvvVb4\n3oJa5zfffMOAAQOYOXMmPXv2pEuXLmzYsAEoWuvr0qULixYt4vbbb6djx4489dRThed6+eWXad++\nPbfddhtvv/02Xbp0KbF877//Pj179qR79+7ceeed7N+/H4CUlBRGjRrFxIkT6dGjB7169WLXrl0A\n7N27l379+hEfH8+4ceNwOp3Fzrtu3Tr69OlTZN/NN9/Mf/7zn3PGoEBKSgpDhw497/U+/fRT+vTp\nQ48ePbj11lvZsWMHAImJifz+++8kJCSQl5dXGFuAN998k169epGQkMCIESM4evRoYWznzJnD3Xff\nTefOnbn77rvJyckp7Z8WgB9//JHExEQSEhK4+eab+fzzzwHIyspi5MiR9OzZk65duzJp0iTy8/NL\n3S/ibUreIufxww8/8PHHH9O2bVteeuklLrnkElauXMkbb7zBzJkzOXDgQLFjtm/fTps2bVixYgV3\n3HEHL730Uonn/vbbb0lOTuaDDz7grbfe4uDBg+zatYv58+ezdOlS3nnnnVJrnUeOHOHxxx/ntdde\n45NPPuHSSy8t0hz8n//8hzvuuINVq1bxl7/8hTfeeAOAGTNm0L59e9asWcNdd93Fxo0bi527ffv2\nHDx4kL179wLuBHzw4EE6dOhQ5hgUKO16DoeD8ePH88QTT7Bq1Sq6dOnC008/DcC0adOIiopi5cqV\nBAYGFp7rv//9L6+++ioLFy5k5cqVNGjQgJkzZxa+vnLlSp577jlWr17N0aNHWb16danlcrlcjB07\nlkGDBrFy5UqmTJnCuHHjyMzM5MMPP6RGjRqsWLGCVatWYbPZ+Pnnn0vdL+JtSt4i5xEXF4fV6v6v\nMmnSJCZPngxAo0aNiIiIYN++fcWOqV69OvHx8QDExMTw+++/l3juPn36YLPZqFevHnXq1OHAgQN8\n++23XHfddURGRhIUFMRtt91W4rF16tTh+++/p379+gBce+21hckWoGnTprRq1QqA6OjowgT73Xff\n0atXLwBat25NkyZNip07MDCQzp07s3btWgDWrFlDfHw8dru9zDEoUNr17HY7X331FVdffXWJ5S9J\namoqPXr0oE6dOgD069ePL7/8svD1uLg4atWqhd1up1mzZue8qdi3bx/p6en07t0bgKuuuooGDRqw\ndetWateuzaZNm/jiiy9wuVw89thjtGzZstT9It6mPm+R86hZs2bh71u3bi2saVqtVtLS0nC5XMWO\nCQsLK/zdarWW+B6A0NDQwt9tNhtOp5MTJ04UuWa9evVKPNbpdDJnzhzWrl2L0+kkKyuLxo0bl1iG\ngnMDZGRkFLlujRo1Sjx/jx49ePPNN7nrrrtYs2YNDzzwwAXFoMC5rrdw4UKWLFlCXl4eeXl5WCyW\nUs8DcPToUSIjI4uc68iRI+f9O5d2rrCwsCLXrFGjBkePHqV3795kZGQwe/Zsfv31V2666SYmTJhA\nz549S9x/ZuuAiDeo5i1yAf7xj3/Qo0cPVq1axcqVKwkPD/f4NUJDQ8nOzi7cPnz4cInvW758OWvX\nruWtt95i1apVjBo1qkznr1GjRpGR9AV9xme7/vrr+fHHH/ntt9/47bffaNeuHXDhMSjtehs3buSV\nV17hpZdeYtWqVUyZMuW8Za9bty7Hjx8v3D5+/Dh169Y973ElqVOnDhkZGZz5bKbjx48X1uoTExN5\n//33Wb58Odu2bePDDz88534Rb1LyFrkAR44coVWrVlgsFpYsWUJOTk6RROsJrVu35ptvvuHo0aPk\n5eWVmhyOHDlCw4YNqV27NseOHWPFihVkZWWd9/xXX311YV/wxo0b2bNnT4nvCwwMpGPHjjzzzDN0\n7doVm81WeN0LiUFp1zt69Ch16tShQYMG5OTksGTJErKzszEMA7vdTnZ2Ng6Ho8i5OnXqxOrVqzl2\n7BgAixYtIi4u7rx/55Jccskl1K9fn+XLlxeWLT09ndatW/PCCy+wePFiwN3ycckll2CxWErdL+Jt\nSt4iF2D06NGMHDmSPn36kJ2dzYABA5g8eXKpCfBitG7dmr59+9K3b1+GDBlC586dS3zfjTfeyPHj\nx+nWrRvjxo1jzJgxHDx4sMio9ZL84x//4LPPPiM+Pp63336bDh06lPreHj16sGbNGnr27Fm470Jj\nUNr1rr/+eiIjI4mPj2fYsGHcddddhIWFMWrUKJo3b07NmjWJjY0tMl6gdevWDB8+nDvvvJOEhARO\nnjzJQw89dM6/b2ksFgvPPvssb731Fj179mTKlCnMnj2bkJAQbr75ZpYuXUqPHj1ISEggICCAm2++\nudT9It5m0fO8RXyPYRiFNbrU1FRmzZql5lkRKaSat4iPOXr0KO3atWP//v0YhsGKFSsKR2SLiIBq\n3iI+6d1332XBggVYLBaaNGnC1KlTCwdSiYgoeYuIiPgZNZuLiIj4GSVvERERP+M3K6ylpZ0s1/Hh\n4SEcO+bZ+bhVkeLoGYqjZyiOnqE4eoYZcYyICCtxf5WpedvttoouQqWgOHqG4ugZiqNnKI6e4c04\nVpnkLSIiUlkoeYuIiPgZJW8RERE/o+QtIiLiZ5S8RURE/IySt4iIiJ9R8hYREfEzfrNIi4iIVB5z\n5z7HTz/t4OjRI5w6dYoGDRpSo0ZNpk175rzHLl++jOrVQ4mLK/lZ97Nnz6Rfv0QaNGh4UWV78MHh\njB37T5o0ueKijvcGJW8RETmvJUvszJoVyM6dVpo1czFmTB59+zou+nx/+9tDgDsR//rrLzz44Jgy\nH9urV59zvj569LiLLpe/UPIWEZFzWrLEzv33Vyvc3rHD9sd2TrkSeEk2bvyORYveIjs7mwcffIhN\nm74nNfVTXC4X7dvHMmzYcF59dR61atWiceOmpKS8h8ViZffu/9GpU1eGDRteWHP+7LNPycrKZM+e\n3ezfv49Ro8bRvn0sb731OmvWfEKDBg1xOBwkJt5J27bXFitLZmYmU6c+SmbmSRwOB2PG/IPmzVsw\na9Yz/PjjDpxOJ3373k6vXn2YNesZfvllJ6dO5RXuM5OS9wXKzITt261s324jJwdCQqBaNYNq1aB6\ndYOaNQ1q1IAaNQxq1zYICKjoEouIlM+sWYEl7p89O9DjyRvgl19+5t13UwgMDGTTpu958cX5WK1W\n+ve/mQED7ijy3u3bt/HOOx/gcrno168Pw4YNL/L64cOHmDFjDl9//RVLl35ATEwrUlLe5913PyAr\nK4vExFtJTLyzxHK8//67xMS0YtCgofz443bmzn2WadOe4auvvuC995bicDhYvnwZJ05k8NVXX/DZ\nZ2s5cOAYy5cv83hMzqbkfR4uF3z+uY3k5AC+/97Gb79ZMAxLmY4NDDRo2dJF69ZO2rRxcfXVTlq2\ndCmhi4hf2bmz5LHNpe0vryuuuJLAQPcNQ3BwMA8+OBybzcbx48c5ceJEkfc2b96C4ODgUs/VuvXV\nAERGRpKZmcm+fXtp0qQpQUHBBAUF07JlTKnH/vjjdoYMuQeAFi2i2bdvLzVq1KRRo8sYP34snTvH\nk5DQm8DAQBo1uowRI0YQG9uJhITe5Q3BeSl5l+LQIQvJyQEsXBjA7t3uD2h4uEFsrJOYGBcxMU5q\n1oScHMjJsZCdDVlZFjIyLJw4ARkZFnbvtrJjh5XNm20sXOg+b3CwwVVXuWjb1knz5i7y8+HUKcjN\ntZCfD6GhRWvu11zj5ByfSxER0zVr5mLHjuIP3WjWzGXK9QL+qOEcPHiA5OS3WbDgbUJCQhg8uH+x\n99ps534YyJmvG4aBYYDVevqmw3KOupjFYsEwjMJtl8v99505cw4//fQjq1evZOXKj3nuuReYOXMO\nhw/v4b33Ugr3mcnU5D1t2jQ2b96MxWJh4sSJtG7dGoBDhw7x97//vfB9e/fuZdy4cfTpY24fwfnk\n5sInn9h5770A1qyx4XRaqFbNYODAfAYNyuPaa13n/IcuSV4e/PSTO4Fv2mRl40YbGzda+fbbsj19\nJjTUoHt3B336OOjSxUG1auc/RkTEk8aMySvS511g9Og8U697/PhxwsPDCQkJ4aeffuTgwYPk5+eX\n65xRUVH8+usvOBwOTp48yY8/7ij1vS1aRLNp03e0anUVP/ywlcaNm3LgwO988cV/6NcvkebNWzBs\n2KDCfQ88cB8PPngpw4YNKlcZy8K05L1hwwZ2795NcnIyv/zyCxMnTiQ5ORmAevXqsfCPqqjD4WDw\n4MF06dLFrKKcV0YGTJ8exOLFARw75s7OrVs7GTgwn3798qlR4+LPHRgIV13l4qqrXAz6498zKwu2\nbnU3wQcFQVCQu0Zut0NmprvmfvKkhb17rSxfbiclJYCUlACqVzfo3NlB9+4O4uOd1K1rnPviIiIe\n4O7XzmH27NOjzUePLt9o87K48spmVKsWwogRw7jqqqu5+eZbmTnzaVq3bnPR56xduw7duiVw331D\nuOyyxkRHx5Rae+/ffyDTpj3GqFF/xeVyMXbsw9StG8EPP2zm008/ISAggN69byrcl5iYCFjp3fum\niy5fWVmMM9sEPGj27Nk0aNCAfv36AZCQkMDixYsJDQ0t8r7333+frKwshg4des7zpaWdLFd5IiLC\nSEs7WWy6w5135vPGGwHs2mUjIsLF7bc7GDAgn+hoc5qDLpRhwJYtVpYts7NsWQD/+5+7ucdiMbjm\nGhePPnqK667zXlkL4ijlozh6huLoGVUtjsuXL6NbtwRsNhtDhiTy7LNziYysV+7zmhHHiIiwEveb\nVvNOT08nJub0QIDatWuTlpZWYvJesGCBWcUooqTpDpMmue+4/vrXPCZPzvW5wWQWC7Rp46JNmzz+\n7//y+PlnK6tW2fjkEzsbNtgYNCiElSuzaNJEtXARkbI4cuQIw4ffRUBAIN27J3gkcXub1waslVTB\n37RpE02aNCmW0EsSHh6C3V62fuLSPP98yR3GDRvCSy8FAiVPh/AlkZHQoQM89hi8/jrcfTfcdVco\nX38NtWp5pwyl3QnKhVEcPUNx9IyqFMexY//G2LF/M+Xc3oqjack7MjKS9PT0wu3Dhw8TERFR5D2p\nqam0b9++TOc7diy7XOWJiAhj+3YDKD7i7NAhg7S0zHKdvyL07g0PPBDEiy8GcsstDt59Nwe7ybdj\nVa15zSyKo2cojp6hOHqGN5vNTXswSWxsLKtWrQJg27ZtREZGFqthb926lRYtWphVhGJKm9Zg1nQH\nb5g8OZfu3R2sW2dn8uSgii6OiIh4gWnJu23btsTExJCYmMiUKVNISkoiJSWF1atXF74nLS2NOnXq\nmFWEYsaMKXlaQ4cOTuLiQoiKCiUuLoQlS/xn+rvNBi+/nEPLlk5efTWQOXMCMWcIooiI+ArTRpt7\nmidHm5853aFDByfz5xfv6543z/Nr9pppzx4LvXuHcOiQldtvz2fGjFOEhHj+Ompe8wzF0TMUR89Q\nHD2jUjSb+6q+fR2kpmbz+++ZpKZm8+WXJQ+Cmz3b9wevnenSSw0++SSba65xsnhxAH36hLBnzwWu\nKCMi4iX33393sQVSXn75ed59960S379x43dMmvRPAMaPH1vs9Q8+SObVV+eVer2ff97Fnj27AUhK\nmkBu7qmLLTq3396H7OzyjcMqryqXvM/m7TV7zRQVZfDhh9kMHpzH1q02uncP4T//Kd8IfRERM3Tr\n1oO1a1cX2Zeaupb4+O7nPfapp5694OutW7eWvXv3APDYY08SFOTf6077T+euSby9Zq/ZgoJg5sxc\n2rRxMWFCEImJ1Zg+PZdBg8q3pKCIiCd17dqdESPu4YEHRgHw4487iIiIICIikm+//Yb5818mICCA\nsLAwHn/8qSLH9u7dlY8//pTvvtvAnDkzqV27DnXq1C18xOfUqY+SlnaYnJwchg0bTv36USxdmsK6\ndWsJDw/nkUcm8OabyWRmnuTJJx8nPz8fq9XK+PGTsVgsTJ36KA0aNOTnn3fRrFlzxo+fXOLf4fDh\nQ0WOnz79Kez2UB5/fDJHjqSTl5fHPffcz7XXXldsX7t2HcoVvyqfvEtbszcjw0JUVKhHHjpfEYYM\nyadZMxdDhwYzdmwwv/5qYdKkPKz+16AgIib7xz8gObm6R8/Zp4+DRx/NLfX18PDaNGjQkO3bfyA6\nuhVr166mW7cEAE6ePElS0hQaNGjIE088wjffrCekhEE88+Y9z+TJT3Dllc34+99H0aBBQ06ePMF1\n17WjZ88b2b9/H5Mnj2fBgrf4y1/a06lTV6KjWxUeP3/+y9x448107dqdzz5bw4IF/+Kee+7np592\n8Nhj0wgPr03fvr04efIkYWHF+57PPv7555+nT5/bycg4zgsvvMLJkydZv/5Lfvnl52L7yqvKf5X3\n7etg3rwcoqOd2O0GDRu6a9z791txOi2FD533pxHoBdq1c7J8eTZNm7p4/vkg7r03mAruphERKdSt\nWwKffupuOv/yy//QqVNXAGrVqsXTT0/hwQeHs2nT95w4kVHi8QcOHODKK5sBcPXVbQEIC6vBjh3b\nGDFiGFOnPlrqsQA//bSDP/3pGgDatr2WXbt+AqBhw0bUqVMXq9VK3boRZGWVvA7I2cdv376dyy67\nnOzsLJ54YjIbN35LfHz3EveVl/9lJBP07esorFnHxYWwf3/x95j10HmzNWlisHx5FnffXY1//zuA\njAwLixfnXPDT0USk8nrmGfjnP7O8ft24uM68+eYCunXrQaNGl1Ljj6dAPfnkEzzzzCwuv7wxzz77\ndKnHn/loz4KJU6tXr+TEiRO88MJ8Tpw4wb33Dj5HCU4/8jM/34HF4j7f2Q8qKX1SVtHjrVYrwcHB\nzJv3Olu3bmHFimV8+eXnTJyYVOK+8qjyNe+zlTZQbft2q1/OAwcID4f33suhSxcHn39uJyXFv8ov\nIpVTSEh1mja9kjfffK2wyRwgKyuTevXqc/LkSTZu/L7Ux4DWrRvBnj2/YRgGmzZ9D7gfIxoV1QCr\n1cq6dWsLj7VYLDidziLHt2wZzcaN3wHw3/9+T4sWLS+o/Gcf36pVq8LnfLdpczV///sEfvvtfyXu\nKy99i5+ltAFsYMHppLAZHfxrHnhgIDz99Ck6dqzO448HkZDgoLpnu7hERC5Yt24JTJmSRFLSE4X7\nbr21HyNG3EOjRpdy551DWLDgXwwf/kCxY4cPf4BJkx6mfv2owoeLdOrUhfHjx7J9+w/07n0TkZGR\nvPbaK7Rp8ydmzXqmSN/5vff+lSeffIJlyz7Ebg9gwoTJOBxl/14/+/gZM54mM9PBvHkvsHRpClar\nlTvuGExUVINi+8qryi3Scj5nP3msNNHRTlJT/a8D+amnAnn22SAeeiiXCRNKXnHuXLSYg2cojp6h\nOHqG4ugZWqSlAp09gA1Kvrfx12b0v/0tjwYNXLz4YiC//aaObxERf6TkXYIzV2Fr2bK0+d4WvxyN\nXr06JCXlkptrISlJDzIREfFHSt7nUdrDTM7mT8up3nKLg3btHKxYEcC6dVqBTUTE3yh5n0dlbEa3\nWGDq1FwsFoP/+78gcktfR0FERHyQkncZVMZm9KuucnH33fns3Gljxgz/aTUQEREl7wtWmZrRJ03K\n5dJLXcydG8imTfooiIj4C31jX6DK1IweGgqzZ5/C5bLwt78Fc+rin5AnIiJepOR9ESpTM3psrJN7\n7slj504bzzzj+60FIiKi5F1ulaEZfdKkXC67zMULLwTy/ff6SIiI+Dp9U5dTZWhGr14d5sxxN5+P\nGhVMKcsIi4iIj1Dy9oDK0Izevr2TwYPz2LXLxrJlvlU2EREpSsnbw/y5Gf1vf8vDajV48cVA/GPF\nexGRqknJ28PK2oxe2qNHK9LllxvceKODLVtsfPGFVl4TEfFVvpdBKoGyNKPbbPhkH/iIEe6Wgxdf\n9L2WARERcVPyNllpzei5ub7ZB37NNS7atXPw6ad2duzQx0NExBfp29lkZzejBwWV3IzuS33gDzzg\nvuF46SXfKZOIiJym5O0FZzajOxwlv8eX+sC7d3dyxRVOPvjAzsGDeua3iIiv8Z2MUUU0a+b7feBW\nK4wYkU9+voX58wMqtCwiIlKckreX+UsfeL9++dSt6+L11wPJzKzQooiIyFmUvL3MX/rAg4Nh2LB8\nTpywsHixat8iIr7E1OQ9bdo0BgwYQGJiIlu2bCny2oEDBxg4cCC33347jzzyiJnF8Dn+0gc+eHA+\ndrvB668HaNEWEREfYlqG2LBhA7t37yY5OZmpU6cyderUIq8/9dRTDBs2jMWLF2Oz2fj999/NKopP\nK60PvLT93lSvnkGvXg62b7exYYMWbRER8RWmJe/169cTHx8PQNOmTcnIyCDzj85Tl8vF999/T5cu\nXQBISkqiQYMGZhXFp5XWB56RYfGJAWx33+1+Sslrr6npXETEV5iWvNPT0wkPDy/crl27NmlpaQAc\nPXqU6tWr8+STTzJw4EBmzpxpVjF83tl94A0bumvc+/dbfWIAW4cOTpo1c7JsmZ20NE0bExHxBV7L\nCMYZnaaGYXDo0CGGDBlCw4YNGT58OKmpqXTq1KnU48PDQ7Dby9d0GxERVq7jzTJ8uPsHoHVrC/v3\nF3/PCy9UK3yPtz34IIwaBUuXhhId7btx9DeKo2cojp6hOHqGt+JoWvKOjIwkPT29cPvw4cNEREQA\nEB4eToMGDbj00ksBaN++Pbt27Tpn8j52LLtc5YmICCMt7WS5zuEN27eHAsVruNu3G6SlVcycrV69\nYPz4UF56yeCf/7Ry9Kjvx9HX+cvn0dcpjp6hOHqGGXEs7WbAtGbz2NhYVq1aBcC2bduIjIwkNDQU\nALvdTqNGjfjtt98KX2/cuLFZRfErvjiArUYNuO22fPbutbJiRYUVQ0RE/mBa8m7bti0xMTEkJiYy\nZcoUkpKSSElJYfXq1QBMnDiRCRMmkJiYSFhYWOHgtaqutAFsHTo4iYsLqbBBbEOHugeuvfiiVy8r\nIiIlsBiGf8zgLW9ThD81Cy1ZYmf27EB27rTSrJmLDh2czJ9ffNGWefNy6Nu3lIniJujVK4Tvv7fx\nzTeZXH65X3xsfJY/fR59meLoGYqjZ1SKZnO5eGcu4pKams2XX5Y8UM/bq7ANHZqHYcDbb2vamIhI\nRVLy9gOlrbbm7VXY+vRxEB4O77wTQH6+Vy8tIiJnUPL2A74yiK1aNRgyBNLSrKxaVbEPThERqcqU\nvP2AL63Cdt997j8XLlTTuYhIRVHy9gO+tApbTAxcd52D1FQbu3drxTURkYqg5O0nzhzEVqNGxT5G\ndPDgfAzDwjvvqPYtIlIRlLz9UEUPYLvpJgc1axoauCYiUkGUvP1QRQ9gq1YN+vXL59AhK6tXa+Ca\niIi3KXn7odIGsI0eXfJ+Mwwa5K5ya+CaiIj3KXn7obMHsEVHO7n33jxmzQr02ujz6GgX117rZO1a\nG3v2aOCaiIg3KXn7qTMHsI0encf8+YHs2GHz6ujzIUPyMAwLixap9i0i4k1K3pXArFkljzI3e/R5\nnz4Oqlc3WLQoAFfFPfRMRKTKUfKuBCpq9Hn16nDLLfns22fl889LXn9dREQ8T8m7EqjI0ecDB7oH\nrr37rprORUS8Rcm7EqjI0ed//rOLK65w8vHHdo4fN/1yIiKCknelUJGjzy0WSEx0kJtrYckS1b5F\nRLxBybuSqMjR5wMG5GOzGWo6FxHxEiXvSsjbo8/r1TPo2tXJf/9rY/t2faRERMymb9pKqCJGn2vg\nmoiI9yh5V0IVMfq8WzcHdeq4WLzYTp73VmkVEamSlLwroYoYfR4YCLff7uDIESuffKKHlYiImEnJ\nuxIqafT5vHk59O3rMPW6/fu7m86XLVPyFhExk5J3JXXm6PPU1GwA4uJCTJ061qqVi0svdbF6tZ3c\nXI+fXkRE/qDkXQUsWWLn/vurmT51zGKBXr0cZGZatFyqiIiJlLyrAG9OHevd2900v3y5ms5FRMyi\n5F0FeHPq2J//7CQiwsXKlXacTo+fXkREUPKuErw5dcxqhZ49HaSnW9mwQU3nIiJmUPKuArw9daxX\nL3fT+ccfq+lcRMQMSt5VgLenjnXs6KRGDYPly+0YhimXEBGp0pS8qwhvTh0LDHSvuLZvn5UtW/QR\nExHxNFPbNadNm8bmzZuxWCxMnDiR1q1bF77WpUsX6tevj83m7hedMWMG9erVM7M48oeCqWMFCqaO\ngedq4717O/jggwCWL7fTpo3WSxUR8STTkveGDRvYvXs3ycnJ/PLLL0ycOJHk5OQi73nllVeoXr26\nWUWQUpxr6pinknfnzg6Cgw0+/tjOhAlK3iIinmRam+b69euJj48HoGnTpmRkZJCZmWnW5eQCeGPq\nWPXq7gS+c6eNXbvUdC4i4kmm1bzT09OJiYkp3K5duzZpaWmEhoYW7ktKSmL//v1cc801jBs3DovF\nUur5wsNDsNvLN/UoIiKsXMdXFtHRsHVrSfstZYpRWeOYmAgrVkBqanU6dLjQUlZ++jx6huLoGYqj\nZ3grjl6by2OcNex41KhRXH9a3l3qAAAgAElEQVT99dSsWZORI0eyatUqEhISSj3+2LHscl0/IiKM\ntLST5TpHZfHgg0X7vAuMHJlDWtq5m80vJI7t20NAQCjvvuvivvvK9+9X2ejz6BmKo2cojp5hRhxL\nuxkwrT0zMjKS9PT0wu3Dhw8TERFRuH3LLbdQp04d7HY7N9xwAzt37jSrKHIWb00dq1ULOnVy8sMP\nNn75pfRWFRERuTCmJe/Y2FhWrVoFwLZt24iMjCxsMj958iT33HMPeXnugUzffvstV155pVlFkRKc\nPXXMrDnfN93kfkzoRx8FmHJ+EZGqyLRm87Zt2xITE0NiYiIWi4WkpCRSUlIICwujW7du3HDDDQwY\nMICgoCCio6PP2WQu/qtnTweBgQYffmjnoYc06lxExBMsxtmd0T6qvP0I6tM5tyVL7MyaFcjOnVaa\nNXMxZkxeibXxi4njkCHBrFwZwBdfZJmynro/0ufRMxRHz1AcPaNS9HmL/zD7ed833eS+CfjoI611\nLiLiCUreYvrzvnv0cBAUZLB0qZK3iIgnKHmL6Yu2hIVB164OfvrJxo8/6iMnIlJe+iYVrzzv++ab\n3U3nqn2LiJSfkrd45Xnf3bo5qFbN3XTuH0MkRUR8l5K3eGXRltBQiI938PPPNrZv18dORKQ81IYp\ngDuBm7VQS4Gbb3awbFkAS5faiYnRnG8RkYulKpB4TXy8g7Awg0WLAsjPr+jSiIj4LyVvKdGSJXbi\n4kKIigolLi7EI3O+Q0Kgf/98Dh608sknavQREblYSt5SjJmLttx1l7vK/frrWutcRORiKXlLMWYu\n2tKihYt27RysW2fn11/1pDERkYuh5C3FmL1oy9Ch7tr3m296ZgU3EZGqRslbijF70ZbevR3Ureti\n0SI7p0555JQiIlWKkrcUY/aiLUFBMHBgPkePWlm2TAPXREQulJK3FOONRVsGD87HYjF4/XU1nYuI\nXChVe6REZi/acvnlBp07O1m71s62bVZiYvScbxGRslLNWyrM0KHuZnhNGxMRuTBK3lImBYu22O14\nbNGW+HgnjRq5SE4OIC1N08ZERMpKyVvOq+iiLXhs0Ra7HR54II9Tpyy88opq3yIiZaXkLedl5qIt\nAwfmU7euiwULAjl5stynExGpEpS85bzMXLQlJASGD8/nxAkLr72mkeciImWh5C3nZfaiLXffnUdo\nqMG8eQHk5HjklCIilZqSt5yX2Yu21KzpTuBpaVYWLVLft4jI+Sh5y3kVXbQFUxZtGT48n6Aggxde\nCMRh3vRyEZFKQclbyqRvXwepqdnk50NqarbHF3CpV89g4MB89uyx8uGHWjtIRORclLzFZ4wcmYfN\nZjB3biCGUdGlERHxXUre4jMuu8ygTx8HO3bY+OorW0UXR0TEZyl5y0UpWHEtKirUYyuuAdx7r3sQ\nnBZtEREpnToX5YIVrLhWoGDFNSj/ILY//9lFmzZOVq60s2ePhUsvVfu5iMjZTK15T5s2jQEDBpCY\nmMiWLVtKfM/MmTMZPHiwmcUQDzNzxTWLxV37drm0aIuISGlMS94bNmxg9+7dJCcnM3XqVKZOnVrs\nPT///DPffvutWUUQk5i54hrALbc4qFvXxVtvBZCV5ZFTiohUKqYl7/Xr1xMfHw9A06ZNycjIIDMz\ns8h7nnrqKR566CGziiAmMXvFtaAgGDIkn4wMCx98oL5vEZGzmdbnnZ6eTkxMTOF27dq1SUtLIzQ0\nFICUlBSuu+46GjZsWKbzhYeHYLeXbwRyRERYuY4Xt0cesTFwYPH9kyfbPBbjceNgzhx47bVgxo4N\nxlIJnxiqz6NnKI6eoTh6hrfi6LUBa8YZE3ePHz9OSkoKr732GocOHSrT8ceOZZfr+hERYaSl6bFV\n5RUREUbXrieZN8/O7NmB7NxppVkzF6NH59G1q4O0NM9cJyAAbropmJSUAFJSsrnhBqdnTuwj9Hn0\nDMXRMxRHzzAjjqXdDJiWvCMjI0lPTy/cPnz4MBEREQB8/fXXHD16lDvvvJO8vDz27NnDtGnTmDhx\nolnFEQ/r29fh8VXWznbvvXmkpATwyiuB3HCDnlgiIlLAtD7v2NhYVq1aBcC2bduIjIwsbDJPSEhg\n+fLlvPfeezz//PPExMQocUsx11zjnja2erWNtLRK2G4uInKRTKt5t23blpiYGBITE7FYLCQlJZGS\nkkJYWBjdunUz67JSiVgscNtt+WzeHMy//23n7rvzK7pIIiI+wWIY/rGKdHn7EdSn4xnniuOSJXZm\nzTrdDz5mTF65m9Z//93C1VeH0qGDgw8/rDxN5/o8eobi6BmKo2dUij5vqVrMWnWtQQOD665zsH69\njYMHLdSv7xf3miIipipTn/cPP/zAZ599BsBzzz3HXXfdxXfffWdqwcS/mLnq2i23ODAMC//+t+41\nRUSgjMl7ypQpNG7cmO+++46tW7cyefJk5syZY3bZxI+Yuepanz4OLBZDz/kWEflDmb5Zg4KCuPzy\ny/n000/p378/V1xxBVarHkgmp5m56lq9egYdOjjZsMHO/v0adS4iUqYMnJOTw4oVK1izZg0dO3bk\n+PHjnDhxwuyyiR8ZMyavxP2jR5e8/0LdfLO73/yjj1T7FhEpU/IeO3Ysy5Yt46GHHiI0NJSFCxcy\ndOhQk4sm/qRvXwfz5uUQHe3EbjeIjnYyb175HxFa4MYbHVitBh99pLXORUTKVI1p164drVq1IjQ0\nlPT0dNq3b0/btm3NLpv4GTNXXatb16BjRyf/+Y+e8y0iUqaa9xNPPMGKFSs4fvw4iYmJvPXWWzz6\n6KMmF02kqFtucd8YLF2q2reIVG1lSt7bt2+nX79+rFixgr59+zJr1ix2795tdtlEiujdOx+73VC/\nt4hUeWVK3gWLsKWmptKlSxcA8vI8MxBJKq8lS+zExYUQFRVKXFwIS5aUL+mGh0Pnzk42b7bx/fea\n7SAiVVeZvgEbN25Mr169yMrKomXLlnz44YfUrFnT7LKJHytYcW3HDhtOp6VwxbXyJvCRI903jTNm\nBHmimCIifqlM36RTpkxh586dNG3aFIArrriC6dOnm1ow8W/nWnGtPIPaOnRwEhvr4NNP7WzcaKVt\n2/LPIxcR8TdlqnmfOnWKtWvXMmrUKEaMGMGXX35JYGD5l72UysvMFdf+/nfVvkWkaivTN+nkyZPJ\nzMwkMTGR/v37k56ezqRJk8wum/gxM1dci4110qGDgzVr7GzapL5vEal6yvTNl56ezsMPP0ynTp3o\n3Lkz//d//8ehQ4fMLpv4MbNXXCuofc+cqdq3iFQ9ZV4eNSfn9LOUs7Ozyc3NNa1Q4v/MXnEtNtZJ\nu3YOPvnEzn//q9q3iFQtZRqwNmDAAHr27EmrVq0A2LZtG6NHjza1YOL/zFxxzWKBf/wjj9tuszNz\nZhALF+ac/yARkUqiTMn79ttvJzY2lm3btmGxWJg8eTILFy40u2wi59Sxo5O//MXBqlV2tm2zEhOj\nkeciUjWUedJtVFQUUVFRhdtbtmwxpUAiZWWxwIMP5vHNN3YWLAhg5kx15YhI1XDRnYUFq66JlJWn\nV1wDiI93cumlLhYvDuDYMQ8UUkTED1x08rZYLJ4sh1RyZq24ZrPBsGF55ORYePttPbBERKqGc35z\nxsXFlZikDcPgmKo5cgHMWnEN4I478pk+PYjXXw9kxIh8bLZynU5ExOedM3m/88473iqHVHJmrrhW\nqxbcdls+CxcG8skndnr2NGeEu4iIrzhn8m7YsKG3yiGVXLNmLnbsKF4l9sSKawD33utO3vPnByh5\ni0ilp9UtxCvMXnGtZUsXHTs6+PxzOz/9pI+1iFRu+pYTrzB7xTWAe+7JB2D+fA1cE5HKrfxzdUTK\nyMwV1wB69HBwySUu3n8/gEmTctEj50WkslLNWyoNux2GDs0nO9vCokWqfYtI5aXkLRXGjEVb7rwz\nn6Agg9deC8Sl1VJFpJIytdl82rRpbN68GYvFwsSJE2ndunXha++99x6LFy/GarXSokULkpKStPBL\nFVKwaEuBgkVboHz94HXqGNxyi4Pk5ADWrbPRubPTA6UVEfEtptW8N2zYwO7du0lOTmbq1KlMnTq1\n8LWcnBw+/vhj3n77bRYtWsSvv/7Kpk2bzCqK+KBzLdpSXsOGuUewL1hQ/nOJiPgi05L3+vXriY+P\nB6Bp06ZkZGSQmZkJQLVq1XjjjTcICAggJyeHzMxMIiIizCqK+CAzF235059ctG3r5JNPbOzZo9Yc\nEal8TEve6enphIeHF27Xrl2btLS0Iu/517/+Rbdu3UhISKBRo0ZmFUV8UGmLs3hq0Za7787DMCy8\n8YYGrolI5WMxTHo82OTJk4mLiyusfQ8cOJBp06bRuHHjIu87deoU9913H2PGjOGaa64p9XwOhxO7\nXYtWVxaLFsHAgcX3v/suJCaW//ynTkGjRmAYsG8fBAeX/5wiIr7CtAFrkZGRpKenF24fPny4sGn8\n+PHj7Nq1iz//+c8EBwdzww03sHHjxnMm72PHsstVnoiIMNLSTpbrHOK5OHbtCvPm2Zk9O5CdO600\na+Zi9Og8unZ1cFYDzUUbODCQuXODeOWVHBITfWvJVH0ePUNx9AzF0TPMiGNERFiJ+01rNo+NjWXV\nqlUAbNu2jcjISEJDQwFwOByMHz+erKwsALZu3VqsRi6VX9++DlJTs/n990xSU7M9voDLXXflY7G4\np42JiFQmptW827ZtS0xMDImJiVgsFpKSkkhJSSEsLIxu3boxcuRIhgwZgt1up3nz5nTt2tWsokgV\ndemlBj16OFi5MoBNm6z86U+a+C0ilYNpfd6eVt6mCDULeYaZcVyyxM6sWaeb0ceMySt3bfzTT20M\nHBjCXXfl8cwzuR4qafnp8+gZiqNnKI6eUSmazUUuRMGiLTt22HA6LYWLtpR31bVOnZzUr+9iyZIA\nTp3yUGFFRCqYkrf4BLMWbbHZoF+/fE6csLBypZ7DIyKVg5K3+AQzF20ZMMDd9K6HlYhIZaHkLT7B\nzEVbmjVzcc01TlJTbRw4oBXXRMT/KXmLTxgzJq/E/aNHl7z/Qg0YkI/LZeH991X7FhH/p+QtPqFv\nXwfz5uUQHe3EbjeIjnYyb175njB2pltucT8q9L337PjH/AoRkdJpBI/4jL59HR5fqKVArVrQs6eD\nDz90z/lu21ZzvkXEf6nmLVVGYmI+oIFrIuL/lLzFZy1ZYicuLoSoqFDi4kLKPec7Lk5zvkWkclDy\nFp9kxqItBXO+MzIsLF+uHiMR8V9K3uKTzFq05c4787FaDWbPDsSlbm8R8VNK3uKTzFq0pUkTg9tv\nd7Bjh42PP1btW0T8k5K3+CQzF20ZOzYXq9VgxgzVvkXEPyl5i08yc9EW1b5FxN8peYtPMnvRFtW+\nRcSfqdohPsvMRVsKat/vvRfAxx/b6dPHnOuIiJhBNW+pslT7FhF/peQtfsPTi7ac2ff973+rEUpE\n/IeSt/gFMxZtAXftOyDAYPLkIE6e9FBhRURMpuQtfsGsRVuaNDEYPTqPAwesPP54ULnOJSLiLUre\n4hfMWrQF3NPPWrRw8sYbgaxfbyv3+UREzKbkLX7BzEVbgoLg2WdPYbEYPPRQMDk55T6liIiplLzF\nL5i5aAvAtde6GD48n19/tTJzZvma4kVEzKbkLX7B7EVbAMaPz+XSS1288EIgW7fqv4aI+C7NjxG/\nYeaiLQDVq8OMGafo3z+Ev/0tmFWrsgnSGDYR8UGqXojf8vS8b4BOnZwMHpzH9u02pk9X87mI+CYl\nb/FLZs37BnjssVwuu8zF888H8vXXGn0uIr5HyVv8klnzvgFCQ+H5508B8OCDwWRmlvuUIiIepeQt\nfsnMed8Af/mLkwcfzGPPHitJSer4FhHfYmrynjZtGgMGDCAxMZEtW7YUee3rr7+mf//+JCYmMmHC\nBFx6MoRcADPnfRf45z/ziI52snBhIJ98ouZzEfEdpiXvDRs2sHv3bpKTk5k6dSpTp04t8vojjzzC\nnDlzWLRoEVlZWXz++edmFUUqIbPnfYN78ZYXXjhFYKDBmDHBpKVZPHZuEZHyMC15r1+/nvj4eACa\nNm1KRkYGmWd0HqakpFC/fn0AateuzbFjx8wqilRC3pj3DRAT42LChFzS06089FAwhuHR04uIXBTT\nknd6ejrh4eGF27Vr1yYtLa1wOzQ0FIDDhw/z5ZdfEhcXZ1ZRpJLq29dBamo2v/+eSWpqtmlzwEeM\nyOf66x188omdN94IMOUaIiIXwmuLtBglVFmOHDnCX//6V5KSkook+pKEh4dgt5ev3zEiIqxcx4ub\nr8Zx0SKYNg22b4foaJg4ERITPXPud96B1q0hKSmYPn2CadGi/Of01Tj6G8XRMxRHz/BWHE1L3pGR\nkaSnpxduHz58mIiIiMLtzMxM7rvvPsaMGUPHjh3Pe75jx7LLVZ6IiDDS0vTA5vLy1TgWzPsusHUr\nDBwIJ054pik9KAieecbOvfdWY8AAJ8uXZxNYjllpvhpHf6M4eobi6BlmxLG0mwHTms1jY2NZtWoV\nANu2bSMyMrKwqRzgqaee4q677uKGG24wqwhShZg577vATTc5SEzMZ8sWG08/rdXXRKTimFbzbtu2\nLTExMSQmJmKxWEhKSiIlJYWwsDA6duzIhx9+yO7du1m8eDEAN954IwMGDDCrOFLJmT3vu8C0aaf4\n+msbc+cG8Ze/OOne3enR84uIlIWpfd5///vfi2y3OKOj8IcffjDz0lLFNGvmYseO4mMiPDnvG9yr\nr736ag69e4cwcmQ1Vq/O4vLLNQRdRLxLK6xJpeCNed8FrrrKxfTpp8jIsDBsWDVycjx+CRGRc1Ly\nlkrBW/O+CyQmOhg8OI8ffrDx8MOa/y0i3qXkLZXG2fO+AY8/MvRMU6fmcvXVThYtCmDhQs3/FhHv\nUfKWSsnMR4YWCA5293+HhxtMmBCkx4eKiNcoeUul5I2pYwCNGhm8+moOhgFDhwbzv/9p/XMRMZ+S\nt1RK3po6BtCxo5Pp03M5etTKoEHVyMjw+CVERIpQ8pZKyRuPDD3ToEH5jBiRx65dNu69txoOc8bJ\niYgASt5SSXlz6liBRx7JpXt3B+vW2ZkwIUgj0EXENEreUil5e+oYgM0GL7/svuYbbwTyxBOBSuAi\nYgolb6m0Snpk6JIldlOnj4WGQnJyDldc4eT554OYOlUJXEQ8T8lbqgxvTB8DqFfPICUlhyZNXMyZ\nE8STTyqBi4hnKXlLleGt6WMA9esbLFmSTePGLmbNCuLpp5XARcRzlLylyvDm9DGAqCh3Ar/8chfP\nPhvExIlBOPUQMhHxACVvqTK8PX0MoEEDg6VLs2nZ0smrrwZy773BepCJiJSbkrdUGRUxfQzcNfCP\nPsomNtbBxx8H0L9/NY4eNfWSIlLJKXlLlVHS9LF7781j1qxA00afF6hZExYtyuGWW/L55hs7HTua\n11wvIpWfvj2kSjlz+tjo0XnMnx9o+ujzAkFB8PLLp/jrX/PYsQPi40N49dUADWQTkQum5C1VljdH\nnxewWuHxx3NJSYFq1WDChGDuvLMahw/rgSYiUnZK3lJleXv0+Zn69oV167Lo1MnBmjV2OnUK4auv\n9EhRESkbJW+psipi9PmZ6tc3WLQohylTTnH8uIV+/aqxeLE5TfYiUrkoeUuVVdro84wMi+kD2ApY\nrTB8eD7JyTlUqwYPPFCN557Tgi4icm5K3lJlnT36vGFDd417/36rVwawnen66538+9/ZXHKJiyef\nDGLs2CDy802/rIj4KSVvqdLOHH1eo0bJ1V0zB7CdqUULFytWZNO6tZO33w4kMbEax4555dIi4meU\nvEX+UJED2ArUq2fw4YfZ9OyZz+ef2+nRo7rmg4tIMfpWEPlDRQ9gKxAaCq+9dooxY3L57TcrPXuG\nsHatRqKLyGlK3iJ/8IUBbAWsVpg4MY+XXsohLw/uuKMaTzwRyMmTXrm8iPg4JW+RP/jSALYCt93m\nYOnSbBo0MJg7N4h27arz1lsBejqZSBWn5C1yBl8awFagbVsXX3yRxfjxuWRlWRg7Npj4+BA+/dSm\nKWUiVZSSt0gpfGEAW4GQEBg7No+vv85iwIB8tm2zMXBgCPHxISxbZldNXKSKUfIWKUVpA9VsNrze\nB16gfn2DuXNPsXZtFrfcks+2bVbuuacaHTtWZ+HCALKzvVocEakgpibvadOmMWDAABITE9myZUuR\n13Jzc3n44Ye59dZbzSyCyEUrbQBbbq6lwvrAC7Rq5eJf/zrFV19lMWhQHnv2WBg3Lpirrw7lsceC\n2LNHDzoRqcxMS94bNmxg9+7dJCcnM3XqVKZOnVrk9enTp9OyZUuzLi9SbmcPYAsK8o0+8DM1aWLw\n7LO5fPddFmPH5mK3G7zwQiDXXVedIUOCWbvWhsu7M91ExAtMS97r168nPj4egKZNm5KRkUFmZmbh\n6w899FDh6yK+6swBbA5Hye/xhUVUoqIMxo/PY9OmLJ5/Poc2bVysXBlAYmII111XnTlzAvXYUZFK\nxLT2vvT0dGJiYgq3a9euTVpaGqGhoQCEhoZy/PjxMp8vPDwEu718C1VERISV63hxq6pxjI6GrVuL\n77fbLURFhREdDRMnQmJi2c5nVhxHjnT/fPcdvPQSvPuulSlTgpgyJYg//Qm6d4du3SA2FoKDTSmC\nV1XVz6OnKY6e4a04eq2zzijnnJZjx8o3EiciIoy0NK1wUV5VOY4PPmjn/vurFdt/6pT7z61bYeBA\nOHEih759S6mm/8EbcbzsMnjqKRg/Ht5/P4AVK+xs2GBj0yYLTz8N1aoZxMY66dLFQefODpo0MbD4\nWeW8Kn8ePUlx9Awz4ljazYBpyTsyMpL09PTC7cOHDxMREWHW5URM507IOcyeHcjOnVZsNvfgtbPN\nnh143uTtTbVqwX335XPffflkZcHXX9tITbWzbp2NNWvsrFnj/hq49FIX8fEOunVz0KGDk2rF71NE\nxEeYlrxjY2OZO3cuiYmJbNu2jcjIyMImcxF/1bevozAxR0WV/Hn2hT7w0lSvDl27Ouna1T0xfP9+\nC599ZmftWhvr1tlZsCCQBQsCC2vl7do5ue46J23aKJmL+BKLUd727HOYMWMG3333HRaLhaSkJLZv\n305YWBjdunVj1KhRHDx4kF27dtGqVSv69+9Pnz59Sj1XeZsi1CzkGYrjaXFxIezYUXwcRlCQgcPh\nnic+ZkxeibVwX4xjfj5s2FBQG7fx00+n/24BAQatW7to3dpJq1YurrrKSYsWrgrvM/fFOPojxdEz\nvNlsbmry9iQlb9+gOJ62ZEnJfeBnmzeveB+4P8Tx4EEL335rY8MG98/WrVYcjtPdBDabQfPmriJJ\nvWVLJzVreq+M/hBHf6A4eoaSdwmUvH2D4ljUkiX28/aBR0c7SU0tOuDSH+N46hT89JOVrVvdiXzr\nVhvbt1vJzi76d46IcNGsmYsrr3QRHe2iTRsn0dEugoI8XyZ/jKMvUhw9Q8m7BErevkFxLF1UVChO\nZ0nDtQ1stqLN6JUljk4n/PyzlS1b3Ml81y4ru3ZZ2bvXgmGcjoXdbtCihYsWLVw0buyiSRP3z2WX\nuQgP56JHuVeWOFY0xdEzKsVoc5GqplkzV4l94GDB6aRwOVXIYfhwb5fOHDYbNG/uonlzF/36ne4a\nyM6GX36xsnWrlc2bbWzZYmPbNis//FA8PiEhBo0aubjkEvdjWBs0MGjQwEVUlEFUlPt3jXUVKUo1\nb7kgimPpytoH7h7QZqFZM2epA9oqI4cD9u2z8OuvVv73P/fPnj0W9u61sm+flYyM0qvfoaEGUVGu\nP5L56d+bNw/Gbs+idm2D2rUNatYEq+8O9vdZ+n/tGap5i/ihs+eBu5dTLZ6QCvrFz6yJV4UEbrfD\n5ZcbXH65Eyj+DNMTJ+D33638/ruFAwcK/nT/fuCAhYMHLezaVVLLRvXC32w2g8hId429fn13gq9X\nz/17ZKT797p1DerUMbDr20/8mGreckEUx7IrbSrZ2Uoa0CYly8lxj4IvSO6ZmdXYuzeXo0ctHDli\n4cgRKwcPuhN9fv65O9Jr1XIn8dq1DerUcf3xp0GNGu6afvXqBqGh7t/Dwk7/7v5xdxlUFvp/7Rmq\neYtUAmPG5JWpGX37ditRUaHnnBcubtWqQePGBo0bu2vuERGQllb80a0uFxw96q65Hz5s4eBBK4cO\nWTh0yEJ6ekGid//89psVp/PCM3G1au4EHxICgYEGgYEQGOjuFqleveAGwN2nHxDgnisfEOB+z9nb\ndru7JaDgp+A1u/30OYOCIDjY/WfBa3a7ezBkwXE228UP/hP/ouQtYpKyLqda0oA2JfDysVqhbl13\nE7lb8Wb6Ai6Xu8m+oPZ+8qSFzEwLmZn88af795Mn3a9lZxfdn51tITvbQl6ehfx895S6M0fae1tB\nQrfZ3HEo+NNica9dX/BT8B6bzX0zYBjVsVqNP957euzAmdsFP2eeB4qf+8zzF9xknH7v6T9LO9+Z\nSmobPvN95W07PrM8Bedyudw/hlH095KOc7/Hgsvlfp7A44+742k2NZvLBVEcL15ZB7Q1bOiiRg2D\nnTutqo2fhy9+Hg3D3byfmWkhK8ud3B0OyMuD/HwLeXn8se1O9nl57il3Dod72+Hgjz/d2/n57pu+\nU6cgNxdycix/vL/gp+TtM38Mo+iPy+V+j8tV8LoVp9NVJGEZhuWs7dO/F5yz4O97+qdqV/vr1IH1\n609Sq5bnzqlmc5EKVrQmbsPhMChpQNv+/Vb273f/rtq4/7FYICTE3Vzu5vv1I/dNUJbHzleQ5E/f\nUJxO8AVcLkuxG4OCYwv+LKlGfvbrUHqN/XxdCGff1BS0BhRtQTCK7Du7rKdbNuCSS8I4duzC43Ux\nVPOWC6I4ekZERBjR0c4yDWiDsq2XXhXp8+gZiqNneHPAmmZEilSQMWOKD7QqTW6uBafTUlgT/9Of\nqhMVFUpcXAhLlqgBTbiFIVgAAAysSURBVKSqUfIWqSB9+zqYNy+H6GgndrtBdLSThg1dZTp2/35r\nkWSuBC5StSh5i1Sgvn0dpKZm8/vvmaSmZvPII7kXdZ5Ro4JVExepQpS8RXzI2bXxoKCyDUlRs7pI\n1aLkLeJjzqyNz5lz6qLOcXazupK5SOWi5C3iw86uiZe1T/xs6iMXqVyUvEV83Jk18U2bsi6qWf1s\nZ/aRT5wYRFxciGrmIn5E87zlgiiOnuGpOJZ11bYL1bChi4MHLT4/r1yfR89QHD1D87xFpEw81ax+\nNvWZi/g2/S8U8XN9+zqK1IyXLLEXPgylXj2D/fvLf49ecI6CZP7tt3l8+aWtcP312FhnkW1frq2L\nVAZqNpcLojh6hjfjeGYyL/3JZp53ZtO7Wcldn0fPUBw9w5vN5kreckEUR8+oqDia1Ud+MTyR3PV5\n9AzF0TOUvEug5O0bFEfPqMg4nlkTb9bMRYcOTr76yubRZnZPOFdyP71to1kzp5rpy0n/rz1DybsE\nSt6+QXH0DF+Ooxl95t5w9gh5gFmzAtUvXwa+/Hn0J0reJVDy9g2Ko2f4UxxLqqnPnx9Y0cXyiLLV\n7su27c83A/70efRlSt4lUPL2DYqjZ/h7HP2l6d3bLuRmoCytA966QfD3z6OvUPIugZK3b1AcPaOy\nx1HJ3TyebC04e+yAJ24wKlOLxIVS8i6BkrdvUBw9o6rHUcm9ajHnhsNzNxievgny5A1LhSTvadOm\nsXnzZiwWCxMnTqR169aFr3311Vc8++yz2Gw2brjhBkaOHHnOcyl5+wbF0TMUx3M7V3Ivum2jXj2X\nkr34nHnzcjy2lkFJTFthbcOGDezevZvk5GR++eUXJk6cSHJycuHrU6ZM4dVXX6VevXoMGjSIHj16\ncMUVV5hVHBHxI2evGlca901QVrFkP3q0uzam2r1UlNmzA03tLjAtea9fv574+HgAmjZtSkZGBpmZ\nmYSGhrJ3715q1qxJVFQUAHFxcaxfv17JW0QuSmnJ/lxfnmWv3Z9/WzcDcradO839PJiWvNPT04mJ\niSncrl27NmlpaYSGhpKWlkbt2rWLvLZ3716ziiIiUkxZa/dldaE3A+drHdANgn9r1swzDwkqjdce\nTFLervXw8BDsdlu5zlFa34FcGMXRMxRHz/CVOA4f7v5xs/3xc67taoXHle0YC4sWwZNPwvbtEB0N\ncXGwbp152xMmuK98sdds0AD27PFMfP3N5Mk2Uz+bpg1Ymzt3LhERESQmJgLQtWtXli5dSmhoKPv2\n7WPcuHGFfeDPP/88tWrVYtCgQaWeTwPWfIPi6BmKo2cojp5hZhw92T3hiRYMc7fdo81HjzZ/tLlp\nNe/Y2Fjmzp1LYmIi27ZtIzIyktDQUAAuueQSMjMz2bdvH/Xr1+ezzz5jxowZZhVFREQqiKe7J851\nnYrmvgnK9sq1TEvebdu2JSYmhsTERCwWC0lJSaSkpBAWFka3bt149NFHGTduHAC9evWicePGZhVF\nRESkUtEiLXJBFEfPUBw9Q3H0DMXRM7y5wpqGLoqIiPgZJW8RERE/o+QtIiLiZ5S8RURE/IySt4iI\niJ9R8hYREfEzSt4iIiJ+xm/meYuIiIibat4iIiJ+RslbRETEzyh5i4iI+BklbxERET+j5C0iIuJn\nlLxFRET8jGnP8/Yl06ZNY/PmzVgsFiZOnEjr1q0rukh+Y/r06Xz//fc4HA7uv/9+rrrqKv75z3/i\ndDqJiIjgmWeeITAwsKKL6RdOnTrFjTfeyAMPPED79u0Vx4vw0UcfMX/+fOx2O6NGjaJ58+aK4wXK\nysri4YcfJiMjg/z8fEaOHElERASPPvooAM2bN+exxx6r2EL6uJ07d/LAAw8wdOhQBg0axIEDB0r8\nHH700Ue88cYbWK1W+vfvT79+/TxXCKOS++abb4zhw4cbhmEYP//8s9G/f/8KLpH/WL9+vXHvvfca\nhmEYR48eNeLi4ozx48cby5cvNwzDMGbOnGm8/fbbFVlEv/Lss88at956q/HBBx8ojhfh6NGjRvfu\n3Y2TJ08ahw4dMiZNmqQ4XoSFCxcaM2bMMAzDMA4ePGj06NHDGDRokLF582bDMAxj7NixRmpqakUW\n0adlZWUZgwYNMiZNmmQsXLjQMAyjxM9hVlaW0b17d+PEiRNGTk6O0bt3b+PYsWMeK0elbzZfv349\n8fHxADRt2pSMjAwyMzMruFT+4c9//jOzZ88GoEaNGuTk5PDNN9/QtWtXADp37sz69esrsoh+45df\nfuHnn3+mU6dOAIrjRVi/fj3t/7+9ewuJaovjOP4dZ2cyOuElp9Au2E0JxEshnJqSgiIIehEfCgui\nHtIIocg0UosgZ0JCtIck7UVSE6N86UaBoKSCGVZWj1EqXbRswlKY5pyHYDjSFGoe5szM7/O215K9\n1/6x5T9rbdjrr7+IiorCZrNx9uxZ5TgLMTExjI2NAeByuYiOjmZoaMi7Iqkcfy88PJzLly9js9m8\nbb6ew/7+flJTU7FarURERJCZmUlfX9+cjSPoi/fIyAgxMTHe49jYWD58+ODHEQUOs9mMxWIBoLW1\nlc2bN/Pt2zfvsmRcXJyynCan00lxcbH3WDnO3ODgIBMTExw6dIg9e/bQ1dWlHGdh586dDA8Ps23b\nNvLy8igqKmLBggXefuX4e4ZhEBERMaXN13M4MjJCbGys92/muvaExDvvf/tbX4Odsfv379Pa2sqV\nK1fYvn27t11ZTs/NmzdJT09n6dKlPvuV4/SNjY1x8eJFhoeH2bdv35TslOP0tLW1kZCQQH19PS9f\nvuTw4cNYrVZvv3L8M7/Kb65zDfribbPZGBkZ8R6/f/+e+Ph4P44osHR0dHDp0iXq6uqwWq1YLBYm\nJiaIiIjg3bt3U5aOxLf29nbevHlDe3s7b9++JTw8XDnOQlxcHBkZGRiGwbJly4iMjMRsNivHGerr\n68NutwOQkpLC5OQkbrfb268cZ87X/7Ov2pOenj5n1wz6ZfONGzdy9+5dAAYGBrDZbERFRfl5VIHh\ny5cvnD9/ntraWqKjowHYsGGDN8979+6xadMmfw4xIFRVVXH9+nVaWlrIzc2loKBAOc6C3W6nu7sb\nj8fDp0+f+Pr1q3KcheXLl9Pf3w/A0NAQkZGRrFy5kt7eXkA5zoav5zAtLY2nT5/icrkYHx+nr6+P\n9evXz9k1Q2JXscrKSnp7ezGZTJSXl5OSkuLvIQWEa9euUVNTQ1JSkrfN4XBw6tQpJicnSUhIoKKi\ngnnz5vlxlIGlpqaGxMRE7HY7J06cUI4z1NzcTGtrKwD5+fmkpqYqxxkaHx/n5MmTjI6O4na7KSws\nJD4+nrKyMjweD2lpaZSUlPh7mP9bz549w+l0MjQ0hGEYLFq0iMrKSoqLi396Du/cuUN9fT0mk4m8\nvDx27do1Z+MIieItIiISTIJ+2VxERCTYqHiLiIgEGBVvERGRAKPiLSIiEmBUvEVERAJM0H+kRSSU\nDQ4OsmPHDjIyMqa0Z2dnc/DgwT8+f09PD1VVVTQ1Nf3xuURk+lS8RYJcbGwsDQ0N/h6GiMwhFW+R\nELV27VoKCgro6elhfHwch8PBmjVr6O/vx+FwYBgGJpOJsrIyVq1axatXrygtLcXj8TB//nwqKioA\n8Hg8lJeX8+LFC8LDw6mtrQXg2LFjuFwu3G43W7ZsIT8/35+3KxJU9M5bJER9//6d1atX09DQwO7d\nu6murgagqKiIkpISGhoa2L9/P2fOnAGgvLycAwcOcPXqVXJycrh9+zbwY7vTI0eO0NLSgmEYdHZ2\n8vDhQ9xuN42NjTQ3N2OxWPB4PH67V5Fgo5m3SJD7+PEje/fundJ2/PhxAO8GFZmZmdTX1+NyuRgd\nHfXu7ZyVlcXRo0cBePLkCVlZWcCPbSXhxzvvFStWsHDhQgAWL16My+Vi69atVFdXU1hYSHZ2Nrm5\nuYSFaa4gMldUvEWC3O/eef/768gmkwmTyfTLfsDn7NlsNv/UFhcXR1tbG48fP+bBgwfk5ORw48aN\nn/ZBFpHZ0U9hkRDW3d0NwKNHj0hOTsZqtRIfH+/ddaqrq8u7jWFmZiYdHR0A3Lp1iwsXLvzyvJ2d\nnbS3t7Nu3TqKioqwWCyMjo7+x3cjEjo08xYJcr6WzZcsWQLA8+fPaWpq4vPnzzidTgCcTicOhwOz\n2UxYWBinT58GoLS0lNLSUhobGzEMg3PnzvH69Wuf10xKSqK4uJi6ujrMZjN2u53ExMT/7iZFQox2\nFRMJUcnJyQwMDGAY+g0vEmi0bC4iIhJgNPMWEREJMJp5i4iIBBgVbxERkQCj4i0iIhJgVLxFREQC\njIq3iIhIgFHxFhERCTD/AN2+v1zrPUO9AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "6MqKDINMN5px",
        "colab_type": "code",
        "outputId": "193fc64a-14f2-450b-f21b-132b6bab8537",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        }
      },
      "cell_type": "code",
      "source": [
        "plt.clf()   # clear figure\n",
        "\n",
        "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAFnCAYAAACPasF4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "zygKqfqlgmX1",
        "colab_type": "code",
        "outputId": "a37904db-c26d-492b-aeff-59a856e3b419",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "cell_type": "code",
      "source": [
        "# Review data\n",
        "  \n",
        "reverse_word_index = dict([(value, key) for (key, value) in word_index.items()])\n",
        "\n",
        "def decode_review(text):\n",
        "    return ' '.join([reverse_word_index.get(i, '?') for i in text])\n",
        "\n",
        "predictions = model.predict(prediction_data)\n",
        "for index, data_item in enumerate(prediction_data):\n",
        "    print(\"Prediction: \" + str(round(float(predictions[index]),2)).ljust(12) , \"Data: \" + decode_review(data_item))"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Prediction: 0.99         Data: <STRING CHANGE> <VALUE UP> . validationxml document browserxml arrangement entry entry entry @label <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "Prediction: 0.0          Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "Prediction: 0.0          Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "Prediction: 0.0          Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Nc-o_deI52p7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "b3dbd672-3357-4dea-978f-de43df0f51d2"
      },
      "cell_type": "code",
      "source": [
        "predictions = model.predict(prediction_data)\n",
        "for index, data_item in enumerate(prediction_data):\n",
        "    print(prediction_ids[index], \"Prediction: \" + str(round(float(predictions[index]),2)).ljust(5) , \"Data: \" + decode_review(data_item))\n"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "idw_atxt_c1110#0 Prediction: 0.99  Data: <STRING CHANGE> <VALUE UP> . validationxml document browserxml arrangement entry entry entry @label <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "idw_atxt_c1110#1 Prediction: 0.0   Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "idw_atxt_c1110#2 Prediction: 0.0   Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n",
            "idw_atxt_c1110#3 Prediction: 0.0   Data: <NUMBER CHANGE> <VALUE UP> . validationxml document displayxml sheets sheet document root node base base node base dlitem dlitem dlitem member @X <PAD> <PAD> <PAD> <PAD> <PAD>\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}